2,3
1
solidi
a
www.pss-a.com
applications and materials science
Properties of spark plasma sintered nanostructured Zn1RxSb 1
status
physica
pss
Phys. Status Solidi A, 1–7 (2011) / DOI 10.1002/pssa.201026665
1
1
1
Christina S. Birkel , Tania Claudio , Martin Pantho¨fer , Alexander Birkel , Dominik Koll , Gregor Kieslich , 4 ,2,3 ,1 Ju¨rgen Schmidt , Raphael Hermann** , and Wolfgang Tremel* 1
Institut fu¨r Anorganische Chemie und Analytische Chemie der Johannes Gutenberg-Universita¨t, Duesbergweg 10-14, 55099 Mainz, Germany 2 Institut fu¨r Festko¨rperforschung, Forschungszentrum Ju¨lich, 52425 Ju¨lich, Germany 3 Faculte´ des Sciences, Universite´ de Lie`ge, 4000 Lie`ge, Belgium 4 Fraunhofer Institut fu¨r Fertigungstechnik und Angewandte Materialforschung IFAM, Institutsteil Dresden, 01277 Dresden, Germany Received 4 November 2010, revised 20 February 2011, accepted 28 March 2011 Published online 6 May 2011 Keywords intermetallics, spark plasma sintering, thermoelectrics, zinc antimonide author: e-mail
[email protected], Phone: þ49-6131-392-5135, Fax: þ49-6131-392-5605
[email protected], Phone: þ49-2461-61-4786, Fax: þ49-2461-61-2610
* Corresponding ** e-mail
Engineering materials with specific physical properties has recently focused on the effect of nanoscopic inhomogeneities at the 10 nm scale. Such features are expected to scatter medium and long-wavelength phonons lowering thereby the thermal conductivity of the system without simultaneously decreasing the charge transport (phonon–glass electron–crystal concept). A new Zn1þxSb nanophase obtained by a wet chemical approach was densified by spark plasma sintering (SPS). Investigations on compounds subsumed as ‘‘Zn4Sb3’’ always suffer from its low thermal stability and the contamination of the nanoparticles with solvents and additives used in the synthesis. In order to gain insight into this compound’s electronic properties we investigated a material free from remnants of the synthesis but contaminated with a small amount of well-characterized decomposition product, i.e., ZnSb. To investigate the influence of the sintering process on the densified samples, different SPS conditions were applied. Four different conditions were used with heating rates between 1608
and 230 8C/min, sintering temperatures between 130 and 190 8C and sintering times between 3 and 6 min. Powders from the surface of the pellets were subject to powder X-ray diffraction (XRD) yielding information about the surface composition. Small pieces of the pellets were also characterized using high-energy synchrotron radiation scattering in order to reveal the phase compositions inside the pellets. Small changes in the sintering conditions of the samples were found to have a large influence on the resulting sample compositions. In addition, the phase compositions on the surface differ significantly from the ones inside the pellets which show a much higher grade of decomposition. The density and morphology of the obtained pellets have been investigated by means of laser microscopy and scanning electron microscopy (SEM). The low density and porosity of the different pellets is a result of the graphite pressing tool which has to be used to ensure the temperature control during the SPS process.
ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1 Introduction Most of the research in the field of thermoelectrics focuses on strategies to increase the dimensionless figure of merit, ZT ¼ a2T/rk, where a is the thermopower, r is the electrical resistivity, k is the thermal conductivity, and T is the temperature. The thermal conductivity is given by the sum of the electronic and the lattice thermal conductivity [1]. Whereas the electronic part is directly related to the electrical conductivity through the Wiedemann–Franz equation [2], the lattice thermal conductivity is related to the crystal structure and the phonon properties of the thermoelectric material.
Among thermoelectric materials, ‘‘Zn4Sb3’’ (more precisely Zn3.9xSb3) exhibits an outstanding figure of merit of ZT 1.3 at 670 K, owing to its exceptionally low thermal conductivity [3, 4]. The structure and precise stoichiometry of ‘‘Zn4Sb3’’ have been studied for a century [2, 4–14]. The proposed structures are variants of a Zn6Sb5 basis structure with structural disorder due to additional Zn interstitials and Zn vacancies [9–13] and have been derived from comprehensive X-ray diffraction (XRD) studies. Recent inelastic neutron scattering investigations related the low thermal conductivity to a soft localized vibration of dumbbell Sb2 units [14]. ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
solidi
status
physica
pss
a
2
C. S. Birkel et al.: Properties of spark plasma sintered nanostructured Zn1þxSb
Recently, a wet chemistry technique was devised for synthesis of thermoelectric ‘‘Zn4Sb3’’ nanocrystals as precursor for phase segregation into ZnSb and a new Zn– Sb intermetallic phase, Zn1þxSb, in a peritectoid reaction. One strategy to enhance the figure of merit is to add dimensions to the tunable physical parameters as the material properties can be optimized independently by nanostructuring. Nanostructures provide a large density of interfaces where phonons can be scattered efficiently thereby decreasing the thermal transport within the material. The first demonstration that a low-dimensional material system could enhance thermoelectric performance was for a 2D superlattice consisting of PbTe quantum wells, Pb1xEuxTe barriers [15] and (Bi,Sb)2Te3 superlattices [16, 17]. Epitaxial-type superlattice structures are formed and their sizes can be controlled by phase separation of metastable ternary compounds into their corresponding binaries [18–21]. Thermoelectric application requires dense/solid materials. When not working with dense/solid bulk materials, one needs to consider the compaction process as a crucial step prior to the thermoelectric characterization. Conventional hot pressing is not applicable to densification of nanocomposites because high temperatures and long heating times lead to crystal growth. Therefore, a more reliable, sensitive technique used these days is spark plasma sintering (SPS) which is characterized by the application of a pulsed electric current while pressing. This leads to a very rapid and efficient heating. While heating rates of up to 1000 8C/min can be achieved, this approach allows the nanoparticular powder to be pressed in a very short time thereby lessening particle growth within the sample. Extensive studies of the influence of sample compaction of bulk Zn4Sb3 on their thermoelectric performance and optimization of pressing conditions have been conducted [22, 23]. Those issues are expected to concern the recently reported Zn1þxSb phase as well. Chemical stability of this compound plays a major role, since it has been reported that Zn4Sb3 partially decomposes into orthorhombic ZnSb and Zn at temperatures higher than 523 K [4]. High mechanical strength is an additional requirement for good performance of thermoelectric materials which decreases with increasing porosity [23]. Besides, Pedersen et al. [22] have shown that density has a large influence on the thermoelectric properties and conclude that sample compaction is a key determinant for thermoelectric performance of Zn4Sb3. In this work we discuss the resulting properties of nanoparticular Zn1þxSb samples after densification by SPS. Different sintering conditions were used and the phases, density and morphology after the pressing process were compared. Powders from the surface of the pellets were subjected to powder XRD with Cu Ka radiation whereas diffraction in transmission geometry through small pieces of the pellets was obtained using high-energy synchrotron radiation. Through these studies, we could evaluate differences in composition on the surface and inside the pellets. Laser microscopy and scanning electron microscopy (SEM) ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
images gave insight into the surface composition and the porosity of the structure. The nanostructure could be preserved during the sintering process. 2 Experimental The synthesis and detailed structural characterization of the Zn1þxSb nanophase are described elsewhere [24]. A combined batch of the Zn1þxSb nanopowder was divided into four samples of equal weight of about 500 mg. These samples were pressed into pellets by SPS using different sintering conditions. In previous experiments we found that the Zn1þxSb nanophase decomposes mainly into ZnSb when heated to temperatures >200 8C [24]. Therefore, sintering temperatures lower than 200 8C were applied. Prior to sintering, the Zn1þxSb nanoparticular compound was heated at 100 8C to remove remainders of surface bound solvent. The sintering conditions are summarized in Table 1. Different sintering temperatures between roughly 130 and 200 8C were chosen and different heating rates were applied. The pressing process was finished within a couple of minutes and the sample was cooled down quickly to prevent crystal growth. The temperature versus time plots of the sintering processes of all four pellets are given in Fig. 1. 3 Instrumentation and measurements 3.1 Spark plasma sintering The SPS was performed using a HP D 5 (FCT Systeme GmbH) system. The samples were heated by a pulsed electric current which flows through the punch–die-sample-assembly using a high current and low voltage. In this study, the pressing tools were made of high-performance graphite and the samples were pressed with a pressure of 76 MPa. According to the SPS technique, the powder mixture is heated stepwise from room temperature to desired sintering temperature with heating rates between 150 and 230 K/min in dynamic vacuum. Temperature measurement was done by a thermocouple inside the die, approximately 1 mm away from the sample. After the isothermal dwell time, the electric current was switched off and the pressing tool including the sinterbody cooled down without additional cooling. 3.2 X-ray powder diffraction X-ray powder diffraction data were collected with a Bruker-AXS D8Discover diffractometer in reflection geometry equipped with a HiStar detector using graphite monochromatized Cu Ka radiation. Samples were glued on top of glass and (111) silicon substrates, respectively, using VP/VA copolymer (vinylpyrrolidone/vinylacetate). Table 1 Sintering conditions of the different pellets. pellet no.
heating rate (K/min)
sintering T (8C)
sintering time (min)
A B C D
160 230 230 180
130 120–130 140–150 190–200
5 6 4 3
www.pss-a.com
Original Paper Phys. Status Solidi A (2011)
3
brass stub using adhesive conductive carbon tape and were additionally contacted on the surface with an adhesive copper band. An acceleration voltage of 7.5 kV was used and the working distance was set to roughly 6 mm.
Figure 1 Temperature versus time plots of the sintering processes of pellets A–D.
3.3 Synchrotron X-ray powder diffraction Synchrotron XRD was performed at the XOR-6 sector (MU-CAT) on the ID-D high energy station of the Advanced Photon Source. The photon wavelength was tuned to 88.4 keV and the scattered radiation collected by a General Electric amorphous silicon 2D-detector with 2048 2048 pixels. Angular rocking of the sample by 38 was used in order to reduce texture effects. The diffraction patterns were recorded in transmission geometry.
4 Results and discussion 4.1 Powder XRD after sintering – surface composition Some loose powder was scraped off of the pellet’s surface to determine the surface composition of the pellets. Quantitative phase analysis of the powder XRD data by Rietveld refinements are shown in Fig. 2. The starting material (bottom in Fig. 2) already shows some decomposition, mainly into the ZnSb phase, due to heating at 100 8C. This provides additional insight into the very low phase stability of the Zn1þxSb compound. With increasing sintering temperature the decomposition of the ‘‘Zn4Sb3’’ compound becomes increasingly prominent. The presence of ZnO can be attributed to the oxidation of Zn which is very likely due to handling of the pressed pellets at ambient conditions. In addition to ZnSb, Zn, and ZnO all surface samples exhibit variable amounts of graphite due to the use of graphite foil in the pressing form. Of the four pellets, pellets C and D show the highest grade of decomposition. In pellet D the original phase is almost completely decomposed (less than 10% starting material left). Therefore, only pellets A, B, and C were subject to further investigation. Individual Rietveld plots and detailed information on the Rietveld refinements (residual factors and errors of the refined phase fractions) of both the surface and the bulk XRD data are presented in the Supporting Information at www.pssa.com (Table SI-1 and Fig. SI-1).
3.4 Quantitative phase analysis Quantitative phase analyses were performed using full pattern profile analysis of the corresponding XRD and synchrotrondiffraction data according to the published structural data of ‘‘Zn4Sb3’’ [10], ZnSb [25], Zn [26], ZnO [27], Sb [11], and graphite [28], respectively, using the fundamental parameter approach as implemented in Topas Academic V 4.1 [29]. 3.5 Laser microscopy Laser microscopic pictures were recorded using a Keyence VK-8710 laser microscope equipped with a movable x/y-stage. Pieces of the different pellets were scanned, combining several single image files to the complete image. The area was subsequently determined using the analysis software of the microscope (VKAnalyzer). The thickness of the pellets was determined in the same way after manually rotating the pieces. The total volume of the pellets war then calculated by using the obtained area and thickness values. 3.6 SEM Scanning electron microscopy data were obtained from small pieces of the pellets using a FEI NovaNano FEG-SEM 630. The pieces were attached to a www.pss-a.com
Figure 2 Rietveld refinements of the powder XRD data of the original material (referred to as ‘‘powder’’) and powders scraped off of the surface of pellets A–D. Major compounds are marked in bold letters. ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
solidi
status
physica
pss
a
4
C. S. Birkel et al.: Properties of spark plasma sintered nanostructured Zn1þxSb
4.2 Synchrotron powder XRD after sintering – bulk composition of the pellets Small pieces of the obtained pellets were subject to XRD using high-energy synchrotron radiation in transmission geometry through the whole object. Therefore, these data reveal the composition of the bulk and not only the surface composition. No graphite is observed because the graphite foil is used in the pressing form which is only attached to the surface of the pellets and does not penetrate the workpiece (Fig. 3). According to these data, the content of the original phase also decreases with increasing sintering time (pellet A vs. pellet B) and sintering temperature. Interestingly, when comparing the data of pellets A and B, already minor changes in sintering conditions result in large effects on the sample composition. Figure 4 summarizes the percentages of the present phases in the different pellets on their surfaces (-S) as well as in the bulk (-B). The data clearly reveal major differences in surface and volume composition of the pellets after SPS. The content of Zn4Sb3 (dark blue) is much lower inside all the pellets than on their surface, in particular in the case of pellet C (30% vs. 1%). As a consequence, the content of the decomposition products (Zn, ZnO, and Sb) is higher in the bulk samples than on the surface of the pellets. The bulk composition of pellets A and B shows major differences despite very similar sintering conditions. A faster heating rate and only one additional minute at the sintering temperature result in a higher degree of decomposition, particularly into Zn and ZnO (see A–B and B–B in Fig. 4). Pellet C, which was sintered at the highest temperature, also shows the most prominent decomposition of the original Zn4Sb3 phase (C–S and C–B). In the bulk, almost no Zn4Sb3 can be found (C–B). The higher grade of decomposition of the bulk sample compared to its surface can be explained by higher temperatures inside the pellets than on their surfaces during the sintering process. This kind of temperature distribution is typically for the SPS of materials with better electrical
Figure 3 Rietveld refinements of high energy synchrotron radiation scattering data of pellets A, B, and C. ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 4 Percentages of the different phases found in pellets A, B, and C (-S detected on the surface, -B in the bulk).
conductivity compared with the graphite of the pressing tool [30]. The Joule heating of the samples is more prominent inside the pellets since the heat on the samples’ surfaces is partly transferred to the graphite of the pressing tool. Since decomposition is mainly influenced by the heat which is applied to the sample, this leads to major decomposition effects inside the pellets. Our findings are in accordance with the results of other research groups concerning the thermal stability of Zn4Sb3. It has been noted that it decomposes into ZnSb and Zn even at temperatures above 373 K [11]. Schlecht et al. [31] have shown that Zn4Sb3 nanoparticles which were prepared by a low temperature solid state reaction, decompose below 200 8C in an open system. Since mechanical and thermal stability are of utmost importance for the performance of the thermoelectric material, more detailed studies have been conducted. Stiewe et al. prepared Zn4Sb3 with various grain sizes by zone melting and quenching which were pressed by using SPS at 400 8C for 3 min. The different samples underwent thermal cycling and were characterized in respect to their thermoelectric properties. They found a higher grade of decomposition for samples with smaller grain sizes due to increased molecular diffusion along a high number of grain boundaries [32]. Therefore, our results are not surprising because the particles in this study are even smaller than those mentioned by Stiewe et al. which should make them even less stable towards decomposition at elevated temperatures. Yin et al. investigated the influence of doping and oxygen impurities on the thermal stability of Zn4Sb3. They concluded an enhanced stability for samples which were doped with 1% Cd and samples which were heated in Ar. The major decomposition products of the doped sample were ZnSb, ZnO, and Sb which is in accordance with our findings. When heated in Ar, much less decomposition took place and no oxidized decomposition products could be detected [33]. Since the samples in this study were sintered under a vacuum, we believe the oxygen is attached to the large surface of the nanoparticles. www.pss-a.com
Original Paper Phys. Status Solidi A (2011)
5
Figure 5 Laser microscopy images of pellets A–C.
5 Density and morphology The morphology after the SPS densification/solidification was determined by means of laser microscopy and SEM images. The density of the pellets was roughly estimated by their mass to volume ratios. The volumes were obtained through laser microscopy and the density calculated to be about 3 g/cm3. This value is independent of the sintering conditions; it is about 50% of the theoretical bulk density of Zn4Sb3. The inhomogeneous surface composition of the pellets is suggested already from the laser microscopy images (Fig. 5). Scanning electron microscopy images provide a more detailed view of the morphology of the pellets. Figure 6 shows images of the surfaces of the three investigated pellets at different magnifications of 10,000 (first row), 20,000 (second row), and 50,000 (third row). All pellets show a
similar morphology and do not differ much from each other. The surfaces are rough and many cracks and voids can be found. However, the nanoscale of the samples seems to be conserved during the pressing process (highest magnification) which is promising considering possibly reduced heat transport. Pellet C seems to consist of larger structures than pellet B (see third row) which could be the result of the higher sintering temperature during the pressing process. All images show the overall character of the morphology is rather porous which is responsible for the low density of the pellets (Fig. 6). The nanoparticles are metastable and show very high surface diffusion rates. Therefore, the sintering of the nanoparticles starts at relatively low temperatures, and it is difficult to separate the sintering from the grain growth. Due
Figure 6 SEM images of the pellets A, B, and C. Magnification are 10,000 (first row), 20,000 (second row), and 50,000 (third row).
www.pss-a.com
ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
solidi
status
physica
pss
a
6
C. S. Birkel et al.: Properties of spark plasma sintered nanostructured Zn1þxSb
to the low decomposition temperature, the use of a high pressure during the sintering is helpful. The limit for applying pressure is systematically given be the material of the punches of in the pressing tool. Graphite can be used up to relatively low pressures (76 MPa during the experiments) while permitting an exact temperature control. Steel or tungsten carbide punches allow very high pressures, but the temperature control is much more complicated due to the much larger thermal mass of the pressing tool. Furthermore, an influence of residual organic components, originated from the wet chemisty process, which decreases the sintering activity of the particles cannot be excluded. The low density accompanied by voids and cracks can also be responsible for the oxidation of small Zn particles. ZnO cannot only be found on the pellet’s surface but also inside the pellets because oxygen can penetrate the bulk structure through the void spaces observed in the SEM micrographs. Note that all pellets were brittle and broke when cut before contacting for subsequent thermoelectric measurements. Due to their very low mechanical stability no adequate objects for thermoelectric characterization could be obtained. 6 Conclusions We reported on the densification/ solidification of a new Zn1þxSb nanophase by SPS. To investigate the influence of the sintering process on the densified samples, different SPS conditions were applied. Four different conditions were applied with heating rates between 1608 and 230 8C/min, sintering temperatures of 130, 150, and 190 8C and sintering times between 3 and 6 min. Powders from the surface of the pellets were subject to ‘‘regular’’ powder XRD yielding information about the surface composition. Small pieces of the pellets were also characterized using high-energy synchrotron radiation scattering in order to reveal the phase compositions inside the pellets. We found that only small changes in the sintering conditions of the samples bear a large influence on the resulting sample compositions. In addition, the phase compositions on the surface differ significantly from the ones inside the pellets which show a much higher grade of decomposition. Due to the direct Joule heating of the sample in SPS, the temperature of the sample is higher in the interior volume than at the surface, in particular as the pressing tool acts as a heat sink. Thus, the finding of a higher degree of degradation of Zn1þxSb inside the pellet is in due to the higher temperature inside the pellet. The density and morphology of the pellets were determined by means of laser microscopy and SEM. Estimated densities of approximately 3 g/cm3 resemble only 50% of the theoretical bulk value and were not influenced by the sintering conditions. Due to the thermal decomposition of the Zn1þxSb nanoparticles even at relatively low temperatures, a critical balance between heat and pressure during the SPS process has to be maintained. Therefore, a graphite tool was used and a rather low pressure applied which did not allow us to gain dense pellets. Top view images ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
show the inhomogeneous composition due to the presence of different phases. A more detailed view on the morphology was provided by SEM images which show rather porous structures resulting in the low density of the materials. On the other hand, the nanostructure of the Zn–Sb compound was preserved which is crucial for an expected reduced heat transport. Due to major decomposition, the presence of multiple phases and low densities we refrained from any thermoelectric measurements. The main challenge will be to achieve high density materials while retaining the Zn1þxSb phase. Acknowledgements The Advanced Photon Source APS Argonne/Chicago is acknowledged for synchrotron radiation beam time. The DFG priority program SPP1386 ‘‘Nanostructured Thermoelectrics’’ is acknowledged for the support of this study. C.S.B. is a recipient of a fellowship from MATCOR, the Graduate School of Excellence of the State of Rhineland-Palatinate. R.H. acknowledges support from the Helmholtz-University Young Investigator Group ‘‘Lattices Dynamics in Emerging Functional Materials’’.
References [1] J. M. Ziman, Prinzipien der Festko¨rpertheorie (Verlag Harry Deutsch, Zu¨rich und Frankfurt am Main, 1972). [2] S. Elliott, The Physics and Chemistry of Solids (Wiley and Sons, Chichester, 1998). [3] S. Bhattacharya, R. P. Hermann, V. Keppens, T. M. Tritt, and G. J. Snyder, Phys. Rev. B 74, 134108-1 (2006). [4] T. Caillat, J. P. Fleurial, and A. Borshchevsky, J. Phys. Chem. Solids 58, 1119 (1997). [5] J. Nylen, S. Lidin, M. Andersson, H. Liu, N. Newman, and U. Haussermann, J. Solid State Chem. 180, 2603 (2007). [6] Y. Mozharivskyj, Y. Janssen, J. L. Harringa, A. Kracher, A. O. Tsokol, and G. J. Miller, Chem. Mater. 18, 822 (2006). [7] K. Yamamoto, Japn. Patent 1905. [8] H. W. Mayer, I. Mikhail, and K. Schubert, J. Less. Common. Met. 59, 43 (1978). [9] G. J. Snyder, M. Christensen, E. Nishibori, T. Caillat, and B. B. Iversen, Nature Mater. 3, 458 (2004). [10] F. Cargnoni, E. Nishibori, P. Rabiller, L. Bertini, G. J. Snyder, M. Christensen, C. Gatti, and B. B. Iversen, Chem. Eur. J. 10, 3862 (2004). [11] Y. Mozharivskyj, A. O. Pecharsky, S. Bud’ko, and G. J. Miller, Chem. Mater. 16, 1580 (2004). [12] J. Nylen, M. Andersson, S. Lidin, and U. Haussermann, J. Am. Chem. Soc. 126, 16306 (2004). [13] J. Nylen, S. Lidin, M. Andersson, B. B. Iversen, H. X. Liu, N. Newman, and U. Haussermann, Chem. Mater. 19, 834 (2007). [14] W. Schweika, R. P. Hermann, M. Prager, J. Persson, and V. Keppens, Phys. Rev. Lett. 99, 125501 (2007). [15] L. D. Hicks, T. C. Harman, X. Sun, and M. S. Dresselhaus, Phys. Rev. B 53, R10493 (1996). [16] J. C. Caylor, K. Coonley, J. Stuart, T. Colpitts, and R. Venkatasubramanian, Appl. Phys. Lett. 87, 023105 (2005). [17] M. Martin-Gonzalez, G. J. Snyder, A. L. Prieto, R. Gronsky, T. Sands, and A. M. Stacy, Nano Lett. 3, 973 (2003). [18] T. Ikeda, V. A. Ravi, and G. J. Snyder, J. Mater. Chem. 23, 2538 (2008). www.pss-a.com
Original Paper Phys. Status Solidi A (2011)
[19] T. Ikeda, L. A. Collins, V. A. Ravi, F. S. Gascoin, S. M. Haile, and G. J. Snyder, Chem. Mater. 19, 763 (2007). [20] P. F. R. Poudeu, J. D’Angelo, A. D. Downey, J. L. Short, T. P. Hogan, and M. G. Kanatzidis, Angew. Chem. Int. Edit. 45, 3835 (2006). [21] G. S. Nolas, J. L. Cohn, G. A. Slack, and S. B. Schujman, Appl. Phys. Lett. 73, 178 (1998). [22] B. L. Pedersen, H. Birkedal, B. B. Iversen, M. Nygren, and P. T. Frederiksen, Appl. Phys. Lett. 89, 242108 (2006). [23] K. Ueno, A. Yamamoto, T. Noguchi, T. Inoue, S. Sodeoka, H. Takazawa, C. H. Lee, and H. Obara, J. Alloys Compd. 384, 254 (2004). [24] C. S. Birkel, E. Mugnaioli, T. Gorelik, U. Kolb, M. Pantho¨fer, and W. Tremel, J. Am. Chem. Soc. 132, 9881 (2010). [25] F. L. Carter and R. Mazelsky, J. Phys. Chem. Solids 25, 571 (1964).
www.pss-a.com
7
[26] F. Weitzer, K. Remschnig, J. C. Schuster, and P. Rogl, J. Mater. Chem. 5, 2152 (1990). [27] E. H. Kisi and M. M. Elcombe, Acta Crystallogr. C 45, 1867 (1989). [28] G. E. Bacon, Acta Crystallogr. 3, 137 (1950). [29] A. Coelho, TOPAS Academic V4.1. [30] J. Schmidt, R. Niewa, M. Schmidt, and Y. Grin, J. Am. Ceram. Soc. 88, 1870 (2005). [31] S. Schlecht, C. Erk, and M. Yosef, Inorg. Chem. 45, 1693 (2006). [32] C. Stiewe, T. Dasgupta, L. Boettcher, B. Pedersen, E. Mueller, and B. Iversen, J. Electron. Mater. 39, 1975 (2010). [33] H. Yin, M. Christensen, B. L. Pedersen, E. Nishibori, S. Aoyagi, and B. B. Iversen, J. Electron. Mater. 39, 1957 (2010).
ß 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
