Chapter 5.30

Contamination Monitoring of MOS-Clean Furnaces

(Tystar1&2, Tylan5&6)

1.0 Title

Contamination Monitoring of MOS-Clean Furnaces (Tystar1&2, Tylan5&6)

2.0 Purpose

Proper pre-furnace cleaning is required for all runs going into any of the Microlab furnace/s, as per guidelines specified in the pertinent Tystar/Tylan lab manual chapters. An additional procedure has also been defined to further safeguard/enhance the performance of our oxidation furnaces. These are TCA or TLC clean step prior to running the oxidation process, and regular monitoring of these furnaces by staff. The SCA surface charge analyzer (SCA) measurements (an electro-optical extension of C-V) are regularly taken on oxide monitor wafers to check the mobile ion contamination level of the furnaces (post TLC/TCA).

3.0 Scope

This manual outlines the three steps taken to test/control contamination monitoring of the MOS clean oxide furnaces.

3.1 TCA/TLC furnace cleaning

3.2 Growth of 300Å dry oxide

3.3 Using SCA for contamination test

4.0 Applicable Documents

Revision History

4.1 Microlab Online Manual Chapter 2.6 - Sink6 (MOS)

4.2 Microlab Online Manual Chapter 5.1 - Tystar1 MOS Clean Gate Oxidation Atmospheric Furnace (4" and 6")

4.3 Microlab Online Manual Chapter 5.2 - Tystar2 MOS Clean Dry/Wet Oxidation and Anneal Atmospheric Furnace (4" and 6")

4.4 Microlab Online Manual Chapter 8.02 - Surface Charge Analyzer (SCA)

5.0 Definitions & Process Terminology

5.1 TCA/TLC cleaning: A furnace cleaning process using a combination of oxygen, chlorine containing vapor (TCA: 1,1,1-Trichloroethane/TLC Trans: 1, 2-Dichloroethene), and high temperature (1100°C) to remove/immobilize the contamination on the quartz ware surfaces in the furnace.

5.2 It is recommended that the dry oxidation process be carried out immediately after the tube cleaning. Since the TCA process (for Tystar1 and Tylan5&6) takes 8 hours and the TLC process (for Tystar2) takes 4 hours, both recipes have a build option to delay the starting time. The cleaning recipe can be loaded with this delay the evening before. So the next morning, it will finish the process just before the dry oxidation.

5.3 D_it: The density of interface traps (unit: number-of-traps/cm²/ev)

5.4 IQF: Interface Quality Factor used for high quality Si/SiO₂ interface when the D_it is below the detection limit (dimensionless).

6.0 Safety

6.1 Do not modify any of the furnace recipes, as well as TCA/TLC recipes.

6.2 Follow the acid sink and furnace safety rules defined in their pertinent chapters, as well as recommendations noted in this chapter (section 4.0).

6.3 Read the Material data sheet for the TCA and TLC posted in the lobby of the Microlab in the blue binders.

6.4 TCA/TLC clean recipes have to be run at the presence of oxygen, which are set up, as such in the corresponding furnace recipe/s. Never attempt running TCA without oxygen, which could generate dangerous phosgene, cocl2 deadly gases.

7.0 Statistical/Process Data

Please refer to the Microlab process monitor Data for Tystar1 and Tystar2 (on the web).

8.0 Available Processes, Process Notes

The dedicated MOS-clean furnaces and specifically the oxidation furnaces must be free of any mobile-ion/metallic contamination. Gate oxidation tubes are most critical in terms of the cleanliness; therefore, special care/handing is required to ensure they are contamination free. This includes; proper pre-furnace cleaning of the process run/s going into these furnace/s, performing regular TCA&TLC cleaning/testing/monitoring of these oxidation furnaces to safeguard against mobile ion and metallic contamination/s. Surface Charge Analyzer is the tool used to test/monitor cleanliness of the MOS-clean oxidation furnaces (see, chapter 8.02 for more detail on surface change analyzer theory and operation).

9.0 Procedures

Furnace Cleaning

9.1 Enable the furnace to be cleaned on the WAND.

9.2 Load the TCA recipe, if you intend to use Tystar1 or Tylan5 or Tylan6 furnaces and the TLC recipe for Tystar2 (see Process Notes in Chapter 5.2).

Test Wafer Cleaning and Oxidation

9.3 Check out three p-type prime wafers from the office. Scribe them as load, center, and source wafers.

9.4 Clean the test wafers in Sink6 using the standard Piranha clean followed by a short HF dip. It is recommended to use fresh chemicals for these cleaning steps. After the wafer cleaning, use the MOS clean vacuum wand to handle the test wafers on their backside, only- DO NOT USE TWEEZERS.

9.5 Load corresponding oxidation recipe into the desired furnace to be tested:

9.5.1 Tystar1: recipe = 1GATEOXA, temperature = 950°C for the time = hour.

9.5.2 Tystar2: recipe = 2DRYOXA, temperature = 950°C for the time = 1 hour.

9.5.3 Tylan 5&6: recipe = SGATEOXA, temperature = 950°C for the time = 1 hour.

9.6 Transfer the test wafers to the furnace in a clean wafer transfer box. Run the recipe and load the test wafers, in the Load, Center, and the Source sections of the furnace tube.

Contamination Measurement

9.7 After the oxidation process is finished, use Nanospec/NanoDUV to measure the oxide film thickness. Be cautious and not to touch the front side of the test wafers.

9.8 Use SCA to measure the IQF for your test wafers. You need to use the OXIDE program, and by entering the oxide thickness obtained from Nanospec/NanoDUV machines (oxide thickness should be around 300 Å).

9.9 The IQF criteria for furnace cleanness is set at 1.5 If the IQF is larger than 1.5, the furnace is considered contaminated. In such case, the whole cleaning process should be repeated again.

9.10 If there is no significant improvement of IQF after the second cleaning, the tube is considered to be gross contaminated. Equipment and Process staff should be informed and the contamination source should be identified.

10.0 Troubleshooting Guidelines

N/A

11.0 Figures And Schematics

N/A

12.0 Appendices

N/A