Characterizing Half-­‐Heusler Thermoelectric Unicouple ProperDes 1
Emily Johnson , Nicholas Kempf , Yanliang Zhang
Department of P hysics, St. O laf C ollege, Northfield, M N; 2 D epartment o f M echanical & Biomedical E ngineering, Boise S tate U niversity, Boise, I D •  Increases in global energy usage have created a growing need for new, efficient energy technologies •  Thermal energy losses pose many problems in current technologies •  Thermoelectric (TE) materials provide a method to convert waste heat directly into energy, making them a strong contender for energy producDon •  Most current TE generators and materials are costly and inefficient •  In order to improve feasibility of thermoelectric energy generaDon we need to improve modules •  OpDmizing fabricaDon methods will help improve TE module efficiency Diagram of TE module internal structure1 Background Thermoelectric Figure of Merit (ZT) •  Strong bond between half-­‐Heusler and conducDng •  Half-­‐Heusler compound material ¡  Nanostructured alloys increase ZT •  Consistent braze joints •  Braze alloy •  Desired electrical and thermal properDes of braze ¡  Diffusion bonding influences joint strength •  High efficiency unicouples and devices ¡  Similar thermal expansion coefficients are ideal ¡  High electrical conducDvity across braze ¡  Braze foil thickness changes joint properDes ¡  Low thermal conducDvity •  ConducDve material ¡  Copper provides high conducDvity electrical connecDon Braze Alloy IncusilTM-­‐ABA CusilTM CusilTM-­‐ABA ¡  Direct Bonded Copper (DBC) provides insulaDon ComposiDon (%wt.) Silver Copper Titanium Indium Solidus(°C) Liquidus (°C) CTE (×10-­‐6K-­‐1) 72 28 -­‐-­‐-­‐ -­‐-­‐-­‐ 779 779 19.6 63 35.25 1.75 -­‐-­‐-­‐ 779 816 18.5 Direct Bonded Copper
κ = thermal conducDvity ΔV = Seebeck voltage ΔT = temperature difference
Efficiency 9 120 DBC Incusil-­‐ABA 100 80 6 60 40 12 20 1 0 100 200 300 Applied Brazing Pressure [MPa] 3 Cusil-­‐ABA Cusil 2.5 2 1.5 1 0.5 0 0 400 0 2 Incusil-­‐ABA d) # 1 # 9 # 3 # 6 # 12 # 5 Cooling Power Output PelDer Effect Heat Flow n-­‐type p-­‐type Heat Rejected Current Source •  Applied current causes a temperature difference •  Charge carriers have different heat carrying capaciDes in different materials •  Charge carries moving between materials cause heat to be absorbed or rejected Acknowledgements # 7 # 9 Half-­‐Heusler Brazed half-­‐Heusler, DBC leg e) Brazed Copper leg f) 30 25 Assembly and Joint Brazing Procedure ~ 6 mm •  Sand parts to appropriate size (approximately 2×2×3 mm) •  Clean components in sonicaDon process •  Assemble half-­‐Heusler legs, braze foil, and direct bonded copper (DBC); or copper, braze foil, and copper in brazing fixture •  Braze components in vacuum furnace (approximately 0.01-­‐0.06 mbar at highest vacuum) Braze Fixture Heat Sink Cusil Copper Half-Heusler Material
Heat Flow p-­‐type 12 ~ 3 mm Braze Foil
Heat Source n-­‐type Sample Number [#] 10 Copper leg braze interface aqer c) tensile; and d) bending tests for various numbered samples shown in red on graphs above (leg dimensions ~2×2 mm) %
1+ ZTMean −1
THot − TCold "
ηmax =
$
'
THot $# 1+ ZTMean + TCold / THot '&
•  Applied temperature difference creates a current •  Temperature gradient across the TE material causes higher energy charge carriers to diffuse toward colder side •  ConDnuously applied heat keeps charge carriers moving 8 ~ 2 mm •  Percent energy output from TE device Seebeck Effect 6 Cusil-­‐ABA Power Output
η=
Heat Input
4 a) Tensile stress test results from Copper – IncusilTM-­‐ABA – Copper joints brazed at 825°C for 5 minutes with various pressures applied to legs; and b) completely reversed bending test results from Copper – CusilTM – Copper and Copper – CusilTM-­‐ABA – Copper joints brazed at 825°C for 5 minutes (samples from data points marked red can be seen below) c) σ = electrical conducDvity ΔV
α =−
ΔΤ
b) Voltage (μV) 59 27.25 1.25 12.5 605 715 18.2 Braze Process ConsideraDons •  Material preparaDon and assembly •  Atmosphere, applied pressure, surface roughness, and cleanliness •  Braze temperature and Dme for different materials •  Diffusion bonding, joint coverage, and braze foil dimensions TesDng Process 800 700 600 500 400 300 200 •  Tensile tests performed on brazed legs using homemade 100 tesDng fixture •  Completely reversed bending cycle test with a full cycle 0 across a maximum angle of approximately ± 24° This research was supported by the Boise State University REU in Materials for Energy and Sustainability and sponsored by NaDonal Science FoundaDon grant 1359344. 0 30 60 References 90 120 150 Time [min] 180 1.  Adapted from Snyder et al., 2008 Nature Materials, 105-­‐114 2.  GMZ Energy, Inc., Voltage Measurement Measured Contact Resistance: 0.827 μΩcm2 20 15 10 250 μm 5 0.2 0.3 0.6 0.5 0.4 PosiDon (mm) 0.7 e) SEM images of Copper – CusilTM – Copper braze joint interface; and f) GMZ Energy example voltage measurement across braze joint2 to measure contact resistance Future DirecDons Typical Furnace Braze Cycle 900 Temperature [°C] a) 140 Materials SelecDon FabricaDon ObjecDves •  Provides a measure of the thermoelectric performance of the TE material α = Seebeck coefficient ασ
ZT =
T
κ
Results Materials and Methods Number of Cycles [#] IntroducDon 2
2 Tensile Stress [MPa] 1 2
210 240 •  Increase material uniformity to create more consistent devices ¡  Decrease variability in part sizes during fabricaDon ¡  Increase braze consistency to overall leg performance •  Improve braze fixture to streamline braze process ¡  Apply pressure uniformly throughout braze cycle ¡  Higher quanDty fabricaDon yield •  Develop beZer tensile tesDng apparatus to aZach TE legs for tesDng •  Improve resistance measurement technique across joint •  Test material limits with current fixtures and fabricaDon techniques •  Implement efficiency measurements, thermal cycling tests, and thermoelectric generator (TEG) tesDng to determine power output and lifeDme of TE devices for various applicaDons