Spatial ALD
Spatial, temporal, and batch ALD reactors all implement the same self-limiting half-reaction chemistry that defines atomic-layer deposition, but they differ in how they physically separate the two half-reactions from each other — in time, in space, or across many wafers at once — and that choice drives throughput more than any chemistry decision does.
Spatial ALD exists specifically to remove the purge-time bottleneck that limits temporal ALD's throughput. In temporal ALD, the reactor has to fully purge reactant A out of the chamber before introducing reactant B, or the two would react in the gas phase instead of at the surface — and that purge step, repeated every single cycle, is often the slowest part of the whole process. Spatial ALD sidesteps the purge entirely by keeping reactant A and reactant B in permanently separate physical zones and moving the wafer back and forth between them, so there's no shared volume that ever needs clearing.
Batch ALD is really temporal ALD's cycling scaled across many wafers at once, inheriting both temporal ALD's purge-time cost and the throughput benefit of processing a full load simultaneously. The individual cycle still pulses reactant A, purges, pulses reactant B, purges — exactly as single-wafer temporal ALD does — but because dozens of wafers go through each cycle together, the total wafers-per-hour can rival or beat spatial ALD even without eliminating the purge step, simply by spreading that fixed per-cycle time cost across a much larger wafer count.
The choice between temporal, spatial, and batch ALD comes down almost entirely to throughput engineering rather than film chemistry — all three can grow the same film with the same atomic-layer precision, and the reactor configuration a fab picks reflects how many wafers it needs through that specific ALD step per hour.