Thermoelectrics for Vehicle Applications

Jihui YangMaterials Science and Engineering Department,

University of Washington

2014 APS/GERA ENERGY RESEARCH WORKSHOPENERGY FOR TRANSPORTATION

Denver, ColoradoMarch 2, 2014

1

Acknowledgement

CollaboratorsJames R. Salvador, Gregory P. Meisner, Francis Stabler, and

Mike Reynolds - GM R&DAndrew A. Wereszczak - Oak Ridge National Lab Xinfeng Tang - Wuhan University of Technology Xun Shi, Lidong Chen, and Wenqing Zhang, Chinese

Academy of Sciences

Sponsored by

State of Washington (STAR program), Inamori Foundation, National Science Foundation, DOE, and General Motors

2

Outline

Magnitude of Waste Heat

Automotive TE Commercial Success and Opportunities

Industrial Technology Development Approach Coupled Materials-Systems Solutions

TE Materials Research Mechanical Property – Battling the Flaws Semiconductor Carriers Engineering and Thermal

Transport

Conclusions 3

Waste heat is one of our most abundant sources of alternative energy

The Magnitude of Our Waste Energymassive quantity of “waste”energy

4

1 Quad (quadrillion btu) of energy is equivalent to 340,000 tank cars of crude oil stretched from Miami to Seattle (3,300 miles)

In the area of energy, scale matters!

How Much is One Quad?

5

Today’s ICE-based vehicles: < 20% of fuel energy is used for propulsion > 60% of gasoline energy (waste heat) is not utilized

Typical Energy Path in Gasoline Fueled Internal Combustion Engine Vehicles

Com

bust

ion

30% Engine

Vehicle Operation

100%

40% Exhaust

Gas

30% Coolant

5% Friction & Radiated

25%Mobility &

Accessories

Gas

olin

eG

asol

ine

Gas

olin

eG

asol

ine

Gas

olin

eG

asol

ine

Note: Charts in this presentation are drawn from multiple sourcesand may have slightly different numbers because of different vehicles & assumptions. Considerthem general estimates, not preciseanalysis.

6

Solid state heat engines with electrons and holes as working fluids

TE Cooling and Power Generation

7

> 10 million thermoelectric-based seat systems sold (Feb. 2013, Gentherm PR )

Commercial Success of TE

8

Vehicular TE Waste Heat Recovery Development

BMW GM

Ford9

Exhaust Heat - City Driving Cycle

0

10

20

30

40

50

60

70

80

0 500 1000 1500 2000 2500

Test Time (s)

kW

The Suburban was selected as a test vehicle because it simplified the modifications and installation of the prototype.

Fuel efficiency improvement will be better in small, fuel efficient vehicles than in large vehicles because the electrical load in small vehicles is a larger portion of the engine output.

Exhaust Data - Chevy Suburban

10

Opportunity for TE Cooling

11

Reduce onboard AC without sacrifice passenger comfort level Improve fuel economy and CO2 emission Work funded by DOE

“If all passenger vehicles had ventilated seats, we estimate that there could be a 7.5 % reduction in national air-conditioning fuel use. That translates to a savings of 522 million gallons of fuel a year,"John Rugh, project leader for NREL's Vehicle Ancillary Loads Reduction Project.

Distributed Cooling for High Efficiency HVAC System

12

Typical Industrial Development - Multi-Scale, Interdisciplinary …

nm µm mm m

log(Length scale)

Engineering

MaterialsScience

Chemistry

Physics

Competency

Electronic

Microstructural

0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.0050

0.1

0.2

0.3

0.4

0.5

0.6

0.7

140612691131994.1856.8719.5582.2444.9307.6170.3

Lp [m]

Wout

PbTe/TAGS85

Continuum

Atomistic

13

$/W (not only ZT) is used for balancing various material, module, and subsystem options

$/W can be readily converted to $/∆mpg, and $/∆mpg < Savings/∆mpg isnecessary to provide consumer value

FE (mpg)0 10 20 30 40 50 60 70

3 Y

ear F

uel S

avin

g pe

r 1m

pg F

E im

prov

emen

t

0

200

400

600

800

1000

1200

1400

1600

1800

2000

$2/gallon$3/gallon$4/gallon

Consumer Fuel Savings/∆mpg ≈ $300-400/∆mpg (15000 mile/yr., 3yrs., 18-20 mpg)

$/W – a Program Metric

14

Approach – a Couple Materials-System Solution

System Model

Pack/Generator Config.Structure Design

Thermal Management

Electrical Management

TE Module DesignMatl. Dev.

TE Materials

Thermo-Mech. Data

Electrochem. Data

Matl. Interaction &Degradation

Vehicle Requirements

Energy and Power (T)

Exhaust and Radiator

Powertrain Control

Elec. & Thermal Load

Cost & Life Target

Mass & Volume

Efficiency

Validation

Module Prototyping& TestsGeneratorPrototyping & TestsDynamometer Tests

Fundamental Research

System Integration and Validation

Vehicle Targets

15

Great Power in Solving Problems Using Coupled Materials-Systems Solutions

System requirements can be quickly translated to revised materials requirementsMaterial advances can be quickly translated into system impact

-

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

- 100 200 300

Avg

. Ou

tpu

t o

ver

cycl

e

ZT=1ZT=2ZT=3ZT=6

0

0

350W GM goal for 2% FE savings

1.65kW GM Estimate for 10% FE

∆THX Hot-Side Heat Exchanger (C)Heat Exchanger Performance

n p

hot flow

HXT∆

-

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

- 100 200 300

Avg

. Ou

tpu

t o

ver

cycl

e [

ZT=1ZT=2ZT=3ZT=6

0

0

350W GM goal for 2% FE savings

1.65kW GM Estimate for 10% FE

∆THX Hot-Side Heat Exchanger (C)Heat Exchanger Performance

-

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

- 100 200 300

Avg

. Ou

tpu

t o

ver

cycl

e

ZT=1ZT=2ZT=3ZT=6

0

0

350W GM goal for 2% FE savings

1.65kW GM Estimate for 10% FE

∆THX Hot-Side Heat Exchanger (C)Heat Exchanger PerformanceHeat Exchanger Performance

n p

hot flowhot flow

HXT∆

J. Yang and T. Caillat, MRS Bulletin 31, 3, 224 (2006)

Power vs. Time - Urban drive cycle

0

200

400

600

800

1000

1200

0 500 1000 1500 2000 2500 3000

time [s]

Pow

er [W

]

Quasi-steady analysis - EES Transient analysis - ANSYS

16

Challenges for TE Technology Development

Materials– High ZT: p-type skutterudites, composites…– Manufacturing and processing for enhanced thermo-

mechanical properties

Modules– Joining, diffusion barrier, …– Design to mitigate thermal and vibrational stress

Systems– Heat exchanger, thermal management, packaging,

electrical interface, mechanical interface, by-pass– Design to mitigate thermal and vibrational stress

Integration and Control– Cycle-averaged output design and control

17

Mechanical Properties of TE Materials- Battling Flaws

Of concern is the long term survival of the TE elements under vehicle operating conditions – vibration, thermal cycle, thermal shock…

TE materials are typical brittle ceramics, Tensile Strength << Compressive Strength - manage tensile stress for conservative design

Must minimize flaw sizes 18

Strength-Limiting Flaw Classification For Brittle Materials

19

Net-Shape Processing

Salvador, et al., Science of Advanced MaterialsVol. 3, 577–586, 2011

20

Power Generation Equation: Cooling Equation:

Z has the dimension of T-1. ZT is usually called dimensionless Figure of Merit

The performance of a TE module is determined by the intrinsic materials ZT

H

CH

CH

TT

ZT

ZTT

TT

++

−+−=

1

11ε11

1

++

−+

−=

ZTTTZT

TTT

COP C

H

CH

c

ρκκρκ )(

22

eLT

SSZ+

==

S - ThermopowerκT - Total thermal conductivityκL - Lattice thermal conductivityκe - Electronic thermal conductivityρ - Electrical resistivity

TE Figure of Merit ZT

21

ZT Mismatch – Device Challenge

1. X. Shi, et al, J. Am. Chem. Soc. 133, 7837 (2011)2. R. Liu, et al., Intermetallics 19, 1747 (2011)3. Y. Wang, et al., Phys. Rev. Lett. 102, 175508 (2009)4. L. Xi, et al.,J. Am. Chem. Soc. 131, 5560, (2009).5. X. Shi, et al., Phys. Rev. Lett. 95, 185503 (2005). 22

Lattice Thermal Conductivity

1. X. Shi, et al, JACS 133, 7837 (2011)2. R. Liu, et al., Intermetallics 19, 1747 (2011)3. Qiu et al., JAP 111, 023705 (2012)4. Cho et al., Acta Mater. 60, 2104 (2012)

n-typep-type

p-type has higher κL, especially at high temperatures 23

Bipolar Thermal Conductivity- A Hallmark of Electron-Hole Coupling in Semiconductors

Electrons and holes (excitons) are coupled at hot end, diffuse collectively through the material, annihilate at the cold end – give up energy ~ Eg

This transport of energy is in addition to that carried by electrons and holes

This has been reported in many semiconductors – Bi2Te3, Si, Ge, PbTe..

∆VCOLD end HOT end

∆T

24

1. B. Davydov and I. Shmushkevitch, Usp. Fiz. Nauk SSSR 24, 21 (1940)2. P. J. Price, Phil. Mag. 46, 1252 (1955)

Bipolar Thermal Conductivity- Theory

For intrinsic Semiconductors n = p, Significant bipolar thermal conductivity when the mobilities of electrons and holes are similar

25

Bipolar Effect for Heavily DopedSemiconductors a Longstanding Puzzle Surprisingly in the case of heavily doped semiconductors, the

mechanisms for this seemingly simple process remains a puzzle after decades of research

For example, in an n-type materials n >> p, the mobilities of the n-and p-type are usually differ by order(s) of magnitude

Noticeable BP effect has been observed in many TE materials

26

κb Numerical Model

We can calculate n(T) and p(T) using semiconductor statistics, provided we have electronic band structure, EF, m*, etc.

27

Mobility vs. n and p- Key Enabling Step

28

Bipolar Thermal Conductivity - Predictable

29

Bipolar Thermal Conduction- a “Conductivity-Limiting” Phenomenon

Dominated by the minority carriers for heavily doped semiconductors

For a given material, variation is negligible

30

Reducing Bipolar Thermal Conductivity- Band Structure Modulation

Conduction band minimum splitting – reduced effective mass (minority carriers)

31

Reducing Bipolar Thermal Conductivity- Nanostructure Effect

Bi2Te3 made by zone-melting (ZM) and melt-spinning/SPS (contains nanoprecipitates)

32

Automotive waste heat recovery and thermal management are ongoing challenges

TE could offer unique technological solutions

Industrial research is driven by application and system requirement, and of course cost

There are still plenty of opportunities for exciting engineering and science.

Conclusions

33