The Home Depot mistake
There’s a homeowner in Tucson who bought four pallets of solar shingles from a clearance liquidator and felt extremely smart about it. The shingles were thin-film CIGS, they came in at $0.45/W, and the math looked perfect. He installed them himself on a studio apartment roof, hooked up the inverters, fired it up. The system generates about 60% of what an equivalent area of monocrystalline would have generated. He covered his entire roof and produced less power than a smaller mono install would have. His payback calculation broke. He’s been on Reddit posting about it for two years.
Thin-film solar is a fascinating technology that solves problems most homeowners don’t actually have. Let’s unpack when it’s the right call and when crystalline silicon will always win.
TL;DR
Crystalline silicon (mono or poly) is what virtually everyone should buy for residential rooftop. Thin-film is the right answer in three specific niches: very high temperature commercial installs, building-integrated applications where flexibility or aesthetics matter more than efficiency, and large-scale utility deployments where land is cheap. For typical residential, thin-film’s lower efficiency means it covers more roof for less power, and that math almost never works.
What’s actually being compared
Crystalline silicon panels are what we usually mean when we say “solar panels” — wafers cut from silicon crystals, assembled into cells, framed into rigid panels. Mono and poly are both crystalline. Efficiency: 15–23% at the panel level.
Thin-film panels deposit a very thin layer of photovoltaic material (just a few micrometers thick) directly onto glass, metal, or flexible substrate. Three main types:
- Amorphous silicon (a-Si): Cheap, low efficiency (6–10%), reasonably common in calculators and small panels.
- Cadmium Telluride (CdTe): First Solar’s specialty. Dominant in utility-scale thin-film. Module efficiency 17–19%.
- Copper Indium Gallium Selenide (CIGS): The high-end of thin-film. Module efficiency 13–16%.
The side-by-side
| Metric | Crystalline silicon | CdTe thin-film | CIGS thin-film | Amorphous silicon |
|---|---|---|---|---|
| Module efficiency | 15–23% | 17–19% | 13–16% | 6–10% |
| Temperature coefficient | -0.30 to -0.40%/°C | -0.20 to -0.28%/°C | -0.30 to -0.35%/°C | -0.20 to -0.25%/°C |
| Cost per watt (wholesale 2026) | $0.20–0.30 | ~$0.18–0.25 utility scale | $0.30–0.45 | $0.20–0.30 |
| Flexible? | No | No (glass) | Yes (some) | Yes |
| Best use case | Residential rooftop | Utility scale, hot climates | Building-integrated, flexible | Calculators, niche flexible |
| Major commercial brands | Jinko, Trina, Longi, Canadian, REC | First Solar | Solar Frontier, others discontinued | Various |
Round 1: Performance
Crystalline wins efficiency by a wide margin. Top-tier mono is 22% module efficiency; top-tier CdTe is 19%; CIGS is 16%; amorphous silicon is 10%. That means for a 6 kW system:
- Mono: ~14–16 panels covering ~30 m²
- CdTe (best thin-film): ~17–20 panels covering ~40 m²
- CIGS: ~22–26 panels covering ~50 m²
- Amorphous silicon: ~50+ panels covering ~90 m²
If you have a small roof, thin-film is essentially disqualified. If you have abundant ground or flat roof space, thin-film might compete on dollars-per-installed-watt — but only specific types like CdTe at utility scale.
The one place thin-film genuinely outperforms crystalline is high heat. Thin-film temperature coefficients are typically -0.20 to -0.30%/°C compared to -0.30 to -0.40 for mono. In a Phoenix-style climate where panels reach 65–70°C, that 0.1%/°C difference means thin-film loses less to heat. CdTe specifically delivers more annual energy than equivalent-rated mono in hot desert climates — measured in real installs by NREL and reported by First Solar.
Round 2: Cost & accessibility
This is the surprising part. CdTe at utility scale is genuinely competitive with mono on dollar-per-watt. First Solar’s Series 7 panels are deployed in gigawatt-scale projects across the US, Middle East, and India. The economics work because manufacturing CdTe doesn’t require polysilicon — sidestepping the supply chain that mono depends on.
Residential thin-film, however, is mostly disappointing. CIGS panels (Solyndra famously made these before going bankrupt; Stion, Miasolé, others have come and gone) usually come at $0.30–0.45/W, similar to mono, but you need 30–40% more of them to deliver the same power. The math doesn’t work for a homeowner.
Amorphous silicon is the only “cheap” thin-film — $0.20–0.30/W — but at 10% efficiency, you’re covering a huge area for very little power. Sometimes worth it if your roof is irregular or shaded, rarely a primary install choice.
Round 3: Real-world fit
Limited roof residential: Crystalline. Always. The efficiency advantage is decisive when every square meter counts.
Abundant roof or ground residential, hot desert climate: CdTe is worth considering. The heat advantage delivers a few percent more annual energy in real conditions. But the availability for residential is limited — First Solar mostly sells to utility scale.
Building-integrated (BIPV) projects: Thin-film wins on flexibility, weight, and aesthetic integration. Solar shingles, solar facades, solar awnings — these are mostly thin-film. They cost much more per watt but aren’t really competing on efficiency; they’re competing on “panels that don’t look like panels.”
Portable / flexible applications: Amorphous silicon and CIGS dominate flexible panels for RVs, boats, backpacks. Crystalline can’t be made flexible.
Utility-scale ground mount: CdTe (First Solar) is a real competitor with mono at this scale. Multi-gigawatt projects in 2024–2025 use CdTe extensively.
The Home Depot clearance bin: Walk away.
The honest verdict by use case
Crystalline is the default residential answer for the next decade. Thin-film exists for specific applications — desert utility, building-integrated, flexible — and it’s genuinely the better choice in those narrow cases. The mistake everyone makes is treating thin-film as a budget alternative to crystalline for the same use case. It almost never is. Mono got cheaper while thin-film stagnated, and the efficiency gap means you cover more roof for less power. If a deal sounds too good — Reddit-pallet-of-CIGS-shingles too good — it’s because the efficiency math hides the catch. Crystalline for your roof. Thin-film for the niches it owns.
FAQs
Are solar shingles always thin-film?
Mostly historically yes, but Tesla’s Solar Roof tiles are actually monocrystalline. The current generation of building-integrated shingles is mixed.
Is CdTe environmentally problematic because of cadmium?
Cadmium is toxic, but in CdTe panels it’s chemically bound and stable. The bigger concern is end-of-life — First Solar runs a recycling program recovering 95% of materials. Worth checking whether your specific manufacturer does the same.
Can I mix thin-film and crystalline in one system?
Only with separate strings or microinverters per panel — different voltage characteristics create bottlenecks if strung together.
Why did Solyndra fail if CIGS is good technology?
Polysilicon prices crashed in 2011 right after Solyndra committed to a tubular CIGS design optimized for higher polysilicon prices. Mono got cheap faster than CIGS could scale. Business problem, not technology problem.
What about perovskite thin-film?
Perovskite is a different category — mostly being commercialized as tandem cells on top of crystalline silicon rather than standalone thin-film. Covered separately.
If thin-film is bad for residential, why is it growing in utility scale?
Because First Solar’s CdTe avoids the polysilicon supply chain, gets domestic content advantages in US policy, and competes on dollar-per-watt installed. None of those advantages translate to a residential customer.