Liquid Thermal - Aero 15x Question (Thermal Paste Barrier)

So I've been doing a lot of research in Liquid Metal thermal compounds. A lot of people seem to have initial success with eventual copper corrosion due to a reaction between the gallium and copper. I have read that some ways of at least slowing down this process is to make sure that there is no oxygen reacting with the gallium and the copper.

I have seen some people using 33+ electrical tape on order to prevent spillage and create an air tight seal. I was wondering... Instead of using electrical tape to create a seal why not used non-conductive thermal paste around the edges of the die?

I have no idea if this is a good idea or not. I was hoping that somebody might have some better ideas on this. My knowledge on Chemistry is limited unfortunately. Perhaps there may be some issues with the liquid metal reacting with conventional thermal paste?

tldr; Should I put non-conductive thermal paste around the cpu/gpu die to create a barrier for the liquid metal thermal compound?

 

If you are going for air tight, then no. Since TIM will remain some what liquid after application it will not create a solid seal. for air tight a thin bead of silicon would be the best bet. This could have some issues as if too much liquid TIM is used the pump out action in tightening down the HS could break the silicon seal/bead.

 

you can't really create an airtight barrier. The gap is less than 1mm thick. you run the risk of causing a reduction in heatsink pressure. However, using a VERY thin and compressed layer of RTV silicone (the same stuff used to reseal a CPU IHS, after delid, applying LM on CPU and re-lidding) could probably work. I'm assuming you're trying to combine an airtight seal and a foam dam type barrier as well. In that case you probably should read Mr Fox's post about how he compressed RTV silicone (by applying super 33+ tape on top of it, a very very thin layer of RTV, then letting it harden for a week of computer use, then removing the tape over the now fully compressed <0.2mm barrier) to act as a barrier. What you're doing here is just creating a 'dam' that stops LM runoff from getting on the PCB. Not an airtight seal.(even if the compressed foam does make it sort of airtight, there will still be "air" between the edges of the barrier and the CPU anyway!)

Anyway the best way to insure a good fit is to make sure the heatsink is fully flat and sanded and not convex or concave. Convex and concave heatsinks will cause coretemp differentials and more absorption/oxidation regardless of the seal used.

I recommend just using basic very light dense foam as barriers as i mentioned here, and trimming them down to <2mm to prevent any resistance to heatsink pressure while still maintaining compressibility.

http://forum.notebookreview.com/thr...r-liquid-metal-safety-insurance-guide.817207/

 

Thanks for replying. I'm not going to go as far as air tight seal the LM. I did some more research on the substance and it seems as long as I keep everything on the die I should be fine. I was worried about long term use with the LM drying out, but it seems like best solution is to clean the previous application and reapply.

I will definitely check out Mr Fox's post. I'm also not sure how substantial of a change air tight ceiling the LM is.