Magnetic structural change of Sr2IrO4 upon Mn doping
ABSTRACT The layered 5d transition metal oxide Sr2IrO4 has been shown to host a novel
Jeff=1/2 Mott spin orbit insulating state with antiferromagnetic ordering,
leading to comparisons with the layered cuprates. Here we study the effect of
substituting Mn for Ir in single crystals of Sr2Ir0.9Mn0.1O4 through an
investigation involving bulk measurements and resonant x-ray and neutron
scattering. We observe a new long range magnetic structure emerge upon doping
through a reordering of the spins from the basal plane to the c-axis with a
reduced ordering temperature compared to Sr2IrO4. The strong enhancement of the
magnetic x-ray scattering intensity at the L3 edge relative to the L2 edge
indicates that the Jeff=1/2 state is robust and capable of hosting a variety of
PHYSICAL REVIEW B 86, 220403(R) (2012)
Magnetic structural change of Sr2IrO4upon Mn doping
S. Calder,1,*G.-X. Cao,2,3M. D. Lumsden,1J. W. Kim,4Z. Gai,5B. C. Sales,3D. Mandrus,2,3and A. D. Christianson1
1Quantum Condensed Matter Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA
2Department of Materials Science and Engineering, University of Tennessee, Knoxville, Tennessee 37996, USA
3Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA
4Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois 60439, USA
5Center for Nanophase Materials Sciences, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA
(Received 10 August 2012; revised manuscript received 26 November 2012; published 26 December 2012)
The layered 5d transition-metal oxide Sr2IrO4 has been shown to host a novel Jeff=1
insulating state with antiferromagnetic ordering, leading to comparisons with the layered cuprates. Here we
study the effect of substituting Mn for Ir in single crystals of Sr2Ir0.9Mn0.1O4through an investigation involving
bulk measurements and resonant x-ray and neutron scattering. We observe a new long-range magnetic structure
emerge upon doping through a reordering of the spins from the basal plane to the c axis with a reduced ordering
temperature compared to Sr2IrO4. The strong enhancement of the magnetic x-ray scattering intensity at the L3
edge relative to the L2edge indicates that the Jeff=1
2state is robust and capable of hosting a variety of ground
DOI: 10.1103/PhysRevB.86.220403PACS number(s): 75.25.−j, 75.70.Tj, 78.70.Ck, 72.15.Eb
The discovery of a novel spin-orbit-coupling (SOC) driven
state in Sr2IrO4 has focused attention on the possibility of
further unique ground states emerging from the strong SOC
found in 5d transition-metal oxides (TMO).1In 5d TMOs
the large radius of the electronic wave function and increased
charge results in a delicate balance between reduced Coulomb
interactions, increased SOC, and crystal field splitting that
are all of comparable strength (∼1 eV). Investigations of 5d
systems have led to the realization of topological insulating
states,2Weyl semimetal behavior,3the Slater metal-insulator
transition,4,5unconventional superconductivity,6and poten-
tially the Kitaev model.7
Sr2IrO4 was revealed to host a Jeff=1
insulating state,1that has subsequently been established to
exist in the iridates CaIrO38–10and Sr3Ir2O7.11–13The Jeff=1
behavior arises from the d-manifold degeneracy, which is split
by the crystalline electric field into the t2gand eglevels, being
degenerate doublet. For the 5d5Ir4+ion this results in a filled
split by even the small on-site Coulomb interactions in 5d
TMOs. The similarity of the magnetic insulating state and the
crystal structure in Sr2IrO4with the layered cuprate La2CuO4
has led to interest in realizing superconductivity via doping
Sr2IrO4.14,15Indeed an important open question is whether or
and consequently able to act as a vehicle from which unusual
ground states can emerge.
Sr2IrO4 crystallizes into the tetragonal I41/acd space
group with the IrO6octahedra rotated around the c axis by
∼11◦.16,17This results in the reduced symmetry of I41/acd
compared to I4/mmm. Sr2Ir0.9Mn0.1O4, the material of inter-
est in this work, has Mn substituted on the Ir site that causes
no symmetry change, as we confirmed by x-ray diffraction
for x ≈ 0.4, with a similar doping value inferred for Fe and
2degenerate quadruplet and a high energy Jeff=1
2manifold and a half filled Jeff=1
2manifold that is
Co.18The end member for our chosen doping, Sr2MnO4,
indeed forms the I4/mmm space group.19
In this Rapid Communication we present a study of Mn-
doped Sr2IrO4. Resonant x-ray scattering (RXS) has emerged
as an important and powerful tool in the investigation of 5d
magnetism, revealing the Jeff=1
long-range magnetic structure.1We employ this technique
to test the robustness of the Jeff=1
Sr2IrO4and to observe alterations to the magnetic structure.
Neutron scattering allows for a direct comparison between the
observed and calculated magnetic Bragg scattering intensities.
We use this technique to distinguish between symmetry
allowed magnetic structures and obtain an ordered magnetic
moment. We report complementary Sr2Ir0.9Mn0.1O4 single-
crystal measurements that consider changes in bulk magnetic
correlations through magnetization and resistivity.
Single crystals of Sr2Ir0.9Mn0.1O4 were grown in a Pt
crucible using the flux method. Neutron scattering was
performed on an 11 mg single crystal, dimensions 0.2 ×
0.2 × 0.02 cm, at the High Flux Isotope Reactor (HFIR)
on the triple axis instrument HB-3 in elastic mode with
λ = 2.36 ˚ A. A pyrolytic graphite (002) monochromator and
analyzer were used and the collimation set to 48?-80?-80?-240?.
The sample was mounted in the (H0L) scattering plane.
Due to the form factor decrease in intensity with scatting
angle, high neutron absorption of iridium and small sample
mass counting times of seconds to several minutes were
employed to obtain sufficient magnetic reflections. While
the flat plate geometry of the sample suppressed neutron
absorption, we calculated the absorption at each Bragg peak
based on the geometry of the sample and subsequent beam
pathway through the volume of the sample. The absorption
factor for Sr2Ir0.9Mn0.1O4is 5.3 cm−1. RXS measurements on
a crystal of approximate size 1 × 1 × 0.5 mm were performed
carried out measurements at both the L2(12.82 keV) and L3
(11.22 keV) resonant edges of iridium. Graphite was used as
the polarization analyzer at the (0010) and (008) reflections
2state in Sr2IrO4 and the
2state in Mn-doped
1098-0121/2012/86(22)/220403(5)©2012 American Physical Society
S. CALDER et al.
PHYSICAL REVIEW B 86, 220403(R) (2012)
FIG. 1. (Coloronline)Field-cooledmagnetizationmeasurements
on single crystals of Sr2Ir0.9Mn0.1O4for applied fields parallel to the
(a) ab plane and (b) c axis. Inset shows the isothermal magnetization
from −7 to 7 T at 2 K. Zero-field resistivity measurements along the
(c) ab plane and (d) c axis.
on the L2and L3edges, respectively, to achieve a scattering
angle close to 90◦. Measurements were taken at several
reflections to investigate possible magnetic structures, with an
analysis of the photon polarization in σ-σ and σ-π allowing
the analyzer and with the detector away from any Bragg peaks
through both absorption energies. The sample magnetization
M(T,H) was measured with a Quantum Design (QD) mag-
netic property measurement system (MPMS) in applied fields
up to 7 T. The electrical resistivity ρ(T,H) was performed
using a QD 14T physical property measurement system
Undoped Sr2IrO4 forms a long-range AFM structure at
240 K with a net ferromagnetic moment arising from canting
of the spins within the basal plane, with the degree of
canting governed by the angle of rotation of the octahedra.1,20
show that the onset of magnetic correlations occurs at ∼170 K
for low field cooled measurements, significantly reduced
compared to Sr2IrO4. Increasing the applied field causes an
increase in the transition temperature. The value of 0.11μB
at 2 K in the maximum field applied of 7 T [see Fig. 1(a),
inset] is close to the saturated moment in the literature for
Sr2IrO4of 0.14μB.21The magnetization anisotropy between
the c axis and the ab plane persists upon Mn doping, with
the magnetization remaining largest for a field in ab plane.
Discrete regions, separated by anomalies in the magnetization
and resistivity, were evident in measurements for Sr2IrO4.22
A similar distinction can be applied to Sr2Ir0.9Mn0.1O4with
anomalies in the magnetization at TI= 170 K, TII= 125 K,
and TIII∼ 55 K for the 0.05 kOe results, highlighted in
Fig. 1(a). Increasing the applied field leads to a removal of
the TIIanomaly. For Sr2IrO4at low temperature (<20 K) the
magnetization increased along the c axis and decreased along
the ab plane, leading to the suggestion of increased canting of
spins along the c axis.22Figures 1(a) and 1(b), however, show
for Sr2Ir0.9Mn0.1O4that Maband Mcboth increase, indicating
no enhanced c-axis canting.
In Sr2IrO4the transition from a high-temperature metallic
phase to low-temperature insulating state occurs without a
well-defined boundary above room temperature, with debate
as to the possibility of a combination of Mott and magnetic
correlations via the Slater mechanism driving the insulating
state.23For Sr2Ir0.9Mn0.1O4the resistivity remains unchanged
through both TIand TII[see Figs. 1(c) and 1(d)]. The lack of
a discernible change in the resistivity at the upper magnetic
transitions indicates that the initial onset of magnetic order
does not affect the electron scattering rate and therefore spin
disorder scattering is not a significant term in the resistivity
in this region. At TIII, however, concurrent with the sharp
drop in the magnetization, the resistivity increases at a greater
rate. The derivative of the resistivity shows this change
to occur more clearly for both the c axis and ab plane,
shown in the insets of Figs. 1(c) and 1(d). Therefore the TIII
region shows a coupling of the magnetization and resistivity
in three dimensions, potentially due to a structural distortion
or a change in the magnetic correlations that localizes the
Having established 10% Mn doping of Sr2IrO4produces
alterations in the magnetic correlations, we consider the long-
range magnetic ordering through a combination of RXS and
neutron scattering. RXS results, presented in Figs. 2(a)–2(c),
show magnetic order for Ir ions is characterized by (10odd)
and (01odd) reflections in Sr2Ir0.9Mn0.1O4. No magnetic
scattering is observed at (00L), (11L), or (1
or at incommensurate positions at 5, 80, 120, 160, or 170 K
for the regions shown in Fig. 2(c). We note that the large
charge scattering in RXS will obscure any potential magnetic
scattering at (004n).
Figure 2(d) shows the temperature dependence of the
integrated intensity for select magnetic peaks. The onset of
long-range magnetic order occurs at ∼155 K. This is reduced
compared to the applied-field magnetization results. This is,
however, consistent with the observation from magnetiza-
tion measurements that the application of even small fields
causes an increase in the magnetic transition temperature
in Sr2Ir0.9Mn0.1O4 and as such the zero-field RXS should
produce a lower TN. To test for x-ray beam heating we
compared the temperature dependence for an attenuator in
the beam providing approximately a factor of 2 attenuation
and no difference in the ordering temperature was observed.
We cannot additionally rule out poor thermal contact or
reflections showing the same magnetic ordering temperature,
the lower temperature region around TIIIshowed anomalous
behavior. To search for potential changes in the long-range
magnetic structure through this region we performed the same
scans presented in Figs. 2(a)–2(c) at various temperatures
between 5 and 170 K and observed no removal or devel-
opment in the measured magnetic Bragg peaks. Therefore,
MAGNETIC STRUCTURAL CHANGE OF Sr2IrO4...
PHYSICAL REVIEW B 86, 220403(R) (2012)
FIG. 2. (Color online) Elastic RXS of Sr2Ir0.9Mn0.1O4at the Ir
L3edge in the σ-π mode. (a) and (b) L scans at constant (H,K)
reveal long-range magnetic order at (10odd) and (01odd) at 5 K.
The same reflections are present at 80 and 120 K. (c) No magnetic
reflections are observed at (00L), (11L), or (0.50.5L) positions at
5, 80, 120, 160, or 170 K. (d) The change in integrated intensity with
temperature for two magnetic reflections.
although the long-range magnetic structure remains unaltered,
the region around TIII appears to host a change consistent
with changes in the moment direction or a small structural
Comparing our results with RXS measurements on Sr2IrO4
structure, as evidenced by the difference in the magnetic
Bragg positions due to a change in the ordering wave
vector. Sr2IrO4has magnetic scattering at (00odd) as well
as (104n + 2) and (014n) reflections. The observation in
Sr2IrO4that (10L) ?= (01L) requires a symmetry transition
from tetragonal to orthorhombic.1Any potential departure
from tetragonal is not observed within the resolution of our
measurements for Sr2Ir0.9Mn0.1O4. The same study of Sr2IrO4
FIG. 3. (Color online) Single-crystal neutron scattering measure-
ments. (a) Rocking scans of the (103) magnetic reflection above
and below the magnetic transition temperature. Similar scans were
performed at the (101), (105), (107), and (109) magnetic reflections.
(b) The normalized experimental intensity for the measured reflec-
tions is compared to instrument resolution and absorption corrected
calculated scattering for both magnetic moments in the ab plane
and along the c axis. (c) The long-ranged magnetic structure for
found an application of a small in-plane field of H > 0.2 T
altered the magnetic structure such that the (104n + 2) peaks
disappear and new peaks appear at (10odd) reflections. This
allows us to speculate that Mn doping of Sr2IrO4simulates the
behavior of Sr2IrO4in a small magnetic field. However, as we
show below, the in-field magnetic structure presented by Kim
et al.1from RXS is not compatible with the neutron scattering
intensities observed for Sr2Ir0.9Mn0.1O4.
The unpredictable effects of multiple scattering and the
it unfeasible to obtain quantitative information regarding the
comparison at different reflections or from an azimuthal scan.
Instead we utilized the scattering values obtainable from
single-crystal neutron scattering. Such results are presented
in Fig. 3. Measurements were performed at (10L) reflections
forL = 1,3,5,7,9andexperimentalpeakintensitiesextracted.
Calculated intensities were obtained for different magnetic
models with the inclusion of the instrument resolution at each
scattering angle using ResLib,24absorption correction using
the sample and scattering geometry at each reflection and
magnetic form factor for Ir4+.25
To explore the specific nature of the long-range magnetic
ordered structure we implemented representational analysis.26
The propagation vector k = (000) was employed in our
analysis as it is consistent with all the observed magnetic
that the symmetry properties of the magnetic structure are
S. CALDER et al.
PHYSICAL REVIEW B 86, 220403(R) (2012)
FIG. 4. (Color online) RXS energy dependence of the iridium L3
edge at 11.22 keV and the L2edge at 12.82 keV. These correspond to
the inflection point at the low-energy side of the enhancement in the
fluorescence scans. Measurements at both edges of Sr2Ir0.9Mn0.1O4
only yield appreciable resonant enhancement at the L3edge for σ-π
magnetic scattering. This behavior is a signature of the Jeff=1
described by only one irreducible representation (IR). The
space group I41/acd and Ir ions on the 8a Wyckoff position
give the IRs ?1, ?3, ?6, ?8, ?9, and ?10(following the num-
bering scheme of Kovalev27). ?9and ?10describe magnetic
structures with spins in the basal plane, as has been presented
for the parent Sr2IrO4 compound.1Magnetic reflections at
(10odd) and (01odd) can be generated using these IRs.
The experimental intensity at the magnetic Bragg peaks is
compared with the calculated intensity for Sr2Ir0.9Mn0.1O4
in Fig. 3(b). Poor agreement is found, indicating a magnetic
structure with the spins in the ab plane is not compatible
with the measured scattering. ?3gives a purely ferromagnetic
magnetic structure and therefore the observed scattering is
not reproduced. ?6 produces antiferromagnetic spins along
the c axis and ferromagnetic interactions in each layer;
this structure does not reproduce the observed scattering at
(10odd) or (01odd). The only remaining IRs, ?1 and ?8,
have antiferromagnetic spins along the c axis. Both models
give identical scattering intensity and symmetry equivalent
magnetic structures through a transformation of a → b. The
in Fig. 3(b), with close agreement between experimental
and calculated intensities for all the reflections measured.
The corresponding magnetic structure is shown in Fig. 3(c).
Considering the ratio between several nuclear and magnetic
peaks yields an ordered magnetic moment of 0.5(1)μB/Ir.
Thus we can conclude that the substitution of 10% Mn on
the Ir site in Sr2IrO4results in a flipping of the spins from
the basal plane to the c axis. The scattering for Mn-doped
Sr2IrO4 occurs at the same reflections as that for Sr2IrO4
in a small applied field of 0.2 T.1The magnetic structure
presented for a small applied field is identical to that for
Sr2Ir0.9Mn0.1O4, with the spins flipped from the ab to the c
from Sr2IrO4(ab plane) to Sr3Ir2O7(c axis) due to a change
in dimensionality.11This suggests that the magnetic ordered
state in Sr2IrO4is potentially unstable and can be influenced
by other perturbations.
The resonant enhancement for Ir not only allows RXS to
be utilized to observe long-range magnetic ordering, but gives
a signature of the Jeff=1
SOC split Ir4+t2gmanifold a large enhancement is expected
at both L2 and L3 edges, however, the scattering at L2 is
forbidden for the Jeff=1
measurement of the intensity through the resonant energies at
magnetic reflections provides direct evidence for the existence
at the (1019) magnetic reflection. A large enhancement is
observed at the L3edge that has its maximum at the inflection
point of the fluorescence scans, as expected. Conversely there
is no appreciable resonant enhancement at the L2energy. The
in magnetic structure. The Jeff=1
exist in iridates other than Sr2IrO4, that host alternative crystal
symmetries,9,11,12and now upon diluting the Ir site with a 3d
magnetic ion that has reduced SOC and cannot itself host the
state. This points to the robustness of the Jeff=1
the apparent ease of altering the magnetic structure it inhabits.
It would be of interest to follow Mn-doped Sr2IrO4, or an
in long-range ordering and the subsequent effect on the Jeff=
Our results show that substituting Mn for Ir in Sr2IrO4
leads to an alteration of the magnetic structure through a
reordering and flipping of the spins. The effect of 10% Mn
doping produces magnetic Bragg positions consistent with
the application of a small field of 0.2 T to Sr2IrO4, however,
the presented magnetic structures differ through a flipping
of each spin from the ab plane to the c axis. The onset
of magnetic ordering is reduced from 240 to ∼155 K and
controllable by applied fields. Even with the altered magnetic
ordering of the Ir ions the Jeff=1
Despite speculation as to the possibilities of doping Sr2IrO4,
few experimental studies exist. The results presented suggest
an unstable magnetic structure in Sr2IrO4that can be altered
by small perturbations, whereas breaking the Jeff=1
appears to require a more dramatic alteration. Further studies
using alternative doping ions and concentrations will lead to a
greater level of understanding in this important 5d material.
2Mott insulating state.1In a non-
2SOC induced state. Therefore a
2state remains in Sr2Ir0.9Mn0.1O4, despite the change
2state has been shown to
2insulating state remains.
We thank Ling Li for supporting characterization measure-
sponsored by the Scientific User Facilities Division, Office
of Basic Energy Sciences, U.S. Department of Energy. Part
of the work (DM, BCS, GC) was supported by the Depart-
ment of Energy, Basic Energy Sciences, Materials Sciences
and Engineering Division. A portion of this research was
conducted at the Center for Nanophase Materials Sciences,
which is sponsored at Oak Ridge National Laboratory by the
Scientific User Facilities Division, Office of Basic Energy
Sciences, U.S. Department of Energy. Use of the Advanced
Photon Source, an Office of Science User Facility operated
for the U.S. DOE Office of Science by Argonne National
Laboratory, was supported by the U.S. DOE under Contract
MAGNETIC STRUCTURAL CHANGE OF Sr2IrO4...
PHYSICAL REVIEW B 86, 220403(R) (2012)
1B. J. Kim, H. Ohsumi, T. Komesu, S. Sakai, T. Morita, H. Takagi,
and T. Arima, Science 323, 1329 (2009).
2D. Pesin and L. Balents, Nat. Phys. 6, 376 (2010).
3X. Wan, A. M. Turner, A. Vishwanath, and S. Y. Savrasov, Phys.
Rev. B 83, 205101 (2011).
4S. Calder, V. O. Garlea, D. F. McMorrow, M. D. Lumsden, M. B.
Y. S. Sun, Y. Tsujimoto, and A. D. Christianson, Phys. Rev. Lett.
108, 257209 (2012).
5Y. G. Shi, Y. F. Guo, S. Yu, M. Arai, A. A. Belik, A. Sato,
K. Yamaura, E. Takayama-Muromachi, H. F. Tian, H. X. Yang,
J. Q. Li, T. Varga, J. F. Mitchell, and S. Okamoto, Phys. Rev. B 80,
6S. Yonezawa, Y. Muraoka, Y. Matsushita, and Z. Hiroi, J. Phys.:
Condens. Matter 16, L9 (2004).
7J. Chaloupka, G. Jackeli, and G. Khaliullin, Phys. Rev. Lett. 105,
8K. Ohgushi, J. Yamaura, H. Ohsumi, K. Sugimoto, S. Takeshita,
A. Tokuda, H. Takagi, M. Takata, and T. Arima, arXiv:1108.4523.
9N. A. Bogdanov, V. M. Katukuri, H. Stoll, J. van den Brink, and
L. Hozoi, Phys. Rev. B 85, 235147 (2012).
10A. Subedi, Phys. Rev. B 85, 020408 (2012).
11J. W. Kim, Y. Choi, J. Kim, J. F. Mitchell, G. Jackeli, M. Daghofer,
J. van den Brink, G. Khaliullin, and B. J. Kim, Phys. Rev. Lett. 109,
12S. Boseggia, R. Springell, H. C. Walker, A. T. Boothroyd,
D. Prabhakaran, D. Wermeille, L. Bouchenoire, S. P. Collins, and
D. F. McMorrow, Phys. Rev. B 85, 184432 (2012).
13S. Boseggia, R. Springell, H. C. Walker, A. T. Boothroyd,
D. Prabhakaran, S. P. Collins, and D. F. McMorrow, J. Phys.:
Condens. Matter 24, 312202 (2012).
14J. Kim, D. Casa, M. H. Upton, T. Gog, Y.-J. Kim, J. F.
Mitchell, M. van Veenendaal, M. Daghofer, J. van den Brink,
G. Khaliullin, and B. J. Kim, Phys. Rev. Lett. 108, 177003
15F. Wang and T. Senthil, Phys. Rev. Lett. 106, 136402
Baca, Z. R. Wang, and D. C. Johnston, Phys. Rev. B 49, 9198
17Q. Huang, J. L. Soubeyroux, O. Chmaissem, I. N. Sora, A. Santoro,
R. J. Cava, J. J. Krajewski, and W. F. Peck, Jr., J. Solid State Chem.
112, 355 (1994).
18A. J. Gatimu, R. Berthelot, S. Muir, A. W. Sleight, and
M. Subramanian, J. Solid State Chem. 190, 257 (2012).
19K. Tezuka, M. Inamura, Y. Hinatsu, Y. Shimojo, and Y. Morii,
J. Solid State Chem. 145, 705 (1999).
20G. Jackeli and G. Khaliullin, Phys. Rev. Lett. 102, 017205
21G. Cao, J. Bolivar, S. McCall, J. E. Crow, and R. P. Guertin, Phys.
Rev. B 57, R11039 (1998).
22M. Ge, T. F. Qi, O. B. Korneta, D. E. De Long, P. Schlottmann,
W. P. Crummett, and G. Cao, Phys. Rev. B 84, 100402
Phys. Rev. Lett. 108, 086403 (2012).
24A. Zheludev, RESLIB-3-axis resolution library for MATLAB.
25K. Kobayashi, T. Nagao, and M. Ito, Acta Crystallogr. Sect. A:
Found. Crystallogr. 67, 473 (2011).
26A. Wills, Physica B 276, 680 (2000).
27O. V. Kovalev, Representations of the Crystallographic Space
Groups Edition 2 (Gordon and Breach Science Publishers,
Switzerland, 1993) .