Organosulfur Compounds Formed by Sulfur Ion Bombardment of Astrophysical Ice Analogs: Implications for Moons, Comets, and Kuiper Belt Objects

The ion beam used for irradiation was generated at the ARIBE low-energy line of the Grand Accélérateur National d'Ions Lourds (GANIL), Large Heavy Ion National Accelerator, facilities in Caen, France. For both Ar⁷⁺ and S⁷⁺, the ion beam was formed by ions at an energy of 105 keV, with a flux of ≈1 × 10¹¹ cm⁻² s⁻¹. The current reaching the vacuum chamber was controlled by a Faraday cup. The main vacuum chamber contained three windows out of ZnSe that can be cooled down to 9 K and could hold one sample each. Ice samples were formed by preparing a gas mixture in a premixing chamber and depositing this vapor using a mobile needle, allowing target deposition onto a desired window. This device effectively enabled the generation and irradiation of three different deposited samples at similar vacuum conditions. Details on beam generation (Lv et al. 2012) and the IGLIAS device (Augé et al. 2018) have been described previously.

The IGLIAS setup also enabled in situ chemical analysis of the sample by a Brüker V70 Fourier Transform Infrared Spectrometer (FT-IR), under primary vacuum while it is exposed to the beam. IR spectra were acquired before deposition (as a reference background), during irradiation (one spectrum after a new layer was deposited and before it was exposed to the beam, Figure 5) and after warming up to room temperature (one spectrum of the residue at 300 K, Appendix A). A residue is defined here as the refractory remaining material from irradiated ices at 300 K.

Ices were formed with an initial gas mixture of H₂O:CH₃OH:NH₃ (2:1:1). The quantity deposited corresponds roughly to 0.5 μm which is thicker than the penetration depths of 105 keV S⁷⁺ or Ar⁷⁺, calculated to be <0.3 μm (Ziegler et al. 2010). Thus, implantation of projectile ions into the ice could be ensured. The windows were kept at 10 K during deposition and irradiation. To increase the yield of formed residue, we deposited several layers of each sample. A new layer is deposited when the underneath irradiated layer reaches a steady state. The evolution during the irradiation is monitored by FT-IR spectroscopy on both methanol and ammonia infrared absorption bands. This procedure was repeated. In total, 15 layers were formed for the argon-irradiated sample (Ar⁷⁺ ions) and 10 layers for the sulfur-irradiated sample (S⁷⁺ ions). The experiment was repeated (technical duplicates) later generating two 15-layer samples to test for reproducibility (for both Ar⁷⁺-irradiated and S⁷⁺-irradiated). Both duplicates show reproducible results.

We performed SRIM simulations (SRIM, The Stopping and Range of Ions in Matter software; Ziegler et al. 2010) to calculate the effective volume of atom implantation. We assume a sample density of 1 g/cm⁻³. We find that sulfur ions stop on average at a depth of 0.21 ± 0.1 μm within the sample (2σ interval). Within the implanted volume, we estimate the signal-to-noise ratio (S/N) and S/C ratio to be 9E-4 at the end of the irradiation experiments (after 20 minutes). Even accounting for uncertainties on SRIM simulations and the sample density probably being below 1 g/cm⁻³, it is readily apparent that the amount of implanted sulfur is small compared to the elements it can react with.

We have calculated that the irradiated volume (from the surface of the sample to a depth of 0.31 μm) receives a dose of 65 MGy (about 11 eV/16 amu).

The residues, resulting from irradiated ices, were kept under argon atmosphere in a stainless steel vessel, to minimize oxidation prior to analysis (de Marcellus et al. 2015).

Electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS), ran in negative ionization mode, was used for in-depth molecular characterization of the organic matter formed within the residue, resulted from irradiated ices. High resolution mass spectrometry has been widely used for in-depth molecular characterization of extraterrestrial organic matter (Schmitt-Kopplin et al. 2010; Ruf et al. 2017, 2018, 2019; Danger et al. 2016; Fresneau et al. 2017). FT-ICR-MS represents the highest performance in mass resolving power and mass precision among all mass spectrometers.

The residue was dissolved in 50 μL methanol (extraction solvent, LC-MS grade; Fluka). This step was repeated four times. The complete residue got dissolved in methanol. Afterwards, 20 μL of the dissolved residue was diluted in 200 μL methanol. The solution was removed with a microsyringe, ready for flow injection into the ESI source. A solvent methanolic blank was measured in accordance to be able to detect the indigenous soluble organic matter in each ice sample.

The experimental study was performed on a high-field FT-ICR-MS from Bruker Daltonics with a 12-T magnet from Magnex. A time domain transient with four MWords was obtained and Fourier-transformed into a frequency domain spectrum. The frequency domain was afterward converted to a mass spectrum by the solariX control program of Bruker Daltonics. The ion excitations were generated in broadband mode (frequency sweep radial ion excitation) and 3000 scans were accumulated for each mass spectrum in a mass range of 147 to 1000 amu (atomic mass unit). Ions were accumulated for 300 ms before ICR ion detection. The pressure in the quadrupole/hexapole and ICR vacuum chamber was 3 × 10⁻⁶ mbar and 6 × 10⁻¹⁰ mbar, respectively. For CID-MS/MS, ions were accumulated for 3 s.

The electrospray ionization source (ESI, Apollo II; Bruker Daltonics) was operated in negative ionization mode. The methanolic solutions were injected directly into the ionization source by means of a microliter pump at a flow rate of 120 μL h⁻¹. A source heating temperature of 200°C was maintained and no nozzle-skimmer fragmentation was performed in the ionization source. The instrument was previously externally calibrated by using arginine negative cluster ions (5 mg L⁻¹ arginine in methanol).

Mass spectra with m/z from 147 to 1000 amu were calibrated externally and internally to preclude alignment errors. Subsequently, mass spectra were exported to peak lists at a signal-to-noise ratio ≥3. Mass resolving power was 400,000 at m/z = 400 with a mass accuracy of <200 ppb, enabling the separate detection of isobars differing by less than the mass of an electron (Ruf et al. 2017). Practically, this approach enables a direct assignment of molecular compositions with C, H, N, O, and S atoms (and isotopologues in natural abundance) for each individual exact mass (m/z value).

Molecular formulas were assigned from exact m/z values by mass difference network analysis for each peak in batch mode by an in-house software tool (Tziotis et al. 2011) and validated via the senior-rule approach/cyclomatic number (Senior 1951). For formula assignment, 50.1% of all formulas have an error of ±0.1 ppm, 78.5% of all formulas have an error of ±0.2 ppm and 100% of formula assignments within ±0.5 ppm (Figure 8). Further details on the assignment of molecular formulas from FT-ICR-MS big data and their visualization in astrochemical context are given in previous studies (Schmitt-Kopplin et al. 2010; Ruf et al. 2018; Bischoff et al. 2019).

Data mining on organosulfur (CHNOS) compounds represent those m/z signals which were uniquely detected in the S⁷⁺-irradiated sample and absent in the Ar⁷⁺-irradiated sample. Double bond equivalent (DBE) was calculated according DBE = C −(H/2) + (N/2) + 1.

IR spectra of S⁷⁺- and Ar⁷⁺-irradiated ices and residues show the presence of organic compounds which have also been previously observed in UV-irradiated ices (e.g., HNCO or H₂CO; Figure 5, Table 1). Nevertheless, no significant difference between both irradiation sources, S⁷⁺ and Ar⁷⁺, could be deciphered with infrared spectroscopy, and neither organosulfur molecules could be identified in the S⁷⁺-irradiated sample.

On a coarse level, mass spectra from FT-ICR-MS analysis show similar features for Ar⁷⁺-irradiated and S⁷⁺-irradiated ices as well (Figures 1(A) and (B)). Nevertheless, the fine structure obtained with ultrahigh resolving power and ultrahigh mass accuracy, FT-ICR-MS unambiguously revealed the detection of organosulfur compounds in S⁷⁺-irradiated ices (Figures 1(C) and (D)). m/z signals which correspond to organosulfur compounds were absent in the Ar⁷⁺-irradiated reference sample (Figures 1(C) and (D)). Within one nominal mass, up to five organosulfur (CHNOS) signals are present as shown for two representative examples (m/z = 248 and m/z = 438, Figures 1(C) and (D)).

Zoom In Zoom Out Reset image size

In general, the mass spectra of S⁷⁺- and Ar⁷⁺-irradiated refractory residues show dense m/z signal patterns over a broad mass range. From 30,000 experimental m/z signals, ranging from 100 to 900 amu (atomic mass unit), 9,616 molecular formulas, based on C, H, N, O, and S, could be unequivocally calculated (5.3% CHO, 82.6% CHNO, and 12.1% CHNOS, Figures 9 and 10). This high molecular diversity indicates a rich and active sulfur chemistry within these processed ices, triggered by high-energy ion implantation. m/z signals appear as regular patterns reflecting a pure chemosynthetic process as observed in UV photon-irradiated ices (Danger et al. 2013, 2016; Fresneau et al. 2017) or meteorites (Schmitt-Kopplin et al. 2010; Ruf et al. 2017, 2019; Figures 1(A) and (B)). Regular patterns are observed both for the global organic signature and for organosulfur (CHNOS) compounds (Figures 1(A) and (B)).

Atomic ratio plots, known as van Krevelen diagram (Van Krevelen 1950), were used to characterize in detail the detected organosulfur (CHNOS) compounds. H/C against O/C representation enables a first screening of complex mixtures with respect to chemical families (Kim et al. 2003; Wu et al. 2004; Danger et al. 2016; Ruf et al. 2018; Schmitt-Kopplin et al. 2019). Organosulfur compounds present in S⁷⁺-irradiated ices show a low degree of unsaturation and a significant degree of oxygenation (O/C > 0.5, H/C > 1.5, Figure 2). The degree of oxygenation of organosulfur compounds is inversely related with DBE (Figures 2 and 12).

Zoom In Zoom Out Reset image size

Basically, three groups of compounds can be extracted (Figure 2). Group 1 (Figure 2, top right corner) basically consists of small mass molecules up to 400 amu. This group of compounds has high H/C and high O/C ratios with a small number of DBE. Group 2 (Figure 2, top left corner) bears molecules with high masses (mostly 500–800 amu). Their DBE is higher than those in group 1 (DBE > 8, Figure 12). In addition, this group of compounds is enriched in nitrogen counts, bearing up to 14 N atoms per molecular formula (Figure 11). Group 3 (Figure 2, bottom left corner) is characterized by high numbers of DBE within a mass range between 400 and 600 amu (Figures 2 and 12).

A large number of identified organosulfur compounds are saturated molecules, having an averaged DBE of 5.6 (Figure 3). Nevertheless, DBEs of up to 17 are observed, especially for high-mass molecules up to 900 amu (Figure 3 and 10). Compounds with high DBE have low H/C and low O/C ratios (Figure 12).

Zoom In Zoom Out Reset image size

The number of atoms for every element varies for the identified organosulfur compounds (Figures 4 and 13). Almost all organosulfur (CHNOS) compounds bear one sulfur atom and only a few molecular formulas have up to five S atoms. The aforementioned high degree of oxygenation is directly related to the number of oxygen atoms. Most CHNOS formulas possess eight O atoms. This might be related to the partially oxidized carbon starting material (CH₃OH, O/C = 1). This enables the formation of a complex organic molecules with a oxygen-rich carbon backbone (Theulé et al. 2013). These findings are in agreement with results from UV-irradiated ices (Danger et al. 2016). Nitrogen counts are also widespread, ranging mostly from 3 to 10 N atoms per CHNOS molecular formula. The carbon backbone counts indicate two local maxima in stability in C counts, for C = 9 and C = 25. Overall, the diversity in CHNOS formulas is related to a diversity in atom counts therein. The observed large diversity in atom counts indicates a rich and active sulfur chemistry.

Zoom In Zoom Out Reset image size

On average, a stoichiometric formula of C₁₃H₂₃O₇N₆S is calculated, having a DBE (average) = 5.6, O/C (average) = 0.6 and H/C (average) = 1.8.

As observed from UV-irradiated ices (Danger et al. 2016), Ar⁷⁺-irradiated and S⁷⁺-irradiated ices (Figure 9), CHNO represents the major chemical family. This implies that pathways for CHNO formation in all cases are likely based on radical chemistry (Theulé et al. 2013; Butscher et al. 2017). Among the organosulfur (CHNOS) compounds observed in S⁷⁺-irradiated residues, almost all of them bear only one sulfur atom (Figures 4, 10, and 13) and all organosulfur species bear nitrogen atoms (only CHNOS and no CHOS compounds got detected, Figure 9). These findings indicate a selective mechanism of organosulfur formation that is not well understood yet. The irradiating agent, S⁷⁺ or Ar⁷⁺, might act chemically inert in a first step (similarly as UV photons), by breaking bonds and triggering radical and ion formation leading to precursor of complex CHNO compounds. Then CHNO precursors might selectively react with sulfur implemented in the water matrix forming CHNOS₁ species. The low number of S in organosulfur compounds can be explained by low S/N and S/C ratios (9E-4) in the implantation zone (see "SRIM simulations," Methods Section 2.1). Thus, the likelihood of implanting a sulfur atom into more complex organic molecules is low. The test of this hypothesis is beyond this manuscript and is part of ongoing work.

Several astrophysical environments reflect parallels with the here presented experiments.

The icy surface of the Jovian satellite Europa is exposed to a bombardment of energetic electrons and ions, including sulfur (Paranicas et al. 2009). This flux is abundant in the 100 keV range which was used in the experiments presented here. A calculation, based on flux (Paranicas et al. 2009), shows that 20 to 30 minutes of exposure to the sulfur beam in the IGLIAS chamber correspond to a few days on Europa's surface in terms of fluence. However, neither methanol nor ammonia have been detected on Europa (Carlson et al. 2009) and the surface temperature of 80–130 K (Spencer et al. 1999) is noticeably higher than the temperature at which irradiation was performed in our experiments. Nevertheless, the parallel with Enceladus allows us to consider the putative presence of both ammonia and methanol within Europa's internal ocean (Hodyss et al. 2009; Waite et al. 2017); and micrometeoroid bombardment represents a possible source of organic matter being transported to its surface (Carlson et al. 2009). However, as CO₂ represents the dominant form of carbon and the presence of NH₃ is uncertain, one should be cautious about direct implications of the results discussed here to Europa.

Further from the Sun, trans-Neptunian objects (TNOs) and KBOs present surface temperatures closer to our experimental conditions (e.g., 44 K on Orcus: Barucci et al. 2008). Ammonia has been detected on several of these objects, e.g., on Charon (Brown & Calvin 2000; Grundy et al. 2016), on Orcus (Barucci et al. 2008), or on Quaoar (Jewitt & Luu 2004). Both ammonia and methanol are present in comets (Bockelée-Morvan et al. 2004; Altwegg et al. 2017) suggesting their presence in the material that has formed TNOs/KBOs. These objects are exposed to irradiation by solar wind including small quantities of sulfur at about 32 keV/nucleus (derived from the solar wind velocity; Von Steiger et al. 2000), an energy comparable to the one used in the experiments presented here. Using the S/O ratios measured in the solar wind at 1 au (von Steiger et al. 2010) and extrapolating to the flux at 50 au, the fluence used in our experiments is approximated to be reached in 2 Myr. This relatively short time (on astronomical timescales) indicates that icy surfaces of objects such as TNOs/KBOs, possibly rich in ammonia and methanol, may have endured the same transformation as our laboratory samples. Interestingly, 2 Myr is shorter than the predicted time of destruction of ammonia hydrates by solar wind (Cooper et al. 2004; Jewitt & Luu 2004). Thus, NH₃ might still be stable enough to be further processed by sulfur ion bombardment to putatively form organosulfur compounds. This time is short enough that the building blocks of these bodies (e.g., small grains, ice) may interact with sulfur ions from solar wind before their accretion (assuming a similar composition of the solar wind at that time).

This is likely applicable to comets as well. The D/H ratio of comet 67P, for example, indicates a formation at a large distance from the Sun (Altwegg et al. 2015). So, ion implantation would have occurred at low temperatures. It should be noted that due to their formation beyond the H₂S snowline, the aforementioned objects (TNOs/KBOs) are likely to include H₂S, another putative source of sulfur chemistry. The presence of H₂S is supported by comparisons between laboratory experiments and IR spectra of KBOs and Trojan asteroids (Poston et al. 2018).

It is critical to note that thermal processing might be a factor driving the formation of complex organosulfur compounds. The surfaces of a KBO or TNO is unlikely to have undergone such processes. However, some of these objects could later get closer to the Sun and lose the more volatile components of their external layers, as has been suggested for Ceres (De Sanctis et al. 2015), undergoing a more modest thermal processing than in our experiments. On the other hand, we may also consider thermal processing of the interior of these objects, combined with the possibility that most of the ice, not only on the surface, has been implanted with sulfur before the object accreted. Accretional heat and radiogenic heating could have provided thermal processing to relatively high temperatures, especially in the larger objects such as Pluto, Charon, or Triton. In the case of Triton tidal heating may also have been a contributor (Ross & Schubert 1990). Contemporary cryovolcanic activity (Jewitt & Luu 2004; Cook et al. 2007; Desch et al. 2009) could then bring possible organosulfur molecules to the surface. Future missions to KBO/TNOs, comets, or even Neptune (if observations of Triton are included) could uncover these compounds.