UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mineral traps for greenhouse gases in mine tailings : a protocol for verifying and quantifying CO₂ sequestration… Wilson, Siobhan Alexandra 2009

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

Item Metadata

Download

Media
24-ubc_2010_spring_wilson_siobhan.pdf [ 56.61MB ]
Metadata
JSON: 24-1.0052719.json
JSON-LD: 24-1.0052719-ld.json
RDF/XML (Pretty): 24-1.0052719-rdf.xml
RDF/JSON: 24-1.0052719-rdf.json
Turtle: 24-1.0052719-turtle.txt
N-Triples: 24-1.0052719-rdf-ntriples.txt
Original Record: 24-1.0052719-source.json
Full Text
24-1.0052719-fulltext.txt
Citation
24-1.0052719.ris

Full Text

MINERAL TRAPS FOR GREENHOUSE GASES IN MINE TAILINGS: A PROTOCOL FOR VERIFYING AND QUANTIFYING CO2 SEQUESTRATION IN ULTRAMAFIC MINES by Siobhan Alexandra Wilson B.Sc. (Honours), McMaster University, 2003 M.Sc., The University of British Columbia, 2005 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Geological Sciences) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2009 © Siobhan Alexandra Wilson, 2009  Abstract  Mineralization of CO2 in ultramafic mine tailings can occur on a scale that is significant relative to the greenhouse gas emissions of a mine. Consequently, some active mining operations may be able to take advantage of carbon mineralization within their tailings to offset part of their greenhouse gas emissions. The secondary Mg-carbonate minerals that form in mine tailings are safe and durable traps for carbon and their presence can represent substantial disposal of atmospheric CO2. Hydrated Mg-carbonate minerals precipitate within mine tailings from the Diavik Diamond Mine, Northwest Territories, Canada, and the Mount Keith Nickel Mine, Western Australia, Australia. An improved understanding of the carbon cycle in mine tailings, and the contribution of mineralogical and geochemical strategies for assessing carbon mineralization in ultramafic mine tailings, are achieved by studying these sites. Quantitative mineralogical procedures, which use X-ray powder diffraction data, are developed for quantifying low abundances of mineral traps for CO2 within mine tailings. Quantitative mineralogical results are used to assess the amount of CO2 stored within hydrated Mg-carbonate minerals at both mine sites, and to assist in determining which gangue minerals are the primary sources for Mg in these minerals. Radiocarbon and stable isotopes of carbon and oxygen are used to identify the sources for carbon in secondary Mg-carbonate minerals. Isotopic analogue experiments are used to study the fractionation of stable carbon isotopes during precipitation of dypingite, a hydrated Mg-carbonate mineral, under conditions that simulate those in the tailings storage facilities at Mount Keith. The results of these experiments suggest that hydrated Mg-carbonate minerals may be precipitating out of isotopic equilibrium with the atmosphere. A carbon isotopic fractionation factor obtained for dypingite, and computational models for isotopic mixing scenarios, are used to interpret stable isotope and radiocarbon data for carbonate minerals. Although models for mixing scenarios can provide convincing fits to stable isotopic data, they are commonly inconsistent with field observations, trends in quantitative mineralogical data, and radiocarbon results. Ultimately, radiocarbon data are used to determine that most of the carbon trapped and  
  ii
  stored within hydrated Mg-carbonate minerals at Diavik and Mount Keith is sourced from the modern atmosphere.  
  iii
  Table of Contents Abstract ............................................................................................................................. ii Table of Contents............................................................................................................. iv List of Tables ..................................................................................................................... x List of Figures .................................................................................................................. xi List of Symbols and Abbreviations .............................................................................. xiv Acknowledgements ....................................................................................................... xvii Dedication....................................................................................................................... xix Co-authorship Statement ............................................................................................... xx Chapter 1: Introduction ................................................................................................... 1 1.1 Climate change and carbon capture and storage....................................................... 1 1.2 Carbon mineralization within mine tailings ............................................................. 4 1.3 Carbon mineralization in tailings from active mines................................................ 5 1.4 Organization and outcomes of this thesis ................................................................. 8 1.5 References .............................................................................................................. 11 
 Chapter 2: Quantifying carbon fixation in trace minerals from processed kimberlite: A comparative study of quantitative methods using X-ray powder diffraction data with applications to the Diavik Diamond Mine, Northwest Territories, Canada ........................................................................................................ 15 2.1 Introduction ............................................................................................................ 15 2.2 Locality and sampling strategy............................................................................... 18 2.3 Experimental method.............................................................................................. 22 2.3.1 Sample preparation and data collection........................................................... 22 2.3.2 Rietveld refinement and quantitative phase analysis....................................... 26 2.3.3 Quantitative analysis of nesquehonite using the RIR method and the internal standard method........................................................................................................ 30 2.4 Results and discussion ............................................................................................ 32 2.4.1 Synthetic mine tailings .................................................................................... 32 
  iv
  2.4.1.1 Rietveld refinement results....................................................................... 32 2.4.1.2 Results of quantitative RIR and the internal standard method ................. 43 2.4.2 Natural mine tailings ....................................................................................... 49 2.5 Implications for neutralization potential and carbon dioxide sequestration........... 54 2.6 Conclusions ............................................................................................................ 59 2.7 References .............................................................................................................. 60 Chapter 3: Carbon isotopic fractionation between dypingite, Mg5(CO3)4(OH)2·5H2O, and aqueous bicarbonate in an evaporative and highly saline system ...................... 70 3.1 Introduction ............................................................................................................ 70 3.2 Review of isotopic fractionation factors for carbonate minerals............................ 71 3.2.1 Equilibrium carbon isotopic fractionation factors for carbonate minerals ...... 73 3.2.1.1 Ca-carbonate minerals .............................................................................. 75 3.2.1.2 Ca-Mg-carbonate minerals ....................................................................... 76 3.2.1.3 Mg-carbonate minerals ............................................................................. 77 3.2.2 Equilibrium oxygen isotopic fractionation factors for carbonate minerals ..... 77 3.2.2.1 Ca-carbonate minerals .............................................................................. 77 3.2.2.2 Ca-Mg-carbonate minerals ....................................................................... 79 3.2.2.3 Mg-carbonate minerals ............................................................................. 80 3.2.3 Equilibrium isotopic fractionation factors for hydrous and basic carbonate minerals .................................................................................................................... 81 3.3 Methods .................................................................................................................. 84 3.3.1 Procedure for precipitating dypingite .............................................................. 84 3.3.2 Qualitative X-ray powder diffraction methods................................................ 86 3.3.3 CHN analyses .................................................................................................. 86 3.3.4 Total inorganic carbon measurements ............................................................. 87 3.3.5 Stable isotopic methods ................................................................................... 87 3.4 Analytical Results................................................................................................... 88 3.4.1 X-ray powder diffraction results...................................................................... 88 3.4.2 Results of CHN determination ........................................................................ 89 3.4.3 Data from monitoring pH, temperature, and mass loss ................................... 93  
  v
  3.4.4 Stable isotopic results ...................................................................................... 95 3.4.5 Results of total dissolved inorganic carbon measurements ............................. 97 3.5 Discussion............................................................................................................. 100 3.5.1 Mineralogy of final precipitates .................................................................... 100 3.5.2 Fractionation of stable isotopes during precipitation of dypingite ................ 102 3.5.2.1 Stable isotopic fractionation in DIC and water ...................................... 102 3.5.2.2 Stable isotopic fractionation of C and O in dypingite ............................ 105 3.6 Implications for CO2 sequestration in mine tailings............................................. 108 3.7 References ............................................................................................................ 111 
 Chapter 4: Carbon fixation in mineral waste from the Mount Keith Nickel Mine, Western Australia, Australia....................................................................................... 121 4.1 Introduction .......................................................................................................... 121 4.2 Locality and sampling strategy............................................................................. 123 4.2.1 The Mount Keith Nickel Mine ...................................................................... 123 4.2.2 Strategy for sampling at Mount Keith ........................................................... 126 4.3 Analytical methods ............................................................................................... 127 4.3.1 Qualitative X-ray powder diffraction methods.............................................. 127 4.3.2 Quantitative X-ray powder diffraction and Rietveld refinement................... 128 4.3.3 Scanning electron microscopy....................................................................... 132 4.3.4 Stable isotopic methods ................................................................................. 133 4.3.4.1 Standard methodology for carbonate minerals....................................... 133 4.3.4.2 Standard procedure for dissolved inorganic carbon ............................... 134 4.3.4.3 Procedures for selective analysis of carbonate minerals ........................ 135 4.3.5 Radiocarbon procedures ................................................................................ 138 4.4 Field results and qualitative mineralogy............................................................... 139 4.4.1 Qualitative mineralogy of Mount Keith mine tailings................................... 139 4.4.2 Timing and depth of tailings deposition ........................................................ 145 4.5 Analytical results .................................................................................................. 148 4.5.1 Rietveld refinement results ............................................................................ 148 4.5.2 Stable isotopic results .................................................................................... 157  
  vi
  4.5.3 Radiocarbon results ....................................................................................... 158 4.6 Discussion............................................................................................................. 160 4.6.1 Abundance and distribution of minerals in Mount Keith mine tailings ........ 160 4.6.1.1 Sulphate minerals ................................................................................... 160 4.6.1.2 Halite and hydromagnesite ..................................................................... 161 4.6.1.3 Primary, bedrock carbonate minerals ..................................................... 162 4.6.1.4 Serpentine, hydrotalcite-group minerals, and brucite............................. 163
 4.6.2 Summary of spatial and temporal changes in mineralogy............................. 165 4.6.3 Rate and scale of hydromagnesite precipitation ............................................ 165 4.7 Carbon reservoir fingerprinting ............................................................................ 174 4.7.1 Fingerprinting with stable isotopes................................................................ 174 4.7.2 Improved discrimination with radiocarbon ................................................... 177 4.8 Conclusions .......................................................................................................... 181 4.9 References ............................................................................................................ 183 
 Chapter 5: Isotopic fingerprinting of mineralized carbon in ultramafic mine tailings............................................................................................................................ 197 5.1 Introduction .......................................................................................................... 197 5.2 Secondary carbonate minerals at the Diavik Diamond Mine ............................... 200 5.3 Analytical methods ............................................................................................... 205 5.3.1 Qualitative X-ray powder diffraction methods.............................................. 205 5.3.2 Stable isotopic methods ................................................................................. 205 5.3.3 Radiocarbon procedures ................................................................................ 207
 5.4 Results of isotopic analyses for the Diavik Diamond Mine ................................. 207 5.4.1 Stable carbon and oxygen isotopic results..................................................... 207 5.4.2 Radiocarbon results ....................................................................................... 208 5.5 Carbon fingerprinting at Diavik ........................................................................... 214 5.5.1 Stable isotopic results .................................................................................... 214 5.5.1.1 Secondary Na and Ca-carbonate minerals.............................................. 214 5.5.1.2 Primary bedrock carbonate minerals ...................................................... 215 5.5.1.3 Secondary nesquehonite, MgCO3·3H2O................................................. 216  
  vii
  5.5.2 Radiocarbon results ....................................................................................... 217
 5.6 Modelling mixing and recycling of carbonate minerals in mine tailings ............ 218 5.6.1 Reservoir values and isotopic fractionation factors used in models.............. 218 5.6.2 Scenario 1: Mechanical mixing between two reservoirs ............................... 221 5.6.3 Scenario 2: Single-event dissolution and reprecipitation of carbonate minerals .................................................................................................................. 222
 5.6.4 Scenario 3: Cyclical dissolution and reprecipitation of carbonate minerals.. 224 5.6.5 Model results ................................................................................................. 226 5.6.5.1 Results for scenario 1: Mechanical mixing ............................................ 226 5.6.5.2 Results for scenario 2: Single-event dissolution and reprecipitation ..... 227 5.6.5.3 Results for scenario 3: Cyclical dissolution and reprecipitation ............ 228 5.7 Application of mixing scenarios to Diavik .......................................................... 233 5.7.1 Application of scenario 1 to Diavik............................................................... 233 5.7.2 Application of scenario 2 to Diavik............................................................... 234 5.7.3 Application of scenario 3 to Diavik............................................................... 235 5.7.4 Implications of modelled results for CO2 sequestration at Diavik ................ 239 5.8 Carbon fingerprinting in ultramafic mine tailings ............................................... 240 5.8.1 The Mount Keith Nickel Mine ...................................................................... 240 5.8.2 Active and historical mines ........................................................................... 246 5.9 Summary and Conclusions ................................................................................... 250 5.10 References .......................................................................................................... 254 Chapter 6: Conclusions ................................................................................................ 262 6.1 Summary of research outcomes............................................................................ 262 6.2 Suggestions for future research ............................................................................ 265 6.3 References ............................................................................................................ 267 
 Appendices .................................................................................................................... 268 A1 Appendix to Chapter 4.......................................................................................... 268 A2 Appendix to Chapter 5.......................................................................................... 300 A2.1 mixing11.m .................................................................................................... 301  
  viii
  A2.2 mixing12.m .................................................................................................... 304 A2.3 F14C_xValues.m ........................................................................................... 307 A2.4 react2f.m ........................................................................................................ 308 A2.5 react3f.m ........................................................................................................ 310 A2.6 react4f.m ........................................................................................................ 312 A2.7 react5f.m ........................................................................................................ 314  
  ix
  List of Tables Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5 Table 2.6 Table 2.7 Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 4.1 Table 4.2 Table 4.3 Table 4.4 Table 4.5 Table 5.1 Table 5.2 Table 5.3 Table 5.4  Compositions of synthetic processed kimberlite renormalized to exclude corundum and taking account of calcite contamination in lizardite ......... 23 Sources of crystal structure data for Rietveld refinement and values used to compute neutralization potential (NP) of processed kimberlite ........... 27 Crystallinity of internal standard phases and estimates of crystallinity from serrated samples suffering from surface roughness effects ...................... 34 Rietveld refinement results for serrated specimens of synthetic processed kimberlite .................................................................................................. 36 Rietveld refinement results for non-serrated specimens of synthetic processed kimberlite ................................................................................. 37 Results of Reference Intensity Ratio measurements on synthetic processed kimberlite .................................................................................................. 44 Results of quantitative phase analysis of natural samples of processed kimberlite, renormalized to exclude corundum ........................................ 51 Starting conditions for experiments dyp-may14-09-1 and dyp-may14-09-2 ................................................................................................................... 85 Results of dypingite precipitation experiment dyp-may14-09-1 .............. 90 Results of dypingite precipitation experiment dyp-may14-09-2 .............. 91 CHN data and H-content of dypingite compared to results of previous studies ....................................................................................................... 93 Sources of crystal structure data for Rietveld refinement ...................... 129 Numbers of samples for which hydromagnesite is quantifiable, imputed (at detection), or below detection............................................................ 132 Carbonate and hydrotalcite-group minerals detected at Mount Keith .... 140 Sulphate and halide minerals detected at Mount Keith........................... 141 Annual rate of CO2 mineralization and cumulative trapping at Mount Keith........................................................................................................ 172 Carbonate mineral phases detected at Diavik ......................................... 202 Sulphate mineral phases detected at Diavik............................................ 204 Stable carbon and oxygen isotopic data and radiocarbon data for carbonate minerals and DIC from Diavik................................................................ 209 Values input to models for scenarios 1 through 3................................... 219  Appendix Table A1.1 Appendix Table A1.2  
  Rietveld refinement results for samples from Mount Keith ........................................................................................ 269 Stable and radiogenic isotope data for carbonate minerals and DIC from Mount Keith ............................................ 295  x
  List of Figures Figure 1.1 Figure 1.2 Figure 2.1 Figure 2.2 Figure 2.3 Figure 2.4 Figure 2.5 Figure 2.6 Figure 2.7 Figure 2.8 Figure 2.9 Figure 2.10 Figure 2.11 Figure 2.12 Figure 3.1 Figure 3.2 Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 3.7 Figure 3.8 Figure 3.9 Figure 3.10 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 4.5  
  Location of Mount Keith Nickel Mine, Western Australia, Australia ....... 6 Location of Diavik Diamond Mine, Northwest Territories, Canada .......... 7 Location of Diavik Diamond Mine, Northwest Territories, Canada ........ 19 Modes in which nesquehonite and other efflorescent minerals have been identified in the processed kimberlite at Diavik ....................................... 20 Sampling locations within the storage facility for processed kimberlite fines at Diavik (fine PKC) ........................................................................ 21 Backscattered SEM images of (a) 05DVK7 and (b) 07lsk3r.................... 30 Results of Rietveld refinements on serrated specimens of synthetic processed kimberlite ................................................................................. 38 Results of Rietveld refinements on non-serrated specimens of synthetic processed kimberlite ................................................................................. 39 Relative error on Rietveld refinement results for each mineral phase in the synthetic processed kimberlites ................................................................ 40 Modeling phlogopite in a synthetic kimberlite (07lsk3r) ......................... 42 RIR and Rietveld refinement results for nesquehonite............................. 45 Calibration curve for nesquehonite ........................................................... 48 XRPD patterns collected for 06DVK53-1................................................ 54 Results for neutralization potential (NP) and maximum potential acidity (MPA) of processed kimberlite................................................................. 56 Equilibrium carbon isotopic fractionation between carbonate minerals and gaseous CO2 compiled from experimental and theoretical studies........... 74 Equilibrium oxygen isotopic fractionation between carbonate minerals and gaseous CO2 compiled from experimental and theoretical studies........... 78 Equilibrium carbon isotopic fractionation between hydrous and basic carbonate minerals and gaseous CO2 ........................................................ 82 Equilibrium oxygen isotopic fractionation between hydrous and basic carbonate minerals and water.................................................................... 83 X-ray powder diffraction patterns of synthetic and natural dypingite...... 92 pH and temperature (˚C) of solutions over time....................................... 94 Water loss from precipitation experiments over time............................... 95 Stable carbon and oxygen isotopic data for synthetic dypingite, dissolved inorganic carbon (DIC), and water ........................................................... 96 Stable carbon isotopic data and concentrations of dissolved inorganic carbon (DIC) over time ............................................................................. 98 Stable oxygen isotopic data and concentrations of dissolved inorganic carbon (DIC) versus mass loss of solution................................................ 99 Location of Mount Keith Nickel Mine, Western Australia, Australia ... 124 The Tailings Storage Facilities (TSFs) at Mount Keith ......................... 125 Backscattered electron images of hydromagnesite in Mount Keith mine tailings..................................................................................................... 142 Mount Keith mine tailings at the surface and at depth within TSF2 ...... 144 The W1 riser in TSF2 ............................................................................. 146 xi
  Figure 4.6  Variation of hydromagnesite abundance with depth beneath the surface of TSF1 and TSF2 over time....................................................................... 150 Figure 4.7 Variation of Na-Mg-sulphate mineral abundances with depth beneath the surface of TSF1 and TSF2 over time ...................................................... 151 Figure 4.8 Variation of Mg-sulphate and Ca-sulphate mineral abundances with depth beneath the surface of TSF1 and TSF2 over time................................... 152 Figure 4.9 Variation of hydromagnesite and halite abundances with depth beneath the surface of TSF1 and TSF2 over time ...................................................... 153 Figure 4.10 Variation of abundance for primary (bedrock) carbonate minerals with depth beneath the surface of TSF1 and TSF2 over time......................... 154 Figure 4.11 Variation of the abundance of select gangue minerals (serpentine-group, hydrotalcite-group, and brucite) with depth beneath the surface of TSF1 and TSF2 over time................................................................................. 155 Figure 4.12 Variation of select mineral abundances with depth for 10-year old tailings from TSF1 ............................................................................................... 156 Figure 4.13 Stable oxygen and carbon isotope data for different modes of occurrence and mineralogy of carbonate minerals at Mount Keith........................... 157 Figure 4.14: Stable carbon (δ13C) and fraction modern carbon (F14C) data for secondary precipitates of hydromagnesite, bedrock carbonate minerals, and soda ash (a process additive) from Mount Keith.............................. 159 Figure 4.15 Development of hydromagnesite (as wt.% abundance) at depth within Mount Keith tailings, from zero to 10 years prior to sampling............... 168 Figure 4.16 Development of hydromagnesite (as wt.% abundance) near the surface of Mount Keith tailings, from zero to 10 years prior to sampling............... 171 Figure 4.17 Estimated cumulative and annual mineralization of CO2 within Mount Keith mine tailings .................................................................................. 173 Figure 4.18 Estimated cumulative and annual mineralization of atmospheric CO2 within Mount Keith mine tailings ........................................................... 180 Figure 5.1 Locations of four mines .......................................................................... 200 Figure 5.2 Backscattered electron SEM images of efflorescences from the surface of the tailings at Diavik ............................................................................... 203 Figure 5.3 Stable oxygen and carbon isotope data by mode of occurrence and mineralogy for Diavik............................................................................. 212 Figure 5.4 Variation of δ13C with F14C for samples from Diavik ............................ 213 Figure 5.5 Models for scenario 1, mechanical mixing of carbonate minerals from two distinct reservoirs .................................................................................... 230 Figure 5.6 Models for scenario 2, batch dissolution and reprecipitation of carbonate minerals with dependence upon the water-rock ratio ............................. 231 Figure 5.7 Models for scenario 3, cyclic dissolution and reprecipitation of carbonate minerals................................................................................................... 232 Figure 5.8 Stable isotope and radiocarbon data from Diavik overlain by modelled predictions for scenario 1 (mechanical mixing)...................................... 236 Figure 5.9 Stable isotope and radiocarbon data from Diavik overlain by modelled predictions for scenario 2 (batch dissolution/reprecipitation)................. 237 Figure 5.10 Stable isotope and radiocarbon data from Diavik overlain by modelled predictions for scenario 3 (cyclic dissolution/reprecipitation)................ 238  
  xii
  Figure 5.11 Figure 5.12 Figure 5.13 Figure 5.14 Figure 5.15  
  Stable isotope and radiocarbon data from Mount Keith overlain by modelled predictions for scenario 1 (mechanical mixing)...................... 243 Stable isotope and radiocarbon data from Mount Keith overlain by modelled predictions for scenario 2 (batch dissolution/reprecipitation)....... ................................................................................................................. 244 Stable isotope and radiocarbon data from Mount Keith overlain by modelled predictions for scenario 3 (cyclic dissolution/reprecipitation)...... ................................................................................................................. 245 Stable oxygen and carbon isotope data from samples for which radiocarbon data are available................................................................. 247 Stable carbon isotope data and radiocarbon data from four ultramafichosted mines............................................................................................ 248  xiii
  List of Symbols and Abbreviations A – area An91 – Anorthite-91 apfu – atoms per formula unit BHPB – BHP Billiton CHN – Carbon-Hydrogen-Nitrogen C – elemental concentration d – Durbin-Watson statistic d – distance between lattice planes D – depth DDMI – Diavik Diamond Mines Inc. DIC – Dissolved Inorganic Carbon DOC – Dissolved Organic Carbon EMP – Electron Microprobe f – fraction of water remaining in an evaporative system fw – fraction of water in a system fr – fraction of rock in a system F14C – Fraction Modern Carbon Fo90 – Forserite-90 FP – Fundamental Parameters h – relative humidity ICDD – International Centre for Diffraction Data ICME – International Council on Metals and the Environment IPCC – Intergovernmental Panel on Climate Change IRMS – Isotope Ratio Mass Spectrometer m or M – mass MAD – Median Absolute Deviation MPA – Maximum Potential Acidity n – molar concentration of an element NAD – North American Datum 
  xiv
  NBS – National Bureau of Standards NP – Neutralization Potential p – fraction of carbon from a isotopic reservoir pfu – per formula unit PCIGR – Pacific Centre for Isotopic and Geochemical Research PDF-4+ – Powder Diffraction File, version 4+ PGE – Platinum Group Element PKC – Processed Kimberlite Containment facility POM – Particulate Organic Matter PV – Pseudo-Voigt Q – water-rock ratio RIR – Reference Intensity Ratio(s) Rwp – weighted pattern statistic SEM – Scanning Electron Microscope SSAMS – Single Stage Accelerator Mass Spectrometer t – time T – temperature TN – Total Nitrogen TOC – Total Organic Carbon TSF – Tailings Storage Facility UNEP – United Nations Environmental Program VPDB – Vienna Pee Dee Belemnite VSMOW – Vienna Standard Mean Ocean Water wt.% – weight percent WGS – World Geodetic System X – weight percent abundance of a mineral XRD – X-Ray Diffraction XRPD – X-Ray Powder Diffraction αx-y – isotopic fractionation factor between substances x and y δ13C – delta carbon-13  
  xv
  δ18O – delta oxygen-18 Δ13Cx-y – approximately 103lnαx-y for C Δ14C – deviation in 14C composition Δ18Ox-y – approximately 103lnαx-y for O  εx-y – isotopic separation between substances x and y θ – scattering angle of X-rays  ρ – density σ – measurement uncertainty  ϕ – fraction of an element held by water χ2 – chi squared statistic  
  xvi
  Acknowledgements This thesis is part of a broader study of CO2 sequestration within mine tailings, conducted by the Mineral Deposit Research Unit in the Department of Earth and Ocean Sciences, The University of British Columbia, and the Geomicrobiology Group in the Department of Earth Sciences, The University of Western Ontario, London, ON, Canada. This project received funding from the Natural Sciences and Engineering Research Council of Canada (NSERC), Diavik Diamond Mines Inc. (DDMI), and BHP Billiton (BHPB) through a Collaborative Research and Development Grant to Profs. Gregory M. Dipple and Gordon Southam. Further support for this project was provided by a NSERC Discovery Grant to Prof. Mati Raudsepp. I have been the extremely fortunate recipient of a University Graduate Fellowship from The University of British Columbia, a Foundation Scholarship and a Student Travel/Research Grant from the Mineralogical Association of Canada (MAC), a Grant from the Edward H. Kraus Crystallographic Research Fund from the Mineralogical Society of America (MSA), and an Alexander Graham Bell Canada Graduate Scholarship from NSERC. The generosity of the support that I have received from MAC, MSA, NSERC, and UBC in the past four years still staggers me. Thank you so much for providing me with the opportunity to focus exclusively on my research and for honouring me by showing such confidence in my abilities and my work. I am immensely grateful. Although I began my doctoral studies in January 2006, I arrived at UBC in the autumn of 2003 to pursue a master’s degree. Having been a part of the Department of Earth and Ocean Sciences at UBC for six years, I would like to thank everyone here for their support, assistance, and kindness. First and foremost amongst these individuals are my research supervisors, Greg Dipple and Mati Raudsepp. I still have difficulty imagining what convinced them to take on a master’s student who couldn’t tell the difference between serpentine and cedar chips. Moreover, it completely escapes me what convinced them to keep her on for a Ph.D. Whatever the reason, as someone who (at least in her own mind) has since mastered the game of “Wood or Fibre?”, I am very thankful for their faith in me. Greg, Mati, I cannot fathom ever finding a way to thank you sufficiently for challenging me to expand the scope of my imagination and for giving me every opportunity to hone my abilities and to indulge my curiosity. Thank you, with every fibre (and all the cedar chips) of my being, for your mentorship, your time, and for challenging me to work harder than I ever have in my life. Greg, you have my undying appreciation for finding the time to help me write my thesis while learning how to manage an academic department! Most importantly of all, thanks are due to Greg for teaching me to appreciate beer and fieldwork and to Mati for his stalwart (albeit unsuccessful) efforts to improve my taste in both coffee and hats. Uli Mayer has been gracious enough to sit on my advisory committee through two degrees. Thank you, Uli, for your time and for providing me with so many excellent papers to read and new ideas to consider. Thank you also for reminding me that minerals do not exist in a vacuum and that water is lurking everywhere. I am indebted to Gordon Southam, at The University of Western Ontario, for sharing his advice and expertise in the course of fieldwork. Thank you, Gordon, for teaching me to always consider the influence of life on geological systems and to remember that minerals and microbes are 
  xvii
  often intimately connected. I am also grateful to Jim Mortensen and Al Lewis for taking the time to help me through my comprehensive exam. Thanks to John Kaszuba at The University of Wyoming and Marc Bustin and Greg Lawrence at UBC for participating in my final oral examination and for asking such thoughtful and challenging questions. Thanks are due to Ben Grguric, Josh Levett, John Tomich, Shona, Red, and all the staff at Mount Keith for their generous and knowledgeable support in the field. Thanks also to Colleen English and her staff at Diavik for giving so freely of their time and expertise. Ian Power and I have been working together on CO2 sequestration in mine tailings since 2004 and have engaged in many field-based adventures and mishaps. It has always been a pleasure and a phenomenal learning experience to work with you, Ian, and I hope to continue doing so for a long time to come. You shall always have my particular thanks for putting up with my annoyingly long, contemplative silences in the field, for allowing me to share in the glow of inventing Golden Bear Coffee at Diavik and that absurd desert at Atlin (the one that required a rock hammer and a bottle of Bailey’s…), and for pulling me out that time I fell thigh-deep into mine tailings at Mount Keith. Abraham, Claire Bear, J.T., Jeremy, Joanne, Julio, Lizzie, Lyle, Nina, Shaun, and Stu: Thank you for being such excellent labmates! It’s been a pleasure working with you and learning from you. Thank you also to all the undergraduate students (past and present) who have assisted me in the lab: Claire Brown, Moira Cruickshanks, Colin Finkbeiner, Vicky Liu, Shelley Oliver, Mandy Tang, and Joanne Woodhouse. Jenny Lai, Edith Czech, and Elisabetta Pani are thanked for their assistance over the years and for being patient with me on the many occasions when I forgot to return equipment to their lab. Thanks to Gen. Stuart Mills for taking the time to pass on some of his ridiculously extensive knowledge of systematic mineralogy. I am grateful to Shaun Barker for giving so freely of his time, for his support in the laboratory, and for reading parts of this thesis in its earliest incarnations. Shaun, I shall always be grateful to you for upholding the unspoken Fancy Coffee Pact and for saving Jesse and me from Mt. St. Helens. I would also like to thank Johanna and Ellen Dipple for the superb artwork on my office door and for the Avenging Narwhal action figure (this is by far the best desk toy I have ever had). I could never have finished this thesis without excellent company and daily trips to Tim Horton’s and other establishments. In this regard, I am particularly grateful to Gareth Chalmers (above all), Shaun Barker, Stuart Sutherland, James Thom, Lyle Hansen, Stephen “Sphagnum” Moss, Shawn “No.2” Hood, Ayesha Ahmed, Shelley Oliver, Curtis Brett, Daniel Ross, Krista Michol, Nils Peterson, Jackie Dohaney, Swati Singh, Steve Johnston, Holly Peterson, R.-E. Farrell, Geneviève Robert, John Dockrey, Gen. Stuart Mills, Mackenzie Parker, and everyone else with whom I have ever enjoyed coffee (and other beverages) at UBC. My gratitude goes to good friends in distant places: Cathy Lovekin, Di Moscu, Steve Johnston, Jeff Hargot, and Swati Singh – I hope to visit all of you soon. A heartfelt thanks goes to everyone at the False Creek Ku Yu Kai Gojuryu Karate dojo, and to Julie Zilber Sensei in particular. It has been a delight and an honour to train with you these past five years. Finally, the love and encouragement that I have received from my parents, Irina and David, my brother, Min, and my partner, Jesse, have sustained me through my Ph.D., and my M.Sc. before that. I could never have done this without your support and your patience. Thank you for everything.  
  xviii
  This work is dedicated to my parents, David and Ira, to Min, my favourite brother, and to Jesse.  
  xix
  Co-authorship Statement This thesis consists of four related manuscripts that have been prepared for publication in peer-reviewed international journals. A version of Chapter 2 has been published in Applied Geochemistry. I am the lead author of this manuscript and Mati Raudsepp and Gregory M. Dipple are my coauthors. As the primary author, I designed and conducted all experiments, performed all analyses, and was responsible for the bulk of the text, tables, and figures. M. Raudsepp provided technical advice and suggestions. G.M. Dipple, who initiated the project, provided oversight on applications and implications of the results of this research. All authors contributed to revision of the manuscript. The reference for the manuscript follows: Wilson, S.A., Raudsepp, M., and Dipple, G.M. (2009) Quantifying carbon fixation in trace minerals from processed kimberlite: A comparative study of quantitative methods using X-ray powder diffraction data with applications to the Diavik Diamond Mine, Northwest Territories, Canada. Applied Geochemistry, 24, 23122331. doi:10.1016/j.apgeochem.2009.09.018. Chapter 3 is intended for submission to a peer-reviewed international journal. I am the lead author and Shaun L.L. Barker, Gregory, M. Dipple, Viorel Atudorei, and James M. Thom are my co-authors. As the first author, I helped to design the experiments and performed most of the sampling. I am responsible for the bulk of the interpretation, for the computational modelling, and for most of the text, figures, and tables. S.L.L. Barker assisted with experimental design, sampling, and interpretation of data. G.M. Dipple provided advice regarding measurement of stable isotopic fractionation factors and assisted with the interpretation of data. V. Atudorei performed all stable isotopic analyses. J.M. Thom conceived of the method by which we precipitated dypingite, by evaporation from a saline brine. S.L.L. Barker, G.M. Dipple, and I have contributed to revision of the manuscript. 
  xx
  Chapter 4 is intended for submission to a peer-reviewed international journal. I am the primary author and Gregory M. Dipple, Shaun L.L. Barker, Ian M. Power, Stewart J. Fallon, Mati Raudsepp, and Gordon Southam are my co-authors. I conducted most of the research and produced most of the text, figures, and tables. I am responsible for much of the interpretation. I received significant assistance with interpretation of results from G.M. Dipple, and further assistance from S.L.L. Barker, I.M. Power, and G. Southam. I.M. Power and I performed most of the fieldwork and were assisted by G. Southam and G.M. Dipple. S.J. Fallon contributed radiocarbon analyses and M. Raudsepp provided advice pertaining to quantitative mineralogical techniques. G.M. Dipple and G. Southam initiated the project. G.M. Dipple, M. Raudsepp, and I have participated in revision of the manuscript. Chapter 5 is intended for submission to a peer-reviewed international journal. I am the lead author and Gregory M. Dipple, Ian M. Power, Shaun L.L. Barker, Stewart J. Fallon, and Gordon Southam are my co-authors. I am responsible for the bulk of the text, figures, and tables. G.M. Dipple and I developed the geochemical models and I implemented them in MATLAB. G.M. Dipple, I.M. Power, S.L.L. Barker, and G. Southam assisted me with the interpretation of isotopic data and model results. G.M. Dipple and G. Southam initiated the project. I.M. Power and I performed most of the fieldwork and were assisted by G. Southam and G.M. Dipple. S.J. Fallon contributed radiocarbon analyses. In addition, J.R. Gare contributed a function to the code for scenario 3 and helped to make some of the scripts more succinct. G.M. Dipple, S.L.L. Barker, and I have contributed to revision of the manuscript.  
  xxi
  Chapter 1 Introduction  1.1 Climate change and carbon capture and storage Current warming of the Earth’s climate has been connected to the emission of anthropogenic greenhouse gases (i.e., CO2, CH4, N2O) into the atmosphere (IPCC 2007). The overwhelming majority of climate scientists support this interpretation (Oreskes 2004) and the national science academies of the G8+5 nations have issued several declarations affirming that current climate change is almost certainly an anthropogenic effect. In 2004, carbon dioxide (CO2) constituted 77% by mass of total anthropogenic greenhouse gas emissions (IPCC 2007). The global atmospheric concentration of CO2 increased exponentially from a pre-industrial (i.e., pre-1750) level of 280 ± 20 ppm to 379 ppm by 2005. Approximately two-thirds of this increase has been attributed to the combustion of fossil fuels and one-third to changes in land use since 1750 (IPCC 2007). Coincident with the increasing concentration of atmospheric greenhouse gases, the average surface temperature of the Earth increased by 0.74º ± 0.18ºC between 1906 and 2005, and sea level rose by 0.17 ± 0.05 m over the course of the 20th century (IPCC 2007). The increased concentration of atmospheric CO2 has also been implicated in a decrease of 0.1 pH units in global average surface ocean water since the pre-industrial era (IPCC 2007). Pacala and Socolow (2004) suggest that less than double the pre-industrial atmospheric CO2 concentration (i.e., 500 ± 50 ppm) should be treated as an upper limit in order to prevent the most damaging and irreversible effects of anthropogenic climate change. Strategies for decarbonising energy sources, increasing efficiency of energy production, and trapping and storing CO2 are required to stabilize concentrations of atmospheric CO2 within the next century (e.g., Hoffert et al. 2002; Lackner 2003; Pacala and Socolow 2004; Broecker 2007). It will take time to increase the efficiency of, and ultimately to replace, the extant fossil-fuel based energy infrastructure that is in use 1  today. In the meantime, geologically stable, energy efficient, and cost-effective strategies for sequestering CO2 must be developed and implemented. The International Energy Agency predicts that nearly 20% of total reductions in greenhouse gas emissions for the next 40 years will need to be achieved through capture and storage of carbon (International Energy Agency 2008). Anthropogenic CO2 could be stored within living biomass or sequestered by injection into deep ocean water, seabed sediments, and subterranean geological formations such as saline aquifers and reservoirs for oil and gas (IPCC 2005). Sequestration of CO2 within biomass has a short characteristic storage time, on the order of 1-100 years, and a relatively limited capacity compared to other technologies (Lackner 2003). Although the ocean’s capacity to store CO2 is significantly larger than that of the biosphere, concerns about leakage and accelerated ocean acidification have curbed development of this option in recent years (Lackner 2003; Sipilä et al. 2008). The large-scale CO2 sequestration projects at Weyburn, Saskatchewan, Canada, and Sleipner, in the North Sea, have demonstrated successful storage of CO2 within subterranean geological reservoirs. At the Weyburn oil field, industrially produced waste CO2 is injected into a reservoir of shallow marine carbonate rocks. Approximately 5000 t/day of CO2 have been injected into the Midale Beds at Weyburn since 2000 (Cantucci et al. 2009). At Sleipner, supercritical CO2 (stripped from mined natural gas) has been injected at a rate of approximately 1 Mt/year since 1996 into the Utsira Sand, which is a ~200 m thick saline aquifer (Bickle 2007). Although storage of supercritical CO2 within underground wells and saline aquifers has been demonstrated to be stable on the decadal timescale, long-term stability of this method is not assured. Significant uncertainty remains regarding seismic instability of trapping and the potential for migration and leakage of buoyant CO2 (Lackner 2003). Consequently, wide-scale use of injection-based technologies could require continuous monitoring at storage sites for thousands of years (e.g., Sipilä et al. 2008). Because this is an extremely difficult, if not impossible, commitment for a society to make, more permanent means of storage must be pursued. Approximately 90% of carbon on Earth is fixed within carbonate minerals (Sundquist 1985; Sudquist 1993) and it is expected that these minerals will be the ultimate sink for most anthropogenic CO2 on a timescale of one million years (Kump et  2  al. 2000). Storage of CO2 in carbonate minerals is recognized as a safe and effective method for the sequestration of anthropogenic carbon (Seifritz 1990; Lackner et al. 1995; Lackner 2003). Furthermore, mineralization of anthropogenic CO2 within carbonate minerals has the largest capacity and longest storage time of any method yet proposed for capturing and storing carbon (Lackner 2003). Since carbon mineralization was first proposed as a method for storing CO2 (Seifritz 1990), most of the work on this subject has focussed on the development of rapid, large-scale methods for trapping and storing CO2 at industrial point sources (reviewed in Huijgen and Comans 2003, 2005; IPCC 2005; Sipilä et al. 2008). Most processes developed to date are based on reaction of naturally occurring silicate minerals such as forsterite [Mg2SiO4], serpentine-group minerals [Mg3Si2O5(OH)4], and wollastonite [CaSiO3] to carbonate minerals like magnesite [MgCO3] and calcite [CaCO3]. In nature, weathering of silicate minerals by dissolution leads to precipitation of carbonate minerals under conditions of atmospheric pressure and temperature. This is one of the most significant mechanisms for geochemical exchange of CO2 between the atmosphere and lithosphere (Schwartzman and Volk 1989; Berner 1990). However, this is typically a slow process and high-temperature pre-treatment of silicate minerals, or extraction of Mg and Ca from these minerals, is often required to accelerate reaction in the laboratory. Furthermore, high temperatures and pressures are needed to induce carbon mineralization reactions on the short timescales (i.e., hours) required for development and deployment of industrial carbonation reactors (Sipilä et al. 2008). As a result, this approach to carbon mineralization remains both costly and energy intensive (Lackner 2003; Sipilä et al. 2008). Recently, there has been a shift toward the development of low-temperature, lowpressure procedures for carbon mineralization that emulate or piggyback upon natural processes that occur at the surface of the Earth. Schuiling and Krijgsman (2006) suggest that olivine-group minerals could be incorporated into acidic soils, where they would act as fertilizer and a neutralization agent while reacting with the atmosphere to produce carbonate minerals. Ferrini et al. (2009) have developed an industrial process by which anthropogenic CO2 may be rapidly and effectively stored within a low-temperature carbonate mineral, nesquehonite [MgCO3·3H2O], by reaction with Mg-rich wastewater  3  from desalination plants and oil fields. Kelemen and Matter (2008) propose that acceleration of carbon mineralization via natural silicate weathering within ophiolites could consume on the order of 1 Gt of CO2 per year within peridotite deposits in Oman alone. Of relevance to this thesis, an expanding body of work suggests that in situ weathering of silicate minerals, within the mineral waste from some mines, may be used to sequester a significant amount of CO2.  1.2 Carbon mineralization within mine tailings Mineral-fluid reactions are greatly accelerated in mine tailings because the milling process leads to a large increase in reactive surface area (e.g., White et al. 1996; Molson et al. 2005). Carbonate minerals are known to precipitate during mineral-fluid reactions in mine tailings at several sites in Canada: the Kidd Creek copper-zinc mine near Timmins, Ontario (Al et al. 2000), the Lower Williams Lake uranium mine near Elliot Lake, Ontario (Paktunc and Davé 2002), and in chrysotile mine tailings at Thetford, Québec (Huot et al. 2003), Clinton Creek, Yukon Territory (Wilson et al. 2004), and Cassiar, British Columbia (Wilson et al. 2005). Wilson et al. (2006) and Rollo and Jamieson (2006) suggest that if atmospheric CO2 were the primary source of carbon in the carbonate minerals that develop within ultramafic mine tailings, the weathering process by which they result could be used to sequester CO2. Also, carbonation of bauxite mine residue to calcite and dawsonite (by reaction with atmospheric CO2) has been proposed as a means by which aluminum mines might neutralize their basic tailings, and offset some portion of their emissions (Jones et al. 2006; Khaitan et al. 2009). Wilson et al. (2009) used a combination of stable carbon and oxygen isotope, and radiocarbon analyses to demonstrate that tailings from the historical chrysotile mines at Clinton Creek, Yukon Territory and Cassiar, British Columbia are trapping and storing CO2 from the atmosphere. Previously, Wilson et al. (2006) developed a technique for use with the Rietveld method and X-ray powder diffraction data (Rietveld 1969; Hill and Howard 1987; Bish and Howard 1988) that enables quantitative phase analysis in samples containing disordered minerals (e.g., serpentine-group minerals). Because  4  ultramafic mine tailings contain abundant serpentine-group minerals, the combination of the quantitative method of Wilson et al. (2006) and the isotopic verification protocol of Wilson et al. (2009) allows the carbonate mineral traps for atmospheric CO2 in ultramafic mine tailings to be identified and quantified. Measured abundances of these mineral traps obtained with the Rietveld method can then be used to estimate the amount of atmospheric CO2 trapped within mine tailings. Deposit types that produce tailings suitable for CO2 sequestration include, but are not limited to: Cu-Ni-PGE deposits hosted by dunite, serpentinite, and gabbro-norite; serpentinite-hosted chrysotile; diamondiferous kimberlite pipes, and podiform chromite deposits in layered mafic intrusions. The only mineralogical prerequisite to efficient CO2 sequestration in mine tailings are the high abundance of Mg-silicate minerals. Although, a low abundance of acid-generating sulphide minerals could be considered a further prerequisite, Power et al. (in review) have demonstrated that the addition of an acid generating substance and a microbial catalyst can accelerate weathering of serpentine minerals, while maintaining a neutral to basic pH (under which most carbonate minerals are stable).  1.3 Carbon mineralization in tailings from active mines Previous investigations into verification and quantification of passive carbon mineralization in ultramafic mines have been undertaken at historical mine sites (Wilson et al. 2006, 2009). There are few data dealing with this phenomenon as it occurs in the tailings from active mines. Active mining operations may be able to take advantage of carbon mineralization within their tailings to offset part of their greenhouse gas emissions. However, the sources and cycles for carbon in active mining operations are, respectively, more plentiful and more complicated than in historical mine tailings environments. In addition, carbonate minerals are not always present at high abundance, and thus not readily detectable, in the recently deposited tailings from active mining operations. In order to enable mining companies to claim credit for CO2 sequestration, there must exist an  5  unbiased scientific protocol for identifying, verifying, and quantifying sequestration of CO2 within mine tailings. This work seeks to improve our understanding of the carbon cycle in mine tailings, and to make a contribution to mineralogical and geochemical strategies for assessing carbon mineralization in mine tailings. Carbon is being mineralized within ultramafic mine tailings from the active mining operations at the Mount Keith Nickel Mine, Western Australia, Australia (Fig. 1.1), and the Diavik Diamond Mine, Northwest Territories, Canada (Fig. 1.2).  Figure 1.1: (a) and (b) Location of Mount Keith Nickel Mine, Western Australia, Australia. (c) Satellite photograph of the open pit at the MKD5 deposit and tailings storage facility, Mount Keith.  At both sites, which are hosted by ultramafic rocks, hydrated magnesium carbonate minerals develop within mine tailings by weathering of Mg-rich gangue minerals. Differences in climate and tailings management practices have resulted in widespread  mineralization  and  preservation  of  secondary  hydromagnesite 6  [Mg5(CO3)4(OH)2·4H2O] at Mount Keith, and limited mineralization of nesquehonite [MgCO3·3H2O], calcite, and Na-carbonate minerals in the tailings at Diavik. These mines are characterized by significant contrasts in climate, rock type, and tailings management practices. The goal of this thesis is to exploit these contrasts in order to develop mineralogical and geochemical strategies for identifying, verifying, and quantifying carbon mineralization in tailings from actively producing mines hosted by ultramafic rocks.  Figure 1.2: (a) and (b) Location of Diavik Diamond Mine, Northwest Territories, Canada. (c) Aerial photograph of the Diavik Diamond Mine (courtesy of Diavik Diamond Mines Inc.).  7  1.4 Organization and outcomes of this thesis An approach to estimating CO2 sequestration within minerals in mine tailings must provide quantitative mineralogical and geochemical information on the sample scale and apply it at the scale of a tailings storage facility. It is necessary to consider the representativeness of samples, the limitations of quantitative mineralogical methods, the potential pitfalls in collecting and interpreting geochemical data, and the physical properties of the materials under study. The goal of this thesis is to address these concerns in order to outline a strategy by which mining companies may account for the capture and storage of carbon within mine tailings. Four manuscripts are assembled here for this purpose. The second chapter of this thesis (Chapter 2) expands the scope of the mineralogical toolbox for assessing carbon mineralization in mine tailings, the subject of my master’s thesis (Wilson 2005), published as Wilson et al. (2006). Chapter 3 presents new data for the fractionation of carbon isotopes in dypingite, a poorly-understood Mg-carbonate mineral which acts as a mineral trap for carbon in mine tailings. In Chapters 4 and 5, the results of the preceding chapters are applied to mine-scale assessments of carbon mineralization at the Mount Keith Nickel Mine, and the Diavik Diamond Mine. Chapter 2 has been published in Applied Geochemistry. Chapters 3, 4, and 5 have undergone internal review and are intended for submission to peer-reviewed international journals. Brief descriptions of these manuscript-style chapters follow. Chapter 2 (published in Applied Geochemistry) is a comparative study of quantitative X-ray powder diffraction (XRPD) techniques applied to measurement of trace amounts of a Mg-carbonate mineral. Carbon is being mineralized within trace amounts of nesquehonite [MgCO3·3H2O] in the kimberlite mine tailings at the Diavik Diamond Mine. Although the abundance of nesquehonite in these mine tailings is low, trapping of CO2 within trace minerals may be offsetting kilotonnes of greenhouse gas emissions at Diavik and other large mines. The applicability and accuracy of three techniques for quantitative phase analysis with XRPD data to the measurement of trace amounts of nesquehonite in kimberlite mine tailings is evaluated. These techniques are tested on synthetic mixtures of pure minerals and are subsequently applied to  8  measurement of trace nesquehonite in natural kimberlite mine tailings from the Diavik Diamond Mine. Results of quantitative phase analysis are used to estimate trapping of CO2 within trace mineralization of nesquehonite at Diavik. Chapter 3 describes the results of mineral synthesis experiments, which are used to obtain a stable isotopic fractionation factor for carbon between the Mg-carbonate mineral, dypingite [Mg5(CO3)4(OH)2·5H2O], and dissolved inorganic carbon. Currently, the only published data for experimentally determined fractionation factors for any Mgcarbonate minerals are those reported by O’Neil and Barnes (1971) for exchange of oxygen between hydromagnesite and water at 0˚ and 25˚C. The lack of isotopic fractionation factors for Mg-carbonate minerals hampers the interpretation of stable isotope data collected from these minerals. This dearth of information also interferes with the development of a consistent and reliable protocol for verifying mineralization of atmospheric CO2 within dypingite and related minerals in ultramafic mine tailings. To begin bridging this gap, open system precipitation experiments are undertaken to study the partitioning of  13  C between dypingite and total dissolved inorganic carbon in an  evaporative and saline system. These experiments are designed to simulate the conditions under which atmospheric CO2 is mineralized within the chemically and structurally related mineral, hydromagnesite, at the Mount Keith Nickel Mine. Chapter 4 is a detailed site study of carbon mineralization at the Mount Keith Nickel Mine. In this chapter, field observations, stable and radiogenic isotope geochemistry, and extensive use of quantitative mineralogy are combined to fingerprint the source of carbon in secondary hydromagnesite [Mg5(CO3)4(OH)2·4H2O], and are used to estimate the amount of CO2 that is fixed within this mineral at Mount Keith. In order to estimate the total amount of CO2 captured within hydromagnesite at this mine, it is necessary that the mineralogy of its tailings storage facilities be well constrained. This requires extensive sampling and the construction of a database of quantitative mineralogical data. Rietveld refinement results for more than 200 samples of mine tailings are used to obtain an empirical rate for mineralization of hydromagnesite, and to estimate the amount of CO2 trapped and stored within mine tailings. The three-isotope system (i.e., δ13C, δ18O, and F14C) employed by Wilson et al. (2009), and the results of Chapter 3, are used to fingerprint the source of CO2 within hydromagnesite and to  9  provide more precise constraints on the amount of CO2 fixed within the tailings at Mount Keith. In Chapter 5, stable and radiogenic isotope data for primary and secondary carbonate minerals from the Diavik Diamond Mine are examined. The diversity of mineral species identified at Diavik and the input of carbon and oxygen from multiple reservoirs give rise to stable isotopic data that are difficult to interpret. Under such conditions, it can be challenging to implicate specific reservoirs in the precipitation of minerals, and therefore, to assess whether atmospheric CO2 is being sequestered into secondary carbonate minerals. In this chapter, I model three scenarios by which carbon may be mixed and cycled between different reservoirs in mine tailings (using conditions based on those at Diavik and Mount Keith). These models make use of measured values from Diavik and Mount Keith, published equilibrium fractionation factors, and the fractionation factor obtained for dypingite in Chapter 3. Model results are used to interpret stable carbon and oxygen isotopic data and radiocarbon data for the Diavik Diamond Mine and three other mines: the Mount Keith Nickel Mine (using data from Chapter 4), and the historical chrysotile mines at Clinton Creek and Cassiar (using data from Wilson et al. 2009). Based on comparison of isotopic data with model results, field observations, and quantitative mineralogical results from Chapters 2 and 4, the impact of mechanical mixing and bedrock carbonate recycling on CO2 sequestration at these four mines is considered. The validity of using stable isotopes of carbon and oxygen as the primary tools for verifying sequestration of atmospheric CO2 within mine tailings is assessed. In Chapter 6, the most significant results of this thesis are summarized, and the implications of these results for sequestration of CO2 within ultramafic mine tailings are discussed. From the conclusions arrived at during this work, suggestions are made for future lines of research.  10  1.5 References Al, T.A., Martin, C.J., and Blowes, D.W. (2000) Carbonate-mineral/water interactions in sulfide-rich mine tailings. Geochimica et Cosmochimica Acta, 64, 3933-3948. Berner, R.A. (1990) Atmospheric carbon dioxide levels over Phanerozoic time. Science, 249, 1382-1386. Bickle, M., Chadwick, A., Huppert, H.E., Hallworth, M., and Lyle, S. (2007) Modelling carbon dioxide accumulation at Sleipner: Implications for underground carbon storage. Earth and Planetary Science Letters, 225, 164-176. Bish, D.L. and Howard, S.A. (1988) Quantitative phase analysis using the Rietveld method. Journal of Applied Crystallography, 21, 86-91. Broecker, W.S. (2007) CO2 arithmetic. Science, 315, 1371. Cantucci, B., Montegrossi, G., Vaselli, O., Tassi, F., Quattrocchi, F., and Perkins, E.H. (2009) Geochemical modeling of CO2 storage in deep reservoirs: The Weyburn Project (Canada) case study. Chemical Geology, 265, 181-197. Department of Indian Affairs and Northern Development (1993) Guidelines for Acid Rock Drainage Prediction in the North: Northern Mine Environment Neutral Drainage Studies No. 1. Prepared by Steffen, Robertson and Kirsten (B.C.) Inc. Ottawa, ON, Canada. Ferrini, V., De Vito, C., and Mignardi, S. (2009) Synthesis of nesquehonite by reaction of gaseous CO2 with Mg chloride solution: Its potential role in the sequestration of carbon dioxide. Journal of Hazardous Materials, 168, 832-837. Hill, R.J. and Howard, C.J. (1987) Quantitative phase analysis from neutron powder diffraction data using the Rietveld method. Journal of Applied Crystallography, 20, 467-474. Hoffert, M.I., Caldeira, K., Benford, G., Criswell, D.R., Green, C., Herzog, H., Jain, A.K., Kheshgi, H.S., Lackner, K.S., Lewis, J.S., Lightfoot, H.D., Manheimer, W., Mankins, J.C., Mauel, M.E., Perkins, L.J., Schlesinger, M.E., Volk, T., and Wigley, T.M.L. (2002) Advanced technology paths to global climate stability: energy for a greenhouse planet. Science, 298, 981-987.  11  Huijgen, W.J.J. and Comans, R.N.J. (2005) Carbon dioxide sequestration by mineral carbonation: Literature review update 2003-2004. Energy Research Centre of the Netherlands, Report ECN-C-05-022. Huijgen, W.J.J. and Comans, R.N.J. (2003) Carbon dioxide sequestration by mineral carbonation: Literature review. Energy Research Centre of the Netherlands, Report ECN-C-03-016. Huot, F., Beaudoin, G., Hebert, R., Constantin, M., Bonin, G., and Dipple, G. (2003) Evaluation of Southern Québec asbestos residues for CO2 sequestration by mineral carbonation; preliminary results. Joint Annual Meeting of the Geological and Mineralogical Associations of Canada, Vancouver, Canada. May 25-28, 2003. International Energy Agency (2008) Energy Technology Perspectives 2008: Scenarios & Strategies to 2050. Organisation for Economic Co-operation and Development Publishing, 648 p. IPCC (2007) Climate Change 2007: The Physical Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., and Miller, H.L., Eds. Cambridge University Press, Cambridge, UK and New York, NY, USA, 996 p. IPCC (2005) IPCC Special Report on Carbon Dioxide Capture and Storage. Metz, B., Davidson, O., de Coninck, H.C., Loos, M., and Meyer, L.A., Eds. Cambridge University Press, Cambridge, UK and New York, NY, USA, 431 p. Jones, G., Joshi, G., Clark, M., and McConchie, D. (2006) Carbon capture and the aluminium industry: Preliminary studies. Environmental Chemistry, 3, 297-303. Kelemen, P.B. and Matter, J. (2008) In situ carbonation of peridotite for CO2 storage. Proceedings of the National Academy of Sciences of the USA, 105, 1729517300. Khaitan, S., Dzombak, D.A., and Lowry, G.V. (2009) Journal of Environmental Engineering, 133, 433-438. Kump, L.R., Brantley, S.L., and Arthur, M.A. (2000). Chemical weathering, atmospheric CO2, and climate. Annual Review of Earth and Planetary Sciences, 28, 611-667.  12  Lackner, K.S. (2003) Climate change: A guide to CO2 sequestration. Science, 300, 16771678. Lackner, K.S., Wendt, C.H., Butt, D.P., Joyce, G.L., and Sharp, D.H. (1995) Carbon dioxide disposal in carbonate minerals. Energy, 20, 1153-1170. Molson, J.W., Fala, O., Aubertin, M., and Bussière, B. (2005) Numerical simulations of pyrite oxidation and acid mine drainage in unsaturated waste rock piles. Journal of Contaminant Hydrology, 78, 343-371. O’Neil, J.R. and Barnes, I. (1971) C13 and O18 compositions in some fresh-water carbonates associated with ultramafic rocks and serpentinites: western United States. Geochimica et Cosmochimica Acta, 35, 687-697. Oreskes, N. (2004) Beyond the Ivory Tower: the scientific consensus on climate change. Science, 306, 1686. Pacala, S. and Socolow, R. (2004) Stabilization wedges: solving the climate problem for the next 50 years with current technologies. Science, 305, 968-972. Paktunc, A.D. and Davé, N.K. (2002) Formation of secondary pyrite and carbonate minerals in the Lower Williams Lake tailings basin, Elliot Lake, Ontario, Canada. American Mineralogist, 87, 593-602. Power, I.M., Dipple, G.M., and Southam, G. Bioleaching of ultramafic tailings by Acidithiobacillus spp. for CO2 sequestration. Environmental Science & Technology, in review. Rietveld, H.M. (1969) A profile refinement method for nuclear and magnetic structures. Journal of Applied Crystallography, 2, 65-71. Rollo, H.A. and Jamieson, H.E. (2006) Interaction of diamond mine waste and surface water in the Canadian Arctic. Applied Geochemistry, 21, 1522-1538. Schuiling, R.D. and Krijgsman, P. (2006) Enhanced weathering: An effective and cheap tool to sequester CO2. Climatic Change, 774, 349-354. Schwartzman, D.W. and Volk, T. (1989) Biotic enhancement of weathering and the habitability of Earth. Nature, 340, 457-460. Seifritz, W. (1990) CO2 disposal by means of silicates. Nature, 345, 486.  13  Sipilä, J., Teir, S., and Zevenhoven, R. (2008) Carbon dioxide sequestration by mineral carbonation: Literature review update 2005-2007. Åbo Akademi University Heat Engineering Laboratory, Report 2008-1. Sundquist, E.T. (1993) The global carbon dioxide budget. Science, 259, 934-941. Sundquist, E.T. (1985) Geological perspectives on carbon dioxide and the carbon cycle. In The Carbon Cycle and Atmospheric CO2: Natural Variations Archaen to Present. Sundquist, E.T. and Broecker, W.S., Eds. Geophysical Monographs 32, Washington, DC, American Geophysical Union, p. 5-60. White, A.F., Blum, A.E., Schulz, M.S., Bullen, T.D., Harden, J.W., and Peterson, M.L. (1996) Chemical weathering rates of a soil chronosequence on granitic alluvium: I. Quantification of mineralogical and surface area changes and calculation of primary silicate reaction rates. Geochimica et Cosmochimica Acta, 60, 25332550. Wilson, S.A., Dipple, G.M., Power, I.M., Thom, J.M., Anderson, R.G., Raudsepp, M., Gabites, J.E., and Southam, G. (2009) Carbon dioxide fixation within mine wastes of ultramafic-hosted ore deposits: Examples from the Clinton Creek and Cassiar chrysotile deposits, Canada. Economic Geology, 104, 95-112. Wilson, S.A., Raudsepp, M., and Dipple, G.M. (2006) Verifying and quantifying carbon fixation in minerals from serpentine-rich mine tailings using the Rietveld method with X-ray powder diffraction data. American Mineralogist, 91, 1331-1341. Wilson, S.A., Thom, J.M., Dipple, G.M., Raudsepp, M., and Anderson, R.G. (2005) Towards sustainable mining: uptake of greenhouse gases by mine tailings. British Columbia and Yukon Chamber of Mines Mineral Exploration Roundup, Vancouver, Canada. January 24-27, 2005. Wilson, S. A., Dipple, G.M., Anderson, R.G., and Raudsepp, M. (2004) Characterization of Clinton Creek mine residues and their suitability for CO2 sequestration. British Columbia and Yukon Chamber of Mines Mineral Exploration Roundup, Vancouver, Canada. January 26-29, 2004.  14  Chapter 2 Quantifying carbon fixation in trace minerals from processed kimberlite: A comparative study of quantitative methods using X-ray powder diffraction data with applications to the Diavik Diamond Mine, Northwest Territories, Canada1  2.1. Introduction Emission of anthropogenic greenhouse gases (e.g., CO2, CH4, and N2O) has been implicated as a cause of current warming of the Earth’s climate. Carbon dioxide (CO2) is by far the most significant of these greenhouse gases, representing 77% of total anthropogenic emissions in 2004 (Solomon et al. 2007). By 2005, the global atmospheric concentration of CO2 had increased to 379 ppm from a pre-1750 (i.e., pre-industrial) value of 280 ± 20 ppm. Approximately two-thirds of this increase is attributed to combustion of fossil fuels and one-third to changes in land-use patterns during the past 260 years (Solomon et al. 2007). It has been suggested that strategies for decarbonizing energy sources, increasing energy efficiency, and trapping and storing CO2 must be developed and implemented in order to stabilize concentrations of atmospheric CO2 and curtail the most damaging effects of anthropogenic climate change (e.g., Hoffert et al. 2002; Lackner 2003; Pacala and Socolow 2004; Broecker 2007; Solomon et al. 2007). Approximately 90% of carbon on Earth is fixed within carbonate minerals (Sundquist 1985; Sudquist 1993) and it is expected that these minerals will be the ultimate sink for most anthropogenic CO2 on a timescale of 106 years (Kump et al. 2000). Storage of CO2 in carbonate minerals is recognized as a safe and effective method for the sequestration of anthropogenic carbon (Seifritz 1990; Lackner et al. 1995; Lackner 2003). Precipitation of carbonate minerals in situ by dissolution of silicate mine residues  1  A version of this chapter has been published. Wilson, S.A., Raudsepp, M., and Dipple, G.M. (2009) Quantifying carbon fixation in trace minerals from processed kimberlite: A comparative study of quantitative methods using X-ray powder diffraction data with applications to the Diavik Diamond Mine, Northwest Territories, Canada. Applied Geochemistry, 24, 2312-2331. doi:10.1016/j.apgeochem.2009.09.018.  15  is one potential implementation of this process. The development of secondary carbonate minerals has been documented in tailings at several mine sites in Canada: at the Kidd Creek copper-zinc mine near Timmins, Ontario (Al et al. 2000), the Lower Williams Lake uranium mine near Elliot Lake, Ontario (Paktunc and Davé 2002), and in chrysotile mine tailings at Thetford, Québec (Huot et al. 2003), Clinton Creek, Yukon Territory (Wilson et al. 2004), and Cassiar, British Columbia (Wilson et al. 2005). The carbon bound within secondary carbonate minerals in the tailings at some of these mines may not have had an atmospheric source. However, Wilson et al. (2009) demonstrate that tailings from the historical chrysotile mines at Clinton Creek, Yukon Territory and Cassiar, British Columbia are trapping and storing CO2 from the atmosphere. Accelerating the uptake of CO2 into tailings from active mines could reduce or offset the net greenhouse gas emissions of many mining operations. Bulk geochemical methods for CO2 abundance cannot distinguish among various carbonate minerals nor can they discern the difference between atmospheric, bedrock, biological, and industrial sources of carbon within minerals. However, the sources of bound carbon can be distinguished using radiocarbon and stable isotopes of carbon and oxygen (Wilson et al. 2009). Also, automated point-counting techniques (e.g. mineral liberation analysis) cannot be used to quantify fine-grained minerals or hydrous minerals that are easily vaporized by an electron beam. As an alternative to point counting, the amount of CO2 trapped within fine-grained, hydrous carbonate minerals can be estimated from weight-percent abundances determined with quantitative phase analysis using X-ray powder diffraction (XRPD) data. At the Diavik Diamond Mine, Northwest Territories, Canada, efflorescent films of Ca, Na, and Mg-carbonate minerals form in the tailings from the fine and coarse Processed Kimberlite Containment facilities (PKC). These minerals precipitate at the surface of kimberlite waste that is beached along the edge of a central pond used for storing process water. Based on our observations, the most common secondary carbonate mineral, and the one best preserved at depth, is nesquehonite (MgCO3·3H2O). Also, geochemical modelling by Rollo and Jamieson (2006) suggests that carbon mineralization may be occurring in waste kimberlite at the nearby EKATI Diamond Mine.  16  Many studies have demonstrated the accuracy and precision of the Rietveld method for determining mineral abundances from XRPD data (e.g., Hill and Howard 1987; Bish and Howard 1988; Bish and Post 1993; Raudsepp et al. 1999; De la Torre et al. 2001; De la Torre and Aranda 2003; Ufer et al. 2004; Omotoso et al. 2006; Wilson et al. 2006). Processed kimberlite from Diavik contains a variety of minerals that are characterized by one or more of the following: (1) extensive solid solution, (2) structural disorder, and (3) severe preferred orientation. Furthermore, processed kimberlite at Diavik generally contains abundant serpentine and forsterite with minor to trace amounts of many other phases, resulting in complicated XRPD patterns that consist of many overlapped peak profiles. The combination of these factors presents a challenge for quantifying carbon mineralization with the Rietveld method (Rietveld 1969). This is chiefly because the Rietveld method requires that the crystal structures and chemistry of the phases being analyzed be known and also gives less reliable results for minerals present at low abundances (e.g., Raudsepp et al. 1999). Alternatively, Chung’s (1974) method of normalized reference intensity ratios (RIR) and similar methods have been used successfully to quantify trace abundances of minerals in multi-phase mixtures (e.g., Bish and Chipera 1991; Omotoso et al. 2006). Calibration curves produced according to the internal standard method (Alexander and Klug 1948) have also been used successfully to measure trace abundances of minerals (e.g., Sanchez and Gunter 2006). These three methods are typically used in isolation on very different systems of minerals and, as such, it is difficult to recommend the use of one method over another for quantification of low abundances of minerals. Here we assess the ability of Chung’s RIR method, the internal standard method, and the Rietveld method to measure minor to trace amounts of nesquehonite using weighed mixtures of pure mineral standards, prepared to simulate processed kimberlite. Based on the results of our tests, we outline a procedure for accurate quantification of CO2 trapping within trace minerals in kimberlite mine tailings. This procedure is subsequently applied to samples of natural processed kimberlite from the fine and coarse PKC at Diavik. Results of quantitative phase analysis are used to estimate trapping of CO2 within nesquehonite at Diavik (from stoichiometry of this mineral) and to determine the contribution of  17  nesquehonite to neutralization potential of the tailings according to the method of Jambor et al. (2007).  2.2 Locality and sampling strategy The Diavik Diamond Mine is located on East Island, in Lac de Gras, approximately 300 km northeast of Yellowknife, Northwest Territories, Canada (Fig. 2.1). There are four mineable kimberlite pipes on the Diavik property; two of which, A154 North and A154 South, are currently being mined from a single open pit. A third pipe, A418, is located to the south of the A154 pipes and began production in 2008. The fourth pipe, A21, is not currently being developed. The kimberlites at Lac de Gras, including those at Diavik, intrude Late Archean granitoids and supracrustal rocks of the Yellowknife Supergroup in the Slave Structural Province (Graham et al. 1998). Approximately 2 Mt/year of kimberlite ore is being mined from the A154 open pit. After undergoing processing to remove diamonds, the kimberlite is transported to one of two locations for permanent storage. Most of the processed kimberlite is piped, suspended in process water, into a natural basin (termed the fine processed kimberlite containment facility or fine PKC) where it is stored beneath a pond of process water. A much smaller amount of coarse-grained waste material is stored subaerially in a pile (called the coarse processed kimberlite containment facility or coarse PKC). The bulk of the processed kimberlite from the fine PKC is characterized by major abundances of serpentine minerals (predominantly lizardite) and high-Mg forsterite with minor amounts of calcite, Cr-rich diopside, Mg-rich garnets, plagioclase, phlogopite, quartz, and clay minerals (predominantly vermiculite and possibly smectite or interstratified clays). Traces of chromite, dolomite, muscovite, perovskite, and amphibole have also been observed. The mineralogy of the coarse PKC is similar to that of the fine PKC, but differs notably in that trace amounts of pyrite and jarosite have been observed in the course of microscope work on the coarse processed kimberlite. Efflorescent crusts of secondary minerals are common in the fine PKC and occur occasionally in the coarse PKC at Diavik. These occur in four distinct modes (Fig. 2.2):  18  (1) white, powdery efflorescences on vertical and horizontal surfaces in the fine PKC, (2) crusts of a white, powdery precipitate just below the surface of the coarse PKC, (3) preserved efflorescent minerals at depth within the fine PKC, and (4) surface crusts in the fine PKC where nutrient-rich waste water from the sewage treatment plant has been deposited. We have observed the presence of nesquehonite (MgCO3·3H2O) in all four types of efflorescence. Nesquehonite is a common low-temperature alteration product of serpentinite and serpentine-rich mine wastes (e.g., Suzuki and Ito 1973; Suzuki and Ito 1974; Inaba et al. 1985; Giester et al. 2000; Wilson et al. 2006; Wilson et al. 2009). Nesquehonite also forms by evaporation in creeks and playa environments that are fed by groundwater derived in part from drainage through ultramafic rock units (e.g., O’Neil and Barnes 1971; Power et al. 2007; Power et al. 2009).  Figure 2.1: (a) and (b) Location of Diavik Diamond Mine, Northwest Territories, Canada. (c) Aerial photograph of the Diavik Diamond Mine (courtesy of Diavik Diamond Mines Inc.).  19  Figure 2.2: Modes in which nesquehonite and other efflorescent minerals have been identified in the processed kimberlite at Diavik: (a) as a white, powdery film on horizontal and vertical surfaces in the fine PKC (Group 1), (b) as horizontally continuous precipitates immediately below the surface of the coarse PKC (Group 2), (c) as preserved efflorescent precipitates at depth beneath the surface of the fine PKC (Group 3, show here as 10-cm diameter core holes), and (d) as discrete deposits of pale-coloured material on some horizontal surfaces in the fine PKC (Group 4, rucksack for scale).  Group 1 efflorescences are widespread at the surface of the fine PKC around the circumference of the pond. They are typically rich in either nesquehonite or sulphate minerals (commonly gypsum). Group 2 efflorescences contain sulphate minerals (variably some combination of anhydrite, gypsum, epsomite, hexahydrite, syngenite, and possibly butlerite) or occasionally nesquehonite. Efflorescences from Groups 1 and 2 are observed as thin films less than 1 mm in thickness. Group 3 comprises efflorescent nesquehonite, preserved in trace amounts at depth within the fine PKC (where trace abundance is defined here as < 0.5 wt.%). Group 4 crusts occur more rarely as thick (> 1 mm), continuous patches near the perimeter wall and road of the fine PKC. These deposits are rich in portlandite and Ca- and Na-carbonate minerals (primarily calcite,  20  gaylussite, natrite, thermonatrite, trona, and vaterite) with occasional nesquehonite, gypsum, and ettringite. Although Group 1 and 4 efflorescences cover a significant portion of the surface of the processed kimberlite beached along the circumference of the fine PKC pond, we have found that only Group 3 efflorescences persist at depth. These preserved crusts of nesquehonite likely represent most of the carbon mineralization in either the fine or coarse PKC at Diavik, and are the main focus of this study. Most of the sampling conducted at Diavik was done in August 2006 in the fine PKC with some additional samples having been collected from the coarse PKC. Limited sampling was carried out in September 2005. Sampling locations for the fine PKC are provided in Figure 2.3. One sample examined in this study was taken from just below the surface of the coarse PKC and three of the samples were collected by trowel from the surface of the fine PKC. The remaining 16 samples were collected with a sedimentcoring device, in 10-cm lengths, from depths of either 0 cm (i.e., from 0-10 cm depth) or 100 cm (i.e., from 95-105 cm depth) below the surface of the fine PKC.  Figure 2.3: Sampling locations within the storage facility for processed kimberlite fines at Diavik (fine PKC). Modified from a report by Reinson (2006) for Golder Associates Ltd. Small arrows indicate locations of discharge spigots. The approximate extent of the tailings pond is shown for late summer 2006. Co-ordinates are given in NAD 83. Bathymetric data were collected in 2005 (from Reinson 2006).  21  2.3 Experimental method 2.3.1 Sample preparation and data collection Thirteen mixtures of pure mineral samples were prepared to simulate processed kimberlite (mine tailings) (Table 2.1). Minerals used in these mixtures represent those commonly found in the tailings at Diavik: lizardite, high-Mg forsterite, diopside, almandine-pyrope series garnet, phlogopite, calcite, quartz, oligoclase, and nesquehonite. The abundance of nesquehonite in the synthetic tailings was varied from 0.10 wt.% to 5.00 wt.% in order to assess the applicability of the Rietveld method (1969), Chung’s (1974) normalized RIR method (also known as the “adiabatic method”), and the internal standard method (Alexander and Klug 1948) for measuring the abundance of this phase for a range of trace and minor abundances. Samples of the constituent minerals were checked for purity using X-ray powder diffraction and energy-dispersion X-ray spectroscopy. The serpentine used in the synthetic mixtures and several samples of serpentine from Diavik were identified to be predominantly lizardite using X-ray powder diffraction and dispersive Raman microspectroscopy according to the method of Rinaudo et al. (2003). All crystalline standards, with the exception of the lizardite, were at least 99% pure. Rietveld refinement results indicated that the lizardite contained 1.1 wt.% calcite. Pure mineral phases were weighed on a scale with ± 0.1 mg precision. Prior to each use, the scale was calibrated and tested for accuracy using a set of weights ranging from 1 mg to 500 mg. Variable amounts of forsterite (30.00, 33.00, 34.00, 34.50, 34.75, and 34.90 wt.%) and nesquehonite (5.00, 2.00, 1.00, 0.50, 0.25, and 0.10 wt.%) were added to the mixtures in Series 1 and 2, such that the abundance of these two phases totalled 35.00 wt.% (Table 2.1). Mixtures containing 0.10-0.50 wt.% nesquehonite were weighed to total 4.00 g, and those containing 1.00-5.00 wt.% nesquehonite totalled 2.00 g (after Bish and Chipera 1991). Samples containing less than 0.10 wt.% nesquehonite were not prepared because large uncertainties would result from weighing and any attempt to homogenize such mixtures.  22  Table 2.1: Compositions of synthetic processed kimberlite renormalized to exclude corundum and taking account of calcite contamination in lizardite. Phase Lizardite Forsterite Nesquehonite AlmandinePyrope Calcite Diopside Oligoclase Phlogopite Quartz Total  Series 1 07lsk3r 07lsk4r 44.50 44.50 34.50 34.00 0.50 1.00 5.00 5.00  07lsk1r 44.50 34.90 0.10 5.00  07lsk2r 44.50 34.75 0.25 5.00  5.50  5.50  5.50  5.00 5.00 100.00  5.00 5.00 100.00  5.00 5.00 100.00  07lsk5r 44.50 33.00 2.00 5.00  07lsk6r 44.50 30.00 5.00 5.00  07lsk7r 34.62 34.90 0.10 5.00  07lsk8r 34.62 34.75 0.25 5.00  Series 2 07lsk9r 07lsk10r 34.62 34.62 34.50 34.00 0.50 1.00 5.00 5.00  5.50  5.50  5.50  5.00 5.00 100.00  5.00 5.00 100.00  5.00 5.00 100.00  5.38 5.00 5.00 5.00 5.00 100.00  5.38 5.00 5.00 5.00 5.00 100.00  5.38 5.00 5.00 5.00 5.00 100.00  5.38 5.00 5.00 5.00 5.00 100.00  07lsk11r 34.62 33.00 2.00 5.00  07lsk12r 34.62 30.00 5.00 5.00  5.38 5.00 5.00 5.00 5.00 100.00  5.38 5.00 5.00 5.00 5.00 100.00  23  Each mixture was ground under anhydrous ethanol using agate grinding elements for six minutes in a McCrone micronizing mill to reduce the mean grain size and to ensure homogenization. At this point, grinding was halted and an internal standard of synthetic corundum (prepared from smelter grade alumina according to Australian Standard AS 2879.3-1991) was added to each sample such that it constituted 10 wt.% of the renormalized weight. Each sample was then ground for an additional four minutes, for a total grinding time of 10 minutes. Samples of synthetic processed kimberlite were dried at room temperature under a fume hood and disaggregated with an agate mortar and pestle once dry. One duplicate mixture (Series 1, 07lsk6r-2) and one additional mixture (Series 2, 07lsk12r) were prepared using 10 wt.% of NIST 676a corundum as an internal standard. Grinding times were chosen to minimize both degradation of the lizardite structure and contamination of the samples by quartz from the grinding elements while optimizing particle-size reduction and homogenization. The micronizing mill was used to reduce the mean particle size of the samples to micron or sub-micron level. Knowledge of the mean grain size of the narrow size distribution afforded by micronizing allows for the use of the Brindley (1945) correction for microabsorption in Rietveld refinements. Diopside and oligoclase were omitted from the Series 1 mixtures to assess the effect of the (110) reflection of diopside (d≈6.47 Å) and the (1 1 0), (110), (020), and (001) reflections of oligoclase (d≈6.30, 6.34, 6.38, and 6.39 Å, respectively) on detection and measurement of the (101) and (10 1 ) reflections €of nesquehonite (d≈6.48 Å and d≈6.52 Å, respectively). In addition, 20 samples € of processed kimberlite from Diavik were prepared to assess the Rietveld, RIR, and internal standard methods on natural samples of kimberlite mine tailings. Samples were left in a drying hood for a minimum of 48 hours and were then homogenized mechanically with a spatula. Once dried and homogenized, an aliquot of each sample (i.e., 50–100 g) was powdered using a tungsten carbide ringmill. Twogram aliquots of natural processed kimberlite were ground under ethanol for six minutes in the McCrone micronizing mill. At this point, 10 wt.% of synthetic corundum was added to each sample and the samples were ground for an additional four minutes for a total grinding time of 10 minutes. 24  Natural and synthetic samples were mounted in a back-loading aluminum cavity holder of the design described by Raudsepp and Pani (2003). Powdered samples were loaded against the roughened surface of a sheet of glass that covered the top of the cavity. Data for Rietveld refinement of synthetic processed kimberlite were collected on specimens prepared in two ways: (1) on specimens that had been serrated with a razor blade along two axes (one parallel to the axis of the diffractometer goniomenter and the second in the perpendicular direction) and (2) on non-serrated specimens. Specimens were serrated to inhibit preferred orientation of crystallites, particularly those of phlogopite. Data for Rietveld refinement of natural processed kimberlite from Diavik were collected from non-serrated specimens. The specimens used to collect data for reference intensity ratios were not serrated with a razor blade so as to preserve the preferred orientation of nesquehonite on {101}, and thus improve detection of the (101) and (10 1 ) reflections. Additional RIR data were collected for three samples (Series 2, 07lsk7r, 07lsk8r, and 07lsk9r) that were front-loaded into the cavity holder to assess the €  impact of mounting procedure on preferred orientation and detection of nesquehonite. XRPD data were collected on a Siemens D5000 θ-2θ diffractometer equipped with a VÅNTEC-1 detector. A long, fine-focus Co X-ray tube was operated at 35 kV and 40 mA and an Fe monochromator foil was employed. All data for Rietveld refinement were collected with a step size of 0.02° 2θ and counting time of 1s/step over a range of 380° 2θ. Data for reference intensity ratios and the calibration curve were collected using a step size of 0.02° 2θ and a counting time of 120 s/step over the ranges 15.2-16.5° 2θ and 43.6-45.1° 2θ, for the (101) and (10 1 ) reflections of nesquehonite and the (110) reflection of corundum, respectively. In order to improve detection of the (101) and (10 1 ) reflections and confirm the presence of trace nesquehonite in the samples of € natural tailings, data were collected on specimens that had been ring-milled and  €  subsequently smear-mounted onto glass slides with anhydrous ethanol. These data were collected using a step size of 0.02° 2θ and a counting time of 40 s/step over the range 15.2-16.5° 2θ. Collection of XRPD data for the RIR method and calibration curve took 6 hours and 12 minutes for the reference peaks of nesquehonite and 6 hours and 34 minutes for the corundum reference peak. Counting times for RIR data were chosen to allow both patterns, for nesquehonite and corundum, to be collected overnight. Data for each 25  Rietveld refinement were acquired in 1 hour and 12 minutes, taking considerably less time. 2.3.2 Rietveld refinement and quantitative phase analysis Rietveld refinements were done with Rietveld refinement software Topas Version 3 (Bruker AXS 2004) using the fundamental parameters approach (Cheary and Coelho 1992). Sources of crystal structure data for the phases in synthetic and natural samples are listed in Table 2.2. Refinements were done using the method of Wilson et al. (2006) to compensate for planar disorder in lizardite. Specifically, peak intensities were extracted independently of atomic scattering from an XRPD pattern of the high-purity lizardite standard with the Pawley method (Pawley 1981). The extracted intensities with the cell parameters and space group of lizardite-1T (Mellini and Viti 1994) were used to fit the lizardite in XRPD patterns of synthetic and natural kimberlites as a peaks phase. Fourth-order symmetrized harmonics were used to model anisotropic peak shape in lizardite (Järvinen 1993). Without reference to atomic co-ordinates, the relative intensities of the calculated intensities in a peaks phase are unconstrained by atomic scattering. To prevent interference of lizardite peaks with peaks from other phases, their relative intensities were initially held constant and were refined only after the peaks of the other phases had been fitted. It is important to note that other serpentine minerals (i.e., chrysotile, antigorite, polygonal serpentine) have been observed in the groundmass of kimberlites (e.g., Mitchell 1986; Mitchell and Putnis 1988; Sharp et al. 1990) and may be present in small amounts in the kimberlite at Diavik. As a consequence of using a peaks phase to model the lizardite component of kimberlite samples, other, minor serpentine phases may also be fitted by the lizardite peaks phase. Furthermore, this method cannot be used to quantify more than one disordered phase per sample (e.g., lizardite plus another serpentine mineral) and does not account for additional, X-ray amorphous phases (Wilson et al. 2006).  26  Table 2.2: Sources of crystal structure data for Rietveld refinement and values used to compute neutralization potential (NP) of processed kimberlite. Mineral  Source of Structure  NP  Source of NP Data  Albite (Oligoclase)  Armbruster et al. (1990)  1  Jambor et al. (2007)  Almandine-Pyrope  Armbruster et al. (1992)  3  Jambor et al. (2007)  Calcite  Maslen et al. (1995)  1000  By definition  Corundum  Brown et al. (1993)  n/a a  n/a a  Diopside  Ahn et al. (1986)  5  Jambor et al. (2002, 2007)  Forsterite  Yu (1997)  38  Jambor et al. (2007)  Lizardite-1T  Mellini and Viti (1994)  32  Jambor et al. (2007)  Nesquehonite  Giester et al. (2000)  723  Phlogopite-1M  Collins and Catlow (1992) Glinnemann et al. (1992)  8  Calculated from Lawrence and Scheske (1997) Jambor et al. (2007)  0  Jambor et al. (2007)  Shirozu and Bailey (1966)  29  Jambor et al. (2002)  Quartz Vermiculite a  n/a - not applicable to computation of neutralization potential. Backgrounds for Series 1 synthetic processed kimberlite were modelled using  first-order Chebychev polynomials for serrated specimens and second-order Chebychev polynomials for non-serrated specimens with an additional 1/x term to aid in the fitting of the background curve at low angles of diffraction. Higher-order Chebychev polynomials (i.e. second-order for serrated specimens and third-order for non-serrated specimens) were necessary to model the backgrounds for the more complex samples in Series 2 and the samples of natural processed kimberlite from Diavik. In each refinement, the zero error was refined and the Lorentzian crystallite size and cell parameters were refined for all phases. Refinement parameters were turned on in the same order for all synthetic and natural processed kimberlites. Preferred orientation of detectable mineral phases was not  27  corrected for because this effect was relatively minor except in phlogopite. Contamination from the agate grinding elements of the micronizing mill was tested on pure mineral phases used in this study. The amount of contamination by agate when grinding corundum was also assessed. Grinding corundum for under four minutes resulted in less than 1 wt.% contamination by agate while longer grinding times gave rise to more significant contamination. Contamination of other pure mineral phases by agate was considerably less than 1 wt.% using a grinding time of 10 minutes. Because quartz is a minor component of the synthetic and natural tailings, this contamination may result in overestimates of the abundance of this phase. The crystallinity of the synthetic corundum, which was used as an internal standard, was assessed by Rietveld refinement with reference to NIST 676a synthetic corundum. Two 50-50 wt.% mixtures were prepared for this purpose and XRPD data were collected for each of the following: (1) NIST 676a corundum and an in-house standard of annealed synthetic fluorite and (2) the same synthetic fluorite and the synthetic corundum. NIST 676a is certified to be 99.02 ± 1.11 wt.% crystalline. Refinement of the pattern with NIST corundum and fluorite gave 50.2 wt.% corundum and 49.8 wt.% fluorite. From these values, the in-house fluorite standard was determined to be 98.1 +1.0/-1.1 wt.% crystalline. Refinement of the pattern with synthetic corundum and fluorite gave values of 50.3 wt.% corundum and 49.7 wt.% fluorite. Based on this result, the crystallinity of the synthetic corundum was determined to be 99.3 +1.0/-1.1 wt.%. It is important to note that the Rietveld refinement results for both mixtures are within error of the expected values, as defined empirically by previous studies on weighed mixtures of pure mineral phases (e.g., Hill and Howard 1987; Bish and Howard 1988; Bish and Post 1993; Raudsepp et al. 1999; De la Torre et al. 2001; De la Torre and Aranda 2003; Ufer et al. 2004; Omotoso et al. 2006; Wilson et al. 2006). Therefore, the actual crystallinities of the in-house fluorite standard and the synthetic corundum may differ from calculated values by several weight percent. However, it can be said with confidence that NIST 676a, the synthetic corundum, and the in-house fluorite standard have comparable crystalline fractions. The amount of lizardite in natural samples and most of the synthetic mixtures was calculated based on a spike of 9.93 wt.% of synthetic corundum. The amount of lizardite was calculated based on a 9.90 wt.% spike for the two  28  mixtures made with NIST 676a corundum (i.e., 07lsk6r-2 and 07lsk12r). When mineral abundances are calculated assuming a 10 wt.% spike of synthetic corundum the values differ only in the third significant figure. Refinement results were recalculated and renormalized for each sample using the method of Gualtieri (2000). Corundum was chosen over fluorite as the spike phase because it gives more peaks over 3-80° 2θ (CoKα) and has fewer overlapped peaks with the constituent minerals of processed kimberlite. A Brindley radius (Brindley 1945) of 2.5 µm was used to correct for microabsorption in all phases. Scanning electron microscopy, done after milling, on natural samples of processed kimberlite revealed occasional sheets of phlogopite with diameters as great as 40 µm (Fig. 2.4a). In synthetic mixtures, sheets of phlogopite were observed that reached diameters up to 100 µm (Fig. 2.4b). Diameters of other phases in both synthetic and natural samples infrequently exceeded 5 µm. For comparison, laserdiffraction based particle-size analysis was done on two samples at the Norman B. Keevil Institute of Mining Engineering, UBC. Each sample was coned and quartered several times to obtain a representative subsample, which was mixed with distilled water and a small amount of Triton X-100 dispersant. A small, representative aliquot was then taken and analyzed using a Malvern Mastersizer laser diffraction instrument with ultrasonic treatment. Average particle diameters of 3.57 µm (r = 1.79 µm) and 7.83 µm (r = 3.92 µm) were obtained for a natural sample of processed kimberlite fines (05DVK7) and a synthetic mixture (07lsk3r), respectively. Observations made with the scanning electron microscope indicate that phlogopite was the only phase resistant to milling. Rietveld refinement results did not change significantly when the Brindley radius for phlogopite was varied within an order of magnitude of 2.5 µm. As a result, the default radius of 2.5 µm was taken to be representative of the natural samples and synthetic mixtures.  29  Figure 2.4: Backscattered SEM images of (a) 05DVK7 and (b) 07lsk3r. Large sheets of phlogopite remain after 10 minutes grinding with the McCrone micronizing mill.  2.3.3 Quantitative analysis of nesquehonite using the RIR method and the internal standard method Reference intensity ratio (RIR) analyses were done using the normalized RIR method (Chung 1974). The RIR is defined as the ratio of the integrated intensity of the overlapping (101) and (10 1 ) reflections of nesquehonite to the integrated intensity of the (110) peak of synthetic corundum in a 50-50 wt.% mixture. The (101) and (10 1 ) reflections were chosen because they are the most intense in the X-ray diffraction pattern € for nesquehonite. As the more commonly used (113) reflection of corundum overlaps € with reflections from other phases in the processed kimberlite, the (110) reflection was used instead. Peak positions and integrated intensities were determined using Topas Version 3 (Bruker AXS 2004). Peak positions were located using a first-derivative peak search or, in the case of the severely overlapping reference peaks of nesquehonite, located manually based on knowledge of the crystal structure. Profile refinement for reference peaks was done in two ways: using the fundamental parameters approach (Cheary and Coelho 1992) and using a Pseudo-Voigt profile. Peak position, integrated intensity, and Lorentzian crystallite size were refined for each reference peak and the Lorentz polarization correction was applied. RIR values were calculated, using both profile types, as the average of six independent measurements (after Bish and Chipera 1991). RIR values for the (101) and (10 1 ) reflections of nesquehonite relative to the (110) reflection of corundum were 0.36 ± 0.02 and 0.37 ± 0.02 for fundamental parameters and Pseudo€  30  Voigt profiles, respectively. Average peak positions (CoKα), using fundamental parameters, were 15.75° ± 0.01° 2θ [(10 1 ), nesquehonite], 15.84° ± 0.01° 2θ [(101), nesquehonite], and 44.10° ± 0.01° 2θ [(110), corundum]. Average peak positions (CoKα), using a Pseudo-Voigt € profile, were 15.77° ± 0.01° 2θ [(10 1 ), nesquehonite], 15.85° ± 0.01° 2θ [(101), nesquehonite], and 44.10° ± 0.01° 2θ [(110), corundum]. RIR values were also determined by fitting the overlapping (101) € and (10 1 ) reflections of nesquehonite using a single peak. These values were 0.41 ± 0.02 for fundamental parameters and 0.41 ± 0.05 using a Pseudo-Voigt profile. Profile refinements were € visibly and statistically better using the more physically correct two-peak fit for nesquehonite. As such, all data reported fit the overlapping reference reflections of nesquehonite with two peaks. Errors reported on quantitative data obtained using the RIR method are as 1σ. Profile refinements were done on the synthetic mixtures in Series 1 and 2 using both profile types (i.e., fundamental parameters and Pseudo-Voigt) and using first, second, and fourth-order Chebychev polynomials to model the background. A first-order background gave the most accurate and reproducible results with fundamental parameters and Pseudo-Voigt profile refinements. Therefore, profile refinements for natural samples of processed kimberlite also were done using first-order Chebychev polynomials. Nearby reflections from phases other than nesquehonite and corundum were also fitted in order to properly partition the intensity from partially overlapping peaks. A calibration curve was constructed by plotting the ratio of the intensities of the (101) and (10 1 ) reflections of nesquehonite to the (110) reflection of corundum versus the nominal abundance of nesquehonite for Series 1 mixtures (after Alexander and Klug 1948). Intensities extracted using the fundamental parameters approach (as described € above) were used to produce the calibration curve.  31  2.4 Results and discussion 2.4.1 Synthetic mine tailings 2.4.1.1 Rietveld refinement results Rietveld refinements were initially done on data collected from serrated specimens of synthetic kimberlite tailings. The results for Series 1 and 2 mixtures gave significant overestimates for the abundance of lizardite while underestimating the abundances of forsterite and the minor phases. This effect bore a resemblance to the pattern of misestimates that would have resulted from a weighing error for the internal standard (e.g., Gualtieri 2000; Wilson et al. 2006). More precisely, it appeared as though there had been significantly less than 9.93 wt.% crystalline corundum (or 9.90 wt.% for the two samples spiked with NIST 676a) in all of the synthetic mixtures. However, the minor weighing error and negligible sample loss monitored throughout preparation and analysis of these samples could not account for the magnitude of the misestimates. Also, that the same misestimates were observed in all synthetic tailings to a similar extent suggested that the abundance refined for lizardite was in fact the abundance of lizardite plus a considerable amount of amorphous material in each sample. Furthermore, each mixture phase (with the exception of lizardite) was underestimated to a different and internally consistent extent. Under this interpretation, the weighed abundances for forsterite and the minor phases would have been underestimated because the refined results only accounted for the crystalline component of each mineral standard. Previous studies have identified X-ray amorphous material in association with weathering features in feldspars (e.g., Rimsaite 1979; Kawano and Tomita 1996). Banfield et al. (1990) observed amorphous silica and nanoscale patches of a smectite-group mineral in olivine crystals from Abert Rim, Oregon, USA. Additionally, work by Wang et al. (1991) suggests that mantle-derived pyrope may be partially amorphous and contain structurally-bound water. However, it is unlikely that considerable amounts of amorphous and/or nanocrystalline material would be associated with the large, relatively unaltered, single crystals chosen for use as mineral standards.  32  Operating under the assumption that the standard minerals could contain undetected material, specimens of each mineral were prepared with the addition of 10 wt.% synthetic corundum to measure the effect of possible amorphous content on Rietveld refinement results. In serrated specimens (and neglecting refinement error), amorphous contents appeared to range from 6.3 +1.2/-0.7 wt.% of the quartz standard to 24.1 +0.9/-0.6 wt.% of the oligoclase (Table 2.3). The known compositions given in Table 2.1 for the synthetic mixtures were recalculated assuming that these misestimates were due to amorphous content associated with the pure mineral standards. Subtracting the supposed amorphous component from the expected concentration of each phase leads to values that are lower than initially expected for forsterite, nesquehonite, almandinepyrope, calcite, diopside, oligoclase, and quartz. Lizardite concentrations increase when assumed to represent the concentration of lizardite plus the supposed amorphous content associated with the other mineral phases. Rerunning non-serrated specimens of the synthetic processed kimberlites and the high-purity mineral standards from which they were made gives very different results. Rietveld refinement results for non-serrated specimens of standard minerals plus corundum were well within error, typically being within ± 0.5 wt.% and not exceeding ± 1.0 wt.% of nominal abundances. This result indicates that there is no detectable amorphous and/or nanocrystalline material in any of the pure mineral standards and demonstrates that underestimates were caused by surface roughness introduced by serration. Surface roughness is known to reduce the intensities of Bragg peaks, with the reduction increasing in severity toward small scattering angles (e.g., de Wolff 1956; Suortti 1972). This effect would have led to the variable intensity loss for different mineral phases.  33  Table 2.3: Crystallinity of internal standard phases and estimates of crystallinity from serrated samples suffering from surface roughness effects.  NIST 676a α-Al2O3 b  Internal Standard n/a  Crystalline (wt.%) 99.02  Uncertainty a (wt.%) 1.11  Amorphous (wt.%) 0.98  Uncertainty a (wt.%) 1.11  UBC CaF2 b,d  NIST 676a  98.1  +1.0/-1.1  1.9  +1.1/-1.0  Synthetic α-Al2O3 b,e  UBC CaF2  99.3  +1.0/-1.1  0.7  +1.1/-1.0  n/a  unknown  n/a  unknown  n/a  Forsterite (PMM 676) c,f  Synthetic α-Al2O3  91.2  +0.7/-1.1  9.8  +1.1/-0.7  Nesquehonite (synthetic) c,g  Synthetic α-Al2O3  89.1  +1.3/-1.9  10.9  +1.9/-1.3  Almandine-Pyrope (PME 20807) c,f  Synthetic α-Al2O3  88.0  +0.7/-1.1  12.0  +1.1/-0.7  Synthetic α-Al2O3  86.7  +0.7/-1.1  13.3  +1.1/-0.7  Diopside c,h  Synthetic α-Al2O3  88.1  +0.7/-1.1  11.9  +1.1/-0.7  Oligoclase (CMN F80-22-7) c,i  Synthetic α-Al2O3  75.9  +0.6/-0.9  24.1  +0.9/-0.6  Phlogopite (Comet Mine, Québec) c,h  Synthetic α-Al2O3  ~100  n/a  n/a  Quartz c,h  Synthetic α-Al2O3  93.7  +0.7/-1.2  none detected 6.3  Phase  Lizardite c,d  Calcite  a b c d e f g h i  c,h  +1.2/-0.7  Calculated assuming crystallinity of spike phase cannot exceed 100%. Non-serrated specimen. Serrated specimen. From our laboratory collection. From C. Kelaart, Bruker Biosciences Pty Ltd (Australia). From The Pacific Museum of the Earth, University of British Columbia, Vancouver, BC, Canada. Courtesy of I.M. Power, University of Western Ontario, London, ON, Canada. As described by Power et al. (2007). From local mineral dealers, Vancouver, BC, Canada. From The Canadian Museum of Nature, Ottawa, ON, Canada.  Results of quantitative phase analysis on serrated and non-serrated specimens of synthetic processed kimberlites are given in Tables 2.4 and 2.5 and Figures 2.5, 2.6, and 2.7. Much of the deviation in the results for serrated specimens is explained by intensity loss due to surface roughness (which is modelled as “amorphous content” in Fig. 2.5). Refined abundances for non-serrated specimens (Table 2.5 and Fig. 2.6) show significantly less deviation from expected abundances for lizardite, nesquehonite, almandine-pyrope, and oligoclase. Also, refinement results for forsterite and calcite are somewhat improved whereas diopside, phlogopite, and quartz tend to be overestimated. On average, the refinement bias (after Omotoso et al. 2006; Tables 2.4 and 2.5) is nearly halved by running profile refinements on non-serrated specimens (i.e., the average bias is reduced from 19.7 to 11.8). Despite the increase in accuracy, the fit statistics for 34  refinements done on non-serrated specimens suffer as a consequence of the poor fit to phlogopite. Also, large overestimates (as much as ~50% relative) for quartz result from overlap of the (003) reflection of phlogopite and the (011) reflection of quartz. Because no correction was made for preferred orientation of phlogopite, the (003) reflection was inadequately modelled and the intensity was fitted by the structure of quartz. Other major sources of error in Rietveld refinement may include (1) the large increase in relative error for refinements of minor phases and (2) compositional variation in phases subject to solid solution. Relative error is known to increase rapidly for Rietveld refinement of the concentrations of minor phases. Raudsepp et al. (1999) and Dipple et al. (2002) found that relative error increases rapidly for concentrations below 6 wt.% for conditions of data collection used in our laboratory. Although relative errors can be high for minor phases (Fig. 2.7), absolute errors are small and seldom exceed 1 wt.% for non-serrated specimens (Table 2.6). Systematic underestimates of some phases, forsterite and almandine-pyrope in particular, may be due to incorrect compositional information. Electron microprobe (EMP) data were not obtained for the pure mineral standards or for grains of minerals from the processed kimberlite at Diavik. Instead, compositions of the almandine-pyrope and forsterite standards were estimated from energy dispersive X-ray spectroscopy and XRPD data. A large compositional variation in the peridotitic, eclogitic, and crustal garnets found at Diavik (McLean et al. 2007) along with a heterogeneous distribution in the processed kimberlite inhibits attempts at constraining the chemical composition of garnet by Rietveld refinement. The composition of forsterite from kimberlite commonly varies from Fo85 to Fo95 (Mitchell 1986). At Diavik this composition is generally in the vicinity of Fo90 (personal communication, R.C. Brett, UBC). Fo90 and almandine-pyrope with Mg/(Mg+Fe) = 0.70 were used in refinements of synthetic mixtures. Because these values also approximate the composition of the forsterite and the predominantly eclogitic garnet at Diavik (McLean et al. 2007), they were retained for refinement of the natural samples. These values are estimates only, and as a result they introduce further uncertainty into the refinements.  35  Table 2.4: Rietveld refinement results for serrated specimens of synthetic processed kimberlite. Phase  Series 1 07lsk2r  07lsk3r  07lsk4r  07lsk5r  07lsk6r  07lsk6r-2 (NIST)  07lsk7r  07lsk8r  07lsk9r  07lsk10r  07lsk11r  07lsk12r (NIST)  Lizardite + Amorphous  53.2  52.4  53.4  51.1  54.6  54.5  55.9  43.5  43.0  44.9  46.3  44.6  48.7  diff a  8.7  7.9  8.9  6.6  10.1  10.0  11.4  8.9  8.4  10.3  11.7  9.9  14.1  Forsterite  30.2  29.9  29.1  30.1  27.0  25.8  25.0  30.3  30.1  29.8  27.6  26.9  24.6  diff  -4.7  -4.8  -5.4  -3.9  -6.0  -4.2  -5.0  -4.6  -4.7  -4.7  -6.4  -6.1  -5.4  Nesquehonite  b/d b  b/d b  0.5  0.8  1.3  2.7  3.4  b/d b  b/d b  0.3  0.7  1.4  4.4  diff AlmandinePyrope  -0.1  -0.3  0.0  -0.2  -0.7  -2.3  -1.6  -0.1  -0.3  -0.2  -0.3  -0.6  -0.6  4.6  4.2  3.8  3.6  4.3  3.6  4.0  4.0  3.8  3.6  3.6  3.6  3.3  diff  -0.4  -0.8  -1.2  -1.4  -0.7  -1.4  -1.0  -1.0  -1.2  -1.4  -1.4  -1.4  -1.7  Calcite  3.7  4.2  3.7  4.4  3.9  4.0  3.7  4.2  4.2  4.0  3.9  4.1  3.8  diff  -1.8  -1.3  -1.8  -1.1  -1.6  -1.5  -1.8  -1.2  -1.2  -1.4  -1.5  -1.3  -1.5  Diopside  4.9  4.9  4.7  4.2  5.0  3.7  diff  -0.1  -0.1  -0.3  -0.8  0.0  -1.3  Oligoclase  3.7  3.9  4.0  4.0  3.8  2.9  diff  -1.3  -1.1  -1.0  -1.0  -1.2  -2.1  Phlogopite  4.1  4.3  4.8  4.7  4.3  4.4  3.4  4.7  4.6  4.1  4.7  5.8  3.8  diff  -0.9  -0.7  -0.2  -0.3  -0.7  -0.6  -1.6  -0.3  -0.4  -0.9  -0.3  0.8  -1.2  Quartz  4.2  4.9  4.7  5.2  4.6  4.9  4.7  4.7  5.5  4.7  5.1  4.8  4.8  diff  -0.8  -0.1  -0.3  0.2  -0.4  -0.1  -0.3  -0.3  0.5  -0.3  0.1  -0.2  -0.2  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  17.4  15.8  17.8  13.6  20.2  20.0  22.8  17.8  17.7  20.5  23.4  21.5  28.1  Rwp d  7.7  7.1  8.7  8.3  7.8  6.9  7.6  6.6  6.9  6.7  6.9  7.5  7.5  de χ2 f Chebychev order  0.8  0.9  0.9  0.9  1.0  1.0  1.0  1.2  1.1  1.1  1.1  0.8  1.0  1.6  1.5  1.6  1.6  1.4  1.5  1.5  1.3  1.4  1.4  1.4  1.6  1.4  1  1  1  1  1  1  1  2  2  2  2  2  2  Total Σ|diff| (bias) c  a b c d e f  Series 2  07lsk1r  Difference between refined and expected weight-percent concentrations. b/d indicates a concentration below the detection limit. Σ|diff| is the sum of absolute deviations from actual compositions, after the “bias” of Omotoso et al. (2006). Rwp is the weighted pattern index, a function of the least-squares residual. d is the Durbin-Watson statistic, a measure of serial correlation for the least-squares fit. χ2 is the reduced chi-squared statistic for the least-squares fit.  36  Table 2.5: Rietveld refinement results for non-serrated specimens of synthetic processed kimberlite. Phase  a b c d e f  Series 1  Series 2  07lsk1r  07lsk2r  07lsk3r  07lsk4r  07lsk5r  07lsk6r  07lsk6r-2 (NIST)  07lsk7r  07lsk8r  07lsk9r  07lsk10r  07lsk11r  07lsk12r (NIST)  Lizardite + Amorphous  49.3  49.9  43.9  48.6  45.8  50.0  49.8  37.8  37.6  39.6  36.2  39.3  33.6  diff a  4.8  5.4  -0.6  4.1  1.3  5.5  5.3  3.1  2.9  5.0  1.5  4.6  -1.0  Forsterite  30.5  30.9  31.5  29.3  29.6  26.2  25.3  31.6  31.1  29.8  30.9  28.4  26.5  diff  -4.4  -3.9  -3.0  -4.7  -3.4  -3.8  -4.7  -3.3  -3.7  -4.7  -3.1  -4.6  -3.5  Nesquehonite  b/d b  b/d b  1.1  1.1  2.1  3.6  4.0  b/d b  b/d b  0.4  1.2  1.5  5.1  diff AlmandinePyrope  -0.1  -0.3  0.6  0.1  0.1  -1.4  -1.0  -0.1  -0.3  -0.1  0.2  -0.5  0.1  4.9  4.8  6.0  5.1  5.7  4.9  4.5  4.3  5.2  4.7  4.5  4.2  5.5  diff  -0.1  -0.2  1.0  0.1  0.7  -0.1  -0.5  -0.7  0.2  -0.3  -0.5  -0.8  0.5  Calcite  4.4  4.2  4.3  4.6  4.4  4.2  4.2  4.5  4.5  4.3  4.3  4.4  4.4  diff  -1.1  -1.3  -1.2  -0.9  -1.1  -1.3  -1.3  -0.9  -0.8  -1.1  -1.1  -1.0  -1.0  Diopside  5.8  5.5  5.6  5.8  5.5  6.2  diff  0.8  0.5  0.6  0.8  0.5  1.2  Oligoclase  4.8  5.1  4.0  4.3  4.5  4.6  diff  -0.2  0.1  -1.0  -0.7  -0.5  -0.4  Phlogopite  5.2  4.6  5.9  5.0  5.8  5.1  5.9  5.8  5.1  5.8  5.8  6.2  6.5  diff  0.2  -0.4  0.9  0.0  0.8  0.1  0.9  0.8  0.1  0.8  0.8  1.2  1.5  Quartz  5.7  5.6  7.3  6.4  6.6  6.1  6.2  5.4  5.9  5.8  7.1  6.0  7.5  diff  0.7  0.6  2.3  1.4  1.6  1.1  1.2  0.4  0.9  0.8  2.1  1.0  2.5  Total Σ|diff| (bias) c  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  11.3  12.0  9.6  11.3  9.0  13.4  15.0  10.4  9.5  14.3  10.9  14.7  11.7  Rwp d  11.2  11.3  14.5  11.4  13.9  12.0  14.6  10.3  12.1  11.7  15.3  11.6  14.3  de χ2 f Chebychev order  0.3  0.3  0.2  0.3  0.2  0.3  0.2  0.4  0.3  0.3  0.2  0.3  0.2  2.8  2.8  3.7  2.9  3.5  3.0  3.7  2.5  3.0  2.9  3.8  2.9  3.5  2  2  2  2  2  2  2  3  3  3  3  3  3  Difference between refined and expected weight-percent concentrations. b/d indicates a concentration below the detection limit. Σ|diff| is the sum of absolute deviations from actual compositions, after the “bias” of Omotoso et al. (2006). Rwp is the weighted pattern index, a function of the least-squares residual. d is the Durbin-Watson statistic, a measure of serial correlation for the least-squares fit. χ2 is the reduced chi-squared statistic for the least-squares fit.  37  Figure 2.5: Results of Rietveld refinements on serrated specimens of synthetic processed kimberlite. Series 1 indicates 07lsk1r through 6r and Series 2 indicates 07lsk7r through 12r. “Adjusted” results model intensity loss (actually due to surface roughness) as “amorphous content”.  38  Figure 2.6: Results of Rietveld refinements on non-serrated specimens of synthetic processed kimberlite. Series 1 indicates 07lsk1r through 6r and Series 2 indicates 07lsk7r through 12r.  39  Figure 2.7: Relative error on Rietveld refinement results for each mineral phase in the synthetic processed kimberlites. Results for Series 1 and 2 (and serrated versus non-serrated specimens) are distinguished by the colour of the symbols.  Phlogopite in kimberlite is commonly distorted, kink banded, and partially altered (Mitchell 1986). The phlogopite standard used in the synthetic kimberlites was chosen because it was highly pure and suffered from anisotropic peak broadening. Anisotropic peak broadening can result from stacking faults or anisotropic distribution of crystallite size or strain, which can be symptoms of deformation (e.g., Jerome and Mohanty 1979; Dutta and Pradhan 2003; Couvy et al. 2004). This phlogopite was best modelled using the March-Dollase correction (March 1932; Dollase 1986) for preferred orientation on {001} and a sixth-order symmetrized harmonic expansion (Järvinen 1993) to model peak anisotropy (Fig. 2.8a). Refined results for phlogopite, from a sample prepared with an internal standard of 10 wt.% synthetic corundum, were introduced as a starting model into refinements of synthetic mixtures. The degree of preferred orientation on {001} was refined, but the coefficients of the symmetrized harmonics were held constant. Although this model gave the best fit to the pattern for phlogopite (when tested on synthetic  40  kimberlite mine tailings), severe overlap of the major reflections of phlogopite with other phases and the use of symmetrized harmonics commonly led to overestimates of several weight percent (Fig. 2.8b). Overestimates worsened when symmetrized harmonics were used to model both preferred orientation and peak anisotropy. Preferred orientation in phlogopite was also modelled using the March-Dollase correction without using symmetrized harmonics to model peak anisotropy (Fig. 2.8c). Again, because of severe overlap with the major reflections of this phase, the model failed to adequately fit the phlogopite. Extreme underestimates (of as much as 4.7 wt.% on a 5.0 wt.% abundance of phlogopite) resulted from overestimates of the degree of preferred orientation. Because of the difficulty in correctly modelling preferred orientation and peak anisotropy for a low abundance of phlogopite in a pattern suffering from severe overlap of peaks, coupled with the difficulty in reducing its particle size, no corrections were applied to phlogopite in the final refinements (Fig. 2.8d). Although the refinement statistics suffered (particularly for data collected from non-serrated specimens), the fit to phlogopite was better constrained and refinement results were more accurate and more consistent. Other possible sources of error include fluorescence from Cr and Mn-bearing phases (using Co radiation), and difficulty in modelling minor phases due to overlap of reflections in the XRPD patterns. In addition, Wilson et al. (2006) observed that the use of a peaks phase to model a serpentine mineral can cause low-abundance mineral phases to be underestimated in Rietveld refinements. They noted that large misestimates are likely to occur for minerals at abundances below approximately 5 wt.% due to limitations in the fundamental parameters approach, difficulty in modelling preferred orientation, and the tendency of complex systems of peaks to dominate the analysis. These effects, together with the number of low-abundance phases in the synthetic samples, may contribute to systematic overestimates for lizardite.  41  Figure 2.8: Modeling phlogopite in a synthetic kimberlite (07lsk3r). (a) Phlogopite was best fit using the March-Dollase correction for preferred orientation on {001} and a sixth-order symmetrized harmonic expansion to model peak anisotropy; (b) the same model applied to a serrated specimen of 07lsk3r; (c) modeling phlogopite with only the March-Dollase correction, and (d) modeling phlogopite with no corrections for preferred orientation or anisotropic peak shape.  42  Nesquehonite was not detected at abundances ≤ 0.25 wt.% in the patterns collected for samples 07lsk1r, 07lsk2r, 07lsk7r, and 07lsk8r. The relative error on refined values for the amount of nesquehonite remained consistent for detectable abundances between 0.50 wt.% and 5.0 wt.%. A maximum relative error of approximately 40% was observed for abundances in this range and can be expected for natural samples. This corresponds to an absolute error of 0.2 wt.% for a sample containing 0.50 wt.% nesquehonite and 2.0 wt.% for a sample containing 5.0 wt.% nesquehonite. No significant overall difference in accuracy was observed for Rietveld refinements of Series 1 and 2 samples and the inclusion of additional, low abundance phases (i.e., diopside and oligoclase) in the Series 2 samples had no adverse effect on refinement results. Also, the choice of NIST 676a corundum or synthetic corundum did not have a significant effect on Rietveld refinement results. However, the disparity in the sizes of these two sample subsets makes the effect of differing internal standards difficult to assess. Despite the significant relative errors for nesquehonite, the absolute errors are generally small and are in keeping with results for low abundance phases from previous studies using the Rietveld method for quantitative phase analysis (i.e., Raudsepp et al. 1999; Omotoso et al. 2006; Wilson et al. 2006). 2.4.1.2 Results of quantitative RIR and the internal standard method Quantitative RIR results for nesquehonite in the synthetic processed kimberlites are given in Table 2.6 and Figure 2.9. No significant difference in accuracy was observed for RIR results using fundamental parameters and Pseudo-Voigt peak fitting. As with Rietveld refinement results (Fig. 2.6), abundances from RIR deviate somewhat from the ideal values. However, RIR results were generally closer to the expected, nominal values than those obtained by Rietveld refinement (Fig. 2.9). Results for the two samples (07lsk6r-2 and 07lsk12r) that used NIST 676a corundum as the spike phase were significantly worse than for those that used synthetic corundum. The (110) reflection of the NIST 676a corundum was less intense than the same reflection in the synthetic corundum, relative to the reference peaks of nesquehonite. Because the reference intensity ratio was determined using synthetic corundum, the abundance of nesquehonite was overestimated in the samples spiked with  43  the NIST corundum standard. Even though the results worsened using the NIST standard, the absolute errors on the abundance of nesquehonite were on the same order as those obtained using the Rietveld method. These results could be improved by measuring and using a RIR for nesquehonite that was determined relative to the NIST 676a corundum. Unfortunately, there was insufficient nesquehonite to make a second 50-50 wt.% mixture to determine this RIR. Table 2.6: Results of Reference Intensity Ratio measurements on synthetic processed kimberlite. Sample Name  07lsk1r  Nominal (wt.%) 0.100  diff a 07lsk2r  0.250 diff  Series 1  07lsk3r  0.500 diff  07lsk4r  1.000 diff  07lsk5r  2.000 diff  Series 2  07lsk6r  a  diff 07lsk6r-NIST diff 07lsk7r diff 07lsk8r diff 07lsk8r-front diff 07lsk9r diff 07lsk9r-front diff 07lsk10r diff 07lsk11r diff 07lsk12r-NIST diff  5.000 5.000 0.100 0.250 0.250 0.500 0.500 1.000 2.000 5.000  Fundamental Parameters  Pseudo-Voigt  Xnesq, (wt.%)  σ(Xnesq) (wt.%)  Xnesq, (wt.%)  σ(Xnesq) (wt.%)  0.096 -0.004 0.081 -0.169 0.430 -0.070 0.843 -0.157 2.010 0.010 4.824 -0.176 5.832 0.832 0.013 -0.087 0.041 -0.209 0.090 -0.160 0.216 -0.284 0.259 -0.241 0.775 -0.225 1.865 -0.135 6.019 1.019  0.005  0.092 -0.008 0.076 -0.174 0.425 -0.075 0.793 -0.207 1.945 -0.055 4.905 -0.095 5.898 0.898 0.012 -0.088 0.041 -0.209 0.087 -0.163 0.205 -0.295 0.250 -0.250 0.722 -0.278 1.710 -0.290 5.829 0.829  0.005  0.004 0.024 0.047 0.112 0.268 0.325 0.001 0.002 0.005 0.012 0.014 0.043 0.104 0.335  0.004 0.023 0.042 0.103 0.261 0.314 0.001 0.002 0.005 0.011 0.013 0.038 0.091 0.310  Difference between refined and expected weight-percent concentrations.  44  Figure 2.9: RIR and Rietveld refinement results for nesquehonite. RIR results are given for Fundamental Parameters (FP) and Pseudo-Voigt (PV) peak fitting.  45  Series 1 and 2 mixtures containing nesquehonite at abundances ≥ 0.50 wt.% gave maximum relative uncertainties of ~20%. Accuracy began to decrease dramatically for concentrations ≤ 0.25 wt.% nesquehonite for the Series 1 mixtures and ≤ 0.50 wt.% for the Series 2 mixtures. The RIR result for the Series 1 sample, 07lsk2r (which contained 0.25 wt.% nesquehonite), was underestimated by 68% relative. Underestimates in the vicinity of 50% to 90% relative were observed for samples 07lsk7r, 07lsk8r, and 07lsk9r in Series 2. The reference peaks of nesquehonite were only just detectable at > 3σ above the intensity of the background for sample 07lsk7r. XRPD data were recollected on specimens of 07lsk7r, 07lsk8r, and 07lsk9r that were front-loaded into cavity mounts and pressed flat with a smooth sheet of glass to maximize preferred orientation of crystallites. The estimates for the concentration of nesquehonite in samples 07lsk8r and 07lsk9r were somewhat improved when the data were recollected in this way (Table 2.6 and Fig. 2.9). The nesquehonite peaks were not detected in the data for the front-loaded specimen of 07lsk7r. It is likely that the underestimates on abundances less than 0.50 wt.% are the result of decreased preferred orientation of nesquehonite. At higher abundances, the elongate crystals of nesquehonite tend to orient themselves along [010], giving rise to the preferred orientation observed on {101}. At lower abundances, crystals of nesquehonite could be packed more randomly into a powdered specimen. Also, underestimates of nesquehonite are more severe for the samples in Series 2, which contain more phases and a lower abundance of (preferentially oriented) phyllosilicate minerals than the samples in Series 1. It appears that at abundances less than 0.50 wt.%, the ability to induce a similar degree of preferred orientation to that seen at abundances of 5.0 wt.% or 50 wt.% (as in the RIR standard mixture), is compromised in the samples from Series 1 and particularly in the samples from Series 2. In some systems, it is possible to use a multireflection RIR method to reduce the impact of preferred orientation on quantitative phase analysis (i.e., Chipera and Bish 1995). Unfortunately, the severe overlap of reflections in XRPD patterns for processed kimberlite precludes this option. Thus, we consider the practical lower limit to quantification of nesquehonite to be in the range of 0.25-0.50 wt.% using this RIR method. Below this limit, RIR quantitative results for nesquehonite should be considered as an order of magnitude estimate at best.  46  At abundances ≥ 0.50 wt.% the RIR method gives results for nesquehonite that are generally more accurate than the Rietveld refinement results (Figs. 2.6 and 2.9). RIR and Rietveld results were typically characterized by relative uncertainties less than 20% and 40%, respectively for nesquehonite in abundances ≥ 0.50 wt.%. For this system of minerals, both the Rietveld and RIR method failed to accurately quantify nesquehonite at abundances below approximately 0.50 wt.%. With the Rietveld method, this is largely a consequence of counting statistics. Rietveld refinement results could be improved and detection limits lowered if a significantly longer time per step was used to acquire XRPD patterns. With the RIR method, the relatively low accuracy for abundances < 0.50 wt.% was likely a consequence of sometimes unpredictable decreases in the preferred orientation of nesquehonite at low abundances. As a readily available alternative to using RIR or the Rietveld method to measure nesquehonite abundances < 0.50 wt.%, a calibration curve was produced. This was done by plotting the ratio of intensities of the reference peaks for nesquehonite and corundum versus the nominal abundance of nesquehonite for the 6 mixtures from Series 1 that were prepared with synthetic corundum (after Alexander and Klug 1948). The Series 1 mixtures were selected for this purpose because they represent the baseline composition of kimberlite mine tailings and have no peaks in the vicinity of the references peaks of nesquehonite. A best linear fit to these data is given in Figure 2.10a. The amounts of nesquehonite in the Series 2 mixtures prepared with synthetic corundum (which are most similar to the mine tailings at Diavik) were calculated using this calibration curve as a test of its accuracy. The use of a calibration curve for mixtures with 1 wt.% or less nesquehonite (Fig. 2.10b) mitigates the systematic underestimates given by the RIR method (Fig. 2.10c). At abundances ≥ 1.00 wt.%, nesquehonite is underestimated by less than 15% relative (in two samples). At lower abundances, maximum underestimates of approximately 50% relative are observed (in three samples). This represents a significant improvement over maximum underestimates of 90% observed using the RIR method. It should be noted that, because of the small but significant underestimates of nesquehonite calculated for samples from Series 2 using the calibration curve for Series 1 data, the amount of nesquehonite in Diavik tailings may also be slightly underestimated.  47  Figure 2.10: Calibration curve for nesquehonite in Series 1 mixtures (a), the calibration curve applied to Series 2 mixtures (b), and FP RIR results (c). An expanded view of Fig. 2.9a is shown in (c) and error bars are 1σ, propagated from repeated measurement of the RIR standard.  48  Based on our results, the abundance of nesquehonite can be measured using either the RIR method or the Rietveld method when present at ≥ 0.50 wt.%. At abundances < 0.50 wt.%, the use of a calibration curve is likely to give more accurate results. 2.4.2 Natural mine tailings Twenty samples of processed kimberlite from the Diavik Diamond Mine were analyzed using the Rietveld method, the RIR method of Chung (1974), and a calibration curve according the internal standard method of Alexander and Klug (1948). Analyses were done on data collected from non-serrated specimens. Samples 05DVK4, 06DVK35, and 06DVK36 are samples of fine processed kimberlite covered by a thin, efflorescent film of nesquehonite (Group 1 crusts). Sample 05DVK7 is from a Group 2 crust from the coarse PKC. Samples 06DG1-1, 06DG3-1, 06DG7-1, and 06DG9-1 are 10-cm long core samples, which were collected at the corners of a 5 m x 5 m grid, from the surface of the fine PKC. Samples 06DG1-3, 06DG3-3, 06DG7-3, and 06DG9-3 are 10-cm long core samples collected from a depth of 100 cm below the corresponding samples from the surface of the fine PKC (e.g., 06DG1-3 was collected from an interval 95 to 105 cm beneath 06DG1-1). 06DVK33-1, 06DVK33-3, 06DVK42-1, 06DVK42-3, 06DVK46-1, 06DVK46-3, 06DVK53-1, and 06DVK53-3 are samples collected from four cores (i.e., cores 06DVK33, 06DVK42, 06DVK46, and 06DVK53) taken at four corners of the fine PKC. Again, core samples were collected from the surface and at a depth of 100 cm. Sampling locations are given in Figure 2.3 and the results of quantitative phase analysis are given in Table 2.7. Rietveld refinement results indicate that samples of fine processed kimberlite from Diavik contain between 40.0 and 76.2 wt.% lizardite and between 12.3 and 46.3 wt.% forsterite. The combined abundance of lizardite and forsterite varies from 76.1 to 93.3 wt.%. Calcite was present at abundances between 1.9 and 4.1 wt.%. Vermiculite, which was detected in 18 samples, was identified by XRPD using a smear-mounted specimen that had been solvated with glycerol for 12 hours at 60˚C, according to the method of Moore and Reynolds (1997). Following solvation, no change in the position of the basal reflection of this clay phase was observed, suggesting that it is vermiculite and discounting the presence of a pure smectite group mineral. However, this does not rule  49  out the possibility that some of the samples from Diavik may contain interstratified serpentine-smectite of the sort described by Sakharov et al. (2004). Plagioclase feldspar, diopside, and quartz were present at abundances less than 5 wt.%. Pyrope-rich garnet and phlogopite were present at abundances of 0-6.4 wt.% and 1.6-3.5 wt.%, respectively. Based on results for the synthetic mixtures, maximum relative errors of approximately 50% can be expected on refined abundances for these minor phases. At higher abundances, in the vicinity of 5 wt.% and above, relative errors between 5% and 20% are anticipated. Nesquehonite was present below the detection limit for Rietveld refinement (approximately 0.50 wt.%) in all samples of processed kimberlite. The reference peaks of nesquehonite were detected at > 3σ above background intensity in three samples from the fine PKC (05DVK4, 06DVK36, and 06DVK53-1) and one sample from the coarse PKC (05DVK7) using the long acquisition time for the RIR method. It was not detected in sample 06DVK35, despite its presence having been confirmed by visual inspection and by XRPD analysis of high-graded crust material. Nesquehonite may have been detected at higher than 2σ above the background intensity in sample 06DG3-3. RIR results for the four samples in which nesquehonite was detected give abundances between 0.039 and 0.063 wt.% nesquehonite (Table 2.7). Considering the difficulty in maintaining a relatively constant degree of preferred orientation in nesquehonite and the underestimates that ensued in the synthetic mixtures, it is very likely that the values obtained for the natural samples are underestimated to a comparable extent. In fact, results of a similar magnitude were obtained for synthetic mixtures with 0.10 and 0.25 wt.% nesquehonite (Table 2.6, Figs. 2.9 and 2.10). Use of the calibration curve suggests that nesquehonite is present between approximately 0.11 wt.% and 0.14 wt.% in the four natural samples from Diavik with measurable abundances (Table 2.7).  50  Table 2.7: Results of quantitative phase analysis of natural samples of processed kimberlite, renormalized to exclude corundum. Phase  05DVK4  05DVK7  06DG1-1  06DG1-3  06DG3-1  06DG3-3  06DG7-1  06DG7-3  06DG9-1  surface  surface  surface  100  surface  100  surface  100  surface  100  Lizardite + amorphous Forsterite  40.0  61.0  72.6  63.8  71.2  61.9  76.2  50.3  75.5  51.4  36.8  15.1  17.5  23.4  18.1  25.4  12.3  35.2  14.5  35.7  Nesq. Rietveld  n/d a  n/d  n/d  n/d  n/d  n/d  n/d  n/d  n/d  n/d  Nesq. RIR  0.058  0.043  n/d  n/d  n/d  n/d  n/d  n/d  n/d  n/d  Nesq. Internal Std.  0.133  0.118  n/d  n/d  n/d  n/d  n/d  n/d  n/d  n/d  Nesq. Present? b  yes  yes  n/d  yes  n/d  yes  n/d  n/d  n/d  n/d  Almandine-Pyrope  6.4  1.9  1.6  2.3  1.5  2.9  1.3  4.4  1.0  4.3  Calcite  2.2  1.9  2.7  2.6  2.3  2.3  2.6  2.5  2.6  2.2  Diopside  2.6  3.2  1.6  2.0  2.1  1.7  1.8  1.9  1.8  2.0  Phlogopite  2.9  2.7  1.6  2.5  1.9  2.0  2.1  2.7  2.1  1.6  Plagioclase  3.4  3.7  1.1  1.3  1.4  1.5  1.5  1.2  0.8  1.0  Quartz  2.7  2.0  0.9  1.2  0.8  1.3  0.9  1.1  0.9  1.0  Vermiculite  2.9  8.4  0.4  0.9  0.8  1.0  1.2  0.6  0.8  0.7  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  Rwp c  5.1  4.8  4.9  4.9  4.8  4.7  5.0  5.1  5.1  4.5  dd  1.0  1.1  1.1  1.1  1.1  1.1  1.0  1.0  1.0  1.2  χ2 e  1.5  1.4  1.4  1.4  1.4  1.3  1.5  1.5  1.5  1.3  NP, Computed f  51.0  47.5  57.7  55.5  55.3  53.2  56.1  54.6  55.7  52.1  NP, Nesq. g  1.0  0.9  0  0  0  0  0  0  0  0  depth (cm)  Total (Rietveld)  a b c d e f g  06DG9-3  n/d indicates that the phase was not detected. Nesquehonite determined to be present if the (101) and (10 1 ) peaks are detected at > 3σ above background in smear-mounted specimens. Rwp is the weighted pattern index, a function of the least-squares residual. d is the Durbin-Watson statistic, a measure of serial correlation for the least-squares fit. χ2 is the reduced chi-squared statistic for the least-squares fit. NP, Computed is calculated using Rietveld refinement results and abundances of nesquehonite determined from the calibration curve. Units are kg CaCO3/t. NP, Nesq. is the contribution of nesquehonite €to the total computed NP. Units are kg CaCO3/t.  51  Table 2.7 (continued): Results of quantitative phase analysis of natural samples of processed kimberlite, renormalized to exclude corundum spike. Phase  06DVK33-1  06DVK33-3  06DVK35  06DVK36  06DVK42-1  06DVK42-3  06DVK46-1  06DVK46-3  06DVK53-1  surface  100  surface  surface  surface  100  surface  100  surface  100  Lizardite + amorphous Forsterite  63.3  65.6  65.1  67.7  47.0  48.5  55.8  55.8  65.8  50.3  27.2  24.0  19.6  15.6  46.3  44.3  32.8  30.7  20.1  33.8  Nesq. Rietveld  n/d a  n/d  n/d  n/d  n/d  n/d  n/d  n/d  n/d  n/d  Nesq. RIR  n/d  n/d  n/d  0.039  n/d  n/d  n/d  n/d  0.063  n/d  Nesq. Internal Std.  n/d  n/d  n/d  0.114  n/d  n/d  n/d  n/d  0.138  n/d  Nesq. Present? b  n/d  n/d  yes  yes  yes  yes  yes  yes  yes  yes  Almandine-Pyrope  1.6  2.2  0.9  2.1  2.0  1.7  4.0  Calcite  2.4  2.6  4.0  4.1  2.5  3.1  2.5  2.0  Diopside  1.8  1.6  2.2  1.9  1.8  Phlogopite  1.8  1.6  2.9  3.5  1.8  2.5  2.8  2.1  Plagioclase  1.1  0.9  2.1  2.4  1.7  1.4  2.5  1.8  Quartz  0.8  1.0  2.2  2.2  1.5  1.2  2.5  1.7  0.6  2.0  3.6  0.1  0.3  0.5  2.3  1.4  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  100.0  Rwp c  4.9  4.7  4.8  5.3  4.9  4.6  4.7  4.9  4.9  4.7  dd  1.1  1.1  1.0  0.9  1.0  1.1  1.2  1.1  1.1  1.1  χ2 e  1.4  1.4  1.4  1.5  1.4  1.4  1.4  1.4  1.4  1.4  54.5  56.2  69.1  71.0  58.0  63.4  55.2  50.3  69.7  57.2  0  0  0  0.8  0  0  0  0  1.0  0  depth (cm)  Vermiculite Total (Rietveld)  NP, computed NP, Nesq. g a b c d e f g  f  2.0  1.6 0.5  06DVK53-3  4.2 3.9  2.7 1.9  n/d indicates that the phase was not detected. Nesquehonite determined to be present if the (101) and (10 1 ) peaks are detected at > 3σ above background in smear-mounted specimens. Rwp is the weighted pattern index, a function of the least-squares residual. d is the Durbin-Watson statistic, a measure of serial correlation for the least-squares fit. χ2 is the reduced chi-squared statistic for the least-squares fit. NP, Computed is calculated using Rietveld refinement results and abundances of nesquehonite determined from the calibration curve. Units are kg CaCO3/t. NP, Nesq. is the contribution of nesquehonite €to the total computed NP. Units are kg CaCO3/t.  52  Additional data for nesquehonite were collected on smear-mounted specimens of the 20 natural samples. Samples were smear-mounted in order to induce preferred orientation of nesquehonite. The (101) and (10 1 ) reflections of nesquehonite were detected in 11 of the 20 samples using a counting time of 40 s/step. RIR data for the reference peaks of nesquehonite were acquired using a counting time of 120 s/step. In € sample 06DVK53-1, the reference peaks for nesquehonite are more intense in the pattern collected in 2 hours and 4 minutes on a smear-mounted specimen than in the pattern from a back-loaded specimen collected in 6 hours and 12 minutes (Fig. 2.11). It can be inferred from Figure 2.11 that the detection limit is significantly lower for smearmounted specimens. Several analytical options are available to improve detection and quantification of nesquehonite. One such option is to use even longer counting times on the specimens loaded into cavity mounts and to measure, empirically, the decrease in preferred orientation with decreasing abundance of nesquehonite. This would need to be done using replicate analyses of standard mixtures. The time and cost involved in this procedure may disqualify its development and routine use. A second option is to induce preferred orientation in standard mixtures and natural samples by smear-mounting onto zero-diffraction quartz plates. This method would improve detection of nesquehonite, but the variation in degree of preferred orientation and the effect of surface roughness would need to be assessed. Also, the very small amounts of sample needed to prepare this type of specimen may result in measurement errors from insufficient homogenization.  53  Figure 2.11: XRPD patterns collected for 06DVK53-1. The reference peaks of nesquehonite are relatively less intense in the RIR pattern (2) than in the pattern acquired from a smear-mounted (oriented) aliquot (1). The left-most peak in (1) and (2) and the only peak in (3), (4), and (5) is from the (101) and (10 1 ) reflections of nesquehonite. The peak to the right in (1) and (2) is from plagioclase.  € 2.5 Implications for neutralization potential and carbon dioxide sequestration Jambor et al. (2007) have demonstrated that the neutralization potential (NP) of a geological sample can be computed from Rietveld refinement results. In their study, they calculated the NP values of geological materials using appropriate values for NP of the individual minerals in each sample. These values were drawn from a library of NP data for pure mineral phases that had been determined using the Sobek method (Sobek et al. 1978). Most of the NP results computed by Jambor et al. (2007) were lower than the measured values, but the authors concluded that they could be used as a baseline estimate of NP. NP values were calculated using the method of Jambor et al. (2007) for the 20 samples of processed kimberlite from Diavik (Table 2.7). Values used for the NP of individual minerals are from Jambor et al. (2002, 2007) or (in the case of nesquehonite)  54  are calculated from Lawrence and Scheske (1997) and are listed in Table 2.2. The mean value of NP for the natural samples was 57.0 kg CaCO3/t. The minimum and maximum values were 47.5 kg CaCO3/t and 71.0 kg CaCO3/t, respectively. The Canadian Department of Indian Affairs and Northern Development (1993) considers mine tailings to be acid generating if they have a ratio of NP to maximum potential acidity (MPA) of 1.2:1 or less. A ratio of 3:1 or less is used to identify waste rock as acid generating. From the calculated NP values, pyrite (which has been observed in the coarse PKC) would need to be present at abundances between 2.4 and 3.5 wt.% for acid mine drainage to be considered a risk in the processed kimberlite at Diavik (Fig. 2.12). Pyrite has not been detected with XRPD in natural samples from either the fine or coarse PKC to a detection limit of approximately 0.5 wt.%. However, gypsum was observed as an efflorescent precipitate at the surface of the processed kimberlites and Baker et al. (2003) have reported gypsum precipitating from dissolved pyrite and calcite in column experiments on freshly processed kimberlite from Diavik. It is not unlikely that some amount of the pyrite initially present in the processed kimberlite had dissolved and been neutralized by calcite to precipitate gypsum. Baker et al. (2003) reported the mean values of NP and MPA for freshly processed kimberlites as 196 ± 131 kg CaCO3/t and 14.8 ± 6.7 kg CaCO3/t, respectively (replotted in Fig. 2.12). The NP values computed as part of this study are considerably lower than those reported by Baker et al. (2003). Some of this decrease may be due to neutralization of calcite; however, the computed NP values should be considered conservative estimates only. Although the calculated NP results are underestimated relative to those of Baker et al. (2003), they can be used (1) to establish whether acid generation may be a risk in localized regions of the tailings or waste rock facilities and (2) to select potentially high-risk samples for kinetic testing.  55  Figure 2.12: Results for neutralization potential (NP) and maximum potential acidity (MPA) of processed kimberlite are replotted from Baker et al. (2003). Ranges for calculated NPs from this study can be used to estimate the minimum threshold at which acid mine drainage could become a concern in the fine PKC at Diavik (i.e., 2.4 wt.% FeS2).  It is important to note that the development of nesquehonite in the tailings from Diavik may contribute to their potential to neutralize acid mine drainage. Nesquehonite, in the four samples from Diavik with abundances measurable using the calibration curve, may contribute NP on the order of 0.8 to 1.0 kg CaCO3/t, an amount that constitutes 1% to 2% of the NP of these samples (Table 2.7). Although this increase in NP is not large, it is significant considering that it is due to mineralization of 0.11 to 0.14 wt.% of nesquehonite. Precipitation of nesquehonite is generally induced by evaporation (e.g., Suzuki and Ito 1974; Jull et al. 1988; Grady et al. 1989; McLean et al. 1997; Kloprogge et al. 2003; Power et al 2007). Correspondingly, efflorescent films of nesquehonite are typically found in dry, subaerial conditions in the fine PKC at Diavik and nesquehonite 56  detected at depth within the processed kimberlite likely represents buried efflorescences from former surfaces in the containment facility. Acero et al. (2007) have used column experiments to demonstrate that efflorescent minerals first precipitate at the surface of unsaturated mine tailings and later begin to infiltrate pore spaces in the underlying material. They attribute this process to the withdrawal of the evaporation front within the column experiments as the mine tailings were drying. Also, Wilson et al. (2009) have described abundant carbon mineralization, including thick crusts (> 1 mm) of efflorescent nesquehonite, in the tailings pile at the Clinton Creek chrysotile mine, Yukon Territory, Canada. Differences in climate and mineralogy of mine residues may play a role in determining the extent to which mineral carbonation is occurring at these sites. However, given that nesquehonite precipitates by evaporation, a further control may be the depth of the vadose zone in a tailings storage facility. The tailings pile at Clinton Creek is located on the slope of a mountainside. Wilson et al. (2009) have excavated within the pile and have described the tailings to be unsaturated by water to a depth of at least 2 m. Contrastingly, most of the volume of the fine processed kimberlite at Diavik is located several metres below the surface of the central pond and is saturated with process water. In addition, samples from several cores of processed kimberlite beached around the perimeter of the pond were located beneath the water table. Values for modal abundance obtained from quantitative phase analysis can be used to estimate the amount of carbon dioxide that is crystallographically bound in nesquehonite and other carbonate minerals. Wilson et al. (2006) have applied this method to bulk samples of chrysotile mine tailings to estimate the amount of secondary carbonate mineralization occurring per kilogram of tailings material. From these results, they were able to estimate the amount of CO2 bound within piles of chrysotile mine tailings. At Diavik, this sort of stoichiometric analysis is made possible by the very limited substitution of other divalent cations for Mg in nesquehonite (assessed using energy dispersive X-ray spectroscopy and scanning electron microscopy). Nesquehonite is present below the quantification limit in all samples taken from at depth within the fine PKC. Also, nesquehonite is unlikely to form in tailings saturated with water. A best estimate for the amount of nesquehonite precipitating in the fine PKC at Diavik can be made (1) using the quantitative results calculated from the calibration curve (for three  57  samples), (2) assuming that samples containing detectable but unquantifiable nesquehonite contain ~0.075 wt.% of this mineral (half way between limits for detection and quantification of nesquehonite, eight samples), and (3) assuming that nesquehonite is present at 0 wt.% in the samples in which it was below detection (eight samples). No estimate was made for CO2 sequestration in the coarse PKC because only one datum was available (for 05DVK7). Eleven of the 19 samples analyzed from the fine PKC were taken from the nesquehonite-rich upper 10 cm of this storage facility. Based on our quantitative XRPD results, between 12 and 24 t of CO2 (depending on packing density) are concentrated within the upper 10 cm of dry, processed kimberlite in the fine PKC. Our results for the deep samples suggest that an additional 1,800 t of CO2 may be stored within nesquehonite preserved at depth. The Diavik Diamond Mine produces approximately 150,000 tonnes of CO2 equivalent greenhouse gas emissions each year (Mining Association of Canada 2007). Relative to this value for annual greenhouse gas emissions, a total of approximately 1,800 t of CO2 trapped within the fine PKC does not seem considerable. In spite of this, it is significant that so much CO2 has been sequestered within minerals at a site like Diavik, where the extremely cold climate and tailings management  procedures  may  have  severely  inhibited  carbon  mineralization.  Consequently, detection and quantification of trace carbon mineralization at Diavik makes it possible to establish an approximate baseline for CO2 sequestration in ultramafic mine tailings that can be applied to subsequent studies. In general, a tailings storage facility has two purposes in an active mining environment: To act as (1) a permanent storage area for tailings and (2) a temporary storage facility for process water that is due to be either recycled or released into the surrounding environment. Draining the process water from the fine PKC and relocating it to a separate collection pond could enhance precipitation of nesquehonite. Unfortunately, this is not an economical option for an operational mine and would increase Diavik’s footprint on the tundra at East Island. However, this strategy could represent a feasible option for promoting CO2 sequestration within minerals at less environmentally sensitive localities and under more temperate conditions.  58  2.6 Conclusions Trace amounts of secondary carbonate minerals in kimberlite mine tailings can be quantified accurately using either the Rietveld method or the method of normalized reference intensity ratios for abundances ≥ 0.5%. A calibration curve, constructed using the internal standard method, provides more accurate results for abundances < 0.5%. Surface roughness effects, when combined with structureless pattern fitting and the use of an internal standard, can produce a pattern of systematic misestimates in Rietveld refinement results that is similar to that caused by X-ray amorphous content. As a result, surface roughness should not be induced as a measure to compensate for severe preferred orientation of phyllosilicates in specimens intended for quantitative phase analysis. It may be useful to promote the formation of secondary carbonate minerals at active mines as (1) a passive method for trapping and storing atmospheric CO2 in sulphide-poor, alkaline tailings or (2) as a defense against acid mine drainage in more sulphide-rich, alkaline or intermediate tailings. At Diavik, nesquehonite was quantifiable in 20% of the samples analyzed (generally surface samples from the fine and coarse PKC) at an abundance of 0.11 wt.% to 0.14 wt.%. Nesquehonite was detectable in another 40% of the samples that were analyzed to a limit of approximately 0.05 wt.%. Based on these results, on the order of 1.8 x 103 tonnes of CO2 may be trapped in the tailings at Diavik. Although nesquehonite is only present at trace abundance in the tailings at Diavik, it may provide 1-2% of the neutralization potential of tailings in some regions of the PKC.  59  2.7 References Acero, P., Ayora, C., and Carrera, J. (2007) Coupled thermal, hydraulic and geochemical evolution of pyritic tailings in unsaturated column experiments. Geochimica et Cosmochimica Acta, 71, 5325-5338. Al, T.A., Martin, C.J., and Blowes, D.W. (2000) Carbonate-mineral/water interactions in sulfide-rich mine tailings. Geochimica et Cosmochimica Acta, 64, 3933-3948. Ahn, Y.P., Kim, B.H., and Ishizowa, N. (1986) Structure refinements of solid solutions in the system CaO-MgO-2SiO2-Al2O3. Journal of the Korean Ceramic Society, 23, 25-34. Alexander, L. and Klug, H.P. (1948) Basic aspects of X-ray absorption in quantitative diffraction analysis of powder mixtures. Analytical Chemistry, 20, 886-889. Armbruster, T., Buergi, H.B., Kunz, M., Gnos, E., Broennimann, S., and Lienert, C. (1990) Variation of displacement parameters in structure refinements of low albite. American Mineralogist, 75, 135-140. Armbruster, T., Geiger, C.A., and Lager, G.A. (1992) Single-crystal X-ray structure study of synthetic pyrope-almandine garnets at 100 and 293 K. American Mineralogist, 77, 512-521. Baker, M.J., Blowes, D.W., Logsdon, M.J., and Jambor, J.L. (2003) Environmental geochemistry of kimberlite materials: Diavik Diamonds Project, Lac de Gras, Northwest Territories, Canada. Exploration and Mining Geology, 10, 155-163. Banfield, J.F., Veblen, D.R., and Jones, B.F. (1990) Transmission electron microscopy of subsolidus oxidation and weathering of olivine. Contributions to Mineralogy and Petrology, 106, 110-123. Bish, D.L. and Chipera, S.J. (1991) Detection of trace amounts of erionite using X-ray powder diffraction: erionite in tuffs of Yucca Mountain, Nevada, and Central Turkey. Clays and Clay Minerals, 39, 437-445. Bish, D.L. and Howard, S.A. (1988) Quantitative phase analysis using the Rietveld method. Journal of Applied Crystallography, 21, 86-91. Bish, D.L. and Post, J.E. (1993) Quantitative mineralogical analysis using the Rietveld full-pattern fitting method. American Mineralogist, 78, 932-940.  60  Brindley, G.W. (1945) The effect of grain or particle size on X-ray reflections from mixed powders and alloys, considered in relation to the quantitative determination of crystalline substances by X-ray methods. Philosophical Magazine, 36, 347-369. Broecker, W.S. (2007) CO2 arithmetic. Science, 315, 1371. Brown, A.S., Spackman, M.A., and Hill, R.J. (1993) The electron distribution in corundum. A study of the utility of merging single-crystal and powder diffraction data. Acta Crystallographica Section A: Foundations of Crystallography, A49, 513-527. Bruker AXS (2004) Topas V. 3.0: General Profile and Structure Analysis Software for Powder Diffraction Data. Bruker AXS, Germany. Cheary, R.W. and Coelho, A.A. (1992) A fundamental parameters approach to X-ray line-profile fitting. Journal of Applied Crystallography, 25, 109-121. Chipera, S.J. and Bish, D.L. (1995) Multireflection RIR and intensity normalizations for quantitative analyses: applications to feldspars and zeolites. Powder Diffraction, 10, 47-55. Chung, F.H. (1974) Quantitative interpretation of X-ray diffraction patterns of mixtures. II. Adiabatic principle of X-ray diffraction analysis of mixtures. Journal of Applied Crystallography, 7, 526-531. Collins, D.R. and Catlow, C.R.A. (1992) Computer simulation of structures and cohesive properties of micas. American Mineralogist, 77, 1172-1181. Couvy, H., Frost, D.J., Heidelbach, F., Nyilas, K., Ungár, T., Mackwell, S., and Cordier, P. (2004) Shear deformation experiments at 11 GPa - 1400°C in the multianvil apparatus. European Journal of Mineralogy, 16, 877-889. De la Torre, A.G. (2003) Accuracy in Rietveld quantitative phase analysis of Portland cements. Journal of Applied Crystallography, 36, 1169-1176. De la Torre, A.G., Bruque, S., and Aranda, M.A.G. (2001) Rietveld quantitative amorphous content analysis. Journal of Applied Crystallography, 34, 196-202. de Wolff, P.M. (1956) Measurement of particle absorption by X-ray fluorescence. Acta Crystallographica, 9, 682-683.  61  Department of Indian Affairs and Northern Development (1993) Guidelines for Acid Rock Drainage Prediction in the North: Northern Mine Environment Neutral Drainage Studies No. 1. Prepared by Steffen, Robertson and Kirsten (B.C.) Inc. Ottawa, ON, Canada. Dipple, G.M., Raudsepp, M., and Gordon, T.M. (2002) Assaying wollastonite in skarn. In Industrial Minerals in Canada, Canadian Institute of Mining, Metallurgy and Petroleum Special Volume, 53, 303-312. Dollase, W.A. (1986) Correction of intensities for preferred orientation in powder diffractometry: application of the March model. Journal of Applied Crystallography, 19, 267-272. Dutta, H. and Pradhan, S.K. (2003) Microstructure characterization of high energy ballmilled nanocrystalline V2O5 by Rietveld analysis. Materials Chemistry and Physics, 77, 868-877. Giester, G., Lengauer, C.L., and Rieck, B. (2000) The crystal structure of nesquehonite, MgCO3·3H2O, from Lavrion, Greece. Mineralogy and Petrology, 70, 153-163. Glinnemann, J., King, H.E., Jr., Schulz, H., Hahn, T., La Placa, S.J., and Dacol, F. (1992) Crystal structures of the low-temperature quartz-type phases of silica and germanium dioxide at elevated pressure. Zeitschrift für Kristallographie, 198, 177-212. Government of Canada (1999) The Canadian Environmental Assessment Act: Comprehensive Study Report: Diavik Diamonds Project. Retrieved June 2008 from the website of The Canadian Environmental Assessment Agency: http://www.acee-ceaa.gc.ca/010/0003/0025/archive_e.htm. Graham, I., Burgess, J.L., Bryan, D., Ravenscroft, P.J., Thomas, E., Doyle, B.J., Hopkins, R., and Armsrong, K.A. (1998) Exploration history and geology of the Diavik kimberlites, Lac de Gras, Northwest Territories, Canada. In J.J. Gurney, J.L. Gurney, M.D. Pascoe, and S.H. Richardson, Eds., Proceedings of the VIIth International Kimberlite Conference, 1, 262-279. Red Roof Design, Cape Town, South Africa. Grady, M.M., Gibson, E.K., Jr., Wright, I.P., and Pillinger, C.T. (1989) The formation of weathering products on the LEW 85320 ordinary chondrite: evidence from  62  carbon and oxygen stable isotope compositions and implications for carbonates in SNC meteorites. Meteoritics, 24, 1-7. Gualtieri, A.F. (2000) Accuracy of XRPD QPA using the combined Rietveld – RIR method. Journal of Applied Crystallography, 33, 267-278. Hill, R.J. and Howard, C.J. (1987) Quantitative phase analysis from neutron powder diffraction data using the Rietveld method. Journal of Applied Crystallography, 20, 467-474. Hoffert, M.I., Caldeira, K., Benford, G., Criswell, D.R., Green, C., Herzog, H., Jain, A.K., Kheshgi, H.S., Lackner, K.S., Lewis, J.S., Lightfoot, H.D., Manheimer, W., Mankins, J.C., Mauel, M.E., Perkins, L.J., Schlesinger, M.E., Volk, T., and Wigley, T.M.L. (2002) Advanced technology paths to global climate stability: energy for a greenhouse planet. Science, 298, 981-987. Huot, F., Beaudoin, G., Hebert, R., Constantin, M., Bonin, G., and Dipple, G. (2003) Evaluation of Southern Québec asbestos residues for CO2 sequestration by mineral carbonation; preliminary results. Joint Annual Meeting of the Geological and Mineralogical Associations of Canada, Vancouver, BC, Canada. May 25-28, 2003. Inaba, S., Minakawa, T., and Noto, S. (1985) Nesquehonite and dypingite from Shiraki, Mie Prefecture, Japan. Chigaku Kenkyu, 34, 281-287. Järvinen, M. (1993) Application of symmetrized harmonics expansion to correction of the preferred orientation effect. Journal of Applied Crystallography, 26, 525-531. Jambor, J.L. and Blowes, D.W. (1998) Theory and applications of mineralogy in environmental studies of sulfide-bearing mine wastes. In L.J. Cabri and D.J. Vaughn, Eds., Modern Approaches to Ore and Environmental Mineralogy, 27, 367-401. Mineralogical Association of Canada, Ottawa, ON, Canada. Jambor, J.L. and Blowes, D.W. (1991) Mineralogical study of low-sulphide, highcarbonate, arsenic-bearing tailings from the Delnite minesite, Timmins, Ontario. In Proceedings of the 2nd International Conference on the Abatement of Acidic Drainage (Montréal, Québec, Canada), 4, 173-197.  63  Jambor, J.L., Dutrizac, J.E., and Raudsepp, M. (2007) Measured and computed neutralization potentials from static tests of diverse rock types. Environmental Geology, 52, 1019-1031. Jambor, J.L., Dutrizac, J.E., Groat, L.A., and Raudsepp, M. (2002) Static tests of neutralization potentials of silicate and aluminosilicate minerals. Environmental Geology, 43, 1-17. Jerome, L.E. and Mohanty, G.P. (1979) Deformation-induced structural effects in cerium. Journal de Physique Colloque, C5, 381-382. Jull, A.J.T., Cheng, S., Gooding, J.L., and Velbel, M.A. (1988) Rapid growth of magnesium-carbonate weathering products in a stony meteorite from Antarctica. Science, 242, 417-419. Kawano, M. and Tomita, K. (1996) Amorphous aluminum hydroxide formed at the earliest weathering stages of K-feldspar. Clays and Clay Minerals, 44, 672-676. Kloprogge, J.T., Martens, W.N., Nothdurft, L., Duong, L.V., and Webb, G.E. (2003) Low temperature synthesis and characterization of nesquehonite. Journal of Materials Science Letters, 22, 825-829. Kump, L.R., Brantley, S.L., and Arthur, M.A. (2000). Chemical weathering, atmospheric CO2, and climate. Annual Review of Earth and Planetary Sciences, 28, 611-667. Lackner, K.S. (2003) Climate change: A guide to CO2 sequestration. Science, 300, 16771678. Lackner, K.S., Wendt, C.H., Butt, D.P., Joyce, G.L., and Sharp, D.H. (1995) Carbon dioxide disposal in carbonate minerals. Energy, 20, 1153-1170. Lawrence, R.W. and Scheske, M. (1997) A method to calculate the neutralization potential of mining wastes. Environmental Geology, 32, 100-106. March, A. (1932) Mathematische theorie der regelung nach der korngestalt bei affiner deformation. Zeitschrift für Kristallographie, 81, 285-297. Maslen, E.N., Streltsov, V.A., Streltsova, N.R., and Ishizawa, N. (1995) Electron density and optical anisotropy in rhombohedral carbonates. III. Synchrotron X-ray studies of CaCO3, MgCO3 and MnCO3. Acta Crystallographica Section B: Structural Science, B51, 929-939.  64  McLean, H., Banas, A., Creighton, S., Whiteford, S., Luth, R.W., and Stachel, T. (2007) Garnet xenocrysts from the Diavik Mine, NWT, Canada: composition, color, and paragenesis. Canadian Mineralogist, 45, 1131-1145. McLean, R.J.C., Jamieson, H.E., and Cullimore, D.R. (1997) Formation of nesquehonite and other minerals as a consequence of biofilm dehydration. World Journal of Microbiology and Biotechnology, 13, 25-28. Mellini, M. and Viti, C. (1994) Crystal structure of lizardite-1T from Elba, Italy. American Mineralogist, 79, 1194-1198. Mining Association of Canada (2007) Towards Sustainable Mining Progress Report 2007: GHG Emissions and Energy Management Progress Report. Retrieved January 25, 2008 from the website of the Mining Association of Canada: http://www.mining.ca/www/Towards_Sustaining_Mining/Technical_Data_and_ Bulletins.php. Mitchell, R.H. (1986) Kimberlites: Mineralogy, Geochemistry and Petrology. Plenum Press, New York, NY, USA. Mitchell, R.H. and Putnis, A. (1988) Polygonal serpentine in segregation-textured kimberlite. Canadian Mineralogy, 26, 991-997. Moore, D.M. and Reynolds, R.C., Jr. (1997) X-ray Diffraction and the Identification and Analysis of Clay Minerals, Second Edition. Oxford University Press, New York, NY, USA. Omotoso, O., McCarty, D.K., Hillier, S., and Kleeberg, R. (2006) Some successful approaches to quantitative mineral analysis as revealed by the 3rd Reynolds Cup contest. Clays and Clay Minerals, 54, 748-760. O’Neil, J.R. and Barnes, I. (1971) C13 and O18 compositions in some fresh-water carbonates associated with ultramafic rocks and serpentinites: western United States. Geochimica et Cosmochimica Acta, 35, 687-697. Pacala, S. and Socolow, R. (2004) Stabilization wedges: solving the climate problem for the next 50 years with current technologies. Science, 305, 968-972. Paktunc, A.D. and Davé, N.K. (2002) Formation of secondary pyrite and carbonate minerals in the Lower Williams Lake tailings basin, Elliot Lake, Ontario, Canada. American Mineralogist, 87, 593-602.  65  Pawley, G.S. (1981) Unit-cell refinement from powder diffraction scans. Journal of Applied Crystallography, 14, 357-361. Power, I.M., Wilson, S.A., Thom, J., Dipple, G.M., and Southam, G. (2007) Biologically induced mineralization of dypingite by cyanobacteria from an alkaline wetland near Atlin, British Columbia, Canada. Geochemical Transactions, 8, article 13. Power, I.M., Wilson, S.A., Thom, J.M., Dipple, G.M., Gabites, J.E., and Southam, G. (2009) The hydromagnesite playas of Atlin, British Columbia, Canada: A biogeochemical model for CO2 sequestration. Chemical Geology, 260, 286-300. Price, W.A. and Errington, J.C. (1998) Guidelines for metal leaching and acid rock drainage at minesites in British Columbia. Retrived June 6, 2008 from the website of the British Columbia Ministry of Energy, Mines and Petroleum Resources: http://www.em.gov.bc.ca/Subwebs/mining/Project_Approvals/guidelines.htm#M etal%20Leaching%20and%20Acid%20Rock%20Drainage. Raudsepp, M., Pani, E., and Dipple, G.M. (1999) Measuring mineral abundance in skarn. I. The Rietveld method using X-ray powder-diffraction data. Canadian Mineralogist, 37, 1-15. Raudsepp, M. and Pani, E. (2003) Application of Rietveld analysis to environmental mineralogy. In J.L. Jambor, D.W. Blowes, and A.I.M. Ritchie, Eds., Environmental Mineralogy of Mine Wastes, 31, 165-180. Mineralogical Association of Canada, Ottawa, ON, Canada. Reinson, J. (2006) Diavik Diamond Mines Inc. management of the processed kimberlite containment facility: short term management plan, revision A. Prepared by Golder Associates Ltd., document number 05-1328-006/059-2012, report #0024. Rietveld, H.M. (1969) A profile refinement method for nuclear and magnetic structures. Journal of Applied Crystallography, 2, 65-71. Rimsaite, J. (1979) Natural amorphous materials, their origin and identification procedures. In M.M. Mortland and V.C. Farmer, Eds., Proceedings of the VIth International Clay Conference, Developments in Sedimentology, 27, 567-577. Elsevier Scientific Publishing Company, Amsterdam, The Netherlands.  66  Rinaudo, C., Gastaldi, D., and Bulluso, E. (2003) Characterization of chrysotile, antigorite and lizardite by FT-Raman spectroscopy. Canadian Mineralogist, 41, 883-890. Robertson, J.D., Tremblay, G.A., and Fraser, W.W. (1997) Subaqueous tailing disposal: a sound solution for reactive tailings. In Proceedings of the Fourth International Conference on Acid Rock Drainage, 3, 1027-1044. MEND, Natural Resources Canada, Ottawa, ON, Canada. Rollo, H.A. and Jamieson, H.E. (2006) Interaction of diamond mine waste and surface water in the Canadian Arctic. Applied Geochemistry, 21, 1522-1538. Sakharov, B.A., Dubinska, E., Bylina, P., Kozubowski, J.A., Kapron, G., and FrontczakBaniewicz, M. (2004) Serpentine-smectite interstratified minerals from Lower Silesia (SW Poland). Clays and Clay Minerals, 52, 55-65. Sanchez, M.S. and Gunter, M.E. (2006) Quantification of amphibole content in expanded vermiculite products from Libby, Montana U.S.A. using powder X-ray diffraction. American Mineralogist, 91, 1448-1451. Seifritz, W. (1990) CO2 disposal by means of silicates. Nature, 345, 486. Sharp, T.G., Otten, M.T., and Buseck, P.R. (1990) Serpentinization of phlogopite phenocrysts from a micaceous kimberlite. Contributions to Mineralogy and Petrology, 104, 530-539. Shirozu, H. and Bailey, S.W. (1966) Crystal structure of a two-layer Mg-vermiculite. American Mineralogist, 51, 1124-1143. Sobek, A.A., Schuller, W.A., Freeman, J.R., and Smith, R.M. (1978) Field and laboratory methods applicable to overburdens and minesoils. US Environmental Protection Agency Report EP-600/2-78-054. Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., and Miller, H.L., Eds. (2007) Climate Change 2007: The Physical Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK and New York, NY, USA. Sundquist, E.T. (1993) The global carbon dioxide budget. Science, 259, 934-941.  67  Sundquist, E.T. (1985) Geological perspectives on carbon dioxide and the carbon cycle. In E.T. Sundquist and W.S. Broecker, Eds., The Carbon Cycle and Atmospheric CO2: Natural Variations Archaen to Present, Geophysical Monographs, 32, 5-60. American Geophysical Union, Washington, DC, USA. Suortti, P. (1972) Effects of porosity and surface roughness on the X-ray intensity reflected from a powder specimen. Journal of Applied Crystallography, 5, 325331. Suzuki, J. and Ito, M. (1973) A new magnesium carbonate hydrate mineral, Mg5(CO3)4(OH)2·8H2O, from Yoshikawa, Aichi Prefecture, Japan. Journal of the Japanese Association of Mineralogists, Petrologists and Economic Geologists, 68, 353-361. Suzuki, J. and Ito, M. (1974) Nesquehonite from Yoshikawa, Aichi Prefecture, Japan: occurrence and thermal behaviour. Journal of the Japanese Association of Mineralogists, Petrologists and Economic Geologists, 69, 275-284. Ufer, K., Roth, G., Kleeberg, R., Stanjek, H., Dohrmann, R., and Bergmann, J. (2004) Description of X-ray powder pattern of turbostratically disordered layer structures with a Rietveld compatible approach. Zeitschrift für Kristallographie, 219, 519527. Wang, A., Wang, W., and Zhang, A. (1991) Microstructural variations of a pyrope inclusion in diamond, as revealed by a micro-Raman spectroscopic study. Canadian Mineralogist, 29, 517-524. Wilson, S.A., Dipple, G.M., Power, I.M., Thom, J.M., Anderson, R.G., Raudsepp, M., Gabites, J.E., and Southam, G. (2009) Carbon dioxide fixation within mine tailings at the Clinton Creek and Cassiar chrysotile deposits, Canada. Economic Geology, 104, 95-112. Wilson, S.A., Raudsepp, M., and Dipple, G.M. (2006) Verifying and quantifying carbon fixation in minerals from serpentine-rich mine tailings using the Rietveld method with X-ray powder diffraction data. American Mineralogist, 91, 1331-1341.  68  Wilson, S.A., Dipple, G.M., Anderson, R.G., and Raudsepp, M. (2004) Characterization of Clinton Creek mine residues and their suitability for CO2 sequestration. British Columbia and Yukon Chamber of Mines Mineral Exploration Roundup 2004, Vancouver, BC, Canada. January 26-29, 2004. Wilson, S.A., Thom, J.M., Dipple, G.M., Raudsepp, M., and Anderson, R.G. (2005) Towards sustainable mining: uptake of greenhouse gases by mine tailings. British Columbia and Yukon Chamber of Mines Mineral Exploration Roundup 2005, Vancouver, BC, Canada. January 24-27, 2005. Yu, S.C. (1997) Effects of pressure on the crystal structure of olivine in harzburgite xenolith of basalt. Proceedings of the National Science Council, Republic of China, 21, 173-179.  69  
  Chapter 3 Carbon isotopic fractionation between dypingite, Mg5(CO3)4(OH)2·5H2O, and aqueous bicarbonate in an evaporative and highly saline system1  3.1 Introduction Dypingite [Mg5(CO3)4(OH)2·5H2O] is the most common intermediate phase in the  decomposition  of  nesquehonite  [MgCO3·3H2O]  to  hydromagnesite  [Mg5(CO3)4(OH)2·4H2O] in abiotic precipitation/decomposition experiments (e.g., Davies and Bubela 1973; Canterford et al. 1984; Han and Lee 1985). Dypingite is also found in nature (1) as a weathering product of mafic or ultramafic rocks and mine tailings (Raade 1970; Suzuki and Ito 1973; Inaba et al. 1986; Gore et al. 1996; Wilson et al. 2006, 2009) and (2) in association with cyanobacterial colonies in Mg-rich alkaline wetlands (Power et al. 2007, 2009). In ultramafic mine tailings, dypingite and other hydrous and basic Mg-carbonate minerals can act as traps for atmospheric carbon dioxide (CO2), potentially storing this greenhouse gas in the long term as a mineral (Wilson et al. 2006, 2009). Wilson et al. (2009) suggest that accelerated carbonation of Mg-silicate mine tailings may be used by some mining operations to offset their greenhouse gas emissions. They propose that a combination of stable carbon and oxygen isotopes and radiocarbon could be used effectively to verify trapping and storage of atmospheric CO2 within secondary Mg-carbonate minerals in mine tailings. However, at this time very little is known about the fractionation of stable carbon and oxygen isotopes in hydrous and basic Mg-carbonate minerals like dypingite, hydromagnesite, and nesquehonite. In fact, the only published data for fractionation effects in any hydrous or basic Mg-carbonate 























































 1  A version of this chapter will be submitted for publication. Wilson, S.A., Barker, S.L.L., Dipple, G.M., Atudorei, V., and Thom, J.M. Carbon isotopic fractionation between dypingite, Mg5(CO3)4(OH)2·5H2O, and aqueous bicarbonate in an evaporative and highly saline system.  70  
 minerals are two equilibrium isotopic fractionation factors for exchange of oxygen between hydromagnesite and water at 0˚ and 25˚C (O’Neil and Barnes 1971). The lack of equilibrium or kinetic isotopic fractionation factors for Mg-carbonate minerals hampers the interpretation of stable isotope data collected from these minerals. This dearth of information also interferes with the development of a consistent and reliable protocol for verifying mineralization of atmospheric CO2 within dypingite and related minerals in ultramafic mine tailings. Here, we give a brief review of available equilibrium fractionation factors for carbonate minerals and suggest which ones may be most applicable to studies of carbon mineralization in mine tailings. To begin bridging this gap, free drift, open system experiments are undertaken to study the partitioning of 13C between dypingite and total dissolved inorganic carbon. The experiments from which dypingite is formed are conducted between 20˚ and 25˚C and the only source of carbon available to the systems is gaseous CO2 from the laboratory atmosphere. These experiments are designed to simulate conditions under which atmospheric CO2 is mineralized within hydrous, basic magnesium carbonate minerals like dypingite and hydromagnesite. More specifically, they simulate evaporation-driven precipitation of these minerals from Mg and Cl-rich, high-pH process water, which can be found in the tailings containment facilities of some ultramafic-hosted mines (e.g., the Mount Keith Nickel Mine, Chapter 4). The results of dypingite precipitation experiments are considered in light of previous studies of carbonate mineral precipitation and an isotopic fractionation factor is reported for exchange of carbon between dypingite and dissolved inorganic carbon. We discuss the implications that these experiments have for verification of CO2 sequestration in mine tailings using isotopic techniques.  3.2 Review of isotopic fractionation factors for carbonate minerals Due to the high abundance of calcite within the Earth’s crust, its utility as an industrial chemical, and the relative ease of its synthesis, the partitioning of stable carbon and oxygen isotopes during precipitation of calcite has been well studied experimentally. Fractionation factors for low-temperature isotopic exchange of carbon and oxygen  71  
 between calcite, gaseous carbon dioxide (CO2) and aqueous carbonate species (i.e., H2CO3, HCO3-, and CO32-) are well-constrained by experiments for equilibrium precipitation of calcite (e.g., Epstein 1953; Rubinson and Clayton 1969; O’Neil et al. 1969; Deines et al. 1974; Friedman and O’Neil 1977; Romanek et al. 1992; Kim and O’Neil 1997; Jiménez-López et al. 2001). A more complete list of references and an indepth review of equilibrium fractionation factors for hydrogen, carbon, and oxygen and their pertinence to geological systems is provided by Chacko et al. (2001). Considerable attention has also been given to the study of kinetic fractionation of 13C between aqueous and gaseous carbonate species (e.g., Vogel et al. 1970; Stiller et al. 1985; Zhang et al. 1995), with particular emphasis on isotopic fractionation during precipitation and dissolution of calcite (e.g., Turner 1982; Kim and O’Neil 1997; Skidmore et al. 2004). Equilibrium fractionation factors have also been determined for related minerals like siderite, Mg-calcite, and mixed-cation (Mg-Ca-Fe) carbonate minerals by precipitation at low temperatures using inorganic procedures (e.g., Jiménez-López et al. 2004, 2006; Romanek et al. 2009). Although isotopic fractionation during low-temperature precipitation of calcite has been well studied, undertaking similar studies of dolomite and magnesite has posed a challenge. The kinetic hurdles to precipitating dolomite and magnesite have yet to be overcome in laboratory experiments under ambient conditions and without mediation by micro-organisms (Land 1998; Romanek et al. 2009). Recently, Vasconcelos et al. (2005) used sulphate-reducing bacteria to mediate the precipitation of dolomite to obtain a temperature-dependent oxygen fractionation factor between dolomite and water over the range of 25˚ to 45˚C. However, carbon isotopic fractionation factors have yet to be reported for dolomite precipitated under these conditions. Comparable experimental results have not been published for magnesite although cyanobacteria have been implicated in mediating its precipitation under ambient conditions (Thompson and Ferris 1990). Traditionally, factors for isotopic fractionation between magnesite and gaseous or aqueous carbon species have been used as proxies for fractionation effects in hydrated Mg-carbonate minerals and hydrated, basic Mg-carbonate minerals (e.g., Kralik 1989; Braithwaite and Zedef 1996; Zedef et al. 2000; Léveillé et al. 2007; Power et al. 2007,  72  
 2009; Wilson et al. 2009). However, the lack of low-temperature, experimentally determined equilibrium fractionation factors for dolomite and magnesite necessitates the use of theoretical results to interpret data for natural occurrences of these minerals. A number of approaches have been used to estimate equilibrium fractionation factors for carbonate minerals, including arguments based on mineral chemistry and observations of natural systems (e.g., Clayton et al. 1968; Sheppard and Schwarcz 1970; Barnes and O’Neil 1971; Zachmann and Johannes 1989; Spötl and Burns 1994; Jiménez-López et al. 2004) and the use of statistical mechanical and quantum chemical modelling (e.g., McCrea 1950; Bottinga 1968; O’Neil et al. 1969; Shiro and Sakai 1972; Golyshev et al. 1981; Kieffer 1982; Chacko et al. 1991; Zheng 1999; Deines 2004; Schauble et al. 2006; Chacko and Deines 2008; Rustad et al. 2008). Although rapid precipitation of carbonate minerals in open systems has often been linked to isotopic disequilibrium (e.g., Fantidis and Ehhalt 1970; Hendy 1971; Turner 1982; Mickler et al. 2004, 2006; Kosednar-Legenstein et al. 2008), the wide range of possible deviations from isotopic equilibrium is difficult to predict for any given geological system. Because of this, equilibrium fractionation factors are best used as a first approximation to explaining isotopic fractionation effects and may be used to identify isotopic disequilibrium (e.g., Mickler et al. 2006). In order to assess which of the available equilibrium fractionation factors are most appropriate for use in interpreting stable carbon and oxygen isotopic data for magnesite and other Mg-carbonate minerals, temperature-dependent equilibrium carbon and oxygen isotopic fractionation factors for a variety of carbonate minerals are plotted in Figures 3.1 and 3.2. All fractionation factors for carbon and oxygen isotopes are given as 103lnαmineral-CO and 103lnαmineral-H O, respectively. These are experimentally and 2  2  theoretically determined equilibrium fractionation factors for aragonite, calcite, dolomite, magnesite, and a number of hydrous (s.s., containing structurally bound water) and basic (s.s., containing structurally bound hydroxyl) carbonate minerals. 3.2.1 Equilibrium carbon isotopic fractionation factors for carbonate minerals Carbon isotopic fractionation factors for minerals are commonly reported for isotopic exchange between a mineral and an aqueous or gaseous carbonate species. In  73  
 order to compare fractionation factors for minerals, values of 103lnαmineral-CO have been 2  calculated from factors reported relative to carbonate species other than CO2. A number of equilibrium carbon isotopic fractionation factors are plotted in Figure 3.1.  Figure 3.1: Equilibrium carbon isotopic fractionation between carbonate minerals and gaseous CO2 compiled from experimental and theoretical studies.  The most recent and widely accepted determination of low-temperature fractionation of carbon isotopes between each of aqueous H2CO3, HCO3-, and CO32- and gaseous CO2 is by Zhang et al. (1995). The results of this experimental study agree very well with previous work by Deuser & Degens (1967), Wendt (1968), Emrich et al.  74  
 (1970), Mook et al. (1974), Turner (1982), and Lesniak and Sakai (1989). The fractionation factors of Zhang et al. (1995) were only determined over the range of 5˚25˚C: however, their results for fractionation between HCO3- and CO2 are in excellent agreement with the earlier results of Mook et al. (1974) in this range of temperatures (Zhang et al. 1995; Chacko et al. 2001). Because Mook et al. (1974) studied the temperature dependence of this fractionation effect from 5˚ to 125˚C, we have applied the Mook et al. (1974) relation to calculations of 103lnαmineral-CO from reported 2  3  3  determinations of 10 lnαmineral-HCO . Consequently, values of 10 lnαmineral-CO were 3  -  2  obtained from fractionation factors between carbonate minerals and aqueous HCO3(Rubinson and Clayton 1969, Rustad et al. 2008 for calcite; Rubinson and Clayton 1969 for dolomite) using the results of Mook et al. (1974) for carbon isotopic fractionation between gaseous CO2 and aqueous HCO3-. Although the experiments of Mook et al. (1974) were conducted at temperatures between 5˚ and 125˚C, we have used their results to extrapolate down to 0˚C. The experimental results of Romanek et al. (1992) for 103lnαcalcite-CO were used 2  to plot theoretical values of 103lnαmineral-calcite from Deines (2004) and Rustad et al. (2008) (Fig. 3.1). For comparison, the theoretically determined equilibrium fractionation between magnesite and calcite from Deines (2004) was also plotted using the experimental results of Deines et al. (1974) for 103lnαcalcite-CO . 2  3  Values of 10 lnαmineral-CO for gaylussite [Na2Ca(CO3)2·5H2O] and pirssonite 2  [Na2Ca(CO3)2·2H2O] were calculated from experimentally determined values of 103lnαmineral-CO for gaylussite (Matsuo et al. 1972) and pirssonite (Böttcher 1994). This 3  2-  was done by extrapolation to higher temperatures of the experimental result of Zhang et al. (1995) for equilibrium fractionation between carbon in aqueous CO32- and gaseous CO2. 3.2.1.1 Ca-carbonate minerals Experimental data from several low-temperature studies of slow, controlled precipitation of Ca-carbonate minerals from supersaturated solutions are plotted in Fig. 3.1. Among these, results for calcite from Rubinson and Clayton (1969), Romanek et al. (1992), and Jiménez-López et al. (2001) are indistinguishable within experimental error. 75  
 The recalculation of Deines et al. (1974) for the temperature dependence of 103lnαcalciteCO2  used data compiled from a large number of early studies, but it is not in good  agreement with more recent results from low-temperature experiments. Also, the theoretical calculations of Chacko et al. (1991) give excellent predictions for fractionation of carbon at high temperatures, but tend to overestimate fractionation effects between calcite and CO2 at low temperatures (Chacko et al. 2001). Quantum chemical calculations done by Rustad et al. (2008) for 103lnαcalcite-CO at 25˚C give a 2  value that is approximately 0.6‰ lower than expected from experiments. Carbon isotopic fractionation factors for aragonite are similarly well constrained by experiments (e.g., Rubinson and Clayton 1969; Romanek et al. 1992). Also, the theoretical result given by Rustad et al. (2008) for 103lnαaragonite-CO at 25˚C is in very 2  good agreement with experimental fractionation factors from Rubinson and Clayton (1969) and Romanek et al. (1992). The experimentally determined fractionation factors of Romanek et al. (1992) for calcite and aragonite cover a wide range of low temperatures from 0˚ to 40˚C and are in excellent agreement with results from other experiments. Because of this, we have made use of the results of Romanek et al. (1992) in subsequent calculations involving these minerals. 3.2.1.2 Ca-Mg-carbonate minerals Ohmoto and Rye (1979) recalculated a fractionation curve for carbon in the system dolomite-CO2 using results from a field-based study of carbon and oxygen isotopes in paired grains of calcite and dolomite that was done by Sheppard and Schwarcz (1970). The results of Ohmoto and Rye (1979) and Rustad et al. (2008) are mutually consistent for dolomite, but no experimental data are available for direct confirmation. Furthermore, Sheppard and Schwarcz (1970) established that their results were valid over the range of temperatures between 100˚ and 650˚C using the Mg-calcite solvus thermometer. As such, the extrapolation of this curve to significantly lower temperatures, as done by Ohmoto and Rye (1979), may not be valid.  76  
 3.2.1.3 Mg-carbonate minerals Although we do not make direct use of fractionation factors for dolomite in this study, several previous workers have proposed that fractionation factors for magnesite might be derived from those for dolomite. Zachmann and Johannes (1989) argued, based on the stoichiometry of calcite, dolomite, and magnesite, that doubling the co-efficients in the expression of Sheppard and Schwarcz (1970) for 103lnαdolomite-calcite would produce an approximate expression for 103lnαmagnesite-calcite. Deines (2004) used the statistical mechanical approach of Chacko et al. (1991) to compute the temperature dependent fractionation effect for carbon between a number of carbonate minerals and calcite. As with dolomite, no experimental results are available for comparison with theoretical determinations of carbon fractionation between magnesite and CO2. However, the theoretical fractionation curve of Deines (2004), which is calculated for a large range of temperatures above 0˚C, and that of Rustad et al. (2008) for 25˚C are internally consistent at this temperature (i.e., 25˚C). As noted previously, the fractionation factors computed by Rustad et al. (2008) are in good agreement with experimental results for aragonite and within 0.6‰ of well-established experimental results for calcite. On this basis, we have used the theoretical fractionation curve of Deines (2004) for 103lnαmagnesite-calcite and the experimental curve of Romanek et al. (1992) for 103lnαcalciteCO2  to estimate the magnitude of carbon isotopic fractionation between magnesite and  gaseous CO2. In the absence of equilibrium fractionation factors, we have adopted the result of Deines (2004) as a rough proxy for the equilibrium fractionation factors of other Mg-carbonate minerals. 3.2.2 Equilibrium oxygen isotopic fractionation factors for carbonate minerals 3.2.2.1 Ca-carbonate minerals Temperature-dependent equilibrium oxygen isotopic fractionation factors are plotted in Figure 3.2. As with carbon, equilibrium oxygen isotopic fractionation between calcite and water is well-understood. As can be seen from Figure 3.2, the experimental results of Epstein et al. (1953), O’Neil et al. (1969), Kim and O’Neil (1997), and Jiménez-López et al. (2001) are in close agreement at 25˚C. There are, however, small  77  
 deviations (on the order of 0.5‰ or less) between several of these curves in the range of 10˚ to 40˚C. The experiments done by Kim and O’Neil (1997) had several advantages over those of previous workers, being stringently controlled to precipitate only calcite, and give internally consistent results for a large number of experiments over a wide range of low temperatures. However, the results of O’Neil et al. (1969) are applicable to a larger range of temperatures. Correspondingly, we have used the results of O’Neil et al. (1969) in any calculations involving 103lnαcalcite-H O for oxygen. 2  Figure 3.2: Equilibrium oxygen isotopic fractionation between carbonate minerals and gaseous CO2 compiled from experimental and theoretical studies.  78  
 Although we do not make use of fractionation factors for aragonite, it is worth noting the magnitude of the isotopic separation between polymorphs aragonite and calcite. Several studies have been made of oxygen fractionation factors for aragonite in biominerals at low temperatures. Grossman and Ku (1986) obtained an equilibrium oxygen fractionation factor between aragonite and calcite from co-existing aragonitic and calcitic foraminifera. Also, Patterson et al. (1993) determined a fractionation factor for oxygen between aragonite and water by measuring the stable oxygen isotopic composition of fish otoliths and the water in which they lived. Outside of these observations of natural systems, one well-controlled laboratory experiment has been done to obtain a fractionation factor for exchange of oxygen isotopes between aragonite and calcite (Tarutani et al. 1969). This last study indicates that aragonite is enriched in 18  O by 0.6‰ over calcite at 25˚C (Fig. 3.2).  3.2.2.2 Ca-Mg-carbonate minerals An oxygen isotopic fractionation curve for 103lnαdolomite-H O from 25˚ to 45˚C has 2  been determined by Vasconcelos et al. (2005) using sulphate-reducing bacteria to overcome kinetic barriers to low-temperature precipitation of dolomite. As expected, and observed for calcite and aragonite (Kim and O’Neil 1997), the temperature dependence (which is indicated by the slopes of the curves) of fractionation factors for dolomite and calcite are very similar at low temperatures. A single estimate of 103lnαdolomite-H O for a 2  natural sample of dolomite precipitating from an alkaline lake at 20˚C (Clayton et al. 1968) is inconsistent with the experimental results of Vasconcelos et al. (2005), overestimating the 103lnαdolomite-H O relative to the experimental fractionation curve. 2  On the basis of stoichiometry, Barnes and O’Neil (1971) argue that 103lnαmineralH2 O  for oxygen in dolomite and magnesite could be approximated by combining (1) the  oxygen isotopic fractionation factor for calcite-H2O of O’Neil et al. (1969) with (2) the observation made by Tarutani et al. (1969) that the fractionation between Mg-calcite and water increases by 0.06‰ for every mole-% of MgCO3 in the solid solution. Estimating the fractionation of oxygen between dolomite and water using the stoichiometric argument made by Barnes and O’Neil (1971) gives an estimate of 103lnαdolomite-H O = 2  3  31.1‰ at 25˚C, which compares favourably to 10 lnαdolomite-H O = 31.0‰, obtained from 2  79  
 the fractionation curve of Vasconcelos et al. (2005). Therefore, we recommend adoption of the fractionation factor of Vasconcelos et al. (2005). 3.2.2.3 Mg-carbonate minerals Theoretical values of 103lnαmagnesite-H O have been proposed by Barnes and O’Neil 2  (1971), Spötl and Burns (1994), Zheng (1999), and Chacko and Deines (2008). The agreement between the theoretical result of O’Neil and Barnes (1971) and the experiments of Vasconcelos et al. (2005) for 103lnαdolomite-H O suggests that the estimate 2  of 103lnαmagnesite-H O made by Barnes and O’Neil (1971) may be an effective proxy for 2  experimental fractionation factors for magnesite (Fig. 3.2). However, more recent work by Jiménez-López et al. (2004) suggests that the equilibrium fractionation between oxygen in Mg-calcite and water may actually be as much as (0.17 ± 0.02)‰ for every mole-% of MgCO3. Estimating 103lnαmagnesite-H O using data from Jiménez-López et al. 2  (2004), assuming ideal magnesite stoichiometry and that the fractionation observed for Mg-calcite and water holds over the entire calcite-magnesite solid solution, gives what may be an unrealistically large 103lnαmagnesite-H O (Fig. 3.2). 2  Spötl and Burns (1994), using an approach proposed by Aharon (1988), calculated an oxygen isotope fractionation curve for magnesite-H2O. This was done by scaling experimentally determined fractionation curves for other divalent metal carbonates according to the known inverse relationship between the ionic radii of divalent cations and the relative oxygen isotopic fractionation of the carbonate minerals that they form. The statistical mechanical computation of Chacko and Deines (2008) for 103lnαdolomite-H O is approximately 2‰ higher than the experimental fractionation factor 2  of Vasconcelos et al. (2005) at 25˚C, suggesting that their fractionation factor for magnesite may also be overestimated. In the absence of experimentally determined fractionation factors for magnesite-H2O, we recommend the use of fractionation factors proposed by either Barnes and O’Neil (1971) or Spötl and Burns (1994). Although these fractionation factors have not been validated by experiments, they were determined based on careful observation of natural systems and experimental results for chemically and structurally related minerals.  80  
 3.2.3  Equilibrium isotopic fractionation factors for hydrous and basic carbonate minerals Isotopic fractionation factors are provided for the following hydrous and basic  carbonate  minerals  monohydrocalcite  (Figs.  3.1-3.4):  [CaCO3·H2O],  hydromagnesite  gaylussite  [Mg5(CO3)4(OH)2·4H2O],  [Na2Ca(CO3)2·5H2O],  pirssonite  [Na2Ca(CO3)2·2H2O], azurite [Cu3(CO3)2(OH)2], and malachite [Cu2(CO3)(OH)2]. Fractionation effects between pairs of more and less hydrous carbonate minerals are shown in Figures 3.3 and 3.4. It can be seen from Figure 3.3, that monohydrocalcite is isotopically lighter in carbon than calcite under equilibrium conditions (Jiménez-López et al. 2001). A similar trend can be seen for pirssonite and gaylussite, with the more hydrous gaylussite being isopically lighter than pirssonite. Oxygen isotopic fraction factors (Fig. 3.4) show the same trend as those for carbon (Fig. 3.3). Measurements of 103lnαhydromagnesite-H O (from O’Neil and Barnes 1971) 2  are consistently lower in magnitude than predictions made for magnesite (e.g., Barnes and O’Neil 1971; Spötl and Burns 1994). The fractionation curves from Melchiorre et al. (1999, 2000) show that malachite is isotopically lighter in oxygen than the less hydrous azurite. In all of these cases, the equilibrium fractionation effects are smaller for more hydrous and more basic mineral phases than for related, anhydrous or less hydrous/basic minerals. There is no statistically significant difference in the 103lnα values for oxygen isotopic fractionation between calcite and monohydrocalcite and H2O (Jiménez-López et al. 2001). However, it follows that the presence of more strongly-bonded hydroxyl groups within a mineral’s structure may give rise to larger differences in fractionation factors than structural H2O. If the relationships observed in Figures 3.3 and 3.4 hold for other carbonate minerals, one would expect basic and hydrous Mg-carbonate minerals to be isotopically lighter in both carbon and oxygen than magnesite formed at isotopic equilibrium under the same conditions. Bonding environment exerts significant control over the energy states of the constituent atoms of a mineral (e.g., Chacko et al. 2001). Because of this, crystal chemistry will influence carbon and oxygen isotopic fractionation in Mg-carbonate minerals. Significant differences in stable carbon and oxygen isotopic fractionation are observed between the CaCO3 polymorphs, calcite (R-3c) and aragonite (Pmcn) (e.g.,  81  
 Romanek et al. 1992; Kim and O’Neil 1997). Because of the structural and chemical differences between magnesite and other Mg-carbonate minerals, it cannot be assumed that carbon and oxygen isotopic fractionation effects in these minerals are similar. Consequently, isotopic fractionation factors for specific Mg-carbonate minerals are required for improved analysis of stable isotopic data for these minerals.  Figure 3.3: Equilibrium carbon isotopic fractionation between hydrous and basic carbonate minerals and gaseous CO2.  82  
  Figure 3.4: Equilibrium oxygen isotopic fractionation between hydrous and basic carbonate minerals and water.  Currently, the only published data for fractionation effects in any hydrous or basic Mg-carbonate minerals are those reported by O’Neil and Barnes (1971) for exchange of oxygen between hydromagnesite and water at 0˚ and 25˚C. It is important to consider here that although O’Neil and Barnes (1971) report precipitation of hydromagnesite within 3 weeks at low temperatures, subsequent studies have not been able to precipitate hydromagnesite below a temperature of 55˚C on such a short timescale (e.g., Botha and Strydom 2001; Li et al. 2003). It should be noted here that the temperature conditions at Mount Keith, and those used in this study, are not as high as  83  
 those generally used to precipitate hydromagnesite in the laboratory. However, it is likely that the relative humidity under which Mg-carbonate minerals precipitate exerts some degree of control over the hydration state of these minerals, as it does with hydrous Mgsulphate minerals. This may explain why hydromagnesite precipitates in the desert at Mount Keith, but dypingite precipitates at comparable temperatures in our laboratory, which is located in a temperate rainforest. At the time when the results of O’Neil and Barnes (1971) were published, dypingite had only just been recognized as a mineral (Raade 1970). Also, the crystal chemistry and structure of hydromagnesite were still under debate (e.g., Murdock 1954; White 1971; Robie and Hemingway 1972) and were not fully described until publication of the work of Akao et al. (1974), Stephan (1974), and Akao and Iwai (1977b). Until then, “hydromagnesite” was commonly used loosely as a descriptive term for a variety of hydrated and basic Mg-carbonate minerals, a practice that still persists to some extent outside of the mineralogical community. It is not improbable that, at least in the case of the experiment conducted at 25˚C, dypingite was produced instead of hydromagnesite.  3.3 Methods 3.3.1 Procedure for precipitating dypingite During this study, dypingite was synthesized in two identical bench top precipitation experiments (dyp-may14-09-1 and dyp-may14-09-2) that were run simultaneously under ambient conditions. These initially contained 1.00 mol/L NaCl (Fisher Scientific, Certified ACS Sodium Chloride), 0.100 mol/L MgCl2 (Sigma-Aldrich, Anhydrous Magnesium Chloride assayed at ≥ 98% purity), and 2.5 x 102 mol/L NaOH (Fisher Scientific, Certified ACS Sodium Hydroxide). Solutions were prepared gravimetrically from water and reagents that were weighed on a scale with ± 0.1 g precision. Before each experiment commenced, a 2000 mL Pyrex® beaker and a magnetic stir bar were sterilized by rinsing with anhydrous ethanol and were then rinsed three times in Nanopure de-ionized water (17.8 MΩ·cm, prepared using a Barnstead™ Epure™ water deionization system). A magnetic stir bar and 1500.0 g of Nanopure water  84  
 were added to each beaker, whereupon the water was allowed to equilibrate with atmospheric CO2 by stirring at 500 rpm for approximately 20 minutes using Fisher Scientific magnetic stirplates. At this point, the rotation of the stirbar was reduced to 60 rpm and reagents were added to each beaker as summarized in Table 3.1. A Cole-Parmer Masterflex® L/S® precision standard pump system fitted with two L/S® Easy-Load® II pump heads was used to pump laboratory air into solution. Laboratory air was the sole input of CO2 into these experiments and is assumed to have had the same composition as the bulk atmosphere (i.e., PCO ≈ 4 x 10-4 atm). Carbon dioxide in air was introduced into 2  solution using two lengths of Tygon® lab tubing (for each beaker) fit with 0.22 µm Millipore Millex GP sterile syringe filters at their open ends to prevent contamination through the pump. The tubing used had an inner diameter of ~6.4 mm and, with the pump system set to 500 rpm, bubbled approximately 2.8 x 103 mL/min of air through solution (based on specifications from the Cole-Parmer website at www.masterflex.com). Experiments were stirred continuously (at either 60 rpm or 100 rpm), maintained at laboratory temperature, and left open to the atmosphere so that evaporation could drive the solutions to supersaturation with respect to dypingite. Table 3.1: Starting conditions for experiments dyp-may14-09-1 and dyp-may14-09-2.  dyp-may14-09-1  mH2O (g) (±0.1 g) 1500.0  mNaCl (g) (±0.1 g) 87.6  mMgCl2 (g) (±0.1 g) 14.4  mNaOH (g) (±0.1 g) 1.5  mtotal (g) (±0.1 g) 2247.9  mequipa (g) (±0.2 g) 646.3  dyp-may14-09-2  1500.0  87.6  14.4  1.5  2249.5  647.5  Experiment  a  Denotes the combined mass of the beaker and stir bar used in the experiment.  Daily measurements were taken for pH, temperature, and total mass of the experiments before and after sampling. Temperature was monitored using a standard mercury laboratory thermometer with ± 0.5ºC precision. A portable Thermo Scientific Orion 4-Star pH/ISE meter was used to measure pH. The pH meter was calibrated daily against three standard buffer solutions that are traceable to NIST standards at pH 4.00, 7.00, and 10.00 (purchased from Fisher Scientific). Both the thermometer and pH electrode were rinsed thoroughly with Nanopure water and carefully dried with a 85  
 Kimwipe before and after each measurement. Samples of water for measurement of total dissolved C, δ18O of water, and δ13C on dissolved inorganic carbon (DIC) were collected using sterile single-use syringes (from Henke Sass Wolf) fitted with 0.22 µm Millipore Millex GP sterile syringe filters to remove precipitate. Millipore filtered water was injected into 0.5 dram (1.85 mL) vials, leaving no headspace, and sealed vials were wrapped with Parafilm M® to prevent sample loss by evaporation. During the first four days of experiments dyp-may14-09-1 and dyp-may14-09-2, 3-mL volumes of wellmixed solution were drawn daily from each experiment to sample the precipitates. Because this provided too little material for mineralogical analysis, the procedure was revised so that 20-mL volumes were drawn for subsequent samples. These samples were filtered through Whatman No. 1 qualitative filter paper in order to isolate the precipitate. Samples of precipitate were not rinsed in water because dypingite is susceptible to partial decomposition in deionized water (Suzuki and Ito 1973; this study). Precipitate samples were dried under a hood for at least 24 hours, removed from filter paper with a sterile spatula, and gently disaggregated using an alumina mortar and pestle. 3.3.2 Qualitative X-ray powder diffraction methods Synthetic mineral phases in all samples of precipitate were identified from X-ray powder diffraction (XRPD) patterns. Finely ground aliquots of sample were mounted as slurry onto a zero-background quartz plate with anhydrous ethanol and allowed to dry at room temperature. XRPD data were collected on a Siemens (Bruker) D5000 θ-2θ diffractometer equipped with a VÅNTEC-1 detector. A long, fine-focus Co X-ray tube was operated at 35 kV and 40 mA and an Fe monochromator foil was employed. Data for mineral identification were collected with a step size of 0.04° 2θ and counting time of 1s/step over a range of 2-45° 2θ. Constituent mineral phases were identified with reference to the ICDD PDF-4+ 2008 database using the program DIFFRACplus EVA 14 (Bruker AXS 2008). 3.3.3 CHN analyses A Carlo Erba 1106 automatic analyzer was used to measure the wt.% of carbon, hydrogen, and nitrogen in five specimens of natural and synthetically produced  86  
 dypingite. On the order of 1 mg of each pulverized mineral specimen was combusted and the resulting gaseous species were scrubbed, reduced, and separated on a gas chromatograph column for measurement at the instrument’s detector. CHN analyses were done in the Microanalytical Unit of the Research School of Chemistry, The Australian National University. 3.3.4 Total inorganic carbon measurements Measurements of total dissolved inorganic carbon were done on 20 samples of filtered water collected from the two precipitation experiments. Measurements were done in the Department of Civil Engineering, The University of British Columbia, using an IL550 TOC/TN analyzer. Each sample of water was decanted into a sparge vessel and treated with acid to release dissolved carbon as gaseous CO2. The amount of CO2 that resulted was measured using an infrared detector. 3.3.5 Stable isotopic methods The δ13C and δ18O values of experimental precipitates and the δ13C values of dissolved inorganic carbon were analyzed using a gas bench attached to a Thermo Finnigan DeltaPlus XL isotope ratio mass spectrometer (IRMS) at the Department of Earth and Planetary Sciences, The University of New Mexico. Specimens of synthetic minerals and filtered water were treated with phosphoric acid to liberate CO2 at 50ºC according to the methods of Révész and Landwehr (2002) and Salata et al. (2000), respectively. Specimens were allowed to react with phosphoric acid for two hours. The external precision (1σ deviation) for isotopic analyses was approximately 0.1‰ for both δ13C and δ18O, as estimated from repeated analysis of calcite standards. Eight specimens of dypingite-rich precipitates were analyzed for δ18O and δ13C compositions and 16 samples of filtered water were analyzed for the δ13C composition of total dissolved inorganic carbon. δ18O values of filtered water were measured using a laser spectroscopic Liquid Water Isotope Analyzer from Los Gatos Research, Inc. at the Pacific Centre for Isotopic and Geochemical Research (PCIGR), The University of British Columbia. Eight specimens of filtered water, including one replicate specimen, were analyzed for their  87  
 18  O concentrations. Tap water was run between analyses of experimental waters as a  precaution against accumulation of salts within the isotope analyzer from combustion of these saline waters. Reproducibility of known values for three water standards was better than 0.3‰ for δ18O.  3.4 Analytical Results 3.4.1 X-ray powder diffraction results Mineral identifications from X-ray powder diffraction patterns are summarized in the last three columns of Tables 3.2 and 3.3. The crystalline phases precipitated in the two experiments are not naturally occurring and, by definition, are not minerals. Nonetheless, for the sake of clarity, synthetic phases will be referred to by the mineral names of their naturally occurring analogues. XRPD patterns of the final precipitates obtained from experiments dyp-may14-09-1 and dyp-may14-09-2 are plotted in Figure 3.5 with three other XRPD patterns for dypingite: one pattern from naturally occurring dypingite from Yoshikawa, Shinshiro Cho, Aichi Prefecture, Japan and two patterns collected on dypingite that was precipitated in association with cyanobacteria (replotted from Power et al. 2007). Precipitates were not obtained in sufficient quantity for collection of XRPD patterns from either experiment on days 1-3 and no sampling was done on day 7. Again, insufficient quantities of precipitates were obtained from experiment dyp-may14-09-1 on day 13 and from experiment dyp-may14-09-2 on day 10. From the fourth until the sixteenth day of both experiments, poorly crystalline brucite (characterized by broad, asymmetric peaks) was the most abundant phase in the precipitates and was accompanied by a lesser amount of synthetic halite. Although synthetic halite did not precipitate in the bulk solutions, it did form along the interior walls of the beakers and as a crust on the tubing. Experimental precipitates were carefully filtered and dried, so most of the halite detected with XRPD was likely contamination from the tubing and interior walls of the beakers.  88  
 Synthetic dypingite was identified in both experiments on day 16: at low abundance in experiment dyp-may14-09-2 and just above detection in experiment dypmay14-09-1. By the seventeenth day of both experiments, dypingite was the most abundant phase, the relative amount of brucite had declined, and halite remained present at relatively low abundance. In experiment dyp-may14-09-1, the amount of brucite continued to decline. On day 19 of this experiment, brucite was observed only at trace abundance and by day 20 it was no longer detectable. Contrastingly, brucite remained detectable in precipitates from experiment dyp-may14-09-2 up to and including the final day (i.e., day 21) of the experiment. Notably, the relative amount of brucite in experiment dyp-may14-09-2 did decline considerably over the final six days, but it was still readily detectable in the XRPD pattern from the final precipitate (Fig. 3.5). 3.4.2 Results of CHN determination CHN data were obtained for four specimens of synthetic dypingite and for the specimen of naturally occurring dypingite from Yoshikawa, Shinshiro Cho, Aichi Prefecture, Japan (Table 3.4). The four specimens of synthetic dypingite are from three precipitation experiments conducted under the same conditions as experiments dypmay14-09-1 and dyp-may14-09-2, but which were not sampled for water and from which only the final precipitate was collected and analyzed. The elemental abundances of the five specimens vary between 8.3 and 10.3 wt.% for carbon and 2.7 and 3.1 wt.% for hydrogen. No nitrogen was detected in any of the specimens. Some of this variability can be explained by the presence of halite in the experimental precipitates (i.e., jtdypingite, dyp-feb13-07-2, and the rinsed and unrinsed aliquots of dyp-sept23-07-2). This is evidenced by the higher abundances of carbon and hydrogen reported for (1) the natural, halite-free specimen of dypingite from Yoshikawa and (2) an aliquot of synthetic dypingite (from dyp-sept23-07-2) that was rinsed with distilled water to remove halite. The majority of experimental precipitates were not rinsed because the structure of dypingite tends to decompose on exposure to distilled water (Suzuki and Ito 1973; this study).  89  
  Table 3.2: Results of dypingite precipitation experiment dyp-may14-09-1. Time (hours)  pH  DIC (mg/L)  T (˚C) (±0.5˚C)  msoli (g) (±0.2 g)  msolf (g) (±0.2 g)  Δmb (g) (±0.4 g)  δ 13CDIC (±0.2‰)  δ 18OH2O (±0.3‰)  δ 13Cdyp (±0.2‰)  δ 18Odyp (±0.2‰)  dypingite present?a  brucite present?a  halite present?a  1  Date and Time of Sampling 5/14/09 18:00  0.0  9.22  –  22.0  1601.6  1592.4  –  -19.2  –  –  –  –  –  2 3 4 5 6 8 9 10 11 12 13 14 15 16 17 18 19 20 21  5/15/09 16:00 5/16/09 12:30 5/17/09 14:15 5/18/09 14:30 5/19/09 16:45 5/21/09 15:30 5/22/09 18:30 5/23/09 16:05 5/24/09 17:40 5/25/09 18:20 5/26/09 16:20 5/27/09 14:30 5/28/09 15:50 5/29/09 16:30 5/30/09 20:30 5/31/09 20:10 6/1/09 15:45 6/2/09 16:30 6/3/09 16:30  22.0 42.5 68.3 92.5 118.8 165.5 192.5 214.1 239.7 264.3 286.3 308.5 333.8 358.5 386.5 410.2 429.8 454.5 478.5  9.35 9.35 9.25 9.31 9.30 9.23 9.16 9.12 9.08 9.01 9.03 9.01 8.96 8.84 8.86 8.84 8.73 8.65 8.48  11.4 – 29.0 – 44.4 – 75.9 – 110.3 – 134.6 – 151.7 – 103.9 – 68.1 – 51.6  20.0 20.0 21.5 21.5 21.0 21.0 23.0 22.0 22.0 23.0 22.5 21.5 22.0 23.5 22.0 22.0 23.0 23.0 24.5  1548.9 1501.8 1444.4 1373.1 1299.2 1190.9 1117.6 1055.1 967.1 899.3 838.3 772.6 704.2 636.3 565.0 505.9 452.0 387.4 324.0  1540.7 1493.5 1414.4 1347.7 1274.6 1166.4 1092.5 1007.3 942.3 873.9 813.3 747.5 678.9 609.8 539.5 480.7 425.7 362.7 6.3  43.5 38.9 49.1 41.3 48.5 83.7 48.8 37.4 40.2 43.0 35.6 40.7 43.3 42.6 44.8 33.6 28.7 38.3 38.7  – -17.4 – – – -16.9 – – – -16.0 – – – -16.2 – -18.2 – -20.0 -20.7  -10.9 -11.3 – – – -10.4 – – – -6.8 – – – -2.0 – – – – – – –  – – – – – – – – – – – – – -14.0 – -15.2 – -15.4 -15.2  – – – – – – – – – – – – – 27.0 – 28.2 – 28.9 30.2  – – – – – – – – – – – – – tr M M M M M  – – M M M M M M M M – M M M m m tr – –  – – m m m m m m m m – m m m m m m m m  Day  a  b  Detected from X-ray powder diffraction patterns of precipitates. “M” for “Major” denotes the most abundant phase in the sample, “m” for “minor” denotes a minor phase, and “tr” denotes a phase present at trace abundance (i.e., near detection). Δm is the mass loss of water due to evaporation between sampling events.  90  
  Table 3.3: Results of dypingite precipitation experiment dyp-may14-09-2. Time (hours)  pH  DIC (mg/L)  T (˚C) (±0.5˚C)  msoli (g) (±0.2 g)  msolf (g) (±0.2 g)  Δmb (g) (±0.4 g)  δ 13CDIC (±0.2‰)  δ 18OH2O (±0.3‰)  1  Date and Time of Sampling 5/14/09 18:00  0.0  9.22  –  22.0  1602.0  1593.9  –  -19.3  2 3 4 5 6 8 9 10 11 12 13 14 15 16 17 18 19 20 21  5/15/09 16:00 5/16/09 12:30 5/17/09 14:15 5/18/09 14:30 5/19/09 16:45 5/21/09 15:30 5/22/09 18:30 5/23/09 16:07 5/24/09 17:40 5/25/09 18:20 5/26/09 16:20 5/27/09 14:30 5/28/09 15:50 5/29/09 16:30 5/30/09 20:30 5/31/09 20:10 6/1/09 15:45 6/2/09 16:30 6/3/09 16:30  22.0 42.5 68.3 92.5 118.8 165.5 192.5 214.1 239.7 264.3 286.3 308.5 333.8 358.5 386.5 410.2 429.8 454.5 478.5  9.34 9.36 9.30 9.30 9.30 9.21 9.16 9.12 9.10 9.01 9.02 9.01 8.94 8.83 8.88 8.84 8.75 8.65 8.52  10.6 – 23.3 – 43.1 – 71.2 – 98.8 – 118.6 – 135.4 – 123.4 – 83.0 – 62.5  20.5 20.0 22.0 22.0 21.5 21.5 23.0 22.5 22.0 23.5 23.0 22.0 22.5 24.0 22.5 23.5 23.5 23.5 24.5  1550.9 1504.1 1453.6 1379.3 1303.7 1190.3 1115.0 1052.5 984.4 916.6 856.0 791.0 727.1 674.0 604.7 555.1 504.8 436.7 372.8  1543.4 1495.7 1423.3 1354.9 1278.9 1164.1 1089.1 1027.4 958.8 890.6 830.3 765.5 701.8 649.0 579.5 529.6 479.1 411.1 4.4  43.0 39.3 42.1 44.0 51.2 88.6 49.1 36.6 43.0 42.2 34.6 39.3 38.4 27.8 44.3 24.4 24.8 42.4 38.3  – -19.0 –
 –
 –
 -18.1 –
 –
 –
 -17.3 –
 –
 –
 -16.9 – -18.2 – -19.8 -20.3  -10.9 -11.3 –
 –
 –
 -9.9 –
 –
 –
 -6.8 –
 –
 –
 -3.4 –
 –
 –
 –
 –
 –
 –
  Day  a  b  δ 13Cdyp (±0.2‰)  δ 18Odyp (±0.2‰)  dypingite present?a  brucite present?a  halite present?a  –
  –
  –
  –
  –
  –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 -12.0 –
 -15.7 –
 -15.3 -16.9  –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 25.8 –
 28.3 –
 29.5 29.5  –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 –
 m M
 M
 M
 M
 M
  –
 –
 M M
 M
 M
 M
 –
 M
 M
 M
 M
 M
 M
 m m m m
 m
  –
 –
 m
 m
 m
 m
 m
 –
 m
 m
 m
 m
 m
 m
 m
 m
 m
 m
 m
  Detected from X-ray powder diffraction patterns of precipitates. “M” for “Major” denotes the most abundant phase in the sample, “m” for “minor” denotes a minor phase, and “tr” denotes a phase present at trace abundance (i.e., near detection). Δm is the mass loss of water due to evaporation between sampling events.  91  
  Figure 3.5: X-ray powder diffraction patterns of synthetic and natural dypingite. The two uppermost patterns are of final precipitates from experiments dypmay14-09-1 and dyp-may14-09-2. The middle pattern is of dypingite from Yoshikawa, Shinshiro Cho, Aichi Prefecture, Japan. The two bottommost patterns are of experimental precipitates formed in association with filamentous cyanobacteria and are replotted from Power et al. (2007).  92  
 Table 3.4:  CHN data and H-content of dypingite compared to results of previous studies.  Sample Name  C (wt.%)a 10.3  H (wt.%)a 2.9  Htotal (apfu)b 14  HOH (apfu)b 2.0  HH2O (pfu)c 12  H 2O (pfu)c 5.8  jtdypingite  8.7  2.7  15  2.0  13  6.4  dyp-feb13-07-2  9.4  3.1  16  2.0  14  6.8  dyp-sept23-07-2 (unrinsed)  8.3  2.8  16  2.0  14  6.9  dyp-sept23-07-2 (rinsed)  9.6  3.1  16  2.0  14  6.8  Raade (1970)  12  2.0  10  5.1  Davies and Bubela (1973)  20  2.0  18  9.1  Suzuki and Ito (1973)  18  2.0  16  7.8  Canterford et al. (1984)  11  2.0  8.7  4.4  Yoshikawa dypingite (natural)  a b c  Measurement error on C and H determination is approximately ±0.3 wt.%. Abbreviation of “atoms per formula unit”. Abbreviation of “per formula unit”.  3.4.3 Data from monitoring of pH, temperature, and mass loss Daily measurements of pH, temperature, and mass loss due to evaporation and sampling are summarized in Tables 4.2 and 4.3 and Figures 3.6 and 3.7. Both experimental solutions began with a pH of 9.22. On the second day of experiments dyp-may14-09-1 and dyp-may14-09-2 this value increased to 9.35 and 9.34, respectively (Fig. 3.6a). The pH of both solutions declined gradually and linearly, reaching values of approximately 8.9 on day 16 when dypingite precipitated. At this point, pH began to change more rapidly – decreasing by approximately 0.4 pH units within six days. On the day the experiments were ended, the pH values for dyp-may1409-1 and dyp-may14-09-2 were 8.48 and 8.52, respectively. Solution temperature was controlled by the ambient temperature of the laboratory in which experiments were conducted and was allowed to vary. Temperatures varied between 20.0˚ ± 0.5˚C and 24.5˚ ± 0.5˚C and a general trend toward increasing temperature was observed with time. A temperature gradient of 0.5˚ ± 1.0˚C, which is 93  
 undetectable within the precision of our thermometer, may have been present in the laboratory as suggested by consistently lower temperatures measured for experiment dyp-may14-09-1 than for dyp-may14-09-2 (Fig. 3.6b).  Figure 3.6: pH and temperature (˚C) of solutions over time. Variation of pH with time is plotted in (a) and variation of temperature with time is plotted in (b). The time corresponding to samples in which dypingite was first identified is marked by a dashed line. Measurement errors in (a) are typically smaller than the symbols employed. Uncertainties associated with temperature measurements in (b) are denoted by the extents of the vertical bars.  Figure 3.7a plots the mass of solutions over the duration of the two experiments. A roughly linear decline in mass is observed for both experiments, with a slightly larger negative slope for dyp-may14-09-1. Mass was lost to both evaporation and sampling of solution and precipitates. In experiment dyp-may14-09-1, 1277.6 g (79.8%) of solution plus precipitate were lost over the course of the experiment. The majority, 51.2%, of mass loss resulted from evaporation and the remaining 28.5% was due to sampling.  94  
 Experiment dyp-may14-09-2 lost 76.7% of its mass – 49.5% to evaporation and 27.2% to sampling. The average rate of evaporative mass loss is plotted against time in Figure 3.7b. Here, the rate of mass loss due to evaporation is measured between sampling events (typically a 20-30 hour interval). Between 1.4-2.0 g/hour of water were lost to evaporation from experiment dyp-may14-09-1 and 1.0-2.0 g/hour of water were lost from experiment dyp-may14-09-2.  Figure 3.7: Water loss from precipitation experiments over time. (a) Plots the mass of solutions with time and (b) gives average hourly mass loss from solutions between measurements. Uncertainties on measured values are considerably smaller than the symbols employed. The time corresponding to samples in which dypingite was first identified is marked by a dashed line.  3.4.4 Stable isotopic results The stable carbon isotopic composition of total dissolved inorganic carbon (DIC) was measured for eight samples of filtered water from each experiment (Tables 3.2 and 3.3 and Fig. 3.8). At the onset of experiments dyp-may14-09-1 and dyp-may14-09-2,  95  
 δ13CDIC values were -19.2‰ and -19.3‰, respectively. In both experiments, δ13CDIC values became progressively more positive over time, reaching maximum measured values on day 16 of the experiments at the onset of dypingite precipitation. Maximum values of δ13CDIC were -16.2‰ and -16.9‰ for dyp-may14-09-1 and dyp-may14-09-2, respectively. After day 16, δ13CDIC values became progressively more negative.  Figure 3.8: Stable carbon and oxygen isotopic data for synthetic dypingite, dissolved inorganic carbon (DIC), and water. Uncertainties on measured values are considerably smaller than the symbols employed. The time corresponding to samples in which dypingite was first identified is marked by a dashed line.  Stable oxygen isotopic compositions were determined for eight samples of filtered water. Results for three samples from each experiment and two identical samples of the Nanopure water used to prepare experimental solutions are listed in Tables 3.2 and 3.3 and plotted in Figure 3.8. Analyses of the duplicate samples of Nanopure water give δ18OH2O values of -10.9‰ and -11.3‰. Water sampled from experiments dyp-may14-091 and dyp-may14-09-2 becomes progressively more enriched in  18  O with time. On day  14 of the experiments, the last day for which there are δ18OH2O data, water from dyp-  96  
 may14-09-1 has a value of δ18OH2O=-2.0‰ and water from dyp-may14-09-2 gives δ18OH2O=-3.4‰. Stable oxygen and carbon isotopic data were obtained for four specimens of dypingite from each experiment. Precipitates of dypingite from days 16, 18, 20, and 21 were analyzed from both experiments (Tables 3.2 and 3.3 and Fig. 3.8). Dypingite from experiment dyp-may14-09-1 gives -15.4‰ ≤ δ13Cdyp ≤ -14.0‰ and 27.0‰ ≤ δ18Odyp ≤ 30.2‰. Experiment dyp-may14-09-2 yielded dypingite with a broader range of values: 16.9‰ ≤ δ13Cdyp ≤ -12.0‰ and 25.8‰ ≤ δ18Odyp ≤ 29.5‰. 3.4.5 Results of total dissolved inorganic carbon measurements The concentration of total dissolved inorganic carbon (DIC) was measured for 10 samples from each experiment (Tables 3.2 and 3.3 and Figs. 3.9 and 3.10). Because only one ~2 mL sample of filtered water was collected each day, measurements of DIC concentrations were made on samples of water collected on alternating days. This approach allowed the remaining samples to be analyzed for stable isotopic concentrations. Over time, DIC concentrations in experiment dyp-may14-09-1 rose steadily from 11.4 mg/L on day 2 to a measured maximum of 151.7 mg/L on day 15. Likewise, DIC concentrations in experiment dyp-may14-09-2 rose from 10.6 mg/L on day 2 to a measured maximum of 135.4 mg/L on day 15. Day 16 saw the onset of dypingite precipitation, and the next measured values of DIC concentration (on day 17) had declined to 103.9 mg/L and 123.4 mg/L for experiments dyp-may14-09-1 and dypmay14-09-2, respectively. DIC concentrations continued to decline until the experiments were concluded. Final DIC concentrations measured for day 21 of the experiments were 51.6 mg/L and 62.5 mg/L.  97  
  Figure 3.9: Stable carbon isotopic data and concentrations of dissolved inorganic carbon (DIC) over time. δ13C values of DIC are given in (a) and concentrations of DIC in solution are given in (b). Uncertainties on measured values are considerably smaller than the symbols employed. The time corresponding to samples in which dypingite was first identified is marked by a dashed line.  98  
  Figure 3.10: Stable oxygen isotopic data and concentrations of dissolved inorganic carbon (DIC) versus mass loss of solution. δ18O values of water are given in (a) and concentrations of DIC in solution are given in (b). Evaporative water loss is modelled as a Rayleigh distillation effect in (a) at three different values of relative humidity (models for experiments 1 and 2 are denoted by crosses and triangles, respectively). Uncertainties on measured values are smaller than the symbols employed. The masses of solution corresponding to the points at which dypingite was first identified are marked by dashed lines.  99  
 3.5 Discussion 3.5.1 Mineralogy of final precipitates Ultimately, both precipitation experiments, dyp-may14-09-1 and dyp-may14-092, produced a synthetic mineral phase that gives XRPD patterns that are in excellent agreement with that for the mineral phase originally characterized as “yoshikawaite” by Suzuki and Ito (1973) and documented as dypingite by Inaba et al. (1985), Gore et al. (1996), and Power et al. (2007). XRPD patterns of the final precipitates from both experiments are plotted in Figure 3.5. Three additional patterns are provided for comparison. One of these was collected from a natural specimen of dypingite from Yoshikawa, Shinshiro Cho, Aichi Prefecture, Japan (purchased from Roger’s Minerals, http://www.rogersminerals.com). The remaining two patterns are replotted from Power et al. (2007) and were collected using micro-XRD techniques. These patterns were collected from dypingite that precipitated in association with filamentous cyanobacteria from (1) groundwater and (2) wetland water that were sampled from the area surrounding a hydromagnesite playa deposit near Atlin, British Columbia, Canada. “Yoshikawaite” is generally considered to be dypingite, and the pattern collected by Suzuki and Ito (1973) is labelled as dypingite in the ICDD PDF 4+ database. Although it suffers from preferred orientation, the XRPD pattern for the sample of naturally occurring dypingite, from the same locality as that studied by Suzuki and Ito (1973), gives an excellent match to their ICDD reference pattern (Fig. 3.5). The XRPD patterns for dypingite precipitated from experiments dyp-may14-09-1 and dyp-may1409-2 are in similarly good agreement with the pattern from Suzuki and Ito (1973) (Fig. 3.5). Both experimental precipitates contain halite, but only the precipitate from dypmay14-09-2 contains residual brucite. Microbial precipitation experiments done by Power et al. (2007) produced a phase that is more similar to the type dypingite, originally described by Raade (1970) and documented by Canterford et al. (1984) (Fig. 3.5). The most notable differences between XRPD patterns of the dypingite described by Raade (1970) and that of Suzuki and Ito (1973) are the extreme loss of intensity from two superstructure peaks at d-spacings of 33.2 Å and 16.4 Å, and a shift in these two peaks and one higher-angle harmonic  100  
 reflection [found at approximately 10.8 Å in the dypingite of Suzuki and Ito (1973)] toward smaller values of d. XRPD data alone, suggest that partial decomposition of “Raade’s dypingite” toward hydromagnesite [Mg5(CO3)4(OH)2·4H2O] may be the cause of this inconsistency. Dypingite is known to decompose to hydromagnesite and to be an intermediate phase in the decomposition of nesquehonite [MgCO3·3H2O] to hydromagnesite (e.g., Davies and Bubela 1973; Suzuki and Ito 1973; Canterford et al. 1984; Han and Lee 1985; Botha and Strydom 2001). Furthermore, Suzuki and Ito (1973) observed the disappearance of higher order peaks and the shift of the peak at 10.8 Å to 9.2 Å on heating at 150˚C, which produces a material with an XRPD pattern consistent with that of hydromagnesite. Therefore, in keeping with previous interpretations, we consider “yoshikawaite” to be dypingite and treat our precipitates as synthetic dypingite. Brucite is known to produce hydromagnesite on weathering under conditions of temperature and pressure that prevail at the Earth’s surface (e.g., Hostetler et al. 1966). In addition, Botha and Strydom (2001) and Xiong and Lord (2008) have synthesized hydromagnesite by slow replacement of brucite in aqueous solution. The gradual decline in the relative abundance of brucite in experiments dyp-may14-09-1 and dyp-may14-092 suggests that dypingite, like hydromagnesite, can form by the carbonation of brucite according to the following reactions: 4CO2(g) → 4CO2(aq)  (Eq. 3.1)  4CO2(aq) + 4OH- → 4HCO3-  (Eq. 3.2)  4HCO3- → 4H+ + 4CO32-  (Eq. 3.3)  4H+ + 4OH- → 4H2O  (Eq. 3.4)  5Mg2+ + 4CO32- + 2OH- + 5H2O → Mg5(CO3)4(OH)2·5H2O  (Eq. 3.5)  which give the net reaction 4CO2(g) + 5Mg2+ + 10OH- + H2O → Mg5(CO3)4(OH)2·5H2O  (Eq. 3.6)  or  101  
 4CO2(g) + 5Mg(OH)2 + H2O → Mg5(CO3)4(OH)2·5H2O  (Eq. 3.7)  Although dypingite is usually referred to by the formula Mg5(CO3)4(OH)2·5H2O (as expressed in Eqs. 3.6 and 3.7) the amount of structural water in its formula is quite variable. CHN data were recalculated on the basis of four atoms of carbon per formula unit (apfu), using the idealized stoichiometry of Raade (1970) for dypingite (Table 3.4). In addition, it was assumed that two H-atoms per formula unit (pfu) formed hydroxyl groups within the structure of dypingite. The number of structurally-bound water molecules in specimens of synthetic dypingite (produced according to the same method used to precipitate dypingite in dyp-may14-09-1 and dyp-may14-09-2) varies from 6.40 to 6.90 pfu. A considerably lower value of 5.80 water molecules per formula unit was determined for natural dypingite from Yoshikawa. This range of values is considerably more narrow than that obtained from a compilation of previous studies (i.e., Raade 1970; Davies and Bubela 1973; Suzuki and Ito 1973; Canterford et al. 1984). Previous determinations of elemental abundances in dypingite, give estimates of 4.35 to 9.05 water molecules per formula unit for different specimens of the same mineral (Table 3.4). In contrast, precipitation experiments of the design described in this study give a relatively narrow range in water content (i.e., 6.40-6.90 H2O pfu). 3.5.2 Fractionation of stable isotopes during precipitation of dypingite 3.5.2.1 Stable isotopic fractionation in DIC and water The evolution of δ13CDIC and δ18OH2O with time is illustrated in Figure 3.8. Initial δ13CDIC values for experiments dyp-may14-09-1 and dyp-may14-09-2 are -19.2‰ and 19.3‰, respectively. In the range of pH under which these experiments were conducted (i.e., 8.48-9.36) bicarbonate is known to be the dominant aqueous carbonate species by several orders of magnitude (e.g., Faure 1986). Using the equilibrium carbon isotopic fractionation factor of Mook et al. (1974), 103lnαHCO --CO (g)=7.9‰ at 25˚C, and assuming 3  2  that CO2 in the laboratory atmosphere has a similar value of δ13C to the bulk atmosphere [i.e. ~ -8‰, e.g., Clark and Fritz (1997) and Keeling et al. (2005)], DIC in equilibrium with the atmosphere should be characterized by δ13CDIC≈0‰. The large deviation from  102  
 this expected equilibrium value could be explained by (1) diffusive fractionation of carbon isotopes caused by preferential uptake of 12CO2(g) into solution or (2) extremely 13  C-deplected CO2 in the laboratory atmosphere. A potential cause of the depleted δ13C signature of DIC in the precipitation  experiments is a kinetic fractionation effect that can sometimes accompany diffusion of CO2 into a solution. O’Neil and Barnes (1971) describe near-instantaneous precipitation of travertines and scums made of calcite and aragonite from high-pH (i.e., pH>11), Ca2+OH--waters. They observed extreme depletion of  13  C in the resulting calcite and  aragonite and were able to explain this depletion with the use of Graham’s Law, αm*m=(m/m*)  1/2  , in which α is the observed diffusive fractionation between a molecule of  mass m* that bears a heavier isotope and a molecule of mass m that bears a lighter isotope. Considering the diffusion of  13  CO2 relative to  12  CO2, and assuming that these  molecules can be modelled as point masses exhibiting simple harmonic motion, O’Neil and Barnes (1971) predicted that a kinetic isotope effect of ~11.2‰ would accompany diffusion of gaseous CO2 into high-pH, Ca2+-OH--water. Assuming that laboratory CO2 has a δ13C value of -8‰, this kinetic diffusion effect would produce water with δ13CDIC≈19.2‰, which is consistent with the initial carbon isotopic composition of DIC in both experiments. Another possibility is that the depleted δ13C signature of the DIC is the result of anomalously light CO2 in our laboratory atmosphere. Operating under the assumption of equilibrium isotopic exchange between DIC and atmospheric CO2, the gaseous CO2 in our laboratory should have a δ13C in the range of -27.2‰ to -27.1‰. In a study of kinetic isotopic fractionation during low-temperature dissolution of calcite, Skidmore et al. (2004) found that the average δ13C of the atmosphere within their laboratory refrigerator was -13.8‰. A study of δ13C compositions of atmospheric CO2 in Paris, France found that laboratory air varied from -9.2‰ to -9.5‰ and that CO2 in a recently used classroom had a high concentration (4630 ppm) of depleted (δ13C=-24.4‰) CO2 that reflected human respiration (Widory and Javoy 2003). These values for δ13C of interior air do not reflect the same extreme  13  C-depletion required to produce an equilibrium value of  13  δ CDIC≈-27‰, but the possible influence of human respiration, or the influence of other  103  
 laboratory sources of 13C-depleted CO2, on the experiments cannot be ruled out until the δ13C of the laboratory air has been measured. It can be seen from Figure 3.8 that values of δ13CDIC become increasingly positive with time, reaching maximum values of -16.0‰ in experiment dyp-may14-09-1 and 16.9‰ in dyp-may14-09-2. This trend toward enrichment in 13C may result from a shift in the carbon isotopic composition of CO2 (in the case of a  13  C-depleted laboratory  atmosphere), very slow equilibration with atmospheric CO2 (of a composition similar to that of the bulk atmosphere), or degassing of light CO2 from solution. Comparison of the trends in δ13CDIC with the concentration of DIC over time (Fig. 3.9) demonstrates a contemporaneous increase in DIC concentration with  13  C-enrichment. As a result the  degassing scenario is not likely. Furthermore, the largest enrichments in δ13CDIC correspond to the greatest uptake of atmospheric CO2 into solution as DIC. After approximately 359 hours (i.e., by day 16 of the experiments), DIC concentrations began to decrease and became increasingly depleted in  13  C. This  behaviour is contemporaneous with the formation of dypingite in both experiments. It would appear that concentration of sufficient DIC into solution was driven by two processes: (1) slow uptake of atmospheric CO2 into solution and (2) evapoconcentration of the solution. The latter effect can be inferred from Figure 3.10, a plot of δ18OH O and 2  DIC concentration against the mass of solution remaining in experiments dyp-may14-091 and dyp-may14-09-2. By the onset of dypingite precipitation, both experiments had lost 37 wt.% of their initial water to evaporation, which would have contributed to the increased concentration of DIC in solution. The fractionation of oxygen isotopes in these experiments can be described by evaporative loss of 16O at moderate humidity. This effect is modelled in Fig. 3.10a as a Rayleigh process using measured values of H2O loss by evaporation, Δm (from Tables 3.2 and 3.3), and Gonfiantini’s (1986) relation for the humidity-dependence of kinetic isotopic fractionation of oxygen during evaporation: δ18OH O = (εH O(g)-H O(l) + ΔεH O(g)-H O(l))·lnf + δ18OH O(0) 2  2  2  2  2  2  (Eq. 3.8)  104  
 where εH O(g)-H O(l) is the equilibrium isotopic separation factor between water vapour and 2  2  liquid water, f is the fraction of liquid water remaining, and δ18OH O(0) is the initial 2  oxygen isotopic composition of the solution. ΔεH O(g)-H O(l) describes the humidity2  2  dependent effect of kinetic isotopic fractionation: ΔεH O(g)-H O(l) = -14.2(1-h)‰ for 2  2  h≡relative humidity (RH) (Gonfiantini 1986). The equilibrium isotopic separation factor was calculated to be -9.4‰ at 25˚C from the results of Kakiuchi and Matsuo (1979). ΔεH O(g)-H O(l) was determined for three different values of relative humidity: (1) ΔεH O(g)2  2  H2O(l)=-4.9‰  2  at 65.4% RH, which is the average for the month of May in Vancouver, BC,  Canada (Environment Canada 2009), (2) ΔεH O(g)-H O(l)=-7.1‰ for RH=50%, and (3) 2  2  ΔεH O(g)-H O(l)=-8.5‰ for RH=40%. The data for δ18OH O are broadly consistent with 2  2  2  kinetic fractionation as a result of evaporation at values of relative humidity between 40% and 65.4%. Mismatches are most likely due to variations in the ambient temperature and relative humidity of the laboratory in which precipitation experiments were conducted. This analysis establishes a kinetic control on the oxygen isotopic composition of the experimental solutions. We are currently unable to determine the extent of kinetic control on the fractionation of carbon isotopes between DIC and the atmosphere without direct measurement of δ13C for laboratory CO2. 3.5.2.2 Stable isotopic fractionation of C and O in dypingite Measurements of δ13C and δ18O from eight specimens of dypingite (four from each experiment) are plotted in Figure 3.8 with δ13CDIC and δ18OH O data. δ13Cdyp was 2  measured to be -14.0‰ and -12.0‰ (in experiments dyp-may14-09-1 and dyp-may1409-2, respectively) in precipitates from the first day that dypingite was detected in XRPD patterns. In the first experiment, δ13Cdyp appears to reach a relatively constant level by the third day of dypingite precipitation (day 18) and remains between -15.2‰ and -15.4‰ for all subsequent measurements. In spite of this, δ13CDIC in the first experiment continues to decline and reaches a value of -20.7‰ when the experiment concludes. As a result, the carbon isotopic separation between dypingite and DIC increases from an initial value of 2.3‰ to a final value of 5.5‰. The average isotopic fractionation between dypingite and DIC in experiment dyp-may14-09-1 is (3.8 ± 1.4)‰. In the second  105  
 experiment, δ13Cdyp declines to a comparable value to that in experiment dyp-may14-091, but continues past this point to end with a lower value of -16.9‰. As the DIC pool becomes increasingly depleted by dypingite precipitation, the value of δ13CDIC decreases from -16.9‰ to -20.3‰. The carbon isotopic separation between dypingite and DIC in experiment dyp-may14-09-2 varies from 2.5‰ to 4.9‰. The average isotopic fractionation between dypingite and DIC in experiment dyp-may14-09-2 is (3.8 ± 1.1)‰, which is indistinguishable from the result obtained from the first experiment. Using the equilibrium isotopic fractionation factor of Mook et al. (1974) for exchange of carbon between aqueous bicarbonate and gaseous CO2, gives an estimated fractionation of 11.7‰ between dypingite and gaseous CO2 (Figs. 3.1 and 3.3), which is consistent with equilibrium fractionation factors for other carbonate minerals. Comparison with factors for magnesite-CO2 shows that it is also consistent with the observed difference between mineral-CO2 fractionation factors for hydrous and anhydrous pairs of chemically related minerals (Fig. 3.3). In the upper portion of Figure 3.8, a relatively constant depletion of δ13CDIC is accompanied by a contemporaneous depletion of δ13Cdyp on a similar scale. Also, as the carbonation reaction progresses, the δ13C of dypingite approaches the initial δ13C of the DIC pool and actually reaches the same value in dyp-may14-09-2. This is inconsistent with the increasing δ13Cdyp and δ13CDIC values that would result from CO2 degassing. Furthermore, these trends in DIC concentration and δ13C suggest that the amount of DIC in solution was not buffered by influx of atmospheric CO2 during precipitation of dypingite. This implies that, within these relatively high-pH, saline solutions, the exchange of CO2 between DIC and the atmosphere is surprisingly slow. Exchange of CO2 is sufficiently slow that – at least on the timescale of dypingite precipitation – the DIC-dypingite system appears to behave as though it were largely closed to the atmosphere for isotopic exchange of carbon. The large standard deviations on the average values for carbon isotopic separation between dypingite and DIC indicate incomplete mixing of product and reactant. Because carbon isotopic equilibrium between DIC and dypingite was only approached from one direction in these non-reversible experiments, it cannot be stated definitively that isotopic equilibrium was reached. However, the value of the fractionation factor obtained for exchange of  13  C between 106  
 dypingite and DIC is consistent with equilibrium fractionation factors for other carbonate minerals. This result provides significant support for the conclusion that equilibrium isotope exchange of carbon was reached between dypingite and DIC. The solubility of gaseous CO2 is notably reduced as the salinity of a solution increases (e.g., Markham and Kobe 1941; Harned and Davis 1943; Onda et al. 1970; Yasunishi and Yoshida 1979). Experiments dyp-may14-09-1 and dyp-may14-09-2 both had an initial concentration of 1.0 M NaCl. As a result of evaporation, the final concentrations of dissolved NaCl could have been as high as 2.1 M in the first experiment and 2.2 M in the second experiment. The high salinity of these experiments may have had an impact on the rate of diffusion of CO2 into solution, but this remains to be tested. Kim and O’Neil (1997) have suggested that equilibrium/disequilibrium fractionation of oxygen isotopes in precipitation experiments like these may be controlled in part by the concentration of aqueous bicarbonate and cations. They observed larger fractionations of oxygen isotopes between carbonate minerals and water with increasing concentrations of aqueous metal-chlorides and sodium bicarbonate, which suggests that higher ionic strengths lead to larger fractionations (at least for oxygen isotopes). Because only two measurements of equilibrium fractionation of oxygen isotopes between hydromagnesite and water have ever been reported (for 0˚ and 25˚C, O’Neil and Barnes 1971), and because of our sparse δ18OH O data, it is difficult to 2  assess whether the concentration of NaCl had an impact on oxygen isotopic fractionation. However, within the resolution of our experiments none was detectable. Due to the limited number of δ18OH O data available, it is not possible to give a 2  factor for the fractionation of oxygen isotopes between dypingite and water under the conditions of the two precipitation experiments. However, the available data and simple Rayleigh distillation models suggest that it is on the order of 30‰. Assuming that the Rayleigh model for δ18OH O using values of ΔεH O(g)-H O(l)=-4.9‰ at 65.4% RH is a good 2  2  2  approximation to the behaviour of the experimental solutions, oxygen isotopic fractionations between 30.4‰ and 31.1‰ and between 30.0‰ and 31.5‰ might be expected for experiments dyp-may14-09-1 and dyp-may14-09-2, during the course of dypingite precipitation. It should be noted that the oxygen isotopic separation between  107  
 hydromagnesite and water would decrease with declining humidity (Fig. 3.10). Perhaps surprisingly, these predictions are consistent with the equilibrium isotopic fractionation factor of 103lnαhydromagnesite-H O=31.2‰ that was measured by O’Neil and Barnes (1971) 2  at 25˚C. Although it cannot be stated decisively whether or not equilibrium exchange of carbon isotopes was reached between dypingite and DIC, the isotopic fractionation measured between these two carbon pools should be applicable to similar natural and built systems from which dypingite, and the closely related mineral hydromagnesite, precipitate. These experiments also suggest that the equilibrium fractionation factor reported by O’Neil and Barnes (1971) for exchange of oxygen between hydromagnesite and water provides a good approximation to oxygen isotopic exchange in evaporative, saline systems. Perhaps the most significant results of these experiments are (1) that the DIC pool in saline, Mg-rich water can have limited exchange with the atmosphere during precipitation of a carbonate mineral and (2) that a fractionation of 103lnαdypingite-DIC=(3.8 ± 1.2)‰ may be expected for exchange of carbon between dypingite and DIC in saline water at 20˚C ≤ T ≤ 25˚C and 8.5 ≤ pH ≤ 9.4.  3.6 Implications for CO2 sequestration in mine tailings The conditions used in these experiments simulate those at the Mount Keith Nickel Mine, Western Australia, Australia (discussed in detail in Chapter 4), where ultramafic mine tailings are stored in an arid environment and are periodically exposed to highly saline, neutral to basic process water. In ultramafic mine tailings environments like that at Mount Keith, sub-potable water with high ionic strength may be used for mineral processing. Once ore has been processed, mineral waste is pumped suspended in process water into tailings impoundments for permanent storage. In this environment, wastewater can become enriched in Mg by dissolution of Mg-rich gangue minerals. Wilson et al. (2009) have used stable carbon and oxygen isotopic data and radiocarbon data to demonstrate trapping of atmospheric CO2 within hydrous Mgcarbonate minerals (i.e., nesquehonite and lansfordite) and hydrous, basic Mg-carbonate  108  
 minerals (i.e., dypingite and hydromagnesite) at two historical chrysotile mines at Clinton Creek, Yukon Territory, Canada and Cassiar, British Columbia, Canada. They found that many of their stable isotopic data for secondary Mg-carbonate minerals fell outside the ranges predicted by available equilibrium fractionation factors for magnesite. In fact, stable carbon and oxygen data for dypingite and hydromagnesite from the Clinton Creek mine were found to be dependent upon the depth at which they had precipitated within the mine tailings pile. Stable isotope signatures of Mg-carbonate minerals that had precipitated at the surface of the tailings were significantly more enriched in both δ13C and δ18O and were commonly consistent with predicted values for precipitation of magnesite in equilibrium with atmospheric CO2 (Wilson et al. 2009). Contrastingly, samples found in the subsurface environment were significantly depleted in 13C and 18O relative to predicted values made using equilibrium isotopic fractionation factors, which suggested input of an isotopically depleted source of carbon or the action of a kinetic isotope effect during mineral precipitation. The stable carbon and oxygen isotopic results of Wilson et al. (2009) are more indicative of the processes by which carbon is cycled in mine tailings environments than the sources from which it is drawn into carbonate minerals. Using 14C as a tracer, they found that atmospheric CO2 was being sequestered into minerals with stable isotopic signatures that were both in and out of accord with predictions made using equilibrium fractionation factors. Ultimately, they relied upon the use of 14C data to confirm trapping of atmospheric carbon within Mg-carbonate minerals. Mine tailings storage facilities, particularly those at active mining operations, are subject to phases of wetting during deposition of new tailings and phases of drying as process water evaporates or is reclaimed. Skidmore et al. (2004) have demonstrated that significant kinetic isotopic fractionation can occur during the first few hours of carbonate mineral dissolution as a result of the influx of atmospheric CO2. In open system calcite dissolution experiments, they report deviations between calculated values for δ13CDIC (from equilibrium fractionation factors) and measured values that are as large as -17.4‰. Based on the results of their experiments, Skidmore et al. (2004) suggest that, in geological systems characterized by short exposures to water and low water-rock ratios, kinetic fractionation processes will determine the carbon isotopic signature of the DIC pool. Although we have considered precipitation of carbonate minerals rather than their  109  
 dissolution, this observation implies that kinetic fractionation processes may also exert significant control over the isotopic signature of carbonate minerals precipitated during wet phases in typically unsaturated mine tailings. Laboratory experiments designed to precipitate dypingite under conditions similar to those in tailings storage facilities demonstrate that stable isotopic data may provide ambiguous or misleading information about the sources from which environmental isotopes like oxygen and carbon are drawn. Perhaps the single most telling example from this study is that equilibrium fractionation factors may not provide an explanation for the δ13C signature of DIC, even though the only available source of carbon was the atmosphere. Had similarly  13  C-depleted dypingite been collected from the tailings  storage facility of an operational mine, this depletion could have been explained equally well by (1) diffusion fractionation of atmospheric CO2 into solution (e.g., O’Neil and Barnes 1971; Van Strydonck et al. 1989; Kosednar-Legenstein et al. 2008), (2) mixing of process waters with oxidized dissolved organic carbon from mine sewage, (3) mixing of process waters with groundwater containing DIC derived from soil CO2 (e.g., Cerling 1984), (4) use of 13C-depleted chemicals during mineral processing, or (5) 13C-depletion of local atmosphere as a result of high biological activity or fossil fuel emissions from mine site generators. The results of this study and the field-based evidence of Wilson et al. (2009) suggest that, in the mine tailings environment, stable isotope data describe the processes by which elements are cycled more than their provenance. In light of this, isotopic analogue experiments, rather than equilibrium precipitation experiments, may be more useful guides to the study of complex, low-temperature geological systems like tailings storage facilities. Consequently, a different tracer like radiogenic  14  C should be  used for the purpose of verifying sequestration of atmospheric carbon within mine tailings.  110  
 3.7 References Aharon, P. (1988) A stable-isotope study of magnesites from the Rum Jungle uranium field, Australia: Implications for the origin of strata-bound massive magnesites. Chemical Geology, 69, 127-145. Akao, M. and Iwai, S. (1977a) The hydrogen bonding of artinite. Acta Crystallographica, Section B: Structural Crystallography and Crystal Chemistry, B33, 3951-3953. Akao, M. and Iwai, S. (1977b) The hydrogen bonding of hydromagnesite. Acta Crystallographica, Section B: Structural Crystallography and Crystal Chemistry, B33, 1273-1275. Akao, M., Marumo, F., and Iwai, S. (1974) Crystal structure of hydromagnesite. Acta Crystallographica, Section B: Structural Crystallography and Crystal Chemistry, B30, 2670-2672. Barnes, I. and O’Neil, J.R. (1971) Calcium-magnesium carbonate solid solutions from Holocene conglomerate cements and travertines in the Coast Range of California. Geochimica et Cosmochimica Acta, 35, 699-718. Botha, A. and Strydom, C.A. (2001) Preparation of a magnesium hydroxy carbonate form magnesium hydroxide. Hydrometallurgy, 62, 175-183. Bottinga Y. (1968) Calculations of fractionation factors for carbon and oxygen isotopic exchange in the system calcite-carbon dioxide-water. Journal of Physical Chemistry, 2, 800-808. Böttcher, M.E. (1994)  13  C/12C partitioning during synthesis of Na2Ca(CO3)2·2H2O.  Journal of the Chemical Society, Chemical Communications, 1485. Braithwaite, C.J.R. and Zedef, V. (1996) Hydromagnesite stromatolites and sediments in an alkaline lake, Salda Gölü, Turkey. Journal of Sedimentary Research, 66, 9911002. Bruker AXS (2008) DIFFRACplus EVA 14 Release 2008. Bruker AXS, Germany. Canterford, J.H., Tsambourakis, G, and Lambert, B. (1984) Some observations of the properties  of  dypingite,  Mg5(CO3)4(OH)2·5H2O,  and  related  minerals.  Mineralogical Magazine, 48, 437-442.  111  
 Cerling, T.E. (1984) The stable isotopic composition of modern soil carbonate and its relationship to climate. Earth and Planetary Science Letters, 71, 229-240. Chacko, T. and Deines, P. (2008) Theoretical calculation of oxygen isotope fractionation factors in carbonate systems. Geochimica et Cosmochimica Acta, 72, 3642-3660. Chacko, T., Mayeda, T.K., Clayton, R.N., and Goldsmith, J.R. (1991) Oxygen and carbon isotope fractionations between CO2 and calcite. Geochimica et Cosmochimica Acta, 55, 2867-2882. Chacko, T., Cole, D.R., and Horita, J. (2001) Equilibrium oxygen, hydrogen and carbon isotope fractionation factors applicable to geologic systems. In Stable Isotope Geochemistry. Valley, J.W. and Cole, D.R., Eds. Reviews in Mineralogy and Geochemistry, 43, 1-81. Clark, I. and Fritz, P. (1997) Environmental Isotopes in Hydrogeology. CRC Press, U.S.A., 328 p. Clayton, R.N., Jones, B.F., Berner, R.A. (1968) Isotopic studies of dolomite formation under sedimentary conditions. Geochimica et Cosmochimica Acta, 32, 415-432. Coleyshaw, E.E., Crump, G., and Griffith, W.P. (2003) Vibrational spectra of the hydrated  carbonate  minerals  ikaite,  monohydrocalcite,  lansfordite  and  nesquehonite. Spectrochimica Acta Part A, 59, 2231-2239. Davies, P.J. and Bubela, B. (1973) The transformation of nesquehonite into hydromagnesite. Chemical Geology, 12, 289-300. Deines, P. (2004) Carbon isotope effects in carbonate systems. Geochimica et Cosmochimica Acta, 68, 2659-2679. Deines, P., Langmuir, D., and Harmon, R.S. (1974) Stable carbon isotope ratios and the existence of a gas phase in the evolution of carbonate ground waters. Geochimica et Cosmochimica Acta, 38, 1147-1164. Deuser, W.G. and Degens, E.T. (1967) Carbon isotope fractionation in the system CO2(gas)-CO2(aqueous)-HCO3-(aqueous). Nature, 215, 1033-1035. Emrich, K., Ehalt, D.H., and Vogel, J.C. (1970) Carbon isotope fractionation during the precipitation of calcium carbonate. Earth and Planetary Science Letters, 8, 363371.  112  
 Environment Canada (2009) National Climate Data and Information Archieve. Vancouver  International  Airport.  Retrieved  01  October  2009.  [http://www.climate.weatheroffice.ec.gc.ca/climate_normals/stnselect_e.html]. Epstein, S. Buchsbaum, R., Lowenstam, H.A., Urey, H.C. (1953) Revised carbonatewater isotopic temperature scale. Geological Society of America Bulletin, 64, 1315-1326. Fantidis, J. and Ehhalt, D.H. (1970) Variations of the carbon and oxygen isotopic composition in stalagmites and stalactites; evidence of non-equilibrium isotopic fractionation. Earth and Planetary Science Letters, 10, 136-144. Faure, G. (1986) Principles of Isotope Geology. Wiley, U.S.A., 589 p. Friedman, I. and O’Neil, J.R. (1977) Compilation of stable isotope fractionation factors of geochemical interest. In Data of Geochemistry. Fleischer, M., Ed. United States Geological Survey Professional Paper 440-KK. Giester, G., Lengauer, C.L., and Rieck, B. (2000) The crystal structure of nesquehonite, MgCO3·3H2O, from Lavrion, Greece. Mineralogy and Petrology, 70, 153-163. Golyshev, S.I., Padalko, N.L., and Pechenkin, S.A. (1981) Fractionation of stable oxygen and carbon isotopes in carbonate systems. Geochemistry International, 18, 85-99. Gonfiantini, R. (1986) Environmental isotopes in lake studies. In Handbook of Environmental Isotope Geochemistry, Vol. 2, The Terrestrial Environment. Fritz, P. and Fontes, J.-C., Eds. Elsevier, Amsterdam, The Netherlands, p.113-168. Gore, D.B., Creagh, D.C., Burgess, J.S., Colhoun, E.A., Spate, A.P., and Baird, A.S. (1996) Composition, distribution and origin of surficial salts in the Vestfold Hills, East Antarctica, Antarctic Science, 8, 73-84. Grossman, E.L. and Ku, T.-L. (1986) Oxygen and carbon isotope fractionation in biogenic aragonite: Temperature effects. Chemical Geology, 59, 59-74. Han, S.-K. and Lee, M.-D. (1985) The formation mechanism of basic magnesium carbonate. Hwahak Konghak, 23, 69-78. Harned, H.S. and Davis, R., Jr. (1943) The ionization constant of carbonic acid in water and the solubility of carbon dioxide in water and aqueous salt solutions from 0 to 50˚. Journal of the American Chemical Society, 65, 2030-2037.  113  
 Hendy, C.H. (1971) The isotopic geochemistry of speleothems – I. The calculation of the effects of different modes of formation on the isotopic composition of speleothems and their applicability as palaeoclimatic indicators. Geochimica et Cosmochimica Acta, 35, 801-824. Hostetler, P.B., Coleman, R.G., Mumpton, F.A., Evans, B.W. (1966) Brucite in alpine serpentinites. American Mineralogist, 51, 75-98. Inaba, S., Minakawa, T., and Noto, S. (1985) Nesquehonite and dypingite from Shiraki, Mie Prefecture, Japan. Chigaku Kenkyu, 34, 281-287. Jiménez-López, C., Romanek, C.S., and Caballero, E. (2006) Carbon isotope fractionation in synthetic magnesian calcite. Geochimica et Cosmochimica Acta, 70, 1163-1171. Jiménez-López, C., Caballero, E., Huertas, F.J., and Romanek, C.S. (2001) Chemical, mineralogical and isotope behavior, and phase transformation during the precipitation of calcium carbonate minerals from intermediate ionic solution at 25˚C. Geochimica et Cosmochimica Acta, 65, 3219-3231. Jiménez-López, C., Romanek, C.S., Huertas, F.J., Ohmoto, H., and Caballero, E. (2004) Oxygen isotope fractionation in synthetic magnesian calcite. Geochimica et Cosmochimica Acta, 68, 3367-3377. Kakiuchi, M. and Matsuo, S. (1979) Direct measurements of D/H and  18  O/16O  fractionation factors between vapor and liquid water in the temperature range from 10˚ to 40˚. Geochemical Journal, 13, 307-311. Keeling, C.D., Bollenbacher, A.F., and Whorf, T.P. (2005) Monthly atmospheric 13C/12C isotopic ratios for 10 SIO stations. In Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tennessee, U.S.A. Kieffer, S.W. (1982) Thermodynamics and lattice vibration of minerals: 5. Application to phase equilibria, isotopic fractionation, and high pressure thermodynamic properties. Reviews in Geophysics and Space Physics, 20, 827-849. Kim, S.-T. and O’Neil, J.R. (1997) Equilibrium and nonequilibrium oxygen isotope effects in synthetic carbonates. Geochimica et Cosmochimica Acta, 61, 34613475.  114  
 Kosednar-Legenstein, B., Dietzel, M., Leis, A., and Stingl, K. (2008) Stable carbon and oxygen isotope investigation in historical lime mortar and plaster – Results from field and experimental study. Applied Geochemistry, 23, 2425-2437. Kralik, M., Aharon, P., Schroll, E., and Zachmann, D. (1989) Carbon and oxygen isotope systematics of magnesites: a review. In Magnesite: Geology, Mineralogy, Geochemsitry, Formation of Mg-Carbonates. Möller, P., Ed. Monograph Series on Mineral Deposits 28, 197-223. Land, L.S. (1998) Failure to precipitate dolomite at 25˚C from dilute solution despite 1000-fold oversaturation after 32 years. Aquatic Geochemistry, 4, 361-368. Lesniak, P.M. and Sakai, H. (1989) Carbon isotope fractionation between dissolved carbonate (CO32-) and CO2(g) at 25˚ and 40˚C. Earth and Planetary Science Letters, 95, 297-301. Léveillé, R.J., Longstaffe, F.J., and Fyfe, W.S. (2007) An isotopic and geochemical study of carbonate-clay mineralization in basaltic caves: abiotic versus microbial processes. Geobiology, 5, 235-249. Li, Q., Ding, Y., Yu, G., Li, C., Li, F., and Qian, Y. (2003) Fabrication of light-emitting porous  hydromagnesite  with  rosette-like  architecture.  Solid  State  Communication, 125, 117-120. Liu, B.-N., Zhou, X.-T., Cui, X.-S., and Tang, J.-G. (1990) Synthesis of lansfordite MgCO3·5H2O and its crystal structure investigation. Science in China, Series B, 33, 1350-1356. Markham, A.E. and Kobe, K.A. (1941) The solubility of carbon dioxide and nitrous oxide in aqueous salt solutions. Journal of the American Chemical Society, 63, 449-454. Matsuo, S., Friedman, I., Smith, G.I. (1972) Studies of Quaternary saline lakes. I. Hydrogen isotope fractionation in saline minerals. Geochimica et Cosmochimica Acta, 36, 427-435. McCrea, J.M. (1950) On the isotopic chemistry of carbonates and a paleotemperature scale. Journal of Chemical Physics, 18, 849-857. Melchiorre, E.B., Criss, R.E., and Rose, T.P. (2000) Oxygen and carbon isotope study of natural and synthetic azurite. Economic Geology, 95, 621-628.  115  
 Melchiorre, E.B., Criss, R.E., and Rose, T.P. (1999) Oxygen and carbon isotope study of natural and synthetic malachite. Economic Geology, 94, 245-260. Mickler, P.J., Stern, L.A., Banner, J.L. (2006) Large kinetic isotope effects in modern speleothems. Geological Society of America Bulletin, 118, 65-81. Mickler, P.J., Banner, J.L., Stern, L., Asmerom, Y., Edwards, R.L., and Ito, E. (2004) Stable isotope variations in modern tropical speleothems: Evaluating applications to paleoenvironmental reconstructions. Geochimica et Cosmochimica Acta, 68, 4381-4393. Mook, W.G., Bommerson, J.C., Staverman, W.H. (1974) Carbon isotope fractionation between dissolved bicarbonate and gaseous carbon dioxide. Earth and Planetary Science Letters, 22, 169-176. Murdock, J. (1954) Unit cell of hydromagnesite. American Mineralogist, 39, 24-29. Ohmoto, H. and Rye, R.O. (1979) Isotopes of sulfur and carbon. In Geochemistry of Hydrothermal Ore Deposits, Second Edition. Barnes, H.L., Ed. John Wiley & Sons, 509-567. O’Neil, J.R. and Barnes, I. (1971) C13 and O18 compositions in some fresh-water carbonates associated with ultramafic rocks and serpentinites: western United States. Geochimica et Cosmochimica Acta, 35, 687-697. O’Neil, J.R., Clayton, R.N., and Mayeda, T.K. (1969) Oxygen isotope fractionation in divalent metal carbonates. Journal of Chemical Physics, 51, 5547-5558. Onda, K., Sada, E., Kobayashi, T., Kito, S., and Ito, K. (1970) Salting-out parameters of gas solubility in aqueous salt solutions. Journal of Chemical Engineering of Japan, 3, 18-24. Patterson, W.P., Smith, G.R., and Lohmann, K.C. (1993) Continental paleothermometry and seasonality using the isotopic composition of aragonitic otoliths of freshwater fishes. In Climate Change in Continental Isotopic Records. Swart, P.K., Lohmann, K.C., McKenzie, J., and Savin, S., Eds. Geophysical Monograph Series, 78, 181-202. Power, I.M., Wilson, S.A., Thom, J., Dipple, G.M., and Southam, G. (2007) Biologically induced mineralization of dypingite by cyanobacteria from an alkaline wetland near Atlin, British Columbia, Canada. Geochemical Transactions, 8, article 13.  116  
 Power, I.M., Wilson, S.A., Thom, J.M., Dipple, G.M., Gabites, J.E., and Southam, G. (2009) The hydromagnesite playas of Atlin, British Columbia, Canada: A biogeochemical model for CO2 sequestration. Chemical Geology, 260, 286-300. Raade, G. (1970) Dypingite, a new hydrous basic carbonate of magnesium, from Norway. American Mineralogist, 55, 1457-1465. Révész, K.M. and Landwehr, J.M. (2002) δ13C and δ18O isotopic composition of CaCO3 measured by continuous flow isotope ratio mass spectrometry: statistical evaluation and verification by application to Devils Hole core DH-11 calcite. Rapid Communications in Mass Spectrometry, 16, 2102-2114. Robie, R.A. and Hemingway, B.S. (1972) The heat capacities at low-temperatures and entropies at 298.15 K of nesquehonite, MgCO3·3H2O, and hydromagnesite. American Mineralogist, 57, 1768-1781. Romanek, C.S., Grossman, E.L., and Morse, J.W. (1992) Carbon isotopic fractionation in synthetic aragonite and calcite: effects of temperature and precipitation rate. Geochimica et Cosmochimica Acta, 56, 419-430. Romanek, C.S., Jiménez-López, C., Navarro, A.R., Sánchez-Román, M., Sahai, N., and Coleman, M. (2009) Inorganic synthesis of Fe-Ca-Mg carbonates at low temperature. Geochimica et Cosmochimica Acta, 73, 5361-5376. Rubinson, M. and Clayton, R.N. (1969) Carbon-13 fractionation between aragonite and calcite. Geochimica et Cosmochimica Acta, 33, 997-1002. Rustad, J.R., Nelmes, S.L., Jackson, V.E., and Dixon, D.A. (2008) Quantum-chemical calculations of carbon-isotope fractionation in CO2(g), aqueous carbonate species, and carbonate minerals. Journal of Physical Chemistry A, 112, 542-555. Salata, G.G., Roelke, L.A., and Cifuentes, L.A. (2000) A rapid and precise method for measuring stable carbon isotope ratios of dissolved inorganic carbon. Marine Chemistry, 69, 153-161. Schauble, E.A., Ghosh, P., and Eiler, J.M. (2006) Preferential formation of 13C-18O bonds in carbonate minerals, estimated using first-principles lattice dynamics. Geochimica et Cosmochimica Acta, 70, 2510-2529.  117  
 Sheppard, S.M.F. and Schwarcz, H.P. (1970) Fractionation of carbon and oxygen isotopes and magnesium between coexisting metamorphic calcite and dolomite. Contributions to Mineralogy and Petrology, 26, 161-198. Shiro, Y. and Sakai, H. (1972) Calculation of the reduced partition function ratios of α-, β-quartzs and calcite. Bulletin of the Chemical Society of Japan, 45, 2355-2359. Skidmore, M., Sharp, M., and Tranter, M. (2004) Kinetic isotopic fractionation during carbonate dissolution in laboratory experiments: Implications for detection of microbial CO2 signatures using δ13C-DIC. Geochimica et Cosmochimica Acta, 68, 4309-4317. Spötl, C. and Burns, S.J. (1994) Magnesite diagenesis in redbeds: a case study from the Permian of the Northern Calcareous Alps (Tyrol, Austria). Sedimentology, 41, 543-565. Stephan,  G.W.  (1974)  De  Kristalstructuur  van  Enige  Magnesiumcarbonaten.  Unpublished Ph.D. thesis, The University of Amsterdam, Amsterdam, The Netherlands. Stephan, G.W. and MacGillavry, C.H. (1972) The crystal structure of nesquehonite, MgCO3·3H2O. Acta Crystallographica, Section B: Structural Crystallography and Crystal Chemistry, B28, 1031. Stiller, M., Rounick, J.S., and Shasha, S. (1985) Extreme carbon-isotope enrichments in evaporating brines. Nature, 316, 434-435. Suzuki, J. and Ito, M. (1973) A new magnesium carbonate hydrate mineral, Mg5(CO3)4(OH)2·8H2O, from Yoshikawa, Aichi Prefecture, Japan. Journal of the Japanese Association of Mineralogists, Petrologists and Economic Geologists, 68, 353-361. Tarutani, T., Clayton, R.N., and Mayeda, T.K. (1969) The effect of polymorphism and magnesium substitution on oxygen isotope fractionation between calcium carbonate and water. Geochimica et Cosmochimica Acta, 33, 987-996. Thompson, J.B. and Ferris, F.G. (1990) Cyanobacterial precipitation of gypsum, calcite, and magnesite from natural alkaline lake water. Geology, 18, 995-998. Turner, J.V. (1982) Kinetic fractionation of carbon-13 during calcium carbonate precipitation. Geochimica et Cosmochimica Acta, 46, 1183-1191.  118  
 Van Strydonck, M.J.Y., Dupas, M., and Keppens, E. (1989) Isotopic fractionation of oxygen and carbon in lime mortar under natural environmental conditions. Radiocarbon, 31, 610-618. Vasconcelos, C., McKenzie, J.A., Warthmann, R., and Bernasconi, S.M. (2005) Calibration of the δ18O paleothermometer for dolomite precipitated in microbial cultures and natural environments. Geology, 33, 317-320. Vogel, J.C., Grootes, P.M., and Mook, W.G. (1970) Isotopic fractionation between gaseous and dissolved carbon dioxide. Zeitschrift für Physik, 230, 225-238. Wendt, L. (1968) Fractionation of carbon isotopes and its temperature dependence in the system CO2-gas-CO2 in solution and HCO3-CO3 in solution. Earth and Planetary Science Letters, 4, 64-68. White, W.B. (1971) Infrared characterization of water and hydroxyl ion in the basic magnesium carbonate minerals. American Mineralogist, 56, 46-53. Widory, D. and Javoy, M. (2003) The carbon isotope composition of atmospheric CO2 in Paris. Earth and Planetary Science Letters, 215, 289-298. Wilson, S.A., Raudsepp, M., and Dipple, G.M. (2006) Verifying and quantifying carbon fixation in minerals from serpentine-rich mine tailings using the Rietveld method with X-ray powder diffraction data. American Mineralogist, 91, 1331-1341. Wilson, S.A., Dipple, G.M., Power, I.M., Thom, J.M., Anderson, R.G., Raudsepp, M., Gabites, J.E., and Southam, G. (2009) Carbon dioxide fixation within mine wastes of ultramafic-hosted ore deposits: Examples from the Clinton Creek and Cassiar chrysotile deposits, Canada. Economic Geology, 104, 95-112. Xiong, Y. and Lord, A.S. (2008) Experimental investigations of the reaction path in the MgO-CO2-H2O system in solutions with various ionic strengths, and their applications to nuclear waste isolation. Applied Geochemistry, 23, 1634-1659. Yasunishi, A. and Yoshida, F. (1979) Solubility of carbon dioxide in aqueous electrolyte solutions. Journal of Chemical & Engineering Data, 24, 11-14. Zachmann, D.W. and Johannes, W. (1989) Cryptocrystalline magnesite. In Magnesite: Geology, Mineralogy, Geochemsitry, Formation of Mg-Carbonates. Möller, P., Ed. Monograph Series on Mineral Deposits 28, 15-28.  119  
 Zedef, V., Russell, M.J., Fallick, A.E., and Hall, A.J. (2000) Genesis of vein stockwork and sedimentary magnesite and hydromagnesite deposits in the ultramafic terranes of southwestern Turkey: a stable isotope study. Economic Geology, 95, 429-446. Zhang, J., Quay, P.D., Wilbur, D.O. (1995) Carbon isotope fractionation during gaswater exchange and dissolution of CO2. Geochimica et Cosmochimica Acta, 59, 107-114. Zheng, Y.-F. (1999) Oxygen isotope fractionation in carbonate and sulfate minerals. Geochemical Journal, 33, 109-126.  120  Chapter 4 Carbon fixation in mineral waste from the Mount Keith Nickel Mine, Western Australia, Australia1  4.1. Introduction Mineralization of carbon dioxide (CO2) in ultramafic mine tailings has been proposed as a method by which the mining industry might reduce its greenhouse gas emissions (Wilson et al. 2006). Storage of CO2 within the crystal structures of carbonate minerals is recognized as a safe and durable method for the sequestration of anthropogenic carbon (Seifritz 1990; Lackner et al. 1995; Lackner 2003). Natural weathering of bedrock silicate minerals by dissolution can lead to precipitation of carbonate minerals under atmospheric temperature and pressure conditions. This weathering process is one of the most significant mechanisms for geochemical exchange of CO2 between the atmosphere and the terrestrial environment (Schwartzman and Volk 1989; Berner 1990). The considerable increase in mineral surface area afforded by mining and milling of ore can result in accelerated trapping of atmospheric CO2 within secondary carbonate minerals in the tailings from some ultramafic-hosted mining operations (Wilson et al. 2009a). Hydrated magnesium carbonate minerals are common low-temperature alteration products of serpentine-rich mine wastes (e.g., Raade 1970; Suzuki and Ito 1973, 1974; Wilson et al. 2006, 2009a, 2009b; Levitan et al. 2009). Previous studies of carbon mineralization in mine tailings have focussed on fixation of greenhouse gases within secondary carbonate minerals in the tailings of historical chrysotile mines (Wilson et al. 2006, 2009a) and active diamond mining operations in northern Canada (Rollo and Jamieson 2006; Wilson et al. 2009b). At the Mount Keith Nickel Mine, Western  1  A version of this chapter will be submitted for publication. Wilson, S.A., Dipple, G.M., Barker, S.L.L., Power, I.M., Fallon, S.J., Raudsepp, M., and Southam, G. Carbon fixation in mineral waste from the Mount Keith Nickel Mine, Western Australia, Australia.  121  Australia, Australia, the hydrated magnesium carbonate mineral hydromagnesite [Mg5(CO3)4(OH)2·4H2O] develops within mine tailings as a result of mineral processing. This hydromagnesite is being mineralized as a direct result of the mining operation at Mount Keith. Consequently, accounting of the amount of atmospheric CO2 that is being trapped and stored within this mineral could be used to reduce the mine’s greenhouse gas emissions in a carbon limited economy. Furthermore, accelerating the uptake of CO2 into mine tailings at Mount Keith could have the potential to completely offset the net greenhouse gas emissions of this mine. The mineral content of mine tailings is generally heterogeneous due to frequent changes in ore mineralogy, processing procedures, and tailings management practices. Mineralogy also tends to vary with time as tailings weather to produce secondary phases. As a result, geostatistical methods that rely on the presence of regular and predictable geological structures cannot be used to predict the mineralogy of a mine tailings pile. In order to estimate the total amount of CO2 captured within secondary carbonate mineral phases at a mine, it is necessary that the mineralogy of its tailings storage facilities be well constrained. This requires extensive sampling and the construction of a database of quantitative mineralogical data. Standard bulk geochemical methods (e.g., Leco analysis) for measuring abundance of CO2 cannot be used to quantify carbon trapping in mine tailings because they cannot distinguish between carbonate mineral species, nor can they discern the source of the carbon in minerals (i.e., atmospheric, bedrock, industrial, or organic). Recent advances in image analysis have improved the accuracy and efficiency of quantitative phase analysis with electron microbeam instruments (e.g., Fandrich et al. 2007). However, these techniques cannot be used to quantify the modal abundance of fine-grained minerals or hydrous minerals that are easily vaporized by an electron beam (e.g., hydrated carbonate minerals, hydrated sulphate minerals, and hydrotalcite-group minerals – all of which are common at Mount Keith). Quantitative crystallographic methods that use X-ray powder diffraction (XRPD) data are not limited by fine grain size and do not damage samples to the same extent as techniques that employ electron microbeams. Quantitative phase analysis using XRPD data provides a measure of the weight-percent contribution of each carbonate mineral in a sample. From a measure of  122  modal abundance, the amount of CO2 stored in specific carbonate minerals, which are known to be trapping atmospheric carbon, can be determined. Several methods of quantitative phase analysis with XRPD data have been tested and calibrated for the purpose of verifying and quantifying CO2 sequestration within mine tailings (Wilson et al. 2006, 2009b). We have used the Rietveld method and X-ray powder diffraction data (Rietveld 1969; Hill and Howard 1987; Bish and Howard 1988) to obtain quantitative mineralogical data for over 200 samples of Mount Keith tailings. Because the ages are known for many of the tailings flows at Mount Keith, we have been able to use our quantitative mineralogical results to obtain an empirical rate for hydromagnesite precipitation and to estimate the amount of CO2 that is being trapped and stored within these tailings. The three-isotope system (i.e., δ13C, δ18O, and F14C) employed by Wilson et al. (2009a) has been used to verify trapping of atmospheric CO2 within hydromagnesite and to provide more precise constraints on the amount of CO2 fixed within the tailings at Mount Keith.  4.2 Locality and sampling strategy 4.2.1 The Mount Keith Nickel Mine The MKD5 orebody at Mount Keith in the North Eastern Goldfields district of Western Australia (Fig. 4.1) is the largest nickel producer in Australia (Grguric 2003). The deposit at Mount Keith occurs in the NNW/SSE-trending Agnew-Wiluna greenstone belt in the Archaean Yilgarn Craton (Hill et al. 1990). The Agnew-Wiluna greenstone belt is at its narrowest in the vicinity of Mount Keith, with a breadth of six kilometres (Grguric et al. 2006), and is surrounded by Archaean granitoids to the east and west (Barrett et al. 1977). The MKD5 orebody is hosted by komatiitic peridotite (primarly dunite), which attained mid-upper greenschist facies as a result of regional metamorphism (Barrett et al. 1977). Retrograde serpentinization and carbonation of the host peridotites resulted from infiltration by H2O-CO2-rich fluids (Barrett et al. 1977; Grguric et al. 2006). Resulting metamorphic assemblages (from proximal to distal) are  123  (1) talc-magnesite, (2) antigorite-magnesite, and (3) lizardite-brucite-hydrotalcite group (Grguric et al. 2006).  Figure 4.1: (a) and (b) Location of Mount Keith Nickel Mine, Western Australia, Australia. (c) Satellite photograph of the open pit at the MKD5 deposit and tailings storage facility, Mount Keith.  Conventional, staged-cutback, open pit mining methods are practiced at MKD5, yielding approximately 11 Mt of ore annually (Grguric 2003). The mining operation at MKD5 produces approximately 370,000 t of greenhouse gases (cited as CO2 equivalent) and approximately 11 Mt of ultramafic tailings each year (BHP Billiton 2005). Ore from the MKD5 deposit is processed using froth flotation methods to concentrate sulphide minerals (Grguric et al. 2006). Additives used in processing include citric and sulphuric acids, guar gum, Na-dithionite, Na-ethyl xanthate, and (historically) soda ash. In 2004, ore reserves contained 0.52 wt.% nickel (Grguric et al. 2006), primarily in high-Ni pentlandite [(Fe,Ni)9S8], godlevskite [(Ni,Fe)9S8], heazlewoodite [Ni3S2], and millerite 124  [NiS]. Recovery of these minerals from the flotation circuit is typically about 70% (Grguric et al. 2006). The material rejected from the processing plant is piped to the second of two tailings storage facilities (TSF2, which is the only facility currently in operation), suspended in the hypersaline process water used in the flotation circuit. Tailings are deposited from spigots located on risers at nine points in TSF2 (Fig. 4.2).  Figure 4.2: The Tailings Storage Facilities (TSFs) at Mount Keith. The roughly circular facility is TSF2 and the smaller, adjoining facility is TSF1. Samples were collected at 59 locations in TSF1 and TSF2. Sampling in TSF2 was done along the perimeter of the facility and on both sides of a radial access road. Sampling in TSF1 was done proximal to the access road in Cell 2 (the right-most cell of TSF1). White circles indicate sites sampled in 2005; yellow circles indicate sites sampled in 2006; red borders on circles indicate sites for which Rietveld refinements have been done; white triangles indicate locations of risers. Yellow shading denotes regions of the TSFs in which deposition of tailings had ceased as of 2006.  125  4.2.2 Strategy for sampling at Mount Keith Limited sampling of the tailings storage facilities at Mount Keith was begun in April 2005. More extensive sampling was done in September and October of 2006. At this time, in excess of 800 samples were collected from TSF2 and the older, nowdecommissioned TSF1. Ideally, a tailings storage facility would be sampled randomly using a grid system; however, modern tailings management practices make this problematic if not impossible. Tailings saturated with water are more easily transported from a processing plant to a storage facility through a network of pipes. All tailings at Mount Keith have been and continue to be transported and deposited as slurry to the tailings storage facilities. Also, subaqueous storage of mine tailings is an effective strategy for preventing acid mine drainage and metal leaching from potentially acid-generating mine tailings (e.g., Robertson et al. 1997; Jambor and Blowes 1998; Price and Errington 1998). Consequently, it is common practise for many mining operations to retain some amount of process water in their stored tailings, either as a result of transportation by water or to mitigate acid generation. At the time of sampling, large regions of TSF2 were saturated with process water. These regions of the tailings facility were not accessible for sampling, making it impossible to collect a random set of samples on a grid. In order to compensate for incomplete access to the tailings at Mount Keith, sampling of the main tailings storage facility (TSF2) was done along a radial maintenance road and around the perimeter of the roughly circular facility (Fig. 4.2). Sampling was done at random intervals along these two paths in TSF2 and within tailings situated along the access road of Cell 2 of the older, non-operational TSF1. Tailings were primarily sampled by coring on 5 m x 5 m and 10 m x 10 m grids (after Roselle et al. 1999) and from vertical profiles on exposed surfaces. A total of 501 samples were collected from 11 grids, three profiles, and one trench excavated using a backhoe. Grid sampling was undertaken to assess local homogeneity of tailings by sampling in a standard pattern, with cores taken at regular intervals of distance (0.25 m and 2.50 m) and depth (10-cm long samples centred at 5, 25, 50, 75, 100, and 125 cm depth). Profiles extended sampling to depths beyond the range accessible using a sediment-coring device to a maximum depth of approximately 4 m. One backhoe  126  excavation allowed otherwise inaccessible tailings from sampling site 30 (Fig. 4.2), which is near the Kidney Wall in TSF2, to be sampled to a depth of ~1.8 m. Twenty-six cores were sampled at random locations, providing an additional 144 samples. Seventynine samples of efflorescent crusts were collected in association with the grids, profiles, and individual cores. Approximately 820 samples of tailings, biofilms, water, and process chemicals were collected from Mount Keith in 2005 and 2006. Of these, 724 samples of mine tailings are suitable for quantitative phase analysis. Quantitative phase analysis with the Rietveld method was done on 204 of these. Sampling locations were selected for Rietveld refinement in a way that optimized coverage of the accessible regions of both tailings storage facilities. A random number generator was used to select subsets of samples collected from 5 m x 5 m and 10 m x 10 m grids for quantitative phase analysis. Sampling did not extend to the base of the tailings deposits in most regions of TSF1 and TSF2 because depths greater than one to four metres were inaccessible using the collection techniques available (i.e., sampling with a sediment coring device, profiling, and limited access to a backhoe). In many of the thinner flows of tailings around the exterior of TSF2, the base of the deposits was accessible. However, basal deposits in the deeper, central regions of TSF1 and TSF2 could not be sampled.  4.3 Analytical methods 4.3.1 Qualitative X-ray powder diffraction methods Qualitative mineralogy was done on selected specimens of efflorescent mineral crusts and microsamples from Mount Keith that were picked with fine tweezers under a binocular microscope. This was done in order to identify sulphate and halide minerals that were present at low abundances within the tailings. Qualitative XRPD was also done on acid-treated specimens of mine tailings prior to stable isotopic analysis. Finely ground aliquots of sample were smear-mounted onto petrographic slides with anhydrous ethanol and allowed to dry at room temperature. XRPD data for mineral identification were collected with a scanning step of 0.04° 2θ and counting time of 0.5 s/step on a Siemens  127  (Bruker) D5000 θ-2θ diffractometer equipped with a VÅNTEC-1 detector. The long, fine-focus Co X-ray tube was operated at 35 kV and 40 mA and an Fe monochromator foil was employed. Mineral phases were identified with reference to the ICDD PDF-4+ database using DIFFRACplus EVA Version 10.0 (Bruker AXS 2004). 4.3.2 Quantitative X-ray powder diffraction and Rietveld refinement Quantitative phase analysis with the Rietveld method (Rietveld 1969; Hill and Howard 1987; Bish and Howard 1988) was done on 204 samples of mine tailings from the Mount Keith Nickel Mine. On return to the laboratory, samples were left in a drying hood under ambient conditions for a minimum of one week and were then homogenized mechanically with a spatula. From each dried and homogenized sample, a 50 g to 100 g subsample was taken and pulverized using a tungsten carbide ringmill. Three-gram aliquots of pulverized sample were weighed to a precision of ± 0.001 g. Ten weightpercent of annealed synthetic fluorite (an in-house standard) was added to each threegram aliquot as an internal standard. Subsequently, the aliquots were ground for 10 minutes in a McCrone micronizing mill to reduce the mean grain size and ensure homogenization. Grinding was done under anhydrous ethanol using agate grinding elements. Ground samples were dried at room temperature under a fume hood and were disaggregated using an agate mortar and pestle once dry. A specimen was prepared from each sample by mounting the powder into a backloading aluminum cavity holder of the design described by Raudsepp and Pani (2003). Powdered samples were loaded against the roughened surface of a sheet of glass that covered the top of the cavity. This was done to reduce preferred orientation of crystallites, particularly those of phyllosilicate minerals, hydrotalcite-group minerals, and hydromagnesite. X-ray powder diffraction data were collected on the Siemens (Bruker) D5000 θ-2θ diffractometer under the same operating conditions used to collect data for qualitative mineralogy. Data for Rietveld refinement were collected with a step size of 0.04° 2θ and counting time of 1s/step over a range of 3-80° 2θ. Rietveld refinements were done with Rietveld refinement software Topas Version 3 (Bruker AXS 2004) using the fundamental parameters approach (Cheary and Coelho  128  1992). Sources of crystal structure data for detectable mineral phases in the tailings samples are listed in Table 4.1. Table 4.1: Sources of crystal structure data for Rietveld refinement. Mineral  Source  Antigorite  Uehara (1998)  Blödite  Vizcayno and Garcia-Gonzalez (1999)  Brucite  Catti et al. (1995)  Calcite  Maslen et al. (1995)  Dolomite  Ross and Reeder (1992)  Epsomite  Calleri et al. (1984)  Fluorite  Batchelder and Simmons (1964)  Gypsum  Cole and Lancucki (1974)  Halite  Nickels et al. (1949)  Hexahydrite  Zalkin et al. (1964)  Hydromagnesite  Akao and Iwai (1977)  Iowaite  Braithwaite et al. (1994)  Kaolinite  Bish (1993)  Konyaite  Leduc et al. (2009)  Lepidocrocite  Zhukhlistov (2001)  Lizardite-1T  Mellini and Viti (1994)  Löweite  Fang and Robinson (1970)  Magnesite  Markgraf and Reeder (1985)  Magnetite  Tsukimura et al. (1997)  Muscovite-2M1  Liang and Hawthorne (1996)  Plagioclase (An91)  Wenk and Kroll (1984)  Quartz  Glinnemann et al. (1992)  Rutile  Meagher and Lager (1979)  Talc  Perdikatsis and Burzlaff (1981)  Trona  Pertlik (1986)  Vermiculite  Shirozu and Bailey (1966)  Woodallite  Grguric et al. (2001) and Braithwaite et al. (1994)  129  The method of Wilson et al. (2006), initially developed for use with chrysotilerich mine tailings, was used to compensate for structural disorder in lizardite and antigorite during refinement. This method was readily applied, with very little change of procedure, because of the mineralogical similarity between chrysotile mine tailings and the tailings from Mount Keith. Mineralogy and relative abundances of the major and minor phases in the tailings at Mount Keith are comparable to those for chrysotile mine tailings. Tailings from chrysotile mines and tailings from Mount Keith both contain one or more serpentine minerals, hydrotalcite-group minerals, minor spinel minerals (i.e., magnetite and/or chromite), occasional hydrated magnesium sulphate minerals, and low but detectable amounts of primary (i.e., bedrock) carbonate minerals and secondary precipitates of hydrated magnesium carbonate minerals. As a result, we were able to apply this established procedure for quantitative mineralogical analysis to samples from Mount Keith. Peak intensities for each serpentine mineral were extracted using the Pawley method (Pawley 1981), independent of atomic scattering, from X-ray powder diffraction patterns of high-purity standards. Fourth-order symmetrized harmonics were used to model anisotropic peak shape in lizardite and antigorite (Järvinen 1993). Without reference to atomic positions, the relative peak intensities of lizardite and antigorite are unconstrained. To compensate, the relative intensities of the serpentine peaks phases were initially held constant to prevent their peaks from interfering with those from other phases. The relative intensities of the peaks for lizardite and antigorite were refined only after the peaks of the other phases had been fitted. The weight-percent abundances of antigorite and lizardite are given together, because the method of Wilson et al. (2006) cannot quantify more than one disordered phase per sample. Several poorly defined minerals are common constituents of Mount Keith tailings:  mountkeithite  {(Mg,Ni)11(Fe3+,Cr3+,Al)3[(OH)24|(SO4,CO3)3.5]⋅11H2O}  and  woodallite [Mg6Cr3+2(Cl2)(OH)16⋅4H2O]. Mountkeithite, occasionally detected in XRPD patterns, is generally present at abundances too low to refine reliably because only its most intense reflection is detectable at ~3σ above background. Woodallite is a minor phase in most samples and, to date, the details of its crystal structure have not been published. However, because of the known structural similarity and the solid solution series between woodallite and iowaite, the structure of woodallite can be approximated 130  using a modified structure based on that of iowaite (using the structure reported by Braithwaite et al. 1994). With this in mind, we replaced the lattice parameters of iowaite [Mg6Fe3+2(Cl2)(OH)16⋅4H2O] with those given for woodallite by Grguric et al. (2001) and substituted the scattering factor for the Fe3+ in the structure of iowaite with that for Cr3+. Woodallite and iowaite in the ore from Mount Keith contain variable amounts of sulphate, carbonate, and chloride in their interlayer sites (Grguric 1999, 2003; Grguric et al. 2001, 2006; Woodhouse 2006). The grain-by-grain variability of chemical species in these interlayer sites commonly gives rise to broad, severely overlapped peaks for woodallite and iowaite. This makes it difficult to distinguish these mineral phases from XRPD patterns without resorting to the use of small angular steps and long counting times to acquire high-resolution XRPD data. Because of this, refined abundances for woodallite and iowaite are given here as a single value. Similarly, due to the structural similarity between magnetite and chromite and the difficulty in distinguishing between them using XRPD data, these phases have been refined using only the structure of magnetite and their abundances are reported together. Second or third-order Chebychev polynomials were used to model the background and a 1/x term was included to assist in fitting the background curve at low angles of diffraction. Cell parameters and Lorentzian crystallite size were refined for all phases and zero error was refined for each XRPD pattern. The method of Brindley (1945) was used to correct for microabsorption contrast between phases, assuming a mean particle radius of 2.5 µm. Preferred orientation of hydrotalcite-group minerals was modeled using the March-Dollase correction (March 1932; Dollase 1986). Quantification of hydromagnesite is of primary importance for estimates of carbon mineralization at Mount Keith. Unfortunately, trace amounts of this mineral are not always measurable using the methodology we employed to collect XRPD data for Rietveld refinement. The detection limit for hydromagnesite was determined to be approximately 0.5 wt.%. Hydromagnesite was successfully refined and quantified at an abundance of ~0.5 wt.% for only one sample, although it was present at close to this value in 22 of 204 samples for which Rietveld refinements were done. In this sample (06MKG10-4-3), the refined cell parameters of hydromagnesite gave reasonable values but the Lorentzian crystallite size refined for hydromagnesite gave an unrealistic value in  131  excess of 1000 nm. Although the most intense peaks of hydromagnesite (the overlapping 011 and 0-1-1 reflections) were detected at > 3σ of the background intensity for 22 samples, our attempts at including (and thereby quantifying) hydromagnesite in 21 of these refinements failed. Hard constraints on cell parameters and crystallite size were required to produce a reasonable fit to the observed data. Because constraints of this sort can bias Rietveld refinements, the crystal structure of hydromagnesite was removed from final refinements and a value of 0.5 wt.% (i.e., the detection limit) was imputed for the abundance of hydromagnesite in these samples. The numbers of samples for which hydromagnesite was quantifiable, imputed, and below detection are summarized in Table 4.2. Although there is error attached to imputed values for hydromagnesite, we consider this to be of an acceptable magnitude because it is almost certainly less than the error attached to refined values for mineral abundances less than 1 wt.% (which can exceed 50% relative, Wilson et al. 2006; 2009b). Table 4.2: Numbers of samples for which hydromagnesite is quantifiable, imputed (at detection), or below detection. Sample Subset  Quantified  All samples  116  At detection (0.5 wt.%) 21  Samples included in statistical analyses  111  20  Below detection (<0.5 wt.%) 67  Total  65  196  204  4.3.3 Scanning electron microscopy Mineral habits and the textural relationships among minerals were characterized using a Philips XL-30 scanning electron microscope (SEM), equipped with a Princeton Gamma-Tech energy dispersive X-ray spectrometer (EDS) system. Selected samples of tailings were impregnated with epoxy and thin-sectioned for petrographic analysis. Backscattered electron (BSE) imaging was used to observe textural relationships in thin section and EDS was used for the identification of minerals.  132  4.3.4 Stable isotopic methods A total of 100 specimens, taken from 45 geological samples and five samples of water, were analyzed for their stable carbon and oxygen isotopic compositions. Eighteen replicate analyses were done to assess reproducibility of the data. Thirty-two of these specimens of tailings were analyzed in bulk before and after treatment with one normal hydrochloric acid (1 N HCl) at room temperature (i.e., approximately 25ºC). Acid treatment was used to preferentially dissolve hydromagnesite, hydrotalcite-group minerals, and bedrock calcite and dolomite. The untreated specimens reflect the bulk isotopic compositions of all carbonate minerals in the tailings (i.e., bedrock and secondary carbonate minerals), while the acid-treated specimens reflect the isotopic composition of the more resistant bedrock magnesite. Stable isotopic data for the resulting 64 specimens, 16 other specimens of mine tailings, and the dissolved inorganic carbon (DIC) in five specimens of water, were collected using standard procedures for stable carbon and oxygen isotopic analyses (described in Subsections 4.3.4.1 and 4.3.4.2) employed at the Pacific Centre for Isotopic and Geochemical Research (PCIGR), The University of British Columbia. The stable isotopic compositions of hydromagnesite in 15 specimens of mine tailings were analyzed using the method for selective acid extraction of Barker et al. (2009), which is described in Subsection 4.3.4.3. 4.3.4.1 Standard methodology for carbonate minerals Stable carbon and oxygen isotopic compositions of bulk samples of tailings and acid-treated tailings were determined using the standard procedure for routine analysis at PCIGR. Stable isotopic compositions were also measured in this way for highly pure subsamples of carbonate minerals that were picked with fine tweezers under a binocular microscope. A total of 80 aliquots of bulk tailings and mineral separates were analyzed using this standard methodology. Specimens were analyzed using a gas bench attached to a Thermo Finnigan DeltaPlus XL isotope ratio mass spectrometer (IRMS). Aliquots weighing approximately 200 µg were loaded into septum vials, flushed with helium, and dissolved in 99% phosphoric acid at 72ºC for a minimum of one hour. Isotopic compositions of the evolved headspace gas were measured in a helium flow. The δ13C compositions are given  133  relative to Vienna Pee Dee Belemnite (VPDB) and the δ18O compositions are relative to Vienna Standard Mean Ocean Water (VSMOW). The external precision (1σ deviation) for isotopic analyses was < 0.1‰ δ13C and < 0.2‰ δ18O, as estimated from repeated analysis of NBS-18, NBS-19, and in-house calcite standards (which are calibrated against the NBS standards). At least three standards were analyzed for every eight specimens. As a result of their relative scarcity, fewer studies of stable isotopes in anhydrous magnesium carbonate minerals (e.g., magnesite, huntite, and dolomite) have been undertaken than for calcium carbonate minerals (Kralik et al. 1989). Even fewer detailed stable isotopic studies have been attempted for the hydrated magnesium carbonate minerals (i.e., O’Neil and Barnes 1971; Grady et al. 1989; Braithwaite and Zedef 1996; Zedef et al. 2000; Léveillé et al. 2007; Power et al. 2007, 2009; Wilson et al. 2009a). Because these minerals have not been studied thoroughly, the factors required to correct for fractionation of  18  O between hydromagnesite and phosphoric acid or other hydrated  magnesium carbonate minerals and phosphoric acid are not known. As such, the acid fractionation factor for magnesite was used as a proxy for hydromagnesite. The δ18O values of calcite, dolomite, and magnesite were corrected for reaction with phosphoric acid using the fractionation factors from Das Sharma et al. (2002), taking into account the approximate proportions of calcium and magnesium carbonate minerals in mixed specimens. 4.3.4.2 Standard procedure for dissolved inorganic carbon Stable carbon isotopic compositions of total dissolved inorganic carbon (DIC) in five specimens of water from Mount Keith were determined using the Thermo Finnigan DeltaPlus XL IRMS. For each specimen, four to five drops of 99% phosphoric acid were loaded into an exetainer, which was then sealed with a rubber septum and flushed with helium for five minutes. A syringe was used to inject 600 µL of each water specimen into a sealed exetainer. The exetainer was shaken and the specimen left to react for a minimum of one hour at 25ºC. As with specimens of carbonate minerals, the isotopic composition of the evolved headspace gas was measured in a helium flow. The δ13C compositions are given relative to VPDB. The external precision (1σ deviation) for  134  isotopic analyses was < 0.1‰ δ13C, as estimated from repeated analysis of NBS-18, NBS-19, and in-house calcite standards. 4.3.4.3 Procedures for selective analysis of carbonate minerals The disseminated nature of hydromagnesite within Mount Keith tailings hinders isotopic analysis using standard analytical procedures. Efforts to manually separate pure hydromagnesite from these tailings have been unsuccessful. Alternatively, bulk dissolution of tailings extracts CO2 from all carbonate minerals in a sample, including secondary carbonate minerals that may be storing atmospheric CO2, bedrock carbonate minerals, and carbonate-bearing hydrotalcite-group minerals. Interpretation of stable carbon and oxygen isotopic data requires accounting of all carbonate-bearing mineral species due to the variability in isotopic fractionation with mineralogy. Stable isotopic results are more easily interpreted for single-phase specimens. Selective analysis of minerals allows the importance of different mineral species to CO2 sequestration to be assessed. Hydrated magnesium carbonate minerals like hydromagnesite are susceptible to dissolution in weak acids (e.g., Power et al. 2007; Wilson et al. 2009a), which can be used to leach them preferentially from samples of tailings. We have used the method of Barker et al. (2009) to determine the carbon and oxygen isotopic compositions of both the secondary hydromagnesite and bedrock magnesite in mine tailings from Mount Keith. This method exploits the difference in reactivity of hydrous and anhydrous magnesium carbonate minerals with phosphoric acid at room temperature (after Epstein et al. 1964; Al-Aasm et al. 1990). The hydrated magnesium carbonate minerals react quickly after addition of phosphoric acid to a bulk sample of pulverized tailings. Within the first 15 to 30 minutes of room-temperature reaction, most of the CO2 evolved from a bulk sample is derived from decomposition of hydrated magnesium carbonate minerals. The evolved CO2 can then be collected using a vacuum line and analyzed to determine the isotopic composition of the hydrated magnesium carbonate minerals in a sample. Once the hydrated magnesium carbonate minerals have reacted completely with the acid, the remainder of the evolved gas can be evacuated, and the acidified sample heated, in order to collect CO2 from the bedrock carbonate minerals in the tailings. Using this  135  method, the CO2 from hydromagnesite and magnesite (± calcite and dolomite) can be analyzed separately for stable carbon and oxygen and radiocarbon isotopic compositions. The resulting isotopic analyses give a more complete description of sources and cycles for CO2 in carbonate minerals than can be gained from bulk analysis of mine tailings. For each of 15 samples of pulverized tailings from Mount Keith, an aliquot was weighed out. The size of the aliquot was dependent upon (1) whether the sample was for stable isotopic or radiocarbon analysis (substantially more sample is required for radiocarbon analysis) and (2) the total percent by weight (wt.%) of carbonate minerals in each sample. Aliquots for stable isotopic analysis were weighed out such that each would provide at least 5 mg (of calcite-equivalent CO2 content) of both hydromagnesite and magnesite. The amount of material required was calculated from the results of quantitative phase analysis with the Rietveld method. Each aliquot was deposited into a sterile, unlined 10 mL BD Vacutainer® with a resealable, 8-mm thick rubber septum. A small amount of glass wool was inserted into the top of each vial. This was done to contain the pulverized tailings within the bottom portion of the vial during reaction with phosphoric acid. It was also used to prevent any potential contamination of the vacuum line by finely powdered material. Following the method of Barker et al. (2009), evacuated specimens were acidified with 99% phosphoric acid and allowed to react at 25ºC for 15-30 minutes. The evolved headspace gas from each vial was collected in the vacuum line described by Barker et al. (2009) for approximately 5 minutes and allowed to remain in contact with a water trap. Evolved gas was passed through a piece of glass tubing, that was cooled by a slush of anhydrous ethanol and dry ice, in order to trap evolved water (as ice) but let pass CO2. Carbon dioxide was collected at the base of a 6-mm diameter glass ampoule, which was chilled in a bath of liquid nitrogen (T = 77 K) so that CO2 could be collected in the solid state. A handheld propane torch was used to weld and seal the top of each ampoule so that the CO2 could be contained and stored prior to analysis. Following collection of CO2 evolved from hydromagnesite, the room-temperature reaction was allowed to continue for two hours until dolomite, calcite, hydrotalcite-group minerals, and any residual hydromagnesite had reacted completely. Although hydrotalcite-group minerals are less susceptible to dissolution in weak acids than  136  hydromagnesite, they are prone to exchange of interlayer anions in some weakly acidic solutions (e.g., Bish 1980). There is also evidence that anionic clays (like the hydrotalcite-group minerals) are able to take up phosphate ions into their interlayers (e.g., Badreddine et al. 1999). The rate at which iowaite and woodallite dissolve in phosphoric acid, and the extent to which they may exchange interlayer CO32- and Cl- for phosphate ions, is not known. The evolved gas from the reaction of hydromagnesite with phosphoric acid was removed from each vial using the vacuum line and vacuum was restored within the sample tubes. The evacuated vials were heated, in a 1 L bath of water on a laboratory heating plate, to 75ºC for 1 hour. Evolved headspace gas produced by reaction of phosphoric acid with magnesite was collected for approximately 5 minutes using the vacuum line. Carbon dioxide evolved from magnesite was trapped and stored within 6mm diameter glass ampoules using the same procedure as for hydromagnesite. Specimens were