Microwave Performance, Microstructure, and Crystallization of (Mg 0.6 Zn 0.4 ) 1 − y Ni y TiO 3 Ilmenite Ceramics

: The sintering behavior, microstructure analysis, crystallization, and microwave performance of (Mg 0.6 Zn 0.4 ) 1 − y Ni y TiO 3 (y = 0.01–0.2) ceramics, processed with raw powders of MgO, NiO, ZnO, and TiO 2 via the conventional solid-state method, are investigated. The main phases of (Mg 0.6 Zn 0.4 ) 1 − y Ni y TiO 3 ceramics were obtained. With partial replacement by Ni 2+ , the (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 could be well sintered at 1200 ◦ C, and the microwave performance was shown to be positively correlated with sintering temperature. The permittivity ( ε r ) saturated at 18.7–19.3, and the quality factor (Q f ) values approached 72,000–165,000 (GHz) as the sintering temperatures increase from 1125 to 1250 ◦ C. The temperature coefﬁcient of resonance frequency ( τ f ) falls in a stable range of − 62.9 to − 66 ppm/ ◦ C as sintering temperature rising. A permittivity ( ε r ) of 19.3, a maximum Q f value of 165,000 (GHz), and a temperature coefﬁcient ( τ f ) of − 65.4 ppm/ ◦ C were measured for the samples at 1200 ◦ C/4 h. (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 material system shows high potential for applications of high frequency-selection components in satellite communication and 5G wireless telecommunication systems.


Introduction
Recent demands for high-frequency electronic components are increasing, and ceramicbased compositions are extensively used in satellite communications, including DBS TV, GPS, Internet of Things (IoT), etc., and 5G telecommunications, cell phone, cell site, etc. Furthermore, microwave dielectric ceramics are widely used for filters, resonators, and mixers in wireless communication systems [1][2][3][4]. In microwave telecommunications, microwave resonators fabricated by dielectric ceramics result in the miniaturization of microwave devices.
Materials possessing the following dielectric properties are suitable for dielectric resonators [5][6][7][8]: (1) Permittivity (ε r ) > 20: The permittivity is an essential factor for component size reduction. (2) Quality factor (tanδ, the inverse of the dielectric loss) > 30,000 (GHz): Lower insertion loss and steeper cut-off can be obtained due to a higher Qf value, which ensured the lower dielectric loss of components. (3) Temperature coefficient of resonance frequency (τ f ) < ±10 ppm/ • C: A near-zero τ f value (<±10 ppm/ • C) is necessary for the stability of device performance even temperature changes enormously in the environment. Generally speaking, three parameters strongly affect the size, frequency selectivity, and temperature stability of the device to the system, respectively. MgTiO 3 -based ceramics, which possessed an ilmenite-type structure and trigonal space group R-3 have caught much attention and typical applications as dielectric resonators [9], filters, antennas, and so on for microwave communication systems due to its good microwave dielectric performance recently. MgTiO 3 sintered at 1350 • C was demonstrated excellent dielectric performance in microwave frequency: ε r~1 7, Qf value~160,000 (at 7 GHz) and temperature coefficients of resonance frequency (τ f )~−51 ppm/ • C [10]. Furthermore, (Mg 0.6 Zn 0.4 )TiO 3 sintered at 1200 • C with ε r~1 9.8, Qf value~144,000 and τ f~− 66 ppm/ • C have also been demonstrated as promising titanate candidate for broad applications in this microwave frequency. Some researchers pay attention to comprehensively investigate the effects of relative contents between Mg and Zn in (Mg y Zn 1−y )TiO 3 system [11,12]. As a result, microwave dielectric performance of (Mg y Zn 1−y )TiO 3 highly depended on the content of Mg.

Experimental Analysis
Conventional solid-state methods were utilized to synthesize samples of (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 (y = 0.01-0.2) from high-purity oxide powders (>99.9%): MgO, NiO, and TiO 2 . Raw MgO powder was fired at 600 • C/1h to avoid moisture retention due to its hygroscopic characteristic. (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 (y = 0.01-0.2) compounds were made from mixed oxides stoichiometrically and ground in distilled water for 24 h in a ball mill with agate balls. The mixed solution was dried in the oven and calcined at 1100 • C for 4 h in a hightemperature furnace. A 3.5 wt% of a 12% PVA solution as a binder (Polyvinyl alcohol 500, Showa) was added into the calcined powder, granulated by sieving through a 100 mesh, and pressed into pellets, 11 mm in diameter and 5 mm in thickness, under 200 MPa pressure. The pellets were sintered at temperatures ranging from 1125 to 1250 • C for 4 h in the atmosphere. The heating and cooling rates of a high-temperature furnace were both set at 10 • C/min to obtain high-quality samples.
The crystalline phases of the calcined powder and the sintered ceramics were distinguished by X-ray diffraction (XRD) using CuKa (λ = 0.15406 nm) radiation with a Siemens D5000 diffractometer (Siemens, Berlin/Munich, Germany). Scanning electron microscopy (SEM; Philips, Amsterdam, The Netherlands) was utilized to observe the sintered surface's microstructure, and an energy-dispersive X-ray spectrometer (Philips) was used to identify the existence of secondary phases. Calculation of lattice parameter was performed with the Rietveld method to fit at least five peaks from measured X-Ray patterns. The apparent densities were measured using the Archimedes method with the three sintered samples under the same process and average measured data to calculate relative densities. The permittivity (ε r ) and the quality factor (Q) at microwave frequencies were measured by the Hakki-Coleman dielectric resonator method [28,29]. This method utilizes parallel conducting plates and coaxial probes in TE 011 mode, where TE represented transverse electric waves, the first two subscript integers denote the waveguide mode, and the third integer subscript indicates the order of resonance in an increasing set of discrete resonant lengths. The measurement system was connected to Anritsu network analyzer with model MS46122B. The measurement technique for the temperature coefficient of resonance frequency (τ f ) is almost the same as that of the quality factor measurement except for using a thermostat. The τ f at microwave frequencies was measured, and note change of resonance frequency from 30 • C to 80 • C. The following formula can calculate the τ f value (ppm/ • C), where f 1 and f 2 mean the resonance frequencies at T 1 = 25 • C and T 2 = 80 • C, respectively.  The formation of (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 Ti 2 O 5 might degrade the microwave dielectric performance of (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 ceramics. As expected, the second phase (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 was enhanced at high temperatures since the Mg 0.95 Ni 0.05 Ti 2 O 5 requires a high sintering temperature of 1450 • C [26,33]. In other words, less (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 Ti 2 O 5 phase would exist at lower sintering temperatures [34]. Similar XRD analysis was identified for the samples with y = 0.05 sintering at 1125-1250 • C for 4 h (Figure 2).

Results and Discussion
The lattice parameters of (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 ceramics sintered at 1200 • C for 4 h with various y values were also calculated for further structure investigation and forming of solid solution (Table 1). A decrease in the lattice parameters was observed for (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 ceramics compared to that of (Mg 0.6 Zn 0.   The micrographs of (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 sintered at different temperatures for 4 h are revealed in Figure 3. The microstructures of samples were not dense at 1125 • C. Grain size gradually increased with temperature and became highly uniform at 1200 • C. The growth of rod-shaped grain was enhanced at temperatures higher than 1225 • C and possibly the second phase. EDS analysis for Spot A-D and E-F identified the quantity composition of (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 sintered at 1200 • C and (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 Ti 2 O 5 sintered at 1250 • C for 4 h , as shown in Figure 4a   The apparent and relative density of the (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 with y = 0.01-0.2 sintered at different temperatures (1125-1250 • C) for 4 h is demonstrated in Figure 5. The relative density of the (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 exceeded 91.6% of the theoretical density (TD) in all specimens. The apparent density of the samples increased when the temperature increased from 1125 to 1200 • C. This increase in the apparent density can be attributed to the formation of dense microstructures, as observed in Figure 3d. The apparent density decreased when the temperature exceeded 1200 • C. The permittivity and quality factor (Qf ) of (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 ceramics sintered at temperatures for 4 h with various y values (0.01-0.2) are shown in Figure 6. The relationship between permittivity (ε r ) values and temperature reveals the same trend between densities and temperature since higher density means lower porosity inside samples. The permittivity increased with increasing temperature, reaching a maximum of 19.3 at 1200 • C and after that decreasing. Thus, increasing sintering temperature does not always result in a higher permittivity (ε r ). The permittivity of the well-sintered (Mg 0. 6  The Qf value is a crucial index for microwave dielectric ceramics applications because a high Qf value represented a lower dielectric loss for microwave devices. Qf values increased when temperature increased from 1125 • C to 1200 • C (Figure 6), consistent with the variation of bulk density, as shown in Figure 5. The increase of Qf value at low temperatures is attributed to the rise in density identical to grain growth uniformity, as shown in Figure 3. A maximum Qf value of 165,000 (GHz) was observed for the (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 ceramics sintered at 1200 • C for 4 h. Qf value reached a maximum at 1200 • C, and after that, it decreased. The Qf value degraded due to the inhomogeneous grain growth, which leads to a reduction of density, as mentioned in Figure 3. There are several possible mechanisms to cause microwave dielectric loss, such as the lattice vibrational modes, the pores, second phases, impurities, and lattice defects [35]. Furthermore, apparent density also closely affects the dielectric loss and has also been demonstrated in microwave dielectric mate-rials [35,36]. The Qf value and the variation of density of the (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 show a similar trend, and hence apparent density is one of the dominant factors to cause the dielectric loss in this ceramic system. As illustrated in Figure 7, the temperature coefficient of resonance frequency (τ f ) of (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 at temperatures ranging from 1125 to 1250 • C with x = 0.01-0.2. τ f values fluctuate slightly at the whole temperature range. As expected, the synthesized composition persisted unchanged, and hence no noticeable fluctuation of the τ f value was shown. As we know, τ f is positively related to the synthesized composition, the second phase, and the material's additive. The measured τ f values were from −62.9 to −66 ppm/ • C as the (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 sintered at 1125-1250 • C. At 1200 • C, a τ f of (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 with 4 h sintering time was measured around −65.4 ppm/ • C. From the above discussion, microwave dielectric performance of (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 with various y value sintered at 1200 • C for 4 h was shown in Table 2 further demonstrated that best performance was found for y = 0.05. We also observed that Qf value is sensitive to Ni content, and however, permittivity and τ f remain slightly fluctuate with varied Ni content. The sintering temperature affects the permittivity and Qf value to some extent. However, τ f is not sensitive to sintering temperature. Table 3  As mentioned previously, a dielectric resonator's thermal stability is defined by the temperature coefficient of the resonance frequency. Dielectric materials systems with near-zero τ f are crucial to fabricate substrates used in microwave frequency. In order to obtain near-zero τ f dielectric material, SrTiO 3 [34] possessed ultrahigh positive τ f (ε r 205, Qf value~4200 and τ f~1 700 ppm/ • C under 1350 • C) was chosen to mix with (Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 (τ f~− 66 ppm/ • C under 1200 • C). According to the mixture rules for τ f , 0.94(Mg 0.6 Zn 0.4 ) 0.95 Ni 0.05 TiO 3 − 0.06SrTiO 3 was fabricated and analyzed in advance. Figure 8 shows τ f ranging from +46.6 to 68.5 ppm/ • C under 1125-1225 • C. Thus, we believe a near-zero τ f dielectric material with excellent microwave performance will be realized possibly after stoichiometry calculation and experiments in future work.

Conclusions
The performance of the (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 at microwave frequency was investigated and analyzed. The dielectric ceramics with an ilmenite-type structure belonging to the trigonal space group R-3 could be founded in all samples. With the substitution of Ni 2+ for Mg 2+ and Zn 2+ , the densification of (Mg 0.6 Zn 0.4 ) 1−y Ni y TiO 3 was significantly enhanced.