Home Knowledge Base Sidewall Image Transfer (SIT)

Sidewall Image Transfer (SIT) is the self-aligned patterning technique that uses the sidewall spacers deposited on a lithographically defined mandrel as the actual etch mask, enabling feature pitches half of (or less than) the minimum lithography pitch — the core mechanism behind all pitch-halving (SADP) and pitch-quartering (SAQP) multi-patterning schemes used at sub-20nm nodes where features must be patterned finer than the optical lithography resolution limit.

Why SIT Is Needed

SIT / SADP Process Flow (Pitch Halving)

1. Deposit mandrel layer (poly, TEOS, or amorphous Si)
2. Litho: Pattern mandrels at 2× target pitch → develop + etch mandrel
3. Spacer deposition: Conformal ALD oxide or nitride (thickness = target half-pitch)
4. Spacer etchback: Anisotropic RIE → removes horizontal spacer, leaves vertical sidewall spacers
5. Mandrel removal: Selective etch (removes mandrel, leaves spacers intact)
6. Spacers now at target pitch (2× the original feature count)
7. Use spacers as etch mask → transfer pattern into underlying material
8. Strip spacers

Pitch Relationship

SADP (Self-Aligned Double Patterning)

SAQP (Self-Aligned Quadruple Patterning)

Spacer Material Selection

Spacer MaterialSelectivity to MandrelSelectivity to Underlying LayerUse
SiO₂High (vs. poly mandrel)ModerateStandard SADP
Si₃N₄ModerateHigh (vs. oxide target)Metal layer SADP
TiO₂High (vs. amorphous Si mandrel)HighAdvanced SAQP

CD Uniformity in SIT

SIT Limitations

Sidewall image transfer is the patterning workhorse that enabled CMOS scaling from 20nm to 5nm — by exploiting ALD thickness as a precision CD ruler and self-alignment to eliminate overlay errors between mandrel and spacer, SIT consistently delivers sub-10nm features without requiring lithography tools beyond their physical capability, making it indispensable to every advanced node manufactured in the last decade.

sidewall image transfersitself aligned spacer patterningspacer lithographysit patterningpitch halving

Explore 500+ Semiconductor & AI Topics

From EUV lithography to CUDA optimization — search the full knowledge base or chat with our AI assistant.