Pollux test bench: from NUV to FUV polarimetric measurements

Pollux is a high-resolution spectropolarimeter proposed by an European consortium for HWO. The current design of Pollux features four spectropolarimetric channels, three of which are in the UV range. For the near-UV (NUV) [236-472 nm] and mid-UV (MUV) [118-236 nm] channels, the polarimeters consist of waveplates and prisms made of MgF2, a birefringent material. However, no such birefringent material can be used for the far-UV (FUV) channel [100-123 nm]. Therefore, the polarimeter for this FUV channel is composed solely of mirrors in an innovative assembly. In this talk, we aim to detail the architecture of the test bench that will allow us to validate the performance of these different polarimeters, as part of the HWO GOMaP. Given that we are working in the vacuum ultraviolet (VUV) range, the test bench operates in a vacuum chamber in a clean room. We will discuss the adaptable architecture of the bench based on wavelength and the measurement methodology that we will implement to test if the polarimeters achieve the precision of $10^{-3}$ required for the Pollux instrument. With this test bench, we will successfully increase the Technology Readiness Level (TRL) of UV spectropolarimeters and, for the first time, develop a means to test FUV spectropolarimetry.

💡 Research Summary

The paper presents the design, construction, and validation methodology of a vacuum ultraviolet (VUV) test bench intended to qualify the polarimetric performance of the proposed high‑resolution spectropolarimeter “Pollux,” which is being considered for the Habitable Worlds Observatory (HWO). Pollux is planned to cover a very broad spectral range (100 nm – 1800 nm) with a resolving power up to R ≈ 100 000, and it includes three ultraviolet channels: near‑UV (NUV, 236‑472 nm), mid‑UV (MUV, 118‑236 nm), and far‑UV (FUV, 100‑123 nm). While the NUV and MUV channels can employ conventional birefringent optics made from MgF₂ (waveplates and prisms), the FUV region lacks any suitable transmissive birefringent material. Consequently, the FUV polarimeter is built exclusively from reflective optics (mirrors) arranged in an innovative configuration that produces the required phase retardance through controlled reflection angles and coating properties.

To verify that both the conventional MgF₂‑based polarimeters and the novel mirror‑based FUV polarimeter meet the stringent polarimetric accuracy requirement of 10⁻³, the authors have constructed a modular test bench that can be re‑configured for the MUV‑NUV or the FUV regime. The bench is divided into five functional blocks:

• Block A & B – Light source and beam conditioning: An unpolarized deuterium lamp (or a deuterium plasma source for the FUV) is coupled through MgF₂ and CaF₂ lenses into an integrating sphere to ensure complete depolarization. A 1 µm pinhole then creates a stable, collimated beam.

• Block C – Polarisation state generator: A Babinet‑Soleil compensator (three‑plate variable retarder) combined with a Wollaston prism allows the generation of any linear or circular state. The compensator is calibrated at 632.8 nm, achieving a zero‑retardance positioning accuracy of 0.01 mm (≈ 4.4 × 10⁻⁵ mm thickness uncertainty).

• Block D – Device under test (DUT): For the NUV/MUV channels this is the MgF₂ waveplate‑prism assembly; for the FUV channel it is the mirror‑only polarimeter. The mirror assembly is mounted on a five‑axis stage with an additional fine‑adjustment piezo, enabling sub‑arcsecond angular alignment.

• Block E – Spectro‑polarimetric analysis: Light exiting the DUT is dispersed by a dual‑grating echelle spectrograph covering 115‑300 nm. The detector is a Teledyne CIS120 CMOS sensor, whose PCB is coated with Mapsil to ensure vacuum compatibility. Simulations show distinct spectral orders and intensity distributions that will be used to extract Stokes parameters.

All components operate inside a stainless‑steel vacuum chamber that reaches ~2 × 10⁻⁶ mbar within 12 minutes of pump‑down. Molecular and particulate contamination monitors are employed continuously. The chamber is mounted on rails, allowing rapid access to internal optics. Alignment of critical subsystems is performed with high‑precision tools: a QWLSI Phasics wave‑front sensor (λ = 632.8 nm) is used to align the collimator mirrors to within 30 arcseconds and 1.5 µm positional tolerance; nine vacuum‑compatible piezoelectric motors provide fine adjustments for all degrees of freedom.

The authors report successful preliminary alignments that satisfy the error budgets required for 10⁻³ polarimetric precision. The modular nature of the bench permits rapid re‑configuration between the MgF₂‑based and mirror‑based polarimeters, thereby raising the Technology Readiness Level (TRL) of both approaches. The FUV configuration, in particular, represents the first experimental platform capable of validating a purely reflective polarimeter in the 100‑120 nm band.

In conclusion, the test bench constitutes a critical step toward qualifying Pollux’s UV polarimetric capabilities for a future space mission. By providing a controlled, repeatable, and high‑precision environment, it enables the systematic characterization of both traditional birefringent and novel reflective polarimeters, paving the way for their eventual integration on the HWO platform and opening new scientific opportunities in far‑UV spectropolarimetry.