Кимберлиты связанные со структурой Lucapa, Ангола

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 delseu 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 isnot authorized (framing). This rights affect to the presentation summary of the thesis as well as to its contents. In the usingor citation of parts of the thesis it’s obliged to indicate the name of the author.

1


by


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

2

Supervisors: Dr. Joan Carles Melgarejo Draper


Dr. Salvador Galí Medina


Dr. Mónica Escayola

CONICET-IDEAN

Committee: Dr. José Mangas Viñuela (President)

Universidad de Las Palmas de Gran Canaria

Dr. Joaquín A. Proenza F. (Secretary)


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.

3

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.

4

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.

5

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

6

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.

7

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

8

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.

9

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.

10

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).

11

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

12

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

13

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”.

14

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

15

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)

16

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.

17

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).

18

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),

19

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

20

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

21

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

22

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).

Province Contract Kimberlite BoreholePresence of diamonds

(YES/NO)

LUNDA NORTE LUEMBA Tchiuzo 34 YESLUNDA NORTE LUEMBA Tchiuzo 44 YESLUNDA NORTE LUEMBA Tchiuzo G10 YESLUNDA NORTE LUEMBA Tchiuzo G18 YESLUNDA SUL CATOCA Catoca 0335 YESLUNDA SUL CATOCA Catoca 0536 YESLUNDA SUL CATOCA Catoca 033/35 YESLUNDA SUL CATOCA Catoca 044/35 YESLUNDA SUL CATOCA Catoca 77/35 UnknownLUNDA SUL CATOCA Catoca CA135 YESLUNDA SUL CATOCA Catoca CA336 YESLUNDA SUL CATOCA Catoca CA515 YESLUNDA SUL CATOCA Catoca CA535 YESLUNDA SUL CATOCA Catoca CA538 YESLUNDA SUL CATOCA Anomaly CAT-116 116 Some prospectivityLUNDA SUL CATOCA Camitongo 28 Some prospectivityLUNDA SUL LAPI Kambundu 216 NOLUNDA SUL ALTO CUILO Alto Cuilo 1 1 NOLUNDA SUL ALTO CUILO Alto Cuilo 16 11 Some prospectivityLUNDA SUL ALTO CUILO Alto Cuilo 254 5 YESLUNDA SUL ALTO CUILO Alto Cuilo 4 4 Some prospectivityLUNDA SUL ALTO CUILO Alto Cuilo 5 5 NOLUNDA SUL ALTO CUILO Alto Cuilo 63 6 YESLUNDA NORTE MUANGA Cucumbi 45 5 NO (not tested)LUNDA NORTE MUANGA Cucumbi 72 MFD07 YESLUNDA NORTE MUANGA Cucumbi 76 MFD03 NOLUNDA NORTE MUANGA Cucumbi 79 MFD01 YESLUNDA NORTE MUANGA Cucumbi 8 MFD06 NOLUNDA 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

23

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).

24

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.

25

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.

26

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

ilmenite, presumably originating by mantle metasomatism at a relatively high fO2. This Fe3+-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

27

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.

28

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.

29

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.

30

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.

31

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).

32

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

variable Fe2+:Fe3+ 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,

33

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-

34

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

35

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.

36

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.

37

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

38

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.

39

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

40

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.

41

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.

42

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.

43

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,

44

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.

45

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.

46

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.

47

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.

48

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.

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.

205 macla nº 9. septiembre ‘08 revista de la sociedad española de mineralogía

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)

IINTRODUCTION.

This study presents results of the initialphase of the research project,“Kimberlites associated to the Lucapastructure, Angola (Africa)”, within theframework of a multilateral agreementbetween the Faculty of Geology-Universitat de Barcelona, the EmpresaNacional de Diamantes de Angola andthe Agostinho Neto University (Luanda-Angola). The research is based on two sets ofcore sampling down to 600 m deep. Thefirst set comes from Catoca pipe andallowed us to identify complete craterand diatreme facies. The second one (18kimberlites) comes from Cucumbi,Cacuilo, Tchiuzo, Alto Cuilo, Camitongoand Kambundu, whose samples weregathered during fall 2008. Currently, weare working on these sets of samples. The kimberlites are ultrabasic rocks withhigh content of volatiles mainly CO2, anda typical inequigranular texturecharacterized by the presence of macro-megacrysts which can be xenoliths orxenocrysts embedded in a fine-grainedmatrix (Mitchell, 1995; Benvie, 2007).These special rocks have a greatimportance, not only in scientific termssince they add valuable informationabout lithospheric mantle but alsobecause they can contain diamond. RREGIONAL SETTING. The area of interest is localized innortheastern Angola (Africa), beingtectonically controlled by the Lucapastructure, a former rift (Guiraud et al.,2005) of early Cretaceous that extendsNE-SW across Angola. Associated to thisstructure there is a magmatic belt,which is composed by kimberlitestoward NE and carbonatites toward SW.At present, over 2000 kimberlites havebeen identified in this structure and their

diamond potential is currently beingstudied. The Catoca kimberlite is themost important primary diamonddeposit in Angola, hosted byPrecambrian rocks and covered byMesozoic-Cenozoic sedimentary deposits(Janse et al, 1995). PPETROGRAPHY. There are some minerals which arefrequently associated to diamond insidekimberlites and they are used asindicator minerals for the diamondexploration. The main indicator mineralsare: magnesian ilmenite (Pell, 1998),garnet and chromite (Wyatt et al, 2004).However, for this instance we will focuson ilmenite since it is the first mineralanalyzed in 2007. Diverse xenoliths, comprising lherzolite,eclogite, harzburgite, carbonatite, gneissand amphibolite are distributed throughthe Catoca and Cucumbi kimberlites.Some shales and sandstones can bepresent in the upper part of thisKimberlite. Accessory minerals andxenocrysts comprise garnet, zircon, Cr-rich diopside, amphibole, phlogopite,chromite and several generations ofilmenite. Secondary minerals includeserpentine-group minerals being themost abundant, calcite, barite,barytocalcite, witherite and strontianite. Based on optical petrographic studiesand BSE images from SEM-ESEM withEDS microanalysis, we have been ableto discriminate up to six textural types ofilmenite in Catoca and Cucumbikimberlite: a) intercumular ilmenite inperidotitic xenoliths (Fig 1); b) anhedralilmenite in carbonatite xenoliths; c)ilmenite unaltered megacrysts; d)nodular xenocryst of ilmenite withdifferent grades of replacement, someof them with symplectitic textures (Fig.2); e) skeletal ilmenite; and f) euhedralcrystals of ilmenite in matrix (Fig. 3).

fig 1. Intercumular ilmenite in peridotitic xenolith.

fig 2. Nodular xenocrysts of ilmenite withreplacement.

fig 3. Euhedral crystals of ilmenite in matrix.

Zircon xenocrysts are partially replacedby fine-grained baddeleyite, and at leasttwo populations exist according to thetrace element distribution. All of thesecrystals are enriched in HREE, but with anoticeable positive Ce anomaly, similarto that reported in zircon in a MARIDxenolith from a southern Africankimberlite (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, butthere is a slight depletion in REE fromthe core to the rim. MMINERAL CHEMISTRY OF ILMENITE. Every texture has been systematicallyanalyzed with EPMA which allowed us toidentify three compositional types ofilmenite (I, II and III). This combinedtechnique –texture and compositionanalysis- has been suitable for analyzingzircon and garnet as well. The primary ilmenite (type I) inmegacrysts (xenocrysts) is generally richin Cr and Fe3+. Its composition is similarto the intercumulus crystals in peridotitexenoliths. This ilmenite is replaced, inthe first instance, by magnesian ilmenite(type II). This process takes place alongmicrodiscontinuities (cleavage, bordergrains, contour subgranes, kink bandplanes, etc.), producing diffusivereplacements. In a more advancedstage, symplectitic replacements occur,involving an early generation ofmagnesian ilmenite (type II) at theexpense of Fe3+-rich primary ilmenite(from texture a to d ). A late generationof Mn-Nb-Zr-rich ilmenite (type III) cutsthe previous ones. Contrastingly, the late euhedral Mn-Nb-Zr ilmenite crystals found in thekimberlite matrix do not present anyevidence of replacement. This ilmeniteis poor in Mg and Fe3+. Theircompositions are identical with the Mn-rich ilmenite produced during latereplacement stages of ilmenitemegacrysts. Compositions of Mn-richilmenite are similar to those found incarbonatite xenoliths. DDISCUSSION AND CONCLUSIONS. Textural evidences indicate a differentcomplex history of growth in thexenocrysts. Unaltered megacrystilmenite (ilmenite type I) rich in Fe3+ (fig.4), indicates crystallization under highfO2 conditions; this ilmenite contains Nb,Cr, Ni and Ta in low contents. Itscomposition is similar to those ilmeniteIntercumular megacrysts that occur inperidotite xenoliths. Hence, most of theilmenite xenocrysts seem to have beenproduced by disaggregation of mantlexenoliths. Ilmenite I is replaced alongdiscontinuities by magnesian ilmenite(ilmenite type II); the elementaldistribution of Mg in these grains pointsto processes of replacement throughsolid-state diffusion in a typical reducing

environment. Magnesian ilmenite is alsoenriched in Cr and Ni. More advancedreplacement produces a symplectiticreplacement of ilmenite II by ilmenite III.

ffig 4. MgTiO3-FeTiO3-Fe2O3 ratio for the differenttypes of ilmenite (I, II and III) discriminated for eachtexture.

A late generation of ilmenite III (Mn-richilmenite) is found rimming all the abovementioned generations, and is stronglyenriched in Nb, Ta, Zr, W, Hf, Th and U,and poor in Mg and Fe3+. Thecomposition of this ilmenite is similar tothat of the fine-grained euhedralilmenite crystals found in the kimberliticmatrix and also to that of the ilmenitecrystals found in the carbonatiticxenoliths. Crystals of Mn-rich ilmenite(ilmenite type III) are not replaced orzoned, and seem to have crystallized inequilibrium with the kimberlitic magma.Both the late generations of ilmeniteand the baddeleyite replacing zircon canbe produced by interaction of acarbonate-bearing kimberlitic magmaenriched in Mn and HFSE. Thereplacement of Fe3+-rich ilmenite by Mg-and Mn-rich ilmenite implies that theearly ilmenite was formed underoxidizing conditions in the mantle, andthe lastest compositions of ilmenitewere produced by reaction with thekimberlitic magma. Megacrysts of ilmenite are frequentlypresent in diamondiferous kimberlites,contrasting with ilmenite observed inbarren kimberlites. This might become anew guide in diamond exploration. In conclusion, the composition of thisilmenite is the result of a set ofreplacement processes with rich fluids inMg and Mn affecting an oxidized primary

ilmenite in a higher or lower grade.These fluids are reducing, especiallythose rich in Mn. Picroilmenite hastraditionally been interpreted as anindicator of kimberlite associations, aswell as an indicator of low fO2, which isnecessary for the preservation ofdiamond. Although Catoca and Cucumbiare diamondiferous kimberlites, theyshow that Mg ilmenite is clearly a latereplacement product, and the grade ofreplacement of the primary grains isvery variable. Therefore the absence ofmagnesian ilmenite in a kimberlite doesnot appear to be a convincing argumentto exclude the presence of diamonds.Accordingly, this work proposes a newinsight into the concept of ilmenite indiamond exploration. ACKNOWLEDGMENTS. This research is supported by the projectCGL2006-12973 of Ministerio deEducación y Ciencia (Spain), the AGAURSGR 589 of Generalitat de Catalunyaand a FI grant sponsored by theDepartament d’Educació i Universitatsde la Generalitat de Catalunya i del FonsSocial Europeu. The authorsacknowledge the ServeisCientificotècnics de la Universitat deBarcelona. REFERENCES. Benvie, B.(2007): Mineralogical imaging of

kimberlites using SEM-based techniques.Minerals Engineering, 20, 435-443.

Dawson, J.B., Hill, P.G., Kinny, P.D. (2001):Mineral chemistry of a zircon-bearing,composite, veined and metasomatisedupper-mantle peridotite xenolith fromkimberlite. Contrib. Mineral Petrol., 140,720-733.

Guiraud, R., Bosworth, W., Thierry, J.,Delplanque, A. (2005): Phanerozoicgeological evolution of Northern andCentral Africa: An overview. Journal ofAfrican Earth Sciences, 43, 83-143.

Janse, A.J.A. & Sheahan, P.A. (1995):Catalogue of world wide diamond andkimberlite occurrences: a selective andannotative approach. Journal ofGeochemical Exploration, 53, 73-111.

Mitchell, R.H. (1995): Kimberlites, Orangeites,and related rocks. New York, PlenumPress, 410 pp.

Pell, J. (1998): Kimberlite-hosted Diamonds,in Geological Fieldwork 1997. BritishColumbia Ministry of Employment andInvestment 1998-1, 24L1-24L4.

Wyatt, B.A., Mike, B., Anckar, E., Grutter, H.(2004): Compositional classification of“kimberlitic” and “non-kimberlitic”

ilmenite. Lithos, 77, 819-840.


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.

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, Spainb ENDIAMA, Major Kanhangulo, 100, Luanda, Angolac Departamento de Geologia, Faculdade de Ciências, Universidade Agostinho Neto, Av. 4 de Fevereiro 7, 815, Luanda, Angola

a b s t r a c ta r t i c l e i n f o

Article history:

Received 2 October 2008

Accepted 16 May 2009

Available online 25 June 2009

Keywords:

Ilmenite

Kimberlite

Diamond

Fluid

Texture

Composition

The Catoca group-I kimberlite, the only currently active diamond-producing mine in Angola, was emplaced in

the northeastern part of the Lucapa structure. We focus here 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 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

size, ranging from micro- to megacrysts. Most ilmenite is found to derive, through a complex process, from

replacement of Fe3+-rich ilmenite, presumably originating by mantle metasomatism at a relatively high fO2.

This Fe3+-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. 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

Angola has become an important producer of diamond (Janse and

Sheahan, 1995; Read and Janse, this issue), with a significant part of the

production being obtained from the Catoca kimberlite, in Lunda Sul

province, in northeastern Angola. Catoca, currently the unique active

kimberlite mine in Angola, is located in the Lucapa structure, a system of

Cretaceous extensional faults trending NE–SW (Reis, 1972; De Carvalho

et al., 2000; Guiraud et al., 2005).

For many years, the composition of ilmenite has been stressed as an

exploration guide for diamondiferous kimberlites and placers (e.g.,

Mitchell, 1989, 1995, 1997; Wyatt et al., 2004). It has been correlated as

well with conditions in the mantle where the kimberlitic magmas

originated (Haggerty and Tompkins,1983; Arculus et al.,1984; Haggerty,

1991a,b; Gurney et al., 1993). Griffin and Ryan (1995) have suggested

that with some patterns of compositional evolution in ilmenite

megacrysts, once can assess the occurrence of mechanisms of fractional

crystallization of single batches of magma associated with extensive

metasomatic alteration of the wallrocks, and hence destruction of

diamond. Thus, reduced kimberlites bearing Mg-rich ilmenite would be

expected to preserve diamond, whereas the presence of Fe3+-rich

ilmenite could indicate oxidizing processes that could destroy crystals of

diamond (Gurney et al., 1993; Gurney and Zweistra, 1995; Kostrovitsky

et al., 2004, 2006; van Straaten et al., 2008). Other compositional

features of ilmenite are of more controversial origin. In particular, Mn-

rich compositions, found in previous surveys of some Angolan

kimberlites (Llusià Queral et al., 2005a,b; Rogers and Grütter, this

issue) have been attributed to supergene processes; others propose

magmatic crystallization under reducing conditions (Hwang et al.,1994)

or metasomatic processes in the mantle (Meyer and McCallum,1986). In

this investigation, we seek explanations of the real significance of

compositional and textural variations of ilmenite in kimberlites. We

used back-scattered electron (BSE) petrography with microanalysis

using energy-dispersion spectroscopy (EDS), quantitative powder X-ray

diffraction (PXRD), and quantitative chemical analyses using an

electron-microprobe (EMP) on a suite of 81 representative thin sections

and 19 probes from two core samples of Catoca pipe. With these new

datasets, we are able to shed new light on the origin of these unusual

compositions.

2. The Catoca kimberlite

The Catoca pipe outcrop, 639000 m2, is found 30 km NNW of

Saurimo, the capital of Lunda Sul. Catoca can be classified as a group-I

Lithos 112S (2009) 966–975

⁎ Corresponding author. Tel.: +34 9340 21344; fax: +34 9340 21340.

E-mail address: [email protected] (S.E. Robles-Cruz).

0024-4937/$ – see front matter © 2009 Elsevier B.V. All rights reserved.

doi:10.1016/j.lithos.2009.05.040

Contents lists available at ScienceDirect

Lithos

j ourna l homepage: www.e lsev ie r.com/ locate / l i thos

kimberlite (Mitchell, 1995). Complete crater and diatreme facies are

recognized (Ganga et al., 2003; Fig. 1), according to the classification

criteria of Clement and Skinner (1985), as modified by Scott Smith et al.

(2008), thus indicating the minimal extent of erosion of the kimberlite.

Crater facies are found up to 230–270 m in depth, and are composed

in the uppermost part by epiclastic sandstones with cross-stratification

in the central part, and coarse debris rimming the crater. Altered crystals

of garnet and diopside may occur as accessory minerals in these

sediments. Ilmenite is rare in this unit. Mostof the cement is ferruginous,

but in some areas, the sandstone has a calcareous-ferruginous cement.

In the lower part of the sequence, the content of volcaniclastic material

increases, and the lower half of the crater facies become dominated by

resedimented volcaniclastic kimberlite facies (unit RVK).

Below the crater facies, Ganga et al. (2003) described classic

tuffisitic kimberlite facies (TK; Mitchell et al., this issue); in the

current terminology, this category could be named massive volcani-

clastic kimberlite (Sparks et al., 2006) or Wesselton-type volcaniclas-

tic kimberlite (Scott Smith et al., 2008). Ganga et al. (2003) also

divided this unit into different subunits based on the size of the

fragments, and pointed out the occurrence of abundant xenoliths

derived from the host rocks, and the scarcity of xenoliths from the

mantle. The extensive drilling allowed us to sample these units down

to 609 m in depth.

The diatreme facies are strongly altered all along the profile; olivine

is completely replaced by serpentine, calcite and saponite (Kotel'nikov

et al., 2005). Xenoliths of the host rocks are common, and comprise

gneiss, amphibolite, granite, sandstone and shale. Mantle xenoliths are

also quite altered and include mainly metasomatic peridotite, and rarely

eclogite as well as xenoliths of carbonatite. Xenocrysts encountered in

the diatreme facies comprise G9 and G10 varieties of garnet according to

the classification of Grütter et al. (2004), zircon (partly replaced by

baddeleyite), chromian diopside, amphibole, phlogopite and ilmenite.

The matrix of the kimberlite contains lizardite, apatite, calcite, ilmenite

and chromite; titanite, zirconolite, baddeleyite, barite, dolomite, with-

erite, barytocalcite, strontianite, sulfides, identified by PXRD and EMP,

and minor minerals are widespread in the matrix and fill small veinlets.

3. Petrography of ilmenite

Optical petrographic studies and back-scattered electron (BSE)

images taken with a SEM-ESEM with EDS microanalysis, coupled with

microprobe analysis, allowed us to discriminate four main variants

of ilmenite based on compositional attributes: a) Fe3+-rich ilmenite,

b) Mg-rich ilmenite, c) Mn-rich ilmenite, and d) near-ideal ilmenite

(Fe2+Ti4O3). These types can be easily distinguished using BSE images,

as the Mg-rich ilmenite displays the darkest shades, and the Mn-rich

ilmenite is the lightest. On the other hand, up to seven textural classes of

ilmenite have been established, based on their paragenetic position and

degree of replacement: 1) intercumulus Fe3+-rich ilmenite grains in

metasomatized peridotite xenoliths, 2) anhedral ilmenite in carbonatite

xenoliths; 3) homogeneous Fe3+-rich ilmenite present as macro- and

megacrysts; 4) partially replaced ilmenite macro- and megacrysts; 5)

symplectitic ilmenite xenocrysts; 6) skeletal ilmenite crystals in a

pelletal matrix; 7) tabular Mn-rich ilmenite crystals in a kimberlite

matrix. The suite of ilmenite that we examined contains grains of

variable size: microcrystals have 5–20 μm in diameter, being some as

large as 50 μm; most macrocrysts have dimensions between 1 and

10 mm; megacrysts are very rare and may exceed 2 cm.

The distribution of these ilmenite textural types is not homogeneous

in the crater and diatreme kimberlite facies. Homogeneous ilmenite

macro- and megacrysts are found in all the kimberlite facies, but the

remainder of the textural variants are restricted to the diatreme facies,

mainly in the volcaniclastic kimberlite (below 250 m in depth). Partly

replaced megacrysts occur in the uppermost part of the diatreme facies,

and strongly corroded macro- and megacrysts (in particular, those with

a symplectitic texture) are only found below 350 m in depth.

3.1. Intercumulus grains of Fe3+-rich ilmenite in metasomatized

peridotitic xenoliths

The intercumulus grains of Fe3+-rich ilmenite are anhedral, 300–

800 μm in diameter, and interstitially distributed among roundish

grains of olivine (Fig. 2A). The olivine is completely replaced by

serpentine, but the mesh texture typical of serpentinized olivine is

clearly recognizable. The ilmenite is polycrystalline, and grains show

polygonally annealed textures with curved borders and triple points.

Similar polygonal textures in ilmenite in kimberlitic suites have been

interpreted as formed by intense annealing of stressed ilmenite

(Mitchell, 1973; Haggerty et al., 1977; Tompkins and Haggerty, 1985).

The intercumulus ilmenite is quite homogeneous in composition and

rich in Fe3+ (as inferred from stoichiometry: see below), although it

may be partly replaced by Mg-rich ilmenite.

Fig. 1. Cross section of the Catoca kimberlite (adapted from Kriuchkov et al., 2000).

967S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975

3.2. Anhedral grains of ilmenite in carbonatite xenoliths

Carbonatite xenoliths are rare (less than 7% of grains) in Catoca, and

accessory ilmenite in them occurs as small crystals (50–90 μm) found

as inclusions in phlogopite crystals, or intergrown with calcite, apatite,

zircon and phlogopite (Fig. 2B). The ilmenite is usually rimmed at grain

margins by Mn-rich ilmenite.

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).

Fig. 3. Partly replaced polycrystalline nodules of ilmenite. SEM image, mode BSE.

(A) Polycrystalline nodule of ilmenite (lightest) showing corrosion on the uppermost

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

subgrain borders and small cracks by Mg-rich ilmenite (darker) and Mn-rich ilmenite

(pale gray to white). (C) Detail of a polycrystalline nodule of Fe3+-rich ilmenite. Ilmenite

(Ilm), serpentine (Srp).

968 S.E. Robles-Cruz et al. / Lithos 112S (2009) 966–975

shape of monocrystalline grains is consistent with an origin by disag-

gregation of polycrystalline grains (Fig. 2C). Homogeneous macrocrysts

of ilmenite or Mg-rich ilmenite are widely represented in kimberlites

(i.e., Mitchell, 1973), but in the Catoca kimberlite, this textural type is

very rare (less than 5% of grains).

3.4. Partly replaced macro- and megacrysts of Fe3+-rich ilmenite

These coarse crystals (0.4–2 cm) are also rounded and they occur

usually in the core of pelletal lapilli; they present different grades of

replacement (Fig. 3A, B, C). These megacrysts consist mainly of Fe3+-

rich ilmenite similar in shape and composition to the last category.

However, they have been corroded and replaced by Mg-rich ilmenite

along old surfaces and discontinuities. Some crystals may display

freshly fractured borders, thus indicating that corrosion and replace-

ment took place before the explosive processes in the kimberlite.

3.5. Macrocrysts of symplectitic ilmenite

Macrocrysts of Fe3+-rich ilmenite from the deepest parts of diatreme

facies are similar in shape and composition to two previous types, but

they show an intense replacement, leading to a symplectitic texture; this

type of replacement is developed only at the border of the crystal

(Fig. 4A) or it can affect the whole crystal (Fig. 4B). The other mineral

originally present in these intergrowths has been fully replaced by

serpentine, but similar unreplaced textures in many kimberlites consist

of pyroxene and ilmenite (i.e., Haggerty et al., 1977). These grains have

complex patterns of replacement, with ilmenite strongly replaced by

Mg-rich ilmenite; the zoned grains are finally overgrown by Mn-rich

ilmenite (Fig. 4C). Manganese-rich ilmenite forms only a thin rim or

veinlet, and it is accompanied by small grains of barytocalcite,

strontianite and baddeleyite. Most of the symplectitic replacement

takes place in crystal discontinuities and is adapted to pre-existing

features as deformation-induced kink bands or subgrains.

3.6. Skeletal crystals of ilmenite in a pelletal matrix

These small crystals (80–150 μm) may be present in pelletal lapilli

and exhibit very irregular shapes (Fig. 4D). The core of the crystals is

generally constituted by ilmenite, which can be partially replaced by

Mn-rich ilmenite. It is the least common variety of ilmenite in the Catoca

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).


3.7. Tabular crystals of Mn-rich ilmenite in the kimberlite matrix

Tabular crystals of Mn-rich ilmenite (1–10 μm in length), slightly Nb-

rich, are set in randomly oriented groups in the kimberlite groundmass

(Fig. 5), and may be associated with euhedral grains of apatite and

chromite. They are internally homogeneous and unaltered, reflecting

equilibrium with the kimberlite matrix, and are found only in the

diatreme facies. Although rare, members of the ilmenite group have

been found in the groundmass of other kimberlite pipes, either Mn-

enriched and slightly Nb-enriched (Chakhmouradian and Mitchell,

1999) or Mg-rich (Nielsen and Sand, 2008).

4. Composition of ilmenite

Representative grains of ilmenite from every class of texture were

selected using as reference the SEM-BSE images, and then analyzed

with a Cameca SX-50 microprobe, with four wavelength-dispersion

spectrometers. All ilmenite crystals were analyzed with an excitation

voltage of 20 keV, beam current of 20.1 nA and a take-off angle of 40°.

We used the following standards, crystals and lines; periclase (Mg,

TAP Kα), orthoclase (Al, TAP Kα), diopside (Si, TAP Kα), wollastonite

(Ca, PET Kα), rutile (Ti, PET Kα), synthetic Cr2O3 (Cr, PET Kα),

rhodonite (Mn, LIF Kα), hematite (Fe, LIF Kα), synthetic NiO (Ni, LIF

Kα), synthetic ZrO2 (Zr, PET Lα), and metallic Nb (Nb, PET Mα). The

ratio Fe2+/Fe3+ is calculated by stoichiometry.

At this phase of the research, we analyzed with EMP about 400

points from 40 grains of 20 representative samples of a suite of 100

core samples (81 thin sections and 19 probes) from two boreholes of

Catoca Kimberlite. Datasets totalling less than 95% after charges were

balanced were rejected. Analyses were made along profiles in order to

evaluate progressive changes among the types of ilmenite.

Correlations among the major and minor components in ilmenite of

the different types are shown in Fig. 6; selected results of electron-

microprobe analyses from Catoca are given in Table 1 under the fol-

lowing headings: (1) intercumulus in peridotite xenolith, (2) anhedral

in carbonatite xenolith, (3) homogeneous macrocryst, (4) partly re-

placed macro- and megacryst, (5) symplectitic macrocryst, (6) skeletal

crystal, and (7) matrix. As can be seen in Fig. 6A, the Mn-rich ilmenite

plots outside the classic kimberlite domain in the TiO2–MgO plot of

Wyatt et al. (2004); moreover, there is a continuous trend of enrichment

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

values in Fe3+ are found in Mn-rich ilmenite, which produces a tight

group (Fig. 6B). However, there is a continuous trend among composi-

tions of Fe3+-rich ilmenite and Mg-rich ilmenite. The level of Fe2+ is

higher in Mg-rich ilmenite than in Fe3+-rich ilmenite, but it also tends to

decrease in Mn-rich ilmenite owing to the substitution of Fe2+ for Mn2+

(Fig. 6C). The contents of Cr are quite variable, but they are higher in Mg-

rich ilmenite and lower in Mn-rich ilmenite; they show a rough positive

correlation with Mg. On the other hand, Mn and Fe3+ very clearly show

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

(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

progressive increase in Mg from Fe3+-rich ilmenite toward Mg-rich

ilmenite (Fig. 6E). Finally, the niobium content tends to increase where

the Mn content increases in Mn-rich ilmenite (Fig. 6F), as described in

other kimberlite fields (i.e., Chakhmouradian and Mitchell, 1999),

although the highest values are found in ilmenite from carbonatite

xenoliths. Groundmass ilmenite is compositionally similar to the outer

rim of ilmenite megacrysts.

5. Discussion

Ilmenite textures and composition are diverse in the Catoca

kimberlite, thus suggesting a complex history for the ilmenite

nodules. The diversity in textures and composition reflects primarily

to the paragenetic position of ilmenite in the kimberlite (accessory

in xenoliths, macro- and megacrysts, matrix) and replacement

processes.

Ilmenite has been described in many kimberlites worldwide as an

accessory mineral in many metasomatized peridotite xenoliths. Some

widespread examples comprise the MARID suite (Dawson and Smith,

1977), MORID veins (i.e., Jones et al., 1982) and some metasomatized

garnet-bearing peridotites (Stiefenhofer et al., 1997; Kopylova et al.,

1999). Some of the Catoca xenoliths can be included in the last

category, and others have similarities with the ilmenite-bearing

dunite xenoliths described by Kaminsky et al. (2002).

The macro- and megacrysts in kimberlites have been interpreted

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

some nodules broken during the explosive processes are not replaced

in the broken corners.Fig. 5. Euhedral platy crystals of Mn-rich ilmenite in matrix. Serpentine (Srp). The

crystals are in equilibrium with the kimberlite matrix in the diatreme facies.


Fig. 6. Correlation between major elements or between major and minor elements for the main textural types of ilmenite. Atoms per formula unit (apfu).


Table

1

Ch

em

ical

anal

yse

sof

ilm

en

ite

cryst

als

from

Cat

oca

.

Textu

re1

11

23

33

44

45

55

66

67

77

Ilm

typ

eI

III

IVI

II

III

III

IIII

II

IIII

III

III

III

I

Bore

hole

535

535

535

535

535

535

535

535

535

535

535

535

535

535

535

535

535

535

535

Dep

th(m

)350

350

350

451

409

409

409

350

350

451

350

451

350

350

350

350

350

350

350

Poin

t26B

sn68

26B

sn67

26B

sn73

36f5

7409a3

2b

40932b21

40932c2

126B

n74

26A

i63

36g59

26A

j17

36a1

326A

j24

26A

q32

26A

e39

26A

e34

26A

a12

26A

a15

26A

n86

(wt.%)

SiO

20.0

00.0

40.0

40.3

30.0

40.0

00.0

30.0

01.

26

0.0

30.0

50.0

50.1

10.0

50.0

50.1

70.4

31.

00

0.1

8

TiO

243.8

84

4.0

650.6

746.5

436.7

836.4

936.6

538.6

951.

64

54.3

047.

25

50.1

051

.30

42.1

346.1

251.

36

48.7

947.

86

49.7

9

Al 2

O3

0.1

90.2

10.2

30.0

20.1

20.0

40.1

40.1

20.1

80.1

60.1

40.0

70.0

50.1

50.1

20.0

20.0

50.1

60.0

6

Nb

2O

50.2

60.3

30.2

73.0

30.3

40.4

80.4

30.4

01.

20

0.2

10.5

10.2

50.1

10.3

00.2

40.0

72.5

41.

59

0.9

6

ZrO

20.1

20.1

40.1

70.0

6–

––

0.1

00.1

70.0

20.0

00.0

00.0

20.1

50.0

80.0

60.0

10.1

20.0

8

Cr 2

O3

0.9

00.8

91.

04

0.0

02.1

42.8

62.8

80.2

80.7

01.

07

0.8

92.0

90.0

80.6

23.8

80.2

90.1

50.3

60.1

7

Fe2O

319.9

918

.93

8.6

85.3

328.5

928.1

927.

56

28.4

39.8

56.2

012.4

49.9

11.

30

20.5

113.3

90.2

31.

26

2.8

11.

73

FeO

28.7

627.

56

30.3

041.

04

27.

35

27.

40

27.

72

28.2

616.6

523.7

630.2

027.

76

43.4

026.9

325.6

533.0

237.

37

35.8

837.

77

Mn

O0.2

00.2

40.2

70.6

80.1

50.1

80.2

20.1

82.0

60.4

41.

88

0.3

92.6

80.2

30.2

513.0

86.4

56.3

56.8

1

MgO

5.8

76.6

68.4

30.1

53.7

43.7

13.6

33.5

716.4

313

.79

5.7

89.4

70.0

66.0

78.7

40.0

70.1

30.7

50.1

5

NiO

0.0

80.0

70.0

80.0

30.0

50.0

40.0

30.0

50.0

40.0

80.0

80.1

00.0

30.0

70.0

90.0

10.0

10.0

30.0

0

CaO

0.0

30.0

20.0

40.2

10.0

10.0

00.0

00.0

10.0

00.0

20.0

60.0

00.0

30.0

00.0

20.0

30.2

10.5

00.0

9

Tota

l10

0.2

899.1

410

0.2

297.

43

99.3

299.3

999.3

010

0.0

810

0.1

710

0.0

899.2

910

0.1

899.1

797.

20

98.6

398.4

197.

40

97.

41

97.

79

Cat

ion

son

bas

isof

thre

eO

atom

s

(apfu)

Si

0.0

00.0

00.0

00.0

10.0

00.0

00.0

00.0

00.0

30.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

10.0

30.0

0

Ti

0.8

10.8

00.9

00.9

70.7

00.6

90.7

00.7

20.8

70.9

30.8

70.8

90.9

80.7

90.8

40.9

90.9

50.9

20.9

6

Al

0.0

10.0

10.0

10.0

10.0

10.0

00.0

10.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

0

Nb

0.0

00.0

00.0

00.0

10.0

10.0

10.0

10.0

10.0

10.0

00.0

10.0

00.0

00.0

00.0

00.0

00.0

30.0

20.0

1

Zr

0.0

00.0

00.0

00.0

0–

––

0.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

0

Cr

0.0

20.0

20.0

20.0

00.0

50.0

60.0

60.0

10.0

10.0

20.0

20.0

40.0

00.0

10.0

70.0

10.0

00.0

10.0

0

Fe3+

0.3

50.3

70.1

50.0

10.5

40.5

40.5

20.5

30.1

70.1

10.2

30.1

80.0

20.3

90.2

40.0

00.0

20.0

50.0

3

Fe2+

0.5

60.5

80.6

00.9

40.5

80.5

80.5

90.5

90.3

10.4

50.6

20.5

50.9

20.5

60.5

20.7

10.8

10.7

70.8

1

Mn

0.0

00.0

00.0

10.0

30.0

10.0

10.0

10.0

00.0

40.0

10.0

40.0

10.0

60.0

00.0

10.2

80.1

40.1

40.1

5

Mg

0.2

40.2

10.3

00.0

00.1

40.1

40.1

40.1

30.5

50.4

70.2

10.3

30.0

00.2

30.3

10.0

00.0

00.0

30.0

1

Ni

0.0

00.0

00.0

00.0

0–

––

0.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

0

Ca

0.0

00.0

00.0

00.0

10.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

10.0

10.0

0

FeTiO

357.

44

59.4

960.9

196.4

158.0

058.0

059.0

059.9

031.

47

45.6

962.9

456.1

292.9

356.8

554.1

771.

72

84.3

879.7

982.2

3

Fe2O

317

.95

18.9

77.

61

0.5

127.

00

27.

00

26.0

026.9

08.6

35.5

811

.68

9.1

81.

01

19.8

012.5

00.0

01.

04

2.5

91.

52

MnTiO

30.0

00.0

01.

02

3.0

81.

00

1.0

01.

00

0.0

04.0

61.

02

4.0

61.

02

6.0

60.0

01.

04

28.2

814.5

814

.51

15.2

3

MgTiO

324.6

221.

54

30.4

60.0

014.0

014

.00

14.0

013

.20

55.8

447.

72

21.

32

33.6

70.0

023.3

532.2

90.0

00.0

03.1

11.

02

Tota

l10

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

010

0.0

0

Textu

ralty

pes:

(1)

inte

rcu

mu

lus

inp

eri

doti

texen

olith

;(2

)an

hed

ralin

carb

on

atit

exen

olith

;(3

)h

om

ogen

eou

sm

acro

cryst

;(4

)p

artl

yre

pla

ced

mac

ro-

and

megac

ryst

;(5

)sy

mp

lect

itic

mac

rocr

yst

;(6

)sk

ele

talcr

yst

al;

(7)

mat

rix.T

he

Fe3+

is

calc

ula

ted

by

char

ge

bal

ance

and

stoic

hio

metr

y.


On the other hand, a late generation of Mn-rich ilmenite is found

rimming all the above-mentioned generations, and is extremely poor

in Mg and Fe3+. 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; Fig. 7A,B).

Accordingly, these compositions could follow the kimberlite reaction

trend of Haggerty et al. (1977), producing enrichment in Fe2+.

However, the most significant aspects in this process are the strong

enrichments in Mn and HFSE. Similar enrichments were 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 instead the deposition of this ilmenite under the influence of a

CO2-rich fluid phase: a) the intimate association of this Mn-rich

ilmenite in open cavities 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), whose 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.

Furthermore, 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 been already 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 mantles other

groundmass minerals such as perovskite and spinel (Tompkins and

Haggerty, 1985). Although Mn-rich ilmenite could be produced during

magmatic crystallization, we contend that it could be also 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 kimberlite are quite

different from those found in the carbonatitic xenoliths at Catoca. In

this case, 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

consequences in 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 represents in the Catoca kimberlite more than 70% of

the volume of the grains, and compositions fall into the domains of “no

preservation of diamond” according the diagram of Gurneyand Zweistra

(1995) Fig. 8. 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); Catoca could be expected to be

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.


barren. Although Catoca is a diamondiferous kimberlite, Mg-rich ilmen-

ite here is clearly a product of late replacement, and the extent of

replacement of the primary grains is very variable. In other words,

textural relations must be taken into account in the application of

discriminants based on composition.

6. Conclusions

The composition of the Catoca ilmenite is complex, as the result of

multiple processes. 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 association with carbonates of Ba and Sr, can be

interpreted in terms of an interaction of the ilmenite crystals with a CO2-

rich fluid.

Although Catoca is a diamondiferous kimberlite, most of its ilmenite

compositions are strongly oxidized and poor in Crand Mg. Therefore, 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. Accordingly, this work proposes new

insight into the concept of ilmenite in exploration for diamond;

compositional attributes must be evaluated in lightof textural attributes.

Acknowledgments

This research was supported by the projects CGL2005-07885/BTE

and CGL2006-12973 of Ministerio de Educación y Ciencia (Spain), the

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

geologists, who kindly allowed us to acquire samples for this study and

gave all facilities for the mine trip. The authors also acknowledge 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. Xavier Llovet). An early version of the

manuscript was improved with the comments of two anonymous

reviewers. Vicki Loschiavo and Prof. Robert F. Martin made further

improvements to the revised version.

References

Agee, J.J., Garrison Jr., J.R., Taylor, L.A., 1982. Petrogenesis of oxide minerals in kimberlite,Elliott County, Kentucky (USA). American Mineralogist 67 (1–2), 28–42.

Arculus, R.J., Dawson, J.B., Mitchell, R.H., Gust, D.A., Holmes, R.D., 1984. Oxidation statesof the upper mantle recorded by megacryst ilmenite in kimberlite and type A and Bspinel lherzolites. Contributions to Mineralogy and Petrology 85, 85–94.

Armstrong, K.A., Nowicki, T.E., Read, G.H., 2004. Kimberlite AT-56: a mantle samplefrom the north central Superior craton, Canada. Lithos 77, 695–704.

Chakhmouradian, A.R., Mitchell, R.H., 1999. Niobian ilmenite, hydroxylapatite andsulfatian monazite: alternative hosts for incompatible elements in calcite kimberlitefrom Internatsional'naya, Yakutia. The Canadian Mineralogist 37 (5), 1177–1189.

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.

Dawson, J.B., Smith, J.V., 1977. The MARID (mica–amphibole–rutile–ilmenite–diopside)suite of xenoliths in kimberlite. Geochimica et Cosmochimica Acta 41 (2), 309–323.

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.


De Carvalho, H., Tassinari, C., Alves, P.H., 2000. Geochronological review of thePrecambrian in western Angola: links with Brazil. Journal of African Earth Sciences31 (2), 383–402.

Ganga, J., Rotman, A.Y., Nosiko, S., 2003. Pipe Catoca, an example of the weakly erodedkimberlites from North-East of Angola. The 8th International Kimberlite Con-ference, Extended Abstract.

Garanin, V.K., Kudryavtseva, G.P., Posukhova, T.V., 1997. Indicators of diamondpreservation in kimberlite. In: Papunen, H. (Ed.), Mineral Deposits: Research andExploration, Where Do They Meet? In: Proceedings of the 4th Biennial SGAMeeting. Turku, Finland, pp. 767–770.

Gaspar, J.C., Wyllie, P.J., 1983. Ilmenite (high magnesium, manganese, niobium) in thecarbonatites from the Jacupiranga Complex, Brazil. American Mineralogist 68 (9–10),960–971.

Gaspar, J.C., Wyllie, P.J., 1984. The alleged kimberlite-carbonatite relationship: evidencefrom ilmenite and spinel from Premier and Wesselton Mines and the BenfonteinSill, South Africa. Contributions to Mineralogy and Petrology 85, 133–140.

Griffin, W.L., Ryan, C.G., 1995. Trace elements in indicator minerals: area selectionand target evaluation in diamond exploration. Journal of Geochemical Exploration53 (1–3), 311–337.

Grütter, H.S., Gurney, J.J., Menzies, A.H., Winter, F., 2004. An updated classification formantle-derived garnet, for use by diamond explorers. Lithos 77, 841–857.

Guiraud, R., Bosworth, W., Thierry, J., Delplanque, A., 2005. Phanerozoic geologicalevolution of Northern and Central Africa: an overview. Journal of African EarthSciences 43, 83–143.

Gurney, J.J., Helmstaedt, H., Moore, R.O., 1993. A review of the use and application ofmantle mineral geochemistry in diamond exploration. Pure and Applied Chemistry65 (12), 2423–2442.

Gurney, J.J., Zweistra, P., 1995. The interpretation of the major element compositions ofmantle minerals in diamond exploration. Journal of Geochemical Exploration 53(1–3), 293–309.

Haggerty, S.E., 1991a. Oxide textures - a mini-atlas. In: Lindsley, D.H. (Ed.), OxideMinerals: Petrologic and Magnetic Significance. Reviews in Mineralogy 25,129–219.

Haggerty, S.E., 1991b. Oxide mineralogy of the upper mantle. In: Lindsley, D.H. (Ed.),oxide minerals: petrologic and magnetic significance. Reviews in Mineralogy 25,355–416.

Haggerty, S.E., 1995. Upper mantle mineralogy. Journal of Geodynamics 20 (4),331–364.

Haggerty, S.E., Hardie III, R.B., McMahon, B.M., 1977. The composition of ilmenite noduleassociations from the Monastery diatreme: Proceedings of the 2nd InternationalKimberlite Conference, vol. 2, pp. 249–256.

Haggerty, S.E., Tompkins, L.A., 1983. Redox state of Earth's upper mantle fromkimberlitic ilmenites. Nature 303 (5915), 295–300.

Haggerty, S.E., Tompkins, L.A., 1984. Subsolidus reactions in kimberlitic ilmenites:exsolution, reduction and the redox state of the mantle. In: Kornprobst, J. (Ed.),Kimberlites I: Kimberlites and Related Rocks. Proceedings of the 3rd InternationalKimberlite Conference.

Head, J.W., Wilson, L., 2008. Integrated model for kimberlite ascent and eruption. 9th

International Kimberlite Conference Extended abstract No. 9IKC-A-00284.Hearn Jr., B.C., 2004. The Homestead kimberlite, central Montana, USA: mineralogy,

xenocrysts, and upper-mantle xenoliths. Lithos 77, 473–491.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, Pennsylva-nia. American Mineralogist 69, 30–40.

Hwang, P., Taylor, W.R., Rock, N.M.S., Ramsay, R.R., 1994. Mineralogy, geochemistry andpetrogenesis of the Metters Bore No. 1 lamproite pipe, Calwynyardah field, WestKimberley Province, Western Australia. Mineralogy and Petrology 51 (2–4),195–226.

Janse, A.J.A., Sheahan, P.A., 1995. Catalogue of world wide diamond and kimberliteoccurrences: a selective and annotative approach. Journal of GeochemicalExploration 53, 73–111.

Jones, A.P., Smith, J.V., Dawson, J.B., 1982. Mantle metasomatism in 14 veined peridotitesfrom Bultfontein mine, South Africa. Journal of Geology 90, 435–453.

Kaminsky, F.V., Sablukov, S.M., Sablukova, L.I., Shchukin, V.S., Canil, D., 2002. Kimberlitesfrom the Wawa area, Ontario. Canadian Journal of Earth Sciences 39 (12),1819–1838.

Kopylova, M.G., Russell, J.K., Cookenboo, H., 1999. Petrology of peridotite and pyroxenitexenoliths from the Jericho kimberlite: implications for the thermal state of themantle beneath the Slave craton, Northern Canada. Journal of Petrology 40 (1),79–104.

Kostrovitsky, S.I., Alymova, N.V., Yakovlev, D.A., Serov, I.V., Ivanov, A.S., Serov, V.P., 2006.Specific features of “picroilmenite” composition in various diamondiferous fields ofthe Yakutian province. Doklady Earth Sciences 406 (1), 19–23.

Kostrovitsky, S.I., Malkovets, V.G., Verichev, E.M., Garanin, V.K., Suvorova, L.V., 2004.Megacrysts from the Grib kimberlite pipe (Arkhangelsk Province, Russia). Lithos 77(1–4), 511–523.

Kostrovitsky, S.I., Piskunova, L.F., 1990. Two groups of symplectites from the sameKimberlite pipe. Doklady Earth Science Sections 306 (3), 176–179.

Kotel'nikov, D.D., Zinchuk, N.N., Zhukhlistov, A.P., 2005. Stages of serpentine andphlogopite transformation in the Catoca kimberlite pipe, Angola. Doklady EarthSciences 403 (6), 866–869.

Kriuchkov, A., Sarychev, I., Vouiko, V., Chipoia, B., Antoniuk, B., 2000. Kimberlite pipe ofCatoca. Geology, substantial analyses of the rocks and characteristics of diamonds.Geoluanda 2000 Abstract volume, 91–92.

Liipo, J.P., Vuollo, J.E., Nykänen, V.M., 1994. Pyrophanite and ilmenite in serpentinizedwehrlite from Ensilä, Kuhmo greenstone belt, Finland. European Journal ofMineralogy 6, 145–150.

Llusià Queral, R.M., Pereira Ginga, S., Galiano, A., Sebastião, P., Melgarejo Draper, J.C., OlimpioGonçalves, A., Proenza Fernandez, J., Morais, E., 2005a. Aluviões asociados ao campokimberlítico da zona de Muanga (Lunda Norte, Angola. XIV semana Geoquímica/VIIICongresso de Geoquimica dos Paises de Lingua Portuguesa, pp. 351–354.

Llusià Queral, R.M., Pereira Ginga, S., Proenza, J.A., José Ribamar, M.R., Sebastião, P.,Olimpio Gonçalves, A., Melgarejo Draper, J.C., Galiano, A.C., Alves Morais, E., 2005b.O campo kimberlítico de Muanga (Lunda Norte, Angola. XIV semana Geoquímica/VIII Congresso de Geoquimica dos Paises de Lingua Portuguesa, pp. 347–350.

Meyer, H.O.A., McCallum, M.E., 1986. Mineral inclusions in diamonds from the Sloankimberlites, Colorado. Journal of Geology 94 (4), 600–612.

Mitchell, R.H., 1973. Magnesian ilmenite and its role in kimberlite petrogenesis. Journalof Geology 81, 301–311.

Mitchell, R.H., 1989. Kimberlites: mineralogy, geochemistry and petrology. New York,Plenum Press. 442 pp.

Mitchell, R.H., 1995. Kimberlites, orangeites, and related rocks. New York, Plenum Press.410 pp.

Mitchell, R.H., 1997. Kimberlites, orangeites, lamproites, melilitites, and minettes: apetrographic atlas. Almaz Press Inc, Ontario.

Mitchell, R.H., et al., 2009, this issue. Tuffisitic kimberlites from the Wesselton Mine,South Africa: Mineralogical characteristics relevant to their formation. Proceedingsof the 9th International Kimberlite Conference. Lithos 112S, 452–464.

Moore, A.E., 1987. A model for the origin of ilmenite in kimberlite and diamond:implications for the genesis of the discrete nodule (megacryst) suite. Contributionsto Mineralogy and Petrology 95, 245–253.

Mücke, A., Woakes, M., 1986. Pyrophanite: a typical mineral in the Pan-African Provinceof Western and Central Nigeria. Journal of African Earth Sciences 5 (6), 675–689.

Nielsen, T.F.D., Sand, K., 2008. The Majuagaa kimberlite dike, Maniitsoq region, WestGreenland: constraints on an Mg-rich silico-carbonatitic melt composition fromgroundmass mineralogy and bulk compositions. The Canadian Mineralogist 46,1043–1061.

Pasteris, J., 1980. The significance of groundmass ilmenite and megacryst ilmenite inkimberlites. Contributions to Mineralogy and Petrology 75 (4), 315–325.

Read, G.H., Janse, A.J.A.(B.), 2009, this issue. Diamonds: Exploration, mines andmarketing. Proceedings of the 9th International Kimberlite Conference. Lithos 112S,1–9.

Reis, B., 1972. Preliminary note on the distribution and tectonic control of kimberlites inAngola: The 24th International Geological Congress - Section 4, pp. 276–281.

Rogers, A.J., Grütter, H.S., 2009, this issue. Fe-rich and Na-rich megacryst clinopyroxeneand garnet from the Luxinga kimberlite cluster, Lunda Sul, Angola. Proceedings ofthe 9th International Kimberlite Conference. Lithos 112S, 942–950.

Schulze, D.J., 1984. Cr-poor megacrysts from the Hamilton Branch kimberlite, ElliottCounty, Kentucky. In: Kornprobst, J. (Ed.), Kimberlites II: The Mantle and Crust–MantleRelationships. In: Developments in Petrology 11B. Amsterdam, Elsevier, pp. 97–108.

Scott Smith, B.H., Nowicki, T.E., Russell, J.K., Webb, K.J., Hetman, C.M., Harder, M.,Mitchell, R.H., 2008. Kimberlites: Descriptive Geological Nomenclature andClassification. 9th International Kimberlite Conference Extended abstract No.9IKC-A-00124.

Smith, C.B., Sims, K., Chimuka, L., Duffin, A., Beard, A.D., Townend, R., 2004. Kimberlitemetasomatism at Murowa and Sese pipes, Zimbabwe. Lithos 76 (1–4), 219–232.

Sparks, R.S.J., Baker, L., Brown, R.J., Field, M., Schumacher, J., Stripp, G., Walters, A., 2006.Dynamical constraints on kimberlite volcanism. Journal of Volcanology andGeothermal Research 155 (1–2), 18–48.

Stiefenhofer, J., Viljoen, K.S., Marsh, J.S., 1997. Petrology and geochemistry of peridotitexenoliths from the Letlhakane kimberlites, Botswana. Contributions to Mineralogyand Petrology 127 (1–2), 147–158.

Tompkins, L.A., Haggerty, S.E., 1984. Koidu Kimberlite Complex, Sierra Leone. GeologicalSetting, Petrology and Mineral Chemistry. In: Kornprobst, J. (Ed.), Kimberlites I:Kimberlites and Related Rocks. Proceedings of the 3rd International KimberliteConference.

Tompkins, L.A., Haggerty, S.E., 1985. Groundmass oxide minerals in the Koidu kimberlitedikes, Sierra Leone, West Africa. Contributions to Mineralogy and Petrology 91 (3),245–263.

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 Northpyroclastic kimberlite, Canada. Journal of Volcanology and Geothermal Research 17,128–138.

Wyatt, B.A., Mike, B., Anckar, E., Grütter, H., 2004. Compositional classification of“kimberlitic” and “non-kimberlitic” ilmenite. Lithos 77, 819–840.



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.

159

mmacla nº 11. septiembre ‘09

revista de la sociedad española de mineralogía

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)

IINTRODUCTION.

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). Cucumbi samples exhibit crater facies along the first 100 m, characterized by volcanoclastic rocks, and diatreme facies, showing typical tuffisitic kimberlite (Mitchell et al., 2009). METHODOLOGY. Thin and polished sections were studied using transmitted and reflected optical microscopy, followed by SEM-BSE-EDS analysis. Chemical analyses were obtained with EPMA. GEOLOGICAL SETTING. Angola has a complex geological history that can be represented by three main stages (De Carvalho et al., 2000; Fig. 1): (1) An important Archaean orogeny, registered by the Central Shield, Cuango Shield and Lunda Shield, most of them

composed by 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 volcanosedimentary groups, gneisses, migmatites, granites and syenites. (3) Unconformably lying Phanerozoic sequences, which are the result of the Pangea formation and the consecutive breaking-up of Gondwana, that contributed to the formation of rift basins associated to fault systems which later allowed the apparition of marine sequences, the origin of the Karoo Supergroup, intraplate magmatism

(alkaline, carbonatitic, kimberlitic) and marginal basins. The Lower Cretaceous regional extension determined the development of deep faults and grabens with trends NE-SW and NW-SE. The Lucapa structure corresponds to the first group, and the NE part concentrates most of the diamondiferous kimberlites in Angola, including Cucumbi, whereas the southwestern zone comprises important outcrops of undersaturated alkaline rocks and carbonatites (Reis, 1972). Other minor kimberlite fields are found in the SW Angola (Egorov et al., 2007). PETROGRAPHY AND COMPOSITION

The Cucumbi samples exhibit all the main characteristic features of Tuffisitic 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]

fig 1. General location of kimberlites in Angola. Modified after De Carvalho et al. (2000) and Egorov et al. (2007).

160

generally massive, poorly sorted, clast-supported rocks with the following main components: anhedral olivine macrocrysts, pseudomorphosed by serpentine and smectite; other mega- and macrocrysts as garnet, ilmenite, clinopyroxene, and phlogopite, whether enclosed in a pelletal assemblage of serpentine or not, often pelletal lapilli, and an interclast groundmass in the matrix, mainly composed by serpentine, less common by chlorite, smectite and calcite. The size and distribution of mega- and macrocrysts is chaotic (Figs. 2, 3).

ffig 2. Cucumbi, a diamondiferous drill hole. A typical pattern of a tuffisitic kimberlite (TK) facies, with macrocrysts containing rounded pseudomorphosed olivine xenocrysts, rounded ilmenite xenocrysts and crustal rock xenoliths, all set in a groundmass of serpentine and phlogopite. Image from the scanned thin section.

fig 3. Xenocrystals of olivine (pseudomorphosed by serpentine (Srp)), phlogopite (Phl), ilmenite (Ilm) in a serpentine groundmass. SEM image, mode BSE.

Magnesian ilmenite (9-13 wt.% MgO) is present as rounded mega- and macrocrysts (fig. 3), as part of xenoliths and as inclusions in phlogopite. In some cases macrocrysts of ilmenite are partially replaced along the borders by perovskite and spinel (Fig. 3). Ilmenite texture is usually either cumulus or homogenous. Symplectite textures are lacking in this kimberlite, in contrast with Catoca.

Garnet and clinopyroxene are usually present as mega- and macrocrysts, and only rarely as part of xenoliths. Garnet composition is diverse. Using the garnet classification of Grütter et al. (2004), it may be stated (Fig. 4) that some garnet derive from lherzolite (G9) and others from pyroxenite and eclogite (G4, G5), only a few of them come from uncommon, unusual or “polymict” mantle lithologies.

DISCUSSION AND CONCLUSIONS. Using the diagram of Grütter et al. (2004) to plot the garnet compositions from Cucumbi, it should be pointed out that all these compositions plot into the graphite domain, out of the diamondiferous field harzburgitic G10 facies. Therefore, this kimberlite could be classified as barren using only that criterion. However, the Cucumbi kimberlite has proven to be diamondiferous. In fact, similar problems were found in the Catoca pipe when using the composition of ilmenite (Robles-Cruz et al., 2009) or the composition of garnets. Therefore, the garnet diagrams can be used to verify the minimum level of diamond content, but some kimberlites may contain diamond samples from deeper sources. Hence, it should be taken into consideration when using these diagrams to assess the potential of kimberlite fields.

fig 4. Classification of the Cucumbi garnets in a plot Cr2O3 versus CaO (wt.%), according with the compositional fields of Grütter et al. (2004).

ACKNOWLEDGEMENTS. This research was supported by the projects CGL2005-07885/BTE and CGL2006-12973 of Ministerio de Educación y Ciencia (Spain), the AGAUR SGR 589 of Generalitat de Catalunya and a FI-2006 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 geologists, who kindly allowed us to acquire samples for this study and gave all facilities for the mine trip. The authors also acknowledge the Serveis Cientificotècnics de la Universitat de Barcelona for assistance in the use of SEM/ESEM-BSE-EDS analyses (E. Prats, Dr. J. García Veigas) and EPMA (Dr. Xavier Llovet). REFERENCES. De Carvalho, H., Tassinari, C., Alves, P.H

(2000): Geochronological review of the Precambrian in western Angola: links with Brazil. Journal of African Earth Sciences 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 Geology and Geophysics 48, 323-336.

Grütter, H.S., Gurney, J.J., Menzies, A.H., Winter, F. (2004): An updated classification for mantle-derived garnet, for use by diamond explorers. Lithos 77, 841-857.

Mitchell, R. H., Skinner, E. M., Scott-Smith, B. H. (2009): Tuffisitic Kimberlites: Mineralogical Characteristics Relevant to their Formation. Lithos, Special Issue 9IK., in press

Reis, B. (1972): Preliminary note on the distribution and tectonic control of kimberlites in Angola: The 24th International Geological Congress - Section 4, 276-281.

Robles-Cruz, S.E., Watangua, M., Isidoro, L., 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.


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

��

��

��

�� !�# �*��<��@��!�#[\]�^��`��^��^{|��

��^�� }�� *�� !�# �*��<��{��!�#[

��{�<�� }�~�� !�# �*��<�� }��!�� }!�#[

��[�� {�� }!��<��!�#[

� �� }�� }�� }�� { �� }� �� }�} �<� �� {}�� <��{ }�� ^�^!� ��*��!� �� <�� }�� ^� � ��{� ��!� �}!� � }� �� }�� #�� <� � � ��*��*�� }��*��{��{}�� } ��<��*�� }� �� <� �� {�� !��^

� � �� {}<� �� *�� <��}� ��*�� }�*�}{��{ }�� *�� }�� {�� !�@�� *�� }��{�� }��<�*��<�� }�� ^�@�� {�\��|�� }� ��{�� \��| � �<�� *� { �� }� �� *�� }��}� �� ^�� ^� @{�� {�� }�*�}{�� {� � }� *��< �� ^�� ^��^��}�� ^ � � �� }� *�� }�� {�� }!� }�� }� �<�� <��!�� }��<�� }��}��*�}��*�� {��^

��! �}� *�� *�� }�� {��!��{�� }�� ^�

��{��{ ��<��}��*��}�� {� �� <�� <� �*�� {��<!� �� *�� {�� <}�� !�� <��*��{�<�� }^��<��*�� }�� *��{�<��{�� }!��!��{}�� }��*��^

�� !��^�^�\��| Earth Planet. Sci. Lett., #�!��^�� !��^� �� ^� \��|�Geology! �$��^� �� !��^�^��^�\��| Chem. Geol.! �%��

�� !�"#��

& '��*��+;��<��<�=�� > ��?��@�+��Z��[�� Z @\��]��^_� ��=�� @��$

��^�}��!�� !��{��}��!�# �*��}�

@�� !�� !�� \]�� }��`{�^�}{|�� { ��

��!��{� }�!�� ^�}��!��{}�}��}�� !�# �*��}�}��

�� !��{� }�!�� {�*�<�� }�!��!�� }�

�� }�� }��{�� }� �{�{�� }� � � �� [�� !� � ��^� �� {�� *�� }� �<� �� }��@��{}��^�

�� <�� { }� � � �� !�� <�� \�� }� �� }��}�|!� � }� �� \�� }� �� }��}�|^� �� { }� �� <� � � ��}��^� � � �� !� �� {�{�� <�� <�� \�� !� }��}�� }� ��|!� �� \�� }�� !� �� }��}�!� � �� }� ��|!� �� {�� \��!�� !�}��}�|�� }��^

�� }�� }�� }� ��}� � �� }�� }� �� ^� �� !� �� {�{�<�� {��}�� }��<�� ^� �� <�� }^� �� }��<� �� !� �� {�� { }� �� }� �� ^� # �� {�{��!� �� *�{��^�� ! �{�� *�{�� <�\��{�� }��*�{�| ��^� � <� �� <�� *�{�� ^�

�� }�� }� �<� �� *�� { }�� }�� {�� }�� }� }�� }�� ^� �� !� �� }��<��{�{�� }� �� {�� *��{�<�}��}��}�� <� � �� ^� � � �� {�{��!� �� {��}�� <�� {��}�� }�� ^

�� !��^�^��^�\��| Lithos! ``!��^�� !��^��^�\��| Can. Mineral., �$!��^ ��!��^��^�\��|�Earth. Planet. Sci. Lett^!�$#!��^


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.

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, Spainb CONICET-IDEAN Instituto de Estudios Andinos, Laboratorio de Tectónica Andina, Universidad de Buenos Aires, C1033AAJ Capital Federal, Argentinac Geological Survey of Canada, 601 Booth Street, Ottawa, Ont. K1A 0E8, Canadad Sociedade Mineira de Catoca, Catoca, Lunda Sul, Angolae Departamento de Geologia, Universidade Agostinho Neto, Av. 4 de Fevereiro 7, 815 Luanda, Angola

a b s t r a c ta r t i c l e i n f o

Article history:

Received 9 August 2011

Received in revised form 2 April 2012

Accepted 4 April 2012

Available online 15 April 2012

Editor: K. Mezger

Keywords:

U–Pb dating

SHRIMP

Geochronology

Zircon

Diamond

Catoca kimberlite

We present 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 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 may

reflect different sources of kimberlitic magma, with some of the grains produced in U- and Th-enriched meta-

somatized mantle units. This idea is consistent with the two populations of zircon identified in this study. 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 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

Kimberlites contain primary minerals crystallized from a kimberlitic

magma, a suite of mega- and macrocrysts (e.g., zircon, diamond), and a

complex variety of xenoliths (e.g., peridotite). According to previous

studies (Moore et al., 1992; Griffin et al., 2000; Pivin et al., 2009; and ref-

erences therein), some zircon megacrysts (crystals larger than 1 cm)

that do crystallize from fractionating magmas in the mantle, are inter-

grown with other megacryst phases (ilmenite, phlogopite, high-Fe

olivine), and contain inclusions of chromian diopside, chromite, and

diamond. Mantle-derived zircon megacrysts have trace element com-

positions that are distinct from zircon derived from the crust. These

megacrysts have lower U and Th concentrations, and a lower total

abundance of rare-earth elements (REE) than zircon of crustal deriva-

tion (Belousova et al., 2002; Heaman et al., 2006; and references there-

in). The U–Pb dating of zircon is by far the most widely used method for

obtaining reliable mineral-growth ages from different types of rocks.

Zircon can provide reliable ages because of its resistance to thermal dis-

turbances. However, the dating of kimberlites is one of the most difficult

applications of this method because crystal growth may have occurred

hundreds of millions of years before kimberlite emplacement.

The aim of this study is to determine the U–Pb ages of zircon from

the diamondiferous Catoca kimberlite, in the Lunda Sul province of

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

for diamond exploration. Trace-element analyses carried out by Laser

Ablation-Inductively Coupled Plasma-Mass Spectrometry (LA-ICP-MS)

Chemical Geology 310-311 (2012) 137–147

⁎ Corresponding author. Tel.: +34 506 8839 3981; fax: +34 506 2242 4411.

E-mail addresses: [email protected], [email protected] (S.E. Robles-Cruz).

0009-2541/$ – see front matter © 2012 Elsevier B.V. All rights reserved.

doi:10.1016/j.chemgeo.2012.04.001

Contents lists available at SciVerse ScienceDirect

Chemical Geology

j ourna l homepage: www.e lsev ie r .com/ locate /chemgeo

were used to determine the chemical composition of the zircon (includ-

ing the REE) from the kimberlite and to suggest potential sources of the

zircon. We also present a comparison of the Catoca kimberlite with

kimberlites in southeastern Brazil that were formed during the Early

Cretaceous.

1.1. Background information and previous geochronological research in

Angola

The kimberlites of Angola are distributed in clusters (Pereira et al.,

2003; Egorov et al., 2007) in the northeastern, central, and southwestern

areas of the country. Most of the kimberlites are concentrated within the

Lucapa structure, with a NE–SW orientation, or along NW–SE faults

(Fig. 1). The Lucapa structure is a major basement fault system with

the highest diamondiferous potential in Angola (Pereira et al., 2003).

The Catoca kimberlite, the fourth largest kimberlite pipe in the world

(639,000 m2), is located in the northeastern part of this structure, and

exhibits rocks of crater and diatreme facies.

In the earliest geochronological studies, Bardet and Vachette (1966)

proposed a main Cretaceous kimberlitic event in the Congo craton

based on stratigraphic relationships. Subsequently, Davis (1977) dated

one crystal of zircon from the Val do Queve kimberlite, located in the

central part of Angola by conventional U–Pb TIMS analysis. He reported

a 206Pb–238U age of 134.0±2.0 Ma. However, Davis noted some unspe-

cified analytical problems, which he attributed to the crystal's low U

content. Later, Haggerty et al. (1983) carried out fission-track dating

of zircon from the same area and reported an age of 133.4±11.5 Ma,

which they interpreted as the time of eruption of the Val do Queve

kimberlite. In contrast, Egorov et al. (2007) published a K–Ar date of

372±8 Ma for phlogopite from the groundmass of the Chicuatite

kimberlite in southwestern Angola and interpreted it as the kimberlite

age. Eley et al. (2008) reported a 206Pb/238U age for zircon from the

Alto Cuilo 55 kimberlite and the Alto Cuilo 197 kimberlite (each date

was based on a single-point analysis) of 113.0±0.8 Ma. The same

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

of 145.1±4.0 Ma (the authors did not specify the technique that was

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

Luxinga kimberlite cluster, located ca. 85 km southwest of Catoca.

According to Eley et al. (2008), the ages indicate that kimberlite intru-

sive activity took place in the Luxinga cluster between approximately

145 and 113 Ma. Recently, Jelsma et al. (2012) reported U–Pb ages for

zircon from kimberlites in central Angola between 252 and 216 Ma

(median age of 235 Ma), using LA-ICP-MS. The authors interpreted

these data as a new age population of kimberlite emplacement in

Angola.

2. Geological setting

The geological history of Angola includes the following three

major phases, which have shaped the country (De Carvalho et al.,

2000; Guiraud et al., 2005; Egorov et al., 2007): (1) the Archean orog-

eny; (2) the Proterozoic orogenic cycles (Eburnian: Paleoproterozoic,

Kibaran: Mesoproterozoic, and Pan-African: Neoproterozoic); and

(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 develop-

ment of basins that are associated with deep fault systems in Angola.

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).

138 S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147

These fault systems facilitated the emplacement of alkaline, carbonatitic,

and kimberlitic magmas (Pereira et al., 2003).

The Late Cretaceous regional extension was associated with older

deep-seated faults and “grabens” with NE–SW and NW–SE trends

(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

Angola along a line that continues southwest to a transform fault of the

Mid-Atlantic Ridge (White et al., 1995). The Lucapa structure has been a

belt of recurring tectonic weakness since the Paleoproterozoic (Jelsma

et al., 2009). Most of the diamondiferous kimberlites in Angola are

located along the Lucapa structure in northeastern Angola, although it

is not clear whether kimberlites were emplaced at the time of rifting

or whether they resulted from post-rifting events. Strike-slip and

shear fault systems in northeastern Angola likely could have led to

decompression (local extension) and compression, resulting in some

control on the distribution of igneous activity within the Lucapa struc-

ture. These processes could also have different expressions within the

Angolan Shield and Kasai craton. Reliable age determinations are very

important for understanding the timing of these intracontinental pro-

cesses. In the southwestern part of Angola, there are outcrops of under-

saturated alkaline rocks and carbonatites along this trend (Reis, 1972),

as well as some minor kimberlite fields (Egorov et al., 2007).

Continental sediments that unconformably overlie the Precambri-

an basement filled the Lucapa structure during the Cretaceous and

Paleogene. An example of this package of sediments is the Calonda

Formation, a fining upward lithostratigraphic unit of the Kwango

Group that was formed by torrential deposits gradually changing to

lagoonal facies and concluding with low-energy deposits associated

with aeolian episodes. The Calonda Formation is the oldest sedimen-

tary unit in Angola that contains detrital diamond and kimberlite

clasts (Pereira et al., 2003; and references therein). It is reported to

be Albian to Cenomanian in age, based on fish macrofossils, palyno-

morphs, and tectonostratigraphic studies (Pereira et al., 2003). It

has become an important target in the exploration for alluvial dia-

mond deposits.

3. Sample description

In this study, nineteen crystals of zircon between 0.6 and 4 mm in

length were taken from three core samples (four crystals) of tuffisitic

kimberlites (TK) and three samples (fifteen crystals) of heavy-

mineral concentrate from the Catoca kimberlite to perform the

SHRIMP analyses (Table 1). The TK have macrocrysts (0.5–10 mm)

of olivine (35–50 modal %), which are, in most cases, completely

replaced by serpentine, calcite, and saponite. Xenoliths of the host

rocks (3–5 modal %), comprising gneiss and amphibolites, are com-

mon. Also present are mantle xenoliths (1–5 modal %) (e.g., altered

metasomatized peridotite) and xenoliths of carbonatite (calcite+

anhedral Mn-rich ilmenite+zircon+apatite+phlogopite). Garnet

(G9 and G10, Grütter et al., 2004), chromian diopside, ilmenite,

amphibole, phlogopite, and zircon are found as mega- and macrocrysts

in an altered kimberlite groundmass. Some zircon crystals exhibit a

reaction rim of baddeleyite. The matrix of the TK rocks contains lizar-

dite, apatite, calcite, ilmenite, and chromite. Titanite, zirconolite, badde-

leyite, barite, dolomite, witherite, barytocalcite, strontianite, and

sulfides have also been identified in the matrix by quantitative powder

X-ray diffraction (powder method) and quantitative chemical analyses

using an electron-microprobe (EMP) (Robles-Cruz et al., 2009). One

additional zircon crystal derived from a heavy-mineral concentrate

from the Tchiuzo kimberlite (15 km north of the Catoca kimberlite)

was added to this study for comparative purposes.

Large grain-size is a characteristic feature of zircon from kimberlites.

According to several authors (e.g., Belousova et al., 1998), zircon crystals

found in kimberlites are relatively large in size (several millimeters) com-

pared to most zircon crystals from other types of igneous rock or from

metamorphic rocks. These crystals vary in color from colorless to

Table 1

Description of zircon crystals analyzed by SHRIMP.

Kimberlite Borhole Sample

name/mount

Type of

sample

Crystal

no.

Width

(μm)

Length

(μm)

Crystal

border

Zonation (using

cathodoluminescence

images)

Presence

of rim of

baddeleyite

Degree of

fracturing

Type of

analyses

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

core

NO M SHRIMP

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

sectors

NO L SHRIMP

14 700 800 Subangular Oscillatory and patchy

in fracture

NO M SHRIMP

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

patchy core/dark-CL

concentric rim

NO M SHRIMP

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.

139S.E. Robles-Cruz et al. / Chemical Geology 310-311 (2012) 137–147

brownish yellow. The twenty crystals analyzed in this study are charac-

terized by an almost complete absence of crystal faces, and most of

them are angular to subangular, presumably as a result of fracturing. A

few of them are subrounded, likely owing to disequilibrium with the me-

dium, causing incipient resorption during interaction with the kimberlitic

magma.

According to back-scattered electron (BSE) images, the twenty

zircon crystals are homogeneous at the major element level. The crys-

tals do not exhibit evidence of metamictization. Nineteen of the

twenty crystals exhibit oscillatory zoning in cathodoluminescence

(CL) (Fig. 2B, C, D, and F), which is usually interpreted to be a result

of crystallization in a melt or fluid (Belousova et al., 1998; Liati et

al., 2004; Page et al., 2007). Crystal no. 18 has dark-CL growth zoning

(Fig. 2E), which is indicative of high U content. Four zircon crystals in

the analyzed set exhibit partial replacement by baddeleyite (ZrO2)

along the borders. In addition, these crystals are fractured, and the

resulting fracture surfaces are not corroded by baddeleyite (Fig. 2A,

B). The baddeleyite crystals are irregular and narrow between 10

and 40 μm in breadth.

Three representative crystals (nos. 19, 21, and 22), between 1 and

3 mm in length and picked from the heavy-mineral concentrate of the

Catoca TK, were analyzed with an LA-ICP-MS. These zircon crystals

exhibit oscillatory zoning in CL. One of these crystals was also ana-

lyzed with SHRIMP (no. 19).

4. Analytical methods

4.1. Determinations of trace element concentrations

Trace-element analyses were carried out by LA-ICP-MS on three

zircon crystals (nos. 19, 21, and 22) from the Catoca kimberlite at the

Geological Survey of Canada. The trace-element determinations were

performed using a New Wave Research UP213 laser-ablation system in

combination with a Perkin Elmer 6100DRC quadrupole inductively

coupled plasma mass spectrometer. The data acquisition and calibration

protocols employed have been described by Longerich et al. (1996) and

Jackson (2008). Operating conditions and data-acquisition parameters

used in this study are summarized in Table 2. Data reduction was

performed using the software GLITTER 4.4.2 (Griffin et al., 2008). The

standard NIST SRM 610 (synthetic glass reference material, National

Institute of Standards and Technology) was used as the primary

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.


georem.mpch-mainz.gwdg.de/and data downloaded February 24,

2010). The stoichiometric SiO2 content of 32.8% was used for internal

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

Survey: Jochum and Stoll, 2008), was used to monitor the precision

and accuracy of the technique. Precision and accuracy were assessed

from repeated analyses of the BCR-2G standard and were usually better

than 10% for concentrations at the ppm level. Detection limits were

better than ~0.06 ppm for all elements reported, with the exception of Ti.

4.2. U–Pb dating

The zircon crystals were mounted in epoxy along with fragments

of laboratory standard zircon z6266 (206Pb/238U age=559 Ma) at

the Geological Survey of Canada. The mid-sections of the zircon crys-

tals were exposed and characterized in back-scattered electron (BSE)

mode utilizing a Zeiss Evo 50 scanning electron microscope and a cold

cathodoluminescence stage mounted on a petrographic microscope

to study internal features within the crystals, such as zoning and

structures. The surfaces of the 2.5-cm mounts were evaporatively

coated with 10 nm of high-purity Au.

The U–Pb analyses were conducted at the Geochronology Laboratory,

Geological Survey of Canada (Ottawa), using a Sensitive High Resolution

Ion Microprobe II (SHRIMP II). Analyses were performed using an 16O−

primary beam, projected onto the zircon crystals at 10 kV with a beam

current of ca. 10 nA. The sputtered area used for analysis was ca. 25 μm

in diameter. The count rates at ten masses including background were

sequentially measured over six scans with a single electron multiplier

with a deadtime of 27 ns. Off-line data processing was accomplished

using the SQUID 2.22.08.04.30 software, rev. 30 Apr 2008. The 1σ exter-

nal errors of the 206Pb/238U ratios reported in the table of data (Table 4)

incorporate a ±1.1% error in calibrating the standard zircon (see Stern

and Amelin, 2003). For the common Pb correction, we utilized the

Pb composition of the surface blank. No fractionation correction was

applied to the Pb-isotope data. The 207Pb method (Williams, 1998) was

used to calculate 206Pb/238U ages and errors.

5. Results

5.1. REE and Ti-in zircon thermometry

On the basis of total REE concentrations in zircon from the Catoca

kimberlite, we identified two ranges of values. One set consisted of

total REE concentrations of less than 25 ppm, and the other set was

characterized by total REE concentrations up to 123 ppm (Table 3). The

zircon from the Catoca kimberlite (Fig. 3) exhibits low concentrations

of light REE (LREE), a positive anomaly in Ce (chondrite-normalized

value up to 10.04 ppm, 9.6 ppm absolute concentration), a lack of an

Eu anomaly, and a positive slope from Pr to Lu, which flattens toward

the heavy REE (HREE). Similar chondrite-normalized patterns and REE

contents have been observed in zircon crystals from other kimberlites

in southern Africa, Yakutia, and Australia (Belousova et al., 1998, 2001).

The REE patterns in zircon indicate that HREE and Ce4+, with a much

more compatible ionic radius (0.97 Å) compared to Ce3+ (1.18 Å), are

preferentially incorporated in the zircon structure (Ballard et al., 2002;

Hoskin and Schaltegger, 2003). This anomalous abundance of Ce4+

could be the result of an increased Ce4+/Ce3+ value in the melt as a

function of high oxygen fugacity (Belousova et al., 2002; Hoskin and

Schaltegger, 2003; Whitehouse and Platt, 2003; and references there-

in). Similarly, the absence of an Eu anomaly may be caused by a high

Eu3+/Eu2+ value under oxidized magma conditions in a feldspar-free

magmatic environment (Hoskin and Schaltegger, 2003). The suggestion

of relatively high oxygen fugacity (fO2) conditions during zircon crystal-

lization seems more favorable for an interpretation of a positive Ce

anomaly combined with the absence of a negative Eu anomaly.

Concentrations of yttrium exhibit distinct ranges of values for each

zircon population (Table 3). In one population, Y concentrations are

Table 2

LA-ICP-MS operating conditions and data-acquisition parameters.

LA

Model New Wave Research UP213

Wavelength 213 nm

Pulse duration (FWHM) ca. 4 ns

Nominal spot sizes used 80–120 μm

Repetition rate 10 Hz

Energy density at sample ca. 5 J/cm2

ICP-MS

Model Perkin Elmer ELAN 6100DRC

Carrier has flow (He) 0.96 L/min

Make-up flow (Ar) 0.70 L/min

Sampler and skimmer Nickel

Data-acquisition parameters

Data-acquisition protocol Time-resolved analysis

Detector mode Pulse counting (b3 M c.p.s.)

Isotopes determined 25Mg, 29Si, 39K, 42Ca, 49Ti, 51V, 53Cr, 55Mn, 57Fe, 59Co,60Ni, 65Cu, 66Zn, 71Ga, 72Ge, 85Rb, 88Sr, 89Y, 90Zr, 93Nb,133Cs, 137Ba, 139La, 140Ce, 141Pr, 146Nd, 147Sm, 151Eu,157Gd, 159Tb, 163Dy, 165Ho, 167Er, 169Tm, 173Yb, 175Lu,177Hf, 181Ta, 205Tl, 206Pb, 207Pb, 208Pb, 232Th, 238U

Scanning mode Peak hopping, 1 point per peak

Dwell time per isotope 10 ms

Time per mass sweep ca. 453 ms

Data acquisition time 180 s (ca. 60 s gas blank, up to ca. 120 s ablation)

Oxide production ThO+/Th+b1%

Standards and calibration

Samples Polished 25 mm round mounts

Data-processing software GLITTER 4.4.2 (Griffin et al., 2008)

Calibration standard NIST SRM 610, GEOREM Preferred Values, Feb. 2010

Internal standard SiO2 (32.8 wt.%)

Secondary standard USGS microbeam standard, BCR-2G

Table 3

Results of Laser-ablation ICP-MS analyses of zircon from the Catoca kimberlite.

Kimberlite Catoca Catoca Catoca Catoca Catoca

Crystal No. 22 22 21 21 19

Type of

sample

Concentrate Concentrate Concentrate Concentrate Concentrate

Spot

location

Core Core Core Rim Core

Sample

number

fe26b05 fe26b06 fe26b09 fe26b10 fe26b15

All values are reported in ppm

Ti 5.4 10.9 1.17 2.66 3.4

Sr 0.045 2.85 0.34 0.61 0.61

Y 46 120 330 480 39

Nb 1.25 4.1 8.8 13.5 0.86

Ba 0.088 4.3 0.59 1.34 0.38

La 0.0181 0.50 0.054 0.168 0.075

Ce 0.79 2.89 5.3 9.6 0.82

Pr 0.031 0.211 0.187 0.39 0.0264

Nd 0.39 1.12 2.64 5.5 0.214

Sm 0.73 1.04 4.4 7.7 0.241

Eu 0.44 0.64 2.89 5.0 0.243

Gd 2.66 4.4 17.3 28.5 1.03

Tb 0.76 1.50 5.2 8.0 0.52

Dy 6.5 16.2 51 79 5.1

Ho 1.96 5.0 15.2 22.8 1.69

Er 6.4 17.7 54 77 6.2

Tm 1.13 3.2 10.4 14.9 1.41

Yb 8.8 25.4 86 123 12.9

Lu 1.42 3.8 11.7 15.1 1.50

Hf 12,800 13,700 9,180 9,970 9,580

Th 2.90 9.3 47 78 2.04

U 7.9 26.8 112 152 10.0


between 39 and 120 ppm, while higher Y concentrations, between

330 and 480 ppm, are found in the other population. Concentrations

of Hf are between 9580 and 13,700 ppm for the first population and

between 9180 and 9970 ppm for the second population of zircon.

These three zircon crystals exhibit low Ti concentration, between

3.4 and 10.9 ppm, for the first population and between 1.17 and

2.66 ppm for the second population. There is no direct evidence that

these zircon crystals coexisted with a Ti-dominant phase. However,

ilmenite is usually found in xenoliths, as mega- and macrocrystals,

and in the groundmass of the Catoca kimberlite.

The Ti-in-zircon thermometer (Watson et al., 2006) was applied

to calculate the temperature at which the zircon crystallized. The

calculated temperatures are between 600 and 750 °C, which is very

low for a kimberlite. One possible explanation for this result is the

lack of coexistence of zircon with a Ti-dominant phase, which leads

to uncertainty in the activity coefficient of Ti. The same problem has

been reported previously, where some doubts have been raised

about the applicability of the Ti-in-zircon thermometer for zircon

from kimberlites (Page et al., 2007).

5.2. SHRIMP U–Pb ages

Forty-one SHRIMP analyses were performed on 19 zircon crystals

from the Catoca kimberlite. Table 4 summarizes the results of all the

SHRIMP analyses. Concentrations of Th and U range between 1 and

654 ppm and between 6 and 326 ppm, respectively. These ranges of

values, along with the REE analyses, appear to support the idea

that the zircon crystals originated from different sources. The Th/U

values range from 0.21 to 2.07. Several analyses show very low U con-

centrations and very high proportions of common Pb, up to 45%. The

large amount of common Pb is evident in a Tera–Wasserburg plot

(Fig. 5). Samples with common Pb concentrations of greater than

15% have higher 206Pb/238U ages, which is attributed to uncertainties

that result from the very large common Pb correction required in

those cases. For this reason, all results with common Pb proportions

greater than 15% were excluded from interpretations (7 of 41 analyses

rejected, Fig. 4). The 34 analyses below the 15% cut-off value define a

single weighted mean age of 117.9±0.7 Ma (Mean square weighted

deviation MSWD=1.3, probability 0.093, Fig. 4B). Quadratic addition

of the systematic error in the mount calibration (0.3% 1 sigma) gives a

total error estimate of 117.9±0.7 Ma.

Five SHRIMP analyses were carried out on one zircon crystal (no.

20) from the Tchiuzo kimberlite, for comparison with the results dis-

cussed above. Concentrations of Th range between 11 and 34 ppm,

and concentrations of U range from 36 to 61 ppm. The Th/U values

range between 0.30 and 0.56. The weighted mean of the five analyses

gives an age of 121±3 Ma (MSWD=1.6; probability of fit 0.17). Two

analyses have common Pb contents above 15%; if these are rejected,

the remaining three analyses have an identical weighted mean age

of 121.2±1.8 (MSWD=0.86; probability of fit 0.45). Thus, the single

crystal from the Tchiuzo kimberlite yields an age slightly older than

the weighted mean age obtained for the Catoca kimberlite.

6. Discussion

6.1. Different sources of zircon

The zircon crystals have well-defined characteristic features in

their chondrite-normalized patterns, total REE abundances (Fig. 3),

and different U and Th concentrations (Table 3). On the basis of their

REE composition, two different populations of zircon crystals have

been identified in the Catoca kimberlite. The first population is charac-

terized by a low concentration of REE (less than 25 ppm), U (less than

30 ppm), Th (less than 10 ppm), and Thzrn/Uzrn values (0.20–0.37). Ex-

perimental and theoretical studies in mafic and ultramafic rocks indicated

that in such rocks, Uzrn/Umelt is approximately 100 and (Th/U)zrn–melt

equals approximately 0.17 (Blundy and Wood, 2003). On the basis of

results obtained in this study (Thzrn/Uzrn≈0.26, Table 4), we estimate a

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.


Throck/Urock value between 2.1 and 3.6 for the ‘first’ population of zircon.

This value is within the limits proposed by Zartman and Richardson

(2005), between ~4 and ~2, for depleted asthenosphere over the last

2.5 Ga. This first population of zircon crystals could be genetically linked

to the kimberlite. The second population is characterized by high concen-

trations of REE (up to 123 ppm), high Th (more than 45 ppm), and high U

(more than 100 ppm). This population is likely derived from an enriched

source.

These two populations of primary zircon were produced by at least

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

typically found in kimberlites (Heaman et al., 2006), and the second is

close to those reported from xenoliths of glimmerite (Rudnick et al.,

1998) or MARID (Kinny and Dawson, 1992; Hamilton et al., 1998).

Both populations of zircon crystals are corroded and overgrown by a

rim of baddeleyite. The presence of such a rim suggests a desilication

reaction as a result of the interaction of zircon with carbonate in the

kimberlitic magma, for example:

ZrSiO4 þ ðCa;MgÞCO3→ZrO2 þ ðCa;MgÞsilicate þ CO2

or as a result of a reaction with other minerals in a carbonated kimberlitic

melt, as has been noted in other kimberlites (Haggerty, 1991; Dawson

et al., 2001; Page et al., 2007). In addition, the crystals were fragmented

after the development of the baddeleyite rim because the broken

surfaces are never replaced by baddeleyite (Fig. 2A, B). Therefore, these

crystals may have been replaced along the borders and then fractured

during eruption or later, such as during treatment in the gridding mill.

6.2. Age data

Two interpretations are possible to explain the presence of the

SHRIMP U–Pb zircon ages obtained in this study. The first interpretation

is that they reflect a period of zircon growth at 117.9±0.7 Ma, which

would represent a maximum eruption age for the Catoca kimberlite.

The second interpretation is that the zircon crystals partially retain an

older, inherited component that was incompletely reset by diffusive

Pb-loss prior to eruption at 117.9±0.7 Ma. Did the zircon form prior

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

(Cherniak and Watson, 2000; Heaman et al., 2006; and references

Table 4

Summary of SHRIMP data for zircon from the Catoca and Tchiuzo kimberlites.

Kimberlite Crystal

no.

Spot name U

(ppm)

Th

(ppm)

Th/U 206Pba

(ppm)

f(206)204% Total238U/206Pb

204Pb corrected ratios 207 corrected

Age (Ma)206Pb/238U

207Pb/206Pb 207/aPb/235U 206aPb/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.


therein). Thus, the zircon thus likely records the time when it was

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

in kimberlitic rocks (Mezger and Krogstad, 1997; Belousova et al.,

2001; Cherniak and Watson, 2003) have suggested that a complete

resetting does not invariably take place. Specifically, according to

Belousova et al. (2001), zircon crystals may retain radiogenic Pb at

lithospheric mantle temperatures between 600 and 1200 °C.

The estimated geotherm for the Catoca kimberlite (Robles-Cruz et al.,

unpublished results), calculated from data on garnet peridotite xenoliths

generated using the Nimis and Taylor (2000) geothermobarometer, gives

a temperature between 900 and 1200 °C at 40–55 kbar (160–200 km, in

the subcratonic lithospheric mantle). At this range of temperatures, the

calculated volume diffusion of Pb in zircon using the Arrhenius relation

(Cherniak and Watson, 2000) is between 3.23×10−26 at 900 °C and

3.14×10−21 m2/s−1 at 1200 °C. Calculations of Pb diffusion in zircon

(DPb) give:

DPb ¼ 1 � 10−1

exp −550 kJ mol−1

=RT� �

m2

s−1

;

R ¼ 8:314472 � 10−3

gas constantð Þ

DPb ¼ 3:23 � 10−26

m2=s

−1; at T ¼ 1173 K 900 BCð Þ

DPb ¼ 3:14 � 10−21

m2=s

−1; at T ¼ 1473 K 1200 BCð Þ:

This Pb diffusion is not significant and precludes the second

interpretation.

Other studies (Schärer et al., 1997; Corfu et al., 2003) suggest that

zircon keeps a record (partial or complete) of one or more thermal

events that it has experienced. Thus each zircon crystal is telling an

“individual” history and the measured U–Pb zircon ages, together

with the REE concentrations, provide insight into different episodes

of crystallization.

6.3. Geotectonic implications

This new interpretation of a maximum age for the kimberlitic erup-

tion at 118±1 Ma (Aptian age) is consistent with the regional tectonos-

tratigraphy of northeastern Angola (Fig. 7). The Catoca kimberlite was

expected to be younger than the carbonatites and alkaline rocks found

in the Lucapa structure. These rocks yielded K–Ar and Rb–Sr ages

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

the Alto Cuilo kimberlites (Eley et al., 2008; and references therein),

which are slightly older than the Calonda Formation and contain eroded

fragments of diamondiferous kimberlite.

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.


Cretaceous kimberlitic events of similar age have also 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 cra-

tons probably were the focus of thermal perturbations and injection of

melt. The Canastra 01 kimberlite in Brazil, located at the border of the

São Francisco craton (Da Costa, 2008), yielded an age of 120±10 Ma

using K/Ar in phlogopite (Chaves et al., 2008; and references therein).

It is associated with a NE–SW general tectonic trend and considered

to be related to lithospheric heating that took place before rifting

(Fleischer, 1998; Read et al., 2004), at a time when the presence of the

Tristan da Cunha mantle plume (133 Ma) would have exerted tectonic

control (Wilson, 1992). In the western part of the Kaapvaal craton, in

South Africa, kimberlites with ages of ca. 120 Ma (Jelsma et al., 2009;

and references therein) are associated with NE–SW preferential tecton-

ic orientation. In the Congo–Kasai craton, the reported ages of the earli-

est episodes of kimberlitic magmatism are between 116 and 70 Ma

(Batumike et al., 2009; and references therein), where NE–SW and E–

W general tectonic trends have been identified. The youngest kimberli-

tic magmatic episode reported, the Mbuji-Mayi kimberlites (70 Ma,

Schärer et al., 1997), located in the east Kasai province in the Democrat-

ic Republic of Congo, have an E–W trend. The implication is that there

was a change in the tectonic direction with which these kimberlites

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

with the two populations of zircon also identified based on

REE abundances. These populations are characterized either

by zircon crystals originating from a depleted mantle source

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.

(4) 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.

(5) The diamondiferous Catoca kimberlite seems to be tectonically

related to other Early Cretaceous kimberlites confined in NE–

SW lineaments in southwestern and southern Africa. This is

consistent with an incipient rifting stage previously proposed

by Jelsma et al. (2009) between 135 and 115 Ma. This under-

standing has important implications for diamond exploration.

The documentation concerning the maximum age of eruption

of the Catoca kimberlite during the Aptian provides precise

data on the age of a diamond-bearing kimberlite pulse in

Angola and should act as an important guide for diamond

exploration.

Acknowledgments

We acknowledge the great contribution of Dr. Bill Davis to the

analytical work. We thank Dr. Anthi Liati and the two more anony-

mous reviewers, as well as the editor, Dr. Klaus Mezger, for their

excellent revision of this manuscript and their valuable comments.

Fig. 5. Tera–Wasserburg (T–W) diagrams with data obtained from zircon crystals from the

Catoca and Tchiuzo kimberlites. Ellipses in gray are those that have a high proportion of204Pb. (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,

intersecting at 120.4±2.8 Ma.


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. Geochemicalcharacteristics of Cretaceous carbonatites from Angola. Journal of African EarthSciences 29 (4), 735–759.

Ballard, J.R., Palin, J.M., Campbell, I.H., 2002. Relative oxidation states of magmasinferred from Ce(IV)/Ce(III) in zircon: application to porphyry copper deposits ofnorthern Chile. Contributions to Mineralogy and Petrology 144, 347–364.

Bardet, M.G., Vachette, M., 1966. Détermination d'âges de kimberlites de l'OuestAfricain et essai d'interprétation des datations des diverses venues diamantifèresdans le monde. Bureau de Recherches Géologiques et Minières, DépartementMinéralogie, 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-isotopeanalyses of zircon from the Kundelungu Kimberlites, D.R. Congo: implications forcrustal evolution. Precambrian Research 156 (3–4), 195–225.

Batumike, J.M., Griffin, W.L., O'Reilly, S.Y., Belousova, E.A., Pawlitschek, M., 2009. Crustalevolution in the central Congo–Kasai Craton, Luebo, D.R. Congo: insights from zirconU–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. MineralogicalMagazine 62A (3), 355–366.

Belousova, E.A., Griffin, W.L., Shee, S.R., Jackson, S.E., O'Reilly, S.Y., 2001. Two agepopulations of zircons from the Timber Creek kimberlites, Northern Territory,as determined by Laser-ablation ICP-MS analysis. Australian Journal of EarthSciences 48 (5), 757–766.

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.

Fig. 7. Summary of geochronological results (SHRIMP) for the Catoca kimberlite eruption.


Belousova, E.A., Griffin, W.L., O'Reilly, S.Y., Fisher, N.I., 2002. Igneous zircon: trace ele-ment composition as an indicator of source rock type. Contributions to Mineralogyand Petrology 143 (5), 602–622.

Blundy, J., Wood, B., 2003. Mineral-melt partitioning of Uranium, Thorium and theirdaughters. In: Bourdon, B., Ribe, N.M., Stracke, A., Saal, A.E., Turner, S.P. (Eds.),Uranium-Series Geochemistry. Rev. Mineral. Geochem., vol. 52, pp. 59–123.

Chaves, M.L.S.C., Benitez, L., Brandão, P.R.G., Girodo, A.C., 2008. Kimberlito Canastra-1(São Roque de Minas, MG): geologia, mineralogia e reservas diamantíferas. REM,Revista Escola de Minas 61 (3), 57–364.

Cherniak, D.J., Watson, E.B., 2000. Pb diffusion in zircon. Chemical Geology 172, 5–24.Cherniak, D.J., Watson, E.B., 2003. Diffusion in zircon. In: Hanchar, J.M., Hoskin, P.W.O.

(Eds.), Zircon. Rev. Mineral. Geochem., vol. 53, pp. 113–143.Corfu, F., Hanchar, J.M., Hoskin, P.W.O., Kinny, P., 2003. Atlas of zircon textures. In: Hanchar,

J.M., Hoskin, P.W.O. (Eds.), Zircon. Rev. Mineral. Geochem., vol. 53, pp. 469–500.Da Costa, G.V., 2008. Química mineral e geotermobarometria de xenólitos mantélicos

do kimberlito Canastra-01. Master thesis (Dissertação de Mestrado no. 239).Instituto de Geociências, Universidade de Brasília.

Davis, G.L., 1977. The ages and uranium contents of zircon from kimberlites and asso-ciated rocks. Extended Abstr., 2nd Int. Kimberlite Conf., Santa Fe, New Mexico.

Dawson, J.B., Hill, P.G., Kinny, P.D., 2001. Mineral chemistry of a zircon-bearing, composite,veined and metasomatised upper-mantle peridotite xenolith from kimberlite.Contributions to Mineralogy and Petrology 140 (6), 720–733.

De Araujo, A.G., Perevalov, O.V., 1998. Carta de recursos minerais — mineral resourcesmap. Ministério Geol. e Minas. Inst. Geol. Angola, Luanda.

De Araujo, A.G., Perevalov, O.V., Jukov, R.A., 1988. Carta geológica de Angola - GeologicalMap 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 Precambrianin western Angola: links with Brazil. Journal of African Earth Sciences 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 Geologyand Geophysics 48 (4), 323–336.

Eley, R., Grütter, H., Louw, A., Tunguno, C., Twidale, J., 2008. Exploration Geology of theLuxinga kimberlite Cluster (Angola) with evidence supporting the presence ofkimberlite lava. Extended Abstr. No. 9IKC-A-00166, 9th Int. Kimberlite Conf., Frankfurt.

Fleischer, R., 1998. A rift model for the sedimentary diamond deposits of Brazil.Mineralium Deposita 33 (3), 238–254.

Griffin, W.L., Pearson, N.J., Belousova, E., Jackson, S.E., van Achterbergh, E., O'Reilly, S.Y., Shee,S.R., 2000. The Hf isotope composition of cratonic mantle: LAM-MC-ICPMS analysis ofzircon megacrysts in kimberlites. Geochimica et Cosmochimica Acta 64 (1), 133–147.

Griffin, W.L., Powell, W.J., Pearson, N.J., O'Reilly, S.Y., 2008. GLITTER: data reductionsoftware for Laser Ablation ICP-MS. In: Sylvester, P. (Ed.), Laser Ablation ICP-MSin the Earth Sciences: Current Practices and Outstanding Issues, Short Course Ser:Mineral. Assoc. Can., Québec, vol. 40, pp. 308–311.

Grütter, H.S., Gurney, J.J., Menzies, A.H., Winter, F., 2004. An updated classification formantle-derived garnet, for use by diamond explorers. Lithos 77, 841–857.

Guiraud, R., Bosworth, W., Thierry, J., Delplanque, A., 2005. Phanerozoic geologicalevolution of Northern and Central Africa: an overview. Journal of African EarthSciences 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 Benguelaand oblique-to-orthogonal-rifted Kwanza basins. Marine and Petroleum Geol-ogy 27, 1040–1068.

Haggerty, S.E., 1991. Oxide mineralogy of the upper mantle. In: Lindsley, D.H. (Ed.), OxideMinerals: Petrologic and Magnetic Significance: Rev. Mineral., vol. 25, pp. 355–416.Chantilly.

Haggerty, S.E., Raber, E., Naeser, C.W., 1983. Fission track dating of kimberlitic zircons.Earth and Planetary Science Letters 63 (1), 41–50.

Hamilton, M.A., Pearson, D.G., Stern, R.A., Boyd, F.R., 1998. Constraints on MARIDpetrogenesis: SHRIMP II U–Pb zircon evidence for pre-eruption metasomatism atKampfersdam. Extended Abstr., 7th International Kimberlite Conference, CapeTown, pp. 296–298.

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, T., Pankhurst, R.J. (Eds.), Magmatism and the Causes ofContinental Break-up: Geol. Soc. Spec. Publ. No. 68, London, pp. 221–240.

Hawkesworth, C., Kelley, S., Turner, S., Le Roex, A., Storey, B., 1999. Mantle processesduring Gondwana break-up and dispersal. Journal of African Earth Sciences 28 (1),239–261.

Heaman, L.M., Creaser, R.A., Cookenboo, H.O., Chacko, T., 2006. Multi-stage modification ofthe northern Slave mantle lithosphere: evidence from zircon- and diamond-bearingeclogite xenoliths entrained in Jericho Kimberlite, Canada. Journal of Petrology 47(4), 821–858.

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.Geochem., vol. 53, pp. 27–62.

Jackson, S., 2008. Calibration strategies for elemental analysis by LA-ICP-MS. In:Sylvester, P. (Ed.), Laser Ablation ICP-MS in the Earth Sciences: Current Practicesand Outstanding Issues, Short Course Ser: Mineral. Assoc. Can., Québec, vol. 40,pp. 169–188.

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 relatedrocks of Southern Africa. South African Journal of Geology 107, 301–324.

Jelsma, H., Barnett, W., Richards, S., Lister, G., 2009. Tectonic setting of kimberlites. Lithos112S, 55–165.

Jelsma, H., 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 casestudy of exploration findings. Extended Abstr. No. 10IKC-42, 10th InternationalKimberlite Conference, Bangalore, pp. 1–5.

Jochum, K.P., Stoll, B., 2008. Reference materials for elemental and isotopic analyses byLA-(MC)-ICP-MS: successes and outstanding needs. In: Sylvester, P. (Ed.), LaserAblation ICP-MS in the Earth Sciences: Current Practices and Outstanding Issues,Short Course Ser: Mineral. Assoc. Can., Québec, vol. 40, pp. 147–168.

Kinny, P.D., Dawson, J.B., 1992. A mantle metasomatic injection event linked to lateCretaceous kimberlite magmatism. Nature 360, 726–728.

Liati, A., Leander, F., Dieter, G., Fanning, C.M., 2004. The timing of mantle and crustalevents in South Namibia, as defined by SHRIMP-dating of zircon domains from agarnet peridotite xenolith of the Gibeon Kimberlite Province. Journal of AfricanEarth Sciences 39 (3–5), 147–157.

Longerich, H.P., Jackson, S.E., Günther, D., 1996. Laser ablation inductively coupled plas-ma mass spectrometric transient signal data acquisition and analyte concentrationcalculation. Journal of Analytical Atomic Spectrometry 11, 899–904.

Mezger, K., Krogstad, E.J., 1997. Interpretation of discordant U–Pb zircon ages: an evaluation.Journal of Metamorphic Geology 15, 127–140.

Moore, R.O., Griffin, W.L., Gurney, J.J., Ryan, C.G., Cousens, D.R., Sic, S.H., Suter, G.F.,1992. Trace element geochemistry of ilmenite megacrysts from the Monasterykimberlite, South Africa. Lithos 29, 1–18.

Nimis, P., Taylor, W.R., 2000. Single clinopyroxene thermobarometry for garnet perido-tites. Part I. Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in-Cpx thermometer. Contributions to Mineralogy and Petrology 139, 514–554.

Page, F.Z., Fu, B., Kita, N.T., Fournelle, J., Spicuzza, M.J., Schulze, D.J., Viljoen, F., Basei,M.A.S., Valley, J.W., 2007. Zircons from kimberlite: new insights from oxygen iso-topes, trace elements, and Ti in zircon thermometry. Geochimica et CosmochimicaActa 71 (15), 3887–3903.

Pereira, E., Rodrigues, J., Reis, B., 2003. Synopsis of Lunda geology, NE Angola: implications fordiamond exploration. Comunicações do Instituto Geológico e Mineiro 90, 189–212.

Pivin, M., Féménias, O., Demaiffe, D., 2009. Metasomatic mantle origin for Mbuji-Mayiand Kundelungu garnet and clinopyroxene megacrysts (Democratic Republic ofCongo). Lithos 112S, 951–960.

Read, G., Grütter, H., Winter, S., Luckman, N., Gaunt, F., Thomsen, F., 2004. Stratigraphicrelations, kimberlite emplacement and lithospheric thermal evolution, QuiricóBasin, Minas Gerais State, Brazil. Lithos 77, 803–818.

Reis, B., 1972. Preliminary note on the distribution and tectonic control of kimberlitesin Angola. Abstr., 24th Int. Geol. Congr., Rep. Sess., 4, pp. 276–281. Montreal.

Robles-Cruz, S.E., Watangua, M., Isidoro, L., 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 112S, 966–975.

Robles-Cruz, S.E., Melgarejo, J.C., Escayola, M., Galí, S., unpublished results. Compara-tive compositions of xenocrysts of garnet, clinopyroxene, and ilmenite from dia-mondiferous and barren kimberlites from northeastern Angola. Oral presentation.Peralk-Carb 2011 workshop on peralkaline rocks and carbonatites, 16–18 June.Tübinguen, Germany.

Rudnick, R.L., Ireland, T.R., Gehrels, G., Irving, A.J., Chesley, J.T., Hanchar, J.M., 1998.Dating mantle metasomatism: U–Pb geochronology of zircons in cratonic mantlexenoliths from Montana and Tanzania. Extended Abstr., 7th International KimberliteConference, Cape Town, pp. 754–756.

Schärer, U., Corfu, F., Demaiffe, D., 1997. U–Pb and Lu–Hf isotopes in baddeleyite andzircon megacrysts from the Mbuji-Mayi kimberlite: constraints on the subconti-nental mantle. Chemical Geology 143, 1–16.

Stern, R.A., Amelin, Y., 2003. Assessment of errors in SIMS zircon U–Pb geochronologyusing a natural zircon standard and NIST SRM 610 glass. Chemical Geology 197,111–146.

Watson, E.B., Wark, D.A., Thomas, J.B., 2006. Crystallization thermometers for zirconand rutile. Contributions to Mineralogy and Petrology 151, 413–433.

White, S.H., de Boorder, H., Smith, C.B., 1995. Structural controls of kimberlite andlamproite emplacement. Journal of Geochemical Exploration 53 (1–3), 245–264.

Whitehouse, M.J., Platt, J.P., 2003. Dating high-grade metamorphism-constraints fromrare-earth elements in zircon and garnet. Contributions to Mineralogy and Petrology145, 61–74.

Williams, I.S., 1998. U–Th–Pb geochronology by ion microprobe. In: McKibben, M.A.,Shanks III, W.C., Ridley, W.I. (Eds.), Applications of Microanalytical Techniques toUnderstanding Mineralizing Processes: Reviews in Econ. Geol., 7, pp. 1–35.

Wilson, M., 1992. Magmatism and continental rifting during the opening of the SouthAtlantic ocean: a consequence of Lower Cretaceous super-plume activity? In:Storey, B.C., Alabaster, T., Pankhurst, R.J. (Eds.), Magmatism and the Causes ofContinental Break-up: Geol. Soc. Spec. Publ. No. 68., London, pp. 241–255.

Zartman, R.E., Richardson, S.H., 2005. Evidence from kimberlitic zircon for a decreasingmantle Th/U since the Archean. Chemical Geology 220, 263–283.

Web references

http://georem.mpch-mainz.gwdg.de/, data downloaded February 24, 2010. GeoReM isa Max Planck Institute database for reference materials of geological and environ-mental interest, such as rock powders, synthetic, and natural glasses as well asmineral, isotopic, biological, river water, and seawater reference materials. It con-tains published analytical data and compilation values (major and trace elementconcentrations, radiogenic and stable isotope ratios). It also contains all importantmetadata about the analytical values such as uncertainty, analytical method andlaboratory. Sample information and references are also included, as well as morethan 2,400 reference materials, 25,400 analyses from more than 5,000 papers,and preferred analytical values (State: 07/01/2011). It complements the threeearthchem databases GEOROC, NAVDAT, PETDB.



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.

Minerals 2012, 2, 1-x manuscripts; doi:10.3390/min20x000x 1

2

minerals3

ISSN 2075-163X 4

www.mdpi.com/journal/minerals/ 5

Article 6

Major- and Trace-Element Compositions of Indicator Minerals 7

that Occur as Macro- and Megacrysts, and of Xenoliths, from 8

Kimberlites in Northeastern Angola 9

Sandra E. Robles-Cruz 1,*, Joan Carles Melgarejo 1, Salvador Galí 1 and Monica Escayola 2 10

1 Department de Cristal·lografia, Mineralogia i Dipòsits Minerals, Universitat de Barcelona, 11

Barcelona 08028, Spain; E-Mails: [email protected]; [email protected] 12 2 CONICET-IDEAN Instituto de Estudios Andinos, Laboratorio de Tectónica Andina, Universidad 13

de Buenos Aires, Buenos Aires C1033AAJ, Argentina; E-Mail: [email protected] 14

* Corresponding author. E-Mail addresses: [email protected], [email protected]; 15

Tel.: +506-883-939-81; Fax: +506-224-244-11. 16

Received: / Accepted: / Published: 17

18

Abstract: In this study, we compare the major- and trace-element compositions of olivine, 19

garnet, and clinopyroxene that occur as single crystals (142 grains), with those derived 20

from xenoliths (51 samples) from six kimberlites in the Lucapa area, northeastern Angola: 21

Tchiuzo, Anomaly 116, Catoca, Alto Cuilo-4, Alto Cuilo-63, and Cucumbi-79. The 22

samples were analyzed using electron probe microanalysis (EPMA) and laser-ablation 23

inductively coupled plasma-mass spectrometry (LA-ICP-MS). The results suggest different 24

paragenetic associations for these kimberlites in the Lucapa area. Compositional overlap in 25

some of the macrocryst and mantle xenolith samples indicates a xenocrystic origin for 26

some of those macrocrysts. The presence of mantle xenocrysts suggests the possibility of 27

finding diamond. Geothermobarometric calculations were carried out using EPMA data 28

from xenoliths by applying the program PTEXL.XLT. Additional well calibrated single-29

clinopyroxene thermobarometric calculations were also applied. Results indicate the 30

underlying mantle experienced different equilibration conditions. Subsequent metasomatic 31

enrichment events also support a hypothesis of different sources for the kimberlites. These 32

findings contribute to a better understanding of the petrogenetic evolution of the 33

kimberlites in northeastern Angola and have important implications for diamond 34

exploration. 35

36

OPEN ACCESS

Minerals 2012, 2

2

Keywords: Angola; kimberlite; olivine; garnet; clinopyroxene; diamond; 1

thermobarometry; mantle xenoliths; REE; Sm/Nd isotopes 2

3

1. Introduction 4

In Angola, more than 700 occurrences of kimberlite are distributed on a trend northeast to 5

southwest from Lunda Sul and Lunda Norte, up through Huambo, Benguela, and Huila provinces. 6

Diamond was first reported in Angola in 1590 [1]. In 1952, the first kimberlite in Angola, Camafuca-7

Kamazambo, was discovered [2]. Since then, and especially after thirty years of civil war, Angola has 8

become known as an important diamond producer. The Catoca kimberlite in Lunda Sul province, the 9

first kimberlite mine in Angola, was for a long time the only kimberlite under production. In 2007, two 10

more pipes came on line, and subsequently other exploration projects have started in the country. By 11

2010, the Catoca kimberlite had produced over 8.36 million carats, valued at US$ 976 million [3]. 12

There are several diamondiferous and barren kimberlites in the northeastern area. However, there 13

are no detailed studies that allow a full understanding of their relationship with the underlying mantle, 14

or their spatial distribution. Six kimberlites from northeastern Angola will be considered for this study. 15

They range from poor to very high level of diamond production: Alto Cuilo 4 (AC4), Cucumbi-79 16

(CU79), Alto Cuilo-63 (AC63), Anomaly 1116 (An116), Tchiuzo (TZ) and Catoca (CA). These 17

kimberlites are located in the Lucapa area and emplaced in Archean metamorphic rocks, in the Kassai-18

Congo Craton. 19

Kimberlites are volatile-rich potassic ultrabasic rocks usually with an inequigranular texture as a 20

result of the presence of crystals, compound clasts, and interstitial matrix [4,5]. Crystals can be: 21

megacrysts, crystals of more than 1.0 cm at maximum dimension (MD); macrocrysts, crystals between 22

0.5 and 10 mm at MD; or microcrysts, crystals less than 0.5 mm at MD. Xenoliths are rare; in these 23

kimberlites, most of them consist of metasomatized peridotite and phlogopite-rich suites. 24

The aim of this paper is to establish the major- and trace-element compositions of indicator 25

minerals: olivine, garnet, and clinopyroxene, that occur as single crystals with those derived from 26

xenoliths from six kimberlites from the Lucapa area, northeastern Angola: TZ, An116, CA, AC4, 27

AC63, and CU79. This area is where most of the diamondiferous kimberlites identified so far are 28

concentrated. We estimate reliable conditions of pressure and temperature for selected samples. Also 29

we have measured Sm/Nd isotopes from the rare xenoliths to better understand their petrogenesis. 30

2. Geological Setting 31

The exposed rocks in Angola range from Archean age to Recent (Figure 1). This geological history 32

may be divided in three main stages [6,7]: (1) the Archean orogeny is recorded by the Central Shield, 33

Cuango Shield and Lunda Shield, mainly composed of gabbro, norite and charnockitic complexes, 34

which constitute the Angolan basement. (2) There are three main Proterozoic cycles (Eburnean in the 35

Paleoproterozoic, Kibaran in the Mesoproterozoic, and Pan-African in the Neoproterozoic), of which 36

the Eburnean is the most important and characterized by complex volcanosedimentary rocks, gneisses, 37

migmatites, granites and syenites. This regional Paleoproterozoic event was followed by the Kibaran 38

Minerals 2012, 2

3

cycle, which was related to extensional events along the border of the Congo craton where later clastic-1

carbonatic sequences and local basic magmatism took place. The Pan-African orogeny was associated 2

with the development of Gondwana and led to the generation of fold belts and granitic intrusions. (3) 3

Phanerozoic sequences covered the older rocks as events associated with the formation of Pangea and 4

the consecutive break-up of Gondwana which contributed to the formation of rift basins associated 5

with deep fault systems and later intraplate magmatism (alkaline, carbonatitic, kimberlitic) and 6

marginal basins. 7

Figure 1. Geological setting of the northeastern Angola kimberlites of this study (after 8

Perevalov et al. [8], Giraud et al. [7], Egorov et al. [9]). 9

10

11

The Lower Cretaceous regional extension caused the development of deep faults and grabens with 12

NE-SW and NW-SE trends. The Lucapa structure, a deep-seated fault system, corresponds to the first 13

group, and the northeastern part is the focus of most of the diamondiferous kimberlites in Angola, 14

whereas the southwestern zone comprises important outcrops of undersaturated alkaline rocks and 15

carbonatites [2]. Other minor kimberlite fields are found in southwestern Angola [9]. This geological 16

configuration sets a tectonic control on the presence of some of the kimberlites in northeastern Angola. 17

The emplacement of the Catoca kimberlite (middle of Aptian and Albian) has been recently correlated 18

with these regional tectonic events [10]. Kimberlite emplacement ages in the Alto Cuilo range between 19

145.0±4.0 (206Pb/238U for perovskite) and 113.0±0.8 Ma (206Pb/238U for zircon) [11]. Synsedimentary 20

continental sediments associated with the filling of the Lucapa structure (Calonda Formation) can also 21

Minerals 2012, 2

4

contain diamond crystals in paleoplacers; alluvial diamond is found in placers associated with rivers 1

draining these diamondiferous areas. 2

3

3. Analytical Techniques 4

Six hundred and fifty samples were studied with an optical petrographic microscope. Then 5

representative samples of mantle xenoliths (51 samples), mega- (3 grains), macro- (116 grains), and 6

microcrysts (23 grains) of olivine, garnet, and clinopyroxene were examined by back-scattered 7

electron (BSE) images using SEM-ESEM with EDS microanalysis and electron-probe microanalysis 8

(EPMA). More than 800 microprobe analyses were carried out to obtain the mineral chemistry of 9

major elements. The major elements were analyzed using a JXA JEOL-8900L EPMA at the 10

Department of Earth and Planetary Sciences, McGill University (Montreal, Quebec), using the ZAF 11

correction method. Acceleration voltage was 20 kV, beam current 20 nA, and beam diameter 5 μm. 12

The counting time for most elements was 20 s on peaks and 20 s on the background. Standardized 13

natural and synthetic minerals were used for calibration. 14

Then 14 representative samples of xenoliths and 25 macrocrysts of garnet, clinopyroxene, and 15

olivine were selected to perform trace-element analyses using laser-ablation inductively coupled 16

plasma mass spectrometry (LA-ICP-MS) at the Geological Survey of Canada, Ottawa. The data were 17

acquired with a Photon-Machines Analyte 193nm Excimer laser ablation in combination with an 18

Agilent 7500cx quadrupole ICP-MS, a powerful technique for reliable solid analysis of samples. Data 19

reduction was performed with the GLITTER 4.4.2 software. The primary calibration standard was the 20

synthetic glass standard of the 610 series (NIST SRM 610) of the National Institute of Standards and 21

Technology, using SiO2 for internal standardization. The GSE-1G (a synthetic reference glass with 22

basaltic major-element composition and trace elements abundance of ca. 500 μg/g) was used as a 23

secondary standard. 24

Finally, accurate high-precision Sm/Nd isotopic compositions of eight samples of xenoliths (1.5 to 25

7 cm in diameter) from kimberlites CA, CU79, and CU80 (for comparison purposes) were carried out 26

at the Pacific Centre for Isotopic and Geochemical Research (PCIGR) at the University of British 27

Columbia, using a Thermo Finnigan Triton thermo-ionization mass spectrometer (TIMS). We used the 28

analytical procedures for sample dissolution, ion exchange, and leaching described by Weis et al. [12]. 29

The normalization procedure has been applied to the Nd isotopic ratios using La Jolla Nd as the 30

reference material (measured ratio normalized to La Jolla 143Nd/144Nd = 0.511858). 31

4. Morphology of the Kimberlite Pipes Studied in Northeastern Angola 32

The CA, TZ, An116, AC4, AC63, and CU79 kimberlites are located within the Kasai craton (Figure 1). 33

The pipes generally appear in clusters along a network of local fractures. They are of variable shape 34

and dimension, some very large up to 900 m in diameter (i.e., Catoca). Some pipes in this area are 35

diamondiferous, and do have an economic grade. Crater and diatreme facies can be recognized despite 36

the intense weathering affecting kimberlites up to a depth of 150-200 m. Because of the weathering the 37

number of fresh xenoliths and crystals is limited, and conditioned the selection of samples for this 38

study. Next to olivine, serpentine is the most abundant mineral in the studied kimberlites, followed by 39

Minerals 2012, 2

5

calcite in the CA, TZ, An116, and AC. Phlogopite is the second most abundant mineral in the Cucumbi 1

cluster. 2

3

4.1. TZ, An116, and CA kimberlites 4

The CA kimberlite has a circular shape on the surface. This pipe exhibits crater and diatreme facies. 5

For this study, we analyzed samples to a maximum of 609 m depth. A new exploration project taking 6

samples to a depth of 800 m started in 2010 [13]. Crater facies are found up to 270 m in depth, and are 7

composed of epiclastic sandstones, coarse debris rimming the crater, and a higher concentration of 8

volcaniclastic material at a lower depth of the crater [14]. Quartz (without a reaction rim) and K-9

feldspar are the most abundant minerals in the volcaniclastic rocks (VR). Altered crystals of garnet, 10

diopside, and rare ilmenite may be present. The diatreme facies rocks, which are more than 700 m 11

thickness, are composed of volcaniclastic kimberlite (VK) and volcaniclastic kimberlite breccias 12

(VKB), with tuffisitic kimberlite (TK) in the deepest zone. There are abundant olivine macrocrysts 13

completely replaced by serpentine and secondary carbonates. Macrocrysts of clinopyroxene, garnet, 14

ilmenite (Fe-rich, Mg-rich, and Mn-rich ilmenite), chromite, magnetite, zircon, phlogopite, hematite, 15

and amphibole also are present. Orthopyroxene has been identified, but is completely replaced by 16

bastite lizardite. The diatreme facies rocks are strongly altered all along the profile. Abundant 17

xenoliths derived from the host rocks are present, (e.g., gneiss, amphibolite, and granite); some 18

carbonatite xenoliths which could be present in the crust or beneath the kimberlite volcano; mantle-19

derived xenoliths (i.e., garnet lherzolites, phlogopite-garnet wehrlite, and very rare eclogite) are sparse 20

and have been intensively altered. The groundmass contains lizardite, smectite, apatite, calcite, 21

ilmenite and chromite. Titanite, zirconolite, baddeleyite, barite, dolomite, witherite, barytocalcite, 22

strontianite, sulfides, and minor minerals are also widespread in the matrix. 23

The TZ kimberlite is characterized by the presence of crater and diatreme facies. The first 30 m 24

contains VR composed of quartz, hematite, K-feldspar, plagioclase, amphibole, and spinel. Diatreme-25

facies rocks are composed of VK and VKB with macrocrysts of replaced olivine and orthopyroxene, 26

garnet, clinopyroxene, spinel (in some cases in a “necklace” shape around pellets of serpentine), 27

apatite, ilmenite, amphibole, phlogopite (some of them with inclusions of ilmenite), and zircon in a 28

groundmass of lizardite, smectite, apatite, calcite, ilmenite and chromite. Xenoliths from amphibolites 29

are very common. Mantle xenoliths are very rare, i.e., garnet lherzolite and carbonatite; and usually 30

altered. This pipe was drilled up to 310 m in depth. 31

The An116 pipe is located on a magnetic anomaly close to the Catoca area, and samples were taken 32

down to a depth of 88 m. These samples describe VR, mainly PK, in which quartz and microcline are 33

the most abundant minerals, and “mafic” xenoliths of ilmenite-calcite-phlogopite are present in the 34

first 10 m. Samples from 10 to 88 m in depth returned VK facies composed of macrocrysts of ilmenite, 35

altered clinopyroxene, phlogopite, hematite, and plagioclase, mafic xenoliths (amphibolites) and 36

metasomatized mantle-derived xenoliths i.e., phlogopite-ilmenite-clinopyroxene suites (PIC suites), 37

altered metasomatic peridotites. 38 39

Minerals 2012, 2

6

4.2. The AC4 and AC63 kimberlites 1

The AC4 and AC63 pipes are also covered for the first 50 and 100 meters, respectively, by 2

sandstone, litharenite, and arkose from the Calonda Formation, Kalahari Group, and Quaternary 3

deposits, as has been observed in most of the kimberlites in the Alto Cuilo cluster [15]. These 4

kimberlites exhibit crater- and diatreme-facies rocks. The crater-facies rocks are mainly composed of 5

VR, mainly pyroclastic rocks (PR) and resedimented volcaniclastic kimberlites (RSVK). Macrocrysts 6

of clinopyroxene, hematite, mica, microcline, and ilmenite are present in the microlitic matrix, which 7

contains clinopyroxene, mica, microcline, ilmenite, and higher content of hematite. The diatreme 8

facies rocks contain VKB composed of macrocrysts of clinopyroxene, garnet, ilmenite with 9

homogeneous and symplectitic textures, phlogopite and rare xenoliths of phlogopite-garnet wehrlite 10

and altered phlogopite peridotites. Mantle xenoliths have not been found in the AC4 pipe. The 11

groundmass is composed of serpentine, calcite, and hematite. 12

4.3. The CU79 kimberlite 13

The CU79 kimberlite exhibits two facies: crater and diatreme. The first 50 m of the crater facies 14

contains VR with ferruginous cement, interbedded with layers of lapilli. The facies then changes to 15

TK, which is generally massive, poorly sorted, and clast-supported, and is present as far down as 200 16

m. The macro- and microcrysts are composed of anhedral olivine replaced by serpentine and smectite, 17

garnet, ilmenite, clinopyroxene, and phlogopite. Some of these crystals are enclosed in a pelletal 18

assemblage of serpentine, but all the crystals have a chaotic distribution in the matrix. The matrix is an 19

interclast groundmass of serpentine, microcrystals of phlogopite, less common chlorite, smectite and 20

calcite [16]. The ilmenite texture is usually either cumulus or homogeneous. Mg-rich ilmenite (9-13 21

wt.% MgO) is present as rounded mega- and macrocrysts, as part of xenoliths, and as inclusions in 22

phlogopite. In some cases, macrocrysts of ilmenite are partially replaced along the borders by 23

perovskite and spinel. Garnet and clinopyroxene are usually present as mega- and macrocrysts, and 24

rarely as part of xenoliths. This kimberlite rarely has mantle xenoliths i.e., garnet lherzolite, 25

phlogopite-garnet wehrlite, and relatively abundant phlogopite-rich (olivine poor or absent, without 26

garnet [17]) xenoliths. 27

5. Major-Element Composition 28

The kimberlites in this suite exhibit differences in their composition and abundance of olivine, 29

garnet, and clinopyroxene depending on the location and type of kimberlite. 30

5.1. Olivine 31

Olivine is abundant in the TZ, CA, and An1116 kimberlites, forming up to 65 % volume of the total 32

mineral components. However, fresh olivine has only been found in the CA kimberlite. The grains 33

generally occur as macrocrysts (most of them anhedral) and microcrysts (subhedral to anhedral), dark 34

and light green to pale greenish white, depending on the degree of alteration. Most of the grains 35

(approximately 95%) are replaced by lizardite, with other alteration minerals i.e., calcite, smectite, 36

chlorite, magnetite, and sulfides. Olivine macrocrysts from the CA kimberlite (Figure 2) occur as: (1) 37

Minerals 2012, 2

7

homogeneous crystals, (2) crystals rimmed by "iddingsite", and (3) zoned macrocrystals. The first 1

population (Table 1, supplementary file) has an average composition of Fo92 and extremely low 2

average values for CaO (0.01 wt.%), and MnO (0.11 wt.%). The second population also has an average 3

composition of Fo92, CaO (0.01 wt.%), and MnO (0.13 rim wt.%). The third population of crystals 4

exhibits zonation with an average composition of core Fo90, middle Fo87, and rim Fo85. The average 5

values varying from core to rim are: CaO (0.02-0.04 wt.%), MnO (0.13-0.19 wt.%), Cr2O3 (0.01-0.06 6

wt.%), TiO2 (0.01-0.03 wt.%), and NiO (0.36-0.13 wt.%). Olivine macrocryst compositions are very 7

similar to archetypal cratonic peridotites, which has a mean Mg# of 92.6 and indicates melt depletion 8

[18]. 9

Figure 2. CaO vs Mg# diagram for olivine macrocrysts from the CA kimberlite. Error bars 10

indicate standard deviation of CaO (wt.%). 11

12

5.2. Garnet as Macrocrysts and within Xenoliths 13

The CA and TZ kimberlite includes dark pink to dark orange anhedral macrocrysts garnet (G9 and 14

G10 after Grütter et al.[19], Figure 3). Some of the garnet macrocrysts are partially altered to chlorite, 15

or replaced by hematite and calcite along fractures. Garnet (Table 2, supplementary file) is also present 16

in eclogite (CA), garnet lherzolite with and without ilmenite (CA, TZ), and phlogopite-garnet wehrlite 17

with and without ilmenite (CA) xenoliths. Garnet macrocrysts from the CU79 kimberlite are mainly 18

pink to orange (G9). Few garnet grains are among the microcrysts. Garnet is also found in garnet 19

Minerals 2012, 2

8

lherzolite, garnet-phlogopite wehrlite xenoliths from CU79, and bimineralic associations of G9-1

diopside. The garnet from AC kimberlite is usually found as a microcryst (AC4) or in garnet-2

phlogopite wehrlite xenoliths (AC63). Garnet from CA, TZ, AC63, and CU79 fall in the mantle-3

derived field (Figure 4), whereas garnet that has been found so far in the AC4 pipe indicates a crust- 4

derived origin. 5

Figure 3. Cr2O3 vs CaO diagram for garnet with superimposed isobars according to the 6

P38 barometer calculation [19]. Graphite-diamond constraint (GDC). 7

8

9

Minerals 2012, 2

9

Figure 4. Ca# vs Mg# for garnet from the TZ, CA, AC63, AC4, and CU79 kimberlites. 1

2

5.3 Clinopyroxene Mega- and Macrocrysts and in Xenoliths 3

Three populations of clinopyroxene can be identified on the basis of an Al-Cr-Na plot (Figure 5): 1) 4

clinopyroxene in a lherzolitic association; 2) low-Cr clinopyroxene that plots between Al and Na 5

extremes; and 3) Low-Na clinopyroxene. Clinopyroxene in a lherzolitic association has been found in 6

all the kimberlites studied with slight variations in the Cr content between pipes (Table 3, 7

supplementary file). Cr-rich diopside is relatively common in the CA pipe, and shows the highest 8

concentrations in Cr. Low-Cr clinopyroxene was found in the CA, some in the TZ, and the CU79 9

pipes. Low-Na clinopyroxene is present in the CA pipe with a relatively high content of Cr, whereas 10

low-Na and low-Cr clinopyroxene is found in the TZ and the CU79 pipes. All the clinopyroxene falls 11

in the “on craton” field, except for a clinopyroxene inclusion in garnet from CU79 (Figure 6). 12 13

Minerals 2012, 2

10

Figure 5. Plot of Al-Cr-Na (apfu) for clinopyroxene from the kimberlites studied. * After 1

Morris et al. [20] 2

3 4

Minerals 2012, 2

11

Figure 6. Cr2O3 vs Al2O3 diagram for clinopyroxene from the An116 (magenta), TZ 1

(cyan), CA (red), AC63 (green), and CU79 (blue) kimberlites. Classification diagram after 2

Ramsey [21] 3

4

6. Geothermobarometry 5

Comparison and evaluation of results obtained from different pressure and temperature (PT) 6

combinations were carried out. The thermobarometer combination used is based on element-exchange 7

reactions between clinopyroxene and garnet that are believed to be in equilibrium (xenoliths), which 8

provides better evaluation of PT results than calculations based on single crystals. First, we selected 9

only the most representative fresh xenoliths from the CA, TZ, CU79, and AC63 kimberlites, then we 10

proceeded to calculate P and T on the basis of EPMA data and by obtaining the average composition 11

for each mineral from each xenolith (i.e., only one average value-point for each type of xenolith). We 12

used the program PTEXL.XLT, prepared by Dr. T. Stachel [22]. The well-calibrated single-13

clinopyroxene thermobaromether of Nimis and Taylor [23] also was applied, in order to compare 14

values of both T and P, which yielded reliable temperature estimates compared to the T-P results 15

Minerals 2012, 2

12

obtained from thermobarometry of xenoliths. We used the following calibration for the single-1

clinopyroxene thermobaromether: (1) clinopyroxene that falls in the garnet peridotite field “on-craton” 2

defined by Ramsay [21] in the diagram Cr2O3 versus Al2O3 (Figure 6), (2) clinopyroxene above the 3

field of Low-Al peridotite (Figure 7). 4

Figure 7. Al2O3 vs MgO diagram for discrimination of no Low-Al peridotite clinopyroxene 5

type for the geothermobarometric calculations based on of single crystals of clinopyroxene 6

(after Nimis [24]). 7

8 9

Minerals 2012, 2

13

Figure 8. Pressure and temperature values calculated for xenoliths using the equation 1

Nimis and Taylor [23] (NT2000): Garnet lherzolite (diamonds) and phlogopite garnet 2

wehrlite (triangles); and for single clinopyroxene macrocrysts (cross) from the TZ (purple), 3

CA (red), AC63 (green), and CU79 (blue) kimberlites. Paleogeotherm calculated using 4

FITPLOT program (blue line, with pink and purple lines representing the error envelope). 5

P-T data calculated from mantle xenoliths from the Bultfontein kimberlite (dark gray 6

shading) and paleogeotherm (dark gray dashed line); and P-T data calculated from mantle 7

xenoliths from the Finsch kimberlite (light gray shading) and paleogeotherm (light gray 8

point line) after Mather et al. [25]. 9

10

11

Calculated temperature and pressure from xenoliths is less scatter than T-P data calculated from 12

single crystals (Figure 8). However, most of the data fall within error of estimate. The calculated 13

northeastern Angola paleogeotherm fit a single value for the CA and the CU79 kimberlites. Only one 14

phlogopite-garnet lherzolite xenolith from the AC63 kimberlite was able to be used and plotted in the 15

same paleogetherm than the CA and CU79 kimberlites. The differences in T-P values between these 16

kimberlites may reflect the different way each kimberlite sampled the lithosphere. The lithospheric 17

thickness calculated from the northeastern Angola paleogeotherm yielded 192 km. A quantitative 18

comparison between Angola lithosphere and reference geotherms in southern Africa (Bultfontein and 19

Finsch kimberlites, after Mather et al. [25]), indicates a slightly cooler (steeper) paleogetherm for 20

Angola than the paleogetherms calculated from southern Africa. 21

Minerals 2012, 2

14

7. Trace-Element Chemistry 1

About one hundred sixty trace-element analyses were performed on garnet and clinopyroxene from 2

representative macrocrysts (about 80 grains) and mantle xenoliths (14 xenoliths) from the CA, TZ, and 3

CU79 kimberlites (Table 4, supplementary file). Three main different trends for garnet can be 4

identified in the Catoca kimberlite on the basis of chondrite-normalized Rare Earth Element patterns 5

(REEN) (Figure 9). (1) Garnet of eclogitic affinity with “normal” [26] REEN patterns, and slightly 6

enriched in Light Rare Earth Element (LREE). (2) Garnet from garnet lherzolite and phlogopite-garnet 7

wehrlite xenoliths can exhibit either “normal” REEN patterns or LREE-enriched patterns. Garnet from 8

phlogopite-garnet wehrlite exhibits the highest LREE-enrichment, with a maximum around the LREE-9

HREE limit and flat HREE. (3) Garnet macrocrysts with “normal” and “sinusoidal” REEN patterns 10

[26]. Garnet macrocrysts from the TZ pipe exhibit both “normal” and "sinusoidal" REEN patterns with 11

lower HREE abundances. In contrast, garnet from xenoliths (garnet-lherzolite and phlogopite-garnet 12

wehrlite) and macrocrysts from the CU79 kimberlite follows the “normal” REEN pattern but with 13

slightly depleted values than garnet from the CA pipe. 14

The data indicate that garnet lherzolite xenoliths found in the CA and CU79 kimberlites were under 15

different equilibration conditions and different degrees of metasomatism. The xenoliths from the CA 16

kimberlite may have been generated by refertilization of a previously depleted peridotite. The 17

xenoliths from the CU79 kimberlite might be the result of a depleted source but with a very limited 18

enrichment in LREE. 19

20

Minerals 2012, 2

15

Figure 9. Chondrite-normalized REE diagrams for representative xenoliths and 1

macrocrysts from the CA, TZ, and CU79 kimberlites. The gray zone represents the 2

“normal” garnet pattern according to McLean et al. [26]. 3

4

Minerals 2012, 2

16

8. Sm/Nd Isotope Results for Xenoliths Whole-Rock 1

These data are the first Sm-Nd isotope analyses carried out in xenoliths of kimberlites from Angola 2

(Table 5, supplementary file). Mantle xenoliths are potentially subject to infiltration and alteration of 3

isotopic signatures from the kimberlite [27]. Consequently, it is important to mention that Sm-Nd 4

model ages can reflect this mixing process. The mantle xenoliths from the CA kimberlite have a 5 143Nd/144Nd value between 0.511288 and 0.511681, whereas the Sm-Nd isotopes in xenoliths from the 6

CU79 kimberlite show higher 143Nd/144Nd values between 0.512274 and 0.512391, as well as a 7

narrower range of values. A single value of 0.512377 was obtained from the CU80 pipe, located 5.5 8

km SSE from the CU79 pipe, and can be used for comparison. Negative ƐNd values from xenoliths of 9

the CA kimberlite indicate an enriched mantle, whereas mantle-derived xenoliths from the CU79 pipe 10

show a slightly depleted mantle signature with positive ƐNd values as well as the sample from the 11

CU80 pipe. The Nd isotope evolution diagram (Figure 10) clearly shows different sources for the two 12

kimberlites (the CA and the CU79). 13

Figure 10. Diagram of ƐNd vs T (Ga) for xenoliths from the CA, CU79, and CU80 14

kimberlites. 15

16

9. Discussion and Conclusions 17

9.1 Discrimination among Kimberlites 18

The trans-lithosphere discontinuity of the Lucapa structure played a very important role by favoring 19

a thermal perturbation, melt production, and mantle upwelling [28,29], and in the evolution of the host 20

rocks. The integration of petrography, geochemistry, and geothermobarometric studies of the less 21

altered samples from six kimberlites in the northeasterm Angola suggests that these kimberlites 22

originated from different sources in spite of the fact that they are all located in the same tectonic 23

corridor. 24

The CA kimberlite has different compositional populations of olivine macrocrysts (homogeneous, 25

"iddingsite" rimmed, and zoning olivine) which suggests different sources for those crystals, where 26

they have likely been modified by subsequent crystallization. It is important to mention that 27

Minerals 2012, 2

17

heterogeneous crystallization is commonly found in volcanic rocks [30]. In the case of Catoca, Fe-Mg 1

zoning in olivine indicates a state of disequilibrium that reflects the physical and chemical conditions 2

that the mineral has experienced. This pipe is also characterized by the presence of pyrope and pyrope-3

almandine (G9 and G10 according to Grütter et al. [19,31]), a relative abundance of Cr-rich diopside, 4

and the presence of eclogite, garnet lherzolite, carbonatite, and phlogopite-garnet wehrlite xenoliths. 5

Orthopyroxene is less common than clinopyroxene in this pipe. Orthopyroxene is usually 6

serpentinized. The very low amounts of orthopyroxene may be explained by rapid dissolution in the 7

kimberlitic melt during transport [32]. Considering that clinopyroxene does not occur as a liquidus 8

phase in Group I kimberlites, kimberlitic magma is the mechanism that brings mantle xenoliths to the 9

surface. The clinopyroxene equilibrates at pressures greater than 30 kbar and the Cr:Na ratio in 10

clinopyroxene from kimberlites usually is around 1. The solubility of kosmochlor (NaCrSi2O6) in 11

diopside decreases with pressure and its limit of solubility may be at about 45 kbar [32], which 12

suggests that the join jadeite (NaAlSi2O6)-kosmochlor is strongly influenced by pressure between 30 13

and 45 kbar. Then there is a difference in the Al-Na component in clinopyroxene from the CA pipe and 14

the other studied kimberlites specially the CU79. 15

The samples from the An116, AC63 and AC4 pipes are so limited that conclusions regarding their 16

origin and evolution should be taken with caution. It is clear that Cr-rich diopside is present in the 17

An116 pipe, indicating a potential “deep” source. The garnet samples from the AC4 pipe suggest 18

shallower pressure conditions, less than 20 kbar. In contrast, garnet from the AC63 pipe shows ranges 19

of pressure between 20 and 43 kbar. The chemical composition of the AC4 garnet suggests crust-20

derived crystals, whereas garnet from xenoliths from the AC63 pipe suggests a mantle-derived source. 21

Samples from the TZ pipe have some similarities with those from the CA pipe, but garnet from the TZ 22

pipe is the only G9 type with different ranges of pressure (less than 43 kbar) and its clinopyroxene has 23

a lower Cr content. Some of the T-P values from Catoca are consistent with data previously published 24

by Aschepkov et al. [33] based on garnet and pyroxene xenocrysts. 25

9.2 The Underlying Mantle in the Northeastern Angola 26

We propose that the Catoca kimberlite was generated from a depleted source. A subsequent 27

metasomatic enrichment event (possibly more than one) incorporated incompatible LREE. Based on 28

normalization to chondrite of REE concentrations in garnet and clinopyroxene, we can track the 29

behavior of these elements in the kimberlite. If garnet equilibrates with clinopyroxene, the result is to 30

shift only LREE in garnet [34], as we observed in the "sinusoidal" pattern of garnet from phlogopite-31

garnet wehrlite xenoliths in the CA pipe. Garnet crystals from the TZ pipe show similar patterns as that 32

of the CA pipe (i.e., "normal" and "sinusoidal" REEN patterns). This also suggests a metasomatic 33

enrichment of a previously depleted source. 34

In contrast, garnet from phlogopite-garnet wehrlite from the CU79 kimberlite shows "normal" 35

REEN patterns. This suggests different degrees of enrichment, likely due to metasomatism. Garnet 36

lherzolites from both kimberlites seem to be derived from a depleted mantle source. 37

Different Sm-Nd TD model ages from the CA and the CU79 pipes may also suggest different 38

mantle sources or metasomatic events that modified the isotopic ratios in the mantle. TDM model ages 39

from the CA kimberlite could indicate different time events than for the CU79 kimberlite. The 40

Minerals 2012, 2

18

differences in diamond production from the CA, TZ, CU79, and AC63 kimberlite pipes may be the 1

result of different mantle sources and metasomatic events, as well as independent subsequent evolution 2

of each kimberlite. 3

We interpret different sources for both kimberlites (the CA and CU79 pipes). These two kimberlites 4

have heterogeneous mantle sources, the CA kimberlite is the more enriched of the two, possibly 5

because of multiple metasomatic events that could explain the "sinusoidal" REEN patterns in garnet. 6

Both kimberlitic mantle xenoliths also have different TDM (Model ages relative to CHUR): the CA 7

kimberlite xenoliths show Mesoproterozoic ages (1220 - 1250 Ma), whereas the CU79 kimberlite 8

xenoliths yield 450 to 390 Ma (Late Ordovician to Devonian) like the TDM from the CU80 kimberlite. 9

These model ages may be interpreted as the age of mantle generation or the age of a metasomatic 10

event that modified the isotopic ratios in the mantle. TDM model ages of 1.2 Ga or the xenoliths from 11

the CA kimberlite could be associated with the Kibaran orogeny, whereas the ages from the CU79 12

kimberlite could imply more juvenile Paleozoic components possibly related to the assembly of 13

Pangea. Based on these data, these two kimberlites with different diamond production grades have a 14

different pattern of evolution despite being from the same tectonic trend, the Lucapa structure. 15

The reactivation of old deep-seated faults during the Paleoproterozoic, the Permo-Triassic, the 16

Cretaceous and the Cenozoic [29] probably is an important factor in some of the different pulses of 17

diamondiferous kimberlites. Thus petrography, geochemistry and Sm-Nd isotopic data from these 18

kimberlites provide interesting tools to recognize possible diamondiferous kimberlites in the area. 19

Acknowledgements 20

This research is funded by the CGL2006-12973 and CGL2009-13758 BTE projects of Ministerio de 21

Educación y Ciencia (Spain), and the AGAUR SGR 589 and SGR444 of the Generalitat de Catalunya. 22

The first author (SERC) received an FI grant and a BE grant, both sponsored by the Departament 23

d'Educació i Universitats de la Generalitat de Catalunya and the European Social Fund. We thank Dr. 24

D.G. Pearson and a second anonymous reviewer for their revision of this manuscript and all their 25

valuable comments. We acknowledge the Geological Survey of Canada (GSC), Ottawa, for all of the 26

support during a six-month Volunteer Assistant visit of SERC, especially to Dr. S.E. Jackson who gave 27

us all the support and guidance for carrying out the LA-ICP-MS analyses. SERC thanks Dr. T. Stachel 28

for providing the PTEXL.XLT program and guidance in the application of the geothermobarometers. 29

SERC also thanks Dr. K.A. Mather who kindly helped her to calculate the Angola paleogeotherm. The 30

authors also acknowledge the Electron Microprobe Laboratory, Department of Earth and Planetary 31

Sciences, McGill University, especially to Mr. Lang Shi for assistance in the use of EPMA. The 32

authors also thank Dr. Robert Martin, emeritus professor at the Earth & Planetary Sciences 33

Department, McGill University, who kindly arranged everything to acquire the EPMA analyses at the 34

McGill University and made valuable improvements to the preliminary version of this manuscript. 35

Thanks to the Pacific Centre for Isotopic and Geochemical Research, Department of Earth and Ocean 36

Sciences, University of British Columbia, Vancouver, especially to Dr. B. Kieffer for all his 37

collaboration in the developing the Sm-Nd analyses. We thank ENDIAMA (Empresa Nacional de 38

Diamantes de Angola), which kindly allowed SERC to acquire samples for her Ph.D. thesis and 39

allowed the use of all facilities for the mine trip, especially to M. Watangua (former Chief Geologist) 40

Minerals 2012, 2

19

and Dr. V. Pervov (petrologist). Also thank to Dr. A. Gonçalves, professor at the Universidade 1

Agostinho Neto, Angola, who helped in all the process of logistics and develop the field trip. The 2

authors also thank the Serveis Cientificotècnics de la Universitat de Barcelona for assistance in the use 3

of SEM/ESEM-BSE-EDS analyses (E. Prats. and J. García Veigas). 4

References and Notes 5

1. ENDIAMA–Empresa Nacional de Diamantes de Angola, E.P. História. Available online: 6

http://www.endiama.co.ao/endiama_historia.php# (accessed on 8 July 2012). 7

2. Reis, B. Preliminary note on the distribution and tectonic control of kimberlites in Angola. In 8

Proceedings of the 24th Int. Geol. Congr., Montreal, Canada, August-September 1972; Rep. Sess. 9

4, 276–281. 10

3. The Israeli Diamond Industry. Diamond news. Available online: 11

http://www.israelidiamond.co.il/english/news.aspx?boneid=918&objid=9920 (accessed on 25 12

September 2011). 13

4. Mitchell, R.H. Kimberlites, Orangeites, and Related Rocks; Plenum Press, New York, 1995; pp 14

410. 15

5. Scott Smith, B. H.; Nowicki, T. E.; Russell, J. K.; Webb, K. J.; Mitchell, R. H.; Hetman, C. M.; 16

Robey, J. V. A.; Skinner, E. M. W.; Robey, J. V. Kimberlite terminology and classification. In 17

Proceedings of the 10th Kimberlite Conf., Bangalore, India, 6-11 February 2012; Abstract 18

Volume, 30–31. 19

6. De Carvalho, H.; Tassinari, C.; Alves, P. H. Geochronological review of the Precambrian in 20

western Angola: links with Brazil. J. Afr. Earth Sci. 2000, 31, 383–402. 21

7. Guiraud, R.; Bosworth, W.; Thierry, J.; Delplanque, A. Phanerozoic geological evolution of 22

Northern and Central Africa: an overview. J. Afr. Earth Sci. 2005, 43, 83–143. 23

8. Perevalov, O.V.; Voinovsky, A.S.; Tselikovsky, A.F.; Agueev, Y.L.; Polskoi, F.R.; Khódirev, 24

V.L.; Kondrátiev, A.I. Geologia de Angola. Notícia explicativa da carta geológica a escala 25

1:1.000.000. Serviço Geológico de Angola 1992. Lunda. 26

9. Egorov, K. N.; Roman'ko, E. F.; Podvysotsky, V. T.; Sablukov, S. M.; Garanin, V. K.; D'yakonov, 27

D. B. New data on kimberlite magmatism in southwestern Angola. Russ. Geol. Geophys. 2007, 28

48, 323–336. 29

10. Robles-Cruz, S.E.; Escayola, M.; Jackson, S.; Galí, S.; Pervov, V.; Watangua, M.; Gonçalves, A.; 30

Melgarejo, J.C. U-Pb SHRIMP geochronology of zircon from the Catoca kimberlite, Angola: 31

Implications for diamond exploration. Chem. Geol. 2012, (310-311), 137–147. 32

11. Eley, R.; Grütter, H.; Louw, A.; Tunguno, C.; Twidale, J. Exploration Geology of the Luxinga 33

kimberlite Cluster (Angola) with evidence supporting the presence of kimberlite lava. In 34

Proceedings of the 9th Kimberlite Conf., Frankfurt, Germany, 10-15 August 2008; Extended 35

Abstr. No. 9IKC-A-00166. 36

12. Weis, D.; Kieffer, B.; Maerschalk, C.; Barling, J.; de Jong, J.; Williams, G. A.; Hanano, D.; 37

Pretorius, W.; Mattielli, N.; Scoates, J. S.; Goolaerts, A.; Friedman, R. M.; Mahoney, J. B. High-38

precision isotopic characterization of USGS reference materials by TIMS and MC-ICP-MS. 39

Geochemistry, Geophysics, Geosystems 2006, 7(8), Q08006. 40

Minerals 2012, 2

20

13. Pervov, V.A.; Somov, S.V.; Korshunov, A.V.; Dulapchii, E.V.; Félix, J.T. The Catoca kimberlite 1

pipe, Republic of Angola: A paleovolcanological model. Geol. Ore Deposits 2011, 53(4), 295–2

308. 3

14. Robles-Cruz, S.; Watangua, M.; Melgarejo, J. C.; Galí, S. New Insights into the Concept of 4

Ilmenite as an Indicator for Diamond Exploration, Based on Kimberlite Petrographic Analysis. 5

Revista de la sociedad española de mineralogía, Macla 2008, 9, 205–206. 6

15. Pettit, W. geophysical signatures of some recently discovered large (>40 ha) kimberlite pipes on 7

the Alto Cuilo concession in northeastern Angola. Lithos 2009, 112S, 106–115. 8

16. Robles-Cruz, S.; Watangua, M.; Isidoro, L.; Melgarejo, J. C.; Galí, S.; Olimpio, A. Contrasting 9

compositions and textures of ilmenite in the Catoca kimberlite, Angola, and implications in 10

exploration for diamond. Lithos 2009, 112S, 966–975. 11

17. Pearson, D.G.; Canil, D.; Shirey, S.B. Mantle samples included in volcanic rocks: xenoliths and 12

diamonds. In The mantle and core. Treatise on Geochemistry; Carlson, R.W., Ed.; Elsevier: 13

Washington, USA, 2007; Volume 2, pp.171–275. 14

18. Pearson, D.G.; Wittig, N. Formation of archean continental lithosphere and its diamonds: the root 15

of the problem. J. Geol. Soc. (London, U. K.) 2008, 165, 1–20. 16

19. Grütter, H.S.; Gurney, J.J.; Menzies, A.H.; Winter, F. An updated classification scheme for 17

mantle-derived garnet, for use by diamond explorers. Lithos 2004, 77, 841–857. 18

20. Morris, T.F.; Sage, R.P.; Ayer, J.A.; Crabtree, D.C. A study in clinopyroxene composition: 19

implications for kimberlite exploration. Geochem.: Explor., Environ., Anal. 2002, 2 (4), 321–331. 20

21. Ramsay, R. R. “Geochemistry of diamond indicator minerals”. Ph.D. diss. University of Western 21

Australia, Perth 1992. 22

22. Stachel, T. Developments in geothermobarometry of mantle rocks based on the PTEXL written by 23

Thomas Köhler in 1994. Unpublished data. 2011. 24

23. Nimis, P.; Taylor, W. R. Single clinopyroxene thermobarometry for garnet peridotites. Part I. 25

Calibration and testing of a Cr-in-Cpx barometer and an enstatite-in-Cpx thermometer. Contrib. 26

Mineral. Petrol. 2000, 139, 514–554. 27

24. Nimis, P. Evaluation of diamond potential from the composition of peridotitic chromian diopside. 28

Eur. J. Mineral. 1998, 10, 505–519. 29

25. Mather, K.A.; Pearson, D.G.; McKenzie, D.; Kjarsgaard, B.; Priestley, K. Constraining the depth 30

and thermal history of cratonic lithosphere using peridotite xenolith and xenocryst 31

thermobarometry and seismology. Lithos 2011, 125, 729–742. 32

26. McLean, H.; Banas, A.; Creighton, S.; Whiteford, S.; Luth, R.W.; Stachel, T. Garnet xenocrysts 33

from the Diavik mine, NWT, Canada: Composition, color, and paragenesis. Can. Mineral. 2007, 34

45, 1131–1145. 35

27. Pearson, D. G. The age of continental roots. Lithos 1999, 48, 171–194. 36

28. White, S. H.; De Boorder, H.; Smith, C. B. Structural Controls of Kimberlite and Lamproite 37

Emplacement. J. Geochem. Explor. 1995, 53, 245–264. 38

29. Jelsma, H.; Barnett, W.; Richards, S.; Lister, G. Tectonic setting of kimberlites. Lithos 2009, 112, 39

155–165. 40

30. Brett, R. C.; Russell, J. K.; Moss, S. Origin of olivine in kimberlite: Phenocryst or impostor? 41

Lithos, 2009, 112S, 201–212. 42

Minerals 2012, 2

21

31. Grütter, H.; Latti, D.; Menzies, A. Cr-saturation arrays in concentrate garnet compositions from 1

kimberlite and their use in mantle barometry. J. Petrol. 2006, 47, 801–820. 2

32. Vredevoogd, J. J.; Forbes, W. C. The system diopside–ureyite at 20 kb. Contrib. Mineral. Petrol. 3

1975, 52, 147–156. 4

33. Aschepkov, I.V.; Rotman; Somov, S.V.; Afanasiev, V.P.; Downes, H.; Logvinova, A.M.; 5

Nossyko, S.; Shimupi, J.; Palessky, S.V.; Khmelnikova, O.S.; Vladykin, N.V. Composition and 6

thermal structure of the lithospheric mantle beneath kimberlite pipes from the Catoca cluster, 7

Angola. Tectonophysics 2012, (530-531), 128–151. 8

34. Stachel, T.; Viljoen, K.S.; Brey, G.; Harris, J.W. Metasomatic processes in lherzolitic and 9

harzburgitic domains of diamondiferous lithospheric mantle. Earth Planet. Sci. Lett. 1998, 159, 1–10

12. 11

© 2012 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article 12

distributed under the terms and conditions of the Creative Commons Attribution license 13

(http://creativecommons.org/licenses/by/3.0/). 14

PAPER VI

Supplementary Files

Tables

Tab

le 1

. Su

mm

ary

of E

PM

A d

ata

from

oliv

ine

mac

rocr

ysts

Kim

ber

lite

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

Bor

ehol

e33

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

5

Oli

vin

e m

acro

crys

t1

11

11

2A2A

2A2B

2B2B

3A3A

3A3B

3B3B

3C3C

3C

Poi

nt

506.

3-I-

Ol-

2 50

6.3-

I-O

l-5

506.

3-I-

Ol-

6 50

6.3-

I-O

l-7

506.

3-I-

Ol-

8 50

5B_e

-Ol3

6 50

5B_e

-Ol4

1 50

5B_e

-Ol4

3 50

5B_a

-Ol1

50

5B_a

-Ol7

50

5B_a

-Ol1

0 50

5.6-

b-o

l-1

505.

6-b

-ol-

250

5,6

C O

L 3

450

5,6

e ol

i 60

505,

6 e

ol 6

150

5,6

e ol

65

505.

6-b

-ol-

1150

5,6

C O

L 3

150

5,6

C O

L 3

2

(wt.

%)

SiO

2

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

TiO

2

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.04

0.0

20

.01

0.0

20

.03

0.0

50

.04

0.0

30

.07

Al2

O3

0

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

20

.00

0.0

00

.02

0.0

0

Cr2

O3

0

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

30

.01

0.0

30

.03

0.0

30

.07

0.1

20

.11

0.1

1

Fe2

O3

0.2

20

.33

0.6

10

.47

0.0

00

.00

0.0

00

.37

0.2

00

.00

0.0

0—

——

——

——

——

FeO

7.4

57

.37

7.1

37

.17

7.6

67

.94

7.9

17

.64

7.8

98

.15

8.0

4—

——

——

——

——

FeO

T7

.66

7.6

97

.74

7.6

57

.66

7.9

47

.91

8.0

18

.09

8.1

58

.04

10

.00

10

.00

8.3

91

1.9

91

1.8

31

2.2

11

4.6

31

4.5

41

4.3

8

Mn

O

0

.11

0.1

00

.11

0.1

00

.12

0.1

50

.13

0.1

20

.13

0.1

40

.15

0.1

60

.15

0.0

70

.12

0.1

80

.15

0.1

90

.20

0.1

4

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

0.0

10

.01

0.0

10

.00

0.0

20

.01

0.0

00

.02

0.0

30

.02

0.0

20

.02

0.0

10

.00

0.0

60

.07

0.0

20

.03

0.0

40

.03

NiO

——

——

——

——

——

—0

.35

0.3

30

.41

0.3

50

.39

0.3

00

.15

0.1

80

.08

Su

m O

x%9

9.9

71

00

.05

99

.83

99

.88

10

0.0

19

9.6

49

9.7

59

9.6

09

9.8

89

9.9

79

9.6

89

9.9

99

9.6

29

9.8

29

9.5

59

9.3

99

9.2

89

9.5

89

9.8

39

9.6

3

(ap

fu)

Si

1.0

01

.00

0.9

91

.00

1.0

01

.00

1.0

01

.00

1.0

01

.00

1.0

00

.99

0.9

90

.98

0.9

90

.98

0.9

80

.98

0.9

90

.99

Ti

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

Al t

otal

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

Cr

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

Fe3

0.0

00

.01

0.0

10

.01

0.0

00

.00

0.0

00

.01

0.0

00

.00

0.0

0—

——

——

——

——

Fe2

0.1

50

.15

0.1

40

.15

0.1

60

.16

0.1

60

.16

0.1

60

.17

0.1

6—

——

——

——

——

FeT

0.1

50

.16

0.1

60

.15

0.1

60

.16

0.1

60

.16

0.1

60

.17

0.1

60

.21

0.2

10

.17

0.2

50

.25

0.2

60

.31

0.3

10

.30

Mn

0

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

0

Mg

1.8

41

.85

1.8

51

.85

1.8

41

.83

1.8

31

.84

1.8

31

.83

1.8

31

.80

1.8

11

.85

1.7

61

.78

1.7

71

.72

1.7

11

.71

Ca

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

Ni

——

——

——

——

——

—0

.01

0.0

10

.01

0.0

10

.01

0.0

10

.00

0.0

00

.00

%T

eph

roit

e0

.11

0.1

00

.11

0.1

00

.13

0.1

50

.13

0.1

30

.13

0.1

50

.15

0.1

50

.15

0.1

00

.15

0.2

00

.15

0.2

00

.20

0.1

5

%F

orst

erit

e9

2.1

49

2.1

39

2.1

09

2.2

19

2.0

79

1.7

49

1.8

09

1.7

49

1.6

39

1.5

49

1.6

38

9.5

88

9.6

39

1.4

18

7.4

68

7.6

28

7.2

38

4.6

28

4.6

58

4.8

3

%F

ayal

ite

7.7

47

.75

7.7

87

.69

7.7

88

.10

8.0

68

.11

8.2

08

.29

8.1

91

0.2

71

0.2

28

.49

12

.39

12

.18

12

.62

15

.18

15

.15

15

.02

Ca-

Ol

0.0

10

.02

0.0

10

.00

0.0

30

.01

0.0

00

.03

0.0

30

.02

0.0

2

Tot

al1

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

0

mg

#9

2.2

59

2.2

49

2.2

19

2.3

09

2.2

19

1.8

99

1.9

39

1.8

89

1.7

99

1.6

99

1.7

98

9.7

28

9.7

69

1.5

08

7.5

98

7.8

08

7.3

68

4.7

98

4.8

28

4.9

6

Tex

ture

typ

es:

(1)

ho

mo

gen

eou

s; (

2A

) cr

ysta

l w

ith

"id

din

gsi

te"_

core

; (

2B

) cr

ysta

l w

ith

"id

din

gsi

te"_

rim

; (3

A)

zon

ed_

core

; (3

B)

zon

ed_

mid

dle

; (3

C)

zon

ed_

rim

Tab

le 2

a. S

um

mar

y of

EP

MA

dat

a fr

om g

arn

etK

imbe

rlit

eA

C4

AC

4A

C63

AC

63C

AC

AC

AC

AC

AC

AC

AC

AC

A

Bor

ehol

e4

46

633

533

533

577

/35

33/3

533

/35

33/3

533

533

5

Tex

ture

/ass

ocia

tion

55

33

11

22

33

44

5

Poi

nt

159A

_a10

_Grt

159A

_b5_

Grt

206B

-a-G

rt5

206B

-a-G

rt9

607B

-a_g

rt2

607B

-i_g

rt24

607

C-C

-Grt

?7 3

98.2

_b-G

rt1

499,

6-B

-Gt-

2 49

9,6-

B-G

t-6

460_

a-G

rt2

551-

ag-

8255

1- c

-71

(wt.

%)

SiO

237.3

637.2

241.5

741.6

341.8

941.9

741.5

441.7

441.6

341.5

641.8

341.9

741.8

2

TiO

20.0

90.0

20.0

90.0

80.0

80.1

20.0

80.1

40.2

80.2

50.1

40.1

80.7

8

Al2

O3

21.3

622.0

720.4

120.1

823.3

123.5

721.0

620.4

220.2

520.3

721.3

620.5

721.4

1

Cr2

O3

0.0

50.0

45.0

75.1

60.0

30.0

44.0

55.0

34.4

64.3

33.5

83.9

70.6

8

Fe2

O3

1.9

33.3

00.4

40.3

80.4

50.0

00.3

00.0

00.0

00.0

20.0

00.9

40.7

2

FeO

26.4

825.9

37.5

27.6

211.3

411.8

98.1

07.7

77.7

47.8

38.2

56.7

410.8

3

Mn

O0.9

31.1

70.4

20.4

30.3

40.3

30.5

10.4

40.4

00.4

00.4

40.4

70.3

9

MgO

3.9

98.8

820.6

520.5

819.3

418.8

919.9

419.6

520.4

820.3

419.3

121.6

419.2

6

CaO

7.9

81.2

33.9

33.9

93.1

23.0

64.3

74.9

14.1

14.2

35.0

83.5

64.0

2

ZrO

2

——

——

——

——

——

——

—

Y2O

3

——

——

——

——

——

——

—

V2O

3

——

——

——

——

——

——

—

Tot

al100.1

699.8

6100.1

0100.0

499.9

099.8

799.9

5100.1

099.3

499.3

399.9

9100.0

399.9

1

(apfu

)

Si

2.9

42.8

92.9

82.9

93.0

03.0

12.9

83.0

03.0

03.0

03.0

02.9

93.0

1

Ti

0.0

10.0

00.0

10.0

00.0

00.0

10.0

00.0

10.0

20.0

10.0

10.0

10.0

4

Al

1.9

82.0

21.7

21.7

11.9

71.9

91.7

81.7

31.7

21.7

31.8

11.7

31.8

2

Cr

0.0

00.0

00.2

90.2

90.0

00.0

00.2

30.2

90.2

50.2

50.2

00.2

20.0

4

Fe3

0.1

10.1

90.0

20.0

20.0

20.0

00.0

20.0

00.0

00.0

00.0

00.0

50.0

4

Fe2

1.7

41.6

80.4

50.4

60.6

80.7

10.4

90.4

70.4

70.4

70.5

00.4

00.6

5

Mn

0.0

60.0

80.0

30.0

30.0

20.0

20.0

30.0

30.0

20.0

20.0

30.0

30.0

2

Mg

0.4

71.0

32.2

12.2

02.0

62.0

22.1

32.1

12.2

02.1

92.0

72.3

02.0

7

Ca

0.6

70.1

00.3

00.3

10.2

40.2

40.3

40.3

80.3

20.3

30.3

90.2

70.3

1

Zr

——

——

——

——

——

——

—

Y

——

——

——

——

——

——

—

V

——

——

——

——

——

——

—

alm

andi

ne

59.1

658.2

515.1

015.2

922.6

123.8

716.2

815.6

915.5

015.6

816.6

213.3

821.3

6

pyro

pe15.8

935.5

573.9

373.5

968.7

267.6

071.4

370.7

273.1

572.6

469.3

676.6

267.7

1

gros

sula

r21.5

13.2

38.5

58.6

57.8

67.8

49.8

710.8

69.1

19.4

311.7

57.7

89.5

3

spes

sart

ine

2.1

02.6

60.8

60.8

60.6

90.6

61.0

30.8

90.8

10.8

20.9

00.9

50.7

8

uva

rovi

te0.0

30.0

01.4

21.4

80.0

10.0

11.2

71.7

91.3

51.3

41.3

21.0

10.2

0

andr

adit

e1.2

40.3

10.1

20.1

00.1

00.0

00.0

90.0

00.0

00.0

00.0

00.2

30.2

1

Ca-

Ti G

t0.0

60.0

00.0

30.0

20.0

20.0

20.0

20.0

50.0

80.0

70.0

50.0

40.2

2

Tot

al100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0

Tex

ture

/ass

oci

ati

on

: (1

) e

clo

git

e;

(2)

grt

lh

erz

oli

te;

(3

) p

hl-

grt

we

hrl

ite

; (4

) m

acr

ocr

yst

; (5

) m

icro

cryst

.

Tab

le 2

b. C

onti

nu

atio

nK

imbe

rlit

eC

AC

U79

CU

79C

U79

CU

79C

U79

CU

79T

ZT

ZT

ZT

ZT

ZT

Z

Bor

ehol

e33

5M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

0134

34G

18G

18G

18G

18

Tex

ture

/ass

ocia

tion

54

42

23

32

24

45

5

Poi

nt

01 -

3dia

-m'-

5711

1,5_

f_4

,8R

EP

_a5_

grt

134B

_b_6

134B

_b_9

d4_g

rt_M

P77

,6B

_g_2

346.

3-a_

grt2

346.

3-a_

grt3

38-a

-538

-a-4

638

-c-5

438

-c-5

8

(wt.

%)

SiO

241.2

442.0

440.3

140.4

641.4

340.7

242.4

441.4

641.5

442.5

842.3

442.4

941.7

4

TiO

20.3

40.6

40.7

20.1

80.1

50.1

30.5

50.2

20.2

30.8

20.8

40.1

10.1

4

Al2

O3

21.4

321.2

619.6

222.7

122.3

321.5

721.2

122.5

522.5

021.1

521.1

823.7

023.6

1

Cr2

O3

2.6

61.7

03.5

51.3

81.4

02.4

22.3

11.4

71.4

21.2

31.3

00.4

20.4

0

Fe2

O3

2.1

71.6

84.4

93.9

61.8

73.3

10.6

52.9

52.4

70.3

70.7

10.4

52.1

2

FeO

6.5

76.4

04.9

96.0

18.1

97.1

26.2

16.9

17.3

48.2

37.9

66.6

25.1

7

Mn

O0.3

90.2

30.3

20.4

30.4

10.5

30.2

20.4

50.5

00.2

70.2

40.2

80.3

2

MgO

20.5

721.5

620.4

520.4

419.9

519.5

021.7

920.6

220.5

421.1

621.1

621.6

121.7

8

CaO

4.6

74.5

15.5

34.4

24.3

04.9

94.6

64.4

34.2

44.2

44.2

64.2

84.4

6

ZrO

2

——

——

——

—0.0

40.0

2—

——

—

Y2O

3

——

——

——

——

——

——

—

V2O

3

——

——

——

——

——

——

—

Tot

al100.0

5100.0

299.9

899.9

9100.0

3100.2

9100.0

3101.1

1100.7

9100.0

599.9

999.9

599.7

4

(apfu

)

Si

2.9

52.9

82.9

02.8

92.9

62.9

23.0

02.9

32.9

43.0

33.0

12.9

92.9

4

Ti

0.0

20.0

30.0

40.0

10.0

10.0

10.0

30.0

10.0

10.0

40.0

40.0

10.0

1

Al

1.8

01.7

81.6

71.9

11.8

81.8

21.7

71.8

81.8

81.7

71.7

81.9

61.9

6

Cr

0.1

50.1

00.2

00.0

80.0

80.1

40.1

30.0

80.0

80.0

70.0

70.0

20.0

2

Fe3

0.1

20.0

90.2

40.2

10.1

00.1

80.0

30.1

60.1

30.0

20.0

40.0

20.1

1

Fe2

0.3

90.3

80.3

00.3

60.4

90.4

30.3

70.4

10.4

40.4

90.4

70.3

90.3

1

Mn

0.0

20.0

10.0

20.0

30.0

20.0

30.0

10.0

30.0

30.0

20.0

10.0

20.0

2

Mg

2.1

92.2

82.2

02.1

82.1

32.0

92.3

02.1

72.1

72.2

42.2

42.2

72.2

9

Ca

0.3

60.3

40.4

30.3

40.3

30.3

80.3

50.3

40.3

20.3

20.3

20.3

20.3

4

Zr

——

——

——

—0.0

00.0

0—

——

—

Y

——

——

——

——

——

——

—

V

——

——

——

——

——

——

—

alm

andi

ne

13.2

512.5

910.2

112.3

816.4

814.5

812.1

113.8

814.7

115.9

315.5

013.0

010.3

4

pyro

pe73.9

075.5

974.6

375.0

671.5

971.2

275.8

073.8

173.3

973.0

373.4

175.6

777.5

9

gros

sula

r10.4

110.1

211.2

310.0

810.0

811.1

310.5

00.0

00.5

69.7

89.7

610.4

910.6

5

spes

sart

ine

0.8

00.4

60.6

60.9

00.8

41.1

00.4

30.9

11.0

10.5

30.4

70.5

60.6

5

uva

rovi

te0.8

70.5

41.3

60.4

10.4

20.8

40.7

74.0

03.8

90.3

80.4

00.1

20.1

2

andr

adit

e0.6

70.5

11.6

41.1

20.5

41.0

90.2

07.6

36.4

40.1

10.2

10.1

30.6

1

Ca-

Ti G

t0.1

10.1

90.2

60.0

50.0

40.0

40.1

70.2

40.2

50.0

30.0

4

Tot

al100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.2

3100.0

0100.0

0100.0

0100.0

0100.0

0

Tex

ture

/ass

oci

ati

on

: (1

) e

clo

git

e;

(2)

grt

lh

erz

oli

te;

(3

) p

hl-

grt

we

hrl

ite

; (4

) m

acr

ocr

yst

; (5

) m

icro

cryst

.

Tab

le 3

a. S

um

mar

y of

EP

MA

dat

a fr

om c

lin

opyr

oxen

eK

imb

erlit

eA

C63

AC

63A

C63

AC

63C

AC

AC

AC

AC

AC

AC

AC

AC

AC

AC

AC

AC

AC

AA

n11

6A

n11

6

Bor

ehol

e6

66

633

/35

33/3

553

638

433

533

533

533

533

577

-35

33/3

533

/35

33/3

533

/35

116

116

Tex

ture

/ass

ocia

tion

88

55

1010

88

99

11

22

1111

55

66

Poi

nt

9A-a

-Cp

x8 A

-a-C

px1

5 6B

-a-C

px5

6B

-a-C

px8

50

5.6-

b-1

450

5.6-

b-1

530

4-j-

3538

4-m

-40

601-

a-11

060

1 a-

113B

_d-C

px2

1 B

_j-C

px3

2 C

-3C

PX

8.2-

d-C

px2

B

_d-C

px1

B

_d-C

px3

6-B

-Cp

x-4

6-B

-Cp

x-5

8A_c

-Cp

x4 8

A_c

-Cp

x6

(wt.

%)

SiO

25

4.7

05

4.4

55

4.8

65

5.0

05

4.1

95

3.9

15

5.0

85

5.2

95

5.0

45

5.0

15

5.3

75

5.4

95

4.7

25

4.7

45

4.3

65

4.6

15

4.3

25

4.4

15

4.7

15

4.5

2

Al2

O3

2.2

42

.30

4.5

24

.48

1.0

41

.07

3.7

72

.66

1.7

82

.33

7.3

06

.69

3.3

83

.17

2.9

53

.11

3.6

63

.70

0.7

60

.76

TiO

20

.28

0.2

90

.12

0.1

20

.11

0.1

30

.49

0.2

10

.26

0.0

90

.43

0.3

60

.11

0.1

90

.11

0.0

90

.31

0.3

00

.07

0.0

7

FeO

0.7

10

.48

0.0

00

.00

0.3

20

.00

3.8

22

.17

1.9

21

.46

0.8

31

.27

0.6

40

.20

0.1

30

.42

0.0

90

.45

0.0

00

.00

Fe2

O3

2.4

82

.88

2.8

22

.40

0.0

00

.00

1.9

30

.27

1.7

61

.31

2.5

52

.15

1.9

32

.36

2.4

42

.07

2.8

12

.49

2.6

93

.68

Cr2

O3

2.8

42

.98

4.5

54

.46

1.9

92

.02

0.2

82

.90

1.8

32

.24

0.0

90

.12

3.0

42

.89

2.9

02

.83

3.5

13

.87

3.2

73

.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

Mn

O0

.09

0.0

90

.07

0.1

00

.05

0.0

70

.18

0.1

30

.07

0.1

20

.05

0.0

50

.07

0.0

80

.07

0.0

70

.10

0.0

90

.07

0.0

6

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

Na2

O2

.89

2.9

94

.91

4.8

61

.58

1.6

12

.88

2.9

72

.52

2.6

04

.89

4.5

63

.22

3.2

73

.02

3.0

03

.76

3.8

92

.45

2.4

7

K2O

0.0

20

.03

0.0

20

.02

0.0

80

.11

0.0

00

.01

0.0

20

.00

0.0

20

.02

0.0

20

.01

0.0

20

.01

0.0

30

.05

0.0

20

.02

Su

m O

x%1

00

.19

10

0.1

21

00

.02

99

.93

99

.72

10

0.5

01

00

.13

99

.98

99

.97

99

.87

99

.96

99

.95

10

0.0

29

9.9

19

9.7

01

00

.01

99

.58

99

.73

99

.93

99

.98

(ap

fu)

Si

1.9

71

.97

1.9

71

.98

2.0

02

.00

1.9

82

.00

1.9

91

.99

1.9

71

.98

1.9

71

.97

1.9

71

.97

1.9

61

.96

1.9

81

.97

Al

0.1

00

.10

0.1

90

.19

0.0

50

.05

0.1

60

.11

0.0

80

.10

0.3

10

.28

0.1

40

.13

0.1

30

.13

0.1

60

.16

0.0

30

.03

Ti

0.0

10

.01

0.0

00

.00

0.0

00

.00

0.0

10

.01

0.0

10

.00

0.0

10

.01

0.0

00

.01

0.0

00

.00

0.0

10

.01

0.0

00

.00

Fe2

+0

.02

0.0

10

.00

0.0

00

.01

0.0

00

.11

0.0

70

.06

0.0

40

.02

0.0

40

.02

0.0

10

.00

0.0

10

.00

0.0

10

.00

0.0

0

Fe3

+0

.07

0.0

80

.08

0.0

60

.00

0.0

00

.05

0.0

10

.05

0.0

40

.07

0.0

60

.05

0.0

60

.07

0.0

60

.08

0.0

70

.07

0.1

0

Cr

0.0

80

.08

0.1

30

.13

0.0

60

.06

0.0

10

.08

0.0

50

.06

0.0

00

.00

0.0

90

.08

0.0

80

.08

0.1

00

.11

0.0

90

.09

Mg

0.8

10

.81

0.7

10

.71

0.8

90

.93

0.8

60

.79

0.8

10

.81

0.6

80

.69

0.8

00

.83

0.8

10

.81

0.8

10

.79

0.8

40

.85

Mn

0

.00

0.0

00

.00

0.0

00

.00

0.0

00

.01

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

0

Ca

0.7

30

.73

0.5

80

.57

0.8

70

.88

0.6

00

.72

0.7

70

.77

0.6

00

.62

0.7

00

.68

0.7

30

.73

0.6

20

.62

0.8

00

.80

Na

0.2

00

.21

0.3

40

.34

0.1

10

.12

0.2

00

.21

0.1

80

.18

0.3

40

.32

0.2

20

.23

0.2

10

.21

0.2

60

.27

0.1

70

.17

K

0.0

00

.00

0.0

00

.00

0.0

00

.01

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

00

.00

Su

m c

at4

.00

4.0

04

.00

4.0

04

.07

4.1

34

.00

4.0

04

.00

4.0

04

.00

4.0

04

.00

4.0

04

.00

4.0

04

.00

4.0

04

.00

4.0

0

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

1.3

70

.93

0.0

00

.00

0.5

50

.00

7.2

44

.13

3.5

32

.72

1.8

82

.79

1.2

60

.40

0.2

50

.81

0.1

80

.96

0.0

00

.00

Tot

al1

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

0

Q9

0.6

28

9.8

18

2.7

38

3.4

59

7.5

09

7.4

78

8.2

19

2.9

19

2.9

89

2.3

27

7.7

07

9.9

88

8.5

98

8.4

28

8.9

18

9.1

68

6.1

18

6.3

39

3.9

09

2.5

4

Jd5

.49

5.6

71

2.3

51

2.3

42

.50

2.5

38

.88

6.6

54

.30

5.6

51

8.2

41

6.6

18

.37

7.8

47

.25

7.6

19

.32

9.5

71

.87

1.8

3

Ae

3.8

94

.53

4.9

24

.22

0.0

00

.00

2.9

10

.43

2.7

22

.03

4.0

73

.41

3.0

53

.74

3.8

43

.24

4.5

74

.10

4.2

25

.63

Tot

al1

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

01

00

.00

10

0.0

0

Tex

ture

/ass

oci

atio

n:

(1)

eclo

git

e; (

2)

grt

lh

erzo

lite

; (

3)

lher

zolite

; (

4)

pyr

oxen

ite;

(5

) p

hl-

grt

weh

rlit

e; (

6)

phl-

rich

suit

e; (

7)

meg

acry

st;

(8)

mac

rocr

yst;

(9)

mic

rocr

yst

; (1

0)

as incl

usi

on in c

hr;

(11)

as incl

usi

on in g

rt.

Tab

le 3

b. C

onti

nu

atio

nK

imb

erlit

eC

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79T

ZT

ZT

ZT

Z

Bor

ehol

eM

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01G

10G

1834

34

Tex

ture

/ass

ocia

tion

22

88

117

79

55

66

44

33

88

22

Poi

nt

178-

b13

175.

5-h

112

,6B

_d_6

1C_d

1_p

x1E

P_a

12_P

x13

4B_i

_213

6,8_

b_3

105,

5_c_

61_

cpx_

MP

63_c

cp

x311

1_c1

111_

c811

1_a1

111

1_a1

3,5

B_a

5_p

xB_e

_cp

x 11

95 f

-106

38-e

-24

6.3-

a_cp

x36.

3-a_

cpx4

(wt.

%)

SiO

253.7

955.0

653.5

954.1

450.5

354.6

854.8

354.0

354.1

353.4

653.2

953.1

453.3

253.3

554.2

353.3

455.4

155.6

954.7

454.7

9

Al2

O3

7.3

82.3

91.4

53.5

35.3

23.5

23.6

94.0

92.3

31.1

21.4

51.5

51.0

90.9

92.4

41.1

82.9

03.4

22.9

53.1

9

TiO

20.3

50.1

50.7

90.3

90.7

20.3

90.3

20.5

10.1

21.1

71.5

11.3

00.4

70.3

40.2

10.2

40.2

00.2

60.1

80.2

3

FeO

1.7

02.8

13.2

82.8

82.7

62.8

83.0

02.6

41.0

02.2

84.0

93.4

64.2

05.2

42.9

01.3

71.3

52.2

40.0

00.0

0

Fe2

O3

1.5

50.5

92.6

92.8

14.7

42.6

72.6

13.5

61.7

42.8

31.5

42.7

02.4

41.9

22.3

03.2

71.0

80.5

14.9

24.4

2

Cr2

O3

0.1

31.3

30.4

50.5

01.0

50.4

70.5

70.0

61.4

80.4

50.1

90.1

90.0

70.0

30.5

90.5

11.3

91.8

00.8

30.8

5

MgO

13.5

018.2

818.3

117.8

517.5

617.7

418.2

215.3

616.3

817.8

216.6

317.7

116.2

016.9

615.5

916.1

915.2

314.3

616.0

415.6

4

Mn

O0.0

60.0

30.2

00.1

40.3

90.1

20.1

70.1

40.0

80.1

70.1

40.1

20.1

60.1

30.1

00.0

60.0

00.0

90.0

00.0

0

CaO

17.8

817.6

218.4

315.5

515.8

415.2

114.3

116.3

120.8

319.7

419.8

318.9

821.4

720.5

019.6

722.5

819.7

718.4

320.5

820.0

2

Na2

O3.4

11.7

01.0

92.2

21.2

32.5

02.5

63.0

21.6

91.1

81.2

51.1

30.7

40.4

81.9

51.0

32.7

03.2

92.6

82.8

9

K2O

0.0

30.0

30.0

00.0

10.2

00.0

10.0

00.0

00.0

10.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

30.0

10.0

20.0

0

Su

m O

x%99.6

299.9

2100.0

199.9

9100.0

599.9

2100.0

399.3

7100.0

199.9

899.9

910

0.0

0100.0

599.9

899.7

599.5

999.9

6100.0

4100.7

2100.5

1

(ap

fu)

Si

1.9

31.9

81.9

41.9

51.8

31.9

61.9

61.9

61.9

61.9

41.9

51.9

31.9

61.9

61.9

71.9

51.9

92.0

01.9

51.9

6

Al

0.3

10.1

00.0

60.1

50.2

30.1

50.1

60.1

70.1

00.0

50.0

60.0

70.0

50.0

40.1

00.0

50.1

20.1

40.1

20.1

3

Ti

0.0

10.0

00.0

20.0

10.0

20.0

10.0

10.0

10.0

00.0

30.0

40.0

40.0

10.0

10.0

10.0

10.0

10.0

10.0

00.0

1

Fe2

+0.0

50.0

80.1

00.0

90.0

80.0

90.0

90.0

80.0

30.0

70.1

30.1

10.1

30.1

60.0

90.0

40.0

40.0

70.0

00.0

0

Fe3

+0.0

40.0

20.0

70.0

80.1

30.0

70.0

70.1

00.0

50.0

80.0

40.0

70.0

70.0

50.0

60.0

90.0

30.0

10.1

30.1

2

Cr

0.0

00.0

40.0

10.0

10.0

30.0

10.0

20.0

00.0

40.0

10.0

10.0

10.0

00.0

00.0

20.0

10.0

40.0

50.0

20.0

2

Mg

0.7

20.9

80.9

90.9

60.9

50.9

50.9

70.8

30.8

80.9

60.9

10.9

60.8

90.9

30.8

40.8

80.8

20.7

70.8

50.8

3

Mn

0.0

00.0

00.0

10.0

00.0

10.0

00.0

10.0

00.0

00.0

10.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

00.0

0

Ca

0.6

90.6

80.7

20.6

00.6

20.5

80.5

50.6

30.8

10.7

70.7

80.7

40.8

40.8

10.7

70.8

90.7

60.7

10.7

80.7

7

Na

0.2

40.1

20.0

80.1

50.0

90.1

70.1

80.2

10.1

20.0

80.0

90.0

80.0

50.0

30.1

40.0

70.1

90.2

30.1

80.2

0

K

0.0

00.0

00.0

00.0

00.0

10.0

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

0

Su

m c

at4.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

04.0

0

Wo

47.0

738.9

439.6

636.4

737.3

336.0

934.0

741.0

446.9

142.6

242.9

640.9

745.4

042.5

445.0

948.9

047.0

645.9

047.9

847.9

2

En

49.4

456.2

154.8

258.2

557.5

858.5

760.3

653.7

751.3

353.5

450.1

253.1

947.6

648.9

749.7

248.7

850.4

449.7

552.0

252.0

8

Fs

3.4

94.8

45.5

15.2

75.0

85.3

35.5

75.1

91.7

63.8

46.9

25.8

46.9

38.4

95.1

92.3

22.5

04.3

50.0

00.0

0

Tot

al100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

010

0.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0

Q80.5

093.7

093.0

387.9

282.2

188.0

087.7

085.0

292.1

493.5

094.5

19

2.7

994.1

995.1

991.0

392.7

891.4

190.6

986.4

986.3

3

Jd17.2

05.4

43.1

98.0

111.3

48.0

98.4

89.6

25.3

22.4

93.2

73.4

22.3

92.1

55.6

02.6

16.9

48.5

06.5

47.2

5

Ae

2.3

00.8

63.7

84.0

76.4

53.9

13.8

25.3

52.5

44.0

12.2

23.7

93.4

22.6

63.3

74.6

11.6

50.8

16.9

76.4

2

Tot

al100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

010

0.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0100.0

0

Tex

ture

/ass

oci

atio

n:

(1)

eclo

git

e; (

2)

grt

lh

erzo

lite

; (

3)

lher

zoli

te;

(4)

pyr

oxen

ite;

(5)

phl-

grt

weh

rlit

e; (

6)

phl-

rich

suit

e; (

7)

meg

acry

st;

(8)

mac

rocr

yst;

(9)

mic

rocr

yst

; (1

0)

as i

ncl

usi

on i

n c

hr;

(11)

as i

ncl

usi

on i

n g

rt.

Tab

le 4

a. S

um

mar

y of

LA

-IC

P-M

S d

ata

from

gar

net

Kim

ber

lite

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

CA

Bor

ehol

e33

533

533

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

33/3

533

/35

335

335

335

33/3

533

/35

33/3

533

/35

33/3

533

553

633

5

Tex

ture

/ass

ocia

tion

11

12

22

33

34

44

11

12

23

33

44

4

Min

eral

grt

grt

grt

grt

grt

grt

grt

grt

grt

grt

grt

grt

cpx

cpx

cpx

cpx

cpx

cpx

cpx

cpx

cpx

cpx

cpx

Sp

ot n

ame

au18

a08

au18

a12

au18

a13

se15

a09

se15

a10

se15

a13

au19

c05

au19

c06

au19

c08

au19

a13

se15

a07

se15

c03

au18

a05

au18

a09

au18

a11

au19

c12

au19

c13

au19

c09

au19

c10

au19

c11

fe25

a15

fe26

a09

fe26

c13

All

va

lues

are

rep

ort

ed

in

pp

m

Sc

62

61

64

10

01

01

96

19

01

90

17

51

69

19

14

72

19

21

30

30

52

53

54

——

—

Ni

28

30

32

93

97

89

49

11

94

71

82

25

26

11

40

33

04

32

44

43

08

31

63

06

33

02

41

37

9

Ga

10

11

12

10

10

10

47

45

41

01

67

16

77

44

48

89

Y1

71

81

72

12

21

94

13

83

51

61

72

44

12

44

67

74

55

Zr

15

.61

7.6

17

.65

1.8

50

.84

8.1

87

.58

2.3

81

.36

9.9

77

.51

2.9

92

.12

1.0

51

.02

0.1

20

.98

3.7

84

.88

5.9

55

.21

9.6

57

.5

La

0.0

1<

0.0

15

0.0

31

.32

1.1

00

.06

1.9

00

.84

0.1

84

.22

0.6

00

.00

21

.45

5.6

91

4.6

13

.52

4.8

23

.07

3.1

34

.25

3.1

52

.08

19

.02

Ce

0.1

70

.22

0.2

02

.65

2.3

00

.32

3.5

91

.76

0.5

57

.93

1.6

20

.06

63

.61

14

.34

39

.56

9.2

61

1.3

41

1.1

61

1.7

41

3.8

51

1.9

87

.66

45

.88

Pr

0.0

70

.08

0.0

50

.31

0.3

10

.10

0.4

60

.21

0.1

11

.13

0.3

30

.03

9.8

62

.09

5.6

81

.54

1.7

62

.19

2.1

32

.44

2.1

21

.29

6.0

4

Nd

0.8

10

.92

0.7

81

.79

2.0

01

.03

2.6

41

.36

1.3

86

.73

3.3

00

.54

41

.00

8.6

82

2.9

28

.67

9.1

21

3.2

01

1.9

81

4.4

91

0.7

47

.05

28

.04

Sm

0.6

80

.56

0.4

01

.03

1.0

80

.94

1.5

01

.36

1.5

43

.61

3.0

80

.91

6.1

91

.22

3.0

62

.24

2.3

33

.33

3.5

13

.45

2.6

91

.89

4.7

8

Eu

0.3

10

.35

0.2

90

.52

0.5

30

.58

0.8

60

.70

0.7

61

.47

1.6

10

.51

1.5

30

.28

0.6

90

.67

0.7

11

.17

1.0

31

.16

0.9

60

.64

1.3

8

Gd

1.2

81

.31

1.0

82

.00

1.8

81

.88

3.5

73

.34

2.9

74

.21

5.2

32

.16

3.4

50

.63

1.6

91

.80

1.9

22

.99

2.8

23

.13

1.8

71

.69

2.9

7

Tb

0.2

70

.26

0.2

90

.48

0.4

70

.47

0.8

50

.73

0.7

10

.56

0.8

60

.50

0.3

00

.06

0.1

50

.24

0.2

50

.40

0.4

20

.42

0.2

50

.24

0.3

1

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

0.6

50

.67

0.6

50

.83

0.8

50

.76

1.5

61

.40

1.3

30

.62

0.6

80

.93

0.1

60

.05

0.1

00

.18

0.1

80

.29

0.3

20

.31

0.1

80

.22

0.2

0

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

Tex

ture

/ass

oci

ati

on

: (1

) e

clo

git

e;

(2)

grt

lh

erz

oli

te;

(3

) p

hl-

grt

we

hrl

ite

; (4

) m

acr

ocr

yst

. (<

) b

elo

w t

he

de

tect

ion

lim

it;

(—)

no

t d

ete

cte

d

Tab

le 4

b. S

um

mar

y of

LA

-IC

P-M

S d

ata

from

cli

nop

yrox

ene

Kim

ber

lite

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79C

U79

CU

79T

ZT

ZT

ZT

ZT

ZT

Z

Bor

ehol

eM

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01M

FD

01G

1834

34G

10G

10G

10

Tex

ture

/ass

ocia

tion

22

23

33

55

52

22

44

45

55

55

5

Min

eral

grt

grt

grt

grt

grt

grt

grt

grt

grt

cpx

cpx

cpx

cpx

cpx

cpx

grt

grt

grt

cpx

cpx

cpx

Sp

ot n

ame

au18

b07

au18

b08

au18

b13

se15

c03

se15

c05

se15

c06

au19

b11

au19

b13

au19

b14

au18

b09

au18

b14

au18

b15

au19

b05

au19

b06

au19

b16

fe27

b07

fe27

c09

fe27

c14

fe27

a08

fe27

a09

fe27

a16

All

va

lues

are

rep

ort

ed i

n p

pm

Sc

80

74

77

47

50

10

27

97

77

83

33

23

44

65

85

4—

——

——

—

Ni

21

21

21

58

83

41

13

10

91

12

26

52

71

26

01

23

17

92

10

12

31

08

12

72

41

24

32

51

Ga

88

81

01

07

12

12

13

77

89

11

12

15

19

14

66

5

Y2

42

32

22

42

53

02

22

12

24

44

77

61

81

81

94

42

Zr

15

.41

9.6

17

.81

2.9

13

.26

6.5

40

.03

9.6

41

.23

4.6

34

.43

9.2

64

.19

3.5

88

.55

4.6

43

.15

4.1

65

.76

9.4

30

.4

La

<0

.00

69

<0

.00

42

0.0

00

.00

0.1

20

.02

0.0

30

.03

0.0

36

.89

6.8

76

.73

5.4

15

.69

5.1

90

.05

1.1

60

.11

21

.82

20

.85

11

.37

Ce

0.0

50

.03

0.0

20

.06

0.2

80

.29

0.2

30

.26

0.2

31

4.1

11

4.3

81

3.9

91

5.6

21

8.1

52

0.0

30

.39

2.9

40

.47

66

.16

61

.78

25

.99

Pr

0.0

20

.01

0.0

10

.03

0.0

70

.12

0.0

90

.09

0.1

02

.21

2.2

12

.19

2.4

93

.07

3.1

30

.12

0.3

30

.12

8.6

58

.27

3.1

4

Nd

0.3

40

.16

0.2

10

.54

0.7

71

.64

0.8

00

.92

0.7

11

0.8

21

0.7

41

0.4

61

3.1

51

6.1

01

5.6

81

.38

1.7

71

.15

36

.97

36

.12

11

.87

Sm

0.4

90

.42

0.4

30

.91

0.9

61

.38

0.6

80

.79

0.7

22

.69

2.4

92

.73

3.2

84

.09

3.7

20

.92

0.7

50

.92

6.0

46

.37

2.1

3

Eu

0.2

90

.30

0.2

80

.51

0.5

40

.83

0.4

50

.45

0.4

30

.79

0.7

30

.78

1.0

41

.26

1.2

80

.52

0.4

50

.51

1.7

31

.71

0.6

1

Gd

1.6

61

.49

1.5

52

.16

2.2

43

.12

1.7

21

.70

1.7

22

.16

2.2

32

.39

2.9

43

.22

2.7

91

.89

1.4

11

.91

3.3

83

.45

1.8

2

Tb

0.3

90

.40

0.3

50

.50

0.5

40

.77

0.4

30

.41

0.4

10

.26

0.2

40

.26

0.3

80

.42

0.3

70

.44

0.3

70

.43

0.3

70

.35

0.2

1

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

0.9

00

.78

0.7

80

.93

1.0

11

.27

0.8

40

.81

0.9

00

.15

0.1

80

.17

0.2

80

.30

0.3

00

.74

0.7

10

.77

0.1

90

.18

0.1

3

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

Tex

ture

/ass

oci

ati

on

: (1

) e

clo

git

e;

(2)

grt

lh

erz

oli

te;

(3

) p

hl-

grt

we

hrl

ite

; (4

) lh

erz

oli

te;

(5)

ma

cro

cryst

. (<

) b

elo

w t

he

de

tect

ion

lim

it;

(—)

no

t d

ete

cte

d

Tab

le 5

. Nd

isot

opic

an

alys

es (

TIM

S)

Age

S

mN

d14

7 Sm

/144 N

d14

3 Nd

/144 N

d14

3 Nd

/144 N

d14

3 Nd

/144 N

d� �N

d�N

dT

DM

TC

HU

R

(Ma)

(pp

m)

(pp

m)

actu

alac

tual

CH

UR

init

ial

actu

alin

itia

lG

aG

a1

12

50

2.0

6

8.1

8

0.2

51

83

4

0.1

51

60

0

0.5

12

53

2

0.0

08

20

9

-0.2

30

.51

10

23

0

.51

12

88

-2

.15

.21

.25

0.3

6

21

25

02

.66

1

5.5

0

0.1

71

61

3

0.1

03

30

0

0.5

12

52

9

0.0

08

20

9

-0.4

70

.51

10

23

0

.51

16

81

-2

.11

2.9

0.7

20

.18

41

25

01

9.9

0

22

7.0

0

0.0

87

66

5

0.0

52

80

0

0.5

11

75

6

0.0

08

20

9

-0.7

30

.51

10

23

0

.51

13

23

-1

7.2

5.9

1.2

20

.93

56

50

1.0

4

6.8

0

0.1

52

04

7

0.0

91

50

0

0.5

12

78

1

0.0

04

26

0

-0.5

30

.51

18

00

0

.51

23

91

2

.81

1.5

0.3

5-0

.21

76

50

1.1

9

7.8

0

0.1

52

17

4

0.0

91

60

0

0.5

12

74

4

0.0

04

26

0

-0.5

30

.51

18

00

0

.51

23

54

2

.11

0.8

0.3

9-0

.15

86

50

1.8

6

9.6

0

0.1

93

75

0

0.1

16

60

0

0.5

12

77

1

0.0

04

26

0

-0.4

10

.51

18

00

0

.51

22

74

2

.69

.30

.45

-0.2

5

96

50

0.8

5

5.3

0

0.1

60

37

7

0.0

96

50

0

0.5

12

72

5

0.0

04

26

0

-0.5

10

.51

18

00

0

.51

23

14

1

.71

00

.43

-0.1

3

106

50

6.1

0

42

.30

0

.14

42

08

0

.08

68

00

0

.51

27

47

0

.00

42

60

-0

.56

0.5

11

80

0

0.5

12

37

7

2.1

11

.30

.38

-0.1

5

No.

Sm

/Nd

e��t

fSm

/Nd

Кимберлиты связанные со структурой Lucapa, Ангола

Documents

Transcript of Кимберлиты связанные со структурой Lucapa, Ангола