Resonance Potentials

Experimental Methods.

The pioneer worker on the subject was Lenard, who in 19o2, long before Bohr's theory was put forward, showed that an electron must possess a minimum energy before it can produce ionisation in a gas. He released the electrons from a metal plate P photoelectrically (see PHOTOELECTRICITY), and accelerated them by means of a parallel gauze E maintained at the desired potential, the gas pressure being low enough for the greater part of the electrons to pass through the space between P and E without a collision (fig. 1). He detected the formation of ions by means of a plate R, charged negatively, so that, while electrons cannot reach it direct, any positive ions formed are at once attracted, and make their presence known by a sudden change in the current from R. In this way he found that no ions were produced unless the accelerating potential exceeded a certain threshold value. However, as pointed out by Bohr and van der Bijl, a change in the current from R does not necessarily indicate ionisation in the gas between E and R, for if the electron impact is not sufficiently energetic to make the atoms struck lose an electron, but merely makes them emit radiation, then this radia tion will act photoelectrically on the plate R, causing it to lose electrons. As far as current effects go, loss of electrons by R or gain of positive ions by R come to the same thing. Lenard's original method has therefore been modified in various ways, to enable a distinction to be made between a resonance potential and an ionisation potential. It is also usual nowadays to produce the electrons by means of a hot wire (see THERMIONICS) instead of photoelectrically. Davis and Goucher introduced a second gauze E' (fig. 2) and arranged the potentials as follows : a potential V R, greater than the accelerating potential VA, acts so as to stop the electrons reaching R, while a small potential which can be reversed, is maintained between E' and R. When E' is negative to R photoelectric electrons cannot escape from R, when E' is positive to R they can, so that the photoelectric effect of radiation on R can be detected at once. Reversal of the direction of the small field between E' and R has, however, little effect on the passage of positive ions to R, because these are accelerated by a comparatively large potential fall, and have sufficient energy to overcome the opposing field. In this way radiation potential and ionisation potential can be clearly distinguished. Lenard's method has also been modified by Franck and Hertz in a famous series of experiments which dealt especially with the resonance po tentials of mercury vapour. This vapour has yielded particularly clear confirmation of Bohr's the ory. There are, for instance, resonance Ntentials at 4.9 volts and 6.7 volts. The wave-lengths which correspond, on the quan tum theory of spectra, to these potentials can be calculated from the fundamental equation when e is the electronic charge, V the potential in volts, h is Planck's constant, c the velocity of light, X the wave length. To 4.9 volts corresponds a wave-length of 2520 A.U., to 6.7 volts a wave-length of 1844 A.U., which agree, within the experimental error of these measurements, with 2536 A.U. and 1849 A.U., the wave-lengths of the two strong lines of the mercury spectrum to be anticipated on the theory.

G. Hertz has more recently worked out some very delicate meth ods of measuring both resonance and ionisation potentials. One of the methods detects, by a skilful disposition of the gauzes, the abrupt loss of velocity of electrons which takes place when the accelerating potential reaches a critical value : this method there fore measures radiation potentials. The other method depends upon an annulment of the so-called space charge, which sur rounds a hot wire, by the positive ions produced when, but not before, the accelerated electrons have sufficient energy : this method clearly detects ionisation potentials.

A different type of experiment, that which relies on the excited radiation for a sign that the reso nance potential has been reached, is represented by the work of Foote, Meggers, and Mohler.

They use the disposition repre sented in fig. 3. The electrons are produced, as usual, by a hot wire, here of hairpin shape, and accel erated by the field between the wire and the grid, constituted by a close spiral coil. The cc mpara tively large cylinder which sur rounds the grid is kept at the same potential as the grid.

The region in which the electrons are accelerated is narrow, so that there is little chance of an impact which would prevent an electron attaining its full velocity : the re gion between the grid and plate is wide, so as to give plenty of opportunity for impacts to produce radiation. The accelerating potential is varied, and the values at which individual spectral lines appear carefully noted. The study of resonance potentials in this way was initiated by Franck and Hertz, who observed the potential which was just sufficient to excite the well-known mer cury line of wave-length 2536 A.U. They were followed by McLennan and his students, who showed that either one or two lines or the whole series could be excited, according to the poten tial. Since then other workers, notably Newman, have succeeded in producing certain spectral series line by line, each new line first appearing at the potential indicated by Bohr's theory.

All these experiments clearly show that energy can be com municated to atoms in definite amounts only, and that the com munication of a definite amount of energy to an atom is followed by the emission of a spectral line or lines, the wave-length of these lines being connected with the energy communicated exactly as indicated by the quantum theory of spectra. It should be added that not only can one electron be completely removed from an atom at the ionisation potential, but that a higher potential can be measured which suffices to remove two electrons, and so excite the so-called spark spectrum. (See SPECTROSCOPY.) Applications of the Theory of Resonance Potentials.— The established fact that a perfectly definite energy is needed to excite any given spectral line, or to separate an electron from an atom and so ionise it, has found wide application. The energy with which one atom strikes another at room temperatures is far below that which corresponds to the first resonance potential of any gas, and so we can expect no luminosity of gases at ordinary temperatures. In a flame the temperature is already sufficient for an appreciable fraction of the atoms to collide with sufficient energy to excite spectral lines. If metal atoms are introduced into a flame, the lines which are detected at the lowest temperatures are the lines with the lowest excitation potentials, and as the temperature of the flame is raised more and more lines appear, in accordance with the theory, the increased energy of the atomic impacts corresponding to the increased energy of electron impact which we get as we raise the potential. Again, elements with spectra of low resonance potentials show in the flame more lines at a given temperature than those with high resonance potentials. More striking still are applications of the conception of ionisation and resonance potentials to astrophysical problems. By consider ing the ionisation of an atom as a chemical problem, in which the ionisation potential takes the place of the heat of dissociation, Saha has worked out the percentage ionisation to be expected under different conditions of temperatures and pressure, and applied his result to the spectrum of the sun, with very interesting results. He has especially considered the ionisation of calcium atoms in the sun's atmosphere, and explained many peculiarities of the apparent distribution of the element. This work has been much extended, especially by R. H. Fowler and E. A. Milne. (See STAR.)