Effect of H
2
S on the Near-infrared Spectrum of Irradiation Residue and Applications to
the Kuiper Belt Object
(
486958
)
Arrokoth
Ahmed Mahjoub
1
,
2
, Michael E. Brown
3
, Michael J. Poston
1
,
4
, Robert Hodyss
1
, Bethany L. Ehlmann
1
,
3
,
Jordana Blacksberg
1
, Mathieu Choukroun
1
, John M. Eiler
3
, and Kevin P. Hand
1
1
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA
2
Space Science Institute, 4765 Walnut Street, Suite B, Boulder, CO 80301, USA
3
Division of Geological & Planetary Sciences, California Institute of Technology, Pasadena, 91125, USA
4
Southwest Research Institute, San Antonio, TX, USA
Received 2021 April 15; revised 2021 May 11; accepted 2021 May 21; published 2021 June 17
Abstract
On 2019 January 1, the New Horizons spacecraft
fl
ew by
(
486958
)
Arrokoth, a small body in the Kuiper Belt that
is the most distant object ever visited by a spacecraft. A strong unidenti
fi
ed absorption band was observed in the
spectrum of Arrokoth at 1.8
μ
m. We report here experimental evidence suggesting that the near-infrared spectrum
of Arrokoth is indicative of sulfur-rich, tholin-like organic residue. The spectra of organic residues produced by
irradiating ice mixtures
“
with H
2
S
”
CH
3
OH:NH
3
:H
2
S:H
2
O
(
3:3:3:1
)
and
“
without H
2
S
”
CH
3
OH:NH
3
:H
2
O
(
3:3:1
)
were measured to study the effect of H
2
S. The
“
with H
2
S
”
sulfur-rich laboratory-synthesized organic residue
displays an absorption band at 1.8
μ
m that is absent in the spectrum of
“
without H
2
S
”
sample. This feature matches
the Arrokoth spectrum better than any other expected material. This suggests the past presence of H
2
S ice on the
surface of Arrokoth and a role for Kuiper Belt objects as a key S reservoir in the early solar system.
Uni
fi
ed Astronomy Thesaurus concepts:
Astrochemistry
(
75
)
;
Solar system
(
1528
)
;
Small Solar System bodies
(
1469
)
;
Laboratory astrophysics
(
2004
)
;
Trans-Neptunian objects
(
1705
)
;
Kuiper belt
(
893
)
1. Introduction
The Kuiper Belt contains primordial icy bodies believed to
be remnants from the reservoir of planetesimals that surrounded
the early solar system. The composition of these Kuiper Belt
objects
(
KBOs
)
, resulting from their starting chemistry and
their history of surface processing, can be read and deciphered
to gain more insight into the formation and evolution of the
solar system. Growing evidence suggests that H
2
S was a highly
abundant molecule in the presolar nebula. Recently, the Rosetta
mission to comet Churyumov
–
Gerasimenko demonstrated that
H
2
S is the
fi
fth most abundant molecule in the coma after H
2
O,
CO, CO
2
, and O
2
(
Rubin et al.
2019
)
. One important question
is the effect of H
2
S on the composition and spectroscopy of
organic residues, called tholins, that are believed to exist on the
surfaces of small icy bodies, including KBOs
(
Cruikshank et al.
2005
)
. Despite the importance of H
2
S and the potential of
sulfur chemistry to considerably affect the chemical reactivity
on the surfaces of airless bodies, little has been published
exploring this chemistry. Also, H
2
S residues have not been
directly detected on solar system bodies to date, although a
recent study has shown evidence for sulfur-bearing species on
Callisto
’
s leading hemisphere
(
Cartwright et al.
2020
)
.We
recently studied the role that H
2
S could play in the chemistry
(
Mahjoub et al.
2016
,
2017
)
, spectroscopy
(
Poston et al.
2018
)
,
and astrobiology
(
Mahjoub & Hodyss
2018
)
of small icy
bodies
—
particularly KBOs and Jupiter Trojans
(
Wong et al.
2019
)
—
and apply these results to Arrokoth.
Arrokoth is a member of the cold classical Kuiper Belt
(
CC-
KBO
)
, a collection of small bodies believed to have formed
beyond the orbit of Neptune and to have never experienced
large-scale change of orbit. Recent New Horizons spacecraft
data show widespread abundant CH
3
OH, a reddish darkening
agent
(
s
)
inferred to be a mixture of amorphous carbon and
tholins, low abundances of water and ammonia, and a strong
1.8
μ
m absorption, to date unexplained
(
Stern et al.
2019
;
Grundy et al.
2020
)
. This absorption feature appears as strong
as the other features identi
fi
ed in Grundy et al.
(
2020
)
. The
band was discussed previously but was not assigned to a
molecular identi
fi
cation.
Here, we present spectra of organic residues produced with
and without H
2
S to infer the effect of sulfur chemistry in the
near-infrared
(
NIR
)
spectral region. We analyze the New
Horizons data in concert with results from our laboratory
experiments and show that S-bearing refractory organics
derived from the products of irradiation of H
2
S on the surfaces
of KBOs are a plausible source of the 1.8
μ
m feature. We also
discuss implication for reservoirs of S in the early solar system.
2. Experimental Methodology
Two residue samples were produced by irradiation of ice
fi
lms under vacuum at 50 K. A tholin
fi
lm
“
without sulfur
”
was
produced from a starting ice of CH
3
OH:NH
3
:H
2
O
(
3:3:1
)
, and
a tholin
fi
lm
“
with sulfur
”
was produced from a starting ice
mixture of CH
3
OH:NH
3
:H
2
S:H
2
O
(
3:3:3:1
)
. These samples are
similar to our previous studies investigating spectroscopic and
chemical properties of H
2
S containing ices
(
Mahjoub et al.
2016
,
2017
; Poston et al.
2018
)
. Electron irradiation experi-
ments were carried out using the Icy Worlds Simulation
laboratory. A detailed description of the facilities and the
capabilities of this laboratory can be found in Hand & Carlson
(
2011
)
. The experimental setup consists of a high-vacuum
stainless steel chamber pumped by a Varian Turbo pump and
backed by oil-free pumps
(
pressure after overnight pumping
about 1
×
10
−
8
torr
)
. The ices were vapor-deposited on a
substrate attached to the cold
fi
nger of a closed-cycle helium
cryostat
(
ARS model DE-204
)
. An external manifold was used
to prepare gas mixtures prior to deposition. The ice
fi
lms were
The Astrophysical Journal Letters,
914:L31
(
4pp
)
, 2021 June 20
https:
//
doi.org
/
10.3847
/
2041-8213
/
ac044b
© 2021. The American Astronomical Society. All rights reserved.
1
grown by leaking the gas mixture into the chamber and forming
ices on the substrate, which was held at 50 K.
High-energy electrons
(
10 keV
)
were directed at the ice with
a typical beam current of 0.5
μ
A. All studied ices were
submitted to the same
fl
uence of electron energy
∼
2
×
10
21
eV cm
−
2
. Radiation
fl
uences were scaled to the
outer solar system based on the electron
fl
ux at 1 au, which was
deduced from values given in Bennett et al.
(
2013
)
. We found
that the total
fl
uence received by our ice samples corresponds to
a timescale of 0.2 Myr for an object at 5 au and 1.8 Myr
at 15 au.
As discussed in Mahjoub et al.
(
2016
)
, the chemical
composition of ice
fi
lms calculated using infrared spectra and
band strengths of deposited molecules is different from the gas
phase mixture. The compositions of the ices solid
fi
lms as
estimated from the respective column densities are:
CH
3
OH
–
NH
3
–
H
2
O
=
73:12:15 for the
“
without H
2
S
”
ice and
H
2
S
–
CH
3
OH
–
NH
3
–
H
2
O
=
7:35:17:41 for the
“
with H
2
S
”
sample. After irradiation for 19 hr at 50 K, samples were
warmed to 120 K at 0.5 K minute
−
1
rate and held there for 1
additional hour under continued electron irradiation. After the
electron irradiation was concluded, the samples were warmed
at 0.5 K minute
−
1
rate to 300 K. The resulting residue
fi
lms
were characterized at room temperature by specular re
fl
ectance
spectroscopy, using a Midac FTIR spectrometer. The residue
samples in both experiments were very thin
fi
lms on top of a
gold mirror, which made it dif
fi
cult to characterize their colors.
The
“
with H
2
S
”
residue appeared darker than the
“
without
H
2
S
”
residue. A network of cracks appeared on the
“
without
H2S
”
sample, while less-pronounced cracks were visible on the
“
with H
2
S
”
sample.
3. Results and Discussion
Figure
1
shows a comparison between the spectrum of the
“
without sulfur
”
and
“
with sulfur
”
residues. The spectrum of
“
without sulfur
”
residue is
fl
at in the 1.8
μ
m region, which
agrees with spectra reported in the literature of CNOH-based
tholins
(
Materese et al.
2015
)
and routinely used to model
spectra of small icy bodies, including Arrokoth
(
Grundy et al.
2020
)
. The spectrum of the
“
with sulfur
”
residue shows a
pronounced band at 1.8
μ
m. The depth of this band is
approximately 2% of the baseline and the width at half
maximum is around 100 nm. The sulfur-bearing residue is a
complex chemical material, composed of a plethora of
molecular species of CNOHS; therefore, the de
fi
nitive assign-
ment of this absorption band to a particular species is not
currently possible. Nevertheless, because it only occurs in the
H
2
S-bearing ice mixture, this absorption could be tentatively
assigned to a combination of
ν
(
SH
)
+
ν
(
CH
)
in a family of
molecules
/
functional groups rich in C
x
S
y
H
z
moieties. This
band could also be due to a combination between the C
=
O
stretching modes in -OCS- groups at 2050 cm
−
1
and the S
=
O
stretching mode around 1400 cm
−
1
:
ν
(
SO
)
+
2
ν
(
CO
)
. A better
characterization of the chemical composition of the residue,
and isotope labeling tests, are needed to better constrain the
sulfurous compounds or family of compounds responsible for
this band. Such in-depth characterization of the sulfur-bearing
residue is beyond of the scope of this Letter and will be
investigated in future studies. Future studies will also
investigate the evolution of the observed 1.8
μ
m band as a
function of thermal processing of the sample between 50 and
300 K. Spectra in the mid-infrared
(
MIR
)
show only limited
chemical evolution of the sample between 50 and 150 K
(
Mahjoub et al.
2016
)
.
Figure
2
shows a comparison between the spectrum of
Arrokoth and the spectrum of
“
with sulfur
”
residue. A good
match between the two spectra at the band 1.8
μ
m can be
visually noted. We
fi
t a linear mixture model including the
spectra of methanol,
“
with sulfur
”
residue and
“
without sulfur
”
residue
(
Figure
2
)
. Methanol and
“
without sulfur
”
residue
spectra were added because in Grundy et al.
(
2020
)
the best
fi
t
was obtained using a model incorporating methanol and tholin.
The model in Grundy et al.
(
2020
)
fi
t quite well the spectrum of
Arrokoth except for the 1.8
μ
m band. Adding the spectrum of
“
with sulfur
”
residue improves considerably the
fi
t, especially
the 1.8
μ
m feature
(
Figure
2
)
. This results in a very small
residual difference between the spectrum of Arrokoth and the
model, which is
fl
at around 1.8
μ
m con
fi
rming the good
spectral alignment between the bands. Adding the
“
without
sulfur
”
residue to the linear regression model improves the
fi
t
of the baseline of the spectrum but obviously is not able to
reproduce the 1.8
μ
m band. We cannot exclude the co-
existence of both
“
with sulfur
”
and
“
without sulfur
”
on the
surface of Arrokoth. We can speculate that, because H
2
Sis
much faster consumed by irradiation chemistry than methanol,
both
“
with sulfur
”
and
“
without sulfur
”
organics could be
formed on the surface of the KBO. It is important to clarify that
the model presented here cannot be used to determine the
relative abundances of each of the compounds. The model-
fi
tting coef
fi
cient is given in the Figure
2
caption. A more
sophisticated model will need the refractive index of
“
with
sulfur
”
residue, a measurement that is not available to date.
Figure 1.
Comparison between spectra of the
“
without sulfur
”
(
red
)
and
“
with
sulfur
”
(
black
)
residues produced by irradiation of ices CH
3
OH:NH
3
:H
2
O
(
3:3:1
)
and CH
3
OH:NH
3
:H
2
S:H
2
O
(
3:3:3:1
)
, respectively.
2
The Astrophysical Journal Letters,
914:L31
(
4pp
)
, 2021 June 20
Mahjoub et al.
The strong bands at 2.27 and 2.34
μ
m were previously
assigned to methanol
(
Grundy et al.
2020
)
. As discussed in
Grundy et al.
(
2020
)
the 1.8
μ
m feature could not be assigned
to any of the simple molecular ices or organic tholins made of
CNOH atoms. Methane, for example, would show additional
bands around this region. Pure H
2
S ice can also be excluded as
origin of the 1.8
μ
m feature because the absorption bands of
H
2
S are located at 1.98 and 2.02
μ
m
(
Fathe et al.
2006
)
and no
bands are observed at 1.8
μ
m
(
Waggener et al.
1969
)
.By
screening spectra from USGS spectral library of most common
minerals, we were able to rule them out. Indeed, while many
minerals have a broad band between 1.9 and 2.0
μ
m, none of
them has the 1.8
μ
m feature. The same arguments eliminate the
orthorhombic sulfur, S
8
. Thus, we propose that a sulfur-rich
residue, produced by irradiation of ice containing H
2
S, is the
source of the 1.8
μ
m feature. This scenario is rather plausible as
the CC-KBOs are believed to have formed and resided beyond
30 au, where H
2
S is stable for the time needed for the formation
of organic crust of heteropolymers
(
Wong & Brown
2016
)
.
The organic layer that is formed as a result of surface
processing on small icy bodies, including KBOs, by photons
and energetic particles is believed to be responsible for the
spectral reddening observed in many of these bodies, including
Arrokoth. The spectral slope in the NIR and visible spectral
regions is a function of the chemical composition of the initial
ices at the surface. As a consequence, spectrally intriguing
characteristics of small bodies, such as the color bimodality
observed in KBOs, Jupiter Trojans, and Centaurs, could be
linked to the initial composition and irradiation history of these
objects. To push this idea to more testable hypothesis, Wong &
Brown
(
2016
)
demonstrated that the sublimation line of H
2
Sis
located within the belt of primordial planetesimals. Therefore,
these objects would have been divided into two groups: those
that retained H
2
S for enough time to develop a sulfur-
containing organic crust, and those that did not. Laboratory
Figure 2.
Comparison between the NIR spectrum of
(
486958
)
Arrokoth and laboratory sulfur-containing
(
CNOHS
)
tholins at 1.8
μ
m. Top panel: spectrum of the
“
with sulfur
”
residue
(
scaled to Arrokoth re
fl
ectance
)
. Middle panel: averaged NIR spectrum of
(
486958
)
Arrokoth
(
data from Stern et al.
2019
)
, as measured by
LESIA spectrograph aboard New Horizons spacecraft
(
black line
)
compared to linear mixture model of methanol
+
“
with sulfur
”
residue
+
“
without sulfur
”
residue
(
red line
)
. Bottom panel: residual difference between Arrokoth spectrum and the model. Previously identi
fi
ed features at 2.27 and 2.34
μ
m are attributed to methanol
(
Grundy et al.
2020
)
. The linear model is obtained by
fi
tting a linear combination of the spectra of the three compounds. This model could not be used for quantitative
determination of relative abundances on the surface of Arrokoth. The model-
fi
tting coef
fi
cients are:
[
0.6, 3.77, 1.46
]
for
[
methanol,
“
with sulfur,
”“
with sulfur
]
and
intercept
=
−
0.94.
3
The Astrophysical Journal Letters,
914:L31
(
4pp
)
, 2021 June 20
Mahjoub et al.
irradiation of ices with and without H
2
S supported the different
NIR spectral reddening after irradiation with high
fl
uence of
electron irradiation: reddening slope is more pronounced in
samples containing H
2
S
(
Poston et al.
2018
)
and these samples
have the 1.8
μ
m feature observed on Arrokoth. Further
rigorous comparison to telescopic and spacecraft spectra will
need laboratory measurements of optical constants for ice
tholins with and without sulfur, requiring additional exper-
imental work.
A broader implication of the organosulfur residue is the role
it could play as a reservoir for sulfur in molecular clouds. The
abundance of sulfur in dense clouds and circumstellar regions
is only 0.1% of its cosmic abundance
(
Tieftrunk et al.
1994
)
.
Recent chemical models attempted to include sulfur residue in
the modeling of the hot molecular cores
(
Woods et al.
2015
)
,
but a lack of accurate measurements of production rates of the
residues hindered accurate evaluation of its role as sink for
sulfur. Sulfur-rich residues have been proposed as an
explanation for the sulfur depletion problem. The fact that
sulfur tholins are not volatile means that they cannot be
detected using ground-based submillimeter telescopic observa-
tions. The low dissociation energy of H
2
S
(
S
–
H bond
energy
=
363 kJ mol
−
1
)
and high reactivity of S atom and
HS radical suggest a rapid conversion of H
2
S by irradiation
chemistry. Garozzo et al.
(
2010
)
have demonstrated that H
2
Sis
almost completely consumed under the effect of ion irradiation
at a dose of 21 eV
/
16 amu, producing multiple molecules,
including a sulfur residue. While the authors in Garozzo et al.
(
2010
)
have not determined the production cross section of the
residue, it is expected to be very high, because all the other
sulfurous molecules explain only 30% of the consumed sulfur.
Thus, our
fi
nding of S-rich residue, likely sourced from H
2
S,
has implications for both the present-day solar system and for
pre-planetary systems. Arrokoth and all CC-KBOs are
classi
fi
ed as ultra-red objects: color slope of 28% per 100 nm
at 550 nm for Arrokoth
(
Grundy et al.
2020
)
. Connecting the
color slope of small icy objects in the solar system to a
particular chemical composition could provide a very powerful
way to trace back the history of the solar system and explicitly
test results suggested by dynamical models
(
Morbidelli et al.
2005
; Nesvorný et al.
2020
)
. The chemistry and spectroscopy
of S-rich residue could be a tracer for speci
fi
c regions and
speci
fi
c environments of the early solar system
(
Wakelam et al.
2011
)
.
This work has been supported by the NASA
/
RDAP
program and by the Keck Institute for Space Studies
(
KISS
)
.
This work has been conducted at the JPL, Caltech, under a
contract with the National Aeronautics and Space Administra-
tion
(
NASA
)
and at the Caltech Division of Geological and
Planetary Sciences.
ORCID iDs
Ahmed Mahjoub
https:
/
/
orcid.org
/
0000-0003-1229-5208
Michael E. Brown
https:
/
/
orcid.org
/
0000-0002-8255-0545
Michael J. Poston
https:
/
/
orcid.org
/
0000-0001-5113-1017
Bethany L. Ehlmann
https:
/
/
orcid.org
/
0000-0002-
2745-3240
Mathieu Choukroun
https:
/
/
orcid.org
/
0000-0001-
7447-9139
Kevin P. Hand
https:
/
/
orcid.org
/
0000-0002-3225-9426
References
Bennett, C. J., Pirim, C., & Orlando, T. M. 2013,
ChRv
,
113, 9086
Cartwright, R. J., Nordheim, T. A., Cruikshank, D. P., et al. 2020,
ApJL
,
902, L38
Cruikshank, D. P., Imanaka, H., & Dalle Ore, C. M. 2005,
AdSpR
,
36, 178
Fathe, K., Holt, J. S., Oxley, S. P., & Pursell, C. J. 2006,
JPCA
,
110, 10793
Garozzo, M., Fulvio, D., Kanuchova, Z., Palumbo, M. E., & Strazzulla, G.
2010,
A&A
,
509, A67
Grundy, W. M., Bird, M. K., Britt, D. T., et al. 2020,
Sci
,
367, eaay3705
Hand, K. P., & Carlson, R. W. 2011,
Icar
,
215, 226
Mahjoub, A., & Hodyss, R. 2018,
ApJ
,
869, 98
Mahjoub, A., Poston, M. J., Blacksberg, J., et al. 2017,
ApJ
,
846, 148
Mahjoub, A., Poston, M. J., Hand, K. P., et al. 2016,
ApJ
,
820, 141
Materese, C. K., Cruikshank, D. P., Sandford, S. A., Imanaka, H., & Nuevo, M.
2015,
ApJ
,
812, 150
Morbidelli, A., Levison, H. F., Tsiganis, K., & Gomes, R. 2005,
Natur
,
435, 462
Nesvorný, D., Vokrouhlický, D., Alexandersen, M., et al. 2020,
AJ
,
160, 46
Poston, M. J., Mahjoub, A., Ehlmann, B. L., et al. 2018,
ApJ
,
856, 124
Rubin, M., Altwegg, K., Balsiger, H., et al. 2019,
MNRAS
,
489, 594
Stern, S. A., Weaver, H. A., Spencer, J. R., et al. 2019,
Sci
,
364, eaaw9771
Tieftrunk, A., Pineau des Forets, G., Schilke, P., & Walmsley, C. M. 1994,
A&A,
289, 579
Waggener, W. C., Weinberger, A. J., & Stoughton, R. W. 1969,
JPhCh
,
73, 3518
Wakelam, V., Hersant, F., & Herpin, F. 2011,
A&A
,
529, A112
Wong, I., & Brown, M. E. 2016,
AJ
,
152, 90
Wong, I., Brown, M. E., Blacksberg, J., Ehlmann, B. L., & Mahjoub, A. 2019,
AJ
,
157, 161
Woods, P. M., Occhiogrosso, A., Viti, S., et al. 2015,
MNRAS
,
450, 1256
4
The Astrophysical Journal Letters,
914:L31
(
4pp
)
, 2021 June 20
Mahjoub et al.