Photonic Crystals: The Bandgap of Light
A semiconductor crystal forbids electrons of certain energies from existing inside it — that's the electronic bandgap that makes transistors possible. A photonic crystal does the same trick for light: a periodic pattern of refractive index can forbid entire colours of light from propagating at all, no matter the angle.
1. From electronic bands to photonic bands
In a semiconductor, electrons moving through a periodic lattice of atoms experience Bragg scattering off the regularly-spaced ion cores. This scattering opens up an energy bandgap — a range of electron energies for which no propagating wave state exists inside the crystal, no matter the direction of travel. This single idea underlies every diode and transistor ever built.
In 1987, Eli Yablonovitch and Sajeev John independently realised that the same physics applies to photons if you replace the periodic atomic lattice with a periodic variation in refractive index at a length scale comparable to the wavelength of light. The result is a photonic crystal: a structured dielectric medium with a photonic bandgap — a band of frequencies that simply cannot propagate through the material, reflected with near-100% efficiency regardless of incidence angle or polarisation (for a full 3D bandgap; a 1D structure like a Bragg mirror only blocks a limited range of angles).
For Bragg-type interference to open a gap, the periodicity of the refractive-index pattern must be on the order of half the wavelength inside the medium. For visible light (400–700 nm) that means structuring matter with features a few hundred nanometres across — a scale only reachable with modern nanofabrication, or found ready-made in certain biological structures.
2. The 1D case: a Bragg mirror
The simplest photonic crystal is a stack of alternating thin layers with refractive indices n₁ and n₂ (say n₁ = 1.45 for silica, n₂ = 2.2 for titania), each a quarter-wave thick at the design wavelength λ₀:
At every interface a fraction of the light reflects. With quarter-wave layers, the reflections from every interface arrive back at the front of the stack exactly in phase — they add constructively no matter how many layers you stack, producing a mirror whose reflectivity approaches 100% as the number of layers grows, for wavelengths near λ₀. Wavelengths far from λ₀ don't satisfy the quarter-wave condition, so their reflected components partially cancel and the stack becomes increasingly transparent — this is the origin of the photonic bandgap in frequency space.
The bigger the index contrast n₂/n₁, the wider the forbidden band. This is why photonic-crystal designers hunt for high-index materials (silicon, titania, gallium arsenide) to pair against low-index ones (silica, air) — air (n=1) paired with silicon (n≈3.5) gives one of the widest gaps achievable.
3. The transfer matrix method
To compute the transmission spectrum of an arbitrary stack of layers exactly (not just at the design wavelength), optical engineers use the transfer matrix method (TMM). Each layer of thickness d and index n contributes a 2×2 matrix relating the forward- and backward-travelling field amplitudes on either side:
where δ = (2π/λ) · n · d · cos θ is the phase thickness and η is the optical admittance of the layer.
The matrices for every layer are multiplied together in order to get a single characteristic matrix for the whole stack, from which the overall reflectance R(λ) and transmittance T(λ) can be read off directly. Because it's just repeated 2×2 matrix multiplication, TMM scales effortlessly to stacks of hundreds of layers and runs comfortably in real time — exactly the algorithm used by our interactive simulation below to plot the live transmission spectrum as you change the number of layers and the index contrast.
4. 2D and 3D photonic crystals
A 1D stack (Bragg mirror) only blocks light travelling roughly along its axis, over a limited angular range. To get a complete bandgap — one that blocks a given frequency band from every direction and every polarisation simultaneously — the periodicity has to exist in two or three dimensions:
- 2D crystals: a triangular or square lattice of holes drilled through a slab (or rods standing in air) can produce a full in-plane bandgap. Used in photonic-crystal fibers, where a hollow or solid core is surrounded by a lattice of air holes running the length of the fiber.
- 3D crystals: structures like the "woodpile" (stacked perpendicular rod layers) or inverse-opal lattices (a face-centred-cubic arrangement of air spheres in a high-index matrix) can open a complete 3D bandgap — the photonic equivalent of a perfect semiconductor gap, though fabricating them at optical wavelengths remains difficult and expensive.
5. Photonic crystals in nature
Evolution discovered photonic bandgaps long before physicists did. The iridescent blue of a Morpho butterfly's wing scales, the shimmering colours of an opal, and the metallic sheen of some beetle shells all come from periodic nanostructures that selectively reflect certain wavelengths through exactly this interference mechanism — not from any pigment. This structural colour is why the blue never fades (there's no dye to bleach) and why it can shift with viewing angle for structures without a full 3D bandgap.
6. Engineering applications
- Dielectric mirrors: laser cavity end-mirrors with reflectivities exceeding 99.999%, impossible with metal coatings.
- Photonic-crystal fiber: a lattice of air holes lets light be guided down a hollow core (avoiding material absorption and nonlinearity entirely) or engineers dispersion in ways ordinary step-index fiber cannot match.
- Distributed feedback (DFB) lasers: a built-in periodic grating forces the laser to emit a single, extremely stable wavelength — essential for telecom transmitters.
- Optical filters and anti-reflection coatings: precisely tuned multilayer stacks on camera lenses and solar panels.
💠 Explore a photonic crystal live
Build your own 1D photonic crystal (Bragg mirror) with the transfer matrix method: adjust layer count, thickness, and refractive-index contrast and watch the transmission spectrum and bandgap update in real time.
Open simulation →