Femtosecond Needle Beam Arrays

• Femtosecond needle beam arrays are ensembles of structured, quasi-nondiffracting ultrafast optical wavepackets engineered via pulsed Bessel-like beams.
• They achieve exceptional spatiotemporal localization and extended depth-of-focus through techniques such as spatial light modulation, axicon arrays, and tailored phase apodizations.
• These arrays enable advanced applications including parallel microfabrication, high-speed optical switching, and orbital angular momentum emission using robust interference and Talbot self-imaging.

Femtosecond needle beam arrays comprise spatially structured ensembles of quasi-nondiffracting ultrafast optical wavepackets, typically engineered through superpositions of pulsed Bessel-like beams. These arrays exhibit extraordinary depth-of-focus, spatio-temporal localization, interference-induced self-imaging (Talbot effect), and spectral-temporal homogeneity. Recent advances enable heterogeneous functionalities, including parallel microfabrication, high-speed beam switching, and orbital angular momentum emission. Their precise generation and control arise from spatial light modulators, axicon arrays, and tailored phase apodizations, as detailed in the analytical formalism for space-time wavepacket construction (Grunwald et al., 3 Dec 2025).

1. Analytical Model of Pulsed Bessel-Like Needle Beams

Quasi-nondiffracting needle pulses are constructed by integrating Bessel-beam contributions across all frequencies near a carrier $\omega_0$. In cylindrical coordinates $(r,\phi,z)$, the scalar field is:

$$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$

where $\Omega = \omega - \omega_0$ (detuning), $S(\Omega)$ is the pulse spectrum (often Gaussian with bandwidth $\Delta\Omega = 2\pi/\tau_0$ for a transform-limited $20$ fs pulse), and $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$, $k_z(\Omega) = (\omega_0+\Omega)/c \cos\theta$.

All spectral components share the same longitudinal phase slope $k_z$ vs. $(r,\phi,z)$0, ensuring the wavepacket resists diffraction both spatially and temporally.

A differential decomposition treats the needle pulse as a bundle of infinitesimal Bessel beams, each subtending cone angle $(r,\phi,z)$1:

$(r,\phi,z)$2

where $(r,\phi,z)$3, and $(r,\phi,z)$4 sets radial apodization and truncation—determined by an axicon or spatial filter.

In $(r,\phi,z)$5-space, the spectral amplitude is restricted to a two-dimensional manifold:

$(r,\phi,z)$6

establishing tight correlation between spatial and temporal frequencies and underpinning nondiffracting, nondispersing propagation.

2. Generation Methodologies and Array Design

Femtosecond needle-beam arrays are produced via optical systems combining collimated Gaussian femtosecond lasers (central wavelength $(r,\phi,z)$7 nm, pulse duration $(r,\phi,z)$8 fs, beam diameter $(r,\phi,z)$9) with axicons, spatial light modulators (SLMs), and apodization filters.

For single needle beams, an axicon of angle $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$0 (thin-film, refractive index $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$1, or SLM-encoded) transforms the beam into a conical wave, generating angle $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$2 and producing a $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$3-space ring of radius $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$4. Typical parameters ($$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$5, $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$6 mm) yield a depth of focus $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$7 mm.

For arrays, the axicon is replaced by an axicon-array (e.g., $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$8 elements) or a hexagonal phase mask on an SLM. Representative experimental configuration (Grunwald & Bock): $$E(r,z,t) = \int_{-\infty}^{\infty} d\Omega\, S(\Omega) J_0\left[k_\perp(\Omega) r\right] \exp\left[i k_z(\Omega) z - i (\omega_0 + \Omega)t\right],$$9, $\Omega = \omega - \omega_0$0 mm, array period $\Omega = \omega - \omega_0$1m, resulting in $\Omega = \omega - \omega_0$2 mm.

Spectral shaping is achieved via angular dispersion (grating-lens pairs) to fine-tune cone angles per frequency, compensating residual group-velocity dispersion. SLMs encode ring-shaped phase (axicon function), higher-order apodization, and arbitrary $\Omega = \omega - \omega_0$3 profiles.

Parameter Summary for Array Generation

Variable	Value (Single/Array)	Context
$\Omega = \omega - \omega_0$4	800 nm	Central wavelength
$\Omega = \omega - \omega_0$5	20 fs	Pulse duration
$\Omega = \omega - \omega_0$6	$\Omega = \omega - \omega_0$750 nm	Bandwidth
$\Omega = \omega - \omega_0$8	0.1$\Omega = \omega - \omega_0$9 / 0.24$S(\Omega)$0	Axicon angle
$S(\Omega)$1	1.45	Refractive index
$S(\Omega)$2	0.5 mm / 5 mm	Input diameter
$S(\Omega)$3	400 $S(\Omega)$4m	Array period
$S(\Omega)$5	$S(\Omega)$6	Elements (array)

3. Interference and Talbot Self-Imaging in Space-Time

Launching $S(\Omega)$7 needle beams in a periodic lattice of spacing $S(\Omega)$8 yields a total field at $S(\Omega)$9:

$\Delta\Omega = 2\pi/\tau_0$0

Beyond the focal region ($\Delta\Omega = 2\pi/\tau_0$1), the discrete spatial spectrum forms a periodic diffraction lattice. Owing to each beam's quasi-nondiffracting temporal character, mutual interference reconstructs the array at regular longitudinal intervals without profile distortion.

The principal Talbot distance (first in-phase revival) is:

$\Delta\Omega = 2\pi/\tau_0$2

with half-Talbot (out-of-phase) at $\Delta\Omega = 2\pi/\tau_0$3. For $\Delta\Omega = 2\pi/\tau_0$4m, $\Delta\Omega = 2\pi/\tau_0$5 nm:

$\Delta\Omega = 2\pi/\tau_0$6

Numerical experiment (hexagonal array, Grunwald & Bock):

• Focal plane $\Delta\Omega = 2\pi/\tau_0$7 mm
• Out-of-phase image $\Delta\Omega = 2\pi/\tau_0$8 mm $\Delta\Omega = 2\pi/\tau_0$9
• First full revival $20$0 mm $20$1

The instantaneous intensity,

$20$2

remains invariant at each Talbot plane, indicating robust space-time self-imaging.

4. Spectral, Temporal, and Geometric Performance

Simulations corroborate superior spectral and temporal uniformity for needle-beam arrays relative to conventional focused Gaussian pulses:

• Spectral homogeneity: Radial shift $20$3 nm over $20$4m for Bessel-Gauss needle pulses, compared to larger shifts in spatio-spectrally shaped Gaussian pulses.
• Temporal robustness: Within the nondiffracting zone $20$5 (up to $20$6 m for $20$7), pulse duration remains $20$8 fs with negligible broadening.
• Depth of focus: $20$9 m (single beam, $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$0, $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$1 mm) and up to $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$2 m (array, $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$3, $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$4 mm).
• Self-imaging fidelity: 2D Talbot carpets map periodic reappearance of hexagonal intensity at Talbot planes ($k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$5, $k_\perp(\Omega) = (\omega_0+\Omega)/c \sin\theta$6) with preserved femtosecond pulse envelopes.

5. Functionalities, Applications, and Extensions

Femtosecond needle-beam arrays facilitate diverse applications and technical enhancements:

• Parallel materials processing: Arrays enable simultaneous, multipoint femtosecond microfabrication, achieving sub-diffraction-limited spots and uniform pulse delivery throughout extended depths.
• High-speed optical switching and tracking: SLM phase masks and ultrafast modulators (or MEMS mirrors) allow dynamic reconfiguration of beam arrays on microsecond and faster timescales.
• Orbital Angular Momentum (OAM): Encoding radial chirps of spiral phases onto axicon masks yields self-torqued OAM arrays, producing propagation-dependent angular velocities in femtosecond pulses.
• Space-time Talbot metrology: Spatial and temporal self-imaging supports precision measurement of group-velocity dispersion and calibration of array periods.
• Prospective directions: Embedding arrays in high-power resonators to produce nondiffracting intracavity modes; extending architectures for attosecond photonic nanojets; and engineering vectorial or accelerating wavepackets within fiber or metamaterial platforms.

A plausible implication is the further maturation of ultrafast optical manipulation through integration of these methods with advanced photonic control devices.

6. Summary and Outlook

Femtosecond needle-beam arrays integrate the nondiffracting properties and spatial-temporal control of Bessel-like wavepackets, enabled by robust analytical frameworks (Eqns. 1–2) and practical generation via axicons and SLMs. Their capacity for coherent Talbot self-imaging, spectral-temporal uniformity, and flexible array architecture catalyzes new methodologies in parallel photonic processing, high-speed switching, OAM metrology, and ultrafast measurement. Future research targets include miniaturization, cavity embedding, and space-time engineering in novel media (Grunwald et al., 3 Dec 2025).