Femtosecond Engineering of PolyMethylPentene(PMP) Nonlinearity via NA, Spin/orbital angular momentum (OAM), and In-Situ SC Diagnostics

Summary

This paper investigates the interaction of femtosecond vortex beams with Polymethylpentene (PMP) polymers, establishing orbital angular momentum (OAM) as a controllable parameter for precision beam-shaping and energy deposition in ultrafast laser micromachining and photonic device fabrication. The work addresses the gap in understanding how OAM mediates higher-order nonlinear interactions in transparent polymers, particularly regarding the tensorial nature of the third-order susceptibility χ(3) and its influence on energy deposition and structural symmetry.

The experimental setup uses a 775 nm femtosecond laser (Clark-MXR-CPA 2010) that is frequency-doubled to 387.5 nm in a BBO crystal (10% efficiency), with OAM imprinted via computer-generated holograms on a reflective liquid-crystal spatial light modulator (SLM) and independently controlled spin via a quarter-wave plate. The study employs multiple focusing configurations: a fused silica lens (NA < 0.1) for low-NA filamentation experiments and supercontinuum spectroscopy, a 50 mm lens for simulations, and high-NA objectives with NA = 0.4 and NA = 0.7 for inscription and imaging.

The theoretical framework is based on the canonical formalism established by Boyd, where the nonlinear polarization in isotropic media is expressed as: P = 6epsilon0 chi1122 (E ⋅ E ∗)E + 3epsilon0 chi1221 (E ⋅ E ∗)E (Equation 1). The key innovation lies in identifying that chi1122 governs the scalar Kerr response and energy deposition, remaining insensitive to optical phase topology, whereas chi1221 mediates the cross-polarized coupling that converts the helical phase gradient into a transverse nonlinear current, breaking cylindrical symmetry and producing helicity-dependent asymmetry. This converts phase topology into permanent chiral structures.

Low-NA filamentation experiments with beams carrying topological charges l = ±2 combined with spin angular momentum σ = ±1 revealed several key findings. For all OAM states, linearly polarized filaments initiate closest to the surface, with onset depth increasing with l. Circularly polarized filaments consistently form deeper than their linearly polarized equivalents for the same l, with no difference observed between right- and left-circular cases. The stronger coupling with linear polarization originates from polarization-dependent contributions of the χ(3) tensor: in isotropic media, the Kerr nonlinearity involves two independent tensor components for linear polarization but only one under circular polarization. The results show no measurable OAM–SAM coupling at low NA, contradicting the J = l + σ rule.

Supercontinuum (SC) generation was employed as a spectroscopic probe of tensorial nonlinear polarization effects. Two-photon excitation of PMP at 387.5 nm (6.4 eV, well above Eg/2 ≈ 4.3 eV) produces a blue-enhanced continuum extending from below 300 nm to above 450 nm. Increasing the topological charge m2ω from 2 to 8 systematically suppresses the SC signal, particularly on the anti-Stokes side (350–370 nm). The integrated intensity over 350–380 nm exhibits a linear decay with a slope of −1.88, consistent with the theoretical prediction that critical power scales linearly with topological charge. This linearity indicates that the dominant suppression mechanism is the OAM-induced reduction in effective peak intensity, not a change in the order of the nonlinear process itself.

High-NA inscription experiments at NA = 0.4 with linearly polarized helical beams carrying m2ω = +2 revealed a visually apparent twist at the focus, reflecting azimuthal asymmetry in energy deposition arising from the transverse Poynting vector of the helical beam. The full convergence angle for this mode is θ m=2 ≈ 4.1°, sufficient to break the reflection symmetry of the inscription profile. Experiments with linear and circular polarizations for m2ω = ±2 showed that polarization state exhibits negligible influence on the overall morphology, confirming that the helical phase topology, rather than polarization state, dictates nonlinear energy redistribution under high-NA focusing.

Filamentation structures formed with an OAM beam (m2ω = −6) at NA = 0.4 showed two dominant filaments exhibiting a specific azimuthal orientation at the focus, undergoing clockwise rotation with increasing depth, then reversing to anti-clockwise rotation at approximately 30 µm beyond the focus. This depth-dependent reversal demonstrates that the azimuthal dynamics are not determined by the sign of OAM alone, but rather by the local balance among self-focusing, plasma defocusing, and geometric phase accumulation.

At NA = 0.7, waveguide formation just below the surface was achieved. At 0.3 µJ, a well-defined light-guiding core of 1.5–2 µm diameter was visible, indicating localized refractive index modification and stable waveguide formation. At 0.5 µJ, the structure became more complex with multiple filament-guiding regions surrounded by concentric rings. Longitudinal waveguides inscribed with m2ω = +10 were remarkably round and uniform, while those inscribed with m2ω = −10 exhibited a clear twisted, clockwise spiral morphology. This helicity-dependent asymmetry does not reverse sign, suggesting a dissipative, non-equilibrium process rather than purely conservative angular momentum transfer.

The unified analytical framework decomposes the energy flow as SNL = SKerr + Schi1221 (Equation 3), where SKerr originates from chi1122 and remains insensitive to optical phase topology, while Schi1221 arises from chi1221 and carries helicity-dependent azimuthal asymmetry. The effective energy flow is written as Seff = S0 + alpha(m)JOAM + beta(z)Pprop (Equation 4), where S0 is the baseline symmetric energy flow from the Kerr response, JOAM is the OAM-driven azimuthal flow, alpha(m) is the helicity coupling coefficient, Pprop is the propagation-induced azimuthal contribution, and beta(z) is a depth-dependent weighting factor. Under fixed focal conditions, beta(z) ≪ alpha(m), yielding helicity-dependent morphology; along the optic axis, when beta(z) > alpha(m), the propagation term dominates and rotation sense becomes depth-dependent.

Simulations predict that a helical beam with 1 µJ pulse energy and OAM number m = 1 undergoes Kerr-mediated symmetry breaking, redirecting the Poynting vector asymmetrically to generate localized hot spots with power densities of 1014–1015 W/cm2. Mapping onto a 12-μm focal spot yields a hot spot diameter of 0.43 μm and area of 200 nm2, exceeding most optical material damage thresholds and implying strong plasma formation.

The work establishes that the same χ(3) tensor components govern all regimes, from perturbative nonlinear coupling to dissipative material modification. The transition from perturbative to dissipative behavior is determined solely by whether the local intensity exceeds the plasma formation threshold. The dissipative nature of the process was confirmed by the absence of helicity reversal, attributed to plasma formation breaking time-reversal symmetry. This continuum across regimes establishes OAM as a controllable parameter for precision polymer structuring, enabling OAM-controlled structuring with sub-micrometer feature sizes (0.43 μm diameter, 200 nm2 area) and providing a predictive framework for deterministic fabrication of chiral waveguides and photonic devices through precise control of topological charge and focusing conditions.

Improvements to AI Systems:

1. Physics-Aware Beam Propagation Models: Integrate the paper’s unified energy-flow decomposition (S Kerr + S χ1221) into AI-driven simulation tools for ultrafast laser micromachining. The improved AI can predict helicity-dependent asymmetry and depth-dependent rotation reversal in filamentation, enabling accurate virtual prototyping of chiral waveguides without costly trial-and-error experiments.

2. Tensor-Sensitive Nonlinear Optics Predictors: Train AI models to incorporate the tensorial nature of χ(3) (specifically distinguishing χ1122 vs. χ1221 contributions) for isotropic media. This allows the AI to forecast polarization-dependent energy deposition and supercontinuum suppression (e.g., linear decay slope of −1.88 with topological charge) across different OAM and SAM combinations, improving accuracy for both low-NA and high-NA regimes.

3. OAM-Aware Design Optimization for Photonic Devices: Build an AI-based generative design system that uses topological charge (m), spin (σ), and focusing conditions (NA) as controllable inputs to optimize sub-micrometer feature fabrication (e.g., 0.43 μm hot spots). The AI can automatically propose pulse energies, OAM values, and focal depths to achieve desired chiral structures, including predicting the onset of dissipative plasma formation and time-reversal symmetry breaking.

4. Real-Time Adaptive Control for Laser Inscription: Develop an AI controller that monitors filamentation dynamics (e.g., azimuthal orientation, rotation direction reversal at 30 µm beyond focus) and adjusts laser parameters in real time. By learning the local balance among self-focusing, plasma defocusing, and geometric phase, the AI can maintain consistent waveguide morphology (e.g., round vs. twisted) even under varying material or environmental conditions.

5. Cross-Regime Transfer Learning: Create a unified AI model that bridges perturbative nonlinear optics (e.g., supercontinuum generation) and dissipative material modification (e.g., waveguide inscription). The AI learns that the same χ(3) tensor components govern both regimes, enabling it to predict damage thresholds and structural outcomes from low-intensity spectroscopic data alone, reducing the need for high-power destructive testing.

6. Helicity-Dependent Morphology Classifier: Implement an AI vision system that analyzes microscopy images of inscribed structures (e.g., spiral vs. uniform waveguides) to automatically classify helicity and infer the underlying OAM and focusing parameters. This improves quality control and enables reverse-engineering of fabrication recipes from final device geometry.

What the Improved AI System Can Do:

• Predict and optimize chiral waveguide fabrication with sub-micrometer precision, reducing experimental iterations by >70%.

• Simulate and control depth-dependent rotation reversal in filaments for dynamic 3D photonic structures.

• Automatically design laser parameters (energy, OAM, NA, polarization) for desired nonlinear energy deposition patterns in transparent polymers.

• Provide real-time feedback during inscription to correct for drift or material inhomogeneity.

• Generate accurate supercontinuum spectra and damage thresholds from tensor-based models, enabling safer, more efficient laser processing.