Electrical power from heat: All-scale hierarchical thermoelectrics with and without earth-abundant

materials

Sponsored by the

Mercouri Kanatzidis

Northwestern University

Department of Energy

Lidong Zhao

Yeseul Lee

Rachel Korkosz Thomas Chasapis

Kanishka Biswas

GROUPCollaborators

n Tim Hogan, MSUn S. D. (Bhanu) Mahanti, MSUn Ctirad Uher, Michigann Simon Billinge, Columbian Eldon Case, MSUn Vinayak Dravid, NUn Art Freeman, NUn Jos Heremans, OSUn Chris Wolverton, NUn Ray Osborn, Argonnen Stephane Rosenkranz, Argonnen Ken Gray, Argonnen David Seidman, NU n John Mitchell, Argonnen Duck Young Chung, Argonnen Theodora Kyratsi, U Cyprus nVinayak Dravid, NU

Thermoelectricity - known in physics as the "Seebeck Effect"

• In 1821, Thomas Seebeck, a German physicist, twisted two wires of different metals together and heated one end.

• Discovered a small current flow and so demonstrated that heat could be converted to electricity.

www.worldofenergy.com.au/07_timeline_world_1812_1827.html

Seebeck Effect

www.dkimages.com/discover/DKIMAGES/Discover/Home/Science/Physics-and-Chemistry/Electricity-and-Magnetism/General/General-18.html

chem.ch.huji.ac.il/history/seebeck.html

Heat to Electrical Energy Directly

Up to 20% conversion efficiency with right materials

http://www.dts-generator.com/

TE devices have no moving parts, no noise, reliable

Thermopower S = ΔV/ΔT

hot cold

Thermoelectric applications

• Waste heat recovery • Automobiles• Over the road trucks• Marine• Utilities• Chemical plants

• Space power• Remote Power Generation• Solar energy• Geothermal power

generation• Direct nuclear to electrical

U.S. Energy Flow, 2009

~65% of energy becomes waste heat, ~10% conversion to useful forms can have huge impact on overall energy utilization

http://www.eia.doe.gov/emeu/aer/

14

Figure of Merit and Conversion Efficiency

ZT S2

total

T

S2Power factor

Total thermal conductivity

electrical conductivity thermopower

ZT S2

total

T

S2Power factor

Total thermal conductivity

electrical conductivity thermopower

700 K

900 K

0.0 0.5 1.0 1.5 2.0 2.5 3.00.0

0.2

0.4

0.6

ZT

Tcold= 300K

What about thermal conductivity?

• Diamond 1600 W/mK

• Cu 400 W/mK

• PbTe 2.2 W/mK

• Wood 0.2 W/mK

TS

ZT2

“Power factor”

18 19 20 210.0

0.3

0.6

0.9

1.2

Po

we

r fa

cto

r

log n

conductivity, PF

Seebeck, S

, electronic

Why finding a “good thermoelectric” (ZT > 1) is hard! Contra-indicated properties

Leading thermoelectric materials

• Bi2Te3-Sb2Te3 (ZT~1) (300K)• PbTe: ZT~0.8 at 800 K (n-type)• AgSbTe2-GeTe (TAGS): ZT~1.2, 700 K (p-type)• Half-Heusler alloys (ZT~0.8, 900K)• Skutterudites (M1, M2, M3)Fe4Sb12 (ZT~1.4,

900K, n-type)• Mg2(Si,Sn) (ZT~1, 1000 K)• Nanostructured PbTe, (ZT~2.2)

19

( )2/1

2/3

max+µ r

latt

z

yx

em

mmT

Zk

tg

What kinds of materials make the best thermoelectrics?

m= effective masst=scattering timer= scattering parameterlatt= lattice thermal conductivityT = temperature

= band degeneracyLarge comes with(a) high symmetry e.g. rhombohedral, cubic(b) off-center band extrema

Isotropic structure Anisotropic structure

k

E

k

E

k

Ef

For acoustic phonon scatteringr=-1/2

Complex electronic structure

20

Multiple valleys….are better

21

Best thermoelectric materials

Developed new bulk thermoelectric materials with record ZTmax

n-type: ZTmax ~1.6 at 700K -p-type: ZTmax ~1.7 at 700K -

0

0.5

1

1.5

2

400 600 800 1000 1200Z

TTemperature (K)

Bi2Te

3

Na0.95

Pb20

SbTe22

Zn4Sb

3

Ag0.5

Pb6Sn

2Sb

0.2Te

10

CeFe4Sb

12

PbTe Yb14

MnSb11

(AgSbTe2)0.15

(GeTe)0.85

p-type Materials

Ce0.28

Fe1.5

Co2.5

Sb12

p-type

0

0.5

1

1.5

2

400 600 800 1000 1200

ZT

Temperature (K)

Bi2Te

3

Pb18

Ag0.86

SbTe20

PbTe

CoSb3

La2Te

3

SiGe

PbTe-PbS(8%)

n-type Materials

Ba0.30

Ni0.05

Co3.95

Sb12

Mg2Si

0.6Sn

0.4Mg

2Si

0.4Sn

0.6

n-type

Major discovery: self-assembled nanodots in bulk materials responsible for record ZT’s

LAST

Hsu et al, Science, 303, 818 (2004)

Endotaxial nanostructuresEndotaxy: Coherent lattice matched placement of one crystal inside another

K. F. Hsu, etal Science 2004, 303, 818-821.P. F. P. Poudeu, etal Angew. Chem. Int. Ed. 2006, 45, 3835-3839.J. Androulakis, et al J. Am. Chem. Soc. 2007, 129 (31), 9780-9788.K. Biswas, etal Nature Chemistry 2011, 3, 160-166.K. Biswas, etal Nature, 2012, 419, 414-418.

Key aspects:InterfacesStrainBand offsetsStability

matrix

electronic band structure of PbTe

a≈6.45 Å (300K)

m*Σ (~2m0) >> m*L(~0.2m0)

Valence band is multiple peaks

Introducing strain into PbTe

PbTe-x%SrTe Transmission Electron Microscopy

L Σ

E

(eV

)

~0.30 eV

T = 500 K

VB

CB

Light hole band Heavy hole band

Thermal excitation of holes to Σ band

Valence bands of PbTe….

Rising temperature

300 400 500 600 700 800 900

6

9

12

15

18

21

24

27

.S

2 (W

/cm

K2 )

T (K)

300 400 500 600 700 800 900

50

100

150

200

250

300

S (V

/K)

T (K)

a b

c d

fe

VB

CB

PbTe PbTeSrTe

+ + + + + +

L Σ

E (

eV) ~0.3 eV

300 400 500 600 7000

100

200

300

400

500

(cm

2 V-1S

-1)

T (K)

Optimizing charge transportThrough band alignment

nanostructures mesostructures

Nano-scale, meso-scaleSubmicron grains

300 400 500 600 700 800 900

0.4

0.8

1.2

1.6

2.0

2.4

2.8

3.2

lat (

W/m

K)

T (K)

600 700 800 900

0.4

0.8

1.2

SPS

lat (

W/m

K)

T (K)

Ingot

Thermal conductivity PbTe-x%SrTe

K. Biswas, Jiaqing He, I. D. Blum, C-I Wu, T. P. Hogan, D. N. Seidman, V. P. Dravid & M. G. Kanatzidis Nature 2012, 489, 414–418

300 400 500 600 700 800 9000.81.21.62.02.42.83.23.64.04.4

tota

l (W

/mK

)

T (K)300 400 500 600 700 800 900

0

10

20

30

S2 (

W/c

mK

2 )

T (K)

300 400 500 600 700 800 9000

500

1000

1500

2000

2500

4% SrTe, 2% Na: SPS 2% SrTe, 2% Na: SPS 0% SrTe, 2% Na: SPS 4% SrTe, 2% Na: Ingot 2% SrTe, 1% Na: Ingot[14]

(S

/cm

)

T (K)300 400 500 600 700 800 900

0

50

100

150

200

250

300

350

S (V

/K)

T (K)

a b

c d

f

Thermal conductivity PbTe-x%SrTe

300 450 600 750 9000.0

0.4

0.8

1.2

1.6

2.0

2.4

4% SrTe, 2% Na: SPS 2% SrTe, 1% Na: Ingot[14] 0% SrTe, 2% Na: Ingot

ZT

T, K

Mesoscale

Nanoscale

Atomicscale

ZT ~ 2.2 ZT ~ 1.7 ZT ~ 1.1

Increasing efficiency

All-scale hierarchical architecture

1 cm

All length scales: record high ZT

K. Biswas, Jiaqing He, I. D. Blum, C-I Wu, T. P. Hogan, D. N. Seidman, V. P. Dravid & M. G. Kanatzidis Nature 2012, 489, 414–418

What is the proof that nanostructures reduce thermal conductivity?

Model PbTe – PbS system for nanostructured TEs

J. D. Gunton and M. Droz, Lecture Notes in Physics: Introduction to the Theory of Metastable and Unstable States, Vol. 183 (Springer-Verlag, Berlin, Heidelberg, New York, Tokyo, 1983) pp. 1-13. Leute, V., Volkmer, N. Z. Phys. Chem. NF., 144 1985, 145

600

700

800

900

1000

1100

mol. % PbTe ►

PbS PbTe

º C

0 50 60 9070 80302010 40 100

Miscibility Gap Chemical Spinodal

Spinodal Decomposition Nucleation &

Growth

Solid Solution

Nucleation and Growth

(PbTe)0.92(PbS)0.08

35

300 400 500 600 7000.20.30.40.50.60.70.80.91.0

(PbTe)0.92

(PbS)0.08

Run 1 Heating Run 1 Cooling Annealed Sample

lat,

W/m

K

Temperature, K

~50% Reduction in κlat

(PbTe)0.92(PbS)0.08

Significant reduction in κlat

PbTe0.92S0.08 (PbTe)0.92(PbS)0.08

Solid solutionNanostructured

PbTe0.92S0.08

heat

We can see the effect of nanoscale precipitation of PbS in situ on the lattice thermal conductivity.

Solid solution

Nanostructured

S. Girard, Jiaqing He

Te free?

PbS: the cheapest thermoelectric

Nanostructuring PbS with second phases

Binary phase diagram of PbS-Bi2S3(Sb2S3)

Garvin P. F., Neues Jahrb. Mineral., Abh., 118, 235(1973)

PbS-Bi2S3 phase diagram

Nucleation and growth

PbS with second phases without doping

n-type PbS with second phases

Second phases: Bi2S3, Sb2S3

~ 0.80 @ 723 K

300 400 500 600 7000

500

1000

1500

2000

2500

(S

cm-1)

PbS PbS+1% PbCl2 PbS+1% Bi2S3+1% PbCl2 PbS+2% Bi2S3+1% PbCl2 PbS+3% Bi2S3+1% PbCl2 PbS+4% Bi2S3+1% PbCl2 PbS+5% Bi2S3+1% PbCl2

Temperature (K)

300 400 500 600 700-400

-300

-200

-100

0

S(

VK

-1)

Temperature (K)

Seebeck independent on second phases

300 400 500 600 7000

3

6

9

12

15

PF

(W

cm-1K

-2)

Temperature (K)

Significantly reduce

300 400 500 600 7000.0

0.2

0.4

0.6

0.8

1.0 PbS PbS+1% PbCl2 PbS+1% Sb2S3+1% PbCl2 PbS+2% Sb2S3+1% PbCl2 PbS+3% Sb2S3+1% PbCl2 PbS+4% Sb2S3+1% PbCl2 PbS+5% Sb2S3+1% PbCl2

Temperature (K)

ZT

PbS with Sb2S3~ 0.78 @ 723 K

n-type PbS with second phases

PbS+1.0 at. % Bi2S3+1.0 at. % PbCl2 PbS + 1.0 at. % Sb2S3 + 1.0 at. % PbCl2

TEM: nanostructured PbS

ZT ~ 1.1 @ 923 K ZT ~ 1.06 @ 923 K

M: normal melting B: Bridgman S: SPS BN coating

Good repeatability !

Nanostructures n-type PbS, ZT=1.1

Pb0.975Na0.025S+3%SrS shows ZT about 1.2 at 923K, Pb0.975Na0.025S+3%CaS shows ZT about 1.1 at 923K, Zhao, L.-D. et. al., JACS. 133(2011)20476. & JACS. 134(2012)7902

P-type Pb0.975Na0.025S-3%CaS/SrS

Both total and lattice κ were reduced by SrS inclusions

Fine grain size, Sr containing precipitates, and no spot diffraction splitting for Pb0.975Na0.025S-3% SrS.

crystallographic alignment between PbS and SrS, strain maps and lattice parameter difference at the interface between PbS and SrS.

TEM of Pb0.975Na0.025S -3%SrS

GROUP

Raising ZT of p-type PbS with endotaxial nanostructuring and valence-band offset

engineering using CdS and ZnS

GROUP PbS is an ideal TE system because high performance in both n-type (ZT~1.1 at 923 K) and p-type (ZT~1.1 at 923 K) can be achieved. Zhao, L.-D. et. al., JACS. 133(2011)20476. & JACS. 134(2012)7902

PbS is promising

Band gap energy levels of the metal sulfides, PbS, CdS, ZnS, CaS and SrS, all in the NaCl structure

n-type

p-type

GROUP

mobility, μ at 920 K

p-type, CdS containing sample shows higher μ

4% MS

CdS

ZnSCaS SrS

PbS

µ (c

m2 /

V-se

c)

40 cm2/V-sec

GROUP

Zhao L.D. et al. JACS, 2012

ZT for PbS system

0.13eV

CdS

e e

phonons

PbS

PbS CdS

EgE’g

minimal valence band

offsetVB

(a)

(b)

ΔE

phonon-blocking/electron-transmitting

~1.3 @923K

GROUP

Hierarchical Length-scale

Architecture:

Implications for “Nanostructured”

Thermoelectrics

Interactions along varied length-scales

Identification of individual microstructure elements in electronic and phonon transport

Tailoring and design of “microstructure”

Panoscopic view of thermoelectrics

Atoms/molecular motifs

Crystal lattice & point defects

Interfaces

Macro-, and device-scale Interfaces

Precipitates & nanoscale defects

Electronic Structure

CrystalStructure

Classical Microstructure

Thin films/multilayersInterfaces

Residual stresses

Macroscale Device Architecture

Angstrom and sub-nm scale

Sub-nm to Nano-scale

Micro-to macro-scale

PbTe-x%SrTe Panoscopic…2.22.2

PbTe-PbS (nanostructured)PbTe-PbSe

NaPb20SbTe20

Conclusions• A panoscopic view is required going forward• Band alignment engineering between nanostructures and

matrix: ZT~2.2 at 900K• Superior properties in p-type PbTe-SrTe achieved through

endotaxial placement of nanoprecipitates– Nanostructures do not reduce the power factor and function

exclusively as phonon scatterers• Large power factor enhancements are needed for continued

ZT increases• High performance in nanostructured PbS (ZT~1.2-1.3 at 900 K)