Multi-junction solar cell
A multi-junction solar cell is a photovoltaic device consisting of multiple p-n junctions fabricated from different semiconductor materials with distinct bandgap energies, stacked monolithically in series to absorb complementary portions of the solar spectrum and minimize energy losses from thermalization and transmission, thereby achieving significantly higher power conversion efficiencies than single-junction cells.[1][2] These cells, primarily based on III-V compound semiconductors such as gallium indium phosphide (GaInP), gallium arsenide (GaAs), and germanium (Ge), were first conceptualized in the mid-20th century but gained prominence in the late 1970s for space applications due to their high efficiency and radiation resistance.[2] The stacking design employs tunnel junctions to electrically connect subcells, ensuring current matching across layers while allowing each to target specific wavelengths: wider-bandgap top layers absorb high-energy blue and green light, while narrower-bandgap bottom layers capture infrared photons.[1][3] Laboratory efficiencies for multi-junction cells have surpassed 40% under concentrated sunlight as of 2025, with a six-junction inverted metamorphic design reaching 47.1% at 143 suns in 2020 and a four-junction design achieving a record 47.6% at 665 suns in 2022, far exceeding the Shockley-Queisser limit of ~29% for single-junction silicon cells.[1][2][4] Non-concentrating (one-sun) efficiencies stand at 39.2% for six junctions and a record 39.5% for triple-junction cells as of 2022.[2][3][5] These performance gains stem from advanced fabrication techniques like metalorganic vapor-phase epitaxy (MOVPE) and molecular beam epitaxy (MBE), which enable precise lattice-matched growth of over 100 layers in devices with more than five junctions.[1] Beyond space power systems for satellites, multi-junction cells are deployed in concentrator photovoltaics (CPV) for terrestrial utility-scale generation, where lenses or mirrors focus sunlight to leverage their high efficiency under intense illumination.[2] Emerging applications include photoelectrochemical hydrogen production and tandem integrations with silicon or perovskites to reduce costs and approach theoretical limits exceeding 50% for hybrid designs.[1][3] However, challenges such as high material costs (often >$10/W) and fabrication complexity persist, prompting research into substrate reuse, silicon-compatible growth, and scalable deposition for broader commercialization.[3][2]Basic Principles
Photovoltaic Effect
The photovoltaic effect refers to the generation of an electric current in a material upon exposure to light, a phenomenon first observed in 1839 by French physicist Alexandre-Edmond Becquerel while experimenting with an electrolytic cell consisting of platinum electrodes in an electrolyte solution.[6] Becquerel noted that the cell produced a voltage increase when illuminated, laying the groundwork for light-to-electricity conversion, though early devices were inefficient and limited to liquid-based systems.[6] In solid-state semiconductors, the effect relies on the absorption of photons by the material, where each photon's energy E is given by E = h \nu, with h as Planck's constant and \nu as the frequency of the light.[7] For absorption to occur and generate charge carriers, the photon energy must exceed the semiconductor's bandgap energy E_g, the minimum energy difference between the valence band and conduction band that allows an electron to transition from a bound state to a free state.[8] When a suitable photon is absorbed, it excites an electron from the valence band to the conduction band, creating an electron-hole pair: the electron in the conduction band and a positively charged hole left in the valence band.[9] This process becomes electrically useful in a p-n junction, formed by doping one side of the semiconductor with acceptors (p-type, hole-rich) and the other with donors (n-type, electron-rich), which establishes a built-in electric field at the junction due to charge diffusion.[10] The field sweeps the photogenerated electrons toward the n-side and holes toward the p-side, preventing recombination and driving a net photocurrent through an external circuit when the cell is connected to a load.[11] The first practical solid-state photovoltaic cell, achieving about 6% efficiency, was developed in 1954 by Daryl Chapin, Calvin Fuller, and Gerald Pearson at Bell Laboratories using a silicon p-n junction. Multi-junction solar cells extend this principle by stacking multiple p-n junctions with different bandgaps to capture a broader range of the solar spectrum.[6]Single-Junction Limitations
Single-junction solar cells, which rely on a single p-n junction with a fixed bandgap energy E_g, face fundamental limitations due to the mismatch between the broad solar spectrum and the narrow energy range that the cell can effectively convert to electricity. Photons with energies below E_g are not absorbed and pass through the material as transmission losses, while those with energies above E_g generate electron-hole pairs, but the excess energy beyond E_g is rapidly dissipated as heat through thermalization, a process governed by carrier relaxation to the band edges. These losses are intrinsic to the photovoltaic effect in a single absorber material and significantly reduce the potential efficiency, as the solar spectrum under standard AM1.5 conditions spans from ultraviolet to infrared wavelengths with varying intensities.[12][13] The Shockley-Queisser limit provides a theoretical upper bound on the efficiency of an ideal single-junction cell by applying the principle of detailed balance, which equates the absorption of photons to the emission of blackbody radiation from the cell under equilibrium. In their seminal 1961 analysis, Shockley and Queisser assumed a step-function absorptivity (perfect absorption above E_g, none below), negligible non-radiative recombination, and the sun modeled as a 6000 K blackbody source, leading to an ultimate efficiency calculation that accounts for the trade-off between short-circuit current density J_{sc} (maximized by lower E_g to capture more photons) and open-circuit voltage V_{oc} (maximized by higher E_g to reduce thermal emission). The derivation involves integrating the photon flux above E_g for J_{sc} = q \int_{E_g}^\infty \phi(E) dE, where \phi(E) is the spectral photon flux, and V_{oc} = (kT/q) \ln(J_{sc}/J_0 + 1), with J_0 as the radiative saturation current derived from the cell's emission. Optimizing E_g at approximately 1.34 eV yields a maximum efficiency of about 33.7% under unconcentrated AM1.5 illumination, incorporating a fill factor FF close to 0.89 for the ideal case; this limit arises primarily from thermalization (around 30% of incident energy) and transmission (about 20%), plus radiative recombination losses.[13][14] In practice, silicon single-junction cells with E_g \approx 1.12 eV achieve efficiencies well below their SQ limit of ~29%, with laboratory records reaching 27.81% as of 2025, primarily due to non-radiative recombination at defects and surfaces, series resistance from contacts and doping, and shading from front metallization that blocks incident light. These parasitic effects reduce V_{oc} and FF, with series resistance alone causing voltage drops that can lower efficiency by 1-2% in high-performance cells, while shading and incomplete light absorption further compound the gap to theoretical ideals. Multi-junction cells mitigate these spectral mismatch losses by employing multiple bandgaps to better utilize the full solar spectrum.[15][16]Multi-Junction Design
Operating Principle
Multi-junction solar cells function by vertically stacking multiple semiconductor p-n junctions, each with a distinct bandgap, to capture a broader portion of the solar spectrum than a single-junction cell. Sunlight enters from the top, where the junction with the widest bandgap absorbs high-energy (short-wavelength) photons, generating electron-hole pairs while transmitting lower-energy photons to the underlying junctions. This spectrum-splitting approach minimizes thermalization losses, where excess photon energy is dissipated as heat in single-junction devices, and reduces the impact of below-bandgap photons that cannot be absorbed.[17][3] The subcells are electrically connected in series through highly doped tunnel junctions, which provide a low-resistance pathway for current flow by enabling quantum tunneling of carriers between adjacent layers. This series configuration results in the total open-circuit voltage being the sum of the individual subcell voltages, expressed as V_{\text{total}} = \sum V_i, where V_i is the open-circuit voltage of the i-th junction. In contrast, the photocurrent is constrained by the subcell producing the lowest current density, as the series connection requires equal current through all layers to avoid bottlenecks; thus, the total current density is J_{\text{total}} = \min(J_1, J_2, \dots, J_n), with J_i denoting the short-circuit current density of each subcell. Current matching is achieved through careful design of layer thicknesses and bandgap selections to balance photon absorption across the stack.[17][3] Typical multi-junction cells incorporate 2 to 6 junctions, balancing efficiency gains against fabrication complexity. A representative example is the triple-junction configuration using GaInP, GaAs, and Ge layers, where the top GaInP junction targets blue and green light, the middle GaAs absorbs red, and the bottom Ge captures infrared, achieving combined spectrum utilization that exceeds single-junction limits.[17][3]Bandgap Selection
Bandgap selection in multi-junction solar cells aims to partition the solar spectrum into segments that each subcell can absorb efficiently, minimizing spectral overlap where higher-energy photons are wasted in lower-bandgap junctions and transmission losses where photons pass unabsorbed through upper junctions. For terrestrial applications, the AM1.5 global spectrum is targeted, while space environments require optimization for the AM0 spectrum, both emphasizing broad coverage from ultraviolet to near-infrared wavelengths (approximately 300–1800 nm) with graded bandgaps decreasing from top to bottom junctions.[17] Numerical models, often based on detailed balance principles adapted for multi-junction configurations, are used to optimize bandgap combinations by maximizing the overall short-circuit current while balancing voltage contributions across subcells. For an ideal three-junction cell under the AM1.5 spectrum, optimal bandgaps are approximately 1.9 eV (top), 1.4 eV (middle), and 0.95 eV (bottom), yielding a theoretical efficiency of around 50% under one-sun illumination. These values ensure each junction captures a distinct portion of the spectrum, with the top cell absorbing high-energy blue-green photons, the middle targeting yellow-red, and the bottom utilizing infrared.[18] A key trade-off in bandgap selection involves lattice matching, which promotes low defect densities and high minority carrier lifetimes for superior performance but restricts combinations to materials with compatible lattice constants, such as the GaAs-based system. To access non-matched bandgaps for better spectral partitioning, metamorphic growth techniques introduce gradual lattice relaxation through buffer layers, enabling higher efficiencies (e.g., up to 46% in inverted metamorphic designs) despite potential increases in dislocation densities that can degrade long-term stability.[17][19] Historically, dual-junction cells evolved from GaAs (1.4 eV)/Ge (0.67 eV) configurations in the 1980s, achieving around 20–25% efficiency for space applications by combining GaAs's high performance with Ge's substrate utility. The addition of a wide-bandgap layer, such as GaInP (∼1.9 eV), formed the dominant triple-junction stack (GaInP/GaAs/Ge) by the 1990s, boosting efficiencies to over 30% under AM0. Further evolution to quadruple-junction designs in the 2010s incorporated an intermediate ∼1.0 eV layer (e.g., dilute nitride GaInNAs), pushing efficiencies toward 40% by finer spectrum splitting.[20] Sensitivity analyses reveal that small bandgap deviations significantly impact efficiency; for instance, a ±0.1 eV shift in the bottom junction bandgap from its optimal value (∼0.9 eV) can reduce overall efficiency by 2–5% in three- to six-junction cells under AM1.5, primarily due to current mismatch and unabsorbed infrared photons. Top-junction bandgaps above 2.5 eV show reduced sensitivity, allowing more flexibility in material selection without substantial losses. Materials like GaInP are commonly used to realize the ∼1.9 eV top bandgap in lattice-matched stacks.[21][20]Structural Components
Tunnel Junctions
In multi-junction solar cells, tunnel junctions serve as critical interconnects between adjacent subcells, functioning as highly doped p-n junctions that facilitate quantum mechanical tunneling of charge carriers. This enables a low-resistance series electrical connection, allowing the subcells to operate in series without incurring a significant voltage drop across the junction, thereby preserving the overall photocurrent matching and efficiency of the stack.[22] The design relies on degenerate doping levels to create overlapping valence and conduction bands, promoting efficient carrier recombination and transport.[23] The physics of these tunnel junctions is rooted in the behavior of an Esaki diode, where band-to-band tunneling dominates due to the heavy doping that narrows the depletion region and aligns the band edges for carrier overlap. Tunneling probability is theoretically described by the Wentzel-Kramers-Brillouin (WKB) approximation, which accounts for the exponential decay of the wavefunction through the potential barrier, but in practice, it is often simplified to models emphasizing high carrier density overlap for direct estimation in device simulations. Kane's model further refines this by incorporating effective masses and bandgap energies to predict peak tunneling currents, typically exceeding those required for multi-junction operation (around 10-20 mA/cm²).[22][24] Materials for tunnel junctions are predominantly III-V compound semiconductors, with GaAs-based structures being traditional choices due to their compatibility with epitaxial growth processes like metal-organic chemical vapor deposition (MOCVD). These junctions feature n⁺⁺ and p⁺⁺ regions doped to concentrations greater than $10^{19} cm^{-3}, often reaching $10^{20} cm^{-3}, using dopants such as tellurium or silicon for n-type and carbon or zinc for p-type to achieve low resistivity and high tunneling rates.[22] In modern designs, InGaP or AlGaAs variants replace pure GaAs to improve lattice matching with overlying subcells and enhance optical transparency.[22] A key challenge in tunnel junction design is minimizing optical absorption losses, as the heavily doped regions can absorb photons intended for lower-bandgap subcells beneath them, reducing short-circuit current. To address this, engineers employ high-bandgap materials like AlGaAs or InGaP for transparency to relevant wavelengths (>700 nm), along with graded doping profiles that progressively vary composition and dopant concentration to reduce free-carrier absorption while maintaining electrical performance.[22][25] The evolution of tunnel junctions traces back to the early 1980s, when the first monolithic multi-junction solar cell incorporated an AlGaAs/GaAs tunnel junction to enable tandem operation, marking a shift from discrete cell assemblies. Over subsequent decades, designs progressed from simple GaAs homojunctions, which suffered from higher absorption, to heterostructure InGaP/GaAs systems in the 1990s for space applications, and further to advanced AlGaAs/InGaP configurations with quantum wells by the 2010s, optimizing for terrestrial concentrator cells with efficiencies exceeding 40%.[22][22]Anti-Reflective Coatings
Anti-reflective coatings (ARCs) are essential for multi-junction solar cells to minimize optical losses at the air-semiconductor interface, where Fresnel reflection occurs due to the significant refractive index mismatch between air (n ≈ 1) and typical III-V semiconductors (n ≈ 3.5–4.0). For normal incidence, the reflectivity R is given by the Fresnel equation:R = \left| \frac{n - 1}{n + 1} \right|^2
This results in approximately 30% reflection loss without mitigation, substantially reducing the light available for absorption in the cell stack.[26][27] Multi-layer ARCs, typically consisting of quarter-wave stacks with alternating layers of high and low refractive index materials, are widely employed to counteract this reflection across the relevant spectrum. Common configurations include double-layer designs such as TiO₂ (high index, n ≈ 2.4) over SiO₂ (low index, n ≈ 1.46), deposited via techniques like electron-beam evaporation or sputtering to achieve optical thicknesses of λ/4 at a central wavelength. These stacks create destructive interference for reflected waves, effectively reducing reflectivity to below 5% in targeted bands. Triple- or quadruple-layer variants, such as TiO₂/SiO₂/Si₃N₄, further optimize performance by addressing multiple wavelengths simultaneously.[28][29] For multi-junction cells operating over a broad wavelength range (300–1800 nm), advanced designs like graded-index (GRIN) or moth-eye structures are implemented to provide ultra-broadband antireflection. GRIN coatings feature a gradual variation in refractive index, often using nanoporous SiO₂ or SiOₓNᵧ layers, achieving average reflectivities as low as 2–3% across 300–1800 nm and enabling efficiency gains of up to 86% in reflection reduction compared to uncoated surfaces. Moth-eye structures, inspired by biological nanostructures, involve subwavelength surface gratings (e.g., etched into the top GaInP layer) that mimic a tapered index profile, suppressing reflection to under 5% over wide angles and spectra in triple-junction devices like GaInP/GaAs/Ge. These designs interact briefly with the top junction's absorption but primarily enhance overall photon entry. The effectiveness of these ARCs is evident in practical implementations, where reflection losses are curtailed from ~30% to less than 5% over the visible and near-infrared regions, contributing 2–4% absolute efficiency improvements in multi-junction cells under AM1.5G illumination. However, trade-offs exist in durability: space applications demand radiation-hard materials like Al₂O₃ or Si₃N₄ to withstand atomic oxygen and high-energy particles, often prioritizing mechanical stability over self-cleaning properties, whereas terrestrial environments favor hydrophobic TiO₂-based coatings for dust resistance and longevity under humidity and UV exposure.[28][31][32]
Contacts and Passivation Layers
In multi-junction solar cells, electrical contacts are essential for efficient carrier collection while minimizing optical and resistive losses. Front contacts typically employ fine metal grid patterns, such as Ti/Pt/Au stacks, to balance current collection with reduced shading of the active area. These grids, often with finger widths around 2-5 μm and spacing optimized for the cell size, limit light blockage to less than 3% of the surface. The shading factor, which quantifies the optical loss, is approximately equal to the fractional grid coverage area f, resulting in a short-circuit current reduction of \Delta J_{sc} \approx f \cdot J_{sc,0}, where J_{sc,0} is the unobstructed current density. Back contacts, usually a continuous metal layer like Au or Ag, provide low-resistance ohmic interfaces to the substrate or bottom subcell without shading concerns. Ohmic contacts in these devices require low specific contact resistance, typically \rho_c < 10^{-6} \, \Omega \cdot \mathrm{cm}^2, to avoid series resistance losses that degrade fill factor. This is achieved through alloying or annealing processes; for instance, AuBe/Ni/Au contacts to n-type InGaP exhibit \rho_c \approx 5 \times 10^{-6} \, \Omega \cdot \mathrm{cm}^2 after annealing at 420°C. Such low-resistance interfaces ensure efficient extraction of photocurrent from each subcell, particularly under high-concentration illumination where current densities exceed 10 A/cm². Passivation layers play a critical role in suppressing surface recombination velocities, which can otherwise limit open-circuit voltage in the subcells. Window layers, such as AlInP with a wide bandgap of ~2.2 eV, are deposited on the front surface of emitters to create a low-recombination heterojunction interface, reducing non-radiative losses at the top subcell. Similarly, back-surface fields (BSF) employ wide-bandgap materials like p+-GaInP or AlInAs, heavily doped to form a potential barrier that reflects minority carriers away from the rear contact, enhancing collection efficiency. These strategies have enabled surface recombination velocities below 100 cm/s in GaAs-based subcells. A key challenge in multi-junction architectures arises from the thin absorber layers (often <1 μm thick), which necessitate effective lateral current flow to reach the grid lines and minimize series resistance. Poor lateral conductivity in these undoped or lightly doped regions can lead to voltage drops exceeding 10 mV across the cell, particularly in large-area devices. Recent advances incorporate transparent conductive oxides (TCOs), such as indium tin oxide (ITO), as front or rear electrodes to reduce the required metal grid area by up to 50%, thereby lowering shading losses while maintaining sheet resistances below 100 Ω/sq. This approach has improved overall cell performance in III-V multi-junctions by enhancing light transmission and carrier transport.Electrical Characteristics
Current-Voltage Behavior
The current-voltage (J-V) characteristics of multi-junction solar cells arise from their series-connected subcells, where the total photocurrent is limited by the subcell generating the lowest current density under illumination, a condition known as current matching. In an ideal current-matched configuration, the overall J-V curve resembles that of a single diode but with the open-circuit voltage summed across junctions, while the short-circuit current density J_{sc} is determined by the minimum J_{ph,i} of the subcells. If subcell currents are mismatched due to spectral variations or fabrication tolerances, the J-V curve exhibits characteristic steps or kinks, reflecting transitions between current-limiting subcells.[33][34] The short-circuit current density J_{sc} is thus set by the current-matched value, typically verified through external quantum efficiency (EQE) spectra that isolate each subcell's response to monochromatic light under bias conditions to account for optical coupling. The open-circuit voltage V_{oc} is the sum of individual subcell contributions, approximated as V_{oc} = \sum_i \frac{kT}{q} \ln \left( \frac{J_{sc}}{J_{0,i}} \right), where J_{0,i} is the saturation current density of the i-th junction, k is Boltzmann's constant, T is temperature, and q is the elementary charge. The fill factor (FF), a measure of curve squareness, is defined as FF = \frac{J_{mp} V_{mp}}{J_{sc} V_{oc}}, where J_{mp} and V_{mp} are the current density and voltage at the maximum power point; deviations from ideality reduce FF.[34][35] Illuminated J-V measurements are performed under standard spectra such as AM1.5G for terrestrial applications or AM0 for space, using solar simulators to ensure 1000 W/m² irradiance at 25°C. EQE spectra for individual subcells are obtained via bias light methods to suppress contributions from non-limiting junctions, enabling precise J_{sc} calculation by integrating over the solar spectrum. Non-ideal behaviors, such as rollover in the J-V curve at high voltages, stem from series resistance effects, where voltage drops across tunnel junctions or contacts limit current, particularly under concentration.[33][35]Efficiency Definitions
The power conversion efficiency (η) of a multi-junction solar cell is defined as the ratio of the maximum electrical power output (P_max) to the incident optical power (P_in), expressed as a percentage: η = (P_max / P_in) × 100%. P_max is determined from the current-voltage (J-V) curve under standard test conditions, typically at the maximum power point where the product of current density and voltage is optimized. For terrestrial applications, P_in is standardized at 1000 W/m² under the AM1.5G spectrum, which represents global solar irradiance on a tilted surface at sea level with the sun 48.2° above the horizon, accounting for both direct and diffuse components as per ASTM G173.[36] For space applications, efficiency is measured under the AM0 spectrum, which simulates the unfiltered extraterrestrial solar irradiance of approximately 1366 W/m², as defined by ASTM E490, to reflect conditions outside Earth's atmosphere. In concentrator systems, efficiencies are evaluated under N suns, where the incident power scales to 1000 × N W/m² using the AM1.5D direct-beam spectrum, enabling higher performance due to increased photon flux but requiring enhanced thermal management.[37] Key metrics beyond overall efficiency include external quantum efficiency (EQE) and internal quantum efficiency (IQE), which quantify wavelength-dependent performance. EQE is the ratio of collected charge carriers to incident photons at a given wavelength, incorporating losses from reflection and incomplete absorption, while IQE measures carriers per absorbed photon, excluding optical losses. The spectral response, often plotted as EQE versus wavelength, reveals how each subcell contributes to current generation across the solar spectrum, with current matching optimized to limit the overall cell's short-circuit current to the minimum subcell value.[38] Efficiencies are reported as either initial (measured immediately after fabrication or annealing) or stabilized (after prolonged operation to account for light-induced degradation). Stabilized values are critical for practical assessments, as initial efficiencies may degrade by 1-5% due to factors like interface recombination or material metastability under illumination and heat. For benchmarking, the National Renewable Energy Laboratory (NREL) verifies records; for example, a six-junction inverted metamorphic III-V cell achieved a verified 39.2% efficiency under 1-sun AM1.5G conditions in 2020, representing a milestone for metamorphic designs.[39] In 2022, a triple-junction III-V cell reached 39.5% under the same conditions, setting a new record for one-sun multi-junction efficiencies.[5]Theoretical Efficiency Limits
Detailed Balance Model
The detailed balance model provides a thermodynamic framework for calculating the ultimate efficiency limits of solar cells by equating the absorption and emission of photons under equilibrium conditions.[13] Developed by William Shockley and Hans-Joachim Queisser in 1961, the model assumes that the solar cell operates as a blackbody radiator, absorbing all incident photons with energy above the bandgap while emitting photons only through radiative recombination, neglecting non-radiative losses.[13] Under these assumptions, the short-circuit photocurrent density J_{sc} is determined by the integral of the external quantum efficiency (EQE) weighted by the incident photon flux \Phi(\lambda): J_{sc} = q \int_0^\infty \text{EQE}(\lambda) \Phi(\lambda) \, d\lambda where q is the elementary charge.[13] For an ideal cell, EQE approaches 1 for photon energies exceeding the bandgap E_g and 0 otherwise, limiting J_{sc} to the flux of absorbable photons from the solar spectrum. The dark saturation current density J_0 arises solely from radiative recombination and is calculated from the blackbody emission spectrum above E_g, given by: J_0 = q \int_{E_g}^\infty \frac{2\pi (E^2 / h^3 c^2)}{e^{E / kT} - 1} \, dE where k is Boltzmann's constant, T is the cell temperature, h is Planck's constant, and c is the speed of light.[13] The open-circuit voltage V_{oc} is bounded by the relation V_{oc} < E_g / q - (kT / q) \ln(\text{constant}), where the logarithmic term accounts for the ratio of photocurrent to dark current, leading to an ultimate efficiency of approximately 33% for a single-junction cell under AM1.5 illumination.[13] This model extends to multi-junction solar cells by treating each sub-cell as an independent absorber with its own bandgap, subject to the same detailed balance principles, while requiring current matching across sub-cells to maximize overall performance under series connection.[40] Such configurations allow efficiencies exceeding the single-junction limit by partitioning the solar spectrum, with detailed balance applied to optimize bandgap selections as explored in subsequent analyses.[40]Optimal Configurations
The optimal configurations for multi-junction solar cells are derived from the detailed balance model by optimizing bandgap combinations to achieve current matching while maximizing the absorption of the solar spectrum and minimizing thermalization and transmission losses. These configurations assume ideal conditions, including radiative recombination only, perfect tunnel junctions, and no parasitic absorption or reflection. For finite numbers of junctions, efficiency increases with additional layers up to a point of diminishing returns, with 6-junction designs approaching the practical upper bound for most applications under standard illumination. For an infinite number of junctions, the detailed balance model predicts a thermodynamic limit of approximately 68% efficiency under 1-sun AM1.5G illumination, as the stack can in principle partition the spectrum into infinitesimally small bandgap steps to capture nearly all incident energy above the blackbody emission threshold. Under maximum concentration, this limit rises to about 86%, reflecting reduced entropy generation from the directional, high-flux photon input that suppresses non-radiative broadening effects.[41] For finite cases under 1-sun AM1.5G illumination, the theoretical maximum efficiencies are approximately 44% for 2-junction cells, 49% for 3-junction cells, and 56% for 6-junction cells, with further junctions offering marginal gains beyond six due to the finite spectral width. These values are achieved through bandgap grading that balances current in each subcell, typically using wider bandgaps for top cells to absorb high-energy photons and narrower ones for bottom cells to utilize longer wavelengths. Bandgap optimization varies slightly between terrestrial (AM1.5G) and space (AM0) spectra, as AM0 has a bluer, more uniform distribution requiring higher average bandgaps to avoid over-absorption in lower subcells. The following table summarizes optimal bandgap combinations and corresponding detailed balance efficiencies for selected configurations under AM1.5G and AM0 spectra (1 sun, assuming current-matched series connection):| Number of Junctions | Optimal Bandgaps (eV) for AM1.5G | Efficiency (%) AM1.5G | Optimal Bandgaps (eV) for AM0 | Efficiency (%) AM0 |
|---|---|---|---|---|
| 2 | 1.65 / 0.95 | 44 | 1.70 / 1.00 | 42 |
| 3 | 1.85 / 1.15 / 0.70 | 49 | 1.90 / 1.20 / 0.75 | 47 |
| 6 | 2.00 / 1.60 / 1.30 / 1.05 / 0.80 / 0.50 | 56 | 2.05 / 1.65 / 1.35 / 1.10 / 0.85 / 0.55 | 54 |
Materials and Substrates
III-V Compound Semiconductors
III-V compound semiconductors are a class of materials composed of elements from groups III and V of the periodic table, such as gallium, indium, aluminum, arsenic, phosphorus, and antimony, which form the backbone of high-performance multi-junction solar cells due to their optoelectronic properties.[17] These materials exhibit direct bandgaps, enabling strong light absorption in thin layers— for instance, gallium arsenide (GaAs) absorbs 90% of incident light within just 1 μm thickness, compared to over 100 μm required for silicon.[17] Bandgap energies can be precisely tuned through alloying, as in the quaternary system \text{Ga}_x \text{In}_{1-x} \text{As}_y \text{P}_{1-y}, allowing optimization for different spectral regions in multi-junction stacks.[44] Common alloys in multi-junction solar cells include gallium indium phosphide (GaInP) for the top subcell with a bandgap of approximately 1.8–1.9 eV, gallium arsenide (GaAs) for the middle subcell at 1.42 eV, and germanium (Ge) for the bottom subcell at 0.67 eV, forming the standard triple-junction configuration.[3] These alloys leverage the direct bandgap nature for efficient carrier generation and high absorption coefficients, with GaInP often lattice-matched to GaAs substrates for seamless integration.[17] The advantages of III-V compounds include exceptionally high minority carrier lifetimes and mobilities, which support long diffusion lengths and low recombination losses, contributing to elevated open-circuit voltages.[44] They also demonstrate superior radiation resistance, retaining up to 88% of initial efficiency after 15 years in space environments, making them ideal for satellite applications.[17] Additionally, their thermal stability allows operation under high temperatures with minimal degradation, as evidenced by less than 5% efficiency loss after 200 hours at 400°C.[17] Despite these benefits, III-V semiconductors face challenges such as high production costs, estimated at approximately $77 per watt for scaled manufacturing as of 2025, due to the need for sophisticated epitaxial growth processes and rare elements like indium and gallium.[45] Toxicity concerns arise from arsenic content, which poses biohazards and necessitates specialized handling and recycling protocols.[17] Historically, the development of III-V solar cells began with early GaAs single-junction devices in the 1970s, including their deployment in Soviet Lunokhod rovers in 1970 and 1972, marking a shift toward space-qualified photovoltaics.[46] The first significant multi-junction milestone came in 1988 with a GaAs-based tandem cell achieving 20% efficiency via improved tunnel junctions.[17] By the 1990s, GaInP/GaAs/Ge triple-junction cells surpassed 30% efficiency under concentrated light, establishing III-V technology as the standard for efficiencies exceeding 40% in modern devices.[3]Lattice-Matched and Metamorphic Structures
In multi-junction solar cells based on III-V semiconductors, lattice-matched structures are employed to minimize crystal defects by selecting materials whose lattice constants closely align with that of the substrate. This approach ensures high structural integrity and low dislocation densities, which are critical for maintaining carrier lifetimes and open-circuit voltages. For instance, GaAs subcells are typically grown directly on GaAs substrates, while Ge serves as a substrate for the bottom junction in triple-junction configurations like InGaP/GaAs/Ge, where the lattice mismatch between Ge and GaAs is only about 0.08%, allowing near-perfect matching across the stack.[47] Such configurations achieve efficiencies up to 40.1% under concentrated illumination due to reduced recombination losses.[47] Metamorphic structures, in contrast, accommodate larger lattice mismatches through the use of compositionally graded buffer layers, enabling greater flexibility in bandgap selection without being constrained by the substrate's lattice constant. These buffers, often made from alloys like GaInP or AlGaInAs, gradually vary in composition over a thickness of several micrometers to relax strain and confine dislocations away from active regions, resulting in threading dislocation densities as low as 10^6 cm^{-2}. A prominent example is the metamorphic InGaP/GaInAs/Ge triple-junction cell, where the middle GaInAs subcell is tuned to a 1.0 eV bandgap via approximately 2% lattice mismatch relative to the GaAs-like lattice, accommodated by a graded buffer; this design has demonstrated 40.7% efficiency under AM1.5 spectrum at 240 suns.[48][49] However, metamorphic growth introduces higher defect levels compared to lattice-matched designs, potentially increasing non-radiative recombination and reducing voltage by 30-60 mV per subcell.[47] Common substrates for these structures include GaAs, Ge, and InP, each offering distinct trade-offs. GaAs substrates provide superior crystal quality and low defect propagation, ideal for high-performance lattice-matched devices, but their high cost limits scalability.[3] Ge substrates are more cost-effective and mechanically robust, with the added benefit of serving as an active IR-absorbing bottom junction in multi-junction stacks, enabling four-junction configurations through metamorphic layers despite slightly elevated defect densities.[3][47] InP substrates support lattice-matched growth of higher-bandgap alloys like InGaAsP for quadruple-junction cells targeting efficiencies beyond 45%, but their higher density, brittleness, and expense pose challenges for large-area fabrication and handling.[50] To optimize light absorption while minimizing series resistance, the active layers in these subcells are typically 1-10 μm thick, with top-junction bases around 0.5 μm, middle junctions 2-4 μm, and bottom junctions up to 5-10 μm, depending on the material's absorption coefficient and current-matching requirements.[51] This thickness range ensures sufficient photon capture—e.g., over 90% internal quantum efficiency in GaAs layers—without excessive carrier transit times that could degrade performance under high illumination.[51]Emerging Hybrid Materials
Emerging hybrid materials in multi-junction solar cells integrate perovskites, quantum dots, and organics with traditional semiconductors to enhance efficiency while reducing costs. Wide-bandgap perovskites, typically with bandgaps of 1.6-1.8 eV, serve as top layers in tandems atop silicon or III-V substrates, enabling four or more junctions to capture a broader solar spectrum.[52] These hybrids leverage the tunable optoelectronic properties of perovskites, which can be solution-processed at low temperatures, contrasting with the high-cost epitaxial growth required for pure III-V stacks.[53] In perovskite-silicon tandems, a notable achievement is the 34.85% efficiency demonstrated by LONGi in April 2025, utilizing a wide-bandgap perovskite top cell on a silicon bottom cell under standard illumination; more recent progress includes a certified 33.6% efficiency for a flexible perovskite/crystalline silicon tandem in November 2025.[54][55][56] For perovskite-III-V configurations, simulations and experiments show promise; for instance, all-inorganic CsPbIBr₂/GaAs four-terminal tandems have reached 30.97% efficiency, benefiting from the high absorption of GaAs in the infrared while perovskites handle visible light.[57] Similarly, wide-bandgap perovskite/GaAs two-terminal tandems have improved efficiencies from 21.68% to 24.27%, with four-terminal variants achieving 25.19%, highlighting the potential for flexible, thin-film applications.[58] These hybrids offer advantages such as lower fabrication costs for perovskite layers and improved spectral utilization, potentially exceeding 45% efficiency in multi-junction setups.[52] However, challenges persist, including perovskite instability under operational conditions and lattice mismatch with III-V materials, which can introduce defects and reduce long-term performance.[59] Ongoing research addresses these through interface engineering and encapsulation to enhance durability.[60] Beyond perovskites, quantum dots enable intermediate band formation in multi-junction cells, allowing sub-bandgap photon absorption for higher current generation. In(Ga)As quantum dot intermediate band solar cells have been optimized to approach theoretical efficiencies by minimizing thermal losses and improving carrier extraction.[61] Organic materials, when hybridized in multi-junction architectures, provide flexible, low-cost options for wide-bandgap junctions, with numerical models showing potential for over 20% efficiency in organic-perovskite tandems through precise layer thickness control.[62] Projections indicate that these hybrid materials could play a key role in surpassing 50% efficiency by 2030, particularly in concentrated light applications, by combining the strengths of perovskites for cost-effective top junctions with III-V or silicon bases for robust bottom cells.[41]Fabrication Processes
Epitaxial Growth Techniques
Epitaxial growth techniques are essential for fabricating the multi-layer structures in multi-junction solar cells, enabling precise control over material composition, thickness, and doping to optimize light absorption across different bandgaps. These methods deposit thin films of III-V compound semiconductors, such as GaInP and GaAs, onto substrates like GaAs or Ge, ensuring high crystal quality and minimal defects for efficient charge carrier collection.[63] Metalorganic chemical vapor deposition (MOCVD), also known as metalorganic vapor phase epitaxy (MOVPE), is the most widely adopted technique for producing multi-junction solar cells due to its scalability and ability to achieve precise alloy compositions. In MOCVD, metalorganic precursors are decomposed in a hydrogen carrier gas at atmospheric or reduced pressure, allowing for uniform deposition over large areas. Typical growth temperatures range from 500°C to 700°C, with rates of 1-10 μm/hr, enabling the formation of complex heterostructures with doping levels controlled to 10^17-10^19 cm^{-3} for n-type and p-type layers using sources like silane and carbon. This method excels in industrial production, supporting wafer sizes up to 8 inches (200 mm), though challenges in multi-layer uniformity can affect yield.[63][64][65][66] Molecular beam epitaxy (MBE) operates in an ultra-high vacuum environment, where elemental beams from effusion cells are directed at a heated substrate to form epitaxial layers atom by atom, resulting in exceptionally sharp interfaces critical for high-performance junctions. Growth occurs at temperatures of 400-600°C with rates of 0.5-1.5 μm/hr, providing superior control over doping profiles for n/p layers via in-situ monitoring techniques like reflection high-energy electron diffraction. Primarily used in research settings for developing advanced multi-junction configurations, MBE's slow growth enables exploration of novel alloys but limits scalability compared to MOCVD, with typical wafer sizes around 4 inches and yield issues arising from vacuum constraints.[67][68] Hydride vapor phase epitaxy (HVPE), particularly the dynamic variant (D-HVPE), offers a cost-effective alternative for thick-layer deposition in multi-junction cells, leveraging chloride-based precursors for rapid growth without metalorganics. Operating at temperatures around 600-800°C, HVPE achieves rates exceeding 100 μm/hr—far higher than MOCVD or MBE—facilitating efficient production of buffer and contact layers with doping control for n/p junctions via precursor modulation. Its high material utilization and potential for in-line processing enhance scalability, supporting 4-6 inch wafers, though uniformity across multi-layers remains a yield challenge due to gas-phase reactions. NREL's D-HVPE systems have demonstrated single-junction efficiencies of 26% as of 2025, underscoring its promise for affordable III-V multi-junction devices.[69][70][17][71]Device Assembly and Testing
Following epitaxial growth, multi-junction solar cell wafers undergo a series of post-processing steps to define device structures and form electrical connections. Photolithography is employed to pattern the front contact grids and delineate the mesa regions, followed by reactive ion etching to create mesas that isolate individual cells by removing material down to the substrate or tunnel junctions.[72] This etching step ensures electrical isolation while preserving the stacked subcell architecture.[73] Ohmic contacts are then formed through metal evaporation techniques, typically using e-beam or thermal evaporation to deposit multilayer stacks such as AuGe/Ni/Au for n-type regions and AuBe/Au for p-type regions.[74] These contacts provide low-resistance interfaces to the semiconductor layers, with front grids designed to minimize shading losses while maximizing current collection.[75] Anti-reflective coatings (ARCs) are subsequently deposited, often via e-beam evaporation or sputtering, using multilayer dielectric stacks like TiO₂/SiO₂ or Al₂O₃/TiO₂ to reduce broadband reflection across the absorption spectrum of the subcells.[76] These ARCs enhance light transmission, with optimized designs achieving average reflectances below 5% over wavelengths from 300 to 1800 nm.[77] Completed wafers are diced into individual chips using mechanical sawing or laser scribing to separate devices without damaging the delicate III-V layers.[78] For terrestrial applications, chips are assembled into modules through soldering to metal frames and encapsulation with polymers like ethylene vinyl acetate (EVA) or silicone to protect against moisture, mechanical stress, and UV degradation, ensuring long-term durability.[79] In space applications, cells are often mounted on coverslips with adhesive or direct bonding, followed by minimal encapsulation to maintain thermal conductivity while shielding from atomic oxygen and radiation.[3] Device performance is rigorously tested to verify functionality and identify issues. External quantum efficiency (EQE) is measured for each subcell using a monochromatic light source, with appropriate bias illumination applied to the other subcells to isolate their contributions and prevent spectral crosstalk.[80] This technique reveals subcell current matching and wavelength-specific responsivity, typically targeting EQE values exceeding 80% in peak regions.[35] Illuminated current-voltage (J-V) characteristics are obtained under standard spectra (e.g., AM1.5G for terrestrial or AM0 for space), providing key metrics like open-circuit voltage, short-circuit current, and fill factor through swept bias measurements.[81] Electroluminescence (EL) imaging under forward bias maps defects by capturing radiative recombination emissions, highlighting shunts, cracks, or non-uniformities with spatial resolution down to micrometers.[82] Yield and quality are assessed through metrics emphasizing low defect densities and high uniformity. Process-induced defect densities are maintained below 10⁴ cm⁻² to achieve commercial viability, primarily through controlled etching and deposition to minimize point defects and dislocations.[83] Wafer-scale uniformity is evaluated via EL or photoluminescence mapping, ensuring variation in efficiency across 4-inch wafers remains under 2% for high-volume production.[84] For space qualification, devices undergo radiation testing to simulate orbital environments, including proton and electron irradiation per ASTM standards such as F1193 for displacement damage equivalence.[85] These tests assess degradation in remaining factor (e.g., <10% power loss after 1 MeV electron fluence of 10¹⁵ cm⁻²), confirming reliability for missions like satellites.[86]Performance Enhancements
Light Concentration
Concentrator photovoltaics (CPV) systems enhance the performance of multi-junction solar cells by using optical elements such as lenses or mirrors to focus sunlight onto a small area, typically achieving concentrations of 100 to 1000 suns. This geometric concentration increases the short-circuit current density (J_sc) linearly with the concentration factor, as more photons are directed to the cell, while the open-circuit voltage (V_oc) rises logarithmically, leading to overall higher power output.[87] Multi-junction cells are particularly suited for CPV due to their high efficiency under intense illumination, minimizing losses from series resistance and enabling better utilization of the concentrated spectrum.[88] The primary efficiency gains in CPV arise from reduced relative impact of thermalization losses, where excess energy from high-energy photons is dissipated as heat; under concentration, the increased photon flux allows the cell to operate closer to its radiative limit, boosting conversion efficiency. For instance, a six-junction inverted metamorphic solar cell achieved a record efficiency of 47.1% under 143 suns concentration in 2020, while a four-junction cell reached 47.6% under concentration as of 2022, demonstrating how CPV can surpass one-sun efficiencies by exploiting the detailed balance limits more effectively.[89][90] These gains are most pronounced in multi-junction architectures, where each subcell can be optimized for the concentrated direct normal irradiance (DNI). Key system components in CPV include primary optics like Fresnel lenses or parabolic mirrors to collect and focus sunlight, secondary optics such as compound parabolic concentrators for uniform illumination, and two-axis solar tracking mechanisms to maintain alignment with the sun's position.[91] High-efficiency multi-junction cells are mounted on heat sinks, with the entire module designed for outdoor deployment in high-DNI environments.[92] Challenges in CPV implementation center on heat dissipation, as concentrated sunlight generates significant thermal loads that can degrade cell performance if temperatures exceed 80°C, necessitating advanced cooling systems like microchannel heat sinks or phase-change materials.[93] Precise optical alignment is also critical, with tracking errors as small as 0.1° potentially reducing output by several percent, requiring robust, low-maintenance mechanisms.[94] Despite higher upfront costs for optics and tracking—estimated at 1.5 to 2 times that of flat-plate systems—CPV offers lower levelized cost of energy (LCOE) in regions with high insolation, such as deserts, due to superior energy yield per unit area and reduced balance-of-system expenses over the system's lifetime. Field tests in high-DNI sites like Morocco have shown CPV modules producing up to 2.5 times more power than conventional PV per active area, supporting economic viability in sunny climates.[95]Spectral Management
Multi-junction solar cells are highly sensitive to spectral variations in the incident sunlight, where deviations from the standard air mass 1.5 (AM1.5) spectrum—caused by factors such as cloud cover, aerosols, or water vapor—alter the photon flux across different wavelength bands.[96] This leads to spectrum mismatch, where the photocurrents generated by individual subcells become imbalanced, as each subcell is optimized for a specific portion of the spectrum based on its bandgap.[97] The overall cell current is then limited by the subcell producing the lowest photocurrent, resulting in reduced fill factor and efficiency losses of up to 4% under non-ideal conditions.[98] To address spectrum mismatch and optimize subcell performance under varying atmospheric conditions, several spectral management techniques have been developed. Luminescent solar concentrators (LSCs) incorporate fluorescent materials or quantum dots to absorb high-energy photons and re-emit them at wavelengths better matched to the subcells' absorption edges, thereby reducing mismatch and enhancing current balance in diffuse or altered spectra.[99] For instance, in tandem configurations, LSCs have demonstrated power boosts by reshaping the spectrum for upper and lower junctions, with outdoor testing showing improved annual energy yield compared to unmodified cells.[100] Dichroic filters and beam splitters provide passive spectrum splitting, selectively transmitting or reflecting wavelength bands to direct light to appropriate subcells or parallel cell types, minimizing thermalization losses and maintaining current matching across spectral shifts. Dichroic filters, in particular, enable high-transmission (>95%) for targeted bands while reflecting others, allowing hybrid systems where different junctions process split spectra independently, as demonstrated in designs optimizing triple-junction III-V cells.[101] Beam splitters extend this by enabling lateral arrangements, where spectral separation improves efficiency in concentrator photovoltaics under fluctuating irradiance.[102] Advanced nanostructures, such as metamaterials, offer precise control through engineered selective absorption, reflection, or upconversion at the nanoscale, tailoring the effective spectrum incident on multi-junction stacks without bulk optics.[103] Photonic metamaterials, for example, have been modeled to enhance light trapping in horizontal multi-junction layouts, redirecting wavelengths to underutilized subcells and reducing mismatch penalties. Recent 2025 studies on wide-bandgap III-V cells highlight improved metamaterial designs for better spectral adaptation.[104][68] Spectral modeling is essential for predicting and mitigating these effects, with the Simple Model of the Atmospheric Radiative Transfer of Sunshine (SMARTS) software widely used to simulate irradiance spectra under diverse conditions like varying aerosol optical depth or precipitable water.[105] SMARTS enables detailed analysis of subcell photocurrents for multi-junction cells at global sites, supporting the design of robust spectral management strategies. Implementation of these techniques has yielded relative efficiency improvements of 5-10% in variable spectral conditions by better aligning the light spectrum with subcell bandgaps, as evidenced in field studies of spectrum-splitter systems and LSC-enhanced tandems.[106]Thermal Management
Multi-junction solar cells, due to their layered architecture and high current densities, are particularly sensitive to temperature rises, which can significantly impact performance and longevity. Elevated temperatures reduce the open-circuit voltage (V_oc) and overall efficiency, primarily through increased non-radiative recombination rates and altered bandgap energies in the semiconductor materials. Under concentrated sunlight, these effects are exacerbated, as the higher photon flux generates more heat, potentially leading to thermal runaway if not managed. The temperature coefficient for V_oc in multi-junction cells typically ranges from -0.2% to -0.3% per °C, while the efficiency loss is approximately -0.05% per °C, depending on the specific material stack and operating conditions. These coefficients arise from the thermodynamic dependence of carrier concentrations and mobilities on temperature, with III-V compounds exhibiting more pronounced drops compared to single-junction silicon cells. For instance, in GaInP/GaAs/Ge triple-junction cells, measurements under AM1.5 spectrum show a V_oc reduction of about 2.4 mV/°C per junction, compounding across the stack. Heat generation in these devices stems mainly from Joule heating due to series resistance in the tunnel junctions and contacts, as well as non-radiative recombination losses in the absorbers, which convert excess energy into phonons rather than photons. These mechanisms are intensified under light concentration, where irradiance levels can exceed 500 suns, raising junction temperatures by 50-100°C above ambient without cooling. In such scenarios, the fill factor also degrades due to increased shunt paths, further compounding efficiency losses. To mitigate these issues, various thermal management strategies have been developed, tailored to the application environment. Heat sinks, often made from high-thermal-conductivity materials like copper or aluminum nitride, are commonly integrated to dissipate heat via conduction and convection in terrestrial setups. Phase-change materials (PCMs), such as paraffin-based composites, provide latent heat storage to buffer temperature spikes during peak illumination, maintaining cell temperatures below 80°C for extended periods. Advanced microchannel cooling systems, employing microfluidic channels etched into the substrate, enable precise heat extraction through liquid circulation, achieving thermal resistances as low as 0.1 K/W under high-flux conditions. Recent integrations of radiative cooling in 2024-2025 have shown potential to reduce operating temperatures by 5-10°C passively, enhancing efficiency in concentrator systems.[107] In space applications, where active cooling is impractical due to vacuum and weight constraints, passive strategies dominate, including radiative heat dissipation via deployable panels coated with high-emissivity materials to reject heat to deep space. These radiators, often combined with multi-layer insulation, keep operating temperatures around 50-70°C for cells like those in the International Space Station arrays. For terrestrial concentrator photovoltaics, active cooling via fans or thermoelectric modules is preferred to handle dynamic loads, ensuring reliability under varying ambient conditions. Long-term reliability is critically affected by sustained high temperatures, with degradation rates accelerating above 100°C due to diffusion-enhanced defect formation and material intermixing at junctions. Lifetime models, such as those based on the Arrhenius equation, predict a factor of 2-5 reduction in operational life for every 10°C rise, with activation energies around 0.7-1.0 eV for III-V degradation mechanisms. These models guide design margins, ensuring 25-30 year lifespans in space missions by limiting maximum temperatures through integrated thermal modeling.Recent Developments
Record Efficiencies
Multi-junction solar cells have seen steady efficiency improvements over decades, driven by advances in materials and design. In the 1980s, early dual-junction cells based on GaInP/GaAs achieved efficiencies around 25% under 1-sun conditions, marking a significant leap from single-junction technologies. By the 2010s, triple-junction configurations, typically lattice-matched GaInP/GaAs/Ge structures, reached approximately 38% efficiency under concentrated sunlight, enabling their adoption in space applications. Recent milestones highlight the potential of higher junction counts and metamorphic growth techniques. In 2020, the National Renewable Energy Laboratory (NREL) demonstrated a six-junction inverted metamorphic (IMM) cell with 47.1% efficiency under concentrated illumination (143 suns), setting a benchmark for concentrator photovoltaics.[89] This was surpassed in 2022 by Fraunhofer ISE, which achieved 47.6% efficiency with a four-junction III-V cell under 665-sun concentration, incorporating advanced anti-reflection coatings to minimize optical losses.[4] Under 1-sun conditions, NREL's triple-junction III-V cell reached 39.5% efficiency in 2022, leveraging quantum well intermixing for broader spectral absorption.[108] Emerging hybrid tandems are also pushing boundaries. In April 2025, LONGi reported a perovskite-silicon tandem cell with 34.85% efficiency under 1-sun illumination, certified by an independent lab, demonstrating the viability of integrating perovskites with established silicon bases.[54] These records are verified by authoritative institutions like NREL and Fraunhofer ISE, which employ rigorous methodologies including spectral mismatch correction, quantum efficiency measurements, and calibration against reference cells under ASTM G173 or IEC 60904-3 standards at 25°C. Efficiencies have trended upward through transitions from rigid lattice-matched designs to flexible IMM architectures, which accommodate strain from lattice-mismatched layers to enable more junctions without degrading performance.[1] Projections indicate that 5- to 7-junction cells could approach 50% efficiency by 2030, approaching detailed balance limits while balancing cost and manufacturability.[109]| Junction Type | Efficiency (%) | Conditions | Year | Institution | Citation |
|---|---|---|---|---|---|
| Dual (GaInP/GaAs) | ~25 | 1-sun | 1980s | Various (e.g., NREL chart) | |
| Triple (GaInP/GaAs/Ge) | ~38 | Concentrated | 2010s | Spectrolab/NREL | |
| Six (IMM III-V) | 47.1 | 143 suns | 2020 | NREL | [89] |
| Four (III-V) | 47.6 | 665 suns | 2022 | Fraunhofer ISE | [4] |
| Triple (III-V) | 39.5 | 1-sun | 2022 | NREL | [108] |
| Perovskite/Si Tandem | 34.85 | 1-sun | 2025 | LONGi | [54] |