UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Trace element analysis of placer gold Tetland, Mikkel 2015

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

Item Metadata


24-ubc_2015_may_Tetland_Mikkel.pdf [ 5.38MB ]
JSON: 24-1.0074427.json
JSON-LD: 24-1.0074427-ld.json
RDF/XML (Pretty): 24-1.0074427-rdf.xml
RDF/JSON: 24-1.0074427-rdf.json
Turtle: 24-1.0074427-turtle.txt
N-Triples: 24-1.0074427-rdf-ntriples.txt
Original Record: 24-1.0074427-source.json
Full Text

Full Text

i  Trace Element Analysis of Placer Gold  by Mikkel Tetland  B.Sc. Honours Geology, University of Saskatchewan, 2012  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE COLLEGE OF GRADUATE STUDIES (Environmental Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Okanagan)  April 2015  © Mikkel Tetland, 2015  ii Abstract  Trace element analysis provides an efficient method of fingerprinting placer gold populations in order to characterize bedrock-source hypogene mineralization style and effects of supergene gold mobilization in surficial settings. LA-ICP-MS analysis, using the AuRM2 gold reference material as an external standard, measured concentrations of Mg, Al, Ti, V, Mn, Fe, Ni, Cu, Zn, As, Se, Rh, Pd, Sn, Sb, Te, Pt, Hg, Pb, Bi, and U in placer gold samples from: the Prophet Mine Australia, Nus River Colombia, Piaba Brazil, and four locations in British Columbia.  Additional reference materials NIST610 and FAU7 were used to quantify the micro-homogeneity of AuRM2, monitor precision, and confirm accuracy. The AuRM2 reference material had not previously been utilized for micro-analysis, but is shown to be sufficiently homogenous at a micro-scale (64-108 µm). Precision of analyses of AuRM2 range between ±10-15% and accuracy is better than ±10%. Semi-quantitative concentrations for V, Hg, and U (0.3, 3.7, and 0.1 ppm respectively) were determined in AuRM2 along with Hg in NIST610 (0.6 ppm). Quantitative concentrations were determined for Se, Rh, Sb, and Te (24.9, 0.1, 0.3, and 2 ppm respectively) in FAU7.  Placer gold in the central Okanagan from Mission Creek and the Winfield paleoplacer share the same hypogene trace element signature indicating that modern-day placer gold in Mission Creek is reworked from Miocene paleoplacers similar to the Winfield occurrence. Samples from Lambly Creek, also in the central Okanagan, exhibit two discrete trace element populations of native gold; one has elevated Cu indicative of greenstone-hosted orogenic gold, the other has a low Cu signature similar to intrusive-hosted mineralization. Both groups contain  iii primary mineral inclusions and limited supergene rims indicating two proximal hypogene gold sources within the catchment of Lambly Creek. The samples from the Prophet Mine contain biologically precipitated supergene gold. Gold-rich rims are up to 100 µm thick indicating significant supergene gold accumulation in this placer occurrence. Analyses of regions of gold grains containing primary hypogene inclusions retain a distinct hypogene signature rich in base metals (Ni, Zn, Pb, and V) whereas others had an indistinct signature of supergene gold richer in Sb and Bi.                  iv Preface All data presented within this work arose from observations and analyses conducted primarily by the author with the exception of some results in Chapter 4.1 entitled: Reference Materials Results. This chapter is the product of a collaborative effort between the FiLTER (Fipke Lab for Trace Element Research) lab in Kelowna, B.C. (British Columbia) and Dr. Brian Fryer at the GLIER (Great Lakes Institute for Environmental Research) lab at the University of Windsor. Results for this section are comprised of standard reference material analyses of NIST610, AuRM2, and FAU7. Of the analyses for AuRM2 and FAU7 compiled in this section half of the data came from the GLIER lab and half was analyzed by the author at FiLTER. The advantage in having analyses from multiple labs lies in the ability to cross-reference results of a standard reference material which has not previously been used for this method of analysis. All data-set compilation and statistical interpretation of the combined standard reference material analyses were conducted by the author.              v Table of Contents Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents .......................................................................................................................... v List of Tables ................................................................................................................................ ix List of Figures ................................................................................................................................ x List of Equations ........................................................................................................................ xiv List of Abbreviations .................................................................................................................. xv Acknowledgements .................................................................................................................... xvi Dedication .................................................................................................................................. xvii Chapter  1: Introduction .............................................................................................................. 1 1.1 Previous Work ................................................................................................................ 3 1.2 Objectives ....................................................................................................................... 8 1.3 Project Samples ............................................................................................................... 9 Chapter  2: Methodology............................................................................................................ 12 2.1 Sample Collection ......................................................................................................... 12 2.2 Sample Characterization ............................................................................................... 13 2.3 Sample Preparation ....................................................................................................... 14 2.4 Petrographic Analysis ................................................................................................... 15 2.5 SEM-EDS Analysis ...................................................................................................... 15 2.6 LA-ICP-MS Analysis.................................................................................................... 18 2.7 Multi-Dimensional Statistical Methods ........................................................................ 20 Chapter  3: Geology and Petrography of Gold Samples ......................................................... 22  vi 3.1 Introduction to Geology and Petrography..................................................................... 22 3.2 Kelowna Area Placer Gold Geology and Petrography ................................................. 22 3.2.1 Overview ................................................................................................................... 22 3.2.2 Historical Production ................................................................................................ 23 3.2.3 Geological Setting of Kelowna Area ........................................................................ 24 3.2.4 Bedrock Geology of Kelowna Area .......................................................................... 27 Winfield Mine ................................................................................................... 30 Mission Creek ................................................................................................... 31 Lambly Creek.................................................................................................... 33 3.2.5 Petrography and Gold Grain Morphology of Kelowna Area Samples ..................... 34 3.3 Geology and Petrography of “World-wide” Assorted Sites ......................................... 39 3.3.1 Overview ................................................................................................................... 40 3.3.2 Geologic Settings and Bedrock Geology of “World-wide” Sites ............................. 40 Prophet Mine, Australia .................................................................................... 40 Nus River, Colombia ........................................................................................ 41 Piaba, Brazil ...................................................................................................... 41 Petrography and Gold Grain Morphology of “World-wide” Sites ................... 44 Chapter  4: Results...................................................................................................................... 49 4.1 Reference Materials Results ......................................................................................... 49 4.1.1 Characteristics of AuRM2 Reference Material ......................................................... 51 4.1.2 Precision of Reference Material Analyses ................................................................ 51 4.1.3 Accuracy of Reference Material Analyses ............................................................... 56 Matrix Effects between Reference Materials .................................................... 60  vii Semi-quantitative Concentrations ..................................................................... 62 4.1.4 Homogeneity of AuRM2 .......................................................................................... 65 Multi-Dimensional Scaling Plots of Standards ................................................. 75 4.2 Kelowna Area Placer Gold Results ............................................................................... 77 4.2.1 Major Element Results for Kelowna Area Samples ................................................. 77 4.2.2 Trace Element Results for Kelowna Area Samples .................................................. 87 4.3 Elemental Variation of Placer Gold in Combined Sample Sets ................................... 93 4.3.1 Major Element Results for Combined Sample Sets .................................................. 93 4.3.2 Trace Element Results for Combined Sample Sets ................................................ 107 Chapter  5: Discussion .............................................................................................................. 136 5.1 AuRM2 Reference Material ........................................................................................ 136 5.1.1 Precision of AuRM2 Analyses................................................................................ 136 5.1.2 Homogeneity ........................................................................................................... 138 Assessment of AuRM2 Homogeneity by Comparison to NIST610 ............... 140 Assessment of AuRM2 Homogeneity using Counting Statistics ................... 142 Possible Sources of Heterogeneity.................................................................. 142 5.1.3 Accuracy of AuRM2 Concentration Values ........................................................... 146 Matrix-Dependent Effects on Trace Element Fractionation ........................... 146 Semi-quantitative Element Concentrations ..................................................... 147 5.2 Kelowna Area Placer Gold ......................................................................................... 150 5.2.1 Winfield and Mission Creek ................................................................................... 150 Hypogene Signature of Mission Creek and Winfield ..................................... 150 Supergene Signature of Mission Creek and Winfield..................................... 152  viii 5.2.2 Lambly Creek.......................................................................................................... 155 5.3 Elemental Trends in Placer Gold ................................................................................ 157 5.3.1 Mode of Trace Element Occurrence ....................................................................... 157 5.3.2 Signatures of Individual Localities ......................................................................... 160 Prophet Mine, Australia .................................................................................. 160 Prophet Mine Hypogene Component ........................................................ 160 Prophet Mine Supergene Component ....................................................... 161 Nus River, Colombia ...................................................................................... 163 Piaba, Brazil .................................................................................................... 164 Northern B.C. .................................................................................................. 165 Chapter  6: Conclusions ........................................................................................................... 166 6.1 Reference Materials Conclusions ............................................................................... 166 6.2 Kelowna Area Gold Conclusions................................................................................ 167 6.3 Elemental Variation in Combined Placer Gold Sample Sets ...................................... 168 Chapter  7: Future Work ......................................................................................................... 170 7.1 Reference Materials .................................................................................................... 170 7.2 Kelowna Area Project ................................................................................................. 172 7.3 Characterization of Native Gold Trace Element Signatures ....................................... 172 References .................................................................................................................................. 174 Appendices ................................................................................................................................. 182 Appendix A Sample List and Petrographic Descriptions ....................................................... 182 Appendix B SEM Major Element Data and LA-ICP-MS Trace Element Data ..................... 189 Appendix C Reference Material LA-ICP-MS Trace Element Data ....................................... 198  ix List of Tables  Table 1.      Instrument parameters used during SEM-EDS analysis of standards and             placer gold samples .................................................................................................... 17 Table 2.      LA-ICP-MS instrument parameters for trace element analysis. ................................ 19 Table 3.      Key characteristics and geologic settings of sample sets used for this project .......... 42 Table 4.      List of trace elements added to AuRM2 certified reference material along             with concentrations and uncertainty values ............................................................... 52 Table 5.      Precision values of each element analyzed for standard reference materials ............ 54 Table 6.      Accuracy statistics for standard reference materials treated as unknowns           against another reference material used as an external standard ............................... 57 Table 7.      Calculation of mass of each element sampled for AuRM2 and NIST 610 ................ 67 Table 8.      Theoretically predicted and actual measured RSD of AuRM2 to quantify heterogeneity .............................................................................................................. 74 Table 9.      Average trace element results from LA-ICP-MS analysis of Kelowna area gold samples ....................................................................................................................... 90 Table 10.   Average major element results for all samples sets analyzed by SEM-EDS .............. 94 Table 11.   Average trace element concentrations by locality separated by quantitative             and semi-quantitative elements. ............................................................................... 108 Table 12.    Average element concentrations for two sample sub-groups of Prophet                 Mine analyses ........................................................................................................... 134   x List of Figures Figure 1.      Gold grain Au-Ag-Cu ternary system developed by Townley et al. (2003) to differentiate porphyry associated deposit types. ....................................................... 5 Figure 2.       Various gold ternary diagrams by selected authors. .................................................. 6 Figure 3.       Belts of the Canadian Cordillera ............................................................................. 25 Figure 4.      Schematic representation of the evolution of the Canadian Cordillera just           prior to the accretion of the Intermonane Belt ........................................................ 26 Figure 5.      Bedrock geology map of the central Okanagan ....................................................... 29 Figure 6.      Weakly lithified gravels outcropping adjacent to Mission Creek ............................ 32 Figure 7.      LAM2 gold grain from Lambly Creek ..................................................................... 35 Figure 8.      LAM8 gold grain from Lambly Creek ..................................................................... 36 Figure 9.      MIS2 gold grain from Mission Creek ...................................................................... 38 Figure 10.    WIN5 gold grain from abandoned Winfield Mine ................................................... 39 Figure 11.    AUS5 gold grain reflected light petrographic image ............................................... 45 Figure 12.    Reflected light image of COL7 gold grain ............................................................... 46 Figure 13.    Piaba Nugget ............................................................................................................ 47 Figure 14.    Reflected light image of Piaba nugget...................................................................... 48 Figure 15.    AuRM2 concentrations calculated using NIST610 as an external standard normalized to accepted concentrations ................................................................... 61 Figure 16.    Compilation of average background corrected counts per second versus           %RSD data by element for AuRM2, FAU7, and NIST610. ................................... 71 Figure 17.    Elemental MDS plot of n=64 analyses of AuRM2 .................................................. 76 Figure 18.    BSE image of LAM8 grain from Lambly Creek ...................................................... 79  xi Figure 19.    SEM-EDS linescan of LAM8 gold grain ................................................................. 80 Figure 20.    SEM-EDS linescan of LAM7 gold grain ................................................................. 81 Figure 21.    Above: Composite SEM-EDS elemental map of LAM8 gold grain.                  Insets: Individual element signals isolated from composite image ......................... 82 Figure 22.    BSE image of MIS2 gold grain from Mission Creek ............................................... 83 Figure 23.    SEM-EDS linescan across MIS4 gold grain ............................................................ 84 Figure 24.    Left: SEM-EDS elemental map composite image of MIS2 gold grain.             Below: Isolated signals for each element detected. ................................................ 85 Figure 25.     Backscatter electron image of WIN6 gold grain ..................................................... 86 Figure 26.    Fine scale SEM-EDS linecan of WIN5 gold grain ................................................... 87 Figure 27.    Probability plots of select trace elements from Kelowna area samples ................... 91 Figure 28.    MDS plot of Kelowna area samples ......................................................................... 92 Figure 29.    Box plot of major element data for grain cores by locality ...................................... 96 Figure 30.    Ranked probability plots of major element concentrations ...................................... 97 Figure 31.    Backscatter electron SEM image within AUS8 gold grain ...................................... 99 Figure 32.    Backscatter electron image of biofilm on Prophet Mine gold grain AUS11 ......... 100 Figure 33.    SEM-EDS linescans of AUS4 gold grain ............................................................... 101 Figure 34.    Top Left: Composite SEM-EDS elemental map of AUS16 gold grain.                  Top Right: Backscatter electron image of the same gold grain.                        Bottom Right: Images of individual element signals ............................................ 103 Figure 35.    SEM-EDS linescan of margin of COL1 gold grain................................................ 104 Figure 36.    SEM-EDS linescan across Piaba nugget ................................................................ 105 Figure 37.    SEM-EDS linescan of BC3 gold sample ................................................................ 106  xii Figure 38.    Common axis box plots of all trace element data by locality................................. 111 Figure 39.    Selected bivariate plots of trace elements in placer gold versus Au ...................... 112 Figure 40.    Ranked probability plots of select trace elements .................................................. 113 Figure 41.    Bivariate Fe vs. Cu plot .......................................................................................... 115 Figure 42.     Bivariate Fe vs. Cu plot. Log-log scale ................................................................. 116 Figure 43.    Bivariate Fe vs. Pt plot. Log-log scale ................................................................... 117 Figure 44.    Bivariate Fe vs. V plot ............................................................................................ 119 Figure 45.    Bivariate Fe vs. Se plot ........................................................................................... 120 Figure 46.    Bivariate Pd vs Sb plot ........................................................................................... 122 Figure 47.    Bivariate log-log Pd vs Pt plot................................................................................ 123 Figure 48.    Bivariate Pt vs Hg plo............................................................................................. 124 Figure 49.    Analyses plotted on a Au-Ag-Cu ternary diagram ................................................. 125 Figure 50.    Analyses plotted on a Au-Ag-Cu ternary diagram similar to Figure 49                   but with Au values divided by 10 to aid in showing the Cu trend. ....................... 126 Figure 51.    Analyses plotted on a Ag-Au-Hg ternary diagram ................................................. 127 Figure 52.    Fe-Cu-Hg ternary diagram ..................................................................................... 128 Figure 53.    MDS plot comparing all samples ........................................................................... 130 Figure 54.    MDS plot showing the northern B.C., Nus River, and Piaba nugget analyses ...... 131 Figure 55.    MDS plot of the Prophet Mine sample set. ............................................................ 132 Figure 56.    Bivariate Se vs Sb plot of Prophet Mine sample set separated into                  inferred hypogene and supergene subsets ............................................................. 135 Figure 57.    Chart of signal vs. RSD  comparing University of Windsor GLIER                       Lab reference material data with data collected for this project ........................... 139  xiii Figure 58.    Reproduction of Figure 16 with field of NIST610 reproducibility shown           within red dashed lines. Black dashed lines define homogeneity            classifications of AuRM2 ...................................................................................... 141 Figure 59.    Elemental MDS plot of NIST610 analyses ............................................................ 144 Figure 60.    Gold grain collected from Mission Creek exhibiting partial reddish-orange              to purple rim .......................................................................................................... 154                   xiv List of Equations Equation 1.    Standardizing formula used for preprocessing of geochemical data for MDS ...... 20 Equation 2.    RSD calculation used to measure variance of elemental concentrations        resulting from multiple analyses of individual standard reference materials. ....... 51 Equation 3.    Formula for semi-quantitative determination of trace element               concentrations in AuRM2 reference material . ..................................................... 62 Equation 4.    Instrument response formula. ................................................................................. 63 Equation 5.    Predicted repeatability of LA-ICP-MS analyses using counting statistics             and based on a Poisson distribution. ..................................................................... 72 Equation 6.    Calculation of heterogeneity in AuRM2 negating the effect of counting       statistics which control repeatability of analyses .................................................. 73 Equation 7.    Linear slope equation from Figure 44 showing the relationship between                Fe and V abundances for most samples. ............................................................. 118 Equation 8.    Linear slope equation from Figure 44 showing steeper V-rich linear        relationship for 5 samples from Winfield and Prophet Mine. ............................. 118          xv List of Abbreviations Accept.  Accepted concentration value Accur.   Accuracy parameter B.C.   British Columbia  BSE   Back Scatter Electron cps   Counts per second EPMA-EDS  Electron Probe Micro-analyzer Energy Dispersive Spectroscopy Ext.   External (standard) FiLTER  Fipke Lab for Trace Element Research g   Gram GLIER  Great Lakes Institute for Environmental Research LA-ICP-MS  Laser Ablation Inductively Coupled Plasma Mass Spectrometer MDS   Multi-dimensional Scaling Meas.   Measured (concentration) oz   Troy ounce PGEs   Platinum Group Elements ppb   Parts per billion ppm   Parts per million RSD   Relative Standard Deviation RSP   Response (of mass spectrometer; instrument specific) SEM-EDS  Scanning Electron Microscope Energy Dispersive Spectroscopy SR-XRF  Synchrotron Radiation X-ray Fluorescence  t   Metric tonne Wt%   Weight percent  xvi Acknowledgements  This research would not be possible without funding from the Natural Sciences and Engineering Research Council of Canada and the support of the University of British Columbia Okanagan. Guidance and direction for this project from my co-supervisors Dr. John Greenough and the late Dr. Robert Kerrich is greatly appreciated along with the other members of the supervisory committee: Drs. Kyle Larson and Yuan Chen. Assistance in collecting and compiling samples for this project was provided by: Dr. Frank Reith, John Fedorowich, Alejandro Velasquez, Mackenzie Plovie, Allan Macdonnell, Garth Macdonnell, James Baker, and Jim Moody. Polishing and mounting of gold grains was performed by Stephen Wood at the University of Western Ontario. Mike Hinds at the Royal Canadian Mint was very helpful in providing access to gold standard reference materials crucial to the success of this project. Contributions of Dr. Brian Fryer and Mohamed Shaheen at the University of Windsor in analyzing reference materials to help determine suitability for this method of analysis are acknowledged. Special thanks are given to Dave Arkinstall at the UBC FiLTER lab for help in SEM and LA-ICP-MS analysis. Funding for the FiLTER lab was provided by a generous donation by Dr. Charles Fipke. CF Mineral Research in Kelowna was also very helpful in providing carbon coating services for samples and aided in polishing gold reference materials.         xvii Dedication To those who stirred me to ask questions and pay particular attention to exceptions to the rule.  1 Chapter  1: Introduction Gold is a fascinating substance; it can be described as an element (Au), a metal, or as a mineral. It is of economic importance worldwide and its unique physical and chemical properties make it extremely useful in scientific and industrial applications. It is a rare element in the Earth‟s crust with an average crustal abundance of around 4 ppb (parts per billion) (Boyle, 1979). Economic concentrations of the metal are rare, with only select geologic, or perhaps even biologic, processes able to concentrate it sufficiently. Natural gold most often occurs in its native form (Boyle, 1979) as free gold grains (often associated with quartz and/or sulfide minerals), micro-inclusions within other minerals (particularly pyrite), or in special circumstances as nano-particulates (Reith et al., 2010). In each of these cases (perhaps excepting nano-particulate form) native gold invariably contains a certain amount of Ag contained in solid solution typically  ranging from 5 wt% (weight percent) (Boyle, 1979) up to 20-80 wt% (called electrum). Less commonly other elements occur in solid solution in native gold at wt% levels including Cu (cuproauride), Hg (gold amalgam), and Pd (porpezite) (Boyle, 1979). Rarely Au may form a mineral other than native gold in the form of a sulfide (ex: uteyenbogaardtite Ag3AuS2), telluride (ex: calaverite AuTe2), antimonide (ex: aurostibite AuSb2) or as a selenide (ex: fischesserite Ag3AuSe2) (Boyle, 1979), but these minerals are not commonly of economic significance. Refractory Au can also often be found in solid solution within another mineral as a lattice constituent (importantly in pyrite and arsenopyrite).    River, stream, creek, and even beach sediment deposits often contain eroded native gold as placer deposits. Placer deposits are defined as detrital mineral occurrences concentrated in surficial environments by physical processes. Many major gold deposits have been discovered by  2 following placer gold occurrences upstream to the bedrock source. However, some placer gold occurrences have elusive source deposits or are associated with sub-economic low grade gold mineralization (Boyle, 1979). Gold particles in placer deposits range in size from flour gold all the way up to nuggets that far exceed the size of native gold grains observed in the bedrock source (Garnett & Basset, 2005). Uneconomic dispersed bedrock gold mineralization may even be collected from a catchment basin and enriched into economic placer gold deposits.  The paradox of gold found in eroded sediments coarser than gold occurring in the bedrock source has been ascribed to various processes including physical amalgamation, chemical precipitation, and biochemical precipitation, under surficial conditions. Secondary gold has also been suggested to be precipitated from organic complexes in such settings as tropical swamps, soils, and lateritic regoliths (Hough et al., 2009).   Many placer gold occurrences have been observed to contain gold nuggets and grains with a highly Au enriched supergene rim (Boyle, 1979). The origin of this enrichment has been contentious; one possibility is that minor alloying elements have been chemically leached leaving the more inert gold. It has been suggested that electro-refining of gold grains may occur where Au and Ag are put into solution due to an Eh potential at the surface of the grain and then almost immediately the more noble metal (Au) re-precipitates on the surface of the gold leaving Ag to be carried away in solution (Groen et al., 1984; McCready et al., 2003). Another theory contends that bacteria living in films on the surface of the grains play a role in precipitating secondary pure gold (Reith et al., 2010).     3 1.1 Previous Work There have been numerous studies utilizing the various physical and chemical characteristics of eroded gold grains to determine their signatures. The end goal of projects using these varying analytical techniques to determine signatures of placer gold is to classify different populations to determine characteristics of the source mineralization (distance upstream to the source, mineralization type, size of mineralization body, stages of mineralization etc.). Researchers have examined gold grain morphology (to estimate transport distance), and mineral or host rock inclusions within the gold (Boyle, 1979).  Mineral inclusion assemblages in placer gold have been used as tracers and signatures for the bedrock source of mineralization (Chapman et al., 2009; Chapman & Leake, 2000; Chapman, 2010; Crawford, 2007; Desborough, 1970; Mortensen & Chapman, 2010). Important minerals for classifying placer gold populations include sufides, sulfarsenides, tellurides, and selenides. These minerals indicate that a placer gold grain has a preserved hypogene core; they also give insight into the composition and characteristics of mineralizing fluids and history of mineralization of the gold. Inclusions inevitably become rarer as distance from the source increases (Chapman et al., 2009). As well, these inclusions tend to be relatively rare features and as such an extremely large sample population (in the thousands) must be taken to properly characterize populations of placer gold. Recently, isotopic signatures of Pb have been used to fingerprint gold in archeological and geologic applications (Kamenov et al., 2013; Standish et al., 2013).   Researchers have tried quantifying trace element concentrations to determine signatures in all types of gold using various techniques. Disciplines interested in this method include archeology, geology, industrial users of the metal, and precious metal refiners. EPMA-EDS  4 (Electron Probe Micro-analyzer Energy Dispersive Spectroscopy) analyses have successfully identified minor alloying components in placer gold samples including Ag, Cu, Pd, and Hg with detection limits ranging from 200 ppm (Cu) to 2,800 ppm (Hg) (Chapman et al., 2009).  Alluvial gold grains in Chile have been linked to their source mineralization using only Au, Cu, and Ag concentrations (Townley et al., 2003). It was found that epithermal mineralizing systems produced gold grains with around 5 wt% Ag and 0.1 wt% Cu. Gold-rich porphyry systems produced grains with higher Ag (8 wt%) and 0.1 wt% Cu. Samples from Cu-Au  porphyry systems were found to have elevated Cu (up to 0.75 wt%) and variable Ag content (Figure 1). In this Chilean example all samples come from broadly the same type of mineralizing system (Au or Cu-Au type porphyry deposits and associated epithermal systems) but major element variation in native gold was still measured within different system components. Comparing these broadly categorized porphyry associated systems with other gold deposit types it appears that they fall on the lower end of Ag contents and the high end of Cu content.  A few other researchers have developed gold composition ternary diagrams similar to Townley‟s (2003; Figure 2). These diagrams can be very insightful but they require extraordinary deposit chemistry to exhibit trace element concentrations detectable by traditional methods. The vast majority of deposits will only have measurable Au and Ag concentrations which limits the conclusions that can be drawn about mineralizing systems.  Quantitative analysis of trace elements at the ppm and sub-ppm level in gold (as opposed to minor alloying components) has proven elusive. A limited number of elements have been quantified by SR-XRF (Synchrotron Radiation X-ray Fluorescence) in archeological applications (Constantinescu & Bugoi, 2008; Guerra et al., 2008). Trace level dopants in gold bonding wire   5  Figure 1. Gold grain Au-Ag-Cu ternary system developed by Townley et al. (2003) to differentiate porphyry associated deposit types. Fields for each deposit type as interpreted by Townley et al. (2003).  6 Figure 2. Various gold ternary diagrams by selected authors. A) Au-Ag-Cu diagram showing data for placer gold grains from Northern Ireland overlain on fields of various other known deposit type chemistries (Moles et al., 2013). B) Phase diagram for Ag-Au-Hg showing position of solid solution, solid solution + liquid, and liquid fields at 600ºC and 650ºC with gold grain compositions from Hemlo deposit plotted. This diagram was used to show effects of partial melting on an existing gold deposit (Tomkins et al., 2004). C/D) Te-Au-Ag and Te-Cu-Ag ternary diagrams of gold samples from epithermal gold mineralization in Romania. This study linked unique gold signatures to archeological gold artifacts (Pop et al., 2011). E/F) Cu-Au-Ag and Cu-Au-Pd ternaries for precious metal alloy grains from the Skaergaard layered mafic intrusion, East Greenland (Bird et al., 1991).  7 have been determined by inductively coupled plasma atomic emission spectrometry (ICP-AES) for industrial applications (Choi et al., 2008). Solution inductively coupled plasma mass spectrometry (Solution ICP-MS) has been successfully applied to quantifying select trace elements in placer gold (McInnes et al., 2008). Issues with solution ICP-MS analysis of placer gold samples, however, include the necessity of dissolving whole gold grains in aqua regia including any supergene rims and non-silicate inclusions contained within.  The technique most employed to determine trace element concentrations in placer gold is laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) (Outridge et al., 1998; Penny, 2001; Brostoff et al., 2009; Constantinescu & Bugoi, 2008; Crawford, 2007; Guerra et al., 2008; Langmuir, 2008; McCandless et al., 1997; Netshitungulwana et al., 2010; Standish et al., 2013). The development of LA-ICP-MS has allowed for in-situ analysis of samples with the ability to discriminate spatial variation of trace element concentrations at the sub-millimetre scale (McCandless et al., 1997). This method has met with limited success when applied to the analysis of native gold specimens, however, and has not been utilized to its full potential. The main issue that arises when applying this technique to gold analysis has been the lack of a suitable external gold reference standard material that is homogenous on a microscopic scale. Such a standard requires a sufficient number, and abundance, of trace elements at known concentrations. This external calibration is required to determine absolute concentrations of trace elements that are reproducible between instruments. There have been studies analyzing gold samples without external calibration that produced recognizable „fingerprints‟ of trace elements using raw counts from individual instruments. These gold analyses prove the presence of many trace elements at detectable levels within natural gold samples (Brostoff et al., 2009;  8 Constantinescu & Bugoi, 2008; Crawford, 2007; Guerra et al., 2008; Langmuir, 2008; McCandless et al., 1997; Netshitungulwana et al., 2010; Outridge et al., 1998).  Part of the significance of this research is that placer gold grains can carry the signature of the bedrock source gold mineralization (Chapman et al., 2009). Some placer samples exhibit an altered supergene rim but the cores of the grains may retain the hypogene signature of gold mineralization. This means that when gold grains are collected during a regional stream sampling survey for a mineral exploration program the unaltered gold signature can be characterized. Regional stream sediment and glacial till sampling are extremely useful techniques commonly used for mineral exploration (McInnis et al., 2008). This is because a wide geographic area can be represented from one sample. If a database were made to geochemically classify known gold deposit types, the chemical signature could be used to target geologically prospective locations within the stream catchment basin (or up-ice direction in the case of till samples) for follow up exploration. Additionally, geochemical variation may be present between gold grains that come from economic mineralizing systems versus uneconomic systems which would allow for targeting of prospective mineral prospects at a regional scale (McInnis et al., 2008). Significant heterogeneity in trace element signature may also be observed in the unaltered cores of placer gold grains. This is significant because it may document the mineralization history and stages of deposit development at a finer, detailed, level.  1.2 Objectives There are three objectives to this research. The main goal is to use the trace element composition to fingerprint hypogene and supergene placer gold populations in order to elucidate  9 their origin. Another goal of this project is to begin to catalogue the unaltered placer gold signature based on the source deposit-type and create a database useful for determining the possible deposit-type of placer gold that has an as yet unknown source. A third goal is to determine if it is possible to characterize supergene placer gold using trace element concentrations in order to resolve the debate about how it forms.   1.3 Project Samples Samples obtained or collected to meet the objectives of this project include 51 placer gold grains from seven different localities. The samples come from three different continents, are sourced from a wide variety of bedrock mineralization styles, and were found in placer systems with varying surficial climate conditions. The purpose of the broad scope of samples is to recognize the differences that may arise from different source deposit types and from different surficial supergene environments resulting from disparate climatic settings. Some samples were obtained because they have been well documented by other researchers providing a large amount of background information. This is important because the relative certainty of their origins can serve as a reference for other samples.  The largest gold sample for this study is a 3.1 gram nugget from northern Brazil. It was dug out of lateritic soil (possibly illegally by artisanal miners) from less than 2 meters depth. It was provided by John Fedorowich who was working at the nearby Piaba orogenic gold deposit which is thought to be the source for the laterite gold. This area is interesting because much of the mobilized gold is apparently not in hydraulic equilibrium with the sediments it is found in. Placer gold accumulating due to physical processes can be expected to be found in association with certain sedimentary environments; this occurrence has been suggested to not have this  10 expected spatial distribution (i.e. out of hydraulic equilibrium).  The gold therefore may be secondarily precipitated in supergene settings chemically, or biochemically. The climate in this region is tropical.  Sixteen gold grains were obtained from Dr. Frank Reith (University of Adelaide) and were sourced from the Prophet placer gold deposit in Australia. The average size of these grains is ~2 millimetres in the largest dimension. They are very important for the research project as they have been convincingly documented as having supergene gold rims that are precipitated exclusively by bacteria living as biofilms on the placer grains (Reith et al., 2010). The bacteria in this case have the ability to take gold cyanide complexes in solution that are toxic to them and nucleate very pure nanoparticulate-form native gold intra-cellularly. This ability has apparently been evolved by bacteria living in these extreme environments as a defensive mechanism. Gold may be dispersed from these bacterial biofilms if they are disrupted (perhaps by mobilization of the gold grain), or they may amalgamate gold onto the placer grain (Reith et al., 2010).  Numerous mostly sub-millimetre scale placer gold grains were collected from the Nus River in Colombia; of these, 8 were prepared for further analysis. These placer grains have likely been eroded from AngloGold's Gramalote gold deposit upstream. The source deposit is an intrusive-hosted, structurally controlled, stockwork gold-silver deposit. This deposit is currently being mined and is extensively documented in NI-43-101 technical reports (Gustavson Associates, 2012). The climate in the region of the Gramalote deposit is classified as 'mildly tropical' at elevations around 2,000 meters above sea level.  Samples were collected by the author from three different placer and paleoplacer gold occurrences in the Kelowna area. At least 9 grains were recovered from each of Mission Creek,  11 Lambly Creek, and the abandoned Winfield Mine. The largest of the gold grains were slightly over one millimetre in the longest dimension.  Mission Creek drains into Lake Okanagan from the east and was mined for placer gold in the late 1800s (Roed et al., 1995). The rusty, possibly interglacial, gravels (part of the Rutland aquifer) that outcrop on the banks of Mission Creek have been suggested as the source of gold. Lambly Creek drains into Lake Okanagan from the west and has minor occurrences of placer gold. Samples were collected from an abandoned paleoplacer gold mine near Winfield. It is located up the side of the mountain east of Wood Lake and was mined sporadically in the 1930s. The basalts overlying the host gravels are currently quarried at this location. The climate of the Kelowna area is temperate and semi-arid.  The preceding sample notes are a basic overview of localities sampled for this project. Further detail, including geologic settings, district geology, mineralization sources, timing, etc. will be discussed in Chapter  3: Geology and Petrography of Gold Samples.   12 Chapter  2: Methodology Methods used for this project were chosen to meet the end objective of the research: quantifying trace element concentrations in placer gold to characterize the processes that formed the gold and ultimately determine its source. In approximate temporal order, the methods detailed below consist of: collection of placer gold, characterization of samples, preparation of samples, petrographic analysis, SEM-EDS (Scanning Electron Microscope Energy Dispersive Spectroscopy) major element analysis, and finally LA-ICP-MS trace element analysis.   2.1 Sample Collection The first requirement for this project was, naturally, obtaining placer gold samples to analyze. The samples, briefly described in the introduction, were obtained or collected over a period of several months. Criteria used in assembling a sample set included: documentation of samples, diversity of sample locations, economic significance, and in some cases accessibility. Samples from exotic locales were collected by other individuals and include gold grains from: Prophet Mine Australia, Nus River Colombia, and a gold nugget from Piaba Brazil. All samples were collected from alluvial and/or fluvial sediments, mostly by hand panning sediment directly or hand panning sediment concentrate derived from sluicing. Samples from the Prophet Mine, Australia were initially collected for a research project by Reith et al. (2010) to examine the possibility of biofilms living on gold grains and actively precipitating gold. Collection methods were undertaken to preserve the biofilms intact. Each grain was rinsed with and stored in a 0.9 wt% NaCl solution, transported over ice, washed in sterilized doubly de-ionized water, and air dried.   13 Samples collected by the author, mainly during the summer of 2012, include placer gold grains from Mission Creek, Lambly Creek, and the Winfield Mine, all in the Kelowna area. Gold samples from Mission Creek and Lambly Creek were hand panned from recent fluvial sediments within 50 centimeters of the surface of the sediment with some samples taken from sediments directly overlying bedrock. Seven gold grains were collected from Mission Creek and ten were collected from Lambly Creek. The samples from the abandoned mine near Winfield were hand panned at the entrance to an approximately 30 meter drift into the side of the hill where loosely consolidated gravels overlay metamorphosed basement rocks. Seven gold grains were collected from this location.   In all cases sediment was panned down to a heavy mineral concentrate using traditional panning methods. Extreme care and patience was used while panning to ensure that even very small gold particles could be recovered (<50 microns). Typically it is not economically worth the effort to recover these small particles, however, for the purposes of this study, it was important to provide an accurate representation of the grain size of the gold population. Once the heavy minerals were concentrated, gold grains were isolated and recovered from the gold pan using adhesive tape. The gold pan was thoroughly cleaned between collection sites to avoid cross-contamination of samples.   2.2 Sample Characterization Samples were organized, catalogued, and whole grains were photographed under microscope. Roundness, particle size dimensions, and grain morphology were recorded. Other aspects such as surface texture and colour were also noted. As there were many hundreds of  14 placer gold grains from Nus River, Colombia, 8 representative samples were isolated for further analysis.  A geography-based sample naming convention was established early on and used consistently throughout the project to avoid confusion. The single gold nugget from Piaba, Brazil was labelled PIABA. Samples from the Prophet mine, Australia were labelled AUS1 through AUS16. Gold from the Nus River Colombia was labelled COL1 through COL8. Winfield Mine samples were assigned WIN1 through WIN7, Mission Creek samples MIS1 through MIS7, and Lambly Creek samples LAM1 through LAM10. In addition three gold grains from an unknown locality in northern British Columbia were assigned BC1 through BC3.  2.3 Sample Preparation Catalogued and organized gold samples were sent to the thin section lab at the University of Western Ontario to be mounted in resin on glass discs and polished with diamond paste. Gold grains were oriented to obtain a cross section perpendicular to the long axis. The large Piaba nugget was sectioned across the long axis and then sectioned again along the short axis. Large samples could be placed together in multiples on a single disc and polished down to the same level. Small gold grains, however, required mounting on individual discs to avoid completely polishing away grains. Some samples were lost during preparation due to their small size and were either polished away or failed to stick in the resin. Of the 52 original gold samples 41 (listed in Appendix A) were successfully prepared for further analysis.    15 2.4 Petrographic Analysis Polished gold grains were examined under reflected light with a petrographic microscope (Nikon Eclipse 50 iPOL) and systematically described. Descriptions include observation and identification of mineral inclusions and heterogeneity of gold alloy composition within gold grains. Heterogeneity of gold alloy composition could be visibly observed under reflected light with lighter yellow sections interpreted as Ag rich and deeply coloured yellow sections Ag poor. These different components of the samples were quantified by modal abundance. Grain shape, morphology and rough cross sectional area were estimated as well.   Any noted mineral inclusions were classified as being likely hypogene inclusions or sediment inclusions picked up during gold grain transport. Features such as folded edges, rough surfaces, and embayments were noted.    2.5 SEM-EDS Analysis After thorough description samples were carbon coated at CF Mineral Research in Kelowna in preparation for SEM-EDS. The carbon coating had an average thickness of ~20 nanometers. SEM-EDS analysis was performed at the University of British Columbia FiLTER Lab in Kelowna. The instrument used was a Tescan Mira3 XMU Scanning Electron Microscope set up with an Oxford Aztec X-max EDS system hosting an 80 mm2 silicon drift detector. Processing of the SEM-EDS data was completed using Oxford Aztec software.  Quality control methods applied during SEM-EDS analyses included duplicate analyses, daily optimization of the beam and instrument parameters using a metallic copper strip, and measurement of known standard reference materials. Preliminary analyses on placer gold grains were made by measuring the exact same area of a gold grain three times. These analyses yielded  16 essentially identical results; as such duplicate analyses were not performed for subsequent runs. Parameters used for analysis are listed in Table 1. The standard reference material used was MAC 80Au - 20Ag. Its composition is 80 wt% Au and 20 wt% Ag; it was chosen because it is the standard with the closest average concentrations to the placer gold samples for this project. For five analyses of the standard the RSD (relative standard deviation) for Au values was 0.44% and Ag values had an RSD of 1.74%. The SEM-EDS method allowed for quantitative elemental concentration data for minor alloying elements and inclusions within the gold grains; the detection limit for most elements was around 0.1 wt%. A total of over 200 SEM-EDS analyses were performed on 41 gold grains. Techniques for measuring the element data included spot samples (172 analyses), line scans (23 analyses), and two dimensional element concentration maps (5 analyses) of the samples. A minimum of two spot analyses were made per gold grain with at least one measurement of the core and one of the rim. Line scans were performed where a supergene rim or heterogeneous alloy was suspected to mark elemental variations. Select gold grains were entirely mapped out by SEM-EDS measurements to show elemental zonation of grains. All observed inclusions (either primary or detrital) were analyzed as well.  SEM analysis was critical for determining silver concentrations in the gold because it was used as an internal standard for LA-ICP-MS analysis. It was also useful for mapping out the supergene gold rich rims on the samples so that points could be picked confidently for subsequent LA-ICP-MS analysis.              17 Table 1. Instrument parameters used during SEM-EDS analysis of standards and placer gold samples SEM-EDS Parameters  Livetime: 14.0s Process Time: 4 Accelerating Voltage: 20.00kV Magnification: 375 x Working Distance: 14.0mm Specimen Tilt (degrees): 0.0 Elevation (degrees): 30.4 Azimuth (degrees): 0.0 Number Of Channels: 2048 Energy Range (keV): 20 keV Energy per Channel (eV): 10.0eV Detector Type: X-Max Window Type: SATW Pulse Pile Up Correction: Succeeded                              18 2.6 LA-ICP-MS Analysis LA-ICP-MS analysis was executed with a Photon Machines Analyte 193 EXCIMER laser coupled with a Thermo Fisher Element XR ICP-MS on July 4th to 5th, 2013 at the FiLTER lab, University of British Columbia Okanagan. Thirty-one different isotopes were measured in low resolution mode during 92 individual analyses on 41 placer gold grain samples from the 5 different localities. Instrumental parameters for the experiment are outlined in Table 2. The triple detector mode was used for all elements except Au which was detected only in Faraday mode to avoid undue wear on the detectors resulting from high counts.  All samples and standards analysed by LA-ICP-MS went through an ultrasonic bath in de-ionized water for one hour to remove surface contamination. Sampling consisted of laser ablation pits with a depth to pit diameter ratio never exceeding 0.5 to avoid elemental fractionation effects (Potts et al., 1995). GLITTER software was used for data reduction with Ag values provided from SEM-EDS analysis used as an internal standard. The external standard reference materials used for quantification included NIST610 glass, London Bullion Market Association‟s AuRM2 gold reference material, and the Royal Canadian Mint‟s FAU7 gold standard. Two analyses of each of the three reference materials were performed at the start and at the end of each run for a total of 12 standard analyses per run; a total of 8 runs were performed.  The primary external standard used during analysis was AuRM2 which is critical for the research project. It had not previously been proven suitable for this method of analysis. A detailed statistical analysis of the accuracy, precision, and homogeneity of this standard is provided in Section 4.1 and represents an important contribution of this work.        19 Table 2.  LA-ICP-MS instrument parameters for trace element analysis.   1  Laser beam size reduced to 64.1  for last two runs due to smaller size of LAM and WIN gold grains.                                            20 2.7 Multi-Dimensional Statistical Methods The exploratory statistical technique MDS (Multi-dimensional Scaling) was applied to geochemical data collected for this study. This statistical method was chosen to uncover populations of samples and examine the association of elements within samples in the data-set.         Multi-dimensional scaling is a powerful tool for data reduction. A whole suite of geochemical data, with as many elements as analytically possible, can be represented in one single graph. Essentially this technique characterizes dissimilarities. This can either be the dissimilarities of elements within a sample set or dissimilarities between groups of samples. Most MDS plots have 2 dimensions however it is also possible to create 3d plots or matrices of 4 or 5 dimensions. The scale of each axis is defined as dimension 1 and dimension 2; these are unit-less and simply represent bulk differences. Elements that follow each other in natural processes tend to cluster together in MDS plots (ex: all chalcophile elements may plot together). These plots can characterize different deposits, metallogenic provinces, deposit types, host rocks, and numerous other characteristics that may control the distribution of trace elements.   The procedure used to create the multidimensional scaling plots involved significant preparation of the original geochemical data. First the data was standardized by Equation 1:                                             Equation 1. Standardizing formula used for preprocessing of geochemical data for MDS  where   is the mean of trace element concentration values and  is the standard deviation. This step dampens the effect of the range in concentrations of different elements. For example Fe may have an orders of magnitude greater abundance than Pt. Next the data were correlated by the Pearson‟s correlation coefficient: r   which measures the similarities between the bivariate plots  21 of each component to the others. The r values were then used for multi-dimensional scaling in the Systat™ statistical program. The program creates the dimensions for the plots by measuring the input distance (r values) which creates a plot showing relative differences or similarities between each of the components.   For all of the plots a Kruskal loss function was used. The R-metric was 2, there were ≤50 iterations, and a convergence value of 0.005 was used. Most plots were created with 2 dimensions, however, if there was significant scatter in the data, plots of 3 or 4 dimensions were attempted to fit the data.                   22 Chapter  3: Geology and Petrography of Gold Samples 3.1 Introduction to Geology and Petrography  The objective of the study is to describe the overall trace element characteristics of, and chemical variation between, placer gold from a number of localities world-wide. A better understanding of the regional geology and thorough petrographic descriptions of the gold grains provides a better basis for interpretation of trace element results. The world-wide results can be placed in perspective by a detailed study of one area, in this case the Kelowna area. This provides insights into local chemical variability. This chapter discusses the geology and petrography of Kelowna area placer gold, and then presents an overview of equivalent information for the placer gold samples from Australia, Colombia, and Brazil.  3.2 Kelowna Area Placer Gold Geology and Petrography 3.2.1 Overview Gold samples from the central Okanagan, Kelowna area, provide a large sample set of gold grains from a spatially restricted area (10s of km) to elucidate the potential extent of chemical variation in placer gold at the district level. Samples come from three localities: Mission Creek, Lambly Creek, and a paleoplacer deposit at Winfield, representing the main known placer gold occurrences in the Kelowna area (BC MinFile). Proximity of these sites to UBC Okanagan allowed for extensive field work by the author on a limited budget. Geological mapping for characterizing central Okanagan geological environments is more accessible than information available for placer samples from the “world-wide” settings used in the study (e.g. Nus River, Columbia; Piaba nugget Brazil). After reviewing historical placer gold production in  23 the area, the regional geological setting and bedrock geology are presented. Next, the local geologic setting at each sampling site is reviewed. Finally, the petrography of the samples is described.  3.2.2 Historical Production Placer gold has historically been of economic importance in the central Okanagan region, specifically the Kelowna area, starting with the discovery of gold at Mission Creek in 1861 (Roed et al., 1995). Records of placer gold mining at Mission Creek begin in 1876 and mining continued until the 1930‟s (BC Ministry of Energy and Mines, 1996b). Mining methods included both open pit and underground operations. During the peak of mining activity (1876-1895) the recorded production, which likely represents a minimum value due to underreporting, is listed at 20.56 kilograms of gold recovered. Eight overburden drillholes tested Mission Creek in 1975 and results were considered very promising yet inconsistent and no further work was conducted (BC Ministry of Energy and Mines, 1996b). Today, the area is encapsulated by the Scenic Canyon Regional Park. The only other placer occurrence in the project area with reported production is the Winfield Mine several kilometers east of Wood Lake. This paleoplacer occurrence was discovered in 1936 and two exploration drifts of 52 and 107 meters were dug. Small scale mining continued until 1945 yielding a total reported 2.33 kilograms of gold out of 1 tonne of ore giving an apparent grade of 75 oz/t (ounces per tonne) (BC Ministry of Energy and Mines, 1993i). In addition, the same sediments that host the paleoplacer gold at Winfield were explored by Union Oil Company for uranium from 1976-1979. Currently at this location, the Chilcotin Basalt, which immediately overlies the paleoplacer gravels, is being quarried for aggregate.   24  3.2.3 Geological Setting of Kelowna Area  The geologic setting of the project area is markedly complex. The study area is located along the division between the Intermontane Belt and the Omineca Belt of the Canadian Cordillera (Monger & Price, 1979) (Figure 3). To the east are Proterozoic and Paleozoic strata of the Omineca Belt that were intensively metamorphosed during the Mesozoic, and to the west are Paleozoic to early Mesozoic arc-related rocks of the Intermontane Belt.   The Omineca Belt occurs east of a suture zone formed during the Mesozoic, between the Precambrian to Paleozoic age continental margin sequence on the western edge of the North American craton (Foreland Belt) and the accreted Slide Mountain and Quesnellia exotic terranes (Intermontane Belt) (Nesbitt & Muehlenbachs, 1995). The Omineca belt formed due to parallel subduction of the Intermontane oceanic plates, subsequent closure of the Slide Mountain Ocean, and accretion of an allocthonous volcanic island arc (Figge & Townsend, 2002) (Figure 4). Numerous post-accretion mid-Jurassic, and Eocene age felsic plutons have intruded along the contact between the Intermontane and Omineca Belts (Armstrong, 1988; Gabrielse & Yorath, 1999).  Structurally, the project area underwent compression during the Mesozoic, strike-slip faulting during the Mesozoic to Paleogene, and extension in recent times (Nesbitt & Muehlenbachs, 1995). Deformation began in the Early to Mid-Jurassic and involved easterly directed thrusting, obduction of oceanic crust, plutonism, and associated metamorphism. A second major metamorphic and plutonic event occurred during the mid-Cretaceous to Paleogene (Nesbitt & Muehlenbachs, 1995).    25  Figure 3. Belts of the Canadian Cordillera. (modified from Digital Geology of Idaho, Idaho State University; Mitchell et al., 2013)  26  Figure 4. Schematic representation of the evolution of the Canadian Cordillera just prior to the accretion of the Intermonane Belt in the Early Jurassic (195 Ma). (From Burke Museum, The Omineca Episode). Missing in this diagram is the extent of the lithospheric mantle.       27  The change to an extensional tectonic regime occurred during the Paleocene-Eocene; the region was affected by east-west ductile to brittle extensional faults. This extension exposed large areas of amphibolite grade orthogneisses and paragneisses (Madsen et al., 2006). An episode of recent uplift of up to 1,000 meters has been suggested to have occurred starting around 2 Ma (Boyle, 1982).   3.2.4 Bedrock Geology of Kelowna Area  The geologic complexity of the area is apparent in the bedrock geology map (Figure 5). The east side of Okanagan Lake is dominated by the Proterozoic to Phanerozoic, predominantly amphibolite grade, metamorphic units of the Omineca Belt (in purple; PrPzShm, PrPzog). On the west side, dominantly in the Lambly Creek (Bear Creek) catchment centre-left of the map, Paleozoic volcanic arc related rocks of the Intermontane Belt are present including: fine grained siliciclastics (DTrHsf; pale blue), volcaniclastics (CPH; medium green), and a small greenstone belt (PCgs; dark blue). The apparent discontinuity along the latitudinal line of the map reflects the merger of two separate maps, and multiple interpretations of the bedrock geology.  Extensive post-accretion plutons (including the Okanagan Batholith) intruded the area during the Middle Jurrasic. They are exposed mostly in the northwest of the map area (MJgd and MJOgd; pink). Eocene age alkali feldspar granite intrusions are also exposed (Egr; red) along with associated volcanic and sedimentary assemblages (EPeMK; tan, EPeK; brown) which form extensive cover rocks on both sides of Okanagan Lake. Remnants of the extensive Miocene to Pliocene Chilcotin Group basalt (MiPiCvb; light green) also act as cover rocks. The youngest rocks in the project area are the Pleistocene Lambly Creek basalts (Roed et al., 2013). These  28 basalts are the product of very recent volcanism indicating that the region is by no means geodynamically stable.   Structural controls in the project area are dominated by the recent, extensional, tectonic regime (Nesbitt & Muehlenbachs, 1995). By far the dominant structural feature in the study area is the terrane-bounding, normal, Okanagan fault which is not exposed in the area but underlies the valley and Okanagan Lake and separates amphibolite grade metamorphic core complexes and volcanic arc units.  Secondary structures in the area include the Mission Creek oblique (normal/strike-slip) fault which runs northeast to southwest in the southern part of the map area near the Mission Creek sample site. This fault may be responsible for the lateral offset of Okanagan Lake at Kelowna (Roed et al., 1995). Various other north-south sub-parallel normal faults are also present in the area.              There are three placer/paleoplacer gold occurrences in the study area highlighted as gold stars in Figure 5. The Mission Creek and Winfield Mine occurrences to the east are on Omineca Belt rocks whereas the Lambly Creek placer occurrence (labeled Bear Creek on Figure 5, the name by which it is also known) is located on Intermontane Belt units. A major difference is that both Mission Creek and the Winfield Mine have been economically exploited whereas the Lambly Creek occurrence has not.    29  Figure 5. Bedrock geology map of the central Okanagan (modified from B.C. Geological Survey 2005 map). Yellow stars indicate field sites sampled for this project. Numbered locations are select MinFile gold mineralization occurrences in the project area that may represent sources of placer gold.     30 Winfield Mine The Winfield paleoplacer camp consists of the mine sampled for this project, along with similar numbered occurrences extending to the northeast, including the Stuart, Aitkens/Stables, and Ribbleworth paleoplacer occurrences (Figure 5). All of these are hosted in Miocene age unconsolidated conglomerates overlying basement rocks predominantly made up of orthogneiss and overlain by Miocene to Pliocene age Chilcotin Group basalt (BC Ministry of Energy and Mines, 1993i).  The paleoplacer host rocks are thought to represent sediments deposited in extensive, structurally controlled, Miocene paleo-river valleys (Boyle, 1982). These sinuous channels can be up to a kilometer wide and are intermittently preserved in the study area from the Winfield Mine in the north down to Hydraulic Lake (off map area in Figure 5, southeast of Mission Creek) with paleoflow indicators showing flow from northwest to southeast. Preservation of these sediments is controlled by capping Chilcotin Basalt which prevented erosion during Pleistocene glaciations. Paleo-climate studies indicate that during deposition of these sediments the region experienced temperate conditions with a moderate amount of uniformly distributed precipitation (Boyle, 1982).  These same Miocene sedimentary units serve as host rocks to significant basal-type uranium mineralization (Boyle, 1982) including the Tyee deposit located near Hydraulic Lake in the extreme southeast (Figure 5). Uranium mineralization is interpreted to be post depositional, and specifically post Chilcotin Basalt, and resulted from leaching of U out of intrusive basement rocks by meteoric water which infiltrated the Miocene sediments through the structures that controlled paleochannels. These oxidised U-bearing fluids encountered reduced organic matter in the sediments depositing mainly urano-phosphate complexes.      31 Mission Creek The Mission Creek placer deposit is the most productive gold deposit in the central Okanagan.  It is hosted in active fluvial sediments of Mission Creek and recent bench deposits of the same fluvial system. The richest placer abundances are found twelve kilometers east of the mouth of Mission Creek, and located just downstream of an exposure of conglomerate that has been postulated as the source of the placer gold (BC Ministry of Energy and Mines, 1996b; Roed et al., 1995). The conglomerate has been interpreted as an interglacial alluvial deposit (BC Ministry of Energy and Mines, 1996b), however the only age constraints are that it is between Eocene to pre-Holocene in age. Clasts are closely packed and mostly well rounded including granitic, dioritic, and argillaceous clasts (BC Ministry of Energy and Mines, 1996b). It is mostly lithified with some friability and extensive limonitic alteration giving it a strikingly rusty appearance (Figure 6). Roed et al. (1995) suggest that this conglomerate represents part of the Rutland Aquifer.        32 Figure 6. Weakly lithified gravels outcropping adjacent to Mission Creek. They have a distinct reddish color due to cementing hematite and have been postulated as the source of Mission Creek placer gold.        33 The paleoplacer conglomerate is underlain by sandstone/siltstone, and dacite of the Eocene Penticton Group. At this location the conglomerate is overlain by till deposits. In the upper reaches of Mission Creek the conglomerate unit is conformably overlain by a dark volcanic unit which is suggested to be related to the Pleistocene Lambly Creek basalt (BC Ministry of Energy and Mines, 1996b). To the south, the Mission Creek fault juxtaposes the Shuswap Metamorphic Core Complex. Lambly Creek Minor placer gold occurrences are present along Lambly Creek but there has been no significant economic exploitation. Host rocks in the creek catchment include faulted Paleozoic volcaniclastics, fine grained siliciclastics, and a small volcanic arc related greenstone belt. Overlying these are Eocene Penticton Group volcanic rocks and the Pleistocene Lambly Creek basalt. A number of bedrock mineralization occurrences, of various types, occur within the drainage basin of the creek (Figure 5).  The Bond Au-Ag vein showing, discovered in 1985, is hosted within a possible greenstone belt and is classified as epithermal (BC Ministry of Energy and Mines, 1993c). It is hosted in decimeter-scale quartz veins and grab samples assayed up to 12 g/t (grams per metric tonne) gold and 7 g/t silver. Fine grained pyrite is associated with the vein along with surface weathering yielding limonitic alteration. The Shear occurrence consists of polymetallic stockwork veins containing Ag-Pb-Zn ± Au (BC Ministry of Energy and Mines, 1996c). Assays up to 18.5 g/t Au are recorded and associated minerals include galena, pyrite and quartz. The Zumar occurrence is another polymetallic stockwork type vein showing similar to Shear but also containing Cu, and showing hematite and sericite alteration (BC Ministry of Energy and Mines,  34 1993j). The Lamb occurrence is possibly a skarn-related Ag-Cu ± Au showing with bulk assays up to 2.4 g/t Ag and 0.19% Cu.      To the northeast of Lambly Creek is the Spod low sulphidation gold mineralization discovered in 1989 (BC Ministry of Energy and Mines, 1996d). It consists of multiple stages of andesite-hosted quartz veining. The alteration assemblage includes propylitic alteration and zones of strong silicification. The best drill intercept was 3 meters of 0.785 g/t Au. In the far northwest of the study area are minor Cu-Mo alkalic porphyry related mineralization occurrences including the AT occurrence (BC Ministry of Energy and Mines, 1993b).   3.2.5 Petrography and Gold Grain Morphology of Kelowna Area Samples Gold grain morphology does not constitute a major component of this project; however, a cursory examination of key morphological features of the gold provides valuable information. Petrography was an important step in determining bulk scale heterogeneity of gold grains (changes in gold fineness and mineral inclusion assemblage). Details of both petrographic and morphological descriptions can be found in Appendix A. Of the Kelowna area placer occurrences, Lambly Creek has the smallest average gold grain size of 0.18 mm in the longest dimension with a range of 0.04 to 0.5 mm. Generally grains are abraded with moderately smooth to well rounded edges. Many grains exhibit embayments and folded edges; most have fine grained detrital inclusions embedded in the surface. The Lambly Creek gold samples range in colour from mainly light yellow to slightly yellow-orange near the rim (Figure 7). In one case several <10 micron grey-pink anhedral sulfide inclusions, perhaps pyrrhotite, were observed in the core of the grain (Figure 8). In addition several large either primary or detrital quartz inclusions were observed.       35                         Figure 7. LAM2 gold grain from Lambly Creek. Grain is 0.5 X 2 mm. Left side is unpolished, right side is after being mounted and polished. Faint Au enriched rim visible.   36       Figure 8. LAM8 gold grain from Lambly Creek. Grain is 0.25 X 0.15 mm. Left side is unpolished, right side is polished grain mount. Red arrow points to primary sulfide inclusion (pyrrhotite?). Note the thin Au enriched rim.  37 The largest gold grains were recovered from Mission Creek where the average longest dimension is 0.62 mm, several multi-millimeter flakes were collected. All grains are extremely flattened and most are smooth and well rounded. Detrital inclusions are limited, and no primary inclusions were observed. Colour is mostly light yellow but some grains exhibit deeply coloured Au-rich rims up to 20 microns thick (Figure 9).  The average longest dimension of gold grains from Winfield mine was 0.49 mm but there is a wide range of sizes from very fine flour gold to flakes well over a millimeter. Edges are subrounded to smooth with evidence of folding, and detrital incorporation. Gold colour is light yellow with thin yellow-orange rims and rare red-orange rims (Figure 10). No definitive primary inclusions were found, but one grain has a subhedral magnetite grain near the core that could be hypogene in nature. Interestingly, the heavy mineral concentrate these grains were found in contains little or no black magnetite sand, but instead contains abundant garnet.             38       Figure 9. MIS2 gold grain from Mission Creek. Grain is 1.2 X 0.6 mm; left side is unpolished, right side is mounted and polished. Note thick deeply coloured Au rich rim on grain. The Au rich rim has an irregular contact with the core of the grain.   39          Figure 10. WIN5 gold grain from abandoned Winfield Mine. Grain is approximately 0.45 X 0.25 mm. Left image is unpolished; note slight red-orange tinge on parts of grain surface. Right is after polishing; thin Au enriched rim present, arrow points to magnetite grain possibly primary in origin.     40 3.3 Geology and Petrography of “World-wide” Assorted Sites  3.3.1 Overview The “World-wide” sample set consists of placer gold from the Prophet Mine Australia, Nus River Colombia, and from Piaba Brazil. The amount and quality of background geological information tied to these sample sets are variable. All information that could be gathered is presented here along with gold grain morphological and petrographic descriptions.    3.3.2 Geologic Settings and Bedrock Geology of “World-wide” Sites Samples for this project were derived from a wide geographic range spanning three continents. Criteria for selecting samples included documentation of gold occurrences, diversity of geologic setting, and availability. Table 3 outlines the key geological characteristics of the study sample groups. Where possible the likely bedrock source of the placer gold is given together with available information on size of occurrence, type of host rock, the style of mineralization, temperature, composition of mineralizing fluids, and associated alteration assemblages. These characteristics are important as any or all may contribute to the trace element geochemical signature of the preserved hypogene cores of the gold grains. Prophet Mine, Australia The Australian Prophet Mine gold samples are of particular interest for this study due to supergene gold rims on samples that have been reliably linked to the growth of bacterial biofilms living on gold grains (Reith et al., 2010). Although origin of the supergene gold rims is well understood, the bedrock source mineralization of grain cores is not well constrained. Mineralization occurrences in the area are broadly associated with the Paleozoic New England  41 Orogen (Little, 1995). The placer deposit occurs on Devonian to Carboniferous greenschist metasedimentary rocks of the Coastal Block (Ashley, 1980). This block consists mostly of quartz-rich siltstone, chert, greywacke, and shale which are intruded by Early Mesozoic tonalitic to dioritic plutons. This geologic setting is favorable for both Cu-porphyry and vein hosted precious metal deposits. Rocks experienced greenschist facies metamorphic conditions, which may constrain source mineralization temperatures. Nus River, Colombia  Samples from the Nus River, Colombia have the best documented hypogene gold mineralization source. They were found several kilometers downstream from AngloGold‟s Gramalote deposit. According to National Instrument 43-101 compliant reports, the deposit contains over 2.5 million ounces of gold at an average grade of 0.81 g/t (Gustavson Associates, 2012). The mineralization style is characterized as a shear-hosted orogenic gold deposit within tonalite/quartz porphyry intrusive rocks. The classification as an orogenic type Au deposit provides significant implications for formational processes. Piaba, Brazil  A single gold nugget from the Aurizona region of Brazil was extracted from laterite soil near the Piaba deposit probably at a depth of less than 2 meters (Fedorowich; pers. comm.). Gold mineralization in the area is classified as typical greenstone belt orogenic gold (SRK Consulting, 2012). Host rocks in the area include intermediate to mafic metavolcanic and metapyroclastic rocks, schists, banded iron formations, and metacherts; some of these units have been intruded by felsic intrusive bodies. The Piaba Deposit has National Instrument 43-101 compliant   42 Table 3. Key characteristics and geologic settings of sample sets used for this project. 1 (Little, 1995); 2 (Ashley, 1980); 3 (Ashley et al., 1978); 4 (Mccourt et al., 1984); 5 (Gustavson Associates, 2012); 6 (SRK Consulting, 2012); 7 (Jia et al., 2003); 8 (Nesbitt & Muehlenbachs, 1995) Placer Gold Sample Set Abbreviation Geographic Location Associated Orogeny Postulated Bedrock Source Mineralization Styles Prophet Gold Mine AUS Kilkivan, south-eastern Queensland, Australia New England Orogen (Paleozoic)1 Porphyry Cu, vein-type precious and base  metal2 Nus River COL Nus River, Providencia, Colombia Andean Orogeny; Central Cordillera (Mesozoic-Cenozoic)4 Gramalote Deposit: intrusive-hosted, structurally controlled,  stockwork gold-silver deposit. Measured and indicated resources: 2.5 Moz Au at 0.81 g/t5 Piaba Nugget PIABA Aurizona, Maranhāo State, Brazil Sao Luis Craton, Gurupi Belt (Paleoproterozoic) ~2240-2080 Ma6 Piaba Deposit in the Aurizona project area: Shear-hosted orogenic gold deposit. Measured and indicated resources: 3.2 Moz Au at 1.26 g/t6 Kelowna Placer Gold LAM, MIS, WIN Central Okanagan, British Columbia Canadian Cordillera (Mesozoic-Cenozoic) Miocene Paleoplacer source or minor gold bearing veins in amphibolite grade metamorphic rocks. Northern BC Gold BC Unknown Canadian Cordillera (Mesozoic-Cenozoic) Unknown  43 Table 3 (continued). Key characteristics and geologic settings of sample sets used for this project. 1 (Little, 1995); 2 (Ashley, 1980); 3 (Ashley et al., 1978); 4 (Mccourt et al., 1984); 5 (Gustavson Associates, 2012); 6 (SRK Consulting, 2012); 7 (Jia et al., 2003); 8 (Nesbitt & Muehlenbachs, 1995)  Placer Gold Sample Set Source Host Rocks Mineralization T (°C) Fluid Composition Prophet Gold Mine Devonian to Carboniferous greenschist metasediments of the Coastal Block intruded by Early Mesozoic tonalitic to dioritic complexes2,3 220-5502 Unknown Nus River Antioquia Batholith: multi-phase calc-alkaline I-type intrusive complex. Tonalite to diorite and granodiorite5 Unknown Au bearing fluid +\- Cu, Mo, Pb, Zn, Ag5 Piaba Nugget Aurizona Group metavolcanosedimentary sequence greenstone belt6 Greenschist Facies (300-450)6 Hydrothermal fluid (oxidized?) likely neutral pH low salinity fluids with high CO2 content7 Kelowna Placer Gold Miocene unconsolidated conglomerates overlying orthogneissic amphibolite grade basement rocks. Unknown Moderate to high salinity CO2 rich hydrothermal fluids8 Northern BC Gold Unknown Unknown Unknown  44 measured and indicated resources of 3.2 million ounces of gold at an average grade of 1.26 g/t (SRK Consulting, 2012). Mineralization may have been controlled by a chemical trap where oxidized Au bearing hydrothermal fluids came into contact with reduced graphitic metasedimentary rocks causing reduction of the fluid and precipitation of gold. Deep saprolitic weathering profiles overprinting the mineralization have been observed to extend to an average of 60 meters in depth. Petrography and Gold Grain Morphology of “World-wide” Sites The Prophet Mine samples show several unique petrographic and morphological features. Before polishing, the surfaces of the gold grains had distinct red-orange staining in places, especially in concave regions of the grain (Figure 11). After polishing, the grains revealed <15 µm thick Au-rich partial rims along concave surface embayments (Figure 11). Cores of grains were visually homogenous except for minor occurrences of primary inclusions of pyrite. These inclusions exhibit an inter-grown textural relationship with the gold and where anhedral pyrite grains contain inclusions of gold. Samples from Nus River, Colombia also exhibit Au-rich rims (Figure 12). These rims extend further into the grain as Au rich pockets in some locations. Some grains show dark-coloured fine-grained inclusions within the Au rich pockets.  The sample from Piaba, Brazil is a, roughly flake shaped, nugget ~1.7 cm by 1 cm (Figure 13). Petrographically no heterogeneity or Au-rich rim was observed (Figure 14). The deep yellow colour indicates high Au purity and the folded edges indicate detrital transport.   45  Figure 11. AUS5 gold grain reflected light petrographic image. Field of view is 0.65mm. Note the darker yellow coloured partial rim on top left of picture. The partial rim is localized mostly to the embayment feature on the grain surface.        46  Figure 12. Reflected light image of COL7 gold grain (width = 0.4mm). On the left margin of the grain a deeply coloured Au rich rim is present with several “pockets” extending deep within the grain. These orange-yellow coloured pockets also contain fine grained, detrital, inclusions.         47  Figure 13. Piaba Nugget. Note the deep yellow coloration and folded edge.       48           Figure 14. Reflected light image of Piaba nugget. Field of view is approximately 0.5mm. No Au-rich rim is apparent.  49 Chapter  4: Results  The results are broken into three interrelated sections that are separated for the purpose of data presentation. The first section presents the results of analysis of the standard reference materials by LA-ICP-MS. The second section gives the results of the geochemical analysis of gold collected from the Kelowna area. The final section provides geochemical data for placer gold grains compiled from all sample sites.    4.1  Reference Materials Results This project required the acquisition or creation of a standard reference material suitable for trace element analysis of natural gold. Although LA-ICP-MS provides an extraordinarily useful tool for trace element analysis of natural gold, it has not been utilized to its full potential because of the lack of a suitable external gold reference standard material that is homogenous on a microscopic scale. The optimal standard would contain all elements likely to be found in native gold at concentrations typical of the higher concentration range found in nature. Concentrations should be accurately known and must be homogenous at the tens of microns-scale (minimum beam diameter used). In addition, the standard must be matrix matched to the sample of interest. This is due to dramatic changes in coupling efficiency between the laser and different material (e.g. metals, sulphides, oxides, and silicates) which can lead to variable ablation yields and fractionation effects as well as matrix-dependent differences in the response curve (Potts et al., 1995). For this reason, the project required a standard with a gold matrix. External calibration with a gold standard is required to yield absolute concentrations of trace elements that are reproducible between instruments in different labs.   50  The external standard that most closely matched the requirements of the project was the AuRM2 standard. Dr. Mike Hinds of the Royal Canadian Mint in collaboration with the London Bullion Market Association (LBMA), and other refiners worldwide, created a gold reference material with known concentrations of 22 trace elements (Murray, 2009). Although created for analyzing refined gold the elements have appropriate concentration ranges to analyze natural gold samples. However, the standard was created for solution analysis and its homogeneity at a microscopic scale had to be checked to ensure it was suitable for LA-ICP-MS analysis.   In addition to AuRM2, NIST610 silicate glass was chosen as a supplementary external standard due to its ubiquitous use with microprobe techniques. It has been convincingly shown to be homogeneous at the micro scale (Jochum et al., 2011) and contains over 70 elements, most with nominal concentrations of ~400 ppm. But, as it is nowhere near matrix matched to placer gold samples, it was not deemed suitable as an external standard for this project on its own. A secondary goal of the research was to evaluate the impact of matrix effects, because if it can be shown to have minimal effect this would make available a vast number of additional elements that can at least be semi-quantitatively determined in natural gold samples.  The other supplemental external standard chosen was FAU7 which was borrowed from the Royal Canadian Mint. A previous attempt at quantitative determination of trace elements in natural gold using LA-ICP-MS used this standard (Penny, 2001). It is limited to 16 elements with known concentrations and has not been tested for homogeneity at the micro-scale, but provides an additional matrix-matched standard with which to cross-reference to AuRM2. If elements not intentionally added during formation of FAU7 can be quantified by use of AuRM2 as an external standard, these elements could be added to the list of trace elements of known concentrations for FAU7.  51 4.1.1 Characteristics of AuRM2 Reference Material Creation of AuRM2 is detailed in Murray (2009). It was rigorously tested for homogeneity on the macro-scale by the manufacturer and then by ten independent laboratories (Murray, 2009). Quantitative verification of trace element concentrations and uncertainty (Table 4) led to certification on July 7th, 2009.   4.1.2 Precision of Reference Material Analyses Precision, defined as the ability to reproduce a result, can be easily represented by RSD (relative standard deviation) which is a percentage value derived from the standard deviation of a group of numbers divided by the mean (Equation 2); this provides a metric of variance. Precision, as represented by RSD, for each reference material is presented in Table 5; this was calculated from 24 individual analyses of each standard reference material using the same reference material as the external standard.    % Equation 2. RSD calculation used to measure variance of elemental concentrations resulting from multiple analyses of individual standard reference materials.   RSD values are consistently lower for NIST610 analyses than either of the gold reference materials; almost all elements in NIST610 are below 10% RSD. Between the gold reference materials, RSD values are variable from element to element but overall it appears that AuRM2 exhibits slightly lower RSD than does FAU7. As Ag was used as an internal standard, the Ag concentration values were manually entered into the GLITTER program based on accepted  52 Table 4. List of trace elements added to AuRM2 certified reference material along with concentrations and uncertainty values (2σ standard deviation). (Modified from Murray, 2009). Element Concentration 2σ Standard Deviation Ag 99.6 5.6 Al 28.3 1.8 As 47.1 2.8 Bi 9.7 0.8 Ca 28.0 2.6 Cr 27.7 2.2 Cu 31.6 2.4 Fe 30.1 2.2 Mg 9.9 0.9 Mn 28.2 1.5 Ni 29.2 2.6 Pb 28.9 2.4 Pd 29.2 1.3 Pt 30.2 2.1 Rh 39.6 2.4 Sb 11.3 1.6 Se 37.4 2.8 Si 28.0 3.8 Sn 29.4 1.8 Te 12.0 3.2 Ti 31.6 1.3 Zn 31.4 2.3  53 values leading to an apparent RSD of 0%. For AuRM2 the RSD values of additional semi-quantitatively determined values for P, V, Hg, and U are also included (methodology for semi-quantitative calculation is included in Chapter These elements were not intentionally added to the standard and therefore their concentrations are low and their RSD is elevated. Semi-quantitative Hg RSD is also included for NIST610. Dashes indicate the elements for which concentrations were neither quantitatively nor semi-quantitatively determined; this includes elements in FAU7 not added to the standard and Au in NIST610 which was not detectible due to the detector being in Faraday mode to avoid overload during analysis of the Au matrix standards.  Quantitatively determined elements that exhibit anomalously elevated RSD values (>30%) in AuRM2 include Si, Ca, and Cr. Moderately elevated (15-30%) values are reported for Mn, As, and Te. All remaining quantitatively determined elements in AuRM2 are below 15% RSD.              54 Table 5. Precision values of each element analyzed (along with the isotope used) for standard reference materials used in this study. RSD is reported for all three indicating degree of reproducibility between analyses. Dashes indicate not analysed or not applicable. N = 24 analyses for each standard. Analyzed July 3-4th 2013 at FiLTER, University of British Colombia, Okanagan. Element/Isotope NIST610  %RSD AuRM2  %RSD FAU7  %RSD Mg24 6.28 6.42 5.91 Al27 6.49 6.74 - Si29 6.36 39.98 201.66 P31 8.40 11.68* - Ca44 6.06 31.90 - Ti48 5.93 8.63 18.35 V51 5.88 16.00* - Cr53 8.33 31.69 59.48 Mn55 7.02 18.69 6.65 Fe57 9.17 11.02 23.56 Ni60 5.97 5.99 4.50 Cu63 5.55 6.24 6.17 Cu65 5.90 6.95 6.06 Zn68 6.70 8.72 11.84 As75 6.93 15.03 49.48 Se77 8.05 10.98 - Rh103 4.33 10.87 - Pd105 5.10 13.08 8.33 Ag107** 0 0 0   55 Table 5 (continued) Element/Isotope NIST 610  %RSD AuRM2  %RSD FAU7  %RSD Sn118 5.44 8.99 14.20 Sb121 5.61 11.21 - Te130 3.31 18.98 - Pt195 3.25 10.88 7.20 Au197 - 11.38 15.66 Hg202* 22.04* 33.82* - Pb204*** 6.26 19.58 11.79 Pb206 5.93 11.58 14.30 Pb207 6.10 11.55 13.39 Pb208 7.66 11.35 14.21 Bi209 7.59 13.37 20.36 U238 6.38 361.29* - *Semi-quantitative data ** Ag used as internal standard using given values ***Strong Hg204 interference                  56 4.1.3 Accuracy of Reference Material Analyses To make accuracy estimates, the three standards were assumed homogenous at the micro-scale of the laser beam. Each reference material was treated as an unknown and element concentrations were calculated using another reference material designated as the external standard. Treating the reference material as an unknown circumvents the bias introduced when, using the same material as an external standard, accepted concentration values are input. When using an alternate reference material as an external standard there is no input of accepted concentration values, only output.  New issues arise because there are compounded uncertainties in the actual composition of all of the standards. Thus, an estimate of poor accuracy could reflect inaccuracies in either standard. Another factor is that NIST610 is not matrix matched to the gold standards. For this reason agreement between measured and accepted values for NIST610 cannot necessarily be expected when one of the gold reference materials is used as the external standard and vice versa. It is a useful comparison however as the degree of matrix effect can be quantified.  The best test of accuracy is the comparison of the two gold reference materials AuRM2 and FAU7. Each was in turn treated as an unknown against the other (Table 6). Accepted values are listed first followed by measured concentrations using other designated external standards. The accuracy is quantified using a simple yet effective percentage formula yielding a plus or minus variation. Dashes indicate elements that do not have accepted concentrations in both standards and therefore cannot be compared statistically.      57 Table 6. Accuracy statistics (Accur.) for standard reference materials treated as unknowns against another reference material used as an external standard (Ext.). Accepted values (Accept.) are given by the manufacturer and are reported in ppm. Measured values (Meas.) are taken by treating the standard as an unknown and are also reported in ppm. Accur. = (|Meanmeas. – Accept.|/ Accept.) * 100% Element /Isotope AuRM2 Accept. (ppm) AuRM2¹ Meas.  (NIST Ext.) AuRM2¹ Accur. (NIST Ext.) AuRM2² Meas. (FAU Ext.) AuRM2² Accur. (FAU Ext.) FAU7   Accept. (ppm) FAU7³  Meas. (AuRM Ext.) FAU7³  Accur. (AuRM Ext.) Mg24 9.9 11.6 16.9% 9.8 1.0% 34.0 35.7 5.0% Al27 28.3 47.1 66.4% - - - 27.8 - Si29 28.0 391.6 1298% 9.2 67.2% 2.7 29.4 989.5% P31 - 465.0 - - - - - - Ca44 28.0 177.2 532.8% - - - 11.3 - Ti48 31.6 17.4 45.1% 28.5 9.7% 12.7 14.6 14.8% V51 - 0.3 - - -  - - Cr53 27.7 47.1 70.1% 137.4 396.2% 32.6 34.9 7.1% Mn55 28.2 38.1 35.2% 26.1 7.4% 58.9 70.6 19.8% Fe57 30.1 1009.3 3253% 16.5 45.3% 11.6 28.3 143.7% Ni60 29.2 28.4 2.8% 28.6 2.2% 32.5 36.4 11.9% Cu63 31.6 25.7 18.8% 29.1 8.0% 98.1 118.0 20.3% Cu65 31.6 31.3 1.1% 30.5 3.3% 98.1 114.1 16.3% Zn68 31.4 39.5 25.6% 29.5 5.9% 54.6 66.5 21.7% As75 47.1 86.8 84.3% 36.5 22.4% 10.0 19.0 89.6% Se77 37.4 134.8 260.3% - - - 24.9 - Rh103 39.6 38.7 2.4% - - - 0.1 -   58 Table 6 (continued) Element /Isotope AuRM2 Accept. (ppm) AuRM2¹ Meas.  (NIST Ext.) AuRM2¹ Accur. (NIST Ext.) AuRM2² Meas. (FAU Ext.) AuRM2² Accur. (FAU Ext.) FAU7   Accept. (ppm) FAU7³  Meas. (AuRM Ext.) FAU7³  Accur. (AuRM Ext.) Pd105 29.2 15.3 47.5% 25.8 11.6% 43.1 51.2 18.7% Ag107** 99.6 99.6 - 99.6 - 20.3 20.3 - Sn118 29.4 26.7 9.2% 30.5 3.7% 33.8 39.3 16.4% Sb121 11.3 12.5 10.6% - - - 0.3 - Te130 12.0 26.5 120.5% - - - 2.0 - Pt195 30.2 36.9 22.3% 30.3 0.3% 87.1 90.4 3.8% Hg202 - - - - - - - - Pb204* 28.9 76.4 164%* 17.4 39.7%* 21.9 63.3 189%* Pb206 28.9 30.1 4.3% 28.5 1.3% 21.9 28.4 29.8% Pb207 28.9 30.6 6.0% 30.5 5.6% 21.9 26.8 22.6% Pb208 28.9 27.6 4.5% 29.9 3.4% 21.9 27.6 26.0% Bi209 9.7 10.0 3.3% 11.0 12.9% 24.0 29.7 23.9% U238 - 0.1 - - - - - - 1 using NIST610 as external standard 2 using FAU7 as external standard 3 using AuRM2 as external standard * Hg204 interference **Ag accuracy could not be calculated as it was used as an internal standard.        59 Accuracy of these standard analyses is quite well constrained. When comparing AuRM2 treated as an unknown with the markedly different matrix of NIST610 used as an external standard (Table 6; column 3 and 4), the values come back mostly within a factor of 2; many elements are <10% different. Exceptions include elements that are at trace levels in the gold standard but are major constituents in NIST610 including Si, Ca, and Fe. Both Se and Te have poor accuracies greater than plus or minus a factor of 2. Pb concentrations calculated using different isotopes are agreeable (~30 ppm) with the exception of 204Pb which yields a significantly higher value (76 ppm) due to 204Hg interference. The calculated concentration uses the combined signal from 204Pb + 204Hg as they cannot be individually resolved. If neither standard contained appreciable Hg the concentration measured in AuRM2 using 204Pb would align with the other isotopes. The higher calculated Pb concentration in AuRM2 using 204Pb, means that there is a significant abundance of Hg in the standard.     A better test of accuracy comes from comparing the results between the two gold reference materials as they have the same matrix. Treating AuRM2 as an unknown with FAU7 used as the external standard (full raw data set can be found in Appendix C) accuracy values are generally better than ±10% (Table 6). Notable exceptions include 29Si and 53Cr which were isotopes that had unacceptably high background values during analysis and which were also observed to have elevated RSD values in Table 5. Other elements with significant variation (>10%) from accepted values include Fe, As, Pd, Bi, and surprisingly, Au. Again, 204Pb is affected by Hg interference, but this time a concentration lower than the accepted value is calculated in AuRM2 (17 ppm). This indicates that 204Hg is a proportionally larger component of the 204Pb + 204Hg signal for FAU7 than for AuRM2.   60  In the reverse situation, with FAU7 treated as an unknown, values deviate slightly more from accepted values but most elements are within 20% (Table 6). In addition, trace elements including Al, Se, Rh, Sb, and Te that are measurable in FAU7 but do not have accepted values can be given approximate concentrations from AuRM2. Matrix Effects between Reference Materials  AuRM2 measured concentrations, using NIST610 as an external standard, divided by accepted concentrations of AuRM2 are plotted by individual analysis (Figure 15). This shows variance from over or underestimating the concentration; a value of 1 would indicate NIST610 has predicted the exact accepted concentration of AuRM2 (no effect from a non-matched matrix). Elements that were excluded in Figure 15 include those that were major element components in either of the reference materials (Au, Si, Ca, and Fe).  The overall flat trend of Figure 15 also demonstrates that there was no significant instrumental drift over the period of analysis. The increased scatter at the far right corresponds to a reduction in beam diameter for analyses during the final two sample runs. A small negative anomaly is present on the left of the graph indicating an anomalous analysis of either AuRM2 or the NIST610 external standard.  For the trace elements it appears that most elements fall between half (meas./accept. = 0.5) and double (meas./accept. = 2) of true values. Ti appears to be the furthest outlier and is underestimated in AuRM2 at approximately meas./accept. = 0.5. As and Al also stand out at close to a ratio of 2.    61                     Figure 15. AuRM2 concentrations calculated using NIST610 as an external standard normalized to accepted concentrations. A value of 1 indicates a perfect match; values above 1 indicate an overestimation and those below indicate an underestimation of concentration.                                                                                                                                                                               62 Semi-quantitative Concentrations In addition to the 20 elements added to AuRM2, background corrected counts for P, V, Hg, and U indicate measurable concentrations of these elements in the standard. NIST610 was used as an external standard to calculate semi-quantitative P, V, and U concentrations in AuRM2. Although the materials are not matrix matched the results from Section indicate that most trace elements will have similar responses in glass and gold matrices because significant deviations were only observed for major elements. Thus, semi-quantitative data for P, V, and U are likely accurate to within a factor of 2. As a secondary verification of these values, semi-quantitative concentrations of these elements were also calculated from the response curve using Equation 3:   Equation 3. Formula for semi-quantitative determination of trace element concentrations in AuRM2 reference material (From Thermo Scientific).    where:  RSP1  = instrument specific response for lighter element RSP2  = instrument specific response for heavier element                          (`new‟ indicates reference material in which element of interest is unknown)                         (`old‟ indicates reference material in which element of interest is known)                     = mass of element of interest                      = mass of lighter element                    = mass of heavier element        63 Instrument specific response was calculated from Equation 4:    Equation 4. Instrument response formula.   where:          = measured intensity of isotope (background corrected counts per second)  a                = natural abundance of measured isotope  c                = concentration of the element in the sample    Semi-quantitative concentrations predicted using this formula are similar to the values derived by treating AuRM2 as an unknown with NIST610 used as an external standard. Using the average of n=24 analyses the above equation yielded a RSP of 604.98 counts/ppm for V with a concentration of 0.305 ppm in AuRM2. This compares to a value of 0.385 ppm measured using NIST610 as external standard. The signal for U was quite low in AuRM2 and therefore has a very high RSD due to the poor counting statistics. The semi-quantitative equation cannot be used for U as there is no heavier element to calibrate to. Using NIST610 as an external standard, however, gave a value of 0.14 ppm in AuRM2.   The P concentration in AuRM2 predicted from the equation yielded a very low RSP value of 0.103 counts/ppm. A low value is to be expected as the ion yield of P is extremely low due to its high electronegativity. It is very difficult for the plasma torch to strip electrons off of P to form P+ which is what is measured by the detector. The concentration calculated from the formula is quite high at 822 ppm P in AuRM2. Estimation of P using NIST610 as an external standard is also high at 465 ppm. This calculated concentration of P is irregular due to the fact that it is over an order of magnitude higher than that of elements intentionally added to the standard.  64           Mercury was a problematic element to determine quantitatively or semi-quantitatively as there are no accepted concentrations for any of the standards used in this experiment. A strong signal for Hg was detected during analysis of the standards and all natural gold samples indicating its significance for this project. The most recent and thorough examination of NIST610 trace element concentrations following ISO guidelines (Jochum et al., 2011) states that no published data exist for Hg concentrations in the widely used reference material. The only reference in the literature to Hg in NIST610 is in regards to Pb interference observations while measuring the Pb isotopic ratio of the standard (Walder et al., 1993). They observed that 204Hg contributed approximately 10% of the intensity of the 204Pb peak. The authors used the assumption that Hg and Pb were ablated with similar efficiency to estimate the Hg concentration in NIST610 to be 4 ppm.   Using a value of 4 ppm Hg in NIST610 as a base point, an attempt to quantify the concentration of Hg in AuRM2 was made. The semi-quantitative formula (Equation 3) yielded a relatively high value of 25.8 ppm Hg in AuRM2; the same formula gave a value of 46.5 ppm Hg for FAU7. This crude approximation has low accuracy due to many sources of uncertainty but is likely within an order of magnitude of the true value. To test this value the approximated Hg concentration was added to the list of elements in the GLITTER software and the data was re-processed. This resulted in rough semi-quantitative concentrations of Hg in all placer gold grains with no analyses below detection limits. The resulting concentrations were compared to known Hg concentrations in a few grains that had detectable Hg from SEM-EDS analyses.  Four SEM analyses of two grains (AUS9 and BC1) contained between 3.1 and 3.5 wt% Hg. The initial LA-ICP-MS semi-quantitative values for these grains ranged between 21 to 25 wt% which is approximately 7 times the true (SEM-EDS) value. Because the RSP equation  65 (Equation 4) is directly proportional to concentration, the AuRM2 Hg concentration can be reverse calibrated by dividing the original value by 7 (factor by which Hg was overestimated). This gives a new refined set of semi-quantitative values of 3.69 ppm Hg in AuRM2 and 14.8 ppm in FAU7. Using these concentrations to back calculate the concentration of Hg in NIST610 gives a value of 0.57 ppm. This is lower than the 4 ppm value estimated by Walder et al. (1993).  4.1.4 Homogeneity of AuRM2 Before LA-ICP-MS geochemical results from natural gold samples can be interpreted, the external standard reference material AuRM2 must be evaluated for homogeneity at the sampling scale of the laser (64-108 microns). A useful benchmark for homogeneity is the exhaustively documented NIST610 standard. By comparing the reproducibility of AuRM2 to NIST610 analyses, the suitability of AuRM2 as a reference material for micro-analysis can be examined. RSD calculations for a large number of reference material analyses can provide some indication of homogeneity for individual elements (Table 5); however, this method does not take into account factors such as: mass of material sampled, concentrations of each element in the material, ablation yield, and ionization yield, all of which affect the statistics involved in calculating reproducibility (precision) and homogeneity. To account for this, the background corrected signals of elements were taken into account to assess the homogeneity of AuRM2 relative to that of NIST610. The nominal trace element concentrations in NIST610 are around 400 ppm which tends to be at least an order of magnitude higher than most trace element concentrations in AuRM2 (Table 4). To aid in comparing these two standard reference materials the total mass of material ablated during lasering and the corresponding elemental concentrations within that mass can be  66 calculated. This is done by first calculating the volume of sampled material based on pit diameter and depth.  Laser ablation pit depths were optically measured by determining the difference in height between the top of the reference material and the bottom of the ablation pits using the height-calibrated focusing knob on a Nikon Eclipse 50iPOL petrographic microscope. Pits were found to average 19 microns in depth in a gold matrix during analysis for this project. The beam diameter used was 108.5 microns which yields a sampling volume of 175,674 cubic microns. Compare this to a calculated pit depth of approximately 90 microns (using a value of 0.3 microns per laser shot; Rob Henry, per. comm.) in NIST610 glass. This corresponds to a sampling volume of ~832,000 cubic microns for NIST 610. These values should be viewed as a maximum sampling volume as significant amounts of sample may not have left the sample chamber; indeed significant ablation ejecta were observed around laser pits.   The sharp contrast between the densities of AuRM2 and NIST610 yields a much closer sample mass. AuRM2 at 19.3 g/cc has an average sample mass of 3.4 micrograms whereas NIST610, at ~2.6 g/cc, has a slightly lower sample mass of approximately 2.2 micrograms. Using these mass values, and the known concentrations of contained elements, the approximate mass of each element sampled can be calculated (Table 7). Given the far higher concentration of most elements in NIST610 a greater elemental mass than sampled in AuRM2 is predicted for most elements. This should be close to proportional to the signal generated for each isotope on the mass spectrometer detector (equivalent to the background corrected counts per second).       67 Table 7. Calculation of mass of each element sampled for AuRM2 and NIST 610.  Mass Sampled (micrograms) Element Stated Concentration (ppm)  Mass         of Element (micrograms) AuRM 3.4  Mg 30.1  1.03E-04 NIST 2.2  Mg 432  9.35E-04 AuRM 3.4  Al 28.3  9.65E-05 NIST 2.2  Al 10319.93  2.23E-02 AuRM 3.4  Si 28  9.54E-05 NIST 2.2  Si 322847.8  6.98E-01 AuRM 3.4  Ca 28  9.54E-05 NIST 2.2  Ca 78902.14  1.71E-01 AuRM 3.4  Ti 31.6  1.08E-04 NIST 2.2  Ti 452  9.78E-04 AuRM 3.4  Cr 27.7  9.44E-05 NIST 2.2  Cr 408  8.83E-04 AuRM 3.4  Mn 28.2  9.61E-05 NIST 2.2  Mn 444  9.61E-04 AuRM 3.4  Fe 30.1  1.03E-04 NIST 2.2  Fe 458  9.91E-04 AuRM 3.4  Ni 29.2  9.95E-05 NIST 2.2  Ni 458.7  9.92E-04 AuRM 3.4  Cu 31.6  1.08E-04 NIST 2.2  Cu 441  9.54E-04 AuRM 3.4  Zn 31.4  1.07E-04 NIST 2.2  Zn 460  9.95E-04 AuRM 3.4  As 47.1  1.61E-04 NIST 2.2  As 325  7.03E-04 AuRM 3.4  Se 37.4  1.27E-04 NIST 2.2  Se 138  2.99E-04 AuRM 3.4  Rh 39.6  1.35E-04 NIST 2.2  Rh 1.29  2.79E-06 AuRM 3.4  Pd 29.2  9.95E-05 NIST 2.2  Pd 1.21  2.62E-06 AuRM 3.4  Ag 99.6  3.39E-04 NIST 2.2  Ag 251  5.43E-04 AuRM 3.4  Sn 29.4  1.00E-04 NIST 2.2  Sn 430  9.30E-04 AuRM 3.4  Sb 11.3  3.85E-05 NIST 2.2  Sb 396  8.57E-04         68                 Table 7 (continued)  Mass Sampled (micrograms)  Element Stated Concentration (ppm)  Mass of Element (micrograms) AuRM 3.4  Te 12  4.09E-05 NIST 2.2  Te 302  6.53E-04 AuRM 3.4  Pt 30.2  1.03E-04 NIST 2.2  Pt 3.12  6.75E-06 AuRM 3.4  Au 999321.6  3.41E+00 NIST 2.2  Au 23.6  5.11E-05 AuRM 3.4  Pb 28.9  9.85E-05 NIST 2.2  Pb 426  9.22E-04 AuRM 3.4  Bi 9.7  3.31E-05 NIST 2.2  Bi 384  8.31E-04                69 Considering that the RSD statistics are strongly dependent upon mass sampled and the proportional background corrected counts (RSP) of the instrument, it is useful to plot these values for each element in AuRM2 to observe the relationship. When all three reference materials are plotted on the same signal versus RSD plot the relationships between them become apparent (Figure 16). NIST610 data extend from 100 all the way to 10,000,000 counts per second. Both of the gold reference materials are limited to a maximum of around 100,000 counts per second and both appear to converge with the NIST610 trend moving from low to high signal. The negative sloping trend of the graph makes it evident that signal intensity has a strong control on reproducibility in AuRM2. There are also three distinct groupings of AuRM2 elements. Two isotopes have very low signals yet moderate RSD. Six isotopes have a moderate signal and very poor reproducibility. The remainder of the isotopes measured have a strong signal and relatively low RSD.   Examination of NIST610 analyses shows a shallower negative trend with less scatter and every isotope has an RSD below 10%. It appears that the RSD values for NIST610 analyses are less signal-dependent and have greater reproducibility. The trend is also more linear than observed for AuRM2.  FAU7 (also shown in Figure 16) mimics the distribution of AuRM2. There are four isotopes grouped at the moderate signal and high RSD portion of the graph with the remaining isotopes trending towards high signal and relatively well constrained RSD values. There is also significant scatter as observed for AuRM2. There are fewer elements to analyze in FAU7, but the ones that are present appear to generally behave in the same manner as in AuRM2.  Examining individual elements the greatest signal was recorded from Si, Ca, Ti, Pb, and Bi in NIST610. Si, Ca, and Ti are all major elements in NIST610 so their high signal is  70 unsurpising. In fact their signals would be even higher if low abundance isotopes were not chosen. Pb and Bi are both trace elements in NIST 610 but they are heavy elements and their prescence at the high signal range is likely due to mass bias of the detector. Gold values for the gold reference materials are not displayed on this graph as they were completely off scale at well over a billion counts per second; this is due to the high concentration (>99.9 Wt%), the mass bias, and the fact that Au is monoisotopic.   Low signal elements generally display a higher RSD and include Fe and Se in both gold reference materials and NIST610. This is due to the necessitiy of selecting low abundance isotopes of trace elements to avoid polyatomic interferences.  There is a distinct grouping of elements from the gold reference materials that are shifted  above the general trend of the elements. For AuRM2 these anomolous elements include Si, Cr, Ca and to a lesser extent As, Te, and Mn. FAU7 anomolous elements include Si, Cr, Ca, As, and perhaps Bi. These elements that are shifted above the general trend may not be homogenously distributed in their respective standards.   71 Analytical Signal vs. Reproducibility for AuRM2, NIST610, and FAU7  Figure 16. Compilation of average background corrected counts per second versus %RSD data by element for AuRM2, FAU7, and NIST610.  72 Another, more rigorous, method of assessing homogeneity incorporates counting statistics by comparing the theoretical reproducibility of analyses from counting statistics with the measured reproducibility (RSD) (Jochum et al., 2011).  There is inherent uncertainty in any analytical method; even a perfectly homogenous material will show some variation between analyses due to counting statistics. The reproducibility observed in a homogenous material is dependent on the strength of the signal used to calculate concentration (stronger signal lowers RSD). This variation can be statistically predicted based on a Poisson distribution to calculate a theoretical RSD (Equation 5; for further information refer to: Microprobe Techniques in the Earth Sciences, by Potts et al. 1995), thus analytical variation can be separated from true heterogeneity. If the measured RSD for an element exceeds the theoretical RSD, then this leads to the conclusion that the element is heterogeneously distributed.  Equation 5. Predicted repeatability of LA-ICP-MS analyses using counting statistics and based on a Poisson distribution.   The counting statistics for LA-ICP-MS analyses of AuRM2 consist of approximately 10-20 unique signal measurements for each of 21 isotopes. These measurements for each AuRM2 analysis were compiled, the reference material signal manually picked out, the background signal subtracted for each, and average signal (background corrected counts per second) calculated for each element. The signal is reported in counts per second and was converted to raw counts (actual number of ions hitting the detector) by multiplying by 100 milliseconds (the dwell time  73 used during analysis). This background-corrected signal underwent a simple square root conversion (Equation 5) to obtain 1 sigma values (assumption of a Poisson distribution).  This sigma value was used to calculate a RSD of repeatability (Equation 2) for these analyses. The calculation for true heterogeneity (Equation 6; Jochum et al., 2011) accounts for variation in analytical reproducibility due to counting statistics. Table 8 lists the true heterogeneity of each element in AuRM2 at the analytical signal levels for this study based on the difference between measured and predicted RSD. This successfully separates analytical noise from true heterogeneity.   Equation 6. Calculation of heterogeneity in AuRM2 negating the effect of counting statistics which control repeatability of analyses (from Jochum et al., 2011). RSDmeasured = the % RSD values for AuRM2 in Table 5. RSDrepeatability = the theoretical % RSD from counting statistics (from Equation 5).            Mercury and U, elements not intentionally added to AuRM2, have heterogeneity RSD values of over 20% and appear significantly inhomogeneous at the sampling scale of the laser. Values between 10 and 20% indicate minor heterogeneity for Mn, Rh, Pd, Pt, Pb, and Bi. The remaining elements have < 10% heterogeneity RSD and for the purposes of this study are considered homogenous. The elevated values for repeatability RSD in some elements including Fe, As, Se, and Te are the result of poorer signal to noise values during analysis.      74 Table 8. Theoretically predicted and actual measured RSD of AuRM2 to quantify heterogeneity. Values for statistically calculated (Repeat. RSD), analytically measured, and true heterogeneity for 21 elements are listed. Si, Ca, and Cr are not included due to very poor signal to noise analytical ratios. Predicted RSD (ie. Repeat. RSD) is based on Poisson distribution of reproducibility of background corrected signal. Measured RSD comes from LA-ICP-MS concentration analysis. Heterogeneity is calculated from Equation 6. Values of 0 were entered where errors in calculation occurred when measured RSD was actually less than predicted repeatability of analysis.  Element Predicted RSD Measured RSD Heterogeneity RSD Mg 4.7 6.4 4.3 Al 3.4 6.7 5.8 Ti 2.7 8.6 8.2 V 31.2 16.0 0.0 Mn 13.9 18.7 12.5 Fe 42.6 11.0 0.0 Ni 5.0 6.0 3.3 Cu 4.4 7.0 5.4 Zn 6.7 8.7 5.5 As 20.0 15.0 0.0 Se 39.0 11.0 0.0 Rh 1.6 10.9 10.8 Pd 3.9 13.1 12.5 Sn 4.6 9.0 7.7 Sb 6.4 11.2 9.2 Te 20.3 19.0 0.0 Pt 3.0 10.9 10.5 Hg 5.8 33.8 33.3 Pb 1.9 11.4 11.2 Bi 2.4 13.4 13.2 U 52 361 357  75 Multi-Dimensional Scaling Plots of Standards Multi-dimensional scaling, an exploratory statistical method, was conducted on AuRM2 reference material analyses using methods outlined in Section 2.7 to construct elemental MDS plots (Figure 17). The purpose of this was to determine what mode of heterogeneity was present in the standards.  This style of graph highlights correlations and dissimilarities in the behavior of elements within the system. These trends can only be seen, however, when a group of samples have all undergone similar processes. In the case of a standard reference material it represents the near ideal conditions for this comparison as the conditions and processes are strictly controlled during production. For this reason even the most subtle differences in elemental fractionation may be teased out of a large data set.   There appear to be some elemental associations evident from MDS plots of the AuRM2 standard (Figure 17). Elements that plot close together behaved similarly during analysis, for example the most striking feature of the plot is the grouping of PGEs (Platinum Group Elements). This means that when one PGE concentration is high in a single analysis the other PGE concentrations are also likely to be high in that analysis. Chalcophile elements are grouped at the opposite side of the chart from the PGEs. This indicates that when PGEs are abundant in an analysis the chaclophiles tend to be low in abundance and vice versa. Sn and Al appear to be following the chalcophile elements. Centrally located are the siderophile elements: Ni and Fe along with lithophile elements: Ti, Mn, and Mg. The elements that had very strong interferences or high backgrounds and did not yield useful data include Cr, Ca, and Si. These elements are grouped away from other trace elements and closer to Ag, which was used as an internal    76  Figure 17. Elemental MDS plot of n=64 analyses of AuRM2. Data came from two different labs. Half of the analyses were collected by the author at the UBC FiLTER Lab, Kelowna. The other half were collected at the University of Windsor GLIER facility (Fryer 2013; pers. comm.).         77 standard. The more abundant elements, Ag and Au, plot on opposite sides of the graph. Interestingly, the trace elements fall in a line closer to Au than to Ag.  4.2  Kelowna Area Placer Gold Results Gold from the central Okanagan region provides a large sample set of placer gold grains from a spatially restricted area (10s of kilometers) to examine variation of placer gold trace element signatures at the district scale. This project can determine the sources of gold in these occurrences, which were not previously known with any degree of certainty.   4.2.1 Major Element Results for Kelowna Area Samples The only elements detected in the Kelowna area gold grains by SEM-EDS analyses (~0.1 wt% detection limit) were Ag and Au. Concentrations of Ag in the hypogene cores vary from 5 to 26 wt% which corresponds to a gold fineness range of 950 to 740 (pure gold fineness is 1000). Au-rich rims are observed from each locality with concentrations of up to 100 wt% (essentially pure gold). Variations in major alloying element concentrations are apparent in some electron backscatter images of grains. Many detrital inclusions of magnetite, quartz, and possibly other silicate minerals were observed during analysis. In addition some grains from Lambly Creek contain primary sulphide minerals that were identified by SEM-EDS.    Gold grain samples from Lambly Creek have an average hypogene Au concentration of 84.8 wt%. Rims with higher Au concentrations (up to 100 wt%) are present in three out of six gold grains. Internally, the cores of individual grains have homogenous major element concentrations (Figure 18).   78 Grains exhibiting Au rich rims show relatively sharp core-rim compositional changes and rim thicknesses of around 5 microns (Figure 19). These features are best shown by SEM-EDS linescans perpendicular to the surface of gold grains. Not all grains had compositional changes at the rim however (Figure 20) and show internal homogeneity. Two dimensional SEM-EDS elemental maps were created from select grains (Figure 21). LAM 8, a composite grain, shows the utility of this technique. A Au-rich and Ag-poor rim is evident as are detrital and primary inclusions including primary sulphide crystals (pyrrhotite), quartz, and possibly other Al-rich silicate minerals. Samples from Mission Creek have an average hypogene core concentration of 83.4 wt% Au (Figure 22). Three out of five grains have rims with higher Au and lower Ag concentrations with some grains exhibiting thick Au-rich rims (Figure 23). These Au-rich rims follow cracks, crevices, and embayments in and along gold grains. MIS2 shows two separate hypogene cores that may reflect the depth that the gold grain was sectioned (Figure 24).   Gold grains collected from the Winfield paleoplacer deposit have an average hypogene core Au concentration of 84.0 wt%. Three out of five grains have a Au-rich, Ag-poor, rim with a maximum Au concentration of 98 wt%. Internally the gold grains are homogenous (Figure 25). Chemical composition boundaries are not as sharp as in gold grains from Mission Creek (Figure 26).         79  Figure 18. BSE image of LAM8 gold grain from Lambly Creek overlain by SEM-EDS analysis results. Gold grain is 0.45 X 0.25 mm. Note numerous primary and/or detrital mineral inclusions.   80                        Figure 19. SEM-EDS linescan of LAM8 gold grain. Results are reported in cps (counts per second). Note distinct increase in Au signal and elimination of Ag signal at edge of grain (3µm-7 µm). C values simply reflect the carbon coating of the samples.  81            Figure 20. SEM-EDS linescan of LAM7 gold grain. Note lack of Au enriched rims and internal homogeneity.  82                        Figure 21. Above: Composite SEM-EDS elemental map of LAM8 gold grain. Colours represent proportional intensity of signal in cps for individual elements detected. Insets: Individual element signals isolated from composite image. Note higher Au signal on rim and various inclusion types identified.  83   Figure 22. BSE image of MIS2 gold grain from Mission Creek with SEM-EDS analysis results overlaid. Faint lighter colour contrast along rim of grain indicates Au rich rim which is denser than the core.  84                 Figure 23. SEM-EDS linescan across MIS4 gold grain. This linescan is created by measuring relative intensity of signal peaks for detectable elements at various points along the line. Above electron backscatter image shows location of the linescan. Note the Au rich rim on both sides of the gold grain. Embayment with detrital fill is present on the left side.    85                        Figure 24. Left: SEM-EDS elemental map composite image of MIS2 gold grain. Note very distinctive Au rich and silver poor rim surrounding cores of grain. Below: Isolated signals for each element detected.   86  Figure 25. Backscatter electron image of WIN6 gold grain. Note distinct fold on upper left. Au concentrations are higher on the rim. The hypogene core is electrum.  87            4.2.2 Trace Element Results for Kelowna Area Samples            Figure 26. Fine scale SEM-EDS linecan of WIN5 gold grain. Ag signal (green) dies off on the outer two microns of the grain.  88 The Kelowna area gold grains were analyzed for trace elements by LA-ICP-MS (Table 9). Differences between Kelowna localities are highlighted in this section with compositions compared with world-wide data in Section 4.3.2.   Averaged data for each placer occurrence (Table 9) shows that the most abundant trace elements are Hg, and Cu which are over an order of magnitude higher than all other trace elements except Fe; Hg is particularly high (up to thousands of ppm). The least abundant element detected was Rh which is present consistently at around 50 ppb. All other elements fall between 0.1 and 10 ppm in the native gold samples.  Mean Cu, As, Rh, Pd, Sn, Sb, and Te do not show large differences between localities though samples from Lambly Creek (LAM) have lower abundances for most trace elements except for slightly higher Cu and Sb. Mission Creek samples are enriched in the lithophile elements Mg, Ti, and U as well as siderophile and chalcophile Fe, Ni, Zn, Se, Pt, Bi, and Hg, but have low Pb. Winfield gold grains have abundant Al, Mn, Pd, V, and high Hg. Probability plots (Figure 27) illustrate differences in concentrations and distribution of selected elements. The advantage of this mode of presentation is that variations in concentration within sample groups due to both natural variations and reproducibility of analyses can be observed. Trends for Fe, Se, and Hg data are sub-parallel with Mission Creek (MIS) distinguished by higher concentration ranges. Winfield samples (WIN) show very similar Cu values but MIS and Lambly Creek (LAM) datasets have quite variable Cu concentrations. In addition, Te concentrations are higher in Winfield samples. Rh and Pd values show similar trends for all three locales whereas Pt is lower and more variable in Lambly Creek samples. Lambly Creek samples have notably low Bi and Hg concentrations.  89   A MDS plot of the Kelowna area samples using only trace element data is shown in Figure 28. Samples with similar trace element signatures plot close together on MDS diagrams. Correlations between 21 different trace elements were used to create this plot making it essentially a 21 dimensional representation as opposed to a bivariate elemental plot which is 2 dimensional.   Mission Creek and Winfield samples plot together although with some scatter. Lambly Creek samples plot in two discrete fields with the smaller field containing two analyses from the same gold grain (LAM7).                        90 Table 9. Average trace element results from LA-ICP-MS analysis of Kelowna area gold samples. All values are reported in ppm. Quantitative values for 18 elements are listed. Si, Ca, and Cr data are not reported due to significant background signal interference or because of possible inhomogeneity in the AuRM2 standard. Where multiple isotopes of the same element were analyzed, the concentration for the isotope with the most reliable signal is reported. Semi-quantitative results are listed at the bottom of columns. Methodology for calculating these values is outlined in Section  Element LAM MIS WIN Mg 0.9 13.7 2.1 Al 6.8 1.3 20.4 Ti 0.5 15.4 9.8 Mn 1.8 3.1 10 Fe 7.5 46.9 15.9 Ni 0.1 1.2 0.4 Cu 97.9 43.3 68.7 Zn 0.7 16.7 1.3 As 9.5 12.3 18.5 Se 8.4 32.7 12.3 Rh 0.04 0.05 0.06 Pd 0.2 0.9 1.8 Sn 0.2 0.3 0.4 Sb 11.8 10 8 Te 1 1 2.6 Pt 0.6 1.9 1.5 Pb 0.1 0.04 0.2 Bi 0.03 0.2 0.1     Semi-quantitative Results  V 0.08 0.2 1.3 Hg 68.7 3088 1247 U 0.1 0.6 0.5  91           Figure 27. Probability plots of select trace elements from Kelowna area samples (concentrations in ppm). These types of plots are created by ordering individual analyses from lowest to highest concentration for each element. A normal score of 0 represents the median value.   92                   Figure 28. MDS plot of Kelowna area samples. Axes are unitless and indicate degree of similarity or difference of trace element patterns using correlations between all trace element data available (Au and Ag not used). Fields illustrate that Lambly Creek (LAM) analyses plot in two separate fields apart from Mission Creek (MIS) and Winfield (WIN) analyses.    93 4.3 Elemental Variation of Placer Gold in Combined Sample Sets This section presents results from the entire placer gold sample set but focuses on the remaining (non-Kelowna) placer gold grains.   4.3.1 Major Element Results for Combined Sample Sets SEM-EDS analysis was conducted on the placer gold samples to determine major element abundances including Ag concentrations, used as the internal standard for subsequent LA-ICP-MS analysis, and Hg which was only detectable in four samples. Averaged data by locality and position in gold grains was calculated (Table 10).  The Piaba nugget and Prophet Mine samples have the most Au-rich cores whereas the northern B.C., gold and electrum, grains have the most Ag-rich cores. Enrichment in Au on grain rims is evident for all localities except for the northern B.C., gold and electrum, grains and the Piaba nugget. The only samples that contain Hg concentrations above the SEM-EDS detection limit come from northern B.C. and the Prophet Mine with detectable Hg limited to the cores of the Prophet Mine samples.    Several primary inclusions in samples from the Prophet Mine were large enough to analyze for major elements. Minerals identified were pyrite: 53.8 wt% S, 46.2 wt% Fe (stoichiometric chemical formula FeS2), and nickeloan pyrite: 54.2 wt% S, 44.2 wt% Fe, 1.6 wt% Ni [stoichiometric chemical formula (Fe0.97Ni0.03)S2].  Box plots (Figure 29) reinforce the conclusion that the purest hypogene gold comes from the Piaba nugget and the Prophet Mine. Also evident is the wide variation in composition in the northern B.C. samples and the remarkable homogeneity of the Piaba nugget. The Au and Ag box   94 Table 10. Average major element results for all sample sets analyzed by SEM-EDS. All concentrations in wt%; n.d. = not detected. Abrv. Average Core Au Average Core Ag Average Core Hg Average Rim Au Average Rim Ag Average Rim Hg KEL 84.1 15.9 n.d. 91.5 9.5 n.d. AUS 90.4 9.4 0.2 94.5 5.5 n.d. BC 77.0 22.3 0.7 76.7 22.5 0.8 COL 83.3 16.7 n.d. 94.4 6.6 n.d. PIABA 92.0 8.0 n.d. 92.0 8.0 n.d.                   95 plots are almost the identical inverse of each other indicating that Au and Ag are the only significant components at the major element scale. A ranked probability plot (Figure 30) illustrates the distribution of Au and Ag concentrations in grain cores at each locality. At the high Au end (85-95 wt%) both the Prophet Mine and Piaba nugget have similar median values (median is located at a normal score of zero), but exhibit diverging trends at the margins of the plot. Piaba analyses show a flat trend that can be explained by the fact that all analyses come from a single nugget but the Prophet Mine analyses reflect 16 different samples.   The moderate Au concentration (75-90 wt%) analyses from the Kelowna area and Nus River, Colombia exhibit a very similar trends with slight divergence between the two at the median values (Figure 30). However, there is an anomalous Kelowna analysis at 97.5 wt% Au which may represent sampling of supergene gold. Analyses of the northern B.C. gold grains have a unique trend extending down to very low Au concentrations of 50 wt%.     96     Figure 29. Box plot of major element data for grain cores by locality. All concentrations reported in ppm. Colours correspond to different sample sets. Central dot in each box represents the mean, horizontal line is the median value, solid box corresponds to 1 sigma uncertainty, outer bracket is 2 sigma uncertainty, and individual points outside of that are considered outlier data points.   97    Figure 30. Ranked probability plots of major element concentrations (core analyses only) by location. Kelowna area samples (Lambly Creek, Mission Creek, and Winfield Mine) are combined. A Au-rich trend is observed for samples from Piaba, Brazil and Prophet Mine, Australia. Kelowna samples and those from Nus River, Colombia have a trend similar to each other. The Northern B.C. placer grains show a very distinct trend extending down to 50 wt% Au. The Ag trend is essentially the inverse of the Au graph.  98 For the Prophet Mine samples, major element concentrations within cores of grains were homogenous except for minor occurrences of primary inclusions of pyrite with variable substitutive Ni contents ( Figure 31). Backscatter electron images of the red-orange films coating the Prophet Mine gold grains (Figure 32) show the presence of abundant Au-rich spheres (inferred Au-rich due to very light colour indicating high density). The matrix of the film on the surface of the gold grain containing the Au rich particles is darkly coloured indicating a relatively low density. The spherical gold particles range in size from less than 0.1 micron (the minimum resolution visible in the image) up to 1 micron. Closer to the surface of the gold grain there are a greater number of gold particles with a coarser size distribution.  SEM-EDS line scans across Prophet Mine gold show an enhanced gold signal from the coatings on concave features of the grains. In these regions, the Ag signal is essentially nil (Figure 33; Left). Some margins have multiple peaks with a high Au signal (with corresponding dampened Ag signal) (Figure 33; Right). The cores of the grains inside of Au signal spikes have relatively constant signals for both Au and Ag though there is a slight change in element signal in the mid left of AUS4 (Figure 33). This corresponds to a slight change in backscatter electron image colour and is an artefact of bevelling of the gold grain edge during sample polishing.        99   Figure 31. Backscatter electron SEM image within AUS8 gold grain. Darker region in center is a Ni rich (1.62 wt% Ni) pyrite inclusion. Also visible are two gold inclusions within the nickeloan pyrite inclusion.      100 Figure 32. Backscatter electron image of biofilm on Prophet Mine gold grain AUS11. Spherical particles of biogenic gold at nanometer-micrometer scale occur on surface of the gold grain. Supergene gold has accumulated, growing the grain.    Bacterial Biofilm on Gold  101  Figure 33. SEM-EDS linescans of AUS4 gold grain. Image on the right is a complete grain transect. One Au rich peak is present on the left margin and at least 5 discrete peaks occur along the right margin of the grain. Left hand image is a separate linescan conducted at a more detailed level which shows that the Au rich signature corresponds with a film along the grain margin. The Ag signal is essentially zero along this film.  102   A two dimensional elemental SEM-EDS map was created for Prophet Mine gold sample AUS 16 (Figure 34). These images show an elevated signal for Au on the grain rim; however, the elevated signal is not uniform, being more pronounced along embayment and crevice features. The Au and Ag signal difference on the margin of the grain is most visibly pronounced in the isolated Ag signal in green (Figure 34; Mid-Right). Also visible in AUS16 is an approximately 200 micron long anhedral primary pyrite inclusion. Internally, the gold grain appears to be relatively homogenous with constant Au and Ag signals.  Line scans across the margins of Colombian gold grains confirm the presence of Au-rich rims, typically 10 µm, but locally up to 20 µm thick (Figure 35). However, no evidence for Au-bearing films was observed on the surface of grains. Internally, element signals are consistent (Figure 35). A SEM-EDS line scan of the Piaba nugget shows no evidence for the development of a Au-rich rim with Au and Ag signals remarkably constant across the nugget (Figure 36). The slight positive slope of Au and Ag signals reflects tilting of the sample in the instrument.  Internally the northern B.C. grains were quite homogenous but with significant major element variation between grains; there is no evidence of Au-rich rims (Figure 37). Signals for the detrital materials show elevated Si, Al, and O typical of common silicate minerals.      103                        Figure 34. Top Left: Composite SEM-EDS elemental map of AUS16 gold grain. Results are semi-quantitative but do show the distribution of elements. Note the increased signal for Au (red) along grain margins. Top Right: Backscatter electron image of the same gold grain. Note the ~200 micron long pyrite inclusion at the bottom of the grain. Bottom Right: Images of individual element signals. The Ag signal is notably lower in select areas of the grain margin including the embayment along the top left of grain. Fe signals are essentially zero on the outside of the grain with constant values within except for a bright feature at the bottom which corresponds to a pyrite inclusion. Oxygen values (purple) correlate with both detrital and possibly primary silicate mineral inclusions.     104                Figure 35. SEM-EDS linescan of margin of COL1 gold grain. A nine micron thick Ag-poor rim is present.  105         Figure 36. SEM-EDS linescan across Piaba nugget confirming the absence of a Au-rich rim. Overall the Ag signal is low throughout grain. The slight slope of the Au signal is likely due to the sample not being perfectly perpendicular to the electron beam.   106  Figure 37. SEM-EDS linescan of BC3 gold sample. Note that through the dark coating on the grain (back-scatter image) there is no gold signal (lower graph). Also, there is no Au-rich rim on the grain. Both Au and Ag signals rise concurrently on entering the grain.  107 4.3.2 Trace Element Results for Combined Sample Sets Over one hundred LA-ICP-MS analyses, restricted mostly to the cores of the gold grains, were conducted with average elemental abundances for 18 trace elements at each locality reported in Table 11 (Si, Ca, and Cr are not reported due to inhomogeneity or poor signal to background ratios; see Section 4.1; full trace element dataset can be found in Appendix B).  Features that stand out are the high semi-quantitative Hg concentrations of the Prophet Mine and northern B.C. gold grains. Also notable are high Sb concentrations in northern B.C. placer grains (758 ppm), and the low Fe concentration in Nus River gold samples. The Piaba Nugget exhibits high Cu with almost all other trace elements at very low levels. The combined averages for the Kelowna area samples have the highest Fe and Se concentrations. The highest U concentrations (semi-quantitative) occur in Kelowna area samples and the Prophet Mine, Australia.  A graphical representation of the analytical results (Figure 38) shows the distribution of elemental concentrations. There are few characteristics that uniquely distinguish localities. There is a significant spread in the concentration of elements at most localities. A notable exception is the Piaba nugget where all analyses are quite similar due to the fact that all are from a single grain. Aluminum appears high in Winfield samples whereas both Mission Creek and Winfield samples have high Ti. Mission Creek analyses show the highest Fe and Ni values. The northern BC grains have elevated Sb with a wide overall range. Se, Te, and Pt values show little variation between analyses for any one locality. Mission Creek and Winfield samples have elevated V (semi-quantitative). Mercury is distinctively higher in the northern B.C. and the Australian samples and overall it appears to be the most abundant trace element in most placer gold samples.    108 Table 11. Average trace element concentrations by locality separated by quantitative and semi-quantitative elements. Element AUS BC COL PIABA KEL Mg 9.6 0.5 0.1 0.04 3.0 Al 7.4 2.9 0 0.03 5.9 Ti 0.6 1.6 0.1 0.05 4.1 Mn 0.7 2.0 1.3 0.6 2.6 Fe 10.4 23.3 11.5 4.2 22.8 Ni 0.4 0.2 0.2 0.2 0.3 Cu 152 219 36.6 227 69.8 Zn 0.5 0.7 0.4 0.12 2.9 As 1.9 6.8 4.6 1.0 10.1 Se 9.7 26.2 11.8 5.2 17.3 Rh 0.03 0.05 0.02 0.02 0.04 Pd 0.14 0.4 0.16 0.12 1.1 Sn 0.04 0.15 0.07 0.03 0.2 Sb 6.2 758 0.2 0.09 9.8 Te 0.7 2.3 1.3 0.6 1.6 Pt 1.2 2.0 0.7 0.9 1.4 Pb 0.14 0.08 0.03 0.02 0.12 Bi 0.08 0.3 0.04 0.01 0.12  Semi-quantitative Results  V 0.14 0.16 1.90 0.14 0.38 Hg 11407 14421 137 198 207 U 0.54 0.11 0.11 0.08 0.38  109 Bivariate plots of trace element concentrations versus Au show that high purity placer gold contains low concentrations of trace elements (Figure 39) whereas Ag-rich and electrum grains contain high trace element concentrations. The only exception is Cu which shows a trend to high concentrations in Au-rich placer gold grains.   Ranked probability plots (Figure 40) show that Fe concentrations have a discrete range at each locality with Mission Creek and the northern B.C. samples showing the highest abundances. The Prophet Mine, Nus River, and most of the Lambly Creek samples have moderate Fe, and the Piaba nugget has the lowest concentrations. Cu values show wide variation within localities with Piaba concentrations consistently high. The northern B.C. grains form two distinct groups with low Cu in the electrum grain (4 analyses) and high concentrations in the other grains. The Prophet Mine samples have a smooth continuous distribution of Cu from around 20 to 400 ppm. Mission Creek, Winfield, and Nus River have intermediate Cu concentrations. Samples from Lambly Creek show a sharp break in trend, most analyses have low Cu values but two samples exhibit high Cu.  Selenium distributions reveal differences between localities. Both Mission Creek and northern B.C. samples have high Se trends. Winfield, Lambly, Australian, and Colombian samples have moderate concentration trends, and the Piaba nugget has consistently low Se.   Antimony values for many sample sets show a stair step pattern (Figure 40). Overall absolute concentrations of the element extend 5 orders of magnitude from 0.1 ppm to over 1,000 ppm (0.1 wt%). Tellurium appears useful for separating sample sets with the northern B.C. and Winfield samples showing the highest values. Mission Creek, Nus River, and Lambly Creek have intermediate concentrations and the Prophet Mine and Piaba nugget exhibit the lowest Te concentrations.   110 Platinum values are discrete for most sample sets as evidenced by their very flat trends. Mission Creek, Winfield, and northern B.C. samples have the highest concentrations. The Prophet Mine distribution trend has a flat low concentration range that transitions into a steep high concentration range. Piaba, Nus, and Lambly Creek have distinctively low Pt concentrations with one anomalously low value from Lambly.   Bismuth shows a wide concentration range with a stair step trend for many sample sets resembling the Sb pattern. Vanadium concentrations are semi-quantitative but appear to mimic the Fe trends for most sample sets. Uranium concentrations are highest for Mission Creek, Winfield, and a couple of anomalous Prophet Mine samples. Semi-quantitative Hg values appear very successful at separating sample sets with the Prophet Mine and northern B.C. samples showing highly elevated concentrations. Piaba, Mission Creek, Winfield, and Nus samples have moderate Hg levels whereas Lambly Creek samples have the lowest abundances.    111  Figure 38. Common axis box plots of all quantitative and semi-quantitative trace element data by locality. Mg-Bi values are quantitative and V-U values are semi-quantitative. Circles in box plots represent means whereas lines correspond to the median value. Solid box is the one sigma field and the whisker extends out to the two sigma field. Isolated points outside two sigma are classified as anomalies.   112  Figure 39. Selected bivariate plots of trace elements in placer gold versus Au concentration. Grey lines are least-squares trend lines fitted to the data. Au concentrations are from SEM-EDS analyses and are on a linear scale. Trace element data are plotted on a log scale. All trace elements exhibit a negative slope versus gold (including elements not shown here) except for Cu which had a positive slope with high Cu concentrations occurring in high fineness gold.   113 Figure 40. Ranked probability plots of select trace elements. Elements are listed in order of increasing mass from Fe to Bi. V, U, and Hg values are semi-quantitative. Note y-axis log scale.  114   Ranked probability plots are essentially one dimensional. To help differentiate trace element signatures, selected bivariate plots are presented to demonstrate which element combinations are most successful in separating sample populations or reveal interesting correlations within the entire data-set. An Fe vs. Cu plot (Figure 41) and log Fe vs. log Cu diagram (Figure 42) shows that non-electrum northern B.C. gold grains distinctly cluster at relatively high Fe and Cu values. The Lambly Creek samples are distinctly separated into two groups: one with low Cu and Fe and one with low Fe and high Cu.  On an Fe vs. Pt plot (Figure 43) the two siderophile elements show an overall positive correlation (R2 = 0.59), and sample groups show some clustering. Two general trends occur in Prophet data and a subset of the Lambly Creek samples show notably low platinum.               115                        Figure 41. Bivariate Fe vs. Cu plot. Linear scales. Fields show clustering of samples by locality. Two distinct groups are present for northern B.C. samples (pink), and only a subset of Prophet Mine samples (blue) trends to high Cu.   116   Error! \                      Figure 42.  Bivariate Fe vs. Cu plot. Log-log scale (see Figure 41 for legend). Hand-drawn fields enclose groups of samples by locality.   117                      Figure 43. Bivariate Fe vs. Pt plot. Log-log scale (see Figure 41 for legend). Hand-drawn fields enclose groups of samples by locality. All northern B.C. samples (pink) are grouped together and there are two separate groups of Prophet Mine samples (blue).   118 An Fe vs. V plot (Figure 44) shows a shallow sloped positive correlation for most analyses (R2 = 0.90) but 6 analyses from Winfield and the Prophet Mine exhibit a steeply sloping linear trend (R2 = 0.98). The equations (Equation 7; Equation 8) for these trends are:  Equation 7. Linear slope equation from Figure 44 showing the relationship between Fe and V abundances for most samples.   Equation 8. Linear slope equation from Figure 44 showing steeper V-rich linear relationship for 5 samples from Winfield and Prophet Mine.    Equation 7 and Equation 8 yield Fe:V ratios of 67:1 and 5:1 respectively. The bulk of the analyses follow Equation 7 including the anomalous Mission Creek samples with high Fe. The two Prophet Mine analyses following Equation 8 are classified as from regions of grains that are hypogene in origin (due to presence of primary mineral inclusions; see Section 0).  A strong positive correlation (R2 = 0.91) is observed between Fe and Se concentrations in the entire sample set (Figure 45). Again the Piaba nugget exhibits the lowest concentrations while Mission Creek, Winfield, and the northern B.C. samples have the highest concentrations.  Two diverging trends are observed on a log-log Pd vs. Sb chart (Figure 46). One sample subset including the Piaba Nugget, some Prophet, some Lambly, some Mission, some Winfield, and the B.C. electrum grain have a shallow slope with the remaining samples exhibiting a relatively steep slope. While not a perfectly correlated trend this divergence is notable.       119          Figure 44Figure 44. Bivariate Fe vs. V plot. Linear scale; V concentrations are semi-quantitative. Two               separate trend lines with slopes given in Equation 7 and Equation 8.  120            Figure 45Figure 45. Bivariate Fe vs. Se plot. Log-log scale (see Figure 47 for legend). Hand-drawn fields  enclose groups of samples by locality. For the entire suite of samples there is a positive correlation (R2 = 0.91).  121  A bivariate plot of two PGEs (Figure 47) shows a positive correlation between Pd and Pt concentrations with significant scatter (R² = 0.56). Analyses significantly off this line tend to be shifted to the right toward Pd enrichment. One Winfield sample is anomalously abundant in Pd with over 10 ppm. A Pt vs. Hg bivariate plot is very good at separating samples by locality (Figure 48). The Piaba nugget falls in a discrete area whereas Prophet Mine samples show wide distribution. Northern B.C., Winfield, and Nus River samples display a somewhat negative correlation whereas Mission Creek samples have a positive correlation between Pt and Hg.  Previous researchers have utilized ternary diagrams to classify natural gold samples (Section 1.1). On the Townley et. al. (2003) Au-Ag-Cu diagram, samples for this project plot along the Au-Ag line (Figure 49), due to relatively low Cu concentrations, they fall within the gold rich porphyry and epithermal gold deposit fields (See Figure 2 for fields). A modification to the Au-Ag-Cu, in which Au concentrations are divided by 10, to illustrate the variation in Cu concentrations (Figure 50) reveals an overall arc-shaped trend with modest separation of sample sets. On a Ag-Au-Hg diagram (Figure 51) all samples show low Hg and plot close to Au on the Au-Ag side of the plot. Comparing this plot with the equivalent phase diagram by Tomkins (Figure 2) reveals that all samples for this project plot below the solidus line of Ag-Au-Hg solid solution at 650 degrees Celsius.  An Fe-Cu-Hg diagram (Figure 52) reveals three distinct sample populations. At the Fe-rich end are Nus River and Kelowna area samples (except two Lambly Creek). At the Cu-rich end are the Piaba Nugget, three Prophet Mine analyses and the two Lambly Creek samples. At the Hg-rich end are almost all of the Prophet Mine samples and all of the northern 7 B.C. samples.     122   Figure 46. Bivariate Pd vs Sb plot. Log-log scale. Two linear trends (y = 3884.4x - 696.38; R² = 0.2737  and 0.205x + 0.328; R² = 0.40). See Figure 44 for legend.   123 Figure 47. Bivariate log-log Pd vs Pt plot. Some correlation with several samples shifted off of the trend to higher Pd values. See Figure 44 for legend.              124            Figure 48. Bivariate Pt vs Hg plot. Log-log scale. Hg concentrations are semi-quantitative. See Figure 44 for legend. Hand-drawn fields show distinct grouping by locality; Winfield and Mission Creek plot together whereas Lambly samples plot in two separate fields.    125                          Figure 49. Analyses plotted on a Au-Ag-Cu ternary diagram; modified from (Townley et al., 2003).   126  Figure 50. Analyses plotted on a Au-Ag-Cu ternary diagram similar to Figure 49 but with Au values divided by 10 to aid in showing the Cu trend.  127  Figure 51. Analyses plotted on a Ag-Au-Hg ternary diagram. Hg concentrations are semi-quantitative. Compare with Figure 2 which shows phase relationships (after Tomkins et al. 2004).  128                         Figure 52. Fe-Cu-Hg ternary diagram. Hg concentrations are semi-quantitative. This ternary separates analyses into three distinct populations (hand-drawn fields).   129 Multi-dimensional scaling plots (see Section 2.7 for methodology) were created for the entire project sample set using all quantitative trace element data (Figure 53). Sample groupings solely reflect trace element signature. Analyses plotting proximal to one another have similar overall trace element signatures and those plotting far apart have dissimilar trace element signatures.  Figure 53 shows that samples from individual localities tend to plot together indicating overall similar chemistry. However the diagram implies that gold from the Prophet Mine (AUS) shows diverse composition. Kelowna area gold is discussed in Section 4.2.2.  On an MDS plot comparing samples from northern B.C., Nus River Colombia, and the Piaba Nugget from Brazil (Figure 54) each of the sample sets plot in mutually exclusive zones.  A MDS diagram comparing Prophet Mine analyses (Figure 55) shows clustering of samples in the upper portion of the graph (dashed circle). Gold samples observed to contain primary mineral inclusions (indicating some hypogene component) plot within this cluster (indicated by arrows).  Note that multiple analyses coming from single grains tend to plot near one another (eg. AUS5, 5B, 5C).      130  Figure 53. MDS plot comparing all samples. Axes are dimensionless scales indicating overall similarities and differences between trace element signatures of samples. Elements used to create this plot include all quantitative trace elements (major elements Au and Ag not used).     131  Figure 54. MDS plot showing the northern B.C., Nus River, and Piaba nugget analyses. Only quantitative trace element data were used in the calculations. Hand-drawn fields show exclusive grouping of samples by locality (two separate fields for northern B.C. samples).   132                         Figure 55. MDS plot of the Prophet Mine sample set. Dashed circle highlights minor clustering of samples. Arrows indicate grains observed to contain primary mineral inclusions.  133   To better understand what characterizes the Prophet Mine sample subsets in Figure 55 analyses were divided into two groups. The first, made up of the analyses clustered in the dashed circle, are classified as inferred hypogene. The second group exhibiting scatter is classified as containing some component of a supergene signature. Average concentrations are listed in Table 12. Inferred hypogene analyses have higher Ni, Zn, Pb, V, and Hg whereas inferred supergene analyses have significantly higher Sb, and Bi.  A Se vs. Sb bivariate plot shows that inferred hypogene analyses have low Sb (<0.4 ppm) and ~ 5-10 ppm Se (Figure 56). Due to the narrow rim of supergene gold on grains, no analyses were of material completely within the supergene rim. Analyses that contain some supergene gold component form a grouping at high Sb (2-16 ppm) with a correlated increase in Se. The analyses with an inferred supergene component that do not follow this trend may have only a minor supergene volume masked by a majority hypogene component.              134 Table 12. Average element concentrations for two sample sub-groups of Prophet Mine analyses. All values are reported in ppm.  Inferred Hypogene Inferred Supergene Fe 10.2 10.6 Ni 0.94 0.04 Cu 101.5 182.4 Zn 1.00 0.18 As 2.04 1.89 Se 8.72 10.2 Rh 0.03 0.03 Pd 0.12 0.16 Sn 0.05 0.05 Sb 0.17 9.87 Te 0.57 0.78 Pt 0.95 1.38 Pb 0.32 0.04 Bi 0.01 0.12 V 0.27 0.07 U 0.20 0.75 Hg 14987 9259  135               Figure 56. Bivariate Se vs Sb plot of Prophet Mine sample set separated into inferred hypogene and supergene subsets. No analyses were completely within supergene gold therefore “inferred supergene” indicates any supergene proportion.  136 Chapter  5: Discussion Discussion is separated into three sections corresponding with the results sections: AuRM2 reference material (Section 4.1), Kelowna area samples (Section 4.2), and elemental variations in the entire collection of samples (Section 4.3).   5.1 AuRM2 Reference Material The AuRM2 standard has not been previously used for micro-analysis of natural gold samples, and, therefore, its characterization at the micro-scale is important to help guide future trace element studies utilizing this standard. AuRM2 was assessed to be suitable for this type of analysis by quantifying precision (Section 4.1.2), and ultimately homogeneity (Section 4.1.4). The accuracy of elemental concentrations (Section 4.1.3) was verified.  Matrix dependent fractionation effects during LA-ICP-MS were calculated ( and determined to be of negligible effect for semi-quantitative determination of trace elements. This study adds value to the standard reference materials analyzed by providing quantitative and semi-quantitative concentrations for additional trace elements not previously reported (Sections 4.1.3;   5.1.1 Precision of AuRM2 Analyses Reproducibility of replicate AuRM2 analyses (Table 5) reveals that, of the 21 trace elements with certified homogeneity at a macroscopic scale, seven have precision (%RSD) better than 10% (Mg, Al, Ti, Ni, Cu, Zn, and Sn), nine are between 10 and 15% (Fe, As, Se, Rh, Pd, Sb, Pt, Pb, and Bi), two fall between 15 and 20% (Mn and Te), and three are between 30 and 40% (Si, Ca, and Cr). Silver was added to AuRM2, but precision cannot be estimated because it was used as the internal standard during LA-ICP-MS analyses. In addition, there are four elements present in AuRM2 at  137 detectable levels, but not intentionally added during manufacture, and therefore not certified for homogeneity at any scale (P, V, Hg, and U). They have precision values of 12%, 16%, 34% and 360% respectively. The precision estimates for most elements (10 to 15%) reflect modest signal (background-corrected cps) for these elements as a result of low concentrations in the AuRM2 standard (Figure 16; Table 4) compared to the nominal 400 ppm concentrations for most elements in NIST610. The moderately elevated RSD values of Mn and Te are primarily due to low signal intensity, but are acceptable for the purposes of this study.  The poor precision of Si, Ca, and Cr reflects extremely low background corrected cps (Figure 16) as a result of poor signal-to-noise ratios; utilization of low abundance isotopes to avoid interferences on higher abundance isotopes (eg. 28Si and 40Ca), and low concentrations for most elements in AuRM2 resulted in low cps (and thus low precision). Attempts were made to use these data, but they provided no information on relationships between samples, and no geochemically explicable relationships between elements, so they are eliminated from discussion. Higher signal, measurement of alternate isotopes, lower background, or reduced interference could provide a suitable signal; therefore, their use in future studies is not precluded.  The sample ablation time for this study was limited to a conservative length to avoid fractionation effects from changes in pit depth which begins to occur at a depth to diameter ratio of ~0.5 (Potts et al., 1995). In retrospect, a longer dwell time would have provided more volume of material and therefore greater sensitivity and reproducibility without a pit depth great enough to cause elemental fractionation. A value of 0.3 µm depth per laser shot (Henry, 2013; pers. comm.) was used to estimate optimal ablation time. However, in a gold matrix, the actual value is approximately 0.06 µm depth per laser shot; this is one fifth of the assumed value.  138 As a validation for these interpretations, precision for elemental data from the University of Windsor GLIER Lab, which are the only other micro-analytical data for AuRM2, is compared with UBCO data on a signal versus RSD graph (Figure 57). The trends of signal versus RSD are very similar between labs. Overall the Windsor data trends to higher signal and better constrained RSD. This is perhaps indicative of greater sample mass analyzed and of the fact that sampling was done by laser rastering, which averages a greater spatial area and is therefore more representative of the bulk sample. Spot analyses, such as conducted for this study, are better suited for quantifying heterogeneity. The Cr data are far better constrained for Windsor analyses and reliable concentrations could be determined in samples there whereas they could not be reliably determined in this study. Their results may reflect the use of 52Cr as opposed to 53Cr or reduced background signal from sample cone contamination.  The reproducibility of AuRM2 element concentrations provides a metric for expected reproducibility of duplicate sample analyses of natural gold. In general, most elements exhibit variation due to analytical precision that is negligible compared to natural geochemical variation in placer gold systems (Figure 38).  5.1.2 Homogeneity   The results in Section 4.1.3 indicate that most elements in the AuRM2 gold reference material are homogenous at the sampling scale of the analytical method (beam diameter of 64 to 108 micrometers). Approaches used to assess homogeneity comprised both comparative (comparison to NIST610; Figure 16), and statistical methods (using counting statistics; Table 8), both of which are discussed below.    139               Figure 57. Chart of signal vs. RSD  comparing University of Windsor GLIER Lab reference material data with data collected for this project with individual elements labeled.   140 Assessment of AuRM2 Homogeneity by Comparison to NIST610 Trace elements in NIST610 are distributed with sufficient homogeneity for analysis by most methods of micro-analysis (Jochum et al., 2011). As such, it is expected that if AuRM2 trace element analyses exhibit a correlative pattern of reproducibility to NIST610 it would indicate that AuRM2 is also suitable for this method.  Reproducibility of analyses is strongly signal-dependent (Potts et al. 1995; Jochum et al., 2011) and, therefore, cannot be directly compared between the two standards. This is due to the lower nominal concentration levels in AuRM2 (Table 4), which affects the mass of each element sampled in the standard (Table 7) and, therefore, the analytical signal. Given this, the relative homogeneity of each element in AuRM2 can be compared to the homogenous NIST610 glass by comparing reproducibility taking into account the background-corrected cps (analytical signal) of each element (Figure 16).  Measurements of NIST610 trace element concentrations are overall more reproducible than AuRM2 for most elements (Table 5; Figure 16). The field of reproducibility for NIST610 on a RSD vs. signal plot (Figure 58; within red dashed lines) represents a baseline for homogenously distributed elements. Elements in AuRM2 contained within this field, and thus indicated to be homogenously distributed, are: Zn, Ni, Mg, Cu, Sn, Al and Ti. Both Fe and Se are within 1% RSD above the NIST610 (homogeneity) trend and are also considered to be homogeneously distributed. Elements between 1% and 5% RSD above the trend of homogeneity are considered to exhibit minor heterogeneity and include: As, Sb, Pd, Pt, Bi, Pb, and Rh. Above the 5% RSD line but less than 10% above the homogeneity trend are Te and Mn which are classified as exhibiting moderate heterogeneity.    141 5% RSD above NIST610  NIST610 Reproducibility  (Homogenous)  (Minor Heterogeneity) (Moderate Heterogeneity) 1% RSD above NIST610   Figure 58. Reproduction of Figure 16 with field of NIST610 reproducibility shown within red dashed lines (logarithmic equation of red solid line given). Black dashed lines define homogeneity classifications of AuRM2 (less than 1% RSD above NIST610 = homogenous; between 1% and 5% RSD above NIST610 = minor heterogeneity; more than 5% RSD above NIST610 = moderate heterogeneity). Si, Ca, and Cr are not included due to analytical difficulties in measuring concentration.     142 Assessment of AuRM2 Homogeneity using Counting Statistics Counting statistics define the inherent variability expected in analytical measurement as described in Section 4.1.4. The effect of counting statistics on uncertainty in the ability to measure element homogeneity (Table 8) limits elements determined to be heterogeneously distributed by comparison to NIST610 in the previous section to those which have non-zero heterogeneity RSD; these elements are: Mn, Rh, Pd, Pt, Pb, and Bi. For these elements, a quantitatively determined heterogeneity can be reported (Table 8) and it is certain that measured heterogeneity is not due to the effect of counting statistics. Semi-quantitatively determined elements in AuRM2 are V, Hg, and U. There are no accepted concentration values for these elements, however, their level of homogeneity can be, and was, determined. Heterogeneity RSD was calculated to be: 0% (homogeneous), 33.3%, and 357% respectively (Table 8).  The apparent heterogeneity of the PGEs is surprising given that these noble metals tend to form alloys with gold (also a noble metal). Pb and Bi are similar in that they are both chalcophile in nature and are the heaviest quantitatively measured elements. While heterogeneities are measurable, most are well below 10% at this scale of analysis, which is acceptable. The fact that Hg and U are grossly heterogeneous isn't surprising as they were not intentionally added to the standard and as such are present as impurities. Vanadium was not added but is apparently homogeneously distributed. Possible Sources of Heterogeneity Examination of the AuRM2 elemental MDS diagram (Figure 17) helps elucidate the nature of heterogeneity in AuRM2. It is expected that a homogenous standard will have an elemental MDS diagram distributed essentially randomly with no association between elemental abundances in  143 individual analyses. Indeed, an elemental MDS plot of NIST610 analyses (Figure 59) exhibits apparently random element distribution confirming trace element homogeneity. If, however, inhomogeneity is indicated, elemental associations may indicate the composition of inclusions, chemical gradients, contamination, or analytical fractionation. This is due to the fact that certain groups of elements have similar inherent chemical characteristics and tend to follow each other during processes. This is seen in natural systems such as igneous processes where fractionation occurs during melting and crystallization. A similar situation may be expected in a manufactured gold reference material if it is not solidified instantaneously. For example, if micro-telluride inclusions were present in a gold standard, tellurium and other chalcophile elements may be expected to cluster together on a MDS diagram.  Elemental variation can also be introduced from the container, the atmosphere, or the method of stirring (Jochum et al., 2011).        144      Figure 59. Elemental MDS plot of NIST610 analyses (n = 32) showing no clear grouping by element. This is indicative of a homogenous standard.  145 AuRM2 element groupings are present when MDS plots are created separately by individual lab and when both lab analyses are combined (as in Figure 17). This means that these element patterns are not an artifact of combining data from different labs or from fractionation in an individual analytical set-up. This leads to the conclusion that, either there is irreconcilable elemental fractionation when ablating a gold matrix, or some elements are heterogeneously distributed in AuRM2.  The most striking feature of the AuRM2 MDS plot (Figure 17) is the grouping of the PGEs. This means that when one PGE concentration is high in a single analysis the other PGE concentrations are also likely to be high in that analysis. This may indicate the presence of micro-inclusions of PGE alloy phases; the similar properties of gold to the other noble metals enables alloying quite easily however, which makes inclusions of a separate PGE phase unlikely. Alternatively, the presence of a spatially controlled gradient in PGE concentrations within the standard could explain the grouping. However, such a gradient is not observed in macroscopic assessment of homogeneity (Murray, 2009).  Chalcophile elements are grouped at the opposite end of the chart from the PGEs. Pb and Bi, which were determined to exhibit minor heterogeneity, are included in this chalcophile group. This may indicate a minor phase of chalcophile-rich inclusions in AuRM2  Since analyses of the external standards for this project were limited to 108.5 micron beam spot sizes the homogeneity of the standard can only be determined at the 108.5 micron and corresponding 3.4 microgram sample mass range. To quantify the limit of the scale of homogeneity analyses would have to be conducted at progressively smaller sampling masses until heterogeneity is observed. For elements inhomogeneous at the 108.5 micron-scale, analyses would have to be conducted at progressively increasing sample masses until sufficient homogeneity was observed.   146 5.1.3 Accuracy of AuRM2 Concentration Values The accuracy of this micro-analytical method was determined by comparison to known concentration values for AuRM2 measured by bulk solution ICP-MS (provided by reference material suppliers). By treating the AuRM2 reference material as an unknown, and using FAU7 as an external standard, an acceptable level of accuracy was confirmed (Table 6) with almost all elements showing less variance (<10%) from given values than the variance in precision of analyses (plus or minus 15%) (Table 5).  New concentration values were quantitatively determined for elements in FAU7 by treating it as an unknown against AuRM2 as an external standard. Elements that have accepted concentrations in AuRM2 but no known value for FAU7 were measured to be: Se (24.9 ppm), Rh (0.1 ppm), Sb (0.3 ppm), and Te (2 ppm). Future gold studies using FAU7 as a reference material could measure these elements in samples by utilizing these values.   Using known concentration values the effect of matrix-dependent fractionation was also quantified (Figure 15) and determined to be of limited effect for determining semi-quantitative trace element concentrations. This allowed for the determination of semi-quantitative concentrations of additional elements in AuRM2 which do not have accepted concentration values. Matrix-Dependent Effects on Trace Element Fractionation  To quantify matrix-dependent fractionation effects during LA-ICP-MS analysis AuRM2 data was processed using NIST610 as the external reference (Table 6). Although elements that were major components of one of the matrices (Au in AuRM2; Si, Ca, and Al in NIST610) have extreme fractionation, trace elements are within 50% of given values for AuRM2 (Figure 15). Nickel, Cu, Rh, Sn, Pb, and Bi are better than 10%. This highlights the usefulness of matrix non-matched standards  147 for gold analysis to obtain semi-quantitative trace element results for numerous elements that are not at sufficient levels in existing gold reference materials. It is difficult to find two solid materials with such drastically different properties as silica-rich glass and gold yet trace element accuracies between the two appear to be acceptable for at least semi-quantitative analyses. Although a matrix matched standard is the ideal, NIST610 may be used to estimate semi-quantitative concentrations for many more elements in natural gold. This can be used as a reconnaissance tool to see which elements are detectable in natural gold. This information could direct future reference material manufacture for gold or any other non-matched matrices (ex: carbonates, sulfides, oxides). The limitation of significant fractionation to elements that are a major component of one of the matrices is likely because the signal of an element that is trace in one matrix has to be extrapolated to far higher concentrations in the other matrix and the instrument response curve (Equation 4) is non-linear when extrapolated that far. Semi-quantitative Element Concentrations The realization that matrix-dependent fractionation between NIST610 and AuRM2 is limited to 50% variation for trace elements allowed semi-quantitative determination of elements in AuRM2 that have no known values. Analyzing AuRM2, using NIST610 as an external standard, yielded measurements of: V (0.3 ppm) and U (0.1 ppm). Vanadium may be an important element for analysis of gold samples due to its redox sensitivity and measurement of U has implications for assumptions used in Pb isotopic signature analyses. There is a long list of additional elements that could be semi-quantitatively determined in AuRM2 using this method.  An initial interest for this study was the possibility that elements associated with biological activity such as P could fingerprint biologically precipitated gold. Phosphorous measurements of  148 AuRM2 were highly anomalous during analysis, however, and unacceptably high estimates of 822 ppm were calculated for AuRM2. As P was not added during preparation of the standard it is likely that this represents an experimental artifact. The signal intensities used for concentration calculation are gas-blank background-subtracted, therefore contamination or interference from the carrier gasses can be excluded. There are no isobaric or polyatomic interferences at monoisotopic 31P with mass of 30.9738 Daltons; 15N16O+, 14N16O1H+, 30Si1H+, and 62Ni++ are all extremely low or non-existent. Analysis at medium resolution mode did not reveal any separation of peaks either.  Volatile elements exhibit an increased intensity compared to less volatile elements during prolonged laser ablation due to fractionation effects (Fryer et al., 1995) however that is not likely the case here as ablation time was limited and pit depth to diameter ratios were <<1. Other possible sources of contamination include hardware such as gas tubes, sample cone, and extraction lens. The gas blank was observed to be uniform, however, and after background subtraction a strong signal was still evident. As sampling consisted of laser pits, surface contamination is likewise ruled out as a source of the signal.  This leaves an unexplained apparent ICP-MS instrumental background signal for P, but the results here are not unique. Müller et al. (2008) analyzed P in quartz samples and got anomalously elevated P values yet no source of contamination or interference could be determined. For the purposes of this study P analyses by LA-ICP-MS are unusable. There are no accepted concentration values for Hg in neither NIST610 nor AuRM2 although a clear Hg signal was detectable in both standards. As neither standard had a known Hg concentration a different approach was required to estimate concentration. Concentration was estimated by assigning rough values for Hg in external standards NIST610 and AuRM2 to predict Hg concentrations in 4 separate natural gold samples that had sufficient Hg abundance to be  149 determined quantitatively by SEM-EDS (Table 10). The rough values were cross referenced with the known values and adjusted accordingly; after several iterations the Hg values entered for both NIST610 and AuRM2 were able to successfully predict Hg concentrations in the gold samples quantitatively determined by SEM-EDS. Mercury was determined semi-quantitatively to be 0.6 ppm in NIST610 and 3.7 ppm in AuRM2.  A significant issue with these estimations is the extremely volatile nature of Hg. During laser ablation Hg is highly sensitive to thermal heating and may have originated not only from the crater but also the heated zone around it (Fryer, 2014; pers. comm.). A gold matrix, with far greater thermal conductivity than silica-rich glass, would be expected to have a greater volume of material affected by thermal heating. This indicates that AuRM2 may have yielded more Hg than NIST610. If this is the case, Hg concentration may be underestimated in NIST610.  The value of 0.6 ppm Hg from this study resembles the results of a semi-quantitative LA-ICP-MS analysis of NIST610 at the University of New Brunswick which used sulfide standards with known Hg concentrations as an external reference material. The experiment yielded a Hg concentration of 0.45 ppm (Dehnavi, 2014; pers. comm.). The close agreement across three drastically different matrix materials (metal, silica-rich glass, and sulfide) indicates that this is a robust concentration value for Hg in NIST610 and that matrix matching of materials has a relatively minor impact for reported trace element concentrations.       150 5.2 Kelowna Area Placer Gold The Kelowna area placer gold samples provide insights into the potential variability that can occur in trace element compositions at the district-scale. Because the geology and mineral occurrences in the area are well documented, it is possible to provide hypotheses for the origin of that variability. The Mission Creek and Winfield occurrences share a single hypogene gold source that has Au-rich rims precipitated on grains in a supergene environment, whereas the Lambly Creek occurrence has two separate hypogene gold sources with minimal to absent supergene Au-rich rims.   5.2.1 Winfield and Mission Creek The morphology, major element, and trace element data (Sections: 3.2.5, 4.2.1, and 4.2.2) all support a common source for the placer gold occurrences in Mission Creek and the Winfield Mine. Neither the Winfield nor Mission Creek samples have any evidence of primary mineral inclusions. They also have a similar size distribution range from flour gold up to millimeter scale flakes. Most grains exhibited a flattened and rounded morphology. All of these properties indicate an extended transport distance (Grant et al., 1991) for both Mission Creek and Winfield gold. In the case of the Mission Creek gold the extent of the current fluvial system cannot account for the transport indicated by morphology. Hypogene Signature of Mission Creek and Winfield The similarity in the major element composition of grain cores is indicative of a common source. Multiple methods of displaying the trace element signature of Mission Creek and Winfield gold including simple averages (Table 9), ranked probability plots (Figure 27), bivariate (Figures  41,  151 42, 43, 44, 46, and especially Figure 47), and MDS plots (Figure 28) show the same signature between the two.   The matching signature of Mission Creek and Winfield indicates that there has been significant reworking and recycling of gold on the east side of Okanagan Lake. The Miocene age paleoplacer gold deposits (Winfield type) have been eroded and reworked in at least one, but possibly more, cycles of erosion into contemporary creeks such as Mission Creek. This theory is also supported by the location of these gold deposits on the eastern side of the westerly dipping Okanagan normal fault. These deposits lay on the upthrown side of the fault which experiences more aggressive and extended erosion. The recycling of placer gold occurrences tends to lead to larger and higher grade placer occurrences over time (Garnett & Basset, 2005). Previously noted bedrock mineral occurrences are a starting point for tracking down the hypogene gold source. Bedrock geology, tectonic history, and the metallogenic endowment of the region may also indicate mineralization types that are likely to be the source of the placer gold.   There are two Mo porphyry type occurrences called Tick and Nova near Wood Lake which is by the Winfield paleoplacer occurrence (BC Ministry of Energy and Mines, 1993e). This is not likely the source for the gold however, as it is not a Au or Au-Cu porphyry. In addition, porphyry gold occurrences tend to have high Cu concentrations (Townley et al., 2003). The proximity does not explain the morphological indications of significant transport distance. Paleo-flow indicators in the northwest to southeast direction (Boyle, 1982) point to a hypogene gold source to the northwest.     Possible hypogene sources are extensive uneconomic quartz-carbonate vein-gold occurrences noted predominantly in greenschist facies of the Omineca Belt (Nesbitt & Muehlenbachs, 1995). The metamorphic grade of the Omineca Belt in the study area around Kelowna is amphibolite facies, however, and contains veins and quartz pods smaller in size at around 5cm in width (Nesbitt &  152 Muehlenbachs, 1995). The Au mineralization styles in this setting are described as mesothermal Au systems (Zhang et al., 1989). Fluid inclusions of Au bearing veins in the Omineca Belt are characterized by mostly H2O-CO2 fluids with minor CH4 at 300-380C (Nesbitt & Muehlenbachs, 1995). This is the most likely source of hypogene gold in the region. Source deposits may have been completely or partially eroded, or gold may be sourced from widespread but small and/or low grade uneconomic occurrences. Repeated recycling of gold in these placer systems, in addition to a significant contribution from supergene gold mineralization, has helped concentrate gold to economic concentrations. Supergene Signature of Mission Creek and Winfield There is a supergene gold contribution to Mission and Winfield placer gold grains. Major element analyses show fairly extensive Au-rich rims on grains indicating an extended time in a surficial environment. The volumetric importance of supergene gold accumulation can be roughly calculated, assuming a spherical grain, using an average grain radius of r = 0.2 mm and a supergene rim thickness of 10 microns (based on results in Section 3.2.5) yielding a supergene gold volume representing 14% of the volume of the entire grain. A supergene rim of equal thickness on a smaller grain would represent a higher percentage of total volume (inversely proportional to r3).  The significant variability in trace element signature in the MDS plot (Figure 28) resembles the scatter observed for the Prophet Mine sample set (Figure 55). This indicates significant re-mineralization and variable destruction of a discrete primary gold signature. The other placer gold sample sets retain a discrete signature indicative of unaltered hypogene gold. Overlap of core and rim regions in laser ablation pits may have led to a scattered trace element signature.  153 The mode or modes of formation of supergene gold are not well established. Hypotheses, outlined in Chapter 1, include electro-refining (Groen et al., 1984; McCready et al., 2003) and bacterial precipitation (Reith et al., 2010). The formation of supergene gold rims on these samples due to dissolution of Ag is unlikely, because as an incipient Au rich rim forms it would shield the dissolution of additional Ag deeper within the grain. Diffusion rates of Ag in gold grains are extraordinarily slow at low temperatures (Groen et al., 1984) and cannot account for the thickness of Au-rich supergene rims. This leaves the possibility that organic (bacterial) precipitation, inorganic precipitation, or electro-refining formed the supergene rim.  The microbially precipitated gold documented at the Prophet gold mine (Reith et al., 2010) is typified by the presence of biofilms containing nanoparticulate biologically precipitated gold that aggregates to the placer gold grain. A reddish tinge similar to the Prophet Mine biofilms was noted on the Winfield samples (Figure 10) and while none of the grains collected from Mission Creek for this study were observed to have these films, they have been reported in the area (Figure 60; Plovie, 2014; pers. comm.). It could be that gold grains mobilized in an active fluvial placer system (Mission Creek) have these films regularly disrupted and as such they are not commonly observed. It is likely that such films would be favored on gold grains that are in an immobile position for a long period of time or precipitation may occur after the placer system becomes dormant (as in Winfield). This is supported by a study at Rich Hill, Arizona where films are only formed on immobile gold grains (Kamenov et al., 2013). The samples for this study were not handled in a manner to best preserve the films (outlined by Reith et al., 2010) and so evidence for microbial precipitation cannot be assessed.   154  Figure 60. Gold grain collected from Mission Creek exhibiting partial reddish-orange to purple rim (top of photo). Photo taken under 60X magnification; provided by Mackenzie Plovie.        155 The U mineralization associated with the sediments hosting the Winfield paleoplacer gold is significant as the redox controlled U mineralization is epigenetic and related to circulation of post depositional fluids (Boyle, 1982). Fluid circulation for an extended period of time after the deposition of the placer gold would provide an opportunity for precipitation of supergene gold. The heavy mineral concentrate at the Winfield site did not contain magnetite grains, though the mineral is typically abundant in placer settings (Boyle, 1979). The absence indicates magnetite destruction by oxidation to hematite from significant volumes of oxidized fluids. Oxidized settings may promote Au mobilization as a soluble urano-organo gold complex. This mode of transportation has been proposed to explain formation of secondary gold in the world class Witwatersrand gold deposits in South Africa (Large et al., 2013). Although U was only semi-quantitatively measured the highest abundances were observed in the Mission Creek and Winfield samples (Table 9, Table 11).  Electro-refining is the process of chemical dissolution of the Au and Ag on the gold grain (possibly due to a redox gradient) and the rapid re-precipitation of only Au back onto the grain. This type of Au rich rim precipitation forms channel enrichment features creating a „coral like texture‟ (Groen et al., 1984) that was only observed faintly in one Mission Creek gold grain (Figure 9). There appears to be a somewhat lobate texture in the Au-rich rim in this case.     5.2.2 Lambly Creek Placer gold grains from Lambly Creek, on the west side of Okanagan Lake, displayed marked differences compared to other Kelowna area samples. Morphologically they were rougher and less flattened, tended to be smaller in size, and carried abundant primary mineral inclusions (Section 3.2.5). The irregular shape and presence of primary mineral inclusions for grains this small imply a proximal gold source.  156 There is little variation in major element chemistry of cores of grains when compared to the other Kelowna samples however, and Au-rich rims were poorly developed or absent on Lambly Creek gold (Section 4.2.1). This indicates very limited supergene influence and possibly less time under surficial conditions.  The Lambly Creek gold sample set carried two distinct trace element signatures that were unique from the Mission Creek/Winfield signature (Figure 28). The most important difference in element concentration between the two Lambly Creek signatures was in Cu values. Two analyses had hundreds of ppm Cu, an order of magnitude more than the other analyses. These high Cu analyses have a trace element signature resembling the Piaba Nugget from Brazil which is sourced from a greenstone hosted orogenic gold deposit (Figure 53). The remaining Lambly Creek analyses share an almost identical signature with the Nus River, Colombia gold which is sourced from the structurally controlled intrusive hosted stockwork Au-Ag Gramalote deposit (Figure 53). Lambly Creek gold may come from two separate gold occurrences within the catchment basin, one being greenstone-hosted orogenic gold, and the other intrusive-hosted resembling Gramalote.  There are several primary gold mineralization occurrences reported in the Lambly Creek catchment (See Section Geologic mapping shows a small greenstone belt within the catchment basin (Figure 5) that contains a vein hosted Au-Ag occurrence assaying up to 12 g/t Au. This represents a likely source for Cu-rich hypogene gold. Sampling of this hypogene occurrence would be required to test this.  The origin of Lambly samples with a similar signature to an intrusive-hosted stockwork Au-Ag deposit (Gramalote) is less certain. Intrusive-associated Au mineralization in the area could be related to emplacement of the Okanagan Batholith in this region. Possible sources include the stockwork Au-Ag-Pb-Zn veins of the Shear or Zumar occurrences, skarn related mineralization of  157 the Lamb occurrence, or the low sulphidation Spod occurrence. The Brenda Cu-Mo porphyry deposit (with Au-Ag as a byproduct; Roed et al., 1995) may be another possible source and is only 25 kilometers to the southwest of Lambly Creek. In addition the Elk deposit, located 55 kilometers to the west of Okanagan Lake, contains free gold in intrusive-hosted quartz veins and represents another possible source deposit-type (301,000 oz Au measured and indicated; Pooley et al., 2011).   5.3 Elemental Trends in Placer Gold Using all sample sets, representing a wide variety of locales, inferences can be made on the mode of occurrence, including general distribution and variation, of trace elements that can be expected in natural gold. In addition, signatures and trends are observed within individual occurrences. Using these initial results a database can begin to be constructed of what „normal‟ abundances are of trace elements in gold and which types of gold are enriched or depleted in certain elements.   5.3.1 Mode of Trace Element Occurrence  Multiple analyses of individual gold grains yielded elemental concentrations that were highly reproducible. The large Piaba nugget is the best example (Figure 38) due to the number of replicate analyses. Other grains large enough for multiple analyses (ex: AUS2, AUS5, MIS1) also have reproducible trace element signatures (Figure 55; Figure 28). Some abundant elements, such as Hg and Cu, can have superior precision values in natural gold than the AuRM2 standard (Table 5). This indicates that most trace elements are distributed evenly in natural gold either in solid solution or as evenly distributed micro-scale inclusions.   158 Significantly more elemental variation is observed between individual gold grains from a single placer occurrence. Elements that most successfully distinguish a placer occurrence are siderophile elements such as Fe and Pt along with chalcophile elements such as Se and Te. These element groups are the most reproducible between grains from the same locality, which suggests that deposit or regional-scale factors control their distribution in native gold. Chalcophile and siderophile group elements have been shown to be important for fingerprinting in other trace element studies of native gold as well (Velasquez, 2014). An exception is Sb which shows several orders of magnitude variation in concentration for grains from the same locality and between localities (Figure 40). An explanation for this is that the element is not present in solid solution and concentrations are controlled by the local abundance of Sb-rich inclusions in the gold.  Copper concentrations are inversely related to Ag concentrations (Figure 39), whereas every other trace element shows a positive correlation with Ag; quantitative trace element analyses of native gold from Timmins, Ontario reproduce this trend (Velasquez, 2014). Differences in Ag and Au atomic size create small interatomic gaps or lattice imperfections during crystal growth of native gold which may make room for the incorporation of trace element impurities. This would explain the higher trace element concentrations in Ag-rich gold. Copper does not follow the same trend as the other trace elements because Au-Ag-Cu forms essentially a continuous-composition alloy and, therefore, does not require lattice imperfections to be readily incorporated. Varying temperatures of crystallization favour either Ag or Cu incorporation into gold. At higher temperature mineralization, Cu concentrations are higher, whereas at low temperatures Ag concentrations are higher (Chapman & Mortensen, 2006). This explains the inverse correlation between Ag and Cu.   159 Many elements appear to be well correlated but none so well as Fe and Se (Figure 45) which, using the entire data set, gives a positive correlation with an R squared value of 0.91. This indicates chemical affinity between these elements in natural gold; the presence of one may help incorporate the other into the gold crystal lattice. Another possibility is that micro-inclusions of Fe and Se bearing minerals control the correlation; however iron selenide minerals such as ferroselite are not common, and this trend spans many different types of mineralizing systems. A third explanation is that Fe and Se concentrations of mineralizing fluids may control this correlation.  The transition metals Fe and V appear to be very well correlated with a subset of 5 samples from Winfield and Prophet exhibiting a separate (more V-rich), but still well correlated trend (Figure 44). It could be that the correlation shown by the bulk of the samples reflects solid solution in gold whereas the V-rich trend results from inclusions of a V-rich mineral. The V concentrations along this trend are the highest observed, whereas Fe values are on the lower end of the distribution. Magnetite is commonly V-rich, however the low Fe concentrations preclude it from explaining high V in these samples. Uranium concentrations were also elevated in the high V samples. These elements are both redox sensitive and some minerals, such as carnotite, bear U and V with some Fe substitution (Kerr et al, 2013). Carnotite is reported as one of the ore minerals in basal-type Uranium deposits at the Blizzard and Tyee deposits (Boyle, 1982) which are hosted in the same Miocene fluvial sediments that host the U-V rich Winfield placer gold (Table 9). Given the U to V stoichiometric ratio in carnotite of 2:1, which corresponds to a mass ratio of 9.3:1, much higher U values than observed for these samples would be expected if this mineral controlled V concentrations. However, these concentrations are semi-quantitative so the presence of carnotite is still a possibility.  Pd and Pt appear well correlated (Figure 47) except in a few analyses where Pd is shifted to higher values. An explanation is that although not many fluids in gold mineralizing systems are able  160 to carry large concentrations of Pt, some fluids such as saline oxidized fluids are able to carry high concentrations of Pd as chloride complexes (Chapman et al., 2009). Thus high Pd analyses may be explained by the presence of chloride complexes in mineralizing fluids. Some Winfield and Mission Creek samples have elevated Pd with respect to Pt (Figure 47) and this gold has been exposed to extended periods of oxidized fluid flow within Miocene fluvial sediments indicating that some gold may have been precipitated by chloride complexes along with Pd.     5.3.2 Signatures of Individual Localities Prophet Mine, Australia The Australian sample set from the Prophet Mine has documentation of biological precipitation of secondary gold (Reith et al., 2010) and is significant because it establishes that microbes can play an important role in placer gold precipitation. A distinct hypogene signature was detected in a subset of samples whereas an indistinct signature was observed for the supergene Au-rich component of the gold grains. Primary bedrock gold sources supplied gold grains, which then underwent supergene processes under surficial conditions, with precipitation of supergene gold coarsening the grains. Prophet Mine Hypogene Component A hypogene gold component was identified by the presence of primary mineral inclusions. Mineral inclusions were deemed primary (hypogene) in origin due to their composition (sulfides not stable in oxidized meteoric waters), anhedral jagged shape (detrital minerals would be abraded), and the presence of gold inclusions within the primary mineral inclusions (indicative of syngenetic precipitation).   161 Although a few elements, most notably Hg, distinguish the Prophet Mine samples (Table 11) most show significant variation leading to a spread-out field on MDS diagrams (Figure 55). Other sample sets had more coherent trace element signatures and therefore plotted in discrete fields on MDS plots. There is, however, a subset of analyses which plot apart on the MDS diagram and these analyses are of regions in gold grains that contained definitive hypogene mineral inclusions. Thus they likely provide a hypogene gold signature unaltered by supergene processes.  The hypogene gold is higher in certain base metal concentrations (Ni, Zn, Pb, and V) and primary inclusions of Ni-rich pyrite were also identified ( Figure 31). This indicates an association with base metal (sulfide) mineralization in the bedrock gold source. The signature of hypogene analyses of Prophet Mine samples does not appear to match any signatures from other sample sets (Figure 53). To geochemically identify the hypogene signature of the deposit type with higher confidence a larger, more comprehensive, database of native gold analyses is required. Prophet Mine Supergene Component All of the morphological and major element observations in this study support the conclusion that the Prophet Mine placer gold has been influenced by supergene gold precipitation. The thicker Au-rich rims on grain embayments as opposed to on protrusions or spread evenly across the surface of the grain, indicates that the supergene gold is the result of precipitation of high purity gold and not dissolution of impurities. Furthermore, the embayments contain films with colloidal gold of varying coarseness that appears to be aggregated to the grain surface (Figure 32). The embayments provide a protected setting in a placer system for microbes to live.   162 The low Hg concentration in supergene gold (Table 12) seems to preclude the possibility of Ag leaching forming Au-rich rims. Mercury, with its strong affinity to Au, would be expected to remain behind in the case of Ag leaching. This also indicates that the Hg present in the cores of grains is from the hypogene source and is not related to mining activity. Observations here show that, not only is supergene gold precipitation occurring, it is a significant component and contributor to placer gold at Prophet. Zones of Au enrichment on placer grains are over 0.1 mm in thickness (Figure 33) and can constitute a significant volume of a gold grain (up to a maximum of ~50%; calculated using dimensions from Appendix A. The thick rims provided the best opportunity to analyze supergene gold for the project. Determination of a supergene gold trace element signature is complicated by the fact that analyses could not be taken wholly within supergene zones resulting in analyses representing a mixture of hypogene and supergene gold. If the supergene gold carries a unique signature, a mixing line between pure supergene and pure hypogene should be determinable. The lack of this and the apparent random composition of the supergene analyses indicate the absence of a unique signature. This may reflect the highly variable conditions for the formation of supergene gold within a spatially large catchment. In addition, mineralizing temperatures of the supergene gold are extremely low which limits diffusion of trace elements through the gold. Chemically variable detrital micro-inclusions are also incorporated into supergene gold resulting in an indistinct trace element signature.  Although the signature of the supergene gold is indistinct, there are consistently higher Sb and Bi concentrations in the inferred supergene gold (Table 12). This may represent some biologic affinity for these elements, or it may represent a detrital inclusion component. Interestingly, the inferred supergene gold has elevated U which was also noted for the Mission Creek and Winfield gold. Thus, there is a possible role for U (as urano-organo-gold complexes) in secondary gold  163 mineralizing processes, which has also been suggested in the Witwatersrand paleoplacer deposit in South Africa (Large et al., 2013). Phosphorous, with its biological role and siderophile tendency, could possibly distinguish biologically precipitated gold (Southam & Beveridge, 1996), however it was not determinable with the analytical methods used in this study. Nus River, Colombia Background information about the Nus River placer gold samples is limited to technical reports on the Gramalote deposit which outcrops upstream in the catchment basin and is the suspected hypogene gold source (Gustavson Associates, 2012). Placer gold grains are moderately rounded but lack a flake like morphology indicating a proximal source (Grant et al., 1991). Supergene rims are thin or absent supporting a limited time in a surficial setting. In addition, several gold grains contain sub-angular primary mineral inclusions indicating a predominantly hypogene source. The Nus River placer gold carries a very distinct trace element signature consistent with a single nearby hypogene source. It appears that an economic gold deposit eroding into a placer system will result in placer gold that preserves the trace element signature of the deposit; analyses of gold directly from the Gramalote deposit could confirm this. Characteristics of the signature are low concentration of lithophile elements (Mg, Al, and Ti) and the base metal chalcophile elements: Pb, Bi, Cu, and Zn. The work of McInnes et al (2008) provides support for the idea that placer gold composition can preserve the trace element signature of the bedrock deposit source.  The Gramalote deposit is an orogenic gold deposit, as is the Piaba deposit, however Gramalote is intrusive-hosted whereas Piaba is hosted in a metavolcanosedimentary sequence (SRK Consulting, 2012) typical of a greenstone belt. Gramalote has low Cu and Piaba has high Cu, thus Cu concentrations in gold may be controlled by host lithology in this case. This could be applied to  164 exploration for orogenic gold deposits; high Cu (200-300 ppm) in gold grains from stream sediment samples would direct further exploration to greenstone belts whereas low Cu (10-50 ppm) would direct efforts toward intermediate to felsic intrusive bodies. Piaba, Brazil The initial theory put forth about the origin of the Piaba nugget when the sample was provided was that it may have formed by in situ supergene gold precipitation under surficial conditions due to a redox gradient within sediments, perhaps biologically mediated (Fedorowich, 2012; pers. comm.). Initial morphological observations including flattened shape, folded edges, and detrital inclusions in the gold nugget indicate significant fluvial transport. If indeed this grain was formed out of hydraulic equilibrium with the surrounding sediment (i.e. not hydraulically sorted by physical properties), it underwent subsequent placer transport. No Au-rich rim or internal elemental variation was observed (Figure 36), and though it is composed of relatively high purity gold (92 wt%) it is within the range of hypogene gold composition (Boyle, 1979). Multiple Piaba nugget analyses were extraordinarily similar supporting a hypogene origin. Supergene gold, such as some Prophet Mine samples, shows great chemical variability even within single grains. Piaba also has very high Cu which is indicative of high temperature hypogene mineralization (Chapman & Mortensen, 2006). Thus, the initial theory of the nugget forming from supergene gold precipitation is not supported by trace element analysis. The most likely candidate for a bedrock source in the region is the Piaba orogenic gold deposit as outlined in technical reports (SRK Consulting, 2012).      165 Northern B.C. The placer gold grains from northern B.C. that are part of the UBC Okanagan gold sample set have little associated background information. The morphology of the grains including roundness, folding, and detrital infill in embayments confirm this is placer gold and that it has likely travelled a significant distance. The lack of any Au-rich rim, and the Ag-rich composition up to 50 wt% (electrum) precludes supergene processes for an origin.  Grains carry a distinct hypogene signature that does not resemble data from any of the other sample sets for this project (Figure 54). The electrum grain has variation in Sb ranging from 0.1 to over 1,000 ppm. A stair-step pattern is observed in Sb ranked probability plots (Figure 40) which matches Bi distribution in Northern B.C. samples at a lower concentration range. This leads to the conclusion that micro-scale mineral inclusions control the concentration of these elements. The range of concentrations could be explained by the relative volumetric abundance of these inclusions in each gold grain. The ratio of Sb to Bi is consistent at approximately 3400:1. This indicates that a mineral containing Sb as a major component, and Bi as a minor component, is controlling this signature. Assuming pure Sb the Bi proportion would equal 0.03 wt%. The most likely mineral is stibnite which contains enough Sb to be responsible for an extremely high concentration in the gold without visible inclusions. Bismuth is easily incorporated into the stibnite crystal structure  in solid solution (Mehrabi, 2008), which accounts for the strong Sb-Bi correlation.        166 Chapter  6: Conclusions 6.1 Reference Materials Conclusions 1. Elements in AuRM2 homogenously distributed at the LA-ICP-MS sampling scale are: Mg, Al, Ti, Fe, Ni, Cu, Zn, As, Se, Sn, Sb, and Te. Thus AuRM2 is a suitable external standard for micro-analysis of these elements.  2. Elements in AuRM2 that exhibit minor heterogeneity are: Mn, Rh, Pd, Pt, Pb, and Bi. Heterogeneity is far exceeded by the natural variation in native gold (Figure 38) meaning that AuRM2 is a suitable external standard for micro-analysis of these elements.  3. Elements with undeterminable homogeneity in AuRM2 due to analytical issues are: Si, Ca, and Cr. Improved analytical methods may allow their measurement in the future.   4. Elemental MDS plots of AuRM2 show two distinct groupings of the chalcophile elements and of the PGEs indicating that heterogeneity may be the result of micro-inclusions of PGE phases and chalcophile-rich phases.  5. Using AuRM2 as an external standard to determine trace element concentrations in silicate-glass NIST610 gives semi-quantitative values within ±1% to ±50% of true values even though they are not matrix matched. This reveals a strength of LA-ICP-MS analysis and means that NIST610 could be used as an external standard with other non-matched matrices (sulfides, carbonates, oxides etc.) to yield reasonable concentration estimates.  6. Semi-quantitative concentrations of elements in AuRM2 not added during manufacture were determined for: V (0.3 ppm; homogenous), Hg (3.7 ppm; heterogeneous), and U (0.1 ppm; heterogeneous). These values may be used to calculate semi-quantitative values in samples.  167 7. Mercury in NIST610, for which there are no accepted values, was semi-quantitatively determined to be 0.6 ppm. Accuracy could be confirmed by solution ICP-MS for values to be considered quantitative.  8. Quantitative concentrations of elements in FAU7 not added during manufacture are: Se (24.9 ppm), Rh (0.1 ppm), Sb (0.3 ppm), and Te (2 ppm). Future trace element studies using FAU7 as an external standard can quantitatively determine concentrations of these elements in gold samples.  6.2 Kelowna Area Gold Conclusions 1. Placer gold in Mission Creek is sourced from Miocene age paleoplacer deposits such as at the Winfield Mine. This indicates there is no direct hypogene gold source for Mission Creek and any mineral exploration should focus on Miocene age fluvial deposits. Some of these occurrences have been historically targeted for uranium but the gold potential has been perhaps under-explored.   2. Extended time (Miocene-present) under surficial conditions with post-depositional circulation of oxidized fluids (also related to U-mineralization) led to supergene gold precipitation on Kelowna area paleoplacer gold. Gold may have been carried as urano-organo-gold complexes, chloride complexes, or precipitated microbially as nanoparticles. 3. There are two distinct trace element chemical signatures for gold grains collected from Lambly Creek indicating ≥2 hypogene sources for placer gold in the catchment. One group has elevated Cu and could carry an orogenic gold signature sourced from the greenstone-hosted Bond occurrence and the low Cu group could have an intrusive-hosted gold signature.   168 6.3 Elemental Variation in Combined Placer Gold Sample Sets  1. Characterizing sample populations of placer gold using LA-ICP-MS analyses requires a much smaller number of samples than other methods such as: morphological characterization, mineral inclusion assemblage classification, and major element analysis.  2. Trace elements, especially siderophile and chalcophile elements, are predominantly present as solid solution in native gold as opposed to as mineral micro-inclusions.  3. Where micro-mineral inclusions are present, they are indicated by strongly correlated elemental trends, such as a consistent Sb-Bi ratio indicating stibnite inclusions in the northern B.C. placer gold. Identification of micro-scale mineral inclusions in placer or glacial till gold can vector exploration for the bedrock source to rock types where such minerals may be expected. Accessory minerals are likely to be more abundant and/or widespread than the gold itself in bedrock mineralization making them an easier exploration target to recognize.   4. Placer gold from a single hypogene mineralization source carries a distinct signature, such as the signature for Nus River and Piaba placer gold, whereas supergene gold, as in the Prophet Mine, does not carry a distinct signature. This is likely due to the purity of the supergene gold where trace elements in solid solution are less abundant and any micro-inclusions are of detrital origin. Trace element analysis of placer gold for exploration purposes with a distinct signature indicates a hypogene bedrock source target.  5. Supergene gold precipitation is more pronounced in settings where gold transport has ceased and the host sediments remain in stable conditions. This allows growth of Au-rich rims on gold grains as a result of inorganic processes or by microbial precipitation in biofilms. Future research on supergene gold should focus on dormant placers or paleoplacers.  169 6. This project confirms the presence of supergene gold precipitation at the Prophet Mine. Furthermore it indicates that supergene precipitation accounts for a significant proportion of the placer gold. The supergene gold tends to have high Sb and Bi. 7. The hypogene component of Prophet Mine gold is characterized by high concentrations of base metals including Ni, Zn, Pb, and V which, along with the presence of Ni-rich pyrite inclusions, indicates a bedrock source with associated base metal sulfide mineralization.  8. Orogenic gold hosted in metasedimentary or metavolcanic rocks appears to have high Cu (~200-300 ppm) whereas orogenic gold hosted in intermediate to felsic intrusive rocks has low Cu (~10-50 ppm).                 170 Chapter  7: Future Work 7.1 Reference Materials LA-ICP-MS Analysis Using AuRM2 Reference Material  There are several simple changes to LA-ICP-MS analytical methods that can improve the quality of future analyses using AuRM2 as an external standard for determination of trace element concentrations in native gold. Due to the decreased rate of laser penetration in a gold matrix (0.06 µm depth per laser shot), a longer sampling time may be used without exceeding a depth to pit diameter ratio of 0.5, thus avoiding elemental fractionation effects (Potts et al., 1995). Also, due to the reduced penetration rate, a slightly higher laser frequency, than the 10 Hz for this project, may be used in the future, which would boost signal. However, care should be taken not to increase frequency to the point where the signal becomes erratic and coarse sized ejecta are created (appears to occur by around ~100 Hz).   Improved signal from the suggestions above could yield usable concentration data in future research for Si, Ca, and Cr (not determinable in this study). In addition, 52Cr should be the isotope used for detection instead of 53Cr due to its higher natural abundance (84%) and the fact that no major interferences were observed at that mass range.           Laser raster lines may be used for subsequent analyses, which would provide a more representative sample by averaging a larger surface area. No evidence for surficial contamination was observed on AuRM2 or the native gold samples. Spot samples, as opposed to raster lines, were used in this study because spot samples quantify homogeneity at a finer spatial resolution.   Improvements to LA-ICP-MS parameters would permit smaller sample sizes to be taken yielding better spatial resolution. This will allow future projects to sample finer details such as supergene Au-rich rims, internal trace element variation, and micro-mineral inclusions. Future work  171 targeting supergene rims may drill laser depth profiles into the surface of unpolished grains to ensure that there is no mixing of supergene and hypogene zones.  Software is available for two-dimensional trace element distribution maps of minerals such as pyrite using LA-ICP-MS (Large & Danyushevsky, 2009) and could be used in future projects analyzing native gold. These imaging techniques are similar to SEM-EDS maps (ex: Figure 21; Figure 24; Figure 34) but would show trace element distributions. A rudimentary attempt was made in this project to create a LA-ICP-MS trace element map of native gold but was hampered due to poor resolution, difficulty in accurately assigning x-y coordinates to individual signal counts, and the labour intensity of pre-processing the data.      LA-ICP-MS Analysis Using FAU7 Reference Material  FAU7 had significant surficial Al contamination observable by the trend of 27Al signal as the laser penetrated into the standard. Visual examination indicated that the sample had been cut with a saw or was roughly polished (probable Al contaminator); this can be remedied by polishing the standard using a fine diamond paste. New quantitative concentrations for trace elements in FAU7, not added, but present at detectable levels (see Section 6.1) can be assessed for homogeneity at a bulk scale using conventional solution ICP-MS methods. Future studies using FAU7 as an external standard can quantitatively measure these elements in native gold samples.    Semi-quantitative Analysis Using NIST610 Reference Material  Because matrix dependent effects were found to be minimal (see Section, semi-quantitative trace element concentrations in native gold samples can be reliably determined using NIST610 as an external standard. A future project should be done to determine which elements with  172 known concentrations in NIST610 (71 elements; Jochum et al, 2011) are detectable in various native gold specimens and at what concentration ranges. This could guide the manufacture of new gold reference materials specifically suited for analyzing natural gold samples. This project underscores the value that Hg concentrations have for classifying native gold populations and any new standards created should have this element added (with controls on homogeneity).  7.2 Kelowna Area Project The work in this project is sufficient to characterize placer gold populations in the central Okanagan and no further work of this nature is recommended at this point. If later development of a database characterizing bedrock gold signatures is done, this data may be compared with signatures from various deposit types to confirm the interpretation of hypogene signatures in this project. Mineral exploration for gold in the Kelowna area should target Miocene paleoplacers on the east side of Okanagan Lake (Omineca Belt) and greenstone belts or intrusive units on the west side of Okanagan Lake (Intermontane Belt).   7.3 Characterization of Native Gold Trace Element Signatures Develop Database of Native Gold Bedrock Deposit-type Signatures  A database should be compiled of bedrock native gold trace element signatures. The database should be of bedrock gold because it can be compared to placer and glacial till gold samples in order to determine source deposit-types which can guide mineral exploration. Such a database should contain signatures of common deposit-types worldwide. Numerous examples of each deposit-type should be analyzed to confirm the repeatability of a signature.    173 Determine Effects on Trace Element Signature from Factors other than Deposit-type Detailed information on the geology of deposits (age, terrane, host lithology, size, grade, fluid composition, temperature of formation, depth of formation, etc.) used for developing a gold trace element database can be used to find patterns that may identify signatures that certain geologic characteristics have in common. An example would be confirming the conclusion from this study that gold from greenstone hosted orogenic deposits is richer in Cu than gold from intrusive hosted orogenic deposits (see Section 6.3). This is significant because these types of characteristics can better guide mineral exploration. More work is required to characterize trace element signatures of Au-rich supergene rims on placer gold grains. Improved analytical techniques will aid this as will a wider selection of samples. The most likely locations to find supergene gold samples are in dormant placers and paleoplacers. This could determine whether processes such as microbial gold precipitation are widespread phenomena or isolated occurrences.            174 References Armstrong, R. (1988). Mesozoic and early Cenozoic magmatic evolution of the Canadian Cordillera. Geological Society of America Special Paper, 55–91. Ashley, P., Billington, W., Graham, R., & Neale, R. (1978). Geology of the Coalstoun porphyry copper prospect, southeast Queensland, Australia. Economic Geology, 73, 945–965.   Ashley, P. M. (1980). Geology of the Ban Ban Zinc Deposit, A Sulfide-bearing Skarn, Southeast Queensland, Australia. Economic Geology, 75, 15–29. BC Ministry of Energy and Mines. (1993a). MINFILE Detail Report Aitkens/Stables (Winfield) Occurence 082LSW142. BC Ministry of Energy and Mines. (1993b). MINFILE Detail Report AT (Dun/Esperon) Occurence 082LSW051. BC Ministry of Energy and Mines. (1993c). MINFILE Detail Report Bond Occurence 082LSW117. BC Ministry of Energy and Mines. (1993d). MINFILE Detail Report Harris Creek 082LSW091. BC Ministry of Energy and Mines. (1993e). MINFILE Detail Report Nova Occurence 082LSW002. BC Ministry of Energy and Mines. (1993f). MINFILE Detail Report Ribbleworth (Winfield) Occurence 082LSW019. BC Ministry of Energy and Mines. (1993g). MINFILE Detail Report Stuart (Winfield) Occurence 082LSW072. BC Ministry of Energy and Mines. (1993h). MINFILE Detail Report Tick Occurence 082LSW003. BC Ministry of Energy and Mines. (1993i). MINFILE Detail Report Winfield (Eley/Hall) Occurence 082LSW093. BC Ministry of Energy and Mines. (1993j). MINFILE Detail Report Zumar Occurence 082LSW111. BC Ministry of Energy and Mines. (1996a). MINFILE Detail Report Jubilation Occurence 082ENW097. BC Ministry of Energy and Mines. (1996b). MINFILE Detail Report Mission Creek Occurence 082ENW105. BC Ministry of Energy and Mines. (1996c). MINFILE Detail Report Shear Occurence 082ENW094. BC Ministry of Energy and Mines. (1996d). MINFILE Detail Report Spod Occurence 082ENW091.  175 BC Ministry of Energy and Mines. (2008). MINFILE Detail Report Hydraulic Lake Occurence 082ENW053. BC Ministry of Energy and Mines. (2012a). MINFILE Detail Report Bear Occurence 082ENW116. BC Ministry of Energy and Mines. (2012b). MINFILE Detail Report Lamb Occurence 082ENW064. Bird, D. K., Brooks, C. K., Gannicott, R. A., & Turner, P. A. (1991). A Gold-Bearing Horizon in the Skaergaard Intrusion, East Greenland. Economic Geology, 86, 1083–1092. Boyle, D. (1982). The formation of basal-type uranium deposits in South Central British Columbia. Economic Geology, 77, 1176–1209. Boyle, R. (1979). The Geochemistry of Gold and its Deposits. Ottawa: Energy, Mines and Resources Canada. British Columbia Department of Mines. (1961). Notes on Placer-Mining in British Columbia. Brostoff, L. B., González, J. J., Jett, P., & Russo, R. E. (2009). Trace element fingerprinting of ancient Chinese gold with femtosecond laser ablation-inductively coupled mass spectrometry. Journal of Archaeological Science, 36(2), 461–466. Butt, C. R. M., & Hough, R. M. (2009). Why Gold is Valuable. Elements, 5(5), 277–280.  Chapman, R. (2010a). Microchemical studies of placer and lode gold in the Klondike district, Yukon, Canada: 1. Evidence for a small, gold-rich, orogenic hydrothermal system in the Bonanza. Economic Geology, 1369–1392.  Chapman, R. (2010b). Microchemical studies of placer and lode gold in the Klondike District, Yukon, Canada: 2. constraints on the nature and location of regional lode sources. Economic Geology, 1393–1410.  Chapman, R., & Leake, R. (2000). The application of microchemical analysis of alluvial gold grains to the understanding of complex local and regional gold mineralization. Economic Geology, 95, 1753–1773. Chapman, R., Leake, R., & Bond, D. (2009). Signatures of gold formed in oxidizing chloride hydrothermal systems and their significance within populations of placer gold grains collected during reconnaissance. Economic Geology, 563–585. Chapman, R., & Mortensen. J. (2006). Application of microchemical characterization of placer gold grains to exploration for epithermal mineralization in regions of poor exposure. Journal of Geochemical Exploration, 91: 1-26.   176 Choi, S., Lee, G., & Han, J. (2008). Determination of trace elements Ge and P in a gold bonding wire by inductively coupled plasma atomic emission spectrometry. Bulletin-Korean Chem. Soc., 29(2), 393–397. Ciobanu, C. L., Cook, N. J., Utsunomiya, S., Kogagwa, M., Green, L., Gilbert, S., & Wade, B. (2012). Gold-telluride nanoparticles revealed in arsenic-free pyrite. American Mineralogist, 97(8-9), 1515–1518. Constantinescu, B., & Bugoi, R. (2008). Micro-SR-XRF and micro-PIXE studies for archaeological gold identification–The case of Carpathian (Transylvanian) gold and of Dacian bracelets. … in Physics Research …, 1–12. Crawford, E. C. (2007). Klondike Placer Gold: New tools for examining morphology, composition, and crystallinity. University of British Columbia. Dehnavi, A., (2014) Graduate student at the University of New Brunswick working on LA-ICP-MS analysis of volatile trace elements in sulfides. Desborough, G. (1970). Silver depletion indicated by microanalysis of gold from placer occurrences, western United States. Economic Geology, 65, 304–311. Fedorowich, J., (2012). Provided the Piaba nugget sample for this project. Figge, J., & Townsend, C. (2002). The Omineca Episode. Burke Museum of Natural History and Culture. Retrieved from http://www.burkemuseum.org/static/geo_history_wa/The Omineca Episode.htm March 2013. Fryer, B. J., Jackson, S. E., & Longerich, H. P. (1995). The Design , Operation and Role of the Laser-Ablation Microprobe Coupled with an Inductively Coupled Plasma - Mass Spectrometer (LAM-ICP-MS) in the Earth Sciences. The Canadian Mineralogist, 33, 303–312. Gabrielse, H., & Yorath, C. (1999). Geology of the Cordilleran Orogen in Canada. Geological Survey of Canada, 677–705. Garnett, R. H., & Basset, N. C. (2005). Placer Deposits. Economic Geology 100th Anniversary Volume, 813–843. Grant, a. H., Lavin, O. P., & Nichol, I. (1991). The morphology and chemistry of transported gold grains as an exploration tool. Journal of Geochemical Exploration, 40(1-3), 73–94.  Groen, J. C., Craig, J. R., & Rimstidt, J. D. (1984). Gold -Rich Rim Formation on Electrum Grains in Placers. Canadian Mineralogist, 28. Guerra, M. F., Radtke, M., Reiche, I., Riesemeier, H., & Strub, E. (2008). Analysis of trace elements in gold alloys by SR-XRF at high energy at the BAMline. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms, 266(10), 2334–2338.   177 Gustavson Associates. (2012). NI 43-101 Technical Report on Resources Gramalote Project Providencia, Colombia. Henry, R. (2013). Greater Denver Area  ICP-MS Specialist at Thermo Fisher Scientific. Hinds, M. (2009). LBMA RM Project Overview. In LBMA Assaying and Refining Seminar (pp. 24–25). Hinds, M. (2013). AuRM3 Reference Material Update. In Royal Canadian Mint (pp. 1–5). Hough, R. M., Butt, C. R. M., & Fischer-Buhner, J. (2009). The Crystallography, Metallography and Composition of Gold. Elements, 5(5), 297–302.  Hough, R. M., Noble, R. R. P., Hitchen, G. J., Hart, R., Reddy, S. M., Saunders, M., … Verrall, M. (2008). Naturally occurring gold nanoparticles and nanoplates. Geology, 36(7), 571.  Hough, R. M., Noble, R. R. P., & Reich, M. (2011). Natural gold nanoparticles. Ore Geology Reviews, 42(1), 55–61. Jia, Y., Kerrich, R., & Goldfarb, R. (2003). Metamorphic origin of ore-forming fluids for orogenic gold-bearing quartz vein systems in the North American Cordillera: Constraints from a reconnaissance study. Economic Geology, 98, 109–123. Jochum, K. P., Weis, U., Stoll, B., Kuzmin, D., Yang, Q., Raczek, I., … Enzweiler, J. (2011). Determination of Reference Values for NIST SRM 610-617 Glasses Following ISO Guidelines. Geostandards and Geoanalytical Research, 35(4), 397–429. Kamenov, G., Melchiorre, E., Ricker, F., & DeWitt, E. (2013). Insights from Pb Isotopes for Native Gold Formation During Hypogene and Supergene Processes at Rich Hill, Arizona. Economic Geology, 108, 1577–1589. Kerr, A., Walsh, J., Sparkes, G., & Hinchey, J. (2013). Vanadium potential in Newfoundland and Labrador: A review and assesment. Newfoundland and Labrador Department of Natural Resources Geological Survey, Report 13-1, 137-165.  Knight, J. B. (1985). A Microprobe Study of Placer Gold and its Origin in the Lower Fraser River Drainage Basin, B.C. University of British Columbia. Langmuir, C. (2008). Forensic Study: Gold Fingerprinting by LA-ICP-MS. Harvard University. Retrieved from http://www.people.fas.harvard.edu/~langmuir/LA-Gold-Fingerprinting.html September 2012. Large, R., & Danyushevsky, L. (2009). Gold and trace element zonation in pyrite using a laser imaging technique: implications for the timing of gold in orogenic and Carlin-style sediment-hosted deposits. Economic Geology, 635–668.  178 Large, R. R., Meffre, S., Burnett, R., Guy, B., Bull, S., Gilbert, S., … Danyushevsky, L. (2013). Evidence for an Intrabasinal Source and Multiple Concentration Processes in the Formation of the Carbon Leader Reef, Witwatersrand Supergroup, South Africa. Economic Geology, 108(6), 1215–1241. Little, T. (1995). 40Ar/39Ar thermochronology of epidote blueschists from the North D‟Aguilar block, Queensland, Australia: Timing and kinematics of subduction complex unroofing. Geological Society of America, (5), 520–535. Madsen, J. K., Thorkelson, D. J., Friedman, R. M., & Marshall, D. D. (2006). Cenozoic to Recent plate configurations in the Pacific Basin: Ridge subduction and slab window magmatism in western North America. Geosphere, 2(1), 11. Mann, S. (1992). Bacteria and the Midas touch. Nature.  McCandless, T., Baker, M., & Ruiz, J. (1997). Trace Element Analysis of Natural Gold by Laser Ablation ICP‐MS: A Combined External/Internal Standardisation Approach. Geostandards Newsletter, 21, 1–8. McCourt, W. J., Aspden, J. A., & Brook, M. (1984). New geological and geochronological data from the Colombian Andes : continental growth by multiple accretion. Geological Society of London, 141, 831–845. McCready, A., Parnell, J., & Castro, L. (2003). Crystalline placer gold from the Rio Neuquen, Argentina: Implications for the gold budget in placer gold formation. Economic Geology, 98, 623–633. McInnes, M., Greenough, J. D., Fryer, B. J., & Wells, R. (2008). Trace elements in native gold by solution ICP-MS and their use in mineral exploration: A British Columbia example. Applied Geochemistry, 23(5), 1076–1085. Mehrabi, B. (2008). Solid Solution in the As2S3 -Sb2S3 Series at Zarshuran Gold Deposit, Iran. Journal of Sciences, Islamic Republic of Iran, 19(2), 137–143. Mitchell, V., Lewis, R., & Larsen, J. Geology of Northern Idaho and the Silver Valley. Idaho State University. http://geology.isu.edu/Digital_Geology_Idaho/Module7/mod7.htm. Retrieved March 2013. Moles, N. R., Chapman, R. J., & Warner, R. B. (2013). The significance of copper concentrations in natural gold alloy for reconnaissance exploration and understanding gold-depositing hydrothermal systems. Geochemistry: Exploration, Environment, Analysis, 13(2), 115–130.  Monger, J., & Price, R. (1979). Geodynamic evolution of the Canadian Cordillera - progress and problems. Canadian Journal of Earth Sciences, 16, 770–791.  179 Mortensen, J., & Chapman, R. (2010). Characterization of Placer-and Lode-Gold Grains as an Exploration Tool in East-Central British Columbia (NTS 093A, B, G, H). Geosciencebc.com, 65–76.  Müller, A., Wiedenbeck, M., Flem, B., & Schiellerup, H. (2008). Refinement of Phosphorus Determination in Quartz by LA-ICP-MS through Defining New Reference Material Values. Geostandards and Geoanalytical Research, 32(3), 361–376. Murray, S. (2009). LBMA Certified Reference Materials: Gold Project Final Update. Alchemest, (55), 11–12. Nesbitt, B. E., & Muehlenbachs, K. (1995). Geochemical studies of the origins and effects of synorogenic crustal fluids in the southern Omineca Belt of British Columbia, Canada. Geological Society of America Bulletin, 107(9), 1033–1050. Netshitungulwana, R., Tredoux, M., & Jacobson, L. (2010). An Investigation of the Trace Element Compositions of Gold from the Zimbabwe and South Africa : Implications for Tracing the Source of Archeological Gold Artifacts. In 7th Inkaba yeAfrica Workshop (p. 2010). Outridge, P., Doherty, W., & Gregoire, D. (1998). Determination of trace elemental signatures in placer gold by laser ablation–inductively coupled plasma–mass spectrometry as a potential aid for gold exploration. Journal of Geochemical Exploration. Penny, G. (2001). Fingerprinting gold using laser ablation microprobe - Inductively coupled plasma - mass spectrometry (LAM ICP-MS): an exploration tool. Memorial University of Newfoundland. Pooley, R., Lomas, S., Hawthorn, G., & Alexander, R. (2011). NI 43-101 Technical Report for a Preliminary Economic Assessment on the Elk Gold Project, Merritt, British Columbia, Canada Pop, D., Ionescu, C., Forray, F., Tămaş, C. G., & Benea, M. (2011). “Transylvanian gold” of hydrothermal origin: an EMPA study in an archaeological provenancing perspective. European Journal of Mineralogy, 23(6), 911–923. Potts, P., Bowles, F., Reed, S., & Cave, M. (1995). Microprobe Techniques in the Earth Sciences (First Edit.). London, UK: Chapman & Hall. Province of British Columbia Ministry of Energy Mines and Petroleum Resources. (1986). Occurrence and Distribution of Platinum-Group Elements in British Columbia. Reith, F., Brugger, J., Zammit, C., Nies, D., & Southam, G. (2013). Geobiological Cycling of Gold: From Fundamental Process Understanding to Exploration Solutions. Minerals, 3(4), 367–394.  Reith, F., Etschmann, B., Grosse, C., Moors, H., Benotmane, M. a, Monsieurs, P., … Brugger, J. (2009). Mechanisms of gold biomineralization in the bacterium Cupriavidus metallidurans.  180 Proceedings of the National Academy of Sciences of the United States of America, 106(42), 17757–62. Reith, F., Fairbrother, L., Nolze, G., Wilhelmi, O., Clode, P. L., Gregg, a., … Brugger, J. (2010). Nanoparticle factories: Biofilms hold the key to gold dispersion and nugget formation. Geology, 38(9), 843–846. Reith, F., Lengke, M. F., Falconer, D., Craw, D., & Southam, G. (2007). The geomicrobiology of gold. The ISME Journal, 1(7), 567–84. Reith, F., Stewart, L., & Wakelin, S. a. (2012). Supergene gold transformation: Secondary and nano-particulate gold from southern New Zealand. Chemical Geology, 320-321, 32–45.  Roed, M., Barendregt, R., Benowitz, J., Smith, C., Sanborn, P., Greenough, J., Huscroft, C., Layer, P., Mathewes, R., & Tessler, D. (2013). Evidence for an Early Pleistocene glaciation in the Okanagan Valley, southern British Columbia. Canadian Journal of Earth Sciences, 51(2), 125-141. Roed, M., Dobson, E., Greenough, Hughes, & Luttmerding. (1995). Geology of the Kelowna Area and Origin of the Okanagan Valley British Columbia. Kelowna: Kelowna Geology Committee. Southam, G. (1998). Quantification of sulfur and phosphorus within secondary gold rims on Yukon placer gold. Geology, 26(4), 339. Southam, G., & Beveridge, T. (1996). The occurrence of sulfur and phosphorus within bacterially derived crystalline and pseudocrystalline octahedral gold formed in vitro. Geochimica et Cosmochimica Acta, 60(22), 4369–4376. Southam, G., Lengke, M. F., Fairbrother, L., & Reith, F. (2009). The Biogeochemistry of Gold. Elements, 5(5), 303–307. SRK Consulting. (2012). NI 43-101 Technical Report Mineração Aurizona S. A. Aurizona Project. Standish, C., Dhuime, B., Chapman, R., Coath, C., Hawkesworth, C., & Pike, A. (2013). Solution and laser ablation MC-ICP-MS lead isotope analysis of gold. Journal of Analytical Atomic Spectrometry, 28(2), 217. Tomkins, A., Pattison, D., & Zaleski, E. (2004). The Hemlo Gold Deposit, Ontario : An Example of Melting and Mobilization of a Precious Metal-Sulfosalt Assemblage during Amphibolite Facies Metamorphism and Deformation. Economic Geology, 99, 1063–1084. Townley, B. K., Hérail, G., Maksaev, V., Palacios, C., Parseval, P. De, Sepulveda, F., … Ulloa, C. (2003). Gold grain morphology and composition as an exploration tool : application to gold exploration in covered areas. Geochemistry: Exploration, Environment, Analysis, 3, 29–38.  181 Velasquez, A. (2014). Trace element analysis of native gold by laser ablation ICP-MS: A case study in greenstone-hosted quartz-carbonate vein ore deposits, Timmins, Ontario. University of British Columbia Okanagan.  Walder, A. J., Abell, I. D., Platzner, I., & Freedman, P. a. (1993). Lead isotope ratio measurement of NIST 610 glass by laser ablation inductively coupled plasma mass spectrometry. Spectrochimica Acta Part B: Atomic Spectroscopy, 48(3), 397–402. Williams-Jones, a. E., Bowell, R. J., & Migdisov, a. a. (2009). Gold in Solution. Elements, 5(5), 281–287. Wrighton, T. M. (2013). Placer Gold Microchemical Characterization And Shape Analysis Applied As An Exploration Tool In Western Yukon. University of British Columbia. Zhang, X., Nesbitt, B., & Muehlenbachs, K. (1989). Gold mineralization in the Okanagan Valley, southern British Columbia: fluid inclusion and stable isotope studies. Economic Geology, 84(1961), 410–424.          182 Appendices  Appendices of analytical data, measurements, and observations used for this project are presented in the following pages. While not exhaustive, they include the key information for the project. The appendices are divided into: Sample List/Petrographic Descriptions, Sample Elemental Concentration Data, and finally Standard Reference Material Elemental Concentration Data.   Appendix A  Sample List and Petrographic Descriptions  This appendix comprises of a comprehensive list of all samples compiled for this project. Many samples were lost at the polishing stage and therefore no further information is presented on them. Samples that did survive polishing have listed empirical data such as grain dimensions (in millimeters), metal alloy colour percentages, and inclusions present. In addition, notes are recorded for features such as roundness, smoothness, mineral inclusion assemblage composition, and any other notable features.     183 Sample Size (mm) Composition Notes LAM 1 0.1 x 0.08 60% light yellow core, 30% yellow-orange rim, 10% inclusions dirty brown plus possibly quartz grains moderately smooth edges with occasional pits LAM 2 0.5 x 0.2 30% light yellow generally closer to centre of grain, 68% yellow-orange, 2% inclusions/filled crevice  dirty brown to red fine grained generally rough edged, angular in places. Grain as a whole exhibits evidence of folding LAM 3 Lost during preparation   LAM 4 Lost during preparation   LAM 5 Lost during preparation   LAM 6 0.25 x 0.15 98% apparently homogenous yellow-orange, 2% inclusion/cavity fine grained clear, also fine grained black coating in large embayment Smooth edges with some embayments, also possible annealed fold on flat edge. LAM 7 0.3 x 0.15 90% light yellow, 10% yellow-orange concentrated near tip of grain, no inclusions at this scale Smooth long edges with relatively ragged ends LAM 8 0.25 x 0.15 70% light yellow core, 5% yellow-orange rim, 5% quartz inclusions, 14% dirty brown to black fine grained inclusions, 1% 3 grey-pink opaque round inclusions in core of grain Relatively smooth edges LAM 9 0.15 x 0.15 98% apparently homogenous yellow-orange, 2% quartz inclusion rough edged with quartz grain stuck into edge LAM 10 Lost during preparation       WIN 1 1 x 0.6 85% light yellow, 3% red orange faint rim in places, 12% inclusion both quartz and very fine grained dark Smooth edges with 4 or 5 pockets filled with sediment, main inclusion seems to be one of these pockets folded inside grain WIN 2 0.35 x 0.2 80% light yellow, 20% yellow-orange Edge smooth to irregular, surface slightly smeared WIN 3 0.2 x 0.15 light yellow well rounded, small fold WIN 4 Lost during preparation    184 Sample Size (mm) Composition Notes WIN 5 0.45 x 0.25 90% light yellow, 10% yellow-orange partial rim, 1 small subrounded opaque inclusion (magnetite?) quite smooth edges with occasional pits WIN 6 0.35 x 0.35 95% light yellow, 5% yellow-orange thin rim Sub rounded edges, 1 major fold WIN 7 Lost during preparation       MIS 1 1 x 1 95% light yellow, 5% fine grained brown to red (hematite?) inclusions Well rounded grain, 2 embayments filled with sediment, possible fold MIS 2 1.2 x 0.6 75% light yellow, 15% yellow orange rim, 8% Very light yellow beneath orange rim on one side, 2% inclusion fine grained brown relatively smooth edge pitted in places, 2 small cavities mid grain MIS 3 1 x 0.5 98% light yellow, 2% trace yellow orange on rim rounded edges with some small pockets MIS 4 0.9 x 0.75 80% light yellow, 15% very light yellow, 2% orange yellow thin rim, 3% dark brown and quartz inclusion Faintly distinguishable lighter yellow color between core and orangish thin rim, evidence for multiple small folds and one large one, well rounded MIS 5 Lost during preparation   MIS 6 0.15 x 0.15 75% light yellow, 25% yellow-orange mostly on small arm L shaped xsection, rounded edges MIS 7 Lost during preparation       COL 1 0.7 x 0.5 90% light yellow, 10% yellow-orange rim, a few dark brown to black fine grained inclusions subrounded irregular shape, some folding COL 2 Lost during preparation   COL 3 0.5 x 0.25 80% light yellow, 18% orange yellow thin rim, 2% dark brown to black inclusions subangular edge, very obvious folding COL 4 2 x 0.5 95% light yellow, 4% yellow-orange, 1% fg dark brown to black inclusions, 1 small subangular light grey opaque inclusion Very elongate (flattened grain), rounded edges pitted in places COL 5 Lost during preparation    185  Sample Size (mm) Composition Notes COL 7 1.25 x 0.75 90% light yellow,5% yellow-orange thin rim and focused around inclusions,5% dark brown to black fg inclusions and crevice fills, also 2 sizeable nonreflective subangular grains (not quartz) subrounded with some rough edges esp near inclusions, inclusions interspersed within grain (suture?) COL 8 1.1 x 0.5 80% yellow-orange, 15% light yellow, 5% black fg inclusions subrounded edges, significant folding       AUS 1 1 X 0.33, 1 X 0.2 95% light yellow, 1% dark yellow rim in places, 4% dirty fine grained brown to black inclusion/coating (film) with specks of gold within Grain broke into 2. film on grain could be biologic in origin. Smooth edges but irregular shaped edges of gold where in contact with film AUS 2 2.5 X 1.5 95% light yellow, 5% black inclusions/voids, trace dark yellow extremely thin rim in places. Most of 'inclusions' appear to be simply voids however some are partially contained by a black substance. There is an ~0.2mm thick rim all around the grain that has the same color but a slightly different texture (almost bevelled appearance. Not sure if due to polishing. Quite smooth edges except where concave in (rough and pitted there) AUS 3 2 X 1.25 98% Light yellow, 2% void/inclusion, trace dark yellow blebs on rim of grain Same 'bevelled' texture as AUS 2, smoothed edges with 2 embayments AUS 4 2 X 1 95% Light Yellow, 4% orange-yellow patch/rim, 1% black to red brown inclusions Composed of 2 lobes, Notable patch of orange yellow on edge of grain, rough edges, some sections completely separated during polishing, bevelled AUS 5 1.8 X 1.25 100% light yellow, trace darker yellow on parts of rim, maybe the odd speck of black inclusion bevelled look, 2 embayments and one very obvious fold, moderately rounded edges          186 Sample Size (mm) Composition Notes AUS 6 2.25 X1.75 99% Light yellow, 1% dark brown to black inclusions and/or voids, trace darker yellow on some rims, dimensions are overall shape but there are multiple sections created due to very irregular shape of original grain. Evidence for multiple folding events of entire grain, rough irregular surfaces of grain, again bevelled texture (variable color in photo is just a fingerprint) AUS 7 1.25 X 0.75 99% Light Yellow, 1% voids bevelled, rounded edges except in embayments, 2 embayments, a finger and a lobe off of tapered end AUS 8 2 X 1.75 95% Light Yellow, 5% voids, trace darker yellow around rim, 3 light gray opaque inclusions (magnetite?) along textured trend so probably folded in detrital material bevelled, overall spherical shape but numerous embayments with rough edges AUS 9 1.2 X 0.75 98% Light yellow, 2% void/ brownish red fine inclusion ,occasional trace dark yellow rim (esp in embayments) bevelled, one large embayment/ folded area, evidence for past 'healed' folds, moderately rounded AUS 10 2.25 X 2 95% Light Yellow, 5% black coarse grained inclusions/infill (some are voids as well) , trace darker yellow on rim especially on ends of fingers Not as much as a bevelled look as the other samples, coarse size of black dull looking inclusions is notable, one large embayment multiple smaller ones, relatively rough edged AUS 11 3.5 X 1.25 95% Light Yellow, 5% voids/ one filled in with single large clear (quartz?) grain, No apparent dark colorations on rim slightly bevelled look, very elongate grain, multiple lobes on each end, moderately rounded AUS 12 1.2 X 0.5 97% Light Yellow, 3% fine grained brownish red infill Most of grain has been lost (fell out of resin) but imprint still preserves shape, slightly bevelled,  2 lobes connected by 'isthmus' rough edged AUS 13 1.1 X 0.8, 2.5 X 0.5 97% Light Yellow, 3% void and 1 large subhedral quartz grain inclusion, trace dark yellow along rim This sample either broke into 2 or was polished down into 2 sections, slightly bevelled, one very elongate almost flattened grain and other pretty much spherical  187  Sample Size (mm) Composition Notes AUS 14 1.25 X 0.9 65% Light Yellow, 30% single clear (quartz?) triangular shaped fully enclosed inclusion, there is a darker grain beneath the clear area as well, 5% very pale yellow colored rim differs significantly from other AUS samples, relatively rough edges and corners AUS 15 2.5 X 1.25 99% Light Yellow, 1% black inclusion, perhaps trace darker yellow on parts of rim bevelled, fold on one end of grain, smooth edges mostly except a bit rougher on inside of fold AUS 16 2.2 X 1.75 97% Light Yellow, 2% voids, 1% single fractured grey opaque inclusion (magnetite?), trace dark yellow on outermost rim bevelled, moderately rounded shape, evidence for one fold and one healed fold     BC1 5.8 X 3.75 96% very pale yellow, 4% debris infill (coarse grained dark red-brown to black), trace medium yellow on very edge of rim in places 2 embayments on either side of narrow part of middle of grain, possibly very small gold flecks occurring in the debris filling in these embayments, fairly smooth edges BC2 4 X 3.2 89% Light yellow, 10% embayments/inclusions filled with void or black fine grained sediment, 1% several grey opaque inclusions  Some of the opaque inclusions have a tinge of pink or red and are weakly fractured and angular shape, bevelled look, almost a bird shaped grain with very numerous embayments and healed folds, relatively smooth edges BC3 4.2 X 2.25 99% light yellow, 1% coarse black infill Elongated tapered grain with only 1 major embayment, rounded shape and slightly bevelled texture  188 Sample Composition     Piaba 95% Light yellow, 5% inclusion/surface contamination, trace blebs of darker yellow mostly concentrated near edge of grain, 'inclusions' are light grey fine grained and widely dispersed, some infill/inclusion on very edges of nugget are clearly detrital material, if they match other inclusions then we can determine if they are contaminants from polishing. Rounded edges for most part     Notes     3.1 g cornflake shaped nugget     1.7cm X 1.0cm X [0.094cm calculated thickness] [1/19.3g/cc/3.1g = 0.16 cc]  high fineness (rich color)     pocked surface     folded edge (0.5cm length fold on narrow end)     small (max 1mm) inclusions (rare but abundant under fold)    inclusions rounded and clear     very small dark minerals are minor component of clear inclusions (mag?)   some inclusions are pitted (dissolved?)     all quartz (acid test performed)          Dark coating 'tarnish' visible on gold     minor abrasions on gold surface     prominent feature on non-folded side     2x1mm flat spot with no inclusions or pits     grooves on surface parallel to long axis     grooves coated with black and red substance          189  Appendix B  SEM Major Element Data and LA-ICP-MS Trace Element Data Appendix B is comprised of the complete geochemical dataset for all analyses of samples reported in this work. Elements are ordered by mass from Mg-Bi and then semi-quantitatively determined elements are listed last by mass from P-U. Major element (Au and Ag) values are from SEM-EDS analysis and trace elements are by LA-ICP-MS; all values are reported as ppm. Where multiple isotopes of the same element were used to calculate elemental concentrations (Cu and Pb) each elemental concentration calculated from each isotope measured is reported. Semi-quantitative trace elements are calculated either from NIST610 (P, V, and U) or calculated using semi-quantitative approximation of the concentration of that element in the AuRM2 standard (Hg) (methodology outlined in Section 5.1.3).  As discussed in Section 5.1, P, Si, Ca, and Cr values are likely erroneous and not used to draw any conclusions for this project; they are included for the sake of completeness.            190  Locale Sample # Mg Al Si Ca Ti Cr Mn Fe Ni AUS AUS1 123.8 64.25 161.13 16.03 3.96 2.34 4.1 10.4 2.34  AUS4 0.099 0.25 40.75 6.21 <0.083 2.83 <0.67 9.49 <0.116  AUS4B 0.072 <0.152 33.58 4.89 0.157 1.01 <0.43 10.45 <0.081  AUS4C 0.143 <0.22 20.69 3.22 0.28 <1.29 1.21 11.78 <0.136  AUS5 0.093 <0.21 15.27 <2.08 0.163 2.8 <0.55 8.67 <0.154  AUS5B <0.057 <0.29 <12.27 4.91 0.21 3.02 1.15 8.15 <0.21  AUS5C 0.104 <0.27 <11.50 <2.95 <0.091 2.42 <0.70 11.87 <0.196  AUS7 <0.056 1.26 <12.85 4.24 0.49 <2.01 1.39 10.3 <0.24  AUS7B 1.4 2.1 22.28 <4.84 0.4 4.55 <1.17 15.44 <0.33  AUS9 0.81 1.29 26.87 7.19 0.34 <1.16 0.69 11.39 <0.175  AUS9B 0.5 1.09 23.95 2.84 0.36 <1.40 1.66 13.76 <0.18  AUS12 176.28 153.58 299.61 29.38 10.43 11.36 3.96 14.09 8.06  AUS14 <0.059 <0.33 50.64 8.67 0.33 1.87 1.88 16.18 <0.24  AUS2 0.071 0.41 17.17 4.57 0.173 <0.76 <0.44 10.3 0.1  AUS2B <0.019 0.95 8.51 2.55 0.227 0.55 0.86 6.19 0.095  AUS2C 0.054 <0.230 6.6 <2.42 0.063 <0.90 0.54 9.13 0.138  AUS2D 0.064 1.22 11.25 3.98 0.202 2.01 <0.56 7.4 0.264  AUS3 <0.021 0.39 5.2 2.49 0.111 0.71 0.83 5.06 <0.074  AUS3B 0.087 0.205 7.4 2.72 0.05 <0.74 <0.35 9.43 <0.088  AUS6 0.076 0.27 11.8 4.28 0.059 2.31 <0.36 7.31 0.109  AUS8 0.046 <0.158 9.35 3.93 0.145 1.95 0.46 6.58 <0.084  AUS8B 1.129 2.22 8.79 1.9 0.165 <0.64 1.03 6.08 0.106  AUS10 2.44 2.05 14.29 <2.00 0.192 4.06 <0.37 7.07 0.159  AUS10B 0.065 <0.33 22.81 <3.49 0.113 3.52 <0.61 15.81 <0.19  AUS11 <0.0168 0.89 6.74 1.18 0.341 0.96 0.73 5.85 <0.066  AUS11B 0.199 0.177 9.39 12.07 0.092 2.42 <0.32 8.64 <0.095  AUS13 0.0331 <0.103 6.59 <1.08 0.031 1.6 0.24 6.55 <0.065  AUS15 0.068 <0.20 14.19 <2.02 0.052 1.13 <0.40 14.75 <0.122  AUS15B <0.033 <0.197 12.67 <2.13 0.117 2.59 1.54 9.15 <0.130  AUS15C 0.09 <0.20 6.47 2.97 0.094 1.54 <0.41 17.93 <0.134  AUS16 0.085 <0.180 6.87 2.63 0.16 <0.72 <0.35 17.76 <0.127  AUS16B <0.020 2.7 14.68 2.17 0.6 <0.55 1.49 10.38 <0.091 BC BC1_1 0.107 0.29 42.53 8.89 0.184 <0.83 <0.52 11.67 0.194  BC1_2 <0.083 1.75 33.29 7.23 0.229 4.5 2.22 13.98 <0.28  BC1_3 0.14 1.36 54.91 14.25 0.435 7.29 3.06 19.5 0.58  BC1_4 0.225 1.67 39.27 12.46 <0.187 <3.19 <1.58 32.35 <0.36  BC2_1 <0.157 1.03 50.28 21.1 15.67 12.76 2.59 20.98 <0.48  BC2_2 0.184 1.59 32.09 13.52 0.39 8.23 1.68 24.97 <0.38  BC2_3 2.19 23.68 42.89 <8.59 1.02 <3.51 10.92 31.5 0.71  BC2_4 0.218 1.2 24.61 5.5 <0.119 4.75 <0.87 26.31 0.25  BC3_1 0.2 <0.59 29.09 <5.31 0.199 6 <1.14 27.06 0.47  BC3_2 0.269 1.07 25.72 <8.29 0.33 8.21 <1.41 29.66 <0.50  BC3_3 0.279 1.43 24.51 <4.80 0.232 3.02 <1.18 29.07 0.41  BC3_4 2.45 <0.31 16.5 8.58 0.217 2.93 2.99 12.85 0.27 COL COL1 0.106 0.49 39.35 4.85 0.116 3.21 2.43 10.9 <0.217  COL1B 0.14 <0.68 17.22 8.24 0.266 <2.36 3.35 21.93 <0.34  191 Locale Sample # Mg Al Si Ca Ti Cr Mn Fe Ni COL COL3 0.058 <0.33 10.87 2.99 0.144 <1.13 0.9 13.39 <0.162  COL3B <0.036 <0.31 16.6 <3.02 <0.071 4.14 1.78 9.05 0.165  COL4 0.044 <0.22 17.38 2.39 0.082 2.61 0.48 7.49 0.195  COL4B 0.082 0.32 15.43 <2.31 0.068 0.86 <0.46 9.76 <0.109  COL4C <0.0187 <0.162 9.34 <1.88 0.068 2.3 1.16 6.47 0.175  COL7 1.28 4948.58 750.85 17.19 54.56 19.92 0.75 11.38 1.56  COL7B 0.099 <1.12 18.65 8.6 <0.28 <2.93 <0.73 12.77 <0.31 KEL LAM2 0.098 0.28 1.4 0.9 <0.090 0.95 <0.25 1.3 <0.063  LAM2B 0.52 3.95 **** <**** 1.12 26.54 3.07 7.99 <0.36  LAM2C 1.23 8.81 <**** <**** 0.45 **** <0.96 6.24 <0.19  LAM6 0.95 11.84 1 7.29 0.48 17.12 2.11 7.88 <0.33  LAM7 <0.155 <0.82 3.76 **** <0.49 10.54 3.16 10.84 <0.41  LAM7B <0.135 0.95 <**** **** <0.42 **** 2.29 10.72 <0.34  LAM8 3.19 21.56 <**** <**** 0.71 <70.68 <1.04 7.28 0.23  MIS1 <0.19 <0.34 33.92 30.08 <0.46 116.32 <0.95 33.77 <0.109  MIS1B 0.32 0.41 <**** <**** 0.76 **** <0.85 27.11 <0.14  MIS1C 0.68 <0.44 <**** <**** 2.08 4834.01 <1.21 39.79 1.62  MIS2 52.64 5.21 <**** <**** 56.61 6240.48 <2.43 76.05 3.54  MIS2B <0.89 <0.86 109.54 194.59 <1.56 <85.51 <2.61 39.36 <0.27  MIS4 1.26 1.77 309.7 391.72 2.1 80.25 8.97 62.53 0.62  MIS4B <0.83 <4.06 147.83 193.51 <2.09 67.66 <4.89 49.95 <0.60  WIN1 <0.47 <0.83 33.62 146.66 <0.99 29.77 2.57 13.02 0.89  WIN1B <0.77 <1.32 28.3 147.07 <1.55 22.04 <3.25 12.76 <0.48  WIN1C 0.59 <0.91 <**** <**** <1.08 <**** <2.32 14.09 <0.36  WIN2 <18.25 <4.58 2768.21 1448.55 <13.13 5273.97 26.59 42.84 <1.31  WIN3 <1.50 10.06 65.21 98.78 <2.44 63.5 6.32 24.38 <0.67  WIN5 <0.53 <0.87 13.92 26.96 <1.00 7.89 <2.15 11.35 <0.41  WIN5B 5.25 60.97 <**** <**** 21.11 0.95 <1.74 7.36 0.35  WIN5C 1.71 10.46 2.1 <**** 8.07 8.71 4.58 7.93 <0.32  WIN6 0.67 <0.88 <**** <**** <0.93 0.86 <2.32 9.77 <0.49 PIABA LongLeft_1 0.0187 <0.114 10.07 16.07 0.078 1.54 <0.24 3.51 <0.059  LongLeft_2 0.0484 0.184 16.86 <2.12 0.081 1.15 <0.31 4.63 <0.082  LongLeft_3 0.0322 <0.122 5.68 <1.61 0.072 0.68 <0.25 3.81 <0.066  LongLeft_4 0.038 <0.120 4.57 1.79 0.091 <0.58 0.34 4.04 <0.063  LongRight_1 0.0456 <0.139 8.57 <1.80 0.082 <0.77 0.64 4.48 <0.079  LongRight_2 0.0367 0.15 <2.54 <1.68 <0.035 <0.66 <0.28 4.36 <0.076  LongRight_3 <0.0176 <0.141 2.87 <1.68 0.089 <0.53 0.58 3.5 0.132  LongRight_4 0.0319 <0.100 <1.60 <1.23 <0.025 2.03 <0.22 3.83 <0.059  Short_1 0.037 <0.23 8.14 <2.66 0.059 2.65 <0.41 7.22 <0.136  Short_2 0.026 <0.145 5.79 <1.70 0.043 1.89 3.06 3.65 0.549  Short_3 0.076 <0.115 3.37 2.31 <0.027 1.65 2.93 3.9 1.018  Short_4 0.04 <0.142 4.09 1.77 <0.034 2.84 <0.24 3.39 0.14  192 Locale Sample # Cu63 Cu65 Zn As Se Rh Pd Ag Sn AUS AUS1 309.2 275.97 5.05 0.47 6.53 0.046 0.066 107618.95 <0.045  AUS4 20.51 19.52 <0.34 <0.60 12.01 <0.0060 0.286 153422.14 <0.060  AUS4B 19.36 19.09 <0.22 4.08 8.06 0.0198 0.205 153422.16 0.174  AUS4C 64.2 61.7 <0.30 5.9 9.11 0.052 0.279 153422.14 0.143  AUS5 245.61 212.73 0.26 0.98 9.57 0.0177 0.217 89372.12 <0.039  AUS5B 457.46 407.7 <0.42 3.61 10.56 0.067 0.228 89372.12 <0.074  AUS5C 301.61 274.94 <0.40 1.2 11.63 0.026 0.195 89372.12 <0.083  AUS7 163.44 154.36 <0.49 3.61 10.74 0.048 0.117 101660.79 0.12  AUS7B 174.28 166.75 <0.67 <1.17 18.76 0.0154 0.163 101660.79 <0.107  AUS9 26.63 25.22 0.34 3.13 9.74 0.025 0.125 80621.09 <0.056  AUS9B 28.84 28.29 <0.35 3.6 13.3 0.033 0.179 85275.89 0.09  AUS12 55.62 54.62 4.99 <0.85 11.62 0.0063 0.192 48689.18 <0.061  AUS14 132.35 126.58 <0.41 5.48 12.64 0.051 0.2 192801.7 0.21  AUS2 91.8 85.49 <0.22 3.06 9.63 0.0277 0.187 75873.17 0.114  AUS2B 92.3 87.08 <0.17 3.68 5.13 0.0392 0.03 75873.18 0.077  AUS2C 81.36 76.96 <0.244 2.36 7.83 0.0292 0.206 75873.18 0.048  AUS2D 75.44 71.18 0.6 <1.15 7.69 0.0103 0.112 75873.18 <0.080  AUS3 397.32 324.68 0.229 2.73 4.79 0.051 0.054 48037.5 <0.033  AUS3B 387.92 330.85 <0.193 <0.60 7.63 0.03 0.155 48037.5 <0.034  AUS6 315.27 276.33 <0.19 <0.60 8.19 0.0254 0.121 81831.32 <0.041  AUS8 61.79 58.91 0.4 1.06 6.52 0.0087 0.064 72987.21 <0.034  AUS8B 53.21 51.47 0.21 3.17 5.57 0.0299 0.056 72987.21 0.083  AUS10 294.34 265.43 0.44 0.43 7.75 0.0112 0.065 105477.68 <0.045  AUS10B 210.99 202.35 0.44 <0.72 15.49 0.0324 0.128 101102.18 <0.063  AUS11 229.91 203.88 0.33 1.94 5.84 0.02 0.056 114042.52 0.057  AUS11B 227.91 201.53 0.42 <0.33 9.34 0.0113 0.15 115438.95 <0.035  AUS13 306 261.95 0.181 <0.34 5.51 0.0167 0.095 64608.57 <0.020  AUS15 114.57 107.83 <0.24 0.24 13.28 0.0078 0.186 90675.43 <0.042  AUS15B 93.21 87.03 0.38 2.28 9.03 0.0326 0.027 89185.91 0.05  AUS15C 87.52 83.97 0.35 0.45 13.96 0.0053 0.194 89185.9 <0.054  AUS16 137.46 125.99 <0.23 3.12 13.82 0.0171 0.227 82110.61 0.096  AUS16B 144.96 136.02 <0.132 4.79 7.96 0.046 0.048 82110.62 0.062 BC BC1_1 9.03 8.64 <0.27 1.48 12.34 <0.0034 0.437 462873.19 0.085  BC1_2 16.82 17.71 <0.66 4.86 15.19 0.057 0.271 462873.19 0.346  BC1_3 4.59 4.66 0.72 8.42 20.71 0.085 0.434 395378.63 0.55  BC1_4 44.59 43.92 0.77 8.79 36.72 0.036 0.68 395378.63 <0.169  BC2_1 434.79 397.42 1.49 <2.58 21.73 <0.026 0.147 114042.58 <0.237  BC2_2 369.13 351.48 1.56 <3.42 36.3 <0.0121 0.206 114042.58 <0.157  BC2_3 350.28 309.62 <0.88 43.22 32.57 0.278 0.165 114042.57 0.54  BC2_4 362.19 328.19 <0.54 <1.37 26.89 0.0223 0.5 114042.56 <0.091  BC3_1 309.69 285.29 1.32 <1.76 31.94 <0.0144 0.54 133778.91 <0.122  BC3_2 320.28 294.25 1.85 1.78 33.53 0.034 0.5 133778.91 <0.218  BC3_3 324.27 294.8 <0.54 8.08 32.81 0.062 0.68 131451.52 <0.120  BC3_4 319.76 286.55 0.81 4.62 14.07 0.0348 0.218 131451.52 0.258 COL COL1 36.58 35.07 0.92 8.23 11.05 0.044 0.075 150256.81 0.195  COL1B 38.87 36.91 <0.66 9.57 19.67 0.056 0.388 150256.83 <0.161  193 Locale Sample # Cu63 Cu65 Zn As Se Rh Pd Ag Sn COL COL3 35.91 33.71 <0.26 7.89 13.66 0.045 0.212 117393.96 0.12  COL3B 34.48 34.14 0.53 6.43 9.41 0.0334 0.069 117393.97 0.159  COL4 49.02 47.37 0.59 <0.36 9.53 <0.0031 0.074 142529.86 <0.041  COL4B 47.38 45.87 <0.203 1.93 10.12 <0.00 0.186 138433.64 <0.029  COL4C 51.22 48.21 0.36 2.44 6.3 0.0135 0.03 131172.14 0.102  COL7 10.02 9.44 2.17 0.97 8.82 <0.0052 0.085 248007.55 0.21  COL7B 12.8 11.5 0.58 <2.87 14.7 <0.0133 0.226 248007.52 <0.075 KEL LAM2 1.95 1.75 0.41 <0.44 2.01 <0.0026 0.016 19270.86 <0.0192  LAM2B 14.68 13 <1.71 11.74 8.93 0.083 <0.057 143460.88 0.29  LAM2C 16.57 15.65 1.07 <2.10 6.23 0.0056 0.141 143460.88 <0.097  LAM6 20.52 19.24 <1.49 <3.50 8.8 <0.0111 0.65 74104.38 0.51  LAM7 471.86 376.45 <1.71 12.99 11.46 0.065 0.27 114228.76 <0.182  LAM7B 324.67 256.34 <1.43 12.12 10.2 0.053 0.22 113763.26 <0.160  LAM8 3.26 2.71 <0.87 8.57 11.31 0.0095 0.16 245028.59 <0.084  MIS1 40.41 39.84 <1.35 2.8 17.17 0.0174 1.7 111994.45 0.198  MIS1B 86.9 82.29 <1.26 <1.28 13.82 0.0075 2.15 111994.45 0.091  MIS1C 67.89 63.4 <1.66 3.8 31.69 0.029 1.84 111994.45 <0.101  MIS2 52.94 49.36 62.61 9.87 41.55 0.045 0.33 177627.11 0.161  MIS2B 54.43 53.97 <4.02 <5.31 29.77 0.0122 <0.057 178278.8 <0.16  MIS4 7.27 8.23 <3.67 43.9 44.16 0.19 0.28 244376.92 1.32  MIS4B 5.81 5.67 <3.94 <8.36 51.03 <0.035 0.19 251359.11 0.23  WIN1 63.54 61.15 <2.50 10.02 10.77 0.037 0.242 179582.08 0.182  WIN1B 67.77 64.21 <3.88 <4.10 11.87 0.0166 0.31 179582.06 <0.218  WIN1C 64.29 61.62 <2.56 <3.44 11.71 0.0148 0.35 180792.31 <0.156  WIN2 101.96 94.7 <34.20 75.88 27.85 0.4 0.33 69821.95 1.38  WIN3 40.82 41.47 <5.27 27.73 17.53 <0.044 0.44 209465.83 <0.22  WIN5 136.59 119.18 <2.24 <2.71 10.2 <0.0099 0.142 120186.84 <0.14  WIN5B 88.32 79.89 <1.33 <2.09 5.84 0.0115 0.132 120186.84 <0.107  WIN5C 24.19 21.1 1.71 9.57 5.52 0.036 0.096 120000.65 <0.080  WIN6 84 74.57 <2.08 3.13 9.15 <0.015 14.26 211979.42 0.24 PIABA LongLeft_1 343.66 268.63 0.194 <0.50 5.59 0.0169 0.11 83507.09 0.025  LongLeft_2 294.56 243.7 0.203 <0.83 5.59 0.0132 0.105 81458.96 <0.033  LongLeft_3 286.4 231.52 <0.137 1.91 4.52 0.0216 0.109 79876.34 0.063  LongLeft_4 282.71 236.54 <0.124 2.36 5.01 0.0334 0.089 79690.16 0.055  LongRight_1 283.82 237.8 <0.154 2.31 5.35 0.0285 0.141 84158.77 0.103  LongRight_2 284.28 233.4 <0.139 1.04 4.84 0.0121 0.135 82576.13 <0.022  LongRight_3 291.19 232.62 <0.128 2.07 3.95 0.029 0.1 82576.13 0.088  LongRight_4 277.97 231.8 0.121 <0.39 4.27 0.0144 0.111 83972.57 0.0162  Short_1 205.16 186.21 <0.27 <0.68 8.98 0.0139 0.18 74569.88 <0.043  Short_2 230.71 202.19 0.46 1.72 4.17 0.0208 0.04 78759.19 <0.022  Short_3 240.81 203.32 <0.118 <0.28 4.76 0.0111 0.166 78759.19 <0.0192  Short_4 246.9 215.74 0.51 <0.00 4.94 0.0152 0.109 86672.35 <0.030  194 Locale Sample # Sb Te Pt Au Pb204 Pb206 Pb207 Pb208 AUS AUS1 0.085 0.57 1.35 892381.05 85030.54 3.33 3.03 2.96  AUS4 4.2 1 1.19 846577.86 4217.63 0.03 <0.0191 0.035  AUS4B 2.7 0.49 1.08 846577.84 13733.14 0.029 0.072 0.042  AUS4C 12.67 0.62 2.12 846577.86 4651.66 <0.0135 0.113 0.019  AUS5 34.27 0.82 1.34 910627.88 1112.69 0.044 <0.0131 0.035  AUS5B 66.57 1.1 2.34 910627.88 2411.42 0.053 0.073 0.026  AUS5C 45.74 0.58 1.62 910627.88 2485.44 <0.0189 0.03 0.03  AUS7 <0.067 0.96 1.55 898339.21 114228.98 0.06 0.083 <0.0150  AUS7B <0.088 1.59 1.96 898339.21 <**** 0.12 <0.037 0.068  AUS9 4.69 0.68 1.83 888378.91 <**** 0.041 0.038 0.047  AUS9B 7.77 0.75 1.44 883724.11 <**** 0.022 0.089 0.044  AUS12 0.33 0.75 1.23 951310.82 <**** 0.51 0.42 0.58  AUS14 11.24 0.99 2 807198.3 <**** 0.066 0.1 <0.0147  AUS2 0.254 0.63 0.87 924126.83 883.17 <0.0110 0.097 0.044  AUS2B 0.09 0.33 0.8 924126.82 898.32 0.0216 0.109 0.0088  AUS2C 0.196 0.38 0.91 924126.82 958.09 0.0222 0.083 0.0338  AUS2D <0.060 0.86 0.85 924126.82 902.61 0.064 <0.0165 0.0325  AUS3 0.125 0.51 0.83 951962.5 4654.71 0.057 0.04 0.0065  AUS3B 0.304 0.5 0.88 951962.5 2948.4 0.0259 0.03 0.0199  AUS6 4.47 0.66 1.03 918168.68 10839.51 0.05 <0.0131 0.0293  AUS8 0.212 0.56 0.96 927012.79 66066.57 0.059 <0.0123 0.0171  AUS8B 0.38 0.56 0.85 927012.79 68682.65 0.047 0.07 0.0085  AUS10 0.067 0.86 1.35 894522.32 41782.65 2.09 0.153 0.229  AUS10B 0.11 1.44 1.14 898897.82 26003.09 0.168 <0.015 0.039  AUS11 0.54 0.3 1.1 885957.48 6875.65 0.077 0.035 0.0267  AUS11B 0.56 0.61 1.2 884561.05 7043.71 0.038 0.0106 0.0307  AUS13 1.15 0.42 0.68 935391.43 4876.34 0.039 <0.0062 0.0117  AUS15 0.171 0.56 0.91 909324.57 3471.23 <0.0100 0.0148 0.049  AUS15B <0.025 0.66 0.84 910814.09 29698.81 0.068 0.023 0.0062  AUS15C 0.154 0.79 0.95 910814.1 26017.05 0.03 0.024 0.033  AUS16 0.239 0.59 0.88 917889.39 2926.05 <0.0112 0.046 0.041  AUS16B 0.102 0.42 0.97 917889.38 3013.06 0.029 0.078 <0.0042 BC BC1_1 2000.35 1.01 1.58 501999.59 95969.65 0.023 0.092 0.064  BC1_2 1003.45 1.74 1.48 501999.59 93653.59 0.144 0.084 0.064  BC1_3 4533.75 1.84 1.7 604596.69 211952.5 0.168 0.189 0.081  BC1_4 1487.19 1.78 1.72 604596.69 131833.66 <0.046 0.154 0.15  BC2_1 20.61 3.03 2.09 884995.56 31012.76 0.268 0.083 <0.047  BC2_2 15.62 2.88 2.2 884995.5 33398.66 0.244 <0.050 0.088  BC2_3 12.61 5.28 2.02 884995.5 26356.06 <0.049 0.69 <0.030  BC2_4 17.82 2.14 2.25 884995.56 26278.55 0.076 0.142 0.115  BC3_1 0.273 2.48 2.24 866995.5 30882.75 0.153 <0.033 0.12  BC3_2 0.47 2.32 2.23 866995.44 25913.47 0.201 <0.089 0.121  BC3_3 0.54 1.91 2.57 866995.5 31114.72 <0.023 0.075 0.144  BC3_4 0.072 1.52 2.1 866995.5 49819.85 0.133 0.053 0.0121 COL COL1 <0.061 1.36 0.532 849743.19 828.2 0.126 0.158 0.0258  COL1B 0.296 1.63 0.63 849743.17 812.83 <0.038 0.155 0.068  195 Locale Sample # Sb Te Pt Au Pb204 Pb206 Pb207 Pb208 COL COL3 0.199 1.28 0.62 882606.04 240.7 <0.0154 0.108 0.057  COL3B <0.046 1.25 0.58 882606.03 264.72 0.076 0.107 <0.0053  COL4 0.069 1.18 0.75 857470.14 58.69 0.08 0.0065 0.033  COL4B 0.172 0.9 0.71 861566.36 52.77 <0.0132 0.029 0.033  COL4C <0.024 0.87 0.64 868827.86 65.87 0.059 0.0238 <0.0044  COL7 0.45 1.85 1.03 751992.45 791.58 3.91 3.64 3.97  COL7B 0.6 2.11 0.88 751992.48 978.17 0.107 <0.092 0.044 KEL LAM2 1.34 0.21 0.103 980729.14 1.75 0.015 <0.0090 <0.0042  LAM2B 13.71 1 0.79 856539.12 18.7 0.069 0.25 <0.028  LAM2C 19.2 0.93 0.35 856539.12 11.09 <0.030 <0.013 0.068  LAM6 0.12 1.31 0.37 925895.62 149.03 0.18 0.062 0.062  LAM7 28.72 2.35 1.24 885771.24 30.85 0.13 0.18 <0.033  LAM7B 19.15 0.76 0.93 886236.74 23.24 0.08 0.12 0.045  LAM8 0.24 0.69 0.45 754971.41 44.59 0.23 0.2 0.22  MIS1 0.76 <0.39 2.07 888005.55 1816.94 0.038 0.038 0.0179  MIS1B 2.39 0.97 1.86 888005.55 2729.64 0.025 0.069 1.24  MIS1C 1.3 0.98 1.65 888005.55 2076.61 0.027 0.038 0.023  MIS2 22.7 <0.91 1.56 822372.89 164.28 <0.045 0.09 <0.038  MIS2B 32.89 1.15 1.27 821721.2 137.42 <0.070 0.024 <0.0221  MIS4 4.71 <1.16 3.15 755623.08 1063.55 <0.076 0.2 <0.035  MIS4B 5.45 2.34 1.91 748640.89 646.44 <0.089 <0.074 <0.059  WIN1 17.87 0.78 1.38 820417.92 273.36 <0.068 0.081 0.064  WIN1B 26.69 1.55 1.59 820417.94 253.41 <0.058 <0.074 <0.042  WIN1C 21.51 1.62 1.44 819207.69 271.16 0.049 0.056 0.046  WIN2 1.56 5.59 1.68 930178.05 1158.44 <0.145 0.49 <0.15  WIN3 0.71 2.21 1.51 790534.17 2646.8 <0.074 0.33 0.151  WIN5 <0.098 1.77 2 879813.16 27.17 0.075 0.037 0.079  WIN5B 0.31 1.16 1.43 879813.16 20.27 0.31 0.151 0.25  WIN5C 0.076 2.48 0.72 879999.35 16.12 0.153 0.24 0.166  WIN6 2.92 5.93 1.76 788020.58 163.61 0.37 0.25 0.23 PIABA LongLeft_1 0.055 0.504 1.062 916492.91 536.54 0.0389 0.0136 0.018  LongLeft_2 0.099 0.52 0.966 918541.04 493.89 0.028 0.0101 0.0205  LongLeft_3 0.074 0.512 1.037 920123.66 560.56 <0.0064 0.0371 0.0144  LongLeft_4 0.086 0.478 0.917 920309.84 587.69 <0.0057 0.0417 0.0154  LongRight_1 0.116 0.499 0.98 915841.23 671.69 <0.0050 0.051 0.0213  LongRight_2 0.111 0.451 0.976 917423.87 679.56 0.0055 0.0093 0.0301  LongRight_3 0.069 0.345 0.935 917423.87 676.22 0.0149 0.0315 0.0118  LongRight_4 0.095 0.53 0.872 916027.43 672.24 0.0152 <0.0052 0.0157  Short_1 0.171 1.17 0.801 925430.12 544.65 0.068 0.032 0.041  Short_2 <0.018 0.71 0.932 921240.81 667.49 0.07 <0.0078 0.0144  Short_3 0.134 0.63 0.882 921240.81 723.99 0.0205 0.0239 0.0287  Short_4 0.066 0.87 0.902 913327.65 781.79 0.071 0.0137 0.0268  196 Locale Sample # Bi P V Hg U AUS AUS1 0.044 113.09 0.83 548434.06 0.087  AUS4 0.0056 184.04 0.121 27702.06 <0.012  AUS4B 0.0035 145.19 0.107 63427.68 0.11  AUS4C 0.0066 145.94 0.091 19548.38 0.15  AUS5 0.132 163.45 0.064 2815.24 <0.015  AUS5B 0.32 181.76 <0.039 4156.05 0.1  AUS5C 0.25 188.42 0.105 2110.4 <0.0195  AUS7 0.0102 194.24 <0.045 22645.92 0.16  AUS7B 0.034 323.01 0.085 24347.15 <0.023  AUS9 0.0074 247.81 0.102 239471.3 0.11  AUS9B <0.0044 263.8 0.167 234603.66 0.11  AUS12 0.0057 237.48 1.84 156390.95 <0.023  AUS14 0.96 278.38 0.072 238682.03 0.15  AUS2 <0.0029 204.64 0.093 7145.83 0.083  AUS2B <0.0017 110.81 0.024 6383.5 0.128  AUS2C <0.0031 174.74 0.068 6198.69 0.108  AUS2D 0.0204 134.19 <0.0266 4910.07 <0.024  AUS3 0.0181 86.31 <0.0163 25064.24 0.12  AUS3B 0.0096 176.78 0.107 16034.86 0.035  AUS6 0.0235 149.11 0.031 45560.14 0.016  AUS8 0.0196 152.69 <0.0140 218677.69 0.059  AUS8B 0.0239 151.98 0.017 213599.73 0.17  AUS10 0.213 183.21 0.031 130340.54 13.78  AUS10B 0.43 395.06 0.059 82828.98 <0.040  AUS11 0.0165 110.91 0.013 21695.03 0.09  AUS11B 0.0339 173.77 0.062 21808.86 0.007  AUS13 0.0119 139.31 0.0176 14928.35 <0.018  AUS15 <0.0024 277.65 0.15 11026.7 <0.049  AUS15B 0.0238 176.45 <0.028 69161.51 0.34  AUS15C 0.0045 310.28 0.135 58498.59 <0.078  AUS16 <0.0016 284.81 0.143 8723.76 0.37  AUS16B <0.00153 181.06 <0.0124 8258.07 1.09 BC BC1_1 0.52 246.35 0.087 242384.97 0.06  BC1_2 0.45 271.45 <0.025 191919.98 0.07  BC1_3 1.3 349.99 0.125 328150.22 0.32  BC1_4 0.55 568.9 0.32 184927.06 0.027  BC2_1 0.061 472.56 <0.062 53577.11 0.3  BC2_2 0.087 467.42 0.103 41361.88 0.016  BC2_3 <0.0053 430.63 <0.078 31845.89 0.53  BC2_4 0.158 378.49 0.36 26124.04 <0.021  BC3_1 0.042 423.65 0.26 30291.55 <0.013  BC3_2 0.051 516.38 0.27 25489.54 <0.00  BC3_3 <0.01 479.79 0.42 25972.87 0.032  BC3_4 0.029 158.88 0.021 29329.08 0.023 COL COL1 0.073 170.1 <0.089 2127.64 0.26  COL1B 0.0197 311.64 0.32 1872.88 0.25  197 Locale Sample # Bi P V Hg U COL COL3 0.0092 211.32 0.28 487.96 0.16  COL3B 0.026 166.09 <0.022 547.87 0.13  COL4 0.071 170.14 0.08 107.24 <0.0103  COL4B 0.0201 156.12 0.19 102.46 0.02  COL4C 0.033 109.92 <0.020 127.75 0.032  COL7 0.156 184.17 16.13 1774.03 0.1  COL7B 0.073 254.17 0.125 2308.28 <0.061 KEL LAM2 0.005 26.67 <0.018 3.58 0.005  LAM2B <0.0122 118.94 <0.089 30.87 0.16  LAM2C 0.017 89.49 0.111 15.85 <0.0174  LAM6 0.045 119.15 0.085 254.98 0.19  LAM7 0.045 183.61 0.13 53.28 0.2  LAM7B 0.021 178.57 0.12 39.17 0.35  LAM8 0.024 122.76 0.12 83.15 <****  MIS1 1.13 629.4 <0.131 5467.4 1.49  MIS1B 0.085 386.36 <0.132 5621.27 <0.38  MIS1C 0.078 929.23 0.23 6015.77 0.55  MIS2 0.169 1328.41 1.42 294.74 0.8  MIS2B 0.124 1009.84 <0.45 335.61 <0.51  MIS4 0.035 1454.46 <0.20 2554.55 1.59  MIS4B 0.081 1081.06 <0.80 1328.73 <0.49  WIN1 0.039 201.66 1.39 608.58 0.27  WIN1B 0.039 201.92 <3.35 536.3 0.09  WIN1C 0.066 246.51 1.51 547.79 0.12  WIN2 0.133 552.51 3.12 2302.67 1.64  WIN3 0.144 252.74 <0.82 6875.53 0.79  WIN5 0.035 122.84 <0.24 38.19 <0.06  WIN5B 0.093 94.91 0.29 30.56 0.05  WIN5C 0.047 121.39 0.2 28.91 0.34  WIN6 0.38 113.81 <0.17 252.47 <0.07 PIABA LongLeft_1 0.0111 90.98 0.029 1342.16 <0.012  LongLeft_2 0.0107 94.34 0.044 1092.78 <0.014  LongLeft_3 0.0052 84.73 0.039 1337.08 0.05  LongLeft_4 0.00147 84.92 0.031 1362.65 0.06  LongRight_1 0.0141 93.17 0.043 1545.13 0.06  LongRight_2 <0.00223 82.38 0.04 1580.93 0.023  LongRight_3 0.0035 62.38 0.033 1630.81 0.039  LongRight_4 0.0069 76.48 0.043 1506.35 <0.0055  Short_1 0.0194 138.05 0.071 964.44 <0.0078  Short_2 0.0207 68.56 0.0115 1301.77 0.017  Short_3 0.0083 74.74 0.047 1503.98 <0.0024  Short_4 0.0381 89.92 0.023 1443.49 0.0039  198 Appendix C  Reference Material LA-ICP-MS Trace Element Data Appendix C is comprised of the geochemical dataset for analyses of standard reference materials reported in this work. All values are recorded in parts per million. Elements are ordered by mass from Mg-U. Major elements in FAU7 and AuRM2 (Au and Ag) are not reported. Trace elements were determined by LA-ICP-MS using AuRM2 as the external standard. Corresponding data sequentially using NIST610 and FAU7 were also calculated but are excluded for the sake of brevity. Where multiple isotopes of the same element were used to calculate elemental concentrations (Cu and Pb) each elemental concentration calculated from each isotope measured is reported. Semi-quantitative trace elements are calculated using the AuRM2 reference material values calculated either from NIST610 (P, V, and U) or calculated using semi-quantitative approximation of the concentration of that element in the AuRM2 standard (Hg) (methodology outlined in Section 5.1.3).  As discussed in Section 5.1, P, Si, Ca, and Cr values are likely erroneous and not used to draw any conclusions for this project; they are included for the sake of completeness. 199 Run  Analysis #   Mg        Al  P  Ti  V Mn             Fe 1 NIST610_1 328.11 3679.39 163.91 477.29 327.03 230.51 5.42 1 NIST610_2 346.71 4140.76 192.89 490.04 331.67 248.1 6.43 1 FAU7_1 28.87 0.83 140.13 13.44 <0.025 50.17 8 1 FAU7_2 30.29 <0.24 272.95 13.8 0.206 60.08 15.08 1 AuRM2_1 9.8 29.26 539.33 32.18 0.44 33.7 33.73 1 AuRM2_2 10.15 28.12 525.75 31.74 0.333 25.16 28.16 1 AuRM2_3 8.84 24.74 472.91 28.3 0.4 24.35 23.41 1 AuRM2_4 15.33 48.11 885.87 50.12 0.3 47.4 54.7 1 FAU7_3 43.06 1.92 667.73 20.3 <0.076 89.92 31.17 1 FAU7_4 55.98 1.07 671.07 19.78 0.16 100.59 38.39 1 NIST610_3 633.54 10405.03 883.68 1315.37 1280.6 460.02 27.64 1 NIST610_4 622.46 10701.51 917.87 1290.13 1439.79 417.98 30.04 2 NIST610_1 396.09 5714.57 269.64 681.66 319.27 249.02 8.55 2 NIST610_2 371.17 5592.03 281.19 664.57 315.16 251.33 7.11 2 FAU7_1 31.53 2.11 216.77 12.33 0.175 53.89 10.52 2 FAU7_2 30.36 2.75 515.28 12.68 0.278 56.29 23.93 2 AuRM2_1 9.87 26.08 534.43 33.58 0.378 34.08 30.48 2 AuRM2_2 9.92 30.5 529.21 30.06 0.383 23.99 30.06 2 AuRM2_3 10.05 29.98 473.74 30.01 0.371 26.34 27.29 2 AuRM2_4 9.79 27.12 580.26 32.89 0.389 29.83 33.07 2 FAU7_3 35.44 0.99 1002 13.47 0.75 76.69 62.75 2 FAU7_4 37.37 1.67 1092.77 13.31 0.11 79.72 59.86 2 NIST610_3 395.65 6171.01 1220.52 752.06 750.52 386.89 37.95 2 NIST610_4 399.36 6090.54 1218.43 804.49 761.72 381.95 37.55 3 NIST610_1 196214.25 2969821 218437.66 423231.41 130274.88 166272.13 6672.21 3 NIST610_2 404.98 6832.05 453.23 838.77 447.33 502.59 14.42 3 FAU7_1 59.72 2.14 441.22 26.46 0.063 118.42 22.22 3 FAU7_2 67.8 931.6 508.78 46.03 0.102 127.6 24.39 3 AuRM2_1 8.91 24.32 491.49 29.44 0.42 27.81 27.02 3 AuRM2_2 11.78 36.8 605.62 35.61 0.333 28.98 36.29 3 AuRM2_3 10.97 32.83 470.74 30.74 0.32 26.25 29.3 3 AuRM2_4 8.99 24.12 585.73 32.16 0.46 30.01 30.53 3 FAU7_3 29.4 1.78 372.46 12.34 <0.14 59.31 21.18 3 FAU7_4 45.69 <0.76 514.34 15.23 1.09 72.63 34.12 3 NIST610_3 394.77 6176.61 281.88 773.15 266.08 304.11 10.72 3 NIST610_4 394.21 6222.94 264.37 757.64 252.65 276.3 9.45 4 NIST610_1 420.37 7056.2 295.71 877.51 979.71 364.06 17.81 4 NIST610_2 424.08 6822.84 337.15 839.02 927.72 359.62 15.36 4 FAU7_1 37.58 1.93 286.46 14.02 <0.086 70.28 21.45 4 FAU7_2 37.28 <0.77 360.12 15.33 <0.124 64.98 25.1 4 AuRM2_1 10.05 27.72 510.89 32.28 0.29 32.36 32.35 4 AuRM2_2 9.66 28.84 571.97 30.83 0.5 23.49 27.81 4 AuRM2_3 10.15 30.85 428.12 30.19 0.41 28.01 26.85 4 AuRM2_4 9.62 25.38 719.82 33.71 0.33 28.92 36.06 4 FAU7_3 33.18 1.61 691.78 12.71 0.27 56.08 25.6 4 FAU7_4 30.98 11.07 640.65 13.26 0.29 56.15 41.89  200 Run  Analysis #  Mg    Al    P   Ti    V   Mn     Fe  4 NIST610_3 384.92 6667.22 399.24 863.22 714.52 261.62 18.17 4 NIST610_4 415.38 6804.07 464.83 890.3 746.8 277.92 18.45 5 NIST610_1 380.14 5910.6 263.75 797.15 835.84 279.93 12.12 5 NIST610_2 382.27 6253.16 273.27 814.09 794.18 288.92 11.5 5 FAU7_1 34.99 <1.63 335.06 14.38 0.95 62.72 19.04 5 FAU7_2 30.29 <1.71 558.76 12.67 <0.32 52.62 25.3 5 AuRM2_1 9.66 24.92 618.35 34.19 <0.47 29.9 28.4 5 AuRM2_2 10.25 34.26 424.62 28.48 <0.51 26.34 32.22 5 AuRM2_3 9.77 23.07 525.09 33.16 0.5 26.17 28.7 5 AuRM2_4 10.05 39.02 548.12 29.64 0.29 32.46 32.24 5 FAU7_3 32.95 <3.53 741.4 10.69 0.53 54.47 37.12 5 FAU7_4 34.42 <4.61 1222.93 12.81 0.78 81.25 61.87 5 NIST610_3 368.22 8243.32 393.3 613.41 305.19 489.25 12.54 5 NIST610_4 355.87 9227.26 390.22 577.57 277.39 469.72 16.02 6 NIST610_1 390.61 6627.91 8275.71 821.34 1602.74 307.94 79.13 6 NIST610_2 394.69 6586.06 5946.28 835.08 1468.34 308.15 66.19 6 FAU7_1 35.63 45.61 1606.84 19.6 0.47 60.76 69.73 6 FAU7_2 43.58 <0.52 1331.97 15.49 0.38 79.21 64.2 6 AuRM2_1 10.54 27.23 <633.35 34.85 0.42 27.66 32.78 6 AuRM2_2 9.42 29.18 <582.83 29.2 0.34 29.09 28.42 6 AuRM2_3 8.93 29.56 509.28 27.4 0.33 22.35 25.92 6 AuRM2_4 12.07 26.35 553.46 41.31 0.49 45.06 39.09 6 FAU7_3 36.55 0.91 528.45 16.82 0.143 87.99 21.54 6 FAU7_4 33.8 <0.80 589.13 4.58 <0.27 112.28 35.77 6 NIST610_3 447.03 6064 648.02 1151.79 615.12 651.73 22.49 6 NIST610_4 380.31 5049.38 473.15 1031.56 529.44 577.08 17.44  201 Run  Analysis #   Ni  Cu63         Cu65         Zn   As          Se   Rh        Pd  1 NIST610_1 370.75 437.21 348.97 248.77 114.12 17.28 1.04 2.16 1 NIST610_2 388.61 455.59 363.45 263.07 118.81 18.9 1.1 2.19 1 FAU7_1 29.05 92.07 86.38 42.26 8.26 8.5 0.033 47.58 1 FAU7_2 30.21 94.23 89.67 46.83 10.61 14.81 0.043 47.31 1 AuRM2_1 30.87 33.34 32.97 31.4 62.3 44.49 38.67 27.15 1 AuRM2_2 28.45 30.96 31.17 32.12 40.07 33.82 40.97 31.46 1 AuRM2_3 25.83 28.11 28.06 26.85 40.84 31.8 36.53 27.06 1 AuRM2_4 47.75 51.34 50.62 58.17 81.51 69.34 54.83 37.92 1 FAU7_3 49.09 153.66 154.35 100.35 11.09 29.03 0.043 69.4 1 FAU7_4 62.37 208.6 204.23 136.02 21.25 25.79 0.019 53.87 1 NIST610_3 799.44 935.86 755.9 734.69 264.12 51.28 1.98 3.17 1 NIST610_4 799.89 924.54 779.39 731.29 250.28 42.15 1.97 3.03 2 NIST610_1 418.69 463.13 368.22 335.81 115.86 18.28 1.11 1.86 2 NIST610_2 408.41 466.47 376.6 334.13 116.05 20.54 1.1 1.64 2 FAU7_1 31.9 99.42 94.27 56.98 2.28 8.84 0.0267 43.3 2 FAU7_2 30.01 88.57 84.75 44.5 12.75 23.78 0.095 49.68 2 AuRM2_1 30.15 31.37 33.89 30.72 46.97 35.68 40.4 29.52 2 AuRM2_2 28.4 31.77 29.66 32.02 46.85 38.58 39.24 29.27 2 AuRM2_3 28.61 32.12 31.34 32.04 51.13 43.73 36.03 25.81 2 AuRM2_4 29.67 31.23 31.83 30.94 44.53 33.45 42.86 32.3 2 FAU7_3 37.09 120.12 117.36 69.48 17.32 49.23 0.032 46.2 2 FAU7_4 35.83 123.48 122.86 77.62 17.05 46.49 0.06 44.05 2 NIST610_3 466.97 558.18 457.98 383.27 195.57 59.5 1.15 1.81 2 NIST610_4 478.98 578.04 487.01 387.7 199.14 66.44 1.13 1.81 3 NIST610_1 217779.13 248120.16 221884.89 167809.92 82460.11 20687.35 643 958.31 3 NIST610_2 467.08 501.79 428.19 394.21 245.58 38.37 1.249 2.23 3 FAU7_1 58.94 187.81 183.16 91.25 22.08 28.89 0.074 102.26 3 FAU7_2 64.5 213.26 215.47 120.44 32.61 32.42 0.117 94.6 3 AuRM2_1 25.87 28.07 27.45 27.76 42.78 31.55 35.26 26.85 3 AuRM2_2 36.13 38.92 40.84 38.75 55.69 50.63 48.72 33.65 3 AuRM2_3 29.47 31.63 31.42 31.4 44.23 35.08 38.48 28.84 3 AuRM2_4 28.49 31.16 31.25 30.96 49.77 39.26 40.14 29.27 3 FAU7_3 29.71 97.26 91.97 52.57 22.73 24.86 0.184 47.93 3 FAU7_4 41.63 140.14 131.25 85.23 19.84 32.56 0.028 57.22 3 NIST610_3 453.32 497.35 429.6 363.66 160.41 21.54 1.42 1.76 3 NIST610_4 454.05 489.31 435.59 353.39 155.83 19.93 1.37 1.64 4 NIST610_1 489.64 543.61 469.5 407.49 228.43 27.56 1.24 1.57 4 NIST610_2 479.88 537.67 455.53 387.01 225 30.1 1.24 1.59 4 FAU7_1 36.04 123.51 115.45 69.7 27.42 21.24 0.114 41.99 4 FAU7_2 36.34 117.87 116.42 70.49 51.12 27.43 0.188 47.81 4 AuRM2_1 28.09 29.81 30.6 28.03 53.51 37.39 40.56 30.1 4 AuRM2_2 30.91 34.51 33.12 37.53 40.08 37.97 38.81 28.29 4 AuRM2_3 29.8 33.33 31.7 32.52 45.89 33.72 35.99 26.75 4 AuRM2_4 28.3 29.49 31.36 29.69 49.41 43.17 46.04 33.17 4 FAU7_3 31.36 101.38 94.01 59.61 10.96 29.28 0.0135 40.07 4 FAU7_4 30.22 92.54 88.77 41.38 30.76 36.51 0.177 48.67  202 Run  Analysis # Ni    Cu63        Cu65        Zn   As    Se Rh Pd  4 NIST610_3 415.34 470.81 388.08 294.69 129.27 24.43 1.41 1.66 4 NIST610_4 440.21 497.12 403.41 310.35 155.76 28.1 1.63 1.46 5 NIST610_1 452.09 486.43 475.6 454.62 135.23 30.09 1.43 4.99 5 NIST610_2 452.46 505.13 474.68 479.07 139.66 21.3 1.46 4.67 5 FAU7_1 32.8 108.94 118.29 65.08 <3.76 24.69 0.061 49.77 5 FAU7_2 29.3 89.62 94.84 58.22 5.27 32.33 <0.033 51.15 5 AuRM2_1 25.81 28.57 29.79 19.88 32.2 35.28 45.53 36.25 5 AuRM2_2 34.91 37.43 34.3 52.01 102.69 41.25 34.08 23.27 5 AuRM2_3 26.49 27.37 31.13 24.85 29.21 30.84 48.35 37.37 5 AuRM2_4 33.46 38.83 31.94 44.56 195.46 51.69 33.65 23.7 5 FAU7_3 34.55 117.7 88.59 62.16 84.16 52.39 <0.118 54.86 5 FAU7_4 38.26 130.42 98.82 75.75 240.33 69.17 0.26 52.75 5 NIST610_3 490.43 558.26 382.48 563.71 1495.25 41.08 1.45 2.54 5 NIST610_4 475.59 556.48 369.68 576.2 2944.56 55.88 1.59 2.63 6 NIST610_1 424.87 557.9 427.88 450.94 219.61 107.12 1.05 6.02 6 NIST610_2 437.96 574.79 440.12 455.02 214.6 106.19 1.07 6 6 FAU7_1 33.79 111.59 106.87 63.14 13.41 23.31 0.042 55.73 6 FAU7_2 38.97 138.92 130.9 86.91 20.95 25.35 0.043 46.77 6 AuRM2_1 29.45 30.98 31.68 32.26 40.84 31.85 46.29 33.42 6 AuRM2_2 29.16 32.32 31.75 30.84 54.7 43.14 36.21 26.89 6 AuRM2_3 26.38 28.57 28.18 27.78 38.95 33.86 34.43 24.17 6 AuRM2_4 34.82 37.51 38.48 38.78 66.37 43.02 54.53 43.84 6 FAU7_3 42.39 141.34 134.3 77.24 31.84 18.12 0.053 64.86 6 FAU7_4 45.93 149.66 150.03 60.59 30.35 40.34 0.087 78.77 6 NIST610_3 624.8 747.31 610.14 515.84 325.58 58.35 2.04 5.21 6 NIST610_4 551.36 715.38 540.27 417.16 293.09 58.06 1.75 4.62  203 Run Analysis # Sn Sb Te Pt Hg Pb204 Pb206 Pb207 1 NIST610_1 330.66 236.16 91.95 2 0.57 102.33 239.23 251.44 1 NIST610_2 340.11 243.03 98.37 2.04 0.79 116.26 259.71 256.82 1 FAU7_1 24.57 <0.108 0.75 87.53 9.42 12.21 15.28 15.37 1 FAU7_2 26.44 0.188 0.75 84.77 12.61 12.09 13.89 12.82 1 AuRM2_1 30.83 13.02 17.14 28.12 22.59 29.29 32.24 32.16 1 AuRM2_2 28.93 10.48 9.66 32.59 31.86 28.59 27.49 27.5 1 AuRM2_3 26.07 9.75 10.24 27.64 25.01 29.55 25.11 25.13 1 AuRM2_4 48.14 21.52 20.97 41.89 31.48 26.44 59.15 57.73 1 FAU7_3 56.41 0.11 2.57 123.16 <**** <**** 47.64 42.15 1 FAU7_4 71.87 <0.15 3.19 102.21 <**** <**** 54.47 47.65 1 NIST610_3 703.72 624.01 210.37 3.64 <**** <**** 758.09 715.4 1 NIST610_4 718.52 657.04 206.01 3.8 <**** <**** 861.71 808.27 2 NIST610_1 416.91 333.54 116.69 2.21 <**** 266.56 434.19 431.62 2 NIST610_2 413.88 331.4 117.36 2.17 129.42 253.29 420.15 407.22 2 FAU7_1 33.52 <0.127 1.55 82.05 2122.48 106.86 29.48 27.11 2 FAU7_2 25.14 0.57 1.54 88.15 1041.99 90.59 22.92 22.23 2 AuRM2_1 27.22 9.79 11.35 30.36 <35.64 21.61 26.48 26.35 2 AuRM2_2 31.74 13.06 12.22 30.33 57.06 35.01 31.5 31.66 2 AuRM2_3 30.99 12.19 18.08 27.6 29.39 33.88 32.09 32.39 2 AuRM2_4 28.33 10.72 9.52 32.44 22.35 24.39 27.01 26.86 2 FAU7_3 41.17 0.65 1.65 79.15 182.36 115.61 22.15 23 2 FAU7_4 48.69 0.11 2.09 76.01 164.42 101.53 28.41 27.63 2 NIST610_3 439.1 330.57 126.88 2.25 <1.14 154.18 385.67 384.81 2 NIST610_4 434.37 327.06 122.86 2.32 <0.94 147.41 396.06 387.28 3 NIST610_1 218902.33 177848.55 68554.68 1273.62 1118.93 205414.81 216989.73 208894.67 3 NIST610_2 458.67 325.52 164.7 2.44 <1.09 116.05 384.23 396.31 3 FAU7_1 54.66 <0.099 2.1 187.24 <**** 254.82 43.44 41.55 3 FAU7_2 73.67 0.149 2.14 175.68 2568.69 232.27 53.84 50.93 3 AuRM2_1 25.34 9.15 9.28 26.65 16.42 22.97 23 22.97 3 AuRM2_2 38.46 17.13 20.14 37.96 34.18 38.21 46.98 47.26 3 AuRM2_3 31.89 11.67 13.14 28.41 44.2 44.87 29.28 28.86 3 AuRM2_4 27.04 10.74 10.75 31.55 11.5 15.06 27.85 28.19 3 FAU7_3 32.61 0.1 2.37 78.63 9.31 16.25 23.3 23.85 3 FAU7_4 51.27 0.8 3.33 98.37 27.14 41.49 35.3 33.61 3 NIST610_3 398.13 319.93 127.21 2.67 <1.63 283.18 417.96 386.95 3 NIST610_4 384.39 302.65 121.13 2.67 2.7 296.51 417.66 390.55 4 NIST610_1 458.87 362.1 147.01 2.29 3.12 175.79 395.95 397.3 4 NIST610_2 463.17 349.05 144.7 2.4 6 176.69 395.3 406.11 4 FAU7_1 40.16 <0.201 2.73 73.94 66.8 46.77 30.83 29.43 4 FAU7_2 36.65 <0.21 2.84 81.53 79.52 48.85 27.74 27.43 4 AuRM2_1 27.67 11.14 11.54 30.92 26.56 26.24 27.31 27.48 4 AuRM2_2 31.83 11.38 12.91 29.51 24.78 31.73 31.21 31.01 4 AuRM2_3 33.8 12.71 10.71 27.86 26.42 33.91 32.55 31.4 4 AuRM2_4 25.21 9.91 14.18 33.96 24.89 21.55 25.35 26.22 4 FAU7_3 41.99 <0.122 4.34 69.19 59.2 50.22 32.22 28.12 4 FAU7_4 27.65 <0.163 3.31 84.41 60.92 47.71 20.46 19.43  204 Run Analysis # Sn Sb Te Pt Hg Pb204 Pb206 Pb207 4 NIST610_3 432.78 309.48 155.77 2.49 63.5 180.33 425.64 399.72 4 NIST610_4 446.22 318.76 166.99 2.61 2.23 190.7 454.88 422.82 5 NIST610_1 387.85 254.95 66.65 2.67 <1.61 145.76 400.4 420.72 5 NIST610_2 413.39 283.61 76.58 2.87 <1.56 138.8 401.61 415.77 5 FAU7_1 33.31 <0.28 1.67 92.43 10.66 23.4 37.06 37.18 5 FAU7_2 28.65 0.77 3.27 92.75 13.1 19.9 21.04 20.2 5 AuRM2_1 24.53 8.97 7.24 34.59 24.86 22.45 21.85 21.72 5 AuRM2_2 41.09 17.57 27.07 25.85 27.04 42.75 63.29 64.13 5 AuRM2_3 23.96 9.51 8.45 35.73 26.28 25.95 22.83 22.31 5 AuRM2_4 39.83 13.96 19.82 25.95 24.62 32.66 41.92 44.6 5 FAU7_3 37.47 1.05 6.68 92.35 16.43 24.29 32.4 28.14 5 FAU7_4 56.67 1.75 11.06 85.51 18.35 21.64 40.3 37.23 5 NIST610_3 599.71 441.82 275.29 2.8 <1.60 114.66 505 519.85 5 NIST610_4 597.12 454.03 308.27 2.6 <1.49 108.43 544.39 547.86 6 NIST610_1 438.01 350.75 240.62 2.18 2.53 129.12 437.07 428.1 6 NIST610_2 444.81 356.43 224.05 2.19 2.05 127.22 446.15 429.27 6 FAU7_1 30.45 <0.110 1.07 102.06 84.53 59.19 22.52 20.78 6 FAU7_2 48.76 0.189 1.08 90.87 91.79 64.61 33.24 31.12 6 AuRM2_1 27.06 10.02 12.94 35.15 27.37 28.93 25.8 25.57 6 AuRM2_2 31.79 12.64 11.23 27.51 24.46 28.94 32.19 32.5 6 AuRM2_3 26.58 10.5 14.26 26.04 23.16 27.53 27 27.06 6 AuRM2_4 34.4 12.46 9.72 41.17 31.44 31.38 31.68 31.48 6 FAU7_3 48.2 <0.081 0.91 110.67 68.85 57.73 31.69 31.07 6 FAU7_4 46.74 0.116 0.94 130.96 90.08 65.44 26.58 25.62 6 NIST610_3 543.48 445.53 121.4 3.75 2.49 159.14 498.73 489.29 6 NIST610_4 532.02 417.61 98.74 3.27 30.63 133.76 439.77 416.58  205 Run  Analysis # Pb208     Bi                    U  1 NIST610_1 210.32 189.44 673.96 1 NIST610_2 222.91 199.74 775.6 1 FAU7_1 15.23 16.24 0.099 1 FAU7_2 13.5 13.07 0.101 1 AuRM2_1 32.33 11.42 0.35 1 AuRM2_2 27.49 8.96 0.1 1 AuRM2_3 25.44 8.44 0.16 1 AuRM2_4 59.28 20.8 <0.15 1 FAU7_3 44.74 48.23 <0.038 1 FAU7_4 50.39 53.39 <0.00 1 NIST610_3 691.84 593.96 1709.94 1 NIST610_4 731.23 638.31 1866.82 2 NIST610_1 370.24 333.59 856.76 2 NIST610_2 405.47 332.41 953.38 2 FAU7_1 28.9 31.84 <0.0127 2 FAU7_2 22.14 21.56 0.35 2 AuRM2_1 26.64 9.32 0.2 2 AuRM2_2 31.35 9.99 0.121 2 AuRM2_3 32.06 11.04 <0.11 2 AuRM2_4 27.07 8.96 1.06 2 FAU7_3 21.85 20.42 <**** 2 FAU7_4 27.13 28.73 <**** 2 NIST610_3 368.17 298.62 <**** 2 NIST610_4 344.95 304.72 <**** 3 NIST610_1 191738.27 166090.59 <**** 3 NIST610_2 539.57 353.24 <**** 3 FAU7_1 39.91 39.86 2.2 3 FAU7_2 48.91 53.65 1.12 3 AuRM2_1 23.12 7.58 0.55 3 AuRM2_2 46.28 16.79 0.07 3 AuRM2_3 28.99 9.54 0.9 3 AuRM2_4 28.11 9.57 0.08 3 FAU7_3 25.1 26.17 0.17 3 FAU7_4 36.04 36.05 <0.010 3 NIST610_3 527.5 520.95 451.9 3 NIST610_4 544.77 558.4 417.37 4 NIST610_1 538.35 539.17 <**** 4 NIST610_2 548.86 450.96 <**** 4 FAU7_1 28.61 31.98 231.79 4 FAU7_2 28.43 29.44 3.64 4 AuRM2_1 27.45 8.87 0.71 4 AuRM2_2 31.01 11.21 0.07 4 AuRM2_3 32.27 10.56 0.13 4 AuRM2_4 25.64 8.75 0.17 4 FAU7_3 30.44 30.91 <0.0201 4 FAU7_4 20.07 19.37 0.18  206 Run  Analysis # Pb208       Bi     U  4 NIST610_3 611.05 391.64 422.16 4 NIST610_4 631.63 477.29 409.55 5 NIST610_1 556.77 714.22 135.48 5 NIST610_2 563.01 668.93 158.82 5 FAU7_1 36.72 46.1 0.042 5 FAU7_2 21.47 22.06 <0.030 5 AuRM2_1 22.66 7.32 0.14 5 AuRM2_2 62.16 25.37 0.17 5 AuRM2_3 23.11 7.4 0.24 5 AuRM2_4 42.55 15.94 <**** 5 FAU7_3 29.05 32.43 <**** 5 FAU7_4 39.47 43.71 <**** 5 NIST610_3 653.3 753.93 <**** 5 NIST610_4 738.06 837.09 <**** 6 NIST610_1 490.51 379.32 <**** 6 NIST610_2 474.64 383.33 <**** 6 FAU7_1 21.54 23.15 <**** 6 FAU7_2 32.33 36.49 <**** 6 AuRM2_1 25.64 8.35 <**** 6 AuRM2_2 32.55 11.28 <1.54 6 AuRM2_3 27.11 9.14 0.11 6 AuRM2_4 31.45 10.4 0.32 6 FAU7_3 31.44 42.84 0.21 6 FAU7_4 26.2 25.42 0.32 6 NIST610_3 783.39 446.34 1900.74 6 NIST610_4 611.08 374.83 1622.81    


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items