Preparation of gallium sulfide nanosheets by liquid exfoliation and their application as
hydrogen evolution catalysts
Andrew Harvey,1,2 Claudia Backes,1,2 Zahra Gholamvand,1,2 Damien Hanlon,1,2 David
McAteer,1,2 Hannah C. Nerl,1,2,3 Eva McGuire,1,2,3 Andrés Seral-Ascaso,1,2,3 Quentin M.
Ramasse,4 Niall McEvoy,1,3 Sinéad Winters,1,3 Nina C. Berner,1,3 David McCloskey,1,2 John
F. Donegan,1,2 Georg S. Duesberg,1,3 Valeria Nicolosi1,2,3 and Jonathan N. Coleman1,2*
1
CRANN & AMBER, Trinity College Dublin, Dublin 2, Ireland
2
School of Physics, Trinity College Dublin, Dublin 2, Ireland
3
School of Chemistry, Trinity College Dublin, Dublin 2, Ireland
4
SuperSTEM Laboratory, STFC Daresbury Campus, Daresbury, WA4 4AD, United Kingdom
*colemaj@tcd.ie
Abstract
Here we demonstrate the production of large quantities of gallium sulfide (GaS) nanosheets by
liquid exfoliation of layered GaS powder. The exfoliation was achieved by sonication of the
powder in suitable solvents. The variation of dispersed concentration with solvent was
consistent with classical solution thermodynamics and showed successful solvents to be those
with Hildebrand solubility parameters close to 21.5 MPa1/2. In this way, nanosheets could be
produced at concentrations of up to ~0.2 mg/ml with lateral sizes and thicknesses of 50-1000
nm and 3-80 layers, respectively. The nanosheets appeared to be relatively defect free although
oxygen was observed in the vicinity of the edges. Using controlled centrifugation techniques,
it was possible to prepare dispersions containing size-selected nanosheets. Spectroscopic
measurements showed the optical properties of the dispersions to vary strongly with nanosheet
size, allowing the elucidation of spectroscopic metrics for in-situ estimation of nanosheet size
and thickness. These techniques allow the production of nanosheets with controlled sizes which
will be important for certain applications. To demonstrate this, we prepared films of GaS
nanosheets of three different sizes for use as hydrogen evolution electrocatalysts. We found a
clear correlation between performance and size showing small nanosheets to be more effective.
This is consistent with the catalytically active sites residing on the nanosheet edges.
1
Introduction
Over the last few years, liquid phase exfoliation (LPE) has become an increasingly
important technique for the production of two-dimensional nanomaterials.1-3 This method
involves the delamination of layered crystals, usually by exposure to ultra sonication4, 5 or high
shear rates,6-9 to form large quantities of two-dimensional nanosheets. These are then stabilized
against aggregation through interaction with appropriate solvents,1, 10-20 or ionic liquids21 or by
coating with surfactants4, 22-32 or adsorbed polymer chains.33-36 The resultant nanosheets range
in lateral size from <100 nm to >2000 nm, depending on the material in question.37, 38 The
nanosheet thickness tends to be broadly distributed between approximately one and 10 layers
with monolayer contents of tens of percent achievable.6, 37 That nanosheets can be stably
dispersed in liquids greatly facilitates further processing. For example nanosheets can be sizeselected by controlled centrifugation23,
37-41
or chemically modified via functionalization
protocols.42-44 In addition, the dispersions can easily be formed into thin films or mixed with
other materials to form composites.4, 6 This has enabled the use of liquid exfoliated nanosheets
in a range of applications from electrocatalysis,37, 45, 46 to composite reinforcement47-51 to inkjetprinted devices.52, 53 Additional advantages of this technique are that it is low-cost, relatively
straightforward to implement and can easily be scaled up to produce very large quantities of
nanosheets.
Liquid phase exfoliation was first used to produce graphene from graphite in 2008.5, 54
Subsequent work has seen the scale up and commercialization of this method as a graphene
production technique.6 Possibly the greatest strength of this technique is its versatility. Liquid
phase exfoliation has been used to produce nanosheets from a wide range of layered crystals
beyond graphite including; boron nitride; transition metal dichalcogenides, from MoS2 to
WTe2; MoO3 and black phosphorus.2, 4, 5, 11, 23, 39, 40, 50, 55-58 However, because there are hundreds
of different types of layered crystals,2 this material set represents only the tip of the iceberg of
materials that could be exfoliated by LPE. We believe it is critically important to use LPE to
exfoliate yet untested layered crystals and to produce new types of nanosheets. These new
materials will be of interest both for basic studies and because of their potential for use in a
range of new applications.
The family of III-VI layered semiconductors is a good example of a largely untapped source
of two-dimensional materials. These materials generally exist in the form MX where M=Ga,
In and X=S, Se, Te (see figure 1A) although other stoichiometries also exist.59 These materials
2
are wide bandgap semiconductors and are of interest for a range of applications from
electrochemistry60 to optoelectronics,61-63 gas sensing64 and nonlinear optics.65 Over the last
thirty years a number of papers have studied the properties of such layered crystals with
common examples being GaSe, InTe and the related compound In2Se3.65-67 However, work on
exfoliation has only started recently with a number of papers reporting production of few-layer
InSe and GaSe by mechanical cleavage or chemical exfoliation as well as vapor phase growth
of thin layers.59, 68-71 Such studies have found these exfoliated materials to be of interest for
applications such as photodetectors and in non-linear optics.59, 68-72
A typical representative example of this family is gallium sulfide, GaS. Although found in
a variety of structures such as nanobelts and tubes,73 this material is most commonly
encountered as a layered crystal and is particularly attractive due to its relatively low cost.
Although a number of papers have studied the layered form of GaS,62,
63, 74
work on its
exfoliation to give nanosheets is in the very early stages. Exfoliation has up to now only been
achieved by micromechanical cleavage61, 64, 75 which suffers from low throughput and can only
produce material quantities suitable for fundamental studies. However, it is clear that the
properties of exfoliated gallium sulfide are interesting and differ from the bulk form.60, 61, 64, 75
For example, exfoliated nanosheets of GaS have been used to fabricate sensitive
photodetectors.61, 64 In addition, it is likely that GaS is useful in applications also beyond
(opto)electronics. Gallium sulfide produced by atomic layer deposition has been combined
with carbon nanotubes to produce high performance anodes in Li ion batteries.60 However, for
such materials to be competitive in applications such as battery electrodes, large quantities
would be needed. Applications such as this highlight the need to develop a scalable method to
produce nanosheets of gallium sulfide and related materials.
Herein we report that layered III-VI semiconductors can be exfoliated in solvents by
bath sonication. The resultant dispersions contain nanosheets which are of high quality and
appear to be defect-free except for an increased oxygen content at the vicinity of edges. In
addition, spectroscopic properties strongly vary as a function of nanosheet size allowing us to
establish quantitative metrics to determine mean length and thickness spectroscopically. This
facilitates the production of dispersions with well-defined nanosheet sizes and so specific
properties. These were subsequently used to fabricate GaS electrodes for electrocatalysis of
hydrogen production. Strong size effects are found with smaller nanosheets performing better.
3
Results and Discussion
Evidence of exfoliation and basic characterization
In this work, we study the exfoliation of layered GaS in a number of solvents. We chose
GaS as a representative member of the family of layered III-VI semiconductors (structure see
Fig. 1A) in part due to its commercial availability. To perform the exfoliation, we added GaS
powder to a variety of solvents and agitated using an ultrasonic bath for a fixed period. The
dispersions were then centrifuged to remove unexfoliated material. In most cases, we obtained
pale yellow colored liquids such as the one shown in figure 1B.
In order to confirm the exfoliation of the layered powder to 2D nanosheets, we
performed low-resolution transmission electron microscopy (TEM) imaging (Fig. 1C). These
measurements showed all dispersions tested to contain large quantities of electron-transparent
2D nanosheets. In addition, bright field (Fig. 1D) and high angle annular dark field dark
(HAADF) (Fig. 1E) scanning transmission electron microscopy (STEM) imaging confirmed
the 2D crystal lattice to be intact.
Typically, the dispersed concentration of stable, exfoliated nanomaterials has been
estimated from measurements of the optical extinction at a given wavelength4-6,
39
(the
extinction, Ext, is defined via the optical transmittance: T 10 Ext ). The measured extinction
spectrum of a typical GaS dispersion is plotted as the black curve in figure 1F (solvent
isopropanol, initial GaS concentration ci=45 g/L, sonication time ts=6 h, centrifugation time
tcf=180 min at 2.5 krpm equivalent to 665 g). This plot shows a near monotonic increase with
decreasing wavelength with no features of note, bar a small peak at 315 nm. As with dispersions
of other 2D materials, such curves are relatively featureless because extinction spectra of
dispersed nano-objects contain a significant contribution from scattering (Sca) in addition to
the actual absorbance (Abs) of the material (NB: extinction, absorbance and scattering are
related by Ext ( ) Abs( ) Sca( ) ).
The absorbance and scattering components can be differentiated from the overall extinction
spectra using an integrating sphere37 as shown in figure 1F. The absorbance is very low in the
high wavelength regime, becoming appreciable only for <400 nm consistent with the
semiconducting nature of GaS. This data does indeed show a significant scattering component,
particularly in the non-resonant regime (>400 nm). The shape of the spectra will be discussed
in more detail below. Using extinction spectra to estimate dispersed concentration has been
4
complicated by the recent realization that both absorbance and scattering coefficients are
generally dependent on nanosheet size.37, 76, 77 The wavelength used has to be chosen with care
and requires an understanding of the size-dependent extinction coefficients as discussed below.
In the case of GaS, we found the extinction coefficient at 365 nm to be relatively nanosheet
size independent with a value of 365nm 3654 L g-1 m-1. This allows the estimation of the
dispersed nanosheet concentration, C, using Ext365nm 365nmCl (l is the cell length).
To identify the most appropriate solvents for the liquid exfoliation of GaS and to investigate
the exfoliation/stabilization mechanism, we have sonicated and centrifuged the powder in 15
solvents to produce nanosheets under identical processing conditions (see methods). We
measured optical extinction spectra for each dispersion, using Ext365nm as a measure of the
dispersed concentration. In this way we found large variations in dispersed concentration
among the solvents studied, with the highest nanosheet contents found for amide solvents such
as N-methyl-2-pyrrolidone (NMP) and N-cyclohexyl-2-pyrrolidone (CHP). To understand the
exfoliation/stabilization mechanism, we plot Ext365nm/l versus the solvent Hildebrand solubility
parameter δS in figure 1G. The concentrations of exfoliated GaS, estimated using 365nm 3654
L g-1 m-1, are shown on the right axis. We find a clear and well defined peak centered at around
21.5 MPa1/2 very similar to studies on other 2D nanomaterials.4, 5, 10, 39
According to classical solution thermodynamics, in the simplest case,55, 78 the saturated
concentration, C, of 2D solutes, such as the GaS nanosheets studied here is given by
v
C exp NS ( S NS )2
3kT
(1)
where NS represents the Hildebrand parameter of the 2-dimensional solute, vNS represents the
unit volume of the solute while the factor of three stems from the solute dimensionality.78 The
dashed line in figure 1G (inset) is a fit to equation (1) and shows very good agreement with the
experimental data, further confirming the validity of solution thermodynamics in the case of
nanomaterial dispersions. The data implies the Hildebrand solubility parameter of GaS
nanosheets to be ~21.5 MPa1/2, close to values of 23, ~22.5, 22, 21 and reported recently for
Graphene,10 BN,4 MoS2 (ref 55) and MoO3 (ref 39), respectively. We note that analysis in terms
of Hildebrand parameters has been shown to be equivalent to analysis in terms of surface
energies.78 Essentially, this means successful solvents are those with surface tensions close to
40 mJ/m2. Taken together, this means that, as has been observed previously for the materials
5
mentioned above, liquid phase exfoliation of GaS is most favorable in solvents whose
interaction with the nanosheets is such that the energy of exfoliation is minimized.5
Even though we found empirically that NMP and CHP are the most suitable solvents
in terms of dispersed concentration, we have used 2-propanol (IPA) for the remainder of the
study (these solvents are marked in red in figure 1G). Although the concentration of GaS which
can be exfoliated in IPA is only a quarter that achievable in NMP, use of IPA brings significant
advantages in processing and analysis due to its low toxicity and low boiling point. We fully
optimized the exfoliation conditions in terms of centrifugation rate and time, sonication time
and initial GaS concentration as presented in the SI figure S1. We found the following
parameters to be optimized for the production of standard dispersions of exfoliated GaS in IPA:
initial GaS concentration Ci = 45 g/L, sonication time ts = 6 h, centrifugation rate f = 2.5 krpm
(equivalent to 665 g), centrifugation time tCF = 180 min. Under these circumstances, we could
produce dispersions with concentrations of roughly C=0.22 g/L of dispersed nanosheets.
It is important to confirm that the exfoliated GaS is pristine and defect free. For this
purpose, samples were subjected to Raman and X-ray photoelectron spectroscopies (XPS).
Figure 1H shows the Raman spectrum (excitation wavelength 532 nm, mean of 20 spectra) of
a filtered film of the standard dispersion. The films are homogenous in appearance as evidenced
by scanning electron microscopy (SEM) (inset in figure 1H). The characteristic lattice
vibrations of GaS are detected at 195, 300 and 370 cm-1 as assigned in the figure.61, 74, 75 Only
minor contributions from other materials are observed such as the ν1(A1) mode of the GaS4
molecular unit as typically found in Ga2S3 which we see at ~240 cm-1.79, 80 The fitted XPS core
level spectra (Ga 3d in Fig 1I and S 2p in Fig. 1J) further confirm that the exfoliated material
is GaS with minor contributions from GaxOy. We note that such oxide species were also found
in the starting powder (Fig. S2). In addition, all XPS peaks measured for the starting powder
are significantly broadened in terms of full width at half maximum of the fit components,
suggesting a lower degree of order and/or purity in the starting powder compared to the
exfoliated material. This shows that the sonication and centrifugation partially purifies the
starting material by removal of unwanted, components such as non-layered impurities. This
can be illustrated by the presence of significant amounts of Ga2S3 in the starting powder as
shown by X-ray diffraction (XRD) (Fig. S3) and Raman spectroscopy (Fig. S4). Since Ga2S3
is not layered, it is not exfoliated and therefore largely removed during centrifugation.
6
Electron energy loss spectroscopy
Since XPS shows that oxides are present in both GaS powder and exfoliated nanosheets
(Fig. 1I and S2) and virtually nothing is known about the long term stability of GaS in the
exfoliated state, it is important to track where oxides species reside. This is particularly crucial
in light of recent investigations on exfoliated black phosphorus nanosheets which have shown
significant degradation under exposure to ambient conditions.81-83 To gain insights into
potential oxidation of GaS, we have analyzed liquid exfoliated nanosheets by STEM imaging
and electron energy loss spectroscopy (EELS). A representative STEM image of a nanosheet
at the edge region is shown in figure 3A. EEL spectra were recorded from the same sample
region to form a map, with the intensity of each pixel in the map (ranging from black to
yellow/white) corresponding to the integrated intensity of the oxygen K-edge in the pixel
location. From this data, an oxygen content map could be constructed from the same sample
region (Fig. 3B) as shown in STEM image (Fig.3A). Hence, the map is color-coded and shows
increasing oxygen content from black/blue (no/low oxygen) to green to red to yellow to white.
It is clear that the oxides reside mostly near the edges of the nanosheets and at step edges
throughout the nanosheets. The EEL spectra corresponding to regions with different oxygen
content are displayed in figure 3C.
Size selection
A great advantage of liquid exfoliation is not only that nanosheets are processable from
liquids, but that they can also be size-selected using well established techniques by controlled
centrifugation.37, 38, 40 This is extremely important considering that potential applications often
require control over both thickness and lateral dimensions. To demonstrate this, we have
performed size-selection by controlled centrifugation with increasing centrifugation velocity
in consecutive steps (see methods). This enables the production of nanosheet dispersions with
varying mean sizes from the same stock dispersion in large quantities. By this procedure, five
different sizes were produced and subjected to detailed characterization. For the main
manuscript we focused on three different dispersions. Further data can be found in the
supporting information. Displayed in figure 3A-C are TEM length histograms and
representative images of three different size-selected GaS dispersions in IPA (data for all five
sizes is given in the SI Fig. S5). The mean lateral dimensions were determined from the
statistical TEM analysis yielding TEM mean lengths <L> ranging from 130 - 405 nm.
7
Since it has previously been reported that optical extinction, absorbance and scattering
spectra change as a function of size, the optical response was measured for each dispersion.37,
40
Any systematic spectral changes can be very valuable as – once calibrated – they can be used
to establish metrics to potentially quantify both lateral dimensions and thickness of the liquid
exfoliated nanosheets.37 Optical extinction (normalized to the local minimum), absorbance
(normalized to the local minimum) and scattering (normalized to the local minimum in
extinction) spectra of the size-selected GaS are displayed in figure 3D-F (plots showing the unnormalized coefficient spectra can be found in the SI, figure S12). All exhibit well defined and
systematic changes as a function of size which will be analyzed below. Of greatest interest are
the absorbance spectra (figure 3E), particularly in comparison to the extinction spectra (figure
3D). As mentioned above, the extinction spectra are almost featureless. However, GaS is
known to display excitonic transitions at ~410 nm (A-exciton, direct transition at the Γ-point)
and ~315 nm (B-exciton, transition at the M-point).84 While the B-exciton is discernible in the
extinction spectra as the local maximum, the A-excitonic transitions are invisible due to
masking by the scattering background. However, in the absorbance spectra, both excitons can
be clearly resolved. In addition, very weak features at higher wavelength are discernible in the
absorbance spectra (inset figure 3E) which may stem from defects.85 However, these are widely
invariant with size so that we suggest they are not edge or exfoliation induced.
In addition to quantifying lateral dimensions by statistical TEM, we have determined
the thickness of the exfoliated GaS nanosheets using atomic force microscopy (AFM) after
deposition onto Si/SiO2 wafers (see methods). We find reasonably thin nanosheets similar in
appearance to the TEM images (figure 3G inset). Since the direct measurement of the number
of layers, N, in the case of liquid exfoliated nanomaterials is complicated by both solvent and
nanosheet contributing to the apparent AFM thickness, we have used previously elaborated
step height analysis6, 37, 40 to convert the thickness to the number of layers (SI figure S6). The
resultant N histogram of the GaS sample with mean length of 130 nm is displayed in figure 3G
showing that the nanosheets are reasonably well exfoliated with mean N of 10 layers.
Histograms and representative images of the other sizes are shown in the SI figure S7.
However, before being certain about conversion of apparent thickness to number of
layers, it needs to be ensured that only individually deposited nanosheets are taken into account.
To test this, we plot the mean AFM length of the counted nanosheets as a function of the TEM
length in figure S8A, finding very good agreement. Furthermore, we can use our previous
8
knowledge on other liquid exfoliated 2D materials that has shown a square root dependence of
N as function of nanosheet area.37, 39 As demonstrated by figure 3H (and Fig. S8B), the same
behavior is observed in the case of GaS strongly supporting the accuracy of the AFM number
of layer determination.
We furthermore note that the structural integrity of the size-selected nanosheets was
confirmed by Raman and XPS (see SI figure S9-11) which gave similar results for both nonsize-selected standard dispersions and the size selected nanosheets described in figure 3. This
shows the nanosheet spectroscopic properties other than the optical response to be generally
independent of size in the regime our nanosheets are produced. For example, we do not observe
layer-number induced shifts in the GaS Raman spectra as previously reported.75 However, this
is not surprising, as these the Raman spectrum of GaS is only sensitive to layer numbers of <5
and the majority of our GaS nanosheets have thicknesses >5 layers.
Spectroscopic metrics to determine size and concentration
The precise quantification of both N and L as described above is important, as it
provides the foundation to establish spectroscopic metrics based on extinction, absorbance and
scattering spectra. In the following, we show a number of these metrics which can be used to
determine information regarding GaS nanosheet size and thickness from optical measurements.
As mentioned above, the scattering spectra are sensitive to the nanosheet lateral size.
Previous work has shown that, in the high wavelength regime, where the absorbance is
negligible, the scattering exponent scales as a power law with wavelength: n , where n
is the size-dependent scattering exponent which can also be used as a potential metric to
quantify nanosheet length.4, 37 To test whether such a metric can be established for liquidexfoliated GaS, we plot the long wavelength scattering exponent, n, as a function of mean
nanosheet length (as measured by TEM) in figure 4A. Importantly, since absorbance of GaS is
negligible at high wavelength, we obtain the same result extracting n from either extinction or
scattering spectra, implying that measurement in an integrating sphere is not essential to
determine <L>. We find a roughly linear relationship (valid in this size range only) between
<L> and n which allows us to use the scattering coefficient to determine the mean length of the
GaS nanosheets in the dispersion according to equation (2).
L( m) 0.67 0.14n
(2)
9
Recent work has also shown that for nanosheet dispersions, the shape of the absorption
spectra is sensitive to nanosheet lateral size due to edge effects.37 These spectral changes in
liquid-exfoliated 2D nanomaterials can be expressed as ratios of absorbance (or extinction) at
two wavelengths. These intensity ratios contain nanosheet length information due to
differences in electronic properties at nanosheet edge and center.37 This effect can clearly be
seen in the case of GaS when plotting the ratio of absorbance at the A-exciton / local minimum
at 290 nm ( A420nm / A290nm ) as a function of <L> (Fig. 4B). This plot shows a well-defined
relationship allowing us to link mean nanosheet length to A420nm / A290nm according to equation
(3).
L(μm) 0.93( A420nm / A290 nm )0.44
(3)
We note that while this shows the validity of the approach, <L> determination from the
scattering exponent in optical extinction spectra is more straight-forward, especially because
the integrating sphere is not required.
While the determination of <L> is useful, for most applications, it is critical to control
concentration/mass of the nanosheets. An accurate determination of the extinction/absorbance
coefficient is therefore required. Since the spectra change as function of size, extinction and
absorbance coefficients are expected to changes as well (as observed for MoS2)37. We therefore
also analyze extinction ε, absorbance α and scattering σ coefficient spectra (SI Fig. S12). The
GaS nanosheet concentration in each case was determined by filtration and weighing.
Depending on the wavelength, the coefficients vary significantly as a function of nanosheet
size (Fig S12-13). However, we observe only minor changes in the extinction coefficient at
365 nm over a wide range of size from 100- 300 nm (Fig. 4C) making this an ideal wavelength
for estimation of the concentration from the extinction spectra using 365nm 3465 Lg-1m-1 .
Since mean nanosheet lengths from a single centrifugation step are typically below 300 nm,
we have used the approximate extinction coefficient for our process optimization (as described
above). If the nanosheet size is >300 nm (where <L> can be determined by equation 2 or 3),
then the extinction coefficient can be found from an empirical fit of the data:
365nm (Lg-1m-1 ) 3465 21 e L(m) /0.128
(4)
In addition to determination of length and concentration from the optical spectra, it would also
be useful to have a metric to assess the mean number of layers. In case of MoS2, the
10
energy/wavelength of the A-exciton provided such a metric due to confinement effects.37 In
the case of GaS, we correlate the mean nanosheet thickness (as measured from AFM) with the
peak position of the B-exciton from the absorbance spectra, λB. We used the B-exciton rather
than the A-exciton in this case, as the intensity of the A-exciton is comparatively weak for all
GaS sizes. Shown in figure 4D is a plot of λB versus N with the correlation being clear. By
applying an empirical fit, we find the mean number of layers of the liquid exfoliated GaS
nanosheets can be determined from the peak position of the B-exciton in the absorbance
spectra, λB, according to equation (5)
N 2.24 1026 eB ( nm)/11.5
(5)
A similar metric from the extinction spectra is presented in the SI (Fig. S14). However, we
note that the position of the B-exciton does not tend to a constant value which might be
associated with bulk for nanosheets thicker than ~10-20 layers as might be expected. The
reasons for this are unclear and will be the subject of a future study.
The presented quantitative in situ spectroscopic metrics to determine <L>, <N> and
concentration of liquid exfoliated GaS underline the strength of LPE, as these will be extremely
useful to prepare dispersions with known dimensions and concentrations to test in applications.
Hydrogen evolution catalysis
The size metrics and precise concentration control described above facilitates the use
of liquid-exfoliated GaS nanosheets in applications where lateral nanosheet size is important.
Such an application may be the hydrogen evolution reaction (HER). It is known that edge sulfur
atoms in MoS2 nanosheets are involved in the electrocatalysis of hydrogen production.86, 87 We
hypothesized that GaS may also have catalytically active edge sites. If this were to be the case,
liquid exfoliated GaS nanosheets should catalyze H2 production with the rate of production
increasing as the nanosheet size decreases. To test whether this was the case, we size-selected
a GaS stock dispersion according to the procedure used throughout this manuscript to give
three size selected dispersions. The mean nanosheet lengths were determined from the
scattering exponent to be 450, 280 and 180 nm, respectively. The dispersions were vacuumfiltered to give thin films (~0.65 mg/cm2) which were transferred onto pyrolytic carbon (PyC)
coated Si/SiO2 substrates (Fig. S15) and characterized for hydrogen evolution catalysis in a
three electrode electrochemical work station (see methods).
11
The current density versus potential relative to reversible hydrogen electrode
(polarization curves) for small, medium and large nanosheets are shown in figure 5A. Clear
size-dependence on hydrogen production is observed with small nanosheets performing much
better than larger ones. This suggests that catalytic sites active in GaS also reside at edges as
found for other 2D materials.86, 87 For example, as the nanosheet size is reduced from 450 to
280 to 180 nm, the onset potential (the potential where J=1 mA/cm2) decreases from 0.62 to
0.53 to 0.48 V while the current density at 0.6 V increases from 0.6 to 6 to 22 mA/cm2 (see
figure S16 and table S2). We note that, while we observe an increased content of oxides in the
vicinity of the edges according to EELS (Fig. 2), this does not necessarily affect the catalytic
reaction, as it was proposed that only the very outer rim of atoms are catalytically active.86, 87
In addition, we do not see a decrease in catalytic performance after repeated measurement of
the polarization curves (Fig. S17) and do not observe a significantly increased oxide content
according to XPS after this procedure (Fig. S18).
Tafel plots of overpotential versus current density (figure 5B) show the expected
behavior in the linear regime with Tafel slopes between 106 and 85 mV/dec. These slopes are
smaller than the values of ~120 mV/dec usually found for the 2H-polytype45, 88 of MoS2 and
indicate the rate limiting step may be different in GaS nanosheets. However, the exchange
current densities were <2.510-6 mA/cm2 for all three GaS samples, much smaller than is
usually found for 2H-MoS2. In addition, the onset potential was at least 480 mV for the samples
studied here, considerably larger than the values of ~300 mV usually found for 2H MoS2.45, 88
Thus, at first glance GaS might appear to be a poor catalyst for hydrogen evolution in
contrast to MoS2 which is commonly accepted as a promising material. This is due to its
relatively high onset potential and low exchange current densities. However, the lower Tafel
slope may offer some compensation for these deficiencies. To test this, we produced 2H MoS2
nanosheets by liquid exfoliation and size selection using previously described procedures.37
We produced a film with near identical mass (~0.7 mg/cm2) to the GaS samples and slightly
smaller nanosheet size (L=120 nm). Polarization curves for LPE GaS and 2H-MoS2 are
compared in Figure 5C. It is apparent that, although GaS requires large overpotentials to initiate
the hydrogen evolution, its smaller Tafel slope causes a more rapid increase of the current with
potential. In fact, both materials reach the same current density at 0.6 V, above which GaS
dominates. This suggests that GaS can indeed be regarded as potentially attractive hydrogen
12
evolution catalyst, especially if problems associated with the low conductivity could be
overcome for example by addition of nanotubes.89
Conclusion
In conclusion, we have demonstrated that layered III-VI semiconductors such as GaS
can be exfoliated in appropriate solvents by sonication. Dispersibility can be well described in
the framework of solution thermodynamics. While a number of solvents can be used to
exfoliate GaS, we have focused on 2-propanol due to its low toxicity and boiling point. We
have fully optimized sonication and centrifugation conditions to yield stable dispersions with
typical concentrations of 0.2 g/L. Raman, XPS and high resolution STEM imaging show the
exfoliated nanosheets to be widely structurally perfect and free of defects except for a higher
oxygen content in the vicinity of edges as shown by EELS.
We utilized one advantage of liquid exfoliation techniques and performed size-selection
by controlled centrifugation. This allowed us to produce liquid-exfoliated GaS nanosheet
dispersions with mean lateral dimensions ranging from ~ 100 nm to > 400 nm and mean number
of layers from 10-40 as quantified by statistical TEM and AFM analysis. Importantly, we found
optical extinction, absorbance and scattering spectra to vary strongly as a function of size and
thickness. This enabled us to establish quantitative spectroscopic metrics to accurately
determine mean length, thickness and concentration of the dispersion.
We subsequently used these metrics to produce nanosheet dispersions with known
concentration and lateral dimensions that were used as electrodes and tested for hydrogen
evolution catalysis. We find a clear size effect with smaller nanosheets performing much better
than larger nanosheets. This suggests that catalytic sites are located at the edges of the
nanosheets. Even though onset potentials are typically larger than for MoS2 of similar lateral
dimensions, the lower Tafel slopes make GaS an attractive material for hydrogen evolution, in
particular in combination with other catalysts that exhibit lower onset potential.
We believe these results are general and that it should be possible to transfer these
procedures almost exactly for the exfoliation of other III-VI layered semiconductors such as
InS and GaSe. This will open up this whole family of 2D materials for exploitation. In addition,
we think it is important to note that this paper describes using LPE to exfoliate a member (GaS)
of yet another family of layered compounds (III-VI layered semiconductors). This underlines
13
the generality and versatility of this method. We believe that many as yet untested layered
materials will be exfoliated using such procedures.
14
Materials and Methods
Materials
Gallium sulfide powder (99.999% Ga-S-05-P) was purchased from American
Elements. All solvents were purchased from Sigma Aldrich at the highest available purity with
all being >99%.
Production of GaS nanosheets
Gallium sulfide powder was sonicated in a solvent using an ultrasonic bath (P30 H
Ultrasonic from Fischer scientific). The sonication was performed with an amplitude of 100%
and a frequency of 37 kHz in 50 mL plastic centrifuge tubes. The water in the sonic bath was
cooled by a water cooling system to 20-30 0C (depending on length of sonication) enabled by
cold water being pumped through piping which was wrapped around the interior of the bath.
Once sonicated, the dispersion was centrifuged in a Hettich Mikro 220R centrifuge with a
fixed-angle rotor 1060 (NB: For this centrifuge, rpm are related to g-force via RCF=106.4 f2
where f is the rotation rate in krpm). The top 60% was then taken from the centrifuged
dispersion (supernatant) for analysis and the sediment was discarded. For the solvent screening,
1 g/L of GaS was sonicated in 20 mL of each solvent for 6 hours and centrifuged for 180 min
at 2.5 krpm. The supernatant was decanted and the absorption and extinction were measured
in a 4 mm path length cuvette using a Perkin Elmer Lambda 650 spectrometer (details see
below). The following solvents were used: Isopropanol (δS=23.6 MPa1/2), N-methyl-2pyrrolidone (δS=23 MPa1/2), methanol (δS=29.6 MPa1/2), chloroform (δS=19 MPa1/2), Ncyclohexyl-2-pyrrolidone
(δS=20.5
MPa1/2),
dimethylformamide
(δS=24.9
MPa1/2),
cyclopentane (δS=16.5 MPa1/2), heptane (δS=15.3 MPa1/2), hexane (δS=14.9 MPa1/2), pentane
(δS=14.4 MPa1/2), acetone (δS=19.9 MPa1/2), acetonitrile (δS=24.3 MPa1/2), 1,3-dioxolane
(δS=21.4 MPa1/2), benzonitrile (δS=22.5 MPa1/2), isopropoxyethanol (δS=21.4 MPa1/2). The
final, optimized exfoliation was performed as follows: 45 g/L of GaS were sonicated for 6
hours in isopropanol in 20 mL vials and then centrifuged for 180 min at 2.5 krpm.
Size selection
We used controlled centrifugation with subsequently increasing rotation speeds as
previously reported.6, 40 10 g/L GaS in 90 mL of isopropanol was sonicated for 6 hours in three
centrifuge tubes. The 90 mL of sonicated dispersion was centrifuged at 0.5 krpm for 60 min.
The sediment was discarded and the supernatant was centrifuged again at 1 krpm for 60 min.
15
The sediment after this centrifugation step was redispersed in fresh IPA (5 min bath sonication)
producing the largest size. The supernatant after the 1 krpm centrifugation step was centrifuged
at 1.5 krpm for 60 min producing the second largest size in the redispersed sediment. These
steps were repeated in further increments of 2 krpm, 2.5 krpm and 3 krpm thus producing 5
sizes.
Characterization and equipment
Optical extinction and absorbance was measured on a Perkin Elmer 650 spectrometer
in quartz cuvettes with a path length of 0.4 cm. To distinguish between contributions from
scattering and absorbance to the extinction spectra, dispersions were measured in an integrating
sphere using a home-built sample holder to place the cuvette in the center of the sphere (NB
cuvettes need to be transparent to all sides and correct positioning is important). The
absorbance spectrum is obtained from the measurement inside the sphere. A second
measurement on each dispersion was performed outside the sphere in the standard
configuration to obtain the extinction spectrum. This allows calculation of the scattering
spectrum (extinction minus absorbance).
Low-resolution bright field transmission electron microscopy imaging was performed
using a JEOL 2100, operated at 200 kV. Holey carbon grids (400 mesh) were purchased from
Agar Scientific and prepared by diluting a dispersion to a low concentration and drop casting
onto a grid placed on a filter membrane to wick away excess solvent. Statistical analysis was
performed of the flake dimensions by measuring the longest axis of the nanosheet and assigning
it as “length”, L. STEM imaging and STEM EELS was carried out on a Nion UltraSTEM100
aberration-corrected dedicated STEM, equipped with a Gatan Enfina electron energy loss
(EEL) spectrometer, at the SuperSTEM Laboratory in Daresbury, UK. The microscope was
operated at an acceleration voltage of 60 kV.
EELS were acquired with a convergence semi-angle of 32 mrad, a collection semiangle of approximately 37 mrad, a dispersion of 0.5 eV/channel and an energy resolution over
vacuum of 0.3 eV. STEM EELS maps were acquired with 0.1s acquisition time per spectra per
pixel. Subsequently, the spectra were aligned to the Oxygen K-edge. The spectra shown here
are representative of a set of 10 STEM EELS maps that were acquired over similar regions.
16
Atomic force microscopy (AFM) was carried out on a Veeco Nanoscope-IIIa (Digital
Instruments) system equipped with a E-head (13 μm scanner) in tapping mode after depositing
a drop of the dispersion (10 μL) on a pre-heated (120 °C) Si/SiO2 wafer with an oxide layer of
300 nm. Typical image sizes were 3-10 μm at scan rates of 0.4-0.6 Hz.
Raman spectroscopy was performed using a WITec alpha 300 with 532 nm excitation
laser in air under ambient conditions. The Raman emission was collected by an Olympus 100×
objective (N.A. = 0.8) and dispersed by 600 lines mm−1 gratings. The laser energy was kept
below 0.2 mW. The mean of 20 spectra is displayed.
X-ray Photoelectron Spectroscopy was performed under ultra-high vacuum conditions
(<510-10 mbar) using monochromated Al Kα X-rays from an Omicron XM1000 MkII X-ray
source and an Omicron EA125 energy analyzer. The analyzer pass energy was set to 100 eV
for survey and 20 eV for core-level spectra, yielding a maximum energy resolution of ~0.65
eV. An electron flood gun was used for charge compensation and the binding energy scale was
referenced to the adventitious carbon 1s core-level at 284.8 eV. After subtraction of a Shirley
background, the core-level spectra were fitted with Gaussian-Lorentzian line shapes. Samples
were prepared by vacuum-filtering the dispersions using porous cellulose filter membranes
(MF-Millipore membrane, mixed cellulose esters, hydrophilic, 0.025 μm, 47 mm) to give thin
films.
Scanning Electron Microscopy was performed with a Carl Zeiss Ultra SEM operating
at 2 kV. Images were acquired using the secondary electron detector. Powder XRD was
performed using a Siemens D500 X-ray Diffractometer equipped with a Cu Kα emission source
(λ = 1.54056 Å) filtered through a graphite monochromator, at ambient temperature.
Hydrogen evolution catalysis
Dispersions of GaS in IPA were vacuum-filtered using porous cellulose filter
membranes (MF-Millipore membrane, mixed cellulose esters, hydrophilic, 0.025 μm) to give
uniform thin films. The deposited films were then cut into pieces (0.65 cm2) and transferred on
PyC electrodes using heat and pressure as described previously.24,
90
The cellulose filter
membrane was then removed by treatment with acetone vapor and subsequent acetone liquid
baths followed by an isopropanol rinse to remove the acetone residue. The mass per area for
all samples was approximately 0.65 mg/cm2.
17
Pyrolytic carbon was grown by chemical vapor deposition (CVD) from an acetylene
feedstock at 950 °C for 30 minutes to a thickness of 300–400 nm on 300 nm thermal SiO2 on
Si substrates in a hot wall quartz tube furnace as previously reported.91
Electrochemical measurements were carried out to evaluate the performance of GaS on
PyC films as electrodes for HER. The measurements were performed in 0.5 M H2SO4 solution
using a three-electrode electrochemical cell, with a reversible hydrogen electrode (RHE)
reference electrode and a graphite rod counter electrode. Electrochemical tests consisted of
linear sweep voltammetry and electrochemical impedance spectroscopy using a Gamry
Reference 3000 potentiostat. Before each test, samples were conditioned at a given voltage for
3 min. The electrocatalysis was measured using linear sweeping from 0 V to -0.8 V (vs. RHE)
with a scan rate of 5 mV/S. The AC impedance was measured within the frequency range of
0.1 to 10 mHz with perturbation voltage amplitude of 10 mV. The equivalent series resistance
of the system was measured by impedance spectroscopy from the high frequency intercept with
the real impedance axis and all the data were corrected by iR compensation.
Supporting Information Available: This includes more detailed descriptions of the basic
characterization, size selection, optical characterization and hydrogen evolution. This
information is available free of charge via the Internet at http://pubs.acs.org/.
Acknowledgement
The research leading to these results has received funding from the European Union Seventh
Framework Programme under grant agreement n°604391 Graphene Flagship. We have also
received support from the Science Foundation Ireland (SFI) funded centre AMBER
(SFI/12/RC/2278). In addition, JNC acknowledges the European Research Council
(SEMANTICS) and SFI (11/PI/1087) for financial support. CB acknowledges the German
research foundation DFG (BA 4856/1-1). HCN, EMcG, AS-A and VN acknowledge ERC
2DNanoCaps, SFI PIYRA and FP7 MoWSeS. GSD, NCB and SW acknowledge SFI for
PI_10/IN.1/I3030. NMcE acknowledges SFI (14/TIDA/2329). STEM experiments were
performed at SuperSTEM, the EPSRC UK national facility for aberration-corrected STEM.
18
Figure 1: A) Structure of gallium sulfide, GaS, in top and side view. Red: S, Green: Ga. B)
Photograph of a typical dispersion (in 2-propanol). C) Representative low-resolution bright
field TEM images of GaS nanosheets exfoliated in IPA. D) Bright-field scanning transmission
TEM image and E) high angle annular dark field (HAADF) STEM image of GaS showing the
intact lattice. F) Optical extinction, absorbance and scattering spectra of a typical GaS
nanosheet dispersion in IPA. G) Dispersed concentration of GaS expressed as optical extinction
at 365 nm divided by cell length plotted as a function of the Hildebrand solubility parameter
δS of the solvent. The concentration (right axis) is estimated by the extinction coefficient at 365
nm. The dashed line in the inset is a fit to eq 1. The dispersions in panel G were prepared with
initial GaS concentration Ci= 1 g/L and each was sonicated for ts=6 h and centrifuged at f = 2
krpm for tCF = 180 min. H) Raman spectrum (excitation wavelength 532 nm, mean of 20
spectra) of the filtered standard dispersion (IPA, Ci = 45 g/L, ts = 6 h, f = 2.5 krpm, tCF = 180
min) showing the characteristic phonons of GaS and a small impurity contribution which we
associate with Ga2S3. Inset: SEM image of the film. I, J) Fitted XPS core level spectra of a
filtered dispersion confirming the chemical nature of the nanosheets. I) Ga 3d core level
spectrum, J) S 2p core level spectrum.
19
Figure 2: A) Scanning transmission electron microscopic image of a GaS nanosheet. B) Colorcoded electron energy loss (EEL) map of the same region. Each pixel in the map corresponds
to the integrated intensity of the energy loss signal of the oxygen K-edge after background
removal to the same edge. Hence, the map shows the variation in the oxygen content across the
region where increasing oxygen content is shown from black to green to red to yellow to white.
The oxygen signal is most prominent on the edges of the nanosheet and at step edges in the
center of the nanosheet as indicated in the figure (yellow/white regions). C) Representative
EEL spectra extracted at different positions of the nanosheet, with the colors of the spectra
corresponding to the colors in the map in panel B. The spectra shown in yellow represent the
highest O content, as found near the edges of the nanosheets, and the spectra shown in black
and blue representing the lowest O content, as found on the surfaces inside the flake and over
vacuum.
20
50 nm
A
0
500
1000
<L> = 260 nm
20
15
100 nm
10
5
B
0
0
500
1000
200 nm
10
5
C
0
0
500
1000
Scattering (au)
Count
15
1.0
0.5
0.0
200
1.5
400
600
800
1
0.1
1.0
0.01
1E-3
0.5
300
600
E
1500
<L> = 405 nm
<L>=130 nm
<L>=260 nm
<L>=405 nm
0.0
200
25
20
1.5
D
1500
25
Count
Extinction (au)
<L> = 130 nm
Absorbance (au)
Count
35
30
25
20
15
10
5
0
400
600
800
1
0.1
F
0.01
1500
200
L (nm)
400
600 800
Wavelength (nm)
Count
G
12
200 nm
8
<L> = 130 nm
4
0
0
5
10
15
N
20
25
Number of layers, N
100
16
H
10
1
1E-3
<L> = 130 nm
<L> = 260 nm
<L> = 405 nm
0.01
0.1
1
2
Area (m )
Figure 3: A-C) Length histograms and representative images of size-selected GaS in IPA as
insets. A) smallest-size GaS, B) medium-sized GaS, C) large GaS. The mean length <L> is
shown as figure legend. Data for additional sizes are given in the SI. D-F) Optical extinction,
absorbance and scattering spectra of dispersions of different sized nanosheets showing
systematic changes in spectral shape. D) Extinction, E) absorbance and F) scattering.
Extinction and absorbance spectra are normalized to their local minimum while scattering
spectra are normalized to the minimum in the extinction spectra. The inset in E shows the
absorbance on a log scale to highlight the A-exciton (~420 nm) and the gap features. G) Typical
AFM number of layers, N, histogram of smallest-size GaS in IPA drop cast onto
Si/SiO2 wafers. The mean number of layers was determined from step height analysis (see SI).
Inset: representative image of a nanosheet. H) Plot of number of layers per nanosheet as a
function of flake area determined from AFM. The dashed line indicates N A behavior.
21
0
0.0
C
3
3
2
Ext, 365nm
1
Sca, 365nm
Abs, 365nm
0.2
0.4
B
0.1
0.01
0.2
0.4
0.6
L from TEM (m)
4
0
0.0
Abs420nm/Abs290nm
From Ext
From Sca
0.6
L from TEM (m)
0.1
B from absorbance (nm)
Scattering exponent, n
2
-1
-1
Coeffcient (*10 L g m )
0.8
A
4
314
0.2
0.4 0.6
L from TEM (m)
D
312
310
308
306
10
20
30
40
Number of layers, N
Figure 4: A) Long wavelength scattering exponent, n, measured from both extinction and
scattering spectra, plotted versus TEM mean length, L, for size-selected GaS in IPA. B) Ratio
of absorbance at 420 nm to that at 290 nm, Abs420nm/Abs290nm plotted versus TEM length, L,
for size selected GaS in IPA. The fit lines plotted in A and B can be used as metrics to determine
L according to eq 2 and 3, respectively. C) Extinction, absorbance and scattering coefficients
as a function of L. The fit line can be used to find the extinction coefficient for a mean
nanosheet length which in turn can be used to determine the dispersion concentration from the
optical extinction spectra according to eq 4. D) Position of the B-exciton, λB (measured from
the absorbance spectra), plotted versus number of monolayers per nanosheet, N (determined
from AFM). The relation can be used as metric to determine N according to eq 5.
22
2
J (mA/cm )
0
-2
-4
-6
-8
-10
-0.8
A
PyC
GaS, L=450 nm
GaS, L=280 nm
GaS, L=180 nm
-0.6
-0.4
-0.2
0.0
Overpotential (V)
V v RHE (V)
0.8
106 mV dec-1
0.7
85 mV dec-1
0.6
85 mV dec-1
0.5
B
0.4
-1.0 -0.5 0.0
0.5
1.0
1.5
2
2
J (mA/cm )
Log(| j |) (mA/cm )
0
C
GaS, L=180 nm, 85 mV dec-1
-20
MoS2, L=120 nm,
120 mV dec-1
-40
-0.6
-0.4
-0.2
V v RHE (V)
Pt, 30
mV dec-1
0.0
Figure 5: A) Linear sweep voltammograms (5 mV s-1) of vacuum-filtered GaS nanosheets
transferred onto pyrolytic carbon with a reversible hydrogen (RHE) reference electrode. The
electrocatalytic response towards the hydrogen-evolution reaction is compared for sizeselected GaS with three different mean sizes. A clear size-dependence is evident. Measured
potentials were subjected to iR correction. Supporting electrolyte was 0.5 M H2SO4. B)
Overpotential versus current density plots showing Tafel slopes expressed as mV per decade
(mV dec-1). C) Linear sweep voltammograms of s-GaS compared to liquid-exfoliated 2H-MoS2
of similar lateral dimensions and mass per area and Pt as benchmark. Tafel slopes are indicated
in the figure. Despite of the high onset potential, s-GaS reaches the same current density as
MoS2 at ~0.6 eV due to its smaller Tafel slope.
References
23
1.
Du, W.; Jiang, X.; Zhu, L., J. Mater. Chem. C. 2013, 1, 10592-10606.
2.
Nicolosi, V.; Chhowalla, M.; Kanatzidis, M. G.; Strano, M. S.; Coleman, J. N., Science 2013,
340, 1420-+.
3.
Ciesielski, A.; Samori, P., Chem. Soc. Rev. 2014, 43, 381-398.
4.
Smith, R. J.; King, P. J.; Lotya, M.; Wirtz, C.; Khan, U.; De, S.; O'Neill, A.; Duesberg, G. S.;
Grunlan, J. C.; Moriarty, G.; Chen, J.; Wang, J.; Minett, A. I.; Nicolosi, V.; Coleman, J. N., Adv. Mater.
2011, 23, 3944-3948.
5.
Hernandez, Y.; Nicolosi, V.; Lotya, M.; Blighe, F. M.; Sun, Z.; De, S.; McGovern, I. T.;
Holland, B.; Byrne, M.; Gun'Ko, Y. K.; Boland, J. J.; Niraj, P.; Duesberg, G.; Krishnamurthy, S.;
Goodhue, R.; Hutchison, J.; Scardaci, V.; Ferrari, A. C.; Coleman, J. N., Nat. Nanotechnol. 2008, 3,
563-568.
6.
Paton, K. R.; Varrla, E.; Backes, C.; Smith, R. J.; Khan, U.; O’Neill, A.; Boland, C.; Lotya, M.;
Istrate, O. M.; King, P.; Higgins, T.; Barwich, S.; May, P.; Puczkarski, P.; Ahmed, I.; Moebius, M.;
Pettersson, H.; Long, E.; Coelho, J.; O’Brien, S. E.; McGuire, E. K.; Sanchez, B. M.; Duesberg, G. S.;
McEvoy, N.; Pennycook, T. J.; Downing, C.; Crossley, A.; Nicolosi, V.; Coleman, J. N., Nat. Mater.
2014, 13, 624-630.
7.
Varrla, E.; Paton, K. R.; Backes, C.; Harvey, A.; Smith, R. J.; McCauley, J.; Coleman, J. N.,
Nanoscale 2014.
8.
Liu, L.; Shen, Z. G.; Yi, M.; Zhang, X. J.; Ma, S. L., RSC Adv. 2014, 4, 36464-36470.
9.
Yi, M.; Shen, Z. G., Carbon 2014, 78, 622-626.
10.
Hernandez, Y.; Lotya, M.; Rickard, D.; Bergin, S. D.; Coleman, J. N., Langmuir 2010, 26,
3208-3213.
11.
Bourlinos, A. B.; Georgakilas, V.; Zboril, R.; Steriotis, T. A.; Stubos, A. K., Small 2009, 5,
1841-1845.
12.
Du, W.; Lu, J.; Sun, P.; Zhu, Y.; Jiang, X., Chem. Phys. Lett. 2013, 568, 198-201.
13.
Liu, W. W.; Wang, J. N., Chem. Commun. 2011, 47, 6888-6890.
14.
Oyer, A. J.; Carrillo, J.-M. Y.; Hire, C. C.; Schniepp, H. C.; Asandei, A. D.; Dobrynin, A. V.;
Adamson, D. H., J. Am. Chem. Soc. 2012, 134, 5018-5021.
15.
Yi, M.; Shen, Z.; Ma, S.; Zhang, X., J. Nanopart. Res. 2012, 14.
16.
Yi, M.; Shen, Z.; Zhang, X.; Ma, S., J. Phys. D: Appl. Phys. 2013, 46.
17.
Zhou, K.-G.; Mao, N.-N.; Wang, H.-X.; Peng, Y.; Zhang, H.-L., Angew. Chem.,Int. Ed. 2011,
50, 10839-10842.
18.
Bang, G. S.; Nam, K. W.; Kim, J. Y.; Shin, J.; Choi, J. W.; Choi, S.-Y., ACS Appl. Mater.
Interfaces 2014, 6, 7084-7089.
19.
Shmeliov, A.; Shannon, M.; Wang, P.; Kim, J. S.; Okunishi, E.; Nellist, P. D.; Dolui, K.;
Sanvito, S.; Nicolosi, V., ACS Nano 2014, 8, 3690-3699.
20.
Sun, L.; Lin, Z.; Peng, J.; Weng, J.; Huang, Y.; Luo, Z., Sci. Rep. 2014, 4.
21.
Nuvoli, D.; Valentini, L.; Alzari, V.; Scognamillo, S.; Bon, S. B.; Piccinini, M.; Illescas, J.;
Mariani, A., J. Mater. Chem. 2011, 21, 3428-3431.
22.
Lotya, M.; Hernandez, Y.; King, P. J.; Smith, R. J.; Nicolosi, V.; Karlsson, L. S.; Blighe, F. M.;
De, S.; Wang, Z.; McGovern, I. T.; Duesberg, G. S.; Coleman, J. N., J. Am. Chem. Soc. 2009, 131,
3611-3620.
23.
Green, A. A.; Hersam, M. C., Nano Lett. 2009, 9, 4031-4036.
24.
De, S.; King, P. J.; Lotya, M.; O'Neill, A.; Doherty, E. M.; Hernandez, Y.; Duesberg, G. S.;
Coleman, J. N., Small 2010, 6, 458-464.
25.
Guardia, L.; Fernandez-Merino, M. J.; Paredes, J. I.; Solis-Fernandez, P.; Villar-Rodil, S.;
Martinez-Alonso, A.; Tascon, J. M. D., Carbon 2011, 49, 1653-1662.
26.
Lin, S.; Shih, C.-J.; Strano, M. S.; Blankschtein, D., J. Am. Chem. Soc. 2011, 133, 12810-12823.
27.
Shih, C.-J.; Paulus, G. L. C.; Wang, Q. H.; Jin, Z.; Blankschtein, D.; Strano, M. S., Langmuir
2012, 28, 8579-8586.
28.
Gao, H.; Shori, S.; Chen, X.; zur Loye, H.-C.; Ploehn, H. J., J. Colloid Interface Sci. 2013, 392,
226-236.
24
29.
Guardia, L.; Paredes, J. I.; Rozada, R.; Villar-Rodil, S.; Martinez-Alonso, A.; Tascon, J. M. D.,
RSC Adv. 2014, 4, 14115-14127.
30.
Samoilov, V. M.; Danilov, E. A.; Nikolaeva, A. V.; Yerpuleva, G. A.; Trofimova, N. N.;
Abramchuk, S. S.; Ponkratov, K. V., Carbon 2015, 84, 38-46.
31.
Wang, S.; Yi, M.; Shen, Z.; Zhang, X.; Ma, S., RSC Adv. 2014, 4, 25374-25378.
32.
Zhang, L.; Zhang, Z.; He, C.; Dai, L.; Liu, J.; Wang, L., ACS Nano 2014, 8, 6663-6670.
33.
May, P.; Khan, U.; Hughes, J. M.; Coleman, J. N., J. Phys. Chem. C 2012, 116, 11393-11400.
34.
Bourlinos, A. B.; Georgakilas, V.; Zboril, R.; Steriotis, T. A.; Stubos, A. K.; Trapalis, C., Solid
State Commun. 2009, 149, 2172-2176.
35.
Chabot, V.; Kim, B.; Sloper, B.; Tzoganakis, C.; Yu, A., Sci. Rep. 2013, 3.
36.
Quinn, M. D. J.; Ngoc Han, H.; Notley, S. M., ACS Appl. Mater. Interfaces 2013, 5, 1275112756.
37.
Backes, C.; Smith, R. J.; McEvoy, N.; Berner, N. C.; McCloskey, D.; Nerl, H. C.; O’Neill, A.;
King, P. J.; Higgins, T.; Hanlon, D.; Scheuschner, N.; Maultzsch, J.; Houben, L.; Duesberg, G. S.;
Donegan, J. F.; Nicolosi, V.; Coleman, J. N., Nature Commun. 2014, 5, 4576.
38.
Khan, U.; O'Neill, A.; Porwal, H.; May, P.; Nawaz, K.; Coleman, J. N., Carbon 2012, 50, 470475.
39.
Hanlon, D.; Backes, C.; Higgins, T. M.; Hughes, M.; O’Neill, A.; King, P.; McEvoy, N.;
Duesberg, G. S.; Mendoza Sanchez, B.; Pettersson, H.; Nicolosi, V.; Coleman, J. N., Chem. Mater.
2014, 26, 1751-1763.
40.
Hanlon, D.; Backes, B.; Doherty, E.; Cucinotta, C. S.; Berner, N. C.; Boland, C.; Lee, K.;
Lynch, L.; Gholamvand, Z.; Harvey, A.; Zhang, S.; Wang, K.; Moynihan, G.; Pokle, A.; Ramasse, Q.
M.; McEvoy, N.; Blau, W. J.; Wang, J.; Sanvito, S.; O’Regan, D. D.; Duesberg, G. S.; Nicolosi, V.;
Coleman, J. N., arXiv:1501.01881 2015.
41.
Kang, J.; Seo, J.-W. T.; Alducin, D.; Ponce, A.; Yacaman, M. J.; Hersam, M. C., Nat Commun
2014, 5.
42.
Backes, C.; Berner, N. C.; Chen, X.; Lafargue, P.; LaPlace, P.; Freeley, M.; Duesberg, G. S.;
Coleman, J. N.; McDonald, A. R., Angew. Chem., Int. Ed. 2015, accepted, 10.1002/anie.201409412R2.
43.
Eigler, S.; Hirsch, A., Angew. Chem.,Int. Ed. 2014, 53, 7720-7738.
44.
Backes, C.; Berner, N. C.; Chen, X.; Lafargue, P.; LaPlace, P.; Freeley, M.; Duesberg, G. S.;
Coleman, J. N.; McDonald, A. R., Angew. Chem.,Int. Ed. 2015, 54, 2638-2642.
45.
Voiry, D.; Salehi, M.; Silva, R.; Fujita, T.; Chen, M.; Asefa, T.; Shenoy, V. B.; Eda, G.;
Chhowalla, M., Nano Lett. 2013, 13, 6222-6227.
46.
Voiry, D.; Yamaguchi, H.; Li, J.; Silva, R.; Alves, D. C. B.; Fujita, T.; Chen, M.; Asefa, T.;
Shenoy, V. B.; Eda, G.; Chhowalla, M., Nat. Mater. 2013, 12, 850-855.
47.
Jan, R.; May, P.; Bell, A. P.; Habib, A.; Khan, U.; Coleman, J. N., Nanoscale 2014, 6, 48894895.
48.
Khan, U.; May, P.; O'Neill, A.; Bell, A. P.; Boussac, E.; Martin, A.; Semple, J.; Coleman, J.
N., Nanoscale 2013, 5, 581-587.
49.
May, P.; Khan, U.; O'Neill, A.; Coleman, J. N., J. Mater. Chem. 2012, 22, 1278-1282.
50.
Zhi, C. Y.; Bando, Y.; Tang, C. C.; Kuwahara, H.; Golberg, D., Adv. Mater. 2009, 21, 2889-+.
51.
Young, R. J.; Kinloch, I. A.; Gong, L.; Novoselov, K. S., Compos. Sci. Technol. 2012, 72, 14591476.
52.
Torrisi, F.; Hasan, T.; Wu, W.; Sun, Z.; Lombardo, A.; Kulmala, T. S.; Hsieh, G.-W.; Jung, S.;
Bonaccorso, F.; Paul, P. J.; Chu, D.; Ferrari, A. C., ACS Nano 2012, 6, 2992-3006.
53.
Finn, D. J.; Lotya, M.; Cunningham, G.; Smith, R. J.; McCloskey, D.; Donegan, J. F.; Coleman,
J. N., J. Mater. Chem. C 2014, 2, 925-932.
54.
Blake, P.; Brimicombe, P. D.; Nair, R. R.; Booth, T. J.; Jiang, D.; Schedin, F.; Ponomarenko,
L. A.; Morozov, S. V.; Gleeson, H. F.; Hill, E. W.; Geim, A. K.; Novoselov, K. S., Nano Lett. 2008, 8,
1704-1708.
55.
Cunningham, G.; Lotya, M.; Cucinotta, C. S.; Sanvito, S.; Bergin, S. D.; Menzel, R.; Shaffer,
M. S. P.; Coleman, J. N., ACS Nano 2012, 6, 3468-3480.
56.
Du, W. C.; Lu, J.; Sun, P. P.; Zhu, Y. Y.; Jiang, X. Q., Chem. Phys. Lett. 2013, 568, 198-201.
25
57.
Yasaei, P.; Kumar, B.; Foroozan, T.; Wang, C.; Asadi, M.; Tuschel, D.; Indacochea, J. E.; Klie,
R. F.; Salehi-Khojin, A., Adv. Mater. 2015, DOI: 10.1002/adma.201405150.
58.
Brent, J. R.; Savjani, N.; Lewis, E. A.; Haigh, S. J.; Lewis, D. J.; O'Brien, P., Chem. Commun.
2014, 50, 13338-13341.
59.
Jacobs-Gedrim, R. B.; Shanmugam, M.; Jain, N.; Durcan, C. A.; Murphy, M. T.; Murray, T.
M.; Matyi, R. J.; Moore, R. L.; Yu, B., ACS Nano 2013, 8, 514-521.
60.
Meng, X.; He, K.; Su, D.; Zhang, X.; Sun, C.; Ren, Y.; Wang, H.-H.; Weng, W.; Trahey, L.;
Canlas, C. P.; Elam, J. W., Adv. Funct. Mater. 2014, 24, 5435-5442.
61.
Hu, P.; Wang, L.; Yoon, M.; Zhang, J.; Feng, W.; Wang, X.; Wen, Z.; Idrobo, J. C.; Miyamoto,
Y.; Geohegan, D. B.; Xiao, K., Nano Lett. 2013, 13, 1649-1654.
62.
Catalano, I. M.; Ferrara, M.; Tantalo, P., physica status solidi (a) 1973, 20, K135-K138.
63.
Kipperman, A. H. M.; van der Leeden, G. A., Solid State Commun. 1968, 6, 657-662.
64.
Yang, S.; Li, Y.; Wang, X.; Huo, N.; Xia, J.-B.; Li, S.-S.; Li, J., Nanoscale 2014, 6, 2582-2587.
65.
Bourdon, A.; Bringuier, E.; Portella, M. T.; Vivières, M.; Piccioli, N., Phys. Rev. Lett. 1990,
65, 1925-1928.
66.
De Blasi, C.; Drigo, A. V.; Micocci, G.; Tepore, A.; Mancini, A. M., J. Cryst. Growth 1989,
94, 455-458.
67.
Parlak, M.; Erçelebi, Ç.; Günal, I.; Özkan, H.; Gasanly, N. M., Cryst. Res. Technol. 1996, 31,
673-678.
68.
Hu, P.; Wen, Z.; Wang, L.; Tan, P.; Xiao, K., ACS Nano 2012, 6, 5988-5994.
69.
Aksimentyeva, O. I.; Demchenko, P.; Savchyn, V. P.; Balitskii, O. A., Nanoscale Res. Lett.
2013, 8, 29.
70.
Li, X.; Lin, M.-W.; Puretzky, A. A.; Idrobo, J. C.; Ma, C.; Chi, M.; Yoon, M.; Rouleau, C. M.;
Kravchenko, I. I.; Geohegan, D. B.; Xiao, K., Sci. Rep. 2014, 4, 5497.
71.
Tamalampudi, S. R.; Lu, Y.-Y.; Kumar U, R.; Sankar, R.; Liao, C.-D.; Moorthy B, K.; Cheng,
C.-H.; Chou, F. C.; Chen, Y.-T., Nano Lett. 2014, 14, 2800-2806.
72.
Eckhoff, W.; Putnam, R.; Wang, S.; Curl, R.; Tittel, F., Applied Physics B 1996, 63, 437-441.
73.
Shen, G.; Chen, D.; Chen, P.-C.; Zhou, C., ACS Nano 2009, 3, 1115-1120.
74.
Taylor, M. J., J. Raman Spectrosc. 1973, 1, 355-358.
75.
Late, D. J.; Liu, B.; Matte, H. S. S. R.; Rao, C. N. R.; Dravid, V. P., Adv. Funct. Mater. 2012,
22, 1894-1905.
76.
Yadgarov, L.; Choi, C. L.; Sedova, A.; Cohen, A.; Rosentsveig, R.; Bar-Elli, O.; Oron, D.; Dai,
H.; Tenne, R., ACS Nano 2014, 8, 3575-3583.
77.
O’Neill, A.; Khan, U.; Coleman, J. N., Chem. Mater. 2012, 24, 2414-2421.
78.
Hughes, J. M.; Aherne, D.; Coleman, J. N., J. Appl. Polym. Sci. 2013, 127, 4483-4491.
79.
Lucazeau, G.; Leroy, J., Spectrochim. Acta, Part A 1978, 34, 29-32.
80.
Ho, C.-H.; Chen, H.-H., Sci. Rep. 2014, 4.
81.
Castellanos-Gomez, A.; Vicarelli, L.; Prada, E.; Island, J. O.; Narasimha-Acharya, K. L.;
Blanter, S. I.; Groenendijk, D. J.; Buscema, M.; Steele, G. A.; Alvarez, J. V.; Zandbergen, H. W.;
Palacios, J. J.; van der Zant, H. S. J., 2D Materials 2014, 1, 025001.
82.
Wood, J. D.; Wells, S. A.; Jariwala, D.; Chen, K.-S.; Cho, E.; Sangwan, V. K.; Liu, X.; Lauhon,
L. J.; Marks, T. J.; Hersam, M. C., Nano Lett. 2014.
83.
Li, L.; Yu, Y.; Ye, G. J.; Ge, Q.; Ou, X.; Wu, H.; Feng, D.; Chen, X. H.; Zhang, Y., Nat.
Nanotechnol. 2014, 9, 372-377.
84.
Ho, C. H.; Lin, S. L., J. Appl. Phys. 2006, 100, 083508.
85.
Aydinli, A.; Gasanly, N. M.; Goksen, K., J. Appl. Phys. 2000, 88, 7144-7149.
86.
Lassalle-Kaiser, B.; Merki, D.; Vrubel, H.; Gul, S.; Yachandra, V. K.; Hu, X.; Yano, J., J. Am.
Chem. Soc. 2015, 137, 314-21.
87.
Jaramillo, T. F.; Jorgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I., Science
2007, 317, 100-102.
88.
Wang, H. T.; Lu, Z. Y.; Xu, S. C.; Kong, D. S.; Cha, J. J.; Zheng, G. Y.; Hsu, P. C.; Yan, K.;
Bradshaw, D.; Prinz, F. B.; Cui, Y., Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 19701-19706.
26
89.
Higgins, T. M.; McAteer, D.; Coelho, J. C. M.; Sanchez, B. M.; Gholamvand, Z.; Moriarty, G.;
McEvoy, N.; Berner, N. C.; Duesberg, G. S.; Nicolosi, V.; Coleman, J. N., ACS Nano 2014, 8, 95679579.
90.
Wu, Z.; Chen, Z.; Du, X.; Logan, J. M.; Sippel, J.; Nikolou, M.; Kamaras, K.; Reynolds, J. R.;
Tanner, D. B.; Hebard, A. F.; Rinzler, A. G., Science 2004, 305, 1273.
91.
McEvoy, N.; Peltekis, N.; Kumar, S.; Rezvani, E.; Nolan, H.; Keeley, G. P.; Blau, W. J.;
Duesberg, G. S., Carbon 2012, 50, 1216-1226.
27
ToC fig
28
Nina Christina Berner