We have studied the evaporation of Na, K and Mn from Al-Na-K- and Mn-rich silicates at various conditions. Total alkali oxide contents ranged from 5 to 20%. The evaporation rate of Na increases with temperature and decreasing oxygen fugacity and decreases with duration of heating. The loss of K is in all cases less pronounced than for Na. Heating in an evacuated vacuum furnace is more effective in removing Na and K from melt droplets than in furnaces with one atm gas flow of air or gas mixtures controlling the oxygen fugacity. The strong pumping required to keep the vacuum removes Na and K atoms very effectively. In all experiments, the rate of evaporation is determined by quasi-equilibrium between a thin layer of Na and K rich gas above the molten silicates. The results of the experiments are in agreement with several other studies.

In experiments with more than one sample in the furnace, equilibration of Na- and K-rich samples with Na- and K-poor samples occurred rapidly, mediated by the ambient gas phase.

The results of experiments with Mn in starting compositions showed much stronger losses of Na than Mn under a variety of conditions.

Thus the nearly chondritic Mn/Na ratios in the Earth cannot be the result of evaporation of Na and Mn in Earth-making materials, as the Mn/Na ratios in evaporation residues would be much higher than chondritic ratios. Such evaporation processes may have occurred in the parent material of Moon, Vesta and Mars.

The data suggest, in agreement with earlier hypotheses, that the high and variable contents of Na and K in chondrules require a gas phase high in Na and K equilibrating with chondrule melts. The volume of nebular gas parental to a certain type of chondrites was heated and Na and K were lost from the chondrule precursors to the gas phase. Subsequently the nebular parcel was compressed leading to higher partial pressures of Na and K. Flash heating then produced chondrule melts which incorporated some of the gaseous Na and K and then cooled rapidly. The large range of Na and K contents in chondrule melts reflects very local enrichments of Na and K in the gas phase. Despite these variations bulk chondritic meteorites have well defined bulk Na and K contents, implying a closed system during formation of chondrules and matrix.

我们研究了在各种条件下从富含Al-Na-K和Mn的硅酸盐中蒸发Na，K和Mn的过程。碱金属氧化物的总含量为5％至20％。Na的蒸发速率随温度的升高和氧气逸度的降低而增加，并随加热时间的延长而降低。在所有情况下，钾的损失都不如钠明显。在抽空的真空炉中加热比从一个大气压的空气或气体混合物控制氧气逸度的熔炉中的熔体液滴中除去Na和K更有效。保持真空所需的强劲抽吸非常有效地去除了Na和K原子。在所有实验中，蒸发速率由熔融硅酸盐上方富含Na和K的气体薄层之间的准平衡确定。实验结果与其他几项研究一致。

在炉中有多个样品的实验中，富Na和K的样品与贫Na和K的样品的平衡迅速发生，这是由环境气相介导的。

在各种条件下，在起始组合物中使用Mn进行的实验结果表明，Na的损失比Mn强大得多。

因此，地球上接近软骨的Mn / Na比不能是造土材料中Na和Mn蒸发的结果，因为蒸发残留物中的Mn / Na比将远远高于软骨的比。这种蒸发过程可能发生在月球，维斯塔和火星的母体中。

数据表明，与早先的假设相一致，球藻中Na和K的含量高而多变，要求气相中Na和K含量高，并与球茎熔体平衡。加热了特定类型球粒陨石的星云气体的体积，从球粒状前体到气相中损失了Na和K。随后，压缩了神经团，导致Na和K的分压更高。然后通过闪蒸产生软骨熔体，该熔体掺入了一些气态的Na和K，然后迅速冷却。软骨熔体中Na和K的含量范围很大，反映出气相中Na和K的局部富集。尽管存在这些变化，但块状软骨陨石具有明确的块状Na和K含量，这意味着在形成软骨和基质时会形成封闭的系统。