Surface Growth Modes

(1)

2008/4/6 EE 235/NSE203 Nanoscale Fabrication; Lec. 11; Prof. Chang-Hasnain 1

Surface Growth Modes

High surface energy

thin wetting layer

High film energy

3D clusters

{101} pyramids relax

~ 50% strain energy

van der Merwe

Volmer-Weber

Stranski-Krastanow

(2)

Quantum-Dot Semiconductor Optical

Amplifiers

MOCVD ALE QDs

100nm

TEM photograph

Input signals Output signals Current

p

n

AFM image

1.3-um CW lasing

QD-SOA Device structure

Optical gain spectrum

K. Mukai et al., CLEO ‘99

Ground

state

Excited

state

Inhomogeneous

broadening

Wavelength

Optical gain

Alternative Growth Method:

MOCVD

MOCVD stands for Metal-Organic Chemical Vapor Deposition.

Also named as OMVPE, MOVPE, …

(3)

MOCVD Basics

• MOCVD is an epitaxial technology used for

growing compound semiconductor-based

epitaxial wafers and devices.

– GaAs, InP, GaN, …

– InGaAs, InGaAsP, …

– Other N-, As-, P-, Sb- based materials.

• Applications

– Compound semiconductor based devices

• Optoelectronic devices such as semiconductor lasers or

LEDs

• High speed electronic devices, e.g. HBT

• Solar cells

• OEIC (Optoelectronic Integrated Circuit)

– Artificial structures for basic research

• Low-dimensional structures (QW, QWW, QD)

• Nonplanar structures

• MEMS

Precursors we are using are …

• As

TBA (tertiary-butyl-arsine)

• P

TBP (tertiary-butyl-phosphine)

• Ga

TEGa (Tri-ethyl-gallium)

• Al

TMAl (Tri-methyl-aluminum)

• In

TMIn (Tri-methyl-indium)

• Zn

DEZn (Di-ethyl-zinc)

• Si

Si

2

H

6

(gas source)

• H

2

Carrier gas

(4)

Source Bubblers

Carrier gas

P

o

precursor

P

_r

(vapor pressure)

P

_i

F

c

F

r

c

i

o

r

P

F

P

=

−

Flow rate can be controlled

by either pressure or bubbler

temperature.

System Configuration

Injection

block (V)

Injection

block (III)

(5)

Principles of MOCVD growth (

cont.

)

Rotating Wafer

Susceptor

(~ 1000-2000 RPM

T ~ 650

o

C)

Reactor Top

(T ~ 50

o

C)

(6)

Chamber Gas Flow Design

Stability vs. Spin Rate in Vertical-Type

MOCVD

(7)

CVD Mechanisms

From EE143

Lecture Note

Chemical Reaction

• Precursors crack due to high temperature on top

of the wafer surface.

• Ga+As (gas phase)

Æ

GaAs (stable solid

compound)

As C C C C TBA

substrate

As Ga C C TEGa C C C C G a C C C

(8)

MOCVD vs. MBE

• Advantages

– Faster growth rate (3-10 um/hr vs. 1 um/hr)

– Scalable to many wafers (100 X 2”)

– Wide temperature control range. Better film uniformity.

– Shorter system downtime (1-2 days vs. 1-2 weeks)

– Possible to grow many compositions (by varying MFC flow rates)

– Excellent morphology and thickness control

• Disadvantages

– Toxic sources

Æ

safety issues

– Consumes large quantity of hydrogen

Æ

also a safety concern

– Some memory effects

– Huge set of parameters. Hard to optimize.

Arsine (AsH

3

) – As Source in Industry

• Less expensive (comparing to TBA)

• Very pure

• Gas phase at room temperature

• Highly Toxic

(9)

Arsine (AsH

3

) vs. Silane (SiH

4

)

Precursors we are using are …

• As

TBA (tertiary-butyl-arsine)

• P

TBP (tertiary-butyl-phosphine)

• Ga

TEGa (Tri-ethyl-gallium)

• Al

TMAl (Tri-methyl-aluminum)

• In

TMIn (Tri-methyl-indium)

• Zn

DEZn (Di-ethyl-zinc)

• Si

Si

2

H

6

(gas source)

• H

2

Carrier gas

(10)

MOCVD

Colloidal QDs

• Grown in solution phase

(11)

Fabrication – colloidal synthesis

• Advantage

– Cheapest, large volume,

benchtop conditions

– Least toxic

• Self assembly of QD using

electrochemical method

– Template created using ionic

reaction at electrolyte-metal

interface

– Results in spontaneous

assembly of QDs

Basic Synthesis Mechanism

•

Temporary and discrete nucleation event,

followed with slower and controlled

growth on existing nuclei

– Nucleation is created by rapidly increasing

reagents into a vigorously stirred flask

containing a hot coordinating solvent

– Nucleation forms to relieve

supersaturation

– Control precursor addition to limit addition

nuclei formation

Æ

focusing of the size

distribution

•

Second Growth Phase: Ostwald ripening

– Smaller nanocrystals (NCs) dissolve due

Thermocouple

Organometallic Precursors Ar

(12)

Ostwald ripening

• A spontaneous process occurs because larger particles

are more energetically favored than smaller particles.

• The formation of many small particles is kinetically

favored,

(i.e. they nucleate more easily),

large particles

are thermodynamically favored.

This is because small

particles have a larger surface area to volume ratio than

large particles and are consequently easier to produce.

(13)

Colloidal QD Synthesis (2)

•

Solvents/Surfactants

– III-V: could be mixtures of alkylphosphines R

3

P, alkylphosphine oxides

R

₃

PO, alkylamines, etc. R=butyl or octyl

– II: metal alkyl (dimethylcadmuim, diethyl-Cd, diethyl-Zn, etc.)

– VI: organophosphine chalcogenides R

₃

PE (E=S, Se or Te), etc.

– Alloys can be made with mixture of solvents

•

Surfactants

– NC surface is coated with a layer of organic molecules called surfactants,

which is used to control the nanocrystal growth.

– Should not bind too strongly, which does not allow growth.

– If too weak

Æ

leads to large NCs or aggregates

•

Uniformity

– InP, InAs

Æ

<10%; growth is slow as Ostwald ripening over 1-6 days is

required

– ~5% for metal NCs

•

Capping

– Provide repulsive force of sufficient strength and range to counteract van

der Waals attraction between NCs

Shape Control: Spherical CdSe

•

Inject TOPO at T>300C

•

Hexagonal (wurtzite) structure

Æ

inherent anisotropy

•

Aspect ration can be tuned by

the ration of TOPO and HPA

– Spheres: 8% HPA

– Nanorods: 20% HPA

50 nm

binary surfactant mixture

Hexyl phosphonic acid

(HPA)

Tri-octyl

phosphine oxide

(TOPO)

(14)

CdSe Rods

Peng, X. G.;

Manna, L.; Yang,

W. D.; Wickham,

J.; Scher, E.

Kadavanich, A. P.

Alivisatos, Nature

2000, 404, 59-61.

Independent Control of Length and

Diameter

(15)

Bandgap vs. Length and Diameter

10 20 30 40 50 3 4 5 6 7 8 1 .8 1 .9 2 .0 2 .1 2 .2 2 .3 D ia_m e te r (n_m )

Li, L. S., J. T. Hu, W. D. Yang and A. P. Alivisatos (2001). "Band gap variation of

size- and shape-controlled colloidal CdSe quantum rods." Nano Letters

1(7):

349-351.

Diameter (nm)

Length (nm)

eV

Wurtzite and Zinc Blende Polytypism

A Mechanism for Branching Nanocrystals

WZ

Stabilized by

surfactant

Growth in WZ

ZB

More stable

by 7meV/unit cell

Nucleation in ZB

(16)

Tetrapods - First Observations

Manna, L.; Scher, E. C.; Alivisatos, A. P., J. Am. Chem. Soc. 2000, 122, 12700-12706.

Control of Branching

•

Manipulation of kinetics and thermodynamics

– Temperature (T) of growth

– Concentration (C) of monomers

•

First set the condition to promote cubic nuclei

•

Larger tetrahedral NC can be obtained by keeping the system at low

T

•

Hexagonal growth can be obtained with high T or high concentration

of monomers

Æ

kinetic growth regime

(17)

CdTe Tetrapods Control of Arm Length and Width

Manna, L.; Milliron, D. J.; Meisel, A.; Scher, E. C.; Alivisatos, A. P.,

"Controlled growth of tetrapod-branched inorganic nanocrystals" Nature Materials 2003, 2, 382.

Cd/Te

1:1

2:1

3:1

5:1

1:2

1:3

1:5

Cd/ODPA

(18)

1

st

Generation Branching off of Rods

-CdS/CdTe

CdS Nanorods

CdS/CdTe Nanorockets

50 nm

50 nm 100 nm

2

nd

Generation Branching off of

Tetrapods

(19)

Electronic Energy Levels of

Tetrapods

Lin-Wang Wang, NERSC/LBNL

Empirical pseudopotential calculations

cb1

cb2

cb5

cb6

vb5,6

vb

vb1,2

Electrical Contact to Individual

Tetrapods

Oxide Degenerately-doped Si gate Source Drain

100 nm

(20)

QD Solar Cell

• Current PV cells are limited by cost of material

and device efficiency

• First generation: silicon

– Efficient but high material processing cost

• Second generation: thin films

– Reduced cost and cheaper, but poor efficiency

Phys. Rev. Lett. 92, 186601 (2004)

Nature Physics 1, 189 (2005)

(21)

Quantum Dot Display

• Motivation

– QD emit lights in very specific Gaussian distribussion

• Results in accurately reflect the colors that human eyes can perceive

– Require very little power (not color filtered) and no backlight needed

– Potential for high contrast

• Prototype

– 32 x 64 pixel, red monochromatic QD display

Image source: QD Visions

QD in Biological Applications

•

Goal – understanding the function and interaction of hundreds of

thousands of proteins in a single cell

•

Problem with fluorescent labels

– Limited fluorescent dyes with distinct emission spectra

– Broad emission line width

– Require a different excitation wavelength

•

QD as fluorescent labels

– Narrow emission spectra – multiple

– Excellent brightness

– excited with single wavelength source

– Long lasting fluorescence

•

Toxicity concern

(22)

Synthesis

•

Preparation of the core

– Choice of materials:

CdS

(UV-blue),

CdSe

(visible), or

CdTe

(near IR)

– Based upon desirable emission wavelength

– Physical dimensions fine-tunes the emission wavelengths

– Unstable emission

•

Crystalline imperfections and surface defects – non-radiative

recombinations

•

Surfaces are reactive and easily polluted

•

Passivating shell

– Transparent but structurally related material

– ZnS – almost unreactive

•

Surface chemistry – water soluble

– Layer of silica

•

Coating layer

– Organic ligands – covalently attached to surface

– Hydropholic/hydrophobic polymers with carboxylic acid derivations

QD Labeling of Breast Cancer Cell

• QD can be conjugated to IgG

and the resulting fluorescent

probe is capable of binding to

specific antibodies

(23)

Comparison with Organic Dye

Molecules

Brightness

Photostability

QD for

in-vivo

Imaging

•

Requirement

– Adequate circulation lifetime

– Minimal nonspecific deposition

– Retain fluorescence for sufficient time

•

The core-shell size of the QDs had no

significant effects on tissue deposition or

circulation lifetimes

•

Immediately after intravenous injection,

fluorescence from all QD emission

wavelengths and surface coatings is

easily seen in the superficial vasculature.

•

Subsequently, QDs are seen to be