Nitrogen diffusion in silicate minerals, with implications for nitrogen transport and cycling in the lithosphere
Nitrogen diffusion in silicate minerals, with implications for nitrogen transport and cycling in the lithosphere
复制标题
DOI：
10.1016/j.chemgeo.2019.04.006
复制
发表时间：
2019-06
期刊：
影响因子：
3.9
通讯作者：
E. Watson;D. Cherniak;M. Drexler;R. Hervig;M. Schaller
中科院分区：
文献类型：
--
作者：
E. Watson;D. Cherniak;M. Drexler;R. Hervig;M. Schaller
采用C-H-O-N蒸气内扩散、NH 4Cl熔融内扩散和离子注入后加热使N迁移的实验方法，确定了氮在正长石、石英、橄榄石和单斜辉石中的扩散规律。大多数实验是在接近1巴的压力下进行的，但也有七个实验是在Ptotal = 1 GPa的活塞-气缸装置中进行的。C-H-O-N源实验是“湿的”（存在H2O）;离子注入样品上的那些是干的。所有氮源都高度富集15 N，以避免表面污染导致的杂散14 N信号。15 N的使用还使得能够通过使用15 N（p，αγ）12 C反应的核反应分析（NRA）进行深度剖析，其具有约10 ppm的检测限。NRA和西姆斯对三个样品的氮分布进行了表征，结果相似。对四种矿物中的每一种进行了总共17到28次扩散系数测量（总共83次），得出的结果符合每一种矿物的阿克里尼乌斯关系式D = D 0 exp（−Ea/RT）。得到了以下指前常数（D 0）和活化能（Ea）的值：矿物合金（D 0，m2/s）Ea（kJ/mol）T范围（°C）正长石−13.9 ± 0.495.4 ± 6.6500- 900石英−12.4 ± 0.3147.3 ± 5.7550- 1150橄榄石−14.2 ± 0.3135.2 ± 6.6650- 1400单斜辉石−14.1 ± 0.3135.6 ± 7.4750- 1300在给定的温度下，各种实验策略产生的结果通常是不可区分的，在整个扩散数据集中，总压或H2O的影响是不可辨别的。实验涉及从外部N源的扩散提供定性的洞察N的结构中的兼容性。氮的浓度在正长石中达到最高水平（~100-8000 ppm原子），石英、橄榄石和单斜辉石的浓度在几十到几百ppm的范围内。新的扩散定律使我们能够模拟在地壳深处获得的N的扩散释放过程中没有流体的折返和冷却。一般来说，正长石是所研究的矿物中保持性最差的，并且在从600 °C以构造典型速率冷却期间将损失其大部分N。橄榄石和单斜辉石对N.石英是中间的镁铁质矿物和长石之间的保留，但在所有情况下，与周围环境的N交换的程度严重依赖于特定的Tpath的矿物的利益。在俯冲设置含N流体释放到地幔楔，扩散是足够快的均匀化个别镁铁质矿物颗粒相对于N在地质上合理的时间尺度，但通过体积扩散的露头或区域尺度上的平衡被排除在我们的数据。
Diffusion laws for nitrogen in orthoclase, quartz, olivine and clinopyroxene were determined using a combination of experimental strategies that included: in-diffusion from C-H-O-N vapor; in-diffusion from molten NH4Cl; and ion implantation followed by heating to mobilize N. Most experiments were conducted at pressures near 1 bar, but seven experiments were also run in a piston-cylinder device atPtotal= 1 GPa. The C-H-O-N-source experiments were “wet” (H2O present); those on ion-implanted samples were dry. All N sources were highly enriched in15N to avoid spurious14N signals resulting from surface contamination. Use of15N also enabled depth profiling by nuclear reaction analysis (NRA) using the15N(p,αγ)12C reaction, which has a detection limit of ~10 ppm. Nitrogen profiles in three samples were characterized by both NRA and SIMS, with similar results. A total of 17 to 28 diffusion coefficient measurements were made on each of the four minerals (83 in all), yielding results that conform for each mineral to an Arrhenius relation of the formD=D0exp(−Ea/RT). The following values for the pre-exponential constant (D0) and activation energy (Ea) were obtained:Minerallog(D0, m2/s)Ea(kJ/mol)T range (°C)Orthoclase−13.9 ± 0.495.4 ± 6.6500–900Quartz−12.4 ± 0.3147.3 ± 5.7550–1150Olivine−14.2 ± 0.3135.2 ± 6.6650–1400Clinopyroxene−14.1 ± 0.3135.6 ± 7.4750–1300The various experimental strategies yield generally indistinguishable results at a given temperature, and no significant effects of total pressure or H2O are discernible in the overall diffusion data set. Experiments that involved in-diffusion from external N sources provide qualitative insight into the compatibility of N in the structures of the four minerals. Nitrogen concentrations attain the highest levels in orthoclase (~100–8000 ppm atomic), with quartz, olivine and clinopyroxene falling in the tens to hundreds of ppm range. The new diffusion laws enable us to model the diffusive release of N acquired at depth in the crust during fluid-absent exhumation and cooling. In general, orthoclase is the least retentive of the minerals investigated, and will lose most of its N during cooling from 600 °C at tectonically typical rates. Olivine and clinopyroxene are the most retentive of N. Quartz is intermediate in retentivity between the mafic minerals and feldspar, but in all cases the extent of N exchange with the surroundings depends critically on the specifict-Tpath of the mineral of interest. In subduction settings where N-bearing fluid is released to the mantle wedge, diffusion is sufficiently fast to homogenize individual mafic mineral grains with respect to N over geologically plausible time scales, but equilibration on the outcrop or regional scale via volume diffusion is precluded by our data.