THESIS This is to certify that the thesis entitled 3‘) \ ")?"\ *3 '. -. *7} j“! ‘71} 'rfinvn l (up?) r)! wt“ U 1 .1 .LLLI. "4.4.1 DI J . 1‘§IJJ'U _ IC-‘l .LA‘i.‘L_J -iu - a 7‘1 Mix?! I‘T‘l‘t‘. r ‘rfv—w T, 17 1'1 ' tram A" C: ...'.;11 J.“"JV\J!1'\4TAII 15.11 iROu 110?. «'1 ,L Ib I, lg" hfivrflx'nrnr‘l 1 :Vrv-w mm '1' my m A a} J‘JA L1 “Ll—.0: .LLL A)I.~) RI'J .,, ‘ T‘ IT ‘P‘ --.LJ-1 AAA.“ presented by . "1 V‘ 77‘ ' s T. ‘ ausan o. ,ltJSnln has been accepted towards fulfillment of the requirements for :‘r’- :73- degree in ’anlngy Major professor Date February 9, 1983 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wilI be charged if book is returned after the date stamped beIow. RARE EARTH ELEMENT DISTRIBUTION IN THE NEGAUNEE IRON FORMATION, MARQUETTE DISTRICT, MICHIGAN By Susan E. Tituskin A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geology 1983 C/aérpag ABSTRACT RARE EARTH ELEMENT DISTRIBUTION IN THE NEGAUNEE IRON FORMATION, MARQUETTE DISTRICT, MICHIGAN BY Susan E. Tituskin The Negaunee Iron Formation, at the Empire Mine, has several lithologies (oxide, carbonate, and silicate iron formation) dependent upon diagenesis weathering, and metamorphic history. The mineralogic content of primary unit(s) from which these lithologies were derived is controversial. The REE abundances progressively decrease in the sequence: clastics— oxides-carbonates-chert rich samples. Distribution patterns are variable within groups but show no systematic lateral or vertical trends. This, coupled with anomalous cerium behavior, suggest deposition near a basin margin from a LREE-enriched solution. The effect of alteration on original REE distributions is difficult to assess, but appears to correlate to ore formation through diagenetic diffusion mechanisms whereby the presence of carbonate complexes permit the stabilization of HREE in solution. The REE trends of chemical sediments are grossly similar to clastic interbeds. This suggests that continental errosion was a major contributor to the solution chemistry and a probable iron source. ACKNOWLEDGEMENTS I wish to give my deepest gratitude to my committee, Drs. D. T. Long and F. W. Cambray, for all of their support and encouragement along the way. I would especially like to thank my principle advisor, John T. Wilband, for his original suggestion of the research topic and his assistance and "brain searching" throughout the project. I would also like to thank my friends for both their support and help during my time here, especially Tom, Sandy, Cheryl, Mick, Mary, Bruce, and Kaz. ii TABLE OF CONTENTS LISTOFFIGURES............. ..... v LISTOFTABLES........................vii INTRODUCTION........................l PurposeandScopeofResearch GeOIOgical setting 0 O O I O O O O O O O O O O O O O O O O O Stratigraphy and Geochemistry of the Empire Mine . . . . . . . UNF— PREVIOUSSTUDIES.......................10 Rare Earth Elements in Sediments and Seawater . ..... . . 10 REE in Chemical Sediments . . . . . . . . . . . . . . 13 Clastic Sediments . . . . ..... . . . . . . . . . . 15 Effects of Metamorphism . . . . . . . . . . . . . . . . . l5 SokomanlronFormation 16 Summary of Constraints of REE Studies in Sediments . . . . . 17 LOCATION AND SAMPLE DESCRIPTION . .......... . . l8 PETROLOGY.................... ...... 20 Carbonates ........ 20 Chert................ ..... 22 Magnetite....... ..... . 22 Clastics......... ..... .. . 25 IronSilicates.... ..... ..25 SiamoSlate................. ......28 MAJOR ELEMENT GEOCHEMISTRY . . . . . . . . . . . . . . . 32 TRACE ELEMENT GEOCHEMISTRY . . . . . . . . . . . . . . . 35 Carbonates........................35 Cherts...... ..... . .......... ....43 Magnetite Ores . . . ....... . . ...... . . . . 46 Clasticlnterbeds......................#6 SiamoSIate........................#9 DISCUSSION 0 O O O O I O O O O O O O O O O O O O O O O O O O O 51 CONTROLS OF REE DISTRIBUTION ..... . . . . . . . . . . 514 Mineralogy and Diagenesis . . . . . . . . . . . . . . . . . 5‘4 Metamorphism.......................63 OXidation. O O O O I O O ......... O O O O O O O O 64 Summary........... ..... . ..... ...64 iii TABLE OF CONTENTS (continued) IMPLICATIONS OF REE CONCENTRATIONS ON CHANGING SEA WATER COMPOSITION WITH TIME . . . . . . . . . . . . . . 66 INFERENCES OF TRACE ELEMENT DATA ON THE FORMATION OF THE NIF AND COMPARISON TO OTHER IRON FORMATIONS . . 73 CONCLUSIONS . . . . . . . . . . . . . . ........... 83 SUGGESTED FURTHER RESEARCH . . . . . . . . ....... 86 REFERENCES . . . . . . . . . ......... . . . . . . . 88 APPENDIX A. ...... . . . . . . . . . . ..... . . . . 93 APPENDIX B . . . . . . . . . . . . . . . . . . . . . . . . . . 95 APPENDIX C. . . . . . . ........... . . . . . . . . 98 APPENDIXD ............ . . . ..... . . . . . . 100 iv Figure 1. Figure 2. Figure 3. Figure l4. Figure 5. Figure 6. Figure 7. Figure 8. Figure 9. Figure 10. Figure 11. Figure 12. Figure 13. Figure 14. Figure 15a. Figure 15b. Figure 16. Figure 17. Figure 18. LIST OF FIGURES Generalized map of the Marquette District, Michigan . . Generalized cross-section of the Negaunee Iron Formation, Empire Mine, Marquette District, Michigan . . . . . . . Study diamond drill hole locations, Empire Mine ..... Siderite matrix with minor chert . . . . . . . . . . . Siderite anhedral mosaic. . . . . . . . . ...... Carbonate-rich section with chert laminae ....... Ankerite partially filling fracture in carbonate- chertmatrix.................... Magnetite ore in carbonate-chert matrix . . . . . . . . Magnetite ore replacing siderite . . . . . . . . . . . Detrital quartz lense in carbonate matrix . . . . . . . Detrital quartz with bladed stilpnomelane overgrowths . . Iron-silicates in carbonate-chert rich laminae ...... Typical quartzite sample of the Siamo Slate ...... Typical graywacke sample of the Siamo Slate ...... Mafic fragments in the Siamo Slate . . . ....... Same as Figure 153 under higher magnification . . . . . Chondrite normalized REE abundances of: a) carbonates, b) chert, c) magnetite ore, d) Clastic interbeds, ande)SiamoSlate NASC normalized REE abundances of: a) carbonates, b) chert, c) magnetite ore, d) clastic interbeds, ande)SiamoSlate................. Average REE concentrations normalized to Chondrite values 0 I O I O O O O O O O O O O O O O O O O O O O 19 21 21 23 23 2’4 24 26 26 27 29 29 31 31 441 47 52 Figure 19. Figure 20. Figure 21. Figure 22. Figure 23. Figure 24. Figure 25. Figure 26. LIST OF FIGURES (continued) Average REE concentrations normalized to NASC values . Correlation of REE content to chromium for: a) chert, b) carbonates, and c) magnetite ore . . . . . . . . . . Ankerite-rich samples normalized to average Empire siderite...................... Plot of Yb/La versus Yb/FezO3 for Empire Carbonate-FICh samples 0 o o o o o o o o o o o o o o REE concentrations for Negaunee hard ores normalized to: a)chondriteandb)NASC. . . . . . . . . . . . . Plot of La versus Th for Empire: a) carbonates, b) chert, and c) magnetite ore . . . . . . . . . . . . Comparison of REE concentrations for Sokoman and Empire average ores normalized to NASC . . . . . . . Plot of Sm/Yb versus La/Sm for Empire samples vi 57 6O 62 65 69 75 80 Table I. Table 2. Table 3. Table 1,, Table 5. LIST OF TABLES Microprobe analyses of carbonates ..... . . . . . . Total iron content reported as percent Fe203 . . . . . . Absolute Rare Earth Element concentration (ppm). . . . . Rare Earth Element ratios ............... Lanthanum-Thorium Ratios ............. vii 33 3Q 36 41 68 INTRODUCTION Pugpose and Scope of Research This investigation deals with the distribution of trace elements, specifically the rare earth elements (REE) in the Negaunee Iron Formation (NIF), Marquette District, Michigan. REE's have been extensively used as petrographic indicators in igneous systems, however, several studies have also been carried out on sediments and sedimentary rocks. The behavior of the REE's in sedimentary systems is not as clearly understood, and this includes their precipitation and occurrence in banded iron formations (BIF). Pioneering investigations by Fryer (1977a, b) and Graf (1978) have indicated that REE distribution patterns are useful for placing constraints on the depositional nature of iron formations. It is generally thought that the BIF's formed as chemical precipitates when iron-rich solutions entered the sedimentary environment, therefore, the REE distribution pattern would be effected by: 1) the chemical nature of the sea water, 2) the REE pattern of the iron source solution, 3) the degree of mixing between these solutions, 4) the amount and type of precipitating or detrital phase, 5) the nature of the basin (i.e., shallow versus deep marine), and 6) the mode of precipitation. The NIF is considered typical of Superior-type Proterozoic iron formations. The Superior-type iron formations in the Marquette-Gogebic ranges have a geological setting for which both turbulent and quiescent conditions of deposition exist. The role of volcanism is uncertain, but most Proterozoic iron formations are not immediately associated with volcanic rocks. This is true of the NIF except for the possible association of the Clarksburg volcanics with the post- Negaunee age iron formations in the Michigamme Formation. Samples have been collected from the Empire Mine located within the lowest-grade metamorphic zone of the NIF. Because the REE are more resistant to fractionation during metamorphism than other trace elements (Cullers, et a1., 1974), they may be considered the most chemically representative of the original unaltered sediments. Therefore, the acquired data will be acceptable for discussing the above six parameters. Trace element data has been collected from #5 core samples. Chert, carbonates, clastics, ores, and the underlying Siamo Slate have been analyzed. These data will be used to investigate mineralogic, lithologic, and facies controls of the REE's within the Negaunee. This information will be used to infer the chemical nature of the solution from which the NIF was deposited, possible sources of the iron-rich solution, and the environment of deposition. The REE imprint will be compared to similar Proterozoic BIF's. Geological Setting The NIF, a member of the Marquette Range Supergroup, consists of a series of metasedimentary rocks occupying the Marquette Trough. This structure is a narrow, westerly trending synclinorium which extends approximately 70 km from the city of Marquette to near the village of Three Lakes. The Empire Mine is located in the south-limb sector in the lowest grade metamorphic zone of the Iron Formation. The south limb folds into a smaller syncline called the Republic Trough in western Marquette County. A second belt trends south from the western portion of the Republic Trough. The Marquette Range extends from Marquette County through Baraga, Iron and Dickinson Counties (Figure l; Gair, 1975). The NIF is the upper member of the Menominee Group within the Precambrian X series of metasedimentary rocks. Van Schmus and Woolsey (1975), using whole rock and mineral Rb/Sr studies, determined that the Anna CEO .3me cmeuE .5235 03033.22 of no amE “63:98on 4 oSwE Ex 111111.]Jlu m o xudsoo hzmzwmg 32:55:32 0 889. 2.: 223232.. 530.. 82:55:52 E 2922:9122. 32382 H 989. 3 255283.. 38.: 3.2.w.0....h Qdou-‘oV-c. I ..... u... ......... .................. ............................. ooooooooooooooo ooooooooooooooooooooooooooooooooo sediments of the Menominee Group were deposited between c. 1.9 and 2.0 Ga. This series is underlain by Lower Precambrian mafic, metavolcanic, and granitic rocks. The primary members of the basement complex are the Mona Schist, Compeau Creek Gneiss, and the Palmer Gneiss. Age relationships of the Marquette Range Supergroup are summarized by Gair (1975). The deposition of the NIF is assumed to correlate to that of other iron bearing sequences in the Lake Superior District (James, 1955, 1958; Trendal, 1968). The iron formation consists of three primary facies: carbonate, oxide, and silicate (James, 1954). Maximum thickness reaches approximately 1300 meters. These facies change laterally in thickness and lithology resulting in difficulty in overall stratigraphic correlations. Correlations have been further complicated by post—depositional changes including diagenesis, metamorphism and post- metamorphic oxidation. The Empire ore body falls stratigraphically in the upper portion of the NIF, where the ore consists predominantly of magnetite with minor hematite. Other common minerals are chert, siderite, iron silicates, and martite. At this location, the NIF reaches a thickness of 900 to 1100 meters. The ore occurs in a monoclinal fold dipping 350 to the northwest. The area is cross-cut by dike- filled transverse faults of Precambrian X and Y ages. The Negaunee is overlain unconformably by the Goodrich quartzite (Baraga Group) and underlain conformably by the Siamo Slate. Gair (1975) notes that this lower contact is, in places, gradational through a thickness of as much as 30 meters. The Siamo contains slate, sericitic, chloritic, and feldspathic quartzite, and graywacke as the major rock types. The Siamo contact with the underlying Ajibik quartzite is also conformable. This transition through the NIF appears to represent a change from detrital facies to essentially detritus-free (NIF) facies. Early investigations suggested an east to west transgressive sequence for the formation of middle Precambrian sediments (Tyler and Twenhofel, 1952). More recently, it has been proposed that early deposition occurred within the stable shelf with later strata deposited after mild uplift in a mildly active miogeosynclinal environment (Van Schmus, 1976; Cannon, 1973). The system of troughs appears to be fault controlled based on the lack of evidence for intervening anticlines. Gair (1968) suggests that faulting has influenced sedimentation and was active during deposition. The area was subsequently metamorphosed and deformed during the Penokean Orogeny (Goldich, et a1., 1961) which has been dated to c. 1.9 t .05 Ga. (Van Schmus and Woolsey, 1975). This has resulted in heavy secondary folding to the west (Cannon, et a1., 1975) and faulting to the east (Gair, 1975; Cannon, et a1., 1975). Metamorphic grade increases from the chlorite facies in the east, near the City of Negaunee to the sillmanite facies in the west near Republic (James, 1955). Haase (1979) suggests that sillmanite grade was barely reached, since only andalusite has been found in situ. A later swarm of east-west trending Keweenawan dikes (c. 1 Ga) were emplaced during a short-lived rifting interval and intrude the Marquette Supergroup. They are considered to have been emplaced during the early opening phase of rifting (Pesonen and Halls, 1979) based on their remnant polarity. Stratigraphy and Geochemistry of the Empire Mine Numerous investigations have been carried out in the Marquette District since the early 1900‘s because of its economic importance. Early studies by Van Hise and Bayley (1897) and Van Hise and Leith (1911) characterized the sedimentation, stratigraphy, structure, and metamorphism of the Marquette Trough. Subsequent investigations have refined the petrology, chemistry, age, and stratigraphic nomenclature of the district (James, 1955, 1958; Anderson, 1968; Cannon and Gair, 1970; Gair, 1975; Cannon, 1973; Cannon, et a1., 1975; Van Schmus and Woolsey, 1975; Cannon, 1976; and Haase, 1979). The most detailed geochemical studies at the Empire Mine have been conducted by Cleveland Cliffs Iron Company (CCIC) (Han, 1962, in Gair, 1975) which characterize mineralogy and lithology of the primary facies and diagenetic replacement. Haase (1979) has extended James (1955) metamorphic investigation by utilizing mineral assemblages, compositions, and reaction mechanisms to discern metamorphic grade and processes. Initial carbon and oxygen isotope data has also been presented by Haase and Rye (1980). Han (in Gair, 1975) categorized the three major lithologic groups at the Empire as follows: Magnetite-carbonate-chert: alternating layers of magnetite chert plus carbonate, and chert plus some carbonate and/or magnetite. At least some of the magnetite is believed to form from the replacement of carbonate. Magnetite-carbonate-silicate-chert: laminae of concentrated magnetite and alternating layers of silicate-chert, carbonate-chert plus some silicate, and carbonate-silicate-chert. Carbonate-chert: laminae of carbonate-chert and chert plus some subordinate carbonate, locally containing appreciable magnetite that apparently formed by the replacement of carbonate. A generalized stratigraphic section of the NIF (Figure 2) at the Empire has been constructed by Boyum (1975). The portion of the formation overlying the Siamo Slate are silicate to silicate-carbonate-oxide types. This is followed stratigraphically by the carbonate-silicate assemblage overlain by the oxide facies. This sequence appears to correlate with the remainder of the formation (Gair, 1975). Han (1978, 1982, personal communication, 1982) believes the "primary precipitates" can essentially be subdivided into two groups: a carbonate (sideritic)-chert association which has altered to the present magnetite-chert "ore facies"; and a hematite-chert association which now is a magnetite + iron-silicate + chert + carbonate (ankerite, siderite) "ore facies". ._o>o_ mom o>onm poo“ 5 38m fimoflco> Ann: .Ezwom 538 5.3222 .3235 881032 6:22 oLEEm 05 pm 83950.2 :0: oocsmwoz 05 Ho cofioomummoco ooszcocoU .N 8sz 100$ :00: 32 Vertical variation of "ore facies" as seen in drill core are complex and from the limited data, one cannot discern the relationship of the original material. The "ore facies", however, has gross vertical patterns from base to top. The Siamo-BIF base is transitional, marked by a decrease in clastics and increase in BIF. This transition is overlain by carbonate and variable amounts of iron silicates. Some clastic lenses persist. The silicates are overlain by the magnetite-chert-carbonate association which in turn gives way to an iron- silicate rich facies. Clastics are then found and cores reveal that carbonates are present. The "ore facies" consists of a series of alternating laminae varying in size from mesa-banding (5 to 250 mm) to microbands (0.5 to 1.5 mm) defined by variations in mineral abundances. Haase (1979) observed that mesobands in the NIF consist predominantly of quartz, quartz and iron-oxides, carbonates, carbonates and iron oxide, and iron oxide and silicate. Microbands are well developed in the quartz and carbonate mesobands. The major element geochemistry conducted by CCIC has been summarized by Han: In general, magnetite-chert-carbonate iron formation contains more Fe203, SiO2 and less FeO, A1203, MnO, CaO, C02, P205, and S than does the magnetite-carbonate-chert-silicate iron formation. Both types appear to contain less FeO, A1203, MgO, and S and more MnO than does the magnetite-silicate-carbonate-chert iron formation that contains Clastic beds. Detailed mineralogic and chemical analyses are presented by Haase. Han (1962) has characterized diagenetic replacement, and metamorphism observed at the Empire Mine. The major processes noted are: l)si1icification, 2) recrystallization, 3) carbonatization, and it) magnetitization (diagenetic replacement). These processes appear to have occurred after lithification. "Silicification" to minnesotaite, stilpnomelane, and chlorite is primarily confined to the upper and lower portion of the formation. Carbonitization is found throughout the formation and may be recognized by the presence of poikiloblasts containing chert, magnetite, silicates, and primary carbonates. "Silicification" is observed primarily in the magnetite-carbon-chert bands by the replacement of earlier carbonates and purification of chert grains. Magnetitization occurs in the magnetite-carbonate-chert facies and is characterized by zonal distribution of minerals, in discrete layers of unreplaced carbonate-chert, a transition zone, and magnetite enriched laminae. This process is believed responsible for the formation of the major ore body. Haase (1979) and Han (1962) place the Empire in the lower greenschist metamorphic facies. Haase characterizes the lowest metamorphic zone by an abundance of siderite and quartz with minor minnesotaite. His geochemical and petrologic data imply that any mass transport of material is restricted to an individual mesoband and that metamorphic recrystallization appears to have been isochemical except for the loss of volatiles. PREVIOUS STUDIES Rare Earth Elements in Sediments and Seawater Generally, the REE's are trivalent, have decreasing atomic radii with increasing atomic weight and behave chemically similar in the sedimentary cycle. However, two elements, cerium and europium, exhibit common oxidation states other than +3: Ce (+3, +14) and Eu (+2, +3). This variation results in the anomalous behavior of these elements in sea water and sedimentary processes making them of primary interest in the study of REE distributions. Present day sea water is depleted in Eu by a factor of 0.7 (Wildman and Haskin, 1973) compared to chondritic abundances. Modern sea water also shows a large negative Ce anomaly (Goldberg, et a1., 1963). REE contents of clastic sedimentary rocks are likely to reflect changing REE compositions of crustal source regions and may have a wide vertical and horizontal range of REE concentration within a stratigraphic sequence. Chemical precipitates, however, should be less erratic in their distribution patterns and thus provide the best evidence for the behavior of REE in the sedimentary environment. It is logical, therefore, that REE data from Precambrian iron formations, which are composed primarily of chemical sediments, have been used as environmental indicators in an attempt to unravel their genesis. Two schools of thought have arisen to explain changes in the relative abundances of Ce and Eu to other REE in the sedimentary cycle. The first attributes changes in Eu and Ce behavior to an overall decrease in the abundance of these elements in sea water with time (Wildman and Haskin, 1973; Nance and Taylor, 1976; Fryer, 1977a). The second view stresses that the changes in 10 11 relative concentrations is controlled by sedimentary and diagenetic processes including; the chemical nature of the solution, the iron source solution, the mixing of these solutions, and the depositional environment, which may serve to mask overall sea water patterns (Graf, 1977, 1978; Shimizu and Masuda, 1977). Fryer (1977a) is the major advocate for a time dependency of REE, and uses BIF's as his major source of evidence. Fryer attributes the present Ce depletion to the selective removal of Ce by manganese nodules. His data show a normal Ce distribution within the chemical sediments of the Archean, slight enrichment or depletion during the Proterozoic, followed by a depletion trend. This implies the oxidation of cerium to Cew in the early Proterozoic. Fryer suggests Ce” is scavenged from sea water during the formation of Mn-nodules and, in so doing, suggests their existence as early as 2.3 Ga. It is believed that even in a highly oxidizing state, the catalytic effect of Mn is needed to oxidize cerium. The negative europium anomaly is attributed to the oxidation of europium to Eu+3 in response to the introduction of free oxygen to the atmosphere at approximately 1 Ga. Fryer, therefore, proposes a time dependence for the anomalous behavior of these elements in iron formations. Graf (1978) tested this hypothesis by comparing the REE distribution of Archean Algoma type and Ordovician iron formations (IF). His data show that Archean and Ordovician IF show a positive anomaly and concludes, in contradiction to Fryer, that Eu cannot be used as an outgassing indicator but reflects the specific environment of the particular basin in which the IF form. REE patterns of iron formations are dependent upon the nature of source solution, the degree of mixing of iron source solution and sea water, geological setting, and the amount and type of precipitating or detrital phase. His data also indicate that overall absolute REE concentrations are related to mineralogic 12 components and sedimentary processes rather than a function of time as has been previously suggested (Nance and Taylor, 1976). Graf‘s speculations were alluded to in 1973 by Wildman and Haskin when the application of REE to sediments, and particularly iron formations, was at an early stage of development. Although their data also show a predominant Eu enrichment for Precambrian sediments, interpretations of igneous, metamorphic, and sedimentary processes associated with these samples could not account for a decrease in the overall concentration of Eu in sea water. Early investigations were hampered by a lack of data for REE in recent sediments. Although the data bank is still minimal, the parameters set forth by Graf may be applied to varying degrees. First, it is necessary to evaluate the scope of the investigations which have been conducted to date. The REE imprint may be indicative of the source of the iron solution, specifically, volcanic exhalation (Graf, 1977) versus a terrestrial origin (Wildman and Haskin, 1973). The data indicate that a magmatic source would result in the enrichment of LREE, overall high concentrations, and would not show a Ce depletion relative to Chondrite data. This model has been applied by Graf (1977) where REE's were used as hydrothermal tracers in massive metal-sulfide deposits. Eu behavior would also be affected by a magmatic origin. For example, feldspars selectively concentrate Eu over other REE (Haskin, 1966). As a result, fractional crystallization of a rock containing feldspars will produce a liquid deficient in Eu relative to REE patterns. The resultant feldspar rich solid would be preferentially enriched in Eu. Subsequent melting of a feldspar-rich rock would result in an Eu enriched liquid. Sediments derived from a feldspathic rich source would contain a large positive Eu anomaly. Conversely, a terrestrial source is indicated by the lack of, or slightly positive Ce anomaly, based on the North American Shale Composite (NASC). The associated chemical sediment l3 REE patterns would also be similar to those of accompanying elastic material. Therefore, a correlation between elastic and chemical sediments has been assumed to be indicative of crustal source regions (McLennan, et al., 1979; 1980). The source solution may interact with sea water. This mixing may serve to mask the overall REE pattern. Although mass balance calculations are possible, at this point, there is no adequate method to determine the degree of mixing of these solutions for Precambrian sediments. Therefore, the REE pattern of the BIF may not be representative of the original source area. REE in Chemical Sediments Goldberg, et al. (1963) stressed the role of separate mineral components in REE studies of sediments. Chert and carbonates are of primary concern in the evaluation of REE data from iron formations. Because Ce is subject to selective removal relative to other REE, it is evident that the absence or presence of a Ce anomaly and the extent of deviation can be a good indicator of aqueous environment of formation for cherty rocks. Cherts generally exhibit overall low REE concentrations and primarily act as a dilutent when interbedded with other lithologies. Shimizu and Masuda (1977) compared deep sea cherts (Deep Sea Drilling Project) to "terrestrial" cherts, including the Gunflint Formation. Samples of both cherty rocks and siliceous microfossil separates from radiolarian oozes were analyzed. Deep sea cherts tended to show a large Ce depletion similar to modern sea water, while terrestrial samples were slightly positive or show no Ce anomaly at all. All samples show a small Eu anomaly. The absolute concentrations of the REE within the cherts is extremely variable and appears to be a function of the degree of diagenesis. They conclude that the lack of a Ce anomaly within terrestrial cherts can be interpreted as suggesting that they formed in coastal areas, marginal seas, and land enclosed seas. They reason that Ce in these environments is mostly involved in suspended solid particles and l4 behaves in the same fashion as other REE's, while in the deep seas, Ce is subject to selective removal by Mn-nodules. Two major theories, biogenic and volcanic, have been advocated to explain the genesis of cherty rocks. If volcanic activity is directly responsible for chert formation, the rocks so formed should not show a negative Ce anomaly because it is expected to be introduced and adsorbed within the sediments in the Ce+3 state, rather than being scavenged in the oxidized state (Gem). Predicted REE abundances in biogenic chert are not conclusive; however, in carbonates formed in organic rich environments there are indications that the HREE's are enriched in organic (bacterial) rich environments (Scherer and Sertz, 1980). .Although REE distribution patterns in carbonates are poorly understood, +3 for Ca+2 +2 theoretically, substitution of REE and Fe could be expected. It has also been suggested that the REE's are adsorbed onto the surface of carbonate minerals rather than incorporated into the crystal structure (Haskin, et a1., 1966). Most previous investigations have shown rather low REE concentrations in carbonate rocks (Haskin, et al., 1966; Jarvis, et al., 1975). McLennon, et al. (1979) found that the distribution of REE in carbonates of the Espanola Formation (Huronian) is similar to clastics of post-Archean age. Ce does not behave anomalously. In more extensive studies a distinct depletion of Ce was noted in marine carbonates. They suggest that this anomaly is inherited from sea water. HREE depletion and LREE enrichment is likely in carbonate minerals. HREE depletion may be explained by the tendency of these elements to form soluble complexes in solutions (Piper, 1974; Parekh, et al., 1977). It has been demonstrated (Scherer and Seitz, 1980) that REE are partitioned differently by different phases of carbonate minerals. Scherer, et al. (1980) studied REE's in Pleistocene and Holocene corals. Their data show variations in distribution coefficients for both aragonite and magnesium calcite. 15 Therefore, it appears that carbonate phases must be treated independently when studying REE distribution patterns. Analysis of the carbonates of the NIF have been conducted by Haase (1979). This work indicates that the carbonates exhibit a solid-solution from ankerite to siderite. A detailed study of the effects of the carbonate components on REE patterns may explain the sporadic distributions found by previous investigators. Clastic Sediments The REE content of shales is strongly affected by parent lithology, local volcanism, weathering rates, and depositional environment (Goldberg, et al., 1966; Piper, 1974; Dypvik and Brunfeld, 1976). Weathered rocks are generally enriched in total REE compared with original unweathered rocks. Strong weathering of the source rocks and later reworking of the sediments result in consistent REE patterns over a large regional area. Conversely, rapid tectonic activity will result in variations in clay patterns. Measured abundances of REE in clay minerals indicate that they can be a significant sink for REE's primarily through adsorption processes (Roaldset, 1973). Therefore, REE patterns in shales are both the result of detrital or authigenic processes. Effects of Meta moLphism The effects of metamorphism on sedimentary rocks must also be taken into consideration. It is generally accepted that REE distribution is not altered at least through the greenschist facies (Cullers, et al., 1974; Herrmann, et al., 1976; Menzies, et al., 1979). The most extensive study has been conducted by Ronov, et al. (1977). Their data show that in progressive regional metamorphism of sedimentary rocks, REE concentrations decrease by less than 1096. Ronov further suggests regional metamorphism is an open system only for the volatiles, to some extent sulfur, chlorine, boron, mercury and uranium, and to a very slight l6 extent the heavy lanthanides. Therefore, Ronov, et al. conclude the regional metamorphism is isochemical and does not distort primary chemical trends of sedimentation. Sokoman Iron Formation The most extensive study of a BIF was conducted by Fryer (1977b) on the Sokoman Iron Formation, Labrador, and its associated sediments. His primary observations are: 1) absolute REE abundances of the iron formation are much lower than the associated shales, 2) the iron formation shows a HREE and Eu enrichment while shales exhibit a HREE and Eu depletion compared to NASC, 3) REE patterns normalized to the average slate composite for the iron formation show no relationship to Clastic sediments of the same age, and 4) within the formation, different REE abundance patterns can be correlated with sedimentary facies and mineralogy. Fryer's (1977b) primary observations were as follows. The iron-silicate- carbonate facies show relatively little fractionation of the LREE compared to associated shales. HREE enrichment was noted with HREE preferentially incorporated into the "dark" silicates and siderite. The absolute REE abundance is inversely proportional to chert content, indicating that chert acts as a dilutant of trace elements. The oxide facies is enriched in the HREE relative to NASC, with enrichment of iron, although trace element patterns are widely variable. The Fe-rich oxide facies samples also show variable REE behavior. This led Fryer to conclude that REE's are adsorbed onto the surface of oxides rather than incorporated into the lattice. Variations are also attributed to: 1) differing rates of precipitation of scavenging iron hydrates, 2) changing pH, 3) post- depositional processes, and 4) a highly unstable chemical system. Therefore, Fryer suggests REE's adsorbed onto iron oxide hydrates abundances may have changed when incorporated into the crystal structure. He also concluded that 17 there is an increasing stability of REE in carbonate complexes with increasing atomic number associated with migration and concentration of iron during diagenesis. These results are consistent with petrographic interpretations of formation. Summary of Constraints of REE Studies in Sediments An experimental data base, such as exists for magmatic rock-forming minerals, is urgently needed in sedimentary and natural hydrothermal minerals. Morgan and Wandless (1980) have investigated hydrothermal mineral systems and suggest that the relative pattern in a hydrothermal mineral is strongly influenced by the respective sizes of major and trace cations. They further suggest that, for continental regions, the REE pattern in hydrothermal fluids may closely resemble the light REE-enriched character of crustal rocks, both igneous and sedimentary. If one is allowed great liberal extrapolation, any "exhalative" solutions associated with BIF environments could be expected to be different from the hydrothermal "exhalations" generated along mid-ocean ridges. For example, mixing of seawater with exhalations along mid-oceanic ridges will show REE patterns consistent with mantle derived material, while those generated with the formation of BIF's in restricted basins will be more typical of continental trace element compositions. LOCATION AND SAMPLE DESCRIPTION Core samples were taken from seven diamond drill holes during the Summer, 1980. All samples are from the Empire Mine located in Palmer, Michigan, operated by the Cleveland-Cliffs Iron Company (CCIC). All selected diamond drill holes fall approximately along two north-south lines for a distance of 1050 meters (Figure 3). For the purpose of this investigation, the drill holes are labelled alphabetically from north to south. Thirty-nine samples representative of the dominant lithologies and facies were analyzed for seven REE's using instrumental neutron activation analysis. Twelve carbonate samples were analyzed for major elements using an electron microprobe. Analytical methodology and sample descriptions are presented in Appendices A and B, respectively. Samples were selected from the following lithologic groups: carbonates, ores, chert, and clastics. Whenever possible, at least one lithologic sample was selected from each facies in order to develop stratigraphic control. In addition, six samples of the underlying Siamo Slate were also analyzed for REE content. Data for the Siamo Slate is only available from two diamond drill holes. 18 l9 «$3 \.. / :. 5“ amoa 0. I / 0&1 M / Study diamond drill hole locations, Empire Mine, sec 19, T47NR26W, Marquette District, Michigan. Figure 3. D it PETROLOGY The dominant mineralogies of the samples collected at the Empire Mine consist of carbonates, chert, and magnetite with minor hematite, jasper, chlorite, and stilpnomelane. Samples are generally well banded with laminae varying from micro- to mesobands, while magnetite ore samples are massive. In places laminae have been disrupted by a series of post-depositional features including fractures, veins, and faults (Figure 4). The major mineral/lithologic assemblages included in this study are: l) carbonate with minor chert, 2) magnetite-carbonate-chert, 3) magnetite- silicate, 4) clastics interbedded with silicate, 5) clastics interbedded with carbonate-ore, 6) chert with minor carbonate, and 7) the underlying Siamo Slate. Magnetite is present to some extent in virtually every sample. A thorough discussion of the major lithologic groups is presented by Han (in Gair, 1975). Samples which contain one dominant mineralogy were chosen for this investigation (Appendix B). Carbonates Carbonate samples consist of alternating laminae of both meso- and microbands of chert and carbonate. Analysis was conducted on those samples which contained at least 7096 carbonate and less than 596 magnetite based on hand sample and thin section examination (Appendix C). Carbonate bands contain small percentages of chert (Figure 5), while magnetite is predominant in the silicate rich laminae. 20 21 Figure 4. Fine grained, anhedral siderite matrix with minor chert. Note fracture filled with secondary quartz. Figure 5. Typical siderite mosaic. Anhedral, fine grained. 22 Generally, siderite is very fine to fine-grained (0.005 to 0.01 mm) anhedral equant crystals (Figure 6). Ankerite is coarser grained (0.01 to 0.05 mm) and ranges from anhedral to euhedral crystals. In most cases, ankerite is found coexisting with siderite in alternating laminae, or filling cross-cutting fractures with magnetite (Figure 7). Ankerite appears to be a diagenetic replacement of siderite. Often, it is associated with coarse grained recrystallized quartz. Two samples show minor oxidation. Q95 Chert-rich samples chosen consist predominantly of quartz, with minor magnetite. Thin laminae rich in carbonate may be locally present. These samples consist of both microcrystalline quartz (0.002 to 0.01 mm) and minor recrystallized quartz (0.01 to 0.25 mm). Grains are anhedral forming an interlocking mosaic pattern. In carbonate laminae chert is scattered throughout the band. Minor magnetite is generally present as euhedral to subhedral crystals. Magnetite Representative samples of magnetite ore consist of magnetite coexisting with chert, or with chert and less than 1096 carbonate. Samples are either massive with intergranular chert or iron-silicates or contain microbands of silicate and carbonate. Magnetite is generally euhedral to subhedral ranging in size from 0.1 to 0.5 mm (Figure 8). In the more massive samples, grains may become anhedral forming a mottled appearance. Coexisting carbonate and chert remain equant and anhedral and are sometimes enclosed by magnetite porphyroblasts. Han (1975) has identified ilmenite as the core of some porphyroblasts. Magnetite appears to replace both carbonates and chert (Figure 9). Hematite is present in minor amounts, generally enclosed in quartz or jasper and closely associated with magnetite. 23 Figure 6. Carbonate-rich section with chert laminae and some intergranular chert. Figure 7. Typical siderite mosaic with minor chert. Note fracture partially filled with coarse grained anhedral ankerite. 24 Figure 8. Magnetite ore in carbonate-chert matrix. Note some grains magnetite euhedral, others elongate. 0.02mm Figure 9. Magnetite ore replacing siderite with nonexisting chert. 25 Clastics Clastic material was noted both as discrete layers or lenses or as scattered grains in the NIF. Microscopic examination indicates that this material consists predominantly of quartz (0.1 to 0.35 mm) and minor magnetite (0.05 to 0.15 mm). Only major clastic lenses were considered for chemical analysis in this study (Figure 10). Detrital quartz grains are poorly sorted, angular to subangular fragments comprising 60 to 7096 of the sample, occasionally being rounded and moderately well sorted. Detrital magnetite consists of poorly sorted subhedral fragments. Two primary matrices are associated with Clastic material: 1) carbonate with magnetite, and 2) iron silicates with minor magnetite. The carbonate-magnetite matrix is generally associated with carbonate or magnetite laminae devoid of clastics. Carbonate is anhedral and fine grained, while magnetite is subhedral to euhedral and medium grained. Iron silicate matrices consist of chlorite and stilpnomelane. Minnesotaite has also been reported (Han, 1975; Haase, 1979). Bedding is lacking in this lithologic grouping, although occasionally large quartz fragments are concentrated in discrete layers. Stilpnomelane may comprise up to 2596 of the sample. It exists as small blades, generally at random orientations or in rosette structures. In one more altered sample (Figure 11), stilpnomelane blades penetrate quartz grains. This sample illustrates a more advanced degree of oxidation than most Clastic samples with a large amount of alteration of magnetite to hematite. Chlorite matrix is either platy or bladed, at random orientations, and generally concentrated into patches. Iron Silicates Quartz and magnetite are occasionally associated with minor silicates in both carbonate and chert samples (Figure 12). Common habits of chlorite and stilpnomelane are similar to those described above. Haase has also reported 26 Figure 10. Interformational Clastic lense. Predominantly poorly sorted detrital quartz in carbonate matrix. . . . I». u a“ .. 5' mp.” 597‘ '5 ,.'.p; - _; ., 1_____J 3 $3” 0.04mm Figure 11. Detrital quartz grains. Note bladed stilpnomelane penetrating quartz. 27 , _ .1 l .I . o ' _ ' . . g. . ‘. _ ‘ - . .4; - I . . . , . . , , .. _. . , . ' - I W '- \\ 1 I ' ' '~ ' . ~ - 8‘ it. ‘ x . —‘ .gfrr‘s. :;_ \ Figure 12. BIaded/platy iron—silicates in chert-carbonate matrix with minor magnetite. 28 minnesotaite, biotite, and riebeckite. Stilpnomelane may occur as an intergrowth with chlorite such that positive optical identification cannot be made. Iron silicates are assumed to be an alteration of Fe-rich muds (Han, 1975). Siamo Slate Samples of slate were selected from the upper unit of the Siamo formation. The upper portions are primarily massive with occasional mesobands of chert or coarse grained clastic material. Samples of both the massive slate and coarse grained laminae were analyzed for REE concentrations (Appendix D). The massive unit can be classified as a graywacke. It is comprised of 45 to 6096 detrital quartz grains with minor plagioclase, granular magnetite, and discrete chloritic concentrations which are interpreted to be mafic interclasts (Figure 13). The matrix is primarily chlorite, although traces of biotite, muscovite, and sericite are present. Samples vary from well sorted and well rounded quartz grains (0.2 to 0.35 mm) to moderately sorted smaller subangular quartz grains (0.1 to 0.06 mm). Evidence of strain and compaction are illustrated by a small portion of quartz fragments. A trace of cryptocrystalline rounded chert fragments were also noted. Plagioclase grains are anhedral and serrated with alteration to sericite along twin planes. The composition ranges from Ab92An05 to AblOOAnO, although this is based on a limited number of grains. Mafic clasts are well rounded with slightly serrated edges and are heavily altered composed predominantly of chlorite, biotite, and possibly epidote (Figure 14). Alignment of the platy minerals is marked and different from that of the chlorite matrix. The chlorite matrix is very fine grained and fibrous showing only limited lineation of elongate minerals. Chlorite may penetrate quartz grains, and this penetration shows a definite orientation. ’ 0.04m m Figure 13. Typical quartzite of Siamo Slate. Note moderately well sorted and rounding of quartzite grains in chlorite matrix. Figure 14. Typical graywacke of Siamo Slate. Poorly sorted subrounded to subangular detrital quartz in chlorite matrix. 30 The coarse grained clastic lenses in the Siamo are distinguished by a higher percentage of detrital quartz (70-8096) approaching a quartzite composition. They are more poorly sorted, and contain higher percentages of mafic interclasts (5 to 1096) and cryptocrystalline quartz (Figure 15). Muscovite is also higher in these samples (approximately 196). Although lamination is not apparent, chlorite is concentrated in zones. 31 Figure 15a. Mafic fragments in Siamo Slate. Note variation in grain size and orientation from surrounding chlorite matrix. Figure 15b. Same as Figure 15a under higher magnification. MAJOR ELEMENT GEOCHEMISTRY Whole rock analysis for total iron content (reported as percent F8203) was determined for carbonate, chert, and magnetite-rich samples using X-ray fluorescence. Ca, Fe, and Mg were determined for the carbonates by electron microprobe analysis (Appendix A). Analysis of carbonate grains indicate that although both ankerite. and siderite are present in the Empire samples, siderite is predominant. This is in agreement with Han (1972) and Haase (1979). In some cases, ankerite coexists with siderite. Chemical analysis are presented in Table l. Siderite consists of Fe.76-.91M8.o3-.2ica.oou-.onco3 and ankerite °f Fe.03-.09M8.37-.46 Ca.39_.55C03. No intermediate members appear to be present. Total iron content, reported as Fe203 is presented in Table 2. Large variations are seen in all samples. Higher concentrations in chert rich samples are due primarily to carbonate contamination. Similarities in the variations of concentrations in magnetites and carbonates is attributed to iron present in the siderite lattice. Therefore, iron content does not appear to be a good indicator of ore content. Trace element data presented in the following sections appears to be more chemically diagnostic of percent ore present. 32 33 00.00-00.00 00.00 000700.00 073 00.730 ~04 00 0.0230 00.00-00.00 00.00 3.07050 00.0 00.0-00.0 00.0 At 0.0001: —0.n0-00.~0 00.00 00473.00 00.3 8.7030 00.0 00 0000\0 0N.00-00.00 00.00 00.0-00.0 00.0 2.0-3.0 00.0 00 0E 05004.00 00.3 00.00-00.00 00.00 00.00-00.00 00.00 83 0000\n— 0N.00-00.~0 00.00 00.0-00.0 070 00.7070 00.0 AN: 0000\n— 0050-0000 5.00 00.0-00.0 00.0 00.0-00.0 00.0 00 3.07070 070 00.00-00.00 00.00 00.00-00.00 00.00 00 00:0 00.00-00.00 00.00 0700-0700 2.00 0.7000 00.0 00 00010 00.0-00.0 00.0 3.00-3.00 00.3 00.0-00.0: 00.00 00 «00:0 00.00-00.00 00.00 00.0-00.0 00.0 R.0-00.0 00.0 00 00.0-00.0 00.0 07000-0700 00.00 0N.u0-0N.00 00.00 3: :10 00.00-00.00 00.00 00.07000 00.0 00.00. 3.0 At 0.0005 0.000500 00.00 00472.0 0.00 00.0-00.0 00.0 00 000?. 00.00-00.00 00.00 $4703.00 0700 00.7070 00.0 :3 0.20:. owcmm 000.530 00.8% owmeo>< 000.90 000.530. 80:52 29:00 Moofi m000.2 .mofimconumo Ho 03me 3800.602 A 200... 34 Table 2. Total iron content reported as percent FeZOB' Carbonates Chert Ores Sample Sample Sample Number Fe203 Number Fe203 Number FeZO3 A/517.5 32. 99 D/88.2 25.64 C/87 27 .96 A/582 26 .01 D/752 9.24 C/582 36.36 B/278.5 34. 80 D/958 12.63 C/929 37 .96 C/41 27.26 F/76 33.91 C/1042 38.61 C/1386 32.28 F/225 5.98 D/518 49.23 D/126 24.01 F/297 29.31 D/733.5 26.97 D/878.5 34.79 G/845 17 .37 E/70.5 22.34 D/1210 11.80 F/695 43.81 F/8 32.75 F/1119 25.02 G/578.5 28 .68 G/761.5 39.31 H/448.5 29 .34 H/710.5 22.01 H/805.5 31. 02 H/849.5 37 . 48 TRACE ELEMENT GEOCHEMISTRY The-REE's (La, Ce, Sm, Eu, Tb, Yb, and Lu), thorium, and chromium of samples from the NIF and Siamo Slate were determined by nondestructive instrumental activation analysis. Concentrations, measured in ppm compared against liquid and U.S.G.S. rock standards (see Appendix A), are presented in Table 3. It is standard procedure to normalize REE concentrations to chondrite values or to the North American Shale Composite (NASC) (Haskin, et al., 1968). For simplicity, values quoted in this investigation will be chondrite normalized unless otherwise stated. Absolute concentrations of REE vary with lithology, but all rock types show a marked fractionation of the LREE. Overall variations with respect to Eu and Ce behavior vary between lithologic groups although all show an Eu depletion compared to chondrite. LREE to HREE ratios and Eu observed to Eu predicted are presented in Table it. Carbonates Twelve carbonate samples from the Empire Mine were analyzed. Chondrite normalized values are plotted in Figure 16a. The carbonate sediments exhibit an average LREE value of 20 times chondrite with a range of values from approximately 4! to #0x. HREE are enriched an average of 2.5 times chondrite, with values ranging from 1 to 4x. The highest abundances are observed in samples containing considerable amounts of magnetite, iron silicates, or hematite. Ankerite samples are depleted compared to siderite and show a substantially smaller enrichment of the LREE. All samples exhibit an europium 35 36 00.0 00.0 000. 00. 0:. 0:. 000. 00.0 00.0 .50 .00 00.0 00.0 00.0 00. 0: a. 000. 2.: 00.0 00802 00.: 00.0 00.0 00.0 00.0 00.0 2.: 00.2 00.0: 0.0005 00.0 00.0 00.0 00.0 0:0 00.0 0:: 00.0: 00.0 0.03: 00.: :0 :.0 0: 00.0 00.0 00.: 00.0: 00.: 0.050 0:.0 00.0 00.0 00.0 00.0 0:.0 00.0 00.0 00.0 0\m 00:0 00.: 00.0 00.0 00.0 0:0 00.0 00.0 0:0 0:25 00.: 00.0 00.0 0: 3.0 00.0 00.: 00.0: 00.0: 0.0005 00.0 00.: 00.0 00.0 3.0 00.0 00.0 00.0 00.0 00:0 00.0 00.0 00.0 00.0 00.0 2.0 00.0 00.0: 3.0 000:5 00.0 0: 00.0 00.0 0:0 :0 00.0 00.0: .00 Eu 00.0 00.0 00.0 00.0 00.0 :.0 00.0 00.0 00.0 0.0003 00.0 R: 00.0 00.0 :0 3.0 00.0 0:0 00.0 0002. 00.0 00.: 00.0 20 0:0 0:0 00.0 0: 00.0 0.202 G i 3 0> e. an em 8 0.. Nmfimm. 322000.00 .AEaav compmhcoucou “coca—m atom 00mm 3280?. .m 2an 37 N04 N04 No.0 No.0 00.0 00.0 3.0 00.0 3.N Son .5 nod :0.N 00.0 3.0 3.0 ~10 0n.0 mmd 3.0 owmco>< 00.0 «0.0 00.0 3.0 3.0 3.0 00.0 00.0 no.0 3&0 $6 000 00.0 00.0 N10 2.0 00.0 01$ 00.0 RN?— wné 3.0 no.0 2.0 3.0 no.0 00.0 00.: 3.0 nNNE nmfi mm.m 00.0 00.0 3.0 010 00.0 00.0: 00.0 00E 00.0 5.0 no.0 0N.0 2.0 00.0 00.: mm; nn.~ 0005 R5 3.: no.0 3.0 3.0 00.0 00.: N06 00.0 «who N00 00.0 00.0 no.0 2.0 2.0 0nd m¢.~ 00.: N00?— 00 E. :4 0; 0:. 3m Em «C a: 80:52 oEEmm hum—.5 6000:0000 0 2000 38 0n.n ::.:: 00.0 n0.0 0:.0 0:.0 00.: 00.0: 00.0 .50 .00 00.0: no.0 00.0 00.0 0:.0 nN.0 :n.: 0n.0: :0.0: 000.630. n0.0: no.0 n:.0 n0.0 0:.0 n:.0 0n.: 0n.0: 00.0: n.000\I 00.0: 00.0: 0:.0 00.: n0.0 00.0 00.0 00.0n no.00 n.0:0\I 00.n: 00.0: 00.0 n0.o 0:.0 00.0 nm: 00.00 00.0: n.:00\u 00.n0 00.00 00.0 00.0 0n.0 00.0 00.0 00.0n 00.:n 0:::\...: 00.0 n0.0 00.0 00.0 ::.0 00.0 00.0 00.:: 0n.0 n00\..: 0:.0 0n.0 00.0 nn.0 no.0 0:.0 00.0 nn.:: n0.n n.00\m 0n.0 00.n 00.0 0n.0 00.0 n:.0 :0.0 00.0: 00.n n.nn0\n: :0.0 00.: 00.0 0n.0 no.0 0:.0 00.0 ::.:: 00.n 0:n\n: 0:.0: :~.n 00.0 0n.0 ::.0 0:.0 00.0 0n.0: 0n.0 00020 00.0: 0n.n no.0 n0.0 00.0 no 00.: 00.0: n00 0N0\U n0.~: 00.n 00.0 00.0 no 00.0 00.0 :0.0: 00.n N030 0:.0 00.: no.0 00.0 0:.0 :00 n0.0 00.0 0n.0 00\U .U E. 0.: 0; 0:. 01”: E0 00 0.: 000E002 0:05:00 mmO 0805208 0 200.0 ’39 00.0 00.:0 0:.0 :0.: 00.0 00.0 00.0 00.00 :0.00 .50 .00 00.0: 00.: 00.0 00.: 00.0 00.0 00.0 00.00 0:.00 0020:: 00.:0 00.:0 :0.0 00.0 00.0 00.0 00.0 00.00 00.:0 000:: 00.0: 00.00 00.0 00.0 00.0 00.0 00.0 00.:0 00.00 0.0005 00.0 00.: 00.0 00.0 00.0 0:.0 00.0 00.:: 0:.0 0.0000": 00.0 :0.0 00.0 00.0 00.0 00.0 00.0 00.0: :0.0 0:000 00.00 00.00 :0.0 00.0 00.: 00.0 00.0 00.:: 00.00 0.00:5 00.0 00.0 0:.0 00.: 00.0 00.0 00.: :0.00 00.0 0.00000: 00.0 00.: 00.0 00.0 00.0 0:.0 00.0 00.:: 0:.0 0.00000: 00.0: 00.00 00.0 00.: 00.0 00.0 0:.0 00.00 00.00 0.0300 5 00 0.: 0> 00 :0: 60 mo 0.: Mummm: 02.0030 .:0000:Eoo: 0 200.0 40 n0.0 n0.0 00.0 :00 00.0 ::.0 0n.0 00.0 00.0 $00 .00 00.0: 00.0 00.0 0n.0 0:.0 00.0 00.0 :N.n: 00.0 00836. $203030 00.0 n0.: :0.0 0n.0 :0.0 00.0 00.0 00.00 0n.0 .>on: .um 0.:0 n.n: :n.0 nn.N 00.0 n0.0 00.:: 0:.0n 00.00 owmoo>< 3:00.530 00.:n 00.0: 0:.0 n0.0 0:.0 0n.0 00.: 00.00 00.0: 0:0}: 00.0 0n.0 No.0 nn.0 00.0 0:.0 00.0 0n.0: nn.n 000).: 00.0n n0.:: 0n.0 :0.N n0.0 00.: 0n.0: 03.00 n0.nn n.000\: 0:.0: 00.n no.0 0n.0 00.0 0:.0 00.0 n0.:: n:.0 n.n00\:.: Nn.00 00.0: :n.0 00.0 n0.0 00.0 00.n n0.0n 00.00 mnm0\u 0n.0: 00.0: 00.0 0n.0 n0.0 00.0 n0.0 0n.0: n0.n m 05 3 nomszLo: :i0 63?: 83.5% 692E 9.20va vcm Btoxcm 6:6 83.29 32372 4N oSwE $952 3205 N6 : 2. 8 8 5 mm mm mm 8 8 5 8 mm mm D mm _ 3 m; m: B :m ”.5 ¢._ .. 3 .S T m .13 no 2 u r 3 11mm //\ ”IIN 000 nd Hm . B 7 r u 3 :{o 4 0 228 e W 938 B m 61 Scherer and Seitz also established two main trends during diagenesis; a strong increase in REE/Ca ratios and a shift towards lighter REE's. They further noted that the highest contents of REE's are present in magmatic and hydrothermal carbonates and that both are enriched in the LREE's. A similar plot for Yb/Fe203 versus Yb/La of NIF carbonates (Figure 22) indicates the reverse trend (i.e., HREE enrichment), although data is not conclusive, again illustrating the effect of magnetite formation in the system. It was also found (Scherer and Seitz, 1980) that REE's have a greater distribution coefficient for Mg-calcite than aragonite. Their conclusions support theoretical predictions that the REE's can substitute more readily for calcium in the calcite structure than for Ca+2 in the orthorhombic structure of aragonite. Therefore, both crystallographic influences and the formation of new minerals during diagenetic processes may cause minor reequilibration of REE's during alteration. Recent work by Chaudhuri and Cullers (1979) indicates moderate REE enrichment with depth in Gulf Coast sediments. They conclude that these changes are primarily in response to changes in source provenance, and diagenesis plays only a minor role. Variations in REE concentrations noted in this study, by Scherer and Seitz, and by Chaudhuri and Cullers may indicate that diagenesis does have at least a minor effect on REE concentration. Based on these preliminary observations, interpretations of anomalous REE behavior appear applicable to trace element studies, but interpretations of concentrations with respect to ancient sea water and comparisons between iron formations will be more meaningful with accurate distribution coefficients for carbonate phases and the determination of their mobility during diagenetic processes. Therefore, modern analogs, or an experimental data base are needed to fully evaluate this problem. 62 0.05 0 CHRBDNRTES 0 1 l 0.04 0.03 YB/FE203 0.02 to 0.01 C3 -1.20 -1.10 -f.00 -0.90 ~0.60 -0.70 -0.60 LOG (10] YB/LR Figure 22. Plot of Yb/ La versus Yb/Fe203 for Empire carbonate-rich samples. 63 Metamorphism Possible controls of metamorphism on REE distribution cannot adequately be evaluated based on this investigation alone. Samples were collected from the lowest metamorphic grade of the Negaunee. Therefore, correlations between metamorphic facies cannot be considered. Previous investigations indicate that metamorphic alteration of sediments has little effect on REE distribution through the green schist facies. Ronov, et al. (1977) data indicate that metamorphic processes may result in a slight HREE depletion. HREE depletion is noted in this study, but appears to be both a product of mineralogic control and the chemical character of the primary solution. Generally, carbonates, oxides and cherts are depleted in HREE (Scherer and Seitz, 1980; Schock, 1979; Goldberg, 1963). Some samples show a more marked HREE depletion which may be enhanced by the loss of volatiles during low grade metamorphism or diagenesis. Since the mobility of REE's have not been fully addressed in sediments, it is possible that the REE's become less mobile with time. Vulcan 1F samples show decreased concentrations suggesting increased mobility with metamorphic grade (Slaughter, et al., 1981). But, this problem cannot be totally evaluated with the present data. Haase (1980) has done preliminary carbon and oxygen isotope studies on Negaunee carbonates. These data show that homogenization during diagenesis and low grade metamorphism was limited and that magnetite formation had no noticeable effect on the isotope geochemistry. Fryer has observed inconsistent patterns in Cr content of oxides in the Sokoman iron formation. He has attributed these fluxuations to the migration of desorbed ions during post- depositional processes. In Negaunee samples, a positive correlation is observed between the REE and chromium concentrations. This appears to follow 64 increased oxide content of the rock samples based on hand sample and thin section examination. If Fryer's assumptions are correct, the Negaunee oxides have not been sufficiently altered during post-depositional or metamorphic processes. One sample (G/578.5) shows minor recrystallization to chlorite which appears to have enhanced REE concentrations similar to iron—silicates of the Sokoman. Oxidation Post-metamorphic alteration has resulted in oxidation of magnetite to geothite and iron carbonates to hematite in the Negaunee. These processes have occurred locally at the Empire and are attributed to the downward circulation of groundwater along fractures, contacts between dikes and iron formations, bedding surfaces, and pore spaces (Han, in Gair, 1972). These processes are more extensive in the "hard ores" of the Negaunee. Samples analyzed from the Empire are primarily from the unoxidized portions of the formation. Two carbonate samples (C/ 1386 and F/578.5) show minor alteration from siderite to hematite and chert to jasper. Oxidation has resulted in a slight overall depletion. Reverse patterns have been found for the hard ores of the Tilden mine, Negaunee Iron Formation and are plotted in Figure 23 (Slaughter, et al., 1981). The high proportion of jasper in these two samples may explain the conflicting results. Summary Experimental determinations of crystallographic versus adsorption controls are urgently needed to access REE behavior in sediments. The effects of post- depositional processes (diagenesis, metamorphism, and oxidation) may be substantially different for each type of REE fixiation. Until such an experimental data base has been obtained, investigations such as this, of natural sedimentary systems, will remain of an empirical nature. 65 O a 1 + NEGBUNEE HRRD DRES o'od 11.1.73 5.“. 1 _J 4 (I 4 t J M 31: w i '- 1 E! 1 ‘3 1 2 2'3. 0 3. 1‘ LR CE _ SM EU TB YB LU O "66 6'7 6'6 6'9 6'0 6'1 6'2 6'3 6'4 6'5 3'8 6'7 66 519 7.0 7+1 72 FITDMIC NUMBER O ‘ + NEGBUNEE HBRD DRES 0°. 1 But: \- fl .J G 2 1 O: O z . CT 8 5'2, 3 J J , LR CE SM EU TB YB LU o '66 ST 6'6 6'9— 6'0 6'1 6'2 6'3 6'4 6'6 6'6 6'7 6'6 6'9 7'0 77 72 RTONIC NUMBER Figure 23. REE concentrations for Negaunee hard ores normalized to: a) chondrite and b) NASC. IMPLICATIONS OF REE CONCENTRATIONS ON CHANGING SEA WATER COMPOSITION WITH TIME Assuming that the above processes have not significantly altered REE concentrations, the chemical precipitates may be used to evaluate the chemical characteristics of the solution from which the NIF was deposited. From these data, we may then infer possible conditions of formation of the NIF. It has been proposed that the chemical composition of sea water has changed through time by several investigators. Sandberg (1975) and Wilkenson (1975) both support an increase in the Mg/Ca ratio with time. It also appears that 018/0l6 values have increased to the present (Veizer and Hoef, 1976). Therefore, it would logically follow that trace element, specifically REE concentrations in sea water may also be time dependent. These geochemical trends will be reflected in chemical precipitates such as carbonates and chert, while clastic sediments will reflect the composition of the continental crust (Taylor, 1964; Haskin, et al., 1966; Piper, 1974). Two different trends are noted for Eu behavior of the NIF chemical sediments relative to NASC. Chert and chert-carbonate samples show a marked negative Eu anomaly while carbonate-rich samples are close to predicted values or slightly depleted. Negative values are observed for all samples in this study compared to chondrite. Ankerite samples generally show a stronger Eu depletion compared to siderite-rich samples. These observations do not correspond to Fryer's results. He attributes the positive europium anomalies in Archaean iron formations to the preferential incorporation of Eu"2 in sediments. Therefore, Archaean iron formations formed 2 3 before the oxidation of Eu+ to Eu+ with the introduction of oxygen into the 66 67 atmosphere. The Sokoman shows fluxuations in Eu (chondrite normalized) behavior which Fryer would interpret as deposited during the transition period. The NIF which was precipitated at approximately the same time as the Sokoman shows negative values suggesting the presence of an aerobic atmosphere or seawater at this time. It has also been suggested (Eriksson and Triswell, 1978) that an oxygenated sea during the formation of Lake Superior-type iron formations resulting in the precipitation of iron in the +3 state and consequently allowed the segregation of iron and manganese. Therefore, it appears that other chemical constituents have influenced REE values observed in the ore body and it does not necessarily represent overall sea water composition of the time. Instead variations in Eu anomalies appear to be the result of different conditions of formation as proposed by Graf (1978), i.e., in response to phases precipitated and the chemical nature of the solution from which the sediments were derived. Clastics and chemical sediments have also been utilized to relate REE compositional changes to crustal evolution. McLennan, et al. (1979) found a decrease of La/Th from 3.6 to 2.7 from Archaean to post-Archaean times in elastic sediments. This is assumed to reflect a change in overall continental crust from a more mafic composition. Similar ratios obtained from the Empire (Table 5) have maximum values comparable to those of post-Archaean sediments for the interbedded clastics. But, in general, clastic interbeds and the Siamo Slate have ratios substantially lower than those observed by McLennan. The chemical sediments also show a similar trend. However, increased Th directly relates to increased REE in all samples (Figure 24). Therefore, if the major source of total REE in sea water is considered to be from continental erosion, data from the NIF does not adequately support a model of REE evolution of the crust with time. 68 mun.~ Q33: mom.~ n.now\: 024 n63} 3w; OwiiI QNNJ QSQU wad n.anD $50 32E me.N mi nmmé meI wwN.N nme :34 Emin— wmmJ n.anU Nmné Coim— wRJ 230 mum; 9330 nmwd 33: ~36 n.nwnE 6on4 manta mmné RNE N34 3:0 3m.m 83: 6mm; mg...” 220 mNJ mNNE 5w; mwmio mwm.~ OSQI MMNJ n.mN:\U 3o.~ Ngzu mum; min. USA :10 on; n.an: mmné 593m 3n; 330 mi .m waD mnmfi mema :n.0 mmmmkv $n.~ n.anm own; anfiv mmw.~ NnQQ owe; an\< Rm; (3&0 ommé w.~3\0 Ruin R30 30.5 N.ww\o nnmé n.2n\< ofimm 59:32 25% LonEzz 25% .8852 03mm 52832 03mm 63:52 29:3 oEEmm 035mm oEEmm oEEmw 32m 33930 0.0 :20 35:3ch .mofimm EatochuEzcmficmq .n 033. 69 3' 6; ° ' o CRRBONRTES o . o 53- _ <> m—i I 0 Hal 0 z °° o. o. vr— 0 0 «9 N... ° l I T I 0 2 4 6 8 10 PPM LR a) b A CHERT A co- A A :r: A '_. v4 2 A a. 0" A N- A ° 1 l r l 0 1 2 3 4 5 PPM LR Figure 24. Plot of Th versus La for: a) carbonates, b) chert, and c) magnetite-ore. 70 S c [I] MRGNETITE ORE o_ E] m . G: ...J 28-1 [1] 0.. 0. [I] g- E] E] @I E] O I I I I ' 0 10 20 30 4O 50 PPM TH Figure 24 (continued). 71 Fryer also indicates an increase of REE concentrations of shales with time. Data from the Siamo show values comparable to those of NASC for quartz-rich samples, and slightly depleted for graywackes. It therefore appears that absolute REE concentration in sediments is dependent on mineral phases present and degree of weathering and transport rather than age. The NIF is of a comparable age to the Sokoman, and shows substantially higher concentrations. Therefore, BIF's do not reflect an increase in the REE concentration, although this may also be a reflection of mixed mineralogies used in both of these studies. Three major conclusions may then be reached: 1) the geochemical nature, specifically relative REE concentrations have not changed through time, 2) post-depositional processes have since altered the concentrations of the REE in the Negaunee, or 3) banded iron formations are not representative of the overall chemical nature of sea water during the Precambrian. The first and second conclusion cannot be adequately evaluated by this study alone, although it appears that oxidation and diagenesis affect overall concentrations in at least minimal amounts. Recent Nd isotope studies in sea water (Hooker, et al., 1981) provide evidence that metalliferous chemical precipitates dated at l490 m.a. have comparable Nd values to those of modern sea water. This is in direct contrast to the implications of major element systems presented by Wilkenson and Sandberg, although major and trace element systems may act independently. These data may also not reflect any changes which are postulated to be in response to the introduction of oxygen into the atmosphere during the Precambrian. The most conclusive proposition concerning BIF's specifically are those presented by Graf (1978). The presence of a major iron- rich ore bodies suggests geochemical properties and sedimentological processes different from those of a normal sea water system. This is supported by evidence that similar iron formations have like REE imprints and formed under 72 analogous sedimentary systems regardless of age. It is this conclusion which is supported in this investigation. INFERENCES OF TRACE ELEMENT DATA ON THE FORMATION OF THE NIF AND COMPARISON TO OTHER IRON FORMATIONS REE patterns observed in this’study are substantially different than those found for other iron formations, specifically the Sokoman (Fryer, 1977b) and Algoma-type iron formations (Graf, 1978). Therefore, it is proposed that the NIF was deposited under different sedimentary conditions. Post depositional alteration processes may also have varied between provenances. The chemical precipitates of the Negaunee show a marked Eu depletion or compatible values compared to NASC while the Sokoman shows an Eu enrichment trend compared to slates. Generally, the Sokoman has a negative Ce anomaly while those of the NIF are variable. NIF chemical precipitates and oxides have similar distribution patterns compared to associated clastic sediments, or are slightly HREE enriched (Figure 18), while Sokoman samples show extreme fractionation and anomalous Eu behavior compared to associated detrital materials Algoma-type iron formations have similar distribution patterns to Empire samples, but have an extreme Eu enrichment. Cerium data in cherts has been used as an indicator of sedimentary environment. Cherts analyzed from the Empire show low absolute abundances compared to all other lithologies (Figure 19). Significantly, the cherts do not have the Ce depletion characteristic of present day sea water but have patterns which Shimizu and Masuda (1977) describe as being "terrestrial", i.e., formed in coastal areas, marginal seas, or within land enclosed areas. Similar patterns have also been found in a chert sample from the Gunflint Formation (Slaughter, et al., 1981). 73 74 Haase (1980) found both lateral and vertical variations in carbon and oxygen isotope values in carbonates at the Empire Mine. These observations, complemented by the complex facies changes which exist, led Haase to suggest that deposition of the BIF occurred close to the paleomargin of a depositional basin. In addition, the REE data, unlike that of the Sokoman BIF, do not appear to be strongly controlled by major facies changes. For example, carbonate mesobands from the silicate facies have similar patterns to carbonate mesobands from the oxide facies. Generally, distribution patterns in the NIF show minor fluxuations in comparison to major changes in REE patterns between facies seen in the Sokoman. Therefore, it is possible that the trace element geochemistry of the solution from which the NIF was precipitated changed locally. This interpretation would lend support to Han's suggestion (personal communication, 1982) that the defined facies at the Empire Mine are primarily mining terminology and may be a result of subsequent alteration (i.e., ore enrichment) rather than a product of major changes in depositional conditions. Therefore, the mineralogy may have originally been much more uniform prior to the occurrence of metamorphic and diagenetic processes. The consistence of REE patterns may be interpreted to mean that primary mineralogic controls of REE uptake dominated the REE abundance at the Empire location and that diagenetic changes or post-metamorphic oxidation have not resulted in significant variation of the original REE imprint. Therefore, it appears from the primary REE patterns that primary mineralogies at the Empire were originally more homogeneous than seen today. Fryer noted the most consistent REE patterns from the Sokoman oxide facies were from samples which were extensively recrystallized and enriched in iron. These samples show strong HREE enrichment (NASC normalized), whereas other oxides show variable REE patterns (Figure 25). In contrast, the Empire 75 Nh .Umm SEEM ocm cmonom c3 mcoflmhcmucou mum Ho coflcmanU .nm 2sz mumzzz o H 20.5 ab Db mm me be mm no em mm mm um om mm mm hm mm — b n p n p — r p — b b — _ L I D .r 3.. m; 3m 2m mu c4 N — - w % q/T .. 3 . N n 0 Lind 0.“ :6 _II I . Z Ina/m. 3 . U MED zczozom m¢¢mw>¢ + wmo mmHmzm u¢¢mw>¢ B 1 .01 ...0 . v 0 .n.....:L C2“ TY.; A A .\ 76 samples are less variable, have a slight LREE enrichment, and exhibit higher absolute REE abundances relative to the Sokoman oxide facies. Fryer suggests that the variation within the Sokoman oxide facies is related to complex post- depositional behavior during the transition from an adsorption on iron oxide hydrates to crystal structure incorporation. - Thus, the processes (post depositional) or solutions operative do not equate to the Negaunee. Indeed, the Negaunee oxides provide evidence that the LREE's behave differently during iron enrichment (siderite to Fe-oxides) than to the iron enrichment process on the Sokoman. Han (1962) states that magnetization or diagenetic enrichment, a process whereby magnetite-chert slightly to completely replace carbonate-chert is believed to be the most important process in forming the chief ore body at the Empire Mine. He reports all gradations from unreplaced carbonate-chert, a transition zone in which poikiloblasts of Ca-rich carbonate develop, and the enriched magnetite. This change is accompanied by a reduction of thickness and generates new-formed magnetite of unusually uniform size. This process is summarized as follows: elimination of fine-grained magnetite and magnesium iron carbonates; the lowering of chert content; development of calcium-rich carbonate; development of magnetite. The amount of magnetite formed from carbonate-chert layers is less than that formed from primary magnetite bearing chert layers. Han's analysis of unreplaced and enriched portions define the process as an addition of FeZOB’ FeO, C30, and A1203, and appreciable subtraction of MgO, MnO, C02, and SiOZ. Hence, it may be stated that the magnetization is not a simple diagenetic oxidation of the magnesium iron carbonate but also involves an extensive ionic diffusion and possibly a nucleation of small magnetite during the enrichment stage. Han further suggests (1978) that the interformational structures (i.e., faulting, folding, fracturing, and 77 brecciation) may have played a major part in channelling oxygen, water, and carbon dioxide through the system. It is noteworthy that the Sokoman iron enriched samples analyzed by Fryer are devoid of any carbonate, and that the one sample which exhibits LREE (NASC normalized) enrichment and high REE abundances is extensively replaced by carbonate. In addition, it has been suggested that the Sokoman was originally deposited as a gel and iron oxide hydrates (Klein, 1974*; Lesher, 1978). Han (1978) proposes that these gels crystallized primarily into iron silicates and hematite. Hydroxides may have also interacted with organic matter resulting in increased mobility during diagenesis and low grade metamorphism. The NIF, on the other hand, appears to have formed directly from precipitated carbonate and/or hematite. Thus, the carbonate formation appears to be a controlling factor in REE abundances. Transport of Fe, REE's, Ca, etc., could take place in carbonate-rich solutions. Precipitation of the scavenged REE's would be different depending on whether carbonates precipitate (Empire) or not (Sokoman). If the REE's migrate as carbonate complexes the solutions would be enriched in the more stable HREE's and precipitate HREE enriched phases so long as carbonate was not precipitated. It appears likely that the formation of carbonate and magnetite does not favor the enrichment of HREE's at the Empire Mine, and tends to increase the absolute REE abundances. Hydrothermal alteration has also been suggested by Cannon (1973) as a mechanism for the formation of the Negaunee hard ores. This type of system would allow for the addition of trace elements into the NIF system, or substantial migration of elements within the system which would not be expected by diagenetic or low grade metamorphic processes alone. 78 Morgan and Wandless' (1980) investigation of hydrothermal minerals concludes that siderite formed during this process will exercise strong crystallographic control over the incorporation of REE. NIF siderite shows distribution patterns markedly different from those expected based on structural considerations. In addition, unpredictable trace element patterns of ankerite also occur. This suggests hydrothermal processes were not operative during formation of Empire carbonates. Hard ores (Cliffs Shaft and Cliffs Drive) from the Negaunee (Slaughter, et al., 1981) show extreme REE enrichment compared to Empire samples. No Ce anomaly is observed, and a slight Eu depletion exists normalized to chondrite abundances (Figure 23). Similar trends are noted compared to NASC (Figure 23). Significantly, a HREE enrichment trend is found with concentrations approaching that of the average shale composite. Cannon (1973) notes that hard ores are located in the upper portions of the NIF. He suggests that fluids, probably derived from dehydration and decarbonization during Penokean regional metamorphism, migrated upwards resulting in the reduction of hematite to magnetite, and the precipitation of magnetite from hydrothermal solution. The REE, and especially the HREE, tend to complex with carbonate ions. Therefore, HREE enrichment trends observed in the NIF hard ores may be explained by their formation by hydrothermal processes. The presence of slump features and broken laminae in Empire samples indicate that during magnetite formation a reduction of volume did occur (Han, 1978). Therefore, the addition of trace elements, iron, and chromium through hydrothermal processes is not necessary for ore formation. These factors suggest that hydrothermal alteration was not a major factor in Empire ore formation. In addition, the plot of La/Sm vs. Sm/Yb cannot be used as a 79 discriminatory for rock type or mineral speciation (Figure 26). The overlap may reflect the common origin. Instead, it appears that magnetite enrichment occurred by diagenetic processes whereby the movement of volatiles occurred along structural features allowing for the concentration of Fe, Cr, and REE during oxide formation. A major controversy remains concerning the source of iron for the formation of massive iron deposits. The two primary theories propose that iron was originally derived from: I) continental erosion (James, 1954) or 2) subaqueous volcanism (Goodwin, 1956). Presently, each argument has major limitations. Limited REE data on the Siamo Slate give information concerning the character of the Negaunee. Two major trends are noted in the data. The quartzite samples (G/92BB and H/867.5) show a marked Eu depletion while the graywackes show a slight Eu depletion on the order of magnitude found in NASC. This is evidence for two possible source regions for elastic material based on the petrology of the samples. The predominant mineralogies consist of quartz and albite, indicative of a sialic source, and chlorite and mafic rock fragments, probably derived from the weathering of a more mafic terrain. Clastic material is considerably enriched compared to later chemical precipitates. Therefore, a mixing of material from two source regions is implied. Possible origins of this material are the Mona Schist and the Compeau Creek Gneiss of the basement complex. Trace element data are not presently available on these formations to correlate to the Siamo. Clastic material contained within the Negaunee show similar REE distribution patterns, specifically with respect to Eu behavior, but are slightly REE enriched compared to the Siamo. Again, two distinct patterns in Eu behavior are noteworthy. Specifically, the higher REE concentrations are LFl/SM 80 S. - ‘” A . .0 o o m, o 3" o omm Em; ‘93 A [I] s_ o 0 .; 49 E] <> luggig? i? 45 §- III ° . . ' o CLFISTICS 2‘ [:1 MRGNETITE 8 <9 CRRBONRTES .:" )ZI SIRMO SLRTE . ACHERT O O .[25 0 :50 0 175 1 :00 1 .125 1 .150 1 55 2 .100 2 :25 SM/YB Figure 26. Plot of Sm/Yb versus La/Sm for Empire samples. 2.50 81 observed in clastic samples (C/#12.9, E/ 1126.5, G/285.5, and H/593) which also have a more marked Eu negative anomalies compared to NASC. Mineralogically, samples are also similar, although petrographic data indicate a lower degree of sorting and rounding. Chromium and thorium values are also within the same range. This suggests that the clastic component of the Negaunee has a similar source area to that of previous shale deposits. Therefore, erosional processes were still periodically active in similar lithologic regions during chemical sedimentation. Higher concentrations in erosional products included in the NIF are probable inherent from a short residence time and exchange with sea water. Alternatively, increased concentrations may be due to less alteration in the younger Negaunee interbedded Clastic sediments. REE patterns of the chemical precipitates are at lower concentrations but mimic patterns of clastic material. This is in direct contrast to observations of Fryer (1977b) in the Sokoman iron formation. Therefore, this correspondence indicates that a detrital source for the iron solution cannot be ruled out. Graf reports similar correlation for chlorite-rich New Brunswick Iron Formation. These similarities are not seen in associated iron sulfide deposits presumed to be of hydrothermal origin. The input of volcanic material in sea water would result in distinct REE patterns of the resultant solution and derived sediments. For example, Graf (1977) suggests that enriched europium values in New Brunswick (Bathurst-New Castle District) Algoma-type iron formations and the Proterozoic Hamersley BIF are the result of the interaction of hydrothermal solutions with felsic material prior to entering sea water. The addition of feldspar-rich material, which preferentially incorporates Eu may cause a change in the chemical character of hydrothermal solutions from which this type of iron formation was formed resulting in distinct Eu enrichment beyond that predicted by crystallographic control. 82 Although samples of known volcanic origin analyzed by Graf show enriched Eu values, in contrast to the NIF this does not preclude a volcanic origin for Fe- solutions in the Negaunee, as has been suggested by some investigators. Although no known volcanics are associated with the Negaunee, input into the hydrologic system, removed from areas of immediate deposition by volcanic exhalation would affect trace element concentrations of the solution and sediments precipitated. Therefore, analysis of volcanic material in the Lake Superior region may give further insight into this problem. CONCLUSIONS The REE data on the Negaunee Iron Formation chemical sediments and associated clastics characterize sedimentary controls and geochemical conditions for its formation during Precambrian times. The following conclusions are drawn from the data: 1. Detrital sediments have marked REE enrichment compared to primary chemical components. Secondary magnetite-rich samples are generally enriched compared to carbonates and chert. 2. Changes in REE distribution patterns between ankerite and siderite and constant chromium content suggest that REE are incorporated into the crystal lattice rather than adsorbed onto the surface. Oxide enriched ores have a wide range of abundances which may reflect original trace element contents were controlled by adsorption processes. The relative REE abundance patterns are grossly similar for all chemical sediments. This may be used to argue for a diagenetic replacement of carbonates as suggested by previous investigators. 3. Metamorphism and post-metamorphic oxidation does not seem to have significantly altered Empire samples. Compared to analyses of samples from the Tilden Mine and Vulcan Iron Formation which illustrate a marked redistribution of REE during these processes, Empire Mine REE patterns are relatively consistent. 83 84 Comparison of REE data from the Sokoman Iron Formation and Empire Mine shows variations in REE behavior between these time equivalent iron formations. This suggests that trace element imprints are not representative of overall sea water composition of the time. Rather, the chemical character is influenced by the processes of formation and types of precipitating phases. In addition, thorium values from clastic and chemical sediments do not reflect values of continental composition seen in other post-Archean sediments. Anomalous cerium behavior in cherts and minor fluxuations in REE trends, both vertically and laterally, suggest deposition close to the paleo-margin of a depositional basin. This is supported by isotope data of Haase. A facies control is not exhibited by NIF samples. This suggests that the major facies present at the Empire are a result of secondary processes and the original deposition of the NIF was homogeneous throughout. Empire oxides have LREE enriched values (NASC normalized) compared to the HREE enriched character of the Sokoman. It is suggested that the precipitation of carbonate phases is a controlling factor on REE behavior. Or, the presence of carbonate complexes in solution allows for the stabilization of HREE and precipitation of these phases. Hydrothermal alteration does not appear to be a major process in the formation of Empire ores in contrast to the hard ores of the NIF. Instead, it appears that magnetite enrichment is a product of diagenetic enrichment through the movement of volatiles. 85 Similarities of REE imprints of chemical sediments to clastic materials indicate that the erosion of continental material remains a plausible source for iron-rich solutions. However, a volcanic input cannot be ruled out without further data from possible volcanic associations. SUGGESTED FURTHER RESEARCH The use of REE as sedimentary indicators appears to be a valid approach to the understanding of the occurrence of banded iron formations. This investigation of the Negaunee has suggested several alternatives to the geochemical conditions of precipitation and alteration. Future research on both the NIF and associated strata may further aid in the evaluation of the occurrence of massive iron ore deposits during Precambrian time. These studies include: 1. Analysis of mineral separates versus whole rock analysis to place constraints on the effects of different mineral phases on REE distribution patterns. These data would aid in further modelling of processes associated with sedimentary systems including source material, crystallographic control, adsorption and diagenetic and post-diagenetic alteration. 2. A better definition of the mobility of REE during diagenetic processes and the relationship to the formation of new minerals using modern analogs. 3. A regional study to quantify trace metal and REE data for the Siamo Slate. Shales are relatively consistent within a single basin unless more than one source of clastic material is present as suggested in this investigation. These data would characterize detrital sedimentation occurring prior to NIF precipitation and give a broader data base for comparison to subsequent chemical precipitation. 86 87 Trace metal investigation of samples from the Empire to correlate to REE data. These data would further aid in evaluating processes of formations proposed in this study. A detailed trace element study of basement complex (i.e., Mona Schist and Compeau Creek Gneiss) for comparison to Clastic and chemical sedimentation. A regional study of the NIF to evaluate the mobility of REE during metamorphism and oxidation. REFERENCES Anderson, G. L., 1968. The Marquette District, Michigan, ore deposits of the United States, in Graton Sales, v.1, J. D. Ridge (ed.): AIME, NY, p. 507-517. Adamchuck, I. P.; Pachadzmanov, D. N.; Melniokova, N. D. and Valezev, Yu. Ya., 1979. The behavior of some rare elements in sedimentation processes, i_n Origin and distribution of the elements, L. H. Ahrens (ed): Pergamon Press, p. 310-352. Balashov, Yu. A.; Rohov, A. B.; Migdisov, A. A. and Turanskaya, N. V., 196‘}. The effect of climate and facies environment of the fractionation of the rare earths during sedimentation: Geochemistry Inter., p. 951-969. Bavinton, O. A. and Taylor, S. K., 1980. Rare earth element geochemistry of Archean metasedimentary rocks from Kambalda, Western Australia: Boyum, B. H., 1975. The Marquette mineral district of Michigan: Cleveland Cliffs Iron Co., in conjunction with the let Ann. Inst. Lake Superior Geol., 59p. Cannon, W. F. and Gair, J. E., 1970. A revision of stratigraphic nomenclature for Middle Precambrian rocks in northern Michigan: Geol. Soc. Amer. Bull., v. 81, p. 2843-28‘16. Cannon, W. F., 1973. High grade magnetite deposits at Republic, Michigan: Their bearing on the genesis of Marquette Range hard ore: 19th Ann. Inst. Lake Superior Geol., Madison, Wisconsin, May. Cannon, W. F.; Gair, J. E.; Klasner, J. S. and Boyum, B. M., 1975. Marquette Iron Range: let Ann. Inst. Lake Superior Geol., Field Guidebook, Fieldtrip 1+, p. 125-1714. Cannon, W. F., 1976. Hard ore of the Marquette Range: Econ. Geol., v. 71, p. 1012-1028. Chaudhuri, S. and Cullers, R. L., 1979. The distribution of rare earth elements in deeply buried gulf coast sediments: Chem. Geol., v. 24, p. 327-338. Cullers, R. L.; Yeh, L. T.; Chaudhuri, S. and Sambhudas, C. V., 1975‘. Rare earth element in Silurian pelitic schist from N. W. Maine: Geochim. Cosmochim. Acta, v. 38, p. 389400. Drever, J. 1., HM. Geochemical model for the origin of Precambrian banded iron formations: Geol. Soc. Amer. Bull., v. 85, p. 1099-1106. 88 89 Dypvik, H. and Brunfelt, A. O., 1976. Rare-earth elements in lower Paleozoic epicontinental and eugeosynclinal sediments from the Oslo and Trondheim regions: Sedimentology, v. 23, p. 363-378. Dypvik, M. and Brunfelt, A. O., 1979. Distribution of rare earth elements in some North Atlantic Kimmeridgian black shales. Nature, v. 278, p. 339-341. Eriksson, K. A. and Triswell, J. F., 1978. Geological processes and atmosphere evolution in the Precambrian, i_n Evolution of the Earth's Crust, D. H. Tarling (ed.): Academic Press, p. 219-228. Folk, R. L., 1962. Spectral subdivision of limestone types, i_n Classification of carbonate rocks, W. E. Ham (ed.): AAPG Memoir 1, p. 62-84. Fryer, B. 3., 1977a. Rare earth evidence in iron formations for changing Precambrian oxidation slates: Geochim. Cosmochim. Acta, v. 41, p. 361-367. Fryer, B. 3., 1977b. Trace element geochemistry of the Sokoman Iron Formation: Can. Jour. Earth Sci., v. 14, p. 1598-1610. Gair, J. F. and Thaden, R. E., 1968. Geology of the Marquette and Sands quadrangles, Marquette County, Michigan: U.S.G.S. Prof. Paper 397, 77p. Gair, J. E., 1975. Bedrock geology and ore deposits of the Palmer quadrangle, Marquette County, Michigan: U.S.G.S. Prof. Paper 769. Goldberg, E. A.; Koide, M.; Schmitt, R. A. and Smith, R. M., 1963. Rare earth distributions in the marine environments: Jour. Geophys. Res., v. 68, no. 14, p. 4209-4217. Goodwin, A., 1956. Facies relations in the Gunflint Iron Formation: Econ. Geol., v. 51, p. 565-566, 588-595. Goldich, S. S.; Nier, A. O.; Buadsgaard, M.; Hoffman, J. H. and Krueger, H. W., 1961. The Precambrian geology and geochronology of Minnesota: Minn. Geol. Surv. Bull 41. Graf, L. J., Jr., 1977. Rare earth elements as hydrothermal traces during the formation of massive sulfide deposits in volcanic rocks: Econ. Geol., v. 72, p. 527-548. Graf, L. J., Jr., 1978. Rare earth elements, iron formations and sea water: Geochim. Cosmochim. Acta., v. 42, p. 1845-1850. Haase, C. S., 1979. Metamorphic petrology of the Negaunee Iron Formation, Marquette District, Northern Michigan: Unpublished Ph.D. Thesis, Indiana Univ. Haase, C. S. and Rye, D. M., 1980. Stable isotope geochemistry of the Ne aunee Iron Formation (NIF), Marquette District, Michigan: Preliminary 6 1 C and 6 1 8O data from carbonates at the Empire Mine: EOS, v. 61, no. 17, p. 399. 90 Han, Tsu-Ming, 1962. Diagenetic replacement of ore of the Empire Mine of Northern Michigan and its effects on metallurgical concentration: 8th Ann. Inst. Lake Superior Geol., Houghton, Michigan, Michigan Coll. Mining and Technology, p. 7. Han, Tsu-Ming, 1978. Microstructures of magnetite as guides to its origin in some Precambrian iron formations: Fortschr. Miner., v. 56, p. 105-142. Han, Tsu-Ming, 1982. Iron formations of Precambrian age: hematite-magnetite relationships in some Proterozoic iron deposits - a microscopic observation, from Ore Genesis - The State of the Art, G. C. Amstatz, A. El. Goresy, G. Frenzel, C. Kluth, G. Moh, A. Warrschkuhn and R. A. Zimmerman (eds), p. 451-459. Haskin, M. A. and Hashun, L. A., 1966. Rare earth elements in European shales: a redetermination: Science, p. 507-509. Haskin, L. A.; Haskin, M. A.; Frey, F. A. and Wildman, T. R., 1968. Relative and absolute terrestrial abundances of the rare earths, i_n_ Origin and districution of the elements, L. H. Ahrens (ed.): Pergamon Press, p. 889-912. Haskin, S. P. and Taylor, S. R., 1974. Excess Europium content of Precambrian sedimentary rocks and continental evolution. Geochim. Cosmochim. Acta., V0 38, p. 739-746. Herrmann, A. J.; Blanchard, O. P.; Haskin, L. A.; Jacob, J. W.; Knoke, D.; Karolev, R. L. and Brannon, J. G., 1976. Major, minor and trace element composition of peridotite and basaltic komatiites from the Precambrian crust of South Africa: Contrib. Mineral. Petrol., v. 59, p. 1-12. Hooker, P. J.; Hamilton, P. J. and Onions, P. K., 1981. An estimate of the Nd composition of Iapetus seawater from Ca. 490 MA metalliferous sediments: EPSL, v. 56, p. 180-188. Jakes, P. and Taylor, S. R., 1974. Excess europium in Precambrian sedimentry rocks and continental evolution: Geochem. Cosmochim. Acta., v. 38, p0 739-7450 James, H. L., 1954. Sedimentary facies of iron formation: Econ. Geol., v. 49, p. 235-293. James, H. L., 1955. Zones of regional metamorphism in the Precambrian of northern Michigan: Geol. Soc. Amer. Bull., v. 66, p. 1455-1487. James, H. L., 1958. Stratigraphy of Pre-Keweenawan rocks in parts of northern Michigan: U.S.G.S. Prof. Paper 314-C, 44p. Jarvis, J. E.; Wildeman, T. R. and Banks, N. G., 1975. Rare earth elements in the Leadville Limestone and its marble derivatives: Chem. Geol., v. 16, p. 27-370 91 Klein, C., Jr., 1974. Greenalite, stilpnomelane, minnesotaite, crocidolite and carbonates in a very low-grade metamorphic Precambrian iron formation: Canadian Mineral., v. 12, p. 475-498. Lesher, C. M., 1978. Mineralogy and petrology of the Sokoman lron Formation ' near Andrea Lake, Quebec: Canadian Jour. Earth Sci., v. 15, p. 480-500. Mancuso, L. J.; Loughud, M. S. and Wygant, T., 1971. Possible biogenic structures from the Precambrian Negaunee (Iron) Formation, Marquette Range, Michigan: Amer. Jour. Sci., v. 271, p. 181-186. McLennan, S. M.; Fryer, B. J. and Young, G. M., 1979. Rare earth elements in Huronian (Lower Proterozoic) sedimentary rocks: composition and evolution of the post-Kemoran upper crust: Geochim. Cosmochim. Acta., v. 43, p. 375-388. McLennan, S. M.; Fryer, S. M. and Young, G. M., 1979. The geochemistry of the carbonate-rich Espanola Formation (Huronian) with emphasis on the rare earth elements: Can. Jour. Earth Sci., v. 16, p. 230-239. McLennan, S. M.; Nance, W. B. and Taylor, S. R., 1980. Rare earth element thorium correlations in sedimentary rocks and the composition of the continental crust: Geochem. Cosmochim. Acta., v. 44, p. 1833-1839. Mel'nik, Yu. P. and Lugoraya, I. P., 1972. The origin of ore minerals in Precambrian ferruginous quartzites from oxygen isotope evidence: Geochem. Internatl., v. 9, p. 808-817. Menzies, M.; Blanchard, D. and Seyfield, W., Jr., 1979. Experimental evidence of rare earth element mobility in greenstones: Nature, v. 252, p. 398-399. Morgan, J. W. and Wandless, G. A., 1980. Rare earth element distribution in some hydrothermal minerals: evidence for crystallographic control: Geochem. Cosmochim. Acta., v. 44, p. 973-980. Nance, W. B. and Taylor, S. R., 1976. Rare earth element patterns and crustal evolution - 1. Australian post-Archean sedimentary rocks: Geochem. Cosmochim. Acta., v. 40, p. 1539-1551. Parekh, P. P.; Moller, P.; Dulski, P. and Bausch, W. M., 1977. Distribution of trace elements between carbonate and non—carbonate phases of limestone: Earth Planet. Sci. Lett., v. 34, p. 39-50. Pesonen, L. J. and Halls, H. C., 1979. The paleomagnetism of Keweenawan dikes from Baraga and Marquette Counties, northern Michigan: Can. Jour. Earth Sci., v. 16, no. 11, p. 2136-2149. Piper, D. Z., 1976. Rare earth elements in the sedimentary cycle: a summary: Chemo G801” V. I“, p. 285-304. Reynolds, R. C., Jr., 1963. Matrix correlations in trace element analysis of x- ray fluorescence: estimation of the mass absorption coefficient by Compton scattering: Amer. Mineral., v. 48, p. 1133-1143. 92 Roaldset, E., 1973. Rare earth elements in Quaternary clays of the Numedal area, Southern Norway: Lithos, 6, p. 349-372. Ronov, A. B.; Migdisov, A. A. and Lobach-Zhuchenko, S. B., 1977. Regional metamorphism and sediment composition evolution: Geochem. Internatl., v. 14, p. 90-112. Ronov, A. B.; Balashov, Yu. A.; Girin, Yu. P. and Bratisko, R. Kh., 1972. Trends in rare earth distribution in the sedimentary shell and in the earth's crust: Geochem. Internatl., v. 9, p. 987-1016. Sandberg, P. A., 1975. New interpretations of Great Salt Lake ooids and of ancient non-skeletal carbonate mineralogy: Sedimentology, v. 22, p. 497-537. Schock, H. H., 1979. Distribution of rare-earth and other trace elements in magnetite: Chem. Geol., v. 26, p. 119-133. Scherer, M. and Seitz, H., 1980. Rare-earth element dristibution in Holocene and Pleistocene corals and their redistribution during diagenesis: Chem. Geol., v. 28, p. 279-289. Shimizu, H. and Masuda, A., 1977. Cerium in chert as an indication of marine environment of formation: Nature, v. 266, p. 346-348. Slaughter, E. L.; Tituskin, S. E. and Wilband, J. T., 1981. Preliminary assessment of rare earth element geochemistry of various iron formations of the Lake Superior District: 27th Ann. Inst. Lake Superior Geol., Abstr. w/programs. Taylor, S. R., 1964. Trace element abundances and the chondritic earth model: Geochim. Cosmochim. Acta., v. 28, p. 1989-1998. Trendal, A. F., 1968. Three great basins of Precambrian iron-formation deposition: a systematic comparison: Geol. Soc. Amer. Bull., v.79, p. 1527-1594. Tyler, S. A. and Twenhofel, W. H., 1952. Sedimentation and stratigraphy of the Huronian of Upper Michigan: Amer. Jour. Sci., v. 250, p.’1-27, 118-151. Van Hise, C. R.; Bayley, W. S. and Smyth, M. L., 1897. The Marquette iron- bearing district of Michigan: U.S.G.S. Monograph XXVIII, atlas, 608p. Van Hise, C. R. and Leith, C. K., 1911. The geology of the Lake Superior region: U.S.G.S. Monograph LII, 641p. Van Schmus, W. R., 1976. Early and Middle Proterozoic history of the Great Lakes area, North America; global tectonics in Proterozoic times: Royal Soc., London, Philo. Trans., A., v. 280, p. 605-628. Wilkenson, B. H., 1979. Biomineralization, paleoceanography, and the evolution of calcareous marine organisms: Geology, v. 7, p. 524-527. Wildeman, T. R. and Haskin, L. A., 1973. Rare earths in Precambrian sediments. Geochem. Cosmochim. Acta., v. 37, p. 419-438. APPENDIX A ANALYTICAL METHODS Analytical methods used in this study include neutron activation, electron microprobe analysis, and X-Ray fluorescence. Neutron Activation Analysis Neutron activation analysis was used to determine whole rock compositions for the rare earth elements La, Ce, Sm, Eu, Yb, and Lu and for Th and Cr. Samples of approximately 1.00000 gm were powdered to pass through a 200 mesh sieve. Powdered samples were placed in polyvinyl vials, sealed, and irradiated at the University Triga Mark 1 nuclear reactor for six hours. Samples were analyzed at the Michigan State University Department of Geology using a Geli detector coupled with a multichannel analyzer directly tied to the University computer system. Liquid standards and U.S.G.S. rock standards (W-l and AGV-l) were used for comparison. Liquid standards were prepared by transferring volumetric aliquots of a prepared liquid of known concentration to the polyvinyl sample holders and were allowed to evaporate to dryness by exposure to IR heat lamps. Electron Microprobe Analysis An ARL-EMX three spectrometer microprobe was used for individual phase analysis. This instrument is operated and maintained by the MSU Center for Electron Optics. A full-time technician is assigned to the probe for maintenance. The ARl is connected directly to the University computer system. LIF, RAP, and ADP detector crystals were used to measure Fe, Mg, and Ca, respectively. Dolomite and siderite standards were used for comparison. 93 94 APPENDIX A (Continued) X-Ray Fluorescence Total iron content was measured by X-ray fluorescence analysis on magnetite, carbonate, and chert-rich samples. A General Electric XRF X-ray generator and detector panel were used. Data were reduced using Compton's method (Reynolds, 1963) with U.S.G.S. standards BCR-l and PCC-l. All data are reported as percent FezO3. APPENDIX B ROCK SAMPLE DESCRIPTIONS SAMPLE NUMBER DESCRIPTION FACIES A/582 Carbonate with fine chert laminae, Carbonate minor magnetite veins A/517.5 Carbonate with fine chert laminae Carbonate B/278.5 Carbonate and chert mesobands Upper series alternating C/41 Carbonate-chert laminae Upper series alternating with chert-magnetite laminae C/87 Magnetite ore with intergranular Upper Series carbonate C/412.75 Clastics intergranular with chert Clastic facies C/ 582 Microbands of carbonate with chert Carbonate facies and magnetite cross cut by magnetite veins C/929 Magnetite ore with fine laminae of Carbonate facies carbonate C/1042 Magnetite ore with intergranular Silicate facies chert and alternating laminae of silicate C/l386 Alternating bands of carbonate Lower series and chert varying in thickness D/88.2 Chert with minor magnetite Upper series some altered to jasper D/126+ Carbonate laminae varying in Upper series thickness with chert laminae D/ 518 Massive magnetite ore Carbonate facies D/733.5 Magnetite ore, massive, with Carbonate facies minor chert and carbonate laminae D/752 Massive chert, minor carbonate laminae Carbonate facies 95 96 APPENDIX B (Continued) minor magnetite SAMPLE NUMBER DESCRIPTION FACIES D/ 878.5 Magnetite ore, massive with Carbonate-silicate carbonate—chert laminae facies change D/958 Chert with carbonate microbands Silicate facies D/ 1210 Carbonate with chert microbands Lower series E/70.5 Magnetite ore with mesobands of Carbonate facies carbonate and jasper E/595.5 Clastics with mesobands locally Silicate facies rich in magnetite and carbonate E/649.5 Clastics in magnetite ore matrix Silicate facies E/1126.5 Clastics intergranular with chert Lower series F/8 Carbonate and chert replacement Carbonate facies minor hematite and magnetite F /76 Chert containing microbands locally Carbonate facies rich in carbonate F/225 Chert containing microbands locally Carbonate facies rich in carbonate F/297 Chert with minor carbonate, Carbonate facies replacement features F/419 Magnetite ore containing clastic Silicate facies interbeds F/585.5 Alternating bands of magnetite with Silicate facies chert and clastics with chert F/695 Alternating bands of magnetite with Silicate facies minor carbonate and carbonate with minor magnetite F / 1119 Magnetite ore with some magnetite Lower series porphyoblasts G/285.5 Clastics interbedded with carbonate, Lower series 97 APPENDIX B (Continued) SAMPLE NUMBER DESCRIPTION FACIES G/578.5 Carbonate with micro- to mesobands of chert Lower series G/761.5 Magnetite with chert rich laminae Lower series G/845 Chert with thin laminae of carbonate Lower series Siamo transition G/923A Slate Siamo G/923B Slate with bands of quartz Siamo H/ 448.5 Carbonate, predominantly massive Lower series locally microbands rich in chert H/ 593 Arkosic sandstone, chert Lower series containing microbands of magnetite H/ 710.5 Alternating mesobands of chert and Lower series magnetite H/ 805.5 Chert with minor carbonate mesobands Lower series H/849.5 Magnetite ore with silicate mesobands Lower series Siamo transition H/ 865.5 Slate containing quartz porphyroblasts Siamo H/867.5 Slate Siamo H/906 Slate Siamo H/916 Slate Siamo APPENDIX C CARBONATE THIN SECTION DESCRIPTIONS SAMPLE NUMBER DESCRIPTION A/5l7.5 Fine crystalline, anhedral siderite in mesoband coexisting with anhedral quartz and trace euhedral magnetite, minor alteration to chlorite. A/582 Very fine crystalline anhedral siderite, coexisting with anhedral chert and euhedral magnetite; some alteration to hematite; well banded. B/27 8.5 Very fine crystalline, anhedral siderite coexisting with anhedral chert and magnetite, alternating with chert microbands. C/41 Very finely crystalline, anhedral siderite coexisting with chert and minor magnetite and finely crystalline anhedral ankerite with hematite and minor chert; cross-cut by fractures filled with magnetite and chert. C/1042 Fine to medium crystalline, anhedral ankerite coexisting with porphyroblastic magnetite; mesoband associated with ankerite chert with minor clastic microbands; cross-cut by fine magnetite filled veins. C/1386 Very fine to finely crystalline siderite, anhedral, minor subhedral, coexisting with anhedral chert and trace euhedral magnetite surrounded by carbonate phorphyroblast; cross-cut by veins filled with quartz and trace chlorite. D/126 Finely crystalline anhedral siderite with anhedral chert and trace magnetite; some chert-carbonate microbands; cross-cut by fractures filled with anhedral, finely crystalline ankerite. D/ 878.5 Massive magnetite ore; subhedral with mesobands containing finely to medium crystalline; anhedral to subhedral siderite; minor chert. D/1210 Finely crystalline anhedral ankerite coexisting with chert and euhedral magnetite associated with bands of chert and magnetite; fractures partially filled with euhedral ankerite. F/8 Finely crystalline anhedral siderite coexisting with chert in mesobands alternating with jasper and magnetite; minor alteration to hematite. 98 99 APPENDIX C (Continued) SAMPLE NUMBER DESCRIPTION G/578.5 Finely crystalline anhedral siderite coexisting with chert; some porphyroblasts; trace euhedral magnetite associated with chert; minor alteration to chlorite; cross—cut by quartz-filled fractures. H/448.5 Very fine, anhedral siderite mosaic with microbands chert; trace euhedral magnetite coexisting with chert; trace chlorite in carbonate rich laminae. H/805.5 Very finely crystalline anhedral siderite coexisting with chert and magnetite in mesobands associated with chert; cross-cut by fractures filled with medium grained anhedral chert. APPENDIX D THIN SECTION DESCRIPTION OF THE SIAMO SLATE SAMPLE NUMBER DESCRIPTION G/923A Moderately to well sorted quartz (2 5596) with trace feldspar in matrix of fibrous to platy biotite. Minor muscovite. G/923B Poorly to moderately well sorted quartz, minor cryptocrystalline (70-7596) possible mafic fragments, fibrous chlorite and biotite, rounded. Matrix fibrous chlorite and minor platy muscovite. H/ 865.5 Poorly sorted detrital quartz (z 70%), possible mafic clasts, consisting of chlorite and biotite. Matrix predominantly chlorite with minor biotite and muscovite, all fibrous to platy. H/ 867.5 Well rounded, well sorted quartz (2 55%) with trace detrital feldspar, possible mafic clastics consisting of chlorite and biotite rounded with serrated boundaries. Matrix predominantly chlorite bladed and fibrous. H/ 906 Well rounded, well sorted quartz (2 60%) with minor feldspar and trace rounded magnetite. Matrix platy chlorite with trace muscovite. Possible mafic clasts. H/9l6 Well rounded, well sorted quartz (2'- 5096) with trace of magnetite. Matrix platy chlorite and minor biotite. 100