Flux Growth and Magnetic Properties of CaB6Crystals

Abstract

Lanthanum doped calcium hexaboride, Ca1 xLaxB6, attracts a lot of interest because of high temperature weak ferromagnetism though the material consists of no magnetic elements. However, there have been some doubts about this material. For example, some of the samples did not have ferromagnetism. Impurities in the samples were not examined well in spite of weak ferromagnetism. In addition, procedures of the sample preparation were not reported in detail in the reports. Therefore, we prepared single crystals of Ca1 xLaxB6 while paying attention to a main impurity, iron, and examined the relation between the impurity and the ferromagnetism. Undoped and lanthanum-doped CaB6 crystals were prepared by the flux method. The CaB6 starting material was synthesized by reaction of Ca(OH)2 and boron. 3) Aluminium metal, Al, was of nominal purity of 4N. The CaB6 and aluminium, 50–52 g in a total weight, were put into an AlN crucible. A molar ratio of CaB6 to Al was 0.6%. Ca was also added as a form of Ca(OH)2 into the crucible in order to suppress formation of AlB2 crystals. The amount of Ca was 20mol% of the CaB6 solute. Lanthanum was doped by adding commercial LaB6 powder. The samples were kept at 1450 C in a flow of argon for 24 h, slowly cooled to 800 C by 15 C every hour, and then quenched. The aluminium flux was removed in NaOH solution. Magnetization of undoped calcium hexaboride CaB6 and lanthanum doped calcium hexaboride Ca1 xLaxB6 (x 1⁄4 0:0051) single crystals were measured by using a SQUID magnetometer up to fields of 55 kG. The obtained crystals were less than several mm in size and 0.2–1mm thick. The shapes were rectangular plate-like or cubic. The large crystals were shown in Fig. 1. The dope of lanthanum did not influence the size and shape of the CaB6 crystals. More than half of both the undoped and La-doped crystals were attracted or slightly moved by a magnet. Especially, the thin crystals were strongly attracted. The cubic crystals were hardly attracted. However, all the crystals were not attracted at all after being kept in HCl solution for a few hours. Table I shows contents of iron impurity, which came from the aluminium flux. The undoped and La-doped CaB6 crystals used for magnetic measurement contained 190 ppm and 100 ppm, respectively, before the HCl treatment. The iron was removed by the HCl treatment, which means that the iron existed on the surface of the CaB6 crystals. This phenomena can be understood as follows. In the process of dissolving the aluminium flux by the NaOH solution, the aluminium matrix acts as a cathode. The crystals which are held on the surface of the aluminium acts as an anode and iron ions were reducted to iron on the crystal surface. That is, the crystals were electrochemically plated with iron. Consequently, the crystals were covered with iron because iron is stable in the NaOH solution. The iron was removed in the HCl treatment because iron reacts with HCl solution. The magnetization of lanthanum doped calcium hexaboride Ca0:0049La0:0051B6 at 300K is shown in Fig. 2. Similar to what has been reported previously, a weak ferromagnetism is observed with hysteresis. The lanthanum doping of 0:51% of our sample is close to the stoichiometry reported for the sample exhibiting the largest magnitude of ferromagnetic component. The magnitude of ferromagnetism ( 2 emu/mol f.u.) is comparable to that reported previously. However, we find that by washing these samples Fig. 1. As-grown CaB6 crystals. One division is 1mm.