A. Results from SEM and TEM

SEM, TEM, and SAED were used to reveal the influence of the base/Zn²⁺ molar ratio and base type on the surface morphologies and microstructures of the samples. As shown in Fig. 1(a), using NaOH as base with α = 1.0, the as-prepared products were 0.5- to 1.0-µm diameter microspheres consisting of nanoparticles with several tens of nanometers in size. When α was increased to 3.0, the products were monodispersed and uniform microspheres were formed with a diameter ranging from 4 to 5 µm [Fig. 1(b)]. These microspheres were composed of numerous densely arranged nanosheets that radially spread and formed well-defined flower-like architectures. Periodical lattice stripes can be observed in almost the whole HRTEM visual field [Fig. 1(e)]. The crystal plane distance was measured with a length of 0.52 nm, which corresponded to the [001] direction of ZnO. The strong and clear diffuse spots in the SAED pattern [inset in Fig. 1(e)] illustrate the single crystal nanosheets.

FIG. 1.

(a–d) SEM and (e–g) TEM images of polymorphous ZnO samples produced at various base/Zn²⁺ molar ratios (α) and base types. (a) Spherical aggregates, α = 1.0 (NaOH/Zn²⁺); (b and e) nanosheet-based flowers, α =3.0 (NaOH/Zn²⁺); (c and f) microrod-composed flowers, α =1.0 (en/Zn²⁺); and (d and g) nanopetal-built flowers, α = 2.0 (NH₃·H₂O/Zn²⁺). The insets in e, f, and g are the corresponding SAED patterns.

When en was used at an en/Zn²⁺molar ratio of 1.0, the hydrothermal precipitates formed a radial structure with microrods projected outwards, exhibiting a flower-like configuration [Fig. 1(c)]. The diameters of the flowers ranged from 8.0 to 10.0 µm. These microrods had aspect ratios of approximately 5.0, and their diameters slightly decreased from 1.0 µm on the body to approximately 600 nm at the tip. The sharp diffraction spots in the SAED pattern [inset of Fig. 1(f)], taken from individual ZnO microrods [Fig. 1(f)], further demonstrated that the [001] direction was the preferred growth direction of crystal formation with a perfect single crystalline nature.

Substituting NaOH with NH₃·H₂O still resulted in the flower-like shape of the ZnO samples, with diameters of about 1.0 µm to 1.5 µm [Figs. 1(d) and 1(g)]. Each flower comprised several petals approximately 1.0 in aspect ratio, radially projecting from one center. The surfaces of the ZnO flowers were not smooth and covered by large amounts of nanoparticles. The irregular SAED pattern in the inset of Fig. 1(g) indicated the polycrystalline nature of the nanopetal and the existence of shear defects. We showed in our previous study ^{Reference Hu, Tong, Wu, Liu, Qian, Guan and Hong6,Reference Tong, Du, Liang, Hu, Wu, Guan and Hu7} that the initially formed nanoparticles can rapidly grow preferentially along the [001] direction and subsequently act as “pioneer crystals” to direct the formation of the nanopetals. The nanoparticles that grew on the nanopetal surface eventually remained parallel to the [001] direction by rotating under hydrothermal conditions. In the present study, the growth of the flower-like ZnO architectures was due to the polar surface-induced oriented attachment growth. When various bases and base/Zn²⁺ molar ratios were introduced to the reaction system, different coordination agents react with Zn²⁺ ions to form zinc coordination compounds with various unstable constant values. These variations affected the attachment and growth behavior of [Zn(OH)₄]²⁻ on the (001) polar planes as well as the nucleation and growth rates of ZnO nanocrystals. ^{Reference Hu, Tong, Wu, Liu, Qian, Guan and Hong6,Reference Tong, Du, Liang, Hu, Wu, Guan and Hu7,Reference Li, Xiong and Xie15}

B. XRD analysis

XRD was used to further study the effects of the base/Zn²⁺ molar ratio and base type on phase structure, intensity ratio of I ₍₁₀₀₎/I ₍₀₀₂₎, crystal size, and internal strain of the products. The XRD patterns in Fig. 2(a) showed that all peaks can be indexed to the hexagonal ZnO with wurtzite structure (space group P6 ₃ mc; a = 3.2498 Å, c = 5.2066 Å; JCPDS 36-1451). No characteristic impurity peaks, such as ZnCO₃·2Zn(OH)₂·H₂O, were observed, indicating that all samples were pure ZnO phases. However, the change in the α and base types initiated an increase in intensities of all diffraction peaks at various degrees. This phenomenon is generally related to the enhanced crystallization, formation of large crystallites, or well-ordered orientation. ^{Reference Tong, Guan and Zhang16} From Fig. 2(b), NaOH displayed the highest intensity ratio of I ₍₁₀₀₎/I ₍₀₀₂₎ among the bases used in this study while keeping the stoichiometric ratio of Zn²⁺ to OH^– unchanged (i.e., at 2.0). With NaOH as base, the intensity ratio of I ₍₁₀₀₎/I ₍₀₀₂₎ peaked at α = 2.0. In general, a small I ₍₁₀₀₎/I ₍₀₀₂₎ ratio corresponds to an intense (002) peak, indicating the preferential growth of rods along the c-axis with (001) or (00 $\overline 1$ ) end faces. Conversely, a very large I ₍₁₀₀₎/I ₍₀₀₂₎ ratio indicated the prior growth of rods along the a-axis ([100] direction). ^{Reference Mclaren, Valdes-Solis, Li and Tsang17} Such variations implied that changing bases and base/Zn²⁺ molar ratios (α) could modulate the growth direction of ZnO nanocrystals. The average crystallite sizes and the internal strain of the samples were determined according to the Hall–Williamson equation: ${\rm{\beta }}\cos {\rm{\theta }} = {\textstyle{{K{\rm{\lambda }}} \over {\rm{\delta }}}} + {\rm{2\varepsilon si}}n{\rm{\theta }}$ , where β is the full width at half-maximum (FWHM) of the samples, λ is the x-ray wave length used, δ is the grain size, K is Scherrer constant, ε is the internal strain, and θ is Bragg angle. β can be given as ${{\rm{\beta }}^2} = {\rm{\beta }}_{\exp }^2 - {\rm{\beta }}_{{\rm{inst}}}^{\rm{2}}$ , where β_inst is the FWHM of the Si powder used for calibration and β_exp is the evaluated FWHM. Figures 2(c)and 2(d) show that microrod- and nanosheet-based flowers had larger crystal size and lower internal strain than spherical aggregates and nanopetal-built flowers. The data indicated that increasing α and using en or NaOH as the base enhanced crystal size and crystallinity.

FIG. 2.

(a) XRD patterns, (b) intensity ratio of I ₍₁₀₀₎/I ₍₀₀₂₎, (c) average crystal size, and (d) internal strain of polymorphous ZnO samples formed at various base/Zn²⁺ molar ratios (α) and base types.

C. BET study

In addition to the morphological and structural modulation, base/Zn²⁺ molar ratio and base type also considerably influenced the specific surface area (S _BET) of the products. The BET S _BET was determined by a multipoint BET method using the adsorption data in a relative pressure (P/P ₀) range from 0.05 to 0.25. Figure 3 shows that the S _BET of the spherical aggregates was as large as 15.60 m²/g, which was much higher than those of nanosheet-based-flowers (11.60 m²/g), microrod-composed flowers (1.68 m²/g), and nanopetal-built flowers (7.73 m²/g). The difference was mainly ascribed to the mode of assembly and crystal size of nanocrystals.

FIG. 3.

Specific surface area (S _BET) of the samples obtained at the various base/Zn²⁺molar ratios (α) and bases.

D. Dielectric properties

Figure 4 showed the frequency dependence of the relative complex permittivity and dielectric loss for the wax-composites containing the various mass fractions and morphologies of ZnO. As shown in Figs. 4(a)–4(c), the ε′, ε″, and tan δ_E values of the nanosheet-built flowers fluctuated within small ranges of 2.1–3.4, 0–0.5, and 0–0.17, respectively. As the mass fraction increased, the ε′ value gradually increased; the ε″ and tan δ_E values also showed less variation from 20 to 40% with a steep increase at 60%. This sharp variation may be related to the formation of the percolation semiconductive network in the matrix under a high mass fraction. ^{Reference Tong, Wu, Hua, Miao, Guan and Qian18} Contrastive analysis (Fig. 4) revealed that the ε′, ε″, and tan δ_E values of nanosheet-built flowers were almost as same as those of the spherical aggregates and nanopetal-composed flowers and conspicuously lower than those of microrod-based flowers. Thus, the flowers composed by high aspect ratio microrods had a higher capacity for storage and loss of electrical energy.

FIG. 4.

Frequency dependence of the real (ε′) and imaginary (ε″) parts of the complex permittivity and the dielectric loss (tan δ_E = ε″/ε′) for the ZnO wax composites: (a–c) various mass fractions of the nanosheet-built flowers and (d) 40% ZnO spherical aggregates, (e) microrod-composed flowers, and (f) nanopetal-built flowers.

The microwave responses of complex nano/microstructures are usually associated with dielectric relaxation, microcurrent attenuation, interface scattering, and microantenna radiation. ^{Reference Fang, Cao, Shi, Hou, Song and Yuan19} First, the dielectric relaxation is a result of the orientation, interfacial, atomic, and space charge polarizations. These types of polarizations strongly depend on the material shapes and microstructures. ZnO is considered as a dielectric loss medium, and its dielectric relaxation is usually related to the characteristic of its polar surfaces. The ZnO polar monomer contains oppositely charged ions. The positively charged (001) Zn-polar and negatively charged (00 $\overline 1$ ) O-polar surfaces generate a normal dipole moment and spontaneous polarization along the c axis. Under an alternating EM field, the electrons from the polar planes or oxygen vacancies migrate back and forth to form relaxation polarization and loss, respectively. ^{Reference Huang, Cao, Li, Jiang, Wang and Li8} According to our previous studies, ^{Reference Tong, Ma, Wu, Hua, Qiao and Qian20} the increased ε′, ε″, and tan δ_E values were due to the enhancement of space charge polarization from the increasing defects, interface polarization from increasing surface area, or orientation polarization from significantly increasing aspect ratio. Thus, the ZnO samples with large S _BET and high internal strain (corresponding to high defects) should display large ε′, ε″, and tan δ_E values. But this was not observed in the permittivity of nanopetal-built flowers and spherical aggregates. Therefore, it can be inferred that the orientation polarization from the significantly increasing aspect ratio is one of the crucial factors for the enhanced permittivity of the microrod-based flowers.

Second, continuous semiconductive micronetworks in the matrix may induce large-scale vibrating microcurrent attenuation when the EM wave travels into the ZnO wax composites. The mass fraction and shape of the ZnO particles are the two key factors in the formation of the nets. Compared with the nanosheet-built flowers, spherical aggregates, and nanopetal-built flowers, the microrod-based flowers with the large aspect ratio are more prone to form such nets in the matrix by linking with one another at a low mass fraction. The nets can also form by elevating the mass fraction of the ZnO particles within certain limits. Flower-like architectures randomly distributed in the matrix can also generate multiple interface scattering and further increase the attenuation of the EM wave. The multiple interface scattering strengthens as the mass fraction and the interface increase.

Third and last, discontinuous flower-like architectures can act as isotropic quasi-antennas, which produce microantenna radiation. The behavior induces EM energy into becoming a dissipative current. With the increasing mass fraction, the discontinuous flower-like architectures decrease and the continuous micronetworks gradually form. Thus, the former decreases the microantenna radiation and the latter enhances the microcurrent attenuation. Overall, increasing mass fraction from 20 to 40% leads to the similar permittivity, as a result of the cooperative effect of the reduced microantenna radiation, the increased interface scattering and dielectric relaxation. However, steeply increased permittivity in the high mass fraction and shape of ZnO microrod-based flowers results from the microcurrent attenuation and interface scattering.

The ε″ and tan δ_E plots presented highly complex nonlinear resonant behaviors relative to the EM energy loss, with a strong and wide peak over 10–14 GHz but low resonant peaks of 2–5, 6–8, and 16–18 GHz. Similar multiresonance behaviors have been found in ZnO nanocombs, ^{Reference Zhuo, Feng, Liang, Liu, Chen, Yan, Feng, Li, Cheng, Geng, Xu, Wang, Wu, Yan and Yue9} ZnO dendritic nanostructures, ^{Reference Zhuo, Feng, Chen, Yan, Feng, Li, Geng, Cheng, Xu and Yan11} nanotrees, ^{Reference Zhuo, Qiao, Feng, Chen, Yan, Wu and Yan21} and micro/nanorod networks, ^{Reference Li, Huang, Sun, Yan, Yang, Wang and Zhang13} which were generally associated with highly conductive and skin effects, ^{Reference Langarkov and Sarychev22} electronic spin and charge polarizations caused by point effects, and polarized centers. ^{Reference Watts, Hsu, Barnet and Chambers23,Reference Che, Peng, Duan, Chen and Liang24} In the present study, two weak resonant peaks at 6–8 and 16–18 GHz were derived from microantenna radiation, whereas a strong wide peak at approximately 12 GHz was caused by the cooperative action of interface scattering, microcurrent attenuation, and dielectric relaxation. ^{Reference Fang, Cao, Shi, Hou, Song and Yuan19} Obviously, a strong resonance peak usually corresponded to a strong reflection loss, which was exhibited by high mass fraction and/or the shape of microrod-composed flowers (Fig. 4).

E. Microwave absorption properties

Figure 5 shows the R _L curves of the wax composites containing the flower-like ZnO powders with various mass ratios and morphologies. Several evident resonance absorbing peaks were observed on all R _L plots of the ZnO wax composites including 20 wt% nanosheet-built ZnO flowers [Fig. 5(a)]. A strong and broad peak was located at 11.0 GHz and two weak ones were situated at approximately 3.0 and 17.0 GHz, respectively. As the sample thickness increased from 1.0 to 6.0 mm, the intensities of three absorption peaks gradually increased and their positions remained almost the same (i.e., no movement to lower frequencies). The nesting absorption peaks also still existed not only at 40 and 60 wt% [Figs. 5(a)–5(c)] but also in spherical aggregates and nanopetal-built flowers [Figs. 5(d)–5(f)]. The peaks gradually broadened and their positions slightly moved to a higher frequency with the increased mass fraction [Figs. 5(a)–5(c)].

FIG. 5.

Reflection loss curves for the ZnO wax composites including various mass fractions and morphologies of ZnO powders. (a–c) Various mass fractions of the nanosheet-built ZnO flowers and (d) 40% ZnO spherical aggregates, (e) nanopetal-built flowers, and (f) microrod-composed flowers.

A similar phenomenon has been reported in spherical Co, hollow Co nanochains, and MWCNTs/ZnO samples, which were ascribed to the effect of dual nonlinear dielectric resonance and negative permeability ^{Reference Shi, Cao, Yuan and Fang25} or the resonant behavior caused by interface and skin effect. ^{Reference Shi, Cao, Yuan and Fang25} In our previous studies, urchin-like Fe₂O₃ took on a nesting absorption peak at approximately 16.5 GHz. ^{Reference Tong, Wu, Guan, Qian, Yuan and Li26} Three nesting absorbing peaks at approximately 12.5, 14.8, and 16.5 GHz can be observed in flower-like Co superstructures, originating from frequency-selective surface-absorption properties. ^{Reference Tong, Yuan, Wu, Hu, Qian, Li and Shen27} The present data showed that the positions of multiresonance absorbing peaks were unexpectedly similar to the resonance peaks on ε″ and tan δ_E plots. This result indicated that the microwave absorption of ZnO flowers originated from the multiresonance loss. By contrast, the absorption peaks of microrod-composed flowers did not remain at a stable frequency but gradually moved to lower frequencies with increased coating thickness. The change is due to the resonant absorption caused by the quarter-wave length attenuation when the reflected waves and incident waves meet in a material. ^{Reference Song, Cao, Wen, Hou, Cheng and Yuan28} Compared with dielectric relaxation and microantenna radiation, microcurrent attenuation and interface scattering are highly important microwave responses for improving the dielectric and microwave absorption properties of ZnO architectures. Obviously, a flower-like architecture built by large aspect ratio microrods is a promising configuration for designing novel microwave absorbing materials.