Introduction to nanophotonics Alexey Belyanin Department of Physics

Introduction to nanophotonics Alexey Belyanin Department of Physics, Texas A&M University

Outline • What is nanophotonics? – motivation • • • Principles of light guiding and confinement Photonic crystals Plasmonics Optical chips and integrated photonics Bio-nanophotonics – Biosensors, nanoshells, imaging, therapy • Terahertz photonics • Exotic stuff: negative index materials, quantum optics of semiconductor nanostructures, etc.

Nanophotonics: control of light at (sub-)wavelength scale near-IR: 700 -2000 nm Optical communications window: 1300 -1600 nm (Why? ) Sub-wavelength scale = nanoscale for visible/near-IR light Violates fundamental laws of diffraction? ? Not applicable to near field Not applicable to mixed photon-medium excitations: polaritons, plasmons

What kind of medium can carry optical frequencies? Air? Only within line of sight; High absorption and scattering Optical waveguides are necessary! Copper coaxial cable? High absorption, narrow bandwidth 300 MHz Glass? Window glass absorbs 90% of light after 1 m. Only 1% transmission after 2 meters. Extra-purity silica glass? !

Loss in silica glasses Loss per km, d. B What is d. B? Increase by 3 d. B corresponds to doubling of power Maximum tolerable loss Wavelength, nm Transmisson 95. 5% of power after 1 km P = P(0) (0. 955)N after N km P = 0. 01 P(0) after 100 km: need amplifiers and repeaters Total bandwidth ~ 100 THz!!

Optical fibers Made by drawing molten glass from a crucible 1965: Kao and Hockham proposed fibers for broadband communication 1970 s: commercial methods of producing low-loss fibers by Corning and AT&T. 1990: single-mode fiber, capacity 622 Mbit/s Now: capacity ~ 1 Tbit/s, data rate 10 Gbit/s

Fibers opened the flood gate Bandwidth 400 THz would allow 400 million channels with 2 Mbits/sec download speed! Each person in the U. S. could have his own carrier frequency, e. g. , 185, 674, 991, 235, 657 Hz.

Limitations of optical communications In optical communications, information is transmitted over long distances along optical fibers However, if we want to modify, add/drop, split, or amplify signal, it needs to be first converted to electric current, and then converted back to photons

Electronic circuits: 45 nm wires, 1 million transistors per mm 2 Computing is based on controlling transport and storage of electric charges Computing speed is limited by inertia of electrons

The interconnect bottleneck • 109 devices per chip • Closely spaced metal wires lead to RC delay • Huge power dissipation due to Ohmic losses

Can electronic circuits and transmission channels be replaced by photonic ones? ! Using photons as bits of information instead of electrons would revolutionize data processing, optical communications, and possibly computing What is wrong with using electric current instead of photonic beams? Good: electrons are small; devices are potentially scalable to a size of a single molecule Bad: electric current cannot be changed or modulated fast enough. Speed is limited to nanosecond scale by circuit inductance and capacitance. As a result, data rate is limited to a few Gb/s and transmission bandwidth to a few GHz. Photons travel much faster and don’t dissipate as much power

THE DREAM: could we replace electric signal processing by all-optical signal processing? IBM website Futuristic silicon chip with monolithically integrated photonic and electronic circuits This hypothetic chip performs all-optical routing of mutliple N optical channels each supporting 10 Gbps data stream. N channels are first demultiplexed in WDM photonic circuit, then rearranged and switched in optical cross-connect OXC module, and multiplexed back into another fiber with new headers in WDM multiplexer. Data packets are buffered in optical delay line if necessary. Channels are monitored with integrated Ge photodetector PD. CMOS logical circuits (VLSI) monitor the performance. Electrical pads are connecting the optoelectronic chip to other chips on a board via electrical signals.

However, dimension of optical “wires” is much larger than that of electric wires Or optical fiber cross-section We need to confine light to at least 1020 times smaller size than the fiber diameter

What is the minimum confinement scale for light at a given wavelength? • Wave equation • Confinement in a metal box • Total internal reflection

EM waves in a bulk isotropic medium E k H Phase velocity - relative dielectric permittivity; refractive index Note: wavelength in a medium is n times shorter than in vacuum

How to confine light with transparent material? ? Total internal reflection! Water: critical angle ~ 49 o

Total internal reflection n 1 > n 2

Dielectric waveguides n > n’ What is the minimum size of the mode confined by TIR?

Basic waveguide geometries

Dielectric waveguides are used in all semiconductor lasers

For integrated photonic circuits we need to use silicon and CMOS-compatible technology Silicon on insulator waveguides nc=1 nw=3. 6 ns=1. 5

The dream No silicon lasers or amplifiers (why? ) No silicon detectors at wavelengths 1. 3 -1. 6 m (why? )

Why there are no silicon lasers k 1 = k 2 + kph ; kph << k 1, 2 k 1 ~ k 2 Only vertical (in k-space) transitions are allowed Only direct gap semiconductors are optically active k 1 k 2 Silicon Ga. As

Si. O 2 doped with active erbium ions and with silicon nanocrystals From L. Pavesi talk 2005

Only simple devices have been built so far: Modulators, beam splitters, etc. Beam A Intel silicon photonic modulator Possible uses: Rack-to-rack, Board-to-board, Chip-to-chip connections Beam B Modulation of light using nonlinear optics: dependence of the refractive index from light intensity I (Kerr effect) By changing n 2, we can shift phases of the beams A and B with respect to each other:

Coupling light into a thin film waveguide can be a problem

Coupling a 5 - m diameter beam from fiber tip into 0. 4 - m thin film (Intel) Tapered channel grating

Guiding light in a low-index core? ! Almeida. OL 2004 Central region is 50 nm, but evanescent field still extends to about 500 nm

Evanescent field can be used for inter-mode coupling and for sensors Intel Cornell group Nature 2004

Evanescent field sensors with substrate sensitized to a specific molecule

Adsorbed molecules change the excitation angle of EM mode

Can we do better than a thin film dielectric waveguide (mode size about 0. 5 m, bending radius a few m)? Photonic crystals! Periodic modulation of dielectric constant blocks the transmission of light at certain frequencies

One dimensional photonic crystal: Bragg grating d Bragg reflection Yablonovitch, Sci. Am. 2001

Yablonovitch, Sci. Am. 2001

Photon momentum conservation d kin + Kg = kout When Kg = 2 kin: incoming wave is reflected

Photonic band gap is formed Light is blocked at certain frequencies: PBG Group velocity tends to 0 at the edge of PBG -> enhancement of light intensity n 1 n 2

Yablonovitch, Sci. Am. 2001

“Photonic crystals – semiconductors of light” Semiconductors Periodic crystal lattice: Potential for electrons Length scale ~ 3 -6 A Photonic crystals Periodic variation of dielectric constant Length scale ~ Natural structures Control electron states and transport From M. Florescu talk (JPL) Artificial structures Control EM wave propagation and density of states

Natural opals Striking colors even in the absence of pigments From M. Florescu talk (JPL)

Yablonovitch, Sci. Am. 2001

Artificial Photonic Crystals Requirement: overlapping of frequency gaps along different directions § § § High ratio of dielectric indices Same average optical path in different media Dielectric networks should be connected Woodpile structure S. Lin et al. , Nature (1998) From M. Florescu talk Inverted Opals J. Wijnhoven & W. Vos, Science (1998)

Some 3 D crystal designs based on diamond lattice By the way, why we don’t see photonic band gap in all crystals? Yablonovitch, Sci. Am. 2001

Photonic crystals can reflect light very efficiently. How to make them confine and guide light? Introduce a defect into the periodic structure!! • • Creates an allowed photon state in the photonic band gap Can be used as a cavity in lasers or as a microcavity for a “thresholdless” microlaser

1 D structure with defect: Vertical Cavity Surface-Emitting Laser (VCSEL) Edge-emitting laser VCSEL

2 D structure: photonic crystal fiber Extra tight mode confinement, high mode intensity, high nonlinearity First commercial all-optical interconnect based on PC fibers (Luxtera)

Photonic circuits? From Florescu talk Intel Note T-intersections and tight bends, as in electric wires. You cannot achieve it in dielectric waveguides!