Star Grazing with Alumina Grass: Antireflection coatings in the visible and near-infrared on IPX-Clear Microlenses assisted by Grass-like Alumina

Abstract

Two-photon polymerization (2PP) enables fabrication of high-precision micro-optics with complex freeform geometries, opening a new parameter space for custom astronomical optics. Among available resins, the newly developed IPX Clear is particularly well suited for visible applications, offering high transmission across the visible-near-IR, low surface roughness, and excellent shape fidelity. However, Fresnel reflections at the air-polymer interface introduce significant optical losses, which are detrimental in low-signal astronomy. Previous studies show grass-like alumina coatings on glass and fused silica can raise average transmission from 91.9% to approximately 99% over 400-900 nm. Here we explore the feasibility of Atomic Layer Deposition (ALD) to apply such coatings to IPX-Clear micro-optics over 400-1700 nm. Grass-like alumina anti-reflective (AR) coatings can approximate the ideal index condition by creating a gradual refractive-index transition from air to bulk IPX Clear, suppressing surface reflections. While grass-like coatings are established on bulk optics and conformal ALD films have been applied to 2PP micro-optics, we demonstrate - for the first time - alumina grass on 2PP microlenses made with the new IPX-Clear resin. We discuss key challenges and process steps, and observe that alumina-grass-coated microlenses lose only approximately 0.3% of photons to reflection in the 400-850 nm range. Future work will test performance across the full 400-1700 nm band and explore improved environmental resilience, e.g., a SiO2 overcoat. Combined with the high optical transparency of IPX Clear, these coatings enable custom-designed, highly efficient microlenses for astronomical applications.

Figures (8)

• Figure 1: A) Side view of micro lenses being developed for the LFAST project, with diameter = $245 \mathrm{\mu m}$, radius of curvature =$245 \mathrm{\mu m}$ and, height = $19.2\mathrm{\mu m}$. Lens material is IPX clear, and Substrate is Fused Silica. B) SEM image of the Alumina Grass AR coating deposited on the surface of the micro lens. The coating is formed by depositing 35 nm of Alumina bloomed at $90^\circ$ C for 10 minutes in DI water
• Figure 2: Change in refractive index of the alumina (y axis) with change in ALD chamber/Growth temperature (x axis). The data points are from Groner. Here, we perform a linear fit to estimate the Alumina refractive index given a deposition temperature. We found the relation to be $n(T) = 1.48 + 7.96\times10^{-4} T$. This study also performed physical measurements using an ellipsometer (Red Star) for alumina deposited at $80^\circ\mathrm{C}$. The results were nearly identical to previous work.Groner
• Figure 3: Reflectance% vs wavelength relation of an uncoated fused silica slide and 3 double sided ALD Alumna coated fused silica slides bloomed at different temperatures. All slides were coated with 35 nm of Alumina using ALD at a chamber temperature of $80^\circ C$. Each slide was then bloomed in a DI water bath for 10 minutes at $50^\circ C$, $70^\circ C$, and $90^\circ C$ respectively. We then measure the Reflectance vs Wavelength relation of the bloomed slides and the uncoated slide. The slight kink observed at the $\sim700nm$ is due to grating change in the Carry 7000.
• Figure 4: Absorbance% of a fused silica slide that was initially coated with 35 nm of alumina on both sides and then bloomed at $90^\circ\mathrm{C}$. where $A=1-T-R$. Residual moisture on the AR coatings. We see two absorption features at $\sim1390 nm$ and $\sim2200 nm$, which indicates the presence of moisture. This can be attributed to the porous nature of alumina grass. To avoid this issue we suggest baking the samples after blooming. It is also been observed that blooming the samples in DI water for longer period of time reduces reflectanceYang.
• Figure 5: A) Back reflection separated Measured Reflectance (red) and Simulated Reflectance profile using optimized parameters of the bump function. B) Ratio of Blue over red curve from A). The Ratio is nearly one for IR and red wavelengths, while it deviates from 1 for blue wavelengths, indicating that the optimization yields an excellent match in the red and less so in the blue. This is likely due to scattering of blue light caused by the irregular structure of alumina grass, which is present in the data, but not accounted for in the simulations.