|
|||||||||||||||||||||||||
| Home > Resources > Application Notes > In Situ | |||||||||||||||||||||||||
|
|
Application Notes:
|
||||||||||||||||||||||||
Characterizing Processes with CompleteEASE®: In Situ Applications |
|
The application of in situ Spectroscopic Ellipsometry (in situ SE) to process monitoring and control has long been recognized. However, did you know that in situ SE can simplify everyday process characterization and development? Recent innovations incorporated into the Woollam Co.’s CompleteEASE software use Virtual Interface (VI) models to readily determine deposition (or etch) rates and optical constants of the top-most film without requiring a detailed model of underlying layers. VI models allow sequential analysis of each layer of a multi-film stack as they are deposited during a single run. By varying process parameters (e.g. gun current, gas flow rates, ...) for each layer during the run, one can easily characterize film deposition rates and optical constants for various conditions. Why use a Virtual Interface?Characterization of film deposition, etching or other processing consists of determining film growth (or etch) rates and optical constants, n(λ) and k(λ), for various process conditions.Obviously, film growth rates can be readily determined from in situ SE data by calculating the change in film thickness over time. Simultaneously, one can also extract the optical constants from in situ SE data. In traditional applications the quality of the analysis depends on optical model accuracy. ‘Virtual Interface’ models simplify analysis by combining the optical response of underlying layers into a ‘pseudo substrate’, as shown in Figure 1. During analysis, the layer above the VI is modeled, while the VI itself remains unchanged. Characteristics of the near surface film can be continuously tracked by constantly updating the location of the VI, placing it at some fixed time interval prior to the most recent in situ SE data. The “depth” of the VI is then determined by the time interval and deposition rate. |
|
![]() Figure 1. Schematic of a Virtual Interface model. |
|
|
|
Figure 2. Evolution of ellipsometric Ψ at five wavelengths acquired during a 160min. deposition of 19 different films of Ta, Si, or SiOx using various gun currents and O2 gas flow rates. |
|
Figure 2 shows the evolution of Ψ at 5 wavelengths, as different layers are sequentially deposited on a starting substrate (consisting of ~25 nm thermally-deposited SiO2 on silicon). The 19 layers included Ta and Si films deposited at sputter-gun currents of 0.1 to 0.5 amps, and the SiOx films deposited using O2 flows of 0 to 16 sccm. Figures 3a and 3b show the GROC layer model for the third and sixth layers of the deposition run. As the figures indicate, the 3rd layer was tantalum deposited at a rate of 2.62 Å/second, while the 6th layer was a-Si deposited at a rate of 0.46 Å/second. In the figures, the optical constants for each film are shown in the bottom-left graphs. |
|
Figure 3a and 3b. GROC layer model for 3rd layer (run time 11 to 14 min.) and 6th layer (run time 25 to 31 min.) respectively, using a Ta gun (ITa-gun= 0.4Amps) and Si gun (ISi-gun = 0.15Amps). Arrows indicate deposition rate and optical constants. |
|
Figures 4a and 4b show the optical constants and deposition rate for the eleventh and twelfth layers of the deposition run. As the figures indicate, the deposition rate was 3.95 Å/sec when the O2 flow rate = 8 sccm, compared to 0.45 Å/sec for an O2 flow rate=16 sccm, a 9x difference that can be attributed to oxygen poisoning of the Si sputter target. Once modeling of all the different layers is completed, chamber performance is analyzed. As an example, the Ta and Si deposition-rate calibration chart is shown in Fig. 5. |
|
a) Rate = 3.95 Å/Sec b) Rate = 0.45 Å/Sec |
|
Figure 5. Ta and a-Si film deposition rate calibration chart. |
|
References1. B. Johs, J. Hale, N.J. Ianno, C.M. Herzinger, T.E. Tiwald, and J.A.Woollam, Proc. SPIE 4449, 41 (2001).2. B. Johs, presented at ICSE-3, accepted for publication (Thin Sol. Films). |
|