Deposition of Cu₂ZnSnS₄ layers by a novel single flash thermal evaporator from mixture element powder

Abstract

This research studies the deposition ability of Cu₂ZnSnS₄ from a mixed element powder and without sulfurization by a novel flash single thermal evaporator technique. Elemental powders of the source, S, Sn, Zn, and Cu, are held in an evaporation boat to mix their vapour together and to evaporate the mixture at a high temperature. The effect of substrate temperature on the thin layers structure properties, optical properties and surface morphology is explored by X-ray diffraction, Raman spectroscopy, scanning electron microscopy, energy dispersive X-ray, atomic force microscopy, and UV-Vis -NIR spectroscopy. Cu₂ZnSnS₄ films of kesterite structures are fabricated successfully at various substrate temperatures by this technique in one step without a sulfurization process. The substrate temperatures affect the composition and optical properties of the prepared layers by changing the concentration of elements and crystallinity. The layer prepared at 250°C exhibits an optimal stoichiometry and an optical bandgap value of 1.51 eV.

KEYWORDS:

• Cu2ZnSnS4 layers
• flash thermal evaporator
• kesterite structures
• solar cells
• thin films

1. Introduction

Cu₂ZnSnS₄ (CZTS) layers have been assumedfavourable layers solar cells devices due to many aspects, such as a high absorption coefficient of 10⁴·cm⁻¹, earth-abundant and suitable bandgap(1.0–1.5 eV), [1,2], low-cost and environmentally friendly materials. The maximum efficiency of CZTS solar cells is theoretically expected to be over 30% [3], whereas the highest experimental efficiency is 11.3 ± 0.3% [4]. Therefore, CZTS solar cells will be economically competitive compared to other conventional solar cells devices in the coming years.

CZTS thin layers can be achieved by either chemical or physical techniques. Chemical procedures include sol–gel methods [5,6], electrodeposition [7,8], spray pyrolysis [9,10], dip-coating [11] and spin-coating [12]. Physical vapour deposition techniques include pulse laser deposition [13], sputtering [14,15], electron beam [16], plasma-assisted coevaporation [17] and thermal evaporation [18].

Most chemical methods are cheaper than physical vapour deposition due to the cost of vacuum systems and their accessories. In contrast, the physical vapour deposition techniques provide very good films in terms of impurity and controlling the uniformity of the thickness and reduce toxic gases. Therefore, most industrial production and high-efficiency solar cells use vacuum processes.

The thermal evaporation system is the simplest among the physical vapour deposition techniques because it does not need an expensive high power supply generator or laser source for evaporating the materials. However, the source for a single thermal evaporation should be a compound of the elements [19]. Nevertheless, there is a possibility that the compound breaks into its elements, which may evaporate sequentially according to their evaporation temperature and vapour pressure, which in turn creates multiple layers rather than one layer of the compound. Therefore, coevaporation systems [20] have been used, which consist of more than one thermal evaporation source.

CZTS thin films can be either deposited by coevaporation only, which is a one-step process, or followed by sulfurization at high temperature.

In this study, we investigated the ability to prepare stoichiometric CZTS thin layers by a single thermal evaporator process direct from elemental powder and without sulfurization. A single thermal evaporation boat is designed to prevent the ejection of materials during raising the boat temperatures and to allow the components to mingle with each other before they sublimate simultaneously at high vapour pressure. Then, the prepared layers’ surface morphology and optical and structural properties at different substrate temperatures are inspected.

2. Experimental procedure

Copper, zinc, tin, and sulfur powders (99.99%) were weighed to 0.124, 0.023, 0.013 and 0.207 g, respectively. The powders were put in a directional evaporation boat (Testbourne Ltd, S17B-.010Ta). After that, the element powders (Cu, Sn, Zn and S) were mixed with each other by shaking the boat and then placed in the end of the boat from the inside, and 10 3-mm-diameter balls of high purity aluminium oxide (Al₂O₃) were inserted between the mixture and the exit hole. Then, the hole of the directional boat was closed by one ball. A weight of 0.725 g of bulk molybdenum metals was placed on the ball, as shown in Figure 1.

Figure 1. Sketch of the directional evaporation boat, (a) from outside and (b) from inside.

The vacuum system was pumped by a turbomolecular pump (Leybold TURBOVAC 191) to start a pressure of about 4.5 × 10⁻⁵ mbar. The temperature of the evaporation boat was increased quickly to around 1600°C. The ball and the weight fell off due to the high pressure of chemicals inside the boat. Then, the elements evaporated at the same time. The layers were prepared onto soda-lime glass at various substrate temperature (S.T.) of 100, 200, 250, and 350°C.

The diffractometer Shimadzu XRD-6000 was used to achieve the X-ray diffraction (XRD) of CZTS layers. The current and the high voltage of X-ray tube were 30 mA and 40 kV, respectively. Scanning electron microscopy (SEM) images were taken by the microscope Shimadzu SUPERSCAN SSX-550. Energy dispersive X-ray analysis (EDX) was achieved by a desktop SEM (JCM-6000). Surface morphology of layers was studied using Atomic force microscope (AFM) Veeco CP-II with Si tips in noncontact mode. The scan rate was 1 Hz.

The Shimadzu 3150 spectrophotometer was used to measure the optical absorption spectra. The spectral range of transmittance, T(λ) and the reflectance R(λ) was between 190 and 1500 nm. Raman spectroscopy was achieved by spectrometer Bruker Senterra. The wavelength of the laser was 532 nm and the power was 5 mW. The range of the spectra, which were collected from the samples, was between 200 and 600 cm⁻¹ with and 3.5 cm⁻¹ resolution.

3. Results and discussion

EDX analyses of the as-deposited layers of CZTS at different S.T. are illustrated in Figure 2. The variation in the atomic percentages of elements is shown in Figure 3. The elemental compositions are calculated in Table 1. It is obvious that increasing the S.T. reduces the atomic percentage of Zn and Sn significantly, which in turn increases the atomic percentage of Cu. The most stoichiometric films are deposited at 200 and 250°C. The optimal elements composition for solar cell device is Zn-rich and Cu-poor [21], which is achieved at a S.T. of 200°C, Copper/(Zinc + Tin) = 0.8 and Zinc/Tin = 1.11. The atomic percentage of sulfur and the atomic ratio of S/metal are very good for all deposited films in terms of stoichiometry. Therefore, the films do not need a sulfurization process.

Figure 2. EDX analysis of as-deposited layers of CZTS.

Figure 3. The variations in element composition of the prepared CZTS samples with substrate temperature.

Table 1. Elemental compositions of the deposited layers at various S. T.

Substrate	Atomic percentage
temperature (^oC)	(at%) of elements				Atomic ratio			Composition
	Cu	Zn	Sn	S	Cu/(Zn + Sn)	Zn/Sn	S/metal
100	2.1	26.9	19.8	51.2	0.045	1.36	1.05	Cu0.2Zn2.2Sn1.6S4.1
200	21.4	14.4	12.9	51.4	0.8	1.11	1.06	Cu1.7Zn1.1Sn1S4.1
250	25.5	15.4	10	49.1	1	1.54	0.96	Cu2Zn1.2Sn0.8S3.9
350	36.5	8	6.6	48.9	2.5	1.2	0.96	Cu2.9Zn0.6Sn0.5S3.9

XRD patterns of the as-prepared samples at different S.T are illustrated in Figure 4. It is clear that the layer prepared at 100°C is amorphous. Crystallinity appears at 200°C, but it is slight since there are no high and sharp peaks. The peaks at 28.4°, 32.7°, 47.3° and 56.1° related to the (112), (200), (220) and (312) planes, respectively. These planes confirm the structure of kesterite CZTS (JCPDS card number 26–0575) [18,22] and they are obvious in the layer deposited at 350°C. The preferred orientation in both layers, which were deposited at 250 and 350°C, is along the (112) crystallographic plane.

Figure 4. X-ray Diffraction patterns of the as-deposited CZTS samples at various S.T.

In the layer deposited at 350°C, the existence of other planes increases such as (200), (220) and (312). Some small peaks clearly appear due to the creation of secondary phases of either Cu₂S or SnS [23].

Lattice constants (a and c) and crystallite size (D) depend on the layer structure. Where the microstrain (ϵ) and dislocation density ρ are indicators of the existence of defects. These parameters are listed in Table 2. The lattice constants for the tetragonal phase structure are deduced using the following formula [24]: (1) $\frac{1}{d^2}=\left(\frac{h^2+k^2}{a^2}\right)+\left(\frac{l^2}{c^2}\right)$(1) d is the interplanar distance between planes and h, k, and l are the Miller indices.

Table 2. Diffraction of X-ray data of layers deposited at 200°C, 250 and 350°C.

Substrate	crystallite size (D)	Lattice Parameter		c/2a	Microstrain (ϵ)	Dislocation density ρ
temperature (^oC)	Å	a (Å)	c (Å)	ratio	×10^−3	(Lines/m^2) × 10^16
200	37.22	5.39	11.29	1.04	9.31	7.2
250	158.35	5.47	11.06	1.01	2.19	0.3
350	245.46	5.45	11.02	1.01	1.41	0.16

Sherrer’s formula is used to calculate the crystallite size (D) [25]: (2) $D=\frac{0.9\lambda }{\beta \text{cos}(\theta )}$(2)

β is the full width at half-maximum (FWHM) in radians, θ is the Bragg angle and λ is the wavelength of the X-ray.

The microstrain (ϵ) [26] and dislocation density (ρ) [27] are deduced by the following equations: (3) $\begin{array}{rl}\epsilon & =\frac{\beta \text{cos}(\theta )}{4}\end{array}$(3) (4) $\begin{array}{rl}\rho & =\frac{1}{D^2}\end{array}$(4)

These values of lattice constants confirm that the structure is tetragonal since the value of c/2a is close to unity. The microstrain and the dislocation density decrease as the substrate temperature increases due to improving the crystallinity. Large grain size of absorber material have important impacts on the solar cell energy conversion solar cells [28].

Raman spectroscopy confirms the existence of the CZTS structure and the effect of substrate temperature on this structure. As shown in Figure 5, there is no peak for the layer deposited at 100°C, confirming that the layer does not have a CZTS structure, as proven by the XRD results. The peaks of the Cu₂ZnSnS₄ structure existed at 329 cm⁻¹, 334 cm⁻¹ and 337 cm⁻¹ for films deposited at 200, 250 and 350°C, respectively. The main peak of CZTS is refer to the vibration of A1 mode arising from the sulfur atoms vibrations of in the Cu₂ZnSnS₄ lattice, whereas the rest of the atoms remain stationary [22]. The intensity of the peak increased from 29.6 to 378 as the S.T. increased from 200 to 250°C and then decreased to 283 at a S.T. of 350°C. Increasing the intensity of the peak indicates that the CZTS structure improved in the layer. Therefore, the film prepared at 250°C has a relatively best CZTS structure. The weakness of the peak at a S.T. of 200°C indicates that CZTS has some structural disorder (e.g. Cu and Zn cation disorder) [29,30]. However, the smallest FWHM value of the film deposited at a S.T. of 350°C suggests a reduction in the structural disorder or anti-site defects and increased crystallinity. The Raman peak shift from 329 cm⁻¹–337 cm⁻¹ can be explained by increasing the bond strength or reducing the mass of the elements according to Keating’s model [31,32].

Figure 5. Raman spectra of CZTS layers prepared at various S.T. of 100, 200, 250 and 350°C.

The SEM images with 2 kx magnification are shown in Figure 6. All films lack voids and cracks with changes in the shape of the grain size and roughness. Most of the surface of the film prepared at a S.T. of 100°C is smooth with some particles on the surface, confirming that the film is amorphous. The grains became larger and interconnected as the S.T. increases. This change is also observed by AFM images, as shown in Figure 7. While the crystallite size increased from 37.22–245.46 Å as listed in Table 2. The value of the root mean square (RMS) roughness increases from 35.4–57.8 nm as the S.T. increases from 100 to 350°C.

Figure 6. SEM surface images of CZTS layers deposited at various S. T.

Figure 7. AFM surface images of CZTS layers deposited at various S.T.

The optical absorbance (A) of the films was calculated from T and R as shown in Figure 8, by the following equation [33]: (5) $A=1-(T+R)$(5) In general, increasing the S.T. increases the absorbance of the layers due to improving the film crystallinity. The absorbance of the film prepared at 200°C is higher than that deposited at 250°C, possibly because of the differences in compositions, as illustrated in Table 1.

Figure 8. Optical absorbance of Cu₂ZnSnS₄ layers prepared at various S.T.

The following equation:[34] was used to deduce the absorption coefficient, α,: (6) $\alpha =\frac{1}{d}\text{In}\left\{\frac{{(1-R)}^2}{2T}+{\left[\frac{{(1-R)}^4}{4T^2}+R^2\right]}^{1/2}\right\},$(6) where d is the layer thickness.

As shown in Figure 9, the absorption coefficient increases as the S.T. increases, especially in the visible range. The high increase in the film prepared at 350°C is not due to the substrate temperature only but is affected more by the change in the composition of the film, which in turn reduces the value of the band gap E_g.

Figure 9. The absorption coefficient of CZTS layers prepared at various S.T.

E_g is deduced from the following equation: (7) $(\mathrm{\alpha h\nu })=A[(\mathrm{h\nu }-E_{\text{g}}){]}^{1/2}$(7) where hν is the photon energy and A is a constant. Extrapolation of the curves down to a zero (αhν)² level yields E_g values of 1.76, 1.7, 1.51 and 1 eV for thin films prepared at 100, 200, 250 and 350°C, respectively, as shown in Figure 10. The variation in the bandgap depends on the S.T. which mainly affects the composition and crystallinity of the layers. The optimal band gap is ∼1.51 eV [35], which for the layer prepared at 250°C had the best composition, as listed in Table 1. Then, thin films with Cu/(Zn + Sn) stoichiometry of ∼1 has promising potential for thin film solar cell application [36]. The reduction of the E_g to approximately 1 eV could be due to the electronic states created in the band gap and serve as carrier trapping recombination centres. These carrier trapping centres existed as a result of the Cu_2-xS secondary phase [37]. This could be the case with the layer prepared at 350°C because the film was copper rich and had a low concentration of zinc and tin.

Figure 10. E_g of the layers deposited at various S.T.

4. Conclusions

In this study, Cu₂ZnSnS₄ layers were prepared successfully by a new process of flash single thermal evaporation. The source was a mixture of the elements without ball milling or sintering. An evaporation boat was designed to confine the elements tightly at high temperatures before they evaporated simultaneously. The layers were prepared at S.T. of 100, 200, 250 and 350°C. The film that was deposited at 100°C was nonstoichiometric and amorphous since it had no X-ray or Raman peaks. Increasing the substrate temperature reduced the concentrations of Zn and Sn to obtain the best stoichiometry and kesterite structures at 200 and 250°C. The crystallite size and roughness increased as the substrate temperature increased, whereas the microstrain and dislocation density decreased. The peak of the Raman shift CZTS structure increased from 329 cm^−1–337 cm⁻¹ for layers as the S.T. increased from 200 to 350°C, whereas the FWHM value decreased. The absorbance and the absorption coefficient of the layers increased with S.T., but the E_g decreased as a result of modifications changes in the crystallinity and the elements concentration of the layers. The optimal band gap was achieved at approximately 1.51 eV for the film deposited at 250°C.

Disclosure statement

No potential conflict of interest was reported by the author(s).