5. Future Directions

• It is proposed to measure the polarization of a reflected electron beam from $Mn/7ML$ $Co/Cu(001)$ as a function of $Mn$ thickness, rather than, as at present, for a single $Mn$ thickness. If the increase in MOKE signal (section 3.1,) observed as the first half monolayer of manganese is deposited, is reproduced as an increase in polarization, this will provide strong evidence that the MOKE signal is a consequence of a genuine increase in overall magnetic moment of the film, rather than merely an increase in magneto-optical response. This in turn would imply either that the manganese atoms were ferromagnetically aligned with each other and with the cobalt, or that the cobalt moments were enhanced by manganese adsorption; the latter explanation seems implausible in the light of the tight-binding theoretical model of Noguera et al. [17], which predicts significantly suppressed $Co$ moments at an $(001)$ interface with manganese.

• It is proposed to undertake a more detailed examination of the form of the polarization of the reflected beam from $Mn/Co/Cu(001)$ as a function of beam energy: certainly with secondary (Auger) electrons and possibly with specularly reflected electrons, this functional form will contain the information to distinguish the polarizing effects of cobalt and manganese.

• It is proposed to expand the measurements presented in figure 3.6 to specify polarization as a function of copper capping layer thickness rather than, as at present, for a single copper capping layer thickness. Examination of the decay of the measured polarization due to the finite mean free path of electron spins in copper may assist in determining how deeply buried the magnetic layer is, and therefore whether the manganese is ferromagnetic.