Introduction by A. Rogalev, ESRF

Over the past decade, X-ray spectroscopies have undergone a continuous expansion. This area has been boosted by the developments in the synchrotron radiation instrumentation, which have made it possible to produce high fluxes of circularly-polarised X-ray photons. The development of X-ray Magnetic Circular Dichroism (XMCD) into a well-established technique with an extensive list of applications and the experimental evidence of X-ray Optical Activity (XOA) in non-centrosymmetric systems are particularly interesting. These spectroscopies, being element specific and orbital selective, proved to be remarkable tools for the investigation of the electronic structure of various materials.

XMCD, with the help of the magneto-optical sum rules, has as its major strength the capability of disentangling the spin and orbital contributions to the total magnetic moment carried by an absorbing atom. Despite several approximations used to derive XMCD sum rules, thorough experimental and theoretical studies have proved their validity for the L-edges of the 3d-, 4d- and 5d- transition metals as well as for M-edges of the 4f- and 5f-elements. However, in the case of the L-edges of rare-earth elements those approximations appear to be too crude and, therefore, straightforward application of the sum rules results in erroneous conclusions. The way to overcome this difficulty is to combine the theory with the experiment. This is nicely illustrated in the contribution by H. Wende et al., where the magnetic moment carried by the 5d electrons of a rare-earth atom has been determined for the first time. This approach extends the range of applications of XMCD to the rare-earth elements and other compounds that are of great technological importance.

Magnetic materials exhibit intrinsic "easy" and "hard" magnetisation directions. The preferred orientation of magnetisation is one of the most important properties of magnetic materials and is determined by the Magnetocrystalline Anisotropy Energy (MAE) which is usually strongly-increased in thin films due to the symmetry breaking. Notwithstanding this, our knowledge about the microscopic origin of MAE is rather scarce since its detailed study has become possible only with the advent of XMCD. Using this technique, P. Gambardella et al. have shown that the MAE of Co clusters, deposited on a Pt(111) surface, vary drastically with the cluster size (1 to 40 atoms), even if the latter is changed by only one single atom.

Dichroic effects can be observed not only in X-ray absorption but also in X-ray Resonant Raman Scattering (RRS); the latter has the potential of giving information that is hardly accessible to other spectroscopies. Using circularly-polarised X-rays incident perpendicularly to the sample magnetisation direction and monitoring the integrated intensity of emitted photons over the energy range of the selected de-excitation channel, L. Braicovich et al. have demonstrated that it is possible to measure both charge and magnetic multipole moments of the absorbing/emitting atom. Moreover, this experimental arrangement is shown to be optimal to test the validity of the RRS sum rules and it allows one to make a direct comparison of the experimental results with ab-initio calculations.

Unlike X-ray magneto optical effects such as magnetic RRS or XMCD which are governed by pure electric dipole (E1E1) or electric quadrupolar (E2E2) transitions, X-ray optical activity is associated with transition probabilities that mix multipole moments of different parities, namely E1E2. Recent work by the ESRF ID12 team and in the Theory Group has revealed that various X-ray optical activity spectra are related to orbital anapole moments and to a whole set of parity-mixing operators [1]. Recently derived X-ray optical activity sum rules may offer unique experimental access to the ground state expectation values of those operators. Furthermore, ab-initio calculations of the anapole moment are also under way. As the first attempt, the Stark-induced anapole moment of alkali atoms have been calculated by the ESRF Theory Group using a relativistic formulation of the current density functional theory with spin-orbit correction.

[1] J. Goulon et al., JETP, 97, 402-431 (2003).