Raman spectroscopy was then applied to further characterize the crystalline quality of as-prepared samples. As shown in
Fig. 4a, D band (1350 cm
-1), the G band (1585 cm
-1) and the 2D band (2700 cm
-1) [
31] of all samples were observed. The D band represented the disorderness in graphite or the edge of the graphite [
36]. The G band was associated with the normal first order Raman scattering process in grapheme [
31]. The 2D band originated from the double resonance (DR) Raman process, and was a robust way to quantify the number of layers in a graphene sample [
13]. First of all, both nitrogen doping and the exfoliating process can significantly enhance the D band [
17,
37], which was due to the doping of nitrogen atom that induced the defects inside the graphene network, and the ball-milling process exfoliated the bulk graphite, and in the meantime, produce the edge defects in the graphene sheets. In
Fig. 4a, it was observed that with the increase of nitrogen content, the D band had a higher intensity as expected. Furthermore, when the nitrogen atom doping increased, the G band peak gradually split into two peaks which was also consistent with what was previously reported by Lin
et al. about the N-doped graphene derived from ammonia plasma treatment of pure grapheme [
17]. This phenomenon could be due to the graphitic N were doped to the lattice of graphene which caused G band split. Besides, the G band peak shifted to lower frequency when the nitrogen content became higher. In sample 1:10, the peak even shifted to 1576 cm
-1 [
38]. The intensity ratio
ID/
IG between the G and the D band was used to quantify the density of defects in monolayers and few-layer grapheme [
39]. In Verónica León's research, the
ID/
IG of bilayer graphene was reported to be about 0.47 [
24]. In our work, however, with the nitrogen atom doping into the bilayer graphene (sample 1–10), the ratio
ID/
IG was calculated to be 1.17 which was much higher than that in the pure bilayer graphene and indicated a low crystallinity of the obtained material. Furthermore, the shape of 2D band peak changed gradually due to the exfoliating process and nitrogen doping. The 2D band peak of graphite was asymmetric. With the nitrogen doping and the number of layer decreasing, the shape of the 2D band peak gradually became symmetric. In
Fig. 4b, we presented the Gaussian fitting 2D Raman band of sample 1–10. There were 4 peaks after the fitting at 2658 cm
-1, 2688 cm
-1, 2706 cm
-1 and 2721 cm
-1 respectively, which were consistent with the reported results from L.M. Malard about Raman spectra for bilayer grapheme [
40]. But the relative intensity of the four peaks were much different from previous work. The peak at 2658 cm
-1 in our research turns out to have a lower intensity but the peak at 2706 cm
-1 turned out to be higher which could be the result of nitrogen doping. This result also proved that the melamine facilitated exfoliation of the graphite.