Transcript Low - The University of Sydney

Nobel lecture: fibre optics
Martijn de Sterke
School of Physics and CUDOS
University of Sydney
[email protected]
Setting the scene
• Most long-distance phone calls or web download
go through optical fibre as an optical pulse
• Low cost and excellent quality
• From Mbit/s  Tbit/s in 40 years
• Fiber optics has transformed society by enabling
internet
2009 Physics Nobel Prize was awarded to Charles
Kao “for groundbreaking achievements concerning
the transmission of light in fibers for optical
communication”
Outline
•
•
•
•
•
Early history
Why light?
What did Kao do?
What others did
Other developments
History
• Light guided
by water
(Colladon,
1880’s)
• Later
incorrectly
attributed to
Tyndall
History
• US patent
taken out to
light house
using single
bright source
(Wheeler
1880’s)
• Never made
to work &
overtaken by
light bulb
Gastroscopes
• A flexible instrument that is put through the mouth
and down the esophagus to view the stomach.
• Early on, straight tube: “one of the most lethal
instruments in the surgeon’s armamentarium”
• Later semiflexible: This tube was 75 cm long and
11 mm in diameter. About 1/3 of the length of the
tube toward the tip could bend to somewhat.
http://www.olympus-global.com/en/corc/history/endo/
Gastroscopes
• Currently fiber bundle in which each fiber
contributes one pixel to picture.
• Required key development: how to isolate different
fibers in bundle
- Uncoated leads to loss and image blurring as light
crosses between fibres
- Silver coating leads to excessive loss
• Solution: coat fibres with low-index
material so light stays in fiber by
total internal reflection (1956)
Total internal reflection
• Almost all fibers still
use same principle:
- Low index cladding
surrounding high-index
core
- Guidance by Total
Internal Reflection at
core-cladding interface
2009 Nobel prize
• This and other applications (military) all involve
short propagation lengths.
• 2009 Physics Nobel prize awarded to Charles Kao
(Standard Telecommunication Labs, UK, Chinese
University of HK for "for groundbreaking
achievements concerning the transmission of light
in fibers for optical communication“
Interlude—why light?
• Why use light for optical communications?
• Need to understand how information is carried
by wave—modulation
• Example: AM modulation
More on AM modulation
• AM modulation only makes sense if signal
frequency is smaller than carrier frequency.
• Otherwise carrier would modulate the signal!
• Very general result
Bandwidth
•
•
•
•
•
•
AM band runs from 520 kHz–1610 kHz
Limited by atmospheric propagation
Radio stations are separated by 9 kHz
Signal limited to frequencies of up to 3-4 kHz
Have (1610-520)/9≈120 radio stations in AM band
What limits it?
What do we conclude?
• Limited by highest AM band frequency 1610 KHz
• Ultimately limited by physical properties of
transmission medium (atmosphere).
• To get substantially more information need to
operate at higher frequency (e.g. FM band)
• How can I maximize the amount of information?
• By using the highest possible carrier frequencies.
Claude Shannon—father of information
theory
What do we conclude?
• What is highest frequency wave we can control?
• Light has a frequency of ~1015 Hz.
• Whether we can use it depends on transmission
medium.
• Atmosphere not always transparent (rain, fog)
• Need another medium—optical fiber
Millimeter waves
• In 60’s light was not deemed viable for telecomms
• Most labs pursued millimeter wave carriers—
frequency much higher than then existing systems
• Need perfectly straight waveguides to limit losses;
perhaps OK for US but not densely populated UK.
“To avoid troublesome mode problems, the waveguides
were buried in trenches as straight as arrows. The tubes
were filed with pure dry nitrogen, because oxygen in the air
absorbs millimeter waves. Nobody pretended that it was
going to be cheap or easy ….”
What did Kao do? 1
• First to have complete vision of light for
telecommunications, consisting of
-
Single-mode, step-index optical fibre
Low loss glasses (100 dB/km to < 20 dB/km)
High speed modulation of lasers
Specs better than microwave system
What did Kao do? 2
• First measurement of loss in optical glasses,
What did Kao do? 3
• Shows conventional wisdom to be wrong using
commercial glass samples
43 dB/km equivalent
to factor 2 in 70 m
What did Kao do? 4
• He sold his vision by traveling the word and
spreading the “gospel” …… including at Corning
in upstate New York.
• Had vision, understood
fundamental physics and
had passion
What Corning did
• Fiber losses were considered most serious
challenge to fiber communications
• Solved by Corning: Keck, Maurer & Schultz
• Current fabrication techniques all directly derive
from their invention
Fibre fabrication
• Fibers are drawn in draw tower from a preform;
well known at the time
Fibre fabrication
• Key step was in the preform
• Used fused silica, poorly researched since high
melting point
• Key insight: SiCl4 in Oxyhydrogen flamewhite
dust (soot) “the purest silica ever made”
• Turn sand into chloride compound; heat and SiCl4
evaporates readily, leaving impurities (iron)
behind
Preform fabrication
• Flame hydrolysis—used vacuum cleaner to suck
flame into the tube and deposit soot.
• Core initially doped with titanium, but brittle fibres
• Later replaced by germanium (via GeCl4)
• All current fiber fabrication processes are
variations on this theme.
Fiber losses
Use low loss region around 1550 nm
Other necessary ingredients
• Semiconductors lasers
• Invented at Bell Labs, Yoffe Institute (GaAs)
and at Lincoln Labs (InGaAsP)
Other necessary ingredients
• Full theory first developed by Kao & Hockham,
but very unwieldy
• Simplified theory later developed (difference
between refractive indices of core and cladding
very small), including by Allan Snyder (ex ANU)
• Provided the physics insight for convenient
design and for building intuition.
SMF 28 fiber
Current situation—WDM
• Wavelength division multiplexing
• Loss region around 1550 nm chopped up in a
grid of width 50-100 GHz (channels)
• Have about 100 channels
• Similar to different radio stations on AM band
• Each channel separately processed at beginning
and end of fibre
Current situation—undersea
Current situation—EDFAs
• Even though losses are very low (0.2 dB/km) they
are not negligible
• Erbium-doped amplifier invented in 1987: amplify
signal while traveling through the fiber
• Example of all-optical processing
Current situation
Capacity
keeps
going up
Conclusions
• Initially optical fibre technology looked like many
other promising technologies
• New technology needs people who understand
basic principles and are passionate
The
Photonic
integration
The
CUDOS
vision revolution
Yesterday
• Expensive
• Space, Power, Heat
• Low bandwidth
Today
• Miniaturization - optics on a chip
• Ultra-high bandwidth
• Limited functionality
• Telecom focused
~ centimetres
The Future (CUDOS)
• High functionality
• Low footprint in space and energy
• Disruptive across all science
micron scale
Paradigm shift in photonics
• Natural optical materials:
– Have limited properties (Positive index of refraction, moderate
nonlinearity)
– And limited performance (Diffraction limits, bandwidth limits)
• Can we do better?
Feynman (1959!): “There’s plenty of
room at the bottom.”
Atoms
10-10 m
Natural
Crystals
10-9 m
Metamaterials
Scale gap
10-8 m
10-7 m
Mid-IR
science
Telecom/
Internet
Visible
10-6 m
10-5 m
How will we do it?
New ways to control light with “meta” materials;
A revolutionary photonic integration “platform”.
Tap water
“Negative”
meta-water
Thank you!
[email protected]
Thank you!
[email protected]