Strain induced superconductivity in the parent compound BaFe2As2
J. Engelmann, V. Grinenko, P. Chekhonin, W. Skrotzki, D. V. Efremov, S. Oswald, K. Iida, R. Hühne, J. Hänisch, M. Hoffmann, F. Kurth, L. Schultz, B. Holzapfel
aa r X i v : . [ c ond - m a t . s up r- c on ] D ec Strain induced superconductivity in the parent compound BaFe As J. Engelmann,
1, 2, ∗ V. Grinenko, † P. Chekhonin, W. Skrotzki, D. V. Efremov, S. Oswald, K. Iida, R. H¨uhne, J. H¨anisch, M. Hoffmann,
1, 2
F. Kurth,
1, 2
L. Schultz,
1, 2 and B. Holzapfel
1, 3 IFW Dresden, Helmholtzstr. 20, 01069 Dresden, Germany TU Dresden, 01062 Dresden, Germany Karlsruhe Institute of Technology (KIT),Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany
The discovery of superconductivity (SC) with a transition temperature, T c , up to 65 Kin single-layer FeSe (bulk T c = 8 K) films grown on SrTiO substrates has attracted specialattention to Fe-based thin films. The high T c is a consequence of the combined effect ofelectron transfer from the oxygen-vacant substrate to the FeSe thin film and lattice tensilestrain. Here we demonstrate the realization of SC in the parent compound BaFe As (no bulk T c ) just by tensile lattice strain without charge doping. We investigate the interplay betweenstrain and SC in epitaxial BaFe As thin films on Fe-buffered MgAl O single crystallinesubstrates. The strong interfacial bonding between Fe and the FeAs sublattice increases theFe-Fe distance due to the lattice misfit which leads to a suppression of the antiferromagneticspin density wave and induces SC with bulk- T c ≈
10 K. These results highlight the role ofstructural changes in controlling the phase diagram of Fe-based superconductors. ∗ [email protected] † [email protected] INTRODUCTION
The discovery of Fe-based superconductors (FeSC) with critical temperatures, T c , exceeding50 K has stimulated many experimental and theoretical efforts to reveal the mechanism for super-conductivity (SC) and to find compounds with even higher T c . [1–3] The band structure of thestoichiometric parent compounds yields two or three small hole pockets around Γ = (0 ,
0) and twoelectron pockets around M = ( π, π ) in the folded zone.[4, 5] The nesting between electron and holebands makes the commensurate spin density wave structure (SDW) with Q = ( π, π ) plausible. Inall FeSCs the transition from the paramagnetic to the magnetic ordered SDW phase is accompa-nied by the structural transition from the tetragonal I /mmm to the orthorhombic F mmm phase.Doping due to injection of charge carriers (electrons or holes), chemical pressure and related disor-der impairs the nesting conditions resulting in reduced structural transition temperature, T s , andN´eel temperature, T N . At some finite doping the magnetic phase gives room to SC. The vicinityto the SDW phase suggests remaining strong spin fluctuations are possibly responsible for thehigh temperature SC. Therefore, the influence of doping on the phase diagram of FeSC is rathercomplex.Another way to realize SC in FeSC is applying external pressure.[6, 7] Finite pressure suppressesthe SDW and restores the tetragonal symmetry. As a result, SC appears with T c ’s comparableto that obtained by doping. However, the existence of SC, the value of maximum T c , and thepressure under which SC occurs can vary considerably, depending on the level of hydrostaticity.[8] By investigating the pressure media with different levels of hydrostaticity, it was found that auniaxial component of the pressure dramatically affects the phase diagram of AEFe As (AE122)(AE = Alkali earth Element) compounds.[7–9] It was suggested that the uniaxial pressure compo-nent is especially efficient for suppressing SDW order and responsible for a high T c of ≈
35 K.[7–9] However, to the best of our knowledge, SC under uniaxial pressure has been only reported inCaFe As with a relatively low T c about 10 K.[10] Therefore, further experimental investigationsare required to clarify the relevance of uniaxial pressure for stabilizing the SC state in FeSC.In this article we demonstrate SC in stoichiometric parent compound Ba122 thin films solelyby strain. We focus on epitaxial FeSC thin films, which are currently of great interest due tothe ability to modify their electronic properties solely by change of the crystal structure in awell defined way. [11] Therefore, this method could aid understanding the conditions which canincrease T c of FeSC. Recently it was demonstrated that the bonding distances in FeAs planesmight be altered by the lattice misfit to the substrate in Co-doped BaFe As (Ba122) [12, 13] andin FeSe . Te . [14]. As a result, variation of T c as a function of bonding distances was observed.In particular, it was discovered that SC with a transition temperature of 65 K in single-layer FeSe[15] can be induced by charge doping from the oxygen-vacant substrate to the thin film [16] (bulk T c = 8 K [17]). However, it was emphasized that the effect of strain also plays an important rolein driving SC.[16] In order to distinguish between the effects of charge doping and strain on thephase diagram of FeSC, further investigations are necessary where these two effects are decoupled.Also, until now all the measurements regarding thin films were done for Fe-pnictide compoundswhich already demonstrate SC in the bulk form. Additionally, we show that our data provide adeeper understanding of the mechanism responsible for controlling the phase diagram of isovalentlyRu-doped Ba122. RESULTSStructural analysis
High resolution reciprocal space maps (RSM) (fig. 1a) of the Ba122 (1 0 9) and (1 0 11) diffractionpeaks as well as the (2 0 6) peak of the MgAl O (spinel) substrate reveal that substrate and thinBa122 layers ( d . d c , where d c is the critical thickness of Ba122) exhibit the same a latticeconstant ( a = 0.404 nm). Increasing the thickness of the Ba122 layer d > d c results in a relaxationof this layer, indicating a decrease of the a lattice parameter (visualized by a shift of the Ba122diffraction peaks to larger Q x values). In the θ -2 θ scans (fig. 1b) a clear trend is observed withincreasing thickness, d , of Ba122. The film with d = 10 nm (S ) has a c lattice parameter of1.267 nm, which is reduced by about 3 % in comparison to the target value ( c = 1.301 nm). For d = 60 nm two different c lattice parameters were found ( c = 1.279 nm and c = 1.301 nm) whichis explained by partial stress relaxation within the Ba122 volume. Further increasing d results incomplete stress relaxation within the whole volume of the Ba122 film. The whole set of latticeconstants depending on the film thickness is summarized in table I.The mechanism of stress relaxation is also seen in bright field micrographs of sample S and S taken by a transmission electron microscope (TEM). The images for the beam direction parallel tothe [0 1 0] direction of the spinel substrate are shown in figs. 1c, d. The thickness of Fe and Ba122layer for S are ≈
27 nm and 10 nm, respectively. The corresponding values for S are 31 nm and64 nm, respectively. In contrast to S , the sample S contains cracks aligned perpendicular to thefilm surface of Ba122 in most cases. Average distances between such cracks are in the range of a few µ m and were also confirmed by atomic force microscopy (see Supplementary Figure S2). Also theroot mean square roughness, R rms , was determined to 3.5 nm for sample S . This is much largerin comparison to the thinner samples ( R rms S = 0.7 nm, R rms S = 0.4 nm). The observed cracksare the result of the relaxation process (shortening of the a -axis of the Ba122) when the criticalthickness of the Ba122 layer is reached. In the vicinity of the cracks the Ba122 layer is partiallyflaked from the Fe buffer layer. This leads to a formation of planar defects (probably several layersof BaO) parallel to the ab -plane as can be seen in fig. 1d. For the films with d > d c the observedrelaxation process results in areas with properties similar to strain-free bulk Ba122 and areas whichare still strained by the underlying layers (see also the sketches in fig. 2b). Further analysis byconvergent beam electron diffraction [18] indicates that the strained state in the Ba122 layer isprimarily determined by the Fe layer beneath. This is in accord with the previous observation ofinterfacial bonding between Fe and the FeAs sublattice of the Ba122. [19] Magnetic properties
The SC properties of the films were investigated using several complementary techniques.The temperature dependence of the dimensionless volume magnetic susceptibility χ SC ( T )=( m ( T )- m (40 K))/ V Ba122 B and the normalized resistance R ( T )/ R (40 K) for films with various thicknessare shown in fig. 2, where m ( T ) is the sample magnetic moment, V Ba122 is the volume of theBa122 layer and B is the applied magnetic field. A decrease of d results in an increase of thesuperconducting critical temperature and the SC volume fraction of the films. The upturn of thesusceptibility in the SC state is probably caused by the paramagnetic Meissner effect (WohllebenEffect) [20–23]. The large value of the diamagnetic signal at low temperatures suggests that anessential volume of the films is in the SC state at low T (see inset of fig. 2a). The tempera-ture dependence of the normal volume magnetic susceptibility χ n ( T ) = ( m ( T )- m (300 K))/ V Ba122 B for films S and S is shown in fig. 2c. As can be seen, the amplitude of the SDW anomaly at T N ∼
130 K is reduced in the thinner film (S ), which is partially strained. Moreover, no signatureof the SDW transition was observed for the films with d <
60 nm, suggesting that in-plane tensilestrain effectively suppresses long-range SDW order, and SC does not coexist with SDW order in thethin films. T N ∼
130 K value of the unstrained thick films is consistent with T N ≈
140 K observedin stoichiometric BaFe As single crystals [27]. Electrical transport properties
The SC transition width, ∆ T c = T c , onset − T c , of the films is very large (fig. 2b, d). For films with d <
30 nm the decrease of resistance starts at relatively high temperatures T c , onset = T filc ≈
35 K,whereas zero resistance is measured at T c ≈
10 K as shown in fig. 2c, d. As can be seen in fig. 2d T filc exactly corresponds to the temperature at which splitting between zero field cooled (zfc) andfield cooled (fc) branches of the susceptibility χ ( T ) curves is observed. In turn, a strong signal insusceptibility data for the thin films with d <
30 nm is observed below T c , only (fig. 2a). From theratio between diamagnetic moments at T >
10 K and in the low- T region we concluded that inthe range between T filc and T c ≈
10 K SC exists only in a small volume fraction of the films. Thisso-called filamentary SC [7, 24] (see Supplementary Figure S3) is attributed to minor regions withslightly different strain state. The values of T c and T filc , given in table I and fig. 4, are deducedfrom the method presented in fig. 2d. Critical current densities
The presence of the ferromagnetic Fe buffer layer and the extremely planar geometry of thefilms (large demagnetization factor) do not allow to estimate directly the SC volume fraction fromthe diamagnetic screening at low temperatures (see fig. 2a). Therefore, we additionally measuredthe transport critical current density, J c , to demonstrate that a large volume of the films is in thesuperconducting state below T c . It is seen in fig. 3a that self-field J c values in the strained film S is comparable with one in optimally Co-doped Ba122 epitaxial thin films having a similar thickness d ≈
30 nm and T c ≈
22 K.[25] We suppose that the different J c and irreversibility field for sampleS and sample S are due to different pinning properties rather than related to grain connectivityor other extrinsic factors. A dominant weak-link behavior is excluded by the relative robustness of J c to external magnetic fields (see figs. 3b, c). Also we exclude a large variation of the cross sectionfor the superconducting currents since we observed a comparable diamagnetic signal for both films(see fig. 2a). Thus, the relatively high J c values obtained from transport current measurementsare an additional strong evidence for a large SC volume of the thin films with d ≤
30 nm.
DISCUSSION
We suppose that the main mechanism of inducing SC and therefore controlling the phase di-agram of Ba122 in our films is tensile in-plane strain. In order to investigate the possible role ofcharge transfer from the Fe buffer to the Ba122 layer similar to the case of FeSe single-layers (seeintroduction), we compared the strained films with non-strained one grown on different substratesbut having similar thickness of Ba122 and Fe-buffer layers. It was found that the unstrained filmsdo not show any signature of a superconducting transition (see Supplementary Figure S4). Thus,a sizable effect of the proximity to the metallic Fe buffer layer on T c is ruled out. We sum up ourmain findings in Figure 4 and Table 1. As can be seen, the films with thickness d < d c ≈
30 nmhaving a large SC volume fraction are completely strained. The Ba122 layer relaxes and SC volumeabruptly reduces when the thickness exceeds d c . The films with d ∼
80 nm have lattice parametersas unstrained bulk Ba122 and a negligibly small SC volume fraction. In order to gain a deeperunderstanding of the relevant parameters for tuning the Ba122 phase diagram, we compare ourresults with data of isovalently Ru-doped Ba122.[26, 27] As can be seen in table I the lattice pa-rameters and T c values of overdoped Ba(Fe − x Ru x ) As with x ∼ . − . T c and the shifting of the SC dome in the phase diagram for different samples of Ru-dopedBa122.[26, 27, 30] Moreover, the comparison between our data on strained Ba122 and the phasediagram of Ru-doped single crystals together with the observed mechanism of stress relaxation (asdiscussed above) shed a light on the filamentary SC with T filc ≈
35 K found in the thinnest films(see fig. 4). The SC regions with T filc can be related to the minor regions with reduced strain(most probably due to low angle grain boundaries) of the Ba122 layer. Therefore, these regionshave slightly smaller a and larger c lattice parameters compared to the main strained phase whichis consistent with an analysis using electron backscatter diffraction [31]. In particular, this showsthat T c ≥
35 K is possible by adjusting the lattice strain of Ba122.Finally we conclude that the coherent interfacial bonding between Fe and the FeAs sublatticeof the Ba122 results in an in-plane straining of Ba122 to the (001) plane of bcc Fe buffer layer, i.e.increasing a and reducing c lattice parameters of the Ba122 layer for thicknesses d < d c ≈
30 nm.This suppresses magnetism and gives rise to bulk SC with T c ≈
10 K. The large superconductingvolume fraction of the films with d ≤ d c is evidenced by a pronounced diamagnetic screening andby measurements of a finite transport critical current. The close similarities of crystal parametersand T c between thin films and Ru-doped Ba122 indicates that structural changes are one of thedominant factors for controlling T c of Ba122. Furthermore, the observation of filamentary SC at35 K in regions with a different strain state suggests that T c &
35 K is possible in Ba122 withoutdoping or applying external pressure.
METHODSSample preparation
The targets used for the thin film growth were prepared by a conventional solid state reactionprocess as described in Refs.12, 13, 32.The thin films were grown using pulsed laser deposition with a KrF excimer laser ( λ = 248 nm).The preparation process took place in an ultra-high vacuum chamber with a base pressure of10 − mbar. Prior to the deposition the substrate was heated up to 750 ◦ C to clean the surface.Subsequently, the Fe buffer layer of about 30 nm thickness was deposited at room temperaturewith a laser frequency of 5 Hz and then heated to 670 ◦ C to flatten the Fe surface.[19, 33] Thistemperature was held for 30 minutes before the Ba122 layer was grown with a frequency of 10 Hz.The layer thickness was adjusted via the pulse number at constant laser energy and was confirmedby TEM for selected samples. To achieve homogeneous samples without thickness gradient thesubstrate was rotated during the whole deposition.
Structural characterization
The c lattice parameters were calculated from X-ray diffractograms (in a Bruker D8 AdvanceDiffraktometer) in Bragg Brentano geometry using the Nelson Riley function, wheras the a latticeparameters were taken out of the RSMs measured in a Panalytical Xpert Pro system.TEM characterization of the samples was performed at FEI Tecnai-T20 TEM operating at anaccelerating voltage of 200 kV. Preparation of the TEM lamellae was done by the focused ion beam(FIB) technique (FEI Helios 600i) using a platinum protection layer and 3 kV accelerating voltagein the last FIB preparation step.Electron diffraction X-ray spectroscopy (EDX) in TEM and auger electron spectroscopy (AES)were performed to check the Ba122 stoichiometry. Sensitivity factors for the AES spectra wereevaluated using single crystals.[34] Parts of the thin film and the single crystal were sputtered, andthe resulting spectra were analyzed in a JEOL JAMP 9500F Field Emission Auger Microprobe(see Supplementary Figure S5). The resulting data are given in supplement and confirm that thethin films are stoichiometric. Transport and magnetic measurements
Transport measurements were performed using a physical property measurement system (Quan-tum Design) with four probe method. Magnetic measurements were performed in a QuantumDesign DC superconducting quantum interference device (SQUID). In the normal state electri-cal currents flow mainly through the metallic Fe buffer layer having a lower resistance than theBa122 layer. Hence, it is not possible to measure the SDW/structural transition via transportmeasurements. For the measurements of the critical current density, J c , the samples S and S were structured using ion beam etching with a stainless steel mask. Bridges of 0.45 mm widthand 0.6 mm length were fabricated. J c was determined from E ( J ) characteristics using the geom-etry parameters of the bridges and applying a 1 µ Vcm − criterion ( J ⊥ B - max. Lorentz forceconfiguration). [1] Stewart, G. R. Superconductivity in iron compounds Reviews of Modern Physics Annual Review of Condensed MatterPhysics − x F x )FeAs (x = 0.05 - 0.12) with T c = 26 K Journal of the American Chemical Society
Physical Review B Low Temperature Physics As Nature Materials As under hydrostaticconditions and its extremely high sensitivity to uniaxial stress Physical Review B Reports on Progress in Physics [9] Duncan, W. J., Welzel, O. P., Harrison, C., Wang, X. F., Chen, X. H., Grosche, F. M. and Niklowitz,P. G. High pressure study of BaFe As - the role of hydrostaticity and uniaxial stress Journal ofPhysics-condensed Matter As single crystals Physical Review B . Te . Thin Films
Physical Review Letters T c dependence for strained epitaxial Ba(Fe − x Co x ) As thin films Applied PhysicsLetters Applied Physics Letters . Te . thin films Journal of Superconductivity and Novel Magnetism Nature Materials thin films Nature Materials Proceedings of the National Academy of Sciences of the United States of America − x Co x ) As thin films accepted in Proceedings of MC Regensburg 2013 (2013)[19] Thersleff, T., Iida, K., Haindl, S., Kidszun, M., Pohl, D., Hartmann, A., Kurth, F., H¨anisch, J., H¨uhne,R., Rellinghaus, B., Schultz, L. and Holzapfel, B. Coherent interfacial bonding on the FeAs tetrahedronin Fe/Ba(Fe − x Co x ) As bilayers Applied Physics Letters Cu O − δ films Physical Review B [21] Dias, F. T., Pureur, P., Rodrigues Jr., P. and Obradors, X. Paramagnetic effect at low and highmagnetic fields in melt-textured YBa Cu O − δ Physical Review B Cu O /La . Ca . MnO superlattices Physical Review B Physical Review B As and BaFe As Physical Review B − x Co x ) As thin films Physical Review B − x Ru x As PhysicalReview B − x Ru x ) As single crystals Physical Review B As ? Physical Review Letters As and Sr(Fe,Ir) As alloys PhysicalReview B − x Ru x As single crystals Physical Review B unpublished [32] Kurth, F., Iida, K., Trommler, S., H¨anisch, J., Nenkov, K., Engelmann, J., Oswald, S., Werner, J.,Schultz, L., Holzapfel, B. and Haindl, S. Electronic phase diagram of disordered Co-doped BaFe As Superconductor Science & Technology − x Co x ) As multilayers and quasi-multilayers with T c = 29 K Physica C As grown using self-flux and Bridgman techniques Journal of Crystal Growth ACKNOWLEDGEMENT
The authors would like to thank Marko Langer, Michael K¨uhnel, Steffi Kaschube and JulianeScheiter for technical assistance and S. L. Drechsler for fruitful discussions. J. Engelmann andP. Chekhonin are grateful to the GRK1621 of the DFG for financial support. Additionally wethank for financial support from the EU (Iron-Sea under project no. FP7-283141, SUPERIRON,Grant No. 283204). We also want to thank A. Saicharan for providing single crystals for the AESmeasurements.
Competing financial interests
The authors declare that they have no competing financialinterests.
Contributions:
J.E. fabricated samples, performed R ( T )-, AFM-, θ - 2 θ -, J c measurements,analyzed the data, designed the experiments and prepared the manuscript. V.G. did magneticcharacterisation, analyzed the data, designed the experiments and prepared manuscript. P.C.carried out the TEM characterisation, M.H. the AES measurements and R.H. the RSM’s. F.K.and K.I. fabricated Ba122 pulsed laser deposition targets for thin-film deposition and helped withpreparation of the thin films. J.H. and D.V.E. prepared manuscript. B.H., S.O. and W.S. super-vised the experiments and contributed to manuscript preparation. L.S. designed and directed theresearch. All authors discussed the results and implications and commented on the manuscript atall stages. FIGURES
64 66 68 70
Log .I n t en s i t y ( a r b . un i t s ) q (°) c = 1.302 nm c = 1.264 nm (a) (b) (1011)Ba122(206)Spinel(109)Ba122 Q z [001] [100][001][001] [110][100] IronSpinelBa122Ba122Iron (d) (c) S S S S S (a)(b) FeBa122FeBa122FeBa122strainedpartially relaxedunstrained n m n m n m Q (rlu) x Q (rlu) x (rlu) (rlu)Q z Figure 1.
Structural characterization by x-ray diffraction and transmission electron microscopy. (a) Reciprocal space maps of samples S and S . Relaxation processes for sample S are clearly seen as ashift of the Ba122 reflections with regard to the substrate peak. The line is a guide for the eyes. (b) Evolutionof the (008) diffraction peak for sample S , S , and S with corresponding sketches of the strain stateof the Ba122 layer. The epitaxial growth of all Ba122 films was confirmed by θ -2 θ x-ray diffraction inBragg Brentano geometry and pole figure measurements (see Supplementary Figure S1). Bright field TEMmicrographs of two thin films: (c) sample S (scale bar 25 nm) and (d) sample S (scale bar 50 nm). -5 -4 -3 -2 -1 T (K) S C
50 100 150 20001234 S n - T(K) T N B ||c= 10 kOe S -2-1012 (d) (c) S S S S S S (b) zfc S C B||c = 20 Oe (a) fc T cfil R ( T ) / R ( K ) T (K) bulk SC T c R ( T ) / R ( K ) T (K) filamentary SC no r m a l s t a t e S C S Figure 2.
Magnetic susceptibility and resistance. (a) Temperature dependence of the dimensionlessmagnetic susceptibility χ SC (T)=( m ( T )- m (40 K))/ V Ba122 B in the superconducting state of the films withdifferent thickness, where m ( T ) is sample magnetic moment, V Ba122 is the volume of the Ba122 layer and B is the applied magnetic field (the data are not corrected by the demagnetization factors). The inset shows theregion of bulk superconductivity. (b) Temperature dependence of the normalized resistance R ( T )/ R (40 K)of the investigated thin films in zero magnetic field. (c) The normal state susceptibility χ n ( T ) = ( m ( T )- m (300 K))/ V Ba122 B . The amplitude of the SDW anomaly at T N ≈
130 K is reduced in the thinner film (S ),which is partially strained. (d) Susceptibility and resistance for sample S indicating two superconductingregions (bulk superconductivity (SC) and filamentary SC). J c ( A / c m ) external magnetic field (kOe) B || c -axis S J c ( A / c m ) external magnetic field (kOe)2 K B || c- axis S Co-doped Ba122/LSAT [
Iida et al. ] J c ( A / c m ) T / T c S S (a) (b)(c) Figure 3.
Critical current densities. (a) Temperature dependence of self-field J c for sample S and S compared with data taken from ref. [25]. The relatively large J c values indicate bulk SC for the strainedfilms. Bulk T c , was taken for normalizing the data. The error-bars are caused by the uncertainty in thefilm thickness due to the surface roughness (minimal and maximal size of Ba122 layer thickness). (b) and(c) provide J c ( B ) data for samples S and S , respectively. filamentarySCbulkSC d c
10 20 30 40 50 60 70 800510152025303540 T ( K ) d (nm) T c T cfil Figure 4.
Schematic T c − d diagram for parent compound BaFe As thin films. Filled symbolsdenote the critical temperature, T c , of the main superconducting phase. Open symbols show the onsetcritical temperature T filc of the filamentary superconductivity (SC). For d c >
30 nm the spin density wave(SDW) transition was observed at T N ∼
130 K. The colours are guide for the eyes. The error-bars wereestimated by a standard deviation of magnetic and transport measurement data.
TABLES
Table I. Lattice parameters of the thin films, the target, and for Ru-doped Ba122. The critical temperaturesare evaluated from the R ( T ) and the χ ( T ) data. The Ba122-target is a polycrytalline sample. The referencesamples [27] are single crystals.sample c -axis parameter a -axis parameter c/a V = a c T filc T c thickness d name (nm) (nm) (˚A ) (K) (K) (nm)Ba122-target [32] 1.30141(1) 0.396 3.29 204(1) S ≈ ≈ ≈ ≈ ≈ ≈ As [27] 1.3022(2) 0.39633(4) 3.29 204.55(4) - 0Ba(Fe . Ru . ) As [27] 1.2749(2) 0.40342(5) 3.16 207.49(5) - ∼∼