Effects of flux treatment on morphology of single-crystalline BaNbO₂N particles

Abstract

The effect of different kinds of alkali halide flux treatments on the morphology of BaNbO₂N particles was investigated. BaNbO₂N powder was first synthesized by heating a mixture of BaCO₃ and Nb₂O₅ under NH₃ flow. Following synthesis, the powder was again heated under NH₃ flow but in the presence of a flux. In the case of a NaCl flux, highly crystalline cubic particles were obtained. On the other hand, a KCl flux led to the formation of truncated octahedral particles. Although particles with clear crystal habits were obtained, it was found that they were actually an aggregation of primary particles, all with the same crystal orientation. A systematic survey of the effect of the different fluxes revealed that the type of cation in the flux played a major role in determining the morphology of the BaNbO₂N particles. It was also revealed that a vaporized flux was also effective in changing the particle morphology.

---

1. Introduction

Oxynitrides have recently attracted much attention for their various potential applications, such as pigments, dielectrics, and photocatalysts.^1–6 They have been intensively studied as photocatalysts because of their appropriate bandgap energy for absorbing visible light and their suitable band alignment for the water splitting reaction. Among these compounds, there has been a particular interest in perovskite-type oxynitrides, generally denoted as AB(O,N)₃.^7,8 BaNbO₂N has a perovskite-type structure and an optical absorption edge at about 740 nm, which is the longest wavelength reported for a transition metal oxynitride.^9,10 In order to achieve high performance as photocatalysts or photoelectrodes, the crystallinity of oxynitrides is one of the most important factors. However, unlike oxide materials, for which several approaches such as hydrothermal synthesis and flux methods can be used to prepare highly crystalline particles,^11–14 for oxynitrides the options are rather limited.

High-quality oxynitride crystals can be prepared using flux synthesis. The flux method can be applied during synthesis of oxide precursors, during nitridation of starting materials or oxide precursors, or during post treatment of oxynitrides. In the first method, a flux is mixed with starting materials such as metal oxides, carbonates and nitrates, which are then heated in air or an inert atmosphere to obtain oxide precursors. Nitridation is then performed to produce the final oxynitrides. Highly crystalline SrTaO₂N and LaTiO₂N have been synthesized using this procedure from Sr₂Ta₂O₇ and La₂Ti₂O₇, respectively.^15,16 In this situation, the oxide precursors have a similar crystal structure to that of the final oxynitrides. However, one of the difficulties in synthesizing well-crystallized BaNbO₂N is that there is no suitable precursor oxide material with a Ba/Nb ratio of 1. This leads to a large change in the crystal structure during the nitridation process. Another means of applying the flux method to prepare oxynitrides is to mix the flux with the starting materials or oxide precursors, which are then heated under an NH₃ flow to promote crystal growth. Takata et al. applied this approach to synthesizing Ta₃N₅,¹⁷ and Kim reported the effects of a KCl flux on CaTaO₂N, SrTaO₂N and LaTaON₂.¹⁸ These (oxy)nitrides contain Ta atoms. However, if this method is used to synthesize BaNbO₂N, it is likely that chemical reduction of Nb species and/or contamination by flux components would occur due to the long reaction time required. The third approach is to apply a flux treatment to an already synthesized oxynitride. The flux is mixed with the oxynitride, which is then heated under an NH₃ flow. This method may be expected to lead to a fairly rapid improvement in crystallinity, because no drastic change in composition or crystal structure occurs. However, there has so far been no report focusing on the application of such a flux treatment to Nb-containing oxynitrides. Hence, in the present study, the effect of different post-synthesis flux treatments on the crystallinity and morphology of BaNbO₂N particles was investigated.

2. Experimental

BaNbO₂N particles were prepared by heating a mixture of BaCO₃ and Nb₂O₅ under an NH₃ flow. The Ba/Nb ratio in the starting material was adjusted to 1.5 in order to suppress the formation of the NbO_xN_y impurity phase due to volatilization of Ba during the nitridation process.¹⁹ Nitridation was performed at 1173 K for 10 h under an NH₃ flow of 150 mL min⁻¹. After the synthesis of BaNbO₂N, different flux treatments were carried out. The BaNbO₂N powder was mixed with either LiCl, NaCl, KCl, RbCl, CsCl, KF, KBr, or KI, and heated at 1173 K for 3 h under an NH₃ flow of 100 mL min⁻¹. The flux/BaNbO₂N molar ratio was fixed at 5 unless otherwise mentioned. After the flux treatment, most of the flux component had evaporated, and the obtained samples were washed with a sufficient amount of distilled water to remove the remained flux. Hereafter, the final flux-treated samples are denoted as BaNbO₂N(flux type), e.g., BaNbO₂N(NaCl). The crystal structure of the samples was identified using X-ray diffraction (XRD; Rigaku, RINT Ultima III, CuKα source). Ultraviolet-visible (UV-vis) diffuse reflectance spectra were acquired using a JASCO V-670 spectrophotometer. The particle morphology was examined by field-emission scanning electron microscopy (FE-SEM; Hitachi, S-4700). Transmission electron microscopy (TEM) images and selected area electron diffraction (SAED) patterns were acquired using a field-emission transmission electron microscope (JEOL, JEM-2800). Elemental analysis was performed using inductively coupled plasma atomic emission spectroscopy (ICP-AES; Shimadzu, ICPS-8100), X-ray fluorescence (XRF; Shimadzu, EDX-800HS), and oxygen/nitrogen combustion analysis (Horiba, EMGA-620W/C).

3. Results and discussion

Fig. 1 shows XRD patterns for BaNbO₂N particles before and after flux treatment using the different alkali chlorides. Before the flux treatment, almost all diffraction peaks were assigned to BaNbO₂N (PDF 01-084-1749), with a trace amount of BaNbO_x being present. Following the flux treatment, although the dominant diffraction peaks were still associated with BaNbO₂N, some other phases, such as NbO_xN_y, were also observed (see Fig. S1 in ESI†). Except for the case of BaNbO₂N(LiCl), the BaNbO₂N peak intensities increased following the flux treatment, and the full width at half maximum (FWHM) of the (110) peak decreased. This indicates that the crystallinity of the BaNbO₂N particles was improved. As shown in Fig. S2,† the (220) peak position shifted to a slightly higher angle in the case of BaNbO₂N(NaCl), and to a slightly lower angle for BaNbO₂N(CsCl). Since the ionic radius of Ba²⁺, Na⁺, K⁺, Rb⁺, and Cs⁺ is 0.16, 0.14, 0.16, 0.17, and 0.18 nm (coordination number: 12), respectively, these flux cations substituted for Ba ions in the oxynitride to a certain extent. Indeed, the results of the elemental analysis shown in Table S1† indicate that although the flux-treated BaNbO₂N was almost stoichiometric, a few mol% of alkali metal species was present. As shown in Fig. S1,† in the case of BaNbO₂N(NaCl), BaNbO₂N(KCl), BaNbO₂N(RbCl), and BaNbO₂N(CsCl), the intensity of the XRD peaks assigned to the NbO_xN_y phase became higher following flux treatment. This suggests that a partial chemical reduction of Nb compounds had occurred.

Fig. 1 XRD patterns for BaNbO₂N particles (a) before and after flux treatment with (b) LiCl, (c) NaCl, (d) KCl, (e) RbCl, and (f) CsCl.

Fig. 2 shows diffuse reflectance spectra of BaNbO₂N particles before and after flux treatment. Apart from BaNbO₂N(LiCl), all of the flux-treated samples exhibited a clear absorption edge at about 740 nm, similar to that for untreated BaNbO₂N, and increased absorption was seen at longer wavelengths. The enhanced absorption at wavelengths longer than the absorption edge suggests the existence of chemically reduced Nb species or anion vacancies. However, since the elemental analysis results indicated that the flux-treated BaNbO₂N was almost stoichiometric, the former explanation is more likely.

Fig. 2 Diffuse reflectance spectra of BaNbO₂N particles (a) before and after flux treatment with (b) LiCl, (c) NaCl, (d) KCl, (e) RbCl, and (f) CsCl.

Fig. 3 shows SEM images of BaNbO₂N particles before and after flux treatment. Before the flux treatment, the morphology was indeterminate, with aggregations of small particles with sizes of a few tens to a few hundreds of nanometers. After the flux treatment, a marked change in morphology occurred. For both BaNbO₂N(NaCl) and BaNbO₂N(CsCl), a cubic morphology was clearly observed. On the other hand, for BaNbO₂N(KCl) and BaNbO₂N(RbCl), most of the particles had a truncated octahedral shape. Similar flux-induced morphology changes were also reported for SrTiO₃.²⁰

Fig. 3 SEM images of BaNbO₂N particles (a) before and after flux treatment with (b) LiCl, (c) NaCl, (d) KCl, (e) RbCl, and (f) CsCl. Scale bars represent 1 μm.

To examine the particle microstructure in more detail, TEM observations of BaNbO₂N(NaCl) were conducted. Fig. 4 shows a TEM image of a typical cubic particle. It is seen to contain clear grain boundaries, indicating that it is actually a secondary particle made up of primary particles with sizes of several hundred nanometers. Fig. 5(A) shows a higher magnified TEM image of the secondary particle, and Fig. 5(B) shows SAED patterns obtained for the regions indicated in Fig. 5(A) and for the entire region. All of the SAED patterns are clearly spot-like, indicating that the individual primary particles are single crystals. In addition, the similarity of the SAED pattern obtained for the entire region in Fig. 5(A) shows that all of the primary particles have the same orientation.

Fig. 4 TEM image of BaNbO₂N particle following NaCl flux treatment.

Fig. 5 (A) TEM image of BaNbO₂N particle following NaCl flux treatment, and (B) SAED patterns for points indicated in (A); (d) shows SAED pattern for entire region in (A).

Next, treatments using potassium-based fluxes with different anions (KF, KCl, KBr, and KI) were performed. Fig. 6 and 7 show XRD patterns and diffuse reflectance spectra, respectively, for BaNbO₂N(KF), BaNbO₂N(KCl), BaNbO₂N(KBr), and BaNbO₂N(KI). For BaNbO₂N (KF), no diffraction pattern or absorption edge associated with BaNbO₂N was observed, suggesting that the BaNbO₂N crystal structure was not maintained after the KF flux treatment. In the case of BaNbO₂N(KCl), BaNbO₂N(KBr), and BaNbO₂N(KI), stronger BaNbO₂N diffraction peaks were observed after the flux treatment, although peaks due to NbO_xN_y were also detected. In addition, the absorption edge at about 740 nm associated with BaNbO₂N was observed, but with higher absorption at longer wavelengths than for the untreated samples. As shown in Fig. 8, SEM observations revealed that all of the flux-treated samples contained particles with a similar truncated octahedral shape, except for BaNbO₂N(KF). These results indicate that the kind of anion in the flux did not affect the morphology of the BaNbO₂N particles.

Fig. 6 XRD patterns for BaNbO₂N particles (a) before and after flux treatment with (b) KF, (c) KCl, (d) KBr, and (e) KI.

Fig. 7 Diffuse reflectance spectra of BaNbO₂N particles (a) before and after flux treatment with (b) KF, (c) KCl, (d) KBr, and (e) KI.

Fig. 8 SEM images of BaNbO₂N particles (a) before and after flux treatment with (b) KF, (c) KCl, (d) KBr, and (e) KI. Scale bars represent 1 μm.

As described in the Experimental section, most of the flux evaporated during the treatment process, which was carried out a temperature that was more than 100 K higher than the flux melting point. To clarify the effects of the amount of flux on the particle morphology, flux treatment using a volatile flux flow was performed using the setup shown schematically in Fig. S3.† It consisted of two alumina boats that were placed in an alumina tube. The upstream and downstream boats contained an NaCl flux and BaNbO₂N particles, respectively. The molar ratio of NaCl to BaNbO₂N was set to 20. During the flux treatment, the boats were heated to the same temperature. After 3 h at 1173 K under an NH₃ flow of 100 mL min⁻¹, a plentiful supply of the NaCl flux still remained in the upstream boat. Fig. 9 shows an SEM image of BaNbO₂N particles after flux treatment with NaCl using this setup. It can be seen that cubic particles similar to those in Fig. 2(c) were observed. Thus, a continuous flow of vaporized flux was also effective for changing the particle morphology. A similar effect of vaporized flux has been reported for BaTiO₃.²¹ The two-boat method used in the present study can also be applied to other synthesis systems such as a two-zone furnace, and will allow more precise control of the flux, and lead to further development of the flux method.

Fig. 9 SEM image of BaNbO₂N particles after NaCl flux treatment using two-boat method.

4. Conclusions

In the present study, the effect of flux treatment using different alkali halides on BaNbO₂N particles was investigated. Using NaCl or CsCl as the flux, highly crystalline cubic particles were obtained, while using KCl, KBr, KI, or RbCl, truncated octahedral shapes were observed. These results indicate that the kind of cation in the flux strongly affects the morphology of BaNbO₂N. The flux-treated BaNbO₂N was almost stoichiometric, particularly with regard to the O/N ratio, although a few mol% of alkali metal species was present. Moreover, the results obtained using a two-boat setup showed that a vaporized flux was also effective for changing the particle morphology.

Acknowledgements

This work was supported in part by Grants-in-Aid for Specially Promoted Research (no. 23000009) and “Nanotechnology Platform” of the Ministry of Education, Culture Sports Science and Technology (MEXT), Japan. The authors would like to acknowledge Dr. Taro Yamada in the University of Tokyo for performing the ICP measurement and O/N analysis.

References

1. M. Jansen and H. P. Letschert, Nature, 2000, 404, 980–982.
2. Y.-L. Kim, P. M. Woodward, K. Z. Baba-Kishi and C. W. Tai, Chem. Mater., 2004, 16, 1267–1276.
3. A. Gomathi, S. Reshma and C. N. R. Rao, J. Solid State Chem., 2009, 182, 72–76.
4. R. Marchand, Y. Laurent, J. Guyader, P. L'Haridon and P. Verdier, J. Eur. Ceram. Soc., 1991, 8, 197–213.
5. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
6. Y. Moriya, T. Takata and K. Domen, Coord. Chem. Rev., 2013, 257, 1957–1969.
7. K. Ueda, T. Minegishi, J. Clune, M. Nakabayashi, T. Hisatomi, H. Nishiyama, M. Katayama, N. Shibata, J. Kubota, T. Yamada and K. Domen, J. Am. Chem. Soc., 2015, 137, 2227–2230.
8. H. Urabe, T. Hisatomi, T. Minegishi, J. Kubota and K. Domen, Faraday Discuss., 2014, 176, 213–223.
9. B. Siritanaratkul, K. Maeda, T. Hisatomi and K. Domen, ChemSusChem, 2011, 4, 74–78.
10. T. Hisatomi, C. Katayama, Y. Moriya, T. Minegishi, M. Katayama, H. Nishiyama, T. Yamada and K. Domen, Energy Environ. Sci., 2013, 6, 3595–3599.
11. J. W. Liu, G. Chen, Z. H. Li and Z. G. Zhang, Int. J. Hydrogen Energy, 2007, 32, 2269–2272.
12. S. Feng and R. Xu, Acc. Chem. Res., 2001, 34, 239–247.
13. J. Yu and A. Kudo, Adv. Funct. Mater., 2006, 16, 2163–2169.
14. J. Boltersdorf, N. King and P. A. Maggard, CrystEngComm, 2015, 17, 2225–2241.
15. Y. Mizuno, H. Wagata, K. Yubuta, N. Zettsu, S. Oishi and K. Teshima, CrystEngComm, 2013, 15, 8133–8138.
16. M. Hojamberdiev, A. Yamaguchi, K. Yubuta, S. Oishi and K. Teshima, Inorg. Chem., 2015, 54, 3237–3244.
17. T. Takata, D. Lu and K. Domen, Cryst. Growth Des., 2010, 11, 33–38.
18. Y.-I. Kim, Ceram. Interfaces, 2014, 40, 5275–5281.
19. T. Hisatomi, C. Katayama, K. Teramura, T. Takata, Y. Moriya, T. Minegishi, M. Katayama, H. Nishiyama, T. Yamada and K. Domen, ChemSusChem, 2014, 7, 2016–2021.
20. H. Kato, M. Kobayashi, M. Hara and M. Kakihana, Catal. Sci. Technol., 2013, 3, 1733–1738.
21. P. M. Rørvik, T. Lyngdal, R. Sæterli, A. T. van Helvoort, R. Holmestad, T. Grande and M.-A. Einarsrud, Inorg. Chem., 2008, 47, 3173–3181.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ce02335a

---