Skip to content

DBT TIR-SIL fiber coupler v1

Waveguide -> printed IP-S TIR-SIL -> 630HP fiber mode coupling workflow.

Tidy3D workflow for coupling an x-oriented DBT dipole through a waveguide stack, a printed IP-S TIR-SIL, and into a 630HP-like fiber mode.

This version is rebuilt around the workflow in:

G:\Shared drives\QLuster\product\v1\tidy3d-out-of-plane\solid_immersion_lens_v6\v6.ipynb

The script follows the same structure as that notebook: build the lens mesh, build the Tidy3D simulation, solve/check the fiber mode, run the forward FDTD job, compute coupling as |a_fiber,+|^2 / flux_dip, export STL, and write a Markdown report with geometry, field, mode, and coupling plots.

Files

  • tir_sil_simulation.py contains the full simulation/report workflow.
  • stl/ receives generated IP-S TIR-SIL STL files.
  • reports/ receives timestamped reports, figures, simulation files, and result JSON.

Geometry

  • SiO2 substrate, n = 1.44
  • 450 nm x 300 nm Si3N4 waveguide, n = 2.0
  • Air beside the waveguide in the waveguide layer
  • 200 nm DBT-host crystal film, n = 1.8
  • x-oriented point dipole on the crystal top surface
  • IP-S TIR-SIL with a central direct lens plus annular TIR reflector
  • The direct lens branch is explicitly limited to rays whose lens-air incidence is below the IP-S/air critical angle with a configurable margin.
  • The report includes angular coverage accounting so unassigned source-ray angles are visible instead of silently ignored.
  • 630HP-like fiber modeled as a step-index core/cladding above the optic

Usage

Dry-run the geometry, STL, simulation object, and report without spending Tidy3D credits:

python .\tir_sil_simulation.py --dry-run

Run the full fiber-coupling simulation:

python .\tir_sil_simulation.py

Sweep the lower TIR source-ray cutoff angle:

python .\tir_sil_simulation.py --dry-run --sweep-tir-theta-min 83.5 84.0 84.5

Useful geometry knobs:

  • --direct-lens-apex-height-um
  • --direct-lens-theta-max-deg
  • --direct-lens-tir-margin-deg
  • --tir-theta-min-deg
  • --tir-theta-max-deg
  • --tir-parabola-constant-um
  • --fiber-gap-um

Results

First sweep point (32.0deg wall, baseline)

DBT TIR-SIL Fiber Coupler v1 Report

  • Generated: 2026-05-28T15:40:08
  • Status: dry run only
  • STL: stl/tir_sil_wall_32.0deg.stl
  • Simulation file: reports/data/simulation_20260528_153959_wall_32p0deg.hdf5
Geometry

Coordinate convention: z = 0 is the SiO2 top surface and the Si3N4 waveguide bottom. The DBT molecule is modeled as a point dipole at the crystal top surface: z = 0.500 um.

  • SiO2 substrate: n = 1.440
  • Si3N4 waveguide: width 0.450 um, thickness 0.300 um, n = 2.000
  • Lateral material beside waveguide: air, n = 1.000
  • DBT host crystal film: thickness 0.200 um, n = 1.800, infinite x/y extent in the simulation
  • IP-S TIR-SIL: n = 1.515, base radius 3.000 um, height 3.000 um, wall angle 32.00 deg
  • IP-S/air critical angle: 41.30 deg measured from the surface normal

TIR-SIL radial profile

Permittivity geometry plots

Fiber Target

The fiber target is Thorlabs 630HP. Thorlabs specifies 630HP as a 600-770 nm single-mode fiber with mode-field diameter 4.0 +/- 0.5 um @ 630 nm. The simulation wavelength is 780 nm, which is slightly outside that listed 630HP operating band, so the 780 nm MFD is treated as an estimate here. This first-pass model uses a 630HP-like step-index fiber stub:

  • Cladding radius: 62.5 um
  • Core radius: 1.750 um
  • NA: 0.120
  • Cladding index: 1.4533
  • Derived core index: 1.4582
  • Estimated 780 nm MFD used for monitor sizing: 4.95 um
  • Mode monitor plane: z = 6.650 um, width 6.400 um
Simulation
  • Center wavelength: 0.7800 um
  • Wavelength samples: 3
  • Simulation window: 8.00 x 8.00 x 10.60 um
  • Grid: Tidy3D auto grid, min 16 steps per wavelength
  • Dipole polarization: Ex
  • Run time: 7.500e-13 s
  • Estimated cost: nan FlexCredits
  • Real cost: nan FlexCredits
Field Plots

Field plots are generated after a full Tidy3D run.

Fiber-Mode Signal

Simulation was not run because --dry-run was used.

Parameters
{
  "wavelength_um": 0.78,
  "bandwidth_um": 0.02,
  "n_wavelengths": 3,
  "n_air": 1.0,
  "n_sio2": 1.44,
  "n_sin": 2.0,
  "n_crystal": 1.8,
  "n_ips": 1.515,
  "fiber_mfd_spec_um_at_630": 4.0,
  "fiber_mfd_est_um_at_780": 4.95,
  "fiber_clad_radius_um": 62.5,
  "fiber_core_radius_um": 1.75,
  "fiber_na": 0.12,
  "fiber_clad_index": 1.4533,
  "fiber_gap_um": 1.5,
  "fiber_stub_um": 3.0,
  "substrate_thickness_um": 1.0,
  "waveguide_width_um": 0.45,
  "waveguide_thickness_um": 0.3,
  "crystal_thickness_um": 0.2,
  "tir_wall_angle_deg": 32.0,
  "tir_base_radius_um": 3.0,
  "tir_cone_height_um": 2.2,
  "tir_cap_sag_um": 0.8,
  "tir_profile_points": 180,
  "tir_angular_points": 160,
  "sim_xy_um": 8.0,
  "pml_pad_um": 0.8,
  "min_steps_per_wvl": 16,
  "run_time_s": 7.5e-13,
  "monitor_plane_xy_um": 7.0,
  "source_polarization": "Ex",
  "tir_height_um": 3.0,
  "tir_apex_z_um": 3.5,
  "fiber_face_z_um": 5.0,
  "fiber_core_index": 1.4582458263269606,
  "ips_air_critical_angle_deg": 41.30490702409531
}
Source Notes
  • Nanoscribe IP-S refractive index was taken from Nanoscribe's printing-materials product folder: n = 1.515 at 589 nm and 20 C: https://www.nanoscribe.com/fileadmin/Nanoscribe/PDF/Product-folder/Folder-Printing-Materials.pdf Dispersion at 780 nm should be replaced by measured data if available.
  • 630HP MFD was taken from Thorlabs' 630HP table: 4.0 +/- 0.5 um @ 630 nm. https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=949&pn=630HP The 780 nm value here is a simple wavelength-scaled estimate for monitor sizing, not a vendor specification.
  • TIR-SIL starting point: https://arxiv.org/abs/2007.06369

Latest sweep point (final)

DBT Waveguide TIR-SIL Fiber Coupling Report

  • Generated: 2026-05-29T08:26:20
  • Status: dry run only
  • Reference implementation style: tidy3d-out-of-plane/solid_immersion_lens_v6/v6.ipynb
  • Cost-estimate task: fdve-2ad3547f-fd3e-47b9-9a77-9d9fd040fe59
  • STL: stl/tir_sil_waveguide_tir40p5_C2p00_m7_Ex.stl
  • Simulation file: reports/data/simulation_20260529_082608_tir40p5_C2p00_m7_Ex.hdf5
Geometry
  • SiO2 substrate: n = 1.440, simulated thickness 1.000 um
  • Si3N4 waveguide: n = 2.000, width 0.450 um, thickness 0.300 um
  • Air beside the waveguide in the waveguide layer
  • DBT host crystal: n = 1.800, thickness 0.200 um
  • Dipole position: (0, 0, 1.500000) um, exactly on the crystal top surface
  • Dipole orientation: x direction (Ex)
  • IP-S TIR-SIL index: n = 1.515
  • Fiber: core radius 1.750 um, cladding index 1.4533, derived core index 1.4582, NA 0.120

Geometry

TIR-SIL Ray Design

The central branch is a Cartesian-oval direct lens for low-angle rays whose incidence at the IP-S/air lens surface is safely below the critical angle. Rays that would TIR at that refractive surface are not assigned to the central branch. The raised annular branch is the reflection-law surface with local normal parallel to k_in - zhat. For the implemented vertical output beam this integrates to the paraboloid condition s - z = C, with the emitter at the focus. The annular roof height is then chosen so the direct and TIR branches differ by an integer number of wavelengths at 780 nm.

  • Direct-lens source angles: 0 to 40.00 deg
  • Direct-lens radius: 1.326 um
  • Max direct-lens internal incidence: 40.5800 deg
  • Direct-lens margin to critical angle: 0.7249 deg (required margin 0.0010 deg)
  • TIR source angles: 40.50 to 89.90 deg
  • TIR radii: 2.003 to 5.421 um
  • TIR incidence range: 45.0500 to 69.7500 deg
  • IP-S/air critical angle: 41.3049 deg
  • Single-bounce vertical TIR source-angle upper limit: 97.39 deg
  • Minimum TIR-surface margin above critical: 3.7451 deg
  • TIR reflection-law error: max 4.965e-16, rms 1.923e-16 against +z
  • Unassigned source-angle gap: 0.500 deg
  • Accounted solid-angle fraction of upper hemisphere: 99.26% (direct 23.40%, TIR 75.87%, gap 0.5638%)
  • Direct-branch OPL variation: 1.776e-15 um
  • TIR-branch OPL variation: 7.994e-14 um
  • Branch phase offset: 7.000000 wavelengths (m = 7)
Simulation
  • Wavelength range: 0.750 to 0.810 um with 31 samples
  • Simulation size: 11.842 x 11.842 x 12.218 um
  • Transverse domain rule: optic diameter 10.842 um plus 1.000 um
  • Simulated substrate thickness: 1.000 um
  • Simulated fiber length above fiber face: 1.000 um
  • Symmetry: (-1, 1, 0) (Ex dipole: odd in x, even in y)
  • Grid: auto, min 15 steps per wavelength
  • Mode monitor z: 11.718 um
  • Mode monitor plane: 8.500 x 8.500 um
  • Flux monitor: flux_dip, same normalization pattern as v6
  • Field plot: Tidy3D sim_data.plot_field('field_xz', 'E', 'abs^2') at the center-frequency field monitor; no time-step selection is involved
  • Estimated cost: 0.6962411590749145 FlexCredits
  • Real cost: nan FlexCredits
Results

Simulation was not run because --dry-run was used.

Field plots are generated after a full Tidy3D run.

Parameters
{
  "wl_um": 0.78,
  "bw_um": 0.06,
  "n_wl": 31,
  "n_air": 1.0,
  "n_sio2": 1.44,
  "n_sin": 2.0,
  "n_crystal": 1.8,
  "n_sil": 1.515,
  "n_fiber_clad": 1.4533,
  "fiber_na": 0.12,
  "t_sio2_um": 1.0,
  "waveguide_width_um": 0.45,
  "waveguide_thickness_um": 0.3,
  "crystal_thickness_um": 0.2,
  "fiber_core_radius_um": 1.75,
  "fiber_clad_diameter_um": 125.0,
  "fiber_gap_um": 2.0,
  "fiber_stub_um": 1.0,
  "direct_lens_apex_height_um": 3.0,
  "direct_lens_theta_max_deg": 40.0,
  "direct_lens_tir_margin_deg": 0.001,
  "tir_theta_min_deg": 40.5,
  "tir_theta_max_deg": 89.9,
  "tir_parabola_constant_um": 2.0,
  "tir_roof_clearance_um": 0.8,
  "transition_clearance_um": 0.08,
  "direct_profile_points": 120,
  "tir_profile_points": 120,
  "angular_segments": 144,
  "min_steps_per_wvl": 15,
  "run_time_s": 5e-12,
  "pml_spacing_um": 0.46799999999999997,
  "sim_xy_um": 0.0,
  "domain_margin_um": 1.0,
  "mode_plane_um": 8.5,
  "eff_inf_um": 10.0,
  "source_polarization": "Ex",
  "n_fiber_core": 1.4582458263269606,
  "z_substrate_top_um": 1.0,
  "z_waveguide_top_um": 1.3,
  "z_crystal_top_um": 1.5,
  "z_emitter_um": 1.5,
  "z_roof_um": 9.21844660194175,
  "z_fiber_face_um": 11.21844660194175,
  "z_mode_monitor_um": 11.71844660194175,
  "domain_xy_um": 11.842474437339622,
  "optic_outer_radius_um": 5.421237218669811,
  "direct_lens_radius_um": 1.3259890361204258,
  "direct_lens_edge_z_um": 3.0802521975370514,
  "tir_inner_radius_um": 2.0034937082264417,
  "tir_outer_radius_um": 5.421237218669811,
  "tir_z_max_um": 7.847453245272696,
  "phase_report": {
    "direct_opl_mean_um": 9.26344660194175,
    "direct_opl_ptp_um": 1.7763568394002505e-15,
    "tir_opl_mean_um": 14.723446601941752,
    "tir_opl_ptp_um": 7.993605777301127e-14,
    "branch_opl_delta_um": 5.460000000000003,
    "branch_phase_offset_wavelengths": 7.0000000000000036,
    "phase_order": 7,
    "direct_lens_incidence_max_deg": 40.579977333803406,
    "direct_lens_incidence_margin_to_critical_deg": 0.7249296902919014,
    "direct_lens_tir_margin_required_deg": 0.001,
    "tir_incidence_min_deg": 45.05000000000002,
    "tir_incidence_max_deg": 69.75,
    "tir_incidence_min_margin_above_critical_deg": 3.7450929759047114,
    "critical_angle_deg": 41.30490702409531,
    "single_bounce_vertical_tir_source_angle_max_deg": 97.39018595180939,
    "tir_reflection_direction_error_max": 4.965068306494546e-16,
    "tir_reflection_direction_error_rms": 1.9229626863835638e-16,
    "unassigned_source_angle_gap_deg": 0.5,
    "direct_solid_angle_fraction_upper_hemisphere": 0.233955556881022,
    "gap_solid_angle_fraction_upper_hemisphere": 0.005638477518947083,
    "tir_solid_angle_fraction_upper_hemisphere": 0.7586606372341327,
    "total_accounted_solid_angle_fraction_upper_hemisphere": 0.9926161941151547
  }
}

20 additional parametrized sweep reports exist under reports/.

Downloads

Source: nanophotonic_devices/fiber_collection_optics/dbt_tir_sil_fiber_coupler_v1/