​​​Research on the origin of gold deposits in the Kibali region of the Democratic Republic of the Congo (DRC) has earned a geology student at Stellenbosch University (SU) the best student presentation award at the biennial meeting of the Society for Geology Applied (SGA) to Mineral Deposits which took place in Zurich, Switzerland, recently.

Yann Mpaka Waku, a PhD student in economic geology and mineral geochemistry in the Department of Earth Sciences at SU, works fulltime as a geologist for Barrick Gold Corporation at Kibali Gold Mine in the north-east of the DRC. At Kibali, regarded as one of the largest gold mines in Africa, his role is to better understand the Archean gold deposit occurring in the Congo Craton shield in order to aid further exploration strategies.

Waku's research, under the supervision of Dr Bjorn von der Heyden, was presented in the session on “Advances in analytical techniques applied for ore deposits research and mineral exploration". In his talk he presents a novel approach to classify pyrites, combining traditional textural analysis, mineral trace elements chemistry, and a new tool for visualising high dimensional data called Uniform Manifold Approximation and Projection (UMAP).  

Waku explains that it is very difficult to decipher the timing and mechanisms of how gold was mineralised within Archean greenstone belts, such as the one found at Barberton in South Africa and at Kibali in the DRC, amongst others. These areas commonly experienced complex and protracted geological histories over millions of years.

That is why geologists have started to rather analyse the mineral chemistry of pyrites, as pyrites are the most abundant mineral in ore deposits hosting the gold. The Kibali gold district is no exception, with an abundant supply of pyrites in the various ore zones.

“We therefore analyse the trace element signatures in pyrites to better understand the conditions that gave rise to ore formation," he explains.

Waku analysed 91 samples from 26 diamond drill holes at eight different sites in the Kibali gold district by using a method called laser ablation inductively coupled plasma mass spectrometry and secondary ion mass spectrometry. In practice, this means that a laser is used to ablate the in-situ minerals, after which their respective chemical and isotopic compositions are measured using a mass spectrometer. The large and rich datasets produced by these analyses were then subjected to further detailed investigation using dimensionality reduction and other “big data" techniques.

Using these big data techniques, he was able to categorise eight different types of pyrites (based on their chemical composition), thereby creating a better understanding of the conditions under which they were originally formed.

According to Waku, this novel approach will prove to become an important method to improve the classification and interpretation of trace elements in pyrite: “I believe our approach has the potential to revolutionise mineral chemistry, especially in the era of increasing big data acquisition," he concluded.

On the photo above, PhD student Yann Mpaka Waku. Photo credit: Wiida Basson​