Cobalt enrichment in Paleoproterozoic African and Brazilian manganese deposits

• Geochemical modeling using PCA to investigate cobalt enrichment in Mn deposits. • Paleoproterozoic Mn deposits have average grades reaching up to 1200 ppm. • High cobalt enrichment factors in Mn-carbonate rocks hosted in Mn-siliciclastic rocks. • Cobalt as a by-product of Paleoproterozoic Mn deposits. This study highlights a new by-product source for cobalt by recycling Paleoproterozoic Mn deposits. We present a geochemical modeling approach utilizing Principal Component Analysis (PCA) for available geochemical data of Paleoproterozoic manganese deposits found in Africa and Brazil, which exhibit anomalous cobalt contents (up to 1200 ppm) along with other metals such as copper, nickel, and vanadium. The PCA results for the correlation coefficient matrix of the Enrichment Factor (EF) values of major and trace elements from samples of eight Mn deposits found in Africa and Brazil (Kisenge-Kamata, Moanda, Nsuta in Africa, and Azul, Buritirama, Lagoa do Riacho, Morro da Mina, and Serra do Navio in Brazil) yielded a cumulative variance of 53.3 % for PC1 (34 %) and PC2 (19.3 %). In PC1, the highest positive loadings correspond to the variables Mn EF , Ni EF , and Co EF , while the highest negative loadings correspond to the variables Si EF , Fe EF , K EF , Ti EF , Cr EF , and Zr EF . PC2 exhibits the highest positive loadings for the variables Ca EF , Mg EF , and P EF , while the highest negative loadings are for Cu EF and V EF . The biplot diagram representation showed that clusters of vectors Mn EF , Ni EF , Co EF , V EF , and Cu EF influence samples of Mn-carbonate rock, Mn-carbonate–silicate rock, Mn-silicate rock, and Mn-carbonate-siliciclastic rock, all with high Co EF values (up to 414). The cluster of vectors Ca EF , Mg EF , and P EF significantly influence carbonate rock and dolomite marble, which have low Co EF values (close to 0). The cluster of vectors Si EF , Fe EF , K EF , Ti EF , Cr EF , and Zr EF strongly influences siliciclastic rock, which exhibits low Co EF values. On the other hand, the cluster of vectors Cu EF and V EF influences oxidized Mn ore, which exhibits Co EF values of up to 108. The results reveal a dichotomy regarding the origin of cobalt and other metal enrichments in these deposits linked to the Mn redox cycle. This process involves the formation of Mn-oxyhydroxides with the adsorption of Co and other metals under oxic conditions, followed by the burial of these Mn oxides in an anoxic diagenetic environment, where microbial sulfate reduction leads to the nucleation of Mn-carbonates and the formation of metal-rich sulfides (Fe, Co, Ni, V). Additionally, detrital input and sulfide phases (e.g., framboidal pyrite) for the formation of Mn-rich siliciclastic rocks associated with Mn-carbonate rocks are evidenced by proxies Si EF , Fe EF , K EF , Ti EF , Cr EF , and Zr EF . This new exploration approach, supported by geochemical modeling through PCA, enhances our understanding of the genesis of these Paleoproterozoic manganese deposits and highlights a new route for cobalt exploration. In the increasing global demand for cobalt, particularly in applications involving electric vehicle batteries and energy storage, exploring these deposits emerges as an alternative source to produce these critical metals