The Volatility Inversion of Silicon Oxide and Magnesium Oxide at Evaporation of Hasp–Glasses on the Moon

Lunar and Planetary Science XLVIII (2017) 1133.pdf

THE VOLATILITY INVERSION OF SILICON OXIDE AND MAGNESIUM OXIDE AT EVAPORATION OF HASP–GLASSES ON THE MOON. O. I. Yakovlev and S. I. Shornikov, Vernadsky Institute of Geochemistry & Analytical Chemistry of RAS, Kosygin st., 19, Moscow, 119991, Russia; e-mail: [email protected].

Acidity-basicity is a universal characteristic of any compositions which are below this limit can be formed melt. To test the effectiveness of its application to the only at evaporation differentiation and thus this limit volatility of melt components we investigated the im- relation is a marker for the HASP–glasses identifica- pact HASP–glass lunar regolith. HASP–glasses formed tion. Fig. 1 also presents selected results of evaporation on the lunar surface through an impact process involv- experiments on lunar aluminous basalt [3]. These show ing high-temperature rock melting and deep selective that the compositions of the most HASP–glasses were evaporation [1]. HASP–glass composition features probably formed in the range of evaporation mass loss- high Al2O3 and poor SiO2 due to the high temperature es of ~20–50 %, corresponding to an evaporation tem- selective evaporation of melts with loss of easily and perature of ~1750–1900 °C [4]. moderately volatile oxides such as Na2O, K2O and Our experiments on the evaporation of chondrites FeO, partial loss of SiO2, and accumulation in the resi- and lunar basalt [3, 5, 6] clearly showed that during the due of CaO and Al2O3. evaporation when the SiO2 content in the melt decreas- es significantly the MgO content increases slightly in- dicating that the MgO is low volatility compared to SiO2. However, this experimental observation was not reflected in the actual compositions of HASP–glasses as higher content of MgO relative to SiO2 in glasses is not observed. On the contrary, there is a very low ratio of MgO / SiO2 (0.01 to 0.1) and very low contents of MgO – an average of 0.8 mass. % [2]. The composi- tional features of HASP–glasses in the main oxides of K2O, Na2O, FeO, SiO2, CaO and Al2O3 are easily ex- plained by evaporative differentiation of the melt components, the MgO behavior contradicts to the evaporative mechanism of HASP–glasses formation. To resolve these questions, we applied the Korgin- sky theory on acid-base interaction components in the melt [7]. According to this theory, the basicity index of a melt is determined by the ion oxygen activity that is Fig. 1. SiO2 / Al2O3 ratio dependence in the rock of always present in the melt due to the partial dissocia- anorthosite-norite-troctolite moon series (ANT series) tion of the oxide components. The effect of ion oxygen (violet circles) and HASP–glasses (red circles) vs. the activity is different for different oxides and may be the total refractory content of melt components of CaO and opposite for oxides contrasting in their individual acid- Al2O3. The experimental evaporation trend of the lunar base properties. For example, an increase in ion oxygen aluminous basalt is shown by triangles. Melt mass loss activity should lead to an increase of the activity of the at evaporation (M) is indicated for the three experi- basic oxide MgO and decrease of the activity of the mental points. The majority HASP–glass compositions acidic oxide SiO2 as the silicon oxide in the melt is not are between the points (~20–50 %). the donor of the oxygen ions but the acceptor. The component activity in the melt determines their component volatility in the vapor according to a general- Fig. 1 compares the SiO2 / Al2O3 ratio in ANT ized Raoult-Henry law: rocks with a population of HASP–glasses [2] vs. the total content of refractory oxides of CaO and Al2O3. pi = p°i ai = p°i xi γi , (1) The rock and HASP–glass compositions fall in an al- most linear decreasing relationship of SiO2 / Al2O3 where pi is the partial pressure of the i-th component which becomes smaller as calcium and aluminum ox- above the melt; p°i is the pressure of the i-th pure com- ides increase. The dotted line indicates the SiO2 / Al2O3 ponent; ai is the activity of the i-th component in the ratio in pure anorthite (equal to 1.18). For the moon melt; xi is the mole fraction of the i-th component in the this may be a petrochemical limit below which the gen- melt; γi is the activity coefficient of the i-th component eration of igneous anorthositic rocks is impossible. All Lunar and Planetary Science XLVIII (2017) 1133.pdf

in the melt. Relation (1) shows that the volatility of the oxide above the melt depends on three parameters: 1) the vapor pressure of the pure oxide; 2) the oxide concentration in the melt; 3) the interaction nature of oxide in the melt. The last two parameters determine the activity of the component and its behavior during evaporation from the melt. The properties of HASP–glass compositions may be modeled as the CMAS oxide system. The main oxides of the system on the acid-basic properties are di- vided into acid – SiO2, amphoteric – Al2O3 and two basic – CaO and MgO. CaO is the major basic oxide and is a major donor of oxygen ions in the melt and therefore its concentration plays a major role in deter- mining cause of the melt basicity. This also implies that the MgO and SiO2 activity depends on the CaO concentration. According to acid-base interaction in high- calcium melts the MgO will have high volatility and it can be higher than the SiO2 volatility [8]. It becomes clear why the MgO low content ob- serves in the HASP–glasses. It’s easy to imagine that during the melt evaporation the CaO content increases (and the melt basicity increase too) and so the MgO activity (and volatility) will increase and the SiO2 activity (and volatility) will fall. As a result of the evaporation process the MgO content in the residual melt will decrease and SiO2 content will increase relative to MgO. For HASP–glasses that means as the CaO content will increase, the MgO / SiO2 ratio will decrease. Fig. 2. The MgO / SiO2 dependence of CAIs in chon- Chemical analyses of the HASP–glasses [2] confirm drites [9] (a), and HASP–glasses (b) vs. the CaO con- this trend (Fig. 2b). tent. The figures show the regression line and correla- It is interesting to note that the evaporation of some tion coefficient (r). CAIs, which also belong to CMAS system, may be also described in the framework of the melt acid-base theory. Fig. 2a shows the same behavior of oxides on the References: [1] Naney et al. (1976) LPS VII, 155. evaporation of CAI melts [9] in spite of HASP–glasses [2] Warren (2008) GCA, 72, 3562–3585. [3] Markova and CAIs have absolutely different genesis. The et al. (1986) Geochem. Int., 19, 1559–1569. HASP–glasses compositions were formed during the [4] Yakovlev et al. (2011) Geochem. Int., 49, 227–238. impact process whereas the CAIs formed in the course [5] Yakovlev et al. (1984) Meteoritica, 43, 125–133 of a complex process including high temperature melt (in Russian). [6] Yakovlev et al. (1987) Meteoritica, evaporation in the event of unknown nature. 46, 104–118 (in Russian). [7] Korginsky (1959) Dokl. AN SSSR, 128, 383–396. [8] Shornikov et al. (2017) The present study was supported by RAS Presidium LPSC XXXXVIII. [9] Grossman et al. (2000), 64, Program #7 (Experimental and theoretical studies of 2879–2894. Solar system objects and star planetary systems. Tran- sients in astrophysics).