Amorphicity, heterogeneity, and the Arrott plots

The Arrott plots make a very powerful and important tool for an accurate determination of the Curie point and of the ‘‘critical’’ exponents near that point. But after being used successfully for almost two decades, the method’s original purpose has been somehow forgotten, and some attempts have recently been made to use it out of its natural region of applicability. Like every tool, this one also has its limitations, and leads to the wrong results when they are ignored. In particular, these plots are based on an equation of state for the saturation magnetization, and cannot reveal any meaningful information for magnetic fields which are too low to saturate the sample. There is thus no point in using the Arrott plots (as some do) to look for the infinite susceptibility without ferromagnetism, predicted for amorphous materials in low fields by a theory which neglects everything that really happens in low fields. On the other hand, too high fields also distort the plots which are based on a first‐order expansion. These, and other limitations will be discussed, and an analysis of their significance will be given. A conclusion of this analysis is that available experimental data do not necessarily indicate a sharp transition at a well defined Curie point in amorphous ferromagnets, as is normally assumed. An equally justified interpretation is that this is only due to an inappropriate extrapolation of the Arrott plots. Therefore, amorphous materials might be just a heterogeneous system of ferromagnets (e.g., the mixture of clusters assumed in some theories), with a region of an ill‐defined transition from ferromagnetism to paramagnetism. For high degree of short‐range order, the ‘‘smearing’’ of the transition region might be smaller than the experimental accuracy, so that the amorphous material might look like a true, homogeneous ferromagnet. Only this is probably rarer than might seem from the literature, because many of the reported sharp transitions are evidently caused by Arrott plots extrapolations from too high fields.