The effect of water on the sulfur concentration at sulfide saturation (SCSS) in natural melts
The effect of water on the sulfur concentration at sulfide saturation (SCSS) in natural melts
复制标题
DOI：
10.1016/j.gca.2015.03.022
复制
发表时间：
2015-07-01
影响因子：
5
通讯作者：
Baker, Don R.
中科院分区：
文献类型：
--
作者：
Fortin, Marc-Antoine;Riddle, Jacqueline;Baker, Don R.
我们实验研究了水对1250摄氏度、1 GPa玄武岩、安山岩和流纹岩熔体成分中硫化物饱和时硫浓度(SCSS)的影响，以及在活塞-圆柱体装置中缓冲在石墨-二氧化碳(CCO)缓冲液或其附近的氧逸度。用拉曼光谱测量了硫化物饱和熔体中的水浓度。实验范围从无水到7.3wt.%的总溶解水。实验表明，在所有熔体成分中添加水后，SCSS每重量增加约100ppm。结合这些数据和以前的研究，建立了两个计算SCSS随温度、压力、熔体组成和总溶解水量的函数的模型。第一个模型，模型A，通过区分铁和铁来考虑氧逸度对熔体成分的影响，并使用广义成分参数MFM值来描述熔体成分。第二个模型，模型B，使用全铁作为FeO，以及每100克熔体中主要氧化物的摩尔分数来描述熔体组成。我们发现，模型B再现了我们校准数据集的SCSS，以及独立的测试集和自然的高镍数据集，相对偏差约为5%，显著好于模型A，并且在几乎所有情况下都在测量不确定度范围内。模型B还复制了一组硫化物饱和的硅质熔体的测试集，其硫浓度接近电子探针检测的极限，相对含量在20%以内。因此，我们倾向于主要氧化物的经验摩尔分数方法，并提出了以下公式来计算SCSS：LN(S，ppm)(SCss)=-34.784-5772.3/T-346.54 P/T-20.393XH(2)O-25.499XSiO(2)-18.344XTiO(2)-27.381XAl(2)O(3)-17.275XFeO-22.398XMgO-20.378XCaO-18.954XNa(2)O-32.194XK(2)O其中T是以开尔文为单位的温度，P是以GPA为单位的压力。和X是每100克熔体中氧化物的摩尔分数。该模型适用于压力从1大气压到5 Gpa，温度从1050℃到1800℃，总熔水浓度从无水到7.3wt.%的陆地熔体，从科马提岩到流纹岩。模型B也适用于火星玄武岩熔体(包括在校准数据集中)和许多月球成分，但不适用于高钛的母质玄武岩。(C)2015爱思唯尔有限公司。保留所有权利。
We experimentally investigated the effect of water on the sulfur concentration at sulfide saturation (SCSS) in basaltic, andesitic, and rhyolitic melt compositions at 1250 degrees C, 1 GPa, and oxygen fugacities buffered at, or near, the graphite-carbon dioxide (CCO) buffer in a piston-cylinder apparatus. Water concentrations were measured in the sulfide-saturated melts using Raman spectroscopy. Experiments varied from anhydrous to 7.3 wt.% total dissolved water. The experiments demonstrate an increase in the SCSS of approximately 100 ppm per wt.% added water in all melt compositions. Combining these data with previous studies, two models for calculating the SCSS as a function of temperature, pressure, melt composition, and total dissolved water were created. The first model, Model A, incorporates the effect of oxygen fugacity in the melt composition by discriminating between ferric and ferrous iron and uses a generalized composition parameter, the MFM value, to describe melt composition. The second model, Model B, uses total iron as FeO and the mole fractions of major oxides per 100 g of melt to describe the melt composition. We find that Model B reproduces the SCSS of our calibrating data set, as well as an independent test set and a natural, high-Ni data set, to within about 5% relative variation, significantly better than Model A and in almost all cases within measurement uncertainties. Model B also reproduces a test set of sulfide-saturated silicic melts with sulfur concentrations near the limit of detection by electron microprobe to within 20% relative. We thus favor the empirical mole fractions of major oxides approach, and present the following equation to calculate the SCSS:ln(S, ppm)(SCSS) = -34.784 - 5772.3/T - 346.54 P/T - 20.393XH(2)O - 25.499XSiO(2) - 18.344XTiO(2) - 27.381XAl(2)O(3) - 17.275XFeO - 22.398XMgO - 20.378XCaO - 18.954XNa(2)O - 32.194XK(2)Owhere T is the temperature in Kelvin, P is the pressure in GPa, andXare the mole fractions of oxides per 100 g of melt. The model is applicable at pressures from 1 atm to 5 GPa, temperatures from 1050 to 1800 degrees C, and total melt water concentrations from anhydrous to 7.3 wt.% in terrestrial melts from komatiitic to rhyolitic in composition. Model B is also applicable to Martian basaltic melts (which are included in the calibration data set) and to many lunar compositions, but not high-Ti mare basalts. (C) 2015 Elsevier Ltd. All rights reserved.