Yeah, a point well made to be very clear about that distinction.
“Run down” does not correctly convey what is going on, and is actually confusing. Even “wear out” is poor. Perhaps “degrade”.
OK, but I think this essentially restates the question. Why do the atoms in a permanent magnet stay lined up? Why don’t they drift into more random, non-magnetic orientations?
Yes, of course. Over a long enough time, the cushion of my chair may flatten out and split, and the wooden legs may crumble. If I want to preserve it longer, I will have to expend energy to do so. Why is a magnet different?
Is it just a question of the time horizon involved?
OK, I hold the magnet close to the fridge, but not touching it. then I let it go, it will pull itself to the surface of the fridge and stick there. If I detach the handle, hold it close to the fridge, and let it go it will fall to the floor.
see the distinction which is difficult for some to understand?
I would say “disorganized” is probably the best term. All the magnetism is still there but it’s no longer neatly arranged.
I agree with your distinction.
But how about we consider the fridge in orbit or deep space? IOW, in free fall. You detach the handle then let go of it and it hovers right there. It accelerating towards either the metal surface of the fridge or towards the floor would be equally surprising and mysterious.
My point is that uneducated humans’ “intuitive” physics sucks because it’s formed at the bottom of an atmosphere deep in a gravity well. If humans were magnetically susceptible such that fridge magnets stuck to their skin and strong magnets could pull them across a room none of this would be counterintuitive in the slightest.
So why does the handle stick to the fridge?
It does that because all the atoms are aligned such that there is an electromagnetic force holding them together. When you pry the handle off the fridge you break those bonds, and the handle falls to the floor. And when you put the handle back in place it just falls off again, because those bonds are permanently broken and randomized. But note something. You didn’t have to do work to keep the handle on the fridge, you had to do work to make it fall off.
Same thing with the magnet. It is held on by electromagnetic forces, the same sort as the forces that keep the handle stuck to the fridge.
The difference is that the individual atoms of the magnet are all lined up in a particular way. And doing work to pull the magnet off the fridge doesn’t randomize those lineups, the way it does when you pry the handle off the fridge. And so when you put the magnet back it sticks back on, unlike the handle.
But the real answer to “magnets–how do they work?” is that magnets work by electromagnetic forces. And these forces are the exact same forces that make solid objects solid. We’re confused when the magnet sticks to the fridge, but we aren’t confused when it doesn’t go right through the fridge, because that’s something we expect.
So here’s Richard Feynman trying to answer that question, and explaining that in order to explain it he has to understand what the real question is. And that electromagnetism is a fundamental property of the universe that can’t be analogized to anything else, because if we explain it like rubber bands or whatever we have to ask what makes rubber bands elastic, and the answer is electromagnetism.
Well all atoms are magnets. But almost no macroscopic materials are magnetic. So right from the start we can guess that there is some special property that a very small subset of materials have makes them, in some sense magnetic.
Cutting to the chase (and ignoring para-magnetism dia-magnetism and anti-ferro-magnetism) we are concerned with ferro-magnetism. There is a clue in the name right away. Iron is magnetic. Not just iron, but it is the first known and probably best known. (Nickel, Cobalt, a slew of rare earths, but Iron is the most important.)
But why? The only useful understanding involves quantum physics. Some atoms interact at very short distance by an interaction called the exchange interaction, one that has a tendency to have the atoms line up in the same direction. Because it is only a short range effect you tend to get small “domains” of atoms in a piece of iron or other ferromagnetic material. Inside a domain all the atoms are lined up. But when the distance gets larger the overall tendency of materials to minimise order (absent any external energy input to prod it into order) results in the domains all pointing in random directions. But the domains are physical alignments. If you do prod a domain with enough energy to get it to change its mind, it will stick in the new direction. The exchange interaction will keep all the atoms in the domain locked together. The trick is that all the atoms in the domain want to stay in the same direction, so if you want to change the direction you have to pump in enough energy to have all the atoms shift, or none will. So a ferro-magnet tends to stay set where it was last left. Heat, physical force or an external magnetic field can all be used to disturb it. Only the latter can be usefully employed to get the domains to line up together, and thus create a larger arrangement. Complicating matters, obviously a domain is affected by the magnetic field of the domains right next to it. So if a domain is surrounded by domains all pointing one direction, there will be a tendency for it to decide to join them. But because the domains flip as a unit, this behaviour of flipping direction in response to a magnetic field is not smooth. The macroscopic behaviour of a 3D volume of domains is seriously messy. Indeed there is no analytic model. In 2D there is. (A mate of mine did a lot of his research work in computational modelling of the Ising model.) This curious behaviour accounts for a lot of the strange ways magnets behave as they are magnetised and demagnetised.
This is a loose use of the term “energy”. Your chair may wear out, it may be subject to corrosion, to microbial attack, and a whole host of other ills. To preserve it longer you will need to protect it from these effects. But this action isn’t measured in energy alone. Although it could be argued that you are, in essence, battling entropy. But the energy you put into preserving your chair’s structure is not directly related to it doing its job of supporting you when you sit upon it. It probably degrades pretty much as fast when not in use as when being sat upon.
Thanks! This is fascinating. So there’s no classical explanation for why the atoms in permanent ferromagnets stay pointed in one direction?
I just meant that eventually something will overcome the electromagnetic energy holding the chair together. It was indeed a loose response.
Correct. Classically, adjacent magnets would rather point in opposite directions, so you need another effect that makes it energetically favorable for them to go against their raw tendency. This other effect (the exchange interaction that Francis Vaughan mentioned) comes directly from quantum mechanics.
(More completely, it’s electrons that carry the relevant magnetic properties, and the exchange interaction means that their spatial distributions are different depending on whether their magnetic properties are aligned with their neighbor(s) or not. This in turn affects more classical aspects of the system like the Coulomb potential energy between the electrons themselves and between the electrons and the positively charged nuclei they orbit. In ferromagnetic materials, these effects conspire to produce a more energetically favorable configuration with aligned electrons rather than anti-aligned electrons, even though the magnetic forces in isolation would always have voted for anti-aligning.)
Yes. It’s one of the fundamental forces.
because it’s bolted to the door.
Yes, but why do bolts stick things to things?
If the “bolt” gets in the way, don’t think of a bolt, just imagine that the entire door including the handle is one piece of extruded metal. So there’s this handle sticking out. Why doesn’t it fall off? Yeah, because it’s one piece of metal. But why does being one piece of metal mean that gravity doesn’t pull it apart? Why doesn’t the handle fall off your coffee cup?
The point is, the door and the cup is held in one piece–the handle is literally stuck on–by the same force that sticks the magnet on.
The reason you can stick a magnet to the fridge but can’t stick a coffee cup to the fridge the same way is that the magnetic forces in the coffee cup holding it together are all randomized in every direction. And the magnetic forces in the magnet have been carefully lined up. You could make the magnetic forces disorganized in various ways in the magnet–heat it up hot enough and those lined up atoms get randomized and you don’t have a magnet any more. Heat up the coffee cup hot enough and the electromagnetic forces holding it together stop holding it together, and the cup melts into a puddle.
Your chair fails, not because it runs out of energy, but because it runs out of chair. What happens to it? Maybe it gets eaten by termites. Maybe it rusts and the rust flakes off. The exact details will depend on what it’s made of. But it’s not a matter of energy.
So I still haven’t seen why a ferromagnetic material is ferromagnetic. And more so, why aluminum or copper bars or silicon crystals cannot become permanent magnets. The micromagnetic fields disorient themselves almost immediately instead of reinforcing. What is special about those?
This wiki article Ferromagnetism - Wikipedia offers some factors:
-
Potential magnetic oomph per atom depends on the degree of unfilled electron shells. A full outermost shell is bad, a mostly unfilled outermost shell is good.
-
Favorable or unfavorable crystalline molecular structure makes or breaks ferromagnetism. I take that to mean the lattice size of e.g. iron is just right to mutually reinforce the individual atomic magnetic domains, whereas the lattice size of Cu is just wrong.
It doesn’t go into the level of detail about why, e.g., copper is wrong and iron is right. But it’s a start.
Here you go.
The electrons of an isolated atom exist in quantum mechanical states called orbitals. Each of these orbitals has a characteristic energy and spatial distribution, and no two electrons of the same spin can be in the same orbital. When two atoms combine to form a molecule, the atomic orbitals are no longer relevant, as now the molecule as a whole has it’s own orbital structure. In this two-atom molecular case the general idea is the same: there are distinct orbitals with distinct energies, and the electrons occupy these in order from lowest energy to highest energy.
When you move to a metallic system with effectively infinite atoms, the nature of the new orbitals changes qualitatively from the discrete case. This is described in the language of electronic bands. Electrons in these energy bands are quasi-free particles that aren’t localized to a single atom and can have energies along a continuum defined by these bands. There can be multiple bands, and electrons of different spins live in different bands, at least in principle. In non-magnetic metals, the bands that differ only in the electron spin are overlapping – i.e., they have the same energies.
If you apply a magnetic field to such a metal, the energy bands for the spin-up and spin-down electrons are shifted up and down accordingly. Since the electrons in the metal will always occupy the lowest available energies states, those electrons at the very top of the “spin-up” band will spill over into the “spin-down” band. This leaves an excess of spin-down electrons over spin-up electrons. The metal now has a net magnetic field of its own. This is paramagnetism.
Necessary digression: Take the magnetic field away for a moment. The morass of overlapping bands in a metal means that each energy level available to electrons has a different capacity for holding said electrons. Since we’re thinking of energies as a continuum, we can say this with calculus. The number N(E) of available states per unit energy is called the “density of states” f(E) = dN(E)/dE. This is a fairly feature-rich function with peaks and valleys as you go from the lowest energies to the highest, with features stemming from atomic orbital filling patterns, lattice structure and spacing, and so on.
As you fill the energy states of the metal with electrons, you eventually get to the top of the energy sea. All electrons are accounted for, and the last ones you put in have the highest energy of all. This energy level is called the Fermi level, E[sub]Fermi[/sub].
Okay, now put the magnetic field back on. The highest spin-up band spilled some number of electrons into the highest spin-down band. How many got spilled depends on the density of states near the Fermi level. If the density of states is small near E[sub]Fermi[/sub], not many will have spilled over when the bands’ energies were shifted by the magnetic field. If f(E[sub]Fermi[/sub]) is very large, though, you could apply the tiniest of external magnetic fields, lift one of the bands by the tiniest of energies, and spill zillions of electrons over from the now higher band to the lower.
In a small number of metals, the Fermi level happens to fall right on a sharp spike in the density of states. In these metals, the spike is so narrow and tall that this spillage actually happens spontaneously, with the spilled-over electrons producing a regional magnetic field that acts as the “external” field. Saying this in energy terms: it costs very little energy to induce a large inbalance of spins, and the gain in energy from having spins aligned with the average magnetic field is worth it.
A possibly less confusing version of the last paragraph (edits embedded):