sign in sign up
<<
Home , CI chondrite

& Planetary Science 54, Nr 9, 2067–2082 (2019) doi: 10.1111/maps.13345

Simulated materials based on carbonaceous

Daniel T. BRITT1, Kevin M. CANNON 1*, Kerri DONALDSON HANNA1,2, Joanna HOGANCAMP3, Olivier POCH4, Pierre BECK4, Dayl MARTIN5, Jolantha ESCRIG4, Lydie BONAL4, and Philip T. METZGER1,6

1Department of Physics, University of Central Florida, Orlando, Florida 32816, USA 2Atmospheric, Oceanic and Planetary Physics, Oxford University, Oxford OX1 4BH, UK 3Geocontrols Systems, Jacobs JETS Contract at NASA Johnson Space Center, Houston, Texas 77058, USA 4Univ. Grenoble Alpes, CNRS, IPAG, 38000 Grenoble, France 5European Space Agency, ECSAT, Fermi Avenue, Harwell Campus, Oxfordshire OX11 0FD, UK 6Florida Space Institute, Orlando, Florida 32826, USA *Corresponding author. E-mail: [email protected] (Received 14 March 2019; revision accepted 30 May 2019)

Abstract–A set of high-fidelity simulated asteroid materials, or simulants, was developed based on the of . Three varieties of simulant were developed based on CI1 (typified by ), CM2 chondrites (typified by Murchison), and CR2/3 chondrites (multiple samples). The simulants were designed to replicate the mineralogy and physical properties of the corresponding meteorites and anticipated asteroid surface materials as closely as is reasonably possible for bulk amounts. The simulants can be made in different physical forms ranging from larger cobbles to fine- grained regolith. We analyzed simulant prototypes using scanning electron microscopy, X- ray fluorescence, reflectance spectroscopy at ambient conditions and in vacuum, thermal emission spectroscopy in a simulated asteroid environment chamber, and combined thermogravimetry and evolved gas analysis. Most measured properties compare favorably to the reference meteorites and therefore to predicted volatile-rich asteroid surface materials, including boulders, cobbles, and fine-grained soils. However, there were also discrepancies, and mistakes were made in the original mineral formulations that will be updated in the future. The asteroid simulants are available to the community from the nonprofit Exolith Lab at UCF, and the mineral recipes are freely published for other groups to reproduce and modify as they see fit.

INTRODUCTION from NEAs and use these resources to support a space- based economy (Metzger et al. 2013; Graps et al. 2017). Only a handful of spacecraft have come close There are compelling science reasons to study asteroid enough to directly interact with the surface material of regoliths. Their compositions and physical properties an asteroid, but this will be changing in the near future. provide valuable information about the nebular Both the Origins, Spectral Interpretation, Resource environment and planetary building blocks, and about Identification, Security, Regolith Explorer (OSIRIS- diverse processes including impact cratering (e.g., REx) and Hayabusa2 missions are attempting to sample Miyamoto et al. 2007), space weathering (e.g., Clark material from the surfaces of C-complex et al. 2002), Yarkovsky and YORP effects (e.g., (Tsuda et al. 2013; Lauretta et al. 2019), the Asteroid Vokrouhlicky et al. 2015), aqueous alteration and Impact and Deflection Assessment mission will use a thermal heating (e.g., Keil 2000), and regolith formation kinetic impactor to deflect a near-Earth asteroid (NEA; itself (e.g., Delbo et al. 2014). For practical applications, Cheng et al. 2015), and a burgeoning group of many C-complex asteroids contain abundant mineral- commercial companies has plans to extract resources bound OH and H2O (Rivkin 2012) that can be used to

2067 © The , 2019. 2068 D. T. Britt et al. produce propellants (Lewis 1996), and regolith of any composition can be used as radiation shielding (Pohl and Britt 2017) or as a construction material for structures in space (Mueller et al. 2016). However, currently there is no bulk asteroid material available to support scientific experiments or technology maturation for asteroid missions. The original Hayabusa mission returned ~1500 grains from the asteroid Itokawa’s surface: these were invaluable for linking the petrology of ordinary chondrites to S-complex asteroids (Nakamura et al. 2011), but material from sample return missions is too precious for bulk regolith studies and for developing and testing hardware for future missions. Additionally, regoliths on S-complex asteroids likely differ substantially from the more volatile-rich (i.e., ≳1wt%equivalent H2O) surfaces of C-complex asteroids that are currently a focus for many current and future missions. The objective of this work is to develop and characterize simulated C-complex asteroid materials, or simulants, based on the mineralogy and physical properties of well-characterized carbonaceous chondrite meteorites. In the past, many regolith simulants with well- known acronyms have been produced for the (e.g., JSC-1; McKay et al. 1994, NU-LHT; Stoeser et al. 2010) and Mars (e.g., JSC Mars-1; Allen et al. 1998, MGS-1; Cannon et al. 2019); a listing is maintained at http://sciences.ucf.edu/class/planetary-simulant-database/. Asteroid simulants have been developed in a more ad hoc way, and no standardized simulants currently exist. Metzger et al. (2019) provided a detailed list of past efforts to simulate asteroid surface materials: in many cases, these are low-fidelity substitutes such as glass beads, lunar simulant, and bread flour, among others. In the first effort to systematically create high-fidelity asteroid simulants, the Fig. 1. Photographs of the asteroid simulants created in this work, in regolith form. a) CI simulant based on the Orgueil University of Central Florida (UCF) and Deep Space . b) CM simulant based on the Murchison meteorite. Industries (DSI; now part of Bradford Space) developed a c) CR simulant based on an average of five CR chondrites. series of prototypes with physical forms including solid (Color figure can be viewed at wileyonlinelibrary.com.) cobbles and regolith. Originally, these simulants were sold for profit by DSI; however, after closing their Orlando SIMULANT DESIGN PHILOSOPHY facility, the simulant production and distribution have been absorbed by the Center for Lunar & Asteroid Surface Composition and physical property measurements Science (CLASS) at UCF and transitioned to a nonprofit of carbonaceous chondrites provide a reference model distribution model. The goal of the rebranded Exolith Lab from which to develop simulated C-complex asteroid (https://sciences.ucf.edu/class/exolithlab/) is to produce and materials, keeping in mind that some asteroids or distribute a line of standardized simulants for asteroids, the portions thereof may not be represented in the known Moon, and Mars (Cannon et al. 2019). In addition, we meteorite collection. Current thinking links carbonaceous maintain a library of raw minerals to rapidly prototype chondrites to the C-complex asteroids as parent bodies customized versions for specific cases. In this contribution, (e.g., Johnson and Fanale 1973; Rivkin 2012). These three asteroid simulants are described (Fig. 1) based on (1) meteorite samples—particularly the CIs, CMs, and highly volatile-rich CI carbonaceous chondrites (CI many CRs—currently provide the best source of Asteroid Simulant), (2) moderately volatile-rich CM information on the anticipated mineralogy and physical carbonaceous chondrites (CM Asteroid Simulant), and (3) properties of volatile-rich asteroid surface material. less volatile-rich CR carbonaceous chondrites (CR Remote sensing of low-albedo asteroids provides Asteroid Simulant). somewhat limited compositional information, especially Simulated asteroid materials 2069 in the visible/near-infrared (VNIR) because reflectance Despite these challenges, moderately to strongly spectra tend to be dark and muted (Bus and Binzel aqueously altered carbonaceous chondrites (petrologic 2002). The 3-lm region and mid-infrared thermal types 1 and 2) are mineralogically dominated by emission spectra add more value, for example, in phyllosilicates (informally, clays) with lesser amounts of determining relative hydration (Rivkin 2012), distinguishing primary refractory phases (Bland et al. 2004; Howard between different phyllosilicate types (e.g., De Sanctis et al. 2009, 2015). The clays formed either in the et al. 2015; Donaldson Hanna et al. 2019), and constraining protoplanetary disk itself (Ciesla et al. 2003) or more physical properties through thermal modeling (e.g., likely through varying degrees of aqueous alteration on Licandro et al. 2011). A link between CM chondrites parent bodies (Krot et al. 2005; Pravdivtseva et al. 2018). and the Ch/Cgh asteroid spectral class has proven Hydrated and hydroxylated matrix species include robust (Rivkin 2012), but aside from this, it has smectite-group minerals (e.g., saponite), serpentine-group generally not been possible to directly tie a given minerals (notably, Fe-rich serpentines in CMs represented asteroid to a specific meteorite or meteorite group. by the cronstedtite endmember), tochilinite (a layered However, every meteorite came from some . metal-hydroxide/sulfide), and iron hydroxides including As such, the collective suite of meteorite samples gives a ferrihydrite. Other common minerals include , (Fe, sense of what types of asteroid materials are likely to be Ni) sulfides, pyroxene, , native (Fe, Ni) encountered in the NEA and main belt populations, metal, sulfate salts, and carbonates. Although technically recognizing that there are strong biases from material not a mineral, insoluble organic matter (IOM) is present strength and orbital dynamics. at weight percent levels (typically 0–5 wt% C) in carbonaceous chondrites (Hayatsu et al. 1977; Cody and Mineral-Based Design Alexander 2005), and given its low grain density, this phase is volumetrically significant enough that it should Minerals are the basic building blocks of planetary be considered an essential part of the simulant materials, and thus, mineral-based simulants offer the formulation. closest potential match to the properties of actual The CI, CM, and CR asteroid simulants are based asteroid materials. The unique combination of minerals on the mineralogy of their respective chondrite groups. present in asteroidal material will determine its chemical For CI and CM, Orgueil (CI1) and Murchison (CM2) composition, grain density, spectral, magnetic, and were chosen as references, respectively, because they are (thermo)physical properties as well as the behavior of well studied and their modal mineralogy is well known regolith formed from the breakdown of originally (Bland et al. 2004; Howard et al. 2009). There is much lithified material. We define simulants based on the more variation among CR mineralogies (Howard et al. mineralogy of a reference material as “high fidelity,” 2015); therefore, the CR simulant was based on an while those based purely on bulk chemistry or average of five Antarctic CR chondrites: Pecora Escarpment geomechanical properties are “low fidelity.” (PCA) 91082 [CR2], Graves Nunataks (GRA) 95229 It is challenging to determine the modal mineralogy [CR2], LaPaz Icefield (LAP) 02342 [CR2], Queen of carbonaceous chondrites by point counting because Alexandra Range (QUE) 99177 [CR3], and Graves they tend to be dominated by a highly complex matrix Nunataks (GRA) 06100 [CR2]. Table 1 lists the original made up of micron- and submicron-sized grains. Matrix modal mineralogy determined by previous authors for fractions exceed 70 vol% in CIs and CMs (Brearley and Orgueil, Murchison, and the CRs listed above. The Jones 1998). Bland et al. (2004) described the challenges mineral recipes for the prototype simulants were based of quantifying phase abundances using conventional on these data, with changes made because of various X-ray diffraction (XRD), and multiple authors have trades involved in sourcing input materials and producing approached this problem using position sensitive detector less hazardous simulants. Some species such as XRD (PSD-XRD) instead (Bland et al. 2004; Howard cronstedtite and tochilinite do not occur in concentrated et al. 2009, 2015). However, even with this technique, it is deposits on Earth, and are difficult to synthesize at scale, difficult to distinguish between interstratified and such that other phyllosilicates and sulfides were complicated alteration phases. Monikers like “poorly substituted. We specifically sourced nonasbestiform characterized phases” abound in older meteorite literature, serpentine (antigorite/lizardite) for safety concerns, rather and many phases require beam techniques to properly than chrysotile. Coal was used as a substitute for interrogate (e.g., Barber 1981; Mackinnon and Zolensky meteorite IOM. After these substitutions, initial recipes 1984). These high-resolution techniques are difficult to were then adjusted in an iterative process between UCF link back to bulk modal mineralogy, and for example, and DSI to arrive at the prototype simulant recipes listed Bland et al. (2004) generally did not distinguish between in Table 2 and described in this work. In some cases, different phyllosilicate species in their modal analyses. there were substantial deviations from the intended 2070 D. T. Britt et al.

Table 1. Quantitative mineralogy of reference meteorites determined by PSD-XRD. Orgueil Murchison Murchison PCA 91082 GRA 95229 LAP 02342 QUE 99177 GRA 06100 Phase (CI1)a (CM2)a (CM2)b (CR2)c (CR2)c (CR2)c (CR3)c (CR2)c Saponite- 64.2 serpentine Magnetite 9.7 0.4 1.1 0.9 0.2 Mg-rich 7.3 22.8 22.2 15.3 8.8 7.1 1.5 11.4 serpentine Olivine 7.2 11.6 15.1 26.5 37.6 43.7 36.7 40.1 Ferrihydrite 5.0 Fe-sulfide 6.6 3.4 1.9 4.7 3.2 3.4 5.3 7.1 Fe-rich 58.5 50.3 serpentine Pyroxene 2.2 8.3 31.6 36.8 28.3 35.9 33.0 Calcite 0.1 1.2 Fe,Ni metal 3.6 6.2 7.1 5.6 2.7 X-ray 18.5 7.8 10.8 14.1 5.8 amorphous Fe-rich Oxide 2.0 7.0 4.0 3.0 6.0 (magnetite + rust) aFrom Bland et al. (2004), given as wt%. bFrom Howard et al. (2009), given as vol%. cFrom Howard et al. (2015), given as vol%. mineral-based approach, and these issues and future density and porosity (Britt and Consolmagno 2003; remedies are described in more detail below. Consolmagno et al. 2008; Macke et al. 2011) and thermophysical properties (Opeil et al. 2010). Grain Physical Properties densities for CIs, CMs, and CRs vary between 2.20–2.38, 2.57–2.87, and 2.92–3.47 g cmÀ3, respectively; average The physical properties of asteroid surface materials porosities for these three types are 11.3%, 23.0%, and are controlling factors for developing spacecraft 6.4% (Britt and Consolmagno 2003). The grain densities hardware that will interact with these bodies: grapple, are much lower than other chondrite types, and thermal anchor, land, rove, and excavate. Physical properties, conductivities are significantly lower as well (Opeil et al. therefore, must be considered for realistic simulants. 2010). Literature for compressive and tensile strengths of Asteroids are covered in an unconsolidated regolith with carbonaceous chondrites is severely limited. The available polymineralic grains ranging in size from microns to data (Pohl and Britt Forthcoming) suggest that they are massive boulders. There appears to be a distinct trend of much weaker than ordinary chondrites and iron decreasing average grain size with increasing mass, or meteorites, as might be expected. Strength varies among gravity field of the asteroid (Gundlach and Blum 2013), different carbonaceous chondrite types, with a general such that smaller asteroids are blockier while larger trend of CM > CI > identified by bodies retain more fines and may closely resemble the Tsuchiyama et al. (2009). Compressive strengths cited for lunar surface in terms of regolith texture. Initial imagery CIs are typically several MPa (similar to Styrofoam), from the Hayabusa2 and OSIRIS-REx missions suggests while CMs can reach 50–100 MPa, stronger than smaller C-complex asteroid surfaces are indeed strewn concrete (Pohl and Britt Forthcoming). There are with large cobbles and boulders (Fig. 2d). currently no strength data available for CRs. Our asteroid simulants are not simply a mixed powder of the minerals in Table 2, but have been DEVELOPING PROTOTYPE SIMULANTS developed to broadly replicate different forms of asteroid surface materials including larger cobbles (Fig. 2c) and To develop asteroid simulant prototypes, we (1) finer grained regolith (Fig. 1). The physical properties of sourced the raw materials (Table 2; Fig. 2a) in as pure carbonaceous meteorites offer clues about the nature of a form as possible from terrestrial mines and chemical coherent, lithified C-complex asteroid materials. These suppliers; (2) comminuted, sieved, and analyzed the raw meteorites have been characterized extensively in terms of materials; (3) mixed the minerals for a given simulant Simulated asteroid materials 2071

Table 2. Initial simulant recipes used to create prototypes, and suggested updates to the recipes based on analyses of these prototypes. All values in wt%. Mineral CI prototype CI future update CM prototype CM future update CR prototype CR future update Mg-serpentine 48.0 51.3 72.5 73.8 9.2 6.8 Epsomite 6.0 ––––– Magnetite 13.5 10.0 10.4 5.0 14.3 2.5 Palygorskite 5.0 5.3 –– –– Olivine 7.0 7.0 7.8 11.2 33.7 33.1 Pyrite 6.5 7.0 2.6 3.3 4.1 5.8 Vermiculite 9.0 9.6 –– –– Coal 5.0 5.0 3.6 3.6 2.0 2.0 Pyroxene –– 2.1 2.1 31.6 29.6 Siderite –– 1.0 1.0 –– Iron metal –– – – 4.6 10.6 Nickel metal –– – – 0.5 – Ferrihydrite – 4.8 –– –– Amorphous silicate –– – – – 9.6

according to the recipes in Table 2 to form a dry mixture raw materials are in the tens of micron range: this is larger (Fig. 2a); (4) bound the minerals with water and in some than carbonaceous chondrite matrix, but this was deemed cases a binder to form a malleable paste (Fig. 2b); (5) to be an acceptable trade given the difficulty in colloidal dried or cured the paste and crushed the resulting grinding for bulk quantities. We analyzed each raw material to create cobbles and finer regolith (Figs. 1 and ingredient using XRD, X-ray fluorescence (XRF), and 2c). These steps are described in more detail below. VNIR reflectance spectroscopy to verify the mineralogy, quantitatively measure the chemistry, and identify any Sourcing Raw Materials contaminants present. Minor contaminants were identified in many of the naturally sourced minerals and are listed in Many of the minerals in Table 2 are available from Table S1. In theory, it would be possible to synthesize pure terrestrial mine sources or chemical suppliers. In some versions of some of these phases from raw oxides, but this is cases, the pure minerals are sold commercially under not practical for large-scale production. The crystal obscure trade names, for example, palygorskite as the chemistry of some natural terrestrial minerals also differs product MIN-U-GELÒ. In other cases, the minerals are from minerals found in meteorites, with higher being quarried for aggregate, or they make up the host concentrations of Mg and Al, and lower amounts of Fe. rock of precious metal deposits and are available as Again, these discrepancies were judged to be reasonable waste. Table S1 in supporting information lists the given the trades involved in using a cost-effective approach sourcing information for all the materials in our to produce large amounts of simulant. asteroid simulants; however, these specific sources are not an inherent property of the simulants and Creating Dry Mixtures mineralogically appropriate alternatives can be used to re-create similar products. After the raw source materials were processed, they were mixed together in the appropriate weight ratios Preparing and Analyzing Raw Materials (Table 2), resulting in homogeneous medium to dark gray powders (Fig. 2a). Mixing can be done using The raw minerals (Fig. 2a; Tables 2 and S1) came from stirring tools or by hand at small scales; for larger suppliers in a variety of physical forms (powders, flakes, scales, a concrete mixer is appropriate. The resulting rocks) and were processed and analyzed prior to combining dry mixtures can be stored in a sealed environment them to create the prototype simulants. We used a variety indefinitely and later processed into different physical of equipment including rock crushers, ball mills, and sieves forms as desired. For the prototypes described here, to comminute and then separate the raw materials into fine- mixing was performed in a Kushlan Products 600 DD grained powders. Materials such as the serpentine and concrete mixer for 6–12 h at 28 rpm. Powders were palygorskite came as micron-sized powders, approximating added all at once sequentially. We did not make specific chondrite matrices and requiring no further processing. For efforts to mitigate against humidity during mixing, but the other minerals, the average grain sizes of our processed final products were stored in sealed watertight buckets. 2072 D. T. Britt et al.

Fig. 2. Production methods for the CI Asteroid Simulant. a) Dry mixture of the powdered raw mineral inputs. b) Paste made from combining the dry mixture with water. c) Dried and crushed cobbles of the CI simulant. d) Image of Ryugu’s surface from the Hayabusa2 mission (contrast enhanced, scale bar 1 m); image credit: JAXA. (Color figure can be viewed at wileyonlinelibrary.com.)

Binding sodium metasilicate pentahydrate in water, mixing this solution with the dry mixture (in a similar 1:4 ratio), In order to create cobbles and regolith grains with and curing at an elevated temperature (at least 75°C). coherent strength, we needed to bind the dry mixtures The final strength of the cured blocks is a function of together somehow. We originally experimented with a the concentration of sodium metasilicate used: 2–5% of wide variety of methods. We found adding water to the the dry mixture mass results in blocks with tens of MPa dry mixtures of CI simulant (~1:4 ratio of water:mix by compressive strength, although more work is ongoing to weight) resulted in a malleable paste (Fig. 2b). After air refine this relationship and compare these materials to drying, the paste lost all added water and formed a actual CMs and CRs. At a 5 wt% level, the sodium solid material (Fig. 2c) with several MPa in unconfined metasilicate pentahydrate increases the SiO2 content of compressive strength, matching measurements from CI the simulants by 1.46 wt% and the Na2O content by chondrites (Metzger et al. 2019; Pohl and Britt 1.42 wt%, while the majority of the added water is lost Forthcoming). The clay species appear to be responsible (outgassed) during curing and polymerization. for most of this strength, because we found mixtures with less clay were much less coherent upon drying. A Forming Cobbles and Regolith modified procedure was needed for the CM and CR simulants, because their respective meteorite types are Once solid blocks were formed, they were crushed considerably stronger than CIs. To increase the strength either with hand tools or mechanized crushing of these two simulant types, we used sodium equipment. Crushing generally results in a power law metasilicate to create an alkali-activated material (also particle size distribution (Turcotte 1986), and the known as a geopolymer). This involved dissolving solid minimum, mean, and maximum particle size can easily Simulated asteroid materials 2073 be adjusted by sieving and by changing the duration/ smaller particles and individual minerals being angular intensity of crushing. We created two classes of and larger particles being subrounded to subangular. physical forms for the asteroid simulants: cobbles Many of the larger particles (>50 lm) are agglomerates (Fig. 2c) that are likely to be prevalent on smaller of smaller particles, and the finest particles (<10 lm) asteroids, and fine-grained regolith or soil (Figs. 1 and adhere to the surfaces of larger particles (Figs. 3a and 3) that should dominate the surfaces of larger asteroids 3b). The CM asteroid simulant has a similar distribution (Gundlach and Blum 2013). Cobbles for this work of particle shapes, but with many agglomerate particles were created by comminution with a mallet, and fine- in the <10 lm range. Individual minerals tend to be grained regolith was produced in a 14″ flail-type rock highly angular or platy, with agglomerate particles crusher. Metzger et al. (2019) performed an in-depth being more subrounded to subangular. On the sprinkled analysis of the particle size distribution of the CI sample stub, the CM simulant fines (<5 lm) adhere Asteroid Simulant in regolith form and compared it to together to form longer “chains” or more angular available constraints from meteorite and asteroid agglomerations than at larger grain sizes (Fig. 3c). The observations. CR asteroid simulant has the greatest abundance of fines (<5 lm), which mostly adhere to larger particles SIMULANT CHARACTERIZATION and appear brighter in SEM images than most other particles (Fig. 3d). Many of the smallest particles We characterized the asteroid simulant prototypes (<0.5 lm) are spherical and strongly adhere to each using a variety of instrumental techniques relevant to other and to larger particles. There are fewer agglomerate remote sensing and future missions including resource particles <50 lm than the CI and CM simulants, but extraction. A companion paper (Metzger et al. 2019) the shape of these larger particles remains comparable describes a rigorous effort to quantitatively assess the (i.e., subrounded to subangular). fidelity of the CI simulant using a modified version of the “Figure-of-Merit” approach developed by NASA Bulk Chemistry for lunar simulants. That paper focused more on physical properties, while the analyses below address The bulk elemental chemistry of the prototype microscopic textures, bulk chemistry, VNIR reflectance simulants was measured using crushed powders with a spectra, thermal emission spectra, and volatile release PANalytical Epsilon XRF spectrometer at UCF, properties for all three simulant varieties. operating in oxides mode. Expressing compositions of chondrites as oxides (SiO2, MgO, etc.) instead of metals Scanning Electron Microscopy (Si, Mg) is common practice because these materials are dominated by -complexed cations. Because the All three simulant prototypes were imaged using a simulants are not meant to reproduce the trace element LEO 1530 VP scanning electron microscope (SEM) at or isotopic compositions of the reference meteorites, the European Space Agency’s Advanced Manufacturing these were not measured. Table 3 reports the major Laboratory to visualize the textures present in the element chemistry of the three prototypes. All three are polymineralic composite particles. The samples were chemically dominated by subequal amounts of SiO2, prepared for imaging using adhesive pads MgO, and FeOT, as is true of carbonaceous chondrites attached to SEM sample stubs. Two methods of in general (Jarosewich 1990). Compared to carbonaceous preparation were used, resulting in a “heavily pressed” chondrites, the simulants have significantly higher Na2O stub and a “sprinkled” sample stub for each simulant. and K2O concentrations, a consequence of using The pressed stub was made by pressing the stub into terrestrial minerals as feedstocks as well as the sodium bulk material and then gently tapping the stub to metasilicate used as a binder in the CM and CR remove any excess material. The sprinkled stub was simulants. Even though the simulants were not created made by gently dropping the sample from a height of as chemical analogs, their bulk chemistry is fairly ~2 cm, with the stub held at a high angle relative to the accurate. However, the CR prototype is a notable infalling material. Preparing the samples using these exception: its unrealistically high magnetite content different methods allows a qualitative assessment of resulted in a bulk chemistry that is strongly enriched in some of the sample’s physical properties, such as FeOT and depleted in SiO2 relative to CR chondrites. cohesiveness. Figure 3 shows two different views of the As discussed below, this issue will be remedied by fine fraction (<1 mm) of the CI Asteroid Simulant: the making changes to the mineral recipe. We did not sprinkled stub (Fig. 3a) and a magnified view of specific measure the speciation of Fe (Fe0,Fe2+,Fe3+) in the particles of the pressed stub (Fig. 3b). Particle shapes in simulants, but this will be done in the future with this simulant are subrounded to highly angular, with Mossbauer€ spectroscopy. 2074 D. T. Britt et al.

Fig. 3. Scanning electron micrographs showing composite polymineralic grains that make up the asteroid simulants in regolith form. a) Diversity of particle morphologies and sizes in the CI Asteroid Simulant. b) Zoom in of a CI composite particle. c) Fine-grained agglomerates in the CM Asteroid Simulant. d) Bright (i.e., heavy) fines coating the surface of a larger grain in the CR Asteroid Simulant. Scale bar is 200 lm in (a), 50 lm in (b), and 20 lm in (c) and (d).

VNIR Reflectance Spectroscopy and wide 3 lm absorption with a minimum at 2.9 lm, attributed to the combination of serpentine and smectite Reflectance spectra of the prototype simulants were clays in the mineral recipe. By contrast, the CM acquired at the Institut de Planetologie et d’Astrophysique simulant (dominated by serpentine) shows a much de Grenoble using the spectrophotometer with variable sharper absorption with a minimum at 2.74 lm, similar incidence and emergence (SHINE) spectro-gonio to the spectral properties of pure serpentines (Beck radiometer. Spectra were acquired between 0.4 and 4.2 lm et al. 2018). The CR simulant, with minimal clay with an incidence angle of 0° and an emergence angle of content, shows a small but noticeable shoulder at 30°. For ambient condition measurements, samples were 2.7 lm. Neither the CM nor CR spectra have noticeable prepared by sieving the original simulant powder to 0.7 lm absorptions as is seen in the corresponding <100 lm, then depositing the sieved powder on a 1-mm meteorites (Beck et al. 2018), but this may be because thick sample holder, and flattening the sample surface with of the Fe-rich nature of asteroidal serpentines that are a spatula. For vacuum conditions, the simulants were not accurately replicated in the simulants. Additional unsieved. Spectral measurements were performed in absorptions are observed around 3.48 lminthe ambient conditions and in vacuum, which is important for simulants and are attributed to aliphatic organic species reducing the effects of adsorbed atmospheric water that present in the coal component (Kaplan and Milliken can strongly affect the 3-lm region in particular. 2016). The CI simulant is the only one that shows The resulting spectra are shown in Fig. 4, which significant changes between ambient and vacuum conditions offsets them for clarity but also presents the absolute (Fig. 4). In vacuum, the 3 lm absorption becomes sharper reflectance at 1.25 lm, outside the region of any major and more asymmetric, attributed to less absorbed absorption. All three simulant varieties have low atmospheric water. The vacuum measurement more reflectance values (<0.1) at all wavelengths. The spectra faithfully captures the inherent spectral properties of the are generally flat, with distinct peaks near 0.8 lm clay mixture in the CI simulant, and is better for attributed to maghemite contamination in the magnetite, comparing to remote spectra from asteroids. and complex hydration absorptions associated with There are broad similarities between the prototype –OH stretching and H2O bending in hydrated silicates simulant spectra and reflectance spectra acquired from between 2.7 and 3.6 lm. The CI simulant has a broad the corresponding carbonaceous chondrite samples, but Simulated asteroid materials 2075

Table 3. Simulant prototype major element bulk chemistry, as measured by XRF. CI Simulant CM Simulant CR Simulant Oxide (wt%) (wt%) (wt%)

SiO2 25.0 32.5 19.3 TiO2 0.5 0.3 0.4 Al2O3 3.1 3.1 1.2 FeOT 25.8 20.2 42.9 MnO 0.3 0.2 0.3 MgO 30.2 32.1 30.1 CaO 3.0 3.1 0.8 Na2O 6.4 6.2 3.3 K2O 0.4 0.2 0.1 P2O5 0.4 0.4 0.4 Cr2O3 0.2 0.2 0.1 SO3 4.9 1.5 0.9 Total 100.0 100.0 100.0 Fig. 4. Visible/near-infrared reflectance spectra of the asteroid simulants measured on the SHINE instrument. Spectra have been offset for clarity; solid lines show ambient condition spectra, while dashed lines show vacuum condition. The inset important differences as well. Carbonaceous chondrite shows absolute reflectance at 1.25 lm for the simulants at reflectance spectra are also dark and flat, and have ambient conditions (squares) compared to meteorite data from complex absorptions in the 3-lm region attributed to the RELAB database (circles = mean, bars = min/max). hydrated silicates (Cloutis et al. 2011a, 2011b; Beck (Color figure can be viewed at wileyonlinelibrary.com.) et al. 2018). However, there are mismatches between the absolute reflectance values of the simulants and the meteorites. The inset in Fig. 4 compares the ambient lunar environment (PASCALE) chamber within the simulant measurements to multiple spectra of each Planetary Spectroscopy Facility at the University of chondrite type taken from the RELAB database, filtered Oxford. The experimental setup and calibration are to those that were measured from similar size splits (0– similar to that of the simulated lunar environment 75, 0–100, or 0–125 lm) as the simulant samples (0– chamber (SLEC), which has been described previously 100 lm). The CM simulant falls within the range of by Thomas et al. (2012, 2014) and Donaldson Hanna meteorite measurements, but the CI simulant and et al. (2019). The environment chamber is mounted to particularly the CR simulant are darker than their the emission port of a Bruker VERTEX 70v Fourier corresponding reference meteorites. This is mostly due to transform infrared (FTIR) spectrometer using a cesium the large amount of magnetite in the prototype recipes iodide (CsI) window. We measured thermal infrared (Table 2) that is unrealistic and deviated from the spectra at a resolution of 4 cmÀ1 from ~2400 to original meteorite mineralogies (Table 1). As discussed 200 cmÀ1 (~5to50lm) using a wide-range TIR beam further below, the excess of magnetite is an important splitter and a deuterated, L-alanine doped triglycine issue to address in the future. It is also responsible for sulfate (DLaTGS) detector. We made three spectral the distinct peaks in the spectra near 0.8 lm because of measurements of each sample with each one including the concomitant maghemite contamination (Fig. 4); such 250 scans, which produces a spectrum with sufficient peaks are not observed in meteorite spectra. As well, the signal-to-noise ratio such that we can uniquely identify detailed shapes of the 3-lm region absorptions differ features of 2% contrast from the noise. The final TIR between simulants and meteorites. This is likely due to spectrum for each sample is an average that has been differences in crystal chemistry of clay species and the calibrated using published methods (e.g., Ruff et al. specific ratios of saponite-like phases and serpentine-like 1997) and spectral measurements of a blackbody phases in CIs, for example. painted with Nextel high emissivity black paint at 340 and 360 K (e.g., Thomas et al. 2012, 2014). Thermal Infrared Emission Spectroscopy Near-surface conditions experienced in the regolith of planetary bodies are simulated in PASCALE by We acquired thermal infrared (TIR) emissivity controlling the atmospheric pressure and temperature measurements under ambient (“Earth-like”) and simulated inside the environmental chamber and the way in which asteroid environment (SAE) conditions using the we heat the samples. Previous laboratory studies have planetary analogue surface chamber for asteroid and shown the environmental conditions under which samples 2076 D. T. Britt et al. are measured impart observable spectral changes (e.g., Logan et al. 1973; Henderson and Jakosky 1994; Donaldson Hanna et al. 2012a, 2012b, 2017, 2019; Thomas et al. 2012). We simulated the Earth’s nearly isothermal near-surface conditions by backfilling PASCALE with dry gas (N2) to an atmospheric pressure ~1000 mbar, maintaining an ambient temperature inside the chamber (~300 K), and heating the sample from below to 353 K. These laboratory conditions are most comparable to emissivity measurements available in other spectral libraries including those in the Arizona State University spectral library (e.g., Christensen et al. 2000). We simulated a thermal gradient similar to that experienced in an asteroid near surface environment by evacuating the chamber to vacuum pressures (<10À4 mbar), by cooling the interior of the chamber to <125 K using liquid nitrogen (LN2), and by heating the samples from above to a maximum brightness temperature of ~350 K and below to 353 K. We crushed and sieved each asteroid simulant to particle sizes <75 lm to measure the thermal emission spectra (Fig. 5) of the fine particulate regolith-like simulant. We prepared each of the samples in a repeatable manner to limit the observed spectral differences to the composition and physical properties of the samples. Our techniques are similar to those used in other laboratory facilities to make our measurements comparable to literature values (e.g., Donaldson Hanna et al. 2017, 2019). We poured each sample into the cup and leveled the material using a flat edge across the top. This technique creates a nearly flat surface without compressing the uppermost portion of the sample. The power of the lamp was held constant to ensure that all of the samples were measured under similar thermal environments. The ambient and SAE spectra of the simulants (Fig. 5a) highlight the complex nature of the mixtures and the spectrally dominant particle size in the simulants. The spectra of the CI Asteroid Simulant (Figs. 5a and 5b) are dominated by diagnostic spectral features Fig. 5. Thermal emission spectra of the asteroid simulants. indicative of phyllosilicates including the Christiansen a) Comparison between the CI Asteroid Simulant measured at feature (CF; an emissivity maximum that is a good ambient conditions (dashed line) and in the simulated asteroid indicator of bulk composition; Conel 1969) and both sets environment (SAE; solid line). b) Comparisons between CI, of vibration bands (also known as the reststrahlen bands CM, and CR simulant spectra all acquired under SAE with particle sizes <75 lm. (Color figure can be viewed at wile or RB). Both the CF and vibration bands are consistent yonlinelibrary.com.) with ambient measurements of Orgueil. In the 1800– 1200 cmÀ1 spectral range, we also observe similarities between the CI simulant spectra and Orgueil including Donaldson Hanna et al. 2019). In the 1800–1200 cmÀ1 absorptions indicative of H2O (e.g., Aines and Rossman spectral range, there are similarities between the CM 1984; Salisbury et al. 1991) and organics. simulant spectra and the CM meteorites, although these Similar to the CI Asteroid Simulant, the spectra of features are at slightly different frequencies in the the CM Asteroid Simulant (Fig. 5b) are dominated by meteorite spectra. The largest spectral differences diagnostic spectral features indicative of phyllosilicates. between the CI simulant spectra and the CM simulant These features are consistent with ambient measurements spectra are the 1800–1200 cmÀ1 spectral range, where the of Murchison and ALH 831000 (both CM chondrites; CM simulant spectra have more subdued features. Simulated asteroid materials 2077

The spectra of the CR simulant (Fig. 5b) seem to anticipated to be present as a physiadsorbed phase represent a more complex mixture of hydrated and (released at low temperatures), and/or as interlayer and anhydrous phases. The diagnostic features in the structural water and hydroxyl groups (released at vibration band regions are similar to those in the CI moderate to high temperatures). High temperature H2O and CM simulant spectra, but the shapes are wider release is a characteristic of dehydroxylation in and more subdued. The CF position is consistent with phyllosilicates. C3H3 is a good indicator of benzene phyllosilicates, but the drop-off in emissivity at higher rings, tracking the thermal breakdown of the organic frequencies is less than what is observed in the CM molecules present in the coal, analogous to meteorite and CI simulants. The differences in the spectral IOM. CO2 is also an anticipated decomposition product slopes in this region are likely due to the CR simulant of organics, as well as a breakdown product of the being of coarser particle size fractions (because it siderite in the CR simulant. SO2 may be evolved by the contains fewer clay minerals). In the 1800–1200 cmÀ1 thermal breakdown of epsomite and organics, as well as spectral range, we do not observe spectral features oxidation of sulfide phases. The TG curves (Fig. 6a) all indicative of mineral-bound H2O and organics. In show modest mass loss from ambient to ~600°C, general, the CR simulant spectra are similar to MIL followed by more rapid losses from 600 to 1000°C 090001 (CR meteorite; Donaldson Hanna et al. 2019), attributed to H2O loss from dehydroxylation of but the most differences are observed between the CR phyllosilicates and decomposition of other phases. The simulant and a CR meteorite. This is most likely due total integrated mass loss was CI (14.3 wt%) > CM to the range of compositions observed in CR (11.2%) > CR (3.2%), which matches both the relative meteorites. trends observed in the respective meteorite types, as well as the range of absolute values for these meteorites Thermogravimetry and Evolved Gas Analysis (Jarosewich 1990; Garenne et al. 2014). Previous studies using TG alone reported values >25 wt% H2Oin Combined thermogravimetry (TG) and evolved gas Orgueil (King et al. 2015); however, without supporting analysis (EGA) were performed at NASA Johnson EGA data, it is not clear that attributing 100% of that Space Center Astromaterials Research and Exploration mass loss to H2O is reasonable. In the simulants, Science (ARES) laboratories to determine the volatile distinct H2O releases are observed at both low and high release patterns of the three prototype simulants. temperatures (Figs. 6b–d), related to the presence of Samples were measured using a Labsys EVO water bound in different ways as described above. A differential scanning calorimeter/thermal gravimeter single sharp spike in the m/z 39 channel (C3H3) was connected to a ThermoStar quadrupole mass observed around 530°C in all the simulants, attributed spectrometer. Measurements were made under a 30-mbar to the breakdown of complex organic molecules present helium ; this provides a neutral low-pressure in the coal component. The release of CO2 showed environment, albeit not a high vacuum as might be complex patterns in all the simulants, with multiple present in future asteroid mining operations. Samples were overlapping releases between 500 and 850°C, a rise heated from ambient to 1000°Cataramprateof toward 1000°C, and a minor release around 400°Cin À1 35°C min and a flow rate of 10 sccm He. Evolved the CR simulant attributed to siderite. SO2 showed only volatile species of particular interest included H2O(m/ minor releases from 450 to 900°C, with a more complex z = 18), C3H3 (m/z = 39), CO2 (m/z = 44), and SO2 (m/ release pattern in the CI simulant which had significant z = 64). These species are predicted to be evolved from epsomite in the prototype recipe. thermal decomposition reactions of the phases in Table 2. The results—mostly the TG—can also be compared to DISCUSSION previous TG measurements of carbonaceous chondrites (King et al. 2015), but we do not have data for all three Mistakes and Future Updates chondrite types (CI, CM, CR) acquired with a similar instrument setup as in this work. Mistakes were made along the development path of The TG/EGA results show that volatile release is the asteroid simulants, and in most cases, these dominated by H2O at both high and low temperatures represented deviations from the original philosophy of due to thermal decomposition reactions. Volatile release sticking as closely to the meteorite mineralogy as patterns for the three prototypes are shown in Fig. 6: possible. The effects of these errors became apparent in Fig. 6a shows TG curves and integrated mass loss for the instrumental results described above. Table 2 lists all three simulant varieties (inset), while Figs. 6b–d both the original mineral recipes used for the prototypes show EGA data for key mass fragments (m/z 18 = H2O, described in this work, and suggested updates for future m/z 39 = C3H3, m/z 44 = CO2, m/z 64 = SO2). H2Ois production that more closely match the original 2078 D. T. Britt et al.

a b

c d

Fig. 6. Thermogravimetry (TG) and evolved gas analysis (EGA) from the asteroid simulants. a) TG for all three simulants; inset shows total mass loss from 0 to 1000°C. b–d) EGA for the CI, CM, and CR simulants, respectively, showing temperature- dependent release of key volatile species (same vertical scale). (Color figure can be viewed at wileyonlinelibrary.com.) meteorite mineralogies (Table 1). In the original mineral mineralogy from Bland et al. (2004) did not identify recipes (Table 2), the CR data from Howard et al. crystalline sulfates. In the updated recipe, we replaced (2015) were interpreted as wt% when they are instead epsomite with ferrihydrite to better reproduce Bland given in vol%. This resulted in modest underabundances et al.’s (2004) CI mineralogy, and to avoid issues with of the dense minerals (metals, sulfides, etc.) and sulfate efflorescence. overabundances of the less dense minerals. We addressed From here on out, the lab at UCF will maintain this issue in the updated recipes (Table 2). Additionally, rigorous internal versioning to track any changes in the amount of magnetite crept up during the iterative mineral recipes, or in suppliers of the source materials. In development stage to large and unreasonable concentrations. effect, the prototypes described in this work are now This was done in an attempt to both darken the frozen in as “version 1.0”; when the updates from Table 2 simulants and make up for perceived Fe deficiencies, come into effect, this will mark a new version although but it strayed from the original goal to make mineral- for simplicity and clarity, the version numbers will likely based simulants. We found that chemically produced not be tacked onto the names, leaving them as “UCF CI magnetite is significantly darker than natural magnetite Asteroid Simulant,” etc. We will thus discard the earlier from mine sources, and switching the supply to chemical DSI versioning scheme that used the naming convention magnetite can help maintain low albedos typical of “CM-2,” easily confused for the metamorphic grade of carbonaceous chondrites at reasonable magnetite the meteorite (these earlier DSI prototypes were not concentrations. The updated recipes (Table 2) have widely distributed). Any discrepancies or questions in the significantly less magnetite, particularly in the CR future will be able to be resolved by cross-checking simulant that was the least chemically and spectrally shipment dates with internal version tracking. realistic(Table3;Fig.4).Another issue involved epsomite in the original CI recipe dissolving during the paste Realism formation process, then precipitating as a white crust on the dried cobbles (Fig. 2c). While sulfates are Overall, the prototype simulants as described are found in carbonaceous chondrites, in some cases, these mostly accurate representations of anticipated volatile- are terrestrial alteration products, and the Orgueil rich asteroid surface materials. In their current form, Simulated asteroid materials 2079 they are appropriate for addressing scientific and most cases, the analyses compared favorably to the engineering questions, and the simulants will be references; however, there were also significant discrepancies improved in the future based on updates to the mineral in bulk chemistry and reflectance spectra in some cases recipes (Table 2) and new information coming from that will be addressed in the future by updating the both meteorite studies and the Hayabusa2 and OSIRIS- simulant mineral recipes. There are a wide range of REx missions. In terms of quantitative metrics, Metzger applications for the asteroid simulants, and they have et al. (2019) adopted the figure-of-merit (FoM) system already seen use in diverse applications. The simulants developed by NASA, and they evaluated the fidelity of will be distributed through the nonprofit Exolith Lab at the CI Asteroid Simulant as an example. In this system, UCF, and the mineral recipes are also published for numerical scores are given from 0 (lowest fidelity) to 1 others to use and modify as they see fit. (highest fidelity) based on comparisons to a reference material, in this case the meteorite Orgueil and asteroid Acknowledgments—This work has been supported by models for some properties like particle size distribution. NASA Award NNA14AB05A and by the NASA Solar The prototype CI simulant received high FoM scores for System Exploration Research Virtual Institute. O.P. magnetic susceptibility (0.96), elemental chemistry (0.94), acknowledges a postdoctoral fellowship from CNES for and mineralogy (0.83). FoMs for other physical properties funding. We thank an anonymous editor for their were somewhat lower, ranging from 0.53 to 0.77. careful comments that improved the manuscript. However, these still compare quite favorably to FoMs for well-known lunar simulants that score as low as 0.25–0.35 Editorial Handling—Dr. Edward Cloutis (Metzger et al. 2019). The changes to the mineral recipes in Table 2 will bring the mean mineralogy FoM for all REFERENCES three simulants up from 0.87 for the prototypes to 0.97 for the improved recipes. We anticipate that the fidelity of Aines R. D., and Rossman G. R. 1984. Water in minerals? A other properties will improve as a result. peak in the infrared. Journal of Geophysical Research 89:4059–4071. Applications Allen C. C., Morris R. V., Jager K. M., Golden D. C., Lindstrom D. J., Lindstrom M. M., and Lockwood J. P. 1998. Martian regolith simulant JSC Mars-1 (abstract #1690). 29th Lunar The asteroid simulants have already been delivered and Planetary Science Conference. CD-ROM. to researchers worldwide, and there are a variety of Barber D. J. 1981. Matrix phyllosilicates and associated applications for these materials in basic science minerals in C2M carbonaceous chondrites. Geochimica et – advances as well as technology maturation (e.g., Boivin Cosmochimica Acta 45:945 970. Beck P. B., Maturilli A., Garenne A., Vernassa P., Helbert etal.2018;Friesetal.2018).Fromascientificstandpoint, J., Quirico E., and Schmitt B. 2018. What is controlling simulated asteroid materials can be used to investigate (1) the reflectance spectra (0.35-150 lm) of hydrated (and regolith behavior in microgravity, including cohesion and dehydrated) carbonaceous chondrites? Icarus 313:124– sorting; (2) impact cratering processes, using simulants as 138. a target for projectiles; (3) thermal cycling effects on Bland P. A., Cressey G., and Menzies O. N. 2004. Modal mineralogy of carbonaceous chondrites by X-ray asteroid material strength properties. In terms of diffraction and Mossbauer€ spectroscopy. Meteoritics & technology development, potential applications include (1) Planetary Science 39:3–16. volatile extraction using radiative energy, (2) curation Boivin A. L., Hickson D., Tsai C., Cunje A., Ghent R. R., protocols for future sample return of volatile-rich asteroid and Daly M. G. 2018. Broadband measurements of the or material, (3) grappling and anchoring devices complex permittivity of carbonaceous asteroid regolith analog materials. Journal of Geophysical Research Planets that can interact with asteroid surfaces. These are all 123:3088–3104. examples of current work underway using our asteroid Brearley A. J. and Jones R. H. 1998. Chondritic meteorites. In simulants; future applications will likely be even richer in Planetary materials, edited by Papike J. J. Albuquerque, scope and creativity. New Mexico: Mineralogical Society of America. pp. 3- 001–3-398. Britt D. T. and Consolmagno G. J. 2003. Stony meteorite CONCLUSIONS porosities and densities: A review of the data through 2001. Meteoritics & Planetary Science 38:1161–1180. We used modal mineralogy and physical properties Bus S. J. and Binzel R. P. 2002. Phase II of the small main- data from carbonaceous chondrites to create a series of belt asteroid spectroscopic survey: A feature-based – simulated C-complex asteroid materials, including taxonomy. Icarus 158:146 177. Cannon K. M., Britt D. T., Smith T. M., Fritsche R. F., and cobbles and regolith. Prototypes based on initial recipes Batcheldor D. 2019. Mars global simulant MGS-1: A were measured using a range of instrumental techniques rocknest-based open standard for basaltic Martian soil and compared to data from reference meteorites. In simulants. Icarus 317:470–478. 2080 D. T. Britt et al.

Cheng A. F., Atchison J., Kantsiper B., Rivkin A. S., Stickle bulk lunar soils: Application to Diviner thermal infrared A., Reed C., Galvez A., Carnelli I., Michel P., and observations of the Moon. Icarus 283:326–342. Ulamec S. 2015. Asteroid impact and deflection assessment Donaldson Hanna K. L., Schrader D. L., Cloutis E. A., Cody mission. Acta Astronautica 115:262–269. G. D., King A. J., McCoy T. J., Applin D. M., Mann J. Christensen P. R., Bandfield J. L., Hamilton V. E., Howard P., Bowles N. E., Brucato J. R., Connolly Jr. H. C., Dotto D. A., Lane M. D., Piatek J. L., Ruff S. W., and Stefanov E., Keller L. P., Lim L. F., Clark B. E., Hamilton V. E., W. L. 2000. A thermal emission spectral library of rock- Lantz C., Lauretta D. S., Russell S. S., and Schofield P. F. forming minerals. Journal of Geophysical Research 2019. Spectral characterization of analog samples in 105:9735–9739. anticipation of OSIRIS-REx’s arrival at Bennu: A blind Ciesla F. J., Lauretta D. S., Cohen B. A., and Hood L. L. test study. Icarus 319:701–723. 2003. A nebular origin for chondritic fine-grained Fries M., Abell P., Brisset J., Britt D., Colwell J., Dove A., phyllosilicates. Science 299:549–552. Durda D., Graham L., Hartzell C., Hrovat K., John K., Clark B. E., Hapke B., Pieters C., and Britt D. 2002. Asteroid Karrer D., Leonard M., Love S., Morgan J., Poppin J., space weathering and regolith evolution. In Asteroids II, Rodriguez V., Sanchez-Lana P., Scheeres D., and Whizin edited by Binzel R. P., Gehrels T., and Matthews M. S. A. 2018. The Strata-1 experiment on small body regolith Tucson, Arizona: University of Arizona Press. segregation. Acta Astronautica 142:87–94. Cloutis E. A., Hiroi T., Gaffey M. J., Alexander C. M. D’. Garenne A., Beck P., Montes-Hernandez G., Chiriac R., D., and Mann P. 2011a. Spectral reflectance properties of Toche F., Quirico E., Bonal L., and Schmitt B. 2014. The carbonaceous chondrites: 1 CI chondrites. Icarus 212:180– abundance and stability of “water” in type 1 and 2 209. carbonaceous chondrites (CI, CM and CR). Geochimica et Cloutis E. A., Hudon P., Hiroi T., Gaffey M. J., and Mann P. Cosmochimica Acta 137:93–112. 2011b. Spectral reflectance properties of carbonaceous Graps A. L., Blondel P., Bonin G., Britt D., Centuori S., chondrites: 2. CM chondrites. Icarus 2016:309–346. Delbo M., Drube L., Duffard R., Elvis M., Faber D., Cody G. D. and Alexander C. M. O’D. 2005. NMR studies of Frank E., Galache J. L., Green S. F., Thimo Grundmann chemical structural variation of insoluble organic matter J., Hsieh H., Kereszturi A., Laine P., Levasseur-Regourd from different carbonaceous chondrite groups. Geochimica A.-C., Maier P., Metzger P., Michel P., Mueller M., et Cosmochimica Acta 69:1085–1097. Mueller T., Murdoch N., Parker A., Pravec P., Reddy V., Conel J. E. 1969. Infrared emissivities of silicates: Sercel J., Rivkin A., Snodgrass C., and Tanga P. 2017. Experimental results and a cloudy atmosphere model of ASIME 2016 white paper: In-space utilisation of asteroids: spectral emission from condensed particulate mediums. “Answers to Questions from the Asteroid Miners.” Journal of Geophysical Research 74:1614–1634. arXiv:1612.00709v2. Consolmagno G. J., Britt D. T., and Macke R. J. 2008. The Gundlach B. and Blum J. 2013. A new method to determine significance of meteorite density and porosity. Chemie der the grain size of planetary regolith. Icarus 223:478–492. Erde— 68:1–29. Hayatsu R., Matsuoka S., Scott R. G., Studier M. H., and De Sanctis M. C., Ammannito E., Raponi A., Marchi S., Anders E. 1977. Origin of organic matter in the early solar McCord T. B., McSween H. Y., Capaccioni F., Capria system–VII. The organic polymer in carbonaceous M. T., Carrozzo F. G., Ciarniello M., Longobardo A., chondrites. Geochimica et Cosmochimica Acta 41:1325– Tosi F., Fonte S., Formisano M., Frigeri A., Giardino 1339. M., Magni G., Palomba E., Turrini D., Zambon F., Henderson B. G. and Jakosky B. M. 1994. Near-surface Combe J.-P., Feldman W., Jaumann R., McFadden L. thermal gradients and their effects on mid-infrared A., Pieters C. M., Prettyman T., Toplis M., Raymond C. emission spectra of planetary surfaces. Journal of A., and Russell C. T. 2015. Ammoniated phyllosilicates Geophysical Research 99:19,063–19,073. with a likely outer Solar System origin on (1) Ceres. Howard K. T., Benedix G. K., Bland P. A., and Cressey G. Nature 528:241–244. 2009. Modal mineralogy of CM2 chondrites by X-ray Delbo M., Libourel G., Wilkerson J., Murdoch N., Michel P., diffraction (PSD-XRD). Part 1: Total pyllosilicate Ramesh K. T., Ganino C., Verati C., and Marchi S. 2014. abundance and the degree of aqueous alteration. Thermal fatigue as the origin of regolith on small Geochimica et Cosmochimica Acta 73:4576–4589. asteroids. Nature 508:233–236. Howard K. T., Alexander C. M. O’D., Schrader D. L., and Donaldson Hanna K. L., Thomas I. R., Bowles N. E., Dyl K. A. 2015. Classification of hydrous meteorites (CR, Greenhagen B. T., Pieters C. M., Mustard J. F., Jackson CM and C2 ungrouped) by phyllosilicate fraction: PSD- C. R. M., and Wyatt M. B. 2012a. Laboratory emissivity XRD modal mineralogy and planetesimal environments. measurements of the plagioclase solid solution series under Geochimica et Cosmochimica Acta 149:206–222. varying environmental conditions. Journal of Geophysical Jarosewich E. 1990. Chemical analyses of meteorites: A Research 117:E11004. compilation of stony and analyses. Donaldson Hanna K. L., Wyatt M. B., Thomas I. R., Bowles Meteoritics 25:323–337. N. E., Greenhagen B. T., Maturilli A., Helbert J., and Johnson T. V. and Fanale F. P. 1973. Optical properties of Paige D. A. 2012b. Thermal infrared emissivity carbonaceous chondrites and their relationship to measurements under a simulated lunar environment: asteroids. Journal of Geophysical Research 78:8507–8518. Application to the Diviner Lunar Radiometer Experiment. Kaplan H. H. and Milliken R. E. 2016. Reflectance Journal of Geophysical Research 117:E00H05. spectroscopy for organic detection and quantification in Donaldson Hanna K. L., Greenhagen B. T., Patterson III W. clay-bearing samples: Effects of albedo, clay type, and R., Pieters C. M., Mustard J. F., Bowles N. E., Paige D. water content. Clays and Clay Minerals 64:167–184. A., Glotch T. D., and Thompson C. 2017. Effects of Keil K. 2000. Thermal alteration of asteroids: Evidence from varying environmental conditions on emissivity spectra of meteorites. Planetary and Space Science 48:887–903. Simulated asteroid materials 2081

King A. J., Solomon J. R., Schofield P. F., and Russell S. S. Mueller R. P., Howe S., Kochmann D., Ali H., Andersen C., 2015. Characterising the CI and CI-like carbonaceous Burgoyne H., Chambers W., Clinton R., De Kestellier X., chondrites using thermogravimetric analysis and infrared Ebelt K., Gerner S., Hofmann D., Hogstrom K., Ilves E., spectroscopy. Earth, Planets and Space 67:198. Jerves A., Keenan R., Keravala J., Khoshnevis B., Lim S., Krot A. N., Hutcheon I. D., Brearley A. J., Pravdivtseva O. Metzger P., Meza L., Nakamura T., Nelson A., Partridge V., Petaev M. I., and Hohenberg C. M. 2005. Timescales H., Pettit D., Pyle R., Reiners E., Shapiro A., Singer R., and settings for alteration of chondritic meteorites. In Tan W.-L., Vazquez N., Wilcox B., and Zelhofer A. 2016. Meteorites and the early solar system II, edited by Lauretta Automated additive construction (AAC) for Earth and D. S. and McSween H. Y. Jr. Tucson, Arizona: University space using In-situ resources. Proceedings of the Fifteenth of Arizona Press. pp. 525–553. Biennial ASCE Aerospace Division International Lauretta D. S., Al Asad M. M., Ballouz R.-L., Barnouin O. Conference on Engineering, Science, Construction, and S.,BierhausE.B.,BoyntonW.V.,BreitenfeldL.B., Operations in Challenging Environments (Earth & Space Calaway M. J., Chojnacki M., Christensen P. R., Clark 2016), American Society of Civil Engineers, Reston, B.E.,ConnollyJr.H.C.,Drouetd’AubignyC.,Daly Virginia, USA. M.G.,DalyR.T.,Delbo M., DellaGiustina D. N., Nakamura T., Noguchi T., Tanaka M., Zolensky M. E., Dworkin J. P., Emery J. P., Enos H. L., Farnocchia D., Kimura M., Tsuchiyama A., Nakato A., Ogami T., Golish D. R., Haberle C. W., Hamilton V. E., Ishida H., Uesugi M., Yada T., Shirai K., Fujimura A., Hergenrother C. W., Jawin E. R., Kaplan H. H., Le Okazaki R., Sandford S. A., Ishibashi Y., Abe M., Corre L., McCoy T. J., McMahon J. W., Michel P., Okada T., Ueno M., Mukai T., Yoshikawa M., and Molaro J. L., Nolan M. C., Pajola M., Palmer E., Perry Kawaguchi J. 2011. Itokawa dust particles: A direct link M.E.,ReuterD.C.,RizkB.,RobertsJ.J.,RyanA., between S-type asteroids and ordinary chondrites. Science ScheeresD.J.,SchwartzS.R.,SimonA.A.,SusorneyJ. 333:1113–1116. C.M.,WalshK.J.,andZoudX.-D.,andtheOSIRIS- Opeil C. P., Consolmagno G. J., and Britt D. T. 2010. The REx team. 2019. OSIRIS-REx arrives at Asteroid thermal conductivity of meteorites: New measurements (101955) Bennu: Exploration of a hydrated primitive and analysis. Icarus 208:449–454. near-Earth asteroid (abstract #2608). 50th Lunar and Pohl L. and Britt D. T. 2017. The radiation shielding potential Planetary Science Conference. CD-ROM. of CI and CM chondrites. Advances in Space Research Lewis J. S. 1996. Mining the sky: Untold riches from the 59:1473–1485. asteroids, , and planets. New York: Basic Books. p. Pohl L. and Britt D. T. Forthcoming. Strengths of meteorites– 274. An overview and analysis of available data. Icarus. Licandro J., Campins H., Kelley M., Hargrove K., Pinilla- Pravdivtseva O., Krot A. N., and Hohenberg C. M. 2018. I- Alonso N., Cruikshank D., Rivkin A. S., and Emery J. Xe dating of aqueous alteration in the CI chondrite 2011. (65) Cybele: Detection of small silicate grains, Orgueil: I. Magnetite and ferromagnetic separates. water-ice, and organics. Astronomy & Astrophysics 525: Geochimica et Cosmochimica Acta 227:38–47. A34. Rivkin A. S. 2012. The fraction of hydrated C-complex Logan L. M., Hunt G. R., Salisbury J. W., and Balsamo S. R. asteroids in the from SDSS data. Icarus 1973. Compositional implications of Christiansen 221:744–752. frequency maximums for infrared remote sensing Ruff S. W., Christensen P. R., Barnera P. W., and Anderson applications. Journal of Geophysical Research 78:4983– D. L. 1997. Quantitative thermal emission spectroscopy of 5003. minerals: A laboratory technique for measurement and Macke R. J., Consolmagno G. J., and Britt D. T. 2011. calibration. Journal of Geophysical Research 102:14,899– Density, porosity, and magnetic susceptibility of 14,913. carbonaceous chondrites. Meteoritics & Planetary Science Salisbury J. W., Walter L. S., Vergo N., and D’Aria D. M. 46:1842–1862. 1991. Infrared (2.1-25 lm) spectra of minerals. USGS Mackinnon I. D. R. and Zolensky M. E. 1984. Proposed Open-File Report 87-263. structures for poorly characterized phases in C2M Stoeser D., Wilson S., and Rickman D. 2010. Design and carbonaceous chondrite meteorites. Nature 309:240–242. specification for the Highland regolith prototype simulants McKay D. S., Carter J. L., Boles W. W., Allen C. C., and NU-LHT-1M and -2M. NASA Technical Reports, TM- Allton J. H. 1994. JSC-1: A new lunar soil simulant. 2010-216438. Engineering, Construction, and Operations in Space IV, Thomas I. R., Greenhagen B. T., Bowles N. E., Donaldson American Society of Civil Engineers 859–866. Hanna K. L., Temple J., and Calcutt S. B. 2012. A new Metzger P. T., Muscatello A., Mueller R. P., and Mantovani experimental setup for making thermal emission J. 2013. Affordable, rapid bootstrapping of the space measurements in a simulated lunar environment. Review of industry and solar system civilization. Journal of Aerospace Scientific Instruments 83:124502. Engineering 26:18–29. Thomas I. R., Bowles N. E., Donaldson Hanna K. L., Metzger P. T., Britt D. T., Covey S., Schultz C., Cannon K. Connolly Jr. H. C., Killgore M., and Lauretta D. S. 2014. M., Grossman K. D., Mantovani J. G., and Mueller R. P. The effects of varying environmental conditions on the 2019. Measuring the fidelity of asteroid regolith and emissivity spectra of meteorites (abstract #1989). 45th cobble simulants. Icarus 321:632–646. Lunar and Planetary Science Conference. CD-ROM. Miyamoto H., Yano H., Scheeres D. J., Abe S., Barnouin-Jha Tsuchiyama A., Mashio E., Imai Y., Noguchi T., Miura Y., O., Cheng A. F., Demura H., Gaskell R. W., Hirata N., Yano H., and Nakamura T. 2009. Strength measurement Ishiguro M., Michikami T., Nakamura A. M., Nakamura of carbonaceous chondrites and using R., Saito J., and Sasaki S. 2007. Regolith migration and micro compression testing machine (abstract #5189). 72nd sorting on asteroid Itokawa. Science 316:1011–1014. Annual Meteoritical Society Meeting. 2082 D. T. Britt et al.

Tsuda Y., Yoshikawa M., Abe M., Minamino H., and Nakazawa Vokrouhlicky D., Bottke W. F., Chelsey S. R., Scheeres D. J., S. 2013. System design of the Hayabusa 2–Asteroid sample and Statler T. S. 2015. The Yarkovsky and YORP effects. return mission to 1999 JU3. Acta Astronautica 91:356–362. In Asteroids IV, edited by Michel P., Demeo F. E., and Turcotte D. L. 1986. Fractals and fragmentation. Journal of Bottke W. F. Tucson, Arizona: The University of Arizona Geophysical Research 91:1921–1926. Press. p. 509.

SUPPORTING INFORMATION Table S1. Sourcing information for the mineral ingredients used in the prototype asteroid simulants. Additional supporting information may be found in the online version of this article: