Stony meteorite thermal properties and their relationship with meteorite chemical and physical states

Authors


Corresponding author. E-mail: gjc@specola.va

Abstract

Abstract– In our ongoing survey of meteorite physical properties, we have to date measured the thermal conductivity for seventeen stony meteorites at temperatures ranging from 5 K to 300 K. Here, we report new results for nine ordinary chondrites, one enstatite chondrite, and the basaltic achondrites Frankfort (howardite) and Los Angeles (shergottite). We find that thermal conductivity is significantly lower than would be expected from averaging the laboratory conductivities of their constituent minerals, with a dependence on temperature different from the expected conductivity of pure minerals. In addition, we find a linear relationship between the inverse of the porosity of the samples measured and their thermal conductivity, regardless of meteorite composition or type. We conclude that thermal conductivity is controlled by the presence of shock-induced microcracks within the meteorites, which provide a barrier to the transmission of thermal energy via phonons. In contrast to conductivity, our first measurement of heat capacity as a function of temperature (on Los Angeles) suggests that heat capacity is primarily a function of oxide composition and is not strongly affected by the physical state of the sample.

Introduction

Thermal properties are essential parameters in modeling a number of important processes in the solar system. For example, the Yarkovsky and YORP effects on asteroid orbital and spin perturbations depend on the inverse thermal inertia, which varies with the square root of thermal conductivity and heat capacity. The thermal evolution of asteroids and the sublimation rates of comets, icy satellites, and other icy bodies thought to have a significant meteorite-like rocky component will depend on the thermal diffusivity, which is linearly dependent on thermal conductivity divided by heat capacity. In addition to its utility to modelers, however, the thermal properties of meteorites provide an important way of characterizing the meteorite’s composition and its physical state (Consolmagno et al. 2008; Opeil et al. 2010).

Thermal conductivity, a sample’s ability to transport heat, is usually modeled as the passage of packets of vibrational energy called phonons. The transmission of phonons depends on the material within the sample, and is expected to be a strong function of temperature especially at temperatures below 300 K. At very low temperatures, below 50 K, theory for crystalline materials (e.g, insulators) where phonons dominate predicts that thermal conductivity should be controlled by the phonon activation energy; as temperature increases, the production of phonons increases, resulting in values of the thermal conductivity increasing as T3. Above 100 K, as the number of phonons increases, they become more likely to scatter off each other, resulting in a reduction in conductivity that varies as 1/T. Thus, one expects conductivity to increase sharply from temperatures near absolute zero to a maximum around 50 K, above which the conductivity drops off by a factor that closely approaches 1/T at temperatures above 200 K. This is in fact the pattern seen in the thermal conductivity we have measured for two enstatite chondrites (out of the 17 samples in this study).

Amorphous materials and metals, on the other hand, show a different behavior; for instance, we found a continually rising conductivity over this temperature range in the iron meteorite Campo del Cielo (Opeil et al. 2010).

Phonons will also be scattered by metal grains, grain boundaries, and isolated pore spaces. Thus, one should expect the thermal conductivity to also be a function of the metal content and porosity of the sample. But perhaps even more importantly, phonons may encounter larger barriers to thermal transport such as sheet cracks and porosity caused by shock (cf. Friedrich et al. 2008); the presence of such cracks should result in a significant drop in thermal conductivity.

Measuring heat capacity allows us access to the intrinsic characteristic thermodynamic property of a particular substance. Specific heat capacity, C, is the fundamental measurable physical quantity that characterizes the amount of heat energy (dQ) required to change the temperature (dT) of a unit mass of substance (m) by a given amount; expressed in simple mathematical terms, C = m−1 dQ/dT with SI units of [J (kg K−1)]. This heat is contained within the minerals via the various vibrational modes of the components present, and thus one expects that the heat capacity of a meteorite should be primarily a function of its composition, with little dependence on its physical state. As with phonons, these vibrational states are suppressed at lower temperatures. For this reason, it is essential for a complete understanding of meteorite formation and asteroid evolution that this quantity be measured over a large temperature range, most importantly at low temperatures that reflect the actual temperature environment of the formation and evolution of this material in the solar system.

Previous Results

In an earlier paper (Opeil et al. 2010), we reported the measurement of the thermal conductivity at low temperatures (5 K to 300 K) of five stony meteorites: Chronstad (H5), Lumpkin (L6), Abee (EH4), Northwest Africa 5515 (CK4 find), and Cold Bokkeveld (CM2). (An iron meteorite, Campo de Cielo, was also discussed in that paper.) The conductivity of all the stony samples except the enstatite chondrite Abee gave values that were lower by as much as an order of magnitude compared with conductivity values reported in the literature for pure minerals (Clauser and Huenges 1995). In addition, we found that all except Abee had conductivities that were nearly constant with temperature above 100 K. The L and CK sample conductivities at 200 K were both close to 1.5 W mK−1, that of the H was 1.9 W mK−1, and that of the CM sample was 0.5 W mK−1; by contrast, the literature value at 300 K for serpentine is 2.5 W mK−1 and that of enstatite and olivine is 4.5 to 5 W mK−1 (comparable to our Abee value).

Our first results were consistent with earlier work on ordinary chondrites by Matsui and Osako (1979) and Yomogida and Matsui (1983), who had derived values of thermal conductivity from measurements of thermal diffusivity. (We note as well ongoing diffusivity measurements by Szurgot and collaborators; c.f. Szurgot [2011]. Their results, published to date in abstract form, promise to extend significantly the number of data points for meteorite thermal diffusivity at room temperature.) However, those authors had found that thermal conductivities ranged from sample to sample over nearly an order of magnitude even within a given meteorite class. Furthermore, as our original results only measured one sample of each meteorite type, it was not possible to draw conclusions about the effects on conductivity of properties other than composition. Thus, our next step has been to measure more ordinary chondrites, including multiple samples of the same meteorite, to look for such variations within a class that can be related to other physical properties of the meteorite.

In addition, with the interest in the thermal evolution of basaltic parent bodies, including Mars and the HED parent body (concomitant with the arrival of the Dawn spacecraft at asteroid 4 Vesta), the measurement of basaltic meteorite conductivities has become of recent interest. Thus, we have included a shergottite, thought to be from Mars, and one of the putative Vesta meteorites in the suite measured here. All meteorite samples were provided from the Vatican meteorite collection.

Measurement Technique

We have measured seven new chondrites and two basaltic meteorites for this study. They include the EL6 meteorite Pillistfer (our previous enstatite meteorite, Abee, is an EH); the H5 chondrites Barbotan, Collescipilli, and Pułtusk and the H6 La Cienega (we had previously measured Cronstad, H5); and the L6 chondrites Bath Furnace and Holbrook (we had previously measured Lumpkin, L6). The two basaltic meteorites included in our study are the shergottite Los Angeles and the howardite Frankfort.

All the ordinary chondrites except La Cienega are falls, but most date from the 19th century (the oldest, Barbotan, fell in 1790; the most recent, Holbrook, in 1912) and show varying degrees of terrestrial weathering, as can be seen in Fig. 1. La Cienega was found in 2007 and shows little evidence of weathering.

Figure 1.

 Sample of Bath Furnace mounted for measurement.

Among the ordinary chondrites, two different samples of the L chondrites Bath Furnace and Holbrook were measured. In addition, one of the Bath Furnace samples was rotated by 90° and remeasured to determine whether there was a significant anisotropy in the conductivity of this meteorite.

As with our previous work (Opeil et al. 2010), all measurements were made using a Quantum Design Physical Property Measurement System, Thermal Transport Option (QD-PPMS–TTO) that allows thermal conductivity measurements [W/m-K] in a temperature range of 2–400 K; the system utilizes a basic cryogenic/field system to establish precise control of temperature (Dilley et al. 2002). A QD option P640 High-Vacuum system adsorption pump in the cryogenic dewar thermally isolates the sample. The accuracy of the measurements has been confirmed on a 7740 Pyrex standard, and temperature calibrations were performed using the Quantum Design Ni-alloy standard.

All samples were cut into 0.5–1 cm prisms, to which gold-coated, oxygen-free, high-conductivity copper (OFHC-Cu) disks were attached with silver (Ag) epoxy. Thermal conductivity is determined by applying a heat pulse “Q” from a heater attached to one end of the sample to create a user-specified temperature difference between two calibrated Cernox thermometers located at the sample ends. Heat flows out of the sample into a cold-foot located on the sample puck. The sample is held in a cryostatic chamber whose temperature is automatically stepped from 300 K to below 5 K at a rate of 0.5 K min−1 in a vacuum (pressure is held to <1.33 × 10−4 Pa).

Results

Enstatite Chondrite

Our results for the EL6 chondrite Pillistfer are consistent with those of the EH chondrite Abee previously reported. In both cases, the characteristic peak conductivity is seen near 50 K, followed by the expected 1/T drop off. The conductivity found at room temperature is comparable to that of enstatite measured at room temperature. Note that above 100 K, there is little difference in the conductivity seen for this sample and that of Abee.

Although both meteorites conduct heat the way that one would expect for samples of pure enstatite, this correspondence is only coincidental; both meteorites are in fact quite different physically and compositionally from pure minerals. According to the literature (c.f. Grady 2000), Abee is a breccia, with a shock state of the various components ranging from S2 to S5, and it is 32.5% by mass iron. The reported metal content for Pillistfer is only slightly lower, 27.8%, and it has a reported shock state of S2. Of the samples from the Vatican collection that we measured, Abee is notable for its jet-black color, evidence of shock blackening, while Pillistfer has a light gray color, probably the least shocked of the enstatite chondrites in the Vatican collection. In spite of its reported lower metal content, in fact, our sample of Pillistfer has a higher magnetic susceptibility, 5.43 (log units), compared with 5.3 measured for our sample of Abee (Rochette et al. 2008). Our sample of Abee has a porosity of 2.19 ± 1.5% (Macke et al. 2010). No porosity data exist for the particular sample of Pillistfer measured (it being too small for reliable volume measurements) but Macke et al. (2010) report porosity measurements for four other samples of Pillistfer, which range from zero to 5.2% with an average of 2.4%. Both porosities are significantly lower than typical porosities for ordinary or carbonaceous chondrites.

Ordinary Chondrites

Matsui and Osako (1979) and Yomogida and Matsui (1983) reported the conductivities of 21 ordinary chondrites by measuring thermal diffusivity at six temperatures from 100 to 350 K, assuming a heat capacity, and calculating from that the thermal conductivity. Their values ranged from less than 0.5 W mK−1 up to nearly 4 W mK−1; L chondrites were generally lower in conductivity than H chondrites, but there was a notable overlap among the two groups. We have plotted their results as shaded regions in Figs. 2 and 3. As seen in Fig. 2, the results from Opeil et al. (2010) fell in the lower range of their values. However, we find (see Fig. 3) that our new measurements fill the entire range of meteorite conductivities previously reported by the Matsui group. There is a wide range, and great overlap, of thermal conductivities among ordinary chondrite classes.

Figure 2.

 Previous results for thermal conductivities of meteorites from Opeil et al. (2010), compared with terrestrial minerals near 300 K indicated with diamonds (data from Clauser and Huenges 1995) and those published by Matsui and Osako (1979) and Yomogida and Matsui (1983), indicated by the shaded regions labeled as Y&M.

Figure 3.

 Thermal conductivity of meteorites as a function of temperature. New results (this paper) are indicated with solid symbols; results previously reported in Opeil et al. (2010) are open symbols. Different classes of meteorites are indicated by the different colors, and the individual samples are identified, to the right, in the order of their conductivity at 300 K, from highest to lowest conductivities.

Furthermore, we find a significant range in conductivities when measuring different samples of nominally the same material, or indeed even the same sample, as is illustrated in Fig. 4. For example, looking at conductivities at 200 K (in the temperature range where conductivity is generally constant with temperature), we find that one sample of Holbrook (L6) has a conductivity of only 0.44 W mK−1, while the conductivity of a second sample of the same meteorite is nearly three times that value, at 1.2 W mK−1. Likewise, one piece of Bath Furnace (L6) has a conductivity at 200 K of 2.3 W mK−1, while a second piece of the same meteorite showed conductivities of 2.7 and 3.2 W mK−1 when measured in two orthogonal directions.

Figure 4.

 Two samples each of the L chondrites Bath Furnace (squares and triangles) and Holbrook (circles) were measured. In addition, one Bath Furnace sample (marked above with triangles as Bath Furnace 2 and Bath Furnace 3) was rotated 90° and measured again. Significantly different results were found, suggesting that thermal conductivity is controlled by the anisotropic physical state of the material, most likely crack sheets, and not by composition.

Note that both the Holbrook and Bath Furnace meteorites are the same chemical and petrographic type, L6, yet one piece of Bath Furnace measured in one direction is more than seven times as conductive as one of the Holbrook pieces. These differences are far greater than any measurement error (which are generally on the order of 0.01 W mK−1), and must represent differences in structure in the conductivity within the samples, including significant anisotropy, as seen in the second Bath Furnace sample. It is clear from these results that the major determinant in the thermal conductivity of stony meteorites is their physical state, rather than their chemical composition.

Basaltic Achondrites

We also measured the thermal conductivity at temperatures ranging from 300 K to 5 K for the basaltic achondrite meteorites Frankfort, a howardite (likely to be typical of material on the surface of asteroid 4 Vesta) and Los Angeles, a shergottite (believed to have originated on Mars). From 300 K to 100 K, the conductivity of Frankfort decreases gradually from 1.6 to 1.2 W mK−1; that of Los Angeles drops from 0.9 to 0.5 W mK−1. At lower temperatures, the thermal conductivity of both meteorites continues to drop, but more rapidly, to values below 0.1 W mK−1 at 10 K. We find that for both meteorites, the conductivity is significantly lower than would be expected from averaging the laboratory conductivities of their constituent minerals.

These results are similar to results from measurements of ordinary chondrites. The monotonic decrease in conductivity with temperature over this range is different from the expected conductivity of pure minerals, which tend to vary as 1/T. This indicates that, in common with the ordinary chondrites, the conductivity we measure is controlled by the physical structure of the meteorites, presumably the presence of shock-induced microcracks that provide a barrier to the transmission of thermal energy via phonons. Thus, these measurements should accurately describe the conductivity of any material in the sampled regolith of their parent bodies. The texture and mineralogy of Frankfort are characteristic of a breccia that mixes surface and subsurface materials in a shock-lithified matrix, which is typical of regolith material. The cooling rates calculated for Los Angeles are significantly slower than other shergottites (Rubin et al. 2000), suggesting that it was formed at depth. However, results from surface samples or samples transported by impact from the interior of a body (and thus presumably strongly shocked in the process) may significantly underestimate the actual conductivity of material deeper in the parent bodies, depending on how characteristically the shock history of these meteorites reflects material still in place inside these bodies.

The thermal conductivities of meteorites measured here (except for the enstatite chondrite Pillistfer) can be fit well by a fourth-order quadratic of the form K = A + BT + CTDT+ ET4. The fits are shown in Fig. 5 and the values of the coefficients A through E for each sample are listed in Table 1.

Figure 5.

 An expanded view of Fig. 3, showing the range of conductivities at the lower end of the thermal conductivity scale. Red data points are H chondrites, blue points are L chondrites, golden points are basaltic achondrites. Newly measured meteorites are indicated with solid symbols and named to the right of their respective datapoints. Lines through the data represent fourth order quadratic fits to the new data (see Table 1). Note that the basaltic achondrite conductivities are very similar to those for ordinary chondrites, in spite of their having significantly different compositions––in particular, little metallic iron.

Table 1.   Quadratic fits to the new thermal conductivity data reported here (except for the E chondrite Pillistfer) can be fit by a quadratic equation of the form: K = A + BT + CT+ DT+ ET4.
MeteoriteTypeAB × 102C × 104D × 106E × 109r2
La CienegaH chondrite−0.2184.975−3.9041.257−1.4380.997
BarbotanH chondrite−0.3677.126−5.4331.751−1.9760.988
CollescipoliH chondrite−0.0701.634−1.1790.378−0.4360.997
PułtuskH chondrite−0.1132.283−1.5220.466−0.5311.000
Bath Furnace 1L chondrite−0.1395.333−4.0631.290−1.4750.996
Bath Furnace 2L chondrite−0.1586.383−5.0121.672−2.0290.997
Bath Furnace 3L chondrite−0.1285.846−3.7441.026−1.0330.999
Holbrook 1L chondrite−0.0240.888−0.6230.191−0.2140.999
Holbrook 2L chondrite−0.1022.485−1.8820.625−0.7530.999
FrankfurtHowardite−0.0843.053−2.4820.844−0.9740.991
Los AngelesShergottite0.0510.774−0.3450.087−0.0860.999

Conductivity, Metal Content, and Porosity

One of the major constituents of ordinary chondrites is metallic iron, which is a much better conductor of heat than other minerals. And while the orientation of the pore spaces obviously would control how they affect conductivity, the likelihood of finding any such cracks should be related to the average porosity of the sample. Thus, it is instructive to compare our conductivity values with those of the average porosity measured for the hand samples, and with magnetic susceptibility, which is directly related to the average metal content of an ordinary chondrite.

We have measured porosities in hand samples of twelve of the meteorites in our data set, and the Matsui group likewise measured porosities for many of their samples (see Table 2). In Fig. 6a, we see that there is a strong linear relationship between thermal conductivity at 200 K (where the values for most of our samples are not strong functions of temperature) and the inverse of the porosity. Even recognizing that this simple relationship does not take into account the anisotropic effects of shock-induced cracks (note, for example, the three values of Bath Furnace seen with the value of 1/porosity = 23), there is a strong correlation (at the R = 0.95 level) between inverse porosity and conductivity in our data set; a line fit through our data gives a relationship of conductivity k = 3.6 + 6.8/P where P is the fractional porosity of the sample. This relationship is also seen when the data from Yomogida and Matsui (1983) are included, as is shown in Fig. 6b; here, however, two of their data points of very low porosity lower the confidence level of the result, which is k = 3.8 + 6.9/P, R = 0.81.

Table 2.   Density, porosity, and magnetic susceptibility compared with conductivity at 200 K.
MeteoriteKindBulk densityGrain densityPorosityMagnetic suscept.Cond. @ 200 KNotes
  1. Notes: NWA = Northwest Africa; ALH = Allan Hills; Y = Yamato; MET = Meteorite Hills.

  2. 1. Data from Macke (2010); average of all samples measured.

  3. 2. Data from Macke (2010); sample from Vatican collection.

  4. 3. Data from Yomogida and Matsui (1983).

New this paper       
PillistferEL 63.613.72.4%5.555.511
PułtuskH 53.443.727.5%5.271.251
BarbotanH 53.493.756.9%5.213.051
CollescipoliH 5   5.370.822
La CienegaH 6    1.90 
Holbrook 1L 63.183.5510.4%4.580.451
Holbrook 2L 63.183.5510.4%4.581.151
Bath Furnace 3L 63.503.664.3%5.053.151
Bath Furnace 2L 63.503.664.3%5.052.721
Bath Furnace 1L 63.503.664.3%5.052.261
FrankfortHowardite2.903.3212.7%3.431.311
Los AngelesShergottite2.833.088.1%3.520.771
Opeil et al. 2010       
AbeeEH 63.523.623.0%5.475.331
CronstadH 5    1.88 
LumpkinL 6    1.47 
NWA 5515CK 42.70   1.482
Cold BokkeveldCM 22.362.7815.0%3.680.501
Yomogida and Matsui 1983      
ALH 77288H 63.69 2.0% 3.533
ALH 77294H 53.35 12.9% 0.753
Gilgoin StationH 53.61 5.0% 3.603
GladstoneH 53.56 5.0% 2.163
MonroeH 43.58 5.9% 2.353
WellmanH 53.58 6.1% 3.853
Y-74156H 43.45 9.2% 1.543
Y-74647H 4.53.49 9.1% 1.153
ALH 769L 62.89 19.4% 0.553
ALH 77231L 63.07 14.3% 1.203
ALH 78103L 63.23 13.4% 0.733
ALH 78251L 63.22 13.2% 0.853
ArapahoeL 53.52 2.5% 2.313
BruderheimL 63.31 8.0% 1.033
FarmingtonL 53.40 5.5% 2.143
KunashakL 63.41 5.2% 1.863
Leedey,AL 63.25 10.4% 0.403
Leedey,BL 63.24 10.6% 0.473
MET 78003L 63.33 7.8% 1.543
New ConcordL 63.27 9.2% 0.783
Y-74191L 33.23 10.3% 1.243
Y-75097L 43.28 10.3% 0.973
Figure 6.

 Conductivity varies linearly with the inverse of porosity. a) Only data collected by the authors (circles). b) The data from Yomogida and Matsui (1983) (diamonds) added to our data; although the scatter is now larger, the fit is essentially unchanged.

The fact that one sample, Bath Furnace, gave two significantly different results when measured in two different directions immediately tells us that one cannot expect a simple, perfect relationship between thermal conductivity and any intensive property such as mineralogy or average porosity, as the conductivity will vary with the orientation of the pore cracks, at least for sample sizes comparable to the extent of the cracks. However, given the multiple impact history that asteroids have experienced, one should expect that on scales much larger than the extent of individual cracks (and much larger than the millimeter-sized samples measured here), the orientation of cracks will be random, and thus an average relationship between porosity and conductivity, as indicated here, should be a reasonable description of the typical conductivity to be found within an asteroid.

Szurgot (2011) reported the thermal conductivity at room temperature for a number of stony iron and iron meteorites as well as stones, and showed that conductivity over this range can be linearly related to density; clearly, the amount of iron present in a stony iron meteorite directly affects its thermal properties. Does this trend continue within the various classes of stony meteorites? The magnetic susceptibility of stony meteorites varies with the amount of metallic iron present, and so one might expect that samples with higher susceptibilities would have higher thermal conductivity. We have magnetic susceptibility data for twelve of the meteorites measured in our sample, and Fig. 7 shows that there is certainly a hint of a relationship, with the highest conductivity meteorites having the highest susceptibility. However, below a certain metal content, the conductivity appears to stay constant even as susceptibility drops further. Furthermore, the two meteorites with the highest metal content, which contribute the most to the visible trend, are also the enstatite chondrites with the lowest porosity. Thus, it is not clear that metal content itself is what leads to the trend seen in Fig. 7, or if it is merely an artifact of the most metal-rich meteorites also being the least porous, and the metal-poor meteorites the most porous.

Figure 7.

 There is a weak correlation in our data between conductivity and magnetic susceptibility, a measure of the iron content of the sample, at least for samples above a threshold metal abundance. However, the most metal-rich samples, the enstatite chondrites, are also the least porous, and so it may be porosity rather than metal content that we see here.

Likewise, our least-conductive meteorite is the water and volatile rich CM sample, Cold Bokkeveld; but this sample is also the most porous one for which a thermal conductivity has been measured. Thus, while it is possible that the volatile and water content may have an effect on its conductivity, as in the case of metal content, we cannot disentangle the effect of composition from that of porosity. More data are needed to explore the relationship further.

Shergottite Heat Capacity

Given that thermal diffusivity and thermal inertia both depend equally on conductivity and heat capacity, it is clear that measuring heat capacity for meteorites is critical. One essential problem in our understanding of heat capacity is that, like thermal conductivity, it is expected to vary strongly (by a factor of two or more) over the temperature range expected for the interiors of asteroids and small bodies. However, to date, heat capacity has been even less well studied than thermal conductivity. The few published values for heat capacity have only been measured at or above room temperature. Matsui and Osako (1979) directly measured the heat capacity of five Yamato meteorites (four ordinary chondrites and a howardite) at 300 K, 350 K, and 400 K, while Yomogida and Matsui (1983) used laboratory data of the constituent minerals of ordinary chondrites to calculate their heat capacities; their calculated values, which they preferred, were 50% higher than their directly measured results.

In the past year, we measured the heat capacity of one meteorite, the shergottite Los Angeles, at the Los Alamos National Laboratory using the Quantum Design Physical Properties Measurement System (QD-PPMS) similar to that used for our thermal conductivity measurements, but equipped with the additional instrumentation (P650 package) that allows the measurement of heat capacity. As with the thermal conductivity measurements, this system utilizes a basic cryogenic/field system to establish precise control of temperature, while specialized components perform the heat capacity measurement. A QD option P640 High-Vacuum system adsorption pump in the cryogenic dewar thermally isolates the sample. The sample was mounted on a compact calorimeter puck with thermal grease and inserted into the PPMS sample chamber. The specific heat capacity option P650 includes hardware and software that allowed fully automated high-sensitivity heat capacity measurements to be taken between 1.9 K and 400 K. The option uses a hybrid adiabatic relaxation method that combines the best measurement accuracy with robust error analysis. Unfortunately, using this system is quite time consuming; a single run, measuring the set-up (platform and thermal grease without the sample, followed by the actual measurement with the sample) requires up to 48 h of chamber time, significantly limiting the number of samples that can be measured.

In contrast to the thermal conductivity, which we have found is controlled by the physical state of the meteorite, one expects that the heat capacity of a meteorite should closely follow its composition. Our results support this hypothesis.

In Fig. 8, we compare our Los Angeles results with data for a variety of pure pyroxenes from the literature (Krupka et al. 1985). Los Angeles is a basalt of roughly equal volume percent abundances of plagioclase (An41Or4 to An58Or1), shocked to maskelynite, and high-ferroan calcium pyroxene (Fe45Wo13 to Fs45Wo37 to Fs72Wo24), with smaller amounts of Fe-rich olivine and other trace minerals (Rubin et al. 2000). Note that Los Angeles itself is only a quarter (by volume) wollastonite. Nonetheless, there is a remarkably close fit between our sample of Los Angeles and that for pure wollastonite, the calcium-rich pyroxene, in contrast to the other pyroxenes. We conclude that this close match is primarily a function of the FeO and CaO content of these samples, rather than the physical arrangement of the oxides within the minerals.

Figure 8.

 The heat capacity of the shergottite meteorite Los Angeles (black dots) compared with literature values for the heat capacities of various pyroxene minerals (Krupka et al. 1985). Note the close correlation between the meteorite and the Ca-rich pyroxene wollastonite.

Note that Los Angeles is a significantly shocked meteorite, as is evidenced by the presence of maskelynite. Min et al. (2004) estimate that it may have experienced a shock of higher than 45 GPa and a shock temperature above 450 °C. Furthermore, the low thermal conductivity reported here (measured on the same sample as that used for the heat capacity measurements) suggests that this sample has significant cracking that inhibits the transport of heat, similar to what is seen for ordinary chondrites. Macke et al. (2011) report a porosity of 8.1 ± 1.0% for Los Angeles (measured on a different sample than the one used here), again similar to ordinary chondrites. Nonetheless, this physical processing, which clearly controls the thermal conductivity of the sample, does not seem to have significantly affected its heat capacity.

This very intriguing result for one basaltic meteorite, that the heat capacity is well modeled by the elemental abundance of the sample and is not significantly affected by its shock state, needs confirmation for other meteorite types, especially ordinary chondrites. This remains for future work.

Conclusions

In the absence of data for various meteorite types, most modelers have assumed that the thermal properties of asteroidal or meteoritic materials could be estimated from the laboratory measurements of minerals typically found in those meteorites. Our results here, confirming earlier work, indicate that this may not be too far off for heat capacity (though further data are needed to confirm this result) so long as one recognizes its strong dependence on temperature, which has been neglected in many models. But this assumption may be seriously in error for thermal conductivity. The thermal conductivity of meteorites is significantly lower, by a factor of three to ten, than that of the pure minerals from which they are made.

The one macroscopic quantity of a meteorite that appears to be a good predictor of its thermal properties is its porosity. We have found a linear relationship between meteorite thermal conductivity and the inverse of the porosity. Even here, however, we have seen an anisotropy in thermal conductivity of 25% within a single sample, and a similar variation in conductivity from sample to sample of the same meteorite. Presumably, however, the parent body experienced numerous shock events resulting in a random orientation of internal cracks throughout the fabric of its material, and so the anisotropies seen in our millimeter-scaled samples will average out over sample sizes significantly larger than the extent of the cracks causing this anisotropy.

For temperatures of interest to planetary scientists, the variation of thermal conductivity with temperature appears to be small. It appears that once phonon activation energies have been reached at temperatures above 50–100 K, the transport of these phonons is controlled primarily by the presence and orientation of cracks, which inhibit the flow of heat more or less uniformly at all temperatures. Above 100 K, for most cases, the thermal conductivity of a meteorite can be assumed to be nearly constant with temperature. However, both literature data on minerals of interest and in our one measurement so far of a meteorite, we find a very strong (roughly linear) relationship between temperature and heat capacity from very low temperatures up to at least 300 K.

Care must, of course, be taken in applying these results directly to modeling the thermal properties of asteroids. The thermal inertia of an asteroid’s surface, which is an important factor in modeling the evolution of asteroid orbits and spins (via the effects related to the Yarkovsky reradiation of thermal energy) will depend strongly on the properties of the asteroid’s surface. Radar results (Magri et al. 2001) suggest that most asteroid surfaces have a macroporosity of 30–70%, which clearly will significantly lower the thermal inertia, consistent with inertias actually measured for asteroid surfaces (Delbo et al. 2007).

Likewise, noting the strong effect we find between sample porosity and thermal conductivity, one must consider whether the microporosity of the samples measured here is actually representative of their state while inside an asteroid. Meteorites now in our labs must have experienced strong shocks both when they were ejected from their parent asteroid and when they impacted the Earth. Thus, our hand samples may be biased toward materials that are more shocked than material within an asteroid itself. Given that the effect of shock on originally unshocked material is to compress and thus reduce meteorite porosity, this suggests that material within an asteroid may be even more porous, and thus lower in thermal conductivity, than the samples we report here. However, we note (Consolmagno et al. 2008) that there is only a very weak dependence of porosity on shock state (above shock state 1). Except for essentially unshocked (state 1) meteorites, once a meteorite has been compressed and cracked by shock, repeated or stronger shocks do not significantly alter its porosity. Given the likelihood that material in asteroids has been shocked many times, it seems unlikely that the final larger shocks which delivered material from the asteroids to our labs would have significantly altered the material’s porosity or thermal properties. Nonetheless, this is an assumption that one must keep in mind when using these data to model asteroid interiors.

With the data presented here, we have roughly doubled the number of meteorites for which thermal conductivities have been directly measured. However, many meteorite types remain unmeasured, and many more ordinary chondrites of varying physical and chemical nature need to be measured to confirm and refine the trends suggested by the data in hand. And the measurement of heat capacities as a function of temperature for many other meteorite types is also a high priority for future work.

Acknowledgments

Acknowledgments–– CO acknowledges support from the Trustees of Boston College and DTB was supported by NASA Grants NX09AD91G and NNG06GG62G from the Planetary Geology and Geophysics Program. We thank Jason Lanshley at the Los Alamos National Laboratory for his assistance with the heat capacity measurements of the Los Angeles meteorite.

Editorial Handling–– Dr. Michael Gaffey

Ancillary