Global TOPCon cell capacity reached 400 GW in 2025, taking roughly 85% of the world's solar cell production and overtaking legacy PERC as the dominant n-type architecture [S5]. Manufacturing cost now sits at $0.08–$0.12/W on converted PERC lines, with each gigawatt of new TOPCon capacity requiring about 35–40 million USD in tool capex depending on the polysilicon deposition route [S5].
For planners, the headline number matters less than the retrofit math: a standard PERC brownfield can be converted to TOPCon for 5–10 million EUR per line, so 2026 capacity additions are overwhelmingly upgrades, not greenfield [S5]. India's MNRE reinforced that direction on 21 Aug 2026 by approving new TOPCon and HJT cell models under the ninth revision of ALMM List-II, explicitly channeling domestic manufacturing toward high-efficiency n-type architectures [S3][S6].
Process stack and the 1.5 nm tunnel oxide
TOPCon (Tunnel Oxide Passivated Contact) cells add an ultra-thin silicon oxide layer of approximately 1–2 nm combined with a doped polycrystalline silicon layer on the rear of an n-type wafer, enabling selective carrier transport and rear-surface passivation [S7]. The 12-step flow runs texturing, boron diffusion, BSG removal, tunnel-oxide plus poly-Si deposition, wraparound removal, edge isolation, AlOx ALD, SiNx PECVD, SE laser doping, screen printing, co-firing, and cell testing, with each wafer spending about 90 minutes on a 10 GW line [S5].
The two competing routes for the rear stack are LPCVD at 12.0 million USD per GW of capex and PECVD at 7.0 million USD per GW, with PECVD typically integrating the phosphorus doping step in-situ and saving a separate 2.0 million USD tool [S5]. Mass-production cell efficiency sits at 24–25.5% on industrial lines, with 26% routinely reported in 2026 production data [S5].
2026 efficiency records and silver reduction
JinkoSolar and the Ningbo Institute of Materials Technology (NIMTE) certified a 26.66% industrial TOPCon cell in February 2026, the highest industrial efficiency for a mass-produced silicon cell [S5]. Trina Solar separately confirmed TOPCon as the bottom cell for perovskite tandems by hitting 32.6% on a full-size perovskite/TOPCon stack in the same month [S5].
Silver consumption, historically a 10–12 mg/Wp drag on cost, fell toward 1.1 mg/Wp after Fraunhofer ISE demonstrated nickel-copper plating on TOPCon in April 2026, and Jolywood started mass production of low-silver "Namic" TOPCon cells exceeding 27% efficiency in May 2026 [S5]. JinkoSolar, JA Solar, Trina, and Tongwei each passed 80 GW of annual cell output in Q1 2026, a scale that lets them absorb silver-reduction retrofits within existing 400 GW of installed TOPCon capacity [S5].
Capex breakdown per gigawatt
A 2026 TOPCon greenfield requires roughly 35–40 million USD in equipment capex per GW, distributed across texturing wet benches (1.5M), boron diffusion tubes (4.0M), BSG removal (1.5M), tunnel oxide plus poly-Si (7.0M on PECVD, 12.0M on LPCVD), phosphorus doping (2.0M, skipped for in-situ PECVD), wraparound removal (1.5M), edge-isolation laser (0.5M), AlOx ALD (2.5M), SiNx PECVD (3.0M), SE laser doping (1.0M), screen printers (1.5M), co-firing furnaces (1.5M), and cell testing plus automation (1.0M) [S5].
JinkoSolar's plan announced in June 2025 was to upgrade over 40% of its existing capacity by year-end 2025, ending with 40–50 GW of high-power TOPCon capacity capable of 670 W modules at 24.8% component efficiency and 85% bifaciality, a 10% price premium over standard TOPCon [S2]. Trina Solar's 2026 cell capacity sits at 75 GW with 95 GW of module assembly, including 40 GW of n-type cells and new Vertex 210R 605 W and Vertex 210 695 W i-TOPCon modules at up to 22.4% module efficiency [S4].
ALMM List-II and India capacity pull
India's MNRE Grid Solar Power Division released the ninth revision of ALMM List-II for solar PV cells on 21 Aug 2026, adding new high-efficiency TOPCon and HJT cell models and removing or updating legacy PERC entries to push approved domestic capacity toward n-type architectures [S3]. The move is designed to crowd in Indian cell manufacturing for utility-scale tenders and rooftop subsidy programs that increasingly specify n-type modules [S6].
For global planners, the ALMM gate is a demand signal: any TOPCon line targeting Indian project revenue must be on the approved list, and the 22 Aug 2026 publication date aligns with tender windows opening in Q4 2026 [S3][S6]. A second-order effect is rising domestic content requirements, which favor lines using n-type wafers and tunnel-oxide deposition tools sourced inside India, even if the underlying equipment IP is foreign [S6].
Comparison: TOPCon vs HJT vs PERC for 2026 fabs
On cell efficiency, TOPCon industrial lines run 24–25.5% in mass production with a 26.66% record [S5], HJT sits a step higher on bifaciality and temperature coefficient but requires different tool sets, and legacy PERC trails at 22.5–23.5% with no realistic path past 24% [S4][S5]. On capex per GW, TOPCon brownfield retrofits are 5–10 million EUR per line, far below HJT greenfield budgets, while PERC lines have no upgrade value [S5].
On silver intensity, TOPCon starts at 10–12 mg/Wp and is dropping toward 1.1 mg/Wp with nickel-copper plating [S5]; HJT historically uses more silver per cell because of low-temperature pastes, and PERC sits between the two. JinkoSolar's upgraded TOPCon modules achieve 85% bifaciality, a 10% premium over traditional TOPCon products [S2]. The honest takeaway for a planner evaluating a 2026 line: TOPCon is the lowest-risk retrofit on a PERC asset, HJT is the right call only on a greenfield, and a fresh PERC line in 2026 is hard to defend.
Selection criteria and who should build TOPCon in 2026
TOPCon fits a planner who already owns a PERC brownfield, has access to an n-type wafer supply, and wants to participate in utility-scale tenders that score on bifacial gain and 30-year warranties [S1][S4]. It does not fit a planner betting on perovskite tandems as the primary 2026 product (use TOPCon as the bottom cell only), nor a planner building a small regional fab below 2 GW, where fixed tool costs compress margin below viable levels [S5].
Key 2026 selection filters are: tunnel-oxide deposition route (PECVD for lower capex, LPCVD for higher throughput and maturity), silver strategy (plating retrofit vs low-temperature paste), module power target (605–705 W class for utility, 410–490 W class for residential and C&I), and bifaciality tested at 85% or better [S1][S2][S5]. Procurement teams sourcing modules should also verify ALMM listing for India-bound volume and 25-year material plus 30-year power warranties [S1][S3].
Limits, failure modes, and what to watch next
The main 2026 risks for TOPCon capacity planners are silver price volatility, which can swing cost of ownership by several percent per cell even at 1.1 mg/Wp, and tunnel-oxide process drift that quietly degrades Voc if SE laser doping or AlOx ALD is not tightly controlled [S5]. LPCVD lines also face wraparound and edge-isolation yield losses that PECVD partly avoids but at the cost of in-situ doping complexity [S5].
Trackable signals through the rest of 2026 are: ALMM List-II subsequent revisions adding newer TOPCon models, JinkoSolar's stated path to higher than 40% capacity upgraded by end-2025 [S2], perovskite/TOPCon tandem announcements from Trina and JinkoSolar, and the next India tender cycle under the PM Surya Ghar program. For context on the upstream material side, see polysilicon market discipline across 2026 and the tier-1 producer share of solar-grade polysilicon, both of which feed directly into TOPCon wafer cost.
The underlying component specifications are covered under load cell, load cell module, and pressure transmitter.