金矿化探(三)——金的颗粒分布与取子样误差关系的研究

GEOCHEMICAL EXPLORATION OF GOLD(Ⅲ)──Distribution of Gold Particles and Subsampling Errorsin Geochemical Samples

当前金矿化探遇到了一个特殊的难题,即地球化学样品中颗粒金的存在造成的取子样代表性问题。这一问题不仅影响到金分析的重视性,而且还影响到对金异常的解释与评价。因此对这一问题的研究和解决不仅具有重要的理论意义,而且还具有重大的实践意义。这篇文章就是作者为了探索解决这一问题长期努力的一部分。本文从理论上和实验上研究了金颗粒分布与取子样误差的关系。发现金分析误差(主要由取子样引起的误差)在低含量和高含量的样品中小,而在中等含量的作品中分析误差大。这是由于(１)低含量(＜10ppb)的样品中金的颗粒非常小,都是显微和越显微金,分布就均匀,故取子样误差小,分析的重现性好。(2)中等含量的作品,不仅有小颗粒的金,而且还有大颗粒金的存在,但由于金的含量不是很高,所以大颗粒金数目并不多,分布也就极不均匀,取子样代表性就差,分析误差也就大。(3)高含量的作品,由于金含量高,大颗粒金相对数目也较多,金颗粒在样品中分布就必较均匀,取子样误差小。通过几个实例的研究证明尽管单个样品金分析的重现性可能很差,但异常的整体可比性还是很好的。

Abstract: Statistical study of the triplicate analytical data shows a general tendency that subsampling error is comparatively low for samples with Au value 1-10 ppb. The error increases in the medium Au value range. For higher Au value range, the subsampling error decreases again. This can be attributed to the follo-wing reasons; ( 1 ) in the samples with low Au value, the gold particles are very small; ( 2 ) in the samples with medium Au value, the subsampling error is dependent on gold particle size: if gold particles are in ultrafine size, the subsampling errors will be insignificant, but if there are unevenly distributed macroscopic particles, the subsampling errors will be great; and(3) in the samples with higher Au value, the macroscopic particles are abundant, so the subsampling errois are low. Although there exist great variations among the triplicate analytical data of individual samples, the trend, shape, intensity and dimensions of an anomaly re-main consistent.