0
$\begingroup$

The susceptibility tensor is defined as

$$ \bf{M}_i = \chi_{ij} \bf{H}_j. $$

From looking at a material's geometry and crystalographic properties, is it possible to say whether the off-diagonal terms of the suceptability tensor will be non-zero? Framed slightly differently, if it was calculated that $ \chi_{i, j} \neq 0 $ for some $i \neq j$, what would this tell us about the structure of the material? Is this ever a property of ideal crystaline materials or does it always suggest impurities?

$\endgroup$
1
  • $\begingroup$ This ties closely to symmetry of the crystal, indeed. Compare with the existence of off-diagonal mechanical property elements. $\endgroup$ – Jon Custer Sep 11 '19 at 16:31
0
$\begingroup$

Indeed, the terms of the susceptibility tensor are related to the symmetry of the material (see Neumann's principle). Some basic facts about the diagonal elements of this tensor are:

  • Cubic crystal has $\chi_{11} =\chi_{22} = \chi_{33}$
  • Uniaxial crystals (tetragonal, hexagonal, trigonal symmetry) have a preferred magnetization direction (easy axis), let's say along the z-axis, such that $\chi_{11} =\chi_{22} < \chi_{33}$
  • Biaxial crystals (orthorhombic, monoclinic, triclinic symmetry) have two preferred magnetization directions, so all diagonal components are different

The off-diagonal elements are more complicated. Of course, these off-diagonal terms can be non-zero for complicated symmetries. Nevertheless, the crystal symmetry might not tell us everything, for example, there might be magnetic interactions (e.g. crystal field) that result in nonzero off-diagonal terms.

NOTE: The relation you stated assumes that the magnetic response of the material is linear, which is the case when the magnetic field is sufficiently small. At high magnetic fields, non-linear magnetic effects might occur.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.