Bridging materials and devices: Selenium-driven bandgap and transport tuning in CdTe1−xSex solar cells

3.1. Structural and morphological analysis of the CdTe_1-xSe_x thin films

XRD measurements of the deposited films are employed to investigate the influence of compositional variation on the crystalline structure of the prepared layers. The diffraction patterns obtained from the as-deposited CdTe thin films exhibit a dominant diffraction peak located at 2θ = 23.9°, which is attributed to the (111) plane of the cubic zinc blende phase of CdTe. This finding is consistent with the standard data reported in the ICDD card No. 15-0770, confirming the formation of a single-phase crystalline structure. The interplanar spacing (d) corresponding to the (111) reflection is determined using Bragg’s law, nλ=2dsin(θ), yielding a value of d₁₁₁=3.7 Å. The lattice constant (a) for the cubic CdTe structure is subsequently calculated from the general relation:1/d² = (h²+k²+l²)/a², where (hkl) represent the Miller indices of the reflecting plane. The calculated lattice parameter is found to be a = 6.44 Å, which is in good agreement with the reported standard value for cubic CdT. In the case of CdTe_1–xSe_x alloy films, the diffraction profiles display reflections originating from both CdTe and CdSe phases, as presented in Fig. 1. The as-deposited CdTe_1–xSe_x samples exhibit two main diffraction peaks at approximately 2θ = 23.9° and 25.6°, which correspond to the (111) plane of CdTe and CdSe, respectively. A noticeable correlation between the Se/Te composition ratio and the relative intensity of these peaks is observed. As the selenium content increases, the peak intensity at 25.6° (CdSe phase) becomes more pronounced, while the CdTe peak intensity gradually decreases. For the composition CdSe_0.6Te_0.4​, the CdSe-related peak intensity slightly exceeds that of CdTe, indicating a shift toward Se-rich structural characteristics. Conversely, in Te-rich films, the diffraction pattern is dominated by the (111) reflection of CdTe at 23.9°, confirming that the crystalline structure is strongly influenced by the relative concentration of Se and Te in the deposited layers. These results clearly demonstrate the compositional dependence of phase formation and the coexistence of CdTe and CdSe phases within the ternary CdTe_1–xSe_x system.

Close

Fig. 1.

(a) XRD pattern of CdSe_xTe_1-x thin films at doping concentrations of (x=0, 0.2, 0.4, 0.6, 0.8, and 1 wt. %). (b-d) shows the deconvolution of main peaks at x = 0.4, 0.6 and 0.8 wt. %, respectively.

Export to PPT

In addition to the dominant diffraction peaks corresponding to the primary CdTe and CdSe phases, minor secondary phases are also detected in the intermediate compositions of the CdTe_1–xSe_x alloy system. For samples with selenium content in the range x = 0.2 to 0.4, weak reflections corresponding to CdSe_0.1Te_0.9​ were identified. These reflections are associated with the cubic zinc blende phase, in agreement with the standard ICDD card No. 41-1324. The presence of these small traces indicates a partial solid-solution formation between CdTe and CdSe during film growth, where selenium atoms partially substitute tellurium sites within the CdTe lattice, giving rise to a slightly modified cubic phase. At higher selenium concentrations (x =0.6 and x =0.8), additional weak peaks are observed, which correspond to CdSe_0.6​Te_0.4​ with a hexagonal wurtzite-type structure, as confirmed by comparison with ICDD card No. 41-1325. This structural transition from a cubic to a hexagonal phase with increasing Se content reflects the gradual evolution of the crystal lattice as the alloy composition shifts toward Se-rich regions. Such coexistence of the cubic and hexagonal phases is a characteristic of CdTe–CdSe solid solutions and often occurs due to the limited solubility of Se and Te in the crystal structure of each other under specific deposition or annealing conditions. Figs. 1(b-d) presents the deconvolution of the main diffraction peaks for compositions x=0.4, 0.6 and 0.8, respectively. The deconvoluted patterns clearly reveal the emergence of minor peaks that correspond to the secondary phases. For x=0.4, subtle traces of CdSe_0.1Te_0.9 can be identified near the main CdTe peak region, confirming the incorporation of a small fraction of Se within the CdTe matrix. Conversely, for x=0.6 and 0.8, the appearance of an additional weak reflection corresponding to CdSe_0.6​Te_0.4​ confirms the formation of a Se-rich hexagonal phase, which indicates that phase segregation begins to occur as the composition approaches higher selenium concentrations.

Fig. 2 illustrates the SEM micrographs of CdTe_1–xSe_x thin films with different selenium concentrations: (a) x =0.0 (pure CdTe), (b) x = 0.4, and (c) x = 1 (pure CdSe). The surface morphology of each film is carefully examined to evaluate the influence of the selenium incorporation on the microstructural features of the deposited layers. The SEM image of the CdTe film (x = 0) shows a relatively uniform grain distribution with a compact and well-adhered surface. The grains appear densely packed with minimal voids, suggesting good film coverage over the substrate. Upon partial substitution of tellurium with selenium (x = 0.4), the film maintains a homogeneous texture with slightly finer grains compared to pure CdTe. This indicates that selenium incorporation may influence the nucleation and growth process during deposition, leading to a smoother and more compact surface morphology. At higher selenium content (x = 1) corresponding to the CdSe film, the surface remains continuous and uniform without any noticeable microcracks or pinholes. The film exhibits a fine-grained structure with improved surface smoothness, which can be attributed to the enhanced crystallinity and better lattice matching associated with Se-rich compositions. Overall, the SEM micrographs confirm that all the investigated films possess dense, uniform, and crack-free surfaces, which indicate good adhesion between the film and substrate. The gradual change in the grain size and texture with increasing the selenium concentration reflects the compositional dependence of the surface morphology in the CdTe–CdSe alloy system. Also, the absence of surface defects or structural discontinuities suggests that the deposition conditions were well optimized, resulting in high-quality ternary alloy films suitable for optoelectronic applications.

Close

Fig. 2.

SEM image of CdTe, CdTe_0.4Se_0.6, and CdSe films.

Export to PPT

3.2. Optical characterization of CdTe₁₋ₓSeₓ thin films

The optical properties of the deposited CdTe₁₋ₓSeₓ thin films are systematically examined through measurements of their spectral transmittance (T) and reflectance (R) over a broad wavelength range extending from 400 to 2500 nm, as presented in Fig. 3. The transmittance spectra exhibit distinct and regularly spaced interference fringes across the transparent and weakly absorbing regions. The appearance of such well-defined fringes confirms that the deposited films possess a high degree of thickness uniformity, surface smoothness, and structural homogeneity. The observed interference pattern corresponds to fringes of equal chromatic order (FECO), which arises from multiple internal reflections within films of nearly constant thickness. The persistence and regularity of these fringes provide strong evidence that the films are optically smooth, continuous, and free from significant thickness fluctuations or surface roughness. To extract the fundamental optical constants of the CdTe₁₋ₓSeₓ thin films, namely the refractive index (n) and the film thickness (d), the Swanepoel envelope method is applied (Soraya 2020). This method is widely recognized for its reliability in determining optical parameters in regions of weak to intermediate absorption. Within this framework, the refractive index can be estimated using the following relation (Soraya 2020):

Close

Fig. 3.

The deposit films’ transmittance (T) and reflection (R) spectra of CdTe_1-xSe_x films as a function of wavelength.

Export to PPT

where T_M and T_m correspond to the transmission maxima and minima at a given wavelength, respectively, and (s) represents the refractive index of the substrate (as illustrated in Fig. 4), which is calculated using the appropriate standard expression for transparent substrates.

Close

Fig. 4.

Illustration of the Swanepoel method using the spectra of T and R for CdTe, showing the maxima T_M and minima T_m, according to the text.

Export to PPT

Once the refractive index values are obtained at successive extrema, the film thickness (d) is determined using the interference condition given by:

where λ₁ and λ₂ denote the wavelengths of two successive transmission maxima (or minima), and n₁ and n₂ are the corresponding refractive indices. This procedure involves constructing the upper and lower envelope curves of the interference fringes and allows for a precise evaluation of both the refractive index dispersion and the film thickness (Soraya 2020). To increase the accuracy, the thickness is calculated using several pairs of adjacent transmission maxima in the transparent region of the spectrum. The final thickness value is obtained by averaging the individual measurements, and the associated standard deviation (σ) is used to assess the uncertainty. The overall experimental error was found to be less than 1.7%, confirming the strength of the adopted method. The average thickness values for the CdTe and CdSe thin films are summarized in Table 1. They are seen to be about 750 nm, which indicates a consistent deposition process and good reproducibility across different compositions.

To complement these results, SE is employed to verify the optical parameters and thickness obtained from the transmittance data. Ellipsometry measures the change in polarization of light upon reflection from the thin film in terms of the parameters ψ (amplitude ratio) and Δ (phase difference). Using the WVASE32 software, the optical response of the films is modeled through a three-layer structure comprising a Cauchy substrate layer, a B-spline layer representing the film itself, and an additional surface roughness layer. This model provides a close fit between the experimental and simulated data, yielding accurate thickness measurements. As shown in Fig. 5, the thicknesses of CdTe and CdSe films are determined to be approximately 757 nm and 761 nm, demonstrating excellent uniformity and consistent deposition conditions across the different compositions.

Close

Fig. 5.

Experimental and modeled ellipsometric optical parameters (ψ and Δ) for calculating the film thickness of (a) CdTe and (b) CdSe thin films.

Export to PPT

The fitting envelops of T_M and T_m can be divided into small intervals to calculate the refractive index that can be interpolated and extrapolated by a two-term Cauchy dispersion relation to give the dispersion of the refractive index n as (Soraya 2020):

where A and B are material-dependent constants obtained through curve fitting. The variation of n(λ) for different compositions (0 ≤ x ≤ 1) is shown in Fig. 6. It can be clearly seen that the refractive index increases progressively as the selenium concentration rises from x = 0, 0.2, and 0.4, reaching a maximum at x = 0.4. This behavior suggests denser microstructure and enhanced packing density in the mixed compositions, which in turn modifies the optical polarizability. Beyond x = 0.6, however, the refractive index begins to decrease, likely due to structural disorder or the presence of secondary phases, which reduces the optical density of films. These observations are consistent with the variations observed in the absorption edge across different compositions.

Close

Fig. 6.

Variation of the refractive index n as a function of wavelength λ for different CdTe_1-xSe_x (0 ≤ x ≤ 1) thin films.

Export to PPT

The absorption coefficient (α), which quantifies the extent of light absorption within the film, is determined from the measured transmittance and reflectance spectra using the following relations (Soraya 2020):

In regions of strong absorption, the optical band gap ($E_g^{opt}$) can be obtained through Tauc’s relation (Makuła et al. 2018) applicable to direct allowed transitions:

where hn is the photon energy, K is a material-dependent constant related to the probability of optical transitions, and $E_g^{opt}$is the optical band gap. Plots of (αhν)² versus hν for different compositions are presented in Fig. 7. The linear portion of each curve is extrapolated to the photon energy axis to determine the band gap value as the intercept at (αhν)² = 0. The variation of the optical band gap with selenium content is presented in Fig. 8, which shows a clear compositional dependence across the CdTe₁₋ₓSeₓ alloy system. As the Se concentration increases from x = 0 to x = 0.4, the optical band gap $E_g^{opt}$ exhibits a gradual reduction from 1.51 eV to 1.43 eV. This decline can be attributed to several interrelated structural and electronic factors. When selenium atoms substitute tellurium atoms within the CdTe lattice, the smaller ionic radius and higher electronegativity of Se induce local lattice distortions and strain fields. These structural modifications enhance the overlap between the atomic orbitals of Cd and Se, thereby altering the band structure and resulting in narrowing the energy separation between the conduction and valence bands. The reduction in $E_g^{opt}$within the mixed composition region (0 < x < 0.4) can also be linked to the formation of localized states near the band edges due to alloy disorder. Such disorder-related states lead to band tailing, which effectively reduces the optical band gap obtained from Tauc plots. This observation is consistent with Vegard’s law deviation often observed in II–VI semiconductor alloys (Jacob et al, 2013), where nonlinear variations in $E_g^{opt}$arise from differences in atomic bonding and lattice mismatch between the end compounds CdTe and CdSe. Beyond x = 0.6, a reversal trend is observed as the optical band gap begins to increase, reaching 1.77 eV for the CdSe-rich composition (x = 1.0). This widening of $E_g^{opt}$can be understood by considering the intrinsic electronic configuration of CdSe, which possesses a naturally larger band gap compared to CdTe. As the alloy composition becomes Se-rich, the electronic structure gradually makes transitions toward that of CdSe with stronger Cd–Se bonds and reduced lattice constants, which leads to higher energy transitions. The increase in $E_g^{opt}$at higher Se concentrations may also indicate improved crystallinity and reduced defect density, as the selenium incorporation beyond a certain limit can promote better atomic ordering and minimize tellurium-related vacancies. Overall, this nonlinear dependence of the optical band gap on selenium content reflects the band bowing effect, which is a characteristic feature of semiconductor solid solutions. The bowing parameter in such systems arises from the difference in electronegativity, atomic size, and bond energy between the constituent anions (Te and Se). The observed variation in $E_g^{opt}$that the alloy composition can be precisely tuned to achieve specific optical properties is suitable for photovoltaic applications. In particular, the composition range x = 0.2–0.4, corresponding to a band gap near 1.43 eV, lies close to the optimal value predicted by the Shockley–Queisser limit for single-junction solar cells (Shockley and Queisser 1961). This makes the CdTe₁₋ₓSeₓ films in this range promising as absorber layers, combining favorable band alignment with high absorption efficiency in the visible region.

Close

Fig. 7.

(αhν)² versus photon energy (hν) for the investigated thin films.

Export to PPT

Close

Fig. 8.

Variation of refractive index and energy gap as a function of Se doping concentration for CdSe_xTe_1-x thin films.

Export to PPT

3.3 Electrical characterization of CdTe₁₋ₓSeₓ thin films

To ascertain the electrical characteristics of the grown CdTe₁₋ₓSeₓ thin films as functions of the Se content, Hall effect measurements are carried out utilizing the Van der Pauw configuration. Sheet resistance, electrical resistivity, electrical conductivity (Fig. 9), Hall coefficient, carrier concentration, and Hall mobility (Fig. 10) are among the metrics under investigation. The sheet resistance (Rₛ) is calculated using the relation Rₛ = 4.53 (V/I) [Ω/sq], where V is the applied voltage (V), I is the current (A), and 4.53 is a geometrical correction factor. As illustrated in Fig. 9, the value of Rₛ decreases progressively with increasing the selenium concentration, attaining a minimum of 93.2 Ω/sq at x = 0.4, followed by an increase at higher Se contents. The electrical resistivity (ρ) is determined from the equation Rₛ = ρ/d, where d represents the film thickness (0.75 μm). The electrical conductivity (σ) is obtained as the reciprocal of resistivity, σ = 1/ρ. The carrier concentration (nₙ) is derived using nₙ = 1/(eR_H), where R_H is the Hall coefficient, and e denotes the electronic charge. The carrier mobility (μ) is evaluated from the relation μ = σ/(nₙe) or equivalently μ = R_H σ. Fig. 9 shows also shows that as Se incorporation increases, ρ decreases, reaching a minimum of 6.99 ×10² Ω·cm at x = 0.4, while σ increases correspondingly to a maximum value of 1.43 × 10^-5 (Ω·cm)⁻¹. Beyond this composition, both parameters exhibit reversed trends. Fig. 10, which depicts variation of the carrier concentration (nₙ) and mobility (μ) on Se composition, indicates that both parameters increase with selenium addition up to x = 0.4, attaining maximum values of 14.501 × 10¹⁸ cm⁻³ and 35.24 cm²·V⁻¹·s⁻¹, respectively, followed by a decline at higher levels of Se. These results agree with the findings in refs. (Xu et al, 2014. Lee et al. 2015, Patil et al. 2003). The optimal selenium doping level in CdTe₁₋ₓSeₓ thin films is therefore identified as x = 0.4, where the material exhibits low resistivity, high conductivity, and enhanced carrier transport characteristics. Consequently, the CdTe₀.₆Se₀.₄ composition is considered most suitable for photovoltaic absorber layer applications.

Close

Fig. 9.

Variation of resistivity (ρ) and conductivity (σ) as a function of Se doping concentration for CdSe_xTe_1-x thin films.

Export to PPT

Close

Fig. 10.

Variation of carrier concentration n_c and mobility (µ) as a function of Se doping concentration for CdSe_xTe_1-x thin films.

Export to PPT

To further evaluate the film’s suitability for solar cell use, the quality of the CdTe₀.₆Se₀.₄ thin film is assessed using the figure of merit (Φ), defined as (Haacke 1976, Ocampo et al. 1995):

where T is the average optical transmittance, and Rₛ is the sheet resistance. The variation of Φ with Se content is plotted in Fig. 11. Consistent with the Hall effect data, Fig. 11 shows that the film doped with 0.4% Se exhibits the highest figure of merit of 2.741 × 10⁻⁵ Ω⁻¹, which confirms the superior electrical conductivity and overall performance potential of for CdTe₀.₆Se₀.₄.

Close

Fig. 11.

Figure of merit as a function of Se doping concentration for CdSe_xTe_1-x thin films.

Export to PPT

3.4. Device characterization of CdTe₁₋ₓSeₓ/CdTe solar cells

To assess the photovoltaic characteristics of the fabricated heterojunctions, two device architectures are investigated: ITO/n-CdS/p-CdTe and ITO/n-CdS/p-CdTe₀.₆Se₀.₄. As shown in Figs. 12(a,b), the CdTe and CdTe₀.₆Se₀.₄ absorber layers possess comparable thicknesses of approximately 750 nm, thereby minimizing thickness-related variations in device performance. The CdS window layer thickness is kept constant for all devices to ensure consistency in the comparative analysis. The dark current–voltage (I–V) characteristics presented in Fig. 13 exhibit a rectifying behavior for both heterojunctions, indicating the formation of functional p–n junctions. Compared to the CdTe-based device, the CdTe₀.₆Se₀.₄ junction shows a higher forward current density, suggesting improved carrier transport under forward bias. This behavior may be associated with reduced bulk recombination and modified defect distributions resulting from moderate selenium incorporation. At higher selenium contents, a reduction in current density is observed, which may reflect an increased disorder or defect-assisted recombination, which agrees with previous studies (Bowman et al. 2024). To further assess the device performance under illumination, Figs. 14(a,b) plot the I–V characteristics of the solar cells under simulated sunlight (AM1.5G). The results reveal that under forward bias, the photocurrent significantly exceeds that in the reverse bias. A noticeable increase in the current density occurs at low voltage regions, confirming efficient charge separation and collection under illumination. Conversely, the reverse current exhibits a lower exponential response in the depletion region, suggesting minimal leakage and strong rectification behavior as indication of a well-formed p–n junction with reduced recombination at the interface.

Close

Fig. 12.

Schematic diagrams of the structure of (a) n-CdS/p-CdTe and (b) n-CdS/p-CdTe_0.6Se_0.4 solar cell. In this way, the device structure is a simple n-p heterojunction with an ohmic contact at the p-CdTe-Se/metal interface.

Export to PPT

Close

Fig. 13.

The dark (I-V) characteristics for the fabricated solar cell at different CdSe_xTe_1-x thin films.

Export to PPT

Close

Fig. 14.

I-V characteristics under standard solar simulated illumination (AM1.5G, 100 mW/cm²) for the p-n junction solar cell in the range of (-2, 2) volts of 1 micron of CdSe_xTe_1-x 1thin films in the presence of light (a) reverse bias and (b) forward bias.

Export to PPT

The photovoltaic performance of the developed solar cells is characterized by several key parameters: the open-circuit voltage (V_oc), short-circuit current density (J_sc), fill factor (FF), and power conversion efficiency (PCE). The open-circuit voltage (V_oc) represents the potential difference across the terminals of the solar cell when no external load is connected, while the short-circuit current density (J_sc) is the current per unit area generated when the voltage across the device is zero. The PCE, a vital figure of merit for solar cells, quantifies the ratio of the maximum electrical output power (P_max) to the incident optical power (Pin) (Ezzeldien et al. 2022).

Similarly, the FF provides a measure of the cell’s quality and is defined as the ratio of the maximum obtainable power (P_max) to the product of V_oc and J_sc (Ezzeldien et al. 2022),

Fig. 15(a) Illuminated current density–voltage (J–V) characteristics measured under standard AM1.5 solar illumination for two solar cell configurations: CdS/CdTe (black points) and CdS/CdTe₀.₆Se₀.₄ (red points). The curves illustrate the photovoltaic response and allow comparison of the electrical performance of the two absorber compositions. Fig. 15(b) Corresponding power–voltage (P–V) characteristics derived from the J–V measurements, showing the output power as a function of the applied voltage for CdS/CdTe (black points) and CdS/CdTe₀.₆Se₀.₄ (red points) solar cells, where the maximum power point for each device can be identified. Under standard AM1.5G illumination (100 mW/cm²) within the voltage range of –2 to +2, the n-CdS/p-CdTe device exhibits J_sc = 27.5 mA/cm², V_oc = 0.822 V, and FF = 63.9%, whereas the n-CdS/p-CdTe₀.₆Se₀.₄ device shows a higher J_sc of 31.5 mA/cm² and an improved FF of 69.8%, accompanied by a slight reduction in V_oc to 0.809 V. The enhancement in J_sc is attributed to improved carrier generation and collection in the Se-alloyed absorber, likely associated with bandgap narrowing and reduced bulk recombination. The increase in FF indicates more efficient charge extraction and lower resistive losses, which is consistent with the reduced leakage current and stronger rectification behavior observed in the dark J–V characteristics. In contrast, the modest decrease in V_oc is interpreted as a consequence of enhanced recombination, potentially originating from interface-related effects at the CdS/absorber junction or from band tail states introduced by selenium incorporation. A qualitative analysis of the dominant loss mechanisms, including bulk recombination, interface recombination, and resistive effects, has therefore been incorporated based on both illuminated and dark J–V measurements. Despite the slight reduction in V_oc, the concurrent gains in J_sc and FF compensate for this loss, resulting in an overall improvement in power conversion efficiency from 17.62% for CdS/CdTe to 20.25% for CdS/CdTe_0.6Se_0.4. These results indicate that moderate selenium incorporation effectively balances optical absorption, carrier transport, and recombination processes, leading to enhanced device performance.

Close

Fig. 15.

(a) Illuminated J-V curves obtained under AM1.5 conditions for: a) CdS/CdTe (black points) and b) CdS/CdTe_0.6Se_0.4 (red points) solar Cells, respectively. (b) The power points of a) CdS/CdTe (black points) and b) CdS/CdTe_0.6Se_0.4 (red points) solar Cells, respectively, against the voltage.

Export to PPT

It is worth highlighting that the present work provides a systematic assessment of CdTe₁₋ₓSeₓ absorbers that clarifies the intrinsic role of Se alloying while introducing a simplified and reproducible fabrication strategy. All films were deposited with a nearly constant thickness (∼0.75 μm) across the full composition range of Se (0 ≤ x ≤ 1), enabling comparison of structural, optical, and electrical properties without the confounding influence of thickness variations commonly present in earlier studies. This fixed-thickness approach allows the composition-dependent trends to be isolated more reliably and facilitates the identification of an optimal alloy composition. The absorbers were fabricated using mechanical alloying of CdTe and CdSe powders followed by single-source thermal evaporation, which offers a distinct alternative to conventional co-evaporation, CSS, or multi-step deposition routes. This method ensures compositional homogeneity, improves process reproducibility, and reduces system complexity, highlighting its potential scalability and practical relevance. Importantly, the composition identified as optimal at level (x ≈ 0.4) of Se, based on crystallinity, bandgap tuning, and charge transport parameters, translates into enhanced solar-cell performance. These gains are achieved without advanced post-deposition treatments, intentional Se grading, or complex band-engineering schemes. The observed improvements arise from improved crystal quality, a favorable bandgap for photon absorption, and enhanced electrical transport, as independently confirmed by Hall measurements.