为了正常的体验网站,请在浏览器设置里面开启Javascript功能!
首页 > 1693_ftp

1693_ftp

2013-05-08 4页 pdf 504KB 12阅读

用户头像

is_315072

暂无简介

举报
1693_ftp www.advmat.de www.MaterialsViews.com C O M M U N IC A TIO N Ye Sun , Mogens Christensen , Simon Johnsen , Eryun Zhang , Anders E. C. Palmqvist , Jørgen Low-Cost High-Performance Z for Thermoelectric Application Risø N...
1693_ftp
www.advmat.de www.MaterialsViews.com C O M M U N IC A TIO N Ye Sun , Mogens Christensen , Simon Johnsen , Eryun Zhang , Anders E. C. Palmqvist , Jørgen Low-Cost High-Performance Z for Thermoelectric Application Risø National Laboratory for Sustainable Energy Rising demand for energy effi ciency and green solutions has led to a strong global interest in thermoelectric (TE) devices. However, TE power generation (heat to electricity) still has few commercial applications [ 1–4 ] since no cheap, stable, non-toxic and high-performing materials made of abundant elements are known. [ 5 , 6 ] The effi ciency of TE materials is evaluated using the dimensionless fi gure of merit, zT = S 2 T / ρ κ , where S is the Seebeck coeffi cient, T is the absolute temperature, ρ is the elec- tice thermal conductivity together with the measured PFs and b 3 n se T ey ve lf- nt re Technical University of Denmark Roskilde, DK-4000, Denmark Dr. Y. Ma , Prof. A. E. C. Palmqvist Department of Chemical and Biological Engineering Chalmers University of Technology Göteborg, SE-41296, Sweden Dr. M. Sillassen , Dr. E. Zhang , Prof. J. Bøttiger Department of Physics and iNANO Aarhus University Aarhus C, DK-8000, Denmark DOI: 10.1002/adma.201104947 κ e ’s, zT was estimated to be ∼ 1.15 and ∼ 0.6 at 573 K for Zn 4 S and ZnSb, respectively. These values are outstanding eve under highly conservative estimations since the true κ L of the fi lms would be much lower than the bulk values, and thus z is higher. These thin fi lms for the fi rst time fulfi ll all the k material requirements for commercial TE application, and ha signifi cant potential to be used in TE sensors, coolers or se powered micro-devices. Three types of zinc antimonide fi lms with differe Zn:Sb atomic ratios of (A) ∼ 50:50, (B) ∼ 54:46 and (C) ∼ 58:42 we Dr. Y. Sun , Dr. M. Christensen , Dr. S. Johnsen , Prof. B. B. Iversen Center for Materials Crystallography Department of Chemistry and iNANO Aarhus University Aarhus C, DK-8000 Denmark E-mail: bo@chem.au.dk Dr. N. V. Nong Fuel Cells and Solid State Chemistry Division trical resistivity, and κ is the thermal conductivity consisting of a contribution from the charge carrier κ e and the lattice κ L . TE materials with zT > 1 are typically required to give a TE device a reasonable effi ciency, but in fact for commercial applications it is equally important that the TE material is made of cheap and abundant elements with a cost-effective synthesis method. Although most research concerns bulk materials for high- power applications such as the recovery of waste heat from a car exhaust, attractive low-power applications of thin fi lms (e.g., as sensors) or self-powered microdevices are evident. The inherent phonon scattering due to the nanostructure of thin fi lms can be exploited to enhance TE properties. [ 7–11 ] Extremely high TE fi gures of merit have been reported for superlattices, [ 12 , 13 ] yet their complicated and expensive production makes large-scale, widespread commercialization problematic. The β phase of Zn 4 Sb 3 is a promising p -type TE material due to its exceptionally low lattice thermal conductivity. [ 14 , 15 ] The highest zT value of Zn 4 Sb 3 was reported to be 1.3 at 670 K, [ 16 ] although later studies have shown that zT degrades signifi cantly with repeated thermal cycling due to thermal decomposition of © 2012 WILEY-VCH Verlag GmAdv. Mater. 2012, 24, 1693–1696 Ngo V. Nong , Yi Ma , Michael Sillassen , Bøttiger , and Bo B. Iversen * inc Antimonide Thin Films s the structure. [ 6 ] In comparison with other high performance TE materials working in the same temperature range, which are predominantly tellurium based, [ 3 , 5 ] Zn 4 Sb 3 has overwhelming advantages in commercial applications owing to the low cost, and the relatively large abundance and low toxicity of Zn and Sb. The ZnSb phase is also a good p -type TE material. [ 14 , 17 , 18 ] with a large power factor (PF = S 2 / ρ ) comparable to Zn 4 Sb 3 , but a high κ restricts zT to ∼ 0.6 at 500 K. [ 14 ] To improve practical application of Zn 4 Sb 3 and ZnSb, numerous studies have exam- ined the TE properties and stability of zinc antimonide bulk materials. [ 6 , 14–19 ] However, few studies have explored the growth and TE properties of zinc antimonide fi lms, and only impure fi lms were obtained. [ 20 , 21 ] The Zn-Sb binary phase diagram has attracted much atten- tion and the stability of the different known phases has been scrutinized. [ 6 , 22 , 23 ] Recently, a new Zn 8 Sb 7 phase was identifi ed experimentally [ 24 ] and theoretical calculations have been used to estimate the physical properties and stability of this phase. [ 25 ] In the present study ZnSb fi lms as well as two completely new meta-stable crystalline phases of zinc antimonide were directly deposited on silica substrates by a co-sputtering method. After post-annealing treatments at 573 K, the two new phases trans- formed into Zn 4 Sb 3 + ZnSb and single phase Zn 4 Sb 3 fi lms, respectively. The phase transitions of the new fi lms were con- fi rmed directly by in situ powder X-ray diffraction (PXRD) and further evidenced through the changes in the electrical proper- ties. Thermoelectric characterization along the in-plane direc- tion revealed that our Zn 4 Sb 3 and ZnSb fi lms exhibit very large PF values, which are comparable to the best reported results of the bulk materials, [ 14 , 15 ] but unlike the bulk materials, [ 19 , 26 ] they also have good thermal stability up to 573 K. This implies a remarkable application potential of these fi lms in TE power generators and sensors. Due to surface and grain-boundary scattering of phonons, TE nanocrystalline thin fi lms potentially have much reduced κ compared with bulk samples, and thus potentially much enhanced z T values. [ 27 , 28 ] We have not been able to measure κ of the present fi lms, but using the bulk lat- bH & Co. KGaA, Weinheim 1693wileyonlinelibrary.com 16 www.advmat.de C O M M U N IC A TI O N 94 wileyonlinelibrary.com © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Wein produced by controlling the growth parameters (Figure S1). The fi lm growth was carried out a very large number of times and the synthesis is highly reproducible. PXRD analyses revealed that the type A samples are single phase ZnSb fi lms ( Figure 1 a), whereas the PXRD patterns of the type B (Figure 1 b) and type C (Figure 1 c) samples were totally different from each other and could not be indexed to any of the well known Zn-Sb phases: Zn 4 Sb 3 , Zn 8 Sb 7 , ZnSb, Zn 3 Sb 2 , Zn, or Sb. This strongly sug- gests formation of two new meta-stable crystalline phases of zinc antimonide. The type A samples were produced at 573 K. Formation of a fl at and continuous fi lm was confi rmed by the top-view scan- ning electron microscope (SEM) image presented in the inset of Figure 1 a. The ZnSb (211) peak dominates the whole PXRD pattern (Figure 1 a), suggesting growth of a < 211 > -textured ZnSb fi lm. These single phase ZnSb fi lms keep their crystal structure under an annealing treatment at 573 K, only the peak intensity in the PXRD patterns is slightly enhanced. Growth of ZnSb fi lms by other methods such as metalorganic chemical vapor deposi- tion [ 29 ] or thermal evaporation [ 21 ] have been reported. However, these studies did not provide clear PXRD patterns of the ZnSb fi lms and/or present rational TE properties. Here, growth of textured and single phase ZnSb fi lms by a sputtering method was demonstrated by PXRD and energy dispersive X-ray analysis (EDX), and their excellent TE properties will be discussed below. patterns. In contr C) are single phas fi lm. The evolutio were followed by evolves into a mi while phase C tra and e. To further prob antimonide fi lms perature ρ and c annealed fi lms w and n of the as-gr 1.9 × 10 19 cm − 3 , slightly changed ρ 1.5 × 10 19 cm − 3 . T of ∼ 650 000 m Ω than that of the an C fi lms, the an from 43.0 m Ω c 3.6 × 10 19 cm − 3 to Figure 2 a–c sh (type A) with a t from cross-section The TE measure cycles, i.e. RT → Figure 1 . XRD patterns of a) a type A sample (ZnSb), b) a type B sample and an annealed type B sample (Zn 4 Sb 3 + ZnSb), and c) a type C sample and an annealed type C sample (Zn 4 Sb 3 ). Top view SEM images of the corresponding samples are presented in the inset of the fi gure. The scale bar is applied to all the SEM images in this fi gure. d) and e) In situ PXRD data from annealing of a phase C fi lm showing the phase transformation to Zn 4 Sb 3 . The large peak at 28–29 ° stems from the graphite dome of the high temperature furnace. is close to that of the type C fi lms (58:42), the structure evolvement of type C fi lms to Zn 4 Sb 3 fi lms during annealing appears to be purely a structural transition. Zhang et al. have reported on the growth of Zn 4 Sb 3 fi lms by sputtering Zn and Sb targets at room temperature ( RT ) and then annealing in Ar atmosphere at 573–673 K. [ 20 ] The obtained fi lms contain zinc impurities and do not have preferred orientation according to the PXRD ast, the present Zn 4 Sb 3 fi lms (annealed type e and with the c -axis in the plane of the thin n of phase transitions in type B and C fi lms in situ PXRD, showing clearly that phase B xture of ZnSb and Zn 4 Sb 3 (See Figure S2), nsforms into single phase Zn 4 Sb 3 , Figure 1 d e the phase transitions of the as-grown zinc during the annealing process, room tem- arrier concentration ( n ) of as-deposited and ere measured (Table S1). The measured ρ own ZnSb fi lms (type A) are 2.6 m Ω cm and respectively. The annealing treatment only and n of this ZnSb fi lm to 3.1 m Ω cm and he as-grown type B fi lm has a high ρ value cm, which is fi ve orders of magnitude larger nealed type-B fi lm (4.9 m Ω cm). For the type nealing treatment led to a decrease of ρ m to 3.4 m Ω cm and increase of n from 2.8 × 10 20 cm − 3 . ows the TE properties of a typical ZnSb fi lm hickness of ∼ 450 nm, which was measured al SEM image of the fi lm (inset of Figure 2 b). ments were subjected to two measurement 573 K → RT → 573 K → RT . It can be seen www.MaterialsViews.com The type B samples were produced at 523 K, but they are not stable at T ≥ 573 K. An annealing treatment under Ar atmos- phere at 573 K for 2 h results in a phase transition of the type B samples into Zn 4 Sb 3 + ZnSb as evident from the PXRD pattern shown in Figure 1 b, where all the PXRD peaks of the annealed sample can be assigned to Zn 4 Sb 3 and/or ZnSb. The annealed type B fi lms are continuous and have relatively smooth top surfaces, as seen in the SEM image in the inset of Figure 1 b. No detectable changes of the Zn:Sb atomic ratio were observed from the as-deposited and annealed type B fi lms. The type C samples, which were produced at 473 K, are also unstable at an annealing temperature of 573 K. Their crystalline metastable phase evolves to pure Zn 4 Sb 3 after a 2 h annealing treatment, as seen in the PXRD pattern in Figure 1 c. The dominating (030) peak implies formation of a Zn 4 Sb 3 fi lm with the c -axis in the plane of the fi lm. Since the Zn:Sb atomic ratio of Zn 4 Sb 3 (4:3) heim Adv. Mater. 2012, 24, 1693–1696 www.advmat.de C O M M U N IC A TIO N was subjected to three TE property measure- ment cycles, and the annealed sample shows a stable TE performance (Figure S3) with PF values consistent with the results shown in Figure 2 f. Figure 3 shows the estimated zT versus temperature for the ZnSb and Zn 4 Sb 3 fi lms using κ L of the corresponding bulk mate- rials. [ 14 – 16 ] To obtain κ , κ e was calculated from measured values of ρ , and added to literature values of bulk κ L . The highest zT of the ZnSb and Zn 4 Sb 3 fi lms were estimated to be ∼ 0.57 and 1.15 at 573 K, respectively. These values are conservatively estimated since the thin fi lms should have a much reduced κ due to their small crys- tallite sizes (about 40 nm for both Zn 4 Sb 3 and ZnSb fi lms estimated by the Scherrer method) and thus the zT values would be higher. The present materials fulfi ll the key ( T ), e) S ( T ), and ge of RT to 573 K lms are shown in Cycle 1 Cycle 2 film 0 500 550 600 re / K www.MaterialsViews.com that the TE properties of the fi lm become stable after the fi rst heating cycle (Figure 2 a–c). In general, ρ slightly decreases, while S ( p -type conducting behavior) increases with increasing temperature. At 573 K, the measured ρ and S values of the ZnSb fi lm were 2.9 m Ω cm and 219 μ V K − 1 , respectively, yielding a PF value of about 16.5 W m − 1 K − 2 , which is con- sistent with the reported results for ZnSb single crystals. [ 30 , 31 ] This indicates a good crystal quality of the ZnSb fi lms. Pre- viously, the TE properties of ZnSb fi lms produced by the MOCVD method have been reported, [ 21 ] but the S ( T ) and ρ are not consistent with the results for ZnSb single crystals. [ 30 , 31 ] In addition, no XRD pattern was provided to confi rm the forma- tion of ZnSb fi lms. Since the Zn 4 Sb 3 fi lms can be obtained by annealing type C fi lms at 573 K for several hours, type C fi lms should par- tially or fully evolve into Zn 4 Sb 3 fi lms during TE property measurements. In Figure 2 d and e, ρ ( T ) and S ( T ) of a type C Figure 2 . a) ρ ( T ), b) S ( T ), and c) PF( T ) of a type A sample (ZnSb). d) ρ f) PF( T ) of a type C fi lm. Two measurement cycles in the temperature ran were performed. Cross-sectional SEM images of the type A and the type C fi the inset of b) and e), respectively. 1 2 3 4 120 160 200 240 ρ / m Ω cm Cycle 1 Cycle 2 ZnSb filma) b) S / μ V/ K 300 350 400 450 500 550 600 12 15 18 c) Temperature / K PF / 1 0-4 W m - 1 K - 2 1 10 100 50 100 150 200 ρ / m Ω cm Type C d) e) S / μ V/ K 300 350 400 45 0 4 8 12 f) Temperatu PF / 1 0-4 W m - 1 K - 2 © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, WeinAdv. Mater. 2012, 24, 1693–1696 fi lm are shown. The fi lm thickness is about 540 nm, which was measured from the cross-sectional SEM image presented in the inset of Figure 2 e. No change of the fi lm thickness was observed before and after the annealing treatment. On heating, ρ ( T ) of the fi rst measurement cycle decreased from 37 m Ω cm at RT to 16 m Ω cm at 460 K, and then quickly dropped to ∼ 3 m Ω cm at T > 460 K. This is attributed to the transition from phase C to Zn 4 Sb 3 , which is in good agreement with the in situ PXRD data with an onset at about 490 K (Figure 1 d). The measured S (553 K) of the Zn 4 Sb 3 fi lms is about 190 μ VK − 1 , and they have a maximum PF value of ∼ 12 × 10 − 4 Wm − 1 K − 2 at 553 K (Figure 2 f). It has been suggested that Zn 4 Sb 3 could be unstable at T ≥ 523 K [ 6 , 32 ] and single-phase Zn 4 Sb 3 nanoparticles are not stable at T ≥ 473 K. [ 33 ] However, the stable PF of our Zn 4 Sb 3 fi lm during the two measurement cycles implies a good phase stability of Zn 4 Sb 3 at T ≤ 573 K in a low pressure inert atmos- phere. In order to further confi rm the stability of the Zn 4 Sb 3 fi lms synthesized in this work, type C samples were annealed at 573 K for 10 h. After annealing, PXRD confi rmed that the fi lms were single phase Zn 4 Sb 3 . Subsequently the annealed sample Figure 3 . zT of ZnS culated from meas κ L . In the literature of bulk Zn 4 Sb 3 hav b and Zn 4 Sb 3 fi lms. In order to estimate κ , κ e was cal- ured values of ρ , and added to literature values of bulk two different values for the lattice thermal conductivity e been given leading to slightly different zT curves. requirements for commercial TE device application: cheap elements, cheap fabrication method, high performance and thermal stability, and they appear ideally suited for TE devices based on thin fi lms. Experimental Section Zinc antimonide fi lms were deposited on polished fused silica substrates by co-sputtering of a Zn target and a specifi cally prepared Zn/Sb compound target. The sputtering system was equipped with two independent magnetron sources with a target-to-substrate distance of 10 cm. The chamber base pressure was approximately 1 × 10 − 5 Pa. Argon (purity 99.9996%) was used as sputter-gas at a fl ow rate of 10 sccm and the Ar pressure in the chamber was fi xed at 0.6 Pa. In order to control the thin fi lm growth, different magnetron power of 5–12 W, substrate temperatures of 473–573 K and deposition times of 1–3 h were applied. The post-annealing treatment of the fi lm samples was performed with an Ar pressure of ∼ 1.0 Pa at 573 K using the vacuum chamber of the 1695wileyonlinelibrary.comheim 169 www.advmat.de www.MaterialsViews.com C O M M U N IC A TI O N [ 9 ] T. C. Harman , P. J. Taylor , D. L. Spears , M. P. Walsh , J. Electron. Mater. 2000 , 29 , L1 . [ 10 ] L. Yan , M. Shao , H. Wang , D. Dudis , A. Urbas , B. Hu , Adv. Mater. 2011 , 23 , 41204 . [ 11 ] C. J. Vineis , A. Shakouri , A. Majumdar , M. G. Kanatzidis , Adv. Mater. 2010 , 22 , 3970 . [ 12 ] R. Venkatasubramanian , E. Siilvola , T. Colpitts , B. O’Quinn , Nature 2001 , 413 , 597 . [ 13 ] T. C. Harman , P. J. Taylor , M. P. Walsh , B. E. LaForge , Science 2002 , 297 , 2229 . [ 14 ] G. J. Snyder , M. Christensen , E. Nishibori , T. Caillat , B. B. Iversen , Nat. Mater. 2004 , 3 , 458 . [ 15 ] E. S. Toberer , P. Rauwel , S. Gariel , J. Tafto , G. J. Snyder , J. Mater. Chem. 2010 , 20 , 9877 . [ 16 ] T. Caillat , J. P. Fleurial , A. Borshchevsky , J. Phys. Chem. Solids 1997 , 58 , 1119 . [ 17 ] P. H. M. Bottger , K. Valset , S. Deledda , T. G. Finstad , J. Electron. Mater. 2010 , 39 , 1583 . sputtering system. The detailed growth parameters of the fi lm samples are given in the Supporting Information. The as-deposited and annealed samples were characterized by SEM (Nova600 NanoLab, FEI) with EDX, PXRD (D8 Discover, Bruker AXS) in θ –2 θ geometry with CuK α radiation, and in situ PXRD (Rigaku Smartlab, see Supporting Information). The in-plane Seebeck coeffi cient and electrical resistivity were simultaneously measured from room temperature to 573 K under He atmosphere ( ∼ 0.09 MPa, purity 99.999% with < 0.5 ppm residual oxygen) using two different ULVAC–RIKO ZEM-3 thermoelectric property measurement systems (instrumental error within 7%). Room temperature Hall measurements were performed using a Quantum Design Physical Property Measurement System. Supporting Information Supporting Information is available from the Wiley Online Library or from the author. [ 1 ] F. J. DiSalvo , Science 1999 , 285 , 703 . [ 2 ] B. C. Sales , Science 2002 , 295 , 1248 . [ 3 ] G. J. Snyder , E. S. Toberer , Nat. Mater. 2008 , 7 , 105 . [ 4 ] M. S. Dresselh
/
本文档为【1693_ftp】,请使用软件OFFICE或WPS软件打开。作品中的文字与图均可以修改和编辑, 图片更改请在作品中右键图片并更换,文字修改请直接点击文字进行修改,也可以新增和删除文档中的内容。
[版权声明] 本站所有资料为用户分享产生,若发现您的权利被侵害,请联系客服邮件isharekefu@iask.cn,我们尽快处理。 本作品所展示的图片、画像、字体、音乐的版权可能需版权方额外授权,请谨慎使用。 网站提供的党政主题相关内容(国旗、国徽、党徽..)目的在于配合国家政策宣传,仅限个人学习分享使用,禁止用于任何广告和商用目的。

历史搜索

    清空历史搜索