I'm using 7604 system for cc in III-V. Somebody recomended using mask with circular pattern (like 1 in the image) to deposit gold contacts. This is very easy becouse you dont get deposition on the side of the sample, like in 2 and 3. The contacts are peffect, the measurement is perfect, but the results are very diferent from SIMS.
Looking at the image can you please let me know which pattern is the most suitable for Van de Pauw method and which are not suitable at all? Is the shape important or the position of the contact?
Last Edit: Feb 28, 2019 2:24:09 GMT -5 by mykhaylo
Post by Lake Shore Jeff L on Feb 28, 2019 11:14:49 GMT -5
Hi mykhaylo This is a very good question. The van der Pauw method assumes point contacts on the edge of the sample. Of course, real samples do not have point contacts and the contacts are not on the edge of the sample. For finite size contacts it is possible to calculate correction factor to correct the measured resistance. These correction factors are difficult to calculate. There are many references in the literature for various geometry of sample and contact size. If you have the manual for your 7600 system, in appendix A there is a summary of some geometries and contact sizes. In addition, here are two references that I find useful CONTACT SIZE EFFECTS ON THE VAN DER PAUW METHOD FOR RESISTIVITY AND HALL COEFFICIENT MEASUREMENT RONALD CHWANG, B. J. SMITH: and C. R. CROWELL Solid-Slate Electronics. 1974. Vol. 17, pp. 1217-1227.
FOUR-POINT SHEET RESISTANCE CORRECTION FACTORS FOR THIN RECTANGULAR SAMPLES DAVID S. PERLOFF Solid-Slate Electronics. 1977. Vol. 20. pp. 681487.
I'll look into those papers you've sugested. Are you avere of a geometry similar to 1 with known corrections?
I'm a novice to Hall effect system as you can see. Can we apply correction coeficients in the software before the measurements?
Hi I am not aware of known corrections for geometry similar to one. The Lake Shore software for Hall Measurements Systems does not support the correnction coefficients for any geometer. The software always assumes, for van der Pauw samples, point contacts on the edge of the sample.