Recent progress in thermoelectric layered cobalt oxide thin films

Abstract

Oxide-based thermoelectric materials that show a high figure of merit are promising because of their good chemical and thermal stabilities and their relative harmlessness compared with chalcogenide-based state-of-the-art thermoelectric materials. Although several high-ZT thermoelectric oxides (ZT > 1) have been reported thus far, their reliability levels are low due to the lack of careful observations of their stabilities at elevated temperatures. Herein, we review the epitaxial film growth and thermoelectric properties of representative p-type layered cobalt oxides: Na_3/4CoO₂, Ca_1/3CoO₂, Sr_1/3CoO₂, Ba_1/3CoO₂, and Ca₃Co₄O₉. Among these specimens, Ba_1/3CoO₂ and Ca₃Co₄O₉ are stable in air at elevated temperatures (~600 °C). The ZT of Ba_1/3CoO₂ reaches ~ 0.55 at 600 °C in air, which is reliable and the highest among thermoelectric oxides. Moreover, this value is comparable to those of p-type PbTe and p-type SiGe.

Thermoelectrics

Today, most energy resources are discharged as waste heat into the environment without being applied. Such exhaust heat reaches approximately 2/3 of the primary energy. Hence, thermoelectric energy conversion technology has attracted great attention for converting waste heat into electricity^1,2. The principle of thermoelectric energy conversion was first discovered by T.J. Seebeck in 1821³. He found that a voltage is generated between two ends of a metal bar by introducing a temperature difference. Thus, when electric loads are connected at both ends of a metal bar, an electric current can be obtained. This phenomenon is called the Seebeck effect. Conversely, in 1834, J.C.A. Peltier discovered that heating or cooling of the junctions can occur during electric current application to a heterogeneous metal circuit. This phenomenon is called the Peltier effect, and it has been commercially applied in electronic refrigerators, among other applications.

Generally, the performance of thermoelectric materials is evaluated in terms of a dimensionless figure of merit, ZT = S²·σ·T·κ⁻¹, where Z is a figure of merit, T is the absolute temperature, S is the thermopower (≡ Seebeck coefficient), σ is the electrical conductivity, and κ is the thermal conductivity. The energy conversion efficiency from transforming the temperature difference into electricity increases as ZT increases. Thus, to realize efficient thermoelectric energy conversion, three physical properties are needed for thermoelectric materials: (1) low κ, which is needed to introduce a large temperature difference into both ends of the material; (2) high σ, which is needed to reduce the internal resistance of the material; and (3) high S, which is needed to obtain a high voltage.

The ZT values of practical thermoelectric materials, such as Bi₂Te₃ and PbTe, are ~1, which is the lowest value needed for practical applications^4,5. Recently, several high-ZT materials (ZT > 1) have been developed sequentially based on heavy metal alloys, including SnSe, PbSe, GeTe, and oxychalcogenides (BiCuSeO)^{5,6,7,8,9,10,11,12,13,14,15,16,17}. Since thermoelectric devices can directly convert a temperature difference into electricity, some automobile companies have developed thermoelectric-assisted hybrid automobiles^18,19,20,21 while considering the outside temperature of the exhaust pipe to be ~700 °C. However, these thermoelectric materials are not appealing, particularly when operating at such high temperatures, because decomposition, vaporization, and melting of the constituents can easily occur. Furthermore, the use of these heavy metals should be limited to specific environments, such as space, because they are mostly toxic, low in abundance as natural resources, and not environmentally benign.

To overcome these issues, metal oxides have attracted much attention as thermoelectric power generation materials at high temperatures based on their potential advantages over heavy metal alloys in terms of chemical and thermal robustness^22,23,24. From the 1950s to the 1970s, there was a boom in the search for oxide thermoelectric materials. At this time, researchers in the United States used thermoelectric effects to investigate the intrinsic properties of grown oxide single crystals. Since the 1990s, the search for oxide thermoelectric materials, which originated in Japan, has spread all over the world, including in Europe, the United States, Asia, and India. To date, it has been reported that some oxides could exhibit parameters of ZT > 1, exceeding that of PbTe.

History of thermoelectric oxides

There is a long history of thermoelectric oxides. From the 1950s to the 1970s (1^st boom), the thermoelectric properties of many simple conducting oxides, including CdO (1949, Hogarth et al.²⁵), NiO (1957, Parravano²⁶), ZnO (1959, Hutson²⁷), In₂O₃ (1962, Arvin²⁸), SrTiO₃ (1964, Frederikse et al.²⁹), rutile-TiO₂ (1965, Thurber et al.³⁰), SnO₂ (1965, Marley and Dockerty³¹), Cu₂O (1969, Young and Schwartz³²), and Fe₃O₄ (1970, Griffiths et al.³³), were studied to obtain the fundamental physical properties of conducting oxides, such as carrier effective mass. After the discovery of cuprous oxide-based high-T_c superconducting oxides in 1986³⁴, the thermoelectric properties of superconducting oxides, including La₂CuO₄ (1987, Cooper et al.³⁵), La-Ba-Cu-O (1987, Chen et al.³⁶), YBa₂Cu₃O_7−δ (1988, Lee et al.³⁷), and Tl-Ca-Ba-Cu-O (1988, Mitra et al.³⁸), were reported sequentially to clarify the superconducting transition (2^nd boom).

After the research boom in the field of high-T_c superconductors, two Japanese researchers, Ohtaki and Terasaki, reported that oxides had good thermoelectric performance, including CaMnO₃ (1995, Ohtaki et al.³⁹), Al-doped ZnO (1996, Ohtaki et al.⁴⁰), and Na_xCoO₂ (1997, Terasaki et al.⁴¹). These reports triggered the 3^rd boom of thermoelectric oxide research in the 2000s. As a result of energetic exploratory research on good thermoelectric oxides, Ca₃Co₄O₉ (2000, Masset et al.⁴² and Funahashi et al.⁴³) and electron-doped SrTiO₃ (2001, Okuda et al., La-doped⁴⁴, 2005, Ohta et al. Nb-doped^45,46) were discovered. In 2012, Fergus reviewed the thermoelectric properties of promising oxides (mostly bulk ceramics), including Ca₃Co₄O₉, Na_xCoO₂, SrTiO₃, CaMnO₃, and ZnO⁴⁷ (also see the references therein). Recently, several scholars that have reported rather high ZT values for oxide ceramics. Acharya et al.⁴⁸ reported that SrTi_0.85Nb_0.15O₃ ceramics sintered with graphite flakes exhibit ZT = 1.4 at 1050 K, and Biswas et al.⁴⁹ reported that Al-doped ZnO ceramics sintered with reduced graphene oxide exhibit ZT = 0.5 at 1100 K. Although these reported ZTs are very attractive as practical thermoelectric materials, there is still no practical application for them, probably due to the low reliability of the ZT values.

To clarify the intrinsic thermoelectric properties of oxides, we focused on high-quality epitaxial films with stepped and terraced surfaces. As a result, we fabricated high-quality epitaxial films of several thermoelectric oxides, including Na_3/4CoO₂^50,51, Sr_1/3CoO₂⁵², Ca_1/3CoO₂⁵³, Ba_1/3CoO₂⁵⁴, Ca₃Co₄O₉^53,55, SrTiO₃:Nb⁴⁶, TiO₂:Nb⁵⁶ and SrO(SrTiO₃):Nb⁵⁷. Among these oxides, we found that Ba_1/3CoO₂ epitaxial films exhibited a ZT value of ~0.55 at 600 °C in air, which is the highest and most reliable value among the reported thermoelectric oxides⁵⁴. In this context, we reviewed the epitaxial film growth and thermoelectric properties of four representative p-type layered cobalt oxide films based on our efforts: Na_3/4CoO₂^50,51, Sr_1/3CoO₂⁵², Ca_1/3CoO₂⁵³, and Ba_1/3CoO₂⁵⁴.

Epitaxial film growth of A _xCoO₂ (A _x = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3)

Na_3/4CoO₂ ^50,51

Figure 1 shows the schematic crystal structure of A_xCoO₂ (A_x = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3). In the case of A_xCoO₂, the rigid CoO₂ layer and mobile A_x layer are alternately stacked along the c-axis. In 1997, Terasaki et al. discovered that a NaCo₂O₄ (≡Na_xCoO₂, x ~ 3/4) single crystal with a two-dimensional layered structure exhibits a very large power factor S² ∙ σ of 5 mW m⁻¹ K⁻² in the in-plane direction at room temperature⁴¹. After this discovery, the electronic structure^58,59,60, crystal structure^61,62, and Na-composition dependence of the thermoelectric properties⁶² of Na_xCoO₂ were energetically studied to understand the origin of the unusually large S. In 2001, Fujita and coworkers fabricated Na_xCoO₂ single crystals and reported that they exhibited ZT values of ~1.2 at 800 K⁶³. This material has attracted much attention because it can be converted into a superconductor (T_c ~ 4.7 K) by introducing H₂O molecules into a layer between the two adjacent CoO₂⁻ layers⁶⁴.

Fig. 1: Schematic illustration of the crystal structure of A_xCoO₂.

Left: A_x = Na_3/4, Right: A_x = Ca_1/3, Sr_1/3, and Ba_1/3.

To clarify the intrinsic thermoelectric properties of Na_3/4CoO₂, we fabricated thin epitaxial films of Na_3/4CoO₂. First, we tried to fabricate Na_3/4CoO₂ epitaxial films by the conventional pulsed laser deposition (PLD) technique, but we failed. Several reports have been written on the thin film growth of Na_3/4CoO₂ by PLD. However, the film quality in terms of crystallographic orientation, surface morphology, and lateral grain size is insufficient. In the case of Na_3/4CoO₂ film growth by PLD, it is very difficult to control the Na concentrations in the films at high temperatures in a vacuum environment due to the high vapor pressure of the Na species.

To overcome this difficulty, we modified the reactive solid-phase epitaxy (R-SPE)⁵⁰ method that was developed to fabricate single-crystal films of InGaO₃(ZnO)_m (m = integer). Figure 2 schematically illustrates the R-SPE procedure. Step 1: A highly (111)-oriented CoO epitaxial film was deposited on a (0001)-α-Al₂O₃ substrate at 700 °C by the PLD technique using a Co₃O₄ sintered disk as a target. Step 2: The surface of the PLD-deposited CoO film was fully capped by an yttria-stabilized zirconia (YSZ) single-crystalline plate to keep the surface clean. Step 3: NaHCO₃ powder was put on the YSZ plate. Step 4: The sandwich specimen was annealed at 700 °C for 1 h in air. Notably, several researchers have used this R-SPE method for fabricating Na_xCoO₂ epitaxial films^65,66,67 since the resultant film quality, especially in surface morphology, is better than that of PLD-grown films^68,69.

Fig. 2: Film growth procedure.

Schematic illustration of reactive solid-phase epitaxy (R-SPE) for the preparation of Na_xCoO₂ epitaxial films.

The out-of-plane (Fig. 3a) and in-plane (Fig. 3b) X-ray diffraction patterns of the resultant Na_xCoO₂ film clearly indicate that epitaxial growth occurred, showing the effectiveness of the R-SPE method. The chemical composition of the obtained film was evaluated to be x = 0.83 by X-ray fluorescence (XRF) measurements. The Na content in the present film was slightly higher than the reported values in the as-grown bulk sample x = 0.7⁷⁰, likely because the amorphous layer at the interface contained some Na ions⁵¹. Figure 3c shows an atomic force microscopy (AFM) image of the Na_xCoO₂ film. A step-like structure composed of several flake-like domains could be observed. The step increment was approximately 3 nm, which was three times longer than the c-axis length of Na_xCoO₂, suggesting that step bunching occurred during annealing at 700 °C.

Fig. 3: X-ray diffraction (XRD) patterns and topographic atomic force microscopy (AFM) images of Na_xCoO₂ epitaxial films.

a Out-of-plane XRD. b In-plane XRD. c Topographic AFM image. This figure was reproduced with permission^50,82.

As described in the next section, the R-SPE-grown Na_0.7CoO₂ epitaxial film was very useful for fabricating epitaxial films of LiCoO₂⁷¹, Sr_0.5CoO₂⁵², Ca_0.33CoO₂^55,72, and Ca₃Co₄O₉⁵⁵. Furthermore, the Na_0.7CoO₂ epitaxial film could be converted into a superconducting sodium cobalt oxyhydrate, Na_0.3CoO₂·1.3H₂O (T_c ~ 4 K), by dipping in HNO₃ for oxidation treatment followed by dipping in NaCl aqueous solution for hydration treatment⁷³. Furthermore, peeling-off of the Na_0.7CoO₂ epitaxial film from the α-Al₂O₃ substrate was possible, and the peeled film could be pasted on the other substrate⁵¹.

Ca_1/3CoO₂

Powder syntheses of Ca_xCoO₂ (x = 0.3, 0.35 and 0.5) were reported in 1996 by Cushing et al.^74,75 The scholars used specimens of the sodium cobalt oxide Na_xCoO₂ (0.6 ≤ x ≤ 1.0) as precursors and performed multivalent ion-exchange reactions.

$${{\rm{Na}}}_{x}{{\rm{CoO}}}_{2}+x/2{\rm{Ca}}{({{\rm{NO}}}_{3})}_{2}\to {{\rm{Ca}}}_{x/2}{{\rm{CoO}}}_{2}+x\,{{\rm{NaNO}}}_{3}$$

Stoichiometric amounts of the A_xCoO₂ precursors were combined with anhydrous Ca(NO₃)₂ in evacuated sealed glass tubes and heated at 350 °C for 48 h.

There is an interesting feature in the Ca_xCoO₂ system: there are two superstructures of cation ordering. In 2006, Yang et al. discovered that there are two common well-defined cation ordered states corresponding to the $2a\,\times \,\sqrt{3}a$ orthorhombic superstructure at approximately x = 1/2 and the $\sqrt{3}a\,\times \,\sqrt{3}a$ hexagonal superstructure at approximately x = 1/3⁷⁶. In 2006, Sugiura et al. fabricated high-quality Ca_0.48CoO₂ epitaxial films by ion-exchange reactions. Na_3/4CoO₂ epitaxial films were heated together with Ca(NO₃)₂ powder at 300 °C for 0.5 h in air⁵⁵. In 2008, Huang et al. directly observed two instances of Ca ordering in Ca_1/3CoO₂ epitaxial films using high-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM)^77,78.

In 2009, Sugiura et al. found that the structural transformation of Ca_1/3CoO₂ occurred at approximately 300 °C⁷². The $\sqrt{3}a\,\times \,\sqrt{3}a$ hexagonal phase transformed into the $2a\,\times \,\sqrt{3}a$ orthorhombic phase when heated in air. The scholars found that the orthorhombic phase showed insulating electron transport, whereas the hexagonal phase showed metallic transport. Interestingly, the temperature dependence of the thermopower of Ca_1/3CoO₂ epitaxial films was similar, independent of the crystallographic phases.

Sr_1/3CoO₂

The most serious drawback of Na_3/4CoO₂ is its low chemical stability against water. Na_3/4CoO₂ is easily decomposed into insulating Co(OH)₂ under high-humidity conditions (temperature, 80 °C; humidity, ~80%) because Na⁺ ions can easily dissolve in water. To address this issue, Sugiura and coworkers hypothesized that modification of the chemical composition improves the chemical stability without degrading the thermoelectric performance.

In 2002, Ishikawa et al. reported that the thermoelectric properties of the Sr_1/3CoO₂ ceramic synthesized by sintering at 400 °C were quite low relative to those of Na_3/4CoO₂. Although a high-density and single-crystal Sr_1/3CoO₂ ceramics are preferable for clarifying the intrinsic properties, this process is extremely difficult because the phase transition of Sr_1/3CoO₂ occurs at a relatively low temperature (~400 °C). To examine the intrinsic thermoelectric properties of Sr_1/3CoO₂, Sugiura and coworkers fabricated high-quality epitaxial films of Sr_1/3CoO₂ because epitaxial films generally exhibit intrinsic carrier transport properties, similar to those of bulk single crystals. The Sr_1/3CoO₂ epitaxial films exhibits better chemical stability than Na_3/4CoO₂ while retaining good thermoelectric properties.

Ba_1/3CoO₂

Recently, we found a reliable high-ZT thermoelectric oxide, Ba_1/3CoO₂. The crystal structure and electrical properties of the Ba_1/3CoO₂ epitaxial films were maintained to 600 °C. The power factor gradually increased to ~1.2 mW m⁻¹ K⁻², and the thermal conductivity gradually decreased to ~1.9 W m⁻¹ K⁻¹ with increasing temperature to 600 °C. Consequently, the ZT reached ~0.55 at 600 °C in air, which was the highest value among oxides and comparable to those of p-type PbTe and p-type SiGe.

Notably, initial investigations on Ba_xCoO₂-based thermoelectric materials have been conducted by Liu et al.^79,80 The scholars fabricated Ba_xCoO₂ (x = 0.19, 0.28, 0.30, 0.33) ceramics by using Ba_xCoO₂ powders synthesized through the ion-exchange technique from Na_0.7CoO₂. The researchers measured the thermoelectric properties to 800 K and found that ZT values ranging from 0.14 to 0.21 were dependent on Ba content. In their research, thermal conductivity did not experience effective suppression from polycrystalline grain boundaries, while electron transport properties deteriorated substantially. Therefore, high-quality epitaxial films are ideal for elucidating the intrinsic properties of Ba_xCoO₂-based thermoelectric materials, which is essential for developing high-performance thermoelectric oxides.

Thermoelectric properties of A _xCoO₂ (A _x = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3) epitaxial films

Systematic investigations on the thermoelectric properties of A_xCoO₂ (A_x = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3) started from our findings that heavy ion substitution at the A-site of A_xCoO₂ effectively reduces the in-plane thermal conductivity⁸¹. By fabricating A_xCoO₂ (A_x = Li₁, Na_0.75, Ca_0.33, Sr_0.33, La_0.3) epitaxial films on (0001) α-Al₂O₃ substrates by conducting R-SPE and an ion exchange process, we clarify the A-site ion mass-dependent thermal conductivity of A_xCoO₂^50,51. As shown in Fig. 4, the in-plane thermal conductivity (κ_||) obviously decreases with the A-site ion mass due to the mismatch of the impedance between the cation layer and CoO₂ layers. This impedance hinders coupling of the vibrational modes, while the cross-plane thermal conductivity (κ_⊥) mainly depends on the interfacial scattering.

Fig. 4: Reduced thermal conductivity through heavy ion substitution at the A-site of A_xCoO₂.

As the mass of the A-site ions increases, the in-plane thermal conductivity (κ_||) and cross-plane thermal conductivity (κ_⊥) show decreasing tendencies. The inset shows a schematic illustration of the phonon propagation of A_xCoO₂. Heavy ions display a stronger thermal conductivity suppression effect than light ions. This figure was reproduced with permission⁸¹.

By conducting heavy ion substitution to reduce thermal conductivity, we further fabricate A_xCoO₂ (A_x = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3) epitaxial films on (0001) α-Al₂O₃ and (111) YSZ substrates and compare their room temperature thermoelectric properties⁸². Fig. 5 presents a summary of the room-temperature electrical conductivity (σ_ip), thermopower (S_ip), power factor (PF_ip), thermal conductivity (κ_ip) and figure of merit (ZT_ip) values along the in-plane direction of A_xCoO₂ epitaxial films. The electrical conductivity and thermopower values of all the films show stable changing patterns, resulting in a consistent power factor (Fig. 5a–c). This consistent value suggests perfect electron–phonon decoupling between the A-site ion layer and CoO₂ layer, where ion substitution has almost no effect on the electrical conductivity of the CoO₂ layers. Moreover, the thermal conductivity along the layered direction decreases with the atomic mass of A_x, thereby enhancing the ZT value (Fig. 5d, e). The highest ZT value of ~0.11 can be obtained in the Ba_1/3CoO₂ epitaxial film, reaching a peak value among layered cobalt oxides. In this research, the in-plane thermal conductivity has been deduced based on the experimental results of cross-plane thermal conductivities for differently oriented epitaxial films by varying the substrate orientations. In our latest report, we have directly confirmed the in-plane thermal conductivity through AC calorimetric measurements by using a freestanding Ba_1/3CoO₂ single-crystalline film, yielding a consistent result⁸³.

Fig. 5: Thermoelectric properties of A_xCoO₂ epitaxial films at room temperature.

a In-plane electrical conductivity (σ_ip), b in-plane thermopower (S_ip), c in-plane power factor (PF_ip), d in-plane and out-of-plane thermal conductivity (κ_ip, κ_op), and e in-plane figure of merit (ZT_ip) as a function of the atomic mass A_x in A_xCoO₂. The κ_ip shows a clear decreasing tendency, whereas the PF shows a slight increasing tendency. These trends result in an increasing tendency of ZT. ZT reached 0.11 when A_x = Ba_1/3. This figure was reproduced with permission⁸².

As an emerging candidate for high-performance oxide-based thermoelectric materials, Ba_1/3CoO₂ has promising prospects in applications at elevated temperatures. To elucidate the high-temperature thermoelectric performance, we further conducted high-temperature characterizations of A_xCoO₂ epitaxial films⁵⁴. First, the thermal stabilities of the Na_3/4CoO₂, Ca_1/3CoO₂, Sr_1/3CoO₂, and Ba_1/3CoO₂ epitaxial films were tested by annealing at an elevated temperature for 0.5 h in air. Figure 6 shows the room temperature XRD patterns after heat treatment. The 0002 Na_3/4CoO₂ diffraction peak shrinks above 450 °C, whereas the 111 Co₃O₄ peak appears due to the evaporation of Na. In contrast, the 0002 Ca_1/3CoO₂, 0002 Sr_1/3CoO₂, and 0002 Ba_1/3CoO₂ peaks appear below 650 °C (Fig. 6a). However, the in-plane XRD patterns (Fig. 6b) demonstrate that a phase transition from hexagonal to orthorhombic occurs in the Ca_1/3CoO₂ film when the annealing temperature is above 200 °C⁷². The Sr_1/3CoO₂ film shows a hexagonal–orthorhombic hybridized phase below 450 °C and a single orthorhombic phase above ~450 °C. Only the Ba_1/3CoO₂ film can maintain a stable phase composition to 600 °C, which suggests a strong thermal robustness and a high potential for high-temperature Ba_1/3CoO₂ applications. We have confirmed a similar temperature-dependent behavior from the resistivity variation after heat treatment.

Fig. 6: Changes in the crystal structures of A_xCoO₂ (A_x = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3) films after annealing at high temperatures for 0.5 hours in air.

a Out-of-plane XRD patterns measured at room temperature after each annealing step. The 0002 Na_3/4CoO₂ diffraction peak intensity decreases above 450 °C due to decomposition into Co₃O₄, whereas those for Ca_1/3CoO₂, Sr_1/3CoO₂, and Ba_1/3CoO₂ remain stable to 650 °C. b In-plane XRD patterns measured at room temperature after each annealing step. The 1/3 and 2/3 diffraction peaks of the Ca_1/3CoO₂ film disappear at approximately 200 °C, and the 1/2 diffraction peak appears above 200 °C. The 1/3 and 2/3 diffraction peaks of the Sr_1/3CoO₂ film disappear at approximately 450 °C, and the 1/2 diffraction peak appears above 450 °C. The 1/3 and 2/3 diffraction peaks of the Ba_1/3CoO₂ film are stable to 600 °C. This figure was reproduced with permission⁵⁴.

Finally, we calculated the temperature-dependent ZT of A_xCoO₂ epitaxial films. The ZT values increase with temperature for all films (Fig. 7a). Due to the strongest thermal robustness, the Ba_1/3CoO₂ epitaxial film displays the highest ZT of ~0.55 at 600 °C, which is higher than those of the Ca_1/3CoO₂ and Sr_1/3CoO₂ films. This high ZT value of Ba_1/3CoO₂ is reproducible and reliable. This value is comparable to those of p-type PbTe and p-type SiGe, indicating that Ba_1/3CoO₂ is a suitable candidate for high-temperature thermoelectric applications (Fig. 7b).

Fig. 7: Temperature dependence of the ZT values of Ba_1/3CoO₂ epitaxial films along the in-plane direction.

a Comparison among the four A_xCoO₂ (A = Na_3/4, Ca_1/3, Sr_1/3, and Ba_1/3) films. In all cases, the ZT values increase as the temperature increases. The Ba_1/3CoO₂ epitaxial film exhibits the highest ZT among the four A_xCoO₂ epitaxial films and reaches a ZT value of ~0.55 at 600 °C. b Comparison against commercially available p-type thermoelectric materials. ZT of the Ba_1/3CoO₂ epitaxial film at 600 °C is comparable to the values of p-type PbTe and p-type SiGe. This figure was reproduced with permission⁵⁴.

Summary and prospects

We have reviewed the thermoelectric properties of representative layered cobalt oxides: A_xCoO₂ (A = Na, Ca, Sr, and Ba) and Ca₃Co₄O₉. Although several high-ZT thermoelectric oxides (ZT > 1) have been reported thus far, their reliability is low due to a lack of careful observation of their stabilities at elevated temperatures. We have explained that Ba_1/3CoO₂ is stable in air even at 600 °C and exhibits a high ZT value of 0.55, which is comparable to p-type PbTe. Bulk crystals (single crystal and sintered) are essential for incorporating Ba_1/3CoO₂ into thermoelectric conversion elements. To date, we are researching the growth of large single crystals and are proceeding with the production of sintered bodies. Moreover, better thermoelectric performance may be realized by optimizing the compositions and nanostructures of these crystals.