3B: Review. Nina Hong. U Penn, February J.A. Woollam Co., Inc. 1

Transcription

1 2014 J.A. Woollam Co., Inc B: Review Nina Hong U Penn, February 2014

2 2014 J.A. Woollam Co., Inc. 2 Review 1A: Introduction to WVASE: Fun Quiz! 1B: Cauchy 2A: Pt-by-Pt and GenOsc 2B: Advanced GenOsc 3A: Non-Idealities

3 2014 J.A. Woollam Co., Inc. 3 Q1 If the left column shows the Psi and Delta for an ideal glass substrate, can you identify the complexity of the right column data? Ideal Surface Roughness

4 2014 J.A. Woollam Co., Inc. 4 Q2 If the left column shows the Psi and Delta for an ideal glass substrate, can you identify the complexity of the right column data? Ideal Backside Reflection

5 2014 J.A. Woollam Co., Inc. 5 Q3 If the left column shows the Psi and Delta for a Si substrate, can you identify the complexity of the right column data? Ideal Backside Reflection

6 2014 J.A. Woollam Co., Inc. 6 Q4 What is the substrate type of the following ellipsometric data? 1Transparent substrate (dielectric) 2 Semi-absorbing substrate (semiconductor) 3 Absorbing substrate (Metal)

7 2014 J.A. Woollam Co., Inc. 7 Q5 What does ellipsometric measurement probe? 1 Size of electric field 2 Shape of electric field Less Intense Y E Different Size (Intensity). More Intense Y E X Same Shape! (Polarization) X

8 2014 J.A. Woollam Co., Inc. 8 Q6 Pseudo Optical Constants <n>, <k> Which one is corresponding to a single reflection on surface? 1 2

9 2014 J.A. Woollam Co., Inc. 9 Q7 Unknown Film on Glass Which spectral range does the film look transparent?

10 2014 J.A. Woollam Co., Inc Data Analysis Strategies Substrates S-1: Opaque Substrates S-2: Semiconductor substrates S-3: Transparent substrates Films F-1: Transparent films F-2: Semi-Absorbing films F-3: Absorbing films

11 Psi in degrees Delta in degrees 2014 J.A. Woollam Co., Inc S-1 Opaque Substrates Psi stays near 45 Delta away from 0 or

12 2014 J.A. Woollam Co., Inc S-1 Opaque Substrates Two Categories Bulk Samples No overlayers. Polished metal. Optically thick films. Fit Strategy Invert psi and delta for n and k. Normal fit from reference values 2 Known: Ψ, Δ 2 Unknowns: n, k Example 1 Optically Thick Cr Film

13 2014 J.A. Woollam Co., Inc S-2 Semiconductor Substrates Psi follows shape of absorption Delta away from 0 or 180 when absorbing Y in degrees D in degrees

14 2014 J.A. Woollam Co., Inc S-2 Semiconductor Substrates Bulk semiconductor n&k are well known Crystalline structure insures repeatable optical/electrical properties. Doping NOT important at VIS wavelengths Doping Is important in the MID-IR, however (5 micron or longer). Fit Strategy Use published index values. Fit for oxide thickness. Example 2 Bare Si Wafer

15 2014 J.A. Woollam Co., Inc S-3 Transparent Substrates Psi flat and smooth- follows shape of index Delta = 0,180 - except for surface films Y in degrees D in degrees nm roughness

16 2014 J.A. Woollam Co., Inc S-3 Transparent Substrates Fit Strategy Cauchy Complexities Surface roughness (Add srough) Backside reflection (Suppress or Model correction) How to obtain small k-values Add Transmission data and fit k only. 25 Example 3 10mm bk7 (Two dat files: SE and T) Experimental Data 200 Experimental Data 1.0 Experimental Data Y in degrees Exp E 55 Exp E 57.5 Exp E 60 Exp E 62.5 Exp E 65 Exp E 67.5 Exp E 70 D in degrees Exp E 55 Exp E 57.5 Exp E 60 Exp E 62.5 Exp E 65 Exp E 67.5 Exp E 70 Transmission Exp pt

17 2014 J.A. Woollam Co., Inc F-1 Transparent Films Each reflected wave will have a different phase and amplitude. ~n 0 ~n 1 ~n 2 Delay caused by both index and thickness

18 2014 J.A. Woollam Co., Inc Interference shifts toward red as film grows. More Interference oscillations as film grows. 100 F-1 Thickness Effects 0nm 25nm 50nm 75nm 100nm 200nm 300nm 400nm 500nm Y in degrees

19 2014 J.A. Woollam Co., Inc F-1 Effect of Film Index Index difference affects Interference oscillations (Mostly Psi amplitude). n=1.5 n=1.75 n=2.0 n=2.25 n=2.5 n=2.75 n=3.0 n=3.25 n=3.5 n=3.75 n=4.0 Y in degrees

20 2014 J.A. Woollam Co., Inc F-1 Using This Information Adjust Index to match Psi peak height. Adjust Thickness to match oscillation period. 90 Experimental Data Y in degrees n=1.5 n=2 n=2.5 n=3 Exp E Y in degrees Generated and Experimental Model Fit Exp E

21 F-1 Cauchy Procedure 1. Adjust An parameter to approximate index. Psi amplitude can help estimate. 2. Adjust thickness to match # of oscillations. 3. Fit Thickness, An and Bn. 4. Add Cn. Does it improve the MSE? 5. Normal dispersion??? Normal Fit IF Model is Close to Answer! RESET if fit fails!!! Example 4 Al2O3 on Si 2014 J.A. Woollam Co., Inc. 21

22 2014 J.A. Woollam Co., Inc F-2 Semi-Absorbing Films If light is absorbed before returning to surface, only top reflection is seen (Film appears as Substrate) D in degrees

23 2014 J.A. Woollam Co., Inc Fit Cauchy to Transparent region determine thickness. F-2 Choosing a Model Fix thickness and fit n,k at all wavelengths using point-by-point fit. When in doubt, convert to Genosc to insure Kramers-Kronig consistency. Surface roughness, grading, and anisotropy can be added if they improve MSE. Example 5 Semi Absorbing Film on Si

24 2014 J.A. Woollam Co., Inc F-2 Surface Roughness / Index Grading Roughness sensitivity when material is absorbing or has high index. Grading and roughness often correlated when both lower index toward surface. When grading increases index toward surface, may also be sensitive to roughness.

25 F-3 Absorbing Films Most metal films are opaque above nm thickness. Absorption at all wavelengths prevents periodic oscillations from thickness. Data will appear similar to absorbing substrates: Psi large and never close to zero. Delta between 0 and 180. Pseudos from different angles the same J.A. Woollam Co., Inc. 25

26 2014 J.A. Woollam Co., Inc F-3 Multiple angles Y in degrees D in degrees Exp E 45 Exp E 55 Exp E 65 Exp E Exp E 45 Exp E 55 Exp E 65 Exp E < 1 > No new information! Exp < 1 >-E 65 Exp < 1 >-E 75 Exp < 2 >-E 65 Exp < 2 >-E < 2 >

27 2014 J.A. Woollam Co., Inc F-3 Modeling? Challenge: More unknown sample properties than measured values. Solution: Measure additional information or reduce number of unknown properties. n(l), k(l) Known Substrate Y(l), D(l) d

28 2014 J.A. Woollam Co., Inc Opaque Layer Transparent Region Optical Constant Parameterization Multiple Angles Interference Enhancement SE + Intensity Multiple-Sample In-Situ F-3 Methods for Absorbing Films Multiple-Ambient J. Hilfiker, Survey of methods to characterize thin absorbing films with Spectroscopic Ellipsometry Thin Solid Films 516 (2008)

29 Preferred Method #1 Multi-Layer Strategies Measure n,k from single-layers: use dispersion models. Fit thickness only. If poor fit, add dispersion parms. for least stable film. Film A Film B Film C Substrate Substrate Film B Film C Film A Substrate Substrate 2014 J.A. Woollam Co., Inc. 29

30 Preferred Method #2 Multi-Layer Strategies Measure n,k for each new layer with previous layers fixed: use dispersion models. If poor fit, add thickness and then dispersion parameters for previous layers. Film C Film B Film B Film A Film A Film A Substrate Substrate Substrate 2014 J.A. Woollam Co., Inc. 30

31 6 Example 6-poly Si on 100nm TOx on Si Two Film layers Unknown thickness poly Si ~100nm Thermal Oxide Build your own method to fit the data J.A. Woollam Co., Inc. 31

32 2014 J.A. Woollam Co., Inc Thank You & Good Luck!