<<

FACULTEIT WETENSCHAPPEN Vakgroep Geologie en Bodemkunde

Geochemistry of the Brent , ,

Bart Vleminckx

Academiejaar 2010–2011

Scriptie voorgelegd tot het behalen van de graad Van Master in de Geologie

Promotor: Prof. Dr. Ph. Claeys Co-promotor: Prof. Dr. M. Elburg Begeleider: Drs. S. Goderis Leescommissie: Prof. Dr. P. Van den haute, Prof. Dr. F. Vanhaecke

I. ACKNOWLEDGEMENTS

I would like to thank Prof. P. Claeys, Prof. M. Elburg and S. Goderis for their answers, remarks, contributions and making this master dissertation possible; J. Spray for providing the Brent samples; F. Paquay for providing the Os isotope data; V. Renson for providing the Pb isotope data; J. Sauvage for contributing to the platinum group element analysis preparations; J. Belza, E. De Pelsmaeker, D. Debruyne, T. Van der Gucht, and I. Smet for contributing to the major element analysis preparations; M. Van Tomme for everything.

Cover art: http://www.mnh.si.edu/earth/text/5_3_2_0.html

PAGE I II. NOTE ON STYLE

When the contents or structure of a (sub)section is in large measure extracted from one or several works, the references are given introductory to the section. In all other cases in‐text references are used. If not stated else, the reference of a figure’s caption is the same as that of the figure’s source. SI units and prefixes are used following BIPM [2006]. Large numbers are written by the short scale naming system and with the dot as decimal separator. Non‐SI units (with SI prefixes) used are astronomical distance in astronomical units (AU) as distance from the Sun (with set at 1 AU), geological date in annum (a) as years before present (i.e. 1950) and time period in minutes (min), hours (h) or years (y). Chemical element and compound notations are used following IUPAC [2007]. Other abbreviations used are wt.% for weight %, vol.% for volume %, USA for United States of America, USD for USA dollar, SD for standard deviation and SE for standard error. Calendar dates are in the international date format ISO [2004]. If not stated else, sizes of , craters and impact structures are expressed as their diameters.

PAGE II III. DUTCH SUMMARY

De eerste vaste stoffen van het zonnestelsel ontstonden 4.567 Ga geleden door condensatie uit de zonnenevel. Door elektromagnetische accretie vormden deze de planetesimalen. Sommige werden zo massief dat ze een kern en mantel met partiële opsmelting vormden. Omdat het zonnestelsel toen nog veel heter was door het samentrekken van de zonnenevel en de jonge zon, konden binnen een bepaalde afstand van de zon, de sneeuwlijn genaamd, geen H‐verbindingen condenseren in ijzen. De planetesimalen die hier ontstonden kregen daarom een samenstelling van voornamelijk metaal en gesteente, en worden asteroïden genoemd. Buiten de sneeuwlijn vormden zich planetesimalen die voornamelijk uit H‐ijzen bestaan met slechts een klein deel metaal en gesteente, kometen genaamd. Onder invloed van de zwaartekracht vormden deze asteroïden en kometen de planeten, dwergplaneten en manen.

Er bleven echter ook een aanzienlijk aantal planetesimalen over. Deze zijn niet willekeurig verspreid over het zonnestelsel te vinden, maar verblijven in bepaalde gebieden die begrens worden door zwaartekracht perturbaties van planeten. Zo zijn de meeste kometen te vinden voorbij de planeten in Oortwolk en Kuiper gordel. De asteroïden bevinden zich voornamelijk in de asteroïdengordel tussen 2.1 en 3.3 AU, maar ook andere verblijfsgebieden zijn bekend zoals de Hungarias, Hildas en Cybeles. Niet alle kometen en asteroïden blijven echter in deze gebieden. Diegene die het meest hun verblijfsgebied verlaten zijn de asteroïden uit de asteroïdengordel. Orbitale resonanties met Jupiter zorgen namelijk voor lege gebieden in deze gordel, Kirkwood gaten genoemd. Omloopbanen in deze gebieden hebben perioden gelijk aan een gehele breuk van Jupiters periode en ontvangen dus meer frequent zwaartekracht perturbaties. Door botsingen tussen asteroïden ontstaan fragmenten die deze Kirkwood gaten kunnen bereiken en zo de asteroïdengordel verlaten naar paden doorheen het zonnestelsel.

Zo ontvangt de Aarde een deel van deze fragmenten. Wanneer ze de atmosfeer binnen treden zal frictie ontstaan waardoor ze vertragen, opbranden en/of exploderen. Alleen diegene vanaf een bepaalde massa zullen het vaste oppervlak van de Aarde kunnen bereiken. Deze worden meteorieten genoemd. Meteorieten met massa’s vanaf ongeveer 10 Mg zullen zelfs een aanzienlijk deel van hun snelheid

PAGE III kunnen behouden en de Aarde inslagen met een dergelijke hoeveelheid kinetische energie, dat deze wordt omgezet in schokgolven. Dit zijn bijna discontinue spanningsgolven met extreme drukken en supersonische snelheden. De resulterende spanningen zijn zo hoog dat de complete meteoriet verdampt en smelt. Ook een hemisferisch volume van doelgesteente, met een straal van 3 tot 4 keer de diameter van de meteoriet, zal smelten. Een nog grotere hoeveelheid van doelgesteente zal geëjecteerd, verplaatst, gebrecciëert en/of gebroken worden. Uiteindelijk zullen de scholgolven een inslagkrater gevormd hebben die 20 tot 30 keer groter is dan de meteoriet. De diepte gaat van 1/3 van de diameter voor kleinere inslagkraters tot 1/6 voor de grotere. Doordat op het einde van de kratervorming instortingen plaatsvinden van de kraterranden, zal de krater bevatten die smelt fragmenten bevatten. Ook smelt lenzen komen voor met een hoog gehalte aan inslagsmelt. Door de oppervlakte processen van de Aarde zullen inslagkraters hun morfologie en lithologie niet behouden. De meesten zijn zelfs onherkenbaar vervormd door tektoniek of totaal geërodeerd.

Wat we nu terugvinden van inslagkraters, worden inslagstructuren genoemd. Op de Aarde zijn er nu 178 gekend. Ze hebben ouderdommen van recent als 1947 tot 2.4 Ga en afmetingen van 0.015 tot 300 km. Studies hebben aangetoond dat de inslagsmelt van deze inslagstructuren een meetbare meteoriet contributie bevat, meestal niet groter dan 0.1 wt.%. Deze is gedetecteerd op basis van bepaalde chemische signaturen die afwijken van het doelgesteente. Zo zijn de siderofiele elementen verarmd in de korst omdat ze eerst zijn aangerijkt in de kern en later in de mantel. Vele meteorieten vertonen deze verarming echter niet en slechts een kleine contributie kan een duidelijk verhoging veroorzaken door de lage achtergrond concentraties. Een onderscheid wordt gemaakt tussen de gemiddeld siderofiele elementen Ni, Co and Cr, and de hoog siderofiele platina groep elementen (PGE). Er bestaan ook isotopische signaturen zoals bijvoorbeeld die van Os, die het toelaten om zeer specifiek de hoeveelheid meteoriet contributie te bepalen. Op basis van al deze signaturen kan een inslagstructuur als zodanig geïdentificeerd worden. Meteorieten vertonen ook onderlinge chemische verschillen, en op basis van de siderofiele element verhoudingen kunnen deze bepaald worden. Hierdoor kan bepaald worden welk soort meteoriet de inslagstructuur heeft veroorzaakt.

De chondrieten zijn de eerste gevormde planetesimalen. Deze vertonen aquatische en thermische metamorfose, maar hebben geen opsmelting ondergaan. Ze worden onderverdeeld in de koolstofhoudende CI, CM, CR, CB, CV, CK, CO en CH groepen, de gewone H, L en LL groepen, en de enstatiet H en L groepen. Ze vertonen verschillende samenstelling naargelang hun afstand tot de zon. Zo zijn de enstatiet

PAGE IV chondrieten het dichts bij de zon gevormd, de gewone iets verder en de koolstofhoudende nog verder, vermoedelijk nabij de sneeuwlijn. Alle meteorieten van een bepaalde groep zijn fragmenten van hetzelfde moederlichaam. Grotere planetesimalen vormden net als de Aarde een kern en ondervonden partieel smelten. Zo zijn er de chondrieten, welke magmatische gesteenten zijn en moeilijk te onderscheiden doordat ze een gelijke geschiedenis hebben als de korst. De ijzermeteorieten zijn fragmenten van de kernen, die zijn vrijgekomen zijn door enorme botsingen.

De studie van inslag en meteoriet identificatie heeft zijn belang. Eerst en vooral hebben inslagstructuren een aanzienlijke economische waarde. Maar liefst 15 % van de inslagstructuren wordt winstgevend geëxploiteerd, waaronder een aantal van de grootste mineraalafzettingen ter wereld, maar ook voor olie, gas en/of bouwstenen. Maar inslagstructuren maken nu ook eenmaal, weliswaar een klein, deel uit van de Aarde en hun identificatie is belangrijk om ze te kunnen bestuderen en zo onze kennis van de Aarde te vervolledigen. Echter, op niet of minder actieve plaatsen in het zonnestelsel, maken inslagkraters het grootste deel van het vaste oppervlak uit. Omdat hun afstand hun studie bemoeilijkt kunnen we onze kennis van op de Aarde extrapoleren naar daar. Gezien het economisch potentieel, is hun studie ook van belang voor toekomstige exploratie van de ruimte. Meteorieten worden voor vele doeleinden bestudeerd. Ze bevatten zeldzame informatie over de vorming, componenten, structuur, dynamiek, fluxen, evolutie en huidige staat van het zonnestelsel. Meteorieten die inslagkraters veroorzaken vertegenwoordigen een meer massief bereik dan andere meteorieten. Hun beter hun identificatie, hoe beter hun fluxen, verblijfsgebied, en mogelijkse relaties en moederlichamen kunnen bepaald worden. Verder hebben inslagen ook biologische en maatschappelijke effecten.

Het opzet van deze masterproef is om een inslag en meteoriet identificatie toe te passen op Brent inslagstructuur. Hierbij zal nagegaan worden welke het nut is van de verschillende chemische signaturen en hoe ze kunnen ingebracht worden in een multi‐signatuur benadering. Twintig stalen werden verzameld van smelt‐fragment breccias, een smelt lens en het doelgesteente. Deze werden geanalyseerd voor hoofdelementen (n = 13), Ni, Co en Cr (n = 20) en PGE (n = 13).

Karakterisatie van de aanwezige alteratie toonde chloritisatie, Au‐verarming en K‐ aanrijking aan in de smelt‐fragment breccias. Terwijl dit de correlatie vernietigt tussen de mobiele Ni, Co en Cr, vertonen de immobiele PGE geen verandering. Ook de smelt lens toont alteratie, al zij het in mindere mate. Ni, Co en Cr tonen hier een

PAGE V goede correlatie. Het gebruik van het Os isotoop systeem leverde meteoriet contributie waarden op tot 1 a 2 gew.% voor het diepste smelt lens staal. Tezamen met de gemiddeld en hoog siderofiele elementen, weerlegt dit de mogelijkheid van een eerder voorgestelde fractionatie, en impliceert het een impact identificatie.

We weerleggen ook een aanname dat er geen (ultra)mafische component aanwezig is. Het is mogelijk dat smelt‐fragment breccias een dergelijke component bevat en zo mee de ontbrekende correlatie verklaart voor Ni, Co en Cr. De smelt lens bevat met zekerheid geen extra component. Een regionale studie van het doelgesteente zou deze mogelijkheid kunnen bevestigen of uitsluiten, en wanneer aanwezig, het effect op de smelt‐fragment breccias bepalen.

Op basis van een multi‐signatuur benadering door het combineren van de gemiddeld en hoog siderofiele elementen kan een precieze meteoriet identificatie worden bepaald tot op het groep niveau: een IA niet‐magmatische ijzermeteoriet. Terwijl de Ni/Cr, Co/Cr en Pd/Ir verhoudingen duiden op een LL of L gewone , zijn de Ni/Co, andere PGE en gecombineerde verhoudingen hier niet mee in overeenstemming. Gebaseerd op een lineair en vergroot PGE patroon, dat representatief wordt geacht voor de meteoriet, zijn enkel de niet‐magmatische IA en IIIC groepen mogelijk. Als alle siderofiele verhoudingen worden samengebracht, wijzen deze duidelijk op een IA.

De Brent inslag structuur is ideaal voor het bestuderen van de effecten van alteratie op inslag en meteoriet identificatie, zeker voor de smelt‐fragment breccias. Door zijn weerstand tegen alteratie en hoge meteoriet contributie levert de smelt lens de beste stalen voor inslag en meteoriet identificatie.

Voor meteoriet identificatie zijn de implicaties duidelijk. De precieze identificatie was niet mogelijk geweest zonder het gebruik van een multi‐signatuur benadering. Terwijl Ni, Co en Cr redelijk gemakkelijk te analyseren zijn, vertonen hun chondritische verhoudingen overlap met meerdere ijzer meteorieten. Langs de andere kant maken ze een verdere discriminatie mogelijk wanneer een chondriet of ijzermeteoriet is bepaald door de PGE.

Het Os isotoop systeem lijkt in staat om alleen een impact te identificeren, maar het omvat een omslachtige analyse. Wanneer deze data met de andere siderofiele elementen en hoofdelementen word gecombineerd, is het mogelijk om andere processen als alteratie of (ultra)mafische componenten, te identificeren.

Een algemene conclusie is dat voor impact en meteoriet identificatie best steeds een multi‐signatuur benadering kan gebruikt worden. Hoe meer data, hoe beter.

PAGE VI IV. CONTENTS

I. Acknowledgements ...... I II. Note on Style ...... II III. Dutch Summary ...... III

IV. Contents ...... VII

1. INTRODUCTION ______1

1.1. Scientific Setting ...... 2 1.1.1. The Impact Theory ...... 2 1.1.2. Impact Geochemistry ...... 3 1.2. Study Objectives ...... 4 1.3. Importance ...... 4 1.3.1. Economic ...... 4 1.3.1.1. Natural Resources ...... 5 1.3.1.2. Tourism ...... 6 1.3.2. Scientific ...... 6 1.3.2.1. The Solar System ...... 7 1.3.2.2. Biological Effects ...... 7 1.3.3. Social ...... 8

2. LITERATURE ______10

2.1. Impact Structures ...... 10 2.1.1. Impact Meteorites ...... 11 2.1.1.1. and Comets ...... 11 2.1.1.2. Residence Regions ...... 12 2.1.1.3. Voyage to the Earth ...... 13 2.1.1.4. Through the Earth’s Atmosphere ...... 14 2.1.2. Impact Craters ...... 15 2.1.2.1. Compression ...... 16 2.1.2.2. Excavation ...... 17 2.1.2.3. Modification ...... 18 2.1.2.4. Post‐impact Development ...... 19 2.1.3. ...... 21 2.1.4. Impact Rocks ...... 22 2.1.4.1. Basement ...... 22

PAGE VII 2.1.4.2. Crater‐fill ...... 23 2.1.4.3. Proximal‐ejecta ...... 25 2.1.4.4. Distal‐ejecta ...... 25 2.2. Classification ...... 26 2.2.1. ...... 28 2.2.1.1. Carbonaceous ...... 30 2.2.1.2. Ordinary ...... 31 2.2.1.3. Enstatite ...... 31 2.2.2. Non‐chondrites ...... 31 2.2.2.1. Achondrites ...... 32 2.2.2.2. Irons ...... 34 2.2.2.3. Stony‐irons ...... 35 2.2.3. Cometary Meteorites ...... 35 2.3. Geochemical Signatures ...... 36 2.3.1. Element Abundances ...... 37 2.3.1.1. Moderately Siderophile Elements ...... 37 2.3.1.2. High Siderophile Elements ...... 38 2.3.1.3. Major and Trace Elements ...... 39 2.3.2. Isotope Ratios ...... 39 2.3.2.1. Os Isotope System ...... 39 2.3.2.2. Cr Isotope System ...... 40 2.3.2.3. Other Isotope Systems ...... 40 2.3.3. Multi‐signature approach ...... 41 2.4. Brent Impact Structure ...... 41 2.4.1. Impact Dimensions ...... 43 2.4.2. Target Rock ...... 44 2.4.3. Lens ...... 45 2.4.3.1. Lithic Breccias ...... 45 2.4.3.2. Melt‐fragment Breccias ...... 46 2.4.3.3. Melt Lens ...... 46 2.4.4. Alteration ...... 48 2.4.5. Geochemical Identification ...... 48

3. METHODS ______50

3.1. Samples ...... 50 3.2. Major Elements ...... 51 3.3. Ni, Co and Cr ...... 51 3.4. Platinum Group Elements ...... 52 3.5. Isotopes ...... 53

4. RESULTS ______54

4.1. Element Abundances and Ratios ...... 54

PAGE VIII 4.1.1. Major Elements ...... 54 4.1.2. Ni, Co and Cr ...... 56 4.1.3. Platinum Group Elements ...... 58 4.2. Isotope Ratios ...... 61 4.2.1. Os Isotope System ...... 61 4.2.2. Pb Isotope System ...... 62 4.3. Multi‐signature Approach ...... 63

5. DISCUSSION ______65

5.1. Alteration ...... 65 5.2. Impact Identification ...... 66 5.3. Meteorite Identification ...... 68 5.4. Non‐magmatic Iron Meteorites ...... 70

6. CONCLUSION ______72

6.1. The Brent Impact Structure ...... 72 6.2. Implications for Identification ...... 73

V. References ...... X VI. Appendices ...... XVII

PAGE IX 1. INTRODUCTION

Besides the , dwarf planets and , the solar system comprises a vast amount of smaller bodies with varied sizes, compositions and provenance regions. They are defined as small solar system bodies [IAU, 2006]. Throughout time, at timescales far beyond ours, many of these obtained dynamical paths, which led them to the solid surface of other solar system bodies. When not hampered by any atmosphere or hydrosphere, or massive enough, they striked the target rocks with such velocity that shock waves originated from their kinetic energies. These caused a huge amount of rocks to vaporise, melt, eject, displace, brecciate and/or fracture, which resulted into the formation of a crater, many times larger than the small solar system bodies themselves and built up out of a variety of distinct impact rocks. In fact, this process of impact cratering occurred on every single solar system body. On most of them, the solid surface is even almost completely formed by impact craters. These show varied diameters, reflecting the varied sizes of the impacted small solar system bodies. Only few solar system bodies have the appropriate hydrosphere or atmosphere, and are active enough, to have surface processes which can defy this shaping force of impact cratering.

One such active body is the Earth. Because it has a dense enough atmosphere, a certain mass range of small solar system bodies was decelerated and landed on the Earth, without . These are called meteorites [IAU, 1961]. The more massive ones however, named impact meteorites, did reach the Earth’s solid surface at high velocity, with impact events as result. But because active, what was once an , is mostly processed beyond recognition as such. However, although not always visible to the naked eye, a minor part does survive in some way. Few of them show still some kind of resemblance to the craters they descent from. Off course, the larger and younger, the higher the preservation. These varied remnants of impact craters are called impact structures. At present, 178 are known, which come in a wide range of ages and sizes: from the 1947 Sikhote Alin up to the 2.4 Ga Suavjärvi impact structure, and from the 15 m Haviland up to the 300 km Vredefort impact structure (see Appendix A) [PASSC, 2010]. They can be studied in many ways and to many ends. Geochemistry is one such way and it forms the topic of this master dissertation.

PAGE 1 1.1. SCIENTIFIC SETTING

Although impact cratering is now considered to be an important process in the solar system – perhaps even the most important at present – and many impact structures are known, this knowledge is rather recent. Before, impact cratering, craters and structures, were often ignored and even ridiculed. Geochemical studies even needed the acquiring of faraway impact rocks from the first, before analyses were conducted on impact structures.

1.1.1. THE IMPACT THEORY

The first appearance of impact structures in literature was by Robert Hooke in 1665. He suggested an impact origin for the Moon’s surface features, which for the first time were described as depressions by Galileo Galilei in 1609. However, his revolutionary impact hypothesis was easily discarded due to (1) the lack of possible projectiles, the interplanetary universe was considered empty back then; and (2) multiple explanations by more conventional, volcanism‐related, processes [Koeberl, 2001].

At the beginning of the 19th century, not long after the first suggestions by E. Chladni in 1794, it became known that small solar system bodies exist and that some of them become meteorites [Howard, 1802 and Vauquelin, 1803]. With this new knowledge, the impact hypothesis was developed further in the 19th and first quarter of the 20th century [von Paula Gruithuisen, 1828; Proctor, 1873; Gilbert, 1893; Öpik, 1916; , 1919; Wegener, 1921; and Gifford, 1924]. But still, it was supressed by the more conventional volcanic theory [Koeberl, 2001 and references therein]. It was only from the 1920s from studies of the Barringer impact structure by D. M. Barringer [PASSC, 2010 and references therein] and theoretical studies, that the hypothesis gained more acceptance [Grieve, 1991; Koeberl, 2001; French & Koeberl, 2010; and references in all]. Combining these different studies by Baldwin [1949], led to a first complete meteorite impact hypothesis that would challenge the volcanic theory increasingly.

In 1965, a special meeting was organised by the New York Academy of Sciences for a debate between the volcanic and impact camps. It illustrates the general acceptance of an impact theory, which occurred in the 1960s and 1970s [Bucher, 1963; Dietz, 1963; French, 1968, 1990; Nicolaysen & Reimold, 1990; Sharpton & Grieve, 1990;

PAGE 2 French & Koeberl, 2010; and references in all]. This by (1) the exploration of the solar system by robotic and manned spacecraft, which revealed impact cratering as an important process in the solar system [Taylor, 1982, 1992; and French, 1998]; and (2) the ability to unambiguously identify impact structures and thus to distinguish them from more conventional geological structures. This impact identification is based on geochemistry, and the presence of permanent petrological and mineral changes, unique to impact cratering, called shock metamorphism [French & Koeberl, 2010 and references therein]. It was only in 1977 that E. M. Shoemaker declared impact cratering as a fundamental process of the solar system, and became the famous pioneer of planetary geology and impact cratering studies, he is now known as [Reimold & Koeberl, 2008]. The suggestion that an impact event caused the mass extinction at the ‐Palaeogene boundary, by Alvarez et al. [1980], made awareness of the importance of impact events complete, and even resulted into the attention of many others than earth and planetary scientists. At present, the study of impact cratering, craters and structures has developed an extended literature that is still increasing at high rate.

1.1.2. IMPACT GEOCHEMISTRY

Geochemical studies on impact structures only started when lunar impact rocks were brought to the Earth by the USA Apollo missions between 1969 and 1972 [Ganapathy et al., 1970a, 1970b, 1972, 1973, 1974; Morgan et al., 1972, 1974a, 1974b; Anders et al., 1973; and Higuchi & Morgan, 1975]. These studies all concluded that during the formation of a lunar impact crater, a small amount of the small solar system body is incorporated into the formed impact rocks. Morgan et al. [1975] confirmed this for the melt‐bearing impact rocks from impact structures on the Earth. Here, the amount of the meteorite contribution is usually less than 1 wt. % [Koeberl, 1998], although sometimes much higher values are found (up to 8 wt.% for the Clearwater East impact structure [McDonald, 2002]). The presence of the meteorite contribution was deduced from multiple geochemical signatures in the impact rocks that were different from the target rock. This difference occurs because meteorites show a distinct difference in chemical composition from the crust, a result of their different differentiation history. These geochemical signatures can be used for (1) meteorite identification: the chemical characterisation of the meteorite, because meteorites show also mutual differences in composition, including their chemical, and are classified this way into multiple levels; and (2) for impact identification: when a meteorite contribution is measured, this identifies or confirms off course, the impact origin of the impact structure.

PAGE 3 The first meteorite identification studies could only identify few impacts and few meteorites, mostly only to a first level [Koeberl, 1998]. But over the last ten years, some improvements have been made, especially regarding to analytical techniques and isotope systems [Koeberl, 2007 and Goderis et al., 2009]. At present, meteorite identification studies are conducted for 51 impact structures, of which only ten delivered a precise classification [PASSC, 2010]. Recently, a new approach of combining the multiple geochemical signatures is used, the so called multi‐signature approach [Koeberl et al., 2007 and Tagle et al., 2007]. This to classify more meteorites to a more precise level.

1.2. STUDY OBJECTIVES

In this master dissertation we conduct a meteorite identification, as precisely as possible, on obtained melt‐bearing impact rocks from the Brent impact structure. In addition, this is used as a case study to review the different geochemical signatures for meteorite identification, and discuss their applicability and place in a multi‐ signature approach and for impact identification.

1.3. IMPORTANCE

The study of impact cratering, craters and structures, including impact and meteorite identification, has its importance, and not only from a scientific point of view. Also social and economic aspects are involved. Especially the economic importance of impact structures is increasing significantly in the capitalised society we live in. In turn, economic prospecting and exploration has it benefits for the scientific part, for example by providing the involved scientist with samples from drill holes or even by discovering impact structures.

1.3.1. ECONOMIC

The economic importance of impact structures, makes impact identification even more significantly. However, this economic part is still addressed only shortly or sometimes even overlooked, when reviewing impact structures. Probably because some impact structures were already exploited since before they were identified as

PAGE 4 such. But the values are substantial. Grieve [2005] estimates that impact structures have a value of more than USD 18 billion per year in North America alone. Most of this economic potential lies in natural resources.

1.3.1.1. NATURAL RESOURCES

Approximately 25 % of the impact structures contain natural resources, two‐third of these are exploited. This means that more than 15 % of all impact structures are of economic profit, which implicates a very high probability in mineral and hydrocarbon exploration standards [Grieve & Masaitis, 1994 and Hawke, 2004]. These numbers are also very important towards future space exploration, because impact craters are the most frequent geological features in the solar system. According to their time of formation, impact‐structure‐related natural resources are pro‐, syn‐ or epigenetic to the impact event [Masaitis, 1989].

Progenetic natural resources were already present before the impact event, but because of it, are displaced, exposed and/or have an increased preservation potential. Notable are the Vredefort impact structure, which is the world’s largest Au deposit and accounts for more than one‐third of the world’s historic production [Phillips & Law, 2000; and Grieve, 2005], and the Carswell impact structure, from which already 590 million kg of U is mined, worth USD 1.5 billion [Grieve, 2005].

The syngenetic natural resources are directly formed by the impact event. Out of the melt that is formed during it, magmatic mineral deposits can form. However, enrichment is only sufficient for economic mining in very large impact structures. That’s why only one such impact structure is known: Sudbury. On the other hand, it is the largest mineral deposit known. Its production and reserves account for approximately 1.655 trillion kg of Ni, Cu, Co, Au, Ag, Se, Te, Ru, Rh, Pd, Os, Ir and Pt [Ames & Farrow, 2007]. Also mineral phase transitions occur during an impact event. This way, when carbon‐rich rocks are present, can form out of or by crystallization of . Such impact diamonds are present in many impact structures and even in large quantities. However, their concentrations are low, which makes exploitation currently not profitable (e.g. the Zapadnaya impact structure contains approximately 17,5 billion kg, worth USD 90 million, but only at 1 ppm [Hawke, 2004]). A more possible place for impact mining are placer deposits. One such place is prospected on the Ebeliakh river, approximately 200 km east of the Popigai impact structure [Shelkov et al., 1998 and Hawke 2004].

Epigenetic natural resources originate after an impact event, but because of conditions that were created during the impact event. Impact structures remain

PAGE 5 heated at deeper levels for thousands of years or even millions if very large [Ames et al., 2000 and Osinski et al., 2001]. This heat can produce hydrothermal systems that form substantial mineral deposits, such as the carbonate and sulphide deposits of the Haughton impact structure [Osinski et al., 2001]. Also a minor part of the Sudbury deposits are of hydrothermal origin [Grieve & Masaitis, 1994]. Even more important are hydrocarbons. Due to the extended fractionation and brecciation, impact structures form excellent reservoirs. Also traps and seals can be created by the structural displacement. For example, in the Gulf of , the major part of the 4.8 billion m³ oil and 425 billion m³ gas of the Cantarell field is related to the Chicxulub impact structure [Grieve, 2005]. In addition, when the crater is preserved, the impact structure itself can form a basin, in which hydrocarbon‐potential layers can be deposited, trapped and sealed. Many impact structure also act as fresh water reservoirs this way. Some even deliver hydroelectricity. For example, the hydro power station at the Manicouagan impact structure produces 4500 GWh per year. This is enough to supply power to a small city and worth approximately USD 400 million per year [Hawke, 2004].

Besides these, impact structures are also exploited for their rocks. They can be as well pro‐, syn as epigenetic. These include building stones, evaporites, diatomites and oil‐shales [Westbroek & Stewart, 1996 and French, 1998].

1.3.1.2. TOURISM

In addition to natural resources, impact structures can also be of economic interest by the relative new form of tourism. Each year, they are visited by tens of thousands of people, all imposed by their scientific newness, dimensions, devastating powers and sometimes beautiful nature. Local economies try to benefit from this and develop tourist centres with information and hiking trails, to attract even more people (e.g. Ries, Tswaing, Sudbury and Barringer impact structures) [Hawke, 2004].

1.3.2. SCIENTIFIC

Impact structures contribute to the Earth’s crust, and although they only constitute a minor part of it, knowledge of them is necessary to complete the understanding of the whole Earth someday, which is what many scientists so desperately seek. In addition, the formation of large impact structures includes a significant rebound uplift, which brings deep seated rocks to the surface. This makes impact structures the ideal place to study these otherwise hard to find rocks. But impact structures have also some scientific importance outside the Earth and even towards life.

PAGE 6 1.3.2.1. THE SOLAR SYSTEM

As already stated above, impact cratering is a fundamental process in the solar system, it plays an important role in the formation and development of most solar system bodies. The surface of our own moon for example, consists for over 80 % out of impact craters, the largest 2500 km in size [Bennet et al., 2008 and Reimold & Koeberl, 2008]. But except for the relatively close Moon, these exotic worlds are still hard to reach or even sample. Therefore, the knowledge we have from our impact structures, is extrapolated to them. Some details will be different, because different gravities, atmospheres and target rocks are present, but the general principles will be similar. In fact, knowledge of impact cratering is probably essential to the success of future space exploration, knowing its economic importance.

Meteorites are studied for many reasons. For example, they provide information on how the solar system is formed, from what it is build, its dynamics and fluxes, but also on its evolution and present state. While meteorites sample a certain mass range of small solar system bodies, impact meteorites are proficient because they sample a more massive range. The more precisely a meteorite is identified, the better its flux rate, provenance region and possible parent body can be determined.

1.3.2.2. BIOLOGICAL EFFECTS

Impact cratering has also some biological effects. In fact, impacts of small solar system bodies are probably even responsible for the origin of life in the first place. A small solar system body that impacts upon another solar system body can spall fragments of this body, of which the most accelerated in turn, can become small solar system bodies. Some scientists suggest that life originated elsewhere and that it was transported to the Earth by such fragments, probably from Mars, or even from out of the solar system. Others say life originated on small solar system bodies. Evidence exists of complex organic molecules that reside on small solar system bodies from the outer parts of the solar system, including amino acids, which are essential for life as we know it. Life could have originated here and then brought to the Earth. A more supported hypothesis is that these small solar system bodies supplied the raw materials, or at least a part of them, that were needed for life. But, the origin of life happened on the Earth. These solar system bodies probably brought also a large portion of the water present on the Earth [Horneck & Rettberg, 2007 and Bennet et al., 2008].

On the other hand, impact events are also destroyers of life. This was first suggested with evidence by Alvarez et al. [1980], who found a meteorite contribution in the

PAGE 7 global clay layer that forms the Cretaceous‐Palaeogene boundary. This implicated the clay to be impact ejecta and thus an impact origin for the ~ 65.5 Ma mass extinction. The meteorite contribution was confirmed in later studies at numerous locations around the world [Schulte et al., 2010 and references therein]. More than ten years later, the impact structure was identified: the 170 km Chicxulub impact structure in Mexico [Hildebrand et al., 1991]. The local and regional effects of the impact event include the air blast and heat from the impact event, tsunamis, and earthquakes. Global effects include forest fires ignited by spalled fragments, the injection of huge amounts of dust into the upper atmosphere, which may have inhibited photosynthesis for as much as two months, and the production of huge quantities of N2O from atmospheric heating, which is toxic, a strong greenhouse gas, and harmful for the ozone layer. However, the most important effect was due to the composition of the target rocks: approximately 3 km of carbonates and evaporites.

This way, the impact event resulted into the release of vast amounts of CO2, also a greenhouse gas, and sulphur species, which are toxic and block sunlight. Al this resulted into immediate devastation and short‐term global cooling. Estimates suggest as much as a 15°C decrease in average global temperatures. This was then followed by a long‐term warming from the release of the greenhouse gasses. Conditions that were too hard to endure for many life forms [Osinski, 2008].

1.3.3. SOCIAL

Understanding the process of impact cratering, should not only be of interest to earth and planetary scientists, but to whole society. This because it is still active. Imagine what would happen to our society if an impact event should take place. Smaller impact meteorites would only cause local effects. But the global effects by larger ones, as described in the previous section, could be catastrophic. Although such large impact events happen at timescales far beyond ours, there is an approximate 1 in 10.000 chance that a small solar system body of 2 km will impact the Earth in the next 100 y, which is enough for devastating effects at regional to global scale [French, 1998].

The scientific findings on a possible impact event, its hazards and consequences, receives more and more public attention these days. This is helped by the more spectacular milestones like the “” extinction by Alvarez et al. [1980] or the 1908 explosion near the Podkamennaya Tunguska river in Krasnoyarsk Krai, . The latter devastated some 2000 km² of forests and the resulting air blast was measured around the world. It is now known that it was the break‐up of an

PAGE 8 approximately 40 m meteorite. If this had happened a few hours later, the whole city of Saint Petersburg would be destroyed [Cohen, 2008 and Steel, 2008]. Even more appealing was the 1994 recorded crash of the fragmented small solar system body Shoemaker‐Levy 9 onto Jupiter. Its resulting marks, some initially as large as the Earth, were visible for several months [Hockey, 1994]. The footage was spread whole over the world. However, too often this public attention on impact events is only exploited for sensational purposes.

Those known small solar system bodies that have the potential for impacting the Earth, have a special status and are referred to as near‐Earth objects (NEO’s). They receive a value of zero to ten on an impact hazard and public warning scale: the Torino Scale. Values eight to ten represent certain impacts with respectively local, regional and global effects. The highest value ever reached was four: an impact probability of 2.4 % with regional effects. Currently, only two NEO’s have a value of one, meaning an impact probability above zero, but still extremely unlikely [Morrison et al., 2004].

The identification of more impact structures could adjust our estimations of the impact rate and extent of impact events. This could improve our knowledge on the impact chances we are subjected to. Meteorite identification can also indicate which solar system bodies to look for.

PAGE 9 2. LITEERATURE

2.1. IMPACT STRUCTURES

As already stated above, impact structures are the remnants of the impact craters once formed on the Earth. A list of all known ones, with location (see also Figure 1), size and age, is given in Apppendix A. The craters comprised impacct rocks that were vaporised, melted, ejected, displaced,, brecciated and fractured. Theyy were formed by shock waves that originated from thhe kinetic energies of small sollar system bodies that reached the Earth’s solid surface relatively unhindered: impact t meteorites.

Figure 1. Location of the 178 impact structures [PASSC, 2010]. A non‐random distribution occurs due to surface processes, the work area of researcheers, and the difficult accessibility of many regions, including the poorlly surveyed marine environment.

PAGE 10 2.1.1. IMPACT METEORITES

Small solar system bodies are defined as those bodies that orbit the Sun, without sufficient mass to obtain a nearly round shape [IAU, 2006]. They range in size from microscopic dust particles [Scott & Krot, 2005] over tens of kilometres to several 100 km with extremes up to approximately 550 km [Tedesco & Desert, 2002]. Many of them, especially the smaller part, are fragments of others, larger ones, originated from spallation by impacts or fragmentation by collisions.

2.1.1.1. ASTEROIDS AND COMETS

Chambers [2007] and Bennet et al. [2008]

A division into asteroids and comets can be made among small solar system bodies. Although the International Astronomical Union uses observational criteria to define their differences, the limitations with this [Brownlee, 2007] prefer us to use a more theoretically distinction based on their location in the solar system and derivative difference in composition.

Following the nebular theory, both asteroids and comets are leftovers from the first rocks that formed in the solar system. These are called planetesimals and accreted by electromagnetic forces from the solids that condensed out of the cooling solar nebula. The difference in location and composition between asteroids and comets, arises from their different place of origin, approximately 4.565 Ga [McKeegan & Davis, 2005]. Due to heating by the gravitational contraction of the solar nebula and the young Sun, the solar system was much hotter back then than today. This caused hydrogen compounds only to condense into ices beyond a certain distance from the Sun, called the snowline, which was situated between the present orbits of Mars and Jupiter. That’s why, theoretically, asteroids are said to be planetesimals formed inside the snowline. They are made out of rock and metal mostly, with almost no ices. Comets in contrast, are planetesimals formed outside the snowline and consist mainly of ices of multiple hydrogen compounds, mixed with minor metal, rock, and complex organic compounds. In addition, asteroids are known to show mutual differences in composition, also related to their distance of formation.

When planetesimals reached sizes exceeding 1 km, gravitational forces took over, which produced even larger planetesimals. This led to the formation of the planets, dwarf planets and moons. During and afterwards, the leftover planetesimals, thus asteroids and comets, experienced gravitational tugs from the planets. With each encounter, these gravitational perturbations increasingly changed their orbits to

PAGE 11 more eccentric and inclined ones. This resulted into being swept up by the planets and the Sun, or swung out of the solaar system. Over time, this process removed most asteroids and comets from the planetary region of the solar system,, leaving now only distinct residence regions in which they are captured.

2.1.1.2. RESIDENCE REGIONS

Most of the inner solar system asterooids can be found in the asterroid belt, between 2.1 and 3.3 AU [McSween, 1999]. Originally, it contained enough material to form a . However, gravitational perturbations by orbital resonances with Jupiter prevented this, leaving now approxiimately one million asteroids larger than 1 km [Chambers, 2007]. The number of smaller ones is assumed to be immense. Orbital resonances also provide the belt with some empty slots, which are called Kirkwood gaps [McSween, 1999]. Asteroids in these regions would have orbits with periods corresponding to a simple fraction of a planet’s period. Thiis results in more frequent gravitational perturbations, which would change the orbits. Besides orbital, also precession resonances exist, which have the same result, but to a lesser extent [Morbidelli et al., 2002].

Figure 2. The Kirkwood gaps of the asteroid belt [Asphaug, 2009]. This histogram cleearly shows the lower number oof asteroids around the 3:1, 5:2, 7:3 and 2:1 orbital resonances with Jupiter. The 2.1 AU side of the asteroid belt is largely bounded by the v6 preceession resonance with Saturn (not shown).

PAGE 12 Other minor asteroid residence regions, all bound by Kirkwood gaps, are that of the Hungarias at 1.8 to 1.9 AU, the Cybeles at 3.3 to 3.5 AU, the Hildas at 3.9 to 4.0 AU and the Trojans [McSween, 1999]. These Trojan asteroids are found on the orbit of Jupiter and consist of two sets, one 60° before Jupiter and one 60° behind. Recently, between 2001 and 2008, also seven Neptune trojans have been discovered [MPC, 2010] and many more are expected to exist. However, they could all resemble comets more than asteroids or be a mix of both [Sheppard & Trujillo, 2006].

Most of the outer solar system comets reside in a vast spherical region at an immense distance of about 50000 AU, called the Oort cloud, which is assumed to consist of the comets that were swung out of the planetary region of the solar system. Perhaps it even contains some asteroids this way [Weissman, 1999]. It is assumed to contain more than 1012 comets larger than 1 km with a total mass of one to fifty Earth masses [Stern & Weissman, 2001]. A smaller number of comets can be found in an ecliptic region beyond the orbit of Neptune, called the Kuiper belt, which is estimated to contain about 105 comets larger than 50 km [Brownlee, 2007], the equivalent of approximately one‐tenth Earth mass [Delsanti & Jewitt, 2007]. Also in the Oort cloud and Kuiper belt, the number of smaller bodies is estimated to be immense.

2.1.1.3. VOYAGE TO THE EARTH

To become meteorites, small solar system bodies have to leave their residence regions and obtain an impact course with the Earth. This requires their orbits to be altered into more eccentric ones that cross the Earth’s orbit. As seen above, this happened frequently in the early days of the solar system by gravitational perturbations from planets, but today, this process has mostly spent its force. It still only accounts for the comets from the Kuiper belt and a minor part of the asteroids [McSween, 1999 and Weissman, 1999].

The comets from the Oort cloud are also being pushed away by gravitational perturbations. However, these are not provided by planets, as the Oort cloud comets are at such great distance. Instead they are being pushed away by passing stars and molecular clouds during the solar system’s orbit in the Milky Way galaxy [Weissman, 1999 and Morbidelli, 2008].

The major part of the asteroids that leave there residence region do so as fragments. Typical relative velocities for asteroids are in the order of 5 km/s, which upon impacts and collisions, can result into fragments with velocities up to approximately 100 m/s [McSween, 1999]. Such velocities will spread the fragments out along the

PAGE 13 orbit of their parent bodies, but are most of the time not enough for ejection into more eccentric orbits and leaving the residence regions this way [Wetherill, 1976]. However, some fragments can reach nearby Kirkwood gaps, by which they are swung away, possibly on an impact course to the Earth. In particular the Kirkwood gap of the 3:1 orbital resonance with Jupiter and the v6 precession resonance with Saturn in the inner asteroid belt are assumed to be responsible for a significant contribution to the asteroid flux reaching the Earth [Morbidelli et al., 2002].

In fact, such resonances are probably responsible for almost all meteorites. This because the major part of the small solar system bodies that reached the Earth are asteroids, which is reflected in the NEO population. The contribution of comets is assumed to be rather low, but the estimations vary. Stokes et al. [2003] estimate this contribution to be only 1 %, while Bottke & Morbidelli [2006] estimate that of the Kuiper belt comets to be less than 10 % and that of the Oort cloud comets to be 5 % at the most. However, Levison et al. [2001] estimate the latter as 1 %, while Weissman et al. [2002] suggest 10 to 30 %.

2.1.1.4. THROUGH THE EARTH’S ATMOSPHERE

When small solar system bodies reach the Earth, they have still one passage to make before becoming meteorites: the Earth’s atmosphere. Although it is not in agreement with the International Astronomical Union’s definition, which is based on size [IAU, 1961], we here use the more workable definition by which any small solar system body is called a as soon as it enters the Earth’s atmosphere.

The maximum velocity of small solar system bodies is approximately 42 km/s. A higher velocity makes a closed orbit around the sun impossible and thus results into leaving the solar system. The Earth’s orbital velocity is approximately 30 km/s, so the entry velocity of a meteoroid is maximum 72 km/s. However, this value assumes a retrograde motion, which is only observed for few comets [McSween, 1999]. These have relative velocities up to 60 km/s [French, 1998]. The minimum velocity is 11.2 km/s, the Earth’s escape velocity. Typical entry velocities are 15 to 25 km/s for asteroids [Chyba et al., 1994] and 30 km/s for Comets [Reimold & Koeberl, 2008].

At altitudes below 100 km, the air density becomes high enough to create significant friction, which causes the meteoroid to decelerate and the surrounding air to ionize. The latter produces light and is called a meteor. The friction also melts the exterior of the meteoroid and this forms a thin layer of glass, called a fusion crust. Less coherent will explode and break up this way into smaller meteoroids, which is often accompanied with an air blast [McSween, 1999].

PAGE 14 Less massive meteoroids burn up by this friction, leaving only some microscopic dust particles to settle. However, although less massive, they account for the major part of small solar system bodies that reach the Earth. It is estimated that the Earth receives between 80 and 215 Mg each day this way [Kortenkamp & Dermott, 1998 and McSween, 1999]. More massive meteoroids will only be decelerated to free fall velocities of a few 100 m/s and can reach the Earth’s surface intact [McSween, 1999]. Upon impact, these meteorites will form mechanical penetration craters, only slightly larger than themselves [French, 1998].

Only meteoroids exceeding approximately 10 Mg [McSween, 1999], depending on velocity, composition and shape, are able to retain a part of their entry velocity. Those that impact with velocities exceeding approximately 3 km/s, depending on the target rock and meteorite lithology, will not only produce a penetration crater at impact, but due to the extreme velocities, also cause shock waves to form. These will excavate a much larger impact crater, around 20 to 30 times the meteorite’s diameter, depending on the impact magnitude (i.e. the meteorite’s mass and velocity, and the impact angle) and target rock [Bjork, 1963; Gault et al., 1968; and French, 1998].

2.1.2. IMPACT CRATERS

Melosh [1989] and French [1998]

In this section, the formation of impact craters and structures is discussed. We will limit us to continental impacts, because only a few marine impact structures are known [PASSC, 2010]. A marine example and the differences with continental impact craters is given by Lindstrom et al. [1996, 2005].

Impact craters form at random intervals, which depend on the solar system dynamics that deliver impact meteorites. Because impact cratering is active over large time‐scales, an average cratering rate can be established. Assuming an approximate constant impact rate for the last 3 Gy [Grieve & Shoemaker, 1994 and Bottke & Morbidelli, 2006], Bland [2005] and Bland & Artemieva [2006] estimate that a 1 km impact crater is formed every 16000 y, a 10 km crater every 200 ky and a 250 km crater every 143 My. However, there are many uncertainties concerning these rates, see Grieve [1984], Morrison et al. [1994], French [1998], Hughes [2000] and Asher et al. [2005] for other calculations.

The formation process of impact craters is still not fully understood. This because it is very complex and can’t be reproduced in total. Also no impact has been recorded

PAGE 15 during human history. In addition, its characteristics are dependent on the impact magnitude and the target rock. Whaat’s known is found by combining theoretical, experimental, and geological studies. Although it is continuous, the formation process can, by dominating forces and mechanisms, be divided into three successive stages: the compression, excavation and modification stage [Gault ett al., 1968].

2.1.2.1. COMPRESSION

When an impact meteorite hits the Earth’s solid surface, it will penetrate the target rock approximately 1 to 2 times its diameter. When it is stopped, ttwo sets of shock waves will have formed from the meteorite’s kinetic energy, which is between 1014 and 1024 J, depending on the impact magnitude [O’Keefe & Ahrens, 1993] (for comparison, 1014 J is the energy of one atomic bomb). These shock waves are nearly discontinuous stress waves with extreme pressures and supersonic velocities. One set propagates into the meteorite, compressing, fracturing, brecciating, and decelerating it. The other set expands hemispherically into the target rock, also compressing, fracturing, and brecciating it, together with a displacement by a down‐ and outwards acceleration. At the surface of the meteorite and targget rock, the shock waves are reflected as rarefaction shock waves. These unload the rocks from the extreme pressures of the initial shock waves to ambient pressures. This process results in such heating (i.e. exceeding 2500 °C) that the complete meteorite vaporizes and melts. But also target rock is melted. About 40 to 60 % of the meteorite’s kinetic energy is transferred into heat, an amount sufficient to melt a hemispherical volume with a radius of three to four times the metteorite’s diameter, proportional to the impact magnitudde. While the vaporized part may escape as a vapour plume, the melted part of the meteorite and target rock forms a layer on the displaced target rocks and violently incorporates fragments of it.

Figure 3. Schematic cross section of a theoretical compression stage (modified from French [1998]).

PAGE 16 The compression stage ends when the rarefaction shock wave that originated at the back of the meteorite, reaches the front of the meteorite. Its duration is very brief. It is approximately given by dividing the diameter of the meteoriite in km by the product of its velocity in km/s and the sine of the impact angle [Melosh & Ivanov, 1999]. For a diameter of 1 km, impact velocity of 20 km/s and imppact angle of 45°, this is merely 0.07 s. So even for a 10 km meteorite, the compression stage would take less than 1 second.

2.1.2.2. EXCAVATION

While the initial shock waves exercised a down‐ and outwards acceleration, the rarefaction waves produce a more upp‐ and outwards movement. This results in the excavation of a parabolic‐shaped depression, the transient crater. TThis comprises the ejection of upper target rock and the still continuing displacement of lower target rock. Both the ejection and displacement zone are approximately of equal size and include impact melt. The higher the impact magnitude, the larger the transient crater will be. Hereby, the horizontal excavation will become greater than the vertical, because of greater resistance with depth. At the edge of the transient crater, an elevated rim is formed by an overturned flap of ejected target materials, which displays inverted stratigraphy [Grieve, 2005].

Figure 4. Schematic cross section of a theorretical transient crater (modified from French [1998] and Hawke [2004]). Shown are the zones of vaporised, melted, spalled, ejected and displaced rocks. The arrows inndicate the trajectories of excavation.. The top of the rectangle represents tthe surface before the impact event. The larger the transient crater, the more the horizontal size will exceed the vertical.

At the surface, target rock fragments are spalled at high velocity by the rarefaction shock waves. The amount of this spallation is proportional to the impact angle and impact magnitude. Some of these fragments can even acquire vellocities exceeding the Earth’s escape velocity and thereby leave the Earth. Also some meteorite fragments can remain intact by spallation from the meteorite’s back.

PAGE 17 When the initial and rarefaction shock waves reach the energy level below which no more rock can be excavated by ejection or displacement, the excavation stage ends. At this point, the transient crater reaches its maximum extent. Calculations indicate that the excavation takes 6 s for a 1 km transient crater and 90 s for a 200 km transient crater.

2.1.2.3. MODIFICATION

All initial and rarefaction shock waves will decay to seismic waves beyond the transient crater rim and will play no further part in the formation of the impact crater. Gravity and rock mechanics take over now. Modification starts with the deposition of the melt and ejecta. But also the transient crater changes, depending on its size and the target rock. This leads to a division into four transitional morphological types of impact craters: simple, complex, ring and multi‐ring.

The smallest transient craters develop simple impact craters. The modification only includes a minor collapse of the walls, making the impact crater up to 20 % larger than the transient. This results into a roughly parabolic shape with a depth to diameter ratio of approximately 1/3. The slumped rocks, together with the melt and ejecta, will fill the crater up to approximately 50 %. For simple impact craters, this crater‐fill is called the breccia lens. It is observed to comprise only a minor part of ejecta.

The modification of larger craters is more complex. In crystalline rocks they form above approximately 4 km and in sedimentary rocks above approximately 2 km [Dence, 1972]. It includes the central rebound uplift of basement rocks by approximately one‐tenth of the final crater’s diameter and a major collapse of the crater walls, resulting in terraces bounded by normal faults, a broad flat floor and an impact crater up to 50 to 70% larger than the transient [Therriault et al., 1997]. This results into a depth to diameter ratio of approximately 1/5 to 1/6 [Koeberl, 2008]. The slumped rocks, melt and ejecta form an annular deposit around the central rebound uplift. The ejecta contribution is larger than with the simple type. Impact craters that are above approximately 25 to 40 km will have a depression in their central peak, forming a central ring, because their uplift overshot gravitational stability [Hawke, 2004].

The largest craters form multiple rings. However, there is still a debate on its definition [Hawke, 2004 and references therein]. Here we follow the systematics of Melosh [1989], which state that, based on lunar impact structures, a multi‐ring impact crater should occur above approximately 100 km on the Earth. However, no

PAGE 18 multi‐ring types are found, hence thhe reason for the debate. Probably, the three impact structures that exceed 100 km, are too eroded (Sudbury and Vredefort) or buried (Chicxulub) to show multiple rings.

Figure 5. Schematic cross sections of the modiification into a (a) simple; (b) compleex; and (c) ring impact crater (modified from Melosh [1989]). The grey parts represent the crater‐fill of melt, and slumped and ejected rocks. The arrows show the movements of these and of the central rebound uplift.

The modification stage gradually passes into that of post‐impact development. However, it is often said to end when rocks stop falling. This takes less than 1 min for smaller craters and 15 min at the most, for the largest ones. Thiis means that the largest impact craters of a few 100 km, are formed in less than only 20 min.

2.1.2.4. POST‐IMPACT DEVELOPMENNT

Related to impact cratering, the Earth’s surface processes of erosion, sedimentation, volcanism and tectonics, to which an impact crater is subjected after its formation,

PAGE 19 are all called post‐impact development. These processes transfer tthe impact crater into the impact structure we observe today. Impact craters are mmost likely to be affected by erosion, which can remove a part or even the entirre impact crater. Especially the meteorite fragments that possibly survived the impact will be affected, giving them a low preservation potenntial. To label the amount of erosion an impact structures experienced, an erosion level classification system is been used [Grieve & Pilkington, 1996 and Hawke, 2004]. The levels for a simple impact crater are (1) ejecta layer largely preserved; (2) ejecta layer partly and rim largely preserved; (3) ejecta not and rim partly preserved; (4) rim not and crater‐fill totaally preserved; (5) crater‐fill partly preserved; (6) cratter‐fill barely preserved; and (7) crater‐fill not preserved (see Figure 6).

Figure 6. Schematic cross section of a theoretical simple impact crater with erosion levels (modified after Grieve & Pilkington [1996] and Hawke [2004]).

A sedimentary basin can be formed inside the crater, by which a sedimentary fill can cover the crater’s interior. When the impact occurred in a sedimeentary basin, even the complete impact crater can be covered by sediments (e.g. the Ries impact structure [von Engelhardt, 1990]). In fact, approximately one‐third of all impact structures is buried [Grieve & Masaitis, 1994]. Impact structures can also be tectonically deformed. Hereby, they can lose their circular shape or even be broken up (e.g. Beaverhead impact structure [Fiske et al., 1994]). Tectonicss also limit the age of marine impact structures, given that oceanic crust is generally younger than 200 Ma [Hawke, 2004].

All this makes is assumable that due to post‐impact development, a large fraction of the impact craters once formed on Earth, are totally destroyed or diifficult to identify, especially to the naked eye. This also implicates a underrepresentation of smaller and older impact structures. When uusing size and age distributions, this bias should be taken into account.

PAGE 20 2.1.3. SHOCK METAMORPHISM

French [1998] and French & Koeberl [2010]

The stresses produced by the shock waves are far higher than the strength of rocks. This results in permanent petrological and mineral changes in the rocks, all are called shock metamorphism [French, 1968]. Because no other processes on the Earth produce such stresses, some of these changes are unique to impact cratering and used for the identification of impact structures. As already stated above, they played an important role in the general acceptance of the impact theory.

Shock pressures can rise up to several 100 GPa at the impact point, depending on the impact magnitude. Because the shock waves lose their energy rapidly, the shock pressures will decrease exponentially from here. In the displacement zone they reach a maximum of approximately 30 GPa. Near the eventual crater rim, the shock waves will have lost more than 99 % of their energy [Gault and Heitowit, 1963] and pass into seismic waves with pressures below 2 GPa and velocities between 5 and 8 km/s, depending on the target rock. These pressures are only sufficient for fracturing and brecciation and not for permanent deformation.

Between 2 and 10 GPa shatter cones form. These are megascopic multiple sets of striated fractures that form partial to complete cones. How they exactly form, is not well understood. They can be up to several metres in size and are observed in all kinds of rock, but the finer the rock is grained, the better the shatter cones are formed. It is assumed that the cones point in the direction of which the shock waves came, independent from any existing bedding. Because the only megascopic ones, shatter cones are the best shock metamorphic effects for impact identification.

Also high pressure polymorphs form. Next to the formation of diamonds, also high pressure polymorphs of quartz occur. It transforms to at pressures of 12 to 15 GPa and to at 30 GPa [Stöffler & Langenhorst, 1994]. This in contradiction with the lithospheric conditions where coesite forms at lower pressures than stishovite. Especially stishovite is useful for impact structure identification, because it has never been found elsewhere. Recent studies also have identified other high pressure polymorphs like reidite, a polymorph of zircon, and a

TiO2 polymorph, possibly derived from rutile or anatase [Glass et al., 2002; Jackson et al., 2006; and Wittmann et al., 2006].

Shock waves also produce microscopic planar structures in many minerals. These are used as standard microscopic impact identification criteria, especially those in

PAGE 21 quartz, which are also best studied. This because quartz is abundant in a wide range of rocks, and is resistant to alteration and metamorphism. Most important are planar deformation structures, which are multiple sets of closed, extremely narrow (typically less than 3 μm), parallel (typically 2 to 10 μm spaced) planar regions of an amorphous phase. They form at shock pressures between 8 and 35 GPa and comprise multiple types, see French [1998] and French & Koeberl [2010] for more on these. Also parallel sets of planar fractures develop. This at shock pressures of only 5 to 8 GPa. The fractures are typically 5 to 10 μm wide and spaced 15 to 20 μm. Rarely, similar cleavage occurs also from non‐impact processes, therefore they are only an indication for a possible impact.

At higher shock pressures, minerals are partial or complete transferred into an amorphous phase, without changing the fabric. These are called diaplectic glasses. They are most common in quartz (35 to 50 GPa) and feldspar (30 to 45 GPa) [Grieve, 2005]. Shock pressures between 40 and 60 GPa produce temperatures above 1000 °C and will cause selective mineral melting. This differs with equilibrium melting, which happens gradually and starts at the mineral boundaries. With shock melting, the melting occurs instantaneously and also inside the mineral. Shock pressures above approximately 60 GPa are sufficient for complete rock melting and those above 100 GPa for vaporising.

2.1.4. IMPACT ROCKS

French [1998] and references therein

The formation of an impact crater produces a wide variety of rocks out of the meteorite and target rock. All are called impact rocks. As already stated above, they experienced processes of fracturing, brecciating, excavating, melting and vaporising. Based on their final location, they are divided into basement, crater‐fill, proximal‐ ejecta and distal‐ejecta rocks. For all breccias counts that being monomict or polymict depends on the variation of target rocks.

2.1.4.1. BASEMENT

The displacement zone consists of target rock, forced down‐and outwards. This happens in large blocks, tens to hundreds of metres in size, inside which the target rock stratigraphy and fabric can be preserved. These rocks are strongly thinned and will form the floor of the impact crater. They are called the parautochthonous basement. Shock pressures can exceed 30 GPa in the centre of the impact crater.

PAGE 22 Here, the high shock metamorphic microscopic planar structures can be found. The major part of the parautochthonous basement are less shocked, often only showing shatter cones, next to fracturing and brecciating.

At deeper levels, the parautochthonous basement will pass into target rock that was not displaced, but fractured and brecciated in situ. These are called the autochthonous basement. They show no shock metamorphic effects, which makes it hard to distinguish them from non‐impact breccias. This autochthonous basement in turn, will pass into rocks, which are not affected by the impact event. This happens at depths of a few 100 m for smallest impact craters and a few km for the largest ones. Such depths implicate that even for impact structures of erosion level seven, some impact rocks can be preserved.

During the modification stage of large transient craters into complex and multi‐ring impact craters, the rapid rebound movements of the basement into a central uplift, can again, cause fracturing and brecciating, which gives them a complex history. Also pseudotachylite can form this way.

Inside the basement, also dikes can be found. These are formed during the excavation stage by rocks from the top of the displacement zone or during the modification stage by crater‐fill rocks, that intruded into fractures of the parautochthonous basement. Examples are known of up to a kilometre long (e.g. the Staten Islands impact structures [Dressler & Sharpton, 1997]).

2.1.4.2. CRATER‐FILL

As already stated above, the crater is filled during the modification stage. In simple craters this forms a breccia lens and in complex craters a more annular deposit, around the central rebound uplift. This crater‐fill consists out of four components: (1) slumped rocks from the lowly shocked displacement zone; (2) highly shocked ejecta; (3) slumped ejecta; and (4) impact melt. The crater‐filling process is both rapid and chaotic, but mixing of the different components is not complete. The crater‐fill therefore contains varied, but distinctive, impact rocks.

Lithic breccias form the major part of the crater‐fill in all impact craters. They are free of melt and consist of rock and mineral fragments in a similar clastic, but finer‐ grained, matrix. The varied shaped clasts are poorly sorted and vary from millimetre to metre scale. Most of the material in the lithic breccias is derived from the lowly shocked regions around the walls and rim of the transient crater, distinctive shock metamorphic effects are therefore only rarely observed.

PAGE 23

Figure 7. Schematic cross sections of a theoretical simple impact crater (modified frrom Grieve [1987] and Hawke [2004]).

Figure 8. Schematic cross sections of a theoretical complex impact crater (modified from Grieve [1987] and Hawke [2004]).

A smaller portion of the crater‐fill, generally 20 to 25 vol.%, conttains melt that is formed during the impact event. This melt component ranges from a few to over 90 vol.%. These melt‐bearing breccias also contain the highly shockked target rocks. Where the melt is present as glass fragments, these are called melt‐fragment breccias. The fragments make out typically 5 to 15 vol.% and have vvaried sizes up to centimetre scale. In larger impact craaters, where a significant ejecctta contribution is present, some of the melt fragments have been ejected and show grooved and lobate flow structures that are evidence of aerodynamic sculpturing. The melt‐fragment breccias occur as lenses in the lithic breccias, and are particularly present in the

PAGE 24 upper parts of the crater‐fill, and sometimes also in the lower centre of simple impact craters.

Where the melt is present as matrix, the breccias are called melt‐matrix. The melt component makes out 25 to 75 vol.% here. It ranges from glass to completely crystalline igneous rock. The target rock fragments and minerals are frequently highly shocked with partial melting. Melt‐matrix breccias often grade into impact melts with little or no target rock clasts. These have the appearance of conventional igneous rocks. For the impact origin of impact structures, this was for a long time an evidence for the volcanic camp, until the discovery of an irrefutable meteorite contribution. Such, almost pure, impact melts surrounded by melt‐matrix breccias are called impact melt lenses. They are rare in simple impact craters but common at the surface of the crater‐fills in larger impact craters.

2.1.4.3. PROXIMAL‐EJECTA

Of all ejected rocks, approximately 50 vol.% is deposited inside a distance of two times the impact crater’s radius from the impact centre. Near the rim, these will form a continuous proximal that may be tens to hundreds of metres thick, depending on the impact magnitude. At greater distances, this blanket becomes thinner and discontinuous. Approximately 90 vol.% of all ejected rocks, is deposited less than a distance of 5 times the impact crater’s radius from the impact centre. Ejecta rocks consist of lithic and melt‐fragment breccias that are highly shocked.

2.1.4.4. DISTAL‐EJECTA

Approximately 10 vol.% of the ejected rocks reaches distances greater than 5 times the radius of the impact crater. These are called distal ejecta and are usually less than a few centimetre thick. The finest part can be transported by the atmosphere to regional or even global distances (e.g. the Chicxulub impact event). Distal‐ejecta is in average higher shocked than proximal‐ejecta. It comprises a peculiar and much‐ studied variety: . These are small (millimetre to centimetre scale) bodies of pure glass that are known to be ejected only from a few impact structures and are spread over areas, called strewnfields, that may be thousands of kilometres in size. Probably, they are formed only with the highest impact magnitudes.

While all impact angles (except the very low, probably below 15°) produce the same circular morphology that impact craters show, the impact angle has much more effect on the distribution of distal‐ejecta, which will show an asymmetry proportional to a lower impact angle [Hawke, 2004].

PAGE 25 2.2.

[McSween, 1999; Haack & McCoy, 2005; Mittlefehldt, 2005; Scott & Krot, 2005; Krot et al., 2007; and references in all]

Although the impact meteorites are destroyed upon impact and the surviving fragments have low preservation potential, a substantial number of the non‐impact meteorites is collected. Currently, approximately 4 x 104 meteorites are known [MS, 2010] and this number is rapidly growing, especially by massive finndings in deserts, where preservation is increased. When recovered, the meteorites observed to fall are called falls, those that cannot be linked to a fall are called finds. Their names have to be approved by the Nomenclature Committee of the Meteoriticall Society and are based on the location of recovery. Numbers are used when multiple meteorites are found on the same location. Meteorites of the same, but fragmented, meteoroid are said to be paired. Most meteorites do not exceed sizes of 0.5 m, the largest one is the Hoba.

Figure 9. The Hoba meteorite [MS, 2010]. Disccovered in 1920, the 60 Mg Namibian iron meteorite (82.4 wt.% Fe and 16.4 wt.% Ni) is estimated to have fallen approximately 80 ky ago. The reason such massive meteorite formed no impact crater is probably due to its odd flat shape, approximately 2.95 by 2.85 by 0.9 m. This must have produced an immense atmospheric friction and thus deceleraattion, by which only a penetration crater formed [Grunert, 2000 and Schneider, 2004].

PAGE 26 A meteorite is said to have multiple aages, each reflecting a moment in the meteorite’s history. Off course there is its formation age, the moment a meteorite’s parent body is formed. But also a gas‐retention age exists. It reflects the last thermal event of a meteorite and can be calculated by measuring the decay of 40K to 40Ar. This because the Ar is released by the thermal event and when the temperature drops below the blocking temperature of Ar, the production of 40Ar starts again. Allso the moment a meteorite is separated from its parent body can be defined. From then, several stable and radioactive isotopes are formed by cosmic radiation on the fragment’s surface, some of them are specific to this process. Their production stops when the fragment becomes a meteorite, because the Earth’s atmosphere absorbs cosmic radiation. By measuring their amounts, the moment of separation can be calculaated. The obtained age from these is called the cosmic‐ray exposure age. When this is aadded to the time the meteorite already spend on the Earth, its moment of separatioon is known. The terrestrial age, or the moment of arrival on the Earth, is also based on the radioactive isotope production by cosmic radiation, but only on the radioactive isotopes.

Figure 10. Classification of meteorites into types, clans and groups (modified froom McSween [1999]; Mittlefehldt et al. [1998]; and Goderis [2006]).

As already stated above, meteorites show mutual differences in composition, including their chemical. But they differ also in mineralogy and petrology. A classification can be made by these differences with a genetic rellevance. Chemical classification includes element abundances and oxygen isotope compositions. There are three types of meteorites: (1) stones are composed mostly of ssilicate and oxide minerals with minor metal; (2) irons consist almost entirely of metallic FeNi; and (3)

PAGE 27 stony‐irons have nearly equal proportions of silicates and metal. Stones are divided into chondrites and achondrites. Achondrites, irons and stony‐irons are also called non‐chhondrites and are differentiateed from chondrites. A further cclassification into groups is used. By convention, a group is defif ned as having at least five unpaired meteorites with comparable chemistry, mineralogy and petrology. Their genetic relevance is that of being fragments of a common parent body. Two to four related meteorites are called grouplets. Meteorites not related to any other meteorite are said too be ungrouped. Appendix B contains the chemical composittions of the most common meteorite groups. The remainder of this section containss a brief review of the meteorite classification. See McSween [1999], Haack & & McCoy [2005], Mittlefehldt [2005], Scott & Krot [2005], Krot et al. [2007], and references in all for a more complete one.

Figure 11. Relative meteorite type abundances of falls (modified from McSween [1999]). Falls are used instead of finds because abundances of the latter are biased due to a difference in preservation potential and finding chance.

As can be seen in Figure 11, the major part of the meteorite fallls consists out of chondrites. Although the relative abundances must have changed over time, this chondrite dominance is probably the case for most of geological time and reflects the dominant chondritic composition of small solar system bodies.

2.2.1. CHONDRITES

By definition, chondrite meteorites are made out of chondrules maainly. These round to irregular shaped inclusions are 0.001 to 10 mm in size and consist t mostly of olivine, pyroxeene and some metallic FeNi. Besides chondrules, chondritess contain diverse

PAGE 28 proportions of three other components. These are refractory inclusions, metallic FeNi and matrix material. Refractory inclusions are tens of micrometre to centimetres in size and can be divided into Ca‐Al‐rich inclusions (CAIs) and amoeboid olivine aggregates (AOAs). The metallic FeNi is present as grains up to a millimetre in size and is also found inside the chondrules. Both chondrules, refractory inclusions and metallic FeNi are formed by high temperature processes involving condensation out of the solar nebula, accretion, – partial – shock‐melting and rapid cooling in minutes to hours. They are the first and thus oldest materials of the solar system, formed between 4.567 and 4.564 Ga [McKeegan & Davis, 2005]. The matrix material is fine‐grained from 5 up to 10 µm and volatile‐rich. It consists of silicates, oxides, sulphides, metallic FeNi, organic material and the rare pre‐solar system grains [Zinner, 2005]. As seen above, the planetesimals formed out of these four components by electromagnetic accretion. This started 4.565 Ga [McKeegan & Davis, 2005] and lasting some few My. All other solids in the solar system originated out of them. Chondrite meteorites represent these first rocks to form in the solar system.

Chondrites are given a petrologic type from one to six, indicating their degree of metamorphism. This occurred approximately in the 60 My following their formation. Chondrites of type three are the least altered. Towards type six heating metamorphism occurred by the short‐lived isotopes 26Al and to a lesser extent 60Fe, but without experiencing melting. Chondrites towards type one suffered aqueous metamorphism with the formation of hydrated minerals. Many chondrites consist out of breccias, formed by collisions and impacts. Herein, almost no mixes of chondrite groups are found, but mixed petrologic types are often. This – and the occurrence of shock metamorphic effects – suggests the existence of parent bodies with a range in metamorphic grades, that were brecciated, fractured and excavated, by collisions and impacts. Cosmic‐ray exposure and gas‐retention ages show that these happen at random and still today. As seen above, they provide a way for the fragments to escape from their parent bodies and residence regions, possibly on a voyage to the Earth.

Chondrite meteorites include fifteen groups and fifteen ungrouped chondrites. However, two groups, the R (Rumuruti‐like) and K (Kakangari‐like), are actual grouplets. In addition to groups, chondrite groups are also placed in clans, to illustrate their related time and place of origin in the solar system. Both the R and K chondrites form a clan, the other thirteen groups belong to the carbonaceous (C), ordinary (O) or enstatite (E) clan. Following the temperature and density dependence of the oxidation state, which increases in the sequence of E, K, O, R and C

PAGE 29 chondrites, E and K chondrites are assumed to have formed closest to the Sun, the O and R chondrites at intermediate distances and the C chondrites the farthest away, probably near the snowline. This is also reflected in their chemiical compositions. Some characteristics of the common carbonaceous, ordinary and enstatite chondrites are listed below.

Figure 12. Bulk oxygen isotope composition of most chondrite groups (modified froom McSween [1999]). The horizontal axis indicates the δ18O and thee vertical axis the δ17O, both relative to the Vienna Standard Mean Ocean Water. The terrestrial sample line is the Earth’s mass fractionation line.

2.2.1.1. CARBONACEOUS

Their name is an historical mismatch, although all carbon bearing, only two groups, the CI (Ivuna‐like) and CM (Mighei‐like), are rich in carbon (i.e. 1.55 to 6 wt.%). The other groups are the CR (Renazzo‐like), CB (Bencubbin‐like), CV (Vigarano‐like), CK (Karoonda‐like), CO (Ornans‐like) and CH (ALH85085‐like). The CI chondrites contain the highest concentrations off volatile elements. They are assumed to be the most primitive chondrite group and best representation of the solar system’s relative chemical composition for most elements. This is why they are often used as a rock reference composition. However, due to intensive aqueous metamorphism (i.e. petrologic type one), they are the only chondrite group that lacks chondrules. All other carbonaceous chondrites havve petrologic types of two and three and are characterized by (1) refractory element abundances that equal or exceed those of the CI; (2) oxygen isotopic composition with δ17O equal or lower than mminus 2 ‰ (with

PAGE 30 some exceptions); and (3) refractory inclusion abundances equal or higher than 0.1 vol.%.

2.2.1.2. ORDINARY

The ordinary chondrite clan contains approximately 80 % of all the meteorite falls. It is divided into three groups: the H, L and LL. These letters refer to the bulk Fe contents, with the H chondrites having high total Fe contents, the L chondrites having low total Fe contents and the LL chondrites having low metallic Fe relative to total Fe as well as low total Fe contents. Also through the H, L and LL sequence, the oxidation state increases and the siderophile element (i.e. those showing affinities with Fe) abundances decrease. All three groups are characterized by (1) Mg‐ normalized refractory lithophile abundances of ± 0.85 times those of CI chondrites; (2) oxygen isotopic composition plotting above the terrestrial fractionation line; (3) high abundance of chondrules with non‐porphyritic and FeO‐rich chondrules being common and Al‐rich chondrules being rare; (4) a rarity of CAIs and AOAs; and (5) a large range in degree of metamorphism, petrologic types three to six, with often also evidence for minor aqueous metamorphism.

2.2.1.3. ENSTATITE

This clan comprises two groups with different contents of metallic Fe: EH and EL. Additionally, there is one ungrouped E chondrite: LEW87223. All are characterized by (1) a unique mineralogy indicating formation under extremely reducing conditions; (2) bulk oxygen isotopic compositions that plot along the terrestrial fractionation line, close to that of the Earth and the Moon; (3) abundant enstatite‐ rich chondrules; (4) a rarity of CAIs; (5) very low abundance of fine‐grained matrix and chondrule rims; and (6) a large range in degree of metamorphism, i.e. petrologic types three to six.

2.2.2. NON‐CHONDRITES

Planetesimals that became larger than 10 km started to differentiate. In the interior a core formed and at the exterior igneous rocks originated by partial melting and fractional crystallization. This caused isotopic homogenization and chemical compositions to deviate to various degrees from chondritic values. Non‐chondrite meteorites contain virtually none of the components found in chondrites. Partial melting occurred in variable degrees, which also led to a distinction between primitive and differentiated non‐chondrites. As already stated above, the non‐

PAGE 31 chondrites are divided into the stony achondrites (including the primitive non‐ chondrites), irons and stony‐irons.

2.2.2.1. ACHONDRITES

The primitive achondrites still have approximately chondritic bulk ccompositions, but show igneous or metamorphic textures. They are thought to be imppact rocks, highly metammorphosed chondrites or residues of very low degrees of parrtial melting. Most primitiive achondrites can be assigned to five groups: the acapulcoite, lodranite, winonaite, brachinite and ureilite group.

Figure 13. Bulk oxygen isotope composition of f most primitive achondrite groups (modified from McSween [1999]). The horizontal axis indicates the δ18O and the vertical axis the δ17O, both relative to the Vienna Standard Mean Ocean Water.

The acapulcoites and lodranites are equigranular rocks composed of orthopyroxene, olivine, Cr‐diopside, Na‐plagioclase, FeNi‐metal, schreibersite, troilite, whitlockite, Cl‐apatite, chromite and graphite. Although this mineral assemblagee is rather similar to that of ordinary chondrites, the mineral compositions and abundances, differ. The acapulcoites are fine‐grained (150–230 mm), whereas the lodraanites are coarse‐ grained (540–700 mm). Cosmic‐ray exposure ages are between 5.5 and 7 Ma for all acapulcoites and most of the lodranites, possibly indicating samplling from a single impact event on a common parent body. Winonaites have a chonndritic mineralogy and chemical composition, but recrystallized textures. They are fine‐ to medium‐ grained equigranular rocks. Their mineral compositions are interrmediate between those of E and H chondrites. FeNi‐FeS veins are common and constitute the first

PAGE 32 partial melts of a chondritic precursor. The brachinites are meedium‐ to coarse‐ grained (0.1–2.7 mm) equigranular dunitic wehrlites, consisting dominantly of olivine and minor augite, plagioclase, orthopyroxene, chromiite, Fe‐sulphides, phosphates and FeNi‐metal. Ureilites are C‐bearing ultramafic rocks composed out of olivine and pyroxene with minor dark interstitial material. They are distinguished from other achondrites by (1) high CaO contents in olivine and piigeonite; (2) high

Cr2O3 contents in olivine; (3) high (up to 5 wt.%) C contents; (4)) the presence of reduced rims on olivine (and sometiimes pyroxene) were in contact with graphite; and (5) a large range in oxygen issotopic compositions which plot along the C chondrite mixing line.

The differentiated achondrites comprise also five groups: the angrites, aubrites, howardites, eucrites and diogenites. In addition also the rare SNC meteorites from Mars (the lherzolitic and basaltic Shergottites, the clinopyroxeniitic and wehrlitic Nakhlites, and dunitic Chassignite) and the Moon (basalts, basaltic breccias and melt‐ fragment breccias) are classified here.

Figure 14. Bulk oxygen isotope composition of most differentiated achondrite groups (modified from McSween [1999]). The horizontal axis indicates the δ18O and the vertical axis the δ177O, both relative to the Vienna Standard Mean Ocean Water.

The angrites are medium‐ to coarse‐grained (up to 2‐3 mm) igneous rocks of generally basaltic composition and consist mainly of Ca‐Al‐Ti‐rich ppyroxene, Ca‐rich olivine, and anorthitic plagioclase. Aubrites are highly reduced breccias that consist mostly out of nearly FeO‐free enstatite, with a minor amount of albitic plagioclase, nearly FeO‐free diopside and forsterite. Similarities in minerallogy and oxygen

PAGE 33 isotopic compositions suggest that aubrites are related to E chondrites. The howardites, eucrites and diogenites are also classified as the HED group because these meteorites originated from the same parent body. Based on spectroscopy studies, this could be the 530 km 4 Vesta, located near the 3:1 Kirrkkwood gap in the asteroid belt. They formed by an immpact and represent different depths from 4 Vesta’s crust [Drake, 2001]. Diogenites sample the deepest parts, being orthopyyroxene cumulates with minoor plagioclase and olivine. The eucrites come from varied, more shallow, pyroxene‐plagioclase basalts. Howarddites are from the brecciated surface regolith, which iss formed by former impacts. They contain high amounts of noble gases implanted by cosmic radiation and consist mostly out of eucrite and diogenite fragments, wiith some minor impact melt and C chondrite xenoliths.

2.2.2.2. IRONS

Figure 15. Bulk oxygen isotope composition of some stony‐iron and iron groups (modified from McSween [1999]). The horizontal axis indicates the δ18O and the vertical axis the δ17O, both relative to the Vienna Standard Mean Ocean Water.

Irons are fragments of the cores formed in the differentiated planetesimals. To exposure such deeply seated cores, the planetesimals must have experienced severe collision and/or multiple large impacts. If any parent body survived and still exists, it must be a core from which, at least partly, the overlying silicate rocks are stripped. Iron meteorites are classified based on their oxygen isotopes and siderophile trace

PAGE 34 element compositions. 85% of them can be categorised into thirteen groups: IAB, IC, IIAB, IIC, IID, IIE, IIF, IIIAB, IIICD, IIIE, IIIF, IVA, and IVB. Numbers I to IV indicate decreasing contents of Ga and Ge, the two most volatile siderophile elements. Chemical evidence indicates that while some groups were formed by fractional crystallization, called the magmatic iron meteorites, others are not.

Probably, some iron groups are genetically related to some achondrite groups, some were possibly derived from a single parent body: (1) silicate inclusions in IAB irons (and probably also those in IIICD irons) are linked through oxygen isotopic and mineral compositions to the winonaites; (2) the bulk oxygen isotopic compositions of silicates in IIE irons are similar to those of H chondrites; and (3) IIF irons are probably formed in the same region of the solar system as some stony‐iron pallasites, some might have a common parent body, others definitely not.

2.2.2.3. STONY‐IRONS

The pallasites and mesosiderites make out the stony‐irons. Both are further divided into groups. Pallasites are composed of roughly equal amounts of silicate (dominantly Mg‐rich olivine), metal and troilite. They are regarded to be derived from the core‐mantle boundary, which presumes a close relationship to the iron meteorites. Mesosiderites are breccias with the same components and proportions as the pallasites, but the silicate component consists mainly out of pyroxene and plagioclase, that have a crustal signature. Probably, they formed by mixing during collisions.

2.2.3. COMETARY METEORITES

[Gounelle et al., 2008 and references therein]

At present, all meteorites are determined to have an asteroid origin, except for those few from the Moon and Mars. It is assumed that comets do not or only for a minor, not known, part, contribute. This because they are already a small fraction of the small solar system bodies that reach the Earth and in addition, have a rather small coherence due to their composition, by which most don’t survive passage through the Earth’s atmosphere. The most massive form impact craters and are destroyed, except for some possible spalled fragments with little preservation potential.

Recently Gounelle et al. [2008] argued that the best candidates for comets are to be found among the CI chondrites. They postulate that there is a continuum in

PAGE 35 composition between asteroids and comets rather than a sharp distinction. This because there is no reason for the snowline to have always occupied the same location, nor to define an abrupt transition between ice‐poor and ice‐rich bodies. CI chondrites might sample this continuum.

2.3. GEOCHEMICAL SIGNATURES

[Tagle & Hecht, 2006 and Koeberl, 2007]

The Earth is formed out of planetesimals with chondritic bulk compositions. But already early in its history, it started to differentiate into a metallic core and primitive silicate mantle by the gravitational separation of Fe. Because trace elements show affinities towards major elements, those which do so towards Fe, were also partitioned into the core [Lorand et al., 2008]. As already stated above, such elements are called siderophile. Relative to the Earth’s chondritic bulk composition, they are enriched in the core and depleted in the primitive mantle. Based on their abundances in the mantle and theoretical values predicted by core‐ mantle equilibrium partitioning models, the later impact‐addition of approximately 0.8 wt.% of chondritic rock to the primitive mantle is suggested [Halliday, 2004]. By partial melting, also the primitive mantle differentiated, by which a mantle and crust was formed. Again, trace elements show different behaviour in this process. Those that prefer to remain in the mantle are called compatible, those that preferentially enter the melt are called incompatible and will rise with the melt to form the crust. Those trace elements that are siderophile and compatible, will now be highly depleted in the crust.

Meteorites are planetesimal rocks with an earlier and/or different evolution than the Earth. Therefore, they do not show such depletion of compatible siderophile trace elements and/or have different isotope ratios of these (see Appendix B). Because depleted, even a small meteorite contribution will create measurable geochemical signatures of these elements this way, which off course, can be used for impact and meteorite identification. These geochemical signatures are mostly measured in the melt‐bearing impact rocks from the crater‐fill, because these contain the highest meteorite contribution [Palme et al., 1978, 1981; McDonald et al., 2001; McDonald, 2002; Tagle and Claeys, 2005]. But also the meteorite contribution in the ejecta is measurable (e.g. Alvarez et al. [1980]). However, for most impact structures, the ejecta rocks are eroded.

PAGE 36 2.3.1. ELEMENT ABUNDANCES

A division can be made between the compatible moderately and highly siderophile elements. While the metal/silicate partition coefficients of the latter are above approximately 10000, those of the moderately are one or two orders of magnitude lower [Lodders, 2003]. They are discussed separately, because they are analysed by different methods and form each an independent way for impact and meteorite identification.

2.3.1.1. MODERATELY SIDEROPHILE ELEMENTS

Ni, Co and Cr are moderately siderophile elements. Cr in addition, also has a lithophile character, this implicates an affinity towards O and because Si is also lithophile, towards silicate rocks. Average crustal values are approximately 51 ppm for Ni, 25 ppm for Co and 119 ppm for Cr [McDonough & Sun, 1995]. The upper continental crust has values of approximately 47 ppm Ni, 17.3 ppm Co and 92 ppm Cr [Rudnick & Gao, 2003]. When a significant enrichment of these elements, relative to the target rocks, is measured, this indicates an impact origin.

When the meteorite contribution exceeds 0.1 wt.%, also the element ratios will significantly be affected. If accounted for the indigenous composition of the target rock, it will be possible to distinguish between a chondrite and iron meteorite [Koeberl, 1998]. This because chondrites show Cr abundances of 2575 to 3810 ppm and Ni/Cr ratios that vary between roughly 2 to 7 [Tagle and Berlin, 2008]. Cr contents in iron meteorites are a few orders of magnitude lower and the Ni contents are higher [Buchwald, 1975]. This means irons display much higher Ni/Cr ratios, up to three orders of magnitude. The same is applicable for the Co/Cr ratios. The absence of a Ni and Co enrichment, accompanied by elevated Cr values, and low Ni/Cr and Co/Cr ratios, could be interpreted as the result of the impact of an achondrite meteorite [Palme, 1980].

A more precise meteorite identification is possible by comparing the Ni/Cr and Co/Cr ratios with meteorite databases (see Appendix B).

Because only moderately siderophile, some crustal target rocks can have relatively high indigenous concentrations of Ni, Co and Cr, which off course is not preferable for impact and meteorite identification. An interfering (ultra)mafic component may also not be present or should be extracted. This might be difficult when multiple types of target rock and high depths are involved and especially when fractionation processes are involved.

PAGE 37 2.3.1.2. HIGH SIDEROPHILE ELEMENTS

The platinum group elements (PGE) Ru, Rh, Pd, Os, Ir and Pt are high siderophile elements and share similar physical and chemical properties. Because of comparable properties, Re and Au are often associated with them. Most meteorites have PGE contents approximately 10000 orders of magnitude higher than the crust (0.022 ppb Ir for the upper continental crust) and 100 higher than the mantle (3.5 ppb Ir for the upper mantle) [Peucker‐Ehrenbrink & Jahn, 2001]. When a significant enrichment of PGE relative to the target rock is present, this implicates an impact origin.

Ir is the best known element when it comes to meteorite identification. This because of the first meteorite contribution studies, which relied on analysis by neutron activation methods. Ir was relative to the other PGE the most easy and sensitive element to analyse. Without dissolution of the samples, it was possible to measure Ir values as low as only 0.1 ppb this way. By inductively coupled plasma mass spectrometry (ICPMS) with isotope dilution or nickel sulphide fire assay pre‐ concentration, all PGE can now be measured together. Because the PGE show an inhomogeneous distribution (they form micrometre scale nuggets), large enough samples are needed (> 1 g).

When PGE concentrations are plotted on a logarithmic‐scale CI‐normalized diagram (most commonly in order of decreasing condensation temperature), a relatively flat pattern points towards a chondritic meteorite, while sloping indicates an iron. However, also a mantle component or the presence of multiple varied target rocks (e.g. the Bosumtwi impact structure [Goderis et al., 2007]), which are not easy to determine, can affect the PGE pattern.

More recently, a different approach based on the determination of PGE ratios has made possible the precise identification of the meteorite, without accounting for the target component [McDonald et al., 2001; McDonald, 2002; Tagle & Claeys, 2005; Tagle and Hecht, 2006; Goderis et al., 2009; and Tagle et al., 2009]. Linear regression analysis can be carried out by plotting the PGE ratios on a X‐Y diagram. This produces a mixing line between the target rock and meteorite. Its slope will only be slightly affected by the target rock, even for low meteorite contributions, because of the extreme differences in PGE abundances between crustal target rocks and thus will be characteristic for the meteorite. Also the nugget effects of the PGE are ruled out this way. Because condensation processes in the solar nebula mainly controlled the fractionation of PGE in chondrites, elements with large differences in condensation temperatures present the strongest variations in element ratios. Element ratios of low condensation temperature elements (Rh or Pd) combined with

PAGE 38 those with higher condensation temperatures (Os, Ir, Ru), therefore offer the best discrimination between different meteorites.

2.3.1.3. MAJOR AND TRACE ELEMENTS

Meteorite contributions due to impact events are too small to significantly affect the major element compositions of the impact rocks. A 5 wt.% contribution of a 5 wt.%

SiO2 iron meteorite would only lower a 65 wt.% SiO2 target rock to a value of 62 wt.% SiO2. Such depletion would probably still not fall significantly outside the uncertainty range. However, together with trace elements compositions, mixing calculations can be made when all target rocks are chemically characterised. Major and trace elements compositions are also needed to detect alteration and unknown, possibly (ultra)mafic, target components which can interfere with the siderophile signatures. After all, impact events excavate and mix a large volume of target rock, also to depths which are often not sampled for target rocks.

2.3.2. ISOTOPE RATIOS

Small isotope variations induced by the addition of almost indiscernible meteorite contributions into impact rocks are now measured with high precision thanks to the development of isotope geology, ion exchange chromatographic separation, and thermal ionization and multi‐collector (MC) ICP‐MS. The Os and Cr isotope systems are the most common used ones.

2.3.2.1. OS ISOTOPE SYSTEM

Os has seven natural occurring isotopes, which are all stable. These are 184Os, 186Os, 187Os, 188Os, 189Os, 190Os, and 192Os. They have relative abundances of respectively 0.024 %, 1.600 %, 1.510 %, 13.286 %, 16.252 %, 26.369 %, and 40.958% [Faure, 1986]. 187Os is radiogenic, it is produced by the beta‐decay of 187Re, which has a half‐ life of 4.16 x 1010 y [Walker et al., 2002]. As already stated, both Os and Re are PGE, thus highly siderophile and depleted in the primitive mantle. But while Os is compatible during partial melting, Re is incompatible. This results in the enrichment of Os in the mantle and Re in the crust. Therefore, crustal rocks have a high Re/Os ratio, and consequently also high 187Os contents relative to other non‐radiogenic Os isotopes (e.g. 188Os). The older the rocks, the higher these ratios. The commonly used 187Os/188Os average for the upper continental crust is 1.4 [Peucker‐Ehrenbrink & Jahn, 2001]. Meteorites have high amounts of both Re and Os, and except for the differentiated achondrites, Os even exceeds Re. This resulted into low 187Os/188Os

PAGE 39 ratios that only changed slowly throughout time. The ratio now ranges from 0.11 to 0.18 in meteorites with an average of approximately 0.13 for the chondrite meteorites [Koeberl & Shirey, 1993 and Tagle & Hecht, 2005].

While the Os meteorite contribution is already significant relative to the low Os contents of crustal target rocks, it induces an even more substantial decrease in the 187Os/188Os ratio of the impact rock, relative to the target rock. This is not used for meteorite identification, but for impact identification. Os isotopes can precisely detect the amount of the meteorite contribution (except for some differentiated achondrites). Off course, older impact rocks usually require a Re correction to obtain the exact initial 187Os/188Os ratio and also a significant mantle component must be excluded from the target rocks. However, because most chondritic and iron meteorites still have two orders of magnitude higher PGE contents than the upper mantle, it takes a 100 times greater mantle than meteorite contribution, for the same ratios. Such significant mantle contribution is easily detected.

2.3.2.2. CR ISOTOPE SYSTEM

Cr has 4 natural occurring isotopes. These are 50Cr, 52Cr, 53Cr and 54Cr. The latter three are stable. They have relative abundances of respectively 4.345 %, 83.789 %, 9.501 % and 2.365 % [Faure, 1986]. 53Cr is radiogenic, it is produced by the decay of 53Mn, which has a half‐life of 3.7 x 106 y. Because of this relative short half‐life, 53Mn has already extinct for a long time. However, it existed during the formation of the first solids and planetesimals. Due to heterogeneities or mass fractionation, these developed differences in their 53Cr/52Cr (52Cr is not radiogenic) which remained unchanged after all 53Mn decayed. Because the Earth formed afterwards, it contains no variations of the 53Cr/52Cr ratio. Any deviation of a 53Cr/52Cr ratio to the Earth’s value, implicates therefore a meteorite contribution.

Because Cr was not completely scavenged from the primitive mantle due to its partial lithophile character, meteorite identification by the 53Cr/52Cr ratio, requires a relatively high meteorite contribution in the order of a few wt.%, depending on the Cr contents of the target rock (approximately 1.2 wt.% of meteorite contribution is needed for 185 ppm Cr in the target rock). If such contribution is present, it can discriminate between the iron and chondrite groups.

2.3.2.3. OTHER ISOTOPE SYSTEMS

Also the isotope systematics of W, Pb, Nd and Sr have been proposed for impact and meteorite identification. However, their uses are still unclear because of the variable

PAGE 40 degrees of success that were made. More studies are needed on these to determine the information their signatures could reveal and how this information will differ from other signatures.

2.3.3. MULTI‐SIGNATURE APPROACH

Combining Ni, Co and Cr with Ir to Ni/Ir, Co/Ir and Cr/Ir ratios, can also discriminate between iron and chondrite groups. But additionally, a strong correlation between the moderately and strongly siderophile elements supports a common origin and can exclude the presence of an interfering target contribution as well as significant post‐ impact fractionation and remobilisation [Palme, 1980]. When no target rock composition is available, the Ni, Co and Cr contents can be retrieved this way by linear regression analysis. This because of the extreme low crustal abundance of Ir, while Ni, Co and Cr have much higher abundances. For Ir, an upper continental crust or even zero concentration is assumed.

Also the isotope systems can be used together with the siderophile element abundances and ratios. Estimations by these on the amount of meteorite contribution can be compared with values obtained by the precise 188Os/187Os ratio. This can validate the conclusions drawn from the siderophile elements abundances and ratios. Recently, together with the PGE ratios, also databases on the 188Os/187Os ratios of meteorites are constructed [Fischer‐Gödde et al., 2010]. The Cr isotope system can be used, together with the absence of moderately siderophile elements and PGE, to discriminate an achondrite meteorite better, or when the siderophile elements provide no decisive answer.

2.4. BRENT IMPACT STRUCTURE

The Brent impact structure is named after the nearby village of Brent and located in Ontario, Canada, near the northern boundary of the Algonquin Provincial Park, 75 km east of Nipissing (46°05’N, 78°29’W). It is visible as a circular depression, approximately 3 km in diameter and at its deepest point 60 m below the surrounding terrain [Dence, 1972]. At its edges, it is partly filled by two curved , named Gilmour and Tecumseh. Already from its discovery by J.A. Roberts in 1951 and first appearances in literature, it was considered to be an impact structure [Beals et al., 1956; Beals, 1958; and Millman et al., 1960]. But before the general acceptance of the

PAGE 41 meteorite impact theory, a few workers still interpreted the Brent impact structure as the product of explosive alkaline volcanism, related to the igneous centre at Callander Bay, 80 km to the west [Currie & Shafiqullah, 1967 and Currie, 1971].

Figure 16. Location of the Brent Impact Structure (modified from PASSC [2010]). The Brent impact structure is marked by a green dot. The red dots represent other impact structures.

In 1955, 1959, 1960, and 1967, the Brent impact structure has been drilled by the Dominion Observatory, which is now called the Earth Physiccs Branch of the Department of Energy, Mines and Resources, Ottawa, Canada. A total of 5035 m of core was recovered from 12 drill hooles (Figure 18) [Dence, 2004]. This extensive drilling revealed a post‐impact sediment fill, which is 264 m thick iin the central drill hole 1‐59. It preserved a simple impact crater which was only eroded for approximately 50 m from the surface at impact, corresponding with an erosion level of 4 [Dence, 1968, 1973; and Dence et al., 1976]. The parautochthoonous basement is reached at a depth of 894 m in 1‐59 [Grieve & Cintala, 1982]. Between the sediment fill and basement, a breccia lens with a maximum thickness of 630 m fills the impact crater [Dence, 1968].

PAGE 42

Figure 17. Aerial Photograph of the Brent Impact Structure (modified from Ryan [2003]). Visible are the circular depression and two curved lakes. The left lake is Gilmour and the right Tecumseh.

Lozej & Beales [1975] dated the Brent impact event at 450 ± 9 Ma by 40Ar‐39Ar analysis. Hartung et al. [1971] found a minimum of 414 ± 20 Ma by y 40K‐40Ar analysis. Recalculations of these dates by PASSC [2010], using the decay constants of Steiger & Jäger [1977] delivered an age of 396 ± 20 My. However, using microfossils, Grahn & Ormö [1995] obtained a minimum age of 452 My, which by Dence [[2004] is assumed to be more reliable.

2.4.1. IMPACT DIMENSIONS

Based on observations and theoretical models, some impact dimensions were determined for the Brent impact structure, all are approximations. The meteorite projectile had a size of 105 to 120 m, a velocity of 27 to 33 km/ss, and penetration depth of 240 m. Its impact released an energy of 1018 J [Dence et al,, 1976 and Grieve & Cintala, 1981, 1982] and formed a simple impact crater with a diiameter of 3.8 km [Dence, 1972] and depth of 1.1 km below the original surface [Dence et al., 1976].

PAGE 43 Observational data by Grieve & Cintala [1981] confirm this diameter and depth, supplementary indicating a rim height of 200 m and volume of 7.4 km3. From this, Garvin & Grieve [1982] estimate a transient crater with a diameter of 3.3 km, 3.1 km at the original surface [Grieve & Cintala, 1982], and rim height of 250 m, resulting in a volume of 5.7 km³. With a bulking factor of approximately 10% for the transition from crystalline rock to breccia [Innes & Beals, 1961], this implies a volume difference between the transient and final crater of 1.9 km³, which represents the amount of rocks that slumped during the modification stage and formed the breccia lens. This is in agreement with the observed 2.1 km³ of breccia lens [Grieve & Cintala, 1982], which possibly also includes a minor, unrecognizable, part of fall‐back ejecta.

Dence et al. [1976] estimate that at least 0.5 km³ of the excavated target rock was exposed to shock pressures above 60 GPa, thus resulting into vaporisation and melting. Of this, approximately 0.022 km³, or 4.4 %, is incorporated in the breccia lens [Grieve & Cintala, 1981]. Robertson and Grieve [1977] estimate that shock pressures were in the order of 23 GPa at the deepest point of the impact crater. At the rim of the impact crater these are estimated to have been 0.3 GPa [Grieve & Cintala, 1982].

2.4.2. TARGET ROCK

The Brent meteorite impacted on an igneous‐metamorphic basement complex of the Grenville structural province of the . The target rock is mesoperthite and microcline gneiss of granodioritic composition with a minor contribution of mafic gneiss [Grieve, 1978]. Rare Cambrian alnoite dikes are also present [Hartung et al, 1971 and Grieve & Dence, 1978]. This target rock can physically be treated as homogeneous and isotropic [Grieve & Cintala, 1982 and Dence, 2004]. Its regional setting is described by Currie [1971] and Dence & Guy‐Bray [1972].

Table 1. Element concentrations of the gneiss (mesoperthite) and alnoite target rocks at the Brent impact structure. The PGE, Re and Au are derived from target rock clasts [Palme et al., 1981]. The others are coming from parautochthonous basement and surface samples [Grieve, 1978]. FeO is total Fe.

wt.% SiO2 TiO2 Al2O3 FeO MnO MgO CaO Na2O K2O Gneiss 63.30 0.91 15.01 06.52 0.20 01.18 01.78 4.61 4.92 Alnoite 37.80 3.37 11.25 10.70 0.24 10.50 11.35 1.80 4.02 ppm Sc V Cr Co Ni Cu Zn Rb Sr Gneiss 04.8 020 012 08.0 010 16 0120 136 136 Alnoite 21.7 300 350 88.1 327 70 1200 040 664 ppb Zr Ba Ge Se Pd Au Re Os Ir Gneiss 900000 938000 0774 027.7 < 5.3 < 0.26 0.021 < 0.040 < 0.06 Alnoite 300000 500000 3586 178.7 < 4.6 < 0.30 0.585 < 0.034 < 0.01

PAGE 44 The chemical composition of the alnoite dikes in Table 1 shows that the presence of an alnoite component could explain possible siderophile enrichment relative to the gneiss target rock, that otherwise would be attributed to the meteorite contribution. However, both Grieve [1978] and Palm et al. [1981] state that thee presence of the alnoite dikes in the target rock is too low to have a significant influence on a siderophile enrichment of the melt‐bearing impact rocks.

2.4.3. BRECCIA LENS

The crater‐fill rocks of the Brent impact structures comprise a 630 m thick breccia lens, as is expected for a simple impaact crater. Although very compllex in detail, three distinct zones with different impact rocks can be distinguished: lithic breccias, melt‐ fragment breccias and a melt lens.

Figure 18. Schematic cross section of the Brennt Impact Structure with drill holes (modified from Grieve [1978] and Dence [2004]). The impact surface is indicated by horizontal dashed llines. The post‐impact sediment fill is marked with horizontal stripes, the lithic breccia zone with triangles, the high melt‐ fragment part of the melt‐fragment lenses witthh vertical striped lenses, the low meltt‐fragment lenses with diagonal striped lenses, and the melt lens is coloured black. The marked part of tthe parautochthonous basement, beneath the melt lens, indicates shock pressures above 23 GPa.

2.4.3.1. LITHIC BRECCIAS

Most of the breccia lens is composed out of lithic breccias. They contain slabs of fracturred and brecciated target rock. The clasts are coarse and are only lowly shocked. In the centre of the breccia lens, especially beneath the melt lens, highly shocked lithic breccias occur. Except for these, most lithic breccias did not experience shock pressures exceedingg 16 GPa [Robertson & Grieve, 1977].

PAGE 45 2.4.3.2. MELT‐FRAGMENT BRECCIAS

At the top of the breccia lens, around the axis, melt‐fragment breccias are present as lenses between the lithic breccias. The target rock clasts here, experienced shock pressures of approximately 25 to 45 GPa. The upper part of the lenses have impact melt contents of average 13 vol.%, but some zones reach up to 30 vol.%. This high melt‐fragment part is intersected between 264 and 427 m in 1‐59 and its radial extent is approximately 1.1 km. The lower part of the lenses, intersected between 427 and 606 m in 1‐59, occurs only occasionally in 1‐67, suggesting a radial extent of about 200 m. They have an average impact melt content of only 6 to 7 vol.% [Grieve, 1978 and Grieve & Cintala, 1981]. Such low melt‐fragment values are also found for the edges of the upper part and for one melt‐fragment breccia layer that is intersected between 741 and 763 m in 1‐59 and 1‐67.

The impact melt is present as alterated inclusion‐rich clasts of glass. Most of them are smaller than 5 cm, but some range up to 25 cm [Grieve & Cintala, 1982]. They have sinuous or contorted shapes and can occur as chilled rinds around target rock clasts. When mixed with the target rock clasts, they were at least in a plastic state, but because they lack distinctive aerodynamic shapes, it is therefore assumed that they have never been ejected [Grieve et al., 1977]. Together with the observation that the melt‐fragment breccias are not continuous throughout the crater, the breccia lens is interpreted as being for the major part derived from the slumping of displaced rocks during the modification stage [Grieve, 1978], as is expected for simple impact craters.

The target rock clasts that suffered the highest shock pressures, are partially melted and vesiculated, and show heavy alteration. Also six zones of melt‐fragment breccia with a vesiculated matrix occur in 1‐59. These range up to 5 m in thickness. Their central portions consist of less than 25 % recrystallized feldspar inclusions in a fine‐ grained matrix of felted feldspar laths up to 0.45 mm in length. The vesicles are up to 1.25 cm in diameter and can be filled with zeolites or less commonly barite [Grieve, 1978]. The vesiculated target rock clasts are distinguishable from the vesiculated matrix by a brick‐red colour and the absence of inclusions [Grieve & Dence, 1978].

2.4.3.3. MELT LENS

[Grieve, 1978]

Near the base of the breccia lens, a 34 m melt lens is intersected in 1‐59, between 823 and 857 m. Its exact radial extent is not known. However, in 1‐67, 200 m to the

PAGE 46 west, it is not found. Instead, a zone of pyroxene hornfels is intersected between 824 and 840 m. Because the pyroxene hornfels present in 1‐59 extends only 10 m away from the melt zone, it is assumed that the melt lens comes close to 1‐67, implying a radial extent of only slightly less than 200 m. Assuming a 190 m radius and lenticular shape, Grieve & Cintala [1982] estimate the volume of the melt lens to be 0.0019 km³, of which 50 to 55 vol.%, or approximately 0.001 km³, is impact melt. This implies that only 4.5 vol.% of all the impact melt in the breccia lens in situated in the melt zone [Grieve & Cintala, 1981], while it was previously estimated to be 1 vol.% by Dence [1971] and 2 vol.% by Grieve [1978].

The impact melt is microcrystalline near the upper and lower edge of the melt zone, but the grain size increases rapidly to the centre. At 4 m from the upper edge, the average modal composition is reached, which is 70 vol.% feldspar, 13 vol.% mesostasis sheet silicate, 7 vol.% pyroxene, 3 vol.% amphibole, 3 vol.% quartz, 3 vol.% opaques and 1 vol.% apatite. The feldspars are tabular and above 0.35 mm in size. They have well‐developed cores and rims. The core consists of an altered intergrowth of K‐ and Na‐feldspar. Between 843 and 850 m, the feldspar cores also contain zones of unaltered feldspar. These range in composition from K‐oligoclase to anorthoclase. The feldspar rims are always unaltered sanidine, often with marginal perthite and needles of apatite. The compositions of the feldspars from the target rock gneiss are within the compositional range of the intergrown K‐ and Na‐feldspar cores. It is believed by Grieve [1978] that the latter are xenocrystal feldspars of the target rock gneiss. The K‐oligoclase is considered to derive from the melt phase. The pyroxene is ferro‐augite and interstitial to the feldspar and often associated with and altered to interstitial sheet silicate. The amphiboles occur as grains with a arfvedsonite core and riebeckite overgrowth. The quartz is anhedral an interstitial and where in contact with the mesostasis, it has a reaction rim. The mesostasis sheet silicate is microcrystalline and of unknown composition. It may represent altered interstitial glass.

The other 45 to 50 vol.% of the melt zone consists out of target rock inclusions. Most of them are microscopic recrystallized feldspar. They make up 20 to 95 vol.% of the upper 5 m and lower 8 m of the melt zone. These rocks can be defined as melt‐matrix breccias. Within the central 21 m of the melt zone, the target rock inclusions vary from 16 to less than 1 vol.%, which makes this zone almost pure impact melt. A number of macroscopic, up to 1 m in size, recrystallized target rock xenoliths are also present.

PAGE 47 2.4.4. ALTERATION

The Brent structure is reported to have experienced a significant amount of alteration, especially resulting in depleted SiO2 and enriched K2O contents [Grieve &

Dence, 1978]. The high K2O contents up to 14 wt.% are encountered in the melt‐ fragment breccias and melt lens. They are not proportional to the amount of melt, but to the shock pressures the target rock clasts and inclusions experienced, with a threshold of approximately 20 GPa. The lower shocked lithic breccias with approximately 5 wt.% (~ target rock) do not show elevated K2O contents. This is reported by Grieve [1978], who states that this K enrichment arose from alkali exchange between feldspars and saline aqueous solutions during post‐impact cooling. Because Lozej & Beales [1975] concluded that the Brent impact event occurred on the edge of a transgressive shallow tropical sea and that the impact crater was inundated soon after formation, it is likely that such saline fluids were present and provided a source of K. The threshold is probably related to the onset of diaplectic glass formation. Palme et al. [1981] also explain depleted Re and Au contents this way by hydrothermal mobilisation.

2.4.5. GEOCHEMICAL IDENTIFICATION

A few geochemical impact and meteorite identification studies have already been conducted for the Brent impact structure. Grieve [1978] reported correlating Ni and Cr values up to respectively 575 ppm and 120 ppm, indicating an impact origin. The target rocks inclusions in the melt lens contained 35 ppm Ni at average. This excess of 25 ppm relative to the target rock reported in Table 1 was explained as contamination of melt in the analysed samples, or as the actual Ni content of the target rock inclusions by incorporation of condensed meteoritic vapour or melt. An average Ni content of 253 ± 133 ppm for the impact melt in the melt lens was obtained this way and a Ni/Cr ratio of 2.76 ± 0.36. Assuming that the meteorite contribution controls this ratio, a chondrite meteorite and more specific an ordinary L was suggested.

Table 2. Some siderophile element concentrations of two samples from the melt lens used by Palme et al. [1981] (more samples were used). The sample name corresponds to the drill hole and depth in foot. Os (ppb) Ir (ppb) Pd (ppb) Au (ppb) Ni (ppm) Co (ppm) Cr (ppm) 59‐2773.3 1.26 2.68 0.055 0.15 151 10.5 73 59‐2800.2 9.22 9.57 18.1 0.93 320 21.1 103

PAGE 48 Based on a flat CI‐normalised pattern of the siderophile elements listed in Table 2, Palme et al. [1981] also suggest a chondrite meteorite. A L or LL ordinary chondrite is derived, based on a Ni/Cr ratio of 2.70 ± 0.26 for the impact melt. A target rock value of 20 ppm Ni and 23 ppm Cr is used to obtain this ratio. Melt lens analyses for PGE by Evans et al. [1993] are listed in Table 3. These PGE values are said to be not in agreement with a L or LL ordinary chondrite. The Pt/Ir ratio of 1.95 suggest a CI or iron meteorite. However, the LL data to compare with are limited and the PGE CI‐ normalised pattern is relatively flat. Together with the data of Palme et al. [1981], an increase of PGE abundances with increasing depth is reported, possibly an indication for a fractionation of the meteorite contribution during post‐impact cooling or by the hydrothermal alteration. This could explain the PGE deviation.

Table 3. Some siderophile element concentrations of the two samples from the melt lens used by Evans et al. [1993]. The sample name corresponds to the drill hole and depth in foot. Pt (ppb) Pd (ppb) Ru (ppb) Ir (ppb) Au (ppb) Rh (ppb) 59‐2778.9 11 12 5.9 1.8 6.7 2.0 59‐2781.0 15 8.8 7.8 3.0 11 2.2

A mixing model by Grieve & Dence [1978], based on major and trace elements, indicates that the melt lens corresponds to a mix of 97.8 % target rock and 2.2 % meteorite, while values of respectively 98.4% and 1.6% were obtained by Grieve & Cintala [1981].

PAGE 49 3. METHODS

To determine the geochemical signatures from the meteorite contribution in the impact rocks of the Brent impact structure, twenty samples from the Dominion Observatory drilling are obtained, including target rock, melt‐fragment breccias and melt lens rocks. For all twenty samples, Ni, Co and Cr are determined. For thirteen of them also the major elements and PGE are analysed. For few selected samples, the whole rock isotope ratios are measured of the common used Os and recently proposed Pb isotope systems.

3.1. SAMPLES

Table 4. The twenty used samples from the Brent impact structures. Given are the depth (m), mass (g) and the kind of impact rock. Sample names are derived from their drill hole and depth in foot, as in literature. * These samples have a dark green colour and are labelled as chloritized based on thin section observations by the Earth Physics Branch of the Department of Energy, Mines and Resources, Ottawa, Canada. Sample Depth Mass Impact Rock 59‐951 289.86 34.47 Melt‐fragment breccia 59‐951.5 290.02 43.90 Melt‐fragment breccia 59‐1050 320.04 51.37 Melt‐fragment breccia 59‐1379 420.32 45.81 Melt‐fragment breccia* 59‐1929.5 588.11 28.32 Melt‐fragment breccia 59‐2762 841.86 33.55 Melt lens 59‐2771 844.60 49.24 Melt lens 59‐2781 847.65 42.46 Melt lens 59‐2794 851.61 40.49 Melt lens 59‐2800 853.44 48.61 Melt lens 59‐2880 877.82 35.00 Target rock 67‐888 270.66 64.95 Melt‐fragment breccia* 67‐926 282.24 51.58 Melt‐fragment breccia 67‐926.5 282.40 30,76 Melt‐fragment breccia 67‐933.5 284.53 26.33 Melt‐fragment breccia 67‐1169 356.31 32.76 Melt‐fragment breccia* 67‐1446 440.75 49.00 Melt‐fragment breccia 67‐2630 801.62 27.12 Target rock 67‐2702 823.57 40.79 Melt‐fragment breccia 67‐3813 1162.20 41.98 Target rock

PAGE 50 The twenty samples have masses between 26.33 and 67.95 g. Eleven are from drill hole 1‐59, five of them are melt‐fragment breccias, another five are from the melt lens. One sample is a target rock clast from the highly shocked lithic breccias beneath the melt lens. The other nine samples are from drill hole 1‐69. Six of these are melt‐ fragment breccias, one is a target rock clasts from a lithic breccia, and one is a target rock from the parautochthonous basement. Sample names are derived from their drill hole and depth in foot. This to be in agreement with sample names used in literature.

3.2. MAJOR ELEMENTS

Thirteen samples are analysed for Si, Ti, Al, Fe, Mn, Mg, Ca, K, Na and P. This by inductively coupled plasma atomic emission spectroscopy at Ghent University. To determine the accuracy, the BCR‐2 (United States Geological Survey, Denver, USA) and SARM‐44 (South African Bureau of Standards, Pretoria, South Africa) certified standards are added to the analysis.

Each powdered sample is dried at 110 °C. Loss on ignition (LOI) is determined by heating at 850 °C for 2 h. Afterwards, to approximately 200 mg, a 1 g flux of 1:1 Li‐ meta‐/‐tetraborate (LiBO2/Li2B407) is added to the samples. The resulting mixtures are homogenised and transferred to high purity graphite crucibles. These are than heated at 800 to 1000 °C. When a melt is formed, this is dissolved by 2 vol.% diluted

HNO3 and stirred to accelerate the reaction.

3.3. NI, CO AND CR

To obtain Ni, Co and Cr concentrations, all twenty samples are quantitatively analysed for the trace elements 52Cr, 59Co, 60Ni, 68Zn, 71Ga, 84Sr, 85Rb, 86Sr, 89Y, 91Zr, 93Nb, 115In, 118Sn, 121Sb, 137Ba, 140Ce, 146Nd, 172Yb, 205Tl, 206Pb, 232Th and 238U. This is done by ICP‐MS with a XSERIES 2 from Thermo Scientific. In and Tl are added as internal standards. To determine the accuracy, the PM‐S (Service d'Analyse des Roches et des Minéraux, Nancy, France) and DNC‐1 (United States Geological Survey, Denver, USA) certified standards are added to the analysis. Each sample is measured twice on two different days to enlarge the analytical confidence. Sample preparation comprises the acid digestion procedure as described by Tagle et al. [2007].

PAGE 51 Approximately 100 mg of each powdered sample is transferred into a 15 ml perfluoroalkoxy (PFA) beaker, discharged with an anti‐electrostatic gun. To dissolve, first 2 ml 14M HNO3 is added slowly and then 4 ml 22M HF. The covered beaker is put on a 125 °C hot plate for 24 to 48 hours to accelerate the reaction. After 45 minutes of cooling, the beaker is opened and put on a 75 to 90 °C hot plate for one night to evaporate the solution. After adding 1 ml of 6 µg/ml In and 1 ml of 3.25 µg/ml Tl as internal standards, the solutions were brought to near‐dryness and this solution is evaporated on a 75 to 90 °C hot plate. The residue is taken up in a 1 ml 7M

HClO4 and 1 ml saturated H3BO3 water solution. This again, is evaporated on a 75 to

90 °C hot plate. The metal rich residue is dissolved in aqua regia (1:3 HNO3:HCl).

After drying down, 2 drops of 5 % H2O2 followed by 2 ml 14 N HNO3 are added to ensure oxidation of all Fe to Fe3+. The solutions are dried down again 30 minutes later. The precipitate is dissolved in 10 ml 14 M HNO3 and heated on a 120 °C hot plate for 30 min to accelerate the reaction. Finally, the solutions are quantitatively transferred into a 250 ml polymethylpentene (PMP) volumetric flask and diluted with ultrapure 18.2 MΩ/cm milli‐q water.

3.4. PLATINUM GROUP ELEMENTS

Thirteen samples are analysed for the PGE 99Ru, 101Ru, 102Ru, 103Rh, 105Pd, 106Pd, 108Pd, 191Ir, 193Ir, 194Pt, 195Pt and 196Pt. Os volatilises during the applied sample preparation. In addition, also 197Au is measured. This is done by ICP‐MS at Ghent University. In and Tl are added as internal standards. To determine the accuracy, the PGE‐poor KTB‐FDO and PGE‐rich WPR‐1A standards are added to the analysis, both are in‐ house standards. PGE concentrations are near or below the detection limits in most crustal rocks. Therefore, a pre‐concentration method is applied to become a precise quantitative analysis. We use the nickel‐sulphide fire assay procedure as described by Plessen & Erzinger [1998] and Tagle & Claeys [2005].

Approximately 20 g (10 g is the minimum to account with certainty for the nugget effect [Tagle & Claeys, 2005]) of each powdered sample is transferred into a polypropylene container and a fire assay flux is added: 12 g of dehydrated sodium carbonate (Na2CO3), 24 g of dehydrated sodium tetraborate (Na2B4O7), 2 g of calcium fluoride (CaF2), 2 g of sulphur (S), and 2 g of nickel (Ni). This mixture is homogenized with an agate pestle and transferred into a fire clay crucible. The crucibles is placed in a furnace and heated to 1150 °C for an hour. Hereby the mixture melts and due to the flux, a heavy NiS phase forms in which the PGE are partitioned. This will sink to

PAGE 52 the bottom and fuse into a button. Once cooled, the crucible is broken with a hammer and the NiS button is removed by hand. The button is dissolved in a 300 ml Erlenmeyer flask in 200 ml of 38 vol.% HCl. The Erlenmeyer flask is covered to avoid splashing and placed in a 90 °C water bath inside a fume hood. After approximately 24 h the dissolution is complete and the cooled solution is filtered through a micro‐ filter disc of glass (brand: Schott; porosity: 4) in a Büchner funnel into a Büchner flask that is connected to a water aspirator. The Büchner funnel is washed with HCl at the end (40 vol.%, twice distilled). The PGE containing residue (only the NiS dissolves in the warm HCl) is dissolved with a mixture of two parts HCl (32 vol.%, twice‐distilled) and one part H2O2 (30 vol.%, analytical grade) in three subsequent steps. Finally, the solution is passed through a filter paper, transferred to a 250 ml beaker and evaporated on a hot plate to a volume of 1 a 2 ml. After adding a final 5 ml of HCl and 2 ml of H2O2, the solution is evaporated to 1 a 2 ml, cooled and diluted with 2% HCl to a volume of 10 ml.

3.5. ISOTOPES

In addition to the major elements, PGE and Ni, Co and Cr, also whole rock ratios for the isotope systems of Os and Pb were obtained for selected samples. These analyses were conducted with MC‐ICP‐MS by respectively F. Paquay at the University of Hawaii and V. Renson at the Vrije Universiteit Brussel.

PAGE 53 4. RESULTS

4.1. ELEMENT ABUNDANCES AND RATIOS

4.1.1. MAJOR ELEMENTS

Table 5 lists the results from the major elements analysis. The concentrations are calculated by constructing a calibration curve with quantitative analyses of a method blank, and the BHVO‐2, AGV‐2, QLO‐1, DTS‐2, GSP‐2 (United States Geological Survey, Denver, USA), JSy‐1 (Geological Survey of Japan, Ibaraki, Japan), Mica‐Fe, BX‐ N, IF‐G (Centre de Recherches Petrographiques et Geochimiques, Nancy, France) and NIM‐L (Mintek, Randburg, South Africa) certified standards.

Table 5. Results of the major element analysis. Standard errors are 0.20 wt.% for SiO2, 0.01 wt.% for TiO2,

0.11 wt.% for Al2O3, 0.03 wt.% for Fe2O3, 0.01 wt.% for MnO, 0.01 wt.% for MgO, 0.01 wt.% for CaO, 0.04 wt.% for K2O, 0.02 wt.% for Na2O, and 0.01 wt.% for P2O5. Limits of detection are 0.003 wt.%. Fe2O3 includes all Fe. wt.% SiO2 TiO2 Al2O3 Fe2O3 MnO MgO CaO K2O Na2O P2O5 LOI 59‐951 54.62 0.75 17.96 10.21 0.07 4.84 0.33 10.28 0.24 0.13 3.50 59‐1050 52.18 1.74 16.57 13.08 0.15 5.50 0.60 8.97 0.66 0.26 3.10 59‐1379 33.77 1.28 17.13 34.31 0.26 11.64 0.95 0.08 0.87 0.36 7.96 59‐1929.5 53.52 0.79 23.05 8.52 0.19 2.50 0.51 8.70 1.17 0.18 2.19 59‐2762 53.58 0.71 24.04 8.35 0.18 1.68 1.90 6.31 3.00 0.18 1.56 59‐2771 56.91 0.75 19.44 9.65 0.28 1.43 2.60 4.89 4.26 0.19 1.04 59‐2781 57.45 0.80 19.19 9.22 0.24 1.48 2.52 5.62 4.11 0.20 1.07 59‐2800 51.66 0.89 24.00 10.03 0.20 2.49 2.37 5.72 3.01 0.23 1.47 67‐888 33.89 0.68 18.02 31.87 0.15 14.25 0.65 0.06 0.76 0.17 8.44 67‐926 54.83 0.59 18.78 9.87 0.05 4.35 0.22 10.46 0.47 0.11 2.75 67‐1169 33.08 1.59 17.68 33.17 0.29 11.46 1.33 0.08 0.89 0.52 7.78 67‐1446 62.03 0.66 14.51 9.87 0.12 2.83 0.94 5.56 2.23 0.14 2.69 67‐2702 55.68 1.02 15.47 11.65 0.19 4.56 0.60 9.59 0.35 0.27 3.12

The obtained values for the BCR‐2 (54.27 ± 0.46 wt.% SiO2, 2.26 ± 0.02 wt.% TiO2,

13.45 ± 0.01 wt.% Al2O3, 13.80 ± 0.10 wt.% Fe2O3, 0.20 ± 0.01 wt.% MnO, 3.58 ± 0.02 wt.% MgO, 7.13 ± 0.05 wt.% CaO, 1.77 ± 0.01 wt.% K2O, 3.15 ± 0.01 wt.% Na2O, and

0.36 ± 0.01 wt.% P2O5) and SARM‐44 (34.8 ± 0.74 wt.% SiO2, 1.86 ± 0.03 wt.% TiO2,

59.81 ± 0.16 wt.% Al2O3, 2.15 ± 0.19 wt.% Fe2O3, 0.03 ± 0.01 wt.% MnO, 0.13 ± 0.05

PAGE 54 wt.% MgO, 0.12 ± 0.09 wt.% CaO, 1.79 ± 0.02 wt.% K2O, 0.03 ± 0.06 wt.% Na2O, and

0.10 ± 0.01 wt.% P2O5) certified standards are in good agreement with the published

BCR‐2 (54.1 ± 0.8 wt.% SiO2, 2.26 ± 0.05 wt.% TiO2, 13.5 ± 0.2 wt.% Al2O3, 13.8 ± 0.2 wt.% Fe2O3, 0.2 ± 0.01 wt.% MnO, 3.59 ± 0.05 wt.% MgO, 7.12 ± 0.11 wt.% CaO, 1.79

± 0.05 wt.% K2O, 3.16 ± 0.11 wt.% Na2O, and 0.35 ± 0.02 wt.% P2O5) and SARM‐44

(34.84 wt.% SiO2, 1.83 wt.% TiO2, 58.80 wt.% Al2O3, 2.06 wt.% Fe2O3, 0.03 wt.%

MnO, 0.10 wt.% MgO, 0.14 wt.% CaO, 0.18 wt.% K2O, 0.05 wt.% Na2O, and 0.10 wt.%

P2O5) values. SARM‐44 was measured together with samples 51‐1379, 67‐888 and 67‐1169. BCR‐2 was measured with the other samples.

Figure 19. Target‐normalised results of the major element analysis. The target rock values used for normalisation are that of Grieve [1978] listed in Table 1. P2O5 is not reported because it is not measured by Grieve [1978]. The values for the upper continental crust (UCC) are added as a reference [Rudnick &

Gao, 2003]. Fe2O3 includes all Fe.

SiO2 concentrations vary from 33.08 to 62.03 wt.%, TiO2 from 0..66 to 1.74 wt.%,

Al2O3 frfrom 14.51 to 24.04 wt.%, Fe2O3 from 9.87 to 34.31 wt.%, MnO from 0.05 to

0.29 wt.%, MgO from 1.43 to 14.25 wt.%, CaO from 0.22 to 2.60 wt..%, K2O from 0.06 to 10.46 wt.%, Na2O from 0.24 to 3.011 wt.%, and P2O5 from 0.11 to 0.52 wt.%. Figure

PAGE 55 19 shows the target‐normalised results of the major element analysis. However, no target rock is measured because no target rock samples were available at the moment of the analysis. Therefore, the target rock used for normalisation is that of Grieve [1978], listed in Table 1. Because lithic breccias, deep parautochthonous basement (≥ 911 m) and surface samples are used for the analysis (n = 10), this is assumed to be a representative target rock.

Three melt‐fragment breccia samples (59‐1379, 67‐888 and 67‐1169; marked red in Figure 19) show a total different composition than the other samples. They have average contents of 33.58 wt.% SiO2, 1.18 wt.% TiO2, 17.61 wt.% Al2O3, 33.12 wt.%

Fe2O3, 0.23 wt.% MnO, 12.45 wt.% MgO, 0.98 wt.% CaO, 0.07 wt.% K2O, 0.84 wt.%

Na2O, and 0.35 wt.% P2O5. Their LOI amounts 8.06 wt% at average and is significantly higher than the other samples. Relative to the target rock of Grieve [1978] these values implicate an enrichment in Fe and Mg, and depletion in Si, Ca, K and Na. As already mentioned in Table 4, these three dark green coloured samples were labelled as chloritized, which is consistent with the values we obtained.

Only the samples from the melt lens (59‐2762, 59‐2771, 59‐2781 and 59‐2800; marked blue in Figure 19) show a flat pattern that resembles the target rock of Grieve [1978], but still a depletion in Si, Ti and Na exists, while the others are enriched. Five melt‐fragment breccia samples (59‐951, 59‐1050, 59‐1929.5, 67‐926 and 67‐2702; marked green in Figure 19) are enriched in Mg and K, and depleted in Si, Ca and Na. Melt‐fragment breccia sample 67‐1446 (marked purple in Figure 19) shows concentrations intermediate to the melt lens and melt‐fragment breccia samples. Only its Si content is distinct with 62.03 wt.% SiO2, being the highest of all samples and closest to the target rock of Grieve [1978].

4.1.2. NI, CO AND CR

Table 6 lists the results from the Ni, Co and Cr analysis. The concentrations are calculated by a one‐point calibration with a quantitative analysis of the BE‐N certified standard (Service d'Analyse des Roches et des Minéraux, Nancy, France). Obtained values for the PM‐S (126 ± 2 ppm Ni, 51 ± 1 ppm Co and 314 ± 3 ppm Cr) and DNC‐1 (269 ± 3 ppm Ni, 59 ± 1 ppm Co and 286 ± 3 ppm Cr) certified standards are in good agreement with the published PM‐S (115 ± 3 ppm Ni, 49 ± 2 ppm Co and 314 ± 6 ppm Cr) and DNC‐1 (247 ± 12 ppm Ni, 57 ± 3 ppm Co and 270 ± 9 ppm Cr) values. The quantified Ni concentrations vary from 6.4 to 311 ppm, Co from 1.4 to 35 ppm, and Cr from 30 to 194 ppm. Figure 20 shows the Ni and Cr values in a Ni/Cr diagram. The three target rock samples (59‐2880, 67‐2630, 67‐3813) all have low Ni,

PAGE 56 Co and Cr concentrations, most of them even below the limit of quantification. Therefore, also the target rock of Grieve [1978] (10 ± 4 ppm Ni, 8 ± 2 ppm Co and 12 ± 2 ppm Cr) is shown in Figure 20.

Table 6. Results of the Ni, Co and Cr analysis. Standard errors are 3 ppm for Ni, 1 ppm for Co and 3 ppm for Cr. Limits of quantification are 3 ppm for Ni, 1 ppm for Co and 5 ppm for Cr. Also the Ni/Cr and Co/Cr ratios are reported. Ni (ppm) Co (ppm) Cr (ppm) Ni/Cr Co/Cr Ni/Co 59‐951 129 4.6 79 1.63 0.06 28.04 59‐951.5 109 5.7 37 2.95 0.15 19.12 59‐1050 81 35 113 0.72 0.31 2.31 59‐1379 9.8 3.6 < LOQ 2.72 59‐1929.5 28 2.1 30 0.93 0.07 13.33 59‐2762 187 12 64 2.92 0.18 15.58 59‐2771 201 13 68 2.96 0.19 15.46 59‐2781 186 13 68 2.74 0.19 14.31 59‐2794 235 15 85 2.76 0.17 15.67 59‐2800 311 20 127 2.45 0.16 15.55 59‐2880 < LOQ 3.4 < LOQ 67‐888 135 1.4 < LOQ 96.43 67‐926 97 7.9 156 0.62 0.05 12.28 67‐926.5 103 8.5 194 0.53 0.04 12.12 67‐933.5 97 12 198 0.49 0.06 8.08 67‐1169 10 2.9 < LOQ 3.45 67‐1446 < LOQ 2.3 < LOQ 67‐2630 6.4 4.1 < LOQ 1.56 67‐2702 94 7.4 64 1.47 0.12 12.70 67‐3813 < LOQ < LOQ < LOQ

This enrichment in Ni, Co and Cr relative to the target rock of Grieve [1978] implicates an impact identification and also precludes a differentiated achondrite as impact meteorite. The samples of the melt‐fragment breccias do not show a correlation, but the melt lens samples do, suggesting a common source. This is also noticed for the Co/Cr and Ni/Co ratios (for the latter also some melt‐fragment breccia samples show correlation with the melt lens samples). Taken the target rock into account, this results into 2.93 ± 0.16 Ni/Cr, 0.10 ± 0.01 Co/Cr and 29.85 ± 3.00 Ni/Co ratios for the impact meteorite. The Ni/Cr and Co/Cr ratios rule out the possibility of an iron meteorite because these all have ratios one to three orders of magnitude higher. The Ni/Cr ratio points to a LL (2.64 ± 0.21) or L (3.22 ± 0.19) ordinary chondrite, while the Co/Cr ratio approaches the LL (0.13 ± 0.02). No chondrite meteorites are known with a lower Co/Cr. The Ni/Co ratio is not consistent with a L or LL ordinary chondrite. It does not fall into the narrow chondrite range (19.34 ± 3.15 to 22.86 ± 2.83) and only corresponds to the iron meteorites [Schmieder & Buchner, 2010]. Of the melt lens samples, 59‐2800 has the

PAGE 57 highest Ni, Co and Cr values and therefore should contain the highest meteorite contribution. With only this sample, a Ni/Cr ratio of 2.62 is found, which is almost equal to that of the LL. Also the Co//Cr ratio of 0.10 is in agreemeent. For the Ni/Co ratio a value of 25.08 is obtained, which still does not significantly fiit the chondrites.

Figure 20. Ni/Cr diagram of the results of the Ni, Co and Cr analysis. The target rock values used are that of Grieve [1978], listed in Table 1.

4.1.3. PLATINUM GROUP ELEMENTS

Table 7 lists the results of the PGE analysis. The concentrations are calculated by constructing a calibration curve with standard solutions. The obtained values for the KTB‐FDO and WPR‐1A in‐house standards are in good agreemeent with previous values. The chloritized samples 59‐1379, 67‐888 and 67‐1169 are not listed because their PGE concentrations were not detected or did not exceed the limits of quantification. Figure 21 shows the CI‐normalised results of the PGE analysis. No target rock is measured because no target rock samples were available at the moment of the analysis. Also the target rock values by Palme et all. [1981], listed in

PAGE 58 Table 1, are not used as a reference in Figure 21, because they are only maxima. Probably they also are not representative for the target rock, because obtained from a melt‐fragment breccia target rock clast. The upper continental crust is shown as a reference instead.

Table 7. Results of the PGE analysis. Standard errors are 0.03 ppb for Ir, 0.09 ppb for Ru, 0.08 ppb for Pt, 0.02 ppb for Rh, 0.03 ppb for Pd and 0.03 ppb for Au. Limits of detection are 0.02 ppb for Ir, 0.03 ppb for Ru, 0.02 ppb for Pt, 0.01 ppb for Rh, 0.05 ppb for Pd and 0.05 ppb for Au. Limits of quantification are 0.06 ppb for Ir, 0.11 ppb for Ru, 0.06 ppb for Pt, 0.03 ppb for Rh, 0.17 ppb for Pd and 0.16 ppb for Au. ppb Ir Ru Pt Rh Pd Au 59‐951 5.02 11.46 13.01 2.15 6.58 1.53 59‐1050 0.22 1.64 1.66 0.31 1.52 0.72 59‐1929.5 1.27 3.48 3.70 0.65 2.21 0.23 59‐2762 2.31 5.53 6.42 1.14 3.27 0.88 59‐2771 2.78 6.42 7.83 1.31 3.94 1.13 59‐2781 2.66 6.25 7.12 1.21 3.57 0.62 59‐2800 6.21 13.37 15.16 2.57 8.27 1.47 67‐926 6.41 17.06 19.63 3.48 8.96 0.75 67‐1446 0.07 0.31 0.13 0.02 0.21 0.18 67‐2702 2.64 4.92 6.30 0.96 3.81 0.86

Ir concentrations vary from 0.07 to 6.41 ppb, Ru from 0.31 to 17.06 ppb, Pt from 0.13 to 19.63, Rh from 0.02 to 3.48 ppb, Pd from 0.21 to 8.96 ppb, and Au from 0.18 to 1.47 ppb. All minimum values are found in sample 67‐1446 and all maxima in 67‐ 926. Only Au has a maximum in sample 59‐2800, which for the other PGE is the second highest sample. Except for sample 59‐951, all samples have Au values below that of the upper continental crust. These seem to be depleted in the melt‐fragment breccia samples, but not in the melt lens samples.

The measured PGE concentrations are significantly higher than any crustal values. This implicates an impact identification and rules out a differentiated achondrite as meteorite. Most samples have a flat PGE pattern, which suggests a chondrite impact meteorite and not an iron. Only samples 59‐1050 and 67‐1446 have an odd pattern. While the latter has no correlation, the former seems to be intermediate to the upper continental crust and might represent a lesser meteorite contribution. When the upper continental crust from [Rudnick & Gao, 2003] is used as target rock and subtracted from sample 59‐1050, it forms a more flat pattern.

A linear regression analysis is carried out on the obtained PGE concentrations. Slopes and intercepts are determined for the Pd/Ir, Rh/Ir, Pt/Ir, Ru/Ir, Pd/Rh, Pt/Rh, Ru/Rh, Pt/Pd, Ru/Pd and Pt/Ru ratios. The results are listed in Table 8 and visualised in Figure 22. Next to the chloritized samples, also sample 67‐1446 is not listed, because it does not show a correlation to the other samples. These other nine

PAGE 59 samples do show a correlation (including sample 59‐1050), which implicates that they have a common source of PGE. Only for Au no correlation existss.

Figure 21. CI‐normalised results of the PGE analysis. The CI values used for normalissation are that of Tagle & Berlin [2008]. The values for the upper continental crust (UCC) [Rudnick & Gaoo,, 2003] and IIIB irons [Goderis,, 2006] are added as a reference.

Table 8. Results of linear regression analysis on the PGE. These data are obtained with the Analysis ToolPak Regression from Microsoft Office 2010 Excel. Nine samples are used.

Y X Slope SE Intercept SE R2 Pd Ir 1.22 0.06 0.68 0.22 0.986 Rh Ir 0.46 0.05 0.04 0.17 0.938 Pt Ir 2.65 0.20 0.30 0.76 0.962 Ru Ir 2.31 0.19 0.22 0.71 0.957 Pd Rh 2.56 0.20 0.76 0.35 0.961 Pt Rh 5.72 0.15 0.21 0.26 0.996 Ru Rh 5.00 0.13 0.13 0.23 0.996 Pt Pd 2.17 0.13 ‐1.19 0.67 0.977 Ru Pd 1.90 0.12 ‐1.09 0.61 0.976 Pt Ru 1.14 0.03 0.08 0.22 0.997

PAGE 60 The Ir/Pd ratio is the best discriminator because of the difference in condensation temperatures. Its value is exactly that of the L ordinary chondrites (1.22 ± 0.14). However, all the other PGE ratios are not in agreement. They all have values that do not fall into the range of the chondrites.

Figure 22. Rh/Ir, Pd/Rh, Ru/Pd and Pt/Ru diagrams showing the correlation of niine melt lens samples and melt‐fragment breccia samples (the chloritized samples and sample 67‐1446 are excluded). The other PGE ratios show similar correlations.

4.2. ISOTOPE RATIOS

4.2.1. OS ISOTOPE SYSTEM

Table 9 lists the results of the Os isotope analysis. Except for sample 59‐1050, pure meteoritic values are obtained. When the 187Os/188Os ratio is plotted against the Os concentrations (Figure 23), all four samples fall onto the mixing line between the upper continental crust and meteorittes, defined by Tagle & Hecht [2006]. Sample 59‐ 1050 plots near the lower continental crust at a 0.01 wt.% meteorite contribution.

PAGE 61 The other samples fall near the primitive upper mantle with samples 59‐2771 and 59‐1050 at 0.5 wt.% and sample 59‐2800 between 1 and 2 wt.%.

Table 9. Results of the Os isotope analysis. 59‐951.5 59‐1050 59‐2771 59‐2800 187Os/188Os 0.1263 0.55251 0.1267 0.1255 2SD 0.0014 0.0067 0.0031 0.0014 Os (ppb) 2.244 0.052 2.976 7.404 2SD 0.26 0.85 0.69 0.44

Figure 23. Determination of the meteorite contribution by plotting the 187Os/1888 ratio against the Os concentration [Tagle & Hecht, 2006]. On thee left the mixing line between the upper continental crust (UCC) and CI meteorites is shown. Also the mid‐ocean ridge basalts (MORB), lower coontinental crust (LCC) and primitive upper mantle are plotted as a reference, together with samples from the Chicxulub and Morokweng impact structures. On the right the values for the Brent impact structurre samples are shown together with the UCC, LCC, PUM and CI meteorrites.

4.2.2. PB ISOTOPE SYSTEM

No conclusions can be drawn from the Pb isotope system alone.

Table 10. Results of the Pb isotope analysis. 59‐951.5 59‐1050 59‐2771 59‐2800 67‐888 67‐926 206Pb/204Pb 20.2300 19.4306 17.3920 18.2904 18.1453 19.9979 SE 0.0010 0.0008 0.0006 0.0011 0.0009 0.0008 207Pb/204Pb 15.6571 15.5626 15.4359 15.4879 15.4762 15.6074 SE 0.0008 0.0008 0.0006 0.0009 0.0008 0.0007 208Pb/204Pb 41.9700 40.7406 37.8185 38.8753 38.4699 41.3150 SE 0.0022 0.0022 0.0019 0.0025 0.0020 0.0020 207Pb/206Pb 0.77393 0.80090 0.88747 0.84673 0.85286 0.78041 SE 0.00001 0.00001 0.00001 0.00001 0.00001 0.00001 208Pb/206Pb 2.07466 2.09678 2.17447 2.12545 2.12014 2.06595 SE 0.00004 0.00005 0.00005 0.00005 0.00005 0.00004

PAGE 62 4.3. MULTI‐SIGNATURE APPROACH

When all siderophile data are combined, a CI‐normalised siderophiile pattern can be constructed to visualise there enrichment (Figure 24). All samplees seem to have a relative flat chondritic pattern, except for sample 59‐1050 which iis intermediate to the crust and the other samples, and sample 67‐1446 which shows a deviating and depleted signature.

Figure 24. CI‐normalised concentrations of all siderophile elements. The CI values used for normalisation are that of Tagle & Berlin [2008]. The values for the upper continental crust (UCC) are from Rudnick & Gao [2003], those for the IB and IIID irons from Tagle [2004].

A lineaar regression analysis is carried out by combining Ni, Co and Cr with Ir. Slopes and intercepts are determined for the Ni/Ir, Co/Ir and Cr/Ir ratios. The results are listed in Table 11. Only the melt lens samples have been used, because the melt‐ fragment breccias show no correlation. All three ratios fall outside the range of the chondrite meteorites, as can be seen in Appendix B.

PAGE 63 Table 11. Results of linear regression analysis on Ni, Co and Cr with Ir. These data are obtained with the Analysis ToolPak Regression from Microsoft Office 2010 Excel. Four samples are used.

Y X Slope (ppm) SE Intercept (ppm) SE R2 Ni Ir 32880 2 106.49 8.25 0.991 Co Ir 2025 1 7.43 0.2 0.999 Cr Ir 16551 1 23.98 2.32 0.997

The results from the Pb isotope systems can be plotted against Ir (and the other PGE) to see if a correlation appears that would suggest a common source of Pb. However, no correlation is found.

PAGE 64 5. DISCUSSION

This master dissertation has the objective to conduct a meteorite identification, as precisely as possible, for the Brent impact structure. If possible, implications are drawn from the results, regarding to the applicability of the different chemical signatures on impact en meteorite identification. Special attention is paid to the multi‐signature approaches. Twenty melt lens and melt‐fragment breccia samples are obtained from the breccia lens and analysed for major elements (n = 13), and the moderately (n = 20) and high (n = 13) siderophile elements.

5.1. ALTERATION

On arrival, three dark green coloured melt‐fragment breccia samples (59‐1379, 67‐ 888 and 67‐1169) were labelled as chloritized, based on thin section observations by the Earth Physics Branch of the Department of Energy, Mines and Resources, Ottawa, Canada. Because no target rock is analysed, these samples are compared with the target rock values obtained by Grieve [1978], which are assumed to be representative because obtained from lithic breccias, deep parautochthonous basement (≥ 911 m) and surface samples (n = 10). A single target rock component is proposed and identified as a mesoperthite gneiss of granodioritic composition, which is confirmed by Palme et al. [1981]. The samples show elevated values of Fe, Mg and the LOI, and depleted Si, Ca, K and Na contents (Figure 19), which is consistent with chloritization. While the Brent impact structure is already reported to be highly altered [Grieve, 1978; Grieve & Dence, 1978; and Dence, 2004], chloritization is not specifically mentioned before.

Also the other melt‐fragment breccias samples (n = 5) show a deviated pattern from the target rock in their major elements composition (Figure 19). Most of this signature is similar to the chloritized samples, albeit to a lower level. However, they strongly differ in K2O contents. While that of the chloritized samples is depleted with

0.07 ± 0.01 wt.%, the other melt‐fragment breccias show 9.06 ± 0.31 wt.% K2O, which is higher than the target rock value (4.92 ± 0.08 wt.%). These elevated K2O

PAGE 65 contents are already explained by Grieve [1978], who found values up to 14 wt.% which are proportional to the experienced levels of shock pressure (with a threshold of 20 GPa). This is probably caused by alkali exchange between feldspars and a saline aqueous solution that was present during post‐impact cooling [Lozej & Beales, 1975].

The major elements analysis of the melt lens samples (n = 4) resembles the target rock the most. They still show a depletion of Si, Ti and Na, and an enrichment of the other major elements, suggesting some alteration, but no spikes like the melt‐ fragment breccia samples (Figure 19). Therefore, the melt lens is assumed to be less altered than the melt‐fragment breccias, which can be expected for its coherent lithology. Grieve [1978] reports that the alteration in the melt lens occurs on the highest shocked target rock clasts. An increase in PGE with depth is noticed by Evans et al. [1993] in the melt lens samples, based on own data and data from Palme et al. [1981]. It is explained by a fractionation of the meteorite contribution during post‐ impact cooling and/or the hydrothermal alteration. While our PGE values for the melt lens samples (n = 4) are in agreement with this observation (Figure 21), we found another explanation, as will be seen below. Following Palme et al. [1981], hydrothermal alteration is also responsible for depleted Au concentrations, which is consistent with our data and seems likely due to the mobility of Au. The lack of correlation for the Pb isotope is probably also because of alteration, taken the mobility of Pb into account.

One melt‐fragment sample (67‐1446) shows deviating values between that of the melt lens and melt‐fragment breccia samples (Figure 19), probably reflecting an intermediate or different alteration.

We can conclude that the Brent impact structure is significantly altered. Probably an early feldspar alteration occurred, made possible by the presence of a saline aqueous solution during post‐impact cooling. This was followed by other alteration processes, including chloritization. The melt lens is the least altered, which can be important.

5.2. IMPACT IDENTIFICATION

A relative easy analysis for impact identification is that of the moderately siderophile elements Ni, Co and Cr. The analysed samples from the Brent impact structure show a significant enrichment of these elements relative to the target rock of Grieve

PAGE 66 [1978]. When the Ni/Cr, Co/Cr and Ni/Co ratios are plotted (Figure 20), the melt‐ fragment breccias (n = 8) show no correlation (probably due to alteration as suggested by the major elements), but the melt lens samples do (n = 5), implicating a common meteorite source. However, ultramafic alnoite and (ultra)mafic gneiss components are reported by Grieve [1978] and Palme et al. [1981], which could be responsible for elevated Ni, Co and Cr contents. While such contribution is discarded by Grieve [1978] and Palme et al. [1981], based on mixing calculations and V contents, we do not find the evidence conclusive because based only on a few samples and also no errors are provided concerning the V concentrations. Perhaps new target rock data should be compiled to explore this possibility of (ultra)mafic components. After all, based on the surrounding rocks, it is difficult to exclude a minor component from the large volume of target rock that contributes to an impact structure. Local enrichments can always occur. Because alnoite clasts are found in the melt‐fragment breccias by Grieve [1978] and Palme et al. [1981], it is always possible that small (ultra)mafic clasts are present in one or more of our samples and absence in all others, partly explaining the scatter of the melt‐fragment breccias. A more uniform contribution (as part of the melt) could be partly responsible for the depleted Si contents that all analysed samples show.

The more difficult to analyse PGE also show high concentrations, except for the chloritized samples (for which they are not detected or quantified) and sample 67‐ 1446, which has very low values, probably due to the alteration (sample 67‐1446 already showed deviating values for the major elements). No target rock is measured to compare with and also the target rock values of Palme et al. [1981] are not used because they are only maxima and obtained from a melt‐fragment breccia target rock clast, thus probably even not representative for the target rock due to alteration. However, all samples (n = 9) show correlating PGE concentrations that are significantly higher than any known crustal values (except for the altered Au) (Figure 21), implicating a common meteoritic source. This means that the PGE concentrations of the included melt‐fragment breccias (n = 5) did not change by alteration (which is consistent with their relative immobility). They do not exclude the presence of (ultra)mafic components because they even exceed the concentrations that are typical for most upper mantle components.

Like the PGE, also the Ni, Co and Cr concentrations increase with depth in the melt lens (Table 6 and Figure 20). An explanation by alteration or an increasing (ultra)mafic component seems unlikely. Higher values would suggest a higher meteorite contribution, but from what is known of impact mechanisms, no process can account for this depth relation. This does not makes it a better option than the

PAGE 67 suggestion of Evans et al. [1993] of fractionation during post‐impact cooling. However, Os isotopes can bring solution.

While it is also relatively difficult to analyse, the Os isotope system is highly straightforward when it comes to impact identification. It is known to be very sensitive in reflecting the exact meteorite contribution. This is done by plotting the 187Os/188Os ratios against the Os concentrations and comparing them with the mixing line from Tagle & Hecht [2006] between the upper continental crust and meteorites (Figure 23). Values are obtained from 0.01 wt.% for sample 59‐1050 (which already showed a PGE signature between the other melt‐fragment breccias and the upper continental crust), 0.5 wt.% for samples 59‐951.5 and 59‐2771, and between 1 and 2 wt.% for sample 59‐2800. This rejects the fractionation option. While it can explain the higher Os concentrations, it cannot account for the difference in the 187Os/188Os ratios.

As a conclusion, a meteorite identification for the Brent impact structure is certain because of a conclusive multi‐signature meteorite contribution detection by the moderately and highly siderophile elements, together with the Os isotope system. While the latter is probably sufficient on its own, the moderately and highly siderophile elements, together with the major elements, can be necessary to characterize the target rock and used samples when this information is not yet available.

5.3. METEORITE IDENTIFICATION

Based on the impact identification, a differentiated achondrite can be excluded as impact meteorite, because this would not result into such distinctive meteorite contributions. No data exists on primitive achondrites in the used databases, probably because these are very rare. The highly siderophile elements show a relative flat CI‐normalised pattern (Figure 21), which is confirmed when including the Os values from the Os isotope system analysis and moderately siderophile elements (Figure 24). Based on the Ni/Cr (2.93 ± 0.16) and Co/Cr (0.10 ± 0.01) ratios of the melt lens samples (n = 5) a LL ordinary chondrite (2.64 ± 0.21 Ni/Cr and 0.13 ± 0.02 Co/Cr) is suggested. This is in agreement with the previous meteorite identification of Grieve [1978] as a L (2.76 ± 0.36 Ni/Cr) and Palme et al. [1981] as a L or LL (2.56 ± 0.26 Ni/Cr). Both fall closer to the LL if compared with recent databases (see Appendix B).

PAGE 68

Figure 25. CI‐normalised PGE concentratioons of the melt lens samples. The CI values used for normalisation are those of Tagle & Berlin [2008], those for the upper continental crust (UCC) from Rudnick & Gao [2003]. The values for the IIIC, IA and IB iron, and LL ordinary chondrite meteorites are from Tagle [2004] and Tagle et al. [2009]. The melt lens samples are magnified by a ffactor of 500.

However, the Ni/Co ratio (29.85 ± 3.00) does not fall into the narrow chondrite range (19.34 ± 3.15 to 22.86 ± 2.83). This can exclude the use of the Ni/Cr and Co/Cr ratios, if Cr is considered not representative for the impact meteorite. Also Ni or Co can be not representative, which would imply their ratio with Cr would point to the LL by coincidence. However, the Ni/Ir, Co/Ir and Cr/Ir all show coorrelation for the melts lens samples, meaning they all share a common source and are representative for the impact meteorite. The best discriminating PGE ratio Pd/Ir (1.22 ± 0.14) matches exactly the L if the meteorite is chondritic, but the other PGE ratios do not. Together with the Ni/Co ratio, they are more indicative for an iron n meteorite, which is a likely possibility because the Ni/Cr, Co/Cr and Pd/Ir ratios, nextt to that of the LL, also fall in the range of the iron meteorites. However, the flat CI‐normalised pattern of the siderophile elements does not agree with an iron meteorite. The preclusion of a L or LL based on PGE concentrations, is already reported by Evans et al. [1993], who propose an iron or CI chondrite based on a 1.95 Pt/Ir ratio. Their suggestion of

PAGE 69 fractionation or alteration would explain the flat CI‐normalised pattern, but we already excluded these possibilities.

Table 12. Comparison with the obtained siderophile ratios for the Brent impact structure (X/Y), and the IA and IIIC non‐magmatic iron meteorites. The Ni/Cr and Ni/Ir values are from Tagle [2004]. Y X X/Y IA IIIC Pd Ir 1.22 1.52 3.02 Rh Ir 0.46 0.60 1.17 Pt Ir 2.65 2.58 3.25 Ru Ir 2.31 2.05 3.03 Pd Rh 2.56 2.57 2.59 Pt Rh 5.72 4.36 2.79 Ru Rh 5.00 3.47 2.60 Pt Pd 2.17 1.70 1.08 Ru Pd 1.90 1.35 1.00 Pt Ru 1.14 1.26 1.07 Ni Cr 2.93 3.07 2.73 Ni Ir 32880 31000 119000

When the CI‐normalised PGE patterns are plotted on a linear scale instead of a logarithmic, and magnified by a factor of 500, a spiked pattern appears (Figure 25). Because of the extreme low target rock PGE contents, the resulting pattern is almost not affected by the target rock. This is shown in Figure 25 for sample 59‐2800. This means the pattern is highly characteristic for the impact meteorite. As can be seen, the flat pattern of the LL does not reflect at all that of the melt lens samples. Only the two non‐magmatic iron meteorite groups IA and IIIC can fit the PGE patterns of the melt samples, while the closely related IB do not. When these two groups are compared to the obtained meteorite discriminating data, almost all ratios do significantly better fit the IA (Table 12). We therefore conclude that the Brent impact structure is produced by the impact of a IA non‐magmatic iron meteorite.

5.4. NON‐MAGMATIC IRON METEORITES

[Goderis, 2009 and references therein]

The non‐magmatic iron (NMI) meteorites are divided into the IA, IB, IIIC, IIID, and IIE groups. They are, as the stony‐irons, composed of an iron and a silicate phase, the latter occurring essentially as large or small inclusions dispersed in the metal phase. Silicate inclusions occur in NMI, which can vary significantly between different meteorites, as well as between different sections of the same meteorite. They

PAGE 70 underwent a complex and poorly understood formation history (including metamorphism, partial melting, incomplete differentiation and crystal segregation) that is reflected in the high variability of the amount of silicate inclusions. After a catastrophic impact break‐up in the asteroid belt, gravitational reassembling of the debris probably produced a heterogeneous mixture of 5 different components: metal, a sulphide‐rich phase, a chondritic‐silicate phase, partial melt, and residues. All NMI show mixed fractions of these components.

At present, four other impact structures are identified as being produced by a NMI: Lockne, Rochechouart, Sääksjärvi and Gardnos.

PAGE 71 6. CONCLUSION

Because all analysis in this master dissertation, as well from this study as from literature, were conducted on melt‐bearing impact rocks from a breccia lens, all following conclusions apply only to such impact rocks and consequently have no implications for impact and meteorite identification with ejecta rocks.

6.1. THE BRENT IMPACT STRUCTURE

Characterization of the alteration present at the Brent impact structure, revealed at least the presence of a chloritization, Au‐depletion and K‐enrichment process in the melt‐fragment breccias. While this totally scattered the mobile Ni, Co and Cr, the immobile PGE are not affected. Also the melt lens shows alteration, albeit to a lesser degree. Ni, Co and Cr still have a strong correlation here. The use of the Os isotope system delivered a meteorite contribution up to 1 a 2 wt.% for the deepest melt lens sample. Together with the moderately and siderophile elements, it rejects the possibility of a fractionation during post‐impact cooling, as suggested by Evans et al. [1993], and implicates an impact identification.

We also reject the assumption of Grieve [1978] and Palme et al. [1981] that no (ultra)mafic components are present. However, if present, they can only be responsible for (a part of) the Ni, Co and Cr scatter in the melt‐fragment breccias. The melt lens is for certain not affected. A more regional target rock study should be conducted to examine the possibility of (ultra)mafic components and their effect on the melt‐fragment breccias.

Based on a multi‐signature approach by combining the moderately and highly siderophile elements, a precise meteorite classification into the IA non‐magmatic irons is possible. While the Ni/Cr, Co/Cr and Pd/Ir ratios point to a LL or L ordinary chondrite, the Ni/Co, other PGE and combined siderophile ratios do not. Based on a linear and magnified PGE pattern that is assumed to be representative for the impact meteorite, the IA or IIIC non‐magmatic irons are the only possibility. When all siderophile ratios are taken into account, the IA is by far the best fitting group.

PAGE 72 6.2. IMPLICATIONS FOR IDENTIFICATION

The Brent impact structure is the ideal case study for the effects of alteration on impact and meteorite identification, especially for melt‐fragment breccias which show consistent PGE values, but scatter for the moderate siderophile elements. Due to its resistance against alteration and high meteorite contribution, the melt lens provides the best samples for impact and meteorite identification.

For meteorite identification, the implications of the Brent impact structure are straight forwarded, no correct meteorite identification would have been possible without the used multi‐signature approach. While Ni, Co and Cr are relatively easy to analyse, the chondritic values of their ratios are shared by multiple iron groups. On the other hand, they make it possible to discriminate further when a chondritic or iron meteorite is determined by the PGE.

The Os isotope system seems sufficient on its own for impact identification. However, its analysis is relative difficult compared to the moderately siderophile elements. When the moderately and highly siderophile, and major elements are added, this is the best way to describe all interfering processes like alteration, (ultra)mafic components and fractionation.

Because the Pb isotope system is scattered due to alteration, other impact structures than Brent should be examined for Pb isotopes to develop its use for impact and meteorite identification. In addition, also the common Cr, and proposed W, Nd and Sr isotope systems could be analysed and compared with the now well established impact and meteorite identification data of the Brent impact structure.

A general conclusion for impact and meteorite identification is: the more data, the better the results. A multi‐signature approach is always recommended.

PAGE 73 V. REFERENCES

Alvarez, L. W., Alvarez, W., Asaro, F. & Michel, H. V. Science 41, 607‐631. [1980]. Extraterrestrial Cause for the Bottke, W. F. & Morbidelli, A. [2006]. The asteroid Cretaceous‐Tertiary Extinction. Science 208, and comet impact flux in the terrestrial planet 1095‐1108. region: a brief history of the last 4.6 Gy. Ames, D. E. & Farrow, C. E. G. [2007]. Metallogeny Planetary Chronology Workshop 2006, Lunar of the Sudbury mining camp, Ontario. In: and Planetary Institute, 2 pp. Goodfellow, W. D. (ed.), Mineral Deposits of Brownlee, D. E. [2007]. Comets. In: Davis, A. M. Canada: A Synthesis of Major Deposit‐Types, (ed.), Treatise on Geochemistry, vol. 1: District Metallogeny, the Evolution of Geological Meteorites, Comets and Planets, Online edition. Provinces, and Exploration Methods. Geological Elsevier, 29 pp. Association of Canada, Mineral Deposits Bucher, W. [1963]. structures Division, Special Publication 5, 329‐350. caused from without or from within the Earth? Ames, D. E., Gibson, H. L. & Watkinson, D. H. [2000]. (“Astroblemes” or “Geoblemes”?). American Controls on major impact‐induced Journal of Science 261, 597‐649. hydrothermal system, Sudbury structure, Buchwald, V. F. [1975] Handbook of Iron Canada. Lunar and Planetary Science XXXI, Meteorites. University of California Press, 1418 1873. pp. Anders, E., Ganapathy, R., Kriihenbiihl U. & Morgan Chambers, J. E. [2007]. Planet Formation. In: Davis, J. W. [1973]. Meteoritic material on the Moon. A.M. (ed.), Treatise on Geochemistry, vol. 1: The Moon 8, 3‐24. Meteorites, Comets and Planets, Online edition. Asher, D. J., Bailey, M., Emel’yanenko, V. & Napier, Elsevier, 17 pp. B. [2005]. Earth in the cosmic gallery. Chyba, C. F., Owen, T.C. & Ip, W.H. [1994]. Impact Observatory 125, 319‐322. delivery of volatiles and organic molecules to Asphaug, E. [2009]. Growth and Evolution of Earth. In: Gehrels, T. (ed.), Hazards Due to Asteroids. Annual Review of Earth and Comets and Asteroids, 9‐58. University of Planetary Sciences 37, 413‐448. Arizona Press, 1300 pp. Baldwin, R. B. [1949]. The Face of the Moon. Cohen, D. [2008]. The day the sky exploded. New University of Chicago Press, 239 pp. Scientist 28, 38‐41. Beals, C. S. [1958]. meteorite craters. Currie, K. L. [1971]. A study of potash fenitization Scientific American 199, 32‐39. around the , Ontario, a Beals, C. S., Ferguson, G. M. & Landau, A. [1956]. A alkaline complex. Canadian Journal of Earth search for analogies between lunar and Sciences 8, 481‐497. terrestrial topography on photographs of the Currie, K. L. & Shafiqullah, M. [1967]. Carbonatite Canadian shield: Part I and Part II. Journal of the and alkaline igneous rocks in the Brent Crater, Royal Astronomical Society of Canada 50, 203‐ Ontario. Nature 215, 725‐726. 211 and 250‐261. Delsanti, A. & Jewitt, D. [2007]. The Solar System Bennet, J., Donahue, M., Schneider, N. & Voit, M. Beyond The Planets. Institute for Astronomy, 27 [2008]. The Cosmic Perspective, 5th edition. pp. Pearson Addison‐Wesley, 864 pp. Dence, M. R. [1971]. Impact melts. Journal of BIPM [2006]. The International System of Units Geophysical Research 76, 5552‐5565. (SI), 8th edition. International Bureau of Dence, M. R. [1968]. Shock zoning at Canadian Weights and Measures, 180 pp. craters: Petrography and structural Bjork, R. L. [1963]. Review of Physical Processes in implications. In: French, B. M. & Short, N. M. Hypervelocity Impact and Penetration. RAND (eds.), Shock Metamorphism of Natural Corporation, 51 pp. Materials, 339‐362. Mono Book Corporation, Bland, P. A. [2005]. The impact rate on Earth. 644 pp. Philosophical Transactions of the Royal Society Dence, M. R. [1972] The nature and significance of A 363, 2793‐2810. terrestrial impact structures. Proceedings of the Bland, P. A. & Artemieva, N. A. [2006]. The rate of 24th International Geological Congress 15, 77– small impacts on Earth. Meteoritics & Planetary 89.

PAGE X Dence, M. R. [1973]. Dimensional analysis of impact Terrestrial Meteorite Impact Structures. Lunar structures. Meteoritics 8, 343‐344. and Planetary Institute, 120 pp. Dence, M. R. [2004]. Structural evidence from French, B. M. & Koeberl, C. [2010]. The convincing shock metamorphism in simple and complex identification of terrestrial meteorite impact impact craters: Linking observations to theory. structures: What works, what doesn’t and why. Meteoritics & Planetary Science 39, 267‐286. Earth Science Reviews 98, 123‐170. Dence, M. R. & Guy‐Bray, J.V. [1972]. Some Ganapathy, R., Keays R. R., Laul J. C. & Anders, E. astroblemes, craters and cryptovo1canic [1970a]. Trace elements in Apollo 11 lunar structures in Ontario and Quebec. Proceedings rocks: implications for meteorite influx and of the International Geological Congress 24, 61 origin of moon. Geochimica et Cosmochimica pp. Acta Supplement 1, 1117‐1142. Dence, M. R., Grieve, R. A. F. & Robertson, P. B. Ganapathy, R., Keays R. R. & Anders E. [1970b]. [1976]. Terrestrial impact structures: Principal Apollo 12 lunar samples: Trace element characteristics and energy considerations. In: analysis of a core and the uniformity of the Roddy, D. J., Pepin, R. O. & Merrill, R. B. (eds.), regolith. Science 170, 533‐535. Impact and Explosion Cratering, 247‐276. Ganapathy, R., Laul J. C., Morgan, J. W. & Anders E. Pergamon Press, 1315 pp. [1972]. Moon: Possible nature of the body that Dietz, R. S. [1963]. Cryptoexplosion structures: A produced the Imbrian basin, from the discussion. American Journal of Science 261, composition of Apollo 14 samples. Science 175, 650‐664. 55‐59. Drake, M. J. [2001]. The eucrite/Vesta story. Ganapathy, R., Morgan, J. W., Krahenbiihl U. & Meteoritics & Planetary Science 36, 501‐513. Anders, E. [1973]. Ancient meteoritic Dressler, B. O. & Sharpton, V. L. [1997]. Breccia components in lunar highland rocks: Clues from formation at a complex impact crater: Slate trace elements in Apollo 15 and 16 samples. Islands, , Ontario, Canada. Proceedings of the Lunar Science Conference 4, Tectonophysics 275, 285‐311. 1239‐1261. Evans, N. J., Gregoire, D .C., Grieve, R. A. F., Ganapathy, R., Morgan, J. W., Higuchi, H., Anders, E. Goodfellow, W. D. & Veizer, J. [1993]. Use Of & Anderson, A. T. [1974]. Meteoritic and volatile Platinum‐Group Elements For Impactor elements in Apollo 16 rocks and in separated Identification ‐ Terrestrial Impact Craters And phases from 14306. Proceedings of the Lunar Cretaceous‐Tertiary Boundary. Geochimica et Science Conference 5, 1659‐1683. Cosmochimica Acta 57, 3737‐3748. Garvin, J. B. & Grieve, R. A. F. [1982]. An analytical Faure, G. [1986] Principles of Isotope Geology, 2nd model for terrestrial simple craters: Brent and edition. Wiley, 608 pp. Meteor. Lunar and Planetary Science 13, 251‐ Fischer‐Gödde, M., Becker, H. & Wombacher, F. 252. [2010]. Rhodium, gold and other highly Gault, D. E. & Heitowit, E. D. [1963]. The Partition of siderophile element abundances in chondritic Energy for Hypervelocity Impact Craters meteorites. Geochimica et Cosmochimica Acta Formed in Rock. Proceedings of the 74, 356‐379. Hypervelocity Impact Symposium 6, 419‐456. Fiske, P. S., Hargraves, R. B., Onstott, T. C., Koeberl, Gault, D. E., Quaide, W. L. & Oberbeck, V. R. [1968]. C. & Hougen, S. B. [1994]. Pseudotachylites of Impact Cratering Mechanics and Structures. In: the Beaverhead impact structure: Geochemical, French, B. M. & Short, N. M. (eds.), Shock geochronological, petrographic, and field Metamorphism of Natural Materials, 87‐99. investigations. In: Dressler, B. O, Grieve, R. A. F., Mono Book Corporation, 644 pp. & Sharpton, V. L. (eds.), Large Meteorite Gifford, A. C. [1924]. The Mountains of the Moon. Impacts and Planetary Evolution, 163‐176. New Zealand Journal of Science and Technology Geological Society of America, 348 pp. 7, 129‐142. French, B. M. [1968]. Shock metamorphism as a Gilbert, G. K. [1893]. The Moon’s Face: A Study of geological process. In: French, B.M. & Short, the Origin of its Features. Bulletin of the N.M. (eds.), Shock Metamorphism of Natural Philosophical Society of Washington 12, 241‐ Materials, 1‐17. Mono Book Corporation, 644 292. pp. Glass, B. P., Lu, S. & Leavens, P. B. [2002]. Reidite: French, B. M. [1990]. 25 years of the impact‐ an impact‐produced high‐pressure polymorph volcanic controversy: Is there anything new of zircon found in marine sediments. American under the sun or inside the Earth? Eos, Mineralogist 87, 562‐565. Transactions of the American Geophysical Goderis, S. [2006]. Geochemie en distributie van de Union 71, 411‐414. platinagroepelementen in de impactstructuren French, B. M. [1998]. : A van Bosumtwi (Ghana, Pleistoceen) en Gardnos Handbook of Shock‐Metamorphic Effects in (Noorwegen, grens Proterozoïcum‐

PAGE XI Paleozoïcum). Universiteit Gent, 103 pp. 420. Goderis, S., Tagle, R., Schmitt, R. T., Erzinger, J. & Grieve, R. A. F. & Shoemaker, E.M. [1994]. The Claeys, P. [2007] Platinum group elements record of past impacts on Earth. In: Gehrels, T. provide no indication of a meteoritic (ed.), Hazards Due to Comets and Asteroids, component in ICDP cores from the Bosumtwi 417‐462. University of Arizona Press, 1300 pp. crater, Ghana. Meteoritics and Planetary Science Grieve, R. A. F., Dence, M. R. & Robertson, P. B. 42, 731‐741. [1977]. Cratering processes: As interpreted Goderis, S., Kalleson, E., Tagle, R., Dypvik, H., from the occurrence of impact melts. In: Roddy, Schmitt, R., Erzinger, J. & Claeys, P. [2009]. A D. J., Pepin, R. O. & Merrill, R. B. (eds.), Impact non‐magmatic iron projectile for the Gardnos and Explosion Cratering, 791‐814. Pergamon, impact event. Chemical Geology 258, 145‐156. 1315 pp. Gounelle, M., Morbidelli, A., Bland, P. A., Spurný, P., Grunert, N. [2000]. Namibia Fascination of Geology. Young, E. D. & Sephton, M. [2008]. Meteorites Klaus Hess Publishers, 17 pp. from the Outer Solar System? In: Barucci, M. A., Haack, H. & McCoy, T. J. [2005]. Iron and Stony‐iron Boehnhardt, H., Cruikshank, D. P. & Morbidelli, Meteorites. In: Davis, A. M. (ed.), Treatise on A. (eds.), The Solar System beyond Neptune, Geochemistry, vol. 1: Meteorites, Comets and 525‐541. Arizona University Press, 592 pp. Planets, 325‐346. Elsevier, 737 pp. Grahn, Y. & Ormö, J. [1995]. Microfossil dating of Halliday, A. N. [2004]. Mixing, volatile loss and the Brent meteorite crater, southeast Ontario, compositional change during impact‐driven Canada. Revue de Micropaléontologie 38, 131‐ accretion of the Earth. Nature 427, 505‐509. 137. Hartung, J. B., Dence, M. R. & Adams, J. A. [1971]. Grieve, R. A. F. [1978]. The melt rocks at Brent Potassium‐argon dating of shock crater, Ontario, Canada. Proceedings of the metamorphosed rocks from the Brent impact Lunar and Planetary Science Conference 9, crater, Ontario, Canada. Journal of Geophysical 2579‐2608. Research 76, 5437‐5448. Grieve, R. A. F. [1984]. The Impact Cratering Rate in Hawke, P. J. [2004]. The geophysical signatures and Recent Time. Journal of Geophysical Research exploration potential of 's meteorite 89, 403‐408. impact structures. University of Western Grieve, R. A. F. [1987]. Terrestrial impact Australia, 314 pp. structures. Annual Reviews of Earth and Higuchi, H. & Morgan, J. W. [1975]. Ancient Planetary Science 15, 245‐270. meteoritic component in Apollo 17 boulders. Grieve, R. A. F. [1991]. Terrestrial impact: The Proceedings of the Lunar Science Conference 6, record in the rocks. Meteoritics 26, 175‐194. 1625‐1651. Grieve, R. A. F. [2005]. Economic natural resource Hildebrand, A. R., Penfield, G. T., Kring, D. A., deposits at terrestrial impact structures. In: Pilkington, M., Camargo, A., Jacobsen, S. B. & McDonald, I., Boyce, A.J., Butler, I.B., Herrington, Boynton, W. V. [1991] : A R.J. & Polya, D.A. (eds.) Mineral deposits and Possible Cretaceous/Tertiary Boundary Impact earth evolution, Geological Society of London Crater on the Yucatan Peninsula, Mexico. Special Publication 248, 1‐29. Geological Society Geology 19, 867‐871. of London, 263 pp. Hockey, T.A. [1994]. The Shoemaker‐Levy 9 Spots Grieve, R. A. F. & Cintala, M.J. [1981]. Brent Crater, on Jupiter: Their Place in History. Earth, Moon Ontario: Observation and theory. Lunar and and Planets 66, 1‐9. Planetary Science 12, 362‐364. Horneck, G. & Rettberg, P. [2007]. Complete Course Grieve, R. A. F. & Cintala, M.J. [1982]. A method for in Astrobiology. Wiley, 434 pp. estimating the initial impact conditions of Howard, E. C. [1802]. Experiments and terrestrial cratering events, exemplified by its observations on certain stony substances, application to Brent crater, Ontario. which at various times are said to have fallen on Proceedings of the Lunar and Planetary Science the earth; also on various kinds of native iron. Conference. 12B, 1607‐1621. Philosophical Transactions 92, 168‐212. Grieve, R. A. F. & Dence, M.R. [1978]. Principle Hughes, D. W. [2000]. A new approach to the characteristics of the at Brent crater, calculation of the cratering rate of the Earth Ontario, Canada. Lunar and Planetary Science 9, over the last 125±20 Myr. Monthly Notices of 416‐418. the Royal Astronomical Society 317, 429‐437. Grieve, R. A. F. & Masaitis, V.L. [1994]. The IAU [1961]. General Assembly XI, Commission 22. economic potential of terrestrial impact craters. International Astronomical Union, 38 pp. International Geological Review 43, 105‐151. IAU [2006]. Definition of a Planet in the Solar Grieve, R. A. F. & Pilkington, M. [1996]. The System, Resolution B5. International signature of terrestrial impacts. Journal of Astronomical Union, 2 pp. Australian Geology and Geophysics 16, 399‐ Innes, M. J. S. & Beals, C. S. [1961]. Profile of the

PAGE XII fossil crater at Brent, Ontario. Journal of the Geological Society of 118, 193‐206. Royal Astronomical Society of Canada 55, 258. Lindström, M., Shuvalov, V. & Ivanov, B. [2005]. IUPAC [2007]. Quantities, Units and Symbols in as a result of marine‐target Physical Chemistry ‐ the IUPAC Green Book, 3th oblique impact. Planetary and Space Science 53, edition. International Union of Pure and Applied 803‐815. Chemistry, 248 pp. Lodders, K. [2003]. Solar system abundances and ISO [2004]. Data elements and interchange formats condensation temperatures of the elements: – Information interchange – Representation of The Astrophysical Journal 591, 1220‐1247. dates and times, 3th edition. International Lorand, J., Luguet, A. & Alard, O. [2008]. Platinum‐ Organization for Standardization, 40 pp. group elements: a new set of key tracers for the Ives, H. E. [1919]. Some Large‐Scale Experiments Earth’s interior. Elements 4, 247‐252. Imitating the Craters of the Moon. Astrophysical Lozej, G. P. & Beales, F. W. [1975]. The Journal 50, 245‐250. unmetamorphosed sedimentary fill of the Brent Jackson, J. C., Horton Jr., J. W., Chou, I. M. & Belkin, meteorite crater, south‐eastern Ontario. H. E. [2006]. A shock‐induced polymorph of Canadian Journal of Earth Sciences 12, 606‐628. anatase and rutile from the Chesapeake Bay Masaitis, V. L. [1989]. The economic geology of impact structure, Virginia, U.S.A. American impact craters. International Geology Review Mineralogist 91, 604‐608. 31, 922‐933. Koeberl, C. [1998]. Identification of meteoritic McDonald, I. [2002] Clearwater East impact component in impactites. In: Grady, M. M., structure: A re‐interpretation of the projectile Hutchinson, R., McCall, G. J. H. & Rothery, R. A. type using new platinum‐group element data. (eds.), Meteorites: Flux with Time and Impact Meteoritics & Planetary Science 37, 459‐464. Effects, 133‐153. The Geological Society Special McDonald, I., Andreoli, M. A. G., Hart, R.J. & Publications 140, 278 pp. Tredoux, M. [2001]. Platinum‐group elements Koeberl, C. [2001]. Craters on the moon from in the Morokweng impact structure, South Galileo to Wegener: A short history of the Africa: Evidence for the impact of a large impact hypothesis, and implications for the ordinary chondrite projectile at the Jurassic‐ study of terrestrial impact craters. Earth, Moon Cretaceous boundary. Geochimica et and Planets 85‐86, 209‐224. Cosmochimica Acta 65, 299‐309. Koeberl, C., [2007]. The geochemistry and McDonough, W. F. & Sun, S. [1995]. The cosmochemistry of impacts. In: Davis, A. M. composition of the Earth. Chemical Geology (ed.), Treatise on Geochemistry, vol. 1: 120, 223‐254. Meteorites, Comets and Planets, Online edition. McKeegan, K. D. & Davis, A. M. [2005]. Early Solar Elsevier, 52 pp. System Chronology. In: Davis, A.M. (ed.), Koeberl, C. & Shirey S. B. [1993]. Detection of Treatise on Geochemistry, vol. 1: Meteorites, meteoritic component in Ivory Coast tektites Comets and Planets, 431‐460. Elsevier, 737 pp. with Rhenium‐Osmium isotopes, Science, 261, McSween, H. Y. Jr. [1999]. Meteorites and Their 595‐598. Parent Planets, 2nd edition. Cambridge Koeberl, C., Shukolyukov, A. & Lugmair, G.W. University Press, 310 pp. [2007]. Chromium isotopic studies of terrestrial Melosh H. J. [1989]. Impact Cratering: A Geologic impact craters: Identification of meteoritic Process. Oxford University Press, 245 pp. components at Bosumtwi, Clearwater East, Melosh, H. J. & Ivanov, B. A. [1999]. Impact crater Lappajärvi, and Rochechouart: Earth and collapse. Annual Review of Earth and Planetary Planetary Science Letters 256, 534‐546. Sciences 27, 385‐415. Kortenkamp, S.J. & Dermott, S.F. [1998]. Accretion Millman, P. M., Liberty, B. A., Clark, J. F., Willmore, of Interplanetary Dust Particles by the Earth. P. L. & Innes, M .J. S [1960]. The Brent crater. Icarus 135, 469‐495. Publications of the Dominion Observatory 24, 1‐ Krot, A. N., Keil, K. & Scott, E. R. D. [2007]. 43. Classification of Meteorites. In: Davis, A. M. Mittlefehldt, D. W. [2005]. Achondrites. In: Davis, A. (ed.), Treatise on Geochemistry, vol. 1: M. (ed.), Treatise on Geochemistry, vol. 1: Meteorites, Comets and Planets, Online edition. Meteorites, Comets and Planets, 291‐324. Elsevier, 52 pp. Elsevier, 737 pp. Levison, H. F., Dones, L. & Duncan, M. J. [2001]. The Mittlefehldt, D. W., McCoy, T., Goodrich, C. A. & origin of Halley‐type comets: Probing the inner Kracher, A. [1998]. Nonchondritic meteorites Oort Cloud. The Astronomical Journal 121, from the asteroidal bodies. In: Papike, J. J. (ed.), 2253‐2267. Planetary materials, 4.1‐4.195. Mineralogical Lindström, M., Sturkell, E. F. F., Törnberg, R. & Society of America, 1059 pp. Ormö, J. [1996]. The marine impact crater at Morbidelli, A. [2008]. Origin and dynamical Lockne, central Sweden. Journal of the evolution of comets and their reservoirs.

PAGE XIII Observatoire de la Côte d’Azur, 86 pp. Haughton impact structure, artic Canada: Morbidelli, A., Bottke, W. F., Froeschlé, C. & Michel, Generation of a transient, warm, wet oasis. P. [2002]. Origin and Evolution of Near‐Earth Meteoritics and Planetary Science 36, 731‐745. Objects. In: Bottke, W. F., Paolicchi, P., Binzel, R. Palme, H. [1980]. The meteoritic contamination of P. & Cellino, A. (eds.), Asteroids III, 409‐422. terrestrial and lunar impact melts and the University of Arizona Press, 785 pp. problem of indigenous siderophiles in the lunar Morgan, J. W., Laul, J. C., Kriihenbiihl, U., Ganapathy highland. Proceedings of the Lunar and R. & Anders, E. [1972]. Major impacts on the Planetary Science Conference 11, p481‐506. Moon: Characterization from trace elements in Palme, H., Janssens, M. J., Takahashi, H., Anders, E. Apollo 12 and 14 samples. Proceedings of the & Hertogen, J. [1978]. Meteoritic material at five Lunar Science Conference 3, 1375‐1395. large impact craters. Geochimica et Morgan, J. W., Ganapathy, R., Higuchi, H. & Anders Cosmochimica Acta 42, 313‐323. E. [1974a]. Proceedings of the Soviet‐American Palme, H., Grieve, R. A. F. & Wolf, R. [1981]. Conference on Cosmochemistry of the Moon Identification of the projectile at the Brent and Planets. The Lunar Science Institute, 764 crater and further considerations of projectile pp. types at terrestrial craters. Geochimica et Morgan, J. W., Ganapathy, R., Higuchi, H., Cosmochimica Acta 45, 2417‐2424. Krahenbiihl, U. & Anders, E. [1974b]. Lunar PASSC [2010]. . Planetary basins: Tentative characterization of projectiles, and Space Science Centre. from meteoritic elements in Apollo 17 boulders. http://www.passc.net/EarthImpactDatabase/ Proceedings of the Lunar Science Conference 5, Peucker‐Ehrenbrink, B. & Jahn, B. M. [2001]. 1703‐1737. Rhenium‐osmium isotope systematics and Morgan, J. W., Higuchi, H., Ganapathy, R. & Anders, platinum group element concentrations: Loess E. [1975]. Meteoritic material in four terrestrial and the upper continental crust. Geochemistry meteorite craters. Proceedings of the Lunar Geophysics Geosystems 4, 8911. Science Conference 6, 1609‐1623. Phillips, G. N. & Law, J. D. M. [2000]. Witwatersrand Morrison, D., Chapman, C. R. & Slovic, P. [1994]. Goldfields: Geology, Genesis and Exploration. The impact hazard. In: Gehrels, T. (ed.), Hazards Society of Economic Geology Reviews 13, 439‐ Due to Comets and Asteroids, 59‐91. University 500. of Arizona Press, 1300 pp. Plessen, H. G., Erzinger, J. [1998]. Determination of Morrison, D., Chapman, C. R., Steel, D. & Binzel, R. P. the platinum‐group elements and gold in [2004]. Impacts and the Public: Communicating twenty rock reference material by inductively the Nature of the Impact Hazard. In: Belton, M. J. coupled plasma‐mass spectrometry (ICPMS) S., Morgan, T. H., Samarasinha, N. H. & Yeomans, after pre‐concentration by nickel fire assay. D. K. (eds.), Mitigation of Hazardous Comets and Geostandards Newsletter 22, 187‐194. Asteroids, 353‐390. Cambridge University Proctor, R. A. [1873]. The Moon: Her Motions, Press, 436 pp. Aspect, Scenery, and Physical Condition. Alfred MPC [2010]. List of Neptune Trojans. Minor Planet Brothers, 314 pp. Center. WWW (2010‐11‐16): Reimold, W. U. & Koeberl, C. [2008]. Catastrophes, http://www.minorplanetcenter.org/iau/lists/N Extinctions and Evolution: 50 Years of Impact eptuneTrojans/ Cratering Studies. Golden Jubilee Memoir of the MS [2010]. Meteoritical Bulletin Database, Geological Society of India 66, 69‐110. Meteoritical Society. WWW (2010‐10‐13): Robertson P. B. & Grieve R. A. F. [1977]. Shock http://tin.er.usgs.gov/meteor attenuation at terrestrial impact structures. In: Nicolaysen, L. O. & Reimold, W. U. [1990]. Roddy, D. L., Pepin, R. O. & Merrill, R. B. (eds.), Cryptoexplosions and catastrophes in the Impact and Explosion Cratering, 687‐706. geological record, with a special focus on the Pergamon, 1315 pp. Vredefort structure. Tectonophysics 171, 1‐422. Rudnick, R. L. & Gao, S. [2003]. Composition of the O’Keefe, J. D. & Ahrens, T. J. [1993]. Planetary continental crust. In: Holland, H. D. (ed.), Cratering Mechanics. Journal of Geophysical Treatise on Geochemistry, vol. 3: The Crust, 1‐ Research 98, 17.011‐17.028. 64. Elsevier‐Pergamon, 702 pp. Öpik, E. J. [1916]. Remarque sur le théorie Ryan, D. [2003]. WWW (2010‐10‐13): météorique des cirques lunaires. Bulletin de la http://www.physics.mcgill.ca/~dominic/updat Société Russe des Amis de l’Etude de l’Univers e03/Flying/ 3.21, 125‐134. Schmieder, M. & Buchner, E. [2010]. Possible iron Osinski, G. R. [2008]. Meteorite impact structures: meteoritic contamination in impact melt the good and the bad. Geology Today 24, 13‐19. particles from the Steinheim basin (Baden‐ Osinski, G. R., Spray, J. G. & Lee, P. [2001]. Impact Württemberg, Germany). Lunar and Planetary induced hydrothermal activity within the Science Conference 41, 2103.

PAGE XIV Schneider, G. [2004]. The Roadside Geology of Meteoriten und Gesteinen irdischer Namibia. Gebruder Borntraeger, 294 pp. Impaktkrater: Identifizierung der Schulte, P., Alegret, L., Arenillas, I., Arz, J. A., Barton, Einschlagskörper. Humboldt‐Universität zu P. J., Bown, P. R., Bralower, T. J., Christeson, G. L., Berlin, 173 pp. Claeys, P., Cockell, C. S., Collins, G. S., Deutsch, A., Tagle, R. & Berlin, L. [2008]. A database of Goldin, T. J., Goto, K., Grajales‐Nishimura, J. M., chondrite analyses including platinum group Grieve, R. A. F., Gulick, S. P. S., Johnson, K. R., elements, Ni, Co, Au, and Cr: implications for the Kiessling, W., Koeberl, C., Kring, D. A., MacLeod, identification of chondritic projectiles. K. G., Matsui, T., Melosh, J., Montanari, A., Meteoritics and Planetary Science 43, 541‐559. Morgan, J. V., Neal, C. R., Nichols, D. J., Norris, R. Tagle, R. & Claeys, P. [2005]. An ordinary chondrite D., Pierazzo, E. , Ravizza, G., Rebolledo‐Vieyra, impactor for the , Siberia. M., Reimold, W. U., Robin, E. , Salge, T., Speijer, Geochimica et Cosmochimica Acta 69, 2877‐ R. P., Sweet, A. R., Urrutia‐Fucugauchi, J., Vajda, 2889. V., Whalen, M. T. & Willumsen, P. S. [2010]. The Tagle, R. & Hecht, L. [2005]. Evaluation of chemical Chicxulub Asteroid Impact and Mass Extinction methods for projectile identification in at the Cretaceous‐Paleogene Boundary. Science terrestrial and lunar impactites. Lunar and 327, 1214‐1218. Planetary Science 36, 1927. Scott, E. R. D. & Krot, A. N. [2005]. Chondrites and Tagle, R., & Hecht, L. [2006] Geochemical their components. In: Davis, A. M. (ed.), Treatise identification of projectiles in impact rocks: on Geochemistry, vol. 1: Meteorites, Comets and Meteoritics and Planetary Science 41, 1721‐ Planets, 143‐200. Elsevier, 737 pp. 1735. Sharpton, V. L. & Grieve R. A. F. [1990]. Meteorite Tagle, R., Claeys P., Öhman, T., Schmitt, R. T. & impact, cryptoexplosion, and shock Erzinger, J. [2007]. Traces of an H‐chondrite in metamorphism: A perspective on the evidence the impact melt rocks from Lappajärvi impact at the K/T boundary. In: Sharpton, V. L. & Ward, structure, Finland. Meteoritics & Planetary P. D. (eds.), Global Catastrophes in Earth Science 42, 1841‐1854. History: An Interdisciplinary Conference on Tagle, R., Schmitt, R. T., and Erzinger, J. [2009] Impacts, Volcanism, and Mass Mortality, 301‐ Identification of the projectile component in the 318. Geological. Society of America, 631 pp. impact structures Rochechouart France and Shelkov, D. A., Verchowsky, A. B., Milledge, H. J., Sääksjärvi, Finland. Implications for the Kaminsky, F. V. & Pillinger, C. T. [1998]. C, N, Ar impactor population for the Earth. Geochimica and He study of impact diamonds from Ebeliakh et Cosmochimica Acta 73, 4891‐4906. alluvial deposits and Popigai crater. Meteoritics Taylor, S. R. [1982]. Planetary Science: A Lunar and Planetary Science 33, 985‐992. Perspective. Lunar and Planetary Institute, 481 Sheppard, S. S. & Trujillo, C. A. [2006]. A Thick pp. Cloud of Neptune Trojans and Their Colors. Taylor, S. R. [1992]. Solar System Evolution: A New Science 313, 511‐514. Perspective. Cambridge University Press, 307 Steel, D. [2008]. Tunguska at 100. Nature 453, pp. 1157‐1159. Tedesco, E. F. & Desert, F. X. [2002]. The Infrared Steiger, R. H. & Jäger, E. [1977]. Subcommission on Space Observatory Deep Asteroid Search. The Geochronology: Convention on the use of decay Astronomical Journal 123, 2070‐2082. constants in geo‐ and cosmochronology. Earth Therriault A. M., Grieve R. A. F. & Reimold, W. U. and Planetary Science Letters 36, 359‐362. [1997]. Original size of the Vredefort Structure: Stern, S. A. & Weissman, P. R. [2001]. Rapid Implications for the geological evolution of the collisional evolution of comets during the Witwatersrand Basin. Meteoritics & Planetary formation of the Oort cloud. Nature 409, 589‐ Science 32, 71‐77. 591. Vauquelin, L. N. [1803]. Memoire sur les pierres Stöffler, D. & Langenhorst, F. [1994]. Shock dite tombees du ciel. Annales des Chimie 45, metamorphism of quartz in nature and 225‐245. experiment: I. Basic observation and theory. von Engelhardt, W. [1990]. Distribution, Meteoritics 29, 155‐181. petrography and shock metamorphism of the Stokes, G. H., Yeomans, D. K., Bottke, W. F., Jewitt, ejecta of the Ries crater in Germany – A review. D., Chesley, S. R., Kelso, T. S., Evans, J. B., Tectonophysics 171, 259‐273. McMillan, R. S., Gold, R. E., Spahr, T. B., Harris, A. von Paula Gruithuisen, F. [1828]. Analekten für W. & Worden, S. P. [2003]. Study to Determine Erd‐ und Himmelskunde, Erstes Heft. Joh. the Feasibility of Extending the Search for Near‐ Palm’schen Buchhandlung, 660 pp. Earth Objects to Smaller Limiting Diameters. Walker, R. J., Horan, M. F., Morgan, J. W., Becker, H., NASA Office of Space Science, 154 pp. Grossman, J. N. & Rubin, A. E. [2002]. Tagle, R. [2004]. Platingruppenelemente in Comparative 187Re‐187Os systematics of

PAGE XV chondrites: Implications regarding early solar of meteorite impact craters. CREWES Research system processes. Geochimica et Cosmochimica Report 8, 34.1‐34.26. Acta 66, 4187‐4201. Wetherill, G. W. [1976]. Where do the meteorites Wegener, A. [1921]. Die Entstehung der come from? A re‐evaluation of the Earth‐ Mondkrater. Friedrich Vieweg & Sohn, 48 pp. crossing Apollo objects as sources of chondritic Weissman, P. R. [1999]. Diversity of comets: meteorites. Geochimica et Cosmochimica Acta formation zones and dynamical paths. Space 40, 1297‐1317. Science Reviews 90, 301‐311. Wittmann, A., Kenkmann, T., Schmitt, R. T. & Weissman, P. R., Bottke, W. F. & Levison, H. F. Stöffler, D. [2006]. Shock‐metamorphosed [2002]. Evolution of Comets into Asteroids. In: zircon in terrestrial impact craters. Meteoritics Bottke, W. F., Paolicchi, P., Binzel, R. P. and and Planetary Science 41, 433‐454. Cellino, A. (eds.), Asteroids III, 669‐686. Zinner, E. K. [2005]. Presolar Grains. In: Davis, A. M. University of Arizona Press, 785 pp. (ed.), Treatise on Geochemistry, vol. 1: Westbroek, H. H. & Stewart, R. R. [1996]. The Meteorites, Comets and Planets, 17‐39. Elsevier, formation, morphology, and economic potential 737 pp.

PAGE XVI VI. APPENDICES

Appendix A. List of all 178 known impact structures on Earth [PASSC, 2010]. Given are the name of the impact structure, its location, diameter in km and age in Ma.

Name Location Diameter (km) Age (Ma) Acraman South Australia, Australia 90 ~ 590 Amelia Creek Northern Territory, Australia 20 600 – 1640 Ames Oklahoma, USA. 16 470 ± 30 Amguid Algeria 0.45 < 0.1 Aorounga Chad 12,6 < 345 Aouelloul Mauritania 0.39 3.0 ± 0.3 Araguainha 40 244.40 ± 3.25 Avak Alaska, USA 12 3 – 95 BP Libya 2 < 120 Barringer Arizona, USA 1.18 0.049 ± 0.003 Beaverhead Montana, USA 60 ~ 600 Beyenchime‐Salaatin 8 40 ± 20 Bigach Kazakhstan 8 5 ± 3 Boltysh Ukraine 24 65.17 ± 0.64 Bosumtwi Ghana 10.5 1.07 Boxhole Northern Territory, Australia 0.17 0.0054 ± 0.0015 Brent Ontario, Canada 3.8 396 ± 20 Calvin Michigan, USA 8.5 450 ± 10 Argentina 0.05 < 0.004 Carswell Saskatchewan, Canada 39 115 ± 10 Charlevoix Quebec, Canada 54 342 ± 15 Chesapeake Bay Virginia, USA 90 35.5 ± 0.3 Chicxulub Yucatán, Mexico 170 64.98 ± 0.05 Chiyli Kazakhstan 5.5 46 ± 7 Chukcha Russia 6 < 70 Clearwater East Quebec, Canada 26 290 ± 20 Clearwater West Quebec, Canada 36 290 ± 20 Cloud Creek Wyoming, USA 7 190 ± 30 Connolly Basin Western Australia, Australia 9 < 60 Couture Quebec, Canada 8 430 ± 25 Crawford South Australia, Australia 8.5 > 35 Crooked Creek Missouri, USA 7 320 ± 80 Dalgaranga Western Australia, Australia 0.02 ~ 0.27 Decaturville Missouri, USA 6 < 300 Deep Bay Saskatchewan, Canada 13 99 ± 4 Sweden 19 89.0 ± 2.7 Des Plaines Illinois, USA 8 < 280

PAGE XVII Dhala India 11 1700 – 2100 Dobele Latvia 4.5 290 ± 35 Eagle Butte Alberta, Canada 10 < 65 Elbow Saskatchewan, Canada 8 395 ± 25 El'gygytgyn Russia 18 3.5 ± 0.5 Flaxman South Australia, Australia 10 > 35 Flynn Creek Tennessee, USA 3.8 360 ± 20 Foelsche Northern Territory, Australia 6 > 545 Gardnos Norway 5 500 ± 10 Glasford Illinois, USA 4 < 430 Glikson Western Australia, Australia 19 < 508 Glover Bluff Wisconsin, USA 8 < 500 Goat Paddock Western Australia, Australia 5.1 < 50 Gosses Bluff Northern Territory, Australia 22 142.5 ± 0.8 Gow Saskatchewan, Canada 4 < 250 Goyder Northern Territory, Australia 3 < 1400 Granby Sweden 3 ~ 470 Gusev Russia 3 49.0 ± 0.2 Gweni‐Fada Chad 14 < 345 Haughton Nunavut, Canada 23 39 Haviland Kansas, USA 0.01 < 0.001 Henbury Northern Territory, Australia 0.15 0.0042 ± 0.0019 Holleford Ontario, Canada 2.35 550 ± 100 Ile Rouleau Quebec, Canada 4 < 300 Ilumetsä Estonia 0.08 ~ 0.0066 Ilyinets Ukraine 8.5 378 ± 5 Iso‐Naakkima Finland 3 > 1000 Jänisjärvi Russia 14 700 ± 5 Jebel Waqf as Suwwan Jordan 5.5 56 – 37 Kaalijärv Estonia 0.11 0.004 ± 0.001 Kalkkop South Africa 0.64 0.25 ± 0.05 Kaluga Russia 15 380 ± 5 Kamensk Russia 25 49.0 ± 0.2 Kamil Egypt 0.045 ? Kara Russia 65 70.3 ± 2.2 Kara‐Kul Tajikistan 52 < 5 Kärdla Estonia 7 ~ 455 Karikkoselkä Finland 1.5 ~230 Karla Russia 10 5 ± 1 Kelly West Northern Territory, Australia 10 > 550 Kentland Indiana, USA 13 < 97 Keurusselkä Finland 30 <1800 Kgagodi Botswana 3.5 < 180 Kursk Russia 6 250 ± 80 La Moinerie Quebec, Canada 8 400 ± 50 Lappajärvi Finland 23 73.3 ± 5.3 Lawn Hill Queensland, Australia 18 > 515 Liverpool Northern Territory, Australia 1.6 150 ± 70 Lockne Sweden 7.5 455

PAGE XVIII Logancha Russia 20 40 ± 20 Logoisk Belarus 15 42.3 ± 1.1 Lonar India 1.83 0.052 ± 0.006 Finland 9 ~ 1000 Macha Russia 0.3 < 0.007 Manicouagan Quebec, Canada 100 214 ± 1 Manson Iowa, USA 35 73.8 ± 0.3 Maple Creek Saskatchewan, Canada 6 < 75 Marquez Texas, USA 12.7 58 ± 2 Middlesboro , USA 6 < 300 Mien Sweden 9 121.0 ± 2.3 Mishina Gora Russia 2.5 300 ± 50 Mistastin Newfoundland and Labrador, Canada 28 36.4 ± 4 Mizarai Lithuania 5 500 ± 20 Mjølnir Norway 40 142.0 ± 2.6 Montagnais Nova Scotia, Canada 45 50.50 ± 0.76 Monturaqui Chile 0.46 < 1 Morasko Poland 0.1 < 0.01 Morokweng South Africa 70 145.0 ± 0.8 Mount Toondina South Australia, Australia 4 < 110 Neugrund Estonia 8 ~ 470 New Quebec Quebec, Canada 3.44 1.4 ± 0.1 Newporte North Dakota, USA 3.2 < 500 Nicholson Northwest Territories, Canada 12.5 < 400 Oasis Libya 18 < 120 Obolon' Ukraine 20 169 ± 7 Odessa Texas, USA 0.16 < 0.05 Ouarkziz Algeria 3.5 < 70 Paasselkä Finland 10 < 1800 Piccaninny Western Australia, Australia 7 < 360 Pilot Northwest Territories, Canada 6 445 ± 2 Popigai Russia 100 35.7 ± 0.2 Presqu'ile Quebec, Canada 24 < 500 Puchezh‐Katunki Russia 80 167 ± 3 Ragozinka Russia 9 46 ± 3 Red Wing North Dakota, USA 9 200 ± 25 Riachão Ring Brazil 4.5 < 200 Ries Germany 24 15.1 ± 0.1 Rio Cuarto Argentina 4.5 < 0.1 Rochechouart France 23 214 ± 8 Rock Elm Wisconsin, USA 6 < 505 Roter Kamm Namibia 2.5 3.7 ± 0.3 Rotmistrovka Ukraine 2.7 120 ± 10 Sääksjärvi Finland 6 ~ 560 Saarijärvi Finland 1.5 > 600 Saint Martin Manitoba, Canada 40 220 ± 32 Santa Fe New Mexico , USA 9.5 <1200 Serpent Mound Ohio, USA 8 < 320 Serra da Cangalha Brazil 12 < 300

PAGE XIX Shoemaker Western Australia, Australia 30 1630 ± 5 Shunak Kazakhstan 2.8 45 ± 10 Sierra Madera Texas, USA 13 < 100 Sikhote Alin Russia 0.02 0.000063 Siljan Sweden 52 376.8 ± 1.7 Ontario, Canada 30 ~ 450 Sobolev Russia 0.05 < 0.001 Söderfjärden Finland 6.6 ~ 600 Spider Western Australia, Australia 13 > 570 Steen River Alberta, Canada 25 91± 7 Steinheim Germany 3.8 15 ± 1 Strangways Northern Territory, Australia 25 646 ± 42 Suavjärvi Russia 16 ~ 2400 Sudbury Ontario, Canada 250 1850 ± 3 N Finland 4 < 1000 Tabun‐Khara‐Obo Mongolia 1.3 150 ± 20 Talemzane Algeria 1.75 < 3 Tenoumer Mauritania 1.9 0.0214 ± 0.0097 Ternovka Ukraine 11 280 ± 10 Tin Bider Algeria 6 < 70 Tookoonooka Queensland, Australia 55 128 ± 5 Tswaing South Africa 1.13 0.220 ± 0.052 Tvären Sweden 2 ~ 455 Utah, USA 10 < 170 Vargeão Dome Brazil 12 < 70 Veevers Western Australia, Australia 0.08 < 1 Vepriai Lithuania 8 > 160 ± 10 Viewfield Saskatchewan, Canada 2.5 190 ± 20 Vista Alegre Brazil 9.5 < 65 Vredefort South Africa 300 2023 ± 4 Wabar Saudi Arabia 0.11 0.00014 Wanapitei Ontario, Canada 7.5 37.2 ± 1.2 Wells Creek Tennessee, USA 12 200 ± 100 West Hawk Manitoba, Canada 2.44 351± 20 Wetumpka Alabama, USA 6.5 81.0 ± 1.5 Whitecourt Alberta, Canada 0.04 <0.0011 Wolfe Creek Western Australia, Australia 0.87 < 0.3 Woodleigh Western Australia, Australia 40 364 ± 8 Xiuyan China 1.8 > 0.05 Yarrabubba Western Australia, Australia 30 ~ 2000 Zapadnaya Ukraine 3.2 165 ± 5 Zeleny Gai Ukraine 3.5 80 ± 20 Zhamanshin Kazakhstan 14 0.9 ± 0.1

PAGE XX Appendix B. Siderophile element abundances and ratios for meteorites [Tagle, 2004; Tagle et al, 2009; and Fischer‐Gödde et al., 2010]. ppm Ni Co Cr Ni/Cr SD Co/Cr SD Ni/Co SD CI 10863 521 2796 3.87 0.25 0.19 0.01 20.87 1.62 CM 12396 583 3059 4.01 0.30 0.19 0.02 21.27 1.78 CO 13564 688 3503 3.96 0.09 0.20 0.02 19.73 1.85 CV 13629 641 3557 3.76 0.12 0.18 0.01 21.26 1.45 CK 12518 647 3627 3.45 0.40 0.18 0.02 19.34 3.15 CR 13794 666 3810 3.72 0.39 0.17 0.02 20.71 2.55 CH 25716 1125 3393 7.65 0.69 0.33 0.06 22.86 2.83 K 17172 786 2575 9.72 5.01 0.31 0.26 21.85 5.06 R 14366 700 3563 3.99 0.19 0.20 0.01 20.53 1.03 H 16846 763 3470 4.38 0.42 0.22 0.03 22.07 2.99 L 12936 571 3650 3.22 0.19 0.16 0.02 22.67 3.33 LL 9585 452 3521 2.64 0.21 0.13 0.02 21.19 3.12 EH 18006 856 3076 5.79 0.36 0.28 0.03 21.04 1.95 EL 14676 727 3148 4.77 1.03 0.23 0.04 20.18 2.86 ppb Os Ir Ru Pt Rh Pd Au CI 502 478 725 970 136 570 140 CO 758 726 1047 1265 0 772 184 CM 671 617 877 1147 156 624 166 CV 830 760 1078 1458 169 717 146 CK 814 770 1119 1274 177 665 130 CR 655 621 918 1115 745 124 CH 1155 1066 1566 1596 249 K 673 628 960 209 R 658 603 908 121 H 835 749 1010 1446 193 789 213 L 565 525 765 1074 127 628 153 LL 391 339 496 686 82 449 119 EH 673 565 937 1207 173 858 331 EL 639 569 830 1103 143 723 240 IIIC 1550 4702 5036 1808 4684 IA 2380 4887 6135 1406 3611 IB 1210 1869 3828 640 3525 IIE‐A 6618 9750 11377 2490 3415 IIE‐M 1313 3633 6370 1580 2963 IIID 91.00 155 801 677 23362

Os/Ir SD Ru/Ir SD Pt/Ir SD Rh/Ir SD CI 1.06 0.04 1.52 0.07 2.03 0.15 0.29 0.014 CM 1.12 0.09 1.50 0.10 1.84 0.09 0.26 0.013 CO 1.08 0.05 1.49 0.08 1.66 0.23 CV 1.06 0.03 1.52 0.08 1.90 0.12 0.28 0.014 CK 1.07 0.05 1.45 0.14 2.02 0.07 0.28 0.014 CR 1.05 0.04 1.49 0.07 1.94 0.16 CH 1.08 0.04 1.47 0.05 1.50 0.27

PAGE XXI Os/Ir SD Ru/Ir SD Pt/Ir SD Rh/Ir SD K 1.07 0.02 1.53 0.02 R 1.06 0.05 1.48 0.07 H 1.07 0.04 1.51 0.07 2.08 0.06 0.31 0.004 L 1.07 0.04 1.49 0.09 2.05 0.12 0.33 0.007 LL 1.08 0.04 1.55 0.07 2.10 0.12 0.34 0.006 EH 1.13 0.04 1.62 0.16 2.06 0.04 0.33 0.017 EL 1.11 0.08 1.53 0.14 2.01 0.04 0.33 0.020

Pd/Ir SD Au/Ir SD Pt/Pd SD Ru/Rh SD CI 1.19 0.10 0.30 0.04 1.69 0.23 4.95 0.25 CM 1.07 0.09 0.27 0.02 1.70 0.13 5.08 0.25 CO 0.94 0.11 0.26 0.01 1.80 0.06 CV 0.93 0.07 0.19 0.02 2.10 0.15 5.01 0.25 CK 0.97 0.05 0.16 0.05 1.82 0.09 CR 1.18 0.17 0.20 0.05 1.69 0.24 CH 0.90 0.05 0.23 0.07 2.00 0.10 K 0.34 0.09 R 0.23 0.09 H 1.10 0.10 0.28 0.03 1.86 0.07 4.73 0.12 L 1.22 0.14 0.30 0.03 1.70 0.19 4.43 0.17 LL 1.49 0.21 0.35 0.03 1.40 0.17 4.40 0.16 EH 1.62 0.09 0.57 0.05 1.27 0.05 4.94 0.25 EL 1.31 0.07 0.43 0.04 1.53 0.09 4.98 0.02

Pt/Ru SD Ru/Pd SD Pt/Rh SD Pd/Rh SD CI 1.37 0.06 1.15 0.01 7.36 0.17 4.18 0.21 CM 1.23 0.14 1.45 0.22 7.18 0.36 4.08 0.20 CO 1.24 0.07 1.46 0.13 CV 1.38 0.17 1.65 0.08 6.93 0.35 3.38 0.17 CK 1.30 0.06 1.41 0.07 7.51 0.38 3.52 0.18 CR 1.30 0.03 1.30 0.20 CH 0.94 0.17 H 1.38 0.04 1.33 0.06 6.77 0.14 3.63 0.09 L 1.38 0.11 1.16 0.10 6.54 0.12 4.08 0.16 LL 1.35 0.11 1.04 0.16 6.46 0.21 4.70 0.24 EH 1.33 0.03 0.96 0.03 6.39 0.32 5.04 0.25 EL 1.32 0.09 1.16 0.10 6.25 0.40 4.11 0.28

Ir/Pd SD Ni/Ir SD Co/Ir SD Cr/Ir SD CI 0.84 0.07 23.02 1.15 1.10 0.07 5.92 0.20 CM 0.93 0.07 20.48 1.29 0.96 0.05 5.06 0.34 CO 1.06 0.13 18.70 1.64 0.95 0.03 4.83 0.38 CV 1.07 0.08 17.97 1.13 0.85 0.02 4.69 0.20 CK 1.03 0.05 16.21 1.97 0.84 0.09 4.70 0.36 CR 0.85 0.12 22.18 2.02 1.07 0.09 6.13 0.63 CH 1.11 0.06 24.13 1.23 1.06 0.12 3.18 0.42 K 27.34 4.86 1.25 0.19 4.10 3.50 R 23.58 1.03 1.15 0.03 5.85 0.23

PAGE XXII Ir/Pd SD Ni/Ir SD Co/Ir SD Cr/Ir SD H 0.91 0.08 22.48 2.20 1.02 0.10 4.63 0.45 L 0.82 0.09 25.84 2.83 1.14 0.11 7.29 0.42 LL 0.67 0.09 28.50 2.41 1.35 0.16 10.47 0.84 EH 0.62 0.03 31.41 2.42 1.49 0.08 5.37 0.40 EL 0.76 0.04 25.92 3.02 1.28 0.10 5.56 0.98

PAGE XXIII