Tuning the static spin-stripe phase and superconductivity in La_(2-x)Ba_xCuO_4 (x = 1/8) by hydrostatic pressure
Z. Guguchia, A. Maisuradze, G. Ghambashidze, R. Khasanov, A. Shengelaya, H. Keller
ppreprint(November 4, 2018)
Tuning the static spin-stripe phase and superconductivity inLa − x Ba x CuO ( x = 1/8) by hydrostatic pressure Z. Guguchia, ∗ A. Maisuradze,
1, 2
G. Ghambashidze, R. Khasanov, A. Shengelaya, and H. Keller Physik-Institut der Universit¨at Z¨urich,Winterthurerstrasse 190, CH-8057 Z¨urich, Switzerland Laboratory for Muon Spin Spectroscopy,Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland Department of Physics, Tbilisi State University,Chavchavadze 3, GE-0128 Tbilisi, Georgia
Abstract
Magnetization and muon spin rotation experiments were performed in La − x Ba x CuO ( x = 1/8)as a function of hydrostatic pressure up to p (cid:39) . Ba . CuO and thatthese phenomena occur in mutually exclusive spatial regions. The present results also reveal thatthe static spin-stripe phase still exists at pressures, where the long-range low-temperature tetrag-onal (LTT) structure is completely suppressed. This indicates that the long-range LTT structureis not necessary for stabilizing the static spin order in La . Ba . CuO . PACS numbers: 74.72.-h, 74.72.Dn, 75.30.Fv, 74.62.Fj ∗ Electronic address: [email protected] a r X i v : . [ c ond - m a t . s up r- c on ] S e p a − x Ba x CuO (LBCO) was the first cuprate in which high- T c superconductivity was dis-covered [1]. The undoped parent compound is an antiferromagnetic (AFM) insulator. Thereplacement of La by Ba ions, through which holes are doped into the CuO planes,causes the destruction of AFM order and superconductivity appears at x = 0.06. Subsequentinvestigations showed that there exists a sharp dip in the T c - x phase diagram, indicatingthat bulk superconductivity is greatly suppressed in a narrow range around a particulardoping concentration x = 1/8 in LBCO [2]. This suppression of T c has attracted a greatdeal of attention and is known in the literature as the 1/8 anomaly (see e.g. , [3, 4]). Latera similar anomaly was also observed in rare earth doped La − x Sr x CuO . Studies of thecrystal structure clarified that the LBCO system undergoes at x = 1/8 a first-order struc-tural phase transition from a low-temperature orthorhombic (LTO) to a low-temperaturetetragonal (LTT) phase [5]. Since the structural transition to the LTT phase appears nearthe Ba concentration x where the strong decrease of T c occurs, it has been suggested thatthere is a correlation between the appearance of the LTT phase and the suppression ofsuperconductivity [5]. Muon spin rotation ( µ SR) experiments detected the appearence ofstatic magnetic order below ∼
30 K in La . Ba . CuO (LBCO-1/8) [6].The discovery of elastic superlattice peaks in La . Nd . Sr . CuO by neutron diffractionprovided evidence of two-dimensional charge and spin order, which was explained in termsof a stripe model where charge-carrier poor AFM regions are separated by one-dimensionalstripes of charge carrier-rich regions [7, 8]. The presence of stripe-like charge and spindensity ordering is believed to be responsible for the anomalous suppression of superconduc-tivity around x = 1/8 in cuprates [7, 8]. The existence of stripes in La . Sr . CuO andBi Sr CaCu O y has also been demonstrated by extended x-ray absorption fine structure(EXAFS) experiments which allow to probe the local structure near a selected atomic site[9, 10].The fascinating issue of charge and spin stripes in cuprate superconductors has attracteda lot of attention for many years (see e.g. , [3, 4]). Experimental results and theoreticalconsiderations show that the modulations of the lattice and of the charge and spin densityappear to be both ubiquitous in the cuprates and intimately tied up with the physics of thesematerials [3, 4]. However, the role of stripes for superconductivity in cuprates is still unclearat present. Therefore, it is important to find an external control parameter which allowsto tune structural and electronic properties of the cuprates and study the relation between2uperconductivity and stripe order. It is known that upon applying hydrostatic pressureboth the LTT and LTO structural phase transition in LBCO-1/8 are suppressed completelyat the critical pressure p c ≈ p c [15]. Hence, it is not known how the static spin-stripe order changes across p c .Here, we report studies of superconductivity and stripe magnetic order in LBCO-1/8under hydrostatic pressure up to p (cid:39) µ SR) experiments. It was observed that the transition temperature of the stripe magneticorder and the size of the ordered moment are not significantly changed by pressure. Butthe volume fraction of the magnetic phase significantly decreases and simultaneously thesuperconducting (SC) volume fraction increases with increasing pressure. This indicatesthat magnetic regions in the sample are converted to SC regions with increasing pressure,providing evidence for a competition between superconductivity and static magnetic orderin the stripe phase of LBCO-1/8. It was also demonstrated that the spin-stripe order stillexists at pressures, where the LTT phase is suppressed.One polycrystalline sample of La − x Ba x CuO with x = 1/8 was prepared by the conven-tional solid-state method. All the measurements were performed on samples from the samebatch. The single-phase character of the sample was checked by powder x-ray diffraction.The magnetic susceptibility was measured under pressures up to 2.1 GPa by a SQUIDmagnetometer ( Quantum Design
MPMS-XL). Pressures were generated using a diamondanvil cell (DAC) [16] filled with Daphne oil which served as a pressure-transmitting medium.The pressure at low temperatures was determined by the pressure dependence of the SC tran-sition temperature of Pb. The temperature dependence of the zero-field-cooled (ZFC) andfield-cooled (FC) magnetic susceptibility, χ ZFC and χ FC , respectively, for LBCO-1/8 in amagnetic field of µ H = 0.3 mT is shown in Fig. 1a. The diamagnetic susceptibility exhibitsa two-step SC transition. The first transition with an onset at T c1 ≈
30 K corresponds toonly about 4 % volume fraction of superconductivity estimated from ZFC magnetizationat 10 K. The second SC transition is observed at T c2 ≈
10 K, with a larger diamagneticresponse. However, the volume fraction of the low temperature SC phase is still small atambient pressure and amounts to about 10 % of full shielding at 2 K. A two-step SC tran-sition, starting at around 30 K with a weak diamagnetic response was observed previouslyin polycrystalline LBCO-1/8 [2, 11]. It was explained as some kind of filamentary super-3
IG. 1: (Color online) Temperature dependence of the magnetic susceptibility of LBCO-1/8measured at ambient pressure without pressure cell (a) and at various applied hydrostatic pressures(b) in a magnetic field of µ H = 0.3 mT. The vertical gray lines and the arrows denote thesuperconducting transition temperatures T c1 and T c2 (see text for an explanation). conductivity due to the presence of a very small fraction of the LTO phase. Recent detailedtransport and susceptibility measurements in single crystal of LBCO-1/8 provided evidenceof the intrinsic nature of the observed two-step SC transition [17]. It was found that a SCtransition at higher temperature T c1 is present when the magnetic field is applied perpen-dicular to the CuO planes. The SC transition at low temperature T c2 is more pronouncedwhen the magnetic field is parallel to the planes ( H (cid:107) ab ). The authors interpreted the4ransition at T c1 as due to the development of 2D superconductivity in the CuO planes,while the interlayer Josephson coupling is frustrated by static stripes. A transition to a 3DSC phase takes place at much lower temperature T c2 (cid:28) T c1 , reflected as a strong increaseof diamagnetism below T c2 for H (cid:107) ab . For polycrystalline samples with random orientationof grains these two temperatures will result in two SC transitions as observed in presentexperiments (see Fig. 1a).We studied the SC transition in LBCO-1/8 as a function of hydrostatic pressure. Mea-surements were performed in the FC mode at 0.3 mT, which was set constant during themeasurements at all pressures in order to avoid a variation of the applied field during themeasurements with different pressures. Figure 1b shows the temperature dependence of χ FC for different pressures after substraction of the background signal from the empty pressurecell. A two-step SC transition is observed at all pressures, except at the highest appliedpressure of 2.1 GPa, where a three-step SC transition is visible. The reason for this isnot clear at present. Further investigations, in particular on single crystals, are needed toclarify this issue. It was found that T c2 increases only slightly with pressure from 10 K toabout 12 K at the maximal pressure applied in our experiments ( p = 2.1 GPa). On theother hand, T c1 shows a significant increase with a rate of 6.2 K/GPa. It is interesting thatthe volume fraction of the corresponding SC phase is also strongly enhanced with appliedpressure (see Fig. 1b). These results are in agreement with previous studies showing thatsuperconductivity in LBCO-1/8 is largely enhanced by applying pressure [11, 13, 18].It is interesting to explore the pressure effect on spin order in the stripe phase and itsrelation to superconductivity. It is also of great interest to study the relation betweenstatic magnetism and the LTT phase in LBCO-1/8. However, to the best of our knowledgemagnetism in LBCO-1/8 was studied only at low pressures [15] where the LTT phase is stillpresent. In order to answer this question we performed zero-field (ZF) µ SR experimentsin LBCO-1/8 at ambient and various hydrostatic pressures, including pressures where thelong-range LTT structure is suppressed. ZF µ SR is a powerful tool to investigate microscopicmagnetic properties of solids without applying an external magnetic field. It is especiallysuitable for the study of weak magnetic order, since the positive muon is an extremelysensitive local probe which is able to detect small internal magnetic fields and orderedvolume fractions.The ZF µ SR experiments were carried out at the µ E1 beam line at the Paul Scherrer In-5
IG. 2: (Color online) ZF µ SR signal A (t) of LBCO-1/8 measured at p = 0 GPa (a), and 2.2GPa (b), recorded for two different temperatures: T = 4 K (circles) and T = 32 K (squares). Thesolid lines represent fits to the data by means of Eq. (1). stitute, Switzerland. Pressures up to 2.2 GPa were generated in a double wall piston-cylindertype of cell made of MP35N material, especially designed to perform µ SR experiments underpressure [19]. As a pressure transmitting medium Daphne oil was used. The pressure wasmeasured by tracking the SC transition of a very small indium plate. The µ SR time spectrawere analyzed using the free software package MUSRFIT [20].Figure 2 shows representative ZF µ SR time spectra for a polycrystalline LBCO-1/8 sam-ple at ambient and at maximum applied pressure p = 2.2 GPa, respectively. Below T ≈
30 K damped oscillations due to muon-spin precession in local magnetic fields are observed,indicating static spin-stripe order [6, 21].A substantial fraction of the µ SR asymmetry signal originates from muons stoppingin the MP35N pressure cell surrounding the sample. Therefore, the µ SR data in the wholetemperature range were analyzed by decomposing the signal into a contribution of the sampleand a contribution of the pressure cell: A ( t ) = A S (0) P S ( t ) + A P C (0) P P C ( t ) , (1)where A S (0) and A P C (0) are the initial asymmetries and P S (t) and P P C (t) are the muon-spinpolarizations belonging to the sample and the pressure cell, respectively. The pressure cellsignal was analyzed by a damped Kubo-Toyabe function [19]. The response of the sample6
IG. 3: (Color online) (a) Temperature dependence of the average internal magnetic field B µ atthe muon site of LBCO-1/8 recorded at various applied pressures. The solid lines represent fitsof the data to the power law described in the text. The arrows mark the transition temperaturesfor the static spin-stripe order T so . The inset shows T so as a function of pressure p . (b) Thetemperature dependence of the magnetic volume fraction V m in LBCO-1/8 at ambient and varioushydrostatic pressures. The solid lines are fits of the data to a similar empirical power law as usedfor B µ ( T ) in (a). consists of a magnetic and a nonmagnetic contribution: P S ( t ) = V m (cid:20) e − λ T t J ( γ µ B µ t ) + 13 e − λ L t (cid:21) + (1 − V m ) e − λ nm t . (2)Here, V m denotes the relative volume of the magnetic fraction, and γ µ / (2 π ) (cid:39) . B µ is the average internal magnetic field at the muon site. λ T and λ L are the depolarization rates representing the transversal and the longitudinalrelaxing components of the magnetic parts of the sample. J is the zeroth-order Bessel7unction of the first kind. This is characteristic for an incommensurate spin density waveand has been observed in cuprates with static spin stripe order [21]. λ nm is the relaxation rateof the nonmagnetic part of the sample. The total initial assymetry A tot = A S (0) + A PC (0) (cid:39) A S (0)/ A tot (cid:39) B µ for different pressures is shown in Fig. 3a. The solidcurves in Fig. 3a are fits of the data to the power law B µ ( T ) = B µ (0)[1-( T /T so ) γ ] δ , where B µ (0) is the zero-temperature value of B µ . γ and δ are phenomenological exponents. Thevalues of the spin ordering temperature T so (cid:39)
30 K and B µ (0) (cid:39)
25 mT at ambient pressureare in good agreement with the values of a previous µ SR study [15, 21]. As evident fromFig. 3a the internal magnetic field B µ (0) is almost pressure independent. This indicatesthat the ordered magnetic moment of the static stripe phase does not depend on appliedpressure. Also T so changes only slightly with pressure as shown in the inset of Fig. 3a. Inthe pressure range of p = 0 - 2.2 GPa, T so ( p ) varies only between 30 and 27 K with a shallowminimum at p (cid:39) p c = 1.85 GPa [14]. Therefore, the present results demonstrate that the spinorder due to static stripes still exists at p = 2.2 GPa, where the LTT phase is alreadysuppressed. Recent high pressure x-ray diffraction experiments showed that also the chargeorder of the stripe phase survives above p c in LBCO-1/8 [14]. Combining these results, onemay conclude that both charge and spin order, and consequently the static stripe phaseitself, still exist at pressures where the LTT phase is suppressed.Here the question arises: What is the effect of pressure on the stripe order in LBCO-1/8? In agreement with the previous low-pressure µ SR results [15], it was found that it isthe magnetic volume fraction V m which is significantly suppressed by pressure. µ SR candetermine the ordered volume fraction and is thus a particularly powerful tool to studyinhomogeneous magnetism in materials. Figure 3b shows the temperature dependence of V m at various pressures. V m increases progressively below T so with decreasing temperatureand acquires nearly 100 % at ambient pressure at the base temperature [6]. An importantresult is that at low temperature V m significantly decreases with increasing pressure (seeFig. 3b). This means that with increasing pressure an increasingly large part of the sample8 IG. 4: (Color online) (a) The pressure dependence of the zero-temperature limit of the magneticand the SC volume fractions, V m (0) and V sc (0), respectively, of LBCO-1/8. Solid lines are linearfits to the data. (b) V sc (0) vs. V m (0). The solid straight line is drawn between a hypotheticalsituation of a fully magnetic ( V m (0) = 1) and a fully SC state ( V sc (0) = 1). remains in the nonmagnetic state down to the lowest temperatures.In order to compare the influence of pressure on the SC and magnetic properties ofLBCO-1/8, the pressure dependences of the zero-temperature limit of the magnetic volumefraction V m (0) and the SC volume fraction V sc (0) = - χ ZFC (0) [22] are plotted in Fig. 4a. Notethat V m (0) linearly decreases with pressure to approximately 50 % at p = 2.2 GPa. A linearextrapolation of V m (0) to higher pressures shows that the magnetic volume fraction should9e completely suppressed at p ≈ µ SR or neutron-scattering experiments. It is evident fromFig. 4a that the decrease of V m (0) is followed by an increase of the SC volume fraction V sc (0). In Fig. 4b we plot V sc (0) as a function of V m (0). The solid straight line is drawnbetween a hypothetical situation of a fully magnetic ( V m (0) = 1) and a fully SC state ( V sc (0)= 1). Remarkably, the experimental data lie on this solid straight line. Thus, the sumof the SC and magnetic volume fractions is constant and is close to one. This stronglysuggests that superconductivity does not exist in those regions where static magnetism ispresent. Thus, superconductivity most likely develops in those areas of the sample whichare nonmagnetic down to the lowest temperatures. The latter implies that in LBCO-1/8magnetism and superconductivity are competing order parameters. It is interesting to notethat a similar scaling was found between the superfluid density and the magnetic volumefraction in the related compound La . − y Eu y Sr . CuO [23]. The tuning of the magneticand SC properties was realized by rare-earth doping.To summarize, magnetism and superconductivity was studied in LBCO-1/8 by meansof magnetization and µ SR experiments as a function of pressure up to p (cid:39) V m (0).Simultaneously, an increase of the SC volume fraction V sc (0) occurs. Furthermore, it wasfound that V m (0) and V sc (0) at all p are linearly correlated: V m (0) + V sc (0) (cid:39)
1. Thisis an important new result, indicating that the magnetic fraction in the sample is directlyconverted to the SC fraction with increasing pressure. The mechanism of this transformation,however, is not clear yet and requires further studies. The present results provide evidence fora competition between bulk superconductivity and static magnetic order in the stripe phaseof LBCO-1/8, and that static stripe order and bulk superconductivity occur in mutuallyexclusive spatial regions. Our findings suggest that a pressure of about 5 GPa would besufficient to completely suppress the static stripe phase and restore bulk superconductivityin LBCO-1/8. 10he µ SR experiments were performed at the Swiss Muon Source, Paul Scherrer Insti-tute (PSI), Villigen, Switzerland. This work was supported by the Swiss National ScienceFoundation, the NCCR MaNEP, the SCOPES grant No. IZ73Z0-128242, and the GeorgianNational Science Foundation grant RNSF/AR/10-16. [1] Bednorz J G and M¨uller K A 1986
Z. P hys. B P hys. Rev. B Rev. M od. P hys. Adv. P hys. P hys. Rev. Lett. P hysica C N ature ( London ) P hys. Rev. B P hys. Rev. Lett. P hys. Rev. B P hys. Rev. B P hys. Rev. B , 104513[13] Zhou J S and Goodenough J B 1997 P hys. Rev. B P hys. Rev. Lett.
15] Satoh K H, Goko T, Takeshita S, Hayashi Y, Arai J, Higemoto W, Nishiyama K and NagamineK 2006
P hysica B Rev. Sci. Instrum. P hys. Rev. B P hysica C P hys. Rev. B P hysics P rocedia P hys. Rev. B χ FC under high pressure. The difference χ ZFC (0)- χ FC (0) was obtained from the measurements at ambient pressure (see Fig. 1a). By assumingthat this difference does not change with pressure, the value of χ ZFC at T = 0 K was determinedat different applied pressures.[23] Kojima K M, Uchida S, Fudamoto Y, Gat I M, Larkin M I, Uemura Y J and Luke G M 2003 P hysica B316-320