Interstellar dust makes up about one percent of the visible mass of the Milky Way and absorbs as much as half the ultraviolet light emitted by hot stars.
How those grains form, what they are made of, and how they evolve are still open problems. Polarimetry is one of the few tools that can answer them directly.
Polarization
Dust grains interact with the electric field of the light and absorb and
re-direct it — depends sensitively on the relative size of the grain and the wavelength of the light.
A typical interstellar grain consists of a core of rocky material (silicates) or graphite, surrounded by a mantle of ices.
Typical grain sizes are 10–8 to 10–7 meters. (1/100 to 1/10 of a micron; by contrast, human hair is about 10–200 microns wide.)
Carbon-rich grains (similar to soot) are diamagnetic and are not expected to align in most
environments, while the silicate grains (similar to sand) aligns. So the polarized signal and the total extinction probe
different populations. To use polarization to study the ISM we need to understand why, where, and how the grain are (and are not) aligned.
Why the ultraviolet
The wavelength dependence tells us about the (relative) grain sizes —to probe very small grains, we need short wavelengths!
PUFFINS can probe the smallest grain sizes — down to about 50 nanometers, or roughly the size of the common cold virus — which are invisible to optical and infrared telescopes.
The gap
Twenty-eight sightlines in thirty years
The Wisconsin Ultraviolet Photo-Polarimeter Experiment (WUPPE) flew on the flew on two NASA Space Shuttle missions: ASTRO-1 and ASTRO-2 (1990 and 1995), and
the Hubble Faint Object Spectrograph added a handful more.
Between them they measured UV polarization along 28 lines of sight. No dedicated UV polarimetry mission has flown since.
Those few measurements produced two genuinely surprising results: Enhanced UV polarization toward 25% of the targets and increased polarization in the 2175Å
extinction feature in 7% of the targets. Sample size was too small (~30 targets) to provide any conclusive test of the grain theory alignment.
Paradigm-shifting developments in the theory and modeling of interstellar grain alignment and the availability of comprehensive multi-band surveys for
target selection now allow conclusive testing of the origins and theoretical mechanisms of small-grain alignment.
UV (open circles) and optical (squares) data for HD 197770 reveal polarization in the 2175 Å
feature, as one of two known examples. The solid and dotted line shows the Serkowski fit, extrapolated into
the UV. The dashed curve adds a Drude function fit representing the 2175 Å feature Wolff et al. 64 . (Right:) A
model fit of the HD 197770 polarization curve, assuming a variably aligned small-grain population for a >
0.005μm, and partially aligned PAHs. Further observations are required to test these theoretical assumptions.
Credit: PUFFINS team.
Objective 1
The enhanced UV polarization
Measure the polarization curve below 3200 Å to establish the nature and
extent of small-grain alignment, and how often it produces the enhanced
signal seen previously.
Objective 2
The 2175 Å feature
Determine when the strongest interstellar extinction feature is polarized
and how prevalent that is — the most direct constraint available on what
carries it.
Objective 3
The survey
Build a well-defined sample across a wide range of extinction, metallicity,
depletion level and radiation environment, with targets in both Magellanic
Clouds.
A: Target and calibration star sample for PUFFINS are shown in galactic coordinates. The diamond
markers for the target candidates correspond to those stars that have been observed before by WUPPE.
B: Location/source of the target candidates chosen to address the three SOs
Credit: Anche et al. 2025, SPIE.
The targets
To achieve PUFFINS’s science objectives, the mission will require a substantial number of lines of sight
with [a] varying extinctions and distances, [b] depletion levels, and [c] ISM radiation field characteristics.
The target sample contains a mixture of O, B, and early-A spectral-type stars with V < 8 mag.
As the polarization observations are limited by the errors introduced because of instrumental effects,
regular calibrations are required to estimate the accurate value of the polarization. These calibrations
are performed by observing well-established targets from the polarized and unpolarized standard catalogs.
What comes after
Combined with ground-based optical spectropolarimetry carried out using SPOL and FiberPOL, the PUFFINS data give
a continuous polarization curve from the UV through the optical. The survey
also serves as a testbed for UV spectropolarimetry on the Habitable Worlds
Observatory, and helps characterise the Galactic dust foreground that limits
searches for primordial B-modes in the cosmic microwave background.
Publications
Papers and proceedings
Optical design and polarimetric performance of a SmallSat
UV polarimeter to study interstellar dust: PUFFINS
Anche et al., SPIE, 2025 · arXiv:2509.02994
