Thermal Paste and the Future?

[confused]

Why is it that we are still using silver pastes? Why doesn’t industry recognize the clear advantage of synthetic diamond or even natural diamond? Synthetic alone has a 7 to 1 thermo conductive advantage over silver, let alone naturally occurring diamond that in some cases has a 10 to 1 ratio over silver. O and the funny thing about diamonds are some have an extremely high resistance also. Wow look at that superior heat transfer and high resistance, could we ask for more? I know that the silver paste has a high resistance also but think if they didn't have to do that because the material itself already isn’t electrically conductive. Here are the numbers straight from my Materials book.

Thermal conductivity (Room temp)
Silver (commercially pure) 428 W/m-K

Diamond
Natural 1450 to 4650 W/m-K
Synthetic 3150 W/m-K

Resistance (Room Temp)
Silver 1.47 x 10^-8 ohms-m

Diamond
Natural 10 to 10^14 ohms-m
Synthetic 1.5 x 10^-2 ohms-m

Now I think industry has the same tables I do so this should be uniform. This concept is clear to even myself, now if I’m missing something or you know more about this please let me know that I’m a moron and for what reasons.

[Tiger]

There are several obvious problems with diamonds;
1. They are one of the hardest materials known;
2. Given that, the cost to produce dust is going to be astronomic
3. Once you have that dust in suspension it is going to have superb abrasion properties. So when applying it to your heat sink you are going to scratch the surface incredibly which is going to be totally counter-productive

[confused]

Tiger here is my rebuttal.

First of all yes the diamond is hard, but the jewelers use something to shape them and cut them so it is possible to cut it. Second, we used diamond embedded cutting wheels in the machine shop I used to work at to shape carbide. Also some of the newer cutting tools for machining metals have a diamond coating, so cost is not the reason. Lastly surface area is to cooling what hot chicks are to porn, the more you want the more you need. That said all those small highly microscopic pores and scratches increases the overall surface area for the heat sink. Besides, it’s not like your going to be grinding your heat sink into your processor die.

confused.

[VegetaQ]

Take a look here

http://forums.overclockers.ws/vb/showthread.php?s=&threadid=51052

Nevin actually gets in theyre and explains the problems (at the very least with manfacturing) Heat Transfer Paste with Diamond content.

[VegetaQ]

In case anyone is too lazy to click on the link. Its ok though I too find myself so out of energy after a full day of crusing on the forums I cannot click links anymore.

I hope Nevin doesnt mind me quoting.

"While the thermal conductivity of diamond is very high, there are other physical characteristics that make it a poor choice for a thermal compound. Microscopic diamond particles have spikes and sharp edges so the physical contact from one particle to the next is extremely small. The heat travels very well through a particle, but does not transfer well to the next particle. When you are talking about 50 to 100 particles to make up 1/1000 of an inch, this lack of particle-to-particle conduction is a major problem.

Also, diamond has a great affinity for diamond. Once the compound sits for a few hours, it forms clumps as the diamond particles bunch together. A very bad thing in a thermal compound.

And finally, the wands on high-speed homogenizers are about $1800 each. The bearings on the wand are submersed in the compound being mixed. The highly abrasive diamond particles ruin the bearings each time a batch is mixed. That's a loss of $1800 per 400cc of compound.

Given all this, I have mixed up several test batches of diamond compound.

They all suck.

Nevin"

[confused]

Well I learned something new. Never thought about the fact the crystals would need to be in such close contact. Well thank you

thanks
confused

[Tiger]

Diamonds are used to cut diamonds. They also use a technique called cleaving (I am ex South African - the home of diamonds - and my family were involved in the industry many years back) i.e. they look for the faults in the stone, which is a crystal structure, and they break the stone along that fault line.
To make the equivalent of AS2 you have to make what's called a colloidal suspension. In normal colloidal suspensions the colloid is achieved because the chemicals being used have electostatic charges on them causing them to repel each other and therefore sustaining the suspension.This is turn means that the solid state of the material can be fairly large. However in the case of diamond this is not possible since it is chemically inert (diamond is just highly compressed carbon). This means that the dust has to be in the region of 1 to 2 microns in diameter to sustain a stable suspension. Unless a colloid can be manufactured it is useless because the dust and vehicle ( the liquid component) will continually separate. In most cases extended storage can result in total separation. Once separated they can never be blended back except by a colloidal mill. To produce dust of this dimension is going to be extremely expensive!