Probing The Thermal Deoxygenation of Graphene Oxide Using High Resolution in Situ X-Ray Based Spectroscopies
Probing The Thermal Deoxygenation of Graphene Oxide Using High Resolution in Situ X-Ray Based Spectroscopies
Probing The Thermal Deoxygenation of Graphene Oxide Using High Resolution in Situ X-Ray Based Spectroscopies
ABSTRACT. Despite the recent developments in Graphene Oxide due to its importance as a host
precursor of Graphene, the detailed electronic structure and its evolution during the thermal reduction
remain largely unknown, hindering its potential applications. We show that a combination of high
resolution in situ X-ray photoemission and X-ray absorption spectroscopies offer a powerful approach
to monitor the deoxygenation process and comprehensively evaluate the electronic structure of
Graphene Oxide thin films at different stages of the thermal reduction process. It is established that the
edge plane carboxyl groups are highly unstable, whereas carbonyl groups are more difficult to remove.
The results consistently support the formation of phenol groups through reaction of basal plane epoxide
groups with adjacent hydroxyl groups at moderate degrees of thermal activation (~400 C). The phenol
groups are predominant over carbonyl groups and survive even at a temperature of 1000 C. For the first
time a drastic increase in the density of states (DOS) near the Fermi level at 600 C is observed,
suggesting a progressive restoration of aromatic structure in the thermally reduced graphene oxide.
KEYWORDS. Graphene Oxide, Thermal Deoxygenation, XPS, NEXAFS, Valence band, Oxygenfunctional Groups and Atomic Structure, carbonyl, phenol.
INTRODUCTION
Since the discovery of Graphene,1,2,3 immense efforts have been focused on Graphene Oxide (GO),4,5
since it is the most promising precursor for obtaining large quantities of this unique material and
because GO is a useful material on its own right. Graphene Oxide can be visualized as individual sheets
of graphene decorated with oxygen functional groups on both basal planes and edges, which has been
prepared by oxidative exfoliation of graphite. The presence of oxygen makes GO amenable to chemical
functionalization, nevertheless it disrupts the extended sp2 network of the graphene hexagonal lattice.
To convert GO back to graphene, the chemical/thermal reduction of GO is so far the most attractive
procedure because of its simplicity, reliability, high yield and low cost.6-14 Chemical treatment,
especially exposure to hydrazine, is the most widely used route to reduce GO in solution.7,9-12 However
the chemically reduced GO (CRGO) suffers from a relatively low C/O atomic content,12 with a
considerable amount of residual O-moieties.7,12 Also, hydrazine-treatment leads to the formation of
nitrogen-functional groups,7,12 along with the inherent toxicity of hydrazine. Improvements have been
accomplished by either post-heating at low temperatures (200-500 C),8,10,11 or by replacing hydrazine
by alternative less toxic reducing agents, e.g. sodium borohydride or alkaline solutions.13,15 The use
nontoxic and biocompatible reducing agents such as vitamin C (L-ascorbic acid),16 green tea,17
melatonin,18 saccharides such as glucose, fructose and sucrose19 or the use of environmentally friendly
processes such as hydrothermal dehadration20 have also been reported. In another approach,21-23 the
reduction of GO was achieved by a TiO2 assisted photocatalytic method, in which the electrons
photogenerated by UVirradiated TiO2 were injected into GO reducing the oxygen-containing
functional groups. Such low temperature deoxygenating processes, although not completely effective in
reducing GO, are amenable to electronic applications of graphene patterned onto glass or plastic
substrates as well as to the synthesis of a wide range of functional hybrids with use in polymer
composites, biosensors, energy storage and conversion technologies.24 In comparison, thermal reduction
at high temperatures (9001000 C), particularly in ultra high vacuum (UHV), is found to be highly
efficient in producing graphene-like films with a significantly high C/O ratio,12 with no introduction of
any contaminants. However, such high-temperature processing is unlikely to be compatible with
fabrication techniques used for most electronic applications. Practically the majority of the studies
report the presence of various amounts of residual oxygen in reduced graphene oxide, with the electrical
conductivity reaching values several orders of magnitude lower than that of mechanically exfoliated
graphene.
To make further progress on optimizing and designing reduction processes, which is key to
numerous applications, GO needs to be well characterized and its thermal deoxygenation needs to be
well understood. Thermal reduction of GO has been shown to involve the removal of oxygen groups by
formation of carboneous species (CO2, CO) thus creating defects in the form of etch holes within the
graphene basal plane.10,11 There is however very little knowledge on how the oxygen containing
functional groups of graphene oxide evolve during thermal reduction.6,11,25 Theoretical and experimental
studies have shown that formation of thermodynamically stable carbonyl and ether groups through
transformation of the initial nearby hydroxyl and epoxy groups during thermal annealing, but little is
known on the survival of the persistent residual oxygen groups.
The specific objectives of the present work were i) to elucidate the evolution of oxygen groups and
probe their survival rate upon heat treatment; ii) to determine the nature of the residual oxygen
containing functional groups that remain after reduction; iii) to clarify how the heat treatment affects the
electronic structure of GO. To do that, we carried out high resolution in situ C 1s and O 1s core level
and Valence band X-ray photoemission as well as X-ray absorption temperature dependent
spectroscopic studies on GO. The analysis of these studies helped to develop a comprehensive view
into the temperature evolution of electronic structure and surface chemistry of GO nanosheets. We
found a predominance of phenol groups, which originate from the reaction of basal plane epoxide
groups with adjacent hydroxyl groups, at moderate temperatures. It was established that these phenol
groups survive even at temperatures of 1000 C.
EXPERIMENTAL
Synthesis: Highly oxidised Graphene Oxide (GO) was produced using a modified Hummers
process.8 The starting material Graphite powder (product: 78391) with particle size 20 m was
purchased from Fluka and is denoted here as pristine graphite. All other chemical and reagents were
purchased from Aldrich.
A mixture of 2.5 g of Graphite and 1.9 g of NaNO3 was placed in a flask cooled in an ice bath. 85 mL
of H2SO4 was added to the mixture and stirred until homogenized. Solution of 11.25 g of KMnO4 in
distilled water was gradually added to the solution while stirring. After 2 hours, the solution was
removed from the ice bath, and further stirred for 5 days. Finally, brown-coloured viscous slurry was
obtained. The slurry was added to 500 mL aqueous solution of 5 wt% H2SO4 over 1 hour while being
continuously stirred. The mixture was stirred for a further 2 hours. Subsequently, 10 ml of H2O2 (30
wt% aqueous solution) was then added to the mixture and stirred for further 2 hours. This mixture was
then left to settle overnight. The mixture was filtered and further purified by dispersing in 500 mL
aqueous solution of 3 wt% H2SO4 and 0.5 wt% H2O2. After two days of precipitation, the supernatant
solution was removed. This process was repeated five times. The solid product obtained after the
rigorous cleaning process was rinsed using copious amounts of distilled water and dried in oven, as
reported in literature.8. The resulting solid was dispersed in water by ultrasonication for 2 h to produce
a GO aqueous dispersion.
After one-day sedimentation, the thick flakes were removed and the
ECSA 300 equipped with monochromatic Al K (h = 1486.6 eV) X-ray source at NCESS Daresbury
Laboratory. Surface charging effects (due to insulating nature of as prepared GO) were compensated
using a Scienta FG300 low energy electron flood gun at 4.0 eV. Step sizes of 1 eV and 0.05 eV were
used for survey and high resolution spectra, respectively. Spectra were collected at room temperature
and then at intervals of 200 C up to 1000 C. Heating for 2 minutes was done inside the chamber under
UHV conditions of the order of 10-7 Torr using electron-beam heater, with electron beam impact on the
back surface of the Si substrate. The samples were cooled before collecting the spectra. High-resolution
in situ Valence band (VB) photoemission spectra were simultaneously collected at each temperature.
For all spectroscopic studies, GO nanosheets were drop dried under infrared lamp to prepare thin films
on Si substrates. Quantification was performed using the data analysis software (the ESCA300 data
analysis software), associated with the SCIENTA ECSA 300 equipment, after performing a Shirley
background correction. Calibration was carried by alignment of the spectra with reference to the C 1s
line at 284.5 0.2 eV associated with graphitic carbon. Binding energies were calibrated by the position
of the Fermi cut-o of a gold foil for valence band data, and by the position of the Au 4f7/2 line (84.0
eV) in the case of core level data.
High resolution in situ Near-edge X-ray absorption fine structure (NEXAFS) spectroscopy was
performed at Synchrotron Radiation Source (SRS), Daresbury Laboratory. Measurements were carried
out at station 5U.1. Spectra at C K-edge and O K-edge were recorded in a total energy yield (TEY)
mode at room temperature, 400 C, 600 C and 800 C. All recorded spectra were normalized to the
signal obtained from a gold covered grid.26
The initial GO sample was characterized by X-ray diffraction (XRD) and Raman spectroscopy, before
and after the in situ temperature dependent XPS studies. XRD data were collected using a Philips
1050/81, for a step size of 0.02 and dwell time of 1 deg/min at standard potential and current settings of
40 kV and 20 mA, respectively, employing a monochromatic Cu K radiation source ( = 1.54 ).
Raman spectroscopy was performed using Argon laser ( = 514.78 nm) at an ISA Lab-Raman system.
Thermo-gravimetric analysis (TGA) was performed using a SDT Q600 V8.3 Build 101 system at a
ramp rate of 1 C/min up to 1000 C in Nitrogen flow of 100 mL/min.
Graphite
(002)
GO
(002)
GO
(b)
(100)
TRGO
10
20
(101)
PG
X 0.2
0.48nm
(a)
30
40
Angle, 2
50
60
(c)
GO
2D
TRGO
G
D
1000
D+G
D'
PG
2D
Figure 1. (a) TEM image of the triple-layered Graphene Oxide (GO) with the corresponding crosssectional profile (bottom inset), and the SAED pattern (top inset). The arrows indicate a sheet separation
of 0.48 nm. (b) XRD spectra and (c) Micro-Raman spectra of as-prepared GO and thermally reduced
GO (TRGO) (vacuum-heat treated at 1000 C), compared with that of pristine graphite (PG). The right
panel of Figure 1(c) shows the magnified the 2D band region. The spectra were shifted in y-scale for
clarity.
graphene oxide sheets,29 with the ss values been reported to vary from ~0.6 nm for dry GO to ~1.2
nm for hydrated samples.7 The difference on the interlayer distance estimated by TEM (0.48A) and
XRD (0.78 A) techniques can be explained bearing in mind the sample preparation and the environment
in which the measurements were carried out. XRD measurements were performed on the samples
comprised of thin films of GO in air, whereas TEM measurements were carried out in vacuum, on
samples produced from dilute dispersion of GO on a TEM grid. The (002) XRD peak for GO shows
considerably larger full width half maximum (FWHM), compared to PG. Upon thermal reduction (at
1000 C in vacuum), the TRGO sample (originated from the GO) exhibits a structure closer to pristine
graphite PG as revealed by the shifting of (002) peak back to 26.4 (ss = 0.337 nm), even though its
FWHM still remained larger than that of PG (Figure 1b), implicating the presence of strains/defects.
Raman Spectroscopy
Raman spectroscopy has played an important role for characterizing graphitic materials since it is able
to provide information on crystalline size, the degree of hybridization, crystal disorder, the extent of
chemical modification, and distinguish single layer graphene or nanotubes from multilayer ones.30-38
The micro-Raman spectra (Figure 1c) of all the samples exhibited three main characteristics peaks: the
G mode, a doubly degenerate (TO and LO) phonon mode (E2g symmetry) at the Brillouin zone center
observed at ~1575 cm-1, originating from in-plane vibration of sp2 carbon atoms; the D mode arising
from the doubly resonant disorder-induced mode (~1350 cm-1) and the symmetry-allowed 2D overtone
mode (~2700 cm-1).31-34 The GO sample shows a prominent D peak with intensity comparable to G
peak, in sharp contrast to the smaller D peak of PG, indicative of significant structural disorder due to
the O-incorporation. Additionally, the D band, attributed to in-plane A1g (LA) zone-edge mode, is
innately Raman-active at the graphitic edges.31-33,35 For small graphene sheets with limited sizes, like the
GO nanosheets synthesized by the harsh Hummers method, the D band is expected to develop
dramatically. Consequently, the sharp increase in ID/IG ratio (from ~0.26 for PG to 0.93 for GO)
indicates a decrease in the in-plane crystal or domain size30 from ~17 nm (PG) to ~4.7 nm (GO). The G
peak of GO is shifted to higher wavenumbers (~ 15 cm-1) and broadens significantly with respect to that
of graphite. Similar upward shifting of the G band has been observed in heavily oxidised carbon
nanotubes39 and was related to the emergence of a new Raman active band (D mode, ~1620 cm-1)
overlapped with the G band.30 The D band, usually inactive, becomes Raman-active due to phonon
confinement caused by defects.40-42 Beside the influence of D band, Kudin et al.35 have also considered
the contributions from the isolated double bonds which yield Raman bands at little higher frequencies
than the G band, for heavily oxidized GO. The vacuum-heat treatment (at 1000 C) tends to shift back
the Raman peaks (~ 12 cm-1 red-shift of G peak with respect to GO), closer to the positions recorded for
PG, indicative of the tendency to recover the hexagonal carbon network. The 2D band at ~2700 cm-1 ,
which originates from a two phonon double resonance Raman process and is indicative of crystalline
graphitic materials, exhibited the most interesting changes. Generally, the position and shape of the 2D
peak are highly sensitive to the number of graphene layers, and has been utilised to distinguish the
single-layer from few-layer graphene.31-34 For the graphene samples prepared by micromechanical
exfoliation of graphite31,33 or by chemical vapour deposition34 on thin metal (e.g. nickel) films, the
single sharp 2D peak of monolayer graphene has observed to become wider and asymmetric with an
upshift in peak position, with the increase in layer number. However in our study it is not possible to
determine the number of layer in the thermally reduced GO, since the GO nanosheets were drop dried
on a Si substrate to prepare a thin film. Hence the Raman spectra is the resultant signal of several
stacked nanosheets, each one consisting of a few layers (2-4, as observed from TEM). The steep
decrease in intensity and broadening of the 2D peak for GO compared with those of PG are mainly
attributed to the steric effects of oxygen moieties on the stacked layers as well as to the partial
amorphization and reduction in sp2 domains.31,38 Interestingly, the TRGO exhibits two distinguishable
peaks close to 2700 and 2950 cm-1 corresponding to the 2D band and to a (D+G) combination mode
induced by disorder.32 Such remarkable features, namely emergence of a sharp 2D band and of a lower
10
intensity distinguishable D+G peak, vividly support the presence of smaller disorder in TRGO when
compared with GO.
Interestingly, after annealing a slight increase in the D peak intensity is observed with the ID/IG ratio
increased to ~1.1, indicating a decrease in the size of sp2 domains upon thermal reduction (~4 nm).
Naturally, a decrease in this ratio would be expected upon annealing since the disorder associated with
the oxygen-defects diminishes. Our experimental observations suggest that for the TRGO sample
(heated at ~ 1000 C) even though the sp2 sites are partially restored, the forced removal of oxygen at
such high temperature leads to the creation of strains and/or topological defects on the C structure,6,27
and hence the isolation of the sp2 clusters forming smaller and dispersed sp2 domains. Similar behavior
i.e. slight increment or no change in ID/IG ratio has been observed in a number of studies7,43 involving
post heating hydrazine reduction. Recent studies36,41 have shown that the intensity ratio I2D/ID+G may be
a powerful indicator for the aromatic C-structural order of the graphitic materials, since the 2D band is
sensitive to the aromatic C-structure, while the combination mode of (D+G) is lattice disorder induced
band for crystalline graphitic materials.37 Well-resolved 2D and (D+G) bands in TRGO sample and its
higher I2D/ID+G ratio (1.3 times higher than GO) indicate the restoration of aromatic C-structure upon
thermal reduction of GO.36,37
11
exhibited well-defined double peak formations, which is a signature of extreme oxidization in GO (at
RT, Figure 2a). The assignment of C1s and O1s components were based on theoretical predictions of
core level shifts and on reported spectra containing the particular oxygen functional groups. The XPS
peaks were fitted to Voigt functions having 80% Gaussian and 20% Lorentzian character, after
performing a Shirley background subtraction. In the fitting procedure, the full width at half-maximum
(FWHM) values were fixed at a maximum limit of 1 eV for all the peaks. The sp2 peak of the C1s
envelope centred at 284.5eV had a FWHM of 1.0 0.2 eV. It was found that the bands appearing at the
higher energy region tended to be much broader (FWHM~1.7 eV) than the sp2 component. In particular,
the FWHM of the components at the tail of the C1s envelopes tended to be much wider than 2eV. In
addition to the sp2 graphite component at 284.49 eV, we found four broad components to account for
the overlapping C1s features. The component at 285.86 eV is assigned to C atoms directly bonded to
oxygen in hydroxyl configurations (shifts of 11.5 eV to higher BE). The component at 286.55 eV is
attributed to epoxide group (C-O-C), and the smaller components at 287.54 and 288.94 eV are related to
carbonyl (>C=O) and carboxyl groups (COOH or HO-C=O). The assignments are in agreement with the
literature,7,10-12,14,27,43-45 even though there is considerable vagueness and subjectiveness. One important
issue is related to the presence of carbonyl >C=O groups: the basic model of GO electronic structure by
LerfKlinowski et al.29 has not accounted for any >C=O moiety, while Dekany model46 and later
Ajayan et al.15 identify its contribution. On the contrary, reports by Jeong et al.27,47 claim the absence of
any experimental evidence of >C=O group, accepting LerfKlinowski model,29 which assumed the
doubly bonded oxygen C=O species exist only as part of the COOH groups at the edge sites of GO
sheets. In a number of reports on GO, the deconvolution of C1s spectra has been performed using four
components, namely sp2, C-OH, C-O-C and COOH, ignoring the presence of >C=O groups.9,27,29,47
Whereas, other reports7,10,11,14 consider only one peak for singly bonded oxygen C O groups,
performing 4-peak-deconvolution for sp2, C O, >C=O and COOH.
12
(b)
Measured spectra
Fit
Fit Components
C-O-C
sp
>C=O
C-OH
C-O-C
C-OH
HO-C=O
(e)
600 C
2
C/O Ratio
Temperature, C
o
400 C
C-O-C
(C = O)
(f)
800 C
2
Raw Graphite
2
sp
sp
sp
35
30
25
20
15
10
5
0
C-OH
sp
40,000
counts
40,000
counts
Intensity/counts
(d)
sp
200 C
20,000
counts
sp
(c)
Thermally Reduced GO
40,000
counts
50,000
counts
RT
1000
counts
Intensity/counts
(a)
C-OH
283
286
289
292
o
(C = O)
C-OH
(C = O)
C-OH
1000 C
(C = O)
Figure 2. High resolution C1s XPS spectra: deconvoluted peaks with increasing reduction temperature
(Tr). Inset of Figure 2c: C/O ratio as a function of Tr (Error bars represent the standard deviation
estimated from six sets of data). Inset of Figure 2f: C1s spectra for starting/precursor graphite (PG).
The symbols specify the following groups:
C-OH Hydroxyl, C-O-C Epoxide, >C=O Carbonyl, HO-C=O Carboxyl.
C O Oxygen doubly bonded to Carbon, C O Singly bonded Oxygen.
* Shake-up satellite peak, sp2 C to C bond in aromatic rings.
The starting graphite (PG) shows 2 main peaks, namely sp2 (284.47 eV) and C-OH (285.53 eV) (inset
of Figure 2f). The presence of a weak C O peak (~285.5 eV) in graphite, associated with atmospheric
oxidation, has previously been observed by Hontoria-Lucas et al.43 Furthermore, Barinov et al.44
assigned the same peak at ~285.6 eV to C O single bond, calculating its close chemical shift of ~1-1.5
13
14
C=C,C-C
C-OH
C-O-C
>C=O
HO-C=O
-*
0.8
0.6
(a)
Graphite
C=C,C-C
C-OH
0.4
0.2
0.0
1.0
(b)
0.6
C-OH
C-O-C
>C=O
HO-C=O
II
III
0.4
0.2
Graphite
C-OH
0.0
0
200
400
600
800
o
Annealing Temperature C
1000
200
400
600
800
1000
Annealing Temperature C
Figure 3. Temperature-dependence of (a) the relative contribution of C1s peak components estimated
by dividing the area under each component by whole C1s peak-area, and (b) the normalised intensity of
O-groups relative to sp2 intensity. The lines shown are guides to the eye only. Dotted lines represent the
values corresponding to C1s peak components, observed for graphite. Error bars represent the standard
deviation estimated from six sets of data.
Thermal Evolution of GO
Restoration of Aromatic C-structure: Upon heating under UHV, the C1s spectrum exhibits a
transformation from a double peak at room temperature to a single sharp peak (~284.5eV) at 1000 C,
resembling the C1s peak of PG and been indicative of a trend to restore the sp2 bonding graphene
character (see Figure S2b). A clear shift of peak-maxima back to lower BE with increasing temperature
signifies the transformation of electrically insulated GO to the conducting nature of graphite. The
evolution of the C/O atomic ratio (inset of Figure 2c and Table S1) reveals an increase in C-content as
reduction temperature (Tr) increases, and an associated decrease of the O groups. A maximum Ccontent of ~97% can be achieved upon heating at 1000 C, with only ~3% of remnant O contribution
(C/O ratio ~ 33.02), alike PG (Table S1). The (*) shake-up satellite peak, observed for PG (inset of
Figure 2f) around ~290.5 eV, appears upon heating at high temperatures 400 C (Figure 2c-f). This
15
indicates that the delocalized conjugation, a characteristic of aromatic C structure, is to some extent
restored in TRGO samples.10,11,15
Thermal Stability of Hydroxyl Groups: Figures 3a and 3b present (i) the relative contribution of the
carbon bonds in GO and (ii) the normalized intensity of O-related peaks (relative to sp2 peak-intensity)
as a function of annealing temperature in UHV respectively. These figures serve as useful guides for
monitoring the evolution of the functional groups and provide insight into the mechanism of thermal
reduction process. At room temperature the determination of separate >C=O and COOH contributions
was possible, however for temperatures higher than 200 C, we denote their combined involvement as C
O (Oxygen doubly bonded to Carbon), since the deconvolution into two separate peaks was not
feasible. The C O component (singly bonded Oxygen, C-OH and C-O-C) remains almost unchanged
(Figure 3b, region I) up to 200 C, most possibly due to insufficient temperature and/or partial
contribution from the transformation of C O to C O. Above 200 C, C O continues to be reduced,
and around 800 C, it almost saturates with a minimal contribution.
Useful information can be obtained from the dispersion of the peak positions as a function of
temperature. It is worth pointing out that at 200 C, the C O peak is centered at ~288.3 eV, between
the >C=O (~287.5 eV) and COOH (288.9 eV) components (Figure 2a and Figure 4a). However at high
temperatures it shifts to BE representative of >C=O. Therefore it is reasonable to postulate that at low
reduction temperatures (Tr < 400 C) the doubly-bonded C O component has contributions from both
>C=O and COOH groups, however at higher temperatures (Tr > 600 C) is dominated by >C=O groups.
In Figure 2b one can note that at Tr >200 C the C-O-C is reduced fast, and at Tr 400 C, it is hard
to be identified (Figure 2d-f). Interestingly, over the temperature range 200-400 C the C-OH increases
rapidly (Figure 3b, region II), followed by a reduction at Tr > 400 C (Figure 3b, region III). C-OH
persists even at 1000 C, at a level similar to that observed in the original graphite material PG (blue
dotted line, Figure 3b). Our observations are consistent with other reports, where the C-OH contribution
16
survives in TRGO showing that annealing at 1000 C is not adequate to completely remove the
oxygen,10-12,14 even though possessing higher C/O ratio (~ 33.02). Complete thermal pyrolysis of C-OH
is thermodynamically difficult,28 mainly due to its intercalating position into the interlayer galleries
between intact conjugated domains.46
100
DSC
80
285
286
C-OH : C-O
C-O-C
287
288
>C=O : C=O
289
OH-C=O
290
Graphite
C=C,C-C
C-OH
291
200
60
TGA
600
800
o
1000
-6
PG
GO
(b)
0
400
-4
40
-*
Annealing Temperature C
-2
II
20
292
0
GO
C=C, C-C : sp
(a)
Weight %
284
200
400
-8
600
800
dW/dT (X 10 ) (mg/ C)
283
1000
Temperature, 0C
Figure 4. (a) Temperature-dependence of the peak-position of C1s peak components; dotted lines
represent the values corresponding to PG. The lines shown are guides to the eye only. (b) TGA results
for GO and PG together with corresponding mass loss rate (dW/dT) for GO.
17
antagonistic activities: either (i) release of by-products and subsequent trapping and/or (ii) loss of
doubly-bonded C O component and simultaneous feeding of C O singly-bonded oxygen groups. In
the same phase, C-OH also shows an increase (Figure 3b, region II) while C O species decrease.
Above 500 C, higher slope in TGA & sharp-rise in dW/dT signal indicate a rapid decomposition of Ospecies, as observed in Figure 3b (region III).
18
I2
400 C
I3
200 C
I3
I3
I1
I1
(c)
10,000
counts
I2
Measured spectra
Fit
Fit Components
(b)
20,000
counts
RT
GO
40,000
counts
Intensity, counts
(a)
I2
I1
I4
I3
I1
(e)
800 C
I2
I1
I3
I2
(f)
1000 C
2000
counts
600 C
5000
counts
(d)
40,000
counts
Intensity/counts
528 530 532 534 536 538 540 528 530 532 534 536 538 540 528 530 532 534 536 538 540
I1
I3
I2
528 530 532 534 536 538 540 528 530 532 534 536 538 540 528 530 532 534 536 538 540
Binding Energy,eV
Binding Energy,eV
Binding Energy,eV
Figure 5. High resolution O1s XPS spectra deconvoluted peaks with increasing Tr.
Formation of phenolic groups is clearly evidenced from the high resolution O1s peaks (Figure 5).
Deconvolution of O1s spectra (Figure 5a) produces 3 main peaks around 531.08, 532.03 and 533.43 eV
assigned to C O (Oxygen doubly bonded to aromatic Carbon-denoted as I1),10,11,14,44 C O (Oxygen
singly bonded to aliphatic Carbon- denoted as I2),6,43 and Phenolic (Oxygen singly bonded to aromatic
Carbon- denoted as I3) 6,43 groups respectively. The pristine GO shows an additional peak at higher BE
(I4 ~534.7 eV) corresponding to the chemisorbed/intercalated adsorbed water molecules.14 Thermal
treatment of GO clearly causes a shift of O1s spectra to the higher energy side and a simultaneous
transformation of O1s spectra from a single peak to a double feature with the development of a
prominent phenolic I3 peak (Figure 5). Although all the O-species decrease in integrated intensity with
increasing Tr (Figure 6a), the relative intensity of phenol group (peak I3) shows a sharp rise around 400
C (Figure 6b), as observed in C1s spectra (Figure 3), followed by a progressive increase relative to
other species (Figure 6b). In addition, the relative intensity of peak I2, O singly bound to aliphatic C,
exhibits an initial increase at 200 C and a subsequent reduction. The initial increase is most possibly
19
due to the internal conversion of C O to C O at low Tr, as described earlier. Undoubtedly, the
evolution of O1s spectra provides further corroborating evidence on the existence of double transition
zone during thermal reduction of GO.
The XPS analysis reveals the OH-moieties on basal plane are the most thermally stable species, in
contrast to observations7,9 made on chemical reduction of GO. Further corroborative evidence on the
evolution of the basal and edge plane O groups, upon thermal reduction, is provided by NEXAFS
(a)
I1
I2
I3
I4
120k
80k
40k
0.75
studies.
(b)
I1
I2
I3
I4
0.50
0.25
0
0
200
400
600
800
o
Annealing Temperature, C
1000
0.00
0
200
400
600
800
1000
Annealing Temperature, C
Figure 6. Temperature-dependence of (a) the integrated intensity of O1s peak components (I1-I4), and
(b) the corresponding relative contribution estimated by dividing the area under each component by
whole O1s peak-area. I1 denotes C O (Oxygen doubly bonded to aromatic Carbon); I2 denotes C O
(Oxygen singly bonded to aliphatic Carbon), I3 denotes Phenolic (Oxygen singly bonded to aromatic
Carbon) and I4 denotes chemisorbed/intercalated adsorbed water molecules groups. Error bars represent
the standard deviation estimated from six sets of data. The lines shown are guides to the eye only.
High Resolution in situ Synchrotron Near-Edge X-ray Absorption Fine Structure (NEXAFS)
Spectroscopy
Soft X-ray absorption spectroscopy probes unoccupied electronic states and is another powerful tool for
20
Intensity, a.u.
C K edge
(a)
I4
I6
1s - * I1
285.216 eV
I7 I8
I5
O K edge
(b)
I9
RT
o
400 C
I3
I2
I3
I4
I6
I7
RT
I8
I9
I5
I1
600oC
400oC
o
800 C
600oC
800oC
1s - *
I2
293.03 eV
Photon Energy, eV
I1 : (1s
*),
*)+
I2 : (1s
*(C-OH),
I3 : *(C-OH)+ *(C-O-C), I4 : *(C = O) : COOH,
I5 : * Excitons, I6 : (1s
*)
525
530
535
540
545
550
Photon Energy, eV
I1 and I2 : *(COOH),
I3 and I4 : *(>C=O) + *(C-O-C)
I5 : *(C-OH), I6 : *(COOH), I7 : *(C-OH)
I8 : *(C-O-C), I9 : *(>C=O)
21
Figure 7. High Resolution (a) C K-edge and (b) O K-edge synchrotron NEXAFS spectra, recorded at
different reduction temperatures. The spectra were shifted in y-scale for clarity.
Upon thermal treatment, a number of important changes can be identified in the spectra of TRGO.
The * resonance (I1) clearly shifts to higher energies, moving towards the position of I2, whereas its
FWHM becomes broader due to presence of mainly C-OH moieties. The two resonances between the *
and * discernible at ~287 (I3) and ~288.7 eV (I4) disappear at 400 C. The assignment of these
intermediate peaks in the absorption spectra of graphene related structures is highly debatable. In
particular, the presence of peak around 288 eV was originally observed in the NEXAFS spectra of
HOPG55 and few-layer graphene56 and was attributed to the free electron like set of bands corresponding
to electronic excitations lying between graphite layers (interlayer states). Others have provided evidence
that this feature (~289 eV) originates from COOH moieties present in single-wall carbon nanotubes,53
carbon fibres49 and GO.9,47 Here, given the prominence of I4 peak in heavily oxidised graphene films
and its complete disappearance upon thermal annealing, we have assigned its origin to COOH consistent
with XPS, where the presence of a sufficient density of carboxylic moieties in GO has been confirmed.
The broad asymmetric peak I3 originates mainly from *(C-O-C) contributions and partly by *(COH).9,47
The * region, which is strongly enhanced due to normal incidence of the polarized beam is
dominated by doubly bonded C O moieties (peaks I7 to I9).53,54 Its temperature evolution reveals a
huge loss of C O groups, which agrees well with our earlier XPS discussion on thermal instability of
C O groups. At 400 C, there is a curious splitting of the * peak (1s*) into 2 peaks, with the
simultaneous emergence of the * exciton peak (I5).49 The appearance of the distinct peak I5 supports
the progressive restoration of graphitic structure in GO, however its reduced/sub-expressed structure,
compared to * peak, indicates the existence of small graphene planes/domains,49 as predicted in XRD
and Raman studies.
22
Remarkably, after thermal reduction, transitions to states of symmetry are seen to decrease
dramatically compared to transitions to states of symmetry. This is visualized in Figure S3, which
depicts the intensity ratio of */* peaks at the C K-edge, as a function of annealing temperature. This
ratio is used to provide an estimate of the relative concentration of sp2 domain configurations in a sp3
matrix consisting of carbon atoms connected to oxygen groups.51 High values of */* ratio around the
temperature zone of 400600 C due to the substantial reduction of edge plane doubly bonded C O
moieties are consistent with the appearance of aromatic characteristic (*) plasmon peak at the same
Tr-range. At higher temperatures Tr > 400 C, the change in */* is less dramatic due to slow
reduction in both edge and basal plane O-moieties.
Further evidence on the significant loss of oxygenated functional groups upon thermal treatment at
400 C is provided by the O K-edge NEXAFS spectra presented in Figure 7b, where the main peaks are
assigned according to literature.9,47,50-52 Two distinctive peaks I1 and I2 appearing at the low energy tail
of O K-edge attributed to the * state of COOH groups located at the GO edge sites,9,47 disappear at 400
C.
23
EF
Graphite
C2p-
800 C
-4 -2 0
2
4
Binding Energy, eV
600 C
C2p-
400 C
o
200 C
-5
1.0
C2s
C2p
RT
(b)
O2s
(c)
0.8
7
6
0.6
10
15
20
25
30
0.4
0
Binding Energy, eV
O2s-C2s Separation, eV
to Ratio
Intensity, a.u.
1000 C
RT
o
400 C
o
600 C
o
800 C
o
1000 C
normalised to -5 eV
(a)
Fermi Edge
Figure 8. (a) High Resolution Valence Band (VB) spectra, recorded at different Tr. The spectra were
shifted in y-scale for clarity. (b) Enlarged view of VB spectra at the vicinity of Fermi level (EF). (c)
Intensity ratios of C2p- to C2p- ( to ) and the separation between O2s and C2s peak-maxima, as a
function of Tr. The lines shown in (c) are guides to the eye only.
A number of interesting features can be identified from the VB spectra. As prepared GO is dominated
by O2s with a band centred at 26.3 eV61 and its Fermi edge shifted to higher energies revealing the
insulating nature of the material. The shoulder located at ~8 eV is associated with strong bonding
states (C2p-) in graphitic like carbon, indicating the presence of a substantial sp2 ordering.44,59,60
Thermal reduction improves the contribution from C2p states and simultaneously reduces the O2s
confirming the deoxygenation of O-moieties. In particular thermal treatment at 600 C, induces the
growth of a new sharp double feature with peaks at 2.3 and 3.25 eV associated with conjugated bonds
24
(C2p- bands) of graphene.44,59,60 An enrichment of the -peak intensity in the valence band
photoemission data implies the formation of sizable graphene domains with 3-fold coordination.60 VB
spectra at 1000 C resemble those of the starting Graphite.
Interestingly, the -derived density of states (DOS) at the vicinity of EF in the range of 0-2.0 eV
(Figure 8b) rises with increasing temperature, with the steepest rise occurring at 600 C. This rise in VB
provides a strong indication for the existence of metallic character in the reduced graphene.44,62 The
simultaneous rise of DOS at EF and the enrichment of -peak intensity in C2p region definitely reflect
the progressive restoration of conjugated aromatic C-structure by de-oxidation via thermal treatment.
Curiously, further increase in temperature (> 600 C) caused a reduction in the -derived DOS
indicating the presence of a high level of defects. The observation is consistent with Raman and XRD
results, which confirmed that thermal treatment at 1000 C is not adequate for the complete restoration
of aromatic C-structure. Presumably the enforced removal of basal O-species produces strains in C=C
bonds and topological wrinkles or hole-like defects on the atomic C structure.6,27 Consistent with the
XPS results the / ratio obtained from C2p-/C2p- bands shows an initial drastic increase followed
by a slower increase (Figure 8c).
Another important observation is the progressive decrease in the O2s-C2s peak separation with
thermal treatment illustrated in Figure 8c. The augmentation of O2s-C2s separation has been associated
with the presence of intermediates between the C O double bond groups and the C O single bonded
O to C, in progressively oxidized graphitic materials, following the order: COC > C O > COH.61
Thus, the smallest separation, observed in highly reduced GO (~800 C), would be attributed to the
prevalence of phenolic/hydroxyls over other O-species.61 The larger separation at low temperatures
would indicate contribution from C-O-C and C O moieties.61
25
(a)
COOH
HOOC
COOH
OH
OH
OH
OH
OH
OH
HOOC
COOH
HO
COOH
HOOC
COOH
: Phenol
OH
: Hydroxyl
: Epoxide
OH
OH
HO
OH
OH
OH
(b)
OH
: Carbonyl
O
OH
HO
OH
OH
OH
HO
HO
COOH : Carboxyl
OH
(c)
HO
OH
OH
OH
OH
CONCLUSIONS
Employing a combination of high resolution in situ temperature dependent spectroscopic techniques
including XPS, VB and NEXAFS we have clarified a number of important issues concerning the
evolution of the electronic structure of GO upon the heat treatment. (i) First it is established, that upon
progressive thermal treatment, the edge-plane COOH groups become highly unstable, whereas >C=O
are more difficult to be removed. (ii) The C-OH group in the phenolic form is the most thermally stable
of all the oxygen species (Figure 9c). (iii) The thermal evolution of C-OH groups exhibits a welldefined transition temperature around 400 C. At lower temperatures, there is an upward trend due to
the formation of phenol groups whereas at Tr > 400 C there is decreasing trend due to their
decomposition (Figure 9b-c). (iv) Valence band spectra reveal a drastic increase in the DOS near EF at
26
600 C, whereas further increase in temperature reduces the DOS to levels similar to those of graphite.
The ability to achieve higher DOS at considerably lower temperatures compared to 1000 C normally
considered for partially restoring the aromaticity of thermally reduced Graphene Oxide holds significant
advantages for nanoelectronics application.
ACKNOWLEDGMENT. The work was supported by The Leverhulme Trust (fellowship for Dr
Ganguly: 1-212-R-0197), The Royal Academy of Engineering, EPSRC funded facility access to
HRTEM at University of Oxford (EP/F01919X/1), NCESS XPS facility (EP/E025722/1) in Daresbury,
and NCESS Synchrotron facility in Daresbury. Also the assistance from Mr. S. Ukleja at FSERTC
(UU); and Dr. D. S. Law and Dr G. Beamson at NCESS is acknowledged. Supporting Information is
available online at http://pubs.acs.org.
27
REFERENCES.
[1] Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Grigorieva, M. I. K. I. V.; Dubonos, S.
V.; Firsov, A. A. Nature 2005, 438, 197.
[2] Du, X.; Skachko, I.; Barker, A.; Andrei, E. Y. Nat. Nanotechnol. 2008, 3, 491.
[3] Geim, A. K. Science 2009, 324, 1530.
[4] Brodie, B.; Ann. Chim. Phys. 1855, 45, 351.
[5] Hummers Jr, W. S.; Offeman, R. E. J. Am. Chem. Soc. 1958, 80, 1339.
[6] Schniepp, H. C.; Li, J. L.; McAllister, M. J.; Sai, H.; Herrera-Alonso, M.; Adamson, D. H.;
Prud'homme, R. K.; Car, R.; Saville, D. A.; Aksay, I. A. J. Phys. Chem. B 2006, 110, 8535.
[7] Stankovich, S.; Dikin, D. A.; Piner, R. D.; Kohlhaas, K. A.; Kleinhammes, A.; Jia, Y.; Wu, Y.;
Nguyen, S. B. T.; Ruoff, R. S. Carbon 2007, 45, 1558.
[8] Eda, G.; Fanchini, G.; Chhowalla, M. Nat. Nanotechnol. 2008, 3, 270.
[9] Lee, V.; Whittaker, L.; Jaye, C.; Baroudi, K. M.; Fischer, D. A.; Banerjee, S. Chem. Mater. 2009,
21, 3905.
[10] Mattevi, C.; Eda, G.; Agnoli, S.; Miller, S.; Mkhoyan, K. A.; Celik, O.; Mastrogiovanni, D.;
Granozzi, G.; Garfunkel, E.; Chhowalla, M. Adv. Func. Mater. 2009, 19, 2577.
[11] Bagri, A.; Mattevi, C.; Acik, M.; Chabal, Y. J.; Chhowalla, M.; Shenoy, V.B. Nature Chem.
2010, 2, 581587.
[12] Yang, D.; Velamakanni, A.; Bozoklu, G.; Park, S.; Stoller, M.; Piner, R. D.; Stankovich, S.; Jung,
I.; Field, D. A.; Ventrice, C. A. Carbon 2009, 47, 145.
28
[13] Shin, H. J.; Kim, K. K.; Benayad, A.; Yoon, S. M.; Park, H. K.; Jung, I. S.; Jin, M. H.; Jeong, H.
K.; Kim, J. M.; Choi, J. Y. Adv. Func. Mater. 2009, 19, 1987.
[14] Akhavan, O. Carbon 2010, 48, 509.
[15] Gao, W.; Alemany, L.B.; Ci, L.; Ajayan, P.M. Nature Chem. 2009, 1, 403.
[16] Gao, J.; Liu, F.; Liu, Y.; Ma, N.; Wang, Z.; Zhang, X. Chem. Mater. 2010, 22, 2213.
[17] Wang, Y.; Shi, Z.; Yin, J. ACS Appl. Mater. Interfaces 2011, 3, 1127.
[18] Esfandiar, A.; Akhavan, O.; Irajizadab, A. J. Mater. Chem. 2011, 21, 10907.
[19] Zhu, C.; Guo, S.; Fang, Y.; Dong, S. ACS Nano 2010, 4, 2429.
[20] Zhou, Y.; Bao, Q.; Tang, L. A. L.; Zhong, Y.; Loh, K. P. Chem. Mater. 2009, 21, 2950.
[21] Akhavan, O. ACS Nano, 2010, 4 4174.
[22] Williams, G.; Seger, B.; Kamat, P.V. ACS Nano 2008, 2, 1487.
[23] Akhavan, O.; Ghaderi, E. J. Phys. Chem. C 2009, 113, 20214.
[24] Sharma, S.; Ganguly, A.; Papakonstantinou, P.; Miao, X.; Li, M.; Hutchison, J.L.; Delichatsios,
M.; Ukleja, S. J. Phys. Chem. C 2010, 114, 1945919466.
[25] Acik, M.; Mattevi, C.; Gong, C.; Lee, G.; Cho, K.; Chhowalla, M.; Chabal, Y. J. ACS Nano 2010,
4, 5861 -5868.
[26] Bhattacharyya S.; Lbbe, M.; Richter, F. J. Appl. Phys. 2000, 88, 5043.
[27] Jeong, H. K.; Lee, Y. P.; Lahaye, R. J. W. E.; Park, M. H.; An, K. H.; Kim, I. J.; Yang, C. W.;
Park, C. Y.; Ruoff, R. S.; Lee, Y. H. J. Am. Chem. Soc. 2008, 130, 1362.
29
[28] Boukhvalov, D. W.; Katsnelson, M. I. J. Am. Chem. Soc. 2008, 130, 10697.
[29] Lerf, A.; He, H.; Forster, M.; Klinowski, J. 1998, J Phys Chem B 1998, 102, 4477.
[30] Tuinstra, R.; Koenig, J. L. J. Chem. Phys. 1970, 53, 1126.
[31] Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.; Piscanec, S.;
Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K. Phys. Rev. Lett. 2006, 97, 187401.
[32] Pimenta, M. A.; Dresselhaus, G.; Dresselhaus, M. S.; Cancado, L. G.; Jorio, A.; Saito, R. Phys.
Chem. Chem. Phys. 2007, 9, 1276.
[33] Calizo, I.; Balandin, A. A.; Bao, W.; Miao, F.; Lau, C. N. Nano Lett. 2007, 7, 2645.
[34] Kim, K. S.; Zhao, Y.; Jang, H.; Lee, S. Y.; Kim, J. M.; Kim, K. S.; Ahn, J.-H.; Kim, P.; Choi, J.Y.; Hong, B. H. Nature 2009, 457, 706.
[35] Kudin, K. N.; Ozbas, B.; Schniepp, H. C.; Prudhomme, R. K.; Aksay, I. A.; Car, R. Nano Lett.
2008, 8, 36.
[36] Su, C. Y.; Xu, Y.; Zhang, W.; Zhao, J.; Tang, X.; Tsai, C. H.; Li, L. J. Chem. Mater. 2009, 21,
5674.
[37] Zhan, D.; Ni, Z.; Chen, W.; Sun, L.; Luo, Z.; Lai, L.; Yu, T.; Wee, A. T. S.; Shen Z. Carbon
2011, 49, 1362.
[38] Dresselhaus, M. S.; Jorio, A.; Hofmann, M.; Dresselhaus, G.; Saito, R. Nano Lett. 2010, 10, 751.
[39] Murphy, H.; Papakonstantinou, P.; Okpalugo, T.I.T. J. Vac. Sci. Technol. B. 2006, 24, 715.
[40] Tung, V. C.; Allen, M. J.; Yang, Y.; Kaner, R.B. Nat. Nanotechnol. 2009, 4, 25.
[41] Paredes, J. I.; Rodil, S.V.; Fernandez, P. S.; Alonso, A. M.; Tascon, J. M. D. Langmuir 2009, 25,
30
5957.
[42] Navarro, C. G; Weitz, T. R.; Bittner, A. M.; Scolari, M.; A. Mews, A.; Burghard, M.;Kern, K.
Nano Lett. 2007, 7, 3499.
[43] Hontoria-Lucas C.; Lopez-Peinado, A. J.; Lpez-Gonzlez, J. D.; Rojas-Cervantes, M. L.;
Martin-Aranda, R. M. Carbon 1995, 33, 1585.
[44] Barinov, A.; Gregoratti, L.; Dudin, P.; Rosa, S. L.; Kiskinova, M., Adv. Mater. 2009, 21, 1916.
[45] Wang, Y.Q.; Zhang, F. Q.; Sherwood, P. M. A. Chem. Mater. 1999, 11, 2573.
[46] Szabo, T.; Berkesi, O.; Forg, P.; Josepovits, K.; Sanakis, Y.; Petridis, D.; I. Dkny, I. Chem.
Mater. 2006, 18, 2740.
[47] Jeong, H. K.; Noh, H. J.; Kim, J. Y.; Jin, M. H.; Park, C. Y.; Lee, Y. H., EPL 2008, 82, 67004.
[48] Scholz, W.; Boehm, H. P.; Z. Anorg. Allg. Chem. 1969, 369, 327.
[49] Zhang, G. S.; Sun, S.; Yang, D.; Dodelet, J. P.; Sacher, E. Carbon 2008, 46, 196.
[50] Banerjee, S.; Hemraj-Benny, T.; Balasubramanian, M.; Fischer, D. A.; Misewich, J. A.; Wong,
S. S. 2004, ChemPhysChem. 2004, 5, 1416.
[51] Roy, S.S.; Papakonstantinou, P.; Okpalugo, T.I.T.; Murphy,H. J. Appl. Phys. 2006, 100, 053703.
[52] Ray, S. C.; Tsai, H. M.; Chiou, J. W.; Bose, B.; Jan,J. C.; Kumar, K.; Pong, W. F.; Dasgupta,
D.; Tsai, M. H. J Phys- Condens Matter. 2004 16, 5713.
[53] Kuznetsova, A.; Popova, I.; Yates, J. T., Jr.; Bronikowski, M. J.; Huffman, C. B.; Liu, J.;
Smalley, R. E.; Hwu, H. H.; Chen, J. G. J. Am. Chem. Soc. 2001, 123, 10699.
[54] Bournel, F; Laffon, C.; Parent, P.; Tourillon, G. Surf. Sci. 1996, 352, 228.
31
[55] Fischer, D.A.; Wentzcovitch, R. M.; Carr, R. G.; Continenza, A.; Freeman, A. J. Phys. Rev. B
1991, 44, 1427.
[56] Pacil, D.; Papagno, M.; Rodrguez, A. F.; Grioni, M.; Papagno, L.; Girit, C.; Meyer, G.C.;
Begtrup, G. E.; Zettl, A. Phys. Rev. Lett. 2008, 101, 66806.
[57] Jeong, H. K.; Yang, C.; Kim, B. S.; Kim, K. J. EPL 2010, 92, 37005.
[58] Ago, H.; KugIer, I. J. T.; Caciaili, F.; Petritsch. K.; Friend, R. H.; Salaneck, W. R.; Ono, Y.;
Yamabe, T.; Tanaka, K. Synthetic Metals 1999, 103, 2494-2495.
[59] McFeely, F. R.; Kowalczyk, S. P.; Ley, L.; Cavell, R. P. O.; Shirley, D.A. 1974, Phys. Rev. B
1974, 9, 5268.
[60] Larciprete, R.; Lizzit, S.; Botti, S.; Cepek, C.; Goldoni, A. Phys. Rev. B 2002, 66, 121402.
[61] Yue, Z. R.; Jiang, W.; Wang, L.; Gardner, S. D.; Pittman, C. U.; Carbon 1999, 37, 1785.
[62] Kim, S. Y.; Lee, J. Y.; Park, J.; Park, C. J.; Lee, C. J.; Shin, H. Chem. Phys. Lett. 2006, 420, 271.
32
OH
O
COOH
HOOC
COOH
OH
OH
OH
OH
OH
OH
HOOC
COOH
OH
HO
COOH
HOOC
COOH
VACUUM HEAT
OH
HO
OH
OH
OH
OH
SYNOPSIS TOC.
Graphene Oxide (GO)
RT
Measured spectra
Fit
Fit Components
sp
C-O-C
>C=O
C-OH
HO-C=O
Thermally Reduced
Graphene Oxide (TRGO)
2
sp
C-OH
800 C
(C = O)
33
Supporting Information
Probing the Thermal Deoxygenation of Graphene Oxide using High Resolution In Situ X-Ray based
Spectroscopies
34
Supplementary Figures
Table S1. High resolution XPS analyses: atomic percentage of carbon (C) and oxygen (O), estimated
from area under C1s and O1s peak, respectively, as a function of temperature.
RT
200C
400C
600C
800C 1000C
PG
67.54
74.52
83.42
86.80
93.66
97.06
96.58
32.46
25.48
16.58
13.20
6.34
2.94
3.42
1.69nm
Figure S1. TEM image of the 4-layered GO with the corresponding cross-sectional profile (bottom
inset): the arrow indicate the total thickness of 4-layers (1.69 nm), implicating the sheet separation of
0.42 nm.
35
(b)
C1s
60k
40k
(O% = 32.46)
O1s
20k
GO(as prepared)
0
0
500
1000
Binding Energy, eV
RT
Intensity(Normalised)
Intensity/ counts
(a)
1500
2000C
4000C
6000C
8000C
10000C
--- Graphite
282
284
286
288
Binding Energy, eV
290
Figure S2. (a) WESS shifted due to charging GO. (b) High-resolution XPS spectra: C1s variation with
temperature. The spectra were shifted in y-scale for clarity.
Intensity Ratio of
- to (a.u.)
C K edge
0.65
0.60
I1 (or I2) / I5
I1 (or I2) / I6
0.55
0
200
400
600
800
o
Annealing Temperature, C
Figure S3. Intensity ratio of * contribution (peak I1 or I2) to * (peak I5 or I6) as a function of
temperature.
36