This study presents an X-ray diffraction analysis and Rietveld structural refinement of selected synthetic fluoroperovskite-type compounds including the Na_1-xK_xMgF₃ solid solutions series in addition to synthetic analogues of cryolite (Na₂NaAlF₆) and simmonsite (Na₂LiAlF₆).

The Na_1-xK_xMgF₃ solid solution series is comprised of three structurally distinct perovskite phases. In order of increasing potassium they are: orthorhombic (Pbnm, a=5.3609(1), b=5.4862 (1), c=7.6661(1), Z=4) in the x=0-0.35 compositional range, tetragonal (P4/mbm, a=5.444(3), c=3.9217(3), Z=2) in the x=0.40~0.46 compositional range and cubic (Pm3m, a=3.9903, Z=1) in the x=0.50-1.0 compositional range. The orthorhombic and tetragonal members are derived from the cubic aristotype by octahedral rotation, a^-a^-c⁺ and a⁰a⁰c⁺ respectively. Introduction of potassium into the Na 4c crystallographic site results in a decrease of octahedral rotation and an overall reduction of structural distortion from f = 4.42 for the neighborite end member to  f = 5.0 for cubic members. This is accompanied by an increase in the pseudocubic cell dimension in addition to a reduction in A-site cationic displacement.  Angular and bond length distortion of the MgF₆ octahedron are at a maximum in the x=0.20 intermediate member of the series and decrease linearly with additional potassium.  These effects are a direct result of cations approaching special positions in the unit cell as potassium is introduced into the neighborite structure.

The  Murun complex contains a number of unusual and mineralogically unique rocks.  Many of them are problematic in terms of their genesis and petrographic interpretation.  These enigmatic rocks include charoite assemblages, as well as unique Ba-Sr-rich carbonatites and alkaline ultramafic dykes referred to as lamproites or lamprophyres.

Specimens of plutonic granitic gneisses have been collected from different locations of the Superior Province.  The study of their Koenigsberger ratio (related to natural remanent magnetization) and the theoretical Koenigsberger ratio (related to maximum artificial remanent magnetization) shows that the theorietical Koenigsberger ratio of the specimens is less dispersed than the Koenigsberger ratio.  There is a power law relation between the remanent magnetization and induced magnetization for the Koenigsberger ratio and the theoretical Koenigsberger ratio and this power law relation is due to variation of amount of ferromagnetic minerals in the specimens.

A structural study of the McKenzie granite (NE of Thunder Bay), the Rice Bay dome and the Sawbill dome (NE of Fort Frances) have also been performed.  The McKenzie granite magnetic fabrics cannot be used as kinematic indicators because of the non-coaxiality of the direction of the magnetic ellipsoid axes and also because a primary fabric has been overprinted by a secondary one.  The Rice Bay dome fabrics are non-coaxial but are clearly related to the regional strain and mineral lineation: they may be related to the diapiric emplacement of the dome.  Sawbill dome fabrics are not related to the diapiric emplacement of the dome but to a later event related to the southern border of the dome adjacent to the Quetico fault.

A copy of the thesis can be downloaded here

The Seagull Intrusion emplaced during formation of the 1.1 Ga Mid-continental Rift centred around Lake Superior was studied in several drill core and surface samples and consists of a lower ultramafic section (~650 m) and an upper mafic section of unknown thickness (<100m). The lower ultramafic section is texturally dominated by cumulate olivine and poikilitic clinopyroxene with lithologies consisting of dunites, peridotites and pyroxenites. The upper mafic section is variable in lithology, with increasing upwards in abundance of cumulate pyroxene and poikilitic feldspar, dominated by olivine gabbro and gabbros.  

Olivine, pyroxene and oxide mineral chemistry from the lower ultramafic section, indicates limited fractionation throughout the section. Forsterite compositions range from Fo75.8 to Fo86.3. Pyroxene also exhibits a limited range of variability with clinopyroxene restricted to the compositions of En42Wo48Fs10 to En54Wo39Fs7. Oxides occur as inclusions in olivine and in pyroxene with compositions reflecting subsolidus re-equilibrium with the host mineral. Olivine compositions from the mafic section exhibit a more differentiated nature and range in composition from Fo57.4 to Fo75.8.

Whole rock geochemistry is consistent with the homogeneous nature of the ultramafic section and more differentiated character of the mafic section. Major elements only exhibit minor variations in abundance through the ultramafics, and exhibit enrichment in Fe and Si only in the mafic section. Trace elements and primitive mantle normalised patterns identified five geochemical units, characterised by total trace element abundances and Eu, Nb and Th anomalies. Sm-Nd and Rb-Sr isotopes indicate the intrusion was variably contaminated by older continental crust. The most extensive contamination is found in the basal section of the intrusion and decreases away from the contact.

Platinum, palladium, nickel and copper mineralization is found at the base of the ultramafic section of the intrusion, formed during initial emplacement and sulphur saturation at that point in time. A second mineralized horizon (RGB Horizon) occurs ~100 to 150m above the basal contact. The RGB Horizon is continuous over 700 m and contains three distinct horizons separated by at least 2 m of unmineralized ultramafic rock. These layers are interpreted to have formed as individual events related to the injection of primitive magma refreshing the chamber.

Geoff is currently working on his PhD at the University of Western Australia in Perth

A copy of this thesis can be downloaded here

Historically wild rice was important economically and spiritually across much of central and eastern Canada, but the antiquity of its use by Native communities is unclear.  Unlike plant macrofossils, which have traditionally been used to identify this plant in prehistoric contexts, silicophytoliths preserve well in archaeological sites and in carbonized food residue encrusted on clay pots.  This proxy, therefore, promises to yield considerable new insight into the antiquity and intensity of wild rice harvesting in this region.

Various phytolith types from various grasses, sedges and aquatic plants were examined.  However, the focus was kept on rondel phytoliths, a form only produced in grasses.  Thirty-eight grass species were examined, including two species of wild rice (Zizania aquatica and Zizania palustris).  A minimum of three hundred rondels from each grass species were counted from various parts of the plant including the inflorescence, the leaf and the stem.  Based on extensive morphological comparisons of phytoliths produced by wild rice (Zizania spp.), and other native Boreal and Prairie grasses and maize (Zea mays), several phytolith morphotypes were identified that are produced only in wild rice (Zizania spp.).  In general, rondels with four spikes, with one and three indentations, are characteristic of Zizania (spp.).  However, differences between the two wild rice species were not established.

Lake sediments from Lulu Lake in the Lake of the Woods area where modern wild rice grows were analysed to determine if the types identified as being diagnostic of wild rice would be present.  As a preliminary analysis, the presence of wild rice (Zizania spp.) can be identified in small quantities in lake sediments.  Therefore, wild rice (Zizania spp.) phytoliths can be a powerful tool in the identification of the plant in Holocene sediments.

Potsherds with encrusted carbonized residues from the Lake of the Woods and surrounding area were also examined for the presence of diagnostic wild rice (Zizania spp.) and maize (Zea mays) phytoliths.  These archaeological samples are attributed to the Laurel (Middle Woodland), Selkirk (Late Woodland), Blackduck (Late Woodland), and Sandy Lake (Late Woodland) cultures.  Based on the use of diagnostic phytolith types for both wild rice (Zizania spp.) and maize (Zea mays), the presence of both these cultigens was identified in the residues of all four cultures mentioned above.

This is the first time maize (Zea mays) and wild rice (Zizania spp.) have been positively identified in prehistoric carbonized food residues from the Boreal Forest.  Based on pottery types, wild rice (Zizania spp.) and maize (Zea mays) were consumed as early as the Middle Woodland period (Laurel phase).  Based on the samples examined, the evidence of maize (Zea mays) phytoliths in the residue is greater than those of wild rice (Zizania spp.).  However, this might reflect sample bias or most likely biases due to processing of the plants before consumption.  Therefore, the absence of a wild rice (Zizania spp.) phytolith signature might not represent that the plant was not consumed, rather that the parts of the plant with the diagnostic phytolith types were removed before consumption.  

Maize (Zea mays) horticulture during the Late Woodland period in the Lake of the Woods and surrounding area does not seem likely because there is no evidence of gardening, or heavy consumption of this plant.  However, the latter might not be necessary for local horticulture.  In contrast, wild rice (Zizania spp.) stands are common in the Lake of the Woods and surrounding area, and therefore local harvesting of this plant is inferred.

A copy of the thesis can be downloaded here

The Hamlin Lake area is located approximately 120 km southwest of the City of Thunder Bay, Ontario within the Shebandowan greenstone belt of the Wawa Subprovince, Superior Province. The purpose of this study was to understand the relationship between an extensive felsic volcanic rock package and the hydrothermal alteration.

Mapping of the area exposed five major lithologies consisting of mafic metavolcanic rocks, intermediate metavolcanic rocks, felsic metavolcanic rocks, felsic intrusive rocks and iron formation. The felsic metavolcanic rocks vary from rhyolites and banded ash to lapilli tuffs and debris flows, whereas the intermediate metavolcanic rocks are made up of dacites and andesites. The only mafic metavolcanic rocks are the debris flows. The felsic intrusive rocks consist of a quartz-eye porphyry and pink breccia. Primary textures, such as fiamme, can still be observed and which are indicative of a subaqueous environment.

Major and trace element geochemical analyses were undertaken on 200 samples of andesitic to rhyolitic volcanic and volcaniclastic rocks. The major elements have become mobile, whereas the trace elements have remained relatively immobile. Hydrothermal alteration affected the primary textures of the Hamlin Lake area so it was necessary to use trace elements to correctly classify the volcanic rocks. Primitive mantle-normalized (PM) and Zr versus Ti plots were used to separate major suites of rocks and also to determine the tectonic setting. The PM plots showed negative anomalies of the Nb and Ti elements, an indication of a supra-subduction zone setting.

A plot of Zr versus Y was used to establish that the rhyolites and andesites are both of a mildly calc-alkaline affinity and the dacites are strongly calc-alkaline. The VMS classifications of Lesher et al. (1986) and Hart et al. (2004) were used to classify the rhyolites and andesites as FII felsic volcanic rocks, whereas the dacites were classified as FI felsic volcanic rocks. FI felsic volcanic rocks are thought to have originated from a deeper source than the FII felsic volcanic rocks suggesting that the Hamlin Lake volcanic rocks were produced in a mature arc-setting.

Mapping and geochemistry was further supported by analysis of Sm-Nd and oxygen isotopes. Nine samples were analyzed for Sm-Nd isotopes and yielded ε_Nd2700 values from -6.59 to +2.62 suggesting some samples had undergone contamination by older material. Several of the samples were close to the value for 2.7 Ga depleted mantle, but one particular sample with a εNd value of -6.59 showed that it had been affected by contamination that could only be explained by a contaminant that was much older than the surrounding 2.7 Ga rocks. The source of the contamination is likely the result of continentally derived sediments being subducted into the mantle wedge, because oceanic arcs are not underlain by older continental crust.

Thirty-seven samples were analyzed for O isotopes to recognize alteration patterns in the Hamlin Lake area. The area mapped at Hamlin Lake was not large enough to clearly show an alteration halo, but it does show that the samples at Hamlin Lake have elevated δ¹⁸O values, the result of interaction with low temperature hydrothermal fluids.

A copy of the thesis can be downloaded here

The Musselwhite gold deposit, 100% owned and operated by Goldcorp Inc., has cumulatively produced in excess of 2 million ounces since opening in 1997 and has a projected mine life through 2013. It is situated in the North Caribou Lake metavolcanic/metasedimentary belt in the central northwestern portion of Superior Province, approximately 430 km northwest of Thunder Bay, Ontario, Canada. The belt occurs along the contact between the North Caribou Terrane and the Island Lake Domain, with a large, crustal-scale deformation zone forming its eastern margin. The Musselwhite gold deposit is hosted by amphibolite facies metamorphic rocks dominated by banded iron formation (BIF). This study primarily focuses on the Northern Iron Formation (NIF) metasedimentary assemblage, host to the majority of gold mineralization at Musselwhite.

Stratigraphic and geochemical analyses suggest that the lithologies of the NIF assemblage were deposited on Mesoarcheaen mafic to ultramafic volcanic rocks forming the ocean floor. The NIF assemblage and another iron formation lower in the stratigraphy, the Southern Iron Formation (SIF), record hydrothermal regimes associated with, and interrupted by, eruptive volcanic activity. The stratigraphically lowest lithologies in the NIF assemblage, meta-argillite, quartz-grunerite BIF, and magnetite-dominant BIF, were deposited in deep, calm water, in association with venting hydrothermal fluids. These ancient chemical sediments are analogous in geochemistry to modern day deposits in places such as the Red Sea and the East Pacific Rise. Differing Eu contents between chert and magnetite layers indicate that rhythmically changing temperature variations drove the hydrothermal system, imparting the banded nature. The chemically pure chert and magnetite layers of the lower portion of the NIF assemblage contrast with silicate-dominant banded iron formation; the silicate-dominant BIF increases in frequency up stratigraphically. It represents a decreasing hydrothermal system and/or an increase in the rate of clastic sedimentation. Hornblende-garnet and biotite-garnet schists were formed by metamorphism of mudstones composed of eroded material. The sediment that formed the hornblende-garnet schist is the same sediment that comprises the siliciclastic component of the silicate-dominant NIF. Similarly the biotite-garnet schist represents a mudstone, but unlike the hornblende-garnet schist, it is primarily derived from intermediate to felsic volcanic rocks. Lastly the garnet-bearing quartzite represents metamorphosed sandstone eroded from the same intermediate to felsic igneous source rocks as the biotite-garnet schist. Just as there is an overall increase in clastic content upwards through the approximately 30 m thick sedimentary succession, there is also a change from more mafic-sourced debris to a more intermediate/felsic source.

The majority of samples collected from Musselwhite did not experience significant remobilization of typically immobile elements. This is indicated by the relatively linear geochemical ratios between the immobile elements (Al₂O₃, TiO₂, Zr, U, Th, etc.). Even elements that are commonly more mobile (K₂O, Na₂O, etc.) appear to have remained relatively immobile at Musselwhite. The only samples that show significant geochemical change are from shear zones. The gold mineralization is primarily associated with shear zones within the siliciclastic-rich, upper NIF assemblage, where pyrrhotite (possibly originally pyrite) replaced iron oxides and iron silicates. This indicates that the control on areas of gold mineralization was a combination of: 1) the presence of structural zones allowing gold-bearing fluids to move through the NIF, which could act as a geochemical trap for gold; and 2) structural conditions in the siliciclastic-rich NIF that favoured hydrothermal fluid involvement with this unit.

Musselwhite Mine is located on the south shore of Opapimiskan Lake, approximately 480 km north of Thunder Bay. Hosted in the  ~2.8Ga North Caribou Lake greenstone belt of the North Caribou Terrane, Superior Province, Musselwhite is currently classified as a shear hosted orogenic gold deposit. The deposit is hosted in an iron formation within a volcanic pile of intermediate to felsic metavolcanic rocks, metabasalts and komatiitic metabasalts and has been metamorphosed to amphibolite grade.

Twenty biotite and 30 quartz samples have been analyzed for δ¹⁵N, δ¹⁸O and δD from Musselwhite Mine as well as 12 biotite samples from the granitoid rocks surrounding the North Caribou Lake Greenstone Belt. Nitrogen isotopes in biotite from Musselwhite Mine are characterized by a δ¹⁵N range from -1.3 to 11.1 per mil. Oxygen and hydrogen isotopes of biotite samples from the mine range from +7.1 to +10.1 per mil for δ¹⁸O and -55 to -100 per mil for δD.

Oxygen isotope signatures from quartz samples from Musselwhite range from +12.4 to +17.1 per mil. Values for the silicate facies iron formation and mineralized zones are consistent with previous work by Otto (2002) and indicate fluid compositions that fall within both the magmatic and metamorphic range.

Biotite samples from granites and metasedimentary rocks adjacent to the deposit have a δ¹⁵N range of -6.9 to +6.1 per mil. Oxygen and hydrogen isotopic ranges for the granitoid plutonic rocks are +2.0 to +4.0 per mil and -59 to -80 per mil respectively; values are typical of felsic plutonic rocks.

The δ¹⁵N, δ¹⁸O and δD stable isotopic data generated for Musselwhite Mine suggest that magmatic fluids played a role in the formation of the deposit.  

Carissa is working on her PhD at the University of Western Australia in Perth

A copy of the thesis can be downloaded here

Diamond bearing Neoarchean metaconglomerates are present in the Michipicoten greenstone belt, Wawa-Abitibi Subprovince, near Wawa, Ontario. They form a portion of the Dore Metasedimentary rocks in the Arliss Lake subbasin, and unconformably overlie a succession of mafic metabasalts. The conglomerates are transitional into argillite and are also overlain by argillite, which in turn is conformably overlain by metabasalts. The conglomeratic succession has a maximum thickness of 454 meters and is confined to what appears to be a deformed paleovalley at the base of the sedimentary succession.

Three lithofacies associations were recognized, and interpreted to indicate that the Leadbetter Conglomerate represents a high-energy, alluvial fan-delta succession. Lithofacies Association One is dominated by weakly sheared and highly viscous debris-flows and represents the upper reaches of the alluvial fan. Lithofacies Association Two is dominated by superimposed longitudinal bars that represent the transformation of the depositional environment to a mid-fan and lower-fan proximal braided system. Lithofacies Association Three is dominated by distributary channel mouth-bars, graded turbiditic density deposits and argillite facies, which represent a transgression and transition to a delta-front and subsequently pro-delta environment.

Whole rock geochemical studies of the Leadbetter Conglomerate indicated that the elements Co, Ni, Sc, TiO₂, V, Al₂O₃, Y, Nb, Zr, Hf, Ta, Th and U are chemically immobile. The immobile elements were used to investigate possible source rock compositions. A mixture of source rocks (mafic, intermediate and felsic igneous rocks, lamprophyre, and possibly kimberlitic occurrences) contributed sediment to the depositional system of the Leadbetter Conglomerate. Of these mafic volcanic rocks and lamprophyres provided the majority of the sediment. Ratio plots used to investigate whether placer accumulation of the heavy minerals chromite and zircon were present indicated that significant placer accumulation had not occurred.

Possible source rocks for the diamonds include either diamond-bearing, ultramafic, lamprophyre dikes and breccias, which have been observed and described near the Leadbetter Conglomerate, or a more commonly diamond associated rock type, such as kimberlite, which have yet to be observed in the area. The presence of kimberlites could be investigated by a study of mineral chemistry, but that was beyond the scope of this thesis.

Corey is currently working as a geologist for North American Palladium, in December 2010 he will start work with Exxon in Houston.

For more details about this thesis contact Dr. Philip Fralick

A copy of the thesis can be downloaded here

Petrography and mineralogy of the C29/30 Candle Lake kimberlite (Saskatchewan, Canada) was studied to understand the characteristics of the kimberlite. Using standard optical microscopy three units where distinguished; pyroclastic kimberlite; resedimented volcaniclastic kimberlite and a crystal tuff kimberlite unit. Each kimberlite unit has been classified based on textural relationships between magmaclasts and the interstitial matrix, as well as phenocrysts, macrocrysts and minor mineral phases within the magmaclasts. Pyroclastic kimberlites are characterized by carbonate or serpentine interclast matrix supporting amoeboid magmaclasts with protruding macrocrysts and phenocrysts of pseudomorphed olivines. Resedimented units have a serpentine-carbonated interclast matrix that is poorly defined against rounded and fractured magmaclasts. The crystal tuff unit is distinguished by fine-grained (<700 μm) clast supported pyrocrysts of pseudomorphed olivine as well as magmaclasts with thin selvages (<100 μm). The interclast matrix is a serpentine-carbonate mixture.

The mineralogy of the C29/30 is typical of an archetype kimberlite. The minerals analyzed include spinels, serpentine, olivine, carbonates, phlogopite, perovskite, apatite, garnets, ilmenite and magnetite. Three types of spinels where identified; type [1] spinels occur enclosed within macrocrysts and phenocrysts with a TIMAC composition; type [2] spinels nucleate along the edges of the macrocrysts and phenocrysts and have a TIMAC core and QUM rim composition; type [3] spinels occur as isolated grains within the matrix and are dominantly QUM in composition but can have a TIMAC core. Atoll spinels are observed in type [2] and type [3] spinels with magnetite rims. The compositions of the spinels are identical to trend T1 spinels from Wesselton. However; they are unlike spinels from Smeaton 169 kimberlite which is also from the Fort à la Corne field which indicates kimberlite within the same kimberlite field may not have the same source.

The eruption was shallow and excavated Mannville sandstones, Paleozoic limestone and any Colorado mudstone if present. The eruption style was probably dry-phreatomagmatic. This is supported by the shallow bowl shape of the kimberlite body with a depth to width ratio of 1:5.

The C29/30 kimberlite is similar to other Fort à la Corne kimberlites by textural and mineralogical comparison.