A simple one pot synthesis of cubic Cu₅FeS₄

Abstract

Reaction of CuCl, FeCl₃ and thiourea in ethylene glycol under refluxing conditions yielded micro flower shaped cubic Cu₅FeS₄ as confirmed from high resolution PXRD, SEM with EDX, UV-visible, Raman spectroscopy and magnetic measurements at room temperature.

---

In addition to holding a special status among copper-age minerals, and serving as copper ore, compositions belonging to Cu–Fe–S family exhibit interesting properties and are used in many areas, ranging from earth science to advanced functional materials. Presence of both Cu and Fe in the same crystal lattice imparts tuneable electrical and magnetic properties such as dilute magnetic semiconducting (DMS) behaviour.¹ Additional control of properties can be achieved by varying the chemical composition/structural arrangements.^2,3 Since both the metal ions present in Cu–Fe–S are earth abundant, they are projected to be low cost alternative for various I–III–VI systems (such as Cu–In–S, Cu–Ag–S, Ag–In–S, Ag–Ga–S). CuInS₂ shows band gap of 1.5 eV in the visible region and therefore has been extensively studied as a photovoltaic material. On the other hand, the optical band gap for CuFeS₂ (0.6 eV) does not lie in the range suitable for photovoltaic applications and this has coerced researchers to explore and identify other compositions of the Cu–Fe–S system, whose band gap is suitable to be used as photovoltaic materials. Plass and co-workers reported the optical band gap of nano particles of Cu₅FeS₄, obtained by the reaction of Cu(acac)₂, Fe(acac)₃, and elemental S in dodecanethiol and oleic acid,⁴ to be 1.25 eV, for the first time. However, bulk solution phase synthesis of Cu₅FeS₄ is essential to further study their properties as well as to examine their photovoltaic functions. To the best of our knowledge, except that of Plass and coworkers, synthesis reports for Cu₅FeS₄ are mostly concerned with indirect synthesis by copper enrichment or by the solid state reaction of binary sulphides at high temperatures for longer durations.^5–7 Existence of large number of chemical compositions, unquenchable (and even moderate) temperature phases, presence of the metal ions, that can show variable oxidation states; and significant difference in chalcophilic character of Cu and Fe, might be the major reasons behind the fewer attempts in the exploration of a synthesis strategy for Cu–Fe–S compositions in general, and Cu₅FeS₄ in particular.^8,9 Specific composition targeted synthetic strategy in desired symmetry, in the Cu–Fe–S system is still a challenging task. Additional difficulty arises from the polymorphic arrangements existing in Cu₅FeS₄ (as high, intermediate, and low bornite). The high-temperature polymorph possesses antifluorite-type structure, consisting of a random distribution of six metal cations and two vacancies in the eight tetrahedral interstices of a cubic close-packed sulphur framework and such an arrangement is stable above 270 °C.^10–13 Based on the single crystal data, Morimoto^10,11 has reported that the cubic symmetry of Cu₅FeS₄ results from the twinning of domains with rhombohedral symmetry in eight different orientations. In this rhombohedral structure, all the metal atoms are positioned in three out of four tetrahedral units in a cubic close packed sulphur sub-lattice as in the case of rhombohedral Cu₉S₅ (Cu_1.8S).^12,13 The intermediate and low forms can be described as superstructures of the high temperature cubic form due to increase in the ordering of vacancies and the metal atoms with reducing temperature.^14–16

Given these complexities, phase selective, direct, reproducible, rapid and one-pot synthesis of Cu₅FeS₄ is justified. As cubic Cu₅FeS₄ and rhombohedral Cu₉S₅ (Cu_1.8S) phases exhibit almost similar kind of atomic and vacancy ordered arrangements, it can be conceptually conceived that a solution based synthetic strategy producing Cu₉S₅ (Cu_1.8S)¹⁷ reliably and reproducibly may be suitable for easy stabilization of Cu₅FeS₄ based on the kinetics and energy considerations. Also, it can be conceived that if ionized iron species can be introduced during the formation of Cu₉S₅ (Cu_1.8S), the possibility of inclusion of iron in it will be quite high. Following our extensive research on copper sulphides and copper containing ternary sulphides,^3,17–22 it was quite clear that Cu₉S₅ can be easily realized from the reaction of copper salts with thiourea in ethylene glycol. Therefore we studied the introduction of iron during its formation leading to Cu₅FeS₄ and the results are reported in this communication.

0.50 g (5 mmol) of freshly prepared CuCl, 0.81 g (5 mmol) FeCl₃ (Thomas Baker, 99%) and 0.76 g (10 mmol) thiourea (SRL, 98%) were refluxed together in ethylene glycol (100 mL) for 1.5 h. After the reaction, product in the form of suspension was separated by centrifugation. Finally, it was washed several times with d.d. water, absolute alcohol, CS₂ and dried at room temperature. The phase composition and structure of the final products were determined from powder X-ray diffraction (PXRD) patterns using a Bruker D8 Discover X-ray diffractometer employing Cu K_α radiation (λ = 1.5418 Å) with a scan rate of 1.2 s per step and step size 0.02° at 298 K. Raman spectra of the samples, in compact form, were collected using a Renishaw spectrophotometer equipped with a microscope and operating with a laser of wavelength 514 nm. Magnetic measurements were carried out at 300 K using a vibrating sample magnetometer (Microsense EV9). Transmission electron microscopy (TEM) and selected area electron diffraction (SAED) were carried out using FEI Technai G² 20 electron microscope operating at 200 kV. The morphology and elemental mapping of the final products were examined from SEM images obtained using an FESEM, FEI Quanta 200F microscope equipped with EDX accessories. UV-visible diffuse reflectance data of the samples were collected over the spectral range of 200–800 nm using a PerkinElmer Lamda 35 scanning double-beam spectrometer equipped with a 50 mm integrating sphere. Photoluminescence (PL) measurements were performed on a Horiba Jobin Yvon Fluorolog 3 Spectrofluorometer at room temperature.

In the PXRD pattern of the product from the reaction of CuCl, FeCl₃ and thiourea in ethylene glycol, sharp reflections in the 2θ range of 10–70° were observed indicating high crystalline nature of the samples (Fig. 1(a)). The presence of less number of reflections suggested that the product possessed higher symmetry. The shift of the observed reflections towards higher 2θ values with respect to Cu₉S₅ (Cu_1.8S) indicated possible inclusion of an additional cation in it. Further, a search of the JCPDS data base for high symmetry compositions consisting of Cu, Fe and S yielded a close match with Cu₅FeS₄ (JCPDS file no. 34-0135) and therefore the observed peaks were indexed in cubic symmetry (space group Fm3m (#225)) with lattice constant of a = 10.98 Å.¹³ Of the many fundamental problems on the super structure of bornite, Fe/Cu ordering plays a significant and decisive role on the final ordering present in the samples. Based on magnetic structure, TEM studies coupled with first principle calculations, two Fe/Cu ordering schemes indicating the filling of Fe in the tetrahedral sites of sulphur atoms in the antifluroite cube, and the vacancies at the copper positions having zinc blende cube of the superstructure have been found to exist in Cu₅FeS₄.¹³ A difference in energy of 52.7 meV per unit cell has been computed between these two orderings.¹³ As Fe and Cu exhibit similar X-ray and electron scattering factors, the unequivocal assignment of the observed phase as Cu₅FeS₄ and not Cu₉S₅ may not be absolutely certain, even though the observed peak at higher 2θ (around 69°) is observed only for Cu₅FeS₄. Therefore, additional evidence has been gathered from Raman spectroscopy measurements. The sample showed one strong peak at 472 cm⁻¹ (shown in Fig. 1(b)) in the Raman spectrum.²³ However, on deconvolution, the presence of two bands at 465 cm⁻¹ and 472 cm⁻¹ was evident (inset of Fig. 1(b)). Since such a deconvolution was not observed in the Raman spectrum of Cu₉S₅ (Fig. S3 ESI†), this was crucial evidence for the formation of Cu₅FeS₄.

Fig. 1 (a) PXRD pattern of the product obtained by reaction of CuCl, FeCl₃ and thiourea in ethylene glycol for 1.5 h and (b) Raman spectrum of Cu₅FeS₄ obtained in the present set of reactions. Inset in (b) shows the deconvolution of the Raman spectrum. JCPDS cards for Cu₅FeS₄ and Cu₉S₅ (Cu_1.8S) are provided for easy comparison.

A low magnification FESEM image of Cu₅FeS₄ is presented in Fig. 2(a) showing uniform flower like structure with an average diameter of 10 μm. The selected area electron diffraction (SAED) pattern recorded from the high resolution TEM showed well defined spots that were indexed corresponding to [800], [444], [440], [331], [311] planes of the cubic cell (Fig. 2). On higher magnification, micro-flower morphology consisting of a hierarchical structure was observed (Fig. 2(b)). The nanosheets intersected with each other resulting in a net like structure with porous surfaces. Interestingly, 3D flower structures assembled by a large number of nanosheets with an average thickness of 14 nm. Similar microflower morphology have earlier been observed for CuS and SnS₂ obtained in ethylene glycol and polyethylene glycol, respectively.^17,24 It is therefore reasonable to state that the ethylene glycol solvent system used in the current study resulted in the microflower morphology. The emission spectrum of the sample at an excitation wavelength of 500 nm showed a low intensity signal centered at 750 nm. Such emission characteristics were not observed or reported earlier for Cu–S compositions. EDX analysis of the sample revealed Cu : Fe : S as 4.9 : 1.4 : 3.8, closely matching with the expected ratio between copper, iron and sulphur (Fig. S2†). However, emissions in this range are usually observed for non-stoichiometric copper containing I–III–VI₂ compounds,^22,25 thereby suggesting the possibility of iron incorporation in the Cu–S system. Paramagnetic behaviour of the sample (inset of Fig. 3(a)), observed at room temperature also offered additional support for the presence of magnetic Fe³⁺ ion in the sample. Copper sulphide (Cu_xS) compositions are known to show weak paramagnetism only for x = 1–1.12.¹⁷ Also, iron sulphide compositions, synthesized by the reaction of FeCl₃ with thiourea were reported to exhibit ferromagnetic behaviour²⁵ and not a simple paramagnetic behaviour. From the UV-visible spectroscopy data (shown in Fig. 3(b)), a band gap value of 1.25 eV was deduced for Cu₅FeS₄, which was quite comparable with the reported value.⁴

Fig. 2 (a) and (b) show the low and high magnification FESEM image of cubic Cu₅FeS₄. Indexed SAED pattern from the TEM experiments is shown in (c).

Fig. 3 (a) and (b) show photoluminescence emission obtained with λ_exc = 500 nm and UV-visible absorbance spectrum spectrum of Cu₅FeS₄, respectively. Plot of magnetic field versus magnetization and plot of photon energy (eV) vs. (αhυ)² are shown in the insets of (a) and (b), respectively.

To answer whether the reaction of other iron salts such as Fe(NO₃)₃·9H₂O or Fe₂(SO₄)₃ with CuCl and thiourea could yield Cu₅FeS₄ or any other ternary composition, experiments were performed under identical conditions using the above salts as iron sources.

PXRD patterns of the products from these reactions showed the formation of binary sulphides of copper and iron (Fig. S4 ESI†). These results hinted the possible mechanism of formation of copper rich Cu₅FeS₄ from the reaction of CuCl, FeCl₃ and thiourea. Independent reactions of CuCl with thiourea and FeCl₃ with thiourea by polyol process have been reported to yield copper rich Cu_1.8S and Fe₃S₄, respectively.²⁶ The occurrence of these two species in higher symmetry along with the highly cation fluidic character of Cu_1.8S possibly might have promoted the formation of copper rich cubic Cu₅FeS₄.²⁷ Also, the existence of Fe²⁺ and Fe³⁺ in Fe₃S₄ increases the probability of cation exchange with Cu¹⁺ and Cu²⁺ present in Cu₉S₅ (Cu_1.8S).

Conclusion

High bornite Cu₅FeS₅ was synthesized in pure form by adopting a target oriented approach, derived from isostructural binary copper sulphide, following a single step and one pot strategy. As the reactions are simple and scalable, they can easily be implemented under industrial conditions for viable applications.

Acknowledgements

The authors thank DST (Nanomission) and DST (SR/S1/PC-07/2011(G)), New Delhi, Govt of India, for the financial support provided for this work. One of the authors MG thanks CSIR, Govt of India for SRF fellowship.

Notes and references

1. K. J. Klabunde, Nanoscale Materials in Chemistry, Wiley Interscience, 2001.
2. L. Barkat, N. Hamdadou, M. Morsli, A. Khelil and J. C. Bernède, J. Cryst. Growth, 2006, 297, 426.
3. P. Kumar, S. Uma and R. Nagarajan, Chem. Commun., 2013, 49, 7316.
4. A. M. Wiltrout, N. J. Freymeyer, T. Machani, P. Rossi and K. E. Plass, J. Mater. Chem., 2011, 21, 19286.
5. W. M. B. Roberts, Econ. Geol., 1963, 58, 52.
6. D. B. Roberts and D. J. Williams, Mater. Res. Bull., 1974, 10, 163.
7. R. Berger and R. V. Bucur, Solid State Ionics, 1996, 89, 269.
8. P. B. Barton Jr, Econ. Geol., 1973, 68, 455.
9. V. Raghavan, J. Phase Equilib. Diffus., 2004, 25, 450.
10. N. Morimoto and G. Kullerud, Am. Mineral., 1961, 46, 1270.
11. Y. Kanazawa, K. Koto and N. Morimoto, Can. Mineral., 1978, 16, 397.
12. N. Morimoto, Acta Crystallogr., 1964, 17, 351.
13. Y. Ding, D. R. Weblen and C. T. Prewitt, Am. Mineral., 2005, 90, 1265.
14. K. Koto and N. Morimoto, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1975, 31, 2268.
15. L. Pierce and P. R. Buseck, Am. Mineral., 1978, 63, 1.
16. A. Putnis and J. Grace, Contrib. Mineral. Petrol., 1976, 55, 311.
17. P. Kumar, M. Gusain and R. Nagarajan, Inorg. Chem., 2011, 50, 3065.
18. P. Kumar and R. Nagarajan, Inorg. Chem., 2012, 50, 9204.
19. M. Gusain, P. Rawat and R. Nagarajan, Mat.Lett., 2014, 133, 220.
20. P. Kumar, R. Nagarajan and R. Sarangi, J. Mater. Chem. C, 2013, 1, 2448.
21. P. Kumar, M. Gusain and R. Nagarajan, Inorg. Chem., 2012, 51, 7945.
22. M. Gusain, P. Kumar and R. Nagarajan, RSC Adv., 2013, 3, 18863.
23. (a) B. Minceva-Sukarova, M. Najdoski, I. Grozdanov and C. J. Chunnilall, J. Mol. Struct., 1997, 410–411, 267; (b) T. P. Mernagh and A. G. Trudu, Chem. Geol., 1993, 103, 113.
24. H. Ke, W. Luo, G. Cheng, X. Tian and Z. Pi, Micro Nano Lett., 2009, 4, 177.
25. (a) M. Uehara, K. Watanabe, Y. Tajiri, H. Nakamura and H. Maeda, J. Chem. Phys., 2008, 129, 134709; (b) C. I. Pearce, R. A. D. Pattrick and D. J. Vaughan, Rev. Mineral. Geochem., 2006, 51, 127.
26. (a) D.-G. Lee, N.-H. Lee, H.-J. Oh, S.-C. Jung, J.-S. Hwang, W.-J. Lee and S.-J. Kim, J. Nanosci. Nanotechnol., 2011, 11, 1434; (b) G. Shen, D. Chen, K. Tang, X. Liu, L. Huang and Y. Qian, J. Solid State Chem., 2003, 173, 232.
27. X. Lu, Z. Zhuang, Q. Peng and Y. Li, CrystEngComm, 2011, 13, 4039.

---

Footnote

† Electronic supplementary information (ESI) available: PXRD patterns of the reaction products using Fe(III) sulfate and Fe(III) nitrate as the iron source with CuCl and thiouera, EDX spectrum of Cu₅FeS₄, Raman spectrum of Cu₉S₅ (Cu_1.8S), comparison of PXRD pattern obtained in the present study with the standard JCPDS files of iron sulphide. See DOI: 10.1039/c4ra06939k

---