DIAMOND SPECIAL — Kimberlite indicator minerals in glaciated

The study of kimberlite indicator minerals in weathered kimberlite and unweathered glacial sediments at the C14 and Diamond Lake pipes (Kirkland Lake, Ont.) has revealed several relationships having important implications for diamond exploration. Some of these relationships are specific to the Kirkland Lake kimberlite field whereas others may be applicable to the search for kimberlite in any glaciated terrain.

The importance of WSW relative to SSE ice flow in the Kirkland Lake district, despite the overwhelming dominance of SSE ice flow indicators on rock outcrops, is reinforced by the WSW orientation of the C14 indicator mineral dispersal train.

Evidence for sustained WSW ice flow from the New Quebec ice dome during deglaciation of the Abitibi region has been growing in recent years, most notably with the identification in Quebec of large buried roches moutonnees by Jean Veillette of the Geological Survey of Canada (Veillette, 1989) and three major base metal and gold dispersal trains by Overburden Drilling Management (unpublished contract surveys).

Final SSE ice flow from the Hudson Bay dome was less influential in terms of mineral dispersal, especially in bedrock depressions over kimberlite pipes and shear zones, apparently because deposition of Matheson Till had already begun before the shift from WSW to SSE ice flow. SSE dispersal was, however, important where the bedrock surface is elevated and the till is thin (for example, McClenaghan, 1990) and also in glaciofluvial sediments because the principal eskers trend SSE. Since most esker sediments were subjected to ice transport before glaciofluvial transport, dog-leg WSW/ESE dispersal patterns may also exist.

Recognition of the importance of WSW ice flow may help resolve certain exploration problems in the Kirkland Lake district such as the concentration of indicator minerals west rather than south of the B30 kimberlite pipe (Averill and Fortescue, 1983) and the lack of a gold dispersal train in till tested by reverse circulation drilling south of the McBean mine (Nichol and MacIntosh, 1986).

Another relationship bearing on the search for kimberlite in the Kirkland Lake area is the abundance of green crustal andradite-uvarovite and low-chrome diopside relative to unequivocal kimberlitic chrome diopside. The andradite-uvarovite grains often have as much emerald green pigment as kimberlitic chrome diopside but can be recognized by their cloudy appearance, microcolloform texture and presence of platy serpentine intergrowths. Crustal low-chrome diopside is so similar to the palest varieties of kimberlitic chrome diopside that it is recognized in the present study only because (a) it is of more distal provenance and the grains show considerable wear, and (b) it is disproportionately abundant in overburden samples that contain few kimberlitic indicator minerals. Therefore it would be prudent for explorationists to regard all low-chrome diopside in overburden in the Kirkland Lake area as crustal unless it shows a sympathetic relationship to other kimberlitic indicator minerals.

An important finding bearing on the global search for kimberlite is that the size and character of indicator mineral grains change very little during glacial and glaciofluvial transport. The minor changes observed are comparable to those reported during fluvial transport in Siberia where the climate is similar (Afanasev et al., 1984). However the good preservation of indicator minerals in both areas may be due less to climate than to the fact that mineral transport was rapid both during glaciation at Kirkland Lake and in the rivers of Siberia.

In contrast, pronounced indicator mineral modification and degradation have occurred over short transport distances in Western Australia (Mosig, 1980; Atkinson, 1989). This may be due more to prolonged transport by soil creep and sheet wash than to the warmer, drier Australian climate. Prior to glaciation, garnet xenocrysts in the upper parts of the C14 and Diamond Lake pipes had been fragmented by weathering under warm climatic conditions, producing many smaller grains as in Australia. Each kilogram of kimberlite supplied up to 5000 chrome pyrope fragments to the ice. During transport, these fragments were subject only to dilution, not degradation as in Australia. Consequently, a 10-kg till or gravel sample is adequate for exploration purposes versus a 20 kg sample in Australia (Atkinson, 1989). Less than 1% of the chrome pyrope fragments are larger than 1 mm and additional savings (up to 30% for coarse gravel samples) can be achieved by processing only the <1 mm fraction.

The small chrome pyrope fragments, when presented to the glacial ice, were angular with freshly broken conchoidal surfaces more abundant than kelyphite-coated resorbed surfaces. Most of the soft kelyphite was removed in the first few hundred metres of transport but subkelyphite matte and orange peel textures are perfectly preserved 2.3 km down-ice from C14 and almost as well preserved on garnet grains that may have travelled more than 10 km in the Misema River Esker.

The color of many chrome pyrope grains seems to have changed from red-purple to blue-purple during transport. It is not clear whether this apparent color change is due to strain, glacial cleansing or some other process. One implication of the color change is that color guidelines used to separate G10 from G9 garnet in kimberlite may not be applicable to chrome pyrope grains recovered from glacial sediments.

Chrome diopside and chromite xenocrysts in the C14 and Diamond Lake kimberlites were little affected by the preglacial weathering but were by nature smaller than the garnet xenocrysts and therefore were equally stable during glacial transport. Chrome diopside xenocrysts dispersed from the C14 pipe remained whole during 2.3 km of ice transport and their surface resorption textures remained intact.

Due to the relatively small size of the xenocrysts, however, their liberation from the kimberlite was retarded during weathering and ice transport and their abundance relative to chrome pyrope increases with distance from source. This contrasts with the Australian situation where chrome diopside is plentiful near source but completely destroyed 3 to 5 km from source. Only picroilmenite, which occurred as larger, whole xenocrysts in the weathered kimberlite, appears to have been subject to fragmentation during transport and this fragmentation may be restricted to the structurally weak, polygranular variety of picroilmenite that occurs at Diamond Lake. Leucoxene coatings were removed from the ilmenite xenocrysts by a few hundred metres of glaciofluvial transport but resorption textures are well preserved. The magnetic to strongly paramagnetic character of the picroilmenite in this pipe argues against the use of a weakly paramagnetic picking fraction for picroilmenite as advocated by McCallum and Vos (1993). Kimberlite pipes such as Diamond Lake in which picroilmenite is both magnetic and the most abundant indicator mineral present could be overlooked if the magnetic and paramagnetic minerals are not examined, at least in the coarser (>0.5 mm) fractions.

Although the present study has shed needed light on the distribution and character of kimberlite indicator minerals in glaciated terrains, its conclusions are based on a small sample population. Data currently being obtained from the remaining samples will be used to test these conclusions but additional samples from other glaciated kimberlite fields are needed to broaden the data base.

Particular emphasis should be placed on comparing indicator minerals from weathered surface till and gravel samples to those from unweathered samples collected at greater depth. With future studies of this nature in mind, special care has been taken to preserve the C14 and Diamond Lake indicator mineral grains.

— From the Geological Survey of Canada, Open File 2819, S.A. Averill and M.B. McClenaghan.