Home >> Encyclopedia-britannica-volume-14-part-2-martin-luther-mary >> 2 Mactaris to Ferromagnetism Magnetism >> Atomic Rays Magnetism 665

Atomic Rays Magnetism 665

magnetic, elements, susceptibility, diamagnetic, paramagnetic, paramagnetism, ions, electronic, constant and salts

ATOMIC RAYS] MAGNETISM 665 were measured over a wide range of temperatures, particular care being taken to eliminate errors which may arise owing to the presence of even small quantities of iron. The results, supple mented by those of later workers, enable a survey of the magnetic properties of the elements to be made. There are no very obvious simple periodic relationships, and the magnetic characteristics are sometimes very different from those of "normal" dia- and para magnetics. There are elements with a diamagnetic susceptibility which varies with temperature, others with a paramagnetic which does not. This, however, is not surprising, for, apart from the fact that in many cases there may be chemical or physical trans formations involving changes in the magnetic carriers present, the conditions for the Langevin theory of dia- and paramagnetism to hold in its original form (or the modified form due to the Weiss conception of the molecular field and by the introduction of the quantum theory) are far from being fulfilled. In a solu tion, or a solid polar salt, there are definite ions present which form electronic systems of the type to which the theories are applicable. It is, however, not generally possible to consider solid crystalline elements as aggregates of simple ions or atoms. As ordinarily investigated, the solid elements consist of aggregates of small crystals, and an elementary metallic, or even non metallic, crystal may best be regarded from many points of view as a complex whole containing the atomic nuclei embedded in an electronic structure, the whole forming a sort of giant molecule. If there are any ions (with magnetic moments) present which are capable of changing their orientation even slightly in a mag netic field, paramagnetism may be manifested. Otherwise the substance will be diamagnetic.

From the theory of diamagnetism already given, the value of the atomic diamagnetic susceptibility is related to the mean area of the electronic orbits in the atom. Values of the mean radius of the orbits ranging from •i to i.o X cm. lead to values for the susceptibility ranging from x= —.014 X to x = —1.4 X Many diamagnetic elements have susceptibilities falling within this range. Bismuth is the most diamagnetic element (to' x= —1.4). Its susceptibility decreases with temperature up to the melting point (io' x= —.06 at 273° C), and then changes abruptly to a value about as great in the liquid form, which remains constant. Antimony behaves somewhat similarly. For diamagnetic elements in the same group in the periodic table, the susceptibility usually increases with increasing atomic number. This is illus trated by sulphur, selenium, tellurium; chlorine, bromine, iodine; zinc, cadmium, mercury; and copper, silver, gold. The suscepti bility of some of the elements decreases on melting (e.g., Bi, Sb, Pb), that of others increases slightly (e.g., Ag, Au).

Among the paramagnetic elements the susceptibility of some remains practically constant with temperature (e.g., sodium, potassium, rubidium, calcium), that of others increases, while among the elements whose susceptibility decreases Curie's law X = is not obeyed. There are, however, indications that

( when corrections are made for the underlying diamagnetism, a Weiss law(x = may be approximately followed with a large negative value for 0, for the rare earths, palladium and platinum. For palladium and platinum atomic magnetic moments corresponding to 9.7 and 9 Weiss magnetons have been deduced. Tin is interesting in that the white tetragonal variety is slightly paramagnetic, while grey tin and the liquid are diamagnetic.

The Magnetic Electron.

If the electron itself has an in trinsic spin and magnetic moment, it might be anticipated that metals, presumably with free electrons responsible for the elec trical conductivity, would be strongly paramagnetic, with a sus ceptibility comparable with that found for salts containing ions each with a magnetic moment corresponding to one Bohr mag neton (giving a Weiss magneton value of about 9). W. Pauli has considered this question from the standpoint of the quantum statistical theory developed by E. Fermi, and has shown that, except at very high temperatures, there will be an almost corn plete balancing in pairs of any free electrons, so that they will make only a very small contribution to the susceptibility. The alkali metals are decreasingly paramagnetic in the order sodium, potassium, rubidium ; while caesium is diamagnetic. This may be attributed to the increase in the underlying diamagnetism with increasing atomic number. It seems quite possible that many of the anomalous magnetic characteristics of metallic elements are partly to be traced to the magnetic properties of free electrons; but there are many curious features which are still very obscure.

Constant Paramagnetism.

The anomalous constant para magnetism occurs not only among elements (e.g., the alkali and alkaline earth metals, vanadium, chromium), but also among oxides of some paramagnetic metals, and, as has been found by Weiss and Collet, for certain salts. Among the salts are potassium bichromate (the solution has a constant susceptibility between and 50° C), permanganate and ferrocyanide. As has been dis cussed under paramagnetism, the salts of manganese, for exam ple, in which the manganese occurs as a positive cation, are para magnetic, the susceptibility leading to a value for the magnetic moment of the ion, which fits in with values for other ions and is in agreement with what is anticipated from the electronic struc ture deduced in other ways. In the anion of potassium per manganate, however, the electronic structure is completely dif ferent, and the electrons can form a symmetrical configuration with no magnetic moment. Although the paramagnetism observed is much smaller than that for normal paramagnetic salts, it is difficult to account for the appearance of paramagnetism at all, independently of the fact that the susceptibility does not vary with the temperature. Although there is a fairly satisfactory theoretical interpretation of normal dia- and paramagnetism, the subject matter of this section shows that there are still many com pletely unsolved problems in the field.