Kimberlites Associated with the Lucapa Structure, Angola
Kimberlites associated with the Lucapa structure, Angola
Sandra Elvira Robles Cruz
ADVERTIMENT. La consulta d’aquesta tesi queda condicionada a l’acceptació de les següents condicions d'ús: La difusió d’aquesta tesi per mitjà del servei TDX (www.tdx.cat) ha estat autoritzada pels titulars dels drets de propietat intel·lectual únicament per a usos privats emmarcats en activitats d’investigació i docència. No s’autoritza la seva reproducció amb finalitats de lucre ni la seva difusió i posada a disposició des d’un lloc aliè al servei TDX. No s’autoritza la presentació del seu contingut en una finestra o marc aliè a TDX (framing). Aquesta reserva de drets afecta tant al resum de presentació de la tesi com als seus continguts. En la utilització o cita de parts de la tesi és obligat indicar el nom de la persona autora.
ADVERTENCIA. La consulta de esta tesis queda condicionada a la aceptación de las siguientes condiciones de uso: La difusión de esta tesis por medio del servicio TDR (www.tdx.cat) ha sido autorizada por los titulares de los derechos de propiedad intelectual únicamente para usos privados enmarcados en actividades de investigación y docencia. No se autoriza su reproducción con finalidades de lucro ni su difusión y puesta a disposición desde un sitio ajeno al servicio TDR. No se autoriza la presentación de su contenido en una ventana o marco ajeno a TDR (framing). Esta reserva de derechos afecta tanto al resumen de presentación de la tesis como a sus contenidos. En la utilización o cita de partes de la tesis es obligado indicar el nombre de la persona autora.
WARNING. On having consulted this thesis you’re accepting the following use conditions: Spreading this thesis by the TDX (www.tdx.cat) service has been authorized by the titular of the intellectual property rights only for private uses placed in investigation and teaching activities. Reproduction with lucrative aims is not authorized neither its spreading and availability from a site foreign to the TDX service. Introducing its content in a window or frame foreign to the TDX service is not authorized (framing). This rights affect to the presentation summary of the thesis as well as to its contents. In the using or citation of parts of the thesis it’s obliged to indicate the name of the author. Kimberlites associated with the Lucapa structure, Angola
Sandra Elvira Robles Cruz
Kimberlites associated with the Lucapa structure, Angola
by Sandra Elvira Robles Cruz
BIENNIUM 2007-2008 Ciencies de la Terra
PhD. Thesis ACADEMIC DISSERTATION
Departament de Cristal·lografia, Mineralogia i Dipòsits Minerals Facultat de Geologia Universitat de Barcelona 2012
1
Supervisors: Dr. Joan Carles Melgarejo Draper
Universitat de Barcelona
Dr. Salvador Galí Medina
Universitat de Barcelona
Dr. Mónica Escayola
CONICET-IDEAN
Committee: Dr. José Mangas Viñuela (President)
Universidad de Las Palmas de Gran Canaria
Dr. Joaquín A. Proenza F. (Secretary)
Universitat de Barcelona
Dr. M. Pura Alfonso Abella (Comm. Member)
Universitat Politècnica de Catalunya
Dr. Maite García Vallès (Alternate) Dr. Fernando Gervilla L. (Alternate)
Universitat de Barcelona Universidad de Granada
Cover: View from northwest of the Catoca mine, Angola.
2
ABSTRACT
Six kimberlite pipes within the Lucapa structure in northeastern Angola have been investigated using major and trace element geochemistry of mantle xenoliths, macro- and megacrysts.
Geothermobarometric calculations were carried out using xenoliths and well-calibrated single crystals of clinopyroxene. Geochronological and isotopic studies were also performed where there were samples available of sufficient quality.
Results indicate that the underlying mantle experienced variable conditions of equilibration among the six cites. Subsequent metasomatic enrichment events also support a hypothesis of different sources for these kimberlites. The U/Th values suggest at least two different sources of zircon crystals from the Catoca suite. These different populations may reflect different sources of kimberlitic magma, with some of the grains produced in U- and Th-enriched metasomatized mantle units, an idea consistent with the two populations of zircon identified on the basis of their trace element compositions.
Calculated temperature and pressure from xenoliths are less scattered than T-P data calculated from single crystals. The calculated northeastern Angola paleogeotherm is consistent with a single value for the CA and the CU79 kimberlites. The differences in T-P values between these kimberlites may reflect the different way each kimberlite sampled the lithosphere. The lithospheric thickness calculated from the northeastern Angola paleogeotherm yielded 192 km.
This research shows that the absence of fresh Mg-rich ilmenite in the Catoca kimberlite (one of the largest bodies of kimberlite in the world), as well as the occurrence of Fe3+-rich ilmenite, do not exclude the presence of diamond in the kimberlite. This is a new insight into the concept of ilmenite and diamond exploration, and leads to the conclusion that compositional attributes must be evaluated in light of textural attributes.
3
The tectonic setting of northeastern Angola was influenced by the opening of the South Atlantic
Ocean, which reactivated deep NE–SW-trending faults during the early Cretaceous. The new interpretation of a kimberlitic pulse during the middle of the Aptian and the Albian, which provides precise data on the age of a significant diamond-bearing kimberlite pulse in Angola, will be an important guide in future exploration for diamonds. These findings contribute to a better understanding of the petrogenetic evolution of the kimberlites in northeastern Angola and have important implications for diamond exploration.
Keywords: kimberlite; Angola; ilmenite; garnet; clinopyroxene; diamond; zircon; xenolith, mantle, Lucapa.
4
RESUMEN
Kimberlitas asociadas a la estructura Lucapa fueron estudiadas mediante geoquímica de elementos mayoritarios y elementos traza tanto en xenolitos del manto, como en macro- y megacristales provenientes de seis chimeneas kimberlíticas localizadas en el noreste de Angola. Cálculos geotermobarométricos se realizaron utilizando xenolitos del manto y cristales individuales de clinopiroxeno bien calibrados. Estudios geocronológicos e isotópicos se realizaron en aquellos casos donde se contaba con muestras de buena calidad disponibles.
Los resultados indican que el manto subyacente experimentó diferentes condiciones de equilibrio.
Eventos posteriores de enriquecimiento metasomático también apoyan la hipótesis de diferentes fuentes para estas kimberlitas. Los valores de U/Th sugieren al menos dos fuentes diferentes para los cristales de circón provenientes de la kimberlita de Catoca. Estas poblaciones diferentes puede reflejar diversas fuentes de magma kimberlítico, donde algunos de los granos podrían haberse producido en unidades del manto metasomatizadas y enriquecidas en U y Th, una idea que es coherente con las dos poblaciones de circón identificados con base en composiciones de elementos traza.
Los valores de temperatura y presión calculados a partir de xenolitos muestran menor dispersión que los datos TP calculados a partir de cristales individuales. La paleogeoterma calculada para las kimberlitas de CA y CU79 se ajusta a un solo rango de valores. En general, las diferencias en los valores de PT entre estas kimberlitas pueden reflejar la forma diferencial como cada kimberlita muestrea la litosfera. El espesor de la litosfera calculado a partir de la paleogeoterma es de 192 km para el noreste de Angola.
Esta investigación también demuestra que la ausencia de ilmenita fresca rica en Mg en la kimberlita de Catoca (una de las kimberlitas más grandes del mundo), así como la presencia de ilmenita rica en Fe3+ no excluye la presencia de diamantes en dicha kimberlita. Esta es una nueva
5 visión sobre el concepto de ilmenita en la exploración de diamantes, y conduce a la conclusión de que los estudios de composición deben estar acompañados de caracterizaciones texturales.
El ambiente tectónico en el noreste de Angola fue influenciado por la apertura del Océano
Atlántico Sur, lo cual reactivó profundas fallas con tren NE-SW durante el Cretácico temprano. La nueva interpretación de un pulso kimberlítico durante la mitad del Aptiense y Albiense proporciona datos precisos sobre la edad de un pulso kimberlítico diamantífero muy significativo en Angola, esta información será una guía importante para futura exploración de diamante. Estos resultados también contribuyen a una mejor comprensión de la evolución petrogenética de las kimberlitas en el noreste de
Angola y tienen importantes implicaciones para la exploración de diamante.
Palabras clave: kimberlita; Angola; ilmenita; granate; clinopiroxeno; diamante; circón; xenolito, manto, Lucapa.
6
TABLE OF CONTENTS
ABSTRACT...... 3
RESUMEN…...... 5
TABLE OF CONTENTS...... 7
LIST OF ORIGINAL PUBLICATIONS AND PARTICIPATION OF SERC IN EACH PUBLICATION……...... 8
PREFACE...... 10
CHAPTER 1 - INTRODUCTION...... 11 1.1 Kimberlites…………...... 11 1.2 Diamond production from kimberlites...... 14 1.3 Diamond production in Angola……...... 16 1.4 Geology of Northeastern Angola...... 17 1.5 The aim of the thesis...... 19 1.6 Methodology...... 21 1.7 Structure of the thesis...... 24
CHAPTER 2 – REVIEW AND RESULTS OF ORIGINAL PUBLICATIONS……………………………………………………………..………...... 25 2.1 Paper I...... 25 2.2 Paper II………………...... 26 2.3 Paper III……………...... 27 2.4 Paper IV……………...... 28 2.5 Paper V……………...... 28 2.6 Paper VI……………...... 30
CHAPTER 3 –DISCUSSION...... 31 3.1 The SCLM beneath Angola and implications for diamond exploration...... 31 3.2 Heterogeneous mantle and metasomatism revealed by subsolidus reactions in ilmenite...... 32 3.3 Diamond potential and regional comparison among diamondiferous and barren kimberlites...... 34 3.4 Future research...... 36
CHAPTER 4 – MAIN CONCLUSIONS...... 37
ACKNOWLEDGMENTS...... 39
REFERENCES...... 41
ORIGINAL PUBLICATIONS...... 49
RESUMEN DE LA TESIS EN ESPAÑOL...... 111
7
LIST OF ORIGINAL PUBLICATIONS AND PARTICIPATION OF SERC IN EACH PUBLICATION
This thesis includes the following six publications:
Paper I. Robles-Cruz, S., Watangua, M., Melgarejo, J.C., Galí, S., 2008. New Insights into the
Concept of Ilmenite as an Indicator for Diamond Exploration, Based on Kimberlite Petrographic
Analysis. MACLA - Revista de la Sociedad Española de Mineralogía, September No. 9, 205-206.
Published.
Paper II. Robles-Cruz, S.E., Watangua, M., Melgarejo, J.C., Gali, S., Olimpio, A., 2009. Contrasting compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in exploration for diamond. Lithos 112S, 966-975. Published.
Paper III. Robles-Cruz, S., Lomba, A., M., Melgarejo, J., Galí, S., Olimpio, A., 2009. The Cucumbi
Kimberlite, NE Angola: Problems to Discriminate Fertile and Barren Kimberlites. MACLA - Revista de la Sociedad Española de Mineralogía, September No.11, 159-160. Published.
Paper IV. Robles-Cruz, S.E., Escayola, M., Melgarejo, J.C., Watangua, M., Galí, S., Gonçalves, O.A.,
Jackson, S., 2010. Disclosed data from mantle xenoliths of Angolian kimberlites based on LA-ICP-MS analyses, in: Acta Mineralogica-Petrographica. Abstract Series, Vol. 6, pp. 553. Published.
Paper V. Robles-Cruz, S.E., Escayola, M., Jackson, S., Galí, S., Pervov, V., Watangua, M., Gonçalves,
O.A., Melgarejo, J.C., 2012. U–Pb SHRIMP geochronology of zircon from the Catoca kimberlite,
Angola: Implications for diamond exploration. Chemical Geology 310-311, 137-147. Published.
Paper VI. Robles-Cruz, S.E., Melgarejo, J.C., Galí, S., Escayola, M., 2012. Major- and trace-element compositions of indicator minerals that occur as macro- and megacrysts, and of xenoliths, from kimberlites in northeastern Angola. Minerals, Special Issue "Advances in Economic Minerals".
Officially accepted for publication.
8
S.E. Robles-Cruz’s contribution to the multi-authored paper was:
Papers I, III, and IV, she participated in the fieldwork and sampling. She carried out the petrography studies, SEM imaging, mineral chemistry analyses (microprobe and LA-ICP-MS), processing, and writing the papers.
Paper II, she participated in the fieldwork and sampling. She carried out the petrography studies,
SEM imaging, mineral chemistry analyses, processing, and writing the manuscript for the most part.
Paper V, she participated in the fieldwork and sampling. She carried out the petrography studies,
SEM imaging, mineral chemistry analyses, LA-ICP-MS analyses, preparation of samples for
SHRIMP analyses, processing and interpretation of raw data from LA-ICP-MS and SHRIMP analyses, and writing the manuscript.
Paper VI, she participated in the fieldwork and sampling. She carried out the petrography studies,
SEM imaging, mineral chemistry analyses, LA-ICP-MS analyses, preparation of samples for Sm/Nd analyses, processing and interpretation of data, and writing the manuscript.
9
PREFACE
Kimberlites are one of the most fascinating types of rocks from the Earth. They are complex rocks and provide significant information about the mantle. As well, the study of kimberlites contributes to a better understanding of the evolution of the planet. The study of kimberlites also has economic relevance, since they can trap diamonds during their ascent.
I began my Ph.D. Project in 2008 at the Departament of Cristal·lografia, Mineralogia i Dipòsits
Minerals, Facultat de Geologia, Universitat de Barcelona, with the financial support of a 3-year FI grant and then a BE 6-month grant, both sponsored by the Departament d'Educació i Universitats of the Generalitat de Catalunya and European Social Fund.
This project was the continuation of the Diploma de Estudios Avanzados (DEA) I presented in
2007 under the supervision of Professor Joan Carles Melgarejo i Draper. The PhD research project was directed by Prof. Joan Carles Melgarejo i Draper and Prof. Salvador Galí, both professors from the Department of Cristal·lografia, Mineralogia i Dipòsits Minerals department, Facultat de Geologia,
Universitat de Barcelona. Dr. Monica Escayola from CONICET-IDEAN Instituto de Estudios
Andinos, Laboratorio de Tectónica Andina, Universidad de Buenos Aires also participated as co- advisor of this Ph.D. thesis. The Ph.D. project was supported by the projects CGL2005-07885/BTE and CGL2006-12973 of Ministerio de Educación y Ciencia (Spain).
10
CHAPTER 1 – INTRODUCTION
1.1 Kimberlites
Kimberlites are relatively rare rocks of great scientific and economic importance. The name
“kimberlite” name was proposed by Professor Henry Carvil Lewis in 1887 and since then that is how the rock has been known as (Lewis, 1887; 1888). Lewis described the rock as a type of volcanic breccia, a porphyritic mica-bearing peridotite (Mitchell 1995). The name followed the type-locality rules of nomenclature at that time; it was named after the locality Kimberley, South Africa. Two large groups of kimberlites (group I and II) were introduced by Smith (1983), based on isotopic studies.
Smith et al. (1985) and Skinner (1986, 1989) proposed that kimberlites could be divided in these two distinct groups: group I (kimberlites sensu stricto), and group II (orangeites, phlogopite-rich
“kimberlites”). Later, several studies clearly established that group I and II “kimberlites” are mineralogically and geochemically quite distinct, and group II rocks have closer affinities to lamproites than to group I kimberlites (Mitchell, 1995, and references therein).
Kimberlites, also known as group I kimberlites (Mitchell 1995), are defined as volatile-rich
(dominantly CO2) potassic ultrabasic rocks that usually show a distinctive inequigranular texture as a result of the presence of crystals (macro- and megacrysts) and xenoliths inside a fine-grained matrix
(Clement and Skinner, 1985; Mitchell, 1986). The mineralogy of kimberlites is very variable and complex. Mega- and macrocrysts are mainly composed of olivine, magnesian ilmenite, Cr-poor titanian pyrope, diopside, phlogopite, enstatite, and Ti-poor chromite; where olivine macrocrysts are a characteristic component except in fractionated kimberlites (Mitchell 1995). Some kimberlites may also contain diamond. Mantle and crustal xenoliths can be also present in kimberlites. The fine- grained matrix may include a second generation of primary euhedral-to-subhedral olivine, monticellite, phlogopite, perovskite, spinel, apatite, and serpentine (Mitchell, 1995). It has been also reported (Kamenetsky et al., 2004) that the groundmass is extremely enriched (at least 8 wt.%) in
11 water-soluble alkali chlorides, alkali carbonates, and sulfates (proportion 5:3:1), and commonly shows immiscibility textures between these phases.
Kimberlites occur as pipe intrusions (Figure 1.1) with an upper crater facies, intermediate diatreme facies, and deep hypabyssal facies (Clement and Skinner, 1985). These facies were produced by explosive emplacement under volcanic and subvolcanic conditions. Crater facies rocks are divided into lavas, pyroclastic rocks, and resedimented volcaniclastic rocks; kimberlite diatremes are cone shaped, composed of clasts of cognate or xenolithic origin with or without matrix, and classified as
“tuffisitic kimberlite” and “tuffisitic kimberlite breccia” (Clement 1982; Clement and Skinner, 1985;
Mitchell 1995); and hypabyssal kimberlites comprise the root zones of diatreme and occur as dikes and sills (Mitchell 1995, and references therein).
The study of xenoliths, megacrysts (crystals greater than 1 cm in their maximum dimension) and macrocrysts (0.5-10 mm) from kimberlites play an important role in the understanding of the characteristics of the mantle and the kimberlite petrogenesis itself. Minerals such as pyrope and eclogitic garnet, chrome diopside, Mg-rich ilmenite, chromite and, to a lesser extent, olivine in superficial materials (tills, stream sediments, loam, etc.) are one of the most important tools, other than bulk sampling, to assess the diamond content of a particular pipe (Pell, 1998), consequently they are called indicator minerals.
Kimberlites are preferentially associated with cratons worldwide (Figure 1.2). Diamondiferous kimberlites have been reported as Proterozoic to Tertiary in age, with diamond crystals that vary from early Archean to as young as 990 Ma (Pell 1998). In 1995 there were already 5000 kimberlites identified, and 10% of them were diamondiferous (Janse and Sheahan 1995).The first kimberlite was discovered in 1869 in South Africa where the first diamond from primary deposit was found. Three years later kimberlites were recognized as primary deposit for diamond (Janse and Sheahan 1995).
Diamond, however, is not genetically related to kimberlites, but rather it is a xenocryst that is formed in the upper mantle. Diamond in kimberlites can be found as sparse xenocrysts or diamondiferous xenoliths hosted by intrusives emplaced as subvertical pipes or resedimented volcaniclastic and pyroclastic rocks deposited in craters (Pell 1998). Most of the natural diamond crystals come from peridotite and in less proportion (33%) from eclogitic sources (Stachel and Harris, 2009, and
12 references therein). The research of kimberlites and natural diamond from kimberlites had a significant impetus since the 1st International Kimberlite Conference in 1973.
Figure 1.1 Idealized diagram of a kimberlite magmatic system (after Mitchell 1995)
Mitchell (1986) defined kimberlites using the typomorphic assemblage of primary minerals and emphasizing their petrologic characteristics as: “Kimberlites are inequigranular alkalic peridotites containing rounded and corroded megacrysts of olivine, phlogopite, magnesian ilmenite and pyrope set in fine-grained groundmass of second generation euhedral olivine and phlogopite together with primary and secondary (after olivine) serpentine, perovskite, carbonate (calcite and/or dolomite) and spinels. The spinels range in composition from titaniferous magnesian chromite to magnesian ulvöspinel-magnetite. Accessory minerals include diopside, monticellite, rutile and nickeliferous sulphides. Some kimberlites contain major modal amounts of monticellite”.
13
Figure 1.2 Kimberlites and major diamond mines worldwide after Janse (2007) and Eckstrand et
al. (1995).
There are three large cratons in Africa: the South African, the West African, and the Central
African cratons. Most of the world’s active major kimberlite diamond mines are located in South
Africa, Botswana, Zimbabwe and Swaziland, on the South African craton which includes the Kalahari
Archon. The West African Craton includes the Man Archon and the Eburnean Proton. The Central
African craton includes two important archons: the Lunda-Kasai (Angola and Congo) and the
Tanzanian (Janse and Sheahan 1995). In Angola, kimberlite pipes and dykes are distributed in the northeast, central, and southwest part of the country. Most of the diamondiferous kimberlites in
Angola are concentrated in clusters in the northeastern area. There are also several alluvial mining areas in Lunda and Cuango, Angola (Llusià et al., 2005).
1.2 Diamond production from kimberlites
Worldwide diamond production from kimberlites is not easy to track since not all values are published and sometimes when they are published they may vary from one publication to another.
The Kimberley Process Certificate Scheme (KPCS) that monitors world rough diamond trade came into effect on 1 January 2003 (Read and Janse, 2009). This was the first real attempt to at least restrain
14 trade in conflict diamonds and to provide surveillance of the rough diamond trade from producers to merchants. Diamond production like other commodities (e.g., gold) depends on demand. Botswana,
Russia, Canada, South Africa, and Angola (in this order) were the top five diamond producing countries by value responsible of the 83% of the total world production in 2009. This represented the
65% of the total diamond production by weight in 2009, since DRC and Australia were in the top five producers by weight but they produce diamonds low in value (Read and Janse, 2009). Figure 1.3 shows the total diamond production for the main 38 kimberlites worldwide until 2009. Currently, the top five diamond producing countries by value are Botswana, Russia, Canada, South Africa and
Angola.
Figure 1.3 Diamond production worldwide (after Janse, 2007; Read and Janse, 2009)
15
1.3 Diamond production in Angola
Kimberlites in Angola are important not only because most of them are slightly eroded so their crater facies are well preserved, but also because there are large and high-grade kimberlites with significant potential diamond reserves (Khar'kiv et al., 1992). The first kimberlite in Angola, the
Camafuca-Camzambo pipe, was discovered in 1947 (De Andrade 1954). However, the first kimberlite that came into production in 1997 was the Catoca pipe, which was discovered in 1985. The civil war in Angola, between 1961 and 2002 (Blore, 2004), hampered progress in the country. The guerrillas controlled the richest diamond provinces and mined them illegally in part to fund their activities. The
National Union for the Total Independence of Angola (UNITA) and the Revolutionary United Front
(RUF), both acted against the international community's objectives of restoring peace in Angola
(Blore 2007). It is clear that “informal” diamond production was much higher than the official values e.g., UNITA’s smuggled production is estimated to have been worth close to $1 billion in 1996 (Janse
2007, and references therein). Angola was the first country to implement a full certificate of origin for diamond exports (at the beginning of 2000) following United Nations sanctions on UNITA’s diamond trading in 1998 and the beginning of investigation into illegal diamond trading in 1999 (Blore, 2004).
The goal of this certificate was to verify the exclusion of conflict diamonds. After 2000 and especially once the civil war ended up in 2002, the mining activities accelerated in Angola. The Catoca pipe passed from 2 Mct/year in 2000 to produce 6.7 Mct/year in 2007 (Read and Janse, 2009). Diamond production in Angola represents the 1% of the gross domestic product (GDP) of Angola (Bermúdez-
Lugo 2004). Estimated reserves in Angola are of 50 million carats in kimberlite pipes (Partnership
Canada Africa, 2004). Ore reserves in the Catoca pipe are given as 84 million tonnes to yield 60 million carats to 150 m depth or as 270 million tonnes to yield 195 Mct to 600 m depth (Read and
Janse, 2009). The mining in Camafuca pipe started in 2007 (low cost operation dredging the river bed) to recover 200,000 ct/yr for five years on a reserve of 13 Mct, which are contained in fluvial mud and sand grading into highly weathered kimberlite (Read and Janse, 2009). The Camatchia–Camagico mine in Angola is developed on two kimberlite pipes, where a reserve of 80 Mct has been estimated
(Read and Janse, 2009). Figure 1.3 includes the diamond production from Angolan kimberlites.
16
1.4 Geology of northeastern Angola
Angola is endowed with mineral resources that are the result of a relatively complex geological history. The study of its cratonic lithospheric mantle is important both for the light it sheds on the physical behaviour of old continents, as well as in contributing to our understanding of Angola's mineral potential.
Angola geology can be represented by three main stages (De Carvalho et al., 2000; Guiraud et al.,
2005, Figure 1.4): (1) An important Archean orogeny, registered by the Central Shield, Cuango Shield and Lunda Shield, most of them composed of gabbro, norite and charnockitic complexes, which constitute the Angolan basement. (2) Three main Proterozoic cycles, Eburnean-Paleoproterozoic,
Kibaran-Mesoproterozoic, and Pan-African-Neoproterozoic; being the Eburnean the most important and characterized by complex volcanosedimentary groups, gneisses and migmatites, granites and syenites. This regional Paleoproterozoic event was followed by the Kibaran cycle, which was related to extensional events that occurred on the border of Congo craton and that later generated clastic- carbonatic sequences and local basic magmatism. The Pan-African orogeny was associated with the development of Gondwana and leaded the generation of fold belts and granitic intrusions. The activation of zones of lithospheric weakness, especially major fault zones, favoured the subsequent break-up of Gondwana. (3) The deposition of Phanerozoic sedimentary sequences resting unconformably on previously eroded surfaces (Pereira et al., 2003). The subsequent break-up of
Gondwana, during the Jurassic to Cretaceous, between 190 and 60 Ma (e.g., Jelsma et al., 2004), caused the development of basins that are associated with deep fault systems in Angola. These fault systems facilitated the emplacement of alkaline, carbonatitic, and kimberlitic magmas (Pereira et al.,
2003).
The Lower Cretaceous regional extension determined the development of deep faults and grabens with trends NE-SW and NW-SE. The Lucapa structure is in the first group (trend NE-
SW).The northeastern part where most of the diamondiferous kimberlites in Angola are found, whereas the southwestern zone comprises important occurrences of undersaturated alkaline rocks and carbonatites (Reis, 1972). More minor kimberlite fields are found in the SW Angola (Egorov et al.,
2007).
17
Figure 1.4 Location map of the area of study. Geological map of northeastern Angola (after
De Araujo et al., 1988; De Araujo and Perevalov, 1998; De Carvalho et al., 2000; Egorov et al.,
2007). Abbreviations: Quaternary (QQ), Cenomanian (CE), Albian (AB), Permian (PP),
Carboniferous (CC), Undifferentiated (Undiff.), Group (Gp), Formation (Fm), sandstone (Sst),
conglomerate (Cgl), limestone (Lst), marlstone (Mrls), argillaceous limestone (ArgLst), claystone
(Clst), granite (Gr), gabbro (Gb), quartzite (Qzt), schist (Sch), granodiorite (Grdr), dolerite (Do),
amphibolite (Am), gneiss (Gns), carbonatites (Cbt), nephelite (Nph), syenite (Syt), ijolite (Ijt),
pyroxenite (Pxt), anorthosite (Ant), troctolite (Trt), Norite (Nrt), epidotite (Epd), granulite (Gnt),
eclogite (Ecl).
Kimberlites from southern Africa, North America, and Russia show similar ages between them. They show alternating periods of abundance or scarcity of kimberlite magmatism (Figure 1.5),
18 especially during Cenozoic/Mesozoic (Heaman et al., 2003; Jelsma et al., 2009). The coincidence of kimberlite occurrences with trans-lithosphere discontinuities may be a result of thermal perturbations.
Such conditions were favoured during rifting and the eventual supercontinent breakup, when a majority of these kimberlites were generated (Heaman et al., 2003). The geological configuration in
Angola, which was consistent with the aforementioned conditions associated with kimberlite generation, apparently set a tectonic control on the presence of kimberlites in Angola. Synsedimentary continental sediments (Calonda Formation) filled the Lucapa structure. The Lucapa structure is an old corridor from an oceanic transform (White et al., 1995), which has been active since Paleoproterozoic
(Jelsma et al., 2009), and is characterized by deep-seated faults associated with carbonatites and kimberlites. The Calonda Formation can also contain diamonds in paleoplacers; alluvial diamonds are found in placers associated with rivers passing across all these diamondiferous areas.
Figure 1.5 Three types of tectonic settings related to kimberlite magmatism. (a) Gondwana
assemblage during Pan-African orogeny. (b) Incipient rifting. Dark gray represents the Karoo basins.
(c) Trans-continental lithospheric discontinuities (gray lines) that have been reactivated, as tectonic
triggers, during the continental extension; and concomitant magmatism (dashed line) in Southern
Africa and South America. The white diamonds represent schematic groups of kimberlites and related
rocks (after Jelsma et al., 2009).
1.5 The aim of the thesis
The Lucapa structure has several hundreds of kimberlites (Figure 1.6). To date, there is no official information about all of them. The knowledge of kimberlites in Angola is low and according
19 to ENDIAMA (2012) only the 40% of the mining resources have been evaluated. Currently, there are
167 mining projects in Angola, 15 of them are active and the most important ones are: Muanga, Alto
Cuilo, Dala, Nhefo, Lunda-Nordeste, Cacuala, and Gango. The kimberlites under current production are: Catoca in Lunda Sul, and Camatchia, Camafuca, Camatue, and Camazuanza in Lunda Norte.
There is no detailed production information available for most of them.
Figure 1.6 Distribution of kimberlites in Angola (including data from Perevalov et al., 1992;
Egorov et al., 2007)
Some mineralogical studies have been carried out in the Catoca pipe to determine the diamondiferous potential of this kimberlite (Ganga et al., 2003; Kotel’nikov et al., 2005). However, there are important questions to address in terms of genesis and evolution of kimberlites in northeastern Angola. Kimberlite magma is considered as derived from the mantle of the Earth at a depth of more than 150 km (Dawson, 1980; Haggerty, 1995). Mantle xenoliths provide information about the subcratonic mantle and lithosphere, as well as melts and fluids associated with mantle
20 metasomatism. The analyses of indicator minerals provide information about the oxygen fugacity conditions, favorable conditions to sample and preserve diamond, and evolution of the kimberlite itself.
Research that I carried out during the “Trabajo de Investigación Tutelado” (TRT - Treball de
Recerca Tutelat) during 2007, established that ilmenite macrocrysts from the Catoca kimberlite exhibit different grades of replacement below 200 m depth, and are almost not visible above this level. This research builds upon the 2007 TRT research to determine if the composition of ilmenite from the Catoca kimberlite indicates one (or alternately multiple) recrystallization events, and explores the relationship between the different types of ilmenite and presence/preservation of diamond in this area. To determine this information, a regional study of kimberlites in the northeastern part of Angola was undertaken using sampling collected from drill cores of kimberlites in the Lunda, Catoca, and Muanga areas.
Another major objective of this thesis is to propose a profile that provides information about the mantle beneath the northeastern Angola based on the study of xenoliths, mega- and macrocrysts from six kimberlites. As well, an evaluation of the conditions that has an influence on diamond distribution along the area of study, based on petrographic and geochemical studies of barren (kimberlite without diamond presence) and diamondiferous kimberlites. Unfortunately, the comparison with barren kimberlites has been hampered because samples from barren kimberlites were the more altered and poor in fresh mantle xenoliths and indicator minerals.
1.6 Methodology
This research was developed in different phases: 1) field work, 2) sampling preparation, 3) analyses, and 4) discussion and writing of manuscripts.
1.6.1 Field work
There was preliminary field work in 2005 when samples from the Catoca kimbelite were collected by J.C. Melgarejo and his research group. Then in 2006, I started my “Trabajo de
Investigación Tutelado” (TRT - Treball de Recerca Tutelat) and I used those samples to start getting
21 an idea about the kimberlites from Angola and the results were presented to obtain the Diploma d'Estudis Avanzats (DEA, Robles-Cruz, 2007). In fall 2007, new field work occurred during which I collected the samples for the Ph.D. thesis. Specifically, about 750 drill cores and heavy-mineral concentrate samples from seventeen kimberlites were obtained (Table 1.1). Only six of the seventeen kimberlites containing samples of good quality were selected to carry out analyses: Catoca (CA),
Tchiuzo (TZ), Anomaly 116 (An116), Alto Cuilo-4 (AC4), Alto Cuilo-63 (AC63), and Cucumbi-79
(CU79).
Presence of diamonds Province Contract Kimberlite Borehole (YES/NO) LUNDA NORTE LUEMBA Tchiuzo 34 YES LUNDA NORTE LUEMBA Tchiuzo 44 YES LUNDA NORTE LUEMBA Tchiuzo G10 YES LUNDA NORTE LUEMBA Tchiuzo G18 YES LUNDA SUL CATOCA Catoca 0335 YES LUNDA SUL CATOCA Catoca 0536 YES LUNDA SUL CATOCA Catoca 033/35 YES LUNDA SUL CATOCA Catoca 044/35 YES LUNDA SUL CATOCA Catoca 77/35 Unknown LUNDA SUL CATOCA Catoca CA135 YES LUNDA SUL CATOCA Catoca CA336 YES LUNDA SUL CATOCA Catoca CA515 YES LUNDA SUL CATOCA Catoca CA535 YES LUNDA SUL CATOCA Catoca CA538 YES LUNDA SUL CATOCA Anomaly CAT-116 116 Some prospectivity LUNDA SUL CATOCA Camitongo 28 Some prospectivity LUNDA SUL LAPI Kambundu 216 NO LUNDA SUL ALTO CUILO Alto Cuilo 1 1 NO LUNDA SUL ALTO CUILO Alto Cuilo 16 11 Some prospectivity LUNDA SUL ALTO CUILO Alto Cuilo 254 5 YES LUNDA SUL ALTO CUILO Alto Cuilo 4 4 Some prospectivity LUNDA SUL ALTO CUILO Alto Cuilo 5 5 NO LUNDA SUL ALTO CUILO Alto Cuilo 63 6 YES LUNDA NORTE MUANGA Cucumbi 45 5 NO (not tested) LUNDA NORTE MUANGA Cucumbi 72 MFD07 YES LUNDA NORTE MUANGA Cucumbi 76 MFD03 NO LUNDA NORTE MUANGA Cucumbi 79 MFD01 YES LUNDA NORTE MUANGA Cucumbi 8 MFD06 NO LUNDA NORTE MUANGA Cucumbi 80 2 NO (not tested) Table 1.1 List of kimberlites sampled for this Ph.D. thesis
1.6.2 Sampling and preparation
This phase started in February 2008, when samples arrived from Angola after passing all the authorization process, and finished in March 2009, before I went to the Geological Survey of Canada
22 to carry out analyses. Kimberlites are very delicate rocks and they need to be prepared properly (low vacuum, no water, and diamond powder for polishing), otherwise they become useless. Unfortunately, some of the samples when revised had to be rejected (bad sample preparation). A set of thin (30 μ) and gross (80-100 μ) sections, and probes was obtained. The classification of the kimberlite textures was followed after Mitchell (1986; 1995), Pearson et al. (2007, and references therein), and Scott
Smith (2012).
1.6.3 Analytical methods
Samples were studied under the optical transmitted and reflective light microscope at the
Department of Cristal·lografia, Mineralogia i Dipòsits Minerals – Faculty of Geology, in order to pick up the best and representative samples to carry out the different type of analyses.
Petrographic studies were performed with a Scanning Electron Microscope – Environmental
Scanning Electron Microscope (SEM-ESEM) with an acopled EDS using BSE (Backscattered electrons) images to identify compositional heterogeneities in the samples (previously dried at 60°C during 24 hours for thin sections and during 72 hours for gross sections, cleaned blowing away powder, and then carbon coating). ESEM Quanta 200FEI, XTE325/D08395 was used to carry out these analyses. High vacuum conditions were preferred to get a precision of less than 0.5 μ in the spot. This equipment has a LINK EDS, which is made up by a Si (Li) crystal, with a Be window. This configuration allows determination of all the elements from Be to U.
Mineral chemistry of major elements were carried out using a Cameca SX-50 microprobe (see parameters at PAPER II) and a JXA JEOL-8900L microprobe (see parameters at PAPER VI). Mineral chemistry of trace elements were accomplished using laser-ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), see parameters in the PAPER V and PAPER VI.
Geochronological U-Pb analyses were conducted using a Sensitive High Resolution Ion
Microprobe II (SHRIMP II), see parameters at PAPER V. Additional Sm/Nd isotopes analyses were carried out using a Thermo Finnigan Triton thermo-ionization mass spectrometer (TIMS).
23
1.7 Structure of the thesis
This PhD thesis presents the results divided in chapters based on the main findings of the research. The second chapter is a revision of the original publications written as part of this Ph.D. thesis. The third chapter will integrate all these analyses and results in a general discussion about the studied kimberlites. Finally, we will present the main conclusions of this research in the chapter fourth and we will suggest the main directions of the future research about kimberlites in Angola. The original publications are included at the end of this thesis.
24
CHAPTER 2 – REVIEW AND RESULTS OF ORIGINAL
PUBLICATIONS
2.1 Paper I
“New Insights into the Concept of Ilmenite as an Indicator for Diamond Exploration, Based on
Kimberlite Petrographic Analysis”
Biannual national journal: MACLA - Revista de la Sociedad Española de Mineralogía,
September No. 9, 205-206.
ISSN: 1885-7264.
Paper I is based on the preliminary findings about ilmenite from the Catoca kimberlite,
Angola. It is a continuation of the “Trabajo de Investigación Tutelado” (TRT – “Treball de Recerca
Tutelat”) carried out by SERC during 2007, and a comparison with ilmenite from the Cucumbi-79 kimberlite. Paper I also includes the regional setting of the area of research. Textural evidences of ilmenite indicate a different complex history of growth in the crystals. Six textural types of ilmenite were identified in Catoca and three compositional types of ilmenite.
The composition of the ilmenite from Catoca is the result of a set of replacement processes with rich fluids in Mg and Mn affecting an oxidized primary ilmenite in a higher or lower grade. These fluids are reducing, especially those rich in Mn. "Picroilmenite" has traditionally been interpreted as an indicator of kimberlite associations, as well as an indicator of low fO2, which is necessary for the preservation of diamond. Although Catoca and Cucumbi are diamondiferous kimberlites, they show that Mg ilmenite is clearly a late replacement product, and the grade of replacement of the primary grains is very variable. Therefore, this paper illustrates that the absence of magnesian ilmenite in a kimberlite does not appear to be a convincing argument to exclude the presence of diamonds. This is a new insight into the concept of ilmenite in diamond exploration.
25
2.2 Paper II
“Contrasting compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in exploration for diamond”
Monthly international journal: Lithos 112S, 966-975.
ISSN: 0024-4937. Impact Factor: 3.246 (2011), 5-Year Impact Factor: 3.691 (Thomson Reuters,
2012). Journal in the Science Citation Index (SCI).
Paper II describes a detailed systematic petrographic characterization of the different types of ilmenite from the Catoca kimberlite. The Catoca kimberlite is emplaced in the northeastern part of the
Lucapa structure. The paper focuses on compositional and textural variations in ilmenite from drill- core material, in the hope of elucidating events before and during the emplacement of the kimberlitic magma. We characterize four main compositional variants of ilmenite, with enrichments in Fe3+, Mg,
Mn and nearly stoichiometric ilmenite, in seven textural classes of ilmenite, and distinguished crystals of variable size, ranging from micro- to megacrysts.
Most ilmenite is found to derive, through a complex process, from replacement of Fe3+-rich
3+ ilmenite, presumably originating by mantle metasomatism at a relatively high fO2. This Fe -rich ilmenite reacted with fluids under reducing conditions, producing Mg-rich ilmenite. The Mn-rich ilmenite is produced by interaction with a late CO2-rich fluid. The Mg-rich ilmenite is here clearly a minor phase and a late product of replacement. The absence of fresh Mg-rich ilmenite and the occurrence of Fe3+-rich ilmenite do not seem to be convincing arguments to exclude the presence of diamond crystals in a kimberlite.
The ilmenite macro- and megacrysts are assumed to be produced by disaggregation of ilmenite- bearing xenoliths (mainly relatively oxidized and metasomatized mantle peridotites and minor carbonatites). The subsequent reaction under disequilibrium conditions with kimberlite-derived fluids produced the replacement of the above macro- and megacrysts by secondary Mg-rich ilmenite. Late subsolidus reactions with the fluids associated with the kimberlite, also in disequilibrium conditions, produced the replacement of the early ilmenite types by highly reduced Mn-rich ilmenite. The enrichment in Nb of this late ilmenite (and in the ilmenite of the matrix), as well as its intimate
26 association with carbonates of Ba and Sr, indicate the interaction of the ilmenite crystals with a CO2- rich fluid.
This work proposes a new understanding of the connection between the search of ilmenite in diamond exploration: compositional attributes must be evaluated in light of textural attributes.
Although Catoca is a diamondiferous kimberlite, most of its ilmenite compositions are strongly oxidized and poor in Cr and Mg. Therefore, an important conclusion of Paper II is that the absence of
Mg-rich ilmenite in a kimberlite, or the absence of its corresponding placers, do not appear to be a convincing argument to exclude the occurrence of economic deposits of diamond in a kimberlite.
2.3 Paper III
“The Cucumbi Kimberlite, NE Angola: Problems to Discriminate Fertile and Barren
Kimberlites”
Biannual national journal: MACLA - Revista de la Sociedad Española de Mineralogía,
September No.11, 159-160.
ISSN: 1885-7264.
Paper III focuses on the petrography and composition of samples from the Cucumbi kimberlite.
The garnet compositions from Cucumbi-79 are plotted using the diagram of Grütter et al. (2004). The compositions plot into the graphite domain, out of the diamondiferous field harzburgitic G10 facies.
Based solely on this criterion the kimberlite would be classified as barren. However, the
Cucumbi kimberlite is diamondiferous. Similar problems were found in the Catoca pipe when using the composition of ilmenite or the composition of garnets. Therefore, the paper concludes that the garnet diagrams can be used to verify the minimum level of diamond content, but some kimberlites may contain diamond samples from deeper sources and that this should be taken into consideration when using these diagrams to assess the potential of kimberlite fields.
It is also important to mention that new diagrams have been proposed (i.e., Grütter et al.,
2006, McLean et al., 2007), where they integrate several attributes at once, and they seem to be a better tool for exploration.
27
2.4 Paper IV
“Disclosed data from mantle xenoliths of Angolan kimberlites based on LA-ICP-MS analyses”
National journal: Acta Mineralogica-Petrographica. Abstract Series, Vol. 6, pp. 553.
Published by the Department of Mineralogy, Geochemistry and Petrology, University of Szeged,
Hungary.
ISSN 0324-6523.
This paper presents preliminary observations of the type of xenoliths found in the Catoca and
Cucumbi-79 diamondiferous kimberlites, and the first set of analyses of Laser Ablation-Inductively
Coupled Plasma-Mass Spectrometry (LA-ICP-MS) from xenoliths. Two main different trends for garnet can be identified in the Catoca kimberlite based on Rare Earth Element (REE) patterns.
Eclogitic garnet has “normal” normalized Rare Earth Element (REEN, McLean et al., 2007) patterns, whereas garnet from lherzolite xenoliths usually has “sinusoidal” REEN patterns and rarely “normal”
REEN patterns. Clinopyroxene from eclogitic associations is Light REE (LREE) enriched. Garnet from the lherzolite xenoliths is characterized by a LREE-enrichment, a maximum around the LREE-
Heavy REE (HREE) limit and flat HREE.
Unlike in Cucumbi-79, garnet from lherzolite xenoliths presents “normal” patterns with lower
REE values. Garnet from phlogopite-rich xenoliths presents “normal” patterns, but their values are significantly (about 10x chondritic value) lower. Only clinopyroxene from phlogopite-rich xenoliths exhibits higher values in LREE than the same xenoliths in the Catoca pipe.
Data indicate that the mantle sampled by these two kimberlites might have been under different equilibration conditions and different degrees of metasomatism.
2.5 Paper V
“U–Pb SHRIMP geochronology of zircon from the Catoca kimberlite, Angola:
Implications for diamond exploration”
Semimonthly international journal: Chemical Geology 310-311, 137-147.
ISSN: 0009-2541. Impact Factor: 3.518 (2011), 5-Year Impact Factor: 4.063 (Thomson Reuters,
2012). Journal in the SCI.
28
Paper V presents the first age determinations of zircon from the diamondiferous Catoca kimberlite in northeastern Angola, the fourth largest kimberlite body in the world. The U–Pb ages were obtained using a Sensitive High Resolution Ion Microprobe II (SHRIMP II) on zircon crystals derived from tuffisitic kimberlite (TK) rocks and heavy-mineral concentrates from the Catoca kimberlite.
The SHRIMP results define a single weighted mean age of 117.9±0.7 Ma (Mean square weighted deviation MSWD=1.3). More than 90% of the results indicate a single age population. There is no evidence for variable ages within single crystals, and no diffusional profiles are preserved. These data are interpreted as the maximum age of the kimberlite eruption at Catoca. The U/Th values suggest at least two different sources of zircon crystals. These different populations appear to indicate different sources of kimberlitic magma, with some of the grains produced in U- and Th-enriched metasomatized mantle units.
This understanding is consistent with the two populations of zircon identified, based on REE abundances determined by LA-ICP MS analyses in this paper. One population originated from a depleted mantle source with low total REE (less than 25 ppm), and the other was derived from an enriched source, likely from the mantle or a carbonatite-like melt with high total REE (up to 123 ppm).
The tectonic setting of northeastern Angola has been influenced by the opening of the south
Atlantic, which reactivated deep NE–SW-trending faults during the early Cretaceous. The eruption of the Catoca kimberlite correlates with these regional tectonic events. The Calonda Formation (Albian–
Cenomanian age) is the earliest sedimentary unit that incorporates eroded material derived from the diamondiferous kimberlites. Thus, the age of the Catoca kimberlite eruption is restricted to the time between the middle of the Aptian and the Albian. This new interpretation will be an important guide in future exploration for diamonds because it provides precise data on the age of a diamond-bearing kimberlite pulse in Angola.
29
2.6 Paper VI
“Major- and trace-element compositions of indicator minerals that occur as macro- and megacrysts, and of xenoliths, from kimberlites in northeastern Angola”
Quarterly international journal: Minerals, Special Issue "Advances in Economic Minerals"
(submitted revised version of the manuscript).
ISSN: 2075-163X. Peer-reviewed open access journal. Published by the Multidisciplinary Digital
Publishing Institute (MDPI).
Paper VI compares the major- and trace-element compositions of olivine, garnet, and clinopyroxene that occur as single crystals (142 grains), with those derived from xenoliths (51 samples) from six kimberlites in the Lucapa area, northeastern Angola: Tchiuzo, Anomaly 116,
Catoca, Alto Cuilo-4, Alto Cuilo-63, and Cucumbi-79.
The samples were analyzed using electron probe microanalysis (EPMA) and LA-ICP-MS.
The results suggest different paragenetic associations for these kimberlites in the Lucapa area.
Compositional overlap in some of the macrocryst and mantle xenolith samples indicates a xenocrystic origin for some of those macrocrysts. The presence of mantle xenocrysts suggests a possibility diamond being present. Geothermobarometric calculations were carried out using EPMA data from xenoliths applying the program PTEXL.XLT. Additional well calibrated single-clinopyroxene thermobarometric calculations were also applied.
Results indicate the underlying mantle experienced different equilibration conditions.
Subsequent metasomatic enrichment events also support a hypothesis of different sources for the kimberlites. These findings contribute to a better understanding of the petrogenetic evolution of the kimberlites in northeastern Angola and have important implications for diamond exploration.
30
CHAPTER 3 – DISCUSSION
3.1 The SCLM beneath Angola and implications for diamond exploration
The characterization of the sub-continental lithospheric mantle (SCLM) is important for identifying the evolution of continents and their mineral potential (Pearson and Wittg, 2008). In particular, the mantle xenolith suite and certain garnet and clinopyroxene xenocrysts provide information regarding the composition and structure of the SCLM. In cases where fresh xenoliths are poor or absent, xenocrysts are very useful. Although they do not provide information as precise as xenoliths, xenocrysts can give a statistically reliable sample of the underlying mantle (Schulze, 1995).
Figure 3.1 Schematic model comparing diamondiferous and barren kimberlites from northeastern
Angola (modified after Haggerty, 1986; Mitchell, 1986; Mather et al., 2011). Lithosphere-
asthenosphere boundary (LAB), graphite (G), diamond (D).
31
The calculated temperature and pressure from xenoliths (PAPER VI), define a single paleogeotherm value for the CA and the CU79 kimberlites, and yielded a lithospheric thickness of
192 km calculated based on this paleogeotherm. A quantitative comparison between Angola lithosphere and geotherms from Bultfontein and Finsch kimberlites in southern Africa indicates a slightly cooler (steeper) paleogetherm for Angola than the paleogeotherms calculated from southern
Africa. This is consistent with the map of the lithospheric thickness of Southern Africa from shear wave velocities (Preistley and McKenzie, 2006), which indicates a thickness >180 km.
3.2 Heterogeneous mantle and metasomatism revealed by subsolidus reactions in ilmenite
Ilmenite from kimberlites in northeastern Angola are very particularly important because it is a common mineral that provides relevant information about their chemical environment and exhibit
2+ 3+ variable Fe :Fe ratios. This ratio is significantly influenced by fO2 conditions, and these conditions can be used to determine the preservation or destruction of diamond. The detailed petrographic study of the Catoca kimberlite (PAPER II) suggests a complex history for the ilmenite nodules. The diversity in textures and composition reflects the paragenetic position of ilmenite in the kimberlite
(accessory in xenoliths, macro- and megacrysts, matrix) and the replacement processes. We propose that most if not all of the ilmenite nodules are produced by disaggregation of ilmenite-bearing metasomatized peridotite xenoliths.
The composition of the early ilmenite is unusual because of its high Fe3+ contents. Similar Fe3+- rich compositions, although rare in kimberlites, have also been found in the Koidu kimberlite, in
Sierra Leone (Tompkins and Haggerty, 1985). The ilmenite from the Catoca pipe is even more strongly oxidized, indicating crystallization under relatively high fO2 conditions. Ilmenite is also replaced along small discontinuities, both in the grain borders and along internal surfaces by Mg-rich ilmenite. The replacement of the Fe3+-rich ilmenite by Fe2+- and Mg-rich ilmenite is indicative of a trend toward more reducing conditions (Haggerty and Tompkins, 1983). This type of sequence is similar to the so-called ilmenite magmatic trend (Haggerty et al., 1977; Pasteris, 1980; Schulze,
32
1984). However, the textural patterns attributable to replacement at Catoca, along grain borders, cracks or other discontinuities, strongly suggest the action of a fluid rather than a magma. It is difficult to ascertain the timing and place of this replacement. Certainly it was produced before kimberlite emplacement, because some nodules broken during the explosive processes are not replaced in the broken corners.
The replacement of the Fe3+-rich ilmenite by Fe2+- and Mn-rich ilmenite is also indicative of a trend toward strongly reducing conditions (Haggerty and Tompkins, 1983). Accordingly, these compositions could follow the kimberlite reaction trend of Haggerty et al. (1977), producing enrichment in Fe2+. The most significant aspects in this process are the strong enrichments in Mn and
HFSE. Similar enrichments have been interpreted in other kimberlites worldwide as produced by crystallization at the expense of a late-stage fraction of melt (Tompkins and Haggerty, 1985,
Chakhmouradian and Mitchell, 1999). In the Catoca case, two facts suggest the deposition of this ilmenite under the influence of a CO2-rich fluid phase: a) the intimate association of this Mn-rich ilmenite along with calcite, witherite, barytocalcite and strontianite; b) the development of this mineral association filling small fractures. In fact, the late stages of kimberlite emplacement are developed under the influence of CO2-rich fluids (Head and Wilson, 2008), which are also responsible for the alteration of host rocks in many kimberlite fields worldwide (Smith et al., 2004); Agee et al.
(1982) also attributed the formation of Mn-rich ilmenite in the Elliott County kimberlite, Kentucky
(USA) to Ca-enriched late fluids.
The composition of this replaced ilmenite is similar to that of the fine-grained euhedral ilmenite crystals found in the kimberlite matrix. Analogous trends have already been described in other kimberlite fields, but in the hypabyssal facies (i.e. Hunter et al., 1984). Similar textures and compositions in the groundmass are not rare in kimberlites. Tompkins and Haggerty, 1985;
Chakhmouradian and Mitchell, 1999 interpreted this type of ilmenite as produced by primary magmatic crystallization in the matrix of the kimberlite. In all these cases, however, Mn-rich ilmenite is produced in late events in the paragenetic sequence at Catoca, and in many cases contains other groundmass minerals such as perovskite and spinel (Tompkins and Haggerty, 1985). Although Mn-
33 rich ilmenite could be produced during magmatic crystallization, we contend that it could also be produced during late hydrothermal processes, during serpentinization. In fact, pyrophanite can be produced during serpentinization of ultrabasic rocks, where it appears as a late mineral in the paragenetic sequence (Mücke and Woakes, 1986; Liipo et al., 1994).
In any case, all of the ilmenite fractions in the specific Catoca kimberlite are quite different from those found in the carbonatitic xenoliths at Catoca. In the case of ilmenite from carbonatitic xenoliths, the growth of ilmenite takes place during the early stages of magmatic crystallization, and there is no evidence of replacement of a precursor ilmenite. Moreover, the crystals are distinct from the other variants of ilmenite in being extremely poor in Mg and Cr and the richest in Nb, thus defining a particular class, more similar to ilmenite found in carbonatites (Gaspar and Wyllie, 1983;
1984).
The existence of many varieties of ilmenite at Catoca has significant implications for mineral exploration. Magnesium-rich ilmenite has traditionally been interpreted as an indicator of kimberlite associations, as well as an indicator of low fO2, which is necessary for the preservation of diamond
(Garanin et al., 1997; Van Straaten et al., 2008). However, the Fe3+-rich ilmenite in the Catoca kimberlite represents more than 70% of the volume of the grains, and compositions fall into the domains of “no preservation of diamond” according to the diagram of Gurney and Zweistra (1995).
Moreover, these compositions of ilmenite are Mg- and Cr-poor, and hence using other criteria for discrimination among fertile and barren kimberlites (i.e., Haggerty, 1995). Based on this understanding the Catoca kimberlite could be expected to be barren. Although Catoca is a diamondiferous kimberlite, Mg-rich ilmenite here is clearly a product of late replacement, and the extent of replacement of the primary grains is very variable. This means, textural relations must be taken into account in the application of discriminates based on composition.
3.3 Diamond potential and regional comparison among diamondiferous and barren kimberlites
The interpretation of a maximum age for the kimberlitic eruption at 118 ± 1 Ma (PAPER V) is consistent with the idea than cretaceous kimberlites in Angola are expected to be younger than the carbonatites and alkaline rocks found in the Lucapa structure (Jelsma et al., 2012). Cretaceous
34 kimberlitic events of similar age have been reported in the São Francisco craton (Brazil), the
Kaapvaal craton (South Africa and Botswana), and the Congo-Kasai craton (the Democratic Republic of Congo), which were all part of Gondwanaland (e.g., Batumike et al., 2007; Jelsma et al., 2009).
Systems of deep faults present in these cratons probably were the focus of thermal perturbations and injection of melt.
Our interpretation of 118 ± 1 Ma for the maximum age of the kimberlitic eruption in Catoca, which is associated with a NE-SW tectonic trend (Lucapa structure), reinforces the hypothesis of
Jelsma et al. (2009) that 120 Ma (Aptian age) kimberlites are preferentially associated with NE-SW tectonic trends, whereas 85 Ma (Santonian age) kimberlites are emplaced in E-W lineaments. Our finding of an Aptian age for the maximum age of the kimberlitic eruption in Catoca is also consistent with a single model for the magmatic province, which would have extended over what is now southeastern Brazil and southwestern Africa, coincident with the opening of the South Atlantic Ocean
(Hawkesworth et al., 1992, 1999; Guiraud et al., 2010). The extensional tectonic setting, rifting, and opening of the South Atlantic during the Early Cretaceous (Pereira et al., 2003; Jelsma et al., 2009) and the reactivation of deep-seated fault systems probably contributed to lithospheric heating (mantle upwelling) and, ultimately, to kimberlitic magmatism in Angola.
The geochronological studies (PAPER V) and geochemical studies suggest that the distribution of kimberlites in Angola is strongly influenced by the tectonic setting. The presence and preservation of diamond depends on the chemical conditions (e.g., fO2, metasomatism) and the rate of ascent of the kimberlite magma which traps and transports diamond to the crust. Haggerty (1986), already proposed that intra- and inter-kimberlite diamond grades differ because of the heterogeneous distribution of potential, differences in the sources, sorting of diamonds during entrainment, flow and mixing of different batches of kimberlites, and varying degrees of resorption of diamond in the ascent magma.
To date there is no information to validate the idea of mineralogical differences between diamond-bearing and diamond-free (barren) kimberlites. The tectonic configuration sets the favorable conditions for diamond presence (kimberlites that pass through craton roots), and the detailed petrographic study of indicator minerals, e.g., ilmenite, and xenoliths provide essential information to determine the conditions for the preservation of diamond.
35
3.4 Future research
Additional work that was not able to be included in this thesis since it is still in progress.
Includes work on the trace element compositions of ilmenites from diamondiferous (CA and CU79) and a barren kimberlite (CU76). The analytical part has been completed and I am currently working on the results and discussion of these data. It is anticipated this research will be ready in January
2013.
Some fluid inclusions in ilmenite were identified during the analyses of PAPER II. Then, ten representative samples of ilmenite with fluid inclusions from the CA, TZ, CU79, and AC4 were selected and analyzed by Dr. D. Kamenetsky. These data will be used for next publications.
Unfortunately, the set of samples of fresh xenoliths arrived after the author of this thesis
(SERC) carried out the analytical phase. These materials will be used for future Ph.D. thesis.
36
CHAPTER 4 – MAIN CONCLUSIONS
The main conclusions of this thesis are:
1. The presence and distribution of the studied kimberlites in northeastern Angola is influenced by
the tectonic setting, as we have been able to determine based on regional and geochronological
studies. The diamondiferous Catoca kimberlite is tectonically related to other Early Cretaceous
kimberlites associated with NE – SW lineaments in southwestern and southern Africa.
2. The maximum age of eruption of the Catoca kimberlite as being during the Aptian provides
precise data on the age of an important diamond-bearing kimberlite pulse in northeastern Angola
and should act as an important guide for diamond exploration.
3. The age of the Catoca kimberlite is restricted to between 118± 1 Ma (the maximum age for the
kimberlite eruption in Catoca) and 112 Ma, the beginning of deposition of diamondiferous clasts
in the Calonda Formation. The eruptive event for the Catoca kimberlite appears to have taken
place in this range of ages.
4. The preservation and differences in diamond grade among kimberlites is influenced by the fO2,
mixing of kimberlite batches, rate of ascent of the magma toward the crust, and different events of
metasomatism.
5. The composition of the Catoca ilmenite is complex, and the result of multiple processes. The
ilmenite macro- and megacrysts were likely produced by disaggregation of ilmenite-bearing
xenoliths (mainly relatively oxidized and metasomatized mantle peridotites and minor
carbonatites). The subsequent reaction under disequilibrium conditions with kimberlite-derived
fluids produced the replacement of the above macro- and megacrysts by secondary Mg-rich
ilmenite.
6. Late subsolidus reactions with the fluids associated with the kimberlite, also in disequilibrium
conditions, produced the replacement of the early ilmenite types by highly reduced Mn-rich
ilmenite. The enrichment in Nb of this late ilmenite (and in the ilmenite of the matrix), as well as
37
its intimate association with carbonates of Ba and Sr, can be interpreted in terms of an interaction
of the ilmenite crystals with a CO2-rich fluid.
7. New understanding in regards to the concept of ilmenite in diamond exploration is proposed. The
absence of Mg-rich ilmenite in a kimberlite or the corresponding placers does not appear to be a
convincing argument to exclude the occurrence of economic deposits of diamond.
8. Some of the zircon crystals from the Catoca kimberlite could have been produced in U–Th-
enriched metasomatized mantle units (MARID or glimmeritic suite assemblages), while others
have chemistries suggestive of a depleted asthenosphere source.
9. The CA and CU79 diamondiferous kimberlites indicate different sources and metasomatic events,
and the diamond present in each one may be derived from different protoliths.
10. The calculated northeastern Angola paleogeotherm is consistent with a single value for the CA
and the CU79 kimberlites. The differences in T-P values between these kimberlites may reflect
the different way each kimberlite sampled the lithosphere. The lithospheric thickness calculated
from the northeastern Angola paleogeotherm yielded 192 km.
38
ACKNOWLEDGMENTS
This doctoral thesis was funded by the projects CGL2005-07885/BTE and CGL2006-12973 of
Ministerio de Educación y Ciencia (Spain), and the AGAUR SGR 589 and AGAUR SGR 444 of
Generalitat de Catalunya. I received an FI pre-doctoral grant and a BE grant both sponsored by the
Departament d'Educació i Universitats de la Generalitat de Catalunya and European Social Fund. I thank ENDIAMA and the Sociedade Mineira de Catoca, LDA, especially Dr. Vladimir Pervov
(petrologist for Catoca), Prof. M. Watagua, and all the mine geologists from Catoca, Alto Cuilo and
Muanga, who allowed to acquire samples for this study as well as facilities during the mine trip. Also thanks to the Universidade Agostinho Neto (Dr. A.O. Gonçanvels) for facilitating the trips in Angola.
I acknowledge the Geological Survey of Canada, Ottawa, especially to Dr. Simon Jackson, for all his help during my six-month research visit and during the writing phase, also thanks to Dr. Bill
Davis who helped me with the revision and interpretation of the SHRIMP data. I also acknowledge the Electron Microprobe Laboratory, Department of Earth and Planetary Sciences, McGill University, especially to Mr. Lang Shi for assistance in the use of EPMA. I also express my thanks to the Serveis
Cientificotècnics de la Universitat de Barcelona for assistance in the use of SEM/ESEM-BSE-EDS analyses (E. Prats, R. Fontarnau†, Dr. J. García Veigas), and EMP (Dr. X. Llovet). Thanks to M.
Rejas (ICTJA) for assistance in separation of some samples.
I am grateful to Prof. Joan Carles Melgarejo Draper and Prof. Salvador Galí Medina who not only directed my Ph.D. thesis but also gave me all their support as good advisors and friends. I also wish to thank Dr. Mónica Escayola (also co-director) who established all the contacts to carry out the
LA-ICP-MS, SHRIMP, and Sm/Nd analyses at the Geological Survey of Canada, Ottawa, and at the
Pacific Centre for Isotopic and Geochemical Research, Department of Earth and Ocean Sciences,
University of British Columbia, Vancouver.
I am indebted to Dr. Robert Martin, emeritus professor at the Earth & Planetary Sciences
Department, McGill University, who reviewed all my manuscripts, arranged for me to acquire the
39
EPMA analyses at the McGill University, helped me and advised me through the whole thesis, and also corrected parts of this Ph.D. thesis. I acknowledge with gratitude the help and advice I received from Vicki Loschiavo. The interesting discussions I had with professors who are kimberlite experts during international conferences are greatly appreciated: they contributed to the development of this thesis. I also value the guidance I received from Prof. Joaquin Proenza during challenging moments of this thesis.
I want to express my enormous gratitude to my partner Rainer and our baby to be born, who are my inspiration for this thesis; also to my mother Perla, brother Wilson, and cousins Lupe and Charli, and to my friends: Rafael David, Hildebrando, Leonardo, Fernando, Andrea, Sebastien, Ignacio,
Jaume, Amaia, and Eder, who helped me a lot in different ways and at different stages during the thesis. Also thanks to the Recursos Minerals research group and to the Cristal·lografia, Mineralogia i
Dipòsits Minerals Department for all their help during these years. Finally, I dedicate this thesis to the memory of my grandparents, who always gave me energy, motivation, and courage to accomplish different goals.
40
REFERENCES
Agee, J.J., Garrison Jr., J.R., Taylor, L.A., 1982. Petrogenesis of oxide minerals in kimberlite, Elliott
County, Kentucky (USA). Am. Mineral. 67 (1–2), 28–42.
Batumike, J.M., O’Reilly, S.Y., Griffin, W.L., Belousova, E.A., 2007. U–Pb and Hf-isotope analyses
of zircon from the Kundelungu Kimberlites, D.R. Congo: Implications for crustal evolution.
Precambrian Res.156 (3-4), 195-225.
Bermúdez-Lugo, O., 2004. The Mineral Industry of Angola. In U.S.Geological Survey Minerals
Yearbook-2004. U.S.G.S.
Blore, S., 2004. The key to kimberlite internal diamond controls, seven case studies. Partnership
Africa Canada and Global Witness Publishing Inc., 1-20.
Blore, S., 2007. Diamond Industry Annual Review. Republic of Angola. Partnership Africa Canada,
1-24.
Chakhmouradian, A.R., Mitchell, R.H., 1999. Niobian ilmenite, hydroxylapatite and sulfatian
monazite: alternative hosts for incompatible elements in calcite kimberlite from
Internatsional'naya, Yakutia. Can. Mineral. 37 (5), 1177–1189.
Clement, C.R., 1982. A comparative geological study of some major kimberlite pipes in the Northern
Cape and Orange Free State. Thesis (Ph. D.), University of Cape Town.
Clement, C.R., Skinner, E.M.W., 1985. A textural-genetic classification of kimberlites. Trans. Geol.
Soc. S. Afr. 88, 403-409.
Dawson, J. B., 1980. Kimberlites and their Xenoliths. Anon. Germany: Springer-Verlag, pp. 252.
41
De Andrade, C.F., 1954. On the discovery of a kimberlite type of igneous rock in the diamondiferous
fields of Luanda. In Association des Services Géologiques Africains. Deuxième Partie. Questions
Diverses et Annexes. Fascicule XXI. Alger.
De Araujo, A.G., Perevalov, O.V., 1998. Carta de recursos minerais - Mineral Resources Map.
Ministério Geol. e Minas. Inst. Geol. Angola, Luanda.
De Araujo, A.G., Perevalov, O.V., Jukov R.A., 1988. Carta geológica de Angola - Geological Map of
Angola. Scale 1 : 1 000 000. Ministèrio da Indústria. Inst. Nac. Geol., Luanda.
De Carvalho, H., Tassinari, C., Alves, P.H., 2000. Geochronological review of the Precambrian in
western Angola: links with Brazil. J. Afr. Earth Sci. 31 (2), 383-402.
Egorov, K.N., Roman’ko, E.F. Podvysotsky, V.T., Sablukov, S.M, Garanin, V.K., D’yakonov, D.B.,
2007. New data on kimberlite magmatism in southwestern Angola. Russian Geol. Geophys. 48 (4),
323-336.
Ekstrand, O.R., Sinclair, W.D., and Thorpe, R.I., 1995. Geology of Canadian mineral deposit types:
Geological Survey of Canada, Geology of Canada, no. 8, and Geological Society of America's
Geology of North America, vol. P-1, 640 p.
ENDIAMA, 2012. Empresa Nacional de Diamantes de Angola, E.P. Accessed on June.
http://www.endiama.co.ao/endiama_operacoes_grupo.php.
Ganga, J., Rotman, A.Y., Nosiko, S., 2003. Pipe Catoca, an example of the weakly eroded kimberlites
from North-East of Angola. The 8th International Kimberlite Conference, Extended Abstract.
Garanin, V.K., Kudryavtseva, G.P., Posukhova, T.V., 1997. Indicators of diamond preservation in
kimberlite. In: Papunen, H. (Ed.), Mineral Deposits: Research and Exploration, Where Do They
Meet? In: Proceedings of the 4th Biennial SGA Meeting. Turku, Finland, 767–770.
Gaspar, J.C., Wyllie, P.J., 1983. Ilmenite (high magnesium, manganese, niobium) in the carbonatites
from the Jacupiranga Complex, Brazil. Am. Mineral. 68 (9–10), 960–971.
42
Gaspar, J.C., Wyllie, P.J., 1984. The alleged kimberlite-carbonatite relationship: evidence from
ilmenite and spinel from Premier and Wesselton Mines and the Benfontein Sill, South Africa.
Contrib. Mineral. Petrol. 85, 133–140.
Grütter, H.S., Gurney, J.J., Menzies, A.H., Winter, F., 2004. An updated classification scheme for
mantle-derived garnet, for use by diamond explorers. Lithos 77, 841–857.
Grütter, H.S., Latti, D., Menzies, A., 2006. Cr-saturation arrays in concentrate garnet compositions
from 1 kimberlite and their use in mantle barometry. J. Petrol. 47, 801–820.
Guiraud, R., Bosworth, W. Thierry, J. Delplanque A., 2005. Phanerozoic geological evolution of
Northern and Central Africa: an overview. J. Afr. Earth Sci. 43, 83–143.
Guiraud, M., Buta-Neto, A., Quesne, D., 2010. Segmentation and differential post-rift uplift at the
Angola margin as recorded by the transform-rifted Benguela and oblique-to-orthogonal-rifted
Kwanza basins. Mar. Pet. Geol. 27, 1040–1068.
Gurney, J.J., Zweistra, P., 1995. The interpretation of the major element compositions of mantle
minerals in diamond exploration. J. Geochem. Explor. 53 (1–3), 293–309.
Haggerty, S.E., 1986. Diamond genesis in a multiply-constrained model. Nature 320, 34 – 38.
Haggerty, S.E., 1995. Upper mantle mineralogy. J. Geodyn. 20 (4), 331–364.
Haggerty, S.E., Hardie III, R.B., McMahon, B.M., 1977. The composition of ilmenite nodule
associations from the Monastery diatreme: Proceedings of the 2nd International Kimberlite
Conference, vol. 2, pp. 249–256.
Haggerty, S.E., Tompkins, L.A., 1983. Redox state of Earth's upper mantle from kimberlitic
ilmenites. Nature 303 (5915), 295–300.
Hawkesworth, C.J., Gallagher, K., Kelley, S., Mantovani, M., Peate, D.W., Regelous, M., Rogers
N.W., 1992. Parana magmatism and the opening of the South Atlantic, in: Storey, B.C., Alabaster,
43
T., Pankhurst, R.J (Eds.), Magmatism and the causes of continental break-up. Geol. Soc. Spec.
Publ. No. 68, London, 221-240.
Hawkesworth, C.J., Kelley, S., Turner, S., Le Roex, A., Storey, B., 1999. Mantle processes during
Gondwana break-up and dispersal. J. Afr. Earth Sci. 28 (1), 239-261.
Hunter, R.H., Kissling, R.D., Taylor, L.A., 1984. Mid- to late-stage kimberlitic melt evolution:
phlogopites and oxides from the Fayette County kimberlite, Pennsylvania. Am. Mineral. 69, 30–
40.
Head, J.W., Wilson, L., 2008. Integrated model for kimberlite ascent and eruption. 9th International
Kimberlite Conference Extended abstract No. 9IKC-A-00284.
Heaman, L.M., Kjarsgaard, B.A., Creaser, R.A., 2003. The timing of kimberlite magmatism in North
America: implications for global kimberlite genesis and diamond exploration. Lithos 71, 153– 184.
Janse, A.J.A., 2007. Global rough diamond production since 1870. Gems and Gemology,
Gemological Institute of America, vol. 43, No. 2, 98–119.
Janse, A.J.A., Sheahan, P.A., 1995. Catalogue of worldwide diamond and kimberlite occurrences: a
selective and annotative approach. J. Geochem. Explor. 53, 73-111.
Jelsma, H.A., Barnett, W., Richards, S., Lister, G., 2009. Tectonic setting of kimberlites. Lithos 112S,
55-165.
Jelsma, H.A., de Wit, M.J., Thiart, C., Dirks, P.H.G.M., Viola, G., Basson, I.J., Anckar, E., 2004.
Preferential distribution along transcontinental corridors of kimberlites and related rocks of
Southern Africa. S. Afr. J. Geol. 107, 301-324.
Jelsma, H.A., Krishnan, U., Perrit, S., Preston, R., Winter, F., Lemotlo, L., v/d Linde, G., Armstrong,
R., Phillips, D., Joy, S., Costa, J., Facatino, M., Posser, A., Kumar, M., Wallace, C., Chinn, I.,
Henning, A., 2012. Kimberlites from Central Angola: a case study of exploration findings.
Extended Abstr. No. 10IKC-42, 10th International Kimberlite Conference, Bangalore, 1–5.
44
Kamenetsky, M.B., Sobolev, A.V., Kamenetsky, V.S., Maas, R., Danyushevsky, L.V., Thomas, R.,
Pokhilenko, N.P., Sobolev, N.V., 2004. Kimberlite melts rich in alkali chlorides and carbonates: a
potent metasomatic agent in the mantle. Geology 32, 845-848.
Khar'kiv, A., Levin, V., Mankenda, A., Safronov, A., 1992. The Camafuca-Camazambo kimberlite
pipe of Angola, the largest in the world. Int. Geol. Rev. 34 (7), 710-719.
Kotel'nikov, D.D., Zinchuk, N.N., Zhukhlistov, A.P., 2005. Stages of serpentine and phlogopite
transformation in the Catoca kimberlite pipe, Angola. Dokl. Earth Sci. 403 (6), 866–869.
Lewis, H.C., 1987. On diamandiferous peridotite and the genesis of the diamond. Geol. Mag. 4, 22-
24.
Lewis, H.C., 1988. The matrix of diamond. Geol. Mag. 5, 129-131.
Liipo, J.P., Vuollo, J.E., Nykänen, V.M., 1994. Pyrophanite and ilmenite in serpentinized wehrlite
from Ensilä, Kuhmo greenstone belt, Finland. Eur. J. Mineral. 6, 145–150.
Llusià Queral, R.M., Pereira Ginga, S., Galiano, A., Sebastião, P., Melgarejo Draper, J.C., Olimpio
Gonçalves, A., Proenza Fernandez, J., Morais, E., 2005. Aluviões asociados ao campo kimberlítico
da zona de Muanga (Lunda Norte, Angola). XIV semana Geoquímica/VIII Congresso de
Geoquimica dos Paises de Lingua Portuguesa, 351-354.
Mather, K.A., Pearson, D.G., McKenzie, D., Kjarsgaard, B., Priestley, K., 2011. Constraining the
depth and thermal history of cratonic lithosphere using peridotite xenolith and xenocryst
thermobarometry and seismology. Lithos 125, 729–742.
McLean, H., Banas, A., Creighton, S., Whiteford, S., Luth, R.W., Stachel, T., 2007. Garnet xenocrysts
from the Diavik mine, NWT, Canada: Composition, color, and paragenesis. Can. Mineral. 45,
1131–1145.
Mitchell, R.H., 1986. Kimberlites: Mineralogy, Geochemistry, and Petrology. Plenum Press, New
York, pp. 442.
45
Mitchell, R.H., 1995. Kimberlites, Orangeites, and Related Rocks. Plenum Press, New York, pp. 410.
Mücke, A.,Woakes, M., 1986. Pyrophanite: a typical mineral in the Pan-African Province of Western
and Central Nigeria. J. Afr. Earth Sci. 5 (6), 675–689.
Nimis, P., Taylor, W. R., 2000. Single clinopyroxene thermobarometry for garnet peridotites. Part I.
Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in-Cpx thermometer. Contrib.
Mineral. Petrol. 139, 514–554.
Pasteris, J., 1980. The significance of groundmass ilmenite and megacryst ilmenite in kimberlites.
Contrib. Mineral. Petrol. 75 (4), 315–325.
Pearson, D.G., Canil, D., & Shirey, S.B., 2007. Mantle samples included in volcanic rocks: xenoliths
and diamonds. In Carlson, R.W. (ed.). The mantle and core. Treatise on Geochemistry, vol. 2,
Washington, Elsevier, 171-275.
Pearson, D.G., Wittig, N., 2008. Formation of archean continental lithosphere and its diamonds: the
root 15 of the problem. J. Geol. Soc. (London, U. K.)165, 1–20.
Pell, J., 1998. Kimberlite-hosted Diamonds. In Geological Fieldwork (1997), British Columbia
Ministry of Employment and Investment, Paper 1998-1, pages 24L-1 to 24L-4.
Pereira, E., Rodrigues, J., Reis B., 2003. Synopsis of Lunda geology, NE Angola: implications for
diamond exploration. Comun. Inst. Geol. Min. 90, 189-212.
Perevalov, O.V., Voinovsky, A.S., Tselikovsky, A.F., Agueev, Y.L., Polskoi, F.R., Khódirev, V.L.,
Kondrátiev, A.I., 1992. Geologia de Angola. Notícia explicativa da carta geológica a escala
1:1.000.000. Serviço Geológico de Angola. Lunda.
Preistley, K., McKenzie, D., 2006. The thermal structure of the lithosphere from shear wave
velocities. Earth Planet. Sci. Lett. 244, 285–301.
Read, G.H., Janse, A.J.A., 2009. Diamonds: Exploration, mines and marketing. Lithos 112S, 1–9.
46
Reis, B., 1972. Preliminary note on the distribution and tectonic control of kimberlites in Angola.
Abstr., 24th Int. Geol. Congr., Rep. Sess. 4, 276-281, Montreal.
Robles-Cruz, S.E., 2007. Textura y química de los minerales del grupo de la ilmenita de la kimberlita
de Catoca (Angola). Treball de Recerca Tutelat (TRT). Diploma d'Estudis Avanzats (DEA).
Universitat de Barcelona. Unpublished.
Schulze, D.J., 1984. Cr-poor megacrysts from the Hamilton Branch kimberlite, Elliott County,
Kentucky. In Kornprobst, J. (Ed.), Kimberlites II: TheMantle and Crust–Mantle Relationships. In:
Developments in Petrology 11B. Amsterdam, Elsevier, 97–108.
Schulze, D.J., 1995. Low-Ca garnet harzburgites from Kimberley, South Africa: Abundance and
bearing on the structure and evolution of the lithosphere. J. Geophys. Res. (Solid Earth), vol. 100,
No. B7, 12,513-12,526.
Scott Smith, B. H.; Nowicki, T. E.; Russell, J. K.; Webb, K. J.; Mitchell, R. H.; Hetman, C. M.;
Robey, J. V. A.; Skinner, E. M. W.; Robey, J. V., 2012. Kimberlite terminology and classification.
In Proceedings of the 10th Kimberlite Conf., Bangalore, India, 6-11 February, Abstract Volume,
30–31.
Skinner, E.M., 1986. Constrasting Group 1 and 2 kimberlite petrology: Towards a genetic model for
kimberlites. Fourth Int. Kimberlite Conf., Perth, Australia. Extended Abstracts, 202-204.
Skinner, E.M., 1989. Contrasting group I and II kimberlite petrology: Towards a genetic model for
kimberlites. In Ross et al. (1989) q.v., vol. 1, pp. 528-544.
Smith, C.B., 1983. Pb, Sr, and Nd isotopic evidence for sources of African Cretaceous kimberlites.
Nature 304, 51-54.
Smith, C.B., Gurney, J.J., Skinner, E.M.W., Clement, C.R., Ebrahim, N., 1985. Geochemical
character of southern African kimberlites: a new approach based on isotopic constraints. Trans.
Geol. Soc. S. Afr. 88, 267-280.
47
Smith, C.B., Sims, K., Chimuka, L., Duffin, A., Beard, A.D., Townend, R., 2004. Kimberlite
metasomatism at Murowa and Sese pipes, Zimbabwe. Lithos 76 (1–4), 219–232.
Stachel, T., Harris, J.W., 2009. Formation of diamond in the Earth’s mantle. J. Phys.: Condens.
Matter 21, 364206, 1-10.
Tompkins, L.A., Haggerty, S.E., 1985. Groundmass oxide minerals in the Koidu kimberlite dikes,
Sierra Leone, West Africa. Contrib. Mineral. Petrol. 91 (3), 245–263.
Thomson Reuters, 2012. Journal Citation Reports. Available online: http://admin-
apps.webofknowledge.com/JCR/JCR (accessed on 24 September).
Van Straaten, B.I., Kopylova, M.G., Russell, J.K., Webb, K.J., Scott Smith, B.H., 2008.
Discrimination of diamond resource and non-resource domains in the Victor North pyroclastic
kimberlite, Canada. J. Volcanol. Geotherm. Res. 17, 128–138.
White, S.H., de Boorder, H., Smith, C.B., 1995. Structural controls of kimberlite and lamproite
emplacement. J. Geochem. Explor. 53 (1-3), 245-264.
48
ORIGINAL PUBLICATIONS
PAPER I
Reprinted from MACLA - Revista de la Sociedad Española de Mineralogía, September No. 9. Robles- Cruz, S., Watangua, M., Melgarejo, J.C., Galí, S., 2008. New Insights into the Concept of Ilmenite as an Indicator for Diamond Exploration, Based on Kimberlite Petrographic Analysis.
macla nº 9. septiembre ‘08 revista de la sociedad española de mineralogía 205
New Insights into the Concept of Ilmenite as an Indicator for Diamond Exploration, Based on Kimberlite Petrographic Analysis / SANDRA ROBLES CRUZ (1*), MANUEL WATANGUA (2), JOAN CARLES MELGAREJO (1), SALVADOR GALI (1)
(1) Departament de Cristal·lografia, Mineralogia i Dipòsits Minerals. Facultat de Geologia. Universitat de Barcelona. Martí i Franquès s/n. 08028, Barcelona (España) (2) ENDIAMA, Major Kanhangulo 100, Luanda (Angola)
INTRODUCTION. diamond potential is currently being studied. The Catoca kimberlite is the This study presents results of the initial most important primary diamond phase of the research project, deposit in Angola, hosted by “Kimberlites associated to the Lucapa Precambrian rocks and covered by structure, Angola (Africa)”, within the Mesozoic-Cenozoic sedimentary deposits framework of a multilateral agreement (Janse et al, 1995). between the Faculty of Geology- Universitat de Barcelona, the Empresa PETROGRAPHY. Nacional de Diamantes de Angola and the Agostinho Neto University (Luanda- There are some minerals which are Angola). frequently associated to diamond inside kimberlites and they are used as fig 1. Intercumular ilmenite in peridotitic xenolith. The research is based on two sets of indicator minerals for the diamond core sampling down to 600 m deep. The exploration. The main indicator minerals first set comes from Catoca pipe and are: magnesian ilmenite (Pell, 1998), allowed us to identify complete crater garnet and chromite (Wyatt et al, 2004). and diatreme facies. The second one (18 However, for this instance we will focus kimberlites) comes from Cucumbi, on ilmenite since it is the first mineral Cacuilo, Tchiuzo, Alto Cuilo, Camitongo analyzed in 2007. and Kambundu, whose samples were gathered during fall 2008. Currently, we Diverse xenoliths, comprising lherzolite, are working on these sets of samples. eclogite, harzburgite, carbonatite, gneiss and amphibolite are distributed through The kimberlites are ultrabasic rocks with the Catoca and Cucumbi kimberlites. high content of volatiles mainly CO2, and Some shales and sandstones can be a typical inequigranular texture present in the upper part of this fig 2. Nodular xenocrysts of ilmenite with characterized by the presence of macro- Kimberlite. Accessory minerals and replacement. megacrysts which can be xenoliths or xenocrysts comprise garnet, zircon, Cr- xenocrysts embedded in a fine-grained rich diopside, amphibole, phlogopite, matrix (Mitchell, 1995; Benvie, 2007). chromite and several generations of These special rocks have a great ilmenite. Secondary minerals include importance, not only in scientific terms serpentine-group minerals being the since they add valuable information most abundant, calcite, barite, about lithospheric mantle but also barytocalcite, witherite and strontianite. because they can contain diamond. Based on optical petrographic studies REGIONAL SETTING. and BSE images from SEM-ESEM with EDS microanalysis, we have been able
The area of interest is localized in to discriminate up to six textural types of fig 3. Euhedral crystals of ilmenite in matrix. northeastern Angola (Africa), being ilmenite in Catoca and Cucumbi tectonically controlled by the Lucapa kimberlite: a) intercumular ilmenite in Zircon xenocrysts are partially replaced structure, a former rift (Guiraud et al., peridotitic xenoliths (Fig 1); b) anhedral by fine-grained baddeleyite, and at least 2005) of early Cretaceous that extends ilmenite in carbonatite xenoliths; c) two populations exist according to the NE-SW across Angola. Associated to this ilmenite unaltered megacrysts; d) trace element distribution. All of these structure there is a magmatic belt, nodular xenocryst of ilmenite with crystals are enriched in HREE, but with a which is composed by kimberlites different grades of replacement, some noticeable positive Ce anomaly, similar toward NE and carbonatites toward SW. of them with symplectitic textures (Fig. to that reported in zircon in a MARID At present, over 2000 kimberlites have 2); e) skeletal ilmenite; and f) euhedral xenolith from a southern African been identified in this structure and their crystals of ilmenite in matrix (Fig. 3). kimberlite (Dawson et al., 2001). The
palabras clave: kimberlita, diamante, ilmenita, manto, Angola. key words: kimberlite, diamond, ilmenite, mantle, Angola.
resumen SEM/SEA 2008 * corresponding author: [email protected] 206
crystals are not optically zoned, but environment. Magnesian ilmenite is also ilmenite in a higher or lower grade. there is a slight depletion in REE from enriched in Cr and Ni. More advanced These fluids are reducing, especially the core to the rim. replacement produces a symplectitic those rich in Mn. Picroilmenite has replacement of ilmenite II by ilmenite III. traditionally been interpreted as an MINERAL CHEMISTRY OF ILMENITE. indicator of kimberlite associations, as well as an indicator of low fO2, which is Every texture has been systematically necessary for the preservation of analyzed with EPMA which allowed us to diamond. Although Catoca and Cucumbi identify three compositional types of are diamondiferous kimberlites, they ilmenite (I, II and III). This combined show that Mg ilmenite is clearly a late technique –texture and composition replacement product, and the grade of analysis- has been suitable for analyzing replacement of the primary grains is zircon and garnet as well. very variable. Therefore the absence of magnesian ilmenite in a kimberlite does The primary ilmenite (type I) in not appear to be a convincing argument megacrysts (xenocrysts) is generally rich to exclude the presence of diamonds. in Cr and Fe3+. Its composition is similar Accordingly, this work proposes a new to the intercumulus crystals in peridotite insight into the concept of ilmenite in xenoliths. This ilmenite is replaced, in diamond exploration. the first instance, by magnesian ilmenite (type II). This process takes place along ACKNOWLEDGMENTS. microdiscontinuities (cleavage, border grains, contour subgranes, kink band This research is supported by the project planes, etc.), producing diffusive CGL2006-12973 of Ministerio de replacements. In a more advanced Educación y Ciencia (Spain), the AGAUR stage, symplectitic replacements occur, fig 4. MgTiO3-FeTiO3-Fe2O3 ratio for the different SGR 589 of Generalitat de Catalunya involving an early generation of types of ilmenite (I, II and III) discriminated for each and a FI grant sponsored by the texture. magnesian ilmenite (type II) at the Departament d’Educació i Universitats expense of Fe3+-rich primary ilmenite A late generation of ilmenite III (Mn-rich de la Generalitat de Catalunya i del Fons (from texture a to d ). A late generation ilmenite) is found rimming all the above Social Europeu. The authors of Mn-Nb-Zr-rich ilmenite (type III) cuts mentioned generations, and is strongly acknowledge the Serveis the previous ones. enriched in Nb, Ta, Zr, W, Hf, Th and U, Cientificotècnics de la Universitat de and poor in Mg and Fe3+. The Barcelona. Contrastingly, the late euhedral Mn-Nb- composition of this ilmenite is similar to Zr ilmenite crystals found in the that of the fine-grained euhedral REFERENCES. kimberlite matrix do not present any ilmenite crystals found in the kimberlitic evidence of replacement. This ilmenite matrix and also to that of the ilmenite Benvie, B.(2007): Mineralogical imaging of is poor in Mg and Fe3+. Their kimberlites using SEM-based techniques. crystals found in the carbonatitic Minerals Engineering, 20, 435-443. compositions are identical with the Mn- xenoliths. Crystals of Mn-rich ilmenite rich ilmenite produced during late Dawson, J.B., Hill, P.G., Kinny, P.D. (2001): (ilmenite type III) are not replaced or Mineral chemistry of a zircon-bearing, replacement stages of ilmenite zoned, and seem to have crystallized in composite, veined and metasomatised megacrysts. Compositions of Mn-rich equilibrium with the kimberlitic magma. upper-mantle peridotite xenolith from ilmenite are similar to those found in Both the late generations of ilmenite kimberlite. Contrib. Mineral Petrol., 140, carbonatite xenoliths. and the baddeleyite replacing zircon can 720-733. be produced by interaction of a Guiraud, R., Bosworth, W., Thierry, J., DISCUSSION AND CONCLUSIONS. Delplanque, A. (2005): Phanerozoic carbonate-bearing kimberlitic magma geological evolution of Northern and enriched in Mn and HFSE. The Textural evidences indicate a different Central Africa: An overview. Journal of replacement of Fe3+-rich ilmenite by Mg- African Earth Sciences, 43, 83-143. complex history of growth in the and Mn-rich ilmenite implies that the Janse, A.J.A. & Sheahan, P.A. (1995): xenocrysts. Unaltered megacryst early ilmenite was formed under Catalogue of world wide diamond and ilmenite (ilmenite type I) rich in Fe3+ (fig. oxidizing conditions in the mantle, and kimberlite occurrences: a selective and 4), indicates crystallization under high the lastest compositions of ilmenite annotative approach. Journal of fO2 conditions; this ilmenite contains Nb, were produced by reaction with the Geochemical Exploration, 53, 73-111. Cr, Ni and Ta in low contents. Its Mitchell, R.H. (1995): Kimberlites, Orangeites, kimberlitic magma. and related rocks. New York, Plenum composition is similar to those ilmenite Intercumular megacrysts that occur in Press, 410 pp. Megacrysts of ilmenite are frequently Pell, J. (1998): Kimberlite-hosted Diamonds, peridotite xenoliths. Hence, most of the present in diamondiferous kimberlites, in Geological Fieldwork 1997. British ilmenite xenocrysts seem to have been contrasting with ilmenite observed in Columbia Ministry of Employment and produced by disaggregation of mantle barren kimberlites. This might become a Investment 1998-1, 24L1-24L4. xenoliths. Ilmenite I is replaced along new guide in diamond exploration. Wyatt, B.A., Mike, B., Anckar, E., Grutter, H. discontinuities by magnesian ilmenite In conclusion, the composition of this (2004): Compositional classification of “kimberlitic” and “non-kimberlitic” (ilmenite type II); the elemental ilmenite is the result of a set of ilmenite. Lithos, 77, 819-840. distribution of Mg in these grains points replacement processes with rich fluids in to processes of replacement through Mg and Mn affecting an oxidized primary solid-state diffusion in a typical reducing ORIGINAL PUBLICATIONS
PAPER II
Reprinted from Lithos, 112S. Robles-Cruz, S.E., Watangua, M., Melgarejo, J.C., Gali, S., Olimpio, A., 2009. Contrasting compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in exploration for diamond.
Lithos 112S (2009) 966–975
Contents lists available at ScienceDirect
Lithos
journal homepage: www.elsevier.com/locate/lithos
Contrasting compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in exploration for diamond
Sandra E. Robles-Cruz a,⁎, Manuel Watangua b, Leonardo Isidoro b, Joan C. Melgarejo a, Salvador Galí a, Antonio Olimpio c a Departament de Cristal·lografia, Mineralogia i Dipòsits Minerals, Facultat de Geologia, Universitat de Barcelona, Martí i Franquès, s/n, E-08028, Barcelona, Spain b ENDIAMA, Major Kanhangulo, 100, Luanda, Angola c Departamento de Geologia, Faculdade de Ciências, Universidade Agostinho Neto, Av. 4 de Fevereiro 7, 815, Luanda, Angola article info abstract
Article history: The Catoca group-I kimberlite, the only currently active diamond-producing mine in Angola, was emplaced in Received 2 October 2008 the northeastern part of the Lucapa structure. We focus here on compositional and textural variations in Accepted 16 May 2009 ilmenite from drill-core material, in the hope of elucidating events before and during the emplacement of the Available online 25 June 2009 kimberlitic magma. We have characterized four main variants of ilmenite, with enrichments in Fe3+, Mg, Mn and nearly stoichiometric ilmenite, and in seven textural classes, and have distinguished crystals of variable Keywords: Ilmenite size, ranging from micro- to megacrysts. Most ilmenite is found to derive, through a complex process, from 3+ Kimberlite replacement of Fe -rich ilmenite, presumably originating by mantle metasomatism at a relatively high fO2. 3+ Diamond This Fe -rich ilmenite reacted with fluids under reducing conditions, producing Mg-rich ilmenite. The Mn- Fluid rich ilmenite is produced by interaction with a late CO2-rich fluid. The Mg-rich ilmenite is here clearly a Texture minor phase and a late product of replacement. The absence of fresh Mg-rich ilmenite and the occurrence of Composition Fe3+-rich ilmenite do not seem to be convincing arguments to exclude the presence of diamond crystals in a kimberlite. Compositional attributes must thus be considered with caution, and only in light of textural studies, in exploration programs. © 2009 Elsevier B.V. All rights reserved.
1. Introduction expected to preserve diamond, whereas the presence of Fe3+-rich ilmenite could indicate oxidizing processes that could destroy crystals of Angola has become an important producer of diamond (Janse and diamond (Gurney et al., 1993; Gurney and Zweistra, 1995; Kostrovitsky Sheahan, 1995; Read and Janse, this issue), with a significant part of the et al., 2004, 2006; van Straaten et al., 2008). Other compositional production being obtained from the Catoca kimberlite, in Lunda Sul features of ilmenite are of more controversial origin. In particular, Mn- province, in northeastern Angola. Catoca, currently the unique active rich compositions, found in previous surveys of some Angolan kimberlite mine in Angola, is located in the Lucapa structure, a system of kimberlites (Llusià Queral et al., 2005a,b; Rogers and Grütter, this Cretaceous extensional faults trending NE–SW (Reis, 1972; De Carvalho issue) have been attributed to supergene processes; others propose et al., 2000; Guiraud et al., 2005). magmatic crystallization under reducing conditions (Hwang et al.,1994) For many years, the composition of ilmenite has been stressed as an or metasomatic processes in the mantle (Meyer and McCallum,1986). In exploration guide for diamondiferous kimberlites and placers (e.g., this investigation, we seek explanations of the real significance of Mitchell, 1989, 1995, 1997; Wyatt et al., 2004). It has been correlated as compositional and textural variations of ilmenite in kimberlites. We well with conditions in the mantle where the kimberlitic magmas used back-scattered electron (BSE) petrography with microanalysis originated (Haggerty and Tompkins,1983; Arculus et al.,1984; Haggerty, using energy-dispersion spectroscopy (EDS), quantitative powder X-ray 1991a,b; Gurney et al., 1993). Griffin and Ryan (1995) have suggested diffraction (PXRD), and quantitative chemical analyses using an that with some patterns of compositional evolution in ilmenite electron-microprobe (EMP) on a suite of 81 representative thin sections megacrysts, once can assess the occurrence of mechanisms of fractional and 19 probes from two core samples of Catoca pipe. With these new crystallization of single batches of magma associated with extensive datasets, we are able to shed new light on the origin of these unusual metasomatic alteration of the wallrocks, and hence destruction of compositions. diamond. Thus, reduced kimberlites bearing Mg-rich ilmenite would be 2. The Catoca kimberlite
2 ⁎ Corresponding author. Tel.: +34 9340 21344; fax: +34 9340 21340. The Catoca pipe outcrop, 639000 m , is found 30 km NNW of E-mail address: [email protected] (S.E. Robles-Cruz). Saurimo, the capital of Lunda Sul. Catoca can be classified as a group-I
0024-4937/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.lithos.2009.05.040 S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975 967
Fig. 1. Cross section of the Catoca kimberlite (adapted from Kriuchkov et al., 2000).
kimberlite (Mitchell, 1995). Complete crater and diatreme facies are microprobe analysis, allowed us to discriminate four main variants recognized (Ganga et al., 2003; Fig. 1), according to the classification of ilmenite based on compositional attributes: a) Fe3+-rich ilmenite, criteria of Clement and Skinner (1985),asmodified by Scott Smith et al. b) Mg-rich ilmenite, c) Mn-rich ilmenite, and d) near-ideal ilmenite 2+ 4 (2008), thus indicating the minimal extent of erosion of the kimberlite. (Fe Ti O3). These types can be easily distinguished using BSE images, Crater facies are found up to 230–270 m in depth, and are composed as the Mg-rich ilmenite displays the darkest shades, and the Mn-rich in the uppermost part by epiclastic sandstones with cross-stratification ilmenite is the lightest. On the other hand, up to seven textural classes of in the central part, and coarse debris rimming the crater. Altered crystals ilmenite have been established, based on their paragenetic position and of garnet and diopside may occur as accessory minerals in these degree of replacement: 1) intercumulus Fe3+-rich ilmenite grains in sediments. Ilmenite is rare in this unit. Most of the cement is ferruginous, metasomatized peridotite xenoliths, 2) anhedral ilmenite in carbonatite but in some areas, the sandstone has a calcareous-ferruginous cement. xenoliths; 3) homogeneous Fe3+-rich ilmenite present as macro- and In the lower part of the sequence, the content of volcaniclastic material megacrysts; 4) partially replaced ilmenite macro- and megacrysts; 5) increases, and the lower half of the crater facies become dominated by symplectitic ilmenite xenocrysts; 6) skeletal ilmenite crystals in a resedimented volcaniclastic kimberlite facies (unit RVK). pelletal matrix; 7) tabular Mn-rich ilmenite crystals in a kimberlite Below the crater facies, Ganga et al. (2003) described classic matrix. The suite of ilmenite that we examined contains grains of tuffisitic kimberlite facies (TK; Mitchell et al., this issue); in the variable size: microcrystals have 5–20 μm in diameter, being some as current terminology, this category could be named massive volcani- large as 50 μm; most macrocrysts have dimensions between 1 and clastic kimberlite (Sparks et al., 2006) or Wesselton-type volcaniclas- 10 mm; megacrysts are very rare and may exceed 2 cm. tic kimberlite (Scott Smith et al., 2008). Ganga et al. (2003) also The distribution of these ilmenite textural types is not homogeneous divided this unit into different subunits based on the size of the in the crater and diatreme kimberlite facies. Homogeneous ilmenite fragments, and pointed out the occurrence of abundant xenoliths macro- and megacrysts are found in all the kimberlite facies, but the derived from the host rocks, and the scarcity of xenoliths from the remainder of the textural variants are restricted to the diatreme facies, mantle. The extensive drilling allowed us to sample these units down mainly in the volcaniclastic kimberlite (below 250 m in depth). Partly to 609 m in depth. replaced megacrysts occur in the uppermost part of the diatreme facies, The diatreme facies are strongly altered all along the profile; olivine and strongly corroded macro- and megacrysts (in particular, those with is completely replaced by serpentine, calcite and saponite (Kotel'nikov a symplectitic texture) are only found below 350 m in depth. et al., 2005). Xenoliths of the host rocks are common, and comprise gneiss, amphibolite, granite, sandstone and shale. Mantle xenoliths are 3.1. Intercumulus grains of Fe3+-rich ilmenite in metasomatized also quite altered and include mainly metasomatic peridotite, and rarely peridotitic xenoliths eclogite as well as xenoliths of carbonatite. Xenocrysts encountered in the diatreme facies comprise G9 and G10 varieties of garnet according to The intercumulus grains of Fe3+-rich ilmenite are anhedral, 300– the classification of Grütter et al. (2004),zircon(partlyreplacedby 800 μm in diameter, and interstitially distributed among roundish baddeleyite), chromian diopside, amphibole, phlogopite and ilmenite. grains of olivine (Fig. 2A). The olivine is completely replaced by The matrix of the kimberlite contains lizardite, apatite, calcite, ilmenite serpentine, but the mesh texture typical of serpentinized olivine is and chromite; titanite, zirconolite, baddeleyite, barite, dolomite, with- clearly recognizable. The ilmenite is polycrystalline, and grains show erite, barytocalcite, strontianite, sulfides, identified by PXRD and EMP, polygonally annealed textures with curved borders and triple points. and minor minerals are widespread in the matrix and fill small veinlets. Similar polygonal textures in ilmenite in kimberlitic suites have been interpreted as formed by intense annealing of stressed ilmenite 3. Petrography of ilmenite (Mitchell, 1973; Haggerty et al., 1977; Tompkins and Haggerty, 1985). The intercumulus ilmenite is quite homogeneous in composition and Optical petrographic studies and back-scattered electron (BSE) rich in Fe3+ (as inferred from stoichiometry: see below), although it images taken with a SEM-ESEM with EDS microanalysis, coupled with may be partly replaced by Mg-rich ilmenite. 968 S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975
3.3. Homogeneous macrocrysts of Fe3+-rich ilmenite
These macrocrysts (0.2–2 cm across) are generally rounded with smooth surfaces, and they can be mono- or polycrystalline. The angular
Fig. 2. Primary grains of ilmenite. SEM image, mode BSE. (A) Intercumulus Fe3+-rich ilmenite (brighter) in peridotite xenolith. (B) Anhedral Mn-rich in carbonatite xenolith. (C) Ilmenite macrocryst without visible signs of replacement. Ilmenite (Ilm), phlogopite (Phl), zircon (Zrn), calcite (Cal), apatite (Ap).
3.2. Anhedral grains of ilmenite in carbonatite xenoliths Fig. 3. Partly replaced polycrystalline nodules of ilmenite. SEM image, mode BSE. Carbonatite xenoliths are rare (less than 7% of grains) in Catoca, and (A) Polycrystalline nodule of ilmenite (lightest) showing corrosion on the uppermost accessory ilmenite in them occurs as small crystals (50–90 μm) found side and replacement to Mg-rich ilmenite (slightly darker) along the subgrain borders. (B) Polycrystalline nodule of Fe3+-rich ilmenite corroded on one side and replaced along as inclusions in phlogopite crystals, or intergrown with calcite, apatite, subgrain borders and small cracks by Mg-rich ilmenite (darker) and Mn-rich ilmenite zircon and phlogopite (Fig. 2B). The ilmenite is usually rimmed at grain (pale gray to white). (C) Detail of a polycrystalline nodule of Fe3+-rich ilmenite. Ilmenite margins by Mn-rich ilmenite. (Ilm), serpentine (Srp). S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975 969 shape of monocrystalline grains is consistent with an origin by disag- they show an intense replacement, leading to a symplectitic texture; this gregation of polycrystalline grains (Fig. 2C). Homogeneous macrocrysts type of replacement is developed only at the border of the crystal of ilmenite or Mg-rich ilmenite are widely represented in kimberlites (Fig. 4A) or it can affect the whole crystal (Fig. 4B). The other mineral (i.e., Mitchell, 1973), but in the Catoca kimberlite, this textural type is originally present in these intergrowths has been fully replaced by very rare (less than 5% of grains). serpentine, but similar unreplaced textures in many kimberlites consist of pyroxene and ilmenite (i.e., Haggerty et al., 1977). These grains have 3.4. Partly replaced macro- and megacrysts of Fe3+-rich ilmenite complex patterns of replacement, with ilmenite strongly replaced by Mg-rich ilmenite; the zoned grains are finally overgrown by Mn-rich These coarse crystals (0.4–2 cm) are also rounded and they occur ilmenite (Fig. 4C). Manganese-rich ilmenite forms only a thin rim or usually in the core of pelletal lapilli; they present different grades of veinlet, and it is accompanied by small grains of barytocalcite, replacement (Fig. 3A, B, C). These megacrysts consist mainly of Fe3+- strontianite and baddeleyite. Most of the symplectitic replacement rich ilmenite similar in shape and composition to the last category. takes place in crystal discontinuities and is adapted to pre-existing However, they have been corroded and replaced by Mg-rich ilmenite features as deformation-induced kink bands or subgrains. along old surfaces and discontinuities. Some crystals may display freshly fractured borders, thus indicating that corrosion and replace- 3.6. Skeletal crystals of ilmenite in a pelletal matrix ment took place before the explosive processes in the kimberlite. These small crystals (80–150 μm) may be present in pelletal lapilli 3.5. Macrocrysts of symplectitic ilmenite and exhibit very irregular shapes (Fig. 4D). The core of the crystals is generally constituted by ilmenite, which can be partially replaced by Macrocrysts of Fe3+-rich ilmenite from the deepest parts of diatreme Mn-rich ilmenite. It is the least common variety of ilmenite in the Catoca facies are similar in shape and composition to two previous types, but kimberlite, and is found only in the uppermost part of the diatreme.
Fig. 4. Advanced replacements of ilmenite. SEM image, mode BSE. (A) An elongate crystal of Fe3+-rich ilmenite affected by kink-band deformation, partly replaced in both sides by Mg-rich ilmenite (slightly darker). It is possible to have symplectitic intergrown of three kinds of ilmenite by a replacement mechanism. The symplectitic intergrowth here is not primary. Intergrowths of ilmenite have previously been described in literature (Tompkins and Haggerty, 1984; Haggerty and Tompkins, 1984; Kostrovitsky and Piskunova, 1990; Haggerty, 1995). (B) Another crystal of Fe3+-rich ilmenite almost completely replaced by Mg-rich ilmenite. This intergrowth is probably of the same origin as in 4A. The gray domains are made of ilmenite–geikielite solid solution. (C) Detail of a symplectitic intergrown of ilmenite. Note that some grains of Mn-rich ilmenite are euhedral. Accompanying minerals include serpentine (Srp), calcite (Cal), witherite (Wth). Note the sharp contact between Mg-rich ilmenite and Mn-rich ilmenite. (D) Skeletal crystals of ilmenite showing similar replacements as in the symplectitic intergrowth. Ilmenite (Ilm), witherite (Wth). 970 S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975
3.7. Tabular crystals of Mn-rich ilmenite in the kimberlite matrix in Mg between Fe3+-rich ilmenite and Mg-rich ilmenite. As expected, there is a clear negative correlation between Ti and Fe3+, and the lowest Tabular crystals of Mn-rich ilmenite (1–10 μm in length), slightly Nb- values in Fe3+ are found in Mn-rich ilmenite, which produces a tight rich, are set in randomly oriented groups in the kimberlite groundmass group (Fig. 6B). However, there is a continuous trend among composi- (Fig. 5), and may be associated with euhedral grains of apatite and tions of Fe3+-rich ilmenite and Mg-rich ilmenite. The level of Fe2+ is chromite. They are internally homogeneous and unaltered, reflecting higher in Mg-rich ilmenite than in Fe3+-rich ilmenite, but it also tends to equilibrium with the kimberlite matrix, and are found only in the decrease in Mn-rich ilmenite owing to the substitution of Fe2+ for Mn2+ diatreme facies. Although rare, members of the ilmenite group have (Fig. 6C). The contents of Cr are quite variable, but they are higher in Mg- been found in the groundmass of other kimberlite pipes, either Mn- rich ilmenite and lower in Mn-rich ilmenite; they show a rough positive enriched and slightly Nb-enriched (Chakhmouradian and Mitchell, correlation with Mg. On the other hand, Mn and Fe3+ very clearly show 1999) or Mg-rich (Nielsen and Sand, 2008). an antithetic behaviour (Fig. 6D), which can be described as a trend of Fe3+ decrease (at low levels of Mn) followed by a trend of Mn increase 4. Composition of ilmenite (without Fe3+). The Zr contents are quite low in all the types of ilmenite. A negative correlation between Fe3+ and Mg can be seen, with a Representative grains of ilmenite from every class of texture were progressive increase in Mg from Fe3+-rich ilmenite toward Mg-rich selected using as reference the SEM-BSE images, and then analyzed ilmenite (Fig. 6E). Finally, the niobium content tends to increase where with a Cameca SX-50 microprobe, with four wavelength-dispersion the Mn content increases in Mn-rich ilmenite (Fig. 6F), as described in spectrometers. All ilmenite crystals were analyzed with an excitation other kimberlite fields (i.e., Chakhmouradian and Mitchell, 1999), voltage of 20 keV, beam current of 20.1 nA and a take-off angle of 40°. although the highest values are found in ilmenite from carbonatite We used the following standards, crystals and lines; periclase (Mg, xenoliths. Groundmass ilmenite is compositionally similar to the outer TAP Kα), orthoclase (Al, TAP Kα), diopside (Si, TAP Kα), wollastonite rim of ilmenite megacrysts.
(Ca, PET Kα), rutile (Ti, PET Kα), synthetic Cr2O3 (Cr, PET Kα), rhodonite (Mn, LIF Kα), hematite (Fe, LIF Kα), synthetic NiO (Ni, LIF 5. Discussion
Kα), synthetic ZrO2 (Zr, PET Lα), and metallic Nb (Nb, PET Mα). The ratio Fe2+/Fe3+ is calculated by stoichiometry. Ilmenite textures and composition are diverse in the Catoca At this phase of the research, we analyzed with EMP about 400 kimberlite, thus suggesting a complex history for the ilmenite points from 40 grains of 20 representative samples of a suite of 100 nodules. The diversity in textures and composition reflects primarily core samples (81 thin sections and 19 probes) from two boreholes of to the paragenetic position of ilmenite in the kimberlite (accessory Catoca Kimberlite. Datasets totalling less than 95% after charges were in xenoliths, macro- and megacrysts, matrix) and replacement balanced were rejected. Analyses were made along profiles in order to processes. evaluate progressive changes among the types of ilmenite. Ilmenite has been described in many kimberlites worldwide as an Correlations among the major and minor components in ilmenite of accessory mineral in many metasomatized peridotite xenoliths. Some the different types are shown in Fig. 6; selected results of electron- widespread examples comprise the MARID suite (Dawson and Smith, microprobe analyses from Catoca are given in Table 1 under the fol- 1977), MORID veins (i.e., Jones et al., 1982) and some metasomatized lowing headings: (1) intercumulus in peridotite xenolith, (2) anhedral garnet-bearing peridotites (Stiefenhofer et al., 1997; Kopylova et al., in carbonatite xenolith, (3) homogeneous macrocryst, (4) partly re- 1999). Some of the Catoca xenoliths can be included in the last placed macro- and megacryst, (5) symplectitic macrocryst, (6) skeletal category, and others have similarities with the ilmenite-bearing crystal, and (7) matrix. As can be seen in Fig. 6A, the Mn-rich ilmenite dunite xenoliths described by Kaminsky et al. (2002). plots outside the classic kimberlite domain in the TiO2–MgO plot of The macro- and megacrysts in kimberlites have been interpreted Wyattetal.(2004); moreover, there is a continuous trend of enrichment worldwide either as xenocrysts (Armstrong et al., 2004; Hearn, 2004)or as produced by primary magmatic crystallization (Moore,1987). At least at Catoca, the similarity in composition of ilmenite in the unreplaced parts of all the macro- and megacrysts and ilmenite from intercumulus positions in peridotite xenoliths suggest that the most if not all of the ilmenite nodules are produced by disaggregation of ilmenite-bearing metasomatized peridotite xenoliths. On the other hand, the composition of this early ilmenite is unusual because of its high Fe3+ contents. Similar Fe3+-rich compositions, although rare in kimberlites, have also been found in the Koidu kimberlite, in Sierra Leone (Tompkins and Haggerty, 1985), but at Catoca, the ilmenite is even more strongly oxidized,
indicating crystallization under relatively high fO2 conditions (Fig. 7A); this ilmenite also contains Nb, Cr and Ni in low contents. The second feature is that ilmenite is replaced along small discontinuities, both in the grain borders and along internal surfaces (cracks, twin planes, cleavages, kink bands), by Mg-rich ilmenite. The replacement of the Fe3+-rich ilmenite by Fe2+- and Mg-rich ilmenite is indicative of a trend toward more reducing conditions (Haggerty and Tompkins, 1983; Fig. 7A). This type of sequence is similar to the so-called ilmenite magmatic trend (Haggerty et al., 1977; Pasteris, 1980; Schulze, 1984). However, the textural patterns attributable to replacement at Catoca, along grain borders, cracks or other disconti- nuities, strongly suggest the action of a fluid rather than a magma. It is difficult to ascertain the timing and place of this replacement. Certainly it was produced before kimberlite emplacement, because Fig. 5. Euhedral platy crystals of Mn-rich ilmenite in matrix. Serpentine (Srp). The some nodules broken during the explosive processes are not replaced crystals are in equilibrium with the kimberlite matrix in the diatreme facies. in the broken corners. S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975 971
Fig. 6. Correlation between major elements or between major and minor elements for the main textural types of ilmenite. Atoms per formula unit (apfu). 972
Table 1 Chemical analyses of ilmenite crystals from Catoca.
Texture 1 1 1 2 3 3 3 4 4 4 5 5 5 6 6 6 7 7 7 Ilm type I I II IV I I I I II II I II III I II III III III III Borehole 535 535 535 535 535 535 535 535 535 535 535 535 535 535 535 535 535 535 535 Depth (m) 350 350 350 451 409 409 409 350 350 451 350 451 350 350 350 350 350 350 350 Point 26Bsn68 26Bsn67 26Bsn73 36f57 409a32b 40932b21 40932c21 26Bn74 26Ai63 36g59 26Aj17 36a13 26Aj24 26Aq32 26Ae39 26Ae34 26Aa12 26Aa15 26An86 (wt.%) SiO2 0.00 0.04 0.04 0.33 0.04 0.00 0.03 0.00 1.26 0.03 0.05 0.05 0.11 0.05 0.05 0.17 0.43 1.00 0.18 TiO2 43.88 44.06 50.67 46.54 36.78 36.49 36.65 38.69 51.64 54.30 47.25 50.10 51.30 42.13 46.12 51.36 48.79 47.86 49.79 Al2O3 0.19 0.21 0.23 0.02 0.12 0.04 0.14 0.12 0.18 0.16 0.14 0.07 0.05 0.15 0.12 0.02 0.05 0.16 0.06 ..Rbe-rze l ihs12 20)966 (2009) 112S Lithos / al. et Robles-Cruz S.E. Nb2O5 0.26 0.33 0.27 3.03 0.34 0.48 0.43 0.40 1.20 0.21 0.51 0.25 0.11 0.30 0.24 0.07 2.54 1.59 0.96 ZrO2 0.12 0.14 0.17 0.06 –– – 0.10 0.17 0.02 0.00 0.00 0.02 0.15 0.08 0.06 0.01 0.12 0.08 Cr2O3 0.90 0.89 1.04 0.00 2.14 2.86 2.88 0.28 0.70 1.07 0.89 2.09 0.08 0.62 3.88 0.29 0.15 0.36 0.17 Fe2O3 19.99 18.93 8.68 5.33 28.59 28.19 27.56 28.43 9.85 6.20 12.44 9.91 1.30 20.51 13.39 0.23 1.26 2.81 1.73 FeO 28.76 27.56 30.30 41.04 27.35 27.40 27.72 28.26 16.65 23.76 30.20 27.76 43.40 26.93 25.65 33.02 37.37 35.88 37.77 MnO 0.20 0.24 0.27 0.68 0.15 0.18 0.22 0.18 2.06 0.44 1.88 0.39 2.68 0.23 0.25 13.08 6.45 6.35 6.81 MgO 5.87 6.66 8.43 0.15 3.74 3.71 3.63 3.57 16.43 13.79 5.78 9.47 0.06 6.07 8.74 0.07 0.13 0.75 0.15 NiO 0.08 0.07 0.08 0.03 0.05 0.04 0.03 0.05 0.04 0.08 0.08 0.10 0.03 0.07 0.09 0.01 0.01 0.03 0.00 CaO 0.03 0.02 0.04 0.21 0.01 0.00 0.00 0.01 0.00 0.02 0.06 0.00 0.03 0.00 0.02 0.03 0.21 0.50 0.09 Total 100.28 99.14 100.22 97.43 99.32 99.39 99.30 100.08 100.17 100.08 99.29 100.18 99.17 97.20 98.63 98.41 97.40 97.41 97.79
Cations on basis of three O atoms (apfu) Si 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.03 0.00 Ti 0.81 0.80 0.90 0.97 0.70 0.69 0.70 0.72 0.87 0.93 0.87 0.89 0.98 0.79 0.84 0.99 0.95 0.92 0.96
Al 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 – Nb 0.00 0.00 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.03 0.02 0.01 975 Zr 0.00 0.00 0.00 0.00 –– – 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Cr 0.02 0.02 0.02 0.00 0.05 0.06 0.06 0.01 0.01 0.02 0.02 0.04 0.00 0.01 0.07 0.01 0.00 0.01 0.00 Fe3+ 0.35 0.37 0.15 0.01 0.54 0.54 0.52 0.53 0.17 0.11 0.23 0.18 0.02 0.39 0.24 0.00 0.02 0.05 0.03 Fe2+ 0.56 0.58 0.60 0.94 0.58 0.58 0.59 0.59 0.31 0.45 0.62 0.55 0.92 0.56 0.52 0.71 0.81 0.77 0.81 Mn 0.00 0.00 0.01 0.03 0.01 0.01 0.01 0.00 0.04 0.01 0.04 0.01 0.06 0.00 0.01 0.28 0.14 0.14 0.15 Mg 0.24 0.21 0.30 0.00 0.14 0.14 0.14 0.13 0.55 0.47 0.21 0.33 0.00 0.23 0.31 0.00 0.00 0.03 0.01 Ni 0.00 0.00 0.00 0.00 –– – 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ca 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.00 FeTiO3 57.44 59.49 60.91 96.41 58.00 58.00 59.00 59.90 31.47 45.69 62.94 56.12 92.93 56.85 54.17 71.72 84.38 79.79 82.23 Fe2O3 17.95 18.97 7.61 0.51 27.00 27.00 26.00 26.90 8.63 5.58 11.68 9.18 1.01 19.80 12.50 0.00 1.04 2.59 1.52 MnTiO3 0.00 0.00 1.02 3.08 1.00 1.00 1.00 0.00 4.06 1.02 4.06 1.02 6.06 0.00 1.04 28.28 14.58 14.51 15.23 MgTiO3 24.62 21.54 30.46 0.00 14.00 14.00 14.00 13.20 55.84 47.72 21.32 33.67 0.00 23.35 32.29 0.00 0.00 3.11 1.02 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Textural types: (1) intercumulus in peridotite xenolith; (2) anhedral in carbonatite xenolith; (3) homogeneous macrocryst; (4) partly replaced macro- and megacryst; (5) symplectitic macrocryst; (6) skeletal crystal; (7) matrix. The Fe3+ is calculated by charge balance and stoichiometry. S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975 973
Fig. 7. Compositions of the different textural types of ilmenite in the Catoca kimberlite: (A) in terms of the geikielite (MgTiO3)–ilmenite (FeTiO3)–hematite (Fe2O3) diagram, after
Haggerty and Tompkins (1983); (B) in therms of the pyrophanite (MnTiO3)–ilmenite (FeTiO3)–hematite (Fe2O3) diagram.
On the other hand, a late generation of Mn-rich ilmenite is found primary magmatic crystallization in the matrix of the kimberlite. In all rimming all the above-mentioned generations, and is extremely poor these cases, however, Mn-rich ilmenite is produced in late events in in Mg and Fe3+. The replacement of the Fe3+-rich ilmenite by Fe2+- the paragenetic sequence at Catoca, and in many cases mantles other and Mn-rich ilmenite is also indicative of a trend toward strongly groundmass minerals such as perovskite and spinel (Tompkins and reducing conditions (Haggerty and Tompkins, 1983; Fig. 7A,B). Haggerty, 1985). Although Mn-rich ilmenite could be produced during Accordingly, these compositions could follow the kimberlite reaction magmatic crystallization, we contend that it could be also produced trend of Haggerty et al. (1977), producing enrichment in Fe2+. during late hydrothermal processes, during serpentinization. In fact, However, the most significant aspects in this process are the strong pyrophanite can be produced during serpentinization of ultrabasic enrichments in Mn and HFSE. Similar enrichments were interpreted in rocks, where it appears as a late mineral in the paragenetic sequence other kimberlites worldwide as produced by crystallization at the (Mücke and Woakes, 1986; Liipo et al., 1994). expense of a late-stage fraction of melt (Tompkins and Haggerty, 1985, In any case, all of the ilmenite fractions in kimberlite are quite Chakhmouradian and Mitchell, 1999). In the Catoca case, two facts different from those found in the carbonatitic xenoliths at Catoca. In suggest instead the deposition of this ilmenite under the influence of a this case, the growth of ilmenite takes place during the early stages of
CO2-rich fluid phase: a) the intimate association of this Mn-rich magmatic crystallization, and there is no evidence of replacement of a ilmenite in open cavities with calcite, witherite, barytocalcite and precursor ilmenite. Moreover, the crystals are distinct from the other strontianite; b) the development of this mineral association filling variants of ilmenite in being extremely poor in Mg and Cr and the small fractures. In fact, the late stages of kimberlite emplacement are richest in Nb, thus defining a particular class, more similar to ilmenite developed under the influence of CO2-rich fluids (Head and Wilson, found in carbonatites (Gaspar and Wyllie, 1983, 1984). 2008), whose are also responsible for the alteration of host rocks in The existence of many varieties of ilmenite at Catoca has significant many kimberlite fields worldwide (Smith et al., 2004); Agee et al. consequences in mineral exploration. Magnesium-rich ilmenite has (1982) also attributed the formation of Mn-rich ilmenite in the Elliott traditionally been interpreted as an indicator of kimberlite associations,
County kimberlite, Kentucky (USA) to Ca-enriched late fluids. as well as an indicator of low fO2, which is necessary for the preservation Furthermore, the composition of this replaced ilmenite is similar to of diamond (Garanin et al.,1997; van Straaten et al., 2008). However, the that of the fine-grained euhedral ilmenite crystals found in the Fe3+-rich ilmenite represents in the Catoca kimberlite more than 70% of kimberlite matrix. Analogous trends have been already described in the volume of the grains, and compositions fall into the domains of “no other kimberlite fields, but in the hypabyssal facies (i.e. Hunter et al., preservation of diamond” according the diagram of Gurney and Zweistra 1984). Similar textures and compositions in the groundmass are not (1995) Fig. 8. Moreover, these compositions of ilmenite are Mg- and Cr- rare in kimberlites. Tompkins and Haggerty, 1985; Chakhmouradian poor, and hence using other criteria for discrimination among fertile and and Mitchell, 1999 interpreted this type of ilmenite as produced by barren kimberlites (i.e., Haggerty, 1995); Catoca could be expected to be 974 S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975
Fig. 8. Representation of the compositions of ilmenite from the Catoca suite in the diagram of Gurney and Zweistra (1995). Note that most of the population of primary compositions of ilmenite from Catoca plots in poorly mineralized domains. barren. Although Catoca is a diamondiferous kimberlite, Mg-rich ilmen- Acknowledgments ite here is clearly a product of late replacement, and the extent of replacement of the primary grains is very variable. In other words, This research was supported by the projects CGL2005-07885/BTE textural relations must be taken into account in the application of and CGL2006-12973 of Ministerio de Educación y Ciencia (Spain), the discriminants based on composition. AGAUR SGR 589 of Generalitat de Catalunya and a FI grant sponsored by the Departament d'Educació i Universitats de la Generalitat de Catalunya and European Social Fund. We also thank ENDIAMA and the mine 6. Conclusions geologists, who kindly allowed us to acquire samples for this study and gave all facilities for the mine trip. The authors also acknowledge the The composition of the Catoca ilmenite is complex, as the result of Serveis Cientificotècnics de la Universitat de Barcelona for assistance in multiple processes. The ilmenite macro- and megacrysts are assumed to the use of SEM/ESEM-BSE-EDS analyses (E. Prats, R. Fontarnau†,Dr.J. be produced by disaggregation of ilmenite-bearing xenoliths (mainly García Veigas) and EMP (Dr. Xavier Llovet). An early version of the relatively oxidized and metasomatized mantle peridotites and minor manuscript was improved with the comments of two anonymous carbonatites). The subsequent reaction under disequilibrium conditions reviewers. Vicki Loschiavo and Prof. Robert F. Martin made further with kimberlite-derived fluids produced the replacement of the above improvements to the revised version. macro- and megacrysts by secondary Mg-rich ilmenite. Late subsolidus reactions with the fluids associated with the kimberlite, also in disequilibrium conditions, produced the replacement References of the early ilmenite types by highly reduced Mn-rich ilmenite. The enrichment in Nb of this late ilmenite (and in the ilmenite of the matrix), Agee, J.J., Garrison Jr., J.R., Taylor, L.A., 1982. Petrogenesis of oxide minerals in kimberlite, as well as its intimate association with carbonates of Ba and Sr, can be Elliott County, Kentucky (USA). American Mineralogist 67 (1–2), 28–42. interpreted in terms of an interaction of the ilmenite crystals with a CO - Arculus, R.J., Dawson, J.B., Mitchell, R.H., Gust, D.A., Holmes, R.D., 1984. Oxidation states 2 of the upper mantle recorded by megacryst ilmenite in kimberlite and type A and B rich fluid. spinel lherzolites. Contributions to Mineralogy and Petrology 85, 85–94. Although Catoca is a diamondiferous kimberlite, most of its ilmenite Armstrong, K.A., Nowicki, T.E., Read, G.H., 2004. Kimberlite AT-56: a mantle sample – compositions are strongly oxidized and poor in Cr and Mg. Therefore, the from the north central Superior craton, Canada. Lithos 77, 695 704. Chakhmouradian, A.R., Mitchell, R.H., 1999. Niobian ilmenite, hydroxylapatite and absence of Mg-rich ilmenite in a kimberlite or the corresponding placers sulfatian monazite: alternative hosts for incompatible elements in calcite kimberlite does not appear to be a convincing argument to exclude the occurrence from Internatsional'naya, Yakutia. The Canadian Mineralogist 37 (5), 1177–1189. of economic deposits of diamond. Accordingly, this work proposes new Clement, C.R., Skinner, E.M.W., 1985. A textural-genetic classification of kimberlites. Transactions of the Geological Society of South Africa 88, 403–409. insight into the concept of ilmenite in exploration for diamond; Dawson, J.B., Smith, J.V., 1977. The MARID (mica–amphibole–rutile–ilmenite–diopside) compositional attributes must be evaluated in light of textural attributes. suite of xenoliths in kimberlite. Geochimica et Cosmochimica Acta 41 (2), 309–323. S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975 975
De Carvalho, H., Tassinari, C., Alves, P.H., 2000. Geochronological review of the Kotel'nikov, D.D., Zinchuk, N.N., Zhukhlistov, A.P., 2005. Stages of serpentine and Precambrian in western Angola: links with Brazil. Journal of African Earth Sciences phlogopite transformation in the Catoca kimberlite pipe, Angola. Doklady Earth 31 (2), 383–402. Sciences 403 (6), 866–869. Ganga, J., Rotman, A.Y., Nosiko, S., 2003. Pipe Catoca, an example of the weakly eroded Kriuchkov, A., Sarychev, I., Vouiko, V., Chipoia, B., Antoniuk, B., 2000. Kimberlite pipe of kimberlites from North-East of Angola. The 8th International Kimberlite Con- Catoca. Geology, substantial analyses of the rocks and characteristics of diamonds. ference, Extended Abstract. Geoluanda 2000 Abstract volume, 91–92. Garanin, V.K., Kudryavtseva, G.P., Posukhova, T.V., 1997. Indicators of diamond Liipo, J.P., Vuollo, J.E., Nykänen, V.M., 1994. Pyrophanite and ilmenite in serpentinized preservation in kimberlite. In: Papunen, H. (Ed.), Mineral Deposits: Research and wehrlite from Ensilä, Kuhmo greenstone belt, Finland. European Journal of Exploration, Where Do They Meet? In: Proceedings of the 4th Biennial SGA Mineralogy 6, 145–150. Meeting. Turku, Finland, pp. 767–770. Llusià Queral, R.M., Pereira Ginga, S., Galiano, A., Sebastião, P., Melgarejo Draper, J.C., Olimpio Gaspar, J.C., Wyllie, P.J., 1983. Ilmenite (high magnesium, manganese, niobium) in the Gonçalves, A., Proenza Fernandez, J., Morais, E., 2005a. Aluviões asociados ao campo carbonatites from the Jacupiranga Complex, Brazil. American Mineralogist 68 (9–10), kimberlítico da zona de Muanga (Lunda Norte, Angola. XIV semana Geoquímica/VIII 960–971. Congresso de Geoquimica dos Paises de Lingua Portuguesa, pp. 351–354. Gaspar, J.C., Wyllie, P.J., 1984. The alleged kimberlite-carbonatite relationship: evidence Llusià Queral, R.M., Pereira Ginga, S., Proenza, J.A., José Ribamar, M.R., Sebastião, P., from ilmenite and spinel from Premier and Wesselton Mines and the Benfontein Olimpio Gonçalves, A., Melgarejo Draper, J.C., Galiano, A.C., Alves Morais, E., 2005b. Sill, South Africa. Contributions to Mineralogy and Petrology 85, 133–140. O campo kimberlítico de Muanga (Lunda Norte, Angola. XIV semana Geoquímica/ Griffin, W.L., Ryan, C.G., 1995. Trace elements in indicator minerals: area selection VIII Congresso de Geoquimica dos Paises de Lingua Portuguesa, pp. 347–350. and target evaluation in diamond exploration. Journal of Geochemical Exploration Meyer, H.O.A., McCallum, M.E., 1986. Mineral inclusions in diamonds from the Sloan 53 (1–3), 311–337. kimberlites, Colorado. Journal of Geology 94 (4), 600–612. Grütter, H.S., Gurney, J.J., Menzies, A.H., Winter, F., 2004. An updated classification for Mitchell, R.H., 1973. Magnesian ilmenite and its role in kimberlite petrogenesis. Journal mantle-derived garnet, for use by diamond explorers. Lithos 77, 841–857. of Geology 81, 301–311. Guiraud, R., Bosworth, W., Thierry, J., Delplanque, A., 2005. Phanerozoic geological Mitchell, R.H., 1989. Kimberlites: mineralogy, geochemistry and petrology. New York, evolution of Northern and Central Africa: an overview. Journal of African Earth Plenum Press. 442 pp. Sciences 43, 83–143. Mitchell, R.H., 1995. Kimberlites, orangeites, and related rocks. New York, Plenum Press. Gurney, J.J., Helmstaedt, H., Moore, R.O., 1993. A review of the use and application of 410 pp. mantle mineral geochemistry in diamond exploration. Pure and Applied Chemistry Mitchell, R.H., 1997. Kimberlites, orangeites, lamproites, melilitites, and minettes: a 65 (12), 2423–2442. petrographic atlas. Almaz Press Inc, Ontario. Gurney, J.J., Zweistra, P., 1995. The interpretation of the major element compositions of Mitchell, R.H., et al., 2009, this issue. Tuffisitic kimberlites from the Wesselton Mine, mantle minerals in diamond exploration. Journal of Geochemical Exploration 53 South Africa: Mineralogical characteristics relevant to their formation. Proceedings (1–3), 293–309. of the 9th International Kimberlite Conference. Lithos 112S, 452–464. Haggerty, S.E., 1991a. Oxide textures - a mini-atlas. In: Lindsley, D.H. (Ed.), Oxide Moore, A.E., 1987. A model for the origin of ilmenite in kimberlite and diamond: Minerals: Petrologic and Magnetic Significance. Reviews in Mineralogy 25, implications for the genesis of the discrete nodule (megacryst) suite. Contributions 129–219. to Mineralogy and Petrology 95, 245–253. Haggerty, S.E., 1991b. Oxide mineralogy of the upper mantle. In: Lindsley, D.H. (Ed.), Mücke, A., Woakes, M., 1986. Pyrophanite: a typical mineral in the Pan-African Province oxide minerals: petrologic and magnetic significance. Reviews in Mineralogy 25, of Western and Central Nigeria. Journal of African Earth Sciences 5 (6), 675–689. 355–416. Nielsen, T.F.D., Sand, K., 2008. The Majuagaa kimberlite dike, Maniitsoq region, West Haggerty, S.E., 1995. Upper mantle mineralogy. Journal of Geodynamics 20 (4), Greenland: constraints on an Mg-rich silico-carbonatitic melt composition from 331–364. groundmass mineralogy and bulk compositions. The Canadian Mineralogist 46, Haggerty, S.E., Hardie III, R.B., McMahon, B.M., 1977. The composition of ilmenite nodule 1043–1061. associations from the Monastery diatreme: Proceedings of the 2nd International Pasteris, J., 1980. The significance of groundmass ilmenite and megacryst ilmenite in Kimberlite Conference, vol. 2, pp. 249–256. kimberlites. Contributions to Mineralogy and Petrology 75 (4), 315–325. Haggerty, S.E., Tompkins, L.A., 1983. Redox state of Earth's upper mantle from Read, G.H., Janse, A.J.A.(B.), 2009, this issue. Diamonds: Exploration, mines and kimberlitic ilmenites. Nature 303 (5915), 295–300. marketing. Proceedings of the 9th International Kimberlite Conference. Lithos 112S, Haggerty, S.E., Tompkins, L.A., 1984. Subsolidus reactions in kimberlitic ilmenites: 1–9. exsolution, reduction and the redox state of the mantle. In: Kornprobst, J. (Ed.), Reis, B., 1972. Preliminary note on the distribution and tectonic control of kimberlites in Kimberlites I: Kimberlites and Related Rocks. Proceedings of the 3rd International Angola: The 24th International Geological Congress - Section 4, pp. 276–281. Kimberlite Conference. Rogers, A.J., Grütter, H.S., 2009, this issue. Fe-rich and Na-rich megacryst clinopyroxene Head, J.W., Wilson, L., 2008. Integrated model for kimberlite ascent and eruption. 9th and garnet from the Luxinga kimberlite cluster, Lunda Sul, Angola. Proceedings of International Kimberlite Conference Extended abstract No. 9IKC-A-00284. the 9th International Kimberlite Conference. Lithos 112S, 942–950. Hearn Jr., B.C., 2004. The Homestead kimberlite, central Montana, USA: mineralogy, Schulze, D.J., 1984. Cr-poor megacrysts from the Hamilton Branch kimberlite, Elliott xenocrysts, and upper-mantle xenoliths. Lithos 77, 473–491. County, Kentucky. In: Kornprobst, J. (Ed.), Kimberlites II: The Mantle and Crust–Mantle Hunter, R.H., Kissling, R.D., Taylor, L.A., 1984. Mid- to late-stage kimberlitic melt Relationships. In: Developments in Petrology 11B. Amsterdam, Elsevier, pp. 97–108. evolution: phlogopites and oxides from the Fayette County kimberlite, Pennsylva- Scott Smith, B.H., Nowicki, T.E., Russell, J.K., Webb, K.J., Hetman, C.M., Harder, M., nia. American Mineralogist 69, 30–40. Mitchell, R.H., 2008. Kimberlites: Descriptive Geological Nomenclature and Hwang, P., Taylor, W.R., Rock, N.M.S., Ramsay, R.R., 1994. Mineralogy, geochemistry and Classification. 9th International Kimberlite Conference Extended abstract No. petrogenesis of the Metters Bore No. 1 lamproite pipe, Calwynyardah field, West 9IKC-A-00124. Kimberley Province, Western Australia. Mineralogy and Petrology 51 (2–4), Smith, C.B., Sims, K., Chimuka, L., Duffin, A., Beard, A.D., Townend, R., 2004. Kimberlite 195–226. metasomatism at Murowa and Sese pipes, Zimbabwe. Lithos 76 (1–4), 219–232. Janse, A.J.A., Sheahan, P.A., 1995. Catalogue of world wide diamond and kimberlite Sparks, R.S.J., Baker, L., Brown, R.J., Field, M., Schumacher, J., Stripp, G., Walters, A., 2006. occurrences: a selective and annotative approach. Journal of Geochemical Dynamical constraints on kimberlite volcanism. Journal of Volcanology and Exploration 53, 73–111. Geothermal Research 155 (1–2), 18–48. Jones, A.P., Smith, J.V., Dawson, J.B., 1982. Mantle metasomatism in 14 veined peridotites Stiefenhofer, J., Viljoen, K.S., Marsh, J.S., 1997. Petrology and geochemistry of peridotite from Bultfontein mine, South Africa. Journal of Geology 90, 435–453. xenoliths from the Letlhakane kimberlites, Botswana. Contributions to Mineralogy Kaminsky, F.V., Sablukov, S.M., Sablukova, L.I., Shchukin, V.S., Canil, D., 2002. Kimberlites and Petrology 127 (1–2), 147–158. from the Wawa area, Ontario. Canadian Journal of Earth Sciences 39 (12), Tompkins, L.A., Haggerty, S.E., 1984. Koidu Kimberlite Complex, Sierra Leone. Geological 1819–1838. Setting, Petrology and Mineral Chemistry. In: Kornprobst, J. (Ed.), Kimberlites I: Kopylova, M.G., Russell, J.K., Cookenboo, H., 1999. Petrology of peridotite and pyroxenite Kimberlites and Related Rocks. Proceedings of the 3rd International Kimberlite xenoliths from the Jericho kimberlite: implications for the thermal state of the Conference. mantle beneath the Slave craton, Northern Canada. Journal of Petrology 40 (1), Tompkins, L.A., Haggerty, S.E., 1985. Groundmass oxide minerals in the Koidu kimberlite 79–104. dikes, Sierra Leone, West Africa. Contributions to Mineralogy and Petrology 91 (3), Kostrovitsky, S.I., Alymova, N.V., Yakovlev, D.A., Serov, I.V., Ivanov, A.S., Serov, V.P., 2006. 245–263. Specific features of “picroilmenite” composition in various diamondiferous fields of Van Straaten, B.I., Kopylova, M.G., Russell, J.K., Webb, K.J., Scott Smith, B.H., 2008. the Yakutian province. Doklady Earth Sciences 406 (1), 19–23. Discrimination of diamond resource and non-resource domains in the Victor North Kostrovitsky, S.I., Malkovets, V.G., Verichev, E.M., Garanin, V.K., Suvorova, L.V., 2004. pyroclastic kimberlite, Canada. Journal of Volcanology and Geothermal Research 17, Megacrysts from the Grib kimberlite pipe (Arkhangelsk Province, Russia). Lithos 77 128–138. (1–4), 511–523. Wyatt, B.A., Mike, B., Anckar, E., Grütter, H., 2004. Compositional classification of Kostrovitsky, S.I., Piskunova, L.F., 1990. Two groups of symplectites from the same “kimberlitic” and “non-kimberlitic” ilmenite. Lithos 77, 819–840. Kimberlite pipe. Doklady Earth Science Sections 306 (3), 176–179. ORIGINAL PUBLICATIONS
PAPER III
Reprinted from MACLA - Revista de la Sociedad Española de Mineralogía, September No.11. Robles- Cruz, S., Lomba, A., M., Melgarejo, J., Galí, S., Olimpio, A., 2009. The Cucumbi Kimberlite, NE Angola: Problems to Discriminate Fertile and Barren Kimberlites.
macla nº 11. septiembre ‘09 revista de la sociedad española de mineralogía 159
The Cucumbi Kimberlite, NE Angola: Problems to Discriminate Fertile and Barren
Kimberlites / SANDRA ROBLES-CRUZ (1,*), ANDRÉ LOMBA (2), JOAN CARLES MELGAREJO (1), SALVADOR GALI (1), ANTONIO OLIMPIO GONÇALVES (3) (1) Departament de Cristal·lografia, Mineralogia i Dipòsits Minerals. Facultat de Geologia. Universitat de Barcelona. Martí i Franquès s/n. 08028, Barcelona (España) (2) ENDIAMA, Angola (3) Departamento de Geologia, Faculdade de Ciências, Universidade Agostinho Neto, Av. 4 de Fevereiro 7, 815, Luanda (Angola)
INTRODUCTION.
A classical key issue in exploration of diamondiferous kimberlites is the accurate use of typical diamond indicator minerals in order to discriminate among fertile and barren kimberlites. In fact, the conclusive criterion is the occurrence of diamond itself which proves the productivity of a given kimberlite. In a previous paper (Robles-Cruz et al., 2009), we pointed out that in the Catoca pipe, the use of ilmenite composition is not suitable to confirm the diamond grade. We have used a new set of samples in Cucumbi area to study the reliance of some of these parameters, in particular, the use of garnet composition as a guide in diamond exploration.
Cucumbi is located in Cacolo, Lunda Sul province, northeastern Angola. This area is notable because of the occurrence of diamondiferous kimberlites (fig. 1).
fig 1. General location of kimberlites in Angola. Modified after De Carvalho et al. (2000) and Egorov et al. Cucumbi samples exhibit crater facies (2007). along the first 100 m, characterized by volcanoclastic rocks, and diatreme composed by gabbro, norite and (alkaline, carbonatitic, kimberlitic) and facies, showing typical tuffisitic charnockitic complexes, which marginal basins. The Lower Cretaceous kimberlite (Mitchell et al., 2009). constitute the Angolan basement. regional extension determined the development of deep faults and grabens METHODOLOGY. (2) Three main Proterozoic cycles, with trends NE-SW and NW-SE. The Eburnean-Paleoproterozoic, Kibaran- Lucapa structure corresponds to the first Thin and polished sections were studied Mesoproterozoic, and Pan-African- group, and the NE part concentrates using transmitted and reflected optical Neoproterozoic; being the Eburnean the most of the diamondiferous kimberlites microscopy, followed by SEM-BSE-EDS most important and characterized by in Angola, including Cucumbi, whereas analysis. Chemical analyses were volcanosedimentary groups, gneisses, the southwestern zone comprises obtained with EPMA. migmatites, granites and syenites. important outcrops of undersaturated alkaline rocks and carbonatites (Reis, GEOLOGICAL SETTING. (3) Unconformably lying Phanerozoic 1972). Other minor kimberlite fields are sequences, which are the result of the found in the SW Angola (Egorov et al., Angola has a complex geological history Pangea formation and the consecutive 2007). that can be represented by three main breaking-up of Gondwana, that stages (De Carvalho et al., 2000; Fig. 1): contributed to the formation of rift PETROGRAPHY AND COMPOSITION basins associated to fault systems which (1) An important Archaean orogeny, later allowed the apparition of marine The Cucumbi samples exhibit all the registered by the Central Shield, Cuango sequences, the origin of the Karoo main characteristic features of Tuffisitic Shield and Lunda Shield, most of them Supergroup, intraplate magmatism Kimberlite TK (Figs. 2, 3). They are
palabras clave: Kimberlita, Mineral indicador, Diamante, Granate. key words: Kimberlite, Indicator mineral, Diamond, Garnet.
resumen SEM 2009 * corresponding author: [email protected] 160
generally massive, poorly sorted, clast- Garnet and clinopyroxene are usually de la Generalitat de Catalunya and supported rocks with the following main present as mega- and macrocrysts, and European Social Fund. We also thank components: anhedral olivine only rarely as part of xenoliths. ENDIAMA and the mine geologists, who macrocrysts, pseudomorphosed by Garnet composition is diverse. Using the kindly allowed us to acquire samples for serpentine and smectite; other mega- garnet classification of Grütter et al. this study and gave all facilities for the and macrocrysts as garnet, ilmenite, (2004), it may be stated (Fig. 4) that mine trip. The authors also acknowledge clinopyroxene, and phlogopite, whether some garnet derive from lherzolite (G9) the Serveis Cientificotècnics de la enclosed in a pelletal assemblage of and others from pyroxenite and eclogite Universitat de Barcelona for assistance serpentine or not, often pelletal lapilli, (G4, G5), only a few of them come from in the use of SEM/ESEM-BSE-EDS and an interclast groundmass in the uncommon, unusual or “polymict” analyses (E. Prats, Dr. J. García Veigas) matrix, mainly composed by serpentine, mantle lithologies. and EPMA (Dr. Xavier Llovet). less common by chlorite, smectite and calcite. The size and distribution of DISCUSSION AND CONCLUSIONS. REFERENCES. mega- and macrocrysts is chaotic (Figs. 2, 3). Using the diagram of Grütter et al. De Carvalho, H., Tassinari, C., Alves, P.H (2004) to plot the garnet compositions (2000): Geochronological review of the from Cucumbi, it should be pointed out Precambrian in western Angola: links with that all these compositions plot into the Brazil. Journal of African Earth Sciences 31 graphite domain, out of the (2), 383-402. diamondiferous field harzburgitic G10 Egorov, K.N., Roman’ko, E.F., Podvysotsky, V.T., Sablukov, S.M., Garanin, V.K., facies. Therefore, this kimberlite could D’yakonov, D.B. (2007): New data on be classified as barren using only that kimberlite magmatism in southwestern criterion. However, the Cucumbi Angola. Russian Geology and Geophysics kimberlite has proven to be 48, 323-336. diamondiferous. In fact, similar Grütter, H.S., Gurney, J.J., Menzies, A.H., problems were found in the Catoca pipe Winter, F. (2004): An updated classification when using the composition of ilmenite for mantle-derived garnet, for use by (Robles-Cruz et al., 2009) or the diamond explorers. Lithos 77, 841-857. Mitchell, R. H., Skinner, E. M., Scott-Smith, B. fig 2. Cucumbi, a diamondiferous drill hole. A typical composition of garnets. H. (2009): Tuffisitic Kimberlites: pattern of a tuffisitic kimberlite (TK) facies, with Mineralogical Characteristics Relevant to macrocrysts containing rounded pseudomorphosed Therefore, the garnet diagrams can be their Formation. Lithos, Special Issue 9IK., olivine xenocrysts, rounded ilmenite xenocrysts and used to verify the minimum level of in press crustal rock xenoliths, all set in a groundmass of serpentine and phlogopite. Image from the scanned diamond content, but some kimberlites Reis, B. (1972): Preliminary note on the thin section. may contain diamond samples from distribution and tectonic control of deeper sources. Hence, it should be kimberlites in Angola: The 24th taken into consideration when using International Geological Congress - Section 4, 276-281. these diagrams to assess the potential Robles-Cruz, S.E., Watangua, M., Isidoro, L., of kimberlite fields. Melgarejo, J.C., Galí, S., Olimpio, A. (2009): Contrasting compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in exploration for diamond. Lithos, Special Issue 9IK., in press.
fig 3. Xenocrystals of olivine (pseudomorphosed by serpentine (Srp)), phlogopite (Phl), ilmenite (Ilm) in a serpentine groundmass. SEM image, mode BSE. fig 4. Classification of the Cucumbi garnets in a plot Cr2O3 versus CaO (wt.%), according with the Magnesian ilmenite (9-13 wt.% MgO) is compositional fields of Grütter et al. (2004). present as rounded mega- and macrocrysts (fig. 3), as part of xenoliths ACKNOWLEDGEMENTS. and as inclusions in phlogopite. In some cases macrocrysts of ilmenite are This research was supported by the partially replaced along the borders by projects CGL2005-07885/BTE and perovskite and spinel (Fig. 3). Ilmenite CGL2006-12973 of Ministerio de texture is usually either cumulus or Educación y Ciencia (Spain), the AGAUR homogenous. Symplectite textures are SGR 589 of Generalitat de Catalunya lacking in this kimberlite, in contrast and a FI-2006 grant sponsored by the with Catoca. Departament d’Educació i Universitats
ORIGINAL PUBLICATIONS
PAPER IV
Reprinted from Acta Mineralogica-Petrographica. Abstract Series, Vol. 6. Robles-Cruz, S.E., Escayola, M., Melgarejo, J.C., Watangua, M., Galí, S., Gonçalves, O.A., Jackson, S., 2010. Disclosed data from mantle xenoliths of Angolian kimberlites based on LA-ICP-MS analyses
! "#
& ' * +; < < = > ? @ + Z Z @ ] ^ _ = @ $ ! # * < @ ! #[ ^ } 9 ! ! \] ^ ` ^ ^{ | { } ! # * } ^ } * ! # * < @ ! ! \] } `{ ^ }{| { ! #[ { { < } ~ ! # * < ! { } ! } ! }! #[ ^ } ! { } } } ! # * } } [ { }! < ! #[ ! { } ! { * < } ! ! } } } } { } } } { } { { } } < {} < } [ ! ^ { { } ^^! * ! < * } < } ^ { ! }! } } @ {} ^ } # < * < { } ! * } * { {} } < * < \ } } < { } } |! } \ } ! ^ } } |^ { } < {}< * < } * } * }{ } ^ ! { { { } * } { < < ! @ * \ ! } } } |! } { } < \ } ! * < } ^ @ { \ | } } ! } |! { } { \ | < * { } \ ! ! } } | } ^ * } } ^ / } } } } ^ @{ { } } ^ } * }{ { } * < ^ O P ! ^ / ^ } { { < O { } P } < ^ İ} * } O P ^ < { }! } } < }^ } < ! } < } } * } < ! { { } * { ^ } ^ # { { ! O P * { ^ O P ! { * { < \ { } * { | ^ < < * { ^ } } < * { } } { } } } } ^ ! O P } < { { } { * { < } } } < ^ { { ! O { } P < { } } ^
! ^ ^ ^ \ | Lithos! ``! ^ ! ! H} * * } ^ ^ \ | Can. Mineral., $! ^ ! ^ { ! { } ^ ^ \ | Earth. Planet. Sci. Lett^! $#! ^
{ { < } * } { < < * {
! ^^ \ | Earth Planet. Sci. Lett., # ! ^ ! ^ ^ \ | Geology! $ ^ ! ^^ ^ \ | Chem. Geol.! % ORIGINAL PUBLICATIONS
PAPER V
Reprinted from Chemical Geology, 310-311. Robles-Cruz, S.E., Escayola, M., Jackson, S., Galí, S., Pervov, V., Watangua, M., Gonçalves, O.A., Melgarejo, J.C., 2012. U–Pb SHRIMP geochronology of zircon from the Catoca kimberlite, Angola: Implications for diamond exploration.
Chemical Geology 310-311 (2012) 137–147
Contents lists available at SciVerse ScienceDirect
Chemical Geology
journal homepage: www.elsevier.com/locate/chemgeo
U–Pb SHRIMP geochronology of zircon from the Catoca kimberlite, Angola: Implications for diamond exploration
Sandra E. Robles-Cruz a,⁎, Monica Escayola b, Simon Jackson c, Salvador Galí a, Vladimir Pervov d, Manuel Watangua e, Antonio Gonçalves e, Joan Carles Melgarejo a a Department de Cristal•lografia, Mineralogia i Dipòsits Minerals, Universitat de Barcelona, Martí i Franquès s/n, 08028 Barcelona, Spain b CONICET-IDEAN Instituto de Estudios Andinos, Laboratorio de Tectónica Andina, Universidad de Buenos Aires, C1033AAJ Capital Federal, Argentina c Geological Survey of Canada, 601 Booth Street, Ottawa, Ont. K1A 0E8, Canada d Sociedade Mineira de Catoca, Catoca, Lunda Sul, Angola e Departamento de Geologia, Universidade Agostinho Neto, Av. 4 de Fevereiro 7, 815 Luanda, Angola article info abstract
Article history: We present the first age determinations of zircon from the diamondiferous Catoca kimberlite in northeastern Received 9 August 2011 Angola, the fourth largest kimberlite body in the world. The U–Pb ages were obtained using a Sensitive High Received in revised form 2 April 2012 Resolution Ion Microprobe II (SHRIMP II) on zircon crystals derived from tuffisitic kimberlite rocks and Accepted 4 April 2012 heavy-mineral concentrates from the Catoca kimberlite. The SHRIMP results define a single weighted mean Available online 15 April 2012 age of 117.9±0.7 Ma (Mean square weighted deviation MSWD=1.3). More than 90% of the results indicate Editor: K. Mezger a single age population. There is no evidence for variable ages within single crystals, and no diffusional profiles are preserved. These data are interpreted as the maximum age of the kimberlite eruption at Catoca. Keywords: The U/Th values suggest at least two different sources of zircon crystals. These different populations may U–Pb dating reflect different sources of kimberlitic magma, with some of the grains produced in U- and Th-enriched meta- SHRIMP somatized mantle units. This idea is consistent with the two populations of zircon identified in this study. One Geochronology population originated from a depleted mantle source with low total REE (less than 25 ppm), and the other Zircon was derived from an enriched source, likely from the mantle or a carbonatite-like melt with high total REE Diamond (up to 123 ppm). Catoca kimberlite The tectonic setting of northeastern Angola is influenced by the opening of the south Atlantic, which reacti- vated deep NE–SW-trending faults during the early Cretaceous. The eruption of the Catoca kimberlite can be correlated with these regional tectonic events. The Calonda Formation (Albian–Cenomanian age) is the earliest sedimentary unit that incorporates eroded material derived from the diamondiferous kimberlites. Thus, the age of the Catoca kimberlite eruption is restricted to the time between the middle of the Aptian and the Albian. The new interpretation will be an important guide in future exploration for diamonds because it provides precise data on the age of a diamond-bearing kimberlite pulse in Angola. © 2012 Elsevier B.V. All rights reserved.
1. Introduction megacrysts have lower U and Th concentrations, and a lower total abundance of rare-earth elements (REE) than zircon of crustal deriva- Kimberlites contain primary minerals crystallized from a kimberlitic tion (Belousova et al., 2002; Heaman et al., 2006; and references there- magma, a suite of mega- and macrocrysts (e.g., zircon, diamond), and a in). The U–Pb dating of zircon is by far the most widely used method for complex variety of xenoliths (e.g., peridotite). According to previous obtaining reliable mineral-growth ages from different types of rocks. studies (Moore et al., 1992; Griffin et al., 2000; Pivin et al., 2009;andref- Zircon can provide reliable ages because of its resistance to thermal dis- erences therein), some zircon megacrysts (crystals larger than 1 cm) turbances. However, the dating of kimberlites is one of the most difficult that do crystallize from fractionating magmas in the mantle, are inter- applications of this method because crystal growth may have occurred grown with other megacryst phases (ilmenite, phlogopite, high-Fe hundreds of millions of years before kimberlite emplacement. olivine), and contain inclusions of chromian diopside, chromite, and The aim of this study is to determine the U–Pb ages of zircon from diamond. Mantle-derived zircon megacrysts have trace element com- the diamondiferous Catoca kimberlite, in the Lunda Sul province of positions that are distinct from zircon derived from the crust. These Angola, using a Sensitive High Resolution Ion Microprobe II (SHRIMP II). The study contributes to a better understanding of the geological evolution of the Catoca kimberlite, which has important implications ⁎ Corresponding author. Tel.: +34 506 8839 3981; fax: +34 506 2242 4411. for diamond exploration. Trace-element analyses carried out by Laser E-mail addresses: [email protected], [email protected] (S.E. Robles-Cruz). Ablation-Inductively Coupled Plasma-Mass Spectrometry (LA-ICP-MS)
0009-2541/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.chemgeo.2012.04.001 138 S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147
Fig. 1. Location map of the area of study. Geological map of northeastern Angola (after De Araujo et al., 1988; De Araujo and Perevalov, 1998; De Carvalho et al., 2000; Egorov et al., 2007). Abbreviations: Quaternary (QQ), Cenomanian (CE), Albian (AB), Permian (PP), Carboniferous (CC), Undifferentiated (Undiff.), Group (Gp), Formation (Fm), sandstone (Sst), conglomerate (Cgl), limestone (Lst), marlstone (Mrls), argillaceous limestone (ArgLst), claystone (Clst), granite (Gr), gabbro (Gb), quartzite (Qzt), schist (Sch), granodiorite (Grdr), dolerite (Do), amphibolite (Am), gneiss (Gns), carbonatites (Cbt), nephelite (Nph), syenite (Syt), ijolite (Ijt), pyroxenite (Pxt), anorthosite (Ant), troctolite (Trt), Norite (Nrt), epidotite (Epd), granulite (Gnt), eclogite (Ecl). were used to determine the chemical composition of the zircon (includ- kimberlite in southwestern Angola and interpreted it as the kimberlite ing the REE) from the kimberlite and to suggest potential sources of the age. Eley et al. (2008) reported a 206Pb/238Uageforzirconfromthe zircon. We also present a comparison of the Catoca kimberlite with Alto Cuilo 55 kimberlite and the Alto Cuilo 197 kimberlite (each date kimberlites in southeastern Brazil that were formed during the Early was based on a single-point analysis) of 113.0±0.8 Ma. The same Cretaceous. authors reported 206Pb/238U ages for perovskite from the Alto Cuilo 139 kimberlite of 135.7±2.1 Ma and from the Alto Cuilo 1 kimberlite 1.1. Background information and previous geochronological research in of 145.1±4.0 Ma (the authors did not specify the technique that was Angola used), and a Rb–Sr age for phlogopite from the Alto Cuilo 254 kimberlite of 115.5±1.1 Ma. All of the Alto Cuilo kimberlites are a part of the The kimberlites of Angola are distributed in clusters (Pereira et al., Luxinga kimberlite cluster, located ca. 85 km southwest of Catoca. 2003; Egorov et al., 2007) in the northeastern, central, and southwestern According to Eley et al. (2008), the ages indicate that kimberlite intru- areas of the country. Most of the kimberlites are concentrated within the sive activity took place in the Luxinga cluster between approximately Lucapa structure, with a NE–SW orientation, or along NW–SE faults 145 and 113 Ma. Recently, Jelsma et al. (2012) reported U–Pb ages for (Fig. 1). The Lucapa structure is a major basement fault system with zircon from kimberlites in central Angola between 252 and 216 Ma the highest diamondiferous potential in Angola (Pereira et al., 2003). (median age of 235 Ma), using LA-ICP-MS. The authors interpreted The Catoca kimberlite, the fourth largest kimberlite pipe in the world these data as a new age population of kimberlite emplacement in (639,000 m2), is located in the northeastern part of this structure, and Angola. exhibits rocks of crater and diatreme facies. In the earliest geochronological studies, Bardet and Vachette (1966) 2. Geological setting proposed a main Cretaceous kimberlitic event in the Congo craton based on stratigraphic relationships. Subsequently, Davis (1977) dated The geological history of Angola includes the following three one crystal of zircon from the Val do Queve kimberlite, located in the major phases, which have shaped the country (De Carvalho et al., central part of Angola by conventional U–Pb TIMS analysis. He reported 2000; Guiraud et al., 2005; Egorov et al., 2007): (1) the Archean orog- a 206Pb–238U age of 134.0±2.0 Ma. However, Davis noted some unspe- eny; (2) the Proterozoic orogenic cycles (Eburnian: Paleoproterozoic, cified analytical problems, which he attributed to the crystal's low U Kibaran: Mesoproterozoic, and Pan-African: Neoproterozoic); and content. Later, Haggerty et al. (1983) carried out fission-track dating (3) the deposition of Phanerozoic sedimentary sequences resting of zircon from the same area and reported an age of 133.4±11.5 Ma, unconformablyonpreviouslyerodedsurfaces(Pereira et al., 2003). The which they interpreted as the time of eruption of the Val do Queve subsequent break-up of Gondwana during the Jurassic to Cretaceous, kimberlite. In contrast, Egorov et al. (2007) published a K–Ar date of between 190 and 60 Ma (e.g., Jelsma et al., 2004), caused the develop- 372±8 Ma for phlogopite from the groundmass of the Chicuatite ment of basins that are associated with deep fault systems in Angola. S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147 139
These fault systems facilitated the emplacement of alkaline, carbonatitic, has become an important target in the exploration for alluvial dia- and kimberlitic magmas (Pereira et al., 2003). mond deposits. The Late Cretaceous regional extension was associated with older deep-seated faults and “grabens” with NE–SW and NW–SE trends 3. Sample description (Jelsma et al., 2009). An example of such a NE–SW trend is the Lucapa deep-seated fault system, which developed a local basin in northeastern In this study, nineteen crystals of zircon between 0.6 and 4 mm in Angola along a line that continues southwest to a transform fault of the length were taken from three core samples (four crystals) of tuffisitic Mid-Atlantic Ridge (Whiteetal.,1995). The Lucapa structure has been a kimberlites (TK) and three samples (fifteen crystals) of heavy- belt of recurring tectonic weakness since the Paleoproterozoic (Jelsma mineral concentrate from the Catoca kimberlite to perform the et al., 2009). Most of the diamondiferous kimberlites in Angola are SHRIMP analyses (Table 1). The TK have macrocrysts (0.5–10 mm) located along the Lucapa structure in northeastern Angola, although it of olivine (35–50 modal %), which are, in most cases, completely is not clear whether kimberlites were emplaced at the time of rifting replaced by serpentine, calcite, and saponite. Xenoliths of the host or whether they resulted from post-rifting events. Strike-slip and rocks (3–5 modal %), comprising gneiss and amphibolites, are com- shear fault systems in northeastern Angola likely could have led to mon. Also present are mantle xenoliths (1–5 modal %) (e.g., altered decompression (local extension) and compression, resulting in some metasomatized peridotite) and xenoliths of carbonatite (calcite+ control on the distribution of igneous activity within the Lucapa struc- anhedral Mn-rich ilmenite+zircon+apatite+phlogopite). Garnet ture. These processes could also have different expressions within the (G9 and G10, Grütter et al., 2004), chromian diopside, ilmenite, Angolan Shield and Kasai craton. Reliable age determinations are very amphibole, phlogopite, and zircon are found as mega- and macrocrysts important for understanding the timing of these intracontinental pro- in an altered kimberlite groundmass. Some zircon crystals exhibit a cesses. In the southwestern part of Angola, there are outcrops of under- reaction rim of baddeleyite. The matrix of the TK rocks contains lizar- saturated alkaline rocks and carbonatites along this trend (Reis, 1972), dite, apatite, calcite, ilmenite, and chromite. Titanite, zirconolite, badde- as well as some minor kimberlite fields (Egorov et al., 2007). leyite, barite, dolomite, witherite, barytocalcite, strontianite, and Continental sediments that unconformably overlie the Precambri- sulfides have also been identified in the matrix by quantitative powder an basement filled the Lucapa structure during the Cretaceous and X-ray diffraction (powder method) and quantitative chemical analyses Paleogene. An example of this package of sediments is the Calonda using an electron-microprobe (EMP) (Robles-Cruz et al., 2009). One Formation, a fining upward lithostratigraphic unit of the Kwango additional zircon crystal derived from a heavy-mineral concentrate Group that was formed by torrential deposits gradually changing to from the Tchiuzo kimberlite (15 km north of the Catoca kimberlite) lagoonal facies and concluding with low-energy deposits associated was added to this study for comparative purposes. with aeolian episodes. The Calonda Formation is the oldest sedimen- Large grain-size is a characteristic feature of zircon from kimberlites. tary unit in Angola that contains detrital diamond and kimberlite According to several authors (e.g., Belousova et al., 1998), zircon crystals clasts (Pereira et al., 2003; and references therein). It is reported to found in kimberlites are relatively large in size (several millimeters) com- be Albian to Cenomanian in age, based on fish macrofossils, palyno- pared to most zircon crystals from other types of igneous rock or from morphs, and tectonostratigraphic studies (Pereira et al., 2003). It metamorphic rocks. These crystals vary in color from colorless to
Table 1 Description of zircon crystals analyzed by SHRIMP.
Kimberlite Borhole Sample Type of Crystal Width Length Crystal Zonation (using Presence Degree of Type of name/mount sample no. (μm) (μm) border cathodoluminescence of rim of fracturing analyses images) baddeleyite
Catoca 535 CA-535-379-29B Core (TK) 1 350 600 Angular Oscillatory YES L SHRIMP 2 120 1000 Subangular Patchy zoning YES M SHRIMP Catoca 335 CA-335-601 Concentrate 3 900 1100 Subrounded Patchy-oscillatory zone NO M SHRIMP core 4 800 1000 Subrounded Oscillatory NO L SHRIMP 5 550 650 Angular Oscillatory NO VL SHRIMP 6 650 1100 Subrounded Oscillatory–patchy rim NO VL SHRIMP 7 800 800 Subangular Oscillatory NO VL SHRIMP 8 600 800 Subangular Oscillatory NO M SHRIMP 9 800 850 Subangular No zoning YES M SHRIMP Catoca 535 CA-535-359-27BA Core (TK) 10 450 850 Angular Oscillatory YES L SHRIMP Catoca 335 CA-335-551 Concentrate 11 600 1000 Subrounded Oscillatory NO L SHRIMP 12 800 850 Subangular Oscillatory NO VL SHRIMP 13 750 1450 Subrounded Oscillatory–patchy by NO L SHRIMP sectors 14 700 800 Subangular Oscillatory and patchy NO M SHRIMP in fracture 15 400 1150 Subrounded Oscillatory NO M SHRIMP 16 500 650 subangular Oscillatory NO VL SHRIMP 17 400 680 Angular Oscillatory NO VL SHRIMP Catoca 535 CA-535-350-26 Core (TK) 18 3800 4000 Subrounded Growth zoning with a NO M SHRIMP patchy core/dark-CL concentric rim Catoca 536 CA-536-304 Concentrate 19 1400 2200 Subangular Oscillatory NO M SHRIMP+LA-ICP-MS 21 1500 1300 Subrounded NA NO L LA-ICP-MS 22 1200 1000 Subrounded NA NO L LA-ICP-MS Tchiuzo G10 TZ-G10-13 Concentrate 20 550 980 Subangular Oscillatory NO M SHRIMP
Middle (M)=40–20% of fractures; Low (L) =5–20%; Very Low (VL)=less than 5%. TK = tuffisitic kimberlite. 140 S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147 brownish yellow. The twenty crystals analyzed in this study are charac- Catoca TK, were analyzed with an LA-ICP-MS. These zircon crystals terized by an almost complete absence of crystal faces, and most of exhibit oscillatory zoning in CL. One of these crystals was also ana- them are angular to subangular, presumably as a result of fracturing. A lyzed with SHRIMP (no. 19). few of them are subrounded, likely owing to disequilibrium with the me- dium, causing incipient resorption during interaction with the kimberlitic 4. Analytical methods magma. According to back-scattered electron (BSE) images, the twenty 4.1. Determinations of trace element concentrations zircon crystals are homogeneous at the major element level. The crys- tals do not exhibit evidence of metamictization. Nineteen of the Trace-element analyses were carried out by LA-ICP-MS on three twenty crystals exhibit oscillatory zoning in cathodoluminescence zircon crystals (nos. 19, 21, and 22) from the Catoca kimberlite at the (CL) (Fig. 2B, C, D, and F), which is usually interpreted to be a result Geological Survey of Canada. The trace-element determinations were of crystallization in a melt or fluid (Belousova et al., 1998; Liati et performed using a New Wave Research UP213 laser-ablation system in al., 2004; Page et al., 2007). Crystal no. 18 has dark-CL growth zoning combination with a Perkin Elmer 6100DRC quadrupole inductively (Fig. 2E), which is indicative of high U content. Four zircon crystals in coupled plasma mass spectrometer. The data acquisition and calibration the analyzed set exhibit partial replacement by baddeleyite (ZrO2) protocols employed have been described by Longerich et al. (1996) and along the borders. In addition, these crystals are fractured, and the Jackson (2008). Operating conditions and data-acquisition parameters resulting fracture surfaces are not corroded by baddeleyite (Fig. 2A, used in this study are summarized in Table 2.Datareductionwas B). The baddeleyite crystals are irregular and narrow between 10 performed using the software GLITTER 4.4.2 (Griffinetal.,2008). The and 40 μm in breadth. standard NIST SRM 610 (synthetic glass reference material, National Three representative crystals (nos. 19, 21, and 22), between 1 and Institute of Standards and Technology) was used as the primary 3 mm in length and picked from the heavy-mineral concentrate of the calibration standard using concentration values from GEOREM (http://
Fig. 2. Representative crystals of zircon from the Catoca kimberlite and one crystal from the Tchiuzo pipe. (A) Back-scattered electron image of zircon (Zrn) and baddeleyite (Bdl) crystals from the Catoca kimberlite, crystal no. 10. (B) Cathodoluminescence image of crystal no. 10 at a different scale than A. (C, D, and E) Cathodoluminescence images of crystals of zircon (nos. 12, 4, and 18, respectively) from the Catoca pipe, and (F) crystal of zircon (no. 20) from the Tchiuzo pipe. S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147 141
Table 2 Pb composition of the surface blank. No fractionation correction was LA-ICP-MS operating conditions and data-acquisition parameters. applied to the Pb-isotope data. The 207Pb method (Williams, 1998)was 206 238 LA used to calculate Pb/ U ages and errors. Model New Wave Research UP213 Wavelength 213 nm 5. Results Pulse duration (FWHM) ca. 4 ns Nominal spot sizes used 80–120 μm Repetition rate 10 Hz 5.1. REE and Ti-in zircon thermometry Energy density at sample ca. 5 J/cm2 On the basis of total REE concentrations in zircon from the Catoca ICP-MS kimberlite, we identified two ranges of values. One set consisted of Model Perkin Elmer ELAN 6100DRC Carrier has flow (He) 0.96 L/min total REE concentrations of less than 25 ppm, and the other set was Make-up flow (Ar) 0.70 L/min characterized by total REE concentrations up to 123 ppm (Table 3). The Sampler and skimmer Nickel zircon from the Catoca kimberlite (Fig. 3) exhibits low concentrations of light REE (LREE), a positive anomaly in Ce (chondrite-normalized Data-acquisition parameters value up to 10.04 ppm, 9.6 ppm absolute concentration), a lack of an Data-acquisition protocol Time-resolved analysis fl Detector mode Pulse counting (b3 M c.p.s.) Eu anomaly, and a positive slope from Pr to Lu, which attens toward Isotopes determined 25Mg, 29Si, 39K, 42Ca, 49Ti, 51V, 53Cr, 55Mn, 57Fe, 59Co, the heavy REE (HREE). Similar chondrite-normalized patterns and REE 60Ni, 65Cu, 66Zn, 71Ga, 72Ge, 85Rb, 88Sr, 89Y, 90Zr, 93Nb, contents have been observed in zircon crystals from other kimberlites 133Cs, 137Ba, 139La, 140Ce, 141Pr, 146Nd, 147Sm, 151Eu, 157 159 163 165 167 169 173 175 in southern Africa, Yakutia, and Australia (Belousova et al., 1998, 2001). Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, 4+ 177Hf, 181Ta, 205Tl, 206Pb, 207Pb, 208Pb, 232Th, 238U The REE patterns in zircon indicate that HREE and Ce , with a much 3+ Scanning mode Peak hopping, 1 point per peak more compatible ionic radius (0.97 Å) compared to Ce (1.18 Å), are Dwell time per isotope 10 ms preferentially incorporated in the zircon structure (Ballard et al., 2002; Time per mass sweep ca. 453 ms Hoskin and Schaltegger, 2003). This anomalous abundance of Ce4+ Data acquisition time 180 s (ca. 60 s gas blank, up to ca. 120 s ablation) could be the result of an increased Ce4+/Ce3+ value in the melt as a Oxide production ThO+/Th+ b1% function of high oxygen fugacity (Belousova et al., 2002; Hoskin and Standards and calibration Schaltegger, 2003; Whitehouse and Platt, 2003; and references there- Samples Polished 25 mm round mounts in). Similarly, the absence of an Eu anomaly may be caused by a high fi Data-processing software GLITTER 4.4.2 (Grif n et al., 2008) Eu3+/Eu2+ value under oxidized magma conditions in a feldspar-free Calibration standard NIST SRM 610, GEOREM Preferred Values, Feb. 2010 magmatic environment (Hoskin and Schaltegger, 2003). The suggestion Internal standard SiO2 (32.8 wt.%) Secondary standard USGS microbeam standard, BCR-2G of relatively high oxygen fugacity (fO2)conditionsduringzirconcrystal- lization seems more favorable for an interpretation of a positive Ce anomaly combined with the absence of a negative Eu anomaly. georem.mpch-mainz.gwdg.de/and data downloaded February 24, Concentrations of yttrium exhibit distinct ranges of values for each
2010). The stoichiometric SiO2 content of 32.8% was used for internal zircon population (Table 3). In one population, Y concentrations are standardization to correct differences in ablation yield between the sample and reference material. A secondary standard, BCR-2G (a homogeneous basaltic reference glass prepared by the U.S. Geological Table 3 Results of Laser-ablation ICP-MS analyses of zircon from the Catoca kimberlite. Survey: Jochum and Stoll, 2008), was used to monitor the precision and accuracy of the technique. Precision and accuracy were assessed Kimberlite Catoca Catoca Catoca Catoca Catoca from repeated analyses of the BCR-2G standard and were usually better Crystal No. 22 22 21 21 19 than 10% for concentrations at the ppm level. Detection limits were Type of Concentrate Concentrate Concentrate Concentrate Concentrate better than ~0.06 ppm for all elements reported, with the exception of Ti. sample
Spot Core Core Core Rim Core – 4.2. U Pb dating location
Sample fe26b05 fe26b06 fe26b09 fe26b10 fe26b15 The zircon crystals were mounted in epoxy along with fragments number of laboratory standard zircon z6266 (206Pb/238U age=559 Ma) at All values are reported in ppm the Geological Survey of Canada. The mid-sections of the zircon crys- Ti 5.4 10.9 1.17 2.66 3.4 tals were exposed and characterized in back-scattered electron (BSE) Sr 0.045 2.85 0.34 0.61 0.61 mode utilizing a Zeiss Evo 50 scanning electron microscope and a cold Y 46 120 330 480 39 cathodoluminescence stage mounted on a petrographic microscope Nb 1.25 4.1 8.8 13.5 0.86 to study internal features within the crystals, such as zoning and Ba 0.088 4.3 0.59 1.34 0.38 La 0.0181 0.50 0.054 0.168 0.075 structures. The surfaces of the 2.5-cm mounts were evaporatively Ce 0.79 2.89 5.3 9.6 0.82 coated with 10 nm of high-purity Au. Pr 0.031 0.211 0.187 0.39 0.0264 The U–Pb analyses were conducted at the Geochronology Laboratory, Nd 0.39 1.12 2.64 5.5 0.214 Geological Survey of Canada (Ottawa), using a Sensitive High Resolution Sm 0.73 1.04 4.4 7.7 0.241 − Eu 0.44 0.64 2.89 5.0 0.243 Ion Microprobe II (SHRIMP II). Analyses were performed using an 16O Gd 2.66 4.4 17.3 28.5 1.03 primary beam, projected onto the zircon crystals at 10 kV with a beam Tb 0.76 1.50 5.2 8.0 0.52 current of ca. 10 nA. The sputtered area used for analysis was ca. 25 μm Dy 6.5 16.2 51 79 5.1 in diameter. The count rates at ten masses including background were Ho 1.96 5.0 15.2 22.8 1.69 sequentially measured over six scans with a single electron multiplier Er 6.4 17.7 54 77 6.2 Tm 1.13 3.2 10.4 14.9 1.41 with a deadtime of 27 ns. Off-line data processing was accomplished Yb 8.8 25.4 86 123 12.9 using the SQUID 2.22.08.04.30 software, rev. 30 Apr 2008. The 1σ exter- Lu 1.42 3.8 11.7 15.1 1.50 nal errors of the 206Pb/238U ratios reported in the table of data (Table 4) Hf 12,800 13,700 9,180 9,970 9,580 incorporate a ±1.1% error in calibrating the standard zircon (see Stern Th 2.90 9.3 47 78 2.04 and Amelin, 2003). For the common Pb correction, we utilized the U 7.9 26.8 112 152 10.0 142 S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147
Fig. 3. Chondrite-normalized REE patterns (in black) for the three crystals of zircon (nos. 19, 21, and 22) from the Catoca kimberlite. Open symbols represent the rim, and closed symbols represent the core of crystals. Gray lines represent average REE trends for zircon from kimberlites and carbonatites reported by others and included were here for comparative purposes.
between 39 and 120 ppm, while higher Y concentrations, between those cases. For this reason, all results with common Pb proportions 330 and 480 ppm, are found in the other population. Concentrations greater than 15% were excluded from interpretations (7 of 41 analyses of Hf are between 9580 and 13,700 ppm for the first population and rejected, Fig. 4). The 34 analyses below the 15% cut-off value define a between 9180 and 9970 ppm for the second population of zircon. single weighted mean age of 117.9±0.7 Ma (Mean square weighted These three zircon crystals exhibit low Ti concentration, between deviation MSWD=1.3, probability 0.093, Fig. 4B). Quadratic addition 3.4 and 10.9 ppm, for the first population and between 1.17 and of the systematic error in the mount calibration (0.3% 1 sigma) gives a 2.66 ppm for the second population. There is no direct evidence that total error estimate of 117.9±0.7 Ma. these zircon crystals coexisted with a Ti-dominant phase. However, Five SHRIMP analyses were carried out on one zircon crystal (no. ilmenite is usually found in xenoliths, as mega- and macrocrystals, 20) from the Tchiuzo kimberlite, for comparison with the results dis- and in the groundmass of the Catoca kimberlite. cussed above. Concentrations of Th range between 11 and 34 ppm, The Ti-in-zircon thermometer (Watson et al., 2006) was applied and concentrations of U range from 36 to 61 ppm. The Th/U values to calculate the temperature at which the zircon crystallized. The range between 0.30 and 0.56. The weighted mean of the five analyses calculated temperatures are between 600 and 750 °C, which is very gives an age of 121±3 Ma (MSWD=1.6; probability of fit 0.17). Two low for a kimberlite. One possible explanation for this result is the analyses have common Pb contents above 15%; if these are rejected, lack of coexistence of zircon with a Ti-dominant phase, which leads the remaining three analyses have an identical weighted mean age to uncertainty in the activity coefficient of Ti. The same problem has of 121.2±1.8 (MSWD=0.86; probability of fit 0.45). Thus, the single been reported previously, where some doubts have been raised crystal from the Tchiuzo kimberlite yields an age slightly older than about the applicability of the Ti-in-zircon thermometer for zircon the weighted mean age obtained for the Catoca kimberlite. from kimberlites (Page et al., 2007).
6. Discussion 5.2. SHRIMP U–Pb ages 6.1. Different sources of zircon Forty-one SHRIMP analyses were performed on 19 zircon crystals from the Catoca kimberlite. Table 4 summarizes the results of all the The zircon crystals have well-defined characteristic features in SHRIMP analyses. Concentrations of Th and U range between 1 and their chondrite-normalized patterns, total REE abundances (Fig. 3), 654 ppm and between 6 and 326 ppm, respectively. These ranges of and different U and Th concentrations (Table 3). On the basis of their values, along with the REE analyses, appear to support the idea REE composition, two different populations of zircon crystals have that the zircon crystals originated from different sources. The Th/U been identified in the Catoca kimberlite. The first population is charac- values range from 0.21 to 2.07. Several analyses show very low U con- terized by a low concentration of REE (less than 25 ppm), U (less than centrations and very high proportions of common Pb, up to 45%. The 30 ppm), Th (less than 10 ppm), and Thzrn/Uzrn values (0.20–0.37). Ex- large amount of common Pb is evident in a Tera–Wasserburg plot perimental and theoretical studies in mafic and ultramafic rocks indicated (Fig. 5). Samples with common Pb concentrations of greater than that in such rocks, Uzrn/Umelt is approximately 100 and (Th/U)zrn–melt 15% have higher 206Pb/238U ages, which is attributed to uncertainties equals approximately 0.17 (Blundy and Wood, 2003). On the basis of that result from the very large common Pb correction required in results obtained in this study (Thzrn/Uzrn ≈0.26, Table 4), we estimate a S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147 143
Table 4 Summary of SHRIMP data for zircon from the Catoca and Tchiuzo kimberlites.
Kimberlite Crystal Spot name U Th Th/U 206Pba f(206)204% Total 204Pb corrected ratios 207 corrected no. (ppm) (ppm) (ppm) 238U/206Pb Age (Ma) 207Pb/206Pb 207/aPb/235U 206aPb/238U 206Pb/238U
Catoca 2 10001-2.1 82 43 0.52 1.3 0.73 53.756 ±0.705 0.0537±0.0017 0.1210 ±0.0068 0.0185 ±0.0002 118.1±1.6 10001-2.2 72 34 0.47 1.1 2.12 54.013 ±0.719 0.0491±0.0017 0.0769 ±0.0234 0.0181 ±0.0003 118.2±1.6 10001-2.3 23 6 0.28 0.3 2.45 54.672 ±0.901 0.0596±0.0036 0.0950 ±0.0480 0.0178 ±0.0005 115.3±2.0 Catoca 3 10002-2a.1 166 137 0.82 2.7 1.69 52.986 ±0.603 0.0536±0.0012 0.1001 ±0.0185 0.0186 ±0.0003 119.8±1.4 10002-2a.2 132 136 1.03 2.1 1.24 53.490 ±0.660 0.0550±0.0014 0.1132 ±0.0241 0.0185 ±0.0003 118.5±1.5 10002-2a-3.1 112 108 0.97 1.8 0.58 53.148 ±0.597 0.0579±0.0016 0.1368 ±0.0060 0.0187 ±0.0002 118.8±1.3 Catoca 4 10002-2b.3 14 3 0.21 0.2 5.14 49.764 ±0.896 0.1095±0.0061 0.1760 ±0.2517 0.0191 ±0.0022 119.1±2.3 10002-2b.4 22 5 0.23 0.3 9.51 51.720 ±0.659 0.0957±0.0046 0.0283 ±0.1018 0.0175 ±0.0009 116.6±1.6 Catoca 5 10002-3a.3 14 5 0.34 0.2 3.33 49.902 ±1.304 0.1212±0.0065 0.2526 ±0.1009 0.0194 ±0.001 117.0±3.2 10002-3a.4 28 11 0.38 0.4 8.94 54.638 ±0.783 0.0585±0.0032 0.0543 ±0.0711 0.0167 ±0.0013 115.5±1.7 Catoca 6 10002-3b.4 28 7 0.27 0.4 13.06 52.131 ±0.644 0.0739±0.0036 0.1137 ±0.0240 0.0167 ±0.0008 118.8±1.6 Catoca 8 10002-4b.1 9 2 0.23 0.2 -8.35 51.392 ±1.433 0.0561±0.0067 0.3509 ±0.0996 0.0211 ±0.0011 123.1±3.5 Catoca 9 10002-4c.1 147 122 0.83 2.3 0.38 54.191 ±0.712 0.0517±0.0013 0.1227 ±0.0182 0.0184 ±0.0003 117.4±1.5 10002-4c.2 177 184 1.04 2.8 0.91 54.663 ±0.613 0.0483±0.0011 0.1014 ±0.0158 0.0181 ±0.0002 116.9±1.3 Catoca 10 10003-1.3 45 18 0.41 0.7 3.72 52.795 ±0.702 0.0557±0.0024 0.0587 ±0.0623 0.0182 ±0.0006 119.9±1.6 10003-1.4 34 15 0.44 0.5 4.78 54.605 ±0.695 0.0483±0.0027 0.0141 ±0.0799 0.0174 ±0.0007 117.0±1.5 Catoca 11 10004-1a.1 11 2 0.21 0.2 11.21 48.272 ±0.706 0.1342±0.0083 0.0967 ±0.3592 0.0184 ±0.003 118.9±2.2 Catoca 12 10004-1b.1 48 18 0.38 0.7 6.65 53.216 ±0.638 0.0722±0.0027 0.0327 ±0.0413 0.0175 ±0.0004 116.7±1.4 10004-1b.2 51 18 0.34 0.8 6.16 54.479 ±0.638 0.0695±0.0025 0.0363 ±0.0357 0.0172 ±0.0004 114.3±1.4 Catoca 13 10004-1c.1 8 3 0.34 0.1 5.80 52.747 ±1.850 0.0740±0.0088 0.0578 ±0.1341 0.0179 ±0.0013 117.4±4.3 Catoca 14 10004-3a.2 35 12 0.36 0.5 6.09 50.972 ±0.752 0.0860±0.0035 0.0851 ±0.0711 0.0184 ±0.0007 119.7±1.8 Catoca 15 10004-3d.1 50 16 0.32 0.8 2.60 53.818 ±0.631 0.0697±0.0026 0.1190 ±0.0473 0.0181 ±0.0005 115.7±1.4 10004-3d.2 50 15 0.31 0.8 4.69 54.477 ±0.965 0.0613±0.0032 0.0490 ±0.0596 0.0175 ±0.0006 115.5±2.1 Catoca 16 10004-4a.1 22 7 0.33 0.3 11.77 51.562 ±0.836 0.0978±0.0046 0.0202 ±0.0805 0.0171 ±0.0008 116.7±2.0 10004-4a.2 12 3 0.29 0.2 9.11 47.685 ±1.070 0.1484±0.0082 0.1934 ±0.2182 0.0191 ±0.0020 118.1±2.9 Catoca 17 10004-4c.1 187 118 0.63 3.0 1.19 52.565 ±0.582 0.0532±0.0012 0.1118 ±0.0203 0.0188 ±0.0003 120.8±1.3 10004-4c.2 154 88 0.57 2.5 1.70 51.665 ±0.603 0.0572±0.0013 0.1121 ±0.0162 0.0190 ±0.0003 122.3±1.4 Catoca 18 10006-1.1 197 320 1.62 3.2 0.95 53.159 ±0.608 0.0516±0.0011 0.1117 ±0.0122 0.0186 ±0.0002 119.7±1.4 10006-1.2 301 623 2.07 4.7 0.40 54.320 ±0.656 0.0538±0.0009 0.1276 ±0.0073 0.0183 ±0.0002 116.8±1.4 10006-1.3 184 315 1.71 2.9 0.67 53.863 ±0.649 0.0577±0.0012 0.1323 ±0.0074 0.0184 ±0.0002 117.3±1.4 10006-1.4 326 654 2.01 5.2 0.87 53.724 ±0.589 0.0508±0.0008 0.1106 ±0.0075 0.0185 ±0.0002 118.5±1.3 Catoca 19b 10007-4.1 22 5 0.24 0.3 8.14 54.344 ±0.694 0.0568±0.0037 0.0407 ±0.0436 0.0169 ±0.0007 116.4±1.6 10007-4.2 79 50 0.64 1.2 3.08 54.340 ±0.627 0.0502±0.0017 0.0574 ±0.0553 0.0178 ±0.0005 117.3±1.4 10007-4.3 36 11 0.30 0.6 2.87 54.263 ±0.758 0.0579±0.0028 0.0820 ±0.0514 0.0179 ±0.0005 116.4±1.7 Tchiuzo 20 10005-1.1 61 34 0.56 0.9 3.30 53.402 ±0.731 0.0486±0.0020 0.0492 ±0.0443 0.0181 ±0.0004 119.6±1.7 10005-1.3 36 11 0.30 0.6 9.14 49.382 ±0.621 0.0937±0.0035 0.0331 ±0.0732 0.0184 ±0.0006 122.4±1.6 10005-1.4 43 13 0.31 0.7 5.56 49.454 ±0.587 0.0974±0.0034 0.1329 ±0.1156 0.0191 ±0.0010 121.7±1.5
Spot name follows the convention x−y.z; where x = sample number, y = grain number and z = spot number. Multiple analyses in an individual spot are labeled as x−y.z.z. Uncertainties reported at 1σ and are calculated by using SQUID 2.22.08.04.30, rev. 30 Apr 2008. f206204 refers to mole percent of total 206Pb that is due to common Pb, calculated using the 204Pb-method; common Pb composition used is the surface blank (4/6: 0.05770; 7/6: 0.89500; 8/6: 2.13840). Calibration standard 6266; U=910 ppm; Age=559 Ma; 206Pb/238U=0.09059. Error in 206Pb/238U calibration 1.1% (included). Standard Error in Standard calibration was 0.30% (not included in above errors but required when comparing data from different mounts). a Refers to radiogenic Pb (corrected for common Pb using measured 204Pb). b Also analyzed with LA-ICP-MS.
Throck/Urock value between 2.1 and 3.6 for the ‘first’ population of zircon. or as a result of a reaction with other minerals in a carbonated kimberlitic This value is within the limits proposed by Zartman and Richardson melt, as has been noted in other kimberlites (Haggerty, 1991; Dawson (2005), between ~4 and ~2, for depleted asthenosphere over the last et al., 2001; Page et al., 2007). In addition, the crystals were fragmented 2.5 Ga. This first population of zircon crystals could be genetically linked after the development of the baddeleyite rim because the broken to the kimberlite. The second population is characterized by high concen- surfaces are never replaced by baddeleyite (Fig. 2A, B). Therefore, these trations of REE (up to 123 ppm), high Th (more than 45 ppm), and high U crystals may have been replaced along the borders and then fractured (more than 100 ppm). This population is likely derived from an enriched during eruption or later, such as during treatment in the gridding mill. source. These two populations of primary zircon were produced by at least 6.2. Age data two different batches of magma that exhibit very similar ages (Fig. 6). The first population is similar in Th/U values to the zircon megacrysts Two interpretations are possible to explain the presence of the typically found in kimberlites (Heaman et al., 2006), and the second is SHRIMP U–Pb zircon ages obtained in this study. The first interpretation close to those reported from xenoliths of glimmerite (Rudnick et al., is that they reflect a period of zircon growth at 117.9±0.7 Ma, which 1998)orMARID(Kinny and Dawson, 1992; Hamilton et al., 1998). would represent a maximum eruption age for the Catoca kimberlite. Both populations of zircon crystals are corroded and overgrown by a The second interpretation is that the zircon crystals partially retain an rim of baddeleyite. The presence of such a rim suggests a desilication older, inherited component that was incompletely reset by diffusive reaction as a result of the interaction of zircon with carbonate in the Pb-loss prior to eruption at 117.9±0.7 Ma. Did the zircon form prior kimberlitic magma, for example: to kimberlite eruption but fail to quantitatively retain Pb owing to high ambient temperatures and diffusive loss of Pb? The Pb closure temperature of zircon is in excess of 900 °C þð ; Þ → þð ; Þ þ ZrSiO4 Ca Mg CO3 ZrO2 Ca Mg silicate CO2 (Cherniak and Watson, 2000; Heaman et al., 2006; and references 144 S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147
Fig. 4. Weighted mean of 206Pb/238U ages of zircon. (A) Zircon from the Catoca and the Tchiuzo kimberlites. (B) Zircon from the Catoca kimberlite.
¼ : −21 2= −1; ¼ ðÞB : therein). Thus, the zircon thus likely records the time when it was DPb 3 14 10 m s at T 1473 K 1200 C transported from the mantle by the kimberlitic magma. Although exposure to mantle temperatures for long periods of time will cause zircon to lose Pb through diffusion, several investigators of zircon This Pb diffusion is not significant and precludes the second in kimberlitic rocks (Mezger and Krogstad, 1997; Belousova et al., interpretation. 2001; Cherniak and Watson, 2003) have suggested that a complete Other studies (Schärer et al., 1997; Corfu et al., 2003) suggest that resetting does not invariably take place. Specifically, according to zircon keeps a record (partial or complete) of one or more thermal Belousova et al. (2001), zircon crystals may retain radiogenic Pb at events that it has experienced. Thus each zircon crystal is telling an lithospheric mantle temperatures between 600 and 1200 °C. “individual” history and the measured U–Pb zircon ages, together The estimated geotherm for the Catoca kimberlite (Robles-Cruz et al., with the REE concentrations, provide insight into different episodes unpublished results), calculated from data on garnet peridotite xenoliths of crystallization. generated using the Nimis and Taylor (2000) geothermobarometer, gives a temperature between 900 and 1200 °C at 40–55 kbar (160–200 km, in 6.3. Geotectonic implications the subcratonic lithospheric mantle). At this range of temperatures, the calculated volume diffusion of Pb in zircon using the Arrhenius relation This new interpretation of a maximum age for the kimberlitic erup- − (Cherniak and Watson, 2000) is between 3.23×10 26 at 900 °C and tion at 118±1 Ma (Aptian age) is consistent with the regional tectonos- − − 3.14×10 21 m2/s 1 at 1200 °C. Calculations of Pb diffusion in zircon tratigraphy of northeastern Angola (Fig. 7). The Catoca kimberlite was (DPb)give: expected to be younger than the carbonatites and alkaline rocks found in the Lucapa structure. These rocks yielded K–Ar and Rb–Sr ages ¼ −1 − −1= 2 −1; DPb 1 10 exp 550 kJ mol RT m s between 138 and 130 Ma (Alberti et al., 1999; and references therein). − The U–Pb ages obtained in this study are similar to those reported for R ¼ 8:314472 10 3ðÞgas constant the Alto Cuilo kimberlites (Eley et al., 2008; and references therein), which are slightly older than the Calonda Formation and contain eroded ¼ : −26 2= −1; ¼ ðÞB DPb 3 23 10 m s at T 1173 K 900 C fragments of diamondiferous kimberlite. S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147 145
are associated between 120 and 70 Ma, around the end of the Early Cretaceous. Our interpretation of 118±1 Ma for the maximum age of the kimberlitic eruption in Catoca, which is associated with a NE–SW tectonic trend (Lucapa structure), reinforces the hypothesis of Jelsma et al. (2009) that 120 Ma (Aptian age) kimberlites are prefer- entially associated with NE–SW tectonic trends, whereas 85 Ma (Santonian age) kimberlites are emplaced in E–W lineaments. Our finding of an Aptian age for the maximum age of the kimberlitic erup- tion in Catoca is also consistent with a single model for the magmatic province, which extends over what is now southeastern Brazil and southwestern Africa, coincident with the opening of the South Atlantic Ocean (Hawkesworth et al., 1992, 1999; Guiraud et al., 2010). The extensional tectonic setting, rifting, and opening of the South Atlantic during the Early Cretaceous (Pereira et al., 2003; Jelsma et al., 2009) and the reactivation of deep-seated fault systems probably contributed to lithospheric heating (mantle upwelling) and, ultimately, to kimberli- tic magmatism in Angola.
7. Conclusions
(1) On the basis of U–Pb-derived zircon dates, petrographic and cathodoluminescence imaging studies, REE data, and the re- gional geological setting, we conclude that the maximum age for the Catoca kimberlite eruption is 118±1 Ma, which is an Aptian age. Almost all of the analyses in this study belong to a single age population, with no evidence for variable ages within single crystals and no diffusional profiles preserved. (2) The U/Th values suggest at least two different sources of zircon crystals. Some of the zircon crystals could have been produced in U–Th-enriched metasomatized mantle units (MARID or glimmeritic suite assemblages), while others have chemistries suggestive of a depleted asthenosphere source. Hence, these different populations can reflect different sources for the kimberlitic magma. (3) The presence of the different sources of zircon is consistent Fig. 5. Tera–Wasserburg (T–W) diagrams with data obtained from zircon crystals from the with the two populations of zircon also identified based on Catoca and Tchiuzo kimberlites. Ellipses in gray are those that have a high proportion of REE abundances. These populations are characterized either 204Pb. (A) T–W diagram with data of crystals from the Catoca kimberlite, intersecting at 117.70±0.94 Ma. (B) T–W diagram with data of crystals from the Tchiuzo kimberlite, by zircon crystals originating from a depleted mantle source intersecting at 120.4±2.8 Ma. with low concentration of REE (less than 25 ppm) or by zircon crystals derived from an enriched source, likely a carbonatitic melt, with high concentrations of REE up to 123 ppm. Cretaceous kimberlitic events of similar age have also been reported (4) The age of the Catoca kimberlite is restricted to between 118± in the São Francisco craton (Brazil), the Kaapvaal craton (South Africa 1 Ma (the maximum age for the kimberlite eruption in Catoca) and Botswana), and the Congo–Kasai craton (the Democratic Republic and 112 Ma, the beginning of deposition of diamondiferous of Congo), which were all part of Gondwanaland (e.g., Batumike et al., clasts in the Calonda Formation. The eruptive event for the 2007; Jelsma et al., 2009). Systems of deep faults present in these cra- Catoca kimberlite appears to have taken place in this range of tons probably were the focus of thermal perturbations and injection of ages. melt. The Canastra 01 kimberlite in Brazil, located at the border of the (5) The diamondiferous Catoca kimberlite seems to be tectonically São Francisco craton (Da Costa, 2008), yielded an age of 120±10 Ma related to other Early Cretaceous kimberlites confined in NE– using K/Ar in phlogopite (Chaves et al., 2008; and references therein). SW lineaments in southwestern and southern Africa. This is It is associated with a NE–SW general tectonic trend and considered consistent with an incipient rifting stage previously proposed to be related to lithospheric heating that took place before rifting by Jelsma et al. (2009) between 135 and 115 Ma. This under- (Fleischer, 1998; Read et al., 2004), at a time when the presence of the standing has important implications for diamond exploration. Tristan da Cunha mantle plume (133 Ma) would have exerted tectonic The documentation concerning the maximum age of eruption control (Wilson, 1992). In the western part of the Kaapvaal craton, in of the Catoca kimberlite during the Aptian provides precise South Africa, kimberlites with ages of ca. 120 Ma (Jelsma et al., 2009; data on the age of a diamond-bearing kimberlite pulse in and references therein) are associated with NE–SW preferential tecton- Angola and should act as an important guide for diamond ic orientation. In the Congo–Kasai craton, the reported ages of the earli- exploration. est episodes of kimberlitic magmatism are between 116 and 70 Ma (Batumike et al., 2009; and references therein), where NE–SW and E– Acknowledgments W general tectonic trends have been identified. The youngest kimberli- tic magmatic episode reported, the Mbuji-Mayi kimberlites (70 Ma, We acknowledge the great contribution of Dr. Bill Davis to the Schärer et al., 1997), located in the east Kasai province in the Democrat- analytical work. We thank Dr. Anthi Liati and the two more anony- ic Republic of Congo, have an E–W trend. The implication is that there mous reviewers, as well as the editor, Dr. Klaus Mezger, for their was a change in the tectonic direction with which these kimberlites excellent revision of this manuscript and their valuable comments. 146 S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147
Fig. 6. Th versus U concentration in zircon from the Catoca (black circles) and the Tchiuzo pipes (gray circles). The area in light gray corresponds to the field of mantle-derived zircon megacrysts, as defined by Heaman et al. (2006). The dashed square represents the characteristic concentration of zircon in kimberlitic rocks reported by Belousova et al. (2002).A black star on the border of some circles indicates that those analyses were performed on zircon crystals with a rim of baddeleyite. Crystals of zircon outside the mantle-derived field are enriched in U and Th. Most of the crystals are larger than 0.6 mm in size, fractured, and yield similar ages. There is no correlation between age and Th/U content.
We greatly appreciate the revision and improvements of Prof. Robert F. Martin to this manuscript. This research is funded by the CGL2006- 12973 and CGL2009-13758 BTE projects of Ministerio de Educación y Ciencia (Spain), and the AGAUR SGR 589 and SGR444 of the General- itat de Catalunya. The first author (SERC) received an FI grant and a BE grant, both sponsored by the Departament d'Educació i Universitats de la Generalitat de Catalunya and the European Social Fund. We thank ENDIAMA, which kindly allowed SERC to acquire samples for her PhD thesis and allowed the use of all facilities for the mine trip. We acknowledge the Geological Survey of Canada (GSC), Ottawa, for all of the support during a six-month Volunteer Assistant visit of SERC, and we thank the Laboratories of Geochemistry and Geochronology (GSC), especially Tom Pestaj, for his collaboration and assistance during the preparation and analysis of samples. The authors also thank the Serveis Cientificotècnics de la Universitat de Barcelona for assistance in the use of SEM/ESEM-BSE-EDS analyses (E. Prats. and J. García Veigas).
References
Alberti, A., Castorina, F., Censi, P., Comin-Chiaramonti, P., Gomes, C.B., 1999. Geochemical characteristics of Cretaceous carbonatites from Angola. Journal of African Earth Sciences 29 (4), 735–759. Ballard, J.R., Palin, J.M., Campbell, I.H., 2002. Relative oxidation states of magmas inferred from Ce(IV)/Ce(III) in zircon: application to porphyry copper deposits of northern Chile. Contributions to Mineralogy and Petrology 144, 347–364. Bardet, M.G., Vachette, M., 1966. Détermination d'âges de kimberlites de l'Ouest Africain et essai d'interprétation des datations des diverses venues diamantifères dans le monde. Bureau de Recherches Géologiques et Minières, Département Minéralogie, Pétrographie, Métallogènie, Géochimie. Rep. No. DS. 66 A 69, Paris, p. 20. Batumike, J.M., O'Reilly, S.Y., Griffin, W.L., Belousova, E.A., 2007. U–Pb and Hf-isotope analyses of zircon from the Kundelungu Kimberlites, D.R. Congo: implications for crustal evolution. Precambrian Research 156 (3–4), 195–225. Batumike, J.M., Griffin, W.L., O'Reilly, S.Y., Belousova, E.A., Pawlitschek, M., 2009. Crustal evolution in the central Congo–Kasai Craton, Luebo, D.R. Congo: insights from zircon U–Pb ages, Hf-isotope and trace-element data. Precambrian Research 170 (1–2), 107–115. Belousova, E.A., Griffin, W.L., Pearson, N.J., 1998. Trace element composition and cath- odoluminescence properties of southern African kimberlitic zircons. Mineralogical Magazine 62A (3), 355–366. Belousova, E.A., Griffin, W.L., Shee, S.R., Jackson, S.E., O'Reilly, S.Y., 2001. Two age populations of zircons from the Timber Creek kimberlites, Northern Territory, as determined by Laser-ablation ICP-MS analysis. Australian Journal of Earth Fig. 7. Summary of geochronological results (SHRIMP) for the Catoca kimberlite eruption. Sciences 48 (5), 757–766. S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147 147
Belousova, E.A., Griffin, W.L., O'Reilly, S.Y., Fisher, N.I., 2002. Igneous zircon: trace ele- Wallace, C., Chinn, I., Henning, A., 2012. Kimberlites from Central Angola: a case ment composition as an indicator of source rock type. Contributions to Mineralogy study of exploration findings. Extended Abstr. No. 10IKC-42, 10th International and Petrology 143 (5), 602–622. Kimberlite Conference, Bangalore, pp. 1–5. Blundy, J., Wood, B., 2003. Mineral-melt partitioning of Uranium, Thorium and their Jochum, K.P., Stoll, B., 2008. Reference materials for elemental and isotopic analyses by daughters. In: Bourdon, B., Ribe, N.M., Stracke, A., Saal, A.E., Turner, S.P. (Eds.), LA-(MC)-ICP-MS: successes and outstanding needs. In: Sylvester, P. (Ed.), Laser Uranium-Series Geochemistry. Rev. Mineral. Geochem., vol. 52, pp. 59–123. Ablation ICP-MS in the Earth Sciences: Current Practices and Outstanding Issues, Chaves, M.L.S.C., Benitez, L., Brandão, P.R.G., Girodo, A.C., 2008. Kimberlito Canastra-1 Short Course Ser: Mineral. Assoc. Can., Québec, vol. 40, pp. 147–168. (São Roque de Minas, MG): geologia, mineralogia e reservas diamantíferas. REM, Kinny, P.D., Dawson, J.B., 1992. A mantle metasomatic injection event linked to late Revista Escola de Minas 61 (3), 57–364. Cretaceous kimberlite magmatism. Nature 360, 726–728. Cherniak, D.J., Watson, E.B., 2000. Pb diffusion in zircon. Chemical Geology 172, 5–24. Liati, A., Leander, F., Dieter, G., Fanning, C.M., 2004. The timing of mantle and crustal Cherniak, D.J., Watson, E.B., 2003. Diffusion in zircon. In: Hanchar, J.M., Hoskin, P.W.O. events in South Namibia, as defined by SHRIMP-dating of zircon domains from a (Eds.), Zircon. Rev. Mineral. Geochem., vol. 53, pp. 113–143. garnet peridotite xenolith of the Gibeon Kimberlite Province. Journal of African Corfu, F., Hanchar, J.M., Hoskin, P.W.O., Kinny, P., 2003. Atlas of zircon textures. In: Hanchar, Earth Sciences 39 (3–5), 147–157. J.M., Hoskin, P.W.O. (Eds.), Zircon. Rev. Mineral. Geochem., vol. 53, pp. 469–500. Longerich, H.P., Jackson, S.E., Günther, D., 1996. Laser ablation inductively coupled plas- Da Costa, G.V., 2008. Química mineral e geotermobarometria de xenólitos mantélicos ma mass spectrometric transient signal data acquisition and analyte concentration do kimberlito Canastra-01. Master thesis (Dissertação de Mestrado no. 239). calculation. Journal of Analytical Atomic Spectrometry 11, 899–904. Instituto de Geociências, Universidade de Brasília. Mezger, K., Krogstad, E.J., 1997. Interpretation of discordant U–Pb zircon ages: an evaluation. Davis, G.L., 1977. The ages and uranium contents of zircon from kimberlites and asso- Journal of Metamorphic Geology 15, 127–140. ciated rocks. Extended Abstr., 2nd Int. Kimberlite Conf., Santa Fe, New Mexico. Moore, R.O., Griffin, W.L., Gurney, J.J., Ryan, C.G., Cousens, D.R., Sic, S.H., Suter, G.F., Dawson, J.B., Hill, P.G., Kinny, P.D., 2001. Mineral chemistry of a zircon-bearing, composite, 1992. Trace element geochemistry of ilmenite megacrysts from the Monastery veined and metasomatised upper-mantle peridotite xenolith from kimberlite. kimberlite, South Africa. Lithos 29, 1–18. Contributions to Mineralogy and Petrology 140 (6), 720–733. Nimis, P., Taylor, W.R., 2000. Single clinopyroxene thermobarometry for garnet perido- De Araujo, A.G., Perevalov, O.V., 1998. Carta de recursos minerais — mineral resources tites. Part I. Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in- map. Ministério Geol. e Minas. Inst. Geol. Angola, Luanda. Cpx thermometer. Contributions to Mineralogy and Petrology 139, 514–554. De Araujo, A.G., Perevalov, O.V., Jukov, R.A., 1988. Carta geológica de Angola - Geological Page, F.Z., Fu, B., Kita, N.T., Fournelle, J., Spicuzza, M.J., Schulze, D.J., Viljoen, F., Basei, Map of Angola. Scale 1 : 1 000 000. Ministèrio da Indústria. Inst. Nac. Geol., Luanda. M.A.S., Valley, J.W., 2007. Zircons from kimberlite: new insights from oxygen iso- De Carvalho, H., Tassinari, C., Alves, P.H., 2000. Geochronological review of the Precambrian topes, trace elements, and Ti in zircon thermometry. Geochimica et Cosmochimica in western Angola: links with Brazil. Journal of African Earth Sciences 31 (2), 383–402. Acta 71 (15), 3887–3903. Egorov, K.N., Roman'ko, E.F., Podvysotsky, V.T., Sablukov, S.M., Garanin, V.K., D'yakonov, D.B., Pereira, E., Rodrigues, J., Reis, B., 2003. Synopsis of Lunda geology, NE Angola: implications for 2007. New data on kimberlite magmatism in southwestern Angola. Russian Geology diamond exploration. Comunicações do Instituto Geológico e Mineiro 90, 189–212. and Geophysics 48 (4), 323–336. Pivin, M., Féménias, O., Demaiffe, D., 2009. Metasomatic mantle origin for Mbuji-Mayi Eley, R., Grütter, H., Louw, A., Tunguno, C., Twidale, J., 2008. Exploration Geology of the and Kundelungu garnet and clinopyroxene megacrysts (Democratic Republic of Luxinga kimberlite Cluster (Angola) with evidence supporting the presence of Congo). Lithos 112S, 951–960. kimberlite lava. Extended Abstr. No. 9IKC-A-00166, 9th Int. Kimberlite Conf., Frankfurt. Read, G., Grütter, H., Winter, S., Luckman, N., Gaunt, F., Thomsen, F., 2004. Stratigraphic Fleischer, R., 1998. A rift model for the sedimentary diamond deposits of Brazil. relations, kimberlite emplacement and lithospheric thermal evolution, Quiricó Mineralium Deposita 33 (3), 238–254. Basin, Minas Gerais State, Brazil. Lithos 77, 803–818. Griffin, W.L., Pearson, N.J., Belousova, E., Jackson, S.E., van Achterbergh, E., O'Reilly, S.Y., Shee, Reis, B., 1972. Preliminary note on the distribution and tectonic control of kimberlites S.R., 2000. The Hf isotope composition of cratonic mantle: LAM-MC-ICPMS analysis of in Angola. Abstr., 24th Int. Geol. Congr., Rep. Sess., 4, pp. 276–281. Montreal. zircon megacrysts in kimberlites. Geochimica et Cosmochimica Acta 64 (1), 133–147. Robles-Cruz, S.E., Watangua, M., Isidoro, L., Melgarejo, J.C., Galí, S., Olimpio, A., 2009. Griffin, W.L., Powell, W.J., Pearson, N.J., O'Reilly, S.Y., 2008. GLITTER: data reduction Contrasting compositions and textures of ilmenite in the Catoca kimberlite, software for Laser Ablation ICP-MS. In: Sylvester, P. (Ed.), Laser Ablation ICP-MS Angola, and implications in exploration for diamond. Lithos 112S, 966–975. in the Earth Sciences: Current Practices and Outstanding Issues, Short Course Ser: Robles-Cruz, S.E., Melgarejo, J.C., Escayola, M., Galí, S., unpublished results. Compara- Mineral. Assoc. Can., Québec, vol. 40, pp. 308–311. tive compositions of xenocrysts of garnet, clinopyroxene, and ilmenite from dia- Grütter, H.S., Gurney, J.J., Menzies, A.H., Winter, F., 2004. An updated classification for mondiferous and barren kimberlites from northeastern Angola. Oral presentation. mantle-derived garnet, for use by diamond explorers. Lithos 77, 841–857. Peralk-Carb 2011 workshop on peralkaline rocks and carbonatites, 16–18 June. Guiraud, R., Bosworth, W., Thierry, J., Delplanque, A., 2005. Phanerozoic geological Tübinguen, Germany. evolution of Northern and Central Africa: an overview. Journal of African Earth Rudnick, R.L., Ireland, T.R., Gehrels, G., Irving, A.J., Chesley, J.T., Hanchar, J.M., 1998. Sciences 43, 83–143. Dating mantle metasomatism: U–Pb geochronology of zircons in cratonic mantle Guiraud, M., Buta-Neto, A., Quesne, D., 2010. Segmentation and differential post- xenoliths from Montana and Tanzania. Extended Abstr., 7th International Kimberlite rift uplift at the Angola margin as recorded by the transform-rifted Benguela Conference, Cape Town, pp. 754–756. and oblique-to-orthogonal-rifted Kwanza basins. Marine and Petroleum Geol- Schärer, U., Corfu, F., Demaiffe, D., 1997. U–Pb and Lu–Hf isotopes in baddeleyite and ogy 27, 1040–1068. zircon megacrysts from the Mbuji-Mayi kimberlite: constraints on the subconti- Haggerty, S.E., 1991. Oxide mineralogy of the upper mantle. In: Lindsley, D.H. (Ed.), Oxide nental mantle. Chemical Geology 143, 1–16. Minerals: Petrologic and Magnetic Significance: Rev. Mineral., vol. 25, pp. 355–416. Stern, R.A., Amelin, Y., 2003. Assessment of errors in SIMS zircon U–Pb geochronology Chantilly. using a natural zircon standard and NIST SRM 610 glass. Chemical Geology 197, Haggerty, S.E., Raber, E., Naeser, C.W., 1983. Fission track dating of kimberlitic zircons. 111–146. Earth and Planetary Science Letters 63 (1), 41–50. Watson, E.B., Wark, D.A., Thomas, J.B., 2006. Crystallization thermometers for zircon Hamilton, M.A., Pearson, D.G., Stern, R.A., Boyd, F.R., 1998. Constraints on MARID and rutile. Contributions to Mineralogy and Petrology 151, 413–433. petrogenesis: SHRIMP II U–Pb zircon evidence for pre-eruption metasomatism at White, S.H., de Boorder, H., Smith, C.B., 1995. Structural controls of kimberlite and Kampfersdam. Extended Abstr., 7th International Kimberlite Conference, Cape lamproite emplacement. Journal of Geochemical Exploration 53 (1–3), 245–264. Town, pp. 296–298. Whitehouse, M.J., Platt, J.P., 2003. Dating high-grade metamorphism-constraints from Hawkesworth, C.J., Gallagher, K., Kelley, S., Mantovani, M., Peate, D.W., Regelous, M., rare-earth elements in zircon and garnet. Contributions to Mineralogy and Petrology Rogers, N.W., 1992. Parana magmatism and the opening of the South Atlantic. In: 145, 61–74. Storey, B.C., Alabaster, T., Pankhurst, R.J. (Eds.), Magmatism and the Causes of Williams, I.S., 1998. U–Th–Pb geochronology by ion microprobe. In: McKibben, M.A., Continental Break-up: Geol. Soc. Spec. Publ. No. 68, London, pp. 221–240. Shanks III, W.C., Ridley, W.I. (Eds.), Applications of Microanalytical Techniques to Hawkesworth, C., Kelley, S., Turner, S., Le Roex, A., Storey, B., 1999. Mantle processes Understanding Mineralizing Processes: Reviews in Econ. Geol., 7, pp. 1–35. during Gondwana break-up and dispersal. Journal of African Earth Sciences 28 (1), Wilson, M., 1992. Magmatism and continental rifting during the opening of the South 239–261. Atlantic ocean: a consequence of Lower Cretaceous super-plume activity? In: Heaman, L.M., Creaser, R.A., Cookenboo, H.O., Chacko, T., 2006. Multi-stage modification of Storey, B.C., Alabaster, T., Pankhurst, R.J. (Eds.), Magmatism and the Causes of the northern Slave mantle lithosphere: evidence from zircon- and diamond-bearing Continental Break-up: Geol. Soc. Spec. Publ. No. 68., London, pp. 241–255. eclogite xenoliths entrained in Jericho Kimberlite, Canada. Journal of Petrology 47 Zartman, R.E., Richardson, S.H., 2005. Evidence from kimberlitic zircon for a decreasing (4), 821–858. mantle Th/U since the Archean. Chemical Geology 220, 263–283. Hoskin, P.W.O., Schaltegger, U., 2003. The composition of zircon and igneous and meta- morphic petrogenesis. In: Hanchar, J.M., Hoskin, P.W.O. (Eds.), Zircon. Rev. Mineral. Web references Geochem., vol. 53, pp. 27–62. Jackson, S., 2008. Calibration strategies for elemental analysis by LA-ICP-MS. In: http://georem.mpch-mainz.gwdg.de/, data downloaded February 24, 2010. GeoReM is Sylvester, P. (Ed.), Laser Ablation ICP-MS in the Earth Sciences: Current Practices a Max Planck Institute database for reference materials of geological and environ- and Outstanding Issues, Short Course Ser: Mineral. Assoc. Can., Québec, vol. 40, mental interest, such as rock powders, synthetic, and natural glasses as well as pp. 169–188. mineral, isotopic, biological, river water, and seawater reference materials. It con- Jelsma, H.A., de Wit, M.J., Thiart, C., Dirks, P.H.G.M., Viola, G., Basson, I.J., Anckar, E., 2004. tains published analytical data and compilation values (major and trace element Preferential distribution along transcontinental corridors of kimberlites and related concentrations, radiogenic and stable isotope ratios). It also contains all important rocks of Southern Africa. South African Journal of Geology 107, 301–324. metadata about the analytical values such as uncertainty, analytical method and Jelsma, H., Barnett, W., Richards, S., Lister, G., 2009. Tectonic setting of kimberlites. Lithos laboratory. Sample information and references are also included, as well as more 112S, 55–165. than 2,400 reference materials, 25,400 analyses from more than 5,000 papers, Jelsma, H., Krishnan, U., Perrit, S., Preston, R., Winter, F., Lemotlo, L., v/d Linde, G., and preferred analytical values (State: 07/01/2011). It complements the three Armstrong, R., Phillips, D., Joy, S., Costa, J., Facatino, M., Posser, A., Kumar, M., earthchem databases GEOROC, NAVDAT, PETDB. ORIGINAL PUBLICATIONS
PAPER VI
Reprinted from Minerals, Special Issue "Advances in Economic Minerals". Robles-Cruz, S.E., Melgarejo, J.C., Galí, S., Escayola, M., 2012. Major- and trace-element compositions of indicator minerals that occur as macro- and megacrysts, and of xenoliths, from kimberlites on the northeastern Angola.
1 Minerals 2012, 2, 1-x manuscripts; doi:10.3390/min20x000x 2 OPEN ACCESS
3 minerals 4 ISSN 2075-163X 5 www.mdpi.com/journal/minerals/ 6 Article
7 Major- and Trace-Element Compositions of Indicator Minerals 8 that Occur as Macro- and Megacrysts, and of Xenoliths, from 9 Kimberlites in Northeastern Angola
1, 1 1 2 10 Sandra E. Robles-Cruz *, Joan Carles Melgarejo , Salvador Galí and Monica Escayola
11 1 Department de Cristal·lografia, Mineralogia i Dipòsits Minerals, Universitat de Barcelona, 12 Barcelona 08028, Spain; E-Mails: [email protected]; [email protected] 13 2 CONICET-IDEAN Instituto de Estudios Andinos, Laboratorio de Tectónica Andina, Universidad 14 de Buenos Aires, Buenos Aires C1033AAJ, Argentina; E-Mail: [email protected]
15 * Corresponding author. E-Mail addresses: [email protected], [email protected]; 16 Tel.: +506-883-939-81; Fax: +506-224-244-11.
17 Received: / Accepted: / Published: 18
19 Abstract: In this study, we compare the major- and trace-element compositions of olivine, 20 garnet, and clinopyroxene that occur as single crystals (142 grains), with those derived 21 from xenoliths (51 samples) from six kimberlites in the Lucapa area, northeastern Angola: 22 Tchiuzo, Anomaly 116, Catoca, Alto Cuilo-4, Alto Cuilo-63, and Cucumbi-79. The 23 samples were analyzed using electron probe microanalysis (EPMA) and laser-ablation 24 inductively coupled plasma-mass spectrometry (LA-ICP-MS). The results suggest different 25 paragenetic associations for these kimberlites in the Lucapa area. Compositional overlap in 26 some of the macrocryst and mantle xenolith samples indicates a xenocrystic origin for 27 some of those macrocrysts. The presence of mantle xenocrysts suggests the possibility of 28 finding diamond. Geothermobarometric calculations were carried out using EPMA data 29 from xenoliths by applying the program PTEXL.XLT. Additional well calibrated single- 30 clinopyroxene thermobarometric calculations were also applied. Results indicate the 31 underlying mantle experienced different equilibration conditions. Subsequent metasomatic 32 enrichment events also support a hypothesis of different sources for the kimberlites. These 33 findings contribute to a better understanding of the petrogenetic evolution of the 34 kimberlites in northeastern Angola and have important implications for diamond 35 exploration. 36 Minerals 2012, 2 2
1 Keywords: Angola; kimberlite; olivine; garnet; clinopyroxene; diamond; 2 thermobarometry; mantle xenoliths; REE; Sm/Nd isotopes 3
4 1. Introduction
5 In Angola, more than 700 occurrences of kimberlite are distributed on a trend northeast to 6 southwest from Lunda Sul and Lunda Norte, up through Huambo, Benguela, and Huila provinces. 7 Diamond was first reported in Angola in 1590 [1]. In 1952, the first kimberlite in Angola, Camafuca- 8 Kamazambo, was discovered [2]. Since then, and especially after thirty years of civil war, Angola has 9 become known as an important diamond producer. The Catoca kimberlite in Lunda Sul province, the 10 first kimberlite mine in Angola, was for a long time the only kimberlite under production. In 2007, two 11 more pipes came on line, and subsequently other exploration projects have started in the country. By 12 2010, the Catoca kimberlite had produced over 8.36 million carats, valued at US$ 976 million [3]. 13 There are several diamondiferous and barren kimberlites in the northeastern area. However, there 14 are no detailed studies that allow a full understanding of their relationship with the underlying mantle, 15 or their spatial distribution. Six kimberlites from northeastern Angola will be considered for this study. 16 They range from poor to very high level of diamond production: Alto Cuilo 4 (AC4), Cucumbi-79 17 (CU79), Alto Cuilo-63 (AC63), Anomaly 1116 (An116), Tchiuzo (TZ) and Catoca (CA). These 18 kimberlites are located in the Lucapa area and emplaced in Archean metamorphic rocks, in the Kassai- 19 Congo Craton. 20 Kimberlites are volatile-rich potassic ultrabasic rocks usually with an inequigranular texture as a 21 result of the presence of crystals, compound clasts, and interstitial matrix [4,5]. Crystals can be: 22 megacrysts, crystals of more than 1.0 cm at maximum dimension (MD); macrocrysts, crystals between 23 0.5 and 10 mm at MD; or microcrysts, crystals less than 0.5 mm at MD. Xenoliths are rare; in these 24 kimberlites, most of them consist of metasomatized peridotite and phlogopite-rich suites. 25 The aim of this paper is to establish the major- and trace-element compositions of indicator 26 minerals: olivine, garnet, and clinopyroxene, that occur as single crystals with those derived from 27 xenoliths from six kimberlites from the Lucapa area, northeastern Angola: TZ, An116, CA, AC4, 28 AC63, and CU79. This area is where most of the diamondiferous kimberlites identified so far are 29 concentrated. We estimate reliable conditions of pressure and temperature for selected samples. Also 30 we have measured Sm/Nd isotopes from the rare xenoliths to better understand their petrogenesis.
31 2. Geological Setting
32 The exposed rocks in Angola range from Archean age to Recent (Figure 1). This geological history 33 may be divided in three main stages [6,7]: (1) the Archean orogeny is recorded by the Central Shield, 34 Cuango Shield and Lunda Shield, mainly composed of gabbro, norite and charnockitic complexes, 35 which constitute the Angolan basement. (2) There are three main Proterozoic cycles (Eburnean in the 36 Paleoproterozoic, Kibaran in the Mesoproterozoic, and Pan-African in the Neoproterozoic), of which 37 the Eburnean is the most important and characterized by complex volcanosedimentary rocks, gneisses, 38 migmatites, granites and syenites. This regional Paleoproterozoic event was followed by the Kibaran
Minerals 2012, 2 3
1 cycle, which was related to extensional events along the border of the Congo craton where later clastic- 2 carbonatic sequences and local basic magmatism took place. The Pan-African orogeny was associated 3 with the development of Gondwana and led to the generation of fold belts and granitic intrusions. (3) 4 Phanerozoic sequences covered the older rocks as events associated with the formation of Pangea and 5 the consecutive break-up of Gondwana which contributed to the formation of rift basins associated 6 with deep fault systems and later intraplate magmatism (alkaline, carbonatitic, kimberlitic) and 7 marginal basins.
8 Figure 1. Geological setting of the northeastern Angola kimberlites of this study (after 9 Perevalov et al. [8], Giraud et al. [7], Egorov et al. [9]).
10 11 12 The Lower Cretaceous regional extension caused the development of deep faults and grabens with 13 NE-SW and NW-SE trends. The Lucapa structure, a deep-seated fault system, corresponds to the first 14 group, and the northeastern part is the focus of most of the diamondiferous kimberlites in Angola, 15 whereas the southwestern zone comprises important outcrops of undersaturated alkaline rocks and 16 carbonatites [2]. Other minor kimberlite fields are found in southwestern Angola [9]. This geological 17 configuration sets a tectonic control on the presence of some of the kimberlites in northeastern Angola. 18 The emplacement of the Catoca kimberlite (middle of Aptian and Albian) has been recently correlated 19 with these regional tectonic events [10]. Kimberlite emplacement ages in the Alto Cuilo range between 20 145.0±4.0 (206Pb/238U for perovskite) and 113.0±0.8 Ma (206Pb/238U for zircon) [11]. Synsedimentary 21 continental sediments associated with the filling of the Lucapa structure (Calonda Formation) can also
Minerals 2012, 2 4
1 contain diamond crystals in paleoplacers; alluvial diamond is found in placers associated with rivers 2 draining these diamondiferous areas. 3 4 3. Analytical Techniques
5 Six hundred and fifty samples were studied with an optical petrographic microscope. Then 6 representative samples of mantle xenoliths (51 samples), mega- (3 grains), macro- (116 grains), and 7 microcrysts (23 grains) of olivine, garnet, and clinopyroxene were examined by back-scattered 8 electron (BSE) images using SEM-ESEM with EDS microanalysis and electron-probe microanalysis 9 (EPMA). More than 800 microprobe analyses were carried out to obtain the mineral chemistry of 10 major elements. The major elements were analyzed using a JXA JEOL-8900L EPMA at the 11 Department of Earth and Planetary Sciences, McGill University (Montreal, Quebec), using the ZAF 12 correction method. Acceleration voltage was 20 kV, beam current 20 nA, and beam diameter 5 μm. 13 The counting time for most elements was 20 s on peaks and 20 s on the background. Standardized 14 natural and synthetic minerals were used for calibration. 15 Then 14 representative samples of xenoliths and 25 macrocrysts of garnet, clinopyroxene, and 16 olivine were selected to perform trace-element analyses using laser-ablation inductively coupled 17 plasma mass spectrometry (LA-ICP-MS) at the Geological Survey of Canada, Ottawa. The data were 18 acquired with a Photon-Machines Analyte 193nm Excimer laser ablation in combination with an 19 Agilent 7500cx quadrupole ICP-MS, a powerful technique for reliable solid analysis of samples. Data 20 reduction was performed with the GLITTER 4.4.2 software. The primary calibration standard was the 21 synthetic glass standard of the 610 series (NIST SRM 610) of the National Institute of Standards and
22 Technology, using SiO2 for internal standardization. The GSE-1G (a synthetic reference glass with 23 basaltic major-element composition and trace elements abundance of ca. 500 μg/g) was used as a 24 secondary standard. 25 Finally, accurate high-precision Sm/Nd isotopic compositions of eight samples of xenoliths (1.5 to 26 7 cm in diameter) from kimberlites CA, CU79, and CU80 (for comparison purposes) were carried out 27 at the Pacific Centre for Isotopic and Geochemical Research (PCIGR) at the University of British 28 Columbia, using a Thermo Finnigan Triton thermo-ionization mass spectrometer (TIMS). We used the 29 analytical procedures for sample dissolution, ion exchange, and leaching described by Weis et al. [12]. 30 The normalization procedure has been applied to the Nd isotopic ratios using La Jolla Nd as the 31 reference material (measured ratio normalized to La Jolla 143Nd/144Nd = 0.511858).
32 4. Morphology of the Kimberlite Pipes Studied in Northeastern Angola
33 The CA, TZ, An116, AC4, AC63, and CU79 kimberlites are located within the Kasai craton (Figure 1). 34 The pipes generally appear in clusters along a network of local fractures. They are of variable shape 35 and dimension, some very large up to 900 m in diameter (i.e., Catoca). Some pipes in this area are 36 diamondiferous, and do have an economic grade. Crater and diatreme facies can be recognized despite 37 the intense weathering affecting kimberlites up to a depth of 150-200 m. Because of the weathering the 38 number of fresh xenoliths and crystals is limited, and conditioned the selection of samples for this 39 study. Next to olivine, serpentine is the most abundant mineral in the studied kimberlites, followed by
Minerals 2012, 2 5
1 calcite in the CA, TZ, An116, and AC. Phlogopite is the second most abundant mineral in the Cucumbi 2 cluster. 3
4 4.1. TZ, An116, and CA kimberlites
5 The CA kimberlite has a circular shape on the surface. This pipe exhibits crater and diatreme facies. 6 For this study, we analyzed samples to a maximum of 609 m depth. A new exploration project taking 7 samples to a depth of 800 m started in 2010 [13]. Crater facies are found up to 270 m in depth, and are 8 composed of epiclastic sandstones, coarse debris rimming the crater, and a higher concentration of 9 volcaniclastic material at a lower depth of the crater [14]. Quartz (without a reaction rim) and K- 10 feldspar are the most abundant minerals in the volcaniclastic rocks (VR). Altered crystals of garnet, 11 diopside, and rare ilmenite may be present. The diatreme facies rocks, which are more than 700 m 12 thickness, are composed of volcaniclastic kimberlite (VK) and volcaniclastic kimberlite breccias 13 (VKB), with tuffisitic kimberlite (TK) in the deepest zone. There are abundant olivine macrocrysts 14 completely replaced by serpentine and secondary carbonates. Macrocrysts of clinopyroxene, garnet, 15 ilmenite (Fe-rich, Mg-rich, and Mn-rich ilmenite), chromite, magnetite, zircon, phlogopite, hematite, 16 and amphibole also are present. Orthopyroxene has been identified, but is completely replaced by 17 bastite lizardite. The diatreme facies rocks are strongly altered all along the profile. Abundant 18 xenoliths derived from the host rocks are present, (e.g., gneiss, amphibolite, and granite); some 19 carbonatite xenoliths which could be present in the crust or beneath the kimberlite volcano; mantle- 20 derived xenoliths (i.e., garnet lherzolites, phlogopite-garnet wehrlite, and very rare eclogite) are sparse 21 and have been intensively altered. The groundmass contains lizardite, smectite, apatite, calcite, 22 ilmenite and chromite. Titanite, zirconolite, baddeleyite, barite, dolomite, witherite, barytocalcite, 23 strontianite, sulfides, and minor minerals are also widespread in the matrix. 24 The TZ kimberlite is characterized by the presence of crater and diatreme facies. The first 30 m 25 contains VR composed of quartz, hematite, K-feldspar, plagioclase, amphibole, and spinel. Diatreme- 26 facies rocks are composed of VK and VKB with macrocrysts of replaced olivine and orthopyroxene, 27 garnet, clinopyroxene, spinel (in some cases in a “necklace” shape around pellets of serpentine), 28 apatite, ilmenite, amphibole, phlogopite (some of them with inclusions of ilmenite), and zircon in a 29 groundmass of lizardite, smectite, apatite, calcite, ilmenite and chromite. Xenoliths from amphibolites 30 are very common. Mantle xenoliths are very rare, i.e., garnet lherzolite and carbonatite; and usually 31 altered. This pipe was drilled up to 310 m in depth. 32 The An116 pipe is located on a magnetic anomaly close to the Catoca area, and samples were taken 33 down to a depth of 88 m. These samples describe VR, mainly PK, in which quartz and microcline are 34 the most abundant minerals, and “mafic” xenoliths of ilmenite-calcite-phlogopite are present in the 35 first 10 m. Samples from 10 to 88 m in depth returned VK facies composed of macrocrysts of ilmenite, 36 altered clinopyroxene, phlogopite, hematite, and plagioclase, mafic xenoliths (amphibolites) and 37 metasomatized mantle-derived xenoliths i.e., phlogopite-ilmenite-clinopyroxene suites (PIC suites), 38 altered metasomatic peridotites. 39
Minerals 2012, 2 6
1 4.2. The AC4 and AC63 kimberlites
2 The AC4 and AC63 pipes are also covered for the first 50 and 100 meters, respectively, by 3 sandstone, litharenite, and arkose from the Calonda Formation, Kalahari Group, and Quaternary 4 deposits, as has been observed in most of the kimberlites in the Alto Cuilo cluster [15]. These 5 kimberlites exhibit crater- and diatreme-facies rocks. The crater-facies rocks are mainly composed of 6 VR, mainly pyroclastic rocks (PR) and resedimented volcaniclastic kimberlites (RSVK). Macrocrysts 7 of clinopyroxene, hematite, mica, microcline, and ilmenite are present in the microlitic matrix, which 8 contains clinopyroxene, mica, microcline, ilmenite, and higher content of hematite. The diatreme 9 facies rocks contain VKB composed of macrocrysts of clinopyroxene, garnet, ilmenite with 10 homogeneous and symplectitic textures, phlogopite and rare xenoliths of phlogopite-garnet wehrlite 11 and altered phlogopite peridotites. Mantle xenoliths have not been found in the AC4 pipe. The 12 groundmass is composed of serpentine, calcite, and hematite.
13 4.3. The CU79 kimberlite
14 The CU79 kimberlite exhibits two facies: crater and diatreme. The first 50 m of the crater facies 15 contains VR with ferruginous cement, interbedded with layers of lapilli. The facies then changes to 16 TK, which is generally massive, poorly sorted, and clast-supported, and is present as far down as 200 17 m. The macro- and microcrysts are composed of anhedral olivine replaced by serpentine and smectite, 18 garnet, ilmenite, clinopyroxene, and phlogopite. Some of these crystals are enclosed in a pelletal 19 assemblage of serpentine, but all the crystals have a chaotic distribution in the matrix. The matrix is an 20 interclast groundmass of serpentine, microcrystals of phlogopite, less common chlorite, smectite and 21 calcite [16]. The ilmenite texture is usually either cumulus or homogeneous. Mg-rich ilmenite (9-13 22 wt.% MgO) is present as rounded mega- and macrocrysts, as part of xenoliths, and as inclusions in 23 phlogopite. In some cases, macrocrysts of ilmenite are partially replaced along the borders by 24 perovskite and spinel. Garnet and clinopyroxene are usually present as mega- and macrocrysts, and 25 rarely as part of xenoliths. This kimberlite rarely has mantle xenoliths i.e., garnet lherzolite, 26 phlogopite-garnet wehrlite, and relatively abundant phlogopite-rich (olivine poor or absent, without 27 garnet [17]) xenoliths.
28 5. Major-Element Composition
29 The kimberlites in this suite exhibit differences in their composition and abundance of olivine, 30 garnet, and clinopyroxene depending on the location and type of kimberlite.
31 5.1. Olivine
32 Olivine is abundant in the TZ, CA, and An1116 kimberlites, forming up to 65 % volume of the total 33 mineral components. However, fresh olivine has only been found in the CA kimberlite. The grains 34 generally occur as macrocrysts (most of them anhedral) and microcrysts (subhedral to anhedral), dark 35 and light green to pale greenish white, depending on the degree of alteration. Most of the grains 36 (approximately 95%) are replaced by lizardite, with other alteration minerals i.e., calcite, smectite, 37 chlorite, magnetite, and sulfides. Olivine macrocrysts from the CA kimberlite (Figure 2) occur as: (1)
Minerals 2012, 2 7
1 homogeneous crystals, (2) crystals rimmed by "iddingsite", and (3) zoned macrocrystals. The first
2 population (Table 1, supplementary file) has an average composition of Fo92 and extremely low 3 average values for CaO (0.01 wt.%), and MnO (0.11 wt.%). The second population also has an average
4 composition of Fo92, CaO (0.01 wt.%), and MnO (0.13 rim wt.%). The third population of crystals
5 exhibits zonation with an average composition of core Fo90, middle Fo87, and rim Fo85. The average
6 values varying from core to rim are: CaO (0.02-0.04 wt.%), MnO (0.13-0.19 wt.%), Cr2O3 (0.01-0.06
7 wt.%), TiO2 (0.01-0.03 wt.%), and NiO (0.36-0.13 wt.%). Olivine macrocryst compositions are very 8 similar to archetypal cratonic peridotites, which has a mean Mg# of 92.6 and indicates melt depletion 9 [18].
10 Figure 2. CaO vs Mg# diagram for olivine macrocrysts from the CA kimberlite. Error bars 11 indicate standard deviation of CaO (wt.%).
12
13 5.2. Garnet as Macrocrysts and within Xenoliths
14 The CA and TZ kimberlite includes dark pink to dark orange anhedral macrocrysts garnet (G9 and 15 G10 after Grütter et al.[19], Figure 3). Some of the garnet macrocrysts are partially altered to chlorite, 16 or replaced by hematite and calcite along fractures. Garnet (Table 2, supplementary file) is also present 17 in eclogite (CA), garnet lherzolite with and without ilmenite (CA, TZ), and phlogopite-garnet wehrlite 18 with and without ilmenite (CA) xenoliths. Garnet macrocrysts from the CU79 kimberlite are mainly 19 pink to orange (G9). Few garnet grains are among the microcrysts. Garnet is also found in garnet
Minerals 2012, 2 8
1 lherzolite, garnet-phlogopite wehrlite xenoliths from CU79, and bimineralic associations of G9- 2 diopside. The garnet from AC kimberlite is usually found as a microcryst (AC4) or in garnet- 3 phlogopite wehrlite xenoliths (AC63). Garnet from CA, TZ, AC63, and CU79 fall in the mantle- 4 derived field (Figure 4), whereas garnet that has been found so far in the AC4 pipe indicates a crust- 5 derived origin.
6 Figure 3. Cr2O3 vs CaO diagram for garnet with superimposed isobars according to the 7 P38 barometer calculation [19]. Graphite-diamond constraint (GDC).
8 9
Minerals 2012, 2 9
1 Figure 4. Ca# vs Mg# for garnet from the TZ, CA, AC63, AC4, and CU79 kimberlites.
2
3 5.3 Clinopyroxene Mega- and Macrocrysts and in Xenoliths
4 Three populations of clinopyroxene can be identified on the basis of an Al-Cr-Na plot (Figure 5): 1) 5 clinopyroxene in a lherzolitic association; 2) low-Cr clinopyroxene that plots between Al and Na 6 extremes; and 3) Low-Na clinopyroxene. Clinopyroxene in a lherzolitic association has been found in 7 all the kimberlites studied with slight variations in the Cr content between pipes (Table 3, 8 supplementary file). Cr-rich diopside is relatively common in the CA pipe, and shows the highest 9 concentrations in Cr. Low-Cr clinopyroxene was found in the CA, some in the TZ, and the CU79 10 pipes. Low-Na clinopyroxene is present in the CA pipe with a relatively high content of Cr, whereas 11 low-Na and low-Cr clinopyroxene is found in the TZ and the CU79 pipes. All the clinopyroxene falls 12 in the “on craton” field, except for a clinopyroxene inclusion in garnet from CU79 (Figure 6). 13
Minerals 2012, 2 10
1 Figure 5. Plot of Al-Cr-Na (apfu) for clinopyroxene from the kimberlites studied. * After 2 Morris et al. [20]
3 4
Minerals 2012, 2 11
1 Figure 6. Cr2O3 vs Al2O3 diagram for clinopyroxene from the An116 (magenta), TZ 2 (cyan), CA (red), AC63 (green), and CU79 (blue) kimberlites. Classification diagram after 3 Ramsey [21]
4
5 6. Geothermobarometry
6 Comparison and evaluation of results obtained from different pressure and temperature (PT) 7 combinations were carried out. The thermobarometer combination used is based on element-exchange 8 reactions between clinopyroxene and garnet that are believed to be in equilibrium (xenoliths), which 9 provides better evaluation of PT results than calculations based on single crystals. First, we selected 10 only the most representative fresh xenoliths from the CA, TZ, CU79, and AC63 kimberlites, then we 11 proceeded to calculate P and T on the basis of EPMA data and by obtaining the average composition 12 for each mineral from each xenolith (i.e., only one average value-point for each type of xenolith). We 13 used the program PTEXL.XLT, prepared by Dr. T. Stachel [22]. The well-calibrated single- 14 clinopyroxene thermobaromether of Nimis and Taylor [23] also was applied, in order to compare 15 values of both T and P, which yielded reliable temperature estimates compared to the T-P results
Minerals 2012, 2 12
1 obtained from thermobarometry of xenoliths. We used the following calibration for the single- 2 clinopyroxene thermobaromether: (1) clinopyroxene that falls in the garnet peridotite field “on-craton”
3 defined by Ramsay [21] in the diagram Cr2O3 versus Al2O3 (Figure 6), (2) clinopyroxene above the 4 field of Low-Al peridotite (Figure 7).
5 Figure 7. Al2O3 vs MgO diagram for discrimination of no Low-Al peridotite clinopyroxene 6 type for the geothermobarometric calculations based on of single crystals of clinopyroxene 7 (after Nimis [24]).
8 9
Minerals 2012, 2 13
1 Figure 8. Pressure and temperature values calculated for xenoliths using the equation 2 Nimis and Taylor [23] (NT2000): Garnet lherzolite (diamonds) and phlogopite garnet 3 wehrlite (triangles); and for single clinopyroxene macrocrysts (cross) from the TZ (purple), 4 CA (red), AC63 (green), and CU79 (blue) kimberlites. Paleogeotherm calculated using 5 FITPLOT program (blue line, with pink and purple lines representing the error envelope). 6 P-T data calculated from mantle xenoliths from the Bultfontein kimberlite (dark gray 7 shading) and paleogeotherm (dark gray dashed line); and P-T data calculated from mantle 8 xenoliths from the Finsch kimberlite (light gray shading) and paleogeotherm (light gray 9 point line) after Mather et al. [25].
10 11 12 Calculated temperature and pressure from xenoliths is less scatter than T-P data calculated from 13 single crystals (Figure 8). However, most of the data fall within error of estimate. The calculated 14 northeastern Angola paleogeotherm fit a single value for the CA and the CU79 kimberlites. Only one 15 phlogopite-garnet lherzolite xenolith from the AC63 kimberlite was able to be used and plotted in the 16 same paleogetherm than the CA and CU79 kimberlites. The differences in T-P values between these 17 kimberlites may reflect the different way each kimberlite sampled the lithosphere. The lithospheric 18 thickness calculated from the northeastern Angola paleogeotherm yielded 192 km. A quantitative 19 comparison between Angola lithosphere and reference geotherms in southern Africa (Bultfontein and 20 Finsch kimberlites, after Mather et al. [25]), indicates a slightly cooler (steeper) paleogetherm for 21 Angola than the paleogetherms calculated from southern Africa.
Minerals 2012, 2 14
1 7. Trace-Element Chemistry
2 About one hundred sixty trace-element analyses were performed on garnet and clinopyroxene from 3 representative macrocrysts (about 80 grains) and mantle xenoliths (14 xenoliths) from the CA, TZ, and 4 CU79 kimberlites (Table 4, supplementary file). Three main different trends for garnet can be 5 identified in the Catoca kimberlite on the basis of chondrite-normalized Rare Earth Element patterns
6 (REEN) (Figure 9). (1) Garnet of eclogitic affinity with “normal” [26] REEN patterns, and slightly 7 enriched in Light Rare Earth Element (LREE). (2) Garnet from garnet lherzolite and phlogopite-garnet
8 wehrlite xenoliths can exhibit either “normal” REEN patterns or LREE-enriched patterns. Garnet from 9 phlogopite-garnet wehrlite exhibits the highest LREE-enrichment, with a maximum around the LREE-
10 HREE limit and flat HREE. (3) Garnet macrocrysts with “normal” and “sinusoidal” REEN patterns
11 [26]. Garnet macrocrysts from the TZ pipe exhibit both “normal” and "sinusoidal" REEN patterns with 12 lower HREE abundances. In contrast, garnet from xenoliths (garnet-lherzolite and phlogopite-garnet
13 wehrlite) and macrocrysts from the CU79 kimberlite follows the “normal” REEN pattern but with 14 slightly depleted values than garnet from the CA pipe. 15 The data indicate that garnet lherzolite xenoliths found in the CA and CU79 kimberlites were under 16 different equilibration conditions and different degrees of metasomatism. The xenoliths from the CA 17 kimberlite may have been generated by refertilization of a previously depleted peridotite. The 18 xenoliths from the CU79 kimberlite might be the result of a depleted source but with a very limited 19 enrichment in LREE. 20
Minerals 2012, 2 15
1 Figure 9. Chondrite-normalized REE diagrams for representative xenoliths and 2 macrocrysts from the CA, TZ, and CU79 kimberlites. The gray zone represents the 3 “normal” garnet pattern according to McLean et al. [26].
4
Minerals 2012, 2 16
1 8. Sm/Nd Isotope Results for Xenoliths Whole-Rock
2 These data are the first Sm-Nd isotope analyses carried out in xenoliths of kimberlites from Angola 3 (Table 5, supplementary file). Mantle xenoliths are potentially subject to infiltration and alteration of 4 isotopic signatures from the kimberlite [27]. Consequently, it is important to mention that Sm-Nd 5 model ages can reflect this mixing process. The mantle xenoliths from the CA kimberlite have a 6 143Nd/144Nd value between 0.511288 and 0.511681, whereas the Sm-Nd isotopes in xenoliths from the 7 CU79 kimberlite show higher 143Nd/144Nd values between 0.512274 and 0.512391, as well as a 8 narrower range of values. A single value of 0.512377 was obtained from the CU80 pipe, located 5.5
9 km SSE from the CU79 pipe, and can be used for comparison. Negative ƐNd values from xenoliths of 10 the CA kimberlite indicate an enriched mantle, whereas mantle-derived xenoliths from the CU79 pipe
11 show a slightly depleted mantle signature with positive ƐNd values as well as the sample from the 12 CU80 pipe. The Nd isotope evolution diagram (Figure 10) clearly shows different sources for the two 13 kimberlites (the CA and the CU79).
14 Figure 10. Diagram of ƐNd vs T (Ga) for xenoliths from the CA, CU79, and CU80 15 kimberlites.
16
17 9. Discussion and Conclusions
18 9.1 Discrimination among Kimberlites
19 The trans-lithosphere discontinuity of the Lucapa structure played a very important role by favoring 20 a thermal perturbation, melt production, and mantle upwelling [28,29], and in the evolution of the host 21 rocks. The integration of petrography, geochemistry, and geothermobarometric studies of the less 22 altered samples from six kimberlites in the northeasterm Angola suggests that these kimberlites 23 originated from different sources in spite of the fact that they are all located in the same tectonic 24 corridor. 25 The CA kimberlite has different compositional populations of olivine macrocrysts (homogeneous, 26 "iddingsite" rimmed, and zoning olivine) which suggests different sources for those crystals, where 27 they have likely been modified by subsequent crystallization. It is important to mention that
Minerals 2012, 2 17
1 heterogeneous crystallization is commonly found in volcanic rocks [30]. In the case of Catoca, Fe-Mg 2 zoning in olivine indicates a state of disequilibrium that reflects the physical and chemical conditions 3 that the mineral has experienced. This pipe is also characterized by the presence of pyrope and pyrope- 4 almandine (G9 and G10 according to Grütter et al. [19,31]), a relative abundance of Cr-rich diopside, 5 and the presence of eclogite, garnet lherzolite, carbonatite, and phlogopite-garnet wehrlite xenoliths. 6 Orthopyroxene is less common than clinopyroxene in this pipe. Orthopyroxene is usually 7 serpentinized. The very low amounts of orthopyroxene may be explained by rapid dissolution in the 8 kimberlitic melt during transport [32]. Considering that clinopyroxene does not occur as a liquidus 9 phase in Group I kimberlites, kimberlitic magma is the mechanism that brings mantle xenoliths to the 10 surface. The clinopyroxene equilibrates at pressures greater than 30 kbar and the Cr:Na ratio in
11 clinopyroxene from kimberlites usually is around 1. The solubility of kosmochlor (NaCrSi2O6) in 12 diopside decreases with pressure and its limit of solubility may be at about 45 kbar [32], which
13 suggests that the join jadeite (NaAlSi2O6)-kosmochlor is strongly influenced by pressure between 30 14 and 45 kbar. Then there is a difference in the Al-Na component in clinopyroxene from the CA pipe and 15 the other studied kimberlites specially the CU79. 16 The samples from the An116, AC63 and AC4 pipes are so limited that conclusions regarding their 17 origin and evolution should be taken with caution. It is clear that Cr-rich diopside is present in the 18 An116 pipe, indicating a potential “deep” source. The garnet samples from the AC4 pipe suggest 19 shallower pressure conditions, less than 20 kbar. In contrast, garnet from the AC63 pipe shows ranges 20 of pressure between 20 and 43 kbar. The chemical composition of the AC4 garnet suggests crust- 21 derived crystals, whereas garnet from xenoliths from the AC63 pipe suggests a mantle-derived source. 22 Samples from the TZ pipe have some similarities with those from the CA pipe, but garnet from the TZ 23 pipe is the only G9 type with different ranges of pressure (less than 43 kbar) and its clinopyroxene has 24 a lower Cr content. Some of the T-P values from Catoca are consistent with data previously published 25 by Aschepkov et al. [33] based on garnet and pyroxene xenocrysts.
26 9.2 The Underlying Mantle in the Northeastern Angola
27 We propose that the Catoca kimberlite was generated from a depleted source. A subsequent 28 metasomatic enrichment event (possibly more than one) incorporated incompatible LREE. Based on 29 normalization to chondrite of REE concentrations in garnet and clinopyroxene, we can track the 30 behavior of these elements in the kimberlite. If garnet equilibrates with clinopyroxene, the result is to 31 shift only LREE in garnet [34], as we observed in the "sinusoidal" pattern of garnet from phlogopite- 32 garnet wehrlite xenoliths in the CA pipe. Garnet crystals from the TZ pipe show similar patterns as that
33 of the CA pipe (i.e., "normal" and "sinusoidal" REEN patterns). This also suggests a metasomatic 34 enrichment of a previously depleted source. 35 In contrast, garnet from phlogopite-garnet wehrlite from the CU79 kimberlite shows "normal"
36 REEN patterns. This suggests different degrees of enrichment, likely due to metasomatism. Garnet 37 lherzolites from both kimberlites seem to be derived from a depleted mantle source. 38 Different Sm-Nd TD model ages from the CA and the CU79 pipes may also suggest different 39 mantle sources or metasomatic events that modified the isotopic ratios in the mantle. TDM model ages 40 from the CA kimberlite could indicate different time events than for the CU79 kimberlite. The
Minerals 2012, 2 18
1 differences in diamond production from the CA, TZ, CU79, and AC63 kimberlite pipes may be the 2 result of different mantle sources and metasomatic events, as well as independent subsequent evolution 3 of each kimberlite. 4 We interpret different sources for both kimberlites (the CA and CU79 pipes). These two kimberlites 5 have heterogeneous mantle sources, the CA kimberlite is the more enriched of the two, possibly
6 because of multiple metasomatic events that could explain the "sinusoidal" REEN patterns in garnet. 7 Both kimberlitic mantle xenoliths also have different TDM (Model ages relative to CHUR): the CA 8 kimberlite xenoliths show Mesoproterozoic ages (1220 - 1250 Ma), whereas the CU79 kimberlite 9 xenoliths yield 450 to 390 Ma (Late Ordovician to Devonian) like the TDM from the CU80 kimberlite. 10 These model ages may be interpreted as the age of mantle generation or the age of a metasomatic 11 event that modified the isotopic ratios in the mantle. TDM model ages of 1.2 Ga or the xenoliths from 12 the CA kimberlite could be associated with the Kibaran orogeny, whereas the ages from the CU79 13 kimberlite could imply more juvenile Paleozoic components possibly related to the assembly of 14 Pangea. Based on these data, these two kimberlites with different diamond production grades have a 15 different pattern of evolution despite being from the same tectonic trend, the Lucapa structure. 16 The reactivation of old deep-seated faults during the Paleoproterozoic, the Permo-Triassic, the 17 Cretaceous and the Cenozoic [29] probably is an important factor in some of the different pulses of 18 diamondiferous kimberlites. Thus petrography, geochemistry and Sm-Nd isotopic data from these 19 kimberlites provide interesting tools to recognize possible diamondiferous kimberlites in the area.
20 Acknowledgements
21 This research is funded by the CGL2006-12973 and CGL2009-13758 BTE projects of Ministerio de 22 Educación y Ciencia (Spain), and the AGAUR SGR 589 and SGR444 of the Generalitat de Catalunya. 23 The first author (SERC) received an FI grant and a BE grant, both sponsored by the Departament 24 d'Educació i Universitats de la Generalitat de Catalunya and the European Social Fund. We thank Dr. 25 D.G. Pearson and a second anonymous reviewer for their revision of this manuscript and all their 26 valuable comments. We acknowledge the Geological Survey of Canada (GSC), Ottawa, for all of the 27 support during a six-month Volunteer Assistant visit of SERC, especially to Dr. S.E. Jackson who gave 28 us all the support and guidance for carrying out the LA-ICP-MS analyses. SERC thanks Dr. T. Stachel 29 for providing the PTEXL.XLT program and guidance in the application of the geothermobarometers. 30 SERC also thanks Dr. K.A. Mather who kindly helped her to calculate the Angola paleogeotherm. The 31 authors also acknowledge the Electron Microprobe Laboratory, Department of Earth and Planetary 32 Sciences, McGill University, especially to Mr. Lang Shi for assistance in the use of EPMA. The 33 authors also thank Dr. Robert Martin, emeritus professor at the Earth & Planetary Sciences 34 Department, McGill University, who kindly arranged everything to acquire the EPMA analyses at the 35 McGill University and made valuable improvements to the preliminary version of this manuscript. 36 Thanks to the Pacific Centre for Isotopic and Geochemical Research, Department of Earth and Ocean 37 Sciences, University of British Columbia, Vancouver, especially to Dr. B. Kieffer for all his 38 collaboration in the developing the Sm-Nd analyses. We thank ENDIAMA (Empresa Nacional de 39 Diamantes de Angola), which kindly allowed SERC to acquire samples for her Ph.D. thesis and 40 allowed the use of all facilities for the mine trip, especially to M. Watangua (former Chief Geologist)
Minerals 2012, 2 19
1 and Dr. V. Pervov (petrologist). Also thank to Dr. A. Gonçalves, professor at the Universidade 2 Agostinho Neto, Angola, who helped in all the process of logistics and develop the field trip. The 3 authors also thank the Serveis Cientificotècnics de la Universitat de Barcelona for assistance in the use 4 of SEM/ESEM-BSE-EDS analyses (E. Prats. and J. García Veigas).
5 References and Notes
6 1. ENDIAMA–Empresa Nacional de Diamantes de Angola, E.P. História. Available online: 7 http://www.endiama.co.ao/endiama_historia.php# (accessed on 8 July 2012). 8 2. Reis, B. Preliminary note on the distribution and tectonic control of kimberlites in Angola. In 9 Proceedings of the 24th Int. Geol. Congr., Montreal, Canada, August-September 1972; Rep. Sess. 10 4, 276–281. 11 3. The Israeli Diamond Industry. Diamond news. Available online: 12 http://www.israelidiamond.co.il/english/news.aspx?boneid=918&objid=9920 (accessed on 25 13 September 2011). 14 4. Mitchell, R.H. Kimberlites, Orangeites, and Related Rocks; Plenum Press, New York, 1995; pp 15 410. 16 5. Scott Smith, B. H.; Nowicki, T. E.; Russell, J. K.; Webb, K. J.; Mitchell, R. H.; Hetman, C. M.; 17 Robey, J. V. A.; Skinner, E. M. W.; Robey, J. V. Kimberlite terminology and classification. In 18 Proceedings of the 10th Kimberlite Conf., Bangalore, India, 6-11 February 2012; Abstract 19 Volume, 30–31. 20 6. De Carvalho, H.; Tassinari, C.; Alves, P. H. Geochronological review of the Precambrian in 21 western Angola: links with Brazil. J. Afr. Earth Sci. 2000, 31, 383–402. 22 7. Guiraud, R.; Bosworth, W.; Thierry, J.; Delplanque, A. Phanerozoic geological evolution of 23 Northern and Central Africa: an overview. J. Afr. Earth Sci. 2005, 43, 83–143. 24 8. Perevalov, O.V.; Voinovsky, A.S.; Tselikovsky, A.F.; Agueev, Y.L.; Polskoi, F.R.; Khódirev, 25 V.L.; Kondrátiev, A.I. Geologia de Angola. Notícia explicativa da carta geológica a escala 26 1:1.000.000. Serviço Geológico de Angola 1992. Lunda. 27 9. Egorov, K. N.; Roman'ko, E. F.; Podvysotsky, V. T.; Sablukov, S. M.; Garanin, V. K.; D'yakonov, 28 D. B. New data on kimberlite magmatism in southwestern Angola. Russ. Geol. Geophys. 2007, 29 48, 323–336. 30 10. Robles-Cruz, S.E.; Escayola, M.; Jackson, S.; Galí, S.; Pervov, V.; Watangua, M.; Gonçalves, A.; 31 Melgarejo, J.C. U-Pb SHRIMP geochronology of zircon from the Catoca kimberlite, Angola: 32 Implications for diamond exploration. Chem. Geol. 2012, (310-311), 137–147. 33 11. Eley, R.; Grütter, H.; Louw, A.; Tunguno, C.; Twidale, J. Exploration Geology of the Luxinga 34 kimberlite Cluster (Angola) with evidence supporting the presence of kimberlite lava. In 35 Proceedings of the 9th Kimberlite Conf., Frankfurt, Germany, 10-15 August 2008; Extended 36 Abstr. No. 9IKC-A-00166. 37 12. Weis, D.; Kieffer, B.; Maerschalk, C.; Barling, J.; de Jong, J.; Williams, G. A.; Hanano, D.; 38 Pretorius, W.; Mattielli, N.; Scoates, J. S.; Goolaerts, A.; Friedman, R. M.; Mahoney, J. B. High- 39 precision isotopic characterization of USGS reference materials by TIMS and MC-ICP-MS. 40 Geochemistry, Geophysics, Geosystems 2006, 7(8), Q08006.
Minerals 2012, 2 20
1 13. Pervov, V.A.; Somov, S.V.; Korshunov, A.V.; Dulapchii, E.V.; Félix, J.T. The Catoca kimberlite 2 pipe, Republic of Angola: A paleovolcanological model. Geol. Ore Deposits 2011, 53(4), 295– 3 308. 4 14. Robles-Cruz, S.; Watangua, M.; Melgarejo, J. C.; Galí, S. New Insights into the Concept of 5 Ilmenite as an Indicator for Diamond Exploration, Based on Kimberlite Petrographic Analysis. 6 Revista de la sociedad española de mineralogía, Macla 2008, 9, 205–206. 7 15. Pettit, W. geophysical signatures of some recently discovered large (>40 ha) kimberlite pipes on 8 the Alto Cuilo concession in northeastern Angola. Lithos 2009, 112S, 106–115. 9 16. Robles-Cruz, S.; Watangua, M.; Isidoro, L.; Melgarejo, J. C.; Galí, S.; Olimpio, A. Contrasting 10 compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in 11 exploration for diamond. Lithos 2009, 112S, 966–975. 12 17. Pearson, D.G.; Canil, D.; Shirey, S.B. Mantle samples included in volcanic rocks: xenoliths and 13 diamonds. In The mantle and core. Treatise on Geochemistry; Carlson, R.W., Ed.; Elsevier: 14 Washington, USA, 2007; Volume 2, pp.171–275. 15 18. Pearson, D.G.; Wittig, N. Formation of archean continental lithosphere and its diamonds: the root 16 of the problem. J. Geol. Soc. (London, U. K.) 2008, 165, 1–20. 17 19. Grütter, H.S.; Gurney, J.J.; Menzies, A.H.; Winter, F. An updated classification scheme for 18 mantle-derived garnet, for use by diamond explorers. Lithos 2004, 77, 841–857. 19 20. Morris, T.F.; Sage, R.P.; Ayer, J.A.; Crabtree, D.C. A study in clinopyroxene composition: 20 implications for kimberlite exploration. Geochem.: Explor., Environ., Anal. 2002, 2 (4), 321–331. 21 21. Ramsay, R. R. “Geochemistry of diamond indicator minerals”. Ph.D. diss. University of Western 22 Australia, Perth 1992. 23 22. Stachel, T. Developments in geothermobarometry of mantle rocks based on the PTEXL written by 24 Thomas Köhler in 1994. Unpublished data. 2011. 25 23. Nimis, P.; Taylor, W. R. Single clinopyroxene thermobarometry for garnet peridotites. Part I. 26 Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in-Cpx thermometer. Contrib. 27 Mineral. Petrol. 2000, 139, 514–554. 28 24. Nimis, P. Evaluation of diamond potential from the composition of peridotitic chromian diopside. 29 Eur. J. Mineral. 1998, 10, 505–519. 30 25. Mather, K.A.; Pearson, D.G.; McKenzie, D.; Kjarsgaard, B.; Priestley, K. Constraining the depth 31 and thermal history of cratonic lithosphere using peridotite xenolith and xenocryst 32 thermobarometry and seismology. Lithos 2011, 125, 729–742. 33 26. McLean, H.; Banas, A.; Creighton, S.; Whiteford, S.; Luth, R.W.; Stachel, T. Garnet xenocrysts 34 from the Diavik mine, NWT, Canada: Composition, color, and paragenesis. Can. Mineral. 2007, 35 45, 1131–1145. 36 27. Pearson, D. G. The age of continental roots. Lithos 1999, 48, 171–194. 37 28. White, S. H.; De Boorder, H.; Smith, C. B. Structural Controls of Kimberlite and Lamproite 38 Emplacement. J. Geochem. Explor. 1995, 53, 245–264. 39 29. Jelsma, H.; Barnett, W.; Richards, S.; Lister, G. Tectonic setting of kimberlites. Lithos 2009, 112, 40 155–165. 41 30. Brett, R. C.; Russell, J. K.; Moss, S. Origin of olivine in kimberlite: Phenocryst or impostor? 42 Lithos, 2009, 112S, 201–212.
Minerals 2012, 2 21
1 31. Grütter, H.; Latti, D.; Menzies, A. Cr-saturation arrays in concentrate garnet compositions from 2 kimberlite and their use in mantle barometry. J. Petrol. 2006, 47, 801–820. 3 32. Vredevoogd, J. J.; Forbes, W. C. The system diopside–ureyite at 20 kb. Contrib. Mineral. Petrol. 4 1975, 52, 147–156. 5 33. Aschepkov, I.V.; Rotman; Somov, S.V.; Afanasiev, V.P.; Downes, H.; Logvinova, A.M.; 6 Nossyko, S.; Shimupi, J.; Palessky, S.V.; Khmelnikova, O.S.; Vladykin, N.V. Composition and 7 thermal structure of the lithospheric mantle beneath kimberlite pipes from the Catoca cluster, 8 Angola. Tectonophysics 2012, (530-531), 128–151. 9 34. Stachel, T.; Viljoen, K.S.; Brey, G.; Harris, J.W. Metasomatic processes in lherzolitic and 10 harzburgitic domains of diamondiferous lithospheric mantle. Earth Planet. Sci. Lett. 1998, 159, 1– 11 12.
12 © 2012 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article 13 distributed under the terms and conditions of the Creative Commons Attribution license 14 (http://creativecommons.org/licenses/by/3.0/).
PAPER VI
Supplementary Files
Tables
Table 1. Summary of EPMA data from olivine macrocrysts Kimberlite CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA CA Borehole 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 Olivine macrocryst111112A2A2A2B2B2B3A3A3A3B3B3B3C3C3C Point 506.3-I-Ol-2 506.3-I-Ol-5 506.3-I-Ol-6 506.3-I-Ol-7 506.3-I-Ol-8 505B_e-Ol36 505B_e-Ol41 505B_e-Ol43 505B_a-Ol1 505B_a-Ol7 505B_a-Ol10 505.6-b-ol-1 505.6-b-ol-2 505,6 C OL 34 505,6 e oli 60 505,6 e ol 61 505,6 e ol 65 505.6-b-ol-11 505,6 C OL 31 505,6 C OL 32 (wt.%) SiO2 41.16 41.15 40.95 41.04 41.34 41.08 41.18 40.88 41.04 41.20 41.04 40.35 40.03 40.16 39.55 39.17 39.13 38.69 39.13 39.23 TiO2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.02 0.01 0.02 0.03 0.05 0.04 0.03 0.07 Al2O3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.02 0.00 Cr2O3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.01 0.03 0.03 0.03 0.07 0.12 0.11 0.11 Fe2O3 0.22 0.33 0.61 0.47 0.00 0.00 0.00 0.37 0.20 0.00 0.00 ————————— FeO 7.45 7.37 7.13 7.17 7.66 7.94 7.91 7.64 7.89 8.15 8.04 ————————— FeOT 7.66 7.69 7.74 7.65 7.66 7.94 7.91 8.01 8.09 8.15 8.04 10.00 10.00 8.39 11.99 11.83 12.21 14.63 14.54 14.38 MnO 0.11 0.10 0.11 0.10 0.12 0.15 0.13 0.12 0.13 0.14 0.15 0.16 0.15 0.07 0.12 0.18 0.15 0.19 0.20 0.14 MgO 51.03 51.09 51.02 51.10 50.86 50.46 50.53 50.57 50.59 50.46 50.43 49.05 49.08 50.74 47.43 47.67 47.37 45.72 45.58 45.61 CaO 0.01 0.01 0.01 0.00 0.02 0.01 0.00 0.02 0.03 0.02 0.02 0.02 0.01 0.00 0.06 0.07 0.02 0.03 0.04 0.03 NiO ———————————0.35 0.33 0.41 0.35 0.39 0.30 0.15 0.18 0.08 Sum Ox% 99.97 100.05 99.83 99.88 100.01 99.64 99.75 99.60 99.88 99.97 99.68 99.99 99.62 99.82 99.55 99.39 99.28 99.58 99.83 99.63 (apfu) Si 1.00 1.00 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.99 0.99 0.98 0.99 0.98 0.98 0.98 0.99 0.99 Ti 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Al total 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Cr 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Fe3 0.00 0.01 0.01 0.01 0.00 0.00 0.00 0.01 0.00 0.00 0.00 ————————— Fe2 0.15 0.15 0.14 0.15 0.16 0.16 0.16 0.16 0.16 0.17 0.16 ————————— FeT 0.15 0.16 0.16 0.15 0.16 0.16 0.16 0.16 0.16 0.17 0.16 0.21 0.21 0.17 0.25 0.25 0.26 0.31 0.31 0.30 Mn 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Mg 1.84 1.85 1.85 1.85 1.84 1.83 1.83 1.84 1.83 1.83 1.83 1.80 1.81 1.85 1.76 1.78 1.77 1.72 1.71 1.71 Ca 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ni ———————————0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.00 0.00
%Tephroite 0.11 0.10 0.11 0.10 0.13 0.15 0.13 0.13 0.13 0.15 0.15 0.15 0.15 0.10 0.15 0.20 0.15 0.20 0.20 0.15 %Forsterite 92.14 92.13 92.10 92.21 92.07 91.74 91.80 91.74 91.63 91.54 91.63 89.58 89.63 91.41 87.46 87.62 87.23 84.62 84.65 84.83 %Fayalite 7.74 7.75 7.78 7.69 7.78 8.10 8.06 8.11 8.20 8.29 8.19 10.27 10.22 8.49 12.39 12.18 12.62 15.18 15.15 15.02 Ca-Ol 0.01 0.02 0.01 0.00 0.03 0.01 0.00 0.03 0.03 0.02 0.02 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 mg # 92.25 92.24 92.21 92.30 92.21 91.89 91.93 91.88 91.79 91.69 91.79 89.72 89.76 91.50 87.59 87.80 87.36 84.79 84.82 84.96 Texture types : (1) homogeneous; (2A) crystal with "iddingsite"_core; (2B) crystal with "iddingsite"_rim; (3A) zoned_core; (3B) zoned_middle; (3C) zoned_rim Table 2a. Summary of EPMA data from garnet Kimberlite AC4AC4AC63AC63CACACACACACACACACA Borehole 4 4 6 6 335 335 335 77/35 33/35 33/35 33/35 335 335 Texture/association5533112233445 Point 159A_a10_Grt 159A_b5_Grt 206B-a-Grt5 206B-a-Grt9 607B-a_grt2 607B-i_grt24 607C-C-Grt?7 398.2_b-Grt1 499,6-B-Gt-2 499,6-B-Gt-6 460_a-Grt2 551- ag-82 551- c-71 (wt.%) SiO2 37.36 37.22 41.57 41.63 41.89 41.97 41.54 41.74 41.63 41.56 41.83 41.97 41.82 TiO2 0.09 0.02 0.09 0.08 0.08 0.12 0.08 0.14 0.28 0.25 0.14 0.18 0.78 Al2O3 21.36 22.07 20.41 20.18 23.31 23.57 21.06 20.42 20.25 20.37 21.36 20.57 21.41 Cr2O3 0.05 0.04 5.07 5.16 0.03 0.04 4.05 5.03 4.46 4.33 3.58 3.97 0.68 Fe2O3 1.93 3.30 0.44 0.38 0.45 0.00 0.30 0.00 0.00 0.02 0.00 0.94 0.72 FeO 26.48 25.93 7.52 7.62 11.34 11.89 8.10 7.77 7.74 7.83 8.25 6.74 10.83 MnO 0.93 1.17 0.42 0.43 0.34 0.33 0.51 0.44 0.40 0.40 0.44 0.47 0.39 MgO 3.99 8.88 20.65 20.58 19.34 18.89 19.94 19.65 20.48 20.34 19.31 21.64 19.26 CaO 7.98 1.23 3.93 3.99 3.12 3.06 4.37 4.91 4.11 4.23 5.08 3.56 4.02 ZrO2 ————————————— Y2O3 ————————————— V2O3 ————————————— Total 100.16 99.86 100.10 100.04 99.90 99.87 99.95 100.10 99.34 99.33 99.99 100.03 99.91 (apfu) Si 2.94 2.89 2.98 2.99 3.00 3.01 2.98 3.00 3.00 3.00 3.00 2.99 3.01 Ti 0.01 0.00 0.01 0.00 0.00 0.01 0.00 0.01 0.02 0.01 0.01 0.01 0.04 Al 1.98 2.02 1.72 1.71 1.97 1.99 1.78 1.73 1.72 1.73 1.81 1.73 1.82 Cr 0.00 0.00 0.29 0.29 0.00 0.00 0.23 0.29 0.25 0.25 0.20 0.22 0.04 Fe3 0.11 0.19 0.02 0.02 0.02 0.00 0.02 0.00 0.00 0.00 0.00 0.05 0.04 Fe2 1.74 1.68 0.45 0.46 0.68 0.71 0.49 0.47 0.47 0.47 0.50 0.40 0.65 Mn 0.06 0.08 0.03 0.03 0.02 0.02 0.03 0.03 0.02 0.02 0.03 0.03 0.02 Mg 0.47 1.03 2.21 2.20 2.06 2.02 2.13 2.11 2.20 2.19 2.07 2.30 2.07 Ca 0.67 0.10 0.30 0.31 0.24 0.24 0.34 0.38 0.32 0.33 0.39 0.27 0.31 Zr ————————————— Y ————————————— V ————————————— almandine 59.16 58.25 15.10 15.29 22.61 23.87 16.28 15.69 15.50 15.68 16.62 13.38 21.36 pyrope 15.89 35.55 73.93 73.59 68.72 67.60 71.43 70.72 73.15 72.64 69.36 76.62 67.71 grossular 21.51 3.23 8.55 8.65 7.86 7.84 9.87 10.86 9.11 9.43 11.75 7.78 9.53 spessartine 2.10 2.66 0.86 0.86 0.69 0.66 1.03 0.89 0.81 0.82 0.90 0.95 0.78 uvarovite 0.03 0.00 1.42 1.48 0.01 0.01 1.27 1.79 1.35 1.34 1.32 1.01 0.20 andradite 1.24 0.31 0.12 0.10 0.10 0.00 0.09 0.00 0.00 0.00 0.00 0.23 0.21 Ca-Ti Gt 0.06 0.00 0.03 0.02 0.02 0.02 0.02 0.05 0.08 0.07 0.05 0.04 0.22 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Texture/association : (1) eclogite; (2) grt lherzolite; (3) phl-grt wehrlite; (4) macrocryst; (5) microcryst. Table 2b. Continuation Kimberlite CA CU79 CU79 CU79 CU79 CU79 CU79 TZ TZ TZ TZ TZ TZ Borehole 335 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 34 34 G18 G18 G18 G18 Texture/association5442233224455 Point 01 -3dia-m'-57 111,5_f_4 ,8REP_a5_grt 134B_b_6 134B_b_9 d4_grt_MP 77,6B_g_2 346.3-a_grt2 346.3-a_grt3 38-a-5 38-a-46 38-c-54 38-c-58 (wt.%) SiO2 41.24 42.04 40.31 40.46 41.43 40.72 42.44 41.46 41.54 42.58 42.34 42.49 41.74 TiO2 0.34 0.64 0.72 0.18 0.15 0.13 0.55 0.22 0.23 0.82 0.84 0.11 0.14 Al2O3 21.43 21.26 19.62 22.71 22.33 21.57 21.21 22.55 22.50 21.15 21.18 23.70 23.61 Cr2O3 2.66 1.70 3.55 1.38 1.40 2.42 2.31 1.47 1.42 1.23 1.30 0.42 0.40 Fe2O3 2.17 1.68 4.49 3.96 1.87 3.31 0.65 2.95 2.47 0.37 0.71 0.45 2.12 FeO 6.57 6.40 4.99 6.01 8.19 7.12 6.21 6.91 7.34 8.23 7.96 6.62 5.17 MnO 0.39 0.23 0.32 0.43 0.41 0.53 0.22 0.45 0.50 0.27 0.24 0.28 0.32 MgO 20.57 21.56 20.45 20.44 19.95 19.50 21.79 20.62 20.54 21.16 21.16 21.61 21.78 CaO 4.67 4.51 5.53 4.42 4.30 4.99 4.66 4.43 4.24 4.24 4.26 4.28 4.46 ZrO2 ———————0.04 0.02 ———— Y2O3 ————————————— V2O3 ————————————— Total 100.05 100.02 99.98 99.99 100.03 100.29 100.03 101.11 100.79 100.05 99.99 99.95 99.74 (apfu) Si 2.95 2.98 2.90 2.89 2.96 2.92 3.00 2.93 2.94 3.03 3.01 2.99 2.94 Ti 0.02 0.03 0.04 0.01 0.01 0.01 0.03 0.01 0.01 0.04 0.04 0.01 0.01 Al 1.80 1.78 1.67 1.91 1.88 1.82 1.77 1.88 1.88 1.77 1.78 1.96 1.96 Cr 0.15 0.10 0.20 0.08 0.08 0.14 0.13 0.08 0.08 0.07 0.07 0.02 0.02 Fe3 0.12 0.09 0.24 0.21 0.10 0.18 0.03 0.16 0.13 0.02 0.04 0.02 0.11 Fe2 0.39 0.38 0.30 0.36 0.49 0.43 0.37 0.41 0.44 0.49 0.47 0.39 0.31 Mn 0.02 0.01 0.02 0.03 0.02 0.03 0.01 0.03 0.03 0.02 0.01 0.02 0.02 Mg 2.19 2.28 2.20 2.18 2.13 2.09 2.30 2.17 2.17 2.24 2.24 2.27 2.29 Ca 0.36 0.34 0.43 0.34 0.33 0.38 0.35 0.34 0.32 0.32 0.32 0.32 0.34 Zr ———————0.00 0.00 ———— Y ————————————— V ————————————— almandine 13.25 12.59 10.21 12.38 16.48 14.58 12.11 13.88 14.71 15.93 15.50 13.00 10.34 pyrope 73.90 75.59 74.63 75.06 71.59 71.22 75.80 73.81 73.39 73.03 73.41 75.67 77.59 grossular 10.41 10.12 11.23 10.08 10.08 11.13 10.50 0.00 0.56 9.78 9.76 10.49 10.65 spessartine 0.80 0.46 0.66 0.90 0.84 1.10 0.43 0.91 1.01 0.53 0.47 0.56 0.65 uvarovite 0.87 0.54 1.36 0.41 0.42 0.84 0.77 4.00 3.89 0.38 0.40 0.12 0.12 andradite 0.67 0.51 1.64 1.12 0.54 1.09 0.20 7.63 6.44 0.11 0.21 0.13 0.61 Ca-Ti Gt 0.11 0.19 0.26 0.05 0.04 0.04 0.17 0.24 0.25 0.03 0.04 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.23 100.00 100.00 100.00 100.00 100.00 Texture/association : (1) eclogite; (2) grt lherzolite; (3) phl-grt wehrlite; (4) macrocryst; (5) microcryst. Table 3a. Summary of EPMA data from clinopyroxene Kimberlite AC63 AC63 AC63 AC63 CA CA CA CA CA CA CA CA CA CA CA CA CA CA An116 An116 Borehole 666633/35 33/35 536 384 335 335 335 335 335 77-35 33/35 33/35 33/35 33/35 116 116 Texture/association 885510108899112211115566 Point 9A-a-Cpx8 A-a-Cpx15 6B-a-Cpx5 6B-a-Cpx8 505.6-b-14 505.6-b-15 304-j-35 384-m-40 601-a-110 601 a-113B_d-Cpx21 B_j-Cpx32 C-3CPX8.2-d-Cpx2 B_d-Cpx1 B_d-Cpx3 6-B-Cpx-4 6-B-Cpx-5 8A_c-Cpx4 8A_c-Cpx6 (wt.%) SiO2 54.70 54.45 54.86 55.00 54.19 53.91 55.08 55.29 55.04 55.01 55.37 55.49 54.72 54.74 54.36 54.61 54.32 54.41 54.71 54.52 Al2O3 2.24 2.30 4.52 4.48 1.04 1.07 3.77 2.66 1.78 2.33 7.30 6.69 3.38 3.17 2.95 3.11 3.66 3.70 0.76 0.76 TiO2 0.28 0.29 0.12 0.12 0.11 0.13 0.49 0.21 0.26 0.09 0.43 0.36 0.11 0.19 0.11 0.09 0.31 0.30 0.07 0.07 FeO 0.71 0.48 0.00 0.00 0.32 0.00 3.82 2.17 1.92 1.46 0.83 1.27 0.64 0.20 0.13 0.42 0.09 0.45 0.00 0.00 Fe2O3 2.48 2.88 2.82 2.40 0.00 0.00 1.93 0.27 1.76 1.31 2.55 2.15 1.93 2.36 2.44 2.07 2.81 2.49 2.69 3.68 Cr2O3 2.84 2.98 4.55 4.46 1.99 2.02 0.28 2.90 1.83 2.24 0.09 0.12 3.04 2.89 2.90 2.83 3.51 3.87 3.27 3.11 MgO 15.16 15.05 13.17 13.23 16.12 16.81 16.16 14.73 15.06 15.10 12.78 13.07 14.94 15.39 14.97 15.03 15.08 14.69 15.48 15.69 MnO 0.09 0.09 0.07 0.10 0.05 0.07 0.18 0.13 0.07 0.12 0.05 0.05 0.07 0.08 0.07 0.07 0.10 0.09 0.07 0.06 CaO 19.02 18.76 14.95 14.91 22.03 21.99 15.72 18.69 19.88 19.75 15.79 16.27 18.10 17.74 18.85 18.87 16.13 15.95 20.52 20.65 Na2O 2.89 2.99 4.91 4.86 1.58 1.61 2.88 2.97 2.52 2.60 4.89 4.56 3.22 3.27 3.02 3.00 3.76 3.89 2.45 2.47 K2O 0.02 0.03 0.02 0.02 0.08 0.11 0.00 0.01 0.02 0.00 0.02 0.02 0.02 0.01 0.02 0.01 0.03 0.05 0.02 0.02 Sum Ox% 100.19 100.12 100.02 99.93 99.72 100.50 100.13 99.98 99.97 99.87 99.96 99.95 100.02 99.91 99.70 100.01 99.58 99.73 99.93 99.98 (apfu) Si 1.97 1.97 1.97 1.98 2.00 2.00 1.98 2.00 1.99 1.99 1.97 1.98 1.97 1.97 1.97 1.97 1.96 1.96 1.98 1.97 Al 0.10 0.10 0.19 0.19 0.05 0.05 0.16 0.11 0.08 0.10 0.31 0.28 0.14 0.13 0.13 0.13 0.16 0.16 0.03 0.03 Ti 0.01 0.01 0.00 0.00 0.00 0.00 0.01 0.01 0.01 0.00 0.01 0.01 0.00 0.01 0.00 0.00 0.01 0.01 0.00 0.00 Fe2+ 0.02 0.01 0.00 0.00 0.01 0.00 0.11 0.07 0.06 0.04 0.02 0.04 0.02 0.01 0.00 0.01 0.00 0.01 0.00 0.00 Fe3+ 0.07 0.08 0.08 0.06 0.00 0.00 0.05 0.01 0.05 0.04 0.07 0.06 0.05 0.06 0.07 0.06 0.08 0.07 0.07 0.10 Cr 0.08 0.08 0.13 0.13 0.06 0.06 0.01 0.08 0.05 0.06 0.00 0.00 0.09 0.08 0.08 0.08 0.10 0.11 0.09 0.09 Mg 0.81 0.81 0.71 0.71 0.89 0.93 0.86 0.79 0.81 0.81 0.68 0.69 0.80 0.83 0.81 0.81 0.81 0.79 0.84 0.85 Mn 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ca 0.73 0.73 0.58 0.57 0.87 0.88 0.60 0.72 0.77 0.77 0.60 0.62 0.70 0.68 0.73 0.73 0.62 0.62 0.80 0.80 Na 0.20 0.21 0.34 0.34 0.11 0.12 0.20 0.21 0.18 0.18 0.34 0.32 0.22 0.23 0.21 0.21 0.26 0.27 0.17 0.17 K 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Sum cat 4.00 4.00 4.00 4.00 4.07 4.13 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00
Wo 46.77 46.81 44.92 44.75 49.28 48.46 38.17 45.73 46.97 47.14 46.15 45.91 45.97 45.13 47.39 47.05 43.39 43.41 48.78 48.62 En 51.86 52.25 55.08 55.25 50.17 51.54 54.59 50.14 49.50 50.14 51.97 51.31 52.77 54.47 52.36 52.14 56.43 55.63 51.22 51.38 Fs 1.37 0.93 0.00 0.00 0.55 0.00 7.24 4.13 3.53 2.72 1.88 2.79 1.26 0.40 0.25 0.81 0.18 0.96 0.00 0.00 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Q 90.62 89.81 82.73 83.45 97.50 97.47 88.21 92.91 92.98 92.32 77.70 79.98 88.59 88.42 88.91 89.16 86.11 86.33 93.90 92.54 Jd 5.49 5.67 12.35 12.34 2.50 2.53 8.88 6.65 4.30 5.65 18.24 16.61 8.37 7.84 7.25 7.61 9.32 9.57 1.87 1.83 Ae 3.89 4.53 4.92 4.22 0.00 0.00 2.91 0.43 2.72 2.03 4.07 3.41 3.05 3.74 3.84 3.24 4.57 4.10 4.22 5.63 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Texture/association: (1) eclogite; (2) grt lherzolite; (3) lherzolite; (4) pyroxenite; (5) phl-grt wehrlite; (6) phl-rich suite; (7) megacryst; (8) macrocryst; (9) microcryst ; (10) as inclusion in chr; (11) as inclusion in grt. Table 3b. Continuation Kimberlite CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 TZ TZ TZ TZ Borehole MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 G10 G18 34 34 Texture/association 228811779556644338822 Point 178-b13 175.5-h1 12,6B_d_61C_d1_px1EP_a12_Px 134B_i_2 136,8_b_3 105,5_c_6 1_cpx_MP 63_c cpx3 111_c1 111_c8 111_a11 111_a13 ,5B_a5_pxB_e_cpx 11 95 f-106 38-e-24 6.3-a_cpx3 6.3-a_cpx4 (wt.%) SiO2 53.79 55.06 53.59 54.14 50.53 54.68 54.83 54.03 54.13 53.46 53.29 53.14 53.32 53.35 54.23 53.34 55.41 55.69 54.74 54.79 Al2O3 7.38 2.39 1.45 3.53 5.32 3.52 3.69 4.09 2.33 1.12 1.45 1.55 1.09 0.99 2.44 1.18 2.90 3.42 2.95 3.19 TiO2 0.35 0.15 0.79 0.39 0.72 0.39 0.32 0.51 0.12 1.17 1.51 1.30 0.47 0.34 0.21 0.24 0.20 0.26 0.18 0.23 FeO 1.70 2.81 3.28 2.88 2.76 2.88 3.00 2.64 1.00 2.28 4.09 3.46 4.20 5.24 2.90 1.37 1.35 2.24 0.00 0.00 Fe2O3 1.55 0.59 2.69 2.81 4.74 2.67 2.61 3.56 1.74 2.83 1.54 2.70 2.44 1.92 2.30 3.27 1.08 0.51 4.92 4.42 Cr2O3 0.13 1.33 0.45 0.50 1.05 0.47 0.57 0.06 1.48 0.45 0.19 0.19 0.07 0.03 0.59 0.51 1.39 1.80 0.83 0.85 MgO 13.50 18.28 18.31 17.85 17.56 17.74 18.22 15.36 16.38 17.82 16.63 17.71 16.20 16.96 15.59 16.19 15.23 14.36 16.04 15.64 MnO 0.06 0.03 0.20 0.14 0.39 0.12 0.17 0.14 0.08 0.17 0.14 0.12 0.16 0.13 0.10 0.06 0.00 0.09 0.00 0.00 CaO 17.88 17.62 18.43 15.55 15.84 15.21 14.31 16.31 20.83 19.74 19.83 18.98 21.47 20.50 19.67 22.58 19.77 18.43 20.58 20.02 Na2O 3.41 1.70 1.09 2.22 1.23 2.50 2.56 3.02 1.69 1.18 1.25 1.13 0.74 0.48 1.95 1.03 2.70 3.29 2.68 2.89 K2O 0.03 0.03 0.00 0.01 0.20 0.01 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.01 0.02 0.00 Sum Ox% 99.62 99.92 100.01 99.99 100.05 99.92 100.03 99.37 100.01 99.98 99.99 100.00 100.05 99.98 99.75 99.59 99.96 100.04 100.72 100.51 (apfu) Si 1.93 1.98 1.94 1.95 1.83 1.96 1.96 1.96 1.96 1.94 1.95 1.93 1.96 1.96 1.97 1.95 1.99 2.00 1.95 1.96 Al 0.31 0.10 0.06 0.15 0.23 0.15 0.16 0.17 0.10 0.05 0.06 0.07 0.05 0.04 0.10 0.05 0.12 0.14 0.12 0.13 Ti 0.01 0.00 0.02 0.01 0.02 0.01 0.01 0.01 0.00 0.03 0.04 0.04 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.01 Fe2+ 0.05 0.08 0.10 0.09 0.08 0.09 0.09 0.08 0.03 0.07 0.13 0.11 0.13 0.16 0.09 0.04 0.04 0.07 0.00 0.00 Fe3+ 0.04 0.02 0.07 0.08 0.13 0.07 0.07 0.10 0.05 0.08 0.04 0.07 0.07 0.05 0.06 0.09 0.03 0.01 0.13 0.12 Cr 0.00 0.04 0.01 0.01 0.03 0.01 0.02 0.00 0.04 0.01 0.01 0.01 0.00 0.00 0.02 0.01 0.04 0.05 0.02 0.02 Mg 0.72 0.98 0.99 0.96 0.95 0.95 0.97 0.83 0.88 0.96 0.91 0.96 0.89 0.93 0.84 0.88 0.82 0.77 0.85 0.83 Mn 0.00 0.00 0.01 0.00 0.01 0.00 0.01 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Ca 0.69 0.68 0.72 0.60 0.62 0.58 0.55 0.63 0.81 0.77 0.78 0.74 0.84 0.81 0.77 0.89 0.76 0.71 0.78 0.77 Na 0.24 0.12 0.08 0.15 0.09 0.17 0.18 0.21 0.12 0.08 0.09 0.08 0.05 0.03 0.14 0.07 0.19 0.23 0.18 0.20 K 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Sum cat 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00 4.00
Wo 47.07 38.94 39.66 36.47 37.33 36.09 34.07 41.04 46.91 42.62 42.96 40.97 45.40 42.54 45.09 48.90 47.06 45.90 47.98 47.92 En 49.44 56.21 54.82 58.25 57.58 58.57 60.36 53.77 51.33 53.54 50.12 53.19 47.66 48.97 49.72 48.78 50.44 49.75 52.02 52.08 Fs 3.49 4.84 5.51 5.27 5.08 5.33 5.57 5.19 1.76 3.84 6.92 5.84 6.93 8.49 5.19 2.32 2.50 4.35 0.00 0.00 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Q 80.50 93.70 93.03 87.92 82.21 88.00 87.70 85.02 92.14 93.50 94.51 92.79 94.19 95.19 91.03 92.78 91.41 90.69 86.49 86.33 Jd 17.20 5.44 3.19 8.01 11.34 8.09 8.48 9.62 5.32 2.49 3.27 3.42 2.39 2.15 5.60 2.61 6.94 8.50 6.54 7.25 Ae 2.30 0.86 3.78 4.07 6.45 3.91 3.82 5.35 2.54 4.01 2.22 3.79 3.42 2.66 3.37 4.61 1.65 0.81 6.97 6.42 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
Texture/association: (1) eclogite; (2) grt lherzolite; (3) lherzolite; (4) pyroxenite; (5) phl-grt wehrlite; (6) phl-rich suite; (7) megacryst; (8) macrocryst; (9) microcryst ; (10) as inclusion in chr; (11) as inclusion in grt. Table 4a. Summary of LA-ICP-MS data from garnet KimberliteCACACACACACACACACACACACACACACACACACACACACACACA Borehole 335 335 335 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 33/35 335 335 335 33/35 33/35 33/35 33/35 33/35 335 536 335 Texture/association11122233344411122333444 Mineral grt grt grt grt grt grt grt grt grt grt grt grt cpx cpx cpx cpx cpx cpx cpx cpx cpx cpx cpx Spot name au18a08 au18a12 au18a13 se15a09 se15a10 se15a13 au19c05 au19c06 au19c08 au19a13 se15a07 se15c03 au18a05 au18a09 au18a11 au19c12 au19c13 au19c09 au19c10 au19c11 fe25a15 fe26a09 fe26c13 All values are reported in ppm Sc 62 61 64 100 101 96 190 190 175 169 191 47 21 9 21 30 30 52 53 54 ——— Ni 28 30 32 93 97 89 49 119 47 18 22 5 261 140 330 432 444 308 316 306 330 241 379 Ga 10 11 12 10 10 10 4 7 4 5 4 10 16 7 16 7 7 4 4 4 8 8 9 Y 17181721221941383516172441244677455 Zr 15.6 17.6 17.6 51.8 50.8 48.1 87.5 82.3 81.3 69.9 77.5 12.9 92.1 21.0 51.0 20.1 20.9 83.7 84.8 85.9 55.2 19.6 57.5 La 0.01 <0.015 0.03 1.32 1.10 0.06 1.90 0.84 0.18 4.22 0.60 0.00 21.45 5.69 14.61 3.52 4.82 3.07 3.13 4.25 3.15 2.08 19.02 Ce 0.17 0.22 0.20 2.65 2.30 0.32 3.59 1.76 0.55 7.93 1.62 0.06 63.61 14.34 39.56 9.26 11.34 11.16 11.74 13.85 11.98 7.66 45.88 Pr 0.07 0.08 0.05 0.31 0.31 0.10 0.46 0.21 0.11 1.13 0.33 0.03 9.86 2.09 5.68 1.54 1.76 2.19 2.13 2.44 2.12 1.29 6.04 Nd 0.81 0.92 0.78 1.79 2.00 1.03 2.64 1.36 1.38 6.73 3.30 0.54 41.00 8.68 22.92 8.67 9.12 13.20 11.98 14.49 10.74 7.05 28.04 Sm 0.68 0.56 0.40 1.03 1.08 0.94 1.50 1.36 1.54 3.61 3.08 0.91 6.19 1.22 3.06 2.24 2.33 3.33 3.51 3.45 2.69 1.89 4.78 Eu 0.31 0.35 0.29 0.52 0.53 0.58 0.86 0.70 0.76 1.47 1.61 0.51 1.53 0.28 0.69 0.67 0.71 1.17 1.03 1.16 0.96 0.64 1.38 Gd 1.28 1.31 1.08 2.00 1.88 1.88 3.57 3.34 2.97 4.21 5.23 2.16 3.45 0.63 1.69 1.80 1.92 2.99 2.82 3.13 1.87 1.69 2.97 Tb 0.27 0.26 0.29 0.48 0.47 0.47 0.85 0.73 0.71 0.56 0.86 0.50 0.30 0.06 0.15 0.24 0.25 0.40 0.42 0.42 0.25 0.24 0.31 Dy 2.6 2.5 2.5 3.7 3.7 3.3 6.7 6.1 5.6 3.5 4.5 3.9 1.3 0.3 0.7 1.3 1.3 2.0 2.0 2.1 1.2 1.2 1.5 Ho 0.65 0.67 0.65 0.83 0.85 0.76 1.56 1.40 1.33 0.62 0.68 0.93 0.16 0.05 0.10 0.18 0.18 0.29 0.32 0.31 0.18 0.22 0.20 Er 2.2 2.2 2.1 2.5 2.5 2.4 5.2 4.6 4.5 1.6 1.5 3.0 0.3 0.1 0.2 0.4 0.4 0.6 0.6 0.6 0.3 0.4 0.3 Tm 0.3 0.3 0.3 0.4 0.3 0.3 0.8 0.7 0.6 0.2 0.2 0.5 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.1 0.0 0.1 0.0 Yb 2.4 2.3 2.7 2.6 2.7 2.3 6.3 5.9 5.5 1.5 1.3 3.6 0.1 0.1 0.1 0.2 0.2 0.3 0.4 0.4 0.2 0.3 0.3 Lu 0.4 0.4 0.4 0.3 0.4 0.4 1.1 0.9 0.9 0.2 0.2 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Texture/association : (1) eclogite; (2) grt lherzolite; (3) phl-grt wehrlite; (4) macrocryst. (<) below the detection limit; (—) not detected Table 4b. Summary of LA-ICP-MS data from clinopyroxene Kimberlite CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 CU79 TZ TZ TZ TZ TZ TZ Borehole MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 MFD01 G18 34 34 G10 G10 G10 Texture/association222333555222444555555 Mineral grt grt grt grt grt grt grt grt grt cpx cpx cpx cpx cpx cpx grt grt grt cpx cpx cpx Spot name au18b07 au18b08 au18b13 se15c03 se15c05 se15c06 au19b11 au19b13 au19b14 au18b09 au18b14 au18b15 au19b05 au19b06 au19b16 fe27b07 fe27c09 fe27c14 fe27a08 fe27a09 fe27a16 All values are reported in ppm Sc 80 74 77 47 50 102 79 77 78 33 32 34 46 58 54 —————— Ni 21 21 21 5 88 34 113 109 112 265 271 260 123 179 210 123 108 127 241 243 251 Ga 8 8 8 10 10 7 12 12 13 7 7 8 9 11 12 15 19 14 6 6 5 Y 24 23 22 24 25 30 22 21 22 4 4 4 7 7 6 18 18 19 4 4 2 Zr 15.4 19.6 17.8 12.9 13.2 66.5 40.0 39.6 41.2 34.6 34.4 39.2 64.1 93.5 88.5 54.6 43.1 54.1 65.7 69.4 30.4 La <0.0069 <0.0042 0.00 0.00 0.12 0.02 0.03 0.03 0.03 6.89 6.87 6.73 5.41 5.69 5.19 0.05 1.16 0.11 21.82 20.85 11.37 Ce 0.05 0.03 0.02 0.06 0.28 0.29 0.23 0.26 0.23 14.11 14.38 13.99 15.62 18.15 20.03 0.39 2.94 0.47 66.16 61.78 25.99 Pr 0.02 0.01 0.01 0.03 0.07 0.12 0.09 0.09 0.10 2.21 2.21 2.19 2.49 3.07 3.13 0.12 0.33 0.12 8.65 8.27 3.14 Nd 0.34 0.16 0.21 0.54 0.77 1.64 0.80 0.92 0.71 10.82 10.74 10.46 13.15 16.10 15.68 1.38 1.77 1.15 36.97 36.12 11.87 Sm 0.49 0.42 0.43 0.91 0.96 1.38 0.68 0.79 0.72 2.69 2.49 2.73 3.28 4.09 3.72 0.92 0.75 0.92 6.04 6.37 2.13 Eu 0.29 0.30 0.28 0.51 0.54 0.83 0.45 0.45 0.43 0.79 0.73 0.78 1.04 1.26 1.28 0.52 0.45 0.51 1.73 1.71 0.61 Gd 1.66 1.49 1.55 2.16 2.24 3.12 1.72 1.70 1.72 2.16 2.23 2.39 2.94 3.22 2.79 1.89 1.41 1.91 3.38 3.45 1.82 Tb 0.39 0.40 0.35 0.50 0.54 0.77 0.43 0.41 0.41 0.26 0.24 0.26 0.38 0.42 0.37 0.44 0.37 0.43 0.37 0.35 0.21 Dy 3.7 3.7 3.2 3.9 4.0 5.7 3.6 3.6 3.5 1.1 1.2 1.3 1.8 2.1 1.8 3.0 2.8 3.2 1.3 1.5 0.8 Ho 0.90 0.78 0.78 0.93 1.01 1.27 0.84 0.81 0.90 0.15 0.18 0.17 0.28 0.30 0.30 0.74 0.71 0.77 0.19 0.18 0.13 Er 3.0 2.9 2.5 3.0 3.1 3.5 2.7 2.8 3.0 0.3 0.3 0.4 0.6 0.7 0.6 2.1 2.3 2.0 0.3 0.3 0.2 Tm 0.4 0.4 0.4 0.5 0.5 0.4 0.4 0.4 0.4 0.0 0.0 0.0 0.1 0.1 0.1 0.3 0.3 0.3 0.0 0.0 0.0 Yb 3.4 3.4 2.8 3.6 3.7 3.1 3.0 3.1 3.1 0.1 0.1 0.2 0.3 0.3 0.4 2.5 3.0 2.4 0.2 0.1 0.1 Lu 0.5 0.5 0.4 0.5 0.5 0.4 0.4 0.5 0.5 0.0 0.0 0.0 0.0 0.1 0.0 0.3 0.4 0.4 0.0 0.0 0.0 Texture/association : (1) eclogite; (2) grt lherzolite; (3) phl-grt wehrlite; (4) lherzolite; (5) macrocryst. (<) below the detection limit; (—) not detected Table 5. Nd isotopic analyses (TIMS) 147 144 143 144 143 144 143 144 Age Sm Nd Sm/ Nd Nd/ Nd Nd/ Nd Nd/ Nd Nd Nd TDM TCHUR No. Sm/Nd e t fSm/Nd (Ma) (ppm) (ppm) actual actual CHUR initial actual initial Ga Ga 1 1250 2.06 8.18 0.251834 0.151600 0.512532 0.008209 -0.23 0.511023 0.511288 -2.1 5.2 1.25 0.36 2 1250 2.66 15.50 0.171613 0.103300 0.512529 0.008209 -0.47 0.511023 0.511681 -2.1 12.9 0.72 0.18 4 1250 19.90 227.00 0.087665 0.052800 0.511756 0.008209 -0.73 0.511023 0.511323 -17.2 5.9 1.22 0.93 5 650 1.04 6.80 0.152047 0.091500 0.512781 0.004260 -0.53 0.511800 0.512391 2.8 11.5 0.35 -0.21 7 650 1.19 7.80 0.152174 0.091600 0.512744 0.004260 -0.53 0.511800 0.512354 2.1 10.8 0.39 -0.15 8 650 1.86 9.60 0.193750 0.116600 0.512771 0.004260 -0.41 0.511800 0.512274 2.6 9.3 0.45 -0.25 9 650 0.85 5.30 0.160377 0.096500 0.512725 0.004260 -0.51 0.511800 0.512314 1.7 10 0.43 -0.13 10 650 6.10 42.30 0.144208 0.086800 0.512747 0.004260 -0.56 0.511800 0.512377 2.1 11.3 0.38 -0.15