Characterization of Cu–Bi–S Powders Synthesized by Polyol Method

Yasuhiro Shirahata; Akira Nagaoka; Hideaki Araki; Takashi Kawakubo

doi:10.2320/matertrans.MT-M2022035

Abstract

Cu–Bi–S powders were synthesized by a polyol method at 180°C and a speed of 600 rpm for 70 min. X-ray diffraction patterns of the Cu–Bi–S powders showed that single-phase Cu₃BiS₃ was obtained at a thiourea concentration of 197 mmol·L⁻¹. Microstructural analysis of the Cu–B–S powders revealed that the sizes of the Cu–Bi–S particles with single-phase Cu₃BiS₃ were 200–400 nm and their composition was Cu-poor, Bi-rich, and slightly S-poor. From the optical measurement by an ultraviolet-visible-near infrared spectroscopy, the optical bandgap of the Cu–B–S powder with single-phase Cu₃BiS₃ was estimated to be 1.24 eV. Based on the obtained results, the influences of thiourea concentration on synthesis of Cu₃BiS₃ were discussed.

1. Introduction

Ternary and quaternary copper (Cu)-based chalcogenide compound semiconductors have attracted much attention due to their high absorption coefficients (>10⁴ cm⁻¹) and appropriate bandgaps for light absorption in the visible spectral region.¹^–⁶⁾ Cu-based chalcogenide compound semiconductors also show considerable promise for a wide range of applications, e.g., solar cells,⁶⁾ light emitting diodes,⁷⁾ bioimaging,⁸⁾ and thermoelectric devices.⁹^,¹⁰⁾ However, the resources of indium (In), gallium (Ga), and tellurium (Te), which are often used in CdTe, CuInSe₂ (CIS), and Cu(In,Ga)(S,Se)₂ (CIGS), are limited in the earth’s crust.¹¹⁾ Moreover, the use of selenium (Se) in CIS and CIGS is not recommended from the perspective of environmental protection and health problem.¹²⁾ Therefore, low-toxicity, earth-abundant, and cost-effective component elements are required.

Among Cu-based chalcogenide compound semiconductors, Cu₃BiS₃ (wittichenite) is a new emerging semiconductor because bismuth (Bi) is an inexpensive and low-toxic element.¹³⁾ The reported optical bandgap of Cu₃BiS₃ is between 1.4 and 1.7 eV, close to the optimum value for efficient solar energy conversion.¹²^,¹⁴^–¹⁸⁾ High optical absorption coefficient (>10⁴ cm⁻¹) of Cu₃BiS₃ enables sufficient light absorption in solar cells.¹⁷^,¹⁸⁾ Several groups have reported the properties of Cu₃BiS₃ films and powders using various different growth techniques, e.g., spray pyrolysis,¹⁶⁾ spin-coating,¹⁷⁾ co-evaporation,¹⁸⁾ reactive sputtering,¹⁹⁾ solid state reaction,²⁰⁾ chemical bath deposition,¹²^,²¹⁾ hydrothermal synthesis method,²²^,²³⁾ and polyol method.²⁴⁾ Among them, polyol method is a chemical technique involving the forced hydrolysis of transition metal salts, which enable to prepare nano- and micro-sized structures of elemental metals, metal oxides, and alloys with controlled size, texture, and shape.²⁵⁾ Polyol method also enables rapid (1–2 h) and low-temperature synthesis (<200°C) of nano- and micro-sized chalcogenide compounds.²⁴^,²⁶^–²⁸⁾ In the synthesis of chalcogenide compounds by polyol method, thiourea is often used as a sulfur source.²⁴^,²⁶^–²⁸⁾

The purpose of the present work is to characterize Cu–B–S powders synthesized by a polyol method. The microstructures and optical properties of the Cu–Bi–S powders were investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDX), and ultraviolet-visible-near infrared (UV-Vis-NIR) diffuse reflectance spectroscopy.

2. Experimental Procedure

The Cu–Bi–S powders were synthesized by a polyol method.²⁴^,²⁶^–²⁸⁾ Figure 1 shows a schematic illustration of synthesis process of the Cu–Bi–S powders. The synthesis was carried out in ambient air, except for drying in vacuum. The Cu–Bi–S precursor solutions were prepared by mixing CuCl (Fujifilm Wako Pure Chemical, 297.2 mg), BiCl₃ (Fujifilm Wako Pure Chemical, 347.8 mg), and thiourea (Fujifilm Wako Pure Chemical) in ethylene glycol (Fujifilm Wako Pure Chemical, 50 mL). The molar concentrations of each starting material are presented in Table 1. Excess of 10% BiCl₃ was added to each solution, aiming to minimize the bismuth loss during the synthesis. The molar concentrations of thiourea were 132, 197, and 263 mmol·L⁻¹, and the samples were labeled as CBS-1, CBS-2, and CBS-3, respectively. The Cu–Bi–S precursor solutions were stirred by a magnet stirrer (As One DP-2S) at 180°C and a speed of 600 rpm for 70 min. During the synthesis, the temperature was checked by a radiation thermometer. The fluctuation of the temperature was within 3°C. After cooling the solutions to room temperature, the solids were filtered and washed with methanol (Fujifilm Wako Pure Chemical). Finally, the resulting solids were dried in vacuum at room temperature for 24 h.

Fig. 1

Schematic illustration of synthesis process of Cu–Bi–S powders.

Table 1 Molar concentrations of starting materials.

XRD patterns of the Cu–Bi–S powders were recorded using an X-ray diffractometer (Rigaku, Miniflex 600) equipped with a CuK_α radiation source (λ = 1.5406 Å) at setting voltage of 40 kV and setting current of 15 mA in the 2θ range of 10–80° with step width of 0.01°. Microstructural analysis were performed using a field-emission scanning electron microscope (Jeol, JSM-6701F) equipped with an EDX detector (Jeol, JED-2300) at accelerating voltage of 15 kV. Diffuse reflectance spectra of the samples were collected using an UV-Vis-NIR spectrophotometer (Shimadzu, UV-2600) equipped with an integration sphere (Shimadzu, ISR-2600 Plus) in the wavelength range of 185–1400 nm. BaSO₄ powder was used as standard sample for the baseline. All the measurements were carried out at room temperature.

3. Results and Discussions

Figure 2 shows XRD patterns of the Cu–Bi–S powders. For CBS-1, diffraction peaks derived from Cu₃BiS₃ were not observed, while secondary phases, such as CuS (ICDD No. 00-006-0464), and Cu_1.8S (ICDD No. 01-073-8624), were observed. In addition to the secondary-phase peaks, some unknown peaks were observed at 2θ = 24.2°, 25.9°, 28.1°, 33.5°, and 36.6°. The secondary-phase peaks and the unknown peaks were disappeared with increasing the molar concentration of thiourea, indicating that the thiourea concentration of 132 mmol·L⁻¹ was not sufficient for the synthesis of Cu₃BiS₃. For CBS-2 and CBS-3, main diffraction peaks derived from Cu₃BiS₃ were observed at 2θ = 23.2°, 28.1°, 29.0°, 31.2°, and 33.9° corresponding to 200, 211, 102/031, 131, and 122, respectively, indicating that CBS-2 and CBS-3 have Cu₃BiS₃ with an orthorhombic wittichenite structure (ICDD No. 01-073-1185). However, the XRD pattern of CBS-3 showed a broad peak structure at 22–36° (gray region). This result can be interpreted that diffraction reflections derived from Bi₂S₃ (ICDD No. 01-074-9437) overlap with those of Cu₃BiS₃, indicating that CBS-3 contains Bi₂S₃. Thus, single-phase Cu₃BiS₃ was obtained only in CBS-2. The lattice constants of CBS-2 were estimated and are presented in Table 2. The estimated lattice constants were slightly smaller than those of an ICDD database (ICDD No. 01-073-1185). The crystallite size (D) and strain (ε) of CBS-2 were estimated and are presented in Table 2. The D and ε were estimated using the Williamson-Hall equation:²⁹⁾

\begin{equation} \beta \cos \theta = 4\varepsilon \sin \theta + \frac{K\lambda}{D} \end{equation}

(1)

where, β, θ, K, and λ are the full width at half maximum, Bragg angle, shape factor, and X-ray wavelength, respectively. In the present work, the K of 0.94 was introduced into eq. (1) because the XRD patterns were fitted by a Gaussian function. Using eq. (1), the values of D and ε were spontaneously estimated. The term β cos θ was plotted with respect to 4 sin θ, and the slope and y-intersect of the fitted line represent ε and D, respectively. The result about ε is suggested that CBS-2 is under small tensile stress.

Fig. 2

XRD patterns of Cu–Bi–S powders. The grey region at 2θ = 22–36° highlights a broad peak structure of CBS-3, derived from diffraction reflections by Bi₂S₃.

Table 2 Lattice constants, crystallite size (D), and strain (ε) of CBS-2.

Figure 3 shows SEM images of the Cu–Bi–S particles. For CBS-1 (Fig. 3(a)), sphere- and flower-like structures were observed. For CBS-2, which has single-phase Cu₃BiS₃ (Fig. 3(b)), sphere-like structures were observed. The particle sizes were 200–400 nm. For CBS-3 (Fig. 3(c)), sphere- and rod-like structures were observed. The lengths of the rod-like structures were 1–2 µm. As reported by some groups, the rod-like structures might be Bi₂S₃.³⁰^,³¹⁾

Fig. 3

SEM images of Cu–Bi–S particles (a) CBS-1, (b) CBS-2, and (c) CBS-3.

The compositions of the Cu–Bi–S powders were analyzed, and the average relative contents of Cu, Bi, and S are shown in Fig. 4. The present EDX analysis was carried out at four different measurement points for each sample, and the EDX data were corrected using a ZAF program included in the EDX software. Although the EDX values contain some errors, the contents of Cu and Bi clearly varied with increasing the molar concentration of thiourea. CBS-1 showed a Cu-rich, Bi-poor, and slightly S-poor composition, attributed to the secondary phases confirmed by the XRD measurement shown in Fig. 2. CBS-2, which has single-phase Cu₃BiS₃, showed a Cu-poor, Bi-rich, and slightly S-poor composition. This result suggests that Cu vacancy (V_Cu) is dominant in CBS-2 because V_Cu is easily formed in Cu-based sulfide or selenide compounds.³²^,³³⁾ CBS-3 showed a Cu-poor, Bi-rich, and S-rich composition. This result supports the result that Bi₂S₃ is contained in CBS-3.

Fig. 4

Average relative contents of Cu, Bi, and S in Cu–Bi–S powders as a function of thiourea concentration.

Figures 5(a) and (b) diffusive reflectance spectra and Kubelka–Munk plots (F(R_∞)) of the Cu–Bi–S powders, respectively. The F(R_∞) was calculated using the following equation:³⁴^,³⁵⁾

\begin{equation} F(R_{\infty}) = \frac{(1 - R_{\infty})^{2}}{2R_{\infty}} \end{equation}

(2)

where, R_∞ = R_sample/R_standard is the reflectance of the examined sample. For CBS-1, peak structures derived from CuS and Cu_1.8S were observed.³⁶^,³⁷⁾ For CBS-2 and CBS-3, spectra derived from wittichenite Cu₃BiS₃ were observed.¹²⁾ The optical bandgap (E_g) value of CBS-2, which has single-phase Cu₃BiS₃, was estimated by the Tauc equation:

\begin{equation} [F(R_{\infty})hv]^{n} = A(hv - E_{g}) \end{equation}

(3)

where, h, ν, A, and n are the Planck constant, light frequency, proportional constant, and power index, which depends on the nature of the transition, respectively. In the present study, n = 2 was used because Cu₃BiS₃ is a direct forbidden transition semiconductor.¹²^,¹⁶^–¹⁸^,²¹⁾ Figure 5(c) displays the Tauc plot of CBS-2. From the intersection of the dashed line, the E_g value of CBS-2 was estimated to be 1.24 eV. This value was smaller than the reported ones (1.4–1.7 eV),¹²^,¹⁴^–¹⁸^,²¹⁾ likely due to the difference in shape.

Fig. 5

(a) Diffuse reflectance spectra and (b) Kubelka–Munk plots (F(R_∞)) of Cu–Bi–S powders. (c) Tauc plot of CBS-2.

Finally, reaction mechanism of the present Cu–Bi–S powders is commented. It has been reported that formation of Cu₃BiS₃ in solution is through the following steps:²²^,²³⁾ first, Cu⁺ and Bi³⁺ combine with thiourea molecules to form [Cu–Thiourea]⁺ and [Bi–thiourea]³⁺ complexes in the solution, which can be expressed as follows:

\begin{equation} \text{CuCl} + \text{thiourea}\to \text{[Cu–thiourea]$^{+}$} + \text{Cl$^{-}$} \end{equation}

(4)

\begin{equation} \text{BiCl$_{3}$} + \text{thiourea}\to \text{[Bi–thiourea]$^{3+}$} + \text{3Cl$^{-}$} \end{equation}

(5)

Second, at a given temperature, the stability of [Cu–Thiourea]⁺ and [Bi–thiourea]³⁺ complexes decrease and undergoes thermal decomposition. Here, thiourea reacts with moisture (H₂O), contained in the air and/or reagents, to give S²⁻.³⁸^,³⁹⁾ Therefore, the S²⁻ reacts with Cu⁺ and Bi³⁺ to form Cu₃BiS₃, which can be expressed as follows:

\begin{align} &\text{3[Cu–thiourea]$^{+}$} + \text{[Bi–thiourea]$^{3+}$}\\ &\quad \to \text{Cu$_{3}$BiS$_{3}$} + \text{thiourea} \end{align}

(6)

In the present work, when the thiourea concentration was 132 mmol·L⁻¹, Cu₃BiS₃ was not formed, and secondary phases such as CuS and Cu_1.8S were formed (CBS-1). Also, when the thiourea concentration was 263 mmol·L⁻¹, Cu₃BiS₃ and Bi₂S₃ were formed (CBS-3). Bi₂S₃ was likely formed by Bi³⁺ and S²⁻ decomposed from Bi–thiourea complex in the precursor solution. Thus, in the present work, the thiourea concentrations of 197 mmol·L⁻¹ (CBS-2) was appropriate to obtain single-phase Cu₃BiS₃. To confirm whether single-phase Cu₃BiS₃ powder is obtained in CBS-2, further investigations by Raman spectroscopy¹²^,²¹⁾ and X-ray photoelectron spectroscopy¹⁷^,²²^,²⁴⁾ are required.

4. Conclusion

Cu–Bi–S powders were synthesized by polyol method at 180°C and a speed of 600 rpm for 70 min, and their microstructures and optical properties were investigated. The XRD patterns of the Cu–Bi–S powders showed that single-phase Cu₃BiS₃ was obtained at a thiourea concentration of 197 mmol·L⁻¹. The sizes of the Cu–Bi–S particles with single-phase Cu₃BiS₃ were 200–400 nm, and their composition was Cu-poor, Bi-rich, and slightly S-poor. The optical bandgap of the Cu–B–S powder with single-phase Cu₃BiS₃ was estimated to be 1.24 eV. From the obtained results, single-phase Cu₃BiS₃ was obtained at the thiourea concentrations of 197 mmol·L⁻¹, whereas single-phase Cu₃BiS₃ was not obtained at 132 and 263 mmol·L⁻¹. These findings can be used to synthesize ternary or quaternary compound semiconductors by polyol method.

Acknowledgments

This work was partly supported by JSPS KAKENHI Grant-in aid for Early-Carrier Scientists (Y.S., JP18K14316) and Collaborative Education Center for Emerging Technology of Kagawa college (Y.S.).

REFERENCES

Corresponding author

Register with J-STAGE for free!