Mo

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Functional Inorganic Materials and Devices

Low contact resistivity and interfacial behavior of p-type NbFeSb/Mo thermoelectric junction Jiajun Shen, Zhenyi Wang, Jing Chu, Shengqiang Bai, Xin-Bing Zhao, Lidong Chen, and Tie-Jun Zhu ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 26 Mar 2019 Downloaded from http://pubs.acs.org on March 30, 2019

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 32 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

Low Contact Resistivity and Interfacial Behavior of p-Type NbFeSb/Mo Thermoelectric Junction Jiajun Shen,† Zhenyi Wang,† Jing Chu,‡ Shengqiang Bai,‡ Xinbing Zhao,† Lidong Chen,‡ Tiejun Zhu*,† †

State Key Laboratory of Silicon Materials, and School of Materials Science and Engineering,

Zhejiang University, Hangzhou 310027, China ‡

State Key Laboratory of High Performance Ceramics and Superfine Microstructure, Shanghai

Institute of Ceramics, Chinese Academy of Sciences, Shanghai 200050, China KEYWORDS: thermoelectric materials, half-Heusler compounds, electrode, contact resistivity, interfacial microstructure, Ohmic contact

ABSTRACT: Half-Heusler compounds are a class of promising thermoelectric (TE) materials for power generation. However, the large contact resistivity at the interface between TE legs and metal electrode of the TE device seriously hinders the full play of the material performance. Here we report an Ohmic contact for the junction of p-type Nb0.8Ti0.2FeSb and Mo electrode with a low contact resistivity of 1, unique mechanical properties and excellent heat stability11-16. The room temperature electrical conductivity of HH compounds is usually higher than 105 S/m17-20. Therefore, the contact resistivity should be on the order of 1 μΩcm2 aiming to minimize the degeneration of output power from a TEG. However, most of the contacts between HH compounds and metal electrodes cannot meet the requirement21, 22. For example, the junction interface between p-type ZrCoSb and Ag shows a contact resistivity as high as 100 μΩcm2,23 while the n-type TiNiSn/Ag junction even presents a value of 350 μΩcm2. 24

Recently, a high zT of ~1.5 for the p-type NbFeSb HH compound was reported12. However, the experimental conversion efficiency of 6.2% is still an unsatisfactory TE performance, which is half of the theoretical calculated conversion efficiency. The subsistent electrical and thermal contact resistances between the semiconductor and metal electrode give rise to the degradation of the output power. Therefore, it is impending to do some deep research on how to reduce the contact resistivity between TE semiconductor and metal electrode in order to give full play to the potential material properties. Up to now, the formation of contact resistivity has been mainly attributed to the change of interface microstructure25-31. Recently, a report about ab initio calculation on the contact resistivity between HH and Ag electrode shows that the Schottky potential generated by the metal-semiconductor contact should also be a significant factor to influence the charge transport at the contact interface32. However, the absence of the fundamental knowledge on interface physical property dramatically hinders the development of TE semiconductor industry. Actually, the issue about interface contact has already been deeply studied in the field of silicon based

ACS Paragon Plus Environment

3

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 32

semiconductor industry, such as MOSFET33, diode rectifier34, 35 and so on. An Ohmic contact is often required to reduce the contact resistivity between Si substrate and metal electrode36, 37. To realize this type of contact, many methods have been successfully applied, such as the matching of work function between semiconductor and metal38, and tunneling effect by the high doping level in the semiconductor39. For TE semiconductors, the theory of metal-semiconductor contact should also be applicable. The biggest difference between TE semiconductors and Si based semiconductors is the relatively lower band gap for TE materials, compared with silicon materials4,

40-42.

Actually, the

relationship between contact resistivity and doping level has already been investigated in the contact of n-type Bi2Te3 and Ni electrode, where the enhanced tunneling electric current due to the surface ion implantation in the n-type Bi2Te3 semiconductor successfully reduces the contact resistivity37. Therefore, it is worthwhile to do some research on the interfacial carrier transport mechanism in the contact of TE semiconductors and metal electrodes. There are generally two methods to realize the joint between HH compounds and metal electrode. The first one is using soldering or brazing materials to make the connection, and the other is that directly connect HH alloys with electrode by SPS or hot-pressing. A good connection can be obtained due to the wetting of brazing materials on both TE materials and metal electrodes. However, the brazing materials also give rise to the strong atomic diffusion and interfacial reactions among the brazing materials, TE materials and electrode, especially for TE devices applied at high temperature. For example, compared with the fast hot-press method with no brazing materials, the interfacial microstructure between n-type TiNiSn and electrode Ag is more complicated and possesses a high contact resistivity (~1000 μΩcm2). Comparatively, the direct joint between n-type TiNiSn and electrode Ag without brazing materials presents a


4

Page 5 of 32 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


relatively lower value of 339 μΩcm2.24 However, this value is still too high to meet the requirement of low contact resistivity. In this work, we choose Mo as the metal electrode and successfully prepare the junction of ptype Nb0.8Ti0.2FeSb HH compound and Mo electrode using the directly connection method without any brazing materials, and the interfacial behavior is analyzed. An Ohmic contact is realized in the junction. The interfacial carrier transport is dominated by field emission and consequently a strong tunneling electric current is obtained due to the high doping level and relatively low dielectric constant for p-type Nb0.8Ti0.2FeSb semiconductor. Therefore, a small contact resistivity of < 1 μΩcm2 is obtained for the as-sintered interface between Nb0.8Ti0.2FeSb and Mo. The heat endurance is characterized by a long time heat treatment. The contact resistivity rises up to 18.4 μΩcm2 after 32 days’ aging due to the increasing content of Nb3Ti, the crack appearing at both sides of Nb3Ti interlayer, and the newly formed FeSb2 interlayer. EXPERIMENTAL PROCEDURES Preparation The p-type Nb0.8Ti0.2FeSb samples were prepared by elemental Nb (foil, 99.98%), Ti (rod, 99.99%), Fe (piece, 99.97%) and Sb (block, 99.999%). The raw materials were weighted according to the stoichiometric ratio and then melted together for three times by levitation melting to obtain alloys of uniform components. The as-melted ingots were pulverized into powders and ball milled (Mixer Mill MM200, Retsch) for 3 hours to refine grain under argon protection. The Nb0.8Ti0.2FeSb powders and the Mo powders were sequentially loaded into a cylinder-shaped graphite die and sintered together by spark plasma sintering (SPS-1050, Sumitomo Coal Mining Co.) at 1073 K/1113 K/1153 K for 20min under 65 MPa in the vacuum environment to obtain the junction of p-type Nb0.8Ti0.2FeSb and Mo electrode. Several as-


5


Page 6 of 32

sintered samples were annealed at 1073 K for 8/16/24/32 days to obtain the interface microstructure undergoing different aging times. Then the samples were cut into cuboids to expose the interface. TE performance measurements The figure of merit zT for the Nb0.8Ti0.2FeSb samples with different sintering temperatures was calculated for comparison. The electrical conductivity and Seebeck coefficient were measured by a commercial LSR-3 system (Linseis, Germany). The thermal conductivity was calculated by κ = DρCP, where D, ρ, and CP represent thermal diffusivity, the sample density and the specific heat, respectively. The thermal diffusivity was measured by a laser flash method on Netzsch LFA457 instrument. The specific heat was calculated with the formula: CP = Cph,H + CD, where Cph,H is the Debye heat capacity, and CD is the item related to thermal expansion43. The thermal expansion coefficient for Nb0.8Ti0.2FeSb was measured on Netzsch DIL 402 PC instrument. The sample’s permittivity was measured with a dielectric constant tester (Turnkey Concept 50, Germany). Characterizations The elemental composition and back scattering image were detected by electron probe microanalysis (EPMA, JEOL JXA - 8100). The element mapping for the interface microstructure was observed by energy disperse spectroscopy (EDS). The contact resistivity was measured by four-probe method26. The work function was measured by Ultraviolet Photoelectron Spectroscopy (UPS, Thermo ESCALAB 250). The I-V characteristic was measured using a semiconductor parameter analyzer (Agilent E5270B). Mechanical property measurements


6

Page 7 of 32 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 bending strength was measured by an electromechanical universal testing machine (CMT 5205). An MH-5 microhardness tester was utilized to measure the microhardness of Nb0.8Ti0.2FeSb employing a load of 100 g with a rate of 0.1 N/s and a loading time of 10 s. The longitudinal (vl) and transverse sound velocities (vt) were measured by an ultrasonic pulse-echo method at room temperature to evaluate the elastic moduli E. The input was generated by a Panametrics 5052 pulser/receiver with the filter at 0.03 MHz. The response was recorded via a Tektronic TDS5054B-NV digital oscilloscope. The elastic moduli E is calculated by the formula:

E

3vt2 vl2  4vt2 vl2  vt2 , where ρ is the sample density.

RESULTS AND DISCUSSION TE performance The TE properties of the p-type Nb0.8Ti0.2FeSb samples with different SPS sintering temperatures are shown in Figure 1. Both the electrical conductivity and Seebeck coefficient agree well with the previous data in Ref. 44. The thermal conductivity of Nb0.8Ti0.2FeSb here shows a slight increase, mainly due to the longer sintering time, which is about twice as much as that in Ref. 44. The longer sintering time is favorable for grains growth and weakening the grain boundary scattering. However, the increase of the thermal conductivity is not distinct and no obvious deterioration of zT value is found for p-type Nb0.8Ti0.2FeSb.


7


Page 8 of 32

Figure 1. Temperature dependences of (a) electrical conductivity, Seebeck coefficient, (b) thermal conductivity and zT value for p-type Nb0.8Ti0.2FeSb. The solid line represents the data from the Ref. 44. Interfacial microstructure Figure 2 shows the microstructure of the junction interface between p-type Nb0.8Ti0.2FeSb and the Mo electrode. A layer of FeMo alloy with a thickness of ~5 μm is formed after SPS sintering at the side of Mo electrode. The existence of a minor interface reaction is conducive to a strong connection between the metal electrode and the TE semiconductor. The EPMA elemental composition analysis illustrates that the content of Fe is about 60%, which refers to the μ phase according to the Fe-Mo phase diagram45. Because the element Fe reacts with Mo, the Nb exceeds the equilibrium stoichiometric ratio of NbFeSb, and segregates and reacts with Ti to form Nb3Ti at the side of p-type NbFeSb. The Nb3Ti forms a crumb structure, which is uniformly distributed in the NbFeSb side. The thickness of the mixing layer of Nb3Ti crumb and NbFeSb is about 25 μm and there are almost no other Nb3Ti crumbs found outside the area. Besides, a slight increase of Sb content is found at the interface between FeMo and NbFeSb according to the EPMA line scanning in Figure 2(b).


8

Page 9 of 32 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. (a) Back-scattering electron image and (b) EPMA line scanning of the interface between as-sintered p-type Nb0.8Ti0.2FeSb and Mo. To further confirm the compositions and microstructure at the interface, the EDS mapping for the as-sintered sample is conducted. The full view of the Nb0.8Ti0.2FeSb/Mo junction is shown in Figure 3(a). The obvious aggregation of Nb and Ti can be observed near the interface, which agrees well with the results of the EPMA analysis. The concentration of Fe shows a slight decrease at the region where Nb3Ti crumbs locate. However, it is difficult to distinguish whether there is a FeMo alloy layer because the thickness of FeMo phase is only 5μm. Therefore, a magnified EDS mapping is performed to reveal the detailed interface structure. As shown in Figure 3(b), the content of Mo shows a decrease when getting close to the interface and simultaneously the Fe concentration increases. Therefore, it can be concluded that the FeMo layer indeed exists at the interface of Nb0.8Ti0.2FeSb/Mo. Besides, the element Sb presents a spontaneous aggregation at the interface between FeMo and NbFeSb, which is consistent with the analysis of line scanning. It is worth noting that the Mo electrode has not been sintered totally. The surface of the electrode looks rugged, as shown in Figure 3(a), mainly because that the melting point of Mo is


9


Page 10 of 32

as high as 2896 K, much higher than the sintering temperature of 1113 K. However, at the interface region, the Mo electrode is smoother and denser because the diffusion of Fe into Mo efficiently decreases the melting point of the alloy and facilitates the sintering process.

Figure 3. EDS mapping of (a) the full view and (b) the detail view of the interface between assintered p-type Nb0.8Ti0.2FeSb and Mo. Contact resistivity


10

Page 11 of 32 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


To clarify the charge transport at the interface, the contact resistivity is show in Figure 4. By linearly fitting the Nb0.8Ti0.2FeSb part of the electrical resistance curve, an electrical conductivity of 4.79×105 S/m at room temperature is obtained, which agrees well with the electrical properties measurement for p-type Nb0.8Ti0.2FeSb in Figure 1(a) indicating that the contact resistivity measurement is reliable. However, there is a big contrast of the electrical conductivity for Mo between the value (~2.06×106 S/m) derived here and the data from the Ref. 46, due to the nondense sintering of Mo during the SPS process. At the joint part of the measured electrical resistance curve, which is shown as the shaded area in Figure 4(a), no abrupt change of the slope is observed. In consideration of the total thickness (~30 μm) of the interfacial reaction layer, the maximum difference of the electrical resistance should be less than 1 μΩcm2, implying that the contact resistivity ρc of the interface between the Nb0.8Ti0.2FeSb and Mo electrode is tiny (

Mo

Recommend Documents