A Diverse, Overlooked Population of Type Ia Supernovae Exhibiting Mid-infrared
Signatures of Delayed Circumstellar Interaction
Geoffrey Mo
1
, Kishalay De
1
,
11
, Eli Wiston
2
, Nayana A. J.
2
, Raffaella Margutti
2
,
3
, Danielle Frostig
4
,
Jesper Sollerman
5
, Yashvi Sharma
6
, Takashi J. Moriya
7
,
8
,
9
, Kevin B. Burdge
1
, Jacob Jencson
10
,
Viraj R. Karambelkar
6
, and Nathan P. Lourie
1
1
MIT Kavli Institute for Astrophysics and Space Research, 70 Vassar St., Cambridge, MA 02139, USA;
gmo@mit.edu
2
Department of Astronomy, University of California, Berkeley, Berkeley, CA 94720-3411, USA
3
Department of Physics, University of California, Berkeley, 366 Physics North MC 7300, Berkeley, CA 94720, USA
4
Center for Astrophysics
|
Harvard & Smithsonian, 60 Garden Street, Cambridge, MA 02138, USA
5
The Oskar Klein Centre, Department of Astronomy, Stockholm University, AlbaNova, SE-10691 Stockholm, Sweden
6
Division of Physics, Mathematics and Astronomy, California Institute of Technology, Pasadena, CA 91125, USA
7
National Astronomical Observatory of Japan, National Institutes of Natural Sciences, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan
8
Graduate Institute for Advanced Studies, SOKENDAI, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan
9
School of Physics and Astronomy, Monash University, Clayton, VIC 3800, Australia
10
IPAC, Mail Code 100-22, Caltech, 1200 E. California Blvd., Pasadena, CA 91125, USA
Received 2024 October 24; revised 2025 January 28; accepted 2025 January 29; published 2025 February 17
Abstract
Type Ia supernovae
(
SNe Ia
)
arise from the thermonuclear explosions of white dwarfs in multiple-star systems. A
rare subclass of SNe Ia exhibit signatures of interaction with circumstellar material
(
CSM
)
, allowing for direct
constraints on companion material. While most known events show evidence for dense nearby CSM identi
fi
ed via
peak-light spectroscopy
(
as SNe Ia-CSM
)
, targeted late-time searches have revealed a handful of cases exhibiting
delayed CSM interaction with detached shells. Here we present the
fi
rst all-sky search for late CSM interaction in
SNe Ia using a new image subtraction pipeline for mid-infrared data from the NEOWISE space telescope.
Analyzing a sample of
≈
8500 SNe Ia, we report evidence for late-time mid-infrared brightening in
fi
ve previously
overlooked events spanning subtypes SNe Iax, SNe Ia-91T, and super-Chandra SNe Ia. Our systematic search
doubles the known sample and suggests that
0.05% of SNe Ia exhibit mid-infrared signatures of delayed CSM
interaction. The mid-infrared light curves ubiquitously indicate the presence of multiple
(
or extended
)
detached
CSM shells located at
10
16
–
10
17
cm, containing 10
−
6
to 10
−
4
M
e
of dust, with some sources showing evidence
for new dust formation, possibly within the cold, dense shell of the ejecta. We do not detect interaction signatures
in spectroscopic and radio follow-up; however, the limits are largely consistent with previously con
fi
rmed events
given the sensitivity and observation phase. Our results highlight that CSM interaction is more prevalent than
previously estimated from optical and ultraviolet searches and that mid-infrared synoptic surveys provide a unique
window into this phenomenon.
Uni
fi
ed Astronomy Thesaurus concepts:
Type Ia supernovae
(
1728
)
;
Infrared excess
(
788
)
;
Supernovae
(
1668
)
;
Common envelope binary stars
(
2156
)
;
Interacting binary stars
(
801
)
1. Introduction
It is well established that Type Ia supernovae
(
SNe Ia
)
arise
from the thermonuclear explosions of white dwarfs
(
WDs
)
,
triggered by interaction with a companion star
(
D. Maoz et al.
2014
)
. Not only do SNe Ia serve as standardizable candles
(
D. Branch & G. A. Tammann
1992
; M. M. Phillips
1993
;
D. Kasen & S. E. Woosley
2007
; S. Dhawan et al.
2018
;
A. Avelino et al.
2019
)
, but they also probe a common end
point in binary stellar evolution
(
B. Wang & Z. Han
2012
;
K. A. Postnov & L. R. Yungelson
2014
; A. J. Ruiter
2020
;
Z.-W. Liu et al.
2023
)
, reveal the fates of mHz gravitational-
wave sources
(
S. Toonen et al.
2012
; A. Rebassa-Mansergas
et al.
2019
; N. Karnesis et al.
2021
; V. Korol et al.
2024
)
, and
provide insights into the universal chemical and dust budget
(
C. Kobayashi & K. Nomoto
2009
; T. Nozawa et al.
2011
;
F. Lach et al.
2020
; L. Wang et al.
2024
)
. However, the nature
of the WD
(
mass, composition
)
and that of the companion
(
degenerate or not; accretion via mergers or mass transfer
)
,as
well as the explosion mechanism
(
detonation or de
fl
agration
)
,
remain unresolved, with clear evidence for multiple channels
(
R. Pakmor et al.
2013
; C. Ashall et al.
2016
; A. Polin et al.
2019
; M. Bulla et al.
2020
; A. A. Hakobyan et al.
2020
)
.
Supporting the diverse nature of progenitors, a rare subclass
of SNe Ia
—
known as SNe Ia-CSM
(
0.2% of SNe Ia;
J. M. Silverman et al.
2013b
; Y. Sharma et al.
2023
)
—
exhibit
strong interaction of the expanding ejecta
12
with dense nearby
circumstellar material
(
CSM
)
. These events are typically
identi
fi
ed via strong narrow emission lines
(
commonly H
α
)
in spectroscopy acquired at or near peak light
(
Y. Sharma et al.
2023
)
, similar to Type IIn SNe
(
N. Smith
2017
)
. This H-rich
material has been ubiquitously attributed to mass supplied by a
The Astrophysical Journal Letters,
980:L33
(
13pp
)
, 2025 February 20
https:
//
doi.org
/
10.3847
/
2041-8213
/
adaf92
© 2025. The Author
(
s
)
. Published by the American Astronomical Society.
11
MIT Kavli Institute Fellow.
Original content from this work may be used under the terms
of the
Creative Commons Attribution 4.0 licence
. Any further
distribution of this work must maintain attribution to the author
(
s
)
and the title
of the work, journal citation and DOI.
12
We note that interaction with very nearby CSM can also be detected as
early-time excesses in optical light curves or high-velocity features
(
e.g.,
U. M. Noebauer et al.
2016
; B. W. Mulligan & J. C. Wheeler
2017
;T.J.Moriya
et al.
2023
)
; however, we do not discuss them here since existing observations do
not conclusively rule out other possibilities
(
e.g., M. J. Childress et al.
2014
;
M. Deckers et al.
2022
)
.
1
nondegenerate companion, allowing for direct constraints on its
composition and con
fi
guration. The proximity
(
10
15
cm
)
and
large mass
(
0.1
M
e
)
of the CSM have been explained via
binary progenitors such as a WD merging with the core of an
asymptotic giant branch
(
AGB
)
star inside
(
or shortly after the
ejection of
)
the common envelope
(
CE; i.e., the core-
degenerate scenario; A. V. Tutukov et al.
1992
; M. Livio &
A. G. Riess
2003
; A. Kashi & N. Soker
2011
; I. Ablimit
2021
)
,
or even as a result of dynamical mass transfer from a main-
sequence star
(
Z. Han & P. Podsiadlowski
2006
)
. Since
systematic searches
(
J. M. Silverman et al.
2013b
; Y. Sharma
et al.
2023
)
have relied on peak-light spectroscopy or peculiar
undulations in early-time light curves to identify candidates for
SNe Ia-CSM, these are biased against SNe Ia with distant CSM
shells, where signatures of interaction would only be detectable
at later phases
(
Y. Sharma et al.
2023
; M. M. Phillips et al.
2024
)
.
B. Dilday et al.
(
2012
)
presented the
fi
rst example of an
SN Ia
(
of the 1991T subclass; see S. Blondin et al.
2012
and
S. Taubenberger
2017
, for a review of SN Ia subtypes
)
that
transitioned to an SN Ia-CSM at late phases
(
>
40 days after
peak
)
, due to the delayed CSM interaction in PTF 11kx.
13
The
multiple, detached CSM shells detected in this object were
suggested to arise from a symbiotic nova progenitor
(
Z. Han &
P. Podsiadlowski
2004
)
, although N. Soker et al.
(
2013
)
have
argued for a core-degenerate scenario where the envelope was
ejected shortly before the explosion. Using a Hubble Space
Telescope
(
HST
)
ultraviolet photometric survey of 72 SNe,
M. L. Graham et al.
(
2019
)
reported evidence for late CSM
interaction in SN 2015cp, which was initially classi
fi
ed as an
SN Ia 91T
–
like object. A search for similar emission in the
Galaxy Evolution Explorer
(
GALEX
)
archive yielded no
additional detection
(
L. O. Dubay et al.
2022
)
. Late-time
spectroscopic signatures of interaction
(
via H
α
emission
)
have
also been reported in SN 2016jae
(
N. Elias-Rosa et al.
2021
)
,
SN 2018cqj
(
J. L. Prieto et al.
2020
)
, and SN 2018fhw
(
J. A. Kollmeier et al.
2019
; P. J. Vallely et al.
2019
)
, while
E. C. Kool et al.
(
2023
)
recently reported the
fi
rst case of an
SN Ia interacting with He-rich CSM. While systematic efforts
have been previously made to identify potential candidates for
late interaction with optical photometry alone
(
J. H. Terwel
et al.
2024
)
, developing a complete census by either continued
spectroscopic monitoring or UV observations is challenging on
large scales owing to cost.
The mid-infrared
(
MIR
)
bands provide an independent and
powerful approach to probing CSM interaction and associated
dust formation
(
T. Szalai et al.
2019
)
. Instead of directly
detecting the high-energy X-ray or UV radiation produced by
the shock interaction with the CSM, MIR observations detect
emission from warm dust, which can either be preexisting
around the progenitor or be produced in the interaction region
after explosion
(
O. D. Fox et al.
2011
; T. Szalai
2013
)
. When
SN ejecta interact with surrounding CSM, dust can be heated
by shock radiation, producing luminous late-time MIR
emission
(
O. D. Fox et al.
2015
)
. MIR brightening episodes
therefore provide a unique tool to detect CSM interaction at late
phases, but systematic searches remain limited without
synoptic MIR capabilities. C. Myers et al.
(
2024
)
have recently
reported a systematic search for late MIR emission in core-
collapse SNe from NEOWISE data
(
E. L. Wright et al.
2010
;
A. Mainzer et al.
2014
)
, demonstrating the potential of all-sky,
slow-cadence surveys in revealing the demographics of CSM
interaction.
In this Letter, we report the identi
fi
cation of an overlooked
population of recent SNe Ia exhibiting delayed CSM interac-
tion. Capitalizing on the rapid growth of large SN samples from
all-sky time-domain optical surveys such as the Zwicky
Transient Facility
(
ZTF; E. C. Bellm et al.
2018
)
, ATLAS
(
J. L. Tonry et al.
2018
)
, ASAS-SN
(
C. S. Kochanek et al.
2017
)
, Pan-STARRS
(
K. C. Chambers et al.
2016
)
, and Gaia
(
S. T. Hodgkin et al.
2021
)
, we present a complete search of
NEOWISE data for MIR excesses using a new image
subtraction pipeline. We begin by characterizing the MIR
behavior of known SNe Ia with delayed CSM interaction and a
search for new similar objects in Section
2
. We use the MIR
emission together with follow-up observations to constrain the
amount and con
fi
guration of the CSM shells in Section
3
.We
conclude by discussing the implications for this overlooked
population in Section
4
. Throughout we assume
Planck18
cosmology
(
Planck Collaboration et al.
2020
)
.
2. Observations
2.1. MIR Behavior of Known SNe Ia with Late CSM Interaction
Using data from the infrared Spitzer Space Telescope
(
M. W. Werner et al.
2004
)
, O. D. Fox & A. V. Filippenko
(
2013
)
reported luminous, brightening late-time MIR emission
in SNe Ia-CSM 2002ic and 2005gj, suggesting renewed shock
interaction with CSM shells. Subsequently, T. Szalai et al.
(
2019
,
2021
)
and Y. Sharma et al.
(
2023
)
also con
fi
rmed
consistently bright late-time MIR emission in larger samples of
SNe Ia-CSM based on Spitzer and NEOWISE observations,
largely consistent with preexisting dust heated by ongoing
CSM interaction. A similar, luminous MIR excess was also
reported in the case of the nearby SN Iax SN 2014dt
(
O. D. Fox
et al.
2016
; J. E. Jencson et al.
2019
)
, although its origin has
been debated
(
as arising from circumstellar dust or possibly a
bound remnant; R. J. Foley et al.
2016
)
. Using a comprehen-
sive spectral sequence and NEOWISE observations, L. Wang
et al.
(
2024
)
provided evidence for multiple dust shells and new
late-time dust formation in the SN Ia-CSM SN 2018evt. To
understand whether MIR searches can be fruitful to look for
delayed CSM interaction in SNe Ia, we
fi
rst investigate
previously known events in NEOWISE data. The NEOWISE
survey
’
s all-sky coverage, MIR photometric bands
(
W
1
≈
3.4
μ
m,
W
2
≈
4.6
μ
m
)
, long baseline
(
≈
13 yr
)
, regular cadence
(
≈
6 months
)
, and sensitivity
(
≈
20 AB mag
)
are ideal attributes
for such a search. When combined with a new difference
photometry pipeline
(
De et al. 2025, in preparation
)
with
unWISE coadded images
(
D. Lang
2014
; A. M. Meisner et al.
2018
)
and a ZOGY-based subtraction algorithm
(
B. Zackay
et al.
2016
; K. De et al.
2020
)
adept at recovering faint
transients on bright hosts, this MIR data set has already proven
to be powerful for core-collapse SNe
(
C. Myers et al.
2024
)
.
NEOWISE observations of PTF11kx begin
≈
3.5 yr after
peak light, exhibiting luminous emission detected for an
additional 2 yr
(
also previously reported with Spitzer data in
M. L. Graham et al.
2017
)
. Similarly, the NEOWISE analysis
of SN 2015cp reveals a long-lasting MIR light curve.
14
We
show a comparison of these light curves with SNe Ia-CSM in
13
C. E. Harris et al.
(
2018
)
name these SNe Ia;n to indicate the late-time
emergence of narrow
(
“
n
”
)
emission lines.
14
M. Thévenot et al.
(
2021a
)
previously also reported the NEOWISE
detection of SN 2015cp.
2
The Astrophysical Journal Letters,
980:L33
(
13pp
)
, 2025 February 20
Mo et al.
Figure
1
and their individual light curves in Figure
A1
.
E. C. Kool et al.
(
2023
)
reported luminous MIR emission from
SN 2020eyj that exhibited delayed interaction with a He-rich
CSM shell. In contrast, NEOWISE light curves of the
subluminous SNe Ia that show late-time H
α
emission
(
SN 2016jae, SN 2018cqj, and SN 2018fhw
)
, as well as the
candidates identi
fi
ed from late-time optical photometry
(
SN 2018grt, SN 2019ldf, and SN 2020tfc; J. H. Terwel et al.
2024
)
, do not show luminous MIR emission comparable to the
other events. We discuss possible reasons in Section
3.3
.As
shown in Figure
1
, many SNe Ia-CSM and SNe Ia with delayed
CSM interaction show luminous late-time MIR emission,
distinct from normal SNe Ia, making MIR searches a promising
avenue for identifying this phenomenon.
2.2. Systematic Search in NEOWISE for Missed Events
We produced NEOWISE MIR difference imaging light
curves for
≈
8500 SNe Ia from 2011 to 2022
15
reported to the
IAU Transient Name Server.
16
These light curves were visually
inspected to search for late-time MIR emission associated with
the SN. We
fi
lter the light curves by
(
i
)
rejecting candidates
with detections prior to the optical onset of the SN
(
to remove
contamination from active galactic nucleus activity
)
or obvious
subtraction artifacts and
(
ii
)
requiring at least two epochs of
NEOWISE detections
(
≈
1yr
)
and rising emission in at least
one band, to conclusively rule out excess MIR emission arising
purely from new dust formation or an SN light echo without
added power from shock interaction and to eliminate detections
deemed to only be associated with photospheric emission from
the optical SN. Most of the identi
fi
ed candidates were known
SNe Ia-CSM; we do not discuss them here, as their light curves
have been presented in Y. Sharma et al.
(
2023
)
. Table
A1
summarizes the cuts at each
fi
ltering stage. Notably, these
selection criteria resulted in the identi
fi
cation of
fi
ve candidates
that exhibit luminous late-time MIR emission despite not being
previously associated with CSM interaction.
17
The candidates span multiple Type Ia subclasses:
SN 2017fra
(
originally classi
fi
ed as an SN Ia-91T
)
,
SN 2017hyn
(
SN Ia-91T
)
, SN 2020yex
(
SN Ia-91T
)
,
SN 2021jun
(
SN Iax
)
, and SN 2022bbt
(
super-Chandra SN
Ia
)
.
18
We summarize these SNe in Table
1
, further describe
each in Appendix
A
, and show reference, science, and
difference image cutouts in Figure
A2
. While half of our
sample are SNe Ia-91T
–
like, strengthening the established
connection between SNe Ia-CSM and SNe Ia-91T
(
G. Leloudas
et al.
2015
)
, we report the
fi
rst detection of late-time MIR
brightenings in a super-Chandra SN Ia,
19
and we report the
second detection of an MIR excess in an SN Iax. We show a
comparison of their MIR light curves in Figure
1
and the
individual optical and MIR light curves of these events in
Figure 1.
The phase space of MIR light curves for SNe Ia. We show NEOWISE
W
2 light curves for the SNe Ia showing delayed CSM interaction from our sample in
the colored stars. For comparison, we also plot previously published SNe Ia with delayed CSM interaction, including PTF 11kx, SN 2015cp, and SN 2020ey
j
(
blue
circles, squares, and triangles, respectively
)
. Plotted with blue diamonds is the Spitzer IRAC Ch2 light curve of SN 2014dt, an SN Iax that exhibits an analogous rise
and fall to the interacting SNe Ia, but at a much lower luminosity
(
O. D. Fox et al.
2016
; J. E. Jencson et al.
2019
)
. The SNe Ia-CSM, for which we show selected
W
2
light curves in dark red
(
these include SN 2016iks, SN 2017eby, SN 2017hzw, SN 2018crl, SN 2018evt, SN 2018gkx, SN 2019agi, SN 2020aekp
)
, display similar
behavior to most of the SNe Ia with delayed CSM interaction. Finally, we show Spitzer Ch2 light curves of normal SNe Ia from T. Szalai et al.
(
2019
)
in magenta.
15
The 2022 cutoff is governed by the availability of unWISE data products.
16
https:
//
www.wis-tns.org
17
SN 2020mvp
(
J. Tonry et al.
2020
)
also emerged as a candidate, but further
astrometric inspection of the NEOWISE imaging revealed that the MIR
emission was likely from temporally coincident nuclear activity.
18
M. Thévenot et al.
(
2021a
)
and M. Thévenot
(
2021b
)
have previously noted
MIR detections for SN 2017hyn, SN 2017fra, and SN 2020yex; N. Jiang et al.
(
2021
)
previously analyzed NEOWISE observations of SN 2017fra in their
search for MIR
fl
ares in nearby galaxies.
19
The super-Chandra SNe Ia SN 2012dn
(
T. Nagao et al.
2017
)
and
SN 2022pul
(
M. R. Siebert et al.
2024
)
have previously reported MIR
excesses, but the lack of a detected MIR brightening is consistent with SN light
echoes or new dust formation in cooling ejecta.
3
The Astrophysical Journal Letters,
980:L33
(
13pp
)
, 2025 February 20
Mo et al.
Figure
2
. Compared to normal SNe Ia, our newly identi
fi
ed
events have signi
fi
cant MIR excesses reaching luminosities of
W
2
≈−
22 Vega mag, lasting hundreds to thousands of days.
Of note is the MIR light curve of the SN Iax SN 2014dt, which
shows similar behavior to but is much less luminous than the
rest of the interacting SNe Ia; its late MIR rise has been
suggested to arise from newly formed dust, but it could also be
attributable to preexisting CSM or a bound remnant
(
O. D. Fox
et al.
2016
; R. J. Foley et al.
2016
; J. E. Jencson et al.
2019
)
.
2.3. Multiwavelength Follow-up Observations
We performed optical spectroscopic follow-up of two of the
youngest events:
20
SN 2020yex
(
z
=
0.087
)
and SN 2022bbt
(
z
=
0.049
)
. We also performed radio follow-up of
SN 2022bbt. Optical spectroscopic observations were carried
out using a Fast Turnaround Program on the Gemini North and
Gemini South telescopes
(
Program IDs GN-2024B-FT-102
/
GS-2024B-FT-102; PI: Mo
)
, while radio observations were
carried out using the Karl G. Jansky Very Large Array
(
VLA
)
under the Director
’
s Discretionary Program
(
VLA
/
24A-498;
PI: Nayana AJ
)
. In addition, we obtained near-infrared
(
NIR
)
spectroscopy of SN 2020yex and NIR imaging of SN 2017fra
using the Magellan Baade telescope and NIR imaging of
SN 2021jun with WINTER
(
N. P. Lourie et al.
2020
; D. Frostig
et al.
2024
)
. Further details about the observations and data
reduction can be found in Appendix
B
. We do not detect
photometric or spectroscopic signatures of CSM interaction in
any of our optical
/
NIR observations, while the VLA observa-
tions also did not detect a radio counterpart. We discuss the
implications of these nondetections in Section
3.3
.
3. Origin of the MIR Emission
We have identi
fi
ed a sample of
fi
ve SNe Ia showing clear
evidence for late-time MIR brightenings. While fading excess
late-time MIR emission can be caused by a light echo from the
SN or dust formation in the cooling ejecta, rebrightening in the
MIR light curve indicates the presence of an internal energy
source
(
i.e., high-energy radiation in the form of UV or X-ray
photons
)
that is heating preexisting or newly formed dust
(
O. D. Fox et al.
2010
,
2011
)
. The lack of observed UV or X-ray
emission from these events is unsurprising given the nonexis-
tence of all-sky UV
/
X-ray surveys and the expected luminosity
of the shock interaction radiation. For the observed peak
W
2
MIR luminosity of
∼
10
42
erg s
−
1
, the equivalent peak X-ray
fl
ux for sources at
200 Mpc would be
∼
10
−
13
erg cm
−
2
s
−
1
,
which is far below the detection threshold of existing X-ray all-
sky surveys
(
∼
10
−
11
erg cm
−
2
s
−
1
for MAXI
/
GSC, K. Hiroi
et al.
2013
;
∼
10
−
10
erg cm
−
2
s
−
1
for Swift
/
BAT, H. A. Krimm
et al.
2013
)
, and may be detectable only with pointed X-ray
follow-up
(
which does not exist for these sources
)
. Furthermore,
targeted X-ray observations of other SNe Ia have had limited
success owing to the rarity
(
one known event in SN 2012ca;
C. D. Bochenek et al.
2018
)
and low luminosity
(
L
X
10
37
erg s
−
1
for normal SNe Ia from R. Margutti et al.
2014
;
L
X
10
40
erg s
−
1
for SNe Ia with delayed CSM interaction from
V. V. Dwarkadas
2023
,
2024
)
of X-ray emission associated with
SNe Ia.
Given that known SNe Ia exhibiting delayed CSM interac-
tion
(
see Figure
1
)
also show late-time MIR brightenings, these
observations strongly suggest that the brightenings are powered
by the onset of CSM interaction, where the shock of the high-
velocity SN ejecta impacting existing CSM creates high-energy
radiation that is absorbed and reemitted by dust. The integrated
MIR energies shown in Table
1
span approximately 1%
–
10%
of the
∼
10
51
erg in kinetic energy released in an SN Ia owing to
the ef
fi
cient conversion of kinetic energy to radiative energy
during the shock; these shocks can convert up to
∼
50% of the
total kinetic energy into radiative energy for interacting core-
collapse SNe with similar CSM interaction
(
N. Smith
2017
)
.
The lack of spectroscopic signatures or optical light-curve
abnormalities at peak light indicate that this sample represents
an overlooked population of SNe Ia undergoing delayed CSM
interaction. In this section, we use the NEOWISE multiband
photometry to constrain the mass, temperature, and location of
the emitting dust to infer the CSM properties. We assume a
spherical con
fi
guration for the dust for simplicity, though our
limited observations cannot rule out complex geometries that
may affect the light curve or inferred parameters only up to
order-unity corrections. In addition to the
fi
ve SNe, we also
fi
t
the NEOWISE data from SN 2015cp, for comparison to a well-
studied event.
3.1. Dust Modeling
We
fi
rst
fi
t the NEOWISE photometry to a pure blackbody,
described by
()()
p
=
nn
F
r
D
BT
,1
bb
2
2
bb
where
F
ν
is the observed spectral
fl
ux density,
r
bb
is the
blackbody radius,
D
is the distance to the source, and
T
bb
is the
blackbody temperature. Given that the dust is not expected to
emit as a pure blackbody source and may be optically thin, the
derived
r
bb
is a lower limit to the true radius of the dust shell.
Assuming the dust to be optically thin, the dust mass and
Table 1
Summary of Our SNe Ia Showing MIR Signs of Delayed CSM Interaction
Name
Type
Host Morphology
Redshift
Distance
Peak
W
2 AB Magnitude
Integrated
W
2 Energy
(
Mpc
)(
erg
)
SN 2017fra
Ia-91T
Interacting spiral
0.03
136
16.8
7.9
×
10
49
SN 2017hyn
Ia-91T
Edge-on disk
0.053
244
18.5
5.4
×
10
49
SN 2020yex
Ia-91T
Blue spheroidal
0.087
425
19.3
2.3
×
10
49
SN 2021jun
Iax
Blue spheroidal
0.040
183
19.3
5.2
×
10
48
SN 2022bbt
Ia-SC
Face-on spiral
0.049
225
19.5
2.8
×
10
49
Note.
The integrated energy is calculated by integrating the detected
W
2
fl
uxes.
20
We did not follow up SN 2021jun, due to its proximity to its host nucleus,
which would make faint object spectroscopy challenging.
4
The Astrophysical Journal Letters,
980:L33
(
13pp
)
, 2025 February 20
Mo et al.