Before 1911, the accepted picture of an atom (J.J. Thomson's "plum pudding" model) treated positive charge as spread out evenly through the whole atom, like plums scattered through a pudding. Ernest Rutherford's gold foil experiment didn't just refine that picture — it demolished it completely, and the specific way it did so is a genuinely instructive example of how a single unexpected result can overturn an entire accepted model.

Alpha particles mostly passing through, one sharply deflected

The setup: firing alpha particles through gold foil

Rutherford's team (Hans Geiger and Ernest Marsden did the actual detection work) fired a beam of alpha particles — positively charged, relatively heavy particles — at an extremely thin sheet of gold foil, then tracked where the particles ended up on a surrounding fluorescent screen. Under the plum-pudding model's evenly-spread-out positive charge, every alpha particle should experience only weak, gradually-accumulated deflections as it passed through — nothing capable of producing a sharp, large-angle bounce.

The result that broke the old model

Most alpha particles did pass through with little to no deflection, consistent with atoms being mostly empty space — but a small fraction, roughly 1 in 8000, bounced back at angles greater than 90°, some nearly straight back the way they came. Rutherford's own famous reaction: it was "as if you had fired a 15-inch shell at a piece of tissue paper and it came back and hit you." A gradually-distributed charge simply cannot produce this kind of sharp, occasional, large-angle deflection — only a concentrated, intense charge could.

Why this specific result demanded a concentrated nucleus

The overwhelming majority of alpha particles passing straight through, undeflected, told Rutherford that atoms are mostly empty space. The rare sharp deflections told him that occasionally, an alpha particle passes extremely close to something small, dense, and intensely charged — concentrated enough to repel a fast-moving alpha particle back the way it came, purely through electrostatic repulsion, with no other force needed. This combination — mostly empty, but with an occasional tiny, dense, charged core — is exactly the nuclear model of the atom, proposed for the first time as a direct, forced consequence of this one experimental result rather than as a hypothesis tested afterward.

Worked check: for a head-on collision, a 5.5 MeV alpha particle approaching a gold nucleus (Z=79) gets deflected straight back at a closest-approach distance of about 41 femtometres — this specific calculated number, matching Rutherford's own experimental estimate closely, is exactly what confirmed the nucleus's genuinely tiny size relative to the atom as a whole.

Why gold foil specifically, and why so thin

Gold was chosen partly for a practical reason often left out of textbook summaries: gold is remarkably malleable, meaning it could reliably be hammered into foil only a few hundred atoms thick — thin enough that most alpha particles interact with at most one nucleus on their way through, keeping the scattering pattern interpretable as single, individual scattering events rather than a hopelessly tangled combination of multiple successive deflections. A thicker or less uniform foil would have muddied the clean statistical pattern Rutherford's team needed to draw their conclusion with any confidence.

What the experiment could NOT tell them

Rutherford's scattering results pinned down the nucleus's existence, its concentrated positive charge, and roughly its size — but said nothing about how electrons were arranged around it, or why those orbiting electrons don't simply spiral into the nucleus the way accelerating charges are expected to radiate away energy and collapse, according to classical electromagnetism. That specific unsolved puzzle — a real, glaring gap left open by Rutherford's own nuclear model — is precisely the problem Niels Bohr set out to fix just two years later with his quantized-orbit model of the atom.

Try it yourself

Rutherford Scattering Bohr Model Davisson-Germer Experiment