Effect of Substitution of Titanium by Tin on Properties of Fe2TiO5

Iron titanate is a pseudobrookite which exhibits many interesting properties such as spin glass behaviour, thermal microcracking, high resistivity, etc. A cluster approach may be used for a description of spin glass behaviour of the pseudobrookites [1]. D. A. Kharmov, et.al [1] have studied the spin glass transition in the Fe2 Ti1-x Snx O5 pseudobrookite and found that Sn4+ ions appear to occupy only M1/ 4c octahedral sites and spin glass transition temperature Tg depends on tin concentration monotonically. According to the XRD analysis and Sn-Mossbauer Spectral (MS) data the maximum solubility of Sn4+ ions in the pseudobrookite structure at 1250oC is x = 0.22 [1], where the unit cell volume increases linearly with the increasing tin concentration. The XRD analysis of the limit of solid solution Fe2 Ti1-x Snx O5 (0

SnO2 (sintered ceramics and also compressed powder under very high pressure) exhibits an astonishingly high value of dielectric constant of the order of 105 [5]. Unlike other oxide semiconductors such as Ni-Zn ferrites etc. the dielectric constant of SnO2 is not markedly dependent on the applied frequency in the audio frequency range. However,in the R. F. range the dielectric constant decreases with frequency [5]. A dielectric hysteresis loop has been observed in SnO2. [5].

The influence of a small amount of SnO2 on the structural changes of nano-crystalline TiO2 has been investigated [6]. It is obsereved that small amount of SnO2 doping enhances the rutilation of anatase and effectively prohibits the grain growth in these powders [6]. The incorporation of about 17% of Sn completely transforms anatase to the rutile form at a calcination temperature as low as 500oC [7]. The optical analysis shows that the band gap and Fermi level of theTi1-x Snx O2 solid solutions increases with increasing x and these solid solutions are expected to be of better photocatalytic properties [7].The as Al, Co, Nb have shown the nonlinear I-V characteristics [8]. The nonlinear coefficient is found to decrease with increase in x. On the other hand breakdown voltage (Eb) shows the increase with the increase in x. The low current in the pre breakdown region implies that the grain boundary resistance is high for the solid solutions. However, the grain boundary barrier or energy barrier is ineffective. The solubility of Ti4+ ions in SnO2 is up to 25 mole %. SnO2 has been supported on TiO2 , Al2O3 , MgO and SiO2 [9]The reduction of the species is easier if it is present on the surface than within the structure of support. It is found that on titania a smaller Quadrupole Splitting (QS) is observed, which might be explained assuming that Sn4+ is inside the structure of the titania. This is in agreement with the findings reported by Bartholomew and Boudart [10], who have reported that the Quadrupole Splitting (QS) for an atom in a crystal is lower than on the surface.

Ceramic bodies have been prepared from the ferric stannates and their dielectric properties have been measured [11]. They are found to be semiconductor. The resistivity of Fe2(SnO3)3 as determined using the direct current at room temperature is 104 Ω-cm. The addition of this stannate to BaTiO3, may be effective in decreasing maturing temperature and improving body density. The substitution of Sn4+ ion for the Ti4+

ion may also be expected to improve the stability of titanate bodies with regard to changes in the state of oxidation during normal firing treatment, since SnO2 is more stable in this respect than TiO2 [12].The Mossbauer Spectral (MS) data of tin doped Fe3O4 recorded at elevated temperature in vacuum shows the decrease in the Curie temperature with increase in concentration of the dopant [12]. The MS data recorded from all tin doped iron oxides show the presence of a hyperfine magnetic field at the Sn4+ sites. The study of Reitveld structure refinement of XRD of tin doped α-Fe2O3 has been reported to contain tin in both interstitial and substitutional sites of the corundum-related α-Fe2O3 structure [13]. Zirconium titanate-stannate solid solutions Zr1-x Snx TiO4 are known to have a high dielectric constant, a high Q-value and a low temperature coefficient of resonant frequency [14].

2. RESULTS AND DISCUSSION:

The structural properties of the samples namely Fe2TiO5 using rutile TiO2 [FTR], Fe2TiO5 using anatase TiO2 [FTA], Fe2 Ti0.75 Sn0.25 O5 using rutile TiO2 [FTSR] and Fe2 Ti0.75 Sn0.25 O5 using anatase TiO2 [FTSA] all synthesized at 1250oC are reported here. The data is indexed in a single phase orthorhombic structure having the space group Bbmm of the pseudobookite. Hence it is concluded

For a given ceramic, the dielectric constant (K‘) depends on the factors such as electron polarization, atomic polarization, interfacial sspolarization (space charge) and dipolar polarization at micro level [4]. Also it is known to depend on the parameters such as particle size, porosity, inter-granular phases, lattice impurities on the grain boundaries, sintering temperature, etc. at macro level [4,16]. Therefore dielectric constants of our samples at room temperature at 1kHz, 10 kHz, 100 kHz, 1000 kHz, Debye particle size, porosity, inhomogeneity and space charge (defined as K‘1kHz-K‘1000 kHz) are included in the Table-1 to facilitate the comparison.

It is observed that the anatase causes the reduction in the space charge. If K‘1000 kHz represents the contribution to the dipolar polarization, then interestingly the anatase enhances the dipolar polarization considerably. On the other hand, the presence of Sn4+ increases the contribution towards both the space charge as well as dipolar polarization. However, the increase in the space charge is significant. The space charge appears to have a weak inverse dependence on the inhomogeneity, porosity, and the Debye particle size. On the other hand, the dipolar contribution (K‘1000 kHz) has a weak direct dependence on these structural parameters.

In order to obtain more information about the space charge the relaxation spectra of the samples are also investigated. The plots (i) K‘ v/s log (frequency)

and (ii) K‖ v/s log (frequency) and (iii) tan () Vs log (frequency), known as relaxation spectra of the samples are given in Figures-1 to 3 respectively. The nature of the curves is similar to that shown by Maxwell-Wagner model for the space charge. The shapes of the curves indicate the presence of the space charge in all the four samples. (K‘1kHz - K‘1000

The variation of dielectric constant K‘ with temperature (300-650 K) is investigated for all the samples. The variation of dielectric constant K‘ with temperature for [FTR], [FTSR], [FTA] and [FTSA] are shown in the Figures-4 to 7 respectively. The dielectric constant K‘ first increases slowly to saturate approximately above temperature 600 K. The heating and cooling cycles exhibit distinct hysteresis curves. The hysteresis curves in the Figures-4and 5 are measured for the samples [FTR] and [FTSR] respectively at 1 kHz and therefore correspond mainly to the space charge. [FTSR] exhibits antihysteresis. Whereas, the hysteresis curves in the Figures-6 and 7 are measured for the samples [FTA] and [FTSA] respectively at 1000 kHz and therefore correspond also to the dipolar contribution. This is because the samples prepared from rutile It is interesting to note that the presence of space charge (which takes part in electrical conduction) means the loops correspond to the thermal diffusivity [17]. In [FTSA] a square loop super imposed on the thermal diffusivity corresponds to the dipolar component. Overall, the loop area is increased by the anatase and Sn4+ has opposite effects on the rutile and anatase. The loop area is better described in terms of the order parameter λ’ which depends on the cation distribution. The hysteresis of space charge decreases in area whereas hysteresis of dipolar polarization increases in area with the increase in order parameter λ‘.

The room temperature measurements of a. c. and d. c. resistivities along with the order parameter λ’, inhomogeneity and thermal (ρa.c.) hysteresis data are reported in the Table -2. The pseudobrookite is an n-type semiconductor [16]. Its conduction is associated with impurities, defects, interstices and therefore is complex. It is interesting to note that the a. c. resistivities increase by the use of anatase. Also the inhomogeneities are larger in the anatase based samples. Thus, the stress may be responsible to reduce the carrier density (space charge) and the mobility. The d. c. resistivities decrease in the anatase based pseudobrookites. Since the (ρd.c.) depends on the grain boundary, perhaps the increased porosity of anatase based samples decreases (ρd.c.) of these samples. The increase in order parameter λ’ (i. e. the increase in lower tetrahedral symmetry at M2 site) appears to increase the a. c. conductivity.

The dependence of a. c. (1 kHz) resistivity (ρa.c.) on temperature is investigated from 300 to 850 K. The graphs of the log (ρ) versus the reciprocal of absolute temperature (1/T) of the samples [FTR], [FTSR], [FTA] and [FTSA] under study are depicted in Figures-8 to 11 respectively. The curves corresponding to heating and cooling are depicted in these figures to facilitate the investigation of ‗Thermal Hysteresis‘ due to microcracking observed in the pseudobrookite [17]. Which affects the thermal conductivity and therefore also the electrical conductivity.

Figure-11.Plot of A.C. Resistivity Vs Reciprocal of temperature (1/T) for the sample[FTSA] The conduction in this region is attributed to be due to the large polarons in the intrinsic region. 2 Ea2 is, therefore, regarded as the band gap which ranges from 0.8 to 1.2 eV. The band gap is larger for the anatase based pseudobrookite. The transition temperature (T) is attributed to the antiferro-paramagnetic transition. The third region from 560 to 850 K is therefore called post transition region and the activation energy (Ea3 ≈ 0.25 to 0.75 eV) for this region is tabulated in the Table-2. It is interesting to note that the loop area of (ρac) hysteresis increases with order parameter λ’ in case of rutile samples whereas it decreases in case of anatase samples. It is also interesting to note that the samples [FTSR] and [FTSA] exhibit antihystereses and Ea3< Ea2 for these samples (Table -2). Hence we may conclude that these antihystereses are due to tunneling effect [21].

It is interesting to note that substitution of Ti 4+ by Sn4+ is responsible to increase the unit cell volume and X-Ray or theoretical density. On the other hand, the use of anatase in the reaction causes the decrease in practical density and the increase in the porosity, Debye particle size and the inhomogeneity. This may be attributed to the vertex –sharing network of anatase which is opposite to the edge –sharing network of the pseudobrookite. However, tin has exactly opposite effect on the anatase as the practical density, porosity, inhomogeneity and the Debye particle size are concern. It is observed from the dielectric measurements that the anatase causes the reduction in the space charge. If K‘1000 kHz represents the contribution to the dipolar polarization, then interestingly the anatase enhances the dipolar polarization considerably. On the other hand, the presence of Sn4+ increases the contribution towards both the space charge as well as dipolar polarization. However, the increase in the space charge is significant. The hysteresis of space charge decreases in area whereas hysteresis of dipolar polarization increases in area with the increase in order parameter λ‘. The post transition activation energies (Ea3) and the transition temperatures (T) decrease with order and Ea3< Ea2 for these samples. Hence we may conclude that these antihystereses are due to tunneling effect.

1) D. A. Khramov, M. A. Glazkova, S.S. Meshalkin (1995). J. Magn. Mag. Materials, 150, pp. 101. 2) S. S. Meshalkin, M. A. Glazkova, D. A. Kharmov, V. S. Urusov (1994). Mater. Sci. Forum, 166-169, pp. 717. 3) G. M. Irwin, E. R. Sanford (1991). Physical Review B, 44, 9, pp. 4423-4430. 4) S. V. Salvi, A. H. Karande, M. A. Madare, S. S. Gurav (2005). J. Ferro. and Inte. Ferro., Netherland, 323, pp. 97. 5) L. V. Deshpande, V. G. Bhide (1960). Lett. Alla Redazione, pp. 816. 6) X. Z. Ding, Z. A. Qi, Y. Z. He (1994). J. Nano Stru. Mater., 4, 6, pp. 663. 7) S. Mahanty, S. Roy, Suchitra Sen (2004). J. Crys. Growth, 261, pp. 77. 8) V. Ravi, S. K. Date (2001). Bulle. Mater. Sci., 24, 5, pp. 483. 9) Noel Nava, Thomas Viveros (1999). Hyperfine Inter. 122, pp. 147. 10) C. H. Bartholomew, M. Budart (1973). J. Catalysis, 29, pp. 278. 11) William W. Coffeen (1953). J. Amer. Cer. Soc., 36, 7, pp. 207. 12) Frank J. Berry, O. Helgason (2000). Hyper. Inter., 126, pp. 269. 13) F. J. Berry, C. Greaves, J. McManus, M. Mortimer (1997). G. Oates, J. Sol. Stat. Chem.130, pp. 272. 14) S. R. Dhage, V. Ravi, S. K. Date (2003). Bull. Mater. Sci. 26, 2, pp. 215. 15) Linus Pauling, Nature of Chemical Bonds, 3rd Edn. Cornell Univ. Press, Ithac., New York. 16) R. S. Singh, T. H. Ansari, R. A. Singh, B. M. Wanklyn, B. E. Watt (1995). Sol. Stat. Comm., 94, 12, pp. 103. 18) T. Birchall, N. N. Greenwood and A. F. Reid (1969). J. Chem. Soc. A) pp. 2388. 19) P. T. Supe and K. R. Rao (1974). Current Science, 43, pp. 377. 20) U. Atzmony, G. Gorodetsky, E. Gurewitz (1979). Phys. Rev. Lett. 43, 11, pp. 782. 21) S. Gueron, M. M. Deshmukh, E. B. Myers and D. C. Ralph (1999). Phys. Rev. Lett., 83(11), pp. 4148.

Department of Physics, K. E S. Anandibai Pradhan Science College, Nagothane, Maharashtra