Facile synthesis of pristine graphene-palladium nanocomposites with extraordinary catalytic activities using swollen liquid crystals

Amazing conductivity, perfect honeycomb sp2 arrangement and the high theoretical surface area make pristine graphene as one of the best materials suited for application as catalyst supports. Unfortunately, the low reactivity of the material makes the formation of nanocomposite with inorganic materials difficult. Here we report an easy approach to synthesize nanocomposites of pristine graphene with palladium (Pd-G) using swollen liquid crystals (SLCs) as a soft template. The SLC template gives the control to deposit very small Pd particles of uniform size on G as well as RGO. The synthesized nanocomposite (Pd-G) exhibited exceptionally better catalytic activity compared with Pd-RGO nanocomposite in the hydrogenation of nitrophenols and microwave assisted C-C coupling reactions. The catalytic activity of Pd-G nanocomposite during nitrophenol reduction reaction was sixteen times higher than Pd nanoparticles and more than double than Pd-RGO nanocomposite. The exceptionally high activity of pristine graphene supported catalysts in the organic reactions is explained on the basis of its better pi interacting property compared to partially reduced RGO. The Pd-G nanocomposite showed exceptional stability under the reaction conditions as it could be recycled upto a minimum of 15 cycles for the C-C coupling reactions without any loss in activity.


1
Different parameters like concentration of surfactant, amount of graphite and duration of sonication were estimated to get the maximum concentration of graphene exfoliated in solution. The optimized conditions were used for preparing the pristine graphene. Typically, the solution of surfactant in water was prepared by dissolving SDS (0.4 g/ml) in a beaker. A desired dispersion of graphene was prepared by sonicating required amount of graphite (0.3 mg/ml) in the SDS solution for 6 h in a bath sonicator at room temperature. The resulting dispersion was left to stand for 24 h for letting the unstable and heavy aggregated flakes of graphene to settle down completely. The undisturbed solution was then centrifuged at 500 rpm for 90 min. The top half of the solution was pipetted out and retained for further use. The yield of graphene was calculated with the help of thermogravimetric analysis (TGA). To get the maximum concentration of graphene exfoliated in solution different parameters like concentration of surfactant, amount of graphite and duration of sonication were studied in detail. For the exfoliation process, surfactant solutions of SDS in water were prepared by dissolving different concentration of surfactant in water (0.1, 0.2, 0.3 and 0.4 g/ml). Accurately weighed quantity of graphite powder (0.3 mg/ml) was then dispersed in the surfactant solutions. The suspension was then ultrasonicated in a bath sonicator for 6h. The resulting dispersion was left to stand for 24 h, for letting the unstable heavy aggregated flakes of graphene to settle down completely. The undisturbed solution was then centrifuged at 500 rpm for 90 min. The top half of the solution was pipetted out and retained for further use. UV-Visible spectra of all the exfoliated graphene solutions were recorded and are shown in, Fig. S1a. The optical density at 660 nm was used as an indicator of the concentration of graphene in the suspension. 2 From Fig. S1b, it is clearly evident that the maximum concentration of graphene was observed where the concentration of surfactant was 0.4 g/ml. However, the increase in concentration of graphene when the graphite concentration was raised from 0.3 mg/ml to 0.4 mg/ml was not significantly high. The optical absorption until 300 nm is predominated by SDS. UV-visible absorption spectra of aqueous SDS solutions having same concentrations as that were used for making graphene are shown in Fig. S1b. Comparison of Fig. S1a and S1b give a clear confirmation of broad range optical absorption by graphene. Duration of sonication was found to be another important factor that affects the exfoliation. We found that concentration of graphene increased when sonication was done upto 6 h. There was no further increase in the concentration of graphene beyond 6 h of sonication as it shown in Fig. S1 c. Finally, it was concluded that best conditions for obtaining maximum graphene dispersion would be a combination of 0.4 g/ml surfactant solution, 0.3 mg/ml of graphite powder and 6 hours of sonication in a bath sonicator. Figure S1. UV-Visible absorption spectra of (a) graphene dispersion in water that are stabilized by varying concentrations of SDS and (b) aqueous solution of varying concentrations of SDS and c) graphene dispersion in SDS solution at different sonication time.

Finding out the yield of graphene
To calculate the yield of graphene synthesized using the optimized procedure, we prepared 400 ml of the graphene suspension. The solution was filtered through an inorganic filter membrane having pore size of 20 nm. The accurate mass of the filter membrane was recorded before filtration. Filtered graphene was washed thoroughly with double distilled water and vacuum dried in a hot air oven for 12 h to remove all traces of water. After weighing the dried graphene collected on the membrane, we could calculate the total mass of exfoliated graphene and residual SDS. Thermogravimetric analysis (TGA) of a portion of the dried samples was recorded to determine the amount of graphene. We used 2 mg of the Figure S2. TGA thermal curves for the exfoliated graphene under nitrogen atmosphere.
sample and it was heated to 800 0 C at a heating rate of 10 0 C min -1 under nitrogen gas flow at the rate of 40 ml min -1 . The TGA thermal curve is shown in Fig. S2. From the TGA thermal curve, it is evident that the mass loss occurs in two stages. The mass loss in the first step from 30 0 C to 450 0 C corresponds to the decomposition of SDS. 3 There was only a minor loss of 5 % weight for the sample till 650º C. This-clearly showed that the sample contained only a small % (w/w) of the surfactant.

Synthesis of GO and RGO
Hummers method was followed for the synthesis of GO and RGO. 4 Typically, graphite powder ( Finally it was filtered and dried in vacuum oven. GO (10 mg) was sonicated for 2 h in water (50 ml). The sonicated dispersion was centrifuged for 1 h. Upper half of the solution was collected to do further reactions. The above suspension of GO (25 ml) was mixed with water (50 ml), NH 3 solution (450 μl, 25%) and hydrazine (80%, 9 μl) to perform the reduction of GO at 95 °C and continued stirring at same temperature for 1 hour. After the completion of the reaction, black coloured RGO was formed. Characterization of synthesized graphene: Preliminary characterization of exfoliated graphene dispersed in SDS solution was done by using UVvisible spectroscopy. Typical absorption spectrum of exfoliated graphene is presented in Fig. S3. A strong peak at 214 nm is due to the characteristic absorption of SDS. The UV-Visible spectra reveal that the optical absorption till 350 nm is dominated by the surfactant. The broad absorption beyond 300 nm in the spectrum showing a notifying difference to the blank confirms the optical absorption by graphene. 2 Further analysis of the sample was performed to characterize the morphology of exfoliated graphene. Further moving on to TEM data analysis we observe lots of folded and wrinkled flakes (Fig. S4a). This is a clear evidence of the thinness of exfoliated sheets. High resolution TEM images ( Fig. S4b and S4c) showed perfect 2-D crystalline structure of the graphene sheet. Furthermore, the selective area electron diffraction (SAED) patterns were recorded and a typical pattern is shown in Fig. 4d. Presence of two bright concentric rings in the diffraction patterns are due to (1100) and (1120) crystallographic planes of exfoliated graphene. 5 The concentric rings in the SAED pattern showed subtle difference in their intensity. The inner ring was weaker in intensity than the outer one. This kind of a pattern is known to be due to multilayer graphene. 5 The lateral dimension of majority of the flakes extended upto many micrometres. AFM imaging and analysis of graphene was used to get information about the thickness of graphene sheets. A typical AFM image is shown in Fig. S5a shows the presence of a large sheet of graphene having lateral dimension more than 1 μm along with some smaller flakes. Height profiling data along with images give information about the number of layers present in the synthesized graphene sheets which could be statistically evaluated to get the information about the quality of graphene. A histogram depicting the distribution in thickness of graphene flakes is shown in Fig. S5c conveying the presence of more than 70% of the sheets having stacked layer less than 5. The thinnest flake had (also thickness of ~1 nm. This corresponds to monolayer graphene. However, in case of exfoliated graphene, increased thickness of the sheets could be due to the residual surfactant molecules on the surface. 6 Detailed analysis of large sheets showed that the thickness of the sheets is maximum at the end and minimum in the middle. This suggested that the graphene sheets were having rolled edges are confirmed proofs about the formation of relatively thin sheets. XRD analyses of graphite and the exfoliated graphene were done and the diffraction patterns are shown in Fig. S6a. XRD pattern of graphite showed a sharp and strong diffraction peak at 2θ=26.6 which was its main characteristic peak corresponding to (200) basal plane along with some small peaks related to (100), (010) (004), (110) planes in graphite. This sharp peak confirmed that the d-spacing in graphite was 3.36Å. 7,8 The exfoliated graphene showed a broad peak at 2θ=24.8. The peak shift is due to the exfoliation of graphite with increase in d spacing from 3.36 to 3.62 Å. This increase in d spacing must be due to insertion of surfactant molecule in between the graphitic layers. Raman spectroscopy has emerged as a powerful tool to study the quality of graphene. Raman spectra of graphite, exfoliated graphene and RGO are shown in Fig.S6b. All the three materials showed most intense peak a t~1354, ~1625 and ~2700 cm -1 corresponding to D, D′ and 2D bands of gra-phene. 9 The D′ band tells us about the graphitic stacking structure. 10 The 2D band was also observed for the three samples at ~2700 cm -1 . Exfoliated graphene and RGO had an additional peak at ~1354 cm -1 corresponding to the D band. The D band correlates to the order/disorder of the graphite edges and was absent for graphite. The presence of D band is expected for RGO because of the residual oxide groups causing defects in sheets. But the appearance of D band in exfoliated graphene does not necessarily confirm the presence of defects. The appearance of D band for exfoliated graphene could be due to the presence of small graphene flakes having lateral dimension less than 1 μm. 6 The size of the laser spot used in Raman spectroscope was larger than 1 μm. Hence, a good deal of small graphene flakes would come under the laser. The smaller flakes will contribute toward the intensity increase of D band as they will be recorded as edge defects. 9 Even, the torn edges within the large graphene flakes as seen in AFM images will also contribute to the intensity of D band. Nevertheless, the intensity of D band for exfoliated graphene was lesser than RGO. Additionally, the degree of apparent disorder can be calculated by the D/D′ intensity ratios. The calculated D/D′ ratio in case of RGO and G are 0.73 and 0.56 respectively confirming the less defective structure of G sheets than RGO. The 2D band in graphene did not broaden apparently compared to the graphite spectrum. This suggests that the structure of the basal plane in G was still preserved which completely agrees with the SAED results.

Preparation and characterization of SLCs containing G, RGO and Pd precursor
Transformation of the mixture of the aqueous phase containing surfactant and graphene and the oil phase containing the Pd salt from an opaque emulsion to a transparent and viscous gel gave the primary indication of the formation of hexagonal mesophase. Colour of mesophases; SLC-G1, SLC-RG and SLC-Pd was pale red whereas SLC-G2 was wine red ( Figure S7). Colour of the mesophases is due to the typical colour of the Pd(dba) 2 complex. The hexagonal mesophases are formed by the self-assembly of infinitely long surfactant-stabilized oil tubes that are regularly arranged in brine solution. 13 The surfactant cylinders have a strong tendency to align and grow parallel to the walls of the container (glass culture tubes here). 13 This growth takes place over a period of few days. Hence, the synthesized mesophases were left undisturbed for few days before further experiments. The mesophases were further characterized by polarized optical microscopy imaging (POM). A typical POM image of the SLC-G1 is shown in Figure S8. The typical striations like texture of hexagonal mesophases can be seen in this figure. Growth and elongation of the tubes causes evolution of the typical patterns and this is evident in the POM image recorded after 24 h that is shown in Figure S8b.
Fan shaped focal conic texture is characteristic of hexagonal mesophases.         Adsorption of p-NP on the nanocomposites was studied by using UV-visible absorption spectroscopy, as shown in figure S17. The nanocomposite (20 mg) was added into aqueous solution of p-NP (50 ml, 3 mM). The mixture was thoroughly stirred at room temperature in dark and time depended adsorption experiment was performed. The amount of p-NP adsorbed per unit weight of the nanocomposite, Q e (mg g -1 ), was calculated from the mass balance equation given below: Q e = (C o -C e ) Vm -1 where C o is the initial p-NP concentration in the liquid phase (mg L -1 ), C e is the p-NP concentration at equilibrium (mg L -1 ), V is the volume of p-NP solution used (L), and m is the mass of adsorbent (g). The maximum adsorption capacity (Q max ) for the nanocomposites RGPd 0.001 and GPd 0.001 were 0.066 and 0.089 mg/g respectively.

Optimization of base and solvents
The catalytic activities of the four different catalysts (GPd 0.001 , GPd 0.01 , RGPd 0.001 and Pd 0.001 ) were tested initially in the Suzuki reaction using Chloro benzene and phenyl boronic acid as reactants. GPd 0.001 M was found to be the best catalyst. Then, the effect of base on the efficiency of GPd 0.001 M was tested using five different bases: NaOH, K 2 CO 3 , Na 2 CO 3 , Na 2 PO 4 and (C 2 H 5 ) 3 N. The results are given in Table S2. The effect of solvent was also studied using four different solvents such as water, ethanol, toluene and water ethanol 1:1 mixture in Suzuki reaction using chloro benzene and the results are given in Table S3.