SiO2 thin film creation in Diffusion furnace - Process Functional Modeling
The oxidation process is performed within the vertical furnace. The furnace temperature is created by three different heaters in three zones using different thermocouples.
The oxidation is performed in different steps:
- Wafers are loaded from FOUP (special plastic box for wafers storage and transformation) into the boat - up to 150 wafers in one run. The furnace is kept open at 300C, N2 gas is supplied to the furnace and coming out of the furnace to prevent penetration of air from outside
- The boat with the loaded wafers is moved up and loaded into the furnace
- The furnace with the loaded boat with wafers dwells at 300C to reach a stable temperature
- After the dwelling, the temperature is elevated up to the oxidation process temperature, and O2 gas is supplied to the furnace to provide the oxidation.
- After the completion of oxidation, the furnace is cooled down to 300C
- The boat is moved down and removed from the furnace
- The wafers are unloaded from the boat to the FOUP
- The thickness of SiO2 on the wafers is measured using the ellipsometric method.
It is proposed to separate the process in the operations as follows:
- Pre-loading - loading wafers onto the boat
- Loading - loading the boat with the wafers into the furnace
- Stabilization - dwelling and increasing the temperature up to the target
- Oxidation - O2 supply
- Unloading - removing the boat from the boat
- Boat cooling & wafers unloading - unloading the wafers from the boat to the FOUP
- SiO2 thickness measurement
PRE-LOADING - Loading wafers to boat
OE 0.34
LOADING - Loading boat to furnace
OE 0.22
STABILIZATION - Dwelling & Temp increase
OE 0.74
OXIDATION - O2 added
OE 0.98
UNLOADING - Boat down from furnace
OE 0.83
COOLING & WAFERS UNLOADING
OE 0.67
SiO2 THICKNESS MEASUREMENT
OE 4.67
Productive operations effectiveness
Effective
Ineffective
Operation types breakdown
Productive
Providing
Corrective
Metrology
| Operation Type | Does it increase cost? | Does it increase product value? | Recommendation |
|---|---|---|---|
Productive | Yes | Yes | Improve |
Providing | Yes | No | Eliminate |
Corrective | Yes | No | Eliminate |
Metrology | Yes | No | Eliminate |
| Operation | Type | Merit | Step time | Recommendation |
|---|---|---|---|---|
| PRE-LOADING - Loading wafers to boat | Providing | OE 0.34 | 1 Hour | Consider eliminating |
| LOADING - Loading boat to furnace | Providing | OE 0.22 | 20 Minutes | Consider eliminating |
| STABILIZATION - Dwelling & Temp increase | Corrective | OE 0.74 | 3 Hours | Do nothing and eliminate it when possible |
| OXIDATION - O2 added | Productive | OE 0.98 | 20 Minutes | Consider improving |
| UNLOADING - Boat down from furnace | Providing | OE 0.83 | 20 Minutes | Consider eliminating |
| COOLING & WAFERS UNLOADING | Corrective | OE 0.67 | 1 Hour | Do nothing and eliminate it when possible |
| SiO2 THICKNESS MEASUREMENT | Metrology | OE 4.67 | 20 Minutes | Consider the necessity of the information and eliminate it when possible |
Two operations were found as the most problematic: LOADING and UNLOADING - operations where wafers and the boat are moving within the furnace:
These two operations should be analyzed to find a solution to the problem
The conditions are as follows:
- Furnace is open
- Heaters keep the temperature of 300C
- Atmosphere - N2 flows from the top of the furnace down and comes out of the furnace
- Boat is empty
- Wafers are in the FOUP (box for keeping and moving the wafers)
- Robot picks the wafers from the FOUP and loads them into the boat
The operation's main purpose is loading the wafers into the boat
Product - Wafers
Operational Effectiveness – OE
Effective
Ineffective
OE 0.34
Operational Perfectness - OP
Basic functions
Components
Supersystems
OP 0.22
Furnace | 7 |
FOUP | 6 |
Boat | 6 |
Robot | 6 |
Heaters | 5 |
Controller | 3 |
Thermocouples | 2 |
Air | 12 |
N2 gas | 91 |
Looks like two separate loops: the boat with wafers and the furnace.
Air and N2 flow have no real function.
The conditions are as follows:
- Furnace is open
- Heaters keep the temperature of 300C
- Atmosphere - N2 flows from the top of the furnace down and comes out of the furnace
- Wafers are in the boat
- Elevator moves the boat with wafers into the furnace
The operation's main purpose is to load the boat with wafers into the furnace
Product - Boat
Operational Effectiveness – OE
Effective
Ineffective
OE 0.22
Operational Perfectness - OP
Basic functions
Components
Supersystems
OP 0.06
Heaters | 8 12 |
Furnace | 7 |
N2 Gas | 4 64 |
Controller | 3 4 |
Thermocouples | 2 3 |
Wafers | 4 |
Air | 24 |
Looks like a very problematic operation. Moving wafers insert the air residue into the furnace. N2 gas pushes the air residue to the bottom zone.
The cold wafers give the wrong reading to the bottom thermocouple (TC) - cold wafer cools the bottom TC, no correct feedback to the bottom TC because the wafers a moving up, a new cold wafer is coming to the bottom TC, etc. The "wrong information" of the bottom TC leads to the overheating of the bottom zone. After the loading, the bottom zone is overheated and consists of residual air.
The conditions are as follows:
- Furnace is closed
- Heaters keep the process temperature
- Atmosphere - N2 is kept within the closed furnace
- Boat keeps wafers
The operation's main purpose is to achieve an equal temperature within the furnace, to keep the wafers at the same temperature.
Product - Wafers within the furnace
Operational Effectiveness – OE
Effective
Ineffective
OE 0.74
Operational Perfectness - OP
Basic functions
Components
Supersystems
Boat | 10 |
Furnace | 10 42 |
Heaters | 9 |
N2 Gas | 6 |
Controller | 3 |
Thermocouples | 2 |
Air residue | 12 |
Unloading can also impact the thermocouple data
The conditions are as follows:
- Furnace is closed
- Heaters keep the process temperature
- Atmosphere - N2 is kept within the closed furnace
- Boat keeps wafers
- O2 is supplied to the furnace to ensure the interaction with wafers to create a SiO2 thin layer on the surface of the wafer.
The operation's main purpose is to create SiO2 layer
Product - SiO2 layer
Operational Effectiveness – OE
Effective
Ineffective
OE 0.98
Operational Perfectness - OP
Basic functions
Components
Supersystems
OP 0.18
O2 gas | 12 |
Furnace | 10 30 |
N2 gas | 9 10 |
Heaters | 8 |
Wafers | 8 |
Boat | 6 |
Air residue | 5 24 |
Controller | 3 |
Thermocouples | 2 |
The air residue is mainly located in the bottom part of the furnace - low zone are more affected
The conditions are as follows:
- Furnace is closed
- Heaters reduce the temperature down to 300C
- Atmosphere - N2 flows from the top of the furnace down and comes out of the furnace
- Wafers are in the boat
- Elevator moves the boat with wafers down out of the furnace
The operation's main purpose is to load the boat with wafers into the furnace
Product - Boat
Operational Effectiveness – OE
Effective
Ineffective
OE 0.83
Operational Perfectness - OP
Basic functions
Components
Supersystems
Elevator | 12 |
Air | 10 |
Furnace | 9 26 |
N2 gas | 8 |
Heaters | 8 24 |
Wafers | 6 20 |
Controller | 3 |
Thermocouples | 2 |
Unloading also can impact the thermocouple data
The conditions are as follows:
- Furnace is open
- Heaters keep the temperature of 300C
- Atmosphere - N2 flows from the top of the furnace down and comes out of the furnace
- Wafers are in the Boat
- Robot picks the wafers from the boat and loads them into the FOUP
The operation's main purpose is loading the wafers into the FOUP
Product - Wafers
Operational Effectiveness – OE
Effective
Ineffective
OE 0.67
Operational Perfectness - OP
Basic functions
Components
Supersystems
OP 0.46
Robot | 4 |
FOUP | 4 |
Air | 4 16 |
Boat | 4 |
N2 gas | 8 |
Furnace | No impact |
Nothing special has been found out. The only small point is the wafers are still hot, and outside air can interact with the wafers that are located within the boat. It is important that bottom-zone wafers will keep the temperature longer than the top-zone wafers because they are more closed to the massive bottom flange.
The thickness of the SiO2 thin film is measured using the ellipsometry method. The scheme of the measurements is shown below (the image was taken from the article):
The purpose of the operation is to generate a signal within the detector.
The product of the operation is the Detector.
Operational Effectiveness – OE
Effective
Ineffective
OE 4.67
Operational Perfectness - OP
Basic functions
Components
Supersystems
OP 0.19
Reflected light | 5 |
SiO2 thin layer | 3 3 |
Wafer | 2 |
Emitted light | 2 |
Stage | 1 |
Light source | 1 |
The measurement operations typically are not problematic. Nevertheless, during the measurement operation, functional modeling raised a concern. The quality of the low-zone wafers is different from the mid-zone and top-zone wafers. For instance, the stoichiometry of SiO2 might change. Even small variations of stoichiometry can result in the same change in optical constants that may affect the model of ellipsometric model and result in SiO2 thickness.
So, should be taken into account.
Similar projects
This project applies Functional Modeling to analyze the single-wafer wet etch process in semiconductor manufacturing. The model shows that wet etch defects are not caused only by chemistry, but also by interactions between the wafer, liquid flow, air, rotation, rinsing, drying, drain, and chamber environment. Key challenges include incomplete wetting, trapped air or bubbles, residue and particle redeposition, evaporation-driven watermarks, and contamination during rinse or drying. The project identifies improvement directions such as better wafer pre-wetting, controlled airflow, reduced evaporation, optimized rinse and drying conditions, and alternative chamber concepts such as face-down processing in a shallow liquid bath.
Wet cleaning is widely used in microchip manufacturing. Single wafer equipment is working as follows. A wafer rotates, and chemistry is poured from a movable nozzle. Water rinsing is performed at the end of the process. Loading of a new batch of the chemistry resulted in excursion - a strongly increased amount of defects was observed on the wafer after the processing. The project is dedicated to the failure analysis and creation of innovative solutions.
This project investigates particle-defect formation during SiO₂ dry etch. Functional modeling revealed that defects can be generated by two coupled mechanisms: chemical formation of SiO₂/SiOFₓ particles from SiF₄ interaction with moisture, and mechanical generation of particles by excessive ion bombardment and sputtering. The proposed improvement direction is to balance chemical etching and ion-assisted etching by properly adjusting the source and bias generators, reducing moisture, improving by-product evacuation, minimizing sputtering, and using pulsed or multi-step process recipes. The goal is to reduce particles and micromasking while maintaining etch rate, profile control, and process stability.