One Time Only: Scientist-on-Scientist TSA Smackdown
Yesterday I quoted a physics professor on why he would never willingly go through the new "enhanced" backscatter-radiation TSA screening machines. Summary: "There is no such thing as a risk-free dose of ionizing radiation."
I heard last night from a reader who shares the physicist's skepticism of the TSA but challenges his explanation of the underlying science. After the jump, because they're so long, are this second reader's message, and then a response from the original physics professor.
This is not the kickoff to further rounds of back-and-forth from these correspondents or others about the scientific principles at stake. I'm uncomfortable hosting extended discussions on topics where I can't independently tell plausible from implausible arguments. But I thought one round was interesting, and I encourage those with the technical standing to pursue the analyses elsewhere.
Before you start, here's an account from another reader, one who has decide to opt out of opting-out. This reader (also, to judge from his email address, a computer-science professional) reports:
>>I tried opting out twice recently.
In the first place (Chicago O'Hare airport, I think), the gentleman from the TSA was very nice and asked me whether I was afraid of the radiation, and said it's much lower radiation than X-rays and so on. I had read about some shouting and hollering to shame opt-outers, but I was a bit nonplussed by this kind approach... So I went thru the scanner. :)
Two weeks ago, I was at the San Diego airport and had plenty of time on my hands, so I opted out again. They made me wait a little longer but then a gentleman from TSA came up and did a fairly thorough pat-down the same way as your previous correspondent wrote -- told me he was using the back of his hand as he came up to a sensitive area, checked the belt area around the waist, etc.
I had the distinct impression that the pat-down caused more discomfort to him than it did to me. Which is how it should be. I'd be very worried if the TSA folks started enjoying such a task or took it to be routine.
While the risks of radiation seem to be high, I think the way these pat-downs are done is very demeaning not just for the traveler but also for the TSA employees. So I'll opt out of opting out for now...<<
Here's the letter of complaint from a reader I've corresponded with frequently, and who explains his scientific background below. I have edited this slightly to remove some (but not all) of the ad-hominem remarks about the original physics professor -- and in the physics professor's reply, I've removed some understandable ad-hominem in response:
>>I was, um, off-put by the science in what you were emailed even though I agree with a lot of the conclusions -- just the route there was _terrible_ science.... I don't mean to be defending the wacky security theater, just hating some pseudo-science...
1) Take it from a guy who's been around a lot of tritium label: there's all kinds of more or less risk-free ionizing radiation. I mean, this guy ... thinks this is all an issue of wavelength. It ain't. There's issues of penetration depth etc. I mean, don't _eat_ that tritium....
2) Seriously, you haven't heard of multiphoton emission/absorption? And it's irrelevant anyway...but you even see it in photoelectrons if you look right.
3) Microwave energies can cause beaucoup molecular change; hell, I've used (um, well, I told my undergrad to use) a microwave source in an organic synthesis, so I'm real sure the emailer doesn't know what he's talking about. I mean, it can't _ionize_ (easily). Big whoop; most of biology is acid/base chemistry.
4) Which brings us to DNA. No, microwaves won't cause the kind of cytidine dimers etc. associated with UV (without a lot of power and just the right resonance). But there is much DNA potential mutation caused by say deamidations, which local heating or bond stretching can cause, and of course single- or double-strand breaks are simple nucleophile-meets-electrophile-with-a-curly-arrow chemistry. Now I think the jury is well out on cell phones but I'm pretty skeptical -- but your emailer is just plain misusing science in his rule-out... "Period." Plus of course there are arguments about stochastic resonance and transport which don't, I think, bear out in experiment but which are certainly theoretically possible, and by the way the argument is generally about changes not in DNA structure but expression. These just aren't arguments theoretical optics is suited to handle.
5) The scanner isn't a hard X-ray scanner so the entire second portion of "things we can conclude" is loopy.
I could go further, but, this isn't a physicist I'd take a lot from, actually. At least, not any biophysics and probably not any optics... Who you're taking _this_ from is a biophysicist and enzymologist who did his Ph.D. thesis in a cyclotron and who has overseen diagnostic sections in a hospital and attended Biophysical Society.... [Some comments here about the software engineer's complaint -- for another time.]
I don't mean to be defending the scanners, which I don't understand to be really useful. There are interesting possible failures which are damaging to health and maybe there are holes in the studies which assess health effects in non-failed systems, although those are going to have to be determined by experiment because they are by no means predetermined by theory....<<
To which the physics professor replies:
>>Not wishing to make this a bigger thing than it is, but some things below are simply not the case."1) Take it from a guy who's been around a lot of tritium label: there's all kinds of more or less risk-free ionizing radiation."Tritium is a beta emitter. Beta radiation is not x-rays (or any other type of photon). Beta radiation is electrons. With x-rays, we are dealing with electromagnetic radiation and that is a completely different issue.
For one thing, electrons have an electric charge and photons do not. That means electrons scatter very strongly off other charged particles, most notably other electrons in the atoms they encounter. Photons are uncharged so, while they do scatter off of charged particles, they do not do so in the same way. You can (very loosely) think of electrons as particles bouncing off of other particles, but it is better to think of photon scattering as the original photon being absorbed and a new photon emitted.
But even for a beta emitter, tritium is extremely low energy. It only releases 18.6 keV total in the decay process, with an average of only about 6 keV carried by the electron. The low energy of the emitted electron makes it difficult to detect by most radiation monitors. Mostly, you have to use liquid scintillators to even find it at all.
This is the entire reason why it is only dangerous when ingested -- the energy of the radiation is simply too low to cause significant damage unless it is literally sitting inside the tissues themselves. Tritium-produced electrons can travel about 6 mm in air and won't penetrate the dead outermost layer of skin at all.
I believe I pointed out the penetration depth issue with regard to UV radiation. But this guy is confusing electrons with photons and they interact with matter in very different ways. In any case, the penetration depth of photons depends on their wavelength, so as long as we're talking about x-rays and microwaves yes, it is all a matter of wavelength. Including beta radiation in the mix is deceptive, to say the least, since that is not a stream of photons and its penetration depth is limited by properties that photons do not have."2) Seriously, you haven't heard of multiphoton emission/absorption? And it's irrelevant anyway...but you even see it in photoelectrons if you look right."To compute absorption/emission probabilities in quantum mechanics, essentially you have to compute the matrix elements of the electromagnetic hamiltonian operator between atomic stationary states. You can see the details of the calculation many places. One easy place to get there on line is here, beginning about page 150.
Here's the important point: there is an exponential operator in that matrix element that can't be dealt with exactly, so it is dealt with perturbatively, by expanding it in a Taylor series. If you simply keep the lowest order term in that series (i.e. you approximate the exponential by unity), that is called an electric dipole approximation. It corresponds physically to the fact that the photons absorbed or emitted by atomic transitions have wavelengths that are very much larger than the atoms they are emitted from or absorbed into. Thus, to a good approximation, the electric field of the photon is almost constant across the space occupied by the atom.
It may sound unlikely, but that turns out to account for pretty much all absorption/emission processes under ordinary circumstances. The next most likely absorption process is a magnetic dipole transition caused by interaction of the electron spin with the oscillating magnetic field of the photon.
Magnetic dipole transition probabilities are about 100,000 times less likely than electric dipole probabilities. Keeping the next order term in the series expansion of the exponential results in electric quadrupole transitions. Electric quadrupole transitions have a probability on the order of 10 million times smaller than electric dipole transitions. The probabilities get correspondingly lower with all higher order transitions.
So there's a few important points to be learned from this:
1. Under the circumstances that prevail in the human body, or any situations with which a typical human body is going to come into close contact, the only thing that matters are electric dipole transitions. These can result in changes of atomic angular momentum of either 0 units or 1 unit. Since photons are spin 1 particles, this means there can be two-photon emission and absorption processes. Two oppositely oriented photons spins, for example, would produce 0 net angular momentum change in the source atom. [JF Note; Preceding two sentences slightly edited from original version, at author's request.] Two photon transitions tend to be lower probability unless there is an intermediate resonant state in the atom. Quadrupole transitions can change angular momentum by as much as 2 units and can easily accommodate multiphoton processes, but we've already seen that they have vanishing small probability. That's the reason why you have to look very hard to see multiphoton processes. Your physician friend seems to be operating under the assumption that all processes that can happen will happen at the same level of probability. That is, to put it kindly, childish.
In any case, even allowing the possibility of two photon transitions, microwave photons have a frequency of around 100 GHz which corresponds to an energy of about half a milli-eV. X-ray photons have frequencies around 10^17 Hz (or about 100 million GHz) for an energy on the order of a thousand eV. Consequently, double microwave photon absorption doesn't even come close to the effect of a single x-ray photon. Your correspondent should know that and his argument is a red herring. He may be aware that it is not relevant, and I am certainly aware that it is not relevant, so why include it as a point of discussion?
2. When we include thermodynamic effects, the probability is almost unity that if an atom is in a state that can transition to a lower state only by emitting more than one photon, it is far more likely to lose energy by collision with another atom than by multiphoton emission. And similarly, if an atom absorbs a photon, it is overwhelming likely to lose that energy by collision or reemission before it can absorb a second photon. In fact, until the late 80's, even the hardest achievable laboratory vacuum was unable to produce quadrupole transitions before the atoms were collisionally deexcited. There are some rare earth lasers that employ quadrupole transitions but they involve a unique crystalline structure that prevents collisional deexcitation. Needless to say, people aren't built that way.
3. With regard to the photoelectric effect, your doctor friend basically doesn't know what he's talking about. Increasing the number of photons in the incident light increases the rate at which photoelectrons are emitted but it does not increase their energy. There exists a minimum frequency below which no photoelectrons will be emitted no matter how many photons you bombard the material with. On the other hand, increasing the frequency of the incident beam does increase the energy of emitted photoelectrons. This much was known by experiment in 1900 and is covered in chapter 1 of every quantum textbook on Earth. Increasing the intensity of an incident beam will *eventually* increase the energy of photoelectrons through sequential multiphoton absorption, and will *eventually* allow below threshold photoelectron emission for the same reason, but that doesn't occur until intensities on the order of several megawatts per square centimeter are reached (see, e.g., A. T. Georges, Physical Review A, volume 66, issue 6, where a laser intensity of 10 MW/cm^2 was needed to make it happen). Again, your friend seems to be unaware of the extreme difference in probabilities between single and multiple photon events. And again he's appealing to phenomena that simply don't apply here. Basically, he's picking nits by appealing to phenomena of marginal importance."3) Microwave energies can cause beaucoup molecular change; hell, I've used (um, well, I told my undergrad to use) a microwave source in an organic synthesis, so I'm real sure the emailer doesn't know what he's talking about. I mean, it can't _ionize_ (easily). Big whoop; most of biology is acid/base chemistry."So does he know what the microwave source actually did? Reaction rates frequently have a strong temperature dependence and a microwave source could give you very fine control over temperature.
However, in that case, the fact that it is a microwave source is inessential. You could (and in the days before microwave sources, often did) achieve the same effect with a simple flame.
Microwaves do not ionize. It is not a matter of whether they do it easily or not. They do not do it at all (unless you've gone to the trouble of specially preparing your sample in an extremely excited state and protected it from collisional deexcitation -- basically, Rydberg atoms in a magneto-optical trap -- and if that is the basis of your argument, it qualifies as pathological).
And yes, when we are discussing radiation induced mutagenic processes, that is a very big whoop. The fact that most of biology is acid/base chemistry is totally irrelevant to the interaction of radiation with matter. Those are independent phenomena. You can have either, or both, or neither.
So this statement is partly wrong, and the part that isn't wrong is irrelevant."4) Which brings us to DNA. No, microwaves won't cause the kind of cytidine dimers etc. associated with UV (without a lot of power and just the right resonance). But there is much DNA potential mutation caused by say deamidations, which local heating or bond stretching can cause, and of course single- or double-strand breaks are simple nucleophile-meets-electrophile-with-a-curly-arrow chemistry. Now I think the jury is well out on cell phones but I'm pretty skeptical -- but your emailer is just plain misusing science in his rule-out: "Period." Plus of course there are arguments about stochastic resonance and transport which don't, I think, bear out in experiment but which are certainly theoretically possible, and by the way the argument is generally about changes not in DNA structure but expression. These just aren't arguments theoretical optics is suited to handle."Basically, what I get from all this is that I'm right but he's going to criticize me anyway.
Listen. Conservation of energy is a fundamental law of the universe. If it requires more energy to initiate a process than is present in the radiation doing the initiating, then the process won't happen. The jury is no more out on this than it is on power lines causing cancer. If you believe you have a mechanism by which low energy photons can conspire to produce high energy effects, then state it and state your evidence that it happens. He's arguing for the unusual event that seems to violate established principles of physics -- the burden of proof is on him.
I'll concede he may have a small point here since I don't know enough about genotype becomes phenotype to say, but unless he can provide a specific instance of a reaction that is initiated by microwave photons and also account for how those microwave photons penetrate to the interior of the body, he's simply blowing smoke. I assume that by "DNA expression" he is referring to something like epigenetic processes. But epigenetic effects are also controlled by molecular bonding, for example chromatin remodeling through the attachment of methyl groups to the DNA molecule. That, to me, seems like a distinction without a difference. Whether we're talking about photon induced modification to the molecular structure of DNA itself, or to the molecular structure of the things attached to DNA that turn genetic sequences on or off is immaterial. You're still talking about photon induced modification of a molecule and that simply won't happen with significant probability if the energy of the photon is significantly less than the energy required by the modification."5) The scanner isn't a hard X-ray scanner so the entire second portion of "things we can conclude" is loopy."That's just stupid.
Soft or hard, any x-ray photons have higher energy than any UV photons and UV photons are well known to cause cancer. Soft x-rays do have small penetration depths, but they penetrate further than UV. That's why TSA uses them. Unless he knows of some special kind of x-ray photon with the energy of a microwave photon, this statement is simply ignorant. And it is ignorant at the level of undergraduate physics.
So let's summarize:
1. This guy seems to think that magnetic dipole and higher order processes occur with the same or similar probability as electric dipole processes. They don't. Under anything remotely resembling ordinary circumstances (i.e. not in a vacuum and not under a multimegawatt photon source, which I assume describes airports pretty well), electric dipole emission and absorption is the ONLY important process. That is at best two photon emission/absorption, and half an x-ray photon is not a microwave photon.
2. He doesn't seem to understand that electrically charged particles interact with matter very differently from electrically neutral particles and treats them as being basically the same thing.
3. Let me put this in simple terms. Einstein showed that electromagnetic radiation behaves like tiny lumps of energy called photons, and that the energy of each photon is determined solely by its frequency. They don't cause any trouble unless their *individual* energies match some natural excitation or resonance. Two photon effects are much lower probability and more than two photon effects basically don't exist under ordinary circumstances. Collective action only occurs in unions, not photons. There isn't much to excite in molecules until you reach the energy of typical molecular vibrations and that is in the microwave frequency domain. This may provide sufficient heat to change a reaction rate, but so would a bunsen burner. That is a function of the heat only, and not the thing that produced the heat. Any reaction that could be produced from the heat of a milliwatt cell phone could equally well be produced by living through an Atlanta summer. Heat is heat. Photons do not begin to eject photoelectrons until you reach the extreme end of the blue part of the spectrum and into the near-UV. Ejecting a photoelectron is something that a bunsen burner won't do*. That is the minimum energy at which a photon can create a mutant strand of DNA. I may be wrong, but as far as I know all known cancer agents create mutant strands of DNA. That was certainly the argument made by Robert Park ten years ago (Journal of the National Cancer Institute, volume 93, issue 3 p166, 2001). And he knows more about it than I do. And even if we allow two photon processes, half the energy of a near UV photon, is in the center of the visible spectrum, not in the microwave. Two microwave photons add up to, at best, the energy of an infrared photon, not an x-ray.
[*Footnote: Well, thermal electron ejection does occur but to make that go the target material has to be heated faster than the rate at which heat is lost to the environment, and that simply isn't going to happen at cell phone radiation levels.]
Most of this is what most physics majors learn in their junior or senior level quantum mechanics course. I'm sure there is a great deal of sophistication in the biological details, but he seems to have gotten lost in those details. The fact remains that regardless of the details, a process that requires energy to go will not go unless that amount of energy is present.
My stepson is in the third year of his residency, aiming to be an orthopedic surgeon, just like his father and his grandfather. Because of this, I've had occasion to encounter a rather large number of physicians and physicians-to-be over the last decade or so. I've found that, by and large, their views on processes of science are myopic and instances of path learning, that they know a great deal less about fundamental principles than they think they know, and that they are quite dogmatic about all the things they thing they know that are not true. Now this does certainly not characterize every medical professional, but the population balance in my experience is clearly shifted in an unfavorable direction....
I've met over the years quite a few Ph.D.'s who managed to cram one idea into their heads, usually some sort of experimental procedure (theory is hard and you can't lose yourself in the technology). They then either got out with a sigh of relief and bolted for the simpler world of a professional school, or they made a career out of doing the same experiment over and over, like a ride at Disneyland. Having a Ph. D. doesn't mean you're smart. It doesn't even mean that you know very much. That's true for me as much as it is for whoever this guy is. The fact that either of us has Ph.D. does not of itself mean that much.
So good luck deciding who is right. I dunno how to advise you on that. I do know that he has conflated charged particles with uncharged particles, included a large number of points that are of marginal relevance or none at all, and argued based on low probability events without revealing to you that they are low probability events....
He's said a lot of "Hey you forgot about this thing even though it really doesn't matter here." Unfortunately, this happens quite often in the world of physics. The need to demonstrate that "I know everything you know and more" is one of the less savory aspects of the community and has caused me to bug out of more than one meeting. Part of the problem, I think, is that he has let his umbrage lead him into the pitfall of trying to disprove all of my points, regardless of the merits, and the requirements of arguing that x-rays may be safe but microwaves may be dangerous, while appealingly counterintuitive, is also intrinsically incoherent. Carving out an exception to energy conservation for one case makes it difficult to carve out a consistent exception for the other.
Now I've spent far more time on this than I really should have. I'm going to have to leave it there.<<
I will leave it there too, with appreciation to readers for taking such time and care to explain these points. I had good physics teachers in both high school and college -- and while they might be disappointed by my inability to referee this discussion, they shared a trait with the professor I'm quoting, in finding clear ways to explain inherently tricky concepts.