Enhancing the Liquid-Phase Exfoliation of Graphene in Organic Solvents upon Addition of n-Octylbenzene

Due to a unique combination of electrical and thermal conductivity, mechanical stiffness, strength and elasticity, graphene became a rising star on the horizon of materials science. This two-dimensional material has found applications in many areas of science ranging from electronics to composites. Making use of different approaches, unfunctionalized and non-oxidized graphene sheets can be produced; among them an inexpensive and scalable method based on liquid-phase exfoliation of graphite (LPE) holds potential for applications in opto-electronics and nanocomposites. Here we have used n-octylbenzene molecules as graphene dispersion-stabilizing agents during the graphite LPE process. We have demonstrated that by tuning the ratio between organic solvents such as N-methyl-2-pyrrolidinone or ortho-dichlorobenzene, and n-octylbenzene molecules, the concentration of exfoliated graphene can be enhanced by 230% as a result of the high affinity of the latter molecules for the basal plane of graphene. The LPE processed graphene dispersions were further deposited onto solid substrates by exploiting a new deposition technique called spin-controlled drop casting, which was shown to produce uniform highly conductive and transparent graphene films.

10 mL. Since the ratio between graphite powder and the solvent was kept constant, i.e. 1 wt %, different masses of graphite flakes were used, and details have been reported in  Table S1. Masses of graphite, solvent and organic molecules used in LPE process. S--4

Concentration of graphene in the different solvents
To quantify the concentration of graphene after centrifugation, a mixture of graphene dispersion and chloroform was first heated up to 50 °C for 30 minutes and then passed through polytetrafluoroethylene membrane filters (pore size 100 nm  S--5

Surface area
Typically NOTBZ does not form any ordered monolayers at the HOPG/graphene surface at room temperature and ambient pressure. Nevertheless, theoretical assumption of NOTBZ packing motif on graphene has been used for calculation the coverage areas. The area occupied by single molecule amounts to A= (5.92± 0.9) nm 2 . The areas A occupied by single dispersion-stabilizing molecules, as well as area of graphene unit cell (G unit cell = (0.052± 0.004) nm 2 ) estimated by theoretical assumption can be used for calculating the mass and number of dispersion-stabilizing molecules needed to cover accessible graphene area. Noteworthy, in our calculations graphite powder has been considered as a single (rectangular) graphene sheet. We found a correlation between the best ratio, i.e. 15 % and the surface area coverage.
Interestingly, only 100 % of graphene coverage gives the highest exfoliation yield and for all the solvents we tested. S--6

UV-Vis-IR spectroscopy
All dispersions were also characterized by UV-vis-IR absorption spectroscopy.
The spectra are as expected being featureless in the visible -IR region. Each of these dispersions was diluted a number of times and the absorption spectra recorded. The absorbance (660 nm) divided by cell length is plotted versus concentration. A Lambert -Beer behavior was observed for all samples.
4.1 Graphene exfoliated in NMP in the presence/absence of NOTBZ with the highest concentration.
Graphene dispersions were characterized by UV-vis-IR absorption spectroscopy.
As expected, the spectra are featureless in the visible -IR region. Each of these dispersions in NMP, i.e. graphene, graphene + NOTBZ was diluted a number of times and the absorption spectra recorded. Graphene dispersions were characterized by UV-vis-IR absorption spectroscopy.
As expected, the spectra are featureless in the visible -IR region. Each of these dispersions in o-DCB, i.e. graphene, graphene + NOTBZ was diluted a number of times and the absorption spectra recorded.
As expected, the spectra are featureless in the visible -IR region. Each of these dispersions in DMF, i.e. graphene, graphene + NOTBZ was diluted a number of times and the absorption spectra recorded. Graphene dispersions were characterized by UV-vis-IR absorption spectroscopy.
As expected, the spectra are featureless in the visible -IR region. Each of these dispersions in TCB, i.e. graphene, graphene + NOTBZ was diluted a number of times and the absorption spectra recorded.
S--8 Figure S6. UV-Vis spectra of graphene dispersion in TCB in the presence/absence of NOTBZ at the highest concentration.

Lambert Beer behaviors
The absorbance at 660 nm divided by cell length is plotted versus concentration.   S--10

X-ray photoelectron characterization (XPS)
XPS analyses were carried out on a Thermo Scientific K-alpha X-ray photoelectron spectrometer with a basic chamber pressure of ~10 -8 mbar and Al anode as the X-ray source (x-ray radiation of 1486 eV). Spot sizes of 400 µm were used and pass energies of 200 eV for survey scans and 50.00 eV for high-resolutions scans were used.
150 µL of dispersions were spin coated on Au substrate for 1 minutes at 1000 rpm and substrates were annealed for 1 day at 100 °C in a oven under vacuum.
We used XPS to analyze graphene + NOTBZ in NMP before and after washing the NOTBZ. The high-resolution C1s XPS spectrum in Fig. S9a of the graphene sheets from NMP dispersions showed a sharp peak at 284.3 eV that corresponded to C-C bonds of carbon atoms in a conjugated honeycomb lattice. Figure S9b represents the highresolution C1s XPS spectrum of graphene + NOTBZ showed a larger peak at 284.6 eV corresponding to delocalized π conjugation from the sp 2 atomic structure of graphite.
Also this peak exhibits a larger behavior as the C-C bond from NOTBZ is also contributed. After the washing step, the peak at 284.4 eV became sharp again ( Figure   S9c) and confirmed the removal of NOTBZ.   S--13

Raman
The first order Raman spectrum is characterized by G and D peak. The D peak is typically observed in graphene obtained by LPE and it is attributed to finite size (i.e. D peak activated by the edges). Therefore, its intensity depends on the size of the flakes (and also on the excitation energy). Table S2 shows the results obtained on thickness distribution in different solvents and I(D)/I(G) obtained from the Raman spectra. Figure   S11 shows the thickness distribution in Table 1. Raman spectroscopy shows that NOTBZ is an effective agent for exfoliation of graphite in o-DCB, while it seems to be less efficient in NMP and TCB. In the case of DMF, the addition of NOTBZ results in a strong decrease of the single-layer content in the dispersion. Those data are not in agreement with TEM. However, one must remember that Raman spectroscopy has been performed with an optical microscope on flakes deposited by drop-casting on a silicon substrate. Therefore, small size flakes (<300 nm) may be difficult to spot, in contrast to TEM analysis, which mainly focuses on small size flakes.

S--15
Surface coverage was estimated by analyzing AFM images for each deposition. In particular, the average surface coverage was measured using Gwyddion's statistical analysis from different images with different scan sizes.

Figure S15
Estimation of the surface coverage.