text stringlengths 1 2.97k |
|---|
furnace |
Message-ID: <YT2PR01MB43997A32C710C649441BA3C689989@YT2PR01MB4399.CANPRD01.PROD.OUTLOOK.COM> |
Hi all, |
We have a Tystar furnace dedicated to thermal oxidation. We do not have a chlorine source for tube cleaning. So far, only bare prime Si wafers cleaned in RCA solutions have been loaded in the furnace. MOS capacitors with 20 nm thermal oxide were fabricated and C-V measurements did not show hysteresis due to mobile ions and I-V measurements showed nice Fowler-Nordheim curves up to 20 V with no breakdown . Quasi-static CV curves showed low density of interface traps. We tried the Zerbst method to determine the minority carrier lifetime generation but somehow it did not work. |
For those of you who care about metal contamination in quartz furnaces, I am curious to know how you evaluate it. I learned about surface charge analyzing that can work on bare and dielectric coated wafers and measure the minority carrier lifetime. I have contacted Berkeley and Georgia Tech for a characterization service. I am also considering TXRF (Total X-ray Fluorescence) but the cost to analyze few wafers with a specialized analytical company is quite high. |
Any suggestion is welcome. |
Best, |
Lino Eugene, P.Eng., Ph.D., |
Micro/nanofabrication process engineer |
Quantum-Nano Fabrication and Characterization Facility |
Office of Research |
QNC 1611 |
University of Waterloo |
200 University Avenue West |
Waterloo, ON, Canada |
N2L 3G1 |
Ph: +1 519-888-4567 #37788 |
Cell: +1 226-929-1685 |
Websites: |
https://uwaterloo.ca/quantum-nano-fabrication-and-characterization-facility/ |
https://fab.qnc.uwaterloo.ca/ |
-------------- next part -------------- |
An HTML attachment was scrubbed... |
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20230413/27e72277/attachment.html> |
From saba.sadeghi at uwaterloo.ca Thu Apr 13 19:55:02 2023 |
From: saba.sadeghi at uwaterloo.ca (Saba Sadeghi) |
Date: Thu, 13 Apr 2023 23:55:02 +0000 |
Subject: [labnetwork] Sputtering issue |
Message-ID: <YT2PR01MB872648B730031AF09CDEFF34F0989@YT2PR01MB8726.CANPRD01.PROD.OUTLOOK.COM> |
Dear Labnetwork community, |
I am having an issue with magnetron-sputtering of tantalum with DC power. After a few minutes of sputtering Ta glow spots appear on the surface of the target that eventually cause the plasma to distinguish. I have loaded two fresh 0.31?-thick Ta targets, but both times the same problem occurred, and both times it generated enough particles in a short amount of time ( about 2 hours of sputtering, accumulatively) that led to shorting of the source. |
I suspected that this might be a cooling issue, but there is enough flow rate at the sputtering source, and I use indium thin foil for thermal contact between the cathode and the target (recommended by the company- MeiVac). |
Have you experienced such a problem? I am wondering if this could be an issue with the sputtering source or could somehow the two Ta targets I loaded have been compromised? |
Thank you very much. |
Best regards, |
Saba Sadeghi |
IQC, University of Waterloo |
-------------- next part -------------- |
An HTML attachment was scrubbed... |
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20230413/c2632cfe/attachment.html> |
From jerry.hunter at wisc.edu Thu Apr 13 20:02:08 2023 |
From: jerry.hunter at wisc.edu (Jerry Hunter) |
Date: Fri, 14 Apr 2023 00:02:08 +0000 |
Subject: [labnetwork] Evaluation of heavy metal contamination in |
oxidation furnace |
In-Reply-To: <YT2PR01MB43997A32C710C649441BA3C689989@YT2PR01MB4399.CANPRD01.PROD.OUTLOOK.COM> |
References: <YT2PR01MB43997A32C710C649441BA3C689989@YT2PR01MB4399.CANPRD01.PROD.OUTLOOK.COM> |
Message-ID: <502669FA-F247-4182-9227-F360FD2C582E@wisc.edu> |
TXRF will detect metallic contamination on the very surface down to ppm levels. |
If you are looking for more the bulk (>5nm) of the film with depth profiling then SIMS is a better choice (as long as the concentration is less than 1%) and will give you detection down to ppb levels. If the concentration is > 0.1% then XPS is a good choice and easier to do than SIMS. |
Jerry Hunter |
On Apr 13, 2023, at 6:42 PM, Lino Eugene <lino.eugene at uwaterloo.ca> wrote: |
? |
Hi all, |
We have a Tystar furnace dedicated to thermal oxidation. We do not have a chlorine source for tube cleaning. So far, only bare prime Si wafers cleaned in RCA solutions have been loaded in the furnace. MOS capacitors with 20 nm thermal oxide were fabricated and C-V measurements did not show hysteresis due to mobile ions and I-V measurements showed nice Fowler-Nordheim curves up to 20 V with no breakdown . Quasi-static CV curves showed low density of interface traps. We tried the Zerbst method to determine the minority carrier lifetime generation but somehow it did not work. |
For those of you who care about metal contamination in quartz furnaces, I am curious to know how you evaluate it. I learned about surface charge analyzing that can work on bare and dielectric coated wafers and measure the minority carrier lifetime. I have contacted Berkeley and Georgia Tech for a characterization service. I am also considering TXRF (Total X-ray Fluorescence) but the cost to analyze few wafers with a specialized analytical company is quite high. |
Any suggestion is welcome. |
Best, |
Lino Eugene, P.Eng., Ph.D., |
Micro/nanofabrication process engineer |
Quantum-Nano Fabrication and Characterization Facility |
Office of Research |
QNC 1611 |
University of Waterloo |
200 University Avenue West |
Waterloo, ON, Canada |
N2L 3G1 |
Ph: +1 519-888-4567 #37788 |
Cell: +1 226-929-1685 |
Websites: |
https://uwaterloo.ca/quantum-nano-fabrication-and-characterization-facility/ |
https://fab.qnc.uwaterloo.ca/ |
_______________________________________________ |
labnetwork mailing list |
labnetwork at mtl.mit.edu |
https://mtl.mit.edu/mailman/listinfo.cgi/labnetwork |
-------------- next part -------------- |
An HTML attachment was scrubbed... |
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20230414/41859065/attachment.html> |
From jeanne.guo at rice.edu Fri Apr 14 10:25:05 2023 |
From: jeanne.guo at rice.edu (Jing Guo) |
Date: Fri, 14 Apr 2023 09:25:05 -0500 |
Subject: [labnetwork] Sputtering issue |
In-Reply-To: <YT2PR01MB872648B730031AF09CDEFF34F0989@YT2PR01MB8726.CANPRD01.PROD.OUTLOOK.COM> |
References: <YT2PR01MB872648B730031AF09CDEFF34F0989@YT2PR01MB8726.CANPRD01.PROD.OUTLOOK.COM> |
Message-ID: <70A3D258-EC7F-4534-A275-539D7F366AA9@rice.edu> |
Hi Saba, |
Was the process a pure Ta sputtering or reactive sputtering? For a pure Ta DC power sputtering process, usually it shouldn?t generate that much particles. Ta coating on the sputter gun house/chimney should have a good adhesion. Any accumulated particles on the Ta target surface after ~ 2 hours sputtering? |
If the chimney or clamp ring for the target has been coated too much materials, you need to clean the whole kit and measure the gun to the ground layer by layer once you install a target. Another issue could be that some magnets got demagnetized which cannot provide strong magnetic field on the target surface. |
What was the strike condition, like pressure, power, gun voltage/current? Did you see whether the voltage/current can stabilize once the plasma was generated? |
There are several ways to figure out it is the sputter gun issue or a Ta target issue. Normally targets should be fine. |
Thanks. |
Best, |
Jing |
--------------------------------------- |
Jing Guo Ph.D. |
Research Scientist |
SEA Cleanroom (SST 017) |
Rice University |
Houston, TX |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.