A First-Order Phase Transition Induced by a Magnetic Field

Abstract

An electronic (metal-to-insulator) phase transition of the first order, which can be caused by an external magnetic field, was discovered in Nd, , , Sr, , MnO, .A clear hysteresis was observed during the increase and decrease of an external magnetic field at a fixed temperature. The hysteretic field region was observed to depend critically on temperature and to drastically expand with a decrease of temperature, perhaps as a result of suppression of the effect of thermal fluctuations on the first-order phase transition. Although it has seldom been observed, this is thought to be a generic feature of the first-order phase transition at low temperatures near 0 kelvin. I n general, a first-order phase transition is accompanied by hysteresis; that is, a difference of the transition points as a thermodvnamic variable that drives a ohase transition is increased and decreased ( 1 ) . A well-known example is the melting of a solid, which often shows a supercooled (or superheated) state as a metastable state below (or above) the normal melting (or solidification) temperature. Here we describe a first-order phase transition induced by an external magnetic field in one of the man-" ganese oxide compounds with perovskitetype structure, Ndl12Srll~Mn03. On the basis of a peculiar phase diagram in the temperature field plane, a novel reentrant, irreversible behavior of the insulator-tometal transition is demonstrated. Figure 1 shows the temperature dependence of the magnetization, lattice parameters, and resistivity of the Ndl12Srl12Mn03 crystal. As a generic feature of hole-doped perovskites of manganese oxide, the crystal undergoes a ferromagnetic transition at the Curie temperature (T,, 255 K) (2, 3). This is attributable to the so-called double-exchange mechanism (4). which is a ferro-. , magnetic coupling between the local 3d electron (t2 state) spins through kinetic exchange of the itinerant 34 electron (e, state). At temperatures above 160 K, the crvstal shows metallic conductivitv: The resistivity shows an appreciable decrease below T,, owing to reduced magnetic scattering of the charge carriers in the spin-ordered ferromagnetic phase. What is unique in the compound Ndl12Srll,Mn0, is that another distinct phase transition occurs at Tco = 158 K (Fig. 1). (Tco represents the critical temperature for the charge-ordering transition.)When this ~h a s e transition occurs at Tco, the resistivity jumps by more than two orders of magnitude from a typically metallic value (5 X l o p 4 ohm. cm), and the ferromagnetic magnetization disappears, which indicates the simultaneous ferromagnetic-to-antiferromagnetic transition. In accord with these changes in electric and magnetic properties, the lattice parameters in the orthorhombicall distorted perovskite lattice (a -b .-d a P and c -la,; ap relating to the cubic perovskite) show a distinct change, as shown in Fig. 1. According to recent studies ( 5 , 6 ) on the analogous crystal Prl12Srl12Mn03, the transition from ferromagnetic metal to antiferromagnetic insulator can be assigned to t...

Colossal Magnetoresistance and Magnetism of NaCr2O4

“…No field hysteresis was observed [1,12]. These are in sharp contrast to MR in Nd 1−x Ca x MnO 3 or Pr 1−x Ca x MnO 3 ; the manganese oxides show the CMR persisting to a low temperature, but the CMR is caused by a first-order phase transition from charge-and orbital-ordered insulator to ferromagnetic metal [16][17][18][19]. Thus, the CMR in the insulating phase is not very large and the field dependence of magnetoresistance show a large hysteresis.…”

Mean-field theory of charge ordering and phase transitions in the colossal-magnetoresistive manganites

“…(2) The charge-ordered AFO phase can be 'melted' by the application of a weak magnetic field as in experiments [18,20].…”

“…Doping-driven, first-order ferromagnet-to-antiferromagnet transitions (presumably associated with charge ordering) have also been reported for La 1−x Ca x MnO 3 [15,16], though the question of first-order phase coexistence has not been investigated sufficiently, as we discuss later. The transition from a charge-ordered antiferromagnetic insulator (AFO) to a chargedisordered (or charge-non-ordered) ferromagnetic metal (FN) can also be obtained by changing the magnetic field H ; e.g., in both Pr 0.5 Sr 0.5 MnO 3 and Nd 0.5 Sr 0.5 MnO 3 [18,20] the transition temperature T AF→F decreases rapidly with increasing H . There is also growing evidence that charge ordering in Ln 1−x A x MnO 3 is related to the width of the e g band, which is in turn determined by the average radius r A of the A-site cations [13]: if r A 1.18, only charge ordering is observed; if 1.18 r A 1.25, charge ordering is observed at low T but a metallic ferromagnet (FN) appears as T is raised; if 1.25 r A , only an FN phase is observed.…”

“…With the localized spins treated as classical, the energy difference between the parallel and antiparallel alignment of the itinerant spin with respect to the localized spin is given by 2J H S, which is about 3 eV from band calculations [24]. In most of our studies we use the unit cell of the well-known 'CE' structure, a ubiquitous chargeand spin-ordered structure in the manganites observed for x 0.5 [2,12,18,19]. We define the scaled couplings J ≡J S 2 , J H ≡J H S, and H ≡ |H|S, and use the following as typical parameters:…”

“…In some cases, as the temperature T is increased, insulator-metal transitions are seen; these are driven by the destruction of charge ordering, reminiscent of the Verwey transition in magnetite [17], but, unlike the case for magnetite, with an accompanying antiferromagnet-to-ferromagnet transition. For instance, at low T , Nd 0.5 Sr 0.5 MnO 3 is a charge-ordered antiferromagnet that has the complex CE structure [3] and undergoes a first-order transition to a charge-disordered ferromagnet when T is raised [2,12,18,19]. Doping-driven, first-order ferromagnet-to-antiferromagnet transitions (presumably associated with charge ordering) have also been reported for La 1−x Ca x MnO 3 [15,16], though the question of first-order phase coexistence has not been investigated sufficiently, as we discuss later.…”