Figure 1 shows the XRD patterns of the PG and FG samples. A very intense and narrow peak is shown referring to X-ray reflection on the (002) plane of graphite. Intriguingly, (002) peak of the FG (2θ = 26.56o) shifts toward a slightly higher angle by approximately 0.04° compared to peak position of the PG (2θ = 26.52°). It indicates a smaller c-axis lattice parameter of the FG (0.3353 nm) than that of the PG (0.3358 nm). Moreover, the full width at half-maximum (FWHM) of the (002) XRD peak is significantly larger in the FG. This is strongly related to the crystallite size according to the Scherrer equation, La = 1.84λ/β.cosθ, where the λ is the X-ray wavelength (0.15406 nm), θ is the position of the (002) peak, and β is the FWHM of the (002) peak in 2θ (rad) units. Thus, the crystallite size of the FG and PG were calculated to be approximately 86 nm and 101 nm, respectively. However, the intrinsic instrumental broadening should also be considered when elucidating exact values. A slight decrease in interlayer spacing suggests the removal of interlayer species that are present in low quantities in pristine graphite samples. As such, XRD provides indirect evidence to assure that the floating technique can aid the enrichment of carbon percentage in graphite samples. The quantitative elemental analysis that has been performed for both samples will attest to this point (vide infra). The XRD analysis of flake graphite (99% Purity, Alfa Aesar) that has been performed by Abdolhosseinzadeh et al. show very similar pattern where the peak positions of flake graphite have been shifted to higher 2θ values. The (002) diffraction appears at 2θ = 26.7o giving a d-value of 0.3340 nm. This is further evidence to show the effect of purity on lattice parameters of graphite crystallites. Although absolute values of intensities (Is) of XRD depend on the amount of diffracting atoms present in the sample, which in turn depends on the amount of material used to get the diffractogrammes, the normalized intensities obtained by dividing I(002) by I(004) are meaningful and independent of the amount of materials used. As revealed by XRD data, the floating of ball-milled graphite has increased this intensity ratio indicating that floating of ball-milled graphite has resulted in increasing the crystallinity of the sample. It is interesting to note that ball-milling of graphite brings the crystallite sizes variable range up to nanometre scale and floating the powder in water further decreases the crystallite size by 15 nm. The decreased crystallite size and the decreased interlayer distance of the FG collectively indicate the shrinking of particles due to floating which may be due to removal of impurities in the pristine graphite when it is milled and floated in water.
We assessed the impurity contents of original graphite, PG and FG samples as well as their carbon content variation from the data given in Table 1, 2 and Fig. 2. Impurities in vein graphite can basically occur either as mineral inclusions such as quartz, pyroxene, pyrrhotite, pyrite, chalcopyrite, sphalerite, marcasite, chlorite, calcite, siderite and dolomite and copper or as elements incorporated into the crystalline lattice3,5,15. Depletion of elements in low-carbon graphite indicates that most of impurities are incorporated as mineral inclusions. However, the rest of remaining such elements indicates the floated graphite still contains fine inclusion of some minerals which can be explained by EDS analysis given in Fig. 2. This shows that the mineral inclusions are present in graphite used for the present study before grinding and the carbon weight percentage around 84%. Silicate minerals such as quartz (SiO2), pyroxene group minerals (Mg,Fe)SiO3 or (CaxMgyFez)(Mgy1Fez1)Si2O6) and chalcopyrite (CuFeS2) as well as native element minerals such as copper (Cu) and iron (Fe) can be suggested from EDS elemental distribution of original graphite sample ( Fig. 2 and Table 1). Further, some silicate minerals with copper and iron also present in raw samples. These minerals may have formed due to the alteration process by hydrothermal fluids. However, the carbon percentage has increased in at least up to 97% as a result of removing mainly of silicate minerals during flotation process. However, the fine inclusion such as iron bearing minerals have not been fully removed and they have floated with graphite (Table 2) as some of them occur as nanometer scale minerals (see Fig. 2-b). Therefore, iron content has increased in FG samples. The enrichment of Ba in floated graphite indicates that the element is incorporated into the crystalline lattice. However, the final FG products in all products contain over 97.3% of carbon content (Table 2). Therefore final FC products are suitable for applications including nuclear reactors, furnaces, advanced materials, specific niche applications, expandable graphite products, composites and electronic applications. The final products can be marketed for USD 4000–6000 per ton according to graphite market in 201915. According to elemental results in Table 2, the PG used for the study have 91-95.9% carbon content and these PG has market value of USD 500–1100 per ton. Therefore after the proposed simple floatation technique reported in this study is capable of adding 4–12 times value addition, which is a huge value addition to graphite. This can be further increased by making PG finer powders and upon development of process such as magnetic separation of iron in final FG products.
Table 1
Elemental analysis of graphite samples by EDS
Elements
|
C
|
O
|
Si
|
S
|
Mg
|
Fe
|
Al
|
Cu
|
Sample FG weight %
|
96.54
|
2.66
|
0.23
|
0.18
|
b.d
|
0.26
|
0.12
|
b.d
|
Sample FG atomic %
|
97.67
|
2.02
|
0.10
|
0.09
|
b.d
|
0.06
|
0.06
|
b.d
|
Sample PG weight %
|
84.69
|
9.77
|
4.99
|
b.d
|
0.18
|
0.22
|
n.d
|
0.19
|
Sample PG atomic %
|
89.81
|
7.78
|
2.26
|
b.d
|
0.06
|
0.05
|
n.d
|
0.04
|
b.d – below the detection limits, n.d – not detected |
Table 2
Compositions of the graphite samples as determined by ICP-MS analysis
Sample
|
Purity
|
Carbon Content together with some minor elemental impurities as Weight%
|
|
|
|
Al
|
K
|
Fe
|
Ca
|
Mg
|
Ctotal
|
|
PG
|
High
|
< 0.01
|
< 0.01
|
0.03
|
0.05
|
0.01
|
95.9
|
|
FG
|
< 0.01
|
< 0.01
|
0.09
|
0.01
|
0.01
|
98
|
|
PG
|
Moderate
|
0.04
|
< 0.01
|
0.11
|
< 0.01
|
0.06
|
92.9
|
|
FG
|
0.04
|
< 0.01
|
0.12
|
< 0.01
|
0.06
|
97.3
|
|
PG
|
Low
|
0.14
|
< 0.01
|
0.21
|
0.09
|
0.14
|
91.9
|
|
FG
|
0.09
|
0.01
|
0.14
|
0.03
|
0.09
|
98
|
|
Trace Elements in ppm (Abundance is too low to be given as Weight %)
|
|
|
|
Ba
|
Bi
|
Cr
|
Cu
|
Mn
|
Mo
|
Pb
|
PG
|
High
|
7.3
|
1.7
|
6
|
24.6
|
6
|
0.2
|
1
|
FG
|
11.7
|
1.6
|
4
|
19.7
|
7
|
0.1
|
1
|
PG
|
Moderate
|
7.7
|
1.7
|
1
|
38.5
|
7
|
0.3
|
1.3
|
FG
|
8.1
|
1.6
|
< 1
|
38.1
|
7
|
< 0.1
|
1.5
|
PG
|
Low
|
6.2
|
0.4
|
14
|
19
|
31
|
0.6
|
1.3
|
FG
|
11.8
|
0.2
|
8
|
10.1
|
20
|
0.3
|
0.9
|
In Raman spectra, given in Fig. 3, three prominent peaks at 1347 cm-1, 1579 cm-1, and 2690 cm-1 correspond to the D band, G band, and 2D band for both PG and FG samples, respectively. Notably, the difference between two samples is the intensity ratio of the D band and G band (ID/IG), in which it was measured to be larger in the FG (0.10) than that of the PG (0.03) suggesting that the higher value of ID/IG results from the edges of graphite. In other words, with the same kind of graphite, the floated graphite has smaller crystalline size (nanographite) compared to the unfloated graphite, which is compatible with the XRD results.
The crystal structure of the PG and FG samples were further investigated by TEM analysis and the results are shown in Fig. 4 and 5, respectively. The size of the sheets ranged from 0.5 µm to 4 µm. Insets in Fig. 4(a) and 5(a) are the corresponding [001] SAED patterns. It is intriguing that distinct six-fold symmetry diffraction spots present more frequently in the FG specimen compared to the PG sample. On the other hand, the electron diffraction pattern of the PG exhibit rather ring-like reflections (polycrystalline), as shown in the inset of Fig. 5(a). It indicates that the FG shows either high crystallinity in nature or/and has no rotational boundaries in this area. The lattice parameters of PG and FG were measured to be 0.247 and 0.246 nm, respectively. To quantify the number of layers in graphite/graphene, a commonly used method is by counting the number of folds at the edge of the flakes/sheets in high resolution TEM images (HRTEM), as shown in Fig. 4(b) and 5(b). It was measured to be more than 68 layers in PG while FG has 16 layers. The interlayer distance of PG was measure to be 0.339 nm, corresponding to (002) graphite crystal spacing. However, FG has a relatively small interlayer spacing of 0.326 nm. This is consistent with the XRD results where the (002) peak of FG shifted toward the higher 2q angle resulting in a smaller in lattice spacing.
In order to distinguish the features more easily and to remove contrast due to the presence of amorphous materials, HRTEM images were filtered by Fast Fourier Transform (FFT) method (Fig. 4(c) and 5(c)). The corresponding FFT images are shown in Fig. 4(d) and 5(d). It should be noted that the HRTEM images were taken at featureless regions, in which these might be area with few-layer graphene. The interplanar spacings along (100) planes were measured to be 0.230 nm and 0.242 nm for the PG and FG, respectively, which are closely matching with the lattice parameter calculated from the diffraction patterns. However, it should be aware that the periodicity in the region might not be the original position of the spots due to the instrumental limitations.
Figure 6 depicts the FT-IR spectra of PG and FG samples. The spectra of both samples are basically very similar with a broad band between 3200 cm− 1 to 3700 cm− 1 resembling the O-H vibrations of adsorbed water molecules on graphite surfaces and narrow band centered at 1622 cm− 1 due to C = C stretching of conjugated double bonds that are present in graphene layers. The absence of bands at 2925 cm− 1 (asymmetric C-H stretching in CH2 groups) 2855 cm− 1 (symmetric C-H stretching in CH2 groups) suggest that there are no detectable amounts of saturated sp3 carbon atoms in both PG and FG samples. In other words, there are no detectable defects due to hydrogenated double bonds in both samples. Absence of C = O vibration centered at 1738 cm− 1 shows that the unsaturated carbon atoms do not contain any carbonyl functionality and hence the materials contain only conjugated C = C in their graphene sheets.
XPS was used to monitor the changes in core level and valence band structure of the PG and FG samples. The survey scan spectra (Fig. 7) of PG (a) and FG (b) samples indicate the presence of C and O atoms. It is found that the oxygen content of the FG sample is much higher, though the absolute percentages are very small in both types, compared to the PG sample. The oxygen to carbon ratio (O/C) was measured to be 0.04 PG sample and 0.05 in FG sample. The C 1s spectra of the PG and the FG samples are shown in Fig. 7 (c) and (d). Both are fitted with four Lorentzian-Gaussian peaks of 20:80 ratios. The most intense peak, at 284.5 eV, is assigned to sp2 C = C bonds atoms, together with the weak component at 290.1 eV that corresponds to its π- π transition (signature of graphitic carbon). The component at 286.5 eV has been usually attributed to C-O-C (ether) bonds. It is noted that two samples exhibit similar patterns. However, the C1s spectrum of floated graphite displays higher intensity of the C-O-C and C-C components and lower intensity of the C = C and π- π bonds. It is generally accepted that C-O-C bonds are formed at the edge of the graphene sheets. As shown in the Raman spectra, the higher D peak represents to higher amounts of edges in FG sample and those edges might be the spots for C-O-C bonds to form. Table 2 shows the different bonds present in both PG and FG. With respect to PG there is 4% reduction of C = C bonds, 3% increase in C-C and 11% increase in C-O-C in FG.
The present study is the first attempt to understand the characteristics of the floated graphite in de-ionized water. This simple floatation technique has been introduced to float ground graphite particles with mineral impurities on the surface of de-ionized water and floatation has been affected with the aid of shaking. This technique does not require any froth formers or floatation aid chemicals and as such it is highly industrially viable and environmentally friendly. The powdered and floated graphite samples have some differences in their crystallite sizes, morphologies, and purities. Floatation has resulted in shrinking of crystallite size due to the removal of mineral inclusions within the interlayer spaces. Number of defects has been increased in floated graphite and both samples have some C-O-C ether bonds in slightly higher amount in the floated graphite. The number of layers present in crystallites has been remarkably decreased in floated graphite when compared to that in powdered graphite that did not subjected to floating in water. The floatation technique can remove many impurities which are present as mineral inclusions in graphite, to some extent. The floatation aids the remarkable enhancement of carbon content in graphite samples and is more prominent when the original samples are impure with lower carbon percentages which leads in 4–12 times high value addition to graphite. This technique can be applied to vein graphite rich in silicate minerals (especially along the wall zones of graphite veins) which are considered as waste of mining products. Further purification of final graphite product is possible by removing iron components of final FG product by magnetic separation as well as using very fine graphite powders for the flotation process.