Optical properties of CdS/CdTe heterojunction prepared by physical vapor deposition technique

The paper presents the study of the optical properties of a thin layer of Cadmium Sulphide deposited on Cadmium Telluride films. CdTe thin films were obtained by vapor phase condensation method using different technological factors, in particular, different thickness (different time of deposition τ) on glass substrates. After deposition the optical properties were analysed by Swanepoel method, using transmission spectra. The upper thin layer of CdS was deposited by thermal evaporation method on CdTe thin films. The change in optical properties of CdS/CdTe heterojunction in comparison with CdTe thin films was investigated. Using a Swanepoel method were calculated the main optical constants, such as refractive index, absorption coefficient and optical conductivity. By this method the thickness of the thin film was determined and compared with the experimental values obtained by the profilometer.


Introduction
Thin film PV solar cell has been considered one of the promising solar cells due to its high energy conversion efficiency, low cost and convenience for large scale production. The most successful thin film solar cells have been cadmium telluride (CdTe), copper indium gallium selenide (CIGS) and amorphous silicon (a-Si) with efficiencies of 18.3 %, 20 % and 12.3, respectively [1]. Typically, the efficiencies of thin-film solar cells are lower compared with silicon (wafer-based) solar cells, but manufacturing costs are also lower. It was reported that [2], CdTe technology costs about 30% less than CIGS technology and 40% less than A-Si technology.
The theoretical efficiency of CdS/CdTe solar cells is predicted to be up to 28 -30 % [3, 4]. However, the real efficiency of PV solar cells based on the n-CdS / p-CdTe heterojunction in a superstrate structure is currently 20.4 % [5], and the efficiency of solar modules with an area > 1 cm 2 is 16.5 % [6,7] The major impact factors for this difference are due to the optical losses, surface recombination, recombination in the space-charge region and rear contact effect.
The best small-area CdTe thin-film cells manufactured show more than 15 % conversion efficiency [8]. Large-area modules with aperture efficiencies in excess of 10 % have also been demonstrated (Ullal et al., 2000). First Solar (the largest CdTe manufacturer) reported fleet average efficiencies increasing from 12.9 % in 2012 to 16.6 % in 2016 for their CdTe modules (First Solar, 2017). For CdTe cells, module efficiency record for the moment is 18.6 %. The best CIGS reported efficiencies so far were 17.5 % for modules (Green et al., 2017). There is also significant industrial production based on CdTe/CdS solutions, represented to a large extent by American corporation First Solar, which is a supplier of PV modules used in the currently largest Agua Caliente Solar Project solar power plant in Arizona [9]. It should be noted that thin film technology based on CdTe is the first technology that has allowed to low the production costs of solar energy up to 0.57 $/W [10]. In spite of the 10 % difference in the lattice constants of CdS and CdTe, they form an electrically excellent heterojunction, as shown by its high fill factors up to FF = 0.75 in the devices made.
Thin film flexible solar cells using CdS/CdTe semiconductor compounds are currently in active research and in the field of interest of several research centers in the world. This is due to the fact that flexible PV cells have a record high power per unit weight -more than 2 kW/kg [11]. Such characteristics are achieved thanks the construction of a flexible SC, where the glass substrate is replaced by a polyamide film. In recent years there have also been reports of the first attempts to manufacture this type of structures on elastic substrates, including both configurations: superstrate [12][13] and substrate [14].
At present, a lot of methods have been developed for the production of thin CdTe films. This paper presents the thermal evaporation method [15]. In order to calculate the refractive index from optical transmittance data the interference patterns in the transmittance spectra should be suppressed by generating the envelope around transmittance maxima and minima [16]. Than the interference pattern free transmittance spectra of deposited films should be used to calculate the refractive index using Swanepoel method [17]. This straightforward method proposed by Swanepoel [18], based on the use of the extremes of the interference fringes of transmission spectrum only, will be used in order to derive the real and imaginary parts of the complex index and also the thickness for the semiconductor film. Through the use of materials with a higher optical absorption coefficient, it is possible to reduce the thickness of active photovoltaic cell layers [19], which affects not only the decrease in production costs, but also the possible weight reduction of photovoltaic devices depending on the substrate used.
There are number of papers theoretically focused on detailed investigation of the dependence of the efficiency of CdS/CdTe thin film solar cells on some properties [20 -24]. However, the results of complex experimental studies of the spectral dependences of the main optical constants have been presented a little. Therefore, our studies are of great importance for the further development of highly efficient devices based on heterojunctions for electronics and solar energy.

Experiment Methodology
First the thin films of CdTe were deposited on a cleaned glass substrates by thermal evaporation method. In the used installation can be to receive series (5 -15 films) in a single cycle for various technological factors: different thickness d = (0.01 -12) μm at a constant temperature of deposition T S = (300 -570) K; uniform thickness d with different T S ; different temperature of evaporation T E (600 -1070) K with constant thickness d or deposition temperature T S .
Thin films of CdTe for research were obtained with different thicknesses (different deposition time τ) at a constant T S and T E ( Table 1). The growth temperature T s was 470 K, the evaporation temperature of presynthesized compounds CdTe was T E = 870 K. The thicknesses of thin films were set by deposition time τ = (60 -180) sec.
Further the next layer of CdS was deposited on CdTe thin films by thermal evaporation method. The technological parameters shows Table I. The samples thickness was analysed using profilometer Bruker Dektak XT. Optical transmission spectra was investigated by measuring transmittance, T at normal incidence and room temperature. The measurements were carried out in the wavelength range of 190 -3300 nm with 1 nm step using Agilent

Results and discussion
Optical characterization of thin films gives information about other physical properties, e.g., bandgap energy, band structure, and optically active defects [25]. The effects of thickness and heterojunction properties on the optical transmittance of the CdTe and CdS/CdTe films have been studied. The region of fundamental absorption was observed in transmission spectra. The transmission spectra of CdTe and CdS/CdTe thin films obtained on the glass substrates with different thicknesses were measured in the wavelength range from 180 to 3300 nm and are shown in Figures 1-2. It can be observed that the films are highly transparent in the near infrared region. Absorption edge is about 800 nm for all samples, which is completely consistent with the width of the bandgap of CdTe thin films [26]. In addition, the observed interference patterns in the optical transmission spectra are the indication for the thickness homogeneity of deposited films [27].
On Figs. 1-2 can be seen that for as grown CdTe films the transmission values reach up to 90 %, which also indicates the high transparency of the films. For a thick film CdTe No21 (Fig.1) is observed a smoother growth of the transmittance with a wavelength compared to the sample CdTe No24 (Fig. 2). Here is worth to note, that the number of "interference maxima" also depends on the thickness of the film. This could be explained by the fact that there is a difference between the refractive indexes of the film and substrate and also due to the interference of multiple light reflections Fig. 3 [28].
Transmission values for CdS/CdTe heterojunction are slightly lower than those for pure CdTe films, which may indicate greater reflection or scattering in film thickness.
One of the most popular methods that uses these interference fringes to determine the optical properties of the material is the Swanepoel method [29]. The shape of a thin film on a transparent substrate looks like on Fig. 3. In this figure, n, α, d and T denote the refractive index, absorption coefficient, thickness and transmission of the film, respectively. The transparent substrate has a thickness of several orders of magnitude larger with index of refractions and absorption coefficient α s = 0 and transmittance T s . The index of refraction for air is n 0 = 1. If the thickness d is uniform then the interference effects generate a spectrum. The interference fringes can be used to calculate the optical constants of the film such as refractive index, film thickness, absorption coefficient and optical conductivity.
The transmission T for the normal incidence resulted from the interference of the wave transmitted from three interfaces can be calculated as [30]: where: Maximum and minimum of interference fringes are determined from the following equations: The refractive index of the substrate is found from the following expression (T s =max): From the equations above the refractive index is determined: where Using the equations (11) Table  II and Table III, respectively, as d 1 .
Comparing the experimental and calculated values of the thickness, can be note that a small difference in value is associated with the error in experimental studies.
The order of interference m at the maxima of transmission spectra for wavelength λ 1 is:      where λ 1 and λ 2 are the wavelengths of two adjacent transmission maxima (λ 1 > λ 2 ).

exp 1485 nm
For a region where is a strong absorption of light, the refractive index is determined as follows: For heterojunction CdS/CdTe takes place a significant increase in the absorption coefficient of light several times, primarily in the short-wave region of the spectrum (Fig. 6-7). This is due to the fact that the thin layer of CdS thanks to the large band gap (2.42 eV) plays the role of an «absorption window» for absorbing light.
The photons absorbed in the window layer do not contribute to the photocurrent, as recombination is very likely to occur, resulting in scattering of light. Therefore, absorption in the CdS layer is a source of significant loss.