Fingerprinting refers to the use of conventionally collected geochemistry data to enhance our ability to explore. We want to know the fingerprint of gold mineralization, so we can use other, better behaved elements as proxies for gold. We also want to understand the genesis of mineralization so we know how to vector in the deposit.
Understanding the fluid source can become very important for targeting. Most Archean gold deposits are hosted in shear zones and are what is generally called orogenic. The answer to “where to go next” is along the host shear zone (see more information in this article). But sometimes, things are less clear. Vectoring can get more complicated in deposits with a stronger intrusive fluid influence – let’s call them intrusion-related. In early days, we often have to rely on very limited data to decide what to do.
To start, we need to know what elements associate with Au and what that means. We can start by looking at the periodic table. Based on their proximity to Au, elements can be expected to behave similarly to it.
OK, that’s something. The reality is much more complicated, however. Partially that is because of electrons, orbits, and other difficult chemistry things. But it’s also because of geology (phew).
How geology fits in
When it comes to understanding elements associated with Au, we need to think about mineralogy, transport processes, fluid sources, host rocks, pathways, temperature of fluids… There are many variables that are going to generate further associations. But let’s keep it simple for now and just look at the temperature and salinity of fluids. The figure below shows temperature and salinity and identifies major deposit types as fields within that space (Wilkinson, 2001).
We can map the diagram to two major groupings for our purposes. One is the orogenic or lode Au deposits, which have fluid inclusions with homogenization temperatures around 200-400 degrees C and <10% salinity. They also, not shown above, record much higher pressure conditions. Porphyry deposits inclusions, which we can read more broadly to mean intrusion-related, have fluid inclusions with hotter homogenization temperatures of 200-800 degrees C and >10% salinity. These different conditions, before even the trace element chemistry of the fluid source region and the petrology of the rocks are considered, can be expected to create different signatures of orogenic and intrusion-related gold systems in Archean rocks. What are those signatures?
Elemental associations are related to deposit genesis
Orogenic deposits are associated with Au-Ag-As-W-B-Sb-Te-Mo (Dubé and Gosselin, 2007). On the intrusion-related side, porphyry deposits have Cu-Mo-Au-Ag-Zn-Pb associations (Sillitoe, 2010). Reduced intrusion-related deposits have a Au-Bi-Te-W-Mo-As-Pb-Zn-Ag signature (Hart, 2007). In greenstone belts specifically, Robert (2001) described syenite-associated gold deposits in the Abitibi and proposed that a signature of Au-Cu-As-Te+/-Pb, Mo, Sb is associated with these deposits. In a study in the Chibougamau area, Mathieu (2019) suggested that Bi and Te were the most effective fingerprints of magmatic-hydrothermal pyrite, but also that every project area could be different.
One of the more well-studied examples in Canada is Côté Gold. Numerous lines of evidence (structure, alteration, lithology, chemistry) support its genesis as a porphyry (Katz, 2016). The association there is Au-Cu-Mo-Ag-Te-Bi-Se-Zn-Pb, although it is zoned within the deposit.
There is obviously a lot of overlap, and that’s fine. What is important is the balance of the evidence. Based on the above, an association with B is suggestive of an orogenic deposits. On the other hand, Bi-Cu-Pb-Zn could be used to interpret a more intrusion-related deposits Now I would add some other important clues to intrusion-related mineralization based on my experience: Sn, Mo, Tl, and Sb. We will have a look at some real data next time, and we can talk about those then. The important thing is we expect that we can see some differences in trace element chemistry.
Fingerprinting with principal components analysis
Principal component analysis or PCA is a dimensionality reduction technique that is commonly used in geology. It is used to remove correlations from a dataset for simulation, but also to simplify data by gathering together properties that behave similarly into principal components. Principal components that are less important can be removed, so we only need to work with a few for modelling while still capturing most of the behaviour of the data.
The way PCA works is it makes a set of vectors that are functions of the original input columns. The relationship between the input column and the output vector or principal component is called the loading. High positive loadings mean an element is positively associated with the PC, negative ones mean it is negatively associated.
It is surprising to geologists who have encountered PCA to hear that in most circles the loadings are not of particular interest. We may be the only field that interprets the loadings to try to learn things about the things the data describe. We can talk more about PCA another day, but for now suffice it to say it can be used to summarize the notion of “what is associated with gold” quite well.
If we take the elements that have higher weights in the principal component most strongly associated with Au, we have a pretty good first pass of the fingerprint of Au in the system.
Using fingerprints
We can look for anomalies in the fingerprint elements instead of just Au because they can be better behaved than Au. Consider a situation where nuggety gold resulted in a low assay when the other elements that are less nuggety but co-occur are high – a good signature of a near-miss hole. It’s easier to interpolate just about anything other than gold, so fingerprint elements make good candidates for visualizing anomalies on grids. Better yet, interpolate the principal component associated with Au!
The other way we can use fingerprints is conceptually. If we see associations in the geochemistry that suggest a magmatic component to the mineralization, then the targeting strategy might change. As Robert (2001) points out, syenite-associated gold deposits tend to be clustered around an intrusion as well as occurring along a continuum from the host intrusion to the surrounding rocks.
Its easy to imagine a situation where relatively low grade mineralization in sheared basalts might represent the distal mineralization to a larger, higher grade center related to a minor intrusion. In that case, following along the grade trend or mapped shear zone might not be the best strategy.
Summary
Knowing what elements covary with Au in an Archean deposit is important for targeting and modelling. A few important elements can help discriminate vein vs intrusion deposits in Archean rocks, which is important for targeting effectively. PCA is a good way of identifying fingerprint elements. Grouped elements can be interpolated and used as a proxy for Au as they are often better behaved in a geostatistical sense.
Next time we will go through the geochemical exploratory data analysis process and look at some real data to see how these guidelines stand up to reality.
References
Dubé, B. and Gosselin, P. (2007) Greenstone-hosted quartz carbonate vein deposits. In Goodfellow, W. D. (2007). Mineral deposits of Canada: a synthesis of major deposit-types, district metallogeny, the evolution of geological provinces, and exploration methods. Geological Association of Canada, Mineral Deposits Division, Special Publication, 5.
Hart, C. (2007) Reduced intrusion-related gold systems. In Goodfellow, W. D. (2007). Mineral deposits of Canada: a synthesis of major deposit-types, district metallogeny, the evolution of geological provinces, and exploration methods. Geological Association of Canada, Mineral Deposits Division, Special Publication, 5.
Katz, L. (2016) Geology of the Archean Côté Gold Au(-Cu) Intrusion-Related Deposit, Swayze Greenstone Belt, Ontario. PhD thesis, Laurentian University. https://laurentian.scholaris.ca/items/08551e2b-bae7-49b4-8d6b-caa8a71fd0b2
Mathieu, L. (2019). Detecting magmatic-derived fluids using pyrite chemistry: Example of the Chibougamau area, Abitibi Subprovince, Québec. Ore geology Reviews, 114. https://doi.org/10.1016/j.oregeorev.2019.103127
Robert, F. (2001) Syenite-associated disseminated gold deposits in the Abitibi greenstone belt, Canada. Mineralium Deposita, V.36:503-615. DOI 10.1007/s001260100186.
Sillitoe, R. (2010) Porphyry copper systems. Economic Geology, v.105, pp:3-41. https://doi.org/10.2113/gsecongeo.105.1.3
Wilkinson, J. (2001) Fluid inclusions in hydrothermal ore deposits. Lithos, v.55:229-272. https://doi.org/10.1016/S0024-4937(00)00047-5