Skip to main content
PNAS Nexus logoLink to PNAS Nexus
. 2026 Nov 6;5(11):pgag385. doi: 10.1093/pnasnexus/pgag385

Light-self-induced geometric phase structures in liquid crystals

Manuel Diaz-Zuniga 1,✉, Mauro Riquelme-Alcantara 2, Marcel G Clerc 3
Editor: Zakya Kafafi
PMCID: PMC13774875  PMID: 43047624

Abstract

The geometric phase, originating from cyclic and adiabatic evolutions in parameter space, is central to both quantum and classical systems. In optics, it enables polarization-based control of light and has been widely exploited in birefringent elements. Here, we show that, through its interaction with optical valves, light can self-induce a localized geometric phase structure. Molecular reorientation forced by light intensity imprints a confined geometric phase with zero orbital angular momentum on the transmitted beam, revealed through interference. An amplitude equation derived from first principles captures the observed phase fields, with analytical, numerical, and experimental results in close agreement. This establishes a universal framework for light-driven geometric-phase localization, bridging fundamental optics with reconfigurable photonic applications.

Keywords: Pancharatnam–Berry phase, liquid crystals, optical valves, out-of-equilibrium systems, self-organization


Significance Statement.

Geometric phases are a fundamental concept in modern physics, with applications ranging from quantum systems to advanced optical technologies. In photonics, they enable precise control and structuring of light for information transport, metrology, and high-precision devices. We show that light–matter interaction in a photosensitive liquid crystal device gives light a dual role: it induces a self-organized, localized molecular reorientation and, through reconfiguration of the medium, imprints geometric phase on itself, with structure shaped by optical intensity distribution. We develop a dynamical model that predicts the emergence of these states and implement an experimental method to estimate and measure the resulting geometric phases, while also demonstrating their controllable manipulation, providing a new route for self-induced geometric-phase engineering in nonlinear optical systems.

Introduction

Since the pioneering observation of geometric phase in optics by Pancharatnam in 1956 (1), and its subsequent generalization by Berry in 1984 (2), the Berry phase has become a fundamental concept across diverse areas of physics. It corresponds to the phase acquired by a vectorial system undergoing a cyclic and adiabatic evolution in its parameter space, explicitly reflecting the geometry of that space. This notion, often referred to as the geometric phase, has driven significant advances in quantum mechanical systems (3). Its applications extend broadly to both quantum and classical phenomena, ranging from the Aharonov–Bohm effect to the precession of the Foucault pendulum’s oscillation plane (4).

In optics, the geometric phase arises from the manipulation of polarization and/or the direction of propagation. When the polarization evolution traces a closed path on the Poincaré sphere representation (see Fig. 1), the final state of a light beam differs from the initial one by acquiring a geometric phase (see Ref. (5) and references therein). Traditionally, birefringent media such as liquid crystals have provided the primary platform for implementing geometric-phase optical elements, enabling a wide range of applications. These include polarization-sensitive diffraction gratings (6), spiral phase elements for laser radiation (7), helical mode generators such as q-plates (8, 9), reflective Bragg mirrors (10, 11), geometric phase arising from propagation in twisted optical waveguides (12), and higher-order geometric phases associated with spatially inhomogeneous (vector) polarization states (13). More recently, efforts to understand and exploit the geometric phase have opened new avenues for designing flat optical elements capable of highly efficient, diffraction-limited beam shaping in compact platforms (14). Metasurface-based implementations have further demonstrated switchable and reconfigurable geometric-phase control in the terahertz regime (15). Geometric-phase elements have also been employed to generate novel light structures, such as perfect vortex beams and vector beams (16), and to achieve efficient beam shaping, including converting Gaussian beams into line beams with high efficiency (17). Extensions into nonlinear optical regimes have been realized in ferroelectric nematic systems (18). Moreover, in inhomogeneously anisotropic media with a transversely rotated optic axis, the evolution of electromagnetic wave packets is governed by the geometric phase, which induces an effective photonic potential and enables these systems to function as geometric-phase-based waveguides (19–21). Recently, experiments on the interaction of light with liquid crystals have shown localized geometry phase (22, 23). Despite recent advances, the understanding of the geometric phase in optics and physics, including the structuring of light, its robustness, and its potential applications, remains open and constitutes an active field of research.

Figure 1.

Seven-panel figure illustrating light-induced geometric phase in a liquid crystal light valve. Panels (A) and (B) show schematic representations of light-induced molecular reorientation and polarization evolution on the Poincaré sphere. Panels (C) and (D) show numerical simulations of the complex amplitude field and corresponding geometric phase distribution. Panel (E) illustrates the experimental optical setup. Panels (F) and (G) show experimental observations obtained using crossed linear polarizers and interferometry, respectively.

Localized geometric phase induced by light in liquid crystal valves. A) Schematic representation of the induction of a localized inhomogeneous molecular orientation in a LCLV, where an incident circularly polarized beam undergoes helicity inversion, acquiring a geometric phase. B) Representation of the helicity inversion of the polarization on the Poincaré sphere, where the average molecular orientation ϕ determines the closed path and the associated geometric phase ΦB=Ω/2=2ϕ. Numerical solutions of Eq. 2: C) complex amplitude field A→=(Re{A},Im{A}) and D) corresponding geometric phase field ΦB=2Arg{A}. Experimental localized geometric phase. E) Schematic representation of the experimental setup. QWPs are quarter-wave plates for linear–circular polarization conversion, P and A are linear crossed polarizers, V0 and I0 are the driving voltage and the light intensity applied to the LCLV, respectively, and CMOS denotes a complementary metal-oxide-semiconductor camera. The setup includes a Mach–Zehnder interferometer with tilt angle β for circularly polarized light. Experimental observation of localized inhomogeneous states for V0=22.5 V rms (f=1 kHz) and I0=8.07 mW/cm 2. F) intensity distribution between crossed linear polarizers and G) corresponding interferogram obtained with a tilted reference beam.

Unlike metasurfaces and q-plates, which rely on prepatterned anisotropic structures to impose a prescribed geometric phase, the phenomenon reported here emerges spontaneously from light-driven molecular self-organization. This distinction is fundamental: instead of encoding the geometric phase through static fabrication, the liquid-crystal valve dynamically generates a localized geometric-phase landscape in response to the optical field itself. Such self-induced geometric-phase localization represents a new physical mechanism, bridging geometric-phase optics with nonlinear light–matter interaction.

In this article, we demonstrate that when a liquid-crystal light valve (LCLV) is illuminated by a Gaussian beam, the light can self-induce a geometric-phase structure by light–matter interaction through a self-organized, localized molecular reorientation of the medium, generating a spatially structured phase without imparting orbital angular momentum. Figure 1A illustrates the localized geometric phase formation in the LCLV. This phase originates from the spin angular momentum inversion of circularly polarized light, and the localized average molecular orientation emerges once the system operates above the reorientation transition, Fréedericksz threshold (24) (see Fig. 1C and D). We experimentally verify the generation and manipulation of this localized geometric phase structure using an interferometric technique, which reveals the effective geometric phase of the electromagnetic field envelope. To account for these observations near the reorientation instability of liquid crystals, the system can be described by the real Ginzburg–Landau equation with topological forcing (25, 26). The analytical and numerical simulations show good agreement with the experimental results.

Results

Experimental localized geometric phase

The experimental setup is schematically shown in Fig. 1E. To investigate localized geometric phase structures, we employ a LCLV consisting of a nematic liquid crystal cell with a photoconductive layer. A complete description of the experimental configuration is provided in Materials and methods. The optical valve is driven by a global AC voltage applied across transparent electrodes. When the applied voltage exceeds the Fréedericksz threshold, V0>VFT, the liquid crystal molecules undergo a reorientation transition toward a homogeneous stationary state. The photoconductive layer introduces an effective spatially inhomogeneous voltage proportional to the incident light intensity. In this regime, a localized optical excitation induces a smooth perturbation of the director field, generating closed loops in the molecular orientation, as schematically illustrated in Fig. 1A. Experimentally, this behavior is observed using polarization optical microscopy, where these localized closed loops appear as closed domains, which correspond to the closed black curves (cf. Fig. 1F).

These orientation loops locally modulate the phase of the transmitted light, giving rise to what we refer to as a localized geometric phase. To elucidate this process, we consider a circularly polarized incident plane wave, E→in=E0e^σei(kz−ωt), where e^σ=(1,iσ) denotes the circular polarization vector and σ=±1 the helicity. After propagation through the nematic layer, which can be approximated as a 2D uniaxial birefringent medium, the resulting envelope of the light beam is reduced to (see SI Appendix)

Eout=E0ei(2σϕ+Φdyn). (1)

This expression shows that the azimuthal director angle ϕ(x,y) induces a geometric phase ΦB=2σϕ (1, 2). The dynamical phase contribution Φdyn depends on the molecular reorientation.

To quantitatively characterize the localized phase shift, we employ a Mach–Zehnder interferometer with a small tilt angle β≪1 between the reference and signal beams, generating a spatial interference fringe pattern. Figure 1G shows the resulting interference pattern. The spatial phase distribution can be inferred from the local displacement of the fringes across the structure. A detailed description of the phase imprint on the electric field and the resulting interference intensity profile is provided in the SI Appendix. The interference pattern is extracted from the experimental measurements through the image-processing procedure outlined in Fig. 2A. We then employ a Hilbert transform method to estimate the geometric phase imprinted on the pattern (see SI Appendix for details). Figure 2B and C shows the measured interference pattern and the corresponding reconstructed transverse geometric phase distribution. The helicity of the incident circularly polarized beam also determines the sign of the geometric phase distribution. A comparison between the two helicities is included in the SI Appendix.

Figure 2.

Three-panel figure illustrating the experimental image-processing procedure used to estimate geometric phase from interference fringes. Panel (A) shows a sequence of images detailing the successive processing steps. Panels (B) and (C) show the raw experimental interferogram and the corresponding reconstructed geometric phase distribution, respectively.

Geometrical phase estimation. A) Image-processing workflow used to extract the interference-fringe pattern from the experimental data. Experimental observations: B) raw interferometric snapshot acquired under tilt-angle interferometry, and C) geometrical phase reconstructed via the Hilbert-transform method (see SI Appendix).

Topologically driven Ginzburg–Landau equation

To figure out the origin of this phenomenon, we model the LCLV as a nematic liquid crystal (NLC) cell subjected to an external electric field. The NLC state is described by the director field n→, whose dynamics follow from the minimization of the Frank–Oseen free energy plus electric energy. This free energy includes elastic, electric, and unit-norm constraint contributions, with |n→|=1 (27–29). Close to the reorientational instability, we can derive from first principles a model minimal amplitude equation that describes the qualitative and quantitative dynamics of the system, called a topologically driven Ginzburg–Landau-type equation (TGL) (26):

∂τA=μ(r)A−|A|2A+∇2A+δ∂ηηA¯−f(r)eiθ, (2)

where A is a complex amplitude describing the dynamics of the nematic director projected onto the transverse x–y plane (see Fig. 1). The magnitude of A measures the director’s reorientation relative to the optical axis, while its phase encodes the director’s azimuthal orientation in the plane, more precisely, n→≈(Re{A}sin(πz/d),Im{A}sin(πz/d),1−|A|2sin2(πz/d)) where d the cell thickness. The linear bifurcation parameter is defined as μ(r)=μ0+μ1I(r), where μ0 denotes the intrinsic bifurcation parameter, set by the balance between elastic and electric contributions, and μ1I(r) accounts for the light-induced modification to this parameter. The cubic term accounts for the nonlinear response, and the term proportional to the Laplacian accounts for the elastic coupling. The parameter δ≡(K1−K2)/(K1+K2) accounts for elastic anisotropy, f(r)=f0dI(r)/dr is the light forcing term, and space and time have been appropriately dimensionless.

A detailed derivation of Eq. 2 from the elastic theory of liquid crystals is provided in Ref. (25). This amplitude model exhibits universal features that arise in other physical systems, such as liquid-crystal media driven by magnetic or oscillatory electric fields (30, 31), highlighting the generality of this theoretical description. The phase of the complex amplitude A corresponds to the average azimuthal angle ϕ of the director in the transverse plane. Accordingly, the induced geometric phase is expressed through the phase of the amplitude A as ΦB=2σArg{A}.

Localized geometric phase solutions

In the absence of optical forcing (I=0), Eq. 2 reduces to the real Ginzburg–Landau equation with elastic anisotropy. For μ0>0, it admits a homogeneous stationary solution A=μ0eiϕ0, with arbitrary constant phase ϕ0, reflecting the translational and phase invariance of the system and corresponding to a uniform director orientation in the transverse plane. When a Gaussian beam, I(r)=e−r2/2ω2, is applied, these symmetries are broken and localized defect-type solutions may emerge. Previous studies have reported light-induced defect localization when the illuminated region operates effectively above the reorientation threshold, that is, for μ0<0 and μ1>|μ0| (25, 26). Unexpectedly, when the entire system lies above the reorientation threshold (μ0>0), localized states can also be sustained as a consequence of light-induced translational and phase symmetry breaking. Simulations of Eq. 2 reveal localized inhomogeneous phase solutions. Figure 1C and D shows a vector representation of the amplitude A=(Re(A),Im(A)) and the corresponding geometric phase ΦB, respectively. Notice the good agreement with the experimentally observed geometric phase. In this regime, to demonstrate that the localized structure corresponds to a stationary state of the system, we derive an approximate analytical solution using perturbation theory in the limit f0≪1. We expand the field as A≃A0+A1 and consider the isotropic limit (δ=0), in which the equations for the real and imaginary parts decouple, allowing for analytical treatment. At zeroth order, assuming a wide excitation beam (ω≫1), so that spatial derivatives can be neglected, we obtain A0=μ0+μ1I(r). The first-order correction is then written as A1=u+iv, where

v=f0ω2sin(θ−ϕ0)r(1−I(r)),u=−f0cos(θ−ϕ0)×∫0∞r′dr′∂r′I(r′)K1(2μ0r>)I1(2μ0r<), (3)

with K1 and I1 denoting modified Bessel functions, and r>=max(r,r′) and r<=min(r,r′). Further details of the derivation are provided in the SI Appendix. Figure 3 shows a comparison between the analytical solution 3 and numerical simulations of Eq. 2. Although the analytical solution is derived in an idealized perturbative regime, it captures the qualitative structure of the localized phase distortion. The numerical simulations, performed beyond this perturbative regime, closely resemble the experimental observations (see Fig. 3C and F).

Figure 3.

Six-panel figure showing analytical and numerical solutions of localized structures that induce geometric phases in light. The upper panels illustrate the effective crossed polarization field, while the lower panels show the corresponding geometric phase distributions. The left column presents the analytical perturbative solution, whereas the middle and right columns display numerical simulations of the Ginzburg-Landau model for different parameter values.

Analytical and numerical localized geometric phase. Upper panels show the polarized field Ψ=Re{A}Im{A}, constructed from the real and imaginary parts of the complex amplitude A. Lower panels show the geometric phase field imprinted onto the light through light–matter interaction, ΦB=2Arg{A}. A, D) The analytical perturbative solution A=[μ0+μ1I(r)+(u+iv)]eiϕ0, obtained from Eq. 3 with I(r)=exp(−r2/2ω2) for ϕ0=0∘, ω=50, μ0=0.2, μ1=0.01, and f0=0.02. B, E) and C, F) Numerical simulations of TGL Eq. 2 for ω=50, μ0=0.2, and μ1=0.01, with f0=0.02 and f0=0.1, respectively.

Global phase effect

The resulting localized state is not a topological defect, since it lacks phase singularities and therefore carries no topological charge. Instead, it corresponds to a localized phase distortion locked to the global molecular orientation ϕ0. Consequently, a rotation of the global orientation (ϕ0≠0) induces a rigid rotation of the localized structure. Additional numerical simulations illustrating the dependence of the solution on the global phase orientation ϕ0 are presented in the SI Appendix. Experimentally, these results indicate a direct relation between the global molecular orientation domain and the anisotropic phase texture imprinted onto the light, which can be controlled externally. Figure 4 presents experimental measurements for different imposed global orientations, demonstrating full rotational control in agreement with the numerical predictions. Further details of the protocol used to control the global phase orientation are provided in the SI Appendix.

Figure 4.

Six experimental snapshots showing localized geometric phase structures with different orientations. Panels (A) to (F) illustrate how the orientation of the localized structures changes with the global molecular orientation, demonstrating experimental control over the induced geometric phase.

Experimental control of the localized geometric phase orientation. The orientation of the final state is selected by the displacement direction imposed in the initial condition for V0=22.5 V rms (f=1 kHz) and I0=8.07 mW/cm 2. Details on the preparation of this initial condition are provided in the SI Appendix. Results are shown for ϕ0=45∘ (A), ϕ0=90∘ (B), ϕ0=135∘ (C), ϕ0=−45∘ (D), ϕ0=−90∘ (E), and ϕ0=−135∘ (F).

Dynamical phase effect

For small refractive-index anisotropy, the dynamical phase can be expressed in terms of the complex amplitude field A as

Φdyn=πd2λ(2no+Δn|A|2). (4)

This relation shows that the dynamical phase is governed by the amplitude of the molecular reorientation, which depends on the optical intensity profile. A dynamical phase is accumulated even without light-induced reorientation, simply due to propagation through the medium. Since the magnitude of A varies little spatially, the variations in the dynamic phase are small. Hence, following Eq. 4, we focus on the spatially varying contribution associated with the light-induced localized structure. Figure 5 compares the geometric and dynamical phase contributions. Note that the dynamic phase slightly distorts the Berry phase structure, consistent with experimental observations (see Fig. 2).

Figure 5.

Three-panel figure showing numerical simulations of the geometric and dynamical phase contributions induced by a liquid crystal system. From left to right, the panels show the geometric, dynamical, and total phase distributions imprinted on the optical field.

Numerical solutions of Eq. 2 for ω=50, μ0=0.2, μ1=0.01, and f0=0.1, illustrating the geometric and dynamical phase contributions. A) Geometric phase induced by the liquid-crystal system, ϕ=Arg(A). B) Spatial variations of the dynamical phase using formula 4. C) Total phase imprinted on the optical field.

Discussion

Our experimental and theoretical findings demonstrate that illuminating an optical valve with a light beam above the reorientational transition generates a localized state that imprints an inhomogeneous, spatially confined geometrical phase on light. We show that light–matter interaction in a photosensitive liquid crystal device gives light a dual role: it spontaneously induces a self-organized, localized molecular reorientation and, through this reconfiguration of the medium, imprints a geometric phase on itself, with the resulting structure shaped by the optical intensity distribution. To elucidate the origin of these phase structures, we employ an amplitude equation that captures the director dynamics of the liquid crystal layer. Analytical and numerical solutions show a quite fair agreement with the experimental observations and reveal that the resulting structures correspond to localized phase distortions locked to the global molecular orientation.

Furthermore, we implemented an interferometric scheme that enables spatially resolved estimation of the geometrical phase. These results demonstrate that light-induced director reorientation in liquid crystal devices can generate controllable geometric phase landscapes without relying on prepatterned optical elements. The ability of optical valves to self-induce such localized geometrical phase patterns opens new avenues for the design of reconfigurable optical devices and for controlled light propagation.

The ability to imprint localized geometric phases enables the design of waveguides, modulators, and reconfigurable optical elements for integrated photonics. Harnessing geometric phase as a resource for encoding and manipulating light states also holds promise for both quantum information and classical optics. Furthermore, our results suggest the development of metasurfaces that can generate localized geometric phases. Altogether, this work establishes a robust framework for light-induced geometric-phase localization, paving the way for reconfigurable photonic devices and bridging fundamental optical science with practical applications.

Materials and methods

Materials and experimental setup

The experimental setup is schematically illustrated in Fig. 1E. To investigate the localized geometric phase structure, we employed a LCLV consisting of a nematic liquid crystal (NLC) layer (MLC-6608, Merck) sandwiched between two thin glass substrates separated by a gap of d=15μm. The assembly incorporates a 1 mm-thick transparent photoconductive slab of bismuth silicon oxide (BSO) and two thin conductive indium tin oxide (ITO) layers. The inner surfaces were chemically treated to ensure homogeneous homeotropic anchoring. The nematic liquid crystal exhibits a negative dielectric anisotropy, Δϵ=ϵ∥−ϵ⊥=−4.2ϵ0, where ϵ∥ and ϵ⊥ denote the dielectric constants parallel and perpendicular to the molecular director n→, respectively. Here, the director n→ refers to the average local orientation of the elongated liquid-crystal molecules. The optical excitation is provided by a linearly polarized DPSS laser (λ0=532 nm, I0=8.07mW/cm2). To visualize the induced molecular reorientation, the LCLV is placed in a circular polariscope composed of two quarter-wave plates and two crossed linear polarizers, namely the polarizer P and the analyzer A. The resulting image is magnified by an optical telescope and recorded with a complementary metal-oxide-semiconductor camera (CMOS).

The LCLV is driven by an external AC voltage V0 applied through the ITO films at a frequency of f=1 kHz. Owing to the photosensitive nature of the BSO slab, illumination generates charge carriers, inducing a localized, nonhomogeneous voltage proportional to the incident light intensity. Consequently, the effective voltage across the liquid crystal layer can be expressed as V=ΓV0+αI0, where 0<Γ<1 is a transfer factor determined by the electrical impedances, and α∼4 is a phenomenological parameter describing the linear response of the photoconductive layer. The competition between the elastic restoring force and the applied voltage within the liquid crystal cell leads to the Fréedericksz transition (24), enabling director reorientation under optical excitation. This instability gives rise to vortex defects, which in turn imprint optical vortices onto the transmitted light beam (32).

Phase measurements were performed using a modified Mach–Zehnder interferometer. In this configuration, the beam transmitted through the LCLV interferes with a reference beam reflected by the first beam splitter, which bypasses the device. A small horizontal tilt angle, β=0.3∘≪1, is introduced between the two beams to generate spatial interference fringes.

Image processing

The protocol for estimating the geometric phase involves analyzing near-field interferometric snapshots acquired using the system described above. The images are processed using ImageJ software. First, the interference pattern is isolated from the experimental image by subtracting the slowly varying background intensity. A Fourier transform is then applied to filter the characteristic spatial modes associated with the fringe pattern. After applying the inverse transform, a threshold operation is used to binarize the pattern and isolate the fringe structure. Finally, a Gaussian blur filter is applied to smooth the pattern and recover an approximately sinusoidal intensity profile suitable for phase retrieval.

Supplementary Material

pgag385_Supplementary_Data

Acknowledgments

M.D.Z., M.R.A., and M.G.C. thank for the financial support of ANID-Millenium Science Initiative Program-ICN17_012, Chile (MIRO) and FONDECYT project 1260307.

Contributor Information

Manuel Diaz-Zuniga, Departamento de Física and Millennium Institute for Research in Optics, FCFM, Universidad de Chile, Casilla 487-3, Santiago 8370448, Chile.

Mauro Riquelme-Alcantara, Departamento de Física and Millennium Institute for Research in Optics, FCFM, Universidad de Chile, Casilla 487-3, Santiago 8370448, Chile.

Marcel G Clerc, Departamento de Física and Millennium Institute for Research in Optics, FCFM, Universidad de Chile, Casilla 487-3, Santiago 8370448, Chile.

Supplementary Material

Supplementary material is available at PNAS Nexus online.

Competing Interest

The authors declare no competing interests.

Funding

This work was financially supported by ANID-Millenium Science Initiative Program-ICN_012, Chile (MIRO) and FONDECYT project 1260307.

Author Contributions

M.G.C. conceived and designed the research. M.D.-Z. and M.R.-A. performed the experiments. M.D.-Z. designed the setup, analyzed the data, and wrote the initial draft. All authors contributed to improving the manuscript.

Data Availability

The data and code supporting the findings of this study are available on Zenodo: https://doi.org/10.5281/zenodo.23226697.

References

  • 1. Pancharatnam  S. 1956. Generalized theory of interference, and its applications. Proc Indian Acad Sci A. 44:247–262. [Google Scholar]
  • 2. Berry  MV. 1984. Quantal phase factors accompanying adiabatic changes. Proc R Soc Lond A. 392:45–57. [Google Scholar]
  • 3. Xiao  D, Chang  MC, Niu  Q. 2010. Berry phase effects on electronic properties. Rev Mod Phys. 82:1959–2007. [Google Scholar]
  • 4. Cohen  E, et al.  2019. Geometric phase from Aharonov–Bohm to Pancharatnam–Berry and beyond. Nat Rev Phys. 1:437–449. [Google Scholar]
  • 5. Bliokh  KY, Gorodetski  Y, Kleiner  V, Hasman  E. 2008. Coriolis effect in optics: unified geometric phase and spin-hall effect. Phys Rev Lett. 101:030404. [DOI] [PubMed] [Google Scholar]
  • 6. Bomzon  Z, Biener  G, Kleiner  V, Hasman  E. 2002. Space-variant Pancharatnam-Berry phase optical elements with computer-generated subwavelength gratings. Opt Lett. 27:1141–1143. [DOI] [PubMed] [Google Scholar]
  • 7. Biener  G, Niv  A, Kleiner  V, Hasman  E. 2002. Formation of helical beams by use of Pancharatnam-Berry phase optical elements. Opt Lett. 27:1875–1877. [DOI] [PubMed] [Google Scholar]
  • 8. Marrucci  L, Manzo  C, Paparo  D. 2006. Optical spin-to-orbital angular momentum conversion in inhomogeneous anisotropic media. Phys Rev Lett. 96:163905. [DOI] [PubMed] [Google Scholar]
  • 9. Nersisyan  SR, Tabiryan  NV, Mawet  D, Serabyn  E. 2013. Improving vector vortex waveplates for high-contrast coronagraphy. Opt Express. 21:8205–8213. [DOI] [PubMed] [Google Scholar]
  • 10. Rafayelyan  M, Tkachenko  G, Brasselet  E. 2016. Reflective spin-orbit geometric phase from chiral anisotropic optical media. Physi Rev Lett. 116:253902. [DOI] [PubMed] [Google Scholar]
  • 11. Barboza  R, Bortolozzo  U, Clerc  MG, Residori  S. 2016. Berry phase of light under Bragg reflection by chiral liquid-crystal media. Phys Rev Lett. 117:053903. [DOI] [PubMed] [Google Scholar]
  • 12. Kugler  M, Shtrikman  S. 1988. Berry’s phase, locally inertial frames, and classical analogues. Phys Rev D. 37:934. [DOI] [PubMed] [Google Scholar]
  • 13. Milione  G, Evans  S, Nolan  DA, Alfano  RR. 2012. Higher-order Pancharatnam-Berry phase and the angular momentum of light. Phys Rev Lett. 108:190401. [DOI] [PubMed] [Google Scholar]
  • 14. Yun  S, Kim  J, Li  Y, Escuti  MJ. 2024. High-efficiency flat Pancharatnam-Berry optical elements for diffraction-limited beam shaping. Opt Express. 32:31107–31119. [Google Scholar]
  • 15. Dong  Y, et al.  2025. Switchable Pancharatnam-Berry phases in heterogeneously integrated THz metasurfaces. Adv Mater. 37:2417183. [DOI] [PubMed] [Google Scholar]
  • 16. Liu  Y, et al.  2017. Generation of perfect vortex and vector beams based on Pancharatnam-Berry phase elements. Sci Rep. 7:44096. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Yun  H, Chen  H, Jiang  M, Wei  Q-H. 2024. Converting Gaussian to line beams with high efficiency by flat liquid crystal optical elements. Opt Express. 32:47076–47085. [Google Scholar]
  • 18. Pan  J-T, et al.  2024. Nonlinear geometric phase coded ferroelectric nematic fluids for nonlinear soft-matter photonics. Nat Commun. 15:8732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Alberucci  A, Jisha  CP, Marrucci  L, Assanto  G. 2016. Electromagnetic confinement via spin-orbit interaction in anisotropic dielectrics. ACS Photonics. 3:2249–2254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Jisha  CP, Alberucci  A, Marrucci  L, Assanto  G. 2017. Interplay between diffraction and the Pancharatnam-Berry phase in inhomogeneously twisted anisotropic media. Phys Rev A. 95:023823. [Google Scholar]
  • 21. Arumugam  SV, Jisha  CP, Marrucci  L, Alberucci  A, Nolte  S. 2023. Exploring the impact of longitudinal modulation on the twisting angle in Pancharatnam-Berry phase-based waveguides. Opt Express. 31:44283–44294. [DOI] [PubMed] [Google Scholar]
  • 22. Chandran Thodika  S, Brasselet  E. 2025. Geometric phase self-focusing from optical spin angular momentum transfer to matter. Phys Rev A. 111:023518. [Google Scholar]
  • 23. El Ketara  M, Kobayashi  H, Brasselet  E. 2021. Sensitive vectorial optomechanical footprint of light in soft condensed matter. Nat Photon. 15:121. [Google Scholar]
  • 24. Fréedericksz  V, Zolina  V. 1933. Forces causing the orientation of an anisotropic liquid. Trans Faraday Soc. 29:919–930. [Google Scholar]
  • 25. Barboza  R, Bortolozzo  U, Clerc  MG, Residori  S, Vidal-Henriquez  E. 2015. Optical vortex induction via light-matter interaction in liquid-crystal media. Adv Opt Photonics. 7:635–683. [Google Scholar]
  • 26. Aguilera-Rojas  PJ, Clerc  MG, Diaz-Zuniga  M, Gajardo-Pizarro  R. 2025. Light-induced matter vortex transitions. Physica D. 476:134599. [Google Scholar]
  • 27. Khoo  I-C. Liquid crystals. John Wiley & Sons, 2022. [Google Scholar]
  • 28. de Gennes  PG, Prost  J. The physics of liquid crystals. Oxford Science Publications, 1993. [Google Scholar]
  • 29. Chandrasekhar  S. Liquid crystals. Cambridge University Press, 1992. [Google Scholar]
  • 30. Brasselet  E. 2018. Tunable high-resolution macroscopic self-engineered geometric phase optical elements. Phys Rev Lett. 121:033901. [DOI] [PubMed] [Google Scholar]
  • 31. Calisto  E, Clerc  MG, Zambra  V. 2020. Magnetic field-induced vortex triplet and vortex lattice in a liquid crystal cell. Phys Rev Res. 2:042026(R). [Google Scholar]
  • 32. Barboza  R, et al.  2012. Vortex induction via anisotropy stabilized light-matter interaction. Phys Rev Lett. 109:143901. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

pgag385_Supplementary_Data

Data Availability Statement

The data and code supporting the findings of this study are available on Zenodo: https://doi.org/10.5281/zenodo.23226697.


Articles from PNAS Nexus are provided here courtesy of Oxford University Press

RESOURCES