Structural transformations in Mg-Ni films induced by hydrogenation

Abstract

We investigated thin film samples of Mg2NiH4 with two intentions. First of all, we wanted to ascertain if the same nanomaterial (Mg2NiH4) prepared by magnetron sputtering and ball milling can exhibit different hydrogen storage properties and to see possible advantages/disadvantages of employing of magnetron sputtering for synthesis of nanometerials for hydrogenstorage. Furthermore, we wanted to see if thin film samples of Mg2NiH4 could be used in a switchable mirror or window device by utilizing the high to low temperature transition at about 510 K. In powder samples, this transition, between a monoclinic conducting low temperature phase to an FCC non-conducting high temperature phase, have been demonstrated in a mechanical reversible conductor–insulator transition [Blomqvist and Nor��us, J. Appl. Phys 91(2002)5141]. The new thin film Mg2NiH4 samples were produced by reacting hydrogen with magnetron sputtered Mg2Ni films on quartz glass or CaF2 substrates. But we could not obtain the monoclinic low temperature phase upon cooling the samples. Instead a cubic phase, related but not identical to the cubic high temperature phase, was formed at temperatures both below and above 510 K. TEM pictures revealed the new cubic phase in the films to have the same cell parameter as the FCC high temperature phase. But the symmetry was lower with similar streaking patterns as observed for the monoclinic low temperature phase. IR-spectroscopy indicated an identical vibrational frequency for the allowed stretching mode of the tetrahedral NiH4-complex both in film and powder samples. Band gap measurements indicated a band gap of 2.3 eV for the films but 1.1 eV for monoclinic Mg2NiH4 powders. The films were transparent with a yellow colour that gradually deepened into red when heating up to 600 K in hydrogen. A typical first order phase transition from a low to high temperature phase could not be observed with DSC when heating through the low- to high temperature phase transition at 510 K. A severely strained film was revealed with optical microscopy in reflected light. This observed microstrain may reflect a prevailing intrinsic structural strain. We assume that strain in the film prevents the NiH4-complexes from fully relaxing back into an ordered arrangement as in the monoclinic structure. Concluding it can be said that magnetron sputtering gave somehow different results from those received for powder material and this “structure locking” can by typical artefact for the thin films not achievable by the ball milling technique.