Mossbauer Effect in the System Srl-,La,Fe03 BY P. K. GALLAGHER AND J. B. MACCHESNEY Bell Telephone Laboratories Incorporated Murray Hill New Jersey Received 1 1 th September 1967 Massbauer spectra at room temperature of samples in the system Srl-,Lax FeOBshow a single absorption peak for x = 0.2-0-6 having an isomer shift intermediate between that of iron@I) and iron(N). This is the result of an electronic exchange which is rapid compared to the lifetime of the excited state of 57Fe. Consequently the Mossbauer effect is indicative of an average oxidation state. At 78"K however Mossbauer spectra of these samples show magnetic hyperfine splitting and the individual oxidation states are readily resolved. The temperature dependence of the Mossbauer effect is indicative of the onset of the rapid electronic transfer and it is discussed with respect to the different phases within the system.In earlier studies of the Mossbauer effect of materials containing iron(IV) the authors have investigated systems in which iron(II1) and (IV) coexist in various crystalline phases. These phases are ideally of the type A~+Fe~40,+2,. Charge compensation for iron (111) present in these systems is achieved by the creation of oxygen va~ancies.''~ It is therefore of interest to compare behaviour in these systems with analogous ones in which compensation by anion vacancies can be eliminated. The system Sr,-,La,FeO was chosen for such a study as it is ideally suited for such an investigation especially since the end number of this series SrFeO, has already been well characterized.l* During the course of this work Shimony and Knudsen described the Mossbauer effect in this system and called attention to the appearance of a rapid electronic exchange between the iron ions.Such an exchange gives rise to an average oxidation state 3- of the same type as that observed for iron(I1) and iron(II1) ions on the octahedral sites in magnetite.6s8 This effect can arise if the electronic exchange is rapid compared to the time of measurement for an individual event in the Mossbauer effect. Shimony and Knudsen were unable to resolve the oxidation states even though they performed some measurements at 78°K. In this work however resolution was observed at this temperature in the initial cursory inspection of the Mossbauer effect in this system. The most obvious examples of samples showing an " average " oxidation state are those where the amounts of iron(II1) and (IV) are nearly equal (0.4 < x < 0.6).These samples are antiferromagnetic with Nee1 temperatures of 190-200°K. Consequently they exhibit magnetic hyperfine splitting at 78°K and it is the primary purpose of this work to investigate the temperature dependence of the Mossbauer effect in greater detail and thereby determine if this resolution of oxidation states is associated with the magnetic ordering process or whether it is simply the result of the temperature dependence of the electronic transfer process. Ceramic samples of Sr,-,La,FeO were prepared with x = 0.2 0-4 0.5 0.6 0.8 and 1.0. Appropriate mixtures of Fe,03 SrCO and La203 were calcined at 1200°C ground and pressed into discs to be fired at 1240°C for final sintering.However the high temperatures needed to prepare these specimens cause oxygen 40 P . K . GALLAGHER AND J . B . MACCHESNEY 41 deficiency especially in strontium-rich specimens. This was restored by anneal- ing at moderate oxygen pressures (lo3 atm) and 618°C for one week. X-ray diffraction patterns of these specimens were obtained with CrKa radiation using straumanis-type Norelco cameras (1 14-59 mm diam.). Cell constants were calculated by the least-squares refinement of Mueller Heaton and Miller. Analyses for available oxygen yielding the ratio of trivalent to tetravalent iron was performed by the method described previ~usly.~ These results are summarized in table 1. TABLE 1 .-COMPOSITION AND CELL CONSTANTS FOR SPECIMENS Srl,La,FeO3 X % Fe4+ symmetry a0 b0 CO Y .O* 100 cubic 3-850 *2 86 cubic 3.865 -4 67 cubic 3.880 *5 48 rhomb 3.889 90-26 -6 45 rhomb.3.896 90.333 -8 34 ortho? 5.510 5-640 7-81 5 1 *ot - ortho. 5.556 5.565 7.862 *ref. (1); tS. Geller and E. A. Wood Acta Cryst. 1956 9 563. Mossbauer spectra of these materials were measured on an apparatus similar to that described by Wertheim.' Temperatures were controlled by using an Andonian Dewar and control system. Temperatures were measured by means of a calibrated platinum resistance thermometer and maintained within & 0.1 "K. The furnace used for above ambient measurements is described elsewhere.' Calibrations were based upon the ground-state splitting of 57Fe and a value of 3-92mmsec-' was used from n.m.r. measurements.'2* l3 Isomer shifts of magnetically split spectra were determined from the centroids.The source was 57C0 in palladium and all values of isomer shift reported are relative to this material. The samples were prepared by making a paraffin slurry of an unweighed amount of material and spreading it on 0.005in. thick aluminum foil. An absorption due to iron impurities in the aluminium foil and beryllium windows was substracted from the spectra. We are concerned principally with the electronic exchange aspects of the Sr,-,La,FeO system therefore samples having values of x = 0.4-0.6 will be stressed and only a brief survey of the results for the samples in which x = 0.2 0-8 and 1.0 will be presented. Fig. 1 shows data used to determine accurately the N6el tempera- ture of LaFeO,. The counting rate was measured at zero velocity for a variety of temperatures.The sudden decrease in counting rate corresponds to the collapse of the magnetic hyperfine structure above the Nee1 temperature. This temperature is confirmed by breaks in the differential thermal analysis trace (10°C min-l) made in this region as shown in the insert of fig. 1. The Nee1 temperature is clearly near 750°K. A summary of the magnetic hyperfine splitting and isomer shift at various temperatures are presented in fig. 2. The temperature dependence of the isomer shift is -7 x mm sec-'OK-'. These data are in good agreement with a recent comprehensive study made of rare earth orthoferrites. l4 Nkel temperatures of Sr,. 6Lao.4Fe03 and Sr,.,La,. ,FeO were determined in a similar fashion and are shown in fig. 3. The N6el temperatures of Sr,.,La,.,FeO and Sro.8Lao.,Fe0 were in this same narrow temperature range while that of Sr,.2Lao.8Fe03 is above room temperature consistent with the high ordering tempera- ture of LaFeO,.Table 2 presents values of the various parameters at room tempera- ture and 78°K for Sr,.,La,.8Fe0 and Sro.8Lao.2Fe03. These materials are not considered in greater detail except to note the distinct oxidation states are resolved 42 SYSTEM Sr,-,La,FeO in the magnetically split samples. The temperature dependence of this resolution for Sr,.8La,.,Fe03 is similar to that for Sr,. ,La,.,FeO which will subsequently be discussed in greater detail. Before proceeding with the results of specimens (x = 0.4-0.6) we consider the phase relations encountered in this system at room temperature. ,- Using X-ray I I I I 740 745 7 50 755 7 60 "K FIG.1.-Determination of the Nkl temperature of LaFeO,. 0 incr. T; 0 decr. T. "K shift mm sec-'. FIG. 2.-Mossbauer parameters of LaFeOs as a function of temperature. 0 Hfs kOe; 0 isomer P. K . GALLAGHER AND J . B. MACCHESNEY 43 diffraction patterns as well as microscopic examinations of pretested specimens we have identified phases generally confirming the earlier results by Shimony et al. A cubic solid solution phase extended from SrFe03 to approximately Sr,. ,La,.4Fe03. An X-ray diffraction pattern of Sr,.,La,.,FeO shows no apparent change in FIG. 3.-Determination of the Nkl temperature of some Srl-xLaxFeO samples a x = 0.4; b x = 06. symmetry at 78°K. A rhombohedral phase exists within narrow limits centred about Sr,.,La,.SFeO,. A narrow range of solid solution was also observed for the orthorhombic perovskite phase whose end member is LaFeO,.Two phase regions separate the regions of solid solution. When considering electronic exchange between iron atoms having different valence states in the samples x = 0.4-0-6 it is important to distinguish between the effects exhibited by the specimens with cubic structure compared to those observed TABLE 2.-MOSSBAUER PARAMETERS FOR Sr,,LaxFe03 WHERE X = 0.2 AND 0.8 iron (IV) iron (111-IV) iron (111) X "K kOe mm sec-1 mm sec-1 kOe mm sec-1 0.2 295 229 - 0 . 4 0 470 +0-13 0.2 78 274 - 0.32 550 + 0-3 1 0.8 295 - 0.06 0.8 78 267 -0.13 395 * *present in too small amount to be determined. for rhombohedral specimens. Values of the isomer shift as a function of temperature for Sro.,La,.,Fe03 are plotted in fig.4. The abrupt discontinuity is apparent at the N6el temperature. Above this temperature the spectra show a single line which has an isomer shift corresponding to an average oxidation state. Repre- sentative spectra are presented in fig. 5. The degree of resolution of the two oxidation 44 0.30 0.20- 0.10 0 % E E \ -0.1 0 - 0'20 - 0.30 I I I - - - \- 0- - - 00 - 0 0 - - O n 00 - I 1 I I 50 I00 I50 2 00 2 50 3 00 I I I I I I I I 1 I I [rn) 0 0 B O 000 a4 8e8 140 160 180 200 220 240 80 channel no. FIG. 5.-Mossbauer spectra of Sro.6T&.4Fe03 at selected temperatures. a 78°K ; Fe(III) 457 kOe +0-23 mm sec-I ; background 1,357,742~ Fe(IV) 259 kOe -0.20 mm sec-I b 187°K; Fe(III) 393 kOe +0.17 mm sec-I background 4,178,104~ - Fe(IV) 237 kOe -0.19 EUII SW-' c 200.7"K ; +0.03 mm sec-l F.W.H.I.0.54 mm sec-l background 376,798~ F.W.H.I. 0.41 mm sec-' ; background 224,268~ d 295°K; -0.01 IIIIII s~c-', P . K . GALLAGHER AND J . B . MACCHESNEY 45 states which would be expected for such a sample can be seen for an analogous sample in fig. 6 of ref. (3). Values of the hyperfine splittings are given in fig. 6. Clearly with Sr,.,Lao.,Fe03 and also for Sro.8La,.,Fe03 the magnetic ordering process gives rise to a marked reduction in the rate of the electron exchange process. Measurements of electrical conductivity of this specimen reflect this process by an abrupt increase in the electrical conductivity upon heating Sr,. 6La,.,Fe03 through its NCel temperatures. Those samples which are rhombohedral behave differently. Fig. 7 shows repre- sentative spectra of Sr,.,Lao.,FeO at various temperatures.Below the N6el temperature two magnetically split spectra are observed corresponding to each oxidation state. Above this temperature the spectra reduce to single broad lines which are also clearly resolved with respect to oxidation state. As the sample approaches room temperature however the electronic exchange rate increases 300 3501 50 I00 I50 2 00 250 "K FIG. 6.-Magnetic hyperfine splitting of Sro. 6Lao.4Fe03 as a function of temperature. so that the two lines eventually become a broad asymmetric smear around 270°K and then begin to narrow into a single line of intermediate isomer shift at room temperature where the electronic exchange time is much less than the lifetimes of the emitting state (1.4 x 10-7sec). It appears that the basic effect in this system involves the temperature coefficient of the rate of electronic exchange Fe+4-0-Fe+3 to Fe+3-O-Fe+4.At room temperature this exchange is rapid in both the cubic and rhombohedral phases. The temperature coefficient is different for the two phases however and the exchange in the rhombohedral phase becomes sufficiently slow to lead to a resolution of oxidation states well before the NCel temperature. The cubic phase would be expected to behave similarly except that the Niel temperature is reached prior to this point of resolution. It appears that the magnetic ordering has a marked effect upon the rate of electronic exchange and in this respect it is interesting to consider a well-known system La,-,CaxMn03.1 Here exchange takes place between Mn3+ and Mn4+ ions having the 3d4 and 3d3 electronic configuration while in the 46 SYSTEM Sr,-,La,FeO present case the iron ions are 3d4 and 3d5.In the Lal-,CaxMn03 system in response to differing concentrations of Mn3f and Mn4+ several different phases are formed having different types of magnetic order. Similar behaviour is expected in this system although such measurements have not yet been completed. It is also valuable to recall that in an antiferromagnet where indirect exchange takes place between magnetic ions i.e. superexchange the spin alignment of adjacent magnetic ions can be antiparallel (G type). If this were the case exchange of c channel no. FIG. 7.-Mossbauer spectra of Sro.5Lao.sFe03 at selected temperatures. a 78°K ; Fe(III) 485 kOe t0.29 mm sec-I ; background 2,227,326~ Fe(IV) 250 kOe - 0.34 mm sec-' background 735 824c Fe(IV) -0.35 mm sec-I ; combined F.W.H.I.0.80 mm sec-' background 215,588~ Fe(IV) -0.37 mm sec-' combined F.W.H.I. 0.80 mm sec-I F.W.H.I. 0.63 mm sec-' ; background 241,864~ F.W.H.I. 0.44 mm sec-I background 2,333,648~ 6 211~7°K; Fe(III) $0.19 mm sec-I c 259°K; Fe(III) f0.15 mm sec-1 ; d 277"K +0.09 mm sec-l e 295°K ; -0.01 mm sec-I electrons between adjacent iron ions would require a change in spin direction. This would be expected to decrease the rate of exchange since at the temperature of magnetic ordering the activation energy for the electronic jump between iron ions would increase abruptly over that due to the simple polarization term because of the addition of an exchange energy term.16* l7 This is not a factor in magnetite because the ions on the B sites are ferromagnetically aligned and consequently rapid electronic exchange is observed below the temperature of magnetic alignment because no change in spin is required.The authors acknowledge the assistance of Messrs. D. N. E. Buchanan for computer programs and J. F. Potter for sample preparation. ' P. K. Gallagher J. B. MacChesney and D. N. E. Buchanan J . Chem. Physics 1964,41 2429. P. K. Gallagher J. B. MacChesney and D. N. E. Buchanan J. Chew. Physics 1965 43 516. P. K. Gallagher J. B. MacChesney and D. N. E. Buchanan J . Chem. Physics 1966,45,2466 J. B. MacChesney J. F. Potter R. C. Sherwood and H. J. Williams J . Chem. Physics 1965 43 1907. P. K. GALLAGHER AND J . B . MACCHESNEY 47 U. Shimony and J. M. Knudsen Physic. Reu. 1966,144,361. R. Bauminger S. G. Cohen A. Marinov S. Ofer and E. Segal Physic. Rev. 1961 122 1447. A. Ito K. Ono and Y. Ishikawa J . Physic. SOC. Japan 1963 18 1465. M. H. Mueller L. Heaton and K. T. Miller Actu Cryst. 1960 13 825. P. K. Gallagher F. Schrey and B. Prescott to be published. ’ K. Ono Y . Ishikawa A. Ito and E. Hirahara J. Physic. SUC. Japan 1962 suppl. Bl 125. l o G. K. Wertheim and R. H. Herber J . Chern. Physics 1963,38,2106. l 2 J. I. Budnick L. J. Bruner R. J. Blume and E. L. Boyd J. Appl. Physics 1961 32 1205. l 3 R. S . Preston S. S. Hanna and T. Heberle Physic. Rev. 1962,128,2207. l4 M. Eibschiitz S. Shtrikman and D. Treves Physic. Rev. 1967 156 562. l5 G. H. Jonker and J. H. Von Santen Physica 1950,16,337. l6 F. J. Morin Physic. Rev. Letters 1959 3 34. l7 D. Adler and H. Brooks Physic. Reu. 1967,155 826.