Thermochemical Equilibrium Modeling Indicates That Hg Minerals Are Unlikely to Be the Source of the Emissivity Signal on the Highlands of Venus

Several of the highlands of Venus exhibit unexpectedly low radar emissivity compared to that of the lowlands. The source has been hypothesized to be a mineral with a high dielectric constant. Recently HgTe (coloradoite) has been suggested to explain the low emissivity signal; however, little research has been completed to verify its stability on Venus. In this project, we used a Gibbs free energy minimization software to investigate whether HgTe, as well as HgS and HgSe, can form at simulated highland conditions. According to our calculations, approximately 1.3 wt% of mercury in the crust needs to be outgassed in order for HgS to be stable at 4 km in altitude. In addition, approximately 250 ppb of tellurium in the crust needs to be outgassed for HgTe to precipitate at the same altitude. The required mercury abundance for HgSe to be stable at this altitude is less, approximately 0.6 wt%; however, this is significantly larger than the 10–90 ppb generally present in basaltic rocks on Earth. Therefore, Hg-bearing minerals are likely not the source of the low radar emissivity signal.


Introduction
Numerous topographic features on Venus exhibit an anomalous change in radar emissivity at some altitudes above the mean planetary radius. First observed by Arecibo (λ = 12.6 cm), it was later observed by Pioneer Venus (λ = 17 cm) and Magellan (λ = 12.6 cm) (Campbell et al. 1976;Ford & Pettengill 1983;Garvin et al. 1985;Arvidson et al. 1991). This unusual phenomenon is present on several geologic features such as coronae, mountain belts, tesserae, and volcanoes Brossier et al. 2020;Brossier & Gilmore 2021;Ford & Pettengill 1983;Garvin et al. 1985;Gilmore et al. 2017;Klose et al. 1992;Pettengill et al. 1992). The change in emissivity is not uniform across Venus and has manifested as various patterns depending on the geologic feature and location. Most of the emissivity features can be divided into two major patterns (Brossier et al. 2020;Brossier & Gilmore 2021). The first group exhibits a gradual decrease in emissivity at altitudes greater than ∼6055 km (mean planetary radius = 6051.80 km; Maxwell Montes, Fortuna Tessera, Artemis Corona), while the second group displays a sudden decrease and then sometimes an abrupt increase in emissivity between ∼6053 and 6056 km in altitude (Sapas Mons, Zisa Corona, Thetis Regio; Brossier et al. 2020;Brossier & Gilmore 2021;Klose et al. 1992;Treiman et al. 2016). Not all geologic features exhibit a low emissivity signal at high altitudes. Different surface compositions, age, or the geologic feature having a relatively low maximum altitude may explain the absence of an emissivity change (Brossier et al. 2020;Brossier & Gilmore 2021).
The source of the low emissivity signal has not been determined but may be attributed to a mineral with a higher dielectric constant than that of the surrounding basaltic lowlands (ε ≈ 5; Ford & Pettengill 1983;Klose et al. 1992;Pettengill et al. 1988;Treiman et al. 2016;Tyler et al. 1991). Based on the two major patterns, it has been posited that the gradual decrease in emissivity with altitude is potentially due to a semiconducting material (Klose et al. 1992;Pettengill et al. 1991;Schaefer & Fegley 2004;Treiman et al. 2016;Brossier et al. 2020;Brossier & Gilmore 2021). Meanwhile, the abrupt decrease and (occasional) increase in signal could be due to ferroelectric minerals Shepard et al. 1994;Treiman et al. 2016;Brossier et al. 2020;Brossier & Gilmore 2021). A ferroelectric mineral displays a sudden increase in dielectric constant, observed as a sudden decrease in emissivity, at a very particular temperature (transition temperature), which is dependent on the mineral (Treiman et al. 2016). Ferroelectric mineral candidates include perovskites, as well as chlorapatite (Ca 5 (PO 4 ) 3 Cl)-a mineral commonly found in igneous rocks on Earth Treiman et al. 2016). Brossier et al. (2020) and Brossier & Gilmore (2021) identified a decrease in emissivity at two different altitudes (~6054 and~6056 km), suggesting the presence of two ferroelectric minerals.
Meanwhile, it is postulated that a semiconducting material on Venus could have formed from a chemical reaction between the surface materials and the atmosphere, or could be a volatile mineral that precipitated from the atmosphere onto the cooler, high-altitude regions (Brackett et al. 1995;Klose et al. 1992;Pettengill et al. 1988Pettengill et al. , 1996Schaefer & Fegley 2004). An assortment of semiconductor material candidates have been suggested to explain this signal. One of the first minerals to gain traction was pyrite (FeS 2 ; Ford & Pettengill 1983;Pettengill et al. 1988;Klose et al. 1992). For several years its stability was highly debated Hong & Fegley, 1997;Wood & Brett 1997), but recent phase equilibria modeling suggests that pyrite may exist on the highlands depending on the fO 2 and fS 2 (Zolotov 2018;Semprich et al. 2020). Volatile minerals and elements, such as tellurium, Bi 2 S 3 , PbS, and HgTe, could potentially precipitate onto the cooler highlands of Venus, causing low emissivity (Brackett et al. 1995;Pettengill et al. 1996;Schaefer & Fegley 2004;Kohler 2016).
Coloradoite (HgTe) is a recent addition to the lowemissivity candidate list (Kohler et al. 2013). It was originally formed in a Venus simulation chamber when a mixture of HgS and tellurium was heated to highland-relevant conditions (380°C/55 bar; CO 2 atmosphere). It has a dielectric constant of 15-21 (Adachi 2004;Secuk et al. 2014). However, further investigation on HgTe's long-term stability, the abundance necessary for it to condense, and the altitude at which it begins to stabilize has yet to be completed. During this project, we sought to thermodynamically model HgTe's (as well as other mercury-bearing minerals') ability to condense on the highlands using elemental abundances applicable to Venus. We completed calculations to determine how much of each of these elements is necessary for a mercury-bearing mineral to precipitate at 4 km from datum (6052 km). A value of 4 km was selected because the emissivity begins to decrease around this altitude on Maxwell Montes, and as stated previously, researchers hypothesize that Maxwell Montes' emissivity anomaly likely originates from a semiconducting mineral (Brossier et al. 2020;Brossier & Gilmore 2021;Treiman et al. 2016). There are several other mountain ranges (Akna, Danu, and Freyja Montes) that also exhibit an emissivity drop around 4 km from datum (Brossier et al. 2020).

Mercury and Tellurium on Venus
The existence of mercury on Venus has been discussed several years before it was associated with the low emissivity signal. It was first considered by Lewis (1968Lewis ( , 1969, who suggested the presence of mercury and mercury halide clouds. In his calculations, he assumed that the abundance of mercury on Venus was similar to that of Earth's crust (8 × 10 −6 wt%). Due to Venus's high surface temperature (∼660-760 K), all the elemental mercury in the crust would have degassed, resulting in 1.4 × 10 19 g of mercury in the troposphere (Lewis 1969;Barsukov et al. 1981). Barsukov et al. (1981) used the same input as Lewis (1969) in their model to investigate Venus's cloud composition. Barsukov et al. (1981) determined that mercury would be gaseous at the surface and liquid/crystallized at 52 km in altitude. However, according to measurements obtained from Venera 12, the concentration of mercury in the aerosols did not exceed the sensitivity limit of 0.05 mg m −3 , meaning that the mixing ratio is <10 −8 at 62 km in altitude. They concluded that the mercury abundance on Venus may be 3.5 orders of magnitude less than that of Earth's crust (Barsukov et al. 1981). Although HgTe was not incorporated into their calculations, tellurium was, and they hypothesized that both tellurium and mercury may have been depleted from Venus early in its formation (Barsukov et al. 1981).
In situ detection of mercury on Venus was attempted using the neutral mass spectrometer on board the Pioneer Venus Large Probe (Hoffman et al. 1980). The instrument obtained a few scattered counts, but there was no strong evidence for mercury. However, it was suggested that mercury could have gone unobserved if it passed through the inlet and condensed on the walls of the neutral mass spectrometer (Seiff 1993). The instrument was pre-calibrated for mercury, and using these data, it was determined that the upper limit of the mixing ratio on Venus was 5 ppm.
Over a decade later, Brackett et al. (1995) modeled the volatile transport of various volatile minerals, including HgS and HgF 2 , from the hotter Venusian lowlands to the cooler highlands using a vapor transport model. Their models showed that these two minerals were highly volatile and could be transported from the surface to >3.5 km in altitude in under 80 yr. Since their model did not include tellurium, Port et al. (2020) amended their Figure 5 to incorporate Te 2 and found that it is also highly volatile. HgTe was not incorporated into the figure.
Tellurium was first postulated to be the source of the low emissivity signal by Pettengill et al. (1996) owing to its high conductivity and volatility. This led to Schaefer & Fegley (2004) incorporating tellurium, as well as mercury, into their chemical equilibrium calculations to investigate the low emissivity signal. Their model assumed that a volatile mineral/element was precipitating onto the surface of the high-altitude regions. Their results indicate that tellurium condenses at 6098 km in altitude (planetary mean radius = 6051.4 km), well above the tallest mountains on Venus. In order for tellurium to condense at 6054 km (the start of the low emissivity signal in their model), approximately 182 ppm, or ∼60,000 times the abundance in the basaltic oceanic crust of Earth, is necessary. They incorporated the mercury abundance, as given by Barsukov et al. (1981), into their model; however, the results for mercury were never discussed, and it is not stated whether HgTe was included in their calculations. Kohler et al. (2013) was the first group to experimentally study the stability of mercury-and tellurium-bearing minerals at simulated Venus conditions. They concluded that at simulated highland conditions (380°C/55 bar; CO 2 atmosphere) HgS will completely volatilize and tellurium will become TeO 2 or, if Hg is present, HgTe. Thus, neither HgS nor tellurium can explain the low emissivity signal (Kohler et al. 2013;Kohler 2016). It was in these experiments that they formed HgTe and concluded that it could potentially explain the emissivity signal.

Methods
This project utilized Thermo-Calc (https://thermocalc.com), a Gibbs free energy minimization modeling software, and their ssub3 and ssub6 thermodynamic databases. The user inputs the temperature, pressure, and bulk composition of the system, and the program minimizes the Gibbs free energy until the system is in equilibrium. This method was selected because it has been used by many researchers to investigate the near-surface environment of Venus (Barsukov et al. 1981;Klose et al. 1992;Zolotov 1996;Schaefer & Fegley 2004;Jacobson et al. 2017). While it is debated that the atmosphere is not in equilibrium Zolotov 2018), kinetic data of chemical interactions on Venus are still limited, which hinders the ability to precisely model the Venusian atmosphere (Jacobson et al. 2017). However, Thermo-Calc is appropriate for this project because our focus is not to precisely model the atmosphere of Venus but to investigate the formation of HgS, HgTe, and HgSe on the highlands of Venus, which is strongly tied to the temperature/pressure and abundance of mercury, tellurium, and selenium.
First, we chose to model the atmosphere without the incorporation of mercury and tellurium to confirm that the model simulates the atmosphere of Venus. These data will then be compared to our data after the incorporation of mercury, tellurium, and selenium to determine which gases may have reacted with these elements. The major gases incorporated into the simulated atmosphere were CO 2 , N 2 , SO 2 , H 2 O, CO, OCS, HF, and HCl (Table 1). Several gases were omitted because their abundance is negligible on Venus and would not affect our results (e.g., H 2 S = 3 ± 2 ppmv). However, gases such as HF and HCl were incorporated because mercury can form minerals with fluorine and chlorine. The inputted abundances for each gas were obtained by in situ measurements, ground-based observation, or thermodynamic models (Arney et al. 2014;Krasnopolsky & Pollack 1994;Zolotov 2018). In our model, we used the abundances given by Zolotov (2018), which are displayed in Table 1 (Column (2)). The only exception was for OCS, which was set to 27 ppm. Currently, there have been no measurements of OCS near the surface; however, spectroscopic analysis of the night side of Venus shows that OCS may increase at lower altitudes to almost 10 ppm at 26 km in altitude (Pollack et al. 1993). Kinetic modeling of the lower atmosphere has shown that it may be around 27-28 ppm (Krasnopolsky & Pollack 1994;Krasnopolsky 2007). Thermochemical modeling of the CO-OCS-CO 2 -SO 2 system shows that, at chemical equilibrium, OCS is 28 ppmv (Zolotov 1996(Zolotov , 2018. Most of our calculated gases ( Table 1, Column (4)) are in good agreement with the input abundances. Only three gases, CO, SO 2 , and OCS, exhibit differences when compared with the input abundances. The abundance of SO 2 was greater than the input abundance, and the OCS and CO abundances were less. This indicates that the elements composing CO and OCS were partially incorporated into SO 2 and CO 2 . The decrease of CO and/or OCS has been observed in other thermodynamic equilibrium calculations Jacobson et al. 2017;Krasnopolsky 2007) and could be a result of the thermodynamic values used in the model  or due to kinetics having a strong role in these reactions (Jacobson et al. 2017). For instance, analysis of some potential OCS-consuming reactions on Venus indicates that they will never reach equilibrium (Jacobson et al. 2017). The temperature/pressure profile used in the model was the Venus International Reference Atmosphere (VIRA), where z = 0 is the planetary radius of 6052 km (Seiff et al. 1985).
Once our simulated atmosphere was confirmed to reasonably match the values from the literature, mercury, tellurium, and selenium were added to this model. Four different scenarios were investigated: (1) mercury, (2) mercury and tellurium, (3) mercury and selenium, and (4) mercury, tellurium, and selenium. We included selenium in this list because, like tellurium and sulfur, it is a chalcogen, it is normally found in close proximity to tellurium on Earth (Ciobanu et al. 2010;Cook & Ciobanu 2004;Cook et al. 2007), and it can be outgassed from volcanoes (Hinkley et al. 1999;Lambert et al. 1988). Lastly, HgSe has a similar dielectric constant to HgTe (static dielectric constant of 25.6; Adachi 2004), and selenium is found in higher abundance in Earth basalt than tellurium (Lodders & Fegley 1998;Lissner et al. 2014). Selenium was also incorporated into the Schaefer & Fegley (2004) and Barsukov et al. (1981) models. Barsukov et al. (1981) hypothesized that selenium was also depleted from Venus. The Schaefer & Fegley (2004) model revealed that selenium condenses at >46.6 km. HgSe was not discussed in either model.
To calculate the abundance of mercury, tellurium, and selenium in the atmosphere, we used the same method demonstrated in Schaefer & Fegley (2004): They used a ratio between the terrestrial oceanic crust and the atmosphere of Venus and determined the elemental abundance as a function of sulfur. Here "ni is the molar abundance of element i in the terrestrial oceanic crust, and Ai is the atomic abundance of element i in ppmv in the Venusian atmosphere." AS and nS are the abundances of sulfur in the atmosphere and crust of Venus, respectively. We chose to use the same terrestrial oceanic crust abundance for nS, 960 ppm (Alt 1995), as Schaefer & Fegley (2004) to more directly compare our results. Our AS was 177 ppmv. The abundances were entered into Themo-Calc, and a calculation was performed at each elevation. The lowest altitude in which the mercury-bearing mineral was stable was recorded.
We chose to start with 90 ppb because it is the upper limit. Assuming 90 ppb in the crust of Venus, HgS (cinnabar) would form at 34.4 km in altitude. Cinnabar is the low-temperature polymorph of metacinnabar. Cinnabar (α-HgS) is red and has a trigonal structure. Meanwhile, metacinnabar (β-HgS) is black and has a cubic structure (Ballirano et al. 2013;Potter & Barnes 1978). According to Kitts & Lodders (1998), a basaltic meteorite (eucrite) could potentially have up to 9120 ppb of mercury ( Table 2). Basaltic meteorites could be an analog for the crust of Venus and were selected as a result of their high upper limit of mercury (Kitts & Lodders 1998;Schaefer & Fegley 2004). When the mercury abundance was adjusted to 9120 ppb, HgS did not begin to form until 24.5 km in altitude (Figure 1, top). The abundance of mercury needed to be increased by an unrealistic amount, up to 1.3 wt%, in order for HgS (metacinnabar) to form at 4 km (Figure 1, bottom).
We compared the thermodynamic values for HgS in Thermo-Calc with a database presented in Barin (1995). The databases are similar, with the largest difference being enthalpy of formation, wherein Thermo-Calc is −59.0 kJ mol −1 at 298.15 K and −53.3 kJ mol −1 in Barin (1995). The slight difference will cause a small change in the results; however, an unrealistic abundance will still be necessary for condensation on the highlands.

Mercury and Tellurium
The average abundance of tellurium measured within MORBs, basalts, and tholeiitic basalts is 4.1 ± 4.4 ppb (Lissner et al. 2014;Yi et al. 2000). This value is very close to the 3 ppb measured in the oceanic crust on Earth (Lodders & Fegley 1998) and was the value used by Schaefer & Fegley (2004) in their calculations studying the low emissivity signal.
Since these values are similar, and to more accurately compare our results with theirs, we chose to start our calculations by setting tellurium = 3 ppb. In our first calculation, we set mercury = 9120 ppb and tellurium = 3 ppb and found that HgTe never forms. Next, we set mercury = 1.3 wt% because our previous calculation (Section 3.1) showed that 1.3 wt% of mercury is necessary for a mercury-bearing mineral to precipitate at 4 km, and we set tellurium = 3 ppb. In our results, HgS precipitates at 4 km; however, HgTe did not begin to form until 12.7 km in altitude (Figure 2, top). Our calculations showed that approximately 250 ppb of tellurium is required for HgTe to form at 4 km (Figure 2, bottom). Note. For simplicity, only elemental abundances for basaltic meteorites are displayed here. Data are obtained from Kitts & Lodders (1998). (N) = noncumulate; (C) = cumulate.
Since HgTe is a rare mineral and does not have wellestablished thermodynamic values, we recreated the data using a second database available through Thermo-Calc (ssub6). In this database, when mercury = 1.3 wt% and tellurium = 3 ppb, HgTe did not condense until 18.5 km in altitude. HgTe will not condense at 4 km unless tellurium = 4.1 ppm. The results obtained using the two different databases highlight the unrealistic abundance necessary for HgTe to precipitate at 4 km.

Mercury and Selenium
The average selenium value for MORBs and basalts is 111.4 ppb ± 56.6 (Lissner et al. 2014). This value is close to the 160 ppb measured in the oceanic crust (Lodders & Fegley 1998) and was the value that Schaefer & Fegley (2004) used in their calculations. Similar to the previous section, we chose to start with 160 ppb in order to compare our results to Schaefer & Fegley (2004). For our first calculation, we set mercury = 9120 ppb and selenium = 160 ppb and found that HgSe does not form until 23.1 km. Using 1.3 wt% of mercury and 160 ppb of selenium as inputs into the model, HgSe began to form at ∼0.4 km (Figure 3, top). For HgSe to begin precipitating at 4 km, either mercury needed to be decreased to 0.6 wt% (in this scenario HgS (metacinnabar) would not precipitate until 7 km; Figure 3, middle), or selenium needed to be decreased to 52 ppb (Figure 3, bottom).
Since HgSe is also a rare mineral, we compared the thermodynamic values in Thermo-Calc with Barin (1995). The databases had different thermodynamic values; the largest difference was the enthalpy of formation where Thermo-Calc has −59.4 kJ mol −1 at 298.15 K, and Barin (1995) has −43.5 kJ mol −1 . This difference will shift the stability of HgSe to even higher altitudes; however, without a better constraint on HgSe's thermodynamic values, we cannot determine the relative altitude of condensation.

Mercury, Tellurium, and Selenium
In our first calculation, we chose mercury = 9120 ppb, tellurium = 3 ppb, and selenium = 160 ppb and obtained, as before, that HgTe would not precipitate, HgSe precipitates at 23.1 km, and HgS (cinnabar) precipitates at 24.6 km (Figure 4, top). We adjusted the abundance of mercury to the lower-end value (90 ppb) and found that the first mineral to precipitate was HgSe at 32 km. Next, we used the abundances necessary for HgS, HgTe, and HgSe to form at 4 km as was determined in the previous sections: mercury = 1.3 wt%, tellurium = 250 ppb, and selenium = 52 ppb; however, instead of HgS, HgTe, and HgSe forming at 4 km, HgS (metacinnabar) forms at 4.1 km, HgTe forms at 4.2 km, and HgSe forms at 4.7 km  ( Figure 4, bottom). In order for these mercury-bearing minerals to form at 4 km, we needed to adjust the input abundances. For HgS to form at 4 km, the abundance of mercury needed to be slightly increased to 1.34 wt%. A more significant change was needed to get HgTe to form at 4 km; either tellurium needed to be increased to 266 ppb, or mercury needed to be increased to ∼1.4 wt%. Lastly, to get HgSe to form at 4 km, either selenium needed to be increased to 71 ppb, or mercury needed to be increased to 1.6 wt%.
We recreated this simulation using ssub6, which contains different thermodynamic values for HgTe. The results for HgS and HgSe formation were unaffected, but when mercury = 1.3 wt%, tellurium = 250 ppb, and selenium = 52 ppb, instead of HgTe forming at 4.7 km, it did not until 9.5 km. An abundance of 4.1 ppm of tellurium was necessary for HgTe to condense at 4 km.
The results presented here only used the thermodynamic data from the internal database (ssub3 and ssub6); therefore, the results for HgSe are only one possible end-member.
Our calculations highlight the competing reactions that can arise when several elements are present. Even slight changes in one abundance can alter the altitude at which several minerals form. Furthermore, differences in thermodynamic data can influence the altitude of stability.

Discussion
The data obtained from these models show that the abundance of mercury is the limiting factor when attempting to form HgS, HgTe, and HgSe. In order for HgS to be present at around 4 km, approximately 1500× more mercury is necessary on Venus than in basaltic meteorites (9120 ppb), and ∼150,000× more than what is commonly present in basaltic rocks on Earth (90 ppb).
Assuming that mercury is abundant (1.3 wt%) on Venus, the abundance of tellurium would need to be ∼83× the abundance in basalt on Earth (assuming 3 ppb), or ∼27× more  if the abundance was the upper limit of a basaltic meteorite (9.4 ppb; Kitts & Lodders 1998). However, using the ssub6 database, even more tellurium would be required for HgTe to form in the highlands.
If the abundance of selenium on Venus is comparable to that of Earth and only the Thermo-Calc internal database is considered, HgSe could potentially form on the highlands, but only if the abundance of mercury is around 0.6 wt%. This abundance is still ∼640× more mercury than present at the upper limit detected in basaltic meteorites (9120 ppb). However, assuming that the data present in Barin (1995) are more accurate, then HgSe is even less likely to be present at 4 km.

Summary
It is highly improbable for such a large quantity of mercury to exist on Venus. Assuming a mercury partial pressure of 50 mbar (the upper limit as claimed by Lewis 1968), past chemical equilibria calculations have shown that elemental mercury (Barsukov et al. 1981;Lewis 1968Lewis , 1969 and mercury sulfides (Lewis 1968(Lewis , 1969 cannot condense on the mountains because of the high surface temperature (Lewis 1968(Lewis , 1969Barsukov et al. 1981;Seiff 1993). The true partial pressure of mercury on Venus may be lower than 50 mbar. Calculations using 50 mbar have shown that mercury condenses at around 32.5 km in altitude (Seiff 1993); however, the neutral mass spectrometer failed to detect mercury in the atmosphere (Hoffman et al. 1980;Seiff 1993). It has been suggested that the altitude of the mercury clouds is suppressed by a convective layer in the troposphere, forcing the clouds to only exist beneath this altitude (Seiff 1993). It has also been hypothesized that mercury may be trapped within Venus owing to inefficient outgassing (Lewis 1968;Schaefer & Fegley 2004). However, even in the case of complete outgassing on Venus, an unrealistic abundance of mercury is necessary for mercury minerals to exist in the highlands: approximately five orders of magnitude greater than present in basaltic rocks on Earth. Furthermore, the abundance of mercury on Venus, as suggested by data obtained by Venera 12, may be 3.5 orders of magnitude less than on Earth (Barsukov et al. 1981). Therefore, HgTe and other mercury-bearing minerals are unlikely to explain the source of the low emissivity signals on the highlands of Venus.
There are several future Venus missions (DAVINCI, VERITAS, and EnVision) that may provide further insight into the source of the low emissivity signal. VERITAS and EnVision will have the capability to obtain the near-IR emissivity of the surface, which can be used to glean the possible surface composition (Garvin et al. 2020;Smrekar et al. 2020;Ghail et al. 2017;Dyar et al. 2020aDyar et al. , 2020b. Meanwhile, DAVINCI will be equipped with a Quadrupole Mass Spectrometer and Tunable Laser Spectrometer, which will collect data on trace gases in the atmosphere and may inform on the existence of volatile compounds in the highlands. This work was funded by the Zonta International Amelia Earhart Fellowship. S.T.P. was supported by an appointment to the NASA Postdoctoral Program at the NASA Glenn Research Center, administered by the Universities Space Research Association under contract with NASA. We would like to thank the anonymous reviewers for their valuable edits. A special thanks to Nathan S. Jacobson for going above and beyond to assist and check our work. All available data can be found on the PI's ResearchGate account: doi:10.13140/ RG.2.2.29605.83689.