Spacer 2 deposition isolates strained S/D regions from spin-on processes, preserving epitaxial stressor geometry .
In depth
Device Context and Integration Logic
The S/D Spacer 2 Deposition step is introduced immediately after selective SiGe source/drain epitaxy to laterally isolate the raised epitaxial regions from subsequent patterning and implantation steps while preserving the effective channel length defined by the gate and first spacer system . Following SiGe epitaxy, the source/drain r
egions exhibit laterally expanded facets governed by crystallographic growth kinetics, as described for FinFET source/drain stressors in , which necessitates an additional dielectric barrier to prevent electrical or physical interaction with later spin-on and hard mask processes (Engineering Practice). This second spacer establishes a controlled separation between the strained S/D regions and the forthcoming S/D Si CSOH spin coat, thereby protecting the epitaxial stressor geometry and dopant distribution during subsequent lithography and etch steps . From a device physics standpoint, maintaining a well-defined spacer-controlled offset is essential to preserve short-channel electrostatics and parasitic resistance optimization in FinFET architectures, consistent with the raised source/drain concepts discussed in and .
Physical and Chemical Deposition Mechanism
S/D Spacer 2 Deposition relies on conformal dielectric film formation on three-dimensional FinFET topography, where uniform coverage along fin sidewalls and around faceted epitaxial regions is required to ensure consistent lateral isolation . Conformal deposition mechanisms, such as surface-reaction-limited growth, ensure that film thickness is governed by surface chemistry rather than line-of-sight transport, a principle consistent with thin-film formation theory derived from surface kinetics and crystal periodicity in . The deposited dielectric acts as a physical diffusion barrier, suppressing dopant migration and metal diffusion in later thermal or chemical steps, following the same isolation logic used in spacer-defined junction engineering described in . By physically separating materials with different chemical reactivities, the spacer mitigates unintended reactions at SiGe interfaces, which is critical given the enhanced diffusion and stress-driven transport phenomena known to occur near strained SiGe regions .
Material and Method Selection Rationale
The spacer material selected for this step prioritizes chemical inertness to SiGe, low diffusivity for common dopants and metals, and mechanical stability under subsequent process stresses . Dielectric materials with strong covalent bonding and high thermodynamic stability are favored because they resist plasma-induced damage and wet chemical attack during CSOH spin coat and hard mask deposition, consistent with integration strategies used for spacer-defined junction control in FinFETs . From a parameter interaction perspective, increased conformality improves sidewall coverage but also enhances parasitic capacitance, whereas higher film density improves diffusion blocking at the cost of increased stress transfer to the fin and channel . These trade-offs mirror the broader spacer optimization problem in scaled MOSFETs, where spacer width and material properties directly influence series resistance and electrostatic integrity, as discussed in .
Node-Specific Considerations for 14 nm FinFETs
At the 14 nm node, fin width and pitch scaling amplify the sensitivity of device performance to spacer-induced variations in effective channel length and stress distribution . The tall, narrow fins characteristic of this node exhibit strong coupling between sidewall processes and channel transport, making spacer uniformity along the fin height critical for maintaining consistent drive current, as demonstrated in high-aspect-ratio FinFET studies . Furthermore, the presence of highly strained SiGe source/drain regions introduces additional stress gradients that can interact with spacer mechanical properties, potentially altering strain transfer efficiency into the channel, a phenomenon rooted in heteroepitaxial strain physics described in . As a result, S/D Spacer 2 Deposition at 14 nm is not merely a geometric isolation step but a stress and electrostatics co-optimization element unique to aggressively scaled FinFET integration .
Risks & Challenges
[High] Spacer Non‑Conformality: Insufficient conformal coverage on fin sidewalls and faceted SiGe surfaces leads to local thinning, enabling unintended dopant diffusion or chemical attack during subsequent processing, driven by surface‑reaction‑limited deposition breakdown on high‑aspect‑ratio features .
[High] Stress‑Induced Channel Degradation: Excessive mechanical stress transferred from a dense spacer film into the fin can locally distort the silicon lattice, altering band structure and carrier mobility in the channel, consistent with strain–mobility coupling principles described in and observed in strained FinFET integrations .
[Medium] Interfacial Chemical Instability with SiGe: Chemical incompatibility between the spacer dielectric and exposed SiGe facets can promote interfacial defect formation or enhanced diffusion pathways, exacerbated by the higher reactivity and strain energy of SiGe compared to pure Si .
[Medium] Parasitic Capacitance Increase: Overly high‑k or excessively dense spacer materials increase gate‑to‑source/drain coupling capacitance, degrading switching speed through enhanced electrostatic coupling, following MOSFET parasitic capacitance principles outlined in .
[Low] Integration‑Induced Pattern Dependency: Variations in local pattern density can modify deposition kinetics and film density, leading to across‑device variability, a common integration challenge in spacer‑defined FinFET processes .
A free account opens the full rationale, risk analysis, and the paper and patent citations behind them.