unnv ‘6 "Shun" ..., ugh“. . 15;. , gym-1‘ a ‘ -.£‘. 13?"; v.3}? - A639; xylfi .‘ 5:1, ” ‘ ‘ gt» 1“ L‘ \'\ ":- wag“: ‘. L‘fifi‘fi. :12?“ *5 “~53?" “a r; 3 _ . “ :‘a .1; ‘ \ ~VS§ . '1‘ 33“,"; an?“ . ‘ PV 4?; '_ ’ :3. j' 3 ,7. > " $3.." 7;? :1! 7 it. ‘4‘ V {$.- : w (19" 5’?! IN 2‘ -. a $5.92 4* , 6» £31,3-: r!” 5.“. '1» - J r' ,. ~v£fihfixg7éxixfi ”45:53.33? "253?: .. 'tI’Vt 0 mg“ new I" :3: "m” . ‘V'fl rr—‘firr:;..rw Warm- « .:.. 4w "“ , .. Ni" ‘ u V'W’f“"> " I 4 ‘ ' ;£?&'.2.-'.".':.“""' Tia;- ‘-" » 5‘13"";E 1.x},— .. ’m ; ' J: :2, ,....-:;y,.,:.’: 1 ,..;~ m- ":J‘“ r ;~';.u5$.5}".'.- «Lafiggr‘nwgi .r w [MICHIGAN STATE UNIVERS Ill/Ill Ill/II/l/l/I/l/I I/l/Il/‘ll/lllllll 3 1293 00900 8826 ll This is to certify that the thesis entitled A STUDY OF THE FIRST ROW TRANSITION ELEMENTS AS A GROUP IN BLACK SHALE presented by David Jerome Piotrowski has been accepted towards fulfillment of the requirements for Masters degree in Geology am. #42? Major professor _ Date fl fmfli €l 0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution r’“ A a LIBRARY Michiflan State University "— A PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE | JL__J W? l MSU Ie An Affirmative AotiorVEquel Opportunity Institution cMMmG—pt STUDY OF THE FIRST ROW TRANSITION ELEMENTS ’ AS A GROUP IN BLACK SHALE BY David Jerome Piotrowski A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geological Sciences 1991 (557—4c3w ABSTRACT A STUDY OF THE EIRST ROI TRANSITION ELEMENTS AS A GROUP IN BLACK SEALS BY David J. Piotrowski The diverse response of the first row transition elements to various chemical conditions gives them the potential for recording chemical changes in geologic processes. This research examines whether there ‘are systematic trends in the relative abundances of the first row transition elements in black shale, and other sedimentary rocks. The techniques used are: sequential selective chemical extractions, data normalization to an average shale, and presentation on Coryell type plots. The results show: 1) some systematic trends in first row transition element abundances are observed in various sedimentary rocks, which can be interpreted based on FRTE chemistry, 2) the FRTE group patterns in the Devonian Antrim Shale suggest that a record is preserved, 3) stratigraphic differences in FRTE patterns are thought to be due mainly to detrital minerals, organic matter and sulfides within the shale, 4) FRTE pattern differences in the Antrim are influenced by more than detrital input. ACKNOWLEDGEMENTS I would like to express my appreciation to all those people who helped make the completion of this project possible. Hewever I would like give special thanks to my advisor, Dr. David T. Long for his comments, suggestions and continued support, to Dr. Thomas A. Vogel for all of the time and effort he put forth on the XRF analysis, and to Dr. Michael A. Velbel for his critical review of my manuscripts. Also, a special thanks to Dr. Duncan F. Sibley for filling in at my defense. I would also like to thank my fellow office mates, Joe McKee, Marcia Schulmeister, and Eric Roth for their continued suggestions and support. Finally, I would like to thank my wife, Jean Ann for her moral support throughout this project. iii TABLE OF CONTENTS Chapter One: Introduction.... ...... ......... ...... ........1 The Problem...........................................1 Goals.................................................5 General Controls on Element Distribution..............5 Rare Earth Elements-Controls on Behavior..............8 First Row Transition Elements-Controls on Behavior....9 Crystal Field Theory (CFT) .......... .... ............. 11 Evidence of Crystal Field Theory in Geologic Systems.16 Chapter Two: Examination of FRTE in Sediment Literature..19 Methods I............................................19 Results and Discussion I.............................21 Conclusion...........................................38 Chapter Three: Examination of FRTE in Black Shale........40 Goals II.............................................40 Rational for using Black Shale ........ ...............41 Rational for using Selective Chemical Extractions on Shale.............................................42 Selective Chemical Extractions.......................43 Effects of Sample Pretreatment.......................46 Chapter Four: Methods II.................................48 Antrim Shale (Description)...........................48 Sample Collection and Preparation....................50 Selective Chemical Extractions-Procedures and Analysis.............................................56 X-Ray Fluorescence-Procedures and Analysis...........62 x-Ray Fluorescence (Sample Preparation)..............68 Chapter Five: Results and Discussion II..................71 Bulk Shale........................... ..... ...........73 Metal Partitioning...................................75 Residual..........................................76 Oxidizable........................................83 Moderately Reducible..............................9o Easily Reducible..................................96 Weakly Acid Soluble...............................99 Summary/Discussion..................................103 Chapter Six: Conclusions................... ............. 106 Recommendations for Further Work....... ..... ........107 iv Appendix A: Shale Data........................ ...... ....110 Appendix B: Total Organic Carbon.......... ....... .......120 List Of ReferencesOOOOO...OOOOOOOOOOOOOOO00......0000000122 LIST OF TABLES TABLE 1: Test analysis: major element data for XRF glass wafer preparation (concentrations in weight percent)............................. ..... .65 TABLE 2: Test analysis: trace element data for XRF glass wafer preparation (concentrations in weight percent)...................................67 TABLE 3: Test analysis: trace element data for XRF pressed powder preparation (concentration in ppm)....67 TABLE 4: Bulk shale major element analysis (concentrations in weight percent)..................111 TABLE 5: Residual shale major element analysis (concentrations in weight percent)..................112 TABLE 6: FRTE abundances in various fractions of the Antrim Shale (concentrations in ppm).........113 TABLE 7: Total organic carbon (in weight percent) in the Antrim Shale...OOOOOOOOOOOOOOOOOIOOOOO...0.0.121 vi LIST OF FIGURES FIGURE 1: REE in various igneous rocks from the Batholith of Southern California ratioed to chondrite (Data from Towell et a1. 1965)..........3 FIGURE 2: FRTE in various igneous rocks from the Basistoppen Sill, East Greenland GPz=gabbro picrite, BGz-bronzite gabbro, PGz=pigeonite gabbro, FDz=fayalite diorite (Data from Naslund 1989).................. ............ 3 FIGURE 3: Spatial orientation of d-orbitals (taken fromfiuheey1983)0.0...OOOIOOOOOOOOOOOOOOOOOOO ..... .012 FIGURE 4: Energy difference of d-orbitals for an atom in octahedral coordination (taken from Bailar et a1. 1978)OOOOOOOCOOOOOOCOOOOCOO... ..... 0.0.12 FIGURE 5: D-orbital splitting for various types of atom-ligand coordination (taken from Burns 1970)OOOOOOOOOOCOOOOOOCOO0.0.0.0... ......... 00.15 FIGURE 6: REE abundance patterns in various Pleistocene Loess ratioed to chondrites (Data from McLennan 1983) ........... . ................ 22 FIGURE 7: FRTE abundance patterns in various Pleistocene Loess ratioed to chondrites (Data from McLennan 1983)............... ............. 23 FIGURE 8: REE and FRTE abundance patterns in major river sediments ratioed to chondrites (Data from Taylor and McLennan 1985).................24 FIGURE 9: REE and FRTE abundance patterns in Australian Phanerozoic and Proterozoic shales ratioed to chondrites (Data from Taylor and ’ McLennan 1985).......................................26 FIGURE 10: REE abundance patterns in greywackes of varying quartz content ratioed to chondrites (Data from Taylor.and McLennan 1985) PAAS=post Archean average shale, SiO contents (in weight percent): m277=56.35, m28 =60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65.................................... ..... 27 vii FIGURE 11: FRTE abundance patterns in greywackes of varying quartz content ratioed to chondrites (Data from Taylor and McLennan 1985) PAAS=post Archean average shale, Sio contents (in weight percent): m277=56.35, m28 =60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65.........................................28 FIGURE 12: REE abundance patterns in various Pleistocene Loess ratioed to PAAS (Data fromucmnnan1983)....OOOOOOOIOOOOO0.0.00.0...31 FIGURE 13: FRTE abundance patterns in various Pleistocene Loess ratioed to PAAS (Data from McLennan 1983).................... ........ 32 FIGURE 14: REE and FRTE abundance patterns in major river sediments ratioed to PAAS (Data from Taylor and McLennan 1985).................34 FIGURE 15: REE and FRTE abundance patterns in Australian Phanerozoic and Proterozoic shales ratioed to PAAS (Data from Taylor and McLennan 1985).......................................35 FIGURE 16: REE abundance patterns in greywackes of varying quartz content ratioed to PAAS (Data from Taylor and McLennan 1985) Sio contents (in weight percent): m277=56.35, m28§=60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65.............................36 FIGURE 17: FRTE abundance patterns in greywackes of varying quartz content ratioed to PAAS (Data from Taylor and McLennan 1985) 810 contents (in weight percent): m277=56.35, m28§=60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65.............................37 FIGURE 18: The Devonian Shale sample locations............51 FIGURE 19: Spatial sample locations and stratigraphic column showing the three units studied, the number of samples taken from each unit, and the maximum distance between samples.................52 FIGURE 20: Flow chart of sample analysis..................57 FIGURE 21: Method of analysis for FRTE from selective Chemical extractionSOOOIIOOOOOOOOOOOOOOOOOOOOOO0.000.61 viii FIGURE 22: Concentration of FRTE in Bulk Antrim Shale ratioed to Post-Archean Average Australian Shale for 3 stratigraphic units in the Paxton Quarry, Alpena County and samples from Livingston and Sanilac Counties........74 FIGURE 23: Concentration of FRTE in the Residual Fraction of Antrim Shale ratioed to Post-Archean Average Australian Shale for 3 stratigraphic units in the Paxton Quarry, Alpena County and samples from Livingston and Sanilac Counties........77 FIGURE 24: Percentage of FRTE in the Residual Fraction...78 FIGURE 25: Concentration of FRTE in the Oxidizable Fraction of Antrim Shale ratioed to Post-Archean Average Australian Shale for 3 stratigraphic units in the Paxton Quarry, Alpena County and samples from Livingston and Sanilac Counties........85 FIGURE 26: Percentage of FRTE in the Oxidizable FractiODOOOOOOOCCOOOOOOOOOOOOOOOOIOOOOOOOOOOOOO00.0.86 FIGURE 27: Concentration of FRTE in the Moderately Reducible Fraction of Antrim Shale ratioed to Post-Archean Average Australian Shale for 3 stratigraphic units in the Paxton Quarry, Alpena County and samples from Livingston and Sanilac Counties............................................91 FIGURE 28: Percentage of FRTE in the Moderately Redqu-ble FraCtionOOOOCOOOOOOOOOOOOOOOOOOOOOO0.00.0092 FIGURE 29: Concentration of FRTE in the Easily Reducible Fraction of Antrim Shale ratioed to Post-Archean Average Australian Shale for 3 stratigraphic units in the Paxton Quarry, Alpena County and samples from Livingston and Sanilac Counties............................................97 FIGURE 30: Percentage of FRTE in the Easily Reducible FractiODOOOOOOOOOOOOOOOOOOOOOOOOOOO00.0.00000000000098 FIGURE 31: Concentration of FRTE in the Weakly Acid Soluble Fraction of Antrim Shale ratioed to Post-Archean Average Australian Shale for 3 stratigraphic units in the Paxton Quarry, Alpena County and samples from Livingston and Sanilac Counties...........................................101 FIGURE 32: Percentage of FRTE in the Weakly Acid Soluble Fraction.......................... ......... 102 ix QEAZI§B_Q!§ IEIBQQQQIIQN This study examines whether there are systematic trends in the relative abundances of the first row ‘transition elements in black shale, and other sedimentary rocks. 233.2892LEE Trace elements in modern and ancient geologic systems are normally studied individually or as ratios (Piper, 1974). Sometimes the abundance of the element itself is important, due to its economic importance or because of its toxicity (Patterson et al., 1986; Patterson et al., 1988). Many times however, element abundances are used to make interpretations about geochemical processes related to the formation and. alteration. of rocks. and sediments (Piper, 1974). This approach can lead to jproblems in complex systems. For example, changes in metal abundances in ancient sediments are considered important in understanding the chemistry of ancient seawater (Holland, 1984). However, there are many processes which affect trace metal deposition, such as precipitation and coprecipitation on oxides, hydroxides, organics and sulfides (Holland, 1984). Thus far only total metal abundances have been examined, which combines the effects of all of these processes and may in turn mask important information. The examination of individual elements has been of limited use (Holland, 1984). The first row transition elements are commonly examined 2 individually in such situations, but may be more beneficial examined as a group, similar to rare earth element studies. The rare earth elements (REE) have been studied as a group in a variety of rock types to make interpretations about geochemical processes (Balashov, 1964: Taylor, 1962; Fryer, 1977: Graf, 1978: Taylor and McLennan, 1985). For example, Buma et al. (1971) described the evolution and differentiation of some New England granites based on the REES. Sedimentary provenance studies were done by Andre et al. (1986) using' REEs concluded. that the dominant rock source of a Belgian shale changed from mafic to felsic between the Cambrian and Ordovician. These interpretations are made based on a knowledge of geochemical processes, chemical controls on group element behavior, and most importantly, changes in relative element abundance patterns in various environmental systems. Changes in element abundance patterns throughout a suite of rocks may reflect changes in formation conditions Coryell et al. (1963). For example, figure 1 is a plot of the REE in mafic and felsic rocks Towell et al. (1965) and demonstrates element fractionation during the formation of these rocks. The abundance of low atomic number REES. increase as the rocks become more felsic. Their enrichment in the felsic end of the batholith is due to their slightly larger ionic size than the higher atomic number REE. This makes them less able to be included in the earlier stages of crystallization. BmMoMMMVSbuflmwwcambwna 1000 E- T l T T T T T T T T l T T T T T E (an: I :1 t .— .5 — -——gamxo — C'. 100 __ “'37 ---*x.\ ' - ‘ tonallte. . O E . ----- granodionte § :- E ‘ ~ , ’ ‘ iii-mi ---~ leucogranite : 0 ~ I T - ‘ \ -1 g r _ o. 10 a“ “a E ” : Q _ —1 Q- .— 1 "‘1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 La Ce Pr Nd Sm Eu Tb Dy Ho Tm Yb Lu Rare Earth Elements FIGURE 1: REE in various igneous rocks from the Batholith of Southern California ratioed to chondrite (Data from Towell et al. 1965) £bsEwaenSflLEfld£fleemhwd 100 __T T 7 T l T T T T T T_1 —A c: | l 001- 153 — ppm/ppm Chondrites O l 1E4 ”‘1 1 1 1 1 1 1 1 1 4 1" Sc Ti V Cr Mn Fe Co Ni Cu Zn First Flow Transition Elements FIGURE 2: FRTE in various igneous rocks from the Basistoppen Sill, East Greenland GPz=gabbro picrite, BGZ=bronzite gabbro, PGZ=pigeonite gabbro, FDZ=fayalite diorite (Data from Naslund 1989) 4 The first row transition elements (FRTE) have not been well investigated as a group, although there is some evidence to indicate that they may be useful (Wilson et al., 1988). Figure 2 is a plot of the FRTE in an Eocene, tholeiitic, layered intrusion (Naslund 1989). The samples plotted are from various zones in the sill, from early forming to late, and include a Gabbro Picrite Zone (GPZ) , two Bronzite Gabbro Zones (BGZ) , a Pigeonite Gabbro. Zone (PGZ), and a Fayalite Diorite Zone (FDZ). There are trends in the FRTE patterns as there were with the REES in figure 1. The FRTE patterns decrease in Cr and Ni and increase in Zn as the rocks become less mafic. Allegre and Michard (1974) demonstrated fractional crystallization in an igneous suite by interpreting similar FRTE patterns based on the geochemistry of the FRTE. Studies have not been done on the FRTE as a group, particularly on ancient sediments, because their chemistries are more individual and complex than the REE. In addition, their mobility is more strongly influenced by changes in geochemical conditions than the REE, and as a result, the number of controls on FRTE patterns are much greater. However, because of the FRTE responsiveness to geochemical changes, FRTE patterns in rocks may contain a more detailed record of environmental chemical changes through time. Unfortunately, this record may also have been altered by post-depositional processes. By examining how metals are partitioned in a sedimentary rock, the FRTE responsiveness to geochemical changes may be determined. . At the least insight could be gained into the extent of post-depositional influences. Determining the extent to which the FRTE patterns have been altered since deposition is the biggest obstacle in interpreting the FRTE group patterns. Therefore it must be determined if FRTE patterns are preserved well enough to make meaningful interpretations. QQALQ The major goals of this study are to: 1) Determine how well FRTE are preserved in ancient sediments by examining the repeatability of FRTE group patterns in the literature of sediments and sedimentary rocks 2) Examine this data to determine if FRTE patterns are interpretable on a global scale 3) Examine this data to. determine if they discern sample differences that the REE do not 4) Examine a black shale over a small area to determine if FRTE patterns are preserved on a local scale 5) Examine how the FRTE are partitioned in the shale so that differences in FRTE patterns can be better understood G RAL CONTROLS ON LE DI TRIS ON The distribution of elements within the earth's crust is a function of their abundance, chemistry, and the geologic processes operating within the earth (Rankama and Sahama, 1960). To understand, or begin to predict element distributions in the earth's crust, element abundances in rocks must be well understood. Element abundance data collected by F.W. Clarke from a large variety of rocks provided initial information necessary to study element geochemistry (Rankama and Sahama, 1960) . From this, some important aspects of element chemistry were described. Goldschmidt (1954) recognized the importance of a trace metal's ionic size and charge in explaining element abundances in minerals and igneous rocks. The majority of elements are present in trace quantities, and their distribution in igneous rocks depends on their ability to substitute for the more abundant elements in common minerals. Goldschmidt (1954) suggested a series of empirical substitution rules based on the ionic radius and charge of the atom. They are known as Goldschmidt’s rules and are reproduced here from Krauskopf (1979): 1) A minor element may substitute for a major element if the ionic radii do not differ by more than about 15%. 2) Ions whose charges differ by one unit may substitute readily for one another, provided their radii are similar. Substitution is generally slight when the charge difference is more than one unit. 3) Of two ions that occupy the same position in a crystal structure, the one that forms the stronger bonds with it's neighbors is the one with the smaller radius and higher charge (or both). These rules work best for "A" period elements in the periodic table, but are unreliable for the transition metals. For example: Goldschmidt’s rules predict well the substitution of Ba+2 (r=1.35A) and Rb+ (r=l.48A) for K+ (r=1.33A). Barium, because of it's +2 charge and small size, substitutes before rubidium in potassium. minerals (Krauskopf, 1979). However, Zn+2 (r=.74A) does not substitute for Mg+2 (r=.65A) or Fe+2 (r=.76A), even though Ni+2 (r=.72A) and Co+2 (r=.74A) do (Burns, 1970). Ringwood (1955) believed the deviation of many metals from Goldschmidt's rules resulted from the assumption that the bonding in the crystal lattice is purely ionic. Many bonds, such as those with transition metals, have combined ionic and covalent components. Ringwood believed that covalent bonds were weaker than ionic bonds, and that a mixed bond character would affect bond strength. A criteria was needed to predict bond strength differences between substituting elements. Idnus Pauling suggested that electronegativity would provide a viable method of comparing bond strength and hence bond character (Ringwood, 1955) . Large differences in electronegativities between a cation and anion, (such as those between an alkali or alkaline earth elements and the halides) produce predominantly ionic bonding (Douglas et al., 1983). Small electronegativity differences between a cation and anion suggest a covalent bond component. Ringwood (1955) suggested that the substituting cation that provided the largest difference in electronegativity with the surrounding anions in the crystal structure would be preferred. The addition of this rule improved the predictions for some metals, ’particularly the REE, but did not reliably predict transition element substitution. Variations in bonding energy due to distortion of the FRTE outer electron orbitals also needed to be taken into account. This is discussed in the section on crystal field theory. BAEE_EABIE_ELE!E!2§:Q§§§IQAL.§QNIBQL§ The rare earths are a group of elements from lanthanum to lutecium, (atomic no. 57-71) that are characterized by the filling of the f-orbitals. The ionic radii gradually decrease across the group (La r=1.15A to Lu r=.93A) due to the lanthanide contraction (Haskin, Frey, Schmitt, and Smith, 1966). The REEs have the same ionic charge, which is +3 across the group, (except for cerium which can be oxidized to +4, and europium which can be reduced to +2). Ionic size and charge are the major controls on REE chemistry (Haskin et al., 1968). Therefore, they partition somewhat. predictably in igneous rocks according to Goldschmidt-Ringwood’s rules. For example, in figure 1 the lighter larger radius REE concentrate in the felsic melts because they are too large to substitute for major elements earlier in the formation sequence. Consequently, the REE also partition similarly in nature, and are found together in comparable abundances in most environments (Haskin et al., 1966). Geochemically, the REE are classified as lithophilic, because they are found associated with silicates. The REE 9 are all relatively immobile, immune to redox changes, and therefore are not easily affected by post depositional processes (Haskin, Frey, Schmitt, and Smith, 1966). Taylor and McLennan (1985) demonstrated that part of the REE's usefulness as a geochemical recorder in sedimentary environments was due to their 1) a short residence time in seawater with respect to seawater mixing, and 2) a low partition ratio from sediment to seawater. This allows them to be deposited quantitatively in sedimentary environments from their source area. I T ON - CA ROL The first row transition elements (FRTE) are a group of metals including Sc (atomic no. 21), Ti, V, Cr, Mn, Fe, Co, Ni, Cu, and Zn (atomic no. 30). They are characterized by the filling of the d-orbitals Allegre and Michard (1974) . The d-electrons in these atoms are not shielded by further outlying electrons and are strongly influenced by neighboring ions (Douglas et al., 1983). These effects play an important role in controlling FRTE chemistry and are described by crystal field theory (Curtis, 1964). Ionic charge is variable across the group. vanadium, chromium, manganese, iron, and copper commonly change valence states in nature. The most common valance states across the group are: Sc+3, Ti+4, V+3,+4,+5, Cr+3,+6, Mn+2,+3, Fe+2,+3, Co+2,+3, Ni+2, Cu+l,+2, and Zn+2 Brookins (1988). Electron repulsion in the d-orbitals of some of the FRTE increase the ionic radii and as a result there is no 10 systematic size decrease across the group (Burns and Fyfe, 1967) . Therefore, the ionic radii are not an important unifying factor' in FRTE chemistry, but do affect their behavior. The FRTE chemistries are more individual and complex than the REE, and less well understood. Goldschmidt and Ringwood did not consider crystal field theory and consequently their rules are poor at predicting transition element partitioning in minerals and igneous rocks (Burns et al., 1967). The geochemical classification of the elements by Goldschmidt (1954) demonstrates the FRTE's range of geochemical affinities. Sc, Ti, V, Cr, Mn, and (Fe), are found in silicates (lithophile) , Co, and Ni accompany Fe (siderophile), and Cu, Zn and (Fe), readily form sulfides (chalcophile) . Many of the FRTE change valence states during changes in redox conditions which can affect their solubility and mobility. Although there are many more factors which affect the behavior of the FRTE in the environment than the REE, it may be that they still can prove useful. The FRTE individual sensitivities may allow them, as a group, to record subtle chemical changes. Virtually all of the FRTE fall within the guidelines Taylor and McLennan (1985) used to determine whether an element could be treated as quantitatively deposited in sediments from the source rocks. Taylor and McLennan (1985) recommend Sc and Co and suggest the rest (excluding Fe and Mn) may be helpful if treated with 11 caution. The sensitivity of Fe and Mn to changes in redox conditions makes them unreliable recorders of geochemical conditions. EBISIAL_ZIELD_IE§QBI_IQIIL The FRTE are the lowest atomic numbered elements with the potential to have electrons filling their d-orbitals. The outermost d-electrons are not shielded by further outlying electrons which makes them susceptible to external electrical influences (Burns, 1970). A transition element not influenced by outside electrical charges has 5 equal energy (degenerate) orbitals (Johnson, 1973). The orbitals, denoted dxy, dyz, dxz, dzz, dxz-yz, each have a unique spatial orientation (see Figure 3). When the atom becomes bonded to one or more anions, as in a crystalline solid, repulsion occurs. between their respective electrons (Burns, 1970). This makes it difficult for transition metal electrons to occupy d-orbitals oriented in or near the bond directions. D-orbitals oriented away from the bond directions require less energy to be occupied than orbitals oriented toward the bonding directions. These electron energies are respectively less and greater than those needed in an unbound transition metal. The energy difference between the d-orbitals is termed 10 Dq (Douglas et al., 1983) (see Figure 4). For an atom in octahedral coordination, the d-orbitals dxy, dyz, dxz are 4Dq lower in energy than the degenerate state, and the dzz, and dxz-y2 are 6Dq higher (Douglas et al., 1983). FIGURE 3: Spatial orientation of d-orbitals (taken from Huheey 1983) d orbitals in free metal ion Energy FIGURE 4: Energy difference of d-orbitals for an atom in octahedral coordination (taken from Bailar et al. ’1’ 500 I __.___._._ ’ 100w 1/ dufimkfifiw1\\ I from ligands was \ 4Dq t ’7’ symmetrical \\ 29 I ‘ _ __ _. ._ __ _ / d orbitals “split“ ' by octahedral field 1978) 13 Therefore, the energy saved by an atom by being bound to anions is a function of the energy saved by the electrons in the low energy orbitals minus the extra energy required by the electrons occupying the high energy orbitals. For example, Fe+2 in octahedral coordination has a crystal field stabilization energy of (4e- x 4Dq)-(2e-x6Dq)=4Dq. When energy is saved overall the transition metal "prefers" incorporation into a solid (or crystal). The greater the energy saved the greater partitioning in a solid (Burns, 1970). If no energy is saved the element will become enriched in the liquid. The energy saved by a FRTE in a crystal lattice is dependent on the transition element and its valence state, its coordination number in a crystal solid (octahedral, tetrahedral, cubic, etc.) , and the type of anions it is bonded to within the crystal structure. The d-orbitals of a metal can be filled one of two ways depending on the strength of the crystal field: 1) all d- orbitals are occupied by an electron before pairing begins (weak field-high spin case), or 2) pairing of electrons in low energy d-orbitals occur before electrons enter the high energy d-orbitals (strong field-low spin case) (Douglas et al., 1983) . Electrons need additional energy for both electron pairing, and entering the higher energy d-orbitals. The lower energy route changes depending on the d-orbital energy gap, which is a function of the crystal field strength. Factors which tend to increase crystal field 14 strength include; large radii transition metal cations with a high valence, and highly electrostatic, covalent anions (Burns, 1970). In a weak electrostatic field less energy is required to start filling the high energy d-orbitals than to pair the low energy d-orbitals (Johnson, 1973). In this case, the electron spin directions coincide, hence the weak field-high (electron) spin. In a strong electrostatic field, electron pairing in the low energy d-orbitals lrequires less energy than entering the high energy d- orbitals. Paired electrons have opposite spins, hence the strong field-low (electron) spin (Johnson, 1973). Other forms of coordination, such as tetrahedral and cubic, result in different d-orbital configurations and crystal field splitting (see Figure 5). The type of coordination a transition metal undergoes depends largely on its ionic size (Burns, 1970). Large ions are more readily able to accommodate six ligands than small ions. The ionic size of a FRTE is a function of its valence and the placement of electrons in the d-orbitals. FRTE have a larger than predicted ionic radii when electrons are in the high energy d-orbitals, because they repel the surrounding anion electrons to a greater degree than the low energy d- orbital electrons. Distortions from "normal" coordination can occur in stressed solids (Burns, 1970), or in cases where the d-orbitals cannot be filled symmetrically, such as (high spin octahedral) d1, d2, d4, d6, d7, and d9 (Johnson, 1973). This results in further deviation from degeneracy. 15 4A, 4,. _ . K A“, 1-' ’ die -yl- 4:! ,. .- do, - ' . d]. l —— .............. 4.. >5 00 3 "._-' S .‘._ . - . d” . d” J... J” x ' ...- 1:4)” d”, d” . N ~ 1 ......... d“, dz." dyz MK“ 4 d" a‘ square planar octahedral tetragonal monoclinic FIGURE 5: D-orbital splitting for various types of atom- ligand coordination (taken from Burns 1970) 16 The Jahn-Teller effect suggests that deviation from degeneracy will occur whenever additional stability will result (Burns, 1970). For example, Cu+2 in malachite (Cu2(OH)2CO3) distorts from octahedral coordination (Burns, 1970). Crystal Field Theory (CFT) has been applied to a variety of geologic systems with varying degrees of success. Aspects of CFT have been used to explain element (partitioning in igneous, sedimentary, and metamorphic environments (Curtis, 1964: Burns et al., 1966; Burns et al., 1967: Schwarcz, 1967: Allegre and Michard, 1974; McKenzie, 1975: Glasby, 1975: and Lewan et al., 1982). The CFT controls on transition. metal partitioning are best illustrated in element partitioning in minerals within igneous systems. Many unusual characteristics of spinel mineral structures have been explained in the context of CFT (Burns, 1970). CFT predictions on the order of transition element uptake during fractional crystallization correspond well to the element distributions found in the Skaergaard intrusion (Burns et al., 1967; Burns, 1970: Curtis, 1964; and Allegre and Michard, 1974). Explanations of transition metal behavior in sedimentary environments has been less successful. The major chemical processes involved in trace element migration include; leaching from a source rock, aqueous transport of the hydrated trace metal, and attachment to sediments via 17 precipitation, co-precipitation or adsorption. CFT appears to be relatively unimportant in explaining trace metal precipitation from the aqueous to solid phase. CFT predicts movement from the liquid to solid phases only if the relative crystal field stabilization energy of the metal is higher in the solid phase. However, the crystal field strength of hydrated transition metals in solution and precipitated transition metal oxides are often comparable. Also, no additional stability is gained by the change in transition metal ligand coordination from liquid to solid. However, there is evidence that once a transition element is adsorbed, its ability to become incorporated into the structure of the substrate is predicted by crystal field theory (McKenzie, 1975). In cases where transition metal valence states change, such as in Mn oxide nodules, crystal field stabilization energy (CFSE) can account for other transition metal enrichments. In addition, CFT has been useful in predicting the leachability of transition metals from a weathering substrate based on the CFSE of the metal. For example, CFT supports the observation that Of”, Ni+2 (high spin), and Co+3 (low spin) are concentrated in laterite deposits. However, this could also be due to the thermodynamic stability of these oxide structures (Burns, 1970). Metamorphic conditions are similar to sedimentary in that aqueous solutions carrying dissolved metals interact with unstable minerals. Similarly, many of the same 18 problems exist using CFT, particularly where CFSE differences are small and the unknown energy effects of thermodynamics are potentially large. The best situations for using CFT are when predicting which transition elements will be most resistant to reorganization as minerals are restructured (Burns, 1970). It must be stressed that in all environments, thermodynamics plays a large role in determining trace element partitioning, but almost no thermodynamic data exists for geologic systems. Therefore, because of this, CFT is still one of the only descriptive tools available, and despite its limitations, continues to be used. QEAETEB TWO IO TE LITERATURE EETEODS I The basis for using the REE as a group is the similarities of their chemical controls. Rare earth elements across the series differ essentially as a function of mass and ionic size (Haskin et al., 1968); therefore, changes in REE patterns can be interpreted in a straight forward manner. The FRTE have a variety of chemical controls, all of which are influenced by other factors. For example: crystal field stabilization energies for transition metals are dependent on their oxidation state, the type and coordination of the surrounding ligands, as well as kinetics and thermodynamics (Burns, 1970) . Because of the large number of factors which influence the accumulation of FRTE during sedimentary rock formation, the processes governing their behavior are difficult to discern (Taylor and McLennan, 1985). It is also unclear how the relative FRTE abundance patterns in sediment or sedimentary rock respond to local variations in chemical conditions. Therefore, data were examined to determine the extent to which reasonable interpretations could be made from FRTE patterns, and the extent to which original FRTE abundance patterns may be preserved. Data sets containing both FRTE and REE data were used so that their group patterns could be compared and the advantages and disadvantages of each contrasted. The data used in this study were taken from 19 20 McLennan et al. (1983) and Taylor and McLennan (1985) and focus on sediments and sedimentary rocks. Samples include: average shales, Proterozoic and Phanerozoic Australian shales, modern ‘river' sediments, loess. deposits and greywackes. Tfl§_!§£_Q£_QQBXELL_£LQI§ Initial studies which compared the abundances of elements with related chemistries used REE, and displayed the data as the log of the absolute element abundance (Coryell et al., 1963). The log of the element abundance was used to conveniently display values which often times vary by orders of magnitude. However the absolute element abundances are related to stellar nucleosynthetic processes, and decrease with increasing atomic number (Brownlow, 1979). The even numbered elements have greater cosmic abundances than the preceding odd numbered element (Brownlow, 1979) . As a result, a confusing zig-zag pattern is produced when a series of element abundances are plotted together, obscuring subtle trends between samples (Coryell et al., 1963). Coryell et al. (1963) removed the cosmic influence by log ratioing REE data to REE abundances in a bronzite ordinary chondrite, which allowed subtle pattern changes to be compared. Ordinary chondrites have since been replaced by carbonaceous chondrites as the standard of choice in systems concerned with element fractionation with respect to cosmic abundances because they are believed to be representative of 21 element abundances within the solar system (Haskin, Frey, Schmitt, and Smith, 1966). The FRTE data are ratioed. both to C1 carbonaceous chondrite (Mason, 1979) and average shale and presented on Coryell type plots. Average shale was included as a standard because it is considered a good estimate of the FRTE abundances in the upper continental crust based on previous igneous chemical analysis (Wedepohl, 1968: Taylor and McLennan, 1985) . Such a standard may be more useful than the C1 carbonaceous chondrite for looking at element fractionation within the upper continental crust. B£§2L1§_AND_DI§£!§§IQN_I The major feature of the chondrite ratioed data (Figures 6-11) is the similarity between the FRTE patterns in all of the examined rock and sediment data. The same is true for the REEs. Loess deposits and average shales have been considered good natural approximations of average trace element abundances in the upper continental crust (Wedepohl, 1968: Taylor et al., 1983). This is reflected by the similarity of REE and FRTE patterns for an average shale and various Pleistocene loess (Figures 6 & 7). The element abundance patterns for both groups are unique, but both are internally consistent. The FRTE and REE patterns in major river sediments (Figure 8) are both repeatable and mimic the PAAS pattern extremely well. This seems to suggest that the source rocks for all these river basins are similar and the sediments 22 Banks Peninsula, New Zealand 1000— T 1 l . 1 T T . v T——T T 100 l LLilill __i_i 1.l_LiLiL i ppm Loess/ppm Chondrites TI‘—T—_T ‘T'TTTTII T_"l— [If T! I ll _e O ‘ ' L .1 l l l l l l l 1 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements North America and China 1 000 [— 1 l T T T T l V T T T U T T T — PAAS : --- Kansas A ---- Kansas 8 ‘ --- Kansas C 100 ----- Nanking .u-—- [1 illLll l .s- ‘D- . H". -~‘~'-"-‘- our,",F-fl‘--"‘.‘." ........ ”r—“T_—T- r‘r‘rrrrr—r—r‘r'r T'T T _A O i J l l 1 l l l L l l Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements ppm Loess/ppm Chondrites l l l 4 La Ce Pr Nd Rhine Valley, Germany 8 1000 [— l . I T T T T T T j T T T T T T__‘ E F E c i’ _ o i" — PAAS ' 1 C Kaiserstuhlt 0 - Kaiserstuhl2 ‘ E Q. 100 [:- -: e .3. a 33 c : CD L 1 O i —J 1- _ g . g 10 LI 1 1 1 1 1 1 1 1"} l ’ 1 Sm Eu Gd Tb Dy Ho Er Yb Lu Flare Earth Elements La Ce Pr Nd FIGURE 6: REE abundance patterns in various Pleistocene Loess ratioed to chondrites (Data from McLennan 1983) ppm Loess/ppm Chondrites ppm Loess/ppm Chondrites ppm Loess/ppm Chondrile 100 10' 0.1 0.01 1E-3 100 10 0.1 0.01 1E-3 23 Banks Peninsula, New Zea/and {—- T 1 T T T T T T T T__] 1 p. ’_1 1 1 1 1 1 1 1 1 1 1—4 So Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements North America and China _ T T T T T T T T T T__ - PAAS — --- KansasA — ----- KansasB ~-—- KansasC " ’ ----- Nanking — \‘v/ A "—1 1 1 1 1 1 1 1 1 1 1 Se Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements Rhine Valley, Germany >_1 T T T T T T T T T —- PAAS - Kaiserstuhh — ..... Kaiserstuhl2 r— r """" —4 l 1 1 1 1 l 1 J 1 I V Cr Fe Co Ni Cu First Row Transition Elements Mn FIGURE 7: FRTE abundance patterns in various Pleistocene Loess ratioed to chondrites (Data from McLennan 1983) 24 ‘ luwmrflherSafimemm 1000 _l T T T T T T T T T T T T T T T_ w E E .‘P. — - 2 r — o _ .c CE) 100;- -:‘ Q E E O. P" _l \ y— .— E CL ._ _ c1 10 —1 1 1 1 1 1 1 1 1 1 1 1 1 L 1 1 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb LU' Rare Earth Elements TMHMVRWGTSflfiMMnfii 100 r A 1 1 r 1 T r T T T T__ w .9 ': 10-— -— 'o c o 1 _ __ .c O E __ _ Q 0.1 o. \ E 0.01— — o. o. 1E’3 — 1 1 1 1 1 1 1 1 — J l Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements FIGURE 8: REE and FRTE abundance patterns in major river sediments ratioed to chondrites (Data from Taylor and McLennan 1985) 25 were homogeneously mixed at the point of sample collection. It would suggest that redox conditions, pollution and climate differences for the river systems do not seem to significantly influence the chondrite ratioed FRTE pattern. This also seems unlikely, since Cu and Zn data were not reported due to the influence of anthropogenic sources (Taylor and McLennan, 1985) . FRTE patterns from shales (Figure 9) from various locations across Australia, throughout the Proterozoic and Phanerozoic, retain the familiar "W" pattern, but exhibit more pattern variability than any of the other sediment/sedimentary rock sample groups examined. The scatter in the FRTE patterns is either due to ‘variations in :metal availability, differences in chemical conditions, or the remobilization of the FRTE after sediment deposition. The REE pattern in Australian shales (Figure 9) are much less variable than the FRTE. The REE vary by less than .5 orders of magnitude, whereas the FRTE vary by more than 1 order of magitude. The REE patterns are as consistent as any of the other REE sediment/sedimentary rock sample groups examined, which illustrates their uniformity in rocks from widely distant locations and ages, and their resistance to remobilization after deposition (Taylor and McLennan, 1985). Taylor and McLennan (1985) have shown an increase in REE (particularly the light ones) in greywackes increasingly rich in quartz (Figure 10). The abundances of FRTE (Figure 11) , however, decrease in greywackes as they become more 26 Phanerozoic and Proterozoic Australian Shales fl} 1000 6 C O .C o E 100 Q. \ 2 (U .C 0) E. E 100 ’8' O 10 .C 0 1 E E E 0.1 E (D 0.01 S Q1E'3 E “E - — Phanerozoic “ ----- Proterozoic —1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1"“ La Ce Pr N SUTEU Gd Tb Dy Ho Er Yb ' Rare Earth Elements FhmmyonMCamthmgmmmk AuflhMEnSWmES r_T T I I I I I I I I T__ —— — Phanerozoic -— ----- anmbmflc '_1 1 1 1 1 1 1 1 1 1 1"“ Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements FIGURE 9: REE and FRTE abundance patterns in Australian Phanerozoic and Proterozoic shales ratioed to chondrites (Data from Taylor and McLennan 1985) 27 Quartz Poor Greywackes 1 OOO ::_ , . . - fl 1 T If f T ‘3: w E 1 52 : *- ': 3. :1 2 1 O 100 :3. *3 .C t: ; O C j a. 5 _. Q. 10 E— E \ ’- -1 E : 3 o. z“ 4 o. 7 J 1 1'— - L 1 1 1 1 1 1 1 L a I 1 1 1 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements Quart Intermediale Greywaclres 1000 E“ f T n r T I I I 7—5: m E : 2 » - .5 _ — PAAS _ c _ a O .c O 100: j E E : Q. - a -E g ‘ E Q- ” ., .:~ - T Q- " ‘ ’ ““3""3%:3‘35:;-...;‘;;;-; - — - 1O b.1 ‘ 1 1 1 1 1 1 1 1 1 1 1 4 1 1— La Ce Pr N Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements Quartz Rich Greywackes 1000 I:- T T T T T T T f T T T T T T__‘ a E 3 g - PAAS 1 c .... ka7 o 1: 9803 ‘ 0 10°: 1 g t : e E 3 E l ‘ :2. '— 1 0- l 10 f— L 1 1 1 l 1 1 1 ' 1 1“ Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements La Ce Pr Nd FIGURE 10: REE abundance patterns in greywackes of varying quartz content ratioed to chondrites (Data from Taylor and McLennan 1985) PAAS=post Archean average shale, sioz contents (in weight percent): m277=56.35, m285=60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65 28' Quart Poor Greywackes 100 :7 _ . . 1 1 T T 7 T .7 w L Q) l E 10 — “o c: .9.- 1 — O E L Q 0.1 -& E 001— O- ‘-\ Q. \‘L 1 E-3 1 1 1 1 1 1 1 1 1 1 1— Sc Ti V Cr Mn Fe Co Ni Cu Zn First Flow Transition Elements Quartz Intermediate Greywaclres 100 _ 1 . T 7 T T T T T T 1 __< 3 ——PN6 E 10— “o r: o 1_ .c 0 . E _ Q 0.1 Q . N. ' . E 001— . , a . - 1E'3 L1 1 1 1 1 1 1 1 1 1 ‘— Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements Quart Rich Greymclres 100 __i T T T T T T T I T T 8 -PN$ E 10 - , --- mk97 - '0 , c: O 1 _ .c O E _ Q 0.1 Q. \ E 001_ E 1E‘3 EX..- - 1 ' 1 1 l l l— “sewn o Cr Mn Fe co Ni Cu Zn First Row Transition Elements FIGURE 11: FRTE abundance patterns in greywackes of varying quartz content ratioed to chondrites (Data from Taylor and McLennan 1985) PAAS=post Archean average shale, sioz contents (in weight percent): m277=56.35, m285=60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65 29 quartz rich. These trends are supported by the enrichment of REE and the depletion of FRTE in increasingly felsic rocks. As the weathering of rocks in the greywacke source area proceeds, the less resistant mafic minerals are preferentially removed, leaving the greywacke enriched in quartz and other felsic minerals. Since the FRTE are more enriched in mafic rocks (Allegre and Michard, 1974) , they decrease in abundance in greywackes as weathering progresses. The REE are more enriched in felsic rocks (Towell et al., 1965), and therefore are more enriched in quartz rich greywackes (Figure 10). The FRTE abundances in loess, average shale, river sediments, and greywackes illustrate that the FRTE patterns, with the possible exception of Australian black shale data, are as repeatable as the REE patterns. However, the use of a chondrite standard for FRTE data may not provide the most informative diagrams. Plots of REE abundances in recent and ancient sediments (Figure 6-11) demonstrate that the REE are enriched in the upper continental crust between 10 and 100 times with respect to chondrites. There is no more than a 10 fold abundance difference between any of the REE in the data, which illustrates their similar behavior in different sedimentary environments. Also evident is the smooth fractionation across the group (except for europium). As a first approximation, the REE abundances in the upper continental crust, though enriched, particularly in the 30 light REEs, reflect relative chondrite abundances fairly well. However, the FRTE do not approximate chondrite abundances in any of the samples. The individuality of the FRTE chemistries are demonstrated by the varying element enrichments across the group when ratioed to chondrites. Titanium is enriched in average shales about 10 times while nickel is depleted by hundreds of times. The 1000 fold abundance difference between elements across the group masks individual element variation within a data set. The FRTE abundances may be more meaningful if they are ratioed to a standard with similar FRTE abundances. The concentration of the FRTE in upper continental crust is closely approximated by the average of compiled shale data. FRTE deviations from this standard should reflect the chemical processes which fractionate the FRTE within the upper continental crust. The FRTE and REE abundances from the previous samples are replotted ratioed to Post Archean Average Shale (PAAS) (Figure 12-17). The lack of REE fractionation within the earth’s upper crust is reflected in ‘the loess samples, (Figure 12), and illustrated by the tendency of the elements to plot near 1. The FRTE (Figure 13) however, are significantly fractionated. The FRTE abundance patterns in loess from various locations (New Zealand, North America, China, and Germany) are distinct. Assuming no influence of. diagenesis, this clearly demonstrates that the loesses were derived from different sources. The REE patterns do not 31 Ban/rs Peninsula, New Zealand 10 L— T fiT T T T T 7 T7 T T T-—1' «n a am 3 E It .. . bpz .J o. bp3 j E t ..... bp4 1 o. l bpS ‘l O- l "133"'ZZZZZ-Z““:.‘1=.=f=,"-~--. ............ \ 1 :- ~-'-35=““'” “”“T “v" i a ,t - 3 cu ” i o E _ -J —1 E t - Q. Q. 01 l ' ' 1 1 1 1 1 l 1 1 1 1 g 1 1 1 La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements North America and China 10 _T T 7 T T l T T T T T T T T T T_‘ CD E ~-- KansasA E :3: - ----- KansasB ~ (1- b ”‘ KansasC ‘ E F ----- Nanking '1 Q P- Q- -. “Mu--- ";.'.— ‘M—u m... \ 1 >— :T:':::"‘"" ’:'—.-:=' -‘~:.“_‘-'.,,‘o . 1 U) : ~"f.T’'..-~-~“""“""'”""""-~-..,..o..‘.'-‘--'.. 777777 .Q'fl.~'.‘l“..-..' " m — 1 ............................... . : a) : —4 O __ _ .J _ . E _ _ a. o. 0-1 b1 1 1 4 1 1 1 1 L 1 1 1 1 1 1 1—‘ La Ce Pr N Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements Rhine Valley, Germany 10 T 1 T T T T T T T T T fl T T T T m E E E F 3 o. C --- Kasennuhn : E _ ----- Kaiserstuhl2 _ Q. Q. \ 1 L- 1 8 C _.......,.‘ ...... H",-,-‘,'..‘» ~~~~~~~~~~~~~~~~~~~~ -‘.‘—"-"U‘"‘-¥'.”.".”.".“-'.' ....... :1 (D C ‘ r : O _ _ J h - E _ _ Q. o. 01 ”'1 1 1 1 1 1 1 1 1— Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements La Ce Pr Nd FIGURE 12: REE abundance patterns in various Pleistocene Loess ratioed to PAAS (Data from McLennan 1983) 32 Ban/rs Peninsula, New Zea/and 1 O _ I , 1 T r T T - . d m g 3 < .. [301 .1' E .” bp2 1 f ‘ bp3 ] E _ bp4 . Q. bps J O. m E 8 : “"”“‘~‘°:. 5:: 222 ‘~-"< 2'” '— ‘s‘ I - ”T ._‘\ 5‘ I —n| L fi“" I T‘:a\:l /’ E l \ t;,;:.a:7 & r 0' 1 ‘T— l 1 1 1 1 1 l l 1 l 1 Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements 1 North America and China 10 E1 T T T T T T T T T _ a) : 5 <1 I - - - Kansas A s at- — ----- Kansas 8 _ _ - Kansas C ‘ E r- ----- Nanking - 0. Cl. \ 1 -- -— <0 E .................. : 8 : 4:51 ................................................... : 1- ’ ‘Q. .I - 3 : : :19 f k ““.‘.-~.."AJ-' ‘_a~‘v‘-5.‘“1‘.‘ ......... .’_ ’1 :1 : ~\"‘o.:122.:"-‘ ;~ 1” O- I— ..... —1 Q. 0-1 ’_I 1 1 1 1 1 1 1 1 1 #— Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements Rhine Valley, Germany 10 L— 1 T T T T T T T T T T m ; 3 < " s < T « a E K; m _ L -~ gamut E ----- Kaiserstuhl2 a ‘1 L s 1 _: m : _ a : 1 o 1‘: ,x’ x ,. I —J ‘ ‘‘‘‘‘‘‘‘‘‘‘ ”Pv” "‘fw. ,:_'~' ““ a'v _ E I - . .1 Q. a l 0.1 f— 1 1 1 1 1 — Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements FIGURE 13: FRTE abundance patterns in various Pleistocene Loess ratioed to PAAS (Data from McLennan 1983) 33 differentiate between the various loess deposits as well as the FRTE. This is probably due to the homogeneity of the REE in the continental crust, or that the physical process of producing loess could not sufficiently fractionate them. It is important to note that the differentiation of FRTE patterns in loess was not as noticeable when ratioed to chondrites. This emphasizes the need to chose an appropriate standard. The individual character of modern river sediments, absent when ratioed to chondrites, is clearly demonstrated when ratioed to PAAS (Figure 14). The Ganges has consistently low FRTE abundances with respect to the other river data, while the Amazon FRTE pattern is very close to average shale. HDwever there is no underlying commonality between the FRTE patterns of major rivers, as one might expect. The FRTE pattern differences are probably due to differences in provenance, weathering rate, rock exposure time, chemical characteristics of the water, and anthropogenic inputs. The REE patterns are less scattered than the FRTE which again demonstrates their consistency under many conditions and over large areas. In addition, the plot of Australian shales (Figure 15), shows relatively little REE pattern scatter, even over a large time period. Their consistency and resistance to remobilization after deposition, even under varying chemical conditions over long periods of geologic time is what makes the REES useful for studying long term changes in the composition of the upper 34 IMQMVHWerSmfiMemm 10 _I I I T I I I I I I I I I I I ___‘ E D Amumon E ~ * Congo ~ U3 ” O Ganges : E T Garonne 0. - O * Mekong ~ ,Q~- Q- -- 8"8 E 1 — 6:=_B“‘ :r"":;'=::‘:.0’;“’: ....... ’;,_‘:;: _‘--_‘ Q r; *’ O “-r-:::’ ’1‘ * : O. I 2 \ .- .4 E " -1 C1. .— _J a — 0-1 ”'1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1'— La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Rare Earth Elements MMMVRWeISmfimems I I I I I I I I I I T— 10 : -—- Aunazon E __ ..... COngO _ I * Ganges : (D _ ----- Garonne - E _ m Mekong ................ _ O. "‘- ;I;‘..r"":— -—-\.‘,;.:--;E':v..‘ ....... I ‘04.’ .~ I”* g 1 __ ><:: ,1 ,,,,, ,. ' m»; _: E' : * ‘.’ * * E 8: _ _ L- - 0-1 I"1 1 1 1 1 1 1 1 1 1 1— Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements FIGURE 14: REE and FRTE abundance patterns (Data from sediments ratioed to PAAS McLennan 1985) in major river Taylor and 35 Phanerozoic and Proterozoic - Australian Shales 1 00 L I I I I I I I I I I I I I I I a) E ' a :E c . « a t —— Phanerozmc .1 E 10 S“ Proterozoic ’5: Q. 1- J E- : 1 i _ m -C 1 S‘ 'E (D E a E l’ : o. i _ o. 0-1 4 1 1 1 L 1 4 1 1 L 1 1 1 1 1 1—T La Ce F% Nd Sm Eu Gd Tb Dy Ho Er Yb Rare Earth Elements IMmmmmmmbandPhMmuawb AuathnShmhs 1 O _ I T I I I T I I I I I_ D. ” _ E 1 E— —;:_ 8- E a a : 1 E . .2 0,1 5— — Phanerozonc .5 (D E ----- Proterozoic E E _ : CL _ _ Q . 0-01 _1 1 1 1 1 1 1 1 1 1— 1 SC Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements FIGURE 15: REE and FRTE abundance patterns in Australian Phanerozoic and Proterozoic shales ratioed to PAAS (Data from Taylor and McLennan 1985) 36 Quartz Poor Greywackes 10 E:— T T Y T 7* . . 1 .' Y ' I I? : :1 (2 I: - m277 3 < L. "1285 0. I {.2 l o. 1 :— , .............. I Q. : I -------- .. . -‘ \ b— E t , j {1 _ Q L _--- i 01 t‘ ' l l l J l l 1 l 1 L l 1 l 1‘} La Ce Pr N Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements Quartz Intermediate Greywaclres 10 T I I I I T I I I I T I I— w E "-4mm3 I g L ..... mk64 0_ ' ~— t82324 ‘ E 1L. ',-----------------—---;: ----- — Q E a‘-”-,-.---o---"‘.‘.f.'-~~~.‘..‘§,.‘,'_‘_3;.~.--~ —---'.; x' E g It - ~~--~---- : o. C I o. 1_ _ l 0-1 L1 1 1 1 1 1 1 1 1 1 1 1 1 1"" La Ce Pr N Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements Quartz Rich Gremckes 10 E? I I I I I I I I I I I 1 1 I? E «- mk97 : g; ; * p39803 - a l . * E 1 L- i"“ I, ........ ~:x’"“‘"‘»~----_--- -_-.* #1 C]. E * t : .2 e : E C q o. .— _ Q. 1__ .1 0-1 L”. 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1— La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Rare Earth Elements FIGURE 16: REE abundance patterns in greywackes of varying quartz content ratioed to PAAS (Data from Taylor and McLennan 1985) Sioz contents (in weight percent): m277=56.35, m285=60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65 37 Quartz Poor Greywaclres 1 f: I I I I I I I I . _3 O : 3 C - m277 '1 g; C rn285 I l o- r ~ , -------- i E L a 1: a Q ; 3 \ .— I —¢ E " “ .II I ‘1 Q- '_ t l I O- 1 L— ' 1 I 1 1 1 1 _j' Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements Qua/a Intermediate Greywackes rfi r I T f I I r I 7 j 10 E :5. I * p40136 s U) .. ..... mk64 .1 :E r -"'t82324 1 a h- n: E .. ....................... [Ax _ 0. 1 F ‘\ ,.. at I ‘x 3; Q >- \‘t'_--’ \ I *\ .1 \ >— ' \\ [I “‘._.‘* ’1‘, : E : \‘V/ "¥;'.;;‘ wfi/I’ ,1 o. P a o. i— --i 01 '— 1 . 1 1 1 1 1 1 1 1 1'.‘ So Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements Quartz Rich Gremckes TI I I T I I I I I I I — 10 E - a U) - mk97 : i 1- p39803 _ Q E 1 1 Q. a 2 E . ...... ‘s * * v *“ [I : E “‘ \\\\\ “\ ’I -- ‘ \‘ ’ —4 Q. “‘ ' ’ “‘~‘ 1’ .4 o. '1. ~ _ 0'1 LL-‘ I l 1 L l L L l 1 1—1 Sc Ti V Cr Mn Fe Co Ni Cu Zn First Row Transition Elements FIGURE 17: FRTE abundance patterns in greywackes of varying quartz content ratioed to PAAS (Data from Taylor and McLennan 1985) $102 contents (in weight percent): m277=56.35, m285=60.78, p40136=71.08, mk64=68.28, t82324=67.5, mk97=81.13, p39803=75.65 38 continental crust. The difference between. FRTE and REE pattern scatter' is :more clearly seen, in ‘the .Australian shales ratioed to PAAS. The FRTE are more easily remobilized after deposition than the REE, and their original depositional concentrations have probably been altered. The REE patterns in greywackes (Figure 16) increase in concentration with increasing quartz content. FRTE patterns in greywackes (Figure 17) again show gradual depletion with increasing quartz, except for Cr and Ni. Since Cr and Ni are the most depleted in felsic and enriched in. mafic igneous rocks, it was suspected that Cr and Ni would decrease in greywackes with increasing quartz (felsic) content. However this may be due to the age and sources of the various greywackes. QQEQLEQIQHE The conclusions from examining data in the literature are as follows: 1) The FRTE abundance patterns ratioed to PAAS show better distinction between samples than when ratioed to chondrites, and therefore PAAS is the better standard for examining FRTE abundance patterns in sediments and sedimentary rocks. 2) Greywacke data showed that, in general, the FRTE abundances decreased with increasing quartz content, while the REE abundances increased with increasing quartz content. This is probably related to the partitioning of the elements 39 in. the parent igneous rocks, but also) demonstrates the ability of the FRTE to provide information about ancient sediments. 3) The FRTE clearly distinguished between Pleistocene loess from different localities, whereas the REE did not. This shows that the FRTE can, at times, provide more subtle sample distinctions than the REE. 4) The FRTE abundance patterns in the loess and greywacke samples appear to be related more strongly to the sample's provenance and physical transport, than to chemical influences during deposition. 5) Proterozoic and Phanerozoic Australian shales are relatively consistent for the REE, but vary greatly for the FRTE. This demonstrates the usefulness of the immobility of the REE after deposition, in that globally collected samples over significantly large portions of geologic time can be compared. The FRTE abundance pattern is too scattered to make interpretations without more information. QEABIEB_I§B§§ F 0' 8 E IN BLACK §§AL§ EQAL§_11 Since the response of FRTE abundances in sediments to changing chemical conditions is unknown, and the influence of post-depositional processes are not readily understood, this portion of the study examines FRTE patterns in a black shale over an area of 0.2 km2 and a depth of about 7m. It is assumed that on this scale the post depositional history of samples should be identical. The goals of this portion of the study are to: 1) examine the FRTE abundances within individual stratigraphic units of a black shale over a small area to determine if FRTE patterns are repeatable 2) examine the FRTE patterns within various black shale stratigraphic units to determine if there are variations over time which can be explained by chemical or physical changes 3) examine predefined shale components to determine their relative importance in influencing the overall FRTE pattern within the shale 4) examine FRTE abundances in predefined shale components within individual stratigraphic units to determine if FRTE patterns are repeatable 4O 41 5) examine the FRTE patterns in predefined shale compOnents to determine if there are variations over time which can be explained by local chemical or physical changes W Black shales are a good starting point for examining FRTE patterns both from economic and scientific standpoints. Heavy metals are known to become concentrated in black shales (Vine and Tourtelot, 1970) . For example, uranium (Bell, 1978) vanadium, (Patterson, 1986), copper (Tourtelot, 1979), and silver (Vine, 1984) can be present in economically important concentrations. Also, the organic component of some black shales is considered for potential oil extraction. IHowever, in this case, enriched trace metals such as As and Cr can exist in toxic quantities and their mobility during and after oil processing is an important concern (Patterson et al., 1986; Patterson et al. 1988) . The elevated abundance of trace metals in black shales has led to the suggestion that shales may have been a source of metals for the formation of low temperature ore deposits (Long and Angino, 1982) . Many current studies interested in this idea have analyzed many of the FRTE (Ripley et al., 1989; Patterson et al., 1986; Patterson et al., 1988). Shales in general comprise 60% of all sedimentary rocks (Potter et al., 1980), and are relatively common through out Phanerozoic time (Tourtelot, 1979). Work by Wedepohl (1968), and Taylor and McLennan (1985) on sedimentary rocks 42 has demonstrated that important information can be obtained through the examination of trace metals in shale. One important reason is that fine grained sediments are more likely to be homogeneously mixed than coarser grained sediments (Wronkiewicz and Condie, 1987). This reduces the difficulty of collecting representative samples. The geographic dispersion of shales throughout Phanerozoic time has led Tourtelot (1979) to suggest that shale formation is controlled by geologic processes rather than geologic setting. If shale deposition is dependent on. geologic processes, the examination of trace element abundances is a good way to gain insight into the geochemical processes involved during shale deposition (Piper, 1974). a. 00., 1°; I: t 3'- 3 "~ ‘51-. . :91; ,ON: - QELLE The FRTE content of any black shale is a sum of metals added from detrital input, and hydromorphic input, that is, metals accumulated while the sediment is exposed to the water column. ~The hydromorphic input is the result of the combined inputs of carbonates, authigenic minerals such as Mn and Fe oxides, clays, organic matter, and sulfides, each of which can accumulate heavy metals including FRTE through precipitation, co-precipitation, or adsorption (Vine and Tourtelot, 1970; Wedepohl, 1971; Tardy, 1975; and Spears et al., 1981). Each one of these fractions are affected differently by various chemical conditions and may contribute unique information about the depositional history 43 of a black shale. It is also possible that the detrital fraction may record different but important information. At the least a separate examination of FRTE abundances in the detrital and hydromorphic fractions may provide clearer information about detrital and oceanic inputs. This may be accomplished through the use of sequential selective chemical extractions. EELlQII!I_£E££I£LL_§ZIBLQIIQ!§ Sequential selective chemical extractions (SCE) are a method to remove metals bonded to a particular compositional fraction of a sediment or shale through the use of chemical agents. Commonly these "fractions", (such as carbonates, Mn or Fe oxides or organic matter) will release metals under particular chemical conditions (Jackson, 1958). For example, organic matter is preserved in sediments during reducing chemical conditions, therefore metals bound to the organics can be extracted using chemicals that produce oxidizing conditions (Rezabek, 1988). Mn-oxides form during oxidizing conditions, and therefore can be destroyed, and the associated metals removed, by using chemicals that produce slightly reducing conditions (Chester and Hughes, 1967). Often metals from 'many different fractions of sediment need to be retrieved, which is done by using gentle chemical reagents to remove metals which are weakly bonded, removing the leachate, and repeating the chemical treatments with stronger reagents to remove more tightly bonded metals (Rezabek, 1988). 44 One of the first selective chemical attacks was designed for soils and was relatively simple. Jackson (1958) used ammonium acetate (NH4OAc) to remove metal cations bonded to exchangeable sites on clays to help address the availability of trace metals needed for plant growth. Other chemical attacks were designed for studying suspended sediments in rivers (Gibbs, 1977) , near shore sediments (Martin, 1987) , pelagic sediments (Chester and Hughes, 1967; and Gupta and Chen, 1975) and shales (Patterson et al., 1986; and Patterson et al., 1988). These later studies used a series of reagents, examined more sediment fractions, and became increasingly more complex. The "ideal" series of attacks is one in which each chemical extraction completely removes of all metals from the targeted sediment fraction without removing, or repartitioning metals from other fractions of the sediment. However, there is evidence that the chemical leaches are not completely selective, and metals may be released from other sediment fractions (Rendell et al., 1980: Tipping et al., 1985; Kheboian et al., 1987; Rezabek, 1988). Metals can be removed from the desired sediment fraction, but then be reabsorbed, which underestimates the metal concentration in that particular phase and overestimates it in the remaining fractions (Rendell et al., 1980; Nirel et al., 1986; Kheboian et al., 1987). In addition, Kheboian et al. (1987) found that extraction distributions for Cu varied between two compositionally identical artificial sediments. The 45 only difference between the two sediments was the amount and distribution of other metals within the sediment fractions. This suggests that the abundances of other trace metals, and how' they are jpartitioned ‘within. the sediment, may also affect the leached distribution of an individual element (Kheboian et al., 1987). Because the SCE are not 100% selective, the leaches are operationally defined rather than specifically defined to the targeted shale fraction. For example, metals targeted to be removed from the Mn-oxide fraction of a sediment by using a weakly reducing leach are termed the weakly reducible fraction. In general the FRTE respond fairly well to the SCE. Martin et al. (1987) performed a leaching experiment using Tessiers et a1. (1979) method on an artificial substrate doped with. metals. They showed the variation in the percentage of metals removed for a number of attacks. However, of the FRTEs tested, chromium, cobalt, iron', and scandium, only iron in the Mn-Fe oxide attack was considered insufficiently removed (20-60% may have remained behind). In addition, Tessier et al. (1979) demonstrated reproducible results for transition elements Mn-Zn with a relative standard deviation of about 110%. Nirel et al. (1986) obtained the same degree of precision using artificial sediments. The above discussion has referred to sediments rather than shales because there have been very few studies which have used SCE on shales (Patterson et al., 1986; Patterson 46 et al., 1988), and no studies which have analyzed potential problems using SCE with ancient sediments. Therefore, concerns about the inaccuracies of SCE can only be extrapolated to black shales, although it is assumed they will react similarly. IZIE9I§_QI_§AH2L£_BBBIBBLIHEEI In addition to the concerns regarding the accuracy of the selective chemical attacks, the handling of the sediments before leaching also plays a role in the reliability of the results. Ideally samples should be treated soon after collection. However this is not generally practical and. a variety of pre-treatment and storage methods have been devised. For sediments, they include; keeping the sample wet, or drying the sample by either freeze drying, or oven drying at various temperatures. The samples are then stored at room temperature, or refrigerated, either under normal atmospheric conditions or in an oxygen free environment. In addition, many researchers recommend carrying out the selective chemical attacks under an oxygen free atmosphere, particularly for anoxic sediments (Rapin et al., 1986). However, the chemical balance of the samples is easily disturbed and some repartitioning will occur under any of these storage methods (Rapin et al., 1986; Kersten and Forstner, 1987). Shale sample collection. is 1much simpler, as it is collected dry and more or less at room temperature. 47 Therefore the problems of metal repartitioning within the sediment due to heat drying or freeze drying are avoided. In addition, samples from outcrops have been exposed to oxygen from the atmosphere for long periods, and storage in anoxic conditions is probably not necessary. HDwever, the necessity of running the SCE under anoxic conditions is not known, although perhaps the accuracy of the SCE could be improved under anoxic conditions. The studies done to test the reliability of selective chemical extractions have pointed out many problems, from sample preparation and storage, to the leaches themselves. Many argue the leaches do not completely remove metals from the desired fractions, leaving some behind due to readsorption, precipitation, and co-precipitation. However SCE are still a useful tool for' probing' into ‘multiple aspects of sediment chemistry as long as the limitations are acknowledged. QEAEI§£_ZQEB EEEEQD§_11 AEIBIH.§EALE.LQ§§££LEELQQL The Antrim Shale in the Michigan Basin is part of a large group of shales deposited over the mid-continent during the late Devonian. It is correlatable with the Ohio Shale, the New Albany Shale, and the Chattanooga Shale. The Antrim is underlain by the Traverse Group, a series of interbedded fossiliferous marine limestones and calcareous greyish-black shales. The shale is overlain in the eastern part of the Michigan Basin by the Bedford Shale. Above the Bedford lies the Berea Limestone followed by the Sunbury Shale. The Antrim grades into the lower Ellsworth Shale to the west which causes difficulties in correlation (Ells, 1979). The Bedford Shale thins to the west and either merges with or overlies the Ellsworth towards the western side of the Michigan Basin (Ells, 1979). The Antrim Shale approaches a thickness of 197 m towards the center of the basin (Ells, 1979) but thins to 30 m or less at the basin margin. The upper and lower Antrim contacts are difficult to determine in the field. However, the Antrim-Bedford contact can be distinguished mineralogically (Routsala, 1980), and by gamma ray logs for the Antrim-Traverse Formation contact (Ells, 1979). Routsala (1980) determined the mineralogy of the Antrim to be: 50-60% quartz, 20-30% illite, 5-10% kaolinite, 0-5% chlorite and 0-5% pyrite. The quartz content is high 48 49 compared to the average shale value of 30% given by Weaver and Shaw (1965) . Hathon (1979) proposes that some of the quartz is authigenic, not detrital, based on the high percentage of small (<25 um) quartz grains and their smooth abrasion-free surface morphology. Pyrite occurs in thin bands along bedding, as lenses or spherical concretions. The shale fractures along bedding when weathered, but breaks conchoidally when fresh. Carbonate concretions 10 cm-1.5 m in diameter are present in the lower sections of the Antrim. The average organic content ranges from 3-14% (Leddy et al., 1980) and is believed to be of marine origin (Cross and Bordner, 1980) . The .Antrim.‘was. deposited. in. an anoxic, low' energy environment. Green shales become more common near the basin margins, suggesting slightly more oxic conditions. Whether the water depth was deep or shallow has been debated, but many' agree that reducing' conditions ‘were. present" The shale' s organic material was derived mainly from marine organisms, although terrestrial input increases toward the margin of the basin (Cross and Bordner, 1980). There were sources of sediments into the basin frmm the northeast and northwest. Tectonics and subsidences ultimately controlled the influences that these two sources had on the shale composition (McGregor, 1954). This perhaps explains why the Ellsworth and Antrim contact is so difficult to determine. Authors have .suggested that the basin was a shallow restricted sea, or a deep basin (Cross and Bordner, 1980; 50 Wardlaw, 1981; and Gutschick, 1987). The burial history, extent of diagenesis, and the temperatures reached are still debated (Vugrinovich, 1988) . Wardlaw (1981) suggests an Antrim burial temperature of 83'C at 2500m based on oxygen isotopes. However, if the Antrim is a source of natural gas, as is suspected, a temp of 150’C is needed unless gas is biogenetically produced (McGregor, 1954). W The samples for this study were obtained from three locations around the eastern rim of the Michigan Basin; Alpena County, Livingston County, and Sanilac County (Figure 18). Seventeen of the twenty samples used in this study were collected from the Paxton quarry (N 1/2 Sec. 30 T31N, R7E), Alpena County. Sampling objectives for the Paxton quarry were, 1) to obtain shale specimens from the same stratigraphic unit across the quarry to determine FRTE abundance variations due to location, 2) to obtain shale specimens from the same geographic location in different stratigraphic units to determine FRTE abundance variations at various times of shale deposition. The first objective was accomplished by collecting samples from three stratigraphically distinct units of the Antrim throughout the quarry (see Figure 19) . These units are referred to by Ells (1979) as unit 1-A, 1-B, 1-C, and will be referred to in this study as A, B, and C. The B unit differs from units A and C in that it is a light grey 51 Alpena I 8 County . Sanilac County Livingston County FIGURE 18: The Devonian Shale sample locations 52 A7,B7 C7 Bz,c2,02 c1,31 N A6 Scale 36 lcm = llOm A3 C4 B4 BS,C5 B3 BB,C8 Meters 'r25 Cé\ J ~20 A Unit 3 samples :]——370 m.—l “- 15 = a 8 Unit figig 7 samples lgug :}———-580 WL-—————% ._ 10 C 5 7 samples C Unit 125% ]——490 m.——| ._ 5 ° Max. Distance Between Samples — 0 (Column from Gutschick, 1987) FIGURE 19: Spatial sample locations and stratigraphic column showing the three units studied, the number of samples taken from each unit, and the maximum distance between samples 53 calcareous shale, whereas unit C is a black shale with carbonate concretions and pyrite seams and nodules. Unit A is a black shale with an interlayering of bioturbated green shale. Contacts between shale units were used as a guide when selecting samples. The second objective was accomplished. by selecting samples as close to the same location as possible for each of the three shale units sampled. Each sample taken fulfills both objectives. Figure 19 shows the number of samples taken from each stratigraphic unit and the maximum distance between them. There were significant sampling limitations. The consistency with which samples were collected from a stratigraphic unit was ultimately controlled by quarry wall talus, which both covered lower exposures and dictated the locations where the upper units could be reached. Therefore samples from the same geographic location varied laterally by up to 12 m and only once were all three shale units sampled at one location. In addition, exposure to weathering varied around the quarry depending on when an area was last worked. The north and northwest faces of the quarry were exposed more recently, and were less weathered than the southern and eastern faces. Within these limitations care was taken to collect the least weathered samples from each location. Samples which had significant iron oxide staining or native sulfur precipitated on the surface, or were easily broken were avoided whenever possible. Even so, trace metal abundances 54 probably have been affected to some degree. .In fact, short periods of exposure to precipitation probably remove FRTE and other trace metals. One hour leaching tests with distilled deionized water by Leung and Plappert (1984) with weathered samples of Devonian Chattanooga Shale showed differential removal of Fe 0-1250 ppm, Mn 0-7 ppm, Cu 0-8.5 ppm, and Zn 0-12 ppm, both laterally and stratigraphically over a 10 county area. Also, the Antrim contains pyrite as nodules (2-50mm in diameter), and in thin. bands. .Any ‘visible pyrite 'was discarded during the initial crushing of the shale to reduce the risk of biasing the sample. The resolution of sampling around the quarry was on the order of decimeters, and could not accurately represent pyrite seams a fraction of a millimeter thick. Also, pyrite nodules are most likely diagenetic (Beier, 1988) and. might. confuse the original depositional pattern. Samples were taken from the quarry in labeled plastic bags. Two samples were also obtained from a core drilled at (Sec.8, T9N, R15E), Sanilac County, Mi. The samples within the care were chosen based on gamma ray log analysis of nearby cores and the Paxton quarry location (Ells, 1979) since gamma-ray information was not available for the actual core used. The samples were taken less than 1.5 m apart so that FRTE abundance variations could be compared in two relatively unweathered samples over a short distance. A 55 third sample was a powdered drilling grab sample from (Sec.16, T2N, R3E) Livingston County, Mi. The Sanilac and Livingston County samples were taken to give a general indication of how trace metal relationships compare over a larger area in the Antrim Shale. Shale samples taken from the quarry averaged 15 cm x20 cm x12 cm. A bagged sample was cushioned and broken down further. The cushioning prevented the sample from shredding the bag during crushing and greatly reduced the potential for contamination. The outer, weathered pieces of shale were selectively discarded, and the chips from the interior of the sample were collected and placed in a new plastic bag for further crushing. Metal at no time came in contact with the shale chips. Pyrite was continuously removed during this process. However, it is probable that microscopic pyrite was present which was not removed. This is one of the inherent sources of error in the sampling method. The shale chips were then placed into a Spex ball mill, using a porcelain canister and milling balls. Samples were powdered and passed through a 180 um mesh sieve, (Tyler equivalent 80) then stored in acid washed plastic bags. The porcelain canister and milling balls were prepared for the grinding process by being rinsed in distilled deionized H20 (DDH20) , placed in a DDHZO ultrasonic bath for about 5 minutes, and dried. No information was available on the influence of oxygen on metal removal during the selective chemical attacks for shale, therefore no attempts were made 56 to extract the FRTE under oxygen free conditions. It was felt. that the shales exposure ‘to the atmosphere before collection would minimize the significance of any repartitioning while stored or during the selective chemical extractions. Patterson et al. (1986), and Patterson et al. (1988) appeared to obtain reliable results under the same conditions. However, analysis under oxygen free conditions is probably still desirable although in this case the mechanical complications of operating within an oxygen free environment would more than offset the increased accuracy due to the oxygen free atmosphere. Five sub-samples were ultimately taken from each powdered sample; (see flow chart Figure 20) 1.5g for bulk shale major element analysis, 1.0g for bulk shale trace metal analysis, 5.9 for sequential selective chemical extractions, 5.g for sequential selective chemical extractions to determine percent shale lost, and 0.2g for determining percent total organic carbon. ‘- 5.1 '1'- asggxr - £314 TO :-"- Hurt: up ' - 8_,: Five grams of each prepared sample were used for the selective chemical attacks. Samples were placed in 250ml glass centrifuge bottles. Polypropylene and polycarbonate plastic bottles were initially used, but rejected due to a sediment film which floated on the leach and tended to wick up and adhere to the sides of the bottle. This film is believed to be a portion of the organic material and did not 57 I CHEMICAL FRACTIONS I SELECTNE CHEMICAL ATTACKS I I LIQUID SOLID SHALE 1 ATOMIC ADS ORPTI ON I I BULK X -RAY FLUORESCENCE TI TRA TI ON I C ORGANIC FIGURE 20: Flow chart of sample analysis 58 remain. in sufficient contact with. the leach in plastic battles. The shale fractions targeted for the selective chemical attacks are; exchange sites on clays, carbonates (weakly acid soluble fraction), Mn-oxides (weakly reducible fraction), Fe-oxides (moderately reducible fraction), organics/sulfides (oxidizable fraction), and detrital minerals (residual fraction). Samples were randomly divided into 4 groups of 5 samples each and kept together throughout the SCE so that the element abundances could be examined for bias in the leaching procedure. The leaching procedure was slightly modified from Gephart (1982) and is summarized below: EXCHANGEABLE: 5 grams of powdered, room temperature shale was reacted with 40 ml of 1M MgClz, pH 7 for one hour with continuous agitation. WEAKLY .ACID SOLUBLE (carbonates) The previously reacted sediment was leached at room temperature with 40 ml of 1M NaOAc (pH 5 with HOAc) for 5 hours with continuous agitation. WEAKLY REDUCIBLE (Mn-oxides) The previously reacted sediment was leached at room temperature with 125 ml of 0.1M NHZOH.HCl in 0.01M HNO3 for 30 minutes with continuous agitation. MODERATELY REDUCIBLE (Fe-oxides) The previously reacted sediment was leached at 96': 3°C with 100 ml of 0.04M NHZOH.HC1 in 20% (v/v) HOAc for 6 hours with 59 occasional agitation. Sediment container was covered loosely with acid washed plastic wrap to prevent leach evaporation, and to release built up pressure during heating. OXIDIZABLE (organics and sulfides) The previously reacted sediment was leached at 85': 2'C with 15 m1 of 0.02M HN03 and 25 ml of 30% H202 (pH 2 with HNOg) for 2 hours with occasional agitation. The H202 was added in 5 ml increments to prevent overflow of leach and sediment during the reaction. Then a second aliquot of 15 ml, 30% H202 (pH 2 with HNO3) was added, again in 5 ml increments, keeping the temperature at 85': 2'C and, agitating occasionally for 3 hours. After cooling (about 10 min.) 25 ml of 3.2 M NH4OAc in 20% (v/v) HNO3 was added and the sample was then agitated continuously for 30 minutes, and centrifuged for 60 min. at 2,000 RPM. The leach was then removed and placed in a 100 ml volumetric flask and brought to volume with DDHZO. (This last step was done to account for possible leach evaporation during the heating process and to give all samples an equal dilution.) RESIDUAL (crystal lattice of detrital and authigenic minerals) The sediments from the last leach were rinsed into an acid washed plastic bag and dried. The sediment was repowdered and element abundances were determined using x- ray fluorescence. After each leach the sediment and leachate were centrifuged for 60 minutes at 2,000 RPM. The supernatant 60 was removed with a pipet, placed in an acid washed polypropylene bottle, and acidified to a pH of less than 2 with nitric acid. The sample was resuspended and rinsed with 25 ml of DDHZO, then centrifuged for 1 hour at 1,500 RPM. The rinse water was removed and discarded, even though slightly discolored. It was assumed that the coloring was due to trace amounts of supernatant which adhered to the powdered shale by cohesion, and did not represent additional metal loss. The samples were stored refrigerated at about 4' C until the next leach. All samples underwent one leach per day until the oxidizable leach, when only 5 samples could be done per day. Therefore the longest sitting time between leaches was four days for the last group of samples. All leachate solutions were analyzed using atomic absorption spectrophotometry (Perkin Elmer Mbdel 560) with either flame or furnace (HGA 2200 graphite furnace with AS 40 auto sampler). See Figure 21. Standards were prepared in the same leaching solution used for the metal extraction. This was done to remove the influence of the matrix solution during AA analysis. The conversion from leachate concentration to shale concentration is as follows: CONCENTRATION OF METAL PER l GRAN OF SHALE = (1111 Of leachate used/ 5 grams shale) * (metal concentration in leachate) * (dilution factor) Each leachate sample was analyzed three times with the AA unit set for three readings at 1 second. 61 FIRST ROW SHALE FRACTION Sc Ti V Cr exchangeable -- -- -- —- weakly acid soluble -- -- fu fu easily reducible -- -- fu fu moderately reducible -~ -- fu fu oxidizable -4 -- fu fu residual —- xg xp xp bulk -- xg xp xp fu= analyzed by furnace atomic absorption TRANSITION ELEMENTS Mn fu fl fl fl X9 xg fl= analyzed by flame atomic absorption xg= analyzed by x-ray fluorescence-glass wafer sample preparation Fe fu fl fl fl X9 X9 Co fu fu fl fl XP XP Ni fu fu f1 fl XP Xp Cu fu fu fl fl XP XP xp= analyzed by x-ray fluorescence-pressed pellet sample preparation FIGURE 21: Method of analysis for FRTE from selective chemical extractions fl fl XP XP 62 Analysis of the exchangeable leachate was not successful due to the high salt concentration of the leaching solution (1M MgClz). Dilution of the leachate resulted in metal concentrations below the detection limits of the machine. Furnace analysis readings were too inconsistent to be reliable. In the future it is suggested that a chelation-extraction technique using methyl isobutal ketone and ammonium pyrrolidine dithiocarbamate (MIBK-APDC) be used to remove metals from the exchangable leachate for easier analysis or reduce the molarity of MgC12 in solution. Scandium analysis was also not successful using atomic adsorption. Scandium concentrations were too low for flame analysis. During furnace atomic adsorption, either incomplete atomization occurred, or Sc carbides formed which carried over from sample to sample. zinc concentrations in the easily reducible leachate could not be determined due to problems with standardization. The samples consistently read lower than the zero standard and blank sample. It was assumed that titanium, which is quite resistant during weathering, would be immobile, and therefore would not be present in the leachates in 'measurable quantities. A comparison of the bulk and residual titanium data supports this assumption. Raw data from the atomic absorption analysis is contained in Appendix A. - U 8 N D 8 ' 8 The standard procedure for sample preparation for XRF analysis is by either, 1) combining a powdered rock sample 63 with binding agent and compressing the mixture to form a disk, or 2) combining a powdered sample with a flux, liquifying the mixture by heating to high temperatures, and pouring it into a mold to form a glass wafer. For most igneous rock analysis the glass wafer method is preferred because better precision is obtained, particularly for major elements. Shale samples cause unique problems however. The high temperature required in preparing the glass wafers (z1000°C) volatilizes significant portions of the shale (s25%). Since the XRF instrument is calibrated for a set sample size, it was not known if the shale loss during preparation would affect the accuracy of the metal analysis. Also, there is no way to accurately determine how much shale was volatilized. In addition, it is difficult to compare volatilization between samples since the heating time required to make a glass wafer varied from sample to sample because the shale mixtures took various times to melt. Therefore, tests were conducted to determine the best preparation methods for this black shale study. The main questions asked when designing these tests were: 1) Would the volatilization of the shale during glass wafer preparation affect the accuracy of the major or trace element analysis? 64 2) Would pre-ignition of the samples before glass wafer preparation allow standard melting times for the glass wafer ingredients? 3) How does pressed powder sample preparation compare to the glass wafer method for trace metal analysis? Two UkS.G.S. shale standards, Cody Shale (Sco-l) and marine mud (MAG-1) were prepared as glass wafers and treated as unknowns. They were analyzed for major element abundances and compared with the U.S.G.S. standard values to determine if shale volatilization affected analysis. The results for Sco-l and MAG-1 are shown in (Table 1). The LOI metal data is adjusted for the increase in metals due to the loss of the matrix. The agreement between analysis and standard values for major elements suggests that LOI did not effect analysis. To examine the effects of loss an ignition (LOI), the U.S.G.S. standards and an Antrim sample were pre-ignited at leOO'C/ls min. By carefully weighing the samples before and after ignition, the amount of shale volatilized can be accurately determined. In addition, the melting time for each sample was consistent. The results, (also table 1), show that major elements for both the U.S.G.S standards prepared with and without LOI are comparable. In fact Antrim shale samples prepared with various pre-ignition temperatures are consistent for the FRTE, suggesting that accurate results can be obtained even if volatilization occurs during LOI or wafer production. However, the 65 TABLE 1: Test analysis: major element data for XRF glass wafer preparation (concentrations in weight percent) Sample Sco-l Sco-l Sco-l USGS w/o LOI lOOO‘C/lS min book value LOI 9.29% 5102 63.39 64.02 61.82 A1203 13.70 13.83 13.57 FeO 4.65 4.33 4.37 M90 2.76 2.89 2.84 CaO 2.64 2.71 2.60 NaZO .95 .85 .84 K20 2.82 2.81 2.69 T102 .62 .62 .59 P205 .22 .21 .20 MnO .05 .05 ..05 TOTAL 91.8 92.32 89.57 Sample MAG-l MAG-l MAG-l USGS w/o LOI 1000'C/15 min book value LOI 14.09% 5102 51.19 52.52 50.35 A1203 16.46 16.55 16.05 FeO 6.21 5.85 5.63 M90 3.13 3.33 3.20 CaO 1.38 1.50 1.43 NaZO 3.91 3.49 3.26 K20 3.72 3.77 3.51 T102 .75 .76 .71 P205 .18 .17 .16 MnO .10 .10 .10 TOTAL 87.03 88.04 84.40 Sample Antrim SS 'Antrim SS w/o LOI , 550°C/30 min 1000‘C/15 min LOI 24.5% LOI 26% SiO 53.09 50.37 50.99 A12 3 11.36 11.12 11.39 FeO 4.40 4.07 4.04 M90 1.40 1.32 1.86 CaO .93 .82 .80 NaZO .43 .38 .38 K20 3.73 3.43 3.48 T102 .60 .54 .54 P205 .12 .11 .11 MnO .01 .01 .01 TOTAL 76.07 72.17 73.60 66 usefulness of LOI is that it standardizes the time of glass wafer production and quantifies the amount of each sample volatilized. When the samples are not pre-heated, various heating times are required to melt the sample-flux mixture, which allows inconsistent time for volatiles to escape from various samples. Another advantage is that since the rock powder used to make the glass wafer is taken from a pre- ignited sample, the ‘metals present. may have been concentrated by #25% which makes it easier to detect major elements at low’ concentrations such. as manganese“ The disadvantage however, is that larger amounts of sample is required, because the rock powder used in the glass wafer is taken from a sample after the loss due to heating has taken place. XRF trace element analysis was preformed on Sco-l, MAG- 1, as glass wafers, and. the .Antrim samples as pressed powders and glass wafers, both with and without LOI at various temperatures (see Tables 2 & 3) . Both U.S.G.S standards prepared as glass wafers with LOI showed a caear' loss of copper and chromium, and zinc for MAG-1. The Antrim samples prepared as glass wafers with LOI were variable in trace metals. Chromium abundances decreased with increasing LOI temperatures, Zn increased, and Ni and Cu were highly variable. U.S.G.S. standards and the Antrim shale sample suggest that some trace metal FRTE were volatilized sometime during LOI and glass wafer production. Therefore a difference in trace element abundances was expected between 67 TABLE 2: Test analysis: trace element data far XRF glass wafer preparation (concentrations in weight percent) Sample Antrim SS Antrim SS Antrim SS w/o LOI LOI 550°C LOI 1000°C 30 min 15 min Cr 167.4 158.4 117.4 Ni 186.1 339.3 267.4 Cu 58.4 194.7 59.6 Zn 21.5 39.0 55.0 Sample Sco-l Sco-l USGS book value LOI 1000’C 15 min Cr 71 43.7 N1 30 29.2 Cu 28 4.6 Zn 105 102.0 Sample MAG-1 MAG-1 USGS book value LOI 1000‘C 15 min Cr 105 79.6 Ni 54 48.5 Cu 27 2.7 Zn 135 44.9 TABLE 3: Test analysis: trace element data for XRF pressed powder preparation (concentration in ppm) Sample Antrim Antrim Antrim Antrim Antrim Antrim Antrim Antrim Antrim ss ss ss 53 55b 88 85d 88 ssf LOI 550'C 1000'C Cr 174.2 149.5 161.9 175.6 176.1 180.3 177.3 175.2 171.9 Ni 286.1 260.6 257.5 300.4 310.5 305.9 315.7 304.3 298.0 Cu 182.0 185.6 183.6 177.6 176.9 174.6 178.0 174.0 173.0 Zn 54.4 52.4 52.0 57.5 52.7 55.4 53.6 53.1 51.7 5a: ma be 68‘ samples prepared as pressed powders without LOI, which are made at room temperature, and those which underwent LOI before pressed powder preparation. Antrim shale samples prepared in this way are listed in table 3, and do show some evidence for trace metal loss for Cr and Ni from the heat of LOI, but no effect on Cu and Zn. However, the same LOI samples prepared as glass wafers exhibit extremely inconsistent concentrations of copper, chromium and zinc. This suggests that heat alone is not the cause for trace metal volatilization. Perhaps the volatilization of the shale during LOI affects how the sample responds during glass wafer preparation. However it must also be considered that the sub-samples used in the glass wafer preparation may not. have been homogeneous. It was concluded. that the pressed powder method either with or without LOI yields more accurate results for the trace elements. The glass wafer method with LOI provided the most accurate results for the major elements. - Y 00 8 C Samples for XRF were prepared following the glass wafer method for the major FRTE elements (Fe, Ti,and Mn), as well as Si, Al, Mg, Ca, Na, K, and P, and the pressed powder method for the trace elements. 69 GLASS WAFER with pre-Loss on Ignition (LOI) 1) Weigh out approximately 1.5g of powdered sample to 5 decimal places in a pre-weighed, acid washed ceramic crucible. Dry at 110°C overnight. Let cool. Reweigh sample and calculate the percent H20 loss of the sample. 2) Heat sample at 1000°C for 15 min., let cool, reweigh. Calculate LOI. 3) In a thoroughly clean platinum crucible, weigh out: a. 9.0000 grams of Lithium tetraborate b. 1.0000 gram of prepared sample c. 0.16 grams of ammonium nitrate 4) Remove the crucible from the balance and stir ingredients until homogeneous. Cover with lid. 5) Heat sample over a gas flame of 1150 to 1180°C for 25 min., agitating the crucible constantly. 6) After about 10 min. stop shaker and, using platinum tipped tongs, remove lid and, rotate the crucible to prevent excess material from clinging to the sides of the crucible. 7) Continue shaking. 8) With 5 minutes of heating left, preheat platinum pouring mold until orange-red to white hat. 9) Remove crucible lid and pour liquid mixture into the mold. Set mold on a hot plate (about 500°C) and let cool until liquid solidifies. 70 10) Remove disk. 11) Store in a desiccator PRESSED POWDER PREPARATION (Without LOI) 1) Use 1 gram of powdered shale to .25 grams avicil (or adjust to 4:1 ratio). 2) Place above mixture in a ball mill and shake for 20 minutes. 3) Prepare metal tin by packing 3/4 full with avicil. 4) Remove sample mixture and fill remainder of metal tin, packing as necessary. 5) Place tin into die press and load to 21 tons psi for 2 minutes. 6) Remove, place in storage container, and store in a desiccator. 938.111.53.111: '1' 8 81 I The data for each sample from all leaches and all modes of analysis are in Appendix A. The FRTE data were prepared for examination as a group by graphically presenting the data on Coryell type plots log ratioed to Post Archean Average Shale (PAAS). This was done for all fractions of the shale. The Coryell type diagram shows the enrichment of the FRTE in the Antrim shale with respect to PAAS. In all of the shale "fractions" examined the abundances of the FRTE are much lower than PAAS, because the PAAS values are bulk shale metal abundances. How the FRTE from various samples plot against each other on this diagram depends on their abundances in the "shale fraction" examined, which may in turn be related to the bulk shale metal abundances. For example, a shale highly enriched in V and Ni, and depleted in Mn with respect to PAAS will probably produce plots with Mn far more depleted than V and Ni in all shale fractions. The importance of Mn in any of the shale fractions is visually lost. Therefore, the percentage of each FRTE in the shale fractions examined are also plotted. This type of diagram shows the relative importance of each shale fraction in sequestering each of the various FRTE, independent of other elements total abundances. The total metal abundances used to calculate metal percentages for each element were taken as the sum of element abundances from the selective chemical extractions. The FRTE patterns from these diagrams 71 SE fJ' at Sc Th. re; 72 can also be compared. Identical overlapping patterns would suggest that metal partitioning in the shale is independent of total metal concentrations. Differences between sample patterns indicate differences in metal partitioning within the shale. FRTE abundance patterns in either type of plot which exhibit the same trends and have correspondingly similar element abundances or percentages will be referred to as "consistent”. The data were collected and the samples were analyzed with only one complication. The metals in the residual fraction of the shale have been concentrated throughout the leaching procedure due to the dissolving of calcium carbonate, the destruction. of’ Mn oxides, Fe oxides and organic matter. This last A mass needs to be accurately determined to find the actual concentration of metals in the residual fraction. However, due to drying difficulties, a final sample weight could not be determined. Therefore a second complete set of samples were run through the selective chemical extractions and dried to determine the weight lost. The percent weight loss from the second set of samples was used with the residual abundance data of the first set of samples to calculate the actual residual metal abundances. The difference .between the FRTE abundances from the SCEs and the bulk shale is less than 15% for most elements. There are large discrepancies for chromium, but no reasonable explanations have been found. 73 BELE_£EALE Plots of the bulk shale data for the three stratigraphic units from the Paxton quarry are shown in Figure 22. The FRTE abundance patterns are similar for the three stratigraphic units in the Paxton quarry. vanadium, niCkel, copper, and in a few samples cobalt are slightly enriched; manganese is depleted; and zinc displays a wide variation with respect to PAAS. The FRTE abundance patterns from the Paxton quarry are consistent within each stratigraphic unit. Similar FRTE patterns occur for Antrim samples collected over approximately 0.2 kmz, despite the potential for inhomogeneity during collecting and sample preparation. This suggests that the methods of sample collection and preparation were satisfactory, and that a chemical record is preserved in the shale. The FRTE abundances and patterns of units A and B are nearly identical, but differ from unit C. It may be that unique FRTE patterns on the scale of decimeters are preserved within the Paxton quarry, and that different stratigraphic units can be distinguished based on FRTE patterns. However, whether these patterns reflect solely depositional processes, or were partially or completely modified during post-depositional processes, requires (additional study. It is also not known whether the FRTE pattern differences between stratigraphic units are due to depositional processes or post-depositional events. Either 74 mafiucsoo awaficmm can coumo:«>wq Baum moamewm 0cm wucsou mammaé .zuumso couxom may ca muss: oflnmmumfiuouum n you mamnm cawamuumsc momum>< :mwnau4Iumom ou omofiumu macaw Banned xasm a“ mama mo cowumuucmocou mEmEQm coEmcmC 26¢ 29E CN 30 .2 00 mm 52 .0 > :. om wha ,- IIJ! I IJI ital III «llllnnlll. .IIIII- .I-.. I .._ Ill _. 7U _ W 3.3 I W 88 - , I :8m.I U T A k C r LI C h F » e p In 3.5360 ghosts pen amazon BEEQw coEmcmC 26m 39E N :0 .z 00 on. :5. .0 > ; um I q q A d 1 4| 4 a a All i TTT'T‘T—‘l—T __‘“ 71 IHIIIY :C p b b H a 35 838838.... b h b » 1111111 L 1 0— To or SWd/BIBIIS )IInB SWd/SIBUS )‘llna "NN mmDUHm 2coEo_m cognac... 26m 5...“. cN so _2 oo 81 :2 .o > _F om « u «I u m _ _ q u a q . I .. I_o m H n U W I. m W . .moe oasxesssssa .mEoEQm cozmcoc. 26m 8.: cm 5 _z . oo 8 £2 5 > : um .11 a 3 a _ _ a q A I w 1111111 I 7 L. _1 11111111 4 l V" C 1 111111 1 1 or p F r P h b — L, » fiq Baum amamfiom can >ucsoo mammad .muuaoo couxom on» CA mafic: ofinmmumfiumuum m you macaw coflamuumo< momum>< camnaucuumom a» owoAumu 77 GHMSM EHHHC4 HO COHHOQHK Hflflflfimmm G89 C.“ mfimm MO COHuMHUFAQOCOU "MN WMDUHW mEmEQw coEmcEA. 26m Ami 25ng coEmcac. 261 mi CN 30 .2 00 on. E). .0 > _.A om cN :0 _z 00 on. 52 .0 > F um I A A A A A A A A A A .1 rod I A A . A .A A A A 4 A .1—00 I . A l 1 H I A.— M 1 A A m. m A A . No A m m D. m .. . x. . mo . m . ml Im v.0 n F A. . A . NU In A O .. DIV. I A .. A A... m0 . I n I S I _. A «o I I A u. n .. x. N0 . A m A W m .A I t 6 m .II.II I... F E “III . . . II IxquI I P .d ,9...\.\. I A W I ,m&. A m A s m . A WI A A A A A A A r A A A Im OF A A r A A A A A A A KW. OF «pagan counts: new case.“ 0 is oEQEuAAEB mAcmEQw c0235»; 30¢ ABE mAcoEQw c0298.... Asom Ami cm :0 .z 00 mm :2 .0 > ; om cN :0 .2 00 on. :2 .0 > :. om I4 A A A I dI 1 A A A A1 FOO . A A A A A I A A A A A1 PO 0 I am I I H T I A mm A w A A H m ..... m m m m ml 8 I- In. to m. m. IM to H A w A A m A W m A "1 IA A m m. Im A I I m I I . A s m WI. A A A A A A A A A A 1 OF P A A A A A A A A r A In OF mESnESESEB «SSAESSSEB SWd/GIBUS [800585 SWd/alelIS IBnDISBH 78 coAuomum AmsuAmmm map :A mama mAcmEQw 8593:. 26¢ 55 CN no .2 8 on. cs. 6 > : 8 A A A A IA A A): ..... «cam -- I 28 I I. I. r A LI A A A A A AI 82:30.0 coAmmAASQ Eta cakes...» mAcmEmAw coEmcSA 30m 55 CN :0 AZ 00 mm 52 AU > ; om NW A % A A 1 A A A A A- 8 I I .8 mm ..... I mm -- #m ..... om I Am I A A A1 m ES QEEAAAEB ON CV 00 om 09 ON ov 00 cm 00.. uouoeJ; lenpgsax u! slelew % e|eus lenpgsaJ ug sle1aw % no mmmucmoumm "AN mmstm 2:955 5293:. 26m Ami 5 :0 _z 00 mm 52 .0 > _.A cm A 1 A A A A A A A A IO Lom No.1 mo. so _ IIov mw,- v , No ,: IAoo A0 I Low I# A A A A A A A A A I OOA oSSAE£E§§G mAcoEQw coEmcSA 26m Ami cw :0 _z 00 on c2 .0 > AA om A A A A A A A A A A I Io Iom A< m< ;- Ilov m<.| loo Iom IA A A A A A A A A A OOA V SS QAEESAER uouoeu |BDDISGJ u! slalaw % uogpeu lenpgsej u! smaw % 79 is consistent with the geochemistry of these elements, which is to partition mainly into silicates. The chalcophile- siderophile elements Co-Cu are sequestered less than the lithophilic elements. However a large percentage of nearly all FRTE are associated with the residual shale fraction, and therefore detrital and other resistant minerals within the shale are the single largest control on metal abundances in the Antrim shale. In addition, the similar patterns for the percentage of each element in the residual fraction suggests that the detrital/authigenic component of the shale has remained equally important in metal sequestering throughout the examined vertical section. The difference in FRTE pattern in unit C from the other Paxton units could be due to a large number of factors. If the SCE were 100% effective in removing metals from all of the other' designated. shale fractions, detrital -and authigenic minerals are the cause of the distinction between units. The detrital products could have been modified by a change in the duration or intensity of parent rock weathering, or a change in provenance all together. The diagenetic conditions which produced the authigenic minerals or recent weathering may also have had an effect on the FRTE patterns. In addition, incomplete removal of metals from the previous chemical attacks would influence the FRTE patterns. It is believed that the samples were not influenced by analytical techniques. Samples were run through the SCE in 80 four randomly selected sub-groups, so that samples from one stratigraphic unit were not grouped together. In this way, slight differences in processing time after leaching could not contribute to the pattern differences between stratigraphic units. The internal agreement for the percent metals removed from each stratigraphic unit in the residual fraction, and the regularity of the unit C FRTE pattern differences, support this idea. The major element abundances in the residual fraction are very similar in each of the stratigraphic units (see Appendix A), which suggests no major change in provenance within the sampled section. If the provenance changed, either the new input was small, or the new source material was very similar to the old. However, a change in the weathering rate of a parent rock could also affect the resulting element composition. Landscapes with steeper gradients erode more quickly, resulting in less thorough weathering of the mineral constituents. It appears that the difference between unit B and unit C’s FRTE patterns could be the result of differences in source weathering intensity. The unit C FRTE pattern is similar to unit B's except that vanadium and nickel are more enriched and manganese more depleted. However, if this were the case one might expect unit C to have a lower concentration of soluble major elements, (Na, K, and Ca) than unit B. There is also the possibility of diagenetic influences. The residual fraction probably includes authigenic as well 81 as detrital minerals. Minor portions of both quartz and illite in the Antrim were authigenically derived (Routsala, 1980; Hathon 1980). Therefore metals from both sources comprise the residual FRTE patterns. It is possible that a portion of the V in the residual fraction is due in part to its incorporation into diagenetically produced illite- montmorillonite interlayers. Patterson et al., (1986), and Patterson et al., (1988) demonstrated with selective chemical extractions and confirmed with mineral analysis that the majority of V in the. Julia Creek and Condor Australian black shales is contained within these mixed- layer clays, rather than incorporated into organic matter. In terms of recent weathering history, all three stratigraphic units were exposed at the same time when the quarry was initally deepened, however presently the southern quarry wall has been exposed longer than the others. Leung and Plappert (1984) did distilled water leaching studies on a Devonian shale and noted that the weathered samples lost more metals than the fresh. Therefore, if recent differential weathering were a factor, one would expect variations ‘within. stratigraphic ‘units based on. the geographic location of the samples. This was not observed. However, it may be that some stratigraphic units lose some metals more easily during weathering than others. Leung et a1. (1984) discovered in distilled water leaching studies that various black shale units lost particular metals more 82 easily than others. This also was not noticeable in the Antrim samples. A component of the FRTE pattern in the residual leach could be due to the ineffectiveness of the oxidizable leach in removing metals. The enrichment of V and Ni in the residual fraction may be due in part to the possible carry over of metals from unreacted insoluble organic matter. Kenny (1980) found that both of these metals were found in petroporphyrins in the Antrm shale. In addition, nickel could also be present due to the carry over of unreacted sulfides from the oxidizable leach. The wide scatter of Co and Cu as well as Ni in the C unit residual pattern could be the result of only a partial removal of metals from the sulfides in the oxidizable fraction. However there is no indication of this in units A and B, as all element abundances are quite consistent. If the chalcophilic metals in units A and B are due strictly to the residual fraction, and the detrital composition of unit C is the same as units 'A and B, then the abundances of Co, Ni, and Cu in unit C should be at least equal, and greater if an additional sulfide component is also included. This is not the case. In addition, unit A has a much higher organic content, and based on observations, more sulfides than unit B. However there is no difference in the residual fraction between the two units in chalcophilic elements. The uncertainty of input from the oxidizable fraction makes it difficult to determine to what extent, if any, 83 sulfide and organic bound metals are influencing the residual fraction FRTE patterns. However, it appears that there is a "real" difference in ‘trace :metal abundances between units C and A & B in the residual fraction. It also seems probable that diagenetic conditions throughout the Paxton quarry were uniform, and that differences in FRTE patterns in the residual fraction are in part due to differences in detrital metal abundances. The samples from southern Michigan reflect the same FRTE patterns as the Paxton quarry samples (Figure 23).. The Livingston county sample and Sanilac county sample Sanz have residual fraction FRTE patterns that are similar to Paxton units A and. BM Sanilac county sample Sanl has lower abundances of Co, Ni, Cu, and Zn in the residual fraction than the A and B Paxton units, and more closely compares with unit C. The difference in residual fraction FRTE patterns for the two Sanilac county samples is unexpected since their bulk FRTE patterns are nearly identical. A plot of percent metal in the residual fraction (Figure 24) shows that the percentage of FRTEs partitioned in the residual fraction is much lower for sample Sanl than San2. This suggests different metal partitioning in two samples with the same bulk composition. This is signficant because the samples would have appeared identical if selective chemical extractions had not been used. OXI DI ZABLE FRACTION-- The FRTE pattern in the oxidizable fraction is distinctly different from the 84 residual pattern (Figure 25). Chromium, cobalt, copper, and nickel are the least depleted, and manganese, vanadium, and for the most part zinc, are the most depleted. Relative, and in many cases absolute, element abundances for all three Paxton stratigraphic units are consistent. Unit C differs from units A and B in that Zn is highly variable. The wide range, and relatively high Zn concentrations suggest are believed to be due to an inhomogeneously distributed Zn mineral, such as sphalerite. The percentage of FRTE sequestered in this fraction generally increase across the group (Figure 26) . This corresponds to the increase in the element's chalcophilic behavior across the group. The lithophilic FRTE, particularly V and Mn (see Figure 25) are the most weakly sequestered, and the siderophile and chalcophile elements Co, Ni, and Cu are the most strongly sequestered in the oxidizable fraction. It also appears that only the abundances of Co, Ni, and Cu in the oxidizable fraction are high enough to influence the bulk shale FRTE pattern. It is important to emphasize that unit C FRTE abundance patterns do not differ significantly from the other two units. In addition, the relative importance of the oxidizable fraction in sequestering metals is similar for all three stratigraphic units except that unit C sequestered a larger portion of Co. This is surprizing because of the variable amounts of total organic carbon (TOC) in the three stratigraphic units (see Appendix B). The A unit, a black 85 macaw Ewuusd no coHuooum manmnaowxo may 2“ mama mo cofiunuusmosou “mm mmDon mEQEBm coEmcmc. 26m 5:”. cm :0 _z 00 on. :2 .0 > ; om We a _ a a _ a A 2 a «.1 -. m W t I U W .5 .m «:3 L W Emm .l L ml. L. p w L p p p F p .1“ mottaou coammtxsh p5. omega mbcmEmm coEmcafi 26¢ FBI 5 so .2 ,oo on. 52 5 > : om Tl q T “If m .m 41., 0m I 1 m , km. 1 H mm ..... H m mm m V cm ..... '1 mm . 1 m E | 1 WIL— _ p h p h p h h b .— I'm mfiSnssseea 5.0 To or 5.0 to o— SWd/ U0!1391:J XO SWd/UOHOBH XO mowucooo Unaflcmm can soummcfi>fiq Eouu mmHQEmm can wussou mammaé .>uumso :ouxmm on» Ca mums: Uficdmumwumuum n you macaw smaamupms< mmwum>¢ smonoudlumom ou omofiumu mecoEmm cozficuc. 26m «mi [HI]! I I T E 3o _z oo on. 22 .o > Om 1.4 a a 2 T _ _ a j a {A W 1 No.1 1 mo. m m0 . u T NU . I wo ul WI_ .. _ p _ _ _ F _ r Um oeSQQS§5§SG 2,8ng coEmcoP 30m 5:... CN 8 _z 8 on. 52 5 > : 8 .l. u d fi : J1 a u 1 — «III... WI. uh T 1 ml n< .Im m< - I4 m< I U p p — b _ _ p _ sesrsafissa ,_ O '— O ,_ Foo To OF SWd/ U0!1391:i XO SWd/UOHGBH XO 86 :OAuowum «Hannacaxo may ca 9.5ng coEwcmfi 26m 6:". cN :0 .2 00 on. 52 .0 > :. om New ..... c -- 2.8 I L TIP P p p p p . p p p b P 8:560 coemucxs pen ommcau. 2cmEm_m coEman 26m Em cN :0 _z 00 m... 22 .0 > :. om . d J a a q q a 1 _ . .1 em I 1 Km . . om ..... mm l a . 8 in 7 E I J T p p L p L, r _ p p p m .23 guesses ON av 00 om 00F om CV 00 cm cow mean no momucmoumm "on amouHm mecoEEw 5235.... 30m 5:“. cN :0 _Z 00 on. E). .0 > ; ow a fi — a 4 u a q _ _ A o I. o/ w 1 w pic. m. 8 I 1 U WW . m ww-: 1 m, N0 .-- my. : 8 l 1 0 U T: p s . s _ . t . s . . I ossnEsfiefa macesom COEmCGC. 30E «em 8. so .2 co a“. 52 .o > : 8 7 _ . j a . d d, g a . .1 o/o w 1 w al.. S l m. 0 ..< m 2 -. m 2 I m. I. l O U IF F r p F t . . s . . I. 3:5 £3288»... 0N CV 00 cm 00* ON 0v oo 00 09 uogzoeu x0 u; smaw % uouoeu x0 u! SIBIGLU °/o 87 shale with interbedded bioturbated green shale averages 8.07% TOC. Unit B is a light calcareous shale with an average of 4.98% TOC, and unit C is a black shale with pyrite and carbonate concretions and contains 6.55% TOC. It has been shown that organic matter attracts trace metals, including many FRTE (Horowitz, 1986) , and that the metal enrichment of some black shales is the result of metals concentrating within the organic matter (Vine and Tourtelot, 1970) . However it appears no relationship exists between the amount of organic material and metal enrichment in the Antrim. Samples with high TOC do not have higher metal abundances in the oxidizable fraction, or the bulk shale. This suggests that some other mechanism is responsible for controlling the FRTE abundances in the oxidizable fraction. Explanations must consider the possibility that previous SCE were not 100% selective or 100% effective in removing their targeted fractions, as well as the uncertainties associated with the oxidizable leach itself. It was considered that the capacity of the oxidizable leach to remove metals may have been exceeded. However, if this were the case, one might expect more random FRTE patterns in the residual fraction, due to the carry over of elements from the oxidizable leach. In addition, Presley et a1. (1972) removed similar amounts of metals from modern sediments with the same chemical extraction with no indications of saturation. Therefore it is believed that 88 the oxidizable FRTE patterns are an actual reflection of the shale, but not necessarily of the targeted shale fraction. This particular chemical extraction is not 100% selective and removes metals from some sulfides (about 60%), as well as a trace amounts from within clays (Tessier et al., 1979). In addition, the organic leach used does not remove all metals bound to organics since "paraffin-like material" and some resistant bitumens are not affected (Tessier et al., 1979). Relatively’ high amounts of Co, Ni, and Cu in the oxidizable fraction are most likely due to the removal of sulfides, as these elements substitute in trace amounts for Fe in pyrite. However a portion of the Ni may have been associated with organic material. Nickel and vanadium are known to be enriched in organic matter derived from oceanic or lacustrine algae, particularly when it is deposited in anaerobic conditions. Nickel and vanadium are strongly bound. in. porphyrin :molecules ‘which. are formed from ‘the breakdown of chlorophyll. The proportions of V to Ni in oil can provide important information about the chemical conditions during the deposition of the organic matter (Lewan, 1984). For example, V+5 forms large quinquivalent anions which cannot fit into the porphyrin structure, resulting in virtually no V associated organics. This would appear to be the case in the oxidizable fraction, assuming all organic associated metals were removed. However studies done on the New Albany and Chattanooga Shales show 89 proportions of V/(V+Ni) to be about 0.7 in organic material (Lewan, 1984). Kenny (1980) showed that porphryins in the Antrim Shale contain between 60—80% V and 40-20% Ni, and that the proportion of V to Ni increased with depth. Vanadium nickel ratios of 80/20 were found within 3 meters depth and 100 meters geographic distance of the Sanl and San2 samples. Since very little V' was found in the oxidizable fraction, it is Ibelieved. that the oxidizable leach did not remove all organic associated metals. Most of the V and Ni trapped in porphyrin molecules were carried into the residual fraction. If this is the case the Ni in the oxidizable fraction is due predominantly to sulfides, and some fraction of Ni in the residual fraction is due to organic matter and perhaps some sulfides. Since V in the residual fraction may be associated with both clays and organics, the proportions can not be determined. Lewan (1984) suggests that the relative amounts of V and Ni within porphryins in crude oil reflect chemical conditions during the binding of metals to organic matter. If the chemical conditions are such that both elements can be incorporated into the organics, then their specific relative abundances depend on their availability in the pore water during shale deposition Lewan (1984). However, Kenny (1980) states that the ratio of V to Ni in porphryins is a function of the age of the shale and suggests that Ni is the primary metal in porphrins, and over time the Ni is replaced by V. 90 The FRTE patterns from Idvingston and Sanilac samples are all similar in the oxidizable fraction and they compare favorably to the FRTE patterns from the Paxton quarry samples, (Figure 25), except for lower abundances of nickel and copper. The concentration of all metals in sample Sanl is slightly greater than the other two southern Michigan samples. The percent metal plot for the oxidizable fraction (Figure 26) shows that the metals in Sanl are much more highly' partitioned than those in sample San2. It ‘was suspected that sample Sanl had a higher TOC value than sample San2, similar to the high TOC values expected for stratigraphic unit C in the Paxton quarry. However, as with the Paxton quarry samples, TOC is .similar; sample, Sanl (6.13% TOC), San2 (5.33% TOC), and Livl (2.18% TOC). The Sanilac county samples also suggest that there is no correlation between TOC and metal enrichment in the Antrim Shale. MODERATELY REDUCIBLE FRACTION-- The FRTE patterns from the moderately reducible fraction are unique for each of the three Paxton quarry stratigraphic units and non-Paxton samples. Variations in individual element abundances within the stratigraphic units for the moderately reducible fraction (Figure 27) are on the same order as the residual and oxidizable fractions. However, the FRTE patterns appear more variable because element variability in the moderately reducible fraction affects the trend of the group pattern. Some samples within the stratigraphic units exhibit 91 Efiuucs no cofiuomum manwospmm waoumumpo: 0:» ca mama no soauwuusoosoo "hm mmDon mEoEmm 5230... 30m .2“. 00 on. $2 .0 > c. 00 q q — 4 _ d 1 c «N 3 ’ O 2 1 {III I Ifi I ‘\ [1111111 I 1 .33 . «cam Emm I 111111 1 1 TITI I I—I h p h p b p p L p h ’— 32280 5.3323 new 68.36.... 2cmEo_w c0265... 26m .mi CN :0 _Z 00 on. £2 .0 > C. om «I a a 4 4 _ 1 I l 1 11111 11] IIIIII I I I 1 IIITTI I I 11111 1 1 mSSnESS€§G 5.0 _..0 5.0 SWd/UOIIOBH BIN SWd/UOIIOBH BIN mmwucsou omawcmm 0cm soumosw>fla Eouu mmamfimm 0cm wuszou mcmaad .uuumno souxwm mnu :H mafic: oflnmmumwumuum n you wamzm smfiamuumsd ommum>< cmmcoudnumom ou Umofiumu mamsm ficoEom cozmcu... 26m 6:... cN so .2 oo o. :2 .o > c u u T 111111 Ij I ITIIIIIT J1111111 1 J11111 1 1 1 oESBESESEB ScmEEw cognac... 26¢ a... CM :0 .z 00 on. 52 .0 > E. I IIIII I I I 1 IIIIIIT I a 1.11 I- >. 1 111J111_1_1 411141111 stS~ESE€§G 5.0 50 F0 SWd/UOHOBH HIN SVVd/U0!1391:I BIN 92 soauomum manfiospmm aamumnmpoz on» c“ ImEmEQm 8:35... 261 7...“. 5 3o _z oo o. :2 .o > : 8 a 4 q . A fl « _ _ fill 1 I . 3... I . mcmm .- - E3 | T. .I L s . p . . p . . 1 3:560 5.39%.: ps6 unfit...» 2coEo_w coEmcm... 26m 6:... cN :0 _z 00 on. :5. .0 > c. ow « d u a a a fi 4 q s 11 m .23 8323.8...» om 0v 00 00 00F 0N CV 00 00 00v uouoeu aw u! SIBIGIU % uouoeu aw u! SIBIGUJ % mama mo wmmucmoumm “mm amouH. m.coEo_w seamen... 26.. .2... .0 > C. om cN 30 .2 00 on. E2 . d u 1 a J a - 4 0m . I NC I I 0... 00 . NO . I no .- 1 00 v0 N . I .w I I 8 TII . . . . _ b . . t LI. 00—. o .23 6.386.226 mar—050$ COEmCfl.F 30m «mtu cN :0 .Z 00 on. cs. .0 > C. om 3 1 . _ . _ q _ a a I 0 ON I. 0v I h< l 0< . 00 0< I I I 00 I. . r _ r _ . s . . . 00—. 3E: ointmentfi uouoeJI aw uI SIBIGUJ % uonoeu 3w u! slelaw % 93 subpatterns, perhaps representing slight Ivariations in conditions. Manganese, cobalt, and zinc variability appear to account for the majority of pattern scatter. However, a plot of the percentage of FRTE in this fraction (Figure 28) demonstrates the general repeatability of the moderately reducible fraction in sequestering the FRTE from sediments deposited under a variety of chemical conditions. There are however a few individual samples which have high percentages of manganese, cobalt or zinc sequestered by the moderately reducible fraction. The high Mn spikes occur in samples B6, B7, and Livl, with relatively high total Mn 700-900ppm, x=293 ppm, and have higher Mn percentages in the moderately reducible fraction, than the other samples. They’ also have the highest bulk CaO concentrations (5.8-9.3% wt.), probably due to calcium carbonate. It may be that the high Mn is due to carry over from the weakly acid soluble (carbonate) leach. However if this were the case one might expect to find elevated Mn concentrations in the easily reducible fraction as well. B6, B7, and Livl samples do not contain as high of a percentage of Mn in the weakly reducible and weakly acid soluble leaches as many of the other samples. Relatively high. Mn concentrations ‘were present in the ‘weakly acid soluble and easily reducible fractions: however, only in samples with relatively high bulk Ca did the Mn concentration. reach anomalously' high ‘values in the moderately reducible fraction. It may be that the easily 94 reducible and weakly acid soluble attacks are not as effective in removing Mn when present in these greater amounts, or when associated with higher amounts of Ca. Perhaps the Mn in these samples is indeed partitioned differently within the shale. The anomalously high V, and Cr in the B6 and B7 samples does not appear to correlate with any other major or trace elements in any simple fashion. The Co peaks in samples (A3, A6, C7) also do not appear to correlate with major or trace element abundances. It is suspected that the higher variability in FRTE patterns is related to Mn and its ability to remobilize during changes in redox conditions. The three stratigraphic units examined in the Paxton quarry appear to represent times when the redox boundary varied with respect to the sediment water interface. Unit A is an organic rich black shale with a fine interbedding of bioturbated green shale and is believed to represent a time when the redox boundary fluctuated above and below the sediment-water interface. The light grey shale of the B unit represents a time when the redox boundary was consistently below the sedimentdwater interface, and the C unit, rich in pyrite and organic matter, represents a period when the redox boundary was above the sediment-water interface. The abundance of Fe in the three Paxton quarry units varies between 12 and 20%, with increasing Fe concentrations from the most anoxic unit, C, to the most oxic unit, B. 95 Manganese oxides are more unstable in near anoxic environments than iron oxides, and are less likely to be preserved (Shaw et al., 1990). It appears that the oxygen available was not sufficient to preserve Mn oxides to any great extent, and consequently, Mn sequestered to an important degree in the moderately reducible fraction (Figure 28) , presumably with Fe oxides. A qualitative correlation exists between the estimated amount of oxygen available per unit during deposition, based on lithology, and the percent. Mn sequestered. From ‘the most anoxic strata, to the least; unit C, contained 15% of the total Mn, unit A, 25%, and unit B, 38%. Cobalt is also sequestered to an important degree (about 10-40%) , and may be associated with Mn. The Sanilac and Livingston County samples have the same general patterns in the moderately reducible fraction as the Paxton quarry samples (Figure 27). The moderately reducible fraction of the Idvingston county sample has highly sequestered Mn where as the Sanilac county samples have not. The relative importance of the moderately reducible fraction in sequestering' of’ Mn (Figure 28) corresponds ‘with. the estimated amounts of oxygen available based on the TOC of the samples. High TOC values suggest that less oxygen was available at the time of sediment deposition, and hence Mn abundances are expected to be low. Sanilac samples average 5.73% TOC and the Livingston County sample, 2.18% TOC. The Sanilac County FRTE patterns in the moderately reducible 96 fraction are similar and the relative FRTE partitioning (Figure 28) is identical. This suggests that the major difference in metal partitioning between the two Sanilac samples is in the residual and oxidizable fractions. Based on the above general FRTE trends, it is thought that at least qualitative information about the Fe oxides is preserved, although element-abundance scatter between samples suggests post depositional processes have modified the FRTE pattern in the moderately reducible fraction. EASILY REDUCIBLE FRACTION-- The easily reducible leach removed between 0-13% of individual FRTE from the shale. The general FRTE pattern is again unique and distinguishable from the other leaches (Figure 29). However, element abundances within stratigraphic units vary considerably. Despite the variability in patterns, unit C can be recognized as different from units A and B, again showing a relative enrichment in V and Ni with respect to the other stratigraphic units. A plot of the percent metals in the easily reducible fraction (Figure 30) demonstrates the consistency of individual metal partitioning, with the notable exceptions being Mn and Co. However, the low overall total abundance of Co (x=23.9 ppm) makes small differences in concentration quite significant in terms of' percent difference. 'The easily reducible fraction sequestered an average of 5-14% Mn in the samples collected. The sequestering ability appears to be independent of paleoredox conditions, as the percent 97 25ng cognac... 26m 7...". cN :0 .2 00. on. 52 .0 > C 00 I . 4 . q I . 1 I q q _ _ l1 7 3:. 1 W «:8 - - I I Emw I u m m T _ . L P . L p _ _ _ pl” «Magoo 5.33.5 p5. 2.5.6 «Ewesm c0265; .52". a... 5. :0 .z 00 on. £2 .0 > ; om I 4 _ a q a a 7 q d A I1 . 1 I 1 w H n no I H ”I B . I“. 00 ..... mm .- - I a . mum . I - E I IIWII I I 11111 1 mssyéssegs 0.9. 5.0 *0 0.m_. 5.0 *0 SWd/UOBCIBH HE! SWd/UOIIOBH 1:13 mmwussoo omHHcmm osm coummcfi>fla Eon“ mmaafimm can aucsou Mommas .muumso souxmm may as mafic: oflnmmumfiumuum m you wanna cmfiamuums<.muaum>< smmnou c. om 1 . c _ . . 3 j q a 41 m.w— 1 u. n I. m. I W I no ..... I F0 0 1 mo.: - r .0 I NU . 1 . .o.I . m n. WI. . . p I I _ p k . .Imw «.0 ossaasisssm 2coEo_w c0330.... .50: 8... CN 8 _z co a. 52 .o > : om . 4.. 4 _ q _ q . . 4 11 m.w_. w H W mFQo I b< 1 . o<. I u 2 I H m . HI» . . _ p _ . . . . I v.0 «SSBESSSIG SWd/ 11090913 1:13 SWd/UOIIOBH 1:13 98 cofiuomum wanaoccmm haammm 0:» ca 2coEo.w cozmcc.» 26¢ a... cN :0 .2 00 on. :5. .0 > :. 00 I4 d _ q 1 u d 4 q q 1 TI 1 3: mcmw I scam I l r p . p b p t . . . +1 «.6338 coRmSS pea 2.3.6.... m.coEo_m coimcofi 26m 8:“. CN 30 ._Z 00 o... :2 .0 > C 00 a _ 4 d W — q a _ F p p L bI _ P P p » m as Sesameefi or me 0m 0.. mp 0N uouoeu 83 u! SIBIauI% uouoeu 83 U! SIBIBUJ % mean no mmmucmoumm "on mmDuH. 25:56 cozfico... 26m 5:... EN 10 .2 00 o... 52 .0 > E. 00 «I a u p _ . _ _ b . s . . 6 SS c.3288...» 2coEo_w 5538.? 30m .2... cN 30 _Z 00 on. 52 .0 > C. 00 u — J _ q I_r I h p P p b P p L 3:3 $328.2.» 0* mp 0m 0* 0? 0m uouoeu 53 u! sleIaw % uouoeu H; u! SIBIauJ % 99 Mn sequestered does not relate to lithology or TOC. Cobalt and nickel, found to be enriched in Mn oxides (Shaw et al., 1990) are relatively enriched with respect to the other FRTE in the easily reducible fraction. The probability of Mn oxide jpreservation is extremely low, in all units, but particularly in unit C, where little if any formed because of the anoxic conditions present during sediment deposition. The Mn oxides ,present in all three units are probably the result of recent Mn oxide formation due to recent weathering. This supports the easily reducible fractions inability to sequester FRTE, particularly Mn, in quantities related to the original paleoredox conditions. However, it is safe to assume that the 3 stratigraphic units from the Paxton quarry have undergone identical post- depositional histories, and that the differences in FRTE patterns between stratigraphic units reflect chemical differences during sediment deposition and shale formation. The samples from Sanilac and Idvingston counties have FRTE patterns in the easily reducible fraction which are comparable, and similar to the Paxton quarry units A and B, (Figure 29). The percent. metals jpartitioned. in easily reducible fraction is identical for the Sanilac samples, but the Livl sample partitioned less Mn (Figure 30). The variation in partitioning pattern is similar to the Paxton samples. WEAKLY ACID SOLUBLE FRACTION-- The patterns for the weakly acid soluble fraction in the 3 stratigraphic units 100 are unique, but internally consistent within the unit in the Paxton quarry (Figure 31) . Manganese, nickel, and cobalt are the least depleted in most samples with respect to PAAS. These metals are also the ones which are partitioned most strongly in the weakly acid soluble fraction (Figure 32) . All FRTE are sequestered in "trace" amounts, 0-5%, except for Mn which is between 1-15%. This is probably due to the substitution of Mn for Ca in calcium carbonate. However, the percentage of metals sequestered in the weakly acid soluble and easily reducible leaches are nearly identical (Figures 30 and 32) . This suggests that the weakly acid soluble fraction has removed metals from Mn oxides as well as carbonates. Tessier (1979) notes that the weakly acid soluble leach has been known to affect Mn oxides in some studies. The FRTE patterns are also comparable except for the relative enrichment of Cu and depletion of V with respect to the other’ elements in ‘the 'weakly acid soluble fraction as compared to the easily reducible fraction (Figures 29 and 31). The FRTE patterns for the Livingston and Sanilac County samples in the weakly acid soluble fraction are similar, (Figure 31). The livingston county sample pattern matches with unit B of the Paxton quarry, although the Sanilac sample patterns do not, as Cu abundances are much lower. The percent metals sequestered by the weakly acid soluble fraction for these samples are equivalent to the Paxton quarry samples. 101 mmwussoo omesmm use coummsfl>aa Scum mmamamm can wussov acmaa< .huumso souxmm on» as muss: 035cmumflumuum m Mam mamnm smfiawuumsd wamum>< smmnou P 00 fi3 3 3 3 3 3 3 A 3 3 31 m .. .. m In J :5 1“ Ncmm - I m 2.3 I I We 3 3 LI IL, 3 3 3 3 p FIJI... 8.33.3360 3.36.23 new 3.33360 mEmEoE 3.02.9333... 30m 39... cm :0 _z 00 on. 3.32 .0 > fi 00 .13 I 3 3 3 3 I 1 3 3 3 J - m mm I I N . 3 a ..... w - mm .- m cm ..... I1 8 -- I rm I .1. n m T. 3 3 b 3 p h 3 F II 3 QSSnEBESEB 8m: mm? 5.0 3.0 cm? Qwr 5.0 3.0 SWd/UONOBH SVM SWd/UOIIOBJd SVM “H m mmDmuHm m3cmEo_w 3.0.3395; 30m 32.“. cm :0 _z 00 a... £2 .0 > ; 00 I3 3 3 a 3 3 3 3 3 I32.-- 3%. «fwr 3. . I m a. H w mmm3 .. NO I I 1 mo. 4 n 50 . u n mo.- m .nl. v0 1 _.0.0 I N0 . 1 I F0 I H WI3 w. . r 3 3 3 h .I 3 Llw r O osSEESSSBG macoegw £239.93». 30m “0.: CN :0 _Z 00 on. :2 .0 > C. om 3 3 3 3 3 q 3 1 3 3 _ . I I v we W mmm3 mu D3. Im 50 I ®< . 1 ml3 3 3 t 3 L 3 3 P 3 m _. O RESBESSSEB SWd/UOIIOBH SVM SWd/UOIIOBH SVM 102 s03uomum m3ns3om U3o¢ >3xmmz 0:» s3 mama no mmmusmoumm “mm HMDUHm 2coE2w c2365.... 26m 3m... 5 8 _z oo on. as. .o > .3 8 I3 3 3 3 3 3 3 3 3 3 3 O I I m I I 03 wk.- ..... mcmm -- I 2.3 I I 2 It 3 3 3 F 3 3 3 3 3 31 ON 3335300 5333335 pea custom. mucmegw COEmCflC. .501 30:“. CN 3.30 32 00 on. :32 .0 > 3.3. 00 I3 3 3 3 3 31 O I I m 33 - I am ..... I 0.. mm a ..... 3| 8 I m3 5 I I3 3 3 3 3 3 3 3 3 3 3 ON 33 3:5 63.33.389.233. uonoeu SVM u! SIBIGUJ % uouoeu SVM uI SIBIaw % 2:23.2m 3.0.3.93... 263. 8.3“. cN no .2 oo o. :32 .o > .3 om T No l I 8 . No ..... mo co . I N0 3. 6 I I3 3 3 3 r 3 3 . 3 3 o 3333 633863386 m3coEo_m 3.0.3.28... 30m 32E 5 3o _z oo 3.. cs. .0 > _3 om 1 3 3 3 3 3 3 3 3 I 2 2 - I 0< I 33.5 633.386.3236 03. 0— 0m 03 9 0m uouoeII SVM u! sIeIaw % uouoeu SVM uI SIBIGUJ % 103 0 ON An examination of FRTE patterns in the literature has demonstrated. that FRTE can. preserve information. about a sediment's depositional history, in some instances, better than the REE. However, these pattern differences or "trends" appear to be related more to sediment provenance and physical transport than a record of depositional chemical conditions. The second part of the study examined FRTE abundances in a sedimentary rock whose metal abundances had a good chance of being influenced by the chemical conditions during sediment deposition. The bulk shale FRTE patterns demonstrate that there is a spatial record preserved in the Antrim shale over the Paxton quarry. This record is also shown to distinguish between stratigraphic unit C and unit B. The use of SCE does help to clarify the various inputs and their relative importance on the bulk FRTE pattern. The various chemical fractions examined also preserved horizontal records in the shale. The consistency of the FRTE patterns within each chemical fraction suggests that the SCE have consistently divided the shale into operationally defined fractions. The residual fraction, designed to be comprised of resistant detrital and authigenic mineral grains, is the dominant influence on the bulk shale FRTE pattern. This suggests that qualitatively correlating bulk. metal abundances in black shale to organic content may be misleading. This also 104 may explain some of the difficulties in trying to correlate metal abundances in shale to metal abundances in ancient seawater. The residual signature strongly influences the shale's trace metal abundances. However it is largely independent of the ocean's element composition or chemistry, and needs to be removed if trace element abundances in shales are to be linked to ancient ocean chemistry. The consistent FRTE patterns for all stratigraphic units in the oxidizable fraction is difficult to interpret due to the variety of sources for metals in this fraction. However the lack of correlation between organic content and chalcophilic metal abundances suggest that perhaps sulfide minerals are controlling the FRTE pattern in this chemical fraction. Even though the oxidizable FRTE pattern can not reflect relative metal concentrations in pore water, due to pore water processing by organic matter and sulfides, the consistent pattern suggests that if the processing of metals remained consistent during the deposition of the Antrim, then the abundances of trace metals in the pore water did not change over the time period studied. The consistency of the FRTE patterns within the stratigraphic units for the residual and oxidizable fractions suggest equal or no influence from post- depositional processes. The scatter of the FRTE patterns in the moderately reducible fraction indicates some metal repartitioning has occurred. Anomalous spikes in the partitioning of Mn, Co, and Zn also support some 105 post-depositional repartitioning. However, the abundance of Mn and the relative importance of the moderately reducible fraction in partitioning Mn increases with lithologic evidence of higher oxygen levels during sediment deposition. As might be expected, the easily reducible fraction FRTE patterns are quite variable, and Mn abundances do not correlate with lithologic evidence of paleoredox conditions. The jpercentage of’ Mn and Co sequestered in. the easily reducible fraction is highly variable. Since Co is found associated with Mn oxides it is believed that the Mn and_Co are indeed from Mn oxides, and that the easily reducible fraction has probably been seriously affected by post- depositional processes. It is possible that the Mn in this fraction is due to the formation of post-depositional Mn oxides during near surface weathering of the shale. The weakly acid soluble fraction is also overprinted by diagenesis, as evidenced by the scatter in the FRTE patterns in each stratigraphic unit. In addition, the metal partitioning patterns in the weakly acid soluble and easily reducible fractions are identical, which suggests that Mn oxides were also removed to some degree in the weakly acid soluble leach. QEB£I§B_§1§ QQEQLQQIQEE The diverse response of transition elements to various chemical conditions gives them potential as a group to fingerprint different geochemical conditions in sedimentary rocks. The extent to which the FRTE abundances in sedimentary rocks can. be treated as a record of these conditions is an important question. The cycling of metals in sedimentary systems is complex with many aspects not well understood. In addition, the examination of ancient sediments must also consider post-depositional processes such as diagenesis and weathering. Before these questions can be addressed, it first. must be determined. if FRTE abundances retain any sort of record, and if so determine how this record is recorded. From this study it can be concluded that: 1) FRTE patterns from sediment data in the literature has demonstrated that FRTE can preserve information about a sediment's depositional history, and at times, makes distinctions that the REEs do not. However, these pattern differences or "trends" appear to be related more to sediment provenance and physical transport than a record of chemical conditions during deposition. 2) Bulk FRTE abundance patterns are consistent geographically throughout the Paxton quarry which suggests that a record is preserved by the FRTE. ‘ 106 107 3) The bulk FRTE abundance pattern vary vertically among the three stratigraphic units sampled, suggesting that "differences" between the stratigraphic units are recorded. 4) The majority of the metals in the shale were sequestered by the residual fraction (detrital 8 authigenic minerals) 30-100%, and oxidizable fraction (organics & sulfides) 10-60%. Consequently the bulk shale FRTE abundance pattern is controlled by these two fractions. 5) Each of the shale fractions exhibit unique FRTE abundance patterns, that are consistent geographically, but differ for the various stratigraphic units. These differences are in agreement with changes in the bulk shale FRTE patterns. 6) Slight but distinguishable differences in the percentage of metals sequestered in the various stratigraphic units, and the unique FRTE abundance patterns, suggest that the FRTE pattern differences between stratigraphic units is, at least in part, shaped by factors other that detrital input. 7) FRTE abundances from Antrim shale samples about 300 km distant from the Paxton quarry show similar patterns and metal partitioning, which suggests that the patterns may be preservable over relatively large distances. EEIEB£_IQB§ The use of selective chemical extractions can help distinguish between detrital and depositional metal sources. 108 However additional work needs to be done to insure that all organic and sulfide related metals are removed during the selective chemical extractions and not carried over into the residual fraction. Also, since the sulfides and organic matter are of prime importance in understanding reducing environments, they need to be separated. If separation is not possible, then sulfide minerals from a variety of the sample areas should be collected, analyzed, and if possible put through the SCE. Information could be obtained about the ”total" sulfide FRTE pattern in addition to how well the various chemical extractions sequester FRTE in sulfides. This may help in interpreting the shale's oxidizable and perhaps residual FRTE patterns. The interpretation of FRTE ratioed to a standard which is an average must be done with cane. Data in this study was ratioed to an average Post Archean Shale (PAAS) . In order to make interpretations about the relative enrichment or' depletion of elements in samples with respect to a standard based on an average, statistical information is needed about the standard. For example, if the standard deviation about the mean for the standard is low, then small deviations in element abundances between samples and a standard can be statistically significant. However, if the standard deviation about the mean is large, then modest to large deviations in element abundances between sample and a standard may not be statistically significant. Therefore care must be taken in making definitive interpretations. 109 There is evidence that the FRTE do record information about a sediment's history, and a closer examination of the FRTE's ability to record information about chemical depositional conditions needs to be addressed. 110 111 000.0 0000.0 5«00.0 .0«00.0 0000.0 00000.0 0000.0 0000.0 0000.0 0000.0 0000.0 0000.0 5«00.0 0000.0 0000.0 0500.0 0«00.0 5000.0 0000.0 0000.0 0:: 0«00.0 0«00.0 0000.0 0500.0 «00.0 0050.0 «000.0 000.0 0050.0 000.0 «000.0 0050.0 0050.0 0«50.0 0500.0 «000.0 000010 «00.0 0500.0 0050.0 memm 505.0 000.0 000.0 000.0 050.0 00«.0 000.0 005.0 005.0 005.0 mos.o mnm.o chm.o 335.0 snn.o mmn.o nms.o 0«0.0 005.0 005.0 ~03e 000.0 000.0 050.0 050.0 000.0 «00.0 00«.0 .000.0 005.0 005.0 00«.0 000.0 05«.0 005.0 000.0 0«5.0 005.0 000.0 000.0 555.0 cmx 000.0 «0«.0 00«.0 50«.0 000.0 0«0.0 000.0 50«.0 000.0 «0«.0 00«.0 000.0 000.0 00«.0 «00.0 000.0 000.0 00«.0 000.0 00«.0 cmaz 000.0 005.0 005.0 000.0 00«.0 000.0 000.0 000.0 000.0 000.0 000.« 000.0 «00.0 000.0 005.0 000.0 05«.0 000.0 «00.0 000.0 000 000.« «00.0 «00.0 050.0 «0«.0 «50.0 «00.0 000.0 500.0 005.0 005.0 050.0 000.0 000.0 000.0 000.0 000.0 000.0 050.0 005.0 om: 000.0 000.0 000.« .000.0 000.« 0«0.0 00«.0 0«0.0 000.« 500.0 mob.3 mom.m ~m3.m so~.3 ¢m~.¢ com.3 305.3 000.0 000.0. 000.« 00% 00.00 00.00 00.00 00.00 00.00 00.0 00.5 00.00 05.00 00.00 00.00 00.00 00.00 00.00 00.00 00.00 «0.«0 00.00 «0.00 «0.00 mo~3< 0000Bmm0 000030 500030 00 00 50 00 «0 00 00 0m 5m 0m 00 «0 0m 0m 54 0d 0d 030500 Aucmouwm 0:0303 C3 mc03umuusmocoov m3m>0aca ucmEm0w uOnwE "m3wnm x390 3« 04009 112 0030.0 0300.0 00000.0 00500.0 0300.0 00«00.0 00«00.0 0030.0 0030.0 0030.0 0«30.0 «000.0 0«00.0 0000.0 0030.0 0030.0 5000.0 «030.0 0000.0 0000.0 0C2 0000.0 30«0.0 0500.0 00«0.0 0000.0 0000.0 0000.0 0000.0 00«0.0 0500.0 0000.0 0000.0 0000.0 0300.0 0000.0 0000.0 0000.0 0000.0 0000.0 50«0.0 memm 005.0 000.0 0«0.0 000.0 000.0 0««.0 050.0 000.0 035.0 500.0 000.0 530.0 «00.0 005.0 305.0 505.0 055.0 000.0 000.0 005.0 «009 000.0 «00.0 300.0 030.0 00«.0 000.0 000.0 500.0 «0«.0 0«0.0 000.0 000.0 000.0 05«.0 000.0 00«.0 000.0 000.0 «00.0 050.0 cam o>~.o ~>~.o ~m~.o 00«.0 00«.0 «50.0 000.0 «0«.0 50«.0 3««.0 000.0 ««0.0 000.0 050.0 000.0 ««0.0 0«0.0 00«.0 000.0 5««.0 cmmz 0«50.0 «0«0.0 00«0.0 00«0.0 00«0.0 00«0.0 00«0.0 00«0.0 00«0.0 00«0.0 0000.0 00«0.0 00«0.0 0000.0 «0«0.0 05«0.0 5««0.0 0350.0 00«0.0 0««0.0 000 000.0 5«0.3 «00.0 000.0 000.3 005.0 0«0.0 0«0.0 003.0 053.3 3«0.3 003.0 000.0 «00.0 000.3 000.3 0«0.0 500.3 000.0 500.3 om: 000.0 050.0 000.0 00«.3 000.0 000.0 000.3 050.0 530.0 000.0 «00.0 05«.0 0«0.0 «00.0 «0«.0 000.0 005.0 «0«.0 000.0 000.0 000 00.03 00.00 00.03 00.03 05.33 00.5 00.5 00.03 «0.03 00.03 05.33 05.0 05.00 00.03 00.00 0«.03 «5.00 00.33 00.00 0«.00 nemda 00.0« 00.00 «0.00 «0.30 «0.0« 0«.50 «0.00 5«.0« 00.00 00.00 00.00 00.0« 00.0« 03.00 05.00 30.00 00.00 00.0« 0«.0« 00.00 N2m 00038000 000330 500030 00 0U 5U 00 «U 0U 3U 0m 5m 0m 00 «0 00 30 54 0d 00 m3QEmm Aucmouma u200m3 :0 mc00umuucwocoov m3m00mcm ucwaw0m Manna "m0mnm Hmsnfimwm "m 00049 113 xo.s 0m.m wo.~u xo.au an.“ *o.on xm.- *~.o w..au .mmuo azmomum o.mm m.oaa m.mm m.m~ .Nmmn¢ .oov o.moH m.~na .mmmn xflsn n< n.mm m.o~H m.om m.9~ 5.0mm¢e «.mmn o.m~a o.nma .vamn mmm 04mm mqmm 04mm m.oa .¢¢mw~ ~+mafl m.oo o.n~a .vamm mmu nc mo.m o.mm .nm «.ma .omnm m.¢~ n.en H.” xo n4 wv.a o.~ v.m m.n .oqu «.mvfl wm.m ~m.o us 54 «0.0 Hm.o ~¢~.o o.oo« om.m~ om.o nm.o um 54 mmu.o m-.H hbn.o 9.an Hm.¢¢ 5mm.o no~.o mm: 54 *0." x.. an.«du xa.nm «5.g: xo.aau ao.un we." wu.n«u .mmHo azmommm o.mm «.9m ~.mm n.v~ .mvmm¢ .mm~ o.mm o.~ma .mmnv xflsn m4 o.mm m.nm «.mh ~.hn n>q~¢ m.n- m.mna H.mma .mmnn zsm ~.m~ v.nm n.0q m.n .mfla¢m 0.00H mwwm «.590 .mmnn mmu 04 o~.m «.mm .on m.~a .oovnfl o.on o.mv m.m xo m4 No.~ o.H o.v 0.00 .oomv m.mv o¢.m on.n ue ma mo.o oo.a mo~.o 0.5mm om.- mb.o «0.0 um 04 0¢H.o HHo.H va~.o o.m- en.h~ 9am.o mma.o was ma xm.oa *m.HH ao.~u *~.’o *«.a x~.«u «5.g. an.na wm.a .mmHo azmommm b.5m m.vm «.mm m.a~ .mmmmn .oma m.aoa m.~¢a .nmmv xasn n4 h.Hv «.mofl o.nm m.mm m.nmomn o.mha n.0mfl «.mdm .mmnq 25m m.b~ vdmm m.¢« m.« xqwodom n+«m m.nm m.oo~ .NNnv mmu nc‘ v~.0H ¢.mn .on v.¢H .oomcfl m.m~ o.m¢ n.m xo n4 vo.n 0.0 m.q o.oH .ommv w.m~ mm.n mm.n us n4 mm.o om.o mqn.o o.mm~ oo.~H m>.o m¢.H um n4 mon.o H90.H mo~.o m.oHH mo.ma m~¢.o no~.o mos n4 cw so «2 00 mm a: no > He mamemm 0600 C0 mc00umuucwocoov m0mcm E0uuc4 ms» 00 macauowuu m90000> c0 mwocmvcsnm 09mm "0 00049 114 00.0 00.0 0«.5u 03.3 05.3 00.0: 00.0« wo.a 00.0: .0030 8200000 «.0« 0.50 0.003 «.00 .50000 .000 0.033 0.300 .«0«« x350 «0 n.0« «.mm 0.303 5.00 m.~«0nn 0.n«~ 0.003 0.300 .mmn« :00 «.«n 0.00 m.«m 0.0 .00003 0.033 0.003 N+Q30 .000« 003 «0 .05.0 0.00 .«« o.«3 .0000 c.03 0.00 0.0 x0 «0 «3.0 0.0 0.03 0.« .000« 0.50 00.0 00.« ME «0 00.0 05.0 0«0.0 m.«0m 00.00 00.0 o0.3 um «0 003.3 005.3 005.0 «.000 30.«0 000.0 «00.0 003 «0 05.5 00.5 00.0: 00.33 00.0 0«.0I 00.50 00.0I «0.0! 0000 8200000 5.00 0.00 0.003 0.00 .000«0 .000 0.003 5.053 .000« x350 00 5.3« 0.50 «.003 0.00 0.000«0 0.000 0.003 0.000. .0«0« £00 5.m3 0.0m 0Jmm 0.5 «00«03 3.003 0.00 0.003 .m«~« may 00 00.0 0.nn .00 0.0 .0000 «.03 «.00 0.0 xo m0 0«.0 «.m 0.03 0.0 .0000 0.00 «0.0 o5.« us 00 00.0 00.0 003.3 0.«00 05.00 ow.o 50.3 um m0 500.0 000.0 003.3 0.050 33.00 000.0 503.0 003 00 03.0 33.5 00.0: 00.0 00.0: xo.33 «0.53 x«.«3 00.3: .mmHa 9200000 n.n« m.«m 0.003 0.00 .50000 .000 3.003 0.003 .005« x350 30 m.m« 0.00 3.50 m.«~ 0.0«000 0.000 0.503 0.003 .000« 200 0.03 0.00 o.«m 0.0 .00030 0.««3 «.00. mqn53 .000« 000 30 ««.0 «.mm .50 0.m .0300 o.o~ 0.00 0.3 x0 30 00.0 m.« 0.33 0.0 .o«00 0.00 00.0 00.5 MB 30 00.0 00.3 0«5.o 0.000 om.0n 03.0 3«.o um 30 005.3 0«5.3 000.0 0.05 03.00 35«.o 000.0 000 30 am so 32 oo 00 :2 no > 39 030500 005:3ucoo "0 00049 115 0«.0 00.03 x3.«1 00.0 00.3: 00.03 «0.0: .00H0 8200000 3.30 0.00 .0000« .««5 .3030 X30n 50 3.00 0.503 0.3000« 0.005 .0030 :00 0.«0 0.00 q00003 «.00 .0030 00H Nu «3.5 «.05 .00303 0.«0 x0 50. 00.3 0.3 .0«003 0.030 ME 50 30.0 0.005 00.0« 00 50 000.0 «.000 30.0« .003 50 x0.« 0«.03 x0.0 £0.0I 00.0: .mmHa 8200000 0.00 «.00 .«000« .005 .0««0 x300 00 0.«0 3.00 0.0050« 0.005 .5000 250 0400 .«.00 .00«53 «.033 0000‘ .45000 000 00 00.5 0.50 .0000 0.00 x0 00 05.3 0.3 .00003 0.00« 08 00 00.0 0.0503 00.00 00 00 000.0 3.000 00.5« 003 00 x3.0 00.03 «0.0 «0.0 *0.0l .hhHG 8200000 0.0« 5.30 .00000 .000 .00«« x300 00 0.0« 0.003 0.03000 0.0«0 .000« :00 0400 0.00 0.0« w.N .30300 3.000 .000« WMHIHQI 00.5 «.00 .0000 0.03 x0 00 00.0 0.0 .0050 0.00 08 00 00.0 0.00« 00.03 00 00 000.3 5.50 05.00 003 00 00 :0 00 a: 39 030600 003:3»:00 "0 00040 116 xv.u xm.ma xv.«u xm.vu xv.v xo.ml wu.05 *a.OH aa.~u .mmHa BZNUmmm m.mm m.~m m.ooa m.mH .mnman .mma v.om ~.oma .Nmmc xasn mo n.0m n.moa «.mm m.na «.mammn a.mma m.¢ma v.5am .mmmv 23m wq¢m mdmm nqu a.n .emwhd m.mm m.~m n.mo~ .mmmv mmu mo vm.m 0.0m .«v v.aa .ombaa «.mm n.mv m.e xo «0 vm.~ m.a u.w ¢.~ .ommn m.oH ~n.n vm.n we «0 ov.o Hm.~ omm.o o.¢m~ om.oa ob.o on.H um no omv.o qnn.a 0««.0 ~.vm Hm.¢a vm¢.o mma.o mm3 mo xa.a xa.n x~.mu «H.51 wa.nl *h.o| «0.65 *«.o xo.~| .mmHo 82mommm m.mma o.moa o.mmm ~.Hn .mahom .oom w.moa m.van .mmwv xasn Ho m.hma n.maa ~.oHN o.mm n.mndmv c.0ma m.omd m.nvn .vmav 25m 0.00 H.mv ~.~HH w.HH .owmnm ~.mm «.QHH m.a~n .¢mHv wwu Ho om.mm 0.«0 .mm m.vH .oebma ~.H~ m.om v.0a x0 H0 vo.mm v.a «.m m.a .oomv «.mn aw.m 05.x uE H0 Hm.o mm.~ mmm.o m.mmv oo.mH mb.o ~m.~ um Ho mom.o 5mm.~ han.o m.mma No.5H moo.o Hmn.o mm: H0 xm.oa an.oa xa.vl xw.dd av.dl an.d . «0.5n. xn.oa »«.nu .hhHa BZHUmmm o.mm m.>> o.m¢a m.mm .cmobn .mem m.mm «.mma .mnmc xaan mm m.mv n.mm m.mna m.Hn H.mammn «.mvm H.wna m.~ma .mmov zbm w.am nwdm m.mn m.> .ommwd u.maa m.nm n.wna .omov man an vn.m m.om .bn w.ma .ommoa m.ha b.wv o.n x0 mm mo.n o.m m.wa m.m .ommm o.nw mh.o mn.n ME wm No.0 mm.n mmb.a o.mav mn.ma mv.o mh.o um mm mmv.a vn~.a vnm.o H.Hb mm.vn wmn.o Haa.o was mm ca 50 flz 00 mm a: no > fie mamEmm nmzcfiucoo no mqm.vn m.HHH m.mma m.~H n.mmo- «.mn o.~ma ~.-n .Onmm Sam A.md n+w~ mqmm odo .nwwafl o.mn n.o> «.omm .onmm mmutmu ma.ma m.mn .aw v.m .ovmh ~.ma ~.¢n o.mH xo mu o¢.v m.¢ «.ma m.~ .ooan o.nH va.m ~n.ma as no ~>.H o¢.v m.o mu¢- m~.> oo.H mh.v um mu vmm.a moc.~ Hm~.o ~.hm ca.oa nan.o who.o maz no ««.u rm.nu xv.a«u xu.vauwu.n ac.uu ««.05 *o.v xo.~u .mmna azmommm o.ano n.5m m.om~ ~.b~ .emonv .Hva m.~oa w.e0n .mamv xflsn v0 o.a>m m.nm m.mo~ ~.nm o.movv¢ m.mna w.m>a h.oan .moav 25m mqanfl nqmn \a.wnfl m‘m .voanm m.>m m¢woa o.00n .moav mmu «u oo.vmm m.om .bm w.oa .oamva o.a~ w.mm H.m x0 «0 em.mva m.a o.aa v.~ .omem ~.>H mw.m mH.m he «0 -.H o~.H ona.o m.oma om.o ov.o mm.o um v0 ooo.m mvn.a ana.o m.m~ an.a omH.o mmo.o was «0 an :0 Hz 00 mm a: no > He mHmEmm nmscflucoo no mqm49 118 an.| xn.u xo.n| xa.vauxo.l *m.vd.u xo.ba wo.~« an. .hmHo 82mommm m.am m.nn m.am n.ma .mqhan .Hma m.mm m.mmH .nmmm xasn bamazo m.Hn ~.nm m.mv o.¢H m.hw¢am m.moa w.ana ~.oma .Hmmn 23m n.ma n.m m.~H1 «.0 .mou¢a 0.«0 «MAN \M.m¢d1 .Hmmm mmu nflwazo «0.x v.5m .mu o.HH .oeaaa o.mm m.mv b.n x0 hamazo om.n «.0 m.n m.a .omnv o.m~ ov.m mn.n us namaxo mm.o vv.a mvv.o o.¢mm m~.a~ mn.a mm.o um banazo Hmo.o mov.a wom.o m.nm~ om.nm m¢o.a 0¢H.o mm: nanazo «n.5mu an.> xh.hn xv.oa ao.«l ««.vc xn.«v *o.d« *~.I .hhHa BZmommm ~.mvm H.m~H w.mom ¢.va .mmmem .ooa m.HHH o.mmn .mawm xaan mo n.moa >.mna o.nma m.ma m.wo~v~ v.noa m.mma m.mmn .bovn 23m m.mm o.H¢ m.vm m.~ .mvnafl n.aw m.moa w.oun .hown mwu mo 00.5 m.am .ms m.HH .onam v.va m.mn «.wa x0 mo vm.>m v.H «.5 ~.H .omnn «.ma mm.> Ho.oa us No no.0 mm.n 5mm.o m.hma m~.¢ mm.o ~n.n um mo vam.o ovn.~ vam.o a.h~ oa.m wmn.o nvm.o .mm3 mo xm.d«w wv.u xa.uu *H.dn an.vd *v.ml *N.vv x«.ud xv.«n .hhHQ BZWUmmm n.mvma m.vma m.am~ m.ma .hmhan .oea ~.nna n.vmv .m¢ov stn mo ~.vamma n.voa m.am~ m.ma n.0mnwm «.mna H.~mH «.mnm .Hmmm 25m w+~n¢ nqwm m.w¢a w.m .mmdou ¢.mm n.wnd \waoam .awmn mmu am! 00.0mmma N.mm .mm v.0H .ovmoa 0.0« m.m¢ m.~H x0 mo vw.amm o.n o.mH o.~ .omom «.ma ~e.m No.na ME mu ~H.~ mm.n mmn.o ~.oo~ mh.v ~m.o mn.~ uw mo mno.v Hm~.n oom.o o.ma Hm.v owa.o mwn.o mm3 mu cN so “z 00 mm a: no > we mHmEmm Umscwucoo "m mqm