Spectrally Complementary Tandem Luminescent Solar Concentrators: Optical, Electrical and Stability Assessment

1. Introduction

The increasing global demand for clean and sustainable energy has accelerated the search for alternatives capable of reducing greenhouse gas emissions and mitigating dependence on fossil fuels. In recent years, geopolitical conflicts affecting major oil- and natural gas-producing regions have exposed the vulnerability of the global energy supply chain, leading to significant fluctuations in energy prices and raising concerns about long-term energy security. These events have reinforced the need to diversify energy sources and accelerate the transition toward low-carbon technologies, particularly those capable of efficiently harvesting abundant renewable resources such as solar energy [1,2].

In this context, luminescent solar concentrators (LSCs) have emerged as a promising approach for solar energy harvesting, especially for building-integrated photovoltaic (BIPV) applications including conventional and smart windows, skylights, and building facades. The concept of the LSC was first introduced by Weber and Lambe in 1976 [3]. Since then, continuous advances in luminescent materials, device architectures, and fabrication techniques have renewed interest in LSC technology for photovoltaic applications [4].

An LSC consists of three main components: a transparent waveguide, luminescent materials embedded within or on the substrate, and photovoltaic (PV) cells positioned at its edges. In typical configurations, fluorescent materials are embedded within polymer or glass waveguide substrates. These materials enable spectral conversion by absorbing short-wavelength photons and re-emitting long-wavelength photons. The emitted light is then guided to the LSC edges via total internal reflection (TIR), where edge-mounted solar cells convert it into electrical energy [5,6]. This architecture offers advantages such as design flexibility, potential transparency, and compatibility with large-area applications. The performance of LSCs is commonly assessed through the optical efficiency (η_opt), which describes the fraction of absorbed photons guided to the device edges, and the power conversion efficiency (η_PCE), which reflects the overall photovoltaic conversion capability of the integrated device. Both parameters are directly influenced by the optical characteristics of the luminophore, while the choice of the host matrix plays a critical role in determining light guiding, waveguide quality, and loss mechanisms occurring during photon transport [7].

A wide variety of luminescent materials have been investigated for LSCs. Traditional semiconductor quantum dots (QDs) provide tunable emission and high absorption coefficients but frequently contain toxic heavy metals such as Cd and Pb [8,9]. Rare-earth (RE) ions offer excellent photostability but generally possess narrow absorption bands and low absorption cross-sections [10,11]. Organic dyes exhibit high absorption coefficients but suffer from photobleaching and limited long-term stability [12]. Perovskite nanocrystals (NCs) exhibit high photoluminescence quantum yields and broad absorption spectra; however, they generally contain lead and suffer from poor stability when exposed to moisture, oxygen, and heat. To overcome some of these limitations, carbon quantum dots (CQDs) have attracted increasing attention due to their favorable optical properties, low toxicity, and environmental compatibility [13].

Despite remarkable progress, the practical implementation of LSCs remains limited by several optical loss mechanisms, including incomplete solar absorption, self-absorption of emitted photons, scattering, escape-cone losses, and long-term photodegradation [14,15,16]. Therefore, developing luminophores that combine high photoluminescence quantum yield, a large Stokes shift, broad absorption, and excellent environmental stability remain one of the major challenges in the field [15,17]. To address this, combining luminophores with complementary optical properties has emerged as an attractive strategy to broaden the harvested solar spectrum and improve photon management [16,18]. In tandem LSCs, luminophores exhibiting complementary absorption and emission spectra are integrated into multilayer architectures to maximize solar photon harvesting while minimizing spectral overlap, photon reabsorption, and transmission losses. Such architectures have the potential to significantly enhance the overall utilization of incident solar radiation.

Although tandem LSC architecture has shown promising optical performance, systematic studies simultaneously correlating optical characterization, photovoltaic response, and degradation under outdoor exposure remain scarce. Furthermore, the long-term stability of tandem devices composed of different host matrices and luminophores has received limited attention [7]. In addition, the influence of combining distinct polymer matrices in tandem configurations on overall device stability has been only marginally explored.

Therefore, this work investigates luminescent solar concentrators based on two spectrally complementary luminophores embedded in different polymer matrices, which were first evaluated individually and subsequently integrated into a tandem configuration. The individual and tandem devices were thoroughly evaluated through optical characterization, photovoltaic measurements, and degradation analyses under real outdoor conditions to establish clear correlations between spectral response, device stability, and overall performance.

2. Materials and Methods

2.1. Materials

Pyrene-derived graphitized carbon quantum dots (CQDs) were synthesized by the Advanced Materials Laboratory (LMA), Federal University of Santa Catarina (UFSC), and used without further purification. Ethylene-vinyl acetate (EVA, Elvax^® 250, DuPont) was used as the polymer matrix. 8K Clear UV-curable blue-emitting resin was supplied by Polaris Microsystems & Nanotechnology LTDA. The blue-emitting resin is an additive acrylate-base for 3D printing with the exact composition not disclosure by the company. Glass substrates (120 × 120 × 6 mm³) were used as waveguides for all LSC devices.

2.2. Fabrication of LSCs

8K Clear resin was drop-cast (29ml) on a glass substrate then, a second slide was placed on top, and a UV light at 400nm was used for curing the resin for 2 minutes. The scheme of the preparation process is shown in Figure 1a.

The CQDs/EVA solution was prepared by mixing CQDs, EVA and toluene by mechanical stirring. CQDs (70.4 mg) and EVA (7.5g) were added into a beaker containing 28 mL toluene. The homogeneous mixed solution was obtained after stirring for 4 hours at 80 °C. The degassed mixture solution was poured into a smooth glass plate mold 120 × 120 × 6 mm³. After drying at room temperature, the CQDs /EVA glass was heated at 80 °C, and A second glass substrate was then placed on top of the film, followed by the application of a 1 kg load to ensure uniform lamination, and the assembly was allowed to cool to room temperature. The scheme of the preparation process is shown in Figure 5a. Pre-cut bifacial PERC silicon solar cells stripes measuring 18 mm x 105 mm were placed at four edges of the LSC. The cells were series connected in a 3D printed case to facilitate the outdoor performance evaluation. The 96 cm² illuminated area was masked by painting the 3D printed case frame with black spray ink.

2.3. Optical Characterization

The photoluminescence (PL) spectra of orange-emitting LSC and Blue-emitting resin LSC were recorded at an excitation wavelength of 405 nm using an LSP-2 spectrometer equipped. Absorption, excitation, and emission spectra were recorded using Duetta fluorescence and absorbance spectrometer (Horiba Scientific). The instrument operated in the wavelength range of 300–650 nm for absorption measurements and 400–800 nm for photoluminescence detection.

2.4. Electrical Characterization

The J-V characteristics were measured with prototypes exposed to outdoor conditions from 11:00 am to 13:00pm, using a Trace2 solar cell energy conversion efficiency testing system and a Keithley 2400 source unit. For that purpose, a fixed structure was used to expose samples to direct sun, without any tracking system. Environmental conditions (irradiance, temperature, and humidity) were also measured using a radiometer Solar Survey 200R. Samples were placed outdoors on a fixed structure with the LSC oriented toward the Sun, ensuring normal solar radiation incidence (90°), without any neighbor shadows. Photovoltaic measurements were performed in Campinas (Brazil) with latitude -22.85° and longitude -47.12°. The outdoor stability test was performed for 2 hours and 30 min.

The power conversion efficiency (PCE) was calculated according to Eq. (1):

$$PCE={\displaystyle \frac{V_{OC }. { J}_{LSC }. FF}{Ws}} ,$$

(1)

where Jsc, is short-circuit current density, V_oc, is open-circuit voltage, FF, fill factor and Ws, is solar irradiance power density.

The optical efficiency (η_opt) was calculated according to Eq. (2):

$$opt={\displaystyle \frac{I_{LSC}}{I_{SC} . G}},$$

(2)

where I_LSC is the short-circuit current of the PV cells coupled with the edge of LSC, I_SC is the short-circuit current of the PV cells under direct solar illumination, and G is the geometric gain factor, defined as the ratio of the LSC top surface area to the PV cell-terminated LSC edge area (G = A_Surface/A_Edge) which is measured to be 1.09.

3. Results

3.1. Blue-Emitting Resin in LSCs

Figure 1a illustrates the fabrication process of the laminated Blue-emitting resin LSC and the operating principle of the LSC–PV system, where the emitted photons are guided by total internal reflection toward the edge-mounted silicon solar cell. The as-prepared laminated LSC is semitransparent (Figure 1b). Upon one UV illumination, a clear concentrated blue-green light can be seen from the edges of the LSC (Figure 1c).

The absorption, excitation and emission of spectrum of the solid-state laminated blue-emitting resin LSC obtained with fluorimeter equipment is displayed in Figure 2. The absorption spectrum extends from 300 to 410 nm, with a maximum centered at 380 nm. The emission spectrum ranges from 400 to 600 nm, peaking at 460 nm, resulting in a Stokes shift of 80 nm.

In Figure 3a presents the emission spectra collected from the edge of the LSC under 405 nm excitation for different optical path lengths (L), as illustrated in the inset. The maximum emission spectrum decreases with the increase of L. This behavior indicates a dependence of the edge emission intensity on the optical path length.

The Blue-emitting resin LSCs are integrated with commercial PV cells to investigate their photo-electrical properties. The current density versus voltage (J-V) characteristic curves of LSC-PV system are shown in Figure. 4. The PCE is determined to be 0.024% (Table 1). The highest short-circuit current density (J_sc) is determined to be 0.04 mA/cm², while the open-circuit voltage

The power conversion efficiency (PCE) was calculated according to Eq. (1), and the optical efficiency (η_opt) was calculated according to Eq. (2), using the measured short-circuit current and the geometric gain factor (G). The photovoltaic parameters obtained for the blue-emitting resin LSC are summarized in Table 1.

3.2. Orange-Emitting LSC

Figure 5a illustrates the fabrication process of the laminated orange-emitting LSC and the operating principle of the LSC–PV system. Detailed preparation procedures are described in the Materials and Methods section. The fabricated device exhibits a homogeneous orange appearance under ambient light (Figure 5b). Under UV illumination, intense reddish-orange emission is observed at the edges of the LSC (Figure 5c).

As shown in Figure 6, the orange-emitting LSC exhibits an absorption spectrum extending from 300 to 600 nm, with two absorption bands centered at 355 and 500 nm. The emission spectrum ranges from 600 to 750 nm, with a maximum at 635 nm corresponding to a Stokes shift of 135 nm.

Figure 7a presents the emission spectra collected from the edge of the orange-emitting LSC under 405 nm excitation for different optical path lengths (L), as illustrated in the inset. Figure 7b summarizes the variation of the maximum emission intensity at 600 nm as a function of the optical path length.

Figure 8 illustrates the current density versus voltage (J-V) characteristic curves of LSC-PV system. The orange-emitting LSC exhibited a short-circuit current density (J_sc) of 0.12 mA cm⁻²², an open-circuit voltage (V_oc) of 1.6 V, a fill factor (FF) of 58.2%, and a power conversion efficiency (PCE) of 0.106%. The complete photovoltaic parameters are summarized in Table 2.

3.3. Tandem LSC

Figure 9a illustrates the fabrication process of the tandem LSC composed of the blue-emitting resin and orange-emitting LSCs, together with the operating principle of the LSC–PV system. Figure 9b presents the tandem LSC under UV illumination from both the orange-emitting and blue-emitting resin sides.

Figure 10 shows the absorption spectrum for tandem LSC extending from 300 to 600 nm, resulting from the combined absorption of the blue-emitting resin and orange-emitting. The emission spectrum for under excitation at 355 nm, ranges from 410 to 750 nm, shows that the tandem LSC exhibits two well-defined emission bands centered at approximately 450 and 650 nm, corresponding to the blue-emitting resin and orange-emitting luminophores, respectively. The excitation spectra corresponding to both emission bands are also presented.

Figure 11a and Figure 11c present the emission spectra collected from the edge of the tandem LSC under 405 nm excitation for different optical path lengths (L), when illuminated from the blue-emitting resin and orange-emitting sides, respectively. Figure 11b and Figure 11d summarize the corresponding maximum emission intensities as a function of the optical path length.

Figure 12a illustrates the current density versus voltage (J-V) characteristic curves of tandem LSC-PV system together with those of the individual blue-emitting resin and orange-emitting LSCs at the initial measurement (T₀) for comparison. Figure 12b compares the PCE and η_opt values obtained for the individual blue-emitting resin and orange-emitting LSCs with those of the tandem device under identical illumination conditions from the blue-emitting resin side. The tandem configuration combines the optical responses of both luminophores, and its corresponding photovoltaic parameters are summarized in Table 3.

4. Discussion

4.1. Blue-Emitting LSC

The blue emission observed under UV illumination (Figure 1c) indicates that the blue-emitting resin was uniformly distributed throughout the waveguide, which is consistent with the well-defined emission spectrum and the efficient light guiding observed in the subsequent optical measurements.

The blue-emitting resin exhibits a broad blue emission band following excitation within its UV absorption region. A Stokes shift of approximately 80 nm is observed, indicating a relatively good spectral separation between the absorption and emission bands. Although a partial overlap is present in the 400–430 nm region, its extent is limited, suggesting reduced self-absorption losses during photon propagation within the waveguide. This characteristic is particularly advantageous for LSC applications, since minimizing the overlap between the absorption and emission spectra decreases the probability of photon reabsorption, thereby improving light transport toward the device edges and potentially enhancing the optical efficiency [19]. This characteristic becomes particularly important when compared with the orange-emitting LSC and the tandem architecture, where the spectral complementarity between the luminophores further reduces reabsorption losses while extending the harvested spectral range.

Following the methodology reported in the literature [20,21], distance-dependent photoluminescence measurements were performed, as illustrated in Figure 3a. The emission intensity of the blue-emitting resin LSC decreases continuously as the distance (L) between the excitation spot and the collecting edge increases. Although the emission intensity decreases markedly, the spectral profile and the maximum emission remain essentially unchanged over the investigated optical path lengths, indicating that photon propagation mainly results in intensity attenuation rather than significant spectral distortion. This behavior consists of cumulative optical losses during photon propagation inside the waveguide, including photon reabsorption, scattering, and other attenuation mechanisms commonly observed in LSCs. The maximum emission intensity follows a typical exponential decay, retaining approximately 20% of its initial value after propagating over an optical path length of 5 cm (Figure 3b). Such exponential attenuation is characteristic of photon transport in waveguide-based LSCs and indicates that reducing the optical path length may minimize propagation losses and improve the optical efficiency of the device. Although the relatively large Stokes shift discussed previously is expected to reduce photon reabsorption, the progressive attenuation observed with increasing optical path length demonstrates that propagation losses remain significant, particularly for larger devices.

The J–V curves (Figure 4) remain nearly unchanged throughout the entire outdoor exposure period in Campinas, Brazil (October 29), indicating that the blue-emitting resin LSC preserves its photovoltaic performance under short-term environmental conditions. Likewise, the photovoltaic parameters (V_oc, J_sc, η_opt, and PCE) remain nearly constant (Table 1), suggesting that neither the luminescent properties of the resin nor the light-guiding capability of the waveguide were significantly affected during the test. These results demonstrate the good short-term photostability of the blue-emitting resin and indicate that no appreciable degradation occurred over the investigated exposure period. Since photodegradation is one of the major limitations for practical LSC applications, The negligible variation in the photovoltaic parameters during the 2.5 h outdoor exposure suggests good short-term photostability of the blue-emitting resin as a host matrix in luminescent concentrators, although longer exposure times are still required to assess its long-term outdoor durability [6]. The nearly constant values of η_opt indicate that photon collection and waveguiding efficiency were preserved during exposure, while the stable PCE demonstrates that the overall optical-to-electrical conversion capability of the LSC–PV system remained unaffected.

4.2. Orange-Emitting LSC

The intense orange emission observed under UV illumination (Figure 5c) demonstrates the strong photoluminescence of the orange-emitting LSC and suggests a homogeneous emission throughout the waveguide. This observation is consistent with the well-defined emission spectrum and efficient light guiding presented in Figure 6, confirming the effective incorporation of the orange-emitting into the EVA matrix.

The orange-emitting LSC exhibits two broad absorption bands located in the UV and green spectral regions (Figure 6), which give rise to an intense reddish-orange emission centered at approximately 635 nm. Compared with the blue-emitting resin (Stokes shift ≈ 80 nm), the orange-emitting LSC exhibits a substantially larger Stokes shift of approximately 135 nm, resulting in nearly complete spectral separation between the absorption and emission bands. Virtually no spectral overlap is observed, suggesting that self-absorption losses during photon propagation are significantly reduced [22]. Consequently, the favorable optical characteristics of the orange-emitting make them particularly attractive for tandem LSC architectures, where minimizing reabsorption losses is essential for maximizing spectral harvesting and photon transport.

The distance-dependent photoluminescence measurements shown in Figure 7a reveal not only a gradual decrease in emission intensity but also a slight red shift of the emission maximum from 608 to 612 nm as the optical path length increases. This behavior can be indicating photon propagation in LSCs where is attributed to successive reabsorption–reemission events occurring during waveguiding [3]. Similarly, the emission intensity of the orange-emitting LSC decreases continuously with increasing distance (L) between the excitation spot and the collecting edge (Figure 7b), indicating cumulative propagation losses during photon transport. Unlike the blue-emitting resin LSC, which maintained an almost unchanged emission maximum during photon propagation, the orange-emitting LSC exhibits a slight but systematic red shift. This behavior suggests that, despite the large Stokes shift of the orange-emitting, residual reabsorption–reemission events still occur over longer optical path lengths.

To evaluate the practical applicability of the laminated orange-emitting LSCs, the photovoltaic performance and stability were investigated under natural sunlight, as shown in Figure 8. Unlike the blue-emitting resin LSC, the J–V curves exhibit a progressive reduction in photocurrent throughout the outdoor exposure period, indicating a deterioration of the photovoltaic performance under short-term environmental conditions. This behavior may be partially associated with the photodegradation of the EVA matrix. Previous studies have shown that prolonged exposure to UV radiation, elevated temperatures, and oxygen can induce chain scission and oxidation of EVA, leading to discoloration, chromophore formation, and the generation of volatile degradation products. These processes reduce the optical transparency of the encapsulant and impair light transmission, ultimately compromising the photovoltaic performance of encapsulated devices [23,24]. The performance decay follows a non-linear trend, with the most pronounced losses occurring during the initial stage of exposure, suggesting that the device is more susceptible to early photodegradation processes [19]. Consistently, all photovoltaic parameters (V_oc, J_sc, η_opt, and PCE) decrease with exposure time (Table 2). In particular, the progressive decrease in the optical efficiency, from 3.52% to 1.45%, is consistent with the reduction in the amount of light effectively guided to the photovoltaic cell, suggesting that changes in the optical properties of the encapsulating matrix and/or the orange-emitting layer may contribute to the observed performance loss. This behavior contrasts with that observed for the blue-emitting resin LSC, whose photovoltaic performance remained essentially unchanged during the same outdoor exposure period. The comparison indicates that, although the orange-emitting exhibit superior optical characteristics, their environmental stability is considerably lower under the investigated conditions.

4.3. Tandem LSC

The distance-dependent photoluminescence measurements of the tandem LSC, shown in Figure 11a,c, reveal that the illuminated face primarily affects the relative contribution of the blue emission associated with the blue-emitting resin. When the excitation light affects the resin side, a higher emission intensity of around 450 nm is observed because the excitation initially interacts with the blue-emitting layer before propagating through the tandem structure (Figure 11a). Conversely, excitation through the orange-emitting side results in a lower blue emission intensity due to the different optical path followed by the incident photons (Figure 11c). In contrast, the reddish-orange emission attributed to the orange-emitting s remains nearly unchanged in both spectral shape and intensity regardless of the illuminated face (Figure 11a–c). This behavior indicates that the orange-emitting layer is efficiently excited in both irradiation configurations, demonstrating that its optical response is largely independent of the excitation direction. Consequently, tandem architecture preserves the complementary contribution of both luminophores while allowing the irradiation direction to modulate mainly the blue-emitting component. Although the irradiation direction influences the relative intensity of the blue emission, the characteristic emission bands remain at essentially the same wavelengths throughout the measurements.

Figure 11b,d quantitatively summarize the variation of the emission intensity as a function of the optical path length for both excitation configurations. In both cases, the emission intensity decreases continuously with increasing distance (L) between the excitation spot and the collecting edge, confirming the cumulative propagation loss occurring during photon transport within the tandem waveguide. Similar attenuation behavior has been widely reported for luminescent solar concentrators and is mainly associated with residual absorption, scattering, and other waveguide losses during photon propagation [5,9]. Although the attenuation behavior is observed for both irradiation configurations, the relative intensities of the blue and reddish-orange emission bands differ depending on the illuminated face, reflecting the sequential interaction of the excitation light with the blue-emitting resin and orange-emitting layers. Nevertheless, both emission bands exhibit similar attenuation trends with increasing optical path length, indicating that photon propagation losses remain the dominant mechanism governing the optical performance of the tandem LSC.

The photovoltaic performance of the tandem LSC was evaluated under natural sunlight and compared with those of the individual blue-emitting resin and orange-emitting LSCs, as well as with the bare silicon solar cell, as shown in Figure 12a. Tandem architecture exhibits higher photocurrent over the entire voltage range than individual LSC, demonstrating that combining the two luminophores enhances the overall photovoltaic response. Among the two irradiation configurations, slightly higher current densities are obtained when the incident light reaches the orange-emitting side first, indicating that the excitation direction influences the efficiency of photon harvesting within the multilayer structure. The photovoltaic parameters summarized in Figure 12b further confirm the advantage of the tandem configuration. Both the optical efficiency (η_opt) and the power conversion efficiency (PCE) increase substantially compared with the individual LSCs, demonstrating that the complementary optical responses of the blue-emitting resin and orange-emitting improve the utilization of the incident solar spectrum. The highest efficiencies are obtained when the orange-emitting face is illuminated, indicating that the sequence in which light interacts with the two luminophore layers plays an important role in the overall photon management of the tandem device. This behavior is attributed to the complementary absorption and emission characteristics of the two luminophores. While the blue-emitting resin efficiently converts higher-energy photons into blue emission, the orange-emitting extends the spectral response toward longer wavelengths. Their integration into a tandem architecture broadens the usable spectral window, increasing the number of photons guided to the photovoltaic cell. Compared with the blue-emitting LSC, the tandem architecture increases the optical efficiency by approximately eightfold, while an improvement of more than twofold is observed relative to the orange-emitting LSC. These results demonstrate that the tandem strategy successfully combines the favorable optical properties of both luminophores, resulting in a device with broader spectral harvesting and superior photovoltaic performance compared with the individual LSCs. Therefore, tandem architecture represents an effective approach for improving the efficiency of luminescent solar concentrators while maintaining the advantages of complementary luminescent materials.

5. Conclusions

This work presented the development and systematic evaluation of individual and tandem laminated luminescent solar concentrators (LSCs) based on a blue-emitting resin and orange-emitting carbon quantum dots (CQDs) embedded in an EVA matrix. The individual devices exhibited complementary characteristics: the resin-based LSC showed excellent short-term outdoor stability, whereas the orange-emitting LSC presented a larger Stokes shift (≈135 nm), minimizing self-absorption losses but exhibiting lower environmental stability under outdoor exposure. By combining both luminophores, tandem architecture broadened the spectral harvesting range and delivered the best photovoltaic performance, achieving a maximum optical efficiency of 8.09% and a power conversion efficiency of 0.306% under orange-emitting LSC side illumination. These findings demonstrate that tandem architecture based on spectrally complementary luminophores effectively enhance solar spectrum utilization while balancing the advantages of different luminescent materials. Future studies should focus on improving the photostability of the orange-emitting LSC through advanced encapsulation strategies and more durable polymer matrices to enable long-term BIPV applications.