Magnetic structure of EuFe2As2 determined by single crystal neutron diffraction
Y. Xiao, Y. Su, M. Meven, R. Mittal, C.M.N. Kumar, T. Chatterji, S. Price, J. Persson, N. Kumar, S. K. Dhar, A. Thamizhavel, Th. Brueckel
aa r X i v : . [ c ond - m a t . s up r- c on ] A ug Magnetic structure of EuFe As determined by single crystal neutron di ff raction Y. Xiao, ∗ Y. Su, M. Meven, R. Mittal,
2, 4
C.M.N. Kumar, T. Chatterji, S. Price, J. Persson, N. Kumar, S. K. Dhar, A. Thamizhavel, and Th. Brueckel
1, 2, 5 Institut fuer Festkoerperforschung, Forschungszentrum Juelich, D-52425 Juelich, Germany Juelich Centre for Neutron Science, IFF, Forschunszentrum Juelich,Outstation at FRM II, Lichtenbergstrasse 1, D-85747 Garching, Germany FRM II, Technische Universitaet Muenchen, Lichtenbergstrasse 1, D-85747 Garching, Germany Solid State Physics Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India Juelich Centre for Neutron Science, Forschungszentrum Juelich,Outstation at Institut Laue-Langevin, BP 156, 38042 Grenoble Cedex 9, France Department of Condensed Matter Physics and Material Sciences,Tata Institute of Fundamental Research, Homi Bhabha Road, Colaba, Mumbai 400 005, India (Dated: October 24, 2018)Among various parent compounds of iron pnictide superconductors, EuFe As stands out due to the presenceof both spin density wave of Fe and antiferromagnetic ordering (AFM) of the localized Eu + moment. Singlecrystal neutron di ff raction studies have been carried out to determine the magnetic structure of this compoundand to investigate the coupling of two magnetic sublattices. Long range AFM ordering of Fe and Eu spinswas observed below 190 K and 19 K, respectively. The ordering of Fe + moments is associated with the wavevector k = (1,0,1) and it takes place at the same temperature as the tetragonal to orthorhombic structural phasetransition, which indicates the strong coupling between structural and magnetic components. The ordering of Eumoment is associated with the wave vector k = (0,0,1). While both Fe and Eu spins are aligned along the long a axis as experimentally determined, our studies suggest a weak coupling between the Fe and Eu magnetism. PACS numbers: 75.25. + z, 75.50.Ee; 74.70.-b I. INTRODUCTION
The recent discovery of pnictide superconductors hasdrawn extensive attention because it provides a new op-portunity to investigate the mechanism of superconductivity[1, 2, 3, 4, 5]. Most of the research on pnictide supercon-ductors has focused on R FeAs(O x F − x )(with R = La, Nd andSm) and A Fe As (with A = Ba, Ca and Sr), the so called’1111’ and ’112’ families. These two families are closely re-lated since both of them adopt a layered structure with a sin-gle FeAs layer in the unit cell of ’1111’ and two such layersin the unit cell of ’122’. The superconducting state can beachieved either by electron or hole doping of the parent com-pounds [6, 7, 8]. Till now, the highest T c attained is 57.4K in the electron doped Ca . Na . FeAsF ’1111’ compound[10], while for ’122’ family the highest T c of 38 K is reachedin the hole doped Ba . K . Fe As [9]. Considering that theelectronic states near the Fermi surface are dominated by con-tributions from Fe and As, it is believed that the FeAs layersare responsible for superconductivity in these compounds.Recent neutron di ff raction experiments reveal that the for-mation of the spin density wave (SDW), originating from thelong range antiferromagnetic (AFM) order of the Fe momentsat low temperature, seems to be a common feature for all theiron pnictide parent compounds [11, 12, 13, 14, 15]. The on-set of the AFM order is also accompanied by the Tetragonal-Orthorhombic (T-O) structural phase transition in the ’122’family and preceded by the T-O phase transition for the ’1111’family. Phase diagrams of some iron pnictides clearly showthat the magnetic order is suppressed with appropriate chargecarrier doping of parent compound. Concomitantly, super- conductivity emerges and reaches a high T c at optimal dop-ing [16], thus exhibiting features similar to high T c cuprates[17]. It is generally believed that the superconductivity in ironpnictides is unlikely due to simple electron-phonon coupling,as demonstrated from extensive studies of phonon dynam-ics [18, 19]. Magnetism seems to play a crucial role in theappearance of superconductivity and AFM spin fluctuationshave thus been suggested to be a possible paring mechanism.Strong evidence on the presence of resonant spin excitationin the superconducting phase has indeed been obtained fromrecent inelastic neutron scattering experiments on several op-timally doped ’122’ superconductors [20, 21, 22, 23].EuFe As is a peculiar member of iron arsenide A Fe As family since the A site is occupied by Eu + , which is an S -state (orbital angular momentum L =
0) rare-earch ion pos-sessing a 4 f structure with the electron spin S = /
2. Thetheoretical e ff ective magnetic moment of Eu + ion is 7.94 µ B .As revealed by M¨ossbauer and magnetic susceptibility mea-surement on single crystals, the Eu + moments order antifer-romagneticlly below T N ∼
20 K [24, 25]. It is also reportedthat the moment of Eu + can be realigned ferromagneticly byapplying a magnetic field [26, 27]. Besides, superconductivitycan also be achieved by replacing Eu by alkali metals, e.g. the T c is observed to be 31 K and 34.7 K for Eu . K . Fe As [28]and Eu . Na . Fe As [29], respectively. Unlike BaFe As , inwhich the superconductivity emerges with the Ni substitutionof Fe, the SDW is suppressed in EuFe − x Ni x As without theemergence of superconductivity [30]. Furthermore, the mag-netic ordering of Eu + moments evolves from AFM to ferro-magnetic at higher levels of Ni doping.Since magnetism and superconductivity appears to be inti-mately related in iron pnictides, it is therefore equally impor-tant to understand the magnetic properties especially for thecompounds that contain the magnetic lanthanide ions. The in-vestigation of the interplay between the lanthanide and ironmagnetism may also be crucial for a deeper understanding ofthe magnetic and electronic properties of iron pnictides. ForEuFe As , the magnetic ordering and the details of magneticstructure have not been clarified so far via single-crystal neu-tron di ff raction due to the extremely large neutron absorptioncross-section of Eu. Here we report neutron di ff raction stud-ies on a high-quality EuFe As single crystal using the hotneutron source. It has been observed that both the Fe + andEu + moments are ordered antiferromagnetically below 190K and 19 K, respectively. A unique propagation vector k = (1,0,1) is determined for the Fe magnetic sublattice with themoment aligned along the a axis. Furthermore, the magneticpropagation vector is determined to be k = (0,0,1) for the Eu + moment, which is also aligned along the a axis. The couplingbetween the Fe and Eu magnetic sublattices has been found tobe rather weak. The determination of the magnetic structureof EuFe As would pave the way for further investigations ofEuFe As under high pressure and strong magnetic fields. II. EXPERIMENT
EuFe As single crystals were grown by the flux method.A small amount of powdered single crystal was examined bymeans of x-ray powder di ff raction (XRD) analysis. The XRDpattern reveals a single phase of EuFe As in the tetragonalThCr Si structure with space group I4 / mmm at room tem-perature. The samples have also been characterized via themeasurements of heat capacity, resistivity and magnetic sus-ceptibility. Two prominent phase transitions can be identifiedrespectively at 190 and 19 K, consistent to those previouslyreported [24, 25]. A 50 mg single crystal with dimensionabout 5 × × was selected for neutron di ff ractionexperiment, which was performed on hot-neutron four-circledi ff ractometer HEIDI at FRM II, Garching (Germany). A Cu(220) monochromator was selected to produce a monochro-matic neutron beam with the wavelength at 0.868 ˚A. AnEr filter was used to minimize the λ / ff raction datawere collected using a He single detector at di ff erent tem-peratures from 300 K down to 2.5 K. A fine collimation ( ∼ ′ ) in front of the sample and a narrow opening of the de-tector slits were adopted to achieve a su ffi cient resolution, inorder to determine precisely the structural splitting due to or-thorhombic twinning and magnetic modulation wave vectors.To ensure the inclusion of the contributions from all possi-ble twinned domains, the integrated intensities were collectedwith the setup adopting a 60 ′ collimation and an angular open-ing of both horizontal and vertical detector slits set at 4.5degree. Furthermore, the integrated intensities for the reflec-tions with 2 θ > o and 2 θ < o were obtained respectively (200) T I n t en s i t y ( c oun t s / s e c . ) (cid:90) (degree) (b) D2+D4 I n t en s i t y ( c oun t s / s e c . ) (cid:90) (degree) (400) O (c) D1+D3 (g) I n t en s i t y ( c oun t s / s e c . ) (cid:90) (degree) (220) T (a) (400) O (040) O I n t en s i t y ( c oun t s / s e c . ) k (r. l. u.) (h) D1 D2 I n t en s i t y ( c oun t s / s e c . ) (cid:90) (degree) (-220) O (d) D3+D4 (e) (f)
FIG. 1: (Color online) (a)(b) Omega scans of tetragonal (220) T and(400) T nuclear reflection at 300 K, respectively. (C) Omega scanof orthorhombic (400) O nuclear reflection at 2.5 K. The splitting ofthe reflection indicates the existence of twining. (d) Triple splittingof the rocking curve of orthorhombic (¯220) O reflection at 2.5 K, asmeasured with the ( hk
0) aligned nearly in the horizontal scatteringplane (e) Schematic presentation of the twinned orthorhombic latticein real space. Four domain patterns are marked as D1-D4. a and b de-note long and short axis of orthorhombic lattice. Red arrows indicatethe direction of the Fe magnetic moment. (f) Schematic presentationof the reciprocal space corresponding to the twinned orthorhombicdomains. (g) The contour map of orthorhombic (400) O reflection at2.5 K. (h) Q scan of (400) O reflection. via the θ -2 θ and the rocking-curve scans. The obtained dataused for the structural refinements were normalized by mon-itor counts and corrected for the Lorentz factor. DATAP pro-gram is used to carry out absorption correction by consideringthe size and shape of crystal [31]. The absorption coe ffi cient µ is calculated to be 2.61 mm − and the transmission factorsare deduced to be only in the range from 2.1% up to 14.2%due to the extremely strong absorption. Determination of boththe nuclear and magnetic structures was performed by usingthe FULLPROF program suit [32]. The scale factor derivedfrom the crystal structure refinement was used to determine (a) (103) M (101) M _ (11L)(20L) I n t e n s i t y ( a r b . un i t s ) L (r. l. u.) (10L) (b) (10-1) M (203) M (202) N (201) M (20-1) M (-112) M (-111) N (-110) M (-11-1) N FIG. 2: (Color online) (a) The schematic diagram of ( h 0 l ) planein the first quadrant of reciprocal space. The circular, rhombic andsquare symbols represent the nuclear reflection as well as the mag-netic reflection attributed to Fe and Eu magnetic sublattices. (b) Long l scan on (01 l ), (02 l ) and (¯11 l ) reflections. the magnitude of magnetic moment from the magnetic reflec-tions. III. RESULTS AND DISCUSSION
First of all, the crystal structure of EuFe As is describedwithin the framework of tetragonal symmetry at 300 K. The ω scans of selected nuclear (220) T and (200) T reflections withmosaic width of ∼ o are shown in Fig. 1 (a) and (b), whichindicate the good quality and homogeneity of the single crys-tal. Upon cooling down, a splitting is observed for orthorhom-bic (400) O and (040) O reflections (Fig. 1(c)). Note that thosetwo reflections are corresponding to the (220) T in tetragonalsetting according to the Tetragonal-Orthorhombic symmetryrelation. The observed splitting of (400) O is the indicationof T-O structural transition and accompanied twinning con-figuration due to the interchange of the orthorhombic a and b axes. It is known that twinning in orthorhombic structurewill result in four di ff erent domain patterns [33, 34], as illus-trated in Fig. 1 (e). Two of domains shared the same (110) plane and formed the domain pairs (D1 and D2), while an-other two shared the (¯110) plane (D3 and D4). In principle, itis possible to observe single peak, two or three or four peaksdepending on the selected reflections and the resolution of theinstrument. In Fig. 1(c), the left and right peaks can be as-signed to the contributions from the domains (D1 + D3) and(D2 + D4) respectively. Note that the ω -scan is performed withopen detector slits. Two Gaussian peaks were used to fit the(400) O and (040) O reflections and the domain population ra-tio is estimated to be around 1:1 for the ( h
00) and (0 k
0) twins, i.e. D1 + D3 ≈ D2 + D4. The ω -scan of (¯220) is examined after-ward with the ( hk
0) aligned in the horizontal scattering planeto obtain more detailed information about domain population(Fig. 1(d)). The occurrence of twinning and T-O structuralphase transition can be confirmed from the clear presence oftriple splitting of (¯220) nuclear reflection. Usually, the reflec-tions with h = k =/ ≈ D3 + D4 ≈ D2. Hence the do-main population for all those four di ff erent domain patternscan be determined roughly as 2:2:1:1. In order to investigatethe distribution of nuclear reflection in reciprocal space anddetermine the lattice parameter accurately, two dimensionalplot of (400) O reflection is shown in Fig. 1 (g). The splittingof (400) O can also be clearly seen. Totally 280 nuclear reflec-tions were collected for nuclear structure refinement withinthe Fmmm space group. Several strong reflections were ex-cluded from the refinement because of the significant extinc-tion. All atoms were refined with the isotropic temperaturefactor. The refinement results of crystal structure are listed inTable 1. The lattice parameters are deduced to be a = b = c = As at low tem-perature, the sample was cooled to 2.5 K, which is well be-low the reported Fe + and Eu + magnetic ordering temper-atures. Considering the existence of the twined ( h
00) and(0 k
0) domains, extensive search of magnetic reflections wasperformed in the a* - c* reciprocal space as schematically il-lustrated in Fig. 2(a). Additional search was also performedin the ( hhl ) reciprocal plane. Fig. 2(b) shows three typicallong l scans in the reciprocal space where in addition to theexpected nuclear reflections, two sets of magnetic superstruc-ture reflections can be clearly identified with two magneticpropagation wave vectors (1,0,1) and (0,0,1) respectively. Asan example, Q scan of (101) M magnetic reflection is plottedin Fig. 3(a) and the same scan at 200 K is also plotted to-gether for comparison. In Fig. 3(b), the θ -2 θ scan of nuclear(002) N and magnetic (003) M reflections show the same peakcenter, which indicates that the magnetic structure is commen-surate in nature. The contour map of (103) M and (401) M re-flections fully illustrated the intensity distribution as shown (101) M *k (r. l. u.) I n t en s i t y ( c oun t s / s e c . ) T = 2.5 K T = 200 K (a) (e)(c) (041) M I n t en s i t y ( c oun t s / s e c . ) k (r. l. u.) (f) I n t en s i t y ( c oun t s / s e c . ) (d) (103) M *k (r. l. u.) -1.0 -0.5 0.0 0.5 1.00200400600800 (003) M M N I n t en s i t y ( a r b . un i t s ) (cid:39)(cid:21)(cid:84)(cid:3)(cid:11) degree (cid:12) (002) N (003) M (b) FIG. 3: (Color online) (a) The comparison of the Q scan of (101) M magnetic reflection at 2.5 and 200 K. The (101) M reflection is ob-served in k scan because of the existence of twining. (b) The θ -2 θ scan of (003) N nuclear and (003) M magnetic reflections at 2.5 K, thesame scan of (003) M magnetic reflection at 300 K is also plotted forcomparison. (c) The contour map shows the Q dependence of the(103) M magnetic reflection. (d) Q scan of (103) M magnetic reflec-tion. The (103) M reflection is observed in k scan because of the exis-tence of twining. (e) The contour map of (041) M magnetic reflectionindicates the contribution of the magnetic reflection of Eu magneticsublattice. (f) Q scan of (041) M magnetic reflection. in Fig. 3(c) and Fig. 3(e). As already discussed, the contourmap of (400) O nuclear reflection (Fig. 1(g)) clearly shows tworeflections attributed to the ( h
00) and (0 k
0) twins. Two peakcenters with k = Q scan of (400) O reflection(Fig. 1(h)). In Fig. 3.(d), the Q scan of (103) M reflection can be fitted by a single Gaussianfunction with k = h0l )type reflections (with h and l equal to odd number) are asso-ciated with the ( h
00) domain and they can thus be describedwith the propagation wave vector k = (1,0,1). This wave vec-tor is exactly the same as observed in other ’122’ pnictides,such as BaFe As [13] and CaFe As [14], which is relatedto the AFM order of Fe + moments. The magnetic structurerefinement was then carried out to determine the magnitudeand direction of Fe + moment. The magnetic structure withFe saturation moment of 0.98(8) µ B aligned along the long a axis is deduced. Note that the origin of AFM order in FeAs-based pnictides is still a matter of controversy. It is arguedthat the AFM order of Fe + may arise from the SDW orderdue to the Fermi surface instability under the itinerant model -0.6 -0.3 0.0 0.3 0.60150300450600 I n t en s i t y ( C oun t s / s e c . ) (cid:90) (degree) (02-9) (a) -0.6 -0.3 0.0 0.3 0.602505007501000 I n t en s i t y ( C oun t s / s e c . ) (cid:90) (degree) (02-7) (b) -0.6 -0.3 0.0 0.3 0.60500100015002000 I n t en s i t y ( C oun t s / s e c . ) (cid:90) (degree) (02-5) (c) -0.6 -0.3 0.0 0.3 0.601000200030004000 I n t en s i t y ( C oun t s / s e c . ) (cid:90) (degree) (02-3) (d) -0.6 -0.3 0.0 0.3 0.601500300045006000 I n t en s i t y ( C oun t s / s e c . ) (cid:90) (degree) (02-1) (e) I n t eg r a t ed I n t en s i t y r a t i o / Reflections
Calculated Experimental (f)
FIG. 4: (Color online) (a)-(e) Omega scans of series of (02 l ) (with l = odd) reflections at 2.5 K. The integrated intensities of (20 l ) and (02 l )can be obtained by fitting the curves with two Gaussian functions. (f)The ratio between (20 l ) and (02 l ) reflections shows good agreementwith the calculated value.TABLE I: Refined results of the crystal and magnetic structures forEuFe As at 2.5 K (space group: Fmmm , Z = / site x y z B ( ˚A )Eu (4 a ) 0 0 0 0.81(3) k , M a ( µ B ) (0,0,1), 6.8(3)Fe (8 f ) 0.25 0.25 0 .
25 0.26(3) k , M a ( µ B ) (1,0,1), 0.98(8)As (8 i ) 0 0 0 . a , b , c ( ˚A): 5.537(2), 5.505(2), 12.057(2)Number of reflections (Nuclear): 280 RF , RF W , RF (%), χ Nuclear: 9.34, 9.67, 6.22, 7.1Number of reflections (Magnetic): 228 RF , RF W , RF (%), χ Magnetic: 9.42, 7.68, 6.53, 5.7 [36]. While other evidences support that the AFM order hasa local moment origin as in Mott insulator [37, 38], such asthe parent compound of high T c cuprates. Recent spin waveexcitation study on CaFe As suggests that the magnetism ofiron arsenide might be resulted from a complicated mixture oflocalized and itinerant properties and it should be understoodby considering both the localized and itinerant electrons [39].Consequently, the magnetic reflections with a propagationwave vector k = (0,0,1) (with h even and l odd) are due to thelong range order of the localized Eu + moments. However,the moment direction of Eu + moments is still indeterminate. F obs. (cid:11) ×1000 (cid:12) Magnetic Reflections
Nuclear Reflections F ca l. (cid:11) × (cid:12) F obs. (cid:11) ×1000 (cid:12) FIG. 5: (Color online) Integrated intensities of the nuclear and mag-netic Bragg reflections collected at 2.5 K are plotted against the cal-culated values. See text for details of the crystal and magnetic struc-ture models.
Symmetry analysis based on the representation theory indi-cates that the magnetic representation Γ for magnetic Eu + on4 a site is decomposed into three one dimensional irreduciblerepresentations: Γ , Γ and Γ . The Eu + moments are alignedin the c , a or b direction according to those three representa-tions. The observation of nonzero intensity of (00 l )(with l = odd) reflections clearly exclude the representation Γ . The ω scans on several ( hk M (with both h and k = odd) reflectionsalso exhibit considerable intensity. Thus, the moment of Eu + is expected to be aligned either along the a or b direction in the ab plane. The Q scan on (041) M reflection (Fig. 3(f)) giving apeak position of k = k value of (400) O nuclear reflection. Therefore, the momentdirection of Eu + can be determined as along the a directionsince the intensity ratio between (041) M and (401) M magneticreflections approximate equals to 73:1 for this arrangement.The structure mode is confirmed further by ω scan of series of(02 l )(with l = odd) reflections as shown in Fig. 4. Similar tosome nuclear reflections, both (20 l ) and (02 l ) magnetic reflec-tions was detected due to the twinning configuration. How-ever, the intensity ratio between (20 l ) and (02 l ) changes grad-ually with the change of the angle between the scattering planeand the c axis. The calculated intensity ratio of (20 l ) / (02 l )for di ff erent l are plotted in Fig. 4(f) and it agrees well withthe observed values which derived from the ω scans directly.By taking into account of twinning components properly, therefinement on Eu + magnetic sublattice was carried out withthe aforementioned magnetic structure model. The calculatedstructure factors are plotted against those observed and shownin Fig. 5. The reliable agreement factors confirms the pro-posed magnetic structural model eventually, i.e. the Eu + mo-ment aligns along a direction with the wave vector k = (0,0,1)and magnitude of 6.8(3) µ B . Thus the magnetic structure ofEuFe As is unambiguously determined as illustrated in Fig.6. The moment direction of Eu + is also consistent with ourresonant x-ray scattering (RXS) measurement [40].Fig. 7(a) shows the temperature dependence of the (112) M FIG. 6: (Color online) Illustration of the magnetic structures ofEuFe As at 2.5 K. The Fe moments align along a direction and orderantiferromagnetically in both a and c directions. The Eu momentsalign along a direction and order antiferromagnetically in c directiononly. The gray line outlines the orthorhombic unit cell. and (003) M magnetic reflections attributed to the ordering ofEu + moments. The onset temperature of Eu + magnetic orderis deduced to be 19 K, which is in good agreement with previ-ous report on electronic and magnetic measurements [24, 25].The magnetic ordering temperature of Fe + moment is esti-mated to be 190 K based on the temperature dependence ofthe (101) M and (103) M magnetic reflections (see Fig. 7(b)).The T-O structural phase transition also takes place at 190 Kas revealed by the sharp change of full width at half maximum(FWHM) in (040) O nuclear reflection. First principle calcula-tions suggest that the nearest and next nearest neighbor su-perexchange interactions between Fe ions lead to a frustratedmagnetic ground state in pnictides with parallel and antipar-allel arrangement of Fe spins in FeAs layer [41]. Usually,the magnetic frustration can be lifted by a structural distortionfrom low symmetry to high symmetry phase. This may be theorigin of the strong coupling between the structural and mag-netic phase transitions observed in EuFe As and other ironpnictides [12, 13, 14].Due to the localized nature of Eu 4 f state, the AFM cou-pling of Eu + moments would be described by the indirect ex-change, e.g. the Ruderman-Kittel-Kasuya-Yosida (RKKY) in-teraction as suggested by Ren et al [30]. Besides of the Eu-Euand Fe-Fe interactions, the strength of the interaction betweenthe Eu and Fe magnetic sublattices is also an interesting issue.Similar to CaFe As , the SDW transition in EuFe As canalso be suppressed continuously by applying the pressure dueto the weakening of nearest Fe-Fe exchange coupling [42].Whereas the AFM ordering temperature of Eu sublattice doesnot change significantly even the compound exhibits the pos-sible reentrant superconducting state. This may suggest theweak interaction between the Eu and Fe magnetic sublattices, (040) O (103) M (cid:11)(cid:104) (cid:12) (112) M (003) M Temperature (K) I n t eg r a t ed I n t en s i t y ( a r b . un i t s ) (a)
120 140 160 180 2000481216 0.160.200.240.280.320.36 I n t eg r a t ed I n t en s i t y ( a r b . un i t s ) F W H M ( a r b . un i t s ) (040) O (101) M Temperature (K) (103) M (b) FIG. 7: (Color online) (a) Temperature dependence of integrated in-tensity of (040) O nuclear reflection as well as the (103) M , (112) M and(003) magnetic reflections below 22 K. (b) Temperature dependenceof integrated intensity of (103) M and (101) M reflections; temperaturedependence of FWHM of (040) o reflection. which is supported by the full potential electronic structurecalculation [43]. In present neutron work, we also did not ob-serve any detectable change in Fe + magnetic moment whentemperature passing through the Eu + magnetic ordering tem-perature (Fig. 7(a)). The coupling between Fe + and Eu + moments in the ordered state varnishes entirely within a meanfield model due to geometrical frustration. Parallel alignmentof the Fe + and Eu + moments might be resulted from thermalor ground state fluctuations, as suggested in an order by dis-order scheme [44, 45]. Those results are in contrast to some’1111’ compounds, such as PrFeAsO, in which the interplaybetween Fe and Pr ordering moments might drive the negativethermal expansion [46]. IV. CONCLUSION
Single crystal neutron di ff raction experiment using a hotneutron source was performed to investigate the crystal andmagnetic structure of EuFe As . With decreasing temper-ature, the antiferromagnetic order of Fe + moments set inat 190 K with the propagation vector k = (1,0,1). Similarto BaFe As and CaFe As , the tetragonal to orthorhombicstructural transition occurs simultaneously with the AFM or-der, which indicates the strong coupling between the latticeand Fe magnetic degree of freedom. Below 19 K, the Eu + moments order antiferromagneticly with the propagation vec-tor k = (1,0,1) and are aligned along the a axis. our studiesalso suggest a weak coupling between the Fe + and Eu + mag-netic sublattice. ACKNOWLEDGMENT
We would like to thank G. McIntyre from the Institut Laue-Langevin (France) for his kind help in the data analysis. ∗ [email protected][1] Y. Kamihara, T. Watanabe, M. Hirano, and H. Hosono, J. Am.Chem. Soc. , 3296 (2008).[2] H. Takahashi, K. Igawa, K.Arii, Y. Kamihara, M. Hirano, andH. Hosono, Nature (London) , 376 (2008).[3] X. H. Chen, T. Wu, G. Wu, R. H. Liu, H. Chen, and D. F. Fang,Nature (London) , 761 (2008).[4] S. Matsuishi,Y. Inoue, T. Nomura, H. Yanagi, M. Hirano, andH. Hosono, J. Am. Chem. Soc. , 14428 (2008).[5] M. Rotter, M. Tegel, and D. Johrendt, Phys. Rev. Lett. ,107006 (2008).[6] H. H. Wen, G. Mu, L. Fang, H. Yang, and X. Zhu, Europhys.Lett. , 17009 (2008).[7] 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, and Z.-X. Zhao, Chin.Phys. Lett. , 2215 (2008).[8] S. Matsuishi, Y. Inoue, T. Nomura, M. Hirano, and H. Hosono,J. Phys. Soc. Jpn. , 113709 (2008).[9] M. Rotter, M. Tegel, and D. Johrendt, Phys. Rev. Lett. ,107006 (2008).[10] Peng Cheng, Bing Shen, Gang Mu, Xiyu Zhu, Fei Han, BinZeng, and Hai-Hu Wen, arXiv:0812.1192 (2008).[11] C. de la Cruz, Q. Huang, J. W. Lynn, J. Li, W. Ratcli ff II, J. L.Zarestky, H. A. Mook, G. F. Chen, J. L. Luo, N. L. Wang, andP. C. Dai, Nature (London) , 899 (2008).[12] Q. Huang, Y. Qiu, W. Bao, J.W. Lynn, M. A. Green, Y. Chen,T. Wu, G. Wu, and X. H. Chen, Phys. Rev. Lett. , 257003(2008).[13] Y. Su, P. Link, A. Schneidewind, Th. Wolf, Y. Xiao, R. Mittal,M. Rotter, D. Johrendt, Th. Brueckel, and M. Loewenhaupt,Phys. Rev. B , 064504 (2009).[14] A. I. Goldman, D. N. Argyriou, B. Ouladdiaf, T. Chatterji, A.Kreyssig, S. Nandi, N. Ni, S.L. Bud’ko, P. C. Canfield, and R.J. McQueeney, Phys. Rev. B , 100506(R) (2008).[15] Y. Xiao, Y. Su, R. Mittal, T. Chatterji, T. Hansen, C.M.N. Ku-mar, S. Matsuishi, H. Hosono, and Th. Brueckel, Phys. Rev. B , 060504(R) (2009).[16] J. Zhao, Q. Huang, C. de la Cruz, S. Li, J. W. Lynn, Y. Chen,M. A. Green, G. F. Chen, G. Li, Z. Li, J. L. Luo, N. L. Wang,and P. Dai, Nature Materials , 953 (2008).[17] J. G. Bednorz, and K. A. Mueller, Z. Phys. B , 189 (1986).[18] R. Mittal, Y. Su, S. Rols, T. Chatterji, S. L. Chaplot, H. Schober,M. Rotter, D. Johrendt, and Th. Brueckel, Phys. Rev. B ,104514 (2008).[19] R. Mittal, Y. Su, S. Rols, M. Tegel, S. L. Chaplot, H. Schober,T. Chatterji, D. Johrendt, and Th. Brueckel, Phys. Rev. B ,224518 (2008).[20] A. D. Christianson, E. A. Goremychkin, R. Osborn, S.Rosenkranz, M. D. Lumsden, C. D. Malliakas, I. S. Todorov,H. Claus, D. Y. Chung, M. G. Kanatzidis, R. I. Bewley and T.Guidi, Nature , 930 (2008).[21] M. D. Lumsden, A. D. Christianson, D. Parshall, M. B. Stone,S. E. Nagler, G.J. MacDougall, H. A. Mook, K. Lokshin, T.Egami, D. L. Abernathy, E. A. Goremychkin, R. Osborn, M.A. McGuire, A. S. Sefat, R. Jin, B. C. Sales, and D. Mandrus,Phys. Rev. Lett. , 107005 (2009).[22] Songxue Chi, Astrid Schneidewind, Jun Zhao, Leland W. Har-riger, Linjun Li, Yongkang Luo, Guanghan Cao, Zhu’an Xu,Micheal Loewenhaupt, Jiangping Hu, and Pengcheng Dai,Phys. Rev. Lett. , 107006 (2009).[23] Y. Su et al. , (in preparation) (2009). [24] Zhi Ren, Zengwei Zhu, Shuai Jiang, Xiangfan Xu, Qian Tao,Cao Wang, Chunmu Feng, Guanghan Cao, and Zhu’an Xu,Phys. Rev. B , 052501 (2008).[25] H. Ra ffi us, M. Moersen, B. D. Mosel, W. Mueller-Warmuth, W.Jeitschko, L. Terbuechte, and T. Vomhof, J. Phys. Chem. Solids , 135 (1993).[26] Shuai Jiang, Yongkang Luo, Zhi Ren, Zengwei Zhu, Cao Wang,Xiangfan Xu, Qian Tao, Guanghan Cao, and Zhu’an Xu, NewJ. Phys. , 025007 (2009).[27] T. Wu, G. Wu, H. Chen, Y. L. Xie, R. H. Liu, X. F. Wang, andX. H. Chen, arXiv:0808.2247.[28] H. S. Jeevan, Z. Hossain, Deepa Kasinathan, Helge Rosner, C.Geibel, and P. Gegenwart, Phys. Rev. B , 092406 (2008).[29] Yanpeng Qi, Zhaoshun Gao, Lei Wang, Dongliang Wang, Xi-anping Zhang, and Yanwei Ma, New Journal of Physics , 094426(2009).[31] P. Coppens, L. Leiserowitz, and D. Rabinovich, Acta Crystal-logr. , 1035 (1965).[32] J. L. Rodrłguez-Carvajal, Physica B , 55 (1993).[33] G. J. McIntyre, A. Renault, and G. Collin, Phys. Rev. B ,5148 (1988).[34] M. A. Tanatar, A. Kreyssig, S. Nandi, N. Ni, S. L. Bud’ko, P.C. Canfield, A. I. Goldman, and R. Prozorov, Phys. Rev. B ,180508(R) (2009).[35] Marcus Tegel, Marianne Rotter, Veronika Weiss, Falko M.Schappacher, Rainer Poettgen, and Dirk Johrendt, J. Phys.: Condens. Matter , 452201 (2008).[36] I. I. Mazin, D. J. Singh, M. D. Johannes, and M. H. Du, Phys.Rev. Lett. , 057003 (2008).[37] Qimiao Si and Elihu Abrahams, Phys. Rev. Lett. , 076401(2008).[38] Chen Fang, Hong Yao, Wei-Feng Tsai, JiangPing Hu, andSteven A. Kivelson, Phys. Rev. B , 224509 (2008).[39] Jun Zhao, D. T. Adroja, Dao-Xin Yao, R. Bewley, Shiliang Li,X. F. Wang, G. Wu, X. H. Chen, Jiangping Hu, and PengchengDai, arXiv:0903.2686.[40] Javier Herrero-Martin, Valerio Scagnoli, Claudio Mazzoli, YixiSu, Ranjan Mittal, Y. Xiao, Th. Brueckel, Kumar Neeraj, S.K. Dhar, A. Thamizhavel and Luigi Paolasini, arXiv:0906.1508(2009).[41] T. Yildirim, Phys. Rev. Lett. , 057010 (2008).[42] C. F. Miclea, M. Nicklas, H. S. Jeevan, D. Kasinathan, Z.Hossain, H. Rosner, P. Gegenwart, C. Geibel, and F. Steglich,arXiv:0808.2026.[43] H. S. Jeevan, Z. Hossain, Deepa Kasinathan, H. Rosner, C.Geibel, and P. Gegenwart, Phys. Rev. B , 052502 (2008).[44] J. Villain, R. Bidaux, J.P. Carton and R. Conte, J. Physique ,1263 (1980).[45] Jan N. Reimers, and A. J. Berlinsky, Phys. Rev. B , 9539(1993).[46] S. A. J. Kimber, D. N. Argyriou, F. Yokaichiya, K. Habicht,S. Gerischer, T. Hansen, T. Chatterji, R. Klingeler, C. Hess, G.Behr, A. Kondrat, and B. Buechner, Phys. Rev. B78