Effect of Ni²⁺ Doping on the Crystal Structure and Properties of LiAl₅O₈ Low-Permittivity Microwave Dielectric Ceramics

Abstract

Low-permittivity microwave dielectric ceramics are essential for high-frequency communication and radar systems, as they minimize signal delay and interference, thereby enabling compact and high-performance devices. In this study, LiAl_5−xNi_xO_8−0.5x (x = 0.1–0.5) ceramics were synthesized via a solid-state reaction method to investigate the effects of Ni²⁺ substitution on crystal structure, microstructure, and dielectric properties. X-ray diffraction and Rietveld refinement reveal a phase transition from the P4₃32 to the Fd$\stackrel{-}{3}$m spinel structure at x ≈ 0.3, accompanied by a systematic increase in the lattice parameter (7.909–7.975 Å), attributed to the larger ionic radius of Ni²⁺ compared to Al³⁺. SEM analysis confirms dense microstructures with relative densities exceeding 95% and grain size increases from less than 1 μm at x = 0.1 to approximately 2 μm at x = 0.5. Dielectric measurements show a decrease in permittivity (ε_r) from 8.24 to 7.77 and in quality factor (Q × f) from 34,605 GHz to 20,529 GHz with increasing Ni content, while the temperature coefficient of the resonant frequency (τ_f) shifts negatively from −44.8 to −69.1 ppm/°C. Impedance spectroscopy indicates increased conduction losses and reduced activation energy with higher Ni²⁺ concentrations.

1. Introduction

Microwave dielectric ceramics are essential components in modern communication and radar systems, enabling high-frequency signal transmission with minimal loss and interference. With the rapid advancement of wireless communication networks, satellite broadcasting, and next-generation radar technologies, there is an increasing demand for materials with well-tailored dielectric properties [1]. Among these, low-permittivity (ε_r) ceramics are particularly critical for applications requiring high signal propagation speeds, reduced electromagnetic interference, and compact device design [2]. Such materials are widely employed in high-frequency electronic components, including substrates, resonators, filters, and antennas. Their ability to reduce signal delay and dispersion makes them indispensable in cutting-edge technologies such as 5G technology, satellite communications, and phased-array radar systems. Therefore, the development of new low-permittivity microwave dielectric ceramics is vital to meet the stringent performance requirements of emerging telecommunication and sensing applications [3,4].

LiAl₅O₈ belongs to the family of lithium aluminates and typically crystallizes in a spinel-related structure. It has attracted considerable attention due to its interesting structural, electrochemical, luminescent, and thermal properties. LiAl₅O₈ exhibits polymorphism, with two distinct structural forms depending on the synthesis temperature and conditions. At high temperatures (above approximately 1295 ± 5 °C), LiAl₅O₈ adopts a cubic spinel-like phase (space group: Fd$\stackrel{-}{3}$m) with a unit cell parameter of approximately 7.921 Å. Upon cooling, it undergoes a reconstructive first-order transition to a primitive cubic phase (space group: P4₃32), with a slightly smaller unit cell parameter of about 7.907 Å. The formation of ordered or disordered polymorphs is highly sensitive to processing parameters such as temperature and duration, making LiAl₅O₈ a structurally tunable material. And it has attracted significant attention in various advanced applications. In lithium-ion and all-solid-state batteries, LiAl₅O₈ acts as an effective coating material because of its low lithium-ion migration barrier and broad electrochemical stability window. Its favorable ionic conductivity facilitates lithium transport across electrode interfaces, thereby improving cycling stability and overall battery performance [5,6]. Additionally, when processed into thin membranes through techniques such as flame spray pyrolysis and tape casting, LiAl₅O₈-based composites exhibit promising ionic conductivities, making them suitable candidates for ceramic electrolytes [7]. In nuclear applications, LiAl₅O₈ forms as a secondary phase within γ-LiAlO₂ ceramics used in tritium-producing burnable absorber rods. First-principles studies of its surfaces have revealed important insights into tritiated water formation and tritium release under irradiation [8]. Furthermore, doping LiAl₅O₈ with transition metal ions such as Fe³⁺ has led to the development of luminescent nanothermometers operating within the first biological optical window, with temperature-sensitive emission properties ideal for non-contact thermal sensing [9].

Recently, LiAl₅O₈ ceramics [10] have been reported as a promising microwave dielectric material due to their low ε_r and relatively good microwave dielectric performance (ε_r  =  8.43, Q × f  =  49,300 GHz and τ_f = −38 ppm/°C), although they require a high sintering temperature of 1600 °C. Subsequently, Ao et al. [11] investigated LiGa₅O₈, in which the Al³⁺ lattice site was occupied by the isovalent and chemically similar Ga³⁺ ion, and demonstrated that this substitution also yields excellent microwave dielectric properties. Yang et al. [12] then optimized the microwave dielectric performance of LiGa₅O₈ by replacing the Ga³⁺ site with the mixed-valence (Zn_0.5Ti_0.5)³⁺ ion. Prior studies have shown that, owing to the similar ionic radii of Ni²⁺ and Al³⁺, partial replacement of Al³⁺ by Ni²⁺ in both MgAl₂O₄ and LiGa₅O₈ matrices significantly enhances their luminescent behavior [13,14]. Moreover, Li et al. [15] explored the substitution of Al³⁺ with Ni²⁺ in BaAl₂Si₂O₈ microwave dielectric ceramics as a strategy to improve their microwave dielectric properties. Inspired by these findings, in this work, LiAl_5−xNi_xO_8−0.5x (x = 0.1–0.5) microwave dielectric ceramics were synthesized via a solid-state reaction by partially substituting Al³⁺ with Ni²⁺ in the LiAl₅O₈ material.

2. Materials and Methods

LiAl_5−xNi_xO_8−0.5x (x = 0.1–0.5) microwave dielectric ceramics were prepared by a solid-state reaction route combined with prolonged mechanical grinding to enhance particle activation before high-temperature sintering. Stoichiometric amounts of Li₂CO₃, Al₂O₃, and NiO reagents were mixed with deionized water and milled at 360 rpm for 5 h using zirconia balls of 2 mm, 3 mm, and 5 mm diameter, in a weight ratio of 5:3:1 (250 g:150 g:50 g). The ball-to-powder weight ratio was maintained at 15:1 to ensure effective homogenization. The dried powder was calcined in a furnace of 1100 °C for 3 h. The calcined powder was mixed with 10% PVA and sieved. The granulated powders were pressed into disks and then sintered at 1500 °C for 6 h.

The crystal structures of LiAl_5−xNi_xO_8−0.5x (x = 0.1–0.5) ceramics were characterized using a Shimadzu XRD-7000 X-ray diffractometer (XRD, Shimadzu Corporation, Kyoto, Japan) with Cu Kα radiation (λ = 1.5406 Å). The diffraction patterns were collected over a 2θ range of 10–80°. For phase identification, data were acquired at a scan speed of 2°/min. For Rietveld full-pattern fitting, high-resolution data were collected with a step size of 0.2°/min to ensure accurate structural refinement. Lattice parameters were refined using the GSAS software package (1st version) [16]. The bulk densities of the ceramic samples were measured by the Archimedes principle, and the relative densities were calculated based on the theoretical densities obtained from the refined lattice parameters. Impedance spectroscopy of the samples was analyzed in the frequency range of 100 Hz to 10 MHz using an impedance analyzer. The microstructures of the sintered ceramics were observed by scanning electron microscopy (SEM, FEI Sirion 200, Eindhoven, The Netherlands). Dielectric properties, including the ε_r and the Q × f values, were measured using the Hakki–Coleman method. In this technique, cylindrical dielectric samples were placed between two parallel conducting plates and operated in the TE01δ resonant mode using a network analyzer (Agilent E8362B, Santa Clara, CA, USA) [17]. Moreover, the τ_f value of the ceramics was calculated using Formula (1):

$${\tau }_f=[f(T_1)-f(T_0)]/[f(T_0)·(T_1-T_0)],$$

(1)

where f(T₁) and f(T₀) denote the resonant frequencies at T₁ (80 °C) and T₀ (30 °C), respectively.

3. Discussion

The central graph in Figure 1 presents the XRD pattern of LiAl_5−xNi_xO_8−0.5x (x = 0.1–0.5) microwave dielectric ceramics under sintering at 1500 °C. Comparison with standard cards indicates that when 0.1 ≤ x ≤ 0.2, the samples crystallize in the P4₃32 space group. When 0.3 ≤ x ≤ 0.5, a significant number of diffraction peaks disappear, as clearly seen in Figure 1b,c. The leftmost panel in Figure 1 shows an enlarged view of the diffraction region of 10–30° for x = 0.2 and x = 0.3. Based on the standard cards, a phase transition is observed at x ≈ 0.3, with samples of 0.3 ≤ x ≤ 0.5 exhibiting a spinel structure (Fd$\stackrel{-}{3}$m space group). Therefore, in LiAl₅O₈ materials, Al³⁺ ions are substituted by Ni²⁺ ions, and as the substitution content increases, a phase transition in LiAl₅O₈ materials is induced, occurring at approximately x = 0.3. The rightmost panel of Figure 1 presents the results of the main peaks measured at diffraction angles between 37° and 38° for different substitution contents. The peaks shift to lower angles with increasing Ni²⁺ content. This is attributed to the fact that the ionic radius of Al³⁺ is smaller than that of Ni²⁺. According to Bragg’s equation, $2d\sin \theta =n\lambda$, an increase in the lattice constant leads to a larger interplanar spacing d, which results in a lower diffraction angle θ, shifting the peaks to lower angles.

To investigate the effect of Ni²⁺ substitution for Al³⁺ ions on the crystal structure of LiAl₅O₈ materials, Rietveld full-pattern refinement was performed on the X-ray diffraction data. Figure 2 presents the Rietveld full-pattern fitting results for LiAl_5−xNi_xO_8−0.5x materials at x = 0.1 and x = 0.3. In the figure, red dots represent the experimental data, the black line corresponds to the calculated diffraction pattern, vertical bars indicate the positions of Bragg peaks, and the purple line shows the difference between the calculated and experimental values. The figure clearly demonstrates that, for samples with different space groups, the fitted values are in excellent agreement with the experimental measurements.

Table 1. Rietveld-refined structural parameters of LiAl_5−xNi_xO_8−0.5x microwave dielectric ceramics. The subscript 1 denotes the sample with x = 0.1, while the subscript 2 denotes the sample with x = 0.3.

Atom	x	y	z	Fraction	Rp	Rwp	χ2
Li11	0.6250	0.6250	0.6250	1.00	0.053	0.081	4.869
Li21	−0.8756	1.1256	0.1250	1.00
Al11	0.0024	0.0024	0.0024	1.00
Al21	0.6250	0.6250	0.6250	1.00
Al31	0.3725	0.1225	0.1250	1.00
Ni11	−0.0024	−0.0024	−0.0024	1.00
Ni21	0.6250	0.6250	0.6250	1.00
O11	0.1135	0.1231	0.3871	1.00
O21	0.4023	0.4023	0.4023	1.00
Li2	0.5000	0.5000	0.5000	1.00	0.048	0.071	4.235
Al12	0.1250	0.1250	0.1250	0.94
Al22	0.5000	0.5000	0.5000	0.94
Ni12	0.1250	0.1250	0.1250	0.06
Ni22	0.5000	0.5000	0.5000	0.06
O2	0.2559	0.2559	0.2559	1.00

summarizes the structural parameters of LiAl_5−xNi_xO_8−0.5x (x = 0.1 and x = 0.3) microwave dielectric ceramics, obtained from Rietveld full-pattern refinement. These parameters include atomic positions, site occupancies, and the R factor and χ² values from the fitting process. The low R factor and χ² values, together with the visual results in Figure 2, clearly indicate that the Rietveld full-pattern refinement method produces highly reliable results.

In addition, the Rietveld full-pattern refinement method enables the determination of the lattice parameters for the LiAl_5−xNi_xO_8−0.5x microwave dielectric ceramics. Given the cubic symmetry of the structure, the unit cell volume was calculated directly from the lattice parameter. Then the theoretical (calculated) densities were obtained from the refined crystal parameter. The actual densities of the ceramic samples, measured using the Archimedes principle, are listed in

Table 2. Lattice parameters and relative densities of LiAl_5−xNi_xO_8−0.5x microwave dielectric ceramics.

x	Lattice Parameter a (Å)	Unit Cell Volume (Å3)	Measured Density (g/cm3)	Calculated Density (g/cm3)	Relative Density (%)
0.1	7.909	494.726	3.528	3.656	96.5
0.2	7.944	501.323	3.552	3.639	97.6
0.3	7.952	502.839	3.565	3.660	97.4
0.4	7.963	504.929	3.536	3.676	96.2
0.5	7.975	507.215	3.582	3.690	97.1

. Based on these values, the relative densities were determined as the ratio of the measured density to the calculated density. Since the LiAl_5−xNi_xO_8−0.5x ceramics are single-phase, the relative density values are considered reliable. The results presented in Table 2 show that the lattice parameters of the LiAl_5−xNi_xO_8−0.5x ceramics increase with the substitution level of Ni²⁺ ions. Specifically, when x increases from 0.1 to 0.5, the lattice parameter a expands from 7.909 to 7.975 Å. Based on the ionic radii of Al³⁺ (0.535 Å) and Ni²⁺ (0.69 Å) reported in reference [18], the substitution of smaller Al³⁺ ions with larger Ni²⁺ ions leads to an expansion of the crystal lattice. Furthermore, all samples exhibit relative densities exceeding 95%, indicating that dense ceramic structures were successfully obtained under sintering at 1500 °C.

Figure 3 shows the SEM micrographs and corresponding grain size distributions of the LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) microwave dielectric ceramics. The micrographs clearly demonstrate that all samples sintered at 1500 °C exhibit minimal porosity, resulting in dense microstructures consistent with the high relative densities reported in Table 2. Additionally, the micrographs indicate that the grain size in the sample with x = 0.1 is less than 1 μm. As the Ni²⁺ content increases, a clear trend toward larger grain sizes is observed; for instance, at x = 0.2, the grain size reaches approximately 2 μm. According to previous studies [10], LiAl₅O₈ ceramics sintered at 1600 °C retain a certain degree of porosity, whereas LiAl_5−xNi_xO₈ (0.1 ≤ x ≤ 0.5) ceramics sintered at 1500 °C exhibit negligible porosity and achieve a significantly higher densification. Therefore, the partial substitution of Al³⁺ by Ni²⁺ in LiAl₅O₈ not only reduces the required sintering temperature but also promotes densification. Moreover, the average grain size increases progressively with higher Ni²⁺ content.

Figure 4 presents the microwave dielectric properties of LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) ceramics, with the black line indicating the variation in the ε_r as a function of x. As the Ni²⁺ substitution level increases from 0.1 to 0.5, the ε_r decreases gradually from 8.24 to 7.77. According to previous studies [19], ε_r is typically affected by factors such as porosity, microstructure, polarizability, and the presence of secondary phases. In this work, the XRD data, relative density measurements, and SEM observations confirm that the ceramics prepared via the solid-state reaction are single-phase and highly densified. Therefore, the variation in the ε_r of LiAl_5−xNi_xO_8−0.5x microwave dielectric ceramics is closely related to their microstructural characteristics. Moreover, according to Shannon’s findings [20], the total dielectric polarizability (${\alpha }_D^T$) of a material can be expressed as the sum of the individual ionic polarizabilities (α_D). Consequently, the overall dielectric polarizability (${\alpha }_D^T$) for LiAl_5−xNi_xO_8−0.5x can be calculated using the following Equation (2):

$${\alpha }_D^T(\mathrm{L}\mathrm{i}\mathrm{A}{\mathrm{l}}_{5-x}{\mathrm{Zn}}_x{\mathrm{O}}_{8-0.5x})=\alpha (\mathrm{L}{\mathrm{i}}^{+})+(5-x)\alpha (\mathrm{A}{\mathrm{l}}^{3+})+x\alpha (\mathrm{Z}{\mathrm{n}}^{2+})+(8-0.5x)\alpha ({\mathrm{O}}^{2-})$$

(2)

Subsequently, employing the Clausius–Mossotti equation (Equation (3)), the calculated dielectric constant ε_c can be calculated by the following expression:

$${\epsilon }_c={\displaystyle \frac{1+2b{\alpha }_{\mathrm{D}}^{\mathrm{T}}/V_{\mathrm{m}}}{1-b{\alpha }_{\mathrm{D}}^{\mathrm{T}}/V_{\mathrm{m}}}}$$

(3)

where b = 4π/3, and V_m is the molar volume obtained from Rietveld refinement. The ionic polarizabilities of Li⁺, Al³⁺, Ni²⁺, and O²⁻ are 1.20 ų, 0.78 ų, 1.23 ų, and 2.00 ų, respectively [19]. The data in

Table 3. The calculated dielectric polarizability (${\alpha }_D^T$), measured dielectric constant ε_r, and calculated dielectric constant ε_c value of LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) microwave dielectric ceramics.

x	${\mathit{\alpha }}_{\mathit{D}}^{\mathit{T}}$ (Å3)	Vm (Å3)	εr	εc
0.1	84.18	494.726	8.24	8.43
0.2	83.96	501.323	8.13	8.04
0.3	83.74	502.839	8.06	7.94
0.4	83.52	504.929	7.92	7.76
0.5	83.30	507.215	7.77	7.60

reveal that the calculated values are in good agreement with the experimental measurements and exhibit a similar decreasing trend.

Figure 4 illustrates the variation in the Q × f value of LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) microwave dielectric ceramics as a function of the Ni²⁺ substitution level, as indicated by the green curve. It is evident that with increasing Ni²⁺ substitution from x = 0.1 to x = 0.5, the Q × f value significantly decreases from 34,605 GHz to 20,529 GHz. The quality factor serves as an indicator of the dielectric loss in ceramic materials, and impedance spectroscopy is a widely adopted technique for analyzing such losses. Figure 5 shows the complex impedance plot for the LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) ceramic samples. Figure 6 presents the impedance analysis for the sample with x = 0.1, where the data are fitted using an equivalent circuit composed of two parallel R-CPE (resistor–constant phase element) units connected in series. Each R-CPE unit represents a different electrical response: one corresponds to the bulk (grain) contribution and is denoted with the subscript b, while the other corresponds to the grain boundary contribution and is denoted with the subscript gb. Since the circuit elements do not behave as ideal capacitors, a constant phase element (CPE) is used to replace the ideal capacitor in the model. The CPE introduces a parameter, n, which reflects the deviation from ideal capacitive behavior and varies between 0 and 1; n = 1 corresponds to an ideal capacitor, and n = 0 corresponds to a pure resistor. The fitted values for grain resistance (R_b), grain boundary resistance (R_gb), and the CPE exponent (n) are listed in Figure 6 for each temperature. Subsequently, using Equation (4), the R_b is converted into the corresponding bulk conductivity σ_b.

$${\sigma }_{\mathrm{b}}={\displaystyle \frac{1}{R_b}}{\displaystyle \frac{4t}{\pi D^2}}$$

(4)

Here, t and D denote the sample thickness and diameter, respectively. The activation energy Ea is determined by fitting the temperature dependence of σ_b to the Arrhenius equation, as shown in Figure 7. As x increases from 0.1 to 0.5, activation energy (Ea) decreases from 2.420 eV to 1.134 eV. For these materials, the intrinsic band gap (Eg) and the Ea are related by Eg ≈ 2Ea [21,22,23]. Therefore, impedance spectroscopy confirms that increasing the acceptor doping of Ni²⁺ ions leads to higher conduction losses, which in turn reduce the Q × f values. Moreover, increasing the Ni²⁺ doping concentration also leads to a higher concentration of oxygen vacancies, thereby contributing to increasing the dielectric loss.

Figure 4 shows the variation in τ_f of LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) microwave dielectric ceramics with increasing x, as indicated by the purple line. As the Ni²⁺ content increases from 0.1 to 0.5, τ_f clearly shifts in the negative direction from −44.8 ppm/°C to −69.1 ppm/°C. Figure 8 further presents the relationship between τ_f and the unit cell volume (V) for these ceramics. The trend in τ_f for Ni²⁺ substitution is analogous to that for the Zn²⁺ substitution of a previous study [10]. This similarity arises because τ_f is inherently linked to the crystal structure, and the structural changes induced by Ni²⁺ are comparable to those caused by Zn²⁺. Therefore, the increasingly negative τ_f with higher Ni²⁺ content is consistent with the observed experimental data.

4. Conclusions

In this work, a series of Ni-doped LiAl_5−xNi_xO_8−0.5x (0.1 ≤ x ≤ 0.5) microwave dielectric ceramics were synthesized via a solid-state reaction method. XRD and Rietveld refinement revealed a distinct phase boundary at x ≈ 0.3: compositions with x ≤ 0.2 retain the ordered P4₃32 structure of LiAl₅O₈, whereas higher Ni contents stabilize the Fd$\stackrel{-}{3}$m spinel phase. The progressive shift of diffraction peaks toward lower angles, along with a linear increase in the lattice parameter from 7.909 Å to 7.975 Å, confirms the successful substitution of larger Ni²⁺ ions for the Al³⁺ sublattice. SEM observations and density measurements indicate that all compositions sintered at 1500 °C exhibit high densification (>95% relative density) with minimal porosity. Dielectric measurements show that Ni²⁺ doping systematically reduces the permittivity (ε_r decreases from 8.24 to 7.77) and deteriorates the quality factor (Q × f drops from 34,605 GHz to 20,529 GHz) with increasing x. Meanwhile, the τ_f becomes more negative (−44.8 to −69.1 ppm/°C), indicating increased sensitivity to thermal expansion and polarizability variations in the Ni-substituted lattice. Impedance spectroscopy further reveals that higher Ni content promotes conduction losses and reduces activation energy, thereby accounting for the observed decline in Q × f. Overall, the controlled Ni²⁺ incorporation into LiAl₅O₈ provides an effective strategy for tuning crystal structure, microstructure, and dielectric performance, rendering these ceramics promising candidates for next-generation microwave substrates and resonator components.