unoccupied electronic structure of ball-milled graphite
TRANSCRIPT
Unoccupied electronic structure of ball-milled graphite
Adriyan S. Milev,*aNguyen H. Tran,
aG. S. Kamali Kannangara
aand
Michael A. Wilsonab
Received 14th December 2009, Accepted 10th March 2010
First published as an Advance Article on the web 23rd April 2010
DOI: 10.1039/b926345d
Changes in electronic and vibrational structure of well characterised macrocrystalline graphite
milled by a planetary ball-mill are investigated by Raman spectroscopy and Near Edge X-ray
Absorption Fine structure (NEXAFS) measurements at the C K-edge. The electronic structure
changes at the surface and in the sub-surface of the particles are examined by comparing
two-different NEXAFS detection modes: total fluorescence yield (TFY) and partial electron yield
(PEY) respectively. When the in-plane crystallite sizes of graphite are decreased to nanosized
(from B160 nm to B9 nm), a new spectral structure appears in TFY at 284.1 eV which is not
present in the macrocrystalline graphite. This feature is assigned to electronic states associated
with zigzag edges. Further the TFY shows a shift of the main graphite p* band from 285.5 to
285.9 eV, attributed to breaking the conjugation and hence the electron localization effect during
milling, The TFY spectra also show strong spectral features at 287.5 and 288.6 eV, which suggest
that the local environment of carbon atoms changes from sp2 to more sp3 due to physical damage
of the graphite sheets and formation of structures other than aromatic hexagons. Complementary
Raman spectroscopic measurements demonstrate an up-shift of the graphite G band from
1575 to 1583 cm�1 en route to nanosize. The changes in TFY NEXAFS and Raman spectra are
attributed to modification of the sub-surface electronic structure due to the presence of defects in
the graphite crystal produced during milling. The discovery of the strong spectral feature at
284.1 eV in nanographite and the 0.4 eV up-shift of the p* band may open up possibilities to
influence the electronic transport properties of graphite by manipulation of defects during the
preparation of the nanographite.
Introduction
The element carbon can exist without bonding with other
elements in sp, sp2 or sp3 hybridised allotropes. Graphite is one
of the most common and interesting allotropic forms which
can have various degrees of order in its crystal structure. In
principle, pure graphite is a 3D layered material formed by
stacking of 2D hexagonal sheets where the sheets are stacked
usually in an alternating (ABAB. . .) sequence giving rise to a
unit cell that consists of four atoms (Fig. 1a and b).1
The bonding within the sheets is due to the overlap of sp2
hybridized 2s, 2px and 2py atomic orbitals of carbon atoms
giving rise to three s-bonds that form the hexagonal network
(s states).
The bonding between the sheets is derived from the overlap
between the non-hybridized out-of-plane 2pz orbitals from the
adjacent sheets (p-states).2 This electronic configuration gives
rise to stronger in-plane bonds (0.142 nm) and weaker out-of-plane
bonds (0.335 nm). At the edges the graphite sheets are
terminated by zigzag or armchair structures which may be
described to have dangling bonds.
A variety of graphitic forms are known in which there are
different degrees of disorder. This could be due to the presence
of small amounts of other elements giving rise to structures
known as anthracites, due to internal packing order sometimes
called amorphous regions or due to the disorder that occurs at
surfaces where the effective forces in the sub-surface are
Fig. 1 (a) 3D graphite lattice. The unit cell is given by dashed lines.
The stacking is called ABAB. . ., where A and B refer to two families of
planes shifted on one another. (b) Top view of 3D graphite lattice
showing the structural difference between the A-type and B-type
carbons.
a School of Natural Sciences, University of Western Sydney,Penrith South, DC NSW 1797, Australia.E-mail: [email protected]
bCSIRO Earth Sciences and Resource Engineering RiversideCorporate Park, 11 Julius Avenue, North Ryde, NSW 2113,Australia
This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 | 6685
PAPER www.rsc.org/pccp | Physical Chemistry Chemical Physics
challenged by surface forces of structure termination. One area
of topical interest is when the surface forces of graphite are
large enough to become substantial and begin to challenge
those present in the sub-surface. This is an interesting
phenomenon when particles become nanosized. Not
surprisingly, the properties of nanographite strongly depend
on the method of preparation because small changes in size
can influence the ratio of surface to sub-surface and they may
also affect the amount of defects in the graphite structure.3–5
Raman spectroscopy and Near Edge X-ray Absorption
Fine Structure Spectroscopy (NEXAFS) are very useful for
studying such changes.3–10 Raman spectra of graphites consist
of a G band peak at B1575 cm�1 and peaks at around
1345 cm�1 and at B1610 cm�1, corresponding to disorder
induced features (called D and D0 bands). The relative
intensities of the G and D peaks called the ID/IG ratio can
be used to establish degree of order. The full widths at half
maximum (fwhm) of the bands can also be measured and
interpreted.
The complementary NEXAFS spectroscopy can give
specific information on the atomic environments and can
discriminate between sp3 and sp2 configurations.9,10 The
carbon K-edge NEXAFS measurements are usually performed
in partial electron yield mode (PEY) which probes the top
1–5 nm layer or total electron yield (TEY) mode that probes
the 1–10 nm of the material. A third detection technique called
total fluorescence yield (TFY) mode, allows details of carbon
X-ray absorption in the sub-surface region 1–100 nm.11–13
The X-ray absorption of the carbon K edge is due to 1s - 2p
electronic transitions and usually consists of two energy
regions associated with 1s - p* transitions at B284–288 eV
and 1s - s* transitions at 4289 eV.9 It thus probes local
order. For all highly ordered graphites, these transitions
give rise to peaks at 285.4 and 291.7 eV, respectively.7,14,15
However, one or two additional resonances in the energy
range of 286 to 289 eV are sometimes observed. The origin
of these peaks is not universally agreed.7,9,16–22 Some
investigators have attributed a peak at B287–290 eV to
atom impurities such as C–O bonds18,19,21 or C–H bonds.9,17
However, experiments by Reihl et al.,22 demonstrated that the
feature at 287.5 eV is not affected by the exposure to activated
hydrogen or activated oxygen or small amounts of water and
independently, Atamny et al.,16 could find no C K-edge
spectral differences between the graphite samples exposed to
oxygen at 1000 K and pristine graphite surfaces. These results
suggest that the alternative assignment by Posternak et al.,20
and Fisher et al.,7 of a peak at B287.5 eV to an additional
three-dimensional interlayer states, may be correct. Recent
theoretical work by Coleman et al.,23 suggests that the peaks in
the range 287–288 eV can be attributed to the presence of
defects such as atomic vacancies in graphite. The interpretation
is that this is due to some sp3 bonded carbon atoms near
defects and can be backed by other evidence as discussed
further below in the bulk of the paper. It is probably
circumspect however, given the variable nature of graphite,
not to exclude C–O and C–H assignments for some graphites.
Our view is that these bonds will give rise to characteristic
vibrations which could be observed using infra red and/or
Raman spectroscopy.
We note that because of the variability of graphite
structures as discussed above, other less ordered or more
ordered structures may behave differently and are worth of
investigation. Elsewhere, we have reported Raman, preliminary
NEXAFS investigations and detailed X-ray diffraction analysis
on the crystallite sizes and their distributions of the graphite
used here as precursor.24–26 In this paper, we first discuss
NEXAFS results for nanographite prepared by ball milling.
Then we compare the surface and sub-surface electronic
structures of the products by measuring electron yield (PEY)
and total fluorescence yield (TFY) simultaneously and relate
them to Raman spectra of the same materials. It will be
demonstrated that the electronic structure of these nano-
graphites is unique due to the accumulation of lattice defects.
Experimental
Sample preparation
The precursor graphite powder (499% purity) was obtained
from Fluka. Graphite nanoparticles were prepared by ball-
milling of 10 g polycrystalline graphite powder in a Rietsch
Planetary Ball mill PM 100. Each sample was loaded into a
stainless-steel container (250 ml) together with 6 stainless-steel
balls (diameter 25 mm). The container was evacuated, then
purged with argon (499.999 purity, rate B3 l min�1 for 5 min)
and later the pressure was increased to about 200 kPa. The
milling was carried out for 3, 10 and 30 h at 400 rpm and the
samples were code named as G3, G10, and G30, respectively.
The precursor graphite was code named as G0. The
milling container was opened in a glove box under nitrogen
(499.99 purity) and the milled powders were transferred into
containers. The containers with the samples were kept sealed
until the NEXAFS measurements.
Transmission electron microscopy
Graphite samples were examined by a Philips Biofilter-120
transmission electron microscope (TEM) operating at 120 kV.
Samples for TEM observation were prepared by mixing
sample powder with acetone and dispersing the sample
aggregates by ultrasound treatment. A drop of the suspension
was transferred onto lacey carbon foils supported on copper
grids for examination.
Raman
The Raman spectra were acquired on a Renishaw inVia Reflex
spectrometer, with a 2400 grooves per millimetre grating,
equipped with a Peltier cooled CCD detector. The measurements
were carried out under a microscope (�50 objective) with the
collection optics based on a Leica DMLM confocal
microscope. The instrument was calibrated against the Stokes
Raman signal of pure Si at 520 cm�1. The 514.5 nm (2.41 eV)
line of an argon laser with power at the sample surface limited
to o1.6 mW was used to excite the samples. This relatively
low laser power was used to minimise the likelihood of sample
heating and thus to prevent downshifts of the Raman peak
position.27,28 Ten scans with acquisition time of 30 s were
co-added over the scan range of 1100–1800 cm�1, where.
All spectra were normalized to the intensity of the G band
6686 | Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 This journal is �c the Owner Societies 2010
peak at B1575 cm�1. Peak positions, peak intensities, full
widths at half maximum (fwhm) were determined by software,
OPUS version 6.5 (Bruker). The peak positions and their full
widths at half maximum (fwhm) and the ID/IG integrated
intensity ratio (D and G peak area ratio) obtained from the
fitting to each spectrum are employed to quantify the changes
which occurred in the samples upon milling. The integrated
peak intensity ratio of the D and G bands (ID/IG) are used to
quantify the changes in-plane crystallite sizes (La).
NEXAFS
Carbon K-edge spectra were acquired using the wide range
beam-line (BL24A, energy approximately 10–1500 eV) at the
National Synchrotron Radiation Research Centre (NSRRC)
in Hsinchu, Taiwan. The carbon K-edge NEXAFS spectra
were acquired in the energy region from 275 to 322.5 eV.
Oxygen K-edge spectra in the energy region of 525–550 eV
were also recorded. To degas the absorbed species during
sample mounting, all samples were baked out at B200 1C in
the ultra high vacuum analysis chamber (B10�9 mbar) for 24 h.
The spectra were acquired at room temperature with the
surface sensitive, partial electron yield (about 1–10 nm
sampling depth for energies o1000 eV) and sub-surface
sensitive, fluorescence yield (sampling depth about 1–100 nm
at 200–500 eV energy range).12 The electron detector operated
in a partial yield mode, with a retarding potential of �70 V at
the entrance of the detector, which was used to repel low
energy electrons originating mainly from the sub-surface.
Spectra calibration
Normally, the calibration is referred to either p* or s*resonances, which are presumed independent of the sample
preparative conditions. The optics of the NEXAFS end-station
used here had a minor potassium contamination which in-fact
provided useful spectral features as an external reference
during calibration.29 First, the photon energy was calibrated
using the precursor graphite sample G0 by taking the value
of the C1s p*-transition to be at 285.5 eV and then the
energy position of the reference, potassium L-edge 2p3/2, was
recorded at 297.6 eV. All the samples, G3–G30, were
calibrated against the value of 297.6 eV. Thus, it was possible
to determine the shift in the position of all graphite-related
peaks for each sample without forcing them at certain energies
as per usual calibration method.19 Calibration, background
subtraction and pre- and post- C K-edge normalisation
procedures were applied on the spectra using an algorithm
developed by Newville et al.,30 and implemented in the
software package ATHENA.31 This removes any ambiguities
in the identification of the background and peak position and
thus provides a reliable criterion for a consistent normalisation
of all spectra.29
Results and discussion
Changes in Raman spectra on ball milling
The main spectral positions of all samples (G0–G30) were
similar to sp2 bonded graphitic materials already reported in
the literature (Fig. 2).32–38 However, some variation in the
relative intensities were observed.
The main features observed in the Raman spectra of our
samples as per the usual are a peak at 1575 cm�1, corresponding
to ordered graphite structure (graphite, G band) and peaks at
around 1345 cm�1 and at B1610 cm�1, corresponding to
disorder induced features (called D and D0 bands).33,34 The
main effects on Raman spectral features after milling of
graphite are; (i) the G peak moves from 1575 to 1583, and
its fwhm increases (ii) D and D0 -peaks intensities and fwhm
increase but no significant peak shifts are observed (Table 1).
The increased intensity of the D band in the Raman spectra of
graphites has been interpreted as decreased in-plane crystallite
sizes.35–38 On an atomic scale, this may be regarded as an
increase of the graphite edges because of the smaller size of the
crystallites, a feature which is expected on milling.
The ratio of the integrated intensities of the G band to D
band is proportional to the average in-plane crystallite size
(La) by the equation; La (nm) = (2.4 � 10�10)l4laser (ID/IG)�1,
where llaser is the frequency of the exciting laser in nm.39 Upon
milling, the average La decreases fromB160 nm (G0) toB9 nm
(G30). This indicates that the number of laterally broken
graphene sheets increases and therefore the amount of boundary
regions along the fracture lines also increases en route to the
nanocrystalline state.
The changes in G band spectral position and the broadening
of the G and D band features are more interesting. The G
band of graphite is characteristic of all sp2 sites, including
alkenic CQC sites and not just those in aromatic rings.4
It always lies in the range 1500–1630 cm�1, as it does in
Fig. 2 Comparison of the Raman spectra for the G0–G30 samples.
The spectra are normalised to the intensity of the G band peak set at 2.
Table 1 Normalized Raman spectra fitted with Lorentzian shapefunction. The in-plane crystallite sizes are also given. Peak intensityareas (Ia) are given
Sample
G D La
cm�1 Ia fwhm cm�1 Ia fwhm nm
G0 1575 62.6 19.8 1345 6.5 34.0 160G3 1575 72.2 22.8 1345 36.9 39.0 33G10 1581 84.2 29.5 1346 136.4 51.5 10G30 1583 99.7 39.7 1347 176.2 60.1 9
This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 | 6687
spectra of other aromatic and alkenic solids. Nevertheless
differences can be delineated. The uncongugated alkenic
CQC bonds are shorter than aromatic bonds, so they
have higher vibration frequencies. Thus, during gradual trans-
formation from sp2 bonded conjugated hexagonal rings to sp2
bonded chains, a shift in the G band to higher wave numbers is
expected. Moreover, it is well established that the width of the
G peak is proportional to the bond-angle or bond-length
disorder at the sp2 sites.4 On the other hand, the D band
intensity is proportional to the carbons only in the aromatic
rings in clusters with small sizes, whereas the D band broadening
is proportional to the distribution of clusters containing
hexagonal aromatic rings with different orders and dimensions.
In other words, the D band intensity is proportional to the
number of hexagonal aromatic rings in the cluster whereas
carbons in non-aromatic bonds do not contribute to the
intensity of the D band.4 That is, if there is any change in
total disorder where there would be no change in the intensity
of the D, but if the distribution of disordered regions changes,
this would still be reflected in D line broadening. For the
samples prepared here, we note that the D-peak intensity and
fwhm increase with increasing milling time. This suggests that
not only more disordered clusters are formed but also they
become randomly distributed. In addition, these data confirm
the presence of more edges due to the presence of smaller
crystallites and the introduction of other forms of disorder in
the crystallites during the ball milling of the graphite.
Changes in NEXAFS spectra on ball milling
Fig. 3a compares the normalized C 1s NEXAFS Partial
Electron Yield (PEY) spectra obtained for the precursor,
G0 and nanocrystalline samples, G3–G30. To minimise the
contribution from electrons from the sub-surface regions, we
adopted the procedure reported previously.9 That is, the
entrance of the electron detector was biased by a negative
voltage to repel the low-energy electrons. The spectrum for
G0 is in agreement with those previously reported.7
The first sharp feature at 285.5 eV corresponds to a
transition from the C 1s initial states to a final p* states which
have pz symmetry polarized along the out-of-plane direction.7
The strong features at 291.7 eV and 292.8 eV are attributed
to excitations from the C 1s initial states to unoccupied
s* orbitals which have pxy symmetry polarized along the
in-plane direction.7,15
Theoretical investigations by McCulloch and Brydson13
demonstrated that the s* peaks at 291.7 eV and 292.8 eV
(called s1* and s2* peaks) are due to partial overlap of the
pxy orbitals with pz orbitals along c-axis of structurally
non-equivalent carbon atoms that belong to two-different
sub-lattices (A- and B-type carbons in Fig. 1). However, in
PEY spectra, these are evident only in those graphite samples
which are highly crystalline where good alignment between the
graphene planes exists.13 Our spectral data show that s1* and
s2* peaks are present in the initial graphite sample (G0) but
change rapidly in their intensity with milling time. The
reduction in intensity of these features is particularly evident
in the spectra for samples G10 and G30 where the s* transitionsare broad and the s1* s2* peaks are no more resolvable.
The latter can be explained by poor ABAB. . . stacking order
of the graphene sheets due to milling. The PEY spectra of all
samples also show two relatively weak peaks at 287.5 eV and
at 288.6 eV which will be discussed in detail below.
Origin of 287.5 and 288.6 eV transitions
The PEY spectra of all samples show two relatively weak
peaks at 287.5 eV and at 288.6 eV. However, in the Total
Fluorescence Yield (TFY) spectra of the milled graphite these
two peaks are the strongest features (Fig. 3 b). We ruled out
the presence of impurities containing oxygen and hydrogen by
observing the near-edge features at energies 4530 eV in the
O K-edge NEXAFS spectra acquired in TFY mode (insert in
Fig. 3b) and analysing the Raman and IR spectral data for
each of the samples.
Fig. 3 Comparison of C K-edge NEXAFS spectra of ball-milled
graphite powders. (a) Partial Electron Yield spectra (b) Total Fluorescence
Yield spectra. X-ray absorption features of potassium L-edge used as
an internal calibrant are marked by the asterisk (K 2p3/2 is at 297.6 eV).
The K 2p1/2 of potassium gives rise to a second peak at 300.2 eV. The
insert in Fig. 3b shows O K edge NEXAFS spectra collected in
fluorescence yield mode. Raw spectra without processing or normalization
are presented. The spectra are compared to the O K-edge spectrum
from the Au electrode. No significant difference between the oxygen
contents of the unmilled and milled samples and the synchrotron
optics is observed.
6688 | Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 This journal is �c the Owner Societies 2010
Recent theoretical work by Coleman et al.,23 suggests that
vacancies in the lattice structure of graphite give rise to peaks in
the range 287–288 eV. For a carbon atom that is far from a
defect, the px and py states will coincide and form the in-plane
s bond as expected in a highly crystalline graphite. However,
close to a defect site, the px and py orbitals will no longer
degenerate due to decreased symmetry and an additional
NEXAFS peak appears for atoms close to a defect.23 The new
peak corresponds to a bond with an s-like character similar to
that found in sp3 bonded carbons such as cyclohexane40 and
diamond.41,42 Therefore, the presence of the peaks at 287.5 and
288.6 eV in our milled samples suggests that the hybridization of
the CQC bonds changes from sp2 to that with more sp3 character
due to physical damage of the graphite sheets and formation of
structures other than aromatic hexagons.
A newly discovered 284.1 eV transition and energy shift of the
transition at 285.5 eV
Fig. 3b shows the NEXAFS spectra acquired in total
fluorescence yield (TFY) mode. The TFY spectra of the
precursor graphite sample is quite similar to data in the
literature,11 but differ considerably for all milled samples,
especially in the energy region below 286 eV. That is, a new
strong feature at 284.1 eV and a shift of the p* transition from
285.5 to 285.9 eV is observed. Because of the different probe
depth of the PEY and TFY detection modes we assign these
new features to electronic states that are present in sub-surface
only. As far as we are aware these electronic states have not
been observed previously in any graphitic structures.
Theoretically, when an atomic vacancy is created in the
A (or B) sub-lattice, new localized p states at the lattice sites
around the zigzag edges appear while no edge states appear for
armchair edges.43–46 These states lie just above the Fermi level,
which for graphite, is set at 284.0 eV.15,47 We assign the new
spectral feature at 284.1 eV to zigzag edge states. If there are
heteroatoms present these new states may not be observed
because the vacancy is quenched by the heteroatom.48 Taking
into account the diminishing sizes of the graphite upon
milling, the enhancement in the density of states at 284.1 eV
is thus suggestive of an increase development of the zigzag
edge states as a function of the milling time.
Of course nanographite sheet has also armchair structures
at the edges. Heggie et al.,49 have suggested that, unlike the
zigzag edges, the armchair edges prefer to hybridize into
sp structures with triple bond character. This suggestion is
supported by our spectra where the position of the peak at
285.9 eV is in excellent agreement with the p* resonance of
triple bonded linear hydrocarbons such as acetylene.50 The
0.4 eV shift suggests that during milling the carbon–carbon
binding energy increases by re-hybridizing towards sp from sp2,
and these carbon pairs can achieve shorter bond lengths closer
to the triple-bond of acetylene (B0.129 nm). This increases the
carbon 1s binding energy due to the p-bond localization and
implies a greater net charge per carbon atom around defects. In
other words, breaking the conjugation which results in electron
localization effect seems to be the mechanism capable of
up-shifting the G Raman peak from 1575 to 1583 cm�1 and
the NEXAFS p* band from 285.5 to 285.9 eV.
It is worth noting that in Raman spectroscopy, the appearance
of theD band is explained by a double resonance (DR) scattering
mechanism due to electron-optical phonon and electron-electron
interactions.33,51 According to Pimenta et al.,5 and Cancado
et al.,52 the DR process can only be fulfilled at an armchair edge
(strong D band), while for a zigzag edge, the resonance process is
forbidden (weak D band). Therefore, the NEXAFS peak at
284.1 eV and the Raman D band atB1345 cm�1 have dissimilar
origin since the electronic transition at B284.1 eV is related to
zigzag edges while the Raman D band is related to armchair
edges. Thus, the electronic and vibrational properties of nano-
graphite are sensitive to particle size and shape particularly in
relation to the peripheral regions.
Physical description of ball milled graphite as the in plane
crystallite size changes
Under prolonged milling, the shear and impact forces
continuously bend, fracture and displace the graphite sheets
relative to each other and the ABAB...stacking sequence is no
longer maintained. Transmission electron micrograph
comparison of the G0 and G30 samples shows that the
graphite particles became progressively disordered, curled
and corrugated along the edges (Fig. 4a and b). The phase
diagram of carbon53 indicates that the graphite can transform
into diamond at high pressures and temperatures. Below room
temperature, the pressure for phase transformation from
graphite to diamond is of several GPa,53 which is in the
same order of magnitude achieved by high energy ball-mills
(up to 4 GPa).54 Although it is difficult to estimate the impact
pressure and temperature fluctuations in graphite during
ball-milling in the current investigation, our data show that
the pressure induced locally changes the bond from sp2 to
sp- and sp3-like character.
Based on the current investigation, a model is proposed
incorporating the structural change to microcrystalline
graphite en route to nanocrystalline graphite via ball-milling
(Fig. 5). The in-plane fragmentation leads to a chemical
transition from aromatic to non-aromatic and the p states
become increasingly localised, i.e. some of the hexagonal rings
change gradually to chains. This result in a less rigid structure,
made of fragmentised graphite sheets bonded laterally via
in-plane sp links (Fig. 5a). The strain energy of sp bond can
be minimised by bond rotation.
When an atom is removed from a graphite sheet, the
structure undergoes bond reconstruction near the vacancy
defect and becomes sp3 like. The planar arrangement around
a vacancy would actually be unstable, and the atoms
surrounding the vacancy are more likely to be displaced
out-of-plane. This increases the probability of inter-sheet inter-
action via sp3 C–C bridges when defect sites from neighbouring
sheets are in close proximity (Fig. 5b) in accord with
calculations by Heggie et al.,49 and simulations by Telling
et al.55 The formation of inter-sheet sp3 bridges serves to relieve
the strain and thereby to stabilise nanographite as a new
metastable 3D structure. The development of the 3D structure
can explain why NEXAFS transitions at 287.5 and 288.6 eV are
intense whereas no detectable levels of heteroatoms are present
in the sub-surface in the nanocrystalline graphite.
This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 | 6689
Conclusions
It is concluded that high-energy milling under an inert
atmosphere damages the graphite sheets resulting in the
formation of structural defects, such as displaced atoms,
vacancies, and inter-plane sp3 bonds. The nanosized domains
contacting sp2-bonded hexagonal carbons are connected to
each other by a random network of sp and sp3 bonded
carbons. Under prolonged milling, the interaction of
these defects with each other and surrounding lattice causes
formation of new 3D structure.
There is considerable difference in the C 1s absorption
spectra at the surface and in the sub-surface of nanocrystalline
samples. The NEXAFS spectra acquired in fluorescence yield
mode demonstrate that nanocrystalline graphite develops new
p* states that give rise to a sharp peak at 284.1 eV from the
sample sub-surface. No such peak is observed from the surface
of the nanocrystalline graphite. The intensity of the new
p* electronic states is strongly influenced by the in-plane
crystallite size of the nanographite particles.
The localization of the p electrons in the nanosized domains
leads to a strengthening of the in plane CQC bonds, thus
leading to the 8 cm�1 up-shift of the G Raman peak and 0.4 eV
up-shift of the main graphite p* band from 285.5 to 285.9 eV.
Because of the close link between the electric conductivity and
the p/p* states, the discovery of this 0.4 eV energy shift may
open up possibilities to engineer the transport properties of
graphite, via the controlled concentration of defects. During
modification of peripheral regions such edges can give rise to a
variety of special practical applications for such nanographites
such as enhanced mechanical strength, electrical conductivity,
energy storage capacity for batteries and super-capacitors, if
they can be harnessed as such.
Acknowledgements
This work was produced as part of the activities of the ARC
Centre of Excellence for Functional Nanomaterials funded by
the Australian Research Council under the ARC Centre of
Excellence Program and by a University of Western Sydney
(UWS) internal grant (#17312) for 2009. The authors would
like to thank Dr L. Fan and Dr Y. Yang from NSRRC,
Hsinchu Taiwan for NEXAFS support and the travel grant
under Access to Major Research Facilities Programme
(ANSTO).
References
1 J. C. Charlier, X. Gonze and J. P. Michenaud, Carbon, 1994, 32,289–299.
2 J. Robertson, Adv. Phys., 1986, 35, 317–374.3 A. C. Ferrari, A. Libassi, B. K. Tanner, V. Stolojan, J. Yuan,L. M. Brown, S. E. Rodil, B. Kleinsorge and J. Robertson,Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 62, 11089–11103.
4 A. C. Ferrari and J. Robertson, Phys. Rev. B: Condens. MatterMater. Phys., 2000, 61, 14095–14107.
5 M. A. Pimenta, G. Dresselhaus, M. S. Dresselhaus,L. G. Cancado, A. Jorio and R. Saito, Phys. Chem. Chem. Phys.,2007, 9, 1276–1291.
6 A. C. Ferrari, Solid State Commun., 2007, 143, 47–57.
Fig. 4 Transmission Electron micrographs of graphite (a) Precursor
graphite, G0 and (b) after 30 h of milling, G30. The milled specimen is
broken into pieces with jagged edges implying a disappearance of the
initial layered structure of graphite along the c-axis but in some areas
still preserves the initial layered structure.
Fig. 5 (a) Schematic representation of the formation of sp bridge as a
result of a missing atom or vacancy. (b) Inter-planar bond-formation
via sp3 bonded carbons. The atoms at the upper edge and those at the
lower edge belong to different sheets.
6690 | Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 This journal is �c the Owner Societies 2010
7 D. A. Fischer, R. M.Wentzcovitch, R. G. Carr, A. Continenza andA. J. Freeman, Phys. Rev. B: Condens. Matter, 1991, 44,1427–1429.
8 S. R. Sails, D. J. Gardiner, M. Bowden, J. Savage and D. Rodway,Diamond Relat. Mater., 1996, 5, 589–591.
9 J. Stoehr, Springer Series in Surface Sciences, Vol. 25, NEXAFSSpectroscopy, 1992.
10 N. H. Tran, M. A. Wilson, A. S. Milev, J. R. Bartlett, R. N. Lamb,D. Martin and G. S. K. Kannangara, Adv. Colloid Interface Sci.,2009, 145, 23–41.
11 D. Arvanitis, U. Dobler, L. Wenzel, K. Baberschke and J. Stoehr,Journal de Physique, Colloque, 1986, C8C8/173–C178/178.
12 E. G. Rightor, A. P. Hitchcock, H. Ade, R. D. Leapman,S. G. Urquhart, A. P. Smith, G. Mitchell, D. Fischer, H. J. Shinand T. Warwick, J. Phys. Chem. B, 1997, 101, 1950–1960.
13 D. G. McCulloch and R. Byrdson, J. Phys.: Condens. Matter,1996, 8, 3835–3841.
14 R. Ahuja, P. A. Bruehwiler, J. M. Wills, B. Johansson,N. Maartensson and O. Eriksson, Phys. Rev. B: Condens. Matter,1996, 54, 14396–14404.
15 P. E. Batson, Phys. Rev. B: Condens. Matter, 1993, 48, 2608–2610.16 F. Atamny, J. Bloecker, B. Henschke, R. Schloegl, T. Schedel-
Niedrig, M. Keil and A. M. Bradshaw, J. Phys. Chem., 1992, 96,4522–4526.
17 N. Benchikh, F. Garrelie, C. Donnet, B. Bouchet-Fabre,K. Wolski, F. Rogemond, A. S. Loir and J. L. Subtil, Thin SolidFilms, 2005, 482, 287–292.
18 A. Gutierrez, J. Diaz and M. F. Lopez, Appl. Phys. A: Mater. Sci.Process., 1995, 61, 111–114.
19 C. Lenardi, M. Marino, E. Barborini, P. Piseri and P. Milani,European Physical Journal B: Condensed Matter Physics, 2005, 46,441–447.
20 M. Posternak, A. Baldereschi, A. J. Freeman, E. Wimmer andM. Weinert, Phys. Rev. Lett., 1983, 50, 761–764.
21 L. Ravagnan, G. Bongiorno, D. Bandiera, E. Salis, P. Piseri,P. Milani, C. Lenardi, M. Coreno, M. de Simone andK. C. Prince, Carbon, 2006, 44, 1518–1524.
22 B. Reihl, J. K. Gimzewski, J. M. Nicholls and E. Tosatti,Phys. Rev. B: Condens. Matter, 1986, 33, 5770–5773.
23 V. A. Coleman, R. Knut, O. Karis, H. Grennberg, U. Jansson,R. Quinlan, B. C. Holloway, B. Sanyal and O. Eriksson, J. Phys.D: Appl. Phys., 2008, 41, 062001/062001–062001/062004.
24 A. Milev, N. Tran, G. S. K. Kannangara and M. Wilson, Sci.Technol. Adv. Mater., 2006, 7, 834–838.
25 A. Milev, M. Wilson, G. S. K. Kannangara and N. Tran, Mater.Chem. Phys., 2008, 111, 346–350.
26 D. E. Smeulders, A. S. Milev, G. S. K. Kannangara andM. A. Wilson, J. Mater. Sci., 2005, 40, 655–662.
27 P. Tan, Y. Deng, Q. Zhao and W. Cheng, Appl. Phys. Lett., 1999,74, 1818–1820.
28 L. G. Cancado, M. A. Pimenta, B. R. A. Neves, G. Medeiros-Ribeiro, T. Enoki, Y. Kobayashi, K. Takai, K.-i. Fukui,M. S. Dresselhaus, R. Saito and A. Jorio, Phys. Rev. Lett., 2004,93, 047403/047401–047403/047404.
29 B. Watts, L. Thomsen and P. C. Dastoor, J. Electron Spectrosc.Relat. Phenom., 2006, 151, 105–120.
30 M. Newville, P. Livins, Y. Yacoby, J. J. Rehr and E. A. Stern,Phys. Rev. B: Condens. Matter, 1993, 47, 14126–14131.
31 B. Ravel and M. Newville, J. Synchrotron Radiat., 2005, 12,537–541.
32 L. M. Malard, M. A. Pimenta, G. Dresselhaus andM. S. Dresselhaus, Phys. Rep., 2009, 473, 51–87.
33 C. Thomsen and S. Reich, Phys. Rev. Lett., 2000, 85, 5214–5217.34 R. Saito, A. Jorio, A. G. Souza Filho, G. Dresselhaus,
M. S. Dresselhaus and M. A. Pimenta, Phys. Rev. Lett., 2002,88, 027401/027401–027401/027404.
35 M. Nakamizo, R. Kammereck and J. P. L. Walker, Carbon, 1974,12, 259–267.
36 K. Nakamura, M. Fujitsuka and M. Kitajima, Chem. Phys. Lett.,1990, 172, 205–208.
37 F. Tuinstra and J. L. Koenig, J. Chem. Phys., 1970, 53, 1126–1130.38 H. Wakayama, J. Mizuno, Y. Fukushima, K. Nagano,
T. Fukunaga and U. Mizutani, Carbon, 1999, 37, 947–952.39 L. G. Cancado, K. Takai, T. Enoki, M. Endo, Y. A. Kim,
H. Mizusaki, A. Jorio, L. N. Coelho, R. Magalhaes-Paniagoand M. A. Pimenta, Appl. Phys. Lett., 2006, 88, 163106/163101–163106/163103.
40 A. P. Hitchcock, D. C. Newbury, I. Ishii, J. Stohr, J. A. Horsley,R. D. Redwing, A. L. Johnson and F. Sette, J. Chem. Phys., 1986,85, 4849–4862.
41 C. Ziethen, O. Schmidt, G. K. L. Marx, G. Schonhense,R. Fromter, J. Gilles, J. Kirschner, C. M. Schneider andO. Groning, J. Electron Spectrosc. Relat. Phenom., 2000, 107,261–271.
42 F. L. Coffman, R. Cao, P. A. Pianetta, S. Kapoor, M. Kelly andL. J. Terminello, Appl. Phys. Lett., 1996, 69, 568–570.
43 K. Nakada, M. Fujita, G. Dresselhaus and M. S. Dresselhaus,Phys. Rev. B: Condens. Matter, 1996, 54, 17954–17961.
44 K. Nakada, M. Igami, K. Wakabayashi andM. Fujita,Mol. Cryst.Liq. Cryst., 1998, 310, 225–230.
45 M. Fujita, M. Igami, K. Wakabayashi and K. Nakada,Mol. Cryst.Liq. Cryst., 1998, 310, 173–178.
46 M. Fujita, K. Wakabayashi, K. Nakada and K. Kusakabe,J. Phys. Soc. Jpn., 1996, 65, 1920–1923.
47 E. J. Mele and J. J. Ritsko, Phys. Rev. Lett., 1979, 43, 68–71.48 M. A. H. Vozmediano, F. Guinea and M. P. Lopez-Sancho,
J. Phys. Chem. Solids, 2006, 67, 562–566.49 I. Suarez-Martinez, G. Savini, G. Haffenden, J. M. Campanera
and M. I. Heggie, Phys. Status Solidi C, 2007, 4, 2958–2962.50 A. P. Hitchcock and C. E. Brion, J. Electron Spectrosc. Relat.
Phenom., 1977, 10, 317–330.51 S. Piscanec, M. Lazzeri, F. Mauri, A. C. Ferrari and J. Robertson,
Phys. Rev. Lett., 2004, 93, 185503–185504.52 L. G. Cancado, M. A. Pimenta, B. R. A. Neves, M. S. S. Dantas
and A. Jorio, Phys. Rev. Lett., 2004, 93, 247401/247401–247401/247404.
53 F. P. Bundy, J. Chem. Phys., 1963, 38, 618–630.54 D. R. Maurice and T. H. Courtney, Metall. Trans. A, 1990, 21,
289–303.55 R. H. Telling, C. P. Ewels, A. A. El-Barbary and M. I. Heggie,
Nat. Mater., 2003, 2, 333–337.
This journal is �c the Owner Societies 2010 Phys. Chem. Chem. Phys., 2010, 12, 6685–6691 | 6691