Narration script

Unit four. The same element can be one perfect crystal, a patchwork of many crystals called grains joined at grain boundaries, or a disordered amorphous network. For solar cells, order matters, because boundaries and disorder create electronic defects. That is why the industry moved from multicrystalline to monocrystalline wafers, and why thin films need special grain-boundary treatments.

You will learn to define the different degrees of order, explain why a grain boundary can recombine carriers, compare grain size with how far carriers travel, and say where each form is used in solar today.

Wafer type was the defining technology choice of the last decade. Cast multicrystalline silicon was cheaper, but monocrystalline wafers gave higher efficiency and textured better. Today mono dominates; n-type monocrystalline alone was about eighty-two percent of crystalline silicon production in twenty twenty-five, according to the I T R P V roadmap. Meanwhile, amorphous silicon came back as the ultra-thin passivation layer of heterojunction cells.

This schematic shows the three degrees of order in two dimensions. On the left, a single crystal, with one orientation throughout. In the middle, a polycrystal, with grains of different orientations meeting at boundaries. On the right, amorphous: each atom still has about four neighbours, so there is short-range order, but no long-range repeating pattern.

The table lists the forms used in solar. Monocrystalline, one crystal for the whole ingot, grown by Czochralski pulling, is the dominant wafer type today. Multicrystalline, with millimetre to centimetre grains made by casting, was the majority until about twenty eighteen and is now marginal. Polycrystalline thin films, with grains around a micrometre, are used for cadmium telluride, C I G S and perovskites. And amorphous hydrogenated silicon, deposited from silane gas, is used in heterojunction passivation.

Where two grains meet, atoms cannot satisfy all their bonds. The boundary contains strained and dangling bonds with energy levels inside the band gap. These act as recombination centres, and they also attract metal impurities. The key comparison is grain size versus diffusion length, how far a carrier travels before recombining. If grains are much larger than the diffusion length and the wafer thickness, boundaries hardly matter. If grains are comparable or smaller, many carriers reach a boundary and are lost.

Some estimates. A good n-type mono wafer may have a diffusion length over a millimetre, far more than its one hundred and thirty micrometre thickness. A multicrystalline wafer might reach one to three hundred micrometres in good grains, but much less near boundaries. A cadmium telluride film with one micrometre grains is only about three micrometres thick, so every carrier is close to a boundary. That is why cadmium telluride needs a cadmium chloride treatment to passivate its boundaries.

In amorphous silicon, each atom still has roughly four neighbours, but bond angles and lengths vary randomly. Many dangling bonds remain, and adding about ten percent hydrogen ties most of them off. Amorphous silicon has a wider gap, about one point seven to one point eight electron-volts, and absorbs strongly, but it transports charge poorly and degrades under light. So amorphous thin-film modules had low efficiency. But as a layer just five to ten nanometres thick on crystalline silicon, it gives superb surface passivation, the basis of heterojunction cells with very high voltage. The catch: after deposition it must stay below roughly two hundred to two hundred and fifty degrees Celsius, or the hydrogen escapes. That forces low-temperature metallization, with silver pastes cured at about two hundred degrees instead of fired.

So why did mono win? Fewer defects, giving higher lifetime and voltage. The one zero zero orientation, which allows alkaline pyramid texture. Diamond wire sawing and cheaper Czochralski growth, which erased the cost advantage of casting. And compatibility with high-efficiency n-type cell designs, which need high lifetime.

Misconceptions. In industry, polysilicon usually means the feedstock chunks, not a type of wafer. Amorphous does not mean no structure at all. And grain boundaries are not always bad; in treated cadmium telluride and perovskites, they can be benign. What matters is their electronic states.

For the job: crystal growth aims for dislocation-free single crystals, and a loss of structure during pulling is a yield event engineers track closely. Photoluminescence imaging and lifetime mapping reveal grain boundaries and defect clusters. And heterojunction lines must carefully manage the thermal budget of amorphous silicon.

Try the review questions. Next unit, we zoom in on defects themselves: what they are, and how a few parts per billion of the wrong atom can ruin a solar cell.