Currently, down in the depths of C-building, there’s a master’s student trying to carry out the Stern-Gerlach experiment. (and also here). This is one of the classic experiments in quantum mechanics – specifically demonstrating the quantisation of angular momentum.
If you look at the text books, it’s simple enough. Pass a beam of atoms through an inhomogeneous magnetic field (i.e. one that is stronger in one region of space than another) and, hey-presto, the beam gets split into two (or more) beams, depending on the magnetic moment of the atoms. The non-uniformity of the field is essential. If the atom has a magnetic moment that lines up with the field, then it will have a negative potential energy due to the field and will move towards the region of strong field, where its energy is most negative. Conversely, if it has a magnetic moment that is against the field, it will have a positive potential energy and will move towards the region of weak field, where its energy is least positive. So the beam splits. The key result, though, is that the beam doesn’t split into a continuum, which would mean any magnetic moment were possible, it splits into discrete beams, showing that only certain values of magnetic moment are allowed. This is what quantization means – things are split into discrete amounts. What the experiment is doing, is measuring the magnetic moment in a particular direction.
Stern and Gerlach did this experiment in 1922. Having seen our poor student struggle with the apparatus, they must have put in some considerable effort, that’s now been glossed over in most books. There are all kinds of issues that need attending to. Preparing a beam of atoms (in our case sodium – we’d like to use potassium but that’s a little bit too exciting from a safety point of view) is tricky. The sample needs to be heated so that atoms are evaporated. We need a high vacuum, meaning that atoms do not collide with air molecules on their way down the apparatus. We need to make sure that we are detecting our sodium atoms not contaminant atoms that are coming from elsewhere. And, most frustratingly this afternoon for student, we need to find where the beam is going an align it so that it falls on the detector.
The stereotypical drawing of the apparatus we see in the quantum textbooks overlooks most of what actually has to go on to get this to work. It’s slow going, tedious, and frustrating, but hopefully the student will nail it in the end. This is all too reminiscent of the reasons why I became a theoretical physicist rather than an experimental one, and the old adage…"Biology experiments wriggle, Chemistry experiments smell, and Physics experiments don’t work"