We report the structural, dielectric and leakage current properties of Co doped MgTiO3 thin films deposited on platinized silicon (Pt/TiO2/SiO2/Si) substrates by RF magnetron sputtering. The role of oxygen mixing percentage (OMP) on the growth, morphology, electrical and dielectric properties of the thin films has been investigated. A preferred orientation of grains along (110) direction has been observed with increasing the OMP. Such evolution of the textured growth is explained on the basis of the orientation factor analysis followed the Lotgering model. (Mg1-xCox)TiO3 (x = 0.05) thin films exhibits a maximum relative dielectric permittivity of ɛr = 12.20 and low loss (tan δ ∼ 1.2 × 10−3) over a wide range of frequencies for 75% OMP. The role of electric field frequency (f) and OMP on the ac-conductivity of (Mg0.95Co0.05)TiO3 have been studied. A progressive increase in the activation energy (Ea) and relative permittivity ɛr values have been noticed up to 75% of OMP, beyond which the properties starts deteriorate. The I-V characteristics reveals that the leakage current density decreases from 9.93 × 10−9 to 1.14 × 10−9 A/cm2 for OMP 0% to 75%, respectively for an electric field strength of 250 kV/cm. Our experimental results reveal up to that OMP ≥ 50% the leakage current mechanism is driven by the ohmic conduction, below which it is dominated by the schottky emission.

Materials with high dielectric permittivity (ɛr ≥ 10) and extremely low losses (tan δ ≤ 10−4) in the form of bulk and thin film nanostructures are playing key role in the development of dynamic random access memories (DRAMs), micro and millimeter wave ICs, and low loss dielectric resonators.1–3 In particular, dielectric systems with high temperature and mechanical stability in the microwave regime are the main requirements for hybrid ICs to use in the global telecommunication technology.4,5 Among many dielectric materials; MgTiO3 (MTO) is one of the best known microwave systems, which exhibits high quality factor Q = |$\frac{1}{{tan\, \delta }}$|1tanδ ≈ 20,000 at 8 GHz frequency, negative temperature co-efficient of resonant frequency τf ≈ −50 ppm/°C and relative dielectric permittivity ɛr ∼ 17.6 Lee et al. reported that dilute substitution (x ∼ 0.05) of Co2+ at Mg2+ site can improve the dielectric properties of MgTiO3 ceramics significantly (ɛr ∼ 16.8, Q × f0 ∼ 244 THz, τf ∼ −54 ppm/°C).7 Also, MTO thin films were used as buffer layer for high purity LiNbO3 grown epitaxially on the c-axis oriented Al2O3 substrates for the applications of integrated optical devices.8 Surendran et al. reported that Zn2+ and Ni2+ substitution into Mg2+ sites in MTO thin films significantly increases the dielectric permittivity and reduces the losses. The corresponding values being tan δ ≈1.1 × 10−2, 1.9 × 10−2 for 5 at % Ni and Zn substitution in MgTiO3.9 Pure MTO thin films prepared by soft chemistry method display the temperature co-efficient of dielectric constant (TCK) ∼ +260 ppm/°C.10 Ho et al. studied the effect of crystal orientation on the photoluminescence spectra of pure MTO thin films deposited on P-type Si(111) oriented substrates. They reported that at 800 °C annealing temperature MTO grain growth occur along (003) orientation at an argon to oxygen ratio of 60/40 sccm.11 However, MgTiO3 thin films grown by PLD method on c-axis oriented Al2O3 substrates are highly oriented along (003) direction and exhibits very high dielectric constant 24 at 1 MHz frequency.12 

Recently, Huang et al.13 reported the electrical properties of the MgTiO3 based metal-insulator-metal (MIM) devices at various temperatures. Consequently the MIM capacitor fabricated at 200 °C with 210 nm thick MgTiO3 shows very high density of capacitance ∼ 1.2 nF/μm2 and low leakage current 1.51 × 10−9 A/cm2 at 5V. Nevertheless, up to now no report is available in the literature on the Co doped MgTiO3 system in the form of 2D thin film nanostructures. Thus, an attempt has been made to investigate the dielectric and leakage constant studies of Co doped MgTiO3 thin films deposited on Pt/TiO2/SiO2/Si substrates using RF magnetron sputtering. This is the continuation of our previous work, on (Mg1-xCox)TiO3 (x = 0−0.07) bulk ceramics synthesized by semi alkoxide precursor method.14 In the present work, we report the growth and characterization of (Mg0.95Co0.05)TiO3 (MCT) thin films under different oxygen mixing percentage (OMP). We report for the first time (110) oriented growth of MCT films at higher OMP (75%) and improved dielectric properties (ɛr ∼12.2), low leakage currents (J = 1.14 × 10−9 A/cm2) and losses (tan δ = 1.2 × 10−3) in 5 at % Co containing MCT thin films deposited by RF magnetron sputtering.

RF magnetron sputtering method has been employed to deposit the (Mg0.95Co0.05)TiO3 (MCT) thin films on platinized silicon (Pt/TiO2/SiO2/Si) substrates at 300 °C deposited at a fixed RF power of 40 W. Initially a sputtering target material of (Mg0.95Co0.05)TiO3 is prepared by semi alkoxide precursor method detailed methodology is reported elsewhere.14 The deposition chamber is maintained up to a base pressure of 1.0 × 10−6 m bar before the deposition. The substrate to target distance was kept at 5 cm. A mixture of high purity argon (99.99%) and oxygen (99.99%) and is then introduced in to chamber using mass - flow controllers. The sputtering pressure of 1 × 10−2 mbar is maintained constantly throughout the deposition process by varying the oxygen mixing percentage (OMP) from 0−100% in the sputtering gas. The target was pre-sputtered in argon ambient for 10 minutes to clean the surface. The rate of deposition under different OMPs has been optimized to achieve the constant thickness of the film. The thicknesses of all the reported films are of the order of 280 ± 10 nm. As deposited MCT thin films are post annealed at 700 °C for 1 hour to obtain good crystallinity.

A Rigaku high power XRD machine with CuKα radiation (λ  =  1.5406 Å) has been employed to study the structural characterization. The X-ray diffraction pattern of sputtering target was refined by using the Rietvield refinement technique and fullprof program.15 The surface morphology of the thin films has been analyzed by using a atomic force microscope (Model 5250, Agilent series) under constant force non contact mode. For dielectric characterization metal-insulator-metal (MIM) electrodes in the form of a typical capacitor geometry has been designed on the top surface and bottom of the samples. Consequently, Aluminum top electrodes of diameter 0.8 mm and 100 nm thick were deposited by thermal evaporation on to the MCT thin films using a shadow mask technique. An LCR meter (Wayne Kerr Electronics Pvt. Ltd., Model 1J43100) has been used to measure the dielectric properties in the frequency range of 100 Hz to 106 Hz. The leakage current measurements have been performed using a parameter analyzer (Keithely 4200 Semiconductor Systems).

The x-ray diffraction pattern along with the Rietveld refinement data of (Mg0.95Co0.05)TiO3 sputtering target, sintered at 1200 °C for 3 hours is shown in Fig. 1. The refinement was carried out by considering |$R\bar 3$|R3¯ space group.16 The lattice parameters, atomic positions of the Mg, Ti, Co and O atoms, and occupancy are refined. The corresponding estimated values being a = b = 5.053 ± 0.001 Å, c = 13.898 ± 0.002 Å. The fitting parameters (i) χ2≈ 2.09, (ii) RBrag factor ≈ 2.86 and (iii) Rf factor ≈ 3.08 for MCT target system.

FIG. 1.

X-ray diffraction pattern after the Rietveld refinement of (Mg0.95Co0.05)TiO3 target material, sintered at 1200 °C for 3 hours in air.

FIG. 1.

X-ray diffraction pattern after the Rietveld refinement of (Mg0.95Co0.05)TiO3 target material, sintered at 1200 °C for 3 hours in air.

Close modal

On the other hand, thin films deposited at 300 °C under various oxygen mixing percentages depicts amorphous nature. However, after annealing at 700 °C for 1 hour in air all the thin films exhibit rhombhohedral crystal structure, which is consistent with the earlier reports.17 Fig. 2 shows the x-ray diffraction pattern of post annealed MCT thin films deposited under various oxygen mixing percentages. Significant changes has been noticed in the peak intensities of (104), (110) and (003) reflections. The peak intensities of (003), (012), (211) and (104) reflections suppressed while (101), (110), (013) and (024) peaks were enhanced with increasing in OMP up to 75% and decreases beyond. The variation in the peak intensities of (104) and (110) reflections as a function of oxygen mixing percentage is shown in Fig. 3. One can clearly observe an opposite trend of peak intensities with an increase in OMP level. The peak intensity of (110) increases progressively up to 75% of OMP confirming the preferred orientation of grains along (110) direction. Such kind of preferred orientational growth with the increase of oxygen content has also been observed in various systems also.18–20 The preferred oriental growth can be explained on the basis of Lotgering analysis21 of orientation factor (F) with respect to a reference plane (abc) defined as

\begin{equation}F = \frac{{\left( {P - P_0 } \right)}}{{\left( {1 - P_0 } \right)}}\end{equation}
F=PP01P0
(1)
\begin{equation}P = \sum {I_{\left( {abc} \right)} } /\sum {I_{\left( {hkl} \right)} } \end{equation}
P=Iabc/Ihkl
(2)

where abc denotes the Miller indices, P represents the ratio of sum of intensities of the investigated reflections (abc) to the sum of all reflections of the textured thin film. The value “P0” stands for the equivalent ratio for the ceramic power of the target with random orientation. Usually, the value of the orientation factor for preferred orientation should lie in the range 0 to1. Thus, F = 0 denotes a film with randomly oriented grains, while F = 1 denotes a perfect epitaxial growth. In the present case, all the diffraction peaks lie in the range 2θ = 15°–60° has been used to calculate the P and P0. The P0 values obtained from the MCT sputtering targets are found to be 0.30 and 0.12 for the (104) and (110) planes, respectively. Consequently, the calculated orientation factors for the (104) and (110) reflections of the MCT thin films are plotted as a function of oxygen mixing percentage which is shown in Fig. 4. For lower OMP values the orientation factor for both (104) and (110) reflections turns out to be close to the 0 indicating that the films grown under pure Ar ambient are randomly oriented. With increase in OMP, the F(104) goes to negative value and F(110) increases continuously, which indicates that the MCT thin films grown under high OMP is textured along (110) direction. The maximum value of F(110) ≈ 0.32 is observed for 75% OMP beyond which F(110) values starts decreasing. Similar results were reported by C. L. Huang et al. previously in the case of highly (110) oriented pure MgTiO3 thin films prepared by using RF magnetron sputtering.22,23

FIG. 2.

X-ray diffraction patterns of post annealed (Mg0.95Co0.05)TiO3 thin films at 700 °C deposited on platinized silicon substrates at different oxygen mixing percentage (OMP). The peak with asterisk mark indicates the reflection from Si substrates.

FIG. 2.

X-ray diffraction patterns of post annealed (Mg0.95Co0.05)TiO3 thin films at 700 °C deposited on platinized silicon substrates at different oxygen mixing percentage (OMP). The peak with asterisk mark indicates the reflection from Si substrates.

Close modal
FIG. 4.

The dependence of orientation factor (F) of MCT thin films as a function of oxygen mixing percentage.

FIG. 4.

The dependence of orientation factor (F) of MCT thin films as a function of oxygen mixing percentage.

Close modal
FIG. 3.

The variation of XRD peak intensity of the (104) and (110) reflections as a function of oxygen mixing percentage.

FIG. 3.

The variation of XRD peak intensity of the (104) and (110) reflections as a function of oxygen mixing percentage.

Close modal

Fig. 5 shows the atomic force microscopic images of as deposited and after annealed MCT thin films deposited under different OMP. Different surface morphologies and roughness have been noticed with the variation in OMP and annealing conditions. Thin films deposited at 300 °C, could enhances the surface diffusion during deposition and forms small and uniform grains. The root means square (rms) surface roughness of as- grown MCT films at 25% and 75% OMP being 8.8 nm and 7.5 nm respectively. However, the post annealed films shows bigger grains with pronounced grain boundaries due to diffusion. With increasing the oxygen content the rms surface roughness of the films decreases from 15.3 nm to 9.3 nm. Under pure Ar ambient one can expect an abnormal grain growth and high surface roughness as compared to the oxygen environment. This is due to large momentum transfer process (heavier Ar atoms) as compared to the lighter gas such as oxygen where relatively less rms roughness sputtering yields can be expected.24 

FIG. 5.

Atomic force microscopic (AFM) images of the (Mg0.95Co0.05)TiO3 thin films deposited on platinized silicon substrates. As grown thin films at (a) 25% OMP, (b)75% OMP. After annealing at (c) 25% OMP (d) 75% OMP.

FIG. 5.

Atomic force microscopic (AFM) images of the (Mg0.95Co0.05)TiO3 thin films deposited on platinized silicon substrates. As grown thin films at (a) 25% OMP, (b)75% OMP. After annealing at (c) 25% OMP (d) 75% OMP.

Close modal

Fig. 6 shows the frequency variation of imaginary part of impedance (−Z″) recorded at different temperatures in the range 300 K–493 K for 25% OMP of MCT. For T < 383 K, Z″ exhibit a very high magnitude which decreases monotonically with increase in ac- driving frequency (f). As the temperature increases beyond 383 K, a cusp like behavior is observed in the Z″(f). The maximum values of Z″ occurs at 2.1 × 103 Hz, 3.1 × 103 Hz, 4.1 × 103 Hz, and 5.1 × 103 Hz at 413 K, 443 K, 468 K and 493K, respectively. The peak value of Z″max shifts towards the higher frequency side with decrease in the magnitude of the peak implying that the electrical relaxation phenomenon is thermally activated.25 For frequencies above 25 kHz all the curves merge into single curve at the investigated temperature range. The relaxation process may be due to the presence of immobile species at low temperature and may be due to the formation of defects at high temperature.

FIG. 6.

Frequency dependence of the imaginary component of impedance (−Z″) spectra recorded at different temperature of (Mg0.95Co0.05)TiO3 thin films deposited at 25% OMP. The inset shows the Nyquist plots (Cole- Cole graphs) of (Mg0.95Co0.05)TiO3 thin films deposited at 25% OMP.

FIG. 6.

Frequency dependence of the imaginary component of impedance (−Z″) spectra recorded at different temperature of (Mg0.95Co0.05)TiO3 thin films deposited at 25% OMP. The inset shows the Nyquist plots (Cole- Cole graphs) of (Mg0.95Co0.05)TiO3 thin films deposited at 25% OMP.

Close modal

The inset of Fig. 6 shows the Nyquist plots of complex impedance spectrum of MCT thin films deposited at 25% OMP, measured at different temperatures. At low temperature range 300–383 K the plots of Z′ vs. −Z″ linearly line up towards the −Z″ axis indicating the higher resistance of the samples. With increasing the temperature above 383 K, these curves become almost semi circular arcs with non-zero high frequency intercept. Generally, the semicircle behavior of complex impedance plots can be explained on the basis of an equivalent circuit model as shown in the inset of Fig. 7. In such equivalent circuit model, the high frequency intercept provides the value of the series resistance and the magnitude of semi circle diameter gives the electrical dc-resistivity of the sample at specified temperature. The maximum value corresponds to the relaxation frequency ω = 1/RC. For the present −Z″ and Z′ Cole-Cole configuration two parallel RC combination circuit [(RgCg) (RgbCgb)] elements connected in series serves as the equivalent circuit. The semi circle arc at low frequency circuit corresponds to grain boundary resistance (Rgb) and the high frequency arc indicates the contrition from grain (Rg). Fig. 7 shows the Z′ vs. −Z″ plots after fitting experimental data with equivalent circuit model. The open symbol represents the fitted data while the solid symbol shows the experimental data. The calculated Rg and Cg values were given in Table I. Using these values the relaxation time (τ) is calculated by substituting it in the equation: τ = 1/ω = RgCg. The inset of Fig. 7 shows the variation of ln τ as a function of 1000/T. All these plots show that the data is following the Arrhenius law.

\begin{equation}\tau = \tau _0 \exp \left( {\frac{{E_a }}{{k_B T}}} \right)\end{equation}
τ=τ0expEakBT
(3)

where τ0 is the prefactor, Ea is the activation energy for the response and kB is Boltzmann constant. From the linear fit, we have obtained the slope value which in turn gives the activation energy (Ea) ∼ 0.14 eV of MCT thin films deposited 25% OMP.

FIG. 7.

Experimental and calculated Nyquist fittings of (Mg0.95Co0.05)TiO3 thin films deposited at 25% OMP at different temperatures for the equivalent circuit model given in the inset. Solid symbol indicates experimental data while open symbol indicates fitted data. The second inset shows the variation of relaxation time (τ) as a function of 1000/T.

FIG. 7.

Experimental and calculated Nyquist fittings of (Mg0.95Co0.05)TiO3 thin films deposited at 25% OMP at different temperatures for the equivalent circuit model given in the inset. Solid symbol indicates experimental data while open symbol indicates fitted data. The second inset shows the variation of relaxation time (τ) as a function of 1000/T.

Close modal
Table I.

The values of grain resistance (Rg) and capacitance (Cg) of 25% OMP deposited MCT films measured at different temperature.

Temperature (K)Grain resistance Rg (kΩ)Grain capacitance Cg (pF)
413 1049.2 58.29 
443 910.5 58.89 
468 682.6 59.63 
493 541.4 60.34 
Temperature (K)Grain resistance Rg (kΩ)Grain capacitance Cg (pF)
413 1049.2 58.29 
443 910.5 58.89 
468 682.6 59.63 
493 541.4 60.34 

Fig. 8 shows the ac-conductivity (σac) as a function of frequency measured at various temperatures for MCT film deposited at 25% OMP. The dependence of ac-conductivity on the frequency is generally described using Jonscher power law.26 

\begin{equation}\sigma \left( \omega \right) = \sigma \left( 0 \right) + A\omega ^s \end{equation}
σω=σ0+Aωs
(4)

where σ(ω) is the total conductivity of the system, σ(0) is the contribution from dc conductivity (frequency independent part), ω is the angular frequency of the ac signal (ω = 2πf) and “s” and “A” are the characteristic parameters. Both “A” and “s” are temperature dependent parameters. The term Aωs contains the ac dependence and characterizes all dispersion phenomenons.27–30 For T < 383 K, the conductivity almost varies linearly with the frequency and no dc-plateau region has been observed within the frequency range 100–106 Hz. On the other hand, above 383 K, frequency independent conductivity (plateau) is observed at low frequencies (f < 2 kHz), due to the contribution of dc conductivity mechanism. Beyond a critical frequency the conductivity follows linearly which is expected due to the contribution of grains.31 The values of exponent “s” and pre-exponent factor “A” for various temperatures above 383 K and are enlisted in Table II. It is observed that, with increasing temperature, the value of “s” increases where as log A decreases significantly. Usually, the value of “s” should lie between 0 and 1. In the present case slightly higher values of “s” than the unity has been noticed which are typically observed in chalcogenide materials.32 It is expected that the higher values of “s” are the indicative of more complicated frequency dependent behavior of the conductivity. Accordingly, it is likely that frequency dependence of conductivity is following the double power law behaviour

\begin{equation}\sigma '\left( \omega \right) = \sigma _0 \left( {dc} \right) + B_1 \omega ^{s_1 } + B_2 \omega ^{s_2 } \end{equation}
σω=σ0dc+B1ωs1+B2ωs2
(5)

(as proposed by Funke and Barranco et al.).33–35 In the above equation the first exponent s1 characterizes the low frequency region connected with the translational hopping of ions. While the second exponent characterizes the high frequency region due to the contribution of localized charge carrier conductivity of the reorientation ionic hopping.36 The inset of Fig. 8 shows the logarithmic variation of conductivity as a function inverse of the temperature measured at constant frequency f ∼ 1 kHz. The linear behaviour of these plots indicates that the dc conductivity follows Arrhenius relation given by

\begin{equation}\sigma _{dc} = \sigma _0 \exp \left( {\frac{{ - E_a }}{{k_B T}}} \right)\end{equation}
σdc=σ0expEakBT
(6)

The activation energy (Ea) calculated from the linear portion of the curve results Ea ≈ 0.132 eV which is in excellent agreement with those values obtained from the relaxation time analysis as discussed in previous section. The oriented growth along (110) direction in-turn causes activation energy to increase from 0.14 to 0.4eV for 25 to 75% OMP respectively.

FIG. 8.

The ac-conductivity (σac) versus frequency (f) at different temperatures. Inset of this figure shows the variation of lnσdc versus 1000/T plot of (Mg0.95Co0.05)TiO3 the films deposited at 25% OMP.

FIG. 8.

The ac-conductivity (σac) versus frequency (f) at different temperatures. Inset of this figure shows the variation of lnσdc versus 1000/T plot of (Mg0.95Co0.05)TiO3 the films deposited at 25% OMP.

Close modal
Table II.

The evaluated values of exponent factor “s” and pre-exponent factor log A measured at different temperatures for the films deposited at 25% OMP.

Temperature (K)413443468493
s 1.39 1.45 1.59 1.66 
logA −9.81 −10.10 −10.79 −11.32 
Temperature (K)413443468493
s 1.39 1.45 1.59 1.66 
logA −9.81 −10.10 −10.79 −11.32 

Figs. 9(a) and 9(b) shows the frequency variation of the relative dielectric permittivity constant (ɛr) and loss tangent (tan δ) of the MCT thin films deposited at different OMP recorded at 300 K. The obtained relative dielectric permittivity and loss tangent values lies in the range 19.5–8.56 and 0.06–0.0012, respectively. These ɛr values are higher than ɛr = 14.5 reported by Chen et al.37 in the case of pure MgTiO3 thin films deposited on n-type silicon substrates using RF magnetron sputtering under O2 rich ambient. These values are still lower than the values (ɛr ≈ 24 at 1MHz) reported by Kang et al. who deposited (003) oriented MgTiO3 on c- axis oriented Al2O3 substrates using PLD.12 Usually dispersion in ɛr with the frequency can be explained on the basis of Maxwell–Wagner approach and koop's theory.38,39 These models demonstrate that as the frequency increases, the probability of electrons reaching the grain boundary decreases which in turn results in the reduction of polarization finally leading to a significant decrease of ɛr.40 It is possible that the interfacial dead layers may exist across the interface of electrodes and the thin film which in turn leads to dielectric losses. The inset of Fig. 9(b) shows the variation of the ɛr and tan δ as a function of different values of OMP, measured at 106 Hz. As the OMP increases from 0% to 75%, the ɛr values increases progressively up to 75% OMP, beyond which it starts decreasing. Such variation in the dielectric properties with OMP strongly suggests the role of oxygen vacancies and crystallographic orientations.41,42 The improvement in dielectric properties of the films deposited up to 75% OMP can be attributed to the following factors: (i) the preferred orientation of grains along (110) direction, may enhance the polarizability and increase in the degree of crystallinity of the films. (ii) From the AFM images, it was clear that the films were highly densified with a uniform grain size, and (iii) films deposited below 75% OMP have some oxygen vacancies and can be explained according to the equation |$O_0 \leftrightarrow V_0^{ \bullet \bullet } + 2e^\prime + 1/2O_2$|O0V0+2e+1/2O2 where O0, |$V_0^{ \bullet \bullet }$|V0, and e′ represents the oxygen ion on its normal site, oxygen vacancy, and electron, respectively. The oxygen vacancies may cause significant changes in the dielectric behaviour of the deposited films. However, for the films deposited at 75% OMP, the number of oxygen vacancies reduces, which causes the improvement in the dielectric properties.42–44 In the present case, the thin films deposited in pure oxygen atmosphere (OMP > 75%) may result in the target poisoning during sputtering, which may cause the reduction in the rate of deposition as a result smaller grain size and lower densities occur. Further, the preferred orientation in films along (110) is less as compared to the films deposited at 75% OMP. These factors may significantly deteriorate the dielectric properties of the films deposited in pure oxygen atmosphere. In the present case the optimum values of ɛr and tan δ being 12.20 and 0.0012 were obtained for the films deposited at 75% OMP and 700 °C for 1 hr.

FIG. 9.

(a,b) The variation of relative dielectric permittivity (ɛr) and loss tangent (tan δ) of MCT thin films deposited at different OMP. The inset of figure 9(b) shows the variation of ɛr and tan δ for different values of OMP measured at 106 Hz.

FIG. 9.

(a,b) The variation of relative dielectric permittivity (ɛr) and loss tangent (tan δ) of MCT thin films deposited at different OMP. The inset of figure 9(b) shows the variation of ɛr and tan δ for different values of OMP measured at 106 Hz.

Close modal

Fig. 10 shows the temperature dependence of ɛr and tan δ for 0% and 75% of OMP of MCT thin film measured at f = 106 Hz in the temperature range 300–433 K. Both relative dielectric permittivity and tan δ linearly increases with temperature. Usually, at low temperatures, the molecules cannot orient themselves in polar dielectrics.45 But as the temperature increases, the orientation of dipoles is facilitated, and cause increase in the relative dielectric permittivity.45 We have estimated the temperature stability of dielectric permittivity by evaluating the temperature coefficients of dielectric constant (TCK) of MCT thin films. The TCK was determined using the following equation,45,46

\begin{equation}TCK = \frac{{\Delta \varepsilon _r }}{{\varepsilon _0 \Delta T}}\left( {ppm/^ \circ {\rm C}} \right)\end{equation}
TCK=Δɛrɛ0ΔTppm/C
(7)

where Δɛr is the change in ɛr with respect to the value ɛ0 (T = 300 K) and ΔT is the change in temperature relative to 300 K. The inset of Fig. 10(b) shows the variation of TCK as a function of oxygen mixing percentage. From this figure one can say that all the MCT films show a positive TCK values across 106 Hz. It is evident that as the oxygen availability increases the TCK of MCT films decreases up to 75% and starts increasing for OMP > 75%. The obtained TCK values lies in the range 350–175 ppm/°C. It is clear from the figure, that the 75% OMP gives better thermal stability (minimum TCK) of the dielectric properties.

FIG. 10.

Temperature dependence of relative dielectric permittivity (ɛr) and loss tangent (tan δ inset (i)) recorded at a constant ac driving frequency of 106 Hz for 0 and 75% OMP. The inset (ii) shows variation of temperature co-efficient of relative dielectric permittivity (TCK) as a function of OMP measured at constant f = 106 Hz.

FIG. 10.

Temperature dependence of relative dielectric permittivity (ɛr) and loss tangent (tan δ inset (i)) recorded at a constant ac driving frequency of 106 Hz for 0 and 75% OMP. The inset (ii) shows variation of temperature co-efficient of relative dielectric permittivity (TCK) as a function of OMP measured at constant f = 106 Hz.

Close modal

On the other hand, the leakage current characteristics show relatively low leakage current density (J = 1.14 × 10−9 A/cm2 at 250 kV/cm2 for OMP = 75%) for the investigated thin films (Fig. 11). Chen et al. reported the leakage current characteristics of pure MgTiO3 thin films deposited on n-type Silicon substrates. They attributed that the low leakage current (I ≈ 1.3 × 10−8 A/mm2) at 75% Ar/O2 ratio) is due to the improvement of grain sizes (less defects) due to the increased annealing temperatures.37 Usually the leakage current density depends on the grain sizes, the average crystallinity, and the rms surface roughness.47,48 We attribute that the decrease in leakage current density with increase in OMP is due to the lower rms surface roughness and more structured grain boundaries in the MCT thin films. In the present case the value of slope for the leakage current versus electric field curves are close to unity (1.08 for 50% OMP and 1.12 for 75% OMP), which indicates that the leakage mechanism is considered to be due to ohmic conduction.49 For OMP values near to 50 and 75% the J versus E curves exhibits linear behaviour confirming the ohmic conduction mechanism. However for OMP values < 50% and more than 75% no linear behavior of J(E) is observed indicating that the ohmic conduction is not responsible for the leakage current in MCT thin films. These dielectric properties of (Mg0.95Co0.05)TiO3 thin films can be suitable gate oxide material for CMOS applications.

FIG. 11.

The I-V characteristics of the (Mg0.95Co0.05)TiO3 films deposited at different oxygen mixing percentage. The inset shows the leakage current density versus electric field of 50%, 75% OMP deposited thin films and its corresponding fits.

FIG. 11.

The I-V characteristics of the (Mg0.95Co0.05)TiO3 films deposited at different oxygen mixing percentage. The inset shows the leakage current density versus electric field of 50%, 75% OMP deposited thin films and its corresponding fits.

Close modal

(Mg0.95Co0.05)TiO3 (MCT) thin films grown on platinized silicon substrates by RF magnetron sputtering at various oxygen mixing percentage (OMP) has been investigated. For OMP approximately 75% provides improved dielectric properties as compared to pure MTO. The structural analysis reveals that all the grains are orientated along (110) preferred direction as OMP increases from (0–75%) with reduced rms roughness (9.3 nm). The MCT films prepared under 75% OMP provides high dielectric permittivity (ɛr = 12.2), low losses (0.0012) and low temperature co-efficient of relative dielectric permittivity (175 ppm/°C) due to the preferred orientation of grain growth along (110) direction. The frequency dependant ac conductivity follows the Jonscher power law, σ(ω) = σ(0) + Aωs. The activation energies estimated from the Arrhenius relation |$\tau = \tau _0 \exp ( {\frac{{E_a }}{{k_B T}}})$|τ=τ0exp(EakBT) are in good agreement with those estimated (0.14 eV for 25% OMP) from dc conductivity analysis. The oriented growth along (110) direction in-turn causes activation energy to increase from 0.14 to 0.4 eV for 25 to 75% OMP respectively. The leakage current analysis (J ∼ 1.14 × 10−9 A/cm2 at 250 kV/cm) reveals that the ohmic conduction mechanism is dominated in MCT films prepared at 50–75% OMP.

This work has been supported by the Board of Research in Fusion Science & Technology (BRFST, India) of National Fusion Programme under the project NFP-RF-A12-01. The authors P. G., S. T. and D. P. acknowledge the financial support from BRFST project NFP-RF-A12-01. D. P. greatly acknowledges the research facilities provided by DRDO [ERIP/ER/0900371/M/01/1264], DAE-BRNS [2010/20/37P/14BRNS], and DST [SR/FTP/PS-109/2009]. T. S. K. acknowledges the infrastructure facility of XRD provided by DST, New Delhi, through FIST program [SR/FST/PSII-020/2009]. P. G. acknowledges Dr. D. K. Goswami, Department of Physics, IIT Guwahati, for providing leakage current measurement facilities.

1.
A. I.
Kingon
,
J. P.
Maria
, and
S. K.
Streiffer
,
Nature
406
,
1032
(
2000
).
2.
W.
Jackson
,
Nature
158
,
547
(
1946
).
3.
O.
Auciello
,
Science and Technology of Electroceramic Thin Films
(
Kluwer Acadamic Publishers
,
Netherlands
,
1995
).
4.
O. G.
Vendik
,
E. K.
Hollmann
,
A. B.
Kozyrev
, and
A. M.
Prudan
,
J. Superconductivity
12
,
235
(
1999
).
5.
A. K.
Tagantsev
,
V. O.
Sherman
,
K. F.
Astafiev
,
J.
Venkatesh
, and
N.
Setter
,
J. Electroceramics
11
,
5
(
2003
).
6.
V. M.
Ferreira
,
J. L.
Baptista
,
S.
Kamaba
, and
J.
Petzdt
,
J. mater. Sci.
28
,
5894
(
1993
).
7.
L.
Li
,
X.
Ding
, and
Q.
Liao
,
J. alloys compd.
509
,
7271
(
2011
).
8.
J.
Lee
and
C. W.
Choi
,
Jpn. J. Appl. Phys.
38
,
3651
(
1999
).
9.
K. P.
Surendran
,
A.
Wu
,
P. M.
Vilarinho
, and
V. M.
Ferreira
,
J. Appl. Phys.
107
,
114112
(
2010
).
10.
B. D.
Lee
,
K. H.
Yoon
,
E. S.
Kim
, and
T. H.
Kim
,
Jpn. J. Appl. Phys.
42
,
6158
(
2003
).
11.
Y. D.
Ho
and
C. L.
Huang
,
J. Am. Ceram. Soc.
96
,
2065
(
2013
).
12.
S.
Kang
,
W.
Lim
, and
J.
Lee
,
Integrated Ferroelectrics
31
,
97
(
2000
).
13.
C. L.
Huang
,
S. Y.
Wang
,
Y. B.
Chen
,
B. J.
Li
, and
Y. H.
Lin
,
Current Appl. Phys.
12
,
935
(
2012
).
14.
T.
Santhosh Kumar
,
P.
Gogoi
,
A.
Perumal
,
P.
Sharma
, and
D.
Pamu
,
J. Am. Ceram. Soc.
(
2014
).
15.
R. A.
Young
,
The Rietveld Method, International Union of Crystallography
(
Oxford University Press
,
London
,
1996
).
16.
B. A.
Wechsler
and
R. B.
Dreele
,
Acta. Cryst.
B45
,
542
(
1989
).
17.
L.
Li
,
X.
Ding
, and
Q.
Liao
,
Ceramics International
38
,
1937
(
2012
).
18.
Y. E.
Lee
,
J. B.
Lee
,
Y. J.
Kim
,
H. K.
Yang
,
J. C.
Park
, and
H. J.
Kim
,
J. Vac. Sci. Technol. A
14
,
1943
(
1996
).
19.
C.
Wang
,
B. L.
Cheng
,
S. Y.
Wang
,
H. B.
Lu
,
Y. L.
Zhou
,
Z. H.
Chen
, and
G. Z.
Yang
,
Thin Solid Films
485
,
82
(
2005
).
20.
H. L.
Chen
and
Y. S.
Yan
,
Thin Solid Films
516
,
5590
(
2008
).
21.
F. K.
Lotgering
,
J. Inorg. Nucl. Chem.
9
,
113
(
1959
).
22.
C. L.
Huang
and
C. L.
Pan
,
J. Vac. Soc. Technol. A
22
,
2440
(
2004
).
23.
C. L.
Huang
and
Y. B.
Chen
,
Jpn. J. Appl. Phys.
44
,
6736
(
2005
).
24.
D. R.
James
,
Optical Thin Films
, (
SPIE
,
Washington
,
1987
).
25.
B.
Behara
,
P.
Nayak
, and
R. N. P.
Choudhary
,
J. Alloys compd.
436
,
226
(
2007
).
26.
A. K.
Jonscher
,
Nature
264
,
673
(
1977
).
27.
A.
Rouahi
,
A.
Kahouli
,
A.
Sylvestre
,
E.
Defay
, and
B.
Yangui
,
J. Alloys Compd.
529
,
84
(
2012
).
28.
S.
Upadhyay
,
O.
Parkash
, and
D.
Kumar
,
J. Phys. D: Appl. Phys.
37
,
1483
(
2004
).
29.
S.
Thota
,
A.
Kumar
, and
J.
Kumar
,
Mater. Sci. Eng. B
164
,
30
(
2009
).
30.
L.
Zhang
,
Appl. Phys. Lett.
87
,
022907
(
2005
).
31.
B.
Louati
,
M.
Gargouri
,
K.
Guidara
, and
T.
Mhiri
,
J. Phys. & Chem. Solids
66
,
762
(
2005
).
32.
S. A.
El Hakim
,
F. A.
El Whab
,
A. S.
Mohamed
, and
M. F.
Kotkata
,
Phys. Stat. Solid A
198
,
128
(
2003
).
33.
K.
Frunk
,
J. Prog. Solid State Chem.
22
,
111
(
1993
).
34.
A. P.
Barranco
,
M. P. G.
Amador
,
A.
Huanosta
, and
R.
Valenzuela
,
Appl. Phys. Lett.
73
,
2039
(
1998
).
35.
A. A.
Ahmed Youseef
,
Z. Naturforsch
57a
,
263
(
2002
).
36.
S.
Mahboob
,
G.
Prasad
, and
G. S.
Kumar
,
Bull. Mater. Sci.
29
,
35
(
2006
).
37.
Y. B.
Chen
and
C. L.
Huang
,
Surf. Coat. Technol.
201
,
654
(
2006
).
38.
J. C.
Maxwell
,
Electricity and Magnetism
(
Oxford University press
,
Oxford
,
1929
).
39.
F. M.
Pontes
,
E. R.
Leite
, and
E.
Longo
,
Appl. Phys. Lett.
76
,
2433
(
2000
).
40.
T.
Li
,
G.
Wang
,
K.
Li
,
N.
Sama
,
D.
Remiens
, and
X.
Dong
,
J. Am. Ceram. Soc.
96
,
787
(
2013
).
41.
L. Z.
Cao
,
W. Y.
Fu
,
S. F.
Wang
,
Q.
Wang
,
Z. H.
Sun
,
H.
Yang
,
B. L.
Cheng
,
H.
Wang
, and
Y. L.
Zhou
,
J. Phys. D: Appl. Phys.
40
,
2906
(
2007
).
42.
D.
Pamu
,
K.
Sudheendran
,
M.
Ghanashyam Krishna
, and
K. C.
James Raju
,
Mat.Sci. & Eng. B
168
,
208
(
2010
).
43.
T.
Santhosh Kumar
,
R.
Bhuyan
, and
D.
Pamu
,
Appl. Surf. Sci.
263
,
184
(
2013
).
44.
M. S.
Tsai
,
S. C.
Sun
, and
T. Y.
Tseng
,
J. Appl. Phys.
82
,
3482
(
1997
).
45.
B.
Taeev
,
Physics of Dielectric Materials
(
Mir publications
,
Moscow
,
1975
).
46.
B. D.
Lee
,
K. H.
Yoon
,
E. S.
Kim
, and
T. H.
Kim
,
Jpn. J. Appl. Phys.
42
,
6158
(
2003
).
47.
L1. P.
Bhattacharya
,
T.
Komeda
,
K.
Park
, and
Y.
Nishioika
,
Jpn. J. Appl. Phys.
32
,
4103
(
1993
).
48.
C. L.
Huang
and
Y. B.
Chen
,
Jpn. J. Appl. Phys.
44
,
6736
(
2005
).
49.
M.
Abazari
and
A.
Safari
,
Appl. Phys. Lett.
97
,
262902
(
2010
).