Enhancement of equilibrium fraction by heterogeneous nucleation in a "phase separated" half-doped manganite
aa r X i v : . [ c ond - m a t . s t r- e l ] O c t Enhancement of equilibrium fraction by heterogeneous nucleationin a “phase separated” half-doped manganite.
P. Chaddah, Kranti Kumar and A. Banerjee
UGC-DAE Consortium for Scientific ResearchUniversity Campus, Khandwa RoadIndore-452017, M.P, India. (Dated: December 15, 2018)
Abstract
Glass-like arrest of kinetics has been observed across many magnetic first-order transitions.By traversing the two control variable H-T space, tunable coexisting fractions of arrested andequilibrium phases have been observed. We report here a fortuitous situation in a half-dopedmanganite sample, where these features occur by varying only temperature (T) along the H=0line. The resistivity at 5 K rises by more than a factor of three provided second cool-down iseffected from a specified intermediate T. This significant enhancement results from heterogeneousnucleation during second cool-down of regions that were kinetically arrested during first cool-down. K of these regions lies above the corresponding T*, but we show thatheterogeneous nucleation of the antiferromagnetic insulating phase can be caused at a highertemperature.Many half-doped manganites show a ferromagnetic-metallic (FMM) to antiferromagnetic-insulating (AFI) transition as T is lowered, but cooling in a large field results in the FMMphase being kinetically arrested and persisting to low-T [9, 11, 14]. By choosing an appro-priate H for the cooling process, the frozen FMM fraction can be tuned. It was conjecturedthat for some samples corresponding to figure 5(c) of ref.[14], this coexistence of an unsta-ble FMM phase would be seen even on cooling in zero field. We present here results forone such half-doped manganite (La-Ca-Mn-O) which shows “phase separation” in zero field[18]. As has been found for various materials earlier [6, 9, 10, 11, 15], the T* and T K areanticorrelated in this sample also. This implies that those regions (see above) which havea higher value of T C or of the supercooling spinodal T*, have a lower value of the kinetic2rrest temperature T K [5]. This is depicted in figure 1, which draws on the similar figuresin refs. [5, 6, 10].As the sample is cooled in zero H towards point A, we have a homogeneous FMM phaseeven though it is metastable. At point A, the entire sample is above the supercoolingspinodal and no homogeneous nucleation takes place. Since regions corresponding to W-band get kinetically arrested at point B, before there is any homogeneous nucleation, theseregions remain FMM at the lowest temperature of point C. But regions corresponding tobands X,Y and Z have got converted to AFI at this point C since their respective spinodalT* is higher than their corresponding T K . We now warm the sample to point B. We retainX, Y and Z in AFI phase, but W is kinetically arrested. We now warm toward point A.W-band is no longer kinetically arrested, is above its supercooling spinodal, but is sittingin an environ where many nuclei of AFI phase exist corresponding to bands X, Y and Z.There is now a possibility of FMM phase of band W undergoing heterogeneous nucleationand converting to the AFI phase. We show through resistivity measurements that sucha heterogeneous nucleation actually takes place. If the fraction in W band is just aroundpercolation threshold, then its conversion from FMM to AFI results in a sharp rise inresistivity.Figure-2 shows the resistivity measurement in zero field. The inset shows the thermalhysteresis across the first-order FMM to AFI phase while cooling from 320 K to 5 K andagain heating from 5 K. However, in this cooling process, the complete transformation tothe AFI state has not taken place even after approaching the lowest temperature. This isevident from the main panel of figure-2 where instead of heating all the way from 5K to 320K the sample is heated up to 150 K and cooled back again to 5K. A spectacular increasein resistivity takes place, giving rise to more than three times increase in resistivity at thelowest temperature. The temperature 150 K was chosen based on detailed data obtained byspanning the H-T space [18]. As discussed above, we attribute this rise in resistivity to theincrease in the AFI phase fraction. This additional AFI phase was formed by heterogeneousnucleation.In this report we have studied the fortuitous situation envisaged in ref. [14], where coolingeven in H=0 gives a coexistence of glassy FMM with equilibrium AFI phase at low-T. Whilea homogeneous (but kinetically arrested) FMM phase can be obtained by cooling in a largeH, a homogeneous AFI phase appears not realizable. By the process of heating to just above3he T K band, we have invoked heterogeneous nucleation and enhanced the equilibrium AFIphase. The de-arrest just above T K used here corresponded to the “softening” of glass.Cooling in high field traps a much larger fraction of FMM glass, and warming in zero fieldwould cause regions Z, Y,.. to cross T K while still below the corresponding T*. The de-arrest would now transform glassy FMM to AFI, and these measurements will report [18]analogies to “shattering” of glass.We thank S. B. Roy for discussion. DST, Government of India is acknowledged forfunding 14 Tesla PPMS. [1] M. A. Manekar et al., Phys. Rev. B , 104416 (2001).[2] K. J. Singh et al., Phys. Rev. B , 094419 (2002).[3] M. K. Chattopadhyay et al., Phys. Rev. B , 214421 (2004).[4] M. K. Chattopadhyay et al., Phys. Rev. B , 180401R (2005).[5] P. Chaddah, A. Banerjee and S.B. Roy, cond-mat 0601095.[6] Kranti Kumar et al., Phys. Rev. B , 184435 (2006).[7] S. B. Roy et al., Phys Rev B , 012403 (2006).[8] A. Banerjee et al., Phys. Rev. B , 224445 (2006).[9] A. Banerjee et al., J. Phys.: Condens. Matter , L605 (2006).[10] S. B. Roy et al., Phys. Rev. B , 184410 (2007).[11] R. Rawat. et al., J. Phys.: Condens. Matter , 256211 (2007).[12] S. B. Roy and M K Chattopadhyay, Euro. Phys. Lett. , 47007 (2007).[13] V. K. Sharma et al., Phys. Rev. B (in press).[14] P. Chaddah and A. Banerjee, cond-mat/0703140.[15] P. Kushwaha et al., Cond-mat/0707.0950.[16] P. Chaddah and S. B. Roy, Phys. Rev. B , 11926 (1999).[17] S. B. Roy et al, Phys. Rev. B , 9191 (2000).[18] A. Banerjee et al., to be published. H * , T * ) H T ( H K , T K ) FMMAFI W XY Z Z Y X c W ABC
FIG. 1: The high-T phase is FMM and the low-T phase is AFI. The phase transition would occurin a (H C , T C ) band, but the FMM phase can be supercooled to the (H*, T*) band. Glass-likearrest of kinetics will occur at (H K , T K ) band. Anticorrelation between supercooling and kineticarrest is assumed (see ref. [5, 6, 10] for details). Points A, B and C are as referred to in text.
23 41