Thermal Conductivity of Asteroid Analogue Material

51st Lunar and Planetary Science Conference (2020) 2354.pdf

THERMAL CONDUCTIVITY OF ASTEROID ANALOGUE MATERIAL. C. M. Gilmour1, J. Freemantle1, and M. G. Daly1. 1Centre for Research in Earth and Space Science, York University, Toronto, ON, Canada. E-mail: cgil- [email protected]

Introduction: Understanding the thermal properties compositional differences [8]. Comparison with our py- of asteroids is fundamental for interpreting their thermal rite measurements reveals that all three samples are evolution throughout the history of the solar system. more similar at 300 K than at 50 K. Considering that our However, acquiring relevant thermal data of these bod- pyrite sample is not homogenous, it is evident that com- ies is challenging due to accessibility; therefore, we position is the cause for lower thermal conductivity val- have to seek alternative practices to develop our under- ues. In this case, the trace minerals have low thermal standing of these properties. Measuring the thermal con- conductivity relative to pyrite, causing a decrease in the ductivity of asteroid analogue material (i.e., minerals bulk thermal conductivity, especially at lower tempera- common in asteroids) is one practical way to overcome tures. Despite the difference between our results and [8], this challenge. There are many thermal conductivity the increase in thermal conductivity as temperature de- studies of minerals available [1-6]; however, these stud- creases is consistent among the results. ies focus on terrestrial applications and measurements are largely conducted at room temperature (300 K) and above. As asteroid surface temperatures reach as low as 50 K [7], acquiring thermal conductivity data at temperatures <300 K is valuable. Our focus for this study is minerals commonly found in carbonaceous asteroids: olivine (forsterite), pyroxene (enstatite), carbonates, pyrrhotite (troilite analogue), plagioclase, graphite (car- bon analogue), magnetite, chromite, and serpentine. By measuring the thermal conductivity of these minerals, we can estimate the thermal conductivity of carbonaceous asteroids based on their known mineral composition and distribution. Methods: Mineral samples are cut into 2 x 2 x 20 mm bars and are connected to a thermocouple, heater,

and thermal probe. Thermal conductivity measurements Fig. 1. Thermal conductivity curve for pyrite. Results from [8] are -4 are collected between 50 and 300 K under a 10 hPa included for comparison. Error bars are one SD. Note that our er- vacuum using a Cryogen-Free Measurement System rors are too small to be seen relative to the scale of the y-axis. (CFMS). A lower limit of 50 K was chosen as asteroid surface temperatures do not reach below 50 K [7]; however, the CFMS can measure as low as 5 K. Composi- tional analysis is also completed via Fourier Transform Infrared (FTIR) spectroscopy, X-Ray Diffraction (XRD), and/or Electron-Dispersive X-ray Spectroscopy (EDS) to aid in the thermal conductivity investigation. Preliminary results: Here we report our initial thermal conductivity findings for olivine and pyrite. Alt- hough pyrite is not found in asteroids, a previous study ([8]) conducted thermal conductivity measurements between 50 and 300 K. Thus, we have chosen to analyze pyrite to help develop our thermal conductivity measurement procedure and ensure our results are valid. Pyrite. A pyrite sample containing traces of calcite and hornblende (determined by EDS) was measured between 50 and 300 K (Fig. 1). Included in Fig. 1 are the Fig. 2. Thermal conductivity curve for olivine. Error bars are one results for two different pyrite samples (Sample 1 and SD. Measurements between 225 and 300 K: n = 1 (hence no er- Sample 2) from [8]. The discrepancy between these two rors). Measurements between 110 and 200: n = 3. The error asso- samples, notably at 50 K, is attributed to chemical ciated with 150 K is small relative to the scale. 51st Lunar and Planetary Science Conference (2020) 2354.pdf

Olivine. We have analyzed a crystalline olivine sam- W/(m K), the bulk thermal conductivity of the olivine ple between 110 and 300 K (Fig. 2). As these are pre- sample is expected to be lower than 4.96 W/(m K). liminary results, we intend to acquire measurements for Analysis⋅ of samples with a greater modal abundance of this sample down to 50 K. This sample was claimed to olivine (>80-90%) will be carried out to further our⋅ in- be nearly pure forsterite; however, XRD analysis re- vestigation. veals the sample is composed of 45% forsterite, 14% Unlike metals and alloys, such as copper and brass, talc, 12% hornblende, 13.5% kaolinite, and 16% amor- which have well-known thermal conductivities [e.g., phous material. For comparison, the calculated average 11,12], it is evident that there will be always been some thermal conductivity of forsterite at 300 K is 4.96±0.21 discrepancy among thermal conductivity values for a W/(m K) [3,4,6]. There is a clear difference between our given mineral due to compositional variations. How- measured value at 300 K (3.02 W/(m K)) and the litera- ever, we can expect that the general shape/pattern of the ture average⋅ . However, the thermal conductivity curve thermal conductivity curves for a given mineral will be of this olivine sample displays a slight⋅ parabolic pattern, similar despite compositional differences (e.g., Fig. 1), similar to what has been observed in fayalite-rich oli- assuming that a sample is comprised mostly of the min- vine between 325 and 1075 K [4]. eral of interest. By studying the behaviour of thermal Discussion: Thermal conductivity is challenging to conductivity curves for minerals found in asteroids, and measure in geological material as it can be affected by factoring in mineral composition and distribution, we several different properties: porosity/density; mineral can formulate a hypothesis to explain the thermal evo- composition, distribution, and anisotropy; grain size and lution of asteroids. boundaries; water content; experimental conditions; and Future work: Once we establish a thermal conduc- sample locality [4,6,9]. Our results demonstrate that the tivity database of asteroid analogue material between 50 mineral composition/distribution of the sample is a ma- and 300 K, we will expand our study to include meteor- jor controlling factor of thermal conductivity. Although ites. To date, only two studies have directly measured minerals are considered to be homogenous substances, meteorite thermal conductivities [13, 14]. Given that they are rarely truly pure as they are often accompanied carbonaceous asteroids are the target of current sample- by other minerals in trace abundances. As seen with the return missions (i.e., Hyabusa-2, OSIRIS-REx), thermal pyrite sample (Fig. 1), which is 95% pyrite, the thermal conductivity measurements of carbonaceous chondrites conductivity at 300 K is more comparable to the litera- are most desired. Obtaining such measurement will ture than the olivine sample (refer to above section), complement the thermophysical results of these mis- which is 45% forsterite. Minerals with a solid solution sions. series (i.e., olivine, plagioclase, pyroxene) are known to Acknowledgements: Thank you to the ERICA have thermal conductivity differences over composi- group, University of Valladolid, Spain for XRD analy- tional ranges [2,3]. In the case of olivine, such differ- sis of olivine and subsequent samples for this project. ences are more notable with compositional changes as- References: [1] Clark S. P. (1966) Geol. Soc. Am. sociated with the forsterite end-member [3], which may Mem., 97, 461-482. [2] Horai K. and Simmons G. potentially contribute to our lower thermal conductivity (1969) EPSL, 6, 359-368. [3] Horai K. (1971) J. Ge- measurement for olivine. However, emphasizing again ophys. Res., 76, 1278-1308. [4] Čermák V. and Rybach that the olivine sample is only half forsterite, it is more L. (1982) Landolt-Börnstein, Group 5, Vol. 1a, pp. 305- likely that the compositional variation of this sample is 343. [5] Diment W. H. and Pratt H. R. (1988) USGS the cause. Taking into consideration the weighted aver- Report, 88-690. [6] Clauser C. and Huenges E. (1995) ages, the accompanying minerals (kaolinite, horn- Rock Physics & Phase Relations, Vol. 3, pp. 105-126. blende, talc, and amorphous material) must have ther- [7] Delbó M. and Harris A. W. (2002) MAPS, 37, 1929- mal conductivities less than forsterite. Reported values 1936. [8] Popov P. A. et al. (2013) Crys. Rep., 58, 319- for kaolinite, hornblende, and talc at ~300 K are 321. [9] Midttømme K. and Roadlset E. (1998) Petrol. 2.64±0.20 W/(m K) [10], 2.42±0.38 W/(m K) [3,5,6], Geosci., 4, 165-172. [10] Brigaud F. and Vasseur G. and 6.57±3.79 W/(m K) [3-6], respectively. Note the (1989) Geophys. J., 98, 525-542. [11] Shackelford J. F. large uncertainty⋅ associated with talc; this is⋅ attributed et al. (2015) CRC Materials Science and Engineering to the preferred orientation⋅ of talc crystals, as well as a Handbook, 4th ed., 634 p. [12] Cverna F. (2002) ASM combination of powdered vs. single-crystal results ready reference. Thermal properties of metals. 560 p. [3,6]. Considering the low thermal conductivities of ka- [13] Opeil C. P. et al. (2010) Icarus, 208, 449-454. olinite and hornblende relative to forsterite, and assum- [14] Opeil C. P. et al. (2012) MAPS, 47, 319-329. ing (1) the thermal conductivity of talc is close to one standard deviation below the mean and (2) the thermal conductivity of amorphous material is less than 3