Hi everybody, we’re back for the next round of Stump the Scientist! This question came in from Facebook community member Forrest Holt. Don’t forget that you can visit www.facebook.com/ge or www.facebook.com/edisonsdesk to interact with us there. We look forward to reading your Stump the Scientist questions!
Question posted by Facebook user Forrest Holt:
How do magnets work?
Response from Chief Scientist Jim Bray:
The answer is “by electricity”. Unless you already know a lot of physics, that answer, while short and sweet, will probably not explain much to you. So let’s go a little deeper.
All magnetism is caused by the flow of electric current (electricity).A broad class of magnets, called “electromagnets”, illustrates this. You can readily make a magnetic field around a wire by flowing an electric current through the wire. If you bundle a lot of wire together, such as in a coil of wire (like a solenoid), you can make a stronger electromagnet, and increasing the amount of electric current also increases the magnet’s strength.
We have to go a little deeper to explain the other class of magnets, permanent magnets (like the “bar” or “horseshoe” shaped materials). In these, the electric currents circle around each atom which compose the magnet. Electricity is compose of moving electrons, and electrons circle around all atoms. Not all atoms have electrons which circle in the proper manner to make a good magnet, but the atoms of permanent magnets (like iron) do. Each atom becomes a very small magnet, but when the very many atoms which make up the permanent magnet line up and cooperate, we get the familiar permanent magnet (which we might use to attach things to the refrigerator door). If the atoms do not line up, then we would say that the magnetic material is demagnetized.
UPDATE:
We received the following comment from one of our readers, Sasikanth Manipatruni, regarding Dr. Jim Bray’s response:
I do want to get some comment from you on permanent magnets. In reality, the Fe/Co/Ni permanent magnets show little orbital magnetic moment. And the contribution is mainly from the spin magnetic moment (one Bohr magenton per uncoupled spin). In-fact, when we do Landau-Lifshitz-Gilbert dynamics of a magnet, the magnetic moment of a single domain magnet is simply the number of unpaired electrons in the magnet multiplied by the value of Bohr magneton. Further confirmation of this is the spin torque effect when electrons with opposite spin injected into a magnet turn it.
So is it still fair to say magnetism is due to electricity ? The magnetic moment of the electron itself cannot be derived as a result of moving electrons forming an Ampere field. What I thought was that if we assumed a spinning electron, we get absurd values of the speed/angular velocity of electron.
Jim requested that we post the following clarification response:
Thank you for the excellent follow-up question. In answering these questions, I am often faced with the dilemma of how deep to go into the physics and still have an answer that is generally correct but understandable by the general public who are not trained in physics. You are quite correct in your observations. My simple answer of “by electricity” is still correct, since the property you point out belongs to the electron, the quantum unit of (most) electricity.
The electronic magnetism in permanent magnets is indeed (as you say), caused by combining both the orbital and spin components of the electron, in varying ways depending on the precise material. Furthermore (as you note), the electron spin magnetic moment cannot be calculated classically by any actual classical spin of the charge cloud of the electron; it is a fully quantum mechanical property. Attempts to do it classically fail by approximately a factor 2. The excellent fully quantum mechanical theory of electromagnetism, called QED (quantum electrodynamics), does allow us to calculate this spin magnetic moment as accurately as we wish. In my original answer, I lumped all these properties into “motion” of the electron around an atom without going into the quantum theory.
Of course, it is possible to go even deeper into the physics of magnetism; e.g., paramagnetism, diamagnetism, antiferromagnetism. The nucleus also has a magnetic moment, and it is small enough compared to the electron that we can neglect it in discussions of most magnets, but it is vital for important fields like MRI and NMR.
Magnetism is a very broad and deep topic in physics. Your follow-up points out some of this additional depth.
Thanks, as always for everyone’s great interest and questions!