Abstract
The results of Matsubayashi et al. on nominally stoichiometric, CaB6 sintered powders arrive at a conclusion concerning our study1 on ferromagnetism in this and related materials that differs in several ways from the conclusions of ongoing studies of aluminium-flux-grown single crystals (Z.F., S. Nakatsuji, F. Drymiotis, M. Bizimis, S. Yeo, J.D.T., A. Bianchi and H.R.O., unpublished results).
Similar content being viewed by others
Young et al. reply
At issue is whether alien Fe–B phases are responsible for all of the observed ferromagnetism. Our conclusion is partly contained in ref. 2, in which it is stated that strongly interacting defects in off-stoichiometric CaB6 (that is, comprising a few-tenths of a per cent) that carry magnetic moments are responsible for the observed ferromagnetic properties. These defects may be iron atoms scavenged during the growth of CaB6 crystals from boron-rich flux growths. We outline experimentally based arguments (Z.F. et al., unpublished results) against the suggestion of Matsubayashi et al. that extrinsic Fe–B phases are the source of the ferromagnetism here. Almost all studies show the surface moments to be removed in acid solution, and these will not be discussed further.
In single crystals of CaB6 grown from molten aluminium flux (starting composition for Ca:B, 1:1–1:9) with added iron (Ca:Fe, 1:0.01–1:1), high-temperature ferromagnetism is generally observed in crystals with Ca:B ratios greater than 1:6; the magnitude of the ordered moment is 3–5 e.m.u. per mol CaB6 for all iron concentrations in the flux (Z.F. et al., unpublished results). This lack of dependence of the measured ordered moment on the iron concentration in the flux suggests that alien Fe–B phases are not the source of ferromagnetism. Also, ferromagnetic CaB6 crystals have a metallic electrical-resistivity characteristic (dρ/dT > 0; ref. 2) that is consistent with the introduction of defects into the material.
If cobalt or nickel is added to the flux in concentrations equal to those of added iron, the crystals of CaB6 show small-moment ferromagnetism only below about 10 K (Z.F. et al., unpublished results), indicating that cobalt or nickel enters CaB6 as low-concentration defects in preference to iron, and that cobalt and nickel do not magnetically interact as strongly as iron defects.
A picture emerges in which iron is involved in the weak high-temperature ferromagnetism of CaB6 and in which off-stoichiometric growth of CaB6 from excess boron melting results in the scavenging of iron from the flux, giving rise to crystals of CaB6 containing about 0.1 atomic per cent Fe. Theoretical3 and experimental4,5 studies of CaB6 indicate that the stoichiometric material is a semiconductor. This suggests that the physics here shares many similarities with that of manganese-doped GaAs. It will be interesting to determine the chemistry of the iron defects, as well as how they interact so strongly and conspire to magnetic order at higher temperatures.
References
Young, D. P. et al. Nature 397, 412–414 (1999).
Fisk, Z., Ott, H. R., Barzykin, V. & Gor'kov, L. P. Physica B 312–313, 808–810 (2002).
Tromp, H. J., van Gelderen, P., Kelly, P. J., Brocks, G. & Bobbert, P. A. Phys. Rev. Lett. 87, 16401/1–4 (2001).
Vonlanthen, P. et al. Phys. Rev. B 62, 10076–10082 (2000).
Denlinger, J. D. et al. Phys. Rev. Lett. 89, 157601/1–4 (2002).
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Young, D., Fisk, Z., Thompson, J. et al. Parasitic ferromagnetism in a hexaboride?. Nature 420, 144 (2002). https://doi.org/10.1038/420144a
Issue Date:
DOI: https://doi.org/10.1038/420144a
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
