An ideal solid-state supermirror (SM) neutron polarizer assumes total reflection of neutrons from the SM coating for one spin-component and total absorption for the other, thus providing a perfectly polarized neutron beam at the exit. However, in practice, the substrates neutron-nucleai optical potential does not match perfectly that for spin-down neutrons in the SM. For a positive step in the optical potential (as in a Fe/SiN(x) SM on Si substrate), this mismatch results in spin-independent total reflection for neutrons with small momentum transfer Q, limiting the useful neutron bandwidth in the low-Q region. To overcome this limitation, we propose to replace Si single-crystal substrates by media with higher optical potential than that for spin-down neutrons in the SM ferromagnetic layers. We found single-crystal sapphire and single-crystal quartz as good candidates for solid-state Fe/SiN(x) SM polarizers. To verify this idea, we coated a thick plate of single-crystal sapphire with a m=2.4 Fe/SiN(x) SM. At the T3 instrument at the ILL, we measured the spin-up and spin-down reflectivity curves with 7.5 A neutrons incident from the substrate to the interface between the substrate and the SM coating. The results of this experimental test were in excellent agreement with our expectations: the bandwidth of high polarizing power extended significantly into the low-Q region. This finding, together with the possibility to apply a strong magnetizing field, opens a new road to produce high-efficient solid-state SM polarizers with an extended neutron wavelength bandwidth and near-to-perfect polarizing power.
Download