Research question: Delidding and Liquid Metal Corrosion?

[gr82banautiger]

So apparently there are reports around the internet that Liquid Metal will corrode onto copper heat blocks, and chemically bond to them, which also degrades the Liquid Metal.

 

Thus, the next obvious question is does delidding run the risk of corroding your copper/nickel plated IHS if using Liquid Metal as the thermal compound between it and the heat spreader?

 

This needs some testing, as there are a bunch of people looking at delidding the new 8th gen processors.  Maybe someone could contact Der8auer, I'm sure he has some older delidded processors he could crack open and take a look.

[Tiwaz]

yes overtime it fucks up anything except for nickel and gold

[gr82banautiger]

IHS's are nickel plated, so maybe it's fine to use between the CPU die and the Nickel plated IHS?

[TheGlenlivet]

Is the solution to not simply use the liquid metal between the die and heat spread and then good quality thermal paste between the heat spreader and block?  I thought I remember seeing a test of liquid metal instead of paste and there wasn't a huge benefit to using it instead of paste.

[gr82banautiger]

Just now, TheGlenlivet said:

Is the solution to not simply use the liquid metal between the die and heat spread and then good quality thermal paste between the heat spreader and block?  I thought I remember seeing a test of liquid metal instead of paste and there wasn't a huge benefit to using it instead of paste.

Absolutely.  If the Liquid Metal won't corrode onto the Nickel IHS, we're golden as long as you use non-conductive TIM between the IHS and the Copper cooler plate.

 

The question is whether the Liquid metal will corrode between the die and the IHS inside the delid.

[CrippledROBOT]

14 minutes ago, gr82banautiger said:

So apparently there are reports around the internet that Liquid Metal will corrode onto copper heat blocks, and chemically bond to them, which also degrades the Liquid Metal.

 

Thus, the next obvious question is does delidding run the risk of corroding your copper/nickel plated IHS if using Liquid Metal as the thermal compound between it and the heat spreader?

 

This needs some testing, as there are a bunch of people looking at delidding the new 8th gen processors.  Maybe someone could contact Der8auer, I'm sure he has some older delidded processors he could crack open and take a look.

With copper? I too have read reports that this will happen. Chemically speaking, gallium (which most if not all "liquid metal" solutions use) WILL chemically react with copper, though not to a degree of critical failure as with Aluminum. 

 

I have heard rumors of liquid metal "drying out" even while attached to the IHS (the nickle platted copper heat spreader that sits on the die). I cannot speak from experience here (though my basic chemistry knowledge says that this shouldn't happen). 

 

Let me bring in the big guns @done12many2. He has been a huge help with my process!

[TheGlenlivet]

Ah I see, my misunderstanding is the copper IHS.  I haven't seen one of those, only Nickle plate.

[gr82banautiger]

IHS's are Copper, they're just Nickel plated.  The question though is whether Liquid Metal will eventually corrode into the nickel plating or through it to the copper?

[CrippledROBOT]

2 minutes ago, gr82banautiger said:

IHS's are Copper, they're just Nickel plated.  The question though is whether Liquid Metal will eventually corrode into the nickel plating or through it to the copper?

Chemically speaking, it should not. What are your temperatures on your Coffee Lake system/system that is heating up?

[gr82banautiger]

 

[gr82banautiger]

Just now, CrippledROBOT said:

Chemically speaking, it should not. What are your temperatures on your Coffee Lake system/system that is heating up?

This is a research question because I'm planning to delid my i7-8700k and just want to make sure using Liquid Metal or Conductonaught won't corrode into the under side of the IHS.  Certainly if you scratch the nickel plating I could see it being a problem.

[CrippledROBOT]

6 minutes ago, CUDA_Cores said:

I'm not sure where you heard that, but aluminum is the only thing that liquid metal will destroy. The reason is because Liquid metal contains gallium, and gallium does THIS to aluminum:

 

 

Now I'm no chemist but as far as I know, Gallium has minimal/zero reaction to copper blocks. If you are really concerned, but a premium CPU cooler like the Phantecs TC-14PE should work perfectly, which has a nickel plated block. 

Agreed. While Copper and Gallium WILL react, it does not drastically CORRODE the material as it does Aluminum (Google Gallium and Copper reaction/corrosion). Mitigating this issue and perhaps entirely avoiding it can be done via using a nickle plated cooling mechanism (completely removing alumimum and copper from the equation).  http://forum.notebookreview.com/threads/something-to-think-about-liquid-metal-compatibility-with-copper-heat-sinks.800890/

 

 

*https://www.osti.gov/scitech/servlets/purl/811932/

 

"Liquid gallium at ~30°C also slowly reacts with copper to leave the copper surface pitted.9 This chemical reaction with copper is a concern if gallium or galinstan spills in the laboratory."

[CrippledROBOT]

Just now, gr82banautiger said:

This is a research question because I'm planning to delid my i7-8700k and just want to make sure using Liquid Metal or Conductonaught won't corrode into the under side of the IHS.  Certainly if you scratch the nickel plating I could see it being a problem.

It certainly may be! Though unless you need the thermal headroom for overclocking purposes (being held back by temperatures), max load temps on the i7 8700k of 75-80c are not the end of the world/damaging. Something to consider before killing your 3 year warranty at a minimum

[gr82banautiger]

1 minute ago, CrippledROBOT said:

Agreed. While Copper and Gallium WILL react, it does not drastically CORRODE the material as it does Aluminum (Google Gallium and Copper reaction/corrosion). Mitigating this issue and perhaps entirely avoiding it can be done via using a nickle plated cooling mechanism (completely removing alumimum and copper from the equation). 

The corrosion I'm talking about is the Liquid metal will degrade, causing it to potentially cause air pockets or not create good thermal contact between the die and the IHS over time.

[gr82banautiger]

6 minutes ago, CrippledROBOT said:

It certainly may be! Though unless you need the thermal headroom for overclocking purposes (being held back by temperatures), max load temps on the i7 8700k of 75-80c are not the end of the world/damaging. Something to consider before killing your 3 year warranty at a minimum

Right, which is why I'm researching it before I do it.  I plan to push a pretty high OC onto the processor if possible.  I'm doing a 10 year anniversary build to replace my 24/7 oc'd QX9650 build from Dec 2007.

[CrippledROBOT]

30 minutes ago, gr82banautiger said:

The corrosion I'm talking about is the Liquid metal will degrade, causing it to potentially cause air pockets or not create good thermal contact between the die and the IHS over time.

Liquid metal such as Coollaboratory Liquid Ultra is quite literally a metal (gallium). If you had a piece of gallium the size of, say a US quarter, it would melt into a liquid in your hand due to its low melting point. Its a metal, quite like iron, aluminum, gold, etc. All matter is subject to degradation and corrosion.

 

I too am worried about doing it and risking something like this, and that is totally OK. If your case allows for it, get a large Corsair H115 or Kraken X62. This way you don't have to delid (saving ANY and ALL risks associated with delidding) and should keep your temperatures in check!  

 

** THOUGH a water cooler will not give the same PERFORMANCE as a delid, but it'll certainly help you to avoid killing your warranty. The ideal situation is to do both.

[gr82banautiger]

2 minutes ago, CrippledROBOT said:

Liquid metal such as Coollaboratory Liquid Ultra is quite literally a metal (gallium). If you had a piece of gallium the size of, say a US quarter, it would melt into a liquid in your hand due to its melting point. Its a metal, quite like iron, aluminum, gold, etc. All matter is subject to degradation and corrosion.

 

I too am worried about doing it and risking something like this, and that is totally OK. If your case allows for it, get a large Corsair H115 or Kraken X62. This way you don't have to delid (saving ANY and ALL risks associated with delidding) and should keep your temperatures in check!

 

Yep, I am doing an EK-S280 kit, so I can eventually expand it to liquid cool a GPU (adding additional radiators at the time) instead of an AIO.

 

And yes, all thermal will eventually degrade, but if it degrades in a few months that's a much different problem than degrading over say 10 years when we're talking about a delid.

[done12many2]

20 minutes ago, CrippledROBOT said:

With copper? I too have read reports that this will happen. Chemically speaking, gallium (which most if not all "liquid metal" solutions use) WILL chemically react with copper, though not to a degree of critical failure as with Aluminum. 

 

I have heard rumors of liquid metal "drying out" even while attached to the IHS (the nickle platted copper heat spreader that sits on the die). I cannot speak from experience here (though my basic chemistry knowledge says that this shouldn't happen). 

 

Let me bring in the big guns @done12many2. He has been a huge help with my process!

 

13 minutes ago, CrippledROBOT said:

Agreed. While Copper and Gallium WILL react, it does not drastically CORRODE the material as it does Aluminum (Google Gallium and Copper reaction/corrosion). Mitigating this issue and perhaps entirely avoiding it can be done via using a nickle plated cooling mechanism (completely removing alumimum and copper from the equation). 

 

Hey bud.  There is a reaction, but it's of a staining nature only.  While technically it is corrosion, it is not to the point of degradation in performance.   

 

To put it simple, use LM on your IHS and it will stain your IHS.  This applies to both sides of the IHS.  The staining can be removed with a metal polish solution, but may result in the OEM Intel etching on the IHS becoming less visible. While I have used and may use LM on the top of the IHS (between IHS and waterblock/cooler), I don't recommend it for anyone that doesn't need those last few degrees of cooling.  It's just not worth the trouble.

 

As for use on the inside between the die and IHS.  Perfectly fine.

[CrippledROBOT]

5 minutes ago, gr82banautiger said:

Right, which is why I'm researching it before I do it.  I plan to push a pretty high OC onto the processor if possible.  I'm doing a 10 year anniversary build to replace my 24/7 oc'd QX9650 build from Dec 2007.

As I too have done a fair bit of research, I have found a mixed bag of results, HOWEVER, if areas are cleaned properly and the process is done correctly, you have done all that you really can to mitigate any issues. Metals like to react with one another, its just inevitable. Some will react more than others. The reaction between Nickle and Gallium is next to none. Read this: http://forum.notebookreview.com/threads/something-to-think-about-liquid-metal-compatibility-with-copper-heat-sinks.800890/

 

This explains the reaction from a chemistry perspective. "Electrochemically speaking, gallium has a potential of -0.53 volts and copper has a potential of +0.334 volts. The difference between the two metals is going to create a "battery" with a potential of 0.864 volts. Gradually this battery will deplete; as the gallium migrates over to the copper. You may have heard some stories about how gallium "dries up" and evaporates. In many of those cases it is likely that it acted like a battery and migrated over to and became part of the copper heat sink. If you take apart the heat sink from the CPU and clean it up you may find that the copper heat sink is colored a silverish-grey that resists efforts to even buff it off with a scrubbing pad. The stuff you can polish off is a corrosive residue of oxidized gallium and the stuff you cannot remove easily is now an alloy. One good piece of news is that if this happens repeatedly the battery action will stop, as it is pretty inefficient to put gallium and have it migrate over to a gallium-copper alloy."

- Credit belongs to original poster at the link attached. 

 

 

In short, if you want that extra overclocking headroom and are willing to disregard the Intel warranty, delidding is certainly a good option, THOUGH a strong water cooler will allow you to get the same or similar overclocks WITHOUT physically altering the chip, possibly saving you a warranty and bringing you piece of mind.

 

As @done12many2 said (and as the article that I linked noted), the reaction with copper is creating an entirely new AND stable alloy. It will likely, over time, show "evidence of corrosion" but this is nothing to be afraid of and is really just a stain. Reactions are occurring, but nothing that would cause catastrophic failure. 

 

 

[gr82banautiger]

2 minutes ago, CrippledROBOT said:

The reaction between Nickle and Gallium is next to none. 

Right, I've read the articles, what I wasn't sure of is how much is "next to none"?  If this process will take 4-5 years or more, then deliding every 4-5+ years isn't a big deal.  Delidding every say 3 months would be a whole different ball game.

[CrippledROBOT]

1 minute ago, gr82banautiger said:

Right, I've read the articles, what I wasn't sure of is how much is "next to none"?  If this process will take 4-5 years or more, then deliding every 4-5+ years isn't a big deal.  Delidding every say 3 months would be a whole different ball game.

Even in your WORST case scenario, you certainly won't be delidding every 3 months. This process SHOULD take about 3-4 years. 

[gr82banautiger]

1 minute ago, CrippledROBOT said:

Even in your WORST case scenario, you certainly won't be delidding every 3 months. This process SHOULD take about 3-4 years. 

So if that number is right, then people using 'delidding services' should be aware that they may need to refresh the LM it in that time frame.  Of course that's probably around the same time frame (or longer) than most people replace their PCs, so probably not an issue for folks that replace their PC's often.

 

 

[done12many2]

1 minute ago, gr82banautiger said:

So if that number is right, then people using 'delidding services' should be aware that they may need to refresh the LM it in that time frame.  Of course that's probably around the same time frame (or longer) than most people replace their PCs, so probably not an issue for folks that replace their PC's often.

 

 

 

@Silicon Lottery has been delidding and selling binned chips for a great deal of time now.  He hasn't had an issue related to LM degrading to date.  Maybe he can chime in as he's done it more than any individual I know of.  

 

https://siliconlottery.com/

 

I swear it's not a plug.  

[CrippledROBOT]

1 minute ago, gr82banautiger said:

So if that number is right, then people using 'delidding services' should be aware that they may need to refresh the LM it in that time frame.  Of course that's probably around the same time frame (or longer) than most people replace their PCs, so probably not an issue for folks that replace their PC's often.

 

 

That is my understanding as well. From what I gather, if you're that passionate about performance, then you're likely getting a new system rather frequently. 

 

Here's my breakdown:

 

* If you want to keep warranty and eliminate ANY chances of unusual chemical reactions, keep the chip as is (don't delid). Get a strong water cooler. THOUGH your temperatures won't be as good as if you de-lidded AND used a water cooler (at 5.0ghz expect the high 70s or low 80s, WHICH, is totally fine for max load gaming). You also won't be second guessing yourself down the road later on.

 

* If you do decide to delid, expect reduced temperatures (anywhere from 10-20c). Your experience with chemical reactions are limited to what has been discussed above. Your warranty is gone out the window and everything that you do with your CPU is COMPLETELY on you.

 

* At the end of the day, its up to you. Both routes have their pluses and minuses. 

 

 

Cheers!

[Silicon Lottery]

5 minutes ago, done12many2 said:

 

@Silicon Lottery has been delidding and selling binned chips for a great deal of time now.  He hasn't had an issue related to LM degrading to date.  Maybe he can chime in as he's done it more than any individual I know of.  

 

https://siliconlottery.com/

 

I swear it's not a plug.  

We've been selling delidded CPUs for over three years now, without a single report of thermal degradation over time. I still have access to a couple Ivy Bridge CPUs delidded near launch, and their temperatures are still the same as the first day they were delidded. Staining will always occur on nickel, but I see no reason to be worried about liquid metal drying out underneath the IHS. As it has been mentioned, the reaction is pretty minimal against nickel and silicon.