BIOFILM ADHESION ON THE SULFIDE MINERAL BORNITE & IMPLICATIONS FOR ASTROBIOLOGY

We present research observing and documenting the model organism, Pseudomonas fluorescens (P. fluorescens), building biofilm on a natural mineral substrate composed largely of bornite (Cu5FeS4), a copper-iron sulfide mineral, with closely intergrown regions of covellite (CuS) and chalcopyrite (CuFeS2). In examining biofilm establishment on sulfide minerals, we investigate a potential habitable niche for microorganisms in extraterrestrial sites. Geochemical microenvironments on Earth and in the lab can also serve as analogs for important extraterrestrial sites, such as sheltered, subsurface microenvironments on Mars. The sulfur cycle has been important to Mars’s surficial and interior processes during most of its geological record (King and McLennan, 2010; McLennan and Grotzinger, 2009). By growing P. fluorescens on polished chips of a sulfur-rich mineral (a Martian microenvironment analog material), we asked which distinct mineral phases bacteria selected as adhesion points. We collected cell count data, applied Raman spectroscopy, and tested FTIR spectroscopy as a complementary tool, to query whether biologically derived signals increase in intensity as bacteria grow in culture, and if the character of the mineral substrate controls cell distribution to some degree. Through use of the ANOVA statistical test on Raman peak intensity data, we found significant differences between the biologically derived signal intensity (i.e., bonds related to EPS, amide I, amide III) for minerals: amide presence on bornite differs from on other substrates. ANOVA tests of cell counts yielded no significant difference between mineral substrate. We report that bornite mineral surfaces are subtly preferred in early adhesion by this model biofilm former.

1 Manuscript Format is in use. The manuscript is prepared for submission to the journal Icarus (ISSN: 0019-1035) by Elsevier.

Scientific background
Research of the last century has illuminated the possibility of life beyond Earth.
Mars is specifically of interest due to its location in the "habitable" zone of the solar system. Water once pooled across its surface, and evidence suggests the surface of early Mars had a variety of geochemical environments suitable for forming a biosphere . In particular, Mars's geologic history is intertwined with its sulfur cycle (see

1.3).
Earth's own sulfur cycle relies heavily on microbial activities, and astrobiologists can use terrestrial sites, such as sedimentary mud rocks and massive sulfide deposits, to inform our understanding of the parameters endoliths require to survive. In planetary environments, endoliths are microorganisms that reside in rock and pores between mineral grains. Many endoliths are also extremophiles--they survive in extreme environments with living conditions such as extreme heat and cold, high acidity, intense pressure, and limited resources . Terrestrial endoliths are of great interest to astrobiologists as they represent the possibility of Martian or other extraterrestrial micro-organisms taking refuge in similar environments .
As Mars transitioned into a colder, dryer planet, life would have had to seek refuge in the subsurface as endolithic chemolithotrophs (lifeforms able to convert inorganic reduced compounds into energy). Fe and S are the two most abundant compounds on Mars's surface , and on Earth we see ecosystems thriving in massive iron-sulfide deposits-some of which are analogue sites for Mars (ex. Rio Tinto, Spain and ancient Meridiani Planum, Mars) Tosca et al., 2005) Benchtop experiments that simulate geochemically extreme environments can serve as analogs for critical extraterrestrial sites, such as sheltered, subsurface microenvironments (Blanco et al., 2013;Wierzchos et al., 2012. We can observe in the lab microorganisms interacting with Mars-relevant minerals to inform our understanding of in which geochemical environments biofilms are most productive, and how biomarker signals may vary in relation to the amount of time a biofilm has to adhere to a surface. The mechanistic process of biofilm adhesion marks the creation of a bacterial community and the various proteins used to achieve this adhesion are all indicative of life and detectable by Raman spectroscopy Yang et al., 2015). With sulfur and sulfides detected across Mars sites that bear astrobiological interest, it is worthwhile to investigate the effectiveness of Raman spectroscopy in biosignature detection on sulfur and iron-bearing minerals: for the purposes of this paper, amides and EPS will be the primary biomarkers investigated for evidence of surface adhesion and the established presence of microorganisms on astrobiologically relevant geological surfaces.
In this study, we develop a novel protocol for studying biofilms on Mars-relevant minerals to (a) determine the effectiveness of Raman spectroscopy in identifying characteristic peaks of adhesion proteins and other biosignatures on astrobiologically relevant minerals, (b) apply spectroscopic data in an astrobiological context, and (c) observe whether fine-scale geochemical heterogeneity in a sulfide mineral matrix may control bacteria adhesion. This protocol uses a model organism, Pseudomonas fluorescens (P. fluorescens) to generate biofilm, establish "proof of concept," and align spectroscopic and direct count observations to determine adhesion site selectivity in Fe/Cu/S-bearing mineral environments.

Sulfur on Mars
Mars has long been considered a sulfur-rich planet . Current evidence suggests that the sulfur cycle may have dominated the planet's surficial processes during the greater part of its geological record. A primitive mantle composition of up to twice the sulfur content of Earth's (>400ppm), would make the differentiated interior an important S reservoir, and sulfur-bearing soil and rocks outcrop in locations of astrobiological interest   .
There are three major mineral forming eras in Martian history: the phyllosian (early Noachian), the theiikian (late Noachian to Hesperian), and the siderikian (3.5 billion years ago to present day) Bibring et al., 2006;Poulet et al., 2005). These eras are derived from the major mineral deposits of hydrated sillicates, sulfates, and iron oxides that we see today. Bibring et al. (2006)  production. While degassing SO2 is thought to be responsible for these minerals, it is also possible that pre-existing sulfides, weathered by acidic conditions, could generate sulfates as well . Sulfides have been identified as accessory phases in shergottite meteorites  and have been proposed as important constituents of some areas near the Martian surface (Bell III, 1996;. Where potential oxidants are concerned, clays and metal oxides composing surface minerals (such as Fe2O3 and FeO4 2-) have been detected on the surface today and were likely widely available during Mars's early clay forming phase (Lasne et al., 2016). H2O2 in the Martian regolith produced via the interaction of minerals and water is suggested to have been available as an oxidant during this time as well (Lasne et al., 2016).

Sample Selection
Primary sulfides disseminated in Martian crust: Sulfides are found as tiny grains in a groundmass of mafic rock. Sulfur is widely distributed on Mars, observed on the oxidized, weathered planetary surface generally as sulfates   ) (Molecular Cube by K. Joseph from the Noun Project)  .
Ore-type unit of massive sulfide: Cu and Fe sulfide ores form through planetary processes. Hydrothermal circulation drives hot, often acidic waters through country rock, and generates predictable packages of sulfide-rich minerals when hot, chemically evolved waters vent into cooler, more oxidizing water reservoirs (e.g., black smoker systems in the seafloor on modern Earth). Magmatic processes generate igneous rocks of particular compositions: as magma cools, mineral constituents crystallize, leaving behind gradually altered residual melts Burns and Fisher, 1993;; sulfides can coalesce as large ore bodies.
Sedimentary mud rocks: On Earth, ocean water containing sulfur as free SO4 2percolates through sedimentary bedding at the ocean floor. This continually feeds oxidized sulfur in the form of sulfate ions through sedimentary substrate containing Fe and Cu, which takes up sulfur to form sulfide mineral deposits . On Mars, we could see similar redox fronts with stable sulfates and sulfides in ancient marine or lacustrine sedimentary systems.
The field of sulfide mineral surfaces and surface reactivity is relatively new, however it is of research interest because of sulfides' role in functionalization of organic molecules (Vaughan, 2006). Chalcopyrite (CuFeS2) is considered the most important copper ore mineral, and during aqueous oxidation the sulfide forms a layer of Fe2O3 at the surface, leaving behind an unoxidized, metastable CuS2 within the chalcopyrite crystal structure (Vaughan, 2006). This has the effect of reducing the chemical reactivity of the mineral surface. The majority of sulfides experience a similar process when reacting with water or any other chemical process; the surface is typically altered or consumed by the reaction (Vaughan, 2006).
Biochemical reactions are also responsible for the cycling of sulfides on Earth's surface. Iron sulfides are considered the most abundant and important sulfides in biosystems, because many microorganisms can heighten sulfide dissolution by oxidizing the metals in sulfide minerals (Vaughan, 2006). One such case of biologically mediated dissolution is in the case of acid mine drainage (AMD). When exposed to oxidative conditions, the sulfide minerals will dissolve and produce acidic waters.
Sulfur-and iron-oxidizing bacteria will often catalyze this reaction to generate metabolic energy (Vaughan, 2006).
For this study, we selected the sulfide mineral bornite (Cu5FeS4) as our astrobiologically relevant substrate (Fig. 2)

Biofilm adhesion to mineral surfaces
Biofilm is an accumulation of microorganisms, such as bacteria that have extruded an extracellular polymeric substance . Before a dense biofilm forms, bacteria create a transient association with a surface Yang et al., 2015). When bacteria form a stable connection, bacteria produce a biofilm, which provides a clement shelter (of particular biogeochemistry) for the microbial community to grow.
A biofilm's extracellular polymeric substance (EPS) is made from mostly polymeric sugars, proteins, lipids, and nucleic acids . Amino acids are organic compounds indicative of biological processes. Composed of carboxyl (-COOH) and amino (-NH2) groups, amino acids are the building blocks for proteins found in all terrestrial life and therefore make an ideal biomarker . Amides are a functional group within amino acids and are used by biofilms, among other things, for surficial adhesion . Raman spectra for amides in general have been well documented, along with complementary peaks for EPS .

Raman spectroscopic applications in astrobiology
Raman spectroscopy is a valuable tool for future planetary exploration endeavors and will be used for detecting biosignatures on Mars within the next two years . In addition, biofilm adhesion is a critical biological process for microbial life, and leaves relatively durable amide and protein deposits on mineral surfaces, albeit at very fine scale.
Previous studies have created libraries of Raman spectra for identifying proteins, bacterial species, and minerals . Efforts to characterize the mineralogy of Jezero Crater, the landing site for Mars 2020, via Raman spectroscopy are underway . Also, ExoMars, a joint mission between the European and Russian space agencies, plans to bring a Raman spectrometer to Mars . A library of Raman spectra for astrobiologically important minerals is in development. We offer data that reflect the resolution power of Raman spectroscopy in detecting markers of biological activity and consider the rigor of the technique through comparison with a classic direct count approach.

Sample preparation
Bornite samples, provided by Ward Science, were first analyzed using powder x-ray diffraction (XRD) and confirmed to contain bornite along with traces of covellite (CuS). Bornite samples were subsampled into approximately 1-cm-wide chips before being secured in a polishing ring with orthodontic resin. Samples were polished first by hand using tap water to wet 8" silicon carbide grinding disks (grits 400, 600, and 1200), followed by monocrystalline diamond suspension (grits 3 μm, 1 μm, 0.25 μm) on 8"

Pseudomonas fluorescens
Chemosynthesis is defined by Merriam-Webster as "the synthesis of organic compounds by energy derived from inorganic chemical reactions." Chemosynthetic organisms rely on these inorganic chemical reactions for their metabolism, whereas photosynthetic organisms need light to generate organic compounds. Chemosynthesis should not be confused with chemotaxis, which refers to how a bacterial cell will move in relation to chemical resources. Obligate aerobic organisms require oxygen to metabolize carbohydrates and grow.
P. fluorescens is an obligate aerobe that can exhibit chemotaxis when necessary, however flagella are not required for biofilm formation (de Weert et al., 2007;Smith & Doetsch, 1968;

Culture preparation and incubation
We dissolved A few drops of the primary culture were added to a smaller bottle of LB Broth, 12 to 24 hours before initiating experiments; cells were transferred to smaller bottle of broth during exponential growth.
Triplicate mineral chips were incubated in the bottles for 2, 6, 12, and 24 hours.
After sterilizing with ethanol, one mineral chip was placed in each well of a 6-well plate (Thermo Scientific, item # 130184) before filling the wells with ~9 ml of LB broth and

Raman spectroscopy analysis
Samples were rinsed in triplicate with filtered tap water and allowed to dry. We used a WiTech confocal Raman spectrometer with a 758 nm laser at ~2.00 mW power.

Statistics
The ANOVA two-factor with replication is a statistical technique that proves differences among group means in a sample and is appropriate for use in this study because we aim to see what effect, if any, slight differences in geochemistry have on biosignature strength . Using ANOVA two-factor with replication, we compared the average intensities of the three mineral phases at each timestep (t at 0, 2, 6, 12, and 24 hours) for three peaks of interest: amide I (1650-1680 cm -1 ), amide III (1294 cm -1 ), and nucleic acids (720-734 cm -1 ). These specific wavenumbers were selected within published ranges for proteins and carbohydrates indicative of DNA . The ANOVA addresses the null hypothesis (H0) that there is no change in organic compound surface loading, as observed in related peak intensity values, across mineral phases; the corresponding hypothesis (Ha) is that the peak intensity values vary significantly across mineral phases, supporting mineralogical control over organic loading.

EMPA analysis
Microprobe data confirmed visual observations that the sample had three distinct sulfide mineral phases: bornite (Cu5FeS4), covellite (CuS), and chalcopyrite (CuFeS2). The electron microprobe scanning electron microscope (SEM) accessory showed that covellite formed along defects and narrow fissures, whereas chalcopyrite crystals were embedded in the bornite matrix.  .  Table 3 shows the three phases identified in geochemical analysis. Note in B, there are dark veins within the mineral (low iron). These are veins of covellite (CuS).

Bornite
Bornite, with a higher weight percent of Cu compared to chalcopyrite or covellite, dominates A, leaving chalcopyrite as the brightest regions in B and C.

Raman spectroscopy
Four plots were generated for each mineral in the amide III range to show the Raman spectra for each mineral phase (bornite, chalcopyrite, covellite) within the sample and the glass control; spectra collected for covellite are presented in Fig Samples incubated for 24 hours were analyzed within 1 week of being harvested. 12 hours were analyzed within 1 to 4 weeks, 6 hours within 3 to 4 weeks, 2 hours within 1 month. Longer wait times for Raman analysis led to a deteriorated intensity, which background processing should correct for. hours. Data has undergone SNV normalization. Covellite and glass have the most well-defined trends, with covellite having a higher intensity in the lower range. Bornite and chalcopyrite have less defined peaks in this range and likely have a lower biomass content compared to covellite and glass. This shows that there could be more biomass loading on glass and covellite. Glass, being more transparent than bornite, likely allowed for a better signal to be resolved, whereas with covellite it could be a potential adhesion preference for the biofilm.

Key changes in biosignature intensity over time
As discussed in section 1.4, as a biofilm evolves, cells excrete increased amounts of EPS and proteins. High concentrations of amides are indicative of increased biofilm production, as they are a major building block of proteins and suggest the presence or former presence of amino acids. Simply put, based on these results, the longer a biofilm grows, the higher the concentration of amino acids and EPS. As a result, biomarker intensity will increase over time.
In all samples, amide III intensity varies statistically significantly between substrate, whereas other biosignatures don't show enough change to be considered significantly different. Amide I and amide III both have strong Raman signals, however, amide I tends to have overlapping peaks with different secondary structures, as does EPS . Secondary protein structures have far more resolved peaks in the amide III region . As in Fig. 6, we can see that the amide I peak (~1650-1680 cm -1 ) is obscured by neighboring signals. The same tends true for all wavenumbers greater than 1000 cm -1 . The ANOVA test of background-corrected spectral intensity rejects the null hypothesis that they are the same. Over time, there is a significant difference between the biosignature intensities.  Table 5 shows the ANOVA analysis of peak intensity at target wavenumber regions as a function of time. For all but amide I (bornite, glass control) & amide III (glass control) regions, signals do not vary significantly from time 0h to time 24h.
Overall, growth seems to be fairly uniform throughout the incubation series, with the control exhibiting the most change over time. This is expected, as SNV normalization of Raman spectra scales the data so samples are comparable. In Table 6, we see the ANOVA results comparing the intensity of each peak between mineral phases at each time step. As was the case in Table 4, amide III intensity seems to be influenced by our variable-in this case, the substrate. The tyrosine/guanine regions also show statistically significant differences between the mineral phases during the lag phase (0-2 hours), where bacteria adapt to their growth conditions, and at 24 hours (see appendix 3, fig. 1) .
The glass control Total cell numbers are fairly diluted in the lag phase and may not be evenly distributed. At 24 hours, biofilm growth has entered its exponential phase; many cells have adhered to the surface, increasing biomass, and in turn increasing the signal from tyrosine/ guanine regions. These statistical results suggest that fine-scale differences in geochemistry are an influencing factor in biofilm growth and adhesion.
Selection of specific mineral phases for adhesion points happens at a critical stage of early growth, and the mineral-specific distribution of adhesion points becomes visible later in the evolution of the biofilm-mineral system, when organic material has accumulated more densely on the mineral surface.
Averaged cell count results suggest that biofilm preferentially associated with covellite (depleted in Fe, higher in Cu and S relative to bornite and chalcopyrite), although cell counts are not statistically significant. Close examination of the polished sample surface via scanning electron microscopy (SEM) revealed that covellite grew along narrow fissures in the mineral. Bacteria have increased success in growth on uneven and rugged surfaces , which may explain why P.
fluorescens seemed to congregate along these covellite veins. Fe makes up a smaller weight percent of bornite than chalcopyrite (11.13% and 30.43% respectively) . Geochemistry may be an important controlling factor in this study. These results indicate P. fluorescens had higher populations of bacterial cells on phases with depleted Fe and enriched in S.

Cell counts
To determine directly the effect geochemistry has on cell adhesion, analysis of variance tests were also performed on cell counts of samples that had been incubated for 24 hours. Twenty-one 5 μm 2 regions of interest on covellite, bornite, and chalcopyrite were used. The following statistics were calculated using ANOVA single factor. research, increasing the total number of regions of interest will yield a more accurate portrayal and clarify the relationship between geochemistry and adhesion preference.

FTIR Spectroscopy
ANOVA tests disproved the null hypothesis that the intensities of peaks in the amide III and tyrosine, guanine ranges are the same across mineral phases.
Considering this, we incubated a 24 hour sample for Fourier-Transform Infrared (FTIR) spectroscopic analysis. FTIR spectroscopy is considered complementary to Raman .Where Raman uses one frequency to generate a spectra, FTIR uses a range of frequencies in the electromagnetic spectrum to identify the chemical makeup of a sample. For this experiment, we generated heat maps of our polished chip before and after incubation at key peaks: amide I (1675, 1685, 1695 cm -1 ), amide III (1240-1310 cm -1 ), nucleic acids (3200), and tyrosine/ guanine (1515, 3200 cm -1 ). The results were complementary to our Raman ANOVA analysis; minerals had varying biomarker signal intensity. Fig. 7 (below) shows the heat maps for amide III (1310 cm -1 ). We can see that bornite initially has little to no signal in this range. The weak but visible signal in the post-incubation analysis verifies the introduction of amide III and thus biological materials onto the mineral surface. Chalcopyrite, which begins with some signal, also sees an increase over time. Covellite, however, is interestingly different. We see some areas that show the same increase in signal strength over time as bornite and chalcopyrite, however, the bottom half of the specimen does exactly the opposite.
Further exploration would be needed to determine precisely why, but for the purposes of our experiment we are able to confirm an overall increase in biomarker signature intensity post-incubation.

Raman Spectroscopy in Astrobiology
In 2020, NASA's Mars Exploration Program will launch a new rover explicitly designed to investigate "high-priority science goals for Mars exploration," including seeking signs of habitable conditions or past microbial life in Mars's ancient history. To do this research, the rover is equipped with a suite of scientific instruments, including cameras, weather sensors, ground-penetrating radar, and spectrometers.

The Scanning Habitable Environments with Raman & Luminescence for
Organics & Chemicals (nicknamed SHERLOC) is one of Mars 2020's many systems.
SHERLOC's mission is to search for biosignatures using a camera, laser, and Mars's first ever Raman spectrometer. Considered complimentary to infrared spectroscopy, Raman is predicated on analyzing a material's spectral fingerprints . These fingerprints are caused by inelastic scattering of light when it interacts with matter and are measured in units of cm -1 , or wavenumbers.
When the light of a single frequency or wavenumber, ν0 interacts with a material, a small amount of it is sent scattering in different directions. The highest intensity in the spectral distribution of scattered light has the same incident wavenumber as ν0; however, there is a visible spectral component at other wavenumbers. This Raman shift is caused by the incident light interacting with atomic and molecular vibration modes-modes which are characteristic of and can be used to identify the material being observed. Raman scattering tends to be a weak effect depending on the material's polarizability. However, Raman is still considered extremely useful and practical, especially for planetary exploration: it is nondestructive, fast, requires no sample preparation, requires no physical contact with the sample, and the spectral bands provide definitive compositional characterization of the sample Franquelo et al., 2009;. On Mars, SHERLOC will use Raman spectroscopy to detect minerals, organic molecules, and potentially biosignatures.
The Raman spectrometer used in this study was a confocal Raman spectrometer, whereas SHERLOC is a Deep UV resonance Raman and fluorescence spectrometer.
SHERLOC uses a type of Raman frequency that is sensitive to aromatic organics, making it highly effective at biomarker detection (10-6 w/w at <100 μm scales). Our spectrometer used a much higher frequency laser (785 versus SHERLOC's 270 nm) at a low power (2mW). Initial trials showed that bornite would be "singed" very easily by the laser, forcing us to use the lower power. This likely had an impact on the weaker intensities in regions of interest, as biomaterial was thinly spread across the mineral surface and often overwhelmed by the mineral (or glass) spectra. SHERLOC is better prepared to circumvent similar issues, since it is more sensitive to organics than mineral substrate, however it is worthwhile to consider how substrate impacts the overall spectra.
We also polished our samples to reduce the amount of laser light absorbed. The objective depth of field (DOF) for our spectrometer is 9 μm for a completely transparent sample. Mars 2020 will be analyzing unpolished particles of sand, clay, and solid rock, which could reduce DOF and add to the difficulties of detecting scarce or weak biomarkers.

Carbon on Mars
Currently, there is no quantitative estimate for total carbon on Mars. However, instruments designed to measure total organic and inorganic carbon (TOC and TI, respectively) have been proposed (Toon et al., 2019;Zent et al., 2006). Much progress has been made concerning Mars's atmosphere and sites accessed by rovers. Owen et al.

Conclusions
Results from this experiment suggest that the model organism P. fluorescens may exhibit chemotaxis with regard to mineralogical substrate in some capacity when generating biofilm. Based upon cell count data, there appears to be a statistically significant preference for cells to adhere to covellite phases over bornite or chalcopyrite.
This could be a result of surficial defects, as most covellite phases tended to appear as veins. Surface properties or overall structural complexity (chalcopyrite is tetragonal, covellite is hexagonal, bornite is orthorhombic) may also play important roles in controlling cell adhesion.
This work advances scientific understanding of the interactions of P.
fluorescens, a proxy for biofilm-forming microbes, with sulfide minerals and provides a foundation for further work categorizing biomarker signals on sulfide minerals for astrobiological research. Additional experiments involving the analysis of biofilmcoated surfaces exposed to simulated Mars surface conditions are an exciting area of future research. For next steps, we would suggest conducting similar experiments with sulfur and iron-oxidizing bacteria instead of P. fluorescens. Future research may also focus on sulfides with a higher weight percent of iron, including end member pyrite.
Mars is dominated by iron and sulfur, and it is therefore imperative that biosignatures on iron-sulfides-not only sulfates, as current literature tends to emphasize-are observed and catalogued.

Polishing protocol
Polishing mineral samples can be an essential first step in many analyses. Getting a high-grade polish can be a challenge for beginners. The following protocol was developed with advice from Dr. Joseph Boesenberg, Brown University, for polishing copper sulfide minerals, but can be used for other mineral and rock samples. It is important to know where your samples lie on the Mohs hardness scale and whether they react with water or any other materials you may be using.

Goals
I aim to have small (5mm-10mm wide) chips of bornite polished so that the surface is relatively scratch-free and a high enough grade for electron microprobe analysis.

Materials
Materials marked with an asterisk need at least one item per mounting ring. One mounting ring can hold 1-3 mineral chips. It is important to determine how many samples you intend to polish beforehand.

Mounting samples for polish
• Using a rock hammer, or other available tools, break apart mineral samples into chips 10 mm or smaller in width. Make sure to add an extra 3-4 chips onto the total number of chips for each mineral you need-as you adjust to the softness or hardness of each mineral, you may polish away entire chips by mistake.
• Place the scrap of cardboard on a flat surface. Lay a strip of double-sided tape across the cardboard, then secure the plastic mounting rings along the tape.
Use the tweezers to place anywhere from 1-3 mineral chips inside the rings.
Try to have the most-flat edge facing down against the tape inside the ring.
Make sure there is some amount of space between each chip when placing multiples in a single ring.
• Next, fill each mounting ring with resin powder. You do not need to cover the mineral chips entirely; a 1-2 mm thick coating should be enough to prevent the resin from breaking during polishing.
• Use the self-curing liquid from the orthodontic resin kit to wet the powder in the mounting rings. Make sure the powder is completely covered (it shouldn't take more than a couple of drops). It's okay if some leaks out of the bottom.
• Allow the resin time to harden. It will take around a half hour to air dry. The cardboard can also be carefully moved to a hot plate set on low to shorten the drying time. To test if the resin is hardened, you can tap it with a scoopula or pen. It should be completely solid.
• Carefully remove the mounting rings from the cardboard.

Polishing Samples
• Wet the 180 grit polishing circle with DI water (you do not need to use a polishing wheel). Take one of the mounting rings and use figure eight movements to drag the sample around. You can polish aggressively to achieve a flat surface. Try to remove any craters or small blemishes that appear on the surface. (Note: depending on the hardness of your mineral, you may need to start this step at a higher or lower grit. This step works well for bornite (hardness 3.5), but is too aggressive for covellite (hardness 1), which should use a 400 grit polishing circle at this stage.) • Throw the 180 grit polishing circle away when you are satisfied with the surface area exposed. Polishing circles should not be reused. Gently rinse off your sample with water. Take out the 400 grit polishing circle and wet it.
Using figure eight movements, polish the sample until all polishing wheel scratches from the 180 grit circle are removed. (Note: depending on the hardness of your mineral, you may need to start this step at a higher or lower grit. This step works well for bornite (hardness 3.5), but is too aggressive for covellite (hardness 1), which should use a 600 grit polishing circle.) • Repeat the previous step using the 600 and 1200 polishing circles, occasionally checking samples under a light microscope to make sure polishing scratches from higher grits are being removed with each finer polish.
• Set your sample aside. Place the TexMet C cloth onto the polishing wheel and wet using 3 μm monocrystalline diamond suspension fluid. Turn the polishing wheel on to its lowest speed setting. Firmly grasp the sample and begin polishing it on the wheel using gentle figure eight movements. Be sure to occasionally rotate the sample in your grip to avoid wearing down only one side of the sample-you want the sample as flat as possible.
• Rinse your sample with water. Throw away the polishing cloth and replace it with another TexMet C cloth. Repeat the last step using 1 μm and 0.25 μm monocrystalline diamond suspension fluid. Regularly check samples under a light microscope to make sure polishing scratches from higher grits are being removed with each finer polish.
• The sample should be extremely reflective with very few scratches or blemishes left on the surface (some polishing scratches are unavoidable, but there should be large areas of unscratched surface). If there are polishing scratches still on the surface, continue polishing until they are gone. If scratches persist, move to a coarser grit grade until they are removed and work back down to a finer polish.
• Soak samples in acetone until resin becomes slimy and loose enough to free mineral chips. If acetone does not dissolve, try cutting the resin to expose a greater surface area.
• Repeat the last two sections (mounting and polishing) so that you are polishing the opposite side of the mineral chip. If you only need one side highly polished, you can simply polish using the 180 grit circle until you reach the desired thickness for the mineral chip. Re-soak samples in acetone to remove resin, rinse with water, and let dry.

Mounting Polished Samples
• While allowing samples to dry, use the glass etcher to cut the microscope slides into approximately 1 cm x 1 cm squares.
• Check that samples are dry before continuing. Place a small amount of superglue onto the center of the glass slide. Using tweezers, place your sample with the side you wish to analyze facing up. Gently press down on the surface using a Kim wipe. Allow to sit under a heat lamp for 5 to 10 minutes. Test the bond of the glue to the glass and sample using tweezers-it should not move.
• Store samples where appropriate and clean work station.

Conclusion
The procedure, as written, works very well for the mineral bornite and chalcopyrite. • Fill glass media bottle with 1 L purified water and add 25 g of culture media.
Screw on the bottle cap and shake and swirl to thoroughly mix the water and media.
• Place glass bottle in an autoclave-safe plastic bin. Unscrew cap and rest it on the edge of the bottle lid so that half the opening is exposed.
• Autoclave at 121⁰C for at least 15 minutes.
• Screw cap back onto bottle while still inside the autoclave to keep sterile.
Allow the bottle and liquid media time to completely cool.
• Once cool, label container and put away. Liquid media can be stored at room temperature.

Incubating Cultures
• Place bottle of LB culture media, sterile 3 ml disposable pipettes, Kim wipes, ethanol, and sterile 6-well plate into the fume hood. Put on latex gloves and wash hands with ethanol. Wipe down the section of the fume hood you plan to work on using Kim wipes and ethanol. Clean the outside of the 3 ml pipettes and 6-well plate with Kim wipes and ethanol. Anytime you accidentally touch a non-sterile surface with your gloves or with a tool, you need to clean it again using ethanol-it is imperative to keep out any unwanted contamination.
• Unwrap your 6-well plate, placing it so that the underside is facing up.
Unscrew the culture media bottle, placing the bottlecap underside face up. • Cover the 6-well plate with its lid and appropriately label. Taping the edges of the well plate together is recommended if you plan to leave the well plate undisturbed for more than a few hours.
• Carefully move the well plate to a safe place where it can be stored at room temperature. Fume hoods can be too cold and dry, evaporating the liquid before the culture has time to grow.
You should begin to see a cloudy biofilm taking form after approximately 24 hours, with best results being between 48 and 72 hours.

Supplemental data
3.1 P. fluorescens growth curves

Sulfur in Jezero Crater
Jezero crater is located along the Martian dichotomy and was once host to a deep lake, as evidenced by a sizable delta, inverted channels, and point bars. The Jezero watershed and floor units have spectral signatures indicative of olivine, pyroxene, carbonates, smectite clays, and sulfates (Goudge et al., 2015;.  averaged areal abundances on TES (Thermal Emission Spectrometer onboard MGS) data from Jezero crater, as seen in table 1 below.