065
Stern-Gerlach
Silver atoms pass through an uneven magnetic field and land in two places only. A smear was expected; nothing lands between.
Field
Detail
Study
065
Chapter
Quantum basics
Equation
μ_z = ±μ_B
What moves
the field gradient ∂B/∂z; the angle θ of the second magnet
Structure
Film
Colour means
blue, atoms deflected up; violet, atoms deflected down; grey, the classical expectation
Posted

Two marks on a screen, and nothing between them. Silver atoms have flown through a magnet, and each one landed in one of two places. This is the Stern-Gerlach experiment, first done in 1922. Each silver atom behaves like a tiny magnet, and almost all of its magnetism comes from a single electron.
Turn the field gradient down to zero and the two marks close into one spot. The gradient is how quickly the field’s strength changes from top to bottom. A field that is equally strong everywhere only turns each atom; it does not push it up or down. A push needs the field to be stronger on one side than the other.
The camera rides along the beam through the gap while the gradient rises. The field is stronger near the pointed upper pole than near the grooved lower one, so each atom is pushed up or down by its magnetic moment: F_z = μ_z ∂B_z/∂z. The magnetic moment is the strength and direction of the atom’s tiny magnet, and its up and down part, μ_z, sets the push. One beam becomes two.
If each atom's moment could point in any direction, the hits would fill the screen as one band from top to bottom, the grey of the classical expectation. Instead, they gather only at its two ends, μ_z = ±μ_B, with nothing between. μ_B, the Bohr magneton, is the natural unit of an electron’s magnetism. Only two values are ever found because the electron’s spin is quantised.
A knife edge, a blade that blocks the lower beam, keeps the upper beam and sends it into a second magnet, turned by θ. At 0° all 200 atoms go up again. At 45°, 169 go up and 31 go down. At 90°, the split is 101 and 99: the first answer says nothing about the second. The chance of up is cos²(θ/2), and every answer is still one of two.
Each measurement gives one of two results. For a single atom, which one cannot be predicted; only the proportion over many can. The screen is drawn 3000 times larger than life, and the hits are seeded draws from the theory, not laboratory data.