On Solder and Superconductivity

[uWave_Matt]

Hi all - I've got a bit of hesitation regarding the next step in my research...

The setup: I've fabricated a uStrip HTSC resonator that I'm trying to characterize on a VNA. The sample is placed in a jig with coaxial feeds, and the jig is cooled via a cryo-cooler.

The problem I'm having is vibration caused by the cryo-cooler, which gives a very ugly trace and has vibrated my sample out from underneath the probes more than once. Given that the cooling process is long and I can't see the sample during the cooling, I don't know if the probes are still in contact with the resonator feed lines until the end of the cooling process when the sample chamber drops below 75-80K.

I'm thinking of soldering the sample into the jig, though I don't know how soldering temperatures and solder itself might affect the superconducting feed lines, nor do I know how solder behaves at low temperatures...

Another thought I had was putting double sided tape on the ground plane of the resonator so that it can't vibrate out from underneath the probes, but once again, I'm not sure what happens to the tapes adhesion at such cold temperature.

Do any of you fine engineers/researchers have any experience with any of the aforementioned issues?

 

[MATLABdude]

Unfortunately, the coolest (and highest) temperature compatible tape I'm familiar with is Kapton (polyimide) tape with silicone-based adhesive. Unfortunately, that's only good for temperatures down to -73 C. Manufactured by 3M and Micronova Manufacturing, if these ever come in handy.

I got a few hits with cryogenic epoxy on Google, but I don't know if you can ensure electrical contact in that case. Unfortunately, unequal thermal expansion coefficients can really do a number on bonds (electrical or not). I know there are a few condensed matter folks in this subforum, so hopefully, they weigh in!

 

[f95toli]

I more or less do this for a living (literally, I do MW experiments at cryogenic temperatures) so I am familiar with the problem.

Yes, solder works fine at cryogenic temperatures (and becomes superconducting at about 5K if you use 60/40). Kapton tape is also fine. Just about any type of glue is also OK; ("cryogenic" epoxies suichs as certain types of stycast are used because of their high thermal conductivity at low temperature, "cryogenic" has nothing to do with how well they work as a glue)

But there is now way you will be able to solder your HTSC sample into the jig unless there is some gold pads on the sample already; you can't solder to HTSC directly.

Your best bet it to use a spring loaded connection; made from something like coppery-beryllium (which is "springy" even at cryogenic temperatures). Alternatively you could use some indium between the sample and the connectors; indium is so soft that you can squash. A third option is to use silver glue instead of solder (silver glue=the stuff you can buy that is really meant to repair PCB traces, i.e. you can get if from most suppliers of electronic components).

There is a whole chapter about this is in Lancaster's book about superconducting microwave devices (including a drawing of a jig for microstrip resonators).

 

[uWave_Matt]

I more or less do this for a living (literally, I do MW experiments at cryogenic temperatures) so I am familiar with the problem.

Yes, solder works fine at cryogenic temperatures (and becomes superconducting at about 5K if you use 60/40). Kapton tape is also fine. Just about any type of glue is also OK; ("cryogenic" epoxies suichs as certain types of stycast are used because of their high thermal conductivity at low temperature, "cryogenic" has nothing to do with how well they work as a glue)

But there is now way you will be able to solder your HTSC sample into the jig unless there is some gold pads on the sample already; you can't solder to HTSC directly.

Thanks for saving me the trouble of trying...I was actually worried about whether or not it would bond in addition to what the heat would do to the HTSC - so thanks for saving me from that cluster.

Your best bet it to use a spring loaded connection; made from something like coppery-beryllium (which is "springy" even at cryogenic temperatures). Alternatively you could use some indium between the sample and the connectors; indium is so soft that you can squash. A third option is to use silver glue instead of solder (silver glue=the stuff you can buy that is really meant to repair PCB traces, i.e. you can get if from most suppliers of electronic components).

My jig is actually spring loaded, but the two probes barely make contact with the feed lines, and I think - as you said - maybe the springs become less springy at cryogenic temperatures. I'll look into the silver glue - I might actually have something similar to it in the lab.

I suppose if I end up using the glue, my question becomes, is there way to free the sample from the jig in such a way as to leave the sample undamaged?



Thanks to both of you for the advice.

 

[MATLABdude]

If you soak your sample in acetone, the acrylic binder of most of the trace pens (silver conductive pens) I'm familiar with should come off:
http://www.mgchemicals.com/products/8420p.html
http://www.chemtronics.com/products/product.asp?id=7

For epoxies, you might need something more aggressive:
http://www.mgchemicals.com/products/8310.html
http://www.chemtronics.com/products/product.asp?id=25

Not sure about low-temperature compatibilities, however.

 

[uWave_Matt]

Your best bet it to use a spring loaded connection; made from something like coppery-beryllium (which is "springy" even at cryogenic temperatures). Alternatively you could use some indium between the sample and the connectors; indium is so soft that you can squash. A third option is to use silver glue instead of solder (silver glue=the stuff you can buy that is really meant to repair PCB traces, i.e. you can get if from most suppliers of electronic components).

I've found some silver paint and silver/gold extender from Ted Pella (part number 16032 and 16021) - are you familiar with these and do you think they'd work? I'm not exactly sure how to apply the extender...the documentation they have on their site says that the paint hardens in seconds. That hardly seems like enough time to mix the paint with the solvent/extender and get it applied to a feed line that is less than 1mm wide...

If you recommend against them, I'm headed to Debco or Radio Shack I suppose.