Highly effective three-dimensional functionalization of graphite to graphene by wet chemical exfoliation methods

Commercial graphite (SkySpring Nanomaterials, Inc.) with high purity (99.5 wt.%) was modified to three-dimensional graphene by means of two procedures. In the first method, 1 g of commercial graphite was added to 3 g of CaCO₃ (purchased from SkySpring Nanomaterials, INC., average particle size of 15–40 nm), and mixed mechanically. Then, 50 ml of 1-methyl-2-pyrrolioline (Sigma-Aldrich) was added and mixed well. Next, the mixture was stored at room temperature for 24 h. After this time, the flask was placed for 1 h in an ultrasonic bath (room temperature) and washed with distilled water using a Büchner funnel. Then, it was dried in an electric furnace at 100 °C for 24 h. After drying, 15 ml of furfuryl alcohol (POCh) was added, mixed with 3 drops of concentrated phosphoric acid aqueous solutions (75%) to each sample and mixed well, followed by heating of the material at 80 °C in an oven for 24 h to allow the polymerization of furfuryl alcohol, and subsequently carbonization at temperatures of 700 to 800 °C for 1 h. The mass was heated under the flow of nitrogen at a rate of 10°C/min in a tube furnace (Thermolyne F21100). After this process, the prepared samples remained in a tube furnace under the flow of nitrogen in order to reached room temperature. After carbonization, the samples were treated with concentrated (34–37%) HCl for 20 min (1 g of carbon was used per 12 ml of HCl) and then washed with distilled water using a Büchner funnel, until the pH of the solution reached 6–7. Then, it was dried in an electric furnace at 100 °C for 24 h. The action of HCl opened pores in the carbon matrix by etching the template. The procedure was found to be the most effective regarding the removal of CaCO₃ or Na₂CO₃. An analogous procedure was performed in the second method where Na₂CO₃ (POCh) was used as a template. In the second method, the aqueous solution of Na₂CO₃ (16.9%) was supplemented with 2 g of commercial graphite, and then mixed well for 1.5 h in a magnetic stirrer at room temperature. Next, 20 ml of 1-methyl-2-pyrrolioline was added, mixed with 0.1 ml of cetyltrimethylammonium chloride solution (CTAC) 25 wt.% in H₂O, and then mixed well for 1.5 h in a magnetic stirrer. After this time, the flask was placed for 1 h in an ultrasonic bath (at room temperature, 40 Hz frequency, 100 W power), washed with distilled water using a Büchner funnel. Then, it was dried in an electric furnace at 100 °C for 24 h. After drying, every 2 g of powder was supplemented with 15 ml of furfuryl alcohol, mixed well with 3–5 drops of concentrated phosphoric acid aqueous solutions (75%). The three-dimensional graphene obtained in these methods was denoted as GF_Na_1:1_T, GF_Ca_1:1_T, GF_Na_1:3_T and GF_Ca_1:3_T where: e.g. GF_Na_1:1—means the mass ratio of graphite (1 g) to Na₂CO₃ (1 g), GF_Ca_1:3—means the mass ratio of graphite (1 g) to CaCO₃ (3 g), T—is the carbonization temperature (°C). The sample denoted as GF_0 means pure graphite used in the research. The only losses were the amounts which were retained on the wall of laboratory vessels. To minimize graphite/graphene losses, all laboratory vessels were rinsed with distilled water and the residues of synthetic material were added to its main portion. Therefore, the weight losses were lower than 5% and the yield can be estimated as about 95% in relation to graphite.

The graphene samples were examined by the low-temperature adsorption of nitrogen method. The relevant isotherms of all the samples were measured at − 196 °C using an automatic adsorption instrument ASAP 2010 (Micromeritics, USA). Prior to gas adsorption measurements, the carbon materials were outgassed under a vacuum at 200 °C for 2 h. The specific surface area (S_BET) was determined based on the Brunauer-Emmett-Teller (BET) method from nitrogen adsorption data in the relative pressure range of 0.02 to 0.2. The micropore volume (V_mi) was calculated by using the t-plot method. The total pole volume (V_t) was determined from the amount of gas adsorbed at the relative pressure of 0.99.

The volumetric elemental composition (carbon, nitrogen and hydrogen) of the materials was analyzed by means of a combustion elemental analyzer (Vario MACRO CHN, Elementar Analysensysteme GmbH).

The morphology of the carbons prepared from graphite was analyzed by scanning electron microscopy (SEM, 1430 VP, LEO Electron Microscopy Ltd.). The obtained carbons were also examined by high-resolution transmission electron microscopy (HRTEM, FEI Europe production, model Tecnai F20 X-Twin). The carbon materials obtained prior to the HRTEM microscopic analysis were dispersed in ethanol and treated with an Inter Sonic IS-1K bath for 15 min and deposit on holey carbon-coated copper grids.

X-ray photoelectron spectroscopy (XPS) measurements were performed using a VG Scientific photoelectron spectrometer ESCALAB-210 with the use of Al Kα radiation (1486.6 eV) from an X-ray source operating at 14.5 kV and 25 mA. Survey spectra were recorded for all the samples in the energy range from 0 to 1350 eV with 0.4 eV step. High-resolution spectra were recorded with 0.1 eV step, 100 ms dwell time and 25 eV pass energy. A 90-degree take-off angle was used in all measurements. The curve fitting was performed using the AVANTAGE software provided by Thermo Electron, which describes each component of the complex envelope as a Gaussian–Lorentzian sum function; a constant 0.3 (± 0.05) G/l ratio was used. The background was fitted using the nonlinear Shirley model. Scofield sensitivity factors and the measured transmission function were used for quantification.

The graphene samples were examined by Raman spectroscopy (microscop Renishaw InVia, Renishaw plc (Gloucestershire, UK), Laser: Modu-Laser Stellar-REN, Multi-Line (max. Power 150 mW), Leica DM1300M camera Infinity 1; objective: Leica, N PLAN L50x/0.5). All of the spectra were collected at ambient temperature.

The elemental composition and surface parameters for the obtained samples of porous 3D graphene are given in Table 1. The elemental analysis revealed that the content of carbon increases with increasing carbonization temperature in the case of Na₂CO₃. Generally, the carbon content in all investigated series of 3D graphene is very high (higher than 90 wt.%). The N₂ sorption analysis for raw commercial graphite materials indicated a specific surface area of 25 m²/g, which, in fact, is too low for plenty of potential applications like electrode manufacturing. The results obtained in the case of the wet chemical exfoliation method combined with the hard template method (two templates: Na₂CO₃) indicated that the increase of the template ratio to graphite from 1:1 to 1:3 causes an increase of the specific surface area up 287 m²/g for both carbonization temperatures (700 °C and 800 °C). In the case of CaCO₃, a more intensive admission of the template resulted in a reversed effect. Nevertheless, the application of CaCO₃ allowed to achieve the ever-highest surface area of 333 m²/g. Generally, for both templates, an increase of the carbonization temperature from 700 to 800 °C causes a decrease of the surface area except for the sample GF_Ca_1:3_800. The templates differ regarding their form when admitted to the reaction mixture. Na₂CO₃ is added as a saturated water solution. Na₂CO₃ nanocrystals precipitate among exfoliated graphene flakes upon drying of the reaction mixture. On the contrary, CaCO₃ is added already as a nano-powder of a specific grain diameter, i.e., 15–40 nm, according to the product specification. Thus, differences in pore size distribution were expected.

The micropore volume calculated for series GF_Na_1:1_T and GF_Na_1:3_T proved that their contribution to the total pore volume was very high and placed in the range of 52–68%. The cationic surfactant (CTAC), which was used for GF_Na_1:1_T and GF_Na_1:3_T, allowed to obtain improved porosity in comparison to the application of sodium laureth ether sulphate (SLES) as described in other works (Ilnicka et al. 2017a). The application of Na₂CO₃ prefers the formation of micropores. In turn, the application of CaCO₃ nano-powder as a template led to the domination of mesopores. In this case, the share of V_mi did not exceed 46% of the total pore volume (Table 1). Nevertheless, the template method allowed to obtain durable 3D graphene structures of a diversified porosity, microporous or mesoporous for Na₂CO₃ and CaCO₃ template, respectively.

In general, nitrogen can positively influence on electric conductivity of graphene. Table 1 proves that the starting material, i.e., commercial graphite is itself the source of nitrogen. The nitrogen content was of 0.5 wt.% but it got reduced after 3D structuration. Thus, electrical conductivity of 3D structured graphene should be less affected by nitrogen then commercial graphite. It is not a positive side effect of 3D structuring procedure.

The results of combustion elemental analysis allowed to determine the overall bulk content of three principal elements: C, N and H, while the specific chemical environment of these elements and oxygen remains unknown. Therefore, 3D porous graphene samples were examined by XPS. The XPS spectra of the representative GF_Na_1:3_T series were determined and are depicted in Fig. 1. Additional detailed data of the spectroscopic analysis are given in Table 2. It is visible that all analyzed samples possess a very high content of carbon ~ 97 at.%. As shown in Fig. 1, the deconvolution of C 1s spectra prove that surface C atoms are occasionally bonded to O, which is confirmed by five major peaks at 284.8, 285.81, 287.61, 289.43, 291.22, 292.95 and 294.4 eV assigned to sp² C (C-1), sp³ C (C-2), C-O (C-3), C=O (C-4), and O=C–O (C-5), respectively. Nevertheless, the overall oxygen content is very low when compared to typical pyrolytic carbons or graphene oxide and reduced graphene oxide (Yang et al. 2018; Pei and Cheng 2012). Therefore, the performed synthesis of 3D structured graphene is definitely a non-oxidative method. The total amount of oxygen in samples GF_Na_1:3_700, and GF_Na_1:3_800 is 2.67, and 2.86 at.%, respectively. It can be noted that the oxygen content increased slightly with the carbonization temperature. Figure 1 shows the nature of O bonding on the carbon surface, where O-1 and O-2 are the major peaks on those spectra. A binding energy of 531.72 eV (O-1) and 533.48 eV (O-2) for these peaks are assigned for molecular oxygen and C–O bonds, respectively.

The morphology of 3D graphene samples was characterized by SEM and HRTEM (Fig. 2). The SEM and HRTEM images were recorded for GF_Na_1:1_700, GF_Ca_1:1_700, GF_Na_1:3_700 and GF_Ca_1:3_800 samples. Figure 2a, b containing images of GF_Na_1:1_700 and GF_Ca_1:1_700 samples, do not reveal major differences as the overall morphology and the level of graphene deglomeration despite different graphene to template ratio. All investigated samples had similar irregular surface characteristics due to the presence of randomly oriented graphene layers. The SEM images do not allow to determine the degree of graphene deglomeration in the examined 3D structures. The HRTEM method allowed for a better insight into the deglomeration process. The high resolution allowed to observe thinny graphene flakes exfoliated previously from graphite grains. The thinny flakes are the building stones of 3D structures under examination. Additionally, it can be concluded that the analyzed materials are pure without a residue of the template. The degree of deglomeration can be defined as DLG (Double Layered Graphene) and FLG (Few Layered Graphene) by using the HRTEM method (Fig. 2c, d).

To confirm the quality of the investigated materials, the 3D graphene samples were analyzed by Raman spectroscopy. Figure 3 shows the Raman spectra recorded for 532 nm for GF_Na_1:3_T and GF_Ca_1:1_T series. The most characteristic graphene/graphite bands occur in the spectral region 500 and 3500 cm⁻¹ (Grodecki 2013; Choi et al. 2014). Figure 3 shows example Raman spectra containing three main bands at 1348 cm⁻¹ (D-band), 1578 cm⁻¹ (G-band), and 2712 cm^{− 1} (G’-band). The D-band intensity is attributed by some authors to the presence of SLG/DLG sheets (Drewniak et al. 2014; Gayathri et al. 2014). The placement of the G’ band is related to the deglomaration degree of graphene layers. Intense deglomation towards double layered graphene (DLG) results in the shift of the peak G’ towards 2700 cm⁻¹, of which effect is observable in the presented spectra, too. The significant intensity of peak D is usually interpreted as a result of an intense deglomeration of graphene towards SLG/DLG and FLG. As for graphite, the intensity of D-band is relatively low. G-band in graphite is located close to 1560 cm⁻¹, while the band for investigated samples is located close to 1580 cm⁻¹, which is a value typical for graphene. The intensity ratio I_2D/I_G and I_D/I_G of investigated samples indicate that the obtained samples consisting multilayer nature of the graphene.

In summary, the presented results of HRTEM images, and Raman spectra clearly indicate that the presented synthesis pathway led to 3D graphene samples consisting of durably fixed DLG and/or FLG/MLG sheets.