Applications of analytical chemistry to geochemistry : thesis presented for the degree of Master of Science in Chemistry, University of Otago

Williams, Xandra Kathleen.

This thesis describes the results of several programmes of geochemical prospecting in New Zealand, during which it was found possible to delimit known deposits and locate new anomalies, and the results of research arising from them. At Copper Queen Mine, Parakao, in Northland, and at the Moke Creek copper lode, near Queenstown in Central Otago, soil prospecting was used to determine the subsurface limits of known copper deposits. The deposit at Parakao appears to be limited, but at Moke Creek the lode formation probably continues for at least twenty chains further than it can be seen in outcrop.
Exploration surveys were made in part of the Shotover River drainage basin and in the Longwood Range in Southland. The source of a cupriferous boulder, found in the Shotover River, may be associated with the soil, and creek water and sediment, anomalies found in Ironstone Creek. The survey in the Longwood Range was more general than that in the/Shotover district. Copper, nickel, and zinc were sought in the water and sediments of all the accessible creeks draining the range. This was followed by soil prospecting in the anomalous areas thus located. Anomalies were found in the soil samples, although it was not possible, from the limited results obtained, to determine the areal extent or possible source of the anomalies.
In several areas studied seasonal effects were evident in the water and soil, and thought to result from changes in rainfall, temperature and organic activity. At Parakao organic acids are thought to affect the soil pH, and this in turn affects the dispersion of copper in the ground water. Organic activity may also affect the metal content of the creek water in the Longwood Range. Since results of water sampling at one locality on different occasions varied, a monthly study was made of the copper and sulphate content and pH of stream water, and the total and cold-extractable copper content of creek sediments. The results show that the total and cold-extractable copper content of the sediments and the copper and sulphate content of the water tend to vary inversely as the pH and creek discharge, although for part of the year the cold-extractable copper content of the sediments varies with the pH.
The copper content of the water and sediments is high in the summer months, then decreases until June when there is a winter maximum while the creek discharge is low. This lasts until September and is followed by a drop in copper content until the temperature rises in November and the creek discharge is again low. In an attempt to overcome the variable metal content of creek water the possibility of using silica as an internal standard was investigated. The results obtained showed that both the silica content and the metal to silica ratio varied significantly, and the method was abandoned.
Much of the soil sampled was over schist and the mode of occurrence of copper in the overlying soil was studied. Of the essential minerals present, namely chlorite, muscovite, albite, and quartz, the chlorite fraction has the greatest concentration of copper initially ( i.e. in the rock) and adsorbes the greatest amount of copper from aqueous solution. The amount of copper associated with organic matter or the iron oxides is negligible.
In an area where soil sampling had proved unsatisfactory, vegetation samples were collected and analysed for nickel. Results from the first group of plants showed the presence of an intense anomaly closely related to a nickeliferous lode. However samples collected from near a second outcrop, with apparently similar mineralogy to the first, all had low nickel contents.
During the exploration work a large number of results were produced; to aid in their interpretation an attempt was made to analyse them statistically. If the cumulative frequency distribution of the analytical results from one area are plotted on probability graph paper a break or point of inflection in the graph indicates the presence of more than one distribution, and is an indication of the boundary between the two groups. Its meaning in terms of the overlap of the background and anomalous distributions can be determined.

1 v. iv. 197 p. ill. 27 cm.

1964Williams