Optical Vortex (Orbital Angular Momentum)

The intuition — a corkscrew of phase

Picture an ordinary laser beam. Its wavefronts — the surfaces of constant phase — are flat, parallel discs marching forward like the pages of a book. Now imagine grabbing one edge of each disc and lifting it a little, so the whole stack becomes a continuous spiral ramp, like a multi-story parking garage with no floors, just one endless helical surface. That is an optical vortex: a beam whose wavefront is a helicoid.

To screw forward smoothly, the phase has to advance from 0 to 2π (or a multiple of it) as you walk once around the beam axis. But that creates an immediate problem at the very center. If the phase is 0 on one side and 2π on the other and everything in between, then on the axis the phase would have to be every value at once. That is impossible — so the only escape is for the field amplitude to drop to exactly zero on the axis. The result is a thread of pure darkness running down the middle of the beam: a phase singularity, also called an optical vortex line.

This is why a vortex beam never lands as a dot. On a screen it shows up as a bright ring — a donut — with a perfectly dark hole punched through the center. And because the phase circulates around that hole, the beam carries angular momentum: orbital angular momentum (OAM), the optical analog of a planet's orbital momentum, distinct from the "spin" you get from circular polarization.

The governing physics

The defining feature is an azimuthal phase factor tacked onto the field. Writing the complex field amplitude in cylindrical coordinates (r, φ, z):

E(r, φ, z) = A(r, z) · exp(i·ℓ·φ) · exp(i·k·z)

The exp(iℓφ) term is the whole story. As φ runs once around the axis (0 → 2π), the phase changes by 2πℓ. The integer ℓ is the topological charge — it counts how many times the phase wraps, and equivalently how many intertwined helical sheets the wavefront has.

Because the on-axis amplitude must vanish, the radial profile A(r, z) goes to zero at r = 0. The exact eigenmodes of a laser cavity or free-space propagation are the Laguerre–Gaussian (LG) modes:

LG(p, ℓ) ∝ (r√2 / w)^|ℓ| · L_p^|ℓ|(2r²/w²) · exp(−r²/w²) · exp(i·ℓ·φ)

where w is the beam radius, p ≥ 0 is the radial index (number of bright rings minus one), and L is an associated Laguerre polynomial. The r^|ℓ| prefactor is what forces the central null and makes the donut wider for larger |ℓ|.

The angular momentum follows from the field. Allen et al. showed in 1992 that each photon in such a beam carries:

L_orbital = ℓ·ℏ   per photon     (orbital angular momentum)
S_spin    = σ·ℏ   per photon     (spin, σ = ±1 for circular polarization)
J_total   = (ℓ + σ)·ℏ            (total angular momentum)

Here ℏ = 1.0546 × 10⁻³⁴ J·s is the reduced Planck constant. Crucially, ℓ is unbounded while σ is locked to ±1 — that asymmetry is the source of the vortex's usefulness.

Spin vs orbital angular momentum of light

Light can be "twisted" in two genuinely different ways, and confusing them is the single most common mistake. Spin is about the direction the field vector rotates; orbital is about the shape of the wavefront in space.

Property	Spin angular momentum (SAM)	Orbital angular momentum (OAM)
Physical origin	Circular polarization — E-field vector rotates	Helical wavefront — phase ramp exp(iℓφ)
Value per photon	σℏ, with σ = ±1 only	ℓℏ, with ℓ any integer
Upper limit	±ℏ (hard cap)	Unbounded — ℓ = 10⁴ demonstrated
Beam cross-section	Ordinary spot (can be Gaussian)	Ring / donut with dark core
Number of states	2 (left / right circular)	Infinite, mutually orthogonal
How to make it	Quarter-wave plate	Spiral phase plate, fork hologram, SLM, q-plate
How to measure it	Polarizer + wave plate	Interference fringes / fork pattern / mode sorter
Interconvertible?	Yes — a q-plate converts SAM ↔ OAM (spin-to-orbital coupling)

A single beam can carry both at once, and the total angular momentum is simply (ℓ + σ)ℏ per photon. They are independent degrees of freedom that both happen to be "rotations" of light.

Topological charge and the donut by the numbers

The charge ℓ is a robust integer: you cannot continuously morph a charge-2 vortex into a charge-1 vortex without the field passing through a singular point, so ℓ behaves like a conserved quantum number for the beam. Higher charge means a faster phase twist, more OAM, and — because the r^|ℓ| factor pushes light away from the axis — a physically larger dark core and a wider bright ring. The peak-intensity radius of an LG mode scales as:

r_peak = w·√(|ℓ|/2)        (ring radius grows ~ √|ℓ|)

Topological charge ℓ	OAM per photon	Phase windings per loop	Ring radius (× beam w)
0 (no vortex)	0	0 — flat wavefront	0 (solid spot)
±1	±ℏ	1 × 2π	≈ 0.71 w
±2	±2ℏ	2 × 2π	≈ 1.00 w
±3	±3ℏ	3 × 2π	≈ 1.22 w
±10	±10ℏ	10 × 2π	≈ 2.24 w
±100	±100ℏ	100 × 2π	≈ 7.07 w

The sign of ℓ sets the handedness — whether the phase spirals clockwise or counter-clockwise — which is why reflecting a vortex off a mirror flips ℓ to −ℓ (the helix that screwed away from you now screws toward you).

A worked number — torque on a trapped bead

How much rotational push does a vortex actually deliver? Consider a 1 mW vortex beam at 1064 nm with charge ℓ = 8, focused onto a micron-sized particle held in optical tweezers. Each photon carries ℓℏ = 8 × 1.055 × 10⁻³⁴ = 8.4 × 10⁻³⁴ J·s of OAM. The photon energy is E = hc/λ = (6.63 × 10⁻³⁴ × 3 × 10⁸) / (1.064 × 10⁻⁶) = 1.87 × 10⁻¹⁹ J, so a 1 mW beam delivers:

photon rate  = P / E      = 1e-3 / 1.87e-19   ≈ 5.3e15 photons/s
max torque   = (ℓℏ)·rate  = 8.4e-34 × 5.3e15  ≈ 4.5e-18 N·m

That atto-newton-meter torque sounds tiny, but at the micron scale viscous drag is also tiny, and such beams routinely spin trapped cells and microgears at tens to hundreds of revolutions per second. The full transfer also depends on absorption and particle shape, but the OAM-per-photon bookkeeping above sets the hard ceiling.

Where optical vortices show up

• Optical spanners / micromachines. The OAM spins absorbing or birefringent microparticles — a light-driven motor. Used to rotate trapped cells, drive microfluidic pumps, and turn microscopic gears.
• STED super-resolution microscopy. A donut-shaped depletion beam (a vortex) switches off fluorescence everywhere except the dark center, shrinking the effective spot far below the diffraction limit (the 2014 Nobel-winning trick).
• High-capacity communications. Distinct ℓ values are mutually orthogonal, so they form independent channels. OAM multiplexing has pushed free-space links past 100 terabits per second and adds capacity to fiber without extra spectrum.
• Vortex coronagraphs. Placing a charge-2 or charge-4 phase mask at a telescope's focus dumps the on-axis starlight outside the aperture, letting astronomers image faint exoplanets and disks right next to a bright star.
• Quantum information. OAM provides a high-dimensional alphabet (qudits) for entanglement and quantum key distribution — one photon can encode far more than a single bit.
• Material processing. Vortex laser pulses drill chiral, helical microstructures and can transfer twist into the structure of fabricated materials.
• Astrophysics and beyond. Light scattered or emitted near rotating bodies (and even from a rotating black hole) can pick up OAM, a proposed probe of spin in extreme environments.

How to make one

Every recipe imprints the exp(iℓφ) phase ramp onto a normal beam:

• Spiral phase plate. A transparent disc whose thickness increases like a spiral staircase, adding a φ-dependent optical path. Simple and efficient, but each plate is fixed to one ℓ and one wavelength.
• Forked diffraction grating / hologram. A grating with an ℓ-pronged "fork" dislocation at its center. The first diffracted order comes out as a charge-ℓ vortex — this fork is also the tell-tale signature when you interfere a vortex with a plane wave.
• Spatial light modulator (SLM). A programmable liquid-crystal array that can paint any phase pattern, so a single device makes any ℓ on demand. The workhorse of modern OAM labs.
• q-plate. A patterned liquid-crystal element that couples spin to orbit: feed in circularly polarized light and it comes out as a vortex, converting σℏ of spin into ℓℏ of OAM.

Common misconceptions and edge cases

• "A vortex is just circularly polarized light." No — polarization is spin (capped at ±ℏ); the vortex is a wavefront/phase property (OAM, unbounded). A linearly polarized beam can still be a vortex.
• "The dark center is a shadow or an obstruction." Nothing blocks it. The darkness is forced by destructive interference at the phase singularity — remove every obstacle and the null is still there.
• "Higher ℓ just means more brightness." Higher ℓ means more twist per photon and a wider donut, not more power. Total power is set by the source, independent of ℓ.
• "Topological charge can be a fraction." A pure, stable vortex requires integer ℓ. "Fractional" vortices exist but are unstable mixtures that immediately break into a stream of integer-charge vortices on propagation.
• "Reflection preserves the charge." A mirror flips handedness: a +ℓ vortex reflects to −ℓ. (This is exactly the edge case the visualization ends on.)
• "OAM and the Poynting vector point along the beam." In a vortex the local Poynting vector spirals — energy flows along a helix, tilted azimuthally — which is the field-level reason the beam carries angular momentum at all.