Controlling Charge Carrier Concentration in Organic Semiconductors

Institut fürAngewandte Photophysik

M. Pfeiffer, J. Blochwitz-Nimoth, G. He, X. Zhou, J. Huang, J. Drechsel, A. Werner, B. Männig, K. Leo

Institut für Angewandte Photophysik, TU Dresden, 01062 Dresden, Germany

www.iapp.de

• Charge carrier control: Doping of organic semiconductors

- Electrical Characterisation- UPS/XPS Spectroscopy: space charge layers

• Application in devices:

Organic light-emitting diodesOrganic solar cells

Outline

• Basics of organic semiconductors

What is an organic semiconductor…..?

Why organic semiconductors…..?

Photovoltaic cellsLight emitters Integrated circuits

Organic Semiconductors: a new class of materials

pzpz

sp2 sp2

σ

σ∗

π

π∗

p lane of thesp 2-orb ita ls

pz-orb ita l π -bond

π -bond

σ -bond

pz-orb ita l

Conjugated π-electron systems

Sp2-hybridised Carbon:

• Spatially extended electronsystems:

• Large dipole moments

• Tunability of energy levels

• Good coupling betweenmolecules in solid state

de loka lis ie rte π −E lektronen

pz

σ∗

sp2

6 x

18 x

LUMO (π*) => EC

HOMO (π) => EV

• saturated π-electron systems• VdW crystals• small π−π-overlap, narrow bands

delocalizedπ-electrons

Molecules with conjugated π-electron system

Film preparation technology: vapor deposition

pz

σ∗

sp2

6 x

18 x

n

n x pz

sp2

VB (π)

CB (π*)

• broad bands• transport limited by

interchain hops

Polymers with conjugated π-electron system

Film preparation technology: solution processing

Inorganic vs. Organic Semiconductors

CB

VB

ED

EA

LUMO

HOMO

≈ 1017Electr. Active impurit. (cm-3)<<10151010-1016 (injection controlled)Charge carrier concentration1015-1018 (doping controlled)

10-6-1 (typ. 10-3)RT mobility (cm2/Vs)102-103HoppingTransport MechanismBand

Organic SemiconductorInorganic Semiconductor

• Charge carrier control: Doping of organic semiconductors

- Electrical Characterisation- UPS/XPS Spectroscopy: space charge layers

• Application in devices:

Organic light-emitting diodesOrganic solar cells

Outline

• Basics of organic semiconductors

• broad bands• small correlation energies (e-h ≈ 4meV) • hydrogen model works

Inorganic Organic

• hopping transport• large correlation (e-h ≈ 1 eV)• polaron effects important

Basics of the doping mechanisms: p-type doping

Dopant Matrix

Quartz monitors

Substrate

p ≈ 10-4 Pa Tevap= 100..400 oCTSubs= -50..150 oCd = 25..1000 nmrM≈ 1 Å/s

Dopant/Matrix ratio of 1:2000 achieved

Institut fürAngewandte Photophysik

Co-evaporation of doped films

Model system for p-doping: ZnPc:F4-TCNQ

Organic Matrix:Zinc-Phthalocyanine (ZnPc)monoclinica = 26.3 Å, b = 3.8 Å, c = 23.9 Å,γ = 94.6o, Z = 4HOMO ≈ -5.3eV

Organic Acceptor:Tetrafluoro-tetracyano-quinodimethane(F4-TCNQ)LUMO ≈ -5.0eV

Institut fürAngewandte Photophysik

e-

ZnPc-F4-TCNQ charge transfer by FTIR

2240 2220 2200 2180 2160

0

10

20

30

40

50

60 >C=C<C Nb2uν32-peak forZ=1 at 2173 cm-1

b2uν32-peakfor Z=1 at 2214 cm -1

b1uν18-peak forZ=0 at 2228 cm-1

b1uν18-peak forZ=1 at 2194 cm-1

trans

mis

sion

[%]

wave number [cm-1]1600 1580 1560 1540 1520 1500

b1uν19-peak for Z=1 at 1501 cm-1

b1uν19-peakfor Z=0 at 1550 cm-1

IR-spectra of a F4-TCNQ - KBr - sample

FF

F F

N

N

N

N

FF

F F

N

N

N

N

10

0

ν−νν−ν

= xZ

Degree Z of the charge-Transfer:

ν0 : neutral moleculeν1 : charged moleculeνx : doped layer

Result in Pc films: Z=1 !

Nominally undoped ZnPc:

≈10-10 S/cm in vacuo

⇒ Doping increases conductivityby orders of magnitude

10-3 10-2 10-110-4

10-3

10-2 ZnPc series 1 ZnPc series 2

T= 30°C

linear dependenceCon

duct

ivity

σ [S

/cm

]

Molar Doping Ratio F4-TCNQ : ZnPc

Conductivity vs. Doping Concentration

• Superlinear behavior ?

• Shallow acceptors ? Coulomb energy too large….

2.6 2.8 3.0 3.2 3.410-4

10-3

10-2

10-1

400:1

100:1

50:135:1

12.5:1

200:1

p/N

µ

1000/T (K)

380 360 340 320 300Temperature [K]

1018

1019

1020

hol

e de

nsity

p (c

m-3) kT

EENTp F µ

µ

−=)(

Hole density is not thermally activated !

M. Pfeiffer et al., Adv. in Sol. State Physics 39, 77 (1999).Institut für

Angewandte Photophysik

Hole densities of ZnPc:F4-TCNQ

Theory: Extension of M.C.J.M. Vissenberg, M. Matters, Phys. Rev. B 57, 12964 (1998)

EF

f(E) g(E)

Density of states:g(E) = g0exp(-E/kBTc)

Variable Range HoppingT < TcEF >> kBTc

Percolation model: σ = σ0 Zc-1

with Zc-1 threshold conductance

Conduction in exponential band tails: Explains superlinear doping efficiency

Assuming shallow acceptors

σ ~ (NA)To/T ; T0 > T

360 320 280 240 2001E-7

1E-6

1E-5

1E-4

1E-3

0.01 theoretical curves

polycrystalline ZnPc

cond

uctiv

ity (S

/cm

)

temperature (K)

experimental data: 1:50 1:100 1:200 1:500

360 320 280 240 200 160

1E-6

1E-5

1E-4

1E-3

0.01

experimental data: 1:150

cond

uctiv

ity (S

/cm

)

temperature (K)

1E-7

1E-6

1E-5

1E-4

1E-3 theoretical curves

polycrystalline ZnPc

mob

ility

(cm

2 /Vs)

material σ (S/cm) Eact (eV) σo(105S/m) To (K) α (Å-1)

ZnPc (polycr.) 5.8 * 10-3 0.18 12 3 485 15 0.37 0.01

ZnPc (amorp.) 4 * 10-4 0.23 11 6 455 15 0.64 0.02

VOPc (polycr.) 2.3 * 10-5 0.32 6 1 485 15 1.00 0.04

TDATA (am.) 5.9 * 10-7 0.34 3 1 515 15 1.39 0.03

σ and Eact for dopant density 1% and T=25°C

Application to field-effect and conductivity data:

Basic parameters are independent of specific material:

Matrix: Naphthalin-tetracarboxylic-dianhydride(NTCDA)

A. Nollau et al., J. Appl. Phys. 87, 4340 (2000)

Donor: Bisethylendithio-tetrathiafulvalene(BEDT-TTF)

N-type doping: conventional approach

charge transfer

HOMO: Donor

LUMO

HOMO

Electron transfer from high-lyingHOMO to matrix LUMO:

ON+

N

Cl-

Pyronin B chloridedonor precursor

charge transfer

reducedpyronin B

donor

HOMO

LUMO

1E-3 0.01

10-5

10-4

undoped NTCDA

σ=3.3*10-8 S/cm

conduct

ivit

y[S

/cm

]

molar doping ratio

A. Werner et al., App.Phys.Lett. 82, 4495 (2003)

dye radicalcreated in situ

from cation

very strongorganic donor

N-type doping: novel approach using cationic dyes

UPS/XPS study of organic/inorganic interfaces

Interested in:• Space charge layers at junctions and their influence on injection

• Barriers and their possible dependence on doping

• Direct observation of Fermi level shift

Institut fürAngewandte Photophysik

matrix dopant

NN

N

N

N

N

NN Zn

FF

F F

N

N

N

NZnPc F4-TCNQ

with N. Armstrong, D. Alloway, P. Lee, Tucson

J. Blochwitz et al., Organic Electronics 2, 97 (2001)

undoped doped

Institut fürAngewandte PhotophysikJ. Blochwitz et al., Organic Electronics 2, 97 (2001)

Creation of ohmic contacts by doping

: blocking : ohmic

• Charge carrier control: Doping of organic semiconductors

- Electrical Characterisation- UPS/XPS Spectroscopy: space charge layers

• Application in devices:

Organic light-emitting diodesOrganic solar cells

Outline

• Basics of organic semiconductors

Institut fürAngewandte Photophysik

Institut für Angewandte Photophysik (www.iapp.de)

• Basic research on novel OLED device concepts

• Organic solar cells

OLED cooperation in Dresden

Novaled GmbH (www.novaled.com)

• Holds University patents

• New Materials for stable doping systems

Fraunhofer-IPMS Dresden (www.ipms.fraunhofer.de)

• Stability of doped organic LED

• OLED-Inline-deposition set-up (30x40cm Substrate)

Organic Light Emitting Diodes

OLEDs: Basic Principles

Device structure

Glass substrate

V

A

Transparent anode

Emissive layer

Cathode

Light emission

Device energy diagram

E

x

Cathode

LUMO

HOMO

Anode+

+ +

---

Thinness

Viewing Angle

Brightness and colour Source: Kodak

OLED Benefits

CostThin

Light Weight

HighResolution

Lifetime

ContrastBrightness

Color Quality

High Response

Wide View Angle

Power Consumption

Suitable for TV and Movies

Suitable for Mobile

a-Si TFT LCD

p-Si TFT LCD

p-Si TFT OLED

Further improvementsexpectedSource: Toshiba 2002

OLED Performance

OLED products on the market

Car Audio Pioneer

Car Audio TDK

Phone SNMD

RGB PM

Passive Matrix

Active Matrix

Sanyo-Kodak

• 2003 market: approx 250 mio US$

• 98% small molecule devices

Sony 13” full color display

• Exponential current-voltage relation

• Flat-band under operation

• Low work-function contacts not needed !

p n

Comparison: Inorganic vs. Organic LED

Inorganic LED (e.g., GaAs/AlGaAs) Organic LED

CB

VB

EFe

EFh

Metal

• space charge limited currents

• low work function metals neededITO-preparation necessary

HTLETL

p-type doping

blocking layers for high efficiency

n-type doping

Ideal OLED: pin-heterostructure

Step 1

Step 2

Steps towards this design:

Step 3HOMO

LUMO

p i n

EF

+ + + + +

-----

+

Institut fürAngewandte Photophysik

N

N

N

N

Starburst = TDATA 4,4',4''-tris(N,N-diphenylamino) triphenylamin

2,8 2,9 3,0 3,1 3,2 3,3 3,4 3,5

1E-6

1E-5

6,86%

5,46%3,87%2.82%2.18%1.35%1.02%

cond

uctiv

ity σ

(S

/cm

)

103/T (K-1)

dopingratio

(mol%)

Eact=0,3eV

• undoped TDATA: 10-10 S/cm

• doped TDATA: up to 10-5 S/cm=> low ohmic losses: 0.1V/100nm@100mA/cm2

Dopant: F4-TCNQ

Step 1: Doping of amorphous wide gap materials

X.Q. Zhou et al., Adv. Funct. Mat.11, 310 (2001)

Y. Shirota et al., APL 65, 807 (1994)

Structure of pin-OLED

Institut fürAngewandte Photophysik

N-doping: Batho-Phenanthroline doped with Li (developed by Kido group)

Lowest voltages reported in literature for small-molecule devicesJ. Huang et al., Appl. Phys. Lett. 80, 139 (2002).

Performance of p-i-n structure

2 3 4 5 6 7 8 9 10

100

101

102

103

104

105

10,000 cd/m2 @5.2V

1,000 cd/m2 @2.9V

100 cd/m2 @2.55V

Lum

inan

ce [c

d/m

2 ]

Voltage [V]

Current efficiency: 5.27 cd/A @1,400cd/m2

(pure Alq3 emitter)

Institut fürAngewandte Photophysik

Institut für Angewandte PhotophysikTechnische Universität Dresden

What determines OLED efficiency ?

opticalrecombIexternal eUhb ηηνη ×××=

• bI: Electron and hole current balance: 1 can be reached

• eU: Operating voltage should be ≈ photon energy

• ηrecomb: 75% Triplet, 25% Singlet excitons: 0.25 for fluorescent emitter

Use phosphorescent (triplet) emitter

• ηoptical: about 20% in flat structure: 80% lost to wave guide modes:

optimize optical outcoupling !

Highly efficient PIN Triplet OLED

2 3 4 5

1

10

100

1000

10000

100000

10000 cd/m2 @ 4.2V26 cd/A; 19 lm/W

1000 cd/m2 @ 3.4V35 cd/A; 33 lm/W

100 cd/m2 @ 3.1V44 cd/A; 44 lm/W

10000 cd/m2 @ 3.79V52.4 cd/A; 43.5 lm/W

1000 cd/m2 @ 3.21V62.4 cd/A; 61.1 lm/W

100 cd/m2 @ 2.95V65.3 cd/A; 69.4 lm/W

Lum

inan

ce (c

d/m

2 )

Voltage (V)

Standard D-doping

Highly efficient PIN Triplet OLED

1E-3 0,01 0,1 1 10 100 10001

10

1001E-3 0,01 0,1 1 10 100 1000

1

10

100

14.5% @ 100 mA/cm2

19.5%

TCTA5TAZ10 TCTA10TAZ10 TCTA10TAZ5

Qua

ntum

effi

cien

cy (%

)

Current density (mA/cm2)

30 lm/W @ 100 mA/cm2

70 lm/W

Pow

er e

ffici

ency

(lm

/W)

Institut für Angewandte PhotophysikTechnische Universität Dresden

LED performance vs. time: linear plot

1970 1975 1980 1985 1990 1995 2000 20050

20

40

60

80

100

120

0

20

40

60

80

100

120

Red Filtered

Tungsten Bulb (unfiltered)

Fluorescent Lamp

InGaNgreen

OLED

PLED

AlInGaP Red/Yellow

Pow

er E

ffici

ency

Year

Institut für Angewandte PhotophysikTechnische Universität Dresden

What determines OLED efficiency ?

opticalrecombIexternal eUhb ηηνη ×××=

• bI: Electron and hole current balance: 1 can be reached

• eU: Operating voltage should be ≈ photon energy

• ηrecomb: 75% Triplet, 25% Singlet excitons: 0.25 for fluorescent emitter

Use phosphorescent (triplet) emitter

• ηoptical: about 20% in flat structure: 80% lost to wave guide modes:

optimize optical outcoupling !

• Three modes:

• External (≈ 20%)

• Substrate (≈ 35%)

• ITO/org (≈ 45%)

Optical outcoupling in OLED

• Charge carrier control: Doping of organic semiconductors

- Electrical Characterisation- UPS/XPS Spectroscopy: space charge layers

• Application in devices:

Organic light-emitting diodesOrganic solar cells

Outline

• Basics of organic semiconductors

LUMO

HOMO

• Absorption dominated byFrenkel excitons

• Large exciton binding energy:approx. 0.5eV

• Free carrier generation requires high fields (=> use in photocopiers)

Problem I: Exciton Separation

Exciton diffusion

LUMO

HOMO

hν

Absorption and exciton formationExciton diffusion

Exciton separation at the interfaceCarrier drift in the electricfield of a pn-heterojunction

DonorAcceptor

Solution: Donor-Acceptor-Heterojunction

Donor: p-typeAcceptor:n-type

C.W. Tang (Kodak)(1986)

• Efficiency 1%• Problems:

- Too low exciton diffusion length- low voltage, no controlled doping

Problem II: Exciton Diffusion

EC

EV

EF

EC

EV

polymer

C60Al

ITO--

+

ultrafast electron transfer

Cells with interpenetrating network: C60 as electron acceptor

p ni1 i2

EF

The piin-heterojunction concept with electron acceptor

• Photoactive interfacebetween two i-layers

• Wide gap transportlayers

• Highly doped layers to maximize Vbi

• Electron acceptor

C60 energy levelsM. Pfeiffer

photoactiveregion

Tang 86

ZnPc*C60

Evolution of the organic solar cell concept

p

n

p-i-n cell

d

Optical mode can be optimized as well

-1,0 -0,5 0,0 0,5 1,0

-20

0

20

40

60

133 mW/cm2

simulated AM 1.5

p-i-n Zelle mit 60nm ZnPc*C60 (1:2)UOC= 0.51 VJSC= 16.7 mA/cm2

FF= 36.4 %η=2.3%

curr

ent d

ensi

ty [m

A/c

m2 ]

voltage [V]

• ca. 80% internal quantum efficiency• Fill factor has to be improved

Pin-structure solar cell

recombinationrecombination centercenter::metal metal clustercluster

(Au)(Au)

Al

n-C60

ITO

p-HTL

ZnPc:C60

p-HTL

n-C60

ZnPc:C60

Tandem Solar Cell Structure

• Full sun spectrum can be used

• Higher voltage, lower currentthan single devices

ITO metal

n

pAu

EQF,e

EQF,h

undoped

Tandem solar cell: Doping is crucial for low losses

-0.4 -0.2 0.0 0.2 0.4 0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

30°C@ 9 mW/cm2

n-C60/i-MeO i-C60/p-MeO n-C60/p-MeO

curr

ent d

ensi

ty [m

A/cm

2 ]

voltage [V]

Au

(n)-C60

ITO

p-MeOTPD

ZnPc:C60

(p)-MeOTPD

• with Doping and Gold-Cluster: quasi-ohmic behavior

Tandem cell: influence of doping

-0.9 0.0 0.9

-20

0

20

40

133+/-3 mW/cm2

simulated AM 1.5

Tandemzelle: optimale Struktur mit 140 nm Spacer

Tandem:UOC= 0.99 VJSC= 10.4 mA/cm2

FF= 46.7 %η=3.5%

Einzelzelle:UOC= 0.51 VJSC= 16.7 mA/cm2

FF= 36.4 %η=2.3%

curr

ent d

ensi

ty [m

A/c

m2 ]

voltage [V]

η=3.6%

Tandem solar cell with 3.6% efficiency

Tandem Zellen: Der nächste SchrittTandem Zellen: Der nächste Schritt

glassglass/ITO/ITO pp--ii--nngapgap 2 2 eVeV

pp--ii--nngapgap 1.5 1.5 eVeV

pp--ii--nngapgap 1.5 1.5 eVeV

0.6 V0.6 V 0.5 V0.5 V 0.5 V0.5 V

AlAl

∑∑ = 1.5 ... 1.6 V ???= 1.5 ... 1.6 V ???VVOCOC

ηη = 5= 5--6%6%

Simulation by H. Hoppe (Uni Linz)

Next step: 3-layer cell

Conclusions:

• Organic semiconductors can be efficiently doped

• Microscopic physics not well understood

• Doping very useful for devices

Future work:

• Molecular n-doping

• Microscopic understanding of doping necessary

• Make solar cells as good as OLED already are

Institut fürAngewandte Photophysik

Conclusions

Acknowledgment

• M. Pfeiffer, J. Blochwitz-Nimoth, G. He, X. Zhou, J. Huang, A. Werner, J. Drechsel, B. Männig, K. Harada, T. Fritz

• J. Amelung, W. Jeroch, M. Schreil, U. Todt (FhG-IMS)

• D. Alloway, P.A. Lee, N. Armstrong, Tucson (XPS/UPS)

• N. Karl (Stuttgart)

• D. Wöhrle (Bremen), J. Salbeck (Kassel), H. Hartmann (Merseburg)

• BMBF, DFG, FCI, SMWK