Spin density wave anomaly at 140 K in the ternary iron arsenide BaFe2As2

Spin-density-wave anomaly at 140 K in the ternary iron arsenide BaFe2As2
Marianne Rotter, Marcus Tegel and Dirk Johrendt∗
Department Chemie und Biochemie, Ludwig-Maximilians-Universita¨t Mu¨nchen,
Butenandtstrasse 5-13 (Haus D), 81377 Mu¨nchen, Germany
Inga Schellenberg, Wilfried Hermes, Rainer Po¨ttgen
Institut fu¨r Anorganische und Analytische Chemie,
Universita¨t Mu¨nster, Corrensstrasse 30, D-48149 Mu¨nster, Germany
(Dated: June 12, 2008)
The ternary iron arsenide BaFe2As2 with the tetragonal ThCr2Si2-type structure exhibits a spin
density wave (SDW) anomaly at 140 K, very similar to LaFeAsO, the parent compound of the
iron arsenide superconductors. BaFe2As2 is a poor Pauli-paramagnetic metal and undergoes a
structural and magnetic phase transition at 140 K, accompanied by strong anomalies in the specific
heat, electrical resistance and magnetic susceptibility. In the course of this transition, the space
group symmetry changes from tetragonal (I4/mmm) to orthorhombic (Fmmm). 57Fe Mo¨ssbauer
spectroscopy experiments show a single signal at room temperature and full hyperfine field splitting
below the phase transition temperature (5.2 T at 77 K). Our results suggest that BaFe2As2 can
serve as a parent compound for oxygen-free iron arsenide superconductors.
PACS numbers: 74.10.+v, 74.70.Dd, 71.27.+a, 75.30.Fv, 61.50.Ks, 61.05.cp, 33.45.+x
The recent discovery of superconductivity in doped
iron-arsenide-oxides [1] has heralded a new era in su-
perconductivity research.[2] After the first report on
LaFeAs(O1−xFx) with a critical temperature (TC) of 26
K, even higher transition temperatures up to 55 K in flu-
oride doped SmFeAs(O1−xFx) followed quickly. [3] It is
meanwhile accepted, that these materials represent the
second class of high-TC superconductors after the dis-
covery of the cuprates more than 20 years ago. [4] The
parent compound LaFeAsO crystallizes in the tetrago-
nal ZrCuSiAs-type structure (space group P4/nmm). [5]
Layers of edge-sharing OLa4/4-tetrahedra alternate with
layers of FeAs4/4 tetrahedra along the c axis, as shown
on the left hand side in Figure 1. This two-dimensional
character of LaFeAsO involves different types of chemi-
cal bonding, which is strongly ionic in the LaO layers and
rather covalent in the FeAs layers, respectively. From this
purely ionic perception, we can assume a charge transfer
according to (LaO)δ+(FeAs)δ−.
It is currently believed, that the superconductivity in
doped LaFeAsO is, like in the cuprates, not of s-wave-
type but intimately connected with magnetic fluctua-
tions and a spin density wave (SDW) anomaly within
the FeAs layers. [6, 7] Undoped LaFeAsO undergoes a
SDW-driven structural phase transition round 150 K, as-
sociated with a reduction of the lattice symmetry from
tetragonal to orthorhombic [8] and anomalies in the spe-
cific heat, electrical resistance and the magnetic suscep-
tibility. Antiferromagnetic ordering of the magnetic mo-
ments (0.36 µB/Fe) was found below TN = 134 K by
neutron scattering. [9] Electron doping with fluoride or
oxygen deficiency, as well as hole doping with strontium
∗Electronic address: johrendt@lmu.de
suppresses the phase transition and the tetragonal phase
becomes superconducting at TC = 25-41 K. [1, 10, 11]
Hence, there is strong evidence that superconductivity
in LaFeAsO emerges primarily from specific structural
and electronic conditions of the (FeAs)δ− layers.
FIG. 1: Crystal structures of LaFeAsO (left, ZrCuSiAs-type)
and BaFe2As2 (right, ThCr2Si2-type).
Another well known structure type provides very simi-
lar conditions. The ternary iron arsenide BaFe2As2 with
the tetragonal ThCr2Si2-type structure (space group
I4/mmm) [12] contains practically identical layers of
edge-sharing FeAs4/4-tetrahedra, but they are separated
by barium atoms instead of LaO sheets. Figure 1 shows
both structures in comparison. In the very large family
of ThCr2Si2-type compounds, superconductivity occurs
scarcely and only at temperatures below 5 K. [13] Exam-
ples are LaIr2Ge2, LaRu2P2, YIr2−xSi2+x and BaNi2P2.
[14, 15, 16, 17] In contrast to this, the related borocar-
bides are superconductors with higher TC ’s up to 26 K
in YPd2B2C. [18] We suggested earlier, that supercon-
ductivity in ThCr2Si2-type compounds may be associ-
ated with direct interactions between the transition metal
atoms, which can lead to structural instabilities. [19]
ar
X
iv
:0
80
5.
40
21
v2
[
co
nd
-m
at.
su
pr
-co
n]
1
2 J
un
20
08

2Regarding the previously described structural and elec-
tronic properties of LaFeAsO, we believe that among the
ThCr2Si2-type compounds especially BaFe2As2 is a very
promising candidate for superconductivity for both struc-
tural and electronic reasons. In addition to the closely
related geometry, also the electron counts of the FeAs lay-
ers in BaFe2As2 and LaFeAsO are identical, because in
both cases one electron is transferred to (FeAs) accord-
ing to Ba2+0.5(FeAs)
− and (LaO)+(FeAs)−, respectively.
Apart from indeterminate magnetic data, [20] no physi-
cal properties of BaFe2As2 are known so far. In this pa-
per, we report on a structural and magnetic phase tran-
sition at 140 K, specific heat, resistivity and magnetic
measurements as well as 57Fe Mo¨ssbauer spectroscopy of
BaFe2As2. We can show that the structural, electronic
and magnetic properties of BaFe2As2 and LaFeAsO are
remarkably similar, which renders BaFe2As2 a potential
new parent compound for oxygen-free superconductors
based on the ThCr2Si2-type structure.
BaFe2As2 was synthesized by heating a mixture of dis-
tilled barium-metal, H2-reduced iron-powder and sub-
limed arsenic at a ratio of 1.05:2:2 in an alumina crucible,
which was sealed in a silica tube under an atmosphere of
purified argon. The mixture was heated to 1123 K at a
rate of 50 K/h, kept at this temperature for 10 h and
cooled down to room temperature. The reaction prod-
uct was homogenized in an agate mortar and annealed
at 1173 K for 25 h. After cooling, the sample was ho-
mogenized again, pressed into a pellet (∅=5mm, 1mm
thick) and sintered at 973K. The obtained black crys-
talline powder of BaFe2As2 is stable in air. Phase purity
was checked by X-ray powder diffraction using a Huber
G670 Guinier imaging plate diffractometer (Cu-Kα1 radi-
ation, Ge-111 monochromator), equipped with a closed-
cycle cryostat. Rietveld refinements of BaFe2As2 were
performed with the GSAS package [21] using Thompson-
Cox-Hastings functions with asymmetry corrections as
reflection profiles. [22] Figure 2 shows the pattern of
BaFe2As2 , which could be completely fitted with a sin-
gle phase. Impurities are, if at all, less than 1 %.
FIG. 2: (Color online) X-ray powder pattern (+) and Rietveld
fit (-) of BaFe2As2 at 297 K (space group I4/mmm).
In order to check for a phase transition at about 150 K,
as known from LaFeAsO, we measured the specific heat
between 3 and 200 K by a relaxation-time method in a
Physical Properties Measurement System (PPMS, Quan-
tum Design Inc.). As it can be clearly seen from Figure 3,
we find a pronounced anomaly of Cp(T ) at ≈ 140 K. The
characteristic λ-like shape of the peak points to a second
order transition, as it is typical for magnetic ordering
or a displacive structural change. From the inflection
point of the λ-anomaly, we extracted the transition tem-
perature of 139.9±0.5 K. In the low-temperature region
the specific heat is of the form Cp = γT + βT 3. The
Debye temperature can be estimated from the equation
β = (12pi4nkB)/(5Θ3D), where n is the number of atoms
per formula unit. From a Cp/T vs. T 2 plot between 3.1
and 14 K, we determined γ = 16(2) mJ K−2 mol−1, β =
2.0(5) mJ K−4 mol−1, and ΘD = 134(1) K.
FIG. 3: Specific heat of BaFe2As2 vs. temperature.
Subsequently we have recorded X-ray powder pat-
terns of BaFe2As2 between 297 and 20 K. Several re-
flections broaden below 140 K and clearly split with fur-
ther decreasing temperature. The patterns below 136
K were indexed with an orthorhombic F -centered unit
cell (aortho =
√
2 · atetra − δ; bortho =
√
2 · btetra + δ;
cortho ≈ ctetra; δ ≈ 5 pm ). The low temperature data
could be refined in the space group Fmmm. Figure 4
shows the Rietveld fit of the data at 20 K. The con-
tinuous transition of the pattern between 150 and 40 K
as well as the variation of the lattice parameters is de-
picted in Figure 5. The space group Fmmm is a sub-
group of I4/mmm, thus a second order phase transi-
tion is in agreement with our data. In terms of group
theory, this transition is translationengleich with index
2 (I4/mmm
t2
−→ Fmmm). Crystallographic data are
summarized in Table I. The main effect of the phase tran-
sition appears in the Fe–Fe distances, where four equal
bonds of 280.2 pm length split into two pairs of 280.8
and 287.7 pm length. This supports the idea, that the
Fe–Fe interactions are strongly correlated with the SDW
anomaly and may play a certain role for the properties
of BaFe2As2.
So far, our results clearly prove a structural distor-
tion in BaFe2As2. The nature of this effect is completely
analogous to that in LaFeAsO, where a transition from

3FIG. 4: (Color online) X-ray powder pattern (+) and Rietveld
fit (-) of BaFe2As2 at 20 K (space group Fmmm).
FIG. 5: Splitting of the 110 and 112 reflections and varia-
tion of the lattice parameters with temperature. Values for
the tetragonal phase above 140 K are multiplied by
√
2 for
comparability.
the tetragonal space group P4/nmm to the orthorhom-
bic space group Cmma occurs. [8] As mentioned above,
the transition of LaFeAsO is driven by a spin density
(SDW) instability within the iron layers and therefore
causes anomalies in the electrical resistivity and mag-
netic susceptibility. We have measured these properties
of BaFe2As2 and found again very similar behavior. The
temperature dependency of the dc electrical resistance is
depicted in Figure 6. BaFe2As2 is a poor metal with a
relatively high resistance around 1 mΩcm at room tem-
perature, which decreases only slightly on cooling. At
140 K, the resistance drops abruptly at first but then
decreases monotonically to 0.2 mΩcm at 10 K. This be-
havior corresponds to undoped LaFeAsO, where the re-
sistance is of the same magnitude at room temperature
and drops in a similar fashion. [6]
Finally we have investigated the general magnetic
properties and the specific behavior at the phase tran-
sition. The magnetic susceptibility was measured with
a SQUID magnetometer (MPMS-XL5, Quantum Design
Inc.) at 0.5 T. BaFe2As2 shows a weak and only slightly
temperature-dependent paramagnetism, as it is typical
for a Pauli-paramagnetic metal. Below 140 K χ drops at
TABLE I: Crystallographic data of BaFe2As2
Temperature (K) 297 20
Space group I4/mmm Fmmm
a (pm) 396.25(1) 561.46(1)
b (pm) = a 557.42(1)
c (pm) 1301.68(3) 1294.53(3)
V (nm3) 0.20438(1) 0.40514(2)
Z 2 4
data points 8700 8675
reflections 50 74
atomic parameters 4 4
profile parameters 4 4
d range 0.979− 6.508 0.981− 6.473
RP, wRP 0.0273, 0.0358 0.0283, 0.0365
R(F2), χ2 0.0522, 1.431 0.0576, 1.392
Atomic parameters:
Ba 2a (0,0,0) 4a (0,0,0)
Uiso = 95(5) Uiso = 69(5)
Fe 4d ( 12 , 0,
1
4 ) 8f (
1
4 ,
1
4 ,
1
4 )
Uiso = 57(6) Uiso = 64(4)
As 4e (0,0,z) 8i (0,0,z)
z = 0.3545(1) z = 0.3538(1)
Uiso = 99(5) Uiso = 65(5)
Bond lengths (pm):
Ba–As 338.2(1)×8 336.9(1)×4, 338.5(1)×4
Fe–As 240.3(1)×4 239.2(1)×4
Fe–Fe 280.2(1)×4 280.7(1)×2, 278.7(1)×2
Bond angles (deg):
As–Fe–As 111.1(1)×2 111.6(1)×2
108.7(1)×4 108.7(1)×2, 108.1(1)×2
FIG. 6: dc electrical resistance of BaFe2As2 (I = 100µA).
Insert: magnetic susceptibility measured at 0.5 T).
first but it increases again below 100 K. The latter fact
may be attributed to traces of ferromagnetic impurities,
which are not detectable by the X-ray powder method.
A 57Co/Rh source was available for 57Fe Mo¨ssbauer
spectroscopy investigations. The BaFe2As2 sample was
placed in a thin-walled PVC container at a thickness
of about 4 mg Fe/cm2. The measurements were per-
formed in the usual transmission geometry at 298, 77,
and 4.2 K, the source was kept at room temperature.
57Fe Mo¨ssbauer spectra of BaFe2As2 at 298, 77, and 4.2

4FIG. 7: (Color online) 57Fe Mo¨ssbauer spectra of
BaFe2As2 with transmission integral fits.
K are shown in Figure 7 together with transmission inte-
gral fits. The corresponding fitting parameters are listed
in Table II. At room temperature we observed a single
signal at an isomer shift of δ = 0.31(1) mm·s−1. Although
the iron atoms have a non-cubic coordination by arsenic,
there was no need to consider quadrupole splitting in the
fits. The observed isomer shift is slightly smaller than in
SmFeAsO0.85 [23] and LaFeAsO. [24, 25] At 77 K, well
below the transition temperature, we observe significant
hyperfine field splitting with a hyperfine field value of
5.23(1) T, which is even slightly larger than the hyperfine
field observed for LaFeAsO (4.86 T). [25] A very small
quadrupole splitting parameter of -0.03(1) mm/s was in-
cluded in the fits. This parameter accounts for the small
tetragonal-to-orthorhombic structural distortion. A sim-
ilar value has been observed for the spin-density-wave
system LaFeAsO below the transition temperature. [24]
The quadrupole splitting parameter slightly increases to
-0.04(1) mm/s at 4.2 K (Table II). The hyperfine field at
the iron nuclei is 5.47(1) T and the corresponding mag-
netic moment was estimated as 0.4µB per iron atom.
TABLE II: Fitting parameters of 57Fe Mo¨ssbauer spec-
troscopy measurements with BaFe2As2.
T (K) δ (mm·s−1) Γ (mm·s−1) ∆EQ (mm·s−1) Bhf (T)
295 0.31(1) 0.32(1) – –
77 0.43(1) 0.33(2) -0.03(1) 5.23(1)
4.2 0.44(1) 0.25(1) -0.04(1) 5.47(1)
In summary, we have shown that the properties of the
ternary arsenide BaFe2As2 with the ThCr2Si2-type struc-
ture are remarkably similar to LaFeAsO, the parent com-
pound of the newly discovered superconductors. Both
materials are poor metals at room temperature and un-
dergo second order structural and magnetic phase tran-
sitions. The 57Fe Mo¨ssbauer data of BaFe2As2 show hy-
perfine field splitting below 140 K, which also hints at
antiferromagnetic ordering. Consequently, BaFe2As2 ex-
hibits the same SDW anomaly at 140 K as LaFeAsO at
150 K. Since this SDW instability is actually believed
to be an important prerequisite for high-TC supercon-
ductivity in iron arsenides, our results strongly suggest
that BaFe2As2 can serve as a parent compound for an-
other, oxygen-free class of iron arsenide superconduc-
tors with ThCr2Si2 type structure. All the signs are
that superconductivity in BaFe2As2 can be induced ei-
ther by electron or hole doping. If this is the case, it
would conclusively prove that superconductivity origi-
nates only from the iron arsenide layers, regardless of the
separating sheets. On the other hand, superconductivity
in doped BaFe2As2 would open new avenues to further
high-TC materials in the large family of ThCr2Si2-type
compounds.
Acknowledgments
We thank Dipl.-Chem. F. M. Schappacher for help
with the Mo¨ssbauer spectroscopy experiments. This
work was financially supported by the DFG.
[1] Y. Kamihara, T. Watanabe, M. Hirano, and H. Hosono,
J. Am. Chem. Soc. 130, 3296 (2008).
[2] D. Johrendt and R. Po¨ttgen, Angew. Chem. Int. Ed. 47,
4782 (2008).
[3] Z.-A. Ren, W. Lu, J. Yang, W. Yi, X.-L. Shen, Z.-C. Li,
G.-C. Che, X.-L. Dong, L.-L. Sun, F. Zhou, et al., Chin.
Phys. Lett. 25, 2215 (2008).
[4] J. G. Bednorz and K. A. Mu¨ller, Z. Phys. B: Condens.
Matter 64, 189 (1986).
[5] V. Johnson and W. Jeitschko, J. Solid State Chem. 11,
161 (1974).
[6] J. Dong, H. J. Zhang, G. Xu, Z. Li, G. Li, W. Z.
Hu, D. Wu, G. F. Chen, X. Dai, J. L. Luo, et al.,
arXiv:0803.3426 (2008).
[7] G. F. Chen, Z. Li, D. Wu, G. Li, W. Z. Hu, J. Dong,
P. Zheng, J. L. Luo, and N. L. Wang, arXiv:0803.3790
(2008).
[8] T. Nomura, S. W. Kim, Y. Kamihara, M. Hirano,
P. V. Sushko, K. Kato, M. Takata, A. L. Shluger, and
H. Hosono, arXiv:0804.3569 (2008).
[9] C. de la Cruz, Q. Huang, J. W. Lynn, J. Li, W. R. R. II,
J. L. Zarestky, H. A. Mook, G. F. Chen, J. L. Luo, N. L.
Wang, et al., Nature, in press, DOI:10.1038/nature07057
(2008).

5[10] W. Lu, X.-L. Shen, J. Yang, Z.-C. Li, W. Yi, Z.-A.
Ren, X.-L. Dong, G.-C. Che, L.-L. Sun, F. Zhou, et al.,
arXiv:0804.3725 (2008).
[11] H.-H. Wen, G. Mu, L. Fang, H. Yang, and X. Zhu, Eu-
rophys. Lett. 82, 17009 (2008).
[12] M. Pfisterer and G. Nagorsen, Z. Naturforsch. B: Chem.
Sci. 35, 703 (1980).
[13] R. N. Shelton, H. F. Braun, and E. Musick, Solid State
Commun. 52, 797 (1984).
[14] M. Francois, G. Venturini, J. F. Mareche, B. Malaman,
and B. Roques, J. Less-Comm. Met. 113, 231 (1985).
[15] W. Jeitschko, R. Glaum, and L. Boonk, J. Solid State
Chem. 69, 93 (1987).
[16] M. Hirjak, P. Lejay, B. Chevalier, J. Etourneau, and
P. Hagenmuller, J. Less-Comm. Met. 105, 139 (1985).
[17] T. Mine, H. Yanagi, T. Kamiya, Y. Kamihara, M. Hi-
rano, and H. Hosono, Solid State Commun., in press,
doi:10.1016/j.ssc.2008.05.010 (2008).
[18] R. J. Cava, H. Takagi, H. W. Zandbergen, J. J. Kra-
jewski, W. F. Peck, T. Siegrist, B. Batlogg, R. B. van
Dover, R. J. Felder, K. Mizuhashi, et al., Nature 367,
252 (1994).
[19] D. Johrendt, C. Felser, O. Jepsen, O. K. Andersen,
A. Mewis, and J. Rouxel, J. Solid State Chem. 130, 254
(1997).
[20] M. Pfisterer and G. Nagorsen, Z. Naturforsch. B: Chem.
Sci. 38, 811 (1983).
[21] A. C. Larson and R. B. V. Dreele, General structure anal-
ysis system (2000).
[22] L. W. Finger, D. E. Cox, and A. P. Jephcoat, J. Appl.
Crystallogr. 20, 79 (1987).
[23] I. Felner, I. Nowik, M. Tsindlekht, Z.-A. Ren, X.-L. Shen,
G.-C. Che, and Z.-X. Zhao, arxiv:0805.2794 (2008).
[24] S. Kitao, Y. Kobayashi, S. Higashitanguchi, M. Saito,
Y. Kamihara, M. Hirano, T. Mitsiu, H. Hosono, and
M. Seto, arXiv:0805.0041 (2008).
[25] H.-H. Klauss, H. Luetkens, R. Klingeler, C. Hess, F. Lit-
terst, M. Kraken, M. M. Korshunov, I. Eremin, S.-L.
Drechsler, R. Khasanov, et al., arXiv:0805.0264 (2008).