Frustration-driven C4 symmetric orders in a hetero-structured iron-based superconductor
Jong Mok Ok, S.-H. Baek, C. Hoch, R. K. Kremer, S. Y. Park, Sungdae Ji, B. Buechner, J.-H. Park, S. I. Hyun, J. H. Shim, Yunkyu Bang, E. G. Moon, I. I. Mazin, Jun Sung Kim
aa r X i v : . [ c ond - m a t . s up r- c on ] J un Frustration-driven C symmetric orders in a hetero-structured iron-basedsuperconductor Jong Mok Ok,
1, 2
S.-H. Baek, ∗ C. Hoch, R. K. Kremer, S. Y. Park, Sungdae Ji,
1, 5
B. B¨uchner, J.-H.Park,
1, 5, 6
S. I. Hyun, J. H. Shim, Yunkyu Bang, E. G. Moon, I. I. Mazin, and Jun Sung Kim
1, 2, † Department of Physics, Pohang University of Science and Technology, Pohang 790-784, Korea Center for Artificial Low Dimensional Electronic Systems,Institute for Basic Science, Pohang 790-784, Korea IFW Dresden, Helmholtzstr. 20, 01069 Dresden, Germany Max-Planck-Institut f¨ur Festk¨orperforschung, Heisenbergstra β e 1, D-70569 Stuttgart, Germany Max Planck POSTECH Center for Complex Phase Materials,Pohang University of Science and Technology, Pohang 790-784, Korea Division of Advanced Materials Science, Pohang University of Science and Technology, Pohang 790-784, Korea Department of Chemistry, Pohang University of Science and Technology, Pohang 790-784, Korea Department of Physics, Chonnam National University, Korea Department of Physics, Korea Advanced Institute of Science and Technology, Daejeon 305-701, Korea Naval Research Laboratory, code 6390, 4555 Overlook Avenue S.W., Washington, DC 20375, USA (Dated: September 3, 2018)A subtle balance between competing interactions in strongly correlated systems can be easilytipped by additional interfacial interactions in a heterostructure. This often induces exotic phaseswith unprecedented properties, as recently exemplified by high- T c superconductivity in FeSe mono-layer on the nonmagnetic SrTiO . When the proximity-coupled layer is magnetically active, evenricher phase diagrams are expected in iron-based superconductors (FeSCs), which however has notbeen explored due to the lack of a proper material system. One promising candidate is Sr VO FeAs,a naturally-assembled heterostructure of a FeSC and a Mott-insulating vanadium oxide. Here, us-ing high-quality single crystals and high-accuracy As and V nuclear magnetic resonance (NMR)measurements, we show that a novel electronic phase is emerging in the FeAs layer below T ∼ C -symmetry in the FeAs layers, whilesuppressing the Neel antiferromagnetism in the SrVO layers. These findings demonstrate that themagnetic proximity coupling is effective to stabilize a hidden order in FeSCs and, more generally,in strongly correlated heterostructures. In strongly correlated electron materials, includingcuprates, transition metal oxides (TMOs), and iron-based superconductors (FeSCs), competing interactionsof spin, charge and orbital degrees of freedom lead tocomplex and rich phase diagrams, extremely sensitive toexternal perturbations. Especially impressive is modifi-cation of the phase diagram via introducing interfacial in-teractions, as intensively studied for the heterostructuresof high- T c cuprates[1–5] or transition metal oxides[6, 7],showing the enhanced T c or new emergent phases thatcannot be stabilized in their constituent layer alone. Thesimilar effect has also been found in FeSCs, for example,in FeSe monolayers on top of nonmagnetic SrTiO [8–10]showing drastically enhanced T c , arguably higher than100 K (Ref. [9]). Although the underlying mechanism isyet to be confirmed, the interfacial coupling is consideredto be critical and may further enhance T c in the super-lattice [11]. Of particular interest is when the proximitycoupled layer is strongly correlated and magnetically ac-tive. As found in heterostructures of high- T c cupratesand magnetic TMOs [3–5], additional interfacial spin in-teraction may also induce novel ground states of FeSCsin proximity of a Mott insulator, which however has notbeen explored so far. Sr VO FeAs is a unique member of the family ofFeSCs, a very rare naturally-assembled superlattice of[SrFeAs] +1 and [SrVO ] − layers[12], as shown in Fig.1a. Initially Sr VO FeAs was thought to have, be-cause of the V bands, an unusual Fermi surface topol-ogy, incompatible with s ± superconductivity scenariodriven by spin-fluctuation [13]. However, it was soonrealized the V 3 d electrons in the SrVO layer arestrongly correlated and form a Mott-insulating state [15–17], while the partially-filled Fe 3 d state in the FeAslayer has the considerable itinerancy and superconductsat T c ∼
35 K [12–14]. These contrasting ground states inSr VO FeAs make this system prototypical for stronglycorrelated heterostructures based on FeSCs and TMOs.Sr VO FeAs has the Fermi surface similar to that inother FeSCs [15, 17], and thus is expected to show eitherthe stripe antiferromagnetic (AFM) order with the wavevector Q =( π, T ∼
155 K with a sizable entropy loss of ∼ R ln2 ( R is the gas constant) [19–21]. With no evidenceof a static magnetic order or another apparent symmetrybreaking, the hidden nature of this phase transition, simi-lar to the famous “hidden-order” in underdoped cupratesor a heavy fermion system URu Si , remains elusive andcontroversial [19–25], posing a challenge to our under-standing of the physics of FeSCs in proximity of a Mottinsulator.Here we report that an emergent electronic phase isdeveloped below T = 155 K in Sr VO FeAs, which ishighly distinct in nature from the transitions found inother FeSCs. Using high-accuracy As and V NMRmeasurements on high-quality single crystals under var-ious field orientations, we unambiguously show that thetransition occurs in the FeAs layer, not the SrVO layer,without breaking either time reversal symmetry and theunderlying tetragonal lattice symmetry. This impliesthat the typical stripe AFM and C nematic phases inthe FeAs layers as well as the Neel antiferromagnetismin the SrVO layer are significantly suppressed by theinterfacial coupling between itinerant iron electrons andlocalized vanadium spins. We propose that the newly-observed phase is a C -symmetric charge/orbital order,which has never been observed iron or vanadium-basedmaterials, triggered by frustration of the otherwise dom-inant Fe stripe and V Neel fluctuations. Our discovery,therefore, offers a new avenue to explore hidden phaseswith unprecedented properties in the proximity-coupledFeSCs or other strongly correlated electron systems inheterostructures. Transport and magnetic properties.
Our trans-port and magnetic measurements on high-quality singlecrystal of Sr VO FeAs shown in Figs. 1b and 1c con-firm that the transition at T is intrinsic. A weak, butdiscernible, anomaly is observed at T ∼
155 K in theresistivity ( ρ ), even more pronounced in its temperaturederivative dρ / dT . The magnetic susceptibility χ ( T ) alsoshows an anomaly at T . Above T , χ ( T ) is several timeslarger than in typical FeSCs and follows the Curie-Weisslaw with a Curie-Weiss temperature T CW ∼ −
100 K (seeSupplementary section 2). The effective magnetic mo-ment is consistent with S = 1 expected for the V ions,suggesting that χ ( T ) is dominated by localized V spins.At T ∼
155 K, χ ( T ) for both H k ab and H k c exhibitsa small jump, which corresponds to ∼ − µ B /f.u.,three orders of magnitude smaller than typical valuesof V ions ( ∼ . µ B ) in vanadium oxides [16] and Feions ( ∼ . µ B ) in FeSCs [26]. Such weak anomalies in ρ ( T ) and χ ( T ), in contrast to a strong one in the specificheat [19–21], question the previous conjectures of a long-range ordering of either V or Fe spins [19–25], and suggestthat this weak ferromagnetic response is only a side ef-fect of the true transition. However, another anomaly at T N ∼
45 K in both χ ab ( T ) and χ c ( T ) turns out to re-flect a long-range ordering of Fe, but still not V spins, asdiscussed below. Notably, neither transition is consistentwith the typical stripe AFM or nematic orders for FeSCs. As and V nuclear magnetic resonance spec-troscopy.
To gain further insight into the transition at
FIG. 1.
Basic properties of Sr VO FeAs. a , The crys-tal structure of Sr VO FeAs as a naturally-assembled het-erostructure of the [SrFeAs] +1 and [SrVO ] − layers. V ionsform a network of corner-sharing tetrahedra, while FeAs lay-ers consist of edge-sharing FeAs tetrahedra as in other iron-based superconductors. The local structure of a FeAs tetra-hedron and a VO pyramid is highlighted at the bottom.FeAs tetrahedra provide a moderate cubic crystal field split-ting, much smaller than the band widths, while the VO unitis missing one O entirely, and thus develops a strong Jahn-Teller splitting. The d xy orbital is pushed up, and the two V d electrons occupy the d xz and d yz states, forming an S = 1spin. The Fe and the V planes are bridged by the As atoms, asindicated by the dashed line. b , The resistivity ρ ( T ) in the ab plane and along the c axis shows a weak anomaly at T ∼ c -axis resistivity, which was scaled down bya factor of 100, is consistent with the quasi-2D nature. c , Themagnetic susceptibility χ ( T ), taken at H = 1 T for H ⊥ c and H k c , shows clear anomalies at T and also at T N ∼ T on a microscopic level, we measured NMR on As and V nuclei as a function of temperature for field orienta-tions parallel to a (100), c (001) and the (110) directions(Fig. 2 and the supplementary Fig. S4). The V probesthe V spin order directly and the As is a proxy forthe Fe sites, which allows us to probe the two magneticions separately. A dramatic change of the As line oc-curs near T ∼
155 K as shown in Fig. 2a, consistentwith the anomalies in ρ ( T ) and χ ( T ). Near 180 K, the As signal starts to lose its intensity rapidly and is notdetectable between 150–170 K due to the shortening ofthe spin-spin relaxation time T (Ref. 25). Strikingly, thesignal recovers below ∼
150 K at substantially higher fre-quencies, in a similar fashion for both field orientations,which contrasts the typical behaviors of As NMR foundin other FeSCs (see Supplementary Fig. S5). This is bet-ter shown in terms of the Knight shift K ≡ ( f − ν ) /ν , FIG. 2.
NMR spectra and their analysis for the Sr VO FeAs single crystal. As ( a ) and V ( b ) NMR spectraas a function of temperature, measured in H = 15 and 12 T, respectively, for the fields oriented along the a and c axes. Theunshifted Larmor frequency ( ν ≡ γ N H ) is marked by the red vertical lines. Whereas the V spectrum is nearly temperatureindependent down to 20 K, the As spectra in both field directions show a sudden shift at T ∼
155 K. c , d , Temperaturedependences of the As and V spectra in terms of the Knight shift ( K ) and the FWHM, respectively. Below T , a nearlyisotropic large jump of the As Knight shift takes place without any magnetic line broadening, contrasting with the V spectrathat remain unchanged. where ν ≡ γ n H with the nuclear gyromagnetic ratio γ n (see Fig. 2 c ). K changes abruptly at T ∼
155 K with-out any peak splitting or broadening of the full-width athalf-maximum (FWHM) across T . Conversely, the Vline barely shifts below T and down to 20 K (Figs. 2band 2d), while its FWHM gradually increases below T .The nearly unchanged V NMR line signals that the Vspins remain disordered down to low temperatures. Thiscontrasting behavior of the As and V spectra unam-biguously proves that the transition at T occurs in theFeAs layer, not in the SrVO layer, contrary to previousclaims[19–24].Having established that the phase transition at T oc-curs in the FeAs layer, we examined the low-energy Fespin dynamics, as probed by the As spin-lattice relax-ation rate T − , which reflects local spin fluctuations. Asshown in Fig. 3, at T &
240 K, ( T T ) − exhibits a typicalCurie-Weiss-like behavior with an anisotropy T − ,a /T − ,c ≈ Q = ( π,
0) andhas been observed in many FeSCs [27, 28] (see Supple-mentary section 6). With lowering temperature, a criti-cal slowdown of the ( π , 0) spin fluctuations usually con-denses into the C stripe AFM phase. For Sr VO FeAs,however, this critical growth is arrested at T ∼
200 K,showing a broad peak of ( T T ) − with an unusually large T − ,a /T − ,c ≈
6, and then the fluctuations harden all theway down to T . Across T , ( T T ) − barely changes and then quickly reaches a constant below T , behaving as aparamagnetic metal. This completely unexpected behav-ior in both K and ( T T ) − confirms that the transitionat T in Sr VO FeAs is unlike any transitions observedin FeSCs so far.Let us now discuss possible orders established below T . First of all, we can eliminate the usual suspects:stripe, double-Q[29–32], and bicollinear[33] AFM orders,observed in other FeSCs. In the first case Q = ( π,
0) andthe Fe spins aligned along the a axis ( s k a ) generate ahyperfine field H hf ∼ c axis. This wouldbe visible in the As NMR spectra as a peak splitting of ∼
10 MHz for H k c , which is far larger than the FWHMof our spectra ( ∼ s k c , a As peak splitting is expected for H k a .Even for s k b , in which case no transferred H hf and thusno peak splitting are expected, considerable line broaden-ing due to the directional fluctuations of Fe spins shouldbe seen in experiments. Neither splitting nor broadeningis observed in our experiments (Fig. 2a). For the double-Q AFM state[29–32], a combination of two spin densitywaves with Q = ( π ,0) and (0, π ), the magnetization van-ishes at one of the two Fe sublattices and is staggeredin the other. Thus, the As peak splitting is expectedfor either H k c ( s k a ) or H k a ( s k c ), as discussedin the Supplementary section 5, which can be ruled outby experiments. The bicollinear AFM order[33] can alsobe excluded with even more confidence. In this case, FIG. 3.
Fe spin fluctuations.
Temperature dependence ofthe As spin-lattice relaxation rate divided by temperature( T T ) − measured at 15 T. At high temperatures, ( T T ) − is well described by a Curie-Weiss law (solid line). Below ∼
240 K, it deviates from the diverging behavior and dropsat lower temperatures, forming a large peak centered at ∼ T T ) − reaches a con-stant value comparable to that observed at the high tempera-ture limit, implying that the spin fluctuations are completelygapped out. ( T T ) − sharply turns up at ∼
50 K, indicatingcritical slowing down of spin fluctuations toward a magneticorder. T N ∼
45 K was identified from the sharp peak ob-served for H k c . ( T T ) − drops at T c , determined by theresistivity measurements under H k ab (red arrow) and H k c (blue arrow), microscopically probing bulk superconductivityin the magnetically ordered state. the As environment is spin-imbalanced (three neighbor-ing Fe spins are aligned in one direction, and the fourthone in the opposite), and already a plain exchange cou-pling would generate two inequivalent As sites and thusa measurable splitting for any direction of external fields.Similarly, other AFM orders with more complicated spinstructures, such as a plaquette AFM order, are excludedas discussed in the Supplementary section 5. This conclu-sion is further supported by the absence of the divergingbehavior in ( T T ) − across T (Fig. 3).Having excluded static magnetic order, we considernow nematic or, as it is occasionally called, vestigial part-ners of various AFM orders. The only nematic order ob-served so far in FeSCs is the stripe-nematic order thatcreates an imbalance in the orbital population betweenFe d xz and d yz states (Fig. 4b). This, in turn, inducesan imbalance between As p x and p y orbitals and dipolarin-plane anisotropy of the As Knight shift in the twinnedcrystals, as observed in e.g. LaFeAsO for H k a belowthe nematic transition temperature[28]. A similar behav-ior is expected for the nematic partner of the bicollinearorder (Fig. 4c), which breaks the C symmetry such that the (110) and (1¯10) directions are not equivalent [34, 35].If the generated imbalance between the correspondingorbital Fe- d xz ± d yz is of the same order as in the stripe-nematic case, a peak splitting for H k (110) should bedetected. And, for the nematic partner of the plaquettemagnetic order, two inequivalent sites and thus a sizablesplitting are expected for every field direction. Yet, noneof these signatures appears in our As spectra for H || a (100), c (001) and (110) directions (Figs. 2a, 2b and theSupplementary Fig. S4). Furthermore, our single crys-tal X-ray diffraction (see Supplementary section 1), aswell as the recent ARPES study [17] do not reveal anysignature of a C symmetry breaking.Since the transition at T retains the C symmetry, andin absence of a long range magnetic order, this transitionmust generate a change in the relative occupations ofthe C orbitals, namely d xy , d z , d x − y , and d xz ± id yz .Given that at high temperature we see clear indicationsof strong spin fluctuations, we looked for a spin-drivenscenario conserving the C symmetry; a good candidateis the vestigial (nematic) partner of the double- Q AFMorder[36]. It can be visualized (Fig. 4d) as a superposi-tion of two charge/orbital density waves with Q =( π ,0)and (0, π ), which preserves the C symmetry withoutunit-cell doubling. This phase has a broken translationalsymmetry in the Fe-only square lattice, but not in theunit cell doubled to include the As atoms[36]. Formationof the intra-unit-cell charge/orbital density wave affectsthe Fe-As hybridization and modifies the hyperfine cou-pling via isotropic Fermi-contact and core-polarizationinteractions, accounting for the nearly isotropic K Knight shift (Fig. 2). One may note that due to dipoleor orbital hyperfine interactions, the Knight shift cansplit for a field parallel to (110), because half of As siteshave paramagmetic neighbors along (110), and half along(1¯10). However, the difference in the d orbital occupa-tions between non magnetic and para magnetic Fe are ex-pected to be small, likely a few % (see Supplementarysection 5), in which case the splitting will be below de-tection, consistent with our experiments.If we assume nonmagnetic origin, another plausiblecandidate could be an orbital-selective Mott transition.In this case, the most correlated Fe orbital state, likely d xy , experiences a Mott-Hubbard transition, becomingessentially gapped, while the other orbitals remain itin-erant. The resulting occupation change in the d xy stateof all Fe sites (Fig. 4e), uniformly changes the hyperfinefield at the As sites, retaining the C symmetry and ex-plaining the nearly isotropic change of K (Fig. 2). In-deed a possibility of such transition has been discussed,but, admittedly, not in undoped pnictides, but in morestrongly correlated chalcogenides [37] and (strongly un-derdoped) KFe As [38, 39].As mentioned, Sr VO FeAs experiences another tran-sition at T N ≈
45 K, which can be identified as a spindensity wave highly distinct from the typical stripe AFM.
FIG. 4.
Fe-V interfacial spin matching and possibleorders retaining a C symmetry without long-rangemagnetism. a , Stripe-type and Neel-type AFM fluctua-tions of Fe and V spins, respectively, at high temperatures.These different types of AFM fluctuations are frustrated viaFe-V spin coupling developed at low temperatures. b-d , var-ious vestigial ordered phases resulting from melting the cor-responding magnetic orders: the typical stripe nematic ( x/y symmetry broken), the bicollinear nematic (( x + y ) / ( x − y ) and the translation symmetry broken) [35], and the vesti-gial double- Q phase (only the translational symmetry bro-ken) [36]. e , d xy orbital order driven by a possible orbital-selective Mott transition. As opposed to b-d , there is nosymmetry breaking at all in this phase compared to the high-temperature phase. In b and c the symmetry breaks becausesome bonds are predominantly ferromagnetic (red) and somepredominantly antiferromagnetic (blue). In d green circlesindicate completely nonmagnetic Fe sites, while open circlescorrespond to fluctuation paramagnetic sites. The As sitesabove (up-triangle) and below (down-triangle) the Fe planeare also shown. Note that only d and e are consistent withthe observed C symmetry, as discussed in the main text, andare our favorite candidates for the hidden order below T . Indeed, ( T T ) − climbs sharply below 60 K ( ≪ T ) forboth H k a and H k c , indicating a critical slow-down ofspin fluctuations toward a magnetic ordering at T N ∼ T − ,a /T − ,c remains isotropic, suggesting thatthe coupling between Fe spins and As is due to hybridiza-tion, which can only generate a magnetic moment on Asif As environment is spin-imbalanced. This excludes suchAFM orders as stripe, Neel or double- Q , but would beconsistent with a longer period AFM order. Also theprogressive broadening of As spectrum at low temper-atures, as shown in Figs. 2a and 2c, suggests a longwavelength, and possibly incommensurate, spin densitywave. Neutron diffraction [22, 23], which observed mag-netic Bragg peaks at Q = (1 / , / ,
0) below T N ∼
45 K,is consistent with this conclusion, although it was incor-rectly attributed to an ordering of V spins in the previousstudies [16, 19–24]. Upon further temperature lowering,( T T ) − abruptly drops at T c . This proves that the su-perconducting gap opens up on the magnetic Fe sites,and emerges on the background of the remaining, butstill strong, spin fluctuations with a C symmetry below T N . How spin density wave competes or cooperate withsuperconductivity remains as an important question.We shall now address an essential question: what sup-presses the expected stripe order in the FeAs layer andthe Neel order in the SrVO layer? The former canbe suppressed via the mechanism in which Neel-typespin fluctuations of the localized magnetic moments arecoupled to the itinerant electrons’ stripe spin fluctua-tion [40]. The stripe order, with Q =( π,
0) or (0, π ), is rel-atively fragile and can give way to bicollinear, double- Q ,and, possibly, plaquette orders, due to AFM fluctuationwith additional Q ’s [36, 40, 41]. Such magnetic frustra-tion is due to the long range magnetic interactions, re-flecting the itinerancy of Fe electrons. Fluctuation at Q =( π, π ), normally weak in FeSCs, can be enhancedthrough coupling to the Q =( π, π ) fluctuations of Vspins[40] (Fig. 4a). This destabilizes the C stripe AFMor nematic orders, but encourages the C symmetric ves-tigial charge/orbital density wave phases [36, 40]. Notethat in Sr (Mg,Ti)O FeAs and Ca AlO FeAs, isostruc-tural compounds with nonmagnetic oxide layers the stan-dard stripe ordering is not suppressed [42, 43]. Clearly,frustration of stripe Fe and Neel V spin fluctuations, viamagnetic proximity coupling, is essential for inducing anunusual hidden phase in Sr VO FeAs.The coupling between the itinerant Fe electrons andthe localized V spins also suppress the Neel order in theSrVO layer. In the SrVO layers, the nearest neigh-bor superexchange interaction would dominate and gen-erate a stable Neel order. In fact, compared to otherV perovskite oxides, such as LaVO , SrVO FeAs shouldhave stronger exchange coupling, because of the morestraight V-O-V bonds. However, the measured Curie-Weiss temperature of T CW ∼ −
100 K in Sr VO FeAs isconsiderably lower than T CW ∼ −
700 K in LaVO [44].The unexpectedly low T CW comes from an additionalferromagnetic coupling between the V spins via indi-rect double-exchange-like interaction mediated by theFe electrons[45]. This frustrates and weakens the VAFM superexchange interaction suppressing the long-range V spin order at low temperatures. Indeed, in ourdetailed LDA+U calculations we found that the calcu-lated magnetic interaction is extremely sensitive to theon-site Coulomb energy U and the Hund’s coupling J .At U − J = 5 eV, the superexchange interaction, whichis inversely proportional to U , is significantly suppressed,while the Fe-mediated one is enhanced, so that the netmagnetic interaction becomes weakly ferromagnetic inthe planes. For U − J = 4, it changes sign and be-comes antiferromagnetic, consistent with a previous re-port [16]. This demonstrates that the SrVO lies on theborderline of competing phases due to a delicate balancebetween the superexchange and the additional indirectinteractions. At the same time, coupling between thestripe fluctuations in the Fe plane at Q = ( π ,0) and Neelfluctuations in the V plane Q = ( π, π ) suppresses bothorders even further [40] and prevents V spins form order-ing. The interfacial Fe-V interaction is again crucial forthe Mott-insulating SrVO layers to remain in a nearlyparamagnetic ground state. Our findings therefore man-ifest that the physics of FeSCs can become even richer inthe proximity of other correlated systems and also offersa new avenue for exploring unusual ground state in thecorrelated heterostructures. Methods
Single crystals of Sr VO FeAs were grown using self-flux techniques as follows. The mixture of SrO, VO ,Fe, SrAs, and FeAs powders with a stoichiometry ofSr VO FeAs FeAs = 1:2 were pressed into a pellet andsealed in an evacuated quartz tube under Ar atmosphere.The samples were heated to 1180 o C, held at this temper-ature for 80 hours, cooled slowly first to 950 o C at a rate of2 o C/h and then furnace-cooled. The plate-shaped singlecrystals were mechanically extracted from the flux. Highcrystallinity and stoichiometry are confirmed by the X-ray diffraction and energy-dispersive spectroscopy. Thetypical size of the single crystals is 200 × × µ m .Single crystal X-ray diffraction patterns were taken us-ing an STOE single crystal diffractometer with imageplate. Single crystal X-ray diffraction (XRD) revealsa good crystallinity in a tetragonal structure with a =3.9155(7) ˚A and c = 15.608(4) ˚A, consistent with theprevious studies on polycrystalline samples. Detailed in-formation about single crystal XRD can be found in theSupplementary Information.Conventional four-probe resistance of single crystalswas measured in a 14 T Physical Property MeasurementSystem. Single crystal magnetizations were measured ina 5 T Magnetic Property Measurement System. The sizeof one crystal was too small ( ∼ VO FeAs single crys-tals (1.2 mg) were stacked together. All single crystalswere carefully aligned along the c-axis or the ab-plane. V (nuclear spin I =7/2) NMR and As ( I =3/2)NMR measurements were carried out at external mag-netic fields of 14.983 T and 11.982 T, respectively. Thesample was rotated using a goniometer for the exactalignment along the external field. The NMR spectrawere acquired by a standard spin-echo technique with atypical π /2 pulse length 2-3 µ s and the spin-lattice re-laxation rate was obtained by a saturation method.Band structure calculations were performed using twostandard codes: an all-electron linearized augmentedplane wave method implemented in the WIEN2k pack-age [46], and a pseudopotential VASP code [47]. In bothcases the gradient-corrected functional of Perdew, Burkeand Ernzerhof was used, and special care was taken toensure proper occupancy of V orbitals in the LDA+Ucalculations. ∗ [Corresponding author:][email protected] † [Corresponding author:][email protected][1] Gozar, A. et al. High-temperature interface superconduc-tivity between metallic and insulating copper oxides. Na-ture , 782-785 (2008).[2] Wu, J. et al.
Anomalous independence of interface super-conductivity from carrier density. Nat. Mater. , 877-881 (2013).[3] Chakhalian, J. et al . Magnetism at the interface betweenferromagnetic and superconducting oxides, Nat. Phys. ,244-248 (2006).[4] Satapathy, D. K. et al . Magnetic Proxim-ity Effect in YBa Cu O /La / Ca / MnO andYBa Cu O /LaMnO δ Superlattices, Phys. Rev.Lett. et al . Long-range transfer of electron-phononcoupling in oxide superlattices, Nat. Mater. et al . Emergent phenomena at oxide inter-faces. Nat. Mater. , 103-113 (2012).[7] Chakhalian, J., Freeland, J. W., Millis, A. J., Panagopou-los, C. and Rondinelli, J. M. Colloquium: emergent prop-erties in plane view: strong correlations at oxide inter-faces. Rev. Mod. Phys. , 1189-1202 (2014).[8] Tan, S. et al . Interface-induced superconductivity andstrain-dependent spin density waves in FeSe/SrTiO thinfilms, Nat. Mater. , 634-640 (2013).[9] Ge, J.-F. et al . Superconductivity above 100 K in single-layer FeSe films on doped SrTiO , Nat Mater. , 285-289 (2015).[10] Lee, J. J. et al . Interfacial mode coupling as the origin ofthe enhancement of T c in FeSe films on SrTiO , Nature , 245-248 (2014).[11] Coh, S., Lee, D. -H., Louie, S. G., and Cohen, M. L.Proposal for a bulk material based on a monolayer FeSeon SrTiO high-temperature superconductor, Phys. Rev.B , 245138 (2016).[12] Zhu, X. et al . Transition of stoichiometric Sr VO FeAsto a superconducting state at 37.2 K, Phys. Rev. B ,220512 (2009).[13] Lee, K.-W. and Pickett, W. E. Sr VO FeAs: A nanolay-ered bimetallic iron pnictide superconductor, Europhys.Lett. , 57008 (2010).[14] Mazin, I. I. Sr VO FeAs as compared to other iron-basedsuperconductors, Phys. Rev. B , 020507 (2010).[15] Qian, T. et al . Quasinested Fe orbitals versus Mott-insulating V orbitals in superconducting Sr VFeAsO asseen from angle-resolved photoemission, Phys. Rev. B ,140513 (2011).[16] Nakamura, H. and Machida M. Magnetic ordering inblocking layer and highly anisotropic electronic struc-ture of high- T c iron-based superconductor Sr VFeAsO :LDA+U study, Phys. Rev. B , 094503 (2010).[17] Kim, Y. K. et al . Possible role of bonding angle and or-bital mixing in iron pnictide superconductivity: Com-parative electronic structure studies of LiFeAs andSr VO FeAs, Phys. Rev. B , 041116 (2015).[18] Johnston, D. C. The puzzle of high temperature super-conductivity in layered iron pnictides and chalcogenides,Advances in Physics , 803-1061 (2010).[19] Sefat, A. S. et al . Variation of physical properties in the nominal Sr V O Fe As , Physica C , 143-149 (2011).[20] Cao, G. -H. et al . Self-doping effect and successive mag-netic transitions in superconducting Sr VFeAsO , Phys.Rev. B , 104518 (2010).[21] Tatematsu, S., Satomi, E., Kobayashi, Y., and Sato, M.Magnetic ordering in V-layers of the superconductingsystem of Sr VFeAsO , J. Phys. Soc. Jpn. , 123712(2010).[22] Tegel, M. et al . Possible magnetic order and suppressionof superconductivity by V doping in Sr VO FeAs, Phys.Rev. B , 140507 (2010).[23] Hummel, F., Su, Y., Senyshyn, A., and Johrendt, D.Weak magnetism and the Mott state of vanadium in su-perconducting Sr VO FeAs, Phys. Rev. B , 144517(2013).[24] Munevar, J. et al . Static magnetic order ofSr A O Fe As ( A = Sc and V) revealed by M¨ossbauerand muon spin relaxation spectroscopies, Phys. Rev. B , 024527 (2011).[25] Ueshima, K. et al . Magnetism and superconductivity inSr VFeAsO revealed by As- and V-NMR under ele-vated pressures, Phys. Rev. B , 184506 (2014).[26] Huang, Q. et al . Neutron-diffraction measurements ofmagnetic order and a structural transition in the parentBaFe As compound of FeAs-based high-temperature su-perconductors, Phys. Rev. Lett. , 257003 (2008).[27] Kitagawa, K., Katayama, N., Ohgushi, K., Yoshida,M., and Takigawa, M. Commensurate itinerant antifer-romagnetism in BaFe As : As-NMR studies on a self-flux grown single crystal, J. Phys. Soc. Jpn. , 114709(2008).[28] Fu, M. et al . NMR search for the spin nematic state ina LaFeAsO single crystal, Phys. Rev. Lett. , 247001(2012).[29] Avci, S. et al . Magnetically driven suppression of nematicorder in an iron-based superconductor, Nat Commun ,3845 (2014).[30] B¨ohmer, A. E. et al . Superconductivityinducedre-entrance of the orthorhombic distortion inBa − x K x Fe As , Nat. Commun. , 7911 (2015).[31] Allred, J. M. et al . Double-Q spin-density wave in ironarsenide superconductors, Nat. Phys. , 493-498 (2016).[32] Wa β er, F. et al . Spin reorientation in Ba . Na . Fe As studied by single-crystal neutron diffraction, Phys. Rev.B , 060505 (2015).[33] Ma, F., Ji, W., Hu, J., Lu, J.-Y., and Xiang, T.First-Principles calculations of the electronic structureof tetragonal α -FeTe and α -FeSe crystals: Evidence fora bicollinear antiferromagnetic order, Phys. Rev. Lett. , 177003 (2009).[34] Bishop, C. B., Herbrych, J., Dagotto, E., and Moreo, A.Possible bicollinear nematic state with monoclinic lat-tice distortions in iron telluride compounds, arXiv 1704-03495 (2017).[35] Zhang, G., Glasbrenner, J. K., Flint, R., Mazin, I. I.,and Fernandes, R. M. Double-stage nematic bond or-dering above double stripe magnetism: Application toBaTi Sb O, Phys. Rev. B , 174402 (2017).[36] Fernandes, R. M., Kivelson, S. A., and Berg, E. Vestigialchiral and charge orders from bidirectional spin-density-waves: Application to the iron-based superconductors,Phys. Rev. B , 014511 (2016).[37] Yu, R., and Si, Q. Orbital-selective Mott phase in multi-orbital models for alkaline iron selenides K − x Fe − y Se , Phys. Rev. Lett. , 146402 (2013).[38] Medici, L., Giovannetti, G., and Capone, M. SelectiveMott physics as a key to iron superconductors, Phys.Rev. Lett. , 177001 (2014).[39] Misawa, T., Nakamura, K., and Imada, M. Ab initio evi-dence for strong correlation associated with Mott proxim-ity in iron-based superconductors, Phys. Rev. Lett. ,177007 (2012).[40] Wang, X. and Fernandes, R. M. Impact of local-momentfluctuations on the magnetic degeneracy of iron-basedsuperconductors, Phys. Rev. B , 144502 (2014).[41] Glasbrenner, J. K. et al . Effect of magnetic frustration onnematicity and superconductivity in iron chalcogenides,Nat. Phys. , 953 (2015).[42] Yamamoto, K. et al . Antiferromagnetic order and su-perconductivity in Sr (Mg . − x Ti . x ) O Fe As withelectron doping: As-NMR Study, J. Phys. Soc. Jpn. , 053702 (2012).[43] Kinouchi, H. et al . Antiferromagnetic spin fluctuationsand unconventional nodeless superconductivity in aniron-based new superconductor (Ca Al O − y )(Fe As ): As nuclear quadrupole resonance study, Phys. Rev.Lett. , 047002 (2011).[44] Mahajan, A. V., Johnston, D. C., Torgeson, D. R., andBorsa, F. Magnetic properties of LaVO , Phys. Rev. B , 10966 (1992).[45] Glasbrenner, J. K., ˇZuti´c, I., and Mazin, I. I. Theoryof Mn-doped II-II-V semiconductors, Phys. Rev. B ,140403 (2014).[46] Blaha, P., Schwarz, K., Madsen, G. K. H., Kvasnicka, D.and Luitz, J. Wien2K: An augmented plane wave + localorbitals program for calculating crystal properties (Techn.Universit¨at Wien, Wien, 2001).[47] Kresse G., and Furthm¨uller, J. Efficient iterative schemesfor ab initio total-energy calculations using a plane-wavebasis set, Phys. Rev. B , 11169 (1996). Acknowledgment:
The authors thank L. Boeri, R. Fernandes, S. Backes,R. Valenti, C. Kim, Y. K. Kim, J.-H. Lee and K. H.Kim for fruitful discussion. This work was supported bythe NRF through the Mid-Career Researcher Program(No. 2012-013838), SRC (No. 2011-0030785), the MaxPlanck-POSTECH Center for Complex Phase Materials(No.2011-0031558) and also by IBS (No.IBS-R014-D1-2014-a02). The work at IFW Dresden has beensupported by by the Deutsche Forschungsgemeinschaft(Germany) via DFG Research Grants BA 4927/1-3 andthe Priority Program SPP 1458. Financial supportthrough the DFG Research Training Group GRK 1621is gratefully acknowledged. J.H.P was also supported bythe National Creative Initiative (No. 2009-0081576). I.Macknowledges funding from the Office of Naval Research(ONR) through the Naval Research Laboratory’s BasicResearch Program, and from the A. von HumboldtFoundation.
Author Contributions:
J.S.K, J.M.O, and S.H.B conceived the experiments.J.M.O synthesized the samples. J.M.O carried out thetransport and magnetization measurements. S.H.B andB.B contribute to the NMR measurements and theanalysis. C.H, R.K.K. S.Y.P, S.D.J and J.H.P con-tribute to single crystal X-ray diffraction measurements.I.M, S.I.H, J.H.S, Y.B and E.G.M contribute to thetheoretical calculations and the analysis. J.M.O, S.H.B,E.G.M, I.M and J.S.K co-wrote the manuscript. Allauthors discussed the results and commented on the paper.
Competing financial interests:
The authors declare no competing financial interests.