50
Rachel A. Segalman UC Berkeley Organic and Nanocomposite Thermoelectrics Shannon Yee, Jonathan Malen, Kevin See, Jeffrey Urban, Arun Majumdar, and Rachel A. Segalman 1

2010-Segalman

Embed Size (px)

Citation preview

Page 1: 2010-Segalman

Rachel A. Segalman UC Berkeley

Organic  and  Nanocomposite  Thermoelectrics  

Shannon  Yee,  Jonathan  Malen,  Kevin  See,  

Jeffrey  Urban,  Arun  Majumdar,    and  Rachel  A.  Segalman  

1

Page 2: 2010-Segalman

Rachel A. Segalman UC Berkeley

Energy  landscape  

Global Power Capacity ~ 13 Terawatts

From Fossil fuels

Nuclear

Hydroelectric

Geothermal + Wind + Photovoltaic

http://www.doe.gov

Page 3: 2010-Segalman

Rachel A. Segalman UC Berkeley

Power  genera8on  

Tout  >Tambient+100  oC  

Power  =  10  TrillionWaBs  Efficiency  =  Power/Heatin~  40%  Heatin  =  25  TW  Heatlost  =  15  TW  

Hot  (Thot)  

Engine  

Heatin  

Power  

Cold  (Tambient)  

Heatlost  

Global Power Capacity ~ 13 Terawatts

From Fossil fuels

Nuclear

Hydroelectric

Geothermal + Wind + Photovoltaic

Page 4: 2010-Segalman

Rachel A. Segalman UC Berkeley

Power  co-­‐genera8on  Power  =  10  TrillionWaBs  Efficiency  =  Power/Heatin~  40%  Heatin  =  25  TW  Heatlost  =  15  TW  

Hot  (Thot)  

Engine  

Heatin  

Power  

Cold  (Tambient)  

Heatlost  Tout  =  Tambient+100  oC  

Extra  Power  

Heat  cold   Efficiency  ~  3  %  Extra  Power  =  0.45  TW  US  Electrical  Capacity  =  1  TW  (2005)  

Page 5: 2010-Segalman

Rachel A. Segalman UC Berkeley

Trends  in  bulk  materials  

Figure of Merit, ZT:

S σ

S2σ Semiconductors Metals

Carrier concentration

kTSZT σ2

=

Bismuth  Telluride  (low  efficiency,  expensive)  

p n p n

R

Qin

Qout

ZT

Frac

tion

of C

arno

t

Bi 2 Te 3

0

0.2

0.4

0 1 2 3 4 5 σ = Electrical Conductivity

κ = Thermal Conductivity

S = Thermopower=V/ΔT

Page 6: 2010-Segalman

Rachel A. Segalman UC Berkeley

Electrical  conduc8vity  in  metals  

σ(E)

E

Ef

Fermi Distribution

0 1

E

Ef

f(E)

E

Ef

−∂f∂E

⎝⎜⎞

⎠⎟

D(E)

σ = σ (E) dE

0

∫ σ (E)dE =

−e2

3τ E( )v E( )2

D E( ) ∂f0

∂EdE

Page 7: 2010-Segalman

Rachel A. Segalman UC Berkeley

Thermopower  in  metals  

S = − 1

eT

σ E( ) E − E f( )dE0

∫σ

σ(E)

E

Ef

E-Ef

E

Ef

E-Ef E

Ef

σ(E)(E-Ef)

Inequality in the size of these humps results from asymmetry in σ(E) at Ef, which results in S

Page 8: 2010-Segalman

Rachel A. Segalman UC Berkeley

Assymmetry  about  Ef  Insulator

Eave

N-type Semiconductor

Eave

Degenerate Semiconductor

Eave

Metal

Ef

E

Eave

Increasing S

Increasing σ Applies as a function of doping in bulk conjugated polymers, too

(Review: Ali Shakouri and Suquan Li, 1999)

Page 9: 2010-Segalman

Rachel A. Segalman UC Berkeley

Trends  in  Bulk  Materials  

Figure of Merit, ZT:

Semiconductors Metals

Carrier concentration

σ = Electrical Conductivity

κ = Thermal Conductivity

S = Thermopower=V/ΔT

kTSZT σ2

=

Bismuth  Telluride  (complex  processing,  expensive)  

p n p n

R

Qin

Qout S σ

S2σ Semiconductors Metals

Carrier concentration

Page 10: 2010-Segalman

Rachel A. Segalman UC Berkeley

kTSZT σ2

=

Strategies  to  achieve  high  ZT  

Heterostructures

S M

M S Incident Energy

Reflected Energy

Transmitted Energy

Interfaces decrease k Large asymmetry in σ(E) increases S

σ(E) of Metal

Effective σ(E)

Met

al

Ef

E

Sem

icon

duct

or

M S

Mahan, Woods, Phys. Rev. Lett (1998)

Page 11: 2010-Segalman

Rachel A. Segalman UC Berkeley

Majumdar, Science 303, 777 (2004)

Venkatasubramanian et al. Nature 413, 597 (2001)

Bi2Te3/Sb2Te3 Superlattices 2.5-25nm

Recent  history  

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

AgPb18SbTe20

AgSb rich

Harman et al., Science 297, 2229 (2002)

PbSeTe/PbTe QD Superlattices

Page 12: 2010-Segalman

Rachel A. Segalman UC Berkeley

Organic-­‐inorganic  hybrid  junc8ons  

•  Scalable  Manufacturing  •  Poten8ally  Inexpensive  Components  

•  Durable/Flexible  Devices  •  Tunable  electronic  proper8es  

•  Bulk  polymers  follow  the  same  trends  as  other  bulk  materials  (S  and  σ  opposing)  

Review: Shakouri and Li (1999) Images from Forrest et al.

Page 13: 2010-Segalman

Rachel A. Segalman UC Berkeley

Molecular  thermoelectrics?  

Heterostructures

S M M S M

Heterostructures

M M M

Physics Unexplored

Inexpensive thermoelectric materials!

Page 14: 2010-Segalman

Rachel A. Segalman UC Berkeley

Metal-­‐molecule  junc8ons  have    unique  poten8al  

•  Metal-­‐molecule  junc8ons  have  unique  proper8es  coming  from  mismatched  DOS.  

•  Predic8on  of  junc8on  proper8es  is  the  subject  of  intense  research    

Page 15: 2010-Segalman

Rachel A. Segalman UC Berkeley

I = e

πτ E( )

−∞

∫ fHot E( )− fCold E( )⎡⎣ ⎤⎦dE

fCold E( )

0 1

E

Ef

fHot E( )

f E( )

E

fHot E( )− fCold E( )

Ef

SJunction = −

π 2kB2T

3e

∂ ln τ Ef( )( )∂ Ef( )

E

Ef

I

Landauer Formula: M. Buttiker, Y. Imry, R. Landauer, and S. Pinhas, Phys. Rev. B, 31, 6207 (1985)

Sommerfield Expansion: P.N. Butcher, J. Phys. Chem. 2, 4869 (1990)

τ(E)

E

Ef

HOMO

LUMO

V

Au Au Hot Cold

HS SH

Page 16: 2010-Segalman

Rachel A. Segalman UC Berkeley

Transport  in  Metal-­‐Molecule-­‐Metal  hybrids  

Energy  

LUMO  

HOMO  

EF  

EF  

G  Slope~-­‐S  

V=-­‐SΔT  

Hot   Cold  

Paulsson  and  Data,  Phys.  Rev.  B    67,  241403  (2003)  

Page 17: 2010-Segalman

Rachel A. Segalman UC Berkeley

Tip approach Tip withdraw

Tip speed: 2 – 40 nm/s

Mica Gold (200 nm)

Gold STM tip

Octanedithiol (ODT) HS SH

Single and multi-molecule Structure-property relationships

B.Xu et.al, Science, 301, pp 1221-1223, 2003 Collaboration: Prof. Arun Majumdar (ARPA-E)

Page 18: 2010-Segalman

Rachel A. Segalman UC Berkeley

0

4

3

2

1

0.5 nm

Con

duct

ance

(Go)

Distance 3Go2Go1Go

Cou

nt (a

u)Conductance (Go)

Go = 2e2

h = 77µS

12.9 kΩ 1

=

> 1000 times

Conductance  of  Au-­‐Au  point  contact  

Conductance Traces Histogram

Maximum Conductance through an energy level

Page 19: 2010-Segalman

Rachel A. Segalman UC Berkeley

Sta8s8cal  analysis  of  molecules  

0

1 nm

Distance

Conductance Traces

Ω==

KheGo 9.12

12 2

HS SH (CH2)6

Jang, Reddy, Majumdar, Segalman, Nano Letters (2006)

Histogram from Last Steps of >2000 traces

R~ 45 MΩ

0 2 X 10-31.5 X 10-31 X 10-35 X 10-4

Cou

nts

(au)

Conductance (Go)Conductance (10-4 Go)

5 10 15 20

Con

duct

ance

(10-

4 Go)

4

8

12

16

Page 20: 2010-Segalman

Rachel A. Segalman UC Berkeley

NH2H2N

H2N NH2

H2N NH2

H2N NH2

OCH3

H3CO

H2N NH2

Cl

Cl

BDA

DBDA

TBDA

DMDBDA

DCDBDA

H2NNH2

H2NNH2

H2N NH2

HAD (C6)

ODA (C8)

DDA (C10)

Molecular  structure/property  rela8onship  

Jang, S. Y.; Reddy, P.; Majumdar, A.; Segalman, R. A. Nano Letters 2006. Similar to: Venkataraman, L. et al. Nature (2006)

βaromatic =0.46 per A

βalkane =0.98 per A

0 5 10 15 20 10

15

20

25

DDA ODA

HDA TBDA

DBDA BDA

ln (R

esis

tanc

e)

Length (A o

)

Aromatics Alkanes

o

o

o

o

Page 21: 2010-Segalman

Rachel A. Segalman UC Berkeley

Length  dependence  of  resistance    Tunnel  junc8ons  

l

Eb

R~ exp(ßN)

) 2 ~ exp( ) ( l mE

E b - τ

Page 22: 2010-Segalman

Rachel A. Segalman UC Berkeley

STM  thermopower  measurements  

22

STM  Tip  withdraw  STM  Tip  approach  

Tip  Movement  

Tambient+ΔT  

Ambient T Metal STM Tip

Hot  Au  Substrate  

Current Amplifier

Voltage Amplifier

Tambient  

Page 23: 2010-Segalman

Rachel A. Segalman UC Berkeley

Experimental  measurement  of  thermopower  

STM  Tip  approach   STM  Tip  withdraw  

Tip speed: 2 – 40 nm/s Reddy,  Jang,  Segalman,  Majumdar,  Science  (2007)  

Page 24: 2010-Segalman

Rachel A. Segalman UC Berkeley

2 2 1 ( ( ))3 ( )

f

Bjunction

E E

k T ESe E E

π ττ =

∂= −

Paulsson & Datta, Phys. Rev. B (2003)

G ~ 2e2

hτ (E) E=E f

Electronic  Structure  of  BDT    

HOMO LUMO

HOMO LUMO

10 -2

10 -1

10 0

Tran

smis

sion

, τ (E)

-12 -11 -10 -9 -8 -100 -80 -60 -40 -20

0 20 40 60

Ther

mop

ower

, S ( µ

V/K

)

Energy, E (eV)

SBDT= 7.2 µV/K

~1.2eV

Fermi Level

Fermi Level

G

Reddy, Jang, Segalman, Majumdar, Science (2007)

Page 25: 2010-Segalman

Rachel A. Segalman UC Berkeley

Temperature  dependent  fluctua8ons  

•  DistribuNons  broaden  at  higher  temperatures  

•  Amount  of  change  is  surprising    

Page 26: 2010-Segalman

Rachel A. Segalman UC Berkeley

Tip

App

roac

hes

to C

ondu

ctan

ce S

etpo

int

Mor

e M

olec

ules

in th

e Ju

nctio

n

For  N  molecules  each  with  transmission  τ1(E)  

GN =G0τN E( )

E=Ef

= NG0τ1 E( )E=Ef

= NG1

SN = −π 2kB

2T3e

1Nτ1 E( )

∂ Nτ1 E( )( )∂E

E=Ef

= S1

Page 27: 2010-Segalman

Rachel A. Segalman UC Berkeley

STM approach STM withdraw STM withdraw

µ1 µ2

Devia8on  between  µ’s  results  from  Junc8on-­‐to-­‐Junc8on  Varia8ons  

Devia8on  about  a  single  µ  results  from  Junc8on  Evolu8ons  (hard  to  measure)  

Two  observed  fluctua8ons  in  measurements  

Page 28: 2010-Segalman

Rachel A. Segalman UC Berkeley

Insight  from  Fluctua8ons  The  FWHM  of  Increases  with  ΔT  

ΔSΔT =VFWHM ΔSS

=2.3 ± .37.7 ± .5

= 0.30 ± .04

Page 29: 2010-Segalman

Rachel A. Segalman UC Berkeley

ΔEEvo

Ef − EHOMO( )~ 0.09 ± .04

ΔEJ −to−J

Ef − EHOMO( )~ 0.21 ± .06

• Junc8on  Evolu8ons  are  unobservable.      

• Junc8on-­‐to-­‐Junc8on  Varia8ons  are  the  primary  source  of  devia8on    

Rela8ve  importance  of    evolu8ons  versus  varia8ons  

Page 30: 2010-Segalman

Rachel A. Segalman UC Berkeley

τ E( ) ≈ Γ2

E − EHOMO( )2 + Γ2

ΔSS≈

ΔEE

f− E

HOMO

≈ 0.30 ± .04

Assume  τ(E)  is  Lorentzian:  

ΔS ≈ ΔE∂S∂E

E =Ef

+ ΔΓ∂S∂Γ

E =Ef

Assume  ∆S  is  a  Perturba8on  to  S:  

ΔΓ  ΔE  ΔE  

Page 31: 2010-Segalman

Rachel A. Segalman UC Berkeley

Effect  of  binding  geometry  has  even  stronger  effect  on  conductance  

•  Experimental  conducNvity  full  width  half  max  ~47%  •  DFT  of  15  possible  binding  geometries  suggests  that  

fluctuaNons  are  due  to  binding  not  molecular  conformaNon  

Venkataraman, Nano Letters 2006 Quek et al, Nano Letters, 2008

Page 32: 2010-Segalman

Rachel A. Segalman UC Berkeley

Transport  devia8ons  increase  with  length  of  molecule  

Page 33: 2010-Segalman

Rachel A. Segalman UC Berkeley

Tuning  electrical  conductance  and  thermopower  

HOMO LUMO

HOMO LUMO

10 -2

10 -1

10 0

Tran

smis

sion

, τ (

E)

-12 -11 -10 -9 -8 -100 -80 -60 -40 -20

0 20 40 60

Ther

mop

ower

, S ( µ V/

K)

Energy, E (eV)

~1.2 eV

Fermi Level

Fermi Level

S

σ κ

σ  =  Electrical  Conduc8vity  

k  =  Thermal  Conduc8vity    

S  =  Thermopower  

kTSZT σ2

=G

S

Bahe8,  Malen,  Doak,  Majumdar,  Segalman,  Nano  LeBers  (2008)  

Page 34: 2010-Segalman

Rachel A. Segalman UC Berkeley

SBDT = 7.7±0.5 µV/K

Electron  withdrawing  subs8tuents  

ClCl

ClClHS SH

Page 35: 2010-Segalman

Rachel A. Segalman UC Berkeley

SBDT = 7.7±0.5 µV/K

Electron  dona8ng  subs8tuents  

HS SH

Page 36: 2010-Segalman

Rachel A. Segalman UC Berkeley

Effects  of  electrode-­‐molecule  bond  

SAu-BDCN-Au ~ -1.3 ± 0.3( )µV / K

N-­‐type  Molecular  Junc8ons    

CNNC

Page 37: 2010-Segalman

Rachel A. Segalman UC Berkeley

Tuning  Electrical  Conductance  and  Thermopower  

HOMO LUMO

HOMO LUMO

10 -2

10 -1

10 0

Tran

smis

sion

, τ (

E)

-12 -11 -10 -9 -8 -100 -80 -60 -40 -20

0 20 40 60

Ther

mop

ower

, S ( µ V/

K)

Energy, E (eV)

~1.2 eV

Fermi Level

Fermi Level

S

σ κ

σ  =  Electrical  Conduc8vity  

k  =  Thermal  Conduc8vity    

S  =  Thermopower  

kTSZT σ2

=G

S

Bahe8,  Malen,  Doak,  Majumdar,  Segalman,  Nano  LeBers  (2008)  

Page 38: 2010-Segalman

Rachel A. Segalman UC Berkeley

2 2

Paulsson and Datta, PRB (2003)

1 ( )3 ( )

f

B

E E

TV EST e E E

π κ ττ

=

⎛ ⎞∂= − = − ⎜ ⎟Δ ∂⎝ ⎠

G ~ 2e2

hτ (E) E=E f

Scaling to materials systems

Fermi Levelinorg

G

Page 39: 2010-Segalman

Rachel A. Segalman UC Berkeley

Design  of  Molecular  Thermoelectrics  

SH HS Gold Gold

Ene

rgy (Ef)

Frequency, ω Den

sity

of S

tate

s, D

(ω) Solid

Molecules

Large Mismatch in the Phonon Density of States

Met

al

Mol

ecul

e Met

al

Δ  

Den

sity  of  States  

Large S

kTSZT σ2

=LUMO HOMO

Page 40: 2010-Segalman

Rachel A. Segalman UC Berkeley

Nanocomposite  Design  

•  High  σ /low k Polymer

•  High S inorganic nanoparticles

•  Processing: –  Water Dispersed –  Amenable to

solution processing

σ∼100 S/cm

S~500 µV/K

Te  Te  

Te  

Te  

Nano Letters ASAP 2010

Page 41: 2010-Segalman

Rachel A. Segalman UC Berkeley

System  Design  

41

Fermi Levelinorg G

Page 42: 2010-Segalman

Rachel A. Segalman UC Berkeley

Te rods, CTAB coated

Synthesis of Composites

Na2TeO3 + PEDOT:PSS + Na2TeO3 + (CTAB) +

H2O

Xi, G Crystal Growth & Design 2006, 6, 2567-2570."

In-Situ Synthesis of Polymer Coated Te Rods

Pedot:PSS Passivation

Te rods, polymer coated!

Polymer

Te  

Nano Letters ASAP 2010

Page 43: 2010-Segalman

Rachel A. Segalman UC Berkeley

Porous  Nanocomposites  

Page 44: 2010-Segalman

Rachel A. Segalman UC Berkeley

Morphology  of  Composite  

Films easily deposited or spin-coated Films are porous, but fully connected

Voc

TH TC Nano Letters ASAP 2010

Page 45: 2010-Segalman

Rachel A. Segalman UC Berkeley

Page 46: 2010-Segalman

Rachel A. Segalman UC Berkeley

Room  Temp  Proper8es  of  Nanocomposite  Films  

Page 47: 2010-Segalman

Rachel A. Segalman UC Berkeley

Room  Temp  Proper8es  of  Nanocomposite  Films  

Highest reported ZT for aqueous processed nanocomposite

• Remarkable combination of properties • Higher σ than conducting polymer • High S of nanocrystals

Nano Letters ASAP 2010

Page 48: 2010-Segalman

Rachel A. Segalman UC Berkeley 48

Summary  

•  Thermopower  from  molecular  juncNons    

•  Measured  thermopowers  provide  insight  towards  molecular  juncNon  physics  

•  ZT  ~  0.2  for  a  nanocomposite  processed  from  H2O  at  room  temperature  

•  Novel  pla\orm  for  soluNon  processable  thermoelectrics    

•  Large  space  to  explore  transport  and  for  opNmizaNon  of  ZT  towards  1    

Page 49: 2010-Segalman

Rachel A. Segalman UC Berkeley 49

Acknowledgements  •  DOE-­‐BES  LBNL  Thermoelectrics  Program  

•  FaciliNes  at  the  Molecular  Foundry  •  Dr.  Shaul  Aloni  (TEM)  

•  Dr.  Arun  Majumdar  

Page 50: 2010-Segalman

Rachel A. Segalman UC Berkeley

Segalman  Group