Space-Time Double-Slit Diffraction

• Space-time double-slit diffraction is a phenomenon that reformulates traditional Young interference by distributing coherent alternatives across both space and time.
• The method employs various formulations—including temporal slits, reciprocal reconstructions, and momentum-space approaches—to reveal interference in phase, energy, and arrival time.
• Experimental realizations with electrons, photons, and matter waves validate the role of coherent phase accumulation and challenge the view that interference implies non-local action.

Space-time double-slit diffraction denotes, in the literature surveyed here, a class of interference phenomena in which the canonical Young configuration is reformulated as a problem of coherent alternatives distributed across spacetime rather than across space alone. The two alternatives may be two spatial apertures sampled at different times, two temporal openings of a shutter, two reciprocal source-side pathways reconstructed through correlations, or two branches of a phase-space or momentum-space distribution already present before far-field transport. In all of these formulations, the observable fringes are traced to coherent phase accumulation between two alternatives and to the measurement of a particular marginal or correlation function, not to a unique commitment to a distant real-space screen (Vedral, 24 Apr 2026, Bauer, 2013, Barbier et al., 2021, Yang et al., 26 Jun 2026).

1. Scope of the concept

The most conservative formulation remains the standard spatial double slit, in which two spatially separated paths interfere at a downstream observation plane. Several works extend this structure without abandoning the underlying two-alternative logic. One line treats temporal double slits as a free-particle initial-value problem, where a single wave packet is prepared with two separated peaks and then propagated by the time-dependent Schrödinger equation. Another line treats absorbing screens with openings that vary in both space and time, so that the aperture itself is a spacetime object rather than a static spatial mask. A third line replaces direct first-order detection by second-order reconstruction, so that the fringe pattern is recovered from joint statistics in frequency-time or source-position space rather than from direct intensity on a remote screen. These formulations suggest that “double slit” is less a fixed geometry than a coherent two-alternative architecture embedded in spacetime (Bauer, 2013, Goussev, 2012, O-oka et al., 2018, Wen, 2024).

A recurrent distinction in this literature is between different notions of temporal interference. In Moshinsky-type diffraction in time, the apparatus is explicitly time dependent, as in shutters or temporal gratings. By contrast, the “swapping space for time” program studies a time-independent Hamiltonian in which one and the same object reaches a single interaction zone at different times, so the interfering alternatives are temporally separated visits to one spatial region. The literature therefore does not treat all temporal analogues as equivalent, even when their observable formulas are Young-like (Czachor, 2018, Barbier et al., 2021).

Another recurrent theme is that the measured variable need not be transverse position. Interference appears in transverse momentum, in frequency, in arrival time, in source coordinate, and in phase-space quasidistributions. This suggests that the standard screen image is only one realization of a more general interference structure whose manifestation depends on which marginal of the state is projected or reconstructed (Yang et al., 26 Jun 2026, Rafsanjani et al., 2023).

2. Local Heisenberg-picture formulation

A particularly explicit spacetime reading is given in the Heisenberg-picture account of the non-relativistic double slit. There the experiment is formulated as two consecutive local measurements separated by unitary evolution, with the slit plane represented by a projector $P_1$ and the detection screen by $P_2$. The joint probability is written as

$$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$

and, for a pure initial state and one-dimensional projectors, factorizes as $p(1,2)=p(1)p(2/1)$. The interference pattern is therefore carried by the conditional probability $p(2/1)$, which can be interpreted as post-selection on particles that pass the slit plane (Vedral, 24 Apr 2026).

With the slits at $x=\pm d/2$ and the detector at $x=s$, the post-slit state is taken to be

$$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$

while the detection state is

$|\psi_2\rangle = |x=s, t_2\rangle.$

The conditional probability becomes

$p(2/1)=|\langle \psi_1|\psi_2\rangle|^2$

and is built from the sum of the two propagators

$P_2$0

From this, the fringe term

$P_2$1

is obtained. In this formulation the fringes are not attributed to a non-local wave “acting across space,” but to the overlap of local Heisenberg-picture position eigenstates at different spacetime points (Vedral, 24 Apr 2026).

The same work makes locality explicit by treating position and momentum as observables depending on both space and time: $P_2$2 with canonical commutator

$P_2$3

For a free particle,

$P_2$4

so the commutator vanishes for $P_2$5. That vanishing is interpreted as the hallmark of locality in the free theory: distinct spatial points are not instantaneously coupled. Within the same framework, the screen-particle interaction may be embedded into the “Church of the Larger Hilbert Space” through the local Hamiltonian

$P_2$6

with $P_2$7 and otherwise nonzero, so that projection emerges from conditioning on the quantized screen state rather than from a primitive non-local collapse (Vedral, 24 Apr 2026).

3. Temporal slits, shutters, and diffraction in time

The temporal analogue of the double slit appears in several mathematically distinct forms. In the free-particle TDSE treatment of space and time double-slit experiments, the spatial case is described by the initial condition

$P_2$8

which yields the standard transverse-momentum interference

$P_2$9

The temporal case is modeled by

$$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$0

with $$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$1. The momentum density then becomes

$$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$2

so the interference now appears in energy or longitudinal momentum space, with consecutive maxima separated approximately by $$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$3. The same treatment emphasizes complementarity between “which-path” and “which-time” marking by introducing unequal weights for the two alternatives (Bauer, 2013).

A more general propagator framework is provided by the Huygens-Fresnel-Kirchhoff construction for matter waves on absorbing screens with spacetime-dependent transmission $$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$4. Its central formula is

$$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$5

This unifies diffraction in time, diffraction in space, and their interplay. Rectangular and elliptical openings in the $$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$6 plane are treated as genuine space-time apertures, suggesting an analogy between $$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$7-dimensional propagation through time-dependent openings and $$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$8-dimensional diffraction through stationary openings (Goussev, 2012).

Phase-space treatments sharpen this temporal analogy. For diffraction in time generated by a succession of Lorentzian-like temporal slits, the Husimi amplitude is analytic in the frozen-Gaussian regime, and a double temporal slit yields

$$p(1,2)=Tr \{( U^\dagger_1P_1U_1) (U_2U_1)^\dagger P_2 (U_2U_1) ( U^\dagger_1P_1U_1) \rho \},$$9

For symmetric openings about the classical hitting time, the fringe loci satisfy

$p(1,2)=p(1)p(2/1)$0

with even $p(1,2)=p(1)p(2/1)$1 bright and odd $p(1,2)=p(1)p(2/1)$2 dark. Here the temporal double slit is not merely a modulation of density versus time; it is a correlated position-velocity interference structure in phase space (Barbier et al., 2021).

A separate temporal program replaces externally opened temporal slits by time-independent dynamics. In the “swapping space for time” construction, a wave packet reaches one localized interaction zone at different times under a stationary Hamiltonian. The emitted-photon probability takes Young-like forms such as

$p(1,2)=p(1)p(2/1)$3

or, in a simplified case,

$p(1,2)=p(1)p(2/1)$4

This is explicitly distinguished from Moshinsky diffraction in time: the interfering alternatives are different arrival times at one spatial “slit,” not a time-dependent apparatus (Czachor, 2018).

4. Reciprocal, ghost, and time-reversed realizations

A major extension of the topic arises when the fringe pattern is reconstructed from correlations rather than from first-order intensity on a detector plane. In reciprocal two-slit diffraction, the source and detector roles are exchanged relative to standard Young geometry. The source-side intensity is

$p(1,2)=p(1)p(2/1)$5

showing that fringes appear as the source is moved, provided the detector is effectively pinhole-like. This reciprocity is then transferred to the frequency-time domain, where a narrow-band chaotic light source and a time-gated detector emulate a reciprocal two-slit configuration (O-oka et al., 2018).

The resulting temporal ghost diffraction is read out through the second-order correlation

$p(1,2)=p(1)p(2/1)$6

With finite spectral bandwidth,

$p(1,2)=p(1)p(2/1)$7

so the visibility obeys

$p(1,2)=p(1)p(2/1)$8

The fringes are recovered by cross-correlation as a function of $p(1,2)=p(1)p(2/1)$9, not by direct first-order modulation in the test arm, and the work explicitly states that legitimate two-photon interference develops even with classical light (O-oka et al., 2018).

A related but distinct construction is the time-reversed Young experiment. There the point source is replaced by a fixed detector $p(2/1)$0, the observation screen is replaced by a laterally extended source plane $p(2/1)$1, and the double slit remains between them. With paraxial approximations, the detected pattern indexed by source coordinate is

$p(2/1)$2

with fringe period

$p(2/1)$3

The defining claim is that this second-order pattern is diffraction-free: there is no $p(2/1)$4 envelope. The pattern plane and source plane coincide on the same side of the slit plane, and the geometry permits programmed and digitized fringes by selective activation or motion of emitters. The same work stresses that the effect does not require entangled photons or any special nonclassical source; what is required is a one-to-one correspondence between each detected event and the emitter position (Wen, 2024).

Taken together, reciprocal temporal ghost diffraction and time-reversed Young schemes show that space-time double-slit diffraction need not be synonymous with first-order interference on a distant screen. A plausible implication is that the essential structure can survive the removal of the conventional “pattern plane,” provided the experiment supplies an alternative correlation coordinate in which the two-path phase is recovered.

5. Phase-space, momentum-space, and action-phase viewpoints

Several treatments relocate Young interference from real space into phase space or reciprocal space. In the hard X-ray momentum-space experiment, the field immediately behind a coherent double slit is described by a Wigner distribution

$p(2/1)$5

whose cross term lies in the empty gap between the slits and oscillates as $p(2/1)$6. The momentum marginal is

$p(2/1)$7

Because free-space propagation shears the Wigner function as

$p(2/1)$8

the momentum marginal is invariant: $p(2/1)$9 The X-ray experiment therefore interprets the usual far-field Young pattern as a real-space manifestation of an interference structure already present in momentum space immediately downstream of the aperture (Yang et al., 26 Jun 2026).

The same work realizes this directly with a perfect-crystal Si(660) analyzer, yielding a reciprocal-space scan

$x=\pm d/2$0

and resolving the complete hard X-ray double-slit fringe structure without a propagation arm, focusing optics, or imaging detector. The measured rocking curve exhibited a cosine fringe period of $x=\pm d/2$1, corresponding to $x=\pm d/2$2, in agreement with the slit geometry (Yang et al., 26 Jun 2026).

For matter waves, path-integral analysis modifies the usual optical analogy. The phase associated with a single matter-wave path of length $x=\pm d/2$3 is

$x=\pm d/2$4

not $x=\pm d/2$5. Nonetheless, the phase difference between two nearby paths is approximately

$x=\pm d/2$6

because paths connecting the same spacetime endpoints generally have different velocities and must be treated as spacetime trajectories rather than identical-speed rays. The path integral is therefore the correct explanation for why the familiar Young fringe condition is usually recovered for matter waves and also why it can fail in near-field or strongly asymmetric geometries (Jones et al., 2015).

These reciprocal-space and action-phase viewpoints converge on a shared conclusion: the interference is not created by the screen. It is encoded either in an invariant marginal such as $x=\pm d/2$7 or in the stationary-action phase relation between spacetime paths, and the experimental geometry determines only how that structure is projected into the measured variable.

6. Experimental realizations, control parameters, and interpretive boundaries

Controlled electron diffraction provides a baseline realization in which slit access is reversibly tuned. In the 600 eV electron experiment, a movable mask placed $x=\pm d/2$8 behind a nanofabricated double slit allows measurement of $x=\pm d/2$9, $x=s$0, and $x=s$1, and the pattern is built from single-electron detections at about $x=s$2 Hz. Intermediate accumulation images were shown for $x=s$3, $x=s$4, $x=s$5, $x=s$6, and $x=s$7 electrons, with the final buildup taking about $x=s$8 hours. The reported average distance between consecutive electrons was $x=s$9 meters, supporting the interpretation that only one electron is present in the $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$0 m system at any given time (Bach et al., 2012).

Time-resolved single-photon interference extends the same event-by-event logic into explicitly resolved spacetime detection. In the birefringent double-slit apparatus based on a calcite beam displacer, the two “slits” are orthogonally polarized beams displaced by $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$1 mm. A $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$2 SPAD array with $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$3 pitch records both detector position and arrival time, with timing uncertainty about $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$4 ps. The interference pattern emerges after the first $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$5 detections and is very clear after $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$6 photons, with visibility $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$7. In the entangled-photon extension, heralding on different idler projections prepares complementary signal-photon fringe patterns shifted in phase (Kolenderski et al., 2013).

The role of controllable which-path registration is explored in atom-cavity schemes. For a 3-level atom crossing a double slit followed immediately by a double cavity, the interaction

$$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$8

allows path information to be recorded in the phase of a quantum field and read out by quadrature measurement. The paper emphasizes that $$|\psi_1\rangle = |x=-d/2, t_1\rangle + |x=+d/2, t_1\rangle,$$9-quadrature can reveal path information, $|\psi_2\rangle = |x=s, t_2\rangle.$0-quadrature acts as a quantum eraser, and the classical field amplitude $|\psi_2\rangle = |x=s, t_2\rangle.$1 acts like a focusing element that makes the same interference or diffraction features appear at shorter propagation times. The visibility-distinguishability-entanglement balance is summarized by

$|\psi_2\rangle = |x=s, t_2\rangle.$2

for the path-internal-state correlations considered there (Rojas et al., 2020).

Space-time structure also appears when the propagation coordinate itself becomes an effective time variable. For diffraction of particles in free fall under the linear potential $|\psi_2\rangle = |x=s, t_2\rangle.$3, the longitudinal coordinate is mapped to the quasi-time

$|\psi_2\rangle = |x=s, t_2\rangle.$4

which converts the transverse diffraction problem into an effective free-particle Schrödinger evolution in $|\psi_2\rangle = |x=s, t_2\rangle.$5. For a double slit, the wave field is a sum of four Fresnel contributions, and the paper emphasizes that classical fall depends on $|\psi_2\rangle = |x=s, t_2\rangle.$6 whereas the fringe structure depends on $|\psi_2\rangle = |x=s, t_2\rangle.$7. This is used to discuss precision tests of the equivalence principle with ultracold neutrons, Bose-Einstein condensates, and molecular beams (Condado et al., 2018).

Entangled-particle generalizations shift the fringes into arrival-time space. In the Bohmian treatment of a double-double-slit experiment with entangled atoms in free fall, the joint spacetime detection law is

$|\psi_2\rangle = |x=s, t_2\rangle.$8

with arrival-time marginal $|\psi_2\rangle = |x=s, t_2\rangle.$9. The numerical analysis finds a complementarity relation between one-particle and two-particle temporal visibilities,

$p(2/1)=|\langle \psi_1|\psi_2\rangle|^2$0

and identifies the nonlocal effect with the joint distribution rather than with any signaling in the marginals (Rafsanjani et al., 2023).

Across these realizations, several misconceptions are treated explicitly. One is that double-slit fringes necessarily imply non-local action at a distance; the Heisenberg-picture analysis rejects that inference for the non-relativistic free-particle case (Vedral, 24 Apr 2026). Another is that temporal interference necessarily requires entanglement or nonclassical light; reciprocal temporal ghost diffraction and time-reversed Young experiments deny that necessity within their respective second-order measurement schemes (O-oka et al., 2018, Wen, 2024). A third is that far-field propagation creates the interference; direct momentum-space X-ray measurements instead treat propagation as a mapping of an already existing reciprocal-space fringe structure (Yang et al., 26 Jun 2026). Collectively, these results suggest that space-time double-slit diffraction is best understood as a family of measurement-dependent projections of one underlying structure: coherent superposition of two alternatives embedded in spacetime.