Achieving Uniform Contact with Silicon Wafer

[Coldhot]

Hi all,

I am currently working to build a Joule Heater that will be used to heat samples to upwards of 1000*C in seconds in vacuum. While the temperature and speed is definitely there, the contacts are always non-uniform -- making the data hardly repeatable. I am using a ~70mmx10mm Silicon wafer that is clamped on either side by 2 contacts. My problem is that on either side there is a "hotspot" where the contact is the best, all of the current flows through these two hot spots. There is 2 problems with this -1- it limits the temperature I can reach because the temperature at the "hotspots" is much much higher than the temperature of the center of the wafer. -2- it extremely prohibits repeatably, which is key.

I have tried stainless steel contacts as well as tantalum contacts, both rough and polished. I have tried several different clamping techniques as well. No matter what I try, there is always a point of "best contact" -- which makes me wonder, is uniform contact even possible (at least with the funds I have available)?

 

[M Quack]

Make sure that there is no SiO2 in the contact area. Etching with HF might help.

Can you evaporate a thin film of gold onto the contact area?

After that I would try clamping polished copper contacts.

 

[Coldhot]

I have tried etching to no avail. As for a thin gold film and copper contacts, I do not believe they can handle the temperatures we're hoping to achieve. I would like to try polished copper at lower temperatures though, to see if there is an improvement.

As for the process of evaporating gold, we don't have the means to do that atm but are actually in the process of building a system that can.

 

[M Quack]

I am not sure about copper handling these temperatures either. The melting point is just a hair above your working temperature. Gold is even closer. With a bit of luck the copper will stay quite a bit cooler than the Si which has much higher resitance. Au will diffuse into Cu, so you might need a blocking layer of Ni on the Cu contacts.

Pt is much better, 1700K. After that you are into more exotic metals like Rh, Ru and Ir.

Graphite electrodes might also be worth looking at. But you might have noticed that I am starting to fish in the dark.

 

[Carl Pugh]

Have you considered heating using an arc lamp?
Peak Systems Inc. (No longer in business) had a patent for Rapid Thermal Heating using an arc lamp. You might be able to get a copy of this patent.