I was looking for IC packages where I can mount a small wafer (3mmx3mm) wire bond to the leads of the package and conduct operation at 500C. I have contacted couple of companies which do ceramic IC packages but they only guarantee operation till 250C. Would these ceramic packages hold till 500C? Also, is there any conductive thermal paste to mount the wafer which can sustain 500C? Any help would be very useful!

  • 4
    \$\begingroup\$ What kind of semiconductor are you using that can survive such temperatures? \$\endgroup\$ Sep 19, 2019 at 22:12
  • 1
    \$\begingroup\$ Do you actually have a wafer that can function at 500C? \$\endgroup\$
    – DKNguyen
    Sep 19, 2019 at 22:14
  • 4
    \$\begingroup\$ I am researching in diamond \$\endgroup\$ Sep 19, 2019 at 23:05
  • 1
    \$\begingroup\$ @HarshadSurdi I don't think diamond is on the market yet. I've seen some research about it but I don't think you'll find anyone with diamond ICs. There are diamond compounds for thermal pasting for heat syncing. Gallium nitride might also be something you should check out. \$\endgroup\$
    – user103380
    Sep 19, 2019 at 23:24
  • 2
    \$\begingroup\$ You have some thoughtful comments and answers. I worked on something we designed to operate using standard commercially available parts, running them at 180 C (above their specs.) The useful running time we promised was on the order of hours, not weeks or months. So one big question I have for you is, "What duration of operation do you need?" The lack of any discussion of this seems to be a glaring hole in your question to me. \$\endgroup\$
    – jonk
    Sep 20, 2019 at 6:37

3 Answers 3


Semiconductors become conductors above a temperature related to the band-gap. A few materials, such as pure carbon (e.g., diamond) and silicon carbide (SiC), have been used to make experimental or custom designed devices such as low power IC's and high power IC's that operate at 500°C.

As far as thermal transmission and insulation, diamond is ne plusa ultra. Since gallium wets diamond, you could make a Ga/C/Ga sandwich to conduct heat and insulate electrically. The diamond probably would be less expensive than the special-purpose IC!

Though you might be able to order (and afford!) such a device, might it be possible to move signal processing to a cooler location? Admittedly, in borehole logging and spacecraft to land on Venus, this might not be a choice, but look for a more practical alternative.


An IC usually consists of these things:

  • Epoxy to encapsulate the IC
  • Pins
  • Wire bonding (usually gold)
  • Chip or Die
  • Solder

I would think that the concern would be with each of the materials of the package. I know that most solders start to lose their structural stability at 250C. I know that IC epoxies don't survive much after 250C either from personal experience. The wire bonds would also need to support high temperatures. All of these components need to support operation higher than 250C.

I would think that the solder would be the most limiting factor, and solders with higher melting points can be found up to 450C, after that its a different process called hard soldering or brazing.

  • 2
    \$\begingroup\$ You forgot to add the important issue of heat expansion. If the coefficient of expansion is not matched it doesn't really matter if the materials survive the temperature change (e.g., don't melt) these can still crack, deform, or detach due to a mismatch on their expansion rates. \$\endgroup\$ Sep 19, 2019 at 22:46
  • \$\begingroup\$ Any idea what company sells these high melting point solders? \$\endgroup\$ Sep 19, 2019 at 23:08

Why do you want to heat the IC to 500°C just for bonding a wire to the leg or whatever of the package? Proper bonding will not heat the package a lot. Especially you said package. You are not even trying to heat the waver.

Best solution would be basically the principle of spotwelding. This is used in batteries, since it is so quick, that barely anything can heat up. This is possible with electronics aswell. But you need the ability to archive the precision required depending on the size of your package. For example bonding an 0402 component would be very difficult. You would need to either build or buy a machine to do so. You can steal with your eyes by watching some Waver bond videos on you tube. This at a slightly larger scale - therefore "simpler" would be the way to go for you. I think someone pulled that of in his own private lab. I guess it was this guy making DIY mosfets at home (actually DIY from the grounds of pure silicon).


Bonus: This might work better with totally cleaned surfaces and nitrogen atmosphere, to prevent oxidation before proper connection. Depending on which way you wanna do it.

  • 1
    \$\begingroup\$ I think you missed the point of the question - the OP has bare-die circuits that they want to electrically test up to 500C. \$\endgroup\$
    – W5VO
    Apr 22, 2020 at 18:26
  • 1
    \$\begingroup\$ You missed the point. The IC needs to operate at 500 degrees. It isn't about the bonding or soldering temperature alone. It's about making an IC for use at 500 degrees. \$\endgroup\$
    – JRE
    Apr 22, 2020 at 18:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.