川滇黔铅锌(锗)成矿区区域地球化学测量在找锗预测中的作用

锗作为一种稀散元素,是现代信息产业最重要的金属之一,然而大多数锗资源是在综合评价煤矿床、铅锌矿床以及铁矿床的过程中发现的,缺乏专门的找锗方法。为了进一步加强锗矿找矿方法的研究,本文以川滇黔接壤处作为研究区,通过采用探索性分析方法(EDA)和基于分形理论的浓度-面积(C-A)方法以及证据权模型,分析研究区内低密度水系沉积物测量数据Zn、Ge元素的数字特征和分布规律。结果表明:(1) Ge元素分布由单个总体构成,基本符合正态分布,数值变化小(变异系数=0. 13),基本围绕地壳丰度值(1. 6×10~(-6))波动,且矿床的分布与地球化学异常之间没有明显的相关关系,因此采用以Ge找Ge的思路是行不通的,而这可能是由于Ge在表生氧化状态下的多亲和性和高度分散性造成的。(2) Zn元素分布明显不符合正态分布,且有多个峰值,数据离散程度大(变异系数=1. 14),可能至少由两个以上总体构成,且在空间上矿床的分布与地球化学异常之间有良好的相关关系,大多数矿床位于相对高值区。另外鉴于在原生状态下,Ge元素往往以类质同象的形式赋存在闪锌矿中,因此在区域上采用以Zn找Ge的思路是可行的。(3)采用累计概率法、直方图法、箱式图法、原生晕法以及基于多重分形的C-A方法,确定的异常下限分别为146. 5×10~(-6)、392. 7×10~(-6)、153. 9×10~(-6)、87. 5×10~(-6)和124×10~(-6)。通过采用证据权方法所提供的空间相关性统计量(t)进行检验,发现异常下限应当落在120×10~(-6)～130×10~(-6)的区间内。因此基于分形理论的C-A方法是最合理的,这可能是其不仅考虑了频数特征,还考虑了空间几何特征的原因。(4)在圈定地球化学块体的基础上,结合热液矿床受构造控制,且往往具有"丛聚效应"以及"鹤立鸡群"的特点,圈定黑区-赤普、大湾子-大桥边、茂租-乐红、毛坪、天宝山-小石房、猫猫厂-白蜡厂、大梁子、会泽、青山-杉树林、猴子厂-顶头山以及富乐作为寻找锗矿的潜力区,建议进一步加强这11个矿田级远景区锗资源的找矿工作。

Role of regional geochemical survey for Ge mineral prediction in Chuan-Dian-Qian Pb-Zn(Ge) metallogenic region

As a dispersed element, Ge is one of the most important metals in the modern information industry. Most of the Ge mineral resources are discovered in the process of comprehensive evaluation of coal, lead-zinc and iron ore deposits, and there is no specific method for predicting Ge ore deposit. In this paper, on the basis of summarizing metallogenic geological conditions and geological characteristics of deposits, we use exploratory data analysis (EDA) method, multifractal concentration-area (C-A) model and weights of evidence model to analyze the characteristics of the low-density stream-sediment data and its distribution of Zn and Ge elements in the junction of Sichuan, Yunnan and Guizhou provinces, aiming to strengthen the research on the Ge ore deposit prospecting method. The following four results are obtained:(1) The Ge element distribution consists of a single population, which basically conforms to the normal distribution, with small numerical variety (coefficient of variation=0.13), and fluctuates around the abundance of the crust (1.6×10^-6). There is no significant correlation between the distribution of deposits and the geochemical anomalies, such as Daliangzi, Huize and Lehong deposits distributed in the area with relatively high Ge concentration, Heiqu-Xuequ located in the area with relatively low Ge concentration, and Tianbaoshan, Xiaoshifang, Maozu, Maoping, Qingshan, Shanshulin and Fule deposits in medium Ge concentration, suggesting that regional geochemical characteristic of Ge element is not a good indicator for Ge deposit. This may be due to the multi-affinity and high dispersion of Ge in the epigenetic oxidation state. Therefore, the idea of using Ge elemental distribution feature to find Ge ore deposit is not feasible. (2) The distribution of Zn element with multiple peaks and high variation is obviously not in accordance with the normal distribution (coefficient of variation=1.14), and the mean value(119.9×10^-6)is obviously greater than the abundance of the crust(79×10^-6), revealing that the Zn distribution may be composed of more than one population. Most of the known Zn ore deposits, including Heiqu-Xuequ, Chipu, Maozu, Lehong, Tianbaoshan, Xiaoshifang, Daliangzi, Huize, Maoping, Shanshulin, Qingshan and Fule, are located in the area with relatively high Zn concentration, whereas, very few deposits are located in the area with relatively low Zn concentration, suggesting a good correlation between the distribution of deposits and the Zn geochemical anomalies. Furthermore, in view of the fact that Ge element exists mainly in the form of isomorphism in sphalerite in the primary ores, it is feasible to use Zn elemental distribution feature to find Ge mineral resources in the region. (3) The lower limit of abnormality extracted from the cumulative probability method, histogram method, boxplot method, primary halos method and C-A method are 146.5×10^-6, 392.7×10^-6, 153.9×10^-6, 87.5×10^-6, 124×10^-6, respectively. The result from the spatial correlation statistic (t) provided by the weights of evidence model shows that the lower limit of abnormality should be falling within the range of 120×10^-6~130×10^-6. It is indicated that the C-A method is the most reasonable method, which may be because it considers not only the frequency characteristics, but also the spatial geometric features. (4) On the basis of delineating geochemical blocks, combined with that hydrothermal deposits are structurally controlled, mainly hosted in Sinian to Lower Permian carbonate rocks, and often have the characteristics of "Clustering effect" and "Stand out" in conjunction with Pb and Ag elemental distribution, we delineate Heiqu-Chipu, Dawanzi-Daqiaobian, Maozu-Lehong, Maoping, Tianbaoshan-Xiaoshifang, Maomaochang-Bailachang, Daliangzi, Huize, Qingshan-Shanshulin, Houzichang-Dingtoushan and Fule ore-field-level tracts as potential areas for the prospection of Ge ore deposits. It is recommended to further strengthen the prospecting work in these eleven prospects.