Thermal Shrinkage-Induced Modifications in Photonic Band Gaps of Two-Photon Polymerized Bragg Reflectors

TL;DR

The study demonstrates that uniform thermal shrinkage of 2PP-DLW–fabricated pillar-supported 1D polymer DBRs can reliably reduce the optical period by up to about φ ≈ 5, enabling pronounced blue shifts of photonic bandgaps into the visible range. Angular-resolved optical measurements via Fourier image spectroscopy (FIS) and full 3D FDTD simulations (including pillar networks) show that higher-order bands move toward shorter wavelengths as the period contracts, while the fundamental bandgap remains beyond the measured spectrum. Bragg-based estimates using $m\lambda = 2d$ and $d = n_1 d_1 + n_2 d_2$ (for normal incidence) informed by SEM-derived dimensions reproduce general trends and explain observed shifts, though pillar-induced scattering and nonuniform shrinkage complicate exact band positions. Overall, the work establishes a scalable approach to tune the photonic band structure of polymer PhCs and provides insights into how 3D pillar networks influence band visibility and strength, guiding future design of shrinkable photonic crystals for visible–NIR applications.

Abstract

One-dimensional (1D) polymer-based photonic crystals (PhCs) in the $1.55~μm$ wavelength range can be easily created using a two-photon direct laser writing system. To achieve shorter period structures, we report the use of thermal shrinkage of two-photon polymerized structures, at elevated temperatures, to eliminate unpolymerised material, leading to the uniform shrinkage of the distributed Bragg reflector (DBR) structures by a ratio of $\sim$ 2.5 to 5. Our Finite Difference Time Domain (FDTD) simulation and the angle-resolved light scattering characterization technique using Fourier image spectroscopy (FIS) measurements show that the low order photonic bandgaps of DBRs blue-shift across the NIR-visible region (850 to 400 nm) as the shrinkage increases.

Figures (8)

• Figure 1: The illustrated 3x3x3 pillar supported DBR PhC, with the unit cell selected at the corner, a is the unit cell size in the x-y period and the pillar period. D is the pillar diameter; $d_1, d_2$ and $n_1,n_2$ refer to the thickness of the layer and the material refractive index, respectively.
• Figure 2: The SEM image of pillar-supported DBR demonstrated for this research. The $d_1$ and $d_2$ are the thickness of air and polymer. In this study, $d_1$ is around 2.8µm and $d_2$ is around 1µm. The detail of the parameter is shown in Table \ref{['table:parameter']}. The actual templates for post-processing obtain more periods.
• Figure 3: Chamber temperature versus processing time for different annealing recipes. In each case, the temperature is first increased to a maximum of 450 ℃ with a heating rate of 10 ℃/min, then remains there for different times depending on the annealing recipe and finally cools back to room temperature at a rate of -20 ℃/min.
• Figure 4: SEM images (top row: a-e) and corresponding reflection images under a white light source (bottom row: f-j) of pillar-supported DBR without shrinkage (a, f) and with shrinkage times of 0 (b, g), 4 (c, h), 8 (d, i), and 12 (e, j) minutes respectively.
• Figure 5: Shrinkage ratios with error bars as a function of different annealing times (See Fig. \ref{['fig:recipe']}), from 0 minutes representing no annealing stage, to 12 minutes annealing time. The shrinkage ratio ascends to 5 times smaller than the original design, approximately.