Talk:Stern–Gerlach experiment

external link
There is a nice explanation in this page: http://rugth30.phys.rug.nl/quantummechanics/stern.htm

but I do not have time to summarize/rewrite. -- ato 21:35, 1 Sep 2004 (UTC)

Description Rewritten
Hey what do you think about the rewritten basic theory/description section? BeardedPhysicist 23:58, 14 May 2006 (UTC)


 * pretty well written overall. nice. Mct mht 06:18, 17 June 2006 (UTC)

History?
Strange that there are no links to articles about Stern or Gerlach or even anything that says when this experiment was done! Michael Hardy 18:19, 9 Sep 2004 (UTC)

OK, now I've added some of that information. Various web pages say it was done in 1920, and others say 1922. Could it be that 1920 was the date of execution and 1922 was the date of publication? Just a guess. Michael Hardy 18:31, 9 Sep 2004 (UTC)


 * I have added a section "Impact" where we can discuss all the related stuff that came AFTER that experiment. this will be usefull to include the links to more modern experiments or physics.

Alain Michaud 01:32, 23 December 2005 (UTC)

Original text was copied from external page
The original edit, as of 6 September, 2004, seems to have been copied from. I've left a note on the user's user page; I'm going to rewrite the article. -- Creidieki 02:44, 12 Sep 2004 (UTC)

Details
none Brianjd

Please excuse me while I dump these here for now








Question related to entanglement and this experiment
Imagine that you have this Stern-Gerlach apparatus but also another one on the right side (mirrored). Now if we send two entangled particles, one on the left and one on the right they will still be deflected up or down for some discrete value. Right? Now imagine that we changed spin of one particle (on the left side) while it was going through inhomogeneous magnetic field. That change will be transferred to the right particle instantaneously. That means that at the end particle will not be deflected at the some of previous 2 discrete values because we changed the spin while it was going through magnetic field. In that way if we want to transfer classical information from the left side to the right we can change states of the particles going through magnetic field on the left and on the right side we will se that particles are not deflected in some of 2 discrete values. In that way we can transfer classical information (change of the states of the left particles) to the right side faster than light. Can anyone explain what is wrong with this?

P.S. Feel free to change my question to meet standards of wikipedia. This is my first entry so I don't really now the rules. Thank you.

many details in the question posed are not necessary, e.g. one doesn't really need the G-S experiment and the question can be phrased in more general and simpler terms. following is the reply.

what's stated is wrong because the way you are proposing to send classical information is impossible. so one party, say A, locally manipulates the system, and through entanglement, causes state change in B. to send classical information means precisely that B performs a measurement, and, if this is going to work, must be able to recover the classical information according to measurement outcome. one can show that no matter how A interacts with the system locally, the measurement statistics on B's subsystem remains the same. this is exactly the statement of no communication theorem. similarly, quantum information can not be transmitted faster than light either. see also quantum teleportation. Mct mht 05:45, 17 June 2006 (UTC)


 * I don't know who is still interested in this question, but David Mermin's Strong Baseball Principle partly explains it, along with his other discussions on QM and entanglement. Gah4 (talk) 17:07, 23 March 2020 (UTC)

The second sentence of your original question is incorrectly written. You would have to run the experiment as it should have been originally: using a cathode ray instead of silver atoms. Silver atoms are not particles, they are composed of groups of particles (protons and neutrons in the nucleus and orbital electrons). I believe another section here deals with this. So the hurdle in designing such an experiment would be to determine an entanglement mode for silver atoms. This may be difficult using the apparatus as given because heat is being added to a sample. The kinetic energy of this heat will probably cause a disentanglement before the atoms exit the crucible. One hurdle of entanglement is getting a correlation between two objects. That is a reason why photons are currently being used to explore entanglement and the Bell hypothesis. Photons can easily be entangled by using a downconverter. Even with this form of generation, there can be some difficulty in determining the exact mechanism leading to entangled pairs. I edited your text for spelling only. The other issue with the original Stern-Gerlach setup as given in the article is the setup is not described very well. For such an important experiment that others will rely on for further research, more specifics should be given, such as crucible temperature and sample size. The field strengths for each of the trial runs should also be given. Another thing I have noticed is the original and subsequent images of the detecting plates have been removed from the article. There are now no plate images for any of the variant experiments listed.BRealAlways (talk) 08:56, 9 September 2021 (UTC)
 * Complicated question. Mostly it doesn't matter that silver atoms are groups of particles. As long as you don't excite them too much, out of the ground state, they act like particles. There might not be a process to create pairs of entangled silver atoms, though, so I suspect it doesn't matter. Gah4 (talk) 07:29, 10 September 2021 (UTC)
 * Complicated question. Mostly it doesn't matter that silver atoms are groups of particles. As long as you don't excite them too much, out of the ground state, they act like particles. There might not be a process to create pairs of entangled silver atoms, though, so I suspect it doesn't matter. Gah4 (talk) 07:29, 10 September 2021 (UTC)

Entanglement requires some form of correlation between the specimens. Vaporizing Ag atoms will surely energize them. A gas has different properties than a liquid. Generating gaseous Ag requires adding heat energy according to the Stern-Gerlach setup. The point is that if you want to model quantum properties and test for them, you need to use the particle as a particle, not as a composite structure. It is possible to design an entangled form of this experiment, but it requires thinking it through. My point is the foundation of science: repeatability. In order to get to repeatability, you need a full understanding of the mechanisms involved. If we hold to the Einstein paradigm that no information can travel faster than light, then there is a tradeoff in quantum correlation (entanglement symmetry). If correlation is given priority, then faster than light travel is meaningless. The question I have is: Why is the Stern-Gerlach experiment not described specifically? What I see is broad generalities that couldn't possibly lead to repeatability. If quantum mechanics is as reliable as is purported, there should be no problem with giving a more full report of the experimental setup and results as run. I see the heavy reliance on modeling, not on the experiment as run. BRealAlways (talk) 22:49, 20 September 2021 (UTC)

Spin numbers
For greater clarity I think the quantum number S and ms should be differentiated between. i.e. that electrons are fermions, and therefore posses a spin number of +1/2, which is a property intrinsic to the electron, and that the electron itself can be 'orientated' to give a projection spin quantity of +/- 1/2, which is a property of a situation or specific electron in an atom. In the experiment, all electrons, by their nature, have a spin number of 1/2, but have angular momentum projection number of +/- half when detected in a system, hence there being two possible (degenerate) quantum states for ground state silver- and the two lines.

Sorry if this doesn't make any sense! This article was recently slated by a university professor, but I feel it's content isn't wrong, just in-explicit.

Alexanderhowell 15:06, 6 March 2007 (UTC)

Purpose of the experiment
From the first paragraph of "Basic theory and description":

"The experiment sought to determine whether individual particles like electrons have any "spin" angular momentum."

Surely this statement is incorrect, as according to the Physics Today source the experiment was originally conducted using silver atoms to test Bohr's theory of the atom (in particular "space quantisation", although I'm not sure what exactly this is supposed to refer to). Electron spin wasn't known about at the time and the experiment was only interpreted in terms of electron spin in 1927.

Pagw 14:46, 28 August 2007 (UTC)


 * I think you are almost certainly right as to the purpose of the original experiment not being about electron spin. Rather than rewrite the sentence to talk about the original purpose, I'd suggest rewriting it to talk about the application to electron spin, which is the most talked about use for the apparatus. Teutanic 15:44, 31 October 2007 (UTC)

>(in particular "space quantisation", although I'm not sure what exactly this is supposed to refer to)

"Thus, depending on the orientation of the magnetic moment relative to the magnetic field there will be either an attractive or repulsive force and the beam will split into two components, exhibiting spatial quantization. ... If the beam is spatially quantized, as Sommerfeld predicted, two spots should be observed on the screen." http://plato.stanford.edu/entries/physics-experiment/app5.html

Is this helpful?

--Beroal (talk) 13:30, 29 December 2007 (UTC)


 * Yes, thanks. Pagw (talk) 13:55, 31 December 2007 (UTC)

I was under the impression that this disputed sentence implied that they were investigating whether there was spin angular momentum in the sense of a rotating object -- not quantum spin which, as mentioned above, was not interpreted until several years later. It would be interesting to know whether any of this was true... I suppose tracking down the original paper describing the experiment might be a good idea. --Qrystal (talk) 16:35, 28 January 2008 (UTC)

not enough information
What will be if one of those magnets in will be removed? Does there will be the same two levels quantization or only one level will remain?

More precisly, what will be if will be removed one of two magnets in Stern-Gerlach Experiment for electron spin? http://hyperphysics.phy-astr.gsu.edu/Hbase/spin.html Does spin then still be quantized in two orentations or only in one? —Preceding unsigned comment added by Fakeup (talk • contribs) 16:06, 21 July 2008 (UTC)

What if one of the magnets were removed(example the N) would the beam still split? What if the beam started lower, if the beam, was closer the the S, would more mass go to the S, would there be a point, where it would not split anymore? How many people repeated this experiment? Answer: In 1925 most people did not even believe there were electrons. There is no spin, no electrons. It is a radiation pressure. Seccond, there is no such thing as having an N one place and an S in anther place. they are apart of the same magnet. How the magnet is connected will affact the experiment, so it needs to be displayed how that is. It is possible the are from two different magnets, in which case there are also magnetic fields around the experiment. Seccond in the regionN S, there is a tension, like a compressed spring. When the radiation leaves the mass, it is effected by the tension. This experiment proves electrons dont have a negative charge. They dont exist. Radiation, has a direct effect on magnetism, and this experiment, needs to be redone, and explored. —Preceding unsigned comment added by 202.89.32.166 (talk) 05:33, 11 October 2008 (UTC)

It's probably better to think about the N and S magnets as iron ends connected to a electromagnet. What is important is one of the poles needs a knife edge, and the other needs a concave troff in it, so the field strength is intensified through one pole, and diffused through the other. Something similar to a E body transformer, with one side wound clockwise and the other counterclockwise, could connect the two iron poles and minimise the rest of the magnetic field. Maybe a couple of layers of Mu metal as a shield. Stern and Gerlach probably did the experiment with a big horseshoe magnet like a lot of Physics departments have for demos, and they probably had the university's machine shop make up the 2 ends of iron or steel. Smaller is better, this is an experiment you could do under a bell jar, vacuum just high enough to silver a mirror should be good enough.

Silver Duality
Did anybody ever collect the two kinds of segregated silver? Based on the results of this experiment, there are two distinct kinds of silver, one of which is diamagnetic and the other which is paramagnetic. Normal silver is obviously a mixture of the two which cancels the magnet effect.

Has anybody ever actually produced the individual polarities of silver in bulk such as to make paramagnetic silver objects? David Tombe (talk) 11:21, 29 December 2008 (UTC)


 * There are precise stern-gerlach experiments done with particle detectors that show the resulting split to be sinusoidal. There are also relatively current Japanese experiments that show the particles go up and down sinusoidally within the magnetic field.  Since results don't agree with current theory people forget them.  I'll post links when I see them again.
 * Daniel.Cardenas (talk) 22:28, 29 December 2008 (UTC)


 * Daniel, Thanks for the reply. I'd certainly like to know more about this experiment. There seem to be conflicting accounts in the literature as regards its applicability to electrons. But as regards silver, it seems strange that the two polarities have never been isolated.
 * If the particles go up and down sinusoidally, that is another mystery. Why would silver atoms move in a sinusoidal path in a magnetic field?
 * David Tombe (talk) 02:54, 31 December 2008 (UTC)


 * I'll post more links as I find them but here is something, sorry if it seems weak: http://www.physics.brocku.ca/People/Faculty/Sternin/teaching/mirrors/qm/stern-gerlach/index.html
 * >Why would silver atoms move in a sinusoidal path in a magnetic field?
 * Because of wave particle duality. Wave has magnetic field component that oscillates up and down.
 * Daniel.Cardenas (talk) 20:06, 31 December 2008 (UTC)


 * I assume the atoms are switching between being paramagnetic and diamagnetic. thus, to answer your first question, it would be impossible to collect one or the other in bulk.
 * just-emery (talk) 14:13, 11 May 2009 (UTC)


 * Here is a page that talks about a spin filter. Don't know if you will find it interesting. http://www.upscale.utoronto.ca/GeneralInterest/Harrison/SternGerlach/SternGerlach.html
 * Daniel.Cardenas (talk) 20:24, 31 December 2008 (UTC)


 * Daniel, Thanks for the links. The Stern-Gerlach experiment is said to produce two distinct patches. One above the centre mark and one below. If the silver atoms are moving in a sinusoidal motion, that would not be the expected result.
 * David Tombe (talk) 09:17, 3 January 2009 (UTC)
 * Yes, and if you find accurate experiments they conclude that most of the distribution is at the top and bottom but there are particle hits all over matching a sinusoidal distribution. I think everyone can agree that the way the historical experiment was done and repeated over the years would not know if the distribution is sinusoidal.  Sorry for repeating myself, just want to make sure it is clear.   Thx,
 * Daniel.Cardenas (talk) 16:35, 3 January 2009 (UTC)


 * Daniel, no matter what the truth is, I still find it hard to come to an explanation. Under the conventional reporting, I would tend to suggest polarization along paramagnetic/diamagnetic lines. That is why I was asking if anybody has ever collected the two kinds and constructed a paramagnetic silver object.


 * But as regards your suggestion that there is actually a sinusoidal motion, would we not then expect more of an even ditribution between the two limits? Why would a sinusoidal motion permit a concentration of hits at the two limits?
 * David Tombe (talk) 05:01, 4 January 2009 (UTC)
 * because that is where each atoms spends most of its time.
 * just-emery (talk) 14:13, 11 May 2009 (UTC)


 * [[Image:ComplexSinInATimeAxe.gif]]
 * If you plastered the dots against a plate where would most of the dots be? Towards the top and bottom.
 * Daniel.Cardenas (talk) 14:40, 11 May 2009 (UTC)

Goudsmith/Uhlenbeck
This sentence doesn't make sense:
 * "Even though the result of the Stern-Gerlach experiment has later turned out to be in agreement with the predictions for a spin ½ particle should the experiment be seen as a corroboration of the Bohr-Sommerfeld theory."

Perhaps it has become mangled during editing. Could somebody please clarify the wording? Thanks.&mdash;RJH (talk) 16:34, 26 April 2009 (UTC)
 * Perhaps this may help?&mdash;RJH (talk)


 * The Stern-Gerlach experiment has been performed before the quantum mechanical notion of spin was established. It tried to test the Bohr-Sommerfeld theory which did not contain that notion.WMdeMuynck (talk) 22:15, 26 April 2009 (UTC)


 * I'm aware of that. Thanks to whomever corrected the sentence in the article.&mdash;RJH (talk) 20:28, 28 April 2009 (UTC)

If the element in experiment is atom, you are measuring properties of atom, not a single electron !
The claim "measurement of the spin of electron" is valid in the sense, that electron is spining around the nucleus, not in the sense that the implicit spin of the electron was measured. Deflected particles are atoms, not electrons. The cause of deflection of atom is orbital direction of the electron motion interacting with magnetic field, or better, interaction of the overall magnetic moment of the atom, caused by the atom as whole.


 * Apart from any theory, it can be stated, as a pure result of the experiment, and as far as the exactitude of our experiments allows us to say so, that silver atoms in a magnetic field have only two discrete values of the component of the magnetic moment in the direction of the field strength; both have the same absolute value with each half of the atoms having a positive and a negative sign respectively (Gerlach and Stern 1924, pp. 690-691, FW)

If the atom is unique discrete system, there are only two configurations of the atom preserving its discrete properties - preserving its identity. These configurations difer in direction of rotations.

Interpreting the Stern-Gerlach experiment as implicit spin of the electron is the same as saying : the outer electron (with lowest binding potential) is draging the heavy atom of the silver, instead of being teared out from the atom. If so, than the force of draging must be less than the force that will tear out the electron ... Softvision (talk) 18:37, 11 July 2009 (UTC)


 * As I understand it, it is an s electron, so no orbital angular momentum. I presume you can also S-G on ones that do have orbital angular momentum. Gah4 (talk) 21:04, 10 May 2018 (UTC)


 * I just realized that I didn't finish this. Yes, the force is small enough not to tear the electron out. Gah4 (talk) 22:41, 14 April 2019 (UTC)

Error: classical electron radius and calculation
From the article "Even if the electron radius were as large as 14 nm (classical electron radius) then it would have to be rotating at 2.3×1011 m/s. The speed of rotation would be in excess of the speed of light, 2.998×108 m/s, and is thus impossible.[2] " - This is in error, both the value and calculation. The classical electron radius as per wiki is 14 femtometers. The calculation is way out.. or am I mistaken? —Preceding unsigned comment added by 75.152.156.5 (talk) 02:03, 24 March 2010 (UTC)

Hypothesis
Suppose you had a stream of random direction magnetically polarized particles that were drifting latterly along and then entered a length of a magnetic field that had a definite orientation of the polarization direction. This encounter would result an interaction of the magnetic lines of force between the two fields such that an aligning torque would occur to both the magnetic field system and the individual moving particles. And this would result in the rotation of the orientation of the direction of the traveling particle's magnetic field so that it could continua to move in its lateral direction. And that might cause the orientation of the particle's magnetic field to change into 2 categories of orientation relative to the fixed magnetic field. Say that if the particle's magnetic field were reoriented to be perpendicular to the fixed magnetic field, then it could get through with the particle's magnetic field being oriented in either direction. But now we have the interaction of 2 perpendicular magnetic fields, with one of them having 2 directions of polarity. And that might result in a deviation in the path of the traveling particles such that they arrive at different spot locations at the end of the travel path of the moving particles.WFPM (talk) 16:09, 2 May 2010 (UTC)

Yes, I also have the same confusion. I think the torque may lead the spin orientation parallel to the magnetic field. In other words, the free electron spin orientation is not decided, which is random before its entering the magnetic field. Siwei Pi Luo (talk) 04:11, 23 October 2013 (UTC)

Magnetic dipoles orient to and seek the greatest spatial concentration of applied magnetic energy.BRealAlways (talk) 05:56, 10 September 2021 (UTC)

Description Inaccurate
I believe the description of this experiment is inaccurate. The article states that the experiment shows that " particles possess an intrinsic angular momentum...that takes only certain quantized values." As noted by other comments and even later in the article, however, spin quantization was known to the experimenters and others at the time of the experiment. Moreover, quantization of intrinsic spin *does not* entirely account for the results of the experiment. The greater significance of this experiment is that it showed the *projection* of the angular momentum vector (and, therefore, the magnetic moment of the particle) onto the coordinate axis parallel to the direction of the applied B-field is quantized and has only two values with equal magnitudes but opposite signs. This is a significant difference from the current text. This phenomenon has been called "amazing" even within the realm of quantum mechanics and is often labelled "quantization of space."

References:

Modern Physics; Kenneth Krane, 1983, pp. 191-192.

The Encyclopaedia of Physics 2nd Edition; Lerner/Trigg, 1991, p. 1162.

"Angular Momentum Quantization: Physical Manifestations and Chemical Consequences"; Michael Fowler; http://galileo.phys.virginia.edu/classes/252/Angular_Momentum/Angular_Momentum.html [This is already a reference to this article.]

Porcelain mouse (talk) 18:30, 5 November 2010 (UTC)

The workaround for that was removing photos of the actual plates of the original and subsequent experiments. They clearly show the point you're making.BRealAlways (talk) 06:02, 10 September 2021 (UTC)

Photons have three different projections??
"as well as photons, W and Z bosons and gluons are spin +1 particles and have three possible values for spin angular momentum." As far as I know we need to leave the photons out in this sentence as they are (which makes them different from the massive bosons) no rest-mass and therefor the projection of the spin angular momentum (the sentence talks about the spin angular momentum itself which is still always +1, or at least the quantum number is +1, the absolute value is sth like sqrt(s(s+1)) right?) So anyways, the projection (depending what defines your quantization axis) might for a photon not have a possible value of 0, namely if the momentum p defines for quantization axis (see helicity for a discussion on that). For Pi light (Zeeman effect) there might be a projection of spin angular momentum m_s=0 as the b-field is the quantization axis defining your atomic level and p can be orthorganal to B.

To wrap this up: 1) "have three possible values for spin angular momentum" is not precise, it needs to be the projection 2) photons ... have three possible values for spin angular momentum is as a general statement WRONG, as explained in above example! —Preceding unsigned comment added by 128.100.93.74 (talk) 19:19, 26 January 2011 (UTC)
 * Thanks. I nuked the rubbish in late 2012 but did not report here. Incnis Mrsi (talk) 13:09, 28 June 2013 (UTC)


 * Photons are spin 1. Spin 1 particles in general can have -1, 0, or +1, but photons can only have -1 or +1. Gah4 (talk) 21:53, 28 February 2018 (UTC)


 * By the way, much of the physics of SG is the same physics as photons and calcite crystals. Calcite will separate photons by polarization into two discrete beams. You can use two calcite crystals rotated at different angles, or even more, and do somewhat similar experiments, except that the two polarizations are 90 degrees apart, instead of 180 degrees for spin 1/2. And calcite crystals are much cheaper. Gah4 (talk) 07:42, 26 June 2019 (UTC)

Basic theory and description
Aside of the scandal above, the section makes redundant references to Δ-baryons and quarks. There were no such words in the time of Stern and Gerlach. Links spin (physics) and spin-½ will be sufficient, and references to all irrelevant particles shall be removed. Comments? Incnis Mrsi (talk) 13:09, 28 June 2013 (UTC)

inhomogeneous magnetic field
How do you create an inhomogeneous magnetic field? The strength of any magnetic field will vary according to your position within it. Seems to me the arrangement of this field is crucial to this experiment. Or are we talking about simply sending the beam closer to one pole than the other? Maybe we don't even the need the second pole near the beam. Pergelator. — Preceding unsigned comment added by 50.43.12.61 (talk) 05:19, 8 August 2013 (UTC)

My problem is that this term is nowhere defined in the article. We get a link to 'homogeneous and heterogenous' but no explanation of what inhomogeneous is. Presumably it's some special kind of homogeneous, but what? More explanation is needed here or a link to where such can be found. 81.178.174.82 (talk) 15:40, 6 September 2015 (UTC)


 * If the magnetic field is homogeneous (i.e. the same everywhere), then the resultant force exerted on the atoms/electrons is nil. They don't move, and you simply get a dot on the screen, as if the field was not there. The expression, for a vertical magnetic field, of the force is $$F_z=\mu_z \frac{\partial B_z}{\partial z}$$, which clearly shows that $$B_z$$ must vary in space. 7dare (talk) 20:38, 29 November 2017 (UTC)

Electric or magnetic dipole
In the basic description section I'm told I can start by treating the particle as "a classical spinning dipole". I'm guessing that this is an electric dipole. Then I'm told it will precess "in a magnetic field because of the torque that the magnetic field exerts on the dipole" which is missleading because the field exerts opposite forces on the ends(moving charges) of the electric dipole and this produces the precession. Then "if it moves through a homogeneous magnetic field, the forces exerted on opposite ends of the dipole cancel each other out". Now the dipole is a *magnetic* dipole that is induced by the spinning *electric* dipole. This is a different dipole. It seems according to my guessed at interpretation that the precession is a side issue and the particle is deflected because of the direction of the electric dipoles spin. Could someone who knows please add the electric or magnetic descriptors to the text? Thanks119.12.50.19 (talk) 02:39, 27 September 2013 (UTC)


 * Consider a spinning charged sphere as the radius goes to zero, and the magnetic moment does not. Gah4 (talk) 01:49, 1 March 2018 (UTC)
 * That is a good description of a dipole, but not a very good profile of an electron moving in a linear path. If there were a true spin for each and every electron, a cathode ray would split (or otherwise diverge) when moving through an applied field, indicating a spin orientation. Mass spectroscopy wouldn't work for electrons. A magnetic dipole is a body-centered property where spin is involved. If there is a spin up/down for silver, that must be explained in view of there being a single valence electron. If there is no pairing, there can be no up/down orientation according to quantum theory. It seems as though a brute force technique was used to mimic a property that wouldn't otherwise exist. As mentioned in the talk question about entanglement, the experiment is not described thoroughly. Many of the design specifics are missing. A dipole will orient itself in an applied field such as the one described. BRealAlways (talk) 10:14, 9 September 2021 (UTC)
 * It is the field gradient, not the field itself, which causes the force. The atom is neutral, so there is no Lorentz (qv cross B) force. The dipole will not (easily) align with the field. It needs a way to lose energy to do that. Also, it has to conserve angular momentum, like what keeps tops up and gyroscopes pointing the right way. Gah4 (talk) 23:58, 9 September 2021 (UTC)
 * It is the field gradient, not the field itself, which causes the force. The atom is neutral, so there is no Lorentz (qv cross B) force. The dipole will not (easily) align with the field. It needs a way to lose energy to do that. Also, it has to conserve angular momentum, like what keeps tops up and gyroscopes pointing the right way. Gah4 (talk) 23:58, 9 September 2021 (UTC)
 * It is the field gradient, not the field itself, which causes the force. The atom is neutral, so there is no Lorentz (qv cross B) force. The dipole will not (easily) align with the field. It needs a way to lose energy to do that. Also, it has to conserve angular momentum, like what keeps tops up and gyroscopes pointing the right way. Gah4 (talk) 23:58, 9 September 2021 (UTC)

The gradient is in the plane of motion, not across the path of motion. I'm not sure I'm understanding what you mean by "it has to conserve angular momentum". Sounds like you're assuming angular momentum is involved for a particle. S-G couldn't possibly determine that. What is your source?BRealAlways (talk) 23:11, 20 September 2021 (UTC)
 * Precesion of (classical) dipoles in a magnetic filed is discussed in Magnetic_moment. Gradient and field are perpendicular to the velocity. For magnetic deflection, as in CRT TV sets, the Lorentz force is so much larger than the deflection force, that it isn't noticed. And besides, the field normally has a small gradient. It has to have a small gradient, so that the horizontal and vertical deflection are, mostly, independent. In S-G, the atoms are neutral, so no q v cross B force, and only the S-G force is seen. — Preceding unsigned comment added by Gah4 (talk • contribs) 01:40, 21 September 2021 (UTC)
 * Precesion of (classical) dipoles in a magnetic filed is discussed in Magnetic_moment. Gradient and field are perpendicular to the velocity. For magnetic deflection, as in CRT TV sets, the Lorentz force is so much larger than the deflection force, that it isn't noticed. And besides, the field normally has a small gradient. It has to have a small gradient, so that the horizontal and vertical deflection are, mostly, independent. In S-G, the atoms are neutral, so no q v cross B force, and only the S-G force is seen. — Preceding unsigned comment added by Gah4 (talk • contribs) 01:40, 21 September 2021 (UTC)

unacceptable edit undone; reasons
I have undone an unacceptable edit. It is unacceptable because it is personal self-promotion of material that is not supported by Wikipedia-standard reliable sources. The editor who posted it, presumably in good faith, apparently does not know the Wikipedia policies on such matters. He could look at WP:RS and at WP:PROMO.Chjoaygame (talk) 17:18, 11 March 2015 (UTC)

The unacceptable edit was posted again by the same editor. I have undone again his post for the same reasons as just above. I have left him on his talk page a pointer to this talk page.Chjoaygame (talk) 02:08, 12 March 2015 (UTC)

faulty new link
This new edit posts a link that says that the Stern–Gerlach beam splits into three. Not as I see it. Where field inhomogeneity is strong, the splitting of the beam into two is big enough to see above the noise. At the edges, where the field inhomogeneity is weak, the splitting of the beam into two is not big enough to see above the noise. I think the linked article is wrong in this bold claim. I think the link should be deleted.Chjoaygame (talk) 14:44, 11 September 2015 (UTC)
 * This figure is correct, you can cite chapter 5 in vol III of The Feynman Lectures on Physics.Earthandmoon (talk) 16:00, 11 September 2015 (UTC)
 * and if you check history of edit this page, i just changed the dead link of this external link.Earthandmoon (talk) 16:03, 11 September 2015 (UTC)


 * Thank you for this response. Yes, on reading the link again I see it is ok. I didn't read it carefully the first time.Chjoaygame (talk) 22:11, 11 September 2015 (UTC)

New edit
There is a new edit just posted.

I don't know the exact rules, but it seems obvious that this is promotion of the poster's own research. If someone knows the exact rules, perhaps they will say what they are. If there is no restriction on self-posting, presumably Wikipedia will grow faster than Topsy.Chjoaygame (talk) 23:01, 27 February 2016 (UTC)
 * I submitted the new edit based on several years of research resulting in publication in a peer-reviewed scientific journal. The fact that it is my own research, not someone else's work, does not, I trust, imply that it is not a sound, scientific argument. The previous argument, I'm sure, is outdated (Scully, Lamb, and Barut). In that analysis (Ref. 12) they employed an obviously non-physical model of the S-G magnetic field. They conceived of their magnet as being very short in the direction of the incident atomic beam, even compared to the separation distance between magnetic pole pieces. That’s an egregious distortion of a real S-G magnet, which has pole pieces very long compared to the distance between the poles. (See the drawing of a typical S-G magnet in the Wiki article.) And, they used B = (-bx, 0, bz) as their field model, with B = 0 at x=y=0, whereas, today, the accepted model is B = (-bx, 0, B0 +bz). Note that the field model employed by Scully et al. shows NO magnetic field at the location of the atomic beam in the magnet: obviously not a realistic model. Other researchers have also noted the same deficiencies in the analysis of Scully et al. (See the article by Hsu et al., referenced in my edit.) My article in the Canadian J. Phys. is also available on the web at https://tspace.library.utoronto.ca/bitstream/1807/69186/1/cjp-2015-0031.pdf. if anyone would like to look at it.


 * D bar x (talk) 19:11, 28 February 2016 (UTC)Michael Devereux


 * Thank you for this response, Editor D bar x. I was asking about the policies of Wikipedia. Properly, they govern what stays or goes. Since asking my question I have revised a little on Wikipedia policy. Your response does not address my query. I can see you are new to this game. You are acting in good faith. Your personal signature is evident. I am not exactly sure about that. It emphasizes the possibility of conflict of interest.


 * Your response gives your views on the validity of your research. In general, a Wikipedia editor, as Wikipedia editor, has no mandate to base editing on his views on the validity of research. He must base it on various factors, the relevant one of which is reliable sourcing. In general, research in refereed journals, no matter of what standing, and no matter how thoroughly refereed, and based on no matter how many years of research, by no matter how expert a researcher, is not ipso facto classified in Wikipedia as reliably sourced. That is the policy. Posted material should, in general, be based on secondary sources. This is because Wikipedia editors are not mandated to judge the validity of research. There are very good reasons for this policy. Of course there are plenty who defy the policy, but defiance is not justification. Two wrongs don't make a right.


 * No secondary source is quoted in the needed way. Merzbacher is cited for a standard point of physics, but not as directly relevant to the new point of the edit. The article by Hsu et al. is not a secondary source. Outside the edit, there is a possibly relevant secondary source, that would count as a reliable source, the account given by French & Taylor (1978) on pages 428–438 and elsewhere. It is not cited in your edit. I guess there must be many other possibly relevant Wiki-reliable sources. 'Its' is twice wrongly written "it's". Some kind editor would likely fix that. It is forbidden to write such things as "We describe ..."


 * There are other problems very nearby in the article, which desperately need attention from a properly motivated Wikipedia editor.


 * Someone posting in Wikipedia articles does so as a Wikipedia editor, not as an expert on the topic, nor as an advocate for a point of view. I think it fair to say that your edit is not up to scratch on these and perhaps other grounds.Chjoaygame (talk) 22:52, 28 February 2016 (UTC)


 * With a view to estimating its relevance and notability for the Wikipedia article Stern–Gerlach experiment, I have read the research paper that Michael Devereux is seeking to promote in the edit in question.


 * Sad to say, I read as the first sentence of the abstract: "Observation of two separated beam spots at a detection screen downstream of a Stern–Gerlach magnet does not, in fact, demonstrate that the wavefunction of a neutral spin one-half particle has remained in a spin superposition while traveling through that magnetic field." This sentence makes it regrettably evident that its author does adequately understand the concept of quantum superposition. Apparently he did not adequately read Dirac or Feynman. They or their like are necessary background for material of this kind, and are absent from the edit. The atoms from the oven are in a mixture of states. Simple observations, such as are considered in the body of the paper, are thoroughly inadequate to decide on the question of superposition for this case. Consequently, the paper is undoubtedly original research not adequately supported by reliable sources. It is largely about the interpretation of reduction of the wave packet. Its tie to the Stern–Gerlach experiment does not endow it with notability for the present article. Though I observe that its author lacks understanding, I do not say that the paper is all wrong. I say that it is not suitable as a source here. I am removing the edit because it is not reliably sourced, and because it does not add value to the article. I am sorry that, as a neutral editor, I need to do this.Chjoaygame (talk) 03:07, 29 February 2016 (UTC)
 * Besides that it is four years, it sounds like such would apply to an article about quantum entanglement. Since it was four years ago, maybe there are good secondary sources by now. There aren't so many things that can be changed in S-G by now. Gah4 (talk) 18:57, 20 July 2020 (UTC)
 * Besides that it is four years, it sounds like such would apply to an article about quantum entanglement. Since it was four years ago, maybe there are good secondary sources by now. There aren't so many things that can be changed in S-G by now. Gah4 (talk) 18:57, 20 July 2020 (UTC)

angular momentum or magnetic moment?
It seems to me that S-G directly shows the quantization of magnetic moment, and only indirectly the quantization of angular momentum. I realize the connection that we now know between them, but it likely wasn't so obvious in 1920. The beginning of this article also doesn't make this so clear. I could just change it, but wanted to see if others have thoughts about it. Gah4 (talk) 22:47, 28 February 2018 (UTC)


 * Not the case.  Magnetic moments are not quantized while angular momentum is.   Remember that in each physical situation there is a "g-factor" that relates the angular momentum of the system (quantized in units of h-bar) to the magnetic moment.   The g-factor can be calculated for simple atomic systems, but not, in general, for many-body systems like nuclei or molecules. There are tabulated values of *measured* g-factors for all sorts of systems, if you look.   Qwerty123uiop (talk) 23:08, 28 February 2018 (UTC)


 * Yes, but once you set the system, which you have to do when setting up a S-G experiment, then g is fixed. S-G measures magnetic moment, and not angular momentum. I suppose you could be unlucky in a variety of ways, though. From, Ag has a nice electron configuration with only one unpaired electron in an s orbital, so no orbital angular momentum. Gah4 (talk) 01:46, 1 March 2018 (UTC)


 * Sure, once you pick a measuring stick from among all possible measuring sticks, then you can measure any length in quantized units of that stick. But that is not the same as saying the stick-length you picked is in any sense "fundamental".  What stays the same in all quantum systems is the size of the angular momentum unit, that is, Planck's constant... you don't get to pick it. Qwerty123uiop (talk) 16:17, 3 March 2018 (UTC)


 * OK, if you measure in sticks then it is always length. But you might measure length by the time it takes something to go that distance. Radar measures length by the time it takes a signal to reach and return from an object. Directly time, indirectly length, and the speed of EM waves as the scale factor. S-G directly measures magnetic moment, and indirectly angular momentum, with appropriate scale factor. But okay, from Electron magnetic moment the first assumption is that the mass and charge distributions are the same. From that, you only need a dimensionless g-factor. But even so, S-G measures the magnetic moment. Knowing that it is proportional, even without knowing the constant, two values of magnetic moment means two values of angular momentum. After S-G comes the problem of measuring the g-factor, not before. Gah4 (talk) 23:44, 10 May 2018 (UTC)


 * From the title of the SG paper, and even though I don't speak German: .. magnetische Moment des Silberatoms, I am pretty sure that means magnetic moment of silver atoms. Since I don't speak German, I can't be sure that magnetische Moment doesn't mean angular momentum, but I still don't think so. They were measuring the quantization of magnetic moment. We now know, though I suspect that they didn't at the time, that that is due to quantization of angular momentum. (And, as noted, g.)  Gah4 (talk) 03:38, 28 June 2019 (UTC)

sequential experiments
In the sequential experiments section, there are no examples that put a split beam back together. As noted, blocking one beam loses polarization on any other axis. If you don't block a beam, and using a S-G apparatus to put the beam together again (that is, opposite polarity), there is no measurement, and no loss of previous polarization state. Gah4 (talk) 04:33, 1 March 2018 (UTC)


 * I think that's wrong: the measurement happens within the S-G apparatus, which outputs two polarized beams. E.g., if you initially have a x+ beam, and you split it in z+ and z- beams using a S-G apparatus and later merge these two beams, you don't have anymore a x+ beam, i.e., if you now measure it with a new S-G polarized along x you'll get both x+ and x-. --151.56.35.40 (talk) 10:41, 23 September 2018 (UTC)


 * According to which interpretation of quantum mechanics do you believe that? Have you done the experiment?  One of the most interesting things about quantum mechanics is when you are sure of something, but it turns out not to be true, or, as Feynman would say, "Nobody understands quantum mechanics". Gah4 (talk) 16:14, 23 September 2018 (UTC)


 * When I find my Feynman vol. III, I will see what he says about it. Gah4 (talk) 16:14, 23 September 2018 (UTC)


 * I finally found it. Read section 5-4 in Feynman's "Lectures on Physics" volume III.  The book is somewhat unusual in the way, and the order in which, it explains things. If you separate the beams, don't block them or otherwise disturb them, then put the back together again, previous state information is not lost.  It isn't so obvious that you can do this with real S-G apparatus, but is otherwise fundamental to QM. You might have to tune carefully to get the beams back together without losing phase.  Gah4 (talk) 01:57, 16 November 2018 (UTC)


 * You might want to check Ellerman https://arxiv.org/pdf/1112.4522.pdf (a revised version is peer reviewed with Springer https://link.springer.com/article/10.1007/s40509-014-0026-2) and it very nicely explains why it is wrong to assume that the S-G apparatus outputs two polarized beams. I assume this misunderstanding is pretty much widespread and strongly suggest you read Ellerman. — Preceding unsigned comment added by 217.95.160.165 (talk) 22:20, 30 June 2019 (UTC)


 * Yes. And to quote from the above: One of the very few texts to consider such a Stern-Gerlach analyzer loop is The Feynman Lectures on Physics: Quantum Mechanics (Vol. III) where it is called a ”modified Stern-Gerlach apparatus”[2, p. 5-2]. It also mentions another book that mostly gets it right, but that most books ignore this. Gah4 (talk) 18:45, 1 July 2019 (UTC)


 * For another way to think about it, consider the strong baseball principle described by Mermin, and quoted by Ivars Peterson. This is meant to explain things such that anyone can understand them. Specifically, the baseball principle says that watching a games on TV, thoughout the season, doesn't change the results.  The strong baseball principles applies to individual games, and the very strong baseball principle to games that are rained out.  (If a rained out game, was not actually rained out, would the result have been different ...)  A quote from Mermin: "These are intrinsically quantum-mechanical data, and the lesson from these data is ... that you have to be extraordinarily careful in talking about what might have happened but didn't. In this case, the numbers demonstrate that there's no way you can make up a picture to account for what might have happened but didn't."  Yet we go through life, planning our days around what might have happened  ...  but didn't.  Gah4 (talk) 18:45, 1 July 2019 (UTC)


 * In fact, Ellerman argues exactly against the misconception expressed by 151.56.35.40 – “if… you split it in z+ and z− beams using a S-G apparatus and later merge these two beams, you don't have anymore a x+ beam”. Spatial deflection of atoms doesn’t affect their spin. Only when the atom hits an obstacle can we count for a kind of measurement of its state. By the way, IMHO all the “baseball” stuff above can be dismissed as ambiguous and cryptic. Incnis Mrsi (talk) 20:02, 1 July 2019 (UTC)


 * The baseball stuff is meant to connect everyday experience to you have to be extraordinarily careful in talking about what might have happened but didn't.  There are more detailed discussions in Mermin's Boojums All the Way through.  Specifically, he uses statistics that could come from SG experiments, either spin 1/2 or spin 1 (photons), which show that no possible results of experiments that weren't done could agree with the ones that were done.  And this is all without what Feynman calls the modified Stern-Gerlach apparatus.  Gah4 (talk) 21:15, 1 July 2019 (UTC)


 * (Excuse my English if it's not good, I'm not a native English speaker.) I would like to remind you that the superposition IS an interpretation, not a fact. We have no means to know whether the spin is determined inside the S-G apparatus with the deflection of the beams or the beams are entangled and each one is in a superposition of both states before touching the screen. The literature is plethoric about thought experiments of putting a split beam back together (which can be called an eraser experiment) but there are quite few actual experiments of the situation. This is what I found : Scully is the 1st one who studied it and he made 3 experiments trying to produce interferences at the recombination of the beams. The absence of interferences express the decoherence of the beams (thus the presence of which-path information, which can be either interpreted as : both paths are superposed as to their spin characteristic but separated in space = not spatially coherent anymore, or as : each path corresponds to a determined spin). The presence of interferences proves the coherence of the beams (the absence of which-path information can be either interpreted as : there is a total entanglement between all possibilities of spin and spacial distribution or : each particle already has a spin but has not undergone spatial decoherence yet). Anyway you will never have as an output a clean beam identical to the one you got as an imput. The passage into the device does something to the beam, whether it is changing it into a mixture or into interferences. But even if you are satisfied with interference because you are only looking for coherence (and not for the impossibility to tell if a measurement was made or not, which I would rather be looking for), it is very difficult to produce them experimentally because of the magnetic field, which is not perfect and has little fluctuations (due to its quantum nature), and these fluctuations provoke systematic decoherence of the beams, and apparently also because of the constraint of irreversibility of processes (linked to the irreversibility of time). I haven't understood the techniques they tried to correct that but it appears to be still a work in process. The more complete article I found on the subject is this one : (open access), he makes a little review in introduction and appears to succeed in recovering the coherence after recombination so if this article holds I think the effect is experimentally confirmed. As you can see it's very recent so I would wait a bit before being sure there is a consensus.Hellena Liseron (talk) 11:44, 25 March 2020 (UTC)
 * I started the discussion based on what Feynman says in his lectures, vol. III. How easy it is to build, is another question. But note that, other than the difference between spin 1/2 and spin 1, the same experiments can be done with calcite crystals and polarized light. (And they are commonly done that way for entanglement experiments.) If putting the beams back together is less than perfect, then the polarization of the result will be less than perfect, but maybe still measurable. This gets very close to Interpretations_of_quantum_mechanics though. Gah4 (talk) 15:59, 25 March 2020 (UTC)

Feynman lectures can be found here []BRealAlways (talk) 03:10, 2 February 2022 (UTC)

Stern-Gerlach Experiment
Imagine i have no clue about this. Lets just assume im currently hearing quantum mechanics and the SG experiment is mentioned. This article would just confuse me.

What causes the forces? The spin interacts with the magnetic field, sure. But how? why? Shouldn't there be some sort of description on why this even happens for the readers that come here to pick exactly that up? when reading up the spin, i cant even find the reason why the spin would cause a magnetic field. an unexperienced reader wouldnt be smarter after reading the article. it sure says the spin causes a magnetic moment, but why would that be? then again, the experiment is useless for understanding if non-commuting operators arent even mentioned. its more confusing than anything.

for some motivated editor: to understand this experiment at all, the page on spin itself must be edited. i think there is a need for a hamiltonian, a minimal coupling, how to come from the minimal coupling to a B-field, how to get from the B-field to the orbital spin, and showing the similarity of orbital spin and spin after introducing the spin in the dirac hamiltonian, which explains why the spin would cause a magnetic field -> aka why the beam would even split in the experiment

on the stern-Gerlach page it must be explained what it means to have non-commutable operators, why it explains the result of the experiment, and why the experiment alone will already tell us much of the structure of the spin matrices (2x2, hermitian, traceless) most importantly though: it should be mentioned that the spin itself causes a magnetic field, making at spin-particle a magnetic dipole, causing a force nabla*(mu*B). as mu can take 2 values (per direction x,y,z), we see 2 dots.

this isnt trivial at all in my opinion.


 * Well, you do have to know that magnets get deflected in magnetic fields, but the whole idea about S-G is that angular momentum and magnetic moment are quantized. In Newtonian mechanics, angular momentum is a continuous variable.  In classical electricity and magnetism, magnetic moment is a continuous variable. In quantum mechanics, they are not continuous variables, but can only have discrete values.  In the case of Ag atoms, only two values. This was one of the early experiments that defined quantum theory.  There is no explanation, it is just the way nature works. We aren't allowed to look inside an electron and ask what the mechanism for spin is. In classical E&M, a current loop, or rotation charged ball, creates a magnetic field. But that doesn't help much for an electron, which is consider a point particle. It is the limit of a rotating charged sphere as the radius goes to zero, but the field does not. There is no why that you can ask. Gah4 (talk) 20:17, 26 March 2018 (UTC)


 * "Well, you do have to know that magnets get deflected in magnetic fields". I assume, you have not understood why the magnetic field has a gradient. Magnets are not simply deflected in a magnetic field. A gradient is necessary, and that is entirely nontrivial. Calculating anything involving magnets is intensely simplified by substituting a complicated dipole field with a constant number: B. Now that we have not two dipole fields interacting, but a gradient field, and a dipole, the exact force parameter deserves an explanation. Just throwing it in there, cannot be the way to go. 2A02:8070:BAB:DB00:292E:9DEC:1E85:1BA3 (talk) 18:09, 18 July 2021 (UTC)SomeGuy


 * The Dirac equation perfectly describes the spin. There is no intuitive way to understand the spin, it basically simply comes from the math behind the Dirac equation. Ofc that all matters what you consider "an explanation".
 * Now, whether the spin and its quantization can be "explained" is one thing. But, having introduced the spin, we can see that the spin causes a magnetic field, looking at the Dirac equation. It just dissappointed me to not even see that in the article on spin itself.
 * Anyways, more importantly:
 * What confused me initially, is the inhomogenous field. In the german article there is some math to look at. I just wondered why I would need an inhomogenous field, my EM lectures where long ago. I just wouldnt have intuitively expected (for example) a bar magnet to rest in a homogenous field, but the german article gave me a familiar formula and everything was clear to me as soon as I saw that. Maybe stupid of me, maybe not, i dont know.


 * It isn't so obvious, as everyday magnets, such as bar magnets, have an inhomogenous (B is not constant) field. The field spreads out from one end, follows curves around to the other end, and then converges into that end. Ferromagnetic objects are attracted to the ends, as the field is higher, and the magnetic energy reduced more. In the case of S-G, you need a long path with inhomogeneous field, so they use a magnet with long, narrow, poles, but with end side pointy, and the other side flat. (The drawings should show this.) The field is higher near the pointy part, and lower near the flat part. Gah4 (talk) 19:09, 27 March 2018 (UTC)

Video
hey guys, idk how to make comments correctly but the video is wrong, the diagonal magnets would be deflected towards their opposite poles — Preceding unsigned comment added by 2600:1700:67D0:C4A0:4177:9510:C275:E59E (talk) 06:29, 11 March 2018 (UTC)

amount proportional to its magnetic moment
It is the component of the magnetic moment in the direction of the magnetic field, that is, $$\mu \cdot B$$. The supposed randomness is in the direction of the magnetic moment. Gah4 (talk) 07:10, 9 May 2018 (UTC)

Did my caption edit involve a bad side effect?
I replaced Much of the text of the caption for the diagram of the experiment, and began to suspect that I had  duplicated at least part of the existing caption. My best guess is that I made no error, but my tool is an iPad 2 which makes my editing somewhat clumsy, so it's hard to be sure. The code I have about it is because of repetitious looking markup, and I am not in a position to do a version to version compared to confirm my best guess to The effect that The repetition is intended and has nothing to do with my edit. I apologize if I am mistaken and hope a colleague will remedy any damage that I did. --Jerzy•t 06:52, 14 August 2018 (UTC)

Larmor precession
I found this sent, and appended the "vague" tag:
 * If the particle is treated as a classical spinning magnetic dipole, it will precess in a magnetic field because of the torque that the magnetic field exerts on the dipole (see torque-induced precession).. My tagging reflects what is surely a colleague's vagueness between two realms of discourse that respectively concern how a topic is dealt with (i.e. "treated" by us or our readers), and what the facts of that topic are. I comment here bcz i lack the patience and interest to fix it, and bcz of the decades that have passed since being actively interested in a precession that isn't occuring in my field of view. And also bcz proposing blanking of the lead, as a means of harm reduction, seems harsh. Who is interested enuf to improve the lead before i shift from moral suasion to  more clearly labelling our dirty laundry? --Jerzy•t (working as an ipUser 09:31, 14 August 2018 (UTC))


 * I find the above comment incomprehensible. Specifically, the referent of "a colleague's vagueness between two realms of discourse" is completely unclear.
 * If no one else understands this either, I suggest removing the "vague" template as it has no clear purpose. Dratman (talk) 19:22, 27 January 2020 (UTC)
 * Removed it.LaurentianShield (talk) 01:29, 28 October 2021 (UTC)

the nuclei of some atoms also have quantized angular momentum
Is the nuclei of some atoms also have quantized angular momentum meant to indicate quantized, as opposed to non-quantized, or as opposed to zero angular momentum? Gah4 (talk) 22:24, 14 April 2019 (UTC)


 * Probably irrelevant in this article, since the nucleus of input molecules contributes almost nothing to the splitting done by the apparatus. Perhaps the article should explain why this is so. David Spector (talk) 20:30, 19 November 2019 (UTC)


 * I wonder now what they knew at the time. When was nuclear spin discovered? In any case, I agree that it could be added to the article. Gah4 (talk) 21:58, 19 November 2019 (UTC)


 * As the article says, even electron spin had not been postulated, when the experiment was done; Pauli was in 1924 the first to propose nuclear spin to explain the hyperfine structure known long before, cf . The article is German, it is cited as the original source, e.g., in Goudsmit, Pauli and nuclear spin, Physics Today 14, 6, 18 (1961);  (to which I don't have access) and Corney, Atomic and Laser Spectroscopy, Clarendon Press (1988), p. 661. Maybe the earliest direct measurements of the nuclear magnetic moment was Rabi (1938).
 * I don't think there's a need to mention nuclear spin in the Stern-Gerlach article: the nuclear magnetic moment is so much weaker that is does not affect the experiment much if there's an electron spin present - and for the experiment the nature of the angular momentum (spin, orbital, or nuclear or some combination) is only quantitatively important. --Qcomp (talk) 00:01, 20 November 2019 (UTC)

Size and price of S-G apparatus?
The article explains that the S-G apparatus is basically two magnets in some sort of bracket, with an optional electric field to counteract the induced deflection force when the input particles are charged. There is no photograph. I have not been able to find any information on how large this apparatus is, which is relevant to how easy it is to rotate to measure spins in different directions. I can imagine it as several cm long or several meters long, and I'd like the actual range of sizes to be included in the article, if someone reading this knows. It might be interesting to include typical prices, too. Thanks, David Spector (talk) 20:27, 19 November 2019 (UTC)


 * As far as I know, it is close to 1m. The atoms are in vacuum, so there is a big vacuum system around them. It is field gradient that matters, but for a given shape, the gradient is proportional to field. The length then depends on how large the gradient is. Gah4 (talk) 21:56, 19 November 2019 (UTC)

What kinds of labs can afford a commercial version of the S-G apparatus? High school? College? Or only well-funded research labs? David Spector (talk) 13:06, 25 August 2021 (UTC)
 * Since the experiment has been done, it is not normally necessary to repeat. I suppose it could be done as a college lecture demonstration (though it might take too long), or upper year college lab experiment. I don't know any that do that. It might be that you can buy the magnet and vacuum system for use in other experiments, and then find or build an appropriate silver atom source. In a well stocked research lab, there would already be vacuum pumps that could be reused, which is a big part of the cost. As well as I know, there are no commercial models available. The magnet might be available from another experiment, too. Gah4 (talk) 16:36, 25 August 2021 (UTC)
 * Since the experiment has been done, it is not normally necessary to repeat. I suppose it could be done as a college lecture demonstration (though it might take too long), or upper year college lab experiment. I don't know any that do that. It might be that you can buy the magnet and vacuum system for use in other experiments, and then find or build an appropriate silver atom source. In a well stocked research lab, there would already be vacuum pumps that could be reused, which is a big part of the cost. As well as I know, there are no commercial models available. The magnet might be available from another experiment, too. Gah4 (talk) 16:36, 25 August 2021 (UTC)

While it is certainly not necessary to repeat a classic experiment, it helps make the subject concrete for students. In graduate school we repeated the Millikan oil drop experiment, which made quantization of charge (carried by electrons) concrete for me and my fellow students. A small but working S-G experiment would be a great addition to any beginning quantum mechanics course, along with the usual cheap polarizing filters. It would make intrinsic spin concrete for students, so it wouldn't remain a foreign or strange concept. Note that an electron source would undoubtedly be cheaper than a silver atom source. And balancing out the electron charge would be interesting to students. Also, the low expense is also needed because two or three S-G sections would be needed for reasoning about orthogonal states. David Spector (talk) 20:30, 15 September 2021 (UTC)
 * Yes. I suspect it isn't done so much (or at all), as the equipment is more expensive than most labs want. We had the Millikan oil drop experiment as one of the choices for upper undergraduate lab, but I didn't do that one. That was on the original apparatus, too. There are some that are done for lecture demonstration, but not for actual students to do, but it might be too expensive for that. Photon experiments are usually easier to set up. I am trying to think if you can make a convincing demonstration of photon spin quantization. Gah4 (talk) 23:19, 15 September 2021 (UTC)
 * Yes. I suspect it isn't done so much (or at all), as the equipment is more expensive than most labs want. We had the Millikan oil drop experiment as one of the choices for upper undergraduate lab, but I didn't do that one. That was on the original apparatus, too. There are some that are done for lecture demonstration, but not for actual students to do, but it might be too expensive for that. Photon experiments are usually easier to set up. I am trying to think if you can make a convincing demonstration of photon spin quantization. Gah4 (talk) 23:19, 15 September 2021 (UTC)

New Image/3D model added to Sequential Experiments
I thought there was lacking images to show what is going on in the Sequential Experiments section, so I created a 3D model to show each of the 3 experiments a bit better. Please let me know if there is something incorrect in the image, maybe I can fix it and re-upload. Also I wasn't sure what to put in the caption, please improve on it if you can. MJasK (talk) 20:32, 25 November 2020 (UTC)

The act of observing (measuring) the momentum along the z-axis
"The act of observing (measuring) the momentum along the z axis corresponds to the operator J_{z}.[specify] "

I think that the operator mentioned should be the pauli operator sigma_z

Kriesell.D (talk) 16:55, 16 January 2021 (UTC)

neutrons
For beginners the use of "neutrons" might be a little bit confusing, because the Stern Gerlach experiment is associated with the spin of electrons. Kriesell.D (talk) 17:19, 16 January 2021 (UTC)
 * That might be, but it has to do with spin, or more specifically magnetic moment, and neutrons have one just like electrons. While they are neutral, they do have a magnetic moment, and actually do have a charge distribution. The charge isn't evenly distributed throughout. Gah4 (talk) 22:11, 16 January 2021 (UTC)
 * That might be, but it has to do with spin, or more specifically magnetic moment, and neutrons have one just like electrons. While they are neutral, they do have a magnetic moment, and actually do have a charge distribution. The charge isn't evenly distributed throughout. Gah4 (talk) 22:11, 16 January 2021 (UTC)

Wrong Explanation
This explanation is wrong: “Video explaining quantum spin versus classical magnet in the Stern–Gerlach experiment”. The classical magnet would rotate and produce only two dots. The original experiment shows that the absolute value of angular momentum has two values - not that it has two orientations! Zyavrik (talk) 21:40, 11 July 2021 (UTC)

Continuous Spectrum Classic Dipole
The explanation that a classic magnetic dipole would cause a continuous spectrum, is not well reasoned. Classically, a magnetic dipole would align with the field, and then yield a 2 point spectrum too. Or is there a specific reason why this would not happen classically? 2A02:8070:BAB:DB00:292E:9DEC:1E85:1BA3 (talk) 17:55, 18 July 2021 (UTC)SomeGuy
 * Dipoles don't easily align with the field. They need a way to lose energy to do that. Normally they precess in a field like that. It is a complication of conservation of angular momentum. Gah4 (talk) 00:07, 10 September 2021 (UTC)
 * Dipoles don't easily align with the field. They need a way to lose energy to do that. Normally they precess in a field like that. It is a complication of conservation of angular momentum. Gah4 (talk) 00:07, 10 September 2021 (UTC)

Pattern image needed
The article seems to refer to the visible image formed by a free electron on a screen in the S-G apparatus, but fails to include a photograph of such an actual image. I see this lack elsewhere as well. I think the article deserves a real photo of the S-G output, along with an explanation of why it is not a pair of separated dots or short lines as expected from theory. If this is not possible, an explanation of why no such image or explanation can be included would be appreciated. Perhaps I have missed something basic about the S-G experiment. David Spector (talk) 13:03, 25 August 2021 (UTC)


 * There is an excellent image of the result from a paper already cited on the page (available here https://plato.stanford.edu/entries/physics-experiment/app5.html). I see no reason not to include it, but what do you mean by "not a pair of separated dots or short lines as expected from theory"? As I see it there are two lines as predicted by the theory, they simply taper out as you go further from center due to the nature of the magnetic field. G1butler615 (talk) 16:33, 29 November 2021 (UTC)
 * Just to be clear, the article makes no claims about free electrons. The experiment involves Silver atoms, not free electrons. Johnjbarton (talk) 18:28, 21 February 2024 (UTC)