FIGURE 6 Dependence of CG on a)tD and b) ω .
concentration, C0 – initial concentration of dispersion, and τ – constant oftD ). On the basis of monitoring data,Cs and τ are determined respectively to be 0.22 mg mL-1 and 25 h, indicating 92% of FLG sheets still remaining in the dispersion even after 250 h of deposition. Such good stability is thought to result mainly from electrostatic repulsion of VIB-co -VI-co -Py adsorbed on the FLG surface. Zeta potential (ξ ), a parameter usually used for evaluation of repulsion, is measured as 37.8 mV for the FLG dispersion, while |ξ| > 30 mV is a widely accepted cut-off range for colloidal stability.51 The dependence of CG on centrifugation speed (ω ) was also evaluated (Figure 6b). After 30 min centrifugation, the dispersion shows a negligible decrease ofCG at ω ≤ 2,000 rpm. It is because the dispersion has already experienced a centrifugation of 2,000 rpm/90 min in its production process. Over 2,000 rpm,CG decreases sharply from 0.228 to 0.073 and 0.049 mg mL-1 at 8,000 and 12,000 rpm, respectively, exhibiting an empirical scaling ofCG ∝ ω -1. Despite undesirable decrease ofCG , a positive phenomenon is observed. That is, the content of single-layer graphene improves to 33% and 76% respectively in 8,000 rpm- and 12,000 rpm-centrifuged supernatants, compared with 11% in 2,000 rpm-centrifuged one (SI: Section IV). This offers an alternative method to produce the single-layer graphene.
FLG-P hybrid prepared from scalably produced dispersion, hereafter named as FLG-P-80 hybrid, displays an unusual ability of being invertibly dried and re-dispersed in pure water (by 5 min sonication) for many rounds. The thus re-obtained dispersion has a maximal CG of 3.87 mg mL-1, still reaching 3.23 mg mL-1even after being deposited for 240 h. Such re-dispersing ability, which is believed to be contributed by VIB-co -VI-co -Py adsorbed on the FLG surface, is absent from reduced graphene oxide (rGO) and most of surfactant-exfoliated FLG, highly facilitating the storage, delivery, and applications of FLG. The content of VIB-co -VI-co -Py inFLG-P-80 hybrid is determined to be 6.8 wt% by TGA (SI: Section V).
Taking the aqueous dispersion of FLG-P-80 hybrid as a typical sample, AFM was carried out to examine the exfoliation state of FLG. Figure 7a presents a typical AFM height image with a cross-section analysis. The analytical results on 107 different flake-like structures reveal that the thickness is mostly within 1.1−2.8 nm (Figure 7b-i). In light of the AFM thickness of pristine graphene usually within 0.6−0.9 nm,52,53 the determined thickness of 1.1−2.8 nm signifies that 90% of detected flakes are single- (9%), double- (17%), or few-layer (3- to 5-layer, 64%) graphene sheets adsorbed with VIB-co -VI-co -Py thin layers. Furthermore, the average lateral size of flakes is determined to be 533 ± 102 nm (Figure 7b-ii), in good agreement with the DLS value of 583 ± 115 nm. Within instrumental resolution, no defects like holes were observed on the basal planes, indicative of high quality of the FLG sheets.