Doping of semiconductor nanocrystals by transition-metal ions has attracted tremendous attention owing to their nanoscale spintronic applications. Such doping is, however, difficult to achieve in low-dimensional strongly quantum confined nanostructures by conventional growth procedures. Here we demonstrate that the incorporation of manganese ions up to 10% into CdSe quantum nanoribbons can be readily achieved by a nucleation-controlled doping process. The cation-exchange reaction of (CdSe)(13) clusters with Mn(2+) ions governs the Mn(2+) incorporation during the nucleation stage. This highly efficient Mn(2+) doping of the CdSe quantum nanoribbons results in giant exciton Zeeman splitting with an effective g-factor of approximately 600, the largest value seen so far in diluted magnetic semiconductor nanocrystals. Furthermore, the sign of the s-d exchange is inverted to negative owing to the exceptionally strong quantum confinement in our nanoribbons. The nucleation-controlled doping strategy demonstrated here thus opens the possibility of doping various strongly quantum confined nanocrystals for diverse applications.

The ferromagnetic semiconductor (Ga,Mn)As has emerged as the most studied material for prototype applications in semiconductor spintronics. Because ferromagnetism in (Ga,Mn)As is hole-mediated, the nature of the hole states has direct and crucial bearing on its Curie temperature T(C). It is vigorously debated, however, whether holes in (Ga,Mn)As reside in the valence band or in an impurity band. Here we combine results of channelling experiments, which measure the concentrations both of Mn ions and of holes relevant to the ferromagnetic order, with magnetization, transport, and magneto-optical data to address this issue. Taken together, these measurements provide strong evidence that it is the location of the Fermi level within the impurity band that determines T(C) through determining the degree of hole localization. This finding differs drastically from the often accepted view that T(C) is controlled by valence band holes, thus opening new avenues for achieving higher values of T(C).

No abstract

We discuss magnetic anisotropy parameters of ferromagnetic body-centered cubic (bcc) Fe films grown by molecular beam epitaxy (MBE) on (001) substrates of face-centered cubic (fcc) GaAs, ZnSe, and Ge. High-quality MBE growth of these metal/semiconductor combinations is made possible by the fortuitous atomic relationship between the bcc Fe and the underlying fcc semiconductor surfaces, resulting in excellent lattice match. Magnetization measurements by superconducting quantum interference device (SQUID) indicate that the Fe films grown on (001) GaAs surfaces are characterized by a very strong uniaxial in-plane anisotropy; those grown on (001) Ge surfaces have a fully cubic anisotropy; and Fe films grown on ZnSe represent an intermediate case between the preceding two combinations. Ferromagnetic resonance measurements carried out on these three systems provide a strikingly clear quantitative picture of the anisotropy parameters, in excellent agreement with the SQUID results. Based on these results, we propose that the observed anisotropy of cubic Fe films grown in this way results from the surface reconstruction of the specific semiconductor substrate on which the Fe film is deposited. These results suggest that, by controlling surface reconstruction of the substrate during the MBE growth, one may be able to engineer the magnetic anisotropy in Fe, and possibly also in other MBE-grown ferromagnetic films.

GaAs/Fe/Au core-shell nanowires were grown on GaAs(111)B substrates by molecular beam epitaxy. Scanning electron microscopy images show that the Fe shell has successfully coated the sidewalls of GaAs nanowires. Magnetic anisotropy of GaAs/Fe core-shell nanowires was studied by ferromagnetic resonance and by superconducting quantum interference device magnetometer. The authors’ results show that the magnetic anisotropy of this novel core-shell nanowire system cannot be simply described by any known theory, as revealed by attempts to use micromagnetic simulation using the Object Oriented MicroMagnetic Framework. The observed features thus suggest the existence of a domain structure that is specific to this new system