Trace element characteristics of sphalerite in the Haobugao Zn-Pb deposit, Inner Mongolia, and their geological significance

ACTA PETROLOGICA SINICA(2023)

引用 0|浏览6
暂无评分
摘要
The Huanggang-Ganzhuermiao metallogenic belt in the south segment of the Great Xing'an Range is one of the important polymetallic belts in northeastern China. It contains several Zn-Pb deposits, among which Haobugao is a representative one. At present, the occurrence mechanism of the trace elements in the Haobugao sphalerite is not clear, and the genetic type of the deposit is controversial. In this study, in situ trace element analysis by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) combined with machine learning method was used to better characterize the sphalerite compositions and explore the mechanism how the trace elements enter into the sphalerite, and furthermore, to reveal the genetic type of the Haobugao deposit. The results show that the Haobugao sphalerite is characterized by enrichment of Fe, Mn, Co, Cu, Se, Ag, Cd, In and Sn, and depletion in Ni, Ga, Ge, As, Mo, Sb, Au, Tl, Pb and Bi. Among these trace elements, Fe, Mn and In are present isomorphously in sphalerite. The concentrations of Cu, Ag and Sn vary widely, indicating that some of them may occur as micro-inclusions. There is a generally positive correlation between the concentrations of In and Cu, suggesting that In substitutes into the sphalerite through substituting for Zn2+ by Cu+ + In3+ <-> 2Zn(2+). Cd is slightly positively correlated with Fe but negatively correlated with Zn, suggesting that Cd mainly replaces Zn rather than Fe. By enumerating the trace element diagrams of sphalerite, it is found that even the Co/Ag vs. Mn diagram with the highest Silhouette Coefficient still has a large part of overlapping areas, so the binary diagram of sphalerite trace elements cannot be simply used to distinguish the deposit types. Through testing the classical support vector machine (SVM) algorithms of four different kernel functions, the best result is a Gaussian kernel SVM classifier with an accuracy of 91.5%, which can be used for the judgment of the genetic types through sphalerite compositions. Using this method, it is demonstrated that the trace elements of the Haobugao sphalerite is significantly different from that of the MVT, SEDEX, VMS and epithermal Zn-Pb deposits, but similar to that of skarn type deposits. Combined with previous mineralogy and geochronology evidence and the present sphalerite geochemical results, the Haobugao deposit could be classified as a typical skarn type Zn-Pb deposit.
更多
查看译文
关键词
Sphalerite, Traceelements, Machinelearning, Skarndeposit, Haobugao
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要