Perovskite vs silicon vs thin-film solar: efficiency, cost, and durability compared

Why It Matters

Solar photovoltaics generated over 2,000 terawatt-hours of electricity globally in 2025, accounting for roughly 7% of total power production, and capacity additions exceeded 590 GW during the year (IEA, 2026). Crystalline silicon has held above 95% market share for over a decade, but certified perovskite-silicon tandem cells surpassed 33.9% efficiency in laboratory settings in late 2025 (NREL, 2026), blowing past silicon's theoretical single-junction limit of 29.4%. Meanwhile, thin-film cadmium telluride (CdTe) panels from First Solar continue to win utility contracts in high-temperature regions where silicon panels lose output. For project developers, asset managers, and procurement teams, the choice of solar technology now directly affects levelised cost of energy (LCOE), bankability timelines, and 25-year degradation risk. This guide compares the three technology families across the metrics that matter most.

Key Concepts

Crystalline silicon (c-Si) panels use wafers sliced from purified silicon ingots. Two main variants exist: monocrystalline (mono-Si), which dominates premium installations with 22 to 24% commercial module efficiency, and polycrystalline (poly-Si), which is being phased out due to lower efficiency and narrowing cost advantages. The technology benefits from six decades of manufacturing learning curves, bankability track records, and a global supply chain centred on China, which produces over 80% of the world's polysilicon and wafers (BloombergNEF, 2025).

Perovskite solar cells use a class of crystalline materials with the chemical formula ABX3, where the most common variant is methylammonium lead halide. Perovskites can be deposited as thin films through solution processing or vapour deposition, enabling potentially dramatic manufacturing cost reductions. The technology's primary commercial pathway is as a tandem layer on top of a silicon bottom cell, combining perovskite's superior blue-light absorption with silicon's red-light harvesting to exceed single-junction efficiency limits. Oxford PV shipped its first commercial perovskite-on-silicon tandem modules in 2024, achieving 24.5% module-level efficiency (Oxford PV, 2025).

Thin-film technologies include cadmium telluride (CdTe), copper indium gallium selenide (CIGS), and amorphous silicon (a-Si). CdTe dominates thin-film deployment, with First Solar operating over 15 GW of annual manufacturing capacity. Thin-film panels use roughly 1% of the semiconductor material required for c-Si, weigh less per watt, and exhibit lower temperature coefficients, meaning they lose less output in hot climates. CIGS offers flexibility for building-integrated applications but has struggled to achieve manufacturing scale.

Degradation rate measures the annual percentage loss in module output. The metric is critical for lifetime energy yield and project finance models. Silicon modules typically degrade at 0.4 to 0.5% per year, backed by decades of field data (Jordan and Kurtz, NREL, 2024). Thin-film CdTe degrades at comparable or slightly lower rates after initial light-induced stabilisation. Perovskite stability remains the technology's primary challenge: early single-junction cells degraded within months, though encapsulation advances have pushed accelerated test lifetimes past 25 years for some tandem architectures (Helmholtz-Zentrum Berlin, 2025).

Bankability refers to whether financiers will lend against a technology at competitive terms. Silicon panels from Tier 1 manufacturers carry 25- to 30-year linear power warranties and are accepted by all major project finance banks. Thin-film CdTe from First Solar is similarly bankable. Perovskite modules are not yet independently bankable for large-scale project finance, though tandem modules with silicon bottom cells benefit from the underlying c-Si bankability foundation.

Head-to-Head Comparison

Efficiency

Laboratory record efficiencies as of early 2026 stand at 26.8% for mono-Si single junction, 33.9% for perovskite-silicon tandem, and 22.3% for CdTe (NREL, 2026). Commercial module efficiencies, which account for interconnection losses, encapsulation, and manufacturing tolerances, lag lab records by 3 to 6 percentage points. LONGi's HJT silicon modules reach 23.3% in production. Oxford PV's tandem modules deliver 24.5%, and First Solar's Series 7 CdTe panels achieve 19.8% (manufacturer specifications, 2025). The tandem efficiency advantage translates directly into higher energy yield per square metre, which matters most in land-constrained or rooftop deployments.

Temperature performance

Silicon output drops by approximately 0.35% per degree Celsius above 25 °C (standard test conditions). CdTe panels have a temperature coefficient of roughly 0.28% per °C, giving them a 2 to 4% energy yield advantage in desert and tropical climates where module temperatures regularly exceed 60 °C (First Solar, 2025). Early data on perovskite-silicon tandems show temperature coefficients between 0.30% and 0.34% per °C, placing them between the two established technologies (Fraunhofer ISE, 2025).

Durability and warranty

Silicon Tier 1 manufacturers offer 25- to 30-year warranties guaranteeing at least 87.4% of nameplate power at year 25. First Solar provides a 30-year warranty on Series 7 CdTe modules with similar degradation guarantees. Perovskite tandem modules from Oxford PV currently carry a 15-year product warranty with a 20-year performance guarantee, reflecting the technology's earlier commercial stage. Moisture ingress and halide ion migration remain the primary degradation mechanisms under investigation, though recent advances in 2D/3D perovskite interfaces have significantly improved intrinsic stability (Nature Energy, 2025).

Manufacturing scalability

Silicon benefits from a deeply scaled global supply chain: over 900 GW of module manufacturing capacity existed worldwide by end of 2025, with wafer costs below $0.03 per watt (BloombergNEF, 2025). First Solar's CdTe manufacturing reached 16 GW of annual capacity across facilities in the U.S., India, and Malaysia, with capex of approximately $0.25 per watt of capacity. Perovskite manufacturing remains pre-scale: Oxford PV's Brandenburg factory has a capacity of roughly 600 MW, and Chinese manufacturers including Longi, Renshine Solar, and Utmolight are building pilot lines of 100 to 300 MW (PV Magazine, 2025). The capital cost of perovskite deposition equipment is projected to be 30 to 50% lower than silicon cell lines at equivalent scale because the process occurs at low temperatures and atmospheric pressure.

Cost Analysis

Silicon module spot prices reached historic lows of $0.09 per watt in late 2025, driven by massive Chinese overcapacity (InfoLink, 2025). Fully installed utility-scale system costs, including balance of system and soft costs, average $0.55 to $0.75 per watt in the U.S. and $0.35 to $0.50 per watt in China and India. This translates to an LCOE range of $20 to $35 per MWh depending on irradiance and financing terms (Lazard, 2025).

First Solar's CdTe modules price at a modest premium of $0.22 to $0.28 per watt, but the company's vertically integrated U.S. manufacturing qualifies for $0.07 per watt in domestic content IRA tax credits, narrowing the effective cost gap. CdTe system LCOE ranges from $22 to $38 per MWh.

Perovskite-silicon tandem module pricing is difficult to benchmark at commercial scale. Oxford PV has indicated target module prices of $0.15 to $0.20 per watt at GW-scale production, which would undercut silicon on a per-watt basis. Because tandems produce more energy per square metre, balance-of-system costs per watt are also lower, potentially reducing installed system costs by 10 to 15%. Projected tandem LCOE at scale ranges from $18 to $28 per MWh, though this remains contingent on demonstrating manufacturing yields above 95% and field degradation below 0.5% per year (Fraunhofer ISE, 2025).

Use Cases and Best Fit

Utility-scale ground mount in temperate climates. Silicon mono-PERC or TOPCon modules remain the default choice. Bankability is unmatched, supply is abundant, and LCOE is at historic lows. Projects above 100 MW can negotiate module prices below $0.10 per watt.

Utility-scale in hot, arid climates. CdTe thin-film gains a measurable energy yield advantage due to its lower temperature coefficient and better spectral response under diffuse light conditions. First Solar's Series 7 panels are specifically designed for desert deployment, and the company holds long-term supply agreements with developers in the Middle East, U.S. Southwest, and India (First Solar, 2025).

Rooftop and space-constrained installations. Perovskite-silicon tandems offer the highest watts per square metre, making them ideal where roof area limits system size. Early adopters including IKEA and Deutsche Bahn have piloted tandem modules on commercial rooftops in Europe (Oxford PV, 2025).

Building-integrated photovoltaics (BIPV). Thin-film CIGS and perovskite can be fabricated on flexible substrates and in semi-transparent configurations, enabling integration into facades, windows, and curved surfaces. Companies like Heliatek (organic PV) and Saule Technologies (perovskite) target this niche.

Off-grid and portable applications. Lightweight thin-film and perovskite modules suit mobile, temporary, or weight-sensitive deployments including disaster relief, military forward bases, and agrivoltaics structures where heavy glass-glass silicon panels are impractical.

Decision Framework

1. Assess site conditions. Evaluate ambient temperature range, irradiance profile, available area, and structural load capacity. Hot climates favour CdTe; constrained areas favour tandems; standard ground-mount suits silicon.

2. Confirm bankability requirements. If the project requires non-recourse debt financing, verify that the chosen technology has Tier 1 classification and independent energy yield assessments from firms like DNV or Black & Veatch. As of early 2026, only c-Si and CdTe meet this bar for large-scale project finance.

3. Model lifetime energy yield. Use location-specific simulation tools (PVsyst, SAM) to compare technologies on an energy-per-dollar basis rather than peak-watt nameplate alone. Include temperature losses, spectral effects, soiling, and degradation trajectories.

4. Evaluate supply chain risk. Silicon supply is concentrated in China, creating tariff and geopolitical exposure. CdTe faces tellurium scarcity constraints beyond 100 GW of cumulative deployment. Perovskite uses lead, raising end-of-life recycling obligations under EU WEEE regulations. Map each risk against project timeline and procurement strategy.

5. Factor in policy incentives. U.S. IRA domestic content bonuses favour First Solar CdTe and emerging U.S. perovskite lines. EU Carbon Border Adjustment Mechanism (CBAM) may eventually apply carbon costs to imported silicon modules manufactured with coal-heavy electricity. Indian PLI scheme subsidies target domestic c-Si manufacturing. Align technology choice with available incentive structures.

6. Plan for technology evolution. For projects with 2028+ commercial operation dates, consider locking in tandem module supply agreements now to secure early allocation. For near-term projects, use proven silicon or CdTe and design mounting systems to allow future repowering with higher-efficiency tandems.

Key Players

Established Leaders

• LONGi Green Energy — World's largest solar manufacturer by shipments, producing over 100 GW of silicon modules annually with leading HJT and TOPCon cell lines.
• First Solar — Sole large-scale CdTe manufacturer with 16 GW capacity, vertically integrated U.S. production, and a fully booked order pipeline through 2028.
• JA Solar — Top-five global silicon module supplier with over 75 GW annual capacity and n-type TOPCon technology at scale.
• Trina Solar — Pioneer in n-type i-TOPCon cells with certified module efficiency above 23%.

Emerging Startups

• Oxford PV — First to commercialise perovskite-silicon tandem modules from its Brandenburg, Germany factory; targeting 5 GW capacity by 2028.
• Caelux (a First Solar subsidiary) — Developing perovskite-on-CdTe tandem technology leveraging First Solar's thin-film manufacturing expertise.
• Saule Technologies — Polish startup producing flexible perovskite solar cells via inkjet printing for BIPV applications.
• Renshine Solar — Chinese perovskite manufacturer with a 150 MW pilot line and reported 28.6% lab-cell efficiency.

Key Investors/Funders

• Breakthrough Energy Ventures — Early investor in perovskite and advanced PV startups, providing capital across the technology readiness spectrum.
• U.S. Department of Energy (DOE) — Funded over $100 million in perovskite R&D through the Solar Energy Technologies Office since 2020, targeting $0.02 per kWh utility-scale solar.
• European Innovation Council (EIC) — Backed multiple perovskite commercialisation projects through Horizon Europe grants exceeding €200 million.

FAQ

When will perovskite-silicon tandems be fully bankable for utility-scale project finance? Industry consensus points to 2028 or 2029 as the likely inflection point. By then, Oxford PV and Chinese manufacturers are expected to have accumulated 3 to 5 years of field performance data, and at least two independent testing laboratories should have completed IEC 61215 and IEC 61730 qualification sequences for tandem modules. Until then, tandems will likely be deployed in corporate PPA structures where the offtaker accepts technology risk, or in hybrid projects where tandem modules occupy a portion of the array alongside conventional silicon.

Is lead in perovskite cells an environmental concern? A standard perovskite-silicon tandem module contains approximately 0.5 grams of lead per square metre, roughly one-tenth the lead content of a car battery. Encapsulation prevents leaching during operation, and recycling processes can recover over 99% of the lead at end of life. The EU Restriction of Hazardous Substances (RoHS) directive currently grants an exemption for photovoltaic panels. Lifecycle assessments by Fraunhofer ISE (2025) conclude that the environmental benefit of higher efficiency and lower manufacturing energy far outweighs the marginal toxicity risk of encapsulated lead.

How does thin-film CdTe compete when silicon prices are at historic lows? First Solar's competitive advantage is not module price alone but total cost of ownership. CdTe panels deliver higher energy yield in hot climates, carry lower degradation rates after initial stabilisation, and qualify for U.S. domestic content IRA bonuses worth $0.07 per watt. First Solar also offers a fully vertically integrated supply chain free from Chinese polysilicon, which mitigates tariff, forced labour compliance, and geopolitical risks. For U.S. utility-scale developers, these factors frequently make CdTe the lowest-LCOE option despite a higher nameplate module price.

What happens to silicon module prices if Chinese overcapacity persists? BloombergNEF (2025) estimates that global silicon module manufacturing capacity exceeded 1,200 GW by end of 2025 against roughly 590 GW of installations, implying a utilisation rate below 50%. This overcapacity is driving consolidation: several second-tier Chinese manufacturers announced production halts or mergers in late 2025. Module prices may recover modestly to $0.12 to $0.15 per watt as unprofitable capacity exits, but structural cost reductions in n-type cell technology will keep long-term prices below $0.18 per watt. Buyers should lock in near-term pricing while maintaining flexibility for technology upgrades.

Can I mix technologies within a single solar project? Yes, and this approach is gaining traction. Developers are beginning to deploy CdTe on south-facing ground-mount arrays and silicon bifacial modules on east-west tracking systems within the same project to optimise energy yield across the day. Tandem modules can be integrated into dedicated subarrays with separate inverter strings. The key constraint is that each technology requires its own maximum power point tracking (MPPT) input, so inverter and string design must accommodate the differing voltage and current characteristics.

Sources

• IEA. (2026). World Energy Outlook: Solar PV Deployment and Capacity Additions. International Energy Agency.
• NREL. (2026). Best Research-Cell Efficiency Chart, January 2026 Update. National Renewable Energy Laboratory.
• BloombergNEF. (2025). Global Solar Market Outlook: Supply Chain, Pricing, and Technology Trends. BloombergNEF.
• Oxford PV. (2025). Commercial Tandem Module Performance and Manufacturing Scale-Up. Oxford PV.
• First Solar. (2025). Series 7 Module Datasheet and Hot-Climate Energy Yield Analysis. First Solar, Inc.
• Lazard. (2025). Lazard's Levelized Cost of Energy Analysis, Version 17.0. Lazard.
• Fraunhofer ISE. (2025). Perovskite-Silicon Tandem Technology Assessment: Stability, Cost, and Lifecycle Analysis. Fraunhofer Institute for Solar Energy Systems.
• Jordan, D. and Kurtz, S. (2024). Photovoltaic Degradation Rates: An Analytical Review, Updated Dataset. National Renewable Energy Laboratory.
• InfoLink. (2025). Solar Module Spot Price Tracker, Q4 2025. InfoLink Consulting.
• Nature Energy. (2025). Long-Term Stability of 2D/3D Perovskite Heterostructures Under Accelerated Aging. Nature Energy.