Absence of static magnetic order in lightly-doped Ti1-xScxOCl down to 1.7 K

0.61 MB·English

by  A. A. Aczel

Absence of static magnetic order in lightly-doped Ti Sc OCl down to 1.7 K
1−x x
A.A. Aczel,1,2,∗ G.J. MacDougall,1,2 F.L. Ning,1,3 J.A. Rodriguez,1,4
S.R. Saha,1,5 F.C. Chou,6 T. Imai,1,7 and G.M. Luke1,7
1Department of Physics and Astronomy, McMaster University, Hamilton, ON, Canada, L8S 4M1
2Neutron Scattering Science Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA
3Department of Physics, Columbia University, New York, NY 10027, USA
4Laboratory for Muon Spin Spectroscopy, Paul-Scherrer Institute, CH-5232 Villigen PSI, Switzerland
5Center for Nanophysics and Advanced Materials, Department of Physics,
University of Maryland, College Park, MD 20742, USA
6Center for Condensed Matter Sciences, National Taiwan University, Taipei 106, Taiwan
1
7Canadian Institute of Advanced Research, Toronto, Ontario, Canada, M5G 1Z8
1
0 (Dated: January 14, 2011)
2
Impurity-induced magnetic order has been observed in many quasi-1D systems including doped

n variants of the spin-Peierls system CuGeO3. TiOCl is another quasi-1D quantum magnet with a
a spin-Peierlsgroundstate,andthemagneticTisitesofthissystemcanbedopedwithnon-magnetic
J Sc. Toinvestigate therole ofnon-magnetic impuritiesin this system,we haveperformed both zero

3 field and longitudinal field µSR experiments on polycrystalline Ti1−xScxOCl samples with x = 0,
1 0.01, and0.03. WeverifiedthatTiOClhasanon-magneticgroundstate,andwefoundnoevidence
for spin freezing or magnetic ordering in the lightly-doped Sc samples down to 1.7 K. Our results

] instead suggest that thesesystems remain non-magneticup to the x = 0.03 Scdoping level.
l
e
- PACSnumbers: 76.75.+i,75.47.Lx
r
t
s
t. INTRODUCTION found an incommensurate lattice distortion along the a
a
andb-axeswhichis quite long-ranged(> 2000˚Ain each
m
case)[13]. Finally, 35Cl NMR measurements detect two
- Low dimensional magnets are of current interest, due
d peaks in the frequency spectrum corresponding to the
totheirpossiblerelevancetohightemperaturesupercon-
n I = −1 to 1 central transition. One of these peaks is
ductivityandtheirpenchantforpossessingexoticground z 2 2
o well-definedbuttheotherismuchbroader[14]. Theseex-
c states[1–4]. One subgroup of these materials is the spin-
perimentssuggestthatwhilethereisdimerizationinthis
[ Peierls (SP) systems, which possess non-magnetic spin-
incommensurate phase, the Ti-Ti intradimer distance is
singlet ground states. These materials consist of quasi-
1 not constant and adjacent Ti chains have a small rela-
v 1D Heisenberg antiferromagnetic chains that dimerize
tive shift to one another along the chain direction. This
4 at low temperatures due to strong magnetoelastic cou-
small relative shift vanishes below T , where the “lock-
5 pling. There are currently three known inorganic SP c1
6 in” transition occurs. At this point, all Ti chains are
systems: CuGeO [5], TiOCl[6], and TiOBr[7]. Organic
2 3 aligned with one another and the Ti-Ti intradimer dis-
SP systems such as TTF-CuS C (CF ) [8] and MEM-
1. 4 4 3 4 tance ina givenchainis constant. Complicating the pic-
(TCNQ) [9] were actually discovered first, but these are
0 2 ture further in the incommensurate phase regime are re-
much more difficult to study due to very low magnetic
1 cent x-ray measurements that have detected commensu-
1 moment densities. rate fluctuations coexisting with incommensurate Bragg
v: Thequasi-1DmagnetTiOClconsistsofmagneticTi3+ peaks[12].
i chains (spin S = 1 in 3d1 state), and undergoes succes-
X 2 It has long been appreciated that substituting a small
sive phase transitions at T = 94 K to an incommen-
c2 amountofnon-magneticimpuritiesforthemagneticsites
r
surate SP phase and T = 66 K to the commensurate
a c1 can lead to long-range magnetic order in conventional
SP ground state[10]. The presence of two phase tran-
SP systems[15], and so doping studies of these materials
sitions is in contrast to conventional SP systems, where
have been of great interest. In particular, it is possi-
onlyonetransitionisobserved. NMR[10]andx-rayscat-
ble to dope the magnetic Cu sites in CuGeO with non-
3
tering measurements[11–13] find a uniform dimerization magnetic Zn2+[5, 16, 18–21], Mg2+[20], or Cd2+[22, 23],
below Tc1 along the b-axis, providing evidence for com- as well as magnetic Ni2+[16, 20], Co2+[24], or Mn2+[16].
mensurate SP behaviour. Systems with Si4+[16, 18] doped in for Ge4+ have also
The nature of the incommensurate SP phase between been created. These materials have generally revealed
T and T is currently not well understood. Mag- phase diagrams with some common features, including
c1 c2
netic susceptibility and NMR measurements provide ev- the loss of SP order at a critical doping concentration
idence that the upper transition is associated with the x , a “dimerized antiferromagneticgroundstate” for the
c
onset of dimerization and the opening of a spin gap be- lightly-dopedcompoundswithx<x ,andauniforman-
c
low T [6, 10]. Recent x-ray measurements have also tiferromagneticgroundstateforsystemswithx>x [25].
c2 c
2
Inaddition,a detailedstudy onthe Cu1−xZnxGeO3 sys-
0.25
tems indicated the presence of impurity-induced long-
range magnetic order down to the lowest doping con-
centration studied (x = 0.001)[21]. This suggests the 0.20
absence of a required critical doping concentration to
achieve magnetic order and is in agreement with theo- y
retical work[26]. etr0.15
m
Oneexceptiontothesepropertieswasfoundinthecase m
y
ofCu1−xCdxGeO3,wherelong-rangemagneticorderwas As0.10 17050 K K
not observed for x ≤ 0.002[22] and the universality class
50 K
wasfoundtochangeupondopingfromthree-dimensional 1.7 K
0.05
XY to mean-field[23]. These features were suggested to
beconsequencesoflocalstrainfieldsinducedbythepres-
(a) TiOCl, ZF (b) Ti0.97Sc0.03OCl, ZF
ence of larger dopant ions, as the ionic radius of Cd2+ 0.00
(0.97 ˚A) is much larger than that of Cu2+ (0.72 ˚A). For 0 2 4 6 8 0 2 4 6 8
Time ( s) Time ( s)
all other doped systems investigated, the ionic radius of
the dopant ionis either smaller or comparableto that of FIG.1: ZF-µSRspectraof(a)TiOCland(b)Ti0.97Sc0.03OCl
measuredatselectedtemperatures. Thesolidlinesarefitsto
the ion being replaced.
thefunctional form described in the text.
Although many detailed studies on doped CuGeO
3
have been performed, very little is currently known
regarding how dopants affect the unconventional SP
ground state of TiOCl. An early report on this topic
discussed susceptibility results of Ti1−xScxOCl[6]. Note vious µSR studies have confirmed the presence of an-
thatsubstitutingnon-magneticSc3+ formagneticTi3+ is tiferromagnetic order in the series of doped spin-Peierls
essentially analogous to substituting a non-magnetic ion compoundsCu1−xZnxGe1−ySiyO3 withyaslowas0.007
for Cu2+ in CuGeO . In both cases, the non-magnetic and x as low as 0.021[18].
3
ions should lead to the destruction of some dimers and
We have performed both zero-field (ZF)-µSR and
create quasi-free spins. Accordingly, the authors found
longitudinal-field(LF)-µSRinthisworkonlightly-doped
thatthesusceptibilityofthedopedsampleswasgoverned
byverylargeCurietailsatlowtemperatures,whichthey samples of Ti1−xScxOCl (x = 0, 0.01,and 0.03). In con-
trasttodopedCuGeO ,wefindnoevidenceformagnetic
attributedtothedopantscreatingfinitechainsofHeisen- 3
order down to 1.7 K in any of the samples investigated.
berg spin-1 moments. Although no SP transition or
2 Our results instead indicate that the ground state re-
impurity-induced order was reported for the doped sam-
mains non-magnetic at these low doping levels.
ples in this study, additional measurements are needed
to definitively address these questions.
Lightly-doped Ti1−xScxOCl systems (x = 0.01 and
0.03) were studied by x-ray scattering very recently[27]
in an attempt to carefully address whether these sys-
tems were subject to a SP transition. These measure- EXPERIMENTAL DETAILS
ments confirmed that Sc-doping prevents the formation
of a long-range SP state down to 7 K even at the x =
0.01doping levelandinsteaddetectanincommensurate, Polycrystallinesamples ofTiOCl, Ti Sc OCland
0.99 0.01
short-range SP state for all temperatures below Tc2. Ti0.97Sc0.03OCl were prepared using the chemical vapor
The second issue of impurity-induced magnetic order transportmethodandtheScdopingconcentrationswere
can be readily addressed by the local probe technique inferred from susceptibility measurements as described
muon spin relaxation (µSR). Due to the large gyromag- in Ref. [6]. We performed zero field (ZF) µSR mea-
netic ratio of the muon, µSR is an extremely sensi- surements on these samples to verify the existence of a
tive probe of magnetism and can readily detect internal non-magnetic ground state in TiOCl and to search for
magnetic fields as small as ∼ 0.1 G. At TRIUMF, the any evidence ofmagnetic orderingin the doped systems.
muons areimplantedinto the sampleone ata time. The We also performedlongitudinal field (LF) µSR measure-
muon spin precesses around the local magnetic field and mentssothattheobservedrelaxationcouldbeattributed
then the muon decays into a positron, which is prefer- to a static or dynamic mechanism. These measurements
entially ejected along the direction of the muon spin at wereconductedontheM20surfacemuonchannelatTRI-
the time of decay (two neutrinos are also produced in UMF,usingaheliumgasflowcryostatinthetemperature
the muon decay process but not detected). The µSR range 1.7 K < T < 150 K with the samples mounted in
method is described in more detail in Ref. [17], and pre- a low-backgroundspectrometer.
3
DISCUSSION AND ANALYSIS
0.24
TiOCl
In systems with spin-singlet ground states, one ex- Ti0.99Sc0.01OCl
0.20 1.0
pects toobserveaZF-µSRsignalcharacteristicofanon- Ti0.97Sc0.03OCl
magnetic state: namely, the relaxation in the singlet -1 s)
0.16
regimeshouldbe small. ThisistruefortheinorganicSP e (
ssyusrteemmenCtsuGsheoOw3e[d28a].smInallT,igOraCdl,uaplreinvcioreuasseZFin-µtShRe rmelaexa-- on Rat0.12 Power 0.8
ation rate below Tc2, and then a much sharper increase axati0.08
intherelaxationratebelowTc1[29];ourresultspresented Rel 0.6
inFig.1(a)areconsistentwiththoseobservations. How-
0.04
ever, unlike previous work[29] we find no evidence that
the relaxation rate saturates at low temperatures. (a) (b)
0.00 0.4
One possible relaxation mechanism in TiOCl may be 0 50 100 150 0 50 100 150
the slowing down of a small concentration of quasi-free Temperature (K) Temperature (K)
spins that are created from defects/impurities. A sec-
ond contribution may be the result of a muon-induced FIG. 2: ZF relaxation rates and β values of Ti1−xScxOCl (x
= 0, 0.01, and 0.03). Note that β was fixed to 1 above Tc2.
effect. If the muon site lies near a Ti-Ti dimer, this may
have the effect of perturbing the local environment and
creating quasi-free spins in close proximity to the muon.
Thesequasi-freespinscanthenslowdownand/orfreeze,
of the ground state hasn’t changed. The large increase
enhancing the relaxation rate at low temperatures. This
in β for the doped samples may then indicate a large
effect has been observed in other singlet systems such
increase in the number of rapidly fluctuating, quasi-free
as SrCu (BO ) [30] and KCuCl [31]. To take these pos-
2 3 2 3 spins as compared to the pure case. This result is con-
sible relaxation mechanisms into account in the present
sistent with recent x-raywork that determined the long-
work,theZF-µSRdataforTiOClwasfittothefollowing
range, commensurate SP state is replaced by a short-
function:
range, incommensurate SP state for doping levels as low
P(t)=A e−(λt)β (1) as x = 0.01[27].
0
Notethatthepowerβwasfixedto1aboveT toprevent ZF-µSRrelaxationcaningeneralbetheresultofstatic
c2
this value from trading off with the relaxation rate as ordynamicprocesses. Todistinguishthesetwocasesone
needs to employ LF-µSR measurements. If the ZF re-
often happens when the latter value is small.
Some selected ZF-µSR spectra for Ti Sc OCl laxation were the result of quasi-static magnetic fields,
0.97 0.03
are depicted in Fig. 1(b). The ZF-µSR spectra for the relaxation would be decoupled in the presence of a
moderate applied longitudinal magnetic field, whereas
Ti Sc OCl, while not shown explicitly, are qualita-
0.99 0.01
dynamic (T ) relaxation would persist to much larger
tively similar to those of the x = 0.03 sample. The ab- 1
applied fields. Fig. 3 shows LF-µSR data for both Ti-
senceofcoherentmuonprecessionandthelackofmissing
OCl and Ti Sc OCl collected at 1.7 K. Assuming a
asymmetryatearlytimesindicatesthereisnolong-range 0.97 0.03
magnetic order in either of the doped materials. In light static field distribution to accountfor the ZF-relaxation,
of this, the ZF-µSR data for these samples was also fit we obtain an estimate for the magnitude of the average
internal field with the relation: B ∼ λ/γ where γ is
to Eq. (1). loc µ µ
The ZF relaxation rates and the β values for all three the muongyromagneticratio. In the casesof TiOCland
Ti Sc OCl, following this procedure leads to static
systems are depicted in Fig. 2. Note that TiOCl is 0.97 0.03
field estimates of ∼ 2.5 and 1 G respectively. An ap-
best described by a root exponential relaxation func-
plied LF of up to one order of magnitude greater should
tion at the base temperature of 1.7 K, as expected
then be enough to completely decouple the ZF spectra.
for a magnetically-dilute system with rapidly fluctuat-
ing spins[32]. This behaviour has also been observed However,bothTiOClandTi0.97Sc0.03OClexhibitsignif-
icant relaxation even with an applied LF of 500 G, and
in other singlet systems including Y BaNiO [33] and
2 5
Sr Cu O [34]. However, the power β increases with the therefore the observed ZF relaxation must be dynamic
2 4 6
dopinglevelatthelowesttemperaturesinvestigated. One in origin. This rules out spin freezing in these systems,
especially when coupled with the lack of a characteristic
possible explanation for this behaviour is that the effec-
peak in the ZF relaxation vs. temperature plots.
tivespindensityisincreasingintheSc-dopedcasesmuch
more than one would expect on the basis of introducing Furthermore, although the increase in β can also be
a small amount of extra impurities into the system. At explained by the spin fluctuations of the systems slow-
theselowdopinglevels,thedeviationfromrootexponen- ing down with increasing x, the LF-µSR measurements
tial behaviour should be minimal assuming the physics rule out this possibility. The increased difficulty in com-
4
are drastically different. A commensurate, long range
0.25
SP state gives way to an incommensurate, short range
SP state at x ≤ 0.01, although it is currently un-
0.20 known whether there is a critical concentration for
this phenomenon. This feature is accompanied by the
y absence of magnetic order for x ≤ 0.03, in contrast to
metr0.15 doped CuGeO3 where impurity-induced magnetic order
m generally seems to persist down to very low doping
y
s
A0.10 concentrations. This “magnetic order by disorder
2 kG
perturbation” effect has actually been proposed as
500 G
100 G a universal feature of quasi-1D spin gap systems, as
0.05 ZF impurity-induced order was also observed in the two-leg
laddersystemSrCu2−xZnxO3[35],thespindimersystem
0.00 (a) TiOCl, T = 1.7 K (b) Ti0.97Sc0.03OCl, T = 1.7 K Pb2−x(Bi, Sr)xV3O9[36], and the Haldane gap system
0 2 Time4 ( s) 6 8 0 2 Time4 ( s) 6 8 PbNi2−xMgxV2O8[37]. Significant interchain coupling
is an essential requirement for this magnetic order; the
FIG.3: LF-µSRspectraof(a)TiOCland(b)Ti0.97Sc0.03OCl uncompensated spins resulting from doping need to be
for selected LF at 1.7 K. The solid lines are fits to the func-
coupled in 3D space. However, the interchain interac-
tional form described in thetext.
tion in TiOCl leads to frustration[38] and it has been
suggested that it is responsible for the unconventional
SP behaviour found in this system. The frustrating
interchain interaction may prevent the formation of
antiferromagnetic long-range order upon doping TiOCl
0.20 with non-magnetic Sc3+.
TiOCl
-1 s) Ti0.99Sc0.01OCl One other possible reason for the absence of mag-
e (0.15 Ti0.97Sc0.03OCl netic order in lightly-doped Ti1−xScxOCl stems from
at the effects of dopant size. The ionic radius of Sc3+
R
on 0.10 (0.745 ˚A) is significantly larger than that of Ti3+
xati (0.67˚A), andthis sizedifference mayleadto localstrain
a
el and lattice distortions that prevent the formation of a
R0.05
magnetically-orderedstate. Thiswaspreviouslyfoundin
Cu1−xCdxGeO3[23], where the dopant ion is also much
0.00
0 500 1000 1500 2000 larger than the host. For this reason, further studies on
Applied LF (G) TiOCl using smaller dopant ions are necessary to help
determine whether impurity-induced order is a universal
FIG. 4: Relaxation rate as a function of applied LF for
feature of SP compounds.
Ti1−xScxOCl (x = 0, 0.01, and 0.03) at T = 1.7 K.
CONCLUSION
pletelydecouplingtheZFrelaxation(i.e. applyingalarge
enoughLFsotherelaxationoftheasymmetryeffectively ZF and LF-µSR measurements have verified that the
becomes zero) for the doped samples is quite evident. ground state of TiOCl is non-magnetic, and reveal
This was quantitatively characterized by fitting the LF the absence of magnetic ordering and spin freezing in
data to Eq. (1); β was fixed to the ZF value for each Ti1−xScxOCl (x = 0.01 and 0.03) down to 1.7 K. The
sample. Fig. 4 displays the resulting relaxation rates as latter result is in sharp contrastto the impurity-induced
a function of applied LF. There is some residual relax- antiferromagnetic order observed in the other inorganic
ation remaining even for the highest applied LFs in the spin-Peierls system CuGeO and many other quasi-1D
3
dopedsamples,possiblyindicating thatthe spinfluctua- spingapsystems. Thedifferencemaybe duetothe frus-
tionsareactuallygettingfasterwithincreasingxinstead trating interchain interaction of TiOCl or the use of a
andsuggestingthatthedopedsamplesremaininthefast dopant ion with a significantly larger ionic radius than
fluctuation regime. The decrease in the relaxation rate the host.
of the doped samples as compared to the pure case may We acknowledge useful discussions with J.P. Clancy
then be a consequence of a motional narrowing effect. and B.D. Gaulin, and we appreciate the hospitality of
Combined µSR, susceptibility[6], and x-ray the TRIUMF Center for Molecular and Materials Sci-
diffraction[27] results have now determined that ence where the µSR experiments were performed. Re-
the Ti1−xScxOCl and doped CuGeO3 phase diagrams search at McMaster University is supported by NSERC
5
and CIFAR. Work at Oak Ridge National Laboratory’s [19] M.C.Martin,M.Hase,K.Hirota,G.Shirane,Y.Sasago,
Spallation Neutron Source was sponsored by the Scien- N. Koide, and K. Uchinokura, Phys. Rev. B 56, 3173
tific User Facilities Division, Office of Basic Energy Sci- (1997).
[20] B. Grenier, J.-P. Renard, P. Veillet, C. Paulsen, G.
ences, US Department of Energy.
Dhalenne,andA.Revcolevschi,PhysicaB259-261, 954
(1999).
[21] K. Manabe, H. Ishimoto, N. Koide, Y. Sasago, and K.
Uchinokura, Phys.Rev.B 58, R575 (1998).
[22] S.Haravifard,K.C.Rule,H.A.Dabkowska,B.D.Gaulin,
∗
author to whom correspondences should be addressed:
Z.Yamani,andW.J.L.Buyers,JournalofPhysics: Con-
E-mail:[[email protected]] densed Matter 19, 436222 (2007).
[1] B.Lake,D.A.Tennant,C.D.Frost,andS.E.Nagler,Na-
[23] M.D. Lumsden, B.D. Gaulin, and H. Dabkowska, Phys.
tureMaterials 4, 329 (2005). Rev. B 58, 12252 (1998).
[2] H.Kageyama,K.Yoshimura,R.Stern,N.V.Mushnikov,
[24] P.E.Anderson,J.Z.Liu,andR.N.Shelton,Phys.Rev.B
K.Onizuka,M.Kato,K.Kosuge,C.P.Slichter,T.Goto, 56, 11014 (1997).
and Y. Ueda,Phys. Rev.Lett. 82, 3168 (1999). [25] K. Uchinokura,J. Phys.Cond. Matt. 14, R195 (2002).
[3] E. Dagotto and T.M. Rice, Science 271, 618 (1996).
[26] H. Fukuyama,T. Tanimoto, and M. Saito, J. Phys.Soc.
[4] C.H. Hsu, J.-Y. Lin, W.L. Lee, M.-W. Chu, T. Imai, Japan 65, 1182 (1996).
Y.J.Kao,C.D.Hu,H.L.Liu,andF.C.Chou,Phys.Rev.
[27] J.P.Clancy,B.D.Gaulin,J.P.Castellan, K.C.Rule,and
B 82, 094450 (2010). F.C. Chou, Phys. Rev.B 78, 014433 (2008).
[5] M. Hase, I. Terasaki, and K. Uchinokura, Phys. Rev.
[28] J.L. Garcia-Munoz, M. Suaaidi, and B. Martinez, Phys.
Lett.70, 3651 (1993). Rev. B 52, 4288 (1995).
[6] A. Seidel, C.A. Marianetti, F.C. Chou, G. Ceder, and
[29] P.J. Baker, S.J. Blundell, F.L. Pratt, T. Lancaster,
P.A.Lee, Phys.Rev.B 67, 020405(R) (2003).
M.L. Brooks, W.Hayes, M. Isobe, Y.Ueda, M. Hoinkis,
[7] T. Sasaki, M. Mizumaki, K. Kato, Y.Watabe, Y. Nishi-
M.Sing,M.Klemm,S.Horn,andR.Klaasen,Phys.Rev.
hata, M. Takata, and J. Akimitsu, J. Phys. Soc. Japan B. 75, 094404 (2007).
74, 2185 (2005).
[30] A.A. Aczel, G.J. MacDougall, J.A. Rodriguez, G.M.
[8] J.W. Bray, H.R. Hart Jr., L.V. Interrante, I.S. Jacobs,
Luke, P.L. Russo, A.T. Savici, Y.J. Uemura, H.A.
J.S. Kasper, G.D. Watkins, S.H. Wee, and J.C. Bonner,
Dabkowska, C.R. Wiebe, J.A. Janik, and H. Kageyama,
Phys.Rev.Lett. 35, 744 (1975). Phys. Rev.B 76, 214427 (2007).
[9] S. Huizinga, J. Kommandeur, G.A. Sawatzky, B.T.
[31] D. Andreica, N. Cavadini, H.U. Gudel, F.N. Gygax,
Thole,K.Kopinga,W.J.M.deJonge,andJ.Roos,Phys. K.Kramer,M.PinkpankandA.Schenck,PhysicaB289,
Rev.B 19, 4723 (1979).
176 (2000).
[10] T. Imai and F.C. Chou, cond-mat/0301425 (unpub-
[32] Y.J. Uemura in µSR Relaxation Functions in Muon Sci-
lished).
ence,EditedbyS.L.Lee,S.H.Kilcoyne,andR.Cywinski,
[11] M. Shaz, S. van Smaalen, L. Palatinus, M. Hoinkis, M.
IOP Publishing, Edinburgh and Briston UK (1999).
Klemm, S. Horn, and R. Claessen, Phys. Rev. B 71,
[33] K.M. Kojima, A. Keren, G.M. Luke, B. Nachumi, W.D.
100405(R) (2005).
Wu, Y.J. Uemura, M. Azuma, and M. Takano, Phys.
[12] J.P.Clancy,B.D.Gaulin,K.C.Rule,J.P.Castellan, and Rev. Lett.74, 2812 (1995).
F.C. Chou, Phys.Rev.B 75, 100401(R) (2007).
[34] K.M. Kojima, A. Keren, L.P. Le, G.M. Luke, B.
[13] E.T. Abel, K. Matan, F.C. Chou, E.D. Isaacs, D.E.
Nachumi, W.D. Wu, Y.J. Uemura, K. Kiyono, S.
Moncton, H. Sinn, A. Alatas, and Y.S. Lee, Phys. Rev.
Miyasaka, H. Takagi, and S. Uchida, Phys. Rev. Lett.
B 76, 214304 (2007). 74, 3471 (1995).
[14] S.R.Saha, S.Golin, T. Imai, and F.C. Chou, Journal of
[35] M. Azuma,Y.Fujishiro, M. Takano, M. Nohara, and H.
Physics and Chemistry of Solids 68, 2044 (2007). Takagi, Phys. Rev.B 55, R8658 (1997).
[15] M. Hase, N. Koide, K. Manabe, Y. Sasago, K. Uchi-
[36] T. Waki, Y. Itoh, C. Michioka, K. Yoshimura, and
nokura,and A.Sawa, Physica B 215, 164 (1995). M. Kato, Phys. Rev.B 73, 064419 (2006).
[16] S.B. Oseroff, S-W. Cheong, B. Aktas, M.F. Hundley, Z.
[37] Y. Uchiyama, Y. Sasago, I. Tsukada, K. Uchinokura,
Fisk, and L.W. Rupp Jr., Phys. Rev. Lett. 74, 1450
A. Zheludev, T. Hayashi, N. Miura, and P. Boni, Phys.
(1995). Rev. Lett.83, 632 (1999).
[17] S.J. Blundell, Contemporary Physics 40, 175 (1999).
[38] R. Ruckamp, J. Baier, M. Kriener, M.W. Haverkort, T.
[18] K.M. Kojima, Y. Fudamoto, M. Larkin, G.M. Luke,
Lorenz,G.S.Uhrig,L.Jongen,A.Moller,G.Meyer,and
J. Merrin, B. Nachumi, Y.J. Uemura, M. Hase, M. Gruninger, Phys.Rev. Lett.95, 097203 (2005).
Y.Sasago,K.Uchinokura,Y.Ajiro,A.Revcolevschi,and
J.-P. Renard,Phys. Rev.Lett. 79, 503 (1997).