It is my understanding that metals are a crystal lattice of ions, held together by delocalized electrons, which move freely through the lattice (and conduct electricity, heat, etc.).

If two pieces of the same metal are touched together, why don't they bond?

It seems to me the delocalized electrons would move from one metal to the other, and extend the bond, holding the two pieces together. If the electrons don't move freely from one piece to the other, why would this not happen when a current is applied (through the two pieces)?

  • 33
    $\begingroup$ I remember hearing something about how this was an issue in space when cutting metal that it could fuse back together because the outer layer didn't oxidize (no oxygen, go figure) or something like that. $\endgroup$
    – zzzzBov
    Commented Nov 19, 2013 at 21:46
  • 2
    $\begingroup$ @cinco, no - I started wondering about it in a physics course on materials a few months ago and it has(had) been bugging me since then. What does this have to do with 3D printing? $\endgroup$
    – jcw
    Commented Nov 20, 2013 at 17:09
  • 1
    $\begingroup$ @jcw I mentioned 3D printing because this principle, basically cold welding, associated with nanoprocessing, has a great potential (in my opinion) in 3D printing of metals. 3D printing is the future of industry, and finding new materials and processes that can give answer to the reals needs are the today's challenge. The fact that metals can cold weld is one of the principles that may be fundamental in this evolution $\endgroup$
    – cinico
    Commented Nov 20, 2013 at 21:17
  • 4
    $\begingroup$ IIRC, you can do this with gold quite well, because of reasons that are mentioned below: 2) gold doesn't oxidise; it's quite soft, which means that 1) the bonds are not as strong (and as difficult to create considering the low energy barrier) and 2) small pressure is enough to reduce the roughness of the surface. $\endgroup$
    – yo'
    Commented Nov 21, 2013 at 10:20
  • 19
    $\begingroup$ Two pieces of mercury bond quite well ;-) $\endgroup$
    – Jitter
    Commented Jan 4, 2014 at 9:19

10 Answers 10


I think that mere touching does not bring the surfaces close enough. The surface of a metal is not perfect usually. Maybe it has an oxide layer that resists any kind of reaction. If the metal is extremely pure and if you bring two pieces of it extremely close together, then they will join together. It's also called cold welding.

For more information:

  • 1
    $\begingroup$ I believe extremely clean and flat glass slabs will also bond together when placed in contact, but for an altogether different reason than delocalized electron bridging. $\endgroup$ Commented Nov 19, 2013 at 22:22
  • 22
    $\begingroup$ There's even an ancient engineering tool, a set of metal parallelepipeds to measure length (something like weights for weighting). They're so polished that they almost weld when touch. Edit: found it! It's called "gauge blocks". en.wikipedia.org/wiki/Gauge_block $\endgroup$ Commented Nov 20, 2013 at 4:03
  • 28
    $\begingroup$ There is a serious lack of videos about this on Youtube $\endgroup$ Commented Nov 20, 2013 at 6:24
  • 2
    $\begingroup$ "It's also called cold welding." It is also part of what happens in the galling of threaded fastener. A point that can come up a lot in assembling ultra-clean parts for low background experiments in particle physics. $\endgroup$ Commented Nov 20, 2013 at 15:21
  • 8
    $\begingroup$ @polkovnikov.ph gauge blocks are as much 'an ancient engineering tool' as a motor car is an ancient form of transport. Both date from the just before the turn of the 20th century, and both are still in current use. $\endgroup$ Commented Nov 20, 2013 at 16:52

They do, as Feynman said. If you have two copper pieces perfectly polished and you put them in contact, they will weld automatically (the copper atoms won't know what piece they belonged to).

But in real life, oils, oxides and other impurities don't allow this process.

Found it! Read Feynman's own words (where $\mu$ = coefficient of friction):

If we try to get absolutely pure copper, if we clean and polish the surfaces, outgas the materials in a vacuum, and take every conceivable precaution, we still do not get $\mu$. For if we tilt the apparatus even to a vertical position, the slider will not fall off—the two pieces of copper stick together! The coefficient $\mu$ , which is ordinarily less than unity for reasonably hard surfaces, becomes several times unity! The reason for this unexpected behavior is that when the atoms in contact are all of the same kind, there is no way for the atoms to “know” that they are in different pieces of copper. When there are other atoms, in the oxides and greases and more complicated thin surface layers of contaminants in between, the atoms “know” when they are not on the same part. When we consider that it is forces between atoms that hold the copper together as a solid, it should become clear that it is impossible to get the right coefficient of friction for pure metals.

The same phenomenon can be observed in a simple home-made experiment with a flat glass plate and a glass tumbler. If the tumbler is placed on the plate and pulled along with a loop of string, it slides fairly well and one can feel the coefficient of friction; it is a little irregular, but it is a coefficient. If we now wet the glass plate and the bottom of the tumbler and pull again, we find that it binds, and if we look closely we shall find scratches, because the water is able to lift the grease and the other contaminants off the surface, and then we really have a glass-to-glass contact; this contact is so good that it holds tight and resists separation so much that the glass is torn apart; that is, it makes scratches.

Source: http://www.feynmanlectures.caltech.edu/I_12.html

  • 6
    $\begingroup$ @jcw A current just means there is a bulk motion of the electron cloud - there are no more or fewer electrons to bond things, and they probably aren't spending any more time than usual bridging the interface. $\endgroup$
    – user10851
    Commented Nov 19, 2013 at 18:46
  • 5
    $\begingroup$ @jcw - free electrons can't form a bond. electrons which are in an orbital create a bond by essentially "time-sharing" orbitals owned by neighboring nuclei. $\endgroup$ Commented Nov 19, 2013 at 18:57
  • 16
    $\begingroup$ I have witnessed this, with extremely precise height-gauge blocks which had an incredibly flat finish. Holding them together for 30 seconds or so made them tough to separate. Quite amazing to see. My lecturer referred to it as 'cold-welding'. $\endgroup$
    – deed02392
    Commented Nov 19, 2013 at 20:21
  • 2
    $\begingroup$ @jcw An electric current can break through the oxidation layers. This is called fritting. Though not an instance of welding, it explains why high voltage electrical contacts don't have to be maintained much. Even a pretty tarnished appliance plug, for instance, will work fine. But, say, small-signal connectors (e.g. audio) will not perform well if they are not clean and polished. (It's good for high voltage connectors to be in good condition too, so that fritting doesn't have to be relied on so much. A high initial contact resistance that has to be broken down could briefly generate heat.) $\endgroup$
    – Kaz
    Commented Nov 19, 2013 at 22:09
  • 3
    $\begingroup$ @deed02392 Note that large flat objects pressed together are also hard to separate simply due to atmospheric pressure. The space between them is a void, and air has to rush in when you separate them. Early in the separation, the crevice through which air can get in is small. $\endgroup$
    – Kaz
    Commented Nov 19, 2013 at 22:12

Two reasons:

  • Oxides
  • The roughness of the surface

If the surface is rough, then the majority of the surface is touching the air gap between the two, not the opposite surface. A bond may form at the touching "peaks", but it will be weak compared to the rest of the metal because a very small fraction of the surface has actually bonded.

In addition, metal surfaces adsorb oxygen and form oxides/oxygen monolayers on the surface. This is actually a visible process with metals like sodium and potassium (the color changes in a short time period). But for all metals, there still is oxide formation to a sufficient extent, because the edge metals have not completely fulfilled their valencies. Even a monolayer of adsorbed oxygen is enough to stop the surfaces from welding.

If two clean, flat metal surfaces are brought together (usually in a vacuum), they do indeed cold weld. This is hard to achieve for macroscopic objects because of the perfect flatness requirement, but is still possible. In practice, it is more commonly used for welding small things.

  • 1
    $\begingroup$ "all metals"? Even gold has an appreciable oxide layer? $\endgroup$
    – user10851
    Commented Nov 19, 2013 at 23:01
  • 1
    $\begingroup$ @ChrisWhite Appreciable=enough to stop welding. Gold doesn't form a whole layer of oxides, but it will form a monolayer of oxygen and other stuff. Which I'm calling "oxide layer" (probably shouldn't, will edit) $\endgroup$ Commented Nov 19, 2013 at 23:03

I believe this is essentially what happens in gilding, owing to the special properties of gold (malleability and lack of corrosion).

Extremely flat surfaces can get stuck together due to Van der Waals forces as well as air pressure. I once accidentially stuck two quartz optical windows together, and had a hell of a time separating them.

  • 14
    $\begingroup$ IIRC It's also how gold nuggets form in rivers via aggregation of particulate gold. $\endgroup$
    – Random832
    Commented Nov 19, 2013 at 21:42
  • $\begingroup$ @Random832 That appears to be false (although it was likely a common theory and does make sense). en.wikipedia.org/wiki/Gold_nugget#Formation It looks like when they examined the grain structure, it was consistent with a fragment of a lode that has been weathered, as compared to particulate grains that cold-welded. $\endgroup$
    – JMac
    Commented Jul 11, 2019 at 13:21

I can't comment since I don't have the reputation for it, but I do have some relevant knowledge from my research in materials science.

To add to what DumpsterDoofus said, it is very easy for two pieces of glass or polymer to bond if you clean them extremely well and ionize the surface. Look up plasma polymerisation.

Moreover, you'd be surprised how much "gunk" is actually on the surface of any given piece of material in standard atmospheric conditions. There is a reason why a lot of surfacial materials characterization techniques require ultra-high vacuum, as I recall it's about 1 atomic layer/second of deposition under $10^{-6}\, \mathrm{torr}$ (source). If you want two metals to bond without applying heat or force you'd need to get a vacuum better than that and then clean off the oxide layer.

You'd also be surprised at how much organic material is covering the surface of everything around you. Your fingers produce oil and they stick to the surface of everything you touch, and your dead skin flakes off all the time and covers the stuff around you. You can notice it if you take a metallic sample to an SEM, and shoot electrons at it to get an image of the surface. If it's got organic material on the surface, after a while the area where you shot electrons will turn dark, you can notice it if you zoom out or pan around. This is due to hydrocarbon contamination, usually from the oils on your fingers.


Metals with perfectly clean surfaces WILL bond together just like you explained, but that isn't the case in real life because there is a thin layer of oxygen blocking the metal's surface.

Much like how rust forms, thin layers of oxygen coat every metallic surface upon contact.

  • 2
    $\begingroup$ It's my understanding that over very long periods of time metals will bond. I'm pretty sure ancient treasure and metal found in tombs have been found bonded. $\endgroup$ Commented Nov 19, 2013 at 21:25
  • 9
    $\begingroup$ @BrandonEnright usually in my understanding what bonds them together is the oxide layers on the individual items merging, not the actual items themselves cold welding together. $\endgroup$
    – jwenting
    Commented Nov 20, 2013 at 8:56
  • $\begingroup$ @jwenting perhaps for ancient treasures yes, but wouldn't reasonably clean surfaces under pressure cold-weld together over time due to diffusion, even if there was an oxide layer? $\endgroup$ Commented Jun 20, 2014 at 14:16

There is much more to bonding than exchange of electrons. Especially because electrons that are part of electron cloud doesn't take part in crystal bonds.

All metals are basically crystals --- they have proper lattice, to weld two parts you'd have them to build common lattice (at least in the are where they are welded). When you weld properly you create liquid phase between (by heat or current) that crystallizes.

I might be mistaken, but this welding can be observed. For example if you have some old iron screws, that are screwed into element also made of iron (and not oiled properly), after time they can be very hard to unscrew, it was supposedly because of diffusion of atoms between both parts which was easy because of the same lattice structure. Once again: I heard this as an anecdote on some Physics lecture, but did not search any further proofs.

  • 1
    $\begingroup$ That is an interesting idea. I wonder if that has more effect than rust. $\endgroup$
    – jcw
    Commented Nov 19, 2013 at 20:25
  • $\begingroup$ I guess that this effect may be facilitated by the fact that if you screw it very tightly there is limited access to oxygen inside, hence oxidation and rust play lesser role. $\endgroup$
    – jb.
    Commented Nov 19, 2013 at 20:29

To add to the other's ideas on this topic, I think the concept of "Surface Potential" also plays a large role in this. Roughness interferes with the surface potential of the material because it creates gaps where the two metals cannot bond. This lowers the surface potential of the material.

Materials such as oxides, oils or other residues that can be found on metals also lower the surface potential of the material. This surface potential can be reduced by Van Der Waal's interactions, ionic interactions and other polar non polar molecular level interactions. Every molecule that comes in contact with the metal, whether it's air found on the surface or between gaps due to roughness or residues, has the potential to reduce it's surface potential.

Ex: Consider bonding in a metal of it's easier for you (in a semiconductor) where you are used to drawing Si-Si bonds to every neighboring atom, however, ON THE SURFACE of the atom the electrons cannot bond because there are no more available electrons. This causes surfaces of metal or other materials like semiconductors to be extremely reactive: thus forming oxide layers, or reducing their surface potential via the aforementioned atom-atom, atom-molecule interactions. (It's easier to consider Silicon as it has electrons that are "considered" attached to it's nucleus rather than a pure transition metal that has "free electrons" that people don't consider attached to the nucleus.

enter image description here


While simple contact between metals isn't enough for most metals to bond, relative motion will achieve the fusion between the metals (at small contacts). A common occurrence is seizing up of mechanical devices due to insufficient lubrication.

I don't think screws stick due to metal-metal bonding- its mostly simple distortion particularly of the threads and body of the screw. Damage a screw and insert in a tight space and you won't that screw out again.


It depends upon the purity of the metal. If the surface is well polished it is indeed possible to make bonds with the adjacent metal pieces. However, if there are oxides and other impurities present at the surface then bonding is not possible. This can be explained by the surface energy of the metal. Well, you can see that metals are bonded in powder metallurgy.


Not the answer you're looking for? Browse other questions tagged or ask your own question.