Folded Time Grating Overview

• The paper demonstrates that a folded time grating is a compact waveguide-based system that uses a single time-modulated strip and its image currents to replicate an infinite periodic temporal aperture.
• It employs a Floquet–Bloch formulation to couple temporal harmonics, unveiling phenomena like temporal Wood’s anomaly and negative-frequency parametric gain.
• The architecture simplifies modulation networks and reduces complexity, offering practical advantages for dynamic filtering, leaky-wave antennas, and pulse fan-out applications.

A folded time grating is a compact waveguide realization of a temporal grating in which a single time-modulated loaded strip, together with the image currents imposed by metallic waveguide walls, synthesizes the response of an infinite periodic time-modulated interface. In the formulation introduced for the experimental observation of temporal Wood’s anomaly, the folded geometry reduces a distributed temporal aperture to one waveguide-enclosed element while preserving the far-field/surface-wave coupling physics of a planar temporal grating; it also reveals a regime in which coupling to negative-frequency surface-wave branches produces tunable parametric amplification (Shaham et al., 7 Sep 2025). The term is narrower than the broader literature on time gratings, temporal dispersion gratings, and space-time gratings, where closely related mechanisms include frequency compensation, Floquet sideband generation, dispersive fan-out, and traveling or standing spatiotemporal modulation, but not necessarily a folded implementation (Yu et al., 2 Apr 2026, Zhu et al., 2023, Sheveleva et al., 2020).

1. Terminology and scope

In the published literature, “folded time grating” is a specific term rather than a generic synonym for any temporally periodic photonic structure. The direct definition appears in the experimental work on temporal Wood’s anomaly, where the phrase denotes a waveguide-based synthesis of an extended time-modulated grating by means of a single time-modulated element and its images (Shaham et al., 7 Sep 2025). Other papers study adjacent objects—such as a temporal dispersion grating coupler, a homogeneous waveguide with refractive index periodic in time, or a time grating for temporal Smith–Purcell radiation—but explicitly do not use the same terminology (Sheveleva et al., 2020, Yu et al., 2 Apr 2026, Zhu et al., 2023).

The main terminological distinction is between time periodicity and folding. A time grating may be realized by periodic modulation of a refractive index, a phase, a boundary impedance, or a coupling coefficient. A folded time grating adds a structural reduction: an effectively extended temporal aperture is reproduced in a compact geometry. In the direct experimental realization, the infinite free-space grating is “folded into a waveguide cross-section” by image theory (Shaham et al., 7 Sep 2025). In several related papers, “folded” is only an interpretation—for example, when a standing space-time modulation is decomposed into opposite traveling gratings, or when temporal Floquet replicas are viewed as frequency-space folding—rather than the authors’ formal label (Horsley et al., 2024, Yu et al., 2 Apr 2026).

Concept	Defining structure	Relation to folded time grating
Folded time grating	Single time-modulated loaded strip in a rectangular metallic waveguide	Direct term and direct implementation (Shaham et al., 7 Sep 2025)
Temporal dispersion grating coupler	Periodic temporal phase profile followed by dispersive propagation	Related temporal-grating mechanism, not folded (Sheveleva et al., 2020)
Time-grating GMR platform	Homogeneous waveguide with $\varepsilon_r(t)$ periodic in time	Related Floquet frequency-shift mechanism, not folded (Yu et al., 2 Apr 2026)
Time grating for t-SPR	Planar interface with refractive index periodic in time	Related free-electron radiation mechanism, not folded (Zhu et al., 2023)
Stationary grating with oscillating amplitude	$2\alpha\cos(gx)\cos(\Omega t)$ standing space-time grating	Conceptually close to a folded traveling grating, but not named as such (Horsley et al., 2024)

2. Folded architecture and unfolded equivalent

The canonical folded time grating consists of a single time-modulated loaded strip placed inside a rectangular metallic waveguide. The strip is thin, $y$-oriented, centered at $x=a/2$, $z=0$, and loaded by a time-varying capacitance $C(t)$. The enclosing waveguide has PEC walls, width $a$, height $b$, and supports a $y$-invariant TE configuration. The folding mechanism is electromagnetic rather than merely geometric: repeated image currents across the PEC walls reproduce the field of an infinite free-space metagrating plane, so one modulated element emulates an unfolded periodic interface (Shaham et al., 7 Sep 2025).

The temporal modulation is written as

$$C(t)=\sum_{l=-\infty}^{\infty}C_l e^{jl\Omega t}, \qquad C_{-l}=C_l^*.$$

For sinusoidal modulation,

$2\alpha\cos(gx)\cos(\Omega t)$0

with $2\alpha\cos(gx)\cos(\Omega t)$1. The incident field is the waveguide $2\alpha\cos(gx)\cos(\Omega t)$2 mode,

$2\alpha\cos(gx)\cos(\Omega t)$3

Time periodicity generates Floquet current harmonics

$2\alpha\cos(gx)\cos(\Omega t)$4

This is the basic temporal-grating action: the incident mode at $2\alpha\cos(gx)\cos(\Omega t)$5 is coupled to harmonics $2\alpha\cos(gx)\cos(\Omega t)$6 through the single modulated load (Shaham et al., 7 Sep 2025).

The experimental implementation uses a reduced-height WR-340 aluminum waveguide with $2\alpha\cos(gx)\cos(\Omega t)$7 mm and $2\alpha\cos(gx)\cos(\Omega t)$8 mm, plus a PCB carrying two narrow printed copper strips of width $2\alpha\cos(gx)\cos(\Omega t)$9 mm. The time-varying capacitance is realized by a balanced bridge of four varactor diodes. A practical advantage of the balanced bridge is modal separation: the DC bias and RF modulation are applied through the differential mode, whereas the coupling to the waveguide field occurs through the common mode. The paper identifies this separation as the reason the structure supports transmissive operation without a ground plane and with negligible parasitic radiation from the modulation network (Shaham et al., 7 Sep 2025).

3. Floquet–Bloch formulation and static surface-wave resonance

In the waveguide formulation, the scattered field is expanded in temporal harmonics $y$0 and odd transverse modes $y$1, because the strip is centered. The scattered field contains factors

$y$2

so each $y$3 channel is either propagating or evanescent depending on whether $y$4 lies above or below the cutoff

$y$5

The reflected coefficient into the $y$6 space-time harmonic is

$y$7

for odd $y$8, and the transmission coefficient is

$y$9

The experimentally measured transmission is essentially $x=a/2$0, i.e. $x=a/2$1-to-$x=a/2$2 conversion across temporal harmonics (Shaham et al., 7 Sep 2025).

Before time modulation is applied, $x=a/2$3, only $x=a/2$4 exists, and the current reduces to

$x=a/2$5

The denominator

$x=a/2$6

contains poles corresponding to eigenmodes of the loaded interface. One of these is a surface-wave resonance at a real frequency below cutoff,

$x=a/2$7

A second feature appears above cutoff near a short-circuit-like resonance $x=a/2$8, but the paper distinguishes it from the true surface-wave pole because radiation contributes an imaginary part to the inductive term (Shaham et al., 7 Sep 2025).

Experimentally, at 6 V reverse bias, the extracted static parameters are approximately

$x=a/2$9

with

$z=0$0

The static transmission spectrum shows a resonant peak below cutoff at $z=0$1 and nearly full transmission above cutoff except for a strong dip at $z=0$2. The folded time grating therefore begins as a static surface-wave platform, to which time modulation adds frequency transitions (Shaham et al., 7 Sep 2025).

4. Temporal Wood’s anomaly and negative-frequency gain

Temporal Wood’s anomaly arises when a frequency-converted harmonic is tuned to the static surface-wave resonance. In the weakly modulated, sinusoidal case, the dominant channels are $z=0$3. The fundamental current can be written in terms of primitive frequency-conversion coefficients $z=0$4, where

$z=0$5

The ordinary anomaly is the regime

$z=0$6

so the downconverted harmonic resonates with the surface wave and feeds back into the transmitted fundamental. In this regime, the folded device reproduces the classical Wood-anomaly interplay between a far-field channel and a surface-wave resonance, but by frequency transitions rather than momentum transitions (Shaham et al., 7 Sep 2025).

The first experimental observation used 6 V reverse bias, modulation frequency

$z=0$7

and modulation depth

$z=0$8

Sweeping the input from 2.2 to 3.5 GHz produced an anomaly at the Wood frequency

$z=0$9

for which

$C(t)$0

The measured signatures were a new transmissive peak in the $C(t)$1 channel and a sharp $C(t)$2 peak at the surface-wave resonance, while the $C(t)$3 harmonic remained much weaker. The paper identifies this as the first direct observation of temporal Wood’s anomaly (Shaham et al., 7 Sep 2025).

The more distinctive regime is the negative-frequency surface-wave branch, defined by

$C(t)$4

Here the feedback condition may be reduced to the pole equation

$C(t)$5

The authors interpret this as a root-locus condition: $C(t)$6 is the open-loop gain, $C(t)$7 is the feedback term, and the pole can be driven toward the real axis. Because the downconverted harmonic lies near a negative-frequency surface-wave branch, the time modulation behaves as an effective negative resistance and permits parametric amplification, with parametric oscillation at threshold (Shaham et al., 7 Sep 2025).

Experimentally, this regime was accessed with

$C(t)$8

again at 6 V reverse bias, and with three modulation depths: $C(t)$9 The resonant feature appeared near the negative-Wood frequency

$a$0

for which

$a$1

As $a$2 increased, the fundamental transmission peak rose from

$a$3

to

$a$4

and then to

$a$5

the last value constituting explicit parametric amplification (Shaham et al., 7 Sep 2025).

5. Relation to broader time-grating literature

The folded time grating sits within a broader family of temporal and space-time grating concepts, but it is not coextensive with them. A useful comparator is the temporal dispersion grating coupler, where a periodic temporal phase modulation followed by second-order dispersion maps discrete spectral lines into equally spaced delayed pulse replicas. In that setting, a periodic phase-only mask $a$6 and dispersive propagation realize the time-domain analogue of a diffractive coupler; the paper is explicit that it is not about a “folded time grating” in that exact terminology (Sheveleva et al., 2020). The closest common element is the interpretation of a temporal grating as a programmable periodic structure whose action appears only after an auxiliary propagation stage.

A second comparator is the homogeneous waveguide time grating used for guided-mode resonance. There the constitutive law is

$a$7

and the coupling rule is

$a$8

This replaces the momentum compensation of a spatial grating with frequency compensation and produces a temporal analogue of GMR. The paper argues that the resulting Floquet sidebands may be viewed as a kind of frequency-space folding, but it does not define a folded time grating; rather, it develops a tunable time-grating platform with $a$9-factors that diverge as $b$0, near-unity reflection in the fundamental harmonic, first-order harmonic reflection values up to $b$1, and Goos–Hänchen shifts exceeding $b$2 (Yu et al., 2 Apr 2026).

A third comparator is temporal Smith–Purcell radiation from a time grating. In that case a periodically time-modulated medium supplies energy compensation rather than the momentum compensation of a spatial grating, and the generalized dispersion law is

$b$3

This again involves temporal Floquet sidebands, but the geometry is a planar free-electron interface rather than a folded waveguide enclosure (Zhu et al., 2023). Likewise, a stationary grating whose amplitude oscillates in time can be decomposed into a pair of opposite traveling gratings,

$b$4

which the authors describe as conceptually close to a folded traveling/time grating, though not under that name (Horsley et al., 2024).

These comparisons delimit a common misconception. Not every time grating is folded, and not every folded interpretation refers to the same architecture. In the strict sense established experimentally, a folded time grating is the waveguide-enclosed, image-theory synthesis of a periodic temporal aperture (Shaham et al., 7 Sep 2025). Other works provide the surrounding theoretical vocabulary—Floquet harmonics, frequency compensation, dispersive fan-out, traveling-modulation equivalence, or space-time branch coupling—but not the same object (Sheveleva et al., 2020, Yu et al., 2 Apr 2026, Zhu et al., 2023, Horsley et al., 2024).

6. Significance, applications, and limitations

The folded architecture matters because it converts a distributed temporal grating into a single modulated element. The direct practical consequences stated for the waveguide implementation are drastically reduced complexity, lower power consumption, a simpler modulation feed network, less concern about modulation-network / EM interference, transmissive operation, and a compact implementation (Shaham et al., 7 Sep 2025). The same paper frames the device as an economical path for universally synthesizing intricate temporal apertures and explicitly points to dynamic filtering and leaky-wave antennas as target application classes.

From a methodological standpoint, the folded time grating is also significant because it makes a previously abstract temporal-grating phenomenon experimentally tractable. The theory combines Floquet current harmonics, waveguide cutoff physics, and a surface-wave pole of a static loaded strip, and the experiment shows excellent agreement with that Floquet–Bloch analysis (Shaham et al., 7 Sep 2025). A broader implication, already suggested by the surrounding literature, is that folded implementations may serve as compact surrogates for more extended temporal or space-time structures: temporally periodic waveguides for resonance engineering (Yu et al., 2 Apr 2026), programmable temporal gratings for pulse fan-out (Sheveleva et al., 2020), and time-grating platforms for free-electron radiation (Zhu et al., 2023).

The limitations are equally specific. The direct folded-time-grating analysis assumes a thin-wire approximation, a $b$5-invariant TE$b$6 waveguide formulation, and usually truncates the modulation to dominant harmonics $b$7. The substrate is not modeled explicitly in the analytical theory, but is approximately incorporated through an effective radius correction. The modulation depth is constrained by positivity of the capacitance,

$b$8

To match experiment, the model includes effective series loss $b$9. The paper also notes minor discrepancies near the cutoff of the $y$0 harmonic, plausibly linked to substrate-related effects not captured by the simplest single-strip model, and it identifies practical ceilings from varactor saturation, breakdown, finite modulation power, and unmodeled nonlinearities (Shaham et al., 7 Sep 2025).

In this sense, the folded time grating is best understood not as a universal label for temporal periodicity, but as a particular synthesis principle: a compact geometry that reproduces the response of an extended temporal grating while retaining the central physics of frequency-converted coupling, surface-wave resonance, and, uniquely, negative-frequency parametric gain (Shaham et al., 7 Sep 2025).