Ptype thermoelectric properties of stoichiometric fullHeusler alloy Fe

P-type thermoelectric properties of stoichiometric full-Heusler alloy Fe 2 Ti. Sn sintered samples Yokohama National University T. Ozaki and H. Nakatsugawa 1

Introduction Thermoelectric PG system High Temp e⁻ n p Thermoelectric materials clathrate Skutterudite Silicide half Heusler h⁺ resistance electric current Low Temp S : Sebeeck coefficient [V/K] ρ : electric resistivity [Ωm] κ : thermal conductivity [W/m. K] Improvement of ZT S ρ κ and more A lot of thermoelectric materials have been studied 2

back ground full-Heusler alloy Intermetallic compounds with composition X 2 YZ When valence electron pseudogap is formed concentration per atom at Fermi level. is six, (ex Fe₂VAl, Fe₂Ti. Sn etc) from Mott’s theory High Sebeeck coefficient is expected 3

aim of study Fe 2 Ti. Sn • From the first principles calculation, it was suggested that S was higher at 300 K than Fe 2 VAl • Fe 2 Ti. Sn has low κ of 7～ 8 W/m. K as compared with other full Heusler alloys ( Fe 2 VAl : 28 W/m. K) But κ is still higher than the currently used Bi-Te based material We aim to improve thermoelectric characteristics by controlling grain refinement, promoting phonon scattering at grain boundaries, and reducing κ for Fe 2 Ti. Sn sintered sample 4

Experimental methods Production method milling time(in Air or Ar) 1 h /3 h /12 h arc melting milling using stainless ball press forming calcining of vacuum annealing at 1073 K/48 h Fe : Ti : Sn = 2 : 1 calcining at 723 K/2 h annealing of vacuum sintered sample Evaluation method ucrystal structure：XRD (Smart. Lab) u S ：steady method u Rietveld analysis : RIETAN-FP program u ρ ：dc four-probe method umicro-structuer ：SEM (VE-8800) u κ ：PEM-2 uparticle size distribution ：Image. J u RH : van der Pauw method 5

(511) (422) (331) (420) (400) (311) (222) (200) (111) (a) 1 h milling in Air (220) XRD patterns S Rwp Re 1 h 7. 7645 8. 454 1. 089 1 h_Ar 6. 3231 6. 062 0. 959 3 h 5. 9214 6. 43 1. 086 3 h_Ar 4. 876 5. 081 0. 984 12 h 5. 9214 6. 43 1. 086 12 h_Ar 5. 8636 5. 636 0. 961 (511) (422) (331) (420) (400) (311) (222) (200) (111) (220) (b) 1 h milling in Ar In Ar samples, formation of the second phase was suppressed. 6

SEM images (a) 1 h milling in Air (b) 3 h milling in Air 1 μm (d) 1 h milling in Ar (c) 12 h milling in Air (f) 12 h milling in Ar (e) 3 h milling in Ar 1 μm 7

particle size distribution (a) 1 h milling in Air Average particle size 0. 86μm (d) 1 h milling in Ar Average particle size 0. 88μm (b) 3 h milling in Air Average particle size 0. 72μm (e) 3 h milling in Ar Average particle size 0. 69μm (c) 12 h milling in Air Average particle size 0. 69μm (f) 12 h milling in Ar Average particle size 0. 66μm 8

electric resistivity ρ 9

Sebeeck coefficient S 10

thermal conductivity κ 11

thermal conductivity κ κph = κ - κcar Wiedemann-Franz rule κcar = LT/ρ (L = 2. 45× 10 -8 WΩ/K 2) 12

dimensionless figure of merit ZT 13

Conclusion Ø By milling in Ar, formation of second phase was suppressed and stoichiometric Fe 2 Ti. Sn composition was maintained, and the decrease of |S| in the Ar sample was suppressed. Ø As milling time longer, dispersion of particle size was suppressed, and average particle size was reduced from 0. 88μm at 1 h to 0. 66μm at 12 h. Ø As milling time longer, lattice thermal conductivity decreased, especially 4. 24 W/m. K@350 K for the sample milled for 12 h in Ar. Ø Maximum dimensionless figure of merit ZT was p-type thermoelectric characteristics of 0. 0013@305 K of milled for 3 h in Ar. Ø As future plans, further improvements of ZT are expected by preparing Fe 2 Ti. Sn sintered sample with stoichiometric composition shifted. 14

Thank you for your kind attention. 15