Behind the scenes My (failed) Stern-Gerlach apparatus

When I chose the topic of this thesis, back in 1999, I was neither an experimental nor a theoretical physicist (to be honest, I am not one even today). At the time I had not even studied Feynman’s celebrated Volume III yet.

But I was armed with colossal naivety: I was genuinely convinced I could physically build a Stern-Gerlach apparatus on my budget, to bring the data to my graduation defence.

I almost immediately hit a wall made of thermodynamics and engineering. Producing a beam of silver atoms requires a furnace at 1000 °C. Making them travel in a straight line demands a high vacuum that is extremely hard to maintain. Aligning them blindly in an asymmetric magnetic field is a nightmare. In short, building it with my own means turned out to be impossible.

Because of this practical “failure”, for 27 years I never put this thesis online. It felt like incomplete work, with the very first experiment never carried out: for me it stayed a thought experiment.

Why did I finally decide to share it online in 2026?

Because, almost thirty years after writing it, I realised that the impossibility of reproducing it in a garage is exactly what makes the original experiment legendary.

In 1922, at the University of Frankfurt, Otto Stern and Walther Gerlach were working in desperate conditions. Germany was in the grip of hyperinflation, there was no money for electricity, and the apparatus melted down every other day. They were literally saved at the last minute by a dollar cheque sent by Henry Goldman — one of the founders of the famous investment bank Goldman Sachs — alerted by the physicists Max Born and Albert Einstein.

But the biggest obstacle was at the end of the tube. Modern electronic detectors did not exist. To find out whether the magnetic field had split the beam of atoms (proving quantum mechanics) or merely scattered it (as classical physics predicted), they used the crudest possible method: they placed an ordinary cold glass slide as a target, hoping the silver would condense on it and form a visible pattern.

After hours of effort, they pulled out the slide. It looked completely blank. The silver deposit was there, but only a few atoms thick, utterly transparent to the naked eye. Convinced they had failed, they dismantled the apparatus.

This is where the cigar comes in. Otto Stern was a heavy smoker of cheap cigars, notorious for being loaded with sulphur. Disappointed, he picked up the slide and brought it close to his face to examine it against the light, breathing on it with the cigar in his mouth.

Without knowing it, he was triggering the same chemical process that blackens old silverware: the sulphur in the smoke reacted instantly with the invisible silver on the glass, turning it into silver sulphide, which is black and opaque.

Under the incredulous gaze of the two physicists, two sharp separate black lines appeared out of nowhere on the slide. The beam had split. The quantisation of spin had just been demonstrated thanks to the breath of a scientist with bad taste in cigars.

I did not have the funds for a high vacuum, but at least I was spared the sulphur-flavoured cigars.

Conceptual, but not merely theoretical

In this thesis the experiment stays on paper — and the one in the next chapter, Stern-Gerlach analysers in series, is even harder to build. But make no mistake: these are not mere thought experiments. Their strength is first of all logical — they dismantle classical intuition piece by piece — yet they are also physically realisable. Not in my garage, of course: you need a well-equipped laboratory and a budget of a few million. But it has actually been done: around 2019 Ron Folman’s group at Ben-Gurion University built, on an atom chip, a full-loop Stern-Gerlach interferometer, splitting and then recombining the beams. In short: what you are about to read is not pure abstraction, but the description of something you can genuinely touch — if you have the means.

One last thing, if you feel like digging deeper: in Chapter 4 the Schrödinger equation is not handed down as a postulate, but obtained from simpler principles. It is the most original part of this work and it is also independent of the rest: it can be read on its own, without having read the other chapters. It is “a little bit” dense, I admit, but in my view it is worth the read all the same: it reveals foundations that — starting from the standard formulations of quantum mechanics, where the equation is simply postulated — remain obscure. Enjoy the read.