UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Composition and structure of titanian andradite from magmatic and hydrothermal environments Hilton, Elizabeth 2000

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2000-0423.pdf [ 7.31MB ]
Metadata
JSON: 831-1.0089519.json
JSON-LD: 831-1.0089519-ld.json
RDF/XML (Pretty): 831-1.0089519-rdf.xml
RDF/JSON: 831-1.0089519-rdf.json
Turtle: 831-1.0089519-turtle.txt
N-Triples: 831-1.0089519-rdf-ntriples.txt
Original Record: 831-1.0089519-source.json
Full Text
831-1.0089519-fulltext.txt
Citation
831-1.0089519.ris

Full Text

Composition and structure of titanian andradite from magmatic and hydrothermal environments by Elizabeth Hilton B.Sc. (Hons), Saint M a r y ' s University, 1998.  A. T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF M A S T E R OF S C I E N C E in T H E F A C U L T Y OF G R A D U A T E STUDIES (Geological Sciences Division, Department o f Earth and Ocean Sciences) We accept this thesis as conforming to the required standard  T H E UNIVERSITY OF BRITISH C O L U M B I A May, 2000. O Elizabeth Hilton, 2000.  In p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f the requirements f o r an a d v a n c e d d e g r e e a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d b y t h e h e a d o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s understood t h a t copying or p u b l i c a t i o n of t h i s thesis f o r f i n a n c i a l gain s h a l l not be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n .  Department o f  (fariV* Oceav\ Sciences,  The U n i v e r s i t y o f B r i t i s h Vancouver, Canada  Date  Kft^  JOOO.  Columbia  Abstract  Titanian andradite provides a wealth o f information about the environments in which they form. Zippa Mountain pluton and the Crowsnest volcanic rocks provide examples o f titanian andradite formed indifferent environments (e.g., magmatic, skarn, volcanic, and hydrothermal). Electron-microprobe, petrographic, and geochemical analysis, coupled with X-ray techniques were used to determine the composition, structure, site occupancies, and to discriminate between titanian andradite formed in different environments. Site occupancies, determined from the study samples, are as follows: C a  and N a are always assigned to the X site; M n  2 +  +  2 +  is preferentially assigned  to the X site, but may also be at the Y site; A l , M g , C r , V , T i , and F e 3 +  assigned to the Y site; F e and H  4 +  3+  2 +  3+  3 +  4+  2 f  are always  may be assigned into the Y site or the Z site; and S i , Zr ^, 4 +  4  are always assigned to the Z site. The titanium substitution mechanism may be  via the T i M g F e . 2  3+  exchange component, indicative o f octahedral T i substitution and  octahedrally controlled cell volume. Chemical zoning o f magmatic and skarn titanian andradite is irregular. Garnet from volcanic samples have irregularly zoned cores and regularly zoned rims as does the magmatic cumulus sample. Hydrothermal samples show regular chemical zonation. E P M A data reveals titanian andradite zoning patterns within different rock types and different formation environments. Dark zones in the volcanic garnet contain more T i 0 and A 1 0 and less F e 0 than lighter zones. Melasyenite samples show positive 2  2  3  2  3  correlation between T i 0 and F e 0 . Conversely, pyroxenite samples show regular 2  2  3  zoning with darker zones having more A 1 0 and less F e 0 than lighter areas. In these 2  3  2  3  Ill  samples, T i 0 and F e 0 are negatively correlated. In terms o f Thompson components, 2  2  3  magmatic samples have small norms, negative M g C a - , very small H 4 S i - component, and positive F e M g - , whereas hydrothermal samples have positive H 4 S i - components, zero to slightly positive FeMg-, and large norms. O n average, skarn samples have equal amounts o f both T i S i - and TiMgFe _ 2  3+  components and are therefore intermediate between  magmatic and hydrothermal samples. Oxygen fugacify and activity o f silica are correlated by T i M g F e . 2  formed under low f  Q2  3f  and T i S i - components and indicate that the magmatic samples  conditions, skarn samples inherited \hef  Q2  signature by interaction  with early magmatic fluids, whereas the hydrothermal sample crystallised from a more evolved fluid which had a h i g h e r ^ and a  s i 0 2  .  Table of Contents  Abstract List of Figures  vi  List of Tables  viii  Acknowledgments  ix  Chapter 1: Introduction  1  1.1 General Introduction  1  1.2 Purpose  2  Chapter 2: Literature Review o f Titanium Andradite  4  2.1 Overview of Titanian Andradite  4  2.2 Anisotropy  5  2.3 Zoning in Titanian Andradite.....  6  Chapter 3: Sample Suite and Petrography  10  3.1 Geological Setting and Sample Suite  10  3.1.1 Geological Setting...,  10  3.1.2 Sample Suite  12  3.1.3 Whole R o c k Geochemistry and Geochemical Methodsl2 3.2 Petrography....  14  3.2.1 Sample Description  14  3.2.2 Petrography  16  3.2.3 Scanning Electron Microscopy  20  Chapter 4: Chemical Characteristics and Representation o f Ti-Andradite....56 4.1 Chemical Mineralogy  56  4.1.1 Electron Microprobe Analyses  56  4.1.2 Compositional Zoning  56  4.2 Representation of Analyses  64  4.3 Summary  :  71  Chapter 5: Other Chemical Techniques Used to Describe Ti-Andradite  101  5.1 Introduction  101  5.2 Wet-chemistry  101  5.3 FTIR and Estimates o f O H  101  5.3.1 Literature Review o f O H Site Occupancy  101  5.3.2 FTIR Methods  103  :  Chapter 6: X-ray Diffraction Analysis  106  6.1 Introduction to Diffractometry  106  6.2 The Andradite Unit Cell by Powder Diffraction  107  6.3 Single Crystal Diffractometry  109  6.4 Summary  111  Chapter 7: A n Analysis o f Site Occupancy in Ti-Andradite  116  7.1 Recapitulation o f Andradite Crystals  116  7.2 Literature Review o f Site Occupancy i n Ti-Andradite.  116  7.3 Results and Ideas from This Study  118  Chapter 8: General Conclusions and Petrogenesis  126  References  130  Appendix A E P M A Error Analysis  137  Appendix B Cation Normalisation Routine: Algorithm and Matlab Code... 163 Appendix C C D R o m o f Complete E P M A Tables  back cover pocket  vi  List of Figures  Figure 1  Zippa location map ( B C )  3  Figure 2  Zippa sample location map  24  Figure 3  C l i f f sample location map  25  Figure 4  Glacier sample location map  26  Figure 5  Bartnick sample location map  27  Figure 6  Geochemical data plotted as alkali-silica, A F M diagrams  30  Figure 6 (cont'd) Trace element variation diagrams  31  Figure 7  Polished thin section pictures  32  Figure 8A  S E M images of zoning o f Zippa magmatic samples  43  Figure 8B  S E M images o f zoning o f cumulus and Crowsnest samples  46  Figure 8C  S E M images o f zoning o f dyke samples  49  Figure 8 D  S E M images o f zoning o f skarn samples  51  Figure 9  X-ray maps for Z M 3 9 B - B  55  Figures 10-24  Line traverses plotted  76  Figure 25  S E M images showing line traverses on grains  91  Figure 26  T C S component plots  96  Figure 27  T C S compilation plots  99  Figure 28  T C S compilation plots for norm o f the vector  100  Figure 29  Comparison o f duplicate FeO analyses  105  Figure 30  Z95-1 2 theta correction factors  113  Figure 31  Measured cell volume plots  115  Vll  Figure 32  Portion o f the garnet structure projected down the c-axis  Figure A l  E P M A analyses for the non-calibration standard plotted over time.. 140  Figure A 2  Grid o f analyses on the non-calibration standard  Figure A 3  Analytical uncertainty plotted against the IS variation for E P M A analyses for the non-calibration standard  Figure A 4  144  145  E P M A analyses o f Andr point 4 from the non-calibration standard, over time  Figure A 8  143  E P M A analyses o f Andr point 3 from the non-calibration standard, over time  Figure A 7  142  E P M A analyses o f Andr point 2 from the non-calibration standard, over time  Figure A 6  141  E P M A analyses o f Andr point 1 from the non-calibration standard, over time  Figure A 5  122  146  E P M A analyses of Andr point 5 from the non-calibration standard, over time  147  Vlll  L i s t of Tables  Table 1  Sample suite from Zippa Mountain and related assemblages  .23  Table 2  Geochemical data with norms  28  Table 3  Petrography  Table 4  Representative microprobe analyses  72  Table 5  Zoning patterns summary  94  Table 6  Wet-chemical analyses o f FeO in select garnet samples  104  Table 7  Calculated unit cells for selected andradite crystals from this study. 114  Table 8  :  37  J Widely accepted cation site occupancies  123  Table 9  Cation site assignments and totals  '.  124  Table A1  E P M A analyses for calibration standards over time  148  Table A 2  Calculated wt% for the andradite non-calibration standard  153  Table A 3  E P M A analyses for the andradite non-calibration standard over time  •.  154  Table A 4  E P M A analyses for the andradite non-calibration standard grid  Table A 5  E P M A analyses for the andradite non-calibration standard, organised by point, over time  Table A 6  155  156  Calculated mean and standard deviation for the andradite non-calibration standard  162  IX  Acknowledgments  The author wishes to thank Dr. Kelly Russell, Dr. Greg Dipple, and Dr. Lee Groat for guidance, supervision, and improvements o f this work. Special thanks to Dr. Mati Raudsepp for guidance and helpful suggestions for instrumentation. Dr. Ian Coulson is thanked for reviewing the thesis and for many helpful suggestions. Most o f all I wish to thank my husband, my sisters, and parents for their support and encouragement.  1  Introduction  1.1 Introduction Titanian andradite forms in both magmatic and hydrothermal environments and is most commonly associated with silica-undersaturated rocks (Howie and Woolley 1968). Titanian andradite forms an important link between magmatic and hydrothermal environments. Variations in its composition can be used to track variables such as f  Q2  and a  S i 0 2  , making it an essential phase for studying the chemical dynamics o f magmatic  and hydrothermal systems and also for establishing records of fluid evolution between the two systems (Russell et al. 1999). Occurrences o f purely magmatic titanian andradite from western Canada include the Crowsnest volcanics, Alberta (Dingwell and Brearley 1985) and the Zippa Mountain pluton, British Columbia (e.g. Lueck and Russell 1994) (Figure 1). Andradite from both these localities is examined in this thesis. The Crowsnest volcanics contain examples o f titanian andradite formed in extrusive alkaline igneous rocks (Dingwell and Brearley 1985). The garnet occurs as phenocrysts along with aegirine-augite, sanidine, analcite, and rare plagioclase in trachyte and phonolite flows, agglomerates and tuffs (Dingwell and Brearley 1985). The garnet is chemically zoned with T i and Fe contents decreasing from core to rim (Dingwell and Brearley 1985). The Zippa Mountain pluton is located in the Iskut River map area, in northern British Columbia about 10 k m southwest o f the confluence of the Iskut and Craig Rivers. The Zippa Mountain pluton (hereafter Z M P ) is roughly 3.5 by 5 k m in size and has an elliptical shape (Lueck and Russell 1994). Chemically, the pluton is strongly silica-  2  undersaturated and alkaline (Coulson et al. 1999). Titanian andradite occurs as a primary magmatic phase in all rock types in the Z M P including syenite, melasyenite, and pyroxenite (Lueck and Russell 1994; Coulson et al. 1999). The pyroxenite contains up to 30% titanian andradite, melasyenite contains 5 to 20%, and syenite contains 5 to 10% (Lueck and Russell 1994). Trachytic K-feldspar, aligned aegirine-augite crystals, titanian andradite and pyroxene-rich layers define a fabric within the syenite and pyroxenite which is interpreted as cumulus in origin (Coulson et al. 1999). A t Zippa Mountain, titanian andradite also occurs in skarn. The skarn is formed by infiltration o f magmatic volatiles into Paleozoic calcite marble and contains garnet o f both grossular and andraditic compositions (Jaworski and Dipple 1996). 1.2 Purpose A suite o f hand samples and polished thin sections from Z M P , including both magmatic and skarn samples, as well as a volcanic sample from Crowsnest, Alberta were analysed. The purpose o f this study was to determine the composition, chemical zonation patterns, and structure o f titanian andradite from magmatic and hydrothermal environments.  Observations and data were collected by electron probe micro-analysis  ( E P M A ) , whole rock geochemical analysis, infrared spectroscopy, and X-ray diffraction. Chemical compositions were compared against environment o f formation. The effect of O H in natural titanian andradite on substitution mechanisms was also studied. The role o f titanium in the crystal chemistry o f natural titanian andradite was determined; specifically the site occupancy and valence(s).  3  Stikine Terrane 1 Butterfly Pluton 2 Zippa Mountain 3 Galore Creek 4 Ten Mile Creek 5 Rugged Mountain Quesnel Terrane 6 Duckling Creek/ Lorraine Complex 7 Mount Polley 8 Rayfield River 9 Kamloops Syenite 10 White Rocks Mountain 11 Averill Pluton 12 Kruger Mountain 13 Crowsnest Volcanics Terranes Wrangellia ] Stikinia ]  Quesnellia  |  Cache  Creek  o  400  km  Figure 1. Map of the British Columbia cordillera showing Mesozoic silica-undersaturated alkaline plutons, and the Cretaceous Crowsnest Formation, Alberta (after Lueck and Russell 1994).  4  Literature Review of Titanian Andradite  2.1 Overview of Titanian Andradite A simplified chemical formula for garnet is X Y (ZO^, i  where X=Ca., M g , Fe ", 21  2  M n ; 7=A1, F e ' , M n , V "', T i ' , C r ; and Z= Si. In titanian andradite, the Z site is 2 +  3  3 +  3  4  3 f  silica-undersaturated which allows for substitution, or coupled substitutions, o f cations such as T i . The structure consists o f alternating Z 0 tetrahedra and Y 0 octahedra which 4  6  share corners to form the three-dimensional framework (Schwartz et al. 1980). The X cation is coordinated by eight oxygens and forms a third kind o f polyhedron, a distorted cube or a triangular dodecahedron. Grossular and andradite garnets are members o f the grandite garnet series. Many andradites have compositions fairly close to the theoretical end-member. Grossular is the next common component because o f the complete solid-solution that exists within the grandite series. Andradite forms a solid-solution series with the chemical endmembers schorlomite and morimotoite. Schorlomite is a species o f garnet containing more than 15 wt.% T i 0 (Peterson et al. 1995). The X site o f schorlomite is almost 2  completely filled with Ca; the Y site contains T i , F e 4 +  filled by Si and F e and F e 31  2+  3+  and F e , and A l ; and the Z site is 2+  (Peterson et al. 1995). Morimotoite is a titanian garnet  which contains more than 50 mol.% o f the morimotoite component, C a T i F e S i 0 , 2 t  3  which is derived from end member andradite by the substitution T i + F e  3  2+  = 2Fe  1 2  3+  ( H e n m i e t a l . 1995). The mechanism by which T i substitutes into the andradite structure remains controversial and is examined in detail in Chapter 7. Whereas previous workers  5 indicated a direct T i - Si substitution (e.g. Manning and Harris 1970; Weber et al. 1975b) more recent work by Armbruster et al. (1998) suggest a complex substitution whereby A l is concentrated on the octahedral site, and T i ' can be either octahedrally or tetrahedrally 4  coordinated, but with O H - , F e , or F e 3+  2!  fdling most o f the tetrahedral vacancies. A l l  titanian andradites have an expanded tetrahedron compared to other andradites because of the hydrogarnet substitution or because of tetrahedral iron incorporation (Armbruster etal. 1998). 2.2 Anisotropy Anisotropy is visible in titanian andradite optically, with the use o f conventional optical microscopy. B y definition, an optically anisotropic garnet cannot be cubic, however techniques which have higher resolution than the optical microscope, such as X-ray diffraction, may indeed indicate an overall cubic habit (Allen and Buseck 1988). Birefringent zones represent intermediate members o f the grandite series whereas isotropic zones are closer to end-member compositions (Ivanova et al. 1998). Early workers commonly attributed anisotropy in andradite to strain or twinning (Chase and Lefever 1960) and lamellae with compositional differences were believed to have been formed after growth (Hirai et al. 1982). More recent workers attribute anisotropy to structural inhomogeneities related to the mechanism o f growth (Gali 1983; A k i z u k i 1984) and most commonly to cation ordering o f F e  3 +  and A l at the octahedral sites  (Akizuki et al. 1984; A k i z u k i 1984). Rossman and Aines (1986) were the first to propose that F e - A l ordering was not the cause o f birefringence in garnet, but that it was the 3+  result o f low-symmetry distribution o f O H groups. Authors in the late 1980's attributed anisotropy to some combination o f strain, cation ordering, and noncubic distribution of  6  O H groups (Allen and Buseck 1988; Kingma and Downs 1989; Hatch and Griffin 1989) and even suggested that low symmetry may be the result o f temperature-induced orderdisorder phase transformation (Hatch and Griffin 1989). Ivanova et al. (1998) suggest that inhomogeneity results from the coexistence o f two types of layers with different F e 7 A l ratios and state that birefringent zones themselves contain very fine zones, 3  usually intermediate in composition within the grandite series. 2.3 Zoning in Titanian Andradite Zoning is found in many minerals, including titanian andradite. Normal zoning in andradite shows an increase in Fe from core to rim, whereas reverse zoning has a decrease in Fe and an increase in A l from core to rim (Vlasova et al. 1985). Zoning can form as a result o f either changing temperature (Vlasova et al. 1985) or external forcings at the site o f the crystal growth (Jamtveit 1999) which include variations i n oxygen fugacity, pressure, p H , and fluctuations in magmatic water content and storage depth (Holten et al. 1997). Oscillatory zoning is a primary growth texture which is commonly superposed on longer-scale zoning, either normal or reversed (Shore and Fowler 1996). Oscillatory zones are narrow with sharp contacts which may indicate either rapid crystal growth or rapid changes in hydrothermal solutions (Lessing and Standish 1973). T w o explanations for the formation o f oscillatory zoning include; extrinsic mechanisms, such as physical or chemical changes within the bulk system that are independent of local crystallisation, or intrinsic mechanisms which link crystal growth to purely local phenomena (Shore and Fowler 1996). In general, small-scale growth mechanisms are responsible for oscillatory zoning, such as u.m-scale large amplitude variation between compatible and incompatible elements (Shore and Fowler 1996). Adsorption o f minor or  7 trace elements has also been proposed as a mechanism for sector zoning (Shore and Fowler 1996). Many workers have studied zonation patterns in titanian andradite. Gomes (1969) found a direct correlation between colour and T i concentration, which was corroborated by Dingwell and Brearley (1985). However, Lessing and Standish (1973), working with samples where 90% o f zones were anisotropic, state that darker zones have higher A l concentrations, which was later refuted by Murad (1976) who found that birefringent zones have higher A l concentrations. Murad (1976) also found that isotropic, andradite-rich zones require a higher temperature o f formation than do birefringent zones based on increasing grossular content with falling temperature in granditic garnet. Samples studied by Gomes (1969) displayed cores that tend to be rich in T i and poor in Si and Fe relative to the rims. He argued that T i replaced both S i and Fe, and showed that the unit cell increased in proportion to the T i content. Cygan and Lasaga (1982) found that Fe increases and M g decreases core to rim, C a and M n exhibit a slight zoning at the edges of grains, whereas S i , A l , and T i remain unchanged. Crowsnest samples (Dingwell and Brearley 1985) have very different zoning patterns which show M n , A l enrichment and T i , Fe depletion in the core, at the boundary o f the core M n and A l decrease with an increase i n T i and Fe, whereas the r i m exhibits oscillatory zoning superimposed on normal zoning with M n and A l increasing and T i and Fe decreasing toward the edge. However, Hickmott and Shimizu (1990) suggest that T i may be less soluble in an intergranular fluid which may lead to slow transport and therefore enrichment near crystal rims during rapid garnet growth. Commonly, the distribution o f  8 M n displays a bell-shaped profile in garnet (Hiekmott and Shimizu 1990; Nakano and Ishikawa 1997). Kerr (1981) describes two different zoning patterns in garnet. The first has an Mg-rich core with Fe, M n , and C a increasing at the margin; whereas the second pattern has a Ca-rich core with other elements antipathetic. Smith and B o y d (1992) state that the most common pattern o f zonation shows C r decreasing from core to rim which could form by either growth o f Cr-poor garnet during cooling at constant pressure, or by diffusion during a pressure decrease. Badar and A k i z u k i (1997) describe two types o f chemical zoning occurring during crystal growth: the first is nearly homogeneous in chemical composition and lacks fine growth lamellae; and the second consists o f fine growth lamellae with various chemical compositions. Ivanova et al. (1998) describe garnet with isotropic zones o f 3 to 4 urn thick that alternate with birefringent zones 10 to 15 u.m thick. These authors state that fine growth zonality originates from non-stationary growth dynamics, and that compositional zoning is the result o f abrupt or continuous changes in the composition o f the solution, as well as in P-T conditions. Zoned titanian andradite also forms in the skarn environment. Concentric zoning is common in both magmatic and hydrofhermally grown crystals (Jamtveit 1999). Vlasova et al. (1985) describe three types o f zoning in skarn garnet: the first is defined by F e 0 contents and is characteristic o f true calcareous skarns; the second is characterised 2  3  by a grossular-rich component which contains significant pyralspite components and varying contents o f C a , M n , and Fe; and the third type contains pyrope components ranging from 4 to 25 m o l % which show reverse zoning.  9 Jamtveit et al. (1993; 1995) state that zoning results from changes in the hydrothermal fluid composition at the site o f garnet growth. These authors describe epitaxial growth o f andradite on preexisting garnet o f higher grossular compositions; a rim ward decrease in M n , T i , Zr, and A l ; and state that the growth rate of andradite rims was larger than the cores. Jamtveit and Hervig (1994) report garnet rims enriched in light R E E ' s , A s , W , M o , Fe, but depleted in Zr, Y , T i , and A l , and state that isotope signatures are controlled externally during all stages o f crystallisation. Jamtveit (1999) and Jamtveit et al. (1995) state that Al-Fe partitioning between garnet and solution is sensitive to variations in temperature, p H , oxygen fugacity, and CT activity. Zonation may be the result o f variable concentration gradients in the pore-fluid near the garnet fluid interface, and variations between layers are controlled by growth rather than by dissolution processes (Jamtveit 1999).  10  Sample Suite and Petrography  3.1 Geological Setting and Sample Suite 3.1.1 Geological Setting The Zippa Mountain Igneous Complex ( Z M I C ) is part o f an alkalic magmatic arc plutonic association, characterised by silica-undersaturated intrusions sometimes associated with silica-saturated intrusions (Lang et al. 1995). Both types o f intrusions were derived from similar sources and formed in similar tectonic settings (Lang et al. 1995). These complexes were emplaced in the Canadian Cordillera o f British Columbia between 210 and 200 M a based on U - P b dating o f zircon (Lang et al. 1995) and include the Averill (Keep and Russell 1992) and Rugged Mountain (Neill and Russell 1993) plutons. A n age o f approximately 210 M a is recorded by U-Pb zircon dating for the Z M P (Jaworski and Dipple 1996). The silica-undersaturated complexes are characterised by associated pyroxenite and syenite, are generally compositionally zoned, and contain aegirine-augite, K-feldspar, biotite, igneous titanian andradite, titanite, and apatite as major phases (Lang et al. 1995). The Z M P grades inward from pyroxenite, which forms the margin o f the pluton, through melasyenite to a core o f K-feldspar syenite (Lueck and Russell 1994). A t the border o f the pluton, pyroxenite forms lenses in calc-silicate rock, and to date five wollastonite localities have been found both within and peripheral to the Z M P . Vishnevite-cancrinite occurs as cm-sized aggregates which weather white in pegmatitic syenite and is pseudomorphic after leucite (Lueck and Russell 1994). The pluton is cut by several different types o f dykes, some o f which have well-developed contacts with the  11  plutonic rocks, and others which have ambiguous contacts, such as porphyritic quartz •if-  dykes and diorite dykes respectively (Lueck and Russell 1994). Previous work by Coulson et al. (1999) proposes that the Z M P is an alkaline intrusion o f syenite and pyroxenite formed from a mantle-derived magma with lowfo  2  and  <3  s i 0 2  .  Coulson et al.  (1999) state that a single pulse o f magma entered a shallow level chamber and that the source o f the magma had affinities with arc-type magmas related to subduction. The parental magma began to fractionate clinopyroxene, then K-feldspar, resulting by physical sorting in side wall, marginal pyroxenite and roof zone syenite. The core continued to fractionate resulting in increased volatile contents and the crystallisation o f feldspathoids. The residual magma buoyed to the roof and partly invaded the syenite to form vishnevite. The pluton intrudes Paleozoic metasediments o f the Stikine assemblage and Triassic volcanics o f the Stuhini Group (Jaworski and Dipple 1996). Wollastonite skarn formation requires a calcite-rich protolith, a high S i 0 content (either from the protolith 2  or from fluid infiltration), and one or both o f high temperature and low C 0 activity 2  (Jaworski and Dipple 1996). Even though the Z M P is silica undersaturated, the fluids carried enough dissolved silica to cause wollastonite skarn to form (as evidenced by the lack o f calcite), but the critical factor for its formation was probably the high temperatures resulting from emplacement o f syenite along with incorporation o f marble xenoliths into the margin o f the intrusion which would further increase the temperature (Jaworski and Dipple 1996). Skarns result from extensive fluid infiltration into marble xenoliths within the pyroxenite border phase (Jaworski and Dipple 1996). The Paleozoic strata adjacent to the pluton are intensely metamorphosed and include calcsilicate and  12 marble (Jaworski and Dipple 1996). The calcsilicate is fine-grained, green, and contains diopside, grossular, biotite, titanian andradite, K-feldspar, and wollastonite (Jaworski and Dipple 1996). The marble is light green to gray, is composed o f recrystallised calcite, and represents a limestone protolith that was unreactive during metamorphism (Jaworski and Dipple 1996). 3.1.2 Sample Suite Nineteen samples are used to represent the full spectrum o f occurrence o f titanian andradite at Zippa, including magmatic (in syenite, melasyenite, and pyroxenite), skarn, and dykes (Table 1). Sample locations are shown in Figures 2-5. From each sample a slab with representative mineralogy was cut and trimmed using a diamond saw and made into polished thin sections. 3.1.3 Whole Rock Geochemistry and Geochemical Methods Whole rock geochemical analysis o f plutonic rock samples was performed using fresh, uncontaminated samples. R o c k chips were produced in a steel-faced crusher, and passed through twice to further reduce the chip size. The samples were subsequently powdered to about 200 mesh using a tungsten carbide shatter box. A total o f nine magmatic samples from the Z M P were analysed for major and trace elements, ferricferrous iron, and water; Determined trace elements include G a , N b , Pb, R b , Sr, T h , U , Y , Zr. Geochemical data are listed in Table 2 along with calculated C I P W normative mineralogy. Major and trace elements, and loss on ignition (L.O.I.) were measured by X-ray fluorescence spectrometry ( X R F ) at M c G i l l University using an automated Phillips P W 2400 spectrometer with a rhodium X-ray tube operating at 60kV, five X-ray detectors, five primary beam filters, eight analysing crystals, two fixed channels for  13  simultaneous measurement o f N a and F, and a 102 sample autochanger. The precision for silica was within 0.5%, for the other major elements, within 1%, and for trace elements within 5%. Total iron was determined by X-ray fluorescence and FeO was determined using ammonium metavanadate titration. Water was determined at 105°C. Ferrous detection limits were 0.01% and water detection limits were 0.01 and 1.0% for H 0 ' and H 0 respectively. +  2  2  The Z M P rocks are alkaline, except for one sample o f pyroxenite which plots just inside the subalkaline field which is indicative o f fractionation (Figure 6). The cumulus sample (1C-98-ZM39B) plots in the alkaline field (Figure 6a). A l l other samples are nepheline normative and have greater normative orthoclase. The A F M diagram (Figure 6b) shows that the two pyroxenite samples and the cumulus sample are F e O enriched, whereas the other magmatic samples plot along a well defined trend from syenite, through trachytic syenite, to melasyenite with increasing F e O enrichment. Pyroxenite samples which show FeO enrichment are indicative o f fractionation and possibly contamination by hydrothermal calc-silicate rich fluids, as they are wollastonite and also diopside normative. The trends for trace elements are shown in Figure 6c. Trends for R b , B a , and Sr trace elements show scatter. Other trace elements (e.g. N b , Z r , and Y ) show somewhat less scatter as most samples lie on a linear trend from syenite through pyroxenite. Y has a negative correlation with S i 0 for magmatic samples. Z M 3 9 B plots separately from 2  the other samples and has more N b , Z r , and Y (compatible), but less Sr, R b , and B a (incompatible) than the other samples. Sr and B a show large variations in concentration both between rock types and samples.  14 3.2 Petrography 3.2.1 Sample Description This section starts with a description o f the hand samples from the nineteen samples used in the present study. IC-98-ZM-42 is a dark green, medium grained pyroxenite containing pyroxene grains 3 to 4 mm in length. M i n o r (less than 10%) K feldspar is also present. JC-98-ZM-26 is a gray, fine to medium grained pyroxenite with visibly zoned titanian andradite which range in size from 3 mm up to 1.3 c m in diameter. Garnet is dark purple to black in color and sub- to euhedral. White calc-silicate composes about 10% o f the rock and is usually found associated with the largest o f the titanian andradite. IC-98-ZM-8 is a medium grained, equigranular melasyenite containing up to 6 mm white to pink K-feldspar, up to 5 mm long green-black clinopyroxene laths, black, anhedral titanian andradite, and 5 m m biotite. IC-98-ZM-16 is very similar to Z M - 8 with the exceptions that the feldspar is white to gray and the clinopyroxene is green in colour. Ti-andradite is smaller (up to 3 mm) and has a subhedral crystal habit. IC-98-ZM-18 is a fine to medium grained syenite. K-feldspar is white to gray in colour and displays trachytic texture. Grains range in size from 2 to 4 mm. M a f i c phases range in size from 1 to 5 mm, and include garnet and biotite. IC-98-ZM-27 is a medium to coarse grained syenite containing approximately 3 % mafic phases comprised o f 0.5 mm garnet and biotite grains. K-feldspar can be up to 1 cm in length and is pink to gray in colour. IC-98-ZM-43 is a gray, coarse grained syenite comprised o f up to 2 cm-sized aggregates o f vishnevite-cancrinite, K-feldspar, and titanian andradite. IC-98-ZM-39 is a titanian andradite cumulus sample. The garnet are dense,  15 small (up to 2 m m in diameter), and euhedral. Some feldspar crystals are visible in the cumulate. The rock itself can be divided into two sections: the " B " part which is the garnet cumulus phase and the " A " phase which is a coarse grained syenite with feldspar laths up to 1 cm in length. The contact between the two phases ("C") is sharp and distinct with minimal overlap between the two. Z95-6 is a sample o f a magmatic dyke with up to 1.2 cm, euhedral titanian andradite which occur with interstitial K-feldspar and light green clinopyroxene. Z95-7 is a coarse grained sample o f a magmatic dyke comprised o f K-feldspar, light green clinopyroxene, and black, subhedral garnet up to 1 cm in size. 69ri B66-5 is a porphyritic volcanic sample from the Crowsnest volcanics, /  Alberta. This sample has been included in this study so that comparisons can be made between titanian andradite from a volcanic suite and those from primary magmatic samples. Black, euhedral titanian andradite crystals are up to 4 mm in size. Z95-1 is a white to green coloured Glacier skarn which contains calc-silicate, clinopyroxene, and anhedral titanian andradite. The calc-silicate is embayed by the garnet. Z95-5 is composed o f wollastonite and calc-silicate with chain garnet in the calcsilicate. Individual subhedral garnet grains are 1 mm in diameter, and can form cm long chains. 97BN1-4 and BN15-1 are from the Bartnick locality. 97BN1-4 is a coarse grained wollastonite skarn with light brown, euhedral, visibly zoned garnet. Garnet crystals can be up to 1 cm in diameter with visible zones up to 1 mm across. BN15-1 is a fine grained skarn which contains light brown, euhedral to subhedral chain gamet. Individual, euhedral garnet can be up to 2 m m in diameter. N o zoning is visible.  16  371 and 393 are from the C l i f f skarn locality. Both samples consist predominantly o f medium grained wollastonite, and K-feldspar. 371 has no visible garnet in hand sample, whereas 393 contains light brown, anhedral garnet up to 3.5 mm in diameter. 3.2.2 Petrography Digital images o f polished thin sections were obtained using a Polaroid SprintScan 35 attached to an Apple Macintosh computer. Figure 7 shows images o f the polished thin sections under plane polarised light. Table 3 summarizes the petrography described below. Pyroxenite Samples Pyroxenite samples include IC-98-ZM-42, IC-98-ZM-26. Both samples consist mainly o f aegirine-augite clinopyroxene. Clinopyroxene is pleochroic clear to green and occurs as anhedral grains in groundmass, and as subhedral to euhedral phenocrystic grains. Titanian andradite occurs as a major phase only in Z M - 2 6 , where it is phenocrystic to megacrystic, euhedral, and some grains show visible zoning. In plane polarised light (PPL) under conventional optical microscopy, garnet is pale red-brown to dark brown in colour. Titanian andradite present i n minor amounts in Z M - 4 2 is interstitial and occurs with titanite. Apatite occurs in all pyroxenite samples and is colourless in P P L , has a euhedral to subhedral crystal habit, and displays first order birefringence. Z M - 4 2 has apatite in major mineral amounts with euhedral grains included in pyroxene. Apatite occurs as fine grained groundmass i n Z M - 2 6 . Titanite occurs as an accessory mineral in all pyroxenite samples. In sample Z M - 4 2 , titanite is interstitial, whereas in Z M - 2 6 titanite is more abundant, grey to pale brown in P P L , fine  17 to medium grained, subhedral to euhedral, and displays first-order birefringence. Abundant, small opaque inclusions litter the titanite. Z M - 4 2 contains K-feldspar in accessory amounts. Groundmass in Z M - 2 6 can be subdivided into two parts based on color and mineralogy: green, apatite-phiogopite-pyroxene-rich areas; and white calcsilicate areas composed o f fine grained feldspars, epidote, and calcite surrounding large garnet crystals. Very minor opaque minerals occur in Z M - 2 6 as inclusions in garnet and pyroxene. Melasyenite Samples Melasyenite samples include IC-98-ZM-8, and 1C-98-ZM-16. Both samples consist predominantly o f K-feldspar, clinopyroxene, titanian andradite, biotite, and apatite. Euhedral titanite occurs in accessory amounts, and zircon occurs in very minor amounts included in other phases. K-feldspar displays some perthitic and vermicular texture and is partly altered to sericite and dusty opaques. Clinopyroxene is pleochroic pale green, occurs as anhedral grains, and is partly altered to dusty opaques. Titanian andradite is dark brown in colour and has an anhedral crystal habit. Biotite is orange to dark green in colour and has a subhedral crystal habit. Apatite is colourless in P P L , has a euhedral to subhedral crystal habit, and displays first order birefringence. Syenite Samples Syenite samples include IC-98-ZM-18, IC-98-ZM-27, and IC-98-ZM-43 and consist predominantly o f K-feldspar, titanian andradite, biotite, with accessory pyroxene, and apatite included in K-feldspar. K-feldspar displays perthitic texture and is partly altered to dusty opaques, sericite, and epidote. Titanian andradite is dark brown in colour with subhedral to anhedral crystal habit. Biotite is anhedral and dark green to  18 brown in colour. Centimetre-size vishnevite-cancrinite aggregates occur in Z M - 4 3 which are clear in P P L and hexagonal in shape. Anhedral muscovite occurs as a major phase in samples Z M - 2 7 and Z M - 1 8 . Accessory euhedral titanite occurs in Z M - 4 3 , whereas Z M 18 contains interstitial calcite and euhedral zircon included in K-feldspar. Z M - 1 8 has some alignment o f mafic phases and is trachytic, Z M - 4 3 also has some alignment o f mafic phases although the K-feldspar is not obviously trachytic, and Z M - 2 7 shows no alignment or fabric.  Zippa Cumulus Sample Samples 1C-98-ZM-39 A , B , and C represent three distinct parts o f a cumulus garnet phase in coarse grained syenite. Z M - 3 9 A represents the syenite above and not in contact with the cumulus garnet, Z M - 3 9 B represents the cumulate garnet, and Z M - 3 9 C is the contact o f the syenite phase with the cumulus garnet. Z M - 3 9 A consists predominantly o f K-feldspar, biotite, and titanian andradite, with accessory apatite, euhedral titanite, and secondary epidote. K-feldspar shows some zoning and is partly altered to sericite and dusty opaques. Anhedral biotite is dark green in colour and is partly altered to chlorite. Titanian andradite is brown in colour, and has a subhedral to anhedral crystal habit, except for some fine-grained crystals included in K-feldspar that are euhedral. Z M - 3 9 B consists predominantly o f titanian andradite, biotite, and K feldspar, with accessory euhedral titanite and anhedral apatite. T w o distinct generations of titanian andradite are discernable based on size: small euhedral garnet which lack inclusions, and phenocrystic, anhedral to subhedral grains which have biotite inclusions. N o zoning is visible in phenocrystic garnet whereas some zoning is visible in the smaller, euhedral grains. Biotite is dark green in P P L and exists in two generations: large  19  euhedral (minor anhedral) biotite in contact with pyroxene and K-feldspar in the groundmass, and garnet inclusions. Biotite shows minor alteration to chlorite. K feldspar is partly altered to dusty opaques. A t the contact between the two phases in Z M 39C skeletal garnet is present in the syenite phase and is replaced by biotite, epidote, and pyroxene. Zippa Magmatic Dykes Z95-6 and Z95-7 are magmatic dykes. Major minerals include K-feldspar, titanian andradite, clinopyroxene, and euhedral to elongate apatite. Interstitial calcite and anhedral mica occur in Z95-6. K-feldspar is perthitic i n texture and partly altered to dusty opaques in both samples. In addition, K-feldspar in Z95-6 is interstitial between garnet grains and displays some vermicular texture. Titanian andradite in both samples is light to dark brown in colour, euhedral to subhedral in habit, but is visibly zoned i n only Z95-6. Titanian andradite in Z95-7 may be embayed with pyroxene or K-feldspar. Clinopyroxene is clear to very pale green in colour, displays subhedral crystal habit, and is partly altered to epidote and dusty opaques. Crowsnest Volcanic Sample 69WB66-5 is a pristine volcanic sample o f the Crowsnest volcanics rich in euhedral titanian andradite. Garnet crystals are green to brown in colour, visibly zoned, and contain fine grained clinopyroxene inclusions. Major K-feldspar is euhedral, shows some zoning, and displays minor perthitic texture. Clinopyroxene is dark green i n colour and has a subhedral to anhedral crystal habit. Plagioclase is fine grained and anhedral. Accessory titanite is euhedral and commonly twinned. M i n o r opaques are also present.  20  Glacier Skarn Samples Z95-1 and Z95-5, both Glacier skarn samples, consist o f titanian andradite, K feldspar, clinopyroxene, calcite, and euhedral apatite. Anhedral wollastonite is present in both samples; in major amounts in Z95-5, and in accessory amounts in Z95-1. Titanian andradite is dark brown to red, shows no visible zoning, is littered with inclusions and, in Z95-5, deeply embayed by calc-silicate. K-feldspar and calcite occur as interstitial grains. Clinopyroxene is light green in colour, partly altered to epidote, and can occur as either large euhedral crystals typically with anhedral titanian andradite inclusions, or itself as anhedral inclusions in other phases.  Bartnick Skarn Samples Bartnick skarn samples (97-BN-1-4 and BN-15-1) predominantly consist o f titanian andradite, anhedral wollastonite, anhedral clinopyroxene, and anhedral K feldspar and calcite. Accessory apatite is present in BN-15-1 and the sample shows minor alignment o f phases. Titanian andradite is very pale brown in colour, euhedral to subhedral, and has visible zoning in 97-BN-1-4 as well as anisotropic zones in some crystals.  Cliff Skarn Samples The C l i f f skarn samples (371 and 393) consist predominantly o f anhedral wollastonite, perthitic K-feldspar, and interstitial calcite, with accessory euhedral apatite and titanite. Light brown, anhedral titanian andradite occurs in 393, but is only found as very minor interstitial grains in 371. Clinopyroxene occurs as subhedral grains intergrown with K-feldspar in sample 371.  21 3.2.3 Scanning Electron Microscopy Scanning electron microscopy was used in conjunction with the optical microscope to further identify and characterise phases (e.g. small grains and inclusions). Thin section traverses were selected by optical microscope. Thin sections were carboncoated and a strip o f copper tape attached to promote electron flow across the sample. Qualitative energy-dispersion X-ray spectrometry (EDS) was done on a PHILIPS X L 3 0 scanning electron microscope equipped with a Princeton Gamma Tech ultrathin window detector. Compositional variation in the samples was studied using digital backscattered electron (BSE) images. Titanian andradite zoning not visible with the naked eye or in thin section was observed in B S E images. Figure 8 A - D show titanian andradite grains from all sample groups which display zoning. For each sample, different grains have been identified using different letters, and different areas on the same grain also have a number. The letters correspond to all grains, not only garnet, selected by optical microscopy for S E M and E P M A study. Qualitative E D S analysis o f zoned garnet grains shows differences in T i , Fe, and A l between zones. Figure 8 A shows zoning visible using B S E imaging for Zippa magmatic samples. A l l magmatic samples display irregular zoning with the exception o f several pyroxenite grains (diagrams a,c,d) which also show some regular zoning. Figure 8B shows zoning from the Zippa cumulus and Crowsnest volcanic samples. Cumulus garnet ( Z M - 3 9 C ) displays a combination o f regular and irregular zoning in all garnet observed, with the most irregular areas concentrated in the cores. The Crowsnest volcanic garnets commonly have a well defined core which has irregular zoning, but from the core edge to the rim displays very regular zoning. Figure 8 C shows  22 zoning from both Zippa dykes samples. The magmatic dyke samples show both regular (Z95-6) and irregular (Z95-7) zoning. Z95-6 is shown in diagrams a-f whereas Z95-7 is shown in diagrams g-i. Figure 8D shows zoning from the Zippa skarn samples. Most skarn samples are very irregularly zoned, with the exception o f 97BN1-4 (diagrams g-r) which has very regular zoning coupled with irregular zones inside otherwise regular zones. Complex, non-periodic zonation patterns o f this sort can be the result o f periodic external forcings, notably changes in the composition o f the hydrothermal fluid (Jamtveit 1991). Some parts o f certain crystals in 97BN1-4 also display sector zoning (diagrams g,ij) and anisotropy. Element maps, produced using S E M / E D S and characteristic X-ray lines, have been used to show element zonation and growth zoning in garnet (Nakano et al. 1989). Element maps o f Z M 3 9 B - B , a small, euhedral, optically zoned garnet, are shown in Figure 9. Figure 9-a shows the grain as a back-scattered electron image. Most elements show little to no discernable zoning with this method, with the exception o f Figure 9-f which shows titanium zoning. Titanium is brightest (yellow) at the centre indicative o f - large amounts, and colour decreases toward the rim which indicates lesser amounts. N o other grains tested show discernable zoning with this method.  23  c o CD c c .o o •9?  TO C O O  CO  CD  3  © 0) XI -O  O  CD  g  co ro  CD  «  g» O  CD  CD  w CD C  1  i  O- Q . E E CO CO CD " CO W _>< CQ) D C C O O  co  C Q. CO r n o ' £ „\ ro .t: co  CD CO CD - </) CO o  °  CO  CJ)  CD  TO 0)  co  x:  CD  Q. CD +_  CD  3 CD F o  "co  E CO  co  E CD c oo C Q.  CNCD (D CO N CO O l O l O l tTC MiO Ol r -i T -iC N t ^,C T O iC Oi C Oi  N N N N N N N N N N c p r ^ CDOOOOOOCOOOOOCOOOOO.T.i. cnc^cncocncocna>a>cn£ £  c  CD X  o i _ >. 0-  CD E CO  co  « in  ro CD  CD  co  N  . a s s  8«  ro<D CO  ±: '= >_-™ 2 'W CDro E  0)  CD  Q- CL|  ro ro  in w c o  >.| c o  N N  T - CO CO  CO  r- 0) CO CO  c:  CO o  o  in CD (I)  E  LO  —  ro  o  If)  to If) CD CD  CD  • • CO  c CD  °roro  ro E  CO  L  c CD >,  CO  c ro ro T3 »r CO CD 01 o > > c o •D CO tO  CD >  CD CD  CD  £  c^ c  < CD O  ro c ro ro E ro,  E CD CO  co  I'  co CO  CD  CD  -  <  co o  1_  So2  o ' o t; >> -O C (J >t- £ CO O > i= C  co  c  •= £ °-  c  ai CD roro ^ 5 S  k.  CD  co F--"=•W -C E  = ro jo  -32 ro c .o  <D  « z CD  CD  CD C  >% Q  c to  I  ro o ro  O  o  u  s o  >  CD CO  O  o 'c -£ ro  DO CD O  O  24  metres Cliff Deposit  •  Pyroxenite Andradite bearing wollastonite  n  Grossular garnet bearing wollastonite +/- andradite  L J Pyroxene bearing wollastonite •  +/- andradite/grossular garnet Coarse wollastonite (contains <5% andradite, no pyroxenite or grossular garnet) Calc-silicate Diorite  V  Figure 3. Cliff sample location map (after Westphal et al. 1999).  A Glacier Deposit Pyroxenite f  Biotite-rich (>90%) pyroxenite  §§§  Andradite bearing wollastonite  j  Grossular garnet bearing wollastonite +/- andradite  j  Pyroxene bearing wollastonite +/- andradite/grossular garnet  7A Coarse wollastonite  (contains < 5 % andradite and no pyroxenite or grossular garnet)  fo ° o] Calc-silicate III  Marble Diorite  0  20 metres  Figure 4. Glacier sample location map (after Westphal et al. 1999).  : mm  (j  if 4  27  +  4  + 4  4  4  4  4  4  4  4 +4  4  4  41 t  I  4+ 4 + 4  n  M  4+ 4  j  4  4  4  4  4  4  4  4  4  jfTf t• , -W+ ^ 4. 4 * - 4 , *4 , 4, *.rtiTlTtTTTT?' + ^ 4 + + + + +• i m 4 4 + 4 4 +7 | i l l 4  4  4  4  4 >  J  ,+  4  +  +  / l » I  i  4  4 j^r*-^.  + 4 4 4 4 +C1 4 4 4 4 4 4 4 4 4 4 4 4 4 + 4/1 • 4 4 4 4 4 4 4 4 4 4 + 4 + + 4 /| i | 4 + 4 4 4 + 4 4 4 4 4 4 4 4 | i | 1 4  M  j j j  4  4  4 4 A+ 4 4 4 4 X  . | •T i ~ r » - J ^ i 4 4 X^+TT. I I I I .1 . I i | 1 I • | • I. | i | '. | 1 | i | 1 | 1 | 1 | 1 | . ] | | • I . I 1 I •I i I 1  1  1 1 Ifrtl  .nil 1  I I I I I I I I I • I 1:3J  I  Bartnik Deposit Pv^ Syenite Stuhini Group Metasediments • Metavolcanics Contour interval = 100m  § i! Assemblage l i s Limestone Shale Wollastonite skarn t  < i n e  0  100 metres  Figure 5. Bartnik sample location map (after Westphal et al. 1999).  28  co CN  O  d  *? * N C  1-  CM OS  CO CJ)  T j - CO •»- m CD CN O •<-  r -  CD CN CO O CO GO CN O  CD  CD  CO  10 0 ) 0 m in o oo d T O) o  o  CD O O iri C N i d  O d  O d  to o  o o  o o o  o o o  CO CO CO CN O  © m oo O  O  o o  o  o  •<- oo  T-  o o  o o O o  i - CN  o  CJ) o o co  o  o  o  ^"  N  iri i -  o  o  o  o  CO  CO  ^ -  CO  d  1  o ^~  csi  CM * - •<- CD CN CN rCM  o in d d TCM  o  CM  T-  o o o o o m T co co co r*~ a> ai C N -I-  T - r-T-  CO  CD  t -  o  CN CN  co  co  CD 00  CD CO  T-  T-  CN c  co  CO  O) T-  CM  d  ai  CO  o  d  oo  CM  ^  CM T-  <» CM CO O CO CO  O o  T  d  o  T-  CO  o  if) O) O) 1^O T- Ti-  iri  CM  «  m co  r-~  o  O  CO  o o o o q o oo co in T -  o o  CN CD  O  r~o m od d d cn  i eo  o o o o o o iri d d t-  o q d  o o  o o  o o  oo o o •>o" d  o  o o  o m  o  o o d  o o  f  d  d  o o  o o  o o  d  O  S  d  m  o  CM CN  o o o o o o — cn r~~ O M O O U h~ d co d C N V  r l f l l O T~  d  o  o  o  X— CM  o  O CO CO CM CN CD I " i r i CM CO CO CJ) CN CJ)  V  N  C CO  CO O)  £y CO  t -  10(0 o 0 0 o O i - 00 o d d d d iri  o  CO  as N  t  rT-  d  i  - o o r -- O CO T O 05  •<-  CD  o  CD  N V 00  T-  o o  o  C CO  £y  o  N -  ..ffi <o  o o d  o •* ( 0 ( D N o i - f~ ir> T d  co  co m  o  ai  oo d co  d  CJ) OD O  o o  O  o o  O  o o  o o  o m  o  o  00  CN CF) V •* CO  T-  d  d  o  o  O O CN O O O O o o o o •>»• d o cd d T— i•>*• CN o v  CM  w N  |S  m o  CO  £  o  0 O •» CN 01 d C M V CO CN CO CO CO  co CN iri  c  co ffl  CD CN  T-  r N  o o o o cn o  o o r>-cd  O o CN in  O O O O o o o q oo d o d h-  00  CO If) o O CD O CD O CO  o o  o o o o o o iri d d  in o  o 00  to co <p o>  i-  o o o  •<- o co oo  •<r o o o o o o o iri oo r-O) CO o CM m T-  00 CN  O  O CM m CM o  T T  o o o  o  CM  o o o o o o m C M o cp CO i - CM iri iri in  CD CN 1 - 1 -  CO  CO  CD  q>.0  CO  £^ I*  IO CD  1^  o  co o CM S  CO CO CO CO  CN  cn d  co ro o -<3-' d  CM  CO CM IO CN CN T d ^ CM T- CN  •«- CM  5  o  O)  CD  CM  O O  O  t  CO CO  o O CM  o q 6  o  o  o  o  o  CO  CD  CO  O)  iri ai cd T-  CN  o  o  CM  t -  o  o  CD  I--' co  o  o  o  o  o  o  ( 0 N ffl IO T- CM N id  s iri cd T-  cd co  T-  00  lit  la  Uj S; v.  ra o  § or  E  Q n  i-  <  u-  rT  o 9 9»° &  ii  C  O) CO  5 5 0  °  z  CL  o o 3 =o o  ^  ai-_imOc/)Nu.Oc/)  | 6  5 S  X! i_ c _ Z Q . K ( 0 r - 3 XI J3  CM  I  29  o o oo t-~  CD CO LO  CD CO CO 00  T -  00 l-~ o  <o  LO  o  O  o  od 1-^  T—  o 00  CO  o  od  CO LO CN  in  00 CO  f1 -  o  CO CO CN  00  o  o  o  d CN  oo O CN  o  d  1— CN 00 00  o r--  oo i n •<*• CO CO CN  o m  o  ^  co O CD  CO  CO CN  iri  CO  o  d  0)  81 £ :|  E a> to  tu O £ io - co c £  c 3 o  h-  CJ)  m CO •<t co  iri  CO  o o  CO  CO 00 CO LO  o  to •<r  1^  CO  o o  LO CO h ~ N CO CO  o  T—  00  O  O  CN  CO CO  d  T—  1—  O  o r>- o o o o CN  O  d  iri  CN CO CO T -  CO CN LO CN T " O  TJ  m *—  ro  -—  o  CO  sz. in  o in o  'c o E E  a>  E  m  tD «-  CO CO  co  a. <o X CO <u in (D v_ CO  c in  •<*  o  od  0)  d  o o  CN CN LO  LO 00 00 O r - CO  O r-  d co  c  CD  E  cn  oj  x>  »  CN  || c  •<*  o o  o  CD  o o  CO CN CN CO CO CO CO CN od  CO CO CO CN CN CO  o d  LO CN T — CO LO  O  CO  0)  o o  CN CN  o o  CO  io  O  CO  s CO* o o o  lO CO t -  od co  T—  r-  CD  d  £  LO co in CN  CO  T—  LO  T—  h~  O  T - CD OO CO  ai  T ~  o  d  O  <D  LO  d  oi  CD O  ° c  CN CO CN  CO  •<r iri T—  o  O  o o o  CO  iri 00  O  CD CN  CN CN  CN  T—  t—  b o CO CN od  TJ CD C "••-»  c o O  xi 03  —  42 c  * « * • CO CN CO CO  o> 1Q*z 6  CO X—  CN  o  h - CN CN LO  5  d  o  in T oo oo  1  ro  e  J.5  \  >- N  -9 X XI c n CD o cn — 2: O < < _ l Z < Z D  O  —>  5  h-  CD  x:  2  CD  Q.  C  ro  o  cn CD <D ro  «l cn rra o „ CO  TJ  CD u CO <D 3 x: 0)  co  S c <S o ™ <n <D  < o !2 3 o ® c  E ° S CD 0) CD xj «>> •c c £ *• > .2 cn £ "S 5_ 3 o _j Q 0- C C <O c  11 I «  E 4) CD  CD  -  If 2  c  CD  C O  ~  c:  Z  It-  cz  's  o  CD" 3 CD II CZ —o <D  u  3 TI C  .t; E  <g  O  o  cn 3 CD C  —  x! fl) "C -S  o ro  sI II  <s  o  <I  CO  £  ID  A  I  TJ in cn  £  c  to  to  tN O  x  < CO • - ro CD JZ £ CD"  LO  ? l o °  TJ CD  JL O  c 'E i_  aj  CD TJ  a; CD"  el  cn In to jo -5 i cn > i cn ll  o  TJ  CN I  : CD  O  Is- . 5  i -  x  t  Q. x* x i .cn  s °  ID  * 18  O  TJ  ro c o. ,_ CD <u ~  IS  <D  s  o m co o — o "5 «  o'O o  C  o  ro  c m c CO 3 — c 'E E 3 2 E TJ  3  CO CD O CO  o  <" >,T J  S;  c  XI  xi  xi  TJ  = CD CD  cn to  1  o  y p  cn xi co  o o  E .co ro x:  ^. c in co T— cn £ in £ CO -  CD CJ cn CD  c 3 . CO 5= in 3 >> a> crCO  10 •=  o" LO CO  Q  is  c tu co  CO Q.  o o  0)  o o m" 3  •7 5  e-  LU  in  o <n CD CD  o o  3  XJ CD C  £ S CO fa  in .co  d  o  CO  o_ m — 3 co in /-)  S  o o o o d CO  o  %  CD  T—  CO  > ro  I  e  "5f  o  c o  CD J£ E O CU CO E °x> ID 3 CD  £O) £ * # — o " B rV  CO TJ nj  c  TJ  2i m  I £  CO  : <D CD  i cn "2 P. -a <?.2  p ®b ®E8  in  i  <o = — CD > to c f •S .2 + § • « CZ -G O o ro o CD _ CN 0 3  « «  II in  CO  to  CL  ro  II Q. < CD" II 3  : z o:  30  1  1  i  1  1  i  1  1  1  • •  +  a)  zm42 zm26 zm8 zml6 zml8 zm27 zm43 zm39a zm39b  A  A  *  -I A  T * * * ^ •  x  •  T  -  Alkaline •  A  • 1 N  35  40  45  50  SubAlkaline l  55  i  60  SICh  i  65  i  70  i  75  i  80 85  (wt.%)  FeO*  N c h O + K2O  MgO  Figure 6. Whole rock geochemical data plotted for magmatic samples. a) Total Alkali-Silica diagram (Irvine and Baragar 1971). Solid line represents division between alkaline and subalkaline compositions. b) A F M diagram (Irvine and Baragar 1971). Diagram shows FeO enrichment.  31  C)  Nb  Zr  Figure 6 continued, c) Concentrations of trace elements (ppm) plotted against Si0 wt.%. 2  IC98ZM54  IC98ZM42  IC98ZM26  IC98ZM8  Figure 7.Scanned digital images of polished thin sections in PPL for all samples from this study.  IC98ZM16  IC98ZM18  IC98ZM27  IC98ZM43  Figure 7. Continued.  Figure 7. Continued.  Z95-7  69WB66-5  Z95-1  Z95-5  Figure 7. Continued.  36  Figure 7. Continued.  37  E  y>  2  •c o o CL 3 O  o  c o  S co CD  CD T3  "CD  c o  c o  _Q  XI  CO TU  co T3  in  if)  — I  I  o X) l _  CD  CD E CD  o  CD  1  CD O  CO OJ 10 CD  •g  a. CD  8-' to  + (D  CD  ~ C to j-  CO <D  8  —  —  CD CD C C !s CO Cis D CD CD  — CL O O) o n a.  to <n 3 CD  CD .ti C CT ~ ro co _ C O " Q . CL ro j2 O < O i— N  CO  CD  *>  c o  2"  i— N  O  to  Q)  ffi  -E CD  O — O CD CL ,h= O < N O  CD  CD CD  ~  CT  ro ro  <  O  <  £  c  S  ro 5 a. Q-  CL .t;  r-  CD  to CD to CD  O C CD O  i  a . CL O <  x  E  §  a . ro -is O O P  CO  Q. X  to  CL  CD CD -S C ~ ~ CD J= • « CD  .Q  OB,  QJ  ^ o o m <  CD CD C  8" 2. ro o  *  O  CD DD  — i CD CD y: CD co o  O  ga CO S-|fe.2 gj 8ro 8  £  2 ^ 2  » >  cCD k_  x:  CL  u. O  CL  c  CD  CD CD  1  L_  to ro o o  CO co  o o  CD  "O CD  CD CD  T3 CD C CD  i_ CD CD C  CD  -a E C  CN  T  CL N  E  00  CD cn CO O  CO  N 00  a> O  N  00 CD  o  CO  oo  N oo CD O  N OO CD O  o  c 'ro s  c  CO  *  E to" o CO o "O CD  E o to„ o  tra hyti  "O CD  O  CD  to o "D CD C CD a> CD 10 v_ CD O O  CD co  cn CD  to  CD O O  CM N oo cr> O  ed  x ro  >  ro E  c  'S o  rse nment  ~ CD <D .t:  CD CD  o o ro  CO  T  N oo CD  o  38  c  1 2 n sz co  CO  ro  o E o c c c  5  c o  c o  CD CO  2 2  CD  XI x>  x:  CO CD — TJ  "9 CO  TJ  ©  CD TJ  2 In  © ©  sz  co  CO  CO  TJ  o in  o m  o o. 3  CD  CD  5 el  a> c  CD  In  X  m2 9-o_^ <  •i  c o to  CO  CD H>  © §  c ca g  '-•3 CD  CT  r- UJ  CO  t- o  <  ro ©  CD  I  CO  CD  ID u  Q..Q  O O ca  i_  £  •55 co c Q- co 8-o io cl| o yO *to h.t: O C  CD  CD a x '75W D . CO  CD .t:  to .2 o. iri CO O  * oo  TJ  CD C  to  g '> x> o "5 c t_ o co Q.  TJ ©  2  -5  TJ  CO  CD  CD C  CO  C  co jfci ° CD  CO  8 © a E  CD  c CO CD  52 CO  8  TJ  CD C  Q. X CO C  c  re I  ~  CO  CL  10  i - CD, CD QJ  * o o  *  < r  TJ  S CO  ©  CO l_ C0  8  _  i_ CD  O 0. O O  *  CD __  © C  ,<S o. to  © C  C  CO  o o  TJ  ©  '5 CD to u. CD  co  '2  CO  E D TJ  © E  CO  x:  o CL  © to t_ CD  8  C  c o  O  CO  J) CO  CO  O  00  00  N  N  CD  N  00 CO  o  CO CD  o  CD  CO CD  o  LO  i  CD CO  <9 m CD  N  CO  LO CD  N  39  2  xi JD  e x: X » cn  x: co  o  o  CN  CN  JD  (0 CD Q.  OJ  1o  CD CD x a 11  a 5  i  S .g E CD  cn  TJ CD  ra  _ a>  ° ° 8  § O  CD _ CL is: O ) O  CO (0 CO O  CO O CD CD "  CL  E n c "O  E o to  o  CD RA  5  c c  CD CD S=  CD  C  Q) c  S .£ s a •e i < H O  -4-4  2  CO CD 3  co 55  £ o-  o JD CD CD "55 0 C L CO, JD CD  x: cn  If) o  CD  C  — i  X)  CD >  to "  o  CO CL  CO QL  z m cn  CO  CL co O O  Z  c o  CD to CD X -£> ™m Q. CO > O C L -= C O CD 2  £  CO  CO  0 5  o  &.§ 8  1 1 to E  to  z CQ  CO «  E  t ?  o  CO  a  CO  CD T3 T3 E E § CO CD CD CO E C D c T3 CD O O CD 'JSc to C vD) C O C O v. (0 — To Q.| CD fe CD CJto  9- 0  CL <  CD  oo C  CD CO  CO  -¥ i JS E  A  CO  JJ to ca 5 o E x> "o  CD CD f-  •S2fe) 8.-E »  CO O) CO  CO  40  CD  1  c  CD CD  '35 •  g  o  ••g  $,£  8  I I c  £ o XI  c TJ c o £  c  CO  N CO JD o XI o 10 (0 0) > CO CD CD  x: 03 i. 3  8  t  Sj cT  C D  S CD  CO  CD  C <D  I  CM 2  CO CO  o  o £  S  x: .ts  3  .ts c: o 2  II  9 fx  CD J=;  •1  o. c x: CD  CD SZ X) 3 to 1 CD L_  XJ 3 to i CD  CD  co-§ •5 CD  TJ  2 ="  x:  <2  5 CD  3  3 CD  CO  5 N  co  CO  o  CO CO  o  N  CO CO  O  13 x: xi  co i_  TJ  3 to  x:  ro E  CD  x: c ro  CO  2  CD  x: c ro  TJ CD  x: c  CD  ro  TJ  TJ  x:  TJ  i  CD = CO CD  CD l_  TJ CD  3  If  CT  CO  CO TJ CD C L  42  v_  = 6  x: .2 ~ x: S 9c o 5 E S  CO CO CD  CO  CD  1N  (0 CO  CD  3 co CO  2  $  c CD — i_ <D 3  3  I I c>- ro  11  CO  v. CD -Q S  *  sz o  o w e  o  S  ?-&  CO  CD  CD CD  CD 3  X) 3 CO  CO  CO  5  CO  N  N  N  O  O  o  T— I  S  CO CO  T CO CO  41  o  n  z fr  8 %  CO CO ~ CD t CO CO  03 cp TJ CD CO _  c  E &  S) g  8-2  o Q.52 -c co 5  2 ©  x: c CO Tj" ©  c  CO L.  CO  © c 3  C  c o O CO  ©  X> CO  I-  Q.  CO  CO  Q  c CO '2 c CQ c © o  ac o •-  1  "§  |  il  C CO  ~ c © o  E « © "7^ CO  o  2 co c c  CO  + to  i^Tosn x:  > c" 2 ©  1  3 .ti  N  i§>8© © £  u £  To ~  ^ -2 -b- © >.  ^.2 © X) 2 x: c  TJ ©  CO  TJ © XZ  X) 3 CO  CO  .£ To  ©  2 x5  c ©  CO  CO ©  £5. C cO o _  22  TJ  © © S>x: ©  x: x: E* 3  ©  CQ CO CO  N  o  C CO CO ^  TJ  ©  x:  © ©  *  s.2 c o1 CO..-  ~c »© *  © - 8"  co c CO o o to x:  c  o" © O _t c > i o ro> a-5cx  JC  s-g  TJ ©  I  © > © ..  CO ©.ti © CO © C ° © o CO OA 1 CO E •E 2 g >2  1o~  TJ C (0  E o g 3)5 To © k" CO ~ c © © E JCa c o CO  TJ ©  o N  CO  Q.  o to  x:  2  2  TJ ©  TJ ©  x: c  x: ©  CO  To °iz © "8 X) 3 js © E  2  TJ ©  x: c CO 1  2  TJ ©  x:  X) 3 CO  O  CD CO  81  S CD  N  S if) CD  N  CO  iCO  8£  CO X ^  £ >» t: ro  X) CO I  3 o  CL c E o X  •e  ©  CO i•- c JS 2  o  co •'  2  If "i  If) CD  N  42  CD  |  '55 o 8 >> >.  CO  x> E ID >. CL CD CD TJ  00 c o '55 c c5 CD  1! c  «  8  s  x:  O) CD c «= c  I TJ  8  W CO  CO TJ  ro  c ca 5  2 x:  © g-i  x: c  2  TJ  X?  c  CO  X)  LO LO  LO  cn N  c CD  1 |  TJ CD  CO  z CO  tszc ro  (0  ro  12  co CO CO  Figure 8A. BSE images of grains showing zoning in Zippa magmatic samples. a-e) pyroxenite; f-g) melasyenite; h-o) syenite.  44  Figure 8A. Continued.  Figure 8A. Continued.  46  Figure 8B. BSE images of grains showing zoning in Zippa cumulus and Crowsnest volcanic samples, a-k) cumulus; l-r) Crowsnest.  47  IC-98-ZM-39C-I  IC-98-ZM-39C-W  Figure 8B. Continued.  48  69WB66-F  Figure 8B. Continued  69WB66-G  Z95-6-M  Z95-6-Mcloseup  w  Z95-6-P1 Figure 8C. BSE images of grains showing zoning in Zippa dyke samples.  Figure 8C. Continued.  Figure 8D. BSE images of grains showing zoning in Zippa skarn samples. a-f) Glacier; g-r) Bartnick; s-u) Cliff.  52  Figure 8D. Continued.  53  Figure 8D. Continued.  Figure 8D. Continued.  55  Figure 9. S E M elemental X-ray maps for IC98ZM39B grain b. a) B S E image; b) Si; c) C a ; d) Fe; e) Al; f) Ti.  56 Chemical Characteristics and Representation of Titanian Andradite  4.1 Chemical Mineralogy 4.1.1 Electron Microprobe Analyses Compositions o f garnet were determined using a Cameca S X 50 electron microprobe. Operating conditions were 15 k V and 20 n A beam current. Estimates o f analytical uncertainties are described in Appendix A . Compositions were measured for all textural varieties o f garnet, including phenocrysts, smaller euhedral garnet, and interstitial grains. Garnet samples were also chosen for analysis on the basis o f zoning that was visible under optical microscopy or during B S E imaging. Straight-line traverses were performed on individual grains and, where possible, multiple grains were analysed within a single sample. Traverses from core to rim were done where possible, however slight deviations from a straight-line were sometimes necessary to avoid inclusions or flaws in the grain. Garnet compositions were converted to structural formulae and Thompson components using Matlab script ti and.m (Appendix B). The program assumes eight cations and twelve oxygens and on that basis calculates iron valences. Cation distributions over the X, Y, and Z sites are forced, within the program, to totals o f approximately 3, 2, and 3, respectively. 4.1.2 Zoning Observed in Microprobe Analyses Representative median microprobe analyses with their corresponding structural formulae and Thompson components are given in Table 4. For full microprobe analyses, structural formulae and Thompson components, inclusions, and bad or unknown analyses  57 (in E x c e l 4.0), see cd-rorri in Appendix C . The process o f eliminating inclusions, bad analyses, and unknowns from the microprobe data is described below. Inclusion analyses that were intentionally spotted and probed were removed from the general data into a separate table labeled "Inclusions". The elimination process for "bad" data included analyses that: a) were taken too near the edge o f a grain or pit or crack, b) have an abnormal ion distribution (ie. not the 3-2-3 site distribution), c) have either very low ion totals or ion totals greater than 8, d) represent other mineral compositions that were not spotted, or e) have oxide totals outside o f the accepted 101 -99 ± 3 wt.% range. The low value of 96, wt.% was established as a cut-off to avoid eliminating any analyses which contain large amounts o f water, though this total for a non-hydroxyl-rich garnet analysis would normally be considered to be too low. Figure 8 A - D contains back-scattered electron images o f zones that were analysed. Zoning is described in terms o f grain by grain patterns within a sample, patterns common to different rock types, patterns in calculated structural formulae, and patterns common to different environments o f formation. Oxides which are excluded from this discussion show no significant zoning. "Major" oxides refer to those oxides present in the samples in the greatest abundance, including S i 0 , T i 0 , A 1 0 , F e 0 , and CaO. Compositional 2  2  2  3  2  3  profiles for selected grains (Figures 10-24) show zoning in most o f the major oxides. The lines along which compositional profiles were constructed for selected grains are shown in Figure 25. Compositional profiles for Z M 3 9 C (Figures 10-16) are described below. In Z M 3 9 B , grains a and h (Figures 10 and 11), high S i 0 content correlates with high F e 0 , 2  2  3  58 and low T i 0 and A 1 0 . In grains i , n, w, and x (Figures 12, 13, 15, and 16), T i 0 shows 2  2  3  2  reverse zoning (decreases from core to rim), and high T i 0 content correlates with low 2  M n O content. In grain n, A 1 0 2  3  and F e 0 are negatively correlated, and in grain x, low 2  3  T i 0 content correlates with low F e 0 , and high A 1 0 2  2  3  2  and S i 0 contents. Figure 15  3  2  shows a low S i O peak which corresponds to a low total oxides peak which may indicate z  the presence of O H . In grain o (Figure 14), high S i 0 correlates with high A 1 0 and 2  2  3  C a O , and low T i 0 and F e 0 . 2  2  3  Compositional profiles for 69WB66-5 (Figures 17-20) show that T i 0  2  has  pronounced oscillatory zoning in all but grain e (Figure 18), in which it shows normal zoning (increases from core to rim). In grains e and p (Figures 18 and 20), high T i 0  2  content correlates with low F e 0 , content, whereas in grain b (Figure 17) high A 1 0  3  2  2  content correlates with low C a O content. Compositional profiles for Z95-5 (Figures 21 and 22) show that both T i 0 and 2  S i 0 show oscillatory zoning and that these pattens are opposite. In Z95-5, grain f 2  (Figure 21), high S i 0 content correlates with high A 1 0 and C a O and low T i 0 content. 2  2  3  2  Compositional profiles for 97BN1-4 (Figures 23 and 24) show that high T i 0 content correlates with high F e 0 2  3  2  and low A 1 0 and C a O contents. In grain b (Figure 2  3  23) high S i 0 content correlates with high A 1 0 and C a O contents; and low S i 0 content 2  2  3  2  corresponds to a low total oxides content which may indicate the presence of O H .  Zoning patterns within rock types Table 5 summarizes the zoning patterns for each sample described below. Chemical zoning in titanian andradite from pyroxenite typically shows regular zoning in which darker zones have more A 1 0 2  3  (and T i 0 in one grain) and less F e 0 . 2  2  3  59 Decreases o f wt.% in T i 0 are accompanied by proportional increases in F e 0 for most 2  2  3  grains, and F e 0 shows the opposite overall zoning pattern to the other major oxides. 2  3  Most subhedral melasyenite titanian andradite grains show no zoning in B S E however, those few that do show irregular zoning contain darker areas which have more T i 0 and less F e 0 wt.% than do light areas. Changes in T i O wt.% may be 2  2  3  z  accompanied by inversely proportional or proportional changes in F e 0 , and slight 2  increases in A 1 0 2  3  3  for interstitial grains, but slight decreases for subhedral grains. F e 0 , 2  3  whether reverse or normal, shows the opposite overall zoning pattern to the other major oxides. T i 0 and A 1 0 may show opposite oscillatory zoning patterns, A 1 0 and C a O 2  2  3  2  3  may show opposite oscillatory zoning patterns, or A 1 0 and F e 0 show opposite 2  3  2  3  oscillatory zoning patterns. Zoning patterns observed in subhedral, anhedral, and interstitial titanian andradite syenite grains include irregular zoning near rims. Dark zones show larger values o f T i 0 and A 1 0 2  3  2  and lesser values o f F e 0 than do lighter zones. T i 0 and F e ^ show the 2  3  2  same oscillatory zoning pattern, but opposite to that of A 1 0 . C a O shows the same 2  zoning pattern as A 1 0 2  3  3  in some grains. Overall from core to rim, A 1 0 2  3  and C a O  commonly have reverse zoning, whereas T i 0 and F e 0 show normal zoning, though 2  2  3  T i 0 may instead show the opposite overall zoning pattern from core to rim as compared 2  to other major oxides. In most grains, A 1 0 2  3  changes by approximately half that o f the  T i 0 change from point to point, however in some grains A I 0 2  2  3  and F e 0 change with 2  3  nearly the same rate from point to point. Zoning patterns observed in subhedral to anhedral, and euhedral titanian andradite cumulus grains include irregular zoning near the core and regular zoning near  60 the rim in B S E . Darker areas have larger amounts o f T i 0 and A 1 0 2  2  3  and lesser amounts  o f F e 0 than do lighter zones. A 1 0 and C a O show the same oscillatory zoning pattern, 2  3  2  3  opposite to that shared by T i 0 and F e 0 . Compositional profiles show that S i 0 2  2  3  the same oscillatory zoning pattern as A 1 0 2  A1 0 2  3  3  2  shares  and CaO. From core to rim, either F e 0 or 2  3  shows the opposite zoning pattern as compared to the other major oxides. T i 0  2  and F e 0 change in wt.% by approximately the same amount from point to point for two 2  3  grains, and A 1 0 changes by some fraction o f the other two. 2  3  The majority of magmatic dyke titanian andradite grains show no visible zoning in B S E . Zoning patterns observed in subhedral to anhedral, or parts o f larger, magmatic dyke grains include irregular zoning near the rim in which darker zones have more A 1 0 , 2  3  but may have either more or less T i 0 , and more or less F e 0 than lighter zones. 2  2  3  Regular zoning is found in half the grains and has darker zones which show more A l 0 , 2  3  less F e 0 , and either more or less T i 0 than light zones. For most grains, T i 0 and 2  3  2  2  F e 0 show the same oscillatory zoning pattern, as do A J 0  3  and C a O but opposite to the  pattern o f T i 0 and F e 0 . However, in several grains A 1 0  3  and F e 0 show the same  2  3  2  2  2  3  2  2  3  oscillatory zoning pattern, which may be the opposite pattern o f either T i 0 or CaO. 2  F e 0 shows reverse overall zoning. Several grains show that A 1 0 2  3  2  3  and F e 0 change 2  3  by approximately the same wt.% from point to point, and T i 0 changes by either the 2  same or half that amount. Zoning patterns observed in subhedral to euhedral volcanic titanian andradite grains include regular zoning in which darker areas show more T i 0 and A 1 0 2  2  3  and less  F e 0 than lighter areas, and irregular cores which are visible optically. In one grain, 2  3  T i 0 and C a O have the same oscillatory zoning pattern, and in compositional profiles 2  61 T i 0 and F e 0 have opposite oscillatory zoning peaks. From core to rim T i 0 shows 2  2  3  2  reverse zoning whereas F e 0 shows normal zoning. The change in wt.% from point to 2  3  point for T i 0 is twice the wt.% change for F e 0 . 2  2  3  Anhedral titanian andradite grains from the Glacier locality contain abundant inclusions and commonly show no zoning in B S E , however those that do show irregular zoning in B S E have darker zones which show more A 1 0 2  3  and less F e 0 and T i 0 2  3  lighter zones. T i 0 and F e 0 share the same oscillatory zoning pattern, as do A 1 0 2  2  3  2  C a O , opposite to that o f T i 0 and F e 0 . 2  2  than  2  and  3  Overall from core to rim, whether normal or  3  reverse, T i 0 and F e 0 have the same zoning pattern, and A 1 0 and C a O share the same 2  2  3  2  pattern, opposite to T i 0 and F e 0 . A l 0 2  2  3  2  3  3  and C a O change in wt.% by approximately  the same amount from point to point, whereas in most grains the change in wt.% for F e 0 and T i 0 from point to point is either half or approximately the same. 2  3  2  The two Bartnick samples have entirely different titanian andradite grain and zoning types. 97BN1-4 has coarse grained, euhedral grains which contain abundant inclusions, and have fine regular zoning with darker zones which show more A 1 0 and 2  3  less T i 0 and F e 0 than lighter zones. T i 0 and F e 0 have the same overall zoning 2  2  3  2  2  3  pattern, and A 1 0 and C a O share the same pattern, opposite to T i 0 and F e 0 . A 1 0 2  3  2  2  3  2  3  and F e 0 consistently change in wt.% by approximately the same amount from point to 2  3  point. BN15-1 has sub-anhedral grains which contain abundant inclusions, and have irregular zoning in which darker zones for anhedral grains show more A 1 0 2  and T i 0  3  and less F e 0 than lighter zones, whereas subhedral grains show more A 1 0 and less 2  3  2  3  T i 0 and F e 0 . 2  2  3  Zoning patterns observed i n anhedral titanian andradite C l i f f grains show  2  irregular zoning in which darker zones show more A 1 0 2  and less T i 0 and F e 0 than  3  2  2  3  lighter zones. Commonly, A 1 0 and C a O share the same pattern, opposite to T i 0 and 2  3  2  F e 0 . Interstitial grains show irregular zoning in which darker zones have more A 1 0 2  3  2  3  and T i 0 and less F e 0 , and from core to rim T i 0 and A 1 0 have the same overall 2  2  3  2  2  3  reverse zoning pattern whereas F e 0 and C a O share the same pattern, opposite to T i 0 2  3  2  and A 1 0 . 2  3  Zoning patterns within different environments Across the sample suite, no systematic trends in zoning occur with respect to brightness in back-scattered electron images, however within different environments o f formation trends do occur as described below. Partly resorbed, euhedral volcanic grains are considered the most primary o f the grains in this study. Cumulus grains, though not resorbed, share many similarities with the volcanic grains (from Crowsnest) including euhedral to subhedral shape; irregular zoning near the core and regular zoning near the rim in which darker zones show more T i 0 and A 1 0 2  2  3  and less F e 0 than lighter areas; A 1 0 and C a O have the same 2  3  2  3  oscillatory zoning pattern, opposite to that shared by T i 0 and F e 0 ; and the change in 2  2  3  wt.% from point to point for T i 0 is either the same or twice the wt.% change for F e 0 . 2  2  Syenite grains also possess these early-formed characteristics, however A 1 0 2  3  3  changes  by approximately half that o f the T i 0 change from point to point, whereas in some 2  grains A 1 0 and F e 0 change wt.% with nearly the same rate from point to point. 2  3  2  3  Melasyenite grains generally show no zoning, or show irregular zoning, in which darker zones have more T i 0 and less F e 0 wt.%. T i 0 and A 1 0 2  2  3  2  2  3  show opposite oscillatory  zoning patterns, A 1 0 and C a O show opposite oscillatory zoning patterns, or A 1 0 and 2  3  2  3  63 F e 0 show opposite oscillatory zoning patterns. Changes in T i 0 wt.% may be 2  3  2  accompanied by proportional increasing or decreasing changes in F e 0 , and slight 2  3  increases in A 1 0 . Pyroxenite grains show only regular zoning in which darker zones 2  3  have more A 1 0 and less F e 0 , but show no change in T i 0 . Decreases o f wt.% in T i 0 2  3  2  3  2  are accompanied by proportional increases in F e 0 . Magmatic dykes show either 2  3  regular or irregular zoning in which darker zones have more A 1 0 , but may have either 2  3  more or less T i 0 , and more or less F e 0 than lighter zones. T i 0 and F e 0 may show 2  2  3  2  2  3  the same oscillatory zoning pattern however, A I 0 and F e 0 may also share the same 2  3  2  3  zoning pattern, as may T i 0 and CaO. In summary, as crystallisation progresses, the 2  grains begin to lose more and more, or to retain only certain parts, o f the early zoning characteristics, however, all magmatic grains do indicate substitution and possibly replacement between T i 0 , F e 0 , and A 1 0 . 2  2  3  2  3  Except for C l i f f samples and Z95-1, skarn samples have a larger content o f A l 0 2  (up to 22 wt.%) and therefore have a larger grossular component than magmatic garnet. C l i f f samples have more A 1 0 2  3  than magmatic samples but less than other skarn  samples, and have the most properties in common with, and are also physically closest to, syenite magmatic samples. C l i f f samples show irregular zoning in which darker zones have the same properties as early formed magmatic zones; A 1 0 2  3  and C a O have  the same oscillatory zoning pattern, opposite to that shared by T i 0 and F e 0 ; and the 2  2  3  change in wt.% from point to point for T i 0 is either the same or twice the wt.% change 2  for F e 0 . C l i f f samples likely formed from early fluids driven off the pluton. Skarn 2  3  with larger grossular components likely formed from infiltration o f later and likely more evolved plutonic fluids, as they possess similar characteristics to later magmatic stages  3  2  64 including irregular to regular, or the absence of, zoning; darker zones which show more A 1 0 , and less T i 0 and F e 0 2  3  2  2  3  tnan lighter zones; T i 0 and F e 0 share the same 2  oscillatory zoning pattern, as do A 1 0 2  3  2  3  and C a O , opposite to that o f T i 0 and F e 0 . In 2  2  3  summary, C l i f f samples formed from early formed plutonic fluids, Bartnick sample BN15-1 was intermediate as it contains properties common to both C l i f f and other skarns, whereas Bartnick sample 97BN1-4 was likely the last to form from more evolved fluid, owing to its large euhedral grains, zoning similar to later magmatic samples, and properties such as anisotropism and twinning which are not found in other samples. 4.2 Representation of Analyses Thompson space representation expresses mineral analyses as end-member mineral chemical formula with variable exchange components (Russell et al. 1999). The twelve linearly independent components used to describe garnet compositions from this study include the additive component andradite, and eleven exchange components: A l F e ( 3 + h CrFe(3+)-, FeMg-, H 4 S i - , M g C a - , M n C a - , NaFe(3+)2Ca-, TiMgFe(2+)-, T i S i - , VFe(3+)-, and ZrSi-. For a complete discussion on how to convert to Thompson components, see Appendix B in Russell et al. (1999). Thompson component representation readily discriminates between titanian andradite derived from magmatic vs. hydrothermal environments, as different substitution mechanisms operate in different environments (Russell et al. 1999). Magmatic garnet always has zero H 4 S i - and positive FeMg-, whereas hydrothermal garnet always has zero F e M g - and positive H 4 S i - (Russell et al. 1999). Russell et al. (1999) state that T i in hydrothermal andradite is expressed as T i S i - and exploits deficiencies at the tetrahedral site, whereas igneous titanian andradite contains both T i S i - and T i M g [ F e ] . and uses coupled substitutions involving octahedral 3+  2  65 cations. The exchange components do not directly represent crystallographic substitution mechanisms (Russell et al. 1999); however, the values o f those components which represent the only exchange mechanism by which an element can enter the formula are equivalent to the number o f atoms per formula unit, whereas two exchange components involving the same substitution (eg. T i ) do not represent an equivalent number o f atoms per formula unit. C a is used as a linearly dependent component and should be zero for all analyses. For those analyses which have non-zero values, the implication is that the C a component is not redundant and that these garnet would lie in another compositional space, possibly grossular space, as compared to the titanian andradite space. The andradite component is always perfect at 1.000. The norm o f the vector o f exchange components provides an idea o f how far the analysis is from the true andradite composition which is an exact zero value. The norm o f each analysis is determined by squaring each component (excluding the andradite component), adding all those values together, then taking the square root o f the sum o f squares. The norms o f the vector for all o f the magmatic samples (as shown in Table 4), including the two dykes, are less than one. A l l magmatic samples have varying amounts o f both T i S i - and TiMg[Fe "]. , zero residual C a , negative M g C a - component, very small 3  2  H 4 S i - component, and positive FeMg-. The pyroxenite sample Z M 2 6 shows more skarn-like component values especially near edges and cracks in crystals, therefore these reflect infiltration o f a latestage skarn fluid and do not reflect the original pyroxenite values. The H 4 S i - component values for Z M 2 6 are quite low (less than 0.06), as compared with skarn H 4 S i - values, but  66 are accompanied by zero F e M g - components. The melasyenite sample Z M 8 for some analyses (randomly distributed across several grains) has a large H 4 S i - component (up to 0.09) and zero FeMg-. Z M 1 6 has a zero or slightly positive M g C a - component, randomly distributed across several grains and across several BSE-visible zones. Syenite sample Z M 1 8 has some very large amounts o f H 4 S i - component, up to 0.16, and zero F e M g - component, mainly located on rim analyses, as does Z M 4 3 with amounts up to 0.32, and Z M 3 9 A has values up to 0.17. Z M 2 7 predominantly has nonnegative M g C a - components, and Z M 3 9 A also has some non-negative M g C a components. Larger H 4 S i - components at the rims may indicate infiltration o f mid-stage skarn fluid. The cumulus garnet sample Z M 3 9 B has predominantly negative M g C a component, but also has many non-negative values (up to 0.05), whereas WB66-5 has positive M g C a - components (up to 0.11). The volcanic analyses have exactly zero H 4 S i for all points (and positive F e M g - values), whereas Z M 3 9 B shows some slightly elevated H 4 S i - amounts up to 0.10, consistently located on grain rims and accompanied by zero FeMg-. The magmatic dyke samples predominantly show small norm values but also have several values o f one or over, and zero or slightly positive FeMg-. Z95-7 has many large values o f H 4 S i - component (up to 0.2), whereas Z95-6 also has some large values of H4Si-, but these are less common than in Z95-7, and are accompanied by large norm values. In both samples, larger H 4 S i values are located near cracks, in dark zones, and near rims o f grains, however entire grains may display these values where crisscrossed by  67 small cracks. C l i f f samples also have norms less than one and therefore share yet another property with the magmatic samples, whereas most other skarn samples have norms greater than one. The C l i f f and Glacier Z95-1 samples generally have equal amounts o f both T i S i - and T i M g [ F e ] . , zero residual C a , negative M g C a - component, zero or 3+  2  positive F e M g - , and small or zero H 4 S i - component, except for several points in both rock types which have large H 4 S i - components located near cracks or rims. The Glacier calc-silicate sample and the Bartnick BN15-1 have intermediate norms greater than one, whereas 97BN1 -4 has the largest norm (1.63) and is therefore furthest in composition from a true andradite. The two samples with the largest norms (Z95-5 and 97BN1-4) also have some o f the largest values for the M g C a - component, up to 0.04. A l l three samples show zero residual C a , relatively no average TiMg[Fe ]_2, 3+  predominantly positive M g C a - component, zero or positive F e M g - , and a mixture o f both small and large (up to 0.20) H 4 S i - components where large values are located at points in dark zones, near cracks, and sometimes at the rim. Figure 26 illustrates the division between magmatic and skarn samples and also the more subtle divisions which exist within some sample types in terms o f T i S i - and T i M g [ F e ] . exchange components. Igneous samples have greater quantities o f both 3+  2  T i S i - and T i M g [ F e ] . , zero H 4 S i - and positive F e M g - , whereas true hydrothermal 3+  2  samples plot in the negative quadrant for both T i M g [ F e ] . and F e S i M g H 4 (FeMg3+  2  component minus the H 4 S i - component) (Russell et al. 1999). Skarn samples plot between the two extremes. The two pyroxenite samples plot separately, especially in terms o f the A l F e - component, likely indicative o f skarn fluid infiltration i n Z M 2 6 as  68 other skarn samples plot in the same position. However, the two samples may have a slightly different origin. A s suggested by Coulson et al. (1999), in addition to the pyroxenite border phase, there may exist two additional pyroxenite units one o f which originated from a liquid not a crystal mush and occurs as fine-grained dykes, whereas the other unit consists o f large tabular bodies that contain stringers o f calc-silicate. Though neither sample is reported as a dyke, the possibility o f different origins should not be discounted as Z M 4 2 is finer grained and Z M 2 6 contains some calc-silicate. The melasyenite samples also plot independently o f one another also likely as a result of skarn fluid infiltration which has affected Z M 1 6 more than Z M 8 , as Z M 8 shows larger amounts o f both T i S i - and T i M g [ F e ] . , and plots lower on the AlFe(3+)- diagram. The 3+  2  dyke samples also show evidence o f minor skarn fluid infiltration and some points plot quite high on the AlFe(3+)- diagrams. However, both dyke samples are clearly magmatic as both have large values o f both T i S i - and T i M g [ F e ] . . The two Glacier samples plot 3+  2  separately, especially in terms o f the AlFe(3+)- component. Z95-1 exhibits a stronger magmatic affinity with larger amounts o f both T i S i - and T i M g f F e ] . ^ and smaller 34  AlFe(3+)- components. Some points for Z95-5 plot in the lower left quadrant on the F e S i M g H 4 diagram which is indicative o f more true hydrothermal affinities. Both Bartnick samples show stronger hydrothermal affinities than do other skarn samples and have almost zero or negative T i M g [ F e j \ , large AlFe(3+)- components, and some points 3+  2  (mostly from BN1-4) plot in the lower left quadrant on the F e S i M g H 4 diagram. Those points, such as B N 1 - 4 , which are "more hydrothermal" than other skarn samples, but do not plot as low in the lower left quadrant on the F e S i M g H 4 diagram as "true hydrothermal" samples (as defined by Thompson component work by Russell et al.  69 (1999)), will hereafter be referred to as hydrothermal. C l i f f samples both show magmatic affinities. 371 plots very low and slightly apart from 393 on the AlFe(3+)diagram. Figure 27 further illustrates the division between magmatic, skarn, hydrothermal, and dyke samples in terms o f Thompson components. The Zippa igneous samples fall into two groups: those points which have not been infiltrated by skarn fluid; and those points which have been infiltrated, in which case these points plot higher on the AlFe(3+)- diagram and near the lower left quadrant on the F e S i M g H 4 diagram. The dykes plot mainly between the two igneous groups. The skarn samples plot between igneous, rich in both T i S i - and T i M g [ F e ] . components, and hydrothermal affinity, for 3+  2  which points plot in the lower left quadrant on the F e S i M g H 4 diagram. The range is best seen on the AlFe(3+)- diagram in which three groups o f skarn are discernable. Those samples richest in the AlFe(3+)- component are hydrothermal and plot near the top o f the diagram, skarn samples intermediate between magmatic and hydrothermal affinities plot in the middle o f the diagram, whereas those skarn samples closest in affinity to the igneous samples plot near the bottom with varying amounts o f the T i S i - component. In summary, magmatic samples show small norms, negative M g C a - components, very small H 4 S i - , positive F e M g - components, and contain both T i S i - and T i M g [ F e ] . 3+  2  components; whereas hydrothermal samples predominantly have positive H 4 S i - , zero to slightly positive F e M g - , norms greater than one, and T i expressed mainly as the T i S i component only (Figure 28). Skarn samples are more similar to the magmatic samples (which plot furthest left on Figure 28f), although some skarn samples show component characteristics intermediate between the magmatic and "true" hydrothermal extremes  ' 7 0 (which plot furthest right on Figure 28f). Magmatic dykes have experienced skarn fluid infiltration as these samples show zero to slightly positive FeMg-, and predominantly show a large H 4 S i - component.  Oxygen fugacity and silica activity Titanian andradite compositions r e f l e c t / ^ conditions at the time o f formation (Virgo et al. 1976a), specifically a strong correlation exists between increasing T i M g [ F e ] . component and decreasing f ,  and T i S i may be considered a proxy for a  3+  2  Q2  s i 0 2  (Russell etal. 1999). Russell et al. (1999) state that the Z M P evolved to a higher f  Q2  state in  conjunction with the evolution o f a fluid phase, and that skarn samples inherit f  from  Q2  the magmatic fluid but have an increased a  SiQ2  as a result o f cooling o f magmatic  volatiles and interaction o f the fluid with quartzite. L o w / ^ values for titanian andradite are indicative o f crystallisation at high pressure, however H diffusion and sulphur loss 2  are two mechanisms which can result in pressure, and c o n s e q u e n t l y ^ , changes (Virgo et al. 1976b). Natural garnet become more and more silicon-deficient with increasing f  Q2  (Huckenholz et al. 1976). Data from this study as shown in Figure 27 ( T i S i - vs.  T i M g [ F e ] . plots) concur with these findings and show that the igneous samples show 3+  2  lower f  Q2  (i.e. they plot higher on the T i S i - vs. T i M g F e - diagram) and lower a  s j 0 2  than the  skarn and hydrothermal samples which plot closer to the origin, indicative o f crystallisation at higher f  Q2  values from a more evolved fluid. a  sio2  is also increased in  the skarn samples, also indicative o f evolution o f the original magmatic fluid to a slightly more oxidizing fluid (negative TiMg[Fe ]_ ) and is even more increased in the 3+  2  hydrothermal samples. Dykes show some effects o f interaction with both the original  71 fluid (low <2 o ) and a slightly more evolved fluid (higher a Si  2  S i 0 2  andf ), but show almost 02  no indication o f interaction with the highly evolved hydrothermal fluid which shows the highest a  s i 0 2  and/o signature. 2  4.3 S u m m a r y Five different titanian andradite types are identified on the basis o f zoning and Thompson components which include: magmatic " A " or primary magmatic samples which include Z M 3 9 B and W B 6 6 - 5 ; magmatic " B " which include most magmatic samples (e.g. Z M 1 6 ) ; skarn " A " which can be further subdivided into two groups including C l i f f samples which show a magmatic affinity in chemical zoning (e.g. 371), and most other skarn samples excluding B N 1 - 4 (e.g. Z95-1) which grade between skarn " A " and skarn " B " ; skarn " B " consists o f only the hydrothermal sample B N 1 - 4 ; and dyke samples which have a magmatic signature overprinted by skarn " A " fluids (e.g. Z95-6).  72  QQ i  CD 2 wlco N co CD 3  O CM  T -  CO  LO LO  CN  O  CM  O -tO t>O O  CN  u<5  LO o LO C O o CD oo co OO CM o o cp o o o CO o CM d d d d d d d d CM LO  oo CO CN oo roo CN CD LO M o od d d C 00 d CD CM  00 CO o CD o CN o CO CO CN o o cp o o O 5) o CM d d d d d d d d CM  CD CD o o CM o CD O O co CO o 00 o d O CN ob CM d d 00 d CD  CD CD O o CM r>- o oo 00 O O o o o o C O CM d d d d d d o d CM  CD  00  OO LO LO co CN co o co N od d d C CO d CD LO  LO  CN  CD  LO  00 T - CD O CM CM O CM O O  to  T -  T—  N 00  T -  CD  O  N. Og  5 CO  CD  00 co o o o co LO h- O ^ CD d CN d d T -  CD CN  T -  T  —  CD co co o O d CD d C CD LO T—  SO  o O oo  T—  T—  CD x— CN O o o CN CD CM o o cp o o o oo CM d d d d d d d d CM T—  T -  ^  CO*  1  CD CM o o o co o LO o CN o o o D CN d d d d d d d o C CM  CO CN CD CO CO CO CD O CO N CO d d C CO d CD  CD o co o CM CM CN LO LO 00 CM o o LO O o o CD C M CM d d d d d o d d CM  CD LO oo CD oo O CO CO Od CM d d C CO CO  CN CO o CO o o oo CM o o o O o o co co o C O CN d d d d d d d d CO  O O CM 00 CD CD I— O CM o CD oo o o ro d cci d O N- d d CO d CD oo CD  CO o o o CO CN o CD o o CO o o CO co D o CM d d d d d CM d d o d C  O o M B N CD I- co r>- C r>- i - co CD oo C O CM „ CD CM S 2 co •sj-' d -r^ O O CM d d CO d CD ^ oj co  CD CN oo o r — CO CN CM O o 00 o o LO o \— oo d d d o oO CM d d d d o d C  CO CD 00 d d  T -  N =S co 5 ©  to  -2  SI  d  LO O LO o o co CN O 00 CO CO ti CM o o  ^3CN  oo co CD oj  CM CD CD hO o co V - O LO d 9> CO LO o o N CO oo  oo  X—  c  SO  CD  T "  SO  CD CO CO CO o CM d 00 OO CD  LO O f O HI o ri d t d  CO  T—  o> co O CO  I  o5 J3> 1  Kl  '£  LO LO ^  05 CO  § s  S  N ><  § s  N  to o CO ct  N  i  1  n  6  O O O O ra CD c D) co ZcotN o u - 22 2 O  ^?  CM  c o f - N < 0 > '  ( -  CO H  N  <  O  >  CD  LL  CD  D)  CO  2 O  73  T— co CO o o o o o T— cn o o o o ci co" co 00 csi  1— CO m CN o o o o o cn o o o o O CO CO oo c\i T—  O O rO •<- O  O l O t O l O T - T - t O  cn  o o o o o o o o o CN o o o o o o o d ci cl  t- CD ci  o m o c n o c M c o o c D O T - T - o c o O  T  -  O  C  N  O  O  O  O  T - o d d d d d d  O  O  O  O CO  O  d d o d d c i  1 —m o o o o o o C J> o o o T— ci d CO CO oo CN  o •<*• o T - O N i n o w m r s o s o o o - - o o o Oo Oo Oo Oo Oo Oo Oo Oo Oo i o-  o CN o o o o o cn o o o T— ci ci CO CO 00 CN  O O O O h - O - r - C N T - T - r - ^ O O O  m CD cn m o o o o o o o o o o o Ci CO •<— CO 00 CN  ocoo-tOt-coo'sj-T-T-Lnoco  o cn CO oo o o o o o C O o o o d ci CO T— CO oo CN  O T - o oo o c N c n i n c N O - t - o o c N o o o o o o o o o c o O r- O CN  cn CD •si- o o o o CD — t o o d O co CO oo CN  o^otoOT-ooi-miDtoo)  CN CD oo CD o o o O o oo o o o ci ci CO cd 00 CN  CM O CO o O  ci O CO  T—  O  O  O  C  N  O  O  O  O  T  -  O  O  O  O  C  O  ^ d d d d d d d d o d d d d  o-<-o^rooooooooo^r r  d  d  d  d  d  d  d  d  d  d  d  d  d  t - o d o d o d o d d o o d o  O  T  -  O  O  O  O  T  -  O  O  T  -  O  O  O  C  N  • > - d d d d d d d d o d o d d  C D O C O O T - C D O - I - C N T - C N O V -  T- o o o o o o o o o o o o d d d d d d d d d d d d d  CM o o  f-r-cnt-cN-'d-OT-cocNOOo  CO 00 CN  d d d d d d d d o o o o d  o o  •  •  -  O  T  -  O  O  T  -  O  O  O  O  O  O  C  O  TJ 0  C C co  CN  "Jo ™U>2CO toX  CO  CO CO -*—>-*—'  o O  •^r X)  eg  CO ^  Z  X  g o3 o2  o to to c c: o o •-  o o  +  _ i tfc_ _ N < O> H  co O CN to o X  CO o  oz  74  o o CO •sr o  oo  o co LO o CO -sr LO CM tooo o \— CO o •sr CO •sr O d CM d d CO d C OO o LO o «r o CO LO o co ^r CM oo o C O CO co o d d LO d d co d o oo C M •sr co •sr o -*— LO coo CM co CM CO o o d o d d d LO o d co CD CO oo  Co it; O CO CM d LO d  LO  .73  oo oo LO co •sr o oo r^- cd d — cd t co h~ o co r-o O co d C •sr co T—  <c CO CQ CQ CO 00 CQ C CO co CD CQ LO LO co N CD CO i-~ co T - .0) CO CO  o CN CO •sr o CD* — I \— iv T— o •sr •sr o o d d oo d o d d CO d co T—  o o CO o CO CO 00 CO o •sr OO o co CM •sr o 1^o •sr O CD d CO d d d d L CO d CD T— T _  CD O OO r-oo 00 o co •sr •sr \— •sr CN CO •sr o oo o T— oo N •sr d d d LO d rd-~ C CO d oo C M N CD N- oo CM •sr o LO LO CO LO CN CO o o CD CM co o O ci CD CO CO d d d •sr d d C CO CO d CD CO CM oo .p CO ^ CD 00 o •si- LO co CM LO LO co o o CM CO -sr •sr o CM CD C M o 00 C O LO co cd d CO d d CM d d CO d OO o> co N LO o co •sr co co CO CN T— CD oo o L O G O L O o o C M C M <? CD C o O •sf d •sr O LO cd co d cd d d d d d CD ^ co co oo CM N Q  1  TJ  CD  c c o O  -sr _CD  n CO  oo l^~ CD CD  co o o LO o co o o o o 00 T - . O LO o o •>- d d d co •sr o o o CM o o CN CM d o CD o d o O O LO CN CN d d d CM o d d d d co o o o o o o o co cor^ocNO-r-'sroT-cNNOOOOCOOOCOOOOCD C N d d - r - o o o d d d c N •<- -sr o T - o T - LO o v - to •sr CD O O — I O O CN O O O CD CM d d d d d d d d CN r- t- o cn oo T - o CN O O 00 O T— CO CN d d T O O CO O O O CD LO  LO  O O  V -  T-  T -  d d d d d d  C D O T - C N . T - T - C Q . O T - C D C OCN O O C O O T - O O L O T T - O O C D  c N d d d d d T ^ d d d c N •<- CO  oo CNT - C O O O L O h ~ N - C O L O o - os roo oo ' s r o o o o o CN d o O O O CN OO o T - O O C M O O O T - L O C N 00 CN o c o o o ' s r o o o o CM d d d d d - r - d d d r o 00 T - i - - s r O ' < - O O C M h ~ C M h ~ co o o CM O o O -Os r O oOo c oo o oo o co  ©, S- -* CO CO  o O O P ct co — i N  o o o ^CDo£ oi ?oro c.o 0  <  o  >  CN  LL  S S O  Z  CO  N  <2_ *A_ &  < O >  ffi  LL  CD  O) CO  2  o  75  o o  CD  o d  o  s  I-  o  co  o  \—  co  , _  cd cci CM  CO  s  O o o co CO CM o o  o o  CM O  o  d  d  o o d  o CD CM I - — o CD o o o o o d CO T— CO cd CM  o  Io  d  d  CO  o  o CM IC-D CM o o o O o CO CO CO CM  o o  co  CM O  o  o o  I-  d  d  CO  o o  LO  o d  Is  CD  o  o o  o o  CO ob CM  s  d  o d  o  o q  d  CO o d  o  o  d  d  c o  o CO T  s  Is  cp  CD  o o  o o  CO CO CM  CN CO o o CD o o o co co CO CM o  CD o o CO o o o T co ~ CO ob CM x— —  O  co  LO  CM o o  CO  I\—  CO cb CM  co o in c o  s  o  o  co o o o d d  o LO  Io  *«— o  d  d  d  O  T  T  -  -  o o  LO  O  I  d  o d  -  d  O  d  O  o  T  ^  d  o o  CN CO  o o  o  d  T—  d  d  o  CD q O o  d  d  d  o  o o  LO  d  d  d  O  O  O  d  O d  O d  O d  CM O  o o  o  o  o o  d  d  d  d  d  O  C  M  d  d  d  CD o d  o o  CM co x—  d  -  >  -  CD CM  o o  d  d  o o  d  -I-  d  d  d  d  d  d  o o  CM o  o o  •ro o  d  d  d  d  d  d  d  T  o o  o  o o  d  d  d  -  CD CM T—  o c o t - c N T - T - r - O T - C M v - o o r - O T - O I - O O I - T - O O O O O C N t  o q  -  d  00 \— d  d  d  CD o  d  d  d  o o o o  d  d  d  d  d  d  d  LO o  Io  Io  o  CM o o  d  d  d  d  d  d  s  s  d  d  d  o o o d  LO  d  d  O C D T - O O C M T - O T - C D T - C O O M O O O C O O O T - O O O O O O CO •  •  -  d  d  d  d  d  d  d  d  d  d  d  d  d  o c t i T - t O i - n o r M t O T - ^ o o O ^  T  -  d  O d  ^ d  T  O d  O d  T  -  d  O d  CM  £ >,| inrox  o o  o o  o  d  OCMOT-Ot-CNOT-COT-h-OTO O O h - O O O O O O O O O O -  ~ 7-  o o  d  -  O  ro  goo + 22  o o  s  o  d  O T - o c n o o o o o c D C M ^ r o L O  \—  d  CM  CO s  s  "O OJ 3 C  CD o  I-  JL.  « co  CD  t N  ,J-> A. ,  CO CD  O d  O d  O  O  d  d  O  L d  O d  CO  o  ro  D CM "O "+~ CD .JL C D J - CO  in 2 , 2- co • i . co co to co oT CD — . !± 1 1 £ 1 CD J= coroI-^r < O > f— LL 2  76  G  400  Microns from core  0  400  Microns from core  Figure 10. Compositional profiles for ZM39C, grain a. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  77  Total oxides  34  ZrO2 _ 0  C a O  4  32  ^ ^ ^ J L j .  30 1.2 MnO  0  400  Microns from core  0.8  0  400  Microns from core  Figure 11. Compositional profiles for ZM39C, grain h. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  78  Figure 12. Compositional profiles for Zm39C, grain i. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  79  Total oxides  CaO  32  MnO  0.8  28 Fe203 24  0  400  Microns from core  20 0  400  Microns from core  Figure 13. Compositional profiles for ZM39C, grain n. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  80  Figure 14. Compositional profiles for ZM39C, grain o. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  81  Total oxides  34 CaO  M n 0  0  400  800  Microns from core  32  0.8  0  400  800  Microns from core  Figure 15. Compositional profiles for ZM39C, grain w. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  82  0  400  Microns from core  0  400  Microns from core  Figure 16. Compositional profiles for ZM39C, grain x. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  83  Figure 17. Compositional profiles for 69WB66-5, grain b. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  84  0  400  Microns from core  0  400  Microns from core  Figure 18. Compositional profiles for 69WB66-5, grain e. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  85  8  100 Total  4  A1203  oxides  0  92  0.8 ZrO2  36 CaO  0 4  1  0  32 28  6  1.5  A  Ti02  96  MnO  0.5 40 Si02  28 Fe203 24  3 6  32  0  400  800  Microns from core  20  0  400  800  Microns from core  Figure 19. Compositional profiles for 69WB66-5, grain n. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  86  Figure 20. Compositional profiles for 69WB66-5, grain p. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  87  Figure 21. Compositional profiles for Z95-5, grain f. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  88  Figure 22. Compositional profiles for Z95-5, grain I. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  89  Figure 23. Compositional profiles for 97BN1-4, grain b. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  90  34 0.4 MnO  40 Si02  3  0.2  !  6  32  r  1  0  500  1000  Microns from core  0  500  1000  Microns from core  Figure 24. Compositional profiles for 97BN1-4, grain c. Oxides are in wt%. Error bars represent 2S analytical uncertainty.  91  Figure 25. S E M images of grains showing the lines along which compositional profiles were constructed. Analyses were done at 20 \in\ intervals, except for 69WB66-F which were done at 10 um intervals.  92  Figure 25. Continued.  94  CM  O  V  CO  CO O LVL. CM < U CM - o CO LVU CO COco" o co" o o O CO CM O co o CM CMCM CM C<M oUV. O CM <'. < "-3. 7? CM < c m " c m " c " cm" CM A Q Q q m o CM o j— t— A cm" i— < O A CO  CO  c o c o sCo CO N  p  CN  o  0) CD  CM CO  A  CO  A  *—  CM CM CO COO O i-V Vp  CM  o o  CO  CO LVL o o tM  co" co"  CO CM CO CO O O 5o o o 3o "A' P £M ACM CO COCO CO CM ft,CM< CM" o o o o O O o CM CM < < < < CM  CM  CM  to 0  c  ra ra  O N T3 CD t~  OS CJ) 3  3  3  TOD) }>  ra o  2;  •2 '5 5 N  c  >. >s ^ 'ro  E  'ro  E  'co  E  >. ^- ^ 2 w E 5 t  'co  _ S  w  t  3 TO CO  "c N  N  t co «> 3 .E £ ra co TO 2 co c <o  O)  o  N  iS  3 ^ TO TO  E*  TO  O)  TO  N  N  N  TO CO  TO (0  '• ~  3  1  g  t  t  o o o  £ sz co cn  c o  ro i—  c  XI  SZ CO  CO  ••g  c CO  ii  •a  CO  ,_ .ti c 2 CO  o c  co  »=  ~ J2  c  «-  ra •E  o  CO  o ra o  o  c o o  xf xf x: "o •o 5 i  xi" x: "5  CD >  O .ti CO o  ra o  L_  TD CD  o •jr= i2  s  x:  co  CO  ra o  CO  3  Q.  - Tro J  CO  o  X) CO  3 to  c CO  3 CO  sz  O C O C O L O C O C O C O C O t O  T> d)  ral  ra  't_ O in CD T> CD  x:  T?  TJ CO  c  -Q  TJ CO  x:  anihe  B  TJ CO  O  xi 3  U>  TJ  ral  ro 1  ro CD  TJ CO  bhi  .3 o  CD O  CO  <D  TJ CO .C C CO  co x: c TO ro £  —  co o •= J 3 Li e 7  L-  ro in CD  c o NI CD  CO  >. CO  co  c c c c g c  JS CO  CO v-  CO CO CO CO > . > - > . > , to CO w 00 CJ) tif O  00  00  E  W) CM CO 2 2 2 2 2 2 2 N N N N N co o> cn C O C O O O o CMCMID -ef  Ni 00  00  r—  NI 00  OO  co O  co O  co O  o O  3 o  <C D o>  E  O) to  S E E  ro  §  ro  -§  O  JS  JS  M i l co  CO  N NI  N  S £ « co co o o o  o  fc .2  rag, > o o m m o o  2 2  00  ro  E  00  cn  5  r-  co  co  io ID ^ T - ro cn s ^ s co IO  NI M ID NINI CO CD CO CO  "O CD  in CD T> X  O  95  C  o3 +  .c '  o3 +  06  + +  08 +  E CD T3  s  E a. a. o  ro O >.  CM <D  co 0  a.  ro O  CM  0 u. Q . CU cf o u. ° 0 == 'o 10 o CM ro O W §  1 0  (U NI  S3 -E  E  o  g  m=  75 =  o ro O CM  c =  ro  O  CM  < CN  1g  20  "0 co 0  £•  0  ro 0 co 0  E  O CD  cu  E  S- u-  CL  co CD  > CO ro ro o n u t^  7C 3 C"O LL. •D c  < TJ  LL. "D  CM  CM  o  o  c c ca ro ro  ra ro CM CM CM o o O  CM O  O  0 0 0 0 CD ro ro ro  o £ ro =  a.  cu  ro ro O O  CD > O  0  CM  8  £ o o  CD >  CO  O O P t= P  2-  0 0 0 o 13 3s JS ro JS 0 0 0 0 0 co CO CO CO CO 0 0 0 O 0 0  ro ro "ro  0  "ro  ro  "ro  CO  c g ro CD  1  8 CD  >  ro cn CD  c  "D  C  "ro  co CD > 'co  o  CL CM  CD  2  M CO  cn O  2  o> ox o>  CD C 1*CD  o  cn ro 1  CD CD  NI  no 10 S  2  co  0 0 0 0 0 cn  o NI  o> NI  o CO  "D m O) NI  10 I  LO O) NI  ro ,Tro h-- ro O  03 +  96  Pyroxenite  -0.20  0 00  0.20  TiSi  0.40  -0-20  000  TiSi  020  Melasyenite 1  1  1  • T  -  zm8 zml6  1 0.00  0 20  TiSi  0*0  -020  0.00  TiSi  1 020  Syenite  0«  -0«  -O20  1  0.60  0.40  O.X 020 FeSiMgH4  040  i  •  i  * zml8 * zm27 * zm43 ^ zm39a  -  * 0.20  +  *  0.00 f  Q.X  020  TiS  0.40  -020  0.00  TGi  020  ,  it f  i  .  i  ,  t  0.00 020 FeSMgB4  Figure 26. Thompson component plots for all samples from this study. For each sample, plots include exchange components TiSi- vs. TiMgFe(3+)2-; TISI- vs. AIFe -; and ((FeMg)-H4Si) vs. TiMgFe(3+)2-. 3+  97 Glacier  Bartnick  2 00  ,  .-  97BN1-4 O BN15-1  »U  3  • mi— fl.QG  020  FeStMgW  Cliff  *  -  -  ,  0.40  Figure 26. Continued.  -0.20  !  -0.2D  0.00 0.20 FeSIMgB4  0.40  0.60  98 Dykes  0 60  Cumulus  0.00  TiSi  0 20  0.40  -0.40  -0.20  -0.20  0.00 0.20 FeSIMgH4  0.40  Volcanic 1  I  1  1  • 69WB66-5 -  -  _  •  0.00  i 0.40  Figure 26. C o n t i n u e d .  .[)20  0.40  •  j-  o*o  •0 20  c  1  0.00 0.20 FeSiMgH4  ,  1  . 0.40  0.60  99 i  Decreasing aSi0 8  , • •  O Igneous • Crowsnest]  2  A  -  Decreasi  _ rr  0  OQ)  o  cP  —  o  o o 0 • °c . 0 00  TISI  °  1 0.20  t  -0.40  i  -0.20  0.40  1  !  0.GO  1  1  A Skarn «  -  -  -  5 •""020  >* 0.*J  -0.20  0.00  TiSi  -0.40  020  1 Decreasing aSiO,  I -0 20  ,  1 •'  1  X  -  Q,  -  CD  ,  1  ,  040  1  ****  0 60  1  Dykes -  040  * u  1  0.00 0.20 FeSiMgW  -  £  ca  a  *~0.20  X  X tt<v  -  0.00  j-—  •O.20  0.00 0.00  TiSi  0.20  i  1 0.40  -0.20  0.00  TiSi  0.20  040  -020  ' , 0.00 0 20 FeSiMgH4  0.40  F i g u r e 2 7 . T h o m p s o n c o m p o n e n t c o m p i l a t i o n p l o t s for a l l i g n e o u s , s k a r n , a n d d y k e s a m p l e s f r o m this s t u d y . Plots a r e a s in F i g u r e 2 6 .  0 60  100  c)  0.40  Q80  120  160  2  Norm  0  040  080  120  Norm  0  040 060  120  160  2  0  Norm  160  2  0 . 040 060 120 Norm  Skarn  040 060  120  160  2  160  2  Norm  160  2  0  040 060 120 Norm  Figure 28. Thompson component compilation plots for all igneous, dyke, and skarn samples from this study, a, b, c plots are norm of the vector vs. H4Si-; d, e, f plots are norm of the vector vs. TiMgFe(3+)2-.  101 Other Chemical Techniques Used to Describe Titanian Andradite  5.1 Introduction To fully characterise titanian andradite, several other chemical techniques were used including wet-chemical analysis, and FTIR spectroscopy. Wet-chemical analyses were performed at M c G i l l University, and infrared spectroscopy was done in the chemistry department at the University o f British Columbia. 5.2 Wet-chemical Analyses Ten samples were analysed for F e O content by wet-chemical analysis for use in establishing site occupancies o f the iron cations. Samples were chosen to represent each o f the different rock types and environments o f formation in this study, also X-ray diffraction was done on those samples on which duplicate analyses were performed; whereas other samples were chosen based on the amount o f material available for analysis as each analysis requires at least 0.25 grams of pure powdered crystal. Analytical precision o f FeO measurements was tested on five o f the analysed samples through duplicate analyses (Table 6). Several o f the duplicate analyses differ outside o f detection limit error because o f the small amount o f sample available for analysis, and because o f slightly inhomogeneous powders caused by small inclusions within the crystals. Figure 29 shows comparisons of duplicate analyses for the five samples. Both analyses for each sample should plot on the 1:1 line. Deviation from this line is caused by the above mentioned factors as well as by analytical discrepancies. 5.3 FTIR and Estimates of O H 5.3.1 Literature Review of O H Site Occupancy  102  Solid solution relationships exist between andradite and hydroandradite, and in such H-rich garnets, OH" likely occupies the tetrahedral sites and substitutes into the site according to the reaction ( O ^ ) " <-»• (Si0 ) " (Basso et al. 1981; Rossman and Aines 4  4  4  1986; Lager et al. 1989, Matsyuk et al. 1998; Amthauer and Rossman 1998) which requires silica undersaturation at the tetrahedral site (Onuki et al. 1982; Lager et al. 1989; Armbruster 1995; Matsyuk et al. 1998) and charge balance is achieved by 0 H (Lager et 4  4  al. 1989). Certain authors propose that OH" substitutes into the octahedral or dodecahedral sites as a coupled substitution, such as with Fe (Amthauer and Rossman 1998; Basso et al. 1981), however Lager et al. (1989) repudiate this idea. Armbruster et al. (1998) indicate that T i  4 h  bearing octahedra are favourably surrounded by O H groups  or by ( F e , A l ) tetrahedra. These authors also state that previous authors who assigned ,+  3 +  A l to the Z site to balance the Si deficiency were unaware o f the hydrogarnet substitution and were therefore in error, and all A l should be assigned to the octahedron. Peters (1965) and Armbruster et al. (1998) state that substitution o f H for Si causes an increase in cell edge, a decrease in refractive index, and possibly birefringence in garnet caused by low-symmetry distribution o f the O H groups (Rossman and Aines 1986). A l l T i must be in the 3+ valence state and at the octahedral site to correct for the increase in cell dimensions (Basso et al. 1981; Onuki et al. 1982; Lager et al. 1989; Armbruster 1995; Armbruster et al. 1998). OH" content in titanian andradite can be up to 2.5 wt% H 0 , and 2  increases with increased Ti-content (Armbruster 1995; Amthauer and Rossman 1998). The OH" content o f titanian andradite from volcanic suites ranges from 0.01 to 0.04 wt% H 0 , whereas OH" content o f skarn titanian andradite falls between the igneous and 2  volcanic extremes (Amthauer and Rossman 1998).  '  •  103  5.3.2 FTIR Methods Infrared spectroscopy is an extremely sensitive method for detecting trace amounts o f water within crystal structures (Wilkins and Sabme 1973). A few parts per million water can be detected in a mineral using this technique which can produce O H stretching bands in the absorption spectrum usually between 3600-3000 cm"' (Wilkins and Sabine 1973; Rossman and Aines 1991). Pressed pellets were made for all samples from this study using 200 mg K B r mixed with 2 mg powdered, hand-picked single crystal titanian andradite. Qualitative infrared spectroscopy was done for several samples from this study by Lee Groat, however no water was detected. The samples either have no water present, or this water was insignificant enough to be masked by background noise.  Table 6: Wet-chemical analyses of FeO in select garnet samples.  Sample  FeO (%)  Fe 0 2  3  (%)  Z951 Z951 ZM16 ZM16 ZM39B ZM39B BN14 BN14 69WB665 69WB665 ZM26 Z956 Z957 Z955 BN151  1.66 1.78 3.17 3.35 2.16 2.12 1.47 1.19 3.12 3.13 1.71 1.75 1.23 0.54 0.68  23.24 23.11 21.16 20.95 23.65 23.69 . 3.28 3.57 19.74 19.73 15.33 18.78 23.16 10.74 11.12  Detection limit(%):  0.01  0.01  Note:  Total Fe as  Fe 0 2  3  25.09 25.09 24.68 24^68 26.05 26.05 4.91 4.91 23.21 23.21 17.23 20.72 24.53 11.34 11.88  Total iron provided by microprobe analyses FeO determined using ammonium metavanadate titration  105  Figure 29. Comparison of F e O concentrations for duplicate analyses from several samples. The 1:1 line indicates identical duplicate analyses of perfectly homogeneous crystals which lack inclusions.  106  X-ray Diffraction Analysis  6.1 Introduction to Diffractometry Atoms are arranged in a periodic way in crystals. A s X-rays fall on a crystal, each atom becomes the centre o f a scattered wavelet which interfere with one another to produce an observable diffraction phenomena on a photographic plate placed normal to the incident beam (Guinier 1952). When X-rays o f known wavelength fall on a crystal, the determination o f the crystal structure is possible (Guinier 1952). To determine the whole crystal structure, it is sufficient to know the nature and positions of the atoms o f one unit cell (Guinier 1952). The Bragg relationship shows how the crystal spacing can be determined by measuring the angles o f diffraction i f the wavelength o f the X-rays is known. The Bragg relationship states " I f an incident ray, o f wavelength X, encounters lattice planes o f spacing d at an angle 9, it gives rise to a diffracted ray in the direction o f the ray reflected by the planes considered, on condition that nA,=2dsin9, where n is a whole number..." (Guinier 1952). Measurements o f diffraction angles o f X-rays allow us to determine only the point lattice o f the crystal but to determine the arrangements o f atoms within the unit cell, measurements o f the intensities of diffracted beams are required (Guinier 1952). Intensity o f X-rays depends on the arrangement o f the atoms of the base and the relative values o f the intensities are usually measured or estimated from lattice planes in the crystal (Guinier 1952). To deduce the structure o f a crystal from its diffraction pattern involves determination o f the reciprocal lattice, the determination o f the crystal lattice, and the determination o f the position o f the atoms in the unit cell. X-rays provide only the  107 measurements o f intensity o f the diffracted beams (Guinier 1952). Either powder or single crystals may be used in X-ray diffraction, however single crystals provide a more complete analysis as the diffraction patterns are simpler (Guinier 1952). There are two methods o f examining single crystals: the rotating crystal method in which a beam o f monochromatic X-rays is used, and the Laue method in which polychromatic radiation is used (Guinier 1952). Polychromatic diagrams are less rich in information than are monochromatic diagrams, they do not give the absolute values o f the lattice constants o f the crystal, and are only easily interpreted i f an important axis is coincident with the direction o f incidence (Guinier 1952). In summary, single crystal diffractometry is used to determine crystal structures which includes the measurement o f unit cell parameters, and the nature of the atomic arrangement. 6.2 The Andradite Unit Cell by Powder Diffraction Five samples, Z M 1 6 , Z M 3 9 B , 69WB66-5, Z95-1, and 97BN1-4, were selected for detailed X-ray work based on different rock types, different environments o f formation, and different chemical compositions o f andradite. The unit cell for sample 97BN1-4 could not be determined as the sample has strong and very fine chemical zoning which produces peaks which are wide and interfere with one another. Powder X R D analyses were carried out at the University o f British Columbia using a Siemens D5000' diffractometer at operating conditions 4 0 m A and 40kV, 0.04 step, 2mm divergence and antiscatter slits, 0.2mm receiving slit, and incident and diffracted beam Soller slits. Hand-picked single crystals from each o f the samples were powdered for X R D analysis using a mortar and pestle. C a F was added (1:2) to the powdered andradite. Clean glass 2  108 slides were moistened with ethanol, a small amount of powder mixture was placed in the centre o f the slide, and a probe was used to spread the powder evenly and in random orientation over the slide to prepare the smear mount. The resulting raw peak files contain both andradite peaks as well as C a F peaks, 2  which are very well constrained and can be used to determine a correction factor based on the difference between the accepted 26 C a F values and the measured 26 values once 2  Ka2 peaks have been stripped using either the E V A or PowderX programs. The correction factor found for C a F can then be applied to the andradite peaks for a 2  particular sample. Usually, an average o f the differences between several strong theoretical and measured peaks is adequate for use as a correction factor however, as was the case for Z95-1, i f each difference is drastically different and the average does not represent the differences for the majority o f the C a F peaks, an alternate course must be 2  taken to determine individual correction factors for each andradite peak. For sample Z95-1, a graph was constructed (Figure 30) o f the measured C a F 26 2  values versus the difference between the measured and theoretical C a F values. A best2  fit line was drawn through this data. The vertical distance from each point to the best-fit line, either positive or negative, is the correction factor for that point. A C a F point and 2  its corresponding correction factor was applied to the nearest measured andradite value to produce corrected andradite values. In this case, each andradite may have a different correction factor. Once the correction factors were determined and applied to the andradite values, the program Celref2 was used to determine the actual unit cell o f the andradite for each sample, as listed in Table 7. Values used in this program include: A"ctl = l .5406;  109 26,=15.000; 26 =148.000; system=cubic; space group=/a3ci; CT=12.045A (for Z95-1, 2  a=12.056A after Deer et al. 1997). A pure quartz-CaF mixture was done under the same conditions as the andradite 2  to monitor the accuracy of the X R D . The quartz unit cell is known to four decimals and is highly reproducible. The results are listed in Table 7. The measured quartz values are within one significant figure and accurate to the third decimal o f the accepted quartz unit cell  :  6.3 Single C r y s t a l X - R a y Diffractometry The methods used to obtain single crystal X-ray diffraction data involve several steps including crystal selection, data collection, and structure refinements. The first step in single crystal X-ray diffractometry is to crush some o f the selected sample and pick six or seven grains which are placed in the air grinder. The crystal selected for analysis must be a true single crystal, round, and 0.2 m m in size. The selected crystal was mounted on the end o f a glass fibre using epoxy. The fibre was mounted in a Siemens P3 automated four-circle diffractometer operated at 50 to 55 k V and 25 to 35 m A , with graphite-monochromatized M o K a radiation. After taking the photographic plate, twenty-five strong reflections distributed in more than one octant o f the reciprocal lattice (Basciano 1999) were chosen and entered into the computer. The four-circle centres these reflections and calculates probable cell parameters and the orientation matrix required to collect a data set (Basciano 1999). The correct unit cell was determined using 50 high-angle reflections in the range 54 to 59 2q and least-squares refinement produced the cell dimensions for each crystal (Lam 1998).  no Intensity data were collected in the 0-20 scan mode, using ninety-six steps with a scan range from [20(MoATa,)-l.l°] to .[26(MoATa,)+l.l°] and a variable scan rate between 2.0 and 29.37min depending on the intensity o f an initial one second count at the centre of scan range. Backgrounds were measured for half the scan time at the beginning and end o f each scan. The stability o f the crystal was monitored by collecting two standard reflections after every 23 measurements.  There were no significant changes in their  intensities during data collection. Data was then collected for the absorption correction. One octant o f reflections was collected from 3 to 60° 29. Ten to 14 strong reflections uniformly distributed with regard to 29 were measured at 5° intervals o f vj; (the azimuthal angle corresponding to rotation o f the crystal about its diffraction vector) from 0 to 355° (Basciano 1999). These data were then used to calculate the absorption correction, which was then applied to the entire data set. The data were also corrected for Lorentz, polarization, and background effects, averaged and reduced structure factors. Structure solution and refinement were done using the Siemens S H E L X T L Version 5.03 system o f programs. For structure solution the heavy atoms o f the mineral were located using the Patterson method followed by inputting the lighter elements. Once a trial structure has been obtained, the refinement process can be carried out. The three fractional coordinates for each atom are refined except where the origin o f the unit cell must be fixed or where an atom is situated on a symmetry element, which causes one or more o f the atom coordinates to be fixed and not refined (Basciano 1999). The lowest R value was found through various steps including setting cations and possible anions anisotropic, restricting observed data to larger a(F) values, omitting individual  Ill reflections, using a weighting factor, and refining atoms versus a substituting element. Once heavy atoms have been located and refined as a trial structure, a difference Fourier wall usually yield the positions of the remaining non-hydrogen atoms. Before the final refinements of the structures, the program STRUCTURE TIDY was used to standardise the atomic positions (Lam 1998). 6.4 Summary Powder X R D analysis for the volcanic sample indicates a cubic cell and a small cell volume, which eliminates the possibility of large amounts of undetected water, as water substitution increases the cell edge of the tetrahedra. ZM39B shows exactly the same cell dimensions and volume as the volcanic sample, which indicate yet more similarities between the two, aside from those determined by E P M A analysis. The melasyenite sample (ZM16) shows slightly larger cell dimensions and a larger volume. The larger volume may be indicative of O H substitution, however Virgo and Huckenholz (1974) statethat both Fe  3+  - O and T i  4 +  - O bond distances are larger than the Si-0 bond  length and substitution of either (or both) cation(s) on the tetrahedral site accounts for the cell parameter increase and also a cell volume increase. The only skarn sample tested with powder X R D shows a volume increase which may be indicative of tetrahedral ly coordinated water. Figure 31 shows the relationship between cell volumes and chemistry. A general linear correlation exists between cell volume and amount of T i 0 (Figure 31a) as well as 2  between cell volume and the TiMg[Fe ]. Thompson exchange component (Figure 3If). 3+  2  The TiMg[Fe ]. correlation is much stronger than the T i 0 correlation. The trend in 3+  2  2  Figure 31 f continues through Ti equal to zero, or ideal andradite. This may indicate that  112  cell volume varies with T i 0 content and that the titanium substitution mechanism may 2  be via the T i M g [ F e ] . exchange component for all andradite crystals. While Thompson 3+  2  components do not represent real mechanisms o f element substitution, they may be correlated to actual mechanisms o f exchange, therefore the T i M g [ F e ] . correlation may 3+  2  indicate that volume is controlled by octahedrally coordinated cation exchanges and possibly that T i substitution is controlled at this site. N o correlation was evident between cell volume and H 4 S i , however most values for the samples examined were zero for this exchange component.  113  Figure 30. Z95-1 2 theta C a F 2 correction factors. Solid line represents best-fit through the data. Vertical axis represents the difference between theoretical and measured values. The correction factor is equivalent to the vertical distance from each point to the best-fit line. Points that lie above the zero line have a positive correction whereas those below are negative.  114  to to  E o  L_ H—  fl- CM CD m o> fl-  jo  ro  h- CD  uo co  "to ^ o  g  ~o c CD X> 0) o _CD 0) (0  II  CD •o  « S  & l  E o  fl"  fl"  •fl-  -fl  fl-  CM -fl CM  O O O CN CM CN  OJ  Q  t_  -fl"  CM fl" C -MCO O O O IT) O CM Csi CM CN I flLO 00 o> a) cn  ro i_  X  flN-  CD 05 05 05  £  Ct  fl"fl  Zr  31,  CO  CD  CJ)  =5  CM CD CD  co Nco  o  5£  05 0 ) 0 1 T  "fl -fl"  CN •fl CM O O O CM CN CN  CJ CD  O CD O C  T>  0  J2 JJ CD  O _CD _Q CD  CO CD 3  § CD  CO CO CN fl" 00 CO  CD  |  CO fl"  00 CO •fl-  P  CD CN  Li. CD  CD  00  O  i  CD CO  CD  gg co cn  O N co N N  115 27  •  26  O  25  i  . b)  -  A  T  24  1735  1740 1745  1750  1755 1760  • i . i 1740 1745 1750 1755 1760  23 1735  Cel\dume  Cell\bLime 1.7  008  •  d)  i  '  i  i  16 15 -  A  T  -  1.4  1735  1740 1745  1750 1755 1760  1.3 1735 1740  ,  i  1750  i 1755 1760  CeBXAbtume  Cell\rfolume 0.16  1735 1740  i 1745  0.16  1745 1750 CellUAffne  1755 1760  1735 1740  1745 1750  1755  1760  CelNAjlume  Figure 31. Measured cell volumes for several samples from this study. Cell volumes are as in Table 7, whereas all other values are as in Table 4. Legend is as in Figure 26. a) Cell volumes vs. Ti02 wt.%; b) cell volumes vs. Fe203 wt.%; c) cell volumes vs. Fe2+ calculated formula unit; d) cell volumes vs. Fe3+ formula unit; e) cell volumes vs. TiSi- Thompson components; f) cell volumes vs. TiMgFe(3+)2-.  116 An Analysis of Site Occupancy in Titanian Andradite  7.1 Recapitulation of Andradite Crystals The general chemical formula for garnet is X Y (Z0 ) , 3  2  where A"=Ca, M g , F e , 2+  4 3  M n ; 7=A1, F e , M n , V , T i , C r ; and Z= Si. In titanian andradite, the Z site is silica2 4  3+  3 +  3 +  4 +  undersaturated which allows for substitution, or coupled substitutions, o f cations, such as T i and O H . The structure consists of alternating Z 0 tetrahedra and Y 0 octahedra which 4  6  share corners to form the three-dimensional framework (Figure 32). The X cation is coordinated by eight oxygens and forms a third kind o f polyhedron that is a distorted cube or a triangular dodecahedron. The polyhedra are distorted as the shared edges are not equal (shortened) to unshared edges because o f the radius o f the X ion and any substitution at theX-site will affect the whole garnet structure. 7.2 Literature Review of Site Occupancy in Titanian Andradite Titanian andradite occurs as a primary magmatic phase in the Z M P . Titanian garnets have been used to discuss the relationships between Si, F e , A l , and T i in zoned 3+  crystals. It is widely accepted (though still debated) that T i octahedral position replacing F e  3+  4 +  occurs mainly in the  and that the relative preference for the tetrahedral site  must be in the order Al>Fe>Ti (Huggins et al. 1977). The site occupancies o f O H and Ti  3 +  are unresolved. Table 8 shows the most widely accepted cation site substitutions for  titanian andradite. Isaacs (1968) states that the principle exchange mechanism i n andradite is T i  4 +  replacing Fe and that no relation o f Si with T i nor Fe is evident. Isaacs (1968) also hypothesizes the possibility o f T i substituting in valence states other than, or in addition  117  to, four but suggests that they would likely appear in chemical analyses as four only. It was previously believed that T i directly substituted for Si, but S i is not the only element being replaced by T i , as the cation imbalance caused by declining Si is better matched by F e and F e  2t  31  at this site than by T i  4 +  (Howie and Woolley 1968) or by combined F e  (Armbruster et al. 1998). Whether T i  the similarity of the bond distances F e  3 +  4 +  3+  or F e substitute at the tetrahedral site, 3+  - O and T i  4 +  -O, both o f which are larger than the  S i - 0 bond length, w i l l produce an increase o f the cell parameter and the cell volume (Virgo and Huckenholz 1974). Substitution o f Si occurs in decreasing preference by A l , F e , and T i 3+  4 +  (Hartman  1969; Huggins et al. 1977a); however at high temperatures, the tetrahedral site preference is argued by Schwartz et al. (1980) to be F e >A1 > T i . Hartman (1969) 3+  shows that T i  4 +  4+  takes on preferentially octahedral coordination and F e takes on 3+  tetrahedral coordination. The role of T i  4 +  has been debated since 1969 by various  authors. Manning and Harris (1970) state that Ti ' " prefers the tetrahedral S i sites which 4 1  was later confirmed by Weber et al. (1975b), Huggins et al. (1976b), Huggins et al. (1977b), Kuhberger et al. (1989), Dingwell and Brearley (1985). Some authors maintain that T i  4 +  substitutes at the octahedral site, thereby displacing F e  3+  and A l  3 +  to the  tetrahedral site (Moore and White 1971; Dowry 1971; Weber at el. 1975a; Amthauer et al. 1977; Gongbao and Baolei 1986; Peterson et al. 1995; Locock et al. 1995) whereas others state that T i  4 +  can be either tetrahedrally or octahedrally coordinated (Armbruster  et al. 1998; Waychunas 1987; Tarte et al. 1979; Huggins et al. 1976b) with A l preferring octahedral coordination (Armbruster et al. 1998). Yet another school o f thought calls for the existence o f T i  3 +  (Isaacs 1968; Dowty 1971; Huggins et al. 1976b; Amthauer et al.  .  118  1977; Gongbao and Baolei 1986; Kuhberger etal. 1989; Locock et al. 1995; Malitesta et al. 1995, Armbruster et al. 1998) which substitutes exclusively at the octahedral sites (Manning and Harris 1970; Dowty and Clark 1973; Basso et al. 1981; Onuki etal. 1982; Dingwell and Brearley 1985; Waychunas 1987; Kuhberger et al. 1989; de Groot et al. 1992; Locock et al. 1995; Malitesta et al. 1995; Merli et al. 1995) or, alternatively, substitutes exclusively at the tetrahedral site (Huggins et al. 1975). Huckenholz et al. (1976) state that natural garnet samples are characterised by Si+Ti > 3.0 per formula unit and therefore it is necessary to consider part of the iron and titanium as F e  2 +  and T i , 3 +  respectively. A s T i is always in excess relative to the calculated S i deficiencies, the presence o f other cations is necessary to ensure that charge is balanced. (1976b) neither confirm nor deny the existence o f T i  3 +  Huggins et al.  in andradite but state that it can  not be distinguished from F e . 2+  7.3 Results and Ideas from T h i s Study Though some o f the experimental attempts herein to solve or further our understanding o f titanian andradite site occupancies have proved unsuccessful, some conclusions and interpretations can be made. Microprobe analysis and the resulting calculated atoms per formula unit show, for the most part, that all o f the C a m u s t be 2+  assigned to the X site and that this site is then approximately 2.9/3.0 full. N a and M n , +  usually combined, add up to fill any vacancies on the X site. M n  2 +  substitution at the X  site would not cause structural distortions. If the X site is filled with both C a then the remaining M n  2 +  2 +  2 f  and N a , +  must be assigned to the Y site.  The presence o f both valences o f Fe is confirmed by wet-chemical analyses. and Mg* must be assigned to the 7 site, as is any excess M n . F e , Cr, V , T i +  2 +  4 +  Al  are also  3 4  119 assigned to this site. Most Fe * can also be assigned to this site, with generally 0.1 ions 3  over the 2.0 Y site total, which are then assigned to the Z site. However, with regards to the assignment o f F e  3+  and T i , it is their combined total that requires some reassignment 4 +  to the Z site, but it is unclear whether the excess is made up o f only one, or a combination o f the cations. Cell volume, as determined by powder X R D analyses, varies with T i 0 content and the titanium substitution mechanism may be via the T i M g [ F e ] . 3+  2  2  exchange component and in octahedral coordination, as opposed to substitution via the T i S i - component and tetrahedral coordination. Therefore T i  4 +  is assigned to the  octahedral site. The author has assigned any excess on the Y site as F e , but with 3+  acknowledgment that the excess may well be caused by T i  4 +  or by a combination o f these  cations. Totals for these cation site assignments are shown in Table 9. With excess Fe ", all Si, Z r and calculated H assigned to the Z site, most totals are 3  only slightly below 3.0, which may be accounted for by a slight underestimation o f the totals for either the iron or titanium valence, or by underestimation o f the total water content. The exceptions to this cation assignment are sample BN1-4 and the volcanic sample which are both over 3.0 total for the Z site, at 3.06 and 3.09, respectively. The volcanic sample total can be explained because of the higher than average total iron (wt%), thereby some o f the iron is likely in the wrong valence, which is evident in the Y site total 2.25/2.0. X R D analysis for the volcanic sample indicates a cubic cell and a small cell volume, which eliminates the possibility o f large amounts o f undetected water, as water substitution increases the size o f the tetrahedra. A s either T i  4 +  or F e  3 +  substitute on the tetrahedral site, the cations w i l l produce an increase o f the cell parameter and the cell volume (Virgo and Huckenholz 1974), therefore most o f the iron  120 is likely in the 2+ valence state, and all o f the f i is in octahedral coordination. The B N 1 - 4 Fsite total is slightly elevated at 2.09/2.0, however the main cause o f the high cation site totals lies with the Z site. Although none o f the samples tested in the 1R trials showed appreciable O H , the presence o f anisotropy in this sample is indicative of either large amounts o f undetectable O H , or cation ordering o f F e  31  and A l at the  octahedral sites. Cation ordering explains the elevated Y site total as there may be more Fe  3+  on this site than is allotted, which brings the Z site total down slightly. A lower Z  site cation total could accommodate the larger amounts o f water necessary to produce anisotropy. Natural garnet become more and more silicon-deficient with increasing  f  02  (Huckenholz et al. 1976) as is the case with B N l - 4 , thereby making even more room on the Z site for water substitution. N o cell volume is available for this sample, however the skarn sample tested with powder X R D showed a slight volume increase which may be indicative o f Z site water: Future Work Future work still to be performed on titanian andradite includes solving the site occupancies o f cations on the Y and Z sites, as well as determining the valence(s) o f titanium. The most likely way to solve for T i valence is with the use of X-ray absorption spectra (de Groot et al. 1992). The presence o f total water in natural titanian andradite. could be determined using either single crystal X-ray diffraction given the appropriate time scales, or thermogravimetric analysis in which the sample is heated on a balance and the resulting weight loss is equivalent to the amount o f water driven off. Electrolytic water determination has also been used to quantitatively measure the water content o f silicates (Wilkins and Sabine 1973). Once the presence and amount of water can be  12.1 established, the site occupancy then needs to be determined, most likely by the use o f single crystal diffraction or of Mossbauer spectroscopy (Gongbao and Baolei 1986).  Figure 32. Portion of the garnet structure projected down the c-axis (after Schwartz et al. 1980). Note the framework of alternating corner-shared Y 0 octahedra and Z 0 tetrahedra, and the chains of alternating edge-shared X 0 dodecahedra and Z 0 tetrahedra. 6  4  8  4  T a b l e 8: W i d e l y a c c e p t e d c a t i o n site o c c u p a n c i e s .  cation Si4+  dodecahedral  octahedral  Xsite  Ysite  Ti4+  2*  Ti3+  1  Zr4+  2  AI3+  1  Cr3+  Fe2+ Mn2+  e . g . - B a s s o et a l . 1981  2  A r m b r u s t e r et a l . 1 9 9 8  cannot  e g . B a s s o et a l . 1981  cannot 1  A r m b r u s t e r et a l . 1 9 9 8  Deer, Howie, Z u s s m a n 1997  H u g g i n s et a l . 1 9 7 7  Deer, Howie, Z u s s m a n 1997  1  L o c o c k et a l . 1 9 9 5  2  cannot  2  2  A r m b r u s t e r et a l . 1 9 9 8  2  A r m b r u s t e r et a l . 1 9 9 8  2 cannot  Ca2+  1  H+  Refetence  1*  cannot  Mg2+ Na+  Z site  2  V3+  Fe3+  tetrahedral  1  3*  2  Locock etal. 1995  1  Lager e t a l . 1989, Labotka 1995 A r m b r u s t e r et a l . 1 9 9 8  3  1 * = e l e m e n t must g o o n this site 2* = e l e m e n t s o m e t i m e s g o e s o n this site 3* = e l e m e n t c o u l d p o s s i b l y g o o n this site  1  L o c o c k et a l . 1 9 9 5 Amthauer + R o s s m a n 1998  124  oo  co co c\i  CO i - |vT - O CM o d d O  O  CD  d  d  T-  O O O  CM CN O CO O h- o o O O O  o  O O d  CM co  CD CO CN T— COCD CN CO CD O CD O O O CD O c i c i CN d d CO CO  o o oo co  CN -r- 0 0 CO CO CO T— T—  d d  i  -  CN  d  00  CD  o  O  CJ)  O  cn O o Oo  CN  xjT O-  o O  co  LO  t -  o o O  CN O  •MO co  r-~  CO CO o T - co  N  co co o o  od  co  d  oo o o CN  d  CO  o> O CSI  C O O T - K - O - t - T - O O C N h - T - C M C O T - O C N O O N - O T - O C O O O C M d d d d d T - ^ d d d c N ' d d  o o o oo o o o o CO cr> oo c cn co  ^  cd  CN  d CO CO T - T d  T -  CM  d  Oi  O 00  o c N i - c o o T - c o o L O L O L n T - L O  oqcNO^roo^rooocnoo  CN  d  d  d  d  d  -  T  '  d  d  d  c  N  o  d  N  OO N CD o o> LO LO LOo o co o o cd d •sr CD T CN- rOcd -r-^ r o CO  CO Oj  O  CO  N co  C D T - O O O C N ^ r c O C N ^ r C D T - O  o  c q c N O c o o o L o o o o c n o o  CN  d  d  d  d  d  -c^ d  d  C M i d  d  d  co  oo oo od d  CD  cr>  CN CO LO LO  n  in ^ ^ o o N  T -  CN  O CO  r -  O)  O  CO  N  v - c o o L n o - r - o o T - ' ^ - i - h - ' M * ' o o c o o o o o c o o o o o o o c N d d d d d - r - d d d c d d d  CD  CD  00  T-  CO  T-  co  T—  co  C N C O O T - O O ^ T O T - C D C D O O c o c N o c o o o o o o o c n o o CN O O O O O O O O CN O O  LO  LO CD  T - ^ J - i — r - C D O r O O  co Tco  CO O)  T-  oo cd  d  o ocn  T- cn d T- OJ  d  CN  O  Q -  ^  co ^)  CD ?  4  C M I - T C c  CT)i—  D C O O T - O O I D O O O N d d d d d t - ' d d d c  O d  O d  O d  o o coo co c  o  LO h- i T— cd  r- O CD O CN d CO CN T -  cn o r>- r- cn CN o CN d  T O  cd  co co  CO LO CD O CN 0 0 T - T cn d d T— CN d CN  T-  CO  o o>  CM O "O C  CD CO  c o  *C  s  CO c o  •o CD - Q  w  o  N  <  O  >  (0  -2 CO  U  0  .  o> c U . 2  N > - N > - > - > -  2  J ? co 2  O  2  co  aw I  + CN 0 U.  >-  N A II  0J S  CO • C  H  «_ 1. A  c  <D  co  0  ^  ±  A II  >-  5 0  ^  X  X  N  c  < + > t  £ CO O < / > « X  0  CO 0  0 X co >-  0 N  fl-  C O  CO CM  C O  m 0 0 T— 0 ci  d  d  0 0 co 0  d  T _  d  0 0  d  in 0 T— co 0 0 0 0 d co d d  00 0  CO  co CM O  0 co 0 0 0 0 d 0  0 fl0 0 co 0 0 T d d ~ d  fl- 0 0 0  co co 0 CN 0  0 0 0  A—  0  co co 0 00 CN 0 0 0 CN CD O CD d CN d d o co  CO CD CN  0 0  0 CO 0 0 0 co d 0  0  m m 0 fl0 cn 0 0 o  CD  x—  CJ)  CN CN  C O  IO  1^ 0 fl- 0  CN T—  0 0 d  O d  00 0 0 CO d  d  T—  d  d  CO  CN 0 0 0 0 0  CN 00 O  co  co  O  CN O O CO d  CD CO CO in fl- o o d — CN d fl"  co 0 co m 00 00 0 0 C N 00 o o CM d co d d CD 0 CD 0 CD fl- co co CM d CN N- O O •r-  QQ fl- 0 0 co 0 T—  d  d  d  d  d  CN  d  d  fl-  0 0 0 0  o o  0 0 cd  d  d  CN  CO 0 CD 0 T - 00 0) OS 1 ^ - co o o c\i d T- d CN d  O)  - •  co CN  0  CO  0  N  0 0 CO CN 0  IO  N  CO  CO  0 00 0 0 CN 0 d 0  0 00 0 0 0 0  co 0 0 T— 0 d ci d  co 0 0 m 0  •*—  T—  00 CN 0 Csi 0 d  i co  d  0 0 0  d  d  CO d  T _  d  d  0 O) 0 CO 0 d 0  m 0  CD CN  d  0  CN  m 0 0  0 0 r-  0 d  d  co  d  d  CN  0 0  0 d  d  co 0 0 00 0 0 d  CN  d  d  \—  o  r- 0 O CD CO co  • < 0 —  CO  co o  cn  m 0 CO  m 0 fl- CD CO CO 0 CO N - T - . CD CM d d CNi o| m CO N 0 CN CN CM OO O 0 d T - CN d cd  m 0 co CM in in 0 CD CM 0 CO CO CN CN O r ( N O  CO  1  O 0 O 0 0 NCN 0 CN 0 0 CN c i d d d  CO  10  O d  IO  in T—  0 0 0 d  d  m 0 d  N 00  co  T—  T—  ro 0 uS CN 0 d o> N r-  fl-  CN  ci  CQ CO 1  O)  co § N 00 o> O  T—  fl"  d  0 0 d  O  O  co 0 0 CO T —  co 0 CO, CO 0 CN 0 CO 0 1— 0 0 d d  CN  d  0 d  8  m 0  d  0 d  0 0 0 co d d CN  O 0 CO d  00 o co 0 CO cd  CO  0 d  T—  CO  cd  CO  0 co m 00 00 0 CO d CO N- O O o CN O m CN 0 v - fl- m m CO 0 d m co T - • O V  d  CO  0 d  CO CD CN  T—  O d  CM  0 d  CD  CO  0 cd  S  O d  N  OJ  0 0 0 0 0 0 0 0 cd d  CO CO  o  1 -  CN O  I  1  CD  00  N  <  0  A "CD CD c o) co ro * > i x U . 2 S O Z X  ' ions tol  ro  +CM  c  CO  CD ro CD — <  CO  X  co 2 ,<•>» CD  CD  +  co  -tS LL -t; co  -b  CD >- i - >- CD  CD  126 Conclusions  8.1 Conclusions Petrogenesis and Summary Nineteen samples containing titanian andradite were analysed from magmatic, volcanic, and hydrothermal environments. This thesis has contributed three things to further the understanding o f titanian andradite. Firstly, chemical compositions o f titanian andradite were determined for the different environments o f formation. Secondly, chemical zonation patterns unique to different environments o f formation were . established. Thirdly, the crystal chemistry o f titanian andradite was examined, specifically the role and substitution mechanism o f T i and, to a lesser extent, the role o f O H on the crystal chemistry. The results from the various tests performed on the wholerock samples from this study are summarised below and these results are applied to a petrogenesis for the sample suite. Geochemical results indicate that pyroxenite samples have undergone fractionation as evidenced by FeO enrichment, and sample Z M 3 9 B contains fewer incompatible trace elements than do other samples, indicative o f an early formation. O f petrographical significance is sample 97BN1-4, which is the only sample from this study that has anisotropic zones, twinning, and visible regular zoning in the titanian andradite. Chemical zoning is visible using B S E imaging. Magmatic titanian andradite samples show irregular zoning, the volcanic sample has irregularly zoned cores and regularly zoned rims, as does Z M 3 9 B . Titanian andradite from magmatic dykes shows  127 both regular and irregular zoning, skarns show only irregular zoning, with the exception o f 97BN1 -4 which is very regularly zoned but contains irregular zones inside o f the regular zones. E P M A data from titanian andradite reveals zoning patterns within different rock types and between different environments o f formation. Both the volcanic and cumulus samples have the same patterns, including: darker zones which have more T i 0 and 2  A 1 0 and less F e 0 than lighter zones; A 1 0 and C a O have the same zoning pattern, 2  3  2  3  2  3  opposite to that shared by T i 0 and F e 0 ; and the change in T i 0 is the same or twice 2  2  3  2  that o f F e 0 . In the syenite samples, A 1 0 changes by half that o f T i 0 , whereas the 2  3  2  3  2  melasyenite samples show that T i 0 change equal the changes o f F e 0 , and generally 2  2  3  shows either no zoning or irregular zoning. The pyroxenite samples show regular zoning with darker zones which have more A l 0 and less F e 0 than lighter areas, as well as 2  3  2  3  equal and opposite changes in T i 0 and F e 0 . The magmatic dykes show regular or 2  2  3  irregular zoning in which the darker areas contain more A 1 0 than lighter areas. C l i f f 2  3  samples share many characteristics with the early-formed magmatic samples, whereas other skarn samples are similar to later magmatic stages and show the absence o f zoning and have darker areas which contain more A 1 0 . 2  3  Thompson components provide definite ways to distinguish between titanian andradite formed in different environments. Magmatic samples have small norms, negative M g C a - , very small H 4 S i component, and positive FeMg-, whereas hydrothermal samples predominantly have positive H 4 S i , zero to slightly positive F e M g - , and larger norms. Magmatic dykes have been infiltrated by an externally derived fluid as they have zero or slightly positive F e M g - components, and many large H 4 S i values. C l i f f and other  128  skarn samples have near equal average amounts o f both T i S i - a n d T i M g [ F e ] . 3+  2  components, small norms, and are therefore intermediate between magmatic and hydrothermal samples, in terms o f fluid evolution.  97BN1-4 has the largest norm and is,  furthest in composition from a true andradite. Five different titanian andradite types identified on the basis o f zoning and Thompson components are as follows: magmatic " A " or primary magmatic samples which include Z M 3 9 B and W B 6 6 - 5 ; magmatic " B " which include most o f the magmatic samples (e.g. Z M 1 6 ) ; skarn " A " which can be further subdivided into two groups including C l i f f samples which show a magmatic affinity in chemical zoning, and most other skarn samples excluding B N 1 - 4 ; skarn " B " consists o f only the hydrothermal sample B N 1 - 4 ; and dyke samples which have a magmatic signature overprinted by skarn " A " fluids. Titanian andradite site occupancies as determined from the study samples are as follows: C a  2 +  and N a are always assigned to the X site; M n +  2 +  is preferentially in  dodecahedral coordination, but may also be in octahedral coordination; A l , M g , C r , 3 +  V , T i , and Fe " are always in octahedral coordination; F e 3 t  4 +  2  or tetrahedrally coordinated; and S i , Zr ", and H 4 +  4  4 +  3+  2 4  3+  may be either octahedrally  are always in tetrahedral  coordination. C e l l volume varies with T i 0 content and the titanium substitution 2  mechanism may be via the T i M g [ F e ] . exchange component, indicative o f octahedral T i 3+  2  substitution and octahedrally controlled cell volume. The petrogenesis o f the Zippa sample suite is as follows, based on the data collected from titanian andradite in this study. The volcanic and cumulus Z M 3 9 B samples were the earliest to form based on euhedral, resorbed grains, identical zoning  129 patterns, and low f  02  and a  Sl02  values which are correlated by the T i M g [ F e ] . and T i S i 3+  2  components, respectively. The magmatic samples formed next and fractionated by sorting from syenite through pyroxenite, and E P M A data indicate substitution and replacement between T i 0 , F e 0 , and A 1 0 . One o f the pyroxenite samples may have 2  2  3  2  3  originated from a liquid, as opposed to other magmatic samples which originated from a crystal mush. The pluton evolved to a higher f  Q2  of a fluid phase. Lowf  state in conjunction with the evolution  values for titanian andradite are indicative o f crystallisation at  Q2  high pressure, however H diffusion and sulphur loss are two mechanisms which can 2  result in pressure, and consequently/^,, changes (Virgo et al. 1976b). C l i f f samples interacted with early fluids driven off the pluton and inherited t h e /  Q2  signature from the  magmatic fluid, and share many zoning and T C S properties with magmatic samples. However the C l i f f samples also have an increased  c3  s i 0 2  as a result o f cooling of the  magmatic volatiles and interaction o f the fluid with quartzite. Glacier and BN15-1 have all the characteristics o f true skarn samples and likely formed after C l i f f and before 97BN1-4, the fluids o f which migrated back into the pluton probably along dykes which bear T C S evidence o f interaction with both the original magmatic fluid and with slightly more evolved skarn fluids. 97BN1-4 was likely the last to fornix based on petrographic evidence, and crystallised from a more evolved fluid, as it has the largest crystals with the most complex zoning patterns, is the furthest in composition from true andradite, shows higher f  Q2  and a  S i 0 2  than the magmatic or skarn samples based on T C S evidence,  and has fine zones with well defined edges, indicative o f long-lived interaction with a continuously or abruptly changing fluid.  130  References  A k i z u k i , Mizuhiko. 1984. Origin o f optical variations in grossular-andradite garnet. American Mineralogist, 69, pp. 328-338. A k i z u k i , M . , Nakai, H . , and Suzuki, T. 1984. Origin of iridescence in grandite garnet. American Mineralogist, 69, pp. 896-901. Allen, F . M . , and Buseck, P.R. 1988. X R D , FTIR, and T E M studies o f optically anisotropic grossular garnets. American Mineralogist, 73, pp. 568-584. Amthauer, G . , and Rossman, G.R. 1998. The hydrous component in andradite garnet. American Mineralogist, 83, pp. 835-840. Amthauer, G . , Annersten, H . , and Hafner, S.S. 1977. The Mossbauer spectrum o f Fe in titanium-bearing andradites. Physics and Chemistry o f Minerals, 1, pp. 399-413. 57  Armbruster, T. 1995. Structure refinement o f hydrous andradite, Ca Fe, M % A l ( S i O ) ( 0 H ) , , , from the Wessels mine, Kalahari manganese field, South Africa. European Journal o f Mineralogy, 7, pp. 12211225. 3  5 4  2 0  0 2 6  4  1 6 5  4  4  5  Armbruster, T., Birrer, J., Libowitzky, E . , and Beran, A . 1998. Crystal chemistry o f T i bearing andradites. European Journal of Mineralogy, 10, pp. 907-921. Badar, M . A . and A k i z u k i , M . 1997. Iridescent andradite garnet from the Sierra Madre Mountains, Sonora, Mexico. Neus Jahrb. Mineral. Monatsh., 12, pp. 529-539. Bascianno, Laurel. 1999. Personal communication. Basso, R., Giusta, A . D . , and Zefiro, L. 1981. A crystal chemical study o f a Ti-containing hydrogarnet. Neus Jahrb. Mineral. Monatsh., 5, pp. 230-236. Chase, A . B . , and Lefever, R . A . 1960. Birefringence of synthetic garnets. American Mineralogist, 45, pp. 1126-1129. Coulson, I.A., Russell, J.K., and Dipple, G . M . 1999. Origins of the Zippa Mountain Pluton: a Late Triassic, arc-derived, ultrapotassic magma from the Canadian Cordillera. Canadian Journal of Earth Sciences, 36:9, pp. 1415-1434. Cygan, R . T . and Lasaga, A . C . 1982. Crystal growth and the formation o f chemical zoning in garnets. Contributions to Mineralogy and Petrology, 79, pp. 187-200.  131 Deer, W . A . , Howie, R . A . , andZussman, J. 1997. Rock-forming Minerals, V o l . 1A Orthosilicates. The Geological Society, London. de Groot. F . M . , Figueiredo, M . O . , Basto, M . J . , Abbate, M . , Petersen, H , and Fuggle, J.C. 1992. 2p X-ray absorption o f titanium in minerals. Physics and Chemistry o f Minerals, 19, pp. 140-147. Dingwell, D . B . , and Brearley, M . 1985. Mineral chemistry o f igneous melanite garnets from analcite-bearing volcanic rocks, Alberta, Canada. Contributions to Mineralogy and Petrology, 90, pp. 29-35. Dowty, E . 1971. Crystal Chemistry of titanian and zirconian garnet: 1. Review and spectral studies. American Mineralogist, 56, pp. 1983-2009. Dowty, E . and Clark, J R . 1973. Crystal structure refinement and optical properties of a T i fassaite from the Allende meteorite. American Mineralogist, 58, pp. 230-242. 3 +  G a l i , S. 1983. Grandite garnet structures in connection with the growth mechanism. Z . Kristallogr., 163, pp. 43-52. Gomes, C. B . 1969. Electron microprobe analysis o f zoned melanites. American Mineralogist, 54, pp. 1654-1661. Gongbao, W . , and Baolei, M . 1986. The crystal chemistry and Mossbauer study o f schorlomite. Physics and Chemistry o f Minerals, 13, pp. 198-205. Guinier, Andre. 1952. X-ray crystallographic technology. Hilger and Watts Ltd., London. Pp. 330. Hartman, P. 1969. Can T i pp.366-369.  4 +  replace S i  4 +  in silicates? Mineralogical Magazine, 37:287,  Hatch, D M . and Griffin, D . T . 1989. Phase transitions in the grandite garnets. American Mineralogist, 74, pp. 151-159. Henmi, C , Kusachi, I., Henmi, K . 1995. Morimotoite, Ca3TiFe2+Si3012, a new titanian garnet from Fuka, Okayama Prefecture, Japan. Mineralogical Magazine, 59, pp. 115-120. Hickmott, D . D . , and Shimizu, N . 1990. Trace element zoning in garnet from the Kwoiek Area, British Columbia: disequilibrium partitioning during garnet growth? Contributions to Mineralogy and Petrology, 104, pp. 619-630. Hirai, H , Sueno, S., Nakazawa, H . 1982. A lamellar texture with chemical contrast in grandite garnet from Nevada. American Mineralogist, 67, pp. 1242-1247.  132 Holten, T., Jamtveit, B . , Meakin, P., Cortini, lyfc, Blundy, J., and Austrheim, H . 1997. Statistical characteristics and origin of oscillatory zoning in crystals. American Mineralogist, 82, pp. 596-606. Howie, R . A . and Woolley, A R . 1968. The role o f titanium and the effect o f T i 0 on the cell-size, refractive index, and specific gravity in the andradite-melaniteschorlomite series. Mineralogical Magazine, 3 6 : 2 8 2 , pp. 775-790. 2  Huckenholz, H . , Holzl, E . , Huggins, F.E., and Virgo, D . 1976. A reconnaissance study of the Ti-garnet stability field at defined oxygen fugacities. Carnegie Institution of Washington Yearbook, 7 5 , pp. 711-720. Huggins, F.E., Virgo, D . , H o l z l , E . , and Huckenholz, H.G. 1975. Distribution o f A l , F e ' , and T i between the octahedral and tetrahedral sites in garnets between C a A l S i 3 0 and C a F e T i , S i . 0 . Carnegie Institution o f Washington Yearbook, 7 4 , pp. 579-585. 3  4 +  3  2  12  3  2  4 1 6  1  5 8 4  1 2  Huggins, F.E., Rosenhauer, M . , and Virgo, D . 1976a. The intrinsic oxygen fugacities and the oxidation states o f iron and titanium in natural melanite and schorlomite garnets. Eos, Transactions, American Geophysical Union, 5 7 : 4 , p. 339. Huggins, F.E., Virgo, D . , and Huckenholz, H.G. 1976b. The crystal chemistry o f melanites and schorlomites. Carnegie Institution o f Washington Yearbook, 7 5 , pp. 705-711. Huggins, F.E., Virgo, D . , and Huckenholz, G. 1977a. Titanium-containing silicate garnets. I. The distribution o f A l , F e , and T i between octahedral and tetrahedral sites. American Mineralogist, 6 2 , pp. 475-490. 3 +  4 +  Huggins, F.E., Virgo, D . , and Huckenholz, G . 1977b. Titanium-containing silicate garnets. II. The crystal chemistry o f melanites and schorlomites. American Mineralogist, 6 2 , pp. 646-665: Isaacs, T. 1968. Titanium substitutions in andradites. Chemical Geology, 3, pp. 219-222. Ivanova, T.I., Shtukenberg, A G . , Punin, Y . O . , Frank-Kamenetskaya, O.V., and Sokolov, P.B. 1998. O n the complex zonality in grandite garnets and implications. Mineralogical Magazine, 6 2 ( 6 ) , pp. 857-868. Jamtveit, Bjorn. 1991. Oscillatory zonation patterns in hydrothermal grossular-andradite garnet: nonlinear dynamics in regions of immiscibility. American Mineralogist, 7 6 , pp. 1319-1327. Jamtveit, Bjorn. 1999, In Press. Crystal growth and intracrystalline zonation patterns in hydrothermal environments. In Growth, dissolution and pattern formation in geo-  133 systems. B . Jamtveit and P. Meahin (eds.), Kluwer Academic Publ., Holland. Jamtveit, B . , and Hervig, R.L. 1994. Constraints on transport and kinetics in hydrothermal systems from zoned garnet crystals. Science, 263, pp. 505-508. Jamtveit, B . , Wogelius, R . A . , and Fraser, D . G . 1993. Zonation patterns o f skarn garnets: records o f hydrothermal system evolution. Geology, 21, pp. 113-116. Jamtveit, B . , Ragnarsdottir, K . V . , and Wood, B.J. 1995. O n the origin of zoned grossularandradite garnets in hydrothermal systems. European Journal of Mineralogy, 7, pp. 1399-1410. Jaworski, B.J. and Dipple, G . M . 1996. Zippa Mountain wollastonite skarns, Iskut River map area. Geological Fieldwork 1995, Paper 1996-1. Keep, M . and Russell, J.K. 1992. Mesozoic alkaline rocks o f the Averill plutonic complex. Canadian Journal o f Earth Sciences, 29, pp. 2508-2520. Kerr, A . 1981. Zoning in garnets from the mainland Lewisian. Mineralogical Magazine, 44, pp. 191-194. Kingma, K . J . , and Downs, J.W. 1989. Crystal structure analysis o f a birefringent andradite. American Mineralogist, 74, pp. 1307-1316. Kuhberger, A . , Fehr, T., Huckenholz, H . G . , and Amthauer, G . 1989. Crystal chemistry o f a natural schorlomite and Ti-andradites synthesized at different oxygen fugacities. Physics and Chemistry o f Minerals, 16, pp. 734-740. Labotka, Theodore C. 1995. Evidence for immiscibility in Ti-rich garnet in a calc-silicate hornfels from northeastern Minnesota. American Mineralogist, 80, pp. 10261030. Lager, G . A . , Armbruster, T., Rotella, F.J'., and Rossman, G.R. 1989. O H substitution in garnets: X-ray and neutron diffraction, infrared, and geometric-modeiing studies. American Mineralogist, 74, pp. 840-851. L a m , Anita Eva. 1998. A n X-ray crystallographic study o f the crystal structures o f four tellurium oxysalt minerals: Dugganite, Choloalite, Rodalquilarite, and Graemite. Unpublished M . S c . thesis, University of British Columbia. Lang, J.R., Lueck, B . , Mortensen, J.K., Russell, J.K., Stanley, C . R . , and Thompson, J.F.H. 1995. Triassic - Jurassic silica-undersaturated and silica-saturated alkalic intrusions in the Cordillera o f British Columbia: implications for arc magmatism. Geology, 23:5, pp. 451-454.  134 Lessing, P. and Standish, R.P. 1973. Zoned garnet from Crested Butte, Colorado. American Mineralogist, 58, pp. 840-842. Locock, A . , Luth, R.W., Cavell, R . G . , Smith, D . G . W . , Duke, M . J . M . 1995. Spectroscopy of the cation distribution in the schorlomite species o f garnet. American Mineralogist, 80, pp. 27-38. Lueck, B . A . and Russell, J.K. 1994. Geology and origins o f the Zippa Mountain igneous complex with emphasis on the Zippa Mountain pluton, northern British Columbia. In Current Research 1994-A, Geological Survey o f Canada, pp. 77-85. Malitesta, C , Losito, I., Scordari, F., and Schingaro, E. 1995. X P S investigation of titanium in melanites from Monte Vulture (Italy). European Journal o f Mineralogy, 7, pp. 847-858. Manning, P.G., and Harris, D . C . 1970. Optical-absorption and electron-microprobe studies o f some high-Ti andradites. Canadian Mineralogist, 10, pp. 260-271. Matsyuk, S.S., Langer, K . , and Hosch, A . 1998. Hydroxyl defects i n garnets from mantle xenoliths in kimberlites o f the Siberian platform. Contributions to Mineralogy and Petrology, 132, pp. 163-179. Matsyuk, S.S., Langer, K . , and Hosch, A . 1998. Erratum to: Hydroxyl defects in garnets from mantle xenoliths i n kimberlites o f the Siberian platform. Contributions to Mineralogy and Petrology, 133, p. 418. M e r l i , M . , Callegari, A . , Cannillo, E . , Caucia, F., Leona, M . , Oberti, R., and Ungaretti, L. 1995. Crystal-chemical complexity in natural garnets: structural constraints on chemical variability. European Journal o f Mineralogy, 7, pp. 1239-1249. Moore, R . K . and White, W . B . 1971. Intervalence electron transfer effects in the spectra of the melanite garnets. American Mineralogist, 56, pp. 826-840. Murad, E . 1976. Zoned, birefringent garnets from Thera Island, Santorini Group (Aegean Sea). Mineralogical Magazine, 40, pp. 715-719. Nakano, T. and Ishikawa, Y . 1997. Chemical zoning o f pegmatite garnets from the Ishikawa and Yamanoo areas, northeastern Japan. Geochemical Journal, 31, pp. 105-118. Nakano, T., Takahara, H . , andNishida, N . 1989. Intracrystalline distribution of major elements in zoned garnet from skarn in the Chichibu M i n e , Central Japan: illustration by color-coded maps. Canadian Mineralogist, 27, pp. 499-507. N e i l l , I., and Russell, J.K. 1993. Mineralogy and chemistry o f the Rugged Mountain  135 pluton: a melanite-bearing intrusion. / ^ G e o l o g i c a l fieldwork 1992, Grant, B . and Newell, J. (eds.), British Columbia Ministry o f Energy, Mines, and Petroleum Resources Paper 1993-1, pp. 149-157. Novak, G . A . and Gibbs, G . V . 1971. The crystal chemistry of the silicate garnets. American Mineralogist, 56, pp. 791 -825. Onuki, H . , Akasaka, M . , Yoshida, T., and Nedachi, M . 1982. Ti-rich hydroandradites from the Sanbagawa metamorphic rocks o f the Shibukawa area, Central Japan. Contributions to Mineralogy and Petrology, 80, pp. 183-188. Peters, T.J. 1965. A water-bearing andradite from the Totalp serpentine. American Mineralogist, 50, pp. 1482-1486. Peterson, R . C , Locock, A . J . , and Luth, R . W . 1995. Positional disorder o f oxygen in garnet: the crystal-structure refinement o f schorlomite. The Canadian Mineralogist, 33, pp. 627-631. Rossman, G.R. and Aines, R . D . 1986. Spectroscopy of a birefringent grossular from Asbestos, Quebec, Canada. American Mineralogist, 71, pp. 779-780. Rossman, G.R. and Aines, R . D . 1991. The hydrous components in garnets: Grossularhydrogrossular. American Mineralogist, 76, pp. 1153-1164. Russell, J.K., Dipple, G . M . , Lang, J.R., and Lueck, B . 1999. Major element discrimination o f titanian andradite from magmatic and hydrothermal environments: A n example from the Canadian Cordillera. European Journal o f Mineralogy, 11:6, pp. 919-935. Schwartz, K . B . , Nolet, D . A . , and Burns, R . G . 1980. Mossbauer spectroscopy and crystal chemistry o f natural Fe - T i garnets. American Mineralogist, 65, pp. 142-153. Shore, M and Fowler, A D . 1996. Oscillatory zoning in minerals: a common phenomenon. The Canadian Mineralogist, 34, pp. 1111-1126. Smith, D . and Boyd, F.R. 1992. Compositional zonation in garnets in peridotite xenoliths. Contributions to Mineralogy and Petrology, 112, pp. 134-147. Tarte, P., Cahay, R., and Garcia, A . 1979 Infrared spectrum and structural role o f titanium in synthetic Ti-garnets. Physics and Chemistry o f Minerals, 4, pp. 55-63. Virgo, D . and Huckenholz, H . G . 1974. Physical properties o f synthetic titanium-bearing grandite garnets. Carnegie Institution o f Washington Yearbook, 73, pp. 426-433. Virgo, D . , Rosenhauer, M . , and Huggins, F.E. 1976a. Intrinsic oxygen fugacities o f  136  natural melanites and schorlomites and ^crystal-chemical implications. Carnegie Institution o f Washington Yearbook, 75, pp. 720-730. Virgo, D . , Huggins, F.E., and Rosenhauer, M . 1976b. Petrologic implications o f intrinsic oxygen fugacity measurements on titanium-containing silicate garnets. Carnegie Institution o f Washington Yearbook, 75, pp. 730-735. Vlasova, D . K . , Podlesskiy, K . V . , Kudrya, P.F., Boronikhin, V . A . , and Muravitskaya, G . N . 1985. Zoning in garnets from skarn deposits. International Geology Review, 27:4, pp. 465-482. Waychunas, G.A. 1987. Synchrotron radiation X A N E S spectroscopy o f T i in minerals: Effects o f T i bonding distances, T i valence, and site geometry on absorption edge structure. American Mineralogist, 72, pp. 89-101. Weber, H.P., Virgo, D . , and Huggins, F.E. 1975a. The role o f titanium in the crystal chemistry o f synthetic melanites and schorlomites. Eos, Transactions, American Geophysical Union, 56:6, p. 462. Weber, H.P., Virgo, D . , Huggins, F.E. 1975b. A neutron-diffraction and " F e Mossbauer study o f a synthetic Ti-rich garnet. Carnegie Institution o f Washington Yearbook, 74, pp. 575-579. Westphal, M . , Jaworski, B . , Dipple, G . M . , and Bingemer, A . 1999, In Press. The Isk wollastonite deposits o f Northwestern British Columbia. In: Industrial Minerals in Canada, CTM special volume. Wilkins, R . W . T . and Sabine, W. 1973. Water content of some nominally anhydrous silicates. American Mineralogist, 58, pp. 508-516.  137 Appendix A : E P M A E r r o r Treatment  Electron microprobe analyses o f titanian andradite were carried out as described in Chapter 4. Spectrometer configurations for elemental analysis were as follows: spectrometer one (crystal LIF) analysed for Fe, M n , C r , spectrometer two (crystal T A P ) analysed for M g , A l ; spectrometer three (crystal P E T ) analysed for Ca, T i , Zr, V ; and spectrometer four (crystal T A P ) analysed for Si, N a , Zr. Calibration standards for the different elements were as follows: albite standard S430 f o r N a , diopside standard S439 for M g , grossular standard S007 for A l and C a , almandine standard S467 for Si and Fe, rutile standard S442 for T i , vanadium metal standard S309 for V , chromite standard S382 standard for C r , rhodonite standard S459 for M n , and zircon standard S387 for Zr. The rutile, grossular, and almandine standards were analysed on the same spot on each crystal at the beginning and end o f each data collection sequence as shown in Table A l to monitor any changes caused by mechanical error within the probe during a collection run. A n y changes in the collection values on the calibration crystals were also monitored from day to day as a check for proper calibration and crystal setup before running unknowns, as the Ix/Istd should ideally always be 1.000. Three experiments were run on an andradite non-calibration standard (SO 17) (Novak and Gibbs 1971). The published formula o f this crystal is [{Caj^MgoojMnoo^tFe, A l 99  00]  ] ( S i ) O , ] (Novak and Gibbs 1971) and the recalculated 3  2  composition is outlined in Table A2. First, the non-calibration standard analyses were recorded through time from which analytical uncertainty was determined. Second, a grid on and around the area analysed on the non-calibration standard was analysed to  138  determine whether the grain itself was zoned. Third, the same five points were analysed at the beginning, middle, and end o f each collection sequence to show any machine variation within a collection sequence and over time. In the first experiment, the andradite non-calibration standard was used to monitor reproducibilty of the machine from day to day, from which analytical uncertainty was estimated.  A n example o f the analyses o f the andradite non-calibration  standard are shown in Table A 3 (see cd-rom in Appendix C for complete table). A l l analyses are plotted against time in Figure A l which show minor, random fluctuations about the mean. The first fifty one points show larger fluctuations about the mean as they were not analysed on exactly the same part o f the crystal as the remaining analyses so were not used to describe analytical uncertainty, as described below. Second, a grid on the non-calibration standard was used to determine whether the grain itself was zoned. The non-calibration andradite was analysed over a small area in this study (Figure A 2 ) and was found to be weakly zoned with respect to C a , A l , Fe, and Si (Figure A 3 , Table A 4 . See cd-rom in Appendix C for complete table). Therefore no direct comparison can be made between the reported Novak and Gibbs (1971) results and those listed herein without knowledge o f the exact location o f their analyses on a zoned crystal. Third, machine variation within a collection sequence and over time was determined. After the first fifty one analyses, five points in row three o f Figure A 2 were each analysed at the beginning, middle, and end o f each run (Table A 5 ) to establish consistency o f analytical technique, as well as to establish the exact analytical uncertainty as shown in Table A 6 . The first 51 points analysed on the andradite standard  139 were not included in the calculation o f the analytical uncertainty as these points were not on exactly the same five points analysed the remainder o f the time, and therefore would contribute sample variance to the probe machine variance because o f the weakly zoned nature o f the crystal. Plots o f oxides for each o f the five repeatedly analysed points are shown in Figures A 4 - A 8 which show that almost all apparent variations in elemental percentages are within analytical uncertainty and are therefore random. The large 26 values associated with A 1 0 and F e 0 (approximately ± 1 wt.%) are possibly a result o f 2  3  2  3  compositional variations in the non-calibration standard at a scale smaller than the probe beam size, and therefore these calculated 2 thetas may contribute sample variance to the probe machine variance. Ideally, the values listed by Novak and Gibbs (1971) should correspond to those analysed in this study, however this study analysed and re-analysed only a small part o f the crystal whereas Novak and Gibbs (1971) do not describe the location, the number o f analyses done, nor their analytical conditions or setup procedures, therefore no direct comparison can be made. The andradite monitor established that all analyses o f unknowns, completed during the whole analysis period, were performed under identical probe conditions, as the andradite standard results did not vary outside o f analytical uncertainty.  140  104  Total  1 0 0  96 36 C a o  35 34 20  Fe203  15 10 16  AI203  1 2  8 08 T . 0 2 04 0  40  80  120  160  Analysis number  Figure A l : Non-calibration standard plotted as oxide concentration against number of analysis (increasing over time). The mean for each set of oxide analyses is represented by a solid horizontal line. Vertical lines demarck analyses collected in a single day.  141  |^  Garnet  Figure A 2 : Grid of analyses on non-calibration andradite. Solid lines represent cracks in the crystal. Points Andr 1-5 were measured on row 3 (end analysis 31). Row-end numbers correspond to analysis numbers of Table A4. Individual analyses are spaced at five microns, both horizontally and vertically.  142  2 e)  row 5 -j  •Fe •Al  whole box  0  0  2  2 ^ ' d)  row 4  row 9 1 ~  •  -  ^  i  0  *  2  rcw 3  row 8 1 *-•  o  0  ^  2  2 b-  row 2  row 7  1  0 fc^  i  L  2  «0|  row 1 -j  row 6  0 0  1  Analytical uncertainty  2  1  o  0  1  2  Analytical uncertainty  Figure A3: Minimum analytical uncertainty for each oxide is plotted against the IS variation for analyses of the non-calibration andradite collected by rows (Fig.A2). All points are as labelled in j). Elements clustered near (0,0) include Ti, Zr, V, Cr, Mn, Mg, Na. The 1:1 line indicates estimated^analytical uncertainty. Points above the 1:1 line indicate oxides with estimated <analytical uncertainty.  143  102  Total oxides 100 98  Zr02  Ti02  40  16  S i 0 2 38  Fe203 12  36  10 20 Analysis  30  10 20 Analysis  30  Figure A4: Oxides, collected thrice per probe session at Andr point 1 (row 3, Fig.A2), are plotted against analysis number. The solid horizontal line is the mean for each set of oxide analyses, and the error bar at the end of each of the lines represents analytical uncertainty.  144  102  Total 100 oxides  AI203  98  Zr02  CaO  Ti02  MnO  40  16  S i Q 2 38  Fe203  12  36 10  20  Analysis  30  10  20  Analysis  30  Figure A5: Oxides, collected thrice per probe session at Andr point 2 (row 3, Fig.A2), are plotted against analysis number. The solid horizontal line is the mean for each set of oxide analyses, and the error bar at the end of each of the lines represents analytical uncertainty.  145  16  AI203  102  Total 100 oxides  u  12  98 36  Zr02  CaO  35 34  Ti02  MnO  40  16  S i 0 2 38  Fe203 12  36  8 0  10 20 Analysis  30  10 20 Analysis  30  Figure A 6 : Oxides, collected thrice per probe session at A n d r point 3 (row 3, Fig.A2), are plotted against analysis number. The solid horizontal line is the mean for each set of oxide analyses, and the error bar at the end o f each o f the lines represents analytical uncertainty.  146  14  102  AI203 12  Total 100  oxides  98  10  36  Zr02  C a O 35 34  MnO  Ti02  a  8  0.6 16  *Si02  Fe203  12 0  10 20 Analysis  30  0  10 20 Analysis  30  Figure Al: Oxides, collected thrice per probe session at A n d r point 4 (row 3, F i g . A 2 ) , are plotted against analysis number. The solid horizontal line is the mean for each set o f oxide analyses, and the error bar at the end o f each o f the lines represents analytical uncertainty.  147  0  10  20  Analysis  30  0  10  Analysis  20  30  Figure A 8 : Oxides, collected thrice per probe session at A n d r point 5 (row 3, F i g . A 2 ) , are plotted against analysis number. The solid horizontal line is the mean for each set o f oxide analyses, and the error bar at the end o f each o f the lines represents analytical uncertainty.  148  XJ-  o d  1-5 ^ TJTJ-  in  ^ CO  ?  O  3  CM  o o  N CO  TJTJ- T <0  o  CO o o o o o o o o o o o o o o o d d d d d d d d d  CO  o  00  o  in  o  o  o  CO CO  o o o T— o o o o CoN o o o o o o o o o  in  Tj-  lO  CO  o  CN  o  _  00  3  •7  co  •A  N-  ^  <o  CO  CN  d d co  CN  7 J 8  w CO O) CN  o  o  T—  ( O r-- o o r-- o o c o r-- o o  TJo  (0  16 <*  co  o m x- o  CO  ^— o \— o CO o o CM o Tj; o o o o o o o o o d o d d d d d d d d d  o  CN \J-  CM  TJ-. x -  CN  «>  O) CN  00 T— c n CO o o o c o o o o CO o o o o o d d d d d d d d d m  CO CO CO  to  o  CO  00  o  o  CN  •? 5 » CN  o d  ^— CO CO CO  K. •<*CO  O CN x-  TJ-  co  -Q Jv CO  CO  Th  CO  CO C o 1--' x-  •3  XJ-  <o  CN  CO  CN  ]3  d  o  xjxj-  CO  o  o  o  K  K t o CO x xj. T-  IN  06  o  O  co  00  CO  CO  TJ-  K-  CO  C  CO  ?  XJ-  00  CN CO  N-  CN CN  CO  CO  O O  O x-  CO  h - CO CN T- O O  00 CO  CO  CN  0  d  0 0  CN  a>  0  0 0  x~  d  10  T— i n  CN  CN  CO  CN  cri  T  m  in  00  m  cri  TJ" r CN i n r -  o o o  CN  cri ro  CM  CO  CO  CO CM  CM  CN CO  co m  cri  CN  co  ~  0  f—  cri  CM  CD  00 CO  T—  00  1^ cri T—  CN  Tf TJ- 0  00  00  CO CD  CD  0  0 0  d  en a> co o 00 o co o d -r^  in 0  XT  d 0 0  0 0 c r i T—  00  CN  0  0  0  CN CN  0 0  T—  d  rI--  00  xj-  CO  0 0  00  T—  co  T—  d  CO  TJ-  CO  oi  r0  CN  CO  7—  Tf  CO  CN  0  CN  <D  in 0  CM  CO  CO  CN  CM  d d co  T— 00  cri  CN  T—  CO  r-  0  d CN  0  d 0 0  T—  d  0 0  in CO  CO CM  in  CN  CN  d d  00  CN  CO  T—  d  TJ- TJ- O 0 CD  CN  CO  cri  x-"  CN  00 r^CD m 00 00  CO CT)  d d  CN  CO  CT) xjm xj00 h CT) CT) d d  CO  a>  CO 0 0  m o cn i n co co co d d  0  CO  GO  xj-  cri T— d  cri  0  m  d  CO  O  o o o '  CO  CN  CN  T-  x-  -~ cn d rCM cn 0 \— d  CM  d d 0  d  CO CN  CN CN  xr 00 CD CO CO CO  0  d d  CM  cri  CM  CN  CN  CM d d Ccri N  CO  CO  CN  CO  CN  CN  T— 0  CM  O  CO  X — TJ-  CN xj-  CO CO CN  06 d m  x- t— 0  O  1  CO  ?  CM  CN  CO  CN CN  O O O  C  00  CN  d d  o  CO  m  CO  o  o 00 o o o o o o o o d CO d d d d d d d d d  O  0  ,00 a> o 3 ?  CO  CO  O  m  CN CM  T?  T—  o  Cvl  TT O T— O  o O d d  O  0 0  CN  m CO o C N o CO o o o o CO o o o o o o o d d d d d d d d d d d o  xj- xj- o o 00 C N o CO o o o o o o o CCON o o o d CT) d d d d d d d d d  0  d d CN d d  00 o (0 xj- o T— T— o o o o o o o o o d d d d d d d d d  CO  t--  t—  CO  SI *  X-  CO  CO  o o CO CO o o o o 00 o o o i — o o o o o o d cri d d d d d d d d d  0 0  d d  O  O  CN CM  TCN  CN CN  o  "  5 CM ^> TJ-  *—  r— co o co o o  3 N. - 5 CO  o o o CO o o O o o o c~ O d CCT> d d d d d d d d d  I CN CN  0  d d  —  o in T- o d d  co CO TJ(0  CO  xt  CN  o  T— o o o o CN o O o o o o o o o O d o d d d d d d d d d  ^t- d <o  ^ ^>  CN  d d  CO l O TJ- CN CO  00 o  ^—  t—  o  00  h> ^  O  d d  in  O  CN  CN  CN  co  00 o CD x— CO CO o o o o cdo o o o o o o o o d o d d d d d d d d d CO T—  CN TJTjCO  I3  <D  o  CO  lO  3  1 -  1—  CO  1-3  o o o o o o CoD o CO o oo o o o d CO d d d d d d d d d CT)  00  00 o m  o d d  o  x— o d o o o CN o o o o o d o d d d d d d d d d  m  <0  CO CO  co CM  s CN  CM CO  0  CO  x-  T—  10 CO  co cn CO co CO d d xf CD  CN  O T—  00 C O CT) CO d d co co  m  o  CT) O CT)  CD CN CT) CO CO T-  o  |-> 00 o O) 00 CO CO  d d CN  0  d  CO  O O  TCN  T-  o o  «*>  co  <u CO CN  O  -I  CN  CN  2 z CO i- < N O  O  CO  O O D! < O  CO n  CN  >  oooS c 5  O ) co CO 5  O  2  <o co Q J  0 0 0 8 8 0 8 0 0 0 8  O O g < J c M < N ^ ^ c o ) r a r o Z i 0 h N < O > u . 5 5 O Z  CO  -a <n w "5 •D  149  5 ^  CN CD  I D N O S O C M C O l f l ^ N r N t o * t c o o ( o m c » o a i d d rCNd d d d d i r i o ' co  io co o co  CD O O O O CO CN  o o to 3 f-~  O l C O O W N N N C M W n c N ( o ^ o m N c o o s m o o  O O O CD O CO  lO i - o • <O) N CD "O V f ) CO TJ- o co o co m o o C O O O T - ' d o O O c i c D CN CO CO CO  co cn T- CO o cn  T CO  o o> CO  N- O o cn o cn •r^  ? o o o w  c o o o o o o o o o  CO  CO  CO CO  o  CN  l i O i - o n c o o i N o n i M i C O ' C f O O l N N O M n C D O  3 N.  cdoo'oooodocdo  CO  CO  CM  O i - L O O i - C O f l - C O C O C O O  ^ co i n t o n n N O M r n o o K N~ • O o>c d o o v ^ o o o o o c c i o CN CO _ o M CO O • co n i n f f l O I N r - O N N N t C O CO  o to  O  O  T - O CN  O  O  O  O  C O O CO  N m c o c o o o o o o o o o ^ t c o  7  o <o  C O CO  o CO  C  CO  ^  O  O  O  O l D N t O N iocoLoot--L T - o ^ c i o o CN  CN  0)  to  CO  TCN  fl- fl-  co cn o  •<- flCN CO  O O  O O  CO LO CD CN  cn o cn o Id ^ IO LO O) CO  C O O CO  o cn o cn  O l O otoo t d o CO  CO CO •r- LO  d  CO  O)O f l - C O C N C O C N C O i CO CO  3  cd d O 1  T - O CN  o o c o o o w n o ^ N o 4 o t ( M m o ( o i o o ) o c j j o o ^ o ' ^ o o o i o o  s J  O  N O O n r--fl-o c o o o CO  1h>  O  fl- co  CO  5  CM <N  co o o o  r O t O O ) C O O I < t O N  35  d  o O T - o o o cd d CO  i-  o  o o IO •>O CO T- CO  o cn  co cn ° ' o o o  o i ^ - n s s T - t N i n m o qtqrvnioofflm^o h>  §  to  CM  CN  co O co O co  ? o N-  aidoV dd CNd r - ' d d d c CO CO T-  O 00 i hco in o d •>- o  CN CO CO  fl- o o d  CN  CN  CO CN CN LO LO O CO CO  C N C N O C N O C N h - l O C O L O r C N f l - O f l - C O L O O C D l O L O O O T- O i CO CN CO  CO T cn in cn o a> o rr~o o  co in o o  O  to  =• cs cu 0)  CN CN CN O O o O O O O  ca o D J ZQ Q Q ^ O ^ I 5  2  |OCN  < PI TJ TJ <n u>  150  CO CM CO •>- t o LO o O O  CO CM 1 - O CM O o o  d d ci ci ci d d  T-  T-  o  o  d d d d  CO CN L O o X—  % ,5 ^ co CN  LO CO CM CO CD O O O  o o CD O  O  O  LO CM O O O  o O  O  O O  T—  55  i N - CO CO X}- ^— CM W  o  t— co  o  i —  c o fl- o o ea o  o o  CM  o  o o  ,—  o  3 h~  ^  o o  N. ^ CM W  i  ^  £ - NCM Co  00  O N. CM  CO o  00  o  oo CM o  o  o  ^  V  •StCO  r~  CN  1 -5 <M l~> fl- C O •flCO  '  a <0  . flT CM CO ^ co  co  r?  CO  co d  to  o o o o T— CN CO 0 0 o CO o o o o o o o <0 — CD o o CM CO o d CD d d d d d d d d d CD  fl- •»CO  ,3  CM  O flv  l  CN CD o CO o o CN T — o o o O o o o o  o o CN o o o  d d co o d d d d d d d d d T—  to CD o o h- o co  co  CO K. fl" CM to  -? O)  LO  o  7 ^ co co co LO N - -ttto  3  « "3;  o CO T— o o CN o o o o o o  co I— LO CO CN fl- CN CO  CD  CD  C  -9 v-  o flea CD o CO  o o  o o  o o  o o  o o  o  CD  CN oo o o CD CN o CM o o o o o o o o o o d d d d d d d d d d d  3Sd * "25  CO O  co CM  CO o o o  d CD d d d d d d d d d  N.  N - CM to  O  T—  o  o CO CN T —  00  co  i —  T—  o  fl  oo CD OO O  00  o CM LO CM CO LO o CM 1 — o o CO d d CM d d CD CN CD  3  ? -»-  T-  <o  CD CD CD CD O  co < r CO CN CD O CD O  co o  d  00  CN O  d  M J I CO f N - CN to  LO  CN CN CD CD CD CD  o  o CD o  d  r»  CM O  d  CO CO CD CD CD T CD O O  fl-  CM  CN CO  T—  T—  o d  LO CN CO CD CM o o o CO CM •*— o o o d d CN d d CD CM CD T— d  T—  fl-  O O  CO CO O O  1^  O O  CO CD CD ,  o d  LO CD CD CD 00 O CD O  CD CN CN CN CO o flo o o flo CM CN O T— CN CD CM CD T ' d d d d d  CN CO LO OO CD CD CD  CM CO o o o  o 1— o LO CN d d CD CM CD  fl-  CM  Id d CO CO oo o o CO d d CM  o CM d d  CD CD T— O CD CM CD  CD CM  LO T —  CO  1-^  o  o  T—  co  00  CO  o T — co LO o CM o o o d d CM d d CD c\i d T— d  T— fl-  CD CM 0 0 O o fl-  CN CN flCN o  d o c\i d d CD CN CD  o o o CM CO T—  LO  T—  o  00  CN  fl-  d d CM d d CD CN d  o o o  d  CO CM CD flLO  CD T — O  CO  o  T— T— r - o o CO o o CM d d CN d d CD CM CD d  so  •o  3  fl-  d d CM d d CD CN CD  CO N fl- CN co  CM CO d cs> d d d d d d d d d  3  N. Co flea  co  CD  o o o T— o T— o o o o o o o d d d d d d d  CD CN  00  3  r - •>-  d  o  o flO 00 o o CN o o o CO o o o o> o o o o o o CD d d d d d d d d d d  CO LO foo o o CD d d CD  T—  d o CM d d CD CN CD t—  o CM  o  CO o o fl- o o o o CD flo o o o o o o o o o d CD d d d d d d d d d  flo  T— co CD flCD CD CO X— LO O o CN d d d d CD CN CD T —  CD o o  so  i-9 ^  CO  CO CN o CO T— CD LO CD CD o O CN o O CO o o o fld d CN d d CD CM CD d  to  o o o o o o o o o o o d d d d d d d d d d d  OO CD  CN  CO CD  -<-  CD CD CM  Id  o  CD o o o o CO O CO o ea CO O o o o o o o o o O) d CD d d d d d d d d d  CD 0 0  ^—  fl-  3 CM  CD 0 0  CD O o fl- o o T— d d CM d d CD CM CD d  00  CM  ca CM  «> fle^a  3 < 9 oN . i CO  d d o d d d d d d d d  CO LO CO CD o o o o o ,— o CO o O o o o o o o o d d d d d d d d d d d  ^  to  f—  o o o o o  CO CM  CM  ^j-  00  CO  h-  CN O  d  CO CD CD  d  O O T-  CN 00  LO  CD o IO ^ CN •<rCD O O O O  o  o  d  c o  o CO I t  0) 0)  CN CN CM  ra ra O O O O Q _ l Z CO i- N  SgoSoo  « 55 CO CO  O  -  ra co o D J Z |W j - N <  co  S  O > li-  TJ TJ  ra ro § O  cn CM CN O  o  O o o  CO  T-  xr CM CO co CN m  CO CN CO m 00 T f o CO CO ob d d T— d CM CO T—  CO  od d |CN  CO  3  N.  m o  CN  CO d CO  co O CN  CN CN CD T  od d O (o  d  CO  ~? co°  I CM  o o  cri d  •5 N  d  1—  CN  r-o d  CM T T CD m  oo co  d  -5  CD T J N . \»- od d O t O CN CO CO  N.  2 § "  d  d  CO d CO  CN o CD d CO  in CN i ^ in CM CM 00 o CO in CN o x— d d d d d CO d CN CO  d  CN  d  CN  d  d  d  P  o o oo d  00 d d  CD  d  CN CO CO CO CM CO o o CN CO 00 o d T— d d d d CN CN CO  o o  I  d  d  00 00 d  O  a  d  CO T — CO o  o O h- CD CD CO T f o in CN r-- o f-- m  CO  ^  ? CO  d  CD T f CO m  o ho CN  CN CO  w  o o o  CO CN CN CO CO 00  CN CO CN CN T— m CM  CD  o CO o o o CN O in o d d d d d d d d  o T— o ro d  CN T— o o X}- o CD d d CO  T— CN  OO CN  d  o m 00 o d  d  CO CM T— CN CO CO CO o CM  d  d  d  d  in d  o m d  CN CO o CO d CO  TT T— tn o CD d  co  CO CO  o  CD d CO  CO CO o o CD m CO o d d CO d CO  in d  lO  d  o o  CD d CO  CO  o TT co xjT t o cd d CO  (0  1-9 K  o  d  T— CO CO CO  CN  d  00 o d d  in  CO CO CO o Tf CO CN CD T f 00 o CO d d T — d d d CN CO CN  00 •<!- CM T — CD  m  CN  d  d  CO d  •*—  o m  d  d  o  co  o  CN o CO* d CO  CO  •5° O CN O CN CO  1  CD T t CO d CO  Iv.  cO  co  S 2  o o d  00 CO o  7 ° 4 S 3  CN o CN CO d CO  CN  o d  CO m CN T J - o CD d d CO  CD 00 CO CN o CN CM oo T— d d d d CN  CD T f CN CO CO 00 CN  d  d  in d  1^- m TJ; t— CN  CD CO 1*-  d  d  d  CO  d  d  CO T f  m co o d  CO d CO  CO CO CM in CO o d CO d CO  m d  o o o o CO d CO  "8*8  c N C N C N o O X U o n n y CO h- N < O > u. 5 "2 O  Z  LO CO T - CO CO O T - O  T—  o d d  d  o cn co cn o o d d  CO CN CN CO CN CO CO CN 1 CO i - T - fCN d o CT> CN d d CN CN  T—  CN •<- CO CO CO fl" O O T- O  CO  T—  o d d d o d d  A-  o d d  "c— CO CN O  o d  O O CO H-  N  CN <  O  o CN  >  Total Ix/lstd Ix/lstd  Q  CN CN CN  CO  Fe203 MnO MgO CaO Na20  CO  co co O O  153  E o  o o m Tm o co  co  o o CO  If) t -  CD  T-  xf (D T" T -  O 00 CM  o o d o  £ "to O CD O  <D  k_  CO CD  T- CD  C h- ooo CD in oM m o oo o oo CD  CD  c 3  co  o in  c o  ^*-» CD  O  L_ .Q  is c o c  c CO o x:  T-  I  o w «cp-  %°  > T3 T3 C CD CO  ™ -£ 3 5  o o CD ; O X)  CD I-  o  Io  O  O  CD O  T-  CN  C\i  T-  T-  T-  O ICO CD oo o CD CD CD CO o d d d CD CD CD o in m 00  »^  CO  O O  o o o o o o  c ,o  .•2 3 -fe £  O  42  6  CO o o CM o o o CO o o CD c\i d d d  o  CO  CM  I c  re > o  c +  E  CO 5c 2co Oco <D  UJ  O  CO  <  LL  154  "fl" fl-  c •  c  CN CN  CO O)  cb co  d  CD  CD  CO  co ci  LO  co  CO  g? SI § 1  LO  =3 CN CN  c co  •fl-  CO  CO  SI §  CD >  o  c  CN  CO C CO  .*—'  c  c o  00  d  co  CO  CN  o d  K  c co  ^3  c  CD  •fl-  CD  d  00  00 CO  CN  CO  T— X—  d  d  co c i co  o  CO  CO  O oo CO  fl-  o  CO  ci  d  LO  LO  o  CN  cd  o o d d  CO d  CN  CO  d  CO T—  CN d  d  co  o CO  CD  co fl-  oo ^r co oo  d  co  CN CO  LO  CN O d  o d  o o d  o d  o  CD CD  T—  o d  o 00  cn  CD CD  CO  00  CD CD  •fl"  CO  d  CD CO  CN  o d  CD O d  O  CO d  d  o  r--  LO  co  d  d  CO  x—  o  CO  CD  T— _  d  CO CO  d  d  o o  o  LO  co o co d  CN  co  o o  LO  d  d  LO  LO  co o  fl-  CO r— T—  co co T—  T—  d  co o o ^—  d  CD fl-  C co  CO  L_  o d  cn CN  5 d  co co  d  fl-  CN  fl"'  o d  CD  co  O  CN d  o  o  CO  — I  0  3  c  O Z  LJ  O i_ CD  fl-  T—  00 LO CD CD d  CN d  CN O  CN CD  •fl" 00  d  CD CD  o o  CN  CD  o  d  d  LO  •fl-  d  CN d  LO CO  o d  aT c Lc o co E  LO d O  \—  o o o o d  d  CD LO T—  fld  o co d  CO LO CO  o o d  fl" fl"  CD CD  d  r— CN  o  o d  o d  LO  oo  CD  CD d  CN CN d  o CD  co  T—  o d  CO  O o d  co c 0 c E  £  LO  1 ^  d  o o  00  CN  d  o o d  flo d  CD CN  CO  CN CN  LO  d  d  CO  LO LO  LO  CN  CD CO  d  d  \—  T—  o o o h- 00 CD d  d  LO  d  OO CN d  flLO  co CO CO  o o d  CD  cri CD  o hd  00  CD  CO co O Q _J 2  CO  "O  81  C CO  «  c I- ° O TO  z h 2 ^ .9 o  o c Q- o ^ c  LO  M_  LP  CO  o  o  £ £  CO  0) CD  T>  O "D  3  ts JD LU co <  CO  s  o 1 ^ co  3  n o  0 E co  0 T3  co CO CD  0 > 0 o  CDco  CO  _>. I  0 E  CO  c  CO  L_  o  co  CO CN  d  CD  O)  3  CD co  E c o  CN  CD d  o  N-  CO CD SZ  o  CD CN  d  LO  CO  3  c  c  d  CD  d  fl-  0)  CO  o o  o  CN CD  CO  ja  •a  o d  d  d  k_  T5  x—  co o  CO  CO  CO  "co o i c o c  cn  CO i —  5 co -  T3  co  o o  —  o  d  d  "7 "oX— co  CN CN  d  o o  •fl-  ~a i  o  CO  co  11  fl-  d  CO  CN CN CN  CO d  CN  o o  CD CO  0 E  CN  o oo CO  j-  N  0  "0  CNI  >  <  ooo  9,2  0 £ j ? co co o S S O Z I-  LL.  155  o CD o CD co o co O O  00  CO CO  O  CM CO CO CN CO CO  o o  CO  o co oo co oo o ' i n O O xj- o o CO  c CO  to T -xO  r»- co o N- o  X—•  O ' O N N O r - CO 00 CD O  O h-  D s d d r i d d i N d d t CO d o iC  C CO  CO xj-  -i o 1c2 °> co CO  o  o m o CD  xj-  d  d  co c CO  00  Tf o  T-  CN  CD  cn  x -  -n co C co CO 00 CO  CO CO  g co CO  o  O  d  •<-  T T  O  d  o o d  CO  00 00 xj- |^ O  d  O  CM  o xj- o  CO CO CD  O  CD  CM CD 00 CM m N- x— o c) oi n s o N Tf O C D d O O 00 C D  o o T  o  CO  CD  r*-~  CM  00  o  o o  •c-  •<- i n n ^ w o o CD co CO T - o CM o o m o o o co o  O co o O co o O 00 o  T - inxj-Tj-cMocD o o c o c o o o o o CM o o m o o xco o  x—  o o  -vr T -  CO CD CD CM  O  O)  C  N - C M O O O O O O C O C O C D O i n CO  x-  x—  CO  CD  o o o - v r o o c D o o c o o o c o d d c d d d c M d d m ' d d *xco o  g CO co CO  c co co CM  Tf  co  m co o  CD  xj- oo o oo co co o m  O o  xj- O xj- O N - O O o - N T C O C O C D O O  h-*  O  d  d  d  cd  CM d x-  d  co  o  d  D  d  O  C  S  I  r  -  d  d  O  O  d  N  d  -  O  d  c  O  C  N  -A c CO  0  C C  D N  O O  O O  C N O O I x - O O O  ^ O O O O C  O  i  O  d  O  c  d  xj- o T co o o N  O  C  O  ( D ' - O C D O O l O t D t t O I N C D O O C O O O C D O O C D O O O  c CO  00  CJ> -rj CO g CO CO  CM d  d  O  O  O  O  O  C  M  O  C  O  C D T - O X J - O O I ^ C D X I - C O T - C O C D O O C O O O C D O O C D O O CM  d d  T—  o d o  d  d  CM  d cd  O o O S CD f oo O CM o O 00 O O CO O O CO O O d d T— o d o d d csi d od  CD T T3  C CO CO  CD CM  CDCMO-r-OOCDcDxtmoCM C D O O C O O O C D O O C D O O •A co CM O O O O O O O C M O C O g oo CO CO g CO CD  CD CO CM  CM O O O OO  CO CM  o o co m o C M OO O O CD O O d d d d CN d od  •6  ("-CNOCDOOx-CD'xrcOOxco q o c N O o r - - o o c D o o c M d d - t ^ d d d d d c N d o d  X*  M N O C D O O  l-O  c CO  g CO CD 00 "CJ c CO CN  o  CM  OO  t C  O M  O O  m i O I  D t ^ O  i O O O C D  C O  M O  OOOOOCMOCO  N CN O CO O o CD O O CN O CN  d d  CO CM CO CN  o d  o o d  T—  m m xj- m o CN d O h~ O O CD O O o d d d CN d od  CD CM x~  o o d  o o d  oo d  CD  o d  CO  o d  CN CD CN  O T— O O d od  CO  s CJ  I D O) ^ M CM N LO N <«) N O f ) CM d d ^ d c o CO  ^ CO co CO  g CO CO  o o To o O o co o d  C  d  h - C N O x - o o h - m x } -  - v r O C M C M O - v T i o c M O T - T } T3-  D  ""O o) C M c CO  O ! - S . O C \ | O O C N C 0 T - T - C 0 C 0  N d d n d d w ' d d t d o i  CM  t^cMocooocMin-NrcooT-  CM x~-  CM  o o o  m TJ- C D O O O C M C M O O O - r - T - m C D C M o i n o o o o c D o o o o - v r r~ d d ^ o o ^ o o i n ' d d xx— co o co  -6 <o co g oo CO  CN O )  od  C O  d  ID T - I f l O ^ co N n m o ^ CM d d ^ d CD  (D  co  xCJ)  o o  xd  00  CM  m o CM co o o d d i - d d d d d csi d od  N - C N O C D O O x - c O x r C O O O C M O O t ^ O O  00  d  o o  CO  o 00 o  CM  d  CO  O  O  Q Q O O ; O  o  0  00 -vt-. C£> CM oo oo m d d -sr oo  s c M o m o o n c D i - t o c M CDOOCMOOI^-OOCDOO d d i - d d d d d c N d e c i c oo c\ (0  OO  CM  : J O O O J  c_ 5  ra O _i 2 C 0 r - N < O > U - ' § ' 2 O Z I -  ^ x t x ^ + cO + C O C M C N c M + 2 I I " i ?2 " n a> c o> co c o o 0 ) I - N < O > I L " > ' > O Z I -  156  CN  LO  C CO  oo CO  •<- T -  CO  d  O  CD  LO  CO  s  LO  C CO  00  co  CN •fl-  7 -fe  CNc o C <M CO  ^^  LO  TJ  d d  00  => t: 7 -o ^  LO  vCD CN  CD 1 -  O  O O  CN I O CD s  cd d d  co CO  CN  T-  o •<- co o o oo o o O  o  o d d  CO  CD  o  co  CO  T-  CN •<- CD  s  n  (DO o ci oi  d  d  CO  1 -  O  o  O  co  o o d d  CN  IIs  CM  CD  co  a  CN  O  00  CO CM  oo  CN  r-^ o  O  O  CN CO  CO  CN T—  d d  o co  CO  CD  CD  _  CM  s  oi  i  LO  CD  s  CD  -flO  O  O  LO CN  T— O  CO d  LO CO  O  CN  O  O 00 O o ^ d  CO <N  T—  CO o  CN O  d  CO OO  CO  N.  T-  CO CN CD  CM  s  ,_ fl•fl  1  CO  O O  O I-  CD  s  oo CN CD o CN o  c  •<- oo O I d d co s  ^  CO  LO  cn  C  T—  00  <u E  CO  3c  o d CD CD  fl  1  co  co CM CD co  3  d  co  > -5 .° ^ C  v-  d  -fl-  d  d  CD  O CD CM co CO CD O A co d d CO d •<—  x—  1  o d o  —  )^  ^  P oQ _£>  N < O  CJ  >  CD* LL  o o c  ^  cn co co o O Z I-  2  o CM CN •fl" O CD 00 co o d d fl- d cri co CD LO  CO CM o T— CD CO o Id • f l d CD CO cn  fl-  s  ed d  d  cn csi d  d  oo1 d fl CO  co  fl-  eg CO o o O o  I-  d  d  CO d  LO  •flo  LO  00  CO 00 CD  s  cb cn  •fl- CD •fl- CO CO CD CN CD co o o O o o LO o co LO d CO d d CO d d •fl d cri  co d d CM co CM  t—  CN  <N  r~  CD  3  CN  co fl- o oo o CD CD d d  o o  CD  CN cn  fl-  CM  o CD co CO fl- o o d CM d d •^ co d CD CD  T— O  LO O  CM d  d  fl  CN o CN o CM d  o CD  1  CD CO floo CM CM o d d •fl d co  LO  fl-  d  43  \— fl-  i^  oo d  IIs  c  co CO o  o  o  T—  CD CD  > -7 5 <^N o  d  c\i d T—  CN  d  CN d  cb d  d  CO d  d  oo CN d d  00 CD oo o CD oo o CD d d  LD  CO CN  o  co d  oo oo CM oo o co o o I-' co d d co d  •flfl-  co  o CD o LO  d  o d  s  I•fl- o  CO CD  co  CD  s  d  —  TD Hi Si  CM  o  CM  o  CM  CM CN cn CM CD CN I- co LO o CO d d d CO oo CD T— s  s  co N. 00 CM CD o CO fl- o oo d CO d d fl1 d  o CM  CD  Is  CD CD  CO  CD  O O O re O o o O c co CM o CM CM  CN  O  o o  CO h- N  CO  00  I-  00  J= CD <D  CD CD  s  s  •*- CD  c re re < Q _i  oo  CO o o co 00 CM •fl- o CM o o cn CO  s  LO  s  CO co CO CN o CN I CO ICD o CN I o CD d d d d CO CD  CD co  CD ^  I-  co  o  — Si -~J c oo  5 a.  "5 O  d  I-'  C t— CD  C  CD  T—  CO T—  LO  o CD oo CM CD CD o o CO o I-  CM CO d  T—  s  CD  •a 3.O. _ sz re re o16" o _i z | » I-  oo  o  o  co o  -fl" -fl" T - CD LO CN 00 CM CN TO O CN N d o 'oo o o oo co co n o d d ^ ' d cd 00 CD oo o o  LO  T—  o co oo CM 00 o co d d CN d d •fl d  s  O O  CN  45  I- C N CO O  CD  d  s  co  CD •fl"  o  s  CD o o CM •fl- o  d d  C CO  co  o o  s  CD CD  s  IO  CD  C CD  d d fl- d  s  o o d d CM d d •fl"  CD CD •fl- CN o •fl; CO o O o r~00 d N- co d d  s  T3  •fl-  d  CO o N. 1^ d d CN CO CN  CN  CD •flo t— o CD CD CN I - o co d CN d d • ^ d  o co co o Io o I - CN CO d d •fl-" d d  o ca oo o o o 00  CN d  00 CD  CD  C  CO  d  cn CT) I- o oo CN co o CD CM d d •^ co d  CO  5  C CN CO LO  c  CN TO  s  co  CD CO  00 •fl- o o o co o cn o o  3 —>  LO  s  CO  CD  CN 1^ d 00 00  co d i CO 00 co CD LO TD  CN I- o  d  ^  oo o o CO CM CN CD o o o CD I- CO I- o o d d c\i d d • f l d  CD  o  ^  CN  co  •flc T— CO  t; o  •fl"  CD  co -co -o CD d d -A d s  CM fl- co CN O co o CD O o d CM d d  ~= CD  ^  h- co o  C  O  T-  CO co co  I-'  co  d cd d d  CD O CM CD CO O  "3 ?  CD o CN O  No  ,  CD \—  <  CO CN  >  CD 2  Ll_  O CO H Z  157  l o o N i n ^ o o) t - CM 00 O CO t O O O O N co t o o d d cri d d C N co d d d oi co oo CO TT-  CD LO  co co o co o  oo c  o o o o o o  T-  O  O  CO  LO  LO  CO  T -  d d  CO  CD  CD  CM  L O C O O O T—  o  N  oo "7 x* c v-  (fl  O  T -  lO  T -  co co o o o o ( o d d t c i d  CO  •8 3  CO O  co o  T -  N  c  t O  o  I t co o o o ; '2 co d c i ri 6 c i " O CM K co C CM CD  CO CD  K  r-  -NT  CO  o  LO CM  O)  CO N  d  d  CO  o d  a>  «2 CMoo ? -Q CM x~ g CM """"" CD N  CI)  LO ^ " N O CD  co  d d  TJ-  CO  C CM <0 CM co d  .t>  x-  O  O d  00  C  c  CO loo  N  O  O  CNd  O i d  CD O  C  oo oo CM  d  co  d  D  O  C M i d T -  rO T  5  T—  d  -  od T-  t  O O  CO t  d  d  oi CD  d  t  co  CD  o d d  O  CO  "° ^  C CO 00  <D 3 C  CU  c o  o  LO  o C T-  5  a CO TJ co C <  <  V  c  T -  N  CM  TJ- C N  co  7 -o T-  o O  CO  CO  x—  CD  N  t  co  LO O  o o  T j - T— C M C D CD O O CM  c o ' d d  CM  O  CD  O  xj"  O  co  TJ-  c  CO I— N  <C O  a  > LL  o  O TO O O) CD CO o O Z h-  d  00  CN  C  O  O  co d  lO O  T  d  O  N LO  O  O  to d  O  CD o o o o o co d o  CO  oo co d co  CN t N CM CO O d d  N N  00  o  0) ^ o 7 *- o v-  Is  I-  CD O  T -  CO  N CD  CD t " 00 CO CD  Q  T -  CD  x—  o o Th- o o  co d  c o ' d d  o  o i  L O CN  co d  d  T -  CM LO  o  CD  TJ- d  CD CN CO T -  O )  oo co co o L O d d t o oi  co  O  o  N  00 d  O CO  d'  CM  TJ-  t  CO  T O  d  co  C N T - 00 N O LO t d oi co co  N  CN T— CO CO  d  d  I  Tj"  O  CD  O  t  o oi  CO  CD  o  LO  co co  CO  CO L  d  c i t d  N ai  CO  CD  co c o  C O C O O N O O t d  o  d  T—  CO CO O CD T- O O Tc o d d co  q  CD  N  d od d  T—  CN CN  CD o  d o T j - ' d c o  CO CDCD  O  d oi CD  CD O CD  CM  O  o  o m  O l (O t d  CN  d  T -  t  co  O  o O) d  o  CO  00 CO TJ- CD CO CO CD o co co CM CO O CD O O T f oo co co N d d co d d C N o o tCO CO T— T—  o  CD  o  T -  O  o  oo CD CD  CD  CU XI (0 C CO -1  <  O O  co d  TJco  co  TJT -  T—  TJ- T J -  I N  CO  O  O  N  t-  co  CD  o  co  T—  CD  CO  -  d csi  co  CO  TJ- d  t CN  CO  o  d  d d  d  T—  CD O C D d 73 Y ICO  C o a  —  d  co  CD  N (N CO CN O 00 C O CO O C D  N  N  CD  co  CO  T—  C co  Tco  t O CD O O CN CD TJ" TJ" T~ CM co co o co o o t S C O ( O O N co d d od d d CM d d TJ- d od  B Acu _ Q o o O g O o ^ Q.. _ ro (0 O Q  O  oo co oo o o TJ-  CN  LO  5  CO  i  3  cd  T -  ^  s  n  o o d d  OO O s  CO CO  s  T -  t  t  CO CD O CO O c oo |N- d d co C N CO  CD CN CD 00 CMN co  o  CD  CO T CO CD O  CD TJCO CO  O CD  T -  N  C CD  co  C N CD N CN N  d  co  T O  co  CD  1  t  CO  O  TO  I c o c o c q T r o o c o d d c o d d c M  CD  CSI  T-  O  O  LO  CO  O N J - T J - C O T - O O  d oi 00  d d - x t d c n  CO C  00  3  N  co  O)  O  c:  O  LO O  co  o o  N d C CN 00 CD  ^ f d co  x— C M LO C O O N O O O C D K C O lO O  T-  lO) CN O | N  CM  o o oo o o LO o co  CO CD  co  " D OM  o  t  TJ  CD  CO  0>  J?  CM x—  o LO d d t d oi  J  LO  t  oo oo i—  CD CO o CN L O co co o d o o cd C D d N d d TT d d T co T— i —  O  P ) ( D i f l t O T - ( J ) ' ( D ( D x f r - s |<o c o o C N q q I D o i t o s o o xf d d CO oo d o xr o d T - d d oo  CD  Y -  C CO  00  O  co T-  T-  LO CM  LO  CN  w t-  -5 "2 7 -o L O  xj"  iv.  O  co  LO  N  T -  K  c c  t o oO co  N  CO CO  CD  oo co s o to d d oo -vt o oi oo  o O t o CM d Oo oo  co t  3 <?  f  LO  O  o o o d d t d d  CD  CD  <"? O O CO | N  P O O  _ o z CO h -  N  O <  O  o ^  O >  LL  9 9,o  4,  S  Z  C  O) CO CD O  158  CO  cn co co fl- T - •<- A -q-T-OT--^ a> co o co o o co oo co io o co id d o ri d d I N d d ^ d oi 00 co Tco o)  -fl- •<- o -fl co co o iO o d co  LO  c  5 "3 "2 7 "D LO C  O CD  <CO  ^ i< d  O O  O  U) LO  O  CO  d d -fl- d d co cn  d  O  o •<o o d co d d  O  •fl- CN fl" T00 CQ N O  CN  CN  CN -fl* 00 CN CO CO N o O O  CO  00  ^  3 "P 7 -D •fl- c x—  co  T-  LO  O  N  O  N  d d ro d  O) •fl"  co  CD CN  CO  O  T-  TCO  CO  O  LO flCO O  (D  CD  co  CD  CD N  ai a>  O  s  O  x -  CD  o cn o o d co d d  csi  O  00  CO  CO  -fl" O  CO  LO  T—  co o C O T - fld oo co o o CLoOO do o T-  00  CN CO  CDCD  T-  CN CO  O O  -si CD  co  CN OO  cn ai  O 00 CO T - CO fl- fl- o CN •fl; o o o o CN N O N O O) o o co o d co d d -fl-' d cri co cn CN  T3 CO CN  C D C N O T - L O C N - < LO  00  N  CO  O)  CO  CO  CO  fl; o o o o o s m N o oi o d co 6 o co d d fl- d cri Tvco co  Ico o o fl- T - o co fl" CN LO T - CO fl co o co o o fl- 00 CO N O it-d d cdi CD d CN CDczi^tcDai  ^~ co co 3  d  cp fl; o T - o o oo N d d co d d csi  sc  c co  O  •fl- co co c o L O cn T o o LO cq r o I - o cb d d c o i d d CN d d c- oA d co T-  LO  co  1  T-  N CS CO N N- CN O h-  LO  c:  LO  M  CO  CO  "D  LO  P)  LO  oo •fl-' d cri co cn  LO  ICO  x  co  t—  cn  co CO CO  LO  led d co  c  3 T3 CD C  c  oo in  < a>  CD  o 7 ac c  co  Q.  CD  co  o  CN  O  O  d d  o o r- d d co co co co co  -5-  O  o d  co  cn o co s 6 if CO o o floo co -fl- oo cn co CN  o LO o o o o fl- o co  co  fl- T - co cn co co  o d  LO  co  5 LO  CO x5 i- c  S re o co  re < Q  N <  O  CN  CD  O O O C  O)  2 2  CO  o  4-* O  159  N CO CN CD CN CO O N N O O CN  OJ  O O  co  *r •fe  C  i"-.'  CN 00 CD  CD  LO N  c  Tj*  N  CO ( D O  CO  TJ-  CD  d d  d  d  O  T-  O  d  d  O  d  d  O  d  CN CD  o Oo oCO o  CN O  TJCO  K  CN  oo  CD  i - O TJ- CN CO CO LO O O O Tj" O  CO CN  O N  xj- O CO O  TO  t  00  T-  •cj c  TJ- CD O T- CD N CN CD O LO O  O  CN CD O O O O TJ- O CD O N d CD T - ' d CO T -  T—  O O  CO CD TT T -  3  o  TJ-  7  00 T - CO T CD CO CD O d d TJ d  co  Tj C  N. CN  3  O CD CD CN T - CD CD TJCD O O TJ„ 00 CD d d T— CD CD TJ*  o  co oo T j x-  CO  TJ- O TJ-  t-  d  7-og  N  T-  o o  T-  CD CO CN LO  oo o d  o  O  N TJ-  o  o  CD CN CO O  co o  TJ-  N  d  d  t  oo oo CO Csi  o  o  CN <0  o)  LO O  TJ" LO CO CO Tj"  CD CO O C CO CO CN N d d CN CO O CO O LO CO O N  CO  CD  o  T-  O  co d  O  d  VLO CO O  T—  o co o  T—  o d  n <D c c o o LO  < XI CO  I-  co  TJCO  B  7  CO  O  TJ-' o  TJ-  00 LO O TJ- CO O N  d  CD O T- O  d co o  O O  o  CD T-  d  00  d  C  N  O  T-  TJ; O  O  O  O  d  co d o  T-  o o  =i  *? CO Y—  o  t: ? c co  T-  d  CO  CD  o>  O  co  o  TJ- LO O CO CO O o ' TJ" O  LO CD  r  c  O  co  -  00  c i c i TJ co  >  LL  H  O  Z  TN  CD CO T CN T - O  CO CD  CO  CD  d d  +* 00  co  d  O  o co  LO  Csi  IN t O CN  T— O  d  d  T-  N  •8 CO  I  -  c <  d  TJCO  d  O  LO  CO CO  d  CO  CD  oo  00 T- LO O CO LO O  N  co d  O  d  d  CO  d  co d  o  N  d  d  O TN O CN  d  o  d  TJ- O O O  co d  CD  N  TT O CD O O O CN O  aOJ 3 XI _ ro I ra O Q CO 1-  o CO  d  d  csi  Tj- O 00  t  d  V_  CD CD  CN T— O CN  d  co  CN CO O CD  d  CN  co oo  o O O CO  —  <  »—  O  CM CU  >  LL  ai  00 CN CN N O (D  t  d  oi CD O  d  o d o  T  T—  N- CO 00 O  00  d  TJCO  co co co d  Ol N  d  d  Tf CO  d  t  N O CO 00 O  d  oi  CD  t o co  t  LO  oi oo  00 Tj- CO CN TJN CO 00 O CD  d  d  t  d  CO  N  00 T-  CD  Tj" CO h- O CO CO Tj- O  o  T—  r— co  co  T— O  Y—  o  CD O  c i d o  TT  CO  O O  o d  d  CD  o o T j - o a i  T-  co d  N  TJ-  N  CN CD N CN CO N CN CO O LO  ci ci  CN CD O T—  od  CD  LO CO CO O N CO N- O  d  d  d  TJ" O CO T N- CO TJ- O  T-  CN O O O  LO  CO N O •st CO O  TJ c  Tj c  t  N CO TN CD O CO CN LO LO O O Tj- CD CO CO O LO  Csi  5! m O  O  co d  CO  N  CN  o a  <  CN  T-  N  co  T -  cid o  d  LO N N CO  co  CN  CN  C  CO I- N  T-  T~  v3  T -  Q. JS (0 Q _i  N  Tt  O O CN T j - T I CO CD O CN — 1  co  d  CO LO O TJ; CO O  3  3  C  N N  CO  CN 7 t to  oi  O  N N  d  O O  o O CO o CD O d CN d  co d  O  CD CN CN CO  TN  TJ- o CN oo co co o co o o TJ- O  CN I - O CD O T— CD CO O CN O O CO N o o co d d CN  cd  CN O  T-  CO  Tj  O  T J CD  T-  CO O O O  CO CO  3  d o  TJ- CD O TJ- o  o 00 CO co d d  co d  LO 00 Tj- LO O CO CN O T— O  CN  d  N. C CN CD  00  s  Tj  o d  T-  co  d  co oo Tj Y -  00 T -  CO LO  d  o o c o o o c o  co d  O  O T - T- CN O0 00 CO N O CO  d  CN  M-  Y — CD  T J * ^ CO CO g CM CD  I  oo  CO  CO  t: <°  OJ CO  CN  C  CD  o  CD CO  CO  c5  CN  N  d  O T-  N  Tj  oi  CO  C  Tj  d  d  O N tCD CO O N d d  LO  N  O  LO O O O T-  co  LO T -  o  CN  CO  |i— oo  NL ^ CN  O  T-  O)  CO  o  <N  •X; CD 3 CN  CN  TJ"  co  5  -O CO 00 c CN CD  co  T-  N  tt CD CM  oo  N  CD T - O CN CO O  CD  CO  6 d oi d c i co  00  N  2  LO CO  d  d  2 CO  CD  o  d  CO N CN CO  CN  00 CO O CN 00 T -  CN TO  d  CD O T- O CD CO TJ-  o o  CN  CO  g  co  co  CO' CD LO T—  CO  CD TJ-  co •§ CN  LO  CO CD CD O O LO CO O LO  CD  CO CN  o  T -  oi  CD  0 0 09,2 c o to co o 2 2 O  Z  I-  160  CD  Oi  2  OO CO CO  CD  co ci  co o d  IS  00 CO  o d  o d  o  CD CM  £ d  CM  00 d  o  S  co CM  co  d  CJ>  3 5 co  oo  c  o 00 CO  d  co  o o  ci  d  Is  co Is  CO  IS  co ci  N . TO CN  cp  CO "O  3  CO  O) Y—  co  C  o  o LO  o 7 C  < "5 JD  n co  Q. CO  oo  *-  n 2 c re < Q  o o  T—  co fl  d  CO T—  o o  CN  co  o d  d  r-- c i  d  CO  d  oo oo ci  o o  co  o o  co S  d  00 d  IS  fl  o o  LO  d  cri  co oo  IS  d  s  CJ)  00  CN  co T— 00 o O d fl d oo CM  c  CM IS  CN  CM  o cri cn  00  CD CM  fl 00 d  00 oo d  CJ)  CM  co o fl d co  d  co  o fl  o o  ci  d  fl  C „O I CN C O CM S  CO  fl hCO  CN  fl  r-CO  OO IS  CN CO  <J) CO  d  r-co  to  hCO  fl o d  CN OO  o o  ci  d  CO  o d  C\i  fl d  CM  5  CJ)  co d  o o  oo  fl d co  cri cn  CJ) CN  cd  T—  o d  d  CM  O  Is  co  d  CO  co d  CN  o  LO CO  o d  co o> o CN d IS  x—  oo  o d  CN  cd  co oo d  co co d  00 o o CO  fl  d  co  IO") S  CN  CN CD  00  o o d  CJ) I-  o d  o o  S  CN  CM  o d  d  CN LO CM  oo 00 d  co oo d  CO  00 d  CM CO  d  CM LO  fl co  o d  CJ)  oo o d  fl  CO  Is  oo ci  h-  o  d  oo  00  00 co  o d  o o  CO CN  d  CN  IS  oo d  CN CO  d  LO  o o  fl  d  cn  CO  LO CO  d  g c  1c  hco  CN CO  -cO  d  d  LO IS  co v-  co co oi T co ro io o d d fl d co  CD  CM O  O O  LO T-  CD h-  CO CO  d  d  CM  d  ci  T-  K cp  cji  CN CO  TO  LO fl  fl  d  cri  CO  CD  CO  IS  fl c  co  T-  oo o  CO O  O O  o fl  O O  CM O  T-  LO  CD  d  d  CM  oo co co o o o fl o  fl  CO  T-  CO  T—  o cri  oo  S  co co d  o o d  o o d  LO CD CO T—  LO O T—  o o  o o  d  d  CO CN CM T  CM  o d  o d  oo co d  CO  d  o 00  O  co d o  o cn  fl co  IS  CO T—  oo d  co co d  CD LO  fl co  o d  C  ? N.  o co  o d o  SI _ re O —J Z  CO  Q o o o 1- N o o o o CN CM  CM  ro  CO  CO  L?  CO  <  CN  >  CM  CD LL  2  co  O  co  z  S  x-  13 CO C  CD  C CD •O  CM CN CM  C co  CJ)  cri  CD  00  v-  0  7  +4  o  o  d  T-  fl  fl  o  o  CO  co  cri CD  CD  CO  O  LO I-  O  i-  oo  d  d  CN  d  d  00  I-  O  O  T-  oo  oo d  T-  T-  Is  oo  S  S  d co d  fl  LO  CN  LO  co o  o o  CO  fl  ICM  I-  d  o o  IS  O  o  o  cj> o oo o  fl  LO  co r--  c o h— L O co o o  CN  O  O  fl  O  OO CJ)  fl CO  1I-  vO  CO O  d  fl d co  d o  oo  co o d  cj) O T - 00 cd d  T-  oo  CO  00  T-  CM  LO  I-  cj) o  oo o  LO  oo  fl  O  S  o  S  fl co  s  oo  s  cn fl d co  o  T—  o cd d T-  o d  cn  o  CM  CN  O  CO  L  O  T  -  N  -  T-  io  d  LO  d  o o  d  oo  d  o o  A  T  O  -  C  r- •<- oo co  CM  00 CO CO d d fl oo  ILO  CM OO  S  CM  o  N  T-  T-  fl CO  o  CN  o d  CN O CD O  fl  CO  CO  o  CO  LO CO  T-  L O co o fl o o co oooocj)Oco d d d c N o ' d f l ' d d T3 fl rco^ d d c -<i co o  c  CD  CD "D  *3  M  T-  I-  N. oo  CN  CN  fl TJ T3 0) 3 C  - O O T -  CO  CD  i  S  h - d d c d d d c N CO  S  d o  O O O OCM re c 2  Tj  ,  co  CD  CM  o d o  s  I-  O) T—  <-. 00 C f~  CM  CD  "o ^ O)  CD O  ro  oo co  fl-  cri cn  CD  co  S  CM CJ)  co  CD  "2: ro "O  cri cn  O  C O C O O I  fl fl fl  fl  cri cn  • " - C D L O C O C M T - f l  c  CD  CO ^  o d  LO  CO  CM  CM  g CM CD  CD  LO fl  CD  I-  CN  CO  c  d  CD O  o o  T—  ^i T"2 2 3 fl  C  o o  CD CM  L O I-' CJ) C O  -^r CD  00 v-  CN  o o  S  3  CM  o d  o oo  I-  CD  CN  CN  CO  T3 C  "O  fl  S  CN  CO  "D CD  CN I-  CJ)  co  C CD  00 ^  o d  cri  CD  3  CN  CO  CD  o d  O f l o o c o ^ T - c o l o o c o o L O O O f l o o oo o d CM  LO  CN  ? -Q £>C O  "5  00  co  CM  CN CD  <N  co  CO  3 CN  S  co  oo c CM  I-  LO  fl  o d  00  CD  CO  CM CO  fl CN  00  CN  IO S  i< d co  Tfl  CD CM  O CD  CM O  O O  d co d  o  O  d  co o  o  fl  cn co d d  co  fl  T-,  I-  o d  s  fl co  co co c o T -  O fl O O LO h - 0 0 - r - O O O C O O O C M O O LO O 00  CO  5  1 * c re re oO a  < a  CD  O O  CO r—  N  o q _<M  <  n 0 ^  CM  O >  q  CD LL  O OOO ^ -  N  c  ra  —  ro  CM  03  5 O Z  75 +rf  o H  161  CI,  t  5 tK  CN  00 CN  co  CO CN  CD  O CO O  c  CO  d cj CM O  CN  TJCO  CN CO  d d  t  TN  Tf  O  d  CO  T—  N CO  CN CO  O N  O O  CO CO  o o d d d d t o oi CM  0 0 0  CN CO  O  CO  1 -  O  N  LO  d d  CM  00 0  CM  CO N  CD  CN O  00  co t o 00 oi d co CD  CO O T - O) tN x- f CM O O O t CO CO L O O O CN t xT co  CM  d d  O  N  co  d d  O  O  T-  CO  CN  co  CD  T-  O  N  s N CO  d d  d o d  co o  d d  T-  10 o o  co  d d  CO O  CO LO  o o  O  CO  O  x—  TCN  Y—  T-  o d d co t d o d LO  CM  CO  00  O O  N CO  CO CO  O  CM O  O CN  d  T-  0  t d  00  1 00  O  co CO  CO  co en Tj Y—  CN  N  o co co 0  CM  O  O  [N*  7 TJ c  T-  CO  TJ-  CO  TJ-  O  LO  o d d co d d  2 <p  N CM  T-  CO  CO  N  CN  CO  Tj  t  CN  co  c co „co o o N  CO CN  T~  LO CD  co CO CD  co  d d  I N  TJ-  T J <"»  O  t o o c o o o t  CO  <0  CN  O  O  CO  O  O  TJ-  CN  CD  t co o o d dot d o  CD N  CM CO  CO CD  CO  CO  CO CD  LO  Tj  c CO  CD CM  O O  o o  CN CM  CD CO  CO O  CO  oo  o> c  CO  CO  CD  00 co 00 TJ"  o o  O  o O  CO  rt o T - 00 co o co 00 o o o 00 co o o T— o o o co d d o o o o O  3 •£>  T-  O O  O  CD  LO  CO  T-  r-  O  CO  CM  CD  CO  O  CO  O  O  N  d d  to  d  0  LO  LO CO  CM  T— CO  co "S  o CN  0J "3  C  XJ C O OJ T 3 c  C CO  o o d co d  CD CN N CM  Tj v-  §13 c o 7 T J 00 o +j o C c:  T-  CO  CM L -O N CO  CO  00 CN  d  T-  o o d d TJ-  O O  CD  O  CO  TJ-  CM  CO  o do  O  O O  TJ-  TJ- TJ-  00  CD O  CO O  O LO CO  O CM  CO CO  CN CO  CM  O  CD  O  CO  Y-  10  < Xl CO  I-  o  CO CD  CM  O  TO  O  t T Td d o tco d o  00 Y~~ CO  N  T-  co00 o o o O  CO O  CD CN  LO  d d  CO  O  co  Q. 10  c <  —  OJ  £» ro ro Q _J  _  o z  Q Q O ^  ^ O  C 0 r - N < O > l X  ^ c  01  co"  S O Z  o r-  162  CD Cl)  c  £ 1c CD  O  cc  sE  CD XT -*—  I  1  c  CO 0)  c  o  o  JD co  T5 _. CD  CD  •o "° x:  c CD CJ  Ii  CO O CD O-  >> CO  *  CD  co o £ i  c i  CD XI — O CO u _ *—  CL «=  o cu  fl LO CM A co CO CM CO  * c i< d o d co  T—  CD CN CD oo r- oo CD o CD CN d d fl d CD CD — oo  •"C— LO  o o d d  fl CM CM o CM CD fl CN fl fl CO o I - o CD o oo o fl 00 I co d d d d CN d d fl co d s  LO  C  s  •i: co co  T—  I I s  s  cri CD  fl |00 O  00 O  fl CM T - 00 00 LO O O CD O  CM i - T O CM O  d d cid cid d d d d d  CO 00  LO CO I - CM T - L O fl O O CN O O CO O s  fl  LO  fl  CM  T - 00 T - O CO O CM O LO CN  d d o d d d d d d cid d  CJ  'E o i . i  o J ?  CD  CO  flcI -fe C° s  CD  ?s . E  OJ  O CD  >t x :  O  CD  c §  > CO CD  43-1 CD c  4w3 c  l  a.  CO  C  13 CD  s  o d  CD CD CD  co fl LO N C N CN I - CO o o CD o o fl O O CM O CO CM co o o o d d o o d d dd o I s  CN  T -  s  CO •+-*  o o d d  o CD CM  o CO  CN CN CD oo  T—  o d d fl d  t—  00  00 LO  CD CD  N C N CO LO CM Tfl o T - fl O O CM O fl O LO poooooo oooo d O CO  00 fl O O  T -  CO CM  ° - COLO  c  CO  Q_  ~ o  CO  13 CD  CO  CD  =  00  CM  ci o d  CD  c\i  o o d d  T—  CM co  fl r-- co LO o fl d d fl d cri co  LO  CD I s  CM  o o d d  o LO  oo  CO  co  CO  I s  CN  o  LO  CD o CO d d fl d s  CO  o o d d  CN LO  oo  •<—  LO I -  CD CN CM I -  \—  o d d fl d s  s  CO  CD CO CO CO T - t - CO C N O O C O O O L O  ( O C N O T O O C M O  d d cid d c i d czi o o c5  CD CO CO CM  CD TJ  fl CD cri CD CD LO  O) CD  _3 fl CO  fl O  CO O  CD CN T - CD I D CN CD r LO O O h O O T - O  ddddddd  I -  co  fl  CN  s  dddd d  Qoo1°d^ooO| N  • • < 0 > u . 2 S O Z i -  O C  CO CD Q_ O i_ TJ TJ —  ff C D C D f l C D i - T - f - O O O O O O O O O  O LO fl T - O C M O  LO CD fl  I s  d o d •<- d d •<- d d d d  ™  co  CD  x:  CD  I -  T—  co cI- d o d  CD  fl  00  s  c3 «  X) CO  I s  co <"  JD '  ci o d  CN CO CD  "5  <  s  I -  T3 ry)" CD CD  C D  OO CD CO I CO CO co o I co ci d co  CM C 1< -fe CD O CD CO  #— CO c  CD CO co CD CD fl CO  o o fl OO co d d CM d d  to  co  o E Q.  CO  x— T—  c  s  ™  CD  LO  T—  CO I - CM CM co  f__ -»-«  eo  CD  s  .§£ CD '5 co  CO  fl co 00 o ci d  9 9 - H CO  y o N < 0 > U . U  M  c ra ro co o II S S O Z H Z  163 Appendix B: Cation Normalisation Routine  % % P R O G R A M TI A N D . M % N E W V E R S I O N W I T H Z r corrected  % % M A T L A B script to compute titanian andradite % mineral formulas from oxide weight percent data.  % % Algorithm is based on 8 cations, 12 oxygens or oxygen and O H equivalents. % Oxides (in order) include: % S i 0 T i 0 A 1 0 Z r 0 C r 0 V 0 F e 0 FeO M n O M g O C a O N a 0 H 0 % 2  2  2  3  2  2  3  2  3  2  3  2  2  % D A T A is input as row vectors o f oxide weight percents [and.dat]  % % Computes cations in the following order: % Si T i A l Z r C r V Fe3+ Fe2+ M n M g C a N a (H)4  % % Computes Thompson Components as follows: % Andradite (Additive component) Ca3Fe(3+)2Si3012 % T i S i - , AlFe(3+>, ZrSi-, CrFe(3+)-, VFe(3+)-, % TiMgFe(2+)-, FeMg-, M n C a - , M g C a - , C a (Linearly Dependent), NaFe(3+)2Ca-, H 4 S i % % O U T P U T includes: % andecho.out echo file o f input data (oxides & total % andcats.out optimized anion distribution (cations & total cats & total O's) % andrecalc.out new oxide totals based on mineral formula (oxides & total)_ % andtcs.out cations represented as thompson component basis  %  % 'PROGRAM TI_AND.M' delete andlog.out; diary andlog.out; echo off; format short G ; % CONSTANTS catn=8; oxyn=T2; % Change this i f you change oxide list (Water input as H 2 0 ) n_cat=13;  164 c=[l 1 2 1 2 2 2 1 1 1 1 2 2]; o=[2 2 3 2 3 3 3 1 1 1 1 1 1]; ox_eq=o./c; % H 2 0 operates as (OH)4 = = S i 0 2 == 2 Oxy's ox_eq(13)=2; atom=[28.086 47.900 26.9815 91.22 51.996 50.942 55.847 55.847 54.938 24.312 40.080 22.9898 1.00797]; mw=atom.*c+ 15.9994*o; % Read in transformation matrix load and_coef.dat; c_matrix=and_coef; % read in data & create matrix of'n' analyses with'm' oxides load and.dat; w=and; rawdat= [w sum(w')']; save echoand.out rawdat -ascii dw=size(w) ncomps=dw(l) n_ox=dw(2) rhs vector=zeros(n cat, 1); x_vector=zeros(n_cat, 1); a _matrix=zeros(n_cat); for i=l:n_cat a_matrix(i,i)=1.0; end % Process individual garnet analyses for j = l :n_comps w(j,7)=w(j,7)+w0,8)*1.1113; w(j,8)=0:0; cats0=(w(j,:).*c./mw)*8/(sum(w(j,:).*c./mw)), ox_tot=sum(catsO. *ox_eq); del_ox= 12-ox_tot; i f d e l o x <= 0 % e.g., h2o not present rhs_vector=catsO; rhs_vector(8)=12.0; a_matrix(7,8)=1.0; a_matrix(8, :)=ox_eq;  165 x_vector=(a_matrix\rhs_vector')'; else cats0=cats0/(del_ox/4 +1); cats0(13)=del_ox/(2 - oxJot/8); x_vector=catsO; end % S T O R E cation normalization catout(j,:)=[x_vector sum(x vector) sum(x_vector.*ox_eq)]; % Recalculate oxide formulas (convert (H4 to H 2 0 ) mol_mass=(x_vector. *mw./c); mol_mass( 13)=x_vector( 13)*2*mw( 13); wt_new= 100*mol_mass/sum(mol_mass); wtnewout(j,:)=[wt_new sum(wtnew)]; % C O M P U T E TCS F R O M Xvector tcs(j, :)=[x_vector*c_matrix]; end % Write out optimized cation formula save andcats.out catout -ascii save andrecalc.out wtnewout -ascii save and tcs.out tcs -ascii  % E N D O F P R O G R A M - J K R 08/99 revised 11/99  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089519/manifest

Comment

Related Items