Passive Cross-Basis Mode Transitions Along a Single Freely Propagating Bessel Beam
Henry P. Evans, Layton A. Hall ∗ · 2026
Read Passive Cross-Basis Mode Transitions Along a Single Freely Propagating Bessel Beam by Henry P. Evans, Layton A. Hall ∗ (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-17200.txt.
Introduction
Evans and Hall show that a single freely propagating Bessel beam can switch transverse modes on its own as it travels. The beam's conical spectrum maps radial position to reconstruction distance, so dividing one static phase-only modulator into rings encodes different modes at chosen depths. They demonstrate ring-lattice fields and a passive sequence through Bessel, vortex, Hermite-Gaussian-Bessel and Airy caustic modes without moving parts.
Catalog record
| CREATOR | Henry P. Evans, Layton A. Hall ∗ |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 16535 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.17200
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Passive Cross-Basis Mode Transitions Along a Single Freely Propagating Bessel Beam Henry P. Evans, Layton A. Hall ∗ Abstract The transverse modal identity of a freely propagating optical beam is ordinarily fixed at the point of generation. We show that the conical angular spectrum of a Bessel beam establishes a one-to-one mapping between radial beam position and axial reconstruction distance. This mapping converts the radial aperture of a single static, phase-only spatial light modulator into a programmable longitudinal-mode register. By partitioning the modulator into independent annular regions, we encode discrete transverse modes at preselected axial positions. We demonstrate this principle with programmable ring-lattice fields of axially varying site number, and with passive transitions that sequence through Bessel, Bessel vortex beam, Hermite-Gaussian-Bessel, and Airy caustic modes within a single beam, without dynamic modulation or cascaded optical elements. Transverse spatial modes of optical beams, including Hermite–Gaussian (HG), Laguerre–Gaussian (LG), Airy, and Bessel families, constitute a rich set of orthogonal state spaces for encoding classical and quantum information, driving light–matter interactions, and structuring optical potentials 1 ; 12 ; 19 ; 2 . Each family is characterized by a distinct underlying spatial symmetry: Cartesian for HG, cylindrical for LG, caustic for Airy, and propagation-invariant for Bessel 10 ; 20 . These modal families are usually treated as distinct bases and are not normally interconverted by simple propagation of a single prepared mode. Transitions between modal families are possible with dedicated optical elements. Astigmatic mode converters continuously deform the Gaussian envelope to map HG onto LG modes 3 ; 1 . Multi-plane light conversion and cascaded diffractive or metasurface elements implement more general unitary transformations between transverse modes 6 , and dynamically reconfigurable SLMs can synthesize arbitrary wavefronts at selected planes 4 ; 15 . In each case, the modal identity at a given plane is set by an active element or by a cascade, not by a passive feature of free-space propagation itself. A complementary line of work tailors the longitudinal degree of freedom while holding the transverse symmetry class fixed. Frozen waves sculpt axial intensity by superposing co-axial Bessel beams with prescribed wavenumber spectra 23 ; 17 ; twisted frozen waves extend this to coherent axial transfer of OAM structure 14 ; and the rotatum of light reveals additional axial structure in the OAM current 8 . A recent perspective consolidates this rapidly growing area but explicitly scopes it to continuous parameters, including intensity, OAM, polarization, and wavelength within a fixed transverse-mode family 22 . Bessel–Gaussian superpositions with discrete longitudinal wavenumbers produce beams whose topological charge \ell steps during propagation 9 , metasurfaces realize longitudinal variation of polarization and OAM 7 , and different-cone-angle Bessel–Gauss superpositions generate optical-needle arrays with independent axial intensity and vortical structure 18 . Two limitations are common: modal evolution stays within one transverse symmetry class (frozen waves, twisted frozen waves, \ell -stepping, needle arrays) or, when crossing classes, needs polarization-sensitive subwavelength elements, cascaded optics, or dynamic modulation (mode converters, MPLC, metasurfaces, reconfigurable SLMs). Neither enables passive, deterministic transitions between incompatible transverse symmetry classes—Cartesian, cylindrical, caustic, and propagation-invariant—along a single beam. Here, we demonstrate that the conical angle of Bessel beams provides a natural physical mechanism for cross-modal changes with propagation. The one-to-one geometrical correspondence between radial beam position and axial reconstruction distance, which is inherent to axicon-generated Bessel fields 11 ; 16 , transforms the radial coordinate of a phase-only SLM into a programmable longitudinal modal register. The novelty of this work is therefore not the finite-aperture axicon mapping itself, which is well established 10 , but its use as a passive longitudinal modal register. That is, independent annular phase holograms on a single static, phase-only, polarization-insensitive SLM encode distinct transverse-mode families at preselected axial planes, and the symmetry class of the beam changes discretely with z . Partitioning the SLM aperture into independent annular regions encodes different transverse modes, creating passive free-space beams with evolving modal and topological states during propagation without dynamic modulation or cascaded elements. We demonstrate programmable ring-lattice fields with axial site variation and discrete cross-basis transitions, sequentially switching through Bessel, vortex, Hermite-Gaussian-Bessel, and Airy modes within one beam. These results establish the longitudinal propagation coordinate as an addressable modal degree of freedom with direct implications for volumetric optical manipulation, axially multiplexed communications 21 ; 5 , and three-dimensional photonic state preparation 13 . Figure 1: Principle and setup. (a) A phase-only axicon \Phi(x_{s},y_{s}) is partitioned into concentric annuli f_{m} ; the Bessel cone angle \theta maps each radius r_{m,s} to a unique reconstruction plane z_{m}=k\,r_{m,s}/k_{r} , so distinct transverse modes appear at z_{1},z_{2},z_{3} . (b) A 635 nm laser is expanded 4\times (L 1 : f_{1}=50 mm,L 2 : f_{2}=200 mm), reflected from the SLM, and relayed through a 4f system (L 3 : f_{3}=400 mm, L 4 : f_{4}=100 mm; 4\times demagnification) onto a translating CCD (C). Principle of Longitudinal Modal Encoding — An ideal Bessel beam is formed by the interference of plane waves distributed on a cone of fixed half-angle \theta [Fig. 1 (a)]. In a finite-aperture axicon geometry, the transverse wave vector k_{r}=k\sin\theta fixes the cone angle and k is the wavenumber, while the launch radius r_{s} determines the axial position at which that annular contribution reconstructs, z\simeq r_{s}k/k_{r} . This geometrical mapping between annular launch radius and axial reconstruction position is the physical foundation of the method presented here. For a conventional zero-order Bessel beam generated by a phase-only axicon, the SLM phase \Phi is \Phi(r_{s},\varphi)=k_{r}r_{s}, (1) where r_{s} is the radial coordinate on the SLM and k_{r}=k\sin\theta is the radial spatial frequency. To encode longitudinally programmable transverse structure, we partition the axicon phase into M independent concentric annular regions, each assigned a distinct phase modulation f_{m}(r_{s},\varphi) : \Phi(r_{s},\varphi)=k_{r}r_{s}+\sum_{m}f_{m}(r_{s},\varphi_{s})\,\Pi\!\left(\frac{r_{s}-r_{m,s}}{\Delta r}\right), (2) where \Pi(\cdot) is the rectangle function, r_{m,s} is the central radius of the m th annulus, and \Delta r is its radial width. Each annular region acts as an independent phase hologram whose encoded transverse structure is reconstructed only at its designated axial position. This can be represented in Fig. 1 (a) on the SLM pattern where the varying colors represent the different regions of f_{m} . Figure 2: Axially programmable ring-lattice fields. (a) Measured longitudinal intensity cross-section I(x,y=0,z) over 25 mm of free-space propagation; dashed lines mark the four axial observation planes. Transverse intensity profiles I(x,y) at (b) z=3 mm ( N=1 ), (c) z=8 mm ( N=6 ), (d) z=14 mm ( N=8 ), and (e) z=20 mm ( N=12 ), demonstrating that the number of ring-lattice sites is independently programmable at each axial depth via the annular SLM geometry. All panels share the normalized intensity scale shown in (a). The mapping between SLM radial coordinate and axial reconstruction distance follows from the finite-aperture axicon geometry. For a fixed transverse wave vector k_{r}=k\sin\theta , an annular zone centered at radius r_{m,s} reconstructs near the axial plane z_{m} determined, in the paraxial limit, by \frac{r_{m,s}}{z_{m}}\simeq\frac{k_{r}}{k}, (3) where k=2\pi/\lambda is the optical wavenumber and z_{m} is the axial plane at which the m th encoded mode is reconstructed. Thus, the transverse wave vector fixes the cone angle, while the annular radius sets the axial address. The axial extent \Delta z over which each encoded mode persists