Hyperfine interaction in K2Ba[Fe(NO2)6]

Spectrochimica Acta Part A 56 (2000) 905 – 913 www.elsevier.nl/locate/saa

Hyperfine interaction in K2Ba[Fe(NO2)6] K. Padmakumar, P.T. Manoharan * Department of Chemistry and Regional Sophisticated Instrumentation Centre, Indian Institute of Technology, Chennai 600 036, India Received 24 June 1999; accepted 21 July 1999

Abstract Magnetic hyperfine splitting observed in the low temperature Mo¨ssbauer spectrum of potassium barium hexanitro ferrate(II), in the absence of any external field, is attributed to the 5T2g state of the central metal atom further split into a ground 5Eg state and a first excited 5B2g state under a distorted octahedral symmetry in contrast to the earlier prediction of 1A1g ground state on the basis of room temperature Mo¨ssbauer spectral and other properties. The central iron atom is co-ordianted to six nitrito groups (NO− 2 ), having an oxidation state of +2. The temperature dependence of Mo¨ssbauer spectra is explained on the basis of electronic relaxation among the spin-orbit coupled levels of the 5Eg ground state. Various kinds of electronic relaxation mechanisms have been compared to explain the proposed mechanism. The observed temperature dependent spectra with varying internal magnetic field and line width can be explained by simple spin lattice relaxation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Hyperfine interaction; Mo¨ssbauer spectroscopy; Spin lattice relaxation

1. Introduction The time dependent magnetic hyperfine interaction produced as a result of slow spin relaxation has been frequently observed and investigated using many techniques including nuclear magnetic resonance (NMR), electron paramagnetic resonance (EPR), and perturbed angular correlation spectroscopies. Mo¨ssbauer spectroscopy has been the most fruitful experimental approach, espe-

* Corresponding author. Tel.: +91-44-4458279; fax: + 9144-2352545/2350509. E-mail address: [email protected] (P.T. Manoharan)

cially for investigating Fe2 + systems because of its inherently rich spectral information. Dipositive iron in its low spin state is described by a nondegenerate orbital and spin ground states when the symmetry at the iron site is close to octahedral in character. This will not give any net unpaired electron density ŽSz at the site of the nucleus, resulting in a single narrow Mo¨ssbauer absorption spectrum. If at all there is a distortion this may lead to the quadrupole splitting of the spectra at low temperatures [1]. As the temperature is increased, the low lying excited states will be populated and this leads to the faster electronic relaxation and broadening of the spectral pattern. Various relaxation mechanisms have been used to explain the observed spectral pattern. However,

1386-1425/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 6 - 1 4 2 5 ( 9 9 ) 0 0 1 9 3 - 6

906

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

the observation of a paramagnetic hyperfine splitting in 57Fe Mo¨ssbauer spectrum of a high spin Fe2 + is rather unusual because often rapid spin lattice relaxation will not allow it to be observed. Ingalls [2] has estimated the relaxation time for Fe2 + in an approximately octahedral environment as about 10 − 9 to 10 − 11 s, much shorter than the nuclear larmor precession time. Thus, if a ferrous complex is paramagnetic down to very low temperatures, its Mo¨ssbauer spectrum is expected to remain as a sharp doublet due to a large quadrupole splitting, and no magnetic hyperfine structure is generally observed in the absence of a strong external magnetic field. For this same reason, the observation of the EPR signal is not possible even down to 4.2 K. The measurement of EPR of such systems is possible only at very low temperatures (millikelvin region) due to the fact that the 5D ground state of Fe2 + is subjected to spin-orbit coupling leading to the orbit lattice interaction and consequent very fast electronic relaxation [3]. However, there are a few Fe2 + systems with a quintet spin ground state where Mo¨ssbauer magnetic hyperfine split spectra have been observed at low temperatures due to slow electronic relaxation even in the absence of an external magnetic field. For example, KFeF3 [4] where the iron in the dipositive state is surrounded by six fluoride ions arranged octahedrally reveals only a single Mo¨ssbauer peak at room temperature with no quadrupole coupling constant or a magnetic hyperfine structure. The magnetic hyperfine structure starts appearing below 77 K where a magnetic phase transition has been observed in the vicinity of 41 K and this is attributed to the structural change from a cubic phase to trigonal phase. The do orbital ground state under trigonal distortion gives an orbital doublet ground state which gives the net magnetic field. The single crystal Mo¨ssbauer measurement has been used to determine the spin state of the system [5]. Similarly the hyperfine split Mo¨ssbauer spectrum observed in KFeCl3 [6,7] in the temperature range of 1.5–30 K in the absence of external magnetic field has been reported to be due to the slow spin-spin relaxation between the three lowest fine structure levels of Fe2 + . Evidence for significant covalency

in KFeCl3 is manifested by the low value of the magnetic hyperfine field (Hc) compared to the free ion as pointed out by several authors [8–10]. In iron doped in ZnCO3 [11], Fe2 + ions possessing a rhombohedral symmetry exhibit magnetic hyperfine splitting at low temperatures in the absence of external magnetic field. The spectra have been fitted using a simple stochastic model and the rate of relaxation of the Fe2 + between the various spin states of the ground orbital doublet at different temperatures were determined. The authors predicted an indirect spin-lattice relaxation mechanism responsible for the temperature dependence of spectral patterns. In a recently reported mixed valent polyiron oxo compound [12,13] the magnetic hyperfine pattern observed is attributed to superparamagnetic relaxation. The Fe3 + component has a longer relaxation and has completely crossed into the slow relaxation regime at 2 K whereas the Fe2 + component which has a short relaxation time is in the intermediate relaxation regime. Thus far, these are probably the few Fe2 + containing systems for which the magnetic hyperfine pattern has been observed in the absence of an external magnetic field. Generally the high spin Fe2 + systems do have a high quadrupole splitting which is also true for KFeCl3 and Fe2 + in ZnCO3. In this context it is necessary to pay attention to a few systems in which a low value for the quadrupole coupling constant (DEQ) has been observed. The low value for DEQ is due to the reduction in spin-orbit coupling for the Fe2 + ions and Jahn-Teller distortions [8,14–16]. Reported values of DEQ for Fe2 + in KMgF3 and in CaO were found to be low in comparison with most other normal Fe2 + systems which are attributed to the reduced spin orbit coupling constant values as a result of high covalency. But the hyperfine splitting in these cases have been observed only in the presence of an applied field. However, we report here the results of a system wherein we observe a low quadrupole splitting especially at high temperature and a Mo¨ssbauer magnetic hyperfine splitting at temperatures from 45 K and down, in the absence of an external magnetic field. The compounds are K2M[Fe(NO2)6] where M=Ca2 + , Ba2 + , Sr2 + , Pb2 + , Zn2 + which have been earlier reported to

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

907

have a singlet ground state (1A1g) [17 – 19]. This present report is restricted to our studies on K2Ba[Fe(NO2)6]. It must be mentioned here that the compound is formed as a powder making it impossible to carry out the single crystal X-ray structure determination. Furthermore, the earlier conclusion of a singlet spin state was derived purely on the basis of room temperature magnetic susceptibility [18] (which still showed a net magnetic moment of 0.916 BM), optical [18,19] and Mo¨ssbauer spectral studies [17]. This led us to carry out a careful low temperature investigation using both magnetic susceptibility and Mo¨ssbauer spectral techniques. It should be remarked here that the compound is quite unstable in presence of water though it is prepared in aqueous medium but by quick filtration.

used as the exchange gas for the sample chamber. This enables the conduction of the heat from the Helium port to the sample chamber. Mo¨ssbauer spectra were recorded in the temperature range from 300 to 4.2 K in the absence of an external magnetic field. All spectral patterns were fitted with a computer program using Lorentzian line shape unless otherwise specified. The isomer shift parameter is reported with respect to the natural Fe foil. Magnetic susceptibility measurements were carried out on a powder sample at a field of 1.4 T using a SQUID in the temperature of 4.2–300 K. The error in the temperature was 9 0.2 K and the susceptibility was within 1%. The magnetic susceptibility was corrected for underlying diamagnetism using Pascal constants.

2. Experimental

3. Results and discussion

2.1. Preparation of the complex

The Mo¨ssbauer spectra were recorded at different temperatures in the region of 4.2–300 K. The values of the isomer shift (d), quadrupole coupling constant (DEQ) and the internal magnetic field (Heff) are given in Table 1. Room temperature spectrum showing a quadrupolar doublet as shown in Fig. 1 (top) is in complete agreement with the earlier report [17]. The shape of the spectrum remains the same except for a small change in the isomer shift down to 45 K on cooling. But as the temperature is further decreased the Mo¨ssbauer spectrum becomes broadened showing additional lines on the wings of the doublet followed by resolution into a magnetic hyperfine sextet at 30 K with considerable broadening. At 15 K though it clearly resolves the integrated intensities of the six lines are not in the ratio of 3:2:1:1:2:3 expected for a normal sextet. At 4.2 K, however, the integrated intensities of the hyperfine components are in the expected simple ratio but the amplitudes are not in the simple ratio due to differences in line width for each component. No EPR signal from this sample could be monitored down to 4.2 K indicating clearly the absence of any ferric ion of low or high spin. In addition the integrity of the sample due to a single

The compound was prepared according to the previously reported procedure [18]. The purity of the compound was checked by the methods of elemental analysis and the IR spectra which concur with the earlier reported results.

2.2. Physical measurements The Mo¨ssbauer spectra were recorded in Canberra S-100 Mo¨ssbauer spectrometer fitted with a Wissel constant acceleration drive. The source is Co/(Pd) (Amersham, 50 mCi) and the absorber was kept stationary in a helium dewar. The source, the absorber holder and the detector were carefully centred and aligned under operating conditions to ensure reliable and reproducible geometry. All the velocity calibrations are done using the natural Fe foil of 25 m thickness. The velocity range is from − 10 to +10 mm s − 1. Variable temperature measurements were done by using a Oxford MD-306 Bath Cryostat which is of gas exchange type. The sample temperature was monitored by using a calibrated Rh-Fe sensor which is connected to the ITC-4 temperature controller. This has an accuracy of 0.1 K. Helium is

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

908

type of species is evident from the absence of multiple spectral characteristics say due to the simultaneous presence of Fe(II) and Fe(III) ions in differing amounts.

3.1. Quadrupole coupling constant(DEQ) Since the compound does not crystallize to allow a single crystal X-ray structure determination earlier structural studies were restricted to powder X-ray diffraction [20] which revealed that the unit cell of these compounds with the general formula K2M[Fe(NO2)6] where M =Ca, Ba, Sr, Pb, Zn contains octahedrally disposed Fe(NO2)46 − ions. Based on the powder XRD data alone with the susceptibility and optical spectral properties [17–19], Manoharan et al. [17] interpreted the small quadrupole coupling constant as due to a small lattice distortion from a t62g configuration of Th point group within the 1A1g ground state. Furthermore, it was assumed that a hindered rotation of all the nitro groups around the iron could lower the symmetry to S6, D2 or T and also a lengthening of the Fe – N distance along one axis could also lower the symmetry. These were suggested to have been responsible for the observation of the quadrupole splitting at room temperature. The present observation of hyperfine splitting at low temperatures cannot, however, be accommodated with a singlet spin state. Hence the spin state ought to be a quintet as supported by recent measurement of temperature dependent suscepti-

bility (vide supra). Hence the ground electronic state of (3d6, 5D) of Fe2 + ion in cubic field is an orbital triplet 5T2g. Furthermore, the octahedral or Th symmetry of Fe2 + ions with high spin multiplicity are expected to undergo Jahn-Teller distortion to reduce the high symmetry environment. Such Jahn-Teller distortion can cause a reduction of the effective spin- orbit coupling for Fe2 + ions in such cubic environment [8,14]. So a strong non cubic component of the crystal field along with or as part of J-T effect may completely lift the orbital degeneracy. Such a distortion may result in a trigonal field, which in turn will split the T2g state expected of a Th environment in an idealized cubic environment into an orbital doublet and a singlet, the wave functions for which can be expressed as follows [21]: 1/2 + 1 2 c1 = (2/3)1/2Y − Y2 2 + (1/3)

(1)

c0 = Y 02

(2)

1 c − 1 = (2/3)1/2Y 22 − (1/3)1/2Y − 2

(3)

where the orbital singlet is c0 and the doublet is due to c1 and c − 1. The doublet each will contribute (2/7)er − 3and the singlet contribute − (4/7)er − 3 to the quadrupole coupling constant. If there is an equal contribution form both the singlet and the doublet the DEQ will be zero. At ambient temperature spin-orbit coupling mixes the doublet and singlet giving a net small electric field gradient. The energy separation between the orbital doublet and the singlet states is expected to be higher than the spin-orbit coupling energy.

Table 1 Mo¨ssbauer parameters of K2Ba[Fe(NO2)6] as function of temperature Temperature (K)

Isomer shift (d)a (mm s−1)

DEQ (mm s−1)

Heff (Kilo Gauss)

300.0 211.0 151.0 102.0 77.0 45.0 30.0 15.0 4.2

0.363 9 0.02 0.4239 0.02 0.4339 0.02 0.4739 0.02 0.5039 0.02 0.4539 0.02 0.4209 0.02 0.3809 0.02 0.3469 0.02

0.74 90.02 0.72 9 0.02 0.74 90.02 0.77 90.02 0.78 90.02 0.70 9 0.02 0.30 9 0.02 0.22 9 0.02 0.0

– – – – – – 369.0 406.0 450.0

a

Isomer shift values are reported with respect to natural iron foil.

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

909

symmetrized vibronic spin orbit states. But at very low temperatures the orbital doublet with its spinorbit and vibronic coupling is effectively populated and magnetic hyperfine splitting is produced. Another possible explanation for this may come from its magnetic property. At high temperatures (prior to the appearance of the hyperfine) the rapid motion of the iron electronic moments is probably responsible for the quenching of the magnetic hyperfine structure.

3.2. Magnetic hyperfine interaction As shown in Fig. 1, the spectrum is made up of a quadrupolar doublet until 45 K. At 45 K, a broadening of the doublet along with the appearance of hyperfine split spectra on the wings results. As the temperature is further lowered, the Mo¨ssbauer sextet shows itself but with the lines of differing integrated intensities and line width as shown in Table 2. Only at 4.2 K the integrated intensity is of the order 3:2:1:1:2:3 though the line width differences persist. The line width differences are expected to disappear at lower temperatures. It is necessary to note that a small component of the original quadrupole doublet seems to persist even at 30 K though it totally disappears at 15 and 4.2 K. This looks like a second order magnetic phase transition though it could be due to the combined effect of relaxation and Jahn-Teller effect. In addition, it is of importance to note here that Hint is a function of temperature as shown in Fig. 2. Fig. 1. 57Fe Mo¨ssbauer spectra of K2Ba[Fe(NO2)6] in the absence of an external magnetic field at different temperatures.

The low quadrupole coupling observed here is reminiscent of Fe2 + /KMgF3 and/or Fe2 + /CaO case- with strong Jahn-Teller distortion, low spinorbit coupling, high covalency and consequent low DEQ. One can say that Fe2 + ion in the ground 5T2g state prior to J-T distortion is strongly coupled either to the T2g or Eg modes of vibrations. One has to solve the problem numerically by considering the above said vibronic coupling and the spin orbit interaction to create

Table 2 Line width (mm s−1) (FWHM) values of the six lines at different temperatures for K2Ba[Fe(NO2)6] Lines

FWHM at 4.2 K

1 2 3 4 5 6

1.99 1.33 0.63 0.69 1.39 1.90

15.0 K

30.0 K

2.05 1.78 1.41 1.46 1.85 2.07

2.67 3.43 1.67 1.70 3.39 2.73

910

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

Fig. 2. Plot of Internal magnetic field (Heff) vs. temperature for K2 Ba[Fe(NO2)6].

The spectral pattern observed at 30 K is very similar to that of the Jahn-Teller distorted theoretically simulated Mo¨ssbauer spectrum [22] but has since not been observed in experimental situations. Secondly the Jahn-Teller effect leading to the observation of magnetic hyperfine structure in Fe2 + high spin system is very unusual [23,24]. Main interactions responsible for the slow spin relaxation in the absence of an applied magnetic field is difficult to establish unequivocally from powder samples. Such magnetic hyperfine interaction may originate from either molecular anisotropy or collective magnetic interactions which produces particle like superparamagnetism [25] or it could be from a three dimensional ferro or anti ferromagnetic ordering [26]. Substantial molecular anisotropy has also resulted in slow paramagnetic relaxation in both Kramers [27] and non-Kramers [28] systems. The overall spectral pattern in the low temperature limit could be explained on the basis of two different mechanisms which come under the general category of relaxation such as

(1) superparamagnetic relaxation and (2) spin-lattice relaxation. The superparamagnetic relaxation [12,13,29,30] is dependent on the volume of the particle and temperature, having a transition point called the blocking temperature. Above this temperature the spin system is in the superparamagnetic relaxed state and below this it is in a superparamagnetically ordered state. Below the blocking temperature the six line pattern due to the magnetic hyperfine interaction will not change the position and so the hyperfine field will be temperature independent throughout the slow relaxation regime [31]. The internal magnetic field is temperature dependent in the case of spin-lattice relaxation mechanism. In the fast relaxation limit the nucleus experiences a zero average magnetic field (in the absence of the external magnetic field) and the hyperfine structure is narrowed to a single line or a quadrupole doublet. In the case of the intermediate relaxation rates, theory predicts a compli-

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

cated continuous spectrum which gradually merges into the two extreme cases of six line pattern or a quadruplole split doublet in the slow and fast relaxation limit. But in the intermediate relaxation regime the spectrum becomes distinctly non-Lorentzian. In the case of K2Ba[Fe(NO2)6] we could see a temperature dependent internal magnetic field as shown in Fig. 2. This is also clearly visible in the variable temperature spectral pattern shown in Fig. 1 where the peak positions change on lowering the temperature. This testifies to our suggestion that the relaxation mechanism operating in this case is of spin-lattice in origin and definitely not of superparamagnetic origin. This is an unusual observation for Fe2 + high spin case. As the temperature is decreased the spin fluctuation becomes slow and comparable to the larmor precessional frequency of the nucleus and the magnetic field is felt by the nucleus through core polarization. The internal magnetic field increases in magnitude as the temperature is lowered until it reaches a saturation value. Yet another proof for the contribution of the spin lattice relaxation for the appearance of the hyperfine lines only at lower temperatures comes from the temperature dependent IR spectrum of the corresponding lead compound [32]. The Infrared spectral lines corresponding to nasym, nsym, d of NO2 groups become considerably sharpened specially below 50 K, though the sharpening process is gradual right from 250 K and downwards. The coincidence of considerable sharpening of FT-IR lines due to the NO2 vibrational modes and the appearance of hyperfine lines in the Mo¨ssbauer spectra, may be interpreted as due to the decoupling of the lattice phonon modes from the electronic wave functions at lower temperatures. This increases the T1 leading to the appearance of the hyperfine lines. The spectra at 4.2 K is still broader indicating that the relaxation processes are still dominant. The peaks 1 and 6 are much broader than the natural line width of Fe. The solid lines in all the spectra represent the fitted spectrum for a given magnetic field, DEQ and d. The internal magnetic field has a value of 450 KG at 4.2 K; it may be noted here that the free ion Fe2 + has an internal

911

magnetic field of 550 KG [33]. This mechanism along with the temperature dependence of field can also be interpreted in terms of explanation given for observing DEQ of small magnitude. The total contribution of Hint can be written as [34]: Hint = 2mB r − 3LZ + 1/2HC SZ  − 1/2mB 3(S . r)r − 2 − SZ r − 3+ Hext (4) while the contact term Hc is the main contributor, the others — the first term being of orbital, third being of dipolar field and the last being exchange interaction in origin — are likely to give fields of sign opposite to HC resulting in a low field at high temperature. In view of the fact that the orbital doublet is the ground state and the first excited singlet state gives a zero magnetic field, the first term is likely to give a reduced contribution to the total Hint. Though the Jahn-Teller distortion is definitely more significant than spin–orbit coupling energy the spin-lattice relaxation will play a dominant role in the relaxation processes.

3.3. Magnetic susceptibility The magnetic susceptibility, measured and interpreted, is a definite indication of the quintet spin state of the compound under investigation. The experimental susceptibility fitted with a Bleaney–Bowers expression as shown in Fig. 3 clearly indicates that the the compound is essentially a paramagnet with a small antiferromagnetic exchange coupling of the order of −14 cm − 1 between the neighbours. This small antiferromagnetic coupling may give a contribution to the Hint but it will change with temperature contributing more (in magnitude) as temperature is lowered towards TN. All these mechanisms will produce a temperature dependent Hint.

3.4. Line width Temperature dependent line width values for all Mo¨ssbauer transitions are given in Table 2. For example the line width values for 15 K spectrum are of the order of 10, 9 and 7 times the natural line width for the outermost, middle, and innermost absorption lines. An analysis of the 30 K

912

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

spectrum reveals its totally non-Lorentzian character and this may be due to the presence of both the doublet and the singlet orbital states being populated with the ground level being populated more. This would distort the ground state and dominate the relaxation process. This also is an indication of the low energy gap (DE) between the J-T distorted states.

observed values are much lower than those for most other Fe2 + high spin systems. Such values, though rarer, have been observed in a few compounds [8,14–16]. This is also an additional evidence for higher covalency.

3.5. Isomer shift

The compound K2Ba[Fe(NO2)6], is proven to have a 5T2g ground state (by Mo¨ssbauer spectroscopy and susceptibility) as against the earlier prediction of 1A1g state. This 5T2g state gets split into an orbital doublet and singlet through JahnTeller distortion. The final states are, however, made up of vibronic spin-orbit states and hence the rationale behind the observation of magnetic hyperfine interaction through relaxation mechanism explainable by Blume–Tjon theory. It is

The isomer shift values of this compound need a special comment. They are temperature dependent with no second order Doppler effect. The temperature dependence of isomer shift reflects the presence of Jahn-Teller distortion in this compound [1]. For iron atom in the +2 oxidation state with net spin S = 2 the lower limit of the isomer shift cannot be exactly predicted [35]. Our

4. Conclusion

Fig. 3. Plot of magnetic susceptibiliy vs. temperature for K2 Ba[Fe(NO2)6].

K. Padmakumar, P.T. Manoharan / Spectrochimica Acta Part A 56 (2000) 905–913

very much possible that the earlier suggestion [18] of the structure of this compound as composed of two pairs of opposite NO2 planes being perpendicular to each other along the three direction may need a revision in order to reduce the ligand field at the iron site. This may be possible by disposing all the NO2 groups in such a fashion as to reduce the p-overlap between the ligand and metal orbitals-in other words in a distorted fashion, may be with an S6 symmetry.

Acknowledgements PTM is thankful to the DST for a project (SP/S1/47/90) and to the CSIR for an Emeritus Scientist Scheme. KPK gratefully acknowledges a fellowship from IIT/Madras.

References [1] M. Tanaka, T. Tokoro, Y. Aiyama, J. Phys. Soc. Jpn. 21 (1966) 262. [2] R. Ingalls, Phys. Rev. A 133 (1964) 787. [3] P. Gu¨tlich, R. Link, A. Trautwin, Mossbauer spectroscopy and Transition Metal Chemistry, Springer-Verlag, Berlin, Heidelberg, 1978, p. 89. [4] A. Ito, S. Morimoto, J. Phys. Soc. Jpn. 39 (1975) 884. [5] G.R. Davidson, M. Eibschutz, H.J. Guggenhein, Phys. Rev. B 8 (1973) 1864. [6] V. Petrouleos, A. Simopoulos, A. Kostikas, Phys. Rev. B 12 (1975) 4675. [7] V. Petrouleos, A. Simopoulos, A. Kostikas, Phys. Rev. B 12 (1975) 4666. [8] F.S. Han, W.M. Schwarz, M.C.M. O’Brien, Phys. Rev. 185 (1969) 548. [9] D.P. Jhonson, R. Ingalls, Phys. Rev. B 1 (1970) 1013.

.

913

[10] Y. Hazony, Phys. Rev. B 3 (1971) 711. [11] D.C. Price, C.E. Jonson, I. Maartense, J. Phys. C Solid State Phys. 10 (1977) 4843. [12] K.L. Taft, G.C. Papaefthymiou, S.J. Lippard, Science 259 (1993) 1302. [13] K.L. Taft, G.C. Papaefthymiou, S.J. Lippard, Inorg. Chem. 33 (1994) 1510. [14] R.B. Frankel, J. Chappert, J.R. Regnard, A. Misetich, C.R. Abeledo, Phys. Rev. B 7 (1972) 2469. [15] T. Ray, J.R. Regnard, Phys. Rev. B 14 (1976) 1796. [16] J.R. Regnard, T. Ray, Phys. Rev. B 14 (1976) 1805. [17] P.T. Manoharan, S.S. Kaliraman, V.G. Jadhao, R.M. Singru, Chem. Phys. Let. 13 (1972) 585. [18] H. Elliott, B.J. Hathaway, R.C. Slade, Inorg. Chem. 5 (1966) 669. [19] K.G. Caulton, R.I. Fenski, Inorg. Chem. 6 (1967) 562. [20] A. Ferrari, L. Cavalca, M. Nardelli, Gazz. Chim. Iral. 81 (1951) 982. [21] B. Bleaney, K.W.H. Stevens, Rept. Progr. Phys. 16 (1953) 130. [22] I.B. Bersuker, S.A. Borshch, I.Y. Ogurstsov, Phys. Stat. Sol. (b) 59 (1973) 707. [23] H.R. Leider, D.N. Pipkorn, Phys. Rev. 165 (1968) 494. [24] F.S. Ham, J. Phys. Colloq. 35 (1974) C6 – 121. [25] A.H. Morrish, The Physical Principles Of Magnetism, John Wiley and Sons, NewYork, 1965. [26] W.M. Reiff, S. Foner, J. Am. Chem. Soc. 95 (1973) 260. [27] K.K. Surerus, E. Mu¨nck, B.S. Snyder, R.H. Holm, J. Am. Chem. Soc. 111 (1989) 5501. [28] K.K. Surerus, M.P. Hendrich, P.D. Christie, D. Rottgondt, W.H. Orme-Johnson, E. Mu¨nck, J. Am. Chem. Soc. 114 (1992) 8572. [29] N.M.K. Reid, D.P.E. Dickson, D.H. Jones, Hyp. Int. 56 (1990) 1487. [30] G.C. Papaefthymiou, Phys. Rev. B 46 (16) (1992) 10366. [31] H.T.H. Le Fever, F.J. Van Steenwijk, R.C. Thiel, Physica. 86 – 88B (1977) 1269. [32] K. Padmakumar, P.T. Manoharan (to be published). [33] R.E. Watson, A.J. Freeman, Phys. Rev. 123 (1961) 2027. [34] O.H. Ok, Phys. Rev. 185 (1969) 472. [35] N.N. Greenwood, T.C. Gibb, Mo¨ssbauer spectroscopy, Chapman and Hall, London, 1971, p. 91.