1

@ MITApril 29, 2003 – Organic Optoelectronics - Lecture 20b

Organic LEDs – part 8

• Exciton Dynamics in Disordered Organic Thin Films

• Quantum Dot LEDs

Handout on QD-LEDs: Coe et al., Nature 420, 800 (2002).

2

Exciton Dynamics in Time Dependant PL

wavelength time

2 ns

25 n

m

time

wav

elen

gth

3

Dynamic Spectral Shifts of DCM2 in Alq3

1.7 1.8 1.9 2.0 2.10.0

0.2

0.4

0.6

0.8

1.0

720 680 640 600

2% DCM2:Alq3

6.2 ns3.4 ns2.2 ns

1.6 ns1.0 ns

0.6 ns0.3 ns

0.1 ns0 ns

time

Nor

mal

ized

Inte

nsity

Energy [ eV ]

Wavelength [ nm ]

DCM2Alq3

~ ~ DCM2 PL red shifts > 20 nm over 6 ns ~DCM2 PL red shifts > 20 nm over 6 ns ~

• Measurement performed on doped DCM2:Alq3 films

• Excitation at λ=490 nm (only DCM2 absorbs)

1.7 1.8 1.9 2.0 2.10

2

4

6

8

10

12

Energy [ eV ]

Inte

nsity

[ ar

b. u

nits

]

4

1.6 1.8 2.0 2.20.0

0.2

0.4

0.6

0.8

1.0

Integrated Spectrum(0-10 ns)

Spectrumat 1 ns

Nor

mal

ized

Inte

nsity

Energy [eV]

750 700 650 600 550Wavelength [nm]

Spectrumat 4 ns

Time Evolution of 4% DCM2 in Alq3 PL Spectrum

5

T1S1

S0FL

UO

RES

CEN

CE

PHO

SPH

OR

ESC

ENC

E

ENERGY TRANSFER

FÖRSTER, DEXTERor RADIATIVE

INTE

RN

AL

CO

NV

ER

SIO

N

AB

SOR

PTIO

N

10 ps

1-10 ns

>100 ns

Ener

gy

density of availableS1 or T1 states

Electronic Processes in Molecules

6

0 1 2 3 4 51.80

1.85

1.90

1.95

2.00

2.05

2.10

2.15

2.20

CH3CN

Acetone

DMSO

CHCl3

CHCl2

BenzeneEn

ergy

[eV

]

Time [ns]

680

660

640

620

600

580

Wav

elen

gth

[nm

]

Time Evolution of DCM2 Solution PL Spectra

7

Wavelength [nm]

Tim

e [n

s]

0

1

2

3

4

5

600 650 700 750

10% DCM2 in Alq310% DCM2 in Alq3

Spectral Shift due to

~ Exciton Diffusion ~~ Intermolecular Solid State Interactions ~

35 nm35 nmwavelength shift

8

Each dye molecule experiences a different local medium

⇒ variations in excitonic energies

Inte

nsit

y

Energy

12

3

Molecular PL Spectra

Freq

uenc

y

Energy

Excitonic Density of States

Non-zero width to excitonicdensity of states

Excitonic Energy Variations

12 3

9

log(Time)

S 1or

T1

Ener

gy

1 2 3 4

Exciton Distribution in the Excited State (S1 or T1)

fwhm

~ Time Evolved Exciton Thermalization ~

EXCITON DIFFUSION LEADS TO REDUCTION IN FWHM

hν

10

Ene

rgy

[eV

]

Time [ns]

680

660

640

620

600

Wav

elen

gth

[nm

]

0 1 2 3 4 5 6 7

High Emidpoint

Low Emidpoint

1.80

1.85

1.90

1.95

2.00

2.05

2.10

T = 15~125 K

Energy

Low Emidpoint

High Emidpoint 4% DCM2 in Alq3

FWHM

11

0 1 2 3 4 5

1.95

2.00

2.05

2.10

2.15

2.20

2.25

2.30En

ergy

[eV

]

Time [ns]

640

630

620

610

600

590

580

570

560

550

540

Wav

elen

gth

[nm

]

x%x% DCM in AlqDCM in Alq33

0.2 %

2 %

0.6 %

0.4 %

Alq3

4.5 %

12

1.96

2.00

2.04

2.08

2.12

2.16En

ergy

[eV

]

% of DCM in Alq3

630

620

610

600

590

580

570

Wav

elen

gth

[nm

]

0.1 1 2 40.50.2

Initial Peak PL

Final Peak PL

X% DCM in AlqX% DCM in Alq33

13

1 10 100

2.25

2.30

2.35

2.40

2.45

Ener

gy [

eV]

Time [ns]

Alq3

Ir(ppy)3

Time Evolution of Peak PL in Neat Thin Films

14

Parameters for Simulating Exciton Diffusion

observed radiative lifetime (τ) Förster radius (RF)

► Assign value for allowed transfers:

0.01

0.1

1

-2 0 2 4 6 8 10 12 14 16

decreasing doping(5% down to 0.5%)

τ = 3.3 ns

Nor

mal

ized

Inte

grat

edS

pect

ral I

nten

sity

Time [ ns ]

1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.40.00.10.20.30.40.50.60.70.80.91.0

750 700 650 600 550

FWHMsSinit(E) : 0.27 eVSmol(E) : 0.22 eVgex(E) : 0.15 eV

Nor

mal

ized

Inte

nsity

[ ar

b. u

nits

]

Energy [ eV ]

Wavelength [ nm ]excitonic density of states (gex(E))

► Assume Gaussian shape of width, wDOS

► Center at peak of initial bulk PL spectrum

► Molecular PL spectrum implied…

∆E0

ΓF61⎟⎠⎞

⎜⎝⎛RR

τF

15

Fitting Simulation to Experiment – Doped Films

1.94

1.96

1.98

2.00

2.02

2.04

2.06

630

620

610

600

0 1 2 3 4 5 6 7 8 9

0.5% DCM2:Alq3

RF=0.88 wDOS=0.185 eVRF=0.98 wDOS=0.161 eVRF=1.08 wDOS=0.146 eVRF=1.18 wDOS=0.136 eVRF=1.28 wDOS=0.130 eV

Ene

rgy

[ eV

]

Wav

elen

gth

[ nm

]

Time [ ns ]1.841.861.881.901.921.941.961.982.002.022.042.06

670

660

650

640

630

620

610

0 1 2 3 4 5 6 7 8 9

5%

2%1%

0.5%

Ene

rgy

[ eV

]

Wav

elen

gth

[ nm

]Peak PL of x%DCM2:Alq3

Time [ ns ]

• Good fits possible for all data sets

• RF decreases with increasing doping, falling from 52 Å to 22 Å

• wDOS also decreases with increasing doping, ranging from 0.146 eV to 0.120 eV

16

0 10 20 30 40 502.26

2.28

2.30

2.32

2.34

2.36

Ener

gy [

eV ]

Time [ ns ]

0 50 100 150 200 250 300

2.24

2.26

2.28

2.30

2.32

2.34

2.36

2.38

2.40

En

ergy

[ eV

]

Time [ ns ]

-2 0 2 4 6 8 10 12 142.72

2.74

2.76

2.78

2.80

Ener

gy [

eV ]

Time [ ns ]

Alq3wDOS~0.15 eV NPD

wDOS~0.10 eV

Ir(ppy)3wDOS~0.21 eV

Fitting Simulation – Neat Films

• Spectral shift observed in each material system

• Molecular dipole and wDOS are correllated: lower dipoles correspond to less dispersion

• Even with no dipole, some dispersion exists

• Experimental technique general, and yields first measurements of excitonic energy dispersionin amorphous organic solids

17

upon excitation both magnitude and directionof lumophore dipole moment can change

FOR EXAMPLE for DCM: µ1 – µ0 > 20 Debye !~ from 5.6 D to 26.3 D ~

following the excitation the environment surrounding the excited molecule will reorganize to minimize the overall

energy of the system (maximize µ • Eloc)

Eloc

µ0

t < 0

Eloc

µ1

t = 0

Eloc

µ1

t ~ 1 ns

Temporal Solid State Solvation

18

∆ E

log(Time)

Ener

gy

1 2 3 4

Exciton Distribution in the Excited State (S1 or T1)

~ Time Evolved Molecular Reconfiguration ~

DIPOLE-DIPOLE INTERACTIONLEADS TO ENERGY SHIFT IN DENSITY OF EXCITED STATES

19

Fusion of Two Material Sets Hybrid devices could enable

LEDs, Solar Cells, Photodetectors, Modulators, and

Lasers

which utilize the best properties of

each individual material.

Fabrication of rational structures has been the main obstacle to dateto date.

Organic Semiconductors

Abs

orpt

ion

[a.u

.]

400 500 600 700Wavelength [nm]

Emissive

Tunable

Nanoscale

Efficient

Flexible

Colloidal QDs

Photo by F. Frankel

20

Inorganic Nanocrystals – Quantum Dots

Quantum Dot SIZE

D

Lowest allowed energy level

E

Abs

orpt

ion

[a.u

.]400 500 600 700

17Å20Å

29Å

33Å

45Å55Å72Å

80Å

D=120Å

Wavelength [nm]

Synthetic route of Murray et al, J. Am. Chem. Soc. 115, 8706 (1993).

CdSeZnS

Photo by F. Frankel

21

Fusion of Two Material Sets

Organic Molecules

Abs

orpt

ion

Spe

ctra

Quantum Dots

TAZ

N

NN

TPD

N

CH3

N

H3C

Alq3

N

OAl

NO

NO

N N

NPD

TUNABLE

PbSe

Wavelength [µm]0.9 1.2 1.5 1.8 2.1 2.4

NANOSCALE

CdSe(ZnS)

Wavelength [nm]400 500 600 700

EMISSIVE

22

ZnS overcoating shell (0 to 5 monolayers)

Oleic Acid or TOPO caps

Integration of Nanoscale MaterialsQuantum Dots and Organic Semiconductors

D = 17-120Å(~50Å pictured)

TAZ

3-(4-Biphenylyl)-4-phenyl-5-tert-butylphenyl-1,2,4-triazole

N

NN

TPD

N,N'-Bis(3-methylphenyl)-N,N'-bis-(phenyl)-benzidine

N

CH3

N

H3C

Alq3

Tris(8-hydroxyquinoline)Aluminum (III)

N

OAl

NO

NO

10Å

Trioctylphosphine oxide

Synthetic routes of Murray et al, J. Am. Chem. Soc. 115, 8706 (1993) and Chen, et al, MRS Symp. Proc. 691,G10.2. PbSe or

CdSe Core

O

HO

Oleic Acid

N N

NPD

N,N'-Bis(naphthalen-1-yl)-N,N'-bis(phenyl)benzidine

N

N

poly-TPD

Phase Segregation

Chemistry SizeMolecular Organics

Differences in:

Aliphatic CapsAromatic

Quantum Dots“Small”“Big”

23

Phase Segregation and Self-Assembly

1. A solution of an organic material, QDs, and solvent…

2. is spin-coated onto a clean substrate.

3. During the solvent drying time, the QDs rise to the surface…

4. and self-assemble into grains of hexagonally close packed spheres.

Organic hosts that deposit as flat films allow for imaging via AFM, despite the AFM tip being as large as the QDs.

Phase segregation is driven by a combination of size and chemistrysize and chemistry..

50nm

~8nm PbSe QDs

Flat TPD background

TPD/solvent

QD Solution TPD Solution

11 22 33 44Substrate Substrate

QDs

TPDclose packed QDs

24

Monolayer Coverage – QD concentration

250nm

500nm

500nm500nm

100nm

As the concentration of QDs in the spin-casting solution is increased, the coverage of QDs on the monolayer is also increased.

QD coverage = 20% 50% 80%

90% 100%

25

400 500 600 1200 1400 1600

Nor

mal

ized

cou

nts

W avelength [nm ]

QD-LED Performance

CdSe(ZnS)/TOPO PbSe/oleic acidGlassITO

50nm Ag50nm Mg:Ag30nm Alq3

40nm TPD

10nm TAZ

EF=3.7

EA=4.6

IE=6.8

QD

EF=4.7

Vacuum Level = 0 eV

ITO

Mg

Elec

tron

Ener

gy [e

V] EA=2.1

IE=5.4

TPD

IE=5.8

EA=3.1

Alq3

EA=3.0

IE=6.5

TAZ

ZnS

CdSe

TOPO

1 10

10-8

10-6

10-4

10-2

100

Voltage [V]

Curre

nt De

nsity

[A cm

-2]

Control

QD-LED

26

4 nm HRTEMHRTEM

AFMAFM

Full Size Series of PbSe Nanocrystalsfrom 3 nm to 10 nm in Diameter

1000 1500 2000 2500

0

1

2

3

4

5

6

7

8

9

1136 nm

1268 nm

1600 nm

1404 nm

1504 nm

1708 nm

1812 nm

2036 nm

2208 nm

Wavelength (nm)

Abso

rban

ce (

arbi

trar

y un

its)

peak absorption

8 nmOrdered Monolayer ofOrdered Monolayer ofPbSe PbSe NanocrystalsNanocrystals

Structure of aSingle Nanocrystal

JonnyJonny SteckelSteckel and Seth Coeand Seth Coe

27

Design of Device Structures

QD

ZnS

CdSe

TOPO capsQD intersite spacing

QDs are poor charge transport materials...

e-

But efficient emitters…

1. Generate excitons on organic sites.2. Transfer excitons to QDs via Förster

or Dexter energy transfer.3. QD electroluminescence.

Isolate layer functions of maximize device performance.

Need a new fabrication methodnew fabrication method in order to be able to make such double heterostructures:

Phase SegregationPhase Segregation.

EF=3.7

EA=4.6

IE=6.8

QD

EF=4.7

Vacuum Level = 0 eV

ITO

MgEl

ectro

n En

ergy

[eV] EA=2.1

IE=5.4

TPD

IE=5.8

EA=3.1

Alq3

EA=3.0

IE=6.5

TAZ

ZnS

CdSe

TOPO

Use organics for charge transport.GlassITO

50nm Ag50nm Mg:Ag30nm Alq3

40nm TPD

10nm TAZ

QD monolayer ZnS shell

TOPO caps

QD

CdSe core

28

A general method?

Phase segregation occurs for different

1) organic hosts: TPD, NPD, and poly-TPD.

2) solvents: chloroform, chlorobenzene, and mixtures with toluene.

3) QD core materials: PbSe, CdSe, and CdSe(ZnS).

4) QD capping molecules: oleic acid and TOPO.

5) QD core size: 4-8nm.

6) substrates: Silicon, Glass, ITO.

7) Spin parameters: speed, acceleration and time.

• This process is robust, but further exploration is needed to broadly generalize these findings.

• For the explored materials, consistent description is possible.

• We have shown that the process is not dependent on any one material component.

3) core

2) Solvent

7) Spin Parameters

Phase segregation QDQD--LED structuresLED structures

6) Substrate/Anode

Cathode

ETL

1) Organic Host

shell

QD monolayer

4) caps

QD

5) QD size

29

GlassAnode

Cathode

exciton density

dept

h

Alq3 (35 nm)increasing current

2% DCM2 in Alq3 (5nm)RF

TPD (40 nm)

Coe et al., Org. Elect. (2003)

Wavelength [nm]

Nor

mal

ized

Inte

nsity

400 500 600 700 800

Alq3

DCM2

130013013

1.30.13

EL F

ract

ion

[%]

log(J[mA/cm2])

100

80

60

40

20

0

100

80

60

40

20

0

-1 0 1 2 3-1 0 1 2 3

J [mA/cm2]DCM EL

Alq3 EL

EL Recombination Region Dependence

on Current

30

Interstitial Space Monolayer Void

Grain Boundary

RF

Device AreaTOP DOWN VIEW of the QD MONOLAYER

Exciton recombination width far exceeds the QD monolayer thickness at high current densityhigh current density.To achieve true monochrome emission, new exciton confinement techniques are needed.

400 500 600 700

1.3

Alq3

QD

Wavelength [nm]

380

130

13

0.38

J [mA/cm2]

Spectral Dependence on Current Density

CROSS-SECTIONAL VIEW of QD-LEDde

pth

GlassAnode

Cathode

HTL

exciton densityRF

QD EL

Alq3 EL

hole

electronincreasing

currentETL

31

Demonstrated:Demonstrated:•Spectrally Tunable – single material set can access most of visible range.

•Saturated Color – linewidths of < 35nm Full Width at Half of Maximum.

•Can easily tailor “external” chemistry without affecting emitting core.

•Can generate large area infrared sources.

Benefits of Quantum Dots in Organic LEDs

Coe et al, Nature 420, 800 (2002).

400 450 500 550 600 650

FWHM 32nm0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

y

x

standard CRTcolor triangle

CIE Diagram

Potential:Potential:•High luminous efficiency LEDs possible even in red and blue.

•Inorganic – potentially more stable, longer lifetimes.

The ideal dye molecule!The ideal dye molecule!

Potential QD-LED colors

QD