Dynamic Article LinksC<Analyst

Cite this: Analyst, 2012, 137, 1639

www.rsc.org/analyst PAPER

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online / Journal Homepage / Table of Contents for this issue

Inside/outside Pt nanoparticles decoration of functionalised carbon nanofibers(Pt19.2/f-CNF80.8) for sensitive non-enzymatic electrochemical glucosedetection

Baljit Singh,a Eithne Dempsey,*a Calum Dickinsonb and Fathima Laffirb

Received 22nd November 2011, Accepted 7th February 2012

DOI: 10.1039/c2an16146j

A highly efficient and reproducible approach for effective Pt nanoparticles dispersion and excellent

decoration (inside/outside) of functionalised carbon nanofibers (f-CNF) is presented. The surface

morphological, compositional and structural characterisations of the synthesised Pt19.2/f-CNF80.8

material were examined using transmission electron microscopy (TEM/STEM/DF-STEM), energy-

dispersive X-ray spectrometry (EDS), thermogravimetric analysis (TGA/DTG), X-ray diffraction

(XRD) and X-ray photoelectron spectroscopy (XPS). Cyclic voltammetry (CV) was employed in order

to confirm the typical electrochemical response for Pt. The aim of the work was to improve the utility of

both the supporting matrix (via the use of both inner/outer surfaces of nanofibers) and precious Pt,

together with the sensitive glucose determination. TEM data indicated successful nanoparticle

decoration with average Pt particle size 2.4 nm. The studies demonstrated that utilisation of the inner

surface of the nanofibers, together with the modified outer surface characteristics using chemical

treatment, enables excellent decoration, effective dispersion and efficient impregnation of Pt

nanoparticles on carbon nanofibers. Pt19.2/f-CNF80.8 exhibited excellent amperometric response

(sensitivity¼ 22.7 mAmM�1cm�2 and LoD¼ 0.42 mM) towards direct glucose sensing, over the range 0–

10 mM glucose, in neutral conditions (pH 7.4). The improved carbon surface area for nanoparticle

decoration, inner surface structure and morphology of nanofibers together with the presence of

functional groups provided strong interactions and stability. These features together with the effective

nanoparticle dispersion and decoration resulted in excellent catalytic response. The decorated

nanoscaled material (Pt19.2/f-CNF80.8) is capable of large scale production, providing sensing capability

in neutral conditions, while eliminating the temperature sensitivity, pH and lifetime issues associated

with glucose enzymatic sensors and holds great promise in the quantification of glucose in real clinical

samples.

Introduction

Nanocarbon materials have been widely studied as supports

because of their unique morphology and reactivity.1–3 Carbon

materials are found in a variety of forms such as carbon blacks,

graphite, diamond, fullerenes, carbon nanofibers (CNFs) and

carbon nanotubes (CNTs). Carbonblack (VulcanXC-72), ismost

widely used as catalyst support (fuel cell applications) because of

its good compromise between electronic conductivity and surface

area,4,5 though CNTs were shown to have higher activities when

used as electrocatalyst support relative to traditional carbon

materials.6,7 Nevertheless, CNTs have relatively small specific

surface area and weak interactions with the supported metals,

aCentre for Research in Electroanalytical Technologies (CREATE),Institute of Technology Tallaght, Tallaght, Dublin 24, IrelandbMaterials Surface Science Institute, University of Limerick, Co. Limerick,Ireland. E-mail: Singh.Baljit765@gmail.com; Eithne.Dempsey@ittdublin.ie

This journal is ª The Royal Society of Chemistry 2012

which restrict further improvement in catalytic activity. These

shortcomings may be overcome by using carbon nanofibers as

they have larger surface areas andmore exposed edge planes along

the surface, which in turn can constitute potential sites for

advantageous chemical or physical interactions.

CNFs possess electronic and mechanical properties similar to

CNTs and are cylindrical nanostructures with graphene layers

arranged as stacked cones, cups, or plates.8 Besides the physico-

chemical properties which are in common to ordinary carbon

materials, CNF exhibit specific structures, high surface areas and

easily tunable mesoporous properties. CNFs exhibit strength to

weight ratio even greater than steel. In addition, vapour grown

CNFs have a lower production cost (100 times lower than single

wall CNTs)9 and are suitable for mass production. The separa-

tion of the catalyst from the reaction mixture has also been

reported to be less cumbersome for CNFs.10 Moreover, the

surface active group to volume ratio for CNFs is much larger

than CNTs.

Analyst, 2012, 137, 1639–1648 | 1639

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

The carbon nanofibers have been intensively investigated and

improved performance has been obtained relative to alternatives.

Chaniotakis’s et al., explored the direct immobilisation of

glucose oxidase onto the nanofibers and established that CNFs

were the best matrix for the development of biosensors, far

superior to CNTs or graphite powder.11 They used three different

types of carbon nanofibers (Electrovac AG, grades: LHT, HTE

and GFE (graphitized fibers)) to make glucose biosensors,11 and

compared them with single-walled carbon nanotubes (CNs,

Aldrich) and graphite powder (Fluka). The amount (units/g) of

the enzyme adsorbed in carbon nanomaterials was found to be 25

(graphite), 19.5 (GFE), 16 (LHT), 14 (HTE) and 5.5 (CNs). The

biosensors were monitored under continuous polarisation for

100 h at 25 �C (+600 mV vs.Ag/AgCl). GFE based biosensor did

not lose any of its initial activity even after 100 h of continuous

operation while LHT, HTE, graphite and carbon nanotube

electrodes presented a remaining activity of 97%, 83%, 81% and

73% respectively. The reproducibility of GFE-based biosensor

was also found to be excellent (RSD < 1%, N ¼ 3). CNFs were

found to be very promising materials for the development of

stable and reproducible amperometric biosensors. Their high

conductivity is ideal for the electrochemical signal transduction,

while at the same time reported to provide a very good surface

chemical environment for enzyme immobilisation. The larger

functionalised surface area of nanofibers compared to nanotubes

induces the adsorption of higher amounts of enzyme on the fiber

structures and the enhanced stability of enzyme molecules on the

GFE surface was attributed to the graphitization treatment to

this type of nanofibers.

Lu et al., have used CNFs based composites as an immobili-

sation matrix for the construction of a reagentless mediator-free

Haemoglobin (Hb) based H2O2 biosensor and results revealed

that Hb retained its essential secondary structure in the CNF-

based composite film.12 The great interest in nanofibers in certain

industry segments is due to the limited performance or much

higher unit prices of alternative materials,13 exemplified by the

utilisation of CNFs as an alternative to CNTs for electron

emitters in flat panel displays.

It is well known that the basic properties of impregnated

catalysts are strongly affected by the impregnation method,

microstructure, surface reactivity and metal precursor.14,15

Synthesis of uniform sized well dispersed platinum nanoparticles

over supporting matrix is highly desirable for effective catalytic

properties and optimum utilisation of precious platinum.

Conventional methods for the synthesis of Pt and Pt based

catalysts include impregnation16 and colloidal approaches.17–19

Impregnation method is limited, as the average particle size is

usually large and the size distribution is broad. Colloidal

methods produce well-homogenized ultrafine Pt electrocatalysts,

but the complexity of synthesis hinders its utilisation.20 A great

deal of effort has been devoted to improve the performance and

utilisation of Pt and surface geometry of carbon support mate-

rials is crucial in this regard. Also a catalyst with higher carbon to

Pt ratio leads to a thick layer of catalyst and concomitantly

higher mass transfer and electric resistance. A strategic approach

to obviate the issues related to mass transfer and conductivity is

to develop catalysts possessing high Pt loading and dispersion.

Tubular morphologies of some carbon nanomaterials with open

tips are interesting in context of developing low carbon loaded

1640 | Analyst, 2012, 137, 1639–1648

and metal-rich catalysts because the morphology offers the

possibility to access both inner and outer carbon surfaces for

effective metal dispersion and deposition. Metal dispersion along

the inner walls of such materials can be accomplished by over-

coming the geometry and surface affinity-related issues. Even

though there have been reports on nanoparticles decoration

inside the tubular morphologies of carbon by different strategies,

in most cases, nanoparticles exist in the form of nanowires, rods,

or particle aggregates by blocking the tubular channels of the

materials.21

Here we demonstrated a highly efficient strategy for the

effective dispersion and excellent decoration of Pt nanoparticles

(inside/outside) on carbon nanofibers, targeted to improve the

nanocatalyst utility and sensitivity for direct glucose determi-

nation. The applied synthetic procedure is simple, capable of

large scale production and controlling the particle size/distribu-

tion, without using any capping agent. In addition, the synthetic

method uses the mixed solution composition (ethylene glycol +

water), which can modify the surface tension and polarity

characteristics for effective inside decoration of nanofibers. The

inside decoration of nanofibers are easily accessible by the

applied synthetic conditions and may be attributed to the

improved viscosity of the solvent system and further improved by

aging time (discussed in experimental section). Acid pre treat-

ments to carbon nanofibers, helps in external/internal nano-

particles decoration, which further improves particle dispersion

and could be very crucial in effective utilisation of precious Pt

during application. The pH of reaction solution was maintained

at 10.5 and it appears that the presence of hydroxyl ions (KOH

addition during synthesis) may helps in the electrostatic stabili-

sation of nanoparticles resulting in excellent nanoparticles

dispersion and decoration.

We have demonstrated previously,22 that the PtAu/C based

catalysts exhibits sensitive and very selective amperometric

responses to glucose compared to common interfering species

(ascorbic acid, uric acid, acetaminophen and dopamine) with

negligible responses towards structurally related sugars over

their physiological levels at neutral pH (7.4). The optimum utility

of Pt together with the synergistic role by Au was proposed for

the excellent response observed. The objective here was to

enhance sensitivity towards glucose sensing by effective Pt phase

dispersion and decoration of nanofibers. An excellent response

(sensitivity ¼ 22.7 mAmM�1cm�2, 0–10 mM glucose) was ach-

ieved amperometrically for direct glucose determination.

Important features such as effective nanoparticle dispersion,

capability of large scale production, excellent sensitivity, low

applied potential, low limit of detection and good reproducibility

together with the ability to respond at neutral pH (7.4),

demonstrated superior performance relative to current

approaches in use. The improved dispersion also increases the

probability that individual Pt nanoparticles will take part during

application rather than an agglomerated group of particles. The

inner surface structure/morphology of nanofibers together with

the presence of functional groups, are also capable of improving

the stability of material during application, by providing

stronger interactions compared to outer side decoration alone. In

addition, the cost effective CNFs which constitutes �80% of the

nanocomposite, makes it more attractive for industrial scale

application. The synthesised nanoscale material is an attractive

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

alternative and of great promise in various applications (cata-

lytic, electrochemical/fuel cells etc.) due to the fact that for a fixed

catalyst loading the available carbon surface area is enhanced,

leading to better Pt dispersion and utilisation.

Experimental

Materials and reagents

Hexachloroplatinic acid (H2PtCl6$6H2O), ethylene glycol, silver

nitrate and potassium hydroxide were used as received from

Sigma Aldrich. Sulphuric acid, phosphate buffer saline (PBS,

0.01 M, pH 7.4), isopropyl alcohol, Nafion� (5% solution in

a mixture of lower aliphatic alcohols and water) were also

purchased from Sigma Aldrich and solutions were prepared

according to requirements. Carbon nanofibers (PR-19-XT-LHT)

were supplied by Pyrograf Products Inc. Ultra pure water puri-

fied with purelab option water equipment (resistivity 17MU$ cm)

was used in all experiments. The glucose stock solution was kept

overnight (at least 24 h) to allow mutarotation.

Functionalisation of supporting matrix (carbon nanofibers)

In view of utilising the outer as well as the inner walls for Pt

dispersion and decoration, carbon nanofibers were milled by the

company (Pyrograf Products Inc., processed by 1-pass (AM-1))

according to our requirement to make the ends of the fibers open

without shortening the fiber length significantly. These milled

nanofibers were used for functionalisation followed by nano-

particle decoration.

Carbon nanofibers (CNFs) were sonicated for 2 h in concen-

trated H2SO4/HNO3 acid solution (3 : 1, 125 ml acid solution

diluted to 250 ml with distilled water) and kept overnight,

diluted, filtered and washed with distilled water in order to

remove the residual catalyst and carbon impurities. The sonica-

tion was repeated again but in concentrated acid mixture for 30

min followed by dilution of acid mixture (3 times, v/v) with

distilled water and again sonicated (1.5 h) for effective func-

tionalisation of the nanofibers. The solution was kept overnight,

diluted, filtered and washed with enough distilled water until

reaching a neutral pH and dried under vacuum at 90 �C for 13 h.

The so obtained functionalised carbon nanofibers are labelled as

f-CNF.

Material synthesis

In order to facilitate the Pt precursor solution all along the inside

regions of the carbon nanofibers (together with the homogeneous

wetting of the nanofibers) and to achieve the effective and

uniform nanoparticle decoration, simultaneous metal ion

dispersion and reduction was applied using the ethylene glycol/

water (3 : 2) solvent system. The aging time (for nanofibers and

Pt precursor solution) and improved viscosity of solvent system

used (ethylene glycol/water, 3 : 2) compared to ethylene glycol

alone, are the important steps in improving the wetting proper-

ties (especially for effective inside decoration), surface tension,

effective metallic dispersion and polarity characteristics. These

parameters are critical particularly to facilitate precursor salt

solution into inner surface regions and hence to achieve excellent

nanoparticles decoration. Further, chemical functionalisation by

This journal is ª The Royal Society of Chemistry 2012

acid treatment (HNO3/H2SO4), is capable of improving/modi-

fying the outer surface of nanofibers for Pt nanoparticles

dispersion and decoration. The proper utilisation of the nano-

fibers surface (both inside/outside) together with the modified

surface characteristics by chemical pre-treatment make it

possible to achieve excellent Pt nanoparticles dispersion and

decoration.

In a typical synthesis (Pt19.2/f-CNF80.8), 80 mg of functional-

ised carbon nanofibers (f-CNF) were well dispersed in the solvent

mixture (ethylene glycol and water, 3 : 2, 100 ml) using sonica-

tion (70 min) followed by stirring and dropwise addition of well

stirred Pt precursor (H2PtCl6$6H2O, 53.1 mg) solution in the

same solvent system (ethylene glycol/water, 3 : 2, 200 ml) at the

maintained rate, over a period of 7 h with continuous stirring.

Subsequently, the mixture was sonicated (15 min) and then kept

for overnight with continuous stirring at ambient temperature to

achieve effective dispersion of the Pt precursor solution all over

the nanofibers. This aging time is important as the precursor

solution needs to reach inside (inner surface regions) the nano-

fibers. The pH of the solution mixture was maintained at 10.5, by

the addition of KOH solution prepared using same solvent

mixture (ethylene glycol/water, 3 : 2). The solution was sonicated

and well stirred before the mixture was refluxed at 135 �C for 8 h

to ensure complete reduction of Pt ions. The reaction mixture

was then cooled with continuous stirring and filtered through

filter paper (Millipore, 0.45 mm) followed by repeated washing

with deionised water and subsequently with small aliquots of

acetone and isopropyl alcohol to remove excess solvent. Finally

enough washing with deionised water was performed and dried

under vacuum at 70 �C for 12 h. The synthesised catalyst was

labelled as Pt19.2/f-CNF80.8, according with the Pt loading

(19.2%, TGA results) on carbon nanofibers.

Material characterisations

The surface morphology and distribution of the synthesised

nanocomposite was characterised using transmission electron

microscopy (TEM) with a JEOL 2011 operated at 200 kV using

a LaB6 filament equipped with a Gatan Multiscan Camera 794

and the dark field scanning TEM (DF-STEM) with a JEOL

2100F operated at 200 kV using a field emission electron source

equipped with a Gatan Ultrascan Camera. For TEM measure-

ments, sample was suspended in isopropyl alcohol, sonicated and

drop casted onto a Cu grid and dried overnight at 30 �C. Theidentity of each element in synthesised material was confirmed by

EDX. Thermogravimetric analysis was performed using

a Thermal Advantage Q50 with a platinum pan, balance N2 flow

40 ml min�1 and sample N2 flow 60 ml min�1. The X-ray

diffraction (XRD) pattern of the catalyst was obtained using

a Rigaku D/MAX-PC 2500 X-ray diffractometer with a Cu-Ka

(q ¼ 1.54 �A) radiation source operating at 40 kV and 200 mA.

The measurements were carried over the scan range (5–90�) withcontinuous scan type. X-ray photoelectron spectroscopy (XPS)

was performed in a Kratos AXIS 165 spectrometer, using

monochromatic Al Ka radiations of energy 1486.6 eV. High

resolution spectra were taken at fixed pass energy of 20 eV. The

surface charge was efficiently neutralised by flooding the sample

surface with low energy electrons. C 1s peak at 284.8 eV was used

as a charge reference to determine core level binding energies.

Analyst, 2012, 137, 1639–1648 | 1641

Fig. 1 Transmission electron micrographs (TEM) showing Pt nano-

particles decoration (images 1–6) of carbon nanofibers, along with view

for the open ends (image 5) and one end closed (image 6) of nanofibers for

Pt19.2/f-CNF80.8.

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

For peak analysis of high resolution spectra, a mixed Gaussian-

Lorenzian function with a Shirley type background subtraction

were used. Relative sensitivity factors used are from CasaXPS

library containing Scofield cross-sections. The sample powder

was dusted onto double sided adhesive tape for measurements.

The electrochemical experiments were performed using an

electrochemical work station CH Instruments Inc. 900, in acidic

(H2SO4) electrolyte, deaerated with high-purity Argon prior to

measurement, using a conventional three-electrode cell at room

temperature. A glassy carbon electrode (geometric area, 0.0707

cm2) modified with synthesised catalyst (Pt19.2/f-CNF80.8) served

as the working electrode, while platinum wire and a standard Ag/

AgCl electrode were used as the counter and reference electrodes

respectively. Prior to electrochemical measurements, a glassy

carbon electrode was polished with 1.0, 0.3, 0.05 micron size

alumina powders, sonicated in acetone and distilled water,

washed with deionised water and dried using argon at room

temperature. A typical suspension of the synthesised catalyst was

prepared by suspending catalyst in water isopropyl alcohol

mixture (3 : 1). The calculated amount of Nafion solution (1%

solution, diluted from 5% Nafion� solution) was added to the

suspension and maintained the catalyst to Nafion ratio 8 : 1. The

mixture was then sonicated for 20 min in order to achieve

a uniform dispersion of catalyst. The suspension (9.6 mg Pt) was

then transferred to the surface of the polished glassy carbon

electrode using a micropipette to form thin film on the surface of

the electrode and was then dried in air at room temperature.

Amperometric measurements were carried out in phosphate

buffer saline solution (0.01 M, pH ¼ 7.4) at Eapp ¼ +0.38 V and

+0.51 V vs. Ag/AgCl for direct glucose electro-oxidation studies.

Results and discussion

Surface morphological, compositional and structural

characteristaions

TEM studies provide clear information regarding the dispersion,

decoration and size of the Pt nanoparticles. The structure and

morphology of the catalyst support (CNF) as well as of the

synthesised catalyst (Pt19.2/f-CNF80.8) were explored using HR-

TEM/STEM/DF-STEM analysis. TEM also provided additional

information regarding the location of the particles, decorated

inside/outside the carbon nanofibers. The Pt nanoparticles

appeared to be well dispersed and embedded within the walls and

outside of the carbon nanofibers (Fig.1, images 1–6). It is also

clear that carbon nanofibers used as supporting matrix have

a large central hollow core and open tips/ends as shown in Fig. 1.

The ends of the nanofibers were opened by a milling process

(milled, without shortening the fiber length significantly) which is

crucial, together with the synthetic process conditions, for

effective dispersion of precursor solution inside the nanofibers.

The open ends of the nanofibers are clearly confirmed from

TEM analysis as can be seen in Fig. 1 (image 5). It can be

concluded that carbon nanofibers exhibit a large central hollow

core and opened ends, which together imposes a significant

portion of exposed and reactive edges in the inner channel

created within the carbon nanofibers. The opened ends by the

milling procedure provide the fibers, the inner channels decora-

tion which improves the nanoparticle dispersion and utilisation.

1642 | Analyst, 2012, 137, 1639–1648

There is also some possibility that one end of nanofibers is

opened and other remains closed (as shown in image 6, Fig. 1),

resulting in nanoparticle decoration from opened to closed end,

i.e. as far as the precursor solution can travel. This clearly

provides an idea for inside/outside nanoparticles decoration as

there is only outside decoration achieved without any inside

nanoparticle decoration (image 6, Fig. 1) in the closed end (area)

of nanofiber.

From high resolution TEM studies, catalyst nanoparticles are

observed as being well separated from each other, maintaining

the monodispersity of particles, indicating that the synthetic

method potentially prevents the agglomeration of Pt nano-

particles. The observed uniformity in size of nanoparticles and

the achieved distribution/decoration, could be due to the

following reasons: (1) the synthetic process conditions (particu-

larly aging time, solvent system and viscosity parameters) makes

for the diffusion and wetting by the precursor metal salt solution

both inside and outside of the nanofibers; (2) the chemical

functionalisation leads to the formation of groups such as OH,

CO and COOH on the surfaces of nanofibers (as evident from

XPS analysis), which are known to act as anchoring sites for

nanoparticles, improved the nanoparticle dispersion and

decoration.

Upon functionalisation, the inactive outer surface of nano-

fibers has also become active, so nanofiber active surface area

available for holding the metal species is almost doubled; (3) the

pH was maintained at 10.5 during the synthetic conditions by

addition of KOH (described in experimental section), which can

improve the dispersion and decoration of nanoparticles and can

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

be attributed to the electrostatic stabilisation of nanoparticles by

the adsorbed hydroxyl ions. It is well known that the pH of

reaction solution affects the size of nanoparticles. All three

reasons can contribute to the nanoparticle dispersion and deco-

ration seen within the sample (Pt19.2/f-CNF80.8).

The Pt19.2/f-CNF80.8 catalyst was also resynthesised under

similar conditions, described in the experimental section, and

showed excellent nanoparticle dispersion and decoration,

demonstrating the reproducibility of the synthetic method

together with the capability of large scale production. The Pt

nanoparticles were crystalline in a face-centered cubic (fcc)

structure, space group Fm3m and cell parameter 3.92 �A, which is

in accordance with the XRD studies. Whilst TEM confirmed the

location of the particles on the nanofibers, as in Fig. 1, dark field

scanning TEM (DF-STEM), as seen in Fig. 2, allows for easier

observation of the Pt particles. As using DF-STEM, contrast

comes from the atomic number (z-contrast), the Pt signal can

easily be observed from that of the carbon nanofibers. Using the

secondary electron (SE) detector on the TEM, the same location

can have both a SE image (image 1, Fig. 2) and DF-STEM image

(images 2 and 4–6, Fig. 2) taken at the same time. Using the

digital micrograph simple math, these images can be added

together (image 3, Fig. 2), allowing for the location of the

particles to be easier seen being inside rather than outside the

nanofibers. If the particles are more common on the inside, as

opposed to the outside, a greater concentration of particles can

be seen inside the tube of nanofibers. Scheme 1 shows the three

situations that can occur with (a) more particles on the inside

Fig. 2 Comparison of (1) secondary electron (SE) image with (2) DF-

STEM image for Pt19.2/f-CNF80.8 and image (3) is the combined math-

ematical image, resulting from the digital micrograph simple mathe-

matical addition of the two images (1 + 2). (4–6) DF-STEM images

allowing for easier observation of Pt nanoparticles decoration (inside/

outside) of the carbon nanofibers.

This journal is ª The Royal Society of Chemistry 2012

than the outside, (b) more on the outside than inside and (c)

equal particles on inside and outside.

Fig. 3a shows particle size distribution of Pt nanoparticles

decorated on carbon nanofibers obtained from TEM analysis for

Pt19.2/f-CNF80.8 nanocomposite. Particle size distribution was

created by measuring 500 particles from randomly chosen areas

(nanofibers) and the average particle size was 2.4 nm with median

(2.42 nm), maximum (4.55 nm), minimum (0.57 nm) and mode

(2.60 nm) respectively. Fig. 3b shows the EDS profile for Pt19.2/

f-CNF80.8 nanocomposite and confirms the identity and presence

of Pt in the synthesised Pt19.2/f-CNF80.8.

In conclusion from all the TEM analysis we can say that the

synthetic procedure (aging time, solvent system and polarity

characteristics), opened ends of nanofibers, active inner wall

morphology (improved due to milling/opened ends) and the

activated outer wall by chemical treatment (acid functionalisa-

tion), make it possible to achieve effective nanoparticle disper-

sion, optimum utilisation (of both nanofibers and precious Pt

content) and excellent decoration of carbon nanofibers. The

monodispersity of nanoparticles, the presence of functional

groups and active inner surface morphology of nanofibers makes

these materials capable of improved stability due to stronger

interactions and therefore can avoid dissolution of precious Pt

which is a common and critical problem during applications. The

synthesised catalyst (Pt19.2/f-CNF80.8) could be of immense

potential and scope in catalytic/electrochemical applications

because for a fixed metallic loading in a catalyst, the available

surface area is enhanced, leading to better dispersion and low

electrode film thickness due to the effective utilisation of both,

carbon nanofiber surface and precious Pt. This obviates the

issues related to mass transfer and conductivity as for a fixed

metallic loading, less material will be required resulting into

lower film thickness during application and so enhanced

conductivity and accessibility. Such materials could also be

excellent alternatives in applications where heavy metal loading

is desired, as the supporting matrix can be decorated from both

sides, so improved loading and dispersion can be achieved.

It can be concluded that the inside surface of nanofibers is

easily accessible using our synthetic process and may also be

attributed to the improved viscosity of the solvent system

Scheme 1 Diagrams showing (top) end-on and (bottom) side view of Pt

nanoparticles (black spheres) decorated carbon nanofibers. The three

situations shown are (a) more particles on inside than outside, (b) more

particles on outside than inside and (c) no preference for particles on

inside or outside.

Analyst, 2012, 137, 1639–1648 | 1643

Fig. 3 (a) Size distribution for Pt nanoparticles decorated on carbon

nanofibers, obtained from TEM analysis of Pt19.2/f-CNF80.8. (b) EDX

profile for Pt19.2/f-CNF80.8.

Fig. 4 Thermogravimetric measurements (a, TGA profile) for Pt19.2/

f-CNF80.8, over the temperature range 20–800 �C at temperature ramp,

20 �C min�1. XRD pattern (b) for Pt nanoparticles decorated carbon

nanofibers (Pt19.2/f-CNF80.8), indexed to platinum with space group

Fm3m and lattice parameter 3.92 �A. The peak marked as * is the (001) of

graphite with d-spacing 3.3 �A.

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

(ethylene glycol + water). The ability of the solvent (ethylene

glycol/water) to pass through the nanofibers by capillary action

should be better compared to ethylene glycol alone, which is

further improved by aging time and continuous stirring during

synthesis, leading to effective utilisation of the inner surface of

carbon nanofibers for Pt dispersion and decoration. Because

of the morphology differences of the inner and outer surface of

nanofibers, the synthetic process appears to be excellent for

proper wetting (by Pt precursor salt solution) of nanofibers

surface and for simultaneous metal ion dispersion and effective

reduction. The surface tension of the solvent medium and the

contact angle between the liquid and nanofiber surface may play

a decisive role in achieving fine particles dispersion all along the

inner walls by significantly eliminating the possibilities of

blocking by any entrapped particles or aggregates. The aging

time seems to be crucial for the effective decoration and disper-

sion of nanoparticles all along the inner and outer surfaces of

carbon nanofibers. In addition, the presence of hydroxyl ions

(pH 10.5) helps in electrostatic stabilisation of nanoparticles,

resulting in improved particle distribution. This (pH) is also

capable of providing uniform and smaller sized nanoparticles,

which further add to the nanoparticle dispersion and decoration

achieved. The capillary action could possibly help to transport

the precursor salt solution into the walls (inner side morphology)

of the carbon nanofibers, which could further be improved by

aging time given to precursor salt solution and carbon nano-

fibers. Overall, studies have demonstrated that the proper uti-

lisation of the inner surface of the nanofibers (by synthetic

conditions) together with activation of the nanofiber surface

characteristics by chemical treatment (functionalisation), makes

it possible to achieve the excellent Pt nanoparticle dispersion and

decoration of carbon nanofibers.

1644 | Analyst, 2012, 137, 1639–1648

Fig. 4a shows the thermogravimetric measurements (TGA

profile) for Pt19.2/f-CNF80.8 nanocomposite, over the tempera-

ture range 20–800 �C at temperature ramp of 20 �C min�1,

demonstrating the oxidation behaviour (carbon degradation) of

the support material and confirm the total metal loading for the

synthesised catalyst Pt19.2/f-CNF80.8. The slight initial weight loss

up to 450 �C, is due to the loss of residual water (desorption of

water and other solvents used during synthesis) and functional

groups followed by slow and rapid weight losses attributed to

carbon degradation/oxidation.

From the TGA analysis (together with the DTG analysis), it is

observed that the material is thermally stable to about 450 �C,followed by slow carbon degradation (weight loss �22.4%, from

approx. 450–600 �C) and fast carbon degradation (weight loss

�49.6%, from approx. 600–650 �C). The measured weight loss

from around 450 to 650 �C is approximately 72%, occurred in the

form of decomposition of CNFs. This is in agreement with the

composition of material because the supporting matrix (CNFs)

constitutes a major component of Pt19.2/f-CNF80.8. Finally, from

the residue content, the amount (wt %) of Pt in Pt19.2/f-CNF80.8

catalyst was quantified to be 19.2%.

XRD pattern (Fig. 4b) for Pt19.2/f-CNF80.8, reveals a high

degree of crystallinity in the synthesised catalyst, exhibiting

strong diffraction peaks at 2q ¼ 26.28�, 40.1�, 46.6�, 67.75�,81.57� and 85.95�. The diffraction peak at 26.28� is attributed to

the carbon nanofibers, (001) of graphite, used as supporting

matrix and other peaks can be indexed as platinum (111), (200),

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

(220), (311) and (222) reflections respectively. The observed

pattern along with the peak positions are the characteristics of

the face-centered cubic (fcc) crystalline Pt in agreement with

TEM analysis, our previous results22,23 and related literature.24

In order to investigate the oxidation states and surface species

together with compositions, X-ray photoelectron spectroscopy

(XPS) analysis was performed. Fig. 5a shows the survey scan for

Pt19.2/f-CNF80.8, which consist of peaks (labelled) corresponding

to Pt 5p, Pt 4f, C 1s, Pt 4d, Pt 4p and O 1s, which is in very good

agreement with literature.25 The C 1s spectra (Fig. 5b) for Pt19.2/f-

CNF80.8 can be deconvoluted into carbon peaks with different

chemical environments. The observed peaks are, graphitic

carbon (284.8 eV), C–C (285.2 eV), C–O–C/C–OH (286.3 eV),

C]O (287.3 eV), COO (289.0 eV) and p–p* satellite (291.3 eV)

peaks. The broad O 1s peaks at 532.1 eV have contributions

representing C]O/C–O/COO species and minor contribution

from Pt oxides. The presence of functional groups on graphitic

carbon as observed from the C 1s spectra, support the proper

functionalisation of carbon nanofibers which is required for

anchoring and effective dispersion of Pt nanoparticles.

Fig. 5 XPS analysis showing (a) survey spectrum (b) C-1s XPS spectrum

and (c) Pt 4f regions, confirming the presence of different oxidation states

for Pt.

This journal is ª The Royal Society of Chemistry 2012

High resolution spectra of Pt 4f (Fig. 5c) appear as a doublet

peak resulting from 4f photoelectron transitions due to spin–

orbit coupling with an intensity ratio of 4f5/2: 4f7/2 of 3 : 4.26 The

asymmetry in the Pt 4f peak of Pt19.2/f-CNF80.8 is indicative of

the presence of more than one oxidation state for Pt. The peaks

were fitted with a doublet separation characteristic of Pt 4f and

full width half maximum (FWHM) is constrained for different

chemical states. Table 1 summarises the results of XPS studies.

The lowest binding energy peak at 71.6 eV (Table 1) may be

assigned to Pt in its zero valent state, although clean Pt appears

at a slightly lower binding energy,27 and the shift in binding

energy was considered to result from the size effect of Pt parti-

cles, interaction with carbon nanofiber support or the influence

of chemisorbed oxygen. The higher binding energy components

at 72.6 and 74.3 eV can be ascribed to oxidised Pt, namely

Pt(OH)2 and PtO/PtO2.23,27 Species corresponding to Pt in zero

valent state (Pt1) and oxidised form (Pt2 + Pt3) together with their

binding energy and relative intensity (%) values are shown in

Table 1. It can be observed that oxidised Pt (Pt2 + Pt3)

contributes approximately 34% to the total platinum intensity

and Pt19.2/f-CNF80.8 have �66% of Pt in the metallic/elemental

state.

Electrochemical and glucose determination studies

The synthesised nanostructured material was also characterised

by cyclic voltammetry to confirm the typical electrochemical

response for Pt. Fig. 6a shows the cyclic voltammetric response

for Pt19.2/f-CNF80.8 in 0.5 M H2SO4 at scan rate 0.1 V s�1 vs. Ag/

AgCl.

The voltammetric profile shows the characteristic hydrogen

adsorption/desorption waves along with Pt oxide formation and

reduction peaks (labelled) and confirms the presence of Pt in the

synthesised Pt19.2/f-CNF80.8 nanocomposite. Fig. 6b shows cyclic

voltammogram for Pt19.2/f-CNF80.8 modified glassy carbon

electrode in the absence and presence (labelled) of glucose in

phosphate buffer saline solution (0.01 M, pH 7.4) at scan rate

0.02 V s�1 vs. Ag/AgCl. The peak current increased over the

potential range of approx. 0.3 V to 0.5 V (in the double layer

region), following glucose additions. From cyclic voltammetric

measurements, the applied potentials of 0.38 V and 0.51 V (vs.

Ag/AgCl) were chosen for further amperometric glucose

determinations.

Researchers have extensively investigated and explored the

glucose behaviour at a platinum electrode in acidic,28 neutral29–31

and basic29 conditions. A common conclusion from related

research is that the sole product of glucose oxidation is glucono-

d-lactone, which hydrolyses to gluconic acid, regardless of the

solution pH (Scheme 2).29,32 However, spectrochemical evidence

regarding intermediate adsorbates has frequently disagreed,

suggesting reduced CO2, COads and fragments of the glucose

molecule to also be present as oxidation products.31,33 The final

and stable product of the two-electron (2e�/2H+ process)

oxidation of glucose is gluconic acid,34,35 whether glucono-d-

lactone is involved as an intermediate or not.

Fig. 7a shows the amperometric response for Pt19.2/f-CNF80.8

using hydrodynamic amperomety in phosphate buffer saline

solution (pH 7.4), over the glucose level (1–20 mM) at

Eapp ¼ 0.38 V vs. Ag/AgCl. The sensitivity (mAcm�2mM�1) value

Analyst, 2012, 137, 1639–1648 | 1645

Table 1 TEM, TGA and XPS results for Pt19.2/f-CNF80.8

Catalyst a<d>/nm bwt/(%) cB.E./eV Relative Intensity/(%) d(Ptox/Pt)/(%)

Pt1 71.6 66.1Pt19.2/f-CNF80.8 2.4 19.2 Pt2 72.6 23.8 34

Pt3 74.3 10.1

a Average particle size (<d>, nm) obtained from TEM analysis. b Weight (%) obtained from TGA analysis (experimental Pt loading is 20%, w/w).c Binding energy (eV). d Ptox/Pt is the (%) of oxidised Pt(Pt2 + Pt3) to total Pt, Pt species are labelled as Pt1(Pt

0), Pt2(Pt+2) and Pt3(Pt

+4).

Fig. 6 (a) Cyclic voltammetric response for Pt19.2/f-CNF80.8 in 0.5 M

H2SO4 at scan rate 0.1 V s�1 vs. Ag/AgCl. (b) Cyclic voltammetric

response for Pt19.2/f-CNF80.8 with and without glucose additions (labelled

as 0 mM, 10 mM and 20 mM) in phosphate buffer solution (PBS, pH 7.4)

at scan rate 0.02 V s�1 vs. Ag/AgCl.

Scheme 2 Glucose electro-oxidation to gluconolactone and further

hydrolysis to gluconic acid.

Fig. 7 (a) Amperometric response for Pt19.2/f-CNF80.8 using hydrody-

namic amperomety in phosphate buffer saline solution (pH 7.4), over the

glucose level (1–20 mM) at Eapp ¼ 0.38 V vs.Ag/AgCl. (b) Amperometric

response for Pt19.2/f-CNF80.8 using hydrodynamic amperomety in phos-

phate buffer solution (pH 7.4) over the glucose level (1–10 mM) at Eapp ¼0.51 V vs. Ag/AgCl.Pu

blis

hed

on 2

0 Fe

brua

ry 2

012.

Dow

nloa

ded

by U

nive

rsity

of

Chi

cago

on

25/1

0/20

14 0

9:45

:40.

View Article Online

was determined by plotting calibration curves and the observed

sensitivity (over 1–10 mM glucose level) for Pt19.2/f-CNF80.8 is

13.6 mAcm�2mM�1, at Eapp ¼ 0.38 V. From the amperometric

data it may be concluded that the modified electrode showed

a sensitive and rapid response towards glucose sensing in neutral

conditions (pH 7.4). The sensitivity response increased with the

applied potential (at Eapp ¼ 0.51 V, Fig. 7b) and the observed

sensitivity for Pt19.2/f-CNF80.8 is 22.7 mAcm�2mM�1 at

Eapp ¼ 0.51 V. The sensitivity response at Eapp ¼ 0.51 V

(compared to response at Eapp ¼ 0.38 V, under similar condi-

tions), demonstrating the importance of the applied potential in

providing sufficient surface active sites for continuous glucose

1646 | Analyst, 2012, 137, 1639–1648

electro-oxidation and in avoiding the accumulation of interme-

diate species on the electrocatalyst surface, resulted in improved

amperometric response.

The catalyst showed excellent response (sensitivity 22.7

mAcm�2mM�1 and lower limit of detection (LoD, S/N ¼ 3) 0.42

mM) towards glucose determination, over 1–10 mM glucose

concentration level at Eapp ¼ 0.51 V vs. Ag/AgCl. The limit of

detection (LoD ¼ 0.42 mM) is the best so far we have achieved in

our studies for glucose determination.22,36 Pt19.2/f-CNF80.8

exhibits reproducible behaviour (intra-electrode response)

towards glucose over the range 1–10 mM glucose in phosphate

buffer solution (pH 7.4), which further highlight its potential for

glucose determination in neutral conditions.

The synthesised material (Pt19.2/f-CNF80.8, Eapp ¼ 0.51 V)

resulted in excellent sensitivity response (22.7 mAcm�2mM�1)

which is much higher (43.65 times) compared to Pt/CNF (Pt

24.07%, TGA analysis) material (outside decoration alone, 0.52

mAcm�2mM�1, over 2–20 mM glucose level, Eapp ¼ 0.55 V)

previously achieved by our group. These results demonstrate the

importance of inside/outside decoration compared to outside

This journal is ª The Royal Society of Chemistry 2012

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

decoration alone. Therefore, the approach is very efficient for

optimum utilisation of precious Pt and for improving sensitivity

towards glucose. The increased performance by Pt19.2/f-CNF80.8

was possibly due to the improved carbon surface area (doubled

due to inside/outside surface availability compared to most

common outside surface decoration) available for holding the

metal species, which leads to optimum dispersion, excellent

decoration and effective utilisation of precious Pt at the nano-

fibers surface. Such decorated materials also provides a lower

catalyst loading on the electrode surface (electrode film thick-

ness) due to the effective utilisation of both, the carbon nano-

fibers surface and precious Pt content. This can obviates the

issues related to mass transfer and conductivity as for a fixed

metallic loading, less material will be required for electrode

fabrication resulting into lower film thickness and so enhanced

conductivity and accessibility. Pt19.2/f-CNF80.8 could also be

excellent alternatives in applications where heavy metal loading

is desired, as the supporting matrix can be decorated from both

sides, so improved loading and dispersion can be achieved.

The other features like monodispersity and small size (2.4 nm)

of Pt nanoparticles, effective functionalisation and inner surface

structure/morphology of carbon nanofibers, collectively can

provide stronger and better nanoparticle-nanofiber interactions

compared to outer side decoration alone. This may avoided the

dissolution of precious Pt and improved the stability of the

material during application, which further explains the observed

improvement in catalytic response. In addition the optimised

applied potential may provided sufficient surface active sites for

direct glucose electro-oxidation and avoided the accumulation of

intermediate species on the electrocatalyst surface, resulted in

excellent amperometric response.

Conclusions

We demonstrated an efficient strategy for fine dispersion and

excellent Pt nanoparticles decoration on the inner and outer

surface of carbon nanofibers. The synthetic conditions (solvent

system with improved viscosity, aging time, pH and polarity

characteristics), opened ends of nanofibers, inner surface

morphology and the activated (functionalised) nanofiber

surfaces by chemical treatment, collectively make it possible to

achieve effective dispersion (Pt nanoparticles, 2.4 nm), optimum

utilisation (of both precious Pt and carbon nanofibers) and

excellent nanoparticle decoration of carbon nanofibers. An

excellent response (sensitivity ¼ 22.7 mAmM�1cm�2 and LoD ¼0.42 mM at Eapp ¼ 0.51 V vs. Ag/AgCl) was achieved ampero-

metrically for direct glucose electro-oxidation over the 0–10 mM

glucose level in neutral conditions (pH 7.4).

The both side (inside/outside) decoration improves the effec-

tive surface area and Pt dispersion which increases the mono-

dispersity of nanoparticles and thereby improved the efficiency

and utilisation of both the carbon nanofibers and precious Pt.

The monodispersity and effective decoration of Pt nanoparticles

resulted in improved utilisation and are capable of providing

stability against poisoning species and during catalytic environ-

ment. The inner surface structure and morphology of nanofibers

in addition to the presence of functional groups could improve

stability by providing stronger interactions compared to outside

decoration alone and resulted in excellent response towards

This journal is ª The Royal Society of Chemistry 2012

glucose electro-oxidation. Another important characteristic is

the capability of such materials to obviate the issues related to

mass transfer and conductivity, as for a fixed metal loading in

a catalyst, both side decoration possesses the total metallic

content on both sides of nanofibers compared to external deco-

ration alone (where all the metal loading is only on the outer

surface of nanofibers). All these advantageous features of Pt19.2/

f-CNF80.8 are collectively responsible for the observed response

towards glucose electro-oxidation. The decorated nanoscale

material is capable of large scale production, providing sensing

capability in neutral conditions, while eliminating the tempera-

ture sensitivity, pH and lifetime issues associated with enzymatic

systems and demonstrate great promise in the quantification of

glucose in real clinical samples considering the high demand of

precious platinum.

The controlled decoration of the internal (e.g. Pt) and external

(e.g. another metal) nanofiber surfaces together with the uti-

lisation of an alloy/bimetallic nanoparticle-nanofiber system will

form part of future work in our group. Pt based bimetallic/alloy

systems are better alternative regarding selectivity from our

previous experience (bimetallic PtAu/C based nano-

composites22). PtAu/C based nanocomposites showed sensitive

and very selective amperometric responses to glucose compared

to common interfering species (ascorbic acid, uric acid, acet-

aminophen and dopamine) with negligible/no responses towards

structurally related sugars over their physiological levels in

neutral pH (7.4).22 The optimum utility of Pt together with the

synergistic role by Au was proposed for the excellent response

observed.22 Detailed selectivity studies using alloy/bimetallic (Pt

based) decorated nanofiber materials is a further goal within the

group.

Acknowledgements

Dr Baljit Singh and Prof Eithne Dempsey acknowledge Tech-

nological Sector Research Strand III, Postgraduate Research

and Development for funding. We also acknowledge DrWynette

Redington at theMaterials and Surface Science Institute (MSSI),

University of Limerick, Ireland for XRD analysis. Dr Calum

Dickinson and Dr Fathima Laffir acknowledge the INSPIRE

programme, funded by the Irish Government’s Programme for

Research in Third Level Institutions, Cycle 4, National Devel-

opment Plan 2007–2013.

References

1 J. M. Planeix, N. Coustel, B. Coq, V. Brotons, P. S. Kumbhar,R. Dutartre, P. Geneste, P. Bernier and P. M. Ajayan, J. Am.Chem. Soc., 1994, 116, 7935–7936.

2 J. Ma, H. Sun, F. Su, Y. Chen, Y. Tang, T. Lu and J. Zheng, Int. J.Hydrogen Energy, 2011, 36, 7265–7274.

3 Y. Chen, G. Zhang, J. Ma, Y. Zhou, Y. Tang and T. Lu, Int. J.Hydrogen Energy, 2010, 35, 10109–10117.

4 Q. Lu, B. Yang, L. Zhuang and J. Lu, J. Phys. Chem. B, 2005, 109,1715–1722.

5 Z. Liu, L. M. Gan, L. Hong,W. Chen and J. Y. Lee, J. Power Sources,2005, 139, 73–78.

6 Y. Lin, X. Cui and X. Ye, Electrochem. Commun., 2005, 7, 267–274.7 J. Wang, M. Musameh and Y. Lin, J. Am. Chem. Soc., 2003, 125,2408–2409.

8 S. George and H. K. Lee, J. Phys. Chem. B, 2009, 113, 15445–15454.9 I. Kang, Y. Y. Heung, J. H. Kim, J. W. Lee, R. Gollapudi,S. Subramaniam, S. Narasimhadevara, D. Hurd, G. R. Kirikera,

Analyst, 2012, 137, 1639–1648 | 1647

Publ

ishe

d on

20

Febr

uary

201

2. D

ownl

oade

d by

Uni

vers

ity o

f C

hica

go o

n 25

/10/

2014

09:

45:4

0.

View Article Online

V. Shanov, M. J. Schulz, D. Shi, J. Boerio, S. Mall and M. Ruggles-Wren, Composites, Part B, 2006, 37, 382–394.

10 P. Serp, M. Corrias and P. Kalck,Appl. Catal., A, 2003, 253, 337–358.11 V. Vamvakaki, K. Tsagaraki and N. Chaniotakis, Anal. Chem., 2006,

78, 5538–5542.12 X. Lu, J. Zhou, W. Lu, Q. Liu and J. Li, Biosens. Bioelectron., 2008,

23, 1236–1243.13 A. Nadarajah, J. G. Lawrence and T. W. Hughes, Key Eng. Mater.,

2008, 380, 193–206.14 L. R. Radovic, F. Rodriguez-Reinoso, CarbonMaterials in Catalysis.

In Chemistry and Physics of Carbon; Thrower, P. A., Ed.; MarcelDekker: New York, 1997, Vol. 25.

15 M. C. Rom�an-Mart�ınez, D. Cazorla-Amor�os, A. Linares-Solano,C. S.-M. n. De Lecea, H. Yamashita and M. Anpo, Carbon, 1995,33, 3–13.

16 A. S. Aric�o, Z. Poltarzewski, H. Kim, A. Morana, N. Giordano andV. Antonucci, J. Power Sources, 1995, 55, 159–166.

17 M. Watanabe, M. Uchida and S. Motoo, J. Electroanal. Chem., 1987,229, 395–406.

18 M. K. Ravikumar and A. K. Shukla, J. Electrochem. Soc., 1996, 143,2601–2606.

19 H. B€onnemann, W. Brijoux, R. Brinkmann, T. Joußen, B. Korall andE. Dinjus, Angew. Chem., Int. Ed. Engl., 1991, 30, 1312–1314.

20 A. S. Aric�o, S. Srinivasan and V. Antonucci, Fuel Cells, 2001, 1, 133–161.

21 X. Pan and X. Bao, Chem. Commun., 2008, 6271–6281.

1648 | Analyst, 2012, 137, 1639–1648

22 B. Singh, F. Laffir, T. McCormac and E. Dempsey, Sens. Actuators,B, 2010, 150, 80–92.

23 B. Singh, L. Murad, F. Laffir, C. Dickinson and E. Dempsey,Nanoscale, 2011, 3, 3334–3349.

24 G. Gao, G. Yang, M. Xu, C. Wang, C. Xu and H. Li, J. PowerSources, 2007, 173, 178–182.

25 B. Choi, H. Yoon, I.-S. Park, J. Jang and Y.-E. Sung, Carbon, 2007,45, 2496–2501.

26 D. Briggs, M. P. Seah, Practical Surface Analysis, 2nd ed., Wiley &Sons: Chichester, 1990.

27 NIST-XPS database, version 3.5 http://srdata.nist.gov/xps/.28 E. Skou, Electrochim. Acta, 1977, 22, 313–318.29 Y. B. Vassilyev, O. A. Khazova and N. N. Nikolaeva, J. Electroanal.

Chem., 1985, 196, 105–125.30 M. L. B. Rao and R. F. Drake, J. Electrochem. Soc., 1969, 116, 334–

337.31 A. E. Bolzan, T. Iwasita and W. Vielstich, J. Electrochem. Soc., 1987,

134, 3052–3058.32 Y. B. Vassilyev, O. A. Khazova and N. N. Nikolaeva, J. Electroanal.

Chem., 1985, 196, 127–144.33 B. Beden, F. Largeaud, K. B. Kokoh and C. Lamy, Electrochim. Acta,

1996, 41, 701–709.34 S. Park, H. Boo and T. D. Chung,Anal. Chim. Acta, 2006, 556, 46–57.35 S. Ernst, J. Heitbaum and C. H. Hamann, Ber. Bunsenges. Phys.

Chem., 1980, 84, 50–55.36 B. Singh and E. Dempsey, ECS Trans., 2011, 35, 75–91.

This journal is ª The Royal Society of Chemistry 2012