NSF-DOE Thermoelectrics Partnership:

Automotive Thermoelectric Modules with Scalable Thermo- and Electro-Mechanical Interfaces

Prof. Ken Goodson Department of Mechanical Engineering Stanford University

Dr. Boris Kozinsky Energy Modeling, Control, & Computation R. Bosch LLC

Prof. George Nolas Department of Physics University of South Florida

This presentation does not contain any proprietary, confidential, or otherwise restricted information

ACE067

1

• Start – January 2011 • End – December 2013 • ~40% complete

• Thermoelectric Device/System Packaging

• Component/System Durability

• Scaleup

• $1.22 Million (DOE+NSF) • FY12 Funding = $423K • Leveraging:

• ONR (FY09-11) • Fellowships (3 NSF, Sandia,

Stanford DARE)

Budget

Barriers (2.3.2)

• K.E. Goodson, Stanford (lead) • George Nolas, USF • Boris Kozinsky, Bosch

Partners

Overview Timeline

2

Relevance: Addressing Key Challenges for Thermoelectrics in Combustion Systems

…Low resistance interfaces that are stable under thermal cycling. …High-temperature TE materials that are stable and promise low-cost scaleup. …Characterization methods that include interfaces and correlate better with system performance.

hot side / heat exchanger

insulation / e.g. ceramic plate

conductor/ e.g. copper

insulation / e.g. ceramic plate

cold side / heat exchanger

conductor/ e.g. copper conductor/ e.g. copper

n p

thermalthermal

electrical& thermal

electrical& thermalthermalthermal

diffusion barrier

joining technology

contact resistanceHeat

Cooling

hot side / heat exchanger

insulation / e.g. ceramic plate

conductor/ e.g. copper

insulation / e.g. ceramic plate

cold side / heat exchanger

conductor/ e.g. copper conductor/ e.g. copper

n p

thermalthermal

electrical& thermal

electrical& thermalthermalthermal

diffusion barrier

joining technology

contact resistanceHeat

Cooling

600 °C

90 °C

Improvements in the intrinsic ZT of TE materials are proving to be very difficult to translate into efficient, reliable power recovery systems. Major needs include…

3

• Combustion TEG systems experience enormous interface stresses due to wide temperature spans.

• Thermal cycling degrades interface due to cracks, delamination, reflow, reducing efficiency.

• Our simulations show importance of thermodynamic stability (chemical reactivity, inter-solubility, etc.) and elastic modulus.

Relevance: Thermoelectric Interface Challenge

Technical Accomplishment 2011 (Bosch)

Gao, Goodson et al., Journal of Electronic Materials, 2010

50 μm50 μm

SEM45,000 Cycles

From our New Paper: Barako, Park, Marconnet, Asheghi, Goodson, “Infrared Imaging and Reliability Study of Thermoelectric Modules under Thermal Cycling,” Proceedings of ITHERM, San Diego, May, 2012.

Infrared

80oC

50oC

20oC

4

Research Objectives & Approach

OBJECTIVES

Develop, and assess the impact of, novel interface and material solutions for TEG systems of particular interest for Bosch.

Explore and integrate promising technologies including nanostructured interfaces, filled skutterudites, cold-side microfluidics.

Practical TE characterization including interface effects and thermal cycling.

APPROACH

Multiphysics simulations ranging from atomic to system scale.

Advanced materials development including CNT and metal nanowire TIMS, and high temperature thermoelectrics. Photothermal metrology including Pico/nanosecond, cross-sectional IR. MEMS-based mechanical characterization.

Removable backer

Nanostructured Film

Mid-temperature binder Adhesion layer

Panzer, Goodson, et al., Patent Pending (2007) Hu, Goodson, Fisher, et al., ASME JHT (2006) Won, Kenny, Fisher, et al., CARBON (2012)

Bosch Prototype TEG in exhaust system

5

Area (emphasis)

Specifics Source

Interfaces 100%

Nanostructured films & composites, metallic bonding Ab initio simulations and optimization

Stanford Bosch

Metrology 100%

(ZT)eff including interfaces, thermal cycling High temperature ZT

Stanford USF/NIST

Materials 100%

Filled skutterudites and half Heusler intermetallics Ab initio simulations for high-T optimization

USF Bosch

Durability 50%

In-situ thermal cycling tests, properties Interface analysis through SEM, XRD, EDS

Stanford Bosch

Heat sink 50%

Gas/liquid simulations using ANSYS-Fluent Novel cold HX using microfluidics, vapor venting

Bosch Stanford

System 50%

System specification, multiphysics code Evaluation of research impacts

Bosch Stanford

Research Approach

Additional Faculty & Staff beyond PIs Prof. Mehdi Asheghi, Stanford Mechanical Engineering Dr. Winnie Wong-Ng, NIST Functional Properties Group Dr. Yongkwan Dong, USF Department of Physics Stanford Students: Michael Barako, Yuan Gao (NSF Fellow), Lewis Hom (NSF Fellow), Saniya Leblanc (Sandia Fellow), Woosung Park, Amy Marconnet (NSF Fellow), Sri Lingamneni, and Antoine Durieux

6

7

Approach: Nanostructured Interfaces

Thermal Resistivity (m K / W)1.00.1

Elas

tic M

odul

us (M

Pa)

Greases & Gels

Phase Change Materials

Indium/ Solders

Adhesives

0.01

Nano-gels

Lifetimethermal cyclingLifetimethermal cycling

GOAL

104

103

102

101

Our Latest CNT Data*Our Latest CNT Data*

Copper Nanowire†

Carbon Nanotube

1 μm

‡ xGNPs:25 µm; ~10 nm

Life time thermal cycling

Nan

ostr

uctu

red

Inte

rfac

es

Year 2 DOE-NSF Project

*Gao, Goodson, et al., J. Electronic Materials (2010). Won, Goodson, et al., Carbon (2012a, 2012b) †In collaboration with group of Prof. Fritz Prinz, Stanford

‡www.xgsciences.com 8

Partial Filling – Optimization of Power Factor & Thermal Conductivity

Heavy-ion Filling Yields Lower Thermal Conductivity. Low Valence Filling Facilitates

Optimization of Power Factor.

Nolas, Kaeser, Littleton, Tritt, APL 77, 1855 (2000), Nolas, JAP 79, 4002 (1996) Lamberton, G.S. Nolas, et al APL 80, 598 (2001)

•Skutterudites with partial filling using heavy, low valence “guest” atoms

George S. Nolas Department of Physics,

University of South Florida

•Half-Heusler alloys: small grain-size provides for disordered state

x= 0.3 ; y = 1.5

x= 0.05 ; y=0

x= 0.6 ; y = 2.4

x= 0.23 ; y=0

x= 0.75 ; y = 2.6

x=y=0

x= 0.9 ; y = 2.4

LaxCo4Sb12-ySny

x= 0.3 ; y = 1.5

x= 0.05 ; y=0

x= 0.6 ; y = 2.4

x= 0.23 ; y=0

x= 0.75 ; y = 2.6

x=y=0

x= 0.9 ; y = 2.4

LaxCo4Sb12-ySny

Approach: Bulk TE Materials for Vehicles

9

Predictive computations of TE materials Electronic conductivity Seebeck coefficient Thermal conductivity

Understanding of transport mechanisms on atomic level and composition trends from ab-initio

Composition screening in skutterudites

Several new compositions predicted with higher Seebeck than base-line CoSb3

Trade-offs with conductivity investigated

Collaborative work with Nolas group focuses on Yb and Eu-filled skutterudites

Approach: Materials Computation Seebeck coefficient

experiment*ab-initio

ab-initio τ(E)

Seebeck

-1000

-800

-600

-400

-200

0

200

400

600

800

1000

0.67 0.69 0.71 0.73 0.75 0.77

µ (Ry)

S [µ

V/K]

CoSb3Composition AComposition BComposition CComposition D

300 K

Fermi level

Seebeck

Research and Technology Center North America 10

courtesy N. Singh-Miller, MIT

Computed electrostatic potential

Approach: Interface Optimization Thermal characterization focuses on interface

engagement, nanotube wetting, and stability Mechanical modeling of interfaces allows screening of

compositions to improve thermo-mechanical stability • Chemical reactivity at interfaces considering phase stability • Ab-initio computations and measurements of modulus, CTE • Q1/2011: Analysis of mechanical stresses at interfaces – in-plane

stress limitations using computed and measured CTE • Q1/2011: Cross-section of leg found to be related to the critical

stress, strong implications for materials strength for cost reduction Electronic transport across contacts • Work function and barrier calculations set up and calibrated • Key numerical screening criteria identified: Fermi level and band

offsets, Schottky barrier heights

Technical Accomplishment Q1 2011

Panzer, Goodson et al., Nanoletters (2010)

11

Technical Accomplishment: Publications 1. Barako, Gao, Marconnet, Asheghi, Goodson, “Solder-Bonded Carbon Nanotube Thermal Interface

Materials,” Proceedings of ITHERM, San Diego, May, 2012. 2. Barako, Park, Marconnet, Asheghi, Goodson, “Infrared Imaging and Reliability Study of

Thermoelectric Modules under Thermal Cycling,” ITHERM, San Diego, May, 2012. 3. Park, Barako, Marconnet, Asheghi, Goodson, “Effect of Thermal Cycling on Commercial

Thermoelectric Modules,” ITHERM, San Diego, May, 2012. 4. Marconnet, Motoyama, Barako, Gao, Pozder, Fowler, Ramakrishna, Mortland, Asheghi, Goodson,

“Nanoscale Conformable Coatings for Enhanced Thermal Conduction of Carbon Nanotube Films,” ITHERM, San Diego, May, 2012.

5. Gao, Kodama, Won, Dogbe, Pan, and Goodson, “Inhomogeneous Mechanical Properties of Vertically Aligned Multi-walled Carbon Nanotube Films,” ITHERM, San Diego, May, 2012.

4. Won, Gao, Panzer, Dogbe, Pan, Kenny, Goodson, 2012, "Mechanical Characterization of Aligned Multi-Wall Carbon Nanotube Films," CARBON, Vol. 50, pp 347-355.

5. Marconnet, Yamamoto, Panzer, Wardle, Goodson, 2011, "Thermal Conduction in Aligned Carbon Nanotube-Polymer Nanocomposites with High Packing Density," ACS Nano, Vol. 5, pp. 4818-4825.

6. Marconnet, Panzer, Goodson, “Thermal Conduction Phenomena in Carbon Nanotubes and Related Nanostructured Materials,” invited and submitted, Reviews of Modern Physics.

7. Gao, Kodama, Dogbe, Pan, Goodson, “Nonhomogeneous Mechanical Properties of Vertically Aligned Multi-Walled Carbon Nanotube Films,” CARBON, accepted and in press.

8. Leblanc, Phadke, Kodama, Salleo, Goodson, “Electrothermal Phenomena in Zinc Oxide Nanowires and Contacts,” Applied Physics Letters, accepted and in press.

9. Garg, Bonini, Kozinsky, Marzari, "Role of Disorder and Anharmonicity in the Thermal Conductivity of Silicon-Germanium Alloys: A First-Principles Study,” Physical Review Letters, 106, 045901 (2011).

10. Volja, Kozinsky, Li, Wee, Marzari, Fornari, "Electronic, vibrational and transport properties of pnictogen substituted ternary skutterudites,” submitted to Physical Review B. 12

5 Full-Length Papers Accepted, after review, for ITHERM 2012

7 Archival Journal Papers Appeared or were Accepted

Infrared

80oC

50oC

20oC

ThermoElectric Module/Pellet Nanostructured Interfaces

Gao, Panzer, Goodson et al., J. Electronic Materials, 2010

1 μm

Indium Bond

Thermal Cycling

50 μm50 μm

SEM45,000 Cycles

Thermal Cycling 100 Cycles (100

C, 6min)

Clamping System

Heat Exchanger

View Port

Recirculating Chiller Input/Output

Clamping System

Heat Exchanger

View Port

Recirculating Chiller Input/OutputHeat Exchanger Interior View

Copper Adapters

Heat

Sin

k

Chilled Water3-axis, 3-angle stage3-axis, 3-angle stage

Load CellLinear

Actuator

Infrared Camera

Sample

TE Heater

TE Cooler

Barako, Gao, Marconnet, Asheghi, Goodson, “Solder-Bonded Nanotube Thermal Interface Materials,” to appear in Proceedings of ITHERM, San Diego, May, 2012.

Barako, Park, Marconnet, Asheghi, Goodson, “Infrared Imaging and Reliability Study of Thermoelectric Modules under Thermal Cycling,” to appear in Proceedings of ITHERM, San Diego, May, 2012.

Marconnet, Motoyama, Barako, Gao, Pozder, Fowler, Ramakrishna, Mortland, Asheghi, and Goodson, “Nanoscale Conformable Coatings for Enhanced Thermal Conduction of Carbon Nanotube Films,” to appear in Proceedings of ITHERM, San Diego, May, 2012.

1 μm

Nanofoil Bond

Nanoscale Conformable Coatings for Enhanced Thermal Conduction of CNT Films

Solder-Bonded Nanotube Thermal Interface Materials

2 μm

Electrodeposited Metal Nanowire TIMs

Michael Barako, Unpublished research

High Temperature IR Imaging & Characterization

Technical Accomplishments: Stanford Overview

13

Technical Accomplishment : Interface Characterization on Thermoelectric with Thermal Cycling

0

2

4

6

8

10

12

1 10 100

R" (

m2 K

MW

-1)

N Cycles

R" TotalR" CNT-SubR" CNT-Pt

R”Total R”CNT-Sub + R”CNT R”CNT-Metal

Resistances for 1.5, 2.5, and 40 micron thick CNT films varied between 0.035 and 0.055 cm2 oC/W, with evidence of decreasing engagement with increasing film thickness.

Cycles (30 to 200

C, 6min)

14

Technical Accomplishment: Mechanical Characterization of CNT Films

Maruyama Lab Samples (SWNT), 95 kg/m3

Monano Samples (MWNT), ~30 kg/m3

Wardle Lab Samples/MIT (MWNT), 45 kg/m3

Zipping/Velcro Model

Thickness (μm)

Modulus (MPa)

Density (kg/m3)

SWCNTTop 1 600 110

SWCNTMiddle 0-25 0.5 95

MWCNTTop 0.4 300 40

MWCNTMiddle 0-150 10 29

Polysilicon 5.8-8.7 155e3 2330

Crust Model

With W. Cai Group, Stanford 15

Won, Gao, Goodson, et al., “Mechanical Characterization of Aligned Multi-Wall Carbon Nanotube Films,” CARBON (2012)

Infrared

80oC

50oC

20oC

Technical Accomplishment: Infrared Thermometry Failure Analysis of TE Modules

Before Cycling After 45,000 Cycles Optical

Infrared SEM courtesy of Yuan Gao

100oC A modified Harman technique was developed to measure the TE figure of merit ZT and the electrical resistance R.

70oC

40oC

100 101 102 103 104 1050.54

0.56

0.58

0.6

0.62

0.64

0.66

0.68

Cycles

R [ Ω

]

100

101

102

103

104

1050.5

0.52

0.54

0.56

0.58

0.6

0.62

0.64

ZT

50 μm50 μm

SEM45,000 Cycles

Fracture

Optical

Thermal conductivity and Seebeckcoefficient remain relatively stable

Reduction of electrical conductivity and ZT

16

•Indium (In) foil† is obtained (25 μm thick). •Cleaned and etched using:

1. Acetone 2. Isopropyl alcohol 3. Deionized water 4. Solder flux 5R

† Indium foil by Indium Corp.

Glass

Si

CNT

In

Hotplate

•The foil is compressed between the CNT film and the glass substrate with light pressure •The stack is placed on a hot plate at 180oC for one minute. This melts and bonds the indium to the adjacent surfaces. (Tmelt = 156.6oC)

Cr/Ni/Au

1 μm1 μmIndium

Cr/Ni/Au

CNT Film

CNT

CNT

Glass

Si

1) Metallize Substrates

3) Remove growth Si

GlassCNTSi

2) Bond surfaces

Metal

Si

Barako, Gao, Marconnet, Asheghi, Goodson, “Solder-Bonded Nanotube Thermal Interface Materials,” Proceedings of ITHERM, May, 2012.

Technical Accomplishment: CNT-Indium Bonding

17

Technical Accomplishment: CNT Nanofoil Bonding

•Nanofoil‡ (NF) is a 40μm Al/Ni superlattice which ignites and exothermically alloys to adjacent surfaces •NF is placed between two gold surfaces. Pressure is applied and the NF is ignited, bonding the two surfaces •Sn-plated NF bonds Au surfaces (forming Sn-Au bonds). The resulting intermetallic is stable up to 1000oC

‡ Nanofoil® by Indium Corp.

b) NF alloys to form Sn-Au bonds to adjacent surfaces

Al/Ni Alloy

Sn Plating

CNT Film

5 μm

a) NF is placed between CNT and adjacent surface

Glass

Si

CNT

Al/Ni

Glass

Si

CNT

Al/Ni

50 μm

Sn-Au

Sn-Au

Cr/Ni/Au

Sn

Cr/Ni/Au

Barako, Gao, Marconnet, Asheghi, Goodson, “Solder-Bonded Carbon Nanotube Thermal Interface Materials,” to appear in Proceedings of ITHERM, San Diego, May, 2012.

18

Thermal Boundary Resistance

Thermal Conductivity

Using conservation of energy, Fourier’s Law, and neglecting convection/radiation, we get:

Using the temperature drop at the interface and Fourier’s Law ref

ref

refCNT

dxdTk

TqTR intint ∆

=′′

∆=′′ −

sample

ref

ref

sample

dxdT

dxdT

kk

=

Technical Accomplishment: Pressure-Dependent Infrared Thermometry of CNT Interfaces

Copper Adapters

Heat

Sin

kChilled Water3-axis, 3-angle stage3-axis, 3-angle stage

Load CellLinear

Actuator

Infrared Camera

Sample

TE Heater

TE Cooler

Compressive Measurement Apparatus

0 0.5 1 1.520

30

40

50

60

70

Tem

pera

ture

[o C]

Position [mm]

CNTSi Glass

dT/dx| CNT

dT/dx| glass

ΔTCNT-Glass

ΔTCNT-Silicon

Gro

wth

In

terf

ace

Bond

ed

Inte

rfac

e

19

0 100 200 300 400 500400

600

800

1000

1200

1400

1600

Pressure [kPa]

R' in

t [mm

2 -K/W

]

0 100 200 300 400 5000.4

0.6

0.8

1

1.2

1.4

Pressure [kPa]

k [W

/m/K

]

Intrinsic Thermal Conductivity, kCNT

Thermal Boundary Resistance, R''int Heat Sink

3. Variation in CNT Heights

CNT-CNT Contacts

1. CNT-CNT Boundary Resistance

Individual CNT Thermal Conductance

Marconnet, Wardle, Goodson, et al., ACS Nano (2011)

Heat Source

2. CNT-metal Dry Contact

2. CNT-metal Dry Contact

1. CNT-CNT Boundary Resistance

2. CNT-metal Dry Contact

Technical Accomplishment: Pressure Dependence Before Bonding

20

0 0.5 1 1.520

40

60

80

100

Distance [mm]

Tem

pera

ture

[C]

Glass

NF

CNT For constant q'', the interfacial temperature drop is reduced by an order of magnitude through solder bonding

Indium wets to CNTs and engages more CNTs in conduction, increasing the bulk thermal conductivity of the film

Thermal Boundary Resistance, R''int

Heat Sink

Nanoscale metal-CNT interface resistance (phonons) Partial nanotube engagement

0 200 400 600 800101

102

103

Axial Pressure [kPa]

R" C

NT-

In-F

S [m

m2 K

/W]

Dry ContactUnbonded (In)Bonded (In)Unbonded (NF)Bonded (NF)

21

Technical Accomplishment: Pressure Dependence After Bonding

22

100 101 102 10310-2

10-1

100

101

102

103

104

Length [µm]

The

rmal

Bou

ndar

y R

esis

tanc

e [m

m2 K

W-1

]

If the CNT-Substrate/ metalization interface resistance could be reduced, the intrinsic thermal resistance of the CNT films would outperform solders

102 1

Yang et al. (2002,2004)Tong et al. (2007)Tong et al. (2006)Pal et al. (2008)Son et al. (2008)Xu et al. (2006)Zhang et al. (2008)Xu et al. (2006a)Xu et al. (2006b)Cola et al. (2008)Hodson et al. (2011)Cola et al. (2007)Aradhya et al. (2008)Cross et al. (2010)Panzer et al. (2008)Hu et al. (2006)Gao et al. (2010)Barako et al. (2012)Marconnet et al. (2011)Marconnet et al. (2012)

RCNT-S2 RCNT RS1-CNT

S1 CNT

S2

The blue arrow shows the magnitude of R_S1-CNT and R_CNT-S2). The green data points indicate the magnitude of the intrinsic R_CNT.

23

Technical Accomplishment: Intrinsic Thermal Resistance

102 1

Yang et al. (2002,2004)Tong et al. (2007)Tong et al. (2006)Pal et al. (2008)Son et al. (2008)Xu et al. (2006)Zhang et al. (2008)Xu et al. (2006a)Xu et al. (2006b)Cola et al. (2008)Hodson et al. (2011)Cola et al. (2007)Aradhya et al. (2008)Cross et al. (2010)Panzer et al. (2008)Hu et al. (2006)Gao et al. (2010)Barako et al. (2012)Marconnet et al. (2011)Marconnet et al. (2012)

RCNT-S2 RCNT RS1-CNT

S1 CNT

S2

10%CNT

8.7%CNT

35%CNT

11.5%CNT

0.8%CNT

2%CNT

40%CNT

Length [μm]

104

101

10-2 100 101 102 103

Ther

mal

Bou

ndar

y R

esis

tanc

e (m

m2 K

W-1

)

24

Technical Accomplishment: Impact of CNT Volume Fraction on Intrinsic Thermal Conductivity

Technical Accomplishment: Electrodeposited Metal Nanowire TIMs

Cu

Cu NW Nominal geometry: •Cylindrical NWs •10 μm film thickness •200 nm diameter

a) Polycarbonate membrane is etched to create cylindrical pores

b) Catalyst Pt/Pd is deposited on one side of the membrane

c) Metal is eletrodeposited into the pores

d) Membrane is etched away, leaving freestanding nanowires

10 μm

SEM of copper NW film:

2 μm

Pt/Pd

1 10 1000.1

1

10

100

1,000

Volume Fraction [%]

k [W

m-1

K-1

Thermal Conductivity vs. Volume Fraction

CopperNickel

• Picosecond thermoreflectance technique. • Uncertainty is due to current sample preparation techniques)

In collaboration with Prinz group, Stanford

25

Where electrode was attached for electroplating

5 μm

4 μm

20 μm

No CNTs at this spot in SEM

10 μm

Uniform coating & Grain Size most of the area except where the electrode is attached

Technical Accomplishment 2011: Nanoscale Conformable Coatings for Enhanced Thermal Conduction of CNTs

Marconnet, Motoyama, Barako, Gao, Pozder, Fowler, Ramakrishna, Mortland, Asheghi, and Goodson, “Nanoscale Conformable Coatings for Enhanced Thermal Conduction of Carbon Nanotube Films,” to appear in Proceedings of ITHERM, San Diego, May, 2012.

26

A custom-fabricated vacuum enclosure with integrated heat exchanger will be built to achieve >500

C temperature gradient across a TE sample and facilitate simultaneous electrical and optical measurements.

Technical Accomplishment: High Temperature Infrared Imaging

Clamping System

Heat Exchanger

View Port

Recirculating Chiller Input/Output

Clamping System

Heat Exchanger

View Port

Recirculating Chiller Input/Output

TE Material

Hot Side Cold Side

Clamping System Top down view

IR Camera

250 W

TE Material

Hot Side Cold Side

TE Material

Hot Side Cold Side

Clamping System Top down view

IR Camera

250 W

Liquid Cooling

Hot Side

Hot Side

Cold Side

BiTe

241 °C

406 °CHot Side

Cold Side

BiTe

241 °C

406 °C

ThermalGrease

Copper

CopperTe

mpe

ratu

re (o C

)ΔT Between Top and Bottom Plates (oC)

Top Plate

Bottom Plate

500

400

300

200

0 100 200

IR

TC

IR Image

Heat Exchanger Interior View

Quantum Focus InstrumentsIR MicroscopeQuantum Focus InstrumentsIR Microscope

27

Technical Accomplishment: Bulk TE Materials for Automotive Applications

p-type partially filled and Fe-substituted Skutterudites: YbxCo4-yFeySb12

Double filled and Fe-substituted Skutterudites: BaxYbyCo4-zFezSb12

n- and p-type Half-Heusler alloys

20 40 60 80

Inte

nsity

(a.u

.)

2-Theta (deg.)

calc. PXRD pattern of CoSb3

measured PXRD pattern of Yb0.1Co3FeSb12

measured PXRD pattern of Yb0.4Co3FeSb12

Yb-filled Fe-substituted CoSb3 by solid state reaction

Thermopower of hot pressed Yb0.4Co3FeSb12 ~ 60µV/K at room temp.

20 40 60 80 In

tens

ity (a

.u.)

2-Theta (deg.)

calc. PXRD pattern of ZrNiSn measured PXRD of arc melted ZrNiSn measured PXRD of arc Melted ZrNiSn measured PXRD of arc Melted ZrNiSn

Bi2Te3-alloys for High Resolution IR Thermometry (in collaboration with Marlow Industries, Inc.)

Survey of other material systems with potential for enhanced thermoelectric properties

ZrNiSn by Arc Melting

In collaboration with GM R&D for melt-spun processing in investigating amorphous and fine-grained Half-Heusler alloys.

28

Ytterbium (Yb) partially-filled skutterudites • Partial filling optimizes lattice thermal conductivity reduction1 • Yb intermediate valence in CoSb3 maximizes filler concentration while minimizing

added carriers2

P-type partially filled skutterudites (high temp measurements at NIST & Clemson U.) Amorphous intermetallic alloys3 (in collaboration with General Motors) Bi2Te3-alloys for High Resolution Infra-Red Thermometry (in collaboration with Marlow Ind.) Survey of other material systems with potential for enhanced thermoelectric properties

Technical Accomplishment: Bulk TE Materials for Automotive Applications

Glen Slack initiated PGEC concept with skutterudites:

Fillers should be loosely bonded to the cage-forming atoms. Fillers should have large atomic displacements. Fillers act as independent oscillators (“rattlers”). Interaction of rattlers with the normal modes should lower lattice thermal conductivity. Phonon-scattering centers (“rattlers”) should not greatly affect electronic properties.

filler Co, Fe Sb, As, P

Prof. G. Nolas, Dr. Yongkwan Dong, University of South Florida

1. G.S. Nolas, et al, Phys. Rev. B 58, 164 (1998) 2. G.S. Nolas, et al, Appl. Phys. Lett. 77, 1855 (2000) 3. G.S. Nolas and H.J. Goldsmid, Phys. Stat. Sol. 194, 271 (2002)

29

Thermal Conductivity of Yb-filled Skutterudites

1. G. S. Nolas, et al, Appl. Phys. Lett. 77, 1855 (2000) 2. P. F. Qiu, et al, J. Appl. Phys. 109, 063713 (2011)

300 400 500 600 700 800 9000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Temperature (K)

Ther

mal

Con

duct

ivity

(W/m

-K)

Yb0.03Co2.70Fe0.91Sb12Yb0.14Co2.51Fe0.61Sb12Yb0.15Co2.65Fe0.80Sb12Yb0.20Co2.58Fe0.91Sb12Yb0.19Co4Sb12 (1)Yb0.94Fe4Sb12.07 (2)

300 400 500 600 700 800 9002.0

2.5

3.0

3.5

4.0

Temperature (K)

Ther

mal

Con

duct

ivity

(W/m

-K)

Ba0.11Yb0.07Co2.80Fe0.96Sb12Ba0.13Yb0.06Co2.66Fe0.80Sb12Ba0.17Yb0.08Co2.83Fe0.88Sb12Ba0.16Yb0.16Co2.87Fe0.91Sb12Ba0.16Yb0.23Co2.1Fe2.3Sb12

300 400 500 600 700 800 9002.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

Ther

mal

Con

duct

ivity

(W/m

-K)

Temperature (K)

Ba0.31Yb0.04Co2.71Fe1.05Sb12Ba0.26Yb0.10Co2.79Fe0.92Sb12Ba0.30Yb0.16Co2.58Fe1.09Sb12Ba0.11Yb0.03Co4Sb12.07 (1)Ba0.05Yb0.09Co4Sb12.13 (1)Ba0.3Co4Sb12 (2)

p-type Ba + Yb filled Skutterudites

3. X. Shi, et al, Appl. Phys. Lett. 92, 182101 (2008) 4. L. D. Chen, et al, J. Appl. Phys. 90, 1864 (2001)

[3] [3]

[4] [1] [2]

Increasing Yb concentration

Increasing Ba concentration

Partial filling optimizes lattice thermal conductivity reduction

Technical Accomplishment: Bulk TE Materials for Automotive Applications

30

Electronic Transport

• The effect of filling distorts the structure locally. – Soft Sb rings accommodate the distortion.

• Electrons from filler open band gap, while volume expansion closes the gap. – Band gap is also sensitive to local distortion

0.15

0.17

0.19

0.21

0.23

0.25

0.27

0.29

0.31

0.33

0.35

0 0.2 0.4 0.6 0.8 1

filling fraction x

Eg a

t Gam

ma

CaxCo4Sb12 SrxCo4Sb12 BaxCo4Sb12 Na1Co4Sb12 K1Co4Sb12

Ca

Ba

Sr

design rules

0.15

0.17

0.19

0.21

0.23

0.25

0.27

0.29

0.31

0.33

0.35

0 0.2 0.4 0.6 0.8 1

filling fraction x

Eg a

t Gam

ma

CaxCo4Sb12 SrxCo4Sb12 BaxCo4Sb12 Na1Co4Sb12 K1Co4Sb12

Ca

Ba

Sr

design rules

.but

,

CaSrBa

CaSrBa

ωωω >>

>> MMM

Ban

d ga

p (e

V)

Filler vibration is localized and strongly hybridized with Sb atoms. Effect of force constants more important than filler mass.

Technical Accomplishment 2011: Effects of Fillers in skutterudites

Research and Technology Center North America

Filling Factor x

31

Electronic Transport Technical Accomplishment: Computational Composition Engineering

Power Factor

-5.00E-04

0.00E+00

5.00E-04

1.00E-03

1.50E-03

2.00E-03

2.50E-03

3.00E-03

3.50E-03

4.00E-03

0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9

µ (Ry)

S2 σ [

W/m

K2 ]

CoSb

CoSnTe

CoSnSe

CoSnS

CoGeSe

CoGeS

Ternary-substituted skutterudites may hold more potential for n-type XCo4B12 B = Sb (Ge,Sn)/ (S,Se,Te)

Higher Seebeck coeffs than CoSb3 Filling of ternary skutterudites weakly affects the Seebeck maxima

Changes the carrier concentration significantly

CoSb3

Seebeck coeff Power factor

Research and Technology Center North America 32

Thermal transport Technical Accomplishment: Nanostructure Design for Thermal Transport

• New method developed for ab-initio thermal conductivity prediction – Grain boundary scattering term included in thermal conductivity

• Effect of scattering noticeable at 300K for 500nm grains • Grain boundary scattering much less effective at high T for skutterudites

Research and Technology Center North America

Smaller grains

d,n)v(

τ phph

q+=

−τ11

33

Accomplishment: Outreach & Engagement Industry Initiatives in Science and Math Education (IISME) – Summer 2011

Undergraduate Thermoelectrics Lab – Fall Quarter

K-12 Educational Outreach – Fall 2011-present

Mentorship of a public high school teacher for summer research experience and curriculum development using thermoelectrics

Stanford’s heat transfer course (ME131A) includes a thermoelectrics laboratory experience. Designed in conjunction with IISME teacher Lab exercise uses infrared microscopy with thermoelectric modules

Designed engineering course which is now taught at a public high school Experienced first-hand application of thermoelectric modules

We are now partnered with a public high school to provide materials and mentors for a TE design lab Introduces high school students to thermoelectric modules and their applications each semester Hands-on design lab to engage students in engineering

34

Collaboration & Coordination

Stanford • Prepares CNTs samples on TE materials • Transport property measurements of CNT-TE pellet combination, thermomechanical reliability tests on interface (300-800 K) •Process development for CNT TIM tape

Bosch • Ab-initio simulations of transport properties of TE materials and interfaces. •System-level simulation and optimization

USF • Develops high-T, high efficiency TE materials • Transport properties (ρ, S and κ) and Hall measurements (10 - 300K) • Structural, morphological and thermal (DTA/TGA) analyses

NIST •Transport properties (ρ, S and κ) and Hall measure- ments (1.8-390K) •Specific heat, Power Factor measurement at 300 K. •Custom-designed precision TE properties measurement system (300 – 1200 K)

1- Interface 2- System-level 3- Durability 4- Materials 5- Heat sink 6- Metrology

1, 3, 4, 6

4, 6

1, 2, 3, 5

Samples Information

35

• Bulk TE Materials: Develop p-type partially/double filled Fe-substituted Skutterudites, n- and p-type half heusler alloys for melt-spun processing, thermal stability tests of materials and joints.

• CNT Thermal Tape Development and Characterization: Bonding extension to 600oC, thermal stability investigation.

• Nanostructured Metal Thermal Interface Materials: Investigate thermal, mechanical, and electrical properties of metal nanowires and meshes, optimization of materials, geometries, and surface treatments for operation at 600oC.

• High-T (ZT)eff Characterization Facility Implementation: Vacuum chamber development with IR transparent window, validation using Bi2Te3-alloys

• Ab-Initio Simulations: Band gap calibration for skutterudite and half-

Heusler families, focussed computatios on phase stability, Seebeck coefficient, and transport properties.

Proposed Future Work

36

• With this award, DOE & NSF are enabling an academic-corporate team to focus on the key practical challenges facing TEG implementation in vehicles: interfaces, system-relevant metrology, and materials compatibility

• We are developing metrology for fundamental properties of nanostructured interfaces, as well as (ZT)eff metrology for half-Heusler and skutterudite thermoelectrics considering interfaces. Simulations include atomistic and ab initio results for TE materials and interfaces, and system & heat exchanger level optimization with the corporate partner.

• Key 2011 results include: (a) process development of CNT tape and several bonding options (Stanford) (b) detailed mechanical characterization of CNT films (Stanford), (c) IR characterization of TE pellets and corresponding interfaces under thermal cycling (Stanford), (d) interface modeling & optimization (Bosch) and (e) process development (arc melting, melt spun) for bulk TE materials (USF)

Summary Slide

37

Technical Backup Slides

• Bonding the CNT films to relevant substrates is a major challenge as not all materials are compatible with the CNT growth procedure.

• Recent progress utilizing a combination of metallizations allows CNT films grown on sacrificial silicon wafers to be successfully transferred to a range of substrates using thin indium foils as binding layers.

• This is a key step towards developing the free standing CNT tape for thermal interface applications.

Diffusion barrier (Ni) 100 nm Surface layer (Au) 150 nm

Adhesion layer (Cr) 100 nm

Substrate (Glass/Quartz)

Surface layer (Au) 150 nm

Solder (In) ~25μm foil

Diffusion barrier (Ni) 100 nm Adhesion layer (Cr) 100 nm

Substrate

CNT Film

Evaporated

Evaporated

Foil ribbon

Technical Accomplishment: CNT Metallization & Bonding Procedure

39

Glass

b) Clean indium foil or apply flux

CNT

1) SAMPLE PREPARATION

Glass

CNT

CNT film bonded to glass, before removal of Si substrate

Si

2) THERMAL BONDING

Hotplate Si

Au/Ni/Cr coated CNT film on Si

a) Evaporate Au/Ni/Cr on CNT and glass substrates

Au/Ni/Cr

Indium

Indium

Technical Accomplishment: CNT Bonding Procedure

SiCNT

GlassIndium

40

Technical Accomplishment: CNT Bonding Procedure

3) CNT FILM RELEASE

CNT bonded to Si

CNT bonded to glass

Original CNT

substrate, with no

CNT remaining CNT

CNT (side is covered with Au from metallization)

20 μm

Indium

Cr/Ni/Au

CNT

Glass

CNT

Glass

CNT

2 μm CNT

Indium

Cr/Ni/Au

20 μm

41

100 1010

0.2

0.4

0.6

0.8

1

Thickness of Thermoelectric [mm]

Z eff/Z

Ideal Solder and Ideal CNT

High Quality CNT

Electrically Insulating Grease

Interface Material

R’’th [W/m2/K]

R’’e [Ω m2]

Solders And

Ideal CNT ~10-7 ~10-12

High Quality CNT ~10-6 ~10-10

Lower Quality CNT ~10-5 ~10-8

Thermally & Electrically Conductive

Grease

~3x10-6 ~3x10-9

Thermal Conductive

Grease ~8x10-6 ~3x10-7

Electrically Insulating

Grease ~8x10-6 >10-5

Relevance: Effect of Interface Resistances on Thermoelectric Device Properties

Using model of Xuan, et al. International Journal of Heat and Mass Transfer 45 (2002). 42

Popular Press

“…Stanford is also working with the National Science Foundation (NSF) on a project with the Department of Energy Partnership on Thermoelectric Devices for Vehicle Applications. Here, the nanotape will facilitate the recovery of electrical power from hot exhaust gases using thermoelectric…”

http://www.eetimes.com/electronics-news/4212401/Nanotape-could-make-solder-pads-obsolete 43