It was an astonishing discovery that LaFeAs(O,F) (La1111) shows a high superconducting transition temperature TC of 26K. 1) The enhancement of TC by substituting lanthanoid elements was also amazing as well as the discovery of La-1111. TC comes up to 56K presently in Sm-1111, and it is a refreshing surprise that such a variety of superconductors with high TC containing a typical magnetic element, i.e., iron, appeared in a short period of time. The first pressure study was performed by Takahashi et al. in La-1111. The most interesting result in this study was the marked increase in TC. They show that the onset of TC is rapidly enhanced to 43K in optimally doped LaFeAsO0:89F0:11 at 4GPa. Further application of pressure causes a gradual decrease in TC, and consequently the onset temperature of superconducting transition is suppressed to 9K at 30GPa. They also investigated the pressure dependence of underdoped LaFeAsO0:95F0:05. According to this study, underdoped La-1111 also shows TC enhancement by pressure. However, the increase in TC is smaller than that observed in optimally doped LaFeAsO0:89F0:11. This result indicates that there is a doping dependence in TC shift under high pressure; in other words, pressure is a parameter independent of doping level. We can refer to other highpressure studies in Ln-1111 (Ln; lanthanoid) except La-1111 by Zocco et al. (Ce-1111) and Yi et al. (Sm-1111, Nd1111). In these studies, TC shows a monotonic decrease under pressure. When we think of the TC of optimally doped oxygen-deficient Ln-1111, TC is almost constant at approximately 52K when we choose Ln 1⁄4 Nd, Sm, Eu, Gd, Tb, and Dy. According to an earlier study, Ce-1111, which has a TC of 41K 7) intermediate between those of La-1111 and Pr-1111, shows a monotonic negative pressure dependence of TC in fluorine-substituted CeFeAsO0:88F0:12. 4) Unfortunately, this result is not clear below 4GPa where La-1111 shows a marked TC enhancement. Moreover, the Ce ion is known as a candidate that induces to pressure-induced valence transition. CeFePO, which has a ZrCuSiAs type crystal structure, the same as Ln-1111, also presents the behavior of a heavy-fermion metal. For these reasons, we have carefully investigated the pressure dependence of TC in oxygen-deficient Ce-1111, especially below 4GPa. We prepared two oxygen-deficient Ce-1111 samples in this study. Both samples are polycrystalline and synthesized by a high-pressure and high-temperature method. One sample (CeFeAsO1 y, ]1) shows superconductivity at TC 1⁄4 40K and the other (]2) shows no superconductivity because of the difference in oxidization between their starting materials. By considering these characteristic features of these two samples, it is reasonable that ]1 is optimally doped and ]2 has a close-to-stoichiometric parent composition. The hydrostaticity of pressure is very important to observe the intrinsic behaviors of physical properties under pressure because Ln-1111 has inhomogeneous compressibility, which originates from its layered crystal structure. Therefore, we employed a cubic anvil apparatus with Daphne 7474 oil as pressure-transmitting medium in this study. Electrical resistivity was measured by a conventional DC 4-wire method to determine the transition temperature TC under a high pressure of up to 15GPa. Pressure always shifts upward 0.2 or 0.3GPa during warming above approximately 200K because of inside friction. We can see systematic resistivity anomalies in our study at approximately 200K caused by such an increase in pressure, which are not intrinsic. Figure 1 shows the resistivity of superconductive CeFeAsO1 y (]1) under high pressures. First, TC decreases very rapidly with pressure below 4GPa with a large transition width. In Nd-1111, we see no broadening and TC decreases much more slowly under pressure. Carrier doping level is not affected by pressure in underdoped Nd1111, TC decreases apparently with crystal structural factor under pressure in Nd-1111. Thus, the superconductivity of Ce-1111 is suppressed principally not by crystal structural factor but by the other reason. At 4.5GPa, the resistivity drop still remains; however, no zero resistivity is observed. Therefore, bulk superconductivity disappears at this pressure. One possible reason for such a distinct behavior is the valence transition of the Ce ion. The Ce ion is known for valence transition from Ce3þ to Ce4þ, which is sometimes induced by pressure, for example, the pressure-induced enhancement of superconductivity takes place in CeCu2Si2. 8,9) This speculation also suggests a possible means of continuous carrier doping using pressure0 50 0 0.1 0.2 0.3
Read full abstract