SiCl₄ Overcapacity Meets Fiber Preform Boom
July 2026
TL;DR
SiCl₄ is simultaneously in chronic oversupply (industrial-grade spot price collapsed from ¥8,000/t in 2021 to ¥2,000–3,000/t by 2026) and in acute shortage at fiber-grade purity (5N–6N). The paradox is not volume — it is purity. The same molecule that polysilicon plants can barely give away must be distilled to parts-per-billion metal impurity levels before optical preform producers will touch it. With global optical fiber demand set to exceed 7 billion fiber-km by 2028, this structural mismatch will persist for years.
Why SiCl₄ Is Oversupplied
The Siemens polysilicon process generates SiCl₄ as an unavoidable byproduct: roughly 15–20 tonnes of SiCl₄ per tonne of silicon produced. China's polysilicon capacity has expanded to approximately 2.5 million tonnes/year (2026 nameplate), implying potential SiCl₄ co-production of 35–50 million tonnes/year — far beyond what downstream markets absorb cleanly.
Price collapse followed capacity addition mechanically:
| Year | SiCl₄ Industrial-Grade Price (¥/t) |
|---|---|
| 2021 | ~8,000 |
| 2022 | ~5,500 |
| 2023 | ~4,000 |
| 2024 | ~2,800 |
| 2026 | 2,000–3,000 |
The dominant disposal route — hydrogenation back to SiHCl₃ (TCS) for polysilicon recycling — absorbs the majority, but is economically circular rather than value-creating. See SiCl₄ downstream outlets for the full breakdown.
Why Fiber Preform Is Booming
Optical fiber preforms are produced by chemical vapor deposition of SiO₂ from SiCl₄ feedstock using MCVD, OVD, or VAD processes (see fiber preform manufacturing). The fiber demand picture is unambiguously expansionary:
| Year | Global Optical Fiber Demand (billion fiber-km) |
|---|---|
| 2022 | 4.7 |
| 2024 | 5.5 |
| 2026 (est) | 6.2 |
| 2028 (forecast) | 7.0+ |
Growth is driven by four simultaneous buildouts: 5G backhaul densification, FTTH (fiber-to-the-home) rollouts in Southeast Asia and India, AI hyperscale datacenter interconnect (800G/1.6T optical modules), and new transoceanic subsea cable programs. Each driver independently sustains multi-year demand; together they create a decade-long cycle.
The Mismatch: Volume Is Not the Problem
The paradox resolves analytically once purity is understood. Industrial-grade SiCl₄ is typically 3N (99.9% SiCl₄) with metal impurities in the parts-per-million range — acceptable for fumed silica flame hydrolysis and hydrogenation recycling, but categorically unacceptable for optical preforms, where metal impurity absorption at ppb levels causes fiber attenuation loss that degrades signal over kilometers.
Fiber-grade SiCl₄ specification requires:
- Purity: 5N–6N (99.999%–99.9999%)
- Fe, Ni, Cu, Cr: each < 1 ppb
- H₂O: < 1 ppm
- Dedicated fractional distillation train + cleanroom handling
Retrofitting a commodity SiCl₄ stream to fiber-grade requires substantial capital. Very few Chinese polysilicon plants have this capability. The result: polysilicon byproduct SiCl₄ piles up at ¥2,000/t, while fiber-grade SiCl₄ trades at a substantial premium with supply constrained.
Preform manufacturing capacity is the second bottleneck. YOFC, FiberHome, Hengtong, and Zhongtian collectively account for 60%+ of global preform output. Their VAD and OVD lines operate at high utilization, with order backlogs extending to 2027–2028 in some configurations.
Strategic Implications
For silicon supply chain participants, the opportunity lies in the purity gap. The companies that invest in SiCl₄ purification infrastructure — not polysilicon expansion — capture the value differential. For fumed silica producers, 3N SiCl₄ remains reliably cheap. For preform producers, sourcing fiber-grade SiCl₄ from non-Chinese suppliers (Japan, Germany) reduces geopolitical exposure at a cost premium.