Enhanced ferromagnetism, metal-insulator transition, and large magnetoresistance in La
12xCa
xMn
12xRu
xO
3free of e
g-orbital double-exchange
M. F. Liu,1Z. Z. Du,1H. M. Liu,1X. Li,1Z. B. Yan,1S. Dong,2and J.-M. Liu1,3,a)
1Laboratory of Solid State Microstructures, Nanjing University, Nanjing 210093, China
2Department of Physics, Southeast University, Nanjing 211189, China
3Institute for Advanced Materials and Laboratory of Quantum Engineering and Materials, South China Normal University, Guangzhou 510006, China
(Received 9 February 2014; accepted 14 March 2014; published online 26 March 2014)
The structure, ionic valences, magnetism, and magneto-transport behaviors of mixed valence oxides La1xCaxMn1xRuxO3 are systematically investigated. The simultaneous substitutions of La3þ and Mn3þ ions by Ca2þ and Ru4þ, respectively, are confirmed by the structural and ionic valence characterizations, excluding the presence of Mn4þ and Ru3þ ions. The enhanced ferromagnetism, induced metal-insulator transition, and remarkable magnetoresistance effect are demonstrated when the substitution level x is lower than 0.6, in spite of the absence of the Mn3þ-Ru4þeg-orbital double-exchange. These anomalous magnetotransport effects are discussed based on the competing multifold interactions associated with the Mn3þ-Ru4þ super-exchange and strong Ru4þ-Ru4þ hopping, while the origins for the metal-insulator transition and magnetoresistance effect remain to be clarified.VC 2014 AIP Publishing LLC.
[http://dx.doi.org/10.1063/1.4869660]
I. INTRODUCTION
A series of exotic physical phenomena in ABO3-type transition metal oxides, with 3dand 4d ions at B sites and thus with strong electron correlation, have been demonstrated in past decades. Extensive researches on these phenomena from various aspects are documented in literature.1,2 The most concerned family in the category of 3d transition metal oxides is probably manganites, i.e., RMnO3 and R1xDxMnO3, whereRis the rare-earth species and Dis the alkaline earth species.3–6For 4dtransition metal oxides, ruth- enates, i.e.,DRuO3whereDis the divalent alkaline-earth spe- cies, have been extensively investigated.7–11It is established that the charge, spin, orbit, and lattice degrees of freedom coexist in these oxides and bring into competitions among multifold interactions which are comparable in energy scale but originate from these different degrees of freedom. These competitions would inevitably result in multi-degenerate states which have similar energy minimal but distinctly differ- ent physical properties. The coexistence and transition of these emergent ordered phases appear to be the common fea- tures in these transition metal oxides and also the core issues of continuous interest.4,12,13
Without loss of generality, one may take two simplest examples for illustration, one from manganites and the other from ruthenates. We start from LaMnO3 which is the end member of RMnO3 family. LaMnO3 has the orthorhombic structure with space groupPbnmand the lattice constants are a¼0.554 nm, b¼0.575 nm, and c¼0.768 nm. The ground state of LaMnO3 is believed to be an A-type
antiferromagnetic (A-AFM) Mott insulator. However, a number of measurements revealed that the A-AFM magnetic structure of LaMnO3is marginal and sufficiently weak per- turbations (disorder, dimensionality, and defects, etc.) enable a transition from the A-AFM order to canted spin structure and partial ferromagnetic (FM) order.14 Furthermore and more importantly, a partial substitution of La3þby a divalent alkaline ion, i.e., La1xDxMnO3(D¼Ba, Sr, Ca etc), drives the Mn3þto Mn4þvalence transitions, leading to the signifi- cantly enhanced ferromagnetism, metal-insulator transition (MIT), and well known colossal magnetoresistance (CMR) effect.4,13,15The Mn3þ-O2–Mn4þdouble-exchange (DE) and electronic phase separation are believed to be responsible for these consequences. Surely, for an over-substitution higher than 0.5 or so, the charge ordering (CO) and orbital ordering (OO) with dominant AFM super-exchange again take over the FM metallic state.4,6 Here, it is noted that the double-exchange acts as the central figure in mediating the magnetic and transport behaviors in La1-xDxMnO3atx<0.5.
Different from the 3d manganites, the 4d ruthenates show even more interesting magnetic and transport proper- ties, partially due to the more extensive spatial extent of the 4d charge distribution and stronger 4d spin-orbit coupling (SOC) effect. In this case, the orbital physics becomes non- negligible. SrRuO3and CaRuO3are the well known ruthen- ates that have been investigated for decades.7–11 While SrRuO3is a well known FM metal with the magnetic Curie point (Tc) at 165 K,7,9,16,17 the ground state of CaRuO3 remains uncertain and it is sensitive to intrinsic fluctua- tions.8,10,18,19 CaRuO3 also has the orthorhombic structure with space group Pnma and the lattice constants are a¼0.536 nm, b¼0.554 nm, and c¼0.766 nm. With the AFM Cuire-Weiss temperature hw 140 K (the AFM
a)Author to whom correspondence should be addressed. Electronic mail:
0021-8979/2014/115(12)/123904/7/$30.00 115, 123904-1 VC2014 AIP Publishing LLC
JOURNAL OF APPLIED PHYSICS115, 123904 (2014)
interaction background), CaRuO3 does not magnetically order until extremely low temperatureT<0.3 K. In addition, the Jahn-Teller (JT) effect in CaRuO3is different from that in LaMnO3and the 4dxyFM orbital ordering induced by the JT effect in CaRuO3was reported.20Therefore, CaRuO3is thought to be a paramagnetic with strong electron correlation on the verge of spin ordering, and thus it is a good platform for understanding the ferromagnetism occurrence in ruthen- ates.8Along this line, substantial efforts have been made in order to reveal the magnetic transitions in CaRuO3upon sub- stitution, pressure, and external stimulus, etc., while so far reported data are very authors-dependent, leaving quite a few of unsolved issues.18,21,22
Given the fact that both LaMnO3and CaRuO3are on the verge states of spin ordering against varying degrees of free- dom, the appearance of any exotic and emergent phenomena in them associated with internal and external stimulus becomes understandable. Recent attention seems to be paid onto the 4d Ru-substitution of LaMnO3and La1xDxMnO3as well as the 3d (Mn, Cr, Fe, etc)-substitutions of CaRuO3.18,21–23 In this case, the substitution induced variations in ionic valence, spin ordering, orbital ordering, lattice distortion, and other local chemical environments are complicated and a clear under- standing of the underlying physics becomes a touch issue. To this stage, an immediate motivation is to begin with an alterna- tive approach to the above mentioned ones so that the substi- tuted materials are as simple as possible in terms of the variations in valence state, magnetic structure, and lattice dis- tortion, etc.
In this work, we intend to address how the magnetic and transport behaviors in (1-x)LaMnO3þxCaRuO3 mixture evolve. This motivation is based on the following three con- siderations: First, this mixture is equivalent to the double-substitution of La3þ and Mn3þby Ca2þ and Ru4þ, respectively. Both LaMnO3and CaRuO3have similar ortho- rhombic lattice structures with the same space groupPnma, and moreover, Ca2þ and Ru4þ have similar ionic sizes as La3þand Mn3þ, respectively. Therefore, a series of homoge- neous solid solutions La1-xCaxMn1-xRuxO3 (LCMRO) over the whole composition range (0x1.0) can be synthe- sized. If this is true, the structural transition/distortion and chemical inhomogeneity if any arising from the substitutions can be avoided to the maximal extent. Second, the valence balance in local lattice units can be maintained without nec- essary valence variations such as Mn3þ-Mn4þtransitions or Ru4þ-Ru5þtransitions for the charge neutrality. Third, both LaMnO3and CaRuO3are on the verges of magnetic states against perturbations and thus their mixtures may generate rich phenomena not available in other manganites and ruthenates.
One expected consequence for mixing LaMnO3 and CaRuO3into LCMRO may be the distinct difference from La1xDxMnO3 in terms of the magnetism and transport behavior. Although the Mn3þhas thet2g3
eg1electronic struc- ture and Ru4þ has the t2g4
eg0
structure, the 4d ions with respect to the 3dions usually have higheregorbital energy.24 Therefore, the Mn3þand Ru4þhave quite differenteglevels although theeglevel of Ru4þis unoccupied. This difference may prohibit any DE sequence associated with the
Mn3þ-O2–Ru4þ pairs. It would be of interest to check whether unusual magnetic and transport behaviors in LCMRO, in terms of the enhanced ferromagnetism, MIT, and magnetoresistance effect, can be observed. We shall investigate systematically the magnetic and transport behav- iors of this simple LCMRO solid solution series over the whole composition range. While our results reveal the absence of Mn and Ru ions other than Mn3þ and Ru4þ in LCMRO solid solutions, the magnetic and transport data indicate the remarkably enhanced ferromagnetism and rela- tively weak MIT upon increasing x up to 0.5. Significant magnetoresistance is identified too in this composition range.
These phenomena suggest that additional mechanisms, dif- ferent from the DE one in La1xDxMnO3 (x<0.5), takes part in mediating the magnetic and transport behaviors in LCMRO.
The remaining part of this article is organized as the fol- lowing. The experimental details are presented in Sec. II.
The structural and valence state characterizations are described in Sec. III with the magnetic and transport data, plus qualitative discussions on the complicated microscopic mechanisms. A brief conclusion is given in Sec.IV.
II. EXPERIMENTAL DETAILS
The La1xCaxMn1xRuxO3 (LCMRO) polycrystalline sample series with x¼0.01.0 were prepared using the usual solid sintering method in air. The highly purified powder of oxides and carbonates was mixed in stoichiomet- ric ratio, ground, and then fired at 1000C for 20 h in air.
The resultant powder was reground and pelletized under a pressure of 1000 psi into disks of 2.0 cm in diameter, and then these pellets were sintered at 1300C for 24 h in air in prior to natural cooling down to room temperature. The sam- ple crystallinity and structure were checked by X-ray diffrac- tion (XRD) with the Cu Karadiation at room temperature.
The refinement of the XRD data was performed using the Rietveld method. The chemical composition and homogene- ity of the samples were characterized by electron probe microanalysis installed with a scanning electron microscopy.
The valence states of Mn and Ru ions were carefully exam- ined using the X-ray photoelectron spectroscopy (XPS) with a photon energy of 1253.6 eV (MgKa) at an energy-step of 0.10 eV for signal probing. Special attention was paid to the Mn2þ, Mn3þ, Mn4þ, Ru3þ, Ru4þ, and Ru5þ. In this case, a thin surface layer for each sample was removed using Ar ion beam in prior to the XPS measurement.
The magnetization (M) as a function of temperature (T) and magnetic field (H) was measured using the Quantum Design superconducting quantum interference device magnetometer (SQUID) in the zero-field cooled (ZFC) and field-cooling (FC) modes, respectively. The cooling field and measuring field are both 1000 Oe. The electro-transport and specific heat were characterized by a physical properties measurement system (PPMS) from the Cryogenic Co. Ltd. and a PPMS from the Quantum Design Inc. The electrical resistivity q as a function ofT and H, including the magnetoresistance MR¼[q(0)-q(H)]/q(0), was measured.
III. RESULTS AND DISCUSSION
A. Structural properties and valence states of Mn and Ru
The lattice units of LaMnO3, CaRuO3, and assumed LCMRO are plotted in Fig. 1, respectively. As mentioned, due to the similar lattice structures, ionic occupations, and one-to-one corresponding ionic sizes between LaMnO3and CaRuO3, the homogeneous LCMRO solid solutions are expected. This scenario is confirmed by the XRDh-2hspec- tra for the samples, as shown in Fig.2(a), while the locally amplified reflections around 2h¼47oand 58oare presented in Figs. 2(b) and 2(c), respectively. Due to the slightly smaller lattice constants of CaRuO3than those of LaMnO3, the gradual and continuous rightward shifting of every reflection with increasingxindicates the homogeneous mix- ing of the two end members, which otherwise would exhibit two separate sets of reflections corresponding to the two members in spite of variations in the diffraction intensity.
The lattice constants evaluated from the XRD spectra with the help of the Rietveld refinement processing are plot- ted in Fig. 2(d) and the unit cell volume V in Fig. 2(e).
Indeed, the three lattice constants (a,b,c) show roughly sim- ilar dependences on x. The constant a remains nearly unchanged in the low-xregime (x<0.5) while bothbandc decrease slightly and linearly. The cell volumeV as a func- tion ofxis basically linear too within the measuring uncer- tainties, following quite well the Vegard’s law.25 It is thus suggested that the valence states of Mn and Ru ions remain unchanged during the mixing, which otherwise would result in an identifiable deviation of the data from the Vegard’s law.
To give further evidence on the invariance of the Mn and Ru ionic valence states, the measured XPS data of eleven samples are plotted in Fig. 3, focusing on the Mn-2p1/2/Mn-2p3/2 and Ru-3p1/2/Ru-3p3/2 binding energy positions.18In the XPS probing, the Mn2O3and RuO2data were taken as the references for identification, as labeled in the plots. While the relative atomic fractions of the Mn3þ/Mn4þand Ru4þ/Ru5þstates may not be quantitatively extracted from these data, we concentrate on the possible peak shifting if any with differentx. Given the fact that the
Mn3þvalence in LaMnO3and Ru4þvalence in CaRuO3are dominant, one can safely conclude that the valence states of these ions remain invariant in all the LCMRO samples because the probed peaks have no identifiable shifting although the peaks intensity varies depending onxin a rea- sonable way. This conclusion is consistent with the above XRD data, and very critical for our qualitative discussion on the magnetic and transport behaviors to be presented below.
FIG. 1. Lattice structures of LaMnO3, CaRuO3, and proposed LCMRO. The ionic sizes are shown for a guide of eyes.
FIG. 2. (a) Measured XRDh-2hspectra for LCMRO, withx¼0.0 to 1.0 as indicated by the dashed arrow. The local amplified (220) and (312) reflec- tions (assigned from LaMnO3) are shown in (b) and (c), respectively. The lattice constants (a,b,c) and unit cell volumeVas evaluated from the data refinement are plotted in (d) and (e), respectively.
B. Magnetic properties
Subsequently, we turn to the magnetic behaviors of these samples. We measured theM(T) dependences of all the samples in the ZFC and FC modes, and the ZFC data are plotted in Figs. 4(a) and 4(b), respectively. As mentioned above, samplex¼0.0 does show weak FM state in addition to the claimed A-AFM ground state, and a FM transition at T160 K is observed. However, the maximal moment is only 1.2lB/f.u., far smaller than 3.8lB/f.u. assuming a fully ferromagnetic Mn3þ alignment. Therefore, a spin-canted AFM structure in samplex¼0.0 can be claimed, as widely reported in literature.26 For the other end sample x¼1.0, a paramagnetic behavior is identified until the lowest T, consistent with earlier report too,8 although CaRuO3 accommodates the AFM interaction background.
Upon increasingx, several features for theM-Tdata can be highlighted. First, in the low-x cases (e.g., x¼0.1), the magnetization is slightly suppressed with respect to the x¼0.0 case, which is reasonable since the Ru4þhas much smaller moment than Mn3þ. Nevertheless, the low-T mag- netization is remarkably reduced, indicating the enhanced AFM tendency. This tendency is expected to be further enhanced due to the AFM interaction of CaRuO3, consider- ing the marginally stable A-AFM state of LaMnO3. Indeed, the low-T magnetization is continuously suppressed with increasing x till x¼0.9 and this reduction can’t be solely attributed to the smaller Ru4þmoment than that of Mn3þ.
Second, in spite of the dominant AFM interaction, clear FM transitions are identified in the high-Trange for samples withxup to 0.7. More surprisingly, the maximal moments for the samples x¼0.1, 0.2, and 0.3 remain similarly 1.11.2lB/f.u., without the predicted continuous decreasing with increasing x. This implies that the FM states in these samples are enhanced rather than being gradually sup- pressed. This serious suppression begins whenxis as high as 0.4, and the maximalM is down to0.25lB/f.u.at x¼0.6, and0.03lB/f.u.atx¼0.8. Third, atx>0.8, the FM state is
completely suppressed and replaced by the marginal AFM state in the low-Trange.
In order to have a quantitative description of the above mentioned effects, we evaluate three characteristic parame- ters as a function ofx, respectively. These parameters are the
FIG. 3. Measured XPS spectra for Mn (2p1/2, 2p3/2) (a) and (b), and for Ru (3p1/2, 3p3/2) (c) and (d) withx¼0.0 to 1.0. The red arrows indicate the bind- ing energy positions for Mn3þ and Ru4þ. The corresponding peaks don’t show any identifiable shifting with varyingx.
FIG. 4. Measured ZFC magnetizationM(T) for LCMRO withx¼0.0 to 1.0 in (a) and (b). The evaluated parametersTc,TN, andhwas a function ofx, respectively, are shown in (c), (d), and (e). The double-head arrows in (c) and (e) indicate the enhanced ferromagnetism regime, while the arrow in (d) indicates the shoulder-like regime of TN(x). The dashed line shows the proposedTN(x), assuming an ideal mixing of LaMnO3and CaRuO3(only for a guide of eyes).
ferromagnetic Curie pointTc, defined as the reflection point of theM(T) curve by differentiating theMagainstTaround the paramagnetic-ferromagnetic transition in the high-T range; the AFM transition pointTN, defined as the reflection point of theM(T) curve by differentiating the M againstT around the AFM transitions in the low-T range; the Cuire- Weiss temperaturehw, determined by fitting the Curie-Weiss formula to theM(T) data in the high-Tparamagnetic state.18 These parameters are plotted, respectively, in Figs.
4(c)–4(e). It is seen that both theTcand the positivehw are surprisingly enhanced at x<0.5, while the TN is decreased continuously, indicating the enhanced ferromagnetism in the regime ofx¼0.00.5, consistent with the above discussion on the magnetization. Abovex0.5, this ferromagnetism is gradually weakened, replaced by the antiferromagnetism characterized by the Tc!0 K and the sign reversal of hw
from positive to negative value at x!1.0. For x¼1.0, hw 80 K is estimated, consistent with earlier report on the dominant AFM interaction in CaRuO3.7,8,18It is noted that a broad bump (or shoulder-like feature) in theTN(x) appears in between x¼0.50.9, consistent with the scenario of enhanced ferromagnetism in betweenx¼0.00.5. The olive dashed line in Fig.4(d)denotes the proposedTN(x) for a sim- ple (1x)LaMnO3þxCaRuO3mixture.
In a short highlight here, the magnetization data reveal enhanced ferromagnetism in the composition region x¼0.00.5 although the two end members are dominated with the AFM interactions. This anomalous behavior sug- gests an additional mechanism which favors the FM inter- action in LCMRO, to be further supported by the transport behaviors presented below.
C. Electrical transport behaviors
We plot the measured electrical resistivityq(T) of a set of samples underH¼0 in Figs. 5(a)and5(b), respectively.
In a general tendency, all the samples except samplex¼1.0 exhibit the insulating/semiconducting behaviors with weak MIT feature aroundT¼200 K in some of them (obvious in samples x¼0.20.5). Besides, sample x¼0.0 shows the well known Mott insulating behavior. The sample x¼1.0 shows the metallic behavior and the q(T) can be well described by q(T)T0.48, consistent with earlier report on CaRuO3.27 It is noted that the q(T) dependence of those high-x samples gives even a broad T-plateau, as identified for samplex¼0.9.
What should be addressed here is the weak but unusual MIT feature in some samples. Obviously due to this MIT, the q(x) dependences in the high-T range are no longer monotonous but exhibit the up-turn or shoulder-like behavior in the high-x regime (x>0.5). The rapid decreasing of the resistivity with increasing x in the low-x regime is under- standable considering that CaRuO3is much more conductive than LaMnO3. To illustrate this unusual shoulder-like fea- ture, we plot theq(x) atT¼100 K and 200 K, respectively in Fig. 5(c). For comparison, the measured M(x) data in the ZFC mode atT¼100 K are plotted in Fig.5(d). This up-turn or shoulder-like effect can be more clearly seen if one looks at theq(x) data atT¼100 K. Assuming that the LCMRO is
an ideal and simple mixture of LaMnO3 and CaRuO3, one may have the dashed lines in Figs.5(c)and5(d)representing the assumed q(x) and M(x) at T¼100 K, which seriously deviate from the measuredq(x) andM(x) in thex¼0.00.6 regime.
An immediate argument as derived from the above description is that the magnetic and transport behaviors in the regimes x¼0.00.5 andx¼0.61.0 are determined by different mechanisms. In the other words, the shoulder-like feature in the q(x) at x>0.5 is induced by the over-rapid drop of q(x) at x<0.5. Similarly, the M(x) plateau in the low-x regime shown in Fig. 5(d) is also unusual. These unusual features suggest the important role of additional Mn3þ-Ru4þ interactions which have not been accessed in earlier relevant works.
D. Magnetoresistance effect
Before discussing the possible mechanisms responsible for the unusualq(x) andM(T) behaviors in the low-xregime (x<0.5), we present the magnetoresistance (MR) data on all the samples. In Fig.6(a)are shown the MR(T) data under a field ofH¼4.0 T for samplex¼0.4 as an example. Theq(T) at H¼0 and M(T) data are inserted for reference. For this sample, significant negative andT-dependent MR effect is recorded. More interestingly, the maximal MR value (30%) appears at the MIT point, around which the FM transition begins. This correspondence finds its counterpart in the CMR manganites La1xDxMnO3, and does suggest the
FIG. 5. Measured zero-field resistivityq(T) for LCMRO withx¼0.0 to 1.0 in (a) and (b). The arrows in (a) and (b) indicate the MITs. The evaluatedq andMas a function ofxat several temperatures are shown in (c) and (e).
The dashed lines guide the proposed dependences, assuming an ideal mixing of LaMnO3 and CaRuO3. The arrow in (c) indicates the shoulder-like regime.
correlations among the FM transition, MIT, and MR effect, bearing in mind no double-exchange between the Mn3þ-Ru4þpairs.
We then look at the MR(x) under H¼4.0 T at several temperatures, and the data atT¼200 K and 250 K are pre- sented in Fig. 6(b). In good consistence with the enhanced ferromagnetism and MIT effect, remarkable MR effect is identified in the regimex¼0.00.5, beyond which it nearly disappears. The maximal MR, appears roughly at x¼0.20.4, very similar to the cases for La1xDxMnO3.
E. Discussion
As well-demonstrated phenomena, the FM transition, MIT effect, and CMR effect in CMR manganites have been understood in the double-exchange scenario.4 The Mn3þ-Mn4þ eg-orbital double-exchange mechanism is believed to be the key for explaining such phenomena. The LCMRO studied here belongs to the alkaline earth doped manganite family, however, theeg-orbital double-exchange process in the Mn3þ-Ru4þ pairs is seriously questioned.
Apart from this, the influences of the strong spin-orbit coupling with the Ru4þ, distinctly different JT effects in LaMnO3 and CaRuO3, and possible Mn3þ-Ru4þ t2g orbital super-exchange may not be excluded.11,14,20,24 Due to the complexity of the interactions, we can only provide a quali- tative discussion on these possibilities and so far no self-consistent conclusion can be reached. In fact, we are not in a position to understand all the observed phenomena, whiles our discussion applies only to some of them.
It is known that Ru4þin DRuO3has its four electrons on thet2g-orbital with spin momentS¼1, and theeg-orbital
is empty, different from Mn3þin RMnO3 which has three electrons on the t2g-orbital and one on the eg-orbital. The Ru4þspins inDRuO3obviously favor the FM interaction so that the electron kinetic energy can be maximized,28 as shown in SrRuO3 and BaRuO3. However, an exceptional case is CaRuO3and the dominant AFM interaction is still puzzling. The Mn3þ-Ru4þinteraction, however, prefers to be AFM in order to stabilize the quantum-confinement effect.29 It is noted that thet2g electrons andegones have strong hybridization with the different 2porbits (prandpP) of O2- so that the Mn3þ-Ru4þ hopping may be possible unless the Mn-O-Ru bond angle is 180. Nevertheless, LaMnO3 is known to have serious lattice distortion with strong quenched disorder.26This effect is somewhat respon- sible for the weak ferromagnetism on one hand, on the other hand allows the Mn3þ-Ru4þhopping to some extent when LaMnO3 is co-substituted by Ca2þ and Ru4þ at the La3þ and Mn3þ sites, respectively, leading to the Mn3þ-Ru4þ super-exchange.
If this scenario is dominant, one expects that the super- exchange can be roughly described by the interaction factor J¼4t2/U witht the exchange bandwidth and Uthe on-site Coulomb energy.30 Two competing consequences will be generated upon the Ru4þ substitution into LaMnO3. One is the reduced Usince the 4dRu4þhas smaller Uthan that of 3dMn3þ. The other is the relatively weak electron hopping, leading to small t. In the low-xregime, the lattice distortion is slightly enhanced with increasingx, as seen from the eval- uated lattice constants in Fig. 2where the in-plane (ab) dis- tortion becomes more serious with increasing x. This effect may reduce the Mn-O-Ru bond angle and thus favor the hop- ping. In this sense, The Ureduction seems more significant than thetreduction, resulting in the gradual enhancement of Tc with increasing x, as shown in Fig. 4(c). In the high-x regime, the rapid decreasing ofTcis understandable consid- ering the weak AFM interaction of CaRuO3. The hw(x) dependence in Fig.4(e) can be understood in the same sce- nario. In accordance, for TN(x), the monotonous and rapid fall in the low-xregime, and the weak rebound (shoulder-like feature) in the high-xregime, also make sense in a qualitative sense.
Theq(x) dependence is however more complicated with respect to the M(x). Basically, the Mn3þ-Ru4þ hopping is very weak in comparison with the Mn3þ-Mn3þhopping, and thus the Ru4þ-substitution of Mn3þwould make the system more insulating, which is contradictory with experimental observation. The reason is essentially associated with the very strong Ru4þ-Ru4þhopping allowing the metallic behav- iors of DRuO3including CaRuO3.9The rapid decreasing of the resistivity with increasingximplies the existence of dom- inant Ru4þ-Ru4þhopping. This suggests that the Ru4þions may have a weak tendency to aggregate into local and dilute clusters although the all the samples are chemically homoge- neous. Therefore, theq(x) shows a monotonous fall, which is confirmed by the data at very low-Trange (T<30 K) shown in Figs.5(a)and5(b).
However, the anomalousq(x) dependences in the inter- mediate T-range (T¼100 K, 200 K), as shown in Fig.5(c), cannot be accounted for by the above mechanism. This
FIG. 6. (a) Measuredq,M, and MR, as a function ofT, respectively, for samplex¼0.4. (b) The evaluated MR(x) curves atT¼200 K and 250 K.
The arrow indicates the regime over which the shoulder-like features of the resistivity andTNappear.
anomaly seems to be related to the weak MIT effect occur- ring in the highT, which together with the MR effect, can’t be reasonably understood at this stage. For the LCMRO, the roles of the SOC effect and JT distortion as well as the con- sequent orbital splitting, would be important in driving the MIT in the high-Trange. These issues definitely deserve for additional investigations.
IV. CONCLUSION
In conclusion, we have investigated systematically the magnetism and electrical transport behaviors of La1xCax Mn1xRuxO3 (0x1) which have been synthesized for excluding the eg-orbital double-exchange. It has been unveiled that the Mn ions and Ru ions maintain the Mnþ3 and Ruþ4 valence states, respectively, excluding the pres- ence of the Mn3þ-Mn4þdouble-exchange. Our data indicate the enhanced ferromagnetism, weak metal-insulator transi- tions, and remarkable magnetoresistance effect in the low substitution levels, although LaMnO3favors the AFM order and CaRuO3is paramagnetic. While the enhanced ferromag- netism may be possibly ascribed to the Mnþ3-Ruþ4hopping and super-exchange, the anomalous metal-insulator transi- tions and associated magnetoresistance remain to be unsolved issues. The present work sheds light into the com- plicated magnetotransport behaviors in 3d-4dmixed perov- skite oxides.
ACKNOWLEDGMENTS
This work was supported by the National 973 Projects of China (Grants No. 2011CB922101), the Natural Science Foundation of China (Grants Nos. 11234005, 11374147, 51332006), and the Priority Academic Program Development of Jiangsu Higher Education Institutions, China.
1L. G. Tejuca, Properties and Applications of Perovskite-Type Oxides (Dekker Publishers, New York, 1993).
2R. H. Mitchell,Perovskites Modern and Ancient(Almaz Press, Thunder Bay, Ontario, 2002).
3J. M. D. Coey, M. Viret, and S. von Molnar, “Mixed-Valence Manganites,”Adv. Phys.48, 167 (1999).
4E. Dagotto, T. Hotta, and A. Moreo,Phys. Rep.344, 1 (2001).
5M. Rini, R. Tobey, N. Dean, J. Itatani, Y. Tomioka, Y. Tokura, R. W.
Schoenlein, and A. Cavalleri,Nature449, 72 (2007).
6S. Cox, J. Singleton, R. D. Mcdonald, A. Migliori, and P. B. Littlewood, Nature Mater.7, 25 (2008).
7J. M. Lango, P. M. Raccah, and J. B. Goodenough,J. Appl. Phys.39, 1327 (1968).
8T. Kiyama, K. Yoshimura, K. Kosuge, Y. Ikeda, and Y. Bando,Phys. Rev.
B54, R756 (1996).
9L. Capogna, A. P. Mackenzie, R. S. Perry, S. A. Grigera, L. M. Galvin, P.
Raychaudhuri, A. J. Schofield, C. S. Alexander, G. Cao, S. R. Julian, and Y. Maeno,Phys. Rev. Lett.88, 076602 (2002).
10H. D. Kim, H. J. Noh, K. H. Kim, and S. J. Oh, Phys. Rev. Lett.93, 126404 (2004).
11M. S. Laad, I. Bradaric, and F. V. Kusmartsev, Phys. Rev. Lett. 100, 096402 (2008).
12V. B. Shenoy and C. N. R. Rao,Phil. Trans. R. Soc. A366, 63 (2008).
13Y. Tokura, “Correlated-Electron Physics in Transition-Metal Oxides,”
Phys. Today56(7), 50 (2003).
14H. Sawada, Y. Morikawa, K. Terakura, and N. Hamada,Phys. Rev. B56, 12154 (1997); R. J. Radwanski and Z. Ropka, J. Mag. Mag. Mater.
272–276, E259 (2004).
15S. Satpathy, Z. S. Popovic´, and F. R. Vukajlovic´,Phys. Rev. Lett.76, 960 (1996).
16D. Kasinathan and D. J. Singh,Phys. Rev. B74, 195106 (2006).
17J. J. Hamlin, S. Deemyad, J. S. Schilling, M. K. Jacobsen, R. S. Kumar, A. L. Cornelius, G. Cao, and J. J. Neumeier,Phys. Rev. B76, 014432 (2007).
18T. Taniguchi, S. Mizusaki, N. Okada, Y. Nagata, S. H. Lai, M. D. Lau, N.
Hiraoka, M. Itou, Y. Sakurai, T. C. Ozawa, Y. Noro, and H. Samada, Phys. Rev. B77, 014406 (2008).
19Y. Shirako, H. Satsukawa, X. X. Wang, J. J. Li, Y. F. Guo, M. Arai, K.
Yamaura, M. Yoshida, H. Kojitani, T. Katsumata, Y. Inaguma, K. Hiraki, T. Takahashi, and M. Akaogi,Phys. Rev. B83, 174411 (2011).
20G. T. Wang, M. P. Zhang, Z. X. Yang, and Z. Fang, J. Phys. Condens.
Matter21, 265602 (2009).
21A. Maignan, B. Raveau, V. Hardy, N. Barrier, and R. Retoux,Phys. Rev.
B74, 024410 (2006).
22V. Durairaj, S. Chikara, X. N. Lin, A. Douglass, G. Cao, P. Schlottmann, E. S. Choi, and R. P. Guertin,Phys. Rev. B73, 214414 (2006).
23V. Markovich, M. Auslender, I. Fita, R. Puzniak, C. Martin, A.
Wisniewski, A. Maignan, B. Raveau, and G. Gorodetsky,Phys. Rev. B73, 014416 (2006).
24H. T. Jeng, S. H. Lin, and C. S. Hsue,Phys. Rev. Lett.97, 067002 (2006).
25A. R. Denton and N. W. Ashcroft,Phys. Rev. A43, 3161 (1991).
26V. Skumryev, F. Ott, J. M.D. Coey, A. Anane, J. P. Renard, L. Pinsard- Gaudart, and A. Revcolevschi,Eur. Phys. J. B11, 401 (1999); Y. S. Hou, H. J. Xiang, and X. G. Gong, e-printarXiv:1310.3590(2013).
27F. Ricci, M. F. Bevilacqua, F. M. Granozio, and U. S. di Uccio,Phys. Rev.
B65, 155428 (2002).
28Y. J. Chang, C. H. Kim, S. H. Phark, Y. S. Kim, J. Yu, and T. W. Noh, Phys. Rev. Lett.103, 057201 (2009).
29B. Kim and B. I. Min, “Termination-dependent electronic and magnetic properties of ultrathin SrRuO3 (111) film on SrTiO3,” e-print arXiv:1311.7533(2013).
30A. Auerbach, Interacting Electrons and Quantum Magnetism(Springer- Verlag, Berlin, 1994).
