Epitaxial Layer Strain Engineering via Stress Memorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for enhancing n-type field effect transistor (NFET) CMOS performance, such as stress engineering, face limitations in simultaneously incorporating substitutional carbon and stress memorization technique (SMT) due to competing substitutionality and irreversible displacement of carbon atoms, leading to stress loss and incompatibility with doping implants.
Innovation Solution
A method involving pre-amorphization implantation of carbon into an epitaxial layer, followed by forming a tensile cap and annealing to recrystallize the layer, enabling stress memorization technique (SMT) and maintaining tensile stress, allowing for higher n-type doping and compatibility with SMT.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If substitutional carbon is incorporated into the epitaxial layer to induce tensile stress and improve electron mobility, then NFET performance is enhanced, but the ability to perform subsequent amorphization for stress memorization technique is lost
Solution Approach 1:
The source/drain structure is divided into multiple epitaxial layers: a first epitaxial layer containing substitutional carbon for stress induction, and a second epitaxial layer without carbon that can be amorphized for SMT. This segmentation allows each layer to fulfill its specific function without interfering with the other process requirements.
Solution Approach 2:
Different regions of the source/drain structure are assigned different material compositions: the first epitaxial layer has high carbon concentration for tensile stress, while the second epitaxial layer has low or no carbon to enable amorphization. This local differentiation resolves the conflict between maintaining substitutional carbon and enabling SMT compatibility.
2Quantity of substance
If high concentrations of substitutional carbon and phosphorus are incorporated simultaneously to optimize doping, then resistance is reduced, but the two elements compete for substitutionality making it impossible to grow both at high concentrations
Solution Approach 1:
The doping function is segmented between two layers: the first epitaxial layer provides carbon for stress engineering, while the second epitaxial layer provides phosphorus for n-type doping. This avoids the competitive substitutionality issue by separating the two elements into different spatial zones.
Solution Approach 2:
The second epitaxial layer acts as an intermediary that enables high-concentration phosphorus doping without interfering with the carbon content in the first layer. This intermediate structure facilitates achieving both high carbon and high phosphorus concentrations in the overall source/drain region.
3Reliability
If amorphization is performed to enable stress memorization technique, then SMT effect is achieved, but substitutional carbon is irreversibly dislocated from lattice positions causing stress loss
Solution Approach 1:
The structure is segmented so that only the second epitaxial layer undergoes amorphization and SMT processing, while the first epitaxial layer with substitutional carbon remains unaffected. This spatial separation protects the carbon lattice positions from dislocation during amorphization.
Solution Approach 2:
The first epitaxial layer with substitutional carbon is formed and protected before the amorphization process is applied to the second layer. This preliminary structuring ensures that when amorphization occurs, the substitutional carbon in the first layer is already in place and shielded from displacement.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances channel electron mobility by maintaining tensile stress through SMT, improving NFET performance by increasing strain and resistance optimization without losing the SMT effect.
Implementation Method 1
performing a pre-amorphization implant (PAI) to form an amorphous layer in at least a portion of the epitaxial layer
Implementation Method 2
crystallizing the amorphous layer by annealing after forming the tensile cap to create an SMT effect
Implementation Method 3
forming a tensile cap over the amorphous layer to enable a stress memorization technique (SMT)
Implementation Method 4
Substitutional carbon induces a tensile stress that improves electron mobility in a channel of the NFET
Data Source
AI summary
A method and structure are disclosed for increasing strain in a device, specifically an n-type field effect transistor (NFET) complementary metal-oxide-semiconductor (CMOS) device. Embodiments of this invention include growing an epitaxial layer, performing a cold carbon or cluster carbon pre-amorphization implantation to implant substitutional carbon into the epitaxial layer, forming a tensile cap over the epitaxial layer, and then annealing to recrystallize the amorphous layer to create a stress memorization technique (SMT) effect. The epitaxial layer will therefore include substitutional carbon and have a memorized tensile stress induced by the SMT. Embodiments of this invention can also include a lower epitaxial layer under the epitaxial layer, the lower epitaxial layer comprising for example, a silicon carbon phosphorous (SiCP) layer.