Magnetic hysteresis and domain structure

2014 A brief survey is made of some of the problems involved in the development of experimental methods of obtaining information on the domain processes involved in the low field magnetization of polycrystalline materials. LE JOURNAL DE PHYSIQUE ET LE RADIUM TOME 20, FÉVRIER 1959, This paper represents an attempt to summarise briefly some of the problems involved in the development of an understanding of the elementary processes involved in low field magnetic hysteresis in polycrystalline materials such as iron and nickel, to discuss some of the methods of obtaining further information, and to indulge in the exercise of estimating orders of magnitude. The magnetic properties of the materials with which this paper is concerned are a coercivity in the range 0.1 to 10 oersted, a remanent magnetization of about one half the saturation value. In addition there is the information from the Barkhausen effect that the discontinuous changes in magnetization take place in steps of 10-7 to 10-5 e. m. u. corresponding to reversals in magnetization in volumes in 10-1° to 10-8 cm3 ; there are probably a considerable number of smaller discontinuities but this range covers those which make up the greater part of the change in magnetization (an average value for hard drawn iron is 1. 6 X 10-6 e. m. u.). The theoretical treatment of coercivity in single crystals and of the mechanism ’whereby the magnetization is reversed has been exhaustively treated particularly by Néel and the processes involved are fairly well understood. The transference of these ideas to polycrystalline materials however presents considerable difficulty and, without any wish to labour this point, it may be of use to give an example of what is involved. The reversal of magnetization in a single crystal of iron is considered as taking place with a 1800 boundary moving across a domain ; this boundary is held up as it unites with a 900 Néel " spike " domain structure around a non magnetic inclusion and as it breaks away produces a discontinuous change in magnetization. This has been confirmed in the well known Bitter patterns of Williams and Shockley (1949) ; from the published photographs it appears that the linear dimensions of the inclusions were about 6 X 103 cm and the length of the Néel spikes about 7 X 10-2 cm. Now in a polycrystalline material, unless one is to postulate such a close coherence in crystal orientation between neighbouring grains that a domain boundary passes with little distortion across a grain boundary, it is necessary to suppose that the whole course of events is carried out in a single grain. A " typical" grain size for a polycrystalline material such as that described earlier, would be about 10-3 cm in linear dimensions, so that the whole scale of the above process is much too large. Even if the scale of the system is reduced to a minimum, it is unlikely that, with a boundary thickness of 1.4 x 10-5 cm (for iron), one could obtain any subsidiary domain structure on a cubic inclusion. with a side of much less than 1 X 10-4 cm ; the cross sectional dimensions of the associated Néel spike would take up the greater part of the cross section of the grain with little room for the 1800 boundary. An alternative suggestion would be that the movement of the 1800 boundary is held up because of the free pole produced at the intersection of the boundary with an inclusion, without the formation of any subsidiary domain structure. It has been shown by a number of workers that in such cases those inclusions with linear dimensions approximately equal to the boundary thickness are of the greatest importance in producing hysteresis (see below). (Néel, in his disperse field theory also allowed for the possibility of regions of strain, instead of non-magnetic inclusion, with the energy of the system dependent on strain and anisotropy). It may be worth mentioning that if the deciding factor is whether or not there is room for the formation of subsidiary domain structure around an inclusion there might well be a close inter-relation between grain size and inclusion size in the control of coercivity. An additional possibility would be that the Néel spike structure itself could provide a means whereby the increase and diminution in the size of the spikes would result in a change in magnetization ; however enough has been said to indicate Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphysrad:01959002002-309800