Solar Cells

What it is

A solar cell (光伏电池片) is a processed wafer that turns light into electric current. Trina describes the steps it takes in-house: polysilicon is grown into ingots and sliced, the wafers are made into cells through processes that include printing the metal contacts, and the cells are laid out and sealed into modules. It buys the polysilicon, silver paste and metal parts from outside (Trina 2026 interim report, p. 10). Drinda, which makes only cells, notes that cell efficiency directly sets module efficiency, and so a plant's output and returns (Drinda 2026 interim report, p. 10).

Three designs compete:

• TOPCon (tunnel oxide passivated contact) became the mainstream technology in 2024, with 70% of output against 30% in 2023 (IEA PVPS Task 1, October 2025, p. 9). Drinda says n-type cells, the family TOPCon belongs to, are now more than 90% of China's mass production (p. 10).
• Back-contact (BC) cells move all the metal contacts to the rear. That removes the shading of a front grid and lets cells sit closer together in a module (PVEducation, Honsberg and Bowden). Trina lists the trade-offs: the rear side generates only 55% to 65% as much as the front, which confines BC mainly to rooftops; converting a line costs several hundred million yuan per GW; and the field is dense with patents (Trina p. 14). The BC makers report the opposite trend: Aiko has launched an ABC module with 80% (plus or minus 5%) bifaciality for utility-scale plants and lists utility projects won abroad and in Chinese state-owned tenders (Aiko 2026 interim report, p. 12), and LONGi says its BC modules were selected for more than 10 GW of domestic state-owned tenders (LONGi 2026 interim report, p. 12).
• Heterojunction (HJT) cells are developed in parallel by makers such as Tongwei (Tongwei 2026 interim report, p. 11).

Why it matters for the chain

The cell sits between solar wafers and solar modules. Drinda says it takes price pressure from both sides (p. 20).

It is also where the silver goes. The IEA's photovoltaic programme puts solar's use at 17% of global silver production (IEA PVPS p. 9). Jinko Solar says that after silver prices surged from late 2025, silver paste overtook polysilicon as the largest cost item in a module (Jinko 2026 interim report, p. 8), and LONGi blamed part of its first-half loss on high silver prices in the first quarter (LONGi p. 12). Makers are cutting it out: Aiko says it has a silver-free metallisation process for its BC cells (Aiko p. 14), and Jinko converted nearly 30 GW of capacity to base-metal contacts by the end of June (Jinko p. 10).

Where China stands

Global cell output was 753 GW in 2024, and China made 90% of it (IEA PVPS p. 9).

• Tongwei (通威股份, 600438.SH) sold 34.78 GW of cells in the first half of 2026. It says its cells have held the largest global market share for many years and that its customers include the world's ten largest module makers (Tongwei pp. 8, 9, 11).
• Drinda (钧达股份, 002865.SZ, 2865.HK), whose operating companies carry the Jietai name, makes only cells, with more than 40 GW of capacity in Chuzhou and Huai'an (Drinda pp. 5, 12).
• Aiko (爱旭股份, 600732.SH) began as a cell supplier and now makes mainly its own all-back-contact (ABC) modules, which brought in 79% of its revenue in the half (Aiko p. 10).
• JA Solar (晶澳科技, 002459.SZ) uses most of its cells itself, but its cell sales to others rose 167.14% to ¥1.44bn (JA Solar 2026 interim report, p. 16).

Cells are the part of the chain other countries lack most. China's cell exports rose 36.8% in the first half of 2026, while wafer exports fell 0.6% and module exports 2.5%. Drinda's explanation: countries are adding module assembly faster than cell capacity, which is harder to build (Drinda p. 11). China ended the export VAT rebate on solar products on 1 April 2026; exports surged before the deadline, and in May cell export volumes fell 35% from April (Jinko p. 9).

Economics

Gross margins in the first half of 2026 depended on the buyer and the destination:

• JA Solar's cell sales to others earned 13.34% (JA p. 16).
• Drinda's cells earned 5.12%, up 3.37 points. Its overseas sales earned 7.89%, against 4.21% at home (Drinda p. 14).
• Aiko's cell sales earned about -1.4% (our calculation from Aiko p. 138).
• Tongwei's cell subsidiary lost ¥1.37bn on revenue of ¥11.50bn (Tongwei p. 17).

What to watch

• Efficiency floors. Mandatory national standards taking effect on 1 January 2027 set minimum module efficiencies of 23.2% for TOPCon and HJT and 23.5% for BC (Trina p. 13). Aiko expects them to force out low-efficiency capacity (Aiko p. 9).
• Cell plants abroad. Drinda's partly owned cell plant in Turkey has started production, while its planned 5 GW plant in Oman has been held back by trade and Middle East uncertainty (Drinda p. 12). JA Solar is bringing a 6 GW cell and 3 GW module project in Oman into production in phases (JA p. 12). Our view: these plants are the next step after exporting cells. If they work, part of the export growth that Chinese cell makers now rely on moves to factories abroad, some of them their own.
• Patents. Patent suits over TOPCon cells are frequent in major markets. In March 2026 the US International Trade Commission opened a Section 337 investigation at First Solar's request, which JA Solar says bears directly on Chinese makers' access to the US market (JA p. 30).