Why solar panels use less silver per watt

What the silver in a solar cell does

The silicon in a photovoltaic cell absorbs light and produces an electrical current. Metal contacts collect that current and carry it into the circuit. Silver paste forms those contacts.

The contacts have to carry current while shading as little of the cell’s surface as possible. Jenny Chase’s Solar Power Finance Without the Jargon describes how better pastes and more precisely shaped contacts have reduced the area occupied by these connections without sacrificing conductivity.¹

Thinner contacts can raise electrical resistance or degrade faster, which reduces the power a module delivers over its life.

Silver use per watt since 2020

Solar panels use less silver per watt than they did in 2020. The International Energy Agency estimates that advances in screen printing, multi-busbar designs and finer metallization lines cut silver use in mainstream crystalline-silicon cells from around 16 milligrams of silver per watt (mg/W) in 2020 to close to 10 mg/W in 2024, a reduction of roughly 38% on those rounded figures.²

Through 2024, installation growth outpaced those savings and total silver demand from solar continued to rise.² Metals Focus estimates that photovoltaic (PV) silver demand then fell from 197.5 million troy ounces in 2024 to 186.6 million in 2025, even as cell and module production continued to expand.³

Finer printing and copper substitution

Manufacturers can cut silver by printing narrower contacts, by replacing part of the silver with copper, or by doing both. In April 2025, Fraunhofer ISE reported silicon heterojunction cells using 1.4 mg/W of silver, made with silver-copper paste on the front, pure copper paste on the rear and fine-line printing.⁴

The cells were more efficient than all-silver reference cells, and in micromodules put through 200 accelerated thermal cycles they degraded at the same rate as the references; further climate-chamber tests were under way when the results were announced. The institute put 1.4 mg/W at about one-tenth of the industrial standard at the time.⁴

In April 2026, Fraunhofer ISE reported TOPCon (tunnel oxide passivated contact) cells with copper contacts electroplated on pilot equipment: a nickel layer stops copper migrating into the cell, copper carries the current and a thin silver layer protects against oxidation. The cells used 1.1 mg/W of silver and matched the 24% efficiency of screen-printed silver references.⁵

The contact still contains some silver, and manufacturers would need to add electroplating equipment to their lines, an investment the institute acknowledges.

Factory adoption has been slower. The International Technology Roadmap for Photovoltaics (ITRPV), an annual survey of solar manufacturers and suppliers published by the German engineering association VDMA, reported in June 2026 that plating had yet to win a significant share of mass production. Heterojunction manufacturers were using copper mainly in modified pastes, such as silver-coated copper, rather than plating.⁶

JinkoSolar reported in its Q2 2026 earnings presentation, dated August 26, 2026, that about 30 GW of its capacity had been upgraded with base-metal metallization technology.⁷ The presentation gives no silver use per watt for those lines and does not name the base metal, so the saving they deliver is not public.

Canadian Solar’s silver-reduction roadmap

Canadian Solar set out a PV technology roadmap on its Q2 2026 earnings call on August 27, 2026. For its third-generation heterojunction and TOPCon products, the company showed module efficiency rising from 23.2% to 24.4% and silver use falling from 6.5 to 3 mg/W between 2026 and 2028.⁸

The roadmap also showed mass production in 2028 of a back-contact design, labeled QBC, aimed at premium residential customers, with module efficiency of 24.8–25.2% and silver use of 1–2 mg/W.⁸

Canadian Solar’s roadmap starts at 6.5 mg/W in 2026; the 3 mg/W and 1–2 mg/W figures are targets for 2028, set in August 2026. Silver-per-watt figures from different stages are not directly comparable. Fraunhofer ISE’s come from research cells, the IEA’s and the ITRPV’s are industry-wide figures, and Canadian Solar’s are company targets.

Cell technologies and First Solar’s silver-free modules

More efficient silicon cell designs have often needed more silver. Figures from 2023 in Chase’s book put PERC (passivated emitter and rear contact) cells at about 10 mg/W, TOPCon cells at about 13 mg/W and heterojunction cells at about 22 mg/W; heterojunction cells are limited to processing temperatures of about 200°C, which leaves their silver contacts less conductive.¹ The IEA attributes TOPCon’s higher loadings — the silver used per watt of cell — to contacts on both the front and rear of the cell, where PERC cells are metallized on one face.² In the ITRPV’s 2025 survey, median cell-level loadings were about 10 mg/W for bifacial TOPCon cells, 12.0 mg/W for heterojunction and 12.2 mg/W for back-contact cells. The June 2026 edition expects new pastes and screens to bring TOPCon down to 6.3 mg/W and heterojunction to 4.3 mg/W within ten years.⁶

First Solar makes thin-film modules from cadmium telluride (CdTe) rather than silicon.

The ITRPV puts thin-film technologies, which include cadmium telluride, at about 2% of the market at the end of 2025, against about 98% for crystalline silicon.⁶ At that share, more cadmium telluride output trims solar silver demand only at the margin; silver use in silicon cells determines the total.

Swipe or scroll to compare all columns →

Perovskite-silicon tandem cells still need metal contacts. Higher efficiency lowers mg/W for the same amount of silver, but the contact material determines how much silver a tandem cell needs.

Production growth against falling loadings

Annual solar silver demand is approximately the watts manufactured multiplied by average silver use per watt across all of that output. Manufacturing and installation figures differ, because modules can sit in inventory or wait for project construction before they are installed.

In the same illustration, smaller reductions leave demand rising. If intensity falls only 10%, total silver requirements increase 8%. For demand to stay flat, intensity must decline by approximately 16.7%.

Metals Focus estimates that silver loadings per unit fell by more than 15% in 2025 as manufacturers facing price competition and higher silver costs adopted thrifting and substitution technologies. In 2025, those cuts in silver per cell outweighed growth in manufacturing, and photovoltaic silver demand fell 6%. Metals Focus’s forecast, published in April 2026, put 2026 demand at 151 million ounces, about 19% below 2025, combining a further 15–20% cut in silver use per cell with slower installations, particularly in China.³

Silver costs and the solar end market

For a manufacturer, reducing mg/W lowers the amount of silver it needs to buy. Whether the cash cost falls also depends on the metal price. Silver started 2025 below $29 an ounce and peaked at $84 in December, with the annual average 42% higher year on year at just over $40. Metals Focus estimates that silver rose from 8–10% of solar cell costs at the start of 2025 to more than 20% during the year.³

LONGi Green Energy said in its 2025 annual report, published in April 2026, that the rise in silver prices had made silver paste the largest cost component for cells and modules. The company also said it honored supply contracts during the silver-price volatility of the second half of 2025 and the first quarter of 2026, despite incurring short-term losses.⁹

Asked about input costs on First Solar’s Q2 2026 earnings call on July 30, 2026, chief executive Mark Widmar said: “We don’t use silver, but obviously our competitors do.”¹⁰

For illustration, a 25% reduction in silver use combined with a 50% increase in price leaves the silver bill per watt 12.5% higher, before paste-processing costs and other changes.

In 2025, photovoltaics accounted for 186.6 million ounces of silver demand, about 28% of industrial demand and about one-sixth of total demand, according to Metals Focus.³