ExoMars Mars Organic Molecule Analyzer (MOMA) Laser Desorption/Ionization Mass Spectrometry (LDI-MS) Analysis of Phototrophic Communities from a Silica-Depositing Hot Spring in Yellowstone National Park,USA

1. Introduction

The search for signs of past or present life on Mars is the primary science goal of the joint ESA-Roscosmos ExoMars Program (Vago et al., 2017), including the ExoMars Rover Mission, which features the rover Rosalind Franklin. A key rover instrument that supports this goal is the Mars Organic Molecule Analyzer (MOMA, Fig. 1), a lightweight (∼12 kg), low-power (75 W average), dual-ion source mass spectrometer–based instrument that utilizes a miniaturized linear ion trap (Goesmann et al., 2017). MOMA uses two modes of operation (i) pyrolysis/gas chromatography–mass spectrometry (pyr/GC-MS), which analyzes heat-evolved volatile and semivolatile compounds carried through a gas chromatograph and ionized with an electron beam, and (ii) laser desorption/ionization mass spectrometry (LDI-MS), which analyzes molecular ions (or their fragments and clusters) desorbed directly from a sample surface at ambient Mars pressure with a pulsed laser (Siljeström et al., 2014; Li et al., 2015; Goetz et al., 2016; Goesmann et al., 2017). Compared to electron ionization, laser desorption is a relatively soft ionization technique that enables detection of mid- to high-molecular-weight (100–1000 Da or more) nonvolatile organic compounds with reduced molecular fragmentation.

OPEN IN VIEWER

FIG. 1.

(Left) MOMA instrument on board the 2022 ExoMars rover. (Right) The ETU instrument for the analog sample study. Mass spectrometer (MS).

Gas chromatography–mass spectrometry (GC-MS) instruments have been used during several previous (Pietrogrande, 2013) and ongoing (Mahaffy et al., 2012) space missions, and there are established databases for data interpretation (e.g., NIST 17 GC Method / Retention Index Library [Babushok et al., 2007]) for this well-known technique. In contrast, LDI-MS will be used in space for the first time as part of MOMA. While commercial databases exist for LDI-MS spectral identification of proteins and other biological molecules found in cells (Stump et al., 2002), typically following purifying sample extraction steps, LDI-MS has not been extensively applied to the analysis of intact geological samples, let alone under martian conditions or with Mars-relevant samples.

Maximizing the effectiveness of LDI-MS on MOMA thus requires study of the technique itself as well as thorough analog sample analysis campaigns conducted under simulated Mars operating conditions. Only a handful of LDI-MS Mars-ambient experiments have been conducted with the use of standards and simple mixtures in prior studies (e.g., Li et al., 2015; Goesmann et al., 2017; Goetz et al., 2017), and such experiments with complex natural samples have received even less attention (Bishop et al., 2013).

End-to-end LDI-MS examination of terrestrial analogs of potential martian materials, prepared and analyzed under simulated Mars operating conditions—for example, under Mars atmospheric conditions (∼6–10 mbar of primarily CO₂)—addresses realistic environmental and sample-specific challenges that will be faced when operating MOMA on Mars. The use of crushed/fine-grained Mars analog samples provides an opportunity to identify how variation in sample characteristics (e.g., surface topology, porosity, texture) impacts mass spectral intensity, stability, fragmentation, and other factors in LDI mode on MOMA. Additionally, measurement of various Mars analog samples that contain different proportions of mineral phases and organics that have experienced different degrees of taphonomic alteration, under Mars conditions, will augment the database required for full interpretation of organics that could be detected in geological materials on the Red Planet.

Here, we report the MOMA LDI-MS analysis of a natural, partly mineralized microbial streamer community as an example of how data from this operational mode can support the overall search for molecular biosignatures on Mars. The analog samples analyzed here are mid-temperature (hydrothermal fluid ∼45°C) phototrophic streamers collected from a slightly alkaline silica-depositing hot spring (pH ∼8.5) in Yellowstone National Park, USA (Siljeström et al., 2017). The living and partly mineralized streamer samples were composed of consortia of indigenous phototrophic thermophilic microbial populations, evidence of which are preserved as entombed remains in siliceous hot spring sinters (Siljeström et al., 2017). As these Mars analog samples were previously characterized (Siljeström et al., 2017), this study utilized the samples to focus on optimizing the LDI-MS acquisition conditions and data analysis protocols.

Siliceous hot spring deposits (aka silica sinters) preserve some of the oldest evidence for terrestrial life on Earth (Djokic et al., 2017) and are considered excellent astrobiology targets (McMahon et al., 2018) due to the rapid entombment of microbial consortia in opaline silica (Walter and Des Marais, 1993; Des Marais and Walter, 2019). Opaline silica sinters have likely been identified on Mars, in situ, during the Mars Exploration Rover mission (Ruff et al., 2020) and remotely in moundlike deposits that rim volcanic calderas in Nili Patera (Skok et al., 2010). These opaline silica deposits on Mars appear spectroscopically not to have experienced the earliest diagenetic phase transformations that alter the primary amorphous opal to thermodynamically more favorable opal phases (Sun and Milliken, 2020). Hence, primary microbial biosignatures may be preserved in aqueously precipitated opaline silica deposits, especially those that are hydrothermally precipitated (Walter, 1996; Hinman and Walter, 2005; Guido and Campbell, 2017; Cady et al., 2018).

At Oxia Planum, where the Rosalind Franklin rover with MOMA will land, opaline silica deposits have been detected in association with the edges of delta lobe deposits (Pan et al., 2019). While these opaline silica deposits appear to have formed in situ via diagenetic rather than hydrothermal processes, low-temperature (<100°C) hydrothermal activity likely occurred in the region given the occurrence of abundant phyllosilicates (Carter et al., 2016; Vago et al., 2017; Quantin-Nataf et al., 2019). Hence, the potential exists for (hydrothermal and non-hydrothermal) opaline silica to occur locally (e.g., as vein and fracture fills and in vugs) and wherever silicifying fluids migrate through detrital materials (McMahon et al., 2018).

Most sediments that MOMA will analyze at Oxia Planum are ∼4 Ga (Tanaka et al., 2014). However, a combination of a lack of plate tectonics, slow surface weathering rates, protection from radiation due to burial by surface regolith, and freeze-drying due to low surface temperatures and pressure is likely to enhance the potential to preserve refractory organic molecules in rocks over long periods of time on Mars (Morrison et al., 1969; Vago et al., 2017; McMahon et al., 2018). Even on Earth, where most ancient rocks have experienced at least some degree of metamorphism, organic biomarkers such as porphyrins and certain isoprenoids can survive >1 Ga (Gueneli et al., 2018; Vinnichenko et al., 2020). It is also worth noting that the martian subsurface, which the ExoMars rover will sample, is far more likely than the exposed surface to host evidence of extant life possibly in the form of freeze-dried remains of cellular life.

2. Method

2.3. LDI-MS analysis

Two different types of green streamer samples were analyzed with LDI-MS: several subsamples of freeze-dried partly mineralized green streamers with no further treatment and the TLE.

2.3.1. Instrumentation

Analog sample measurements were conducted on the MOMA Engineering Test Unit (ETU, Fig. 1), which matches the functionality, environmental conditions, and performance of the MOMA flight investigation. The ETU further mimics the flight sample interface but for ease of use in development includes a multisample carousel that holds numerous user-accessible and interchangeable cups or stubs versus the single “refillable container” sample tray featured on the flight rover, into which crushed drill samples will be dosed for analysis by the payload's analytical laboratory drawer (ALD) instruments, including MOMA LDI-MS. The flightlike ETU further enables analog sample testing by using flight controls and protocols, while such analyses were quite limited on the flight model due to contamination requirements. Operational sequences of MOMA LDI-MS described previously (Goesmann et al., 2017) have been wholly duplicated in this study.

2.3.2. Sample preparation

Approximately 10 μL of the lipid standards and the TLE were drop cast onto a blank metal sample stub and dried.

Several thin layers were broken off the green streamer fragments and affixed with sterilized tweezers onto the surface of a sample stub with double-sided tape (3M 966, acrylic adhesive tape). The sample stubs were then loaded onto the sample carousel, which was subsequently evacuated to Mars ambient pressures (0.5–1 kPa ∼4–8 torr).

2.3.3. LDI-MS measurement protocol

Once loaded, the sample was moved under the capillary ion inlet of the ion trap mass spectrometer. A laser beam was then introduced at an incident angle of ∼45° relative to the sample surface from a solid-state, frequency-quadrupled Nd:YAG laser with 266 nm wavelength, ∼1.5 ns pulse duration, and adjustable energies up to 130 μJ per pulse. The laser was operated at a time-averaged pulse repetition rate ≤2 Hz, with an available burst mode of 50 Hz, and focused to an elliptical spot size with a major axis diameter of 600 μm.

At Mars ambient pressure, laser-desorbed ions were drawn into the linear ion trap through a combination of gas flow and applied DC voltage. An inline fast-actuating aperture valve closed off the inlet shortly following ion collection, and ions were then maintained in stable orbits at a low trapping voltage until the chamber was pumped down to ≤10⁻³ mbar, within a few hundred ms, after which ions were ejected to the electron multiplier detector for recording of mass spectra. Due to the mission architecture limitations, only positive ions are detected in the MOMA design.

Laser desorption/ionization (LDI) at MOMA fluence levels is predominantly a surface analysis method and does not remove significant material or “drill” to depths beyond several to tens of micrometers. The green streamer samples were thick enough to avoid measurement of the mounting tape located between the sample and the stub. Additionally, the tape has been confirmed to produce negligible LDI signals at the wavelength and energies employed. Therefore, no substrate background signal from the tape was detected in the acquired spectrum.

Structural characterization of complex molecules is enabled on the MOMA instrument with the stored waveform inverse Fourier transform (SWIFT) method to isolate a specific mass-to-charge range (from “coarse” ∼100 Da down to “fine” isolation window ∼10 Da or below) and with tandem mass spectrometry for the fragmentation of isolated molecules.

Mass spectra were collected in two different surface modes for this study: single-point mode and step mode (Fig. 2). In single-point mode, the laser was focused on a fixed location on the sample surface, and 10 or more individual mass spectra (at ∼5 laser shots per spectrum) were collected, after which the data were nominally summed together to obtain a single representative spectrum for that point on the sample. Raw value sums of stable signals in this step were found not to degrade mass resolution. While such online summing would reduce net data volume on the flight instrument, the baseline for MOMA during the mission will be to record and transmit each individual “few”-shot spectrum. The same measurements will then be repeated on a different nearby surface location to check for sample homogeneity. All the mass spectra of the lipids and pigment standards analyzed in this study were acquired in single-point mode.

OPEN IN VIEWER

FIG. 2.

The single-point mode versus the step mode during MOMA ETU data collections.

Alternatively, in step mode, the sample carousel was moved 100 μm (adjustable in the flight protocol) to a new position after each individual 2 laser shot spectrum, so that several data sets comprising a surface scan were acquired (Fig. 2). Sensitivity to the organic material in these samples was found to be higher in step mode than in single-point mode; less average laser energy was required (20 μJ in step mode compared to >35 μJ in single point mode) to produce comparable spectral peak signal-to-noise ratios. Empirically, it has been observed that surficial organic material is more readily desorbed, with signal levels dropping with each successive laser shot fired at a fixed energy. Because each position (nominally each spectrum) probes a partially pristine sample surface, the step mode is preferred for surveying an area for organics. Although the sample can move to a fully pristine surface after each measurement, moving in smaller step is better for reducing operation time and power consumption on flight. Given the lower sensitivity of the single-point mode for these samples, all the data shown in this paper for the TLE and the freeze-dried green streamer samples were obtained in step mode. The lipid standards were analyzed with LDI separately from the TLE and green streamer samples, and the cleanliness of the instrument was frequently checked between experiments to ensure no cross contamination between samples.

3. Results

3.1. LDI-MS results of the lipid and pigment standards and the TLE of green streamers

The LDI-MS spectrum of the chlorophyll a standard (Fig. 3a) contains the expected parent molecular ion at m/z 893 and a strong peak at m/z 614, which can be assigned to the fragment ion formed from loss of the phytyl side chain from the parent. This result is in agreement with earlier studies by LDI-MS using a 337 nm laser (Suzuki et al., 2009). Additionally, a set of fragment ions at m/z 439, 453, 467, 481, 497, and 555 are observed in the spectrum of the chlorophyll a. These peaks can be assigned to fragment ions that originate from the sequential loss of various CO₂H and CH₂ groups from the porphyrin structure of chlorophyll a. These peaks have not been reported previously (Suzuki et al., 2009) but have been observed in mass spectra of other techniques such as ToF-SIMS (Leefmann et al., 2013a, 2013b; Siljeström et al., 2017).

OPEN IN VIEWER

FIG. 3.

The mass spectra of each lipid standard including (a) chlorophyll a, (b) MGDG, (c) DGDG, and the spectrum of (d) the TLE of green streamer sample. Monogalactosyldiacylglycerol (MGDG); digalactosyldiacylglycerol (DGDG).

The mass spectrum of the monogalactosyldiacylglycerol (MGDG) standard (Fig. 3b) contains strong sodiated [M+Na]⁺ ions at m/z 781 (C₄₃H₈₂O₁₀Na) and 809 (C₄₅H₈₆O₁₀Na). These peaks are consistent with earlier studies of MGDG with LDI-MS that showed that MGDG is most often detected as a sodiated molecular ion (Al-Saad et al., 2003; Vieler et al., 2007). There are also fragment ions present, not observed previously (Al-Saad et al., 2003), which most likely formed by the loss of a sodiated sugar head group (m/z 202.04, C₆O₆H₁₁Na) from the parent ion at m/z 579 (C₃₇H₇₁O₄) and at m/z 607 (C₃₉H₇₅O₄) (Kim et al., 1997; Siljeström et al., 2017).

The mass spectrum of the digalactosyldiacylglycerol (DGDG) standard (Fig. 3c) contains [M+Na]⁺ ions at m/z 939 (C₄₉H₈₈O₁₅Na) and 963 (C₅₁H₈₈O₁₅Na). No strong fragments are found in the positive spectrum (not shown). This is consistent with what has been found in earlier LDI-MS studies of DGDG (Al-Saad et al., 2003; Vieler et al., 2007). The spectrum of the DGDG also contains peaks at m/z 788, 814, and 930 of unclear origin.

Additional standards were tested, but they produced either weak or unstable signals, or both (e.g., β-carotene exhibited low-intensity molecular ions at m/z 536, and hopane was not observable under the limiting standard analysis conditions). Standards reported here included those compounds readily detected with the same analytical protocols as the those used for the streamer samples.

In the LDI mass spectrum of the TLE (Fig. 3d), an intense peak at m/z 614 is consistent with the major fragment ion peak of chlorophyll a. Other, weaker peaks can be assigned to fragment ions from chlorophyll at m/z 481. In addition, a peak at m/z 871 can be assigned to the chlorophyll molecule that lacks the central Mg⁺. This peak is typically observed in MALDI spectra when using an acid matrix as acidity promotes loss of the Mg⁺ and might explain the lack of the molecular ion of chlorophyll a at m/z 893 (Fuchs et al., 2010).

Peaks at m/z 751, 779, and 793 in TLE spectrum (Fig. 3d) can be assigned to the [M+Na]⁺ ions of MGDG. The peak at m/z 751 is most likely a MGDG_16:0/16:1 (see Section 2.2 for the naming convention), while the peak at m/z 779 is a MGDG_16:0/18:1. The peak at m/z 793 can be assigned to MGDG_16:0/19:1. Similarly, the peaks at m/z 941 and 955 can be assigned to the [M+Na]⁺ ions of DGDG (Fig. 3d). The mass of m/z 941 originates from an intact DGDG_16:0/18:0, while the mass of m/z 955 is produced by an intact DGDG_16:0/19:0. It should be noted that the observed peaks of MGDG and DGDG in the TLE spectrum are different from their corresponding standard spectra because MGDG and DGDG in the streamer sample have different fatty acid compositions compared with the lipid standards analyzed in this study (Siljeström et al., 2017). In addition to the main compounds, there is a peak at m/z 551 (C32:0) in the TLE spectrum that might represent a fragment ion from diacylglycerols (DG) (Passarelli and Winograd, 2011). Additionally, the molecular ion of β-carotene possibly occurs at m/z 536 (C₄₀H₅₆) (Leefmann et al., 2013b).

3.2. LDI-MS results of freeze-dried partly mineralized green streamers

The LDI-MS spectra of the freeze-dried green streamers (Fig. 4) are dominated by a peak at m/z 795 of unknown origin. There is also a weak peak at m/z 614, which can be assigned to the fragment ion of chlorophyll a. This peak becomes an intense peak at the same mass when SWIFT is applied to the mass region m/z 480–720 (Fig. 4). When SWIFT was applied to other mass ranges such as m/z 720–1000, where the peaks of MGDG and DGDG should be observed, there are many peaks in the range generating relatively noisy spectra. For example, in the mass isolation of the m/z 700–900 range, a small peak at m/z 779 that can be tentatively assigned to MGDG was detected (Supplementary Fig. S1).

OPEN IN VIEWER

FIG. 4.

The mass spectra of (top) the freeze-dried partly mineralized green streamers and (bottom) the enhancement of the chlorophyll a detection while using narrower mass range isolation (m/z 500–700 window).

4. Discussion

The Yellowstone sample chosen here represents a complex real-world example of samples that contain pristine and partly mineralized and degraded organics, and the measurements and data analysis performed were among the first attempts for us to systematically study a challenging sample with MOMA. The LDI mass spectra of the standards, the TLE sample, and the freeze-dried partly mineralized streamers indeed showed that MOMA LDI-MS detects a variety of lipids and pigments that include chlorophyll a, DG, MGDG, DGDG, and β-carotene (Table 2).

Other compounds known a priori to be present in the green photosynthetic streamer consortia, such as sulfoquinovosyldiacylglycerol (SQDG) and phospholipids, were not detected by the ETU version of MOMA. These compounds typically produce their most diagnostic ions in negative mode, as indicated by analyses performed previously with ToF-SIMS (Siljeström et al., 2017) and with commercial LDI-MS instruments in our laboratory (Supplementary Fig. S2, Table 2). Additionally, MGDG was by far the most abundant lipid (11, 804 μg/g dry weight) in the TLE of the streamers when compared to the combined concentration of DGDG and SQDG (367.9 μg/g dry weight) and the phospholipids (1230 μg/g dry weight) (Siljeström et al., 2017).

The LDI-MS mainly detects high-mass, nonvolatile species, since many of the volatile species are removed during the vacuum setup required for sample analysis. For example, the LDI-MS did not detect any signals from waxy esters (m/z 450–550) in the TLE. These compounds are known from our previous conventional GC-MS biomarker analyses to be abundant in the green streamer samples due to the presence of the non-oxygenic phototroph Chloroflexus spp. Yet these compounds were typically only detected in ToF-SIMS spectra of the TLE when analyzed at lower temperature (< -20°C), indicating their loss under low analytical pressures (10⁻⁹ mbar inside the ToF-SIMS instrument) (Siljeström et al., 2017). This suggests that it may be possible to detect more volatile compounds on Mars as the sample temperature in the rover is expected to be lower than in our laboratory.

A comparison of the results from the TLE (Fig. 3d) and the freeze-dried green streamer samples (Fig 4) revealed that it was more challenging to detect the lipids and pigments in the streamer samples than in the TLE (Table 2). A notable exception was the ability to detect chlorophyll a in the streamer samples, though the LDI-MS spectra of these samples were noisier than that of the TLE. This finding is interpreted as resulting from the lower concentration of the lipids in the partly mineralized freeze-dried green streamers versus that in the TLE, combined with the high ionization potential of chlorophyll a, the latter of which interferes with the detection of other compounds in non-extracted samples (Vieler et al., 2007). In addition, the combined mixture of cells and mineral precipitates in fine spatial association often generates a strong mineral signal baseline, above or within which possible organic biosignatures may be partially masked until isolated and examined by using adjustable instrument parameters.

As a test of one such parameterization and tuning approach, we performed a mass isolation measurement that has been observed to significantly enhance the resolution and signal-to-noise ratios of observed but weak signals (like that of the chlorophyll a). The consequent reduction of total charge in the ion trap at a given partial pressure and trapping voltage for such a measurement tends to reduce space charge and trap capacity-induced effects as remaining ions are limited to a smaller number of potential m/z values. This type of signal isolation is expected to enable a broad “window scan” that divides the full m/z range into a sequence of SWIFT spectra if required during initial examination of martian samples (Li et al., 2017). Mass isolation can also be used in an iterative analysis of broader m/z ranges, which can reveal correlations in molecular fragments with their parent molecule, and narrower m/z ranges to focus on individual molecular detection and structural characterization.

In either standard or isolation window operation, the total and relative spectral signals attributable to organic species were found to be higher in step mode than in single-point mode. The depletion of the signal from desorbed surficial organic material with increasing laser pulse number, over a relatively constant signal level from inorganics, was particularly observed in single-point mode. In contrast, the step mode was predominantly performed on partially “fresh” surfaces and resulted in steadier and, thus, more inter-comparable signals over the nominal multiscan data set.

Step mode can be implemented on the flight rover, in principle with an arbitrary number of positions on the sample surface addressed over multiple cycles (subject to the several-centimeter spatial extent of the rover's refillable container). In practice, however, this is not likely to be the baseline for most samples given the extra operation time and energy involved in micro-stepping the sample, which is controlled by the rover and requires uninterrupted communication with MOMA. Energy and thermal considerations also lower the throughput of a true step mode when run in conjunction with other required phases of analysis within the operation constraints. However, by recording and transmitting spectra that contain as few as one laser shot, step mode data can effectively be extracted during postprocessing. That is because the LDI-MS spectra of the first few laser shots from each sample area where data are acquired during nominal multiple “single-point mode” locations can be selected from the data set and analyzed as an isolated summed spectrum (Fig. 5). The main concern for collecting and/or extracting data from multiple points on the same sample during rover operations is the a priori uncertainty of the spatial chemical heterogeneity of the crushed particulate sample (Vago et al., 2017), especially when data analysis is constrained to a narrow m/z window. This concern can be addressed prior to rover operations in part through testing with a variety of types and preparations of known mineralogical/mineraloid matrices. Additional information about the spatial chemical heterogeneity of the samples will also be obtained from the other rover instruments through planned rover team collaboration. For example, a combination of the results of various instruments, such as those from MOMA and the IR imaging spectrometer MicrOmega (Bibring et al., 2017), will provide fine-scale mineral maps of the delivered sample.

OPEN IN VIEWER

FIG. 5.

LDI spectra of the freeze-dried green streamers taken by 25 laser shots compared to the spectra of the first 5 shots from the same data set. The chlorophyll a is much clearer in the 5-shot spectra.

This study showed that chlorophyll a for the partly mineralized microbial streamer samples was the organic molecule most easily detected by LDI-MS on the MOMA ETU. Chlorophylls and bacteriochlorophylls are the dominant photosynthetic pigments used by phototrophic bacteria to absorb energy from light. Other types of porphyrins, such as heme, are also abundant in non-phototrophic organisms, including anaerobic bacteria, as part of the electron transport chain in the cell membrane. Due to the ubiquity of porphyrins across the three domains of life, it is hypothesized that they developed early in life's evolutionary history (Bosak et al., 2013; Fox and Strasdeit 2013). The peaks we detected from chlorophyll do not represent the intact molecule but are mainly produced by the porphyrin ring of the chlorophyll. Porphyrins other than chlorophyll should, therefore, be readily detected through similar porphyrin ring substructures. The ring structure of the chlorophyll molecule is known to survive in the geological record for extended periods of time. Porphyrins have been detected in 1.1 Ga sediments on Earth (Gueneli et al., 2018) and have also shown resistance to degradation by the harsh conditions on Mars, including photolysis caused by radiation, especially if protected in rocks (Stromberg et al., 2014; Baqué et al., 2016). They have been proposed as potential biomarkers in the search for extraterrestrial life (Suo et al., 2007) as there is no known way of producing them abiotically, although this hypothesis has of late been challenged (Fox and Strasdeit, 2013).

Analyses of complex analog samples such as those presented here enable the preparation of effective MOMA LDI-MS measurements on Mars, an important consideration given the limited time and energy resources of any rover instrument. By utilizing single-point and scanning (or effectively scanning by using multi-point) modes along with more advanced operations that involve mass isolation and tandem mass spectrometry, the LDI of MOMA provides a means to obtain representative and diagnostic data from multiphase samples that contain a range of variable concentrations of organics and minerals. Collaborative operations and data processing among MOMA LDI-MS, the Raman Laser Spectrometer (Rull et al., 2017), and MicrOmega (Bibring et al., 2017) on the same sample in the Rosalind Franklin rover's refillable container will provide a more comprehensive assessment of both nonvolatile organic and inorganic compositions. It will also advise the potential follow-on analysis with MOMA's GCMS mode that utilizes sample pyrolysis and chemical derivatization in the instrument's single-use ovens.

5. Conclusions

In the current study, the LDI mode of the MOMA instrument on Mars was conducted at Mars ambient conditions and utilized to detect organic compounds preserved in partly mineralized microbial streamer samples collected from a hot spring in Yellowstone National Park, USA. These samples were some of the first complex real-word materials analyzed by MOMA LDI-MS. The results from our analysis of the lipid extract of the streamer demonstrates the detection of high-molecular-weight molecular compounds, such as chlorophyll a, β-carotene, DG, MGDG, and DGDG, which are diagnostic of the cellular material in the samples. In the untreated partly mineralized streamer samples only chlorophyll a was detected, likely because of the high background and the relative high ionization potential of chlorophyll a. This work underscored the value of utilizing selected ion accumulation and step mode to improve the quality of the spectra and the fidelity of the data interpretation, especially in the presence of high background peaks. This information will provide guidance for future MOMA in situ surface operations and mass spectral data interpretation. Additional study and analysis of Mars analog samples will continue in preparation for the 2023 landing of the ExoMars rover on Mars.