Transcript Fan, Shanhui - Quantum Electronics Group

Broad-band nano-scale light propagation in plasmonic
structures
Shanhui Fan, G. Veronis
Department of Electrical Engineering and Ginzton Laboratory
Stanford University
In collaborations with Professors Mark Brongersma and Peter Peumans
Supported by the Stanford-GCEP, and NSF-NIRT
Organic Solar Cell
• Silicon and Compound Semiconductor Cells
• High efficiency (~30%), but high cost.
• Organic Solar Cell
• Low cost, but low efficiency (<5%)
Operational Principle of Organic Solar Cells
D
A
Photon absorption
Exciton diffusion
Charge-separation
Charge-transfer
Charge-collection
Exciton Diffusion Process
Optical
absorption
~ 100nm
D
Exciton
Diffusion
~ 10nm
A
• Deliver light directly to the DA interface.
• Enhance light absorption at the DA interface.
• Efficiently extract carriers once they are generated.
 Nanoscale manipulation of light and electrons using metals.
From single-wavelength to deep sub-wavelength scale
1 m
Scale of SOI waveguide
Vlasov et al, IBM, 2004
Scale of a transistor, < 100nm
Core of a single mode fiber: ~ 10 m
Kobrinsky et al, Intel, 2004
The need for nano-photonics in optical interconnect
Micron scale dielectric waveguide
Nanoscale photodetector or latch
Stanford MURI on Plasmonics (Brongersma, Miller, Fan)
•The relevant length scales here:
modal diameter ~ 50-100 nanometer; propagation distance ~ 10 micron
• Broadband width.
Two-conductor configuration: perfect metal
d

Perfect metal
Air
Frequency /p
1.5
1
0.5
0
0
0.5
1
1.5
Wavevector k/kp
2
d

Plasmonic metal
 2p
    1
   i 
Air
Frequency /p
Two-conductor configuration: plasmonic metal
1.5
1
Band 2
0.5
Band 1
0
0
E. N. Economu, Physical Review B, 182, 539 (1969)
0.5
1
1.5
Wavevector k/kp
2
Band 1
Magnetic field
Low
Frequency
Intermediate
Frequency
(infrared and visible)
High
Frequency
(ultra-violet)
Electric field
Band 1
Magnetic field
Low
Frequency
Intermediate
Frequency
(infrared and visible)
High
Frequency
(ultra-violet)
Electric field
Plasmonic slot waveguide
air (n=1)
metal
metal
SiO2 (n=1.5)
slot dimension: 50~100 nm
• The corresponding microwave structure does not support a true
bound mode in this asymmetric geometry.
• Intermediate regime showing both microwave and plasmonic
behaviors.
G. Veronis and S. Fan, Optics Letters, 30, 3359 (2005)
Bound-mode in plasmonic slot waveguides
w = 50 nm
 =1550nm
 =1550nm
• Calculated using tabulated experimentally determined dielectric
function of silver at all frequencies.
• True bound mode.
• Guiding bandwidth exceeding 100THz.
Modal Diameter << Wavelength
y
 =1550nm
• Mode diameter is small even when the phase index approaches
that of silica.
• Mode diameter ~ 90 nm at 1.55 micron wavelength.
• Mode diameter weakly dependent upon frequency.
G. Veronis and S. Fan, Optics Letters, 30, 3359 (2005)
Far field v.s. near field
y
• Modal size determined by the near field.
• Exponential decay only appears far from waveguide, where the field
amplitudes are already negligible.
Nano-scale waveguide bends
Z0
Z0
Ag
air
Ag
50 nm
•Complete transmission through sharp bends from microwave to
optical wavelength range
G. Veronis and S. Fan, Applied Physics Letters, 87, 131102 (2005).
Coupling between dielectric and MDM guide
• Direct butt coupling.
• 70% Coupling efficiency.
• Non-adiabatic taper.
• Designed with micro-genetic
algorithm.
• 93% Coupling efficiency.
G. Veronis and S. Fan, Optics Express (submitted)
Summary
• Proper design of metallic nano-structures leads to subwavelength propagating modes with very broad bandwidth.
• Such modes might be exploited for nano-scale manipulation of
light in energy and information applications.
• Plasmonic crystals may also be used to substantially modify
optical absorption and thermal emission properties.