Impure Hydrogen Valorization for Chemicals Production in a

Impure Hydrogen Valorization for Chemicals

Production in a Tubular Reactor

Dissertation presented for the Doctor of Philosophy degree in

Refining, Petrochemical and Chemical Engineering.

by

Clara Sofia Rodrigues Sá Couto

Supervison: Professor Luís Miguel Madeira

Professor Clemente Pedro Nunes

Doutor Paulo Araújo

Porto, February 2016

Trabalho financiado pela Fundação para a Ciência e Tecnologia e

pela CUF - Químicos Industriais no âmbito do Programa Doutoral

em Engenharia da Refinação, Petroquímica e Química

(SFRH/BDE/51794/2012)

http://www.bing.com/images/search?q=compete+2020+&view=detailv2&&id=1CF7F68343E2DAFFB76851318C5D271AA5F8493E&selectedIndex=6&ccid=+8MHnNr1&simid=607999728708881450&thid=OIP.Mfbc3079cdaf59a1810ee002bbc97ea58o0

http://www.bing.com/images/search?q=Portugal+2020&view=detailv2&&id=C5797E43564FE78F8BA3ABE0066B680AC7CEF06F&selectedIndex=2&ccid=/nyL9xp8&simid=608041364117262277&thid=OIP.Mfe7c8bf71a7c2c1b7f763b1b1d3d2d1ao0

http://www.bing.com/images/search?q=FCT&view=detailv2&&id=4D653375AAA6BB2D5D5BC0DFE38B20598D4B14B5&selectedIndex=0&ccid=HHDrp/zS&simid=608012355905257619&thid=OIP.M1c70eba7fcd29fe3b2b813d2d00b89b6H0

i

Agradecimentos

Ao concluir mais uma etapa do meu percurso académico e pessoal, gostaria de

escrever umas palavras de agradecimento a todos os que me ensinaram, guiaram,

apoiaram e partilharam comigo esta experiência. Tendo em conta tudo o que vivi nesta

fase, as palavras que escreverei serão sempre poucas para demonstrar a minha gratidão e

apreço.

Aos meus orientadores, ao Professor Luís Miguel Madeira, ao Professor Clemente

Pedro Nunes e ao Doutor Paulo Araújo, em primeiro lugar, por terem acreditado em mim

e por terem aceitado orientar-me nesta jornada. Obrigada por partilharem comigo os

vossos conhecimentos, o vosso tempo e por me ensinarem a ser melhor profissional e a

encarar as coisas de prespetivas diferentes. Obrigada pelo apoio, dedicação, paciência,

empenho e por me terem sempre feito acreditar que seria possível levar este projecto até

ao fim com sucesso.

À CUF – Químicos Industriais, nomeadamente ao Eng.º Mário Jorge, por ter dado

a possibilidade de este doutoramento fosse concretizado. Uma ideia diferente, com

desafios novos e uma instalação por construir requer investimento e empenho, em algo

que apenas na teoria se tem a certeza que funciona. Agradeço a confiança da CUF e

principalmente, o empenho do Doutor Paulo Araújo, por me ajudar a levar este

doutoramento a bom porto.

À Fundação para a Ciência e Tecnologia – FCT, pelo apoio financeiro com uma

bolsa de doutoramento em meio empresarial, ao abrigo do Programa Doutoral em

Engenharia da Refinação, Petroquímica e Química (SFRH/BDE/51794/2012). Ao

Laboratório de Engenharia de Processos, Ambiente, Biotecnologia e Energia (POCI-01-

0145-FEDER-006939) – LEPABE – financiado pelo Fundo Europeu de Desenvolvimento

Regional (FEDER), através do COMPETE2020 – Programa Operacional

Competitividade e Internacionalização (POCI) e por fundos nacionais através da FCT I.P.

A todos os colegas da CUF, principalmente: Fernando Mendes, Hugo Pedreiras,

Marco Prior, Rui Andrade e Susana Caldas, por me terem recebido tão bem e por estarem

ii

sempre disponíveis para ajudar e esclarecer dúvidas. Um agradecimento especial ao

Alejandro Ribeiro e à Dulce Silva pelo apoio, ajuda e disponibilidade.

À Professora Filipa Ribeiro, pelo apoio, por me incentivar a aceitar este desafio e

por me abrir sempre as portas do Laboratório para fazer análises e para aprender como

funciona o CataTest.

Aos amigos que fui guardando ao longo dos anos: Filipa Henriques, André Neves,

Miguel Pinto, Rita Sousa, Inês Leal, Telmo Duarte, Diana Fernandes, Margarida Vilhena,

que estiveram tão presentes nesta fase final. À Leonor Alves, Marta Silva e ao Pedro

Brântuas.

À Raquel Bértolo, a pessoa que (quase) se adequa ao meu perfil melhor do que eu

mesma…uma amiga sempre presente, que sempre me apoiou e com a qual tive a sorte de

partilhar as mesmas experiências, apesar de distantes geograficamente.

À Família do Palacete, que quer tendo vivido nele ou não, são parte integrante:

Daniel Marcos, Diogo Afonso, Joana Azevedo, João Dionísio Sousa, José Gomez,

Mariana Cardoso, Rita Tavares, Sérgio Terras, Sofia Vilaça, Tiago Couchinho, pelas

horas de distração, pela animação e pelos bons momentos partilhados. Um agradecimento

especial ao João Martins (Escravo) e ao João Silva (Jonas) pelo apoio e ajuda, pela

enorme paciência que tiveram e pelos momentos de discussões e gargalhadas no

contentor / laboratório. À Joana Duarte pelo apoio constante, principalmente nesta fase

final (e mais difícil) e pela companhia (quase) constante nas longas viagens entre

Estarreja – Lisboa – Estarreja.

Anabela Nogueira…Sem sabermos crescemos na mesma zona, estudámos e

acabámos por nos encontrar no IST. A partir daí tornámo-nos quase inseparáveis, Lisboa,

Lyon, Estarreja e quem sabe o futuro? Ajudaste a tornar a decisão de vir para Estarreja

muito mais fácil e estiveste sempre presente, apoiaste, chamaste à razão e acreditaste em

mim, principalmente quando eu duvidei.

À Cristina Rodrigues, João Sáude, João Maria, Vitor Rodrigues, Ana Castro,

Carolina Rodrigues, Afonso Rodrigues, Fernanda Pimenta, Rui Pimenta, João Pimenta e

Tiago Pimenta, obrigada pela motivação e apoio.

iii

Ao meu irmão, André Sá Couto, por tudo que sempre partilhámos, por seres meu

irmão. Por seres chato e por seres tão amigo, por ter a certeza que estarás sempre

presente, que poderei contactar contigo incondicionalmente.

Aos meus pais, Maria Alice e José Sá Couto, por Tudo. Por me terem sempre

ensinado que devemos lutar por aquilo em que acreditamos, respeitando sempre o

próximo. Pelos valores que sempre me incutiram, pelo apoio incondicional e por sempre

me motivarem a desafiar-me a mim mesma, por sempre me terem colocado à frente de

tudo e por serem sempre o meu porto de abrigo. Aquilo que sou é o reflexo do que

sempre me transmitiram. As palavras nunca serão suficientes para agradecer tudo o que

fizeram e fazem por mim.

Ao Nuno Amorim, pela pessoa extraordinária que és, por sempre me teres

apoiado, por teres estado sempre a meu lado ao longo de todos estes anos. Nesta longa

caminhada, que implicou várias deslocações, nunca puseste em causa se eu seria capaz de

ultrapassar os desafios a que me propunha e estiveste sempre lá a incentivar-me. Obrigada

pelas muitas lágrimas que tiveste de limpar e alguma tristeza que tiveste de afugentar,

pelas muitas alegrias que partilhámos, pelos inúmeros sorrisos, por nunca duvidares, por

acreditares sempre em mim.

v

Abstract

Aniline (ANL) is an aromatic amine mainly consumed in the production of

methylene diphenyl diisocyanate (MDI). MDI is in turn a key raw material in the

polyurethane industry for the automotive and construction sectors. Worldwide ANL

capacity was around 5.4 million tons per annum in 2011, and between 75 to 85 % was

consumed for the production of MDI. There are around 30 companies producing ANL of

which 8 account for 66 % of the total production. Among them is CUF-QI that owns 4%

of the ANL production worldwide. ANL production is mainly done through nitrobenzene

(NB) hydrogenation. This reaction can be carried out either in liquid or in vapor-phase.

For that reason, several technological processes were developed to perform this industrial

production, which basically differ on the type of reactor; however, the most common are

the fixed-bed or the fluidized-bed for vapor-phase and the slurry reactor for liquid phase.

Catalyst development is also a key aspect for the NB hydrogenation and several

papers are available for both phases. In the NB hydrogenation into ANL, there is the

formation of several secondary products that leads to a lower productivity. Trying to

understand the formation of those compounds is very important and some information is

available in the open literature, although consensus has not yet been reached; moreover,

the species reported to be formed are different from work to work.

Catalysts to be later tested in a fixed-bed reactor were acquired and a multiphase

continuous stirred tank reactor (CSTR) operating in batch mode was firstly used to test

them. The first step was to study the mechanism of ANL and secondary products

formation as well as analysing the effect of the main operating conditions in this

multiphase reaction system. It was found that there are more by-products than those

referred in the literature and both NB consumption and selectivity are extremely

dependent on temperature. The first catalyst used, designated as I.1 (1 wt.% Pd/Al2O3)

proved to be selective to ANL formation. Besides, a new reaction network was proposed

for ANL and secondary products formation, where benzene (Bz) was included, since it

was not considered in a quantitative manner by any other previous authors in the

literature.

vi

The effect of the reaction products and of the use of different solvents, in this

reaction, was also analized over catalyst I.1. It was found that using p-toluidine (p-tol) as

solvent prevents the formation of secondary products when compared with ANL. The

presence of secondary products (namely Bz and water) in the feed mixture leads to a

decrease in the NB conversion.

Afterwards, the four commercial catalysts supplied, three Pd-based and one Ni-

based, were compared and it was found that they present different performances,

particularly different activities in what concerns the NB conversion and ANL selectivity.

Chemical and physical characterization of the catalysts used, namely catalyst I.1, catalyst

I.2, catalyst I.3 and catalyst II.1, was crucial to better understand their quite distinct

performances. Based on those results one of the catalysts was chosen, the one that

presented the highest NB consumption rate with a low secondary products formation,

catalyst I.2 (0.3 wt.% Pd/Al2O3).

One of the main objectives of this thesis was the design, construction and testing of

a laboratorial unit comprising a tubular reactor for the hydrogenation of NB into ANL.

The unit was designed and constructed and some preliminary tests were carried out to

ensure its proper functioning, and to evaluate the adequate temperature control and the

pressure drop in the catalytic bed.

Using the catalyst chosen before, several catalytic tests were performed in the

laboratorial trickle-bed tubular reactor. A parametric study was carried out to analyse the

effect of the operating conditions in the catalyst performance, namely on NB conversion

and selectivity towards ANL and secondary products. It was found that catalyst age is

extremely important as it changes the selectivity to the products formed along time-on-

stream, although NB conversion remains stable. In what concerns the influence of the

operating conditions, it was found that temperature and pressure are important and critical

parameters.

Then, it was decided to focus on some issues that are of paramount importance from

the perspective of industrial process implementation. In particular, it was decided to

evaluate the influence of the solvent and also to test if the catalyst was still active at mild

conditions of pressure and temperature. Cyclohexane (CH) seemed to be a good solvent,

however it promotes the formation of heavy secondary products. Relatively to the

operation under mild conditions, the Pd-based catalyst showed to be active but on the

vii

other hand it also leads to the formation of dicyclohexylamine (DICHA). In addition, the

influence of some reaction products, namely water (H2O) and cyclohexylamine (CHA),

were analysed to determine their influence on ANL selectivity and on secondary products

formation. Neither H2O nor CHA seem to have a significant influence on NB conversion,

although selectivity to ANL decreases.

To verify if it is possible to valorise the industrial stream of impure H2, some

analysis to that stream were carried out, in order to define the methodology to be used

when studying the effect of contaminants in the reaction. It was shown that the

contaminants that are present in higher quantities are ammonia (NH3), carbon dioxide

(CO2) as well as some organic compounds, mainly benzene (Bz). Among all, it was

decided to use NH3 as contaminant (because it is present in larger quantities) and it was

possible to conclude that NH3 concentrations up to 1 wt.% do not have a negative

influence in NB hydrogenation.

The ultimate goal was to test the industrial H2 stream, available at low pressures and

with the contaminants referred above. No major effect was detected in NB conversion at

any of the temperatures used (120 ºC and 150 ºC), nor in selectivity towards ANL. It was

also seen that heavy products formation is low.

Summarizing, it was proved that the industrial H2 stream available at CUF-QI can

actually be valorized to produce ANL in the range of operating conditions studied.

Nevertheless, some attention must be given to the composition of this stream, mainly to

the organic compounds eventually present, which can have some impact in the results

obtained namely in the composition of the outlet stream. More tests should be performed

to validate these conclusions and further explore this topic; however, it was demonstrated

that the trickle-bed tubular reactor can be used to produce ANL, by using an active Pd

supported catalyst, with good selectivity and high levels of NB conversion.

ix

Resumo

A anilina (ANL) é uma amina aromática consumida principalmente na produção de

metileno difenil diisocianato (MDI). O MDI é uma das principais matérias-primas da

indústria dos poliuretanos para os sectores automóvel e da construção. Em 2011, a

capacidade mundial de produção de ANL rondava os 5,4 milhões de toneladas por ano,

sendo que entre 75 % a 85 % era consumida na produção de MDI. Existem cerca de 30

empresas a produzir ANL, das quais 8 totalizam 66 % da produção global. Entre elas

encontra-se a CUF-QI que detém 4 % da cota de produção mundial de ANL. A produção

de ANL é essencialmente realizada através da hidrogenação de nitrobenzeno (NB), a qual

pode ocorrer quer em fase líquida, quer em fase gasosa. Assim, foram desenvolvidas

diversas tecnologias para esta reação, essencialmente relacionadas com o tipo de reator

mais adequado. Não obstante, os tipos de reatores mais comuns em fase gasosa são os de

leito-fixo ou de leito fluidizado, enquanto em fase líquida são os reatores agitados de

“lamas”.

O desenvolvimento de catalisadores é também um aspeto fundamental na reação de

hidrogenação de NB, sendo que existem inúmeros documentos disponíveis onde esta

temática é estudada, quer em fase gasosa, quer em fase líquida. Durante a hidrogenação

de NB a ANL existe a formação de produtos secundários, que conduzem a uma menor

produtividade. Compreender a formação desses compostos secundários é extremamente

importante e verifica-se que existe alguma literatura disponível, apesar de não existir

consenso sobre o esquema reacional. Além disso, também se constata que as espécies

identificadas diferem de estudo para estudo.

No âmbito desta tese, foram adquiridos alguns catalisadores comerciais para

hidrogenação de NB em leito-fixo, tendo-se recorrido numa primeira fase a um reator

agitado (CSTR), a operar em modo descontínuo, para os testar. O primeiro passo

consistiu no estudo do mecanismo de formação de ANL e produtos secundários, assim

como na análise do efeito das principais condições operatórias neste sistema reacional

multifásico. Constatou-se que existem mais compostos secundários do que os que são

referidos na literatura e que quer a velocidade de consumo de NB, quer a formação de

x

produtos secundários, são extremamente dependentes da temperatura. O catalisador

primeiramente testado, designado I.1 (1 % m/m Pd/Al2O3) demonstrou ser seletivo

relativamente à formação de ANL. Adicionalmente, foi proposto um novo esquema

reacional para a formação de ANL e dos compostos secundários, onde o benzeno (Bz) foi

incluído, uma vez que a sua formação não foi avaliada quantitativamente, por nenhum

autor na literatura existente.

O efeito dos produtos de reação e o uso de diferentes solventes, nesta reação, foram

também avaliados, usando o catalisador I.1. Verificou-se que o uso de p-toluidina (p-tol),

como solvente, evita a formação de produtos secundários quando comparado com o

solvente ANL. A presença de produtos secundários, na corrente de alimentação

(nomeadamente Bz e água), conduz a uma menor conversão de NB.

Posteriormente, os quatro catalisadores adquiridos, três à base de Pd e um à base de

Ni, foram comparados, tendo-se concluído que apresentam diferentes desempenhos,

nomeadamente diferentes atividades no que se refere à conversão de NB e seletividade à

ANL. A caracterização química e física dos catalisadores utilizados, catalisador I.1,

catalisador I.2, catalisador I.3 e catalisador II.1, foi crucial no entendimento dos seus

desempenhos tão distintos. Com base nestes resultados foi escolhido um dos

catalisadores, tendo-se optado pelo catalisador que apresentou maior velocidade de

consumo de NB e baixa formação de produtos secundários, ou seja, o catalisador I.2 (0.3

% m/m Pd/Al2O3).

Um dos principais objetivos desta tese consistiu no projecto, construção e validação

de uma unidade laboratorial compreendendo um reator tubular para hidrogenação de NB

a ANL. A unidade foi concebida e construída e alguns testes preliminares foram

efetuados com o intuito de assegurar-se o bom funcionamento da instalação e avaliar-se o

controlo de temperatura, assim como a queda de pressão no leito catalítico.

Usando o catalisador escolhido anteriormente, catalisador I.2, foram realizados

diversos testes catalíticos no reactor tubular. Foi efetuado um estudo paramétrico, de

forma a analisar o efeito das condições operatórias no desempenho do catalisador,

nomeadamente, na conversão de NB e seletividade à ANL e aos produtos secundários.

Constatou-se que o tempo de uso do catalisador (idade) é extremamente importante, uma

vez que ao longo do tempo há alterações na seletividade aos produtos secundários, apesar

da conversão de NB se manter estável. Relativamente à influência das condições

xi

operatórias, observou-se que a temperatura e a pressão são parâmetros importantes e

críticos.

Posteriormente, o foco do estudo foi direcionado para aspectos de elevada

importância do ponto de vista da implementação do processo a nível industrial. Mais

especificamente, decidiu-se avaliar a influência do solvente e também testar se o

catalisador permanece ativo em condições mais suaves de pressão e temperatura. O ciclo-

hexano (CH) demonstrou ser um bom solvente, contudo conduz a uma maior formação de

produtos secundários. Quanto às condições de operação mais suaves, o catalisador de Pd

demonstrou que é ativo, mas por outro lado conduz à formação de diciclo-hexilamina

(DICHA). Além destes ensaios, também se estudou a efeito de alguns produtos de reação,

nomeadamente água (H2O) e ciclo-hexilamina (CHA), com o objetivo de determinar a

sua influência quer na seletividade à ANL, quer na formação de produtos secundários.

Nenhum dos compostos, H2O ou CHA, parece exercer qualquer tipo de influência na

conversão de NB apesar de se registar uma diminuição na seletividade à ANL.

Para verificar a possibilidade de valorizar a corrente industrial de H2 impuro,

realizaram-se análises a essa mesma corrente, por forma a definir qual a metodologia a

seguir no estudo do efeito dos contaminantes da corrente gasosa. Verificou-se que os

principais contaminantes são o amoníaco (NH3), o dióxido de carbono (CO2) assim como

alguns compostos orgânicos, nomeadamente o Bz. Optou-se por estudar o efeito do NH3

(uma vez que está presente em quantidades elevadas) e concluiu-se que com

concentrações de NH3 até 1 % m/m não existe uma influência negativa na reação de

hidrogenação de NB.

O objetivo central desta tese consistiu no teste de uma corrente industrial de H2, que

está disponível a baixa pressão e com os contaminantes referidos anteriormente. Não foi

detetado qualquer tipo de influência na conversão de NB, independentemente da

temperatura utilizada (120 ºC ou 150 ºC), nem na seletividade à ANL. Além disso,

também se observou que a formação de produtos secundários pesados é baixa.

Concluindo, foi demonstrado que a corrente industrial de H2 existente na CUF-QI

pode efetivamente ser valorizada na produção de ANL, na gama de condições operatórias

estudadas. Não obstante, é necessário ter especial cuidado com a composição desta

corrente, nomeadamente, ter atenção aos compostos orgânicos presentes, que poderão ter

impacto nos resultados obtidos, principalmente na composição da corrente de saída. Para

xii

validar estas conclusões deverão ser realizados mais testes; todavia, foi demonstrado que

o reator tubular de leito fixo pode ser utilizado na produção de ANL, usando um

catalisador de Pd suportado ativo, obtendo-se boas seletividades e elevados níveis de

conversão de NB.

xiii

Table of Contents

List of Figures ....................................................................................................................... xix

List of Tables ..................................................................................................................... xxvii

Nomenclature ...................................................................................................................... xxix

Part I - Introduction and State of Art

Chapter 1 - Introduction ........................................................................................................... 3

Chapter 2 - State of the Art ...................................................................................................... 7

2.1 Aniline industrial production and applications .................................................................. 7

2.2 Technological aspects of the Industrial Production of Aniline ........................................ 13

2.2.1 Aniline .................................................................................................................... 13

2.2.2 Reaction mechanisms for aniline production and by-products formed ................. 14

2.2.3 Hydrogenation in Gas-phase ................................................................................. 28

2.2.4 Hydrogenation in Liquid-phase ............................................................................. 30

2.2.4.1 DuPont Process ............................................................................................... 30

2.4.2.2 Huntsman Process ........................................................................................... 31

2.4.2.3 Mitsui Process ................................................................................................. 33

2.4.2.4 Chematur Process ............................................................................................ 33

2.4.2.5 CUF-QI Process .............................................................................................. 34

2.4.2.6 Bechamp Process ............................................................................................ 35

2.2.5 Catalysts for Aniline production............................................................................ 37

2.2.5.1 Catalysts for vapor-phase processes ............................................................... 40

2.2.5.2 Catalysts for liquid-phase processes ............................................................... 42

2.2.6 Types of reactors ................................................................................................... 44

References .............................................................................................................................. 54

xiv

Part II - Preliminary catalytic tests in a Continuous Stirred-Tank Reactor (CSTR)

Chapter 3 - Hydrogenation of Nitrobenzene over a Pd/Al2O3 Catalyst –

Mechanism and Effect of the Main Operating Conditions. ................................................... 63

Abstract .................................................................................................................................. 63

3.1 Introduction ...................................................................................................................... 64

3.2 Material and Methods ...................................................................................................... 67

3.3 Results and Discussion .................................................................................................... 69

3.3.1 Influence of initial nitrobenzene concentration ..................................................... 69

3.3.2 Influence of Pressure ............................................................................................. 75

3.3.3 Influence of Temperature....................................................................................... 77

3.4 Conclusions ...................................................................................................................... 82

References .............................................................................................................................. 83

Chapter 4 – Study of Effects of the Solvent and Reaction Products in the Catalytic

Hydrogenation of Nitrobenzene. ............................................................................................ 85

Abstract .................................................................................................................................. 85

4.1 Introduction ...................................................................................................................... 86

4.2 Material and Methods ...................................................................................................... 90

4.3 Results and Discussion .................................................................................................... 93

4.3.1 Influence of the solvent .......................................................................................... 94

4.3.2 Influence of the presence of reaction products in the feed .................................... 97

4.3.2.1 Effect of H2O ........................................................................................................... 98

4.3.2.2 Effect of Benzene .......................................................................................... 101

4.3.2.3 CHA hydrogenation ...................................................................................... 102

4.3.2.4 ANL hydrogenation ...................................................................................... 104

4.4 Conclusions .................................................................................................................... 107

References ............................................................................................................................ 108

Chapter 5 - Commercial Catalysts Screening for Liquid Phase Nitrobenzene

Hydrogenation...................................................................................................................... 111

Abstract ................................................................................................................................ 111

5.1 Introduction .................................................................................................................... 112

5.2 Material and Methods .................................................................................................... 113

xv

5.2.1. Catalyst samples ................................................................................................. 113

5.2.2. Catalysts Characterization ................................................................................. 114

5.2.3. Catalytic Reaction .............................................................................................. 115

5.3 Results and Discussion .................................................................................................. 118

5.3.1 Catalysts Characterization .................................................................................. 118

5.3.2 Nitrobenzene Hydrogenation ............................................................................... 122

5.3.2.1 Catalysts activity ........................................................................................... 124

5.3.2.2 Catalysts selectivity ...................................................................................... 128

5.4 Conclusions .................................................................................................................... 134

References ............................................................................................................................ 136

Part III - Catalytic Tests in a Tubular Reactor

Chapter 6 - Tubular Reactor Laboratorial Unit .................................................................... 141

6.1 - Introduction ................................................................................................................. 141

6.2 - Unit conception............................................................................................................ 142

6.2.1 - Unit purpose ...................................................................................................... 143

6.2.2 - Unit description ................................................................................................. 143

6.2.2.1 – Liquid feed section ..................................................................................... 146

6.2.2.2 – Gas feed section ......................................................................................... 147

6.2.2.3 – Reaction section ......................................................................................... 148

6.2.2.4 – Separation section ...................................................................................... 151

6.3 – Preliminary tests.......................................................................................................... 152

6.3.1 - Test with catalyst support, H2O and H2 ............................................................ 153

6.3.2 - Test with catalyst support, ANL and H2 ............................................................ 155

References ............................................................................................................................ 157

Chapter 7 - Hydrogenation of Nitrobenzene in a Tubular Reactor: Parametric

Study of the Operating Conditions Influence ...................................................................... 159

Abstract ................................................................................................................................ 159

7.1 Introduction .................................................................................................................... 160


xvi


7.3.1 Reproducibility tests ............................................................................................ 167

7.3.2 Influence of Total Pressure.................................................................................. 169

7.3.3 Influence of Temperature..................................................................................... 172

7.3.4 Influence of Liquid Feed Flow Rate .................................................................... 174

7.3.5 Influence of NB Concentration in the Feed ......................................................... 175

7.4 Conclusions .................................................................................................................... 176

References ............................................................................................................................ 178

Chapter 8 - Industrial Perspective of Nitrobenzene Catalytic Hydrogenation in a

Tubular Reactor – Impure H2 valorization........................................................................... 181

Abstract ................................................................................................................................ 181

8.1 Introduction .................................................................................................................... 182



8.3.1 Influence of the solvent ........................................................................................ 189

8.3.2 Influence of H2O .................................................................................................. 191

8.3.3 Influence of CHA ................................................................................................. 193

8.3.4 Reaction at mild conditions (T and P) ................................................................. 195

8.3.5 Influence of impure H2 ......................................................................................... 196

8.3.5.1 Influence of NH3 ........................................................................................... 197

8.3.5.2 Industrial H2 .................................................................................................. 199

8.4 Conclusions .................................................................................................................... 202

References ............................................................................................................................ 204

Part IV - General Conclusions and Future Work

Chapter 9 - General Conclusions ......................................................................................... 209

Chapter 10 - Future Work .................................................................................................... 213

10.1 Catalysts ....................................................................................................................... 213

10.2 Tubular reactor ............................................................................................................. 214

10.3 Kinetic studies .............................................................................................................. 214

xvii

Appendixes

Appendix A – Supporting Information of Chapter 3. .......................................................... 217

Appendix B – Supporting Information of Chapter 5 ........................................................... 223

Appendix C - Resume of the operating conditions used in the catalytic tests with

the tubular reactor (Chapters 7 and 8). ................................................................................. 229

Appendix D - Complementary results of the parametric study in Chapter 7....................... 231

xix

List of Figures

Figure 2.1 - ANL market share for 2010 [2]. .......................................................................... 8

Figure 2.2 - Global ANL capacity by producer in 2011, adapted from [1]. ............................ 9

Figure 2.3 – Network of chemical complex of Estarreja [3]. ................................................ 11

Figure 2.4 – Main world Producers of ANL (2013) [3]. ....................................................... 12

Figure 2.5 – Schematic diagram of CUF-QI plant [3]. .......................................................... 12

Figure 2.6– Reaction network involved in nitrobenzene hydrogenation, Haber

mechanism [9]. ....................................................................................................................... 15

Figure 2.7 – Reaction network of nitrobenzene hydrogenation, as proposed by

Wisniak and Klein [13]. ......................................................................................................... 16

Figure 2.8 – Scheme of components transformation on catalytic surface, proposed

by Makaryan [14]. .................................................................................................................. 16

Figure 2.6 - Reaction network of nitrobenzene hydrogenation proposed by Gelder

et al. [7]. ................................................................................................................................. 17

Figure 2.10 – Proposed reaction pathway for the hydrogenation of aromatic nitro

compound to aniline [15]. ...................................................................................................... 18

Figure 2.11 - Scheme of NB catalytic hydrogenation in the presence of Pd-

containing heterogeneous catalyst, [19]. ................................................................................ 19

Figure 2.12 – Supplemented reaction mechanism for NB hydrogenation

considering Haber’s and Gelder’s reaction mechanism, proposed by Turáková et

al. [20]. ................................................................................................................................... 20

Figure 2.13 – Reaction network for the formation of ANL and secondary products

proposed by Nagata [21]. ....................................................................................................... 21

Figure 2.14 - Reaction network proposed by Narayanan et al. [23]. ..................................... 22

Figure 2.15 – Reaction network proposed for ANL and secondary products

formation by Relvas [24]. ...................................................................................................... 23

Figure 2.16 – Reaction network proposed for secondary products formation from

ANL hydrogenation by Králik et al. [25]. .............................................................................. 24

Figure 2.17 – Reaction network proposed in liquid phase hydrogenation of NB by

Králik et al. [24] ..................................................................................................................... 25

Figure 2.18 – Reaction network proposed for the Pd/C catalyzed hydrogenation of

NB by Rubio-Marqués et al. [26] .......................................................................................... 26

xx

Figure 2.19 – Reaction network for the formation of secondary products during

NB hydrogenation in the presence of Ni supported catalyst, proposed by Sousa

[27]. ........................................................................................................................................ 27

Figure 2.20 – Fluidized-bed ANL process in vapour-phase [1]. ........................................... 28

Figure 2.21 – DuPont ANL Process via liquid-phase [1]. ..................................................... 31

Figure 2.22 – Huntsman ANL Process via liquid-phase [1]. ................................................. 32

Figure 2.23 – Chematur ANL Process [1]. ............................................................................ 34

Figure 2.24 – CUF-QI ANL process...................................................................................... 35

Figure 2.25 – Typical concentration profiles during hydrogenation of NB [4]. .................... 38

Figure 2.26 – The two modes of reactants introduction in a catalytic membrane

reactor [59]. ............................................................................................................................ 46

Figure 2.27 – Catalytic wall reactor configuration [75]. ....................................................... 47

Figure 2.28 – Configuration proposed in US 2000/6040481 [43]. ........................................ 49

Figure 2.29 – Process flow by Huntsman [76]. ..................................................................... 50

Figure 3.1 – Reaction network for the formation of ANL and secondary products

as proposed by a) Nagata et al. [22]; b) Narayanan and Unnikrishnan [23]. ......................... 66

Figure 3.2 –Relvas [24] (*very reactive and unstable compounds)....................................... 67

Figure 3.3– Influence of initial nitrobenzene concentration in the secondary

products formation (Bz, CHA, CHOL, CHONA, NB and DICHA) vs. time, runs

B4, B7 and B11. ..................................................................................................................... 70

Figure 3.4– Influence of initial nitrobenzene concentration in the secondary

products formation (CHENO and CHANIL) vs. time, runs B4, B7 and B11. ...................... 71

Figure 3.5 – Influence of initial nitrobenzene concentration in the ANL formation

a) and NB conversion b) vs. time, runs B4, B7 and B11. ...................................................... 72

Figure 3.6 – Comparison between total secondary products formation (closed

symbols) and NB consumption (open symbols) as a function of reaction time for

different initial NB concentrations; runs B4, B7 and B11. .................................................... 73

Figure 3.7 - Influence of nitrobenzene initial concentration in the secondary

products formation for NB dimensionless concentration, runs B4, B7 and B11. .................. 74

Figure 3.8 - Influence of nitrobenzene initial concentration in the secondary

products formation for NB dimensionless concentration, runs B4, B7 and B11. .................. 75

Figure 3.9 - Influence of reaction pressure in the secondary products (Bz, CHA,

CHOL, CHONA, ANL, DICHA, CHENO and CHANIL) vs. time, runs B2, B3

and B4. ................................................................................................................................... 76

xxi

Figure 3.10 – Comparison between a) ANL formation and b) total of secondary

products formation (closed symbols) and NB consumption (open symbols) as a

function of reaction time; runs B2, B4 and B5. ..................................................................... 77

Figure 3.11 - Influence of reaction temperature in the ANL and by-products

formation (Bz, CHA, CHOL, CHONA and DICHA) vs. reaction time, runs B4,

B5, B9 and B10. ..................................................................................................................... 78

Figure 3.12 - Influence of reaction temperature in the ANL and by-products

formation (CHENO and CHANIL) vs. reaction time, runs B4, B5, B9 and B10. ................ 79

Figure 3.13 – Comparison between a) NB conversion and b) total secondary

products formation (closed symbols) and NB consumption (open symbols) as a

function of reaction time for different reaction temperatures; runs B4, B5, B9 and

B10. ........................................................................................................................................ 80


formation including Bz (*very reactive and unstable compounds). ...................................... 81

Figure 4.1 - Reaction network involved in nitrobenzene hydrogenation illustrating

intermediary species proposed by a) Haber [8] and b) Turáková et al. [14]. ........................ 87


secondary products formation proposed by Relvas [18]........................................................ 88


secondary products formation proposed by Sousa [21]. ........................................................ 89

Figure 4.4 – Scheme of the reactor and set-up used in the experiments. ............................... 91

Figure 4.5 – Evolution of a) NB and b) ANL as a function of reaction time for

different solvents - ANL and ANL + 28 wt. % P-tol (runs TB7 and TB8). .......................... 94

Figure 4.6 – Evolution of secondary products concentration as a function of

reaction time for different solvents - ANL and ANL + 28 wt. % P-tol (runs TB7

and TB8). ............................................................................................................................... 95

Figure 4.7 – Evolution of the concentration of a) light products and b) heavy

products, c) secondary products with ANL as solvent and d) secondary products

with ANL + 28 wt.% p-tol as solvent, along reaction time for different solvents

(runs TB7 and TB8). .............................................................................................................. 96


formation including Bz (*very reactive and unstable compounds). ...................................... 97

Figure 4.9 – Evolution of a) ANL concentration, b) secondary products

concentration, c) light products concentration, d) heavy products concentration, e)

secondary products concentration distribution for ANL in the reactor feed and f)

secondary products concentration distribution for ANL+ 1 wt.% H2O in the

reactor feed, along reaction time (runs TC1 and TC4). ......................................................... 99

xxii

Figure 4.10 – Evolution of a) NB concentration, b) ANL formation, c) secondary

products concentration, along reaction time (runs TC3 and TC6). ...................................... 101

Figure 4.11 – Evolution of a) secondary products concentration, b) light products

concentration, c) heavy products concentration, along reaction time (runs TC1 and

TC5). .................................................................................................................................... 103

Figure 4.12– Evolution of a) ANL concentration, b) secondary products

concentration c) light products concentration and d) heavy products concentration,

along reaction time (runs TC1 and TC2). ............................................................................ 104

Figure 4.13– Evolution of a) secondary products concentration distribution for 150

ºC and 14 barg and b) secondary products concentration distribution for 200 ºC

and 20 barg, along reaction time (runs TC1 and TC2). ....................................................... 105

Figure 4.14– Evolution of a) CHONA concentration, b) CHENO concentration

along reaction time (runs TC1 and TC2). ............................................................................ 106

Figure 5.1– X-ray diffraction patterns of the fresh catalysts studied: a) catalyst I.1,

b) catalyst I.2, c) catalyst I.3 and d) catalyst II.1. ................................................................ 119

Figure 5.2 – Particle size distribution of fresh group I catalysts determined by

HRTEM................................................................................................................................ 120

Figure 5.3 – Temperature programmed reduction profiles for the fresh Pd-based

(a) catalyst I.1, b) catalyst I.2, c) catalyst I.3) and Ni-based (d) catalyst II.1)

materials studied. ................................................................................................................. 121

Figure 5.4 – Reproducibility tests, showing NB consumption as a function of

reaction time at 150 ºC, 14 barg and 10% NB for each catalyst. ......................................... 122

Figure 5.5 – Reproducibility tests, showing NB consumption as a function of

reaction time at 150 ºC, 14 barg and 10% NB for each catalyst. ......................................... 123

Figure 5.6– Reaction network proposed for formation of ANL and secondary

products [10]. *very reactive and unstable compound. ....................................................... 124

Figure 5.7– Effect of reaction total pressure on NB consumption as a function of

reaction time for the different catalysts: a) P = 6 barg, b) P = 14 barg and c) P = 30

barg. ..................................................................................................................................... 125

Figure 5.8- Effect of reaction temperature on NB consumption as a function of

reaction time for the different catalysts: a) T = 150 ºC, b) T = 180 ºC and c) T =

240 ºC. .................................................................................................................................. 126

Figure 5.9 – Comparison of NB consumption rate for all operating condition used

a) per gram of catalyst and b) per gram of metal. ................................................................ 127

Figure 5.10 - Light products and Heavy products concentration at Tref as a

function of reaction time for different pressures: a) and b) P = 6 barg. ............................... 129

xxiii

Figure 5.11 - Light products and Heavy products concentration at Tref as a

function of reaction time for different pressures: a) and c) P = 14 barg and b) and

d) 30 barg. ............................................................................................................................ 130

Figure 5.12 – Total secondary products concentration at Tref as a function of

reaction time a) P = 6 barg, b) P = 14barg and c) 30 barg. .................................................. 131

Figure 5.13 – Light products and Heavy products concentration at Pref as a

function of reaction time at: a) and d) T = 150 ºC, b) and e) T = 180 ºC and c) and

f) 240 ºC. .............................................................................................................................. 132

Figure 5.14 – Total secondary products concentration at Pref as a function of

reaction time a) 150 ºC Tref and b) 180 ºC. ......................................................................... 133

Figure 5.15 – Total secondary products concentration at Pref as function of

reaction time: 240 ºC............................................................................................................ 134

Figure 6.1 – Tubular reactor unit P&ID............................................................................... 145

Figure 6.2 – Photos of the liquid feed section. .................................................................... 147

Figure 6.3 – Photos of the gas section. ................................................................................ 148

Figure 6.4 – Photos of the reaction section, with closed (left) and open (right)

views of the oven. ................................................................................................................ 149

Figure 6.5 – Tubular reactor: a) reactor bed distribution and b) thermocouples

positions. .............................................................................................................................. 150

Figure 6.6 – Photos of the separation section. ..................................................................... 151

Figure 6.7– Tubular reactor unit overview. ......................................................................... 152

Figure 6.8 – Oven program for preliminary test1. ............................................................... 154

Figure 6.9 – Results obtained for: a) Reactor and oven temperatures, b) Reactor

temperatures, c) Pressure and d) Gas flow rate in test1. ...................................................... 154

Figure 6.10 – Oven program for preliminary test2. ............................................................. 155

Figure 6.11 – Results obtained for a) Reactor and oven temperatures, b) Reactor

temperatures, c) Pressure and d) Gas flow rate in test2. ...................................................... 156

Figure 7.1 – Scheme of the tubular reactor used for the catalytic tests. .............................. 162

Figure 7.2 – Evolution of a) NB conversion and b) Selectivity to ANL and

secondary products, as a function of reaction time for all tests of the parametric

study. .................................................................................................................................... 166

Figure 7.3 - Evolution of NB conversion as a function of reaction time for the

reproducibility tests. ............................................................................................................. 167

xxiv

Figure 7.4 - Evolution of a) Temperature of thermocouple TTr2, b) Pressure, c)

NB conversion and d) H2 consumption in transient state for reproducibility tests

TR5a) and TR10a). .............................................................................................................. 168

Figure 7.5 - Evolution of a) NB conversion and b) selectivity to ANL for different

total pressures....................................................................................................................... 169

Figure 7.6 - Evolution of a) selectivity to secondary products and b) Secondary

products selectivity distribution for different total pressures. ............................................. 170

Figure 7.7 - Reaction network proposed for ANL and secondary products

formation including Bz (*very reactive and unstable compounds). .................................... 171


temperatures at 14 barg. ....................................................................................................... 172

Figure 7.9 - Evolution of a) selectivity to secondary products and b) Secondary

products selectivity distribution for different temperatures at 14 barg. ............................... 173

Figure 7.10 - Evolution of a) NB conversion, b) selectivity to ANL, c) selectivity

to secondary products and d) secondary products selectivity distribution, for

different feed flows rates at 150ºC and 14 barg. .................................................................. 174

Figure 7.11 - Evolution of a) NB conversion and b) selectivity to ANL for

different NB concentrations at 120 ºC and 14barg. ............................................................. 175

Figure 7.12 - Evolution of a) selectivity to secondary products and d) Secondary

products selectivity distribution, for different NB concentrations at 120 ºC and

14barg. ................................................................................................................................. 176

Figure 8.1 – Scheme of the set-up and tubular reactor used for the catalytic tests. ............. 185

Figure 8.2 – Evolution of a) NB conversion, b) selectivity to ANL at 120ºC and 14

barg. ..................................................................................................................................... 189

Figure 8.3 – Evolution of a) selectivity to secondary products and b) secondary

products selectivity distribution for different solvents (ANL and CH) at 120ºC and

14 barg. ................................................................................................................................ 190


formation including Bz (*very reactive and unstable compounds). .................................... 191

Figure 8.5 - Evolution of a) NB conversion, b) selectivity to ANL, c) selectivity to

secondary products and d) secondary products selectivity distribution for different

H2O concentrations at 120ºC and 14 barg. .......................................................................... 192

Figure 8.6 - Evolution of a) NB conversion, b) selectivity to ANL in the presence

of CHA c) selectivity to secondary products and d) secondary products selectivity

distribution in the presence of CHA at 120ºC and 14 barg. ................................................. 194


pressures at low temperature (75 ºC). .................................................................................. 195

xxv

Figure 8.8 - Evolution of a) selectivity to secondary products and b) secondary

products selectivity distribution for different pressures at low temperature (75 ºC). .......... 196

Figure 8.9 - Evolution of a) NB conversion, b) selectivity to ANL, c) selectivity to

secondary products and d) secondary products selectivity distribution for different

NH3 concentrations at 120ºC and 14 barg. .......................................................................... 198

Figure 8.10 - Comparison of a) NB conversion, b) selectivity to ANL, c)

selectivity to secondary products, at 120 º C and 150 ºC, as a function of pressure

with pure hydrogen and impure industrial hydrogen grade. ................................................ 200

Figure 8.11 - Comparison of a) selectivity to light products, b) selectivity to heavy

products at 120 and 150 ºC and c) secondary products selectivity distribution at

120 ºC and d) Secondary products selectivity distribution at 150 ºC as a function

of pressure with pure hydrogen and impure industrial hydrogen grade. ............................. 201

xxvii

List of Tables

Table 2.1– West Europe ANL capacity by producer in 2011 [1] .......................................... 10

Table 2.2– Main chemicals produced at CUF-QI in 2014 and their applications [3]. .......... 10

Table 2.3 – ANL properties [1].............................................................................................. 13

Table 2.4 – Typical ANL sales specification [1]. .................................................................. 14

Table 2.5 – Main ANL Vapour-phase Processes. .................................................................. 30

Table 2.6 – Summary of ANL liquid-phase processes. ......................................................... 36

Table 2.7 – Industrial ANL applications [41]. ....................................................................... 39

Table 2.8 – Experimental conditions used in the several tests of the Hunstman patent [76]. 52

Table 2.9 – Experimental results of the tests described in Table 2.7 [76]. ............................ 53

Table 3.1– Main catalysts studied for NB hydrogenation. .................................................... 64

Table 3.2 - Initial conditions of the experiments performed. ................................................ 68

Table 4.1 - Initial conditions of the experiments performed. ................................................ 92

Table 4.2 – ANL/H2O system solubility [31] ........................................................................ 98

Table 5.1– Catalysts main physical characteristics. ............................................................. 114

Table 5.2 - Initial conditions of the experiments performed. .............................................. 117

Table 5.3 – Textural parameters for the catalysts samples studied...................................... 122

Table 6.1 – Main instruments characteristics. ..................................................................... 146

Table 6.2 – Main equipment characteristics. ....................................................................... 146

Table 8.1 – ANL/H2O system solubility [27] ..................................................................... 192

Table 8.2 – Composition of industrial H2. .......................................................................... 197

xxix

Nomenclature

The nomenclature used in the manuscript will vary since it depends on the different

authors.

Aniline ANL / Ar-NH2

Arylhydroxylamine Ar-NHOH / PHA

Azobenzene Ar-N=N-Ar / AZB

Azoxybenzene Ar-NO=N-Ar / AZXB

Benzene Bz

Catalytic wall reactor CWR

Coke oven light oil COLO

Cyclohexane CH

Cyclohexanol CHOL

Cyclohexanone CHONA

Cyclohexylamine CHA

Cyclohexyldeneaniline CHANIL

Dicyclohexylamine DICHA / DCHA

Diphenylamine DPA

Diphenylmethane diamine MDA

Direct methanol fuel cells DMFC

Hydrazobenzene Ar-NH=NH-Ar / HB

Hydroxyapatite HAP

kilotons per annum Kta

Methylene diphenyl diisocyanate MDI

N-cyclohexylaniline CHENO

Nitrobenzene NB

Nitrosobenzene Ar-NO / Ph-NO / NSB

N-phenylcyclohexylamine NPCHA

Supercritical carbon dioxide ScCO2

Toluidine TLD / p-tol

Water-gas shift reaction WGSR

Part I

Introduction and State of the Art

3

Chapter 1 - Introduction

The catalytic hydrogenation of nitrobenzene (NB) is an important industrial

reaction used in the commercial production of aniline (ANL), for subsequent use mainly

in the polyurethane industry. A mechanism for the reaction was first proposed by Haber

in 1898 and has been widely accepted despite never being fully delineated. This reaction

can be carried out in gas or in liquid-phase, and both alternatives are widely used by

world producers.

In this work, a review of the ANL industry, as well as of the technologies available

for its production, will be firstly done with the purpose of contextualizing the objective of

this PhD thesis. The main goal of the thesis is to valorize an industrial stream of hydrogen

that is available at CUF-QI at low pressures and has some contaminants. In this way, very

active catalysts must be used (e.g. consisting in supported noble metals). If the catalyst is

not in powder form, the most suitable reactors are those with a fixed-bed. Consequently, a

tubular reactor with a fixed-bed configuration was chosen to perform the catalytic

hydrogenation of NB into ANL.

Chapter 2 is dedicated to the presentation of the ANL market, the CUF-QI position,

the technological aspects of the ANL production, such as the formation of intermediary

compounds and of secondary products, and the type of reactors used in this process

(either for vapor as for liquid-phase). Most used and appropriated catalysts for this

reaction will be also discussed and it will be carried out a description of some new reactor

configurations that have been proposed.

Chapter 3 is related with the first results obtained with a commercial catalyst for the

NB hydrogenation in liquid-phase. The catalyst used was a 1 wt.% Pd/Al2O3 in pellets

form and it was tested in a batch reactor. The main goal is to evaluate the performance of

this type of catalysts in this multiphase reaction and also to understand the mechanism

behind ANL and secondary products formation. The influence of the main operating

conditions is also analyzed, namely of temperature, pressure and NB concentration.

Impure Hydrogen Valorization for Chemicals Production in a Tubular Reactor

4

In Chapter 4 will be analyzed the effect of the solvent as well as of the presence of

reaction products in the reaction mixture, for the hydrogenation of NB into ANL using

the same catalyst as in Chapter 3 (1 wt.% Pd/Al2O3). Besides, direct ANL and CHA

hydrogenation studies will also be presented. The main goal is to evaluate the influence of

those parameters in the catalyst performance, activity and selectivity to both ANL and

secondary products.

Chapter 5 shows the catalytic behavior of several commercial catalysts that were

supplied by different manufacturers. In order to have a better know-how about the

performance of those catalysts and select the most active one, with low formation of

secondary products, catalytic tests are performed in the batch reactor unit. Operating

conditions, like temperature and pressure, are varied and a catalyst screening is done with

the purpose of selecting the best on for further works. Moreover, all the catalysts are

characterized by different chemical and physical techniques and their relationship with

the hydrogenation performance is discussed.

Chapter 6 presents the design and construction of the tubular reactor aimed at

testing the possibility of producing ANL using the impure H2 stream that is available in

the plant. In this section, a detailed presentation of the unit design and construction is

done: unit conception, unit purpose and unit description, as well as technical and

operational details. It will be also presented some preliminary tests that were performed

with the objective of evaluating the temperature control and pressure drop issues in the

trickle-bed reactor.

In Chapter 7, the chosen catalyst of Chapter 5 is tested in the tubular reactor that

was built. The influence of several parameters is analyzed, like temperature, pressure,

liquid feed flow rate and NB concentration in the feed. Catalyst performance and

selectivity towards ANL and secondary products are important questions that are

discussed and analysed in detail.

Chapter 8 presents the results obtained on the trickle-bed tubular reactor with the

industrial H2 stream. In this section, the same catalyst sample that was tested on Chapter 7

is used to study the hydrogenation reaction; some keys factors are analyzed from an

industrial perspective. The effect of the solvent, the presence of some reaction products in

the liquid feed stream as well as of some contaminants present on the industrial H2 is

Chapter 1 - Introduction

5

discussed. The feasibility of using the industrial stream, that is available at low pressures,

is also investigated.

Chapter 9 presents the main conclusions that were achieved with this work in terms

of commercial catalysts performance and understanding of the mechanism behind ANL

and secondary products formation, using this type of catalysts. Conclusions related with

the use of a tubular fixed-bed reactor in the NB hydrogenation into ANL are also

presented. Finally, response is given to the main objective of this thesis: the possible

valorization of an industrial H2 stream, which results from other industrial processes and

is available at low pressures.

In Chapter 10 some suggestions will be put forward for future work.

7

Chapter 2 - State of the Art

2.1 Aniline industrial production and applications

Aniline (ANL) is mainly consumed in the production of methylene diphenyl

diisocyanate (MDI), which is a raw material for polyurethanes, that are mainly used in the

automotive and construction sectors. Polyurethanes have very different formulations and

can thus be used in the form of flexible or rigid-foams, elastomers, coatings, adhesives

and low molecular weight additives.

Legislation for energy-efficient buildings is pushing up the use of polyurethane-

based building materials as they are more insulating than the competitor products

(mineral fiber and polyester). Therefore, more extensive use of MDI in building

insulation will provide additional drivers for market growth. This is particularly the case

of Europe where the energy usage in buildings accounts for almost half of all energy

consumption [1], and so legislation is being implemented to meet EU targets for energy

efficiency of new buildings. Use of insulation is estimated to reduce energy usage by 30

to 50% when retrofitted into existing buildings and by as much as 90 to 95% in new

buildings, offering the possibility of significantly lower utility bills to the domestic

consumer at a time of inflationary pressures and economic instability [1].

Worldwide ANL capacity reached about 5.4 million tons in 2011. Depending on the

geographical location, around 75 to 85% is consumed for the production of MDI via

condensation of ANL with formaldehyde to give diphenylmethane diamine (MDA) that is

then reacted with phosgene. Other uses of ANL are predominantly in rubber processing

chemicals, such as vulcanization accelerators, antioxidants, antiozonates, and stabilizers.

Smaller uses include agrochemical intermediates and chemicals, pesticides (fungicides)

and herbicides. Miscellaneous uses for ANL include cyclohexylamine (CHA) for boiler

treatment, rubber chemicals, pharmaceuticals, textile chemicals, photographic developers,

amino resins, explosives, and specialty fibers (Kevlar, Nomex) [1]. Azo-dyes were once a

substantial consumer of ANL but now only account for a small fraction of demand. A

new interesting area for ANL consumption is the preparation of fuel cell membranes as in


8

the direct methanol fuel cells (DMFC). Oxidative polymerization of ANL adsorbed on a

perfluorosulfonic acid membrane gives a polyaniline layer which acts as a barrier towards

methanol without loss of proton conductivity.

In 2010, global ANL market distribution was the one shown in Figure 2.1 [2].

Figure 2.1 - ANL market share for 2010 [2].

In China, integrated coal to ANL facilities are under construction including those of

Jilin Connell and Shanxi Tianji Coal Chemical. Raw material hydrogen will be produced

from coal gasification and the benzene (from refining of the Coke Oven Light Oil -

COLO), as a by-product from coke production, since COLO production increased

dramatically (in line with the growth of coke demand for the burgeoning iron and steel

industry in China). Then, hydrogen will be used in ammonia manufacture, which is the

raw material for nitric acid manufacture. Sulphur from coal is also used to make sulphuric

acid. Nitrobenzene (NB) is produced from the nitration of benzene with a mixture of

nitric and sulphuric acid and is then hydrogenated to make ANL. In this case, all the

feedstocks for NB and ANL can be derived from coal, however outside China, on a

global basis the majority of the feedstocks still come from natural gas and oil.

The global capacity for ANL in 2011 was estimated at 5357 kilotons per annum

(kta). There are around 39 companies producing ANL of which 8 account for 66% of the

production, as shown in Figure 2.2. By 2016, it is estimated that the global capacity for

MDI 75%

Others, 7%

Rubbers, 11%

Dyes, 7%

Chapter 2 – State of the Arte

9

ANL will increase to 6647 kta, a growth between 2012 and 2016 of about 4.4% per year

[1].

Figure 2.2 - Global ANL capacity by producer in 2011, adapted from [1].

In North America DuPont is the largest manufacturer with 41% of the capacity,

while Rubicon has the largest single plant (420 kta) and owns 37% of the installed

capacity.

In West Europe, ANL capacity in 2011 amounted to 1574 kta. Analyzing Table 2.1

it is possible to conclude that Bayer is the largest producer with 38% of the capacity.

BASF has 22%, Huntsman 19%, CUF-QI 13% and Dow 8%.

CUF, 4%

Other , 34%

Bayer, 17%

BASF, 11%DuPont, 9%

Rubicon, 8%

Yantai Wanhua

Polyurethane, 7%

Hunstman, 5%

Tosoh Corporation,

5%


10

Table 2.1– West Europe ANL capacity by producer in 2011 [1]

Company Location Capacity (thousand tons per annum)

BASF Antwerp 342 (22%)

Bayer Antwerp 165 (10%)

Bayer Antwerp 185 (12%)

Bayer Brunsbuettel 100 (6%)

Bayer Krefeld-Uerdingen 152 (10%)

Dow Bohlen 130 (8%)

CUF Estarreja 200 (13%)

Huntsman Wilton 300 (19%)

In China the estimation for ANL capacity was 1767 kta in 2011, spread by 19

suppliers being Yantai Wanhua the largest producer with 20% of the capacity. Japan has

only 5 producers, with a total capacity of 448 kta (2011), being Tosoh Corporation the

largest manufacturer with 67% of the capacity. Companies in the rest of the world are

estimated to have accounted for 534 kta of capacity in 2011. The largest single supplier is

Yantai Wanhua via its Borsodchem subsidiary’s plants in Ostrava, which has a total

capacity of 190 kta [1]

CUF-QI, SA is one of the companies owned by José de Mello, SGPS group

developing its activities in the chemical industry area. CUF-QI is located at the chemical

complex of Estarreja, Portugal. The chemicals produced at Estarreja are nitric acid, NB,

ANL, sulphanilic acid, CHA, hydrochloric acid, among others (Table 2.2).

Table 2.2– Main chemicals produced at CUF-QI in 2014 and their applications [3].

Compound Sales Volume Application

ANL 69.5% MDI production, rubber industry, paints and

pigments, special fibers

NB 7.5% ANL production, chemical and pharmaceutical

industry

Liquid Chlorine 6.7% PVC production, polyurethanes, water treatment

Sodium

Hydroxide 8.5%

Chemical, textile, cellulose, food, detergents and

soap industry.

Hypochlorite 3.3% Water treatment, hygiene and cleaning products,

textile blanching


11

In CUF-QI in Estarreja, the organic compounds are exclusively destined to external

markets, either directly or indirectly through DOW, and a considerable amount of the

inorganic compounds is also for exportation. In Figure 2.3 is presented the network of the

chemical complex of Estarreja:

Figure 2.3 – Network of chemical complex of Estarreja [3].

CUF-QI is the leader in terms of sales of ANL for the “open” market in Europe,

being the 4th producer. Currently, CUF-QI is one of the main non-integrated ANL

producers, with a quota of approximately 3% of the global production capacity, as

illustrated in Figure 2.4.

1 The only flows represented here are those in Estarrejawhere CUF participates (there are other entities and flows at the site)

LEGEND:

Key

Suppliers

Key

Customers

CUF Operations

in Estarreja

ORGANICS

INORGANICS

HCL

H2SO4

Hydrogen

Salt

Chlor., NaOH

HCL

Other Suppliers

(“Market”)

Other Customers

(“Market”)

• Aveiro Port

• SGPAMAG

NOVA AP

QUIMITÉCNICA

• Aveiro Port

• SGPAMAG

Aniline

Ammonia

Benzene

Aniline, MNB

Steam

Electri-city

Aniline, MNB, Nitric Acid, SulphanilicHypochlor.

Chlorine, NaOH, HCL

NaOH

MDI

DCP

AluminiumSalts

Hydrogen

Chlor., NaOH

HCL

Over-the-Fence

Inputs CUF Estarreja

Outputs CUF Estarreja

Other flows


12

Figure 2.4 – Main world Producers of ANL (2013) [3].

ANL produced at CUF-QI is mostly sold to DOW for MDI production. The process

begins in the plant of nitric acid, the 1st plant. Then the nitric acid is sent to the NB plant,

2nd plant, where it reacts with benzene (Bz). The NB formed goes to the 3rd plant, where it

is hydrogenated in the presence of a catalyst and ANL is formed, Figure 2.5.

Figure 2.5 – Schematic diagram of CUF-QI plant [3].

Global capacity share (%)

Integrated with MDI

Non-integrated with MDI

Nitric Acid

Plant

NH3

Nitrobenzene

Plant

Benzene

Nitrobenzene

Nitrobenzene

Aniline

PlantH2

Aniline

Sulphanilic Acid

Plant

Aniline

H2SO4

Sulphanilic Acid

Nitric Acid

Nitric Acid


13

2.2 Technological aspects of the Industrial Production of Aniline

2.2.1 Aniline

Aniline (C6H7N) when freshly distilled is a colorless, oily liquid with a

characteristic “fishly” amine-like odor. It is manufactured by gas and liquid phase

hydrogenation of NB using base or noble metal catalysts. If exposed to air and light, gains

a brown color. In industrial use, color formation can be minimized by storage and

processing under an inert atmosphere. The color might be removed by distillation just

prior to use in color-critical applications. It is miscible with a large number of organic

solvents, and forms soluble salts in the presence of strong acids in water. The main

properties of ANL are shown in Table 2.3.

Table 2.3 – ANL properties [1].

Property Value

Molar Mass (g/mol) 93.1

Boiling Point (ºC) 184

Flash Point (ºC) 70

Auto Ignition Temperature (ºC) 615

Densityliquid 20ºC (g/cm3) 1.02

Viscosity 20ºC (cP) 4.4

Solubility20ºC ANL in water 3.6 wt %

water in ANL 5.5 wt %

ANL is slightly corrosive to some types of metal, particularly amphoteric materials

such as aluminium, copper, tin, zinc, and alloys containing any of these metals. These

materials should be excluded from ANL service. For normal applications, carbon steel or

cast iron are satisfactory materials for ANL storage and handling. If product discoloration

must be kept to a minimum, then ANL should be stored and handled in 400-series

stainless steel equipment with proper nitrogen blanketing.

Typical aniline sale specifications are shown in Table 2.4. For some special

applications, the concentrations of trace impurities like CHA, cyclohexanol (CHOL),


14

cyclohexanone (CHONA), phenol, toluidine (TLD) and dicyclohexylamine (DICHA)

may be specified.

Table 2.4 – Typical ANL sales specification [1].

Color maximum (APHA) 100

Freezing point, minimumdry (ºC) -6.2

Purity, minimum 99.9%

NB content, maximum (ppm) 2

Water content, maximum 0.15 wt %

2.2.2 Reaction mechanisms for aniline production and by-products formed

Traditional method for ANL production involves multiple reactions, as it happens

in processes for preparing other aromatic amines. Typically, ANL is produced through the

Bz conversion into a derivative, such as NB, phenol or chlorobenzene, which is then

converted to ANL [4]. In the CUF – QI, SA unit, at Estarreja, the production is realized in

three different plants. In the 1st plant occurs the HNO3 formation, then in the 2nd plant,

takes place the Bz nitration with nitric acid in the presence of sulphuric acid (which is the

catalyst and dehydrating agent) to produce NB, equation 2.1:

𝐵𝑒𝑛𝑧𝑒𝑛𝑒 + 𝐻𝑁𝑂3 → 𝑁𝐵 + 𝐻2𝑂…….. (2.1)

In the 3th plant occurs the NB hydrogenation in the presence of a Ni catalyst, at mild

conditions (equation 2.2):

𝑁𝐵 + 3𝐻2 → 𝐴𝑛𝑖𝑙𝑖𝑛𝑒 + 2𝐻2𝑂……….. (2.2)

The reaction for ANL production is conducted at temperatures between 120-200ºC

and pressures between 10-20 bar, with yields higher than 99% [5]. The highly exothermic

catalytic hydrogenation of NB, with a heat of reaction of about 544 kJ/mol, is carried out

commercially in the presence of excess hydrogen in either the vapor or in the liquid phase

[6].


15

Commercially, ANL production is done through NB hydrogenation, however a

route involving phenol amination was previously used by Sunoco Chemical but is no

longer employed.

The catalytic hydrogenation of nitrobenzene is commonly employed as a standard

reference reaction for testing and comparing the activity of hydrogenation catalysts for a

range of applications, because its transformation is extremely easy and is carried out

under relatively mild conditions [7, 8].

Although there is a large volume of literature references available studying and

citing this reaction, there is not much information about the reaction mechanism. The first

explicative mechanism for ANL formation through NB hydrogenation was proposed by

Haber, in 1898 [9], and is widely accepted. The first step is a hydrogenolysis of N-O bond

giving a nitrosobenzene (NSB), followed by the formation of arylhydroxylamine (PHA) –

Figure 2.6.

Figure 2.6– Reaction network involved in nitrobenzene hydrogenation, Haber mechanism [9].

However, this mechanism does not fully explain all the experimental results,

although a number of studies had reported the identification of the suggested reaction

intermediates during hydrogenation [10 – 12]. Consequently, more studies were done, and

another mechanism was proposed by Wisniak and Klein [13] that is slightly more

complicated than the Haber’s mechanism. They also consider that probably the real

mechanism is even more complex and should consider the phenomena at the surface of

the catalyst (Figure 2.7).

Ar - NO2

nitro / NB

Ar - NO

nitroso / NSB

Ar - NHOH

arylhydroxylamine / PHA

Ar - NH2

ANL

Ar - NO = N - Ar

Ar - NHOH

Ar - N = N - Ar Ar - NH = NH - Ar

hydrazo / HZBazo / AZBazoxy / AZXB

Direct route

Condensation route


16

Figure 2.7 – Reaction network of nitrobenzene hydrogenation, as proposed by Wisniak and Klein

[13].

Beyond that, it was concluded that hydrogenation of nitro compounds and

disproportionation proceed on different sections of catalytic surface and so, a scheme for

the process was suggested, as shown in Figure 2.8.

Figure 2.8 – Scheme of components transformation on catalytic surface, proposed by Makaryan

[14].

In 2005, Gelder et al. [7] suggested that the number of steps involved in ANL

formation was higher and substantially different from those previously reported. Figure

2.9 shows the mechanism proposed by these authors, using a Pd on carbon catalyst.

Analyzing the results obtained, they concluded that NSB is not an intermediate in aniline

formation. They also concluded that the new understanding of the mechanism had

NO2 NH2

NB

NO

NSB

N

ON N N N

H

N

H

AZXB AZB HZB

PHA

NHOH

ANL

k4

k1

k1'

k5

k5'

k3'

k3k2

k6k8 k7

centre I centre II centre I

ArNO2

ArNHOHArNH2

ArNOM

H H

M

N

HH

Ar O

H

+

-

N

Ar

O

H

+

-M

H H

ArNHOH


17

implications for both catalyst and reactor design and that to obtain a high activity and

selectivity it is essential that the hydrogen flux at the surface is maintained at a constant

level, with good access to the reaction site and no diffusion limitations [7].

Figure 2.9 - Reaction network of nitrobenzene hydrogenation proposed by Gelder et al. [7].

Summarizing, when NB is hydrogenated to ANL, the reaction mechanism is

complex and there are some common intermediates, not depending on the mechanism

proposed, such as NSB, azoxybenzene (AZXB), azobenzene (AZB), PHA and

hydrazobenzene (HZB).

Corma et al. [15] reported that Au on TiO2 or FeO3 catalyzes the selective reduction

of a nitro group without the need to add metal salts and thus acts as a highly selective and

environmentally friendly catalyst. Some experiments were carried out over Au/TiO2

catalyst and it was verified that under the reaction conditions the NSB and hydroxylamine

compounds formed react before desorbing. That explanation was consistent with the fact

that the NSB and PHA derivatives were not detected in the reaction media. Their

proposed mechanism is shown in Figure 2.10.

Ph - NO2

Ph - NOH (a)

Ph - NO

+ Ph - NOH (a)

Ph - N(OH)H Ph - N(O) = N - Ph

Ph - N = N - Ph

Ph - NH - NH - Ph

Ph - NH2

Ph - NH

NB NSB

PHA

ANL

HZB

AZB

AZXB


18

Figure 2.10 – Proposed reaction pathway for the hydrogenation of aromatic nitro compound to

aniline [15].

Thus, on Au/TiO2 catalysts, PHA is formed as both a primary product (from NB)

and a secondary product (via NSB) at the active sites. The gradual accumulation of this

intermediate on the catalyst surface showed that the transformation of PHA into aniline is

the rate-determining step of the whole process. Makosch et al. [16] also evaluated the

influence of the support in the reaction route of NB hydrogenation, using Au/TiO2 and

Au/CeO2 catalysts. Both catalysts rapidly convert NSB but while over Au/TiO2

hydrogenation proceeds through the direct route, over Au/CeO2 proceeds through the

condensation route (Figure 2.6). For the condensation route to occur, a high surface NSB

is necessary. In the case of Au/TiO2, PHA is rapidly formed from NSB, accumulates on

the surface and is then transformed to ANL. With Au/CeO2, hydrogenation rate is

considerably lower and the conversion NB NSB is slower, which leads to an

accumulation of NSB and to the formation of condensation intermediates. These authors

concluded that the support has a direct impact on the reaction mechanism and actively

changes the reaction route.

Selective hydrogenation of NB over Ni/γ-Al2O3 was also studied, using different

media (dense phase carbon dioxide, ethanol, n-hexane) [17]; it was found that conversion

of NB was higher in CO2 than in ethanol and selectivity to ANL was almost 100%. This

might be explained by the interactions of dense phase CO2 with reacting species (NB,

NSB and PHA): NB reactivity is decreased while NSB is increased and the

NO2

+ H2

- H2O

NO

+ H2

NHOH

Fast Step + 2 H2

- H2O

- H2O

+ H2Slowest Step

NH2

NB NSB

ANL

PHA


19

transformation of PHA to ANL is likely promoted. Thus, hydrogenation of NB should

occur through the direct hydrogenation route (NB NSB PHA ANL), being NB

NSB the rate determining step. On the other hand, using a Pd based catalyst in

supercritical carbon dioxide (scCO2), Chatterjee et al. [18] concluded that the most

probable route to the ANL formation is (i) NB ANL (as no intermediate compounds

was detected even in short reaction times) and (ii) NB PHA ANL (through some

calculations of the initial rate of ANL formation from NSB and NB, NSB presented the

slowest rate of hydrogenation indicating that it could not be the possible intermediate

specie involved in the ANL formation).

In 2014, Rakitin et al. [19] investigated the catalytic hydrogenation of NB using Pd

catalysts in a scCO2 medium and proposed a scheme for the catalytic process, shown in

Figure 2.11, based on the analysis effectuated to the mixture of the hydrogenation

reaction either using scCO2 or isopropanol.

Figure 2.11 - Scheme of NB catalytic hydrogenation in the presence of Pd-containing

heterogeneous catalyst, [19].

Pd/C catalyst was used as well to study the hydrogenation of NB in methanol [20];

according to the experimental results obtained and mechanistic considerations, an

extended reaction scheme for NB hydrogenation to ANL was proposed, illustrated in

Figure 2.12.

NO2

NB

[Cat]

H2

NO

NSB PHA ANL

AZXB AZB

[Cat]

H2

NHOH

[Cat]

H2

NH2

[Cat] H2

NN

O

[Cat]

H2

NN[Cat]

H2

NH NH

[Cat] H2

HZB


20

Figure 2.12 – Supplemented reaction mechanism for NB hydrogenation considering Haber’s and

Gelder’s reaction mechanism, proposed by Turáková et al. [20].

Turáková et al. through the analysis of catalytic results and taking into account both

Figure 2.4 and Figure 2.8, concluded that the intermediate Ph-NOH condensates to

AZXB, that is going to react with hydrogen chemisorbed on the metal surface and

subsequently Ph-NOH and chemisorbed form of nitrene Ph-NH are formed. Ph-NH

chemisorbs near a chemisorbed hydrogen atom and then is desorbed from the surface as

ANL. Ph-NOH can again enter condensation reaction and therefore higher concentrations

of AZXB, in comparison to AZB, were measured and its temporary accumulation in the

reaction mixture was observed. Other way, PHA is not formed directly from NSB but is

formed through Ph-NOH. Authors did not observed accumulation of HZB and so ANL

formation via direct AZXB hydrogenolysis was hypothesized as the preferred reaction

NO2

NB NO

NOHNHOH

NH2

NN

O

NOH

NN

NH NOH

NH2

NH2

NH

+

NSB

PHA

ANL

AZXB

AZB

HZB

NH NH

ANL

ANL


21

path. In an industrial point of view, the authors stressed that at higher temperature, ANL

is formed via a condensation path (AZXB, AZB and HZB intermediates) rather than by a

direct route (NSB and PHA intermediates).

In the case of secondary products formation, there is not much information about

the issue. Nagata et al. [21] proposed a mechanism to explain the formation of some of

these compounds (Figure 2.13), for the reaction in vapor-phase in the presence of a

palladium or palladium/platinum catalyst. Although this mechanism includes N-

cyclohexylaniline (CHENO), does not include the formation of CHOL and DICHA that

are observed in experimental tests performed at CUF.

Figure 2.13 – Reaction network for the formation of ANL and secondary products proposed by

Nagata [21].

In 1995, Narayanan et al. [22] reported the ANL hydrogenation under vapor phase

conditions over nickel-alumina catalysts. The reaction products identified were

cyclohexane (CH), CHA, DICHA and N-phenylcyclohexylamine (NPCHA). They

observed that in the case of supported nickel catalysts, NPCHA is a major-product and

that depending upon the conditions of the experiment and metal content, other products

such as CHA and DICHA are also formed. In a latter article, Narayanan et al. [23]

NO2 NH2

H2

[ 1 ]

NH2 NH2NH2

NHNH2

NNH

O

H2

[ 2 ]

CHA

NH2

H2

[ 3 ]

[ 4 ]

N-phenylcyclohexylamine

[ 7 ] H2

CHANIL

- NH3

+ H2O

H2

[ 8 ]

[ 5 ]

CHONA

NH2

[ 6 ]

CHENO

NB ANLIMINE


22

compared the catalytic properties of Co/Al2O3 and Ni/Al2O3 materials, containing the

same metal content and prepared under similar conditions. It was concluded that on

increasing the contact time of ANL, the conversion increased and CHA and NPCHA were

the two major products. CHA formation was slightly favored at low contact time and

NPCHA was formed in roughly similar amounts at all contact times. They concluded that

there was hardly any selectivity difference with respect to feed rate. However, at high

contact time DICHA and CH were also formed. Based on these results they proposed the

next mechanism for the formation of the secondary products, shown in Figure 2.14.

Figure 2.14 - Reaction network proposed by Narayanan et al. [23].

In 2008, Relvas [24] proposed a mechanism, based on the Nagata mechanism

(Figure 2.13). For the elaboration of such mechanism, several laboratorial tests were

made, in which temperature, pressure, NB and catalyst concentration effects were studied.

DICHA and CHOL (through CHONA hydrogenation) were included in the proposed

mechanism, since they were detected in the experimental tests. This mechanism is

depicted in Figure 2.15.

NH2NH2

Co / Al2O3

Ni / Al2O3

+ H2

CHA

NH

CHANIL

+ NH3

- NH3

DICHA

+ CHA

+ H2

NH

ANL

ANL

CH


23

Figure 2.15 – Reaction network proposed for ANL and secondary products formation by Relvas

[24].

Nevertheless, as it was said by Relvas [24], this mechanism does not fully explain

the formation of all secondary products in the NB hydrogenation. So, further work in this

topic was still required.

Hydrogenation of NB and ANL, among other nitro compounds, was studied by

Králik et al. [25] over Pd, Pt and Ru catalysts. For ANL hydrogenation, it was observed

that CHA and DICHA are the main products and cyclohexane (CH) was also detected.

Moreover, for ANL conversion to CHA a total hydrogenation of the aromatic ring is

required, which is more complicated in the liquid phase. Therefore, a higher stability of a

single bond will need longer reaction time and consequently, a wide number of side

reactions are observed. So, for direct ANL hydrogenation, authors proposed the following

reaction network, shown in Figure 2.16.

Cat

+H2,-H20

Cat

H2

+ANL

-NH3

Cat

H2

+H2 -NH3

+H2

+H2

-NH3, -H2O

+ANL -H2O

+H2

+H2 CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

ANLNB

Amine

N-[1-(Amino)cyclohexyl]-N-phenylamine

N

O

OH

NO2 NH2

NH2

NHNH2

NH

NH


24

Figure 2.16 – Reaction network proposed for secondary products formation from ANL

hydrogenation by Králik et al. [25].

The authors [25] also investigated the production of ANL from NB, verifying that

the extension of side reactions increases with temperature, acidic-basic features of

reaction environment and properties of the support. NB hydrogenation reaction route was

observed to be via formation of NSB and hydroxylamine intermediates, being possible

those intermediates undergo side reactions. ANL formed during reaction can react further

and generate other compounds such as cyclohexenyl amine which rearranges to more

stable imine. The presence of H2O allows formation of CHONA, Figure 2.17. A novelty

in this proposal is the detection of benzene (Bz) formation, via ANL deamination. In the

NH2

ANL

NH2

+ - NH3 NH

+ H2

+ 2 H2

NH N

NNH

+ H2 + H2

NH2

+

NH2

- NH3

NH2

NH2

+ - NH3

+ H2

+ H2

+ H2

NH2

NH

+ H2

NH

+

NH2

- NH3

+ H2

NH

+ H2

+ H2

+ H2

NH2

+

- NH3

NH2

-

NH

N

N

+ H2

[hydrogenolytic products] [hydrogenolytic alkylation products]

+ n H2

ANL

ANL

CHENO

DICHA

CH


25

determination of reaction conditions for hydrogenation of NB to ANL energetic aspects

are important. For maximum industrial exploitation of the reaction heat released, higher

temperatures are preferred, however that will implicate a careful analysis of catalyst

selectivity and lifetime.

Figure 2.17 – Reaction network proposed in liquid phase hydrogenation of NB by Králik et al.

[25]

Commercial catalyst, Pd supported on carbon, was used by Rubio-Marqués et al.

[26] with the purpose of studying the kinetics of NB hydrogenation. Results showed that

ANL is rapidly formed as a primary and unstable product that, by partial hydrogenation,

gives CHANIL as a secondary product. By further hydrogenation generates DICHA in

minor amounts, as a tertiary product. Diazo products were not detected. Nevertheless,

CHA was not observed among the reaction products, and since the hydrogenation of ANL

is very likely to occur, a possible way to form CHANIL would be through the reaction

between ANL and CHA. More catalytic tests were performed to confirm this theory and a

NO2

NH2

+ 3 H 2

+ 2 H 2

- NH3

OH

+ H2O - NH3

NH2

+ 2 H 2

NH

NH2

O OH

N

N

H

NH NH

+ H2

- NH3

NH2

+- NH3

NH2

+

+ H2O

- NH3

+ H2

+ H2

+ 3 H 2

High molecular weight compounds

NB

ANL

Bz

CHA

CHONA CHOL

CHENO

CHANIL DICHA

Phenol


26

mechanism was proposed, which is consistent with the kinetic experimental results,

Figure 2.18.

Figure 2.18 – Reaction network proposed for the Pd/C catalyzed hydrogenation of NB by Rubio-

Marqués et al. [26]

In 2015, Sousa [27] reported the formation of secondary products in the NB

hydrogenation to ANL during transient operation, over a Ni/SiO2 catalyst, through the

addition of secondary products in the feed stream. This addition had the purpose of

observing the interactions between all the species present in the system. Some

preliminary studies were carried out at room temperature and it was observed that

CHONA reacts reversibly with ANL to form CHENO and H2O. CHONA was also added

to a pure ANL sample and CHENO formation was immediately detected. Furthermore, it

was observed that H2O creates two different scenarios. When H2O is not present,

formation of CHA, CHANIL, CHENO and DICHA is favored while no significant

amounts of CHONA and CHOL are detected. On the other hand, when H2O is present,

CHONA and CHOL increase considerably but the amount of all the other compounds is

reduced. Proposed mechanism is shown in Figure 2.19.

NO2 NH2

+ 2 H 2

- 2 H2O

+ 3 H 2

NH2NH

NHNH2

- NH3

N+ H2

- H2

ACID CATALYSIS

NHNH+ 3 H 2

NB ANL CHA

CHENOCHANILDICHA


27

Figure 2.19 – Reaction network for the formation of secondary products during NB

hydrogenation in the presence of Ni supported catalyst, proposed by Sousa [27].

According to Sousa [27], H2O phenomenon explains why Narayanan did not

observe any CHONA or CHOL. Small amounts of Bz were also detected in experiments

performed, suggesting that hydrogenolysis reactions occur in the reactor, by withdrawing

the nitro (from NB) or the amine (from ANL) group. However, CH was not detected

which might mean that there are no hydrogenolysis reactions with molecules with

saturated carbon cycles, such as CHA, CHOL, CHONA and DICHA.

Knowledge of all these mechanisms, either of intermediaries or secondary

compounds, in NB and ANL hydrogenation are quite relevant, since the most important

manufacturing processes for ANL are based on the continuous catalytic hydrogenation of

nitro compounds. In large scale processes heterogeneous catalysis are employed, whereas

for smaller scale homogeneous catalysis is achieving more importance [28].

Industrially, NB hydrogenation processes can be done in liquid-phase or in vapor-

phase. A comparison between those two processes shows little difference in yield and

product quality. The liquid-phase process saves energy by eliminating gas recycle and

offers a higher space-time yield in the reactor. The vapor phase process has other

NO2 NH2

- 2 H2O

+ 3 H2

N

CHENO

NBANL

CHANIL

NH

NH2

- NH3

NH2 CHA

+ H2

+ 3 H2

DICHA

NH

NH2

- NH3

+

+

NH2 + H2O

- NH3-

+ 3 H2

O OH

+ H2

CHONA CHOL


28

advantages: very effective utilization of reaction heat in order to generate pressure steam,

elimination of catalyst separation from the product, and extended catalyst life [1].

In the next sections it will be presented the vapor-phase and liquid-phase processes

as well as the companies where each of these processes is used.

2.2.3 Hydrogenation in Gas-phase

In gas phase aniline production processes two types of reactors can be used, fixed

bed (Bayer) or fluidized bed reactors (BASF). Generally, a catalyst constituted of a non-

noble metal supported on a suitable carrier is used for high reaction temperatures (gas-

phase processes) [1].

In Figure 2.20 is presented a process flow scheme for a gas phase process in a

fluidized-bed reactor.

Figure 2.20 – Fluidized-bed ANL process in vapour-phase [1].

NB and hydrogen are charged to the NB vaporizer and then the gas-mixture (10 to 1

molar ratio of hydrogen to NB) leaves the vaporizer as a superheated vapour to prevent

any condensation of NB and is charged to the bottom of the reactor, entering into the

fluidized bed of catalyst. Reaction temperature and pressure are controlled at about 270


29

ºC and 3.4 to 3.7 bar, respectively. Contact time in the catalyst bed is of several seconds.

Essentially complete NB conversion is attained at selectivity to ANL of 98 %. The heat of

reaction is removed via a heat transfer fluid circulating through tubing networks located

at key points in the reactor. The catalyst is slowly deactivated and so it is necessary to

regenerate it about every 6 months. Catalyst life in commercial plants is believed to be up

to five years.

The ANL and water in the gaseous effluent of the reactor are then condensed in the

product gas condenser (Figure 2.20). Hydrogen is removed and is then compressed and

recycled to the reaction system. The water/ANL stream leaving the product gas condenser

is a two-phase system that forms a minimum boiling heteroazeotrope at about 98 ºC with

a concentration of 78.6 wt.% of water. The mixture is first gravity separated in the

ANL/water separator. The water-rich phase, containing 2-3 wt.% of ANL, is fed to the

water removal column where water is removed as the bottom stream. The overhead

stream, consisting of the azeotropic composition, is then recycled to the separator. The

recovery column bottoms contain mostly ANL, plus small amounts of residue and heavy

ends. ANL product is taken overhead from the refining column, which is operated under

vacuum to retard the formation of degradation products, and sent to storage.

Companies referred in Table 2.5 are the most important ones for processes in vapor-

phase. However, other companies have also made an important contribution in this area,

such as Chemopetrol [36].


30

Table 2.5 – Main ANL Vapour-phase Processes.

Company Reactor

Type/Process

Experimental

Conditions Observations

BASF

[1, 29 - 32]

Fluidized-bed

reactor

Temperature:

300 ºC

Pressure:

4 to 10 bar

-Reaction heat is recovered for steam

generation

-US Patent 2011/8044244,

improvements in the reactor in order to

avoid poor mass transfer

Bayer

[1, 33 – 35] Fixed-bed, Adiabatic

Temperature:

250 - 350 ºC

Pressure:

1 to 7 bar

-Cost conversions savings of 25 %

-High purity product

-Maximum outlet temperature is 460

ºC, providing for heat recovery by

generation of high-pressure steam

Borsodchem/

Yantai Wanhua

[1]

Non-isothermic,

water cooled Tubular

reactor connected in

series with an

Adiabatic reactor

Sinopec

[1]

Two-stage fluidized

bed

Temperature:

240 - 300 ºC

-Includes a novel distributor that

avoids the temperature spikes which

would normally lead to by-product and

coke formation

2.2.4 Hydrogenation in Liquid-phase

Hydrogenation in liquid-phase can be done in a slurry reactor or in a trickle-bed

reactor. At CUF-QI, hydrogenation of NB to ANL is realized in liquid-phase using a

slurry reactor. This process, in liquid-phase, is also operated by other companies such as

DuPont, Huntsman and Mitsui.

2.2.4.1 DuPont Process

Information presented in this section is mostly based in some available references

[1, 37, 38].

DuPont has a long history with this type of technology, which has started in 1940

with a batch process. In 1958 the company developed a proprietary aniline catalyst and

then moved to a continuous process. Successive improvements were made to the catalyst

during the 1970s.

The liquid-phase NB hydrogenation can be represented in a simple schematic flow-

sheet as shown in Figure 2.21.


31

Figure 2.21 – DuPont ANL Process via liquid-phase [1].

Today, the DuPont process feeds NB (produced by dehydrating nitration process)

together with hydrogen into a liquid-phase, plug-flow hydrogenation reactor that contains

the Pt-Pd supported catalyst with an iron modifier. The catalyst selectivity and NB

conversion per pass is 100 %. The reaction conditions are optimized to achieve essentially

quantitative yields, and the reactor effluent is free of NB. Excess hydrogen from the

reactor effluent is vented, and the reactor liquid product is sent to a dehydration column to

remove water, followed by a purification column to produce high quality aniline product.

The liquid-phase hydrogenation catalytic system is very simple, efficient, compact

and robust, avoiding the complexity of a catalyst regeneration step, typical of vapour-

phase, fluidized-bed technologies.

DuPont claims that aniline produced from this technology maintains constant

product purity without experiencing a drop-off in product quality as the catalyst ages.

2.4.2.2 Huntsman Process


[1, 39].

Nitrobenzene

H2

Vent

Fuel Gas

Reaction Water

Aniline

Reaction Water

Aniline Purification

Hydrogenation

Reactor

Dehydrating

Nitration


32

Huntsman inherited ANL production technology with its acquisition of ICI

polyurethanes in 1999. ICI patents disclosed a process for a continuous liquid-phase ANL

production process in which the catalyst is suspended in a 95 wt. % solution of ANL. In

Figure 2.22 is presented the Huntsman scheme process.

Figure 2.22 – Huntsman ANL Process via liquid-phase [1].

NB and hydrogen are introduced into the continuous slurry reactor and the

hydrogenation is carried out at 165 ºC. Process parameters are controlled to ensure that

the liquid-phase concentration of ANL remains > 95 wt %. The hydrogenation reactor

contains an agitator for dispersing the hydrogen gas. Heat from the strong exothermic

reaction is controlled partly by allowing evaporation of water and ANL in the reaction

mixture as well as using cooling coils and jackets. A proportion of the ANL contained in

the condensed effluent is returned to the reaction to ensure steady state conditions are

maintained.


33

2.4.2.3 Mitsui Process


[21, 23, 40].

Mitsui is believed to have used a liquid-phase process. In their US Patents

1994/5,283,365 and 1997/5616806, it is claimed a process for the preparation of aniline

by hydrogenating NB in an ANL solvent in the presence of a catalyst. The temperature

used is from 150 to 250 ºC, the reaction occurs substantially in the absence of water, with

aniline and water produced in the reaction being continuously distilled off. NB is

maintained at 0.01 wt % or less in the reaction solution. High purity ANL is produced,

with CHOL, CHONA and CHENO present at 10, 20 and 20 ppm, respectively.

The process is described in the patents referred above and the objective is to prepare

high-purity ANL containing low level of impurities. NB, catalyst suspended in the solvent

(ANL) and hydrogen gas, are continuously fed to a stirred reactor. Afterward, the vapour

exiting the reactor is condensed, and the water is separated from ANL and then removed

from the system. Part of the ANL can be returned to the reactor so that the volume of the

solution in the reactor may be kept at substantially constant level during the reaction.

Using this process, a high purity ANL > 99.9 wt.% can be produced, containing traces of

CHOL (<10 ppm), CHONA (<50 ppm), CHANIL (<20 ppm) and MNB (<5 ppm).

2.4.2.4 Chematur Process


[1].

Chematur Engineering is marketed as offering a high product purity, safe handling

of high pressure hydrogen, and improved energy efficiency through improved heat

recovery.


34

Figure 2.23 – Chematur ANL Process [1].

The process flow sheet shown in Figure 2.23 presents the liquid-phase process of

Chematur. The NB is preheated and fed to a well stirred hydrogenator for conversion into

ANL. Catalyst slurry is also fed to the reactor. Hydrogen is efficiently dispersed in the

liquid and reacts with NB to form ANL and water. The reaction occurs in a single pass of

hydrogen. The mixture ANL/catalyst/water is circulated through a slurry thickener where

the crude ANL is removed as filtrate. The heat of reaction is removed by a closed loop

water cooling system. The heated cooling water flashes at two different pressure levels to

produce steam.

The crude ANL and water is separated in a gravimetric decanter, after being cooled

down in order to allow the separation. The wet crude ANL is transferred to storage for

refining.

2.4.2.5 CUF-QI Process

In 1978, the chemical company owned by CUF-QI started with the production of

ANL through NB hydrogenation, in liquid phase, using the Tolochimie process. Initially,


35

production capacity was 50.000 ton/year. However, since then two major plant revamping

were carried out and nowadays the capacity is 200.000 ton/year.

In Figure 2.24 is presented a simplified scheme of CUF-QI process.

Figure 2.24 – CUF-QI ANL process.

NB, hydrogen, recycled ANL and suspended catalyst are fed to a well stirred

reactor, where NB hydrogenation occurs. After, the reaction mixture is fed to a decanter

with the purpose of separating the catalyst from the reaction products. The decanter top

stream, rich in ANL, is supplied to another unit for the separation between the aqueous

phase (H2O with solubilized ANL) and the organic phase (ANL with solubilized water

and secondary products). The aqueous phase goes to water treatment processes and the

organic phase is then divided in two streams, one goes to an ANL purification unit and

the other stream is recycled to the reactor to act as a solvent.

2.4.2.6 Bechamp Process


[1].

This process uses iron metal and iron (II) chloride as reducing agents:

4𝐶6𝐻5𝑁𝑂2 + 4𝐻2𝑂 + 9𝐹𝑒𝐹𝑒𝐶𝑙2,𝐻𝐶𝑙→ 4𝐶6𝐻5𝑁𝐻2 + 3𝐹𝑒3𝑂4..................(2.3)

ANL+Catalyst

MNB

H2

H2O+ANLresidual

H2O+ANLresidual ANLpure

Heavy ends

Reactor

Decanter

Separator

Destillation

Column

Destillation

Column

ANL

Tank

Tank


36

Bechamp discovered that nitro compounds could be reduced by iron and acetic acid.

In 1857, Perkin applied this reaction to the manufacture of aniline [1]. Improvements

were made first by substituting hydrochloric by acetic acid. Subsequently, it was

discovered that the ferrous salt of the acid has catalytic activity, so that the reduction

could be effected with less than the theoretical quantity of acid. This process co-produces

iron oxide colored pigment. Bayer still uses this process at its plant in West Virginia, US.

NB and water reactants are charged to an agitated vessel along with iron (II) choride

solution and ground iron filings. Typically, the process is operated batch-wise, with

sequential additions of iron and NB in order to control temperature and pressure build-up.

The vessel is heated to 100 ºC for 2 h after all iron and NB has been added. Total batch

wise is 8 to 10 h.

In Table 2.6 are summarized the most important processes for ANL production in

liquid-phase. Besides the NB hydrogenation vapour and the liquid-phase processes, ANL

can be also produced by Ph amination, as it was said before, in a fixed-bed reactor in the

presence of silica-alumina catalyst at about 370 ºC and 17 bar. Some research was also

done in order to produce ANL from biomass or from coal [1].

Table 2.6 – Summary of ANL liquid-phase processes.

Company Reactor

Type/Process

Experimental

Conditions Observations

DuPont Plug-flow reactor Temperature: 100-300ºC

Pressure: 4.4-21.7bar

-Catalyst inventory is low and a

continuous bleed of catalyst is taken to

metals recovery and reprocessing

Huntsman Slurry reactor Temperature: 165-170ºC

Pressure: < 5bar.

-Steam generated by the exotherm

reaction is sold, recycled and used on

the MNB plant

Mitsui Slurry reactor Temperature: 150-250ºC

Pressure: 1.5 -10bar

-Inhibition of side reactions and the

formation of by products

Chematur Slurry reactor -Steam generation

CUF-QI Slurry reactor Temperature: 80-250ºC

Pressure: 10 – 60 bar

Bayer Agitated vessel Temperature: 100ºC

Batch time:8-10h

-ANL is a byproduct of colored iron

oxide pigment production.

Other companies/researchers published some patents proposing processes for the

production of aromatic amines in liquid-phase, such as Solvay Process Company [42],

Rhodia Chimie [43], Alberts et al. [44], or HRD Corporation [45].


37

2.2.5 Catalysts for Aniline production

The use of catalysts allows not only to increase the rate of chemical transformations

but also to selectively orientate them towards the formation of the desired product(s).

Thus, catalytic processes are more efficient in terms of profitability and also from an

environmental point of view (less by products). Industrial catalysts can be divided in two

groups: bulky catalysts and supported catalysts. The catalysts are composed by active

substances that include catalytic agents, metals and metallic oxides and also promoters,

species which dramatically increase the activity or selectivity of the catalyst.

Hydrogenation catalysts can be of two types: heterogeneous or homogeneous.

Heterogeneous catalysts for this application are usually constituted by metals in their

fundamental state. Their formulation depends on their utilization: those that are used in

fixed-bed reactors and those that are used in slurry or fluidized-bed reactors.

Homogeneous catalysts are very attractive for small-scale and synthetic applications

involving multifunctional starting materials. The metal, the ligand, the solvent and the

reaction conditions are variables that have great influence in the homogeneous catalysis

[28].

For the hydrogenation of nitro groups, the most used metals are palladium, platinum

and nickel, supported or not. In the case of NB hydrogenation there is a wider choice for

metals, including copper, cobalt, palladium, platinum, and nickel, depending on the

compound to be produced. The catalyst selection is very important for maximizing ANL

production but the knowledge of mass transfer problems is also very critical.

The catalytic hydrogenation of nitrobenzene to aniline is a complex chemical

process. A number of competing mass transfer and kinetic rate processes contribute to the

overall observed reaction rate. In Figure 2.25 are shown typical concentration profiles

during NB hydrogenation, which are the result of simultaneous and competing

phenomena: the rate of hydrogen mass transfer from the gas to the liquid phase; the rate

of hydrogen and NB mass transfer from the bulk liquid phase to the outer surface of the

catalyst; the rate of hydrogen and NB mass transfer inside the porous catalyst; and the

adsorption and kinetic rates of the hydrogen and NB on the inner catalytic surface of the

catalyst particle [4]. In addition, desorption and diffusion of the products in the opposite

direction should be also taken into account.


38

Figure 2.25 – Typical concentration profiles during hydrogenation of NB [4].

For ANL production in vapour-phase at industrial scale, copper and palladium in

combination with other metals (as modifiers/promoters), on activated carbon or on an

oxidic support, have been shown to be effective catalysts leading to high activities and

selectivities. Nevertheless, catalysts need to have adequate attrition resistance [1].

In the case of liquid-phase hydrogenation, palladium and nickel catalysts have great

success. DuPont’s liquid-phase process uses platinum-palladium catalyst on a carbon

support, with iron as a modifier. The modifier provides a good catalyst, with high activity

and protection against hydrogenation of the aromatic ring. For the Huntsman process, one

preferred catalyst is finely divided nickel on diatomaceous earth, known as kieselgur [1].

Mitsui claims the use of a catalyst of palladium or palladium-platinum deposited on a

lipophilic carbon support [40]. At CUF-QI is used a commercial solid catalyst, Ni

supported on SiO2.

In Table 2.7 are summarized the most successful industrial applications, according

to Králik et al. [41].


39

Table 2.7 – Industrial ANL applications [41].

Phase Catalyst Company

Gas Cu – support BASF

Gas Cu – support First Chemicals

Gas Cu – support BC – MCHZ

Gas Pd – alumina Bayer

Gas Cu – Cr2O3 Sumitomo

Gas Ni - support NIOPIK Russia

Liquid Ni - support Tolo Chimie

Liquid Ni - support ICI

Liquid Pd – Pt – Fe Du Pont

Many studies have been made and patents published concerning the best catalyst to

use in this reaction. The first process used for the ANL production was the Bechamp

process and this method had the disadvantage of requiring large quantities of iron and

also the difficulty of separating the aromatic amine from the large amounts of water and

iron sludge formed. Other reducing agents, such as ferrous salts, tin, zinc, soluble

sulfides, sulfur, and carbon monoxide were proposed. The hydrogenation of NB using a

homogeneous catalyst in combination with water-gas shift reaction (WGSR) was also

studied, being the catalyst carbonyl complexes of iron, ruthenium and rhodium, employed

in strongly basic solutions [28]. In vapor-phase processes it was proposed to use catalytic

materials such as nickel, iron, copper, silver, platinum and zeolites.

In 1942, Kise [42] proposed an improved process for the catalytic hydrogenation in

liquid-phase of aromatic nitro compounds, and in which a high activity of the catalyst was

maintained when the process was carried out under conditions such that the finely divided

catalyst body was maintained uniformly in suspension. They suggested that the catalyst

comprised Ni, Co or Cu, supported on a finely divided carrier such as kieselguhr,

asbestos, pumice or other inert material.

Even other catalytic technologies where studied, for instance, Sheng et al. [46] used

Cu/CuxO and Pt nanoparticles supported on multi-walled carbon nanotubes as

electrocatalysts for the reduction of NB. It was verified that Cu/CuxO has a much better

activity than Pt nanoparticles. Moreover, 44 % of conversion was achieved after 52h of

reaction in Cu electrocatalyst being a promising candidate for the development of a fuel

cell that will be able to generate electricity and products from NB hydrogenation.


40

2.2.5.1 Catalysts for vapor-phase processes

For vapor-phase processes, it was found that under optimum conditions, Cu

catalysts were stable and produced nearly theoretical yields of ANL from MNB. Gharda

and Sliepcevich [47] tested several Cu-based catalysts (CuCrO2, CuCrO2 + CuO

supported in silica-gel and CuO supported on alumina). It was observed that the catalysts

activity increased with temperature; however, the stability decreased and it was verified a

catalyst deactivation due to the presence of dinitrobenzenes. Also Rihani et al. [48] tested

Cu catalysts supported on activated charcoal impregnated with various metallic chlorides,

activated bauxite and Ag-impregnated activated bauxite and Cu, Ni and Cr impregnated

on inert carriers (and mixtures thereof); they concluded that the catalyst consisting of

Cu+Ni+Cr (and mixtures thereof) on an inert carrier give encouraging results. It was also

concluded that Cu-Ni promoted with Cd supported on asbestos presents the best results,

with a conversion of 79 %.

In 1981, Birkenstock et al. [49] experienced multi-component supported catalysts

containing a noble metal and a transition metal. These catalysts were composed by

Pd+V+Pb, Pd+Mo+Pb, Pd+Cr or Pd+V+Zn supported on α-Alumina and the yield of

ANL was always higher than 99.8 %. Some years later, Immel et al. [50] tried palladium

catalysts supported on graphite or petrol coke combined with Ir and Rh. Before reaction,

the catalyst must be activated and the hydrogenation was carried out at 350 - 420 ºC and

normal pressure. The results obtained in terms of ANL yield were higher than 99 % for

all the catalysts.

Narayanan et al. [22] prepared Ni/Al2O3 catalysts with different weight percentages

of Ni and tested the materials in the vapour phase ANL hydrogenation. The reaction

temperature was varied between 200 and 350 ºC. ANL conversion generally increased

with Ni content and the product selectivity was influenced by the degree of conversion. In

2007, Sangeetha et al. [51] studied a Pd catalyst supported on Mg-Al oxide hydrotalcite,

with different loadings of Pd, for MNB hydrogenation at atmospheric pressure and 225 -

300 ºC. It was concluded that the effect of dispersion and particle size of Pd supported on

hydrotalcite is significant in the hydrogenation activity of these catalysts in the

conversion of NB to ANL, and the catalyst with less Pd loading presented higher

selectivity and activity. However, it was detected that the formation of water in the

reaction could be a poison for the catalyst. The effect of the support was also studied,


41

using three types of support: hydrotalcite, MgO and γ-Al2O3. As the temperature was

increased the rate of NB was found to decrease in all the catalysts, probably due to the

coke formation or due to water poisoning. Nevertheless, the product selectivity towards

ANL was around 98 % for all materials. Hydrotalcite-supported Pd catalyst was found to

be more active than MgO and γ-Al2O3-based materials [52].

Amorphous carbon supported Ni catalysts were tested in gas-phase hydrogenation

of NB, at 250 ºC and atmospheric pressure, generating ANL as a sole product.

Hydrogenation activity was found to be insensitive to Ni particle size and exhibited a

proportional increase with increasing surface activity [53]. In 2012, Mohan et al. [54]

studied the advantages of Ni/SBA-15 over Ni/MgO catalyst (experimental conditions:

225 - 300 ºC and atmospheric pressure). Due to the exothermic nature of NB

hydrogenation reaction, it was verified a decline in the activity for both catalysts at higher

temperatures (above 250 ºC). However, the selectivity to ANL remained at 100% on both

catalysts. Ni/SBA-15 yielded high and steady activity unlike Ni/MgO.

More recently, Sudhakar et al. [[55] carried out the catalytic hydrogenation of NB

over different metals (Ru, Ni, Pt, Pd) supported on hydroxyapatite, HAP (as porous solid

possesses both acidic and/or basic sites in its surface). It was verified that among the

various metals supported on HAP, Pd displays good to excellent yields towards ANL

under mild conditions, without the formation of intermediate products unlike Ru, Pt and

Ni. Varkolu [56], in 2015, tested a Ni/TiO2 catalyst and found that high Ni dispersion,

small Ni particles along with strong metal-support interactions leads to NB conversions to

ANL higher than 99%. On the other hand, it was verified a decrease in conversion during

time on stream analysis that was ascribed to the condensation of reaction intermediates on

the catalyst surface, which means that more efforts should be carried out in order to get

efficient catalytic systems for a consistent catalytic performance for longer lifetime.

The choice of the best solvent to be used in the NB hydrogenation is also very

important since its use may play a crucial role in the stabilization of reactive

intermediates and has a decisive influence on chemical reactions. Therefore, supercritical

carbon dioxide (scCO2) has been considered since hydrogen is completely miscible with

it which might enhance hydrogenation reaction rate. The most used catalysts in these

conditions have been transition metals (Pt, Pd, Ru and Rh) supported on carbon (C) and

Al2O3. In 2004, Zhao et al. [57] reported that in their work all the catalysts showed 100 %


42

selectivity towards ANL in scCO2 at 20 ºC, without detection of intermediate products.

Moreover, both conversion and selectivity to ANL were higher in scCO2 than in ethanol

and the order of activity was Pt/C > Pd/C > Rh/C, concluding that scCO2 is a suitable

replacement for organic solvent in this reaction, being easy to separate the organic,

aqueous and gas phase. Relatively to the supports, it was found that conversion is higher

in C than in Al2O3. Using a Pd/MCM-41 catalyst, Chatterjee et al. [18] also concluded

that the use of scCO2 as a reaction solvent could have a significant advantage toward the

efficiency of the reaction. In fact, it was demonstrated that it enhances the catalytic

activity substantially by lowering the mass transfer limitations between the solid and gas

phase. Besides, the reaction was free from any intermediate product formation and long-

term catalytic activity was achieved. Several Pd-containing catalysts were tested [19] in

scCO2 and isopropanol media at 90 ºC and it was found that the formation rate of ANL in

the scCO2 medium was 3.5 to 5 times higher than in the other solvent and the selectivity

was between 92 to 95%. Effect of the support on activity was also analyzed and it seems

that it decreases in the presence of scCO2, indicating that accessibility to Pd-centers was

enhanced.

In 2009, Meng et al. [17] decided to study a Ni/γ-Al2O3 catalyst in dense phase CO2

and noted that this multiphase reaction system enables the selective hydrogenation of NB

to ANL under mild conditions. Selectivity to ANL was almost 100 % at any conversion

level ranging from 0 % to 100 % and the reactivity of NB was decreased.

2.2.5.2 Catalysts for liquid-phase processes

In liquid-phase hydrogenation, Turek et al. [58] studied a Ni/Al2O3 catalyst, with a

reaction temperature of 50 - 150 ºC and a pressure of 2 - 20 bar. They concluded that the

catalyst activity decreases with time of operation and the ANL yield depends on the

reaction temperature, decreasing with the increase of the temperature.

In 1995, Peureux et al. [59] selected a Pt/(γ-Al2O3+α-Al2O3) catalytic membrane to

perform the hydrogenation of NB, at 25 - 60 ºC and gas pressures up to 5 bar. The

catalytic membrane appeared to be very stable without showing significant changes in the

catalytic performance; however, its behaviour strongly depended on the mode of feeding

the reactants. A catalyst constituted of polymer anchored Pd(II) complexes was also


43

studied for the reaction, at atmospheric pressure and 30 - 45 ºC, using methanol as

solvent, which was found to be active for the NB hydrogenation [60]. Li et al. [61]

investigated the catalytic performance of Pt supported on carbon nanotubes (1 wt. % and

3 wt. %) under mild conditions, 25 ºC and 1 bar, concluding that these catalysts presented

extraordinary activity for the reaction, even at those experimental conditions. NB was

directly hydrogenated to ANL, reaching a yield of 90 % after 40 min (3 wt. % Pt).

Industrial Ni/SiO2 supported catalyst was studied by Relvas et al. [62] at 50 - 250

ºC and 5 – 50 bar. In all the experiments realized, the selectivity in ANL was higher than

99 %, and it was also observed that the presence of large excess of NB may cause some

kind of inhibition. Nickel nanoparticles stabilized by filamentous carbon were also tested

in the NB hydrogenation, at 150 ºC and 15 bar, and exhibited excellent performance in

the reaction, producing clean ANL (~ 99 % yield) without accumulation of side products

[63].

Pd on carbon are the most widely used hydrogenation catalysts both in research

laboratories of academia and by the chemical industry. Main advantages of using Pd

based catalysts are related with their activity, as Pd is the most active metal for

hydrogenolysis reactions and for the saturation of double bonds in conjugation with an

aromatic ring. At low temperatures Pd is inactive for the hydrogenation of most aromatic

rings but its catalytic behavior can be tuned by adding organic or inorganic modifiers

[64]. Numerous studies have been performed with the purpose of developing the best Pd

catalyst for this reaction. One of the key points is centered in the support, for instance, El-

Hout et al. [65] compared Pd nanoparticles supported on reduced graphene oxide

(Pd/RGO) with a commercial Pd/C and concluded that Pd/RGO presents a greater

catalytic activity because Pd/C has an uneven size distribution of Pd nanoparticles and a

significant amount of Pd agglomeration on the surface of the support. Pd/RGO could be

reused for two times without any lose in the activity.

Pd particles size is extremely important having a direct influence in the catalytic

activity. Over carbon nanofibers, when the size of Pd particles increase from 1.5 nm to 30

nm, catalytic activity decreases by 2.5 orders of magnitude, which might be related with a

substantial increase in the fraction of surface Pd atoms [66]. On the other hand, different

results were achieved for Pd supported on carbon nanofiber coated monoliths, i.e., NB

conversion depended just on the Pd dispersion. In fact, a good linear relationship between


44

the catalytic rate and the Pd dispersion was determined, indicating that activity is mainly

controlled by the dispersion but does not depend on the size of active metal or on the

crystalline planes of it [67]. Turáková et al. [68] studied the influence of Pd content on

carbon support in the catalytic activity and concluded that with lower loadings (1 – 4

wt.%), Pd particles are uniformly and separately dispersed on the catalyst surface, only

increasing their surface density. For higher Pd contents (> 4 wt.%), a second layer of Pd

appears on the surface, that causes a decrease in accessibility of previously existing

catalyst centers. This decrease in accessibility of catalytic centers leads to a drop in

reaction rate. According to the authors, for technological praxis, catalysts with higher

metal content than 4 wt. % should not be used.

Other catalysts were studied either for vapour or liquid-phase processes: the

reduction of nitroaromatics to amine with ruthenium catalysts, reducing 80 to 95 % of

mononitroaromatic to the corresponding amine [69]; the NB hydrogenation coupled with

ethylbenzene dehydrogenation using activated carbons [70]; the selective hydrogenation

of nitro aromatics with Pt nanocatalyst under ambient pressure [71]; the catalytic

hydrogenation of NB with date pit active carbons as Pd supports in [72]; the NB

hydrogenation with supported Pt catalysts in supercritical carbon dioxide [73].

2.2.6 Types of reactors

In terms of reactors, it is possible to classify them taking into account the following

characteristics:

- Mode of operation: batch or continuous (or even semi-continuous);

- Phases present: homogeneous or heterogeneous;

- Reactor geometry, flow pattern and manner of contacting the phases:

(i) Stirred tank reactor (slurry)

(ii) Tubular reactor

(iii) Packed bed, fixed and moving bed

(iv) Fluidised bed

Homogeneous gas phase reactors will always be operated continuously; whereas

liquid phase reactors may be batch or continuous (or even semi-continuous). Tubular

(pipe-line) reactors are normally used for homogeneous gas-phase reactions. Both tubular


45

and stirred tank reactors are used for homogeneous liquid-phase reactions. In a

heterogeneous reaction two or more phases exist, and the overriding problem in the

reactor design is to promote mass transfer between the phases.

Most common reactors used in hydrogenation reactions are the slurry reactors and

the fixed-bed reactors. The majority of the hydrogenation processes are performed in

slurry reactors, however productive processes at large scale can also use fixed-bed

reactors. The objective of hydrogenation reactors is to promote the contact between the

hydrogen, the catalyst and the compound to be hydrogenated. This type of equipment can

be very different, from micro-reactors up to huge vessels, designed to work from

atmospheric pressure to hundreds of bars.

For elevated production of intermediates, such as ANL, there are advantages in the

production process being in continuous mode. In this type of processes there is also the

possibility of working in the vapor or in the liquid phase, as described above.

As stated before regarding the hydrogenation processes that are available

industrially, in vapor phase the principal reactor configurations used are the fixed bed and

the fluidized bed, whereas for liquid phase hydrogenation slurry reactors are more often

employed. However, fixed bed has not experienced an extensive industrial-scale

development, because the major difficulty with this configuration lies in the removal of

the heat evolved during the reaction; if this heat removal is not sufficient, it can lead to a

runaway situation, and also to a lowering in the performance of the process and

degradation of the catalyst used.

Research has been done in order to develop more economic processes with overall

energy saving and, if possible, with fewer pollutants, using new types of catalysts and

reactors. In the literature it is possible to observe that this effort is being developed by

many researchers that are trying to adapt the best reactor configuration to the more

efficient process.

In 1995, Peureux et al [59] studied the potentialities of a catalytic membrane reactor

in a gas-liquid-solid reaction. The hydrogenation of NB was chosen as a model reaction

and it was carried out in a catalytic membrane reactor, constituted by Pt/(γ-Al2O3+α-

Al2O3), especially designed for gas-liquid experiments. The tubular catalytic membrane


46

divided the reactor in two parts and in both were separately introduced gas or liquid

reactants. Two modes of feeding were explored, as illustrated in Figure 2.26.

Figure 2.26 – The two modes of reactants introduction in a catalytic membrane reactor

[59].

In mode 1 it was verified that an increase of reaction temperature had only a limited

effect on the activity, whatever the pore size of the catalytic membrane, suggesting a

conversion controlled by diffusional limitations. It was also observed that the activity was

proportional to the hydrogen pressure in good agreement with a control by hydrogen

transfer. In mode 2, the effect of the reaction temperature on the activity showed to

clearly depend on the average pore size of the catalytic membrane, being the limitations

attributed to the diffusion of NB in the porous γ-Al2O3 membrane. Contrarily to mode 1,

hydrogen transfer from the gas phase to the catalyst is very efficient.

Authors concluded that this type of reactor presents potential interest for three-

phase reactions because there should be no problem with thermal stability of the

membrane and the catalyst is a part of the reactor, so that its recovery does not need any

separation from the liquid medium, as is the case in conventional slurry reactors. This

specific configuration should also limit the catalyst losses and allow in-situ regenerations.

Amon et al. [75], a few years later, investigated the deactivation by coking of a

noble metal catalyst in the catalytic hydrogenation of NB using a catalytic wall reactor.

The kinetic measurements for highly exothermic reactions, used in this case for

quantitative description of coke formation, should not be performed in fixed-bed reactors

since isothermal conditions cannot be ensured. Differential recycle reactors show


47

isothermal behavior, but they are often very complicated to use. Besides this, the presence

of a large internal surface area and a high free volume allows homogeneous reactions to

proceed, which often interfere with the kinetic measurements at high temperatures.

Catalytic wall reactor (CWR) combines the advantages of both the plug flow and recycle

reactor. In the CWR the catalyst is coated on the reactor wall and the heat of reaction can

be removed directly through the wall. The quite simple construction of the CWR enables

one to easily use the renewable reactor tubes. An industrial (1.1 wt. %) Pd-Al2O3 catalyst

was used, being the inner tube wall-coated with the catalyst. The coatings were thermally

and mechanically stable for use in the CWR. The CWR consists of several tubular reactor

segments that are of varying length, as shown in Figure 2.27. The experimental conditions

used were temperature range 275 - 425 ºC and pressure range 0.02 – 0.07 bar.

Figure 2.27 – Catalytic wall reactor configuration [75].

Measurements of the conversion at variable bulk velocities and catalyst thickness

showed no significant change in conversion. The selectivity to ANL was always higher

than 99 % and no by-products except coke on the catalyst could be determined

quantitatively. It was also concluded that hydrogen adsorbs much stronger to palladium

than NB, inhibiting the reaction rate. The coking of catalyst strongly depends on the

temperature and below 325 ºC coke formation is slow.


48

In the US Patent 2000/6040481 [43] a method for hydrogenating aromatic nitro

compounds is proposed by reacting at least one aromatic nitro compound with hydrogen

in two adiabatically operated fixed-bed reactors arranged in series, wherein a part of the

reaction mixture from the first reactor is recirculated therein while the other part is fed

into the second reactor (Figure 2.28). Thus, it was proposed a tubular reactor with a fixed

bed, injection of the reactants co-currentwise downwards in the reactor, in which the

supply flow comprises some of the recycled reaction mixture coming from the reactor.

However, the reactor dimensions are such that this reactor cannot be considered as

operating adiabatically since 60 to 70% of the heat exchanges take place by means of

losses through the walls of the reactor. Consequently, extrapolation of this process to the

industrial scale would result in having reactors whose height would be considered greater

than their cross-section, so as to conserve a reasonable level of heat exchange through the

walls. However, such reactors would need to use very large amounts of catalyst, which

would detract from one of the advantages of the process with a fixed bed compared with

catalyst suspensions. Another possibility for the industrial-scale exploitation of this type

of process would be to increase the heat exchanges with the recycled reaction flow. In

other words, this would consist in increasing the rate of recycling of the reaction medium

considerably. In this case, the production efficiency of such a process would be much too

low, on account of the large dilution required for the compounds to be nitrated and the

need to convert these compounds completely.

The process according to the invention is easy to carry out and it incurs in lower

capital costs since it is no longer necessary to provide complex systems for separating out

the catalyst once the reaction is complete, nor even to provide stirring systems, with all

the problems of leak tightness inherent therein. The inventors proposed two reactors in

series, where each of the reactors are preferably supplied co-currentwise with hydrogen

and one aromatic nitro compound and/or the recycled or non-recycled reaction mixture,

such that the continuous phase in the reactors is liquid. Such result is obtained by

supplying the reactors from the bottom upwards. In this way, the catalyst bed is immersed

in the reactants and/or reaction mixture.


49

Figure 2.28 – Configuration proposed in US 2000/6040481 [43].

One of the essential characteristics of the invention lies in the fact that some of the

reaction mixture, coming from the first reactor, is recycled into the supply flow of this

reactor, the other part being introduced into the second reactor (ratio flow rate of supply

of first reactor to reaction mixture recycled is between 0.1 % and 10 %). The temperature

at the foot of the first reactor is about 80 to 170 ºC. On leaving the first reactor, the rise in

temperature is less than 120 ºC (difference between the outlet temperature of the reaction

mixture and the inlet temperature of the nitro compounds and/or of the recycled reaction

mixture). The reaction mixture coming from the first reactor is introduced into a gas-

liquid separator. The resulting gaseous fraction is separated into two parts, one of which

is recycled into the supply flow of the first reactor. Prior to the recycling in the first

reactor and introduced into the second reactor, the reaction mixture is cooled in order to

have the desired temperature after mixing with reactants and separated from the gases it

contains.

The reactors in which the process is carried out are preferably cylindrical tubes

fitted with standard means for retaining the fixed bed of catalyst and for distributing the

liquid and gas flows. On account of the high level of recycling, the behavior of the first

reactor is of the stirred type. As regards the second reactor, a piston type (plug flow)

reactor is used. It is advantageous to combine the two reactors since this makes possible

to minimize the overall amount of catalyst and thus the size of the reactors.

Most of the hydrogenation reaction is carried out in the first reactor. A degree of

conversion of greater than or equal to 90 % of hydrogenated aromatic nitro compounds


50

relative to the aromatic nitro compounds fed into the recycling loop is achieved. The

second reactor corresponds to a finisher by means of which degrees of conversion of 100

% are achieved.

The hydrogen used is more particularly pure hydrogen. The hydrogen is supplied in

the stoichiometric amount, preferably in excess relative to the aromatic nitro compounds

(5 to 50 % mole). The hydrogen pressure in the reactor ranges between 10 to 50 bar. The

reaction is carried out in the presence of a hydrogenation catalyst which is conventional in

this field.

In 2011, Hunstman company [76] patented a new process for the conversion of NB

into ANL. The main objective was to reduce the amount of byproducts in the reactor

effluent. It was proposed a process that comprises a fixed-bed catalytic reactor, in fact a

trickle-bed reactor, to which is fed a hydrogen gas stream and NB liquid stream (Figure

2.29). The effluent is constituted by ANL and water and one part is recycled to the inflow

of the reactor and the other is sent to a separation unit. The process can be run providing

an industrially acceptable production yield.

Figure 2.29 – Process flow by Huntsman [76].

An excess of hydrogen is used, typically 5 to 30% mole above the stoichiometric

requirement to convert all the NB into ANL. It is believed that the higher hydrogen partial


51

pressure at the inflow side of the reactor causes the hydrogenation reaction to take off

quickly, initiating a high conversion of NB into ANL, whereas the lower partial pressure

at the outflow side of the reactor decreases the formation of by-products.

Through control of the water (or inert solvent), fed carefully at the inflow side of

the reactor, is possible to achieve a selectivity of NB to ANL of 99.8 %, with a

conversion of NB of 99.98 %. There may remain a potential for an incomplete reaction of

the NB, as it passes through the fixed-bed reactor, due to channeling of either gas or

liquid or both. Consequently, the NB concentration exiting the catalyst bed may reach

several thousand ppm. This residual NB can be efficiently removed from the ANL

product by use of an in-line polishing device. This polisher takes the ANL product at 200

– 300 ºC and 30 – 40 barg from the exit of the reactor cooler.

In order to avoid channeling in the main reactor, it is possible to split the bed of the

fixed-bed reactor into several bed sections, where the gas and liquid from each bed are

redistributed. The hydrogen partial pressure may be controlled by controlling the amount

of water recycled (or inert solvent), because those compounds will evaporate during their

passing throughout the reactor, and as such gradually increase their contribution to the

pressure, possibly reducing or influencing the partial pressure of the hydrogen. Another

way to deal with the hydrogen partial pressure may be by adjusting a temperature profile

through the reactor. In the inflow side it should be provided a temperature in the range of

160 – 200 ºC and in the outflow side in the range of 240 – 280 ºC. The temperature

difference over the reactor should be, preferably, between 40 to 100 ºC. Concerning the

pressure, it may be kept substantially constant, at 30 – 40 barg, being the pressure drop

between the inflow and the outflow of 0.2 to 4 barg.

The reactor effluent is typically a liquid product and a gaseous product and can be

cooled/condensed separately. The gas product, which has a higher concentration of light

impurities, and a lower concentration of heavy impurities than the total reactor product,

can become the principal reactor product, used as product for further processing

(purification process). Some part of the liquid phase may be purged to evacuate the

impurities from the process, whereas another part of the liquid phase can be used to

recycle to the inflow of the reactor, being the water added to the reactor inflow that acts

like a solvent.


52

The fixed-bed reactor is preferably a trickle-bed reactor that is operated

adiabatically and at a constant pressure throughout the reactor. The inflow in the reactor

may be in vertical direction, using preferably a top flow. In order to better control the

hydrogen partial pressure, the reactor may be provided with cooling equipment such as

shell and tube, plate and/or spiral heat exchanger systems. The advantage of the mixture

obtained, according to this patent, is related with the fact that it does not need further

purification of the mixture prior to using it in a process for producing isocyanates.

The authors made several tests, using various H2O/H2 ratios in the inflow of a NB

to ANL trickle-bed reactor. Two different catalyst were tested, the first one, CAT(I), was

as Fe promoted Pd catalyst on a Al2O3. The second, CAT(II), was a Pd/Al2O3

commercially available from Johnson Matthey. The test conditions are presented in Table

2.7.

Table 2.8 – Experimental conditions used in the several tests of the Hunstman patent [76].

Test 1 2 3 4 5 6

Catalyst CAT(I) CAT(I) CAT(I) CAT(I) CAT(II) CAT(II)

H2 (mol/h) 0.277 0.277 0.277 0.277 0.277 0.277

MNB (mol/h) 0.072 0.072 0.072 0.072 0.072 0.072

ANL (mol/h) 0.582 0.582 0.582 0.582 0.582 0.582

H2O (mol/h) 1.667 0.833 0.417 0 1.667 0

H2O/ H2 molar 6 3 1.5 0 6 0

Set-point Temp. (ºC) 250 250 250 250 250 250

Pressure (barg) 20 20 20 20 20 20

Mass of Catalyst (g) 4 4 4 4 4 4

The amount of over-reduced species in the reactor effluent, such as Benzene, CHA,

CHOL, CHONA, DICHA, CHENO and CHANIL is presented in Table 2.9.


53

Table 2.9 – Experimental results of the tests described in Table 2.7 [76].

Test 1 2 3 4 5 6

H2O/ H2 molar 6 3 1.5 0 6 0

Benzene (ppm) 80 86 79 98 68 40

CHA (ppm) 22 149 235 1515 0 546

CHOL (ppm) 58 123 253 1018 40 46

CHONA (ppm) 952 2289 2968 3818 610 2772

DICHA (ppm) 0 0 0 249 0 0

CHANIL (ppm) 101 436 1256 2428 0 761

CHENO (ppm) 75 713 1800 11039 19 4583

DPA (ppm) 135 316 484 704 124 441

Total (ppm) 1423 4112 7075 20869 861 9189

It is clear that when decreasing the molar ratio of water over hydrogen an increase

of over-reduced species (CHA, CHOL, CHONA, CHANIL and CHENO) is found.


54

References

[1] CUF-QI Internal Report not published, May 2012.

[2] Sonnenschein, M., Polyurethanes: Science, Technology, Markets and Trends, John

Wiley & Sons, Inc, 2014

[3] CUF Corporate Presentation, June 2013.

[4] Machado, M., Fundamentals of Mass Transfer and Kinetics for the Hydrogenation of

Nitrobenzene to Aniline, Air products and Chemicals, No 1, 2007, 1-14.

[5] Nexant, PERP 02/03-2 Nitrobenzene/Aniline, Nexant CHEM SYSTEMS, 2003.

[6] Wegener, G., Brandt, M., Duda, L., Hofmann, J., Klesczewski, B., Koch, D., Kumpf,

R., Orzesek, H., Pirkl, H., Six, C., Steinlein, C., Weisbeck, M., Trends in industrial

catalysis in the polyurethane industry, Applied Catalysis A: General 221 (2001) 303-335.

[7] Gelder, E., Jackson, S. and Lok, C., The hydrogenation of nitrobenzene to aniline: a

new mechanism, Chemical Communication (2005) 522-524.

[8] Gelder, E., Jackson, S. and Lok, C., A study of nitrobenzene hydrogenation over

palladium/carbon catalysts, Catalysis Letter 84 (2002) 205-208.

[9] Haber, F., On the stepwise reduction of nitrobenzene with limited cathode potencial,

Z. Elektrochem., 4 (1898) 506-514.

[10] Burge, H., Collins, D., Burtron H. Davis, Intermediates in the Raney Nickel

Catalyzed Hydrogenation of Nitrobenzene to Aniline, Industrial & Engineering Chemistry

Product Research and Development 19 (1980) 389-391.

[11] Höller, V., Wegricht, D., Yuranov, I., Kiwi-Minsker and, L., Renken, A., Three-

phase nitrobenzene hydrogenation over supported glass fiber catalysts: kinetics study,

Chemical Engineering & Technology 23 (2000) 251-255.

[12] Figueras, F., Coq, B., Hydrogenation and hydrogenolysis of nitro-, nitroso-, azo-,

azoxy- and other nitrogen-containing compounds on palladium, Journal of Molecular

Catalysis A: Chemical 173 (2001) 223-230.

[13] Wisniak, J., Klein, M., Reduction of Nitrobenzene to Aniline, Industrial &

Engineering Chemistry Product Research and Development 23 (1984) 44-50.

[14] Makaryan, I.A., Savchenko, V.I., n-Arylhydroxylamines transformation in the

presence of heterogeneous catalysts, 10th International Congress on Catalysis, 19-24 July,

1992, Budapest, Hungary.

[15] Corma, A., Concepción, P., Serna, P., A different reaction pathway for the reduction

of Aromatic Nitro compounds on gold catalysts, Angewandte Chemie 119 (2007) 7404-

7407.

http://www.sciencedirect.com/science/article/pii/S0926860X01009103


http://pubs.acs.org/action/doSearch?action=search&author=Wisniak%2C+Jaime&qsSearchArea=author


55

[16] Makosh, M., Sá, J., Kartusch, C., Richner, G., van Bokhoven, J., Hungerbühler, K.,

Hydrogenation of nitrobenzene over Au/MeOx catalysts – a matter of the support, Chem

Cat Chem 4 (2012) 59-63.

[17] Meng. X., Cheng, H., Akiyama, Y., Hao, Y., Qiao, W., Yu, Y., Zhao, F., Fujita, S.,

Arai, M., Selective hydrogenation of nitrobenzene to aniline in dense phase carbon

dioxide over Ni/γ-Al2O3: Significance of molecular interactions, Jornal of Catalysis 264

(2009) 1-10.

[18] Chatterjee, M., Chatterjee, A., Kawanami, H., Ishizaka, T., Suzuki, T., Suzuki, A.,

Rapid hydrogenation of aromatic nitro compounds in supercritical carbon dioxide:

Mechanistic implications via experimental and theoretical investigations, Advanced

Synthesis & Catalysis 354 (2012) 2009-2018.

[19] Rakitin, M., Doluda, V., Tereshchenkov, A., Demidenko, G., Lakina, N., Matveeva,

V., Sul’man, M., Sul’man, E., Investigating the catalytic hydrogenation of nitrobenzene

in supercritical carbon dioxide using Pd-containing catalysts, Catalysis in Industry 7

(2014) 1-5.

[20] Turáková, M., Salmi, T., Eränen, K., Wärna, J., Murzin, D., Králik, M., Liquid phase

hydrogenation of nitrobenzene, Applied Catalysis A: General 499 (2015) 66-76.

[21] Nagata, T., Watanabe, K., Kono, Y., Tamaki, A., Kobayashi, T., Process for

preparing high-purity aniline, US Patent 5283365, 1994.

[22] Narayanan, S., Unnikrishnan, R., Vishwanathan, V., Nickel-Alumina prepared by

constant and varying pH method: Evaluation by hydrogen-oxygen chemisorption and

aniline hydrogenation, Applied Catalysis A: General 129 (1995) 9-19.

[23] Narayanan, S., Unnikrishnan, R., Comparison of hydrogen adsorption and aniline

hydrogenation over co-precipitated Co/Al2O3 and Ni/Al2O3 catalysts, Journal of the

Chemical Society, Faraday Transactions 93 (1997) 2009-2013.

[24] Relvas, J., Optimization of the hydrogenation reaction of Nitrobenzene to aniline in a

3-phase hydrogenation reactor, Dissertation presented for the Doctor of Philosophy

degree in Chemical Engineering, 2008.

[25] Králik, M., Turáková, M., Macák, I., Wenchich, S., Catalytic Hydrogenation of

Aromatic Compounds in the liquid phase, Journal of Chemistry and Chemical

Engineering 6 (2012) 1074-1082.

[26] Rubio-Marqués, P., Leyva-Pérez, A., Corma, A. A bifunctional palladium/acid solid

catalyst performs the direct synthesis of cyclohexylanilines and dicyclohexylamines from

nitrobenzenes, Chemical Communications 49 (2013) 8160-8162.

[27] Sousa, J., Intensification of Aniline production: kinetic study of Nitrobenzene

hydrogenation, Dissertation presented for the Doctor of Philosophy degree in Refining,

Petrochemical and Chemical Engineering, 2015.

[28] Downing, R., Kunkeler, P., van Bekkum, H., Catalytic syntheses of aromatic amines,

Catalysis Today 37 (1997) 121-136.


56

[29] Sperber, H., Poehler, G., Pistor, H., Wegerich, A., Production of Aniline, US Patent

3136818, 1964.

[30] Langer, R., Buysch, H., Pentling, U., Wagner, P., Process for the production of

aromatic amines, US Patent 5877350, 1999.

[31] Seidemann, L., Koenigsmann, L., Schneider, C., Schwab, E., Stuetzer, D., Liekens,

C., Process for preparing aromatic amines in fluidized-bed reactor, US Patent 8044244,

2011.

[32] Koenigsmann, L., Schwab, E., Hesse, M., Schneider, C., Heidemann, T., Liekens, C.,

Bickelhaupt, T., Theis, D. Method for producing aromatic amines, US Patent 0065431,

2012.

[33] Langer, R., Buysch, H., Wagner, P., Pentling, U., Process for producing aromatic

amines, US Patent 5808157, 1998.

[34] Dugal, M., Gehlen, F., Wershofen, S., Lago, A., Lehner, P., Backer, W., Marotz, B.,

Brinkschulte, H., Process for preparing aniline, US Patent 0238901, 2007.

[35] Sommer, K., Wilke, K., Lehner, P., Gehlen, F., Mleczko, L., Schubert, S.,

Bellinghausen, R., Hoffmann, E., Process for the preparation of aromatic amines, US

Patent 0280271, 2010.

[36] Chemopetrol, Process for the production of aniline and apparatus therefor, Patent

1452466, 1976.

[37] Gonzalez, R., Hydrogenation of aromatic nitro compounds, US Patent 3499034,

1970.

[38] Lawrence, F. Process for preparing aniline, US Patent 4415754, 1983.

[39] Coke, E. Thurlow, H., Catalytic hydrogenation of nitro aromatic compounds to

produce the corresponding amino compounds, US Patent 3270057, 1966.



[41] Králik, M., Turáková, M., Macák, I., Lehocký, Aniline – Catalysis and Chemical

Engineering, In Proceedings of the 41st International Conference of Slovak Society of

Chemical Engineering, Tatranské Matliare, Slovakia, 723-733, 2014.

[42] Kise, M., Production of aromatic amines, US Patent 2292879, 1942.

[43] Chambost, B., Marion, P., Mathieu, C., Method for hydrogenating aromatic nitro

compounds, US Patent 6040481, 2000.

[44] Alberts, P., Auer, E., Gross, M., Krauter, J., Packruhn, U., Catalyst for the

hydrogenation of aromatic nitro compounds, US Patent 0077504, 2002.

[45] Hassan, A., Bagherzadeh, E., Anthony, R., Borsinger, G., Hassan, A., System and

process for production of aniline and toluenediamine, US Patent 8153076, 2012.

http://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=19761013&DB=EPODOC&locale=en_EP&CC=GB&NR=1452466A&KC=A&ND=4


57

[46] Sheng. X., Wouters, B., Breugelmans, T., Hubin, A., Vankelecom, I., Pescarmona,

P., Cu/CuxO and Pt nanoparticles supported on multi-walled carbon nanotubes as

electrocatalysts for the reduction of nitrobenzene, Applied Catalysis B: Environmental

147 (2014) 330-339.

[47] Gharda, K., Sliepcevich, C., Hydrogenating Nitrobenzene to Aniline, Industrial &

Engineering Chemistry 52 (1960) 417-420.

[48] Rihani, D., Narayanan, T., Doraiswamy, L., Kinetics of catalytic vapor-phase

hydrogenation of nitrobenzene to aniline, Industrial & Engineering Chemistry Product

Research and Development 4 (1965) 403-410.

[49] Birkenstock, U., Lachmann, B., Metten, J., Schmidt, H., Process for the catalytic

hydrogenation of nitrobenzene, US Patent 4265834, 1981.

[50] Immel, O., Waldmann, H., Braden, R., Fröhlich, C., Friedhofen, G., Niemeier, W.,

Catalyst for the preparation of aniline, US Patent 5304525, 1994.

[51] Sangeetha, P., Seetharamulu, P., Shanthi, K., Narayanan, S., Rama Rao, K.S Studies

on Mg-Al oxide hydrotalcite supported Pd catalysts for vapor-phase hydrogenation of

nitrobenzene, Journal of Molecular Catalysis A: Chemical 273 (2007) 244-249.

[52] Sangeetha, P., Shanthi, K., Rama Rao, K.S., Viswanathan, B., Selvam, P.,

Hydrogenation of nitrobenzene over palladium-supported catalysts - Effect of support,

Applied Catalysis A: General 353 (2009) 160-165.

[53] Nieto-Márquez, A., Gil, S., Romero, A., Valverde, J., Gómez-Quero, S., Keane, M.,

Gas phase hydrogenation of nitrobenzene over acid treated structure and amorphous

carbon supported Ni catalysts, Applied Catalysis A: General 363 (2009) 188-198.

[54] Mohan, V., Pramod, C.V., Suresh, M., Hari Prasad Reddy, K., David Raju, B., Rama

Rao, K.S., Advantage of Ni/SBA-15 catalyst over Ni/MgO catalyst in terms of catalyst

stability due to release of water during nitrobenzene hydrogenation to aniline, Catalysis

Communication 18 (2012) 89-92.

[55] Sudhakar, M., Kantam, L., Kishore, R., Kumar, S., Venugopal, A., Vapour phase

hydrogenation over metal (Ru, Ni, Pt, Pd supported on Ca5(PO4)3(OH) catalysts, Indian

Journal of Chemistry 53A (2014) 550-552.

[56] Varkolu, M., Velpula, V., Pochamoni, R., Muppala, A., Burri, D., Kamaraju, S.,

Nitrobenzene hydrogenation over Ni/TiO2 catalyst in vapour phase at atmospheric

pressure: the influence of preparation method, Applied Petrochemical Research (2015) 1-

9.

[57] Zhao, F., Zhang, R., Chatterjee, M., Ikushima, Y., Arai, M., Hydrogenation of

nitrobenzene with transition metal catalysts in supercritical carbon dioxide, Advanced


[58] Turek, F., Geike, R., Lange, R., Liquid-phase Hydrogenation of Nitrobenzene in a

Slurry reactor, Chemical Engineering and Processing: Process Intensification 20 (1986)

213-219.


58

[59] Peureux, J., Torres, M., Mozzanega, H., Giroir-Fendler, A., Dalmon, J-A.,

Nitrobenzene liquid-phase hydrogenation in a membrane reactor, Catalysis Today 25

(1995) 409-415.

[60] Patel, D., Ram, R., Hydrogenation of nitrobenzene using polymer anchored Pd(II)

complexes as catalyst, Journal of Molecular Catalysis A: Chemical 130 (1998) 57-64.

[61] Li, C., Yu, Z., Yao, K., Ji, S., Liang, J., Nitrobenzene hydrogenation with carbon

nanotube-supported platinum catalyst under mild conditions, Journal of Molecular


[62] Relvas, J., Andrade, R., Gama Freire, F., Lemos, F., Araújo, P., Pinho, M., Pedro

Nunes, C. Ramôa Ribeiro, F., Liquid phase hydrogenation of nitrobenzene over an

industrial Ni/SiO2 supported catalyst, Catalysis Today 133-135 (2008) 828-835.

[63] Mahata, N., Cunha, A.F., Órfão, J.J.M. Figueiredo, J.L., Hydrogenation of

nitrobenzene over nickel nanoparticles stabilized by filamentous carbon, Applied

Catalysis A: General 351 (2008) 204-209.

[64] Blaser, H., Indolese, A., Schnyder, A., Steiner, H., Studer, M., Supported palladium

catalysts for fine chemicals synthesis, Journal of Molecular Catalysis A: Chemical 173

(2001) 3-18.

[65] El-Hout, S., El-Sheikh, S., Hassan, H., Harraz, F., Ibrahim, I., A green chemical

route for synthesis of graphene supported palladium nanoparticles: A highly active and

recyclable catalyst for reduction of nitrobenzene, Applied Catalysis A: General 503

(2015) 176-185.

[66] Eremenko, N., Podyacheva, O., Ismagilov, Z., Obraztsova, I., Eremenko, A., Kibis,

L., Svintsitskiy, D., Highly dispersed palladium on carbon nanofibers for hydrogenation

of nitrocompounds to amines, Euroasian Chemico-Technologicaç Journal 17 (2015) 101-

103.

[67] Du, B., Su, H., Wang, S., Palladium supported on carbon nanofiber coated moniliths

for three-phase nitrobenzene hydrogenation: Influence of reduction temperature and

oxidation pre-treatment, Journal of Industrial and Engineering Chemistry 21 (2015) 997-

1004.

[68] Turáková, M., Králik, M., Lehocký, P., Pikna, L., Smrcová, M., Remeteiová, D.,

Hudák, A., Influence of preparation method and palladium content on Pd/C catalysts

activity in the liquid phase hydrogenation of nitrobenzene to aniline, Applied Catalysis A:

General 476 (2014) 103-112.

[69] Knifton, J., Suggitt, R., Homogeneous ruthenium catalysts useful in the selective

reduction of nitroaromatics to amines, US Patent 3906045, 1975

[70] Malaika, A., Rechnia, P., Krzyżyńska, B., Kozłowski, M., The influence of texture of

activated carbons on their catalytic activity in the process of ethylbenzene

dehydrogenation coupled with nitrobenzene hydrogenation, Microporous and

Mesopouros Materials 163 (2012) 300-306.

http://www.sciencedirect.com/science/article/pii/0920586195001283


59

[71] Xu, K., Zhang, Y., Chen, X., Huang, L., Zhang, R., and Huang, J., Convenient and

selective hydrogenation of nitro aromatics with a platinum nanocatalyst under ambient

pressure, Advanced Synthesis & Catalysis 353 (2011) 1260-1264.

[72] Bouchenafa-Saïb, N., Grange, P., Verhasselt, P., Addoun, F., Dubois, V., Effect of

oxidant treatment of date pit active carbons used as Pd supports in catalytic

hydrogenation on nitrobenzene, Applied Catalysis A: General 286 (2005) 167-174.

[73] Zhao, F., Ikushima, Y., Arai, M., Hydrogenation of nitrobenzene with supported

platinum catalysts in supercritical carbon dioxide: effects of pressure, solvent and metal

particle size, Journal of Catalysis 224 (2004) 479-483.

[74] Sinnott, R., Coulson & Richardson’s Chemical Engineering, Volume 6, Third

edition. Burlington: Butterworth-Heinemann, 1999.

[75] Amon, B., Redlingshöfer, H., Klemm, E., Dieterich, E., Emig, G., Kinetic

investigations of the deactivation by coking of a noble metal catalyst in the catalytic

hydrogenation of nitrobenzene using a catalytic wall reactor, Chemical Engineering and

Processing: Process Intensification 38 (1999) 395-404.

[76] Mitchell, C., Stewart, D., Process fot the production of aromatic nitro compounds

into amines, WO Patent 113491, 2011.

http://www.sciencedirect.com/science/article/pii/S0255270199000379



Part II

Preliminary catalytic tests in a Continuous

Stirred-Tank Reactor (CSTR)

*Adapted from: Clara Sá Couto, Luís M. Madeira, Clemente Pedro Nunes, Paulo Araújo, Hydrogenation of

Nitrobenzene over Pd/Al2O3 Catalyst – Mechanism and Effect of the Main Operating Conditions,

Chemical Engineering & Technology, vol. 38, No. 9, 2015, 1625-1636 (DOI:

http://dx.doi.org/10.1002/ceat.201400468). 63

Chapter 3 - Hydrogenation of Nitrobenzene over a

Pd/Al2O3 Catalyst – Mechanism and Effect of the

Main Operating Conditions*.

Abstract

The catalytic hydrogenation of nitrobenzene (NB) was studied in a three-phase

basket reactor and the catalyst used was a commercial 1 wt.% Pd/Al2O3 sample. The

kinetic experiments allowed a better understanding of the mechanism behind aniline

(ANL) and by-products formation, a topic not yet well understood. The effect of some

operating conditions was studied and it was found that there are more by-products than

those referred in the literature; specifically, benzene formation was verified. It was also

found that both the reaction kinetics and selectivity are strongly dependent on the

temperature, while the effect of total pressure is not that pronounced. Moreover, the high

selectivity of the catalyst used in the present work was put in evidence, and as such the

deep hydrogenation of ANL to form by-products only occurs in considerable extension

when NB concentration in the reaction mixture becomes negligible.


64

3.1 Introduction

The catalytic hydrogenation of nitrobenzene (NB) is an important industrial

reaction used in the commercial production of aniline (ANL) that is in turn a major

intermediate in the polyurethane industry. ANL is mainly consumed in the production of

methylene diphenyldiisocyante (MDI) [1].

ANL production can be carried out in gaseous or in liquid phase [2]. For the vapor

phase reaction, fluidized and fixed-bed reactors are used, typically at temperatures of 200

to 400 ºC and pressures in the range of 1-10 bar, whereas for liquid phase the common

reactor used is the slurry one operating in the temperature range of 100-300 ºC and at

pressures between 1.5 and 22 bar [3-8]. The choice of the catalyst to be used will depend

on the type of reactor and also on the possibility of being poisoned by nitro-compounds.

For instance, it was detected that in the case of Ni-based catalysts the presence of large

excess of NB may cause some kind of inhibition, since NB will be adsorbed preferentially

on the most active sites, remaining only a few sites available to promote the

hydrogenation reaction [2].

At industrial scale, for the reaction in liquid phase the main catalysts reported are

based on noble metals (Pd, Pt, Ru) [5-6]. Nevertheless, many studies have already been

made, in order to obtain the best catalyst to use in this reaction (Table 3.1).

Table 3.1– Main catalysts studied for NB hydrogenation.

Process type Catalyst type Ref

Vapour phase

Cu-based catalysts [9]

Ni-based catalysts [10-11]

Pd-based catalysts [12-13]

Liquid phase

Ni-based catalysts [2]

Pt-based catalysts [14-15]

Pd-based catalysts [6], [15-19]

Although there is a large volume of literature available describing this reaction,

there is not much information about the reaction mechanism. This mechanism is complex

and can be divided in two main steps: the first one is the formation of ANL and the

second one is the production of secondary products from the deep hydrogenation of ANL.

Chapter 3 – Hydrogenation of Nitrobenzene over a Pd/Al2O3 Catalyst –

Mechanism and Effect of the Main Operating Conditions

65

The first step was explained by the mechanism for ANL formation through NB

hydrogenation that was first proposed by Haber, in 1898 [20]. However, this mechanism

does not fully explain all the experimental results. Consequently, more studies were done,

and another mechanism was proposed by Wisniak and Klein [21] that is slightly more

complicated than the Haber’s one. Some years later, Gelder et al. [22] suggested that the

number of steps involved in ANL formation was higher and substantially different from

those previously reported.

There is more information available for the mechanism of ANL production than for

the formation of secondary products (second step), for which the lack of information is

higher. Nagata et al. [23] proposed a mechanism to explain the formation of some of

these compounds (Figure 3.1a)). However, this mechanism does not include the formation

of cyclohexanol (CHOL) and dicyclohexylamine (DICHA) that are observed in

experimental tests carried out by several authors and also in those performed in the

present work.

In 1995, Narayanan et al. [10] identified the following reaction products:

cyclohexane (CH), cyclohexylamine (CHA), DICHA and N-phenylcyclohexylamine

(NPCHA). They observed that in the case of supported nickel catalysts, depending upon

the conditions of the experiment and catalyst metal content, products such as CHA and

DICHA are formed. In a latter article, Narayanan and Unnikrishnan [24] compared the

catalytic properties of Co/Al2O3 and Ni/Al2O3 materials. It was concluded that by

increasing the contact time of ANL, the conversion increased and CHA and NPCHA were

the two major products. CHA formation was slightly favored at low contact time and

NPCHA was formed in roughly similar amounts at all contact times. However, at high

contact times DICHA (DCHA in their notation) and CH were also formed. Based on

these results they proposed the mechanism for the formation of secondary products shown

in Figure 3.1b).

In 2008, Relvas [25], using a Ni catalyst, proposed another mechanism, based on

the Nagata one. For this proposal, several laboratorial tests were made, in which

operating condition effects were studied. DICHA and CHOL (through cyclohexanone –

CHONA – hydrogenation) were included in the proposed mechanism (Figure 3.2)), since

they were detected in the experimental tests, as well as N-cyclohexylaniline (CHANIL)

and cyclohexyldeneaniline (CHENO). Nevertheless, as recognized by Relvas [25], this


66

mechanism does not fully explain the formation of all secondary products in NB

hydrogenation. So, further work in this topic is still required.

a)

b)

Figure 3.1 – Reaction network for the formation of ANL and secondary products as proposed by

a) Nagata et al. [23]; b) Narayanan and Unnikrishnan [24].

NO2 NH2

H2

[ 1 ]

NH2 NH2NH2

NHNH2

NNH

O

H2

[ 2 ]

CHA

NH2

H2

[ 3 ]

[ 4 ]

N-phenylcyclohexylamine

[ 7 ] H2

CHANIL

- NH3

+ H2O

H2

[ 8 ]

[ 5 ]

CHONA

NH2

[ 6 ]

CHENO

NB ANLIMINE

NH2NH2

Co / Al2O3

Ni / Al2O3

+ H2

CHA

NH

CHANIL

+ NH3

- NH3

DICHA

+ CHA

+ H2

NH

ANL

ANL

CH



67

Figure 3.2 –Relvas [25] (*very reactive and unstable compounds).

Summarizing, the information available in the literature about the overall reaction

mechanism involved in the NB hydrogenation is not sufficient. It is important to refer that

the knowledge and understanding of this mechanism has implications for both catalyst

and reactor design. Consequently, the present study has the objective of trying to

determine all the compounds obtained from this reaction, and also the overall mechanism

that may explain both ANL and secondary products formation during the NB

hydrogenation. The experiments were performed using a Pd-based catalyst and under

different experimental conditions.

3.2 Material and Methods

Hydrogenation of NB, in liquid phase, was carried out in a 1L-capacity Parr batch

reactor provided with an air-impelled stirrer to promote appropriate mixing and gas

distribution during the experiments. The stirrer was equipped with a basket, where the

catalyst was placed. The catalyst used was a commercial 1 wt.% Pd/Al2O3 material in

spherical form with a diameter of 2-4 mm (catalyst I.1).

A known amount of Pd/Al2O3 catalyst was loaded into the reactor. The material was

pre-treated in situ, as described in the Appendix A. NB was charged in a vessel that is

Cat

+H2,-H20

Cat

H2

+ANL

-NH3

Cat

H2

+H2 -NH3

+H2

+H2

-NH3, -H2O

+ANL -H2O

+H2

+H2 CHONA

CHENO

CHOL

DICHA

CHANIL

CHAANL

Amine


N

O

OH

NO2 NH2

NH2

NHNH2

NH

NH

NB

*


68

connected to the reactor and subsequently was pushed into the reactor by using a high-

pressure hydrogen stream (cf. Figure A.1 in the Appendix A). The procedures used in the

catalytic studies and in the analyses along reaction time were the same as described by

Relvas et al. [2] and are summarized in the Appendix A. The reference values for

temperature, pressure and nitrobenzene concentration during the parametric study are T =

150 ºC, P = 14 barg and Cref = 10 wt. % NB (100 000 ppm), respectively. The

experiments performed and the conditions used are given in Table 3.2. Test B12 was

performed only with ANL, in order to study the formation of secondary products. The

catalyst to liquid volume ratio was in all cases 80.0 g L-1.

Table 3.2 - Initial conditions of the experiments performed.

Test Experimental conditions

B1 150 ºC, 14 barg, 10 wt. % NB

B2 150 ºC, 20 barg, 10 wt. % NB

B3 150 ºC, 30 barg, 10 wt. % NB

B4 150 ºC, 14 barg, 10 wt. % NB

B5 240 ºC, 14 barg, 10 wt. % NB

B6 150 ºC, 14 barg, 10 wt. % NB

B7 150 ºC, 14 barg, 30 wt. % NB

B8 150 ºC, 14 barg, 10 wt. % NB

B9 180 ºC, 14 barg, 10 wt.% NB

B10 210 ºC, 14 barg, 10 wt. % NB

B11 150 ºC, 14 barg, 3 wt. % NB

B12 150 ºC, 14 barg

The concentration of all the compounds is presented as a relative dimensionless

value that was calculated as follows:

𝐶𝑖 = 𝐶𝑖,𝑡 (𝑝𝑝𝑚)

𝐶𝑟𝑒𝑓 (𝑝𝑝𝑚) (3.1)

where Ci is the dimensionless concentration of component i, Ci,t is the concentration of

component i at time t and Cref stands for the NB reference concentration.



69

3.3 Results and Discussion

As mentioned above, the aim of this work was to better understand the mechanism

for the formation of aniline and secondary products in the reaction of nitrobenzene

hydrogenation. Thus, experiments were done in batch mode, where the influence of NB

initial concentration, temperature and pressure were studied in detail. The liquid phase

analysis confirmed the presence of the following products: NB, ANL, CHA, CHOL,

CHONA, CHANIL, DICHA, CHENO and Bz – cf. Nomenclature section. Relatively to

the carbon mass balance, all the compounds identified allowed to close the balance within

the analytical uncertainty (± 6 %).

Some tests were performed in the same operating conditions (10 wt. % NB, 14 barg

and 150 ºC) that allowed the confirmation of the reproducibility of the experiments (data

not shown). It was also concluded that the catalyst is stable during the runs made and

within the conditions studied.

The comparison between tests B4 and B12 as indicated in Figure A.2 of Appendix

A, clearly confirms that the formation of secondary products, i.e. sum of all species other

than ANL, only begins to occur in a considerable extension after the significant decrease

of NB concentration in the reaction mixture, ca. 90-100 min, although both NB and ANL

were initially loaded into the reactor (run B4). On the other hand, if there is initially only

ANL (run B12) in the reactor, the formation of secondary products starts from the

beginning of the reaction. So, when the NB is also present in the reactor, this formation

will be delayed and only happens when NB concentration becomes very small, which

means that this catalyst is very selective for the NB hydrogenation.

3.3.1 Influence of initial nitrobenzene concentration

Tests B4 (10 wt. % NB), B7 (30 wt. % NB) and B11 (3 wt. % NB) were realized at

the same operating conditions, 150 ºC and 14 barg, and only the NB concentration was

varied (cf. Table 3.2 and Figure 3.3 to Figure 3.5). Comparing the three tests, it is

observed specially in the case of CHONA (Figure 3.3d)), that in the beginning of the

reaction the concentration is higher when the NB concentration is 30 wt. %, unlike what


70

is observed for the other compounds where the concentration is very similar even for high

NB concentration values.

a)

b)

c)

d)

e)

f)

Figure 3.3– Influence of initial nitrobenzene concentration in the secondary products formation

(Bz, CHA, CHOL, CHONA, NB and DICHA) vs. time, runs B4, B7 and B11.

0 50 100 150 200 250 300 350

0.0

6.0x10-5

1.2x10-4

1.8x10-4

Ci /

C re

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

Bz

0 50 100 150 200 250 300 350

0.0

1.0x10-3

2.0x10-3

3.0x10-3

4.0x10-3

Ci /

Cre

ftime (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHA

0 50 100 150 200 250 300 350

0.0

5.0x10-4

1.0x10-3

1.5x10-3

2.0x10-3

2.5x10-3

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHOL

0 50 100 150 200 250 300 350

0.0

3.0x10-3

6.0x10-3

9.0x10-3

1.2x10-2

1.5x10-2

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHONA

0 50 100 150 200 250 300 3500.0

0.5

1.0

1.5

2.0

2.5

3.0

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

NB

0 50 100 150 200 250 300 350

0.0

4.0x10-4

8.0x10-4

1.2x10-3

1.6x10-3

2.0x10-3

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

DICHA



71

a)

b)

Figure 3.4– Influence of initial nitrobenzene concentration in the secondary products formation

(CHENO and CHANIL) vs. time, runs B4, B7 and B11.

According to the mechanism shown in Figure 3.2, CHA, CHANIL and DICHA, are

formed from ANL hydrogenation, which was also observed by Nagata et al. [23]

(although they did not identify the formation of DICHA). The results obtained are thus in

agreement with Relvas’s mechanism; for short reaction times, their concentration is small

but when the NB concentration starts to decrease along each experiment, which means an

increase of ANL concentration and high NB conversion (see also Figure 3.2), their

concentration also increases; this is particularly noticeable for CHA and CHANIL, as

shown in Figure 3.3 and 3.4, because DICHA is a final hydrogenation product. CHENO

shows a similar trend, but its formation is anticipated by another route – cf. Figure 3.2.

As indicated in Figure 3.5a), in the beginning of the reaction the ANL concentration

decreases due to the entrance of NB in the reactor. Then, NB starts to convert mainly into

ANL and consequently its concentration increases steadily, being the hydrogenation of

NB the preferential reaction (thus putting into evidence the selectivity of the catalyst used

in this reaction). In fact, ANL hydrogenation is verified mostly when NB concentration

inside the reactor is quite small. Although this is not very clear in these tests, because the

runs were stopped before the hydrogenation of ANL becomes the preferential reaction, in

other tests a subsequent decrease of ANL concentration is clearly seen (e.g. Figure 3.11e)

at higher temperatures).

0 50 100 150 200 250 300 350

0.0

3.0x10-3

6.0x10-3

9.0x10-3

1.2x10-2

1.5x10-2

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHENO

0 50 100 150 200 250 300 350

0.0

2.0x10-3

4.0x10-3

6.0x10-3

8.0x10-3

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHANIL


72

a)

b)

Figure 3.5 – Influence of initial nitrobenzene concentration in the ANL formation a) and NB

conversion b) vs. time, runs B4, B7 and B11.

Narayanan and Unnikrishnan [24] observed under vapor phase conditions the

formation of cyclohexane (CH) and ammonia (NH3) through the hydrogenolysis of CHA

at 250 ºC. In the current case, CHA conversion was not observed in some additional

studies that were done (data not shown for brevity), and CH was not detected. This could

mean that CH and DICHA are not formed from CHA conversion as it was proposed by

those authors, but from ANL hydrogenation as it is proposed by Nagata et al. [23] and

Relvas [25]. Nevertheless, Bz was herein identified and quantified. To the authors’

knowledge, the quantification of Bz obtained as a by-product is not referred nor presented

in any other previous studies. The presence of Bz and also of NH3, which was detected

during samples collection, could be related to the hydrodenitrogenation of ANL.

Figure 3.6 demonstrates the time evolution of the total of secondary products

formed, i.e., excluding ANL, for the three runs, as well as the NB concentration.

Analyzing this figure, it is observed that the total of secondary products exhibits in

general a greater increase when the NB concentration starts decreasing significantly, and

so it can be concluded that their formation is mostly due to the ANL hydrogenation

reaction. This can be also related with the adsorption affinity, because NB adsorbs much

easier on the catalyst surface than ANL. Moreover, it is seen that in general the higher the

initial concentration of NB, the higher is the formation of secondary products, although

the maximum in their concentration is shifted towards longer reaction times. The reason

is that a longer reaction time is required to reduce the NB to a level for increasingly initial

loads of this reactant in the reactor.

0 50 100 150 200 250 300 3506

7

8

9

10

Ci /

Cre

f

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB

ANL

0 50 100 150 200 250 300 3500

20

40

60

80

100

NB

co

nv

ersi

on

(%

)

time (min)

3 wt.% NB

10 wt.% NB

30 wt.% NB



73

Figure 3.6 – Comparison between total secondary products formation (closed symbols) and NB

consumption (open symbols) as a function of reaction time for different initial NB concentrations;

runs B4, B7 and B11.

In the results of NB consumption that are illustrated in Figure 3.6 and as will be

shown below, it is noteworthy that the maximum expectable value of its concentration is

never reached because the compound is not inside the reactor at the initial instant as

detailed in the experimental / Appendix A section). Consequently, when it is loaded into

the reactor, there is immediately a consumption of NB and so the maximum expectable

concentration is not achieved, e.g. CNB/Cref of 3 for run B7 is not reached, neither CNB/Cref

of 1 for run B4, etc.

Figure 3.7 and 3.8 present the same results as the ones in Figure 3.3 and 3.4, but in

this case the species concentrations are represented as a function of the dimensionless NB

concentration. Thus, it is possible to confirm and better understand the discussion above

and also to compare the results at the same level of NB concentration. The results should

be analyzed taking into account that the evolution of the reaction time is done from the

right to the left, which corresponds to the decrease of NB concentration. Consequently, it

is observed that the secondary products are formed in greater quantities when the NB

concentration is very low. Because ANL is also initially loaded into the reactor (see

Appendix A), this means that the Pd/Al2O3 catalyst is selective towards NB

hydrogenation, but the undesired reactions become more significant when the NB

concentration is very low. Moreover, it is noted that some of the compounds are already

present in the reactor at the start of the reaction, which is due to the fact that the ANL

used, namely, industrial grade ANL, contains already traces of secondary products. It is

0 50 100 150 200 250 300 3500.0

0.5

1.0

1.5

2.0

2.5

3.0 3 wt.% NB 10 wt.% NB 30 wt.% NB

3 wt.% NB 10 wt.% NB 30 wt.% NB

time (min)

CN

B /

Cre

f

0.0

1.0x10-2

2.0x10-2

3.0x10-2

4.0x10-2

C to

tal

seco

ndar

y p

roduct

s / C

ref


74

finally noted that, in general, the amount of secondary products formed is increased for

higher initial NB concentrations (run B7–30 wt. % NB > run B4–10 wt. % NB > run

B11–3 wt. % NB).

a)

b)

c)

d)

e)

f)

Figure 3.7 - Influence of nitrobenzene initial concentration in the secondary products formation

for NB dimensionless concentration, runs B4, B7 and B11.

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

3.0x10-5

6.0x10-5

9.0x10-5

1.2x10-4

1.5x10-4

1.8x10-4

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

Bz

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

1.0x10-3

2.0x10-3

3.0x10-3

4.0x10-3

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHA

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

5.0x10-4

1.0x10-3

1.5x10-3

2.0x10-3

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHOL

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

3.0x10-3

6.0x10-3

9.0x10-3

1.2x10-2

1.5x10-2

CNB

/ Cref

Ci /

Cre

f

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHONA

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

2.0x10-4

4.0x10-4

6.0x10-4

8.0x10-4

1.0x10-3

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

DICHA

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

3.0x10-3

6.0x10-3

9.0x10-3

1.2x10-2

1.5x10-2

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHENO



75

a)

b)

Figure 3.8 - Influence of nitrobenzene initial concentration in the secondary products formation

for NB dimensionless concentration, runs B4, B7 and B11.

3.3.2 Influence of Pressure

The tests in this section were performed at the same temperature, 150 ºC, and with

the same initial concentration of nitrobenzene, 10 wt. % NB. The total pressure in the

reactor was increased from 14 barg (run B4) by 6 barg (20 barg in run B2) and 16 barg

(30 barg in run B3). Pressure variation is due to the increase of hydrogen pressure; so

there will be more hydrogen in the gas phase and also solubilized in the reaction mixture.

In Figure 3.9 are presented the results obtained as a function of the reaction time. It

is noticed that Bz formation is almost independent of the total pressure (Figure 3.9a)). For

the other secondary products, it seems that for higher pressures, higher quantities of these

compounds are obtained, although this effect varies from product to product. Even so, the

effect of pressure on the total formation of secondary products is not very pronounced

(Figure 3.9b)).

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

2.0x10-3

4.0x10-3

6.0x10-3

8.0x10-3

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

CHANIL

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0

1.0x10-2

2.0x10-2

3.0x10-2

4.0x10-2

Ci /

Cre

f

CNB

/ Cref

3 wt.% NB

10 wt.% NB

30 wt.% NB

Csecondary products


76

a)

b)

c)

d)

e)

f)

Figure 3.9 - Influence of reaction pressure in the secondary products (Bz, CHA, CHOL,

CHONA, ANL, DICHA, CHENO and CHANIL) vs. time, runs B2, B3 and B4.

Since these by-products are mostly formed after ANL formation/NB complete

consumption, Figure 3.9, it is concluded that the results obtained are in agreement with

0 50 100 150

0.0

2.0x10-5

4.0x10-5

6.0x10-5

8.0x10-5

1.0x10-4

Ci /

C re

f

time (min)

14 barg

20 barg

30 barg

Bz

0 50 100 150

0.0

2.0x10-3

4.0x10-3

6.0x10-3

8.0x10-3

1.0x10-2

1.2x10-2

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

CHA

0 50 100 150

0.0

4.0x10-4

8.0x10-4

1.2x10-3

1.6x10-3

2.0x10-3

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

CHOL

0 50 100 150

0.0

3.0x10-3

6.0x10-3

9.0x10-3

1.2x10-2

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

CHONA

0 50 100 1508.0

8.4

8.8

9.2

9.6

10.0

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

ANL

0 50 100 150

0.0

4.0x10-4

8.0x10-4

1.2x10-3

1.6x10-3

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

DICHA

0 50 100 150

0.0

4.0x10-3

8.0x10-3

1.2x10-2

1.6x10-2

2.0x10-2

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

CHENO

0 50 100 150

0.0

1.0x10-2

2.0x10-2

3.0x10-2

Ci /

Cre

f

time (min)

14 barg

20 barg

30 barg

CHANIL



77

Relvas [24] mechanism. Moreover, at the same reaction time, when the pressure is higher,

NB concentration is lower and, consequently, the ANL concentration rises. Therefore,

ANL hydrogenation will be favored and the concentration of the secondary products will

increase sooner, as indicated in Figure 3.10. Although for higher pressures it seems that

more secondary products are formed, the effect of an increase in pressure from 20 to 30

barg is less significant.

a)

b)

Figure 3.10 – Comparison between a) ANL formation and b) total of secondary products

formation (closed symbols) and NB consumption (open symbols) as a function of reaction time;

runs B2, B4 and B5.

The influence of pressure is also visible in the reaction rate; when the total pressure

in the reactor rises, there is also an increase in the reaction rate and consequently, the NB

is consumed in a shorter period of time and the ANL concentration will increase more

rapidly.

Figure A.3 of Appendix A presents the same data but as a function of the

dimensionless NB concentration.

3.3.3 Influence of Temperature

The influence of temperature – runs B4 (150 ºC), B9 (180 ºC), B10 (210 ºC) and B5

(240 ºC) – was also studied at the same pressure, 14 barg, and initial NB concentration,

10 wt.% NB. It was concluded that temperature is the parameter with greater influence in

the production of secondary products, as shown below (Figure 3.11 to Figure 3. And

Figure A.4 of Appendix A).

0 50 100 1500.0

0.2

0.4

0.6

0.8

1.0 14 barg 20 barg 30 barg

14 barg 20 barg 30 barg

time (min)

CN

B /

Cre

f

8.0

8.4

8.8

9.2

9.6

10.0

C A

NL /

Cre

f

0 50 100 1500.0

0.2

0.4

0.6

0.8

1.0 14 barg 20 barg 30 barg

14 barg 20 barg 30 barg

time (min)

CN

B /

Cre

f

0.0

2.0x10-2

4.0x10-2

6.0x10-2

8.0x10-2

C to

tal

secondary

pro

ducts /

Cre

f


78

Figures 3.11 and 3.12 presents the results obtained in these experiments, showing

the concentration of by-products and ANL as a function of the reaction time.

Concentration profiles show the same trend for all the tests; indeed, it is possible to see

that for higher temperatures the concentration of secondary products increases in the

order 240 ºC > 210 ºC > 180 ºC > 150 ºC.

a)

b)

c)

d)

e)

f)

Figure 3.11 - Influence of reaction temperature in the ANL and by-products formation (Bz, CHA,

CHOL, CHONA and DICHA) vs. reaction time, runs B4, B5, B9 and B10.

0 50 100 150

0.0

1.0x10-4

2.0x10-4

3.0x10-4

4.0x10-4

Ci /

C re

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

Bz

0 50 100 150

0.0

2.0x10-2

4.0x10-2

6.0x10-2

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

CHA

0 50 100 150

0.0

2.0x10-3

4.0x10-3

6.0x10-3

8.0x10-3

1.0x10-2

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

CHOL

0 50 100 150

0.0

1.0x10-2

2.0x10-2

3.0x10-2

4.0x10-2

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

CHONA

0 50 100 1508.0

8.4

8.8

9.2

9.6

10.0

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

ANL

0 50 100 150

0.0

1.0x10-3

2.0x10-3

3.0x10-3

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

DICHA



79

a)

b)

Figure 3.12 - Influence of reaction temperature in the ANL and by-products formation (CHENO

and CHANIL) vs. reaction time, runs B4, B5, B9 and B10.

Bz belongs to the compounds that seem to be more influenced by temperature

variation. In the temperature range of 150 – 210 ºC, the concentration profiles obtained

are similar to those shown above. While at lower temperatures Bz is detected only in

small amounts, at 240 ºC, when the dimensionless NB concentration is < 0.012 (after ca.

1 h), its concentration starts to increase considerably (Figure 3.11a)) which had not been

observed before. Some tests of ANL hydrogenation, which were performed at different

operational conditions (data not shown), confirmed the Bz formation from ANL and that

its concentration is higher when the conditions are more severe, namely for temperature;

see also Figure A.4 of the Appendix A.

In the case of CHA, DICHA and CHANIL, it is shown that when temperature rises,

there is an increase in their concentration when NB concentration is low (high

conversion). As it was observed in the tests with pressure variation, in the first hour of

reaction the concentration profile is described by a plateau. This plateau corresponds to

the period when NB is present in high quantities and when the conversion of NB into

ANL mainly takes place. After that period, although the NB conversion still proceeds,

ANL hydrogenation becomes more pronounced and, consequently, the concentration of

secondary products increases. This reaction, ANL hydrogenation, starts to be more

important with higher temperature.

Both CHONA and CHENO generally present concentration profiles similar to the

compounds discussed above, but at the end of the reaction their concentration seems to

start to stabilize, reaching a plateau as indicated in Figure 3.11d) and Figure 3.12a). For

CHOL similar profiles are anticipated; however, longer reaction times are probably

required to observe such a pattern.

0 50 100 150

0.00

1.50x10-2

3.00x10-2

4.50x10-2

6.00x10-2

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

CHENO

0 50 100 150

0.0

2.0x10-2

4.0x10-2

6.0x10-2

8.0x10-2

1.0x10-1

Ci /

Cre

f

time (min)

150 ºC

180 ºC

210 ºC

240 ºC

CHANIL


80

In relation to NB, there is, effectively, an important growth of the rate of the

hydrogenation reaction with the temperature increase, which means that for higher

temperatures the same NB conversion is achieved in much less time, as shown in Figure

3.13.

a)

b)

Figure 3.13 – Comparison between a) ANL formation and b) total secondary products formation

(closed symbols) and NB consumption (open symbols) as a function of reaction time for different

reaction temperatures; runs B4, B5, B9 and B10.

Once more is noticed that the concentration of secondary products is strongly

dependent on the temperature of the reaction (Figure 3.13b)). Indeed, the formation of

secondary products is tremendously accelerated as a consequence of the exponential

temperature effect in the kinetics according to the Arrhenius law. The different activation

energies for each reaction in the complex reaction scheme also explain the various effects

of temperature in each product formed. For those steps with higher activation energies,

reaction rates are more favored for the same increase in the temperature.

Narayanan and Unnikrishnan [24] also studied the influence of temperature on a

Co/Al2O3 and a Ni/Al2O3 catalyst and concluded that with higher temperature the ANL

conversion increases over Co/Al2O3 but over Ni/Al2O3 the increase in temperature

reduced the conversion, due to the decrease of the hydrogen adsorption. Besides, they

also observed that above a certain temperature, hydrodenitrogenation is favored leading to

the formation of CH. In the current case, instead of CH formation, the formation of Bz

was observed and the same tendency was noticed, namely, a higher temperature favors

the Bz formation, Figure 3.12a)). Relatively to the catalyst herein used, its behavior was

similar to the Co/Al2O3 one; with higher temperatures the ANL conversion increased.

0 50 100 1500.0

0.2

0.4

0.6

0.8

1.0 150 barg 180 ºC 210 ºC 240 ºC

150 barg 180 ºC 210 ºC 240 ºC

time (min)

CN

B /

Cre

f

8.0

8.4

8.8

9.2

9.6

10.0

C A

NL /

Cre

f

0 50 100 1500.0

0.2

0.4

0.6

0.8

1.0 150 barg 180 ºC 210 ºC 240 ºC

150 barg 180 ºC 210 ºC 240 ºC

time (min)

CN

B /

Cre

f

0.0

5.0x10-2

1.0x10-1

1.5x10-1

2.0x10-1

2.5x10-1

3.0x10-1

C to

tal

seco

nd

ary

pro

du

cts /

Cre

f



81

Analyzing all the results obtained, it can be stated that the data obtained are in

general agreement with the results described by Relvas [25], and all the compounds

mentioned were identified in our experiments. Furthermore, even using a different type of

catalyst, Pd in this work vs. Ni in Relvas work, and different experimental conditions, the

mechanism was found to be very similar. Another important issue which is worth

mentioning in the results herein presented, is the formation of Bz, which is not considered

by Relvas nor by any other authors in the literature, except Králik et al. [26]; however,

the authors did not quantify Bz. Consequently, and reflecting the results obtained, the Bz

formation should be considered in the mechanism through the reaction of ANL

hydrodenitrogenation, as detailed in Figure 3.14.

Figure 3.14 – Reaction network proposed for ANL and secondary products formation including

Bz (*very reactive and unstable compounds).

NO2

Cat

+H2

NH2

+ NH3

H2

Cat

H2

+ANL

-NH3

Cat

H2

NH2

NHNH2

+H2 -NH3

NH

+H2

NH

O

+H2

-NH3

+ANL -H2O

N

+H2

+H2OH CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

Bz

ANLNB

Amine


-H2O

-H20

*


82

3.4 Conclusions

The hydrogenation of NB to ANL over a Pd catalyst was studied. A parametric

study was performed in which the effects of the initial NB concentration, temperature and

pressure were tested. It was found that temperature is the parameter with greater influence

in the production of secondary products, for the ranges analyzed.

It was also observed that the formation of secondary products is strongly dependent

on the concentration of NB in the reaction mixture: their concentration rises more when

the NB concentration starts decreasing significantly, or becomes almost negligible, and

consequently it can be concluded that the formation of secondary compounds is mostly

due to the ANL hydrogenation. This highlights the high selectivity of the used Pd-based

catalyst.

Some mechanisms were already proposed in the literature; nevertheless, the one

proposed by Relvas, using a Ni/SiO2 catalyst, is the one that better describes the majority

of the experimental results herein obtained for both ANL and formation of secondary

products, despite the fact that a Pd-based catalyst was now used. Furthermore, the

formation of Bz was also identified, which was not considered in a quantitative manner

by any other previous authors in the literature.



83

References

[1] R. Lawrence, W.J. Marshall. Aniline. Ullmann's Encyclopedia of Industrial

Chemistry, 6th Edition (Print), John Wiley & Sons, New York, 1998.

[2] Relvas, J., Andrade, R., Gama Freire, F., Lemos, F., Araújo, P., Pinho, M., Pedro

Nunes, C. Ramôa Ribeiro, F., Liquid phase hydrogenation of nitrobenzene over an

industrial Ni/SiO2 supported catalyst, Catalysis Today 133-135 (2008) 828-835.

[3] Burge, H., Collins, D., Burtron H. Davis, Intermediates in the Raney Nickel Catalyzed

Hydrogenation of Nitrobenzene to Aniline, Industrial & Engineering Chemistry Product

Research and Development 19 (1980) 389-391

[4] Collins, D., Smith, A., Davis, B., Hydrogenation of Nitrobenzene over a Nickel Boride

Catalyst, Industrial & Engineering Chemistry Product Research and Development 21

(1982) 279-281.




[6] Yu, X., Wang, M., Li, H., Study on the nitrobenzene hydrogenation over a Pd-B/SiO2

amorphous catalyst, Applied Catalysis A: General 202 (2002) 17-22.

[7] Klemm, E., Amon, B., Redlingshöfer, H., Dieterich, E., Emig, G., Deactivation

kinetics in the hydrogenation of nitrobenzene to aniline on the basis of a coke formation

kinetics * investigations in an isothermal catalytic wall reactor, Chemical Engineering

Science 56 (2001) 1347-1353.



[9] Rihani, D., Narayanan, T., Doraiswamy, L., Kinetics of catalytic vapor-phase

hydrogenation of nitrobenzene to aniline, Industrial & Engineering Chemistry Product

Research and Development 4 (1965) 403-410.








[12] Sangeetha, P., Seetharamulu, P., Shanthi, K., Narayanan, S., Rama Rao, K.S Studies

on Mg-Al oxide hydrotalcite supported Pd catalysts for vapor-phase hydrogenation of

nitrobenzene, Journal of Molecular Catalysis A: Chemical 273 (2007) 244-249.


84

[13] Sangeetha, P., Shanthi, K., Rama Rao, K.S., Viswanathan, B., Selvam, P.,

Hydrogenation of nitrobenzene over palladium-supported catalysts - Effect of support,

Applied Catalysis A: General 353 (2009) 160-165.

[14] Li, C., Yu, Z., Yao, K., Ji, S., Liang, J., Nitrobenzene hydrogenation with carbon

nanotube-supported platinum catalyst under mild conditions, Journal of Molecular


[15] Höller, V., Wegricht, D., Yuranov, I., Kiwi-Minsker and, L., Renken, A., Three-

phase nitrobenzene hydrogenation over supported glass fiber catalysts: kinetics study,

Chemical Engineering & Technology 23 (2000) 251-255.

[16] Harraz, F., El-Hout, S., Killa, H., Ibrahim, I., Palladium nanoparticles stabilized by

polyethylene glycol: Efficient, recyclable catalyst for hydrogenation of styrene and

nitrobenzene, Journal of Catalysis, 286 (2012) 184-192.

[17] Bouchenafa-Saïb, N., Grange, P., Verhasselt, P., Addoun, F., Dubois, V., Effect of

oxidant treatment of date pit active carbons used as Pd supports in catalytic

hydrogenation on nitrobenzene, Applied Catalysis A: General 286 (2005) 167-174.

[18] Gelder, E., Jackson, S. and Lok, C., A study of nitrobenzene hydrogenation over

palladium/carbon catalysts, Catalysis Letter 84 (2002) 205-208.

[19] Mitchell, C., Stewart, D., Process fot the production of aromatic nitro compounds

into amines, WO Patent 113491, 2011.


Z. Elektrochem., 4 (1898) 506-514.

[21] Wisniak, J., Klein, M., Reduction of Nitrobenzene to Aniline, Industrial &

Engineering Chemistry Product Research and Development 23 (1984) 44-50.













Engineering 6 (2012) 1074-1082.

http://pubs.acs.org/action/doSearch?action=search&author=Wisniak%2C+Jaime&qsSearchArea=author

85

Chapter 4 – Study of Effects of the Solvent and

Reaction Products in the Catalytic Hydrogenation

of Nitrobenzene.

Abstract

The effect of the type of solvent and the presence of reaction products in the

reaction mixture on the nitrobenzene (NB) hydrogenation towards aniline (ANL) was

studied over a commercial catalyst, Pd-based (1 wt.% Pd/Al2O3). Hydrogenation of ANL

and cyclohexylamine (CHA, a main sub-product) were also analysed. All the catalytic

tests were carried out in a stired tank basket reactor either on batch or on continuous

mode, to reproduce laboratorial and industrial conditions, respectively. It was found that

the use of ANL + para-toluidine (p-tol) leads to a decrease of secondary products

formation but NB conversion is also lower, when compared with the use of ANL alone as

solvent. The presence of H2O in the reaction mixture was evaluated and it was concluded

that its presence decreases the ANL formation, and particularly heavy by-products

formation. Influence of benzene (Bz) was analyzed, when co-feeded in the liquid stream,

being shown that it leads to a slight decrease in NB conversion. Moreover, direct

hydrogenation of ANL using the Pd-based catalyst allowed to conclude that an increase in

operating conditions severity (namely in terms of pressure and temperature) leads to a

higher formation of secondary products, mainly CHA, dicyclohexylamine (DICHA) and

cyclohexylaniline (CHANIL). Hypothesis of DICHA being formed via condensation of

two molecules of CHA was not confirmed and it was verified that CHA, by itself, is not a

precursor of any secondary product of NB/ANL hydrogenation reaction. All these tests

together were quite useful to provide further insights about NB hydrogenation

mechanism.


86

4.1 Introduction

Aniline (ANL) has been a major chemical product for many decades, being used as

intermediate in the production of polyurethanes, dyes, herbicides, fungicides, rubber

chemicals, pharmaceuticals, among others. ANL can be produced by several reactions;

however, the nitrobenzene (NB) route continues to dominate and gives the highest

selectivity [1].

NB hydrogenation into ANL is a highly exothermic catalytic reaction, ΔH = -544

kJ/mol (at 200 ºC), and can be performed both in the vapor and in the liquid phase in

commercial processes [2]. Industrial use of vapor-phase hydrogenation is limited by the

thermal stability of the nitro compounds. Most used catalysts are Cu/SiO2 and Pd/Al2O3

[3 - 5]. According to Vogt and Gerulius [3], it is preferable to hydrogenate most aromatic

nitro compounds in the liquid-phase, since in this case pressure and temperature can be

changed independently. Nevertheless, temperature is limited by the hydrogenation

reaction of the aromatic ring, which occurs above 170 – 200 ºC, so normally the reduction

is carried out at 100 – 170 ºC. In the case of pressure, values between 1 – 50 barg are the

most used. Typical catalysts used in liquid-phase hydrogenation are Ni, Pd and Pt

supported catalysts [6, 7].

In the NB hydrogenation reaction there is the production of ANL, unstable

intermeadiary species and secondary products. In fact, some authors have studied this

reaction with the objective of describing the mechanism and also of knowning all the

compounds envolved. In the case of intermediaries species, Haber [8] proposed a

mechanism for ANL formation where nitrosobenzene (NSB) and arylhydroxylamine

(PHA) are formed. Haber claimed that the reaction occurs through a direct route and a

condensation route, Figure 4.1 a). Gelder et al. [9], using a Pd/C catalyst, suggested that

the number of steps involved in ANL formation was higher and substantially different

from those previously reported. Other authors studied intermediate products formation

when hydrogenating NB into ANL, namely Corma et al. (Au/TiO2 catalyst) [10],

Makosch et al. (Au/MeOx catalyst) [11], Chatterjee et al. (Pd/MCM-41 catalyst) [12],

Rakitin et al. (Pd-based catalysts) [13] and Turáková et al. (Pd on carbon catalyst) [14],

the latter shown in Figure 4.1 b).

Chapter 4 – Study of Solvent and Presence of Reaction Products Effects

in the Catalytic Hydrogenation of Nitrobenzene.

87

a)

b)

Figure 4.1 - Reaction network involved in nitrobenzene hydrogenation illustrating intermediary

species proposed by a) Haber [8] and b) Turáková et al. [14].

Relatively to secondary products formation, Nagata et al. [15], using a Pd and

Pd+Pt supported on carbon catalysts, and later on Narayanan et al. [16, 17], with catalysts

of Ni and Co supported on Al2O3, proposed a mechanism to explain their formation. In

2008, Relvas [18] using a Ni-based catalyst proposed a mechanism, based on the Nagata

one, where dicyclohexylamine (DICHA) and cyclohexanol (CHOL) - through

cyclohexanone (CHONA) hydrogenation - were included, Figure 4.2. Králik et al. [19]

Ar - NO2

nitro / NB

Ar - NO

nitroso / NSB

Ar - NHOH

arylhydroxylamine / PHA

Ar - NH2

ANL

Ar - NO = N - Ar

Ar - NHOH

Ar - N = N - Ar Ar - NH = NH - Ar

hydrazo / HZBazo / AZBazoxy / AZXB

Direct route

Condensation route

NO2

NB NO

NOHNHOH

NH2

NN

O

NOH

NN

NH NOH

NH2

NH2

NH

+

NSB

PHA

ANL

AZXB

AZB

HZB

NH NH

ANL

ANL


88

observed more recently that cyclohexilamine (CHA) and DICHA are the main products

and cyclohexane (CH) is also detected, over the Pd, Pt and Rh catalysts on carbon, while

Rubio-Marqués et al. [20] showed that ANL is rapidly formed as a primary and unstable

product that, by partial hydrogenation, gives cyclohexylaniline (CHANIL) as a secondary

product, when using Pd/C. By further hydrogenation it generates DICHA in minor

amounts, as a tertiary product. Nevertheless, CHA was not observed among the reaction

products, and since the hydrogenation of ANL is very likely to occur, a possible way to

form CHANIL would be through the reaction between ANL and CHA. In 2015, Sousa

[21] using a Ni-based catalyst suggested that H2O creates two different scenarios. When

H2O is not present, formation of CHA, CHANIL, n-cyclohexylideneaniline (CHENO)

and DICHA is favored while no significant amounts of CHONA and CHOL are detected.

On the other hand, when H2O is present, CHONA and CHOL increase considerably but

the amount of all the other compounds is reduced; the proposed mechanism is shown in

Figure 4.3.

Figure 4.2 - Reaction network involved in nitrobenzene hydrogenation illustrating secondary

products formation proposed by Relvas [18].

Cat

+H2,-H20

Cat

H2

+ANL

-NH3

Cat

H2

+H2 -NH3

+H2

+H2

-NH3, -H2O

+ANL -H2O

+H2

+H2 CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

ANLNB

Amine


N

O

OH

NO2 NH2

NH2

NHNH2

NH

NH



89

Figure 4.3 - Reaction network involved in nitrobenzene hydrogenation illustrating secondary

products formation proposed by Sousa [21].

NB hydrogenation into ANL involves more steps and compounds than what was

initially thought, therefore it is clear that the comprehension regarding the formation of

these secondary products is extremely important, which may vary with the operating

conditions and catalyst employed. Moreover, the appearance of some compounds favors

the formation of other secondary products. It is known that industrially, ANL is used as

solvent [22] and it might be recirculated into the reactor, meaning that secondary products

can also be fed into the reactors. In this concern, Mohan et al. [23] analysed the effect of

co-feeding H2O with NB over a Ni/MgO and a Ni/SBA-15 (Santa Barbara amorphous

silica) catalyst, concluding that H2O might affect the catalyst support, having influence in

the catalytic performance. According to various authors [16, 17, 19], CHA is the main

product resulting from ANL hydrogenation. In this way, influence of CHA is another

important question that should be analysed.

Besides the operating conditions, particularly temperature and pressure, and

presence of reaction products in the reactor feed, the solvent used has also a great impact

in the catalyst performance and in the secondary products formation/distribution. The

hydrogenation of NB is highly exothermic, so when using a solvent, the reaction can be

performed in more stable conditions and it also avoids the formation of two phases during

the reaction, which may lead the reaction to stop. A large number of solvents can be used

in this reaction, each one with its own advantages and disadvantages, being the selection

traditionally based on solvent performance/effect in the reaction. Figueras and Coq [6]

NO2 NH2

- 2 H2O

+ 3 H2

N

CHENO

NBANL

CHANIL

NH

NH2

- NH3

NH2 CHA

+ H2

+ 3 H2

DICHA

NH

NH2

- NH3

+

+

NH2 + H2O

- NH3-

+ 3 H2

O OH

+ H2

CHONA CHOL


90

refer that the solubility of hydrogen depends on the solvent, so that volatile apolar

solvents dissolve hydrogen better.

In this work the influence of the presence of reaction products in the reactor feed

and the nature of the solvent in NB hydrogenatin towards ANL will be studied over a

commercial Pd-based catalyst. The main goal is to try to understand how those

parameters influence the reaction as well as the formation of secondary products.



reactor provided with an air-impelled stirrer to promote appropriate mixing and gas

distribution during the experiments. The stirrer was equipped with a basket, where the

catalyst was placed. Catalyst used was a commercial 1 wt.% Pd/Al2O3 material in

spherical form.

A known amount of the catalyst was loaded into the reactor, Figure 4.3, and the

material pre-treated, in situ. Pre-treatment of the catalyst was firstly performed, at 150 ºC

and under hydrogen pressure (20 barg) for 2 hours. Several temperature-programmed

reduction (TPR) experiments were performed with the fresh catalyst and the results have

shown that it was fully reduced, under the pre-treatment conditions employed (data not

shown).

After this activation step, a certain volume of aniline (industrial grade) was loaded

into the reactor, as is commonly done in the industry, with two main goals:

i) to act as a solvent for the water that is produced during the reaction, in order to

avoid the formation of two phases (organic and aqueous) that would lead to the

interruption of the reaction, and also to avoid strong NB adsorption;

ii) to help to dissipate the excess heat generated due to the high exothermicity of the

reaction.

The reactor was enclosed in an electric furnace regulated by a temperature

controller (SHIMADEN SD20) and the initial temperature was established. The heat

produced by the nitrobenzene hydrogenation was removed by a cooling water stream

whose flow rate was controlled with a set of ball valves, as shown in Figure 4.4. The



91

reactor temperature was constant with a maximum ΔT of 4 - 5 % and it was continuously

measured throughout the experiments.

Figure 4.4 – Scheme of the reactor and set-up used in the experiments.

NB was charged in a vessel and subsequently was loaded into the reactor by

pushing it using a high-pressure hydrogen stream (cf. Figure 4.4). The NB was loaded

into the reactor instantaneously, when the desired reaction temperature was achieved

(time = 0 min). This procedure was adopted to ensure that the NB hydrogenation does not

start before reaching the desired temperature (beginning of the experiment) and also to

avoid any strong NB adsorption on the catalyst, blocking accessibility to active sites; this

is also ensured by using ANL as solvent. In batch mode, all the experiments were done up

to a nearly complete consumption of NB (which was considered to correspond to the

instant at which NB concentration was below 1000 ppm).

The sampling of the liquid phase was performed at selected time intervals and

analyzed by gas chromatography, in an Agilent 6890A chromatograph equipped with two

flame ionization detectors (FID). The column used was a HP-1 one (100%

dimethylpolysiloxane 30 m x 320 µm x 4 µm). The temperature in the injector and in the

detector was 250 ºC, the pressure in the column was 14 bar and the carrier gas used was

helium. The column oven was temperature-programmed with a 1 min initial hold at 120

ºC, followed by an increase of temperature until 230 ºC at a rate of 15 ºC min-1 and then

kept at 230 ºC for 9 min.

H2

H2V1

V3

V4N2

V6

V5

Vessel

V7

reactor

Samples

V14

V11 V12

V9

V8

H2OPI4

PI2

V2

PI1 PI3

V10

V13


92

All the compounds were previously identified using the external standard method.

Calibration curves were plotted for all the compounds to be analyzed, which were easily

identified since their retention times are well known. Several samples were injected and

the standard deviation associated with this method was found to be below 10%.

The reference values for temperature, pressure and nitrobenzene concentration are T

= 150 ºC, P = 14 barg and Cref = 10 wt. % NB (100 000 ppm), respectively. The

experiments performed and the conditions used are given in Table 4.1.


Test Catalyst Mode Experimental conditions

TC1

catalyst I.1

Continuous 150 ºC, 14 barg, ANL, 5 ml/min

TC2 Continuous 200 ºC, 20 barg, ANL, 5 ml/min

TC3 Continuous 150 ºC, 14 barg, 10 wt.% NB+ 0.7 wt.% Bz in

ANL, 5 ml/min

TC4 Continuous 150 ºC, 14 barg, 1 wt.% H2O in ANL, 5 ml/min

TC5 Continuous 150 ºC, 14 barg, CHA, 5 ml/min

TC6 Continuous 150 ºC, 14 barg, 10 wt.% NB in ANL, 5 ml/min

TB7 Batch 150 ºC, 14 barg, 10 wt.% NB in ANL

TB8 Batch 150 ºC, 14 barg, 10 wt.% NB in 62 wt.% ANL +

28 wt.% P-Tol

In Table 4.1 are presented all the catalytic testes that were carried out. Some of the

tests were performed in batch mode, others in continuous mode (CSTR). The reactor

always operates in a semi-continuous mode for the gas phase (hydrogen) and in a batch /

continuous mode relative to the liquid phase. The total pressure inside the reactor is kept

constant along each run due to the continuous admission of hydrogen to compensate what

is being consumed.

The concentration of NB is presented as a relative dimensionless value that was

calculated as follows:

𝐶𝑁𝐵 = 𝐶𝑁𝐵 (𝑝𝑝𝑚)

𝐶𝑟𝑒𝑓(𝑝𝑝𝑚) (4.1)

where CNB (ppm) is the concentration of NB (ppm) in a given instant and Cref stands for

the NB reference concentration (ppm). The secondary products will not be presented



93

individually but in groups: Light products – Benzene (Bz), Cyclohexilamine (CHA),

Cyclohexanol (CHOL) and Cyclohexanone (CHONA); and Heavy products:

Dicyclohexylamine (DICHA), N-cyclohexyldeneaniline (CHENO) and Cyclohexylaniline

(CHANIL). It is important to refer that ANL, used as solvent, is of industrial grade as

well as CHA.


Usually, in the laboratory, hydrogenation reaction is carried out in batch mode with

the purpose of studying secondary products formation as well as catalyst activity and

selectivity. However, industrially, this reaction is performed in continuous mode and in

some configurations it may exist an ANL recirculation stream, which is composed by

ANL and secondary products formed during the reaction. So, the effect of such products

will be analysed in this chapter, at least for the most critical ones / those that exist in

larger quantities.

In Chapter 3 it was done a parametric study, in batch mode, with Catalyst I.1 (1 wt.

% Pd / Al2O3). Temperature, pressure and NB concentration were varied and their

influence was analyzed. Nevertheless, another important parameter is the solvent used,

which may avoid the formation of two different phases and also interfere in the reaction

rate. Therefore, in this chapter and in order to continue that study, the solvent effect will

be also analyzed. The solvent chosen is P-toluidine (P-tol), since it is a molecule similar

to ANL. Thus, ANL and a mixture of ANL + 28 wt.% P-tol will be tested as solvent and

compared.

H2O and Bz are reaction products and the knowledge of their influence in the

hydrogenation of NB is crucial to analyse the formation of other reaction products, since

they may be present in the recycled ANL. Hydrogenation of CHA will also be tested in

order to verify if there is the formation of reaction products through CHA transformation,

thus providing further insights about reaction mechanism. In Chapter 3 it was verified that

the secondary products formation begins when NB concentration is low, which means

that ANL hydrogenation and condensation are predominant. In this way, some additional

experiments will be performed to better understand secondary products formation (during


94

ANL hydrogenation) at reference conditions and with higher temperature and pressure

using only ANL in the reactor feed.

4.3.1 Influence of the solvent

The solvent has an important influence in the hydrogenation reaction, for instance

in the hydrogen solubility. In the case of NB hydrogenation, the solvent has also the

function of avoiding the formation of two phases in the reaction mixture, and help

dissipating reaction heat. Nevertheless, the choice of the solvent should also take into

account if the compound is inert in the reaction conditions.

In this section catalytic tests were carried out in batch mode and solvent effect was

evaluated, with the purpose of analysing catalyst performance and also the formation of

secondary products.

P-toluidine (p-tol) is an aromatic amine with a chemical structure similar to ANL.

This compound is available in powder and it was decided to dissolve it in ANL. In

Figures 4.5 and 4.6 are presented the results obtained for runs TB7 with ANL and TB8

with 62 wt.% ANL + 28 wt.% P-tol. Both tests were performed at 150 ºC, 14 barg and 10

wt.% NB (cf. Table 1).

a)

b)

Figure 4.5 – Evolution of a) NB and b) ANL as a function of reaction time for different solvents -

ANL and ANL + 28 wt. % P-tol (runs TB7 and TB8).

0 20 40 60 80 100 1200

2

4

6

8

10

NB

con

centr

atio

n (

%)

time (min)

62 wt.% ANL + 28 wt.% p-tol

ANL

0 20 40 60 80 100 1200

2

4

6

8

10

AN

L f

orm

ed (

%)

time (min)


ANL



95

Figure 4.6 – Evolution of secondary products concentration as a function of reaction time for

different solvents - ANL and ANL + 28 wt. % P-tol (runs TB7 and TB8).

Analysing Figure 4.5 a) it is observed that the dimensionless NB concentration is

not 1 for t = 0 min and that it increases in the beginning of the reaction runs. This occurs

due to the fact that NB is loaded into the reactor from the external vessel according with

the procedure described above, starting then to be consumed; at the same time ANL

concentration decreases (Figure 4.5 b)) since ANL is already inside the reactor. Then,

after approximately 15 min, NB concentration starts to decay as it is being transformed

into ANL and also into secondary products. ANL used as solvent is of industrial grade,

which explains why secondary products are already present for t = 0 min.

Comparing the results for the two solvents, it is observed that, when only ANL is

used, NB consumption is slightly faster than with the mixture ANL + P-tol, being this

also demonstrated through ANL formation, Figure 4.5. In the case of secondary products

formation, it is however seen that their formation is lower for the mixture ANL + P-tol,

mainly after ca. 60 min. In Figure 4.7 are presented the results of secondary products,

divided in light and heavy products, as well as the evolution of secondary products

distribution along reaction time.

0 20 40 60 80 100 1200

400

800

1200

1600

2000

C se

con

dar

y p

rod

uct

s (p

pm

)

time (min)


ANL


96

a)

b)

c)

d)

Figure 4.7 – Evolution of the concentration of a) light products and b) heavy products, c)

secondary products with ANL as solvent and d) secondary products with ANL + 28 wt.% p-tol as

solvent, along reaction time for different solvents (runs TB7 and TB8).

It is seen that, globally, light products formation does not seem to be affected by

the solvent used, since the results are similar (Figure 4.7 a)). However, CHA formation is

lower in the run with ANL + P-tol (Figure 4.7 c) vs. Figure 4.7 d)), which might mean

that in the presence of P-tol, direct hydrogenation of ANL into CHA is not favoured (cf.

proposed reaction mechanism in Figure 4.8), which might be also a consequence of the

lower concentration of ANL when P-tol is added as co-solvent. In the case of heavy

products, their formation is higher when ANL is used as solvent, Figure 4.7 b). CHONA

formation seems to be increased with solvent ANL + P-tol while heavy products, mainly

CHENO, are increased when the solvent is ANL. Analysing Figure 4.8, one can see that

CHENO is formed through reaction of ANL with CHONA and it appears that in the

presence of P-tol this transformation is not favored, either due to the presence of the co-

solvent or to the lower concentration of ANL.

0 20 40 60 80 100 1200

200

400

600

800

1000

C li

ght

pro

du

cts (

pp

m)

time (min)


ANL

0 20 40 60 80 100 1200

200

400

600

800

1000

C h

eavy p

roduct

s (p

pm

)

time (min)


ANL

0

300

600

900

1200

1500

1800

0 5 10 15 20 30 45 60 90 120

Co

ncentr

atio

n (

ppm

)

BzCHACHOLCHONADICHACHENOCHANIL

ANL

time (min)

0

300

600

900

1200

1500

1800

0 5 10 15 20 30 45 60 90 120

Co

ncentr

atio

n (

ppm

)

BzCHACHOLCHONADICHACHENOCHANIL

62 wt.% ANL + 28 wt.% P-tol

time (min)



97

Figure 4.8 – Reaction network proposed for ANL and secondary products formation including Bz

(*very reactive and unstable compounds).

Using ANL + P-tol as solvent seems therefore to have the advantage of not

allowing the production of secondary products, in particular the heavy ones. On the other

hand, NB consumption is slightly higher in the presence of the solvent ANL than with

ANL + P-tol.

4.3.2 Influence of the presence of reaction products in the feed

As it was said, industrially NB hydrogenation reaction is carried out in continuous

mode and usually there is the recirculation of an ANL stream. Consequently, in this

section, effects resulting from the presence of several compounds in the reaction mixture

will be studied.

This study will be comprised by two parts: the 1st where secondary products are

co-fed into the reactor, simulating the recirculation stream, and the 2nd where it will be

considered that only ANL or CHA are fed into the reactor.

Heavy products

Light products

Light products

NO2

Cat

+H2

NH2

+ NH3

H2

Cat

H2

+ANL

-NH3

Cat

H2

NH2

NHNH2

+H2 -NH3

NH

+H2

NH

O

+H2

-NH3

+ANL -H2O

N

+H2

+H2OH CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

Bz

ANLNB

Amine


-H2O

-H2O

*-NH3

+H2O


98

4.3.2.1 Effect of H2O

Water is one of the by-products formed in higher quantities. In fact, for each

molecule of ANL formed, there is the formation of two H2O molecules, cf. eq. 4.2 that

illustrates conversion of NB into ANL. Moreover, H2O is also a key product in

undesirable reactions related with by-products formation, as it can be produced in the

reaction along with CHONA (eq. 4.3) or can be a reagent for CHONA formation (eq. 4.4)

– see also the reaction mechanism proposed in Figure 4.7.

𝐶6𝐻5𝑁𝑂2 + 3𝐻2 → 𝐶6𝐻5𝑁𝐻2 + 2𝐻2𝑂 (4.2)

𝐶6𝐻5𝑁𝑂2 + 5𝐻2 → 𝐶6𝐻10𝑂 + 𝑁𝐻3 +𝐻2𝑂 (4.3)

(4.4)

The presence of H2O might affect the catalyst performance in different ways. In

this case, it was decided to analyse the influence of H2O in the secondary products

formation via direct ANL hydrogenation. Thus, an important topic is the solubility of

H2O in ANL and vice-versa, as shown in Table 4.2, as that can lead to the formation of

different phases. Consequently, H2O was added to the feed mixture in a concentration that

guarantees compounds miscibility.

Table 4.2 – ANL/H2O system solubility [31]

Temperature (ºC) % (wt. / wt.) of ANL in H2O % (wt. / wt.) of H2O in ANL

25 3.5 5.0

90 6.4 9.9

The main objective was to hydrogenate ANL with and without H2O in the reactor

feed and to evaluate the influence of H2O in the ANL transformation and specially in the

formation of secondary products. This aims to provide further insight into the reaction

mechanism. In Figure 4.9 are shown the results of the tests performed at 150 ºC, 14 barg,

liquid feed flow rate of 5 ml/min with ANL in the reactor feed (run TC1) and ANL + 1

NH2 NH

Amine

+ H2O

- NH3

O



99

wt.% H2O (run TC4) – cf. Table 4.1. Experiments were carried out until steady-state has

been reached.

a)

b)

c)

d)

e)

f)

Figure 4.9 – Evolution of a) ANL concentration, b) secondary products concentration, c) light

products concentration, d) heavy products concentration, e) secondary products concentration

distribution for ANL in the reactor feed and f) secondary products concentration distribution for

ANL+ 1 wt.% H2O in the reactor feed, along reaction time (runs TC1 and TC4).

As it was expected, ANL hydrogenation occurs in the presence of the Pd-based

catalyst. However, observing Figure 4.9 a) it is seen that with H2O, ANL hydrogenation

also occurs but is less marked. If ANL hydrogenation is less pronounced with H2O, it is

evident that secondary products formation will be lower.

0 100 200 300 400 50098.0

98.5

99.0

99.5

100.0

AN

L C

once

ntr

atio

n (

%)

time (min)

ANL + 1 wt.% H2O

ANL

0 100 200 300 400 5000

1500

3000

4500

6000

7500

C se

con

dar

y p

rod

uct

s (p

pm

)

time (min)

ANL + 1 wt.% H2O

ANL

0 100 200 300 400 500

0

400

800

1200

1600

2000

C li

gh

t p

rod

uct

s (p

pm

)

time (min)

ANL + 1 wt.% H2O

ANL

0 100 200 300 400 5000

1000

2000

3000

4000

5000C

hea

vy

pro

du

cts (

ppm

)

time (min)

ANL+ 1 wt.% H2O

ANL

0

1500

3000

4500

6000

7500

Co

ncentr

atio

n (

ppm

)

Bz CHA

CHOL CHONA

DICHA CHENO

CHANIL

ANL

time (min)

0

1500

3000

4500

6000

7500

Co

ncentr

atio

n (

ppm

)

Bz CHACHOL CHONADICHA CHENO

ANL + 1 wt. % H2O

time (min)


100

According to the proposed reaction network (Figure 4.8), and since there is no NB

in the mixture, when H2O is present the equilibrium affecting the intermediate Amine is

shifted towards CHONA formation, eq. 4.4. In this case, the equilibrium

CHENO/CHONA is also shifted towards the CHONA side (Figure 4.8). Consequently,

CHONA concentration should increase. In fact, that is observed, when comparing both

tests.

Among all the by-products, only CHONA (and CHOL) present higher

concentrations when H2O is present in the reactor feed, Figure 4.9 e) and f). On the other

hand, CHA concentration is lower in the presence of H2O, what is natural since ANL

hydrogenation is not favoured in those conditions and the Amine will more easily yield

CHONA. Analysing light products distribution, it is seen that they are similar with or

without the presence of water; however, heavy products curves do not present the same

tendency: in the presence of H2O, formation of those compounds is clearly inhibited,

Figure 4.9 d) – CHENO presents no major differences, while mostly CHANIL and also

DICHA concentrations decrease in the presence of H2O.

The effect of co-feeding H2O with NB over a Ni/MgO and a Ni/SBA-15 (Santa

Barbara amorphous silica) catalyst was also analysed by Mohan et al. [23]. It was

observed that NB conversion has a drastic decrease when is used the Ni/MgO catalyst,

due to the poisoning effect of H2O, while NB conversion over Ni/SBA-15 was high and

steady. Figueras and Coq [6] studied over a Pd/C catalyst, concluding that H2O does not

seem to have any effect over the catalyst due to its hydrophobicity.

Apparently, the presence of H2O does not affect reaction rate or products

formation directly but it might have an influence in the catalyst support. Consequently,

when working with higher quantities of H2O special attention must be paid to the choice

of the support to be used (possible more critical than the active phase), as it will affect the

activity of the catalyst [3]. In this study, it was demonstrated that the presence of H2O

helps avoiding ANL hydrogenation, mostly preventing the heavy products formation,

particularly CHANIL. This is an important information since industrially H2O can be

present in the reaction feed, eventually in still higher concentrations than those tested

here.



101

4.3.2.2 Effect of Benzene

In Chapter 3 a new reaction network was proposed for ANL and secondary

products formation including Bz, through ANL deamination – Figure 4.8, which should

be confirmed with further tests. Besides, it is known that industrially Bz can be present in

the reactor as a secondary product, that is recirculated with ANL or as a contaminant of

NB. Having these issues in mind, catalytic tests were made using a feed mixture of 0.7

wt.% Bz + 10 wt.% NB in ANL (TC3) and a feed mixture using 10 wt.% of NB in ANL

(TC6) – cf. Table 4.1. The tests were carried out at 150 ºC, 14 barg and using a liquid

feed flow rate of 5 ml/min, Figure 4.10. It is important to refer that Bz concentration was

not included in the light products, since it is a reagent in run TC3.

a)

b)

c)

Figure 4.10 – Evolution of a) NB concentration, b) ANL formation, c) secondary products

concentration, along reaction time (runs TC3 and TC6).

It was found that NB consumption is only very slightly detrimentally affected by

the presence of Bz, leading to a small decrease in NB conversion. Moreover, secondary

products formation is not affected by the presence of Bz (Figure 4.10).

0 100 200 300 400 5000.0

0.4

0.8

1.2

1.6

2.0

NB

Co

nce

ntr

atio

n (

%)

time (min)

10 wt.% NB + 0.7 wt.% Bz

10 wt.% NB

0 100 200 300 400 5000

3

6

9

12

AN

L f

orm

ed (

%)

time (min)

10 wt.% NB + 0.7 wt.% Bz

10 wt.% NB

0 100 200 300 400 5000

1000

2000

3000

4000

C se

con

dar

y p

rod

uct

s (ppm

)

time (min)

10 wt.% NB + 0.7 wt.% Bz

10 wt.% NB


102

Concluding, if Bz concentration has a huge increase in the reaction mixture, a

decrease in the NB conversion will occur, meaning less ANL formation, which is a

drawback from an industrial point of view.

4.3.2.3 CHA hydrogenation

CHA is one of the most important secondary products in the ANL hydrogenation,

since it is formed directly from ANL via the Amine (Figure 4.8). In fact, CHA can

achieve concentrations up to 2 wt.% in the reactor mixture. Beyond that, some authors

believe that CHA is not an end reaction product, contributing for the formation of some

heavy products. One of the compounds that might be formed through CHA is DICHA,

Eq. 4.5.

𝐶6𝐻13𝑁 + 𝐶6𝐻13𝑁 → 𝐶12𝐻23𝑁 + 𝑁𝐻3 (4.5)

Considering the scenario where only CHA is present in the liquid feed, the

catalytic hydrogenation of this compound was carried out. CHA used in this test was of

industrial grade and contained around 0.26 wt.% of CHOL and 0.24 wt.% of Bz. To

identify if CHA is a terminal product or not, the test was performed with CHA as reactant

(TC5) and results shown in Figure 4.11 were compared with ANL hydrogenation (TC1).

Both tests were performed at 150 ºC, 14 barg and with a liquid feed flow rate of 5 ml/min.



103

a)

b)

c)

Figure 4.11 – Evolution of a) secondary products concentration, b) light products concentration,

c) heavy products concentration, along reaction time (runs TC1 and TC5).

Data obtained (Figure 4.11) clearly demonstrated that there is no hydrogenation of

CHA along the reaction test. In fact, it was not detected the formation of any compound,

neither light nor heavy products.

According to eq. 4.5, two molecules of CHA react and there is the formation of

DICHA. According to Narayanan et al. [16], CHA can couple with ANL to form N-

phenylcyclohexylamine (NPCHA) or it can undergo dimerization to form DICHA, which

means that the ANL produced will be consumed. In this case, DICHA formation was not

detected. Relatively to the other heavy products, CHENO and CHANIL, their formation

was not observed.

Although CHA is one of the most important compounds resulting from ANL

hydrogenation, it was demonstrated that, by itself, it is not a precursor of any compound.

Direct formation of DICHA through CHA condensation was not found.

0 100 200 300 400 5000

1500

3000

4500

6000

7500

C se

condar

y p

roduct

s (ppm

)

time (min)

ANL

CHA

0 100 200 300 400 5000

500

1000

1500

2000

C li

gh

t p

rod

uct

s (p

pm

)

time (min)

ANL

CHA

0 100 200 300 400 5000

1000

2000

3000

4000

5000

C h

eav

y p

rod

uct

s (ppm

)

time (min)

ANL

CHA


104

4.3.2.4 ANL hydrogenation

In Chapter 3 it was concluded that NB hydrogenation can be divided in two parts:

the 1st part where NB hydrogenation into ANL is the main reaction (high NB

concentrations) and the 2nd part where the predominant reaction is ANL hydrogenation

(low NB concentrations). Consequently, it is important to study another extreme case,

when only ANL is fed into the reactor; moreover, one should assess, in this condition,

how high pressures and temperatures will influence the hydrogenation reaction.

In Figure 4.12 and 4.13 the results of tests performed for ANL hydrogenation are

shown. TC1 and TC2 were carried out with ANL in the feed, at a flow rate of 5 ml/min.

TC1 was performed at 150 ºC and 14 barg and TC2 at 200 ºC and 20 barg. It is important

to refer that the ANL used is of industrial grade, containing some of the secondary

products and H2O.

a)

b)

c)

d)

Figure 4.12– Evolution of a) ANL concentration, b) secondary products concentration c) light

products concentration and d) heavy products concentration, along reaction time (runs TC1 and

TC2).

0 100 200 300 400 50095

96

97

98

99

100

AN

L C

on

cen

trat

ion

(%

)

time (min)

200 ºC, 20 barg

150 ºC, 14 barg

0 100 200 300 400 5000

5000

10000

15000

20000

25000

C se

condar

y p

roduct

s (p

pm

)

time (min)

200ºC, 20 barg

150 ºC, 14 barg

0 100 200 300 400 5000

3000

6000

9000

12000

15000

C li

ght

pro

duct

s (p

pm

)

time (min)

200 ºC, 20 barg

150 ºC, 14 barg

0 100 200 300 400 5000

3000

6000

9000

12000

15000

C h

eavy p

roduct

s (p

pm

)

time (min)

200 ºC, 20 barg

150 ºC, 14 barg



105

a)

b)

Figure 4.13– Evolution of a) secondary products concentration distribution for 150 ºC and 14

barg and b) secondary products concentration distribution for 200 ºC and 20 barg, along reaction

time (runs TC1 and TC2).

Analysing data presented in Figures 4.12 and 4.13, it is well demonstrated that

ANL hydrogenation occurs in both operating conditions. Besides that, the hydrogenation

is more noticed when operating conditions are more severe. While at 150 ºC and 14 barg

ANL concentration is always higher than 99 wt.%, for 200 ºC and 20 barg the ANL

concentration decreases, being around 98 wt.% at steady-state, Figure 4.13 a).

For higher temperature and pressure, ANL concentration decreases and secondary

products increase, as expected. This increase is more noticed in the formation of light

products than of heavy products (increase at steady-state by a factor of ca. 7.7 and 3.2,

respectively). In fact, augmentation occurs for all secondary products, except for CHONA

and CHOL. In the case of CHONA, since there is no NB in the feed mixture, its

formation should be through the Amine + H2O, Figure 4.8. In what concerns the effect of

time on stream, it is detected a slight increase in CHONA concentration in the beginning

of the test followed by a decrease and stabilization at low concentrations. This variation is

more pronounced for more severe operating conditions, Figure 4.14. In the case of

CHOL, it was not detected its formation. CHENO also presents an augmentation in the

beginning of the test, higher for 200ºC and 20 barg, followed by a stabilization in

concentration in similar values for both tests.

0

4000

8000

12000

16000

20000

24000

Co

ncentr

atio

n (

ppm

)

Bz

CHA

CHOL

CHONA

DICHA

CHENO

CHANIL

150 ºC, 14 barg

time (min)

0

4000

8000

12000

16000

20000

24000

Co

ncentr

atio

n (

ppm

)

Bz

CHA

CHOL

CHONA

DICHA

CHENO

CHANIL

200 ºC, 20 barg

time (min)


106

a)

b)

Figure 4.14– Evolution of a) CHONA concentration, b) CHENO concentration along reaction

time (runs TC1 and TC2).

CHA, DICHA and CHANIL are the compounds where the increase in

concentration with the operating conditions severity is more pronounced. These

compounds result directly from ANL hydrogenation and condensation. At 20 barg and

200 ºC, there is more H2 solubilized in the reaction mixture and reaction rate is

accelerated, so it is expected that compounds resulting from hydrogenation reactions have

a greater increase, Figure 4.13. It is the case of CHA, that afterwards reacts with ANL and

forms CHANIL, that in turns is hydrogenated into DICHA.

Králik et al. [19] also studied hydrogenation of ANL over a Ru catalyst

concluding that CHA and DICHA are the main products obtained, in this experiment,

CHA was also a major compound but instead of DICHA, it was CHANIL to be detected

in higher quantities.

The increase of temperature and pressure in the hydrogenation of ANL shows that

more secondary products are formed. This information is very important in order to

choose the better operating conditions to be used in order to avoid the formation of

secondary products when the reaction is in the 2nd part, i.e., low NB concentrations,

particularly in a fixed-bed reactor (where there is an NB concentration profile along the

reactor length).

0

40

80

120

160

200

0 100 200 300 400 500

C (

ppm

)

time (min)

CHONA

150 ºC, 14 barg

200 ºC, 20 barg

0

500

1000

1500

2000

2500

0 100 200 300 400 500

C (

ppm

)

time (min)

CHENO

150 ºC, 14 barg

200 ºC, 20 barg



107

4.4 Conclusions

Influence of solvent and reaction products of NB hydrogenation into ANL was

studied over a Pd-based catalyst, 1 wt.% Pd/Al2O3. This analysis was also extended to

direct ANL and CHA hydrogenation.

A molecule similar to ANL was chosen to be used as co-solvent (p-tol) and so a

mixture of ANL + p-tol was tested and compared with ANL alone. It was found that the

presence of p-tol does not leads to the formation of other compounds and prevents the

formation of secondary products, in particular the heavy ones. However, NB conversion

and ANL formation is slower.

Reaction products influence in the NB hydrogenation was analysed out by co-

feeding H2O and Bz. It was observed that ANL hydrogenation decreases in the presence

of H2O, and that the formation of secondary products is lower, particularly of heavy

products. Relatively to Bz co-feeding, it was concluded that when Bz is present in higher

quantities it leads to a decrease in NB conversion. From an industrial point of view this

represents less ANL formation, which is negative; effect in secondary products formation

is negligible.

By direct hydrogenation of CHA it was concluded that if the only compound

present is CHA, no other compounds are formed; this confirmed the reaction mechanism

postulated in Chapter 3. By performing direct ANL hydrogenation, it was demonstrated

that when operating conditions are more severe, in terms of temperature and pressure,

secondary products formation has a considerable increase, particularly the light ones.


108

References

[1] Rase, W., Hydrogenation of Nitroaromatics, Nitrobenzene Aniline, Handbook of

Commercial Catalyst: Heterogeneous Catalysts, CRC Press LLC, 2000.

[2] Kahl, T., Schröder, K.-W., Lawrence, F.R., Marshall, W.J., Höke, H., Jäckh, R.,

Aniline, Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH, 2011.

[3] Vogt, P.F., Gerulis, J.J., Amines, Aromatic, Ullmann’s Encyclopedia of Industrial

Chemistry, Wiley-VCH, 2000.



[5] Wegener, G., Brandt, M., Lothar, D., Hofmann, J., Klesczewski, B., Koch, D., Kumpf,


catalysis in the polyurethane industry, Applied Catalysis A – General 221 (2001) 303-

335.




[7] PERP Program – New Report Alert, August 2003, Nexant Chem Systems.


Z. Elektrochem., 4 (1898) 506-514.




of aromatic nitro compounds on gold catalysts, Angewandte Chemie 119 (2007) 7404-

7407.



Cat Chem 4 (2012) 59-63.








(2014) 1-5.

[14] Turáková, M., Salmi, T., Eränen, K., Wärna, J., Murzin, D., Králik, M., Liquid phase

hydrogenation of nitrobenzene, Applied Catalysis A: General 499 (2015) 66-76.



109














Engineering 6 (2012) 1074-1082.






Petrochemical and Chemical Engineering, 2015.

[22] Du, B., Su, H., Wang, S., Palladium supported on carbon nanofiber coated

monoliths for three-phase nitrobenzene hydrogenation: Influence of reduction

temperature and oxidation pre-treatment, Journal of Industrial and Engineering

Chemistry 21 (2015) 997-1004.







activity in the liquid phase hydrogenation on nitrobenzene to aniline, Applied Catalysis

A: General 476 (2014) 103-112.

[25] Pikna, L., Hezelová, M., Demcakova, S., Smrcová, M., Plesingerová, B., Stefanko,

M., Turáková, M., Králik, M., Pulis, P., Lehocký, P., Effect of support on activity of

palladium catalysts in nitrobenzene hydrogenation, Chemical Papers 68 (2014) 591-598.

[26]Torres, G., Jablonski, E., Baronetti, G., Castro, A., Miguel, S., Scelza, O., Blanco,

M., Jiménez, P., Fierro, J., Effect of the carbon pre-treatment on the properties and

performance for nitrobenzene hydrogenation on Pt/C catalysts, Applied Catalysis A:

General 161 (1997) 213-226.


110

[27]Li, J., Li, X., Ding, Y., Wu, P., Pt nanoparticles entrapped in ordered mesoporous

carbons: An efficient catalyst for the liquid-phase hydrogenation of nitrobenzene and its

derivatives, Chinese Journal of Catalysis 36 (2005) 1995-2003.

[28] Relvas, J., Andrade, R., Freire, F., Lemos, F., Araújo, P., Pinho, M., Nunes, C.,

Ribeiro, F., Liquid Phase hydrogenation of nitrobenzene over an industrial Ni/SiO2

supported catalyst, Catalysis Today 133–135 (2008) 828–835.

[29] Zhao, F., Zhang, R., Chatterjee, M., Ikushima, Y., Araic, M., Hydrogenation of



[30] Zhao, F., Zhang, R., Chatterjee M., Ikushimab, Y., Araic, M., Hydrogenation of

Nitrobenzene with Supported Transition Metal Catalysts in Supercritical Carbon

Dioxide, Preprints of Papers - American Chemical Society, Division of Fuel Chemistry 49

(2004) 13-14.

[31] Kirk-Othmer, Amines, Aromatic, Encyclopedia of Chemical Technology, M. Howe-

Grant. New York, John Wiley & Sons 426-482, 1992.

*Adapted from: Clara Sá Couto, Luís M. Madeira, Clemente Pedro Nunes, Paulo Araújo,

Commercial Catalysts Screening for Liquid Phase Nitrobenzene Hydrogenation, Applied Catalysis

A: General 522 (2016) 152-164 (DOI: http://dx.doi.org/10.1016/j.apcata.2016.04.032). 111

Chapter 5 - Commercial Catalysts Screening for

Liquid Phase Nitrobenzene Hydrogenation.

Abstract

In this work, a series of commercially available materials was screened for the

catalytic hydrogenation of nitrobenzene (NB). The materials revealed different

performances, particularly different activities in what concerns the NB conversion, and

notably diverse selectivities towards the industrially desired reaction product, aniline

(ANL). The catalysts’ active phases are based on Pd and Ni (respectively groups I and II),

namely 1 wt. % Pd/Al2O3 (Catalyst I.1), 0.3 wt. % Pd/Al2O3 (Catalyst I.2), 0.3 wt. %

Pd/Al2O3 (Catalyst I.3), and 50 wt.% NiO/(Al2O3+SiO2) (Catalyst II.1). The fresh and

used materials were characterized by several physical-chemical techniques, specifically

scanning electron microscopy (SEM), high resolution transmission electron microscopy

(HRTEM), nitrogen adsorption (with BET surface area determination), X-ray diffraction

(XRD), H2 temperature-programmed reduction (TPR), inductively coupled plasma mass

spectrometry (ICP-MS) and elemental (CHNS) analysis. It was shown that the catalysts

are stable in the conditions studied and no deactivation was found. The characterization

results allowed explaining the catalytic behavior of the tested materials. In particular,

catalyst I.1 was found to be the less active, probably due to its much lower BET surface

area (and larger Pd particle size). On the other hand, catalyst I.2 was the more active,

which was well correlated to the smaller average particle size (along with narrower Pd

particle size distribution) and smaller pellet size, although the active metal content is low.

Finally, it was observed that catalyst II.1 is the most selective towards light by-products

(benzene (Bz), cyclohexylamine (CHA), cyclohexanol (CHOL) and cyclohexanone

(CHONA)), probably due to its lower pore size dimensions

http://dx.doi.org/10.1016/j.apcata.2016.04.032


112

5.1 Introduction

Aniline (ANL) is an important raw material for the polyurethane industry, being

used mainly in the production of methylene diphenyl diisocyanate (MDI) [1, 2].

Commercial ANL is predominantly produced by the catalytic hydrogenation of

nitrobenzene (NB) – eq. (1), which can be performed in gaseous or in liquid phase [3, 4].

𝐶6𝐻5𝑁𝑂2 + 3𝐻2 → 𝐶6𝐻5𝑁𝐻2 + 2𝐻2𝑂 (5.1)

One of the advantages of using liquid phase hydrogenation is to avoid the hot-spots

(as a consequence of the reaction exothermicity); moreover, when compared with the

vapor-phase and for a given reactor size, it usually shows a greater production capacity

and allows using reaction heat to produce steam. Typically, NB liquid-phase

hydrogenation is operated at 80 – 250 ºC under pressure with yields of 98 – 99 %. In

vapor-phase, although yields of 99 % or higher could be achieved, it is operated under

pressure at slightly higher temperatures [1, 5].

For NB hydrogenation many catalysts offering high activity and selectivity are

available, being the choice of the right catalyst to be used in the industrial ANL synthesis

a key issue. Usually, for the hydrogenation of nitro groups, the most used metals are

palladium, platinum and nickel, supported or not [6]. In the case of NB hydrogenation

there is a wider possible choice, including also the use of copper and cobalt [7]. Other

reducing agents, such as ferrous salts, tin, zinc, soluble sulfides, sulfur, and carbon

monoxide were also proposed [1-8]. Catalyst selection is very important for maximizing

ANL selectivity, keeping secondary products formation low.

Palladium and nickel catalysts have shown great success in the NB hydrogenation

[9, 10]. For instance, it has been reported the use of platinum-palladium catalyst on a

carbon support, with iron as a modifier [11], as well as palladium or palladium-platinum

deposited on a lipophilic carbon support [12], palladium supported on iron oxides [13], or

using a gel entrapped palladium catalysts [14]. Noble metals can catalyze the NB

hydrogenation under mild conditions, since the nitro group was found to be one of the

most suitable to be reduced using this type of catalysts. Nevertheless, their use in large-

Chapter 5 – Commercial Catalysts Screening for Liquid

Phase Nitrobenzene Hydrogenation.

113

scale production has not been explored due to their high costs [15].

The advantage of Pd-based catalysts is related with their high activity and also to

the fact that they do not attack, or even disrupt, the aromatic ring [16]. In the open

literature, the most studied catalysts for NB hydrogenation in liquid phase are palladium

supported in carbon [17-20]. However, upon consulting the list of commercial catalysts

available in the market, it is possible to verify that the most common is palladium

supported on Al2O3, with the variations restricted to their shape and metal content [21].

Ni-based catalysts are also well-known and widely studied in the liquid phase

reaction, either being supported or on a Raney form [22, 23]. Industrially, Ni catalysts are

also very used due to their low cost and high yields [3].

The main goal of this work was to test available commercial catalysts suitable for

the NB hydrogenation in liquid phase that are active in mild conditions of temperature

and pressure. Pd-based catalysts are the most appropriate, and so it was decided to acquire

several Pd catalysts (from now on called group I). Industrially, Ni based-catalysts are the

most used for ANL production being decided to acquire one of this kind to use as a

reference (group II) to compare activity and selectivity to the desired reaction, with less

secondary products formation.


5.2.1. Catalyst samples

Materials used in this work are available in the market and are presented as

catalysts for the hydrogenation of aromatic nitro groups into aromatic amines. The main

characteristics of the as-received catalysts are presented in Table 5.1 (group I – Pd-based;

group II – Ni-based).


114

Table 5.1– Catalysts main physical characteristics.

catalyst I.1 catalyst I.2 catalyst I.3 catalyst II.1

Composition 1 wt. % PdO /

Al2O3

0.3 wt. % PdO /

Al2O3

0.3 wt. % PdO /

Al2O3

50 wt. % NiO /

(Al2O3+SiO2)

Size (mm) 2 - 4 1.2 – 2.5 4 - 6 0.8 – 3.2

5.2.2. Catalysts Characterization

The commercial catalysts were characterized by XRD (X-ray Diffraction), SEM

(Scanning Electron Microscopy), TEM (Transmission Electron Microscopy), HRTEM

(High Resolution Transmission Electron Microscopy), H2-TPR (Temperature

Programmed Reduction), nitrogen adsorption for BET surface area determination, ICP-

MS (Inductively coupled plasma mass spectrometry) and elemental (CHNS) analysis.

Two samples of each catalyst were characterized: without suffering any reaction (fresh

samples) and after the catalytic tests (used samples), except for H2-TPR, where only the

fresh materials were analysed because the used samples have some organic compounds,

which interfere in the TCD signal.

The XRD patterns were obtained on a Rigaku diffractometer, Geigerflex, in the

angular range of 10° to 100° (2θ) with a scan rate of 3° / min. The morphology and

surface of the catalysts were analyzed by SEM, performed on a Hitachi SU-70

microscope. TEM, performed on a Hitachi 8100 with ThermoNoran light elements EDS

detector and digital image acquisition, was used to monitor the morphological properties

of the different samples. HRTEM was performed on a JEOL 2200FS apparatus to

determine the Pd particle size distribution in the support.

H2-TPR experiments were performed on a Micromeritics AutoChem II 2920

apparatus, using 130 mg of catalyst. Ni catalyst was pre-treated, under argon (flow rate of

25 mL/min NTP), from room temperature (RT) up to 250 °C (heating rate of 10 °C/min),

kept at 250 °C for 1 h and then cooled to RT. This pre-treatment was realized with the

objective of removing water and impurities from the catalyst surface. Then, the reactor

was purged with argon for 10 min at RT. Only the Ni catalyst was pre-treated, because Pd

particles are reduced in the presence of Argon. H2-TPR was carried out under a mixture

of 5% H2/Argon with a flow rate of 30 mL/min (NTP), from RT up to 900 °C at a heating

rate of 10 °C/min. Hydrogen consumption was measured with a TCD; water formed

during the reduction processes was trapped in a dry ice trap.



115

Nitrogen adsorption and desorption measurements were carried out at 77 K with

an automatic Micromeritics ASAP 2000 apparatus. Prior to analysis, the samples were

pretreated at 448 K under vacuum for 6 h. The BET surface area (Sext), the total pore

volume (Vtotal), calculated from the adsorbed volume of nitrogen for a relative pressure

P/P0 of 0.99, and the average pore diameter (Daverage), were estimated.

Elemental (CHNS) analysis was performed on a TruSpec Micro equipment with a

nominal sample weight of 2 mg. The determination of elements by ICP-MS was

performed on an ICP-MS Thermo X Series apparatus. The sample to be analyzed was

rigorously weighed (ca. 0.05 g) and it was added 1 ml HNO3 + 3 ml HCl + 1 ml de HF.

Then the sample was digested under microwave heating (180 ºC) for 5 min. The sample

was finally taken up to 100 ml with ultrapure water and analyzed.

5.2.3. Catalytic Reaction


reactor provided with an air-impelled stirrer (Figure B.1 of the Appendix B section). The

stirrer was equipped with a basket, where the catalysts were placed. The catalysts used are

those indicated in Table 5.1. Preliminary results have shown that above 1000 rpm no

major differences were observed in the catalytic results (data not shown); so, the stirrer

was set at such value in all runs to guarantee that external resistances to mass transfer do

not exist.

A known amount of catalyst was loaded into the reactor and the material pre-

treated, in situ, at 150 ºC under hydrogen pressure (20 barg) for 2 hours. A certain volume

of ANL was loaded into the reactor, in order to avoid the formation of two phases

(aqueous and organic) that would stop the reaction; ANL addition also helps in

dissipating the excess heat generated due to the high exothermicity of the reaction.


controller (SHIMADEN SD20) and the initial temperature was then established. The heat

produced by the nitrobenzene hydrogenation was removed by a water stream whose flow

was controlled with a set of ball valves, as shown in Figure S.1. The reactor temperature

was constant with a maximum ΔT of ± 6 ºC and it was continuously measured throughout

the experiments.


116

Nitrobenzene (NB) was charged in a vessel and subsequently loaded into the reactor

by pushing it using a high-pressure hydrogen stream (cf. Figure B.1). NB was loaded into

the reactor instantaneously, when the desired reaction temperature was achieved (time = 0

min). This procedure was adopted to ensure that NB hydrogenation does not start before

the beginning of the experiment and also to avoid any strong NB adsorption on the

catalyst, blocking accessibility to active sites; this is also ensured by using ANL as

solvent. All the experiments were done in batch mode up to a nearly complete

consumption of NB (which was considered to correspond to the instant at which NB

concentration was below 1000 ppm).

The reactor operates in a batch mode relative to the liquid phase but in a semi-

continuous mode for the gas phase (hydrogen). The total pressure inside the reactor is

kept constant along each run due to the continuous admission of hydrogen as it is being

consumed.

The reference values for temperature, pressure and initial nitrobenzene

concentration during the catalytic screening tests will be represented by Tref (ºC) = 150

ºC, Pref (barg) = 14 barg and Cref = 10 % (100 000 ppm), respectively. For catalysts

screening, the effect of reaction temperature and total pressure was analysed, tests

indicated as T, and all the experiments were done in batch mode. For all the catalysts,

some preliminary tests were performed in the same operating conditions (Cref, Tref and

Pref) – Table 5.2, tests E – to check reproducibility.

The experiments performed and the conditions used are given in Table 5.2. In all

runs, the catalyst-to-liquid volume ratio was 80.0 g / L and the initial reactants ratio, NB

to ANL, was 0.11 (w / w).



117


Test Catalyst Experimental conditions

E1

catalyst I.1

1 wt. % PdO / Al2O3

150 ºC, 14 barg

T2 150 ºC, 6 barg

E3 150 ºC, 14 barg

T4 150 ºC, 30 barg

E5 150 ºC, 14 barg

T6 180 ºC, 14 barg

T7 240 ºC, 14 barg

E8

catalyst I.2

0.3 wt. % PdO / Al2O3

150 ºC, 14 barg

T9 150 ºC, 6 barg

E10 150 ºC, 14 barg

T11 150 ºC, 30 barg

E12 150 ºC, 14 barg

T13 180 ºC, 14 barg

T14 240 ºC, 14 barg

E15

catalyst I.3

0.3 wt. % PdO / Al2O3

150 ºC, 14 barg

T16 150 ºC, 6 barg

E17 150 ºC, 14 barg

T18 150 ºC, 30 barg

E19 150 ºC, 14 barg

T20 180 ºC, 14 barg

T21 240 ºC, 14 barg

E22

catalyst II.1

50 wt. % NiO / (Al2O3+SiO2)

150 ºC, 14 barg

T23 150 ºC, 6 barg

E24 150 ºC, 14 barg

T25 150 ºC, 30 barg

E26 150 ºC, 14 barg

T27 180 ºC, 14 barg

T28 240 ºC, 14 barg

The sampling of liquid phase was performed at selected time intervals and collected

samples were analysed by gas chromatography, in an Agilent 6890A chromatograph

equipped with two flame ionization detectors (FID). The column used was a HP-1 one

(100% dimethylpolysiloxane, 30 m x 320 µm x 4 µm). The temperature in the injector

and in the detector was 250 ºC, the pressure in the column was 14 bar and the carrier gas

used was helium. The column oven was temperature-programmed with a 1 min initial

hold at 120 ºC, followed by an increase of temperature until 230 ºC at a rate of 15 ºC/min

and kept at 230 ºC for 9 min.


Calibration curves were plotted for all the compounds to be analysed, which were easily

identified since their retention times are known. Several samples were injected and the

standard deviation associated with this method was found to be below 10%.


118


It is important to refer that all the catalysts were analyzed in their original shape,

because the main objective of this study is to perform a catalytic screening of commercial

catalysts. Consequently, with the exception of TEM analysis, all the other techniques

were performed with the catalysts as received.

5.3.1 Catalysts Characterization

All the catalysts, fresh and used, were analyzed by ICP and elemental (CHNS)

analysis. It was verified that Pd content is equal in fresh and used samples. Consequently,

it can be concluded that Pd leaching does not occur. Also, some samples of reaction

mixture were analyzed, which confirmed the absence of Pd and Ni in the liquid phase.

However, in the case of catalyst II.1, with Ni, it was verified some losses of metal. In all

used catalysts it was detected the presence of carbon (by elemental analysis), amounting

in catalyst I.1 to 1.2 wt.%, in catalyst I.2 to 4.6 wt.%, in catalyst I.3 to 1.2 wt.% and in

catalyst II.1 to 3.0 wt.%, indicating that some of the reaction compounds became trapped

in catalysts pores/surface, blocking porosity. However, the amounts quantified are not

expected to interfere with the catalysts stability, as shown below (cf. section 5.3.2).

The X-Ray diffraction patterns of the different catalysts are shown in Figure 5.1. It

is observed that only catalyst I.1 has a crystalline nature while the others present an

amorphous structure. The samples used were also analyzed by XRD (Figure B.2 of

Appendix B), but no significant differences between them (fresh and used samples) were

observed. In the case of catalyst I.1, crystallinity loss was less than 10%. This loss is

calculated through the determination of the diffractogram area of each sample (fresh and

used). The species found, in both fresh and used catalysts, were aluminum oxide (Al2O3)

and corundum (Al2O3, all the other peaks). Species detected for catalyst II.1 were

aluminum oxide (Al2O3) and gibbsite (Al(OH)3), while for catalyst I.3 only aluminum

oxide (Al2O3) was identified. Catalyst II.1 presents peaks for metallic nickel (Ni) and

bunsenite (NiO). For catalysts of the group I, Pd metal was not detected. This is probably

related with the fact that the samples were not previously activated (Pd was not reduced).

PdO was also not found due to the very small load / particle size of Pd.



119

Figure 5.1– X-ray diffraction patterns of the fresh catalysts studied: a) catalyst I.1, b) catalyst I.2,

c) catalyst I.3 and d) catalyst II.1.

The morphology of the catalysts samples was analyzed by SEM. Fresh and used

samples (Figure B.3 of the Appendix B) evidenced no major differences between them.

TEM micrographs of the catalysts are shown in Figure B.4 of the Appendix B, also

evidencing no major differences between fresh and used samples, except for the Ni-one.

In fact, in catalyst II.1 it was observed that Ni particles agglomerate when subjected to the

reaction, but the impact in catalytic activity was not evident in the few tests performed;

possibly this would be noticed after long-term use.

Particle size distribution determined by HRTEM for the group I samples revealed

that in catalyst I.2, Pd particles are homogeneously dispersed in the support with a

particle size distribution in the range of 0.5-5.5 nm (Figure 5.2). In catalyst I.3 average Pd

particle size is slightly higher, being the highest one for catalyst I.1, which distribution is

not so homogeneous.

a)

b)

c)

d)

2 40

ntensit

(a.u

.)

heta (degrees)

*

*

α

α

α

α

α

α

α

α

αα α αα α

αα

Al2O

Al(OH)

NiO

Niα orundum (Al2O )


120

Figure 5.2 – Particle size distribution of fresh group I catalysts determined by HRTEM.

H2-TPR profiles of fresh catalysts are shown in Figure 5.3. The profiles of all the

catalysts containing Pd (catalyst I.1, catalyst I.2 and catalyst I.3), show a negative peak

(H2 desorption) at about 70 ºC, which is due to the decomposition of Pd hydride (formed

in the reduction of PdO under 5 % H2 / Ar before the ramp was started at RT), typical for

palladium-based catalysts. Catalyst I.1 has the highest Pd content (1 wt. % vs. 0.3 wt. %

for catalyst I.2 and catalyst I.3), being observed that its peak is the most intense, in

agreement with what is reported in the literature [24, 25]. In the case of catalyst I.1 and

catalyst I.3 it was also detected a slight H2 consumption from 230 to 450 ºC, which could

be assigned to the reduction of stable PdO species that strongly interact with the support

[26, 27]. Catalyst II.1, the Ni-based, presents three peaks in the H2-TPR, centred at 150

ºC, 245 ºC and 650 ºC. The two reduction peaks at low temperature may be ascribed to

the reduction of Ni2O3 to metallic nickel or to Ni2+ [28, 29]. The peak centred at 650 ºC

can be assigned to the reduction of dispersed Ni2+ species that are stabilized by a strong

interaction with the support [30]. Based on the H2-TPR results it is possible to conclude

that all the catalysts are completely reduced after the pre-treatment described in section

5.2.3.

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.510.00

10

20

30

40

50

60

Fre

qu

ency

(%

)

Particle size (nm)

catalyst I.1 dpaverage

= 5.59 nm


= 1.03 nm


= 1.18 nm



121

Figure 5.3 – Temperature programmed reduction profiles for the fresh Pd-based (a) catalyst I.1,

b) catalyst I.2, c) catalyst I.3) and Ni-based (d) catalyst II.1) materials studied.

The textural parameters of all catalysts are summarized in Table 5.3. Globally, none

of the catalysts seems to lose its porosity after the catalytic reaction, at least in a

significant extent. Catalyst I.1 presents very low BET surface area and total pore volume.

Catalyst I.2 has the same formulation as catalyst I.3, but they are from distinct suppliers;

through the analysis of their textural parameters it is possible to see that the results for

fresh samples are only slightly different, which becomes more evident after being used in

the hydrogenation reaction. While catalyst I.3 loses 12.5 % of its total pore volume,

catalysts I.2 loses 7.8 % of its BET surface area. In the case of catalyst II.1, one can

conclude that the reaction has also some influence in the textural parameters, except in

the BET surface area. The parameter that, globally, appears to be more influenced by the

hydrogenation reaction is the total pore volume (Vtotal pore). It is also observed that the

average pore diameter in the case of catalysts I.1 and I.2 increases after reaction, while for

catalyst I.3 and II.1 it decreases. This might be related with the types of pores that are

blocked upon the hydrogenation reaction. Although a decrease in the surface area and

volume of pores is exhibited by all the used catalysts, it was not observed any catalyst

deactivation during the parametric study, as discussed below.

0 50 100 150 200 250 300 350 400 450 500 550

Temperature (ºC)

c)

b)

H2 c

onsu

mpti

on (

a. u

.)a)

0 100 200 300 400 500 600 700 800 900

Temperature (ºC)

H2 c

onsu

mpti

on (

a.u.)

d)


122

Table 5.3 – Textural parameters for the catalysts samples studied.

Sample BETsurface area

[m2 g-1] [a]

ΔBETsurface area [%][b]

Vtotal pore

[cm3 g-1] [c]

ΔVtotal pore

[%][d]

Daverage

[Å] [e]

ΔDaverage

[%][f]

catalyst I.1 Fresh 8

12.5 0.020

5.0 103

-2.9 Used 7 0.019 106


7.8 0.566

3.4 223

-4.9 Used 94 0.547 234


4.7 0.553

12.5 209

8.1 Used 101 0.484 192

catalyst II.1 Fresh 212

3.8 0.298

16.4 56

12.5 Used 204 0.249 49

[a] BET surface area. [b] BET surface area loss. [c] Total pore volume. [d] Total pore volume loss. [e] Pore diameter

average. [f] Pore diameter average loss.

5.3.2 Nitrobenzene Hydrogenation

The catalytic tests performed had the objective of defining and choosing, among the

commercial industrial catalysts, which one is the most selective, which means the one that

presents higher selectivity towards ANL, having less secondary products formation, while

being also as active as possible in NB conversion. In Figure 5.4 and 5.5 are presented

some preliminary tests, which allowed confirming the reproducibility of the experiments.

It is also possible to conclude that the catalysts are stable in the runs made and within the

conditions studied. This is in agreement with the results obtained from the

characterization: no major differences were detected through textural or morphological

parameters, between fresh and used samples.

Figure 5.4 – Reproducibility tests, showing NB consumption as a function of reaction time at 150

ºC, 14 barg and 10% NB for each catalyst.

0 30 60 90 120 150 180

0

20000

40000

60000

80000

100000

CN

B (

pp

m)

time (min)

E1

E3

E5

Catalyst I.1

0 30 60 90 120 150 180

0

20000

40000

60000

80000

100000

CN

B (

pp

m)

time (min)

E8

E10

E12

Catalyst I.2



123

Figure 5.5 – Reproducibility tests, showing NB consumption as a function of reaction time at 150

ºC, 14 barg and 10% NB for each catalyst.

The liquid phase analysis confirmed the presence of NB and of the industrially

desirable product, ANL, as well as of the by-products cyclohexylamine (CHA),

cyclohexanol (CHOL), cyclohexanone (CHONA), N-cyclohexylaniline (CHANIL),

dicyclohexylamine (DICHA), cyclohexyldeneaniline (CHENO) and benzene (Bz) – cf.

Nomenclature section. Nevertheless, the secondary products will not be presented

individually but in groups: Light products – Bz, CHA, CHOL and CHONA; and Heavy

products: DICHA, CHENO and CHANIL. Figure 5.6 illustrates their formation,

according to the reaction mechanism proposed in a previous work [10], either directly

from NB or through ANL hydrogenation. Relatively to the carbon mass balance, all the

compounds identified allowed to close the balance within the analytical uncertainty (± 6

%).

0 30 60 90 120 150 180

0

20000

40000

60000

80000

100000C

NB (

pp

m)

time (min)

E15

E17

E19

Catalyst I.3

0 30 60 90 120 150 180

0

20000

40000

60000

80000

100000

CN

B (

pp

m)

time (min)

E22

E24

E26

Catalyst II.1


124

Figure 5.6– Reaction network proposed for formation of ANL and secondary products [10].

*very reactive and unstable compound.

5.3.2.1 Catalysts activity

The influence of pressure in the catalysts performance was studied by varying this

variable between 6 and 30 barg (Table 5.2). The total pressure is increased in the reactor

due to the increase of the hydrogen pressure, meaning that more hydrogen will be present

in the gas phase above the liquid and, consequently, solubilized in the reaction mixture.

Therefore, it is expected that catalytic activity increases with the total pressure. In Figure

5.7 are presented the results for the NB consumption along reaction time for all the

catalysts, at different pressures. It is noteworthy that the maximum expected value of NB

concentration is never reached because the reactant is not inside the reactor at the initial

instant (cf. Experimental section 5.2.3), and so, when loaded into the reactor, there is

immediately a consumption of NB (this explains why CNB never reaches the value of

100 000 ppm).

Heavy products

Light products

Light products

NO2

Cat

+H2

NH2

+ NH3

H2

Cat

H2

+ANL

-NH3

Cat

H2

NH2

NHNH2

+H2 -NH3

NH

+H2

NH

O

+H2

-NH3

+ANL -H2O

N

+H2

+H2OH CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

Bz

ANLNB

Amine


-H2O

-H2O

*-NH3

+H2O



125

a)

b)

c)

Figure 5.7– Effect of reaction total pressure on NB consumption as a function of reaction time for

the different catalysts: a) P = 6 barg, b) P = 14 barg and c) P = 30 barg.

It is clearly seen that an increase in pressure increases the NB consumption rate,

decreasing the time required to reach complete conversion (one should notice the

different time scales in Figures 5.7a to 5.7c). Relatively to the catalyst screening, and

whatever the operating conditions used, catalyst I.2 is the one that presents higher

reaction rates while catalyst I.1 always presents the slower ones. Catalyst I.3 and catalyst

II.1 present very similar performances at 14 barg (Figure 5.7 b)), however at 6 bar (Figure

5.7a)) and 30 barg (Figure 5.7 c)) catalyst I.3 is slightly better than catalyst II.1.

Analysing the temperature effect, it is observed in Figure 5.8 that catalyst I.2 is

again the one with the highest activity at all temperatures tested, while catalyst I.1

continues to be the one with the slowest activity. At 240 ºC (Figure 5.8c)), NB

consumption rate has a great increase for all the catalysts. It is shown that the profiles of

0 30 60 90 120 150 180 210 240 270 300 330 360

0

20000

40000

60000

80000

100000C

NB (

pp

m)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 30 60 90 120 150

0

20000

40000

60000

80000

100000

C N

B (

pp

m)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 30 60 90 120

0

20000

40000

60000

80000

100000

C N

B (

pp

m)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1


126

catalyst I.3 and II.1 are very similar and, despite still being the slowest one, catalyst I.1

has a NB profile closer to them.

a)

b)

c)

Figure 5.8- Effect of reaction temperature on NB consumption as a function of reaction time for

the different catalysts: a) T = 150 ºC, b) T = 180 ºC and c) T = 240 ºC.

Comparing the results shown in Figure 5.7 and Figure 5.8, it can be concluded that

temperature has a higher influence in the performance of all catalysts than pressure (in the

ranges studied), and that the higher is the temperature (or pressure), the higher will be the

activity.

0 30 60 90 120 150

0

20000

40000

60000

80000

100000

C N

B (

pp

m)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 30 60 90 120

0

20000

40000

60000

80000

100000

C N

B (

pp

m)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0

20000

40000

60000

80000

100000

C N

B (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1



127

a)

b)

Figure 5.9 – Comparison of NB consumption rate for all operating condition used a) per gram of

catalyst and b) per gram of metal.

In Figure 5.9 are shown the results of NB consumption rate, at all the conditions

studied, normalized either a) per gram of catalyst or b) per gram of metal. These reaction

rates were computed from previous figures, in the range where NB concentration

decreases nearly linearly along reaction time. The results are in agreement with what was

stated above when normalized per mass of catalyst, i.e. catalyst I.2 has the highest NB

consumption rates and catalyst I.1 the lowest ones, whatever the reaction temperature and

pressure. The Ni-based catalyst (sample II.1), similar to others used industrially, has NB

consumption rates in the range of 0.37-1.75 gNB / gcat h, close to others reported in the

literature (values in the range of 0.865 to 2.214 gNB / gcat h have been found, but they

depend on the operating temperature, total pressure, catalyst composition and particle size

6 barg - 150ºC 14 barg - 150 ºC 30 barg - 150 ºC 14 barg - 180 ºC 14 barg - 240 ºC0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

NB

consu

mpti

on r

ate (

g N

B /

h g

cat

alyst

)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

6 barg - 150ºC 14 barg - 150 ºC 30 barg - 150 ºC 14 barg - 180 ºC 14 barg - 240 ºC

0

200

400

600

800

1000

1200

1400

NB

con

sum

pti

on

rat

e (g

NB

/ h

g m

etal

)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1


128

[22, 31, 32]). For the catalysts of group I, Pb-based, NB consumption rate is in the range

of 0.21 – 3.25 gNB / gcat h; again, these values are in agreement with those found in

literature [32 – 33].

When normalized per gram of active phase, the results (Figure 5.8b) change

drastically in the case of catalyst II.1. This catalyst becomes the one with the slowest

consumption rate since it has the highest content of metal, as shown in Table 1. The

performance reached with catalyst I.1 slightly decreases, comparing with the other

catalysts, for the same reason. In the case of catalysts I.2 and I.3 the results remain

proportional between them, since they have the same metal content (and the same

formulation). However, the NB consumption rates are quite different among them.

HRTEM data allowed understanding that in catalyst I.2 Pd nanoparticles are slightly

smaller while in catalyst I.3 Pd particle size distribution is slightly wider. Another

possible reason for this difference is related with the pellet size (Table 5.1) – catalyst I.2

is smaller and thinner than catalyst I.3 and, consequently, if internal resistances to mass

transfer exist, they will be less relevant for catalyst I.2 and so the activity will be higher.

5.3.2.2 Catalysts selectivity

To analyse catalysts selectivity, it is necessary to evaluate the formation of

secondary products (grouped as light and heavy, as stated above) because their formation

will have an influence in the evolution of ANL concentration. Moreover, it should be

referred that light products are over-hydrogenated and hydrogenolytic substances, while

heavy products result from ANL condensation reactions.

Secondary products formation is influenced by the operating conditions used;

therefore, selectivity results will be presented as a function of pressure (Figure 5.10 and

Figure 5.11) and temperature (Figure 5.12 to Figure 5.14) along reaction time.

Catalyst II.1 is the one that produces more light products (Figure 5.10), mainly

CHONA, and unlike what would be expected, pressure increase has a negative influence

in their formation. Analysing the textural parameters it is possible to see that catalyst II.1

is the one that has the smallest Daverage (average pore diameter), which might explain its

larger selectivity to mostly smaller (i.e., light) by-products. On the other hand, catalysts

I.2 and I.1 produce more heavy compounds than the other catalysts.



129

The selectivity plots results can be divided in two parts, with different patterns: the

first one for high NB concentrations (short reaction times – where NB hydrogenation

prevails) and the second for low NB concentrations (longer reaction times – where ANL

hydrogenation is the dominant one). For high NB concentration (1st part), all the catalysts,

except catalyst II.1 for light products, present similar results. The formation of by-

products is low and constant (their presence since the beginning of the experiments is due

to the use of industrial level NB/ANL mixture, with low purity). When the NB

concentration is low, the formation of secondary products has a huge increase that is more

noticed for higher pressures (30 barg). To better illustrate this issue, Figure B.5 (and B.6)

in the Appendix B shows the selectivity of each catalyst towards the total of secondary

products at the same NB conversion level. Because the catalysts have different activities,

as reported above, in such figure one has, in the xx axis, not reaction time but rather NB

concentration; this means that in such plots time evolution of selectivity in a given

experiment should be analysed from the right to the left.

a)

b)

Figure 5.10 - Light products and Heavy products concentration at Tref as a function of reaction

time for different pressures: a) and b) P = 6 barg.

0 50 100 150 200 250 300 350

0

1x103

2x103

3x103

4x103

5x103

6x103

CL

ight

Pro

duct

s (ppm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 50 100 150 200 250 300 350

0

1x103

2x103

3x103

4x103

5x103

CH

eavy P

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1


130

a)

c)

b)

d)

Figure 5.11 - Light products and Heavy products concentration at Tref as a function of reaction

time for different pressures: a) and c) P = 14 barg and b) and d) 30 barg.

Figures 5.10 and 5.11 show that catalyst II.1 is the one that leads to a higher

formation of light products, mainly CHONA. According to Figure 5.6, CHONA

formation might occur through the direct hydrogenation of NB (the stage that prevails at

low reaction times) or through the reaction between the intermediate amine and H2O (the

2nd stage that dominates at longer reaction times). It seems that with catalyst II.1,

CHONA formation occurs via NB hydrogenation (1st part), and eventually through the

ANL route also (via the amine), while in the Pd-based catalysts its formation is via the

amine reaction (2nd part, which dominates at long reaction times).

When comparing the Ni-based with the Pd-based catalysts (group I), it is seen that

catalyst II.1 promotes the formation of light products, in detriment of the heavy ones

(Figures 5.10 and 5.11).

0 20 40 60 80 100 120 140

0

1x103

2x103

3x103

4x103

5x103

6x103

CL

ight

Pro

duct

s (ppm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120 140

0

1x103

2x103

3x103

4x103

5x103

CH

eavy P

roduct

s (ppm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0

1x103

2x103

3x103

4x103

5x103

6x103

CL

ight

Pro

duct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0

1x103

2x103

3x103

4x103

5x103

CH

eavy P

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1



131

a)

b)

c)

Figure 5.12 – Total secondary products concentration at Tref as a function of reaction time a) P =

6 barg, b) P = 14 barg and c) 30 barg.

Summarizing, in the case of pressure variation, catalyst II.1 is the one that has

higher production of secondary products at high NB concentrations – Figure 5.12 –

benefiting hydrogenolysis and overhydrogenation of ANL. At low NB concentrations,

either catalysts I.1/I.2 (at high pressures) or I.2/II.1 (at low to intermediate pressures)

appear to produce more secondary products, leading to ANL condensation reactions, and

to be more influenced by the variation of this parameter. Catalyst I.3, when NB

concentration is high, does not seem to be affected by pressure variation and mantain the

secondary products formation low, i.e., high selectivity towards ANL.

Relatively to temperature variation, its effect in secondary products formation is

shown in Figure 5.13. It is seen again in the 1st part of the reaction (short reaction times)

that catalyst II.1 produces more light products (influenced once more by CHONA

0 50 100 150 200 250 300 350

0.0

2.0x103

4.0x103

6.0x103

8.0x103

CT

ota

l se

condar

y p

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120 140

0.0

2.0x103

4.0x103

6.0x103

8.0x103

CT

ota

l se

condar

y p

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0.0

2.0x103

4.0x103

6.0x103

8.0x103

CT

ota

l se

condar

y p

roduct

s (pp

m)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1


132

formation). In the 2nd part, catalyst II.1 has the highest yields in light products and

catalysts I.1/I.2 exhibit a great production of heavy products.

a)

d)

b)

e)

c)

f)

Figure 5.13 – Light products and Heavy products concentration at Pref as a function of reaction

time at: a) and d) T = 150 ºC, b) and e) T = 180 ºC and c) and f) 240 ºC.

0 20 40 60 80 100 120 140

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

CL

ight

Pro

duct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120 140

0.0

5.0x103

1.0x104

1.5x104

2.0x104

CH

eavy P

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

CL

ight

Pro

duct

s (ppm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0.0

5.0x103

1.0x104

1.5x104

2.0x104

CH

eavy P

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 30 60 90 120

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

CL

ight

Pro

duct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20 40 60 80 100 120

0.0

5.0x103

1.0x104

1.5x104

2.0x104

CH

eavy P

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1



133

The increase in the reaction temperature leads to a higher formation of over-

hydrogenated products, mainly CHA. According to the proposed mechanism (Figure 5.6),

there is a condensation reaction between the amine and ANL resulting in the formation of

CHANIL that can be hydrogenated into DICHA. This influence is more noticed in the 2nd

part of the reaction (long reaction times), when NB concentration is low, and particularly

at higher temperatures.

Figures 5.14 and 5.15 clearly demonstrates that, in the 1st part of the reaction,

catalyst II.1 produces high quantities of secondary products although this production does

not increase significantly with the temperature. This trend is verified for all the catalysts.

Therefore, it can be concluded that in this part of the reaction (low reaction times and

high NB concentrations), the increase of temperature has a more significant effect in NB

consumption not significantly affecting the ANL selectivity (less secondary products

formation). In the 2nd part, with low NB concentrations, the effect is the opposite – high

temperature leads to high secondary products formation and consequently lower ANL

selectivity.

a)

b)

Figure 5.14 – Total secondary products concentration at Pref as a function of reaction time a) 150

ºC Tref and b) 180 ºC.

0 30 60 90 120 150

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

3.5x104

CT

ota

l se

condar

y p

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 30 60 90 120

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

3.5x104

CT

ota

l se

condar

y p

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1


134

Figure 5.15 – Total secondary products concentration at Pref as function of reaction time: 240 ºC.

5.4 Conclusions

Catalytic screening for the hydrogenation of NB into ANL was performed using 4

different industrial catalysts at different operating conditions; characterization of fresh

and used catalyst samples was also carried out.

In general, it was concluded that there are no major differences between fresh and

used samples, i.e., the hydrogenation reaction does not have a great influence in the

morphological and textural properties of all catalysts. Nevertheless, it was possible to

infer why catalysts I.2 and I.3 have such a different catalytic behavior, although having

the same formulation. HRTEM analysis allowed verifying that although Pd particles are

well distributed in both catalysts, the size of Pd particles is slighlty smaller in catalyst I.2.

Higher catalytic activity of catalyst I.2 might be also ascribed to its smaller pellet size in

the as-received commercial materials.

The results obtained allowed to divide the reaction in two parts: at high NB

concentrations NB hydrogenation is the predominant reaction (1st part – short reaction

times) and at low NB concentrations, which means high ANL concentrations, where ANL

hydrogenation and condensation become the main reactions (2nd part – long reaction

times). It was concluded that in the 1st part, catalyst I.2 presents the highest NB

consumption rates and a low formation of secondary products. On the other hand, at low

NB concentrations the best catalyst to use, in terms of selectivity, will depend on the

operating conditions to be employed: at low pressure or temperature the best would be

0 30 60 90 120

0.0

5.0x103

1.0x104

1.5x104

2.0x104

2.5x104

3.0x104

3.5x104

CT

ota

l se

condar

y p

roduct

s (p

pm

)

time (min)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1



135

catalyst I.1 or I.3; at high pressures and/or high temperatures the best would be catalyst

I.3.

Another important conclusion is that the operating conditions have a great impact in

the catalyst perfomance, mainly temperature, and thus they should be chosen very

carefully, in order to have a high NB consumption rate, while maintaning at a low level

the formation of secondary products.


136

References

[ ] R. Lawrence, W.J. Marshall. “Aniline”, Ullmann's Encyclopedia of Industrial

Chemistry, 6th Edition (Print), John Wiley & Sons, New York, 1998.

[2] K. Othmer, “Nitrobenzene and Nitrotoluenes”, Encyclopedia of hemical

Technology, vol. 17 (2007) 1-30.

[3] J. Relvas, R. Andrade, F. Gama Freire , F. Lemos, P. Araújo, M. Pinho, C. Pedro

Nunes, F. Ramôa Ribeiro, Catal. Today 133-135 (2008) 828-835.

[4] K. Othmer, “Amine by Reduction”, Encyclopedia of hemical Technology, vol. 2

(2007) 476-498

[ ] K. Othmer, “Aniline and its derivatives”, Encyclopedia of hemical Technology, vol.

2 (2007) 783-809.

[6] PERP Program – New Report Alert, August 2003, Nexant Chem Systems..

[7] M. Kise, US Patent 2292879, 1942.

[ ] G. Maxwell, “ hapter 2 : Synthetic Nitrogen Products”, A Practical Guide to the

Products and Processes, Springer, 2005.

[9] K. K. yeong, A. Gavriilidis, R. Zapf, V. Hessel, Catal. Today 81 (2003) 641-651.

[10] C. Sá Couto, L. M. Madeira, C. P. Nunes, P. Araújo, Chem Eng Technol 38 (2015)

1625-1636.

[11] F.Lawrence, US Patent 4415754, 1983.

[12] T. Nagata, K. Watanabe, Y. Kono, A. Tamaki, T. Kobayashi, US Patent 5616806,

1997.

[13] Y. Wang, Y. Deng, F. Shim, J Mol Catal A - Chem 395 (2014) 195-201.

[ 4] S. B. Tong, K. F. O’Driscoll, G. L. Rempel, an J hem Eng ( ) 4 -345.

[15] R. Downing, P. Kunkeler, H. van Bekkum, Catal Today 37 (1997) 121-136.

[16] H-U. Blaser, A. Indolese, A., Schnyder, H. Steiner, M. Studer, J Mol Cat A - Chem

173 (2001) 3–18.

[17] M. Turáková, T. Salmi, K. Eränen, J. Wärna, D. Yu. Murzin, M. Králik, Appl Catal

A - Gen 499 (2015) 66-76.

[18] M. Králik, M. Turáková, I. Macák, P. Lehocký, Aniline – catalysis and chemical

engineering, Editos: J. Markos, In Procedings of the 41st International Conference of

Slovak Society of Chemical Engineering, Tatranské Matliare, Slovakia, 723-733, 2014.

[19] M. Turáková, M. Králik, P. Lehocký, L. Pikna, M. Smrcová, D. Remeteiová, A.

Hudák, Appl Catal A - Gen 476 (2014) 103-112.

[20] P. R. Marqués, A. L. Perez, A. C. Corma, Chem Commun 49 (2013) 8160-8162.



137

[21] B. Amon, H. Redlingshöfer, E. Klemm, E. Dieterich, G. Emig, Chem Eng Process 38

(1999) 395-404.

[22] F. Turek, R. Geike, R. Lange, Chem Eng and Process 20 (1986) 213-219.

[23] N. Mahata, A. F. Cunha, J. J. M. Órfão, J. L. Figueiredo, Appl Cat A: Gen 351

(2008) 204-209.

[24] P. Sangeetha, K. Shanthi, K. S. Rama Rao, B. Viswanathan, P. Selvam, Appl Cat A -

Gen 353 (2009) 160-165.

[25] S. C. Shekar, K. S. Rama Rao, E. Sahle-Demessie, Appl Cat A - Gen 294 (2005)

235-243.

[26] S. K. Matam, E. H. Otal, M. H. Aguirre, A. Winkler, A. Ulrich, D. Rentsch, A.

Weidenkaff, D. Ferri, Catal Today 184 (2012) 237-244.

[27] V. Ferrer, A. Moronta, J. Sánchez, R. Solano, S. Bernal, D. Finol, Catal Today 107-

108 (2005) 487-492.

[28] B. Mile, D. Stirling, M. A. Zammitt, A. Lovell and M. Webb, J Catalysis, 114 (1988)

217-229.

[29] B. Mile, D. Stirling, M. A. Zammitt, J Mol Cat 62 (1990) 179-198.

[30] K. Al-Dalama, A. Stanislaus, Thermochim Acta 520 (2011) 67-74.

[31] D. Rihani, T. Narayanan, L. Doraiswamy, Ind. Eng. Chem. Prod. Res. Dev. 4

(1965) 403-410.

[32] G. Maria, D-N. Stefan, J Loss Prevent Proc 23 (2010) 112-126.

[33] E. Gelder, The Hydrogenation of Nitrobenzene over Metal, Dissertation presented

for the Doctor of Phylosophy degree in Chemical Engineering, University of Glasgow,

2005.

Part III

Catalytic Tests in a Tubular Reactor

141

Chapter 6 - Tubular Reactor Laboratorial Unit

This chapter covers all the details regarding the laboratorial tubular reactor unit

design and construction, as well as technical and operating details.

6.1 - Introduction

In order to minimize the dimension of reaction units, tubular catalytic reactors have

been selected in many applications, and this was the case again for nitrobenzene (NB)

hydrogenation. As the hydrogen streams, referred above, are available at CUF-QI at low

pressures, very active and selective catalysts have to be chosen. Moreover, the active

metal must be supported, so that the catalyst should be easily separated from the liquid

stream. In Part II catalysts that were tested proved to be active and selective towards

aniline (ANL). In this part (Part III) the chosen catalyst will be tested in a packed-bed

reactor and its resistance to the impurities existing in the industrial hydrogen streams will

be evaluated in the tubular reactor.

Tubular reactors or fixed bed reactors are characterized by continuous gradients of

concentration in the direction of flow that ideally approaches plug flow. The reactants are

charged continuously at one end and products are removed at the other. Normally, after

some time a steady state is attained, which represents an important fact for automatic

control and laboratory work [1]. The main advantages of this type of reactors are (i)

highest conversion per weight of catalyst of any conventional catalytic reactor, (ii) better

defined residence time of molecules in the reactor and (iii) easy to maintain as there is no

agitator or moving part [2]. Principal disadvantage is related with the difficulty of

temperature control within the reactor. This reactor configuration is especially suited for

cases needing considerable heat transfer, where high pressures and either very high or low

temperatures occur and where relatively short reaction times are needed.

The stainless steel reactor chosen is a multiphase tubular reactor, containing NB

(liquid phase) and H2 (gaseous phase) as reactants and the solid catalyst. Therefore, it is


142

crucial to take into account the competition between the reaction phenomenon and the

gas/liquid, liquid/solid mass transfer typical of this process [3].

This unit was conceived to operate in a wide temperature, pressure and feed flow

rate range. All the equipment was selected based on the type of compounds that will be

used and also knowing that it will be operated with high H2 pressure (up to 100 bar) and

moderate / high temperature (up to 300 ºC).

6.2 - Unit conception

In the research laboratory of CUF-QI, it was only available a CSTR that was used

in Part II to study the effect of the operating conditions over 1 % Pd / Al2O3 catalyst and

afterwards to carry out the screening of commercial catalysts. Nevertheless, a tubular

reactor unit was the main objective and had to be constructed. The first step consisted on

the state of the art review and design of the extended tubular process unit. This consisted

on discovering if any information about tubular reactors for NB hydrogenation was

available and finding out which companies develop/design this type of units, in order to

get in contact with them. At a laboratory or pilot scale, only Huntsman patent [4]

described a similar unit to what was pretended in this project.

Operating conditions for the laboratorial fixed-bed reactor for the hydrogenation of

NB to ANL were previously defined, based on the existing knowledge:

Feed: gas (H2) and liquid (NB+ANL).

Liquid maximum flow rate: 50 ml/min.

Gas maximum mass flow rate: 20 g/h.

Typical liquid feed composition: 10% NB + 90% ANL.

Operating temperature: 100 – 300 ºC.

Maximum operating pressure: 100 barg.

Typical catalysts mass: 50 g.

Reactor needs to be heated (and eventually cooled).

This research stage was very important and helpful since it was possible to conclude

that, despite what was thought regarding the high reaction exothermicity, it would be

Chapter 6 – Tubular Reactor Laboratorial Unit

143

possible to carry out catalytic tests without needing a cooling system (in the range of

desired operating conditions).

Since in CUF-QI a tubular reactor was available and ready to be used and there was

some previous knowledge about this kind of units, it was decided to design and construct

the unit at CUF-QI, which is also a less costly solution.

Once it was decided to construct the laboratory unit, the next step was to collect all

the information necessary to design, develop and purchase the required material,

equipment and instruments.

The development, design and construction included several stages:

- Definition of the equipment that would be necessary.

- Determination of equipment size based on design basis.

- Procurement of all the instruments and equipment.

- Piping and Instrumentation Diagram (P&ID) and Isometric Diagram.

- Unit construction, installation and commissioning.

- Preliminary tests.

6.2.1 - Unit purpose

The primary objective of the tubular reactor unit is to perform catalytic tests with

different types of supported catalysts in order to evaluate their activity and selectivity

during NB hydrogenation into ANL. The designed unit is a conventional “ atatest”

consisting mainly of: a fixed bed reactor, a system for reactants feed and a system for

products separation. All liquid samples are manually collected and analyzed by gas-

chromatography (GC).

6.2.2 - Unit description

The reaction system is a continuous flow reactor unit suitable for performing

experimental studies on catalytic hydrogenation of model compounds such as NB in a


144

tubular fixed bed reactor. The fixed bed reactor is designed for down flow operation,

although up flow can also be possible to perform.

Typical application of the unit is the evaluation of catalyst performance, namely:

Investigation of catalyst deactivation (time-on stream tests).

Comparison of catalysts from different suppliers (catalyst

screening).

Study of the effect of temperature, pressure, space velocity,

reactants concentration and H2 partial pressure on the reaction

conversion, selectivity and yield (parametric analysis).

The fixed bed reactor is heated by a hot shell, which is regulated by a cascade

control. Both gas and liquid feeds are controlled by regulation instruments and

measured/recorded on-line. In the part of product separation, the reaction mixture

pressure and the outlet gas flow are measured and recorded. All the other data

(temperature, H2 flow rate and total pressure) are acquired on an executable application

developed by Termolab in Labview, exclusively for this unit.

In Figure 6.1 is presented the Piping and Instrumentation Diagram (P&ID) for the

tubular reactor unit.


145

Figure 6.1 – Tubular reactor unit P&ID.

N2

Gas-liquid

separator

Sample

collection Product

FI

14

Reactor

12

TI

10TI

11

Liquid

Feed

FI

04

PI

05

06

PI

TIC

PI

01

PI

03

PI

07

PI

08PI

09

gas

13

PI

gas

H2

Industrial H2

PI

02


146

The tubular reactor unit can be divided in four main sections:

- Liquid feed section.

- Gas feed section.

- Reaction section.

- Separation section.

Each section is composed by several instruments, such as filters, flow meters, ball

valves, needle valves, relief valves, check valves and pressure gauges, among others. In

Table 6.1 and Table 6.2 are summarized the main instruments and equipment of the unit,

respectively.

Table 6.1 – Main instruments characteristics.

Measurement type Instrument -Tag Operating conditions

Generally range used Maximum

Pressure

(Pressure transmitter PT)

PTe1 (before reactor) 2 - 14 barg 50 barg

PTs1 (after reactor)

Temperature

(thermocouple transmitter TT)

TTf1 (slave) (in oven) 60 – 140 ºC 600 ºC

TTr1 (inside

reactor)

60 – 140 ºC

800 ºC TTr2 (master) 70 – 150 ºC

TTr3 80 – 160 ºC

Gas mass flow controller &

transmitter (FT)

FTe1 (gas line inlet) 2 g/h 20 g/h

FTs1 (gas line outlet) 2 g/h

HPLC pump (Norleq) Q 2.5 – 20 ml/min 50 ml/min

Table 6.2 – Main equipment characteristics.

Equipment Volume

Temperature Pressure

Operating

conditions Maximum

Operating

conditions Maximum

Tubular reactor 300 cm3 < 200 ºC 550 ºC 2 – 14 barg 150 barg

Heat exchanger 25 cm3 < 200 ºC 550 ºC 2 – 14 barg 100 barg

Separator 964 cm3 Ambient temperature 150 ºC 2 – 14 barg 100 barg

Relief valve (reactor safety) - Ambient temperature 121 ºC 52 barg 103 barg

Relief valve (H2 line safety) - Ambient temperature 121 ºC 60 barg 103 barg

6.2.2.1 – Liquid feed section

This section is composed by a feed tank, which is loaded with the liquid reactants

before each test (e.g. NB / solvent mixture). The connection between this tank and the

reaction section is done through the HPLC pump. Although the pump has a filtering system,


147

a 60 µm filter was placed after it to guarantee that particles do not pass to the reactor. In

order to isolate and protect the feed injection system from the reactor section, a check valve

was placed (thus avoiding a back flow from the reaction to the feed section).

All the feed line, after the check valve, is heated by a heating tape that is covered by

insulating material, as illustrated in Figure 6.2, with the purpose of guaranteeing that such

feed line is at a temperature as close as possible to that of the reactor.

Figure 6.2 – Photos of the liquid feed section.

6.2.2.2 – Gas feed section

In this part of the unit there are three distinct gases: H2 and N2 (from gas cylinders)

and industrial H2. These gases are in separated lines until they get into the unit. N2 has

always an independent line until reaching the reactor (Figure 6.1), while H2 from gas

cylinder and industrial grade H2 have the same line in the unit. In both lines (H2 and N2)

there is a check valve, and they are connected just before the reactor inlet line and after both

check valves.

The H2 line is the main line; before the check valve there is a 60 µm filter to protect

existing instruments, from eventual solid impurities. There is also a pressure regulator and a

mass flow controller. After these instruments there is the check valve, a pressure transmitter


148

and a system of purges and relief valves to guarantee that the system does not go over

pressure and all the security procedures are followed. Figure 6.3 shows a photo of this

section.

Figure 6.3 – Photos of the gas section.

6.2.2.3 – Reaction section

The tubular reactor has 15 mm of internal diameter (dt) and 400 mm of length (L). In

the top it has a mixture head, where gas and liquid flows get together, mix and go into the

reactor. In this mixture head there is also a cannula that crosses the reactor and where three

thermocouples are placed inside.

According to open literature, for fixed bed catalytic reactors the idealized flow

pattern is generally well approximated when the catalyst particle size, dp, is small enough

with respect to the internal diameter of the reactor, dt, to have an essentially uniform void

fraction over the cross section of the tube [5]. So, according to the rule of thumb the

following ratio should be observed:

𝑑𝑡𝑑𝑝 ≥ 10 (6.1)

Besides this rule, it is also important to guarantee that backmixing is eliminated. The

tubular reactor usually operates in plug flow as long as length-to-tube diameter (L / dt) is

much greater than unity and the flow is turbulent. In the case of packed- or fixed-bed


149

catalytic reactors, the criterion for negligible backmixing (thus approaching ideal plug-flow

pattern) is [6]:

𝐿

𝑑𝑝 ≥ 50 (6.2)

These rules of thumb have been followed in the reactor design. Figure 6.4 shows a

photo of the reaction section, with the reactor placed inside the oven.

Figure 6.4 – Photos of the reaction section, with closed (left) and open (right) views of the oven.

The catalytic bed represents a volume of 34 cm3 and is located in the reactor center, as

shown in Figure 6.5 a). In this case, catalytic bed length, L, is 120 mm and particle diameter

of catalyst I.2 (that will be tested in Chapter 7 and 8) is dp = 1 mm, so 𝐿

𝑑𝑝 = 120; the

internal diameter of the reactor is 15 mm, so 𝑑𝑡

𝑑𝑝 = 15. Therefore, to have a catalytic bed

length of 120 mm, there is a minimum catalyst mass, which is of 10 g in the case of catalyst

I.2.

Heating and temperature control are done by a program linked to the oven

thermocouple (TTf1). As it was previously referred there are three thermocouples inside the


150

cannula (Figure 6.5 b)) and a fourth one in the oven near to the reactor wall. The oven

temperature control is performed in cascade mode where two thermocouples are used as

reference (TTf1 is the slave and TTr2 the master).

Figure 6.5 – Tubular reactor: a) reactor bed distribution and b) thermocouples positions.

The master thermocouple is in a middle position, inside the reactor cannula, while the

slave thermocouple is at the same level as the master, but outside the reactor. The other two

thermocouples are placed at the beginning (TTr1) and at the end (TTr3) of the catalytic bed.

Cascade control is a strategy that allows to handle load changes more effectively with

respect to the manipulated variable. A cascade control structure has two feedback

controllers with the output of the primary (or master) controller changing the setpoint of the

secondary (or slave) controller. There are two purposes for cascade control: (i) to eliminate

the effect of some disturbances and (ii) to improve the dynamic performance of the control

loop [7]. In this case, the setpoint is defined in the master thermocouple (TTr2), which will

define the setpoint for TTf1 and so provide the power needed to achieve TTr2. With this

type of control, it is possible to have a better and more accurate control of the reactor

temperature and also to get a better response to some perturbations of the operating

conditions that have influence in the reactor temperature.

400 mm

120 mm

100 mm

100 mm

Ca

taly

stC

arb

uru

nd

um

Ca

rbu

run

du

m

Glassspheres

Glassspheres

TTr1 TTr2 TTr3a) b)


151

6.2.2.4 – Separation section

To ensure that no catalyst particle goes forward in the unit a 60 µm filter was placed.

Pressure is measured and controlled by a pressure transmitter, which is located after the

filter and before the heat exchanger (Figure 6.1).

The reaction mixture passes in the tube of the heat exchanger (a simple tubular heat

exchanger is used) and cool water circulates in the shell, in concurrent flow. Between the

heat exchanger and the gas-liquid separator there is a zone for sample collection, where the

samples are collected and then analyzed by GC. This zone consists of a set of valves that

allows to collect samples without depressurizing the entire unit, and thus without

interrupting the test.

Gas-liquid separation occurs in the separator, which also has a water cooling system.

The liquid is collected in a nearby vessel at the same flow rate as the feed one, in order to

maintain the unit at a constant pressure and also to keep the liquid at a constant level. The

gas (mainly H2) goes through the mass flow meter and through the back pressure valve that

regulates the pressure in the reaction unit. After the back pressure valve, gaseous effluent

pressure is reduced until atmospheric pressure and the outlet gas is released in a safe and

controlled way. Figure 6.6 is a photo of all this section.

Figure 6.6 – Photos of the separation section.


152

To conclude, in Figure 6.7 is presented the final result of the design, development and

construction of the tubular reactor unit at CUF-QI.

Figure 6.7– Tubular reactor unit overview.

6.3 – Preliminary tests

After the lab unit construction, it was necessary to perform some preliminary tests (not

catalytic), with the purpose of evaluating the hydrodynamics, the possibility of occurring

non-catalysed (homogeneous) reactions and also to quantify the pressure drop in the fixed


153

bed. All these assessments are very important for the pilot unit development, in particular

the pressure drop.

In the laboratorial reactor selected catalyst(s) will be tested and its (their) activity,

selectivity and eventual deactivation will be evaluated. At first, these preliminary catalytic

tests should be carried out using model reaction mixtures and catalyst performance assessed

in a wide range of conditions. Then, several tests with different pressures and residence

times (few minutes) and at temperatures not higher than 250 ºC should be carried out in the

tubular reactor. This will be addressed in Chapter 7.

For this reactor, it will be crucial to measure and control the temperature along the

reactor, in order to avoid the formation of hot spots (because the hydrogenation reaction is

extremely exothermic and this can lead to runaway situations). The use of high temperatures

should be carefully considered because, besides the safety aspects, there is also the

possibility of selectivity decrease (with formation of undesirable by-products as CHA and

other organics).

6.3.1 - Test with catalyst support, H2O and H2

Four temperature values (TTr1, TTr2, TTr3 and TTf1), two pressure values (PTe1 and

PTs1) and two gas flow rate values (FTe1 and FTs1) are recorded during the entire test.

In this test, preliminary test1, operating conditions used were as follows:

PTe1 and PTs1 = 36 barg

FTe1 and FTs1 = 0.5 g/h

QH2O= 5 ml/min;

Oven program:


154

Figure 6.8 – Oven program for preliminary test1.

In Figure 6.9 is presented the data obtained during the test.

a)

b)

c)

d)

Figure 6.9 – Results obtained for: a) Reactor and oven temperatures, b) Reactor temperatures, c)

Pressure and d) Gas flow rate in test1.

This test was performed using H2O in the liquid feed. PTe1 and PTs1 are measured at

the entrance and at the exit of the reactor, respectively. It is possible to verify their values’

are nearly identical (Figure 6.9c)), which means that almost no pressure drop is detected in

10 ºC /min

10 ºC /min

150 ºC , 1h00

175 ºC , 1h00

5 ºC /min

200 ºC , 1h00

0 1 2 3 4 50

100

200

300

400

Tem

per

atu

re (

ºC)

time (h)

TTr1 TTr2

TTr3 TTf1

0 1 2 3 4 590

120

150

180

210

Tem

per

ature

(ºC

)

time (h)

TTr1

TTr2

TTr3

0 1 2 3 4 530

35

40

45

50

Pre

ssu

re (

bar

g)

time (h)

PTe1

PTs1

0 1 2 3 4 50

1

2

3

4

5

Gas

flo

w (

g /

h)

time (h)

FTe1

FTs1


155

the catalytic bed. Temperatures along the reactor / oven were also registered and a quick

response and stabilization are obtained when temperature is increased, as shown in Figure

6.9 a) and b) (see also temperature program in Figure 6.8). TTr1 presents lower values than

TTr2 and TTr3 that was corrected subsequently. In Figure 6.9 d) it is presented the H2 flow

(g/h) at the unit entrance and exit being visible that the gas flow is constant.

To guarantee that all the thermocouples are registering correctly all the information,

they were placed at the same level as TTr2, in preliminary test2.

6.3.2 - Test with catalyst support, ANL and H2

For preliminary test2, the operating conditions used were as follows:

PTe1 and PTs1 = 24 barg

FTe1 and FTs1 = 17.5 g/h

QANL= 5 ml/min;

Oven program:

Figure 6.10 – Oven program for preliminary test2.

Thermocouples were all at the same level, however TTr1 was still displaying a

different value from TTr2 and TTr3, Figure 6.11. After the supplier has been contacted, it

was found that TTr1 was misconfigured (Figure 6.11b) TTr1 before) and the problem was

then solved (Figure 6.11b) TTr1). Apart from this problem, different values between

thermocouples are expected since it is not an isothermal reactor.

10 ºC /min

5 ºC /min

150 ºC , 2h00

200 ºC , 2h40


156

a)

b)

c)

d)

Figure 6.11 – Results obtained for a) Reactor and oven temperatures, b) Reactor temperatures, c)

Pressure and d) Gas flow rate in test2.

No major differences were detected between feeding H2O (test1) or ANL (test2). Once

more, pressure drop was not detected and the gas flow rate presented the same behavior (0.5

g/h or 17.5 g/h). Figure 6.11 c) and d) demonstrate slight oscilations that ate due to

sampling. Another important conclusion is that non-catalyzed homogeneous reactions were

not observed when using ANL as reactant (as inferred from the sample analyses).

At this point no problems related to temperature control were detected and all the

measured parameters presented the expected behavior. As no operating issues were

detected, the unit was considered ready to work in real conditions, with a real feed.

0 1 2 3 4 50

100

200

300

400

Tem

per

atu

re (

ºC)

time (h)

TTr1 TTr2

TTr3 TTf1

0 1 2 3 4 5120

150

180

210

Tem

per

atu

re (

ºC)

time (h)

TTr1

TTr2

TTr3

TTr1 before

0 1 2 3 4 520

22

24

26

28

30

Pre

ssu

re (

bar

g)

time (min)

PTe1

PTs1

0 1 2 3 4 515

16

17

18

19

20

Gas

Flo

w (

g /

h)

time (h)

FTe1

FTs1


157

References

[1] S. Walas, Chemical Process Equipment – Selection and Design, Chapter 17, page 583,

2nd Edition, Elsevier Inc, 2005.

[2] V. Ranade, R. Chaudhari, P. Gunjal, Trickle Bed Reactors - Reactor Engineering &

Applications, page 9, Elsevier, 2011.

[3] M. Machado, Fundamentals of Mass Transfer and Kinetics for the Hydrogenation of

Nitrobenzene to Aniline, Air products and Chemicals, No 1, 2007, 1-14.

[4] C. Mitchell, D. Stewart, Process fot the production of aromatic nitro compounds into

amines, WO Patent 113491, 2011.

[5] G. Froment, K. Bischoff, Chemical Reactor Analysis and Design, Part II, Chapter 9 page

395, John Wiley & Sons, 1979.

[6] J. Carberry, Chemical and Catalytic Reaction Engineering, Chater 4 page 170, Dover

Pucblications, 2001

[7] W. Luyben, Process modeling simulation and control for chemical engineers, Part III,

Chapter 8 page 255, 2nd Edition, McGraw-Hill, 1990.

159

Chapter 7 - Hydrogenation of Nitrobenzene in a

Tubular Reactor: Parametric Study of the

Operating Conditions Influence

Abstract

Industrially, nitrobenzene (NB) hydrogenation into aniline (ANL) is usually made

in slurry or fixed-bed reactors being more common to use slurry reactors for liquid phase

reactions. In fact, fixed bed reactors for liquid phase hydrogenation, have not experienced

an extensive industrial-scale development, mostly due to the difficulty in removing the

heat generated during reaction. If not removed, this heat can lead to runaway situations

lowering process performance and catalyst lifetime. On the other hand, this type of

reactors is the most appropriate to use if the objective is to minimize the dimension of

reaction units and to use more active catalyst. At laboratory scale, fixed-bed reactors have

not been so explored (at conditions near to the industrial ones) and literature available

only reports catalytic tests at lower pressures or temperatures. Nevertheless, in this

chapter NB hydrogenation in liquid phase was carried out in a tubular fixed-bed reactor

using a Pd/Al2O3 commercial catalyst. The influence of some operating conditions was

analyzed either by assessing catalyst performance in conversion of NB or by selectivity

towards ANL and secondary products. It was found that catalyst age is extremely

important as it changes the selectivity to the products formed, while NB conversion

remains stable. Although ANL selectivity increases with the catalyst use, selectivity to the

formation of secondary products (initially representing as much as 50 % of the products

formed) has a huge decrease until almost disappearing after ca. 100 h of time-on-stream.

This is as important finding even from an industrial prespective. In what concerns the

operating conditions influence (temperature, pressure, NB concentration and liquid feed

flow rate), it was found that temperature and pressure are the most important parameters.


160

7.1 Introduction

In liquid phase, hydrogenation reactions are usually carried out in slurry reactors,

being less common the use of fixed-bed reactors [1]. Catalysts used in the hydrogenation

of nitro compounds are typically palladium, platinum and nickel, supported or not [2 - 4].

Králik et al. [5] reported that one factor that has influence in the lifetime of the catalyst is

its mechanical deterioration that subsequently will limit the reactor selection.

Fixed-bed reactors have not experienced an extensive industrial-scale development

in the case of nitrobenzene (NB) hydrogenation due to the difficulty of removing heat

generated during reaction that can lead to runaway situations, hot-spots and a decrease in

process performance. Nevertheless, this type of reactor is the most appropriate to use

when the objective is to minimize the dimension of reaction units while using more active

catalysts that afterwards do not need to be separated from the reaction mixture [6].

Another advantage of this configuration is related to the well-specified residence time

with minimum back-mixing [7]. In fact, it is one of the most common reactor

configurations used in other hydrogenations reactions, such as of cumene hydroperoxide

[8], 1,3 butadiene at high conversions [9], or acetylene in large-scale, which is generally

conducted in a series of two adiabatic, fixed-bed reactors to minimize the temperature

increase through the bed [10]. Fixed bed reactors are also appropriate for benzene

hydrogenation, although some authors verified that mass transfer limitations appeared to

have a considerable impact on the reactor performance [11]. In fixed-bed reactors usually

gas and liquid phases flow downward but for some selected reactions they can flow also

upward, although some studies also focused their attention in the development of a

periodic modulation of gas or liquid flow rate with significant improvements in the

reactor performance, namely in the hydrogenation of either 2-ethylanthraquinones [12] or

phenylacetylene [13].

Aniline (ANL) is a major chemical product that can be produced by many routes.

However, catalytic hydrogenation of NB is the one that dominates and gives the highest

selectivity [14]. Most of the processes are carried out in gas-phase in catalytic fixed-bed

reactors while liquid-phase involves suspended, highly active metal-supported catalysts

(slurry reactors) [4]. Nevertheless, new configurations for ANL production through NB

hydrogenation in liquid-phase have been studied. For instance, in 1995 Peureux et al. [15]

tried this reaction in a membrane reactor and concluded that it is an active contactor

Chapter 7 - Hydrogenation of Nitrobenzene in a Tubular Reactor –

Parametric Study of the Operating Conditions Influence

161

between gaseous and liquid reactants. A microstructured falling film reactor with Pd

catalyst deposited as films was also tested for the hydrogenation of NB to ANL in ethanol

and proved to be feasible, although deactivation has been detected mainly caused by the

formation of organic compounds on the catalyst surface and due to Pd loss [16].

The use of tubular reactors in liquid-phase ANL production has also been studied,

but the number of reports found is quite limited. A different tubular catalytic apparatus in

which the heat release and heat-exchange surfaces are not spatially separated was

presented by Kirillov et al. [17] and was applied to NB hydrogenation (coolant

temperature 200 ºC, feed reagent ratio: 0.86 / 0.14 and gas velocity 1 m/s) obtaining

conversions of 98.2 %; they also concluded that higher NB concentrations could be used

under compatible conditions. Du et al. [18] worked on carbon nanofiber coated monoliths

for three-phase NB hydrogenation. Reaction with monolith catalyst was performed in a

continuous flow reactor (stainless steel tube, 420 mm length and 6.5 mm inner diameter)

and its performance was evaluated. For testing a catalyst based on ruthenium, Bombos et

al. [19] used a continuous fixed bed catalytic reactor in the total pressure range of 10-40

bar, temperature range of 45-75 ºC, in concurrent with downward flow of reagents,

having concluded that temperature rise favors the increase of the NB conversion and yield

in total aromatic compounds and in ANL.

In the present work, the main goal is to study and evaluate NB hydrogenation

reaction in liquid-phase in a tubular fixed-bed reactor using an active and selective

catalyst that was chosen in Chapter 5. An important issue is related with catalyst

performance along the catalytic tests in order to determine if there is any deactivation.

Moreover, a parametric study will be effectuated aiming the determination of the

operating conditions impact on both NB conversion and aromatic compounds production

(selectivity towards ANL and secondary products).


Hydrogenation of NB, in liquid phase, was carried out in a tubular reactor with 15

mm of internal diameter and 400 mm long (with a catalytic bed of 120 mm long), in

continuous downflow mode (under H2 pressure). The catalyst used was a commercial 0.3

wt.% Pd/Al2O3 material in extrudate form, Catalyst I.2. A known amount of Pd/Al2O3


162

catalyst was loaded into the reactor, Figure 7.1, and the material pre-treated, in situ. Pre-

treatment of the catalyst was firstly performed at 150 ºC and under hydrogen pressure (20

barg), gas flow between 2 g/h, for 2 hours. As detailed before, in Chapter 5, temperature-

programmed reduction (TPR) experiments have shown that the catalyst used was fully

reduced under the pre-treatment conditions employed. Chapter 6 contains more details

about the experimental set-up.

Figure 7.1 – Scheme of the tubular reactor used for the catalytic tests.

All the tests were performed with the same sample of catalyst, about 10 g. The

catalytic bed, at the center of the reactor, was positioned between two layers of SiC (21 g

each, granulometry of 1.68 mm) and glass spheres. The upper layer of spheres and SiC

served both as a mixer for the reactants and as a pre-heater. Moreover, SiC layer had the

objective of equalizing temperature along the reactor and to help avoiding hot-spots.

Temperature regulation is made by an oven of Termolab equipped with one

thermocouple that regulates the reactor heating. Inside the reactor there is a cannula

Aromatic

compound

FeedH2

Sample

collection Product

Hydrogenated

aromatic compound

ReactorCatalytic

Bed

SiC

SiC

Glass spheres

Glass spheres



163

where three thermocouples where positioned, being that the one in the middle carries the

control of temperature in the reaction zone (cf. more details in section 6.2.2.3).

H2 (ALPHAGAZ 1 – AirLiquide) is supplied from a gas cylinder. Liquid feed and

gas are mixed before entering the reactor, being fed by the top. Products leave the reactor

at the base, so it has a down-flow regime. The liquid reactant is pumped to the reactor

with a HPLC pump, JASCO PU-2087, at a required flow rate. The total pressure inside

the reactor is kept constant along each run using a back pressure regulator, that guarantees

a constant exit gas flow. The inlet gas flow however is dependent of the H2 consumption

in the unit, being supplied with an excess of about 90 %.



flame ionization detectors (FID). The column used was a HP-1 (100%


detector was 250 ºC, the pressure in the column was 14 bar and helium was used as

carrier gas. The column oven was temperature-programmed with a 1 min initial hold at

120 ºC followed by an increase until 230 ºC (15 ºC min-1 rate) and then kept at 230 ºC for

9 min.


Calibration curves were plotted for all the analyzed compounds which were easily



The reference values for temperature, pressure and nitrobenzene concentration

during the parametric study are: T = 120 ºC, P = 14 barg and Cref = 1.2 wt.% NB. The

experiments performed and the conditions used are given in Table C.1 (Appendix C).

The NB conversion was calculated based on the data obtained from GC analysis:

𝑋𝑁𝐵 = 𝑁𝐵0 (𝑝𝑝𝑚) − 𝑁𝐵𝑜𝑢𝑡,𝑡 (𝑝𝑝𝑚)

𝑁𝐵0 (𝑝𝑝𝑚) (7.1)

where 𝑁𝐵0 (𝑝𝑝𝑚) is the reactor feed NB concentration (ppm) and 𝑁𝐵𝑜𝑢𝑡,𝑡 (𝑝𝑝𝑚) is the NB

concentration at the reactor outlet at any time instant t.


164

The liquid phase analysis confirmed the presence of the following compounds: NB,

ANL, as well as the by-products cyclohexylamine (CHA), cyclohexanol (CHOL),

cyclohexanone (CHONA), N-cyclohexylaniline (CHANIL), dicyclohexylamine

(DICHA), cyclohexyldeneaniline (CHENO) and benzene (Bz) – cf. Nomenclature

section. Nevertheless, the secondary products will not be presented individually but in

groups: light products – Bz, CHA, CHOL and CHONA; and heavy products: DICHA,

CHENO and CHANIL. It is important to refer that ANL, used as solvent, is of industrial

grade, thus containing some by-products.

The values used, for results discussion, correspond to what is formed during

reaction, therefore, to each point the calculation made is as follows:

𝐹𝑖 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) = 𝐹𝑖 𝑒𝑥𝑖𝑡,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) − 𝐹𝑖,0(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (7.2)

where Fi represents the molar flow rate of species i (where i might be ANL, secondary

products or its division: light and heavy products), being the amount “formed” obtained

by the difference between the one that “exits” the reactor and what is fed (subscript “ ”)

Selectivity towards ANL, light products, heavy products and secondary products is

based on the amount of each product/group of products produced as compared to all

products formed during the reaction:

𝑆𝑖 =𝐹𝑖 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡 (𝑚𝑜𝑙 𝑚𝑖𝑛⁄ )

∑𝐹𝑗 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡 (𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (7.3)

∑𝐹𝑗 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡 (𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) = 𝐹𝑠𝑒𝑐𝑜𝑛𝑑𝑎𝑟𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) + 𝐹𝐴𝑁𝐿 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (7.4)

𝐹𝑠𝑒𝑐𝑜𝑛𝑑𝑎𝑟𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) = 𝐹𝑙𝑖𝑔ℎ𝑡 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) + 𝐹ℎ𝑒𝑎𝑣𝑦 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (7.5)


In Chapter 5, a catalytic screening study was performed with the objective of

defining and choosing among the commercial catalysts available which one was the most



165

active with higher selectivity to ANL and lower secondary products formation, within the

ranges of operating conditions defined.

Among all the catalysts tested it was found that catalyst I.2, with 0.3 wt. %

Pd/Al2O3, is the one with the highest NB consumption rate and with a low secondary

products formation.

Therefore, a parametric study for assessing the influence of the main operating

conditions was now performed in a fixed-bed reactor, with a new sample of the 0.3 wt. %

Pd/Al2O3 catalyst being tested for approximately 162 h of time-on-stream. Temperature,

pressure, liquid feed flow rate and feed NB concentration were varied and their impact in

the catalyst performance was analyzed.

An overview of the catalyst performance for all the runs carried out is displayed in

Figure 7.2 (each data point corresponds to steady-stade conversion in a given test of

Table C.1). It can be observed that NB conversion remains nearly constant along the

reaction time (Figure 7.2a)) – being that time corresponds to the accumulated duration of

runs in Table C.1, i.e. it represents catalyst age. Some points indicated lower NB

conversions, but they correspond to some of the perturbations that were carried out in the

parametric study – cf. condictions in Table C.1, runs TR4, TR5, TR6, TR9, TR10 and

TR13 and in Table C.2, runs TR14, TR15 and TR17.


166

a)

b)

Figure 7.2 – Evolution of a) NB conversion and b) Selectivity to ANL and secondary products, as

a function of reaction time for all tests of the parametric study.

Although NB conversion remains constant, data obtained for selectivity present a

different tendency. Analyzing Figure 7.2b), it becomes evident that selectivity to ANL

increases with time-on-stream, while that to secondary products decrease (with a few

exceptions due to the reasons mentioned above); that means that with its use, the catalyst

becomes more selective towards ANL and secondary products formation is nearly null

after ca. 100 h. In the point of view of this study, this is an important fact since the goal is

to industrially produce ANL with a high purity, which means low secondary products

formation, mainly DICHA (due to the difficulty in separating it from ANL in the

downstream distillation process).

0 20 40 60 80 100 120 140 1600

20

40

60

80

100

NB

Conver

sion (

%)

time (h)

0 20 40 60 80 100 120 140 1600

20

40

60

80

100

Sel

ecti

vit

y (

%)

time (h)

S ANL

S secondary products



167

An important conclusion from the results up to now is related with the age of

catalyst. In the beginning NB conversion is 100 % but selectivity to secondary products is

predominant; thus results obtained in the first hours will be discarded. On the other hand,

for runs carried out in short periods of time, results are comparable, and the effect of

some parameters can be compared and analyzed.

7.3.1 Reproducibility tests

In order to verify if, at the same operating conditions, the Pd-based catalyst presents

the same performance, or in other words, if NB conversion is identical, independently of

the catalyst age, several reproducibility tests were carried out. Therefore, several tests

were effectuated in reference conditions (120 ºC, 14 bar, 5 ml/min and NB concentration

1.2 wt.%), presented on Table C.1 of Appendix C (highlighted tests), intercalated with

other in different conditions (parametric study).

Results obtained for this reproducibility tests are presented in Figure 7.3,

corresponding to the runs described above.

Figure 7.3 - Evolution of NB conversion as a function of reaction time for the reproducibility

tests.

Analyzing the results in Figure 7.3, it is possible to conclude that catalyst activity

remains nearly stable (NB conversion) along the tests, with values in the range 95.6 –

0 20 40 60 80 100 120 140 16080

85

90

95

100

NB

Co

nv

ersi

on

(%

)

time (h)


168

99.8 % (although selectivity to the compounds formed during reactions changes over

time, as mentioned above, for the reasons described below).

During the tests some parameters are monitored and registered, namely temperature

in different positions along the catalytic bed, total pressure at the reactor inlet and outlet,

and H2 flow rate. Data are collected and analyzed and the results are presented in Figure

7.4, which correspond to the runs TR5b) and TR10a), cf. Appendix C.

a)

b)

c)

d)

Figure 7.4 - Evolution of a) Temperature of thermocouple TTr2, b) Pressure, c) NB conversion

and d) H2 consumption in transient state for reproducibility tests TR5a) and TR10a).

This merely illustrative example is valid for all the tests performed: temperature is

stable during the tests as well as total pressure, without any significant pressure drop in

the catalytic bed. Even when conditions are changed, system response is fast and

stabilization is achieved after a short time. In the case of H2 consumption, it corresponds

to the difference between gas flow measured before and after the reactor, ant it can be

seen that even with a difference of 51 h (catalyst age at TR5b) = 39.8 h and at TR10a) =

90.8 h) results are reproducible.

0.0 0.5 1.0 1.5 2.0115.0

117.5

120.0

122.5

125.0

Tem

per

atu

re (

ºC)

time (h)

TR10a)

TR5b)

0.0 0.5 1.0 1.5 2.010.0

12.5

15.0

17.5

20.0

Pre

ssure

(bar

g)

time (h)

TR5b) in

TR5b) out

TR10a) in

TR10a) out

0.0 0.5 1.0 1.5 2.080

85

90

95

100

NB

Co

nv

ersi

on

(%

)

time (h)

TR5b)

TR10a)

0.0 0.5 1.0 1.5 2.00.00

0.25

0.50

0.75

1.00

H2 c

on

sum

pti

on

(g

/ h

)

time (h)

TR5b)

TR10a)



169

In the sections below, the effect of some relevant operating conditions will be

presented, using the data when the reactor has reached steady-state (usually ca. 2 h only

are required after changing operating conditions).

7.3.2 Influence of Total Pressure

Influence of total pressure in the catalytic hydrogenation of NB is relevant as it was

concluded in the studies presented in Chapter 3 and Chapter 5 for a slurry stirred batch

reactor. In this section influence of total pressure will be studied and this will be done

through the H2 pressure increase, which means that more H2 will be available in the gas

phase and also solubilized in the reaction mixture.

In Figures 7.5 and 7.6 are presented the results of series TR4 and TR6 that include

runs TR4d) to f) and TR6b), where the conditions used were: 120 ºC, 5 ml/min, 1.2 wt. %

NB in ANL and 4 barg (TR6b)), 7 barg (TR4f)), 10barg (TR4e)) and 14 barg (TR4d)). It

is observed that NB conversion increases with the pressure while a strong decrease in

ANL selectivity (due to an increase in secondary products formation) was found.

a)

b)

Figure 7.5 - Evolution of a) NB conversion and b) selectivity to ANL for different total pressures.

2 4 6 8 10 12 1470

80

90

100

NB

Co

nv

ersi

on

(%

)

Pressure (barg)

2 4 6 8 10 12 1450

60

70

80

90

100

AN

L S

elec

tiv

ity

(%

)

Pressure (barg)


170

a)

b)

Figure 7.6 - Evolution of a) selectivity to secondary products and b) Secondary products

selectivity distribution for different total pressures.

NB conversion clearly increases from 4 barg to 7 barg, being this gain less

notorious between 7 barg and 14 barg because nearly complete conversion was reached.

Further, selectivity to reaction products formed during reaction presents two opposite

behaviors: selectivity to ANL decreases with the increase of pressure and inherently

selectivity to secondary products increases, which are mainly composed by heavy

products (the amount of light products amount represents only 13 - 20 % of all products

formed). According to Figure 7.7, the reaction network for ANL and secondary products

formation proposed in Chapter 3, heavy products are mostly formed through ANL

hydrogenation (ANL condensation reactions), as it also occurs with Bz and CHA (over-

hydrogenated and hydrogenolytic substances). Some tests demonstrated that CHONA

only appears in high quantities when NB is present (data not shown). This means that

when only ANL is hydrogenated, CHONA concentration is very small or not detected.

Therefore, it can be concluded that with this Pd-based catalyst, CHONA is essentially

formed through NB hydrogenation. CHOL is formed through CHONA hydrogenation,

consequently, it will be produced only when CHONA is formed.

2 4 6 8 10 12 140

10

20

30

40

50

Sel

ecti

vit

y (

%)

Pressure (barg)

S light products

S heavy products


4 barg 7 barg 10 barg 14 barg0

10

20

30

40

50

Sel

ecti

vit

y (

%)

S Bz S CHA

S CHOL S CHONA

S DICHA S CHENO

S CHANIL



171

Figure 7.7 - Reaction network proposed for ANL and secondary products formation including Bz


In the first place, if there is more H2 available, i.e., solubilized in the liquid phase

(as a consequence of pressure increase), ANL production should increase. However, a

decrease in ANL formation is verified. On the other hand, secondary products selectivity

increases with pressure, mainly heavy products and CHA (that results from ANL deep

hydrogenation); so, it can be concluded that the decrease in ANL formation is due to its

reaction and hydrogenation into other compounds. Probably, ANL formation also

increases with pressure (as NB conversion increases) but since there is more H2 available

in the reaction medium, the reaction will proceed and ANL will be transformed mostly in

CHANIL, CHENO and DICHA.

In the case of light products, Bz and CHA should also appear due to ANL

transformation, Figure 7.7. CHONA results from NB direct hydrogenation and

deamination and CHOL from CHONA hydrogenation. If there is more H2 available, it is

expected that CHONA concentration increases because reaction will be shifted into that

direction.

Heavy products

Light products

Light products

NO2

Cat

+H2

NH2

+ NH3

H2

Cat

H2

+ANL

-NH3

Cat

H2

NH2

NHNH2

+H2 -NH3

NH

+H2

NH

O

+H2

-NH3

+ANL -H2O

N

+H2

+H2OH CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

Bz

ANLNB

Amine


-H2O

-H2O

*-NH3

+H2O


172

Heavy products formation is higher for all range of pressure studied. However,

analyzing selectivity values, light products augmentation with pressure is by a factor of 4

(2 % at 4 barg to 8 % at 14 barg) while for heavy products this increase is of an order of

2.5 (13 % at 4 barg to 32 % at 14 barg). From an industrial point of view, the important is

the overall selectivity, that is higher for heavy products which mean that ANL that is

produced, is being consumed.

It can be concluded that with pressure increase, although ANL formation increases,

its conversion rate to secondary products also increases.

7.3.3 Influence of Temperature

Several runs were performed at different temperatures, 75ºC (TR4c)), 100 ºC

(TR3d)), 120 ºC (TR3c)) and 150 ºC (TR3a)), but at the same: pressure, 14 barg; liquid

feed flow rate, 5 ml/min; and NB concentration diluted in ANL in the feed, 1.2 wt. %.

Results at steady-state are shown in Figure 7.8 and Figure 7.9. In this case it was chosen

to analyse catalyst performance with an age of 1.1 days, but in Appendix D - Figure D.1

are presented the same graphics for the catalyst with an age of 2.4 days; conclusions are

however the same. Main differences are related with selectivity - with 2.4 days’ catalyst is

more selective to ANL, with less formation of secondary products.

a)

b)

Figure 7.8 - Evolution of a) NB conversion and b) selectivity to ANL for different temperatures

at 14 barg.

70 80 90 100 110 120 130 140 15060

70

80

90

100

NB

Co

nv

ersi

on (

%)

Temperature (ºC)

70 80 90 100 110 120 130 140 15040

50

60

70

80

90

100

AN

L S

elec

tiv

ity

(%

)

Temperature (ºC)



173

a)

b)

Figure 7.9 - Evolution of a) selectivity to secondary products and b) Secondary products

selectivity distribution for different temperatures at 14 barg.

As shown in Figure 7.8a), an increase in NB conversion is noticed when

temperature is increased from 75 ºC to 100 ºC, reaching almost 100%. Analyzing

selectivity results, and as it was expected, temperature raise leads to an enhancement in

secondary products formation, principally for heavy products.

Catalyst behavior due to temperature variation is similar to that of pressure.

However, in terms of amounts of products formed this influence is more noticed in the

temperature range of 120 ºC to 150 ºC because of the higher formation of secondary

products. Selectivity to light products presents a slight enhancement from 100 ºC to 120

ºC and then remains nearly constant, up to 150 ºC whereas heavy products formation

strongly augments with temperature variation in the whole tested range, mainly due to the

formation of CHANIL. While light products selectivity increases 3.2x (2.2 % at 75 ºC to

7 % at 150 ºC) heavy products selectivity presents an increase of 18x (2.5 % at 75 ºC to

45 % at 150 ºC).

The results obtained so far help realizing that with this type of reactor it is possible

to work in mild conditions (100 ºC) having almost 100% of NB conversion. On the other

hand, catalyst selectivity to ANL decreases for higher temperatures, and this might be an

important drawback from an industrial point of view. Usually, in industrial conditions,

reaction temperature is higher in order to be valorized through vapor production. In this

case, to have a high selectivity to ANL, reaction temperature should be lower, not being

possible to produce vapor.

70 80 90 100 110 120 130 140 1500

10

20

30

40

50

60

Sel

ecti

vit

y (

%)

Temperature (ºC)

S light products

S heavy products


75 ºC 100 ºC 120 ºC 150 ºC0

10

20

30

40

50

60

Sel

ecti

vit

y (

%)

S Bz

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL


174

7.3.4 Influence of Liquid Feed Flow Rate

Liquid feed flow rate is another parameter that can have a significant impact in the

NB hydrogenation reaction. The increase in liquid feed flow leads to a decrease in the

residence time (τ), which in turn implies that NB will have less time to be in contact with

the catalytic active sites, and thus, it is expected that NB conversion will decline.

Figure 7.10 presents results obtained for test TR9 that includes runs TR9a) to

TR9d) with liquid feed flow rates of 2.5 ml/min, 5.0 ml/min, 12.5 ml/min and 20 ml/min,

respectively, and where other operating conditions used were 150 ºC, 14 barg and 1.2 wt.

% of NB in ANL feed. The same test was performed at 120 ºC and the results can be seen

in Appendix D, Figure D.2.

a)

b)

c)

d)

Figure 7.10 - Evolution of a) NB conversion, b) selectivity to ANL, c) selectivity to secondary

products and d) secondary products selectivity distribution, for different feed flows rates at 150ºC

and 14 barg.

As it was expected, NB conversion decreases with the feed flow rate; conversely,

selectivity to ANL increases with the increase of the feed flow rate. For secondary

0 5 10 15 2060

70

80

90

100

NB

Co

nv

ersi

on

(%

)

Liquid Feed flow (ml / min)

0 5 10 15 2040

50

60

70

80

90

100

AN

L S

elec

tivit

y (

%)


0 5 10 15 200

10

20

30

40

50

Sel

ecti

vit

y (

%)


S light products

S heavy products


2.5 ml/min 5 ml/min 12.5 ml/min 20 ml/min0

10

20

30

40

50

Sel

ecti

vit

y (

%)

S Bz

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL



175

products, it is observed that the impact is higher in the formation of heavy compounds,

essentially CHANIL. Besides DICHA is also detected in considerable amounts at low

feed flow rates and this is one of the compounds which formation should be avoided.

Therefore, in order to produce ANL with low formation of secondary products, at 150 ºC,

flow rates equal to or higher than 5 ml/min should be used. At 120 ºC (Appendix D,

Figure D.2) the same tendency was obtained although a small decrease in ANL selectivity

was detected for 20 ml/min, mainly due to the formation of CHONA.

These results also prove that secondary products formation is most likely due to the

hydrogenation of ANL formed in the reaction and that for higher residence times, ANL

instead of being released from catalyst active sites remains there and so the hydrogenation

reaction proceeds (ANL Secondary products).

7.3.5 Influence of NB Concentration in the Feed

The last parameter to be analysed is the NB concentration in the reactor feed. All

the runs were performed in the same operating conditions, 120 ºC, 14 barg, 5 ml/min but

with different NB loadings (diluted in ANL). NB concentration was varied between 1.2

wt. % to 8.2 wt %: TR13c) 1.2 wt. % NB, TR15a) 1.7 wt. % NB, TR14a) 2.4 wt. % NB,

TR10b) 4.2 wt. % NB and TR17b) 8.2 wt. % NB) – Table C.2. Results are shown in

Figures 7.11 and 7.12.

a)

b)

Figure 7.11 - Evolution of a) NB conversion and b) selectivity to ANL for different NB

concentrations at 120 ºC and 14barg.

0 2 4 6 8 1060

70

80

90

100

NB

Co

nv

ersi

on

(%

)

NB concentration (%)

0 2 4 6 8 1080

85

90

95

100

AN

L S

elec

tivit

y (

%)



176

a)

b)

Figure 7.12 - Evolution of a) selectivity to secondary products and d) Secondary products

selectivity distribution, for different NB concentrations at 120 ºC and 14barg.

Although NB conversion markedly decreases for higher NB concentrations

(because more NB is being fed to the reactor), ANL selectivity remains nearly constant

(selectivity to ANL, in the operating conditions studied, is about 98 %). For secondary

products, selectivity is therefore quite low in the conditions tested (< 3 %). CHONA is the

main product formed.

In this case, attention should be centred also in the reactor temperature because

when increasing the NB concentration in the feed, more heat will be released. NB

hydrogenation is extremely exothermic and to stabilize the temperature in the desired set-

point (120 ºC), more time will be necessary to achieve that set-point. Nevertheless, all the

tests were carried out at the desired set-point, 120 ± 3 ºC (data not shown). Beyond that,

the possibility of hot-spots occurrence is higher for higher NB concentrations and

consequently an additional care should be taken into account.

7.4 Conclusions

Catalytic NB hydrogenation into ANL was studied in a tubular fixed-bed reactor.

This type of reactor configuration, for this reaction in liquid phase, is not very common

and the information available in literature is scarce. This study represents therefore a setp

forward in this technology.

Using the catalyst chosen in Chapter 5 (0.3 wt.% Pd/Al2O3), a parametric study was

done in which the influence of pressure, temperature, liquid feed flow rate and NB

concentration were evaluated at steady state. Catalyst lifetime and performance were also

0 2 4 6 8 100

1

2

3

4

5

Sel

ecti

vit

y (

%)


S light products

S heavy products


1.2 wt. % NB 1.7 wt. % NB 2.4 wt. % NB 4.2 wt. % NB 8.2 wt. % NB0

1

2

3

4

5

Sel

ecti

vit

y (

%)

S BZ

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL



177

analyzed; it was concluded that globally, NB conversion remains constant along the

reaction time (along all the tests carried out, corresponding to ca. 160 h of time-on-

stream). However, data obtained for selectivity present a different tendency: with its use,

the catalyst becomes more selective towards ANL and secondary products formation is

nearly inexistent after about 100 h. This conclusion is of a great importance either for this

study as from an industrial point of view, since the goal is to produce ANL with a high

purity, which means low secondary products formation, mainly DICHA.

The effect of operating conditions can be divided in two groups: i) pressure and

temperature and ii) liquid feed flow rate and NB concentration in the feed. In the first

group, an increase of the parameters has a positive influence in NB conversion but not in

ANL selectivity; in fact, with the increase of either pressure or temperature, ANL

selectivity decreases and secondary products formation increases. In the second group, a

decrease in NB conversion is noticed for both parameters (either increasing liquid feed

flow rate or NB concentration), but ANL selectivity increases with flow rate and remains

approximately constant for different feed NB concentrations (in the conditions tested).

This study allowed to understand that, also in a fixed-bed tubular reactor, operating

conditions should be chosen carefully and their effects balanced taking into account the

tradeoff between high NB conversion and high selectivity to ANL with minimum

secondary products formation. Temperature and pressure are the parameters with

paramount importance in secondary products selectivity. It was also possible to verify

that the reaction network proposed in Chapter 3, for ANL and secondary products

formation, is in agreement with the results obtained.


178

References




[2] PERP Program – New Report Alert, August 2003, Nexant Chem Systems.

[3] Downing, R., Kunkeler, P., van Bekkum, H., Catalytic systhesis of aromatic amines,


[4] Wegener, G., Brandt, M., Lothar, D., Hofmann, J., Klesczewski, B., Koch, D., Kumpf,


catalysis in the polyurethane industry, Applied Catalysis A – General 221 (2001) 303-

335.

[5] Králik, M., Turáková, M., Macák, I., Lehocký, Aniline – Catalysis and Chemical

Engineering, In Proceedings of the 41st International Conference of Slovak Society of

Chemical Engineering, Tatranské Matliare, Slovakia, 723-733, 2014.

[6] Datsevich, L., Muhkortov, D., Multiphase fixed-bed technologies Comparative

analysis of industrial processes (experience of development and industrial

implementation, Applied Catalysis A: General 261 (2004) 143-161.

[7] Eigenberger G., Ruppel, W., Catalytic Fixed-Bed Reactors, Ullmann's Encyclopedia

of Industrial Chemistry, Wiley VCH, 2012.

[8] Ma, Y., Zun, Q., Cumene Hydroperoxide Hydrogenation on a Pd/Al2O3 Catalyst in a

Trickle Bed Reactor – Kinetics of Hydrogenation and Deactivation, Chemical

Engineering Technology 35 (2012) 1849-1856.

[9] Pérez, D., Oliver-Fuentes, C., Curbelo, S., Rodríguez, M., Zeppieri, S., Study of the

selective hydrogenation of 1,3-butadiene in three types of industrial reactors, Fuel 149

(2015) 34-45.

[10] Bartholomew, C., Farrauto, R., Hydrogenation and Dehydrogenation of Organic

Compounds, in Fundamentals of Industrial Catalytic Processes, 2nd Edition, John Wiley &

Sons, Inc., Hoboken, 2005.

[11] Metaxas, K., Papayannakos, N., Kinetics and Mass Transfer of Benzene

Hydrogenation in a Trickle-Bed Reactor, Industrial & Engineering Chemistry Research

45 (2006) 7110-7119.

[12] Liu, G., Duan, Y., Wang, Y., Wang, L., Mi, Z., Periodically operated trickle-bed

reactor for EAQs hydrogenation: Experiments and modelling, Chemical Engineering

Science 60 (2005) 6270-6278.

[13] Wilhite, B., Huang, X., McCready, M., Varma, A., Effects of Induced Pulsing Flow

on Trickle-Bed Reactor Performance, Industrial & Engineering Chemistry Research 42

(2003) 2139-2145.

http://onlinelibrary.wiley.com/doi/10.1002/14356007.b04_199.pub2/abstract



179

[14] Rase, W., Hydrogenation of Nitroaromatics, Nitrobenzene Aniline, Handbook of

Commercial Catalyst: Heterogeneous Catalysts, CRC Press LLC, 2000.

[15] Peureux, J., Torres, M., Mozzanega, H., Giroir-Fendler, A., Dalmon, J.,

Nitrobenzene liquid-phase hydrogenation in a membrane reactor, Catalysis Today 25

(1995) 409-415.

[16] Yeong, K., Gavriilidis, A., Zapf, R., Hessel, V., Catalyst preparation and

deactivation issues for nitrobenzene hydrogenation in a microstructure falling film

reactor, Catalysis Today 81 (2003) 641-651.

[17] Kirillov, V., Kuzin, N., Mescheryakov, V., Droboshevich, V., Catalytic heat-

exchanger reactor for strongly exothermic reactions, Chemical Engineering Science 56

(2001) 381-386.




Chemistry 21 (2015) 997-1004.

[19] Bombos, D., Argesanu, C., Doukeh, R., Bombos, M., Vasilievici, G., Nitrobenzene

hydrogenation by catalysts based on ruthenium, Bulletin of Romanian Chemical

Engineering Society 2 (2015) 2-11.

181

Chapter 8 - Industrial Perspective of Nitrobenzene

Catalytic Hydrogenation in a Tubular Reactor –

Impure H2 valorization

Abstract

The activity and performance of a commercial Pd/Al2O3 catalyst was studied with

the final goal of being used in a fixed-bed reactor with a real stream of industrial

hydrogen, which is available at CUF-QI at low pressure. In Chapter 7, a parametric study

was performed and some important conclusions were taken: catalyst age is extremely

important, so that nitrobenzene (NB) conversion remains nearly stable along the time

while selectivity to aniline (ANL) increases. Moreover, pressure and temperature were

found to be the critical parameters. In this Chapter, in an industrial perspective of NB

hydrogenation, it was decided to evaluate the influence of the solvent and also to test if

the catalyst was still active at mild conditions of pressure and temperature. Besides, some

reaction products, namely water and cyclohexylamine (CHA), were added to the feed

mixture to determine their influence either on ANL selectivity or on secondary products

formation and distribution. It was found that even at mild conditions (75 ºC and 9 barg)

NB conversion is higher than 65 % with high selectivity to ANL (above 98%). Regarding

the effect of the studied reaction products addition to the reactor feed, in both cases (H2O

and CHA) it appears they cause a decrease in ANL selectivity but NB conversion is not

significantly affected. Hydrogen contaminated with ammonia was used to simulate the

industrial stream; it was found that NH3 presence does not affect the catalytic reaction on

terms of NB conversion but on the other hand it helps decreasing the formation of

secondary products. The industrial H2 stream was used to hydrogenate NB into ANL and

it was demonstrated that even containing contaminants and at low pressures, this stream

can be used to produce ANL. No major differences were detected on NB conversion and

selectivity to ANL, when using the pure or the industrial H2 stream.


182

8.1 Introduction

Nitrobenzene (NB) hydrogenation processes can be done in liquid or in vapor-

phase. A comparison between those two processes shows little difference in yield [1].

Common reactors for vapor-phase are the fluidized ones whereas for liquid-phase the

slurry reactors are more often used.

Fixed or fluidized-bed reactors are generally used for exothermic reactions.

Although the fixed-bed tubular reactor configuration commonly used (due to its high

efficiency and relatively low cost), it shows some limitations when removing the large

amounts of heat generated along the reactor axis [2]. On the other hand, it is the most

appropriate to use when the objective is to minimize the dimension of reaction units using

more active catalysts that afterwards do not need to be separated from the reaction

mixture [3]. Hwang and Smith [2] developed and compared various reactor

configurations with fixed-bed reactors and used nitrobenzene (NB) hydrogenation as a

case study. It was concluded that diluting the catalyst with inert material and employing

side-stream injection of NB generates large amounts of aniline (ANL) and an effective

temperature control between certain boundaries is achieved.

NB hydrogenation is the main process used for ANL production. Only small

quantities of ANL are used as final product, the main demand being the production of

isocyanates required for polyurethane synthesis [4]. As mentioned above, this reaction is

carried out both on vapor and on liquid-phase, and several catalysts have been studied for

both cases. In liquid-phase the process is generally operated at temperatures between 90 –

200 ºC and pressures of 1 to 30 bar. Relatively to the catalyst, focus have been directed to

Ni, Pd, Pt and Ru supported on C, SiO2 and Al2O3. Hassan et al. [5] worked on a Ru

supported on fullerene obtaining 100 % conversion with Ru loading of 15 wt.% at 150 ºC

and 22.33 bar with a reaction time of 180 min. Ru-Cu-Ni supported on C and Ru

supported on H-ZSM5 were also investigated and it was found that although Ru/H-ZSM5

presented a higher selectivity to ANL, Ru-Cu-Ni/C is more active [6]. Industrial Ni/SiO2

was studied by Relvas et al. [7] and results obtained indicated the possibility of the

catalyst surface changing with time due to the contact with the reaction mixture. An

amorphous Ni-P catalyst doped with Mo was studied by Qin et al. [8], concluding that

addition of Mo increased the conversion of NB achieving 98.8 % of selectivity to ANL.

Chapter 8 – Industrial Perspective of Nitrobenzene Catalytic

Hydrogenation in a Tubular Reactor

183

Au-based catalysts were also tested in NB hydrogenation, being concluded that all the

catalysts were active in the hydrogenation reaction although Au deposited on TiO2

performed better than SiO2 due to stronger metal/support interactions [9].

In the last years, the most studied catalysts for liquid-phase hydrogenation have

been Pd-based materials on different supports though their use in large-scale production

has not been explored due to their high costs [10]. The advantage of Pd catalysts is related

with their high activity and also to the fact that they do not attack, or even disrupt, the

aromatic ring [11]. For instance, Turáková et al. [12] analyzed the influence of the

preparation method and palladium content on Pd/C concluding that for Pd contents higher

than 4 wt.% a significant drop in catalytic activity is verified. Pd supported on activated

carbon and on a mixture of activated carbon and multi-walled carbon nanotubes were also

tested [13]. Pd supported on C seems to dominate the scientific research, however, upon

consulting the list of commercial catalysts available in the market, it is possible to verify

that the most common is Pd supported on Al2O3, with the variations restricted to their

shape and metal content.

Besides the catalyst, the solvent to be used is another parameter of great

importance, since it avoids the formation of two phases in the reaction mixture. The

hydrogenation of NB is highly exothermic, so in the presence of a solvent, the reaction

can take place under milder conditions than in the gas phase, and also in more stable

conditions [14]. Different solvents can be used, each with its own advantages and

disadvantages, being the selection traditionally based on solvent performance / effect in

the reaction. However, solvent characteristics like chemical, physical and biological

properties for further treatment may affect both operational conditions and dictate product

separation and catalyst recovery procedures [15]. Suitable solvents for this reaction are

methanol [12, 13], ethanol [9, 16, 17], hexane [18], toluene [5, 19, 20], aniline [7, 21] and

supercritical CO2 [22 - 24], among others.

The knowledge of the influence of reaction products in the NB hydrogenation is

another important issue. Mohan et al. [25] analyzed the effect of co-feeding H2O along

with NB and found out a drastic decrease in the conversion of NB due to the poisoning

effect of H2O over Ni/MgO while over Ni/SBA-15 (santa barbara amorphous silica) the

NB conversion is high and steady [25]. Figueras and Coq [26] refer that Pd/C catalysts

are preferred due to their hydrophobicity since water is produced during the reaction.


184

Influence of cyclohexylamine (CHA) is also relevant, because according to Narayanan et

al. [27], CHA can couple with ANL to form N-phenylcyclohexylamine (NPCHA) or can

undergo dimerization to form dicyclohexylamine (DICHA). At high temperatures it also

favors deammoniation of CHA to give cyclohexane (CH). The authors concluded that in

the absence of CHA, secondary products formation via NPCHA and DICHA are not

possible.

Industrial H2 streams resulting from purges, solubilisation processes or other

applications are usually impure and are available at low pressures. Typically, these type

of streams are energetically valorized by burning or are simply released into the

atmosphere, and so, their use to produce chemicals might be an alternative with added-

value. The knowledge of their main contaminants and the understanding of how they

affect the hydrogenation reaction or the catalyst performance are key aspects to be

evaluated towards the viability and valorization of those streams.

Concluding, in this chapter a commercial Pd-based catalyst was tested in the NB

hydrogenation into ANL, in a tubular fixed bed reactor. In this way, several points will be

analyzed from an industrial point of view, namely the influence of solvent and presence

of reaction products: in the NB conversion and selectivity to both ANL and secondary

products; it will be also assessed the performance of the catalyst when using mild

conditions. The impact of H2 stream contaminants in the catalyst performance will be

evaluated as well. The main novelty of this work is to prove the possibility of successfully

using an industrial H2 stream, resulting from other processes and that is available at low

pressures, in the catalytic hydrogenation of NB into ANL.


Hydrogenation of NB, in liquid phase, was carried out in a tubular reactor with an

internal diameter of 15 mm and length of 400 mm, with a catalytic bed of 120 mm long,

in continuous downflow mode (under H2 pressure). The catalyst used was a commercial

0.3 wt.% Pd/Al2O3 material in extrudate form (catalyst I.2 selected in chapter 5). A

known amount of Pd/Al2O3 catalyst was loaded into the reactor, Figure 8.1, and the

material pre-treated, in situ. Pre-treatment of the catalyst was performed at 150 ºC and

under hydrogen pressure (20 barg), with a gas flow of 2 g/h, for 2 hours. As detailed



185

before, in Chapter 4, temperature-programmed reduction (TPR) experiments have shown

that the catalyst used was fully reduced under the pre-treatment conditions employed.

Chapter 6 contains more details about the experimental set-up.

Figure 8.1 – Scheme of the set-up and tubular reactor used for the catalytic tests.

All the tests were performed with the same sample of catalyst (used in chapter 7),

weighting about 10 g. The catalytic bed, at the center of the reactor, was positioned

between two layers of SiC (21 g each, granulometry of 1.68 mm) and glass spheres. The

upper layer of spheres and SiC served both as a mixer for the reactants and as a pre-

heater. Moreover, SiC layer had the objective of equalizing temperature along the reactor

and to help avoiding hot-spots.

Temperature regulation is made by a control system coupled with an oven from

Termolab equipped with one thermocouple that regulates the reactor heating. Inside the

reactor there is a cane where three thermocouples where positioned, one carrying the

control of temperature in the reaction zone (cf. more details in section 6.2.2.3).

Aromatic

compound

Feed

Sample

collectionProduct

Hydrogenated aromatic

compound

ReactorCatalytic

Bed

SiC

SiC

Glass spheres

Glass spheres

H2

Industrial H2


186

H2 is supplied from a gas cylinder. The liquid and gas feeds are mixed before

entering the reactor, being fed by the top. Products leave the reactor at the bottom, so it

has a down-flow regime. The liquid reactant is pumped to the reactor with a HPLC pump

at a required flow rate. The total pressure inside the reactor is kept constant along each

run using a back pressure regulator.

When the industrial H2 stream was used, each test was initiated with H2 from the

gas cylinder, in order to have always the same starting point and to evaluate catalyst

performance under reference conditions. In the meantime, industrial H2 line is purged to

clean and also with the objective of having the smallest possible amount of condensates.

In order to guarantee that the quantity of contaminants that goes into the unit is the

smallest one, a cylinder was placed in the H2 line, Figure 8.1. As referred tests were

initiated with pure H2, then pressure was decreased until it equalizes industrial H2

pressure. Afterwards, H2 feed stream was changed and the catalytic tests performed at the

pretended conditions.





detector was 250 ºC, the pressure in the column was 14 bar and helium was used as

carrier gas. The column oven was temperature-programmed with a 1 min initial hold at

120 ºC followed by an increase until 230 ºC (15 ºC min-1 rate) and then kept at 230 ºC for

9 min.


Calibration curves were plotted for all the analyzed compounds which were easily



The reference values for temperature, pressure and nitrobenzene concentration

during the parametric study are: T = 120 ºC, P = 14 barg and Cref = 1.2 wt.% NB. The

experiments performed and the conditions used are given in Table C.1 and Table C.2

(Appendix C).

The NB conversion was calculated based on the data obtained from GC analysis:



187

𝑋𝑁𝐵 = 𝑁𝐵0 (𝑝𝑝𝑚) − 𝑁𝐵𝑜𝑢𝑡,𝑡 (𝑝𝑝𝑚)

𝑁𝐵0 (𝑝𝑝𝑚) (8.1)

where 𝑁𝐵0 (𝑝𝑝𝑚) is the reactor feed NB concentration (ppm) and 𝑁𝐵𝑜𝑢𝑡,𝑡 (𝑝𝑝𝑚) is the NB

concentration at the reactor outlet at any time instant t.

The liquid phase analysis confirmed the presence of the following compounds: NB,

ANL, as well as the by-products cyclohexylamine (CHA), cyclohexanol (CHOL),

cyclohexanone (CHONA), N-cyclohexylaniline (CHANIL), dicyclohexylamine

(DICHA), cyclohexyldeneaniline (CHENO) and benzene (Bz) – cf. Nomenclature

section. Nevertheless, the secondary products will not be presented individually but in

groups: Light products – Bz, CHA, CHOL and CHONA; and Heavy products: DICHA,

CHENO and CHANIL. It is important to refer that ANL, used as solvent, is of industrial

grade, thus containing some by-products.

The values used, for result discussion, correspond to what is formed in the reaction,

therefore, to each point the calculation is as follows:

𝐹𝑖 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) = 𝐹𝑖,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) − 𝐹𝑖,0(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (8.2)

where Fi represents the molar flow rate of species i (it can be ANL, secondary products or

its division: light and heavy products), being the amount “formed” obtained by the

difference between the one that “exits” the reactor and what is fed (subscript “ ”).

Selectivity towards ANL, light products, heavy products and secondary products is based

on the amount of each product/group of products produced as compared to all products

formed during the reaction:

𝑆𝑖 =𝐹𝑖 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡 (𝑚𝑜𝑙 𝑚𝑖𝑛⁄ )

∑𝐹𝑗 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡 (𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (8.3)

∑𝐹𝑗 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡 (𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) = 𝐹𝑠𝑒𝑐𝑜𝑛𝑑𝑎𝑟𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) + 𝐹𝐴𝑁𝐿 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (8.4)

𝐹𝑠𝑒𝑐𝑜𝑛𝑑𝑎𝑟𝑦 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑠 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) = 𝐹𝑙𝑖𝑔ℎ𝑡 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) + 𝐹ℎ𝑒𝑎𝑣𝑦 𝑓𝑜𝑟𝑚𝑒𝑑,𝑡(𝑚𝑜𝑙 𝑚𝑖𝑛⁄ ) (8.5)


188


In Chapter 7 a parametric study was carried out and the influence of several

parameters was analyzed. Once more, temperature and pressure have been found to be

parameters with a great influence in the catalyst performance and selectivity towards

ANL/secondary products. Since it was concluded that NB conversion remains stable

along the extended series of reaction tests performed, it is possible to perform another

type of catalytic tests, such as the analysis of the influence of the solvent or of the

presence of reaction products, employing the same catalyst sample.

Industrially, NB hydrogenation is performed in continuous mode and in some

configurations it may exist and ANL recirculation stream, which is composed by ANL

and secondary products formed during the reaction. Among all the secondary products,

CHA and H2O are those that are formed in higher quantities.

Moreover, the main goal of this thesis is to prove the feasibility of using an

industrial stream of H2 that is available at low pressures at CUF-QI facilities and has

some contaminants like NH3, CO2, H2O, Bz, ANL and CHA. The effect of H2O and CHA

will be studied by adding these compounds into the feed mixture. Those compouds were

chosen because they are present in the industrial H2 and also because they might be

present in the ANL recirculation stream in high quantities. In the case of gas

contaminants, it will be studied the effect of the presence of NH3 since this compound is

present in the industrial H2 in high quantities, by using a mixture of H2 with different

loadings of this compound. All these tests will be done at reference conditions: 120 ºC, 14

barg, liquid flow rate of 5 ml/min and gas flow of 2 g/h.

As it was said, industrial H2 is available at low pressures, so it was decided to test

the catalyst response at mild conditions, i.e. 75 ºC and pressures of 14 barg and 9 barg, to

verify if there is any NB conversion and the influence in the selectivity to ANL and

secondary products. The industrial stream of H2 was tested in the tubular reactor,

although it was observed that the line pressure is not constant, with a variation between

2.5 to 5 barg.

The data that will be presented in the sections below was determined when the

reactor reached steady-state, Appendix C – Table C.2 show the conditions employed (it



189

should be noted that after changing operating conditions are necessary ca. 2 h to reach

steady-state).

8.3.1 Influence of the solvent

Since NB hydrogenation is an extremely exothermic reaction, the right choice of the

solvent is very important, since it helps to maintain the reaction temperature at the desired

value. Another function of the solvent is to avoid the formation of two phases, which may

lead to the stop of the reaction. On the other hand, having as solvent a non-existent

compound in the reaction (“strange” substance) can cause another problem such as the

formation of new and unknown compounds or difficulties in the analysis due to phase

separations. Most used solvents are alcohols (methanol, ethanol) or other organic

compounds like hexane or toluene.

In this case, it was chosen cyclohexane (CH) as solvent since it is a saturated

organic compound, it is not detected as a reaction product and NB has a good solubility

on it.

Figures 8.2 and 8.3 presents the results of the catalytic tests TR14b) and TR15a),

with a feed mixture of 1.8 wt.% NB using CH as a solvent and a feed mixture of 1.7 wt.%

NB using ANL as a solvent, respectively, which was employed as reference due to its

commom use in industrial practice [20]. Tests were done at reference conditions: 120 ºC,

14 barg and 5 ml/min.

a)

b)

Figure 8.2 – Evolution of a) NB conversion, b) selectivity to ANL at 120ºC and 14 barg.

ANL CH70

80

90

100

NB

Co

nv

ersi

on

(%

)

ANL CH80

85

90

95

100

AN

L S

elec

tiv

ity

(%

)


190

a)

b)

Figure 8.3 – Evolution of a) selectivity to secondary products and b) secondary products

selectivity distribution for different solvents (ANL and CH) at 120ºC and 14 barg.

The first conclusion is that using CH as solvent, NB conversion is higher than using

ANL. Selectivity to ANL is also higher with CH. So far, CH seems to be a better solvent

to use in this reaction than ANL. In the NB hydrogenation into ANL, for each molecule

of ANL formed two molecules of H2O are also produced (cf. Eq. (8.6)); this fact might be

a problem, because it can lead to the formation of two different phases, an organic and an

aqueous one, and consequently the reaction stops. None of the solvents led the reaction to

stop, but in the case of CH, it seemed that at certain point it started to have two different

phases.

Selectivity to secondary products is clearly favored by using ANL as solvent,

mainly forming light products. Analyzing Figure 8.3b), it is seen that CHONA is the

predominant compound with ANL solvent and CHA, CHONA, DICHA and CHANIL

with CH solvent. It can be observed that in the presence of ANL as solvents, NB

hydrogenation into CHONA is favoured (see proposed reaction mechanism in Figure

8.4), but on the other hand with CH, although there is formation of CHONA, ANL

hydrogenation into CHA, DICHA and CHANIL (Figure 8.4) is quite marked as well.

ANL CH0

1

2

3

Sel

ecti

vit

y (

%)

S light products

S heavy products


ANL CH0.0

0.5

1.0

1.5

2.0

2.5

3.0

Sel

ecti

vit

y (

%)

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL



191

Figure 8.4 – Reaction network proposed for ANL and secondary products formation including Bz


DICHA formation is a huge disadvantage since it is the most difficult product to

separate from ANL. Another drawback of the use of CH is related with the fact that

although selectivity to ANL is quite high, it also promotes ANL hydrogenation. In the

case of using ANL as solvent, secondary products formation is, in the conditions

employed, mostly due to the reaction NB → HONA and not because ANL is being

hydrogenated.

8.3.2 Influence of H2O

Water is one of the compounds that results from the hydrogenation of NB either

into ANL or in CHONA (Figure 8.3):

𝐶6𝐻5𝑁𝑂2 + 3𝐻2 → 𝐶6𝐻5𝑁𝐻2 + 2𝐻2𝑂 (8.6)

𝐶6𝐻5𝑁𝑂2 + 5𝐻2 → 𝐶6𝐻10𝑂 + 𝑁𝐻3 +𝐻2𝑂 (8.7)

Heavy products

Light products

Light products

NO2

Cat

+H2

NH2

+ NH3

H2

Cat

H2

+ANL

-NH3

Cat

H2

NH2

NHNH2

+H2 -NH3

NH

+H2

NH

O

+H2

-NH3

+ANL -H2O

N

+H2

+H2OH CHONA

CHENO

CHOL

DICHA

CHANIL

CHA

Bz

ANLNB

Amine


-H2O

-H2O

*-NH3

+H2O


192

H2O solubility in ANL at 25 ºC is around 5 wt.%, cf. Table 8.1. Consequently, with

the objective of analyzing the effect of this compound in the NB hydrogenation and also

in the secondary products formation, it was added to the feed mixture 4 wt.% of H2O.

Table 8.1 – ANL/H2O system solubility [28]

Temperature (ºC) %(wt. / wt.) of ANL in H2O %(wt. / wt.) of H2O in ANL

25 3.5 5.0

90 6.4 9.9

Catalytic tests were performed at reference operating conditions, 120 ºC, 14 barg, 5

ml/min, solvent ANL and the runs are TR11b) with 1.2 wt.% of NB + 4 wt.% of H2O in

ANL and TR11c) with 1.2 wt. % of NB in ANL (no water). Results are shown in Figure

8.5.

a)

b)

c)

d)


products and d) secondary products selectivity distribution for different H2O concentrations at

120ºC and 14 barg.

The presence of water does not seem to influence NB conversion, Figure 8.5a), as it

was also observed by Gelder [29]. Selectivity to ANL decreases however when H2O is

NB NB + 4wt.% H2O90

92

94

96

98

100

NB

Co

nv

ersi

on (

%)

NB NB + 4wt.% H2O90

92

94

96

98

100

AN

L S

elec

tivit

y (

%)

NB NB + 4wt.% H2O0

1

2

3

4

5

Sel

ecti

vit

y (

%)

S light products

S heavy products


NB NB + 4wt.% H2O0

1

2

3

4

5

Sel

ecti

vit

y (

%)

S Bz

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL



193

present in the feed mixture. Likewise, selectivity to secondary products increases when

H2O is present in the reactor feed; without H2O the two main compounds formed are

CHONA and CHANIL. CHENO is also detected as well as CHA. Nevertheless, CHENO

selectivity decreases in the presence of H2O, while CHA formations remains nearly

stable. When H2O is added, CHANIL formation decreases and a huge increase in

CHONA formation is verified.

CHONA selectivity increases by a factor of 3 and the result is in agreement with

what was expected (cf. reaction mechanism in Figure 8.4): in the presence of H2O the NB

hydrogenation reaction is shifted into the formation of CHONA and not towards ANL

(because for each molecule of ANL formed, there is the formation of two molecules of

H2O, whereas for each CHONA molecule there is only the formation of one molecule of

H2O – cf. Eqs 8.6 and 8.7). Sousa [30] using a Ni based catalyst also tested the influence

of 4 wt.% H2O in the secondary products formation and verified that CHENO is formed

and consumed rapidly to CHONA. CHONA and CHOL concentrations increased

considerably while the concentration of other products was reduced. In the case of this

study, it is also observed a considerable augmentation of CHONA and a slight decrease in

CHENO concentration.

8.3.3 Influence of CHA

CHA is one of the main compounds in the ANL hydrogenation and it is believed to

react with ANL being a precursor of CHANIL (CHA + ANL), Eq 8.8 and DICHA (CHA

+ CHA), Eq 8.9:

𝐶6𝐻7𝑁 + 𝐶6𝐻13𝑁 → 𝐶12𝐻17𝑁 +𝑁𝐻3 (8.8)

𝐶6𝐻13𝑁 + 𝐶6𝐻13𝑁 → 𝐶12𝐻23𝑁 + 𝑁𝐻3 (8.9)

To verify if effectively CHA is a precursor of CHANIL and DICHA, and its

influence in the hydrogenation reaction, 1 wt. % of the compound was added to the

reference liquid feed mixture, 1.2 wt.% NB in ANL. Other operating conditions were 120

ºC, 14 barg and 5 ml/min. These tests correspond to runs TR16a) and TR16b), and the

results are shown in Figures 8.6.


194

a)

b)

c)

d)

Figure 8.6 - Evolution of a) NB conversion, b) selectivity to ANL in the presence of CHA c)

selectivity to secondary products and d) secondary products selectivity distribution in the presence

of CHA at 120ºC and 14 barg.

NB conversion slightly increases in the presence of CHA and selectivity to ANL

decreases from 97 % to 92 %, Figure 8.6. Presence of CHA in the reactor feed seems to

slightly improve hydrogenation of NB but not necessarily into ANL. Observing

secondary products distribution, one can conclude that the only products that seem to

benefit with the existence of CHA in the feed mixture are CHONA and particularly CHA.

CHONA is comprehensible; if NB conversion increases and ANL formation decreases, it

should mean that NB is being converted in CHONA (cf. Figure 8.4). CHA selectivity

increasing in the presence of CHA was not expected and it is not well understood.

Formation of DICHA and CHANIL through CHA was not proved and actually,

DICHA was not even detected, while the concentration of CHENO and CHANIL

decreased in the presence of CHA, Figure 8.6. In this case, CHA seems to have a slight

influence in the NB hydrogenation reaction, improving to some extent the reaction rate,

but selectivity to ANL is smaller. Gelder [29] using Pd supported on carbon catalysts,

verified that CHA is adsorbed on the acidic surface sites and its adsorption does not

interfere with NB hydrogenation, as the two molecules do not compete directly for the

NB NB + 1 wt.% CHA90

92

94

96

98

100

NB

Co

nv

ersi

on

(%

)

NB NB + 1 wt.% CHA90

92

94

96

98

100

AN

L S

elec

tiv

ity

(%

)

NB NB + 1 wt.% CHA0

2

4

6

8

10

Sel

ecti

vit

y (

%)

S light products

S heavy products


NB NB + 1 wt.% CHA0

2

4

6

8

10

Sel

ecti

vit

y (

%)

S Bz

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL



195

catalytically active sites. Therefore, according to that author, CHA does not act as an

inhibitor and appears to have no effect on the hydrogenation of NB. In our case, some

effect was observed through a slight increase in NB conversion, but particularly in

selectivity.

8.3.4 Reaction at mild conditions (T and P)

The catalyst tested, 0.3 wt.% Pd/Al2O3, was chosen for being the most active in the

NB hydrogenation with low secondary products formation, in the CSTR, and in a wide

range of temperatures (150 – 240 ºC) and pressures (6 – 30 barg) – cf. Chapter 4. This

catalyst proved to be the most active at 6 barg and reaction temperature of 150 ºC.

In the parametric study of Chapter 7, it showed also to be active at low pressures (4

barg, 120 ºC) and at low temperatures (75 ºC and 14 barg), with NB conversions higher

than 85 % and selectivity to secondary products of 15 % (4 barg) and 5% (75 ºC).

Therefore, it was decided to prove that the catalyst is really active and rather selective to

ANL by testing at mild industrial conditions, i.e., 9 barg and 75 ºC.

In Figures 8.7 and 8.8 are exhibited results obtained in the operating conditions

referred, corresponding to the runs TR7b) and TR7c).

a)

b)

Figure 8.7 - Evolution of a) NB conversion and b) selectivity to ANL for different pressures at

low temperature (75 ºC).

9 barg 14 barg50

60

70

80

90

100

NB

Conver

sion (

%)

9 barg 14 barg50

60

70

80

90

100

AN

L S

elec

tivit

y (

%)


196

a)

b)

Figure 8.8 - Evolution of a) selectivity to secondary products and b) secondary products

selectivity distribution for different pressures at low temperature (75 ºC).

As it was expected, NB conversion is lower at 9 barg than at 14 barg (because

reaction kinetics and hydrogen solubilisation are detrimentally affected), however it is

around 65 %, Figure 8.7a). On the other hand, selectivity to ANL slightly decreases, with

pressure increase, from 98.4 % to 97.2 %, Figure 8.7b). Surprisingly, selectivity to heavy

products remains stable and it is observed higher selectivity to DICHA at 9 barg.

Selectivity to light products has an increase with pressure augmentation, mainly due to

the formation of CHONA, Figure 8.8b). Such results are in agreement with the reaction

mechanism in Figure 8.3 (NB conversion into CHONA is favoured at the expense of

ANL).

Through the analysis of this results it can be concluded that the catalyst is active at

this operating conditions and it can be used at mild conditions of pressure and

temperature with high selectivity to ANL.

8.3.5 Influence of impure H2

As stated, the main aim of this work is to valorise an industrial stream of H2, which

results from industrial processes and is available at low pressures, at CUF-QI.

Nevertheless, the composition of this stream was unknown. The knowledge of impurities

present in the H2 is very important, in order to determine the strategy that will define the

methodology to be used. That methodology resulted in the introduction of the main

impurities of the industrial gas, through the use of H2 gas cylinder with those

contaminants.

9 barg 14 barg0.0

0.5

1.0

1.5

2.0

2.5

3.0

Sel

ecti

vit

y (

%)

S light products

S heavy products


9 barg 14 barg0.0

0.5

1.0

1.5

2.0

2.5

3.0

Sel

ecti

vit

y (

%)

S Bz S CHA

S CHOL S CHONA

S DICHA S CHENO

S CHANIL



197

Thus, analyses to the industrial H2 stream were made and results are presented in

Table 8.2.

Table 8.2 – Composition of the industrial H2 stream.

Gas Concentration

NH3 > 1373 ppm

N2 2.3 %

CO2 < 0.1 %

He 1.24 %

O2 0.2 %

H2 73.8 %

CO < 5 ppm

ANL > 26 ppm (GC-FID)

12 ppm (GC-MS)

Bz 212 ppm

CHA 0.5 ppm

CHOL 4.7 ppm

CHONA 0.9 ppm

Contaminants present in higher quantities (excluding inert gases) are NH3, CO2 and

O2. Some organic compounds were also detected, being Bz the one with higher

percentage.

Martins [31] studied the NB hydrogenation, on a Ni based catalyst, using H2 with

CO and CO2 as contaminants, at CUF-QI. It was concluded that the effect of those

compounds does not appear to be significant either on ANL formation as on secondary

products formation.

Therefore, according to those results, it was decided to perform catalytic tests only

with NH3.

8.3.5.1 Influence of NH3

NH3 effect on NB hydrogenation was evaluated using H2 gas cylinder with this

contaminant. In fact, two different concentrations of NH3 were tested, 0.05 wt.% (500

ppm) and 1 wt.% (10 000 ppm). Each catalytic test was initiated with pure H2, and after

stabilization (i.e., reaching steady-state), the gas feed was changed to the one with the

contaminant. In the end, to clean the unit and the catalyst surface, the gas feed was

changed again to pure H2.


198

These tests were those of TR12 and TR13 series, performed at 120 ºC, 14 barg, 1.2

wt.% NB in ANL, liquid feed flow rate of 5 ml/min and gas flow rate of 2 g/h. Results

obtained are presented in Figure 8.9.

a)

b)

c)

d)


products and d) secondary products selectivity distribution for different NH3 concentrations at

120ºC and 14 barg.

NH3 concentration in industrial H2 is around 1400 ppm (0.14 wt.%), which is within

the range of concentrations experienced. Observing NB conversion and selectivity to

ANL data, it can be concluded that almost no effect is detected due to the augmentation

of NH3 concentration (only NB conversion slightly increased in the presence of NH3),

Figure 8.9.

In the case of secondary products formation, a positive effect, from an industrial

point of view, is verified: it decreases with the increase of NH3 concentration in the gas

feed. The effect is slightly more noticed in heavy products than in light ones, which

means that NH3 somehow inhibits ANL hydrogenation. DICHA formation is not detected

in any of the tests. CHENO and CHANIL concentrations decrease in the presence of the

contaminant. Light products detected are CHA, CHOL and CHONA. CHOL formation

H2 + 0 wt.% NH3H2 + 0.05 wt.% NH3H2 + 1 wt.% NH390

92

94

96

98

100

NB

con

ver

sion

(%

)

H2 + 0 wt.% NH3H2 + 0.05 wt.% NH3H2 + 1 wt.% NH30

20

40

60

80

100

AN

L S

elec

tiv

ity

(%

)

H2 + 0 wt.% NH3 H2 + 0.05 wt.% NH3 H2 + 1 wt.% NH30

1

2

3

4

5

Sel

ecti

vit

y (

%)

S light products

S heavy products


H2 + 0 wt.% NH3 H2 + 0.05 wt.% NH3 H2 + 1 wt.% NH30

1

2

3

4

5

Sel

ecti

vit

y (

%)

S BZ

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL



199

disappears with the augmentation of NH3 content and CHA concentration decreases.

Interestingly, CHONA formation is not influenced by the presence of NH3, remaining its

concentration the same in the three tests.

Since it is intended to have the highest possible selectivity towards ANL with

minimum formation of secondary products, while not affecting NB conversion, these

results are very interesting and positive. It can be concluded that NH3 concentrations up

to 1 wt.% do not have influence in NB conversion, but helps decreasing the formation of

secondary products, mainly heavy.

8.3.5.2 Industrial H2

In this section, it will be shown catalytic results of the tests performed using the

industrial impure H2 stream. Up to the author knowledge, there is no available

information in the literature about this type of tests or about the use of industrial H2

streams in the NB hydrogenation at low pressures.

To guarantee that catalyst surface and the unit are as clean as possible, tests were

initiated at reference conditions and with pure H2. Two tests with industrial H2 were

carried out, one at 120 ºC and the other one at 150 ºC. The objective of increasing the

temperature was to try to obtain the highest NB conversion since the pressure on the

industrial stream was low and not constant, varying between 2.5 – 5 barg.

Data displayed on Figure 8.10 correspond to runs TR19a) to TR19d) and TR20a to

TR20e). Tests at 2 barg, 4 barg and 14 barg were carried out with pure H2 from the gas

cylinder.


200

a)

b)

c)

Figure 8.10 - Comparison of a) NB conversion, b) selectivity to ANL, c) selectivity to secondary

products, at 120 º C and 150 ºC, as a function of pressure with pure hydrogen and impure

industrial hydrogen grade.

In terms of NB conversion, the use of industrial H2 does not seem to have influence

at any of the temperatures used; values obtained are within the range of conversions

reached with pure hydrogen at the pressures studied (one should recall that the industrial

hydrogen stream had, along the tests, fluctuating pressure in the range of 2.5 – 5 barg).

Relatively to ANL selectivity, at 150 ºC, using industrial or pure H2 no major differences

are noticed, Figure 8.10b). In the case of the test carried out at 120 ºC, selectivity to ANL,

with industrial H2, is much lower to what was expected, when compared with pure one. It

might be thought this is due to operating conditions used, mainly the stream pressure that

is variable and low. Still, it is important to analyze selectivity to secondary products.

Light products formation is superior at 120 ºC than at 150 ºC, with pure H2, but with

industrial H2 the difference between them is huge, Figure 8.10c).

The industrial stream of H2 has a high quantity of organic compounds, since it

results from a process in the ANL plant. Therefore, since the line of industrial H2 is at

room temperature, some of those organic compounds will condensate, mainly the heavy

ones, and another part will continue in the gas-phase, entering in the unit.

2.5 - 5 barg 2 barg 4 barg 14 barg0

20

40

60

80

100

NB

Co

nv

ersi

on (

%)

120 ºC 150 ºC

Industrial H2


85

90

95

100

AN

L S

elec

tivit

y (

%)

120 ºC 150 ºC

Industrial H2


2

4

6

8

10

12

14

Sel

ecti

vit

y t

o s

econ

dar

y

pro

du

cts

(%)

120 ºC 150 ºC

Industria H2



201

As it was referred, in the laboratorial unit, there is a cylinder where industrial H2

goes through to minimize the quantity of condensates present in the tubular reactor. At

the end of each test (TR19 and TR20) the condensates were collected and analyzed by

GC. Main compounds of both samples are Bz, CHA, CHOL and ANL. CHONA, NB,

DICHA, CHENO and CHANIL were also detected but in smaller amounts.

Observing Table 8.2, one concludes that Bz is the organic in higher quantities in the

industrial grade hydrogen; in the GC analysis made to the condensate phase Bz is also a

main compound. Thus, this compound should be one of the light products in greater

quantity in the industrial H2 and that will have influence in the results obtained. In Figure

8.11 are presented selectivity for light and heavy products as well as their distribution in

the several compounds.

a)

b)

c)

d)

Figure 8.11 - Comparison of a) selectivity to light products, b) selectivity to heavy products at

120 and 150 ºC and c) secondary products selectivity distribution at 120 ºC and d) Secondary

products selectivity distribution at 150 ºC as a function of pressure with pure hydrogen and

impure industrial hydrogen grade.

Selectivity to light products at 120 ºC is really superior with the industrial H2, while

selectivity to heavy products is within the range of the results obtained with pure H2,


2

4

6

8

10

12

14

Sel

ecti

vit

y t

o l

igh

t p

rod

uct

s (%

) 120 ºC 150 ºC

Industrial H2

2.5 - 5 barg 2 barg 4 barg 14 barg0.00

0.05

0.10

0.15

0.20

0.25

0.30

Sel

ecti

vit

y t

o h

eavy p

rodu

cts

(%)

120 ºC 150 ºC

Industrial H2


2

4

6

8

10

12

14

Sel

ecti

vit

y (

%)

S Bz

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL

Industrial H2 120 ºC

2.5 - 5 barg 2 barg 4 barg 14 barg0.0

0.5

1.0

1.5

2.0

Sel

ecti

vit

y (

%)

S BZ

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL

Industrial H2

150 ºC


202

Figure 8.11a) and 8.11b). At 150 ºC, selectivity for secondary products is within the range

of results reached with pure H2.

The problem related with the high quantity of light products at 120 ºC with

industrial H2 might be explained with Figure 8.11c) and d). At 120 ºC selectivity to light

products is around 13 % and at 150 ºC is 2%. Observing secondary products selectivity

distribution, the main compound is Bz for both temperature, although at 120 ºC its

selectivity is about 12 %. In this case, minor selectivity to ANL might be related with a

higher concentration of Bz in the industrial stream of H2 than with a lower catalyst

activity. When the catalytic test at 120 ºC was carried out, industrial H2 possibly had more

organic compounds than when was performed at 150 ºC. This fact might be the reason

why selectivity to ANL is lower at 120 ºC when compared with results with pure H2.

Nevertheless, it is important to refer that from an industrial point of view the

problem related with this high quantity of light products is not a big question. ANL

produced in those conditions might be directed into the industrial process and so,

secondary products can be separated from ANL. Alternatively, it may be considered

purifying the H2 stream before entering into the fixed-bed reactor.

8.4 Conclusions

Hydrogenation of NB into ANL over a commercial 0.3 wt.% Pd/Al2O3 catalyst was

studied in a fixed-bed reactor and the influence of several compounds and contaminants

that might be present in the feed (liquid or gas streams) was analyzed. In addition, mild

industrial conditions of pressure and temperature were also tested with the purpose of

evaluating catalyst activity and performance.

Some research was made related with the better solvent to be used. Several studies

are published and it is widely accepted that solvent is an important question to be

considered. In this work, CH was chosen as an alternative solvent to ANL. Both NB

conversion and selectivity to ANL are higher using CH than ANL, however it also

conducts to the formation of DICHA. Influence of the presence of some reaction products

in the feed was also evaluated, most precisely of H2O and CHA. H2O does not seem to

have influence on NB conversion but selectivity to ANL decreases, although a huge

increase in CHONA formation is verified. In the case of CHA, it was concluded that NB



203

conversion slightly increases in its presence while selectivity to ANL decreases.

Moreover, formation of DICHA and CHANIL through CHA was not proved and DICHA

was not even detected.

Catalyst activity and performance was also evaluated by using mild conditions, 75

ºC and 9 barg. As expected, NB conversion is lower at 9 barg but, on the other hand,

selectivity to ANL slightly decreases with pressure increase. Unfortunately, DICHA

formation is verified at 9 barg even though selectivity to heavy products is not affected by

pressure variation. In the case of light products, selectivity has an increase with pressure

augmentation, mainly due to CHONA formation.

The ultimate objective is to use an impure stream of industrial H2. Analyses to this

stream showed that contaminants present in higher quantities are NH3, CO2 and O2. Some

organic compounds were also detected, mainly Bz. From previous studies, CO2 had

proved to have no influence in the reaction, and for that reason only the effect of NH3 was

tested in different concentrations. NB conversion slightly increases in the presence of

NH3 while secondary products selectivity decreases. It was possible to conclude that NH3

concentrations up to 1 wt.% do not have a negative influence in NB hydrogenation, by the

contrary. The industrial H2 stream was tested and no effect was detected in NB

conversion at any of the temperatures used, nor in selectivity to ANL at 150 ºC. Heavy

products formation is low. Relatively to light products, it was shown that the composition

of industrial H2 has some influence and no conclusions can be clearly taken.

Finally, with these catalytic tests it was proved that the industrial H2 stream

available at low pressures and with contaminants can be valorized to produce ANL.

Nonetheless, some attention must be taken to the composition of the stream, since it can

have influence in the formation of by-products. Besides, it was also possible to attest that

a tubular reactor can be used to produce ANL, in an active Pd supported catalyst, with

good selectivity and high levels of NB conversion.


204

References

[1] CUF-QI Internal Report not published, May 2012

[2] Hwang, S., Smith, R., Heterogeneous Catalytic reactor design with optimum

temperature profile I: application of catalyst dilution and side-stream distribution,

Chemical Engineering Science 59 (2004) 4229-4243.

[3] Datsevich, L., Muhkortov, D., Multiphase fixed-bed technologies Comparative

analysis of industrial processes (experience of development and industrial

implementation, Applied Catalysis A: General 261 (2004) 143-161.

[4] Lloyd, L., Hydrogenation Catalysts, Chapter 3, Handbook of Industrial Catalysts,

Fundamental and Applied Catalysis, Springer US, 2011.

[5] Hassan, K., Mohammad, N., Alimorad, R., Masoud, S., Application of Response

Surface Methodology for Catalytic Hydrogenation of Nitrobenzene to Aniline Using

Ruthenium Supported Fullerene Nanocatalyst, Iranian Journal of Chemistry and

Chemical Engineering 34 (2015) 21-32.

[6] Bombos, D., Argesanu, C., Doukeh, R., Bombos, M., Vasilievici, G., Nitrobenzene

hydrogenation by catalysts based on ruthenium, Bulletin of Romanian Chemical

Engineering Society 2 (2015) 2-11.

[7] Relvas, J., Andrade, R., Freire, F., Lemos, F., Araújo, P., Pinho, M., Nunes, C.,

Ribeiro, F., Liquid Phase hydrogenation of nitrobenzene over an industrial Ni/SiO2

supported catalyst, Catalysis Today 133–135 (2008) 828–835.

[8] Qin, Z., Liu, Z., Wang, Y., Promotion Effect of Mo in Amorphous Ni-P Catalysts for

the Liquid-Phase Catalytic Hydrogenation of Nitrobenzene to Aniline, Chemical

Engineering Communications 201 (2014) 338-351.

[9] Torres, C., Campos, C., Fierro, J., Oportus, M., Reyes, P., Nitrobenzene

Hydrigenation on Au/TiO2 and Au/SiO2 Catalyst: Synthesis, Characterization and

Catalytic Activity, Catalysis Letter 143 (2013) 763-771.

[10] Downing, R., Kunkeler, P., van Bekkum, H., Catalytic synthesis of armatic amines,


[11] Blaser, H., Indolese, A., Schnyder,A., Steiner, H., Studer, M., Supported palladium

catalysts for fine chemical synthesis, Journal of Molecular Catalysis A: Chemical 173

(2001) 3–18.



activity in the liquid phase hydrogenation on nitrobenzene to aniline, Applied Catalysis

A: General 476 (2014) 103-112.

[13] Pikna, L., Hezelová, M., Demcakova, S., Smrcová, M., Plesingerová, B., Stefanko,

M., Turáková, M., Králik, M., Pulis, P., Lehocký, P., Effect of support on activity of

palladium catalysts in nitrobenzene hydrogenation, Chemical Papers 68 (2014) 591-598.



205

[14] Wegener, G., Brandt, M., Duda, L., Hofmann, J., Klesczewski, B., Koch, D., Kumpf,


catalysis in the polyurethane industry, Applied Catalysis A: General 221 (2001) 303-335.

[15] Wolfson, A., Grinberg, I., Tavor, D., Solvent Selection in Transfer Hydrogenation

and Suzuki Cross-coupling, Energy and Environmental Engineering 1 (2013) 17-23.

[16] Torres, G., Jablonski, E., Baronetti, G., Castro, A., Miguel, S., Scelza, O., Blanco,

M., Jiménez, P., Fierro, J., Effect of the carbon pre-treatment on the properties and

performance for nitrobenzene hydrogenation on Pt/C catalysts, Applied Catalysis A:

General 161 (1997) 213-226.

[17] Li, J., Li, X., Ding, Y., Wu, P., Pt nanoparticles entrapped in ordered mesoporous

carbons: An efficient catalyst for the liquid-phase hydrogenation of nitrobenzene and its

derivatives, Chinese Journal of Catalysis 36 (2005) 1995-2003.





of Aromatic Nitro compounds on gold catalysts, Angewandte Chemie 119 (2007) 7404-

7407.



Cat Chem 4 (2012) 59-63.




Chemistry 21 (2015) 997-1004.

[22] Zhao, F., Zhang, R., Chatterjee, M., Ikushima, Y., Arai, M., Hydrogenation of






(2014) 1-5.












206







[28] Kirk-Othmer, Amines, Aromatic, Encyclopedia of Chemical Technology, M. Howe-

Grant. New York, John Wiley & Sons 426-482, 1992

[29] Gelder E. The Hydrogenation of Nitrobenzene over Metal, Dissertation presented for

the Doctor of Philosophy degree in Chemical Engineering, University of Glasgow, 2005.



Petrochemical and Chemical Engineering, Instituto Superior Técnico/CUF-QI, S.A.,

2015.

[31] Martins, J., Evaluation of Industrial Catalysts, Dissertation presented for the Master

Degree in Chemical Engineering, Faculdade de Ciências e Tecnologia da Universidade de

Coimbra/ CUF-QI, S.A., 2013.

Part IV

General Conclusions and Future Work

209

Chapter 9 - General Conclusions

This chapter has the objective of giving a general overview of the work done,

highlighting the main results obtained and the most relevant conclusions.

The main objective of this thesis is to prove the technical feasibility of valorising an

industrial H2 stream, which results from processes within the plant and that is available at

low pressure. At CUF-QI one of the relevant reactions that uses H2 as a reactant is the

aniline (ANL) production; so, it was decided to use that stream in the nitrobenzene (NB)

hydrogenation. The great question is related with the H2 stream, which is not pure and is

available at low pressure; this means that to produce ANL, it will be needed a very active

catalyst. The most active catalysts are those in which the active phase is composed by a

noble metal, commonly supported on carbon or on an oxide matrix.

The catalyst choice is fundamental in most chemical reactions. In the case of the

NB hydrogenation, catalysts used are those supported with metals as the active phase,

including Cu, Co, Pd, Pt, and Ni. The catalyst selection is very important for maximizing

ANL production but the knowledge of mass transfer problems is also very critical. By-

products formation is also very important, since their appearance might mean that the

reaction is not occurring as it would be expected and less ANL is being formed or is

being directly consumed. Both NB and ANL might being converted into other products.

The identification of those products, their quantification as well as the way in which they

are formed is a case-study and along the years several papers have been published trying

to identify all the compounds and their formation.

Besides that, efforts have also been devoted to the construction of alternative

reactor configurations, both at laboratorial and at pilot scale. However, more attention

should be directed to this question since this reaction has some challenges, such as the

high exothermicity, the formation of certain compounds that are difficult to separate or

that may lead to the creation of two different phases in the reaction mixture.

The trickle-bed tubular reactor was the configuration chosen to valorize the

industrial H2 stream. Therefore, it was needed to acquire a suitable catalyst for this type


210

of reactor and for the NB hydrogenation. Four commercial catalyst samples were supplied

presenting different compositions, metal loadings and shapes. It was however decided to

start with the catalyst with higher Pd loading and since the tubular reactor was not

constructed yet, it was used a multiphase CSTR operating in batch mode to study the

reaction mechanism and the effect of the main operating conditions. Secondary products

formation proved to be very dependent on the concentration of NB in the reaction

mixture: when NB concentration decreases below a certain threshold (ca. 2 wt.%) their

formation has a huge increase. Consequently, it is possible to infer that the formation of

secondary products is mostly due to the ANL hydrogenation. Thus, it was also possible to

conclude that the Pd-based catalyst used is extremely selective to ANL. This study also

contributed to the comprehension of the ANL and secondary products formation

mechanism in this type of catalyst and it was added an important contribution to the

formulated mechanism: the formation of benzene (Bz). This compound had already been

identified as a secondary product, however it had not been quantified. Finally, and

according to the study performed at Chapter 3, it was possible to conclude that

temperature and pressure are the operating conditions with greater influence in the

hydrogenation reaction. It was also concluded that the solvent used has a direct effect in

the NB conversion and secondary products formation, as it helps to decrease the

formation of secondary products but it also leads to a decrease in NB conversion. The

same conclusion was achieved when studying the presence of reaction products in the

feed mixture.

Then, a catalytic screening of the four commercial catalysts available was carried

out. Those catalysts were Pd and Ni based, all supported (catalyst I.1: 1 wt.% Pd/Al2O3,

catalyst I.2: 0.3 wt.% Pd/Al2O3, catalyst I.3: 0.3 wt.% Pd/Al2O3 and catalyst II.1: 50 wt.%

Ni/(Al2O3+SiO2)); they were subjected to the NB hydrogenation and their morphological

and textural properties were analysed by different techniques (XDR, SEM, HR-TEM,

TEM, H2-TPR, nitrogen adsorption for BET, ICP-MS and CHNS). No major differences

were observed between fresh and used samples, which means that reaction does not have

a great influence in the catalysts’ properties. Once more it was shown that the reaction

(conducted in a batch reactor) can be divided in two parts, at high and low NB

concentrations, and that secondary products formation occurs predominantly in the 2nd

part (low NB concentrations) for all the catalysts. Comparing all the results, catalyst I.2

showed to be the one with higher NB consumption rate at all operating conditions

Chapter 9 – General Conclusions

211

studied, with a low secondary products formation. It was reinforced that a special

attention should be taken when choosing operating conditions, mainly temperature.

A tubular reactor was designed and constructed at CUF-QI, using internal know-

how. Before testing the selected catalyst in real conditions with a real feed, some

preliminary tests were performed, where the catalytic bed pressure drop and temperature

control were evaluated. No significant issues were detected and the unit was considered

ready to work.

In the catalytic screening, catalyst I.2 showed to be the most active with low

secondary products formation and for this reason it was the chosen one to be evaluated in

the tubular reactor. A parametric study was carried out where the influence of some

operating conditions was analyzed. The first conclusion is that catalyst age is an

important parameter that should be taken into account. Along the reaction time, NB

conversion remains stable whereas selectivity to ANL increases and inherently to

secondary products decreases. This conclusion is extremely important from an industrial

point of view, since the objective is to produce ANL with high purity. It was also found

that temperature and pressure have a great influence in the selectivity towards secondary

products, as previously determined in the discontinuous reactor configuration.

Nevertheless, it was demonstrated that ANL can be produced through NB hydrogenation

in a tubular reactor, without major issues related with temperature control, in the range of

conditions studied.

The last part of this work consisted on an industrial perspective of the

hydrogenation reaction. It was decided to analyse the catalyst performance in some

industrial conditions and, for that, some of the reaction products were co-fed (that might

be present e.g. when recirculating ANL) and mild conditions of temperature and pressure

were also used. The effect of the solvent used in the reaction was also analysed. It was

demonstrated that cyclohexane (CH) is a good solvent for this reaction, although it leads

to the formation of secondary products, mainly dicyclohexylamine (DICHA - which is

difficult to separate from ANL); for that reason, ANL was chosen as solvent. Both CHA

and H2O were fed with the reference mixture and no major differences between tests with

and without them were observed. In the case of CHA, it helped to conclude that even if

CHA is present in higher quantities, it will not result on an augmentation of CHANIL and

DICHA concentrations. Catalyst I.2 proved to be active at mild conditions, although


212

formation of DICHA has also been detected. Analysis made to the industrial gas stream

revealed that NH3, CO2 and light organics are the main contaminants. A previous study

was carried out using CO2 as a contaminant and it was concluded that its presence does

not have a significant influence either on ANL formation or on secondary products

formation. Consequently, some catalytic tests were carried out with H2 from a gas

cylinder containing NH3 (important contaminant). Surprisingly, both NB conversion and

selectivity to ANL slightly increased in the presence of this contaminant and inherently

secondary products selectivity decreased. Thus, it was possible to conclude that NH3

concentrations up to 1 wt.% do not seem to have a negative influence in NB conversion

into ANL.

As it was previously referred, the main goal of this PhD thesis was to assess the

technical feasibility of valorising an industrial H2 stream by using an active catalyst to

produce ANL in a tubular reactor. This was experimentally tested and no detrimental

effect was detected in NB conversion at any of the temperatures used, nor in selectivity to

ANL, when such industrial stream was employed, as compared to pure hydrogen.

Consequently, it was shown that the industrial H2 stream available at low pressures and

with contaminants can be valorised to produce ANL and the catalyst used is suitable for

this type of process, with good selectivity and high levels of NB conversion.

213

Chapter 10 - Future Work

Several suggestions might be presented for future work since there are still some

important questions that should be analysed and answered or further explored. Those

suggestions are essentially related with the catalysts, the tubular reactor and with kinetic

studies and modelling.

10.1 Catalysts

It was concluded that catalyst age is an important issue and this issue should be

carefully analysed. Consequently, more catalytic tests should be performed in a

continuous mode in order to evaluate if NB conversion remains constant as well as the

selectivity to ANL for prolonged times-on-stream, coupling with detailed physico-

chemical catalyst characterization. In particular, changes in catalyst textural and

morphological properties, pore dimensions, and reaction selectivity issues should be

related. A new parametric study, with more drastic conditions (higher temperatures and

different NB concentrations) should also be carried out to analyse the possibility of

having, for instance, steam production.

Characterization of the catalyst is an important instrument that might help to

understand both the catalyst behaviour and the lifetime / selectivity changes that could be

expected. In this way more detailed studies focused on characterization should be carried

out together with catalytic tests.

In the trickle-bed tubular reactor, it was only possible to work with one of the

catalysts that were available and that was chosen for being the most active. Nevertheless,

this catalyst is a Pd-based one and costs associated to its acquisition might have an

important impact in the total costs of the unit. Among the catalysts that were tested in

Chapter 5, there is one made of Ni (catalyst II.1), that should be cheaper than those of Pd.

Therefore, at least catalyst II.1 should be tested in the tubular reactor and a detailed

evaluation of its performance and selectivity to ANL and secondary products should be


214

performed, and complemented by a cost analysis. The use of mixed catalyst beds should

also be considered.

10.2 Tubular reactor

The reactor that was constructed has a 30 cm3 catalytic bed, being a laboratorial

unit. The unit was not conceived and is not prepared for working in a continuous mode of

24 over 24 h. This is a challenge that could be considered in the near future since it would

allow to work in industrial conditions and would lead to a better evaluation of the catalyst

performance, namely for prolonged times of operation.

The construction of a pilot unit, working 24 / 24 h, is another strategy to be

considered for implementation. However, some precautions must be taken in the scale-up;

certainly it would need a cooling system in order to have a better control of the

temperature inside the reactor, or alternative strategies should be considered (e.g. catalyst

diluted with inert, multi tubular configuration, etc.). Problems of mass transfer in the

multiphase (G-L-S) reaction system should also be taken into account. In this case, a new

parametric study should be performed where the influence of higher temperatures,

pressures and of both gas and liquid feed flow rates would be required. In this part,

special attention should be paid to the temperature control in order to avoid hot-spots and

run-away situations, guarantying that all the safety issues would be assured and respected.

10.3 Kinetic studies

In this work it was not possible to perform a kinetic study of the NB hydrogenation

into ANL, which should include and account for the formation of secondary products,

using catalysts suitable for fixed-bed reactors. The issues of mass and heat transfer in

such a complex G-L-S system should be considered as well.

Moreover, such studies would also allow to develop robust models of the tubular

reactor both at laboratorial and at pilot scale levels, that should be experimentally

validated. This study would be very interesting and important since it would allow to

more easily optimize operating conditions.

Appendixes

Appendix A – Supporting Information of Chapter 3.

217

Appendix A – Supporting Information of Chapter 3.

Materials and Methods

For the catalytic runs, carried out in the set-up shown in Fig. A.1, pre-treatment of

the catalyst was firstly performed, at 150 ºC and under hydrogen pressure (20 barg) for 2

hours. Several temperature-programmed reduction (TPR) experiments were performed

with the fresh catalyst and the results have shown that the most important peaks appeared

at 80 ºC (PdO species). This means that the catalyst used was fully reduced, under the

pre-treatment conditions employed.

After this activation step, a certain volume of aniline (industrial grade) was loaded

into the reactor, as is commonly done in the industry, with two main goals:

i) to act as a solvent for the water that is produced during the reaction, in order to

avoid the formation of two phases (organic and aqueous) that would lead to the

interruption of the reaction, and also to avoid strong NB adsorption;

ii) to help to dissipate the excess heat generated due to the high exothermicity of the

reaction.


controller (SHIMADEN SD20) and the initial temperature was established. The heat

produced by the nitrobenzene hydrogenation was removed by a cooling water stream

whose flow was controlled with a set of ball valves, as shown in Figure A.1. The reactor

temperature was constant with a maximum Δ of 4 - 5 % and it was continuously

measured throughout the experiments.

The NB was loaded to the reactor as fast as possible, when the desired reaction

temperature was achieved (time=0 min), in order to ensure that the NB hydrogenation

does not start before the beginning of the experiment and also to avoid any strong NB

adsorption on the catalyst, thus blocking the accessibility to the active sites; this is also

safeguarded by using ANL as solvent. All the experiments were done in a batch mode up

to a nearly complete consumption of NB (which was considered to correspond to the

instant at which NB concentration was below 1000 ppm).


218

The reactor operates in a batch mode relative to the liquid phase but in a semi-

continuous mode for the gas phase (hydrogen). The total pressure inside the reactor is

kept constant along each run due to the continuous admission of hydrogen to compensate

what is being consumed.





detector was 250 ºC, the pressure in the column was 14 bar and the carrier gas used was

helium. The column oven was temperature-programmed with a 1 min initial hold at 120

ºC, followed by an increase of temperature until 230 ºC at a rate of 15 ºC min-1 and then

kept at 230 ºC for 9 min.


Calibration curves were plotted for all the compounds to be analyzed, which were easily



Figure A.1 – Scheme of the batch reactor and set-up used in the experiments.

H2

H2V1

V3

V4N2

V6

V5

Vessel

V7

reactor

Samples

V14

V11 V12

V9

V8

H2OPI4

PI2

V2

PI1 PI3

V10

V13

Appendix B – Supporting Information of Chapter 5.

219

Results

a)

b)

Figure A.2 – a) Secondary products formation vs. time for runs with aniline and nitrobenzene

(run B4) or only aniline (run B12) in the reaction mixture and b) NB consumption vs. time for the

run with aniline and nitrobenzene in the reaction mixture (run B4).

a)

b)

c)

d)

Figure A.3 – Aniline and secondary products formation vs. NB dimensionless concentration for

different reaction pressures, runs B2, B3 and B4.

0 50 100 150

0

1x10-2

2x10-2

3x10-2

4x10-2

5x10-2

Cto

tal

seco

nd

ary

pro

du

cts /

Cre

f

time (min)

ANL (B12)

10% NB + ANL (B4)

0 50 100 150

0.0

0.2

0.4

0.6

0.8

1.0

Ci /

Cre

f

time (min)

10% NB + ANL (B4)

NB

0.0 0.2 0.4 0.6 0.8 1.0

0.0

2.0x10-5

4.0x10-5

6.0x10-5

8.0x10-5

1.0x10-4

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

Bz

0.0 0.2 0.4 0.6 0.8 1.0

0.0

2.0x10-3

4.0x10-3

6.0x10-3

8.0x10-3

1.0x10-2

1.2x10-2

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

CHA

0.0 0.2 0.4 0.6 0.8 1.0

0.0

5.0x10-4

1.0x10-3

1.5x10-3

2.0x10-3

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

CHOL

0.0 0.2 0.4 0.6 0.8 1.0

0.0

3.0x10-3

6.0x10-3

9.0x10-3

1.2x10-2

CNB

/ Cref

Ci /

Cre

f

14 barg

20 barg

30 barg

CHONA


220

a)

b)

c)

d)

Figure A.4 – Aniline and secondary products formation vs. NB dimensionless concentration for

different reaction pressures, runs B2, B3 and B4.

a)

b)

Figure A.5 - Aniline and secondary products formation vs. NB dimensionless concentration for

different reaction temperatures, runs B4, B5, B9 and B10.

0.0 0.2 0.4 0.6 0.8 1.08.0

8.4

8.8

9.2

9.6

10.0

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

ANL

0.0 0.2 0.4 0.6 0.8 1.0

0.0

3.0x10-4

6.0x10-4

9.0x10-4

1.2x10-3

1.5x10-3

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

DICHA

0.0 0.2 0.4 0.6 0.8 1.0

0.0

5.0x10-3

1.0x10-2

1.5x10-2

2.0x10-2

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

CHENO

0.0 0.2 0.4 0.6 0.8 1.0

0.0

6.0x10-3

1.2x10-2

1.8x10-2

2.4x10-2

3.0x10-2

Ci /

Cre

f

CNB

/ Cref

14 barg

20 barg

30 barg

CHANIL

0.0 0.2 0.4 0.6 0.8 1.0

0.0

1.0x10-4

2.0x10-4

3.0x10-4

4.0x10-4

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

Bz

0.0 0.2 0.4 0.6 0.8 1.0

0.00

1.50x10-2

3.00x10-2

4.50x10-2

6.00x10-2

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

CHA


221

a)

b)

c)

d)

e)

f)

Figure A.6 - Aniline and secondary products formation vs. NB dimensionless concentration for

different reaction temperatures, runs B4, B5, B9 and B10

0.0 0.2 0.4 0.6 0.8 1.0

0.0

3.0x10-3

6.0x10-3

9.0x10-3

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

CHOL

0.0 0.2 0.4 0.6 0.8 1.0

0.0

1.0x10-2

2.0x10-2

3.0x10-2

4.0x10-2

CNB

/ Cref

Ci /

Cre

f 150 ºC

180 ºC

210 ºC

240 ºC

CHONA

0.0 0.2 0.4 0.6 0.8 1.08.0

8.4

8.8

9.2

9.6

10.0

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

ANL

0.0 0.2 0.4 0.6 0.8 1.0

0.0

1.0x10-3

2.0x10-3

3.0x10-3

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

DICHA

0.0 0.2 0.4 0.6 0.8 1.0

0.00

1.50x10-2

3.00x10-2

4.50x10-2

6.00x10-2

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

CHENO

0.0 0.2 0.4 0.6 0.8 1.0

0.00

2.50x10-2

5.00x10-2

7.50x10-2

1.00x10-1

Ci /

Cre

f

CNB

/ Cref

150 ºC

180 ºC

210 ºC

240 ºC

CHANIL


222


223

Appendix B – Supporting Information of Chapter 5

Material and Methods

Catalytic Reaction

Figure B.1 – Scheme of the batch reactor and set-up used in the experiments.

H2

H2V1

V3

V4N2

V6

V5

Vessel

V7

reactor

Samples

V14

V11 V12

V9

V8

H2OPI4

PI2

V2

PI1 PI3

V10

V13


224

Results and Discussion

Catalysts Characterization

a)

b)

c)

d)

Figure B.2 – X-ray diffraction patterns comparison between fresh and used catalyst: a) catalyst

I.1, b) catalyst I.2, c) catalyst I.3 and d) catalyst II.1.

10 20 30 40 50 60 70 80 90 100

Catalyst I.1 fresh

Catalyst I.1 used

2 Theta (degrees)

10 20 30 40 50 60 70 80 90 100

Catalyst I.2 fresh

Catalyst I.2 used

2 Theta (degrees)

10 20 30 40 50 60 70 80 90 100

Catalyst I.3 fresh

Catalyst I.3 used

2 Theta (degrees)

10 20 30 40 50 60 70 80 90 100

Catalyst II.1 fresh

Catalyst II.1 used

2 Theta (degrees)


225

a)

b)

c)

d)

e)

f)

g)

h)

Figure B.3 – SEM micrographs of the catalysts studied: catalyst I.1 a) fresh and b) used, catalyst

I.2 c) fresh and d) used, catalyst I.3 e) fresh and f) used and catalyst II.1 g) fresh and h) used.


226

Fresh Used

a)

b)

c)

d)

e)

f)

g)

h)

Figure B.4 – TEM micrographs of the catalysts studied: catalyst I.1 a) fresh and b) used, catalyst

I.2 c) fresh and d) used, catalyst I.3 e) fresh and f) used, and catalyst II.1 g) fresh and h) used.


227

Catalysts selectivity

a)

b)

c)

Figure B.5 – Total secondary products concentration at Tref as a function of nitrobenzene

concentration a) P = 6 barg, b) P = 14 barg and c) 30 barg.

0 20000 40000 60000 80000 1000000

2000

4000

6000

8000

CT

ota

l se

con

dar

y p

rod

uct

s (p

pm

)

C NB

(ppm)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20000 40000 60000 80000 1000000

2000

4000

6000

8000

CT

ota

l se

con

dar

y p

rod

uct

s (pp

m)

C NB

(ppm)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20000 40000 60000 80000 1000000

2000

4000

6000

8000

CT

ota

l se

con

dar

y p

rod

uct

s (pp

m)

C NB

(ppm)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1


228

a)

b)

c)

Figure B.6 – Total secondary products concentration at Pref as a function of nitrobenzene

concentration a) 150 ºCTref, b) 180 ºC and c) 240 ºC.

0 20000 40000 60000 80000 1000000

5000

10000

15000

20000

25000

30000

35000

CT

ota

l se

con

dar

y p

rod

uct

s (p

pm

)

C NB

(ppm)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20000 40000 60000 80000 1000000

5000

10000

15000

20000

25000

30000

35000

CT

ota

l se

con

dar

y p

rod

uct

s (p

pm

)

C NB

(ppm)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

0 20000 40000 60000 80000 1000000

5000

10000

15000

20000

25000

30000

35000

CT

ota

l se

con

dar

y p

rod

uct

s (pp

m)

C NB

(ppm)

Catalyst I.1

Catalyst I.2

Catalyst I.3

Catalyst II.1

Appendix C - Resume of the operating conditions used in the catalytic tests

with the tubular reactor (Chapter 7 and 8).

229

Appendix C - Resume of the operating conditions used in the catalytic tests

with the tubular reactor (Chapters 7 and 8).

Table C.1 – Operating conditions used in the catalytic tests performed on the tubular reactor.

Test run Temperature

(ºC)

Pressure

(barg)

Liquid feed

flow rate

(ml/min)

Composition of

liquid feed Gas

TR1

TR1a)

150 14 5

ANL

H2 TR1b) 4.5 wt. % NB in ANL

TR1c) 7.5

TR2 TR2a)

150 14 5 4.5 wt. % NB in ANL


TR3

TR3a) 150

14

5

1.2 wt. % NB in ANL H2 TR3b) 150 12.5

TR3c) 120 5

TR3d) 100

TR4

TR4a) 120

14

5 1.2 wt. % NB in ANL H2

TR4b) 100

TR4c) 75

TR4d)

120 TR4e) 10

TR4f) 7

TR5

TR5a)

120 14

5 1.2 wt. % NB in ANL H2 TR5b)

TR5c)

TR5d) 7

TR6

TR6a)

120

14

5 1.2 wt. % NB in ANL H2 TR6b)

4 TR6c)

TR6d)

TR7

TR7a) 120 14

5 1.2 wt. % NB in ANL H2 TR7b) 75

TR7c) 9

TR8

TR8a)

120 14

2.5

1.2 wt. % NB in ANL H2 TR8b) 5

TR8c) 12.5

TR8d) 20

TR9

TR9a)

150 14

2.5

1.2 wt. % NB in ANL H2 TR9b) 5

TR9c) 12.5

TR9d) 20

TR10

TR10a)

120 14 5

1.2 wt. % NB in ANL


TR10c) 8.5 wt. % NB in ANL

TR11

TR11a)

120 14 5

1.2 wt. % NB in ANL

H2 TR11b) 1.2 wt. % NB + 4 wt.

% H2O in ANL


TR12

TR12a)

120 14 5 1.2 wt. % NB in ANL

H2

TR12b) H2 + 0.05 wt. %

NH3

TR12c) H2

TR13

TR13a)

120 14 5 1.2 wt. % NB in ANL

H2

TR13b) H2 + 1 wt. % NH3

TR13c) H2 *Highlighted blue tests correspond to reproducibility tests mentioned in Chapter 7, section 7.3.1.


230

Table C.2 – Operating conditions used in the catalytic tests performed on the tubular reactor (continuation).

TR14

TR14a)

120 14 5

2.4 wt. % NB in ANL

H2 TR14b) 1.8 wt. % NB in CH


TR15 TR15a) 120 14 5 1.7 wt. % NB in ANL H2

TR16

TR16a)

120 14 5

1.2 wt. % NB in ANL

H2 TR16b)

1.2 wt. % NB + 1.2

wt.% CHA in ANL

TR17 TR17a)

120 14 5 1.2 wt. % NB in ANL


TR18 TR18a)

120 14

5 1.2 wt. % NB in ANL H2 TR18b) 2

TR19

TR19a)

120

14

5 1.2 wt. % NB in ANL

H2 TR19b) 2

TR19c) 2.5-5 H2 Industrial

TR19d) 4 H2

TR20

TR20a) 120

14

5 1.2 wt. % NB in ANL

H2 TR20b) 4

TR20c)

150 TR20d) 2.5-5 H2 Industrial

TR20e) 14 H2 *Highlighted blue tests correspond to reproducibility tests mentioned in Chapter 7, section 7.3.1.

Appendix D - Complementary results of the parametric study in Chapter 7.

231

Appendix D - Complementary results of the parametric study in Chapter 7.

Study of the influence of temperature, for a catalyst with an age of 2.4 day at 14

barg, 1.2 % NB in ANL and feed flow rate of 5 ml/min.

a)

b)

c)

d)

Figure D.1 - Evolution of a) NB conversion, b) selectivity to ANL, c) selectivity to secondary

products and d) secondary products selectivity distribution for different temperatures.

60%

70%

80%

90%

100%

70 90 110 130 150

Convers

ion

temperature (ºC)

NB

60%

70%

80%

90%

100%

70 80 90 100 110 120 130 140 150S

ele

ctivity

temperature (ºC)

ANL

0%

5%

10%

15%

20%

70 80 90 100 110 120 130 140 150

Sele

ctivity

temperature (ºC)

S Light products

S Heavy products


0

5

10

15

20

75 ºC 100 ºC 120 ºC 150 ºC

Sel

ectivity (%

)

S BZ

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL


232

Study of the influence of liquid feed flow, at 120ºC, 14 barg and 1.2 % NB in ANL.

a)

b)

c)

d)

Figure D.2 - Evolution of a) NB conversion, b) selectivity to ANL, c) selectivity to secondary

products and d) secondary products selectivity distribution for different flow rates at 120 ºC and

14 barg.

40%

60%

80%

100%

0 5 10 15 20

Convers

ion

Liquid Feed flow (ml/min)

NB

70%

80%

90%

100%

0 5 10 15 20

Sele

ctivity


ANL

0%

5%

10%

15%

20%

25%

30%

0 5 10 15 20

Sele

ctivity


S Light products

S Heavy products


0

10

20

30

2.5 ml/min 5 ml/min 12.5 ml/min 20 ml/min

Sel

ectivity (%

) S BZ

S CHA

S CHOL

S CHONA

S DICHA

S CHENO

S CHANIL

Documents

Impure Hydrogen Valorization for Chemicals Production in a