Synthesis of HTC

1.5 g of glucose [D-(+), Sigma-Aldrich] and 20 mL of deionized water were placed into a Teflon-lined stainless steel autoclave. The autoclave was then sealed and placed in an oven which was set to 200 °C for 48 h. After the autoclave was cooled down, obtained carbons were washed with deionized water several times, filtered, and dried at 105 °C for 4 h.

Synthesis of active materials

40 mg HTC from glucose and 120 mg (NH₄)₆Mo₇O₂₄.4H₂O (Sigma-Aldrich) were added in 10/10 mL distilled water/ethanol solution. After stirring for 30 min the mixture was transferred into the autoclave, and it was placed into an oven which was set to 200 °C for 18 h. After filtration, the product was dried at 105 °C for 12 h, and then calcined at 600 °C for 6 h in an argon atmosphere. The hybrid material was labeled HTC–MoO₂. Another active material was obtained by using 40 mg HTC of glucose, 120 mg (NH₄)₆Mo₇O₂₄.4H₂O, and 120 mg (NH₂)₂CS (Merck) with the same procedure and it was labeled HTC–MoS₂. The last active material was obtained by just calcinating HTC from glucose at 600 °C for 6 h in an argon atmosphere, and it was labeled HTC.

Materials characterization

The morphology of the carbon materials was characterized by SEM (Nova Nano SEM 430, FEI) and TEM (Philips CM 300 FEG/UT, 300 kV). For elemental analysis, Hekatech CHNS EuroEA 3000 Analyzer was used. Hybrid materials were analyzed by XRD (STOE STADI MP) using Cu K_a radiation (λ = 1.5406 Å). XPS was performed with an ESCA M-Probe (Al K_α rays, λ = 8.33 Å). Survey scans were performed with a detector pass energy of 158.9 eV, high-resolution spectra were recorded with a pass energy of 22.9 eV. All spectra were referenced to adventitious carbon signal at 284.8 eV. All spectra were fitted using GL(30) line shape with CasaXPS software from Casa Software Ltd.

Cell fabrication and electrochemical measurements

For the electrochemical measurements, three different active materials (HTC, HTC–MoO₂, HTC–MoS₂) were separately prepared in vials. Active material, graphite (Super P Li, Timcal) and polyvinylidene fluoride (PVDF, Alfa Aesar) were added in a weight ratio of 70:20:10. After addition of several drops of N-methyl-2-pyrrolidione (NMP, Sigma-Aldrich) on this solid mixture, a viscous slurry was obtained and stirred at the room temperature for 48 h. Thus, the slurry was spread on copper foil using doctor blade and dried in a vacuum oven at 110 °C for 6 h, to construct anodes for battery testing.

Coin half-cell batteries were prepared inside of an argon-filled glove box (O₂ and H₂O levels <1 ppm, MBraun) using anodes, 1 M LiPF₆ in a mixture of ethylene carbonate (EC), ethyl methyl carbonate (EMC), and diethyl carbonate in a weight ratio of 37:46:17 (BASF), a polyethylene separator membrane and lithium metal foil (99.9%, Sigma-Aldrich) as a counter electrode. The galvanostatic measurements were performed with current density of 0.1 A/g and within a potential range of 0.01–3.00 V. The cyclic voltammetry measurements were performed with 0.1 mV/s scan rate and within 0.01–3.00 V potential limits. For the cyclic voltammetry, we use the Metrohm Autolab PGSTAT 101. Galvanostatic measurements were performed with the MTI Corp. BST8-MA battery analyzer.

Structural characterization of carbon materials

The morphology of the as-prepared HTC derivatives was studied with SEM and TEM. Figure 1 shows that the HTC particles are spherical in shape [Fig. 1(a)]. Calcination of HTC samples did not show any significant changes on the particle morphology based on our previous records.^[Reference Simsir, Eltugral and Karagoz^14] The size of most spheres was in the range of 300–400 nm. Besides, particles with sizes about 700–800 nm and over 1 µm were observed which are in the form of hard spheres as defined previously.^[Reference Wang, Li, Chen and Huang^15, Reference Titirici, White, Brun, Budarin, Su, del Monte, Clark and MacLachlan^16]

Figure 1.

Representative SEM and TEM images (the insets) of (a) HTC, (b) HTC–MoO₂, and (c) HTC–MoS₂.

It can be seen from the SEM and TEM images given as the inset in Fig. 1(b) that MoO₂ nanocrystals are aggregated with irregular shape and size on the carbon spheres. Judging from this image, the thickness of the MoO₂ coating is about 25 nm. As it can be seen from the images, MoO₂ nanocrystals do not cover the whole surface as a shell with a uniform coating, instead, they are aggregated in the nodular form. Similar results were recorded elsewhere as well.^[Reference Chen, Di, Ma, Zhu, Gao, Li, Sun and Ouyang^9, Reference Zeng, Zheng, Deng, Ding and Wei^10, Reference Ihsan, Wang, Majid, Yang, Kennedy, Guo and Liu^17] On the other hand, it was observed that MoS₂ nanocrystals were uniformly distributed in HTC without having any uncoated space over the HTC surface. SEM photo depicts that the morphology of the MoS₂ aggregation in HTC can be described as a uniformly dispersed wormlike structure with a coating thickness, judging from SEM image, about 46 nm making the HTC–MoS₂ sphere core–shell structures [Fig. 1(c)]. TEM image of this structure revealed coating of MoS₂ as nanosheet on HTC.

Table I shows the elemental compositions of the glucose (starting material) and its hydrothermal derivatives before and after calcination. The carbon ratio which was 39.64 wt% initially increases to 66.80 wt% with lowered hydrogen content as the hydrothermal carbonization occurs at 200 °C. After the calcination of as prepared HTCs, carbon content further increases to a level of 92.06 wt%.

Table I.

Elemental composition (wt%) of glucose and HTCs.

^a By difference.

^b Atomic ratio.

Furthermore, XRD analysis also supports the results obtained from SEM analysis. In Fig. 2(a), all the peaks, 26°, 37°, 53°, 60°, and 66° of the MoO₂ (JCPDS 73-1249) are observed in the XRD pattern of the HTC–MoO₂ sample. Similarly, looking at the HTC–MoS₂ XRD pattern [Fig. 2(b)], all reflections at 14°, 33°, 36°, and 58° of the MoS₂ (JCPDS 75-1539) are clearly visible in the structure. However, at 14°, 33°, and 36° are shifted compared with the reference. This may be due to the formation of thin MoS₂ nanosheets, as seen from the corresponding TEM image.^[Reference Liu, Chen, Deng, Su, Guo and Zhu^12] The XRD analysis revealed that HTCs were coated with MoO₂ and MoS₂ nanocrystals, respectively.

Figure 2.

XRD patterns of (a) HTC–MoO₂, (b) HTC–MoS₂.

XPS measurements were carried out for HTC–MoO₂ and HTC–MoS₂ samples to investigate the composition of the sample surface. The XPS spectra were given in the Supplementary Material section. The results for HTC–MoO₂ are shown in Fig. S1. In the survey spectrum carbon (63.51%), oxygen (25.17%), molybdenum (10.53%), and minor amounts of sulfur (0.70%) were detected [Fig. S1(a)]. The high-resolution spectrum of the O 1s region showed signals for molybdenum trioxide at 530.72 eV as well as for hydroxides, organic oxygen species and defect oxide at 531.84 eV [Fig. S1(b)]. Contrary to the results from the bulk analysis via XRD the Mo 3d/S 2s region revealed that only MoO₃ with its components at 232.62 and 235.77 eV was present on the surface of the sample, which was mainly due to sample storage under ambient conditions [Fig. S1(c)]. The position of the S 2s peak at 229.7 eV indicated that trace amount of sulfur detected in its elemental form and not bound to molybdenum. The most probable explanation for this contamination was the usage of the same teflon vessels for the hydrothermal reactions of HTC–MoS₂ and HTC–MoO₂ samples.

The results for HTC–MoS₂ are shown in Fig. S2. The survey spectrum showed also the presence of carbon (19.28%), oxygen (16.18%), molybdenum (26.37%), and sulfur (38.17%) [Fig. S2(a)]. In comparison with HTC–MoO₂, the significantly lower amount of carbon could be explained by the much higher surface coverage of the MoS₂ sheets according to the SEM results. The O 1s region contained again a peak for molybdenum oxide (MoO₃ and Mo₂O₅) at 530.75 eV, which was an evidence for the oxidation of the particle surface under ambient conditions [Fig. S2(b)]. Next to the oxide peak also signals for the hydroxides, organic oxygen, and defect oxides at 531.60 eV on the one hand, and adsorbed water and further organic oxygen species at 532.66 eV one the other hand could be observed. The S 2p peak showed components for sulfides (63.36%) at 162.57 and 163.75 eV as well as for elemental sulfur (36.64%) at 164.20 and 165.38 eV [Fig. S2(c)]. The sulfides could be related to the duplet for MoS₂ at 229.98 and 233.13 eV in the Mo 3d/S 2s region and again be confirmed by the position of the sulfidic S 2s component at 227.34 eV. In addition, the S 2s component for the elemental sulfur could be found at 229.72 eV [Fig S2(d)]. Due to surface oxidation under ambient conditions, further duplets for Mo₂O₅ at 231.59 and 234.74 eV and MoO₃ at 233.5 and 236.65 eV, respectively, were present in the sample. The ratio between the MoS₂, Mo₂O₅, and MoO₃ was observed to be nearly 2:2:1.

Electrochemical performance of carbon anodes

The charge–discharge measurements of the prepared Li-ion batteries were made at potential values between 0.01 and 3.00 V, at a current density of 100 mA/g. It was observed that in the 1st cycle the Li-ion battery using HTC as the anode material has discharge and charge capacities in the 1st cycle of 755 and 381 mAh/g, respectively. The irreversible capacity loss is 49%. In the 2nd cycle, those capacities become 382 and 348 mAh/g, respectively, resulting in a much higher coulombic efficiency of 91% which reaches up to 98% at the end of the 50th cycle [Fig. 3(a)]. In the case of the Li-ion battery prepared by using HTC–MoO₂ anode material, the initial discharge and charge capacities are 734 and 454 mAh/g, respectively. In the 2nd cycle, these capacities become 462 and 505 mAh/g, respectively. The coulombic efficiency is 89%, which goes up to 98% at the end of the 50th cycle with the discharge, and charge capacities of 266 and 263 mAh/g, respectively [Fig. 3(b)]. HTC–MoO₂ was found to have better cyclic performance than that of bulk MoO₂ at similar conditions from the previous reports where bulk MoO₂ was shown to have the initial discharge capacity of 543 mAh/g and then decreases to 73 mAh/g at the end of 50 cycles.^[Reference Zeng, Zheng, Deng, Ding and Wei^10, Reference Sun, Hu, Luo and Huang^18, Reference Yang, Liu, Zhu, Wang and Wu^19] On the other hand, the discharge capacity of the Li-ion battery made by using HTC–MoS₂ material is 868 mAh/g, while the charge value is 642 mAh/g in the 1st cycle. They are 551 and 540 mAh/g, respectively, in the 2nd cycle. The coulombic efficiency is 98%, and it goes up to 99% at the end of the 50th cycle with the discharge and charge capacities of 484 and 483 mAh/g, respectively [Fig. 3(c)]. From previous reports, initial discharge and charge capacities of pure MoS₂ were given as 451 and 400 mAh/g, respectively. However, its capacity sharply drops to 110 mAh/g after 60 cycles.^[Reference Liu, Chen, Deng, Su, Guo and Zhu^12] Therefore, our results indicate that HTC–MoS₂ shows improved specific capacity compared with the pure MoS₂.

Figure 3.

Charge–discharge curves of (a) HTC, (b) HTC–MoO₂, (c) HTC–MoS₂.

The electrochemical performance of the molybdenum-supported hybrid composites of the HTC was comparable with those reported in the literature as shown in Table II. The specific capacities of synthesized HTC hybrid composites seemed to be lower than that of the composites of graphene or multiwalled carbon nanotubes (MWCNT). Notably, HTC can be prepared from any type of biomass with a green and simple process. Hence, it is more likely in the industrial area of applicability of such material.

Table II.

Electrochemical performance comparison of the present work with some previously published reports

The irreversible capacity loss is due to the irreversible reaction of Li-ions and formation of SEI layer that deposits on the anode surface.^[Reference Whittingham^2–Reference Shi, Guo, Corr, Shi, Hu, Heier, Chen, Seshadri and Stucky^4] By observing the charge–discharge measurement, the irreversible capacity loss appears to be the largest at HTC. This effect was reduced when the HTC surface was coated with MoO₂ or MoS₂. For all measurements, a very high coulombic efficiency value was obtained after the 2nd cycle. Particularly, in the HTC–MoS₂, 98% efficiency was obtained at the 2nd cycle. Our findings indicate that all the tested anode electrodes exhibited good cycling performance at a current density of 100 mA/g. HTC and HTC–MoO₂ exhibited similar reversible discharge capacities of 296 and 266 mAh/g, respectively after 50 cycles. Furthermore, the reversible discharge capacity of the tested anode material, HTC–MoS₂ at the same condition was 484 mAh/g which is higher than that of graphite. When we switch from HTC to HTC–MoS₂ anode material, the reversible capacity increases from 296 to 484 mAh/g and the capacity value is retained for 50 cycles. The reason for this can be attributed to the fact that MoS₂ in the form of nanosheets is dispersed uniformly all around the carbon surface and they offer more sites for Li-ion intercalation. This intrinsic structure provides a more facile Li-ion storage as well as shortened pathways for efficient diffusion of the ions and transportation of the electrons.^[Reference Zeng, Zheng, Deng, Ding and Wei^10, Reference Liu, Chen, Deng, Su, Guo and Zhu^12]

Figure 4.

CV measurements of (a) HTC, (b) HTC–MoO₂, (c) HTC–MoS₂.

We investigated the electrochemical changes in the structure during the charge–discharge of the Li-ion battery obtained by using the HTC anode with the cyclic voltammetry measurements [Fig. 4(a)]. According to the results obtained from these measurements, the peaks observed between 0.10 and 0.21 V are due to the electrolyte deterioration. The peak at 0.50 V seen in the 1st cycle and not visible in the other cycles is due to the solid–electrolyte interaction.^[Reference Liu, Chen, Deng, Su, Guo and Zhu^12, Reference Li, Xu, Tan, Wang, Holt, Stephenson, Olsen and Mitlin^24] The peaks observed at 0.96 and 1.14 V^[Reference Xing, Ju, Zhao, Wan, Zhu, Qiang and Qian^25] correspond to reduction and oxidation of Li-ion, respectively, on the electrode surface. Lithium insertion/extraction at carbon structures can be written as Eq. (1).

(1)$${\rm Li} + 6{\rm C}\; \leftrightarrow \; {\rm Li}{\rm C}_6. $$

The Li-ion battery obtained using the HTC–MoO₂ anode shows two reduction/oxidation peak pairs at about 1.20/1.52 and 1.46/1.73 V which are attributed to the phase transition of partially lithiated Li_xMoO₂ during insertion/extraction of lithium at the anode side [Fig. 4(b)].^[Reference Auborn and Barberio^26] At potential values below 1.00 V, it is known that peaks arise from electrolyte depletion and solid–electrolyte interaction. Therefore, the potential values between 1.00 and 3.00 V were measured.^[Reference Liu, Chen, Deng, Su, Guo and Zhu^12, Reference Li, Xu, Tan, Wang, Holt, Stephenson, Olsen and Mitlin^24] There are some differences in the intensity of the peaks obtained at the end of the ongoing conversions. These differences are due to the degradation of the anode which is caused by lithium insertion and extraction. Lithium insertion/extraction on the MoO₂ structure can be written as Eq. (2).^[Reference Zeng, Zheng, Deng, Ding and Wei^10]

(2)$$x{\rm L}{\rm i}^ + + xe^ - + {\rm Mo}{\rm O}_2\; \leftrightarrow {\rm L}{\rm i}_x{\rm Mo}{\rm O}_2. $$

Figure 4(c) shows the first three discharge/charge profiles of the HTC–MoS₂. In the first cathodic scan, the peak at 1.22 V can be ascribed to the phase transformation of MoS₂ from trigonal prismatic (2H) to octahedral (1T) structure. This peak was observed at 1.22 V shifts to 1.33 V in the following cycles.^[Reference Wang, Wei, Ozawa, Cai, Cheng and Kimura^23, Reference Ma, Zhou, Xu, Xu, Zhang and Chen^27] Continuously, the peak at 0.52 V can be assigned to the formation of SEI. Furthermore, this peak was absent in the subsequent cycles. In the 2nd and 3rd cycles, a broad new peak at 1.84 V shows that MoS₂ is reduced to Mo metal and Li₂S according to the following reaction [Eq. (3)].^[Reference Yuan, Wang, Wang and Bai^21]

(3)$${\rm Mo}{\rm S}_2 + 4{\rm Li}\; \leftrightarrow {\rm Mo} + 2{\rm L}{\rm i}_2{\rm S}{\rm.} $$

The peaks at 1.71 and 2.21 V potential values are oxidation peaks and can be attributed to the oxidation of Mo to form MoS₂ and oxidation of Li₂S to S, respectively.^[Reference Xiao, Choi, Cosimbescu, Koech, Liu and Lemmon^28, Reference Liu, Su, Zhou, Sun, Wang and Qiao^29] Lithium insertion/extraction on MoS₂ structure can be written as Eq. (4).^[Reference Wang, Wei, Ozawa, Cai, Cheng and Kimura^23]

(4)$${\rm Mo}{\rm S}_2 + x{\rm L}{\rm i}^ + + xe^ - \leftrightarrow \; {\rm L}{\rm i}_x{\rm Mo}{\rm S}_2. $$