Get documents about "Breaking the Bandwidth Bottleneck in Telecommunications
Document Sample
10th Foresight Conference on Nanotechnology
October 11-13, 2002
Breaking the Bandwidth Bottleneck in
Telecommunications & Information Processing:
New Electro-Optic Materials
Larry Dalton
Departments of Chemistry, Materials Science & Engineering,
and Electrical Engineering
University of Washington & University of Southern California
Acknowledgements
Financial support provided by the National Science Foundation and the Air
Force Office of Scientific Research
Critical to Next Generation Computing
•Semiconductor Research Corporation
Workshop on Optical Interconnects
http://www.src.org/member/sa/nis/E002117_Op
to_wksp.asp
•British House of Lords Select Committee on
Science & Technology Study of Innovations in
Computer Processors
•Forthcoming article in IEEE Computing
•High frequency, ultra high stability clocks
•On-chip signal distribution
•Chip-to-chip interconnection
•Module-to-module interconnection
Critical to Telecommunications
Industry
From: "PARK,CHRIS (A-England,ex1)" <chris_park@agilent.com>
To: "'Larry Dalton'" <dalton@chem.washington.edu>
Cc: "MEADOWCROFT,SIMON (A-England,ex1)" <simon_meadowcroft@agilent.com>
Subject: Collaboration with Agilent Technologies
Date: Tue, 2 Jan 2001 15:54:07 +0100
Dear Prof Dalton,
Agilent Technologies would like to meet with you to find out more about your work on high speed polymer
modulators. Our interest is based on a need to manufacture low cost 100 Gigabit Ethernet optical components in
approximately 3 years time using technologies which are compatible with high volume and low manufacturing
cost. The work you have published on high speed modulators is currently one of the best alternatives for a low
component count, low modulation voltage 100GbE transmitter. Agilent Technologies would therefore be interested
in discussing your work and the options for collaboration. These options can include research sponsorship and/or
business development including support for new start-up activity. Simon and I will be attending OFC in Anaheim
and would like to meet with you that week, or if you are not attending OFC we could visit Washington early the
following week (w/c 26th March). Please let us know whether you are available at this time.
Best regards
Dr Chris Park
Research ManagerAgilent Technologies
Whitehouse RdIpswichUK
Tel 44 1473 465628
e-mail chris_park@agilent.com
Electro-Optic Devices: The on-ramps & interchanges of the
information superhighway
(The Metro Loop and Fiber to the Home)
Critical to Defense Industry
Caltech U Washington
Electro-Optics: The Phenomena
An electro-optic material (device) permits electrical and optical
signals to ―talk‖ to each other through an ―easily perturbed‖
electron distribution in the material. A low frequency (DC to 200
GHz) electric field (e.g., a television [analog] or computer
[digital] signal) is used to perturb the electron distribution (e.g., p-
electrons of an organic chromophore) and that perturbation alters
the speed of light passing through the material as the electric field
component of light interacts with the perturbed charge
distribution.
Because the speed of light is altered by the application of a control
voltage, electro-optic materials can be described as materials with
a voltage-controlled index of refraction.
Index of refraction = speed of light in vacuum/speed of light in material
Electro-Optic Devices: The on-ramps & interchanges
of the information superhighway
The electro-optic effect can be used to transduce electrical
information (signals) onto the internet (in to optical signals). By
slowing light down in one arm of the Mach Zehnder device shown
below, the interference of light beams at the output can be
controlled. Electrical information appears as an amplitude
modulation on the optical transmission. This works equally well
for analog or digital data.
DC bias electrode
ground electrode
Modulated
Light In Light Out
RF electrode
Substrate
What are the critical requirements for EO
materials and devices?
Low halfwave voltage is a critical requirement in
externally modulated photonic systems:
Analog systems:
For RF transparency:
Link gain 1/Vp2
For high dynamic range:
NF Vp2
(low level signal detection limited by noise floor)
Digital systems:
High speed digital circuits have low output voltage
Digital amplifiers very costly
Bandwidth is the other critical requirement!
Why Organic Electro-Optic Materials (Devices)?
•Intrinsic material bandwidths of several hundred gigahertz.
The response time (phase relaxation time) of p-electrons
in organic materials to electric field perturbation is on the
order of femtoseconds. Operational bandwidths of 150
GHz have been demonstrated for modulators & switches
.
•Organic electro-optic coefficients are currently 2-4 times
higher than lithium niobate and getting larger.
Theoretically-inspired rational design of materials will keep
electro-optic activity improving for several years. Device
operational voltages of less than 1 volt are routine.
•Organic EO materials are highly processable into 3-D circuits
and can be easily integrated with semiconductor VLSI
electronics and silica fiber optics. Low loss coupling
structures can be straightforwardly fabricated.
Comparison of Material Performance
Property Polymer (at 1.3m) Polymer (at 1.55m) Lithium Niobate
EO Coefficient (r) 130 pm/V 60 pm/V 32 pm/V
Optical Loss 1 (0.7) dB/cm 1 (0.2) dB/cm 0.2 dB/cm
Bandwidth•Length >350 GHz•cm >350 GHz•cm 10 GHz•cm
Product
Dielectric Constant () 2.5-4.0 2.5-4.0 28
Refractive Index (n) 1.5-1.7 1.5-1.7 2.2
Figure of Merit (n3r/) ~100 ~100 6
Thermal Stability 85 ºC 85 ºC (90 ºC)
Comparison of Lithium Niobate and Polymer Electro-
Optic Modulators
State-of-the-art High Speed Infrared Modulators
Commercial Lithium Niobate Devices—The Competition
Vp: 6 V @1550 nm, 30 GHz Bandwidth, $6000/per unit
Commercially Available Polymer Devices
Vp: 1.2 V @ 1300 nm, 1.8 V @1550 nm
20 GHz and 30 GHz Bandwidth (3dBe)
Published Prototype Device Results
Vp: 0.77 V @ 1300 nm
100 GHz operation 10 Modulator Chips on 3 Inch Wafer
Recent Dendrimer Device Results
Vp: 0.5 V @ 1550 nm
Recent MR Device Results
Vp: < 1 V @ 1550 nm 2 Push-Pull MZ Modulators on One Chip
Why Nanostructured Electro-Optic Materials?
•Noncentrosymmetric ordering of chromophores (all pointing in the
same direction) in the material lattice is required for electro-optic
activity. Chromophore dipole-dipole interactions oppose this ordering.
Forces must be used to achieve the desired order and chromophores
must be positioned precisely in space to minimize the undesired effects
of dipole-dipole interactions.
•A uniform chromophore distribution (and high concentration) is
necessary not only to maximize electro-optic activity but also to avoid
optical loss from scattering off of material density (index of refraction)
variations.
•Several routes to nanostructured electro-optic materials are being
pursued including (1) the electric field poling of dendritic materials and
(2) sequential (layer-by-layer) synthesis from an appropriate substrate
(which also serves as a cladding material).
Theoretically inspired rational improvement of
organic electro-optic materials
•Theory (quantum and statistical mechanics have guided the systematic
improvement of the hyperpolarizability (b) of organic chromophores
and the electro-optic activity of macroscopic materials, e.g.,
Systematic Improvement in Molecular
Electro-Optic Activity: Variation of b
b(x10-48 esu) b(x10-48 esu)
R R
N NO 2 80
R N NC
NA S 9,800
R CN
CN
R
N N 580 TCV
R N NO 2 R
N
DR, 30 wt%, r33 = 13 pm/V 13,000
NC
R
R S CN
O O
N N CN
S TCVIP
2,000
Ph CN
R R
NC 15,000
ISX N S SO2
R
NC R
N S CN 3,300 SDS
R CF 2(CF 2)5CF 3 R NC NC
CN
FCN N S 18,000
O
R R
N
FTC, 20 wt%, r33 = 55 pm/V
R 4,000 R
O
N
Ph O
N
APTEI R
CN
R R'
O 30,000
NC
N S CN 6,100 NC CN
R NC
CLD NC
TCI
New Advances in Chromophore
Development
Quantum mechanical calculations permit the optimization of the p-
electron structure that defines molecular hyperpolarizability.
New Synthesis Techniques: Microwave synthesis techniques
permit dramatic enhancement in reaction yields and synthesis of
. new materials. HO
A, B, C = NO , CN,
2
N S
O2CF3, etc.
N
S
S D D
New Paradigm: N O
N
Gradient-
C A
Bridge, Mixed- OH
B
Ligand-Acceptor
D = CF3, etc.
Chromophores
Why Microwave Synthesis?
•Microwave synthesis has permitted dramatic enhancement in
reaction yields, reducing time devoted to purification. It has
also permitted many materials to be synthesized for the first
time and has permitted greater flexibility in reaction conditions.
. •Microwave synthesis techniques obviously permit more
uniform heating of reaction mixtures. The absence of thermal
gradients and ―hot spots‖ helps minimize decomposition and
side reactions. Microwave synthesis permits the use of a wider
range of solvents.
•We have found this approach to be particularly effective for
condensation, addition, and de-protection reactions.
Comparison of Microwave & Reflux
Synthesis of CF3-TCF acceptor
-Hydroxyketone Condensation
CN
O
Li OH CN
O O
i,
OEt CF3
CN F 3C O CN
OH 2
CF3 CN
CF3 CF3 -TCF
ii, dilute HCl
70%
Conditio n Base Reaction time Yield (%)
Reflux LiOEt 48 h 30
Microw ave NaOEt 20 min 55
Table. Comparison of conventional and microw ave
methodologie s
Microwave Synthesis: Examples of Syntheses of
New Acceptors
CN
O CN
OH CN Microw av e 20 W CN Microw av e 20 W CN
+ +
NH O
CN EtONa/EtOH O COOEt EtONa/EtOH COOEt
1
2
O Et CN O
CN
N Microw av e 20 W Et
+ N
S
NH O
O N EtONa/EtOH S
N
O Et O
Et
1 3
. CN CN
CN Microw av e 20 W CN
+
O NH NO2 O
EtONa/EtOH NO2
4
1
O N
N Microw av e 20 W
OH CN Microw av e 20 W CN
+ + CN
N EtONa/EtOH O NH CN EtONa/EtOH O
CN
5 6
CN
O CN
OH CN Microw av e 20 W CN Microw av e 20 W CN
+ +
NH CN F 3C O
CF3 CN EtONa/EtOH F 3C O EtONa/EtOH CN
7
8
Coupling Reactions
NC
CN NC
Bu O cat. P P
y. iper. CN
CN
N + Bu O
Bu F 3C O CN THF, CHCl3, reflux N
Bu CF3
1
NC
NC
CN
O CN
y. iper.
cat. P P O
Bu CN Bu
+
N CF3
F 3C O CN THF, CHCl3, reflux N
Bu Bu
2
S O
. CN
Bu NC
CN
N
Bu S O
EtOH, ref lux
1.5 hr. Bu CF3
CN
N
CN Bu
O LMAJ 22
F 3C CN
S O
CN
Bu NC
CN
N
Bu OTBDMS S O
CN , 20W, 8 min. Bu CF3
N
CN EtOH Bu OTBDMS
O
F 3C CN
LMAJ 24
Translating Microscopic to Macroscopic Electro-Optic Activity
NFb
reff c os3 q
n4
Chromophore-poling Thermal Randomization Chromophore-Chromophore
Field Interaction Electrostatic Interaction
E
Acentric Ordering Isotropic Centric Ordering
< cos3q> = F/5kT < cos3q> =
= f(0)Ep/5kT
(F/5kT)[1-L 2(W/kT)]
Comparison of Potential Functions from Analytic Theory &
Monte Carlo Calculations
Points—Monte Carlo Calculation
Solid Line—Analytic Theory
P A exp
0.3wcos
2 2
2
N 2
Centric Order w s 3 s
. r kT kT
Acentric Order
Comparison of Theory & Experiment
Experiment—Solid
Diamonds
.
kT
Nmax 2
kT
0.48 0.28 4.8 f 2 2
Prediction of the Dependence on Electric Poling Field
.
Theory-Guided Nano-Engineering: Generalization
of the Use of Dendronized Chromophores
New Paradigm: The Concept of Dendronized Chromophores Can
Be Generalized.
.
Statistical Mechanics Guides the Optimization of
Macroscopic Electro-Optic Activity
New Paradigm: Dendrimer synthesis of theoretically-predicted
optimum chromophore shapes—nano-architectural engineering.
: Core moiety
: NLO chromophore moiety
.
: Dendritic moiety
: Crosslinkable moiety
With electric field poling and crosslinking, multi-chromophore
dendrimers assume partially closed umbrella-like shapes. Also, these
dendrimers don’t interpenetrate. These two observations are
supported by theoretical calculations and experimental observations.
Control of Intermolecular Electrostatic Interactions Using
Multi-Chromophore Dendrimers
F F Twice the EO activity of same
F F
O F
chromophore in polymer matrix—
O F O
O
record value at 1.55 microns.
Factor of 2 in thermal stability.
CN
NC N
O O
NC S O O
NC O O
O O
O 1.2
O O O CN
O O S CN
O
O N
NC
CN 1
O
(0)
O F O
O
F F 0.8
33
F O
(t)/r
F F
O O
0.6
33
0.4
N O F
Thermal stability of
r
O O
F F
S
0.2 EO activity at 85 C
O
NC
CN
NC CN
O O
O F
0
O F F 0 20 40 60 80 100
O
Time (hr)
Jen, Dalton et al., J. Am. Chem Soc, 123, 986 (2001)
Dendronized Chromophores: An example
Dendronized chromophore yields 3 times the electro-optic activity
and reduced optical loss (next figure).
F F
F F F F
F F
F O O F
F F
F F F F
. O O N
O N O
F F
O O
F F
S S
F
O
F NC
NC O
O
NC
NC F F
NC CN NC CN
F
O F
F
TCBD
FLDR
F F
F
Perfluorodendron-substituted Chromophore Contributes
Little to Optical Loss in Guest-Host APC Polymer
0.85 dB/cm
at 1.55 mm
0.68 dB/cm
at 1.3 mm
Perfluroinated Chemophore-Containing
Dendrimers: Low Total Optical (Absorption and
Scattering) Loss
Optical Loss (dB/cm)
0.3
0.25
0.2
0.15
0.1
0.05
0
1480 1500 1520 1540 1560 1580 1600
Wavelength (nm)
THERMAL STABILITY—The Need to Lock-In Poling
Induced Acentric Order: Intermolecular Crosslinking
x y z
OH
free-radical copolymerization
with methyl methacrylate and 1. spin cast with
hydroxyethylmethacrylate diisocyanate crosslinker
2. electric field poling
3. thermal crosslinking
HO OH HO OH
3-D crosslinked network
Optimizing Photostability
•Photochemical stability can be improved by chromophore design.
Lumera has demonstrated this.
•Photochemical stability can be improved by the use of scavengers
(see below), packaging, and lattice hardening.
Photo Stability of Different FTC Samples
.
120
100
Intensity Ratio (%)
FTC in Air
80
FTC Sealed
60
FTC w/ Quencher in Air
40
FTC w/ Quencher
20 Sealed
0
0 50 100 150 200
UV Exposure Time (minute)
Improvement in Photostability by Simple Packaging
5
4
Photostability--Packaged in Argon
3
50 mW (1550nm) at the output fiber
Vp
Exposed over 30 days, Vp change negligible
2
Reduce free O2 . Clearly some oxygen is
1 present in this test.
0
0 5 10 15 20 25 30
Time [days]
Processability: An Advantage of Organic
Electro-Optic Materials
•The tailorability of organic materials and particularly of
dendrimers permits integration of organic EO materials with
virtually any material (silicon, silicon dioxide, Mylar, III-V
semiconductors, metals, etc.)
•Hardened organic EO materials are amenable to reactive ion
etching (RIE) and to various photolithographic processes.
Processing is very compatible with semiconductor processing
techniques.
•Organic materials are quite robust (high dielectric breakdown,
good thermal stability at most processing temperatures, high
radiation (gamma, high energy particle) damage thresholds, etc.
•Likely amenable to high volume manufacturing using processing
techniques such as spin casting and dry etching.
•Straightforward fabrication of an array of prototype devices.
Reactive Ion Etching of 3-D Optical Circuits
UV Spin-Casting
Preserves
Oxygen Surface Contour
Ions
Photoresist
Cladding
Core
Cladding
Substrate
Variable Photoresist RIE Slope Waveguide
Exposure Transfer Completion
Fabrication of Vertical Slope Using
Gray Scale Mask Lithography
UV
• Computer Generated Layout
• Variable Transmission Exposure Mask
– Height Exposure Level
– Angles: 0.1-3°
Photoresist
– Heights: 1-15m
– Lengths: 100-2,000m 6
• Entire Device Contoured m)
Height ( 4
– Complex Patterns Possible
– 10m Resolution 2
• Precision of Mask Aligner 0
0 50 100 150 200
• Repeatable Quality
Length ( m)
Fabrication: Shadow Etch
Oxygen Ions
• Shadow Masking of Ions Mask
– Angle RF Power, Gas Pressure,
Time, Mask Dimensions Offset
– Angles: 0.1-3°
– Heights: 1-9m
– Lengths: 200-2,000m Polymer
6
• Fast Prototyping
Height ( m)
– Various Angles From Single Mask 4
– No Extensive Fabrication Steps
2
• Repeatable Quality
0
0 400 800 1200 1600
Length (m)
Tapered Transitions: Minimization of Coupling
Loss
n(active) > n(passive)
Length
small length material loss
large length radiation loss
Fabrication
Lower Electrode Vertical Slope
Upper Coatings Waveguide Ridge
3-D Modulators
Vertical Integration of EO Circuitry with
VLSI Electronics
Polymer EO
Modulator
Electronics
Silicon
Modulation Intensity (#7)
8
Volts
4
0 Vout #7
Modulation Intensity (#1)
8
Volts
Vout #1
4
0
time
Vertical Integration of EO Circuitry with
VLSI Electronics
Vertical Integration of EO Circuitry with
VLSI Electronics
1. 2. 3. CF4 Plasma
4. O 2 Plasma + Metal
Meter = Photoresist
= Metal
= Spin On Glass
= Glass Substrate
= Planarizer
IMPROVED PROCESSABILITY: POLYMER
MICRO-PHOTONIC RING RESONATORS
1 2 3
1, 2, 3
Integrated wavelength add-drop filter
Re-configurable optical waveguide cross connect.
The streets and avenues are fabricated on
different levels with the ring resonator switches
in between at each junction.
Modulates 1 Modulates 2 Modulates 3
WDM modulation module.
Each wavelength modulated by
separate resonate modulator. Laser
1, 2, 3
POLYMER MICRO-PHOTONIC RING
RESONATOR USING ELECTRO-OPTIC POLYMERS
Au upper modulation Complementary
electrode modulated output Au Au
UFC 170 3m
SU-8
4.5m
CLD1 CLD1
UV15 5m
Input Modulated output
Au
Si
CROSSECTION
GND
Why Polymers?
-Wide range of indices of refraction
-Easy fabrication on multiple levels and integration with other devices
-Voltage tunable filter or switch/ modulator using electro-optic polymers
-Compact structure; size limited by index contrast
-Temperature tuning, 0.1nm/C (use as an advantage or eliminate by athermal design
in which thermal expansion of polymer substrate balances dn/dT of waveguide)
INTEGRATED WDM TRANSMITTER-
RECIEVER
Modulates 1 Modulates 2 Modulates 3 1 2 3
1, 2, 3
Laser
1, 2, 3
Transmitter Receiver
Gold
ground
GND
Eye diagram
1 Gb/s, Vpeak = 1 V
Device has ~2GHz BW
Au
BW Kvo n 3 r33
Electrode
VFWHM 2ne d
SU-8 = 2 GHz/V
Large Angle, Fast Response
Spatial Light Modulator (SLM)
Schematic Diagram Experimental Results
Literature Citations
• Dalton, Steier, et al., “Polymeric waveguide
prism based electro-optic beam deflector,” Opt.
Eng., 40, 1217-22 (2001)
Photonic Band Gap Fabrication
Recording beam #1 Recording beam # -1
• Dalton, Steier, et al., “Beam deflection with
Z electro-optic polymer waveguide prism array,”
hologram pattern
being formed # -1'' # 1'
Proc. SPIE, 3950, 108-116 (2000)
# -1 #1
# -1' # 1'' X • Dalton, Steier, et al., “Polymeric waveguide
silicon substrate Y beam deflector for electro-optic switching,”
cladding polymer
w aveguide layer formed by
photopolymer
hexagonal w avevector lattice of the
combination of recording beams Proc. SPIE, 4279, 37-44 (2001)
Phased Array Radar with Photonic Phase
Shifter (1 of 3 approaches)
Dalton, Steier, Fetterman, et
al., IEEE W & Guided
Wave Lett., 9, 357 (1999)
High Bandwidth, Ultrastable Oscillators
(Signal Generators)
• Dalton, Steier, Fetterman, et al., “Photonic control of terahertz systems,”
Terahertz Electronic Proceedings, 102-5 (1998)
• Dalton, Steier, Fetterman, et al., “Electro-optic applications,” in Encyclopedia of
Polymer Science and Technology (J. Kroschwitz, ed) Wiley & Sons, NY,
2001
100 Gbit/sec Analog-to-Digital Converter
(1 of 2 approaches)
• Dalton, Steier, Fetterman, et
al. “Time stretching of 102
GHz millimeter waves using
a novel 1.55 mm polymer
electrooptic modulator,”
IEEE Photonics Technology
Letters, 12, 537 (2000))
• Dalton, Steier, Fetterman, et
al. “Photonic time-stretching
of 102 GHz millimeter waves
using 1.55 mm polymer
electro-optic modulator,”
Proc SPIE, 4114, 44 (2000).
High Bandwidth Optical Modulators
and Switches (The Electrical Problem)
Two bands approach:
• DC-65 GHz direct modulation, use one modulator section;
• 65-130 GHz using upconversion scheme, RF applied to one
modulator section, and LO applied to the other section.
Steier, Bechtel, Dalton et al., Proc. SPIE, 4114, 58-64 (2000).
HYBRID INTEGRATION POLYMER
PHOTONIC MODULE
Electro-optic SSB modulator
Si Electronics
Low loss passive guide
Electro-optic guide
4
1, 2, 3 Amplifying guide
filter
EO phase shifter
Amplifier
OBJECTIVE – Develop photonic modules which integrate multiple waveguide
devices and Si electronics into single package.
APPROACH – Use 3D integration concepts to integrate different photonic polymers
into single photonic circuit. Use adiabatic coupling in tapered guides for low loss
coupling between various materials. Fabricate polymer devices on top of processed
Si integrated electronics. Reduce fiber coupling loss by symmetric design of
passive waveguides
Related docs
