Optimization and Analysis of Thermoelectric Properties of Unfilled Co1

Subscriber access provided by GAZI UNIV

Article

Optimization and analysis of thermoelectric properties of unfilled Co1-x-yNixFeySb3 synthesized via a rapid hydrothermal procedure Ahmad Gharleghi, Yu-Hsien Chu, Fei-Hung Lin, Zong-Ren Yang, Yi-Hsuan Pai, and Chia-Jyi Liu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.5b09327 • Publication Date (Web): 11 Feb 2016 Downloaded from http://pubs.acs.org on February 12, 2016

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 free 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 accessible to all readers and 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.

ACS Applied Materials & Interfaces 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 57

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

Optimization

and

analysis

of

thermoelectric

properties

of

unfilled

Co1-x-yNixFeySb3 synthesized via a rapid hydrothermal procedure

Ahmad Gharleghi, Yu-Hsien Chu, Fei-Hung Lin, Zong-Ren Yang, Yi-Hsuan Pai, and Chia-Jyi Liu*

Department of Physics, National Changhua University of Education, Changhua 500, Taiwan. [email protected] Abstract A series of nanostructured co-doped Co1-x-yNixFeySb3 were fabricated using a rapid hydrothermal method at 170°C for a duration of 12 h, followed by evacuated-and-encapsulated heating at 580°C for a short period of 5 h. The resulting samples are characterized using powder x-ray diffraction, field emission scanning electron microscopy, bulk density, electronic and thermal transport measurements. The power factor of Co1-x-yNixFeySb3 is significantly enhanced in the high temperature region due to significant enhancement of the electrical conductivity and absolute value of thermopower. The latter arises from the onset of bipolar effect being shifted to higher temperatures as compared with the nondoped CoSb3. The room-temperature thermal conductivity falls in the range between 1.22 Wm-1K-1 and 1.67 Wm-1K-1 for Co1-x-yNixFeySb3. The thermal conductivity of both the (x,y) =

ACS Paragon Plus Environment


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

(0.14,10) and (0.14,12) samples are measured up to 600 K and found to decrease with increasing temperature. The thermal conductivity of the (0.14,10) sample goes down to ~1.02 Wm-1K-1. As a result, zT = 0.68 is attained at 600 K. The lattice thermal conductivity is analyzed to gain insight into the contribution of various scattering processes that suppress the heat transfer through the phonons in Co1-x-yNixFeySb3. The effect of simultaneous presence of Co, Ni and Fe elements on the electronic structure and transport properties of Co1-x-yNixFeySb3 is described using the quantum mechanical tunneling theory of electron transmission among the potential barriers.

Keywords: thermoelectrics; cobalt skutterudites; hydrothermal synthesis; co-doping effects; bipolar effects; low thermal conductivity


Page 2 of 57

Page 3 of 57

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


I.

Introduction The efforts of harvesting waste energy using thermoelectricity have recently

attracted great interest among interdisciplinary scientists and engineers. Fabrication of materials with high thermoelectric efficiency is one of the requirements for satisfactorily achieving this goal. The efficiency of a thermoelectric material is commonly evaluated from the dimensionless figure of merit, zT = σS2T/κ, where S, σ, κ and T are the thermopower (Seebeck coefficient), electrical conductivity, thermal conductivity and the absolute temperature, respectively. The cobalt skutterudite is one of the materials exhibiting phonon-glass and electron-crystal (PGEC) features, which have attracted many ongoing research activities for enhancing zT values.1-4 Skutterudites with chemical formula of TPn3 (T = Co, Rh, or Ir and Pn = P, As, or Sb) crystallize in the cubic lattice of CoAs3-type with space group Im3 , which possesses two intrinsic

Sb-icosahedral voids per unit cell.5 The Sb-icosahedral voids can be partially filled by inserting rare earth and some other metal elements to suppress the lattice thermal conductivity while retaining the large power factor of σS2.2,3,6-8 The void of the CoSb3 skutterudite is large enough to allow random rattling of the accommodated filler ions that may act as phonon scattering centers.9,10 The transport properties of nondoped CoSb3 could be optimized via tuning its electronic band structure, since its estimated direct band gap is ~0.8 eV being larger than 10 kBT at room temperature.11,12 Partial



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

substitution of metal ions in CoSb3 is one of the well-known strategies towards optimizing thermoelectric properties of PGEC materials and is widely practiced using solid state synthesis.5, 13-17 The Co/Sb site doping might decrease the lattice thermal conductivity via introducing aliovalent impurities or point defects to cut down a range of phonon frequencies.18,19 The majority carrier type of CoSb3 strongly depends on the dopant species. Partial substitution of Fe, Ru or Os on the Co site, and Ge or Sn on the Sb site have been reported to enhance thermoelectric properties of p-type CoSb3,11,20 while light substitution of Ni, Pd or Pt on the Co site, and Te on the Sb site is adopted to enhance thermoelectric properties of n-type CoSb3.13,21-23 From the aspect of fabrication procedure, it is worth noting that the Co/Sb site doping or filling the voids of CoSb3 using solid state reactions is employed more commonly than wet chemical methods.5,11,13,15-18,20-24 Wet chemical methods have not been very common for synthesizing co-doped cobalt skutterudites probably due to difficulties in obtaining as-desired products.25-28 Before partially filling the voids in the structure, we first carry out the studies of co-doping effects on the transport properties of unfilled cobalt skutterudite. In the present work, we adopt a rapid hydrothermal method29 to synthesize a series of unfilled Co1-x-yNixFeySb3 and study their transport properties. As a result of optimizing the Ni and Fe doping content, the power factor is markedly enhanced. The room-temperature thermal conductivity of Co1-x-yNixFeySb3 is even


Page 4 of 57

Page 5 of 57

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


lower than that of our nondoped CoSb3 in some cases. In addition, the total thermal conductivity of Co1-x-yNixFeySb3 decreases with increasing temperature. The lattice thermal conductivity is analyzed so that we can estimate the contribution from various scattering processes to the thermal conductivity of Co1-x-yNixFeySb3. The electrical conductivity is also analyzed for understanding the predominant conduction mechanism. In addition, the effects of simultaneous presence of Co, Ni and Fe elements on the electronic structure and transport properties of Co1-x-yNixFeySb3 is described using the quantum mechanical tunneling theory of electrons transmission among the potential barriers. All these results indicate that co-doping of Fe and Ni on the Co site leads to decoupling between the power factor and thermal conductivity of Co1-x-yNixFeySb3.

II. Experimental Details A series of Co1-x-yNixFeySb3 skutterudites (x = 0.0, 0.07, 0.14 and y = 0.0, 0.03, 0.05, 0.10, 0.12, 0.14) was fabricated using the hydrothermal synthesis. The starting materials of SbCl3, CoCl2·6H2O, NiCl2·6H2O, and FeCl3·6H2O were weighted and loaded into a Teflon cup with a total volume of ~185 mL containing 110 mL deionized water. In the reaction, we used 10.65 millimole SbCl3 with a ratio of Sb/(Co+Ni+Fe) = 3. The solution was then sonicated at 47°C for 30 min, followed by adding 74.26



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

millimole of NaOH with further sonication for another 15 min. To create a reductive environment, 78.56 millimole NaBH4 was added to the above solution, followed by sonication at 47°C for 30 min. The cup containing the solution was then loaded into an autoclave and heated in an oven with heating rate of 2.6°C /min at 170°C for 12 h.29 The precipitated product was washed using ethanol and deionized water. Drying of the wet powders was carried out using a rotary evaporator vacuum dryer. To avoid oxidation of as-synthesized fine powders during drying, argon gas was used to purge the flask before evacuation and introduced into the flask while breaking the vacuum after drying. The dried powders were then cold-pressed at 16.88 MPa to form a parallelepiped. The prepared parallelepipeds were then loaded into a Pyrex ampoule, which was evacuated using a diffusion pump to reach 10-5 Torr to 10-6 Torr and then sealed.30 The parallelepipeds inside the encapsulated ampoule were then heated in a tubular furnace at a rate of 2.6°C/min to 580°C, and hold for 5 h. The phase identification of as-synthesized powders and as-heated Co1-x-yNixFeySb3 samples was carried out using a Shimadzu XRD-6000 diffractometer equipped with Fe Kα radiation. The lattice constants for all the samples were refined using the collected XRD data. The morphology of the samples was examined using a JEOL JSM-6700F field emission scanning electron microscope (FE-SEM). Electrical resistivity and thermopower measurements for parallelepipeds


Page 6 of 57

Page 7 of 57

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


(10mm×1.5mm×1.2mm) were simultaneously carried out from liquid nitrogen temperature up to 700 K. Thermopower measurements were carried out using steady-state techniques. The temperature difference between hot and cold ends of the sample was measured using a type E differential thermocouple using a Keithley 2000 multimeter. The thermally generated Seebeck voltage across the sample was measured using a Keithley 2182 nanovoltmeter.31 The thermopower of sample was obtained by subtracting the thermopower of the Cu. Electrical resistivity was measured using standard four-probe techniques with reversing the current sources to cancel thermoelectric voltages. Thermal conductivity measurements were carried out using transient plane source techniques with very small temperature perturbations of the sample material by the Hot Disk thermal constants analyzer as described in detail elsewhere.30 The uncertainty for the electrical resistivity, thermopower, and thermal conductivity is about ±3%, ±4% and ±5%, respectively. Hall measurements was performed using the van der Pauw method under an applied magnetic field of 0.55 T (ECOPIA:HMS -3000). The relative bulk density of all the samples was measured using the Archimedes’ method with an uncertainty of about ±2%.

III.

Result and Discussion Fig. 1 shows the x-ray diffraction (XRD) patterns of hydrothermally synthesized



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

powders of CoSb3 and Co0.9Ni0.07Fe0.03Sb3. The peak occurring at 2θ ≈ 36° is identified as the most intense (012) peak of the Sb phase. The XRD patterns in Fig. 1 also show the presence of CoSb3, CoSb2 and CoSb as minority phases. In order to obtain single phase of CoSb3, the hydrothermally synthesized powders were heated in an evacuated-and-encapsulated ampoule at 580°C for 5 h. The XRD patterns of as-heated Co1-x-yNixFeySb3 samples with (x,y) = (0.0,0.0), (0.07,0.03), (0.07,0.05), (0.14,0.10), (0.14,0.12), (0.14,0.14) are displayed in Fig. 2. Except the two samples with the composition of y = 0.12 and 0.14, all the Co1-x-yNixFeySb3 samples are of single phase with the CoSb3 structure. The (012) peak intensity of the Sb phase for the sample with (x,y) = (0.14,0.12) is lower than that for the one with (x,y) = (0.14,0.14). The refined lattice constants from XRD data for all the heat-treated Co1-x-yNixFeySb3 are listed in Table 1. The lattice constant increases for all of the Ni and Fe co-doped samples as compared with that of the nondoped CoSb3. The density of all the Co1-x-yNixFeySb3 bulks falls in the range of 74 and 79 % of the theoretical density as shown in Table 1. Figs. 3a and 3b display fractured-surface FE-SEM of as-fabricated Co1-x-yNixFeySb3 bulks taken for the (x,y)=(0.14,0.10) and (0.14,0.12) samples, respectively. The obtained micrographs illustrate similar polycrystalline structures with particle sizes in the range from ~50 to ~400 nm for the (x,y)=(0.14,0.10) and (0.14,0.12) samples.


Page 8 of 57

Page 9 of 57

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 Hall carrier concentration and mobility are shown in Table 1. The Hall data show that electrons are the majority charge carriers for all the Co1-x-yNixFeySb3 samples. Compared with the nondoped CoSb3 sample, the electron concentration significantly increases for all the Ni and Fe co-doped samples. In particular, the electron carrier concentration of the (x,y) = (0.07,0.05) and (0.14,0.14) samples is -1.84 × 1020 cm-3 and -1.67 × 1020 cm-3, respectively, which are the two largest among the samples. As shown in Table 1, the electron mobility increases with increasing Ni dopants, while decreases upon Fe co-doping for each set of samples. Reduction of conduction band deformation potential might be responsible for the enhanced mobility upon Ni doping, in particular for both the (x,y) = (0.07,0.0) and (0.14, 0.10) samples.22 The reduction of mobility upon Fe doping seems to be similar to the case in p-type Co1-xFexSb3. Nevertheless, the electron concentration increases in Co1-x-yNixFeySb3 for a given x at high y content, which is in sharp contrast to an increase of hole concentration upon increasing Fe content in p-type Co1-xFexSb3.32 Fig. 4 shows the temperature dependence of electrical conductivity for Co1-x-yNixFeySb3 with the composition of (x,y) = (0,0), (0.07,0.0), (0.07,0.03), (0.07,0.05), (0.14,0.10), (0.14,0.12), (0.14,0.14) from liquid nitrogen temperature up to 700 K. As compared with the nondoped CoSb3 (Fig. 4a), the room-temperature electrical conductivity monotonically increases with increasing Ni and Fe doping



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

content for Co1-x-yNixFeySb3. In addition to the influence of fine-size particles on limiting the electron mobility, the existence of porosity in the bulk material would also limit the mobility (Fig. 3), which decreases electrical conductivity. Higher porosity is then expected to reduce the electrical conductivity to a larger extent. Even though there only exists a small difference in density among the materials, it seems that higher porosity in the material correlates with its smaller mobility for Co1-x-yNixFeySb3. However, the electrical conductivity of Co1-x-yNixFeySb3 is apparently dominated by its carrier concentration. The electrical conductivity of the nondoped CoSb3 exhibits non-metallic behavior from 77 to 700 K, which will be discussed later. Conduction behavior depends on the temperature and co-doping level of Ni and Fe. For the series of Co0.93-yNi0.07FeySb3 with y = 0.0, 0.03 and 0.05, metal-like conduction is observed for 83≦T≦125 K, 86≦T≦182 K and 78≦T≦222 K, respectively. At temperatures higher than the above temperature regimes, the conduction switches to non-metallic behavior for all the three Co0.93-yNi0.07FeySb3 samples, which can be readily seen in Fig. 4a. Apparently the transition from metal-like conduction to non-metallic behavior shifts to higher temperatures with increasing Fe content. For Co0.76Ni0.14Fe0.10Sb3, the metal-like conduction occurs at the both temperature intervals of 77≦T

Optimization and Analysis of Thermoelectric Properties of Unfilled Co1

Recommend Documents