First Observed Metal-like to Insulator Transition in the vacant 3d orbital Quantum Spin Liquid Tb 2 Ti 2 O 7
11 Abstract
We report the first observed metal to insulator transition (MIT) in the 3 d - pyrochlore oxide Tb Ti O at 603 K due to the interaction of empty 3 d orbitals of Ti with O ions, evidenced by the transition in resistivity. Magnetic susceptibility, specific heat capacity and differential scanning calorimetry supports the MIT and the transition is of second order. A broad change in magnetic moment, without any magnetic phase transition, is a behaviour typical in this family of compounds - a phenomenon that is seldom observed in an empty d orbital pyrochlores. The possible mechanism that supports the MIT in Tb Ti O is discussed. Another, magnetic transition is also evident at 696 K supported by thermal studies. Thermogravimetric analysis confirms, the observed MIT and magnetic transition are not due to oxygen vacancy. First Observed Metal to Insulator Transition in the vacant 3 d orbital Quantum Spin Liquid Tb Ti O B. Santhosh Kumar , and C. Venkateswaran *
1. Department of Nuclear Physics, University of Madras, Guindy campus, Chennai – 600 025, India. * E mail: [email protected] I. Introduction
A little frustration in the spin of a material gives remarkable properties also making its physics more interesting. This is true in geometrically frustrated materials which have unusual phenomena like random alignment of spins, interaction with neighbouring atoms, corner sharing tetrahedra and many more [1-8]. The term ‘geometrical frustrated’ refers, the spin of individual ions are well ordered but with the absence of individual interaction due to the competing magnetic behaviour of neighbour ions [9]. A large number of compounds in the frustrated family of materials crystallize in the stoichiometry A B X (X = F, O), where A is a rare earth trivalent ion and B is a transition metal ion (A = Tb, Y, Dy, Ho, Gd, Er, Pr and B = Mo, Ti, Sn, Ir, Ru, Os, Mn) with eight-fold and six-fold oxygen co-ordination, respectively [7,10]. This class of compounds form two interpenetrating lattice network of corner sharing tetrahedra of A and B which makes them thermal and chemically inert [7,10]. The electrons in the 4d/5d pyrochlore compounds are localised compared to the 3d pyrochlores due its greater hybridisation with O anions. The higher spatial degree of 4d/5d wave function over 3d leads to a decrease in the intra-atomic Coulomb repulsion and increases the wave function of the elements/ compounds. The degree of correlation, in electron correlated systems, is defined as the ratio of intra-atomic Coulomb repulsion (Hubbard U) to the bandwidth (W) [11-13]. The electron wave function will increase from 3d to 5d leading to a larger bandwidth that follows the order W > W > W and the Columbic repulsion will decrease accordingly as U < U < U . Typically, U/W >> 1 for 3d oxides, for metals U/W << 1 and for 4d/5d systems U/W ~ 1. This property makes the physics of 3d pyrochlores more interesting. Metals are different from insulators/semiconductors by their variation in resistivity (ρ) as a function of temperature. For a metal, dρ/dT > 0 and for insulator /semiconductor dρ/dT < 0. If a compound is associated with MIT, then as a function of temperature, the resistivity increases and beyond a critical temperature the resistivity decreases. But the resistivity decreases with rise in temperature for both the semiconductor and insulator. Hence it becomes difficult to distinguish between metal to insulator transition (MIT) and Metal to semiconductor transition (MST). An example is the pyrochlore Cd Os O in which a semiconductor to metal transition is reported by A.W. Sleigh. et al. in 1973 [14] and later (2000) D. Mandrus. et al. discussed the same observation as metal to insulator transition [15]. In a short review of A B O pyrochlores, Cd Re O is the first superconductor at 1 K [16-18], evident from the heat capacity measurements revealing a clear second and first order phase transition with change in its electrical resistivity at 200 K and 120 K [16]. Consequently, heat capacity of Pr Ru O shows a λ-type divergence at 162 K which is due to the antiferromagnetic ordering of Ru ion in the B-site [19]. Even in the last decade, Hg Ru O was shown to exhibit a metal to insulator transition (MIT) at 108 K by Wilhelm Klein et al. [20] and Ayako Yamamoto [21]. It was observed, in general, for most of the oxide pyrochlores A-site ion is responsible for low temperature magnetic properties and B-site ion is responsible for electrical transition [22]. a. Effect of A-site ionic radii on the electrical property of pyrochlore
The ionic radii of A site in pyrochlore leads to many interesting electrical properties. For example, the iridates, A Ir O show both metallic and insulating nature. If A= Pr , Nd , Sm and Eu whose ionic radii decreases from 1.13 Å to 1.087 Å, metallic nature is exhibited [23-25]. If A= Gd , Tb , Dy and Ho , ionic radii varying from 1.078 Å to 1.041 Å, insulating/semiconducting property is exhibited [25,26]. The ionic radii of A cation decide the bond length and bond angle of Ir-O. This kind of a variation is because the larger A (Pr , Nd , Sm and Eu ) cation promotes the electron transfer via Ir-O-Ir hybridisation that makes the compound metallic. Smaller cation (Gd , Tb , Dy and Ho ) will not promote electron transfer via Ir-O-Ir and therefore the iridates remain insulating, and the larger cation promotes metallic nature [26]. The ionic radii of A cation also play a significant role in Metal to insulator transition: the MIT observed in Nd Ir O (Nd3+: 1.12 Å) is at 36 K, for Sm Ir O (Sm3+: 1.098 Å) and Eu Ir O (Eu3+: 1.087 Å) it is 117 K and 120 K, [25,27,28] respectively, which will be further discussed. b. B-site electronic configuration on the electrical property of pyrochlore
Table I shows some of the pyrochlores and their associated electrical property. It may be noted that, if B-site ion has unpaired electron then the compound exhibits fascinating electrical properties like Metal to Insulator Transition (MIT), Superconducting (SC), Magneto-resistive (MR) and Metal to semiconductor transitions (MST). The compound Tl Mn O is exceptional as it shows MR property due to the hybridisation of 3 d -orbital of Mn with orbital of Tl [29-31]. It should also be noted that Tl- orbital and Mn- orbital energy bands are narrow which aids the MR property [31]. Pyrochlore oxides having or transition metal ions in the B- site usually consociate with MIT along with the magnetic properties of the A-site ion [25]. MIT in the pyrochlore family is relatively due to the electron-electron interaction of B-site ions that play an important role in electrical and magnetic transitions. As discussed earlier, A Ir O (A = Nd, Sm and Eu) compounds were found to exhibit metallic behaviour at low temperatures, while A΄ Ir O (A΄= Gd, Tb, Dy) exhibit semiconducting behaviour [26]. This type of electrical property is due to the hybridisation of oxygen ions of the t energy level of Ir [32]. Nd Ir O , Sm Ir O , Eu Ir O have sharp MIT at 36, 117 and 120 K, respectively. Similarly, Cd Os O , Pr Mo O are also associated with electrical transition in the low temperature regime. Except Cd Os O MIT is peculiarly associated with magnetic ordering, a second order phase transition confirmed from heat capacity measurements [15]. Nd Ir O and Sm Ir O also show MIT in the range of 50 to 150 K, which is due to the hybridisation of 5 d electron of Ir . Even molybdenum pyrochlore (A Mo O ) series also show a sharp transition at low temperatures, due to Mo interaction with oxygen and A-rare earth ion. As mentioned above, a new pyrochlore compound – Hg Ru O also exhibits a MIT with magnetic transition [20,21]. Studies on the single crystal of pyrochlore compounds further confirm MIT with electrical and magnetic transitions, as seen reported in Sm Mo O , Gd Mo O and Ho Mo O [33]. Among all the pyrochlore family’s, A Ti O is exceptional because the compound does not show any sign of electrical transition due to the vacant 3 d orbital of Ti ion. By comparing Ti ion with rest of the B-site electronic configuration as in Ru , Ir , Os and Mo , Ti has vacant d ( ) configuration. Roth et al . in 1956 first reported the synthesis of lanthanide titanates like Ln Ti O (Ln: Sm, Yb and Y) [34] and L.H. brixner in 1964 reported the synthesis of Tb Ti O and other pyrochlore oxides [35]. The electrical resistivity of most of the lanthanide titanates, La Ti O , Nd Ti O , Sm Ti O , Ho Ti O and Er Ti O were measured (by Brixner in 1964) upto 1000°C which showed p - type semiconducting nature at the higher range of temperature [35] . However, the electrical resistivity of Tb Ti O is still considered a fruit under the ice-berg, as interesting properties are further expected in these pyrochlore oxides. Tb Ti O is novel because of the (i) absence of long range ordering even down to 17 mK [6,36] (ii) absence of superconducting behaviour unlike Cd Re O (iii) strong magnetic anisotropy along [111] axis [37] (iv) co-operative Paramagnetism, which is contrary to the long range spin order , a property of spin ice and spin glass [6] (v) presence of glassy behaviour even at 200 mK [38] and (vi) absence of lattice deformation which makes Tb Ti O a perfect pyrochlore [39]. The level of magnetic frustration in a magnetic system is called frustration index ( f), defined as the ratio of Curie Weiss temperature (θ CW ) to critical temperature ( T c ) [40]. For Tb Ti O , f-index is calculated as 157 where θ CW =3000 K and T c = 19 K, which indicates high frustration. Also, Tb Ti O is not associated with any structural transition upto 600 K, suggesting that this compound is a perfect pyrochlore lattice [41]. This is proved by S.W.Han.et.al through the neutron powder diffraction (NPD) studies [41]. The effect of d -electron (B site) in this compound is unexplored, due to the non-degenerate ground state of Tb similar to Ti in Ho Ti O [42]. Pyrochlore irridates and molybendates like R Ir O and R Mo O , where R =Y, Ho, Tb, Dy, Gd, Sm, Eu, Nd, Yb, show the lanthanide effect due to Ir electron density from f and d orbitals that play an important role in magnetic and electrical transitions [25,43]. But the major contribution will be due to f electron. But in Tb Ti O , Ti has empty d -electron density. Therefore, absence of f -orbital electron prompts the investigation of the property of vacant d -electron, in detail. This article provides first information on the formation of Tb Ti O phase at a relatively low synthesis temperature, prepared by high energy ball milling and subsequent firing. a Metal to Insulator Transition in this Ti based pyrochlore at 603 K supported by magnetic, heat capacity and differential scanning calorimetry measurements. a magnetic transition at 696 K, in Tb Ti O II.
Experimental
Preparation of phase pure pyrochlore oxides is difficult due to the formation of secondary oxygen deficient phases like TbTi O and some oxygen rich phases like Tb O . First attempt to prepare Tb Ti O was carried out by Brixner in 1964 using wet ball milling by mixing the corresponding oxides and subsequent firing at 1050 °C for 10-14 hour followed by firing again at 1350 °C for 10 to 14 hour [35]. Thus, the pyrochlore oxides are prepared usually by solid state reaction method by firing the corresponding oxides of rare earth and transition metals in the temperature range of 1350 °C (1623 K). For example, some oxides like Ln Ir O , Ln: Pr, Nd, Sm and Eu, were prepared in sealed tubes for several days (more than 5 days) with intermediate grinding, whereas Tl Mn O was prepared using high pressure techniques [25,34,41,44-47]. In the present study, polycrystalline Tb Ti O is prepared by high energy ball milling (dry medium) followed by firing the powder in a furnace. Tb O (Alfa Aesar, purity – 99.99 %) and TiO (Merck, Purity – 99.9 %) taken in stoichiometry are milled in zirconia vial and balls at 500 rpm for 5 hour in a Fritsch Pulverisette - 7 planetary micro ball mill with a ball to powder ratio of 5:1. The sample is then scraped and subsequently fired at 950 °C (1223 K) in an alumina crucible for 5 hour. The equation representing the reaction is Tb O + 4TiO Ti O The prepared sample is analysed using X-ray diffraction (XRD) study at room temperature (D8 Advance, Bruker) using Cu-K α radiation of wavelength 1.5406 Å with 2θ ranging from 10° to 80° at a scan rate of 0.02° steps per second. The electronic state of Tb and Ti is determined through X-ray photoelectron spectroscopy using Al-K α radiation (omicron nanotechnology) and the low temperature and high temperature magnetisation measurements are performed in a Lakeshore 7410 model – Vibrating sample magnetometer at 1000 Oe and 7000 Oe, respectively. Solatron 1260 impedance/gain phase analyser is used to investigate the electrical property with Pt-probes and heat capacity measurement was performed in Neztech DSC-204F1. III.
Results and discussion a.
Structural analysis – X-Ray diffraction
The crystallographic structure is analysed by Rietveld refinement of the XRD pattern using FULLPROF program. Figure 1 shows the diffraction peaks fitted using Pseudo-Voigt function and the background with 12- coefficient polynomial function. The compound adopts cubic structure with space group
Fd-3m and the unit cell parameters and reliability factors obtained from the fit are tabulated in table II. The refined unit cell parameters are a = b = c = 10.1504(2) Å with the cell volume of 1045 Å , in agreement with the reported values [7,35,41,48,49]. The density of prepared sample is calculated using lattice parameters from XRD data using the relation 𝜌 = ∑ 𝐴𝑁 𝐴 𝑉 Where ρ is density of the compound (g/cm ), ∑ 𝐴 is the product of number of atoms in the unit cell to the atomic weight of the compound, 𝑁 𝐴 – Avogadro number and V is the volume of the unit cell (1045 Å – calculated from XRD). By substituting the value, density is found to be 6.6 g/cm [48,49] which is in agreement with the values reported. The bond lengths and bond angles are tabulated in Table III. Tb ions bond with O1 and O2 oxygen ions in the axial and equatorial position. The axial Tb-O2 (2.508(1) Å) bond length is found to be larger than that of the equatorial Tb-O1 (2.197(2)) Å. This is due to a larger spatial occupancy of Tb towards axial than the other. But in the case of Ti-O2, all the six oxygen atoms are equidistant with a bond length of 1.959(1) Å. This is in agreement with S.W. Han et.al., from the neutron powder diffraction studies, upto the second decimal [41]. b. Electronic Analysis – X-ray Photoelectron spectroscopy
The XPS spectra is shown in figure 2. The survey spectrum in figure 2.a shows the binding energy peaks corresponding to Tb , Ti and O ions. The peak obtained for Ti (fig 2.b) is deconvoluted and the corresponding binding energies obtained are 464 and 458 eV. It is well known that the peak at 464 and 458 eV correspond to Ti oxidation states 2 P and 2 P respectively [50-53]. Figure 2.c shows spectrum of Tb 3 d and 3 d corresponding to 1278.1 and 1241.6 eV respectively. High resolution XPS spectrum of Tb, in fig 2.d, shows the peak corresponding to 150 eV due to the 4 d electron [54]. Therefore, comparison of the XPS binding energy with standard data, confirms Tb, Ti in 3+ and 4+ oxidation states, without any oxygen deficiency. c. Low Temperature Magnetic studies - from Vibrating Sample Magnetometer
The dc-magnetic susceptibility measurement is carried out at a constant magnetic field of 1000 Oe. Figure 3 shows the inverse susceptibility as a function of temperature. The data is fitted using Curie-Weiss law where the parameters were adopted from Gingras et al. ie., C = 23.0 emu K mol -1 , θ CW = -18.9 K [55] and the effective Bohr magnetron (P eff ) is calculated using electronic configuration ( F ) of Tb ion as 9.7μ B per Tb ion and the plot shows a good agreement with experiment and Curie Weiss law. This confirms that the Curie Weiss law describes the experimental data till 50 K as reported by Gingras et al. and the negative Curie temperature indicates the interaction as anti-ferromagnetic in nature [55]. Therefore, dc susceptibility further reinforces the magnetic phase purity of the prepared Tb Ti O sample. d. Transport Analysis – MIT in Tb Ti O The prepared pyrochlore, Tb Ti O is compressed into a pellet of thickness 0.9 mm and diameter of 8 mm and heat treated at 500 °C for 5 hour in order to remove the moisture adsorbed. The pellet is subjected to impedance analysis in the frequency range of 10 MHz to 1 Hz. Dielectric constant, at room temperature, is observed as 40.8 which is very close to the reported values [35]. Resistivity (resistance) of the sample is obtained by fitting Z' vs Z" (real and imaginary part of impedance). Figure 4 (a) and (b) shows normalised resistivity and resistivity as a function of temperature from 403 to 803 K respectively. The plot reveals that I. the resistivity of the sample increases from 403 K to 603 K (i.e.) ρ α T (resistivity is proportional to temperature) - which confirms the metallic/metal nature of the sample II. beyond 603 K, the resistivity of the sample decreases (i.e.) ρ α 1/T (resistivity is inversely proportional to temperature) - which is a property of insulator as discussed in previous sections. Therefore, the plot (figure 4) confirms the metal to insulator transition (MIT) in Tb Ti O at 603 K. Figure 5 (Z" spectra) shows the Z" (imaginary part of impedance) vs log f where a suppression in the peak from 403 K to 603 K is seen confirming the metallic nature of the sample. Beyond MIT (603 K), the peak shifts towards the high frequency regime – which is characteristics of an insulator. The inset of fig 5 shows unity in τ (relaxation time) which decreases by 1/100 th of order beyond MIT. Figure 6 shows the electric modulus spectra from which it is inferred, I. a decrease in magnitude in M" Vs log f till 603 K and II. a shift in peak towards higher range of frequency from 693 K Both (electric and impedance) the modulus spectra show a decrease in magnitude till MIT (603 K) and a shift in peak towards higher frequency beyond MIT. Similar kind of MIT is also found in other pyrochlores like Nd Ir O , Sm Ir O , Eu Ir O ,Gd Ir O , Tb Ir O , Dy Ir O , Ho Ir O , Nd Mo O , Hg Ru O , Tl Ru O and Tl Mn O which is due to d -orbital electron in B-transition metal ions like Ir , Mo , Ru and Mn respectively [21,25,28,56-58]. MIT is one of the common properties exhibited by iridium, molybdenum, ruthenium and manganese pyrochlore families, but no reports are available in Ti based pyrochlores. But the major difference in Tb Ti O is, the temperature of MIT is around 600 K. To the best of our knowledge this is the first report of the metal to insulator transition in a Ti based compound of the pyrochlore family. Even though the pyrochlore family exhibits a series of metal to insulating transitions in the low temperature regime, no reports are available for the transition in the high temperature regime in spin glass, spin ice and spin liquid materials (pyrochlores). The observed MIT in Tb Ti O is due to hybridisation of the vacant 3 d orbital of Ti ion with O ion. Due to hybridisation there is a relative increase in the electron density of Ti (3 d ) similar to other pyrochlore compounds. As a result, the bandgap of the material decreases which in turn decreases the resistivity with rise in temperature. On the other hand, B cation is in six-fold coordination with oxygen due to which the d -state of B cation (empty d- state of Ti ion) show a significant covalent interaction with 2 p state ion. Further, 6p orbital is absent in Tb ion and hence hybridisation of Ti and O ions alone are possible. Therefore, transfer of electrons from 2 p orbital of O to vacant 3 d orbital of Ti ion (figure 7), which has relatively closer energy levels, is responsible for the metallic nature [59] observed. Figure 8 shows the local oxygen environment of Tb, Ti and Tb Ti O Tb and Ti has 8 and 6 oxygen ion co-ordination. At higher temperatures, due to thermal vibration the bond distance between Tb and O1 increases. O1 of Tb repels O2 of Tb thereby decreasing the bond length between Ti and O2- facilitating the electron transport from O2 and Ti ions, which is responsible for the decrease in the resistivity of the compound. Therefore, ρ(T) of Tb Ti O shows a metallic behaviour at T > T MIT . The above discussion is also consistent with the bond length and angle shown by S.W.Han et al. using NPD taken at 45 K. The bond length of Ti and O2 reported is 1.9699 Å at 45 K (Ref no: 38). But in the present work, at 300 K, it is found to be1.9590 Å (table I). Also, the bond angle increases from 132.12° (ref 38) to 132.70° which favours the hybridisation of Ti ion with O ion. Also, the bond length of Tb and O1 increases from 2.1939 Å (ref 38) to 2.1978 Å which supports our claim on MIT. We strongly believe that Ti-O2 bond length further decreases above 600 K that favours the observed MIT. The MIT in Tb Ti O is the only observation so far in the high temperature regime, whereas all other compounds exhibit MIT in the low temperature regime as discussed in the previous sections. The stability of this compound is already verified by Han et.al using neutron powder diffraction and X-ray absorption fine structure and confirm that Tb Ti O is not associated with any structural transition in the range of 4.5 to 600 K. More than these, in order to confirm that the observed MIT is not due to oxygen vacancy at elevated temperatures and to also confirm that the compound is not associated with any secondary phase, the prepared Tb Ti O is subjected to Thermogravimetric analysis (TGA) (inset of fig 6) in the temperature range of 450 K to 800 K. Two affirmative results: (i) no weight loss (from TGA) in the entire range of measurement confirms absence of any oxygen vacancy (ii) no additional peaks in XRD pattern, other than the characteristics, are observed showing absence of secondary phase in the prepared compound. This strongly reinforces, the observed metal to insulator transition we observe in Tb Ti O is also not associated with any structural transition. e. High Temperature – Magnetic studies
The prepared Tb Ti O is subjected to high temperature magnetic measurement from 500 to 800 K at a constant magnetic field of 1000 G (Figure 8). Two major conclusions are drawn: (i) At 600 K there is a small difference in magnetic moment, which is clearly shown in the inset as a first derivative of magnetic moment – the change in magnetic moment is associated with MIT and (ii) above 696 K, the peak is broad, diffused and consistent with the peak shift obtained in M’’ vs log f plot as shown in figure 6. Also, an associated magnetic transition (T M ) at 696 K is seen. Therefore, the high temperature magnetic measurement confirms (i) change in magnetic moment is associated with MIT (603 K) and (ii) another magnetic transition at 696 K due to the transfer of considerable electrons from O ion to Ti . f. Heat Capacity and Differential Scanning Calorimetry
Specific heat capacity measurements in the range from 450 to 800 K is shown in figure 9. A broad hump at 603 K supports the MIT observed in electrical property and another peak at 696 K supports the magnetic transition observed in magnetic measurements. Differential Scanning Calorimetry (DSC) is also performed (a sensitive tool to detect transitions) and is shown in figure 10 that confirms MIT through the sharp peak at 603 K, and another sharp peak at 696 K also confirms the observed magnetic transition. Thus, the Heat capacity and DSC confirm both the (i) MIT at 603 K and (ii) magnetic transition at 696 K.
IV.
Conclusion
To conclude, a cubic pyrochlore Tb Ti O is synthesised by solid state reaction method at a relatively low phase formation temperature effected by the prior high energy ball milling of the precursors. Metal to insulator transition (through electrical resistivity) is found at 603 K. Nyquist plot, relaxation time and electric modulus plots (real and imaginary) confirm MIT. DC magnetic susceptibility and heat capacity measurements prove the MIT observed. The plots are also continuous and exhibit no thermal hysteresis at the transition temperature (T MIT ) - which indicates the second order phase transition. The MIT in Tb Ti O is a consequence of d -orbital unlike other pyrochlore oxides. This is the first report on Ti- based pyrochlore oxides, which shows MIT driven by the empty 3d orbital of Ti4+, and a magnetic transition observed at 696 K. This study opens up new avenues for further research in Ti based pyrochlore oxides. Acknowledgment
BSK thanks DST-Inspire for the award of SRF (IF-140582) and Mr. B. Soundararajan for his support in the measurements. NCNSNT, University of Madras, is acknowledged for XPS measurement. Table 1: Materials belonging to the pyrochlore family exhibiting physical properties including superconductivity (SC), metal to insulator (MIT), magneto-resistance (MR) and semiconductor to metal (SMT) transition with the corresponding B metal electronic configuration and temperature (K).
Compound Electrons of B site ions Property T(K) Ref Cd Re O d SC 1 K [16-18,29] Nd Ir O d MIT 36 K [25] Sm Ir O d MIT 117 K [25] Eu Ir O d MIT 120 K [25] Gd Ir O d MIT 127 K [28] Tb Ir O d MIT 132 K [28] Dy Ir O d MIT 134 K [28] Ho Ir O d MIT 141 K [28] Nd Mo O d MR 90 K [60] Hg Ru O d MIT 108 K [20,21] Tl Ru O d MIT 120 K [57] Tl Mn O d MR 142 K [58] Cd Os O d SMT 225 K [15] Table 2: Structural parameters obtained from Rietveld refinement of XRD pattern of Tb Ti O . χ = 2.68, R p = 3.32, R wp =2.09, R exp = 1.27, a = b = c = Fd-3m . Label Atom Position x y Z
Occupancy Tb1 Tb ½ ½ ½ 1 Ti1 Ti
0 0 0 1 O2 O Table 3: Important bond length and bond angle obtained from Rietveld refinement of XRD pattern of Tb Ti O , (at room temperature). Bond length (Å) Bond angle (°)
Tb1 – O1 2.197 (2) Tb1 – O1 – Tb1 109.38(3) Tb1 – O2 2.508 (1) O2 – Ti1 – O2 132.70(2) Ti1 – O2 1.959 (1) O2 – Ti1 – Tb1 90.00(1) Ti1 – O1 3.777 (2) Tb1 – O2 – Ti1 108.96(3) Ti1 – Ti1 3.589 (2) O1 – Tb1 – O1 180.00(1) O2 – O1 3.621 (2) Figure 1: Rietveld refinement of XRD pattern of polycrystalline Tb Ti O sintered at 950° C for 5 hour. Figure 2: XPS spectra showing (a) survey spectrum of Tb Ti O (b) binding energy spectrum of Ti2p revealing +4 oxidation state of Ti (c) & (d) binding energy spectra of Tb- 3d and 4d confirming the +3 oxidation state of Tb Figure 3: Temperature dependent DC susceptibility for Tb Ti O at a constant magnetic field of 1000 Oe. Figure 4: Variation of resistivity with temperature from 400 K to 800 K-Tb Ti O . The top panel shows normalised resistivity and bottom with magnitude, the metal to insulator (MIT) transition at 603 K. T M indicates the magnetic transition around 696 K. Figure 5: Imaginary spectrum (Z'') as a function of applied frequency - Tb Ti O . Inset shows the difference in relaxation time before and after transition. Figure 6: Electric modulus spectrum vs log (f) of Tb Ti O showing the shift in peak and the corresponding region before and after MIT. (The red arrow indicates the shift in peak towards high frequency regime beyond 703 K). Figure 7: 3 d orbital interaction (Ti) and 2 p orbital (O) possible interactions in Tb Ti O . Figure 8: (a) Local environment of Oxygen ion around Tb, in Tb Ti O . O1 and O2 indicate the oxygen atom at axial and equatorial directions respectively, (b) Ti environment of Oxygen confirms that all the six oxygen ion are equidistant from Ti ion and (c) shows overall picture of Tb and Ti with oxygen. Figure 9: Temperature dependent dc-susceptibility of Tb Ti O . Inset shows the first derivative of magnetic moment with notable change in emu/g at 600 K. Figure 10: Heat capacity of Tb Ti O from 450 to 800 K. The T MIT arrow indicates Metal to insulator transition around 600 K and T M indicate the magnetic transition around 696 K. Inset shows the first derivative of heat capacity, which clearly shows a change in magnitude in the region of 603 K (MIT) and 696 (magnetic transition). Figure 11: Differential scanning calorimetry measurement of Tb Ti O from 450 to 725 K, which clearly shows metal to insulator transition (MIT) and magnetic transition (T M ) References [1] W. F. Giauque and M. F. Ashley, Physical Review , 81 (1933). [2] M. Gingras, B. Hertog, M. Faucher, J. Gardner, S. R. Dunsiger, L. J. Chang, B. D. Gaulin, N. P. Raju, and J. E. Greedan, Phys. Rev. B (2000). [3] S. Rosenkranz, A. P. Ramirez, A. Hayashi, R. J. Cava, R. Siddharthan, and B. S. Shastry, Journal of Applied Physics , 5914 (2000). [4] B. D. Gaulin, J. N. Reimers, T. E. Mason, J. E. Greedan, and Z. Tun, Physical Review Letters , 3244 (1992). [5] M. J. P. Gingras, C. V. Stager, N. P. Raju, B. D. Gaulin, and J. E. Greedan, Physical Review Letters , 947 (1997). [6] J. S. Gardner et al. , Physical Review Letters , 1012 (1999). [7] M. A. Subramanian, G. Aravamudan, and G. V. Subba Rao, Progress in Solid State Chemistry , 55 (1983). [8] P. Schiffer, Nature , 35 (2002). [9] L. Balents, Nature , 199 (2010). [10] J. S. Gardner, M. J. P. Gingras, and J. E. Greedan, Reviews of Modern Physics , 53 (2010). [11] T. Kashima and M. Imada, Journal of the Physical Society of Japan , 3052 (2001). [12] H. Morita, S. Watanabe, and M. Imada, Journal of the Physical Society of Japan , 2109 (2002). [13] B. Kyung and A. M. S. Tremblay, Physical Review Letters , 046402 (2006). [14] A. W. Sleight, J. L. Gillson, J. F. Weiher, and W. Bindloss, Solid State Communications , 357 (1974). [15] D. Mandrus et al. , Physical Review B , 195104 (2001). [16] M. Hanawa, Y. Muraoka, T. Tayama, T. Sakakibara, J. Yamaura, and Z. Hiroi, Physical Review Letters , 187001 (2001). [17] H. Sakai, K. Yoshimura, H. Ohno, H. Kato, S. Kambe, R. E. Walstedt, T. D. Matsuda, Y. Haga, and Y. Onuki, Journal of Physics: Condensed Matter , L785 (2001). [18] R. Jin, J. He, S. McCall, C. S. Alexander, F. Drymiotis, and D. Mandrus, Physical Review B , 180503 (2001). [19] M. Tachibana, Y. Kohama, T. Atake, and E. Takayama-Muromachi, Journal of Applied Physics , 09D502 (2007). [20] W. Klein, R. K. Kremer, and M. Jansen, Journal of Materials Chemistry , 1356 (2007). [21] A. Yamamoto, P. A. Sharma, Y. Okamoto, A. Nakao, H. Aruga Katori, S. Niitaka, D. Hashizume, and H. Takagi, Journal of the Physical Society of Japan , 043703 (2007). [22] H. Guo, C. Ritter, and A. C. Komarek, Physical Review B , 144415 (2017). [23] S. Nakatsuji et al. , Physical Review Letters , 087204 (2006). [24] H. Kumar, R. Chaurasia, P. Kumari, and A. K. Pramanik, AIP Conference Proceedings , 130058 (2018). [25] K. Matsuhira, M. Wakeshima, R. Nakanishi, T. Yamada, A. Nakamura, W. Kawano, S. Takagi, and Y. Hinatsu, Journal of the Physical Society of Japan , 043706 (2007). [26] D. Yanagishima and Y. Maeno, Journal of the Physical Society of Japan , 2880 (2001). [27] J. J. Ishikawa, E. C. T. O'Farrell, and S. Nakatsuji, Physical Review B , 245109 (2012). [28] K. Matsuhira, M. Wakeshima, Y. Hinatsu, and S. Takagi, Journal of the Physical Society of Japan , 094701 (2011). [29] M. A. Subramanian, B. H. Toby, A. P. Ramirez, W. J. Marshall, A. W. Sleight, and G. H. Kwei, Science , 81 (1996). [30] A. P. Ramirez and M. A. Subramanian, Science , 546 (1997). [31] M. Venkatesan, P. Velasco, J. A. Alonso, J. L. Martínez, and J. M. D. Coey, Journal of Physics: Condensed Matter , 3465 (2004). [32] N. Aito, M. Soda, Y. Kobayashi, and M. Sato, Journal of the Physical Society of Japan , 1226 (2003). 6 [33] N. Cao, T. Timusk, N. P. Raju, J. E. Greedan, and P. Gougeon, Journal of Physics: Condensed Matter , 2489 (1995). [34] R. S. Roth, Journal of Research of the National Bureau of Standards , 9 (1956). [35] L. H. Brixner, Inorganic Chemistry , 1065 (1964). [36] J. S. Gardner et al. , Physical Review B , 180401 (2003). [37] A. Sazonov, A. Gukasov, H. Cao, P. Bonville, E. Ressouche, C. Decorse, and I. Mirebeau, (2013). [38] E. Lhotel, C. Paulsen, P. D. de Réotier, A. Yaouanc, C. Marin, and S. Vanishri, Physical Review B , 020410 (2012). [39] O. Ofer, A. Keren, and C. Baines, Journal of Physics: Condensed Matter , 145270 (2007). [40] A. P. Ramirez, Annual Review of Materials Science , 453 (1994). [41] S. W. Han, J. S. Gardner, and C. H. Booth, Physical Review B , 024416 (2004). [42] M. J. Harris, S. T. Bramwell, D. F. McMorrow, T. Zeiske, and K. W. Godfrey, Physical Review Letters , 2554 (1997). [43] N. Ali, M. P. Hill, S. Labroo, and J. E. Greedan, Journal of Solid State Chemistry , 178 (1989). [44] L. G. Shcherbakova, L. G. Mamsurova, and G. E. Sukhanova, Russian Chemical Reviews , 228 (1979). [45] R. Senis, B. Martı ́nez, J. Fontcuberta, W. Cheikh -Rouhou, P. Strobel, P. Bordet, M. Pernet, and M. Nuñez-Regueiro, Journal of Magnetism and Magnetic Materials , 259 (2000). [46] P. Velasco, J. Mira, F. Guinea, J. Rivas, M. J. Martínez-Lope, J. A. Alonso, and J. L. Martínez, Physical Review B , 104412 (2002). [47] J. S. Gardner, B. D. Gaulin, S. H. Lee, C. Broholm, N. P. Raju, and J. E. Greedan, Physical Review Letters , 211 (1999). [48] B. V. I. e. L.M.Ershova, L.N.Kusalova,, V. P. M. E.E.Lomanova, V.M.Tatarintsev,, and a. L.G.Shcherbakova, Izv.Akad.NaukSSSR, Ser. Neorg.Materialy (1977). [49] N. I. T. K.I.Portnoi, S.E.Salibekov, and and I.V.Romanovich, Izv.Akad.NaukSSSR, Ser.Neorg. Materialy , 6 (1970). [50] D. Simon, C. Perrin, and J. Bardolle, Journal de Microscopie et de Spectroscopie Electroniques , 175 (1976). [51] V. I. Nefedov, Y. V. Salyn, A. A. Chertkov, and L. N. Padurets, Zhurnal Neorganicheskoj Khimii , 1443 (1974). [52] C. D. wagner, Riggs, W.M.Davis, L.E.Moulder, J.F.,Mullenberg, G.E, Handbbok of X-ray Photoelectron Spectroscopy (Perkin-Elmer Corporation, Physical Electronics Division, Eden Prairie. MN 55344, 1979). [53] V. I. Nefedov, D. Gatin, B. F. Dzhurinskij, N. P. Sergushin, and Y. V. Salyn, Zhurnal Neorganicheskoj Khimii , 2307 (1975). [54] D. D. Sarma and C. N. R. Rao, Journal of Electron Spectroscopy and Related Phenomena , 25 (1980). [55] M. Gingras, B. Den Hertog, M. Faucher, J. Gardner, S. Dunsiger, L. Chang, B. Gaulin, N. Raju, and J. Greedan, Physical Review B , 6496 (2000). [56] W. Klein, R. K. Kremer, and M. Jansen, Journal of Materials Chemistry , 1356 (2007). [57] T. Takeda, M. Nagata, H. Kobayashi, R. Kanno, Y. Kawamoto, M. Takano, T. Kamiyama, F. Izumi, and A. W. Sleight, Journal of Solid State Chemistry , 182 (1998). [58] Y. Shimakawa, Y. Kubo, and T. Manako, Nature , 53 (1996). [59] N. Cao, T. Timusk, N. P. Raju, J. E. Greedan, and P. Gougeon, Journal of Physics: Condensed Matter , 2489 (1995). [60] Y. Taguchi, Y. Oohara, H. Yoshizawa, N. Nagaosa, and Y. Tokura, Science , 2573 (2001)., 2573 (2001).