Print 24 comment(s) - last by jtemplin.. on Oct 4 at 11:44 AM

Purdue Researchers Develop Carbon Nanotube Thermal Interface Material  (Source: Purdue University)
Purdue researchers develop thermal interface material using carbon nanotubes

Traditional thermal interface materials include greases, waxes and a foil made from a metal called indium. The most common type of thermal interface material most computer enthusiasts are used to is a paste with a white or silver color. Most heat sinks for computer use also include a pink colored pad on them to conduct heat from the CPU to the heat sink.

A group at Purdue University has demonstrated a way to make a more efficient thermal interface material using carbon nanotubes. The researchers on the project include Placidus B. Amama, a postdoctoral research associate at the Birck Nanotechnology Center in Purdue's Discovery Park, doctoral student Baratunde A. Cola, Timothy D. Sands, director of the Birck Nanotechnology Center and the Basil S. Turner Professor of Materials Engineering and Electrical and Computer Engineering; and Xianfan Xu and Timothy S. Fisher, both professors of mechanical engineering.

The researchers have demonstrated a method of growing what’s described as a forest of tiny carbon nanotube cylinders on the surface of a computer chip to enhance the flow of heat to a critical point where the chip connects to a cooling device according to Purdue University News. The nanotube thermal interface material being grown directly on the surface of the chip allows the material to completely fill gaps and gouges on the surface of the chip.

Cola claims, “The method developed by the Purdue researchers enables them to create a nanotube interface that conforms to a heat sink's uneven surface, conducting heat with less resistance than comparable interface materials currently in use by industry." Cola also said, "The tubes bend like toothbrush bristles, and they stick into the gaps and make a lot of real contact.”

The carbon nanotubes are grown directly on the surface of the chip using templates created from branching molecules called dendrimers. The dendrimers are then seeded with catalyst particle-laden silicon inside a chamber and exposed to methane gas. The application of microwaves breaks down the methane that contains carbon. The catalyst particles prompt the nanotubes to assemble and grow vertically on the chip.

Carbon nanotubes have already proven their use in computing thermaldynamics.  OCZ Technology announced earlier this year its intention to use a CNT-based materials in high-end consumer heatsinks.

There is no word on when or if this process might see commercial applications. However, the better the thermal interface material is at transferring heat from the chip to the heat sink, the smaller the heat sink can be. This will lead to smaller cooling systems, smaller computers, and less power required to cool the chip.

Computer manufacturers should be keen on exploring this technology further. Carbon nanotubes are being looked at for a variety of applications from nanotube batteries to using nanotubes to send impulses to nerves inside the body.

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

By scrapsma54 on 10/3/2007 8:46:20 AM , Rating: 1
Wouldn't this make the chip unsafe? I don't mean to break the positive feedback, Lol, but Nanotubes can easily be broken by the human hand. This will have to go with a lot of approval or it wont make it into comercial production.

RE: RoHs
By Martin Blank on 10/3/2007 10:00:33 AM , Rating: 1
You must have some really, really tiny hands, since the longest carbon nanotubes yet manufactured are only 18mm long.

That aside, carbon nanotubes have a tensile strength 50 times greater than that of high-carbon steel. Unless you're something out of DC or Marvel Comics, I doubt you're going to be able to pull apart a nanotube.

RE: RoHs
By ADDAvenger on 10/3/2007 10:15:01 AM , Rating: 2
Pulling it apart and snapping it are two very different things. I can easily snap a piece of wood, but there's no way I can pull it apart. We're talking about shear strength, you're thinking tensile.

RE: RoHs
By Cogman on 10/3/2007 11:02:41 AM , Rating: 2
Sheer strength will be EXTREMELY high, in fact, it will be the longer nano-tubes that will be the safest because of how high their sheer strength. Think of Steel, it is basicly a mess of loosly connected materials that just happens to like to stick together.

Now think of CNT's, they are CHEMICALLY bonded together. Do you have any Idea how strong these chemical bonds are? Yeah, go read up on that. In the meantime you should learn how to use the Wiki before posting. CNT's are extremely strong in all stress applications. Not only that, but they are generally 10 times stronger then steal as far as stress measurements go. The only place they where not so hot was compression strength, they tend to buckle (not break) when pressure is put on them.

As far as the cancer thing goes, well, what doesn't cause cancer now a days? I can hardly believe that one or 2 carbon fibers in the skin are going to be any worse then 1 or 2 seconds in the sun.

RE: RoHs
By Meaker10 on 10/3/2007 11:12:53 AM , Rating: 2
Eh? Everything is bonded together chemically, they are sharing electrons and thus being bound together (not getting to in depth here) and while there are different types of bonding, its more to do with the structure that makes a substance harder.

Even water molecules are chemically bound to each other ^-^.

RE: RoHs
By Meaker10 on 10/3/2007 11:15:45 AM , Rating: 2
By everything I mean anything in contact with another object. I was mostly talking about a single substance binding to itself to form a mass.

RE: RoHs
By Golgatha on 10/3/2007 2:48:05 PM , Rating: 4
Water molecules are hydrogen bonded together, but carbon nanotubes are covalently bonded together, as in they have atomic orbitals overlapping.

Hydrogen bonds are on the order of 5kcal/mol and covalent carbon-carbon single bonds are on the order of 80kcal/mol. This is why water flows and rocks don't. It's all about the energy it takes to break the bonds that are holding the substance together.

RE: RoHs
By geddarkstorm on 10/3/2007 5:34:31 PM , Rating: 2
Usually "chemically bonded" is a term saved solely for covalent bonds. As even totally uncharged, non protonated molecules can have electrostatic interactions (van der Waals and London dispersal) that can hold them together. For instance, geckos stick to any surface through van der Waals, even though those are the weakest molecular interaction there is.

But that's the key, it's a molecular interaction. Ionic, hydrogen, "hydrophobic", and van der Waals are all interactions; only covalent bonds are actual physical linkages between atoms and thus are "chemically bonded"

RE: RoHs
By ADDAvenger on 10/3/2007 12:10:48 PM , Rating: 3
I'm not saying one way or the other whether or not their SHEAR (not sheer) strength is high, simply because I don't know, my point was simply that there are different kinds of strengths, and it's not a good idea to confuse them. And pay attention to your spelling, it makes you look like a doofus when you confuse simple things like steal and steel.

RE: RoHs
By Cogman on 10/3/2007 10:11:11 PM , Rating: 2
:) spelling was never my high point. But, the information is correct, again, a simple wiki will verify what I was saying. You could do a google search for more extensive results.

RE: RoHs
By ADDAvenger on 10/4/2007 10:00:14 AM , Rating: 2
Yeah I've wiki'd it since then. CNTs buckle instead of breaking, but wouldn't a disformed heatsink be just as worthless as a split one? I'm imagining trying to seat the thing and smashing the block into some kind of parallelogram. Also, even if it was a real parallelogram and the end was still flat, the buckling changes the properties of the tubes, so its heat transferring abilities would be shot either way.

RE: RoHs
By scrapsma54 on 10/3/2007 10:18:39 AM , Rating: 2
Put your finger on the surface and I am willing to bet that 1,000,000 nanotubes aka nano particles will pass your skin and get into your blood stream. I am not talking about breaking the tubes themeselves I am talking about breaking them off their base or the chip. Also nano particles have been proven to cause neurological brain damage in humans and in rats. They can deteriorate cells and even cause cancer because they are capable of being embeded in dna.

RE: RoHs
By scrapsma54 on 10/3/2007 10:26:38 AM , Rating: 2
Think of the finger as an 18 wheeler dropped vertically (hey god is that strong, what can I say) on top of a feiled titanum poles that are sticking deep in the ground (no cement just dirt) and that truck is dragged along that feild (vertically), chances are those poles will get knocked over and become ebeded in the grill. Now the human hand is comparably considerably larger and will easily break those with a single press of a finger.

RE: RoHs
By Cogman on 10/3/2007 11:19:12 AM , Rating: 2
Again, tensile, sheer, and elastic moduli are all extremely high on CNT, it is not as simple to break them as you are making it too be.

BTW, did you ever get soot on your hand? Good Lord! You have had cancer causing CNT's embedded in you! Do a quick google search of CNT's and cancer. Most articles you will see are about using CNT's to combat cancer, none that I could see where about it causing cancer (some byproducts might though). I need your help, where is it again that carbon nanotubes cause neurological brain damage? I really am struggling to find an article that suggests that (most say they can be used to treat that too...)

RE: RoHs
By masher2 on 10/3/2007 4:09:44 PM , Rating: 2
> "I need your help, where is it again that carbon nanotubes cause neurological brain damage?"

He's probably confused CNTs with buckyballs. Fullerenes have been shown to cause a mild form of brain damage in certain species of fish-- at least when directly exposd to brain tissue. Its unclear whether they'd be able to cross the blood-brain barrier to reach that brain tissue....and of course buckyballs and CNTs are two totally different things.

RE: RoHs
By geddarkstorm on 10/3/2007 5:39:33 PM , Rating: 2
Exactly. And there has been work on using CNTs to fight cancer by coating them with folic acid that cancer cells preferentially take up, and then using infrared lasers to selectively super heat the CNTs. I have no idea how that research has turned out however; but it seems incredibly promising. See

RE: RoHs
By jtemplin on 10/4/2007 11:44:07 AM , Rating: 2
He may have confused the fish studies with CNT, however there have been studies documenting CNT toxicity in the lungs of mice. In this post I will use (---) to differentiate between my thoughts and the articles/abstracts that support them.
Pulmonary Toxicity of Single-Wall Carbon Nanotubes in Mice 7 and 90 Days After Intratracheal Instillation
Chiu-Wing Lam*,{dagger},1, John T. James*, Richard McCluskey* and Robert L. Hunter

Nanomaterials are part of an industrial revolution to develop lightweight but strong materials for a variety of purposes. Single-wall carbon nanotubes are an important member of this class of materials. They structurally resemble rolled-up graphite sheets, usually with one end capped; individually they are about 1 nm in diameter and several microns long, but they often pack tightly together to form rods or ropes of microscopic sizes. Carbon nanotubes possess unique electrical, mechanical, and thermal properties and have many potential applications in the electronics, computer, and aerospace industries. Unprocessed nanotubes are very light and could become airborne and potentially reach the lungs. Because the toxicity of nanotubes in the lung is not known, their pulmonary toxicity was investigated. The three products studied were made by different methods and contained different types and amounts of residual catalytic metals. Mice were intratracheally instilled with 0, 0.1, or 0.5 mg of carbon nanotubes, a carbon black negative control, or a quartz positive control and euthanized 7 d or 90 d after the single treatment for histopathological study of the lungs. All nanotube products induced dose-dependent epithelioid granulomas and, in some cases, interstitial inflammation in the animals of the 7-d groups. These lesions persisted and were more pronounced in the 90-d groups; the lungs of some animals also revealed peribronchial inflammation and necrosis that had extended into the alveolar septa. The lungs of mice treated with carbon black were normal, whereas those treated with high-dose quartz revealed mild to moderate inflammation. These results show that, for the test conditions described here and on an equal-weight basis, if carbon nanotubes reach the lungs, they are much more toxic than carbon black and can be more toxic than quartz, which is considered a serious occupational health hazard in chronic inhalation exposures.
CNT may have uses in drug delivery to malignant tissue, but it seems in their raw form they pose a significant hazard to the lungs. Before CNTs are used as drug carriers, more work needs to be done to figure out how to get them to a non-reactive state, so that they will only bind to the cancer itself and not wreak havoc anywhere else.

Regarding C-60 toxicity, it has been shown to be cytotoxic to neural cells of largemouth bass when in aqueous solution. 500ppb was shown to lead to a 17-fold increase in lipid peroxidation in neural tissue. 500ppb is a fairly minute amount of material present to cause such a significant result. Imagine if a CNT spill led to a contamination of local groundwater. I would not want to be drinking that water my friend. C-60 is apparently very soluble.
Manufactured nanomaterials (fullerenes, C60) induce oxidative stress in the brain of juvenile largemouth bass.
Oberdörster E.
Duke University Marine Laboratory, Beaufort, North Carolina, USA
(Full text: )
C-60 has also been shown to be cytotoxic through lipid peroxidation to astrocytes(cells that underlie the BBB), human dermal fibroblasts(skin tissue), and human liver cancer cells.
Nano-C60 cytotoxicity is due to lipid peroxidation
(Full text:
Christie M. Sayesa, Andre M. Gobinb, Kevin D. Ausmanc, Joe Mendeza,
Jennifer L. Westb,c,, Vicki L. Colvina,c
Not only can the fullerenes get to the brain tissue and cause the lipid peroxidation, they can damage the astrocytes and the BBB itself. These are not insignificant findings.

The article below also supports the notion that nanomaterials are can permeate the BBB as well as damage the astrocytes themselves.
Nanoparticle surface charges alter blood-brain barrier integrity and permeability.
Lockman PR, Koziara JM, Mumper RJ, Allen DD.
Department of Pharmaceutical Sciences, School of Pharmacy, Texas Tech University Health Sciences Center, Amarillo, TX 79106-1712, USA.

PURPOSE: The blood-brain barrier (BBB) presents both a physical and electrostatic barrier to limit brain permeation of therapeutics. Previous work has demonstrated that nanoparticles (NPs) overcome the physical barrier, but there is little known regarding the effect of NP surface charge on BBB function. Therefore, this work evaluated: (1) effect of neutral, anionic and cationic charged NPs on BBB integrity and (2) NP brain permeability. Methods: Emulsifying wax NPs were prepared from warm oil-in-water microemulsion precursors using neutral, anionic or cationic surfactants to provide the corresponding NP surface charge. NPs were characterized by particle size and zeta potential. BBB integrity and NP brain permeability were evaluated by in situ rat brain perfusion. RESULTS: Neutral NPs and low concentrations of anionic NPs were found to have no effect on BBB integrity, whereas, high concentrations of anionic NPs and cationic NPs disrupted the BBB. The brain uptake rates of anionic NPs at lower concentrations were superior to neutral or cationic formulations at the same concentrations. CONCLUSIONS: (1) Neutral NPs and low concentration anionic NPs can be utilized as colloidal drug carriers to brain, (2) cationic NPs have an immediate toxic effect at the BBB and (3) NP surface charges must be considered for toxicity and brain distribution profiles.
I think any reasonable person would acknowledge there is plenty of evidence that nanomaterials are not harmless. While they have amazing properties that we should exploit, we need to be sure that we can harness them safely and in a way that will not jeopardize our health and the health of the animals we share this planet with.

variable results
By mindless1 on 10/3/2007 6:55:59 AM , Rating: 2
IMO, the superiority of this depends quite a bit on how imperfect the mating surfaces are. While I've sold plenty of non-o'c systems that needed no further thermal margin than the provided 'sink allowed, on personal use systems where I'd o'c them, the surface irregularities were lapped and polished out to the point where i find it unklikey that this accomplishment will have any cost effective application.

In other words, I feel it less important than that heatsink manufacturers provide a good flat machined polished finish. Carbon nanotubes cost < heatsink with good enough finish that the difference is negligible.

I don't discount that there may be other useful applications, rather it is not so useful in a PC.

RE: variable results
By Blight AC on 10/3/2007 9:20:59 AM , Rating: 2
Well, PC's aren't the only place a Processor is used, this might be quite useful for battery powered devices like Laptops, PDA's, PMP's and UMPC's and other pre-assembled hardware like gaming consoles. Lower power cooling solutions will help extend battery life or even just help the Next Gen gaming consoles be quieter and a bit more tolerable as a living room multimedia device (although, I really have no problem with the Xbox 360's fan noise as is).

It might also be nice to have heat sinks come with CNT TIM on it, and all you have to do is install it, instead of trying to apply a nice thin TIM layer from paste (I have no luck with this). Especially if it will help get a better overclock off of air cooling.

Course, that depends on if the CNT's are durable, like toothbrush bristles and not brittle. If they are durable, it would also be useful for folks who swap out heat sinks as well (reviewers).

By Moishe on 10/3/2007 8:46:02 AM , Rating: 2
It almost sounds like the mating between the heatsink and the CNT will be so perfect that it'll almost be like permanently attaching the chip to the sink.

Would you be able to get the heatsink back off?

Does the heatsink effectively become part of the chip? (At what point do the two separate pieces become one?)

RE: cool
By geddarkstorm on 10/3/2007 5:42:48 PM , Rating: 2
No, it's all van der Waal interactions between the CNTs and the heat sink (weakest molecular interactions there are, around 1-2Kcal per mole). You might feel a light "stickiness", like with a gecko, but it should come right off.

I like the idea.
By Misty Dingos on 10/3/2007 7:51:14 AM , Rating: 2
I think that this may have a greater impact than some think. They keep the shuttle from burning up on re-entry with a carbon based heat sink. Pure carbon has the highest heat conduction of any substance on the planet. I don’t think that we are going to see the water cooled systems changing over to carbon NT heat sinks but I would not be surprised to see a CNT film interface between chips and heat sinks. Artic Silver might be come Artic Black and be a film instead of a paste.

You might also be able to improve everything from water blocks to regular heat sinks with the inclusion of CNT in the design. In water systems it should also have the advantage not having galvanic corrosion problems.

OK this is way out there. But this stuff, if you could make a sheet of it large enough, would be great as a cooking surface.

And for the really far out there. It will help with the mitigation of global warming. Hey it is carbon and the carbon is being locked up so it can't enter the atmosphere.

limited usefullness
By markitect on 10/3/2007 8:05:09 AM , Rating: 2
I would say that for at least 10 years the technology will have no cost effective applications, not even NASA or military. Also by its nature the tubes have to deform when the heatsink is applied, I wonder how many times you can do that before they start breaking off.

Forget the paste
By Shadowmaster625 on 10/3/2007 4:12:09 PM , Rating: 2
what we need is a better integrated heat spreader. I imagine they will be able to make better IHS's using nanotubes. Imagine instead of having a flat chip with a flat IHS , you'd have a thicker chip with ribs or fins built right into the IHS. Then the HSF (with a mating fin pattern) would mount onto the IHS, taking advantage of 10~50 times the surface area for heat transfer. Perhaps they will be able to run these nanotubes directly from the silicon up into the fins of these newly designed IHS's , resulting in much better core-to-case thermal transfer characteristics. I do think that is what needs to be improved, especially since the cores in my Q6600 run at 70C even though my cpu case temperature reads 48C. This is unacceptable to me because it shows that my cooling solution is working just fine (under 50C is quite cool for a loaded temp) but the chip itself simply cannot efficiently transfer heat out of the cores and to the IHS. That sounds like a job for these nanotubes! I also think they'd be able to start stacking layers and layers of transistors into a 3D matrix structure if they could use these nanotubes as "cooling pipes"... but now I'm just rambling...

"If a man really wants to make a million dollars, the best way would be to start his own religion." -- Scientology founder L. Ron. Hubbard
Related Articles

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki