Amorphization

Amorphization occurs when a heavy enough ion implant dose displaces more silicon atoms than the crystal lattice can tolerate while remaining crystalline, causing the ordered lattice structure to collapse into disordered amorphous silicon. Rather than being purely an unwanted side effect of ion implantation, amorphization is deliberately exploited through pre-amorphization implant (PAI) techniques that intentionally destroy crystal structure before the actual dopant implant, specifically to eliminate the channeling problem that would otherwise distort a subsequent dopant's depth profile.

The damage-cascade mechanism underlying amorphization is straightforward in principle: each individual implanted ion displaces a cascade of target atoms along its path, and once the local density of displaced atoms crosses a critical threshold — roughly 10 percent displacement — the crystalline structure simply can't maintain its ordered arrangement any longer and collapses into an amorphous state. This threshold-crossing behavior is exactly why amorphization is a genuine phase transition rather than a gradual degradation — below the critical damage density the lattice remains recognizably crystalline despite accumulated defects, and above it the material transitions wholesale into amorphous silicon, with the transition itself depending on ion mass, energy, dose, and substrate temperature all acting together.

Pre-amorphization implant deliberately exploits this threshold behavior by using germanium or silicon ions specifically to amorphize the near-surface region before the actual dopant implant runs, which eliminates channeling for that subsequent dopant implant because amorphous material, by definition, has no ordered crystal structure containing the open channels dopant ions would otherwise travel through. This is exactly why PAI has become standard practice specifically for shallow boron implants, where channeling is a particularly severe problem given boron's light mass — PAI addresses the channeling problem at its physical root by removing the crystal structure boron would otherwise channel through, rather than merely compensating for channeling through tilt angle adjustments alone.

Solid-phase epitaxial regrowth (SPER) is the mechanism by which an amorphized layer recovers crystalline order during subsequent annealing, with recrystallization propagating upward from the still-crystalline substrate interface beneath the amorphous layer, typically at anneal temperatures of 500 to 700°C — a notably fast, low-temperature activation mechanism relative to what would otherwise be required to activate dopants in undamaged crystalline silicon. This low-temperature, fast-regrowth characteristic is a genuine practical advantage of the PAI-plus-dopant-implant approach: the resulting amorphous layer actually recrystallizes and activates dopants more efficiently at a lower thermal budget than implanting directly into crystalline silicon would require, converting what started as implant damage into a net processing benefit.

End-of-range (EOR) defects, forming as dislocation loops at the depth beyond where the amorphous-to-crystalline interface originally sat, represent damage that solid-phase epitaxial regrowth cannot fully heal, because SPER recrystallizes the amorphous region itself but doesn't repair damage that extends past that region's original boundary into the still-crystalline substrate below it. These EOR defects matter enormously in practice because they can cause genuine junction leakage if they end up positioned too close to where the actual electrical junction forms — which is exactly why amorphization depth has to be carefully controlled to stay safely above the intended junction depth, ensuring any EOR defects that do form stay comfortably separated from the junction rather than interfering with it.

Implant temperature directly controls how readily amorphization occurs, because implanting at elevated substrate temperature allows a degree of dynamic self-annealing to happen concurrently with the implant itself, continuously repairing some fraction of the displacement damage as it accumulates and thereby raising the effective dose threshold needed to actually achieve amorphization — while cryogenic implant temperatures suppress that same dynamic annealing and therefore maximize amorphization at a given dose. This temperature-dependent tradeoff gives process engineers a genuine additional control knob beyond ion species, energy, and dose alone: a process specifically wanting to maximize amorphization depth and uniformity for PAI purposes can use cryogenic implant conditions, while a process wanting to minimize unwanted amorphization from an otherwise unrelated implant step can instead implant at elevated temperature to suppress it.