Monolayer Deposition System Multi-Orifice Injector
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Atomic Layer Deposition (ALD) processes are limited by slow film growth rates and lack of in-situ measurement capabilities, leading to suboptimal process control and throughput in semiconductor device manufacturing.
Innovation Solution
A monolayer deposition (MLD) processing system utilizing intelligent set points for controlled precursor and purge processes, combined with multi-orifice injectors and dynamic modeling to optimize gas flow rates and chamber pressure, enabling faster cycle times and improved uniformity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional ALD processes are used with single-orifice injectors, then film uniformity is maintained, but cycle time increases and throughput decreases
Solution Approach 1:
The single-orifice injector is divided into multiple orifices (e.g., 4-8 orifices per injector) that are distributed across the reactor chamber. This segmentation allows simultaneous precursor delivery to multiple wafer positions, reducing cycle time while maintaining uniformity through controlled distribution of precursor flux across the chamber.
Solution Approach 2:
The invention transitions from a single-point precursor injection approach to a distributed spatial distribution approach by positioning multiple orifices at different locations within the reactor chamber. This dimensional expansion of the injection system enables parallel processing across multiple wafer positions without compromising film uniformity.
2Device complexity
If single-orifice injectors are used, then system complexity is low, but the time to fill reactor with precursors increases
Solution Approach 1:
The injector system is segmented into multiple orifices that can operate simultaneously, reducing the time required to fill the reactor with precursor. Each orifice contributes to the overall precursor flux, allowing faster saturation of the reactor volume while maintaining manageable complexity through modular design.
Solution Approach 2:
Multiple orifices enable continuous and simultaneous precursor delivery to different regions of the reactor chamber, eliminating the sequential filling process required by single-orifice systems. This continuous action accelerates the precursor filling time without introducing significant additional complexity.
3Reliability
If ALD processes operate at lower temperatures, then flux-independent deposition is achieved, but deposition rate decreases
Solution Approach 1:
By segmenting the precursor injection into multiple orifices distributed throughout the chamber, the system achieves more uniform precursor distribution at lower temperatures. This spatial segmentation compensates for the reduced thermal energy by ensuring every surface receives adequate precursor flux, maintaining deposition rate while achieving flux-independent behavior.
Solution Approach 2:
The invention changes the spatial distribution parameter of precursor delivery by introducing multiple orifices at different positions and orientations. This parameter change allows the system to maintain effective deposition rates at lower temperatures by optimizing the distribution of precursor molecules across the reactor chamber and wafer surfaces.
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
The MLD system significantly reduces cycle time and enhances film uniformity and throughput by dynamically controlling gas flow and pressure, addressing the limitations of conventional ALD methods.
Implementation Method 1
The technique of ALD is based on the principle of the formation of a saturated monolayer of reactive precursor molecules by chemisorption
Implementation Method 2
This is followed by injecting precursor or reactant B (RB) into the chamber, also for a period of time, to combine B with A thus forming the layer AB on the substrate
Data Source
AI summary
An adaptive real time thermal processing system is presented that includes a multivariable controller. The method includes creating a dynamic model of the MLD processing system and incorporating virtual sensors in the dynamic model. The method includes using process recipes comprising intelligent set points, dynamic models, and/or virtual sensors.


