Fabrication and Thermoelectric Properties of n-Type CoSb2.85Te0.15

Subscriber access provided by Kaohsiung Medical University

Functional Inorganic Materials and Devices

Fabrication and thermoelectric properties of ntype CoSb2.85Te0.15 using selective laser melting Yonggao Yan, Hongquan Ke, Jihui Yang, Ctirad Uher, and Xinfeng Tang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b01564 • Publication Date (Web): 04 Apr 2018 Downloaded from http://pubs.acs.org on April 5, 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 29 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

Fabrication and thermoelectric properties of n-type CoSb2.85Te0.15 using selective laser melting Yonggao Yan1, Hongquan Ke1, Jihui Yang2, Ctirad Uher3, Xinfeng Tang1

1, State Key Laboratory of Advanced Technology for Materials Synthesis and Processing, Wuhan University of Technology, Wuhan, Hubei 430070, China *Email: [email protected]; [email protected] 2, Department of Materials Science and Engineering, University of Washington, Seattle, Washington 98195, United States 3, Department of Physics, University of Michigan, Ann Arbor, Michigan 48109, United States

ABSTRACT: We report on a non-equilibrium fabrication method of n-type CoSb2.85Te0.15 skutterudites using selective laser melting (SLM) technology. A powder of CoSb2.85Te0.15 was prepared by self-propagating high-temperature synthesis (SHS) and served as the raw material for the SLM process. The effect of SLM processing parameters, such as the laser power and scanning speed, on the quality of the forming CoSb2.85Te0.15 thin layers has been systematically analyzed, and the optimal processing window for SLM has been determined. A brief post annealing at 450 oC for 4 hours, following the SLM process, has resulted in a phase-pure CoSb2.85Te0.15 bulk material deposited on a Ti substrate. The Seebeck coefficient of the annealed SLM prepared bulk material is close to that of the sample prepared by the traditional sintering method, and its maximum ZT value has reached 0.56 at 823 K. Moreover, a Ti-Co-Sb ternary compound transition layer of about 70 μm in thickness was found at a dense interface between CoSb2.85Te0.15 and the Ti substrate. The contact resistivity was measured as 37.1 μΩcm2. The results demonstrate that SLM, coupled with post annealing, can be used for fabrication of incongruently melting skutterudite compounds on heterogeneous substrates. This lays an important foundation for the follow-up research utilizing energy efficient SHS and SLM processes in rapid printing of thermoelectric modules. Keywords : Thermoelectric materials, Selective Laser Melting (SLM), CoSb3, Self-propagating High-temperature Synthesis (SHS), contact resistivity.

ACS Paragon Plus Environment

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

1. Introduction With the depletion of traditional fossil fuels and the increasing environmental pollution, it has been a key exploring new and efficient ways to utilize energy and therefore achieving the goal of saving energy and reducing emission. In many traditional industrial processes, including production of cement, steel and metallurgy, there has always been a large amount of waste heat generated in the medium temperature range between 200-500 ℃.1 Therefore, it is of great significance to recycle at least a fraction of this waste heat and thus limit the ever rising heat pollution. Thermoelectric energy generation based on the Seebeck effect can directly convert waste industrial heat to electricity. Such thermoelectricity-based energy conversion has many advantages, including pollution-free operation, use of no moving parts, and consequently exceptionally high reliability that benefitted many fields spanning from aerospace, military and other areas.2-4 Thermoelectric materials that are particularly efficient in the 200 to 500 oC range are CoSb3-based skutterudites. They have been intensively researched during the last two decades and their main attributes are the high carrier mobility, high electrical conductivity, excellent Seebeck coefficient, and good operational stability.5,6 While skutterudite materials have been shown time and again to possess high ZT values,7-12 their deployment in large-scale industrial applications has been hampered by inadequate development of thermoelectric modules. The traditional fabrication processes, starting with the development of thermoelectric materials and culminating in the assembly of reliable thermoelectric modules are rather complex. Bulk materials of both p- and n-type must be sliced, electroplated, diced, assembled and soldered, and this is time consuming, wasteful of materials and energy and, therefore, carry a rather high cost penalty. These drawbacks of the traditional materials synthesis and module assembly fabrication have been a serious constraint on the large-scale deployment of TE power generation technology.13-15 Therefore, it is vitally important to seek more efficient TE module fabrication technology that has high utilization rate of raw materials and consumes minimal amounts of energy.


Page 2 of 29

Page 3 of 29 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


Selective Laser Melting (SLM), as a novel and rapid manufacturing technology, has become one of the most developed in the field of additive manufacturing.16 SLM utilizes a laser to melt a thin layer of dispersed powder materials, which solidifies rapidly to form a certain shape. 3D objects can be built by successive steps of powder deposition followed by laser-induced melting.17 Compared with the traditional synthesis processes, the SLM method avoids the material loss during the dicing and slicing processing steps. The laser spot size is so small that one can obtain parts with the desired finish quality and, more importantly, highly dense products of complex shape can be fabricated quickly.18-20 During the laser-induced fusion and solidification, the heating and cooling rates are extremely fast, leading to materials with the highly refined grain structure resulting in a reduced thermal conductivity. Hence, provided that this technology can be applied to the rapid laser forming of p- and n-type TE legs and their bonding to the electrode strips, a one-step rapid manufacturing process of thermoelectric modules would be realized. This would then greatly shorten the production cycle and cut down the cost; moreover, the TE performance of materials and modules would likely be enhanced due to the rapid non-equilibrium processing. At present, the SLM technology is being widely applied to the printing of metallic structural materials.21-26 However, it is rarely used in the preparation of TE and other functional materials and their modules. In contrast to the metal-based structural materials, thermoelectric materials typically possess much lower melting points, are poor heat conductors, and have weak mechanical strength and thermal shock resistance. Thus, it is a considerable challenge to adopt the SLM technology to the preparation of TE materials, which must have a well-controlled chemical composition, structure, and excellent TE transport properties. In 2016, an attempt was made by El-Desouky et al.27 to utilize the SLM technology to prepare Bi2Te3-based thermoelectric materials. The authors focused on investigations of the effect of laser processing parameters on the depth of the melt zone in the powder bed of Bi2Te3. However, they did not study how variations in the chemical composition affect thermoelectric properties following laser processing. In 2017, our team28 systematically investigated the effect of SLM processing parameters on the



microstructure, chemical composition, and the overall quality of n-type Bi2Te2.7Se0.3 layers formed during the process. This was the first successful use of the SLM technology in preparing n-type bulk Bi2Te2.7Se0.3 materials. The ZT values attained in these structures reached 0.84 at 400 K. As one of the best medium-temperature thermoelectric materials, n-type CoSb2.85Te0.15 skutterudite was selected as the research object of this work. A powder of CoSb2.85Te0.15, which served as the raw material for the SLM processing, was prepared by crushing an ingot synthesized by the self-propagating high-temperature synthesis (SHS).29,30 The effect of laser processing parameters on the quality of individual layers forming during the process has been systematically investigated, and the processing window for well-formed layers has been determined. A stack of individually processed CoSb2.85Te0.15 layers was deposited on a Ti substrate and subjected to post annealing. The resulting bulk structure possessed a sub-micron grain size and attained a maximum ZT value of 0.56 at 823 K. The bonding interface between the Ti substrate (an electrode) and CoSb2.85Te0.15 was also analyzed. This paper documents the feasibility of using SLM to process incongruently melting compounds, such as skutterudites, on a metallic substrate, which are critically important for future fabrication of thermoelectric modules by the one-step 3D printing technique. 2 Experimental Method Co powder (99.9%, 200 mesh), Sb powder (99.99%, 200 mesh) and Te powder (99.999%, 200 mesh) were weighed according to the CoSb2.85Te0.15 stoichiometric ratio. After being ground and well-mixed, the powder was cold-pressed to a pellet. An ingot of CoSb2.85Te0.15 was produced by the SHS reaction ignited by torch heating in an evacuated quartz tube. The SHS-prepared ingot was ground into powder (400 mesh) by ball milling. Figure 1 shows the morphology and size distribution of the obtained powder; the average size is around 4.3 μm. The powder was dispersed in ethanol to obtain a slurry with a solid content of 30%. Ti is selected as the substrate due to its similar coefficient of thermal expansion to CoSb3. A high-purity Ti powder (99.99%,


Page 4 of 29

Page 5 of 29 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


300 mesh) was sintered by SPS at 900 oC for 8 minutes under 40 MPa to obtain a cylinder with the diameter of 16 mm, which served as a substrate in the SLM experiment. The slurry was spread on the substrate with a blade, when the ethanol evaporates, a flat powder bed layer of 30-40 μm thickness on the substrate is ready for further laser processing. The SLM processing used a commercial fiber laser (MFSC-100W), with the maximum output power of 100 W, laser wavelength of 1064 nm, and the focal spot size of about 100 μm. The scanning strategy used is a single parallel scan shown in figure 2 (b). All SLM processing was carried out under 0.5 atm of argon. The schematic diagram illustrates the customized apparatus used for the SLM experiments are presented in the Supporting Information (Figure S1). The particle size and distribution of the powders were measured by a Malvern particle size analyzer (Mastersizer 2000, Marlven). The microstructure of SLM-processed layers was observed through field emission scanning electron microscopy (S-4800, Hitachi). The phase composition of the resulting bulk samples was analyzed by powder X-ray diffraction (Empyrean, PANalytical). The chemical composition of the samples was determined by electron probe micro-analysis (JXA-8230, JEOL EPMA). The surface Seebeck coefficient and its distribution, and the contact resistance between the material and the electrode were measured with a Potential Seebeck Microprobe (PSM, Panco). The electrical conductivity and the Seebeck coefficient of the bulk sample were measured with a ZEM-3 instrument from Ulvac-riko, Inc.

The thermal conductivity of the SLM-prepared bulk samples was

calculated from the relationship  = DCpd, where D is the thermal diffusivity obtained by the laser flash method (LFA-457, Netzsch), Cp is the specific heat measured by a differential scanning calorimeter (DSC Q20, TA Instrument), and d is the density measured by the Archimedes method. 3.1 Processing Window Figure 2 shows the relationship between laser parameters and the quality of the layer formed during the SLM process. Usually, we used the laser volume energy density EV to comprehensively assess the effect of processing parameters on the powder bed during SLM. The relevant formula is expressed as:



EV 

P  dh

Page 6 of 29

(1)

where P is the laser power (W), ν is the scanning speed (mm/s), d is the hatch spacing, which is the distance between two neighbor scan vectors (mm), and h is the thickness (mm) of the powder layer. Figures 2(a)-2(d) show the results of four kinds of single layer forming planes of 2×2 mm2 processed with different laser powers (P = 4~18 W) and scanning speed (ν = 30~300 mm/s). The typical morphologies include: vaporization, flat surface, instability, and balling. The hatch spacing d = 0.06 mm and the powder bed thickness h = 0.04 mm are kept fixed. The laser scanning path is shown by the red arrow in Fig. 2(b). According to the above four typical morphologies, the processing window can be divided into four distinct regions, delineated in Fig. 2(e). When the laser power is relatively large (P > 16 W) and the scanning speed is low (ν ≤ 40 mm/s), there are notable pores on the surface of the forming layer and the boundary around the forming surface is not smooth, as shown in Fig. 2(a). The reason is that the laser energy density EV in this case is too large so that the molten pool’s temperature is too high. This causes severe evaporation of the material, leaving voids on the formed surface. Moreover, heat will dissipate from the overheated molten pool to the powder bed and substrate, resulting in complete melting of the powder even in un-scanned areas around the forming surface. This gives rise to uneven edges on the forming surface. When the laser power is somewhat reduced but remains comparatively large (P > 12 W), and the EV decreases (58 J/mm3 < EV < 130 J/mm3) because the scan rate increases (ν > 50 mm/s), the boundary around the forming surface is straight and un-melted powder is found around the forming zone. Such energy density EV assures that the powder melts where it should, the evaporation is not too severe, and resulting surface is relatively flat and dense, as shown in Fig. 2(b). When the scanning speed is too high, v > 100 mm/s, and the corresponding EV decreases to 25~58 J/mm3, the forming surface becomes uneven, as shown in Fig. 2(c). The voids can still be seen in figure 2(c) even EV is much decreased compared with EV in figure (a). Due to the laser density EV may not be homogenous inside the laser spot,


Page 7 of 29 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


some area in the layer can be still overheated leading to evaporation of materials and voids on the layer surface. When the laser power falls below about 12 W and the scanning speed is too high so that the energy density becomes less than about EV < 25 J/mm3, the laser energy is not large enough to completely melt the powder bed. The too low temperature of the melt and its high viscosity results in a large wetting angle between the molten pool and the substrate, eventually giving rise to balling, as depicted in Fig. 2(d). Above results show that, in order to obtain the high-quality n-type CoSb2.85Te0.15 material, the laser power in the SLM process should be selected somewhere between 12 W and 16 W, and the laser energy density should be controlled within 58~130 J/mm3 by selecting the proper scanning speed. It is critical to select the optimal forming process parameters in order to be able to fabricate CoSb2.85Te0.15 skutterudites using the SLM process.

3.2 Phase, microstructure and properties of bulk skutterudites after SLM Bulk samples of CoSb2.85Te0.15 with the thickness of 1.7 mm were fabricated by individually processing 80 layers on top of each other. Figure 3(a) shows the X-ray diffraction pattern of the as-prepared bulk sample processed by SHS-SLM. The pattern indicates that the main phases in the sample are CoSb2, Sb and only a small amount of CoSb3. We ascribe this chiefly to the phase separation driven by the rapid heating and cooling process during SLM. Meanwhile, the broadening of the diffraction peaks is associated with the small grain size produced in the non-equilibrium process. Figure 3(b) shows the XRD pattern of the bulk sample after annealing at 450 oC for 4h. The sample shows an almost pure CoSb3 phase following this short annealing. The rapid phase transition is due to a greatly reduced diffusion distance between CoSb2 and Sb that favors the formation of the CoSb3 phase, which is the consequence of already mentioned highly refined grain structure. Figure 4 shows cross-sectional FESEM images of bulk CoSb2.85Te0.15 samples prepared by both SHS-SPS and SHS-SLM. The grain size of the bulk material prepared by SPS sintering is mostly between 1~5 µm (Fig. 4a), but the grain size of



the SLM-prepared CoSb3 is much reduced as clearly seen from figure 4(a) and 4(c). The difference in the grain size is mainly due to the rapid cooling rate of the molten pool during the SLM process that provides insufficient time for the grain growth. Furthermore, during the brief post annealing process, the mass transfer is significantly promoted, which reduces the usual one-week long annealing process needed to complete the peritectic reaction in the traditional synthesis to just a few hours. Meanwhile, the sample’s grain size becomes more uniform after annealing, and the grains are bonded tightly together with clean grain boundaries. However, there exist a few micro-cracks in the annealed samples. They form as a consequence of the shrinkage associated with the peritectic reaction of CoSb2 and Sb during the annealing process, which produces tensile stresses in the body of the material. The stress is released by the formation of cracks. Figure 5 shows the distribution of surface Seebeck coefficients measured at room temperature on bulk samples prepared by SHS-SLM. Figures 5(a) and 5(b) depict the test results before annealing, while Figs. 5(c) and 5(d) show test results after annealing for 4 hours. Prior to annealing, the Seebeck coefficients ranged from 0 to -60 μV/K, with the average Seebeck coefficient of -26.1 μV/K. After annealing, the Seebeck coefficient has dramatically increased and covered the range from -80 to -140 μV/K with the average value of the Seebeck coefficient of -107.3 μV/K. The as-prepared sample (no annealing) contained a large amount of the CoSb2 phase and of pure Sb, which both have a rather low Seebeck coefficient and therefore highly degraded thermoelectric properties. After annealing, the sample transformed into an essentially pure skutterudite phase (CoSb2.85Te0.15) with the Seebeck coefficient close to that of the SHS-SPS-processed skutterudite. Moreover, as depicted in Figs. 5(c) and 5(d), the Seebeck coefficient of the annealed sample is distributed in a significantly narrower range, indicating that the composition and structure of the annealed sample is more uniform. Figure 6 shows the temperature dependence of the electrical conductivity, the Seebeck coefficient, the thermal conductivity, and the thermoelectric figure of merit ZT before and after annealing. One notes an abrupt change in the transport properties


Page 8 of 29

Page 9 of 29 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


of the SHS-SLM sample at temperatures between 300 and 400 oC. This is due to the presence of a large amount of heterogeneous phases, such as CoSb2 and Sb, in the un-annealed sample. When the temperature rises to 300 oC, CoSb2 and Sb gradually fuse to become the CoSb3 phase through a solid-state reaction, the process that is substantially completed by 400 oC. Above 400 oC, the Seebeck coefficient of the annealed and un-annealed samples is essentially the same because testing at these temperatures has effectively annealed the originally un-annealed sample. An interesting situation is observed in the case of the electrical conductivity in Fig. 6(a). Up to about 300 oC, the electrical conductivity of the annealed sample is significantly higher than that of the un-annealed one. This is not surprising as the heterogeneous nature of the un-annealed sample strongly limits the carrier mean free path. However, above about 300 oC, the electrical conductivity of the originally un-annealed sample is nearly twice as large as that of the sample subjected to 4 hours of annealing. This suggests that the forced 4 hour-long annealing process at a fixed temperature of 450 o

C has introduced micro-cracks in the structure while the gradual temperature ramp up

during the property measurement appears as a more gentle annealing process, leaving behind a less defected structure. However, both the annealed and un-annealed SLM samples show much reduced electrical conductivities as compared with the sample prepared by SHS-SPS. A similar situation, but with a twist, is observed in the thermal conductivity. Here, the initially un-annealed sample shows a dramatically higher thermal conductivity than the annealed structure on account of the former’s highly heat conducting phases of CoSb2 and pure Sb. When CoSb2 and Sb fuse to form the skutterudite structure, the thermal conductivity precipitously falls but its magnitude remains about 30% higher than in the annealed sample, likely reflecting its higher electrical conductivity and less defected structure. The annealed SLM-prepared skutterudite shows a lower thermal conductivity than the SHS-SPS-synthesized material. This is partly due to the highly refined grains of the SLM-processed skutterudite, but it also likely indicates the presence of micro-cracks that degrade its transport properties. Below about 450 oC, the unannealed SLM-processed sample has a much higher thermal conductivity than the annealed sample and its thermal



conductivity actually rises as a function of temperature. Above 450 oC, upon fusing of CoSb2 and Sb and the formation of the skutterudite CoSb3 phase, its thermal conductivity dramatically decreases below the SHS-SPS prepared sample but remains above the thermal conductivity of the annealed SLM-processed sample. The trend in the individual transport parameters is also reflected in the behavior of the thermoelectric figure of merit. Up to about 400 oC, the ZT values of the annealed SLM sample are much higher than that of the un-annealed sample. Above 450 oC, the two ZT values become substantially identical but are considerably below (only about one half) those of the SHS-SPS-fabricated skutterudite. We also note that we have tested the thermoelectric properties of the SHS-SLM-AN samples repeatedly and the results did not differ significantly. The highest ZT obtained on annealed SHS-SLM-fabricated skutterudite was 0.56 at 823 K.

3.3 Buk SLM-fabricated Material Attached with Ti Electrodes Figure 7 shows a back-scattered electron (BSE) image of the interface between CoSb2.85Te0.15 and Ti, and EDS elemental distribution maps. It follows that the CoSb2.85Te0.15 material is tightly bonded to the Ti substrate. At the interface of CoSb2.85Te0.15 and Ti, there is a transition layer about 70 μm thick. The EDS analysis shows this transition layer to be a Ti-Co-Sb ternary compound, and no Ti is detected in the CoSb2.85Te0.15 region, which indicates that the ternary compound layer effectively acts as a blocking layer to Ti diffusion.31 Figure 8 plots the resistance value of the CoSb2.85Te0.15/Ti interface as the position of the test probes is changing. There is a clear resistance jump in the vicinity of the interface, and the measured contact resistivity was about 37 μΩcm2. The interface contact resistivity is related to the pore structure near the interface diffusion layer, see Fig. 7(a), and possibly also due to the presence of micro-cracks.

4 Conclusion N-type CoSb2.85Te0.15 bulk skutterudite was obtained by SHS combined with the SLM technology. The phase structure, morphology, and thermoelectric transport


Page 10 of 29

Page 11 of 29 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


properties were systematically studied. The results show that when the laser energy density (EV) is within the range of 58~130 J/mm3, one can obtain good quality surfaces. In the process of SLM, non-equilibrium heating and cooling gives rise to a heterogeneous material consisting of mostly CoSb2 and pure Sb phases. However, a highly refined grain structure resulting from the rapid molten pool solidification facilitates a phase transformation at about 450 oC, whereby CoSb2 and Sb fuse to form the skutterudite CoSb3 phase in a short time. The annealed SLM-processed material has a lower thermal conductivity than the material prepared through the traditional sintering method. However, micro-cracks and other microscopic defects within the material have an adverse impact on its electronic properties. Although further improvements and optimization of the SLM process are needed to make the SLM-synthesized bulk skutterudite competitive with the best SHS-SPS-fabricated skutterudites, we have demonstrated that the SLM technology makes it feasible to print thermoelectric materials directly on metal substrates, such as Ti. These are important advances toward the realization of one-step rapid manufacturing of thermoelectric materials and the assembly of thermoelectric modules.

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected] ORCID Yonggao Yan: 0000-0002-0370-5917 Xinfeng Tang: 0000-0001-7555-919X Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes The authors declare no competing financial interest.

ACKNOWLEDGMENTS



We wish to acknowledge support from the Natural Science Foundation of China (Grant Nos. 51772232, 51521001, 51401153) and the 111 project of China (Grant No. B07040).

References (1) Lave, L. Real Prospects for Energy Efficiency in the United States; The National Academies Press: Washington DC, 2009. (2) Tritt, T. M. Holey and Unholey Semiconductors. Science 1999, 283, 804-805. (3) Chen, G.; Dresselhaus, M. S.; Dresselhaus. G.; Fleurial, J. P.; Caillat, T. Recent developments in thermoelectric materials. Int. Mater. Rev. 2013, 48, 45-66. (4) Zhang, Q. H.; Huang, X. Y.; Bai, S. Q.; Shi, X.; Uher, C.; Chen, L. D. Thermoelectric Devices for Power Generation: Recent Progress and Future Challenges. Adv. Eng. Mater. 2016, 18, 194-213. (5) Sales, B. C.; Mandrus, D.; Williams, R. K. Filled skutterudite antimonides: a new class of thermoelectric materials. Science 1996, 272, 1325. (6) Tan, G. J.; Liu, W.; Wang, S. Y.; Yan, Y. G.; Li, H.; Tang X. F.; Uher, C. Rapid preparation of CeFe4Sb12 skutterudite by melt spinning: rich nanostructures and high thermoelectric performance. J. Mater. Chem. A 2013, 1, 12657-12668. (7) Su, X. L.; Li, H.; Yan, Y. G.; Wang, G. Y.; Chi, H.; Zhou, X. Y.; Tang, X. F.; Zhang, Q. J.; Uher, C. Microstructure and thermoelectric properties of CoSb2.75Ge0.25−xTex prepared by rapid solidification. Acta Mater. 2012, 60, 3536-3544. (8) Li, H.; Tang, X. F.; Su, X. L.; Zhang, Q. J. Preparation and thermoelectric properties of high-performance Sb additional Yb0.2Co4Sb12+y bulk materials with nanostructure. Appl. Phys. Lett. 2008, 92, 202114. (9) Moure, A.; Rull-Bravo, M.; Abad, B.; Del Campo, A.; Muñoz-Rojo, M.; Aguirre, M. H.; Jacquot, A.; Fernandez, J. F.; Martin-Gonzalez, M. Thermoelectric Skutterudite/oxide nanocomposites: Effective decoupling of electrical and thermal conductivity by functional interfaces. Nano Energy 2017, 31, 393-402. (10) Li, G.; Bajaj, S.; Aydemir, U.; Hao, S.; Xiao, H.; Goddard, W. A.; Zhai, P.; Zhang, Q.; Snyder G. J. p-Type Co Interstitial Defects in Thermoelectric Skutterudite CoSb3 Due to the Breakage of Sb4-Rings. Chem. Mater. 2016, 28, 2172-2179. (11) Zheng, Z.; Wei, M.; Li, F.; Luo, J.; Ma, H.; Liang, G.; Zhang, X.; Fan, P. Optimization in fabricating skutterudite CoSb3 thermoelectric thin films. J. Mater. Sci-Mater. El. 2017, 28, 17221-17226. (12) Liang, T.; Su, X.; Yan, Y.; Zheng, G.; She, X.; You, Y.; Uher, C.; Kanatzidis, M. G.; Tang, X. Panoscopic approach for high-performance Te-doped skutterudite. NPG Asia Mater. 2017, 9, e352. (13) Wang, S. Y.; Tan, G. J.; Xie, W. J.; Zheng, G.; Li, H.; Yang, J. H.; Tang, X. F. Enhanced thermoelectric properties of Bi2(Te1−xSex)3-based compounds as n-type legs for low-temperature power generation. J. Mater. Chem. 2012, 22, 20943-20951. (14) Bell, L. E. Cooling, heating, generating power, and recovering waste heat with thermoelectric systems. Science 2008, 321, 1457-1461. (15) Zhao, L. D.; Zhang, B. P.; Li, J. F.; Zhang, H. L.; Liu, W. S. Enhanced thermoelectric and


Page 12 of 29

Page 13 of 29 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


mechanical properties in textured n-type Bi2Te3 prepared by spark plasma sintering. Solid State Sci. 2008, 10, 651-658. (16) Vandenbroucke, B.; Kruth, J. P. Selective laser melting of biocompatible metals for rapid manufacturing of medical parts. Rapid Prototyping J. 2007, 13, 148-159. (17) Kruth, J. P.; Leu, M. C.; Nakagawa, T. Progress in Additive Manufacturing and Rapid Prototyping. CIRP Ann-Manuf. Techn. 1998, 47, 525-540. (18) Martin, J. H.; Yahata, B. D.; Hundley, J. M.; Mayer, J. A.; Schaedler, T. A.; Pollock, T. M. 3D printing of high-strength aluminium alloys. Nature 2017, 549, 365-369. (19) Bremen, S.; Meiners, W.; Diatlov, A. Selective Laser Melting. Laser Technik Journal 2012, 9, 33-38. (20) Hassanin, H.; Finet, L.; Cox, S. C.; Jamshidi, P.; Grover, L. M.; Shepherd, D. E. T.; Addison, O.; Attallah, M. M. Tailoring selective laser melting process for titanium drug-delivering implants with releasing micro-channels. Additive Manufacturing 2018, 20, 144-155. (21) Almangour, B.; Grzesiak, D.; Yang, J. M. Rapid fabrication of bulk-form TiB2/316L stainless steel nanocomposites with novel reinforcement architecture and improved performance by selective laser melting. J. Alloy. Compd. 2016, 680, 480-493. (22) Cherry, J. A.; Davies, H. M.; Mehmood, S.; Lavery, N. P.; Brown, S. G. R.; Sienz, J. Investigation into the effect of process parameters on microstructural and physical properties of 316L stainless steel parts by selective laser melting. Int. J. Adv. Manuf. Tech. 2015, 76, 869-879. (23) Campanelli, S. L.; Contuzzi, N.; Ludovico, A. D.; Caiazzo, F.; Cardaropoli, F.; Sergi, V. Manufacturing and Characterization of Ti6Al4V Lattice Components Manufactured by Selective Laser Melting. Materials 2014, 7, 4803-4822. (24) Li, X. P.; Ji, G.; Chen, Z.; Addad, A.; Wu, Y.; Wang, H. W.; Vleugels, J.; Humbeeck, J. V.; Kruth, J. P. Selective laser melting of nano-TiB2 decorated AlSi10Mg alloy with high fracture strength and ductility. Acta Mater. 2017, 129, 183-193. (25) Sing, S. L.; Wiria, F. E.; Yeong, W. Y. Selective laser melting of lattice structures: A statistical approach to manufacturability and mechanical behavior. Robot. CIM-INT. Manuf. 2018, 49, 170-180. (26) Khorasani, A.; Gibson, I.; Goldberg, M.; Littlefair, G. Production of Ti-6Al-4V Acetabular Shell Using Selective Laser Melting: Possible Limitations in Fabrication. Rapid Prototyping J. 2017, 23, 110-121. (27) El-Desouky, A.; Carter, M.; Andre, M. A.; Bardet, P. M.; Leblanc, S. Rapid processing and assembly of semiconductor thermoelectric materials for energy conversion devices. Mater. Lett. 2016, 185, 598-602. (28) Mao, Y.; Yan, Y. G.; Wu, K. P.; Xie, H. Y.; Xiu, Z. K.; Yang, J. H.; Zhang, Q. J.; Uher, C.; Tang, X. F. Non-equilibrium synthesis and characterization of n-type Bi2Te2.7Se0.3 thermoelectric material prepared by rapid laser melting and solidification. Rsc Adv. 2017, 7, 21439-21445. (29) Liang, T.; Su, X. L.; Yan, Y. G.; Zheng, G.; Zhang, Q.; Chi, H.; Tang, X. F.; Uher, C. Ultra-fast synthesis and thermoelectric properties of Te doped skutterudites. J. Mater. Chem. A 2014, 2, 17914-17918. (30) Su, X. L.; Fu, F.; Yan, Y. G.; Zheng, G.; Liang, T.; Zhang, Q.; Cheng, X.; Yang, D. W.; Chi, H.; Tang, X. F.; Zhang, Q. J. Uher, C. Self-propagating high-temperature synthesis for compound thermoelectrics and new criterion for combustion processing. Nat. Commun. 2014, 2, 4908.



(31) Fan, J. F.; Chen, L. D.; Bai, S. Q.; Shi, X. Joining of Mo to CoSb3 by spark plasma sintering by inserting a Ti interlayer. Mater. Lett. 2004, 58, 3876-3878.


Page 14 of 29

Page 15 of 29 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 1. (a) SEM image and (b) particle size distribution of the SHS-prepared CoSb2.85Te0.15 powder.

Figure 2. SEM images of four typical surface morphologies of SLM-prepared layers: (a) vaporization; (b) flat; (c) unstable; and (d) balling. (e) SLM processing window for a single layer of CoSb2.85Te0.15. The SLM processing parameters are d=0.06 mm, h=0.04 mm, P=4-18

W, and v=30-300 mm/s.



Figure 3. XRD patterns of : (a) CoSb2.85Te0.15 bulk prepared by SLM; (b) SLM-prepared bulk samples were annealed in vacuum for 4 hours at 450 oC, and SHS-prepared CoSb2.85Te0.15 powder after ball milling.

Figure 4. (a) SEM image of a cross section of the SHS-SPS sample; (b) SEM image of the as-prepared SHS-SLM sample; (c) SEM image of the SHS-SLM sample annealed at 450 oC for 4 hours; (d) the same as (c) but magnified, showing clearly the presence of a micro-crack.


Page 16 of 29

Page 17 of 29 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 5. (a) Spatially resolved Seebeck coefficient of the as-prepared SLM sample at room temperature; (b) distribution of its Seebeck coefficient. (c,d) the same, but the sample is annealed at 450 ℃ for 4 hours.



Figure 6. Temperature dependence of thermoelectric properties of the SHS-SLM sample, the SHS-SLM-AN sample and the SHS-SPS sample: (a) electrical conductivity, (b) Seebeck coefficient, (c) thermal conductivity, and (d) ZT.


Page 18 of 29

Page 19 of 29 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 7. (a) BSI image of the interface of CoSb2.85Te0.15 with Ti. Note the presence of pores near the interface. (b) – (d) EDS elemental maps of the interface of CoSb2.85Te0.15 with Ti.



Figure 8. Electrical resistance as a function of the distance across the interface of CoSb2.85Te0.15 with Ti.


Page 20 of 29

Page 21 of 29 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


highlight 150x46mm (300 x 300 DPI)



Figure 1. (a) SEM image and (b) particle size distribution of the SHS-prepared CoSb2.85Te0.15 powder. 80x28mm (300 x 300 DPI)


Page 22 of 29

Page 23 of 29 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 2. SEM images of four typical surface morphologies of SLM-prepared layers: (a) vaporization; (b) flat; (c) unstable; and (d) balling. (e) SLM processing window for a single layer of CoSb2.85Te0.15. The SLM processing parameters are d=0.06 mm, h=0.04 mm, P=4-18 W, and v=30-300 mm/s. 80x40mm (300 x 300 DPI)



Figure 3. XRD patterns of : (a) CoSb2.85Te0.15 bulk prepared by SLM; (b) SLM-prepared bulk samples were annealed in vacuum for 4 hours at 450 ℃, and SHS-prepared CoSb2.85Te0.15 powder after ball milling. 80x35mm (300 x 300 DPI)


Page 24 of 29

Page 25 of 29 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 4. (a) SEM image of a cross section of the SHS-SPS sample; (b) SEM image of the as-prepared SHSSLM sample; (c) SEM image of the SHS-SLM sample annealed at 450 ℃ for 4 hours; (d) the same as (c) but magnified, showing clearly the presence of a micro-crack. 109x82mm (300 x 300 DPI)



Figure 5. (a) Spatially resolved Seebeck coefficient of the as-prepared SLM sample at room temperature; (b) distribution of its Seebeck coefficient. (c,d) the same, but the sample is annealed at 450 ℃ for 4 hours. 80x67mm (300 x 300 DPI)


Page 26 of 29

Page 27 of 29 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 6. Temperature dependence of thermoelectric properties of the SHS-SLM sample, the SHS-SLM-AN sample and the SHS-SPS sample: (a) electrical conductivity, (b) Seebeck coefficient, (c) thermal conductivity, and (d) ZT. 80x67mm (300 x 300 DPI)



Figure 7. (a) BSI image of the interface of CoSb2.85Te0.15 with Ti. Note the presence of pores near the interface. (b) – (d) EDS elemental maps of the interface of CoSb2.85Te0.15 with Ti. 99x72mm (300 x 300 DPI)


Page 28 of 29

Page 29 of 29 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 8. Electrical resistance as a function of the distance across the interface of CoSb2.85Te0.15 with Ti. 50x44mm (300 x 300 DPI)


Fabrication and Thermoelectric Properties of n-Type CoSb2.85Te0.15

Recommend Documents