Improved Performance of Bifacial Photovoltaic Modules with Low-Temperature Processed Textured Rear Reflector

Abstract

Bifacial photovoltaic (PV) modules can capture both front and rear incident light simultaneously, thereby enhancing their power output. Achieving uniformity in rear incident light is crucial for an efficient and a stable operation. In this study, we present a simple, yet effective textured rear reflector, designed to optimize the performance and stability of bifacial PV modules. The three-dimensional textured surface was created using an ethylene vinyl acetate sheet (EVA) through a hot-press method at 150 °C. Subsequently, the textured EVA surface was coated with solution-processed silver ink, increasing the reflectance of the textured reflector through a low-temperature process. The integration of the developed textured rear reflector into bifacial crystalline silicon (c-Si) PV modules resulted in an additional 6.9% improvement in power conversion efficiency compared to bifacial PV modules without a rear reflector, particularly when the rear reflector is close to the PV module. Furthermore, the textured rear reflector may mitigate current mismatch among cells by randomizing incident light and uniformly redistributing the reflected light to the PV cells. Consequently, the proposed textured reflector contributes to the enhanced performance and stability of bifacial PV modules.

1. Introduction

Bifacial photovoltaic modules (bPVs) efficiently capture incident light from both front and rear sides, thereby enhancing their power output [1,2]. Scattered light and photons penetrating bPVs can rebound off rear reflectors and be harvested at the rear side of bPVs. The advantages of bPVs are particularly evident in environments with reflective surfaces, such as snow-covered ground, water bodies, or light-colored surfaces [3,4]. The albedo of the rear surface plays an important role in determining an additional energy yield when integrating bPVs. Empirical field tests confirmed that bPVs with highly reflective rear surfaces demonstrate approximately 20% improvement in power output compared to that of mono facial modules, verifying bPVs as a predominant technology in the current photovoltaic (PV) market [5,6,7]. These tests also show that additional power can be generated through bPVs installed on rooftops, walls, agrivoltaic systems, and carports [7,8,9,10,11]. This increased power generation has been consistently observed across various types of bPVs, regardless of cell type. Multiple types of bifacial solar cells, including those using crystalline silicon (c-Si), dye, quantum dots, perovskite, chalcogenide, and organic materials, have been developed [12,13,14,15,16]. Notably, bPVs optimize power output within confined spaces and exhibit reduced sensitivity to installation angles, allowing for substantial energy generation even in the case of non-south-facing modules [17,18,19]. As a result, bPVs emerge as compelling candidates for building-integrated photovoltaics (BIPVs), particularly in situations where non-south-facing PV modules must be installed in constrained areas [20,21,22,23].

To optimize the power output of bPVs, it is desirable to ensure uniformly distributed rear incident light with strong intensity. Uniformly distributed rear incident light helps suppress losses from current mismatching among cells, while the rear surface shadow and shading generally result in power loss in bPVs [10,11,24,25]. For example, the bifacial gain of field-installed bPVs is limited to below 4% due to inhomogeneous rear reflection, which is significantly lower than what is achievable under ideal conditions [10,11,25]. The intensity of rear incident light on bPVs is typically influenced by factors such as the albedo of the rear reflector, the distance between the module and the rear reflector, the angle of rear incident light, the installation angle of bifacial PV, and other variables [3,19,23,26]. Thus, the rear surface and installation conditions should be taken into careful consideration during the design of PV power plants with bPVs.

In field-installed bPVs, negligible current mismatch induced by non-uniform rear irradiation commonly takes place due to the elongated distance between bifacial cells and a rear surface (D_RR). Self-shading by PV cells and non-uniform reflectance from a rear surface can be compensated during the extended path to reach the rear side of the bPVs [5,27]. As the fraction of reflected light to reach the rear side of bPVs increases proportionally with the D_RR, theoretical simulations and experimental results indicate that a 1 m of D_RR is high enough to minimize the deviation of rear incident light among bifacial cells in bPVs. However, a short distance between the rear surface and bPVs gives rise to non-uniform rear reflection issues when integrated into BIPV systems. Non-uniform rear reflection is an inherent challenge in bPVs equipped with a nearby specular surface [23]. The adverse effects of non-uniform rear incident light result in power loss and local heating of PV cells. Bifacial PV cells with non-uniform rear incident light may restrict the operating current, leading to energy generation losses and potential safety hazards through cell heating [10,24,28]. Particularly, current mismatch among cells is expected to occur more frequently in a BIPV with bPVs due to a shorter D_RR. Yusufoglu et al. observed that the deviation of rear incident light exceeds 200 W/m² in a bPV with 10 cm of D_RR under standard test conditions [5]. Simulation studies also indicate that the non-uniformity of rear incident light exceeds 50% when the D_RR is less than 50 cm [25]. Additionally, non-uniform rear incident light can cause localized heating of bifacial cells, raising their temperature by more than 15 °C [24]. Although the bypass diodes are applied in bPVs to maximize the power output during periods of non-uniform rear illumination, the continuous current flow through the bypass diode might cause leakage in currents and pose fire issues due to a massive reverse current [24,29,30,31]. Most BIPVs replacing roofs, curtain walls, and spandrels with solar panels are normally installed close to the envelope of buildings, where the D_RR is small (less than 10 cm) [24,32]. Moreover, controlling the albedo and D_RR of the rear surface is nearly impossible, increasing the likelihood of unevenly distributed rear incident light [23]. Consequently, BIPVs incorporating bPVs may experience uneven rear incident light, leading to a diminished energy generation and to an increase in reliability issues.

To address issues arising from non-uniform rear incident lights in BIPVs with bPVs, a rear reflector has been introduced. The double glass-structured BIPV comprising of bPVs and a rear reflector enhances energy generation and minimizes heat dissipation from buildings, significantly improving the energy efficiency of buildings with them. Furthermore, selective and Bragg reflectors provide a BIPV with bPVs with the functionality of a glass window and recycle unabsorbed photons [33,34,35]. These reflectors redirect some of the photons that strike the gaps between cells to the rear side of the bPVs, thereby increasing their efficiency. However, selective reflectors typically consist of multi-layer stacks of precisely thickness-controlled dielectric and metal layers, requiring vacuum-based deposition processes. The high production costs and difficulties in scaling up these reflectors hinder their widespread application in bPVs. Furthermore, selective reflectors are not particularly effective at altering the trajectory of photons, making them inefficient at transferring photons that impinge on the gaps between cells. Given these limitations, reflective metal film or conventional polymer-based backsheets were widely applied in bPVs as rear reflectors [35]. Through a stack of polymer films containing scattering micro-particles and asymmetrical structure, Lambertian reflectors randomize the trajectory of reflected light and distribute photons almost evenly to the rear side of bPVs [36,37,38]. Since the reflectance of polymer-based film is nearly unity, most photons reaching the polymer-based reflector are efficiently reflected [39,40]. Additionally, solution-processed micro-particles selectively casted on the gap between cells can redirect incident light to the rear side of bPVs. Through the solution and the printing-based process, the polymer film-based reflector can be produced at a low cost. However, the poor stability of the polymer-based Lambertian reflector against humidity and UV radiation limits its implementation as a rear reflector of bPVs for BIPV. For instance, it has been reported that polymer-based reflectors tend to be yellowish after a long-term exposure to UV [23].

Conversely, metal-based reflectors typically withstand harsh outdoor conditions despite their high cost and the requirement for high-temperature processing. Nevertheless, the high specular reflection of a metallic reflector is not effective in transferring photons impinging on its surface to the rear side of bPVs. Theoretical calculations have shown that energy generation in bPVs decreases as the specularity of the rear surface increases [41,42,43]. Addressing the issue of strong angular dependence in metallic reflectors is a significant challenge, and this problem becomes more pronounced in bPVs with a short D_RR, particularly in BIPV applications. Although textured metallic surfaces randomize the direction of reflected light similar to a Lambertian reflector, the dry process for patterning microstructures and depositing metallic layers over large areas are highly energy-intensive and involve complicated procedures [44,45,46]. Additionally, wet etching processes for creating microstructures inevitably generate toxic chemical waste, making them unsuitable for large-scale bPVs. While high-precision patterning techniques for metallic layers widely accepted in the semiconductor industry offer accuracy and resolution, they are difficult to implement in bPVs due to high energy consumption, waste management challenges, scalability issues, and cost concerns. Regarding the pros and cons of the previously suggested rear reflectors for BPVs, a low-cost, textured metal reflector would be beneficial in increasing the energy generation of BIPVs consisting of bPVs.

In this study, we demonstrate a simple, yet effective textured metallic reflector for bPVs to unlock their full potential. The textured reflector is comprised of printed silver ink on an embossed ethylene vinyl acetate (EVA) film, fabricated through the hot-press method. The maximum process temperature required for the textured reflector is below 150 °C, allowing for energy savings in its fabrication. Both experimental results and theoretical analyses indicate that the textured reflector contributes to the improved performance of bPVs by randomizing the direction of rear reflected light. We believe that the proposed textured reflector in this work has the potential to contribute to the design of highly efficient bPVs for BIPV applications.

2. Experiments and Methodology

2.1. Fabrication of bPV with Textured Reflector

The fabrication process of the textured reflector is illustrated in Figure 1a. The textured surface was made by an Ethylene vinyl acetate (EVA) sheet positioned on a clean glass surface. EVA is commonly employed as an encapsulant for PV modules due to its robust mechanical adhesion to glass and high resistance to moisture, elevated temperatures, and UV light. As a result, the production of a textured surface with EVA can be obtained at the current photovoltaic module production line without a significant change. Following the removal of any dust particles, the EVA film underwent annealing at 150 °C for 10 min at a low vacuum. Subsequently, the molten EVA was pressed by a metal mesh (thickness of 1 mm with a 10 × 10 mm pattern) under high pressure (760 mm Hg) for 5 min. Throughout the hot-press process, the part of EVA film pressed by the metal mesh descended, causing the residual EVA to migrate towards the center of the mesh where pressure was not applied. This resulted in the center of the mesh protruding and forming a textured surface with a height difference of 1.5 mm between the peak and valley, as depicted in Figure 1a. Here, we used EVA for the textured surface, but we believe that the textured surface can be obtained using various stable thermoplastics. After cooling the EVA film at room temperature for 1 h, the truncated cone-shaped EVA was covered by a printable silver ink (TEC IJ-060 from InkTec, Ansan, Republic of Korea) [47]. The film was then annealed at 150 °C for 30 min to dry the solvent of silver ink. The thickness of the silver layer was ~100 μm. The key advantage of this proposed process lies in the ability to generate mechanically and chemically stable textured reflectors through a low-temperature process. Due to the elimination of complicated photolithography and the high-temperature vacuum processes, the energy consumption for demonstrating the suggested textured reflectors is significantly reduced. Another notable advantage of this approach is the facile modification of the textured reflector’s dimension by controlling a dimension of mold for hot pressing. While a metal mesh was used in this experiment, we anticipate that a convex-shaped mold could enable the demonstration of a hemispheric reflector [48]. Lastly, the reflective silver ink can be selectively deposited where the reflector is needed, thereby substantially minimizing silver ink wastage and rendering unintended light blocking from the rear reflector to the rear incident light negligible.

Finally, the textured reflector was integrated with a 50 × 50 mm c-Si bPV module (laser-cut from a cell with a p-passivated emitter rear contact structure from Jinko Solar, Shanghai, China), whose bifaciality was 0.7. The bPV module structure was glass (3 mm)/EVA (~200 μm)/c-Si bPV cell (150 μm)/EVA (~200 μm)/Glass (3 μm), laminated under 150 °C for 10 min with a pressure of 760 mm Hg. To comprehensively evaluate the impact of both flat and textured reflectors on the performance of the bPVs, a flat reflector was also fabricated. The flat reflector, composed of solution-processed silver ink printed on a flat EVA without hot pressing, was fabricated using identical processing conditions to the textured reflector. Reflectance measurements for both reflectors conducted under an integrating sphere to capture the sum of diffusive and specular reflections (Please see Figure 1b) revealed that the reflectance of both reflectors exceeded 90%. In contrast, the specular reflectance of the proposed reflector (~10%) markedly differed from that of the flat reflector (~80%). The specular measurement of the reflector was monitored using a setup with a detector evaluating reflected light whose incident angle is less than 5 degrees.

2.2. Measurement of bPVs

The current-voltage (I–V) characteristics of bPVs with flat and textured reflectors were monitored through a source meter (Keithley 2420, Tektronix, Beaverton, OR, USA) under AM 1.5 G conditions (intensity of 1000 W/m²) using a solar simulator (K201, McScince, Suwon, Republic of Korea) equipped with an Xe lamp and a collimated lens. For quantifying the improvement ratio, the measurement was systematically conducted at varying D_RR and gaps among cells (D_Gap), as shown in Figure 1c. Here, the textured reflectors were incorporated with bPVs, where the edge of bPV cell was aligned with the edge of a periodic textured structure. Notably, the BIPV case with a large D_Gap, designed for enhanced transmittance, was also taken into account to evaluate the impact of the textured reflector on the bPVs. The size of bifacial c-Si PV cell (A_Cell) was fixed with 2500 mm², while D_Gap was controlled by a black mask (in the range of 5–12.5 mm). Thus, the ratio between the area of the gap (A_Gap) and the entire bPV area ranged from 17.3 (D_Gap of 5 mm) to 36% (D_Gap of 12.5 mm), as summarized in

Table 1. The configuration of bPVs with various D_Gap.

DGap (mm)	5	7.5	10	12.5
Configuration
ACell (mm2)	2500	2500	2500	2500
AGap (mm2)	525	806	1100	1406
AGap/ACell+AGap (%)	17.3%	24.3%	30.5%	36%

, which is similar to the general BIPV design [21].

Moreover, the performance of the bPV was evaluated under various D_RR ranging from 10 to 30 mm. As the ventilation through the D_RR normally contributes to the decreasing temperature of the bPV during the daytime, such a large D_RR is commonly adopted for BIPV installed at rooftops and curtain walls. Additionally, a large D_RR guarantees easy access to rear reflectors and bPVs, which is essential for maintaining and operating the BIPV. The measurement was repeatedly conducted 3 times at each condition in a dark room to avoid errors during the measurement.

3. Results and Discussion

Figure 2 illustrates the I–V characteristics of a bifacial PV module depicted with and without a reflector, considering varying distances between the cell and the reflector (D_RR: 10–30 mm) and cell gap widths (D_Gap: 5–12.5 mm). The short-circuit current (I_SC) of bPVs is normally the sum of the current generated from the front (I_front) and the rear (I_rear) incident light, as follows:

$${\mathrm{I}}_{\mathrm{S}\mathrm{C}}={\mathrm{I}}_{\mathrm{f}\mathrm{r}\mathrm{o}\mathrm{n}\mathrm{t}}+{\mathrm{I}}_{\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{r}}=\int {\mathrm{A}}_{\mathrm{C}\mathrm{e}\mathrm{l}\mathrm{l}}\times {\mathsf{\varphi }}_{\mathrm{F}\mathrm{r}\mathrm{o}\mathrm{n}\mathrm{t}}\left(\mathsf{\lambda }\right)\times {\mathrm{E}\mathrm{Q}\mathrm{E}}_{\mathrm{F}\mathrm{r}\mathrm{o}\mathrm{n}\mathrm{t}}\left(\mathsf{\lambda }\right)\mathrm{d}\mathsf{\lambda }+\int \left(1-{\mathsf{\eta }}_{\mathrm{s}\mathrm{h}\mathrm{a}\mathrm{d}\mathrm{i}\mathrm{n}\mathrm{g}}\right)\times {\mathrm{A}}_{\mathrm{G}\mathrm{a}\mathrm{p}}{\times \mathsf{\varphi }}_{\mathrm{G}\mathrm{a}\mathrm{p}}\left(\mathsf{\lambda }\right)\times \mathrm{R}(\mathsf{\lambda }){\times {\mathsf{\eta }}_{\mathrm{O}\mathrm{p}\mathrm{t}}\times \mathrm{E}\mathrm{Q}\mathrm{E}}_{\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{r}}\left(\mathsf{\lambda }\right)\mathrm{d}\mathsf{\lambda }$$

(1)

where the Φ_Front(λ) and Φ_Gap(λ) represent the photon flux irradiating the front of the bifacial photovoltaic cell and the gap between cells, respectively, and are equal under a solar simulator. EQE_Front(λ) and EQE_Rear(λ) denote the external quantum efficiency of the bPV when illuminated from the front and the rear, respectively. η_shading is the loss ratio due to self-shading from the cell and mask. R(λ) is the reflectance of the rear reflector, and η_Opt is the efficiency of the reflected light that reaches the rear surface of the bPV [6,49]. In this study, we assume that the EQE_Rear(λ) is 0.7 times of EQE_Front(λ) considering the bifaciality of the cell. In addition, the R(λ) is set to 0.8 for all cases based on the reflectance spectrum of the reflectors used. While the I_front remains constant regardless of a rear reflector, the I_rear varies depending on the module configuration (A_Gap: 17–36%) and rear reflector conditions (η_shading and η_Opt). We set the η_shading as 0.2 (D_Gap = 10 mm and D_RR = 30 mm)–0.8 (D_Gap = 5 mm and D_RR = 10 mm), considering the dimension of bPVs used in the experiment.

In the case of a bPV with its rear surface covered by black paper (None), consistent short-circuit current (I_SC) and power conversion efficiency (PCE) values were observed, irrespective of changes in D_RR and D_Gap. In this configuration, the I_rear is nearly negligible due to the very low R(λ), leaving I_SC primarily governing the I_front. Thus, the I_SC and PCE of the none group ranged from 0.97 to 0.99 A and 17.6 to 18.2%, respectively, in all scenarios, indicating that the I_front of the bPV is approximately 0.98 A (current density of 39.2 mA/cm²). The limited rear incident light resulting from the highest absorption of the incident light by the black surface contributes to these similar values. Here, the small deviations in PCE and I_SC observed in the none group might be attributed to the imperfect black surface, which does not absorb the entire incident light.

Conversely, the implementation of a rear reflector improves the I_SC and PCE of bPVs. Photons passing through the gap between cells bounce off the rear reflector, providing additional opportunities to harvest in the rear side of bPVs. The improved I_rear contributes to improvement in I_SC and PCE of the bPV with the proposed textured reflector (Textured). The textured reflector exhibits a more effective transfer of transmitted photons to the cells compared to the flat reflector with increased η_Opt in all cases. Although the I_SC enhancement ratio introduction of the textured reflector to the bPV varies depending on the D_RR and D_Gap, the textured reflector consistently outperforms the one with the flat reflector (Flat). For example, under conditions where both D_RR and D_Gap are set at 10 mm, the I_SC and PCE of the flat reflector are 1.016 A and 18.8%, respectively. In the same conditions, those of the textured reflector increased to 1.034 A and 19.3%, as shown in Figure 2c. If the η_shading is around 40% due to off-axis incident light from the imperfect optical system of the solar simulator and bPV including experimental system, η_Opt values for the textured and flat reflectors, as calculated from Equation (1), are 0.53 and 0.35, respectively. This indicates that the textured reflector enables 1.5 times more light to be transferred to the rear side of the bPV.

These effects are more pronounced in the bPVs with larger D_Gap and longer D_RR. As the D_Gap increases, more light passes through the gap between cells through the enlarged A_Gap in the bPVs and reaches the reflector. The textured surface alters the trajectory of the transmitted light, increasing the chances of unabsorbed photons returning to the rear side of the bPV. The extended D_RR allows more transmitted light to be collected at the rear side of the bPV, broadening the possible reflection angle to reach the rear side of the bifacial cell. On the other hand, it is challenging to change the direction of reflective light in the flat reflector. Most normal incident photons impinging on the flat reflector escape from the module through the gap between cells without altering their trajectory. Although the specular reflective flat reflector redirects part of the incident light to the rear side of the bPVs, the ratio of redirected light to the bPV is lower than that of the textured reflection, resulting in reduced I_SC and PCE of the flat reflector compared to those of the textured reflector. The textured reflector shows better PCE and I_SC (19.6% and 1.060 A, respectively), when D_RR and D_Gap are at 30 and 10 mm, respectively. Assuming an η_shading of 30%, the calculated η_Opt for the textured reflector is 0.58. The improved η_Opt for the textured reflector reflects the enlarged subtended angle, which allows more reflected light to reach the rear side of the bPV. Under the same D_RR and D_Gap conditions, the flat reflector achieves a PCE of 19.2%, an I_SC of 1.033 A, and an η_Opt of 0.37. The textured reflector directs more photons to the rear side of the bPV, leading to a 7.6% improvement in I_SC, as shown in Figure 2d. This result indicates that the textured reflector enables unabsorbed light to be recycled by guiding it to the rear side of the bPVs.

However, the narrow D_Gap (<5 mm) limits the impact on the reflector due to the lack of passing photons. As illustrated in Figure 2a,b, the I_SC improvement ratio arising from incorporating textured and flat reflectors into the bPV with a D_Gap of 5 mm is less than 1.5%. We believe this reduced enhancement is primarily due to self-shading and lower incident light from the narrow A_Gap. Moreover, as most of the light rebounding to the rear side of the bPVs comes from near the cell regardless of reflector type, the impact of the proposed textured reflector on the bPV is not distinctive in the bPVs with small D_Gap [49,50,51,52]. This suggests that the proposed system is particularly beneficial for BIPV applications, where the larger D_Gap maximizes the reflector’s effect, while the impact of the rear reflector is diminished in field-installed PVs due to the reduced D_Gap. Thus, while the textured reflector significantly enhances the PCE of bPVs designed for BIPVs, its effect is constrained when the D_Gap is small.

To quantify the impact of the textured reflector on the performance of bPV for BIPVs, we analyzed its I–V characteristics under various D_RRs and D_Gaps in three cases: the None, the Flat, and the Textured reflectors. Figure 3 shows contour plots of the I_SC improvement ratio for the flat and the textured reflectors. Here, the I_SC enhancement ratio is derived by dividing the I_SC of a bPV with the reflector by the I_SC of a bPV without the reflector (the none group). For both flat and textured reflectors, no significant enhancement ratio in I_SC is observed when the D_RR and D_Gap are small. This is primarily attributed to a reduction in transmitted photons through the gap between the cells and the self-shading from the mask and the cells.

However, as the D_RRs and D_Gap increase, the enhancement ratio boosts in the cases of the flat and the textured reflectors. Notably, the textured reflector exhibits a higher enhancement ratio and reaches a saturation point at a higher value compared to the flat one. Specifically, the flat reflector demonstrates an I_SC improvement ratio of 4.8% where the D_RR and the D_Gap are 30 and 12.5 mm, respectively. Despite the enlarged A_Gap, the η_Opt decreases to 0.29. Furthermore, its enhancement ratio is saturated by approximately 5% despite the elongated D_RR and the enlarged D_Gap. No additional improvement was observed in the flat reflector when the D_RR was higher than 20 mm and the D_Gap was larger than 10 mm due to a trade-off relationship between A_Gap and η_Opt.

On the other hand, the I_SC enhancement ratio of the textured reflector continuously increases, reaching over 8.4% where the D_RR and D_Gap are 30 and 12.5 mm, respectively. The calculated η_Opt was 0.51, which is similar to that of the textured reflector with 10 mm of D_Gap. Additionally, employing the suggested textured reflector with D_RR = 20 mm and D_Gap = 10 mm achieves an enhancement ratio of 6%, which is 2.0% higher than that of the flat reflector. While an 8.4% improvement in the I_SC may seem modest, the cumulative additional energy over an expected lifetime of bPV (over 25 years) would sufficiently compensate for the additional cost of the reflector. To enhance the durability of the proposed reflector under harsh outdoor conditions for the lifetime of a BIPV, further development is required. This includes incorporating a protective layer over the silver layer and optimizing the processing temperature to improve the stability of EVA [47,53]. In this study, we evaluated the performance of bPVs incorporating a truncated cone-shaped reflector on a single size due to the limitations in the available mold for the reflector. Future enhancements in the performance of bPVs could be achieved by further modifications to the mold of the textured reflector. Smaller textured structures in the reflector design may lead to a more diffused reflection, which could randomize the direction of reflected light and enhance bPV performance [38]. Considering the wavelength of incident light, micro-sized reflective structures are more desirable, involving complicated patterning and high temperature processes. Despite these challenges, we believe that the proposed method offers a cost-effective and viable solution for bPVs.

The ray-tracing simulation results point out that the textured reflector facilitates more photons reaching the rear side of bPVs compared to the flat reflector, aligning well with the aforementioned empirical result. The increased number of reflected photons move towards the edges of bPVs cells, thereby serving as an origin of improved I_SC in the textured reflector. Figure 4 presents the rear incident electrical field intensity of the none, flat, and textured reflectors. Here, the calculation was conducted through a finite-difference, time-domain method simulation using commercially available software (Setfos 5.4, Fluxim Inc., Winterthur, Switzerland), where D_RR and the D_Gap are 10 and 10 mm, respectively. In our simulation, we assume that the textured reflector consists of truncated cones covered by silver, whose dimension is the same as the experimentally obtained structure. The simulation result reveals that reflected light mainly converges on the edge of the rear side of bPV cells because of a self-shading effect in all cases. Notably, the intensity of rear incident light varies depending on the shape and reflectance of the rear surface. Only a limited amount of reflected light reaches the edges of cells in the case of None, due to the low reflectance of the rear glass. As the reflectance of the rear surface increases, more light redirects toward the rear side of the cells. The shape of the reflector also influences the intensity and distribution of the rear incident light. Compared to the flat reflector, the textured reflector exhibits a more successful redirection of reflected light toward the edge of the rear side of the cell, marked as enlarged red and yellow regions at the cell’s edges (Please see Figure 4c). The calculated sum of electric field distribution in the rear side of the textured reflector is 4.5% higher than that of the flat reflector, which is slightly higher than the experimental data. We believe that the small discrepancy between the experimental and simulation results might be caused by the flatness of the rear reflector. Furthermore, the enhanced rear incident light is theoretically calculated at the center of bPVs in the case of the textured reflector, revealing that the suggested texture reflector is more effective in randomizing the trajectory of reflected light. These findings indicate that the textured reflector provides randomly distributed reflected light with high intensity to bPVs, offering potential benefits such as increased I_SC and reduced current mismatch among cells.

4. Conclusions

In conclusion, we propose a simple, yet effective textured rear reflector for bPVs. By employing a low-temperature, hot-press method and silver ink, we have developed textured reflectors without the need for high energy consumption or complex processes. The textured reflector successfully randomizes the trajectory of rear-reflected light, redirecting more reflected light towards the edges of the bPV’s rear side. As a result, bPVs equipped with textured reflectors show an 8.4% increase in current compared to bPVs without reflectors, outperforming flat specular reflectors, which offer only a 4.8% improvement. Our experimental and simulation results indicate that the effect of textured reflectors is particularly significant in bPVs with larger gaps, especially in the context of building-integrated photovoltaics (BIPVs). However, the proposed reflector is less suitable for field-installed bPVs, where reduced light irradiation and shading within the gap diminish its effectiveness.

We believe that further modifications to the size and shape of the textured structure, along with improvements in its durability, could make this design a strong candidate for enhancing BIPV performance. Moreover, if future studies explore selective patterning of the textured layer in the gap between cells, the cost-effective textured reflector could significantly contribute to the power generation of bPVs in BIPV applications.