High-pressure effects on single crystals electron-doped r$_{2-x}$Ce$_{x}$CuO$_{4}$

0.33 MB·English

by  C. R. Rotundu

High pressure effects on single crystal electron-doped Pr Ce CuO
2−x x 4
C. R. Rotundu,1,2,∗ V. V. Struzhkin,3 M. S. Somayazulu,3 S. Sinogeikin,4 Russell J. Hemley,3 and R. L. Greene1
1Center for Nanophysics & Advanced Materials and Department of Physics,
University of Maryland, College Park, MD 20742, USA
2Materials Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA
3Geophysical Laboratory, Carnegie Institution of Washington, Washington, DC 20015, USA
4HPCAT, Geophysical Laboratory, Carnegie Institution of Washington,
Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois 60439, USA
3 (Dated: January 10, 2013)
1
We present high pressure diamond anvil cell synchrotron X-ray, resistivity, and ac-susceptibility
0
measurements on the electron-doped cuprate Pr2−xCexCuO4 to much higher pressures than previ-
2 ′
ouslyreported. At2.72GPabetween88and98% ofthesuperconductingT phaseoftheoptimally

n dopedPr1.85Ce0.15CuO4transformsintotheinsulatingphaseT.Withapplicationofpressure,theT
a phasebecomesmoreinsulating,sowepresentherewhatmaybethefirstexampleofelectron-doping
J intheTstructure. Theresultshaveimplicationsforthesearchforambipolarhigh-Tccupratesuper-
9 conductors. The Tc of theremaining 2-12% T′ phaseis suppressed continuously from 22 K to 18.5
K at about 14 GPa. Remarkably, the Tc of the overdoped Pr1.83Ce0.17CuO4 remains practically
] unchangedeven at 32 GPa.
n
o
PACSnumbers: 61.50.Ks,74.25.Dw,74.72.Ek,74.25.fc,74.62.Fj
c
-
r
p I. INTRODUCTION run on a small sample of approximately 40 × 40 × 10
u µm3 cleaved from a few mm size crystal. The measure-
s
. Although hole-doped cuprates are the most studied ments were performed using a standard four-probe Van
t
a class of high-T materials, attention has been drawn re- der Pauw configuration, and the schematic of the setup
c
m cently to the electron-doped cuprates1,2 in the effort to is shown in Fig. 1a.
- achieve a unified understanding of the high temperature
d Pressure was achieved using a lever-arm system with
superconducting mechanism in cuprates. High pressure
n two300µmculetdiamondsmountedontungstencarbide
experiments are important for understanding the super-
o supports. Ontheculetofoneofthediamondsfourradial
c conductivity and to help identify ways for increasingTc. platinum-basedpolymerconductiveleadsweredeposited
[ Experiments on hole-doped cuprates showed an increase using focused ion beam (FIB) lithography10. The inner
of T when pressure is applied, with the record belong-
2 c endoftheseleadspassedoverthe sample,therebyassur-
v ing to HgBa2Ca2Cu3O8+δ3 forwhichthe Tc is enhanced ing electrical contact and mechanical attachment of the
1 from133 K to 164K when compressedto 30 GPa. Pres-
sample to the diamond (Fig. 1b). A stainless steel gas-
7 sures up to 2.5 GPa showed no (or extremely small)
ket was indented firstto 40 µm thickness and a centered
8 changes in structural4 and other physical properties6,7 ∼ 100 µm hole in the indentation was drilled. Cubic
2 ofelectron-dopedcuprates. We presentherea highpres-
. boronnitride (BN) powder was indented in the hole and
1 sure study, to pressures higher than previouslyreported,
on the conical side of the gasket,creating a thin insulat-
1 of the structural and other physical properties of single
ing layer. Four electrodes 5 µm made of thin platinum
2 crystals of electron-doped Pr2−xCexCuO4. We explain foilwereindented ina radialpositionto assureelectrical
1
the close relation between the structural and supercon-
: contactwiththeFIBdepositions(Fig. 1d). TheBNwas
v ducting properties. To our best knowledge, there are no
drilled in the center to match the gaskets hole, and the
i high pressure (> 2.5 GPa) studies of the superconduct-
X spacecreatedformedthesamplechamber. Beforeclosing
ing properties of electron-doped cuprates, except for the
r the DAC, ruby spheres were placed next to the sample
a resistivity study on polycrystalline Ln1.85Ce0.15CuO4−y for in situ pressure determination based on a calibrated
to 10 GPa by J. Beille et al.8. fluorescence shift11. Finally, the DAC was closed in pre-
compressed Ne gas at about 0.2 GPa, which served as a
pressure-transmitting medium. During the course of the
II. EXPERIMENTAL METHODS experiment, it was essential that the diamond anvils do
notpressdirectlyonthesample. Thus,thethicknessesof
Single crystals of Pr2−xCexCuO4, x=0.15 (optimally- the indented gasket and sample were 40 µm and 10 µm,
doped) and 0.17 (over-doped) were synthesized via a respectively. Measurements were stopped on compres-
flux method refined by Peng et al.9. The x=0.15 crys- sionwhenthe gasketthinned downsuchthatthe sample
tal had T = 21 K under normal pressure conditions was in direct contact with the diamonds. We note that
c
as determined from magnetization measurement in 20 the material under study cannot withstand an uniaxial
Oe, in agreementwith literature values9. Diamond anvil pressurelargerthan0.5GPa12. Inthe presentresistivity
cell (DAC) high pressure resistivity measurements were measurement the pressure corresponding to gasket col-
2
lapse was larger than 43 GPa.
DAC magnetic ac-susceptibility measurements were
carriedoutusingadouble-frequencymodulationmethod
with details given elsewhere13–15. The DAC consisted
of two pairs of diamonds inside of a larger primary coil,
a secondary signal coil (encircling the pair of the dia-
monds with the sample), and secondary compensating
coil - identical to the secondary signal coil but with no
sample. The gasket, made of a NiCrAl non-magnetic al-
loy, was pre-indented by the two pairs of diamonds and
then drilled the center of the indentations. The crystal
was cut to approximately 50 × 50 × 20 µm3 and placed
inside of one of the drilled holes, and the other was left
blank intentionally.
HighpressureX-raydiffractionwasperformedonpow-
der from crushed crystals at HPCAT (Sector 16) at the
AdvancedPhotonSource(APS),ArgonneNationalLab-
oratory. Inthese experiments,a DAC with400µmculet
diamondswithasamplechamberhavinga50mmdiam-
eterwasused. As forthe resistivityandac-susceptibility
measurements,Ne wasused as the pressuremedium. Ne
provides an quasihydrostatic environment for pressures
up to about 15 GPa;bove this value, the pressure gradi-
entsremainverysmall: at50GPathestandarddeviation
of pressure is less than 1%16.
III. RESULTS AND DISCUSSION
An early high pressure X-ray (to 0.6 GPa) study by
Kamiyamaet al.4 showedaverysmallbutcleardecrease
of the lattice parameters with pressure of the undoped
Nd CuO and optimally-doped Nd Ce CuO .
2 4 1.835 0.165 4
Higher pressure experiments showed that in the parent
Nd CuO a T′5 to T structural transition takes place at
2 4
21.5 GPa17, but the transition is found to take place at
15.1 GPa in the parent Pr CuO 18. To our best knowl-
2 4
edge we are the first to show X-ray data for an electron- FIG. 1: a) Schematic of the diamond anvil cell resistivity
dopedcupratetoapressurehigherthan0.6GPa. Figure highpressuresetup. b)Thediamondculetwiththeplatinum-
2showthelatticeparametersaandcandvolumecellver- basedpolymercontactsonthesample. c)IndentedBoronNi-
tridewithplatinum-foilmadeleads. Thecenterholebecomes
sus pressure of the optimally-doped Pr Ce CuO .
1.85 0.15 4
the sample space. d) View of the sample with the electrical
Our first pressure data point is 0.8 GPa. Interestingly,
′ contacts after DAC assembly was closed under compressed
the T to T transition takes place at a much lower pres-
′ neon gas and ready for theexperiment. (details in thetext)
sure,2.72 GPa,when 88-98%of the T phase transforms
to the T phase (Fig. 3). This is of interest because for
the case of the undoped Pr CuO at 37.2 GPa there is
2 4
still 50% of the T′ phase surviving18. While we believe
the differences are mostly intrinsic, the different pres- 15] GPa for LaNdCuO418. Upon applying pressure, the
′
sure media used (N2 gas in Wilhelm et al.17,18 vs. the latticeconstantsoftheT phaseofPr1.85Ce0.15CuO4 are
more hydrostatic Ne gas in the present study) may have continuously suppressed through the phase transition as
asizableinfluence. Thestandarddeviationofpressureis seen in Fig. 2. One other observation is that 16 GPa
about3-4%inN2gasat25GPa,whileforthecaseofNe pressure produces a more drastic shrinkageof the lattice
′
is less than 1% even at 50 GPa16. One question remain- parametersoftheT phasethana23%Cesubstitutionof
ing to be addressed is up to what pressure the phase Pr19. In fact, the 23% Ce doping (maximal solubility9)
T′ coexists with T in the optimally doped cuprate. In produces lattices changes equivalent to about 2 GPa of
Pr CuO bothT′ andT arepresentina50%ratioupto pressure.
2 4
37.2 GPa. In Nd CuO the phases coexist for the [21.5, Figure 4 shows resistivity versus temperature of
2 4
29.5]GPapressureinterval17andshortenfurtherto[11.4, Pr Ce CuO at 4.5, 7.0, 13.7,34, and 43 GPa. The
1.85 0.15 4
3
1.0
T phase
0.8
n0.6
o
cti
a
r
f0.4
0.2
0.0 T’ phase
0 2 4 6 8 10 12 14 16
P (GPa)
FIG. 3: Fractions of phases T′ and T versus pressure for
′
Pr1.85Ce0.15CuO4. InsetshowsrepresentationsofbothT and
T structures20.

1.4 13.7 GPa
1.0
16 1.3
14 ()cm0.9 ()cm11..12 7 GPa
0.8 4.5 GPa 1.0
0.9 4.5 GPa
FIG. 2: Lattice parameters versus pressure of 12 0.7 0.8
Pr1.85Ce0.15CuO4. From top to bottom: Volume cell m) 0 50 100 150 200 250 300 10 20 30
versus pressure (P), c axis versus P, and a axis versus P. c 10 T (K) T (K)
Solid symbols are for the T phase, hollow for the T′ and the ( 8
lines are a guide to the eye. At 2.72 GPa between 88 and
98% of theT′ phasetransforms to the T phase. 6
43 GPa
4
34 GPa
2
attempttocompresstheDACtothenexthigherpressure
0 50 100 150 200 250 300
resulted in the collapse of the metallic gasket and there-
T (K)
fore end of the resistivity experiment. At relatively low
pressure the resistivity versus temperature curves show FIG.4: ResistivityρversustemperatureofPr1.85Ce0.15CuO4
what resembles a superconducting transition (but with at 4.5, 7, 13.7, 34, and 43 GPa.
non-zero resistivity below T ) and enhancement of resis-
c
tivityclosetoT (leftinsetofFig. 4). Thisenhancement
c
of the resistivity near T , is due in part to a slight incli-
c
nationofthesamplecleavedfacefromtheCuO planes21 resistivity above T at normal pressure (where the ma-
2 c
and in part to granular effects within the crystal22. It is terial is in T′ phase) is a fraction of mΩcm9. The same
unlikely that this resistivity enhancement near T is due mechanismmostlikelyisresponsibleforthehighpressure
c
to inhomogeneities in Ce doping (as proposed by Klim- non-zeroresistivitydataofBeilleet al.8 belowthesuper-
czuk et al.23) given that the enhancement in the x=0.15 conducting transition in Ln1.85Ce0.15CuO4−y (Ln = Nd,
crystal(asseeninresistivitydataat4.5GPa)ismeasured Sm, Eu), although no high pressure X-ray are available
on a 10 µm thickness crystal while the inhomogeneities for these compositions.
in Ce were found to appear more in crystals of thickness T is suppressed by pressure and at 34 GPa we can-
c
greater than 300 µm24,25. not detect any sign of superconducting transition in the
Thenon-zeroresistivitybelowT forpressuresgreater resistivity data, and the shape of the resistivity versus
c
than the 2.72 GPa of the T′ −→ T transition can be ex- temperature curves are consistent with aninsulating be-
plained based on the 88-98%insulating T phase. This is havior. At higher pressure (43 GPa), the resistivity ver-
consistent with the magnitude of resistivity at 4.5 GPa sus temperature curve show two broad peaks. These
that is of an order of a fraction of a Ωcm, while typical mysterious features are perhaps due to the complicated
4
of Crusellas et al.29 were obtained from data using 1:1
isoamyl and n-pentane alcohol, that is a completely dif-
ferent pressure medium than the neon gas used in the
10
presentstudy. Therefore,lowerpressuresresistivitymea-
surements using the same Ne gas pressure media will be
needed to settle if T is monotonically suppressed with
5 c
applyingpressureorifthatbeyond2.7GPa(correspond-
ing to the T′ −→ T transition) T is suppressed at a
c
0 higher rate. Regardless, the rate of suppression of Tc
forthe optimally-dopedsample decreasesbeyonda pres-
surethatis somewherebetween7 and13GPashowinga
-5 “saturation” to certain T .
c

-10
0 10 20 30 40 50
Superconducting
22
phase diagram
FIG. 5: Real component of the ac-susceptibility versus tem-
peratureoftheover-dopedPr1.83Ce0.17CuO4 atvariouspres-
sures during a) compression and b) decompression. Data are
20
vertically displaced for clarity. The arrow (shown for the 7.5
GPa data) points to theTc.
)
K
(c18
T
effect of the pressure on the magnetic ordering (spin
orientation)26,27, and a better understanding of this will
require a careful high pressure neutron scattering study.
Figure 5 shows DAC ac-susceptibility data (real com- 16
ponent) for the over-doped Pr Ce CuO at com-
1.83 0.17 4
pression (Fig. 5a) and decompression (Fig. 5b), with
maximum pressure of 32.1 GPa. The arrow points to
the T . The shape of susceptibility data and how T
c c
14
isdeterminedwhenadouble-frequencymodulationtech-
nique is used have been discussed in detail in a review
paper by Struzhkin et al.13 and are based on the Hao-
0 5 10 15 20 25 30
Clemm theory for reversible magnetization in type II
superconductors28. Basically, T in the ac-susceptibility P (GPa)
c
data for Pr Ce CuO is marked by the higher tem-
1.83 0.17 4
perature “end” of the peak as shown in Fig. 5. The FIG.6: Tc versuspressureforPr1.85Ce0.15CuO4 (Htheonset
extremely sensitive ac-susceptibility proved to be an ex- of diamagnetism from χ under normal pressure, ⋄ adapted
from resistivity on crystal by Crusellas et al.29, N ρ during
cellent probe for detecting and monitoring the evolution
of T versus pressure given the small fraction of the T′ compression)andPr1.83Ce0.17CuO4 (•χduringcompression,
c ◦ χ during decompression). The dotted lines are a guide to
superconducting phase beyond the structural transition.
theeye.
Remarkably,fortheover-dopedPr Ce CuO ,T re-
1.83 0.17 4 c
mains unaltered all the way up to 32.1 GPa.
Finally, the phase diagram T versus pressure Lastly we discuss the significance of the resistivity of
c
is drawn in Fig. 6, for both Pr Ce CuO the optimally-doped (Fig. 4) “moving” into a more in-
1.85 0.15 4
and Pr Ce CuO . T for the optimally-doped sulating regime with application of pressure in the con-
1.83 0.17 4 c
Pr Ce CuO is given by the temperature at the text of search for ambipolar30–32 high-T cuprate super-
1.85 0.15 4 c
peak of resistivity versus T, and T for the over-doped conductors. One such example of an ambipolar high-
c
Pr Ce CuO from the ac-susceptibility versus T as T cuprate superconductors has been reported recently
1.83 0.17 4 c
described earlier. We also included in the supercon- by K. Segawa and Y. Ando30. They reported successful
ducting phase diagram T versus pressure for [0-2] GPa doping of n-type carriers by La substitution for Ba in
c
as determined from resistivity measurements by Crusel- YBa Cu O , such that Y La Ba La Cu O is
2 3 y 0.38 0.62 1.74 0.26 3 y
las et al.29 on an optimally-doped PCCO crystal. It 2% electron-doped. It has been known for a long time
should be noted here that the high-pressure data points thattheT-structurecanbeonlyeasilyhole-doped,while
5
the T′-structure can be easily only electron-doped33. In practically unchanged even at 32.1 GPa.
the present study, since 88-98% of the normal pressure One very interesting and surprising result is that with
′
T phase (that is electron-doped) transforms into the T application of pressure, the T phase becomes more in-
phase, it is natural to assume that excess electrons were sulating, and so we present here the first example of
doped in the T phase. We believe the significance of the electron-doping in the T structure. One of the most im-
resistivityofthe T phase becoming moreinsulating with portant questions is if by applying even larger pressure
application of pressure is that we successfully doped for the T phase can be driven to electron-doped supercon-
the first time n-type carriers in the T structure. From ductivity. Most certainly the present study will spark
the X-ray data it can be seen that T structure is stable interest and further experiments on the affect of high
upto16GPa,so,onequestionisifthestructureisstable pressure on the electron-doped cuprates.
at much higher pressures.
IV. SUMMARY Acknowledgments
We studied the evolution of superconductivity and We thank P. Fournier and S. Uchida for useful dis-
structure (and the relationship between) with pressure cussions. The FIB deposition contacts were done at In-
of electron-doped Pr2−xCexCuO4. At 2.72 GPa be- stitute for Research in Electronics and Applied Physics,
′
tween 88 and 98% of the superconducting T phase of UniversityofMaryland. The workwas supportedby the
the optimally-doped Pr Ce CuO transforms into State of Maryland and the NSF through grant DMR-
1.85 0.15 4
′
the insulating T phase. T of the remaining 2-12% T 1104256 (C.R.R. and R.L.G.), and DOE through DE-
c
phase is suppressed from 22 K to 18.5 K at a pressure FG02-02ER45955 (V.V.S). X-ray diffraction was per-
of about 14 GPa. The non-zero resistivity below T can formedatHPCAT(Sector16),AdvancedPhotonSource
c
be explainedbasedonthe 88-98%insulatingT phase for (APS), Argonne National Laboratory. HPCAT opera-
pressures beyond 2.72 GPa. This is in accord with the tions were supported by CIW, CDAC, UNLV and LLNL
highmagnitude (orderof Ωcm) of resistivityat 4.5GPa, through funding from DOE-NNSA and DOE-BES, with
while the typical resistivity above T at normal pressure partial instrumentation funding by NSF. APS was sup-
c
′
(at which the material is in T phase) is a fraction of ported by DOE-BES, under Contract No. DE-AC02-
mΩcm9. T oftheover-dopedPr Ce CuO remains 06CH11357.
c 1.83 0.17 4
∗
On leave from Lawrence Berkeley National Laboratory; 4673 (1986).
corresponding author: [email protected] 12 Y.Kaga,T.Sasagawa, S.Takahashi,K.Unosawa,H.Tak-
agi, Physica B 359, 442 (2005).
1 N.P.Armitage,P.Fournier,andR.L.Greene,Rev.Mod. 13 V.V.Struzhkin,E.Gregoryanz,H.K.Mao,R.J.Hemley,
Phys. 82, 2421 (2010). & Y. A. Timofeev, New methods for investigating super-
2 K.Jin, N. P.Butch,K.Kirshenbaum,J. Paglione, andR. conductivityatveryhighpressures,inHighPressurePhe-
L. Greene, Nature476, 73 (2011). nomena, edited by R. J. Hemley, M. Bernasconi, L. Ulivi
3 L. Gao, Y. Y. Xue, F. Chen, Q. Xiong, R. L. Meng, D. andG. Chiarotti, pp.275-296 (IOSPress/Societ`a Italiana
Ramirez, C. W. Chu, J. H. Eggert and H. K. Mao, Phys. di Fisica, Amsterdam, 2002).
Rev.B 50, 4260 (1994). 14 E. A. Gregoryanz, V. V. Struzhkin, R. J. Hemley, M. I.
4 T.Kamiyama,F.Izumi,H.Takahashi,J.D.Jorgensen,B. Eremets, H.K.Mao, &Y.A.Timofeev, Phys.Rev.B65,
Dabrowski, R. L. Hitterman, D. G. Hinks, H. Shaked, T. 064504 (2002).
O. Mason, M. Seabaugh, Physica C 229, 377 (1994). 15 X.-J.Chen,V.V.Struzhkin,Y.Yu,A.F.Goncharov,C.
5 T′ - structure is I4/mmm, Nd2CuO4-type and T - struc- -T.Lin,Ho-kwangMao, &R.J.Hemley,Nature466,950
ture is I4/mmm,K2NiF4 - type(see theinset of Fig. 3). (2010).
6 C. Murayama, N. Mˆori, S. Yomo, H. Takagi, S. Uchida, 16 S.Klotz,J.-C.Chervin,P.Munsch,andG.LeMarchand,
and Y.Tokura, Nature339, 293 (1989). J. Phys. D: Appl.Phys.42, 075413 (2009).
7 J. T. Markert, J. Beille, J. J. Neumeier, E. A. Early, C. 17 H.Wilhelm,E.Cros,E.Reny,G.Demazeau,andM.Han-
L. Seaman, T. Moran, and M. B. Maple, Phys.Rev.Lett. fland,J. Matter. Chem. 8, 2729 (1998).
64, 80 (1990). 18 H.Wilhelm,E.Cros, E.Reny,G.Demazeau,andH.Han-
8 J.Beille, A.Gerber,Th.Grenet,M.Cyrot,J.T.Markert, fland,J. Solid StateChem. 151, 231 (2000).
E. A. Early, and M. B. Maple, Solid State Commun. 77, 19 E. Maiser, P. Fournier, J. -L. Peng, F.M. Araujo-Moreira,
141 (1991). T. Venkatesan, R. L. Greene, G. Czjzek, Physica C 297,
9 J.L.Peng,Z.Y.Li, andR.L.Greene,PhysicaC177, 79 15 (1998).
(1991). 20 W. Kraus & G. Nolze, J. Appl.Cryst. 29, 301 (1996).
10 J. Melngailis, Nucl. Instrum.Meth. B 80, 1271 (1993). 21 M. Suzuki,Phys. Rev.B 50, 6360 (1994).
11 H. K. Mao, J. Xu and P. M. Bell, J. Geophys. Res. 91, 22 M.A.Crusellas, J.Fontcuberta,andS.Pin˜ol ,Phys.Rev.
6
B 46, 14089 (1992). (1991).
23 T. Klimczuk, T. Plackowski, W. Sadowski, M. Pleb´aczyk, 29 M.A.Crusellas,J.Fontcuberta,S.Pin˜ol,J.Beille,andT.
Physica C 387, 203 (2003). Grenet, Phys. Rev.B 48, 615 (1993).
24 M. Matsuda, Y. Endoh, Y. Hidaka, Physica C 179, 347 30 K. Segawa and Y. Ando, Phys. Rev. B 74, 100508(R)
(1991). (2006).
25 E.F.Skelton,A.R.Drews, M.S.Osofsky,S.B.Qadri, J. 31 M. Ikeda, M. Takizama, T. Yoshida, A. Fujimori, K.
Z. Hu, T. A. Vanderah, J. L. Peng, R. L. Greene, Science Segawa,andY.Ando,Phys.Rev.B82,020503(R)(2010).
263, 1416 (1997). 32 K. Segawa, M. Kofu, S. -H.Lee, I. Tsukada, M. Fujita, S.
26 S. Katano, N. Mˆori, H. Takahashi, T. Kobayashi, and J. Chang, K. Yamada, and Y. Ando, Nature Physics 6, 579
Akimitsu, J. Phys. Soc. Jpn. 59, 1928 (1990). (2010).
27 S. Katano, R. M. Nicklow, S. Funahashi, N. Mˆori, T. 33 G. Burns, High-temperature superconductivity (Aca-
Kobayashi, and J. Akimitsu, Physica C 215, 92 (1993). demic, San Diego, 1992).
28 Z. Hao, J. R. Clemm, M. W. McElfresh, L. Civale, A.
P. Malozemoff, and F. Holtzberg, Phys. Rev. B 43, 2844