Long-term stability of Arctic Silver/Ceramique

Another inadequacy of the thermal grease testing/comparisons found all over the web is the absence of long-term reliability testing or at least an accelerated reliability testing. This is often overlooked but ultimately of utmost importance, since thermal greases have well-documented failure mechanisms ("pump-out") leading to a progressive deterioration of the thermal interface, and consequently higher die temperatures.

Arctic Silver and Ceramique claim to have a "controlled triple-phase viscosity", and that they "thicken slightly over the next 100 to 300 hours". This seems to imply that these products CURE, even if only partially. I have to bring to question the validity of these claims. Arctic Silver has no datasheets that characterizes the modulus of the cured and uncured product.

The curing mechanism is also undisclosed (i.e. is it heat-curable?, aerobic?, anaerobic?, moisture-cure?). A curable product also implies a very short shelf life. What is the shelf life of AS?

What "activates" the curing mechanism?

Being single-component (i.e. not 2-component ala epoxy) also implies certain storage conditions, typically refrigeration/freezing. How do you guarantee that AS won't cure inside the syringe while sitting in that hot store shelf space? (cured product won't spread well)

Or is the claimed viscosity change just a marketing gimmick???

This is important since the viscosity change is key to long-term stability.

So ultimately the question is: has Arctic Silver conducted long-term stability validation of their products? (is AS an ISO-certified company???)

I had AS II on my CPU for 18 months and used rubbing alcohol to remove it and it was just as pliable ( soft) as the day I put it on. I saw no REAL difference in tempsfrom month 1 to 18. Only about a 2°C change in that time.
I usually reapply the TIM every year just to be on the safe side.
The AS Tim cures via heat.
I had a tube for about 2 years and it is still alright.
Hope that helps

It doesn't really "cure" as most people think the word means. It is the arrangement of the particles in the product to get the best heat transfer properties. All that happens is minute particles move just a tiny bit to provide a better medium for heat transfer.

I just realized I just repeated myself.

Anyway, AS never really harden cures.

[url]www.arcticsilver.com[/url] said:

During the system's initial use, the heat from the CPU lowers the viscosity of the compound to enhance the filling of the microscopic valleys and ensure a minimum bond line between the heatsink and the CPU core. Then the compound thickens slightly over the next 100 to 300 hours of use to its final consistency designed for long-term stability.

...

Due to the unique shapes and sizes of the particles in Ceramique, it will take a minimum of 12 hours to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. (This period will be longer in a system without a fan on the heatsink.) On systems measuring actual internal core temperatures via the CPU's internal diode, the measured temperature will often drop slightly over this "break-in" period.

Click to expand...

It seems like the viscosity of the suspension fluid changes with temperature, but this is only a real factor after initial application. The viscosity reduces, allowing the particles to more easily settle in. The second paragraph I quoted shows that this "curing" time is affected by fans, e.g. vibration. That means that it's related to the settling of the particles. Apparently, the thickening of the compound is related to the squeezing out of excess suspension fluid, leaving the interface junction with an optimal amount of fluid and particles.

Yes of course the suspension fluid decreases in viscosity as temps go up, and is beneficial in achieving a thin bond line.

But you seemed to have confused "curing" with "settling"....

And no, the claimed thickening is not due the separation of the suspension fluid. When grease is squeezed out the particles get squeezed out too ;-)

Particles can't get squeezed out if they're settled into the grooves of the heatsink...

And who ever said this really was "curing"? All I've seen is that Arctic Silver calls it "thickening", which in the marketing world, could mean curing, settling, evaporation of the fluid, expansion of the particles...

And who ever said this really was "curing"? All I've seen is that Arctic Silver calls it "thickening", which in the marketing world, could mean curing, settling, evaporation of the fluid, expansion of the particles...

Click to expand...

That's exactly what this thread is all about. Does AS really "THICKEN" (increase viscosity) or is it just marketing hoopla?

This is important since the "thickening" is integral to the long-term stability of a TIM. A "thick" TIM is less likely to pump-out.

Have you ever noticed that +/-3 months after AS application your temps begin to rise??? i.e. Thermal performance degradation...

I would appreciate if Colin or Nevin can comment on this...

RonnieG said:

That's exactly what this thread is all about. Does AS really "THICKEN" (increase viscosity) or is it just marketing hoopla?

This is important since the "thickening" is integral to the long-term stability of a TIM. A "thick" TIM is less likely to pump-out.

Have you ever noticed that +/-3 months after AS application your temps begin to rise??? i.e. Thermal performance degradation...

I would appreciate if Colin or Nevin can comment on this...

Click to expand...

Why don't you send them an e-mail and ask them, they are friendly people with an open door for any and all questions. Posting a specific question on a message board sometimes doesn't necessarily get you a rapid answer.

Heres their E-mail addy [email protected]

Maybe after you speak to them let us know what they had to say.

(maybe they will read this thread but its usually faster to e-mail them. Hey It's faster to get cheese at the supermarket than buying a cow and trying to make your own, meaning its usually faster to go right to the source rather than tip toe through the tulips......so to speak)

I can answer one question, I've got a tube of Arctic silver II thats at least a year or more old and its just as workable as when I bought it.

Why don't you send them an e-mail and ask them, they are friendly people with an open door for any and all questions. Posting a specific question on a message board sometimes doesn't necessarily get you a rapid answer.

Heres their E-mail addy [email protected]

Maybe after you speak to them let us know what they had to say.

(maybe they will read this thread but its usually faster to e-mail them. Hey It's faster to get cheese at the supermarket than buying a cow and trying to make your own, meaning its usually faster to go right to the source rather than tip toe through the tulips......so to speak)

I can answer one question, I've got a tube of Arctic silver II thats at least a year or more old and its just as workable as when I bought it.

Click to expand...

Yes of course I emailed them first before I posted here (what were you thinking?). Nevin did claim the shelf life to be "several decades" in a refrigerator. But he didn't answer my questions regarding the "thickening" mechanisms... nor the pump-out issues...

My subsequent queries went unanswered. Cheese is out at the supermarket... so to speak...

It could be they're just plain busy, or could it be that they're evading this reliability issue???

RonnieG said:

Yes of course I emailed them first before I posted here (what were you thinking?).

Click to expand...

I think Silversinksam was thinking about being courteous and informative.

I have seen a 1-2c improvement in temps for the first 12-48 hours after the application of the arctic silver 3 or ceramique. After that it stays the way it is, I can't speak for 6 months down the line or more though. Sufficing to say as far as thermal interface compounds, arctic silver 3 or ceramique is as good as you are going to get.

RonnieG said:

Yes of course I emailed them first before I posted here (what were you thinking?). Nevin did claim the shelf life to be "several decades" in a refrigerator. But he didn't answer my questions regarding the "thickening" mechanisms... nor the pump-out issues...

My subsequent queries went unanswered. Cheese is out at the supermarket... so to speak...

It could be they're just plain busy, or could it be that they're evading this reliability issue???

Click to expand...

I doulbt they are avoiding answering your question, and Im sure they are very busy, prior to a product launch of AS5 and all

For what it's worth ... I've found Arctic Silver to work very well "straight out of the tube", but it seems touchy if you disturb your HS after installation and it also seems to take time to "resettle" if you do disturb your HS after it's been in a while ... long term, I don't know if it dries out or not, and I don't know if it will again work as well as when new. On the other hand, I've had really good luck with Shin-Etsu grease, G-751 or G-765 ... pretty stable and pretty forgiving long term. When first installed, it may be just a bit under Arctic Silver in performance, but it seems to be more reliable once installed. My 2 cents anyway. Good luck!

KK

Koooler King said:

For what it's worth ... I've found Arctic Silver to work very well "straight out of the tube", but it seems touchy if you disturb your HS after installation and it also seems to take time to "resettle" if you do disturb your HS after it's been in a while ... long term, I don't know if it dries out or not, and I don't know if it will again work as well as when new. On the other hand, I've had really good luck with Shin-Etsu grease, G-751 or G-765 ... pretty stable and pretty forgiving long term. When first installed, it may be just a bit under Arctic Silver in performance, but it seems to be more reliable once installed. My 2 cents anyway. Good luck!

KK

Click to expand...

Actually this isn't right, its common knowledge that by jiggling (wiggling) the heatsink once installed on the cpu, it actually cuts down the break in period significantly. Without jiggling the heatsink it could take weeks or months for the compount to settle.

Silversinksam said:

Actually this isn't right, its common knowledge that by jiggling (wiggling) the heatsink once installed on the cpu, it actually cuts down the break in period significantly. Without jiggling the heatsink it could take weeks or months for the compount to settle.

Click to expand...

Well, I agree and disagree ... it is true the "jiggle" method is the way to go to initially seat the HS and new grease. I've done that with Arctic Silver, and it's the best way to reduce the settle-in time. However, I have found if you later disturb it, it will lose some contact and take a while to lower temps again, repeat "jiggling" seems less effective with Artic Silver once it's been on there a while. Maybe it dries out and flows less well? I was just saying that the Shin Etsu grease seems more forgiving in that sense ... maybe better particles or liquids? Maybe just my imagination? Just commenting on my experience based on temp changes I've seen with both.

KK

What's this "jiggling" stuff? I'm not exactly a newb to this but this one is news to me.

How exactly do you "jiggle" your heatsink? I run PIII's and P4's and generally use SLK800's and SLK900's. What do I do, back waay off on the screws and twist it around a few times, then retighten them?

"Jiggling"... LOL

Could anyone please explain how "jiggling" is done?

AFAIK you shouldn't disturb them once the surfaces have been brought together, otherwise you could introduce air into the junction...

once you set the hs, you twist it left and right (i do it once in each direction) then attach the hs clip..

Jiggling- the newest technical term on the block

But I find this to be a VERY effective method as well, primarily on first installation but also after moving a system.

My method:
First install- install compound and sink (or water block) normally and then slightly twist left and right a few times. If the mounting method is too tight at full tension then EVENLY loosen it just to the point where a slight twist is possible, Jiggle it (lol), and retighten. Complete system build and start normally.

After moving a system:
Start the system and run it under load for long enough to get things up to temp- CAREFULLY jiggle the sink or block slightly.

After messing around a lot, those methods work best for me.

Back to the original topic-
I have seen no data and have no formal data either, just my own observations and impressions, but here they are:

There is thickening that occurs in AS2, AS3 and Ceramique; cheapo silicon based stuff seems to thicken very slightly as well, but even less than the AS products. I have had systems running for more than 6 months on one application with no performance degradation- BUT if the heatsink/waterblock is disturbed cpu temp does go up. A Jiggle while the compound is hot fixes this. (I suspect that the movement damages the TIM joint and the HotJiggle repairs it.)

I have removed cooling devices that were in place over 6 months and can state that AS2 and AS3 afetr that time are definitely NOT the same consistency as they were initially- exactly what the process that changes them is, I do not know.
Heat, vibration, pressure and time= thicker compound is what I have seen though.

NOTE- my Jiggling methods may not be for all- especially the HotJiggle.
Messing with the cooling device on a modern cpu while running is dangerous and can lead to almost instant cpu death if something goes wrong- beware!

Last comment: thermal compounds are so good now that virtually any of them will get you good results. Time and many reviews have proven to me beyond all shadow of doubt that application is more critical for performance than which compound you use. I've had RadioShack white junk keep a TBird 1000 under 45c (indicated) when run at 1400mhz and 1.85 vcore.

second rogerdugans' post...i've noticed very little difference between as3, the vantec goop and generic stuff...all seem to fall w/in 1-1.5C of each other if i take my time...

I emailed Nevin again:

>Nevin,
>
>one last thing. Does AS really THICKEN? Please characterize the thickening,
>viscosity over time. Or at least the viscosity of the un-thickened and
>fully-thickened product. Please state also the ASTM number, for reference.
>
>Thank you.
>
>BTW you might want to put datasheets on your website.

Click to expand...

His response:

I don't think you will find any thermal compound that quotes exact
viscosity. Shin Etsu, for example, quotes "less than 3000 poise" or
something like that.

As the surface areas and other characteristics of each type of filler vary
from batch to batch, the suspension oils, and filler ratios are adjusted
for each batch to maintain maximum thermal performance. The viscosity
changes from batch to batch are irrelevant to the performance of the
compound in its intended application. This is quite different from a
lubricating grease where viscosity is a critical specification.

Yes, AS3 does thicken slightly over time in use. How and how much is not
information we care to hand to our competitors. Since you have asked this
same question several times and do not seem to appreciate that we spend
hundreds of hours developing formulas with proprietary secret combinations
of materials and oils, I recommend that you use thermal compound from a
different company that provides more complete specifications. AOS comes to
mind. None of their thermal compounds will perform up to the level of AS3
or Ceramique, but they do provide fairly extensive specifications.
http://www.aosco.com/nonsilicone.html

Nevin House
Arctic Silver, Inc.

Click to expand...

Actually ALL the REPUTABLE TIM manufacturers quote viscosity in their datasheets. e.g.

Bergquist TIC-7500:
viscosity @ 25*C - 500,000cps @ 2RPM

Thermagon T-grease 401:
viscosity @ 22*C - 700,000 (cP) @ 2RPM


Apparently he doesn't know the viscosity of his product (instead gives excuses on batch-to-batch variations)... and goes on to say that viscosity is irrelevant to the performance!!!


Would it be safe to assume that the thickening is NEGLIGIBLE and is nothing but a marketing gimmick?