Article pubs.acs.org/JPCC
Annealing-Induced {011}-Specific Cyclic Twins in Tetragonal Zirconia Nanoparticles Wentao Hu, Shaocun Liu, Yang Zhang, Jianyong Xiang, Fusheng Wen, Bo Xu, Julong He, Dongli Yu, Yongjun Tian,* and Zhongyuan Liu* State Key Laboratory of Metastable Materials Science and Technology, Yanshan University, Qinhuangdao 066004, China S Supporting Information *
ABSTRACT: Zirconia (ZrO2) nanocrystals with average size of 4 nm are fabricated by oxidation of the nonstoichiometric ZrC0.6 with ordered carbon vacancies at 450 °C under atmosphere. The nanocrystals are predominantly tetragonal (t) phase and spherical in shape, and their exposed surfaces are constructed by the {011} and {001} facets. After annealing at 700 °C under atmosphere, the coalescence of adjacent t-ZrO2 nanocrystals is observed, and most of the annealed t-ZrO2 nanoparticles are found to exhibit the {011}-specific twins. The dominant cyclic twins as well as a small number of the single and lamellar twins are recognized in the twinned nanoparticles. The cyclic-twinned nanoparticles are identified to have the 5-fold symmetry of either decahedron or icosahedron. In contrast to the single and lamellar twins which are formed via the coalescence of adjacent nanocrystals on the well-developed {011} surfaces, the cyclic-twinned nanoparticles are developed from the coalescence on the disoriented contact surfaces, in which the emission of partial dislocations and induced deformation are recognized to play the key role.
1. INTRODUCTION With the rapid expansion of nanostructured materials, multiply twinned particles (MTPs) have been attracting great attention because of their significance in many research fields such as heterogeneous catalysis,1 optoelectronic properties,2 spontaneous ferromagnetism,3 magnetocrystalline anisotropy,4 and hardness.5 MTPs, especially those with 5-fold symmetry of either decahedron or icosahedron, are usually observed in the fcc or dc nanomaterials, for example, Au,2 FePt,4 BC,5 Cu,6 Pd,7 Ni,8 Si,9 diamond,10 TiN and SiC,11 B6O,12 etc. To date, the formation mechanism of MTPs still remains unclear, although a huge amount of experimental and theoretical studies have been done. It is not possible to affirm a unique formation mechanism due to the diversity of involved material and the difference in the preparation processes. Instead, several formation mechanisms have been developed. For example, the 5-fold twins have been attributed to nucleation-based and growth-mediated mechanisms. The nucleation-based mechanism involves the noncrystallographic packing of atoms, which are transformed into quintuples of twins. In the growth-mediated mechanism, MTPs are formed by cyclic twinning operations owing to either mis-stacking of atoms (growth twinning) or mismatch of lattices (deformation twinning) during growth.13 Recently, Karkin et al. have performed the molecular dynamics (MD) investigations of the influence of disoriented contact surfaces on the formation of MTPs.14 The MD simulation predicted that two nanoparticles agglomerated on the disoriented contact surfaces can evolve into a MTP of 5-fold symmetry at proper temperature. Zirconia (ZrO2) has three polymorphs, i.e., monoclinic (m), tetragonal (t), and cubic (c) phases. For bulk ZrO2, the m © 2012 American Chemical Society
phase is stable at room temperature (RT), while the hightemperature t and c phases can not be quenched to RT despite their more important technological applications. In general, RT stabilization of the t and c phases can be realized mainly through two routes. One is via the dopants of divalent or trivalent cations,15 and the other is via the reduction of size to nanometer scale.16 For undoped zirconia nanoparticles, the tZrO2 is usually considered to be stable below a critical size of about 10 nm at 298 K. Otherwise, the martensitic phase transformation would be triggered. However, via carefully controlling the capping agents and reaction time in a solvothermal routine, m-ZrO2 with the size of 2.8−8 nm and t-ZrO2 with the size of 0.8−3 nm have been selectively synthesized by Xu and Wang. The presence of m-ZrO2 nanoparticles with the size smaller than the critical value of 10 nm is considered to be owing to the high intrinsic compressive strain.17 The martensitic phase transformation is one important characteristic of ZrO2. Its occurrence always leads to the shape change and is usually accompanied by the formation of twins to accommodate the local strain induced by the shape change.18,19 Extensive investigations have been performed on the twin formation in constrained t-ZrO2 particles in a matrix phase. However, similar studies on twins in free isolated t-ZrO2 nanoparticles are rare, and no experimental investigation has been reported on multiple twin formation in isolated spherical t-ZrO2 nanoparticles to the best of our knowledge. Received: June 15, 2012 Revised: August 29, 2012 Published: September 6, 2012 21052
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058
The Journal of Physical Chemistry C
Article
Here we present the experimental study on the annealinginduced twins in isolated spherical t-ZrO2 nanoparticles. The primary monodisperse spherical t-ZrO2 nanocrystals were produced via low-temperature oxidation of the nonstoichiometric ZrC0.6 with ordered carbon vacancies. The as-prepared nanocrystals were subjected to the annealing at high temperature in a homemade tubular furnace. After the annealing, investigations were carried out on the annealing-produced tZrO2 nanoparticles with transmission electron microscopy (TEM).
2. EXPERIMENTAL DETAILS Materials and Methods. The starting nanopowders of ZrCx were synthesized by high-energy ball milling of Zr powders (∼325 mesh, purity 99.5%) in toluene (HPLC grade, 99.7%, Alpha Aesar, USA) in a planetary ball miller (Fritsch P4, Germany). The nonstoichiometric ZrC0.6 with ordered carbon vacancies was prepared by spark plasma sintering of the nanopowders of ZrCx obtained after the 6 h milling. The preparation was performed in a DR. SINTER type SPS 3.20MK-IV (Sumitomo Coal Mining, Japan). The detailed information was introduced in our previous work.20,21 The sample with 10 mm thickness and 20 mm diameter was finally obtained and then polished to remove any surface carbon contamination. The nanocrystals of t-ZrO2 were obtained by oxidation of prepared ZrC0.6 at 450 °C under atmosphere for 1 h. The annealing was performed at 700 °C under atmosphere for 1 h in a homemade tubular furnace. Characterizations. The XRD patterns were obtained by using a diffractometer of D/Max-2500PC with Cu Kα radiation (λ = 1.5406 Å) at 200 mA and 40 kV. The scan speed was set to be 2°/min. The transmission electron microscope (TEM) images and high-resolution transmission electron microscope (HRTEM) images were acquired by using a JEM-2010 transmission electron microscope (JEOL, Japan) at 200 kV. Scanning electron microscopy (SEM) images were obtained by using a scanning electron microscope (S4800, HITACHI, Japan) with accelerating voltage of 20 kV. The details about the superstructure of ordered carbon vacancies in ZrC0.6 were introduced in our previous work.22 Some of the results are shown in Figure S1 (Supporting Information). The lowtemperature oxidation of ordered ZrC0.6 will appear elsewhere.
Figure 1. (a) XRD pattern for the as-prepared ZrO2 nanocrystals. The nanocrystals are demonstrated to be predominantly tetragonal phase. (b) The TEM image of the as-prepared ZrO2 nanocrystals, showing the spherical shape. (c) The Fourier-filtered HRTEM image and the corresponding FFT pattern (inset) for an as-prepared t-ZrO 2 nanocrystal. (d) The XRD pattern after the isothermal annealing at 700 °C in air for 1 h. (e) The TEM image for the annealed nanoparticles. (f) The Fourier-filtered HRTEM image and the corresponding FFT pattern for an annealed t-ZrO2 nanoparticle with the irregular shape.
and theoretical investigations on the surface of t-ZrO2.23,24 In nanometer scale, the flat {011} and curved {001} surfaces are more favorable instead of the {111} surfaces, which are more stable for large t-ZrO2 single crystals.25 The as-prepared t-ZrO2 nanocrystals were annealed at 700 °C under atmosphere for 1 h before fast cooling down to RT. The XRD pattern (Figure 1d) indicates the size growth of the annealed t-ZrO2 nanoparticles since the t-phase peaks become sharper after the annealing. The TEM image of the annealed nanoparticles (Figure 1e) confirms the size growth and reveals more morphology features. Most of the annealed nanoparticles keep the spherical shape with size smaller than 10 nm. Moreover, some nanoparticles display an irregular shape, in which lattice distortion and defects such as dislocations are observed. As revealed clearly in the HRTEM image of an irregular nanoparticle (Figure 1f), the annealing induced the coalescence of adjacent nanocrystals. The joined adjacent nanocrystals would undergo self-recrystallization after the coalescence and evolve to a bigger spherical nanoparticle. The presence of these irregular nanoparticles implies that the self-recrystallization does not have enough time to be completed owing to the fast cooling to RT after the annealing. Some intermediate states of self-recrystallization were frozen to RT, giving rise to the observed irregular nanoparticles. Although the irregular nanoparticles were poorly developed, the t-phase crystallinity can still be revealed in most cases in the
3. RESULTS AND DISCUSSION The XRD pattern shown in Figure 1a confirms that the dominant t-ZrO2 nanocrystals were obtained after the oxidation of sintered ZrC0.6. An average size of 4 nm is estimated using the Scherrer equation. The TEM measurement (Figure 1b) displays the spherical shape of t-ZrO2 nanocrystals, and the average diameter is consistent with the estimated one from the XRD result. As shown in Figure 1c, the high-resolution TEM (HRTEM) image reveals the good crystallinity of t-ZrO2 nanocrystals. In the corresponding fast Fourier transform (FFT) pattern (inset of Figure 1c), the reflections can be indexed as the t-phase {002} and {011} with the corresponding interplanar spacings of 0.257 and 0.298 nm, respectively. As shown in Figure 1c, the exposed surfaces of these as-prepared tZrO2 nanocrystals are predominantly {011} and {001} facets. It is noted that the {011} facets are well developed and flat, while the {001} facets are less developed and curved. These observations are consistent with the previous experimental 21053
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058
The Journal of Physical Chemistry C
Article
Previously, the {111}-specific coalescence twinning via oriented attachment was reported in a small fraction of the isolated t-ZrO2 nanoparticles with the size varying from 10 to 400 nm, and it was pointed out that only relatively large particles with well-developed {111} facets allow the {111}specific coalescence twinning.25 In our case, however, the annealing-produced nanoparticles are smaller than 10 nm in size, and the exposed surfaces are flat {011} and curved {001} facets. The observed {011}-specific twinning can thus be ascribed to the smaller crystal size. Moreover, it was reported that for the isolated HfxZr1−xO2 nanocrystals the annealinginduced martensitic phase transformation is accompanied by the shape change from sphere to rod, and the lamellar-twinned nanorods are observed.26 Among the annealed t-ZrO2 nanoparticles, in addition to some twinned bicrystals with the nonspherical shape as shown in Figure 2e, some nanorods are also observed (Figure S3, Supporting Information). However, the occurrence of martensitic phase transformation is not observed. Most of the annealed nanoparticles are dominantly cyclictwinned with the 5-fold symmetry of either Marks decahedron or icosahedron. Figures 4a and 4d show the HRTEM images for two cyclic-twinned nanoparticles with the 5-fold symmetry of Marks decahedron. These images were taken by focusing the electron beam along the 5-fold axis of the decahedron. The twin boundaries can be easily recognized and labeled out with white arrows in the HRTEM images. In the FFT patterns of the corresponding HRTEM images (Figure 4b and 4e), only the reflections of {002} (marked with white squares) and {011} (marked with white circles) are identified, and the FFT pattern is formed with five overlapping sets of the reflections with a rotation angle of ∼72° relative to each other around the [100] zone axis. Figures 4c and 4f give the schematic diagrams of these decahedral nanoparticles. Figure 4g and 4j show the HRTEM images for two cyclictwinned nanoparticles with the 5-fold symmetry of the icosahedron. The corresponding FFT patterns (Figure 4h and 4k) confirm the characteristics of 5-fold symmetry, in which only reflections of {002} and {011} appear with a slight split. As illustrated in Figure 4i and 4l, the icosahedron consists of 20 tetrahedrons and is bounded by 20 triangular facets, sharing six 5-fold axes and one common point at the center. The HRTEM image of Figure 4g was taken with the electron beam along one of the 5-fold axes of the icosahedron, while the HRTEM image
FFT patterns (for instance, see inset of Figure 1f). By carefully checking the TEM image in Figure 1e, it is noted that most of the spherical nanoparticles are imaged with nonuniform contrast, which can be attributed to the formation of twinning structures. The single, lamellar, and cyclic twins are identified with the help of the HRTEM images. Among the twinned t-ZrO2 nanoparticles, only a small fraction of them exhibit the single twins. Figure 2 displays the
Figure 2. (a), (c), (e) Fourier-filtered HRTEM images for three single-twinned t-ZrO2 nanoparticles, in which the twin boundaries are marked with white arrows. (b), (d), (f) The corresponding FFT patterns of (a), (c), and (e), respectively, in which the twinning relations are labeled out.
HRTEM images and the corresponding FFT patterns for three twinned bicrystals. In the FFT patterns, the reflections can be indexed along the [100] zone axis, and the twinning relations are clearly displayed. The twin plane and the twin angle are determined to be (011̅ ) and 70.7°, respectively. Similarly, the {011}-specific lamellar twins are observed in a few t-ZrO2 nanoparticles. Figure 3 shows the HRTEM images and the corresponding FFT patterns for two lamellar-twinned nanoparticles. The twinning relations are clearly labeled out in the FFT patterns. In the enlarged HRTEM images of Figure 3c and 3f, the twin boundaries are marked by white dashed lines, and the same twin angle of 70.7° is determined. The statistics on the size distribution of single and lamellar twinned nanoparticles indicates that those twinned nanoparticles have an average size of ∼13 nm (Figure S2a, Supporting Information).
Figure 3. (a), (d) Fourier-filtered HRTEM images for two lamellar-twinned t-ZrO2 nanoparticles. (b), (e) The corresponding FFT patterns of (a) and (d), respectively, in which the twinning relations are labeled out. (c), (f) The enlarged HRTEM images of (a) and (d), respectively, in which the twin boundaries and twinning relations are marked with white dashed lines. 21054
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058
The Journal of Physical Chemistry C
Article
Figure 4. (a), (d), (g), (j) Fourier-filtered HRTEM images for four cyclic-twinned t-ZrO2 nanoparticles. (b), (e), (h), (k) The corresponding FFT patterns of the HRTEM images of (a), (d), (g), and (j), respectively. In the FFT patterns, only the diffraction maxima of {002} (marked with white circles) and {011} (marked with white rectangles) appear. It is identified that the cyclic twins in (a) and (d) display the 5-fold symmetry of the Marks decahedron as displayed in the schematics of (c) and (f), respectively, while the cyclic twins in (g) and (j) show the 5-fold symmetry of icosahedron as shown in the schematics of (i) and (l), respectively.
facets, only a small fraction of the twinned nanoparticles show the {011}-specific single and lamellar twins, while most of them display the 5-fold cyclic twins. These observations can be rationalized by considering the crystallographic orientations of the coalesced surfaces of adjacent nanocrystals. The primary tZrO2 nanocrystals were put into a corundum cell and annealed under atmosphere, free of solute and impurities. The random motion of nanocrystals is ignorable without the help of solution. The attachment of adjacent nanocrystals is determined by agglomeration. As a result, the crystallographic orientations of the contact surfaces of adjacent nanocrystals are random. In most cases, the contact surfaces are more or less mismatched. For two nanoparticles attached over the disoriented surfaces, they are expected to evolve into a MTP with the 5-fold symmetry at proper temperature.14 As mentioned above, the fast cooling to RT produced a small amount of the undeveloped nanoparticles with the irregular shape, inside which severe lattice distortion and defects such as dislocations are captured. The observed lattice distortion and defects are related to the annealing-induced coalescence process of adjacent t-ZrO2 nanocrystals. When two or more adjacent nanocrystals are initially coalesced together, the newly produced nanoparticle is not in a stable state and will undergo the adjustment of morphology or self-recrystallization.31 Before self-recrystallization, the chemical energy at the joint is negative, and thermodynamically the atoms elsewhere prefer to move to the joint.32 The observed lattice distortion and defects can thus be attributed to the thermally driven motion of atoms during self-recrystallization of the coalesced t-ZrO2 nanocrystals. Although it is hard to catch the moment when two or more individual particles just coalesce together, detailed information can be inferred about the coalescence-twinning process by carefully analyzing the HRTEM images of these undeveloped nanoparticles. It is recognized that the cyclictwinned nanoparticles are formed in a different way from the single and lamellar-twinned ones.
of Figure 4j was taken with a slight angle between the electron beam direction and the 5-fold axis. Since the tetrahedral units are stacked one above another, the overlying of lattice fringes leads to the complicated contrast patterns in the HRTEM images (Figure 4g and 4j). For both the decahedral and icosahedral nanoparticles with cyclic twins, the exposed surfaces are constructed of the {101} facets, and the twin boundaries and angles are confirmed to be the same as those of the abovementioned single and lamellar twins, i.e., (01̅1) and 70.7°, respectively. The statistics on the size distribution of cyclictwinned nanoparticles indicates that those twinned nanoparticles have an average size of ∼7.6 nm (Figure S2c, Supporting Information). In a multiply twinned nanoparticle with 5-fold symmetry, an angular gap usually remains to achieve a 360° circle.13 The accommodation of the angular gap involves some kind of structural modification or lattice defects, for instance, transformations of the tetrahedral subunit,27 distortions of lattice at twin boundaries,28 stacking faults, and secondary twin boundaries,29 etc. In the {011}-specific 5-fold twinned t-ZrO2 nanoparticles, theoretically, an angular gap of 6.5° should always remain due to the geometrical angle of 70.7° between (011̅) and (011) or (01̅1̅) and (01̅1) planes. By checking the HRTEM images of the decahedral t-ZrO2 nanoparticles with cyclic twins, the lattice distortion and stacking fault are identified at the twin boundaries to accommodate the angular gap (see Figure S4 in Supporting Information). However, the angular gap is not equally assigned among the five twin boundaries, implying that the formation of 5-fold twinned tZrO2 nanoparticles is dominated by the growth-mediated mechanism. Otherwise, the gap would be uniformly accommodated into the five twin boundaries.30 It is noteworthy that among the annealed t-ZrO2 nanoparticles only a few single nanocrystals are observed. Most of them are found to have the twinning structures. Although the primary t-ZrO2 nanocrystals display the well-developed {011} 21055
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058
The Journal of Physical Chemistry C
Article
The primary t-ZrO2 nanocrystals are found to possess the flat {011} and curved {001} surfaces. The formation of {011}specific twinned bicrystals are supported by the presence of well-developed {011} surfaces. Figure 5a shows the HRTEM
central region I (marked with a dashed ellipse), and inside this region obvious lattice mismatch and distortion are present. Although the two adjacent nanocrystals seem to be joined together over the (001) facet in the frozen state, the observed lattice mismatches imply that, in the initial stage of coalescence, the attached surfaces should be highly mismatched. Otherwise, they would develop into a single nanocrystal as shown in Figure S3 (in Supporting Information). The coalescence over the disoriented contact surfaces leads to the observed lattice distortion in the central region. Outside the central region, five regions can be distinguished. On the left and right sides of the central region, there exist two nearly symmetrical regions IIa and IIb, where numerous dislocations and severe lattice distortion are observed. In regions IIIa and IIIb, however, the crystal lattice is less distorted. It is interesting to note that the lattice is severely bent in region IV (marked with green arrows), indicating the coalescence-induced large elastic strain. Recently, 5-fold deformation twins have been observed in nanocrystalline fcc metals and alloys synthesized by severe plastic deformation, and the emission of partial dislocations has been considered to play the key role in their formation.33 Theoretically, the occurrence of large deformation has also been proposed in the evolvement of coalesced individual nanocrystals into a spherical nanoparticle.34 In the present work, severe deformation has been observed in the coalescence of adjacent t-ZrO2 nancrystals. From the observed large number of the dislocations in the regions of IIa and IIb, it can be inferred that the dislocations are emitted on the opposite directions from the coalesced boundary during the coalescence, and most of them pass quickly through the two joined nanocrystals. The motion of the emitted dislocations leads to the severe plastic deformation and the shape change of the coalesced nanocrystals observed in regions IIa and IIb. Owing to the small size, the coalesced nanocrystals would tend to evolve into a spherical nanoparticle for lowering the surface energy. The lattices in other regions would thus be forced to be severely curved inward due to the severe plastic deformation in regions IIa and IIb, generating the elastic strain as observed in region IV. Figure 5c shows the FFT pattern of the corresponding HRTEM image, in which the appearance of four pairs of the {011} reflections suggests the polycrystalline characteristics of the undeveloped nanoparticle. The contributions to the {011} reflections can be recognized as follows: IIIb→A, (IIa, IIb, IIIa)→B, (IIa, IIb)→C, (IIIb, IV)→D. The relative orientations of the different contributing regions can be represented by the relative angles between the four pairs of {011} reflections. As shown in Figure 5c, the relative angles are determined to be close to 35°, i.e., the relative angle between the {011} reflections in the FFT pattern of a well-developed cyclic-twinned nanoparticle. This consistency implies that the undeveloped nanoparticle be very close to a cyclic-twinned structure. In view of energy, the small undeveloped nanoparticle of just several nanometers in size would also prefer to develop into a MTP with the 5-fold symmetry since a single crystal has a higher surface energy than a MTP.35 Figure S5 in the Supporting Information shows another undeveloped t-ZrO2 nanoparticle, which was developed much better than the one shown in Figure 5b. As shown in the HRTEM image, although the coalescence-induced deformation and defects such as dislocations are still observable, this undeveloped nanoparticle exhibits the embryo of a cyclic-twinned particle in the structure. If the annealing was long enough, these two similar
Figure 5. (a) Fourier-filtered HRTEM image and the corresponding FFT patterns (inset) for an undeveloped t-ZrO2 nanoparticle. (b) The Fourier-filtered HRTEM image (inset) and the enlarged one of poorly developed t-ZrO2 nanoparticles. It is demonstrated that the undeveloped nanoparticle originates from the coalescence of two tZrO2 nanocrystals. The coalesced boundary can be identified in the central region I (marked with a dashed ellipse). Severe lattice distortions and many defects such as dislocations are observed in the nearly symmetric regions of IIa and IIb on the left and right sides of region I. In the regions of IIIa and IIIb, the lattice is less distorted. In the region of IV, the lattice is severely bent inward. (c) The corresponding FFT pattern of the HRTEM image in (b), in which four pairs of {011} reflections appear.
image and the corresponding FFT pattern (inset) of an undeveloped t-ZrO2 nanoparticle. Obviously, this undeveloped nanoparticle originates from the coalescence of two t-ZrO2 nanocrystals. A {011}-specific boundary exists between the two coalesced nanocrystals. In this frozen intermediate state of selfrecrystallization after the coalescence, the two coalesced nanocrystals are found to deviate noticeably from the original spherical shape, and strong lattice distortion and defects such as dislocations are simultaneously observed inside. The observed shape change, lattice distortion, and defects can be attributed to the thermally driven motion of atoms to the attached boundary during the coalescence. With the annealing long enough, it can be anticipated that this undeveloped nanoparticle would finally evolve to a {011}-specific twinned bicrystal via self-recrystallization. Considering the small size (less than 10 nm), the spherical shape would be favorable in view of energy. Similarly, the observed {011}-specific lamellar-twinned spherical nanoparticles should result from the coalescence of several adjacent nanocrystals via the {011} facets. For the cyclic-twinned t-ZrO2 nanoparticles, although the twin planes and the twin angle are determined to be the same as those of the single and lamellar twins, their formation originates from the coalescence of adjacent nanocrystals on the disoriented contact surfaces instead of the well-matched ones. Figure 5b shows the HRTEM image (inset) and the enlarged one for an undeveloped nanoparticle frozen in the early stage of self-recrystallization. The undeveloped nanoparticle can be recognized to result from the coalescence of two adjacent nanocrystals. The attachment boundary is identified in the 21056
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058
The Journal of Physical Chemistry C
Article
cyclic twins via self-recrystallization. It should be emphasized that this process is also possible to occur in the coalescence twinning of more than two nanocrystals, but it would be more complicated.
undeveloped nanoparticles would eventually evolve into the cyclic-twinned particles via self-recrystallization. The coalescence-induced deformation indicates that the zirconia nanoparticles with such small size become “softened”. This phenomenon is usually observed in nanostructured materials36,37 with the particle size below a certain value and named as “inverse Hall-Petch” behavior.38 Moreover, the size distributions of those twinned t-ZrO2 nanoparticles indicate that single and lamellar twinned nanoparticles have a bigger average size than cyclic twinned particles (Figure S2, Supporting Information). It has been reported that the size and surface properties usually play important roles in the subsequent assembly process.39 Thus, it can be reasonably speculated that cyclic-twinned nanoparticles can be formed dominantly by the coalescence of the smaller nanoparticles, while single and lamellar twinned nanoparticles result mainly from the coalescence of nanoparticles with the bigger size. On the basis of the discussions above, a possible formation process is proposed for the cyclic-twinned t-ZrO2 nanoparticles, as schematically shown in Figure 6. (I) Before the coalescence,
4. CONCLUSIONS In summary, the nonstoichiometric ZrC0.6 with ordered carbon vacancies was prepared by spark plasma sintering of the mechanochemically synthesized zirconium carbide nanopowders. It is found that low-temperature oxidation of the nonstoichiometric ZrC0.6 produces the predominant spherical t-ZrO2 nanocrystals with the average size of 4 nm. In such small size, the nanocrystals are found to have the exposed surfaces of flat {011} and curved {001} facets. By annealing of the asprepared t-ZrO2 nanocrystals at higher temperature, the coalescence of adjacent nanocrystals is observed to occur. Most of the annealed t-ZrO2 nanoparticles are found to have the {011}-specific twins. Three types of twins are recognized, i.e., single, lamellar, and cyclic twins. The formation of single and lamellar twins is occasional and occurs via the coalescence of adjacent larger nanocrystals on well-developed {011} facets. The formation of dominant cyclic twins originates from the coalescence of adjacent smaller nanocrystals on the mismatched surfaces. The emission of partial dislocations and severe deformation or lattice distortion are identified to play the key role in the formation of cyclic-twinned nanoparticles. Finally, a possible formation process is proposed for the cyclic twins.
■
ASSOCIATED CONTENT
S Supporting Information *
XRD pattern, bright and dark field TEM images, SEM image, and HRTEM image for the sintered ZrC0.6 (Figure S1). The size distributions of single and lamellar twins, twinning width, and cyclic-twinned nanoparticles (Figure S2). The HRTEM images for some t-ZrO2 nanoparticles with the irregular shape induced by annealing (Figure S3). The enlarged HRTEM image for a t-ZrO2 nanoparticle with cyclic twins (Figure S4). The HRTEM image and the corresponding FFT pattern for an undevoloped t-ZrO2 nanoparticle (Figure S5). This material is available free of charge via the Internet at http://pubs.acs.org.
Figure 6. Schematic diagram that shows the formation process of cyclic twins via the coalescence of two t-ZrO2 nanocrystals on the disoriented contact surfaces. (I) Two adjacent t-ZrO2 nanocrystals contact with each other on the mismatched surfaces before annealing. (II) The nanocrystals are coalesced together on the disoriented contact surfaces due to the annealing. Owing to the disorientation of the contact surfaces, severe deformation is induced around the coalesced boundary, from which partial dislocations are emitted on the opposite directions. (III) The emitted dislocations pass quickly through the coalesced nanocrystals, producing the severe deformation and the shape change. After the coalescence, an intermediate nanoparticle with the polycrystalline characteristics is formed. (IV) Via self-recrystallization, the intermediate nanoparticle eventually evolves to a cyclic-twinned nanoparticle with 5-fold symmetry.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected] and
[email protected]. Fax: +86 335 8074545. Notes
The authors declare no competing financial interest.
■
two adjacent t-ZrO2 nanocrystals contact with each other on the mismatched surfaces owing to agglomeration. (II) By the annealing, the nanocrystals start to be coalesced together on the disoriented contact surfaces, producing the severe lattice distortion around the attachment boundary. The partial dislocations are thus induced to be emitted on the opposite directions from the coalesced boundary. (III) The emitted dislocations pass quickly through the coalesced nanocrystals, generating severe deformation and thus the shape change. (IV) The coalesced nanocrystals evolve to an intermediate nanoparticle with the polycrystalline characteristics because of the emission of partial dislocations and the induced deformation. The emission of partial dislocations would play an important role in the formation of twin boundaries. After the coalescence, the intermediate nanoparticle finally develops into the 5-fold
ACKNOWLEDGMENTS We would like to thank the financial support from the National Basic Research Program of China (Grant No. 2010CB731605, No. 511CB808205), the National Science Fund for Distinguished Young Scholars (Grant No. 51025103), and Natural Science Foundations of China (Grant No.51272225, No. 51121061, No. 51102206).
■
REFERENCES
(1) Mohr, C.; Hofmeister, H.; Claus, P. J. Catal. 2003, 213 (1), 86− 94. (2) Sánchez-Iglesias, A.; Pastoriza-Santos, I.; Pérez-Juste, J.; Rodríguez-González, B.; García de Abajo, F. J.; Liz-Marzán, L. M. Adv. Mater. 2006, 18 (19), 2529−2534. 21057
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058
The Journal of Physical Chemistry C
Article
(3) Sampedro, B.; Crespo, P.; Hernando, A.; Litrán, R.; Sánchez López, J. C.; López Cartes, C.; Fernandez, A.; Ramírez, J.; González Calbet, J.; Vallet, M. Phys. Rev. Lett. 2003, 91 (23), 237203. (4) Gruner, M. E.; Rollmann, G.; Entel, P.; Farle, M. Phys. Rev. Lett. 2008, 100 (8), 087203. (5) Wei, B. Q.; Vajtai, R.; Jung, Y. J.; Banhart, F.; Ramanath, G.; Ajayan, P. M. J. Phys. Chem. B 2002, 106 (23), 5807−5809. (6) Silly, F.; Castell, M. R. ACS Nano 2009, 3 (4), 901−906. (7) Huang, X. Q.; Zheng, N. F. J. Am. Chem. Soc. 2009, 131 (13), 4602−4603. (8) Hall, C. R.; Fawzi, S. A. H. Philos. Mag. A 1986, 54 (6), 805−820. (9) Takeguchi, M.; Tanaka, M.; Yasuda, H.; Furuya, K. Surf. Sci. 2001, 493 (1−3), 414−419. (10) Wang, W. N.; Fox, N. A.; Davis, T. J.; Richardson, D.; Lynch, G. M.; Steeds, J. W.; Lee, J. S. Appl. Phys. Lett. 1996, 69 (19), 2825−2827. (11) Cheng, H. E.; Lin, T. T.; Hon, M. H. Scr. Mater. 1996, 35, 113− 116. (12) Hubert, H.; Devouard, B.; Garvie, L. A. J.; O’Keeffe, M.; Buseck, P. R.; Petuskey, W. T.; McMillan, P. F. Nature 1998, 391, 376−378. (13) Hofmeister, H. Fivefold Twinned Nanoparticles. In Encyclopedia of Nanoscience and Nanotechnology; Nalwa, H. S., Ed.; American Scientific: Stevenson Ranch, CA, 2004; Vol. 3, p 431. (14) Karkin, L.; Karkina, L.; Gornostyrev, Y. Mater. Sci. Forum 2008, 584−586, 1033−1038. (15) Abbas, H. A.; Hamad, F. F.; Mohamad, A. K.; Hanafi, Z. M.; Kilo, M. Diffus. Fundamentals 2008, 8, 7.1−7.8. (16) Garvie, R. C. J. Phys. Chem. 1965, 69 (4), 1238−1243. (17) Xu, X. X.; Wang, X. Nano Res. 2009, 2, 891−902. (18) Kelly, P. M.; Francis Rose, L. R. Prog. Mater Sci. 2002, 47, 463− 557. (19) Deville, S.; Guénin, G.; Chevalier, J. Acta Mater. 2004, 52, 5697−5707. (20) Xiang, J. Y.; Hu, W. T.; Liu, S. C.; Chen, C. K.; Zhang, Y.; Wang, P.; Wang, H. T.; Wen, F. S.; Xu, B.; Yu, D. L.; He, J. L.; Tian, Y. J.; Liu, Z. Y. Mater. Chem. Phys. 2011, 130, 352−360. (21) Xiang, J. Y.; Liu, S. C.; Hu, W. T.; Zhang, Y.; Chen, C. K.; Wang, P.; He, J. L.; Yu, D. L.; Xu, B.; Lu, Y. F.; Tian, Y. J.; Liu, Z. Y. J. Eur. Ceram. Soc. 2011, 31, 1491−1496. (22) Hu, W. T.; Xiang, J. Y.; Zhang, Y.; Liu, S. C.; Chen, C. K.; Wang, P.; Wang, H. T.; Wen, F. S.; Xu, B.; He, J. L.; Yu, D. L.; Tian, Y. J.; Liu, Z. Y. J. Mater. Res. 2012, 27, 1230. (23) Eichler, A.; Kresse, G. Phys. Rev. B 2004, 69, 045402. (24) Morterra, C.; Cerrato, G.; Ferroni, L.; Negro, A.; Montanaro, L. Appl. Surf. Sci. 1993, 65−66, 257−259. (25) Shen, P.; Lee, W. H. Nano Lett. 2001, 1 (12), 707−711. (26) Tang, J.; Zhang, F.; Zoogman, P.; Fabbri, J.; Chan, S.; Zhu, Y.; Brus, L. E.; Steigerwald, M. L. Adv. Funct. Mater. 2005, 15, 1595− 1603. (27) Bagley, B. G. Nature 1965, 208, 674−675. (28) Johnson, C. L.; Snoeck, E.; Ezcurdia, M.; Rodríguez-González, B.; Pastoriza-Santos, I.; Liz-Marzán, L. M.; Hÿtch, M. J. Nat. Mater. 2007, 7, 120−124. (29) Hofmeister, H.; Bardamid, A. F.; Junghanns, T.; Nepijko, S. A. Thin Solid Films 1991, 205, 20−24. (30) Gryaznov, V. G.; Heydenreich, J.; Kaprelov, A. M.; Nepijko, S. A.; Romanov, A. A.; Urban, J. Cryst. Res. Technol. 1999, 34, 1091− 1094. (31) Yagi, K.; TaKayanAgI, K.; Kobayashi, K.; Honjo, G. J. Cryst. Growth 1975, 28, 117−119. (32) Zhang, J.; Huang, F.; Lin, Z. Nanoscale 2010, 2, 18−34. (33) Han, W. Z.; Wu, S. D.; Li, S. X.; Zhang, Z. F. Appl. Phys. Lett. 2008, 92, 221909. (34) Hendy, S.; Brown, S. A.; Hyslop, M. Phys. Rev. B 2003, 68, 241403(R). (35) Marks, L. D. Philos. Mag. A. 1984, 49 (1), 81−93. (36) Schiotz, J.; Jacobsen, K. W. Science 2003, 301, 1357. (37) Detor, A. J.; Schuh, C. A. Acta Mater. 2007, 55, 371. (38) Trelewicz, J. R.; Schuh, C. A. Acta Mater. 2007, 55, 5948−5958. (39) Shen, S.; Wang, X. Chem. Commun. 2010, 46, 6891−6899. 21058
dx.doi.org/10.1021/jp305881r | J. Phys. Chem. C 2012, 116, 21052−21058