Single crystal EPR studies of Cu(II) doped in cadmium sodium sulphate hexahydrate: a case of low hyperfine coupling constant

Solid State Communications 128 (2003) 137–142 www.elsevier.com/locate/ssc

Single crystal EPR studies of Cu(II) doped in cadmium sodium sulphate hexahydrate: a case of low hyperfine coupling constant C. Shiyamala, R. Venkatesan, P. Sambasiva Rao* Department of Chemistry, Pondicherry University, Pondicherry 605 014, India Received 18 March 2003; accepted 30 July 2003 by H. Akai

Abstract Single crystal electron paramagnetic resonance studies of Cu(II) doped cadmium sodium sulphate hexahydrate are carried out at room temperature. Angular variation of copper hyperfine lines in the three orthogonal axes shows the presence of a single site in a substitutional position. The spin Hamiltonian parameters calculated from the spectra are: g11 ¼ 2:070; g22 ¼ 2:191; g33 ¼ 2:366; A11 ¼ 5:96 mT; A22 ¼ 1:65 mT and A33 ¼ 8:62 mT: The low value of A33 has been explained by considering considerable admixture of d2x 2 y2 ground state with d2z excited state. The admixture coefficients of ground state wave function are: a ¼ 0:281; b ¼ 0:957; c ¼ 0:057; d ¼ 0:032; e ¼ 20:032; where a and b correspond to admixture coefficients for d2z and d2x 2 y2 respectively. EPR powder spectrum at room temperature and 77 K gives identical spin Hamiltonian parameters, which matched fairly well with the single crystal data. Parameters k ¼ 0:2018; P ¼ 142:62 £ 1024 ; a2 ¼ 0:7149; a ¼ 0:8455 and a0 ¼ 0:5982 have also been calculated. q 2003 Elsevier Ltd. All rights reserved. PACS: 76.30. 2 v; 33.35. þ r; 71.20.B Keywords: A. Crystal growth; C. Crystal structure and symmetry; D. Crystal and ligand fields; D. Spin-orbit effects; E. Electron paramagnetic resonance

1. Introduction The electron paramagnetic resonance (EPR) technique is used to study paramagnetic ions in host lattice such as probes. These kinds of studies give valuable information about the site symmetry of the transition metal ions [1]. EPR of Cu(II) has been investigated in wide symmetry environments, viz., octahedral [2 – 4], tetrahedral [5], square planar, square pyramidal [6 –9], trigonal bipyramidal [10, 11] etc. The diamagnetic host ions possess a closed outer electronic shell, thereby causing the local symmetry of the host lattice to be high. When the Cu(II) ions are introduced in such host lattices, substituting for the diamagnetic ions, local distortions will take place, because of the mismatch of Cu(II) ion size to that of the host ions and dynamic effects such as the Jahn– Teller effect [12]. However, Cu(II) ion is * Corresponding author. Tel.: þ91-413-2665991; fax: þ 91-4132655265. E-mail address: [email protected] (P. Sambasiva Rao). 0038-1098/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0038-1098(03)00675-6

the simplest paramagnetic probe that enters easily into a number of host lattices and one gets an idea about the ground state (compressed/elongated octahedron) and type of Jahn– Teller distortion (static/dynamic/tunnelling) [13 – 16]. EPR of low symmetry complexes with d2x 2 y2 ground state have been extensively studied and only a few copper complexes are known with d2z ground state [17]. From the literature, it has been found that a small number of Cu(II) systems, whose ground state is not a pure d2x 2 y2 are reported. Due to low symmetry, the d2x 2 y2 gets admixed with other orbitals present, leading to low hyperfine coupling constant values for these systems. These complexes are also characterized by orthorhombic g-tensors which indicate the greater extent of anisotropic effect in the xy plane. The EPR studies of the low symmetry complexes have been thoroughly discussed by Pilbrow [18] and Viswanath et al. [19]. For the low symmetry complexes, the direction of g-tensors does not depend upon the metal ligand bonds. Examples for low symmetry complexes are Cu(II) in [Zn(v-nitroacetophenone)2 (4-methyl pyridine)]

138

C. Shiyamala et al. / Solid State Communications 128 (2003) 137–142

[20], [Zn(hexafluoro acetyl acetonato) (2,20 -bipyridine)] [20], zinc fluoride [21], zinc antipyrine nitrate [22], ciscatena-m-sulphatoaqua tris-(imidazole) cadmium(II) [23], etc. In our search for systems with low hyperfine coupling constant, we have come across a system, i.e. Cu(II) doped cadmium sodium sulphate hexahydrate. This system has very low hyperfine coupling constant and the results obtained for this ion are reported in this communication.

2. Experimental Cadmium sodium sulphate hexahydrate (CSSH) single crystals are grown at room temperature by slow evaporation of aqueous equimolar solutions of cadmium sulphate and sodium sulphate. To this, a few drops of copper sulphate solution are added as paramagnetic impurity. Good blue colored crystals are obtained in about 20 days. The single crystals of Cu(II)/CSSH are studied at room temperature in the JEOL JES-TE100 ESR spectrometer, having a modulation of 100 kHz and operating at X-band frequency. DPPH with a g value of 2.0036 is used for g factor calculations. Low temperature measurement has been made using a quartz Dewar. A single crystal of proper size is mounted on to the goniometer by means of quick-fix to carry out rotations at room temperature in the three mutually orthogonal ab, bc p and ac p planes. Powder spectrum of the sample is recorded by taking the sample in a quartz tube.

3. Crystal structure The crystal lattice CSSH belongs to the group of Tutton’s salts. The Tutton’s salts have monoclinic crystal structure with space group P21=n : The Tutton’s salt has been found to have the general formula of M00 M20 (XO4)2·6H2O, where M00 is a divalent cation like Co, Cu, Ni, Mg, Zn, Cd; M0 is a monovalent like K, Cs, Rb, Na, NH4; and X is S or Se. The lattice parameters of CSSH (M00 is Cd, M0 is Na and X is S) are: a ¼ 0:613 nm; b ¼ 1:223 nm; c ¼ 0:909 nm; b ¼ 104:788 and Z ¼ 2: Crystal a axis is considered perpendicular to crystal b axis in ab plane, whereas the third axis c p is considered perpendicular to both a and b axes. The cadmium ion in CSSH is surrounded by six water molecules in the form of a distorted octahedron. In all Tutton’s salts, the shortest Cd– O bond is unique, whereas the longest bond depends on the nature of M0 and M00 [24].

4. Results and discussion Single crystal EPR studies are carried out at room temperature for Cu(II)/CSSH. Fig. 1 shows the EPR spectrum of Cu(II)/CSSH crystal in bc p plane at two different orientations with respect to the magnetic field (top

Fig. 1. Top two ones correspond to single crystal EPR spectra of Cu(II)/CSSH in bc p plane at two orientations at RT ðn ¼ 9:12093 GHzÞ: Bottom figure corresponds to ac p plane ðn ¼ 9:12336 GHzÞ:

two) and at a particular orientation in ac p plane (bottom in Fig. 1). The recorded EPR spectra consists of four lines characteristic of Cu(II) with S ¼ 1=2 and I ¼ 3=2: As expected, along a crystallographic axis, only four lines are noticed, indicating the incorporation of copper in the host lattice. As the crystal is rotated in bc p plane, two sets of four lines are seen, as shown in Fig. 1. As mentioned in the crystal structure, the unit cell contains two molecules per unit cell and hence two sets of four lines are observed. The angular variation studies have been performed in the three planes. Figs. 2 – 4 show the variation of hyperfine lines in the three planes bc p, ac p and ab respectively. Fig. 2 confirms that the system contains two sites, which become magnetically equivalent along b and c p axes. However, in Fig. 3, the second site is not fully separated, but appeared only at a few orientations. In addition, at angles between 140 and 160,

C. Shiyamala et al. / Solid State Communications 128 (2003) 137–142

Fig. 2. Angular variation plot of Cu(II)/CSSH in the bc p plane ðn ¼ 9:12093 GHzÞ:

the hyperfine value has become very small (bottom of Fig. 1). Also, in Fig. 4, the second site does not appear at any orientation. These results can be explained later, by considering the g matrix (see below). The EPR spectrum of Cu(II) ion in CSSH are fitted to the following spin Hamiltonian with orthorhombic symmetry, H ¼ g11 bBx Sx þ g22 bBy Sy þ g33 bBz Sz þ A11 Sx Ix þA22 Sy Iy þ A33 Sz Iz where the symbols have their usual meaning. By making use of EPR – NMR program [25] and Schonland procedure, the spin Hamiltonian parameters g and A are calculated from the isofrequency plots and are given in Table 1. The direction cosines of g and A tensors are almost coincident, as evidence from the Table 1 and isofrequency plots. From the crystal data of CSSH, the direction cosines of various cadmium – oxygen direction cosines have been calculated and are also given in Table 1. One of the principle values of g and A tensors has a direction cosine value, which is very close to that of Cd –O3. This

139

Fig. 4. Angular variation plot of Cu(II)/CSSH in ab plane ðn ¼ 9:12256 GHzÞ:

further confirms that the paramagnetic impurity Cu(II) has entered the lattice CSSH, substitutionally in the place of Cd(II). The EPR spectra have been simulated at a few orientations using the EPR – NMR program [25], where the agreement is good. The three principal g and A values are 2.070, 2.191, 2.366 and 5.96, 1.65, 8.62 mT. It can be labeled as gy ; gx ; gz and Ay ; Ax ; Az respectively. During the isofrequency plot in bc p plane (corresponding to yz plane), the difference between gy ; gz and Ay ; Az values is appreciable and hence, the second site is observed in the isofrequency plot (Fig. 2). Table 1 Principal values and direction cosines of Cu(II)/CSSH and direction cosines of Cd –oxygen direction in the crystal lattice g Matrix 2.2004

Principal g values 20.0031 2.1232

0.0568 0.1039 2.3027

A matrix (mT) 3.02

2.0699 2.1906 2.3658 Principal A values

20.19 6.02

2.68 0.72 7.19

1.65 5.96 8.62

Direction cosines of g matrix 20.2950 0.1063 20.8797

20.9316 20.3168 0.1780

20.2123 0.8721 0.4410

Direction cosines of A matrix

Fig. 3. Angular variation plot of Cu(II)/CSSH ac 9:12336 GHzÞ:

p

plane ðn ¼

20.2117 0.4160 20.8844 Cd–O1 Cd–O3 Cd–O4

20.9707 20.1942 0.1410 0.0000 20.7314 20.7143

20.1131 0.8884 0.4449 20.8902 0.3344 20.3265

0.4556 0.5944 20.6190

140

C. Shiyamala et al. / Solid State Communications 128 (2003) 137–142

However, in the case of xz plane of rotation, the difference in the g values is not that much and hence appeared only at a few orientations. Finally, in the case of xy plane of rotation, the difference is so small that it did not appear at any orientation. This roughly explains the behaviour of g and A values in the three isofrequency plots. In order to confirm the single crystal analysis, powder spectrum of the sample is recorded at room temperature and is given in Fig. 5. It clearly indicates the orthorhombic nature of the impurity. The g and A values have been calculated from this spectra and are given in Table 2. Using these parameters and Bruker Simfonia program, the powder spectrum has been simulated and included in Fig. 5 itself. The agreement is very good and the values have matched fairly well with the single crystal analysis. As the central hyperfine lines are not resolved at room temperature in the powder spectrum, the sample is cooled to 77 K. The 77 K EPR spectrum of the powder sample does not show any change in g=A values and in resolution. A close look at the hyperfine values indicates that one of the principal values ðA33 Þ is appreciably smaller than a generally observed value of Ak for a copper ion. A few systems, having a smaller hyperfine value, have been selected from the literature and are given in Table 2, for comparison. One of the previously studied system of ours, i.e. Cu(II)/ZPPH [26] and the present system have the lowest A33 values, reported in the literature. This low value of A has been explained by considering the admixture of d2x 2 y2 (ground state) with d2z (excited state) orbital and metal – ligand covalent character [20]. The spin orbit coupling will further modify the weights of the mixing of the eigenstates. Generally, the contribution will be small. However, a significant contribution is needed to explain the observed very low hyperfine coupling constant. In other words, the magnitude of A33 gets decreased because of the opposite sign of hyperfine coupling value for the electrons in d2x 2 y2 and d2z orbital. The total dipolar coupling gets vanished to first order, for 1:1 admixture of d2x 2 y2 and d2z :

Table 2 Spin Hamiltonian parameters for Cu(II) in different host lattices System

g Values

A Values

CSSH

2.191 2.070 2.366

6.0 1.6 8.6

Present study

Powder

2.178 2.054 2.421

6.1 2.3 7.5

Present study

Cadmium ammonium sulphate hexahydrate Site I 2.142 4.5 2.052 3.5 2.414 10.4

Reference

[1]

Site II

2.137 2.057 2.419

4.7 3.3 10.3

ZPPH

2.188 2.032 2.372

6.5 5.0 8.0

[26]

2.101 2.081 2.392

2.1 1.0 10.8

[27]

4.6 1.8 13.0

[29]

Zn(AP)2(NO3)2

Di ammonium D -tartrate Site I 2.093 2.060 2.337 Site II

2.092 2.043 2.330

3.6 1.9 13.1

Cd(stpy)3(NO3)2·1/2stpy 2.108 2.066 2.298

5.4 2.3 10.7

[30]

Zn(stpy)3(NO3)2·1/2stpy 2.111 2.067 2.296

5.5 2.3 10.8

[30]

ZPPH: zinc potassium phosphate hexahydrate (hyperfine values are in units of mT).

Fig. 5. Experimental and simulated powder spectra of Cu(II)/CSSH ðn ¼ 9:40406 GHzÞ:

Hence it is found that a 10% admixture of d2x 2 y2 and d2z results in a 20% reduction in dipolar anisotropy. The low hyperfine value is also due to a direct mixing of metal 4 s orbital with the ground state orbitals of Cu(II) having low symmetry. The admixture opposes the core polarization of isotropic coupling, which occurs due to the involvement of outer 4 s orbital, results in the reduction of hyperfine values [27,28]. This is mainly due to the presence of electron in a d2x 2 y2 orbital, which is the ground state. The admixture coefficients of d2x 2 y2 and d2z can be evaluated using the

C. Shiyamala et al. / Solid State Communications 128 (2003) 137–142

method given in the literature [21]. It is well known that the spin orbit coupling mixes the ground state with the excited state. From g values, the coefficients of the d-orbitals of the Kramer’s doublet are determined. The d2x 2 y2 orbital gets admixed with dxy ; dyz and dxz : The d2z gets admixed with dyz and dxz : The Kramer’s doublet wave functions for the ground state can be expressed as,

c ¼ af1 a þ bf3 a þ icf2 a 2 id f4 b 2 ef5 b; cp ¼ iðaf1 b þ bf3 b 2 icf2 b 2 id f4 a þ ef5 aÞ where f1 ¼ d3z2 2r2 ðAÞ; f2 ¼ dxy ðB1 Þ; f3 ¼ dx2 2y2 ðAÞ; f4 ¼ dyz ðB3 Þ and f5 ¼ dxz ðB2 Þ: This equation represents that if a ¼ 1; the system has d2z ground state and lowest A33 value and if b ¼ 1; the system is in dx2 2y2 ground state and maximum hyperfine value. The constants a; b; c; d and e indicate the mixing of d-orbitals brought about by metal spin –orbit coupling. In terms of admixture coefficients, the expression for the g and A values are given as,

141

Table 3 Admixture coefficients (a; b; c; d and e) for various crystal lattices System

Admixture coefficients

Reference

CSSH ZnF2 SAIC ZPPH ZAPN SCB SCC

0.281, 0.957, 0.057, 0.032, 20.032 0.975, 0.198, 0.048, 0.061, 20.061 0.151, 0.988, 0.041, 20.014, 0.019 0.330, 0.941, 0.051, 0.032, 20.032 0.142, 0.988, 0.050, 0.022, 20.017 0.128, 0.989, 0.057, 0.016, 0.033 0.144, 0.988, 0.041, 20.041, 0.019

Present study [21] [23] [26] [27] [31] [31]

SCB: sarcosine cadmium bromide; SCC: sarcosine cadmium chloride; ZAPN: Zn(AP)2(NO3)2; SAIC: cis-catena-m-sulphatoaquo tris(imidazole) cadmium.

gz ¼ 2 2 4d 2 2 4e2 þ 8bc þ 4de;

and are given in Table 4, along with some known values. The ratio of Pcomplex to Pfree ion is around 60%, indicating the delocalisation of the d-electron. The percentage of unpaired spin density on copper ion is 40% and the remaining density is being distributed onto the ligands. The molecular orbital coefficient a2 ; which gives a measure of covalent nature of s-bonding is given as,

p gy ¼ 2 2 4c2 2 4e2 þ 4 3ad 2 4ce þ 4bd;

a2 ¼ Ak =0:036 þ ðgk 2 2:0023Þ þ 3=7ðg’ 2 2:0023Þ þ 0:04

p gx ¼ 2 2 4c2 2 4d 2 þ 4 3ae 2 4be þ 4cd; Az ¼ P{8bc þ 4de þ ð6j 2 kÞð1 2 2d 2 2 2e2 Þ2; 2

2

2

p

3j½4c þ 4b 2 e þ ð3Þaðd þ eÞ þ 3ðd 2 eÞðc 2 bÞ }; p Ay ¼ P{4 ð3Þad 2 4ce þ 4bd þ ð6j 2 kÞð1 2 2c2 2 2e2 Þ p p 2 3j½ð ð3Þa þ bÞ2 2 c2 þ 4d 2 2 e2 2 ð3Þaðe þ 2cÞ þ 3dc 2 3be 2 3de }; p Ax ¼ P{4 ð3Þae þ 4dc 2 4be þ ð6j 2 kÞð1 2 2c2 2 2d 2 Þ p p 2 3j½ð ð3Þa þ bÞ2 2 c2 2 d2 þ 4e2 2 ð3Þaðd 2 2cÞ þ 3ce 2 3db þ 3de } a; b; c; d and e are the coefficients of f1 ; f3 ; f2 ; f4 and f5 [22] respectively. j is a constant and depends on the electronic configuration of the ion and the value of j ¼ 2=21 for Cu(II) ion. P is the gyromagnetic ratio of copper and its free ion value is 360 £ 1024 cm21 and k is the Fermi contact term, which is a measure of bonding effects on the Cu(II) in the crystal lattice. Assuming that d ¼ 2e; the coefficients a; b; c and d have been calculated and given in Table 3. Using the similar procedure, the coefficients have been calculated for Cu(II)/ZPPH also and are included in Table 3, along with some other values. As expected, an increase in the coefficient of a is noticed whenever a decrease in hyperfine is observed. From the above equations, P and k have been calculated

where g’ is the average of g11 and g22 : Another parameter a0 is also evaluated from the expression,

a0 ¼ ð1 2 a2 Þ1=2 þ aS where S is the overlap integral between dx2 2y2 orbital and normalized ligand orbital. The value of S is given as 0.076 for a copper complex with water ligands. The complex is found to be partially covalent in nature. These values are also given in Table 4. Yet, another parameter R; the ratio difference between second and third values to first and second values of the increasingly arranged g value is also calculated, which gives an idea about the ground state nature of the paramagnetic impurity on the host lattice. In this case, the R value is found to be 0.51 and so the ground state is of dx2 2y2 type.

5. Conclusions The room temperature single crystal EPR spectra of Cu(II)/CSSH show the presence of single site namely substitutional site. The angular variation of the EPR spectra reveals that Cu(II) ions have been substituted at Cd(II) site. The very low parallel component of hyperfine coupling constant has been explained by considering the admixture of dz2 with dx2 2y2 : The bonding parameter indicates that the metal– ligand bondings are fairly covalent in nature. From the observed spin Hamiltonian parameters, the admixture coefficients of the ground state have been calculated. The EPR spectrum recorded at liquid nitrogen temperature does

142

C. Shiyamala et al. / Solid State Communications 128 (2003) 137–142

Table 4 Molecular orbital coefficient of some Cu(II) systems System

k

P ( £ 1024 cm21)

a2

Reference

CSSH ZPPH Cd(stpy)3(NO3)2·1/2stpy Zn(stpy)3(NO3)2·1/2stpy

0.202 0.281 0.271 0.272

142.6 165.5 227.3 226.5

0.715 0.678 0.249 0.257

Present study [26] [30] [30]

not show any change compared to that of room temperature spectrum.

Acknowledgements The authors thank DST, UGC, AICTE and CSIR {01(1771)/02/EMR-II} for financial support.

References [1] F. Koksll, I. Kartal, F. Ucun, Solid State Commun. 98 (1996) 1087. [2] V.S. Xavier Anthonisamy, R. Murugesan, Chem. Phys. Lett. 287 (1998) 353. [3] R. Murugesan, S. Subramanian, Mol. Phys. 52 (1984) 129. [4] B.J. Hathaway, Struct. Bonding (Berlin) 57 (1984) 55 and references therein. [5] J. Gouteron, S. Jeanin, Y. Jeanin, J. Livage, C. Sanchez, Inorg. Chem. 23 (1984) 3387. [6] N. Wei, N.N. Murthy, K.D. Karlin, Inorg. Chem. 33 (1994) 6093. [7] M. Velayutham, B. Varghese, S. Subramanian, Inorg. Chem. 37 (1998) 5983. [8] G.A. McLachlan, G.A. Fallon, R.L. Martin, L. Spiccia, Inorg. Chem. 34 (1995) 254. [9] P. Comba, T.W. Hambley, M.A. Hitchman, H. Stratemeier, Inorg. Chem. 34 (1995) 390. [10] J.G. Lozono, J.S. Carrio, E. Escriva, J.V. Foldado, C. Molla, L. Lezama, Polyhedron 16 (1997) 939. [11] D. Reinen, C. Friebel, Inorg. Chem. 23 (1984) 791. [12] T.K. Kundu, P.T. Manoharan, Chem. Phys. Lett. 241 (1995) 627. [13] P. Sambasiva Rao, A.K. Viswanath, S. Subramanian, Spectrochim. Acta 48A (1992) 1745 and references therein.

[14] B.L. Silver, D. Getz, J. Chem. Phys. 61 (1974) 630–638. [15] S.N. Rao, Y.P. Reddy, P. Sambasiva Rao, Solid State Commun. 78 (1991) 1025. [16] R.M. Krishna, J.L. Rao, S.V.J. Lakshman, Radiat. Eff. Defects Solids 132 (1994) 67. [17] R. Murugesan, S. Subramanian, J. Magn. Reson. 36 (1979) 389 and references therein. [18] J.R. Pilbrow, Transition Ion Electron Paramagnetic Resonance, Clarendon Press, Oxford, 1990. [19] A.K. Viswanath, S. Radhakrishna, J. Phys. Chem. Solids 52 (1991) 227. [20] D. Attansio, J. Magn. Reson. 26 (1977) 81. [21] J.D. Swalen, B. Johnson, H.M. Gladney, J. Chem. Phys. 52 (1970) 4078. [22] D. Srinivas, M.V.B.L.N. Swamy, S. Subramanian, Mol. Phys. 57 (1986) 55. [23] R. Murugesan, Ph.D Thesis, I.I.T. Madras, 1981. [24] K.K. Kannan, M.A. Viswamitra, Z. Kristallogr. 16 (1965) 122. [25] EPR–NMR Program developed by F. Clark, R.S. Dickson, D.B. Fulton, J. Isoya, A. Lent, D.G. McGavin, M.J. Mombourquette, R.H.D. Nuttall, P.S. Rao, H. Rinnerberg, W.C. Tennant, J.A. Weil, University of Saskatchewan, Saskatoon, Canada, 1996. [26] P. Sambasiva Rao, T.M. Rajendiran, R. Venkatesan, N. Madhu, A.V. Chandrasekhar, B.J. Reddy, Y.P. Reddy, R.V.S.S.N. Ravikumar, Spectrochim. Acta 57A (2001) 2781. [27] D. Attanasio, G. Dessy, V. Fares, J. Chem. Soc., Dalton Trans. (1979) 28. [28] J.A. Guckert, M.D. Lowery, E.J. Solomon, J. Am. Chem. Soc. 117 (1995) 2817. [29] R. Tapramaz, B. Karabulut, F. Koksal, J. Phys. Chem. Solids 61 (2000) 1367. [30] C. Karunakaran, K.R. Justin Thomas, A. Shanmugasundaram, R. Murugesan, Spectrochim. Acta 57A (2001) 441. [31] D. Pathinettam Pandiyan, C. Muthukrishnan, R. Murugesan, Cryst. Res. Technol. 35 (2000) 595.