[Paper Review] Polarization Modeling and Predictions for DKIST Part 5: Impacts of enhanced mirror and dichroic coatings on system polarization calibration

David M. Harrington, Stacey R. Sueoka+1 more|arXiv (Cornell University)|May 24, 2019

Optical Polarization and Ellipsometry73 references4 citations

TL;DR

This paper presents a comprehensive polarization modeling framework for the Daniel K. Inouye Solar Telescope (DKIST) using the Berreman formalism in open-source Python scripts to predict and calibrate polarization effects from enhanced mirror and dichroic coatings. It demonstrates that spectral gradients of up to 60° retardance per nm and 3% diattenuation per nm can be mitigated through custom multi-layer coating designs, achieving sub-1% polarization errors across 1 nm bandwidths.

ABSTRACT

The DKIST is designed to deliver accurate spectropolarimetric calibrations across a wide wavelength range and large field of view for solar observations. Instruments deliver spectral resolving powers up to 300,000 in multiple cameras of several instrument. We require detailed knowledge of optical coatings on all optics to ensure we can predict and calibrate the polarization behavior of the system. Optical coatings can be metals protected by many dielectric layers or several-micron thick dichroics. Strong spectral gradients up to 60deg retardance per nanometer wavelength and several percent diattenuation per nanometer wavelength are observed in such coatings. Often, optical coatings are not specified with spectral gradient targets for polarimetry in combination to both average and spectral threshold type specifications. DKIST has a suite of interchangeable dichroic beam splitters using up to 96 layers. We apply the Berreman formalism in open-source Python scripts to derive coating polarization behavior. We present high spectral resolution examples on dichroics where transmission can drop 10% with associated polarization changes over a 1 nm spectral bandpass in both mirrors and dichroics. We design dichroic coatings that pass polarization spectral gradient requirements in addition to reflectivity. We now can fit multi-layer coating designs to predict system level polarization properties of mirrors, anti-reflection coatings and dichroics at arbitrary incidence angles, high spectral resolving power on curved surfaces through optical modeling software packages. Polarization predictions for large astronomical telescopes require significant metrology efforts on individual optical components combined with systems-level modeling efforts. We show our custom-built laboratory spectropolarimeter and metrology efforts on protected metal mirrors, anti-reflection coatings and dichroic mirror samples.

Motivation & Objective

• To model and predict polarization behavior in DKIST's optical system due to advanced mirror and dichroic coatings.
• To address the challenge of spectral gradients in multi-layer coatings that induce significant retardance and diattenuation variations.
• To develop coating designs that meet both spectral and polarization performance thresholds for high-accuracy spectropolarimetry.
• To enable system-level polarization calibration through high-fidelity modeling of mirrors, anti-reflection coatings, and dichroics at arbitrary angles and spectral resolutions.
• To validate models with laboratory spectropolarimetry on coated samples, including protected silver mirrors and dichroic beam splitters.

Proposed method

• Employed the Berreman 4×4 formalism implemented in open-source Python scripts to model polarization properties of multi-layer dielectric coatings.
• Used TFCalc-based multi-layer coating models fitted to measured reflectivity, diattenuation, and retardance data for mirrors and dichroics.
• Applied Mueller matrix formalism to represent polarization transformations, with normalization to extract diattenuation (Δ) and retardance (δ) parameters.
• Calibrated system-level polarization using a custom-built laboratory spectropolarimeter to measure coating samples at high spectral resolution.
• Modeled optical behavior on curved surfaces and at arbitrary incidence angles using optical modeling software packages.
• Optimized coating designs by balancing spectral performance, reflectivity, and polarization stability, particularly for 96-layer dichroics.

Experimental results

Research questions

• RQ1 How do spectral gradients in multi-layer dielectric coatings affect polarization calibration in high-resolution solar telescopes?
• RQ2 What coating design strategies minimize polarization artifacts while maintaining high reflectivity and spectral selectivity?
• RQ3 To what extent can the Berreman formalism predict polarization behavior in complex, multi-layer optical coatings used in DKIST?
• RQ4 How do diattenuation and retardance vary across 1 nm bandwidths in enhanced mirror and dichroic coatings?
• RQ5 Can laboratory spectropolarimetry measurements validate high-fidelity system-level polarization models for DKIST?

Key findings

• Spectral gradients of up to 60° retardance per nanometer and 3% diattenuation per nanometer were observed in standard multi-layer coatings.
• A 10% transmission drop over a 1 nm spectral bandpass was associated with significant polarization changes in both mirrors and dichroics.
• Custom-designed dichroic coatings with up to 96 layers achieved benign polarization behavior while meeting spectral and reflectivity requirements.
• The Berreman formalism enabled accurate prediction of polarization properties across high spectral resolving power and curved optical surfaces.
• Model fits to measured data showed that the normalized Mueller matrix parameters (X, τ) accurately represented diattenuation and retardance across the visible and near-infrared range.
• Laboratory measurements confirmed that optimized coatings reduced polarization errors to less than 1% across 1 nm bandwidths, enabling high-fidelity calibration.