Marcel Schlaf Z. Conrad Zhang Editors
Reaction Pathways and Mechanisms in Thermocatalytic Biomass
Conversion II Homogeneously Catalyzed Transformations, Acrylics
from Biomass, Theoretical Aspects, Lignin Valorization and
Pyrolysis Pathways
Green Chemistry and Sustainable Technology
Series editors Prof. Liang-Nian He State Key Laboratory of
Elemento-Organic Chemistry, Nankai University, Tianjin, China
Prof. Robin D. Rogers Department of Chemistry, McGill University,
Montreal, Canada
Prof. Dangsheng Su Shenyang National Laboratory for Materials
Science, Institute of Metal Research, Chinese Academy of Sciences,
Shenyang, China and Department of Inorganic Chemistry, Fritz Haber
Institute of the Max Planck Society, Berlin, Germany
Prof. Pietro Tundo Department of Environmental Sciences,
Informatics and Statistics, Ca’ Foscari University of Venice,
Venice, Italy
Prof. Z. Conrad Zhang Dalian Institute of Chemical Physics, Chinese
Academy of Sciences, Dalian, China
Aims and Scope The series Green Chemistry and Sustainable
Technology aims to present cutting- edge research and important
advances in green chemistry, green chemical engineering and
sustainable industrial technology. The scope of coverage includes
(but is not limited to):
– Environmentally benign chemical synthesis and processes (green
catalysis, green solvents and reagents, atom-economy synthetic
methods etc.)
– Green chemicals and energy produced from renewable resources
(biomass, car- bon dioxide etc.)
– Novel materials and technologies for energy production and
storage (biofuels and bioenergies, hydrogen, fuel cells, solar
cells, lithium-ion batteries etc.)
– Green chemical engineering processes (process integration,
materials diversity, energy saving, waste minimization, effi cient
separation processes etc.)
– Green technologies for environmental sustainability (carbon
dioxide capture, waste and harmful chemicals treatment, pollution
prevention, environmental redemption etc.)
The series Green Chemistry and Sustainable Technology is intended
to provide an accessible reference resource for postgraduate
students, academic researchers and industrial professionals who are
interested in green chemistry and technologies for sustainable
development.
More information about this series at
http://www.springer.com/series/11661
Reaction Pathways and Mechanisms in Thermocatalytic Biomass
Conversion II Homogeneously Catalyzed Transformations, Acrylics
from Biomass, Theoretical Aspects, Lignin Valorization and
Pyrolysis Pathways
ISSN 2196-6982 ISSN 2196-6990 (electronic) Green Chemistry and
Sustainable Technology ISBN 978-981-287-768-0 ISBN
978-981-287-769-7 (eBook) DOI 10.1007/978-981-287-769-7
Library of Congress Control Number: 2015951980
Springer Singapore Heidelberg New York Dordrecht London © Springer
Science+Business Media Singapore 2016 This work is subject to
copyright. All rights are reserved by the Publisher, whether the
whole or part of the material is concerned, specifi cally the
rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfi lms or in any
other physical way, and transmission or information storage and
retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed. The use
of general descriptive names, registered names, trademarks, service
marks, etc. in this publication does not imply, even in the absence
of a specifi c statement, that such names are exempt from the
relevant protective laws and regulations and therefore free for
general use. The publisher, the authors and the editors are safe to
assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the
publisher nor the authors or the editors give a warranty, express
or implied, with respect to the material contained herein or for
any errors or omissions that may have been made.
Printed on acid-free paper
Springer Science+Business Media Singapore Pte Ltd. is part of
Springer Science+Business Media (www.springer.com)
Editors Marcel Schlaf The Guelph-Waterloo-Centre for Graduate Work
in Chemistry (GWC)2
Department of Chemistry University of Guelph Guelph , ON ,
Canada
Z. Conrad Zhang Dalian National Laboratory
for Clean Energy Dalian Institute of Chemical Physics Dalian ,
China
Pref ace
Short carbon chain molecules (C 2 –C 9 ) obtained either directly
from sugars, the hydrolysis of starch, or preferably by the
controlled breakup of lignocellulosic bio- mass into soluble
components are the only conceivable sustainable source of carbon on
the planet that could ultimately replace the fossil hydrocarbons
that currently form the base of the chemical industry and hence our
technological civilization at large. In particular, the production
of polymer components and polymers that are chemically or at least
functionally equivalent to those derived from the refi ning of
crude oil would offer ecologic and environmental as well as
economic advantages.
The use of sugars, starch, and ultimately lignocellulosic biomass,
i.e., forestry (e.g., wood and bark chips, etc.) and agricultural
(e.g., straws, husks, stovers, etc.) residues, as a renewable
carbon resource will, however, require careful life-cycle analyses
of the processes involved. This in turn is critically dependent on
a deep and detailed understanding of the mass and energy fl ows in
these processes and hence their mechanisms at the molecular level.
Almost “by defi nition” these processes will have to be catalytic
in nature to be ecologically sustainable and economically
viable.
The development of new catalysts and catalytic processes that are
specifi cally designed for and adapted to the unique properties of
the biomass-derived carbon substrates poses a unique challenge. Due
to the abundance of oxygen-containing functional groups, the
pentose and hexose sugars and their furanic condensates obtainable
from (hemi-)cellulose as well as the phenol propanoid units of
lignin are characterized by a high polarity and reactivity that is
very different – one could say almost opposite – to that of the
traditionally employed alkane and arene sources available from refi
ned crude oil. The fundamental study of the reaction cascades and
mechanisms involved in the transformation of oxygenated biomass to
value-added chemicals is the fi rst step to meet this
challenge.
Focusing on the use of thermochemical and acid-/base- or
metal-catalyzed pro- cesses only, the two volumes of this book
attempt to give an overview of and insights into the specifi c
aspects of this challenge as perceived and formulated by expert
contributors research-active in this fi eld.
vi
Volume I is comprised of chapters that address the nanoscale
structure of lignocellulose, the application of acid-base reactions
and catalysts to the depoly- merization of cellulose, the use of
heterogeneous hydrogenation catalysts for its direct conversion to
polyols, as well as chapters that explore pathways for the metal-
catalyzed dehydration and oxidation of sugars and sugar alcohols to
furans and carboxylic acids, respectively.
The chapters of Volume II cover the hydrodeoxygenation of
sugar-derived sub- strates by homogenous catalysts systems; the
valorization of carboxylic acids, nota- bly lactic acid and its
derivatives; a theoretical approach to the elucidation of the
conversion pathways of sugars and sugars condensates and their
decomposition to humins; as well as mechanistic and practical
aspects of the conversion and pyrolysis of lignin to functionalized
monocyclic aromatics and the pyrolysis of biomass to synthesis
gas.
We hope that the insights provided by the different and varied
perspectives offered here will convince the readers that a switch
to renewable biomass as a key carbon source for the chemical
industry will be feasible and does indeed offer a way forward to a
more sustainable future.
Guelph , Canada Marcel Schlaf Dalian, China Z. Conrad Zhang
Preface
vii
Contents
1 Deoxydehydration (DODH) of Biomass- Derived Molecules
................. 1 Shuo Liu , Jing Yi , and Mahdi M.
Abu-Omar
2 Homogeneous Catalysts for the Hydrodeoxygenation of Biomass-
Derived Carbohydrate Feedstocks
...................................... 13 Marcel Schlaf
3 Valorization of Lactic Acid and Derivatives to Acrylic Acid
Derivatives: Review of Mechanistic Studies
.................................. 39 Elodie Blanco , Stéphane
Loridant , and Catherine Pinel
4 Computational Chemistry of Catalytic Biomass Conversion
............... 63 Guanna Li , Emiel J. M. Hensen , and Evgeny A.
Pidko
5 Humin Formation Pathways
....................................................................
105 Jacob Heltzel , Sushil K. R. Patil , and Carl R. F. Lund
6 Catalytic Hydrodeoxygenation of Lignin Model Compounds
.............. 119 Basudeb Saha , Ian Klein , Trenton Parsell , and
Mahdi M. Abu-Omar
7 Oxidation of Lignins and Mechanistic Considerations
......................... 131 Adilson R. Gonçalves , Priscila Benar
, and Ulf Schuchardt
8 Pyrolysis Mechanisms of Lignin Model Compounds Using a Heated
Micro-Reactor
................................................................
145 David J. Robichaud , Mark R. Nimlos , and G. Barney
Ellison
9 Catalytic Gasification of Lignocellulosic Biomass
................................. 173 C. V. Pramod and K.
Seshan
ix
Contributors
Mahdi M. Abu-Omar Department of Chemistry and the Center for
Catalytic Conversion of Biomass to Biofuels (C3Bio) , Purdue
University , West Lafayette , IN , USA
Spero Energy, Inc. , West Lafayette , IN , USA
Priscila Benar Instituto Agronômico de Campinas , Campinas , SP ,
Brazil
Elodie Blanco Institut de Recherches sur l’Environnement et la
Catalyse de Lyon (IRCELYON) , UMR 5256, CNRS – Université Lyon 1 ,
Villeurbanne Cedex , France
G. Barney Ellison Department of Chemistry and Biochemistry ,
University of Colorado , Boulder , CO , USA
Adilson R. Gonçalves Pontifícia Universidade Católica de Campinas,
PUCCAMP , Campinas , SP , Brazil
Jacob Heltzel Department of Chemical and Biological Engineering ,
University at Buffalo , Buffalo , NY , USA
Emiel J. M. Hensen Inorganic Materials Chemistry Group, Schuit
Institute of Catalysis, Department of Chemical Engineering and
Chemistry , Eindhoven University of Technology , Eindhoven , The
Netherlands
Ian Klein Department of Chemistry and the Center for Catalytic
Conversion of Biomass to Biofuels (C3Bio) , Purdue University ,
West Lafayette , IN , USA
Spero Energy, Inc. , West Lafayette , IN , USA
G. Li Inorganic Materials Chemistry Group, Schuit Institute of
Catalysis, Department of Chemical Engineering and Chemistry ,
Eindhoven University of Technology , Eindhoven , The
Netherlands
Guanna Li Inorganic Materials Chemistry Group, Schuit Institute of
Catalysis, Department of Chemical Engineering and Chemistry ,
Eindhoven University of Technology , Eindhoven , The
Netherlands
x
Shuo Liu Department of Chemistry , Purdue University , West
Lafayette , IN , USA
Stéphane Loridant Institut de Recherches sur l’Environnement et la
Catalyse de Lyon (IRCELYON) , UMR 5256, CNRS – Université Lyon 1 ,
Villeurbanne Cedex , France
Carl R. F. Lund Department of Chemical and Biological Engineering ,
University at Buffalo , Buffalo , NY , USA
Mark R. Nimlos National Renewable Energy Laboratory, National
Bioenergy Center , Golden , CO , USA
Trenton Parsell Spero Energy, Inc. , West Lafayette , IN ,
USA
Sushil K. R. Patil Advanced Module Engineering , Globalfoundries ,
Malta , NY , USA
Evgeny A. Pidko Inorganic Materials Chemistry Group, Schuit
Institute of Catalysis, Department of Chemical Engineering and
Chemistry , Eindhoven University of Technology , Eindhoven , The
Netherlands
Institute for Complex Molecular Systems , Eindhoven University of
Technology , Eindhoven , The Netherlands
Catherine Pinel Institut de Recherches sur l’Environnement et la
Catalyse de Lyon (IRCELYON) , UMR 5256, CNRS – Université Lyon 1 ,
Villeurbanne Cedex , France
C. V. Pramod University of Twente , Enschede , The
Netherlands
David J. Robichaud National Renewable Energy Laboratory, National
Bioenergy Center , Golden , CO , USA
Basudeb Saha Department of Chemistry and the Center for Catalytic
Conversion of Biomass to Biofuels (C3Bio) , Purdue University ,
West Lafayette , IN , USA
Marcel Schlaf The Guelph-Waterloo-Centre for Graduate Work in
Chemistry (GWC)2, Department of Chemistry , University of Guelph ,
Guelph , ON , Canada
Ulf Schuchardt Instituto de Química – UNICAMP , Campinas , SP ,
Brazil
K. Seshan University of Twente , Enschede , The Netherlands
Jing Yi Department of Chemistry , Purdue University , West
Lafayette , IN , USA
Contributors
1© Springer Science+Business Media Singapore 2016 M. Schlaf, Z.C.
Zhang (eds.), Reaction Pathways and Mechanisms in Thermocatalytic
Biomass Conversion II, Green Chemistry and Sustainable Technology,
DOI 10.1007/978-981-287-769-7_1
Chapter 1 Deoxydehydration (DODH) of Biomass-Derived
Molecules
Shuo Liu , Jing Yi , and Mahdi M. Abu-Omar
Abstract Deoxygenation of vicinal diols and polyols, common
moieties in biomass- derived molecules, represents an important
chemical pathway for making chemicals from renewable biomass
resources. Catalytic deoxydehydration (DODH) is a promising
deoxygenation reaction that removes two adjacent hydroxyl (−OH)
groups from vicinal diols in one step to generate alkenes. Since
the fi rst catalytic DODH with Cp*Re(O) 3 report by Cook and
Andrews in 1996, a number of metal complexes based on rhenium,
ruthenium, vanadium, and molybdenum have been investigated.
High-valent oxorhenium complexes are among the most effi cient
cata- lysts for DODH reactions and have been studied using various
reductants including organic phosphines, molecular hydrogen (H 2 ),
sulfi te, and alcohols. These com- plexes exhibit intriguing
oxophilic performance, which facilitates selective C-O bond
cleavage of polyols. A fl urry of investigations have appeared in
the literature over the past few years on the scope and mechanism
of the DODH reaction in the context of biomass conversion and
sustainable chemistry. In this chapter, we briefl y review the
development of DODH reactions with a focus on homogenous
Re-catalyzed transformations. Several heterogeneous and other metal
catalysts are included for comparison.
Keywords Biomass • Deoxygenation • Deoxydehydration • Sustainable
chemistry • Polyols • Rhenium • Alkenes
S. Liu • J. Yi Department of Chemistry , Purdue University , West
Lafayette , IN 47907 , USA
M. M. Abu-Omar (*) Department of Chemistry and the Center for
Catalytic Conversion of Biomass to Biofuels (C3Bio) , Purdue
University , West Lafayette , IN 47907 , USA
Spero Energy, Inc. , 1281 Win Hentschel Blvd. , West Lafayette , IN
47906 , USA e-mail:
[email protected]
1.1 Introduction
Biomass-derived molecules can be used as a sustainable platform
feedstock to pro- duce high-value chemicals (HVCs) and biofuels [ 1
– 6 ]. Since carbohydrates, the major component of biomass, contain
one oxygen atom per carbon, deoxygenation of carbohydrates and
their derivatives such as sugar alcohols (polyols) represents a
major path forward for making renewable compounds that can
potentially replace petroleum-based chemicals. Current
deoxygenation methods include high- temperature pyrolysis [ 7 – 9
], acid-catalyzed dehydration [ 10 – 13 ], hydrogenolysis reactions
[ 14 , 15 ], and deoxydehydration (DODH) reactions (Fig. 1.1 ).
This chapter focuses on deoxydehydration (DODH), which removes two
adjacent hydroxyl (−OH) groups from vicinal diols in one step and
represents an effi cient approach to making olefi ns from renewable
feedstock [ 16 , 17 ].
1.2 DODH of Diols and Polyols Catalyzed by Rhenium Complexes
1.2.1 High-Valent Oxorhenium and Rhenium Carbonyl Catalysts
In a typical DODH reaction, the substrate (vicinal diol or polyol)
is reduced to pro- duce an alkene or allylic alcohol, while a
reducing agent is oxidized via oxygen atom transfer (OAT). Rhenium
complexes are among the most effi cient and well- studied
catalysts. Cook and Andrews reported in 1996 the fi rst catalytic
DODH of aromatic diols catalyzed by Cp*ReO 3 , employing
stoichiometric amount of PPh 3 as the reducing agent (Fig. 1.2a ) [
18 ]. The conversion was quite low; nevertheless, the addition of
Brønsted acid enhanced the rate of reaction signifi cantly. In the
presence of p -toluenesulfonic acid (TsOH), glycerol was reduced to
allylic alcohol using
Fig. 1.1 General strategies for deoxygenation of biomass-derived
sugar alcohols (polyols)
S. Liu et al.
3
Cp*ReO 3 , PPh 3 , and chlorobenzene in biphasic medium at 125 °C.
When unpro- tected tetritol was used as substrate in this system,
the main product was the fully deoxygenated butadiene (80 % yield),
with the rest as 3-butene-1,2-diol and cis-2- butene-1,4-diol in an
85:15 ratio (Fig. 1.2b ). The cis-2-butene-1,4-diol isomerized to
3-butene-1,2 diol under the catalysis conditions. Previously, the
Gable [ 19 – 25 ] and Herrmann groups [ 26 , 27 ] independently
synthesized and characterized a num- ber of oxorhenium(V) diolate
complexes, investigated alkene extrusion from rhe- nium diolate
complexes, and shed light on the mechanistic pathway of these
reactions. The mechanism of the DODH reaction of diols by
oxorhenium was sug- gested to proceed through three steps: (1)
reduction of Cp*ReO 3 by phosphine to generate Cp*ReO 2 , (2) diol
condensation with Cp*ReO 2 to generate Re(V)-diolate intermediate
and water, and (3) extrusion of alkene from the Re(V)-diolate to
regen- erate Cp*ReO 3 and complete the catalytic cycle.
In 2009 the Abu-Omar group reported on methyltrioxorhenium
(MTO)-catalyzed DODH/deoxygenation of vicinal diols and epoxides to
alkenes and alkanes with the more practical reductant H 2 (Fig. 1.3
) [ 28 ]. Under lower H 2 pressure, the products were dominated by
alkenes, whereas under higher pressure, by the alkane. This
indicated an extension of the DODH reaction to generate saturated
products. Several biomass-derived substrates were also tested in
the system. 1,4-anhydroerytheritol, formed by acid-catalyzed ring
closure of erythritol, yielded 25 % 2,5-dihydrofuran and 5 %
tetrahydrofuran. However, erythritol gave signifi cant charring.
The authors also noted that only cis-cyclohexane diol gave the
desired products, while trans-
a
b
Fig. 1.2 DODH reactions of aromatic diol ( a ) and erythritol ( b )
catalyzed by Cp*ReO 3 employing PPh 3 as a reductant
Fig. 1.3 DODH/deoxygenation of diols and epoxides by MTO using H
2
1 Deoxydehydration (DODH) of Biomass-Derived Molecules
4
cyclohexane diol did not react, consistent with observations in
other reports [ 18 ]. The specifi c stereoselectivity indicates the
formation of a metal diolate intermedi- ate, which would require
cis-vicinal diols in the two adjacent hydroxyl groups.
Soon afterwards, the Nicholas group reported MTO and perrhenate
salts cata- lyzed DODH of diols using sulfi te as the reductant
(Fig. 1.4 ) [ 29 , 30 ]. Both aromatic and aliphatic diols were
converted to the corresponding alkenes with good to mod- erate
yields, while the latter required longer reaction times. Addition
of the crown ether 15-crown-5 as a phase transfer agent was found
to signifi cantly shorten the reaction time and increase
conversions. [NBu 4 ][ReO 4 ] was a superior catalyst to MTO under
these conditions in terms of conversion of erythritol.
1,3-butadiene (27 % yield), 2,5-dihydrofuran (6 % yield), and
cis-2-betene-1,4-diol (3 % yield) were obtained using [NBu 4 ][ReO
4 ] as the catalyst, while substantial charring was observed for
MTO.
The Bergman group employed rhenium carbonyl as the catalyst (Re 2
(CO) 10 or BrRe(CO) 5 ) and a secondary alcohol as solvent and
reductant (Fig. 1.5 ) [ 31 ]. In the presence of simple alcohols
such as 3-octanol or 5-nonanol, both terminal and inter- nal
vicinal diols were deoxygenated to olefi ns with good yields, and a
stoichiomet- ric amount of the alcohol was oxidized to the
corresponding ketone. In the presence of TsOH, erythritol could be
converted to 2,5-dihydrofuran (62 % yield). Interestingly, the
system required air and high temperature for activation, which
indicated that the actual active catalyst is probably an oxidized
rhenium species and shares a similar reaction mechanism with
high-valent oxorhenium complexes.
Our group applied the MTO-catalyzed DODH reaction to glycerol,
under either neat conditions or in the presence of a sacrifi cial
alcohol (Fig. 1.6 ) [ 32 ]. The sub- strate, glycerol, participated
in transfer hydrogenation and deoxygenation to form volatile
products, for example, allyl alcohol, propanal, and acrolein,
leaving the nonvolatile dihydroxyacetone (DHA) as by-product in the
residual liquid reaction mixture. Under neat conditions, the
glycerol itself can function as reductant. Utilization of NH 4 ReO
4 as the catalyst gave similar results to MTO, with allyl alco- hol
as the major product. The addition of acid (HCl, NH 4 Cl) increased
the conver- sion rate and yield.
Fig. 1.4 DODH of diols by MTO using sulfi te
Fig. 1.5 DODH of diols by rhenium-carbonyl complexes in the
presence of sacrifi cial alcohol
S. Liu et al.
5
We have also investigated the MTO-catalyzed DODH reaction with
sacrifi cial alcohol as reductant and proposed a reaction mechanism
based on detailed kinetics and spectroscopic studies [ 33 ]. The
reaction kinetics were zero order in [diol] and half order in
[MTO]. The different induction periods of MTO and MDO
(methyldioxorhenium(V)) indicated that the active form of the
catalyst was MDO, which was formed by reduction of MTO by alcohol
or via a novel C-C bond cleav- age of an MTO-diolate complex. The
rate-determining step involved reaction with alcohol, and the
majority of the MDO-diolate complex was present in dinuclear form,
giving rise to the [Re catalyst] 1/2 (half-order) dependence.
Furthermore, the Re(V)-diolate did not extrude alkene at rates that
are consistent with the catalytic rate unless a reductant was
present. Thus we proposed a catalytic cycle in which MDO-diolate
was reduced further by 3-octanol to a transient rhenium(III)
diolate, which was responsible for alkene extrusion and
regeneration of MDO (Fig. 1.7 ). It
ReO O
Fig. 1.6 Rhenium-catalyzed DODH reactions of neat glycerol and
erythritol
Fig. 1.7 Mechanism of the major pathways for MTO-catalyzed DODH
reaction of vicinal diols in the presence of sacrifi cial
alcohol
1 Deoxydehydration (DODH) of Biomass-Derived Molecules
6
is worth noting that depending on the reducing agent, the
rate-limiting step may shift under different conditions.
The reaction mechanism(s) of the MTO-catalyzed DODH of diols was
also investigated by density functional theory (DFT) calculations [
34 , 35 ]. The DFT results supported the original reaction
mechanism proposed by Gable and Cook, alkene extrusion from
rhenium(V) diolate being the key rate-determining step.
The Toste group extended the reaction scope to larger sugar
alcohols and sugars, demonstrating the general effi ciency and high
selectivity of DODH [ 36 , 37 ]. C 3 -C 6 sugar alcohols can be
readily obtained by pretreatment of naturally abundant sugars, such
as hydrogenation, fermentation, and decarbonylation. These sugar
alcohols were converted to the corresponding olefi nic products
with good yields using either secondary alcohols or primary
alcohols (1-butanol) as reductants, albeit the former is more
favorable than the latter (Fig. 1.8 ). The linear alkene products
can be used as drop-in chemicals because they can serve as
precursors for polymers and fuels that are already in use in the
chemical industry. MTO displayed a higher activity than
rhenium-carbonyl complexes previously reported by the Bergman
group. When using a primary alcohol as the reducing agent,
rhenium-carbonyl catalysts didn’t generate any product, while
high-valent oxorhenium complexes afforded alkene products with high
yield and selectivity.
Following their initial work (Fig. 1.8 ), Toste and Shiramizu
expanded the reac- tion to 1,4-DODH and 1,6-DODH via tandem [1,
3]-OH shift-DODH process [ 36 ]. This process merged different
reaction intermediates into one product and thereby increased
product selectivity, providing a new strategy for DODH reaction
develop- ment for real biomass-derived sugar conversion (Fig. 1.9
). The reaction was also
Fig. 1.8 DODH reactions of C 3 -C 6 sugar alcohols catalyzed by
MTO
S. Liu et al.
7
applied to the conversion of sugar acids into unsaturated esters,
employing HReO 4 as both DODH catalyst and Brønsted acid. The
DODH/acid dual-catalyst strategy was applied to produce plasticizer
precursors from tartaric acid and erythritol.
1.2.2 Heterogeneous Rhenium Catalysts
There are few reports on heterogeneous DODH reactions. Jentoft and
Nicholas reported on heterogeneous polyol-into-olefi n DODH
reactions catalyzed by carbon- supported perrhenate, employing both
H 2 and hydrogen-transfer reductants with moderate yield [ 38 ].
Interestingly, in 2011 the Schlaf group reported that stainless
steel reactors could catalyze the deoxygenation of glycerol and
levulinic acid in aqueous acidic medium [ 39 ]. Ferdi Schuth
reported an iron oxide-catalyzed conver- sion of glycerol to
allylic alcohol and proposed a mechanism through dehydration and
consecutive hydrogen transfer [ 40 ]. Andreas Martin also reported
glycerol deoxygenation reaction in the gas phase using a series of
heteropolyacid catalysts [ 41 ]. Ir-ReOx/SiO 2 was also reported as
the catalyst for the production of butanediol from erythritol via
hydrogenolysis, by the Tomishige group [ 42 ]. More recently,
Nicholas reported on deoxydehydration of glycols using
heterogeneous elemental reductants such as zinc, iron, manganese,
and carbon [ 43 ]. The molecular identity of the catalyst is more
diffi cult to discern in these systems and the contribution from
the metal surface in the reactor per Schlaf’s study remains an open
question. Nevertheless, these studies demonstrate the feasibility
to translating the DODH reaction into a more practical process in
which continuous fl uidized bed reactors can be employed.
1.3 Other Transition Metal Catalysts for DODH
In addition to rhenium, other transition metal complexes based on
vanadium [ 44 ], molybdenum [ 45 , 46 ], and ruthenium [ 47 , 48 ]
have been used in DODH of vicinal diols. Nicholas reported a DODH
reaction of diols to olefi ns catalyzed by
Fig. 1.9 1,4-DODH and 1,6-DODH reactions via tandem [1, 3]-OH
shift-DODH process
1 Deoxydehydration (DODH) of Biomass-Derived Molecules
8
inexpensive metavanadate (VO 3 − ) and chelated dioxovanadium
derivatives, with phosphine or sulfi te as reductants [ 44 ].
Dioxomolybdenum(VI) complexes with acylpyrazolonate ligands were
synthesized by the Pettinari group, and these com- plexes showed
moderate activity toward diol DODH reactions with PPh 3 as reduc-
tant [ 45 ]. The Fristrup group also reported another DODH reaction
catalyzed by a series of Mo-oxo complexes under neat conditions [
46 ]. In addition, Schlaf and Bullock have pioneered the use of
organometallic ruthenium catalysts for the deox- ygenation of
alcohols [ 49 , 50 ]. Two other Ru-catalyzed DODH reactions were
reported by Srivastava and Nicholas, independently, using
[Cp*Ru(CO) 2 ] 2 or Ru(II)- sulfoxides as catalysts for
hydrodeoxygenation (HDO) and hydrocracking of diols and epoxides [
47 , 48 ].
1.4 Conclusion
As one of the promising strategies to selective removal of oxygen
atoms from biomass- derived molecules, DODH shows tremendous
potential to producing important fi ne and commodity chemicals from
renewable feedstock. Compared to conventional deoxygenation methods
such as hydrogenation and hydrogenolysis, the advantages of DODH
reaction are that the products are deoxygenated while remaining
synthetically versatile through the retention of a double bond,
making the most use of every atom in biomass. The successful
development of stable and highly active homogenous and
heterogeneous catalysts provides access to various syn- thetic
procedures to produce specifi c molecules with tailored properties
to be used as drop-in chemicals for making polymers and other
useful products. However, many challenges remain in utilizing
biomass-derived molecules and feedstock. Some of these challenges
include the ability to use carbohydrates and polysaccha- rides
directly and developing cost-effective, effi cient, and active
earth-abundant catalysts that can be employed under continuous
reaction conditions with facile product/catalyst separation.
Acknowledgment Our research in deoxygenation of biomass-derived
molecules has been sup- port by the Department of Energy (DOE),
Basic Energy Sciences (BES) grant no. DE-FG-02-06ER15794.
References
1. Huber GW, Iborra S, Corma A (2006) Synthesis of transportation
fuels from biomass: chemis- try, catalysts, and engineering. Chem
Rev 106(9):4044–4098. doi: 10.1021/cr068360d
2. Serrano-Ruiz JC, Luque R, Sepulveda-Escribano A (2011)
Transformations of biomass- derived platform molecules: from high
added-value chemicals to fuels via aqueous-phase pro- cessing. Chem
Soc Rev 40(11):5266–5281. doi: 10.1039/C1CS15131B
S. Liu et al.
3. Alonso DM, Bond JQ, Dumesic JA (2010) Catalytic conversion of
biomass to biofuels. Green Chem 12(9):1493–1513. doi:
10.1039/C004654J
4. Gallezot P (2012) Conversion of biomass to selected chemical
products. Chem Soc Rev 41(4):1538–1558. doi:
10.1039/C1CS15147A
5. Huber GW, Corma A (2007) Synergies between bio- and oil refi
neries for the production of fuels from biomass. Angew Chem Int Ed
46(38):7184–7201. doi: 10.1002/anie.200604504
6. Chheda JN, Huber GW, Dumesic JA (2007) Liquid-phase catalytic
processing of biomass- derived oxygenated hydrocarbons to fuels and
chemicals. Angew Chem Int Ed 46(38):7164– 7183. doi:
10.1002/anie.200604274
7. Zhang Q, Chang J, Wang T, Xu Y (2007) Review of biomass
pyrolysis oil properties and upgrading research. Energy Convers
Manage 48(1):87–92. doi: http://dx.doi.org/10.1016/j.
enconman.2006.05.010
8. Mohan D, Pittman CU, Steele PH (2006) Pyrolysis of wood/biomass
for bio-oil: a critical review. Energy Fuel 20(3):848–889. doi:
10.1021/ef0502397
9. Mettler MS, Vlachos DG, Dauenhauer PJ (2012) Top ten fundamental
challenges of biomass pyrolysis for biofuels. Energy Environ Sci
5(7):7797–7809. doi: 10.1039/C2EE21679E
10. Jin F, Enomoto H (2011) Rapid and highly selective conversion
of biomass into value-added products in hydrothermal conditions:
chemistry of acid/base-catalysed and oxidation reac- tions. Energy
Environ Sci 4(2):382–397. doi: 10.1039/C004268D
11. Tong X, Ma Y, Li Y (2010) Biomass into chemicals: conversion of
sugars to furan derivatives by catalytic processes. Appl Catal Gen
385(1–2):1–13. doi: http://dx.doi.org/10.1016/j.
apcata.2010.06.049
12. Rinaldi R, Schuth F (2009) Design of solid catalysts for the
conversion of biomass. Energy Environ Sci 2(6):610–626. doi:
10.1039/B902668A
13. Akien GR, Qi L, Horvath IT (2012) Molecular mapping of the acid
catalysed dehydration of fructose. Chem Commun 48(47):5850–5852.
doi: 10.1039/C2CC31689G
14. Corma A, Iborra S, Velty A (2007) Chemical routes for the
transformation of biomass into chemicals. Chem Rev
107(6):2411–2502. doi: 10.1021/cr050989d
15. Du X-L, Bi Q-Y, Liu Y-M, Cao Y, He H-Y, Fan K-N (2012) Tunable
copper-catalyzed che- moselective hydrogenolysis of biomass-derived
[gamma]-valerolactone into 1,4-pentanediol or
2-methyltetrahydrofuran. Green Chem 14(4):935–939. doi:
10.1039/C2GC16599F
16. Dutta S (2012) Deoxygenation of biomass-derived feedstocks:
hurdles and opportunities. ChemSusChem 5(11):2125–2127. doi:
10.1002/cssc.201200596
17. Metzger JO (2013) Catalytic deoxygenation of carbohydrate
renewable resources. ChemCatChem 5(3):680–682. doi:
10.1002/cctc.201200796
18. Cook GK, Andrews MA (1996) Toward nonoxidative routes to
oxygenated organics: stereo- specifi c deoxydehydration of diols
and polyols to alkenes and allylic alcohols catalyzed by the metal
oxo complex (C5Me5)ReO3. J Am Chem Soc 118(39):9448–9449. doi:
10.1021/ ja9620604
19. Gable KP, Phan TN (1994) Extrusion of alkenes from rhenium(V)
diolates: energetics and mechanism. J Am Chem Soc 116(3):833–839.
doi: 10.1021/ja00082a002
20. Gable KP (1994) Condensation of vicinal diols with the oxo
complex {Cp*Re(O)}2(.mu.-O)2 giving the corresponding diolate
complexes. Organometallics 13(6):2486–2488. doi: 10.1021/
om00018a048
21. Gable KP, Juliette JJJ (1996) Hammett studies on alkene
extrusion from rhenium(V) diolates and an MO description of metal
alkoxide−alkyl metal oxo interconversion. J Am Chem Soc
118(11):2625–2633. doi: 10.1021/ja952537w
22. Gable KP, Zhuravlev FA (2002) Kinetic Isotope effects in
cycloreversion of rhenium (V) dio- lates. J Am Chem Soc
124(15):3970–3979. doi: 10.1021/ja017736w
23. Gable KP, AbuBaker A, Zientara K, Wainwright AM (1998)
Cycloreversion of rhenium(V) diolates containing the
hydridotris(3,5-dimethylpyrazolyl)borate ancillary ligand.
Organometallics 18(2):173–179. doi: 10.1021/om980807o
1 Deoxydehydration (DODH) of Biomass-Derived Molecules
24. Kevin PG, Brian R (2006) Improved catalytic deoxygenation of
vicinal diols and application to alditols. In: Feedstocks for the
future, vol 921, ACS symposium series. American Chemical Society,
Washington, DC, pp 143–155. doi: 10.1021/bk-2006-0921.ch011
25. Gable KP, Juliette JJJ (1995) Extrusion of alkenes from
rhenium(V) diolates: the effect of substitution and conformation. J
Am Chem Soc 117(3):955–962. doi: 10.1021/ja00108a012
26. Herrmann WA, Marz D, Herdtweck E, Schäfer A, Wagner W, Kneuper
H-J (1987) Glycolate and thioglycolate complexes of rhenium and
their oxidative elimination of ethylene and of glycol. Angew Chem
Int Ed Engl 26(5):462–464. doi: 10.1002/anie.198704621
27. Herrmann WA, Marz DW, Herdtweck E (1990) Mehrfachbindungen
zwischen hauptgrup- penelementen und übergangsmetallen: LXXVIII.
Über oxo- und methylimido-komplexe des rheniums mit sauerstoff-,
schwefel- und stickstoffchelaten: Synthese, spaltungsreaktionen und
strukturchemie. J Organomet Chem 394(1–3):285–303. doi:
http://dx.doi. org/10.1016/0022-328X(90)87239-A
28. Ziegler JE, Zdilla MJ, Evans AJ, Abu-Omar MM (2009) H2-driven
deoxygenation of epoxides and diols to alkenes catalyzed by
methyltrioxorhenium. Inorg Chem 48(21):9998–10000. doi:
10.1021/ic901792b
29. Vkuturi S, Chapman G, Ahmad I, Nicholas KM (2010)
Rhenium-catalyzed deoxydehydration of glycols by sulfi te. Inorg
Chem 49(11):4744–4746. doi: 10.1021/ic100467p
30. Ahmad I, Chapman G, Nicholas KM (2011) Sulfi te-driven,
oxorhenium-catalyzed deoxydehy- dration of glycols. Organometallics
30(10):2810–2818. doi: 10.1021/om2001662
31. Arceo E, Ellman JA, Bergman RG (2010) Rhenium-catalyzed
didehydroxylation of vicinal diols to alkenes using a simple
alcohol as a reducing agent. J Am Chem Soc 132(33):11408– 11409.
doi: 10.1021/ja103436v
32. Yi J, Liu S, Abu-Omar MM (2012) Rhenium-catalyzed transfer
hydrogenation and deoxygen- ation of biomass-derived polyols to
small and useful organics. ChemSusChem 5(8):1401– 1404. doi:
10.1002/cssc.201200138
33. Liu S, Senocak A, Smeltz JL, Yang L, Wegenhart B, Yi J,
Kenttämaa HI, Ison EA, Abu-Omar MM (2013) Mechanism of
MTO-catalyzed deoxydehydration of diols to alkenes using sacri- fi
cial alcohols. Organometallics 32(11):3210–3219. doi:
10.1021/om400127z
34. Bi S, Wang J, Liu L, Li P, Lin Z (2012) Mechanism 1: mechanism
of the MeReO3-catalyzed deoxygenation of epoxides. Organometallics
31(17):6139–6147. doi: 10.1021/om300485w
35. Qu S, Dang Y, Wen M, Wang Z-X (2013) Mechanism 2: mechanism of
the methyltrioxorhenium- catalyzed deoxydehydration of polyols: a
new pathway revealed. Chem Eur J 19(12):3827– 3832. doi:
10.1002/chem.201204001
36. Shiramizu M, Toste FD (2012) Deoxygenation of biomass-derived
feedstocks: oxorhenium- catalyzed deoxydehydration of sugars and
sugar alcohols. Angew Chem Int Ed 51(32):8082– 8086. doi:
10.1002/anie.201203877
37. Shiramizu M, Toste FD (2013) Expanding the scope of
biomass-derived chemicals through tandem reactions based on
oxorhenium-catalyzed deoxydehydration. Angew Chem Int Ed
52(49):12905–12909. doi: 10.1002/anie.201307564
38. Denning AL, Dang H, Liu Z, Nicholas KM, Jentoft FC (2013)
Deoxydehydration of glycols catalyzed by carbon-supported
perrhenate. ChemCatChem 5(12):3567–3570. doi: 10.1002/
cctc.201300545
39. Di Mondo D, Ashok D, Waldie F, Schrier N, Morrison M, Schlaf M
(2011) Stainless steel as a catalyst for the total deoxygenation of
glycerol and levulinic acid in aqueous acidic medium. ACS Catal
1(4):355–364. doi: 10.1021/cs200053h
40. Liu Y, Tuysuz H, Jia C-J, Schwickardi M, Rinaldi R, Lu A-H,
Schmidt W, Schuth F (2010) Iron oxide: from glycerol to allyl
alcohol: iron oxide catalyzed dehydration and consecutive hydro-
gen transfer. Chem Commun 46(8):1238–1240. doi:
10.1039/B921648K
41. Atia H, Armbruster U, Martin A (2008) Poly acid: dehydration of
glycerol in gas phase using heteropolyacid catalysts as active
compounds. J Catal 258(1):71–82. doi: http://dx.doi.
org/10.1016/j.jcat.2008.05.027
S. Liu et al.
43. Michael McClain J, Nicholas KM (2014) Elemental reductants for
the deoxydehydration of glycols. ACS Catal 4(7):2109–2112. doi:
10.1021/cs500461v
44. Chapman G, Nicholas KM (2013) Vanadium-catalyzed
deoxydehydration of glycols. Chem Commun 49(74):8199–8201. doi:
10.1039/C3CC44656E
45. Hills L, Moyano R, Montilla F, Pastor A, Galindo A, Álvarez E,
Marchetti F, Pettinari C (2013) Dioxomolybdenum(VI) complexes with
acylpyrazolonate ligands: synthesis, structures, and catalytic
properties. Eur J Inorg Chem 2013(19):3352–3361. doi:
10.1002/ejic.201300098
46. Dethlefsen JR, Lupp D, Oh B-C, Fristrup P (2014)
Molybdenum-catalyzed deoxydehydration of vicinal diols. ChemSusChem
7(2):425–428. doi: 10.1002/cssc.201300945
47. Stanowski S, Nicholas KM, Srivastava RS (2012)
[Cp*Ru(CO)2]2-catalyzed hydrodeoxygen- ation and hydrocracking of
diols and epoxides. Organometallics 31(2):515–518. doi: 10.1021/
om200447z
48. Murru S, Nicholas KM, Srivastava RS (2012) Ruthenium (II)
sulfoxides-catalyzed hydroge- nolysis of glycols and epoxides. J
Mol Catal Chem 363–364(0):460–464. doi: http://dx.doi.
org/10.1016/j.molcata.2012.07.025
49. Thibault ME, DiMondo DV, Jennings M, Abdelnur PV, Eberlin MN,
Schlaf M (2011) Cyclopentadienyl and pentamethylcyclopentadienyl
ruthenium complexes as catalysts for the total deoxygenation of
1,2-hexanediol and glycerol. Green Chem 13(2):357–366. doi:
10.1039/ C0GC00255K
50. Ghosh P, Fagan PJ, Marshall WJ, Hauptman E, Bullock RM (2009)
Synthesis of ruthenium carbonyl complexes with phosphine or
substituted Cp ligands, and their activity in the catalytic
deoxygenation of 1,2-propanediol. Inorg Chem 48(14):6490–6500. doi:
10.1021/ic900413y
1 Deoxydehydration (DODH) of Biomass-Derived Molecules
Chapter 2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass- Derived Carbohydrate Feedstocks
Marcel Schlaf
Abstract The use of homogeneous rather than heterogeneous catalysts
for the hydrodeoxygenation of sugars, sugar alcohols, and their
condensates such as furfu- ral, 5-hydroxymethylfurfural, levulinic
acid, and iso sorbide may offer reaction pathways that have
distinct advantages, notably with respect to catalyst deactivation
by coking and fouling as observed on many heterogeneous systems
with these highly reactive and polar substrates. Homogeneous
systems, however, also face unique challenges in ligand, catalyst,
and process design. The catalyst systems employed will have to be
stable to the required aqueous acidic high-temperature (T > 150
°C) reaction conditions while exhibiting activities that make them
kineti- cally competent over acid-catalyzed decomposition and
oligo- and polymerization reactions leading to humin formation. For
each of the hydrodeoxygenation reaction cascades for the C3
(glycerol), C4 (erythritol), C5 (xylose and derivatives or
levulinic acid), and C6 (glucose and derivatives) value chains,
comparatively few homogeneous catalyst systems have been evaluated
to date. Key issues remain the thermal and redox stability of the
complexes employed against decomposition and reduction to bulk
metal acting as a heterogeneous catalysts and the recovery and
recycling of the catalyst from the often very complex reaction and
product mixtures.
Keywords Homogeneous catalysis • Hydrodeoxygenation • Sugar
alcohols • Carbohydrate biomass • Ruthenium-based complexes
M. Schlaf (*) The Guelph-Waterloo-Centre for Graduate Work in
Chemistry (GWC)2, Department of Chemistry , University of Guelph ,
Guelph , ON , Canada e-mail:
[email protected]
2.1.1 The Fundamental Challenges of Biomass
Hydrodeoxygenation
Carbohydrate biomass, i.e., essentially sugars and their polymers
and derivatives, is, as the name implies, constituted of molecules
of the general composition C n + n H 2 O and therefore
characterized by a high oxygen content – mainly in the form of
hydroxyl functions with a lower relative abundance of carbonyls and
ether or acetal and ketal linkages. This high oxygen content
imparts a high and typically nonspe- cifi c reactivity and at the
same time comparatively low energy density on the material.
Hydrodeoxygenation effects the net removal of oxygen from a
biomass-derived substrate to yield products of lower oxygen content
and reactivity and higher energy density that – for a complete
removal of oxygen – is equivalent to that of fossil hydrocarbons
used as liquid fuels.
An actual hydrodeoxygenation of carbohydrate-derived substrates
faces two fun- damental challenges that have to be met for an
economically viable and ecologically sustainable use of biomass as
a source of chemicals and fuels:
(a) The depolymerization and separation of recalcitrant biomass, in
particular the most abundant carbohydrates cellulose and
hemicellulose, into molecularly well-defi ned and soluble
substrates that can then be targeted for conversion to value-added
products.
(b) The partial selective or total catalytic hydrodeoxygenation of
soluble carbohy- drate monomers – glycerol (C3),
erythrose/erythritol (C4), xylose/xylitol (C5), and
glucose/sorbitol (C6) – or their condensates, e.g., furfural,
levulinic acid or 5-hydroxymethylfurfural (HMF), and isosorbide,
respectively, to less polar and less reactive products that can
serve directly as polymer components, solvents, or fuels and so
seamlessly integrate into existing large-scale (petro-) chemical
feed streams.
The application of transition metal-based homogeneous catalyst
systems to the second challenge is the subject of this chapter.
From a purely chemical viewpoint, this challenge is limited to only
a handful of reactions detailed below that for a given substrate
must typically occur in an iterative reaction cascade and specifi c
value chain to yield the desired deoxygenated products.
2.1.2 Reaction Patterns for the Hydrodeoxygenation of Carbohydrate
Biomass-Derived Substrates to Chemicals and Fuels
An effective removal of oxygen from highly oxygenated substrates is
only possible by rejection of either H 2 O or CO 2 . The former is
applicable to (poly-)alcohols by acid-catalyzed dehydration and the
latter to carboxylic acids, e.g., levulinic acid,
M. Schlaf
15
which is ultimately obtainable from cellulose via
5-hydroxymethylfurfural ( vide infra ). In a second step the
unsaturation or ring formation caused by the loss of water is then
followed by metal-catalyzed hydrogenation or hydrogenolysis result-
ing in a net removal of an oxygen atom from the substrate. Using
the linear sugar alcohols as an example, the reaction principle is
illustrated in Fig. 2.1 and can – when carried out in an iterative
reaction cascade – result in the total deoxygenation of the
substrates to alkenes or alkanes.
The type of functional groups present in carbohydrates and
substrates derived from them by acid-catalyzed dehydration is
limited to hydroxyl, carbonyl, carboxyl, alkene, and cyclic ether
functions, which in turn limits the number of fundamental reaction
patterns required for hydrodeoxygenation to the comparatively short
and deceptively simple (!) list presented in Fig. 2.2 [ 1 ].
Ideally any catalyst system tar- geting the hydrodeoxygenation of
such substrate should be promiscuous, i.e., be at least be capable
of catalyzing reactions i–iii shown in this fi gure, as this would
enable an actual process in a single reaction vessel. An obvious
potential limitation of hydrodeoxygenation reaction cascades
assembled from these fundamental reac- tions is the propensity of
the carbonyl and alkene intermediates formed to react with
themselves leading to uncontrolled oligomerization or
polymerization side reac- tions resulting in humin formation. The
side reactions are catalyzed by the necessar- ily present acid,
which then effectively reverse the depolymerization of (hemi-)
cellulose required to obtain soluble hydrodeoxygenation target
substrates in the fi rst place, i.e., negating any effort expended
to meet challenge (a) discussed previously.
Any viable catalyst system and process must therefore be
kinetically competent with respect to these side reactions by
converting the reactive alkene and carbonyl intermediates to more
stable oxacycles, primary alcohols and alkanes faster than they can
nonspecifi cally (re-)condense to insoluble macromolecules.
Regardless of their homo- or heterogeneous nature, this constitutes
the main problem in the design and development of catalyst systems
and processes for a successful hydrodeoxy- genation of highly
reactive biomass-derived carbohydrate substrates.
Fig. 2.1 Hydrodeoxygenation as an iterative reaction of
acid-catalyzed dehydrations and metal- catalyzed hydrogenations
resulting in the net loss of oxygen
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
16
2.2 Why Homogeneous?
2.2.1 Historic Perspective
Beginning with the historic development of processes for the
production of xylitol and sorbitol by hydrogenation of xylose and
glucose over Raney Nickel [ 2 ], the reductive cyclization of
levulinic acid to γ-valerolactone over PtO 2 by Sabatier and
Schuette [ 3 , 4 ], and the more recent seminal paper by Descotes
describing the hydrogenation of 5-hydroxymethylfurfural in aqueous
medium over Ni, CuO•CrO 3 , Pd/C, Pt/C, Ru/C, PtO 2 , as well as Ru
and Pt metal [ 5 ], the development of hydro- deoxygenation
catalysts has mainly focused on heterogeneous systems employing
various combinations of the same hydrogenating metals. Typically
the metals are supported on metal oxides, e.g., Al 2 O 3 , SiO 2 ,
ZrO 2 , or Nb 2 O 3 , which often also play the role of the acid
component of the system, as can carbon supports bearing acidic
surface functionalities (e.g., carboxylic or sulfonic acids).
Alternatively a solid acid such as WO 3 can be added separately .
Substantial progress has been made in this endeavor and the fi eld
has been extensively reviewed [ 6 – 13 ]. Comparatively much
Fig. 2.2 Fundamental reactions for the hydrodeoxygenation of
carbohydrate-derived substrates
M. Schlaf
17
fewer efforts have been directed at the development of
hydrodeoxygenation catalyst systems, in which both the acid and the
hydrogenating metal are present in homo- geneous molecular
dispersed phase [ 14 ].
2.2.2 Heterogeneous vs. Homogeneous Catalysts for
Hydrodeoxygenation
As in any catalytic reaction or process, the use of either hetero-
or homogeneous systems has distinct advantages and disadvantages.
Adopting and expanding a com- parison originally given by Cornils
and Herrmann [ 15 ], Table 2.1 lists these with special
consideration given to the unique properties of the highly polar
carbohydrate substrates targeted.
As a consequence of the necessarily aqueous acidic reaction
conditions encoun- tered in acid/metal-catalyzed hydrodeoxygenation
reaction cascades (Fig. 2.1 ) and the undifferentiated high
reactivity of the substrates, heterogeneous catalysts are
susceptible to challenges that are typically not – or only to a
much lesser extent – encountered in their applications to fossil
carbon-derived hydrocarbon feeds for which most of them were
originally developed. In particular, they are vulnerable to
leaching of the active hydrogenating metal into solution,
degradation, or complete destruction of oxide supports such as SiO
2 , Al 2 O 3 or similar by the corrosive aqueous acidic reaction
medium under what are effectively hydrothermal conditions and/or
rapid fouling/coking and hence deactivation due to condensation of
the highly polar substrates on the also polar catalyst surface [ 13
]. The latter is in principle reversible
Table 2.1 Advantages and disadvantages of hetero- vs. homogeneous
catalysts applied to the hydrodeoxygenation of biomass-derived
carbohydrate substrates
Heterogeneous catalysis Homogeneous catalysis
Activity (rel. to metal content) Variable High Selectivity Variable
Variable Reaction conditions Harsh Moderate to harsh Service life
Long Variable Sensitivity to poisons High Low Diffusion limitations
Can be a limiting factor None Catalyst recycling Not necessary Very
challenging Temperature stability High Needs to be “designed in”
Stability against aqueous acids
Typically low due to leaching and support degradation
Needs to be “designed in”
Sensitivity to coking High None Sensitivity to fouling High None
Coordinative inhibition None Can be a limiting factor Electronic
and steric design Not possible Rationally changeable Mechanistic
understanding Very diffi cult Plausible and achievable
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
18
by oxidative reactivation (i.e., burn-off of deposited carbonaceous
material followed by reduction), but is however – by defi nition –
not applicable to the very promising carbon-supported catalysts
systems, e.g., Ru/C.
Homogeneous catalysts will not suffer from any of these
limitations, but face their own challenges when applied to
hydrodeoxygenation reaction cascades. Empirically the
acid-catalyzed loss of water from polyalcohol substrates (reaction
i in Fig. 2.2 ) or hydrolysis of furan rings (as encountered, e.g.,
in HMF and its deriva- tives) to α,γ-diones requires temperatures
in excess of 150 °C, while a direct ring opening of THF-type
substrates will require 200–300 °C at elevated pressures [ 16 ],
with all of these reactions occurring in an aqueous acidic medium.
Together these reaction conditions – listed as “moderate to harsh”
rather than “mild” in Table 2.1 – will require the rational design
and synthesis of transition metal complex catalysts of
unprecedented thermal , acid , and general chemical stability while
at the same time offering satisfactory activity toward the
hydrogenation of C = C and C = O bonds at reasonably accessible
pressures of hydrogen gas. This constitutes the fi rst major
challenge to the use of homo- rather than heterogeneous catalyst
systems for hydrodeoxygenation and can be addressed through the use
of highly chelating ligands that result in very high complex
formation constants [ 17 ]. The main role of the ligand in this
instance is the stabilization of the metal center against reduction
to oxidation state zero and precipitation of bulk metal
(effectively generating a hetero- geneous catalyst) rather than the
induction of chemo-, regio-, or stereoselectivity as usually sought
after in homogeneously catalyzed reactions.
The second major challenge to an economically and technically
viable use of homogeneous catalysts is the requirement for their
effective and facile reuse and recycling, in particular when
considering that the catalysts will likely have to be based on
platinum group metals, e.g., Ru, Ir, Pd, or Pt. The desired
hydrodeoxygen- ation reaction products will – especially if a total
deoxygenation to alkenes or alkanes is realized – have a much lower
polarity and solubility in polar solvents than the sugar(-derived)
starting material. This very feature, identifi ed as a problem for
heterogeneous systems, can work in favor of homogeneous catalyst
systems, provided they are designed as polar solvent-soluble salts
(ideally in water), e.g., by combining a cationic
hydrogen-activating transition metal complex with a hydrolysis-
stable non-coordinating counter anion such as phosphate, sulfate,
or trifl uoromethanesulfonate (“trifl ate”) [ 18 ].
Figure 2.3 illustrates a – to date unrealized – vision, in which
the polarity and solubility differences between the highly polar
substrates and catalyst(s) and the nonpolar deoxygenated products
are exploited leading to an “automatic” product isolation and
catalyst system recovery and recycling by phase separation using
the total deoxygenation of sorbitol to hexane as a conceptual
example. Even in cases where the hydrodeoxygenation reaction
cascades lead to high-value-added products that retain one or two
oxygen atoms, e.g., tetrahydrofuran, α,ω-diols, or mono-ols, this
strategy may still be viable, if the deoxygenates can be recovered
by (azeo- tropic) distillation from the aqueous mixture
regenerating the acid/catalyst solution. With product isolation and
catalyst recycling by phase separation, even a continu- ous process
may become possible.
M. Schlaf
2.3.1 Hydrodeoxygenations of C2 to C6 Substrates
2.3.1.1 Ethanol
Due its high production volume, to date mainly from corn starch
(US) and sugar- cane (Brazil), but anticipated to shift to
cellulosic feeds in the near future, ethanol is an attractive
starting material. While not strictly a hydrodeoxygenation reaction
in the sense defi ned in Fig. 2.1 , the recent revival of the
Guerbet reaction [ 19 – 21 ], which is conducted under basic rather
than acidic conditions, opens a pathway for the conversion to
n-butanol. As a fuel this is substantially superior to ethanol and
arguably more importantly can serve as an entry point into the C4
petrochemical product manifold. A conceivable simplifi ed catalytic
cycle for the Guerbet reaction of ethanol to 1-butanol with [Ru(H)
2 (dppm) 2 ] as an actually used catalyst is proposed in Fig. 2.4 [
22 ]. The reaction employs the concept of borrowed hydrogen [ 23 ],
in which a transition metal complex fi rst acts as the
dehydrogenation catalysts for the
Fig. 2.3 A vision for a homogeneously catalyzed hydrodeoxygenation
of sugar(-derived) sub- strates by homogeneous catalysts systems
with product isolation and catalyst recycling effected by phase
separation using the sorbitol to hexane transformation as an
example
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
20
conversion of the alcohol to acetaldehyde, temporarily stores
hydrogen in form of hydride ligands, and then acts as a
hydrogenation catalyst for the re-addition of hydrogen to the
α,β-unsaturated crotonaldehyde resulting from the aldol condensa-
tion of two equivalents of acetaldehyde. Iridium – as well as
ruthenium-based diphosphine and phenanthroline chelate complexes –
has been successfully employed realizing selectivities > 90 % at
up to 20 % conversion [ 22 , 24 – 26 ].
2.3.1.2 Glycerol
Due to the large amounts of glycerol generated as the by-product of
biodiesel pro- duction by the transesterifi cation of triglycerides
with methanol or ethanol, glycerol is a very attractive target for
conversion to value-added deoxygenated, carbonylated, or otherwise
modifi ed products. A hydrodeoxygenation reaction cascade and the
products obtainable in principle from glycerol (excluding
polymerization) are shown in Fig. 2.5 . Glycerol (in very pure,
i.e., food-grade form) itself has direct applications, e.g., in
cosmetics; however, its singly deoxygenated derivatives,
1,2-propanediol and in particular 1,3-propanediol [ 27 ], the key
component for the manufacture of polytrimethylene terephthalate
(PTT) and polytrimethylene glycol, offer substantial value
addition, while the doubly or totally deoxygenated products
Fig. 2.4 Proposed mechanism of the Guerbet reaction of ethanol to
1-butanol using [Ru(H) 2 (dppm) 2 ] as an example homogeneous
catalyst
M. Schlaf
21
propanol and propene/propane are, while still useful, less
desirable due the relatively higher amount of hydrogen required for
their production and lower relative value.
Evident from Fig. 2.5 is that a successful acid-/metal-catalyzed
glycerol hydro- deoxygenation process resulting in conversion and
selectivity to a single product in high yield is challenging and
requires a very careful adjustment and tuning of the catalyst and
reaction parameters in order to select a single dominant reaction
chan- nel from the cascade. Under aqueous acidic conditions, the
double or total deoxy- genation via the acrolein pathway leading
beyond the desired diol products is of special concern. Research
efforts have again been mainly focused on heterogeneous catalysts
and processes [ 28 – 32 ]; however, glycerol was also the subject
of the earli- est work on a homogeneously catalyzed
hydrodeoxygenation with a solid acid and the seminal work of Braca
et al. employing ruthenium iodocarbonyl complexes (Fig. 2.6 ) [ 33
, 34 ]. For both reactions relatively high pressures of carbon
monoxide serve as a source of carbonyl ligands stabilizing the
metal against reduction to oxi- dation state zero. A further
similar example is a system composed of a palladium diphosphine
complex and (trifl uro)methane sulfonic acid patented by Shell (E.
Drent) that achieves comparable conversions to 1,3-propanediol in
water/sulfo- lane mixtures under similar reaction conditions (170
°C, 6 MPa CO/H 2 = 1:2) [ 35 ].
The generation of small amounts of THF by the ruthenium system
reported by Braca (Fig. 2.6 , bottom) is intriguing, as it points
the way for a possible synthesis of a very valuable C4 product
manifold by the deoxycarbonylation rather hydrodeoxy- genation of
cheap and abundant glycerol. This approach was originally described
in a Japanese patent by Nakamura and has recently been further
explored using
Fig. 2.5 Reaction cascade and pathways for the hydrodeoxygenation
of glycerol to value-added products
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
22
rhodium or iridium iodocarbonyl complexes generated in situ from
the corresponding chlorides and methyl iodide in acetic acid/water
mixtures at 130–180 °C and CO pressures of 30–120 bar. Depending on
the actual conditions using various blends of allyl acetate,
(iso-)butyric acid and vinyl acetic and vinyl crotonic acid can be
obtained [ 36 – 38 ].
A series of ruthenium-based complexes known or postulated to
operate as ionic hydrogenation catalysts [ 39 ], i.e., by a
heterolytic activation of hydrogen gas into a hydride ligand and,
under aqueous conditions, H 3 O + as the strongest possible
solvent- leveled acid in the reaction mixture was developed by
Bullock and the author and tested as catalysts against terminal
diols as model systems for glycerol, achieving – depending on
reaction conditions – partial or total hydrodeoxygenation to either
primary alcohols or alkenes/alkanes, respectively. The
deoxygenation of diols proceeds by the reactions (i) and (ii) in
Fig. 2.2 . Figure 2.7 proposes a heterolytic dihydrogen activation
mechanism, in which an aqua ligand, e.g., in the complexes cis
-[(6,6′-Cl 2 -2,2′-bipyridine) 2 Ru(H 2 O) 2 ][OTf] 2 [ 40 – 42 ]
or [(4′-Ph-terpyridine)Ru(H 2 O) 3 ](OTf) 2 [ 43 ], is displaced by
a dihydrogen ligand forming a transient highly acidic (η 2 -H 2 )
complex that is deprotonated by water to give the corresponding
hydride as the active reductant. In this context it should be noted
that the metal–ligand bond enthalpy of the M-(η 2 -H 2 ) unit is
estimated to be very similar to that of M-(OH 2 ) and that the
direct generation of nonclassical dihy- drogen complexes from aqua
complexes has been demonstrated even in water as the solvent [ 44 ,
45 ]. Water-soluble cationic aqua complexes with weakly
coordinating
Fig. 2.6 First examples of homogeneously catalyzed
hydrodeoxygenation transformations of glycerol
M. Schlaf
23
counter ions may therefore be regarded as almost ideal
pro-catalysts for the homogeneously catalyzed hydrodeoxygenation of
polyalcohols in aqueous medium. Displacement of a second water
ligand (in either sequence) gives a hydride ketone/ aldehyde (or
alkene as relating to reaction ii in Fig. 2.2 ) complex that reacts
by insertion of the C = O (or C = C) bond into the metal-hydride
bond. Protonation of the resulting alkoxide (or alkyl) complexes
releases the product restarting the cycle.
The complexes tested however all failed to achieve the most
desirable conver- sion of glycerol to 1,3-propanediol [ 40 , 43 ,
46 – 49 ]. Instead the complex [(4′-Ph-terpyridine)Ru(H 2 O) 3
](OTf) 2 effects – depending on reaction temperature, hydrogen
pressure, solvent, and added HOTf acid concentration – a partial
deoxy- genation to n -propanol in up to 35 % or total deoxygenation
of glycerol to propene/ propane in quantitative yield [ 43 ].
A successful selective deoxygenation to 1,3-propanediol was
ultimately realized by heterogeneous systems developed by Tomishige
[ 50 ] and relevant to this chapter using an iridium pincer system
patented by Heinekey and Goldberg [ 51 , 52 ]. In an aqueous acidic
1,4-dioxane solution at 200 °C and 8.1 MPa, the complex [(κ 3 -C 6
H 3 - 1,3-[OP( t Bu) 2 ] 2 )Ir(CO)] reaches conversions of up to 45
% with a product selectiv- ity of 1:4 of 1,3-propanediol/ n
-propanol. As with the ruthenium complexes, this system is
postulated to operate as an ionic hydrogenation catalyst with a
heterolytic activation of hydrogen gas split into a proton, taken
up by a base from the reaction medium (water, solvent, substrate)
and hydride ligand on the iridium center. The proposed catalytic
cycle with and without explicit formulation of an acidic η 2 - H 2
ligand is shown in Fig. 2.8 and begins with the protonation of the
neutral pro- catalyst to what is formally an Ir III complex as the
hydrogen-activating species.
Fig. 2.7 Proposed mechanism for the heterolytic activation of
dihydrogen and hydrogenation of carbonyl and alkene bonds by
ruthenium aquo complexes
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
24
A remarkable metal-free glycerol hydrodeoxygenation, in which
formic acid acts as both the acid catalyst and the reductant,
achieves a conversion to allyl alcohol in 80 % yield at 230 °C [ 53
]. The reaction is also applicable to the conversion of erythritol
to 2,5-dihydrofuran, and a proposed mechanism is shown in Fig. 2.9
.
2.3.1.3 Erythritol
The comparatively lower availability and hence higher price of
erythritol, which is typically prepared by yeast fermentation of
glucose (e.g., by Moniliella pollinis ), appear to date have
limited its use as a hydrodeoxygenation target. However, if
Fig. 2.8 Proposed mechanism for the conversion of glycerol to
1,3-propanediol by [(κ 3 -C 6 H 3 -1,3- [OP( t Bu) 2 ] 2 )Ir(CO)]
in aqueous acidic solution under hydrogen pressure
Fig. 2.9 Proposed mechanism of the metal-free hydrodeoxygenation of
glycerol to allyl alcohol
M. Schlaf
25
competitive sources of erythritol became available, its
hydrodeoxygenation would offer a direct route to the important C4
building block 1,4-butanediol, a key compo- nent of the elastic
polyester–polyurethane copolymers LYCRA TM and SPANDEX TM and of
course THF solvent (Fig. 2.10 ).
Again most efforts toward the hydrodeoxygenation reactions shown in
Fig. 2.10 have been focused on the use of heterogeneous systems,
e.g., the production of THF in ~ 50 % yield over a Re/Pd/SiO 2
/Nafi on catalyst patented by Manzer [ 54 ] or the more recent
development of the already mentioned Ir–ReOx/SiO 2 systems by
Tomishige [ 43 ]. The only homogeneous hydrodeoxygenation of this
substrate to the potentially very valuable 2,5-dihydrofuran (Fig.
2.11 ) was achieved by formic acid via 1,4-anhydroerytrhitol in
analogy to the already discussed glycerol transforma- tion [ 53
].
2.3.1.4 Xylose, Furfural, and Xylitol
The high abundance of hemicellulose, its comparatively facile
hydrolysis to xylose followed by hydrogenation to xylitol, and its
well-established direct conversion to furfural make these three C5
units attractive hydrodeoxygenation targets [ 2 , 55 ]. However,
starting with a carbon-chain length of fi ve (or longer – see
below), an overall very complex hydrodeoxygenation reaction can
cascade result, in which the actual pathways and observed
intermediates followed and observed depend on (a) whether the
reaction cascade begins with hydrogenation of xylose to xylitol or
dehydration and rearrangement to furfural and (b) the relative
concentrations of substrate, water, acid, and catalysts employed.
Excluding any substrate or product
dimerization/oligomerization/polymerization pathways, e.g., those
leading to humin formation [ 56 ], Fig. 2.12 shows the overall
hydrodeoxygenation reaction cascades with the conceivable
intermediates and pathways ending in 2-methy- tetrahydrofuran,
tetrahydropyran, and the total hydrodeoxygenation product pen- tane
as stable terminal products under reducing conditions [ 57 ].
Neither hetero- nor homogeneous catalysts have to date realized
this entire C5 reaction cascade, value chain, and pathways.
Intermediates and products observed with homogeneous systems are
limited to furfural → furfuryl alcohol,
Fig. 2.10 Hydrodeoxyge- nation of erythritol to 1,4-butanediol and
THF
Fig. 2.11 Hydrodeoxyge- nation of erythritol 2,5-dihydrofuran by
formic acid
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
26
furfuryl alcohol → 2-methylfuran and furfural → tetrahydrofurfuryl
alcohol, while the potentially very valuable 1,5-pentanediol has to
date only been generated using heterogeneous catalysts [ 58 – 60 ].
In addition to the pathways shown in Fig. 2.12 , heterogeneous
catalysts have also been used to generate 1,2-pentanediol and
cyclo- pentanone directly from furfural [ 61 – 65 ].
Employing the complex cis -[(6,6′-Cl 2 -2,2′-bipyridine) 2 Ru(H 2
O) 2 ][BArF] as a homogeneous catalyst, Lapido et al. have
demonstrated part of the reaction cascade as shown in Fig. 2.13
leading from furfural to furfuryl alcohol, 2-methylfuran, and
tetrahydrofurfuryl alcohol [ 66 ]. This catalyst already mentioned
above was origi- nally developed as an alkene hydrogenation
catalyst operating in aqueous medium by Lau [ 41 , 42 ] and has
also (as the trifl ate salt) been successfully employed for the
partial or complete hydrodeoxygenation of terminal diols to primary
alcohols or alkanes in water [ 40 ]. In ethanol solvent and with
perfl uoro tetra-aryl borate as the counterion, it achieves very
high conversions of furfural to furfuryl alcohol (93 %) and
essentially quantitative conversion to tetrahydrofurfuryl alcohol
(99 %) under 5.16 MPa hydrogen at 85 and 130 °C, respectively.
Under similar conditions, the trifl ate salt catalyst generates
20–25 % of 2-methylfuran, effectively a benzylic hydrogenolysis,
underlining the potentially large infl uence of solvent and
counter- ion effects on catalyst reactivity and selectivity in
homogeneous systems assuming that the catalyst is in fact still
homogeneous in this case [ 66 ].
Fig. 2.12 Hydrodeoxygenation reaction cascade for
hemicellulose/xylose showing all conceiv- able intermediates and
pathways
M. Schlaf
27
Sen has applied the RhCl 3 /HI (1/1.5) system to the conversion of
xylose in a biphasic water/chlorobenzene mixture at 140 °C and 2
MPa hydrogen realizing 80 % yield of 2-methyl-tetrahydrofuran at 95
% substrate conversion. This system is further discussed in more
detail in the context of the C6 hydrodeoxygenation targets in Sect.
2.3.1.6 .
2.3.1.5 Levulinic Acid
A second C5 value chain starting from levulinic acid – obtained by
rehydration and deformylation of 5-hydroxymethylfurfural (HMF) –
can lead to the same terminal stable products as when starting from
hemicellulose. Figure 2.14 shows all conceiv- able intermediates
and pathways for this substrate, again all of which have been
realized with heterogeneous catalyst systems [ 4 , 67 – 69 ]. For
homogeneous cata- lysts, the cascade has to date focused on the
comparatively facile conversion to γ-valerolactone, which has been
considered as a key platform molecule that can serve as solvent,
fuel, or precursor to other value-added chemicals as laid out in
Fig. 2.14 and has been generated in high yield by ruthenium- and
iridium-based hydrogenation catalysts supported by phosphine and
pincer ligands [ 70 – 73 ]. The complexes can be prepared in situ
from “standard” precursors such as [Ru(acac) 3 ] or [Ir(COE) 2 Cl 2
], which for the latter with the PNP pincer ligand
2,6-di(bis-tbutyl- phosphine)methyl-pyridine can reach TON of
70,000 [ 74 , 75 ]. Using formic acid as the reductant, which
eliminates the need for hydrogen pressure, Horvath has successfully
applied the metal–ligand bifunctional Shvo catalyst to this
transformation achieving TON of ~ 1000 with repeated recycling of
the catalyst without loss of activity [ 76 – 78 ].
A fi rst example of the use of a homogeneous catalyst system
reaching beyond γ-valerolactone is part of the conceptual value
chain from sucrose to alkanes devel- oped by Horvath and given in
Fig. 2.15 [ 79 ], which integrates the acid-catalyzed hydrolysis,
dehydration, and deformylation of sucrose or HMF to levulinic acid
with
Fig. 2.13 Actually realized (homo- and heterogeneously catalyzed)
hydrodeoxygenation reaction cascades for furfural
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
28
Fig. 2.14 Hydrodeoxygenation reaction cascade for levulinic acid
showing all conceivable intermediates and pathways
Fig. 2.15 Integrated levulinic acid centered value chain leading
from sucrose and HMF to alkanes by combination of homo- and
heterogeneously catalyzed reactions
M. Schlaf
29
three alternative homogeneously catalyzed pathways for the
hydrodeoxygenation of levulinic acid and as a fi nal step the
heterogeneously catalyzed conversion of 2-methyl-THF to alkanes
using Pt(acac) 2 as the pro-catalyst that in the absence of
supporting ligands likely forms Pt 0 particles as the active
system. Starting from levulinic acid, the right branch of the
reaction cascade in Fig. 2.15 describes the use of a water-soluble
catalyst formed in situ from sulfonated triphenylphosphines P(m- C
6 H 4 SO 3 Na ) 3 and Ru(acac) 3 for the initial conversion of
levulinic acid to γ-valerolactone followed – after product
isolation – by a second hydrogenation to 2-methyl-THF under
slightly more forcing and acidic (NH 4 PF 6 additive) solvent- free
conditions using PBu 3 as the supporting ligand. At higher
temperatures and starting directly from levulinic acid as the
reactant and solvent (left branch of the reaction cascade in Fig.
2.15 ), the same catalyst realizes hydrogenolysis of the ini- tial
γ-valerolactone product to the potentially valuable 1,4-pentanediol
in 63 % yield. Alternatively, and avoiding the requirement for H 2
pressure, [(η 6 -C 6 H 6 ) Ru(1,10-phenanthroline)]SO 4 can act as
a transfer hydrogenation catalyst using sodium formate as the
reductant and HNO 3 as the acid cocatalyst [ 80 ].
Another unique homogeneous system capable of deoxygenating
levulinic acid beyond the initial product γ-valerolactone was
developed by Leitner, Klankenmayer, and coworkers [ 81 , 82 ]. The
combination of the complex [Ru(triphos)(CO)H 2 ], gen- erated in
situ under reducing conditions from [Ru(acac) 3 ] and the triphos
ligand (1,1,1-tris(diphenylphosphinomethyl)ethane), and an acidic
imidazolium-based ionic liquid and/or NH 4 PF 6 as acid cocatalysts
enables the generation of 2-methyl- THF in up to 92 % yield. Figure
2.16 summarizes the transformation, catalyst struc- tures, and
reaction conditions.
The authors also performed a preliminary engineering and
techno-economical analysis of a continuous process that indicated
that with levulinic acid as the sub- strate and solvent, i.e.,
under effectively solvent-free conditions, effi cient catalyst
recycling should be possible and that the overall costs of the
process would be dominated by those of the catalyst and raw
material making geographically distributed small-scale operations –
a key point for any biorefi nery – potentially economically
attractive.
Fig. 2.16 Homogeneously catalyzed hydrodeoxygenation of levulinic
acid to 2-methyl-THF
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
30
2.3.1.6 Glucose, Fructose
With glucose units as the repeat unit of cellulose and therefore
likely the most abun- dant biomolecule on the planet, the C6
substrates are arguably the most important hydrodeoxygenation
targets. Similar to the C5 substrates, a very complex reaction
cascade can result and Fig. 2.17 shows the intermediates and
pathways ending in 2,5-dimethyl-THF, hexane and methylcyclopentane,
and conceivably dianhydrosor- bitol [ 83 ] as the terminal stable
products under reducing conditions with HMF, 2,5-dimethylfuran,
2,5-hexanedione, and iso sorbide as the key reactive
intermediates.
As with the C5-based substrates, approaches to the
hydrodeoxygenation of glu- cose and its derivatives have to date
been dominated by the use of heterogeneous catalysts with only a
handful of examples attempting to use homogeneous systems described
in the literature.
A simple example of the application of a homogeneous catalyst to a
biomass- derived substrate is the use of the metal–ligand
bifunctional Shvo catalyst to the hydrogenation of HMF to
2,5-dihydroxymethyl-furan [ 76 , 77 , 84 ]. The reaction proceeds
quantitatively (99 % yield) under comparatively mild conditions (90
°C, 1 MPa H 2 ), and the catalyst can be recycled without loss of
activity at least nine
Fig. 2.17 Hydrodeoxygenation reaction cascade for cellulose/glucose
showing all conceivable intermediates and pathways
M. Schlaf
31
times by isolating the product through precipitation and fi
ltration from the toluene reaction solution.
Sen et al. have reported the use of HI as a reductant capable of
effecting both dehydration and simultaneous partial reduction of C6
sugars, most notably fructose to 5-methylfurfural in up to 47 %
with the concomitant production of I 2 . The HI reductant can then
be regenerated in situ by reaction of I 2 with hydrogen over a sup-
ported noble metal (Pd/C, Rh/C, Ru/C) [ 85 ]. More relevant to this
chapter is the extension of this concept to the use of a catalyst
system consisting of RhCl 3 , HI, and NaI that results in the
conversion of fructose, HMF, 5-methyl-furfural, 2,5- hexanedione,
and 2,5-hexanediol to 2,5-dimethyl-THF in up to quantitative yield
following the same reaction pathways as laid out in Fig. 2.17 [ 86
– 88 ]. While the true identity of the catalyst could not be
established, fi ltration of the biphasic toluene/water reaction
mixture gave an aqueous phase that maintained its catalytic
activity and could be recycled multiple times by adding fresh
fructose substrate. When HCl rather than HI was used as the acid
cocatalyst, precipitation of Rh(0) with lower and different
catalytic activity was observed with no formation of
2,5-dimethyl-THF. This suggests that in the presence of HI, the
catalyst formed in situ from RhCl 3 is in fact homogeneous, likely
a complex or soluble cluster of unknown structure supported by
iodide ligands [ 86 ]. Notably the presence of HI rather than HCl
also almost completely suppressed the hydrogenation of the toluene
solvent phase to methylcyclohexane that is observed with the
heterogeneous Rh(0) catalyst generated in situ from RhCl 3
/HCl.
The author’s group has demonstrated that the complex
[(4′-Ph-terpyridine) Ru(H 2 O) 3 ](OTf) 2 , previously applied to
terminal diols and glycerol (see Sect. 2.3.1.2 ), is also effective
in the hydrogenation of 2,5-hexanedione, either targeted directly
or generated in situ by hydrolysis of 2,5-dimethylfuran (Fig. 2.17
) to 2,5-hexanediol and 2,5-dimethyl-tetrahydrofuran and also
evaluated the iso - electronic iridium system
[(4′-Ph-terpyridine)Ir(H 2 O) 3 ](OTf) 3 in the same reaction [ 89
]. Depending on reaction conditions, the ruthenium system realizes
yields of 2,5-hexanediol (69 % at 175 °C) or
2,5-dimethyl-tetrahydrofuran (80 % at 200–225 °C) and also
generates small amounts of 2-hexanone and hexane, while the iridium
system is slightly less active and also has a lower temperature
stability. Both cata- lysts work best in water as the reaction
medium and are inhibited by the addition of acid (e.g., HOTf) as
the cocatalyst or the use of 1,4-dioxane or sulfolane as solvents.
They are deactivated by formation of the substitutionally inert
bis-chelate complexes [M(4′-Ph-terpy) 2 ] n+ (M = Ru, Ir; n = 2, 3)
and a deposition of a metal coating in the reactor body, which for
iridium is acting as a heterogeneous catalyst emphasizing the need
for carefully executed control reactions in order to establish true
homogeneous catalyst activity under the fairly harsh reaction
conditions (T > > 150 °C, acidic medium) typically required
for hydrodeoxygenation reactions.
An effective rapid total deoxygenation of various C6 sugars to
hexane and its isomers and in some cases hexene(s) was achieved by
using either the iridium phosphine pincer complex
[(POCOP)Ir(H)(acetone)][B(C 6 F 5 ) 4 ]; POCOP = 2,6-[OP( t Bu) 2 ]
2 C 6 H 3 or – in a completely metal-free system – B(C 6 F 5 ) 3 as
the catalyst and excess diethylsilane (Et 2 SiH 2 ) as the
reductant [ 90 – 92 ]. This in
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
32
principle promising result is however completely negated by the
high cost and extremely low atom effi ciency of both the use and
synthesis of the reducing agent, which must be produced via the
corresponding silyl chloride obtained through the Müller–Rochow
process followed by metathesis with metal hydride or Grignard
reagents [ 93 , 94 ]. While the authors also suggest that
polymethylhydrosiloxane, a by-product of silicone manufacture, is
also effective as a reductant, the overall pro- duction volumes of
silicones and hence its by-products are miniscule compared to the
actual needs for alkanes either as solvents, let alone fuels. An
actual large-scale use of the process would therefore be
prohibitive from both an economic and eco- logic perspective. The
proposed reaction sequence does however have enormous potential and
merit for the synthesis of partially deoxygenated silylated
polyalco- hols, e.g., 1,2-deoxyglucitol obtainable when
stoichiometric amounts of sterically more demanding silanes (e.g.,
Me 2 EtSiH) are used. Such deoxygenated sugars are otherwise very
diffi cult to access and very valuable synthons for
enantioselective synthesis with a well-defi ned stereochemistry of
the starting material. By the use of different, naturally abundant
hexoses, i.e., mannose, galactose, etc., different pre- cursors
with four stereocenters should be accessible by this method.
2.3.2 Homogeneous Catalysts for Hydrodeoxygenation Upgrading of
Pyrolysis Bio-Oil
Pyrolysis bio-oil, obtained from the fast pyrolysis of
lignocellulosic biomass with exclusion of oxygen, is characterized
by the presence of a multitude of reactive carbonyl compounds,
i.e., aldehydes, ketones, and hydroxyl aldehydes and hydroxy
ketones, anhydrosugars, and phenolics (originating from the lignin
content of the biomass) up to 10 % w/w of formic and acetic acid as
well as a high and variable (15–30 % w/w) water content. The oil is
therefore self-reactive, i.e., unstable against condensation and
polymerization reactions. These properties make a typical bio-oil
unusable as a fuel “as is,” requiring a reductive upgrading process
that lowers the overall content of reactive oxygen functionalities
and increases the energy density of the oil. The identifi cation of
any catalyst capable of effi ciently hydrodeoxygenat- ing bio-oil
is challenging and the subject of current intense research mainly
employ- ing heterogeneous systems [ 95 – 97 ]. The use of
homogeneous systems for this purpose may arguably be even more
diffi cult, in particular with a view to catalyst inhibition by the
(acidic) substrates and catalyst recycling from an anticipated very
complex reaction and product mixture. To date only two studies that
attempt this approach have appeared in the literature. Heeres et
al. reported the use of the well- established hydrogenation
catalyst RuCl 2 (PPh 3 ) 2 [ 98 , 99 ] for the hydrogenation of the
ketone and aldehyde model compounds acetol and
hydroxyl-acetaldehyde (both typically major components of bio-oil)
and the aqueous extract of a bio-oil in a biphasic mixture with
toluene as the organic solvent. The catalyst achieved a selec- tive
conversion to the corresponding diols in ~ 60 % yield after < 30
min. at 60–90 °C
M. Schlaf
33
under 4 MPa H 2 and tolerated the presence of acetic acid with only
a small decrease in activity [ 100 ]. The catalyst hydrogenates the
aldehyde substrate faster that the ketone and the authors suggested
that catalyst recycling from the organic phase of the reaction
mixture may be possible with further process optimization.
The already mentioned Shvo catalyst (Sects. 2.3.1.5 and 2.3.1.6 )
was also applied to a set of model compounds consisting of lignin
models such as vanillin, cinnam- aldehyde, and methylacetophenone
as well as the acetol and hydroxyl-acetaldehyde also targeted by
Heeres [ 101 ]. In toluene as the reaction medium at 90–145 °C
under 1 MPa H 2 , the catalyst achieved high conversions of these
substrates to cor- responding alcohols and tolerated the presence
of acetic acid as was also observed with RuCl 2 (PPh 3 ) 2 . When
applied to an actual bio-oil (derived from white poplar), the
catalyst was completely soluble in the oil and effected an almost
complete con- version of the aldehyde components (by NMR analysis)
indicating a substantial stabilization of the oil. However,
catalyst recovery or reuse was not possible, which as previously
stated is one of the major challenges of applying homogeneous sys-
tems to bio-oils and may in fact prevent their larger-scale use for
this purpose.
2.4 Conclusion and Outlook
From the literature survey presented here, it is apparent that
compared to the multi- tude of studies on heterogeneous catalyst
systems, the use of homogeneous systems for biomass
hydrodeoxygenation remains relatively unexplored, which may in part
be rooted in the fact that the research fi eld is dominated by
chemical engineers that often do not venture into the synthetic
procedures typically required to make homo- geneous catalysts
available for study. However, from the author’s viewpoint, this
represents an opportunity rather than a problem, as collaborations
between syn- thetic (organometallic) chemists well-versed in ligand
and catalyst design and syn- thesis and chemical engineers as
experts in process design and optimization may indeed turn out to
be very fruitful to the fi eld. In particular, the perennial
problem of recycling homogeneous catalysts should benefi t from
such collaborations, while the chemist’s role will lie in
synthesizing catalysts that maintain their activity under the
high-temperature aqueous acidic conditions necessarily required for
the reaction cascades laid out in the Figs. 2.5 , 2.10 , 2.11 ,
2.12 , 2.14 , and 2.17 given above. Since the overall goal of the
hydrodeoxygenation reactions is the defunctionalization of the
substrates to products that are thermodynamically stable and
unreactive under the reducing conditions (hydrogen atmosphere), the
main role of the ligand frame- work around metal center will in
these reaction not be the induction of chemo- or stereoselectivity
typically sought with homogeneous catalyst systems, but the maxi-
mization of deoxygenation activity and the stabilization of the
metal center against reduction to bulk metal leading to the in situ
generation of heterogeneous catalysts defeating the purpose and
potential advantages of maintaining a homogeneous sys- tem. A
logical approach to achieve this goal is the design and use of
highly chelating ligands that will result in very high complex
formation constants due to the
2 Homogeneous Catalysts for the Hydrodeoxygenation of
Biomass-Derived…
34
macrocyclic effect while maintaining at least one free or labile
coordination site for hydrogen and/or substrate activation [ 17 ].
The successful use of homogeneous cata- lysts for the
hydrodeoxygenation of biomass-derived substrates to value-added
chemicals and possibly alkane fuels will be contingent on meeting
these criteria.
References
1. Schlaf M (2006) Selective deoxygenation of sugar polyols to
alpha,omega-diols and other oxygen content reduced materials – a
new challenge to homogeneous ionic hydrogenation and hydrogenolysis
catalysis. J Chem Soc Dalton Trans 39:4645–4653
2. Schiweck H et al (2000), Sugar alcohols. In: Ullmann’s
encyclopedia of industrial chemistry. Wiley-VCH Verlag GmbH &
Co. KGaA
3. Sabatier P, Mailhe A (1909) New applications of the general
method of hydrogenation on divided metals. Annales De Chimie Et De
Phys 16:70–107
4. Schuette HA, Thomas RW (1930) Normal valerolactone. III. Its
preparation by the catalytic reduction of levulinic acid with
hydrogen in the presence of platinum oxide. J Am Chem Soc
52:3010–3012
5. Schiavo V, Descotes G, Mentech J (1991) Catalytic-hydrogenation
of 5- hydroxymethylfurfural in aqueous-medium. Bull Soc Chim Fr
128(5):704–711
6. Huber GW, Iborra S, C