Grain Boundary Phase Segregation for Dramatic Improvement of the

Subscriber access provided by University of Sunderland

Functional Inorganic Materials and Devices

Grain Boundary Phase Segregation for Dramatic Improvement of the Thermoelectric Performance of Oxide Ceramics Xueyan Song, Sergio Paredes Navia, Liang Liang, Cullen Boyle, Cesar-Octavio RomoDe-La-Cruz, Bryan Jackson, Alec Hinerman, Megan Wilt, Jacky Prucz, and Yun Chen ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b12710 • Publication Date (Web): 24 Oct 2018 Downloaded from http://pubs.acs.org on October 24, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Grain Boundary Phase Segregation for Dramatic Improvement of the Thermoelectric Performance of Oxide Ceramics Xueyan Song,* Sergio A. Paredes Navia, Liang Liang, Cullen Boyle, Cesar-Octavio Romo-De-LaCruz, Bryan Jackson, Alec Hinerman, Megan Wilt, Jacky Prucz, Yun Chen Department of Mechanical & Aerospace Engineering, West Virginia University, Morgantown, WV 26506, USA KEYWORDS. Thermoelectric, Ceramics, Oxide, Grain Boundaries, Segregation.

ABSTRACT: This work presents a novel approach of dramatically increasing the energy conversion efficiency of thermoelectric CaMnO3- ceramics through the combination of lattice dopants substitution and secondary phase segregation at the grain boundaries. The oxide ceramic samples are with the nominal composition of Ca1-xBixMnCuyO3- (x=0, 0.02, 0.03; y=0.02, 0.04). When Cu is introduced into the Ca1-xBixMnCuyO3- samples, the grain growth from Bi-doped CaMnO3- grains is accompanied by the limited solubility of Cu ions in the grain interior, while Cu mainly formed a CuO secondary phase at the grain boundaries. Cu nonstoichiometry addition subsequently resulted in the increase of the Seebeck coefficient and decreases of electrical resistivity simultaneously. The sample with designed chemistry of Ca2.97Bi0.03MnCu0.04O3- exhibits the power factor of 2.4 mW m-1K-2 at 337 K and Figure of Merit ZT of 0.67 at 773 K. This ZT of 0.67 is by far the highest ZT reported for various perovskites oxide ceramics. Such enhancements in electrical power factor and the overall ZT are attributed to the synergistic effect of decreasing the carrier concentration to increase the Seebeck coefficient and simultaneously increasing the carrier mobility through the existence of CuO phase at the grain boundaries.

I. INTRODUCTION To promote the development of high performance thermoelectric devices, in the past decades, intense effort on thermoelectric research has been concentrated on designing new materials with high energy conversion efficiency.1-7 Transition metal thermoelectric oxides, such as cation doped CaMnO3-, are of significant interest in various energy applications because those are earth-abundant materials, nontoxic, environmental friendly, and having high thermal stability in air. However, the transition metal oxides have been generally regarded as poor thermoelectric materials because of their low energy conversion efficiency. The thermoelectric conversion efficiency is characterized by the figure of merit ZT = S2σ/κ, where S, σ, S2σ, and κ are the Seebeck coefficient, the electrical conductivity, power factor, and the thermal conductivity, respectively. A good thermoelectric material should have large |S| and high σ together with low κ. The state-of-the-art heavy metal based thermoelectric materials typically have a ZT >1, that is corresponding to ~10% energy conversion efficiency. So far, CaMnO3- has highest reported ZT value of about 0.32. 8 The oxide CaMnO3- polycrystalline possess very low thermal conductivity, and the low energy conversion efficiency results from the low power factor of S2σ. So far, the highest reported electrical power factor 9 is ~0.4 mWm-1K-2, and it is much lower in comparison with the power factor of ~2 mW m-1K-2 from that of the well-developed thermoelectric materials such as SiGe. 10,11 To improve the ZT value of CaMnO , there is an urgent 3- need of improving the electrical power factor. However, increasing the power factor of CaMnO3- has been very challenging, since the Seebeck coefficient and the electrical

conductivity are both as a function of carrier concentration and strongly correlated. The electrical conductivity increases as the carrier concentration and carrier mobility increase. Meanwhile, the Seebeck coefficient decreases as the carrier concentration increases. As a consequence, conventional doping approaches for increasing the carrier concentration generally results in the simultaneous increase of electrical conductivity and a decrease of Seebeck coefficient, and eventually leads to very limited increase in the overall power factor. In the present work described herein, significant increase in power factor in CaMnO3- were achieved through a synergistic approach of dopant substitution in the lattice and secondary phase of CuO segregation at the grain boundaries of polycrystalline ceramics. The Bi substitution of Ca in the CaMnO3- lattice results in the concurrent decrease of both the Seebeck coefficient and the electrical resistivity, producing an overall moderate increase in power factor. When the CuO phase is introduced to grain boundaries of Bi-doped CaMnO3- ceramics, it promotes the thermoelectric parameters (S, σ) decoupling, and leads to the dramatic decrease of the electrical resistivity and simultaneous increase of Seebeck coefficient. The overall energy conversion efficiency ZT for the best performed sample with the chemistry of Ca1-xBixMnCuyO3- is 0.67 at 773 K.

II. MATERIALS AND METHODS The precursor powders with the nominal formulation composition of Ca1-xBixMnCuyO3- (x=0, 0.02, 0.03; y=0.02, 0.04) are prepared by a sol-gel chemical solution route. Citric

ACS Paragon Plus Environment


cracking during the high temperature sintering. The pellets are easily broken into pieces even without any machining or further thermal cycle treatment at high temperatures. When the Cu is introduced to the Bi-doped CaMnO3- pellets, the pellets exhibit superior toughness and remain intact after machining and many cycles of thermal treatment from room temperature to 1073 K in various gas atmospheres. The high density of the pellets and the superior mechanical toughness make the ceramics suitable for practical device applications. Significant increase of electrical power factor and energy conversion efficiency Figures 1(a and b) show the temperature dependence of the electrical resistivity of samples with different chemistry. In comparison with the pristine CaMnO3-, the Bi substitution of Ca results in the decrease of the electrical resistivity. a

1000

100

10 400

b

Enhancement of mechanical integrity and toughness of bulk oxide ceramics with designed chemistry The sintered pellets are with diameter of ~13 mm in diameter and ~2 mm in thickness. The sample apparent density is calculated using pellet mass divided by volume. Table 1 Apparent densities of the samples Ca1xBixMnCuyO3- (x=0, 0.02, 0.03; y=0, 0.02, 0.04)

CaMnO3-

3.64

Ca0.98Bi0.02MnO3-

4.25

Ca0.97Bi0.03MnO3-

4.37

Ca0.98Bi0.02MnO3-Cu0.02

4.51

Ca0.97Bi0.03MnO3-Cu0.04

4.53

Table 1 shows the apparent densities of the oxide ceramic pellets. Compared with the baseline pure CaMnO3-, the density of the Bi-doped pellets with chemistry of Ca1-xBixMnO3- increases with increases of Bi doping level. The density of the pellets further increases as the Cu is further introduced in the samples with the chemistry of Ca1-xBixMnCuyO3-. For the thermoelectric device application, the bulk oxide ceramics should have high density to ensure the mechanical strength during device fabrication machining and operation. The samples with Cu non-stoichiometric addition have increased density, and most importantly, the overall mechanical toughness. The ceramic CaMnO3- pellets usually suffer from

600

700

800

900

1000

1100

1000

1100

CMO Baseline CMO+Bi 2% Sub CMO+Bi 3% Sub CMO+Bi 2% Sub+Cu 2% CMO+Bi 3% Sub+Cu 4%

240 200

Density (g/cm3)

500

Temperature (K)

III. RESULTS AND DISCUSSION

Chemical Formula


10000

Resistivity ( m)

acid (BDH Chemical), ethylene glycol (BDH Chemical), polyethylene glycol, nitric acid (68%-70% concentration), and nominal amounts of Ca(NO3)2·4H2O (99%, Acros Organics), Bi(NO3)3·5H2O (99%, Acros Organics), Mn(NO3)2·6H2O (99%, Acros Organics), and Cu(NO3)2 (99.5%, Strem Chemical) are used as starting raw materials. The above listed raw materials with different mass ratio to achieve the designed chemistry were then dissolved in deionized water with stir at 353 K for 3 hours to form the gel. The gel is ashed at 773 K in a box furnace. The ashed product is ground using mechanical ball-milling, and then calcined at 1173 K in a tube furnace with oxygen flow for 2 hours to form the precursor powders. The powders are then uniaxially pressed into pellets under 0.75 GPa at 423 K for 2 minutes. The pellets are sintered at 1373 K in a tube furnace with oxygen flow for 2 hours to obtain the bulk samples. The absolute Seebeck coefficient S and electrical resistivity  were measured from 320 K up to 1080 K using a Linseis LSR1100 in a He environment. Thermal conductivity was measured from 323 to 973 K using a Laser Flash Analyzer Linseis-1200. The thermal conductivity is calculated as ktotal = λCpρ, in which Cp is specific heat, λ is the thermal diffusivity, ρ is the density. X-ray diffraction (XRD) analysis was carried out using a PANalytical X’Pert Pro X-ray Diffractometer for phase identification. A JEOL JSM 7600F Scanning electron microscope (SEM), and a JEM-2100 transmission electron microscope (TEM) equipped with energy dispersive X-ray spectroscopy (EDS) were used to examine the structure and chemistry from micron to atomic scale.

Resistivity (µ m)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 12

160 120 80 40

400

500

600

700

800

900

Temperature (K)

Figure 1 a, Temperature dependence of electrical resistivity of Ca1-xBixMnCuyO3- samples. b, Enlarged portion of the lower resistivity regime in Figure 1a.

The electrical resistivity from the pristine CaMnO3- shows semiconductor behavior and decrease as the temperature increase (Figure 1a). On the other hand, the samples with Bi substitution are with metallic behavior and their resistivity increases as the measurement temperature increases to 1083 K (Figure 1b). The electrical resistivity of the samples Ca1xBixMnO3- decreases as the Bi concentration increases from x = 0.02 to x=0.03. When Cu is further added to sample of Ca0.98Bi0.02MnO3-, the sample Ca0.98Bi0.02MnCu0.02O3- exhibits the same metallic behavior with resistivity increasing as the measurement temperature increases. But, the absolute value of the electrical resistivity is decreased than that of the sample Ca0.98Bi0.02MnO3-, from the entire measurement temperature regime as shown in Figure 1b. Likewise, the resistivity of sample Ca0.97Bi0.03MnCu0.04O3- is much lowered by about a factor of 3 than that of the sample Ca0.97Bi0.03MnO3-. Remarkably, the electrical resistivity of sample Ca0.97Bi0.03MnCu0.04O3- reached value of ~17 µ m at 377 K, which is the lowest electrical resistivity ever reported for CaMnO3-.


Figures 2 (a and b) show the temperature dependence of the Seebeck coefficient. All samples are n-type semiconductor with negative Seebeck coefficient values. S of the samples Ca1xBixMnO3- decreases as the Bi concentration increases, indicating the increased carrier concentration induced by Bi substitution of Ca. By contrast, when Cu non-stoichiometric addition is applied to the Bi doped samples, the Ca1xBixMnCuyO3- samples exhibit a slightly increased absolute Seebeck coefficient in comparing with the corresponding parent sample without Cu addition (in Figure 2b). a

CMO Baseline CMO+Bi 3% Sub CMO+Bi 3% Sub+Cu 4%

Absolute Seebeck (µV K-1)

-100

which is with the highest power factor, is higher than that of the baseline and slightly lower than that of the Bi-doped CaMnO3- samples.

2.0

1.6

1.2 400

500

600

700

800

900

1000

1100

Temperature (K)

Figure 4 Temperature dependence of thermal conductivity of Ca1xBixMnCuyO3- samples.

-300

-400 0.7 -500

0.6 0.5 500

600

700

800

900

1000

1100

ZT 0.3


-160 -180


0.4

Temperature (K)


2.4

-200

400

b


2.8

CMO+Bi 2% Sub CMO+Bi 2% Sub+Cu 2%

-600

0.2 0.1

-200 0.0 400

-220

500

600

700

800

900

1000

1100

Temperature (K) -240

Figure 5 Temperature dependence of ZT of Ca1-xBixMnCuyO3- samples.

-260 -280

Figure 2 a, Temperature dependence of Seebeck coefficient of Ca1-xBixMnCuyO3- samples. And b, Enlarged portion of the lower Seebeck coeficient regime in Figure 2a.

Benefitting from the low electrical resistivity and increased Seebeck coefficient, the sample of Ca0.97Bi0.03MnCu0.04O3- exhibits a high power factor of 2.4 mW m-1K-2, which is about factor of 5 increase compared to that from the sample Ca0.97Bi0.03MnO3-, as shown in Figure 3. The power factor of 2.4 mW m-1K-2 from Ca0.97Bi0.03MnCu0.04O3- sample is the highest power factor so far reported in the literatures for thermoelectric CaMnO3- synthesized using different methods. 3,9,12-17 This value of 2.4 mW m-1K-2 is also comparable with that from the state-of-the art SiGe and Bi-Te thermoelectric semiconductors. 10,11 2.4


2.0

Overall, the sample of Ca0.97Bi0.03MnCu0.04O3- has the maximum ZT value of 0.67 at 773 K (Figure 5) which is about factor of 2 higher than that highest ZT value of up to 0.32 reported in the literatures. 8,9,14,16,18 The thermoelectric oxide from the sample Ca0.97Bi0.03MnCu0.04O3- also exhibits a high plateau of the ZT from room temperature to 1073 K. Furthermore, from room temperature to 773 K, the ZT of the ntype bulk oxide is higher than that from the state-of-the art thermoelectric SiGe bulk alloys. 19 Microstructure changes induced by chemistry modification and existence of CuO phase at grain boundaries of Bi-doped CaMnO3- CMO Baseline CMO+Bi 2% Sub CMO+Bi 2% Sub+Cu 2% CMO+Bi 3% Sub CMO+Bi 3% Sub+Cu 4%

123

1100

222 311

1000

202

900

212

800

201 111/200 (CuO) 112 220 131

700

Temperature (K)

121

600

101

500

Intensity (a.u.)

400

Power Factor (mW m-1K-2)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Thermal conductivity (Wm-1K-1)

Page 3 of 12

1.6 1.2 0.8

24

0.4

28

32

36

40

44

48

52

56

60

64

2 ()

0.0 400

500

600

700

800

900

1000

1100

Temperature (K)

Figure 3 Temperature dependence of electrical power factor of Ca1-xBixMnCuyO3- samples.

The thermal conductivity of the samples is shown in Figure 4. The thermal conductivity of sample Ca0.97Bi0.03MnCu0.04O3-,

Figure 6 XRD diffraction patterns from the samples with different chemistry.

Figure 6 shows the X-ray diffraction spectra from samples with different chemistry of Ca1-xBixMnCuyO3- (x=0, 0.02, 0.03; y=0.02, 0.04). The diffraction peaks from CaMnO3- with the Pnma pace group (ICPD 01-073-6702) were present for all


ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 12

samples. In addition, for the samples with designed chemistry of Ca2.97Bi0.03MnCu0.04O3-, the existence of (111) peak from CuO with monoclinic structure (PDF#45-0937) is apparent. The plan-view SEM images in Figures 7 a, b and c were taken from the pressed plane of the samples.

Figure 7 SEM images showing the significant grain growth due to Cu addition and the existence of grain boundary phase. a, CaMnO3-; b, Ca0.97Bi0.03MnO3-; c, Ca0.97Bi0.03Cu0.04MnO3-.

Figure 10 HRTEM imaging taken from the Bi-doped CaMnO3- samples, and Fourier transformation from the corresponding nano domains with different orientations.

High resolution TEM (HRTEM) (Figure 10) indicates those domain boundaries are separating the nano-domains with different crystal orientations. In the sample Ca0.97Bi0.03MnCu0.04O3- with Cu addition, such domains (Figure 11) persist. Cu addition in the Bi-doped CaMnO3- samples does not result in the obvious nanostructure changes.

Figure 8 SEM elemental mapping illustrating the distribution of the sparsely distributed CuO phase at the grain boundary of sample Ca1-xBixCu0.04MnO3-.

Bi substitution of Ca did not introduce the apparent grain size and grain morphology changes in comparison with that from the baseline. By contrast, when Cu is introduced into the Bi-doped CaMnO3- as nonstoichiometric addition, there is significant grain growth from ~2 µm up to ~ 20 µm. The SEM images also clearly demonstrate the formation of the secondary phase at the grain boundaries. The SEM elementally mapping in Figure 8 indicate the secondary phase at the grain boundaries are CuO.

Figure 11 HRTEM imaging taken from the Cu-added Bi-doped CaMnO3- samples and Fourier transformation from the corresponding nano domains with different orientations.

Nanostructure changes of oxide ceramics introduced by sample chemistry modification

Figure 12 TEM diffraction contrast imaging and electron diffraction taken from the grain boundaries of Ca1-xBixMnCuyO3- samples.

Figure 9 TEM diffraction contrast imaging from a, baseline pure CaMnO3-, and b, Ca0.97Bi0.03MnO3- samples.

The samples of CaMnO3-, Ca0.97Bi0.03MnO3- and Ca0.97Bi0.03MnCu0.04O3- were subjected to TEM examination. Figure 9a depicts the representative morphology of granular pristine CaMnO3- that is without much crystal defects. The Bi substitution of Ca resulted in the formation of domain boundaries within the CaMnO3- grains (Figure 9b).

The TEM diffraction contrast image of Figure 12 is taken from the grain boundary region from the samples with Cu addition. Consistent with the SEM analysis, the grain boundary phase is about 2 m in width; the electron diffraction from the secondary phase was indexed as CuO with monoclinic crystal structure. Except for CuO, no other impurity phases were identified from all samples in the present study. The TEM image in Figure 12 also illustrates that the CuO phase has a strong and solid bonded interface with Ca1-xBixMnO3- grain matrix. There is no nano-cracking or nano-voids existing at the interface regions. EDS results also show that there is limited of solubility of Cu ions with the Cu / (Ca+Bi+Cu+Mn) ratio up to about 0.03. 20 Such limited solubility of Cu ions in the CaMnO3 grains has apparently triggered the CaMnO3- grain growth and


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the saturation of Cu in the CaMnO3- lattice rendered the formation of the CuO phase at the grain boundaries. On other hand, Ca and Mn each has limited solubility of 0.02 for Ca / (Ca+Mn+Cu) and 0.03 for Mn / (Ca+Mn+Cu) respectively in the CuO grain boundary phase as well. No Bi was detected in the CuO grain boundary phase. Effect of Bi substitution of Ca in the lattice of CaMnO3- on their electrical properties and nanostructure. Bi substitution does result in the increased pellet density, and the pellet density increases with the increase of Bi doping level, as shown in Table 1. But Bi doping did not result in the grain growth. The samples exhibited the same morphology and grain size upon Bi substitution of Ca, as demonstrated by the SEM images in Figure 7. On the other hand, Bi substitution clearly affects both the electrical resistivity and Seebeck coefficient. Consistent with literature reports, 18 the current result of simultaneous decreases in resistivity and Seebeck coefficient indicates an increased carrier concentration, resulting from Bisubstitution of Ca. The increased carrier concentration is due to the substitution of divalent Ca2+ with trivalent Bi3+ into the structure of Ca1-xBixMnO3- that will generate one electron per cation dopant. Effect of the Cu nonstoichiometric addition on the microstructure and properties of the bulk ceramics Cu addition apparently affects the morphology of the CaMnO3- grain matrix. As shown in the SEM (Figures 7a and 7b), the grain size of the CaMnO3- and Bi doped CaMnO3- is all about 1 m. In the samples with Cu addition, the grain size increase significantly to about 20 m. Meanwhile, Cu addition also triggers the formation of CuO at the grain boundaries. The evolution of the grain morphology and formation of the grain boundary network upon Cu addition is depicted in the threedimensional schematic in Figure 13.

Figure 13 Evolution of the grain morphology and formation of the grain boundary network upon Cu addition in Ca1xBixMnCuyO3- samples. a, CaMnO3-; b, Bi doped CaMnO3- with Cu addition.

To understand the formation of the grain boundary phase, the CuO precursors were made using the same chemical sol-gel route as described previously. At the same sintering temperature of 1373 K and under the same sintering atmosphere that Ca1-xBixMnCuyO3- samples are sintered, no final pellets with well-defined geometry are formed. This result indicated that CuO is probably a liquid phase during the sintering of Ca1xBixMnCuyO3- samples at high temperature of 1373 K. A separate CuO green pellet was then sintered at a lower temperature of 1233 K, and its electrical resistivity and Seebeck coefficient was measured.

800


Page 5 of 12

600 400 CMO+Bi 3% Sub CuO

200 0 -200 -400 380

390

400

410

420

430

440

Temperature (K)

Figure 14 Cu oxide is a p-type conductor, while the matrix Ca1xBixMnO3- is a n-type conductor.

The electrical resistivity (not shown) of CuO is over 10,000 times higher than that of the Bi-doped CaMnO3-. Different from CaMnO3- samples with negative Seebeck coefficient, the CuO has a positive Seebeck coefficient (Figure 14). CuO is thus confirmed to be a p-type semiconductor 21 with low electrical conductivities. Even the CuO has very high electrical resistivity, the electrical resistivity values of the Cu-added Ca1-xBixMnO3- samples were all lower than that of the samples without Cu. The decreased electrical resistivity could be partially ascribed to the decreased grain boundary density. The grain size from the samples of Cu addition increased by a factor of 20, thus the grain boundaries density decreased, which significantly reduced the scattering of electrons by the grain boundaries. On the other hand, the p-type CuO at the grain boundaries form the electrical connections between the neighboring n-type CaMnO3- grains. Therefore, the carrier scattering at the grain boundaries can be further reduced and lead to the increased carrier mobility that is evidenced from the simultaneously increased Seebeck Coefficient and decreased electrical resistivity. The present work clearly demonstrates the effect of increased carried mobility due to the presence of CuO at the grain boundaries of CaMnO3- ceramics. The present work is also consistent with the work on the Pr-doped SrTiO3 for which Pr forms Pr5O9 at the grain boundaries and increases the carrier mobility and the electrical power factor in the SrTiO3 ceramics. 22 The observed phenomenon of increased electrical conductivity of n-type CaMnO3- caused by the existence of sparsely distributed p-type CuO at the grain boundaries is also similar to what has been reported for the p-type materials. There is increased electrical conductivity of various p-type ceramics of Ca3Co4O9, 23 SmBaCuFeOx, 24 and NaxCo2O4 25 caused by the addition of n-type Ag particles. Furthermore, what has been achieved by Ag addition has also been obtained through the addition of alkaline earth elements into different thermoelectric materials.26,27,28,29 In the present study, CuO phase formation also has influence of the thermal conductivity of the samples. The thermal conductivity is calculated as ktotal = λCpρ, in which Cp is specific heat, λ is the thermal diffusivity, ρ is the density. As shown in Figure 4, the thermal conductivity of the baseline sample is lowest, due to its small grain size and low density. In comparison with pristine CaMnO3-, the Bi doped samples with the similar grain size and morphology has much higher thermal conductivity largely due to the increased density. On the other hand, as mentioned previously, the grain size from the samples of Cu addition increased by a factor of 20 apparently decreases the grain boundaries density, which significantly reduced the



scattering of phonons as well. But, the Cu-added sample has lower conductivity from that of the Bi-doped CaMnO3-. The reduced thermal conductivity in the sample with Cu addition could be attributed to the existence of the CuO grain boundary phase for additional phonon scattering. Overall, benefitting from the significantly increased power factor and reduced thermal conductivity, the sample Ca2.97Bi0.03MnCu0.04O3- has a high ZT of 0.67 at 773 K.

IV. CONCLUSIONS This study provides a novel approach of engineering the grain boundaries to tailor electrical power factor and improve the thermoelectric performance of perovskite CaMnO3-. Presence of CuO at the grain boundaries of Bi-doped CaMnO3- leads to the large increase of carrier mobility and resulted in the simultaneous increase of both the Seebeck Coefficient and electrical conductivity. For the strongly correlated CaMnO3- oxide, the present work demonstrated the effective approach of decoupling the Seebeck coefficient with electrical conductivity to significantly increase the overall power factor. The peaking power factor S2 of sample Ca2.97Bi0.03MnCu0.04O3- is 2.4 mW m-1K-2 at 377 K, and the peaking ZT is 0.67 at 773 K. The optimized thermoelectric oxide from this study also exhibits a high plateau of the ZT values from room temperature to 1073 K, and makes it versatile for thermoelectric application over a wide temperature range.

AUTHOR INFORMATION Corresponding Author * Xueyan Song, Email: [email protected], Tel: +1-304293-3269.

Author Contributions The manuscript was written through contributions of all authors. X.S, J.P and Y.C. proposed the concept and led the experiments. S. P, L.L, C.B, C.R, B.J, A. H., M. W. contributed to the thermoelectric performance analysis, microscopy imaging and the rendering the art work for schematics.

ACKNOWLEDGMENT X. Song, Y. Chen, A. H, L. L and S. P. acknowledges the financial support from FE0024009. M. Wilt acknowledges the support from NSF-DMR 1559880. B. Jackson, L. Liang, C. Boyle, C. Romo-DeLa-Cruz, X. Song acknowledge the support from NSF-DMR 1254594.

REFERENCES (1) He, J.; Tritt, T. M., Advances in Thermoelectric Materials Research: Looking Back and Moving Forward. Science 2017, 357, eaak9997. (2) Li, S. W.; Funahashi, R.; Matsubara, I.; Ueno, K.; Sodeoka, S.; Yamada, H. Synthesis and Thermoelectric Properties of the New Oxide Materials Ca3-xBixCo4O9+delta (0.0 < x < 0.75) Chem. Mat. 2000, 12, 2424-2427. (3) Bocher, L.; Aguirre, M. H.; Logvinovich, D.; Shkabko, A.; Robert, R.; Trottmann, M.; Weidenkaff, A. CaMn1−xNbxO3 (x ≤ 0.08) Perovskite-Type Phases As Promising New High-

Page 6 of 12

Temperature n-Type Thermoelectric Materials, Inorg. Chem. 2008, 47, 8077-8085. (4) Taran, S.; Chatterjee, S.; Chaudhuri, B. K. Magnetotransport properties of (La1-yPry)0.65Ca0.35MnO3 (y=0.00.7) showing crossover from A-type to CE-type Antiferromagnetic State, Phys. Rev. B 2004, 69, 184413. (5) Romo-De-La-Cruz, C.; Liang, L.; Navia, S. P.; Chen, Y.; Prucz, J.; Song, X. Role of Oversized Dopant Potassium on the Nanostructure and Thermoelectric Performance of Calcium Cobaltite Ceramics. Sustain. Energ. Fuels 2018, 2, 876-881. (6) 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. (7) Boyle, C.; Carvillo, P.; Chen, Y.; Barbero, E. J.; McIntyre, D.; Song, X. Grain Boundary Segregation and Thermoelectric Performance Enhancement of Bismuth Doped Calcium Cobaltite. J. European Ceram. Soc. 2016, 36, 601-607. (8) Fergus, J. W., Oxide Materials for High Temperature Thermoelectric Energy Conversion. J. Eur. Ceram. Soc. 2012, 32, 525-540. (9) Zhu, Y.-H.; Su, W.-B.; Liu, J.; Zhou, Y.-C.; Li, J.; Zhang, X.; Du, Y.; Wang, C.-L., Effects of Dy and Yb Co-doping on Thermoelectric Properties of CaMnO3 Ceramics. Ceram. Int. 2015, 41, 1535-1539. (10) Bathula, S.; Jayasimhadri, M.; Singh, N.; Srivastava, A.; Pulikkotil, J.; Dhar, A.; Budhani, R. Enhanced thermoelectric figure-of-merit in spark plasma sintered nanostructured n-type SiGe alloys. Appl. Phys. Lett. 2012, 101, 213902. (11) Wang, X.; Lee, H.; Lan, Y.; Zhu, G.; Joshi, G.; Wang, D.; Yang, J.; Muto, A.; Tang, M.; Klatsky, J. Enhanced thermoelectric figure of merit in nanostructured n-type silicon germanium bulk alloy. Appl. Phys. Lett. 2008, 93, 193121. (12) Bocher, L.; Aguirre, M. H.; Robert, R.; Logvinovich, D.; Bakardjieva, S.; Hejtmanek, J.; Weidenkaff, A., Hightemperature Stability, Structure and Thermoelectric Properties of CaMn1-xNbxO3 Phases. Acta Mater. 2009, 57, 5667-5680. (13) Wang, Y.; Sui, Y.; Fan, H.; Wang, X.; Su, Y.; Su, W.; Liu, X., High Temperature Thermoelectric Response of Electron-doped CaMnO3. Chem. Mat. 2009, 21, 4653-4660. (14) Huang, X.; Miyazaki, Y.; Kajitani, T. High Temperature Thermoelectric Properties of Ca1−xBixMn1−yVyO3−δ (0≤ x= y≤ 0.08). Solid State Commun. 2008, 145, 132-136. (15) Reimann, T.; Töpfer, J., Thermoelectric Properties of Gd/W Double Substituted Calcium Manganite. J. Alloy. Compd. 2017, 699, 788-795. (16) Zhu, Y.; Wang, C.; Su, W.; Li, J.; Liu, J.; Du, Y.; Mei, L., High-temperature Thermoelectric Performance of Ca0.96Dy0.02RE0.02MnO3 Ceramics (RE= Ho, Er, Tm). Ceram. Int. 2014, 40, 15531-15536. (17) Wang, Y.; Sui, Y.; Su, W., High Temperature Thermoelectric Characteristics of Ca0.9R0.1MnO3 (R= La, Pr,…, Yb). J. Appl. Phys. 2008, 104, 093703. (18) Kabir, R.; Tian, R.; Zhang, T.; Donelson, R.; Tan, T. T.; Li, S., Role of Bi Doping in Thermoelectric Properties of CaMnO3. . J. Alloy. Compd. 2015, 628, 347-351. (19) Elsheikh, M. H.; Shnawah, D. A.; Sabri, M. F. M.; Said, S. B. M.; Hassan, M. H.; Bashir, M. B. A.; Mohamad, M., A Review on Thermoelectric Renewable Energy: Principle


Page 7 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Parameters That Affect Their Performance. Renew. Sust. Energ. Rev. 2014, 30, 337-355. (20) Ravi, S.; Kar, M.; Borah, S.; Krishna, P., Neutron Powder Diffraction Studies in CaMn1‐xCuxO3 (x= 0, 0.2). Cryst. Res. Technol. 2008, 43, 1318-1322. (21) Hartung, D.; Gather, F.; Hering, P.; Kandzia, C.; Reppin, D.; Polity, A.; Meyer, B.; Klar, P., Assessing the Thermoelectric Properties of CuxO (x= 1 to 2) Thin Films as a Function of Composition. Appl. Phys. Lett. 2015, 106, 253901. (22) Mehdizadeh Dehkordi, A.; Bhattacharya, S.; Darroudi, T.; Graff, J. W.; Schwingenschlögl, U.; Alshareef, H. N.; Tritt, T. M., Large Thermoelectric Power Factor in Pr-Doped SrTiO3−δ Ceramics via Grain-Boundary-Induced Mobility Enhancement. Chem. Mat. 2014, 26, 2478-2485. (23) Mikami, M.; Ando, N.; Funahashi, R., The effect of Ag Addition on Electrical Properties of the Thermoelectric Compound Ca3Co4O9. J. Solid State Chem. 2005, 178, 21862190. (24) Zeng, C.; Butt, S.; Lin, Y. H.; Li, M.; Nan, C. W., Enhanced Thermoelectric Performance of SmBaCuFeO5+δ/Ag Composite Ceramics. J. Am. Ceram. Soc. 2016, 99, 1266-1270.

(25) Ito, M.; Sumiyoshi, J., Enhancement of Thermoelectric Performance of NaxCo2O4 with Ag Dispersion by Precipitation from Ag+ Aqueous Solution. J. sol-gel Sci. Technol. 2010, 55, 354-359. (26) Sotelo, A.; Costa, F.; Ferreira, N.; Kovalevsky, A.; Ferro, M.; Amaral, V.; Amaral, J.; Rasekh, S.; Torres, M.; Madre, M. Tailoring Ca3Co4O9 Microstructure and Performances Using a Transient Liquid Phase sintering Additive. J. European Ceram. Soc. 2016, 36, 1025-1032. (27) Li, Y.-N.; Wu, P.; Zhang, S.-P.; Chen, S.; Yan, D.; Yang, J.-G.; Wang, L.; Huai, X.-L. Thermoelectric Properties of Lower Concentration K-doped Ca3Co4O9 Ceramics. Chinese Physics B 2018, 27, 057201. (28) Chere, E. K.; Zhang, Q.; Dahal, K.; Cao, F.; Mao, J.; Ren, Z. Studies on Thermoelectric Figure of Merit of Na-doped p-type Polycrystalline SnSe. J. Mater. Chem. A 2016, 4, 18481854. (29) Karakaya, G. Ç.; Özçelik, B.; Nane, O.; Sotelo, A.; Rasekh, S.; Torres, M.; Madre, M. Improvement of Bi2Sr2Co2Oy Thermoelectric Performances by Na Doping. J. Electroceram. 2018, 40, 11-15.



Table of Contents artwork


Page 8 of 12

Page 9 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10 234x177mm (307 x 307 DPI)



Figure 11 188x153mm (300 x 300 DPI)


Page 10 of 12

Page 11 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 12 250x130mm (300 x 300 DPI)



Figure 13 289x136mm (300 x 300 DPI)


Page 12 of 12

Grain Boundary Phase Segregation for Dramatic Improvement of the

Recommend Documents