We recently grew some 100um wide, 150nm high, mm-scale long metal contacts on P-doped Si wafer but found that they are conducting with each other through the Si wafer at resistances around several hundred ohms. We didn't find this phenomenon when we directly apply multimeter tip onto the Si wafer surface to test. I guess this is because the 'large' area of metal contacts eased conduction through the wafer area between the contacts. Does anybody have good idea that what kind of wafer we should use in order to largely avoid this problem? We don't care whether it is Si or pyrex or anything else, only if it is flat enough and insulating. And any good provider of such wafers?