Compound-target sputtering toward high-performance polarizing multilayer neutron optics : Process optimization and microstructural control for nextgeneration neutron instrumentation

Sammanfattning

Polarizing neutron mirrors are important components in neutron instrumentation, where efficient control of the neutron spin state is required. Fe/Si-based multilayers are promising for this purpose because Fe provides magnetic contrast, while Si acts as a non-magnetic contrast layer. The incorporation of boron carbide can improve the structural stability of such multilayers by reducing crystallization and interdiffusion, but practical deposition routes must still provide control over composition, interface quality, magnetic behaviour, and neutron-optical performance. This work investigates compound-target sputtering as a route for depositing Fe–B–C/Si–B–C multilayers for polarizing neutron optics. The motivation for this approach, rather than continued co-sputtering, is partly practical and industry-oriented. Compound targets are directly compatible with standard planar magnetron systems, where each source typically accommodates only one target material at a time. This makes the approach more relevant for industrial implementation, since Fe–B–C and Si–B–C layers can be deposited from single targets rather than requiring multiple synchronized sources. Fe+B4C and Si+B4C targets were therefore used to deposit single layers and multilayers by DC magnetron sputtering. The sputtering and gas-phase transport processes were studied using SRIM and SiMTRA simulations, while the deposited films were characterized using X-ray reflectivity, X-ray diffraction, elastic recoil detection analysis, SQUID/VSM magnetometry, and polarized neutron reflectometry. The results show that B and C are incorporated into both Fe-based and Si-based films when using compound targets. However, the measured film compositions differ from the simulated sputtered and transported fluxes, indicating that the final film composition is influenced by growth-related effects and contamination. In addition, limitations in the measurement techniques restrict how precisely the composition and depth distribution can be determined. X-ray reflectivity shows that lower working pressure and moderate ion assistance improve multilayer quality, while high pressure and insufficient or excessive energetic assistance lead to broader effective interfaces. The magnetic measurements show that very thin Fe-based layers have reduced magnetic performance, while thicker multilayers retain a clearer ferromagnetic response. Polarized neutron reflectometry confirms that selected multilayers produce spin-dependent neutron reflectivity, with the best-performing sample showing strong spin selectivity at the multilayer Bragg peak. Overall, the results demonstrate that compound-target sputtering is a promising route for fabricating Fe–B–C/Si–B–C multilayers for polarizing neutron optics. Compared with the earlier Fe/Si+11B4C multilayers reported by Zubayer et al., the present samples do not yet reach the same neutron-optical performance in terms of interface quality, spin selectivity, and magnetic behaviour [47, 42, 43, 46]. However, the compound-target approach addresses a different and more practical objective, namely to explore a deposition route that is more compatible with industrial planar magnetron systems, where each source typically uses one target material at a time. The results should therefore be viewed as a first step toward a more scalable alternative to co-sputtering, rather than as a direct performance replacement. Further optimization is needed to improve composition control, reduce contamination and interface broadening, and increase neutron-optical performance, particularly through the use of isotopically enriched 11B4C.