Giant Enhancement in High-Temperature Thermoelectric Figure-of

May 3, 2017 - Related Content: Bi4TaO8Cl Nano-Photocatalyst: Influence of Local, Average, and Band Structure. Inorganic Chemistry. Bhat, Swain, Feygen...
23 downloads 12 Views 5MB Size
Article pubs.acs.org/IC

Giant Enhancement in High-Temperature Thermoelectric Figure-ofMerit of Layered Cobalt Oxide, LiCoO2, Due to a Dual StrategyCoSubstitution and Lithiation Md. Mofasser Mallick and Satish Vitta* Department of Metallurgical Engineering and Materials Science, Indian Institute of Technology Bombay, Mumbai 400076, India ABSTRACT: The chemical composition of LiCoO2, a layered oxide commonly used as electrode in batteries, was changed to Li1+yCo1−xNixO2 by a combination of substitution and lithiation to enhance the thermoelectric figure-of-merit at high temperatures. Substitution of Ni as well as lithiation does not change the crystal structure, R3̅m. The lattice parameters c and a are found to increase slightly but maintain a nearly constant ratio, 4.99, indicating no lattice distortion. The trivalent Co was substituted with divalent Ni to synthesize LiCo1−xNixO2 series of p-type compounds with x varying up to 0.15. The high-temperature thermopower decreases drastically from ∼600 to 300 μV K−1, while the electrical resistivity drops by an order of magnitude from 1 × 10−2 to 1 × 10−3 Ω m due to substitution of 15 atom % Ni. The total thermal conductivity also decreases from ∼3 to 1.5 W m−1 K−1. Increasing the amount of Li in LiCo0.85Ni0.15O2 changes the thermophysical properties further and leads to enhancement of figure-of-merit. The power factor is found to change from 37.6 μW m−1 K−2 for the base compound to 120 μWm−1 K−2, a significant improvement for a p-type compound. The overall figure-of-merit as a result increases to 0.12 at ∼1100 K due to substitution and lithiation, a giant increase of ∼760% compared to 1 × 10−2 for the pure compound LiCoO2. These substituted and lithiated compounds are found to be extremely stable even after six months and exhibit totally reproducible thermophysical properties.



INTRODUCTION Thermoelectric materials development has gained significant interest in the recent times because of its potential to contribute to reduction in greenhouse gas emissions by facilitating conversion of waste heat into useful electrical power.1 Development of these materials, which have high conversion efficiency, is a significant challenge, as maximizing the thermoelectric figure-of-merit requires optimization of seemingly contradicting material properties.2 It mainly requires optimization of charge carriers concentration vis-à-vis heat transport. Conventionally several classes of materials varying from chalcogenide alloys to half-Heusler alloys and clathrates have been exhaustively investigated.3 These materials however have several limitations such as • stability at high temperatures including low oxidation resistance. • are made of toxic elements. • are made of expensive, less earth-abundant elements. • most alloys have significantly high density. Hence, development of materials that will not have these limitations for practical application is a significantly difficult but not an unsurmountable problem. Oxides are one class of materials that do not have the above-mentioned limitations. They are chemically inert and thermally stable at high temperatures and are generally made of nontoxic, earthabundant elements. Because they are made of mostly light elements, they tend to have low densities, which is an extremely © 2017 American Chemical Society

useful parameter for device application. The single most disadvantageous or limiting factor for their usage, however, is their extremely low electrical conductivity. This can be increased by suitable substitutions or doping, and it forms the prime objective of oxide materials research for thermoelectric application. Oxide materials such as perovskites have been extensively investigated, and they exhibit a predominantly n-type behavior with high thermoelectric figure-of-merit for oxide materials.4 Another interesting class of oxide compounds is layered cobalt oxides that exhibit a predominantly p-type behavior.5 The interest in layered cobalt oxide systems for thermoelectric application is largely due to NaxCoO2−δ, which has a figure-ofmerit of ∼1.0 at 800 K in single crystal state and reduces to ∼0.25 for polycrystalline state. These oxides, both perovskites and layered cobalt oxides, have several salient features(i) the electrons are strongly correlated; that is, charge, spin, orbital, and lattice are interrelated, which gives rise to a variety of electronic phenomena, (ii) the resistivity can be tuned from being weakly semiconducting to barely metallic by suitable substitutions, (iii) thermopower and thermal conductivity can be independently varied due to spin and orbital degrees of freedom, and (iv) spin fluctuations due to variable valence states of cations change the overall properties. In the case of the layered cobalt oxide LiCoO2, the spin state of the compound Received: February 25, 2017 Published: May 3, 2017 5827

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry varies dynamically due to fluctuations between Co3+low spin state, Co3+intermediate spin state, and Co4+high spin state. These fluctuations can also be controlled by suitable substitutions.6 The fluctuations between different valence states influences the charge carrier concentrations and hence influences the transport properties. Delithiation of LiCoO2 is known to change the Co3+/Co4+ ratio and hence lead to semiconductor-to-metal transition.7 In the present work, rhombohedral α-NaFeO2-type layered cobalt oxide system LiCoO2 was chosen for investigation. This compound along with its variants have been extensively investigated as an electrode material for solid-state batteries.8 On the one hand, it has not been investigated for hightemperature thermoelectric application.7a,9 On the other hand, its contemporary oxide NaxCoO2−δ has been studied for hightemperature thermoelectric application due mainly to its “metallic” electrical conductivity.10 This compound however is environmentally unstable and degrades in the presence of moisture.11 Hence LiCoO2 was chosen in the present work, which has a semiconducting behavior with a direct band gap that has been reported to vary from ∼1.5 to 2.5 eV.12 To improve its electrical conductivity without significantly affecting the thermal conductivity a dual strategy has been adopted in the present work. The trivalent Co in the stoichiometric LiCoO2 compound has been substituted with divalent Ni until 15 atom % to improve its electrical conductivity. The electrical conductivity of this substituted compound is further increased by having excess Li in the compound up to 10 atom %. Although Ni substitution has been studied earlier, a systematic investigation of the effect of Ni substitution combined with lithiation on the high-temperature thermoelectric properties has not been reported so far. The high-temperature thermophysical properties have been investigated up to 1100 K together with its crystal and microstructures. The environmental stability of the compounds that exhibit good thermoelectric properties was studied after a period of six months. It was found that the compounds are extremely stable and show reproducible behavior even at high temperatures.



structural refinement can be performed to determine the presence of even extremely small quantities of different phases. The density of the pellets after sintering was determined using the standard Archimedes principle and is given in Table 1 for all the compounds. The density of

Table 1. Density of Sintered Discs Determined by the Archimedes Principal Are Given

a

Sl. No.

chemical composition

density,a g cm−3

1 2 3 4 5 6 7 8

LiCoO2 LiCo0.98Ni0.02O2 LiCo0.96Ni0.04O2 LiCo0.92Ni0.08O2 LiCo0.85Ni0.15O2 Li1.01Co0.85Ni0.15O2 Li1.04Co0.85Ni0.15O2 Li1.10Co0.85Ni0.15O2

4.15 4.098 4.085 4.077 4.028 3.32 3.43 3.55

The density is lower than the theoretical value in all the cases.

Ni-substituted compounds was ∼90%, while the lithiated compounds exhibit lower density for identical processing parameters. Since increasing the temperature to increase the density could lead to loss of Li, the processing parameters were not changed. The microstructure of the compounds was investigated both in the scanning electron microscope as well as in high-resolution transmission electron microscope. To determine the amount of Li present in different compounds, chemical composition was determined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES). The variation of electrical resistivity and Seebeck coefficient in the temperature range from 300 to 1100 K was determined using bar samples in ZEM 3-M8 (ULVAC) system. The bar sample was cut from the circular disc, and electrical conductivity and Seebeck coefficient were simultaneously measured along a direction perpendicular to disc pressing direction. The thermal diffusivity variation of circular disc in this temperature range was measured using the laser flash system LFA 457 (Netzsch) system. The thermal diffusivity of the pellets was measured along the disc pressing direction. Since the discs are polycrystalline with no texturing or orientation, this does not affect the measured properties. The specific heat capacity was determined using sapphire as the standard in a differential scanning calorimeter, 204 F1 (Netzsch). The typical experimental uncertainties associated with these thermophysical properties, electrical resistivity, Seebeck coefficient, and thermal conductivity are ∼4%, 5%, and 8%, respectively. The charge carriers concentration and mobility were determined from the Hall coefficient at room temperature measured using the Lake Shore model 8404 system.

EXPERIMENTAL METHODS

The substituted and lithiated cobalt oxides Li1+yNixCo1−xO2 (0 ≤ x ≤ 0.15; 0 ≤ y ≤ 0.1) were synthesized using standard solid-state reaction methods. High-purity (99.9%) LiCO3, Co2O3, and NiO were mixed in required quantities in an agate mortar for ∼30 min, and the resulting powder mixture was calcined at 1073 K for 12 h in air atmosphere. The calcined powder was pressed into two separate pellets of circular disc shape and sintered at 1123 K for 12 h in ambient air atmosphere. The calcination and sintering temperatures were determined by thermogravimetry/differential thermal analyzer. The formation of layered cobalt oxides takes place as per reaction given below.



RESULTS The X-ray diffraction patterns from the LiCo1−xNixO2 series of compounds are shown in Figure 1a, while those from Li1+yCo0.85Ni0.15O2 series are shown in Figure 1b. In all the compounds, Li1+yCo1−xNixO2 with 0 ≤ x ≤ 0.15; 0 ≤ y ≤ 0.1 the main phase is found to be the rhombohedral lattice layered cobalt oxide with an R3̅m structure. A Rietveld refinement of the diffraction patterns, shown in Figure 1a,b, gives the lattice parameters a and c of the rhombohedral crystal along with other structural parameters, which are given in Table 2. On the one hand, the Li+ ions in this structure are interleaved between edge-sharing CoO6 octahedra, and they occupy the (0,0,0) 3a sites in the crystal. The Co3+ ions, on the other hand, occupy the (0,0,1/2) 3b sites with the O2− ions occupying the (0,0,k) 6c sites. The lattice parameters of the parent compound LiCoO2 are found to be 0.2814(9) nm and 1.4048(7) nm in complete agreement with the literature values.13 The substitution of Co3+ with Ni2+ increases both the lattice parameters a and c by a small magnitude as seen in Figure 1c.

4(1 + y)LiCO3 + 2(1 − x)Co2O3 + 4x NiO → 4Li1 + yCo1 − x NixO2 + 4(1 + y)CO2 + (1 − x + 2y)O2 (1) The density of the sintered pellets was determined by specific gravity bottle method before cutting one of the pellets into a rectangular bar for electrical resistivity and Seebeck coefficient/ thermopower measurement. The remaining piece of this pellet was ground into fine powder and used for structural and chemical analysis. The phases present and crystal structure of the phases present in the compound were determined by high-resolution X-ray diffraction using Cu Kα1 radiation obtained from a 9 kW rotating anode source. The diffractometer is equipped with both incident and diffracted beam monochromators to obtain high-resolution diffraction data, so that 5828

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 1. Powder X-ray diffraction patterns from the two sets of compounds (a) LiCo1−xNixO2 and (b) Li1+yCo0.85Ni0.15O2 together with the refined structure results. The structure in both series of compounds conforms to R3̅m space group, which has been reported for the pure compound. The lattice parameters a and c increase with increasing Ni, while the change in these lattice parameters for lithiation is small (c, d).

Table 2. Structural Parametersa Obtained by Rietiveld Refinement of High-Resolution Diffraction Patterns of All the Different Compounds lattice parameter, nm

site occupancy factor

Sl. No.

chemical composition

a

c

3a

1 2 3 4 5 6 7 8

LiCoO2 LiCo0.98Ni0.02O2 LiCo0.96Ni0.04O2 LiCo0.92Ni0.08O2 LiCo0.85Ni0.15O2 Li1.01Co0.85Ni0.15O2 Li1.04Co0.85Ni0.15O2 Li1.10Co0.85Ni0.15O2

0.2814(9) 0.2815(9) 0.2816(2) 0.2816(8) 0.2819(7) 0.2817(9) 0.2819(4) 0.2821(7)

1.4048(7) 1.4049(1) 1.4050(6) 1.4051(7) 1.4064(8) 1.4056(8) 1.4065(1) 1.4065(1)

Li Li Li Li Li Li Li Li

3b Co Co, Co, Co, Co, Co, Co, Co,

Ni Ni Ni Ni Ni Ni Ni

6c

Li

Co

Ni

O

O O O O O O O O

1 1 1 1 1 1.01 1.04 1.10

1 0.98 0.96 0.92 0.85 0.85 0.85 0.85

0 0.02 0.04 0.08 0.15 0.15 0.15 0.15

0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

a The compounds have a rhombohedral R3̅m structure with a and c as the two lattice parameters. The refinement is performed using the FULLPROF program.

This enhancement is due to the ionic size of Ni2+ 83 pm, which is larger than the ionic size of Co3+ 68.5 pm. However, note that the c/a ratio remains nearly constant 4.99, indicating that the octahedral units do not undergo any distortion due to Ni2+ substitution. Lithiation of the Ni-substituted compounds Li1+yCo0.85Ni0.15O2 does not show any appreciable change in the lattice parameters a and c with increasing amount of excess Li as shown in Figure 1d. The excess Li ions above the

stoichiometric limit were allowed to occupy either their regular 3a (0,0,0) or the 3b (0,0, 1/2) Co3+ sites during structural refinement of X-ray diffraction data. It was observed that the goodness-of-fit in both cases, that is, excess Li occupying 3a or 3b sites, was very similar, ∼1 indicating that structurally Li+ ions can occupy either of the positions. The Li+ ions however are large compared to either Co3+ or Ni2+ ions, and hence positioning these in the 3b sites should result in significant 5829

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 2. Electron microscopy of Li1+yCo1−xNixO2 compounds shows formation of large defect free grains of size 4−5 μm as seen in scanning electron micrographs, (a) LiCoO2, (b) LiCo0.85Ni0.15O2 and (c) Li1.1Co0.85Ni0.15O2. A high-resolution transmission electron micrograph together with selected area diffraction pattern is shown in (d) and indicates formation of defect free grains. The inter-planar spacing shown is for (015) planes and is 0.184 nm.

changes to the lattice parameters and formation of a distorted lattice. Since the lattice parameters are not significantly changed and there is no lattice distortion Figure 1d, the excess Li+ ions are considered to be in 3a position as indicated in Table 2. It is observed that the Ni-substituted compounds LiCo1−xNixO show the presence of a small amount of minority second phase, NiO, together with the layered cobalt oxide phase, Figure 1a. The presence of this minority phase could be observed only in high-resolution X-ray diffraction pattern indicating that this phase is indeed extremely small in quantity. This minority NiO phase however is completely absent in the Li1+yCo0.85Ni0.15O2 compounds as seen in Figure 1b, indicating that excess Li enhances the solubility of Ni2+ in the parent compound. The Li1+yCo0.85Ni0.15O2 compounds are therefore completely single phase in nature without any impurity phases. The microstructure of the compounds observed in both scanning and transmission electron microscopes, Figure 2, shows that the grains are large, 4−5 μm, in all the cases with varying levels of density. They do not show the presence of any second phase, in agreement with X-ray diffraction results, which show a minority phase only in LiCo1−xNixO2 compounds. High-resolution transmission electron micrograph, together with a selected area diffraction pattern, shows the formation of defect-free large grains. Synthesis using conventional solid-state methods is generally known to result in the formation of defectfree large grains as observed here. The amount of Li present in the Li1+yCo0.85Ni0.15O2 compounds was determined by ICPAES, and the results are given in Table 3. It is seen that in the second set of compounds in which excess Li was incorporated,

Table 3. Chemical Composition of the Compounds Li1+yCo0.85Ni0.15O2 Determineda by ICP-AES Method amount of excess Li+ designed (atom %) experimental (atom %)

x = 0.15; y=0 0

x = 0.15; y = 0.01 1

x = 0.15; y = 0.04 4

x = 0.15; y = 0.1 10

0

1

3.36

9.92

a

This was done mainly to determine the amount of excess Li present in the compounds.

a slight loss is observed compared to the designed excess amount added. This is due to possible loss due to evaporation during sintering at high temperatures. Loss of Li due to hightemperature processing has been reported earlier, which is generally compensated by taking excess.14 Since lithiation of substituted compounds was the objective, no such additional steps were necessary in the present work. To determine the oxidation states of various species present in the compounds, Xray photoelectron spectroscopy was performed on typical compounds, x = y = 0; x = 0.15, y = 0; and x = 0.15, y = 0.1, and the results are shown in Figure 3a−c. A peak at 529 eV in Figure 3c corresponds to O1− s core level and indicates the presence of O2− in all the compounds.15 The 2P3/2 and 2P1/2 peaks at 779 and 794.2 eV in Figure 3a correspond to Co ions, while the peaks at 855.2 and 872 eV correspond to Ni ions.16 The binding energy difference of 15.2 eV for the Co ions is a signature of trivalent state of Co. The binding energy difference of 17.2 eV in Ni ions corresponds to its presence in the divalent 5830

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 3. X-ray photoelectron spectra from the three compounds LiCoO2, LiCo0.85Ni0.15O2, and Li1.1Co0.85Ni0.15O2 show peaks corresponding to (a) Co3+, (b) Ni2+, and (c) O2 in the compounds. These are the expected peaks as per the composition.

before increasing with increasing excess Li. The Li1.01Co0.85Ni0.15O2 compound has the lowest κ of ∼1 W m−1 k−1 in the entire temperature range from 300 to 1100 K. The electrical resistivity ρ of the compounds shows a complementary T dependence. Increasing Ni substitution increases the room-temperature resistivity by 2 orders of magnitude before decreasing for a substitution of 15 atom % Ni2+. This compound also exhibits the lowest ρ at 1100 K, Figure 5a. The high-temperature ρ of this compound is found to be ∼6 times lower than the unsubstituted compound. Increasing the Li content in this compound increases the electrical conductivity at all T with the x = 0.15, y = 0.1 compound exhibiting the highest electrical conductivity in the whole range, Figure 5b. All the compounds, independent of the extent of Ni substitution or lithiation, have a negative temperature coefficient of resistivity, a semiconducting thermally activated behavior. The nature of charge carriers as well as their concentration and mobility at room temperature were determined by measuring the Hall coefficient, and the results are given in Table 4. It can be seen that the charge carriers concentration increases by an order of magnitude for 15 atom % Ni substitution to 8.4 × 1016 from 8.8 × 1015 cm−3 and increases further to 1.12 × 1017 cm−3 upon lithiation. The charge carriers mobility, however, is found to decrease with the addition of Ni but remains constant for the excess Li compound. The convergence of high T resistivity of the

state. These results clearly show that Co and Ni are present in their expected oxidation states in all the different compounds without any deviation. The thermophysical properties of the two series of compounds LiCo1−xNixO2 and Li1+yCo0.85Ni0.15O2 were investigated in the temperature range of 300−1100 K to determine the thermoelectric efficiency of energy conversion. The specific heat capacity Cp of both the series of compounds determined using differential scanning calorimetry is shown in Figure 4a and is found to be in the range from 0.6 to 1.0 J g−1 K−1. It increases with T to ∼460 K before reaching saturation in almost all the compounds. The Debye temperature θD of stoichiometric LiCoO2 has been reported to be ∼448 K,17 and the present results agree with this θD indicating that substitution and lithiation do not significantly change the heat capacity behavior of the compounds. The total thermal conductivity κ determining by measuring the density as well as thermal diffusivity, however, is significantly different for the two sets of compounds, Figure 4b,c. The room-temperature thermal conductivity drops from ∼8 W m−1 K−1 for the unsubstituted compound to 2 W m−1 K−1 for 2% Ni substitution and then decreases further on increasing Ni substitution. The κ of all the substituted compounds shows very little T dependence and converges to ∼1 W m−1 K−1 at high T, indicating the phonondominated behavior. Addition of excess Li to the Ni-substituted compound results in lowering the thermal conductivity first 5831

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 4. Specific heat capacity Cp of all the compounds (a). The total thermal conductivity κ of (b) LiCo1−xNixO2 and (c) Li1+yCo0.85Ni0.15O2 compounds. The κ decreases drastically due to Ni2+ substitution and remains low on lithiation of the substituted compounds.

Figure 5. Electrical resistivity ρ variation with T up to 1100 K shows a semiconducting behavior in both sets of compounds, LiCo1−xNixO2 (a) and Li1+yCo0.85Ni0.15O2 (b). The absolute ρ drops both due to Co3+ substitution with Ni2+ as well as due to lithiation of the substituted compounds.

compound LiCoO2 at room temperature is ∼350 μV K−1 and increases steeply with temperature to ∼1100 μV K−1 at 375 K. This sharp increase around room temperature has also been observed earlier based on low-temperature thermopower measurements.6 The thermopower decreases beyond 375 K with increasing T to ∼600 μV K−1 at 1100 K. This behavior of α changes drastically due to the substitution of Co with Ni. The

different compounds clearly shows that charge carrier scattering by phonons is the limiting factor to achieve high electrical conductivity. The variation of thermopower or Seebeck coefficient α with T for the two sets of compounds is shown in Figure 6a,b. The thermopower is positive in all the compounds indicating that holes are predominant charge carriers at all temperatures. The thermopower of the parent 5832

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

imation together with the Boltzmann transport equation is used to estimate charge carriers mass, concentration, and band gap from α and ρ. This approximation is valid only for a limited number of materials, whose band structure is not complex. However, most of the materials including oxides have a band structure whose topography is more complex compared to the simple parabolic approximation, and hence this model cannot be applied.19 Another limitation of this model is that it does not take into account the minority carrier’s contribution to transport and hence is not applicable to predict the peak in thermopower due to bipolar transport effects.20 An added complication in the case of oxide systems, specifically cobalt oxides, is the strong correlation effects between charge, spin, orbital, and lattice. The spin state of Co3+ ions depends strongly on the overall chemical composition and also the temperature. Charge disproportionation into Co2+ and Co4+ as well as spinstate transitions within the t2g and eg d-levels of Co3+ are known to occur leading to changes in not only magnetic behavior but also in charge transport, conductivity, and thermopower.7b,21 An understanding of these variations quantitatively requires a detailed calculation of the electronic band structure, which is beyond the scope of the current work. Hence in the present work the variation of α and ρ with substitutions and T is qualitatively analyzed. LiCoO2 is a p-type conductor with the net conductivity being a function of the dynamic balance between Co3+ and Co4+ given by

Table 4. Various Transport Parameters, Charge Carriers Concentration nH and Mobility μH Determined by Measuring the Hall Coefficient of Representative Compounds Sl. No.

chemical composition

carriers type

Hall carrier concentration nH, cm−3

Hall mobility μH, cm2 V−1s−1

1 2 3 4

LiCoO2 LiCo0.92Ni0.08O2 LiCo0.85Ni0.15O2 Li1.10Co0.85Ni0.15O2

p p p p

8.75 × 1015 6.4 × 1015 8.39 × 1016 1.12 × 1017

0.0557 0.009 28 0.002 83 0.002 99

room-temperature α decreases to ∼300 μV K−1 and exhibits a peak around 600 K. The decrease in α at high T is due to bipolar conduction. The most dramatic reduction in α, however, is observed for the x = 0.15 compound, whose α value does not exceed 400 μV K−1 in the whole temperature range. Lithiation of this substituted compound results in further reduction of α over the complete temperature range with very little temperature dependence. These results are in complete agreement with electrical resistivity of the compounds that exhibit a drop in ρ with both Co substitution as well as lithiation. The changes in α and ρ with both substitution and lithiation is seen in the power factor α2σ and the thermoelectric figure-of-merit shown in Figure 7a−d. The power factor increases from ∼40 to 65 μW m−1 K−2 due to Ni substitution at 1100 K, which is further enhanced to 120 μW m−1 K−2 by the addition of excess Li to this compound. The thermoelectric figure-of-merit correspondingly is increased by an order of magnitude from ∼0.01 at 1100 K to 0.12 for the polycrystalline compounds. These results clearly show that they are comparable to polycrystalline NaCoO2,18 which has, however, a lower thermal and environmental stability compared to Li− Co-oxides.

Co3 + ↔ Co4 + + 1‐hole

(2)

Co4 + + 1‐electron ↔ Co3 +

(3)

Substitution of trivalent Co with divalent Ni increases the concentration of effective number of holes, and this changes the charge carrier transport properties. The substitution can be represented by the following charge balance scheme.



(Co3 +)Ni 2 + → Ni 2 + + 1‐hole

DISCUSSION The charge carrier transport parameters, thermopower α, and electrical resistivity ρ in general can be determined using the Boltzmann transport equation. These parameters depend on the electronic band structure of the materials, and for thermoelectric materials a simplified parabolic band approx-

(4)

The charge carrier concentration therefore is anticipated to increase with increasing Ni2+ substitution and result in simultaneous decrease of thermopower α and increase of conductivity σ, which have been observed in Figures 6a and 5a, respectively. The overall resistivity decreases by ∼2 orders of magnitude at room temperature, and the decrease is ∼5.6 times

Figure 6. Thermopower α variation with T shows that bipolar transport becomes effective due both to substitution as well as lithiation. The α of base compounds decreases with T, while that of (a) LiCo1−xNixO2 and (b) Li1+yCo0.85Ni0.15O2 compounds exhibit a maximum in the 550−800 K range. 5833

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 7. Thermoelectric power factor α2σ and figure-of-merit zT of the two sets of compounds (a−d). The power factor increases by ∼200%, while the figure-of-merit increases from 0.015 to 0.12 at 1100 K, a giant increase of 760% compared to unsubstituted, stoichiometric compound LiCoO2.

at 1100 K for x = 0.15. Note that the room-temperature resistivity increases first with increasing Ni substitution until 8 atom %, after which it drops. These results are in agreement with the carrier concentrations and mobility in these compounds given in Table 4. The carriers concentration remains nearly constant, while their mobility drops from 6 × 10−2 to 1 × 10−2 cm2V−1s−1 for 8 atom % substitution. Although the effective mobility drops to 3 × 10−3 cm2 V−1 s−1 on increasing substitution to 15 atom %, an order of magnitude increase in carrier concentration results in increasing the electrical conductivity. A further increase in room-temperature electrical conductivity on lithiation is primarily due to an increase in carrier concentration. The thermopower exhibits a bipolar conduction behavior with a peak whose magnitude decreases and shifts to higher temperatures with increasing Ni substitution and lithiation.22 These results show that the effective band gap for minority carriers increases with substitution and that the compounds become monopolar in terms of charge transport. The Ni ions in stoichiometric LiNiO2 should exist in Ni3+/Ni4+ electronic configurations, but since it is highly unstable, the compound becomes nonstoichiometric with the possibility of Ni existing in Ni3+/Ni2+ electronic configurations. Added to this, the splitting of Co and Ni d-band into t2g and eg levels will result in a hierarchy of bands at the Fermi level and lead to a complex band structure formation in the substituted compounds.23 The effective band gap for majority carriers in the substituted compounds reduces

compared to the unsubstituted compound with the Fermi level becoming closer to the valence band maximum. These band structure changes lead to changes in the transport behavior. The high-temperature charge transport, however, is completely dominated by phonon interactions as well as Li ions migration. The room-temperature thermal conductivity exhibits a sharp drop due to substitution of Co3+ with Ni2+. The hightemperature behavior of all the compounds however is nearly identical, indicating that phonon scattering dominates at these high temperatures. The drop in room-temperature κ however is possibly due to a combination of reduced charge carrier mobility and enhanced point defects formation due to substitution of Co with Ni. The total thermal conductivity κ is due to a combination of heat transported by the charge carriers and that by the phonons. The charge carriers concentration has been found to increase with Ni substitution, but their mobility is found to decrease. The decreased carriers mobility together with scattering of phonons by the point defects could result in the observed thermal conductivity drop with Ni substitution. Lithiation of the substituted compounds however results only in a slight increase in the overall k showing that phonon scattering is still the dominant scattering mechanism. The charge carriers contribution can be estimated using the Weidmann-Franz law assuming a lower value of 1.5 × 10−8 W Ω K−2 for the Lorentz number. Although the Weidmann−Franz law is not strictly valid for highly correlated cobalt oxides with relatively low charge carriers concentration, 5834

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 8. Charge carrier contribution to total thermal conductivity κe determined using the Weidmann−Franz Law is shown for the different compounds in (a, b). The lattice or phonon contribution κl thus determined shows that it is dominant only after ∼600 K in all the compounds and is shown in comparison to total thermal conductivity in (c, d).

substitution for Co (with 58.93 g mol−1 atomic mass) will not change the phonon dispersion and also grain boundary scattering will not contribute with 4−5 μm grains. The umklapp scattering process is defined by the Gruneisen parameter γ given by the relation

it still can be applied to determine the relative magnitude of contribution of charge carrier vis-à-vis phonons to κ. It is found that the charge carrier contribution to κ is at least 2 orders of magnitude lower at the highest T, while it is 3 orders of magnitude lower at room temperature as shown in Figure 8a,b. As a result, the lattice contribution to thermal conductivity becomes significant at all temperatures as shown in Figure 8c,d, and hence it can be concluded that phonons are the main heat transport medium. This result clearly demonstrates that in these compounds the charge and heat transport are decoupled. The heat transport is limited by the various phonon scattering mechanisms such as acoustic umklapp phonon−phonon scattering, alloy−phonon scattering, structural disorder− phonon scattering, and magnetic scattering. The low value of κ indicates that all these scattering phenomena are present in the substituted cobalt oxides. The heat transported by acoustic phonons assuming a linear dispersion and a relaxation-time approximation is given by24 kl =

3 kB ⎛ kBT ⎞ ⎜ ⎟ 2π 2 ϑ ⎝ h ⎠

∫0

θD/ T

τC

x 4e x dx (e x − 1)2

−1

γ = (3βBmVm/Cv)

(6)

where β, Bm, Vm, and Cv are thermal expansion coefficient, isothermal bulk modulus, molar volume, and specific heat at constant volume, respectively. Using the different elastic properties of pure LiCoO225 the Gruneisen parameter γ at room temperature can be determined, and it is found to be 3.2, which is high. This parameter is known to increase with increasing temperature and lead to a reduction in the thermal conductivity of the material. Another parameter that can play an important role in reducing the thermal conductivity in these materials is the layered structure, similar to that observed in chalcogenide alloys.26 Additionally, on the one hand, the relatively low density of the pellets results in lowering not only the total thermal conductivity but also the electrical conductivity. Spark plasma sintering to obtain high density, on the other hand, can lead to both higher thermal and electrical conductivites. Since these factors have opposite effects on the figure-of-merit, the net change in zT will be negligible. Unlike in the case of alloys, porosity in these oxides cannot lead to materials degradation.

(5)

where x = hω/kBT is the reduced energy with ω the phonon frequency, kB the Boltzmann constant, h the modified Planck’s constant, ϑ the velocity of sound, and τC the total phononscattering time given by τC−1 = ∑iτi−1 i the different scattering processes. Among the different scattering processes mentioned above umklapp process will be the most relevant, as alloy scattering due to Ni (with 58.69 g mol−1 atomic mass) 5835

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry

Figure 9. Stability of Li1+yCo1−xNixO2 compounds tested as a function of time. (a) Variation of thermal diffusivity of Ni-substituted compound prepared in two separate batches together with repeating the measurements. (b) Variation of power factor of Li1.1Co0.85Ni0.15O2 compound studied after six months and compared with data of freshly prepared compound.



ENVIRONMENTAL STABILITY A good thermoelectric material should have good environmental and thermal stability apart from having a high figure-ofmerit. This is an extremely important criterion for all classes of thermoelectric materials including oxides, which can exhibit sensitivity to atmospheric moisture content. Hence in the present work the environmental stability and reproducibility of thermophysical properties of Li1+yCo1−xNixO2 compounds was investigated after a time period of six months in ambient conditions. The thermal diffusivity of Ni-substituted compound that exhibits lowest electrical resistivity was synthesized in two separate, independent batches. The thermal diffusivity of batch I compound was determined immediately after synthesis. The second batch of compound was stored under normal conditions for a period of six months, and then its thermal diffusivity was studied for two cycles of heating and cooling, Figure 9a. The variation of thermal diffusivity is found to be within the reproducibility limit of LFA-457 system, both for the two batches as well as for two separate measurement runs. The electrical resistivity and thermopower of Li1.1Co0.85Ni0.15O2 compound, which exhibits the maximum figure-of-merit, were investigated after a period of six months, and the resulting power factor variation with T is shown in Figure 9b. It is clearly seen that the power factor variations are again within the reproducibility limit over the complete T range of 300−1100 K. These results clearly show that the Li1+yCo1−xNixO2 compounds are extremely stable and reliable for device fabrication and use.

conductivity and hence a giant increase in the high-temperature figure-of-merit. The thermopower as well as total thermal conductivity of the Li1.1Co0.85Ni0.15O2 is found to be optimal, but the electrical conductivity, however, needs to be increased further so as to realize high power factors and hence figure-ofmerit. Although the maximum power factor obtained in the present series of compounds, 120 W m−1 K−2, is high for an oxide thermoelectric material, for application in devices this needs to increase by ∼2 orders of magnitude. A possible methodology could be to reduce Li to below stoichiometric limit and thus enhance charge carriers concentration as well as mobility. The LiCoO2 compound is known to undergo a semiconductor-to-metal transition on decreasing the Li content to below stoichiometric values.



ASSOCIATED CONTENT

Accession Codes

CCDC 1546683−1546690 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing [email protected], or by contacting The Cambridge Crystallographic Data Centre, 12, Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID



Satish Vitta: 0000-0003-4138-0022

CONCLUSIONS LiCoO2, a semiconducting electrode material used in solid-state batteries, has been modified both by substitution and lithiation to make it suitable for thermoelectric application at high temperatures. Substitution of nominally trivalent Co with divalent Ni increases the net charge carriers concentration and results in increasing the electrical conductivity while reducing the thermopower. This increase in charge carriers concentration, however, does not increase the total thermal conductivity of the compounds indicating that charge transport and heat transport are effectively decoupled. The thermal conductivity therefore is predominantly regulated by phonon transport. Lithiation of the substituted compounds leads to a further decrease in electrical resistivity as well as total thermal

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors wish to acknowledge Nano mission, Govt. of India, and Indian Space Research Organization for financial assistance and the Indian Institute of Technology Bombay Central facilities for the provision of structural characterization facilities.



REFERENCES

(1) (a) Riffat, S. B.; Ma, X. Thermoelectrics: a review of present and potential applications. Appl. Therm. Eng. 2003, 23, 913−935. (b) Zheng, X. F.; Liu, C. X.; Yan, Y. Y.; Wang, Q. A review of thermoelectrics research − Recent developments and potentials for

5836

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry sustainable and renewable energy applications. Renewable Sustainable Energy Rev. 2014, 32, 486−503. (c) He, W.; Zhang, G.; Zhang, X.; Ji, J.; Li, G.; Zhao, X. Recent development and application of thermoelectric generator and cooler. Appl. Energy 2015, 143, 1−25. (2) Snyder, G. J.; Toberer, E. S. Complex thermoelectric materials. Nat. Mater. 2008, 7, 105−114. (3) (a) Chung, D. Y.; Lordanidis, L.; Choi, K. S.; Kanatzidis, M. G. Complex chalcogenides as thermoelectric materials: A solid state chemistry approach. Bull. Korean Chem. Soc. 1998, 19, 1281−1293. (b) Fardy, M.; Hochbaum, A. I.; Goldberger, J.; Zhang, M. M.; Yang, P. Synthesis and Thermoelectrical Characterization of Lead Chalcogenide Nanowires. Adv. Mater. 2007, 19, 3047−3051. (c) Yan, X.; Joshi, G.; Liu, W.; Lan, Y.; Wang, H.; Lee, S.; Simonson, J. W.; Poon, S. J.; Tritt, T. M.; Chen, G.; Ren, Z. F. Enhanced thermoelectric figure of merit of p-type half-Heuslers. Nano Lett. 2011, 11, 556−60. (d) Chen, S.; Ren, Z. Recent progress of halfHeusler for moderate temperature thermoelectric applications. Mater. Today 2013, 16, 387−395. (e) Downie, R. A.; MacLaren, D. A.; Smith, R. I.; Bos, J. W. Enhanced thermoelectric performance in TiNiSnbased half-Heuslers. Chem. Commun. 2013, 49, 4184−6. (f) Fu, C.; Bai, S.; Liu, Y.; Tang, Y.; Chen, L.; Zhao, X.; Zhu, T. Realizing high figure of merit in heavy-band p-type half-Heusler thermoelectric materials. Nat. Commun. 2015, 6, 8144. (g) Mudryk, Y.; Rogl, P.; Paul, C.; Berger, S.; Bauer, E.; Hilscher, G.; Godart, C.; Noël, H. Thermoelectricity of clathrate I Si and Ge phases. J. Phys.: Condens. Matter 2002, 14, 7991. (h) Christensen, M.; Johnsen, S.; Iversen, B. B. Thermoelectric clathrates of type I. Dalton Trans. 2010, 39, 978−992. (4) (a) Wang, Y.; Sui, Y.; Su, W. High temperature thermoelectric characteristics of Ca0.9 R0.1 MnO3 (R = La,Pr,Yb). J. Appl. Phys. 2008, 104, 093703. (b) Lee, K. H.; Kim, S. W.; Ohta, H.; Koumoto, K. Thermoelectric properties of layered perovskite-type (Sr1‑x Cax) 3 (Ti1‑y Nby) 2 O7. J. Appl. Phys. 2007, 101, 083707. (c) Reimann, T.; Töpfer, J. Thermoelectric properties of Gd/W double substituted calcium Manganite. J. Alloys Compd. 2017, 699, 788−795. (d) Kabir, R.; Zhang, T.; Wang, D.; Donelson, R.; Tian, R.; Tan, T. T.; Li, S. Improvement in the thermoelectric properties of CaMnO3 perovskites by W doping. J. Mater. Sci. 2014, 49, 7522−7528. (5) (a) Terasaki, I.; Sasago, Y.; Uchinokura, K. Large thermoelectric power in NaCo2O2 single crystals. Phys. Rev. B: Condens. Matter Mater. Phys. 1997, 56, R12685−R12687. (b) Miyazaki, Y.; Kudo, K.; Akoshima, M.; Ono, Y.; Koike, Y.; Kajitani, T. Low-temperature thermoelectric properties of the composite crystal [Ca2CoO3.34]0.614[CoO2]. Jpn. J. Appl. Phys. 2000, 39, L531−L533. (c) Fujita, K.; Mochida, T.; Nakamura, K. High-Temperature Thermoelectric Properties of Na x CoO 2‑δ Single Crystals. Jpn. J. Appl. Phys. 2001, 40, 4644. (d) Carvillo, P.; Chen, Y.; Boyle, C.; Barnes, P. N.; Song, X. Thermoelectric Performance Enhancement of Calcium Cobaltite through Barium Grain Boundary Segregation. Inorg. Chem. 2015, 54, 9027−9032. (e) Nam, S. W.; Choi, J. W.; Hwang, H. K.; Park, K. Improvement in high-temperature thermoelectric properties by adding Mn for Co in Ca3Co4O9. J. Nanosci. Nanotechnol. 2010, 10, 7689−93. (f) Park, K.; Choi, J. W. High-temperature thermoelectric properties of Na(Co0.91, Ni0.09)2O4 fabricated by solution combustion method for power generation. J. Nanosci. Nanotechnol. 2012, 12, 3624−8. (g) Koshibae, W.; Tsutsui, K.; Maekawa, S. Thermopower in cobalt oxides. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 62, 6869−6872. (h) Liu, P.; Chen, G.; Cui, Y.; Zhang, H.; Xiao, F.; Wang, L.; Nakano, H. High temperature electrical conductivity and thermoelectric power of NaxCoO2. Solid State Ionics 2008, 179, 2308−2312. (i) Peleckis, G.; Motohashi, T.; Karppinen, M.; Yamauchi, H. Enhanced thermoelectric properties of NaxCoO2 whisker crystals. Appl. Phys. Lett. 2003, 83, 5416−5418. (j) Xu, L.; Li, F.; Wang, Y. High-temperature transport and thermoelectric properties of Ca3Co4−xTixO9. J. Alloys Compd. 2010, 501, 115−119. (k) Tajima, S.; Tani, T.; Isobe, S.; Koumoto, K. Thermoelectric properties of highly textured NaCo2O4 ceramics processed by the reactive templated grain growth (RTGG) method. Mater. Sci. Eng., B 2001, 86, 20−25.

(6) Levasseur, S.; Ménétrier, M.; Delmas, C. On the Dual Effect of Mg Doping in LiCoO2 and Li1+δCoO2: Structural, Electronic Properties, and 7Li MAS NMR Studies. Chem. Mater. 2002, 14, 3584−3590. (7) (a) Menetrier, M.; Saadoune, I.; Levasseur, S.; Delmas, C. The insulator-metal transition upon lithium deintercalation from LiCoO2: electronic properties and 7Li NMR study. J. Mater. Chem. 1999, 9, 1135−1140. (b) Sugiyama, J.; Nozaki, H.; Brewer, J. H.; Ansaldo, E. J.; Morris, G. D.; Delmas, C. Frustrated magnetism in the twodimensional triangular lattice of LixCoO2. Phys. Rev. B: Condens. Matter Mater. Phys. 2005, 72, 144424. (8) (a) Chen, J. Recent Progress in Advanced Materials for Lithium Ion Batteries. Materials 2013, 6, 156. (b) Nitta, N.; Wu, F.; Lee, J. T.; Yushin, G. Li-ion battery materials: present and future. Mater. Today 2015, 18, 252−264. (c) Oz, E.; Altin, S.; Demirel, S.; Bayri, A.; Altin, E.; Baglayan, O.; Avci, S. Electrochemical effects and magnetic properties of B substituted LiCoO2: Improving Li-battery performance. J. Alloys Compd. 2016, 657, 835−847. (9) Motohashi, T.; Sugimoto, Y.; Masubuchi, Y.; Sasagawa, T.; Koshibae, W.; Tohyama, T.; Yamauchi, H.; Kikkawa, S. Impact of lithium composition on the thermoelectric properties of the layered cobalt oxide system LixCoO2. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 83, 195128. (10) (a) Kuroki, K.; Arita, R. Pudding Mold” Band Drives Large Thermopower in NaxCoO2. J. Phys. Soc. Jpn. 2007, 76, 083707. (b) Singh, D. J. Electronic structure of NaCo2O4. Phys. Rev. B: Condens. Matter Mater. Phys. 2000, 61, 13397−13402. (c) Tanaka, T.; Nakamura, S.; Iida, S. Observation of Distinct Metallic Conductivity in NaCo2O4. Jpn. J. Appl. Phys. 1994, 33, L581. (11) (a) Schneider, C.; Schichtel, P.; Mogwitz, B.; Straubinger, R.; Beyer, A.; Rohnke, M.; Volz, K.; Janek, J. Thermomigration and Soret effect in NaxCoO2 as thermoelectric material: Preparation and characterization of sodium cobaltate thin films. Phys. Status Solidi A 2016, 213 (a), 1284−1295. (b) Namhongsa, W.; Singsoog, K.; Paengson, S.; Pilasuta, P.; Seetawan, T.; Muthitamongkol, P.; Thanachayanont, C. Effects of Magnetic Field on Synthesis and Thermoelectric Properties of NaCoO2. Integr. Ferroelectr. 2015, 165, 81−85. (c) Mendels, P.; Bono, D.; Bobroff, J.; Lang, G.; Collin, G.; Colson, D.; Blanchard, N.; Alloul, H.; Mukhamedshin, I.; Bert, F.; Amato, A.; Hillier, A. D. Series of bulk magnetic-phase transitions in: A study. Phys. B 2006, 374−375, 278−281. (12) (a) Ghosh, P.; Mahanty, S.; Raja, M. W.; Basu, R. N.; Maiti, H. S. Structure and optical absorption of combustion-synthesized nanocrystalline LiCoO2. J. Mater. Res. 2007, 22, 1162−1167. (b) Kushida, K.; Kuriyama, K. Narrowing of the Co-3d band related to the order−disorder phase transition in LiCoO2. Solid State Commun. 2002, 123, 349−352. (c) Rosolen, J. M.; Decker, F. Photoelectrochemical behavior of LiCoO 2 membrane electrode. J. Electroanal. Chem. 2001, 501, 253−259. (13) Shao-Horn, Y.; Croguennec, L.; Delmas, C.; Nelson, E. C.; O’Keefe, M. A. Atomic resolution of lithium ions in LiCoO2. Nat. Mater. 2003, 2, 464−467. (14) (a) Sathiya, M.; Prakash, A.; Ramesha, K.; Shukla, A. NitrateMelt Synthesized HT-LiCoO2 as a Superior Cathode-Material for Lithium-Ion Batteries. Materials 2009, 2, 857. (b) Antolini, E.; Ferretti, M. Synthesis and Thermal Stability of LiCoO2. J. Solid State Chem. 1995, 117, 1−7. (15) Majumdar, D.; Chatterjee, D. X-ray photoelectron spectroscopic studies on yttria, zirconia, and yttria-stabilized zirconia. J. Appl. Phys. 1991, 70, 988−992. (16) (a) McIntyre, N. S.; Cook, M. G. X-ray photoelectron studies on some oxides and hydroxides of cobalt, nickel, and copper. Anal. Chem. 1975, 47, 2208−2213. (b) Mansour, A. N. Characterization of NiO by XPS. Surf. Sci. Spectra 1994, 3, 231−238. (c) Khawaja, E. E.; Salim, M. A.; Khan, M. A.; Al-Adel, F. F.; Khattak, G. D.; Hussain, Z. XPS, auger, electrical and optical studies of vanadium phosphate glasses doped with nickel oxide. J. Non-Cryst. Solids 1989, 110, 33−43. 5837

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838

Article

Inorganic Chemistry (17) Kawaji, H.; Takematsu, M.; Tojo, T.; Atake, T.; Hirano, A.; Kanno, R. Low temperature heat capacity and thermodynamic functions of LiCoO2. J. Therm. Anal. Calorim. 2002, 68, 833−839. (18) Wang, L.; Wang, M.; Zhao, D. Thermoelectric properties of caxis oriented Ni-substituted NaCoO2 thermoelectric oxide by the citric acid complex method. J. Alloys Compd. 2009, 471, 519−523. (19) Xing, G.; Sun, J.; Ong, K. P.; Fan, X.; Zheng, W.; Singh, D. J. Perspective: n-type oxide thermoelectrics via visual search strategies. APL Mater. 2016, 4, 053201. (20) (a) Gibbs, Z. M.; Kim, H.-S.; Wang, H.; Snyder, G. J. Band gap estimation from temperature dependent Seebeck measurement Deviations from the 2e|S|maxTmax relation. Appl. Phys. Lett. 2015, 106, 022112. (b) Wang, S.; Yang, J.; Toll, T.; Yang, J.; Zhang, W.; Tang, X. Conductivity-limiting bipolar thermal conductivity in semiconductors. Sci. Rep. 2015, 5, 10136. (21) Mukai, K.; Aoki, Y.; Andreica, D.; Amato, A.; Watanabe, I.; Giblin, S. R.; Sugiyama, J. Thermally activated spin fluctuations in stoichiometric LiCoO2 clarified by electron paramagnetic resonance and muon-spin rotation and relaxation measurements. Phys. Rev. B: Condens. Matter Mater. Phys. 2014, 89, 094406. (22) Wang, S.; Sun, Y.; Yang, J.; Duan, B.; Wu, L.; Zhang, W.; Yang, J. High thermoelectric performance in Te-free (Bi,Sb)2Se3 via structural transition induced band convergence and chemical bond softening. Energy Environ. Sci. 2016, 9, 3436−3447. (23) (a) Julien, C.; Mauger, A.; Zaghib, K.; Groult, H. Comparative Issues of Cathode Materials for Li-Ion Batteries. Inorganics 2014, 2, 132. (b) Manthiram, A.; Vadivel Murugan, A.; Sarkar, A.; Muraliganth, T. Nanostructured electrode materials for electrochemical energy storage and conversion. Energy Environ. Sci. 2008, 1, 621−638. (24) Morelli, D. T.; Slack, G. A. High Lattice Thermal Conductivity Solids. In High Thermal Conductivity Materials; Springer: New York, NY, 2006; pp 37−68. (25) Cheng, E. J.; Taylor, N. J.; Wolfenstine, J.; Sakamoto, J. Elastic properties of lithium cobalt oxide (LiCoO2). J. Asian Ceram. Soc. 2017, DOI: 10.1016/j.jascer.2017.03.001. (26) Zhao, L.-D.; Lo, S.-H.; Zhang, Y.; Sun, H.; Tan, G.; Uher, C.; Wolverton, C.; Dravid, V. P.; Kanatzidis, M. G. Ultralow thermal conductivity and high thermoelectric figure of merit in SnSe crystals. Nature 2014, 508, 373−377.

5838

DOI: 10.1021/acs.inorgchem.7b00476 Inorg. Chem. 2017, 56, 5827−5838