Solar Panel Supply Chain: Where Antimony and Copper from Pakistan Actually Fit

Antimony trioxide was trading around $39,500 a tonne last time I checked the Rotterdam quotes. Two years ago it was under $12,000. That single number tells you most of what you need to know about why my phone keeps ringing from Frankfurt, Shenzhen and, more recently, Phoenix.

But the story people are telling themselves about the solar supply chain — that it's mostly a polysilicon and silver problem — misses two metals that quietly sit inside every panel and every inverter cabinet. Antimony and copper. And Pakistan, specifically Gilgit-Baltistan, has meaningful deposits of both.

Let me walk through where we actually fit. Not where a pitch deck says we fit.

The antimony piece nobody explains properly

Solar glass is the thing. When you laminate a crystalline silicon module, the front cover glass has to resist solarisation — that slow yellowing under UV that kills transmittance over a 25-year warranty period. The industry fix, for two decades now, has been antimony trioxide as a fining and clarifying agent in the melt. Roughly 0.1 to 0.25% by weight in the glass batch. Doesn't sound like much until you multiply it across 500 GW of annual module output.

A rough working number the glass makers use: about 200 to 250 grams of Sb2O3 per kilowatt of module. So a gigawatt of panels pulls somewhere in the region of 220 tonnes of antimony trioxide, which back-calculates to around 180 tonnes of contained antimony metal. Do the math on projected 2027 solar deployment and you're looking at 90,000 to 110,000 tonnes of antimony demand from solar glass alone. Global mine supply last year was about 83,000 tonnes total, across all end uses.

That's the squeeze. And China's export controls on antimony (in force since September 2024) mean the material that used to move freely to Xinyi, Flat Glass Group and the Turkish and Indian glassmakers now moves under license, if at all.

Our stibnite occurrences in the Chilas and Astore belts run coarse-grained massive sulphide in quartz veining, with the better channel samples returning 18-42% Sb. The concessions aren't in production at commercial scale yet — I want to be honest about that — but the trenching data and the historical Soviet-era reconnaissance work (there's a 1978 UNDP report I keep on my desk) point to something that justifies a proper drill program. We're in discussion with two off-takers on pre-financing that work against forward tonnage.

Honestly, the constraint for antimony solar panels isn't demand. It's the number of politically stable jurisdictions with real stibnite geology. That list is short. Tajikistan, Turkey, Bolivia, Myanmar (complicated), Australia (small), and us.

Copper is the boring one that actually breaks first

Everyone talks about antimony because the price chart looks dramatic. But if you're building solar farms, copper is the metal that will quietly wreck your schedule.

A utility-scale PV plant uses roughly 2.8 to 4 tonnes of copper per megawatt when you count module ribbons, string cables, DC combiners, inverter windings, transformer coils and the MV cabling out to the substation. Rooftop and C&I systems use less per MW but more per panel. BloombergNEF's copper-in-solar estimate for 2030 sits above 1.1 million tonnes annually just for new PV builds. Add wind, grid expansion and EVs and the copper solar supply chain is competing with everything else that got electrified this decade.

Gilgit-Baltistan sits on the northwestern extension of the Kohistan island arc — same tectonic package that hosts the porphyry copper systems being drilled at Reko Diq down in Balochistan, just uplifted and dissected differently. Our concession work has identified porphyry-style alteration in two blocks with surface grab samples in the 0.4 to 1.1% Cu range, with associated Mo credits (which matters for the moly story I've written about separately). It's early-stage. I won't pretend we have a resource statement. But the geological argument for copper here is not speculative — it's the same arc that hosts Saindak.

Here's the thing about copper from Pakistan that overseas buyers keep missing: the logistics for concentrate out of the north are genuinely difficult. Karakoram Highway to Karachi is 1,750 km on a road that closes for landslides. The alternative — trucking north to Khunjerab and into Kashgar for rail onwards — actually works better for Chinese smelters, and that's part of why the concentrate flow tends to move east rather than west. Any European or American buyer needs to price that reality in from day one.

What buyers should actually ask

I've had maybe 40 serious inquiries this year from solar-adjacent buyers. The good ones ask three things:

Chain of custody documentation for the mining lease and the export permit — because critical minerals solar energy procurement in the EU is going to require it under the CRMA by 2026, and US buyers want it now for IRA compliance. We can provide it. The GB Mineral Concession Rules 2016 give us clean title and the export process through the Ministry of Commerce is documented.

Assays from a third-party lab, not from us. ALS Loadstone in Islamabad or SGS Pakistan. I'd rather a buyer trust an independent number than argue about ours.

Realistic tonnage. If you need 5,000 tonnes of antimony metal in Q1, I'm not your supplier and neither is anyone else in Pakistan. If you're building a 3-to-5 year off-take structure with pre-payment against development, that's a conversation worth having.

The solar industry spent fifteen years assuming the input metals would just show up. That assumption is now visibly wrong on antimony, getting shaky on copper, and already broken on tellurium and high-purity silver paste. What replaces it is slower, more relational, and more geographically awkward — which is fine by me, because that's the game we're actually set up to play.

If you're sourcing for a 2026 or 2027 module line and you haven't mapped your antimony exposure back to the mine, you're going to find out the hard way what the glassmakers already know.