Laser-Induced Melting and Crystal Growth of Sodium Ion Conductive β-NaFeO₂

Abstract

Laser irradiation of β-NaFeO₂ shows local melting and crystal growth from the melt pool and sodium ion conductivity. Local melting, densification, and crystal growth were confirmed on the β-NaFeO₂ surface through laser irradiation with a wavelength of 1 µm. Scanning electron microscopy and electron backscatter diffraction analysis confirmed anisotropic grain growth and grain boundary shrinkage in the laser-irradiated area. X-ray diffraction results suggested that the laser-irradiated area was in a low crystalline state. In addition, the precipitation of Fe₃O₄ in laser-irradiated β-NaFeO₂ indicates a reduction reaction, suggesting that the laser-irradiated area is heated to 1500°C or higher. We confirmed that the structure change region could be controlled down to a thickness of about 10 µm by laser irradiation with varying laser power to β-NaFeO₂.

Fig. 3 (a) (b) β-NaFeO₂ and laser-irradiated β-NaFeO₂ (P = 30 W, S = 500 mm s⁻¹) of surface and (c) (d) cross-sectional SEM images.

1. Introduction

In recent years, lithium-ion batteries (LIBs) have been in increasing demand as energy storage devices suitable for portable electronic devices and electric vehicles due to their high energy efficiency, long cycle life, and high power density.^1,2) However, the material prices of rare metals such as lithium and cobalt, primary raw materials for LIBs, have been continuously rising, causing serious concerns about resource shortages. In addition, flammable organic electrolytes are used in LIBs as electrolytes, and there are also concerns over safety, such as the risk of ignition. Sodium-ion batteries (SIBs) and all-solid-state batteries (ASSBs) are proposed for alternative LIBs. SIBs use sodium ions as conduction ions instead of lithium ions. ASSBs are nonflammable batteries in which an inorganic solid electrolyte replaces the flammable electrolyte.^3–13) We have proposed an oxide-based Na-ion ASSB to simultaneously solve cost and safety issues. In previous studies, sodium iron phosphate-based materials, which are composed of inexpensive raw materials, have attracted much attention as active cathode materials for SIBs; Homma et al. reported that Na₂FeP₂O₇ glass-ceramics synthesized by glass-ceramic method showed good battery properties as active cathode materials.^14–17)

Oxide ASSBs exhibit high chemical stability in the air. However, because oxides generally have a high elastic modulus, forming by pressurization at room temperature is impossible, and sintering at high temperatures is inevitable. In the conventional fabrication of oxide-based ASSBs, bonding methods by pressurization and high-temperature sintering have been commonly used. Yamauchi et al. took advantage of the fact that Na₂FeP₂O₇ glass exhibits softening flow during crystallization and succeeded in preparing oxide-based Na-ion ASSBs without applying mechanical pressure by simultaneously firing Na₂FeP₂O₇ glass and β-alumina.^11,12) They also reported that the fabricated all-solid-state batteries have reduced internal resistance to 120 Ω and can be operated under temperature conditions of −20°C. We proposed a new ASSB fabrication process in which Na₂FeP₂O₇ slurry deposited on a NASICON-type Na₃Zr₂Si₂PO₁₂ (NZSP) solid electrolyte substrate is irradiated with a fiber-coupled continuous wave (CW) Yb:YVO₄ laser beam (λ = 1080 nm). The active material and the solid electrolyte were integrated to form a dense bonding interface by laser irradiation.¹⁸⁾ Fe²⁺ ions have absorption at 1 µm wavelength and are converted to thermal energy by non-radiative relaxation, causing heating around Fe²⁺.¹⁹⁾ We found that laser-irradiated Na₂FeP₂O₇ undergoes glassy state by rapid cooling that does not cause recrystallization.

Nakata et al. reported that sodium iron phosphate glass can be used as an active cathode material for SIBs and that 30Na₂O–40FeO–30P₂O₅ glass exhibits a high reversible discharge capacity of 115 mAh/g.²⁰⁾ Phosphate crystals containing Na⁺ ions and Fe²⁺ include Na₂FeP₂O₇, Na₄Fe₃(PO₄)₂P₂O₇, and NaFePO₄.^21–24) Of these, maricite-type NaFePO₄ (= 25Na₂O–50FeO–25P₂O₅), which has the highest Fe²⁺ content, has a high theoretical capacity (155 mAh g⁻¹) but is inactive as an active material because the isolated PO₄³⁻ blocks conduction of sodium ions. In addition, the composition of NaFePO₄ makes it difficult to vitrify by the conventional melt-quenching method.^25–27) However, it has recently been shown that maricite-type NaFePO₄ exhibits good battery performance by nanosizing and carbon coating.^28–32) We recently reported for the first time that amorphization of NaFePO₄ by laser irradiation improves battery performance.³³⁾ In previous studies, we have reported the local melting of sodium iron phosphate-based glass-ceramics by laser irradiation and their vitrification by rapid cooling.

In this study, we focused on β-NaFeO₂, one of the transition metal oxides NaMO₂ (M: transition metal), which has been actively studied as an active cathode material for SIBs. β-NaFeO₂ crystals contain Na and redox-paired iron atoms at 1:1, and more than two oxygen atoms, resulting in low mass and high. Therefore, it is expected to have a high energy density. Recently, nano-sized crystals have been reported to exhibit good battery performance³⁴⁾ Oxide ASSBs using transition metal oxides as cathode materials have not been studied.

In this study, we report the characterization of the local melting, solidification, and crystallization of β-NaFeO₂ by laser irradiation. Fe³⁺ ions in the β-NaFeO₂ crystal act as laser absorbers and Fe³⁺ ions are known to have strong absorption bands at 530–800 nm due to the d-d transition of ⁶A₁ → ⁴T₁, ⁴T₂.¹⁹⁾ The absorption spectrum of Fe³⁺ has a broad absorption band, and the band tail also exists at 1000 nm. β-NaFeO₂ crystals are expected to heat, melt and quench in the laser irradiation region due to the high content of Fe³⁺ ions.

2. Experimental Procedure

2.1 Synthesis of β-NaFeO₂ crystals

β-NaFeO₂ was synthesized by the solid-state reaction method. Na₂CO₃ (Nakarai tesque Co.) and α-Fe₂O₃ (Kojyundo chemicals Co.) were mixed as raw materials and formed into pellets (0.3 g batch) by a circular mold (12 mm diameter) and a hand press. Heat treatments were performed at 900°C for 6 h in an electric furnace.

2.2 Laser irradiation and characterization

We used a fiber laser with λ = 1064 nm infrared light emission for the laser irradiation experiment. In addition, the laser consists of an f-θ lens. The lens’s effective focal length is 290 mm, and the minimum line width spot diameter is approximately 60 µm. β-NaFeO₂ sintered body was laser irradiated. The laser power (P) and scanning speed (S) were P = 15, 18, 21, 24, 27, and 30 W, respectively, and S = 500 mm s⁻¹. Structural changes in the laser-irradiated region were evaluated by X-ray diffraction (XRD) (Rigaku ULTIMA IV), electron beam backscatter diffraction (EBSD) analysis (Oxford), scanning electron microscope (SEM) (VE-8800, Keyence) and confocal scanning laser microscope (Shimadzu OLS3000).

3. Results and Discussions

Figure 1 shows a photograph of β-NaFeO₂ and laser-irradiated β-NaFeO₂ (P = 30 W, S = 500 mm s⁻¹). The confocal laser scanning micrograph of the laser-irradiated sample surface is shown in Fig. 2. Structural changes accompany the laser-irradiated area. The laser-irradiated β-NaFeO₂ seems more irregular texture than the initial surface of pristine sample.

Fig. 1

Photographs of β-NaFeO₂ and laser-irradiated β-NaFeO₂.

Fig. 2

Confocal laser scanning micrographs of β-NaFeO₂ and laser-irraiated β-NaFeO₂.

Figure 3 shows SEM images of pristine and laser-irradiated samples. Figure 3(a) shows the microstructure of the pristine NaFeO₂ surface. Figure 3(b) shows the surface microstructure observed from the laser irradiation direction. Figure 3(c) and 3(d) show the microstructures of pristine and laser-irradiated samples cut and observed from the direction perpendicular to the laser irradiation direction. Figure 3(a) and 3(c) show that isolated grains a few micrometers in size are clearly observed in the pristine β-NaFeO₂. There are voids and grain boundaries between the particles, indicating that grain growth due to the solid-state reaction is in progress and that the ceramics densification that fills the voids is not in progress. On the other hand, no distinct particles or grain boundaries can be seen on the surface of the laser-irradiated area shown in Fig. 3(b), indicating that the β-NaFeO₂ has completely melted. In Fig. 3(d), a stripe pattern perpendicular to the surface is observed, suggesting that the crystals may have grown anisotropically during the process of rapid cooling of the melt.

Fig. 3

(a) (b) β-NaFeO₂ and laser-irradiated β-NaFeO₂ (P = 30 W, S = 500 mm s⁻¹) of surface and (c) (d) cross-sectional SEM images.

Figure 4 shows the EBSD results of (a) reflection electron image, (b) band contrast, (c) crystallographic phase, and (d) crystal orientation diagram of the fractured surface of β-NaFeO₂ ion milling after laser irradiation. The crystal orientation diagram shows high crystallization and orientation development in the area circled by the square. The laser-irradiated area is reported to be hottest in the center and cooler as one moves away from the center.³⁵⁾ The melting of β-NaFeO₂ by laser irradiation may have caused anisotropic crystal growth with grain growth toward the high-temperature side of the sample surface as it cooled in the direction away from the laser focal point.

Fig. 4

(a) backscattered electron image (b) band contrast (c) crystal phase and (d) crystal orientation map of the fracture surface of β-NaFeO₂ ion milled after laser irradiation (P = 30 W, S = 500 mm s⁻¹).

Figure 5 shows fracture surface SEM images of laser irradiated β-NaFeO₂ at laser power P = 15, 18, 21, 24, 27, and 30 W and scanning speed S = 500 mm s⁻¹. Figure 6 shows the dependence of the thickness of the laser-irradiated structural change region on the laser power. Since the center of the laser irradiation has the most significant thickness of the structure change region, this center was treated as the thickness. The thickness of the structure change region increases almost linearly with increasing laser power. We obtained similar results from our previous laser irradiation of maricite type NaFePO₄ using a fiber-coupled CW Yb:YVO₄ laser (λ = 1080 nm). In this study, we confirmed that the structure change region could be controlled down to a thickness of about 10 µm.

Fig. 5

SEM images of fracture surface of laser-irradiated β-NaFeO₂ (P = 15, 18, 21, 24, 27, 30 W, S = 500 mm s⁻¹).

Fig. 6

Dependence of the laser reach distance on the laser power.

Figure 7 shows the XRD pattern of the surface of the synthesized β-NaFeO₂. Many sharp diffractions were obtained, and we confirmed that they originated from β-NaFeO₂ crystals (COD: 1009034). Figure 7 shows the surface XRD patterns of laser-irradiated β-NaFeO₂. Many sharp diffractions were obtained for laser-irradiated β-NaFeO₂. This diffraction pattern was the same as that of the synthesized β-NaFeO₂, confirming the presence of β-NaFeO₂ crystals in all laser-irradiated β-NaFeO₂. In addition, the diffraction intensity of laser-irradiated β-NaFeO₂ is much lower than that of pristine β-NaFeO₂. This suggests a decrease in crystallinity or amorphization of β-NaFeO₂, and diffraction originating from Fe₃O₄ (COD: 1526955) was confirmed in the XRD pattern of laser irradiated β-NaFeO₂. Pristine β-NaFeO₂ is composed of Fe³⁺, while Fe₃O₄ is composed of Fe²⁺ and Fe³⁺. The results mean that a reduction reaction causes the precipitation of Fe₃O₄ due to heating by laser irradiation. Such laser-induced reduction of metal oxides has been actively studied.^36–39) Kihara et al. exposed α-Fe₂O₃ in distilled water to an unfocused beam (6 mm diameter) of the second harmonic (532 nm) of a nanosecond Nd³⁺:YAG laser (Continuum, Surelite I, pulse width: 6 ns FWHM, repetition rate: 10 Hz).⁴⁰⁾ The redox reaction of Fe₂O₃; 4 Fe₃O₄ + O₂ = 6 Fe₂O₃, the Fe₂O₃ phase is thermally stable at room temperature. However, according to the Ellingham diagram, above 1500°C, the Gibbs free energy of the reaction becomes positive, so the reaction equilibrium tends toward the reduction side, namely Fe₃O₄.^40,41) The Rietveld analysis was performed for laser irradiated β-NaFeO₂ by the Rietveld refinement program PROFEX⁴²⁾ to quantify the volume fraction of crystals formed. In this study, the area integral of each component was used as an indicator of volume fraction. β-NaFeO₂ and Fe₃O₄ XRD signal area fractions are shown in Fig. 8. In the pristine sample, the area intensity of β-NaFeO₂ (I_Area) is I_Area = 100%, indicating that no unreacted sample remains. The ratio of magnetite Fe₃O₄ increases with increasing laser power. It is inferred that at all laser powers, the laser irradiated area is heated above 1500°C.

Fig. 7

XRD patterns of pristine β-NaFeO₂ ceramic and laser-induced part (P = 15, 18, 21, 24, 27, 30 W, S = 500 mm s⁻¹).

Fig. 8

Area fractions of XRD signals of β-NaFeO₂, Fe₃O₄, and Na₂CO₃ in pristine β-NaFeO₂ and laser-irradiated β-NaFeO₂.

The lattice parameter of laser-irradiated β-NaFeO₂ was determined by Rietveld analysis. The dependence of the lattice constant on laser power is shown in Fig. 9. The lattice constants in the a- and c-axis directions decrease with increasing laser power, while those in the b-axis direction increase. It should be emphasized that, as mentioned above, the laser-irradiated area is instantly at a higher temperature than 1500°C. Under such high-temperature conditions, the possibility of Na volatilization from β-NaFeO₂ must also be considered. The change in lattice parameter with increasing laser irradiation is most likely due to the volatilization of Na. Figure 10 shows a schematic illustration of the crystal structure of β-NaFeO₂ drawn by the program VESTA.⁴³⁾ β-NaFeO₂ is orthorhombic, and its framework structure is FeO₄ tetrahedron. It has a tunnel structure toward the a-axis direction and is a diffusion path for Na⁺ ions. Observation of the crystal structure from the b-axis direction shows that the a- and c-axis directions are densely packed with vertex-sharing FeO₄ units. On the other hand, observation from the a-axis and c-axis directions shows that the structure in the b-axis direction is such that there are many spaces between FeO₄ units. These structural features may be due to an increase in the lattice constant in the b-axis direction. However, further study must clarify the cause of the change in lattice constant.

Fig. 9

Dependence of lattice parameter on laser power.

Fig. 10

A schematic illustration of the crystal structure for β-NaFeO₂ drawn using the program VESTA.⁴³⁾

4. Conclusion

Densification of β-NaFeO₂ by laser irradiation through melting and quenching caused a reduction of grain boundaries and anisotropic growth of crystals. In addition, the laser-irradiated β-NaFeO₂ has low crystallinity and the precipitation of Fe₃O₄ indicates that a reduction reaction has occurred. Melting and solidifying active cathode material without a glass-former by laser irradiation is expected to provide bonding with solid electrolytes and reduce interface resistance, which is difficult to achieve by co-firing. We are also interested in elucidating the mechanism of recrystallization of NaFeO₂ by laser irradiation.

Acknowledgments

This study has received funding from Nagaoka University of Technology (NUT) presidential research grant, the research collaboration project to develop high performance secondary battery materials between NUT and Nippon Electric Glass Co. Ltd., and JSPS KAKENHI Grant numbers 22H00260, 22H05279 and 22J10210.

REFERENCES

• 1)   Y.  Ding,  Z.P.  Cano,  A.  Yu,  J.  Lu and  Z.  Chen: Electrochem. Energy Rev. 2 (2019) 1–28. doi:10.1007/s41918-018-0022-z
• 2)   A.  El Kharbachi,  O.  Zavorotynska,  M.  Latroche,  F.  Cuevas,  V.  Yartys and  M.  Fichtner: J. Alloy. Compd. 817 (2020) 153261. doi:10.1016/j.jallcom.2019.153261
• 3)   K.  Kubota and  S.  Komaba: J. Electrochem. Soc. 162 (2015) A2538–A2550. doi:10.1149/2.0151514jes
• 4)   N.  Yabuuchi,  K.  Kubota,  M.  Dahbi and  S.  Komaba: Chem. Rev. 114 (2014) 11636–11682. doi:10.1021/cr500192f
• 5)   M.D.  Slater,  D.  Kim,  E.  Lee and  C.S.  Johnson: Adv. Funct. Mater. 23 (2013) 947–958. doi:10.1002/adfm.201200691
• 6)   S.W.  Kim,  D.H.  Seo,  X.  Ma,  G.  Ceder and  K.  Kang: Adv. Energy Mater. 2 (2012) 710–721. doi:10.1002/aenm.201200026
• 7)   J.Y.  Hwang,  S.T.  Myung and  Y.K.  Sun: Chem. Soc. Rev. 46 (2017) 3529–3614. doi:10.1039/C6CS00776G
• 8)   P.  Albertus et al.: ACS Energy Lett. 6 (2021) 1399–1404. doi:10.1021/acsenergylett.1c00445
• 9)   Y.  Kato,  S.  Hori,  T.  Saito,  K.  Suzuki,  M.  Hirayama,  A.  Mitsui,  M.  Yonemura,  H.  Iba and  R.  Kanno: Nat. Energy 1 (2016) 16030. doi:10.1038/nenergy.2016.30
• 10)   S.  Shiraki,  T.  Shirasawa,  T.  Suzuki,  H.  Kawasoko,  R.  Shimizu and  T.  Hitosugi: ACS Appl. Mater. Interfaces 10 (2018) 41732–41737. doi:10.1021/acsami.8b08926
• 11)   H.  Yamauchi,  J.  Ikejiri,  F.  Sato,  H.  Oshita,  T.  Honma and  T.  Komatsu: J. Am. Ceram. Soc. 102 (2019) 6658–6667. doi:10.1111/jace.16607
• 12)   H.  Yamauchi,  J.  Ikejiri,  K.  Tsunoda,  A.  Tanaka,  F.  Sato,  T.  Honma and  T.  Komatsu: Sci. Rep. 10 (2020) 9453. doi:10.1038/s41598-020-66410-1
• 13)   A.  Unemoto,  K.  Yoshida,  T.  Ikeshoji and  S.I.  Orimo: Mater. Trans. 57 (2016) 1639–1644. doi:10.2320/matertrans.MAW201601
• 14)   T.  Honma,  T.  Togashi,  N.  Ito and  T.  Komatsu: J. Ceram. Soc. Jpn. 120 (2012) 344–346. doi:10.2109/jcersj2.120.344
• 15)   T.  Honma,  N.  Ito,  T.  Togashi,  A.  Sato and  T.  Komatsu: J. Power Sources 227 (2013) 31–34. doi:10.1016/j.jpowsour.2012.11.030
• 16)   T.  Honma,  A.  Sato,  N.  Ito,  T.  Togashi,  K.  Shinozaki and  T.  Komatsu: J. Non-Cryst. Solids 404 (2014) 26–31. doi:10.1016/j.jnoncrysol.2014.07.028
• 17)   C.  Akatsuka,  T.  Honma and  T.  Komatsu: J. Ceram. Soc. Jpn. 126 (2018) 563–567. doi:10.2109/jcersj2.18040
• 18)   M.  Hiratsuka,  T.  Honma and  T.  Komatsu: Int. J. Ceram. Eng. Sci. 2 (2020) 332–341. doi:10.1002/ces2.10072
• 19)   K.  Nagamine,  K.  Hirose,  T.  Honma and  T.  Komatsu: Solid State Ionics 179 (2008) 508–515. doi:10.1016/j.ssi.2008.03.028
• 20)   S.  Nakata,  T.  Togashi,  T.  Honma and  T.  Komatsu: J. Non-Cryst. Solids 450 (2016) 109–115. doi:10.1016/j.jnoncrysol.2016.08.005
• 21)   H.  Kim,  I.  Park,  D.-H.  Seo,  S.  Lee,  S.-W.  Kim,  W.J.  Kwon,  Y.-U.  Park,  C.S.  Kim,  S.  Jeon and  K.  Kang: J. Am. Chem. Soc. 134 (2012) 10369–10372. doi:10.1021/ja3038646
• 22)   J.N.  Bridson,  S.E.  Quinlan and  P.R.  Tremaine: Chem. Mater. 10 (1998) 763–768. doi:10.1021/cm9704847
• 23)   X.  Pu,  H.  Wang,  T.  Yuan,  S.  Cao,  S.  Liu,  L.  Xu,  H.  Yang,  X.  Ai,  Z.  Chen and  Y.  Cao: Energy Storage Mater. 22 (2019) 330–336. doi:10.1016/j.ensm.2019.02.017
• 24)   P.  Barpanda,  T.  Ye,  S.  Nishimura,  S.-C.  Chung,  Y.  Yamada,  M.  Okubo,  H.  Zhou and  A.  Yamada: Electrochem. Commun. 24 (2012) 116–119. doi:10.1016/j.elecom.2012.08.028
• 25)   K.  Zaghib,  J.  Trottier,  P.  Hovington,  F.  Brochu,  A.  Guerfi,  A.  Mauger and  C.M.  Julien: J. Power Sources 196 (2011) 9612–9617. doi:10.1016/j.jpowsour.2011.06.061
• 26)   S.P.  Ong,  V.L.  Chevrier,  G.  Hautier,  A.  Jain,  C.  Moore,  S.  Kim,  X.  Ma and  G.  Ceder: Energy Environ. Sci. 4 (2011) 3680–3688. doi:10.1039/c1ee01782a
• 27)   S.N.  Yadav,  S.J.  Rajoba,  R.S.  Kalubarme,  V.G.  Parale and  L.D.  Jadhav: Chin. J. Phys. 69 (2021) 134–142. doi:10.1016/j.cjph.2020.11.020
• 28)   L.  Zhao,  D.  Zhou,  W.  Huang,  X.  Kang,  Q.  Shi,  Z.  Deng,  X.  Yan and  Y.  Yu: Int. J. Electrochem. Sci. 12 (2017) 3153–3165. doi:10.20964/2017.04.35
• 29)   Y.  Liu,  N.  Zhang,  F.  Wang,  X.  Liu,  L.  Jiao and  L.-Z.  Fan: Adv. Funct. Mater. 28 (2018) 1801917. doi:10.1002/adfm.201801917
• 30)   J.  Kim,  D.-H.  Seo,  H.  Kim,  I.  Park,  J.-K.  Yoo,  S.-K.  Jung,  Y.-U.  Park,  W.A.  Goddard, III and  K.  Kang: Energy Environ. Sci. 8 (2015) 540–545. doi:10.1039/C4EE03215B
• 31)   J.  Shao,  X.  Li,  J.  Wei,  H.  Pang and  C.  Chen: Adv. Sustain. Syst. 2 (2018) 1700154. doi:10.1002/adsu.201700154
• 32)   X.  Ma,  J.  Xia,  X.  Wu,  Z.  Pan and  P.K.  Shen: Carbon N. Y. 146 (2019) 78–87. doi:10.1016/j.carbon.2019.02.004
• 33)   M.  Hiratsuka,  T.  Honma and  T.  Komatsu: J. Alloy. Compd. 885 (2021) 160928. doi:10.1016/j.jallcom.2021.160928
• 34)   L.  Durai,  A.  Gopalakrishnan and  S.  Badhulika: Mater. Lett. 270 (2020) 127739. doi:10.1016/j.matlet.2020.127739
• 35)   K.  Kioka,  T.  Honma and  T.  Komatsu: Opt. Mater. (Amst) 33 (2011) 267–274. doi:10.1016/j.optmat.2010.08.028
• 36)   M.S.  Yeh,  Y.S.  Yang,  Y.P.  Lee,  H.F.  Lee,  Y.H.  Yeh and  C.S.  Yeh: J. Phys. Chem. B 103 (1999) 6851–6857. doi:10.1021/jp984163+
• 37)   M.  Kawasaki and  N.  Nishimura: J. Phys. Chem. C 112 (2008) 15647–15655. doi:10.1021/jp8056916
• 38)   Y.  Ishikawa,  N.  Koshizaki and  A.  Pyatenko: Electron. Commun. Jpn. 99 (2016) 37–42. doi:10.1002/ecj.11898
• 39)   Y.  Ishikawa,  N.  Koshizaki and  S.  Sakaki: J. Phys. Chem. C 123 (2019) 24934–24942. doi:10.1021/acs.jpcc.9b06949
• 40)   R.  Kihara,  A.  Shigetaka,  T.  Isshiki,  H.  Wada,  S.  Yamamuro and  T.  Asahi: Chem. Lett. 49 (2020) 413–415. doi:10.1246/cl.190947
• 41)   H.J.T.  Ellingham: J. Soc. Chem. Ind. 63 (1944) 125. doi:10.1002/jctb.5000630501
• 42)   N.  Doebelin and  R.  Kleeberg: J. Appl. Crystallogr. 48 (2015) 1573–1580. doi:10.1107/S1600576715014685
• 43)   K.  Momma and  F.  Izumi: J. Appl. Crystallogr. 41 (2008) 653–658. doi:10.1107/S0021889808012016