Electrical transport properties of manganite powders under pressure

0.4 MB·English

by  M. G. Rodríguez

x
Electrical transport properties of manganite powders under pressure
M. G. Rodr´ıguez, A. G. Leyva, and C. Acha 1∗
Laboratorio de Bajas Temperaturas, Departamento de F´ısica, FCEyN, UBA,
and IFIBA (CONICET), Ciudad Universitaria, (C1428EHA) Buenos Aires, Argentina
We have measured the electrical resistance of micrometric to nanometric powders of the
La5/8−yPryCa3/8MnO3 (LPCMO with y=0.3) manganite for hydrostatic pressures up to 4 kbar.
2 Byapplyingdifferentfinalthermaltreatmentstosamplessynthesizedbyamicrowaveassisted den-
1 itration process, we obtained two particular grain characteristic dimensions (40 nm and 1000 nm)
0 whichallowed ustoanalyzethegrain sizesensitivity oftheelectrical conductionpropertiesofboth
2 the metal electrode interface with manganite (Pt / LPCMO) as well as the intrinsic intergranular
interfaces formed by the LPCMO powder, conglomerate under the only effect of external pressure.
n
a We also analyzed the effects of pressure on the phase diagram of these powders. Our results indi-
J cate that different magnetic phases coexist at low temperatures and that the electrical transport
propertiesarerelatedtotheintrinsicinterfaces, asweobserveevidencesofagranularbehaviorand
7
an electronic transport dominated bythe SpaceCharge limited Current mechanism.
1

]
i I. INTRODUCTION morphologycorrespondstoanon-sinteredpowder. These
c
s samples were synthesized following a microwave assisted
l- Up to now there have been many studies trying to denitrationprocess3 and a final heattreatment atdiffer-
r shown which is the particular influence of grain size on ent temperatures to obtain grain sizes nearly two orders
t
m the properties of ceramic manganites.1,2 It is clear that, of magnitude different: Sample A, with a final sintering
◦
. by reducing the grain size of the sample, the surface to treatmentof4hoursat1400 C,withagrainsizediame-
at volume ratio increases,havingclear implications in their ter (Dg) of1-2micronsandSample B,treatedat800◦C
m physicalproperties. Consideringnearestneighbors,mag- for 10 minutes with Dg close to 40 nm. From roomtem-
- netic interactions are different in bulk than in the sur- perature down to 50 K this compound presents several
d face layer. Also impurities and oxygen content as well magnetictransitionsandacomplicatedphaseseparation
n as the crystal structure can also be different, modify- scenario. A magnetic characterization and the effects of
o ing particularly transport properties for their large sen- hydrostatic pressure on the different magnetic orderings
[c sitivity to the intergrain coupling. Finally, the phase can be found in a previous work.4
transitions may be modified, affecting particularly most In order to study the electrical transport properties
1
of the physical properties of these phase separation sys- of the powders at different pressures two different tech-
v
2 tems. Regarding transport properties, it is well known niques were used: for small pressures (P¡0.4 kbar) the
4 fact that manganites in the paramagnetic state (PM) powders were introduced inside a 2 mm diameter py-
6 haveasemiconducting-liketemperaturedependentresis- rophyllite ring with Pt wires, the whole arranged in a
3 tance. Ontheotherhand,ifwearedealingwithsamples Bridgman-likeconfiguration,whereasmalluniaxialpres-
1. in powder form, only held together by the application of surewasappliedjusttoensuretheelectricalconductivity
0 external pressure (P), we would expect that this gran- of the powder. For higher pressures, the inner volume
2 ularity will dominate the behavior of the electrical con- of a plastic (Teflon) cylinder was filled with the sam-
1 duction, by determining an extrinsic mean free path of ple’s powder (shown in Fig. 1) and pressurized inside a
v: carriers or the tunneling mechanism between grains. In piston-cylinder hydrostatic cell. The plastic cylinder has
i thispaperwepresentresistancemeasurementsasafunc- tworemovablelidsandfourequidistantholesthroughits
X
tion of temperature of La5/8−yPryCa3/8MnO3 (LPCMO walls,rangingfromsidetosidealongitsdiameter,where
ar with y=0.3)powders with two distinct grainsizes (1 µm Pt leads were introduced to perform the resistance mea-
and40nm)Thus,bycomparison,wewanttoexplorethe surements using a standard four terminal DC technique.
influenceofnanostructurationontheelectricaltransport To avoid overheating and non-linear effects the exciting
properties, to determine the main mechanism that regu- currentwasregulatedtokeepthevoltagedropacrossthe
lates the conductivity between grains and to determine sample below 10 mV. IV characteristics were also mea-
the sensitivity to external pressure of the magnetic or- suredusing the same configuration. The cylinder andall
dering of this compound. its holes were carefully sealed by using a low tempera-
ture resistent epoxy in order to avoidthat the pressuriz-
ing fluid (50%-50% kerosene and transformer oil) mixes
II. EXPERIMENTAL
with the sample preventing its electrical continuity. The
proper increase of pressure inside the cylinder with in-
In this paper we study the ceramic manganite creasingthepressureofthefluidwassuccessfullychecked
La5/8−yPryCa3/8MnO3 (LPCMO with y=0.3), whose bymeasuringtheelectricalresistanceofamanganinwire
2
placedinside the cylinderwithsteatitepowderasthein- temperature range. Nevertheless, by analyzing the loga-
ner transmitting medium. Only a small pressure range rithmic derivative of the resistance (no shown here), the
can be studied; up to 0.4 kbar for the Bridgman-like ferromagnetic onset of T seems to be located over 300
c1
setup,duetothemaximumalloweduniaxialstresses. For K. At T the resistance substantially increases, reach-
N
the hydrostatic cell, a minimum pressure (in some cases ingvaluesoverourexperimentalcapacity,preventingthe
up to ≃ 2 kbar) was required to have low contact re- measurement down to low temperatures.
sistances in order to allow the resistance measurements,
and, due to the pressure cell inner dimensions, the max-
imum attained pressure was 4 kbar. Sample A Sample B
5 P = 1.9 kbar P = 3.3 kbar
5x10 P = 1.7 kbar 106 P = 4 kbar
5 Tc1 TN
4x10
m] 5
h 5 10
R [o 3x10 Tc2
5
2x10 4
10
5
1x10
TN 3
0 10
0 50 100150200250300 100 150 200 250 300

FIG. 2: (Color online) TemperTa t[Ku]re dependence of the resis-
tanceofSampleAandBatdifferentpressures. Thetransition
temperaturesareindicatedforeachcase(Tc1,TN andTc2for
Sample A,and only T for Sample B).
N
FIG.1: (coloronline)Setuptomeasuretheresistanceofpow-
der samples under hydrostatic pressure. Left panel: Part of
theinnersampleholderofthehydrostaticcellwiththeTeflon
cylinder on top and its sketch. Right panel: Different views
of the Teflon cylinder filled with the LPCMO powders.
240
Sample A
200
III. RESULTS AND DISCUSSION K] Tc1 (R)
T [ Tc1 (Xac)
Tc2 (Xac)
160 TN (R)
The temperature dependence of resistance of both
samples(AandB)atdifferentpressurescanbeobserved
in Fig.2. Resistance measurements shown here were 120
0 2 4 6 8 10
taken on warming. Measurements on cooling showed a P [k bar]
different path due to hysteresis effects associated with FIG.3: (coloronline)Phasediagram ofSampleA.Thetran-
thefirst-order-typetransitioninasystemwithtwo-phase sitiontemperaturesobtainedpreviouslybyACsusceptibility4
coexistence.5 Both samples show a high resistance value are also included for comparison.
whencomparedtosinteredceramicsamplesofsimilarge-
ometric factors.5 Sample A shows a semiconducting-like In Fig.3 and Fig.4 the obtained phase diagram can be
dependence in the paramagnetic state and very well de- seen. We also included for comparison the curves ob-
finedtransitionstotheferromagneticorderingandtothe tained in a previous paper4 where we studied the pres-
chargeorderedantiferromagneticstate,correspondingto sure sensitivity of the magnetic ordering transitions by
the temperatures labeled T ,T and T , respectively. ACsusceptibility. Thetransitiontemperaturesmeasured
c2 c1 N
A metallic-like conduction can be observed in two inter- by different techniques not necessarilycoincides, but the
mediate temperature regions, arbitrarily given rise to a general trend is maintained. In this sense, the CO-AF
peak in the resistance, not necessarily associated with a transition is quickly reduced by increasing the pressure
magnetic order transition but rather to the temperature while the FM transition increases softly. Here, by mea-
dependence of a competing distribution of metallic and suring the pressure sensitivity of the resistance, we con-
insulatorregionsdevelopedasaconsequenceofthephase firm these previous results. Interestingly, from the mag-
separation scenario. netic studies, the FM transition developed at T com-
c2
Asacleardifferencewiththeresistancedependencefor pletely mask our capacity to detect the existence of the
sinteredsamples, the groundstate of Sample A seems to AFtransitioninthecasethatT ≤T . Incontrast,the
N c2
be insulator, instead of metallic.6 On the other hand, resistance measurements reveal that the CO-AF phase
Sample B shows a quite different resistance dependence, is still present up to 4 kbar, even for temperatures in
which is essentially semiconducting-like for the whole that range, at least for sample B. The pressure range of
3
ourmeasurementsshouldbeextendedtogeneralizethese canbeperformed,assumingthat∆φisinthe60-200meV
results for both samples. range (from the conductivity activation energy9,10) and
ǫ≃10-100. 11Theobtainedvalueisinthe1-10nmrange,
in a very good agreement with previous estimations.12

220 Sample B TN (Xac)
TN (R)
Tc2 (Xac)
200
Sample B
180 P=0.3 kbar
T [K]160 10-3 ~ V2
140 A]
I [
120
1.1
1000 2 4 6 8 10 10-4 ~ V
P [kbar]
FIG.4: (coloronline)Phasediagram ofSampleB.Thetran-
sitiontemperaturesobtainedpreviouslybyACsusceptibility4 0.1 1
V [V]
are also included for comparison. There were not evidences FIG.6: (Coloronline)IVcharacteristicsatroomtemperature
in theresistivity of the transitions at Tc1 and Tc2. for Sample B at 0.3 kbar. A crossover from a nearly ohmic
dependence to a I ∼ V2 law can be observed, typical of a
SCLC conduction mechanism at granular interfaces.
We alsoanalyze the IV characteristicsinorderto gain
105 Sample B insightontheconductionmechanismbetweeninterfaces.
We present in Fig.6 and Fig.7 results for Sample B, al-
though a similar behavior was obtained for Sample A.
m] In both figures a non-linear dependence can be observed
h
R [o 104 twhiatht eIv∼olvVes2.from a nearlyohmic relationto a powerlaw
Y= A + B * X
A = (-1,998 + 0,007)
B = (91,4 + 0,1) K1/2
R = 0,99980983

-2
103 10
0,06 0,07 0,08 Sample B
1/2 1/2 P=4 kbar
1/(T ) [1/(K )]
FIG.5: (Color online) Resistanceof SampleBat 4kbarasa
functionof 1 intheparamagnetictemperaturerange. The 10-3 ~ V2
T0.5
line is a fit using Eq.1
A]
I [ -4
From the temperature sensitivity of the resistance, it 10
1.5
was not possible to elucidate the dominant electrical ~ V
transportmechanismforSampleA,asboththefitsbased
-5
on a semiconductor-like conduction or on the granular 10 1 10
type gavereasonableresults. However,for Sample B, we
FIG. 7: (Color online) IV chVa r[Vac]teristics at room tempera-
obtained a better match considering the granular con-
tureforSampleBat4kbar. AsinFig.6theobtaineddepen-
duction model7,8 instead of the semiconducting one, as
denceisrepresentativeofaSCLCmechanism,whereelectrons
can be observed in Fig.5. The expression we use to fit
are injected by good conducting zones into semiconductors
our data corresponds to: in close electrical contact. Here, the crossover to a I ∼ V2
occurs at higher voltages indicating that pressure favors the
conduction of thesemiconducting zones.
C (χS2e2)(1+ 1 )
ln(R)= 0 =4{ 2χS }1/2, (1)
T1/2 ǫD(12D+S)kbT Thisbehavioristypicalofthespacechargelimitedcur-
rents(SCLC) mechanism,obtainedfor interfaces formed
where χ2 = 2m∆φ, m and e are the electron’s mass and by good metals in contact with semiconductors.13 It is
h¯2
charge, respectively, ∆φ the energy difference between interestingtonotethatthecrossovertothequadraticde-
the effective barrier height and the electrons energy, k pendenceisshiftedtohighervoltagevalueswithincreas-
b
the Boltzmann constant, ǫ the dielectric constant of the ing pressure, indicating that pressure extend the ohmic
insulator, D its characteristic grain size and S the grain behavioroftheinterface,probablybyincreasingthecon-
separation distance, associated with the insulator layer ductivity of the semiconductor.14 As the IV characteris-
between the conducting grains. A crude estimation of S tics were measured in a four terminal configuration,it is
4
clearthattheSCLCconductionisnotonlyrelatedtothe powders samples with grain sizes in the micrometric to
interface formed by the Pt electrodes with the powder nanometric scale. Both samples show a conduction be-
but it is essentially associated with the intrinsic inter- havior dominated by the interfaces, with non-linearities
faces formed between grains. As in the granular model basedonthe SCLC mechanism anda conductionregime
thatdescribesthetemperaturedependentresistance,the determined by their extrinsic granularity. Within the
SCLC bulk conductionis revealingthe existence of good granular model, we made a crude estimation of the in-
metalliczones(grains)verywellconnectedtoasemicon- sulator layer obtaining a value in the 1-10 nm range.
ducting layer (surface of grains). Considering the magnetic order transitions, we obtain
a phase diagramthat confirms previous results obtained
by magnetic measurements. Moreover, new results were
IV. CONCLUSIONS obtained as we observed the coexistence of the CO-AF
phasewiththelowtemperatureFMphaseconfirmingthe
We have measured, as a function of temperature and complex phase diagram of these phase separated com-
pressure, the electrical transport properties of LPCMO pounds.
∗ Corresponding author: [email protected] 8 P. Sheng,B. Abeles, and Y. Arie, Phys.Rev.Lett. 31, 44
1 R. D. S´anchez, J. Rivas, C. V´azquez-V´azquez, A. L´opez- (1973).
Quintela, M. T. Causa, M. Tovar, and S. Oseroff, Appl. 9 E.Dagotto,T.Hotta,andA.Moreo,PhysicsReports344,
Phys. Lett. 68, 134 (1996). 1 (2001).
2 A. Dutta, N. Gayathri, and R. Ranganathan, Phys. Rev. 10 G. Garbarino, C. Acha, D. Vega, G. Leyva, G. Polla,
B 68, 054432 (2003). C. Martin, A.Maignan, and B. Raveau,Phys.Rev.B70,
3 A.Leyva,P.Stoliar, M.Rosenbusch,V.Lorenzo, P.Levy, 014414 (2004).
C.Albonetti,M.Cavallini,F.Biscarini, H.Troiani,J.Cu- 11 J. L.Cohn,M. Peterca, andJ. J.Neumeier, Phys.Rev.B
riale, et al., Journal of Solid State Chemistry 177, 3949 70, 214433 (2004).
(2004). 12 J.Curiale,M.Granada,H.E.Troiani,R.D.Snchez,A.G.
4 C. Acha, G. Garbarino, and A. G. Leyva, Physica B 398, Leyva, P. Levy, and K. Samwer, Appl. Phys. Lett. 95,
212 (2007). 043106 (2009).
5 M. Uehara, S. Mori, C. H. Chen, and S.-W. Cheong, Na- 13 G. Dietz, W. Antpohler, M. Klee, and R. Waser, J. Appl.
ture (London) 399, 560 (1999). Phys. 78, 6113 (1995).
6 J. A. Collado, C. Frontera, J. L. G.-M. noz, C. Ritter, 14 P.C.JoshiandS.B.Krupanidhi,J.Appl.Phys.73, 7627
. M. Brunelli, and M. A. G. Aranda, Chem. Mater. 15, (1993).
167 (2003).
7 P. Shengand B. Abeles, Phys. Rev.Lett. 28, 34 (1972).