SPS sintering of nanocrystalline zinc oxide—Part II: Abnormal grain growth, texture and grain anisotropy

SPS sintering of nanocrystalline zinc oxide—Part II: Abnormal grain growth, texture and grain anisotropy

Journal of the European Ceramic Society 36 (2016) 1221–1232 Contents lists available at www.sciencedirect.com Journal of the European Ceramic Societ...

3MB Sizes 0 Downloads 124 Views

Journal of the European Ceramic Society 36 (2016) 1221–1232

Contents lists available at www.sciencedirect.com

Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc

FAST/SPS sintering of nanocrystalline zinc oxide—Part II: Abnormal grain growth, texture and grain anisotropy Benjamin Dargatz a,c,∗ , Jesus Gonzalez-Julian a , Martin Bram a , Yutaka Shinoda b , Fumihiro Wakai b , Olivier Guillon a a

Institute of Energy and Climate Research, Materials Synthesis and Processing, Forschungszentrum Jülich, Wilhelm-Johnen-Straße, 52425 Jülich, Germany Secure Materials Center, Materials and Structures Laboratory, Tokyo Institute of Technology, Yokohama, Kanagawa 226-8503, Japan c Otto Schott Institute of Materials Research (OSIM), Friedrich-Schiller-University of Jena, 07743 Jena, Germany b

a r t i c l e

i n f o

Article history: Received 30 September 2015 Received in revised form 6 December 2015 Accepted 11 December 2015 Available online 21 December 2015 Keywords: Grain boundary complexion Nano-crystalline zinc oxide Morphological anisotropy Crystalline texture Abnormal grain growth

a b s t r a c t This second part describes the retention of nanocrystallinity during sintering of ZnO by means of Fieldassisted Sintering Technique/Spark-Plasma-Sintering (FAST/SPS), whereas the first part [doi: 10.1016/ j.jeurceramsoc.2015.12.009] concentrated on hydroxide-ion-enhanced densification and defect stoichiometry. Interface design by surface bound water on zinc oxide offers a novel method to control in a new way diffusion in nanocrystalline polycrystals. Therefore, zinc oxide powder was humidified or dried and afterwards heated quickly (100 K/min) by FAST/SPS. Interestingly, the densification is strongly promoted in presence of water reducing the sintering temperature to 400 ◦ C. Thus, grain growth is decreased by one order of magnitude while achieving full densification. The crystalline texture developed irrespective of temperature or presence of water. Moreover, the formation of hydroxide complexion at grain boundaries is discussed as it might modify grain boundary mobility and lead to pronounced grain anisotropy perpendicular to the uniaxial applied load. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Since the majority of physical properties are affected by the residual porosity of the material, it is essential to obtain nearly full densification. Indeed, the sintering of nanocrystalline dense bulk materials with grain sizes below 100 nm is still a challenging task [1]. In general, the retention of nanocrystallinity depends on the competition between densification and grain growth, which driving forces both depend on reciprocal grain size [2,3]. Typically, grain growth is suppressed by the application of an external pressure [4], grain boundary pinning (addition of a second phase/doping), very high heating rates [5–7] or two-step sintering [4,8]. Nevertheless, extensive coarsening takes place for sintering of pure ZnO, resulting into either significant residual porosity [9] or large grain size of microns [4,10]. Moreover, grain boundary mobility is decreased by a large amount of small pores, resulting into a pinning force on the boundary. Increasing particle agglomeration will raise the mean pore size due to the formation of intra-agglomerate pores, while

∗ Corresponding author at: Institute of Energy and Climate Research, Materials Synthesis and Processing, Forschungszentrum Jülich, Wilhelm-Johnen-Straße, 52425 Jülich, Germany. E-mail address: [email protected] (B. Dargatz). http://dx.doi.org/10.1016/j.jeurceramsoc.2015.12.008 0955-2219/© 2015 Elsevier Ltd. All rights reserved.

decreasing the relative amount of pores at grain boundaries [11,12]. Thus, the retention of nanocrystallinity also depends on a high green density and homogeneity of the material [2,3]. In this context it is worth to mention that dry molding of powders is typically restricted to 50–61%TD [4–14], but aqueous casting, e.g., pressure filtration [8], offers much higher green density by benefiting from reduced friction between particles and prevent large agglomerates [2,15]. Moreover, water is typically chemically adsorbed by the oxide surface and forms hydrated interface [16]. Recently, Quach et al. [17] suggested that driving forces for grain coarsening are decreased by the adsorption of hydroxides onto the oxide surface due to a reduction of surface energy. In order to achieve temperatures while water is still participating at densification process, high heating rates are required. These high heating rates (>100 K/min) can be achieved by Field-Assisted Sintering Technology (FAST), which is also referred to as Spark-Plasma-Sintering (SPS) [18]. The present study will highlight the influence on adsorbed water on the microstructural development of nanocrystalline zinc oxide. Interestingly, the presence of adsorbed water on nanocrystalline ZnO strongly promotes diffusion resulting into significant decrease of sintering temperatures necessary for full densification, which offers a reduction of grain boundary mobility [19]. Furthermore, the growth of polycrystalline microstructure may result in the for-

1222

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

mation of texture, which affects electrical [20–23] and mechanical [24] properties. In general, texture is a preferred alignment of grains or crystallographic axis [22], but crystalline texture and morphologic texture (anisotropy of grain shape) do not necessarily need to coincide [23]. Indeed, the formation of texture during coarsening is promoted by anisotropy of surface energy. ZnO exhibits a higher surface energy for (0 0 0 2) plane compared to (1 0 1¯ 0) or (1 0 2¯ 0) planes, due to its polar crystal structure [25–27]. It is noteworthy that Dillon and Harmer [28,29] investigated characteristic structure and composition at grain boundaries. These so called complexions at grain boundaries exhibit characteristic grain boundary mobility [30]. Moreover, different types of complexions can co-exist and modify the grain boundary mobility of specific grains, which delivers a reasonable explanation for abnormal grain growth [31]. Therefore, investigations on interface engineering e.g., by complexions is declared to provide one of the most interesting challenges for ceramic research for the next decade [32]. Moreover, ZnO is known for preferred growth direction along the polar c-axis of the crystal lattice offering rod-like particles [33–35] or structured thin films [36]. However, countless sintering studies of pure ZnO showed that polyhedral grains with quasi globular shape always result from isotropic (spherical) powder particles [4,10,37–42]. This isotropic morphology was found to be independent from sintering process (free sintering, hot pressing, FAST/SPS) or initial particle size (nano to macro scale). In contrast, anisotropic grain growth for polycrystalline ZnO is known to occur only for sintering of powders which originally contain a certain fraction of non-spherical particles [43]. Here, for pure ZnO no study exists that reports the development of grain anisotropy from isometric powder particles. From the application point of view a texture will affect the electrical [20] and optical [44] properties, whereas anisotropy of electrical properties seems promising for ceramic in electric industry [20,23]. While part I [19] concentrated on the effect of surface bound water on the densification behavior and defect stoichiometry, the current part II highlights the microstructural development of hydroxide doped ZnO during sintering.

2. Experimental procedure 2.1. Materials and methods The standard ZnO powder utilized for the experiments is referred as NG20 (Nanogate AG, Quierschied- Göttelborn, Germany). The NG20 data sheet announces a purity of >99.99 wt% and a primary particle size of 20–50 nm. The as received zinc oxide powder was either humidified or dried directly before sintering in order to investigate the effect of bound water on the microstructural behavior. For that purpose, a total amount of 3 g zinc oxide powder was stored in a glass beaker inside an environmental chamber (KBF 240, Binder GmbH, Tuttlingen, Germany) at 20 ◦ C with 14 g/m3 of moisture or the powder was dried in a drying cabinet at 120 ◦ C with «0.5 g/m3 («0.1% relative humidity). Afterwards, the stored powder was directly purred inside a FAST/SPS graphite die with an inner diameter of 20 mm and further precompacted for 1 min at 50 MPa by uniaxial pressing. The sintering study of nanocrystalline ZnO is mainly focusing on sintering by a FAST/SPS device (HP-D5, FCT Systeme, Rauenstein, Germany). The inner part of the graphite die (type 2233, Mersen, Paris, French) and the graphite punches were covered with a 0.4 mm thick graphite sheet in order to improve electrical contact and to prevent reaction between the compact and the tool. Moreover, a low density graphite dumping covered the graphite die to reduce the thermal gradient during sintering. The sintering chamber was evacuated to 1 mbar, flushed with pure Argon (99.999% purity) and evacu-

ated again before the start of the temperature schedule in order to remove the residual foreign gas species. The sintering conditions were set with an external pressure of 50 MPa, 100 K/min heating rate and maximum sintering temperatures of 400 ◦ C and 800 ◦ C under aqueous and dry conditions, respectively. The temperature measurement and control was performed with a type K thermocouple, which was placed in a radial hole at a distance of 5 mm to the ZnO body. The pulsing pattern of the electric current was set to 25 msec: 5 msec (on: off). Moreover, the relative sintering density could be determined as a function of temperature and processing time by measurement of the axial displacement, as sintering geometry is constrained in radial direction due to the die and shrinkage could only occur in axial direction. Therefore, thermal and mechanical expansion/compression had to be considered by correcting the measured displacement by the same thermal and compressive schedule with a fully densified specimen. Here, accuracy of displacement measurement is ±10 ␮m. In addition, single experiments were performed with electrically insulation of the ceramic body, in order to investigate the effect of electrical current on grain growth behavior. Therefore, the tool setup was modified by installing alumina discs (Rubalit 710 Alumina, CeramTec GmbH, Plochingen, Germany) between NG20 body and graphite punches. For the insulating alumina discs the data sheet of the producer announces an electrical resistance of 1013 cm and 109 cm at 25 ◦ C and 900 ◦ C, respectively. 2.2. Microstructural characterization The microstructural analysis on sintered ZnO specimen (only NG20) focused on two aspects: grain shape and crystalline orientation. Sintered disc specimens were cut in bar geometry (18 mm × 1.5 mm × 1.5 mm), which allowed to evaluate microstructural features parallel and perpendicular to uniaxial applied pressure, which are further referred to z- (axial) and x, ydirections (radial), respectively. The bars were grinded with SiC paper and finally mirror polished using diamond paste (MasterPolish, Buehler, Minnesota, USA). Transmission electron microscopy (TEM, JEM-3010 JEOL Ltd., Tokyo, Japan) and high-resolution scanning electron microscopy (HRSEM, Auriga60, Carl Zeiss AG, Oberkochen, Germany) were performed in order to evaluate size distribution and orientation of grains and pores. The grain size measurement was carried out on polished samples using the line interception method with the software Lince (v. 2.31, Ceramics Group, TU Darmstadt) using a factor of 1.56 [45], evaluating at least 300 grains in different micrographs. In addition, grain size and shape were evaluated by means of ImageJ software (v1.45s, National Institute of Health, USA) applying threshold value method at 1000–1400 grains per distribution. For shape evaluation, each grain was fitted by ellipses of the same orientation, area and ratio of length to width, which enables to determine the aspect ratio and the orientation angle (˛) referring to specimen geometry. High resolution transmission electron microscopy (HRTEM) was performed with a field emission transmission electron microscope (JEM2100F, JEOL Ltd., Tokyo, Japan). Secondly, crystalline texture was investigated by electron beam scattering diffraction (EBSD) and X-ray diffraction (XRD) by means of Rietveld refinement. The crystallographic phase of sintered specimen was characterized by means of X-ray diffraction (XRD). XRD measurements were carried out using a D8-Discover (Bruker AXS, Billerica, USA) with Cu-K␣ radiation at  = 1.54056 Å, operated at 40 kV and 40 mA, a step size of 0.02◦ and a counting time of 1.6 s. Furthermore, the crystallite size was also determined by means of Scherrer analysis for partially sintered specimen, which showed a crystal size smaller than 100 nm by SEM. Here, the (1 0 1¯ 0), (1 0 2¯ 0) and (0 0 0 2) Bragg peaks were fit with the Pseudeo-Voigt function, a shape factor K = 0.94 was applied to Scherrer equation [46] and

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

full width half maximum was corrected in respect to instrumental broadening. Furthermore, Rietveld-refinement [47] was carried out in order to determine the degree of texture considering the March–Dollase parameter [48]. EBSD measurement was performed on polished specimen at an acceleration voltage of 15 keV with an angle of 70◦ and a step size of 50 nm, considering 300–500 grains for each distribution. 3. Results and discussion The aim of the present study is to achieve full densification of ZnO, retaining the nano-grain size. Fully densified samples were obtained after sintering at 400 ◦ C and 800 ◦ C under aqueous and dry conditions, respectively. Interestingly, densification behavior of nanocrystalline ZnO is strongly promoted by the presence of surface bound water, which is discussed in part I of this paper [19]. As a main benefit, grain boundary mobility is expected to be drastically reduced at a sintering temperature of 400 ◦ C, while full densification. In contrast, densification is stagnating at the same temperature in absence of bound water. Thus, much higher sintering temperatures are necessary for full densification under dry condition. 3.1. Coarsening of microstructure The microstructural development under both conditions was evaluated by high resolution scanning electron microscopy (HRSEM). Fig. 1 shows the sintering path for aqueous (Tmax = 400 ◦ C) and dry (Tmax = 800 ◦ C) sintered nanocrystalline ZnO (NG20) under constant uniaxial pressure of 50 MPa. For that purpose, grain size (G) was normalized by a mean grain size of 58.8 nm (G/G = 72%TD ) at a sintering density of 72%TD. Up to a sintering density of 90%TD the mean grain size develops independent of the presence of bound water or sintering schedule. Above 90%TD, grain size increases drastically under dry condition (G/G = 72%TD = 43), whereas in presence of bound water the coarsening is very moderate (G/G=72%TD = 3.8). However, it is noteworthy that a much higher sintering temperature of 800 ◦ C is necessary to achieve nearly full densification in absence of bound water. Typically, sintering temperature of 800–1300 ◦ C are reported for full densification of ZnO [4,10,37,39,42,49]. In sum, an increase in sintering temperature causes excessive coarsening which explains the main difference between both sintering conditions. Interestingly, sintering in presence of water at 800 ◦ C will result into the same grain size, as sintering in absence of water. However, the overall coarsening can be defined as grain growth factor, with the ratio of final grain size to initial primary particle size. Hence, aqueous sintering results into a grain growth factor of 7.3 with a final mean grain size of 220 nm, which is the smallest grain size reported so far for dense polycrystalline bulk ZnO. In Fig. 2 the grain growth factor is given as a function of final sintering density for several reference values for free sintering and pressure assisted sintering of ZnO. The application of external pressure promotes densification and suppresses grain growth, which is why lower grain growth factors are found at high final sintering density for sintering at 50 MPa in comparison to free sintering. Here, the values for labeled “dry” and “aqueous” represent the results from the present study. Other authors used similar initial nanoparticle size compared to the current study, but always a much higher grain growth factor or a much lower final sintering density was gained, although promising approaches e.g., two- step-sintering [50] or FAST/SPS sintering at high heating rate [51] were used. However, the present study reveals the smallest final grain size for full densification by means of low temperature densification, which is promoted by the presence of bound water (Fig. 2). Interestingly, high heating rates

1223

(>50 K/min) are required to keep water bound to the nanoparticle surface while temperatures are achieved, which are necessary for activation of densification mechanism. Thus, water may participate during sintering due to interface interaction and promote densification, which is discussed in detail in the companion paper [19]. Hence, residual adsorbed water from aqueous powder synthesis or storage at ambient atmosphere is likely to promote densification at lowered sintering temperatures resulting into decreased final grain size. Surprisingly, Gao et al. [51] reported a similar small grain growth factor for SPS sintering of nanocrystalline ZnO at a high heating rate of 600 K/min. However, Gao et al. did not consider the adsorption of water species, which will clearly affect the sintering of nanocrystalline ZnO at those very high heating rates. 3.2. Retardation of nanocrystallinity during sintering Typical attempts in order to further suppress final grain size are (1) doping, (2) application of high pressure and (3) reduction of initial powder particle size. However, these strategies exhibit drawbacks, as (1) final properties are altered by doping or (2) the sample size has to be reduced in order to realize high pressure, limiting the potential applications. Solely, the strategy (3) will not necessarily result into smaller final grain size, as the scale law by Herring [52] announces that the densification rate and the grain growth rate behave indirect proportional to the grain size. This fact is emphasized by a comparative study of Langer et al. [39] who sintered ZnO submicron (160 nm) and nanopowder (30 nm) by FAST/SPS and found a much lower grain growth factor in the case of the submicron powder. Nevertheless, final grain size and density were both found to be similar (Fig. 2). Hence, the reduction of initial particle size is no guarantee for the retention of nanocrystallinity, as grain growth factor depends on the temperature-time history such as isothermal dwell and maximum sintering temperature. Chen and Wang [53] were the first to demonstrate this attempt in the case of free sintering of nanocrystalline yttrium oxide, by the application of two sintering steps with optimized temperatures. Thus, full densification with hardly any grain growth was achieved by this so called “two-step sintering” (TSS) method. On kinetic considerations, Chen and Wang [53] concluded that suppression of grain growth can be achieved by benefiting from the difference between mobility and diffusivity of grain boundary. Nevertheless, the main disadvantage of TSS is the large time consuming due to increased isothermal dwell and the necessity to optimize two temperature steps, otherwise no full densification is achieved or strong coarsening occurs [50]. The reduction of sintering temperature by typically 50 K indeed seems promising with regard to energy saving, but the energy cost would not be amortized due to raised isothermal dwell of 5–20 h. The optimized TSS temperatures are 800 ◦ C and 750 ◦ C with a dwell of 20 h for hot pressing of nanocrystalline ZnO [50]. However, nanocrystallinity could not be preserved since grain growth factor was only reduced from 45 [4] to 24 [50] for single step sintering and TSS, respectively. In comparison to the present study, the sintering temperature was drastically reduced to 400 ◦ C with an isothermal dwell of 10 min, which emphasizes the overall impact of surface bound water on the grain boundary diffusivity. Here, truly energy saving potential is found, whereas concepts of FAST/SPS sintering of multiple parts highlight the possible application for industry [18]. The grain boundary mobility is also reduced due to pinning effect by small intergranular pores [54]. Therefore, high homogeneity and density of the initial green body is needed, as large interagglomerate pores will be substituted by smaller intercrystallite pores [12]. Moreover, it was shown for ZnO that an increasing green density (39–73%TD) enhances the densification rate [41].

1224

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

Fig. 1. Sintering path (grain size as a function of relative density) for nanocrystalline ZnO (NG20).

Fig. 2. Grain growth factor as a function of final density for sintered ZnO. The values for “aqueous” (1.6 wt% H2 O/400 ◦ C) and “dry” (800 ◦ C) sintering condition refer to the results of the present work. The references are from Chu et al. [42], Gao et al. [51], Gupta and Coble [37], Han et al. [74] , Hynes et al. [9], Kim et al. [75], Langer et al. [39], Mazaheri et al. [4], Rahaman et al. [41] and Roy et al. [10].

However, a high green density will not necessarily entail a reduction of coarsening neither for free sintering nor under pressure assisted sintering. This is especially interesting as the production of aqueous slurries is cost and time consuming, as nanoparticles tend to agglomerate requiring their stabilization by dispersants. More interestingly, the comparison of Fig. 2 with the initial green density exhibits that sintering under aqueous condition with a moderate green density of 50%TD provide the best result for retention of nanocrystallinity for ZnO so far, as much higher initial green density of >63%TD resulted into higher grain growth factor [10,37,41] . This example highlights that the reduction of temperature while increasing densifying mechanism is more powerful than grain boundary pinning by pores or pressure. Nevertheless, FAST/SPS is done under pressure. Moreover, the present processing method is easy and quick to implement, whereas the production of stable slurries becomes superfluous.

3.3. Effect of surface bound water on development of morphological anisotropy and abnormal grain growth The evaluation of shape and size distribution of grain microstructure was performed on nearly full densified NG20 (98%TD) by means of SEM micrographs. Fig. 3 emphasizes that the shape and the size of grains strongly depend on presence of bound water during sintering. Under aqueous condition large and elongated grains were found among a large fraction of smaller grains with isometric shape. This morphological anisotropy is less pronounced in measurement direction perpendicular to applied load. In contrast, under dry condition only isometric shaped grains were found independent of measurement direction. Moreover, no preferred orientation for anisotropic grains was found in axial measurement direction independent from temperature, sintering density or the presence of bound water (Fig. 4). These findings were

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

1225

Fig. 3. SEM micrograph of FAST/SPS sintered NG20 (98%TD) under (a), (b) aqueous and (c), (d) dry condition. The specimen were measured in direction (a), (c) parallel and (b), (d) perpendicular to uniaxial load.

Fig. 4. Grain orientation of FAST/SPS sintered NG20 for (a)–(c) aqueous and (d) dry condition. Specimen were sintered to maximum temperature of (a), (b) 400 ◦ C and (c), (d) 800 ◦ C.

also confirmed on transgranular fracture surfaces, where the shape of the grains was revealed and elongated grains were found solely under aqueous condition. Fig. 5 shows the grain size distribution for grains with equivalent spherical radius (3D) of aqueous sintered NG20 at 400 ◦ C with 88%TD and 99%TD. The frequency of grains was normalized by the

fractional area of the grains, whereas a bimodal size distribution is obtained. This finding was already attributed to the occurrence of abnormal grain growth [55]. Here, the size distribution was fit with logarithmic normal function in order to determine the mean grain size for abnormal and normal grain growth. The amount of abnormally grown grains decreases with increasing sintering density,

1226

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

Fig. 5. 3D size distribution of grain radius for aqueous sintered NG20 with (a) 88%TD and (b) 99%TD.

whereas the ratio of the radius between abnormally to normally grown grains (Rna /Rn ) remains constant in the first approximation with Rna /Rn = 2.8 up to full densification. This result indicates that grain growth rate is similar for both species between intermediate and final stage of sintering. However, the surface area decreases with increasing sintering density resulting into a smaller effective interface for interaction of surface bound hydroxides. Hence, the ratio Rna /Rn remains constant. In addition, the occurrence of large elongated grains corresponds to the large fraction of grains. There exist several reasons for abnormal grain growth to occur. Fisher et al. [56] reported about enhancement of abnormal grain growth for niobate under reducing hydrogen atmosphere. Doping may results into abnormal grain growth of ZnO [57], but pure ZnO was used in the present study. Moreover, theoretical calculations indicate that abnormal grain growth is promoted by anisotropy of interface energy [58]. The higher interface energy is known to cause an increase of grain boundary mobility, whereas different growth rates result for different crystallographic directions [59]. ZnO exhibits an anisotropic surface energy, which is much higher for the polar (0 0 0 2) face with 3.0 J/mol compared to the nonpolar (1 0 1¯ 0) and (1 0 2¯ 0) with 1.4 J/mol [25,27]. However, no faceting or abnormal grain growth is observed for NG20 sintered under dry conditions. In contrast, a raised fraction of faceted grain boundaries was recently observed for sintering of nanocrystalline ZnO in presence of water [60], which is discussed in context of occurrence of abnormal grain growth [61,62]. The formation of faceting is probably due to enhancement of surface diffusion by increased water vapor pressure. Thus, faceting would be absent without bound surface water during sintering, which might be a reason, why abnormal grain growth was not jet observed for ZnO. Part I of the present paper emphasized that hydroxide ion diffusion along grain boundaries (from dissociated surface bound water) is identified as the origin for promoted densification of nanocrystalline ZnO. Therefore, preferred segregation of hydroxide ions at the grain boundary has to be considered, which may act as grain boundary complexion. It is noteworthy that complexions at grain boundaries modify the grain boundary energy [30] , whereas different types of complexions may coexist in the same microstructure each with

characteristic grain boundary mobility of different planes. Dillon and Harmer [28,29] investigated doped alumina and introduced a reasonable mechanism for the occurrence of abnormal grain growth due to difference of grain boundary mobility by characteristic complexions a specific planes. A study about grain growth kinetic of Nb-doped alumina clearly shows that normal and abnormal grown grains exhibit complexions mono- and bi-layers of Nb, respectively [31]. However, it is still unclear why some grain boundary planes are preferential occupied by bi-layered complexions, but it is suggested that the grain boundary with the highest energy preferred [63]. Theoretical investigation suggest preferential occupation of the polar (0 0 0 1) of ZnO by hydrogen [64]. In sense of the present study, it seems conclusive that surface bound water is preferential dissociated at the (0 0 0 1) plane, resulting into enrichment of hydroxide complexion. Thus, grain boundary mobility is increased and results into pronounced coarsening for single grains with preferential grain growth in crystallographic direction of caxis during sintering. Dillon and Harmer [28] concluded that the number of abnormal grains increases with grain size, as the relative solute concentration at the grain boundary area is increased. However, hydrogen concentration in ZnO can be completely removed after thermal treatment at 500 ◦ C [65]. Here, the proton-related hydroxide concentration should already decrease during intermediate stage of sintering at lower temperature of 400 ◦ C, as water vapor pressure shows a maximum at 350 ◦ C and steadily decreases at higher temperatures [19]. Thus, the concentration of complexion by hydroxide at grain boundaries should decrease during sintering process. Therefore, the driving force for abnormal grain growth should decrease, which is emphasized by the constant ratio Rna /Rn between intermediate and final stage of sintering (Fig. 5). In this context, the desorption of water and hydrogen rises with increasing temperature, whereas the difference between abnormal and normal grain growth should be negligible and no morphological anisotropy is observed for maximum sintering temperature up to 800 ◦ C (Fig. 3(c) and (d)). This finding coincides with the absence of hydrogen after sintering at 800 ◦ C [19]. In addition, the grain boundary character was analyzed by means of high-resolution transmission electron microscopy

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

1227

Fig. 6. Transmission electron microscopy of sintered ZnO under (a), (b) aqueous and (c), (d) dry sintering conditions with overview (a), (c) and (b), (d) high resolution images. Faceting is visible under aqueous sintering condition as indicated by a red line (b). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(HRTEM) on dense sintered specimen in order to analyze the grain boundary character. The fraction of rough grain boundaries is found to be smaller under aqueous than under dry sintering condition (Fig. 6). Moreover, the TEM study confirms the presence of elongated shaped grains under aqueous sintering condition. These elongated grains might be related to abnormal grain growth. Faceted grain boundaries were found under aqueous condition during HRTEM investigation (Fig. 6(b)), but not under dry condition (Fig. 6 (d)). This grain boundary structure could provide an indication that explains the abnormal grain growth due to sintering of ZnO in presence of surface bound water and/or hydroxide ions. Jung et al. [66] investigated the sintering of BaTiO3 under various oxygen partial pressures and found a decreasing fraction of abnormal grown grains with decreasing oxygen partial pressure, which is accompanied by a decreasing fraction of faceted grain boundaries. This behavior is attributed to a critical driving force for appreciable migration of the boundary. A hill-and-valley shaped grain boundary is only observed during the stage between abnormal and normal grain growth (with an intermediate oxygen partial pressure), whereas faceting and rough grain boundaries were attributed to strong abnormal grain growth and normal grain growth, respectively. However, the lateral scale of grain boundaries with faceting or hill-and-valley shape show one order of magnitude smaller size for the present study in comparison to the findings of Jung et al. [66]. Thus, a minority of grain boundaries is believed to exhibit faceting causing abnormal grain growth in combination with preferred coarsening along c-axis crystallographic direction due to a critical driving force for appreciable migration of the grain boundary. Sintering in absence of surface bound water will result into rough grain boundaries with normal grain growth exhibiting solely isometric morphology. Interestingly, a non-stoichiometric (higher) ratio of Zn:O is observed for sintering in presence of bound water, whereas the ratio of Zn:O is approximately stoichiometric under dry sintering condition [19]. Surprisingly, a stoichiometric ratio of Ti:O at the grain boundary was attributed to normal

grain growth, whereas the non-stoichiometric ratio (higher Ti:O ratio) was observed for abnormal grain growth in BaTiO3 [66,67]. Thus, the increased ratio of Zn:O might result from hydroxide complexions at the grain boundaries which could form under aqueous sintering condition. Fig. 7(a) shows the frequency of anisotropic shaped grains and pores as a function of orientation angle for aqueous sintered NG20 at 400 ◦ C. Here, anisotropic grains are preferential found perpendicular to uniaxial applied load. Moreover, the residual porosity enabled the evaluation of pore size and orientation. Fig. 7(b) shows pore orientation for aqueous sintered NG20 at 400 ◦ C. Here, anisotropic pores are slightly preferably oriented in radial direction, although the significance is marginal. Nevertheless, the orientation of anisotropic pores and grains coincides, as they are both preferential oriented perpendicular to uniaxial applied load during FAST/SPS sintering. However, the analysis by SEM micrographs does not indicate any preferential direction of these anisotropic grains. Hence, EBSD analysis was performed to reveal the crystalline orientation of the anisometric grown grains. The EBSD mapping as shown in Fig. 8 clearly exhibits the coincidence between the growth direction of the elongated crystallites and their crystallographic c-axis. Quasi isometric shaped grains exist among these large orientated grains, which confirm the observation by SEM from Fig. 3. However, this finding is unusual, as typically no elongated grains are formed during sintering from isotropic ZnO powder particles. 3.4. Development of crystalline texture Fig. 9 shows the development of crystallite size with (0 0 0 2) and (1 0 1¯ 0) Bragg reflection. Both Bragg reflections reveal the same crystallite size at a given sintering density for dry processed NG20. These values are in good agreement with the grain size determination by means of line intercept method from TEM and SEM micrographs. In contrast, the (0 0 0 2) crystallite size increases stronger than the one for (100) Bragg reflection if water is present,

1228

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

Fig. 7. Microstructural analysis of aqueous sintered NG20 (98%TD) concerning (a) grain and (b) pore orientation. The SEM micrographs were evaluated by ImageJ perpendicular to uniaxial load.

Fig. 8. EBSD analysis of aqueous FAST/SPS sintered NG20 (98%TD) at 400 ◦ C.

Table 1 Development of the degree of preferred orientation during FAST/SPS sintering of NG20. Sintered density [%TD] 98 96 83

Axial measurement (01–10) Dry

Aqueous

Radial measurement (0 0 0 2) Dry

Aqueous

32.1 30.3 1.1

28.6 26.8 1.8

15.0 12.8 5.4

14.2 8.7 5.4

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

1229

Fig. 9. Crystallite sizes from FAST/SPS sintered NG20 determined by Scherrer analysis.

Fig. 10. XRD patterns for FAST/SPS sintered NG20 (98%TD) under (a) aqueous and (b) dry condition. The arrows indicate the shift of intensity maxima during the change from axial to radial measurement direction. The points refer to the theoretical maxima of XRD intensity.

beginning with a sintering density of >73%TD. Interestingly, the (0 0 0 2) Bragg reflection corresponds to the crystallographic direction along the c-axis of ZnO, whereas the (1 0 1¯ 0) Bragg reflection refers to a direction perpendicular to the c-axis. The values for dry conditions are exactly between the values from aqueous sintered NG20. These results support the findings from Fig. 5 and further indicate continuous development of anisotropic grain growth in

direction of the crystallographic c-axis. Interestingly, grains grow in preferred crystallographic direction referred to uniaxial applied pressure. Fig. 10 shows the detailed XRD patterns of fully densified NG20 with (1 0 1¯ 0), (1 0 2 0) and (0 0 0 2) Bragg reflections for measurement directions axial and radial to applied load. The theoretical maxima of intensity for unaligned crystallites are represented by points. The arrows indicate the shift of intensity maximum after

1230

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

Fig. 11. Rietveld refinement of sintered NG20 (a) without and (b) with simulation of preferred orientation. The difference between measured and simulated data is given below the plot.

Fig. 12. Pole figure of sintered NG20 (98%TD) for (a) 400 ◦ C, aqueous condition and (b) 800 ◦ C, dry condition. Measurement was performed in direction parallel to uniaxial load. The color scale is in multiples of uniform (pole) density (m.u.d.).

changing from axial to radial measurement direction. These results indicate the formation of a crystallographic texture, due to the large deviation of measured to theoretical intensity maxima. Here, the degree of preferred crystalline orientation was evaluated by Rietveld refinement from XRD measurement by means of the March–Dollase approach. Fig. 11 shows the calculated and measured XRD spectra with and without simulated texture. A good agreement was only found under consideration of preferred orientation. The degree of preferred orientation was determined as a function of sintering density (by means of interrupted sintering) and is listed in Fig. 11. The degree of preferred orientation for (1 0 1¯ 0) Bragg reflection was measured in axial direction and is higher than the one for (0 0 0 2) Bragg reflection, which was

measured in radial direction. In general, the degree of preferred orientation increases strongly with sintering density indicating that crystalline texture is formed due to sintering process. These results represent a mean value for preferred orientation on the macroscopic scale, whereas EBSD measurement allows local texture analysis. Fig. 12 shows the pole figures of (a) aqueous and (b) dry sintered NG20. The (0 0 0 2) Bragg reflection corresponds with the crystallographic c-axis of ZnO, which is underrepresented in axial direction for aqueous and dry condition. Thus, there is preferred growth of grains occurs with c-axis perpendicular to uniaxial applied load during FAST/SPS sintering. Interestingly, morphological texture corresponds to crystallographic texture under aqueous condition, whereas under dry condition grains are shaped isomet-

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

ric although crystalline texture is present as well. The (1 0 1¯ 0) and (1 0 2¯ 0) Bragg reflections have their faces perpendicular to crystallographic c-axis. Grains with these orientations are preferential oriented parallel to applied pressure, which is coherent to hitherto results. The degree of preferred orientation develops during sintering as shown in Table 1. This crystalline texture is similar under dry and aqueous conditions, although sintering temperature is much higher under dry condition. However, the occurrence of morphological and crystalline texture does not necessarily need to coincide. Here, a pronounced crystallographic texture develops independent from presence of bound water or sintering temperature, whereas morphological texture is only observed under aqueous condition at 400 ◦ C maximum sintering temperature. Díaz-Chao et al. [23] created a crystalline texture during SPS sintering of needle-like shaped ZnO powder particles, whereas the final microstructure consists of isometric grains. Díaz-Chao et al. [23] concluded that the crystallographic preferred orientation is a direct consequence of the needle-like morphology, which reoriented under the uniaxial pressure. Similar investigations about the sintering of needle-like ZnO particles also resulted into crystalline texture [20,43,57,68]. These particles may rotate under mechanical pressure into a direction perpendicular to it and finally align, which is an important mechanism for this formation of texture [69]. On the other hand, the particles may already be aligned during the casting process of the green body prior to sintering. This procedure is often referred to template growth [43] and may be realized via slip casting [68] or tape casting of platelet [22] and rod-like [20,43] crystallite. However, in the present work only isometric particles are initially present, which represents no good explanation for the occurrence of crystalline texture. Interestingly, sintering under aqueous conditions and electrical insulation generates the same grain size at 400 ◦ C, as without electrical insulation (Fig. 1). This gives an indication that the electrical current has a negligible effect on sintering ZnO. Under the present experimental conditions the field strength in FAST/SPS across the ZnO body is determined to be 1.2–2.6 V/mm, but might be even lower due to the voltage loss through the electrodes, punches, etc. Moreover, the electric field during FAST/SPS sintering shows no effect on densification behavior of nanocrystalline ZnO, which is emphasized in part I of this paper [19]. These findings are in good agreement with the report by Schmerbauch et al. [70], who investigated the free sintering of nanocrystalline ZnO under external electrical field with and without current through the ZnO specimen. The electrical field suppressed coarsening only for a high electrical field of >4 V/mm is much higher than for FAST/SPS sintering. In addition, these results are consistent to comparative sintering studies between FAST/SPS and hot pressing, where no significant impact was found neither on densification nor on grain growth behavior for insulating [71], semiconducting [39] or ion conducting materials [72]. Shinoda et al. [73] showed that crystalline texture occurs during liquid phase sintering of SiN3 only if specimen is not electrically insulated. Here, precipitation of nuclei with preferred orientation occurs due to presence of electric field. However, for pure ZnO solely solid state sintering is expected. Moreover, electrical insulation of the NG20 body gave no impact on microstructural development. The sintered body was only insulated against the punches, whereas electric current might affect the local microstructure at the edge of the body which is in contact with the conducting die (nevertheless separated by a graphite foil). Hence, higher density and grain size is expected, if a liquid aqueous film would significantly increase the resistance of the ZnO and cause intrinsic Joule heating. However, this is not the case as no gradient in microstructure was observed neither under aqueous nor under dry sintering condition in the present study. In sum, the effect of electric field during FAST/SPS sintering may be regarded as negligible in case of nanocrystalline ZnO.

1231

4. Conclusions Sintering of ZnO at 400 ◦ C in presence of surface bound water resulted into fully densified specimens with a mean grain size of 220 nm. The present study reveals the smallest final grain size for full densification by means of low temperature densification. The reduction of grain size is attributed to the decrease of sintering temperature due to hydroxide-ion-enhanced densification. The presence of surface bound water and the resulting hydroxide ion diffusion not only affects densification but also microstructural development. A pronounced morphological anisotropy occurs only in presence of bound water for sintering at 400 ◦ C. The elongated grains exhibit a preferred growth direction with its polar c-axis perpendicular to uniaxial pressure. In general, the formation of hydroxide layered complexion at the grain boundary is suggested as the origin for the morphological anisotropy for sintering in presence of surface bound water. In contrast, a crystalline texture develops independently from the presence of bound water with the (0 0 0 2) plane orientated preferential perpendicular to uniaxial applied load. Furthermore, the electrical current shows no effect on microstructural development. Moreover, the effect of electric field is considered to be negligible for FAST/SPS sintering of nanocrystalline ZnO. Acknowledgments This work was partially supported by the ‘Deutsche Forschungsgemeinschaft’ (Emmy Noether Program GU993-1). The authors thank Yoo Jung Sohn for Rietveld refinement, Clemens Reuther for assistance at EBSD measurement and Markus Rettenmayr for general support and the access to the SEM. We gratefully acknowledge the partial financial support of Collaborative Research Project of Materials and Structures Laboratory, Tokyo Institute of Technology, the Deutsche Forschungsgemeinschaft (DFG) with grant reference INST 275/241-1 FUGG, and the Thüringer Ministerium für Bildung, Wissenschaft und Kultur (TMBWK) with grant reference 62-4264 925/1/10/1/01. References [1] K. Lu, Sintering of nanoceramics, Int. Mater. Rev. 53 (1) (2008) 21. [2] J. Binner, B. Vaidhyanathan, Processing of bulk nanostructured ceramics, J. Eur. Ceram. Soc. 28 (7) (2008) 1329. [3] M.J. Mayo, Processing of nanocrystalline ceramics from ultrafine particles, Int. Mater. Rev. 41 (3) (1996) 85. [4] M. Mazaheri, S.A. Hassanzadeh- Tabrizi, S.K. Sadrnezhaad, Hot pressing of nanocrystalline zinc oxide compacts: densification and grain growth during sintering, Ceram. Int. 35 (3) (2009) 991. [5] U. Anselmi- Tamburini, J.E. Garay, Z.A. Munir, Fast low-temperature consolidation of bulk nanometric ceramic materials, Scr. Mater. 54 (5) (2006) 823. [6] Z. Guo- Dong, J. Kuntz, W. Julin, J. Garay, A.K. Mukherjee, A novel processing route to develop a dense nanocrystalline alumina matrix (<100 nm) nanocomposite material, J. Am. Ceram. Soc. 86 (1) (2003) 200. [7] R. Chaim, Superfast densification of nanocrystalline oxide powders by spark plasma sintering, J. Mater. Sci. 41 (23) (2006) 7862. [8] S. Schwarz, O. Guillon, Two step sintering of cubic yttria stabilized zirconia using Field Assisted Sintering Technique/Spark Plasma Sintering, J. Eur. Ceram. Soc. 33 (4) (2013) 637. [9] A.P. Hynes, R.H. Doremus, R.W. Siegel, Sintering and characterization of nanophase zinc oxide, J. Am. Ceram. Soc. 85 (8) (2002) 1979. [10] T.K. Roy, D. Bhowmick, D. Sanyal, A. Chakrabarti, Sintering studies of nano-crystalline zinc oxide, Ceram. Int. 34 (1) (2008) 81. [11] J.R. Groza, Nanosintering, Nanostruct. Mater. 12 (1999) 987. [12] S.M. Sweeney, M.J. Mayo, Relation of pore size to green density: the Kozeny equation, J. Am. Ceram. Soc. 82 (7) (1999) 1931. [13] A.S.A. Chinelatto, R. Tomasi, Influence of processing atmosphere on the microstructural evolution of submicron alumina powder during sintering, Ceram. Int. 35 (7) (2009) 2915. [14] P.L. Chen, I.W. Chen, Sintering of fine oxide powders 2. Sintering mechanisms, J. Am. Ceram. Soc. 80 (3) (1997) 637. [15] R. Chaim, A. Levin, A. Shlayer, A. Estournes, Sintering and densification of nanocrystalline ceramic oxide powders: a review, Adv. Appl. Ceram. 107 (3) (2008) 159.

1232

B. Dargatz et al. / Journal of the European Ceramic Society 36 (2016) 1221–1232

[16] H. Noei, C. Woll, M. Mahler, Y.M. Wang, Activation of carbon dioxide on ZnO nanoparticles studied by vibrational spectroscopy, J. Phys. Chem. C 115 (4) (2011) 908. [17] D.V. Quach, A.R. Bonifacio, R.H.R. Castro, Water adsorption and interface energetics of zinc aluminate spinel nanoparticles: insights on humidity effects on nanopowder processing and catalysis, J. Mater. Res. 28 (15) (2013) 2004. [18] O. Guillon, J. Gonzalez-Julian, B. Dargatz, T. Kessel, G. Schierning, J. Räthel, M. Herrmann, Field- Assisted Sintering Technology/Spark Plasma Sintering: mechanisms, materials, and technology developments, Adv. Eng. Mater. 16 (7) (2014) 830. [19] B. Dargatz, J. Gonzalez-Julian, M. Bram, P. Jakes, L. Schade, R. Röder, C. Ronning, O. Guillon, FAST/SPS sintering of nanocrystalline zinc oxide—Part I: enhanced densification and formation of hydrogen-related defects in presence of adsorbed water, J. Eur. Ceram. Soc. 36 (2015) 1207. [20] J.L. Ning, D.M. Jiang, K.H. Kim, K.B. Shim, Influence of texture on electrical properties of ZnO ceramics prepared by extrusion and spark plasma sintering, Ceram. Int. 33 (1) (2007) 107. [21] A.B. Haugen, M.I. Morozov, M. Johnsson, T. Grande, M.-A. Einarsrud, Effect of crystallographic orientation in textured Ba0.92Ca0.08TiO3 piezoelectric ceramics, J. Appl. Phys. 116 (13) (2014) 134102. [22] J.L. Jones, B.J. Iverson, K.J. Bowman, Texture and anisotropy of polycrystalline piezoelectrics, J. Am. Ceram. Soc. 90 (8) (2007) 2297. [23] P. Díaz- Chao, F. Giovannelli, O. Lebedev, D. Chateigner, L. Lutterotti, F. Delorme, E. Guilmeau, Textured Al-doped ZnO ceramics with isotropic grains, J. Eur. Ceram. Soc. 34 (16) (2014) 4247. [24] D. Pravarthana, D. Chateigner, L. Lutterotti, M. Lacotte, S. Marinel, P.A. Dubos, I. Hervas, E. Hug, P.A. Salvador, W. Prellier, Growth and texture of spark plasma sintered Al2O3 ceramics: a combined analysis of X-rays and electron back scatter diffraction, J. Appl. Phys. 113 (15) (2013). [25] N.L. Marana, V.M. Longo, E. Longo, J.B.L. Martins, J.R. Sambrano, Electronic and structural properties of the (10-10) and (11-20) ZnO surfaces, J. Phys. Chem. A 112 (38) (2008) 8958. [26] R.H.R. Castro, R.B. Torres, G.J. Pereira, D. Gouvea, Interface energy measurement of MgO and ZnO: understanding the thermodynamic stability of nanoparticles, Chem. Mater. 22 (8) (2010) 2502. [27] J.I. Sohn, W.-K. Hong, S. Lee, S. Lee, J. Ku, Y.J. Park, J. Hong, S. Hwang, K.H. Park, J.H. Warner, S. Cha, J.M. Kim, Surface energy-mediated construction of anisotropic semiconductor wires with selective crystallographic polarity, Sci. Rep. 4 (2014). [28] S.J. Dillon, M.P. Harmer, Relating grain-boundary complexion to grain-boundary kinetics I: calcia-doped alumina, J. Am. Ceram. Soc. 91 (7) (2008) 2304. [29] S.J. Dillon, M.P. Harmer, Relating grain boundary complexion to grain boundary kinetics II: silica-doped alumina, J. Am. Ceram. Soc. 91 (7) (2008) 2314. [30] S.J. Dillon, M. Tang, W.C. Carter, M.P. Harmer, Complexion: a new concept for kinetic engineering in materials science, Acta Mater. 55 (18) (2007) 6208. [31] S.J. Dillon, M.P. Harmer, G.S. Rohrer, The relative energies of normally and abnormally growing grain boundaries in alumina displaying different complexions, J. Am. Ceram. Soc. 93 (6) (2010) 1796. [32] G.S. Rohrer, M. Affatigato, M. Backhaus, R.K. Bordia, H.M. Chan, S. Curtarolo, A. Demkov, J.N. Eckstein, K.T. Faber, J.E. Garay, Y. Gogotsi, L.P. Huang, L.E. Jones, S.V. Kalinin, R.J. Lad, C.G. Levi, J. Levy, J.P. Maria, L. Mattos, A. Navrotsky, N. Orlovskaya, C. Pantano, J.F. Stebbins, T.S. Sudarshan, T. Tani, K.S. Weil, Challenges in ceramic science: a report from the workshop on emerging research areas in ceramic science, J. Am. Ceram. Soc. 95 (12) (2012) 3699. [33] L.E. Greene, B.D. Yuhas, M. Law, D. Zitoun, P. Yang, Solution-grown zinc oxide nanowires, Inorg. Chem. 45 (19) (2006) 7535. [34] S. Baruah, J. Dutta, pH-Dependent growth of zinc oxide nanorods, J. Cryst. Growth 311 (8) (2009) 2549. [35] S. Baruah, J. Dutta, Hydrothermal growth of ZnO nanostructures, Sci. Technol. Adv. Mater. 10 (1) (2009) 013001. [36] T.A. Vijayan, R. Chandramohan, S. Valanarasu, J. Thirumalai, S. Venkateswaran, T. Mahalingam, S.R. Srikumar, Optimization of growth conditions of ZnO nano thin films by chemical double dip technique, Sci. Technol. Adv. Mater. 9 (3) (2008) 035007. [37] T.K. Gupta, R.L. Coble, Sintering of ZnO: I, densification and grain growth, J. Am. Ceram. Soc. 51 (9) (1968) 521. [38] T.K. Gupta, R.L. Coble, Sintering of ZnO: II, density decrease and pore growth during the final stage of the process, J. Am. Ceram. Soc. 51 (9) (1968) 525. [39] J. Langer, M.J. Hoffmann, O. Guillon, Electric Field-Assisted Sintering and hot pressing of semiconductive zinc oxide: a comparative study, J. Am. Ceram. Soc. 94 (8) (2011) 2344. [40] T.K. Roy, A. Ghosh, Sintering and grain growth kinetics in undoped and silica doped zinc oxide ceramics, Ceram. Int. 40 (Part B (7)) (2014) 10755. [41] M.N. Rahaman, L.C. De Jonghe, M.-Y. Chu, Effect of green density on densification and creep during sintering, J. Am. Ceram. Soc. 74 (3) (1991) 514. [42] M.Y. Chu, M.N. Rahaman, L.C. Dejonghe, R.J. Brook, Effect of heating rate on sintering and coarsensing, J. Am. Ceram. Soc. 74 (6) (1991) 1217. [43] E. Suvaci, I˙ .Ö. Özer, Processing of textured zinc oxide varistors via templated grain growth, J. Eur. Ceram. Soc. 25 (9) (2005) 1663. [44] B. Jungsik, Y. Heesun, H.H. Paul, Enhanced and stable green emission of ZnO nanoparticles by surface segregation of Mg, Nanotechnology 17 (4) (2006) 973.

[45] M.I. Mendelson, Average grain size in polycrystalline ceramics, J. Am. Ceram. Soc. 52 (8) (1969) 443. [46] K.A. Zak, A.W.H. Majid, M.E. Abrishami, R. Yousefi, X-ray analysis of ZnO nanoparticles by Williamson–Hall and size-strain plot methods, Solid State Sci. 13 (1) (2011) 251. [47] L.B. McCusker, R.B. Von Dreele, D.E. Cox, D. Louer, P. Scardi, Rietveld refinement guidelines, J. Appl. Crystallogr. 32 (1) (1999) 36. [48] E. Zolotoyabko, Determination of the degree of preferred orientation within the March–Dollase approach, J. Appl. Crystallogr. 42 (3) (2009) 513. [49] K.G. Ewsuk, D.T. Ellerby, C.B. DiAntonio, Analysis of nanocrystalline and microcrystalline ZnO sintering using master sintering curves, J. Am. Ceram. Soc. 89 (6) (2006) 2003. [50] M. Mazaheri, A.M. Zahedi, S.K. Sadrnezhaad, Two-step sintering of nanocrystalline ZnO compacts: effect of temperature on densification and grain growth, J. Am. Ceram. Soc. 91 (1) (2008) 56. [51] L. Gao, Q. Li, W. Luan, H. Kawaoka, T. Sekino, K. Niihara, Preparation and electric properties of dense nanocrystalline zinc oxide ceramics, J. Am. Ceram. Soc. 85 (4) (2002) 1016. [52] C. Herring, Effect of change of scale on sintering phenomena, J. Appl. Phys. 21 (4) (1950) 301. [53] I.W. Chen, X.H. Wang, Sintering dense nanocrystalline ceramics without final-stage grain growth, Nature 404 (6774) (2000) 168. [54] M.J. Mayo, Grain Growth in Polycrystalline Materials II, Pts 1 and 2, Transtec Publications Ltd., Zurich, 1996, pp. 389–397. [55] Y.-W. Kim, M. Mitomo, H. Hirotsuru, Grain growth and fracture toughness of fine-grained silicon carbide ceramics, J. Am. Ceram. Soc. 78 (11) (1995) 3145. [56] J.G. Fisher, S.-Y. Choi, S.-J.L. Kang, Influence of sintering atmosphere on abnormal grain growth behaviour in potassium sodium niobate ceramics sintered at low temperature, J. Korean Ceram. Soc. 48 (6) (2011) 641. [57] I˙ .Ö. Özer, E. Suvaci, S. Bernik, Microstructure–property relationship in textured ZnO-based varistors, Acta Mater. 58 (12) (2010) 4126. [58] A. Kazaryan, Y. Wang, S.A. Dregia, B.R. Patton, Grain growth in anisotropic systems: comparison of effects of energy and mobility, Acta Mater. 50 (10) (2002) 2491. [59] A.D. Rollett, W.W. Mullins, On the growth of abnormal grains, Scr. Mater. 36 (9) (1997) 975. [60] S. Schwarz, A.M. Thron, J. Rufner, K. Benthem, O. Guillon, Low temperature sintering of nanocrystalline zinc oxide: effect of heating rate achieved by Field Assisted Sintering/Spark Plasma Sintering, J. Am. Ceram. Soc. 95 (8) (2012) 2451. [61] S.-Y. Chung, D.Y. Yoon, S.-J.L. Kang, Effects of donor concentration and oxygen partial pressure on interface morphology and grain growth behavior in SrTiO3 , Acta Mater. 50 (13) (2002) 3361. [62] S.J.L. Kang, Sintering: Densification Grain Growth and Microstructure, Elsevier Butterworth-Heinemann, Oxford, 2004. [63] G.S. Rohrer, Grain boundary energy anisotropy: a review, J. Mater. Sci. 46 (18) (2011) 5881. [64] M. Valtiner, M. Todorova, J. Neugebauer, Hydrogen adsorption on polar ZnO(0001)-Zn: extending equilibrium surface phase diagrams to kinetically stabilized structures, Phys. Rev. B 82 (16) (2010). [65] K. Ip, M.E. Overberg, Y.W. Heo, D.P. Norton, S.J. Pearton, C.E. Stutz, B. Luo, F. Ren, D.C. Look, J.M. Zavada, Hydrogen incorporation and diffusivity in plasma-exposed bulk ZnO, Appl. Phys. Lett. 82 (3) (2003) 385. [66] Y. Jung, S. Choi, S. Kang, Effect of oxygen partial pressure on grain boundary structure and grain growth behavior in BaTiO3 , Acta Mater. 54 (10) (2006) 2849. [67] S.-M. An, S.-J.L. Kang, Boundary structural transition and grain growth behavior in BaTiO3 with Nd2 O3 doping and oxygen partial pressure change, Acta Mater. 59 (5) (2011) 1964. [68] K. Keskinbora, T.S. Suzuki, I.O. Ozer, Y. Sakka, E. Suvaci, Hybrid processing and anisotropic sintering shrinkage in textured ZnO ceramics, Sci. Technol. Adv. Mater. 11 (6) (2010) 065006. [69] T. Kimura, T. Yoshimoto, N. Iida, Y. Fujita, T. Yamaguchi, Mechanism of grain orientation during hot-pressing of bismuth titanate, J. Am. Ceram. Soc. 72 (1) (1989) 85. [70] C. Schmerbauch, J. Gonzalez-Julian, R. Röder, C. Ronning, O. Guillon, Flash sintering of nanocrystalline zinc oxide and its influence on microstructure and defect formation, J. Am. Ceram. Soc. 97 (6) (2014) 1728. [71] J. Langer, M.J. Hoffmann, O. Guillon, Direct comparison between hot pressing and electric Field-Assisted Sintering of submicron alumina, Acta Mater. 57 (18) (2009) 5454. [72] J. Langer, M.J. Hoffmann, O. Guillon, Electric Field-Assisted Sintering in comparison with the hot pressing of yttria-stabilized zirconia, J. Am. Ceram. Soc. 94 (1) (2011), 131. [73] Y. Shinoda, Y. Suzuki, K. Yoshida, Electric field effect on texture formation of mullite in spark plasma sintered SiC with Al2 O3 -SiO2 additive, Scr. Mater. 100 (51) (2015). [74] J.P. Han, P.Q. Mantas, A.M.R. Senos, Densification and grain growth of Al-doped ZnO, J. Mater. Res. 16 (2) (2001) 459. [75] J. Kim, T. Kimura, T. Yamaguchi, Effect of bismuth of oxide content on the sintering of zinc oxide, J. Am. Ceram. Soc. 72 (8) (1989) 1541.