ARTICLE IN PRESS
Journal of Crystal Growth 310 (2008) 665–670 www.elsevier.com/locate/jcrysgro
Single-crystal growth of NaxCo2O4 via a novel low-temperature flux method Xiaofeng Tanga,, Jian Heb, Kelvin Aaronb, Ed Abbottc, Joseph K. Kolisc, Terry M. Trittb a
Department of Materials Science and Engineering, Clemson University, Clemson, SC, USA b Department of Physics and Astronomy, Clemson University, Clemson, SC, USA c Department of Chemistry, Clemson University, Clemson, SC, USA
Received 2 September 2007; received in revised form 25 October 2007; accepted 2 November 2007 Communicated by M. Schieber Available online 12 November 2007
Abstract NaxCo2O4 single-crystals have been grown using a novel low-temperature flux method. As compared to the formerly reported recipe, the growth temperature has been lowered almost by half to 550 1C via using pure Co as the Co source and the NaOH/NaCl admixture as the flux. The issue of significant loss of Na due to its high vapor pressure at elevated temperatures is much better controlled. Various growth conditions, such as the NaOH to NaCl ratio in the flux, the ratio of Co element to the flux, the firing temperature,the soaking time,the warming/cooling rates and the atmosphere control, have been optimized. The as-grown NaxCo2O4 single-crystals, in form of platelets with growth direction along the ab-plane and with a typical size of few mm on one edge, are found to be Na deficient (x0.3–0.8), which is consistent with the observed electrical resistivity, thermopower and magnetic susceptibility behavior. Despite the single-crystal nature, the directly measured in-plane thermal conductivity k exhibited an amorphous-like temperature dependence; with the room temperature k value is 7 W/m-K. In addition, the thermal stability and the associated surface micro-morphology changes have been studied by means of the differential scanning calorimetry–thermogravimetry analysis and scanning electron microscope. The results showed that the as-grown NaxCo2O4 single-crystals start losing Na above 900 1C. r 2007 Elsevier B.V. All rights reserved. PACS: 81.05.Je; 81.10.Fq; 72.15.Jf; 75.20.Ck; 65.40.b Keywords: A2. Low-temperature flux method; B1. NaxCo2O4 crystals; B1. NaOH and NaCl flux; B2. Electrical transport and magnetic properties; B2. Thermal conductivity
1. Introduction As a member of the alkali ternary oxide group AxMO2 (A ¼ Na, K; M ¼ Cr, Mn, Co, etc.), NaCo2O4 was first reported by Fouassier et al. in 1973 for its interesting transport and structural properties [1]. Since then, a number of papers have been published considering this oxide as a promising cathode material due to its good electronic and electrochemical properties [2–4]. Adopting a space group P6322, NaCo2O4 compound has a layered hexagonal crystal structure, namely the CoO2 layers formed by the 2D edge-sharing MO6 octahedra, and Corresponding author. Tel.: +1 864 650 6178; fax: +1 508 351 7540.
E-mail address:
[email protected] (X. Tang). 0022-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jcrysgro.2007.11.013
intercalating Na layers alternately stacked along the c-axis. The ratio of Na/Co defines four distinct bronzetype phases (a, a0 , b and g) with different oxygen packing orders in hexagonal sheets [1]. Microscopically there are three different oxygen environments between CoO2 sheets for Na+ ions to intercalate, octahedral (O), trigonal prismatic (P) and tetrahedral (T). Note that the Na sites are not fully occupied. For example, the occupancy of Na sites in NaCo2O4 and Na1.5Co2O4 are 50% and 75%, respectively. Stoichiometric NaCoO2 is a semiconductor while non-stoichiometric NaxCo2O4 compounds with x0–1.5 reportedly have a metallic character [3]. In 1997, Terasaki first reported the rather unexpected and quite favorable thermoelectric (TE) properties in NaCo2O4 [5]: low metallic in-plane (ab-plane) resistivity
ARTICLE IN PRESS 666
X. Tang et al. / Journal of Crystal Growth 310 (2008) 665–670
r0.2 mO-cm, large thermopower a100 mV/K, despite a low mobility 13 cm2/V-s, at room temperature. As known, the thermoelectric performance of a material can be evaluated by the dimensionless figure-of-merit ZT ¼ a2 T=rðkL þ ke Þ, where a is the Seebeck coefficient (also named thermopower), r the resistivity, kL and ke lattice and electronic thermal conductivity, respectively. So the resultant power factor (a2T/r) value, 1.5 W/m-K, is even larger than the conventional TE material Bi2Te3, 1.2 W/m-K at 300 K. To the best of our knowledge, NaxCo2O4 is to date the most promising p-type TE oxide material. In 2003, Takada discovered that Na-deficient hydrated compound Na0.35CoO2 1.3H2O, which was synthesized from the parent g-phase Na0.7CoO2, displayed superconducting behavior at a critical temperature (TC) below 5 K [6]. Recently great attention has been given to this low TC superconducting oxide [7–9]. The investigation on the role of intercalating thick H2O layer in the hydrated Na0.35CoO2 1.3H2O may provide valuable information in order to fully understand the transport mechanisms in the high TC superconducting cuprates. Traditionally, NaxCo2O4 single-crystals were synthesized via a high-temperature NaCl flux method using Na2CO3 and Co3O4 as starting materials and firing up to 1050 1C followed by slow cooling at a rate of 0.5–5 1C/h [5,10]. The typical size of these as-grown crystals is 1–2 mm and it is difficult to grow bigger samples. Therefore measurements of the TE properties, especially thermal conductivity, are difficult to perform. Of a greater concern is the loss of Na during the growth due to the high vapor pressure of Na at elevated temperatures. Even in the case that the crucibles are specifically arranged and embedded in the protection powders, a considerable amount of Na is lost with the gas product CO2. It should be noted that an air-tight seal cannot be the option, as extra oxygen is required to oxidize Co from the valence states Co2+,3+ in the starting material Co3O4 to Co3+,4+ in the final product of NaxCo2O4. As part of the effort to grow large single-crystals, an optical floating-zone technique was employed [11–13]. However, the growth is still performed in the high-temperature regime, it is difficult to stabilize the molten zone as considerable amount of Na vaporizes during the synthesis. In this paper, we report the successful synthesis of NaxCo2O4 single-crystals, for the first time, via a lowtemperature flux approach using pure Co as the Co source and the NaOH/NaCl admixture as the flux. The crystal growth conditions, chemical compositions, surface micromorphology, TE and magnetic properties, and thermal stability of as-grown NaxCo2O4 single-crystals are discussed.
was welded utilizing an electric Arc. The Ag tubes were then placed in a larger size Ag tube or quartz tube with the top end open to air. The whole assembly was positioned in a programmable vertical tube furnace. Depending on the NaOH to NaCl ratio, the firing temperature was set up to 550, 700 and 780 1C, respectively. Slow cooling rates at 0.3–5.0 1C/h were used. The phase and crystallinity of the obtained crystals and powders have been checked by X-ray powder diffraction (XRD). The electrical and thermal transport properties were measured on the custom-designed systems in the temperature range of 10–310 K [14,15]. The in-plane thermal conductivity has been determined by a direct measurement method, for the first time, via a novel technique developed in our labs called the ‘‘parallel thermal conductance’’ technique, specifically developed for measuring the thermal conductivity of small single-crystals [16,17]. Magnetic properties as a function of temperature were measured on a SQUID (Quantum Designs) with a magnetic field of 0.01–0.5 T. The surface micro-morphology and chemical composition were analyzed with Olympus BX60M optical microscopy and scanning electron microscope Hitachi-3500N equipped with energy-disperse X-ray spectroscopy (EDX). The high-temperature thermal stability of as-grown single-crystals was studied by a TAs Instrument SDT-2960, simultaneous with the differential scanning calorimetry–thermogravimetry analysis (DSC–TGA) system. 3. Results and discussion In this study, the mixture of NaOH/NaCl powders was employed as the flux in order to lower the growth temperature and thus prevent the Na loss. After firing the mixture of Co and flux NaOH/NaCl powders at 550 1C and then soaking for 6–24 h followed by slow cooling down to 420 1C, many shiny and thin plate-like NaxCo2O4 crystals with sizes up to 5 mm were obtained. Compared to the typical NaCl flux method, this novel approach lowers the crystal-growth temperature almost by half. The typical raw materials (Na2CO3 with Co3O4) used in the formerly reported NaCl flux method did not work with this new method, only fine black powders and micro-crystallites instead of NaxCo2O4 single-crystals were obtained.
2. Experimental procedure The Ag tubes with diameter of 3800 were used as the reaction containers for the starting materials of Co powders (Aldrich, 99.9%), NaOH (Alfa Aesar, 499%) and NaCl (Alfa Aesar, 499%). One end of the Ag tube
Fig. 1. Typical plate-like NaxCo2O4 single crystals. (a) scale: mm.
ARTICLE IN PRESS X. Tang et al. / Journal of Crystal Growth 310 (2008) 665–670
(002)
the two strong peaks (0 0 2) (0 0 4) suggest that the asgrown crystals are inclined toward c-axis, which implies that the flat surface (growth direction) is parallel to abplane. This is also verified by the single-crystal XRD results (not shown). The chemical composition of as-grown crystals NaxCo2O4 was analyzed by the EDX equipped in SEM Hitachi-3500N. Due to the complex nature of crystal growth via the flux method, the results from different batches vary quite differently from each other, even for the batches synthesized under the completely same conditions. The as-grown crystals were found to be Na deficient with Na concentration x0.3–0.8. It is found that the composition of single-crystals is not very sensitive to the ratio of Co: flux, in other words, the amount of flux cannot greatly help increase the Na content. This is the first report of direct synthesis of NaxCo2O4 single-crystals with low Na content. It was formerly reported that the hydrated superconducting compound Na0.35CoO2 1.3H2O was prepared from the parent phase Na0.7CoO2 powder, which was first immersed in Br2/ CH3CN solution for five days in order to deintercalate Na+ ions, then filtered and washed with CH3CN and distilled water, and finally dried in an ambient atmosphere [6]. (Please note that Na0.35CoO2 1.3H2O can be denoted as Na0.7Co2O4 2.6H2O as the manner used in this paper.) The Na content in the hydrated compound is in the range
NaxCo2O4 crystal is inclined to c-axis
(004)
Intensity
Fig. 1 presents the images of some selected NaxCo2O4 single-crystals. Due to its hexagonal layered crystal structure, the crystals are thin platelets and exhibit the expected hexagonal shape. The typical size of NaxCo2O4 crystals are 0.2–2.0 mm 0.2–2.0 mm 10–50 mm. Many small crystals have a nearly perfect hexagonal shape and shiny surface, such as the one shown in Fig. 1(b). On the other hand, most large crystals with sizes up to 5 mm do not exhibit the perfect hexagonal shape, and there are some small cracks and/or holes observed on the surface. This is not a surprise for the crystals possessing a layered structure grown by the flux method. During the crystal growth, due to the thermal or composition fluctuation, the atomic deposition and arrangement to form the crystal at the melt–crystal interface are very easily interrupted. In order to obtain larger crystals with better quality, variant processing parameters, such as the ratio of NaOH to NaCl in the flux medium, ratio of Co to NaOH/NaCl flux, firing temperature, soaking time, cooling rate and firing atmosphere have been further tuned, as summarized in Table 1. According to the NaOH/NaCl phase diagram [18], depending on the ratio of NaOH:NaCl in flux, three different firing temperatures (550, 700, 780 1C) were employed. Slow cooling rates, 0.3–5.0 1C/h, were applied. In addition, the syntheses were conducted under three different firing atmospheres: air, oxygen and vacuum. The slower cooling rates and extension of the soaking time are found to enhance the crystal yield. The elevated processing temperature helps increase the overall crystallinity of the obtained crystals, as evidenced in the narrowing of the Bragg peaks in XRD patterns, however it does not greatly facilitate larger crystal size in this case. On the other hand, batches II and III (see Table 1), that were fired at higher temperatures, not only yielded smaller crystals but also had a much smaller amount of flux leftover than batch I. So we believe that the volatilization of Na is getting worse at elevated temperatures for batches II and III. Other Co sources in various forms were also used in the starting materials, such as Co pellets, Co wires and Co bulk pieces, but it was found not to facilitate the crystal growth in size and yield. Actually, only their very surfaces were dissolved in the flux at such low processing temperatures. As shown in Fig. 2, the XRD pattern indicates that the crystal has only one single phase, g-NaCo2O4. Apparently
667
10
20
30
40
50
60
70
80
2θ Fig. 2. XRD pattern of NaxCo2O4 single-crystals.
Table 1 Processing conditions for NaxCo2O4 crystal growth Batches
I
II
III
NaOH:NaCl (molar ratio) Corresponding Tm (1C) from phase diagram Setup firing temperature (1C) Co: flux (mass ratio) Co source Soak time (h) Cooling rate (1C/h) Atmosphere
6:4 4:6 2:8 500 650 740 550 700 780 1:1, 1:5, 1:8, 1:10, 1:15, 1:20, 1:40 Co powders, cold-pressed Co pellets, Co pieces, Co wires, as-grown NaxCo2O4 crystals 6, 24, 72, 120 0.3, 0.6, 1.2, 2.0, 5.0 air, oxygen, vacuum
ARTICLE IN PRESS X. Tang et al. / Journal of Crystal Growth 310 (2008) 665–670
Obviously oxygen is needed to oxidize Co in this reaction. That is consistent with the observation that almost no crystals were obtained from the batch fired under vacuum. Actually more oxygen is needed in this reaction than the typical high-temperature NaCl flux method using Na2CO3 and Co3O4 as the starting materials. Here the Co is oxidized from Co0+ in Co powders (instead from the mixed valence Co2+ and Co3+ in Co3O4 in the typical NaCl flux method) to Co3+ and Co4+ in NaCo2O4. However, the experimental results revealed that the oxygen-firing atmosphere does not greatly aid the crystal synthesis. Transport properties of the as-grown NaxCo2O4 singlecrystals were measured by our custom-designed system and the results are shown in Fig. 3. The thermopower a (52 mV/K at 300 K) is smaller than the crystals (x1) grown via the typical high-temperature NaCl flux method, while the resistivity r is higher than that of those crystals. The r displays a metallic-like temperature behavior with a positive temperature coefficient. The higher r for this Nadeficient crystal (x0.54 as revealed by EDX) is likely due to the lower carrier concentration, since Na is the donor contributing electrons to the conducting CoO2 layers. The relatively small a can also be explained by the low-Na composition, since the a limit at high-temperature was found to be related to the ratio of Co3+/Co4+ ions, which 60
ρ 0.8
30
α 0.6
20 10
0.4
5 4 3 2 1
0
Lattice thermal conductivity KL
40
Resistivity ρ (mΩ-cm)
1
6
0
1.2
50 Thermopower α (μV/K)
7
50
100
150
200
250
160
6
130
5
100
4 70 3 2
0.2 0.1
40
0 10
1
110
210
T (K)
100
150
200
250
0.2 300
Temperature (K) Fig. 3. In-plane resistivity r and thermopower a of Na-deficient NaxCo2O4 single-crystal.
310
10 -20
0 50
350
7
0
0
300
Heat capacity Cv
NaCl 2Co þ xNaOH þ 2 x4 O2 !Nax Co2 O4 þ x2 H2 O:
is directly associated with the Na composition [19]. If we take Na content x ¼ 0.5, the high-temperature limit of a which can be estimated by the equation in Ref. [19] will be 60 mV/K, which is consistent with the measured a value in Fig. 3. To our knowledge, there are only a couple of reports of thermal conductivity k measurement for single-crystals of NaxCo2O4, because it is absolutely a technical challenge to measure k precisely (considering serious heat loss) for a thin and small crystal [10,20]. Fujita et al. [10] reported inplane k of NaxCo2O4 crystals above 300 K, which was calculated via measuring the in-plane thermal diffusivity, density and heat capacity. Satake et al. [20] employed a revised Harman method to measure the in-plane k of NaxCo2O4 crystals below 300 K, which is an indirect measurement approach through analyzing the heat-flow balance (based on Peltier effect) across the sample and combing the Seebeck values of the sample (must be
KL/Cv ~ lph
of Na composition of the Na-deficient crystals synthesized by this new flux method. Therefore, this novel lowtemperature flux method may be adopted to directly prepare the low-Na content hydrated superconducting compound Na0.35CoO2 1.3H2O with no need of deintercalating Na+ ions from its parent phase. The possible chemical reaction for the crystal growth via this new low-temperature flux method is believed to be as the following:
In-plan ethermal conductivity κT (W/m-K)
668
100
200 T (K)
300
Fig. 4. (a) In-plane thermal conductivity kT of Na-deficient NaxCo2O4 single-crystal. (b). Lattice thermal conductivity kL and heat capacity Cv of a Na-deficient NaxCo2O4 single-crystal. Inset is the temperature dependence of phonon mean free path lph.
ARTICLE IN PRESS X. Tang et al. / Journal of Crystal Growth 310 (2008) 665–670
0.014 Magnetic susceptibility χ (emu/molar)
measured using other technique) [20]. However, using the custom-designed ‘‘Parallel Thermal Conductance’’ (PTC) system developed in our lab [15,16], we were able to directly measure the in-plane k of as-grown NaxCo2O4 single-crystal sample with size of 2 0.4 0.03 mm3. As shown in Fig. 4(a), the measured total thermal conductivity kT monotonously increases with temperature increases and does not have the usual behavior for crystalline materials where kT reaches a peak and then starts to decrease with temperature. The measured in-plane kT at 300 K is 7 W/m-K, close to the data reported in Ref. [20]. It should be noted that this is the first report of direct measurement of in-plane k of NaxCo2O4 single-crystal below 300 K. We feel that our method is a more reliable measurement technique with better determination of the in-plane k. From the measured kT and the electronic thermal conductivity ke estimated from the r via Weidemann– Franz relation, the lattice thermal conductivity kL is found to dominate kT. As we know, kL is proportional to heat capacity Cv, sound velocity vs and phonon mean free path lph via equation kL C v vs l ph . As the vs is usually weak temperature dependent, the similar temperature dependences of Cv and kL shown in Fig. 4(b) hence suggests that the lph get saturated fast with elevated temperature, which we believe is due to the partially occupied disordered Na layer as strong phonon scatter. The temperature dependence of lph for a crystalline material is typical of a U-process; it reaches a peak at 10–20% Debye temperature and follows 1/T at the higher temperature side, finishes with a roughly constant value. This is exactly what we see in the inset of Fig. 4(b), served as an evidence of the crystalline nature of the sample and indicates that the heat transport is dominated by the ‘‘Umklapp-process’’ in the NaxCo2O4 system, where the Na vacancy is the entity of the phonon scattering. Therefore, in this case the temperature dependence of the kL is not a signature of crystal nature, but the temperature dependence of the lph is. Magnetic properties of the synthesized NaxCo2O4 singlecrystals were investigated on a SQUID with magnetic field of 0.01–0.5 T at temperature range of 5–310 K. As presented in Fig. 5, the magnetic susceptibility w of Nadeficient NaxCo2O4 single-crystals obeys the Curie–Weiss law w ¼ wo+C/(Ty), where both temperature-independent Pauli paramagnetic component wo and Curie–Weiss paramagnetic component are included. As a comparison, the w values of polycrystalline Na1.2Co2O4 and Na1.5Co2O4 samples are also plotted [21]. The temperature dependence of w of Na-deficient NaxCo2O4 single-crystals is similar to that of polycrystalline samples, but the magnitude is much larger. It is understandable that the magnetic moment in NaxCo2O4 is attributed to the magnetic Co4+ ions with 12 spin, the low Na content in the as-grown NaxCo2O4 singlecrystals means the large ratio of Co4+ to Co3+ ions, which consequentially leads to the large value of w. The thermal stability of the Na-deficient NaxCo2O4 single-crystals (30 mg) were investigated by DSC–TGA
669
0.012 0.01 as-grown Nax Co2 O4 single crystals
0.008 0.006 0.004
polycrystalline Na1.2Co2 O4
0.002 0 polycrystalline Na1.5Co2 O4 -0.002 0
50
100
150 200 Temperature (K)
250
300
Fig. 5. Magnetic susceptibility w of Na-deficient NaxCo2O4 single-crystals. As comparison, the polycrystalline samples of Na1.2Co2O4 and Na1.5Co2O4 are also plotted.
system, the results are shown in Fig. 6. The relatively large weight loss 7.6% at low temperature is likely due to the vaporization of water. Before running the DSC–TGA experiment the single-crystals were thoroughly washed using deionized water and only dried under the high beam heat lamp for 30 min, so probably some residual water remained. In the temperature region of 350–900 1C, there is no clear weight loss which suggests the crystals are quite stable. However, a large weight loss (6.1 wt%) starting from 900 1C is observed. This is most likely due to the Na vaporization. After the first run, these crystals were subsequently run one more time in the DSC–TGA system. As shown in Fig. 6(b), since water is totally removed from the crystals in the first run, there is almost no clear weight loss at low temperature in the second run. Again the weight loss starts from 900 1C. Therefore, based on the DSC–TGA results, we can roughly consider that in the NaxCo2O4 crystals the sodium begins to vaporize at 900 1C and above. The microstructure and composition of NaxCo2O4 single-crystals before and after running the DSC–TGA experiments were studied by SEM Hitachi-3500N equipped with EDX. As displayed in Fig. 7(a), the as-grown crystals have dense and fairly smooth surface, despite some growing steps are observed. After firing crystals up to 1000 1C in DSC–TGA, the surface shows an unusual pattern with many continuous grooves and discrete holes, as shown in Fig. 7(b). EDX showed that there is only trivial amount of Na in the crystal, which suggests that the formation of holes is due to the volatilization of Na. 4. Conclusions For the first time, NaxCo2O4 single-crystals were successfully synthesized at low temperature 550 1C by
ARTICLE IN PRESS X. Tang et al. / Journal of Crystal Growth 310 (2008) 665–670
670
20
22.5
0.5
22
Heat flow
7.6 wt% -20
21 20.5
-40
20
weight 6.1 wt%
-60
-0.5
19.5 19
-80 0
200
400 600 800 Temperature (°C)
900°C
0
21.5 Weight (mg)
Heat flow
18.5 1000
weight loss (%)
0
-1 -1.5 -2 -2.5 -3 -3.5 0
20
40
60 80 100 Time (minutes)
120
140
Fig. 6. DSC–TGA curves of Na-deficient NaxCo2O4 single-crystals. (a) 1st run, (b) 2nd run.
References
Fig. 7. SEM microstructures of NaxCo2O4 crystals before and after the DSC–TGA experiments: (a) before DSC–TGA experiment, (b) after DSC–TGA runs.
using NaOH/NaCl as the flux and Co powders as the Co source. The hexagonal shape of the as-grown NaxCo2O4 platelets reflects its hexagonal layered crystal structure. Variant growth conditions, such as the ratio of NaOH to NaCl, ratio of Co to NaOH/NaCl flux, fire temperature, soaking time and cooling rate were tuned and optimized. Longer soak times and slower cooling rates aid the crystal growth; and the chemical composition of crystals are not greatly related to the ratio of Co: flux. The as-grown crystals are found to be Na-deficient, which is in agreement with the observed low thermopower a, high resistivity r and large magnetic susceptibility w. The in-plane total kT of Na-deficient single-crystal was directly measured via our custom-designed ‘‘PTC’ system to be 7 W/m K at 300 K. The studies of thermal stability and microstructure performed by DSC–TGA and SEM suggests that the Nadeficient NaxCo2O4 crystals are stable at temperature up to 900 1C. Acknowledgment The authors would like to appreciate the financial support from ONR DEPSCoR program ONR ]N0001403-0787 and SC EPSCoR/Clemson University cost share.
[1] C. Fouassier, G. Matejka, J.M. Reau, P. Hagenmuller, J. Solid State Chem. 6 (1973) 532. [2] C. Delmas, A. Maazaz, C. Fouassier, J.M. Hagenmuller, Mater. Res. Bull. 14 (1979) 329. [3] C. Delmas, C. Fouassier, P. Hagenmuller, Physica B 99 (1980) 81. [4] J. Molenda, C. Delmas, P. Hagenmuller, Mater. Res. Bull. 9–10 (1983) 431. [5] I. Terasaki, Y. Sasago, K. Uchinokura, Phys. Rev. B 56 (1997) 12685. [6] K. Takada, H. Sakurai, E.M. Takayama, F. Izumi, R.A. Dilanian, T. Sasaki, Nature 422 (2003) 53. [7] R.E. Schaak, T. Klimczuk, M.L. Foo, R.J. Cava, Nature 424 (2003) 527. [8] K. Takada, H. Sakurai, E.M. Takayama, F. Izumi, R.A. Dilanian, T. Sasaki, Physica C 412 (2004) 14. [9] H. Sakurai, K. Takada, T. Sasaki, F. Izumi, R.A. Dilanian, M.E. Takayama, J. Phys. Soc. Jpn. 73 (2004) 2590. [10] K. Fujita, T. Mochida, K. Nakamura, Jpn. J. Appl. Phys. 40 (Part 1) (2001) 4644. [11] S. Bayrakci, C. Bernhard, D.P. Chen, B. Keimer, R.K. Kremer, P. Lemmens, C.T. Lin, C. Niedermayer, J. Strempfer, Phys. Rev. B 69 (2004) 100410. [12] F.C. Chou, J.H. Cho, P.A. Lee, E.T. Abel, K. Matan, Y.S. Lee, Phys. Rev. Lett. 92 (2004) 157004. [13] D. Prabhakaran, A.T. Boothroyd, R. Coldea, N.R. Charnley, J. Crystal Growth 271 (2004) 74. [14] A.L. Pope, R.T. Littleton, T.M. Tritt, Rev. Sci. Instrum. 72 (2001) 3129. [15] X. Tang, K. Aaron, J. He, T. Tritt, Phys. stat. sol. (RRL), submitted. [16] B.M. Zawilski, R.T. Littleton, M. Terry, T. Tritt, Rev. Sci. Instrum. 72 (2001) 1770. [17] B.M. Zawilski, R.T. Littleton, T.T. Tritt, Appl. Phys. Lett. 77 (2000) 2319. [18] E.M. Levin, C.R. Robbins, H.F. McMurdie (Eds.), Phase Diagram for Ceramists, vol. 7, American Ceramic Society, (1989, Fig. 7099). [19] W. Koshibae, K. Tsutsui, S. Maekawa, Phys. Rev. B 62 (2000) 6869. [20] A. Satake, H. Tanaka, T. Ohkawa, et al., J. Appl. Phys. 96 (2004) 931. This method is not a direct measurement of thermal conductivity of single crystals and it uses a technique that is sometimes employed for well-characterized TE materials. [21] X. Tang, J. He, E. Abbott, J. Kolis, T. Tritt, in preparation.
