Structure and microwave dielectric characteristics of Sr(La1−xSmx)2Al2O7 ceramics

Zhi Qi Yuan, Bing Liu, Xiao Qiang Liu and Xiang Ming Chen*
Laboratory of Dielectric Materials, School of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, China. E-mail: xmchen59@zju.edu.cn

Received 17th August 2016 , Accepted 29th September 2016

First published on 30th September 2016


Abstract

Sr(La1−xSmx)2Al2O7 (0 ≤ x ≤ 1) ceramics were synthesized by a standard solid-state reaction method, and the microwave dielectric characteristics were investigated together with the densification behavior, microstructure and phase composition. X-ray diffraction analysis indicated that the matrix phase with Ruddlesden–Popper structure of n = 2 was obtained in the entire composition range. Dense ceramics with relative density of over 96% theoretical density were obtained by sintering at over 1575 °C in air for 3 h. With increasing Sm3+ substitution, the relative dielectric constant increased from 18.8 to 21.3, and the temperature coefficient of resonant frequency was adjusted from −21.7 to 4.8 ppm per °C, while though the Qf value should theoretically decrease with Sm substitution because of the increased interlayer polarization charge, the high Qf was maintained for x < 0.5 and even increased slightly at x = 0.5 due to some extrinsic reasons such as large grain size. The best combination of microwave dielectric characteristics was achieved at the composition of x = 0.5: εr = 21.0, Qf = 115[thin space (1/6-em)]900 GHz, τf = −3.7 ppm per °C.


Introduction

Over the past few decades, with the rapid development of modern wireless communication technology such as satellite communications, base station and mobile communication systems, high-performance microwave dielectric ceramics which play the key roles in microwave resonators, filters and antennas have attracted numerous attentions of researchers.1,2 It is well-known that three key characteristics are required in microwave dielectric ceramics: appropriate relative dielectric constant (εr), high unloaded quality factor (Q) and near-zero temperature coefficient of resonant frequency (τf).3,4 As the important commercial microwave dielectric ceramics with low εr (εr < 30), ceramics in MgTiO3–CaTiO3 system (εr = 21, Qf = 56[thin space (1/6-em)]000 GHz, τf = 0 ppm per °C) with low cost are well-known as the materials for dielectric resonator and patch antenna,5 but their Qf values are not high enough for high-performance microwave applications. On the other hand, Ba-based complex perovskite ceramics such as Ba(Mg1/3Ta2/3)O3, Ba(Zn1/3Ta2/3)O3 and Ba(Zn1/3Nb2/3)O3 are well known as the ultralow loss microwave dielectric ceramics.6–9 However, their high cost due to the noble elements such as Ta and Nb, poor sintering ability and strong processing dependence because of the volatile element such as Zn are serious limitations, considering their continued practical applications. Therefore, searching for a new candidate with desired properties and low cost is a crucial issue.

In the past few decades, Ruddlesden–Popper (R–P) compounds with the general formula of A′2[An−1BnO3n+1] (A, A′ = alkaline or rare earth metal ions, B = metal ion)10 have attracted much attentions because of their rich properties such as low thermal conductivity, oxide-ion conductivity, superconductivity, ferroelectricity, colossal magnetoresistance, catalytic activity and magnetic transport property.11–15 The stability of the structure depends on the crystal-chemical constraint that is the n/1 intergrowth of n ABO3 perovskite layers (Pn) alternating with one AO rock-salt layer (RS) along c axis. Recently, MLnAlO4 (M = Sr, Ca; Ln = La, Nd, Sm) ceramics with R–P structure of n = 1, namely K2NiF4 structure without noble elements such as Ta and Nb have been investigated, and they have attracted wide attentions as potential candidates for low-loss microwave dielectric applications.16–20 Wise et al.21 reported the relation between the structure and property of (SrxCa1−x)n+1TinO3n+1 with different n value. Thus, the microwave dielectric characteristics of R–P compounds could be controlled by adjusting the n value.

In the previous work, SrLn2Al2O7 (Ln = La, Nd, Sm) ceramics with R–P structure of n = 2 were prepared, and their excellent microwave dielectric properties (εr = 18.2–21.6, Qf = 64[thin space (1/6-em)]680–71[thin space (1/6-em)]680 GHz and τf = −22.1 to +4 ppm per °C) were obtained.22 Referring to the MLnAlO4 (M = Sr, Ca; Ln = La, Nd, Sm) ceramics, their microwave dielectric characteristics can be improved by cation substitution. Zvereva et al.23,24 have researched the formation of (Sr1−xCax)Ln2Al2O7 (Ln = La, Nd) solid solution with x < 0.5, and have found that the solid solution with Ca substitution was formed easily. According to Yi and Liu et al.,25,26 the more excellent microwave dielectric characteristics of SrLn2Al2O7 (Ln = La, Nd, Sm) ceramics can be expected by proper amount of Ca2+ and Ti4+ substitution for Sr2+ and Al3+, respectively. SrLa2Al2O7 possesses the best Qf value among SrLn2Al2O7 (Ln = La, Nd, Sm) ceramics due to the minimum interlayer polarization charge,22 but its τf is too negative for practical application. Therefore, SrSm2Al2O7 is chosen to adjust the Qf and τf values of SrLa2Al2O7 because of its same crystal structure as SrLa2Al2O7 and positive temperature coefficient of resonant frequency.

In the present work, Sr(La1−xSmx)2Al2O7 (x = 0, 0.25, 0.5, 0.75, 0.85, 0.95, 1) microwave dielectric ceramics are prepared by a standard solid-state reaction method. The microwave dielectric properties are investigated together with their structure, and the relationship between them are also discussed. The effects of interlayer polarization charge, grain size, secondary phase and stacking fault on the properties are emphasized.

Experimental

Sr(La1−xSmx)2Al2O7 (x = 0, 0.25, 0.5, 0.75, 0.85, 0.95, 1) ceramics were prepared using high purity SrCO3 (99.95%), La2O3 (99.99%), Sm2O3 (99.99%) and Al2O3 (99.99%) powders as raw materials. Here, La2O3 and Sm2O3 raw powders were preheated at 900 °C in air for 2 h before weighing because of the hygroscopicity of the rare earth oxide. The stoichiometric raw powders were mixed by ball-milling with zirconia media in ethanol for 24 h. The slurries were dried and then calcined at relevant temperature range from 1425 °C to 1575 °C in air for 6 h. Then, the calcined powders were mixed by ball-milling and dried again. After mixing and grinding with the organic binder (about 8 wt% polyvinyl alcohol), the powders were pressed into cylinders of 12 mm in diameter and 2–6 mm in height under a uniaxial pressure of 98 MPa. Then the cylinders were sintered at 1525–1600 °C in air for 3 h to form the dense ceramics with the heating rate of 5 °C min−1. After sintering, the cylinders were cooled to 1000 °C at a rate of 2 °C min−1 and further cooled inside the furnace.

The bulk densities of sintered ceramics with regular shape were measured by the Archimedes method. After crushing and grinding the sintered ceramics, their crystal structures were characterized by powder X-ray diffraction analysis (XRD) (RIGAKU D/max 2550PC, Rigaku Co., Tokyo, Japan) with CuKα (λ = 0.15418 nm) radiation. After polishing and thermal etching at 1525 °C for 30 min, the microstructures of the sintered ceramics were observed by scanning electron microscopy (SEM) (SIRION-100; FEI Co., Eindhoven, Netherlands). For Rietveld refinement analysis, the XRD data were collected over the 2θ range from 8° to 130° with a step width of 0.02° and a count time of 2 s at room temperature. FULLPROF program was chosen to calculate the Rietveld structure. Samples for transmission electron microscopy observation were cut ultrasonically to disks of 3 mm in diameter and then mechanical polished to a thickness of about 10 μm. The final perforation of the samples was conducted by precision argon-ion milling. The selected area electron diffraction (SAED) patterns and high-resolution TEM (HRTEM) images were obtained using a transmission electron microscopy (Tecnai G2 F20, FEI Co., Hillsboro, OR).

The relative dielectric constant (εr) and the temperature coefficient of resonant frequency (τf) were determined by the Hakki–Coleman method,27 using a vector network analyzer (E8363B, Agilent Technologies, Palo Alto, CA). The τf value was measured in the temperature range of 20–80 °C at microwave frequency and calculated by the following equation:

 
image file: c6ra20776f-t1.tif(1)
where, the f20 and f80 were the TE011 resonant frequencies at 20 °C and 80 °C, respectively. The quality factor (Q) was determined at around 7 GHz by the closed cylindrical resonant cavity method, using a silver-coated cavity connected to the network analyzer.28 Here, the product Qf was used to evaluate the dielectric loss instead of Q alone, because the Q factor generally varied inversely with frequency (f) in the microwave range.

Results and discussion

Fig. 1 shows the relative densities of Sr(La1−xSmx)2Al2O7 (0 ≤ x ≤ 1) ceramics as functions of the sintering temperature. In the entire composition range, the relative density increases with increasing sintering temperature and over 96% theoretic density (T.D.) is reached at 1575 °C. Moreover, the maximum relative density of over 99% T.D. is obtained at x = 0.5. Fig. 2 shows the microstructures of the present ceramics sintered at 1575 °C in air for 3 h. The grains generally exhibit long strip-shaped and plate-shaped morphology in the first three compositions (x = 0–0.5) and the different shapes indicate the different sections of the grains. In contrast, the rest compositions exhibit short rod-shaped morphology. As for the grain size, the lengths of the grains are nearly 10–15 μm at x = 0–0.5, however that of the rest compositions (x = 0.75–1) are about 2–5 μm. Besides, visible pores exist in the thermally etched surfaces of all ceramics, but the minimum amount is obtained at x = 0.5.
image file: c6ra20776f-f1.tif
Fig. 1 Relative densities of Sr(La1−xSmx)2Al2O7 ceramics as functions of sintering temperature.

image file: c6ra20776f-f2.tif
Fig. 2 SEM micrographs of thermally etched surfaces of Sr(La1−xSmx)2Al2O7 ceramics with various compositions sintered at 1575 °C in air for 3 h: (a) x = 0, (b) x = 0.25, (c) x = 0.5, (d) x = 0.75, (e) x = 0.85, (f) x = 0.95 and (g) x = 1.0.

Fig. 3(a) shows the XRD patterns of the present ceramics sintered at 1575 °C in air for 3 h. In the entire composition range, the reflection peaks of the major phase along with the relevant indices of the crystallographic planes can be matched with those for SrLa2Al2O7, indicating that the structure of the major phase is tetragonal Ruddlesden–Popper structure in space group I4/mmm. Different secondary phases appear in different compositions: Al2O3 in space group R[3 with combining macron]c at x = 0–0.5; SrSmAl3O7 and SmAlO3 in space group P[4 with combining macron]21m and Pbnm, respectively at x = 0.75–1. As shown in Fig. 3(b), the diffraction patterns of the enlarged (006) and (110) peaks shift slightly toward higher angle with the increasing x, indicating that the lattice parameters (a, c and V) are decreasing.


image file: c6ra20776f-f3.tif
Fig. 3 (a) XRD patterns of Sr(La1−xSmx)2Al2O7 ceramics with various compositions sintered at 1575 °C in air for 3 h; (b) enlargement of the peaks at (006) and (110) crystallographic plane.

Fig. 4 illustrates the profile fit of the calculated and observed X-ray diffraction patterns of the ceramic (x = 0.5) together with the Bragg positions of the secondary phases. The Rietveld refinement results are listed in Table 1, including refined the structural parameter, reliability factor, ion occupation distribution and bond length. According to the results, as shown in Fig. 5, with the increment of Sm3+ substitution, the lattice parameters (a, c and V) of the present ceramics decrease linearly due to the decreasing effective ionic radius (RLa3+ > RSm3+), which is in consonance with the variations of the corresponding peaks in XRD patterns.


image file: c6ra20776f-f4.tif
Fig. 4 Profile fit for the Rietveld refinement of Sr(La1−xSmx)2Al2O7 (x = 0.5) ceramics sintered at 1575 °C in air for 3 h.
Table 1 The refined structural parameter, reliability factor, ion occupation distribution and bond length for Sr(La1−xSmx)2Al2O7 ceramics sintered at 1575 °C for 3 ha
  0 0.25 0.5 0.75 0.85 0.95 1
a The refined atomic fractional coordinate positions are A(1)[2b] (0, 0, 0.5), A(2)[4e] (0, 0, z), Al[4e] (0, 0, z), O(1)[2a] (0, 0, 0), O(2)[8g] (0, 0.5, z), O(3)[4e] (0, 0, z) in space group I4/mmm.
a (Å) 3.7741(4) 3.7611(5) 3.7460(4) 3.7316(5) 3.7252(5) 3.7192(5) 3.7165(5)
c (Å) 20.2082(2) 20.1227(3) 20.0369(2) 19.9589(3) 19.9261(3) 19.9026(3) 19.8901(3)
V (per unit cell) (Å3) 287.845 284.653 281.161 277.930 276.515 275.305 274.722
Z(Sr, Ln)(2) (Å) 0.3185(3) 0.3184(3) 0.3183(3) 0.3184(2) 0.3185(2) 0.3185(2) 0.3185(2)
ZAl (Å) 0.0954(12) 0.0953(2) 0.0950(15) 0.0949(12) 0.0946(11) 0.0942(12) 0.0941(12)
ZO(2) (Å) 0.0961(14) 0.0975(3) 0.0979(19) 0.0988(15) 0.0994(14) 0.0993(15) 0.0992(15)
ZO(3) (Å) 0.1944(3) 0.1962(2) 0.1983(3) 0.1995(2) 0.1998(2) 0.2005(2) 0.2003(2)
Rp 5.73% 5.30% 6.18% 4.95% 4.54% 4.76% 4.83%
Rwp 8.13% 7.91% 8.45% 6.71% 6.21% 6.54% 6.59%
RB 3.57% 3.28% 3.97% 4.27% 3.75% 3.91% 3.96%
χ2 5.20 5.80 4.52 2.47 2.74 2.80 2.77
[thin space (1/6-em)]
Ion occupation distribution of A sites
Sr12+ 0.232 0.252 0.365 0.482 0.509 0.566 0.577
La13+ 0.768 0.739 0.532 0.232 0.239 0.061 0
Sm13+ 0 0.009 0.103 0.286 0.252 0.373 0.423
Sr22+ 0.768 0.747 0.634 0.517 0.491 0.433 0.423
La23+ 1.232 0.761 0.468 0.268 0.061 0.041 0
Sm23+ 0 0.492 0.898 1.215 1.448 1.526 1.577
[thin space (1/6-em)]
Bond length
(Sr, Ln)(1)–O(1) × 4 (Å) 2.6687(20) 2.6595(2) 2.6488(20) 2.6387(2) 2.6341(2) 2.6299(2) 2.6279(2)
(Sr, Ln)(1)–O(2) × 8 (Å) 2.708(2) 2.717(2) 2.711(3) 2.715(2) 2.719(2) 2.714(2) 2.710(2)
(Sr, Ln)(2)–O(3) × 1 (Å) 2.496(6) 2.451(4) 2.405(6) 2.374(4) 2.365(4) 2.352(4) 2.349(4)
(Sr, Ln)(2)–O(3) × 4 (Å) 2.6851(5) 2.6721(6) 2.6696(8) 2.6628(5) 2.6592(5) 2.6569(6) 2.6545(6)
(Sr, Ln)(2)–O(2) × 4 (Å) 2.5575(20) 2.530(2) 2.516(3) 2.492(2) 2.4793(19) 2.4766(20) 2.4762(20)
Al–O(3) × 1 (Å) 1.990(7) 2.037(5) 2.071(7) 2.087(5) 2.097(5) 2.113(5) 2.115(5)
Al–O(1) × 1 (Å) 1.927(2) 1.922(3) 1.903(3) 1.895(2) 1.884(2) 1.876(2) 1.871(2)
Al–O(2) × 4 (Å) 1.8871(4) 1.8810(10) 1.8739(15) 1.8674(16) 1.8651(18) 1.8624(2) 1.8610(2)



image file: c6ra20776f-f5.tif
Fig. 5 Lattice parameters (a, c and cell volume) of Sr(La1−xSmx)2Al2O7 ceramics with various compositions.

Crystal structure of the present ceramics is shown in Fig. 6(a) according to the refined results. From the diagrammatic sketch, there are two kinds of A sites occupied by Sr2+, La3+ and Sm3+ ions with different proportions—one is (Sr, Ln) ion with twelve coordination number (A1) in the middle of the P2 layer and the other is (Sr, Ln) ion with nine coordination number (A2) in the RS layer, and three kinds of oxygen sites—O(1), O(2), O(3) shown in Fig. 6(c). Because of the layered structure of Sr(La1−xSmx)2Al2O7 ceramics, it can be regarded as different parallel charged layers stacking alternately along the c axis, i.e., -AlO2-(Sr, La, Sm)O-(Sr, La, Sm)O–AlO2-(Sr, La, Sm)*O–AlO2-(Sr, La, Sm)O-(Sr, La, Sm)O–AlO2- (see Fig. 6(b)). Benabbas29 have handled the layers along c axis in the R–P structure with n = 1 as uniform blocks according to the superposition principle and have summarized their interlayer polarization charges. In this case, because the difference between Sr2+ and Ln3+ in charge and radius, the ratio of Sr2+ to Ln3+ at different A site is not exactly equal to 1[thin space (1/6-em)]:[thin space (1/6-em)]2 (see Table 1). Therefore, the formal charge of each layer can be obtained by eqn (2) on the basis of the corresponding formal ionic charges. The results are shown in Fig. 6(b).

 
image file: c6ra20776f-t2.tif(2)
where, C represents the polarization charge of corresponding layer, Ci and C0 represent the valence of corresponding cation and oxygen, respectively, and xi represents the proportion of corresponding cation on the basis of the ion occupation distribution shown in Table 1. In the present ceramics, every layer experiences the electric field created, as a first approximation, by the neighboring layers, and the interlayer polarization creates as a consequence of these polarization charges. Moreover, with increasing the substitution of Sm3+ ion with smaller radius, Ln3+ ions trend to occupy A2 site, while Sr2+ ion trends to occupy A1 site according to the refined results, resulting in the increasing interlayer polarization especially in RS layer.


image file: c6ra20776f-f6.tif
Fig. 6 (a) Crystal structure of Sr(La1−xSmx)2Al2O7 ceramics; (b) interlayer polarization charges of Sr(La1−xSmx)2Al2O7 ceramics with various compositions ((Sr, Ln) ion with twelve coordination number is marked by asterisk); (c) atom displacements of Sr(La1−xSmx)2Al2O7 ceramics along c axis.

In order to minimize the interlayer polarizations, the corresponding ions drove by the electric field created and the signs of their charges will move along c axis until they reach their equilibrium (real) positions. As shown in Fig. 7, the normalized bond length (NBL) is used to describe the state of cations. NBL is defined as the ratio of the actual bond length to the ideal bond length which is calculated using the effective ionic radii.30 As shown in Fig. 6(c), because the polarization charges of neighboring layers in RS layer are positive, the A2 cations move towards neighboring -AlO2- layer along c axis, on the contrary, the apical oxygen (O(3)) anions move closer along c axis. As a result, the A2 ion is in “compressed” state along c axis with NBL of A2–O(3) shorter than 1, while the Al ion is in “elongated” state along c axis with NBL of Al–O(3) larger than 1. In summary, the A1O12 polyhedron is compressed along and perpendicularly to c axis, and the AlO6 octahedron is elongated along c axis and compressed perpendicularly to c axis, while the A2O9 polyhedron is opposite to AlO6 octahedron.


image file: c6ra20776f-f7.tif
Fig. 7 Normalized bond length of Sr(La1−xSmx)2Al2O7 ceramics with various compositions.

The stability of the structure can be characterized by tolerance factor (t). The tolerance factors are calculated first from two extreme aspects based on the assumptions that the cations of A site in one perovskite block are all twelvefold (represented by tA1) or ninefold (represented by tA2) by the following eqn (3).22 Because the ratio of A1 to A2 is 1[thin space (1/6-em)]:[thin space (1/6-em)]2, the final tolerance factors are calculated by the following eqn (4).22 The result as a function of x is shown in Fig. 8. The coordination numbers and effective ionic radii of atoms are based on the research of Shannon.30

 
image file: c6ra20776f-t3.tif(3)
 
image file: c6ra20776f-t4.tif(4)
where, RSr2+, RLa3+, RSm3+, RAl3+, RO2− are the effective ionic radii of corresponding ions; x, y and z are the proportions of Sr2+, La3+ and Sm3+ ions on the basis of the ion occupation distribution shown in Table 1. As shown in Fig. 8, with increasing x, the tolerance factor decreases linearly. The stability of the structure could be related to the internal stresses. With increasing x, the distortions of most of the bonds and the three polyhedrons become serious (see Fig. 7), which can lead to the increased internal stress. As a result, the stability (characterized by the tolerance factor) decreases.


image file: c6ra20776f-f8.tif
Fig. 8 Tolerance factor of Sr(La1−xSmx)2Al2O7 ceramics with various compositions.

Fig. 9 shows the microwave dielectric characteristics of Sr(La1−xSmx)2Al2O7 ceramics sintered at 1575 °C for 3 h as functions of composition. As shown in Fig. 9(a), with increasing x, the relative dielectric constant increases slightly from 18.8 to 21.3. On the basis of the structure parameters, the molecule polarizability (α) of the present ceramics is calculated by the microscopic Clausius–Mosotti equation31 as follow:

 
image file: c6ra20776f-t5.tif(5)
where, V represents the volume per unit cell, Z represents the number of molecule per unit cell. As shown in Table 2, with increasing x, variation trend of the calculated molecule polarizability (αc) based on eqn (5) is similar to that of predicted molecule polarizability (αp) according to the oxide additivity law.32 There is a deviation between αc and αp because αp is given by ion polarizability of oxide. The cations in actual Sr(La1−xSmx)2Al2O7 structure are in “elongated” or “compressed” state which are very different from that in oxides. According to Shannon, the cation polarizability will be increased or decreased as a result of the presence of “elongated” or “compressed” state, respectively.32 In this case, most of NBL are shorter than 1, which means the αp listed in Table 2 is larger than that in the actual condition, resulting in the negative sign of Δ. In addition, the larger deviation should be related to larger volume of A1O12, AlO6 and A2O9 polyhedrons which also can reflect the surroundings of A1, Al and A2 cations.


image file: c6ra20776f-f9.tif
Fig. 9 Microwave dielectric characteristics of Sr(La1−xSmx)2Al2O7 ceramics sintered at 1575 °C in air for 3 h as functions of composition: (a) relative dielectric constant, (b) Qf value and (c) temperature coefficient of resonant frequency.
Table 2 The calculated and predicted molecule polarizability in Sr(La1−xSmx)2Al2O7 ceramics sintered at 1575 °C for 3 ha
x αp3) Observed (Z = 2) Δ (%)
εr Vunit cell3) αc3)
a Δ (%) = ((αcαp)/αc) × 100.
0 32.03 18.8 287.845 29.42 −8.88%
0.25 31.37 19.5 284.653 29.25 −7.22%
0.5 30.70 21.0 281.161 29.20 −5.14%
0.75 30.04 21.1 277.930 28.88 −3.99%
0.85 29.77 21.2 276.515 28.75 −3.53%
0.95 29.50 21.3 275.305 28.65 −2.99%
1 29.37 21.3 274.722 28.58 −2.75%


As shown in Fig. 9(b), the Qf value of Sr(La1−xSmx)2Al2O7 ceramics trends to decrease in general with increasing the substitution of Sm3+ ion but indicates a slight increment at x = 0.5. The Qf values of the first three compositions (x = 0–0.5) are much larger than that of the rest compositions (x = 0.75–1). Fan et al. have done lots of research about the effect of interlayer polarization charge on dielectric loss.16,33,34 In this case, the theoretical decreasing trend of Qf value is associated with the interlayer polarization effect and the stability of the structure (characterized by the tolerance factor). As x increases, the interlayer polarization charge increases, leading to the increasing damage to the charge balance between each layer. Besides, the degree of the actual bond length away from ideal value, namely the deviation of NBL from 1 increases, especially the bond lengths of Al–O(3), A2–O(3) (see Fig. 8) in the RS layer, resulting in the increasing internal stresses. Meanwhile, the stability of the structure decreases. In conclusion, the Qf value should decrease theoretically with increasing x.

Actually, high Qf value is maintained for x < 0.5, and the Qf value even increases slightly at x = 0.5 (Qf = 115[thin space (1/6-em)]900 GHz), then decreases rapidly at x = 0.75 (Qf = 76[thin space (1/6-em)]590 GHz). It is well known that the extrinsic factors such as grain size, secondary phase, porosity and stacking fault also affect the Qf value significantly, and even play the dominating role in some cases.6,35–38 The extraordinary high relative density at x = 0.5 (about 99% T.D.) may contribute to the slight increment at x = 0.5. Kim et al. have pointed out that the Qf value decreased obviously with the decreasing grain size.35 In this case, the smaller grain size at x = 0.75–1 (see Fig. 2) also has adverse effect on the Qf value. Meanwhile, the increasing amount and type of secondary phases with lower Qf value such as SmAlO3 (εr = 20.4, Qf = 65[thin space (1/6-em)]000 GHz, τf = −74 ppm per °C)39 could also lead to the deterioration of Qf value at x = 0.75–1. Besides, Fig. 10 shows the HRTEM images viewed along the [010] direction which is the most readily interpretable zone axis for direct illustration of the stacking fault. As shown in Fig. 10(a) and (b), compared to the ceramic at x = 0.5, the stacking faults appeared at x = 0.75 are obvious. Using FFT analysis, as shown in Fig. 10(c) and (d), the diffraction spots of x = 0.75 elongate along c axis, while that of x = 0.5 are round dots, which indicates one-dimensional (1D) disorder as the appearance of the stacking faults where the n values of the layered structure are not equal to 2.25 The stacking faults can be regarded as the secondary phases implanted to the matrix phase, resulting in the inhomogeneous microstructure of the ceramics, and they could do harm to Qf value.


image file: c6ra20776f-f10.tif
Fig. 10 TEM microanalysis for Sr(La1−xSmx)2Al2O7 ceramics (x = 0.5, 0.75). (a) and (b) HRTEM images viewed along [010] zone axis, and the insets are the corresponding selected diffraction patterns taken along the same zone axis; (c) and (d) the simulated diffraction patterns of white square areas of (a) and (b) by using FFT respectively.

As shown in Fig. 9(c), the temperature coefficient of resonant frequency (τf) increases from −21.7 to 4.8 ppm per °C with increasing x. It is correlated to the temperature coefficient of εr (τε) and the linear thermal expansion coefficient (αL) through the following equation.3

 
image file: c6ra20776f-t6.tif(6)

For the perovskite-type ceramics, the effect of αL can be ignored because it is almost a constant about 8 to 10 ppm per °C.40 Therefore, τf is mainly affected by τε which is related to εr. According to Reaney et al.,3 the variation of τε with t is strongly affected by the structural phase transition. However, no diffraction spot which represents inphase or antiphase tilting of oxygen octahedron is observed (see Fig. 10), even though the t is less than 0.985. Consequently, τε decreases with the decreasing t, resulting in the increment of τf according to eqn (6), and the relationship between εr and τf is monotonous. Therefore, the τf value of SrLa2Al2O7 ceramic can be adjusted to near-zero by Sm3+ ion substitution.

Conclusions

Low-loss and temperature-stable Sr(La1−xSmx)2Al2O7 solid solution ceramics with R–P structure of n = 2 can be prepared through a standard solid-state sintering method. The crystal structure of matrix phase evolves linearly with x. With increasing Sm3+ substitution, εr value increases slightly, while τf value can be adjusted from negative towards positive. Though the Qf value should theoretically decrease with increasing interlayer polarization charges which is influenced by the enhancive occupation ratio of Ln ions in the RS layer, the extrinsic factors such as large grain size and high relative density could benefit to the Qf value, resulting in the high Qf value for x < 0.5 and even leading to a slight increment at x = 0.5. The best combination of microwave dielectric characteristics is achieved at the composition of x = 0.5: εr = 21.0, Qf = 115[thin space (1/6-em)]900 GHz and τf = −3.7 ppm per °C, making this material potential for commercial microwave applications.

Note added after first publication

This article replaces the version published on 10th October 2016, which, owing to an editorial office error, referred to the product of the quality factor (Q) and the frequency (f) by the term Qf rather than the correct form Qf.

Acknowledgements

The present work was financially supported by the Chinese National Basic Research Program under grant number 2015CB654601.

References

  1. T. A. Vanderah, Science, 2002, 298, 1182 CrossRef CAS PubMed.
  2. W. Wersing, Curr. Opin. Solid State Mater. Sci., 1996, 5, 715 CrossRef.
  3. I. M. Reaney and D. Iddles, J. Am. Ceram. Soc., 2006, 89, 2063 CAS.
  4. M. T. Sebastian, Dielectric Materials for Wireless Communication, Elsevier Science Publishers, Oxford, 2008 Search PubMed.
  5. C. L. Huang, C. L. Pan and S. J. Shium, Mater. Chem. Phys., 2002, 78, 111 CrossRef CAS.
  6. N. Ichinose and T. Shimada, J. Eur. Ceram. Soc., 2006, 26, 1755 CrossRef CAS.
  7. H. Matsumoto, H. Tamura and K. Wakino, Jpn. J. Appl. Phys., 1991, 30, 2347 CrossRef CAS.
  8. S. F. Wang, Y. R. Wang, C. Y. Liu and W. S. Hsieh, Ceram. Int., 2012, 38, 1127 CrossRef CAS.
  9. M. R. Varma and M. T. Sebastian, J. Eur. Ceram. Soc., 2007, 27, 2827 CrossRef CAS.
  10. E. S. Raymond and E. M. Thomas, Chem. Mater., 2002, 14, 1455 CrossRef.
  11. C. Wan, T. D. Sparks, P. Wei and D. R. Clarke, J. Am. Ceram. Soc., 2010, 93, 1457 CrossRef CAS.
  12. G. Amow, I. J. Davidson and S. J. Skinner, Solid State Ionics, 2006, 177, 1205 CrossRef CAS.
  13. Y. Moritomo, A. Asamitsu, H. Kuwahara and Y. Tokura, Nature, 1996, 380, 141 CrossRef CAS.
  14. J. Feng, B. Xiao, R. Zhou and W. Pan, J. Appl. Phys., 2013, 113, 143907 CrossRef.
  15. K. I. Shimizu, S. Itoh, T. Hatamachi, T. Kodama, M. Sato and K. Toda, Chem. Mater., 2005, 17, 5161 CrossRef CAS.
  16. X. C. Fan, X. M. Chen and X. Q. Liu, Chem. Mater., 2008, 20, 4092 CrossRef CAS.
  17. H. X. Yuan, X. M. Chen and M. M. Mao, J. Am. Ceram. Soc., 2009, 92, 2286 CrossRef CAS.
  18. M. M. Mao, X. C. Fan and X. M. Chen, Int. J. Appl. Ceram. Technol., 2010, 7, E156 CrossRef.
  19. B. Liu, L. Yi, X. Q. Liu and X. M. Chen, J. Electroceram., 2015, 34, 114 CrossRef CAS.
  20. B. Liu, L. Li, X. Q. Liu and X. M. Chen, J. Mater. Chem. C, 2016, 4, 4684 RSC.
  21. P. L. Wise, I. M. Reaney, W. E. Lee, T. J. Price, D. M. Iddles and D. S. Cannell, J. Eur. Ceram. Soc., 2001, 21, 1723 CrossRef CAS.
  22. L. Yi, X. Q. Liu and X. M. Chen, Int. J. Appl. Ceram. Technol., 2013, 10, E177 CrossRef CAS.
  23. I. Zvereva, Y. Smirnov and J. Choisnet, Int. J. Inorg. Mater., 2001, 3, 95 CrossRef CAS.
  24. I. A. Zvereva, S. R. Seitablaeva and Yu. E. Smirnov, Russ. J. Gen. Chem., 2003, 73, 31 CrossRef CAS.
  25. L. Yi, L. Li, X. Q. Liu and X. M. Chen, J. Am. Ceram. Soc., 2014, 97, 3531 CrossRef CAS.
  26. B. Liu, X. Q. Liu and X. M. Chen, J. Mater. Chem. C, 2016, 4, 1720 RSC.
  27. B. W. Hakki and P. D. Coleman, IRE Trans. Microwave Theory Tech., 1960, 8, 402 CrossRef.
  28. X. C. Fan, X. M. Chen and X. Q. Liu, IEEE Trans. Microwave Theory Tech., 2005, 53, 3130 CrossRef.
  29. A. Benabbas, Acta. Cryst. B, 2006, 62, 9 CrossRef PubMed.
  30. R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751 CrossRef.
  31. S. Roberts, Phys. Rev., 1951, 81, 865 CrossRef CAS.
  32. R. D. Shannon, J. Appl. Phys., 1993, 73, 348 CrossRef CAS.
  33. X. C. Fan, M. M. Mao and X. M. Chen, J. Am. Ceram. Soc., 2008, 91, 2917 CrossRef CAS.
  34. X. C. Fan and X. M. Chen, J. Am. Ceram. Soc., 2009, 92, 433 CrossRef CAS.
  35. E. S. Kim and K. H. Yoon, J. Mater. Sci., 1994, 29, 830 CrossRef CAS.
  36. C. L. Huang, M. H. Weng and H. L. Chen, Mater. Chem. Phys., 2001, 71, 17 CrossRef CAS.
  37. B. Liu, L. Yi and X. M. Chen, Mater. Res. Bull., 2015, 67, 230 CrossRef CAS.
  38. M. M. Mao, X. M. Chen and X. Q. Liu, J. Am. Ceram. Soc., 2011, 94, 3948 CrossRef CAS.
  39. S. Y. Cho, I. T. Kim and K. S. Hong, J. Mater. Res., 1999, 14, 114 CrossRef CAS.
  40. H. S. Park, K. H. Yoon and E. S. Kim, J. Am. Ceram. Soc., 2001, 84, 99 CrossRef CAS.

This journal is © The Royal Society of Chemistry 2016
Click here to see how this site uses Cookies. View our privacy policy here.