nanostructured thermoelectric materials: from basic ... · nanostructured thermoelectric materials:...

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Nanostructured Thermoelectric Materials: From Basic Physics to Applications

Gang Chen

Nanoengineering GroupRohsenow Heat and Mass Transfer Laboratory

Massachusetts Institute of Technology

Cambridge, MA 02139

Email: gchen2@mit.eduhttp://web.mit.edu/nanoengineering

Acknowledgments:J.P. Fleurial

DOE, NSF, and Industry

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Outline

•

Why nanocomposites?•

What we have achieved?

•

What we do not understood?•

Application opportunities

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Nanoscale Effects for Thermoelectrics

Phonons=10-100 nm

=1 nm

Electrons=10-100 nm=10-50 nm

Electron Phonon

Interfaces that Scatter Phonons but not Electrons

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

0

20

40

60

80

10 1 10 2 10 3 10 4

THER

MA

L C

ON

DU

CTI

VITY

(W/m

K)

LAYER THICKNESS (Å)

BULK

SPECULAR (p=1)

p=0.95

p=0.8DIFFUSE (p=0)

Yao (1987)Yu et al. (1995)0

20

40

60

80

10 1 10 2 10 3 10 40

20

40

60

80

0

20

40

60

80

10 1 10 2 10 3 10 410 1 10 2 10 3 10 4

THER

MA

L C

ON

DU

CTI

VITY

(W/m

K)

LAYER THICKNESS (Å)

BULK

SPECULAR (p=1)

p=0.95

p=0.8DIFFUSE (p=0)

Yao (1987)Yu et al. (1995)

Heat Conduction Mechanisms in Superlattices

Periodic Structures Are Not Necessary, Nor Optimal!

Major Conclusions:

•

Ideal superlattices do not cut off all phonons due to pass-bands

•

Individual interface reflection is more effective

•

Diffuse phonon interface scattering is crucial

10-1

100

101 102 103

Capinski et al. 1999Capinski et al. 1999

Yao 1987 Yu et al. 1995

Nor

mal

ized

The

rmal

Con

duct

ivity

Period Thickness (Å)

T=300KAlAs/GaAs

In-Plane

Cross-Plane

LD

LD

Ideal Superlattices

In-Plane

Cros

s-Pl

ane

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Nanocomposites Approach

–

Increase interfacial scattering by mixing nano-sized particles.

–

Enable batch fabrication for large scale application.

Nanoparticle(a) (b)

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Low-Cost Synthesis

AGraphite cylinder

Graphite piston

Current for heating

Force for pressing

Sample powder A

Graphite cylinder

Graphite piston

Current for heating

Force for pressing

Sample powder

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Thermoelectric Properties: Bi2

Te3

Poudel et al., Science, 320, 634, 2008

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Six

Ge1-x

System

Joshi et al., Nano Letters 8, 4670 (2008)

Wang, et al.

Appl. Phy. Lett. 93, 193121 (2008)

P-type SiGe

N-type SiGe

0 200 400 600 800 10000.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4 RTG SiGe Si80Ge20B5 (run-1) Si80Ge20B5(run-1)- annealed 7 days Si80Ge20B5(run-2)

ZT

Temperature ( oC)

0

5

10

15

20

25

30

35

40

45

200 400 600 800 1000 1200 1400

Ther

mal

Con

duct

ivity

(W.m

-1K-1

)

Temperature (K)

No large cold welded agglomeratesShort Pressure Sintering at 1275 KLong Pressure Sintering at 1475 K

Undoped SiSingle Crystal

Doped Nanobulk Si using only "cold

welded" agglomerates of nanoparticles

Undoped Nanobulk Si

Doped Nanobulk Si

Bux et al., Adv. Func. Mat., 2009

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Understanding Electron and Phonon Transport in Random Nanostructures

5 nm

Phonons: •

What is the mean free path in bulk materials Phonon-phonon Dopants Alloy Electron scattering Defects

•

How phonons are scattered at an interfaces Interface roughness Crystallographic directions Anharmonic effect

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Molecular Dynamics Simulation of Phonon Mean Free Path

10-3

10-2

10-1

100

101

102

103

0

10

20

30

40

50

60

70

80

90

100

Mean Free path (m)

Per

cen

t A

ccu

mu

lati

on

CallawayAseMD

Henry and Chen, J. Comp. Theo. Nanosci., 5, 141 (2008).

Si

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.00

0.05

0.10

0.15

0.20

0 30 60 90

SV

TO

SV

WA

VE S

V TO

L WA

VE

INCIDENT ANGLE (DEGREE)

REFLECTIVITYTRANSMISSIVITY

REFLECTIVITY

TRANSMISSIVITY

Interface Reflectivity and Transmissivity

L(v)1(v)2

sv

L

svsv

We know how to do it only in the continuum limit, with perfect interfaces!

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Thermal Interface Resistance Between Perfectly Contact Solids

Cahill et al., JAP, 2003

•

No model can predict interface thermal conductance except at very low temperatures.

•

Phonon transmissivity and reflectivity at interfaces are crucial parameters to model heat conduction in thermoelectric materials.

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Electron Transport

•

Where thermoelectric effect happens?

•

Interfacial barrier height

•

Momentum conservation at interfaces

•

Scattering inside particles

•

Energy dependency of scattering

5 nm

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Thermionic emission at grain boundaries

SiGe nanocomposite,

Eb

=0.2eV, ND

=4x1020

cm-3

Need: Em

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Nanoparticle/Nanograin Scattering

D/2carrier

Dominant electron wavelengths

* ~ 6 10nmc

hm v

Grain size

m~ 20 n30D

Grain boundary thicknessm~1ntDebye screening length

1/2

2 ~ 0.5 m1 nDkTLe N

Nanograin

D/2D/2carrier

Dominant electron wavelengths

* ~ 6 10nmc

hm v

Grain size

m~ 20 n30D

Grain boundary thicknessm~1ntDebye screening length

1/2

2 ~ 0.5 m1 nDkTLe N

Dominant electron wavelengths

* ~ 6 10nmc

hm v

Grain size

m~ 20 n30D

Grain boundary thicknessm~1ntDebye screening length

1/2

2 ~ 0.5 m1 nDkTLe N

Nanograin

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Electrons: MFP vs λ

5 10 15 20 25 301

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6

MF

P (

nm

)

Electron wavelength (nm)

Minnich et al., Energy & Env. Sci., 2009.

SiGe

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Potential Applications

Industrial Waste Heat (DOE/ITP)

Renewables

Transportation

Main AC: TE Local cooling of seats: TE

Catalytic converter: TE

Radiator: TE

Main AC: TE Local cooling of seats: TE

Catalytic converter: TE

Radiator: TE

BuildingsCombustor: TE

Solar Panel: PVSolar Panel: PV

Furnace: TE

Refrigerator

HVAC: TE

And Entropy is Generated

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

System Consideration

TobjectTcold

Thot

TambientT

Tte

Sometimes, thermal systems more expansive

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Heat to Electricity Recovery from Gas Stream

Hot Gasm, Th

,i

Hot GasTh

,o

Heat Transfer Surfaces

Coolant

InsulationHot Side TemperatureTH

Cold Side TemperatureTc

•

For thermoelectric devices, TH

higher is better•

However, maximum heat intercepted from hot gas stream, mcp

(Th,i

-TH

), decreases with TH

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

Efficiency Expectations• Th,i

=600 oC, Tc

=50 oC• Optimal Hot Side Temperature ~ 270 oC

0

0 . 0 5

0 . 1

0 . 1 5

0 . 2

0 . 2 5

0 . 3

0 .5 1 . 5 2 . 5

F I G U R E O F M E R I T ZT

EFF

ICIE

NCY

Ideal Device 50-600 oC

Ideal Device ~50-270 oC

Line: System Efficiency 50-270 oCDots: System Efficiency 50-250 oC

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT

0

0.05

0.1

0.15

0.2

0.25

0.3

0.5 1 1.5 2 2.5 3

EFFI

CIE

NC

Y

FIGURE OF MERIT ZT

Ideal Device 50-600 oC

Ideal Device 50-270 oC

Line: System Efficiency 50-270 oCDots: System Efficiency 50-250 oC

Ideal System

0

0.05

0.1

0.15

0.2

0.25

0.3

0.5 1 1.5 2 2.5 3

EFFI

CIE

NC

Y

FIGURE OF MERIT ZT

0

0.05

0.1

0.15

0.2

0.25

0.3

0.5 1 1.5 2 2.5 3

EFFI

CIE

NC

Y

FIGURE OF MERIT ZT

Ideal Device 50-600 oC

Ideal Device 50-270 oC

Line: System Efficiency 50-270 oCDots: System Efficiency 50-250 oC

Ideal System

Locally Optimized Systems

TEG2

T2

TEG3

T3

TEG1

T1

TEG2

T2

TEG3

T3

TEG1

T1

Hot Gas • 1-3% Absolute Efficiency• Engineering Room• Materials Development

––WARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MITWARREN M. ROHSENOW HEAT AND MASS TRANSFER LABORATORY, MIT