---
title: Spatiotemporal Optical Vortex (STOV)
url: https://www.emergentmind.com/topics/spatiotemporal-optical-vortex-stov
type: topic
---

# Spatiotemporal Optical Vortex (STOV)

Spatiotemporal optical vortices (STOVs) are structured electromagnetic wave packets whose phase singularity is embedded in a joint space–time domain rather than in the transverse spatial plane alone. In the most common free-space formulation, a STOV transports a phase line singularity perpendicular to its propagation direction and carries orbital angular momentum (OAM) with a transverse component; in nonlinear media, closely related objects appear as toroidal nulls that track a propagating pulse and mediate internal energy flow during collapse and filamentation [2301.09105] [1604.01751]. Since the first experimental evidence in nonlinear pulse propagation and the later linear free-space demonstrations, STOVs have developed into a research area spanning singular optics, ultrafast pulse shaping, metasurfaces, nonlinear frequency conversion, plasma photonics, and quantum structured light [1909.11079].

## 1. Physical concept and emergence

A conventional optical vortex is usually defined for a monochromatic beam whose complex field contains an azimuthal phase factor such as $\exp(i l \phi)$ in the transverse plane; its phase singularity lies along the propagation axis, and its OAM is longitudinal. A STOV instead localizes the vortex in space and time. In one common geometry, the phase singularity lies in an $(x,t')$ plane, with $t'=t-z/c$ or $t-z/v_g$ the local time in a co-moving frame, so that the associated OAM is transverse rather than longitudinal [1909.11079] [2301.09105].

The earliest experimental evidence connected STOVs to nonlinear collapse and filamentation of short pulses in material media. In that setting, a STOV was identified as a ring-shaped null in the electromagnetic field about which the phase is spiral, forming a dynamic torus concentric with and tracking the pulse. Depending on the sign of material dispersion, the local electromagnetic energy flow was found to be saddle or spiral about the STOV. These vortices are born and evolve conserving topological charge; they can be simultaneously created in pairs with opposite windings, or generated from a point null [1604.01751].

Subsequent work showed that STOVs are not confined to nonlinear self-action. Linear generation and free-space propagation were demonstrated with pulse-shaping systems, and single-shot amplitude and phase retrieval confirmed that the defining space–time phase circulation survives propagation and mediates internal energy transport within the pulse [1909.11079]. This transition from an emergent nonlinear structure to a deliberately engineered ultrafast field was central to the modern STOV literature.

## 2. Field representations and transverse orbital angular momentum

A standard quasi-monochromatic representation writes the real field as
$$
E(x,y,z,t)=\mathrm{Re}\{\psi(x,y,z,t')\,e^{-i\omega_0 t'}\},
$$
with $t'=t-z/c$ in free space. A prototypical elliptical STOV at $z=0$ is written in normalized coordinates $\tau=t'/t_0$ and $\xi=x/x_0$ as
$$
\psi(\tau,\xi,y)=f(\rho)e^{-i l \phi}\,Y(y),
$$
where
$$
\rho=\sqrt{\tau^2+\xi^2}, \qquad \phi=\mathrm{atan2}(\xi,\tau),
$$
and the ellipticity is
$$
\gamma=\frac{c t_0}{x_0}.
$$
Near the singularity, $f(\rho)\propto \rho^{|l|}$ enforces the intensity zero, and the integer $l$ is the topological charge [2301.09105].

A central issue in the theory of STOVs is the distinction between total, intrinsic, and extrinsic transverse OAM. For free-space or nondispersive propagation, the cycle-averaged transverse OAM flux density may be written as
$$
\langle M_{yz}\rangle=z\langle T_{xz}\rangle-x\langle T_{zz}\rangle,
$$
with
$$
\langle T_{xz}\rangle=\frac{\varepsilon_0}{2k_0}\mathrm{Im}\{\psi^* \partial_x \psi\}, \qquad
\langle T_{zz}\rangle=\frac{\varepsilon_0}{2}|\psi|^2.
$$
The integrated transverse OAM obeys
$$
J_y=J_y^{(i)}+J_y^{(e)},
$$
where $J_y^{(i)}$ is intrinsic OAM about a moving transverse axis through the pulse center and $J_y^{(e)}$ is the opposite extrinsic contribution about a fixed transverse axis. Unlike longitudinal vortices in monochromatic beams, STOVs do not carry any net transverse OAM about a fixed transverse axis crossing the center. Their physically relevant conserved quantity is the intrinsic transverse OAM per photon [2301.09105].

For an elliptically symmetric STOV of frequency $\omega_0$ and charge $l$, the intrinsic transverse OAM per photon is
$$
\frac{J_y^{(i)}}{W}=\frac{\gamma l}{2\omega_0},
$$
with $W$ the pulse energy. Circularly symmetric STOVs therefore carry half the intrinsic longitudinal OAM of circularly symmetric monochromatic light beams with the same $l$ and $\omega_0$. The same analysis rejects the previously proposed expression
$$
\frac{J_y^{(i)}}{W}=\frac{(\gamma+1/\gamma)l}{2\omega_0},
$$
because it diverges as $\gamma\to 0$ or $\infty$ and is not conserved under free-space propagation for a particular STOV [2301.09105].

Propagation does not alter this intrinsic quantity even when the visible singularity structure changes. Under paraxial diffraction,
$$
\partial_z \psi=\frac{i}{2k_0}\nabla_\perp^2\psi,
$$
an initially elliptic STOV may lose elliptical symmetry, split into several charge-1 vortices, or show singularities that disappear and reappear with reversed sign, while $J_y^{(i)}$ remains fixed at $\gamma l/(2\omega_0)$ [2301.09105]. This point addresses a recurrent misconception: singularity counting in propagated intensity-phase slices does not by itself determine the conserved transverse OAM.

The free-space theory has been extended in two directions. In a simple dispersive medium, a modified per-photon transverse OAM
$$
\langle L_{\rm EM}\rangle_y=\frac{\ell}{2}\Bigl(a-\frac{\beta_2}{a}\Bigr)\hbar
$$
was reported for the STOV polariton, with $\beta_2$ the dimensionless group-velocity dispersion and $a$ the pulse eccentricity parameter [2502.17413]. More generally, a quantum theory beyond the paraxial limit models STOVs with an arbitrary tilt of the helical phase front in $\mathbf{k}$-space, thereby interpolating continuously between purely transverse and purely longitudinal OAM directions [2403.01054].

## 3. Generation and diagnostic platforms

Linear free-space generation is commonly implemented with a zero-dispersion $4f$ pulse shaper based on gratings and cylindrical lenses. In this architecture, a spiral phase plate or a single $2\pi$ step in the spatio-spectral Fourier plane imposes the desired phase singularity, and propagation to the far field yields a STOV-carrying pulse. The development of transient grating single-shot supercontinuum spectral interferometry (TG-SSSI) made it possible to recover both the space–time intensity envelope and the full spatiotemporal phase map in a single shot, directly exposing the $2\pi l$ circulation around the singularity [1909.11079].

STOV generation does not require mode-locked laser pulses. A theoretical and experimental study demonstrated STOV formation from a source with partial temporal coherence and fluctuating temporal phase. The implementation used an Yb-doped single-mode fiber laser driven below threshold to produce broadband amplified spontaneous emission (ASE) with FWHM $\simeq 7.8$ nm, corresponding to a coherence time $T_c\approx 450$ fs. A two-dimensional hologram on a spatial-light modulator imprinted the spiral phase in the spectrally dispersed beam, and interferometric fringe scanning verified the topological charge through the number and direction of $\pi$-phase jumps for $\ell=\pm 1$ and $\ell=+2$ [2103.12640]. This established partially coherent STOVs as a distinct source class.

Integrated and metasurface realizations recast STOV generation as a transfer-function problem. In a dielectric slab waveguide with buried metal strips, reflection from an integrated metal–dielectric resonator was modeled by a vectorial spatiotemporal transfer function
$$
H(k_y,\tilde\omega)=R\bigl(k_{y,s}(k_y,\tilde\omega),\omega_0+\tilde\omega\bigr).
$$
When $H(0,0)=0$ and its linear expansion satisfies
$$
H(k_y,\tilde\omega)\approx c_y k_y+c_t\tilde\omega, \qquad
\arg(c_t/c_y)=\pm \frac{\pi}{2},
$$
the reflected envelope becomes proportional to $(\tilde y\mp i\tilde t)e^{-\tilde y^2-\tilde t^2}$, i.e. a charge-$\pm 1$ STOV in all nonzero field components [2308.11319].

Nonlocal metasurfaces introduce a complementary route. In a mirror-symmetry-broken metasurface, the complex transmission coefficient $T(k_x,\omega)$ acquires a spectral vortex in the $(k_x,\omega)$ plane. The topological charge
$$
Q=\frac{1}{2\pi}\oint_C \nabla_{(k_x,\omega)}\phi(k_x,\omega)\cdot d\ell
$$
is quantized, and the corresponding STOV generation is topologically protected because vortices in the synthetic $(k_x,\omega,h_1)$ space are created or annihilated only in opposite-charge pairs [2206.01925]. A related all-dielectric bilayer metagrating uses a lateral shift $\Delta$ to convert a $\Gamma$-point bound state in the continuum into a quasi-BIC with directional radiation and asymmetric coupling. Its transmission zero and branch cut in frequency–momentum space seed a stable STOV under spatiotemporal Gaussian-pulse excitation, with measured phase winding $\ell=-1$ for $\Delta>0$ and $\ell=+1$ for $\Delta<0$ [2603.02837].

At much higher pulse energies, a large-scale four-grating compressor with a phase mask between the second and third gratings has been proposed to generate ultra-intense STOVs. Numerical simulation reported a wave packet with 60 fs duration and 83 J energy in the far field, with a peak intensity $\simeq 5\times 10^{21}\,\mathrm{W/cm^2}$, while a proof-of-principle experiment obtained 1.1 mJ single-pulse STOV energy [2508.14452]. This places STOV generation within the design space of high-peak-power chirped-pulse-amplification systems.

## 4. Propagation, diffraction, focusing, and extended field families

Free-space propagation of STOVs is governed by spatiotemporal diffraction rather than by the mode invariance familiar from monochromatic vortex beams. Experiments and simulations showed that near-field two-lobe or flying-donut patterns transform into far-field STOVs with a donut-shaped envelope while conserving the phase singularity and the associated space–time OAM. The internal Poynting flow is not merely decorative: it redistributes energy from one side of the pulse to the other as a function of local time, thereby mediating the morphological conversion between different field profiles during propagation [1909.11079].

Exact, nonparaxial solutions have also been constructed directly from Maxwell’s equations. Closed forms were obtained by applying suitable differential operators to complex-focus pulses with Poisson-like frequency spectra, producing scalar and vector STOVs that satisfy the free-space wave equation exactly. In this framework, a transverse vortex is imposed by an operator
$$
\widehat W=\frac{1}{k_0}\partial_x \mp i\Bigl(\frac{\alpha}{\omega_0}\partial_t+\frac{\beta}{k_0}\partial_z+i\gamma\Bigr),
$$
and full electromagnetic consistency is recovered through a Hertz-potential operator. A simple ray model in the $xz$ plane explains the characteristic shearing and hyperbolic deformation of the vortex loop under propagation [2406.02382].

Focusing geometry strongly affects STOV fidelity. Conventional focusing introduces spatiotemporal astigmatism because the beam diffracts while the pulse duration remains constant. Temporal focusing overcomes this by using angular dispersion so that the pulse compresses only at the geometric focus and spatial and temporal widths focus and defocus together. In this regime, the effective spectral width obeys
$$
\Delta\omega_{\rm eff}(z)=\Delta\omega\Bigl[1+\frac{z^2}{Z_{Rt}^2}\Bigr]^{-1/2},
$$
and the pulse duration follows
$$
T(z)=T_0\sqrt{1+\frac{z^2}{Z_{Rt}^2}}.
$$
The resulting STOVs evolve self-similarly over an extended focal region, and the OAM vector can be continuously steered from purely longitudinal to strongly tilted orientations by varying the spatial dispersion, focal length, or input beam size [2603.25154].

Diffraction provides both a probe of STOV physics and a practical detection method. For STOV pulses diffracted by a grating, the far-field intensity exhibits a multi-lobe structure in which
$$
N_{\rm lobes}=|\ell|+1,\qquad N_{\rm gaps}=|\ell|.
$$
Each gap corresponds to one topological charge, and the sign of $\ell$ is encoded in the orientation of the lobe array. This “D-method” permits direct, non-interferometric charge readout from a single image [2201.12474].

Several generalized STOV families have been introduced. Cylindrical-vector two-dimensional STOVs carry two orthogonal transverse OAMs in the $x$–$t$ and $y$–$t$ planes. Under Richards–Wolf focusing, radially polarized 2D-STOVs produce wavepackets toward transverse-magnetic toroidal topology, whereas azimuthally polarized 2D-STOVs produce wavepackets toward transverse-electric toroidal topology; their spatiotemporal intensity traces “Yo-Yo” trajectories in $(r,t)$ [2202.07326]. Two 2025 studies developed “perfect” STOVs whose ring size is decoupled from charge. One formulation obtains a perfect STOV by the spatiotemporal Fourier transform of a polychromatic Bessel–Gaussian beam, yielding
$$
\psi(r,\theta)=\frac{i^{\ell-1}}{w^2}(r-r_0)\exp\Bigl[-\frac{(r-r_0)^2}{w^2}\Bigr]e^{-i\ell\theta},
$$
with $r_0$ and $w$ independent of $\ell$ [2501.11426]. Another introduces perfect spatiotemporal optical vortices whose intensity distribution is nearly independent of topological charge and whose azimuthal-dependent phase modulation converts the annular profile into arbitrary polygonal shapes [2501.10055].

## 5. Nonlinear conversion, plasma interaction, and matter-coupled modes

The free-space OAM analysis implies an important constraint on mechanical action. Because STOVs carry no net transverse OAM about a fixed transverse axis through the center, they cannot induce rotation of particles initially at rest about such a fixed axis; however, they can transfer intrinsic transverse OAM to other optical fields and, plausibly, to material degrees of freedom that couple to the moving intrinsic axis [2301.09105].

Second-harmonic generation provided the first direct test of OAM conservation for STOVs in nonlinear optics. In the undepleted-pump and perfect phase-matching limit,
$$
E_{2\omega}(r_\perp,z,t)=A(x,t)^2 e^{i\,2\ell\,\chi_s(x,t)} e^{-i2k_s z} E_0(r_\perp,t)^2,
$$
so the second harmonic doubles the spatiotemporal charge. Experimentally, TG-SSSI measurements of the fundamental and second harmonic confirmed that the average OAM per photon at $2\omega$ is, within experimental uncertainty, twice that at $\omega$, verifying conservation of STOV-based OAM [2012.10806].

High-order harmonic generation extends this transfer to shorter wavelengths. Macroscopic theory and experiment have shown that infrared STOV drivers can generate extreme-ultraviolet STOVs and conjugate spatiospectral vortices. In the simplest scaling picture, the harmonic phase multiplies the driver phase, giving $\ell_q=q\ell$. By increasing the driver charge to 2 or 4, harmonics up to $\ell_q=60$ were reported, and near- and far-field characterization demonstrated that these EUV fields form conjugated spatiotemporal and spatiospectral vortex pairs rather than propagation eigenmodes [2412.01716].

A relativistic version appears in single-slit diffraction from plasma. In the relativistic oscillating-window mechanism, the spatiotemporal structure of a high-intensity STOV driver induces differential electron oscillations on the screen, and the generated harmonics inherit the transverse OAM. The resulting scaling is
$$
l_n=n l_0,
$$
with $l_0$ the driver charge and $n$ the harmonic order. By rotating the slit, the STOV exerts a torque on the plasma and reorients the transverse OAM axis of the generated harmonics [2505.03215].

In dispersive plasma, the field can hybridize with collective response into a STOV polariton. Particle-in-cell simulations of reflection and transmission through a fully ionized hydrogen slab confirmed the existence of the STOV polariton and the value of its transverse OAM, with linear theory matching the simulations up to near-critical densities and near-relativistic field strengths. For higher intensities, optical transverse OAM becomes invested in collective modes of the plasma, and the dynamics are naturally described in terms of a spatiotemporal ponderomotive torque [2502.17413].

## 6. Topological protection, many-singularity dynamics, and quantum formulations

Topological language has become central to the modern understanding of STOV generation. In mirror-symmetry-broken metasurfaces, the zeros of $\mathrm{Re}\,T$ and $\mathrm{Im}\,T$ form nodal lines in the synthetic $(k_x,\omega,h_1)$ space. A fixed-$h_1$ slice intersects these nodal lines in isolated vortices of charge $\pm 1$, and because the total charge on the slice is inherited from the same nodal loop, vortices can only be created or annihilated in opposite-charge pairs. This gives a concrete sense in which STOV generation is robust against disorder and fabrication perturbations [2206.01925].

Internal singularity dynamics within a single wave packet have only recently been resolved. For a three-singularity STOV string, tuning the temporal dispersion produces pronounced oscillations of vortex positions. The outer like-charge vortices approach each other as the group-delay dispersion approaches the Fourier-transform limit, a behavior interpreted as a spatiotemporal attractive effect related to their separation. Replacing the central vortex with an antivortex leads to stretching into dark filaments and eventual annihilation. These behaviors were retrieved experimentally with the Full Interferometric Retrieval of Spatiotemporal Tomography (FIRST) method, which reconstructs the full $E(x,y,t)$ in a single shot [2602.09941]. This suggests that STOVs can support an internal defect dynamics comparable, in formal structure, to other vortex-bearing media.

The quantum theory of STOVs generalizes the semiclassical picture to arbitrary phase-front tilt and nonparaxial regimes. A single STOV mode is defined by a wavepacket operator
$$
\hat a^\dagger_{\xi,\lambda}=\int d^3k\,\xi_l(\mathbf k)\,\hat a^\dagger_{\mathbf k,\lambda},
$$
and the photon-field OAM operator is
$$
\hat{\mathbf L}=-i\hbar\int d^3k\sum_{\lambda=\pm1}\hat a^\dagger_{\mathbf k,\lambda}\,(\mathbf k\times\nabla_k)\,\hat a_{\mathbf k,\lambda}.
$$
Within this framework, Fock and coherent STOV states exhibit non-vanishing longitudinal OAM fluctuations that are absent in conventional monochromatic twisted pulses, and these fluctuations possess a spatial “texture” that can, in principle, be isolated experimentally [2403.01054]. The quantum theory therefore reframes STOVs not merely as shaped classical pulses but as a platform for structured-photon states with directionally tunable OAM and distinctive fluctuation signatures.

Across these developments, several points are now comparatively well established. A STOV is not simply a spatial vortex with short-pulse dressing; its singularity is fundamentally spatiotemporal. Its conserved angular-momentum content in free space is intrinsic and referenced to a moving transverse axis, not a fixed laboratory axis. Its observable manifestations range from nonlinear collapse and free-space diffraction to metasurface phase singularities, exact Maxwell solutions, harmonic up-conversion, and quantum OAM textures. The field continues to broaden, but the unifying object remains the same: a localized optical wave packet with a helical defect in space–time and a correspondingly transverse form of orbital angular momentum.

Source: https://www.emergentmind.com/topics/spatiotemporal-optical-vortex-stov