Monocrystalline vs Polycrystalline vs Thin-Film Solar Panels

One material, three ways of arranging it

Start with what is actually different, because the names obscure it. Monocrystalline and polycrystalline cells are both silicon, and differ only in how orderly the atoms are. Thin film is a family rather than a material — amorphous silicon, cadmium telluride, copper indium gallium selenide — united by being deposited as a layer a few microns thick rather than sliced from a block.

Drag sideways to see the rest

Enlarge

The manufacturing effort runs in the same order. Growing a single crystal is slow and exacting; letting a melt solidify into many crystals is neither; depositing a layer avoids the wafer entirely and uses perhaps a hundredth of the semiconductor. That effort is not spent on vanity — it is spent on the one thing that decides how much of the arriving light becomes usable current.

Why the boundaries cost you

A photon absorbed in the cell creates a mobile charge carrier. That carrier has to reach the junction and be collected before it recombines and gives its energy back as heat. Everything in a cell's efficiency comes down to how long it survives on that journey.

Drag sideways to see the rest

Enlarge

At a grain boundary the lattice is interrupted, leaving bonds unsatisfied. Those dangling bonds create energy states inside the bandgap, and those states are efficient traps: a carrier that finds one recombines, and its energy becomes heat. More boundaries means more chances to be trapped, which is why a single-crystal wafer collects a larger share of what it absorbs.

Amorphous material has no long-range order at all, so this problem would be crippling — except that the layer is so thin the carrier has very little distance to travel, and the material is deliberately hydrogenated to satisfy most of the dangling bonds. Different problem, different solution.

Even a flawless crystal has a ceiling that no manufacturing removes. Photons with less energy than the bandgap pass through unabsorbed, and the surplus energy of those with more is shed as heat, so a single-junction cell has a theoretical maximum fixed by physics: the Shockley–Queisser limit, first derived by William Shockley and Hans-Joachim Queisser in 1961, which comes to about 33% for an ideal bandgap under the standard solar spectrum. For crystalline silicon, whose bandgap is close to but not at that ideal and which loses further energy to Auger recombination, the practical theoretical ceiling is about 29.4% (Richter, Hermle and Glunz, Fraunhofer ISE, 2013). It is the photovoltaic counterpart of the Betz limit on wind turbines.

Efficiency is area efficiency

Here is the distinction that dissolves most arguments about which is best. A module's efficiency is its power divided by its area under standard test conditions. It is not a measure of how much of a kilowatt-hour you keep, or how well it works on cloudy days.

Drag sideways to see the rest

Enlarge

So it matters exactly when area is the binding constraint: a small roof with a power target it can only just meet. Where there is more surface than the target needs — a large flat roof, a ground mount, a field — the higher-efficiency module buys nothing that can be used, because the limit was never area in the first place.

This is also why comparing panels by efficiency alone is misleading. Two modules rated at the same power produce the same power. The efficient one simply does it in a smaller rectangle.

What each does in the real world

Standard test conditions are a laboratory. Outdoors, three behaviours separate the technologies more usefully than efficiency does.

Drag sideways to see the rest

Enlarge

Heat. Every cell loses voltage as it warms, the effect described in why panels lose efficiency in hot weather. Thin-film technologies generally have a gentler temperature coefficient than crystalline silicon, so they surrender less on the afternoons when the modules are hottest.

Weak light. In overcast and diffuse conditions — the situation examined in whether panels work on cloudy days — thin film typically holds a slightly larger fraction of its rating. The mechanism is partly spectral and partly the shape of the low-irradiance response.

Drag sideways to see the rest

Enlarge

Ageing. Crystalline silicon loses a small amount in its first hours of exposure through light-induced degradation, then declines slowly for the rest of its life. Amorphous silicon has a larger initial drop — the Staebler-Wronski effect — before stabilising, which is why its rating refers to the stabilised state.

Which differences actually decide anything

Set against each other, the picture is less dramatic than the marketing around it.

Two rows are worth reading against each other. Semiconductor used is why thin film exists at all: a hundredth of the material is a structural advantage, not a detail. And shape and substrate is the one difference no amount of efficiency closes — a rigid wafer cannot follow a curve, and some installations are curves.

Everything else on that list is a matter of degree. None of the three changes what a photovoltaic system is, none of them is a fix for a system producing less than expected, none of them escapes the way partial shade on a few cells drags down a whole module, and none of them changes the fact that getting the tilt and orientation right will move more energy than choosing between them ever will.

The honest summary

Choose by area first: if the roof is tight against the target, area efficiency is the constraint and monocrystalline resolves it. Choose by environment second: a hot climate or a persistently overcast one shifts the balance toward thin film. Choose by surface third: if the mounting is not flat and rigid, the decision has already been made for you.

And when none of those three applies — a plain roof with room to spare in a temperate climate — the difference between them is smaller than the difference a week of dust makes, and keeping the glass clean without damaging it is a more useful thing to spend attention on.