Gustav Hertz (1968) - Memories of James Franck and the Electron Scattering Experiments (German Presentation)

Ladies and gentlemen, First of all, I don’t know if you can hear anything. There is a switch here and none of the experimentalists present was able to tell me whether it should be switched to the right or to the left. Can you hear anything? On the red button. There it is. So I am hoping that you can hear something. Now, Herr Weisedel was kind enough to say that I have been here often. This is the fifth time. I have already had many opportunities to speak here and it is a special pleasure for me to be able to do that once more today. The first times that I was here it was especially stimulating to again meet many old friends of my generation, some of whom I had not seen for many decades. Unfortunately our generation has in the meantime shrunk more and more, and I myself unfortunately And since this year Herr Hahn and Herr Born were unable to come, to the regret of all of us, I am in the remarkable position of being the oldest of the Nobel Prize winners present. Consequently, I am of course in no position to report on new research, but must rather look back at the past and I would like to use the opportunity to keep alive the memory of my dear friend Franck, who we still could see in our circle here six years ago, by talking about our old experiments together. But when Herr Weisedel just now referred to Herr Franck as my employee, the matter is really the other way around. In those days I was - of course working together is naturally something mutual, but Franck was originally the elder; I was a young doctoral candidate when I had the great luck to meet him and to be invited by him to work with him. Perhaps I may show you a photograph of him which I myself managed to take a few weeks before his death. It is however rather hard to see. That is now the old Franck as he was shortly before his death. When we did our experiments he was 30 and I was 25. Now it is really extremely difficult to think oneself back into those days. And if I want to report to you about the experiments, I will tell a very simple story. And I am not doing it to tell you what we did, because it is extremely simple and primitive seen from the present, and the interpretation is clear to everybody today, but contrariwise to show how difficult it was to recognize the facts of the matter and in particular the mistakes we made then which, as we can see, were numerous. This is what I want to go into here. Naturally, the whole situation was completely different then. A physicist mainly worked with his hands. I mean also with the head a bit, but most of the day it was real manual work. The apparatus had to be stuck together. I mean that certain pieces of apparatus were built correctly by mechanics, but in our area sealing wax and mercury were the main things, and you had to know how to handle them. And when I think back to the early days in the Physics Institute on the banks of the Reichstag, then I envisage Franck the way he stood by the glass bulbs, this cobbled-together apparatus. He stood on one leg, and with the other leg he trod on a bellows for this torch flame. In his right hand he had the flame, in his left hand he had a piece of glass pipe, and in his mouth a tube for blowing. And that is how the apparatus was put together. And it was really the case that a good measure of manual skill was required there and a large part of the time was really occupied with manual work and also with direct observation. One saw directly, one saw the spectral lines in a spectroscope or one saw the appearance of the gas discharge. One measured curves point by point and experienced it much more directly, in my view, than is the case today. Insofar, I am now in that age where one of course finds everything from former times I don’t think it would be so much fun today if I just worked on one project, and something would be built and afterwards something comes out of some computer or other, and then I can have a look at it. But of course this development is unavoidable and even necessary, because direct subjective observation also involves direct subjective errors. Now, I don’t want to say any more about these things, only to point out that the attitude to the problems is really also different, and that today one often sees the questions of these experiments quite wrongly posed. The result of the experiments was that one could really measure the energy condition of the atoms as quantities of energy. But that was by no means the intention. When we began with the experiments, Bohr’s theory did not exist and even we ourselves, as I will show later, did not understand what was involved at all. The starting point was the theory of gas discharges. Franck was a pupil of Warburg, in the institute a lot of work was done in those days on gas discharges, but also Wehnelt whose name is also still well known. And the matter referred directly to a theory of Townsend on disruptive electrical discharges. Franck was a man possessed of very unusual degree of the gift of physical intuition, if one may call it so. Meaning that he grasped the relationships, simply from the enormous intensity which he applied to problems; everything he knew and which he might bring into relation with an object, he connected together unconsciously and came to a conclusion. He used to say "my feeling is that that must be like this and cannot be like that". And there he was usually right. And he had this gift to a pronounced degree and he got excited about particular ideas which were built into Townsend’s theory and which he thought were wrong. Now I hope you will allow me to present Townsend’s theory briefly. I will not go into any detail. The idea was simply this: One knew already about the process of impact ionization, that starts with the multiplication of the electrons and that the discharge mechanism then builds up further. And Townsend had tried to construct a quantitative theory and also did it. And that was an unusually significant achievement by him that he did that. But he had introduced into the assumptions, as one so often does, simplifying assumptions, merely in order to be able to make calculations. Perhaps I can explain that a bit and hope that I can still be heard when I am now here. The quantity that interested Townsend was… We let an electron run into a field ... So, under the influence of a field strength the electrons should run into a gas and the quantity (inaudible 08:26), the quantity alpha. That simplifies the number of electrons which are newly produced when an electron moves a distance of, let us say, 1 cm. And he made a picture of this for himself, so that he said that ionization, meaning the removal of an electron from an atom, occurs when the energy of the impacting electron exceeds the work function, and he characterized this with the ionization voltage, so that was the work function in electron volts as we say today. So if the energy of the electron measured in electron volts is higher than the ionization voltage, meaning if the potential difference that the electron has traversed is greater, then ionization should occur. If that is not the case, then it does not. And now in relation to the - I can try to speak a bit louder - Now of course he had to make assumptions about the collisions. He imagined it in this way, that gas, the gas molecules; in those days they were still balls, ... So he introduced the terms of the kinetic gas theory and in particular the term for the mean path length of the electrons. And then he used these two quantities for the derivation of the theory of the quantity alpha. And now he made the following assumption that in every collision between an electron and a molecule the electron loses the whole of its energy. And that is natural, it was unlikely from the very beginning and that was exactly the point which was later shown by our investigations to be false, and which Franck had got excited about from the start. The reason was naturally this: When the electron loses energy each time, then it starts on a new path and then the theory is extremely simple. There I simply have to observe the various successive free paths of the electron. They are sometimes shorter and sometimes longer. And I only need to say that ionization occurs if the path is long enough for the electron to gain the energy on this path. So the length of the individual path must be greater than U divided by the electric field strength. And that is, of course, very simple. The kinetic theory of gases supplies the law directly. The number of paths, for one centimetre I have an average of 1 over lambda paths, and of these a percentage equal to e to the power of minus x * lambda has the property of being greater than x. And by replacing x with this quantity etc. one arrives at the theory. So you can see that Townsend naturally chose these assumptions to allow a simple calculation. Otherwise it is an enormously complicated problem. And that is a legitimate method. In this way he actually represented all the decisive aspects of the breakdown correctly. Still, one can see here once more that if a theory represents reality correctly, it does not need to be correct in every detail. And now comes the essential point. By making measurements – perhaps we can see the next slide now – in order to measure this quantity alpha. This is an image from Townsend’s old book. You can also tell that it is old because the wires here are still braided. Today they are straight lines. That is an apparatus in a vacuum. Here he has a plate that serves as photoelectron. From below ultraviolet light is shone here through a semi-transparent plate. Then electrons are emitted here. And now a voltage is applied between this plate and this plate. Here you can see, by taking different numbers of batteries from a large number, then the separation is changed and the field strength kept constant and in this way he can determine the quantity alpha. Then the number of electrons is, of course, simply n = n(0) * e to the power of alpha * x, that increases exponentially. And by taking the quantity alpha that he has now determined and applying his theory, which I sketched earlier, to calculate the quantities, namely U and lambda, he now determines the ionization voltage of the relevant gases and the average path length of the electrons. And it is interesting to see what came from this – thank you, I don’t need that image any more – that for the noble gases, for which one knows that they have low breakdown voltages, particularly low ionization voltages were found, for the others they were greater. Now, Franck said, that cannot be right. Franck had worked with very pure noble gases. He said with the noble gases it is most definitely not like this. He had in fact not done such experiments, but it was clear to him by analogy with other things, and he said we should now have a go at directly determining these two quantities U and lambda, which Townsend had calculated there in this way from his measurements, in order to check whether it is correct or not. And you can see that we arrive here at our experiments. The starting point therefore is purely a question of gas discharge. The aim, the result, later became a statement about atomic physics. Now, I think that in consideration of the Bavarian breakfast I don’t want to go into too much detail for the performance of the measurements, since they are not really so terribly important. I would like firstly to say something about the measurements of the ionization voltage which we started with. There was already a method provided by Lenard. Perhaps I could have the next image. This is the apparatus which we used there. The improvement compared with other measurements of this kind was that we strove for the first time to produce very clean conditions. So what we then called a vacuum would be seen as a considerable pressure in modern terms, and what we called a metal surface would be seen today as a filthy mess. A metal surface with oxides and fatty remains and I don’t know what else. And if we now wanted to make clean measurements for the various gases, then the conditions had at least to be as reproducible as possible. And that is the reason for building this apparatus. Everything was made of platinum and the gas was boiled in nitric acid. The process of baking out was not known in those days. I personally first came across it later at the Philips bulb factories and it is, of course, absolutely necessary today. Now, the principle is very simple. Here in the middle is an incandescent wire which emits electrons - it was also made of platinum then – and the electrons are accelerated here, now enter another chamber and this attracting plate is brought to such a voltage that the electrons are slowed down again there, so they cannot reach it. If the electrons now produce ions here in this way, then the ions will move there and we have a positive current here. You can see up here the electrometer; we measured currents in those days with electrometers. So that was a very cumbersome procedure with the old Dolezalek electrometer, measuring the charge. Now, this method, it has one error which we did not see back then, which we overlooked as an error. The one, that if light is produced here by electron collisions, in particular ultraviolet light, that this ultraviolet light produces a photo effect here and also produces a positive charge. This current may be weak, but precisely because we used such a sensitive current-measuring instrument we were actually measuring this photo effect here and not the ionization voltage. The next picture please. This is the one main error, which continued to the end of our work together. Now, you can see that we obtained really sharply-defined curves. Here below is the accelerating voltage, and this is the current on the pick-up electrode, and from this we derived the ionization voltage. The next picture please. No, wait a moment. Here oxygen is around 8 and nitrogen is around 7. And perhaps I could have the next image. And you can see that argon is now over 10 and helium is even 20. And Franck was of course very content with this. He said, aha, you can see that Townsend was wrong. It is exactly the noble gases which have the greatest ionization voltage. Now the fact that what we measured was in reality not the ionization voltage at all, but, as we say today, the exciting voltage, that caused no harm in reality, because it is just a matter of the order and the magnitudes are always in the same order. But it was of course a mistake, which one should perhaps not take so seriously, because back then we simply did not know about this phenomenon. So that was the ionization voltage and one saw that the ionization voltages, the actual values, are in contradiction with Townsend's theory. So then it could then be the path lengths, the mean path lengths. The next picture please. And there we made some more measurements of the mean path lengths. Simply here with an apparatus. Here is a metal plate, here a grid, a filament. Electrons are fired into this chamber and are then received by an attractor with a reversed field where one only measures those which carried on further unchanged, unchanged in direction and speed. Then the number of electrons which arrive here is an exponential function, similar to the one above, and we can derive the mean path length from this. And from this also nothing special arose, and it is again interesting from today’s viewpoint what we wrote there. What it said was that we obtained the values predicted by the kinetic theory of gases, that is 4 * root 2 * the mean path length of the gas molecules. That was supposed to be the mean path length of the electrons. And you can see that this is once again this quite naive idea, because one mainly worked with elastic spheres in the gas theory of those days, and we were after all young people and relatively harmless, so we said: Right, if two gas molecules are going to collide, the distance between them must be 2r. It must be less than 2r if they are going to fly past each other. An electron has a diameter which is practically 0. So if the separation here is r, then the path length is 4 times greater. And root 2 comes from the fact that the gas molecules are both moving, whereas with the electrons the single electron is much faster, so that the molecule can be taken to be at rest. You can see that when one looks at it today it is all relatively harmless. And if we had measured everything exactly then we would have had to find all sorts of other things. But it was later that Ramsauer found the Ramsauer Effect. In reality it is, the process of the scattering of an electron with an atom, quite different from the collision of a hollow ball with a billiard ball, as we roughly represented it. But in those days one worked with suchlike primitive ideas and we still made headway. Now particular things remain there, particular errors are contained in it. Now, the result was that the path length was also normal, that is in the way we understood it, and as a result we had to find out how it really was in reality. Townsend’s assumption is wrong. What happens in reality? Now, I want to cover the next experiments rather briefly, just to run through the pictures, please. We already know that more or less. Perhaps we can move on to the next one? This is an experiment mainly to see whether gas molecules are reflected. This here is an electron wave. The electrons are injected into the space through this plate which was covered in soot to stop them being reflected. With this we established how many electrons came out of there. Then it was turned to the side and then measured, what arrives here, and there it was found that the electrons were reflected. And one could more or less determine that the energy loss was very low with noble gases. With other gases a certain energy loss was present. The evaluation of this measurement was difficult, of course, because the electrons are either parallel, or if they are fully undirected one can draw no conclusions at all about the speed distribution from such reverse voltage curves. Here it is a sort of middle condition. It was initially impossible to make any definite statements. Now, the important conclusion, especially with reference to gas discharge, was modern. If the electrons do not lose all their energy in each collision, they can obtain the energy they need for ionization from a whole series of collisions. They can collide many times, move on a zigzag path among the molecules and still keep gaining energy. And to test this we now performed a further experiment - the next slide please. That is another similar apparatus. Incidentally, I would like to point out the braided wires. Here, it has a purpose because this separation was changed. This hung on a chain. The chain went through a U-pipe and was then wound on a grease-coated pin. The U-pipe was in liquid air in order to keep the grease vapours away. And here a very simple experiment. A homogeneous field. The field was held constant. So if the separation was increased – no, excuse me, not with this experiment. In this one, the voltage was held constant. The separation was altered. The voltage was held constant and each time the energy distribution of the electrons here was measured with reverse field curves. Maybe we could have a look at one of these curves. The next picture please. You see these curves up there, these are the curves that one gets there. So they are not very pronounced. They go up to this point. That is connected with the fact that the reverse field measures only one component of the velocity, the component very roughly in the direction of the field, and most electrons hit at a large angle. We then distinguished these curves graphically with a result something like this. Here at the top you can see, that was helium – two curves. One time a particular separation, accelerated with 18 volts. Then the separation was increased by 13 path lengths and this yielded practically the same curve again. That shows that helium could collect the energy from very many collisions. Despite the many collisions, the energy of the electrons corresponded to the voltage applied. It is considerably different with oxygen. There energy losses are present. We don’t want to talk about that now. And here below are a few differentiated groups which essentially only show us the fact that electrons in noble gases can collect energy from many free paths. And we now concluded from this, we knew that the ionization voltage, when the electrons have reached the ionization voltage, then ionization appears. Then they lose all the energy. But before that, none at all. And so we said: Good, in future things will be much easier. We don’t need ionization to measure the ionization voltage, which is difficult, but we simply measure the smallest quantity of energy which the electron can lose. And then we initially changed to metal vapours, because metal vapours are also monatomic like the noble gases, and we expected similar effects and then did the experiments with mercury vapour. The next picture please. Now the apparatus is here very simple. It is essentially the same as what we had before. Only this one is cylindrical. Simply here again for reasons of cleanliness. Again it is all made of platinum and is very carefully cleaned. Of course platinum and mercury is again a questionable combination, but at that temperature it was in any case a well-defined condition of the surfaces where the difficulty still remained that voltages could be produced between the hot filament and the remaining cold surfaces. Now, we wanted to do it like that with mercury too. That means applying some voltage here and then measuring the energy distribution here with the reverse field. And here there is a little trick, a small point that is very typical, probably for many experiments. You saw earlier, these reverse voltage curves, they were not very characteristic or differentiated etc., and drawing a clear conclusion from them was difficult. And then we had the idea of doing it differently. One should not vary this voltage, keep this voltage constant for the measurement and vary this one, but we could apply a constant reverse field here. Perhaps a weak reverse field. And then see, it always happens when the electrons arrive exactly here with the speed 0, when they have lost their energy here just in front of the wire, then they cannot make contact. If they do however have energy, then they can make contact. And then it can also happen that they e.g. in the intermediate space, they become accelerated, and reach the ionization voltage for the first time here. I keep saying ionization voltage here, I am thinking back to some extent to the times back then. In reality though it is the work voltage. Lose energy here, start again etc. And as a result that we did that, we suddenly had the possibility of making a clean measurement. The next picture please. And that is now these curves which one sometimes sees. So when we started, here is the accelerating voltage. Then it starts – it is always 0 at the beginning because of back diffusion and such things - the current increases. Here for the first time the electrons lose their energy, cannot make contact against the reverse field, the current falls. Then it increases again because the electrons gain energy again. Falls again. Increases again and falls again. And as you see, that is now such a typical example, that what I say, that one experienced earlier in physics. This point and the points were measured one after the other. And when one now measures and measures again, what comes next? Now it really falls. And how it then increases again and now reduces again for the second time. That is really an experience, one experiences that more directly than when one some completed recording chart. Or at least that’s how it affects me. So now that was this typical experiment and the result was that for mercury vapour the threshold value was a quantity of energy of 4.9 electron-volts. That the mercury atom absorbed not less than 4.9 volts from the electrons and, as one can see from these curves, also not more. Because if it were more, then the further maxima would be smeared. Now the question is, what we concluded from this. Now we come to exactly what I mentioned earlier, the interesting errors. First of all, at that time there was the quantum theory, various experiments had already been made to relate the ionization voltage as a typical quantity of energy, as a quantity of energy typical for the atoms, to relate it to some frequency or other. And of course it was a matter of seeing what frequency one obtained from the quantities h x Ny = U x E where U is the ionization voltage. And there the most remarkable thing happened, that we obtained for the frequency now the frequency of the 2536 angstrom mercury resonance line. And that was now a highly remarkable fact, since this frequency was a highly characteristic frequency for the mercury atom. According to Hood’s experiments, that is the resonance line. In relation to radiation of this frequency, mercury atoms behave exactly like electric dipoles, electric oscillators in terms of the classical theory. All the energy that they absorb is scattered with the same frequency. In every respect the same. And our conclusion was that: We have here a typical case. We have an oscillator of this frequency and the oscillator takes no other quanta from the electrons than those with the value h * Ny. And we saw that as confirmation of Planck’s theory. We could also determine the h very precisely because we were now able to measure the separation of two maxima. There is always the difficulty with all such experiments that the true energy of the electron does not correspond to the voltage applied, because Volta potential differences still play a role. They are here completely neutralized. Now, that is the picture that we made from it for ourselves. We also readily drew another conclusion: If that is the case, then – the next picture please – mercury vapour, excited by electrons of this energy, must emit this line and no other. And there we also provided a few primitive pieces of evidence. That is a quartz chamber, the mercury is put on the filament as an electrode, and is simply pumped out. And we knew, because we applied a suitable voltage, that electrons of higher energies were completely absent and we got the excitation of the mercury resonance line, only that and no other spectral line of the mercury vapour. And now it is interesting to compare what we back then concluded from this with what we now know. In reality the whole thing became completely clear with Bohr's theory which appeared at about the same time, and it is quite interesting that this is often not seen today. In Herr von Laue’s book, in the little History of Physics, these experiments are described and there follows: “The interpretation was obvious”. And this is followed by the quantum theoretical interpretation. It would be correct to say “the interpretation is obvious”, but it was not obvious back then, since we had in any case not yet understood it. Well, I would like to say some more about it now. I would now like to – yes, perhaps the next picture. That is a little small. That is the nuclear diagram for mercury. And just to compare, how does it look from the point of view of Bohr’s theory? Here we have the ground state. Here is the state that leads to the mercury resonance line. And here is one which is slightly lower in energy. And which in reality, as Franck later showed, leads to the metastable condition. So, what we have determined is that electrons can transfer the quantum of energy necessary for this and so bring it into this state, where it then naturally emits the line again. But if we say that it is the smallest quantity that a mercury atom can absorb, that is in actual fact false. There is an even smaller one. And that results from, that we did not notice that, results simply from the fact that we treated it as all or nothing, either it emits energy or it does not emit energy. In reality, the probability, or as we say today, the effective cross-section for this energy transfer, is very much smaller than for that, so that under the condition of our experiment this plays practically no role, so that that here came out cleanly. So it is also rather by chance that we had just the experimental conditions where this quantum was very clearly absorbed and this particular line emitted. Maybe I should just read the summary of the last work which we see there. It says: “The results of our two studies on the collisions of electrons and mercury atoms may be summarized in the following way: than the quantity h * Ny, since Ny is that of the higher frequency corresponding to the resonance line. this energy quantum will be transferred in one of the next collisions to the electron of the frequency Ny present in the atom. An electron of the frequency Ny must be present there, and it is interpreted like this. That is also wrong. The second statement also wrong.“ Still, an important statement of fact is contained in the study. And then it goes on to say“ So now, the most interesting thing that I wanted to represent here is how extremely difficult it is to recognize such new facts when they appear quite novel, if the other ideas do not yet conform to it. The instant Bohr’s theory was there, everything became completely clear. The misfortune was, if you like, that at this moment we had to end our cooperation because the war broke out. And when we could have started again six years later, it had of course been explained from another direction. We continued to pursue our line of research and this of course provided the possibility of precisely verifying Bohr’s theory. Now, perhaps that is also a question that one can still pose. Bohr’s theory appeared in 1913, and this is 1914 after all. Why did you not see that? That is actually rather odd. But when I think back on it today, for us experimenters the most important thing in Bohr’s theory was the hydrogen atom. Bohr’s theory provided two fundamental insights in those days: That was a very fundamental matter which applied to all atoms, naturally including our mercury atoms. And the second was the special theory of the hydrogen atom, and as experimental physicists we naturally said “a theory of electrons which move in circular orbits and do not radiate, we can't take those seriously for our respectable experiments" - or something like that. One can understand that. And the fact that the Rydberg constant was calculated, the fact showed that that was right for us. But we just did not connect it with our mercury. Now, maybe I can jump to the answer which Franck himself gave to the question in our last conversation, as I mentioned shortly before his death: Then I said to him too: And then he said to me: “Yes, Hertz, we were just too dumb.”

Gustav Hertz (1968)

Memories of James Franck and the Electron Scattering Experiments (German Presentation)

Gustav Hertz (1968)

Memories of James Franck and the Electron Scattering Experiments (German Presentation)

Comment

Gustav Hertz participated in five of the Lindau Nobel Laureate Meetings and this is the last one. His earlier lectures had all been concerned with his more recent research interest in nuclear isotopes, but for this lecture he had chosen the story behind the 1925 Nobel Prize that was awarded to him and James Franck in 1926. It was not so unusual at that time, that a Nobel Prize was reserved from one year to the next. The same procedure was followed, e.g., for Max Planck and Albert Einstein. At the time of the Nobel awarded discovery made by Franck and Hertz, the 5 year younger Hertz was a PhD student with Franck in Berlin. At the time of the Lindau lecture, Franck had passed away and Hertz was 80+, but still going strong. He delivered the lecture at a sprinters speed, in a lively way with many jokes. As he told the story of the discovery of electron impact excitation of atoms, “it was mainly a task for the hands and not for the head”. With this he referred to the construction of the apparatus, which initially was planned to be used for studies of electrical discharges in gases. It was well known that a minimum electric potential was needed to get a flow of ions and electrons through the gas and this potential was identified with the ionization potential of the atoms. But when it was found that there was a stepwise increase in the electric current also below the ionisation potential, it was realized that this showed the quantized excitation of the atoms, just as predicted by the then recently published Bohr theory. In this way their “hands on” experiment became an important cornerstone for the “in heads” quantum theory and explains why their work was rewarded with a Nobel Prize!

Anders Bárány

Cite


Specify width: px

Share

COPYRIGHT

Cite


Specify width: px

Share

COPYRIGHT


Related Content