Lamb and The Photoelectric Effect Without Photons

[bcrowell]

Lamb and "The Photoelectric Effect Without Photons"

I've recently been pointed by two different people to this paper by Lamb (yes, that Lamb) and Scully: http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19680009569_1968009569.pdf . The title is "The Photoelectric Effect Without Photons." Despite its famous first author, I think the paper is clearly bunk (and therefore it's not surprising it was never published in a refereed journal). Below is my analysis, cut and pasted from a FAQ I maintain at http://www.lightandmatter.com/cgi-bin/meki?physics/faq . I may, of course, be wrong. If you think so, let me know :-)

Didn't Lamb and Scully show that you don't need photons to explain the photoelectric effect?

No.

There is a 1968 paper by Willis Lamb, Jr. and M.O. Scully, "The Photoelectric Effect Without Photons," that seems to have developed a life of its own on the internet. It was never published in a refereed journal. What you can find on the web is a pdf file made from scans of an internal publication of the Center for Theoretical Studies at the University of Miami.

It may help to understand something about the historical context. Einstein hypothesized the photon in 1905, but his paper was ahead of its time and was not widely accepted. For decades afterward, even once the quantum-mechanical nature of the *atom* was assumed by all physicists, the quantum-mechanical nature of *light* was considered suspect. Bohr was influential in pushing a theory in which atoms were quantized, but the light they absorbed and emitted was classical. Lamb began his career during this era.

If you read the Lamb-Scully paper, the first thing you notice is that they explicitly state that photons are absolutely necessary in order to explain phenomena such as blackbody radiation, Compton scattering, spontaneous emission, and the Lamb shift. Any internet kooks who are trying to quote Lamb as an authority against quantization of light are way off base.

As in Bohr's old-fashioned dead-end approach, they then treat the atom as a quantum-mechanical system and the electromagnetic field as a classical one. They are able to reproduce the Einstein relation E=hf-W, where E is the maximum energy of the electron once it leaves the cathode, h is the quantum-mechanical Planck's constant, f is the frequency of the light, and W is the energy required for the electron to escape through the surface of the cathode. This is not particularly surprising or impressive in a bastardized quantum/classical calculation like this one; essentially it just says that the light wave has to have the energy taken out of it at a resonant frequency that matches its own frequency.

They also show that the transition rate is nonzero even when the light is first turned on, saying that their transition rate "certainly does not imply the 'time delay' which some people used to expect for the photoelectrons produced by a classical e.m. field." This result is not as impressive as they make it sound, since the classical prediction is what one expects for a classical light wave impinging on *classical* atoms. In fact, the transition rate they derive shows the real problem with their calculation. Their calculation treats every atom as *independent* of all the other atoms. Therefore if a single photon illuminates the cathode, it may ionize more than one atom, violating conservation of energy. This unphysical result shows the opposite of what they claim; it shows that their mixed quantum-classical Frankenstein fails to provide a physically acceptable explanation of the photoelectric effect. What they really need is a quantum-mechanical entanglement between the different parts of the photon's wavefunction, so that if the photon is observed at atom A, it is guaranteed not to be observed at atom B. Without this quantum-mechanical "spooky action at a distance," their theory violates conservation of energy.

 

[conway]

First of all, thanks to bcrowell for posting the link to the paper which was referred to elsewhere in the discussion minus the actual link. I found the paper to be good and bad in ways that are different from the criticism posted here. What I found missing from the paper was a good physical description of how the process goes. This is disappointing because the great strength of the semi-classical approach is that it does indeed give real-time description of the process from start to end, unlike the SQM idea of the quantum leap where you're not allowed to ask how the system gets from A to B. To some extent I have to forgive the authors for this lapse because what they present instead is also something very necessary: a careful mathematical calculation using standard methods of SQM for what should be very straightforward problem: evaluate the transition rate from the ground state to the unbound states for an atom in the presence of a oscillating electric field.

Is this not the kind of calculation which a qualified practitioner of quantum mechanics should be able to carry out? The author's look for the well-known identifying characteristics so often talked about in the debate over waves vs particles: the frequency dependence, the cutoff frequency and prompt emission. They find that these phenomena all appear in their model just as they do in the SQM version involving photons.

BCrowell criticizes this as being the same as "Boh'rs old-fashioned dead-end approach". I don't see his point at all. In quantum mechanics you have an atom with its bound and unbound states, and you can also have an external potential; if you choose, surely you can allow this potential to oscillate sinusoidally. Then you take what is widely known as the "most accurate theory known to man" and apply it to this problem. Isn't that what the authors have done?

 

[ZapperZ]

I still find this approach of treating the photoelectric effect as being the interaction of a photon with an atom to be rather naive. The photoelectric effect is done on metals, and the collective effect of the metal to produce the conduction band seems to have been ignored in here. The photoelectrons in a standard photoelectric effect comes from this conduction band, not from a particular atom of the metal.

Zz.

 

[conway]

If you read the Lamb-Scully paper, the first thing you notice is that they explicitly state that photons are absolutely necessary in order to explain phenomena such as blackbody radiation, Compton scattering, spontaneous emission, and the Lamb shift.

That's one of the things I didn't like about the paper. This statement is made without any backup and doesn't have anything to do with the body of the article.

 

[sylas]

That's one of the things I didn't like about the paper. This statement is made without any backup and doesn't have anything to do with the body of the article.

I disagree. I think this is a very important point, lest the paper be misunderstood as arguing that light is not quantized. The statement here doesn't need backup; it is elementary background information, properly cited if anyone actually needs it, given as a useful and relevant context to what is in the paper.

Having such comments in a paper like this is crucial, because otherwise the paper becomes far too easy for the cranks to abuse in support of something that the paper is not actually saying.

The actual claim of the paper appears to be that the photoelectric effect itself is insufficient on its own to establish the quantum nature of light. This claim can be evaluated on its own merits, and it's good to have that done here.

The title of the paper is unfortunate, as it lends itself to misunderstanding by those who want to deny the quantum nature of light altogether. So it's a very good thing that the paper explicitly shows this to be invalid, and that the author recognizes the well established quantum nature of light.

By the way, although the paper was not published in a journal, it was published, I understand, as part of a volume of papers. The reference is:

• Lamb W E and Scully M O (1969) "The photoelectric effect without photons", in "Polarization, Matière et Rayonnement", Volume in Honour of A. Kastler (Presses Universitaires de France, Paris, 1969)

Felicitations -- sylas

 

[atyy]

I still find this approach of treating the photoelectric effect as being the interaction of a photon with an atom to be rather naive. The photoelectric effect is done on metals, and the collective effect of the metal to produce the conduction band seems to have been ignored in here. The photoelectrons in a standard photoelectric effect comes from this conduction band, not from a particular atom of the metal.

Zz.

Does Einstein's first treatment address this objection?

 

[ZapperZ]

Does Einstein's first treatment address this objection?

Einstein first treatment, i.e. the photoelectric effect equation, isn't the issue that I was addressing.

Zz.

 

[heldervelez]

That's one of the things I didn't like about the paper. This statement is made without any backup and doesn't have anything to do with the body of the article.

The paper has its own merits and I agree with Conway.
It is abbusive the statement that others effects can not have a nonquantum explanation.
Because of the strong momentum of QM it was simpler to made those remarks.

" they explicitly state that photons are absolutely necessary in order to explain phenomena such as blackbody radiation, Compton scattering, spontaneous emission, and the Lamb shift."

one can find :
http://www.springerlink.com/content/t7n71u8161824842/"
abstract "We prove that a correct nonquantum treatment of electromagnetic radiation leads to Planck''s distribution (including zero-point energy). The only principles that we invoke are general invariance principles or, alternatively, the impossibility of violating the second law at a macroscopic level."

 

[heldervelez]

Semiclassical Compton scattering:

http://www.iop.org/EJ/article/0049-1748/7/7/A10/QEL_7_7_A10.pdf"
V A Dubrovskiĭ and B G Tsikin 1977 Sov. J. Quantum Electron. 7 832-836

http://www.springerlink.com/content/m80x2818r44850h3/"
Summary It is shown that the classical interaction between a charged particle and a radiation field (Thomson scattering) involves a hierarchy of processes analogous to that for photon-electron scattering (Compton scattering) in quantum electrodynamics. Statements that the double and multiple Compton effects are intrinsically quantum-mechanical processes are incorrect; it is shown that double Thomson scattering is the classical counterpart of the double Compton effect. However, it is only the inverse of double Compton scattering (two photons scattered into one photon) which has a classical counterpart; the direct process does not. Familiar examples of the interaction of three waves in nonlinear plasma theory are shown to be particular cases of double Thomson scattering.

http://www.springerlink.com/content/r10am90am8v1a18p/"
Abstract

Scattering of a plane electromagnetic wave by free and bound wave packets is analyzed by semiclassical radiation theory. It is shown that the theory gives the correct answer to the question of radiation intensity in the photoelectric effect and to the correlation problem in Compton scattering. The expression for the intensity of the scattered radiation differs from the cross section which is derived from the model based on the particle nature of the electromagnetic field. The meaning of this difference is discussed. Low frequency spectrum of the scattered radiation on a bound charge is obtained.

 

[heldervelez]

"[URL shift semiclassical
http://www.opticsinfobase.org/abstract.cfm?URI=josab-16-1-173"
Abstract
It is often remarked that an explanation of spontaneous emission and the Lamb shift requires quantization of the electromagnetic field. Here these two quantities are derived in a semiclassical formalism by use of second-order perturbation theory. The purpose of the present paper is not to argue the validity of QED but rather to develop a semiclassical approximation to QED that may nonetheless have certain computational advantages over QED. To this end, the vacuum of QED is simulated with a classical zero-point field (ZPF), and as a consequence the resulting theory is entitled semiclassical random electrodynamics (SRED). In the theory, the atom is coupled to the ZPF and to its own radiation-reaction field through an electric dipole interaction. These two interactions add to produce exponential decay of excited states while they cancel each other to prevent spontaneous excitation of the ground state; the Lamb shift appears in the theory as an ac Stark shift induced by the ZPF. The spontaneous decay rate of an excited-state derived in SRED is equal to the Einstein A coefficient for that state, and the Lamb shift agrees with that of nonrelativistic QED. Moreover, SRED is shown to be useful for the numerical simulation of spontaneous decay.

Spontaneous Emission in Semiclassical Radiation Theory[/URL]
It has recently been shown by Jaynes and collaborators that semiclassical radiation theory contains a description of spontaneous emission of radiation and of radiative level shifts. The present paper gives a gauge-invariant derivation of the radiant energy production rate, using only Maxwell's equations and the usual definitions of electric current and charge density for a many-particle material system described by Schrödinger's equation. A complementary derivation, using the time-dependent Schrödinger equation, verifies the instantaneous conservation of energy and probability during the radiation process. The semiclassical rate of spontaneous emission differs from the usual formula (Einstein's A coefficient) because it depends on the occupancy of both initial and final states of the material system. The implications of this with regard to thermal equilibrium and Planck's law are examined. If a new hypothesis is introduced, postulating the decomposition of the equilibrium radiation into incoherent components, each interacting with a specific pair of energy levels of the material system, then Planck's law is shown to hold for the total intensity of radiation. If this hypothesis is not introduced, the equilibrium conditions for different transitions are incompatible, and the semiclassical radiation theory is incapable of describing thermal equilibrium.

 

[ZapperZ]

But I can describe charge transport in metal via semiclassical description as well. It doesn't mean that it works ALL THE TIME!

Again, the problem here is that many of these are tackling the NAIVE phenomenon known as the "photoelectric effect". It is ignoring the fact that this phenomenon has gone on to greater heights (photoemission phenomenon) in which there are no classical or semi-classical formalism to describe things such as angle-resolved photoemission spectroscopy (did anyone come up with a semiclassical anything to replace the Spicer 3-step model?), multiphoton photoemission, resonant photoemission, etc. In other words, there is MORE information that resides in the energy and momentum of the emitted photoelectrons that tells you of the mechanism of the emission. These seems to have been overlooked in all of the so-called classical and semiclassical description.

Zz.

 

[jtbell]

http://www.opticsinfobase.org/abstract.cfm?URI=josab-16-1-173"
Abstract
[...] The purpose of the present paper is not to argue the validity of QED but rather to develop a semiclassical approximation to QED that may nonetheless have certain computational advantages over QED. [...]

Just to single out that sentence from the abstract [I added the boldface emphasis].

 

[sylas]

Even more explicit...

http://www.springerlink.com/content/m80x2818r44850h3/"
Summary [...] However, it is only the inverse of double Compton scattering (two photons scattered into one photon) which has a classical counterpart; the direct process does not. ...

heldervelez's comments actually demonstrate just how important it is for an author like Lamb to include appropriate statements of their own real position. This prevents helderverlez and such ilk from distorting the author's own real position; he is reduced to merely denigrating Lamb for refusing to share his own curious views. Most readers, I expect, will recognize that Lamb actually means what he says. He actually disagrees with helderverlez on the nature of electromagnetic radiation. Attempts to paint this as a lack of integrity in the face of the "momentum" of QM are transparently self-serving.

Sylas

 

[Physics Monkey]

Suppose we ignore questions of history and educational merit.

I think we might agree on the following statements?

1. The notion that the photo-electric effect (in its most primitive form) simply proves the existence of photons is scientifically a bit too naive. Again, I'm ignoring the history (maybe it was taken historically as really solid evidence, I don't know) and the educational question.

2. Lamb and Scully's model does appear to account for some aspects of the photo-electric effect. There is also a cartoon in terms of photons which predicts many of the same features. Given that we accept that photons are the ultimately more useful description, it is still an interesting question ask to just how far the classical model can take you. Perhaps the classical model is much more practical for many calculations.

3. The most relevant physics criticism may be ZapperZ's since he points out that the simplest experimental photo-electric effect is observed in metals, etc. An interesting question is then does Lamb's method work here as well, with some simple model of the metal? For example, if we treat the electrons in the metal as classical a la Drude (which is adequate for some transport phenomena) can we still recover the photo-electric affect without assuming photons?

 

[atyy]

Suppose we ignore questions of history and educational merit.

I think we might agree on the following statements?

1. The notion that the photo-electric effect (in its most primitive form) simply proves the existence of photons is scientifically a bit too naive. Again, I'm ignoring the history (maybe it was taken historically as really solid evidence, I don't know) and the educational question.

2. Lamb and Scully's model does appear to account for some aspects of the photo-electric effect. There is also a cartoon in terms of photons which predicts many of the same features. Given that we accept that photons are the ultimately more useful description, it is still an interesting question ask to just how far the classical model can take you. Perhaps the classical model is much more practical for many calculations.

3. The most relevant physics criticism may be ZapperZ's since he points out that the simplest experimental photo-electric effect is observed in metals, etc. An interesting question is then does Lamb's method work here as well, with some simple model of the metal? For example, if we treat the electrons in the metal as classical a la Drude (which is adequate for some transport phenomena) can we still recover the photo-electric affect without assuming photons?

Agree with 1 and 2 fully, and 3 partially. For 3, I would say relevant wrt to what? If it's with respect to 1, then we should also ask the criticism would apply to Einstein's analysis; but with respect to 2, then some of the other stuff that Lamb (spontaneous emission, lamb shift etc) mentioned would be equally relevant. Also, Drude is entirely classical isn't it? I think Lamb's point is that while the (naive) photoelectric effect is quantum, the quantumness could be in either the material or light. What's one step up from the Drude model?

 

[heldervelez]

... he is reduced to merely denigrating Lamb for refusing to share his own curious views.

It was not my intention. Just to say that the papper has a positive assertion about one issue, and it is simpler to verify the steps and assumptions.
The negative assertions he made about others issues, not in analyse, are just an oppinion not necessary to the article.
Just to prove my point I did a google search and came up with those papers that apparently claim differently.

Not all statements are equally strong (verifiable).
As an example: I, living in a world of black tables, can say 'this table is black' but can not verify 'there are no non black tables' (without checking every single one).

Attempts to paint this as a lack of integrity in the face of the "momentum" of QM are transparently self-serving.

We are humans and the context always matters, even in physics. It is not about integrity.

 

[sylas]

It was not my intention. Just to say that the papper has a positive assertion about one issue, and it is simpler to verify the steps and assumptions.
The negative assertions he made about others issues, not in analyse, are just an oppinion not necessary to the article.

We disagree there. The value of those statements is particularly clear from this thread.

Just to prove my point I did a google search and came up with those papers that apparently claim differently.

Well, you got at least two of them wrong, even quoting what is in the extracts you provided here.

We are humans and the context always matters, even in physics. It is not about integrity.

If you think it is not about integrity, then you should not have said "Because of the strong momentum of QM it was simpler to made those remarks." That's a pretty clear imputation of a lack of integrity. How about assuming he made the statements because he felt it was true and useful context?

That's what I certainly think!

A more reasonable assumption is that Lamb made those statements precisely to clarify that light certainly DOES have a quantum nature, even if in his estimation this cannot be proved just from the photoelectric effect. This is important to help clarify that he is not giving reasons to doubt the quantum nature of light.

Cheers -- Sylas

 

[ZapperZ]

Agree with 1 and 2 fully, and 3 partially. For 3, I would say relevant wrt to what? If it's with respect to 1, then we should also ask the criticism would apply to Einstein's analysis;

I don't think it does. Einstein's photoelectric effect model makes no assumption about the mechanism at the cathode other than the requirement of some minimum energy for emission (the work function). There's no attempt to model "atoms".

but with respect to 2, then some of the other stuff that Lamb (spontaneous emission, lamb shift etc) mentioned would be equally relevant. Also, Drude is entirely classical isn't it? I think Lamb's point is that while the (naive) photoelectric effect is quantum, the quantumness could be in either the material or light.

Then that makes the Lamb's model even MORE severely wrong, because the quantumness of the material is definitely very unlike an atom, which has no conduction band.

Zz.

 

[bcrowell]

The notion that the photo-electric effect (in its most primitive form) simply proves the existence of photons is scientifically a bit too naive. Again, I'm ignoring the history (maybe it was taken historically as really solid evidence, I don't know) and the educational question.

I would actually turn your argument around a little bit.

Historically, it is certainly much too naive to see the photoelectric effect as proof of the quantum nature of light.This is clear because people like Bohr were arguing for a classical theory of light several decades after Einstein's 1905 paper. Einstein's paper didn't really propose a complete photon model; it was (as others have pointed out) a cartoon model.

Scientifically, I don't think the statement is too naive at all. Lamb and Scully produced a model that violated conservation of energy. If a classical flash of light comes along with an energy equal to 1.5W, where W is the work-function of the metal, then conservation of energy dictates that the number of electrons ejected through the surface is either 1 or 0. Lamb and Scully's model says that with quite high probability, the number of electrons ejected will be 2 or more. Since nobody has been able to produce a model that simultaneously (1) has a classical electromagnetic field, and (2) satisfies conservation of energy, I think it is absolutely fair to say that the photoelectric effect suffices to prove the quantized nature of light.

Educationally, what I really abhor is the textbooks that introduce the photon via blackbody radiation. It's a horrible mess to try to get that across to students. And what are the advantages? Scientifically, it's not any better than the photoelectric effect, because either observation suffices to prove that light is quantized. Historically, it's also no better than the photoelectric effect, because Planck didn't interpret his calculation in terms of quantization, and the success of his calculations was no more successful than the photoelectric effect in convincing people like Bohr.

 

[bcrowell]

I still find this approach of treating the photoelectric effect as being the interaction of a photon with an atom to be rather naive. The photoelectric effect is done on metals, and the collective effect of the metal to produce the conduction band seems to have been ignored in here. The photoelectrons in a standard photoelectric effect comes from this conduction band, not from a particular atom of the metal.

Zz.

Maybe I'm missing your point, but I'm not sure that their model even really assumes atoms. They do use the *word* "atoms," but all they assume is the following: (1) each electron is independent of the others; (2) each electron has available to it a ground state plus a quasicontinuum (the latter representing the vacuum, not the metal). I don't know if it matters whether you consider the ground state to be a state localized to an atom or a state in the conduction band.

I'm actually confused by your statement that the PE effect comes from the conduction band. The PE effect can't occur for an electron in a vacuum, because it would violate conservation of mass-energy. (In the center of mass frame, the final mass-energy would be less than the initial mass-energy.) I've always understood this as the essential reason for the sharp Z-dependence of the photoelectric effect. Z=0 would be a vacuum. The fact that it's proprtional to a power of Z, and not of a screened charge, would suggest to me that it's a phenomenon that preferentially affects the innermost electron shell.

In the specific case of metals, we observe that the ability to absorb gammas via the PE effect is highly correlated with Z, so, e.g., we use lead as shielding, not aluminum.

Is there perhaps some aspect of this situation that is different for visible light photons than for gammas? My experience is with gammas.

 

[atyy]

Lamb and Scully produced a model that violated conservation of energy.

Yes, but if this is an objection to their model, wouldn't it also be an objection to all time-dependent Hamiltonians in quantum mechanics?

 

[ZapperZ]

I'm actually confused by your statement that the PE effect comes from the conduction band. The PE effect can't occur for an electron in a vacuum, because it would violate conservation of mass-energy. (In the center of mass frame, the final mass-energy would be less than the initial mass-energy.) I've always understood this as the essential reason for the sharp Z-dependence of the photoelectric effect. Z=0 would be a vacuum. The fact that it's proprtional to a power of Z, and not of a screened charge, would suggest to me that it's a phenomenon that preferentially affects the innermost electron shell.

Actually, the reason why PE cannot occur for "free electrons" is the conservation of momentum. For conduction electrons, the presence of the lattice ions solved that problem. However, it also means that, by default, these are not truly "free electrons", i.e. the free-electron approximation in the Drude model would not do if want these conduction electrons to couple (even weakly) to the lattice ions or their potential.

It certainly is not the innermost electrons, because we know what that looks like when we do Xray photoemission, i.e. we will then do what we call core-level photoemission. This looks nothing like, say VUV photoemission that probes roughly the first eV below the Fermi energy of the material. And certainly the STANDARD photoelectric effect uses even lower-energy photons than this. So it can't probe any deeper than a few meV below the Fermi energy, i.e. the conduction band.

Zz.

 

[conway]

The title of the paper is unfortunate, as it lends itself to misunderstanding by those who want to deny the quantum nature of light altogether. So it's a very good thing that the paper explicitly shows this to be invalid, and that the author recognizes the well established quantum nature of light.

Again, the authors do nothing of the sort. They do not "explicitly show" the wave model to be invalid for the Compton effect, spontaneous emission, etc; they merely declare it to be invalid without giving reasons.

I found helderverez's speculations as to why they would do this to be as reasonable as any other suggestion; his posslible lack of correct nuance can probably be excused on account of Englsih not being his first language. But surely in any field there is a certain "political correctness" that people don't want to step too far outside, and the authors' nod to SQM can easily be understood as a polite nod to conventional thinking. It is clearly not backed up by any explicit reasoning.

 

[Physics Monkey]

I think Lamb's point is that while the (naive) photoelectric effect is quantum, the quantumness could be in either the material or light.

I also suspect this may have been Lamb's point, or at least part of it.

What's one step up from the Drude model?

Haha, according to Ashcroft/Mermin it's the Sommerfeld Model. Sommerfeld is Fermi statistics but no lattice. For some aspects of the real photoelectric effect one needs to include the lattice or something else to suck up momentum, but that is perhaps more detail than I need at the moment.

Along the lines of your comment, I think that if I mostly declare ignorance about the metal, as Einstein did, and just approximate it as some slushy thing where quantized energy levels aren't too important, then one might make a stronger case for the quantization of the electromagnetic field from the photoelectric effect.

 

[sylas]

The title of the paper is unfortunate, as it lends itself to misunderstanding by those who want to deny the quantum nature of light altogether. So it's a very good thing that the paper explicitly shows this to be invalid, and that the author recognizes the well established quantum nature of light.

Again, the authors do nothing of the sort. They do not "explicitly show" the wave model to be invalid for the Compton effect, spontaneous emission, etc; they merely declare it to be invalid without giving reasons.

What I mean is that passage given shows explicitly that the author does recognize the quantum nature of light. Of course it isn't trying to prove all the background. But this paragraph does show it is invalid to cite this work as denying the quantum nature of light, and that's useful context.

Equivalently, the paragraph makes explicit what the paper is not showing. The paper is not showing that the semi-classical treatment works in all cases. It is not showing evidence against the quantum nature of light.

It would be redundant to "show" that the classical or semi-classical wave treatment is incomplete. That would be repeating a whole related line of work! It is not redundant at all to state relevant knowledge as context, in a paper that could otherwise be misrepresented as refuting or denying the quantum nature of light.

Thinking that light is adequately represented with a semi-classical model for addressing all aspects of the phenoma listed is flatly wrong -- and ironically heldervelez' own selected extracts are explicit on this with respect to the Compton effect at least!

I found helderverez's speculations as to why they would do this to be as reasonable as any other suggestion; his posslible lack of correct nuance can probably be excused on account of Englsih not being his first language. But surely in any field there is a certain "political correctness" that people don't want to step too far outside, and the authors' nod to SQM can easily be understood as a polite nod to conventional thinking. It is clearly not backed up by any explicit reasoning.

True, the English issue means I should allow for unintended nuance. Suffice to say; there's no reason to speculate that the author of the paper did not consider the paragraph as well established and valid relevant context.

I certainly think it is well established relevant context, and this thread has made crystal clear why context like this is so important. I often wish more papers took better care to clarify such things, given the ways in which mavericks are sometimes inclined to distort published work as support for absurd claims.

Cheers -- sylas

 

[Physics Monkey]

Scientifically, I don't think the statement is too naive at all. Lamb and Scully produced a model that violated conservation of energy. If a classical flash of light comes along with an energy equal to 1.5W, where W is the work-function of the metal, then conservation of energy dictates that the number of electrons ejected through the surface is either 1 or 0. Lamb and Scully's model says that with quite high probability, the number of electrons ejected will be 2 or more. Since nobody has been able to produce a model that simultaneously (1) has a classical electromagnetic field, and (2) satisfies conservation of energy, I think it is absolutely fair to say that the photoelectric effect suffices to prove the quantized nature of light.

Since I don't know what Lamb would say to your criticism I will just make a guess. I think that your setup is somewhat artificial in the sense that the light beam is seriously modified (i.e. totally absorbed) in the electronic emission process. In the photon model, your light beam would carry a few photons assuming it was at the relevant frequency. I think this is not the situation Lamb intended to address. He was trying to model what I assume was the experimentally relevant (at the time) situation of a semiclassical light beam containing many photons.

Of course, I agree that energy is ultimately conserved. But I like to ask myself how important it is to keep track of this for the decription of the experiment in question. I think Lamb regards it as relatively unimportant in this case.

Educationally, what I really abhor is the textbooks that introduce the photon via blackbody radiation. It's a horrible mess to try to get that across to students. And what are the advantages? Scientifically, it's not any better than the photoelectric effect, because either observation suffices to prove that light is quantized. Historically, it's also no better than the photoelectric effect, because Planck didn't interpret his calculation in terms of quantization, and the success of his calculations was no more successful than the photoelectric effect in convincing people like Bohr.

My personal opinion is that it's bad to teach students that one experiment proves or disproves something. I think we know by now that this just isn't how science works. It may nevertheless be how science is taught. I appreciate your frustration with black body radiation, so perhaps a more balanced presentation of multiple experiments is better?

 

[bcrowell]

Actually, the reason why PE cannot occur for "free electrons" is the conservation of momentum.

Well, not to belabor a point, but that depends on the frame. I had in mind the center of mass frame. Of course the real issue is not whether you can conserve p or E but whether you can conserve both.

Anyway, my main point was that I just don't see anything in their model that really has to be interpreted as an atomic orbital, rather than as a conduction band.

Yes, but if this is an objection to their model, wouldn't it also be an objection to all time-dependent Hamiltonians in quantum mechanics?

The question is whether the energy-nonconservation caused by the time-dependence is an incidental feature that we know we could clean up with a more careful treatment, or a fundamental flaw in the approach. Here it's a fundamental flaw in the approach. The real reason that two different atoms can't be ionized by the same photon is that the photon has quantum-mechanical correlations. In the Copenhagen picture, detecting the photon at atom A collapses the wavefunction, so it can't be detected at atom B. There is no way to fix up their model so that it avoids this problem. It's not an incidental issue, it's fundamental. Wavefunction collapse (or entanglement, in pictures other than the Copenhagen one) is a definitive characteristic of what it means for a field to be quantized.

To see the extreme naturalness of raising this issue, take a look at the book that you pointed us to, George Greenstein and Arthur Zajonc, "The quantum challenge: modern research on the foundations of quantum mechanics," Jones and Bartlett, 2005. Immediately after presenting a summary of the Lamb-Scully argument, they talk about quantum-mechanical correlations as the smoking gun that proves the electromagnetic field is quantized.

I think that your setup is somewhat artificial in the sense that the light beam is seriously modified (i.e. totally absorbed) in the electronic emission process. In the photon model, your light beam would carry a few photons assuming it was at the relevant frequency. I think this is not the situation Lamb intended to address. He was trying to model what I assume was the experimentally relevant (at the time) situation of a semiclassical light beam containing many photons.

I don't think we can really say that Lamb was trying to mock up the possible description of the photoelectric effect "at the time," if "at the time" means 1905; in 1905 there was no quantum-mechanical model of the atom. Lamb is specifically trying to address the question of whether this experiment probes the quantum or classical nature of the electromagnetic field. If he's saying that it can be explained with a classical EM field, then it's absolutely natural to ask whether the theory gives reasonable results in the limit of low-energy fields. A classical field is supposed to be okay when you take the low-energy limit, because it's not granular. If you get nonsense answers in the low-energy limit, that tells you that you're not succeeding with a description in terms of a classical field.

My personal opinion is that it's bad to teach students that one experiment proves or disproves something. I think we know by now that this just isn't how science works

I would distinguish between proving and disproving a theory. You can never really prove a theory with any finite number of experiments. But you can certainly disprove a theory with one experiment. E.g., the Rutherford alpha-scattering experiment conclusively disproved the raisin-cookie model of the atom.

 

[conway]

To see the extreme naturalness of raising this issue, take a look at the book that you pointed us to, George Greenstein and Arthur Zajonc, "The quantum challenge: modern research on the foundations of quantum mechanics," Jones and Bartlett, 2005. Immediately after presenting a summary of the Lamb-Scully argument, they talk about quantum-mechanical correlations as the smoking gun that proves the electromagnetic field is quantized.

They also say that the Compton effect can be explained without photons.

I'm going to speculate that any experiments involving thermal light can be explained equally well by the photon model or by the semi-classical approach, and that all the remaining sticking points ultimately boil down to some version of the anti-bunching phenomena as seen in so-called non-classical light.

 

[heldervelez]

quantization results of electronic orbits beeing resonances (*), which is why radiation is emitted or absorbed in quanta not because that is the nature nature of light.
I think that it does not exist (or devise) any experiment where the resonant nature of emitter and absorver are absent. This make the issue very hard to solve.
What is in the middle, photons, have also a definite amount of energy, but IMO it is only field (analog) not a particle.
How can energy be localized, bounded, without coherence, resonance?
IMO all particles are field.
The closer we 'look' to an electron the smaller is its radius. Is it there a 'particle' at all?
QM is good to deal with the results of the observations but not enough to grasp under the skin.
The quantization of the reality (time, space, field ,(gravity,static field,...)) gives me deep philosophical concerns. An exploration of other views (change of paradigm) may bring us a model where masses are derived (I'm reading Douglas Pinnow) instead of postulated.
Why QM can not explain the influence of light polarization on the direction of emited electrons?

(*) standing waves, bounded by 'c', geometry and energy give a limited set of possible states

 

[ZapperZ]

quantization results of electronic orbits beeing resonances (*), which is why radiation is emitted or absorbed in quanta not because that is the nature nature of light.
I think that it does not exist (or devise) any experiment where the resonant nature of emitter and absorver are absent. This make the issue very hard to solve.

Again, what is "quantized" for the conduction BAND?! This is a continuum energy states!

The rest of your post continues to perpetuate the faulty idea that the photoelectric effect is done on isolated atoms, which it is not! Photoionization is not the same as the photoelectric effect. If you truly think that it is the material that's responsible, then pay attention to the physics of the material!

Zz.

 

[Physics Monkey]

Lamb is specifically trying to address the question of whether this experiment probes the quantum or classical nature of the electromagnetic field.

I agree with you here, this does seem to be part of Lamb's point. My point is that the experiments Lamb is trying to explain are not of the type you describe. The body of experiments Lamb is referring to used, to my knowledge, semiclassical beams with many photons. So Lamb doesn't need to have a description of the low energy situation to discuss the experiments.

If he's saying that it can be explained with a classical EM field, then it's absolutely natural to ask whether the theory gives reasonable results in the limit of low-energy fields. A classical field is supposed to be okay when you take the low-energy limit, because it's not granular. If you get nonsense answers in the low-energy limit, that tells you that you're not succeeding with a description in terms of a classical field.

I disagree with you here. It's not necessarily natural to ask what happens in a certain limit if your model is assuming you aren't in that limit. Perhaps we can say Lamb's paper is unclear, since his assumptions are not clearly stated. However, the assumption that the beam is large and classical is visible. For example, supposing Lamb were trying to do everything classically even for small fields, he would at least have needed to include the classical dynamics of the field to have a chance of describing the extreme situation you are interested in. I think this makes it plausible that he was intentionally neglecting certain things, and hence he would agree that you can't push his theory too far. He's not saying light isn't quantized, or that some extreme version of the photoelectric effect wouldn't clearly show this, only that the experiments with large beams can be modeled without invoking the quantized field.

On the other hand, I think your statement is a good thought experiment which pushes us towards a more complete understanding. If one did these experiments and had Lamb's description in hand but with no knowledge of photons, then a natural direction for further experimental exploration would be decreasing the energy of the light beam. One might then discover photons.

I would distinguish between proving and disproving a theory. You can never really prove a theory with any finite number of experiments. But you can certainly disprove a theory with one experiment. E.g., the Rutherford alpha-scattering experiment conclusively disproved the raisin-cookie model of the atom.

I roughly agree here. With sufficient background information and context, the Rutherford experiment is clean enough to strongly disfavor certain ideas. My concern is that students often lack the relevant background and context to make such a determination. If you've never done the classical mechanics calculations for different types of scattering, if you've never thought hard about the assumptions and mechanism underlying scattering experiments, etc then just being told that Rutherford conclusively proved this or that isn't necessarily very convincing.

 

[Cthugha]

I'm going to speculate that any experiments involving thermal light can be explained equally well by the photon model or by the semi-classical approach, and that all the remaining sticking points ultimately boil down to some version of the anti-bunching phenomena as seen in so-called non-classical light.

I do not think it is that easy. I am not aware of any sensible classical or semiclassical explanation for single mode photon bunching.

 

[conway]

The normal explanation for bunching as in e.g Hanbury-Brown and Twiss is that a "photon" is more likely to be detected when there is a thermal fluctuation in the field, therefore the detection event also tends to be associated with higher fields hence more "photons". Once you get rid of thermal fluctuations, the bunching disappears e.g. laser light with pure Poisson detection statistics. These are all what you'd expect for classical light. I know the antibunching is problematic but I don't exactly know what you mean by single mode bunching.

 

[Cthugha]

Yes, this classical description works for most cases, but it gets nontrivial if you have a single thermal mode and not a superposition of modes and even worse if you consider complicated experimental situations like ghost imaging. I do not intend to turn the discussion around, but "Can Two-Photon Correlation of Chaotic Light Be Considered as Correlation of Intensity Fluctuations?" (PRL 96, 063602 (2006)) by G. Scarcelli, V. Berardi and Y. Shih (and the comments and replies) give a good overview of where it gets problematic, if you are interested in that topic.

 

[harrylin]

[..]
If you read the Lamb-Scully paper, the first thing you notice is that they explicitly state that photons are absolutely necessary in order to explain phenomena such as blackbody radiation, Compton scattering, spontaneous emission, and the Lamb shift. Any internet kooks who are trying to quote Lamb as an authority against quantization of light are way off base. [..]

That can be misleading, as you here refer to an old publication. Arnold Neumaier quotes Lamb as a Nobel prize winner who is against photons:

"the photoelectric effect does not
require the quantization of the radiation field,
a misconception perpetuated by the mills of textbooks".
Some people, such as the Nobel prize winner
Willis Lamb 1995 even take this as an indicator
that "there is no such thing as a photon"."
- http://arnold-neumaier.at/ms/lightslides.pdf

That quote is from a more recent paper:

W.E Lamb, Jr., Anti-Photon, Applied Physics B 60 (1995),
77{84. Reprinted in: W.E Lamb, Jr., The interpretation of
quantum mechanics, Rinton Press, Princeton 2001.

In that paper Lamb argues against the idea that radiation (I take him to mean EM radiation) consists of particles. As I understand it after a quick glance, he now argues against the use of the word "photon" (in contrast with the older paper that you cited) because it is often associated with the particle interpretation of "photon".

The Abstract:

"It should be apparent from the title of this article that the author does not like the use of the word "photon", which dates from 1926. In his view, there is no such thing as a photon. Only a comedy of errors and historical accidents led to its popularity among physicists and optical scientists. I admit that the word is short and convenient. Its use is also habit forming. Similarly, one might find it convenient to speak of the "aether" or "vacuum" to stand for empty space, even if no such thing existed. There are very good substitute words for "photon", (e.g., "radiation" or "light"), and for "photonics" (e.g., "optics" or "quantum optics"). Similar objections are possible to use of the word "phonon", which dates from 1932. Objects like electrons, neutrinos of finite rest mass, or helium atoms can, under suitable conditions, be considered to be particles, since their theories then have viable non-relativistic and non-quantum limits. This paper outlines the main features of the quantum theory of radiation and indicates how they can be used to treat problems in quantum optics."

Cheers,
Harald