Photovoltaic plant losses: spectral mismatch

Multijunction solar cells

MJ and tandem solar cells mitigate spectral mismatch by increasing the system's absorption range and harvesting more energy from across the incident spectrum as compared to standalone solar cells. In MJ solar cells, a series of cells with optimized individual absorption ranges is combined into cell stacks. The resulting overall absorption range of the cell stack is thus widened compared to the individual absorption ranges of the subcells, as visualized in Figure 4. In these cells, spectral mismatch is therefore directly mitigated through energy harvesting of an extended range of incident photons by the respective subcell whose bandgap best matches the photon's energy. The cells can be integrated into a unified structure using either mechanical interconnection or monolithic integration approaches^[16]. In organic tandem solar cells, for example, combining ITO/P₃HT: PCBM/Al processed as a diffused bilayer with an absorption range between 375 and 630 nm as the bottom cell, and ITO/ZnPc/ZnPc: C₆₀/C₆₀/Cr/Al as the top cell with an absorption range between 600 and 800 nm. This combination increases effective total absorption range of the hybrid system to [375-800 nm]^[30].

Figure 4. Layered tandem solar cell structure and the respective absorption spectra of the absorber materials. The absorber materials are combined to widen the overall absorption range and thus mitigate spectral mismatch.

Due to their better matched, broader absorption pattern, the limiting efficiency of MJ solar cells exceeds the Shockley-Queisser limit of conventional, single junction SSCs^[31]. For a hypothetical MJ cell of an infinite number of cells, the maximum PCE under one sun illumination is 68.2%, and as much as 86.8% under highly concentrated sunlight^[32]. In this case, spectral mismatch is reduced to a minimum as there would exist an optimized subcell for the absorption of each specific photon in the incident solar spectrum and its associated energy.

Popular cell types for integration into tandem architectures include III-V solar cells, cells with c-Si, the highly promising solid-state PSCs, and thin-film solar cells (TFSCs). III-V MJ solar cells are advanced solar cells that combine semiconductor materials from the third (III) and the fifth (V) groups of the periodic table in the form of multiple stacked layers, each optimized to absorb a specific portion of the solar spectrum.

Under unconcentrated (one sun) AM1.5G irradiance at 1,000 W/m², a remarkable PCE of 39.2% has been achieved by a monolithically grown inverted metamorphic six-junction III-V solar cell in 2020^[33], with a notable open-circuit voltage V$$ _{oc} $$ of 5.55 V. The high PCE can in parts be attributed to the reduced internal resistances from the usage of a reverse heterojunction AlGaInP subcell. It is to be noted that an aperture-area of 1 cm² is required for PCE record recognition, while this six-junction device featured an aperture area of merely 0.25 cm².

Silicon with a 1.1 eV bandgap is close to the optimum for dual-junction and triple-junction devices aiming to reduce spectral mismatch losses. Therefore, aiming to combine the high PCE of modern c-Si solar cells with the high potential of III-V MJ solar cells is another focus of current MJ device research. A promising attempt at manufacturing high-performance, high-reliability III-V silicon tandem solar cells at competitive costs is direct growth of III-V absorber layers using methods such as metalorganic vapor-phase epitaxy (MOVPE). This methodology utilizes a precise crystal growth technique where metal-organic and hydride gases are used to deposit thin absorber layers on a substrate^[34]. However, for direct growth of III-V absorber layers onto silicon, achieving low enough defect densities proves challenging due to the large difference in the thermal expansion coefficient as well as the 3%-4% difference in the lattice constant between the III-V layers and silicon. One such triple-junction III-V silicon tandem solar cell in which III-V layers were epitaxially grown onto GaAs, featuring a top tandem wafer-bonded to a TOPCon (tunnel oxide passivated contact) silicon bottom cell, achieved a remarkable PCE of 36.1% under unconcentrated (one sun) AM1.5G irradiance^[35].

PSCs, which use perovskites as light absorbers, are another popular cell type for tandem configurations. Perovskite materials consist of crystal structures with the chemical formula ABX₃. Examples for perovskite materials include lead halides such as formamidinum lead iodide (CH₅I₃N₂Pb, also denoted FAPbI₃)^[36] or the archetypal methylammonium lead iodide (CH₃NH₃PbI₃, also denoted MAPbI₃). However, due to concerns related to the long-term stability, as well as the toxicity of the lead^[37], lead-free, all-inorganic alternatives such as the caesium antimony iodide Cs₃Sb₃I₉^[38] or caesium tin iodide Cs₂SnI₆^[39] are gaining popularity. There exist two main PSC cell architectures. The first is a sensitized solar cell architecture with a liquid electrolyte, often referred to as a liquid-state PSC, where perovskites serve as photon-absorbing sensitizers for carrier excitation. The second, more actively studied PSC architecture, is a PV cell structure employing perovskites as solid state absorbers as visualized in Figure 5.

Figure 5. Schematic structure of a solid state PSC.

MJ devices based on stacked solid state perovskite absorbers represent a key focus in PV cell research. Numerical simulations investigating thickness-optimized PSC tandem devices with maximal light absorption across unconcentrated (one sun) AM1.5G irradiance have returned highly promising PCE's as high as 28.15%^[40]. A recent experimental study of a monolithic perovskite-perovskite-silicon triple junction cell with optimized light management and maximized photocurrent has achieved a PCE of 24.4%^[41], approaching simulated values. Beyond its high PCE, the device exhibited excellent thermal stability, retaining 96.6% of its initial PCE after exposure to a N₂ atmosphere at 85 °C for around 1,000 h. See Table 1 for more details.

TFSCs are another promising cell type for the mitigation of spectral mismatch losses owing to their bandgap tun-ability enabling the optimization of the absorption characteristics with respect to the incident spectral composition. Common TFSCs include CIGS (Cu(In, Ga)Se₂), CdTe (cadmium telluride), CZTS (Cu₂ZnSnS₄), CZTSSe (Cu₂ZnSn(S, Se)₄), or silicon TFSCs. Current research trends highlight the integration of TFSCs as narrow-bandgap bottom subcells within tandem PV devices. For instance, the bandgap tunability of CZTSSe thin-films in the range between 1.0 and 1.5 eV may allow this material to maximize the PCE when utilized in tandem cell architectures, where the ideal bottom cell bandgap is around 1.1 eV^[54]. Recent studies have also explored TFSC-based tandem architectures, such as a numerical study investigating the design and optimization of a double-absorber-based tandem structure comprising CZTS and CZTSSe TFSCs, resulting in a simulated PCE of 26.31%^[52]. A life cycle assessment on commercial and emerging TFSC systems revealed a superior environmental performance over conventional c-Si solar cells, despite a longer energy payback time due to their PCE deficit compared to recent performances achieved by other cell types such as PSCs^[55]. A notable CIGS MJ TFSC has achieved a remarkable PCE of 22.30% under non-concentrated AM1.5G irradiance at a cell temperature of 25 °C, indicative of the strong performance of state-of-the-art TFSCS developed to increase the spectral absorption range of the system and effectively mitigate spectral mismatch^[56].

Other recent experimental or numerically simulated results for each of the named MJ solar cell types under unconcentrated (one sun) AM1.5G irradiance are summarized in Table 1, highlighting the competition surrounding R & D of modern MJ and tandem PV technologies and their potential to achieve ultra-high PCEs.

As of today, MJ solar cell technology is mainly limited to regions on Earth with high direct normalized irradiance or for use in satellites due to the cells' potential for a low weight to electrical power ratio^[16]. For terrestrial use, MJ cells are most practical with the addition of optical concentrators, which redirect and concentrate light rays from a large area onto a small area to enable higher efficiency^[57]. Such MJ concentrator solar cells have achieved PCEs of over 40% since 2006^[58]. The practicality of MJ cells in conjunction with optical concentrators is in part due to the cells' far higher cost compared to conventional Si solar cells. New inefficiencies such as the introduction of resistive losses related to the lateral interconnection between top and bottom strings increases design complexity and thus hinders scaling MJ/tandem technology to larger areas. For example, while the prices for mainstream c-Si solar cells or cadmium telluride (CdTe) modules are at around $0.30–$0.50/W, current fabrication costs of III-V//Ge tandems are significantly higher at values between $70/W to $100/W^[31]. Thermal management plays a critical role for concentrator cells and high power irradiance. Using interface materials such as low-index oxides or fluids with high thermal conductivity is a viable approach for enhancing heat dissipation^[59].

In 2024, a wafer-bonded four-junction III-V concentrator solar cell under 665 sun AM1.5D irradiance, has experimentally demonstrated an ultra-high PCE of 47.6%^[60]. Operation of the MJ device at such high concentrations was enabled by the low specific series resistance of the device. Other electrical characterizations of III-V MJ concentrator cells reported in experimental studies demonstrating outstanding PCEs are presented in Table 2. These results show the potential of concentrator MJ solar cells to surpass the 50% PCE milestone in the near future.

While MJ cells are more efficient than their single-junction counterparts, they feature certain loss mechanisms that partially counteract their ability to mitigate spectral mismatch. Among such loss mechanisms is the slight increase in optical reflection and parasitic absorption as a result of the supporting layers or interfaces. With spectral overlap of some of the utilized absorber materials, photons of a certain energy may be absorbed by narrow-gap materials instead of wide-gap materials, leading to an undesired increase in thermalization losses and a lower generated voltage. Furthermore, the output voltage may decline due to tunnel recombination effects^[63], an effect in MJ solar cells where charge carriers recombine within the tunnel junctions. Tunnel junctions use quantum tunneling to electrically connect adjacent sub-cells in MJ solar cells by allowing the transfer of charge carriers between sub-cells without a significant barrier.

Substantial progress in MJ/tandem solar cell technology is required to enable large-scale grid integration and enhance cost scalability^[64]. While MJ architectures efficiently mitigate spectral mismatch by enabling substantially widened absorption ranges, the integration of subcells that respond to unique parts of the incident spectrum increases this technology's susceptibility to operating temperature and spectral variability over the course of a year. This leads to further challenges in yield prediction and overall system design. MJ solar cell grid integration therefore requires more consistent energy yield models and spectral resource datasets integrated into PV design tools.

Furthermore, the integration of multiple absorber materials may introduce new degradation mechanisms. Hence, long-term reliability demonstrations are required to lower the Levelized Cost of Energy (LCOE) of MJ cells and enhance economical scaling. Additional grid-related concerns of MJ technology include the increased embodied energy and greenhouse-gas footprint at a cell-level, which necessitate significantly increased efficiencies and longer module lifetimes.

Co-sensitization of dye-sensitized solar cells

DSSCs are promising alternatives to conventional solar cell systems, offering the potential for designs mitigating the effects of spectral mismatch. Unlike in conventional systems, light is absorbed by sensitizers attached to the surface of a wide bandgap semiconductor photoanode of nanocrystalline morphology^[65], commonly titanium dioxide (TiO₂). The schematic structure of DSSCs as well as the photoelectric processes are shown in Figure 6. After photons pass through a layer of transparent conductive oxide (TCO), typically fluorine doped tin oxide (FTO) and the porous TiO₂ layer, charge separation takes place via photo-induced electron injection from the dye sensitizer into the TiO₂ conduction band. In the conduction band of the semiconductor, carriers are then transported to the charge collector. The transportation of electrons back into the dye is achieved through the oxidation of triiodide within the iodide/triiodide electrolyte located between the photoanode/sensitizer and a counter electrode on the back side of the DSSC. Having passed through consumer devices, electrons arrive at a counter electrode commonly made of platinum. Finally, carriers from the counter electrode are used for the reduction of the iodide to complete the electrolyte reaction cycle.

Figure 6. Schematic structure of co-sensitized DSSCs. The co-sensitization is performed by combining multiple dyes (green and orange circles) with complementary absorption ranges, widening the cell's overall absorption range and effectively mitigating spectral mismatch losses.

Through the association of light-absorbing dyes (sensitizers) with wide-bandgap semiconductor materials, DSSCs can harvest a large fraction of the incident sunlight. Dyes used as sensitizers in DSSCs are classified into metal complex dyes, metal-free organic dyes, and natural dyes. The most common and extensively investigated sensitizers in DSSCs are ruthenium (Ru) based dyes such as N719 or N3. With such Ru-based dyes posing toxicity-related hazards to the environment during the manufacturing and disposal stage, recent developments have shown growing interest in less environmentally hazardous alternative sensitizers. Consequently, increasing research efforts have shifted toward metal-free organic dyes, quantum-dot sensitizers, perovskite-based sensitizers and even natural dyes in natural DSSCs (NDSSCs)^[66]. The development of such alternative sensitizers is also economically motivated, with organic or metal-free alternative sensitizers being less expensive than Ru-based dyes.

Similar to how spectral mismatch losses are mitigated in MJ solar cells by increasing the absorption range and pattern through the integration of multiple absorber materials, the absorption characteristics of DSSCs may be improved by broadening the absorption range via co-sensitization using multiple dyes with complementary optical absorption properties^[67] as shown in Figure 6. With co-sensitization, limitations imposed by the narrow spectral range of individual sensitizers in DSSCs can be effectively overcome^[68]. To further broaden the absorption range and thus further improve solar energy harvesting, co-sensitized DSSCs may be arranged in tandem configurations. In addition to enhancing performance by broadening the absorption range, co-sensitization fosters improvements in efficiency through the suppression of charge recombination and increases in V$$ _oc $$ and I$$ _sc $$.

A recent experimental study reporting on the electrical characterization of an N3-sensitized DSSC co-sensitized with the novel metal-free organic N-1 dye resulted in a near doubling of the device's PCE from 5.42% achieved by the N3 dye alone to 10.18% for the co-sensitized system (PCE enhancement of 87.8%)^[69]. Co-sensitization of two metal-free organic dyes, MS5 and XY1b, resulted in a similarly impressive PCE enhancement of 68.8% (from 8.0% for MS5 alone to 13.5% for MS5 + XY1b)^[70]. Furthermore, recent research literature has investigated ternary systems, in which multiple co-sensitizers are employed. An example for such a ternary system is the co-sensitization of the ruthenium-based metal complex dye N-719 with two metal-free organic dyes, MO-1 and MO-2, resulting in a PCE enhancement of 49.1% (from 7.35% for N719 alone to 10.96% for the ternary system)^[71]. Test conditions, electrical characterizations, and PCE enhancements reported in recent experimental studies investigating co-sensitization of DSSCs are summarized in Table 3.

DSSCs are fabricated using low cost techniques such as inkjet/screen-printing or the roll-to-roll technique^[77], making this cell type an economically viable alternative to other solar cell technologies. DSSCs are fabricated at ambient temperatures and are less prone to contamination^[78]. Several technological advantages point towards the high potential of DSSCs to play a major role in future PV developments: DSSCs exhibit shorter payback times for both energy and CO₂. Depending on the utilized sensitizers, DSSCs additionally hold the potential to minimize toxicity^[79]. DSSCs achieve relatively high PCEs even under low light intensity conditions at dawn, dusk, or during cloudy weather. Such spectral robustness makes DSSCs a promising candidate particularly for indoor applications^[80]. DSSCs exhibit remarkable power generation characteristics in environments and times of low light intensity. Furthermore, DSSCs exhibit a reduced sensitivity to changes in the incident light angle on the cell, potentially eliminating the need for a mechanical light tracking system^[81]. Despite improved spectral robustness, spectral variations throughout the course of a day due to changing solar zenithal angles impact the DSSC's power generation in real-world integration cases, facilitating either spectral gains or losses.

NDSSCs use organic dyes at different degrees of purification as an alternative to synthesized dyes. With their cost-effectiveness and environmental friendliness, NDSCCs hold potential not only in producing green energy but also in establishing an environmentally friendly production of solar power plants^[76]. Common examples for natural dyes employed in NDSSCs include natural pigments from fruits such as anthocyanins (which are responsible for fruits' different colors)^[82], natural dyes extracted from cereal grains such as black rice^[83] or from flowers such as Begonias, Perillas, or different rose types^[84], or natural carotenoids from plants algae such as kelp (which are yellow, orange, and red pigments)^[85]. Among the major challenges that NDSSCs face is their low conversion efficiency. Future research may lead to efficiency breakthroughs in NDSSCs up to 14% or higher, which have already been achieved by synthetic dye-based DSSCs^[86].

Two major challenges of DSSCs compared to silicon-based technologies are their generally lower efficiency and shorter life-times^[87]. Particularly, the relatively low V$$ _{oc} $$ of DSSCs is an efficiency-limiting factor^[88]. Furthermore, harvesting solar energy towards the low-energy part of the solar spectrum has proven challenging^[78]: after excitation of electrons in red-absorbing dyes via low-energy photons, the excess energy of such electrons past the conduction band of the TiO₂ electrode is only small. This leads to a reduction in the amount of charge injections per absorbed low-energy photon.

To reduce transmission of low-energy photons and increase the chance of absorption, the path length of light within the cell can be increased either by increasing the film thickness beyond 10 μm or employing light trapping techniques. For DSSCs, confining incident light can be achieved by structures with dispersed TiO₂, which cause multiple scattering^[89]. The cell's performance can be improved by installing a special silver back-reflecting layer^[90]. Using graphene-TiO₂ composite photoanodes has achieved a relative improvement of 59% to reach a PCE of 4.28%^[91]. Concerning ruthenium-based dyes, however, further breakthroughs in cost and efficiency have appeared to be increasingly less likely^[78].

Hot carrier and hybrid thermoelectric solar cells

HCSCs have the potential for breaking the Shockley-Queisser limit^[92]. The idea of HCSCs was first proposed in 1982. HCSC technology attempts to harvest the excess energy of so-called hot-carriers (electrons excited past the bandgap energy by high energy photons) by establishing a steady-state hot carrier distribution that can capture the excess kinetic energy of hot carriers. By extracting hot carriers and delivering them to the electrodes of a given cell prior to having fully cooled down, their surplus energy would partially contribute to an additional electrical output^[93]. This reduces thermalization losses and effectively mitigating spectral mismatch.

In the case of complete harvesting of surplus energy in hot carriers, the collected energy per extracted electron-hole pair is larger than the bandgap energy^[20]. HCSCs may achieve a maximum PCE of 66% according to thermodynamic calculations for single junction cells with an ideal bandgap of 0.7 eV to 0.9 eV, assuming full absorption of photons with greater energy than the bandgap energy under 1 sun illumination if the energy of such hot carriers is fully harvested before the cells cool back to the lattice temperature^[94]. Under the same assumptions, the PCE of single junction solar cells may reach ~85% at maximum solar concentration.

The extraction of high-energy hot carriers can only be achieved when the cooling process is greatly slowed and/or the extraction is greatly accelerated. Hot carrier collection is achieved by introducing selective energy contacts around the absorber material which only allow the passing of hot electrons and hot holes within a narrow energy window, respectively, and in doing so prevent heat flow back toward a cold carrier population^[93]. The conceptual design of the HCSC is visualized in Figure 7, where E$$ _{gap} $$ is the absorber's bandgap energy, E$$ _{ext} $$ is the extraction energy of an energy selective contact, and $$ \delta $$E is the transmission range (or energy window) of an energy selective contact.

Figure 7. Schematic design of HCSC's, visualizing the hot carriers (electrons). Spectral mismatch losses are reduced by capturing the hot carrier's excess thermal energy.

Simulations have shown that for a 10,000× concentrated AM1.5D solar spectrum, both the incident power that is being absorbed and that which is being lost through thermalization decrease significantly with decreasing absorber thickness^[20]. Therefore, to enable and maintain sufficient absorption and allow for efficiency improvements in ultrathin absorber layers, enhanced light trapping techniques are required.

Unfortunately, as of today, experimental work to realize the concept of HCSCs is scarce due to the difficulty of finding suitable absorber materials in which cooling is sufficiently slowed down to allow for the extraction of hot carriers while maintaining the absorber's ability to absorb a wide spectral range. Energy selective contacts with a narrow density of states at an appropriate energy level need to be engineered in order to suppress cooling through the contacts and leakage of hot carriers^[92]. HCSCs were shown to have higher spectral robustness compared to tandem devices, implying a reduced sensitivity to spectral variations due to changing solar zenithal angles throughout the day^[95]. Substantially reduced current density as caused by insufficient extraction of widely distributed hot electrons through the requisite energy selective contact can severely limit the performance of HCSCs^[96].

Another promising technology that has been studied in conjunction with PV's is thermoelectric (TE) generators. These hybrid solar systems work by repurposing unused heat generation from solar cells into electricity using the TE generators. TE's generally work best in the IR range of the solar spectrum, which is typically below the bandgap of standard PV cells, thereby effectively extending the usable solar spectrum range of a PV system^[97]. These generators work by utilizing the Seebeck effect, first discovered by Thomas Seebeck in 1821, which operates on the principle that a temperature difference between two different materials creates a potential difference relative to the temperature gradient, inducing a current flow.

A TE generator is made of multiple TE couples connected in series electrically and in parallel thermally consisting of p-type and n-type semiconductor materials^[98]. A temperature difference across the TE generator causes electrons and holes to travel from the hot end to the cold end of the couple, where they recombine and cause a current flow through the connected external load. This technology has been applied in conjunction with PV technology to harvest unused heat from photons under the bandgap within the IR spectral region, as well as for collecting heat due to thermalization within the solar cell itself. To promote the temperature difference needed, a heat sink is added to the opposite side of the TE generator. This in essence allows the system to make use of a wider spectral range, thereby increasing overall system efficiency.

One study utilizing this hybrid setup found an efficiency increase from 12.5% to 16.3% by incorporating TE's into c-Si solar cells to harvest losses from both thermalization and transmission^[99]. Another study found that incorporating TE's into MJ solar cells also helped to offset losses due to thermalization. This hybrid MJ system achieved maximum efficiency values of ~32% under concentrated sunlight^[100]. A major contribution to the efficiency increase of this system was from the TE's cooling effect on the PV, while only a smaller contribution to efficiency increases were from energy generation from the TE itself. This presents the notion that TE's can be utilized to improve PV efficiency by other methods besides from their own power generation. In addition to this, research has been proposed to incorporate heat storage devices into hybrid solar cell systems in order to continue power generation during the night. During the day, heat storage devices collect heat either by raising temperature or with the use of phase change materials^[101]. At night this heat can be used by TE generators to generate power when solar cells are unable to, allowing for full-time power generation. Currently there are limited studies on utilizing this technology with hybrid solar systems. Furthermore, the costs involved by incorporating TE technology into PV's are high. The addition of BiTe TE modules increases the cost per meter of a standard Si panel from ~400 $/m² to ~2,100 $/m^2[102].

Quantum dot solar cells

IBSCs were proposed in 1997 as an alternative to tandem solar cells, promising ultra-high solar-electricity energy conversion^[103]. IBSCs reduce spectral mismatch losses that occur due to non-absorption (transmission) by introducing absorber materials with one or more distinct intermediate bands between the conduction and valence bands of an otherwise conventional semiconductor material^[104]. The schematic working principle of IBSCs based on multi-photon excitation is visualized in a bandgap diagram in Figure 8. Here, E$$ _{vi} $$ is the energy of the first excitation step of an electron from the valence band to the newly introduced intermediate band and E$$ _{ci} $$ is the energy for the excitation of an electron from the intermediate band energy state to the conduction band via a second low-energy photon. E$$ _{cv} $$ is the energy difference for direct excitation from the valence band to the conduction band corresponding to the bandgap energy E$$ _{gap} $$ of an equivalent cell without the intermediate band. By introducing narrow bands of electronic states, the absorption spectra of the semiconductor material is extended to previously sub-bandgap photons, effectively leading to an increase in PCE over conventional single-band solar cells^[105].

Figure 8. Energy band diagram of IBSCs allowing for excitation processes to intermediate energy bands, effectively allowing the absorption of lower energy photons and thus enhancing the cell's absorption range.

Optimizing the energy levels of the bands for IBSCs reveals a thermodynamic efficiency limit of 74.6% when four intermediate bands are included. IBSCs with one intermediate band exhibit a limiting efficiency as high as 63.2% under maximum solar concentration (46,050 suns) for optimum energy separations of E$$ _{iv} $$ = 1.24 eV (valence to intermediate band) and E$$ _{ci} $$ = 0.71 eV (intermediate to conduction band)^[106]. Despite experimental progress having been made within each technological approach, the benefits of the intermediate band have so far not been fully exploited by today's IBSC implementations^[103].

Quantum dot solar cells (QDSCs) mitigate spectral mismatch by utilizing Quantum Dots (QDs) that offer tunable bandgaps to optimize the cell's absorption spectrum. QDs are nanometer-sized semiconductor particles that confine charge carriers in all three dimensions, create discrete energy levels, enable precise energy selection, and act as real energy selective contacts. Bandgap tuning is achieved via size variation of the QDs.

Popular QD materials investigated in literature include lead-based materials such as lead chalcogenides (PbS, PbSe, and ternary alloys such as PbS_xSe$$ _{1-x} $$), lead halide perovskites (PVK), and lead free materials such InAs, InP, or CdTe^[107]. While a variety of QD-based PV architectures have been realized, the two main types are Quantum Dot Sensitized Solar Cells (QDSSCs) and Colloidal Quantum Dot Solar Cells (CQDSCs)^[108]. QDSSCs function like other sensitized solar cell types such as DSSCs but, instead of dyes, rely on quantum dots as sensitizers for light absorption and charge separation. In CQDSCs, one or multiple thin films embedding colloidal QDs (quantum dot superlattices) are integrated as active absorber layers between electron and hole-selective layer. The schematic structure of a CQDSC integrating two stacked QD-doped thin films is shown in Figure 9.

Figure 9. Common structure of a colloidal QDSC in which spectral mismatch is mitigated by fine-tuning the bandgap of the QDs.

To improve photostability of the QDSC, enhance carrier mobility and decrease the negative effects of carrier recombination, surface passivation layers may be added to the QDs, effectively leading to advances in the achieved PCE. For example, by optimizing the surface passivation of PbS QDs, an overall PCE of 9.6% was achieved^[109].

Provided a sufficient density of QDs embedded or grown within a host material, energy separation of the QD levels from both the valence and conduction band, and strong dot-to-dot coupling, the discrete quantum dot energy levels can hybridize into discrete intermediate bands. Such a distinct energy level can be achieved through variation of size, density and coupling of the QDs. If the distinct QD-generated energy band is entirely within the host bandgap acting as a long-lived carrier reservoir and supporting two-step photon absorption, the device effectively acts as an IBSC. In this case the device would be referred to as a quantum dot intermediate band solar cell, or QDIBSC.

Investigating the impact of different Sb compositions in a type-II GaAsSb capping layer of a QDSC incorporating an InAs QD layer, an Sb composition of 18% yielded the high measured PCE of 17.31% under unconcentrated AM1.5G irradiance. This PCE increase can be attributed to a rising GaAsSb valence band with increased Sb compositions. The obtained PCE corresponds to an 11.25% increase compared to type-I InAs/InGaAs QDSCs^[110].

Inadequate light collection efficiency and the influence of charge recombination could be the primary reasons for low PCEs of QDSCs compared to conventional SSCs or even emerging PSCs. Aiming to minimize charge recombination, passivation layers may be incorporated directly around the QDs or between layers of the QDSC structure. Better utilization of sunlight may lead to further efficiency improvements for QDSCs. For example, by additionally depositing black phosphorus QDs onto a ZCISSe QD-sensitized TiO₂ electrode, the photoanode's light capture ability was enhanced, resulting in a PCE of 15.66% under unconcentrated (one sun) AM1.5G irradiance^[111].

In order to optimize QDSCs for operation under concentrated sunlight, a QD based InAs/GaAs device was equipped with a InGaP top cell in order to prevent extreme over-generation of electric current on account of the very low transition energy of InAs QD's of only around 0.3 eV^[112]. By using this design approach, the short circuit current density J$$ _{sc} $$ was decreased to 13.0 mA/cm² under one sun concentration, a 50% reduction compared to non current-constrained InAs/GaAs QD based IBSCs, resulting in a PCE of 22.80% PCE under 93 sun concentration.

Table 4 lists electrical characterizations featuring notable PCE achievements reported in recent experimental studies investigating a variety of QD based PV devices.

Quantum well solar cells

Quantum Well Solar Cells (QWSCs) incorporate Quantum Wells (QWs) in the form of thin layers of semiconductor material within otherwise conventional wider bandgap semiconductor materials. QWs enable spectral mismatch mitigation by creating new, tunable energy states, effectively allowing the absorption of previously sub-bandgap photons towards the IR end of the incident spectrum. State of the art QW based PV devices further enhance absorption characteristics by employing multiple QWs. The effective bandgap of QW structures is influenced by quantum size effects, quantum confined stark effects, and stress effects^[117]. QWs do not act as isolated intermediate bands due to their inability to sustain independent quasi-Fermi levels, and do not act as long-lived carrier reservoirs. Therefore, unlike in IBSCs using supplementary two-step photon absorption, QW based devices generally rely on standard photogeneration followed by carrier escape/extraction from QWs into the surrounding conduction band via thermionic emission or (thermally assisted) tunneling^[118]. As opposed to QDs, which confine charge carriers in all three dimensions, QWs restrict charge carrier movement to two dimensions in the plane of the well. Because QWs merely act as potential steps, a lower maximum PCE is expected than for QD-based energy selective contacts. Both QW and QD nanostructures can be applied to reduce entropy loss^[92].

By incorporating strain-balanced GaAsP/GaInAs QWs into III-V two-junction GaInP/GaAs and single-junction GaAs PV devices, PCEs of 32.92% and 27.18% were achieved, respectively^[119]. Recently, exploration of high-efficiency solar cells that integrate QDs into QW structures (commonly referred to as DWELL, meaning Dot in Well) has shifted into focus. In 2024, a remarkable PCE of 31.8%^[120] was numerically simulated using the SCAPS-1D software by effectively creating intermediate energy levels using CdTe QDs embedded in Al_0.3Ga_0.7As/GaAs QWs with an open circuit voltage V$$ _{oc} $$ of 1.36 V and a fill factor between 77% and 85%. The authors suggest the use of CdTe QDs given their low cost and suitability in terms of photocurrent generation. Table 5 summarizes electrical characterizations featuring notable PCE achievements reported in recent experimental and numerical studies for QW-enhanced PV devices.

Upconversion

Several methods exist for the facilitation of UC technology in PV devices. Common approaches include adding UC layers or UC material-doped glass layers at the rear of bifacial solar cells, as well as the doping of photoanodes with UC nanorods or nanoparticles (NPs). Transmitted sub-bandgap photons are absorbed via sequential ground state absorption/excited state absorption processes in a three-level system. Principal UC processes include excited state absorption, energy transfer, and the photon avalanche process. Other UC processes such as cooperative UC and energy migration mediated UC are being researched. Figure 10 visualizes the principal excited state absorption UC process, where E₁ is the energy for excitation from the ground state to the intermediate energy level of the upconverter material, E₂ is the energy for excitation from the intermediate energy level to the top energy level, and E_g is the energy of the photon emitted upon relaxation of the electron from the top energy level to the ground state, which corresponds to the bandgap energy of the solar cell. Since the upconverting material outputs one photon for two absorbed photons, the maximum quantum efficiency of the upconverter can only be 50%. Transmission losses are generally higher for solar cells with a wider bandgap, and lower for narrower bandgaps. UC pushes the theoretical limiting efficiency of single junction SSCs, which have a bandgap of 1.12 eV, to 40.2%^[127]. This is a relative improvement of almost 26% to the Shockley-Queisser limit of 32% for the same SSCs. The highest limiting efficiency of solar cells using UC systems is 50.69% for a bandgap of 2.0 eV.

Figure 10. Band diagram for upconversion process visualizing the conversion of two low-energy photons into one photon of higher energy tuned to the bandgap of the employed solar cell.

UC systems are commonly employed in PV devices including SSCs, amorphous Si TFSCs, III-V solar cells and DSSCs^[128]. Working principles include UC via rare earth ions, UC via triplet-triplet annihilation, UC via QDs, or UC via thermal radiation^[16]. As previously mentioned, the impact of UC is greatest in solar cells with a bandgap above 1.25 eV. This is due to the fact that for cell's with a bandgap above 1.25 eV, transmission losses (which are mitigated by UC) outweigh losses due to thermalization of hot carriers.

While the bandgap of crystalline SSCs of ~1.12 eV (corresponding to a wavelength of 1,100 nm) is below the mark of 1.25 eV, sub-bandgap transmission are still significant in this cell type at 20%^[129]. Given the magnitude of these losses and the dominance of SSCs in the solar cell market, developing cost-effective, PCE-enhancing UC technologies holds significant economic potential.

The concentration factor of the incident irradiance has a considerable impact on UC efficiency^[17]. For NaYF₄: Er³⁺, Yb³⁺, one of the most investigated upconverter materials, saturated quantum UC efficiencies of around 10% have been reported at about 40 W/cm^-2 (400 sun concentration). At lower concentration factors of e.g. six (ca. 0.6 W/cm^-2), Gd₂O₂S: Er³⁺, Yb³⁺, another UC material, experiences a saturated quantum efficiency of less than 1%. Aiming to increase the UC efficiency under conditions of low light intensity, the use of photoluminescent QDs in combination with UC materials has been proposed. Such photoluminescent QDs would absorb photons with energies outside of the UC response range before re-emitting UC-usable photons. Demonstrating this concept, photoluminescent PbS-QDs were dissolved together with UCs in a layer above the back reflector of a bifacial SSC, resulting in a 60% increase in photocurrent when compared to the incorporation of UC only^[130]. Photonic crystal nanostructures can positively impact the direction and wavelength of a photon undergoing UC. Experiments show that PbSe QDs with InP graphite lattice photonic crystals increased vertical emissions by a factor of 7.8, significantly increasing the amount of photons that reach the cell after UC^[131].

DSSCs sensitized with ruthenium-based dyes such as N719 or N749 possess bandgap energies exceeding 1.8 eV (corresponding to a wavelength of 700 nm), leading to approximately 52% of the solar energy in the incident solar spectrum not being harvested by the absorber material due to sub-bandgap transmission^[129]. Consequently, considerable research efforts have been directed toward incorporating UC materials into DSSCs. One such UC material extensively investigated in contemporary research is trivalent erbium (Er³⁺). By doping the TiO₂ membrane of an investigated DSSC with 1 wt% NaYF₄: Yb³⁺/Er³⁺ nanorods annealed at 500 °C, a substantial PCE enhancement of 95.58% from 3.17% to 6.2% was achieved^[132]. Besides the photoluminescent processes, the authors attribute this performance increase to improved charge separation facilitated by the nanorods.

The bandgap of PSCs is ~1.5 eV, corresponding to near-infrared light with a wavelength of 800 nm. Hence, similar to DSSCs, UC is a promising approach for the development of ultra high-efficiency PSCs. For UC-integrating PSCs under unconcentrated (one sun) AM1.5G illumination a limiting efficiency of 47.0% was calculated^[127], corresponding to a relative increase of 42.8% compared to the limiting efficiency of 32.91% for the same devices without UC NP^[133]. Experimental studies on UC technologies incorporated into PSCs have shown promising PCE enhancements. For example, by mixing high-fluorescence NaYbF₄: Ho³⁺ UC NPs into ZrO₂ paste at 40 wt%, drying, annealing, and finally incorporating the resulting film as a scaffold layer in a FA_0.4MA_0.6PbI₃ PSC under unconcentrated (one sun) AM1.5G irradiance, the device's PCE was increased by 28.8% from 11.12%, resulting in a UC-enhanced PSC of 14.32%^[134]. It is important to note that this PCE enhancement cannot be solely attributed to UC processed, but also to a decrease in recombination rate and trap-state density, as well as accelerated charge transfer and extraction. Further electrical characterizations of DSSCs and PSCs incorporating a variety of UC materials reported in recent literature yielding notable PCE enhancements are summarized in Table 7.

Downconversion

Normally, high energy photons above the cell's bandgap energy would potentially cause significant thermalization losses due to the waste of excess energy contained in hot carriers. Aiming to mitigate these thermalization losses, DC layers transform such high energy photons into two lower energy photons tailored to the cell's bandgap energy prior to absorption^[143]. Schematic DC photoluminescent processes comprising excitation and emission are visualized in a band diagram in Figure 11, where E_g is the solar cells bandgap energy and the energy difference between each respective energy level in the DC material. Because of the 1-to-2 conversion ratio between absorbed and emitted photons, the quantum efficiency of the DC process cannot exceed 200%^[144].

Figure 11. Band diagrams for downconversion process, showing the conversion of one high-energy photon into two photons of lower energy tuned to the employed solar cell's bandgap.

Common approaches for facilitating DC in PV devices include doping its antireflective coating layer (commonly made of ethyl vinyl acetate (EVA)) at the top surface with DC particles, incorporating the DC particles into the device's glass cover sheet, embedding DC core-shell particles into a cell's photoanode, or inserting individually function DC layers between photoanode and absorber layers^[123]. In addition to PCE enhancements through mitigation of thermalization losses, DC layers placed at the front surface of a PV device based on materials such as Tb³⁺-Yb³⁺, employed e.g. as co-dopants in silicon nitride matrix layers, may consequently further improve the device's absorption characteristics through beneficial anti-reflective coating characteristics^[16].

Calculations have shown a maximum efficiency of up to 38.6% for a solar cell with a bandgap energy of 1.1 eV incorporating a downconverting layer located on the front surface of the solar panel^[143]. Typical DC mechanisms include quantum cutting using host lattice states, quantum cutting on single rare-earth ions, or DC using rare-earth ion pairs^[123]. In quantum cutting using host lattice states describes, a high-energy UV photon excites the host lattice. The energy is then transferred to luminescent centers, where two photons in the visible or infrared range are emitted. Quantum cutting on single rare-earth ions typically involves sequential energy transitions within the ion's multiple excited states. A UV photon excites a single rare-earth ion to a high-energy state, which then relaxes in two steps, emitting two photons with lower energies. Finally, DC using rare-earth ions involves the excitation of a donor ion, which then transfers part of its energy to a neighboring acceptor ion, resulting in the emission of lower energy photons by both ions.

Downconverting materials have been applied to a wide range of PV technologies, including SSCs, DSSCs, or QDSCs^[145]. DC is commonly facilitated by lanthanide and transition-metal ion activated materials. The absorption range is broadened by introducing sensitizers with spin-allowed transitions and by combining the advancing perovskite QDs, nanostructures, and organic dye molecules^[146].

Remarkable PCE enhancements in SSCs facilitated by luminescent DC processes have been reported in contemporary research. By installing a hydrothermally treated vinyl acetate ethylene (VAE) coating with 0.6 wt% dispersed carbon QDs on the top surface of a SSC, a remarkable PCE increase from 6.5% to 8.1% (corresponding to a 24.6% increase) was achieved under unconcentrated AM1.5G irradiance produced by a solar simulator^[147].

Substantial PCE enhancements have also been reported in recent literature where DC technology is employed in PSCs. For example, the embedment of ZnGa₂O₄: Eu³⁺ nanophosphor particles into the TiO₂ electrode of an investigated PSC increased its initial bare-cell PCE of 10.67% to 13.8% under unconcentrated (one sun) AM1.5G irradiance, which corresponds to a relative PCE enhancement of 29.33%^[145]. This PCE enhancement is accompanied by a strong increase in short-circuit current from 20.12 mA/cm² of the bare cell to 23.68 mA/cm² of the nanophosphor-enhanced cell. In addition to the positive impact of DC, the authors attribute the reported performance enhancement to improved exciton generation rates. The cell retained roughly 66% of its initial efficiency when stored for nine days in ambient air between 21-23 °C at a relative humidity of 40%-45%.

Finally, by embedding YVO₄: Eu³⁺, Bi³⁺@SiO₂ core-shell particles at a 10% weight ratio within a nanostructured TiO₂ layer, substantial PCE enhancements were reported for an investigated DSSC. In addition to facilitating DC, the particles, which were prepared using the sol-gel method, lead to notably enhanced light scattering, extending the path light takes within the cell and thus effectively increasing the probability of absorption. The combined effects of light scattering and DC led to a significant relative increase in PCE of 64% from the initial 3.6% of the bare cell to 5.9% under simulated, unconcentrated (one sun) AM1.5G irradiance.

Emerging research trends emphasize the development of hybrid nanomaterials which merge UC and DC technology, enabling simultaneous conversion of both near-infrared and UV photons while emitting visible light tailored to a cell's absorption range. UC/DC hybrid nanomaterials are commonly lanthanide-doped materials. For example, employing NaYF₄: Yb³⁺/ Er³⁺@ NaYF₄: Eu³⁺ core-shell hybrid UC/DC NPs in a N719-sensitized DSSCs enhanced the device 's PCE from 6.726% to 7.664% under unconcentrated (one sun) AM1.5G illumation, corresponding to a relative increase of 13.95%^[148]. A portfolio of recent electrical characterizations reported in experimental studies on DC technologies employed in SSCs, PSCs, and DSSCs is presented in Table 8.

Downshifting

In the DS process, a high energy photon is absorbed by the DS material followed by the emission of one lower energy photon at a longer wavelength (typically in the visible/near-infrared region). Therefore, similar to DC, DS aims to mitigate thermalization losses caused by high energy photons with unusable excess energy beyond the bandgap of a PV absorber material. Similar to the UC and DC photoluminescent processes, efficiency enhancements can be achieved through DS since incoming high-energy photons are tailored to the cell's absorption characteristics prior to reaching the cell's photoconductive layers^[161].

The DS process consists of the initial electron excitation, followed by a two-step sequential relaxation process, as visualized in the bandgap diagram in Figure 12. The first relaxation step is non-radiative from the top energy level to the intermediate band of the DS material. During the second, radiative relaxation, a photon of the bandgap energy E_g is emitted for absorption by the solar cell.

Figure 12. Band diagram for DS process visualizing the conversion of one high-energy photon into one low-energy photon tuned to the employed solar cell's bandgap.

Since only one photon is emitted for each absorbed photon, the quantum efficiency of a DS process followed by ideal abosrption cannot exceed 100% (one excited electron per one incident photon)^[144].

By fabricating two SiO₂ layers containing unique DS-facilitating phosphors on the top surface of a SSC via a spin-on coating process, a pronounced increase in EQE in the spectral range between 350 and 1,000 nm from 52.93% for the bare cell to 68.17% for the DS-enhanced cell was obtained^[162], indicative of the ability of DS to widen the cell's effective light harvesting range. This EQE enhancement manifests itself in the form of a substantial 28.91% increase in short-circuit current and 31.63% PCE enhancement (from 11.14% to 14.66%). While the authors attribute this PCE enhancement exclusively to the DS phosphors, the integration of DS technology into host materials commonly facilitates a range synergetic effects beneficial to the PV device's performance. For example, by combining plasmonic scattering effects of silver NPs with the luminescent DS effects of Eu-doped phosphor, the PCE of an investigated SSC was increased from 11.31% to 14.27%, which corresponds to a substantial relative PCE enhancement of 26.17%^[163].

DS may also be integrated into multipurpose textured anti-reflection films^[164]: incorporating (Ba, Sr)₂SiO₄: Eu²⁺ micron phosphor as a DS material into textured polydimethylsiloxane films at the front of a perovskite/silicon tandem solar cell yielded PCE enhancement through combined effects of the antireflective control for the top cell, enhanced light trapping, and DS benefits. A 14.9% relative PCE enhancement from 20.1% to 23.1% was observed. The antireflective film alone raised the efficiency from 20.1% to 22.3%, which was then further increased by another 3.6% through the DS phosphors.

Recent literature has investigated hybrid nanomaterials merging UC and DS. In a recent 2025 study, a pronounced PCE enhancement of 22.19% was achieved in a PSC through simultaneous and spatially selective doping of Er³⁺ and Eu³⁺ lanthanides into the core and shell of nanomaterials^[165]. The resulting hybrid NaGdF₄: Yb³⁺, Er³⁺@NaGdF₄: Eu³⁺ NPs were incorporated in the PSC as an interfacial layer, facilitating both UC and DS. The authors attribute this improvement to a highly efficient DS process and reduced nonradiative recombination, while pointing out that the contribution of UC requires further confirmation.

Table 9 summarizes additional electrical characterizations reported in experimental studies of DS technology employed in SSCs, showing the potential of DS to significantly enhance the PCE of fabricated PV devices.