Colloidal synthesis of marcasite FeS₂ nanoparticles with improved electrochemical performance

Abstract

Marcasite FeS₂ nanoparticles were synthesized for the first time in colloidal solution via a hot-injection protocol. In comparison with the previous iron sulfides, marcasite FeS₂ presents better lithium ion storage and charge–discharge performance as the anode materials in lithium ion battery application.

---

Iron sulfide nanomaterials are promising anode materials for lithium ion batteries (LIBs) owing to the advantages of high-capacity, environmental friendliness, low cost, and so forth.¹ Although lots of works have been devoted to troilite FeS, greigite Fe₃S₄, and pyrite FeS₂ as anode materials, the development of high performance iron sulfides anodes is still challenging.^2–4 For instance, pyrite FeS₂ nanoparticles (NPs) are one of the most potential anode materials for LIBs, because of the high theoretical capacity of 890 mA h g⁻¹.⁵ However, the real discharge capacity of pure pyrite NPs is only half of the theoretical capacity. This is attributed to the dependence of LIBs performance upon intercalation/extraction-related lithium ion diffusion pathways, which is very sensitive to the crystalline structure/phase of electrode materials.^6,7 In this scenario, pyrite FeS₂ possesses cubic crystal structure (space group Pa3̄, Z = 4) with Wyckoff position 4a(0,0,0) for Fe and 8c(u,u,u) for S.⁸ According to the previous analysis, such structure is not the best for lithium ion diffusion.^9,10 To improve the LIBs performance of FeS₂, the NPs with optimized crystal structure should be investigated.

Marcasite is another natural mineral of FeS₂, which commonly distributes in hydrothermal systems and sedimentary rocks.¹¹ Until now, marcasite FeS₂ nanomaterials have never been synthesized by chemical method. The investigations merely rest on theoretical prediction and the direct utilization of natural marcasite.^12,13 As to the orthorhombic structure of marcasite FeS₂, the space group is Pnnm, Z = 2 with Wyckoff position 2a(0,0,0) for Fe and 4g(u,v,0) for S.¹⁴ Such structure is considered better than pyrite for lithium ion diffusion.¹⁰ However, limited by the rare synthesis of orthorhombic FeS₂ nanomaterials, the LIBs performance of marcasite has not been tested yet. In the conventional synthesis of FeS₂ NPs, oriented attachment (OA) and fusion of small FeS₂ clusters is the major process for NP growth, which favors the formation of pyrite NPs with cubic structure.¹⁵ According to the reports in synthesizing II–VI semiconductor NPs, Ostwald ripening (OR) process via monomer diffusion may facilitate the growth of orthorhombic NPs.^16,17 Therefore, marcasite FeS₂ NPs are considered to be produced by converting the growth kinetics of FeS₂ NPs from OA to OR.

In this work, we demonstrate the facile and controlled synthesis of marcasite NPs with orthorhombic crystalline structure for the first time in colloidal solution via hot-injection protocol. The as-synthesized NPs are well dispersible in non-polar solvents. The formation of marcasite FeS₂ NPs rather than pyrite is attributed to the OR-determined growth of NPs, which is quite different from the previous OA-determined process. The as-synthesized marcasite NPs are further fabricated to coin cell batteries, which exhibit low electrochemical resistance and high activity in lithiation/delithiation cycling process.

As shown in the Experimental section of ESI,† marcasite FeS₂ NPs are synthesized by injecting a sulfur diphenyl ether (DE) solution into the mixtures of Fe(acac)₃, DE, and dodecanethiol (DT) at 220 °C and stirring for 1 h. After purification, the products are characterized in Fig. 1. As revealed by transmission electron microscopy (TEM), the products appear as spherical particles with the average diameter of 7.1 nm and a size deviation of 40% (Fig. 1a). Slight aggregation of NPs is observed by TEM, which may be attributed to two reasons. One is the high surface energy leading from the small size of NPs. The other is the intrinsic ferromagnetism of FeS₂. Under high-resolution TEM (HRTEM), an interplanar spacing of 0.276 nm is observed (Fig. 1b), which corresponds to the (101) plane of marcasite FeS₂.¹⁸ No the sign of oxide coating or amorphous overlay is found. The selected area electron diffraction (SAED) pattern indicates that the NPs are single-phase marcasite FeS₂ without other iron sulfide signal (Fig. 1c). Low resolution TEM image and the corresponding energy-dispersive X-ray spectroscopy (EDS) elemental maps of Fe and S are displayed in Fig. 1d–f, which reveals the homogenous distribution of Fe and S elements. The products possess the Fe/S molar ratio of 1/1.97 from the EDS data (Fig. 1g), which consists well with the stoichiometric ratio of FeS₂. Fig. 1h presents the X-ray diffraction (XRD) pattern. All the diffraction peaks can be exclusively attributed to the marcasite FeS₂ with orthorhombic crystal phase.¹⁸ This crystal phase consists thus of an hcp stacked sulfur atoms, where the tetrahedra between a pair of hexagonal sheets is empty and the octahedra is alternatively empty or occupied by iron atom.⁸ Such atomic arrangement is greatly different from pyrite.

Fig. 1 TEM (a) and HRTEM (b) images, SEAD pattern (c), STEM-HAADF image (d) and element maps (e and f), EDS spectrum (g), and XRD pattern (h) of the as-synthesized marcasite FeS₂ NPs. Inset in (a): size distribution of the NPs. Inset in (h): the schematic atomic arrangement of marcasite FeS₂ NPs.

The as-synthesized marcasite NPs are well dispersible in non-polar solvents due to the capping ligands of DT. The uniform dispersion is revealed by the classical Tyndall effect (Fig. 2a).¹⁹ In addition, the NPs exhibit weak ferromagnetism that can be separated from the solution with the help of a magnet (Fig. 2b). The saturation magnetization is 6.95 emu g⁻¹, which is among the highest level of ever reported small iron sulfides NPs. This further confirms the products are marcasite FeS₂ rather than pyrite, because pyrite FeS₂ has no magnetism. As revealed by TEM, the as-synthesized marcasite NPs are stable as heating at 300 °C for 2 h under Ar atmosphere (Fig. S1a†). TGA profile shows that the weight loss at 100, 217, 376, and 900 °C is 2.3, 9.8, 18.1, and 27.6%, respectively. The weight loss of the first three is attributed to water, ligand, and the phase transformation from marcasite to pyrite. Accordingly, the calculated content of marcasite FeS₂ and the capping ligand of DT in the NPs is 90.2 and 7.5%, respectively.

Fig. 2 (a) UV-vis absorption spectrum of marcasite FeS₂ NPs. Inset: the optical image of NPs chloroform solution with Tyndall effect. (b) The hysteresis loop of marcasite FeS₂ NPs. Inset: the optical image of NPs chloroform solution without and with a magnet.

The formation of marcasite FeS₂ NPs is led from the hot-injection protocol, which converts the growth kinetics of FeS₂ NPs from OA to OR, because the crystal phase of FeS₂ NPs greatly depends on the formation mechanism.²⁰ The temporal evolution of the NPs is monitored by TEM images and absorption spectra (Fig. S2 and S3†). Before the injection of sulfur solution, the nucleus already form by the reduction of Fe(acac)₃ and the S released from DT decomposition (Fig. S2a†). As revealed by TEM, the average diameter of nucleus is 2.7 nm (Fig. S2b†). Subsequently, the S solution is injected into the Fe solution to supply the further growth of FeS₂ NPs. Accompanied with the growth of NPs, the number of original nucleus rapidly reduces within 10 min (Fig. S2c–f and S3†). According to the numerous reports about the growth of colloidal NPs, such property is typical for OR process. Namely, the growth of bigger NPs is supplied by the dissolution of smaller ones through a monomer diffusion controlled process (Fig. S2g†).²¹ In comparison, OA mechanism involves the attachment of the specific lattice plane of small particles and further fusion, which are not observed in the current synthesis of marcasite FeS₂ NPs.²⁰ On the basis of OR growth mechanism, the size of marcasite FeS₂ NPs is further controlled by regulating the reactivity of Fe monomers. As a dominant factor in controlling Fe reactivity, the growth temperature is altered from 180 to 260 °C, while other parameters are fixed. Accordingly, the diameter of NPs increases from 3.5, 5.0, 10, to 18 nm (Fig. S4†). Another factor that influences Fe reactivity is the amount of DT, which acts as the capping ligand (Fig. S5†). As increasing DT amount from 0.5 to 5 mL (2 to 20 mmol), the NPs size increases from 3.4 to 8.1 nm.

The good structural stability of the as-synthesized marcasite FeS₂ NPs under heating treatment permits the application as LIBs anode materials (Fig. S1a†). The electrochemical properties of marcasite FeS₂ NPs based electrode are further evaluated (Fig. 3). Electrochemical impedance spectroscopy (EIS) of an unused coin cell is firstly examined from 10 kHz to 10 MHz (Fig. 3a). In the higher frequency region, there is a semicircle that is attributed to the charge transfer resistance, which relates to the reaction between the active materials and electrolyte interface.³ The charge transfer resistance of the electrode is approximately 50 ohm, which is the smallest value of ever reported FeS₂ materials.²² The straight line, named Warburg line, in the low frequency region is attributed to the ion diffusion from electrolyte to the electrode interface.²³ On the basis of the precondition of good conductivity, marcasite FeS₂ NPs should be ideal electrode materials in the coin cell.

Fig. 3 (a) EIS profile of a coin cell with marcasite FeS₂ NPs. (b) Galvanostatic charge–discharge curves of marcasite FeS₂ electrodes in the voltage range of 0.005–3.0 V (versus Li) at the current density of 100 mA g⁻¹ for the 1^st, 2^nd, 10^th and 25^th cycles. (c) CV profiles of the first four cycles for the electrode at a scanning rate of 0.5 mV s⁻¹ in the voltage window of 0.01–3.0 V (Li⁺/Li). (d) Cycle performance of the electrode at a current density of 100 mA g⁻¹.

In order to verify the performance and the reaction inside of a coin cell, the galvanostatic charge–discharge curve with the selected cycles is evaluated in the voltage window of 0.005–3 V (vs. Li⁺/Li) at a current density of 100 mA g⁻¹. Fig. 3b shows that the open-circuit voltage of the cell is 2.42 V. At the reaction of the charge–discharge stage, the voltage rapidly declines at round 1.36 V to reach a plateau set until the capacity of 348 mA h g⁻¹. Both the reduction from Fe²⁺ to Fe⁰ and the lithium transfer to FeS₂ are capable to cause the discharge plateau.⁴ Another voltage plateau appears at about 0.8 V, which is observed only for the first cycle. It is associated with the formation of solid electrolyte interlayer (SEI) film.²⁴ To completely understand the reaction on the electrode, the cyclic voltammetry (CV) is evaluated in the voltage range of 0.05–3.0 V with the scanning rate of 0.5 mV s⁻¹ (Fig. 3c). The reduction peaks at 0.67, 1.16, 1.29, and 1.67 V are indicative of the single-phase insertion of Li and reduction of Fe²⁺, accompanying with the formation of the SEI film.⁵ The peaks at 1.51 and 2.83 V correspond to the oxidation of Fe⁰.⁴ At the first cycle, the internal reaction of a coin cell is complicated. Subsequently, the current intensity decreases rapidly, implying the existence of an irreversible process in the charge transfer.

The cycling performance of the coin cell at a current density of 100 mA g⁻¹ is shown in Fig. 3d. The discharge capacity decreases continuously from 1240 mA h g⁻¹ at the first cycle to 210 mA h g⁻¹ at the 16^th cycle. The huge first cycle loss is mainly attributed to the formation of solid electrolyte interlayer (SEI) film between the electrolyte and FeS₂-based electrode, which possesses larger specific surface area due to the small size of marcasite FeS₂ NPs.³ This SEI film consumes a large amount of Li⁺. As a result, the charge–discharge capacity decreases for the first cycle. The irreversible loss at first ten cycles is attributed to the side reaction between Li⁺ and active materials.⁵ Another reason for capacity decrease is attributed to the defects in marcasite FeS₂ NPs before the 16^th cycle. After the 16^th cycle, the system builds a balance between active materials and Li⁺ diffusion. So, the capacity does not decrease any longer. Afterwards, the discharge capacity indicates a steady rising to 641 mA h g⁻¹ until 100^th cycle, accompanying with almost 100% coulombic efficiency that suggests the good cyclability of the marcasite FeS₂ electrode.²⁵ According to the previous reports, the increase of capacity from the 16^th to the 100^th cycle indicates an activation process in the FeS₂ electrode.²⁶ As far as we know, both the initial discharge capacity of 1240 mA h g⁻¹ and the balance discharge capacity of 641 mA h g⁻¹ are highest for ever reported iron sulfides as LIBs anode.^2–5 Furthermore, the marcasite FeS₂ NPs with different sizes are employed for electrochemical tests. The NPs with average diameter of 7.1 nm exhibit better electrochemical performance than the NPs with other sizes. The reason is attributed to the dependence of battery performance on the surface area and migration distance of the building blocks.²⁷ Associated with the low resistance and stable charge–discharge performance, the marcasite FeS₂ NPs are promising electrode materials in LIBs application.

In summary, we demonstrate the hot-injection synthesis of marcasite FeS₂ NPs in colloidal solution. The formation of marcasite FeS₂ rather than conventional pyrite is attributed to the OA growth mechanism. Most importantly, marcasite NPs exhibit better electrochemical performance as LIBs anode than pyrite and other iron sulfides.

Acknowledgements

This work was supported by the 973 Program of China (2014CB643503), NSFC (51425303, 21374042, 21174051), Natural Science Foundation of Jilin Province (20140101048JC), the Special Project from MOST of China, and the Fundamental Research Funds for the Central Universities.

Notes and references

1. X. Rui, H. Tan and Q. Yan, Nanoscale, 2014, 6, 9889.
2. C. Xu, Y. Zeng, X. Rui, N. Xiao, J. Zhu, W. Zhang, J. Chen, W. Liu, H. Tan, H. H. Hng and Q. Yan, ACS Nano, 2012, 6, 4713.
3. G. Li, B. Zhang, F. Yu, A. A. Novakova, M. S. Krivenkov, T. Y. Kiseleva, L. Chang, J. Rao, A. O. Polyakov, G. R. Blake, R. A. de Groot and T. T. M. Palstra, Chem. Mater., 2014, 26, 5821.
4. L. Li, M. Cabán-Acevedo, S. N. Girard and S. Jin, Nanoscale, 2014, 6, 2112.
5. J. Liu, Y. Wen, Y. Wang, P. A. van Aken, J. Maier and Y. Yu, Adv. Mater., 2014, 26, 6025.
6. S. I. Nishimura, G. Kobayashi, K. Ohoyama, R. Kanno, M. Yashima and A. Yamada, Nat. Mater., 2008, 7, 707.
7. H. Wang, X. Wang, L. Wang, J. Wang, D. Jiang, G. Li, Y. Zhang, H. Zhong and Y. Jiang, J. Phys. Chem. C, 2015, 119, 10197.
8. V. K. Gudelli, V. Kanchana, S. Appalakondaiah, G. Vaitheeswaran and M. C. Valsakumar, J. Phys. Chem. C, 2013, 117, 20120.
9. H. Tang, J. Wang, H. Yin, H. Zhao, D. Wang and Z. Tang, Adv. Mater., 2015, 27, 1117.
10. Y. Tang, Y. Zhang, J. Deng, D. Qi, W. R. Leow, J. Wei, S. Yin, Z. Dong, R. Yazami, Z. Chen and X. Chen, Angew. Chem., Int. Ed., 2014, 53, 13488.
11. D. Rickard and G. W. Luther, Chem. Rev., 2007, 107, 514.
12. R. Sun, M. K. Y. Chan and G. Ceder, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 83, 235311.
13. J. B. Murowchick and H. L. Barnes, Geochim. Cosmochim. Acta, 1986, 50, 2615.
14. M. Nespolo and B. Souvignier, Cryst. Res. Technol., 2015, 50, 442.
15. Y. Bai, J. Yeom, M. Yang, S. H. Cha, K. Sun and N. A. Kotov, J. Phys. Chem. C, 2013, 117, 2567.
16. D. V. Talapin, A. L. Rogach, E. V. Shevchenko, A. Kornowski, M. Haase and H. Weller, J. Am. Chem. Soc., 2002, 124, 5782.
17. W. Niu, S. Zheng, D. Wang, X. Liu, H. Li, S. Han, J. Chen, Z. Tang and G. Xu, J. Am. Chem. Soc., 2009, 131, 697.
18. J. Puthussery, S. Seefeld, N. Berry, M. Gibbs and M. Law, J. Am. Chem. Soc., 2011, 133, 716.
19. S. Hu and X. Wang, J. Am. Chem. Soc., 2010, 132, 9573.
20. L. Zhu, B. J. Richardson and Q. Yu, Chem. Mater., 2015, 27, 3516.
21. X. Peng, J. Wickham and A. P. Alivisatos, J. Am. Chem. Soc., 1998, 120, 5343.
22. M. R. Gao, Y. F. Xu, J. Jiang and S. H. Yu, Chem. Soc. Rev., 2013, 42, 2986.
23. W. Ai, Z. Luo, J. Jiang, J. Zhu, Z. Du, Z. Fan, L. Xie, H. Zhang, W. Huang and T. Yu, Adv. Mater., 2014, 26, 6186.
24. R. Mo, S. O. Tung, Z. Lei, G. Zhao, K. Sun and N. A. Kotov, ACS Nano, 2015, 9, 5009.
25. Y. Tang, Y. Zhang, W. Li, B. Ma and X. Chen, Chem. Soc. Rev., 2015, 44, 5926.
26. Y. Du, Z. Yin, J. Zhu, X. Huang, X. J. Wu, Z. Zeng, Q. Yan and H. Zhang, Nat. Commun., 2012, 3, 1177.
27. Q. Zhang, E. Uchaker, S. L. Candelaria and G. Cao, Chem. Soc. Rev., 2013, 42, 3127.

---

Footnote

† Electronic supplementary information (ESI) available: TGA profile, absorption spectra, and TEM images of the as-synthesized FeS₂ NPs. See DOI: 10.1039/c5ra22610d

---