Deposition of SiN with PECVD4/NEW TESTS QC

Unless otherwise stated, this page is written by DTU Nanolab internal (Feb 2025), for more info contact mfarin.

New QC tests on Silicon nitride - PECVD4

Due to recent instability in the High-Frequency (HF) generator, the thickness deposited during the QC HF Nitride process has not been very reliable. To better control it, new tests were performed. Since the recipe uses 25W of HF, and the main variation was the reflected power—oscillating between ~5W and 10W—higher power levels were tested. When using 100W, the reflected power was significantly more stable. Below, some of the data is presented, comparing 25W and 100W of HF power, as well as results from different characterization tools (Ellipsometer and FilmTek).

Data with HF generator=25W

Data from november to february of 15min QC nitride using 25W. The graph showcases thickness with standard error across time.

Wafers processed with 25W had the thickness measured in both Filmtek and Ellipsometer, comparing the results.

Data from november to february of 15min QC nitride using 25W. The graph showcases the refractive index with standard error across time.

Wafers processed with 25W had the refractive index measured in both Filmtek and Ellipsometer, comparing the results.

Reflected power of different tests, with 25W as the demanded power.

Data with HF generator=100W

Data from november to february of 15min QC nitride using 100W. The graph showcases thickness with standard error across time.

Wafers processed with 100W had the thickness measured in both Filmtek and Ellipsometer, comparing the results.

Data from november to february of 15min QC nitride using 100W. The graph showcases the refractive index with standard error across time.

Wafers processed with 100W had the refractive index measured in both Filmtek and Ellipsometer, comparing the results.

Reflected power of different tests, with 100W as the demanded power.