We demonstrate that total energy calculations based upon only the local density approximation of density functional theory in combination with an orbital polarization correction can be used to derive the magnetocrystalline anisotropy energy (MAE) for the ferromagnetic metals, bcc Fe, hcp Co, fcc Co, and fcc Ni. In the case of bcc Fe, hcp Co, and fcc Co the calculations reproduce the experimental easy axis as well as the size of the MAE. However, for fcc Ni we obtain the wrong easy axis. PACS numbers: 71.25.Pi, 71.20.Ad, 71.25.Cx, 75.30.Gw The calculation of the magnetocrystalline anisotropy energy (MAE), i.e. , the difference in energy when the magnetization is pointing in two different directions, has for a long time been an outstanding problem for the 3d ferromagnetic elements [1]. It was suggested earlier that the coupling between the magnetization and the lattice, arising through the spin-orbit interaction, in combination with an energy band picture, will provide a MAE of the right order of magnitude for bcc Fe, hcp Co, and fcc Ni [2]. Several investigations have been reported within this approach, and one of the conclusions drawn from these studies is that the technical problems in resolving the extremely small energy differences which are needed are very hard to overcome. Previously, an investigation of the magnetocrystalline anisotropy energy for the late 3d transition metals, using a standard energy band calculation method in conjunction with the local spin density approximation (LSDA), was reported [3].While the magnitude of the MAE (of the order of p, eV) was correctly obtained for all three metals, the wrong easy axis was obtained for hcp Co and fcc Ni, and, moreover, the number of k points was unduly large.The calculations presented by Daalderop, Kelly, and Schuurmans [3] were based on a geometrical constraint on the charge and spin density of the crystal as well as of the shape of the crystal potential through the use of the atomic sphere approximation (ASA). Further, the socalled force theorem [4] was used to evaluate the energy difference between two spin directions. However, due to the development of more efficient and faster computers in combination with recent advances in accurate total energy methods, we will in this Letter demonstrate that it has now become possible to calculate the MAE for the 3d elements directly from the total energy in a self-consistent way. This is a most desirable development since from previous theoretical work one could not draw conclusions about whether the disagreement between experiment and theory was due to limitations of the calculational method or due to the errors associated with the LSDA. Previously, the total energy difference approach was used to calculate the MAE for uranium sulphide (US) [5]. However, its MAE is 10 times larger than for the 3d elements and does not call for the special treatment we address in this present Letter.Here we present results that go beyond the above mentioned approximations.As a matter of fact, the only major approximation mad...

The spin and orbital moments of the surfaces of Fe, Co, and Ni as well as of overlayers of these metals deposited on Cu have been calculated using a full-potential linear muffin-tin orbital method in a slab geometry. With one exception, Ni on Cu͑001͒, the calculated spin moments are considerably enhanced at the surface, which is in agreement with previous theoretical and experimental work for some of the presently studied systems. We argue that the Ni-d -Cu-d hybridization is responsible for the reduced Ni spin moment in the case of a Ni monolayer on a Cu͑001͒ substrate. The orbital moment is enhanced at the surface for all of the studied cases, sometimes by as much as a factor of 2 relative to the value found in the bulk. Based on our theoretical calculations we propose that an experimental confirmation of an enhanced orbital magnetic moment on a surface is most likely to be found for a monolayer of Co on a Cu͑001͒ substrate or for the Fe͑001͒ surface layer. Experimental work on Co on Cu͑001͒ confirm this behavior.

The optical properties of cubic and hexagonal GaN films in the region of the fundamental gap are studied with spectroscopic ellipsometry at temperatures between 110 and 630 K. It is verified that the gap of hexagonal GaN is higher than that of the cubic polytype. The parameters of the gaps are determined against temperature and the temperature shifts are found to be lower than and close to those of GaAs and GaP in the cases of cubic and hexagonal GaN, respectively. Additional theoretical calculations of the electronic structure of both polytypes using the full-potential linear-mu5n-tin-orbital method reveal a significant contribution to the Eo gap from the 8~10 transitions. The resulting gap energies are compared with the literature and the difference between the two GaN polytypes is discussed. The dielectric function c& (co) is directly calculated from the band structure and its features at energies up to 9.5 eV are discussed and compared to experiment.

Theoretical calculations in combination with experiments for the * and * x-ray absorption edges are reported for graphite. Theory and experiment agree well for the leading * and * resonances. By comparing theoretical calculations for a single graphene layer that include the effect of the core hole to similar calculations that do not, we find that 1s x-ray absorption in graphite is to be associated with an excitonic effect. Both the * and * excitons are localized primarily on the core-excited atom, but have significant weight on the nearest-neighbor atoms. The results are closely related to the electronic structure of a N impurity.

The prevalence of Chlamydia psittaci infection in household, feral and farm cats in Britain was investigated. Chlamydia were isolated from 30 per cent of conjunctival swabs collected from 753 household cats with conjunctivitis. The prevalence of active chlamydial infection was highest in cats in the age group five weeks to nine months. Males were more frequently infected than females. Cats with chlamydial conjunctivitis usually had antibody titres greater than 1024 as assessed by indirect immunofluorescence. Chlamydia appeared to be endemic in two out of three feral cat colonies on the basis of serological evidence and occasional isolations. Cats on 10 of 22 sheep farms (45 per cent) had serological evidence of chlamydial infection, and this was confirmed on two farms by isolation of the organism from conjunctival and, or, rectal swabs. This is the first survey of infection with Chlamydia psittaci in cat populations in Britain.