Addendum D
Experimental facts reported in three recent articles in Nature 1,2,3 indicate need for important correction to the current interpretation of magnetization by magnetic field. It is believed to be a "rotation" ("switching", "reversal") of spins in the intact crystal lattice. The new interpretation, put forward earlier 4, states that spin vector is an orientation characteristic of its carrier (atom, molecule), therefore magnetization involves turning the carriers. The latter proceeds by the universal mechanism of crystal growth in liquids and solids: by nucleation and molecule-by-molecule filling "kinks" (steps) at the interfaces to complete layers and building successive layers in this manner.
Lavrov et al.1 note that while magnetic field affects orientation of spins, it, it seems, should have little impact on the crystal structure; nevertheless, changes of crystal orientation were recently reported. But their own observation of structural rearrangement in an antiferromagnet they find even less expected. According to the current views, spins in antiferromagnets strongly interact to each other, making a magnetically neutral system and should not be affected by magnetic field. Yet, the authors observed generation and motion of crystallographic twin boundaries and kinks moving along the crystal surfaces, resulting in reorientation of the crystal. While their findings are inconsistent with magnetization by "switching", they are in accord with the magnetization mechanism presented in Ref.4. Actually, Lavrov et al. dealt with the antiferromagnet à ferromagnet phase transition when every second spin carrier was turned during relocation at twin domain interfaces to make all spins parallel to the field. Evidently, spins were strongly bound to their carriers rather than to each other.
Novoselov et al. 2 recorded magnetization with the high resolution attained never before. They found that ferromagnetic domain interface propagated by distinct jumps matching the lattice periodicity, the smallest being only a single lattice period. Some results also suggested that "kinks" were running along the interface. The authors interpreted the interface movements as following the Peierls potential of crystal lattice and stated that further theoretical and experimental work is needed to understand the unexpected dynamics of domain walls. The phenomenon, however, had been described, predicted to be traced to the molecular level, and illustrated with a molecular model in Ref.4 as a crystal rearrangement by the above-mentioned mechanism. In fact, the same mode of interface movement (running kinks and layer-by-layer) was observed by Lavrov et al. 1 , only on more macroscopic scale, and there crystal rearrangement was firmly established.
Tudosa et al. 3 estimated experimentally the ultimate speed of "magnetization switching" in tiny single-domain particles - an important issue in developing magnetic memory devices. The speed turned out three orders of magnitude lower than was predicted and, besides, was not the same in the effected particles. The error of that prediction is hidden in the term "switching", in other words, in the assumption of spin rotation in the crystal structure. The lower speed had to be expected, considering that magnetization is not a "switching", but occurs by nucleation and growth in every individual domain. Nucleation is heterogeneous, requires specific crystal defects and not simultaneous in different particles. It is nucleation that controls re-magnetization of small single-domain particles 4.
Recognition that magnetization resulted from structural rearrangement will be an essential advancement in understanding of magnetism.
1. A. N. Lavrov, Seiki Komiya & Yoichi Ando, Nature 418, 385 (2002).
2. K. S. Novoselov, A. K. Geim, S. V. Dubonos, E. W. Hill
& I. V. Grigorieva, Nature 426, 812 (2003).
3. I. Tudosa et al., Nature 428, 831 (2004).
4. Y. Mnyukh, Fundamentals of Solid-State Phase Transitions, Ferromagnetism and Ferroelectricity, Authorhouse, 2001.
*The article was submitted to journal Nature in July 2004. It was rejected by editor Rosalind Cotter while stating that "we are not questioning the validity of your innovative theory, or its interest to others in the field". The same editor recently promoted publication in Nature of article on the quantity of ways to tide up shoe laces.