Heat Conductor

Thread Starter


Joined Sep 10, 2007
Hi Guys,
I've asked this on more than one forum.. but here goes...

I have to assume that ordinary silicone sealant doesn't have the same
heat conducting property as the silicone grease used on CPU heatsinks,
otherwise there would be no reason to spend the extra on a tiny tube of silicone grease.

I'm looking to fill a small area that would usually call for Bostik foam filler that comes
in a spray can an expands prior to setting, but that is a heat insulator, not conductor.

Any ideas of something that could be poured through a hole,
or added to the mix with the foam?
Or is silicone sealant a good conductor as well?
Cheers, Art.


Joined Apr 20, 2004
Actually, the transistor pad grease isn't that great a heat conductor in anything but very thin films. Most of the effect comes from filling voids between the transistor and the surface of the heat sink to maximize heat transfer. That's why you use a big expensive aluminum fin and a dab of silicon grease.

The fun fact is that if it conducts heat well, it also conducts electricity.

Tell us more about the volume you're filling, and why it has to conduct heat.

Thread Starter


Joined Sep 10, 2007
I have constructed some solar panels like this:

with my own added aluminium backing, and angle to complete an enclosure for
mounting a stalk or mast in the centre, and also to aid in anti tamper
since there is a serial data loopback wire as well.

Solar panels work better when cooler, so I'd like to transfer as much heat
as possible to the aluminium at the back.

The unit pictured is only about 10mm deep, but some larger panels with bigger
frames end up with a larger gap of 20 - 30mm.

Last time I just used filler, but it occured to me that a thermal conductor
could make use of the frame I have made.

Of course a dedicated solar forum would seem appropriate, but they aren't so much
help with irregular/unconventional issues.

Cheers, Art.


Joined Jul 17, 2007
Interesting project, Art.

A few random thoughts about it:
Better bonding of the solar cells to the backing would likely improve heat transfer; but after it gets transferred, there isn't an effective way to dissipate the heat. The aluminum stock used presents a fairly flat surface, except for some ribs. These small ribs would increase the surface area by only a very small amount.

I don't know what alloy your aluminum backing plate is, but aluminum has a pretty high coefficient of thermal expansion, and silicon wafers' COTE is pretty low. Therefore, if you bonded the two tightly together and then heated them up, you would likely wind up with an aluminum plate with silicon shards stuck to it.

Aluminum really isn't all that wonderful for conducting heat. Copper is twice as good. Water is even better. Diamond is the best that I know of, but that would get pricey in a hurry.

Perhaps a better thing to do would be to look at alternative cooling methods - how are you going to remove the heat from the panel?

Heat pipes are one way to go. They're used in many items nowadays; even laptop computers and satellites orbiting the earth. They're quite simple in theory; a copper tube is evacuated of atmosphere, and a small amount of liquid, such as alcohol or water, is introduced. Since the atmosphere is evacuated, the liquid will boil at a much lower temperature than it would at normal sea level atmospheric pressure, and would condense at the coolest point in the tube. In a spacecraft, that would be an end of the tube that was external to the craft and had a large radiating surface, normally with a thin black coating. A black heat radiator transfers the heat more efficiently due to a better matching of the infrared spectrum.

Since your solar cell array will undoubtedly be located outdoors, you'd want to use a liquid with a low freezing point; that leaves water out, although you could possibly design a triangular-shaped highly polished tube with a vee of the tube downwards; if the water froze during a cold snap, the expansion would simply rise up the sides of the vee rather than destroying the shape of the tube.

A triangularly-shaped tube would also simplify the bonding of the solar cells to the back of the tube.

Carrying away the heat; seems like another household item that uses power is a water heater. Might as well use the waste heat from the solar array to warm your bath water, eh? A very low-volume pump would be enough to keep the water moving; during cold spells you could evacuate (drain) the pipes, or have a microcontroller do it for you.

Well, enough ramblings for the moment - something to chew on anyway. ;)

Thread Starter


Joined Sep 10, 2007
how are you going to remove the heat from the panel?
No active method... I hope it is sunny and windy ;)

Thanks, some things to think bout there.
I have thought about making a water heater this way,
since many people using solar panels are absorbing enough heat
to make use of with copper pipes, etc.

But this is a mere camping setup, a few panels at best.

In common practice you would never cover the rear of any PV panel.

This is a rather unique approach to anti theft, and mounting a camping
panel by driving a stake into the ground, and placing an assembly over it.
If there is any rotator, it is to be part of the panel assembly for this reason.

In fact, here is a picture of that idea.. not the same as most solar trackers.
The panel is always fixed at a 45 degree angle.



Joined Jun 30, 2006
A two part silicon rubber can have a metal powder or ceramic powder added to it to enhance its' thermal conductivity.

If you'd not worried about electrical conductivity silver coated copper powder is what I'm familiar with. Another is aluminum oxide. Exotics are boron nitride powder and there's newer ones coming. Some RF applications work well with thermally and electrically conductive silicon rubber. I've seen flexible circuitry done with them, it wasn't long lasting but was cool for a while.


Another compound is metal filled expoxy. 3M makes a Scotchweld variety that's excellent and incredibly tough.

Where I used to work we bonded chips to metal frames to dissipate heat. The frame was a heat exchanger box that had air forced into it under high pressure. Old late '70 military radars, I bet you could have replaced the 'cpu' with a DSP/CPU chip from TI and replaced the power supply with an off the shelf unit and saved about 150 pounds weight. We still repaired them to the original documents *exactly*, no deviation.

For thermal grease for CPUs I just saw an article where silicon grease and industrial diamond powder worked better than the best available currently.