Composition for BiSb
Although Bi needs to be grown at a growth rate of < 0.06 A/s and a substrate temperature < 25 to prevent the formation of Bi surface droplets, BiSb can be grown at a much higher growth rate. I think this is because Sb reduces the adatom mobility, preventing the formation of the (104) and etc orientations that correspond to the surface droplets. I still grow the BiSb between 20 - 25 C to be safe.
Because the Sb sublimator temp is optimized for III-V growth, my normal Bi BEP is ~ 3E-8, and we're mostly targeting < 22% Sb, I use a higher growth rate for the BiSb films in order to be able to use a large enough BEP for the Sb so that I can reliably repeat it.
For BiSb growths B170928A - present, my approach has been to set the total BEP to be 2E-7, and to adjust the Bi / Sb BEP ratios appropriately, so that for a ~5% Sb composition the Sb BEP is 1E-8, which I feel is safely repeatable in the STO measurement.
In April 2018, Sarah's XPS data showed that the actual measured composition (including both the oxide and the metallic film), is ~ 2x the intended BEP. Her calibration is listed below.
Sample ID | Thickness | Orientation' | targeted % Sb from BEP ratio | Measured % Sb center of film XPS |
B171019C | 10 nm | (001) | 3 | 3.8 |
B170928B | 10 nm | (012) | 5 | 8.8 |
B170928A | 10 nm | (012) | 10 | 23.3 |
B171019B | 10 nm | (012) | 15 | 26.9 |
B171016C | 50 nm | (012) | 10 | 28.5 |
B171102A | 50 nm | (012) | 15 | 28.1 |
By averaging the measured composition for samples with the same targeted composition, we get the formula:
Measured % Sb = 2.03 * (BEP ratio % Sb)
So, to determine the BEP ratio for your growth:
Decide what actual film composition you want (i.e. 10% Sb)
Determine the BEP % Sb from the equation above (i.e. BEP % Sb = 10/2.03 = 4.92)
Determine the Sb BEP using the condition that the total BEP is 2E-7 (or whatever number you decide) (i.e. .0492(2E-7) = 9.8E-9 )
Determine the Bi BEP using the same method (i.e. 2E-7 - 9.8E-9 = 1.9E-7)
Notes:
I believe that the Sb composition is about 2x what is expected because in III-V growth half of the Sb molecule is desorbed due to the high substrate temperature, but at BiSb growth temperatures both are incorporated. Further pure Sb growth and/or growth at different substrate temperatures can help to clarify this
The growth rate also varies slightly with the % Sb