How do you find specific resistance of a contact on semiconductor?

Answer 1

The problem that you face is that there is a resistance derived from the contact and a resistance derived from he semiconducting material itself. The trick is to separate the two resistance values.

To do this you will need to set up a series of contacts. Each of the pairs of contacts will need to be separated by a different distance through the semiconducting material. I would suggest at least three pairs but the more you use the more accurate your contact measurement will be.

Use a Kelvin, or 4-point measurement technique. This is done by using, as the name suggests, four probes. with two probes apply a small but known current through the contact pairs and semiconductor. Using the other probes measure the voltage difference from one contact to the other. The voltage drop, and current flow will give you the resistance across the contact pair (Resistance = Volts / Current).

This gives you the resistance of two contacts and a resistance of a small amount of semiconductor. Now do this to all three different contact pairs with different spacings of semiconductor. Of course, the longer the semiconductor spacing the more the total resistance will be. If you plot these values as a line (using the spacings as an X-axis, resistance on the Y-axis), you will be able to extrapolate what the resistance would be if there were the theortical value of zero spacing, or no contribution from the semiconductor (the Y-intercept).

Now this is the value of two contacts so you will have to divide this value by 2. And it is also the value of the specific area of the contact you used. The larger the area the less the contact resistance. So Divide the contact resisance by the area. The typical units to report this number is in Ohms/cm^2 (ohms per centimeter squared).

By the way, this in theory could be done using less points, however you need to be aware that the linearity of your measurements is important. In practice the contacts (even good ones called 'Ohmic Contacts') are not perfect. Measuring at least three (I use five) pairs of contacts will give you a way to see if you are getting a good reading of the contact resistance. If you get a correlation coefficient, (R-squared available on an Excel spreadsheet), for your data to a best-fit line of 0.9 or better you are getting a 'reasonable' contact resistance measurement.