Why don’t magnets stick to aluminum? –Les, Los Angeles

It all has to do with electron shells. In a newspaper of general circulation, however, it is always risky to jump straight into a discussion of electron shells. Better we should sort of edge into this.

First some facts. Fact number one: magnets only stick to other magnets. Fact number two: big magnets are made up of jillions of tiny magnets. Fact number three: so are the metals the magnets stick to, notably iron, nickel, and cobalt, which are called ferromagnetic materials. The difference is that in the big magnets the tiny magnets are organized–that is, they’re all lined up with their north poles in one direction and their south poles in the opposite direction–while in ferromagnetic materials, the tiny magnets are scattered every which way, and their magnetic fields cancel each other out.

But suppose we enterprisingly place a ferromagnetic material in a strong magnetic field. Voila, the formerly scrambled atoms line up parallel with one another. The material as a whole becomes magnetized and sticks to the magnet. Aluminum doesn’t contain tiny magnets, so there’s nothing to get organized and nothing for the big magnet to stick to.

Certain restless intellects out there may now be wondering: what’s with this tiny magnet crap, anyway? That’s where the electron shells come in. As you may have guessed by now, the tiny magnets we’re talking about are individual atoms. Some atoms, such as those in iron, have individual magnetic fields, while others, such as those in aluminum, do not. It all has to do with the electrons.

Electrons may be thought of as spinning, much as the earth does. They spin one way, they develop a magnetic field with north on top and south on the bottom; they spin the opposite way, they develop a magnetic field with north on the bottom and south on top. For convenience, we call the two directions of spin positive and negative.

Most atoms, such as those in aluminum, have half their electrons spinning in one direction and half in the opposite direction. That means the magnetic fields of the individual electrons cancel each other out. But in ferromagnetic materials things are different. Take a gander at the third subshell of the M shell of iron, for example. (A shell is an electron’s orbit. Electrons are rigidly organized into layers of shells, with so many electrons per shell.) What a wacky sight! We find five electrons with a positive spin and one with a negative spin. This gives the iron atom a pronounced magnetic field. You get those iron atoms lined up, by cracky you’ve got yourself a magnet. There’s more to it than that, but that should be enough to see you through next time the question comes up at the corner saloon.


In his recent treatise on whether singers can break glasses with their voices (May 11), Cecil mentioned “forced oscillation resonance,” in which an external force amplifies the natural vibration of an object, sometimes with destructive results. As an example he cited the 1940 collapse of the Tacoma Narrows bridge. The usual explanation for this disaster is that the wind gusted (“generated a train of vortices,” to be precise) in perfect synch with the bridge’s natural rate of bounce, causing its demise.

Reader Wilbur Pan has alerted us to a recent report in Science News heaping abuse on this widely held view. Mathematicians Joseph McKenna and Alan Lazer doubt that a storm could produce the perfectly timed winds required. They’re working on a “nonlinear” model of bridge behavior they hope will provide a better explanation. The main problem apparently is that when the roadway of a lightly constructed suspension bridge flexes, the cables supporting it go slack, introducing an element of unpredictability in which little causes (i.e., the wind) produce big results (i.e., a collapsing bridge). They hope to have the mathematical model describing this effect finished in five years. (This is obviously government work.) You’ll read about it here first.

Art accompanying story in printed newspaper (not available in this archive): illustration/Slug Signorino.