Electron paramagnetic resonance and optical absorption studies of Cu2+ ions in alkali barium borate glasses

Electron paramagnetic resonance and optical absorption studies of Cu2+ ions in alkali barium borate glasses

Journal of Alloys and Compounds 265 (1998) 29–37 L Electron paramagnetic resonance and optical absorption studies of Cu 21 ions in alkali barium bor...

295KB Sizes 1 Downloads 40 Views

Journal of Alloys and Compounds 265 (1998) 29–37

L

Electron paramagnetic resonance and optical absorption studies of Cu 21 ions in alkali barium borate glasses R.P. Sreekanth Chakradhar, A. Murali, J. Lakshmana Rao* Department of Physics, S.V. University, Tirupati-517 502 (A.P.), India Received 28 May 1997; received in revised form 29 July 1997

Abstract Electron Paramagnetic Resonance (EPR) and optical absorption studies have been made for Cu 21 ions in alkali barium borate glasses. The EPR spectrum consists of four hyperfine lines characteristic of Cu 21 ions. From the observed spectra the spin-Hamiltonian parameters have been evaluated. The values of spin-Hamiltonian parameters indicate that the Cu 21 ions in alkali barium borate glasses were present in octahedral coordination with a strong tetragonal distortion. The linewidth of the parallel and perpendicular hyperfine peaks increases in order with magnetic quantum number of copper nuclei and this has been attributed to the microenvironmental fluctuations around Cu 21 ion which is intrinsic to the glassy state. The optical absorption spectra of Cu 21 ions doped glasses show a single broad absorption band which is attributed to 2 B 1g → 2 B 2g transition. By correlating the EPR and optical data, the molecular orbital coefficients have been evaluated. The theoretical values of optical basicity (Lth ) of these glasses have also been evaluated.  1998 Elsevier Science S.A. Keywords: Electron paramagnetic resonance; Optical absorption spectrum; Copper ions; Glasses; Optical basicity

1. Introduction In recent years glasses doped with transition metal ions have attracted a great deal of attention, because of their memory and photoconducting properties. They also find potential applications in the solid state lasers, luminescent solar energy concentrators (LSCs) and fibre optic communication devices [1]. B 2 O 3 is one of the best known glass formers and is present in almost all glasses. It is often used as a dielectric and insulating material and it is known that borate glass is a good shield against IR radiation. They are also of academic interest because of the occurrence of the boron anomaly [2]. EPR investigations of Cu 21 ions in glasses are interesting and have received a considerable attention because the EPR parameters are very sensitive to the local symmetry [3–9]. No EPR studies of Cu 21 ions in alkali barium borate glasses have been reported so far. We are interested to know the site symmetry around Cu 21 ions in these glasses. Hence we studied the EPR and optical absorption spectra of Cu 21 ions in these glasses. The present work *Corresponding author. 0925-8388 / 98 / $19.00  1998 Elsevier Science S.A. All rights reserved. PII S0925-8388( 97 )00437-4

describes the EPR and optical absorption spectra of Cu 21 ions in alkali barium borate glasses.

2. Experimental The starting materials used for the preparation of glasses were analar grade R 2 CO 3 (R5Li, Na or K), BaCO 3 , H 3 BO 3 and CuCO 3 . The composition of glasses used in the present study were 20 Na 2 CO 3 1 (252x) BaCO 3 155 H 3 BO 3 1x CuCO 3 (0.1#x#5 mol %) and 20 R 2 CO 3 1 24.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3 (R5Li, Na or K). Table 1 gives the batch composition of glasses studied in the present work. The chemicals were weighed and ground to a fine powder and mixed thoroughly. The batches were melted in an electric furnace at 1223 K by placing them in porcelain crucibles. The melt was then quenched at room temperature in air by pouring it on a polished brass plate and pressing it with another brass plate. The glasses thus obtained were transparent and bluish in colour. The EPR spectra were recorded at room temperature on a EPR spectrometer (JEOL-FE 1X) operating in the Xband frequencies with a modulation frequency of 100 kHz. The microwave frequency was kept at 9.205 GHz. The

30

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

Table 1 Composition of glasses (mol %) studied in the present work System

Composition

(a) Lithium Barium Borate Glass (Li Ba B) (b) Sodium Barium Borate Glass (Na Ba B) (c) Potassium Barium Borate Glass (K Ba B)

20 Li 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3

magnetic field was scanned from 220 to 420 mT. The microwave power was set at 20 mW. A powdered glass specimen of 100 mg was taken in a quartz tube for EPR measurements. Polycrystalline DPPH was used as a standard field marker. EPR spectra of 20 Na 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 1 0.5 CuCO 3 glass sample was recorded at different temperatures (123–433 K) using a variable temperature controller (JES-UCI-2AX). A temperature stability of 61 K was easily obtained by waiting approximately half-an-hour before recording a spectrum at each temperature. Optical absorption spectrum of 20 R 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3 (R5Li, Na or K) glass samples were recorded at 300 K on a Perkin–Elmer

20 Na 2 CO 3 1(252x) BaCO 3 155 H 3 BO 3 1x CuCO 3 (0.1#x#5 mol %) 20 K 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3

spectrophotometer (model 551) in the wavelength region 720–850 nm.

3. Results and analysis No EPR signal was observed in the spectra of undoped glasses indicating that no paramagnetic impurities were present in the starting materials. When Cu 21 ions are introduced into the alkali barium borate glasses a–c (see Table 1) all the investigated samples exhibit resonance peaks. The EPR spectra were obtained even at very low concentrations (x#0.5 mol % of CuCO 3 ) of paramagnetic ions. Fig. 1 shows the EPR spectra of 0.5 mol % of Cu 21

Fig. 1. EPR spectra of 0.5 mol % Cu 21 ions doped in (a) lithium barium borate glass (Li Ba B) (b) sodium barium borate glass (Na Ba B) (c) potassium barium borate glass (K Ba B) at room temperature. [Microwave frequency59.205 GHz; Microwave power520 mW; Field modulating frequency5100 kHz; Field modulation50.1 mT; Field sweep rate525 mT min 21 ; Recorder time constant50.3 sec. The gains are different for different glasses.]

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

ions in different alkali barium borate glasses at 300 K. The EPR spectra observed in the present work are similar to those reported earlier for Cu 21 ions in many other glass systems [10–12]. The EPR spectra of sodium barium borate glasses (Na Ba B) with different mol % of Cu 21 ions have been studied and the spectra are shown in Fig. 2. The number of spins participated in the resonance can be calculated using the formula [13]. N ¯ Ipp (DHpp )2 This can be used to calculate the number of Cu 21 ions participated in the resonance in sodium barium borate glass samples by taking Ipp and DHpp as the peak to peak height and peak to peak width of the resonance line respectively. Ipp were measured from the height of the first and fourth perpendicular hyperfine lines. DHpp were measured from the maximum and minimum hyperfine positions at the lowest and highest fields of the first derivative curve of the perpendicular components. The number of spins are thus calculated for various mol% of Cu 21 ions doped in sodium barium borate glass samples and is shown in Fig. 3. The variation in linewidth DH(m) of parallel and perpendicular hyperfine peaks as a function of magnetic quantum number (m) of copper nuclei is shown in Fig. 4. It is found

31

that the linewidth of parallel and the perpendicular hyperfine peak increased in order of m. A similar observation for hyperfine parallel components was reported by Hosono et al. [10] for Cu 21 ions in silicate glasses. The EPR spectra at different temperatures (123–433 K) were also recorded for Na Ba B glass to study the temperature dependence of the intensity of the resonance line at g' 52.049 i.e., Ig' 52.049. The EPR spectra recorded at different temperatures are shown in Fig. 5. Fig. 6 shows the dependence of intensity on the temperature for a derivative spectrum. The measurement of this is associated with the area under the absorption curve (i.e.,) intensity is proportional to the product of peak-topeak height and (peak-to-peak width)2 . From Fig. 6 it is clear that the intensity of the resonance signal at g' 5 2.049 decreases with increasing temperature according to 1 /T law. This is in accord with the expectation (Ia 1 /T ) for paramagnetic centres. The optical absorption spectra of Cu 21 ions in 20 R 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3 (R5Li, Na and K) (see Table 1) glasses at 300 K are shown in Fig. 7. It is seen that there exist only one broad absorption band between 720–850 nm. This band has been assigned to 2 B 1g → 2 B 2g transition of Cu 21 ions in a distorted octahedral site. A band due to 2 B 1g → 2 E g transition at higher

Fig. 2. EPR spectra of different mol % of Cu 21 ions doped in sodium barium borate glass sample at room temperature. [Microwave frequency59.205 GHz; Microwave power520 mW; Field modulating frequency5100 kHz; Field modulation50.1 mT; Field sweep rate525 mT min 21 ; Recorder time constant50.3 sec and Gain5320.]

32

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

Fig. 3. Variation of number of spins with mol % of Cu 21 ions in sodium barium borate glass (the line is drawn as a guide for the eye).

energy level side could not be observed. The observed optical absorption bands obtained in the present work are in good agreement with those reported by earlier workers [8,11]. Fig. 8 shows the dependence of A uu and guu as a function of temperature for 0.5 mol % of Cu 21 ions in sodium barium borate glasses.

4. Discussion The Cu 21 ion (S51 / 2) has a nuclear spin I53 / 2 for both 63 Cu (the natural abundance 69%) and 65 Cu (natural abundance 31%) and therefore (2I 11) i.e., four perpendicular and four parallel hyperfine components could be observed. In the recorded spectra for the glass (a) three weak parallel components are observed in the lower field region and the fourth parallel component is overlapped with the perpendicular component; the perpendicular components in the higher field region however are not well resolved. For the glasses (b) and (c), three weak parallel components are observed in the lower field region but the perpendicular components in the higher field region are well resolved. For Cu 21 ions in a crystalline environment a regular octahedral site may not exist, because the cubic symmetry is disturbed by electronic hole in the degenerate d x 2 2y 2 orbital and this produces the tetragonal distortion. The EPR spectra of Cu 21 ions in glasses could be best analysed by using an axial spin-Hamiltonian of the form:

* 5 b guu Hz Sz 1 b g' (Hx Sx 1 Hy Sy ) 1 A uu Sz Iz 1 A ' (Sx Ix 1 Sy Iy )

Fig. 4. Variation of linewidth DH(m) of parallel and perpendicular hyperfine peaks as a function of magnetic quantum number (m I ) of Cu 21 nuclei for sodium barium borate glass at room temperature.

(1)

where z has been taken as the symmetry axis of individual Cu 21 complex. The symbols in Eq. (1) have their usual meaning and the nuclear quadrupole and nuclear Zeeman interaction terms are ignored [14]. Two sets of hyperfine lines each consisting of four peaks are designated as parallel and perpendicular hyperfine

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

33

Fig. 5. EPR spectra of 0.5 mol % of Cu 21 ions doped in sodium barium borate glass sample at different temperatures. [Microwave frequency59.205 GHz; Microwave power520 mW; Field modulating frequency5100 kHz; Field modulation50.1 mT; Field sweep rate525 mT min 21 ; Recorder time constant50.1 sec and Gain5400.]

peaks. The solution of the spin-Hamiltonian gives the expressions for the peak positions related to the principal g and A tensors as [15]: A 2' ]] hn 5 guu b H 1 mA uu 1 (15 / 4 2 m ) 2guu b H

(2a)

A 2uu 1 A 2' ]]] hn 5 g' b H 1 mA ' 1 (15 / 4 2 m ) 4g' b H

(2b)

2

2

for the parallel and perpendicular hyperfine peaks respectively. Here m is the nuclear magnetic quantum number of the copper nucleus with values 3 / 2, 1 / 2, 21 / 2 and 23 / 2 and n is the microwave frequency at resonance. Using 2a and 2b the spin-Hamiltonian parameters have been evaluated and are presented in Table 2. From Table 2, it is observed that the g values are monotonically decreasing from lithium barium borate glass to potassium barium borate glass. A change in g values is attributed to the structural change in the environment of Cu 21 ions. The observed guu and g' values are characteristic of Cu 21 ions co-ordinated by six ligands which form an octahedron elongated along the z-axis [3,11]. As guu .g' .

ge we consider that the ground state for paramagnetic 2 21 electrons is d x 2 2y 2 orbital ( B 1g state); the Cu ion being located in distorted octahedral sites (D 4h ) elongated along the z-axis [3]. It is clear from Table 2, that the guu values decreases from Li to K glasses whereas much difference is not observed in g' , A uu and A ' values. From Fig. 3, it is clear that as the concentration of Cu 21 ion is increased, the number of spins participating is increased monotonically and no cluster formation is observed till 5 mol % of Cu 21 ions. The number of spins for 5 mol % of Cu 21 ions is not shown in Fig. 3. The linewidth of parallel and perpendicular hyperfine lines increases in order of m. This broadening may be attributed to the microenvironmental fluctuations around Cu 21 ion which is intrinsic to the glassy state. The structural distribution in glass causes fluctuation in ligand field and in turn, it is reflected in the distribution of the spin-Hamiltonian parameters. Hosono et al. [10,11] calculated the fluctuations of guu and A uu . In a similar way the authors calculated the fluctuations of guu and A uu (dguu and dA uu ) which gives rise to that of the resonating field (dH ) which is derived from the Eq. (2a):

34

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

1 dH 5 2 ]] guu b

F

S

dguu dA uu dguu hn ] 1 mA uu ]] 2 ] guu A uu guu

DG

The peak width DH (m) is related to the variation of guu (i.e.,) dguu . hn 1 m( guu P 2 A uu ) DH(m) 5 dguu ¯ dguu ]]]]]] g uu2 b ¯ dguu (1285.2 1 283 m) (gauss)

Fig. 6. Intensity dependence on inverse temperature of the resonance line at g' 52.049 for 0.5 mol % of Cu 21 ions in sodium barium borate glass sample.

Here n 59.205 GHz and P (50.036 cm 21 ) is a constant. The authors also calculated the fluctuations of g' ie., (dg' ). The variation of linewidth DH (m) of parallel and perpendicular hyperfine peaks with magnetic quantum number of copper nuclei is shown in Fig. 4. The variation of guu and g' ie., (dguu and dg' ) was determined from the slopes of the lines in Fig. 4. The dguu and dg' values thus obtained are 0.13 and 0.07 respectively. The magnitude of dg uu obtained in the present work is higher compared to the value reported by Hosono et al. [10]. This has been attributed to the decrease in the strength of the B–O bonds resulting in the increase of covalency of Cu(II)–O bonds. The dependence of guu and A uu value with temperature is shown in Fig. 8. It is observed that guu increases with temperature whereas A uu decreases with temperature. No significant change is observed with temperature for g' and A ' values. The guu and A uu values are sensitive to both temperature and alkali content. On the other hand g' and A ' values are almost independent of alkali content and temperature. When temperature changes we can expect a change in ligand field, due to surrounding ligands which reflects a change in g and A values. As there is no change in g' and A ' values and there are changes in guu and A uu values, it is understood that the ligands that are situated in z-axis are vibrating more effectively with temperature causing a change in the internuclear distances with the central metal ion which results in a change in guu and A uu values. Duffy and Ingram [16] proposed that the ideal values of optical basicity can be predicted from the composition of glasses and basicity moderating parameters of various cations present. The theoretical value of the optical basicity of glasses can be estimated using the formula: Zr O] 2g i i

Lth 5

i

i

where Zi is the total oxidation number of the cation i, r i is the molar ratio of the cation i to the total number of oxides and gi is basicity moderating parameter. gi for cation is

gi 5 1.36(x i 2 0.26)

Fig. 7. Optical absorption spectra of 20 R 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3 [R5(a) Li (b) Na and (c) K] glasses at room temperature.

where x i is Pauling electronegativity [17] of the cation. The theoretical value of optical basicity thus calculated are also given in the Table 2. It is observed that the theoretical values of Lth increases from lithium to potassium.

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

35

Fig. 8. The dependence of guu and A uu as a function of temperature for 0.5 mol % Cu 21 ions in sodium barium borate glass. (The line is drawn as a guide for the eye.)

The bonding coefficient a 2 (i.e., the in-plane s bonding) can be calculated from the EPR data using the expression given by Kuska et al. [18]. 7 a2 5] 4

F

A uu A 2 5 6 ] 2 ] 1 ] guu 2 ] g' 2 ] P P 3 21 7

G

where P50.036 cm 21 and A5(1 / 3 A uu 12 / 3 A ' ). The a 2 value calculated for alkali barium borate glasses are shown in Table 3. The EPR and optical absorption spectral data can be correlated to evaluate the bonding coefficients as follows [19]: guu 5 2.0023 g' 5 2.0023

F F

G G

4la 2 b 21 1 2 ]]] E1

(3)

la 2 b 2 1 2 ]] E2

(4)

Table 2 Spin-Hamiltonian parameters and optical basicity for 0.5 mol % of Cu 21 ions in different alkali barium borate glasses Glass

guu

g'

A uu 310 24 cm 21

A ' 310 24 cm 21

Lth optical basicity

Li Ba B Na Ba B K Ba B

2.284 2.262 2.259

2.053 2.049 2.048

131 137 140

25 24 24

0.4860 0.4960 0.5107

The errors in g and A values are 60.002 and 65310 ly.

24

cm

21

respective-

E1 and E2 are the energies corresponding to the transitions B 1g → 2 B 2g and 2 B 1g → 2 E g respectively, and l is the spin orbit coupling constant (5 2828 cm 21 ) [20]. The parameters a 2 and b 12 represent the contribution of 3d atomic orbitals of the cupric ion to the B 1g and B 2g anti bonding orbitals respectively. The bonding coefficients a 2 , b 21 and b 2 (51.00) characterize respectively, the in-plane s bonding, in-plane p bonding and out of plane p bonding of the Cu 21 complex in the glasses. Their value lie between 0.5 and 1.0, the limits of pure covalent and pure ionic bondings [21]. The value of b 2 may be expected to lie sufficiently close to unity so as to be indistinguishable from unity in the bonding coefficient calculations [3]. The expression a 2 given in Eq. (3) is the bonding coefficient due to the covalency of the s bonds with the equatorial ligands which measures the electron density delocalized on 2 the ligand ions and b 1 accounts for covalency of p-anti 2 bonding between ligands and the excited B 2g state. From the Eqs. (3) and (4) it can be seen that to

2

Table 3 Observed band positions and molecular orbital coefficient parameters for 0.5 mol % Cu 21 ions doped in alkali barium borate glasses. The error in the observed band position is 620 cm 21 Glass system

2

B 1g → 2 B 2g cm 21

a2

b 12

Gp

Li Ba B Na Ba B K Ba B

12545 12465 12420

0.650 0.656 0.658

0.816 0.744 0.730

36.6 51.2 54.0

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

36

determine Cu 21 bonding coefficients, one needs in addition to the EPR parameters, the energy positions of the absorption bands of Cu 21 which indicates the values of E1 and E2 . Since the authors have observed only one absorption band corresponding to 2 B 1g → 2 B 2g transition, the position of second band can be estimated by the approximation [21]. 2k 2 l E( 2 B 1g → 2 Eg) 5 ]]]] 2.0023 2 g' 2

2

where k is the orbital reduction factor (k 50.77) and l is the spin orbit coupling constant. The observed band positions for different alkali barium borate glasses and their corresponding calculated molecular orbital coefficient values ( b 21 ) are given in Table 3. From the observed band positions it is found that the bands are shifted towards red side in Cu 21 : K Ba B glass as compared to Li Ba B and Na Ba B glass samples. This confirms that heavier the molecule the more the bands are shifted towards the lower frequency side [8]. 2 If a 51 the bond would be completely ionic. If the overlapping integral were vanishingly small and a 2 50.5, the bond could be completely covalent. However, because the overlap integral is sizeable we cannot speak strictly of covalent versus ionic bonds but we can say that the smaller the value of a 2 the greater the covalent nature of the bond. The trend is in the expected direction. Cupric ion is a network modifier and B 2 O 3 is network former in respect of Cu 21 –O–B bond. The b 21 reflects the competition between the cupric ion and its neighbouring network former cations for attracting the lone pairs of the intervening oxygen ions. The value of b 21 depends strongly on network former. In the present work, the b 21 value decreases with alkali content. This change in b 12 is related to the change in B–O bonds. The values of the calculated parameter a 2 and b 21 obtained for various glasses indicate that the in-plane s bonding is covalent whereas in-plane P bonding is significantly ionic in nature. The ionic nature of in-plane P bonding decreases from Li to K whereas the covalent nature of in-plane s bonding slightly decreases from Li to K. The normalized covalency of the Cu(II)–O in plane bondings of s and P symmetry are expressed [11] in terms of bonding coefficients a 2 and b 12 as follows. 200(1 2 S)(1 2 a 2 ) Gs 5 ]]]]]] % 1 2 2S

GP 5 200(1 2 b 21 )% where S is the overlapping integral (Soxygen 50.076). The normalized covalency values of the Cu(II)–O of in-plane bonding of p symmetry (GP ) thus calculated are given in Table 3, and it is seen that in-plane bonding of p symmetry increases from Li to K in this set of alkali barium borate glasses. This also receive support from optical basicity studies. In general optical basicity in-

creases with GP value [22]. The changes in b 21 are related to the changes in B–O bonds; there is a decrease in the strength of B–O bonds resulting in the increase of covalency of Cu(II)–O bonds. The Cu(II)–O bonds may be affected by the direct adjacent B–O bonds as well as by the little more distant ones [9]. The optical absorption spectra of Cu 21 ions in our glasses show a single broad absorption band ranging from 720–850 nm. For Cu 21 ion in a regular octahedral complex involving six equivalent ligands one absorption band is expected. The Cu 21 lacks cubic symmetry and this cubic symmetry is disturbed by electronic hole in the degenerate d x 22y 2 orbital which produces the tetragonal distortion. The Jahn–Teller theorem requires that any nonlinear system with a degenerate ground state must distort in order to remove the degeneracy. Then two structures are to be expected: one elongated and the other compressed structure. Experimental data show that the cupric ion generally exists in solutions, solids and gases in octahedral symmetry with a strong tetragonal distortion [23–26]. According to the present EPR studies, Cu 21 ions in alkali barium borate glasses have octahedral symmetry with tetragonal distortion.

5. Conclusions From the EPR and optical absorption studies of copper ions doped in 20 R 2 CO 3 124.5 BaCO 3 155 H 3 BO 3 10.5 CuCO 3 (R5Li, Na and K) the authors conclude the following 1. The Cu 21 ions are in tetragonally distorted octahedral sites. 2. The guu and A uu values changes with alkali content in alkali barium borate glasses. This is attributed to the structural changes with alkali content. 3. The linewidth of parallel and perpendicular hyperfine peaks increases with the magnetic quantum number (m) of copper nuclei, this has been attributed to the microenvironmental fluctuations around a Cu 21 ion which is intrinsic to the glassy state. The dguu and dg' values measured in this case are compared with others. The magnitude of dguu obtained in the present work is higher compared to the value reported by Hosono et al. This has been attributed to the decrease in the strength of the B–O bonds resulting in the increase of covalency of Cu(II)–O bonds. 4. The molecular orbital values a 2 and b 12 obtained for various glasses in the present work indicate that the in-plane s bonding is covalent and in-plane p bonding is significantly ionic in nature. 5. As the alkali content is changed from Li to K, all the observed bands are shifted towards lower energy side confirming that the heavier the molecule, the more the bands are shifted towards lower energy side.

R.P. Sreekanth Chakradhar et al. / Journal of Alloys and Compounds 265 (1998) 29 – 37

Acknowledgements One of the authors (JLR) thanks University Grants Commission (New Delhi) for providing financial assistance. AM thanks Council of Scientific and Industrial Research (New Delhi) for the award of Senior Research Fellowship.

References [1] J. Kreidl, J. Non-Cryst. Solids 123 (1990) 377. [2] D.L. Griscom, Mater. Sci. Res. Borate Glasses, in: L.D. Pye, V.D. Frechette, N.J. Kreidl (Eds.), Plenum, New York, 12, 1978, p. 36 [3] H. Imagawa, Phys. Stat. Solidi 30 (1968) 469. [4] L.D. Bogomolava, T.F. Dollgolenko, V.A. Jachkin, V.A. Lazukin, J. Magn. Reson. 15 (1973) 283. [5] H. Hosono, H. Kawazoe, T. Kanazawa, J. Non-Cryst. Solids 34 (1979) 339. [6] J.M. Dance, J.P. Darnandery, H. Branoby, M. Monveraye, Solid State Commun. 39 (1981) 199. [7] O. Cozar, I. Ardelean, J. Non-Cryst. Solids 92 (1987) 278. [8] A.S. Rao, J. Lakshmana Rao, S.V.J. Lakshman, J. Phys. Chem. Solids 53 (1992) 1221. [9] N.C. Biswas, R. Dayal, P. Chand, Phys. Chem. Glasses 37(1) (1996) 31.

37

[10] H. Hosono, H. Kawazoe, T. Kanazawa, J. Non-Cryst. Solids 29 (1978) 173. [11] H. Kawazoe, H. Hosono, T. Kanazawa, J. Non-Cryst. Solids 33 (1979) 103. [12] A.K. Bandyopadhyay, J. Mater. Sci. 15 (1980) 1605. [13] D.W. Moon, A.J.M. Aitken, R.K. McCrone, C.S. Cieloszyk, Phys. Chem. Glasses 16 (1975) 91. [14] A. Abragam, B. Bleaney, Electron Paramagnetic Resonance of Transition ions, Clarendon Press, Oxford (1970) p. 175. [15] B. Bleaney, K.D. Bowers, M.H.L. Pryce, Proc. Roy. Soc. A 228 (1955) 147. [16] J.A. Duffy, M.D. Ingram, J. Inorg. Nucl. Chem. 37 (1975) 1203. [17] L. Pauling, The Nature of Chemical Bond, 3 rd edition, Cornell University Press, New York, (1960), p. 93. [18] H.A. Kuska, M.T. Rogers, R.E. Durllinger, J. Phys. Chem. 71 (1967) 109. [19] D. Kivelson, R. Neiman, J. Chem. Phys. 35 (1961) 145. [20] F.M. Mabbs, D.J. Machin, Magnetism and Transition Metal Complxes, Chapman and Hall, London, (1973), p. 154. [21] A. Klonkowski, Phys. Chem. Glasses 24 (1983) 116. [22] A. Klonkowski, Phys. Chem. Glasses 26 (1985) 13. [23] D.S. Mc Clure, in: Solid State Phys., F. Seitz, P. Turnbull (Eds.), Academic Press, New York, 9, (1959), p. 359 [24] T. Bates, Modern aspects of the vitreous state, in: J.D. Mackenzie, (Ed.), Butterworths, London, 2,(1962), p. 195 [25] C.K. Jorgensen, Acta Chem. Scand. 9 (1955) 1362. [26] R. Harani, C.A. Hogarath, K.A.K. Lott, J. Mater. Sci. 19 (1984) 1420.