Bauxite Formation on Tertiary Sediments and Proterozoic Bedrock in Suriname

Bauxite formation on Tertiary sediments and Proterozoic bedrock in Suriname

Bauxietvorming op Tertiaire sedimenten en Proterozoïsche gesteenten in Suriname (met een samenvatting in het Nederlands)

Proefschrift

ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op woensdag 24 januari 2018 des middags te 2.30 uur

door

Dewany Alice Monsels Geboren op 27 mei 1979 te Paramaribo, Suriname

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 1 Promotoren: Prof. dr. M.J.R. Wortel Prof. dr. Th.E. Wong

Copromotor: Dr. M.J. van Bergen

This thesis was accomplished with financial support from the Suriname Environmental and Mining Foundation (SEMIF).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 2 Utrecht Studies in Earth Sciences 147

Bauxite formation on Tertiary sediments and Proterozoic bedrock in Suriname

Dewany A. Monsels

Utrecht 2018

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 3 Thesis assessment committee:

Prof. dr. P.L. de Boer Faculteit Geowetenschappen Universiteit Utrecht Utrecht, Nederland

Prof. dr. S.B. Kroonenberg Faculteit der Technologische Wetenschappen, Studierichting Delfstofproductie Anton de Kom Universiteit van Suriname Paramaribo, Suriname

Prof. dr. M. Lima da Costa Instituto de Geosciências Universidade Federal do Pará Belém, Brasil

Prof. dr. P.R.D. Mason Faculteit Geowetenschappen Universiteit Utrecht Utrecht, Nederland

Dr. H. Théveniaut Bureau de Recherches Géologiques et Minières Orléans, France

Utrecht Studies in Earth Sciences: 147

Printed in the Netherlands by Ipskamp Printing, Enschede.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 4 Table of Contents

Chapter 1 Introduction 7

Chapter 2 Bauxite deposits in Suriname: Geological context and resource development 27

Chapter 3 Trace-element analysis of bauxite using laser ablation-inductively coupled 47 plasma-mass spectrometry on lithium borate glass beads

Chapter 4 Bauxite formation on Proterozoic bedrock of Suriname 69

Chapter 5 Bauxite formation on Tertiary sediments in the coastal plain of Suriname 111

Chapter 6 Synopsis 165

Nederlandse samenvatting 171

Acknowledgements 175

About the Author 177

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 5 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 6 Chapter 1 Introduction

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 7 8 | Chapter 1

1.1 | Lateritic bauxites

Bauxite is the world’s main economic source of aluminium, the second most abundant metallic element in the Earth’s crust after silicon. Bauxite was discovered in 1821 by the French scientist Pierre Berthier (Berthier, 1821) who originally named the material “beauxite” after a hill (“Colline des Beaux”) near the community of Les Baux in the Alpilles Mountains (Bouches du Rhône) in Southern France where it was found. In 1861 the term was recast into “bauxite” (Bárdossy, 1997). The ore belongs to the family of lateritic rocks that are products of strong weathering, accompanied by significant chemical and mineralogical modification occurring at or near the surface (Figure 1.1a). Hence, the following definition of laterites (Schellmann, 1983) also applies to bauxites: “Laterites are products of intense subaerial rock weathering. They consist predominantly of mineral assemblages of goethite, hematite, aluminium hydroxides, kaolinite minerals and

quartz. The SiO2 : (Al2O3+Fe2O3) ratio of a laterite must be lower than that of the kaolinized parent rock in which all the alumina of the parent rock is present in the form of kaolinite, all the iron in the form of iron oxides, and which contains no more silica than fixed in the kaolinite plus the primary quartz. This definition includes all highly weathered materials, strongly depleted in silica and enriched in iron and alumina, regardless of their morphological and physical properties (fabric, colour, consistency, etc.)”.

Throughout this thesis we will use “laterite” loosely as a geologic term for a rock that experienced intense subaerial weathering, consists predominantly of goethite, hematite, aluminium

hydroxides, kaolinite and quartz, and has a SiO2/(Al2O3 + Fe2O3) ratio lower than that of the associated kaolinized parent rock. The term “bauxite” will be referred to using a combination of definitions (Bárdossy and Aleva, 1990; Aleva, 1994; Tardy, 1997): Bauxite is an Al-rich laterite characterized by a particular enrichment of free aluminium hydroxide minerals such as gibbsite, boehmite, diaspore and nordstrandite. It can form from many different parent rocks, which preferably have a high aluminium content and relatively low iron content. Bauxite is basically an economic term, as it is a raw material for the aluminium industry when its grade and other economic parameters allow profitable extraction.

Different classification schemes for bauxites have been developed over the years, based on factors such as chemical and mineralogical composition, age, genetic history, shape and location (Valeton, 1972; Bárdossy and Aleva, 1990; Patterson et al., 1997; Tardy, 1997; Bogatrev et al., 2009). Three main genetic types can be distinguished when considering mineralogy, chemistry and host-rock lithology (Bárdossy and Aleva, 1990): 1. Lateritic bauxites — residual deposits derived from underlying alumosilicate rocks. 2. Karst-bauxites — deposits overlying karstified surfaces of carbonate rocks. 3. Tikhvin-type bauxites — detrital deposits of eroded lateritic bauxites. Bogatyrev et al. (2009) classified this type as “sedimentary bauxites”.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 8 Introduction | 9

Approximately 88% of the global bauxite resources belong to the lateritic bauxites, 11.5% to the karst bauxites and approximately 0.5% to the Tikhvin-type deposits (Bárdossy and Aleva, 1990).

Figure 1.1 | (a) Ternary compositional diagram, illustrating the nomenclature of weathering products in humid tropical climates (modified after Valeton, 1972; Bárdossy and Aleva, 1990; Aleva, 1994); (b) Sketch of a typical bauxite-bearing laterite profile.

A typical complete weathering profile of lateritic bauxite starts with a weathered parent rock at the bottom (saprolite, also known as kaolin), grades into an Al-accumulation horizon (bauxite) and is topped by a hard iron-rich cap (duricrust) and soil (Figure 1.1b). Sometimes horizons are absent in weathering profiles because of erosion of elevated areas in a landscape. Transition zones may separate horizons, but abrupt changes between bauxite and parent rock have also been described. The formation of lateritic bauxite requires a set of favourable conditions that include climate, time, hydrogeology, parent rock composition, geomorphology, petrophysics and biological factors (Aleva, 1984; Bárdossy and Aleva, 1990; Tardy, 1997; Bogatyrev et al., 2009 and references therein). Key requisites are: – Wet tropical and subtropical conditions with mean annual temperature above 20oC and rainfall above 1200 mm/yr, is the reason why many Cretaceous and Tertiary lateritic bauxite deposits in South America, Africa, India and Australia are found in (paleo-)coastal plains where humidity is usually higher than in continental interiors.

– Parent rock with suitable composition (e.g., Al2O3-content ≥ 10%, SiO2-content is variable but preferably quartz-free) and sufficient porosity and permeability to allow fluids to percolate, leach out and transport dissolvable components.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 9 10 | Chapter 1

Figure 1.2 | Relationship between bauxitization, climate conditions and sea-level changes during Phanerozoic times (modified after Bogatyrev et al., 2009).

Figure 1.3 | Average Al2O3 -content and reserves of important lateritic bauxite deposits worldwide. Surinamese deposits are indicated by circular symbols and italic labels. See Table 1.1 for details and references.*= resource.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 10 Introduction | 11

– Long-term tectonic stability with quiet vertical uplifts as the only disturbance, favouring the formation of an elevated fault-dissected landscape with well-drained flat areas such as plateaus or peneplains (planation surfaces), since mature bauxite deposits mainly develop when the chemical weathering rate is higher than the rate of mechanical erosion. Passive continental margins often fulfill this requirement. – Favourable pH and Eh gradients that allow fluid-rock interactions involving mineral dissolution and precipitation. Vegetation and metabolic products of micro-organisms may provide the acidity needed to accelerate the decomposition of aluminosilicate minerals (pH < 5), whereas new aluminium phases precipitate at pH 5.5–8.0. Likewise, the oxidation state of the system exerts a strong control on the (im-)mobility of Fe. 1 Certain periods in the Earth’s history were clearly more favourable for bauxitization than others, as is demonstrated by the strong link with climate conditions and regressions (Figure 1.2). A productive period was the Cretaceous–Eocene epoch when bauxite formed in Gondwana and Laurasia.

1.2 | Geomorphological setting of bauxite in the Guiana Shield region

The Precambrian core of the Guiana Shield is surrounded and unconformably overlain by Phanerozoic sedimentary units. During and following the Atlantic openingen, the craton and its margins have undergone slow vertical movements, which together with climate conditions, were major controls of landscape evolution and pervasive laterite weathering on both sides of the Central Atlantic Ocean (Peulvast and De Claudino Sales, 2004; Chardon et al., 2006). The combined effects of epeirogenic uplift, climate change and denudation resulted in the formation of stepped planation surfaces in generally low-relief landscapes. Ages of the surfaces are difficult to determine and polycyclic evolution is common, which explains existing inaccuracies in the number of bauxite forming episodes and uncertainties in regional correlations. Table 1.1 summarizes the five planation levels distinguished in and around the Guiana Shield with distinct nomenclatures used in Suriname, Guyana, French Guiana, Venezuela and northern Brazil (Théveniaut and Freyssinet, 2002). The surfaces range in age from Late Cretaceous to Quaternary (King et al., 1964; McConnell, 1968; Bárdossy and Aleva, 1990; Théveniaut and Freyssinet, 2002; Bogatyrev and Zhukov, 2009). In general, their current elevations tend to drop from the core parts of the Guiana Shield towards the rims, i.e., in northern direction to the Atlantic Coast and in southern direction to the Amazon Basin. Associated bauxite deposits have been distinguished in the (1) Guiana Shield, (2) Coastal Plain and (3) Amazon Basin Subprovinces of the South American Platform Province (Bárdossy and Aleva, 1990) (Figure 1.4). Stratigraphic and petrologic similarities and differences between bauxite deposits of the Guiana Coastal Plain and the Lower Amazon Basin have been discussed by Grubb (1979) and Aleva (1981). The principal bauxite-covered horizon of roughly Paleocene-Eocene age developed on Precambrian crystalline basement rocks at relatively high altitudes (up to 750 m) and on Late Cretaceous–Early Tertiary sedimentary parent material in peripheral areas at lower levels (less than 100 m). This horizon has been recognized in the five countries mentioned earlier. Bauxites in coastal areas are often buried by younger sediments as a result of marine transgressions and/ or tectonic movements.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 11 12 | Chapter 1

Early Velhas level Early Velhas 5-100 m (Amapá) Aporema Aporema 5-100 m (Amapá) hills (laterite) and Paradão Late Velhas level Velhas Late 5-70 m (Amapá) Pararguaçu level / Amazon / Amazon level Pararguaçu Floodplain 5-50 m King, 1962; Bárdossy & Aleva, & Aleva, 1962; Bárdossy King, 1990; Carvalho 1997; et al., & Roquin Tardy Melfi, 1997; 2002; et al., 1998; Montes 2006; da Horbe & Peixoto, et 2009; Leonardi et al., Costa 2014 et al., 2011; Costa al., BRAZIL (North+ Amazone)

Intermediate Intermediate level 400-500 m Guiacas Los Llos Llanos Llos Llanos level 80-150 m Middle Caroni level 200-350 m Orinoco Plain Plain Orinoco 0-50 m Bárdossy Bárdossy 1990 & Aleva,

Kaeiteur laterite 250-300 m 400-600 m Kuyunini 230-260 m Rupunini Rupunini laterite 100-180 m North & Central Savanna Mazaruni Mazaruni 75-90 m Bárdossy Bárdossy 1990 & Aleva, GUYANA VENEZUELA

Late Velhas Late Kaw Mt. surface) (lower Unit 2 Cayenne

Théveniaut & & Théveniaut Freyssinet, 2002

Deuxième Péné- plaine / Surface II 210-370 m Tortue, Mt. Plomb, Mt. Trios Piton Mt. KawMt. Troisième Troisième / Pénéplaine Surface III 210-260 m 250 m Forgassié Mt. Gabrielle Mt. 240 m MatouryGrand 230 m Terrases fluviatiles fluviatiles Terrases 0-60 m péné-Quatrième plaine 150-170 m Mahury 150 m Choubert, 1957; Blancaneaux, 1981 FRENCH GUIANA

Late Velhas Velhas Late Moengo 3, (Adjoema 1) Lobato

Freyssinet, 2002 Freyssinet,

First Late Tertiary First Tertiary Late surface 300 m Buried by Coesewijne Formation Second Late Late Second surface Tertiary floors Valley between Lelydorp- Onverdacht

Pollack, 1983Pollack, & Théveniaut

Foothill level level Foothill 390 m Mts. Lely Bakhuis Mts. 200 m Kwakoegron 280 m Pediplain level level Pediplain 50-150 m Valley floor level Valley Aleva, 1984; 1984; Aleva, Janssen, 1970, 1979 SURINAME Overview of planation surfaces and geomorphological correlations levels in the countries of the Guiana Shield (modified after Théveniaut and Freyssinet, 2002). Freyssinet, and of the Guiana Shield (modified Théveniaut afterOverview in the countries levels correlations of planation surfaces and geomorphological Lower Miocene + Oligocene hiatus Miocene Quarter- nary

Table 1.1 | Table Age

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 12 Introduction | 13

Gondwana surface Sul-Americano / South / South Sul-Americano peneplanation American surface 200 m Mts (Amapá) Tartarugal Rio Jari (kaolin) 200 m 240-320 m Pitinga 160-190 m trombetas Porto Juruti 100-170 m 375 m Paragominas 650 m Carajas BRAZIL (North+ Amazone)

Nuria level Nuria level 600-700 m Bolivar Cerro

Kanuku laterite Kanuku laterite 950 m surface” “Old Kopinang Kopinang bauxite 650-750 m Pakaraima Orenoque Mts. laterite 275-305 m GUYANA VENEZUELA

Sul-Americano Kaw Mt. (upper surface) Unit 1 Cayenne

Première Péné- Première plaine / Surface I 500-800 m Massif Trinité 350 m Mt. Lucifer Mt. Decou-Decou Atachi-Baka Mt. GalbaoMt. FRENCH GUIANA

Sul-Americano Sul-Americano Bakhuis Mts. 10) (Area

Gondwana Gondwana surface C1-C2 1000 m Bakhuis C1 Mts. Juliana Top Juliana Top (Wilhelmina Mts.) Early Tertiary Early Tertiary surface 500 m Bakhuis Mts. Brownsberg Nassau Mts. Mts. Lely Moengo Hills Onverdacht

Summit level Summit level > 1000 m Main bauxite Main bauxite level 690 m Mts. Lely Nassau 570 m Brownsberg 508 m Mtn. Wintiwaai 480 m Bakhuis Mts. 400 m Jones Hill 100 m Onverdacht 0-30 m SURINAME

Upper Upper Cretaceous Eocene Paleocene- Table 1.1 | Continued Table Age

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 13 14 | Chapter 1

Table 1.2 | Al2O3 grade, reserves and parent rocks of some lateritic bauxite deposits worldwide.

Symbol Country Reserve AL2O3 Parent rock References (MT) (%) Nassau NAS Suriname 32 42 Greenschist Bárdossy and Aleva, 1990; Bauxtite Institute, 2009; Monsels, 2016; Monsels and van Bergen, 2017

Bakhuis BAK Suriname 70 43.8 Various meta- Bárdossy and Aleva, 1990; morphic rocks Monsels, 2016; Monsels and van Bergen, 2017

Paranam, PAR Suriname 100 59.5 Sedimentary Bárdossy and Aleva, 1990 Onverdacht rock

Moengo, Ricanau MGO Suriname 127 54.2 Sedimentary rock Bárdossy and Aleva, 1990

Coermotibo CRB Suriname 18 51.3 Sedimentary rock Bauxite Institute, 2009

Linden, Berbice LIN Guyana 524 55 Sedimentary rock Bárdossy and Aleva,1990

Porto Trombetas TRO Brazil 955 47.5 Sedimentary rock Melfi, 1997

Paragominas PGM Brazil 1249 48.5 Sedimentary rock Bárdossy and Aleva, 1990; Bogatyrev and Zhukov, 2009

Rondon do Para RON Brazil 642 42.7 Sedimentary rock de Oliveira et al., 2016

Pocos de Calda POC Brazil 70 54 Syenite Bárdossy and Aleva, 1990

Los Pijiguaos PGU Venezuela 570 44 Rapakivi Granite Meyer et al., 2002

Boké (Sangaredi) SAN Guinea 990* 50.1 Reworked bauxite, Patterson et al., 1986; shale, siltstone Abzalov, 2016.

Fria-Kimbo FRI Guinea 363 42.5 Slate and clastic Bárdossy and Aleva, 1990 material

Kindia KIN Guinea 72 48 Slate Bárdossy and Aleva, 1990

Nyinahin NYH Ghana 730 47.5 Schist, green-stone, Bárdossy and Aleva, 1990; greywacke Amegashie, 2015

Panchpatmali PPM India 316 46.1 Khondalite Bogatyrev and Zhukov, 2009

Tayan TAY Indonesia 271 30.4 Monzonite, Bárdossy and Aleva, 1990 monzodiorite

Weipa WEI Australia 1922 55.1 Sand-, clay- and Eggleton et al., 2008; siltstone, schist Bogatyrev and Zhukov, 2009

Worsley WOR Australia 973 31.9 Granites (85%) Bárdossy and Aleva, 1990; and dolerite dikes Bogatyrev and Zhukov, 2009

Gove GOV Australia 252 50.4 Sandstone Bárdossy and Aleva, 1990

*= resource estimation.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 14 Introduction | 15

1.3 | Bauxite deposits and mining in northern South America

Important bauxite deposits of the Guiana Shield and surroundings are listed in Table 1.2 (see also Figures 1.3–1.5). Deposits with a significant mining history in Suriname and Guyana belong to the so-called “bauxite belt” in the lowlands, which runs subparallel to the “Old Coastal Plain”, an accumulation of continental sediments that were deposited along a paleo-coastline in Early Cenozoic times (Montagne, 1964; Valeton, 1983; Krook, 1979; Wong, 1989; Aleva and Wong, 1998). These lowland bauxites developed on unconsolidated Tertiary sediments, although in Guyana some may be underlain by Precambrian basement rocks (Van Kersen, 1956; Grubb, 1979; Patterson et al., 1986 and references therein). Suriname and Guyana have the longest 1

exploitation history of high-Al2O3 bauxite (Figure 1.3) in the region, and were together the world’s leading producer in the mid-20th century (Patterson et al., 1986; Bárdossy and Aleva, 1990; Aleva and Wong, 1998). Guyana was the first country outside Europe and the U.S. to start commercial bauxite mining about a century ago. Some seven groups of (buried) deposits are geographically distributed along the coastal bauxite belt. Mining commenced in 1917 near MacKenzie (Linden) and operations were expanded to Ituni and Kwakwani during the Second World War and in the 1950s. Modest production is still ongoing. Extensive bauxitic and ferruginous laterite deposits are also found in the Pakaraima Mountains along the border with Venezuela, where they cover Precambrian crystalline lithologies (e.g., Patterson et al., 1986). In Suriname bauxite mining has been restricted to deposits covering Tertiary sediments in the coastal lowlands although considerable reserves have also been discovered as plateau- type bauxites in the highlands, which had formed at higher topographic levels on various Proterozoic metasedimentary and meta-igneous rocks of the Guiana Shield. Bauxite exploitation commenced around 1920 in the Moengo-Ricanau-Jones District in the eastern part of the coastal lowlands, which remained the country’s mining center until the 1940s. The installation of an alumina plant and aluminium smelter in Paranam, together with the construction of a hydro-electric dam near Afobaka for power supply in the early 1960s, ultimately gave Suriname an integrated bauxite-alumina-aluminium industry. From the 1940s to the 1990s the bauxite sector was the mainstay of the Surinamese economy. In recent years, bauxite mining gradually came to a halt. In 1999 the smelter closed and the refinery in 2015. Bauxite deposits of French Guiana have been documented in the vicinity of Cayenne as well as in the country’s interior. The best explored and economically most interesting deposits are those of the Montagnes de Kaw, which developed on Proterozoic metavolcanics (Patterson et al., 1986; Egal et al., 1994). Studies of O and H isotopes in lateritic profiles of French Guiana have been used to reconstruct paleo-weathering processes and associated climate conditions (Girard et al., 1997, 2000, 2002). On the northwestern margin of the Guiana Shield in Venezuela, the major Los Pijiguaos bauxite deposits are situated on dissected plateaus belonging to a 600–700 m high planation surface that possibly represent a Late Cretaceous–Early Tertiary erosion event (Soler and Lasaga, 2000; Meyer et al., 2002). The bauxite formed on the Parguaza granite, a Precambrian rapakivi- type granite batholith, and has been mined since 1987. Smaller deposits developed on igneous and metamorphic rocks elsewhere in the Guiana Shield and on sediments in the Orinoco Delta region (Patterson et al., 1986).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 15 16 | Chapter 1

Figure 1.4 | Tectonic map of South America showing the outcrops of the Guiana Shield, Central Brazil Shield, Atlantic Shield and their Phanerozoic cover (modified after Bárdossy and Aleva, 1990; Bogatyrev and Zhukov, 2009).

Owing to its widespread lateritic bauxite deposits, Brazil hosts one of the world’s largest reserves. In the Amazon region and adjacent areas, the world-class bauxite deposits of Trombetas, Juruti, Paragominas and Rondon do Pará developed on Upper Cretaceous to Paleogene siliciclastic sediments (Melfi, 1997; Boulangé and Carvalho; 1997; Kotschoubey et al., 1997; Lima da Costa et al., 2014; De Oliveira et al., 2016). Bauxite mining activities started in 1979 in the Trombetas district, in 2007 in the Paragominas district and in 2009 in the Juruti district (De Oliveira et al., 2016). Most of these deposits are overlain by yellow kaolinitic clay (Belterra), which can reach a thickness of 20 m (average thickness 9–10 m) (Bárdossy and Aleva, 1990). Bauxite and aluminous laterite deposits with high contents of guano-derived phosphate occur near the Atlantic Coast, east of Belém (Patterson et al., 1986). A small volume of north Brazilian bauxite formed on Precambrian crystalline rocks of the Guiana Shield (Melfi, 1997; Costa et al., 1997; Horbe and Anand, 2011).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 16 Introduction | 17

Figure 1.5 | Geological sketch map of the Guiana Shield with planation surfaces (modified after Théveniaut and Freyssinet, 2002). Planation names follow those listed for Suriname (Aleva, 1984) in Table 1.1. Labeled arrows point to the areas investigated in this thesis.

1.4 | Previous research on Surinamese bauxite

The bauxites of Suriname have received considerable attention in the past. A comprehensive overview of previous research is given in Aleva and Wong (1998), mentioning that the first descriptions of what presumably were bauxite-topped hills, dates back to the late 17th century. Interest in Suriname’s bauxite started in the late 19th century when studies were conducted by Martin (1888), Du Bois (1901, 1903) and Van Bemmelen (1903, 1904) (references in Aleva and Wong, 1998). This early work mainly concerned petrographic descriptions. Du Bois (1903) and Van Bemmelen (1903) most likely reported the first chemical analysis of Surinamese bauxite. IJzerman (1931) was the first to distinguish “plateau type bauxite” and “lowland type bauxite”, based on their geomorphological positions, and was also the first to report on heavy minerals in bauxite. His supposition that the Nassau Mountains were covered with bauxite has proven to be correct, whereas his attempt to explain the genesis and age of the bauxite deposits suffered from a lack of relevant data. Since then, many workers focused on petrographic descriptions, the stratigraphic position and the formation history of Suriname’s bauxite deposits (Bakker et al., 1953; Van Kersen, 1956; Moses and Mitchell, 1963; Montagne, 1964; Aleva, 1965, 1979, 1981, 1984; Krook, 1969a, 1969b, 1975, 1994; Valeton, 1971, 1983; Pollack, 1981, 1983). Among the unsolved issues, much uncertainty still exists on the nature of the parent rock of the Coastal- plain bauxites and their genetic history.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 17 18 | Chapter 1

Because research in Suriname was mostly connected with exploration of bauxite from an economic perspective, studies of deposits often dealt with the mineralogy, size, shape, grade or reserve estimation (Doeve, 1955; Janssen, 1970, 1979; Pollack, 1981; Aleva and Hilversum, 1984). The Surinamese bauxites are found in several international classifications as the ”Suriname type“ or “trihydrate bauxite” that is gibbsite-rich with < 3% boehmite (Patterson et al., 1986; Hill and Robson, 2016). The mineralogy is an important aspect, as significant amounts of Al-goethite and boehmite would result in substantial Al loss during refining of bauxite in a Bayer process at low temperature. High sulphur contents can also be an adverse property, as is the case at the Coermotibo deposit in the Moengo area. Surinamese lateritic material has also been used to develop international standard reference materials intended for use in evaluating chemical and instrumental methods of bauxite analyses. The BAK-1 standard, assembled by the Geological Survey of Suriname, is a composite

of four lateritic bauxites with different Al2O3 concentrations from the Bakhuis Mountains (Burke, 1985; LaBrecque and Schorin, 1992), the SLB-1 standard is a subportion of the BAK-1 sample (LaBrecque, 1990), and the NIST-SRM 696, a bauxite from Suriname provided by Alcoa, has been certified by the US National Institute of Standards and Technology (Reed, 1991). Their major and trace-element compositions have been determined with different analytical techniques for various purposes such as comparison with Caribbean karst bauxites (Roelandts, 1989; Potts and Rogers, 1991; Grant et al., 2005). Topp et al. (1984) and Logemerac (1969) examined the distribution of trace elements in lateritic profiles and red mud. Geomorphological aspects relevant for the origin of laterites and bauxites have been treated by King et al. (1964), Zonneveld (1969, 1982) and Kroonenberg and Melitz (1983). Information on ages of bauxites has been obtained from stratigraphic dating, based on the analysis of pollen from overlying and underlying sediments (Van der Hammen and Wijmstra, 1964; Wijmstra and Van der Hammen, 1964), and from paleomagnetic constraints for the Bakhuis and Moengo deposits (Théveniaut and Freyssinet, 2002). Diko et al. (2001) conducted 3D geostatistical modelling to quantify the spatial distribution of major elements in deposits near Moengo. The search for new undiscovered bauxite deposits was vigorous in the 1950s when more than 1,000 exploratory drill holes were drilled in the western part of Suriname’s coastal plain but the campaign only revealed a few shows (Van Lissa, 1975). Nevertheless, Patterson et al. (1986) did not exclude the presence of bauxite there, as the drill holes had not reached depths at which buried deposits in the eastern coastal areas had been found. Given the downwarp of Precambrian basement along the Atlantic Coast (Bárdossy and Aleva, 1990), it is conceivable that any covered bauxite deposits are deeper positioned.

1.5 | Research objectives and thesis outline

With few exceptions no academic geological studies of bauxites of Suriname have appeared since the early 1980s. Hence, despite the long preceding history of research on the country’s bauxite deposits, important questions as to their origin and evolution remain unsolved. Key issues concern the significance of the parent rock for the chemical, mineralogical and textural properties of bauxite-bearing weathering profiles, the extent to which element re-distributions, driven by leaching and re-precipitation, were influenced by consecutive or cyclic weathering

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 18 Introduction | 19

processes, and the timing, duration and climate conditions of distinct bauxitization events. Since these questions relate to the origin of bauxite as economic commodity in terms of quality, abundance and geographic distribution, they have direct relevance for future exploration strategies. The lateritic bauxites of Suriname have formed on precursor lithologies ranging from Precambrian low- to high-grade metamorphic igneous and sedimentary rocks to unconsolidated siliciclastic sediments and clays of Tertiary age. This wide diversity of precursor material and the notion that the bauxites commonly formed from in-situ weathering, are a unique starting point to investigate the influence of parent rock composition, relative to mineralogical and other controls of element mobility and fractionation, in creating the spectrum of bauxitic weathering 1 products as observed in the field today. From a petrological and geochemical perspective, this thesis will address the relationship between bedrock lithology, element mobility and properties of bauxite-bearing weathering profiles within the context of Suriname’s geological history. Chapter 2 introduces the principal bauxite deposits of Suriname, presents an overview of their geological setting, including a comparison with West African bauxites, and briefly summarizes the exploitation history and economic significance for the country.This chapter has been published as: Monsels, D.A. (2016). Bauxite deposits in Suriname: Geological context and resource development. Netherlands Journal of Geosciences 95, 405–418. Since this research has a strong focus on geochemistry, accurate determination of trace element concentrations is of critical importance. Bauxite material is difficult to process for this purpose, because rare earth and high field-strength elements are often hosted by accessory minerals that are difficult to dissolve. Through analytical work on international standard reference materials it is shown in Chapter 3 that laser ablation ICP-MS analysis of lithium-borate glass beads, similar to those used for X-ray fluorescence analysis, is a reliable and efficient method to obtain high-quality data for trace elements in bauxite. The results of this analytical part of the study have been submitted for publication in Geostandards and Geoanalytical Research as: Monsels, D.A., Van Bergen, M.J. and Mason, P.R.D. – Trace-element analysis of bauxite using laser-ablation inductively coupled plasma-mass spectrometry on lithium borate glass beads. Chapter 4 presents a detailed study of lateritic plateau (or highland) bauxites in Suriname. Textural, mineralogical and geochemical information, obtained from weathering profiles on Proterozoic crystalline basement rocks in different parts of the country, is used to demonstrate a conspicuous influence of bedrock lithology on the compositional properties of in-situ developed bauxite. These parent rocks range from (ultra)high-temperature metamorphic gneissic and amphibolitic rocks in the Granulite Belt in west Suriname to greenschist-facies meta-igneous rocks in the Greenstone Belt in east Suriname. It is shown that distinct trace element abundances are attributable not only to lithological differences, but also to the control exerted by accessory mineral phases on element mobility during weathering. This chapter has been published as: Monsels, D.A. and van Bergen, M.J. (2017). Bauxite formation on Proterozoic bedrock of Suriname. Journal of Geochemical Exploration 180, 71–90. Chapter 5 examines lateritic bauxite deposits that developed on Tertiary sediments in Suriname’s coastal lowland. Their textural, mineralogical and geochemical evolution is reconstructed based on field evidence and compositional data acquired from materials of former mining areas near Moengo and Onverdacht, with special attention to the recently abandoned Successor Mines also known as Successor deposits (Klaverblad, Kaaimangrasie,

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 19 20 | Chapter 1

Caramacca), where vertical sections were studied in detail. Using geochemical signatures it is demonstrated that the nature of siliciclastic sedimentary precursor material of the Coastal- plain bauxites is diverse and probably controlled by provenance and fluvial transport of eroded material from Precambrian rocks of the Guiana Shield. Compositional variations in profiles are not only attributable to weathering-induced re-distribution of chemical components, but also to original stratigraphic heterogeneity. Questions concerning the nature of the sedimentary precursor (e.g., arkosic sandstone versus kaolinitic clay) and trace element signatures are addressed. An adapted version of this chapter has been submitted for publication in Journal of South American Earth Sciences as: Monsels, D.A. and Van Bergen, M.J. – Bauxite formation on Tertiary sediments in the coastal plain of Suriname. Chapter 6 presents a synthesis of the results.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 20 Introduction | 21

References

– Ab’Sáber, A. (2000). Summit surfaces in Brazil. Revista Brasileira de Geociências 30 (3), 515-516. – Abzalov, M. (2016). Reconciliation with new data. Applied Mining Geology. Springer International Publishing, 315-320. – Aleva, G. (1965). The buried bauxite deposits of Onverdacht, Suriname, South America. Geologie en Mijnbouw 44, 45-58. – Aleva, G. (1979). Bauxitic and other duricrusts in Suriname. A review. Geologie en Mijnbouw 58, 321-336. – Aleva, G. (1981). Essential differences between the bauxite deposits along the southern and northern edges of the Guiana Shield, South America. Economic Geology 76, 1142-1152 1 – Aleva, G. (1984). Lateritization, bauxitization and cyclic landscape development in the Guiana Shield. In: Jabob Jr., L., ed., Bauxite [Bauxite Symposium, Los Angeles, California, 1984, Proceedings]; New York, American Institute of Mining, Metallurgical, and Petroleum Engineers, 297-318. – Aleva, G. (1994). Laterites. Concepts, geology, morphology and chemistry. ISRIC, Wageningen, 169 pp. – Aleva, G. and Wong, Th. (1998). The history of bauxite exploration and mining in Suriname. In: The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences, and Netherlands Institute of Applied Geoscience. TNO, 275-310. – Amegashie, A. (2015). Ghana: A developing bauxite source. 21st Bauxite and Alumina Conference, 24-26 February 2015, Miami, USA. – Bakker, J., Kiel, H. and Müller, H. (1953). Bauxite and sedimentation phases in the northern part of Surinam (Netherlands Guiana). Geologie en Mijnbouw 15, 215-226. – Bárdossy, G. (1997). Berthier, les Baux and the history of bauxite (Berthier, les Baux et l’histoire de la bauxite). Comptes Rendus de l’Académie de Sciences - Serie IIa: Sciences de la Terre et des Planetes 324 (12), 1031-1040. – Bárdossy, G. and Aleva, G. (1990). Lateritic bauxites. Developments in Economic Geology. Elsevier Science Publishing 27, 624 pp. – Bauxite Institute (2009). Suriname as a source of bauxite and alumina. In: Prospects for growth? 15th Bauxite and Alumina Seminar, February 2009, Montega Bay, Jamaica. – Berthier P. (1821). Analyse de l’alumine hydratée des Beaux, département des Bouches-du-Rhóne. Annales des mines, 1st series (6), 531-534. – Blancaneaux, P. (1981). Essai sur le milieu naturel de la Guyane Française. Travaux et Documents de ORSTOM 137, 126 pp. – Bogatyrev, B., Demina, V. and Zhukov, V. (1995). Influence of Partial pressure of CO2 on bauxite formation in the Cenozoic, Geol. Rudn. Mestorozhd 37, 3, 208-219. – Bogatyrev, B. and Zhukov, V. (2009). Bauxite provinces of the World. Geology of ore deposits 51 (5), 339-355. – Bogatyrev, B., Zhukov, V. and Tsekhovsky, Y. (2009). Formation conditions and regularities of the distribution of large and superlarge bauxite deposits. Lithology and Mineral Resources 44 (2), 135-151. – Burke, K.E. (1985). First co‐operative study on four tropical soil samples and one lateritic bauxite (Suriname) proposed as geochemical reference samples. Geostandards Newsletter 9 (1), 69-78. – Carvalho, A., Boulangé, B, Melfi, A. and Lucas, Y. (1997). Brazilian bauxites. Orstom, São Paulo-Paris, 331 pp. – Chardon, D., Chevilotte, V., Beauvais, A., Grandin, G. and Boulangé, B. (2006). Planation, bauxites and epirogeny; one or two paleosurfaces on the West African margin. Geomorphology 82, 273-282. – Choubert, B. (1957). Essai sur la géomorphologie de la Guyane. Mémoire, Carte géologique détaillée de la France: Département de la Guyane. Imprimerie Nationale, Paris, 43 pp. – Costa, M., Lemos, V. and Villas, R. (1997). The bauxite of the Carajás Mineral province. In: Carvalo, A., Boulangé, B., Melfi, A. and Lucas, Y. (eds.), Brazilian bauxites.USP/FAPESP/ORSTOM, São Paolo- Paris, 137-159.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 21 22 | Chapter 1

– Da Costa, M., Sousa, D. and Angélica, R. (2009). The contribution of lateritization processes to the formation of the kaolin deposits from eastern Amazon. Journal of South American Earth Sciences 27, 219-234. – Diko, L., Vervoort, A. and Vergauwen, I. (2001). Geostatistical modelling of lateritic bauxite orebodies in Suriname: effect of the vertical dimension. Journal of Geochemical Exploration 73, 131-153. – Doeve, G. (1955). De bauxietexploratie op het Nassaugebergte. Internal Geologische Mijnbouwkundige Dienst Suriname report, 68 pp. – Doeve, G. and Groeneveld Meijer, W. (1963). Bauxite deposits of Britsch Guiana and Suriname in relation to underlying unconsolidated sediments suggesting two-step origin. Economic Geology 58, 1060-1062. – Egal, E., Milési, J., Ledru, P., Cautru, J., Freyssinet, P., Thieblemont, D. and Vernhet, Y. (1994). Ressources minérales et évolution lithostructurale de la Guyane. Carte thématique minière à 1/100000. Feuille Cayenne. Rapport BRGM R38019, 59 pp. – Girard, J., Razanadranorosoa, D. and Freyssinet, P. (1997). Laser oxygen isotope analysis of weathering goethite from the lateritic profile of Yaou, French Guiana: paleoweathering and paleoclimatic implications. Applied Geochemistry 12, 163-174. – Girard, J., Freyssinet, P. and Chazot, G. (2000). Unraveling climatic changes from intraprofile variation in oxygen and hydrogen isotopic composition of goethite and kaolinite in laterites: An integrated study from Yaou, French Guiana. Geochimica et Cosmochimica Acta 64, 409- 426. – Girard, J., Freyssinet, P. and Morillon, A. (2002). Oxygen isotope study of Cayenne duricrust paleosurfaces: implications for past climate and laterization processes over French Guiana. Chemical Geology 191, 329– 343. – Grant, C., Lalor, G., and Vutchkov, M. (2005). Comparison of bauxites from Jamaica, the Dominican Republic and Suriname. Journal of Radioanalytical and Nuclear Chemistry 266 (3), 385-388. – Grubb, P. (1979). Genesis of bauxite deposits in the Lower Amazon Basin and Guianas Coastal Plain. Economic Geology 74, 735-750. – Hill, V. and Robson, R. (2016). The classification of bauxites from the Bayer plant standpoint. In: Essential readings in light metals, Springer International Publishing, 30-36. – Horbe, A. and Peixoto, S., (2006). Geochemistry of Pitinga bauxite deposit- Amazonian region- Brazil. In Regolith-Consolidation and dispersion of ideas. In: The CRC Leme Regolith Symposium, Hahndorf Resort, South Australia, Proceedings, 144-146. – Horbe, A. and Anand, R. (2011). Bauxite on igneous rocks from Amazonia and Southwestern of Australia: Implication for weathering process. Journal of Geochemical exploration 111 (1-2), 1-12. – Ijzerman, R. (1931). Outline of the geology and petrology of Suriname (Dutch Guiana). Thesis Utrecht, Kemick & Zn. (Utrecht)/ Martinus Nijhoff (The Hague), 519 pp. – Janssen, J. (1970). Preliminary report on the Coermotibo Exploration. Internal Suralco L.L.C. report, Paramaribo, Suriname, 1-8. – Janssen, J. (1979). Bauxite and laterite hard caps in Suriname. Unpublished internal Grassalco report, 1-12. – King, L., Hobday, D. and Mellody, M. (1964). Cyclic denudation in Surinam. Internal Geologisch Mijnbouwkundige Dienst Suriname report, 12 pp. – Klaver, M., De Roever, E., Nanne, J., Mason, P. and Davies G. (2015). Charnockites and UHT metamorphism in the Bakhuis Granulite Belt, western Suriname; Evidence for two separate UHT events. Precambrian Research 262, 1-19. – Kotschoubey, B., Truckenbrodt, B. and Hieronymus, B. (1997). Bauxite deposits of Paragominas. In: Carvalho, A., Boulangé, B., Melfi, A. and Lucas Y. (eds.) Brazilian bauxites: São Paulo, Universidade de São Paulo, FAPESP- ORSTOM, 75-119. – Kroonenberg, S. and Melitz, P. (1983). Summit levels, bedrock control and the etchplain concept in the basement of Suriname. In: Van den Berg, M. and Felix, R. (eds): Special issue in the honor of J. De Jong, Geologie en Mijnbouw 62, 389-399. c `1 – Krook, L. (1969a). Investigations on the mineralogical composition of the Tertairy and Quaternary sands in northern Suriname. In: Proc.7th Guiana. Geol.Conf., Paramaribo, 1966. Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootschap 27, 89-100.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 22 Introduction | 23

– Krook, L. (1969b). The origin of bauxite in the coastal plain of Surinam and Guyana. Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo 20, 173-180. – Krook, L. (1979). Sediment petrographical studies in northern Suriname. Doctoral thesis, Free University of Amsterdam, 155 pp. – Krook, L. (1994). De geologische en geomorfologische ontwikkeling van Noord Suriname. In: van der Steen, L. (ed.), Recente Geologische en Mijnbouwkundige Ontwikkelingen in Suriname. Publ.Found. Sci.Res., Caribbean Region, 23-40. – Krook, L. and De Roever, E. (1975). Some aspects of bauxite formation in the Bakhuis Mountains, western Suriname. Anais 10a Conferencia Geologica Interguianas, Bélem 1, 686-695. – LaBrecque, J. (1990). The comparison of the results of two independent intercomparison studies (BAK-1 and SLB-1) from the same bulk material of a lateritic soil. Fresenius’Journal of Analytical Chemistry 338 (4), 498-500. 1 – LaBrecque, J., and Schorin, H. (1992). Certification of SLB-1, a Suriname lateritic bauxite material as a standard reference material. Fresenius’ Journal of Analytical Chemistry 342 (4-5), 306-311. – Leonardi, F., Ladaira, F. and dos Santos, M. (2011). Paleosurfaces and bauxite profiles in the Poços de Caldas Plateau, São Paulo/ Minas Gerais, Brazil. Geociênces 30 (2) 147-160 – Lima da Costa, M., da Silva Cruz, G., de Almeida, H. and Poellmann, H. (2014). On the geology, mineralogy and geochemistry of the bauxite-bearing regolith in the lower Amazon Basin: evidence of genetic relationships. Journal of Geochemical exploration 146, 58-74. – Logemerac, V. (1969). The distribution of rare earths and other minor elements in Surinam bauxite and laterite and in the red mud obtained from them. Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootschap 27, 155-162. – Mc Connell, R. (1968). Planation surfaces in Guyana. The Geographical Journal 134, 506-520. – Melfi, A. (1997). Brazilian bauxite deposits: A review. In: Carvalho, A., Boulangé, B., Melfi, A.,and Lucas Y. (eds.) Brazilian bauxites. São Paulo, Universidade de São Paulo,FAPESP-ORSTOM, 3-22. – Meyer, F., Happel, U., Hausberg, J., and Wiechowski, A. (2002). The geometry and anatomy of the Los Pijiguaos bauxite deposit, Venezuela. Ore Geology Reviews 20, 27-54. – Moses, J. and Michell, W. (1963). Bauxite deposits of British Guiana and Suriname in relation to underlying unconsolidated sediments suggesting two-step origin. Economic Geology 58 (2), 250-262. – Monsels, D.A (2016). Bauxite deposits in Suriname: Geological context and resource development. Netherlands Journal of Geosciences, Geologie en Mijnbouw 95 (4), 405-418. – Monsels, D.A and Van Bergen, M.J. (2017). Bauxite formation on Proterozoic bedrock in Suriname. Journal of Geochemical Exploration 180, 71-90. – Montage, D. (1964). New facts on the geology of the ‘Young’ unconsolidated sediments in northern Suriname. Geologie en Mijnbouw 43, 499-515. – Montes, C., Melfi, A., Carvalho, A., Veira-Coelho, A. and Formoso, M. (2002). Genesis, mineralogy and geochemistry of kaolin deposits of the Jari River, Amapá State, Brazil. Clays and Clay Minerals 50 (4), 494-503. – Patterson, S., Kurtz, H., Olson, J. and Neeley, C. (1986). World Bauxite Resources; Geology and resources of aluminum. U.S. Geological Survey professional paper, 1076-B, United States Government Printing Office, Washinghton. – Peulvast, J.P. and De Claudio Sales, V. (2004). Stepped surfaces and palaeolandforms in the northern Brazilian “Nordeste”: constrains on models of morphotectonic evolution. Geomorphology 62 (1), 89-122. – Pollack, H. (1981). Bauxites and laterites of the Bakhuis Mountain Zone, western Suriname; a general description with emphasis on geomorphology and chemistry. In: Lateritization processes, Proc. Int. seminar on lateritization processes, December 1979, Oxford and IBH Publishing Company, New Dehli, 270-268. – Pollack, H. (1983). Land surfaces and lateritization in Suriname. In: Melfi, A., and Carvalho, A. (eds.) Proc. 2nd Int. Sem. on Lateritisation Processes, 1982, São Paulo, Brazil, 295-308. – Potts, P. and Rogers, N. (1991). Determination of trace elements in selected geological reference materials by instrumental neutron activation analysis. Geostandards Newsletter 15 (1), 111-116.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 23 24 | Chapter 1

– Reed, W. (1991). Certificate of NIST-SRM 696 Bauxite. National Institute of Standards and Technology, U.S. Department of Commerce. – Roelandts, I. (1989). Geological reference materials. Spectrochimica Acta Part B: Atomic Spectroscopy 44 (1), 5-29. – Soler, J. and Lasaga, A. (2000). The Los Pijiguaos bauxite deposit (Venezuela): A compilation of field data and implications for the bauxitization process. Journal of South American Earth Sciences 13 (1-2), 47-65. – Tardy, Y. (1997). Petrology of laterites and tropical soils, A.A. Balkema publishers, Rotterdam, Brookfield, 377 pp. – Topp, S., Salbu, B., Roaldset, E., Jørgensen, P., (1984/1985). Vertical distribution of trace elements in laterite soil (Suriname). Chemical Geology 4 (15), 9-174. – Théveniaut, H. and Freyssinet, P. (2002). Timing of lateritization on the Guiana Shield: synthesis of paleomagnetic results from French Guiana and Suriname. Palaeography, Palaeoclimatology, Palaeoecology 178, 91-117. – Valeton, I. (1971). Tubular fossils in the bauxites and the underlying sediments of Suriname and Guyana. Geologie en Mijnbouw 50 (6), 733-741. – Valeton, I. (1972). Bauxites. Developments in Soil Science 1, Elsevier, Amsterdam, 244 pp. – Valeton, I. (1983). Palaeoenvironment of lateritic bauxites with vertical and lateral differentiation. Geological Society, London, Special Publications 11 (1), 77-90. – Van der Hammen, T. and Wijmstra, T. (1964). A palynological study on the Tertiary and the Upper Cretaceous of British Guiana. Leidse Geologische Mededelingen 30, 183-241. – Van der Laan, S. (1998). Mineralogy and textures of Suriname bauxites. Internal Billiton BHP report. Hoogovens Research and Development, IJmuiden, 1-15. – Van Kersen, J. (1956). Bauxite deposits in Suriname and Demerara (British Guiana), Thesis Leiden; also published in Leidse Geologische Mededelingen 21, 247-375. – Van Lissa, R. (1975). Review of bauxite exploration in the coastal plain of Suriname. In: Contributions to the geology of Suriname 4. Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo 23, 250- 259. – Wijmstra, T. and Van der Hammen,T. (1964). Palynological data on the age of the bauxite in British Guiana and Surinam. Geologie en Mijnbouw 43, 143. – Wong, Th. (1989). Revision of the stratigraphy of the coastal plain of Suriname. Mededelingen Natuurweten- schappelijke Studiekring Suriname en Nederlandse Antillen 123, 64 pp. – Zonneveld, J. (1969). Preliminary remarks on summit levels and the evolution of the relief in Surinam. In: Proc.7th Guiana Geol. Conf., Paramaribo, 1966. Verhandelingen van het Koninklijk Nederlands Geologisch, Mijnbouwkundig Genootschap 27, 53-60. – Zonneveld, J. (1982). Summit levels in Suriname. ITC Journal 3, 237-242.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 24 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 25 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 26 Chapter 2

Bauxite deposits in Suriname: Geological context and resource development

This chapter has been published as: Monsels, D.A. (2016). Bauxite deposits in Suriname: Geological context and resource development. Netherlands Journal of Geosciences 95, 405-418.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 27 28 | Chapter 2

Abstract

Bauxite, the raw material of aluminium, has been one of the economically vital natural resources for Suriname. Mining operations started about a century ago, and subsequent development of a refinery industry and hydro-electric power made Suriname one of the foremost bauxite and alumina producers worldwide for a long period. This paper presents a concise survey of main geological attributes of its bauxite deposits and examines significant aspects in the development of mining in the country where alumina has dominated the export revenues until a decade ago. The lateritic bauxite deposits are spread across the northern part of the country and developed on various parent rocks during Late Cretaceous–Early Tertiary times. Bauxites in the coastal lowlands formed on Cenozoic sedimentary deposits, whereas plateau bauxites originated on various crystalline rocks in inland regions of the Precambrian Guiana Shield. The composition of parent rocks and timing of bauxitization point to a genetic correspondence with West African bauxites, and a strong control of paleoclimate conditions on the distribution and properties of bauxite in both regions. The more accessible bauxite deposits in the coastal lowlands have been mined out, whereas the plateaus bauxites have been extensively explored but have not been brought into production to date. For economic and environmental reasons, the future of the bauxite industry in Suriname is currently uncertain.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 28 Bauxite deposits in Suriname: Geological context and resource development | 29

2.1 | Introduction

Suriname has been one of the leading bauxite and alumina producers in the world for over more than 90 years (Patterson et al., 1986; Aleva and Wong, 1998; Gurmendi et al., 2012). Its lateritic bauxite deposits are distributed in two major geographic areas with different origins, properties and exploitation histories. Coastal-plain bauxites formed on sedimentary parent rocks in the coastal zone and have been mined since the early 20th century, whereas Plateau bauxites originated on metamorphic rocks in interior parts of the country and have not been productive to date. Bauxite deposits of economic interest occur in four bauxite districts (Bárdossy and Aleva, 1990) (Figure 2.1): 1. Paranam-Onverdacht-Lelydorp District (coastal-plain) which includes the Lelydorp-1, Kankantrie Noord, and Para Noord bauxite deposits. 2. Moengo-Ricanau-Jones District (coastal-plain) with the Coermotibo bauxite deposit. 3. Bakhuis District (plateau) with the Bakhuis bauxite deposit. 4. Nassau District (plateau), which encompasses the Nassau, Brownsberg and Lely bauxite 2 deposits and the Wintiwaai laterite deposit.

A detailed history of Suriname’s bauxite can be found in Aleva and Wong (1998), as well as Pollack (2016). The first academic descriptions of bauxite in Suriname were written at the end of the 19th century. The ‘Surinaamsche Bauxiet Maatschappij’ (SBM) was established in 1916 after discovery of the Moengo bauxite hills in 1915 by the Pittsburg Reduction Company, later renamed as Aluminum Company of America (Alcoa). In 1939, the N.V. Billiton Maatschappij obtained a concession for bauxite exploration in the Para district which led to the discovery of the Onverdacht bauxite deposit. Suriname was one of the leading bauxite producing countries in the world during the Second World War (1939–1945), which generated a boost in bauxite exploration and research on the local deposits (Patterson et al., 1986) (Figure 2.2). In 1941, the SBM opened the Paranam bauxite processing plant, named after the Para and Suriname Rivers bordering the mining concession areas. In 1958, SBM signed the Brokopondo Agreement with the Surinamese Government to create a fully integrated aluminium industry in the country, and became the Suriname Aluminum Company (Suralco L.L.C.). The most important objectives of this agreement were the construction of a dam in the Suriname River and a hydro-electric power facility at Afobaka, the establishment of an aluminium smelter and alumina refinery at Paranam. In 1993, Suralco L.L.C. and BHP Billiton signed a joint venture about mining and refining activities but BHP Billiton halted its exploration and mining activities in Suriname in 2009, while Suralco L.L.C. ceased all of its activities in 2016. This chapter presents a geological overview of Suriname’s bauxite deposits, including a comparison with West African bauxites, and a summary of the exploitation history and economic impact for the country.

2.2 | General geological background of Surinamese bauxite deposits

Suriname is located on the Guiana Shield in the northeastern sector of South America. Precambrian rocks make up about 80% of the country, forming a crystalline basement, and the remaining 20% consists of Cretaceous to Recent sediments that were deposited along the northern fringe of the Guiana Shield also known as the Guiana Coastal Plain (Figure 2.3).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 29 30 | Chapter 2

Figure 2.1 | Bauxite districts of the Coastal-plain and Plateau bauxite deposits in Suriname. *=mined out.

Figure 2.2 | Cumulative bauxite production (Mt) between 1935 and 1980. Note the peak during the Second World War (WW II) (modified after Patterson et al., 1986).

The bauxite deposits in South America are situated in four subprovinces of the Amazon Platform. The Guiana Shield subprovince includes bauxite deposits of Venezuela, Guyana, Suriname, French Guiana and Brazil (Figure 2.4a). The Coastal Plain subprovince contains bauxite deposits of Suriname and Guyana, while Brazilian bauxite deposits occur in the Guiana Shield, Amazon Basin and Central Brazil subprovinces (Chapter 1). Five planation levels can be distinguished on the Guiana Shield (King et al., 1964; Bárdossy and Aleva, 1990): (1) Summit Level, Jurassic to Cretaceous; (2) Main Aluminous Laterite Level, Early Tertiary; (3) Foothill Level, Oligocene to Early Miocene; (4) Pediplane Level, Pliocene; (5) Valley Floor Level, Pleistocene to Recent (Chapter 1 and Figure 2.4b). The major bauxite and aluminous laterite horizons on the Guiana Shield are related to the well-developed and widespread Main

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 30 Bauxite deposits in Suriname: Geological context and resource development | 31

Figure 2.3 | Geological map of Suriname with boundary line between the sedimentary coastal area to the north and the crystalline basement to the south (modified after www.staatsolie.com and Kroonenberg et al., 2016).

Aluminous Laterite Level (Van Kersen, 1956; Wong, 1989; Aleva and Wong, 1998). The age of this palaeosurface is well-documented by Paleocene pollen from unconsolidated sediments (arkosic sands or kaolin) below, and Miocene sediments on the top of some of the major economic bauxite deposits in Guyana and Suriname (Van der Hammen and Wijmstra, 1964). A long period of non-deposition during the Late Eocene to Oligocene is known as the ‘Bauxite Hiatus’, during which intense weathering resulted in bauxitization of the upper part of the Onverdacht Formation (Aleva and Wong, 1998; Wong et al., 2009). In Suriname, bauxite deposits formed on two different types of parent rock (Figure 2.1): 1. Sedimentary parent rocks in the coastal area (coastal-plain or lowland bauxites). The coastal-plain bauxites formed at the expense of Cenozoic sediments in the “Bauxite Belt” running subparallel to the Old Coastal Plain, an accumulation of continental sediments that were deposited along a paleo-coastline during Early Cenozoic times (Valeton, 1983; Aleva and Wong, 1998). As such, they belong to the global elongate belts of lateritic bauxite deposits in Cretaceous and Tertiary coastal plains, following Lower Tertiary shorelines of India and South America (Valeton, 1983; Aleva and Wong, 1998). The Surinamese Coastal-

plain bauxites are relatively Fe2O3- and SiO2-poor (2% and 6% respectively), while they are

Al2O3-rich (> 50%) with an average thickness of 6 m. 2. Crystalline parent rocks in the hinterland (plateau or highland bauxites). The plateau bauxites are mostly developed on intermediate to basic Precambrian igneous or metamorphic rocks of the Guiana Shield (Van Kersen, 1956; Bárdossy and Aleva, 1990). They occur on plateaus (250–650 m above mean sea level) and have an average thickness

of 4 m with little or no overburden (< 1 m). The deposits are relatively Fe2O3-rich (15–20%),

SiO2-poor (2–4%), with an average available alumina content of 45%.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 31 32 | Chapter 2

Figure 2.4 | (a) Overview of the most important bauxite deposits of the Guiana Shield (Guiana Shield subprovince) with its Phanerozoic cover (Coastal Plain subprovince) (modified after Bárdossy and Aleva, 1990); (b) The planation levels, stratigraphy and pollen zones of Suriname (Van der Hammen and Wijmstra, 1964; Wong, 1989: Bárdossy and Aleva, 1990; Wong et al., 2009)

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 32 Bauxite deposits in Suriname: Geological context and resource development | 33

2.3 | Coastal-plain deposits

2.3.1 | Coermotibo bauxite deposit The Coermotibo bauxite deposit, named after the Coermotibo River, is a buried deposit belonging to the Moengo-Ricanau-Jones (MRJ) District, also known as the Moengo Group of deposits (Bárdossy and Aleva, 1990) (Figure 2.5a, b). It is located in the eastern part of the Guiana Coastal Plain at 5032’N, 54018’W. Here, a bauxite-capped plateau with a surface area of approximately 20 km2 is split over 25 hills, varying in size between 0.03 and 7 km2. The bauxite layer is 3–6 m thick. All MRJ deposits but one (Coermotibo), stood out 25–70 m above the surrounding plain. The original geological reserve was approximately 127 Mt, and all of the exposed flat-topped bauxite hills have been mined out except the Coermotibo deposit, which has a proven reserve of 18 Mt and approximately 37 Mt of resources (Bauxite Institute Suriname, 2009). This bauxite deposit, with an overburden of approximately 40 m, was discovered in June 1959 during reconnaissance core drilling along the Coermotibo River, but was neglected due to 2 its high sulfide content and resilication (Bárdossy and Aleva, 1990) despite a high average Al2O3

content (51%) and a low average Fe2O3 content (4%). The high sulfur content of the bauxite is

linked to large quantities of marcasite (FeS2). Much of the deposit is located beneath a swamp, which probably created the reducing conditions favourable for marcasite formation in the grey bauxite.

2.3.2 | Successor Mines (Klaverblad, Kaaimangrasie, Caramacca) The Klaverblad, Kaaimangrasie, and Caramacca bauxite deposits, better known as the Successor Mines or Successor deposits, are located in the vicinity of the Paranam-Onverdacht-Lelydorp bauxite district (5020’N, 55020’W; Figure 2.6a). Only five of the thirteen deposits of the Paranam- Onverdacht-Lelydorp bauxite district had outcrops (nrs. 1, 3, 5, 6, and 7 in Figure 2.6a), whereas the remaining deposits were covered by a thick packet of sediments. Most the bauxite-capped hills are underlain by Early Eocene and Paleocene sediments. Exploration of this bauxite district started in 1939, while mining commenced in 1941. Most of its bauxite deposits have been

mined out. The original reserves of 100 Mt were based on a cutoff grade of Fe2O3 < 30% and

TSiO2 < 15%. The Klaverblad (KLB), Kaaimangrasie (KMG) and Caramacca (CRM) deposits were discovered during the Brokopondo bauxite exploration campaign in the early 60’s and 70’s of the last century (Figures 2.6b, c). The overburden at KLB is approximately 15 m thick and consists of Neogene and Quaternary sediments, while the sedimentary cover of the KMG and CRM deposits consists of Quaternary sediments with thicknesses of 4 and 9 m, respectively (BHP Billiton feasibility study, 2004) (Figure 2.6b). Mining of the Successor deposits started in 2004, and the remnants of the Caramacca deposit were mined out in 2015. The initial reserve of this group of deposits was 16 Mt.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 33 34 | Chapter 2

Figure 2.5 | (a) Distribution of the bauxite deposits in the Moengo-Ricanau-Jones district; (b) North-south cross section through the bauxite deposits in Figure 2.5a (modified after Bárdossy and Aleva, 1990).

2.4 | Plateau deposits

2.4.1 | Bakhuis Mountains The Bakhuis Mountains form a chain of strongly dissected plateaus in the district of Sipaliwini (4045’N, 56040’W). They are an expression of a 25 km wide and 95 km long northeast trending horst, one of the many structural features of the Bakhuis-Kanuku zone in the heart of the Guiana Shield (Kroonenberg and De Roever, 1975; Kroonenberg, 1976). It covers an area of approximately 2400 km2 (Figure 2.7a). Hilltops reach a height of approximately + 480 m MSL. The climate is tropical humid with an average temperature of 27.50 C and 1700–2200 mm of rainfall, divided over a long and a short rainy season. The basement of the Bakhuis Mountains consists of high-grade metamorphic rocks which include banded (mafic and intermediate) granulites and sillimanite gneisses with mafic and felsic intrusive bodies (Figure 2.7a) (De Roever et al., 1976, 2003; Klaver et al., 2015; Kroonenberg et al., 2016). The banded granulites and sillimanite gneisses underwent ultrahigh-grade (UHT) metamorphism between 2.055 and 2.072 Ga (De Roever et al., 2003). The charnockite intrusions in the southwestern part of the horst were formed at 1984.4 to 1992.5 Ma (Klaver et al., 2015). Variable compositions of bauxite weathering profiles reflect the large diversity of parent rocks (Aleva and Hilversum, 1984) (Figure 2.7a, b), which, together with different drainage conditions is also responsible for variable thicknesses of the bauxite body. Locally, lenses of kaolinite-rich material occur within the bauxite horizon, and lenses and boulders of bauxite

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 34 Bauxite deposits in Suriname: Geological context and resource development | 35

Figure 2.6 | (a) Location of the original bauxite deposits in the Paranam-Onverdacht-Lelydorp bauxite district (modified after Bárdossy and Aleva 1990); (b) Schematic relative position of the Successor deposits in a south- north section; (c) Lateritic weathering profile in the Klaverblad deposit.

material within the kaolinitic saprolite (Figure 2.7b) (Pollack, 1981; Aleva and Hilversum, 1984). Exposures of fresh bedrock and boulders along slopes, hill tops, creek beds and within the lateritic bauxite body are common. The proven bauxite reserves for Areas 10 and 5 are 70 Mt at

43.8% Al2O3 and 2% SiO2 (Janssen, 1963), while the resources of the entire Bakhuis Mountain are estimated to be larger than 500 Mt at an average of 34% available alumina and 2% reactive silica (Bauxite Institute Suriname, 2009).

2.4.2 | Nassau Mountains The Nassau Mountains in the Marowijne district in northeast Suriname (4046’–4054’N, 54030’– 54038’W) form an isolated, U-shaped mountain ridge (20 x 20 km2), which is bordered by the Professor W.J. Van Blommenstein Lake to the west and the Marowijne River to the east (Figure 2.7c). The ridge comprises four steep-sided, laterite-capped plateaus (A-D) at elevations between 500 and 564 m above MSL (Alonso and Mol, 2007). The mountains receive some of the highest rainfall in the country (2750–3000 mm/year, even more on the plateaus) due to orographic effects and mist interception (Alonso and Mol, 2007 and references therein).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 35 36 | Chapter 2

Figure 2.7 | (a) Geological map of the Bakhuis Mountains horst (modified after Klaver et al., 2015); (b) Approximately 2.4 km long WE-profile in Area 10. Note the highly variable grade and ore thickness in horizontal direction (modified after Janssen, 1963).

The Nassau Mountains and surrounding areas belong to the NW-SE striking Trans-Amazonian Marowijne Greenstone Belt with zircon ages between 2.26 and 2.10 Ga (Delor et al., 2003; Klaver et al., 2015; Kroonenberg et al., 2016). The Nassau Mountains contain rocks from the Paramaka Formation, which consists of metabasalts, metagabbro, meta-andesites, meta-cherts and other intermediate and felsic metavolcanic rocks (De Vletter, 1984; De Vletter et al., 1998; Bárdossy and Aleva 1990).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 36 Bauxite deposits in Suriname: Geological context and resource development | 37

Figure 2.7 continued | (c) Elevation map of the Nassau Mountains showing plateaus and concession limits (modified after Van den Bergh, 2011).

Figure 2.8 | Distribution of bauxite in South America and Africa (depicted in blue) since Jurassic times (modified after Tardy et al., 1991). The location of Suriname is highlighted in red.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 37 38 | Chapter 2

The presence of pure bauxite in the midst of highly hematitic laterite was discovered in 1918 by Douglas and Beems (Doeve, 1955). The deposit belongs to the Nassau bauxite district which also includes other bauxite and/or laterite-covered high plateaus (Brownsberg, Wintiwaai and Lely Mountains) in eastern Suriname (Bárdossy and Aleva 1990, and references therein). These mountains are tentatively ascribed to an Late Cretaceous or Early Tertiary surface (King et al., 1964). It has been estimated that the Nassau deposit contains about 40 Mt bauxite, most of which is located in approximately 1.7 m thick lateritic deposits below an overburden with an average thickness of 1.5 m. The most promising mining areas are Plateaus A and C (Figure 2.7c). One of the many obstacles to extract this deposit is the environmental concern about the presence of critically endangered endemic species of catfishes and frogs in these mountains (Ouboter et al., 2007; Alonso and Mol, 2007).

2.5 | Comparison of Surinamese and West African bauxite deposits

Large parts of South America and Africa are covered by a thick lateritic mantle (Tardy et al., 1991; Tardy, 1997) (Figure 2.7), which includes bauxites, ferricretes and nodular soils. The continents were separated in Jurassic–Cretaceous times during the break-up of Pangea (Pletsch et al., 2001). There are conspicuous similarities and differences between the Surinamese and West African bauxite deposits.

2.5.1 | Similarities and differences between Surinamese and West African bauxites – Most bauxites on both sides of the Atlantic are Paleocene-Oligocene in age (Prasad, 1983; Bárdossy and Aleva, 1990; Chardon, 2006; Théveniaut and Freyssinet, 2002) and formed during a bauxitization phase when conditions were favourable worldwide (Figure 2.8). – In both cases the bauxites are related to planation surfaces, in Suriname to the Main Aluminous Laterite Level and in West Africa to the African Level (King et al., 1964; Aleva, 1983; Wong et al., 1998). – The Plateau bauxites of Suriname and most of the West African countries have a metamorphic Proterozoic parent rock (Prasad, 1983; Mutakyahwa et al., 2003; Chardon, 2006). – An important difference is the mineralogical signature of the bauxites, as the Surinamese bauxites are dominantly gibbsitic with traces of boehmite in some deposits (e.g., in the Nassau deposit), while the West African bauxites are also characterized by high gibbsite contents but generally have higher boehmite contents (Figure 2.9). The boehmite content in the West African bauxite deposits increases northward towards the warmer and more arid conditions of the Sahara (Tardy et al., 1991). – Mineralogical changes from gibbsite and goethite into boehmite and hematite are accompanied by the formation of nodular and pisolitic structures and induration of ferricretes and bauxites (Bárdossy and Aleva, 1990; Tardy et al., 1991). This explains the frequent pisolitic texture of the West African bauxites compared to the generally massive texture of the Surinamese bauxites.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 38 Bauxite deposits in Suriname: Geological context and resource development | 39

Figure 2.9 | Mineralogical composition of the West African bauxites. Based on data from Tardy (1997).

2.5.2 | Climate control on bauxite distribution and mineralogy The most favourable climatic conditions for bauxite formation are found in tropical to humid subtropical zones with a mean annual temperature higher than 200C, a mean annual rainfall of more than 1700 mm and less than 4 months of dry season, which are currently located in a latitudinal belt approximately between 300 North and 300 South (McFarlane; 1983; Valeton, 1983; Bárdossy and Aleva, 1990). Bauxite deposits worldwide have formed in hot and humid (paleo-)tropical or (paleo-) equatorial regions since Devonian times (Valeton, 1972; Patterson et al., 1986; Bárdossy and Aleva, 1990; Tardy et al., 1991; Tardy, 1997). Bauxites have been retained on separate fragments of a post-Gondwanan surface in South America, Africa and India, i.e., in the present hot and humid tropical zone (Prasad, 1983). West-African bauxites in humid zones (Guinea, Nigeria, and Cameroon) and in drier areas (Burkina Faso) are considered to have formed during Jurassic, Cretaceous or Eocene times, when conditions were generally more humid than today (Tardy et al., 1991). Contrasting chemical, mineralogical and textural characteristics of laterites on the African and South American continents can be attributed to latitudinal differences in present- day climates as well as to different paleoclimatic histories since the opening of the Atlantic Ocean (Tardy et al., 1991). From Jurassic to Present time, the previously arid climates of South America became progressively more humid, whilst the formerly humid climates of West Africa progressively became more arid. In Africa, the development of ferricretes gradually decreases northwards, as humidity decreases, while the Saharan influence and the altitude increase. The presence of fossil ferricretes in the Sahara indicates that ancient climates were more humid than today (Tardy et al., 1991 and references therein).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 39 40 | Chapter 2

The mineralogical composition of bauxite can also be controlled by climate, as boehmite and hematite (dehydrated minerals) are often associated with relatively arid climates, while gibbsite and goethite (hydrated minerals) are related to constantly humid climates (Bárdossy and Aleva, 1990; Tardy, 1997). During the paleoclimatic history of West Africa, boehmite formation could be considered as secondary and contemporaneous to hematite (Tardy et al., 1991). In tropical climates with a marked dry season and relatively high temperatures, humidity is usually sufficient to induce a strong weathering during rainfall, but the temporary aridity of the dry season results in dehydration of gibbsite and goethite into boehmite and hematite, respectively (Tardy et al., 1991; Tardy, 1997).

2.6 | Bauxite and alumina production in Suriname and the economic outlook

Bauxite has been one of the most important sources of income for Suriname since 1930 (Aleva and Wong, 1998; Bauxite Institute Suriname, 2009). Bauxite was exported as raw material till 1985, whereas the export of aluminium, as end product started in 1965 and ceased in 2000 (Figure 2.10a). The export of alumina, the intermediate refinery product of bauxite, also commenced in 1965 and ceased in 2015. The leading geological commodities (gold, petroleum and alumina) contribute 95% to the total export and accounted for 35% of the country’s revenues, making the Surinamese economy highly vulnerable to market price volatility (Figure 2.10a). The bauxite sector in Suriname lost its position as main foreign exchange earner in 2004. Mining activities were significantly affected by the global economic recession that started in 2008. The bauxite revenues declined (Figure 2.10b), following a drop in alumina demand and market prices (Hoefdraad, 2014). The future economic potential of bauxite in Suriname is uncertain and will be determined by the quality and size of the reserves, geological conditions, available production technology, targeted locations in relation to transport facilities, world market price and governmental factors. The unexploited Surinamese bauxite deposits contain a total reserve of 580 Mt (Bauxite Institute Suriname, 2009; Lee Bray, 2015). These deposits have not been mined to date for different reasons. The Coermotibo deposit has not been economically attractive due to the high sulfur content in the bauxite and the 40 m thick overburden. The heterogeneous properties of the Bakhuis and Nassau bauxite deposits will make it necessary to drill a very close-spaced grid for mine planning. The remote location of the Nassau Mountains is also an economic and strategic problem. Governmental and economic factors have also had a negative influence on exploiting the Bakhuis deposit. The most sensitive obstacle is the environmental impact, as the Nassau and Lely Mountains have a unique flora and fauna including critically endangered endemic species (Ouboter et al., 2007; Bánki et al., 2008; Alonso and Mol, 2007).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 40 Bauxite deposits in Suriname: Geological context and resource development | 41

Figure 2.10 | (a) Real total export of commodities and real gross domestic product of Suriname, 1960–2010; (b) Mining exports of Suriname, 2007–2013 (modified from Hoefdraad, 2014). 2

2.7 | Concluding remarks

The lateritic bauxites of Suriname are derived from a variety of sedimentary and crystalline parent rocks. A diversity of geological conditions has yielded a multitude of bauxite deposits with variable quality, grade and accessibility, which are scattered across Cenozoic sediments of the coastal lowlands and Precambrian basement rocks of the interior highlands. During a century- long mining history and associated refining operations, the country became one of the world’s leading producers of bauxite, alumina and aluminium. The Al-based mineral commodities were a central pillar of Suriname’s economy for many decades. At present, the easily accessible high- grade deposits of the coastal plain have been mined out and significant reserves are restricted to the Plateau bauxite deposits in the more remote highlands. Despite a continuous growth of the annual global production of primary aluminium in recent years, Suriname’s production of alumina has ceased, largely due to unfavourable economic conditions. The future of the bauxite industry in Suriname is therefore uncertain.

Acknowledgements The author would like to thank M. Van Bergen and two anonymous reviewers for their valuable comments and suggestions to improve this manuscript. This work was supported by the Suriname Environmental and Mining Foundation (SEMIF).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 41 42 | Chapter 2

References

– Aleva, G. (1994). Laterites. Concepts, geology, morphology and chemistry. ISRIC (Wageningen), 169 pp. – Aleva, G. (1979). Bauxites and other duricrusts in Suriname: A review. Geologie en Mijnbouw 58, 321-336. – Aleva, G. and Hilversum, A. (1984). West Suriname: known deposits and potential. In: Jacob Jr., L. (ed): Bauxite: Proceedings of 1984 Bauxite symposium (Los Angeles). American Institute of Mining Metallurgical and Petroleum Engineers (New York), 319-348. – Aleva, G. and Wong, Th. (1998). The history of bauxite exploration and mining in Suriname. In: The history of Earth Sciences in Suriname- Royal Netherlands Academy of Arts and Sciences & Netherlands Institute of Applied Geoscience. TNO, 275-310. – Alonso, L. and Mol, J. (eds.) (2007). A rapid biological assessment of the Lely and Nassau Plateaus, Suriname (with additional information on the Brownsberg Plateau). RAP Bulletin Assessment 43. Conservation International, Arlington, 279 pp. – Bánki, O., Ter Steege, H., Jansen-Jacobs, M., Raghoenandan, U. (2008). Plant diversity of the Nassau Mountains, Suriname; Report of the 2003 Expedition. NHN-Utrecht branch. Utrecht University (Utrecht), BBS, Anton de Kom University of Suriname (Paramaribo), 1-52. – Bárdossy, G. and Aleva, G. (1990). Lateritic bauxites. Developments in Economic Geology. Elsevier Science Publishing 27, 569 pp. – Bauxite Institute Suriname (2009). Suriname as a source of bauxite and alumina. Prospects for growth? 15th Bauxite and Alumina Seminar, February 2009, Montega Bay, Jamaica. – BHP Billiton Suriname (2004). Feasibility study Successor Mines project. Volume 2 of 3, section 5, 1-7. – Chardon, D., Chevilotte, V., Beauvais, A., Grandin, G., Boulangé, B. (2006). Planation, bauxites and epirogeny; one or two paleosurfaces on the West African margin. Geomorphology 82, 273-282. – Delor, C., De Roever, E., Lafon, J., Lahondre, D,. Rossi, P., Cocherie, A., Guerrot, C ad Potrel, A. (2003). The Bakhuis ultrahigh-temperature Granulite Belt (Suriname): II. Implications for late Transamazonian crustal stretching in a revised Guiana Shield framework. Géology de la France, (2-3-4), 207-231. – De Roever, E., Kieft, C., Murray, E, Klein, E and Drucker, W. (1976). Surinamite, A new Mg-Al silicate from the Bakhuis Mountains, western Suriname: I. Description, occurrence and conditions of formation. American Mineralogist 61, 193-199. – De Roever, E., Lafon, J., Delor, C., Cocherie, A., Rossi, P., Guerrot, C. and Potrel, A. (2003). The Bakhuis ultrahigh- temperature Granulite Belt (Suriname): I. Petrological and geochronological evidence for a counterclockwise P-T path at 2.07-2.05 Ga. Géology de la France, (2-3-4), 175-205. – De Vletter, D. (1984). Economic geology and mineral potential of Suriname. In: De Vletter, D. (ed.). Geology of Suriname 8. Mededelingen Geologische Mijnbouwkundige Dienst Suriname 27, 91-129pp. – De Vletter, D., Aleva, G. and Kroonenberg, S. (1998). Research into the Precambrian of Suriname. In: Wong, Th., De Vletter, D., Krook, L., Zonneveld, J. and Van Loon, A., (eds.). The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences & Netherlands Institute of Applied Geoscience. TNO, 15-64 pp. – Doeve, G. (1955). De bauxietexploratie op het Nassaugebergte. Internal report Geologische Mijnbouwkundige Dienst Suriname, 68 pp. – Gurmendi, A. (2014). The mineral industries of French Guiana, Guyana, and Suriname. In: US Geological Survey Minerals Yearbook 2012, 10.6-10.9. – Hoefdraad, G. (2014). Leading sectors of Suriname: the impact of mining, agriculture and tourism activities on the economy, 1970-2012, Central Bank of Suriname (CBVS), Paramaribo, 147 pp. – Janssen, J. (1963). Reconnaissance exploration for bauxite in the Adampada-Kabalebo Area 1961-1963. Internal report Geologische Mijnbouwkundige Dienst, 45 pp. – Klaver, M., De Roever, E., Nanne, J., Mason, P. and Davies G. (2015). Charnockites and UHT metamorphism in the Bakhuis Granulite Belt, western Suriname: Evidence for two separate UHT events. Precambrian Research 262, 1-19.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 42 Bauxite deposits in Suriname: Geological context and resource development | 43

– King, L., Hobday, D. and Mellody, M. (1964). Cyclic denudation in Surinam. International report Geologische Mijnbouwkundige Dienst Suriname, 12 pp. – Kroonenberg, S. (1976). Amphibolite facies and granulite facies metamorphism in the Coeroenie Lucie Area, SW Suriname. Mededelingen Geologische Mijnbouwkundige Dienst van Suriname 25, 109-209. – Kroonenberg, S. and De Roever E. (1975). Dumortierite in cordierite pseudomorphs and in shear zones in high- grade metamorphic rocks from western Suriname. Mededelingen Geologische Mijnbouwkundige Dienst Suriname 23, 255-259. – Kroonenberg, S., De Roever, E., Fraga, L., Reis, N., Faraco, T., Lafon, J, Cordani, U and Wong, Th. (2016). Paleoproterozic evolution of the Guiana Shield in Suriname: A revised model. Netherlands Journal of Geosciences, Geologie en Mijnbouw, 95 (4), 491-522. – Lee Bray, E. (2015). U.S Geological Survey Mineral Commodity Summaries, Aluminum 2015, 16-17. – McFarlane, M. (1983). Laterites. In: Chemical sediments and geomorphology: precipitates and residua in the near surface environment. Goudi A., and Pye K. eds., Academic Press (London), 7-59. – McFarlane, M. (1983). Laterites. In: Goudi, A. and Pye, K. (eds.): Chemical sediments and geomorphology: precipitates and residua in the near surface environment. Academic Press, London, 7-59. – Mutakyahwa, M., Ikingura, J. and Mruma, A. (2003). Geology and geochemistry of bauxite deposits in Lushoto 2 District, Usambara Mountains, Tanzania. Journal of African Earth Sciences 36, 357-369. – Ouboter, P., Jairam, R. and Wan Tong, Y. (2007). Additional records of amphibians from the Nassau Mountains, Suriname. In: A rapid Biological assessment of the Lely and Nassau Plateaus, Suriname (with additional information on the Brownsberg Plateau). Alonso, L. and Mol, J. (eds). RAP Bulletin of Biological Assessment 43. Conservation International, Arlington, 128-129 – Patterson, S., Kurtz, H., Olson, J. and Neeley, C. (1986). World bauxite resources, geology and resources of aluminum. U.S Geological Survey Professional Paper 1076-B. US Government Printing Office, Washington, 125 pp. – Pletsch, T., Erbacher, J., Holborn, A., Kuhnt, W., Moullade, M., Oboh-Ikuenobede, F., Söding, E. and Wagner, T. (2001). Cretaceous separation of Africa and South America: the view from the West African margin (ODP Leg 159). Journal of South American Earth Sciences 14, 147-174. – Pollack, H. (1981). Bauxites and laterites of the Bakhuis Mountain Zone, western Suriname; a general description with emphasis on geomorphology and chemistry. In: Lateritization processes, Proceedings of International seminar on lateritization processes, December 1979, Oxford and IBH Publishing Company (New Delhi), 270-268. – Pollack, H. (2016). Suriname’s Bauxite Industry 1898-2009 and the Brokopondo agreement. Vaco, Paramaribo, Suriname, 372 pp. – Prasad, G. (1983). A review of the early Tertiary bauxite event in South America, Africa and India. Journal of African Earth Sciences 1 (3-4), 305-313. – Priem, H. (1998). Isotope geochronological research in Suriname. In: Th.E Wong, de Vletter, D., Krook,L., Zonneveld, J., van Loon, A., (eds.). The history of earth sciences in Suriname. Royal Netherlands Academy of Arts and Sciences & Netherlands Institute of Applied Geoscience. TNO, 65-72. – Tardy, Y. (1997). Petrology of laterites and tropical soils. Translated by Sarma, V., A.A.Balkema (Rotterdam- Brookfield), 376 pp. – Tardy, Y., Kobilsek, B., Paquet, H., (1991). Mineralogical composition and geographic distribution of African and Brazilian periatlantic laterites. The influence of continental drift and tropical paleoclimates during the past 150 million years and implications for India and Australia. Journal of African Science Sciences 12 (1-2), 283-295. – Théveniaut, H. and Freyssinet, Ph. (2002). Timing of lateritization on the Guiana Shield: synthesis of paleomagnetic results from French Guiana and Suriname. Palaeography, Palaeoclimatology, Palaeoecology 178, 91-117. – Valeton, I. (1972). Bauxites. Elsevier (Amsterdam), 226 pp. – Valeton, I. (1983). Palaeoenvironment of lateritic bauxites with vertical and lateral differentiation. Geological Society, London, Special Publications 11, 77-90.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 43 44 | Chapter 2

– Van der Bergh, J. (2011). Executive summary for the Nassau Plateau bauxite project. Internal report Suralco L.L.C., 1-14. – Van der Hammen, T. and Wijmstra. T. (1964). Palynological data on the age of the bauxite in British Guyana and Suriname. Geologie en Mijnbouw 43, 143. – Van Kersen, J. (1956). Bauxite deposits in Surinam and Demerara Britisch Guiana. Leidse Geologische Mededelingen 21 (1), 247-375. – Wong, Th. (1989). Revision of the stratigraphy of the coastal plain of Suriname. Mededelingen Natuurwetenschappelijke Studiekring voor Suriname en de Nederlandse Antillen 123, 64 pp. – Wong, Th., Krook, L. and Zonneveld, J. (1998). Investigations in the coastal plain and offshore area of Suriname. In: Wong, Th., De Vletter, D., Krook, L., Zonneveld, I. and Van Loon, A. (eds.): The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences, Netherlands Institute of Applied Geoscience (TNO), 73-100. – Wong, Th., De Kramer, R., De Boer, P, Langereis, C. and Sew-A-Tjon, J. (2009). The influence of sea-level changes on tropical coastal lowlands; the Pleistocene Coropina formation, Suriname. Sedimentary Geology 216, 125- 137.

Visited websites – www.staatsolie.com (visited on December 17th 2014)

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 44 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 45 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 46 Chapter 3 Trace-element analysis of bauxite using laser ablation-inductively coupled plasma-mass spectrometry on lithium borate glass beads

The results of this study have been submitted for publication in Geostandards and Geoanalytical Research as: Monsels, D.A., Van Bergen, M.J. and Mason, P.R.D. – Trace- element analysis of bauxite using laser ablation-inductively coupled plasma-mass spectrometry on lithium borate glass beads.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 47 48 | Chapter 3

Abstract

The determination of trace-element contents in bauxite by solution ICP-MS or other wet chemical analytical techniques relies on complete digestion of samples, which can be problematic, in particular since important geochemical tracers such as HFSE and REE are often hosted in resistant mineral phases. Fusion treatment with alkali fluxes is an efficient alternative to acid digestion in decomposing bauxite for analytical purposes. This paper explores the applicability of laser ablation ICP-MS on lithium borate glass beads, commonly prepared for XRF analysis, as a viable method for the determination of trace elements in bauxite samples. The use of laser ablation overcomes the problems of measuring a heavy sample matrix encountered in solution nebulization ICP-MS, increases sample throughput and facilitates prior acquisition of complementary XRF data without additional sample preparation. The method was validated using four bauxite reference materials (ANRT BX-N, NIST-SRM 69b, NIST-SRM 696 and NIST-SRM 698) and an iron formation reference (CCRMP FeR-2). We demonstrate that most trace elements of interest can be typically measured to within 20% of reference values with an external reproducibility of < 20% RSD, with many elements showing much better levels of accuracy and precision.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 48 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 49

3.1 | Introduction

Bauxite is the most important raw material for commercial aluminium production. The natural weathering processes that cause Al accumulation also promote enrichment of other relatively immobile constituents of original parent rocks to the extent that waste products of industrial processing receive increasing attention as potential economic sources of trace elements such as gallium, scandium and rare earth elements (Ochsenkühn-Petropoulou et al., 1994; Zhang et al., 2005; Paramguru et al., 2006; Borra et al., 2015; Ujaczki et al., 2017). Analysis of the major and trace-element composition of bauxite not only serves to constraining ore grade, but also provides information on parent rocks and the nature and conditions of genetic processes, which can be used as a guide for exploration or processing strategies. Trace-element concentrations in bauxites have been determined by a variety of methods including instrumental neutron activation analysis (INAA) (Vukotic 1983, Korotev 1996, Grant et al., 2005), X-ray fluorescence spectrometry (XRF) (Mordberg et al., 2001) and inductively coupled plasma-atomic emission spectrometry (ICP-AES) (Boulangé and Colin 1994; Ochsenkühn-Petropoulou et al., 1990). In recent years, inductively coupled mass spectrometry (ICP-MS) has become the technique of choice, because it allows analyzing a comprehensive set of elements that includes transition metals, high field strength elements (HFSE) and rare earth elements (REE) at low concentrations (e.g., Horbe and Anand, 2011; Gu et al., 2013; Wang et al., 2013; Da Costa et al., 2014; Mongelli et al., 2014; Liu et al., 2016; Ahmadnejad et al., 2017). However, a key concern in obtaining 3 reliable trace-element concentrations by solution ICP-MS is the difficulty in achieving complete digestion of the bauxite, as it often contains refractory minerals such as zircon, anatase/rutile, monazite, xenotime, titanite, thorite and tourmaline (Bárdossy and Aleva 1990; Horbe and Da Costa, 1999; Da Costa et al., 2014; Monsels and Van Bergen, 2017), which are usually inherited from the parent rock and host many of the elements of interest. Mixed acid attack may result in incomplete recoveries of analytes if such phases are present

or upon the formation of insoluble AlF3 precipitates that could incorporate significant amounts of trace elements (Yu et al., 2001; Cotta and Enzweiler, 2012; Chen et al., 2017). Hence, special optimization of the acid digestion method is recommendable to accurately determine trace- element contents of bauxite samples by ICP-MS (Zhang et al., 2016). An alternative method to ensure dissolution of refractory minerals, is fusion with lithium borate followed by acid dissolution of the glass material (e.g., Yu et al., 2001; Panteeva et al., 2003; Awaji et al., 2006), with the additional advantage that the same sample can be analysed for major elements by XRF and for trace elements by ICP-MS (De Madinabeitia et al., 2008; Amosova et al., 2016). Although fusion treatment with an alkali flux is efficient in decomposing refractory minerals, typical drawbacks for ICP-MS are the introduction of a large amount of total dissolved solids (TDS) causing suppression/enhancement effects and extensive instrumental drift, relatively high blank levels, as well as a risk of contamination from impure reagents or metal crucibles (Totland et al., 1992; Yu et al., 2001; Awaji et al., 2006). Several of these problems are overcome when using laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS), which is a widely used technique for the microanalysis of solid materials and has been extensively tested for the analysis of XRF fusion beads (Nesbitt et al., 1997; Günther et al., 2001; Eggins, 2003; Orihashi and Hirata, 2003; Petrelli et al., 2008; Regnery et al., 2009; Leite et al., 2011; Park et al., 2016). The relatively small amount of solid particles

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 49 50 | Chapter 3

introduced by this method is not susceptible to the same matrix effects as observed when aspirating solutions rich in lithium borate flux. LA-ICP-MS also offers several major advantages compared to more conventional analytical methods for the analysis of solid fusion beads such as XRF including the multi-element analysis capability coupled with high detection power. On the other hand, accuracy and precision in LA-ICP-MS can be limited by elemental fractionation at or close to the site of ablation, incomplete ionization of the particles in the plasma, relatively poor stability of the plasma source, spectroscopic interferences, difficulties in correcting for ablation yield and the lack of suitable matrix-matched standards (e.g., Longerich et al., 1996; Mason and Kraan, 2002; Yu et al., 2003; Jochum, 2007; Regnery et al., 2009; Miliszkiewicz et al., 2015; Shazzo and Karpov, 2016; Zhang et al., 2016; Chen et al., 2017). In this study, we investigate the utility of laser ablation ICP-MS after lithium borate fusion for bauxite analysis. The method was validated using international reference materials for bauxite and iron formation, as well as USGS basaltic glass BCR-2G as external control standard.

3.2 | Experimental procedures

3.2.1 | Materials Four international geological reference materials for bauxite, NIST-SRM 69b (Arkansas, USA), NIST-SRM 696 (Suriname), NIST-SRM 698 (Jamaica) and BX-N (France), and one reference for iron formation (FeR-2 from Canada) to account for possible Fe-enrichments in samples from bauxite profiles, were prepared as lithium borate glasses following standard sample preparation procedures for XRF analysis. The glass beads were prepared by mixing 1 gram of ignited powder

of the reference material with 4 grams of flux consisting of 66.5 wt% lithium tetraborate (Li2B4O7)

and 33.5 wt% lithium metaborate (LiBO2), heating the mixture in a platinum crucible (75% Pt – 25% Au) at ca. 1100 0C, and pouring the melt into platinum molds. Test samples of Surinamese bauxite deposits, analyzed to demonstrate the effectiveness of the method, were treated in the same way.

3.2.2 | Analytical methods Laser ablation analysis Trace elements were determined by laser-ablation inductively coupled plasma-mass spectrometry (LA-ICP-MS) using a ThermoFischer Scientific Element 2 magnetic sector field ICP-MS instrument, coupled to a Lambda Physik excimer laser (193 nm) with GeoLas optics (Mason and Kraan, 2002). Ablation parameters and operating conditions are given in Table 3.1. Calibration was performed against NIST-SRM 612 following the methodology of Longerich et al., (1996) with double-standard measurements bracketing each of six unknown samples. Silicon 29 ( Si) was adopted as internal standard using the recommended/working values for SiO2 of 7.4, 3.79, 0.68, 13.41 and 49.24 wt% for BX-N, NIST-SRM 696, NIST-SRM 698, NIST-SRM 69B and FeR-2 (Govindaraju, 1994), respectively. Reported compositions are averages of three measurements for each sample. Instrumental performance and accuracy of results was monitored by analyzing USGS standard BCR-2G after each six samples. For assessing the quality of analytical results on the bauxite and iron formation standards we used the compilation values of Govindaraju (1994,

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 50 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 51

1995) for comparison. The available values for the BX-N and FeR-2 reference materials include sets of recommended and provisional concentrations, whereas no uncertainty information on the trace-element working values were reported for the NIST-SRM 69b, NIST-SRM 696 and NIST- SRM 698 standards. For FeR-2, we also employed a set of compilation values that were certified or provisionally recommended by the Canadian Certified Reference Materials Project (CCRMP) under the auspices of Natural Resources Canada (NRCan). Additional literature data included for comparison are solution ICP-MS concentrations reported for the three NIST standards (Wang et al., 2016) and FeR-2 (Yu et al., 2001; Sampaio and Enzweiler, 2015) and INAA data for BX-N (Bédard and Barnes, 2002), NIST-SRM 69b (Korotev, 1996), NIST-SRM 696 and NIST-SRM 698 (Grant et al., 2005). All reference data were extracted from the GeoReM database (georem. mpch-mainz.gwdg.de).

3.3 | Results and discussion

3.3.1 | Bauxite and iron formation standards The potential problem of intra-sample inhomogeneity of synthesized lithium borate glass beads warrants testing for the reference materials analyzed. Measurement precisions from multiple spot analyses in the individual glass beads are a guide for the degree of homogeneity, provided that uncertainties caused by counting statistics can be ignored. 3 Average precisions for three independent spot analyses made on each of the five discs were better than 4% RSD, except for Rb, Cs and Zn (average 6–8% RSD), which are present in concentrations relatively close to detection limits. This finding demonstrates a homogeneous distribution of analytes in the glasses (Tables 3.3–3.7).

Table 3.1 | Instrumental operating conditions.

ICP-MS Plasma power 1300 W Cool: 16 l min-1 Ar Gas flows Auxiliary: 1.0 l min-1 Ar Carrier: 0.685 l min-1 Ar, 0.696 l min-1 He Internal standard Silicon (29Si) Laser Wavelength, pulse rate 193 nm, 10 Hz Energy density at target 5 J cm-2 Ablation crater diameter 120 µm Ablation time total of 60 seconds + 30 seconds gas blank

The average relative standard deviation (RSD) of the elements measured in the BCR-2G basaltic glass standard (Table 3.2), used to monitor instrumental and procedural performance, was typically better than 10%, except for Zn, As, Zr, Hf (12–13%), whereas reproducibility was

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 51 52 | Chapter 3

Figure 3.1 | Comparison of trace element concentrations measured by LA-ICP-MS in BCR-2G relative to GeoReM values (Jochum et al., 2016) and other LA-ICP-MS results (Gao et al., 2002; Jochum et al., 2005), expressed as percent difference (%dev.). Data from Table 3.2.

relatively poor for Pb (23 %). Accuracy, measured against the GeoRem results (Jochum et al., 2016) was excellent. The deviation from recommended values, determined in multiple sessions, was ≤ 12% for all reported trace elements in BCR-2G (Figure 3.1), except for As, which is present at a level of ca. 1 ppm. Comparison with other LA-ICP-MS results obtained on BCR-2G (Gao et al., 2002; Eggins, 2003; Jochum et al., 2005: Andrade et al., 2014) shows a good concordance as well. A rigorous assessment of the quality of our data obtained on the reference bauxites and iron formation was hampered by the lack of certified/recommended values for trace elements in these materials. Reference BX-N is the only bauxite standard with a virtually complete set of recommended or provisional values available for comparison (Govindaraju, 1994, 1995). Relative to these data, our results generally show good agreement (Tables 3.3–3.7 and Figure 3.2). Relative deviations (%dev.) are better than 10% for Zn, Sr, Y, Zr, Nb, most of the REE, Hf, Pb and Th, while %dev. values are between 10 and 20% for As, Rb, Ba, Eu, Gd, Tb, Ho, Ta and U. Poor accuracy (% dev. > 20) is shown by Sc, V, Cr and Cs. The iron-formation standard FeR-2 has a substantially different trace-element composition, as it contains some 40 times less As, 10–20 times less REE and HFSE, 2–5 times less Zn, Sr, Eu, Cr, and 7–20 times more Ba, Cs and Rb than BX-N. Nonetheless, relative to recommended or provisional values (Govindaraju, 1994, 1995), our results for FeR-2 show roughly the same accuracies as for BX-N, although %dev. values tend to be enhanced for elements with much lower concentrations (e.g., As, Sc, La, Pb and U), while accuracies are strongly improved for elements with much higher concentrations (e.g., Ba, Cs and Rb).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 52 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 53

Table 3.2 | Summary of replicate analyses of USGS BCR-2G glass compared to GeoReM preferred values and other

LA-ICPMS results. Measured values of BCR-2G are based on Si as internal standard using SiO2= 54.00% m/m from Jochum et al., 2016. (*) Average of 61 spot measurements taken in 5 different sessions. (**) Reference values, except for As, La and Ho, which are information values. Analytical uncertainties for the GeoReM values expressed at the 95% confidence level (CL).

Measured GeoReM values** Other LA-ICPMS results %dev. (this work)* Jochum et al. Gao et al. Jochum et al. GeoReM Gao et al. Jochum et al. (2016) (2002) (2005) values (2002) (2005) ppm %RSD ppm Uncert. Mean %RSD Mean %dev. %dev. %dev. 95%CL ppm ppm Sc 34.4 7 33.53 0.4 32 6– 38 – V 466 6 417.6 4.5 425 2– 12 10 – Cr 17.4 8 15.85 0.38 17 12 – 10 2 – Zn 134 13 129.5 1.8 153 6– 4 -12 – As 1.09 13 0.86 0.22 –– – 27 – – Rb 48.7 3 46.02 0.56 51 6 45.1 6 -4 8 Sr 323 6 337.4 6.7 321 2 332 -4 0.6 -3 Y 33.6 8 36.07 0.37 31 6 36.9 -7 9 -9 Zr 179 12 186.5 1.5 167 5 188 -4 7 -5 3 Nb 11.4 5 12.44 0.2 10.9 6 12.3 -8 5 -7 Cs 1.13 7 1.16 0.023 1.17 7 1.1 -2 -3 3 Ba 698 8 683.9 4.7 641 2 650 29 7 La 24.5 6 25.08 0.16 25 4 25.5 -2 -2 -4 Ce 51.9 5 53.12 0.33 52 4 50.4 -2 -0.2 3 Pr 6.20 5 6.827 0.044 6.3 6 6.59 -9 -2 -6 Nd 27.4 6 28.26 0.37 27 4 28.5 -3 1.5 -4 Sm 6.33 7 6.547 0.047 6.3 8 6.58 -3 0.5 -4 Eu 1.83 6 1.989 0.024 1.91 5 1.95 -8 -4 -6 Gd 6.40 8 6.811 0.078 6.5 9 6.74 -6 -2 -5 Tb 0.95 9 1.077 0.026 0.95 7 1.07 -12 -0.2 -11 Dy 6.24 8 6.424 0.055 67 6.59 -3 4 -5 Ho 1.22 9 1.313 0.011 1.2 6 1.34 -7 2 -9 Er 3.54 9 3.67 0.038 3.3 6 3.76 -4 7 -6 Yb 3.34 8 3.392 0.036 3.2 9 3.59 -2 4 -7 Lu 0.48 8 0.5049 0.0078 0.47 9 0.531 -5 2 -10 Hf 4.66 12 4.972 0.034 4.5 9 5.07 -6 3 -8 Ta 0.70 4 0.785 0.018 0.63 10 0.761 -11 11 -8 Pb 11.5 23 10.59 0.17 10.9 5 10.6 85 8 Th 5.55 7 5.828 0.05 5.5 4 6.12 -5 1.0 -9 U 1.58 4 1.683 0.017 1.7 5 1.67 -6 -7 -5

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 53 54 | Chapter 3

Table 3.3 | Measured values of ANRT BX-N compared to reference values. Trace-element concentrations (ppm) and precisions expressed as 1 SD and %RSD for LA-ICPMS measurements on lithium borate glass beads, compared with reference values obtained by compilation, INAA or solution ICPMS. The SD values for measured concentrations are based on three spot analyses on the same lithium borate glass bead. %dev. values represent relative difference between measured and reference concentrations expressed in percentage. R (recommended value) and P (provisional value) (Govindaraju 1994, 1995).

ANRT BX-N Measured (this work) Govindaraju (1994) Bédard and Barnes (2002) Element Compiled %dev. INAA %dev. ppm SD (n=3) %RSD ppm ± (1SD) ppm ± (1SD) Sc 72.1 1.8 2.4 60 10 R 20 61.55 0.09 17 V 435 4 0.9 350 77 P 24 Cr 346 2 0.5 280 75 R 24 290 2 19 Zn 81.7 3.1 3.8 80 39 R2 As 131 6 5.0 115 9R14 120.5 18 Rb 2.91 0.08 2.8 3.6 11 P -19 Sr 109 1 0.9 110 19 P -1 Y 118 2 1.5 114 40 P3 Zr 564 8 1.3 550 89 P2 Nb 54.5 0.8 1.5 52 5P5 Cs 0.30 0.02 7.9 0.4 –R-24 Ba 33.4 0.2 0.5 30 26 P 11 La 377 5 1.3 355 76 R6386.5 0.7 -3 Ce 558 2 0.4 520 43 R7574 2 -3 Pr 51.2 0.4 0.8 54 –R-5 Nd 156 1 0.9 163 31 R -4 178 2 -12 Sm 21.0 0.3 1.5 22 3.4 R -5 22.64 0.03 -7 Eu 3.98 0.10 2.6 4.4 0.5 R -10 4.44 0.2 -10 Gd 17.4 0.3 1.7 20 6.4 R -13 Tb 2.67 0.02 0.9 3 0.3 R -11 3.15 0.1 -15 Dy 18.6 0.1 0.7 18.5 –R1 Ho 3.67 0.07 1.8 4.1 –R-10 Er 11.0 0.1 0.6 11 –R0.2 Yb 11.9 0.4 3.1 11.6 2.1 R312.4 0.1 -4 Lu 1.67 0.06 3.3 1.8 0.4 R -7 1.89 0.02 -12 Hf 13.9 0.2 1.7 15.2 4P-9 15.6 0.3 -11 Ta 3.7 0.05 1.3 4.6 0.8 R -19 4.2 0.2 -11 Pb 129 2 1.7 135 76 R -4 Th 46.4 0.5 1.0 50 11 P -7 54.1 0.2 -14 U 7.83 0.07 0.9 8.8 2.4 P -11 10 0.2 -22

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 54 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 55

Table 3.4 | Measured values of NIST-SRM 69b compared to reference values. See Table 3.3 for explanation. avg. = average.

NIST-SRM69b Measured (this work) Govindaraju 1994 Korotev 1996 Zhang et al. 2016 Element Compiled %dev. INAA %dev. ICPMS %dev. ppm SD (n=3) %RSD ppm ppm 95%CL(rel) average SD (n=3) Sc 15.5 0.7 4.5 8R 93 8.32 3 86 V 167 3 1.8 160 R4 147 8 13 Cr 86.1 2.0 2.2 75 R 15 73 10 18 59.8 3 44 Zn 36.5 4.6 12 28 R 30 21.2 1 72 As 26.2 0.2 0.9 24.4 38 Rb 3.52 0.17 4.6 4.4 30 -20 3.33 0.03 6 Sr 126 2 1.6 134 10 -6 126 4 0.1 Y 67.2 0.68 1.0 78 – -14 69.0 1 -3 Zr 2882 20.6 0.7 2150 R 34 2670 10 8 2632 39 10 Nb 861 5 0.5 868 8 -1 Cs 0.09 0.02 17 0.15 30 -38 0.15 0.01 -36 Ba 78.7 1.8 2.2 72 R9 77 30 2 78.0 41 3 La 81.4 1.4 1.7 75 3872.2 4 13 Ce 241 4 1.4 240 R1 249 3 -3 242 2 -0.3 Pr 10.5 0.2 1.6 10.6 0.3 -0.1 Nd 31.5 0.7 2.1 31 10 2 31.0 12 Sm 5.37 0.06 1.1 5.73 3 -6 5.70 0.3 -6 Eu 0.79 0.02 2.6 0.866 3 -9 0.84 0.03 -6 Gd 5.26 0.02 0.4 5.7 – -8 5.80 0.3 -9 Tb 1.07 0.01 0.6 1.21 10 -11 1.23 0.04 -12 Dy 8.99 0.21 2.2 9.2 – -2 9.40 0.2 -4 Ho 2.08 0.05 2.5 2.4 – -13 2.25 0.1 -8 Er 7.65 0.16 2.0 8–-4 7.71 0.2 -1 Yb 10.5 0.3 2.5 10.4 10 1 10.4 0.4 1 Lu 1.54 0.03 2.0 1.58 10 -3 1.56 0.1 -1 Hf 52.1 1.4 2.5 63 R -17 58.2 10 -10 54.1 3 -4 Ta 41.5 0.5 1.3 43.9 3 -6 45.1 1 -8 Pb 41.2 0.8 1.8 47.2 3 -13 Th 79.4 1.4 1.7 83.9 3 -5 83.5 4 -5 U 10.8 0.2 1.8 12.9 – -16 12.8 0.4 -15

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 55 56 | Chapter 3

Table 3.5 | Measured values of NIST-SRM 696 compared to reference values. See Table 3.3 for explanation. avg. = average. NIST-SRM 696 Measured (this work) Govindaraju (1994) Grant et al. (2005) Zhang et al. (2016) Element Compiled %dev. INAA %dev. ICPMS %dev. ppm SD (n=3) %RSD ppm ppm avg. SD (n=3) Sc 12.1 0.9 7.6 8 51 8.2 47 V 447 6 1.4 400 12 383.8 16 381 6 17 Cr 362 2 0.7 320 13 310.3 17 288 7 26 Zn 17.7 3.0 17 11 61 11.7 51 6.30 0.20 181 As 14.9 0.4 3.0 14.4 4 Rb 0.40 0.02 4.0 0.31 0.01 30 Sr 21.5 0.6 2.6 179.5 -88 21.1 0.3 2 Y 13.4 0.4 2.6 14.5 1.1 -8 Zr 1076 10 1.0 1040 3 1023 13 5 Nb 51.4 1.0 2.0 53.5 0.5 -4 Cs 0.04 0.001 2.0 0.2 -82 0.030 0.003 17 Ba 12.6 0.8 6.2 36 -65 12.5 0.3 1 La 29.9 0.8 2.7 26.4 13 23.8 0.3 26 Ce 36.2 1.1 2.9 41 -12 42.1 -14 36.1 0.1 0.5 Pr 3.30 0.08 2.5 3.26 0.0 1 Nd 10.4 0.4 3.6 9.4 10 10.2 0.1 1 Sm 1.72 0.05 2.6 2.1 -18 1.76 0.04 -2 Eu 0.33 0.02 6.0 0.5 -33 0.37 0.01 -10 Gd 1.41 0.09 6.5 1.41 0.03 -0.1 Tb 0.25 0.01 4.6 0.3 -16 0.28 0.01 -9 Dy 2.01 0.08 4.0 2.1 -4 2.11 0.07 -4 Ho 0.46 0.03 5.5 0.48 0.02 -3 Er 1.57 0.08 5.1 1.58 0.06 -1 Yb 2.24 0.12 5.4 3.4 -34 2.25 0.08 0 Lu 0.35 0.01 3.9 0.4 -12 0.39 0.01 -9 Hf 24.6 1.1 4.3 32 -23 34.8 -29 29.0 1.6 -15 Ta 3.75 0.20 5.2 4.19 0.10 -10 Pb 18.1 1.1 6.1 21.5 0.41 -16 Th 62.2 2.5 4.1 77.5 -20 70.4 0.9 -12 U 4.01 0.20 4.9 5.1 -21 4.72 0.07 -15

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 56 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 57

Table 3.6 | Measured values of NIST-SRM 698 compared to reference values. (*) For most elements in NIST-SRM 698 the measured concentrations would fit better with comparison values if Ce is used as internal standard (see text). avg. =average.

NIST-SRM 698 Measured (this work)* Govindaraju (1994) Grant et al. (2005) Zhang et al. (2016) Element Compiled %dev.* INAA %dev.* ICPMS %dev.* ppm SD (n=3) %RSD ppm ppm avg. SD (n=3) Sc 52.1 0.8 1.5 51 2 51.1 2 V 363 2 0.4 360 1 342.5 6 339 13 7 Cr 562 3 0.4 550 2 513.8 9 502 5 12 Zn 182 2 1.3 230 -21 232.1 -21 268 3 -32 As 38.1 0.7 1.9 39.9 -5 Rb 0.67 0.13 19 0.61 0.02 9 Sr 124 2 2.0 308.3 -60 152 3 -19 Y 252 3 1.0 313 8 -20 Zr 427 5 1.2 450 -5 482 9 -12 Nb 44.0 1.0 2.2 53.4 2 -18 Cs 0.14 0.01 10 0.9 -84 0.20 0.01 -28 Ba 44.9 1.0 2.3 72 -38 63.7 2 -30 3 La 226 4 1.6 272.2 -17 257 5 -12 Ce 251 8 3.0 300 -16 293.2 -14 290 2 -14 Pr 33.8 1.0 2.8 40.9 1 -18 Nd 134 3 2.5 233 -43 160 2 -17 Sm 22.7 0.5 2.2 30.3 -25 28.8 1 -21 Eu 5.12 0.19 3.6 5.8 -12 6.25 0.04 -18 Gd 24.0 0.7 3.1 24.2 0.3 -1 Tb 3.30 0.07 2.1 3.7 -11 4.46 0.0 -26 Dy 23.2 0.7 3.1 29.8 -22 28.5 0.4 Ho 4.78 0.16 3.3 6.04 0.1 -21 Er 13.1 0.4 2.9 15.7 0.2 -17 Yb 10.3 0.3 2.9 9.2 12 12.2 0.4 -15 Lu 1.38 0.03 2.1 1.8 -23 1.70 0.1 -19 Hf 10.5 0.3 2.9 15 -30 14.1 -25 14.3 0.1 -26 Ta 2.84 0.14 4.7 3.72 0.1 -24 Pb 59.1 2.8 4.7 85.5 1 -31 Th 26.5 1.0 3.7 34.7 -24 32.5 2 -18 U 6.81 0.25 3.6 9.8 -31 9.55 0.2 -29

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 57 58 | Chapter 3

Table 3.7 | Measured values of FeR-2 compared to reference values. C (certified value) and PR (provisionally recommended value) (CCRMP, 2003). S. and E. (2015)= Sampio and Enzweiler (2015). avg.= average

CCRMP FeR-2 Measured (this work) Govindaraju (1994) Yu et al. (2001), CCRMP (2003) S. and E. (2015) Element Compiled %dev. ICPMS Compiled %dev. ppm SD (n=3) %RSD ppm avg. SD n ppm Sc 10.0 0.3 2.9 6P67 5.55 0.34 26C67 V 50.0 0.5 0.9 36 R 39 34.5 6.4 2 37 C 35 Cr 62.1 0.9 1.4 46 R 35 45.5 2.1 2 47 C 32 Zn 39.6 0.5 1.2 44 R -10 36.5 0.7 2 43 C -8 As 3.14 0.2 7.6 2.1 P 49 – 2 PR 57 Rb 62.7 0.2 0.3 67 P -6 64.3 3.5 2 66 C -5 Sr 60.2 1.3 2.2 58 R464.0 3.2 4 58 C4 Y 13.8 0.3 2.3 16 R -13 12.7 1.1 5 15 C -8 Zr 40.7 0.8 2.0 39 R439.8 4.1 4 39 C4 Nb 4.24 0.08 1.8 3.01 0.65 4 Cs 4.21 0.05 1.2 4.5 P -6 4.89 0.08 25PR -16 Ba 216 3 1.6 230 R -6 227 17 4 240 PR -10 La 22.8 0.4 1.9 12 P 90 12.4 0.71 5 14 C 63 Ce 23.5 0.4 1.6 25 R -6 25.2 1.79 5 Pr 2.75 0.08 2.9 3P-8 3.04 0.20 5 Nd 11.7 0.2 2.1 12 P -3 12.3 0.93 5 Sm 2.54 0.15 6.0 2.5 R22.60 0.28 5 2.6 C -2 Eu 1.23 0.05 4.3 1.25 P -1 1.29 0.12 5 Gd 2.17 0.08 3.6 2R82.37 0.30 5 Tb 0.32 0.02 4.9 0.32 P -1 0.36 0.03 5 Dy 2.24 0.17 7.5 2P12 2.27 0.19 5 Ho 0.46 0.01 3.0 0.6 P -23 0.48 0.05 5 Er 1.40 0.08 5.8 1.5 P -7 1.44 0.14 5 Yb 1.39 0.03 1.8 1.25 R 11 1.38 0.15 5 1.3 C7 Lu 0.20 0.01 7.0 0.2 P -2 0.21 0.02 5 Hf 0.98 0.04 3.7 1P-2 1.16 0.13 4 Ta 0.18 0.01 3.3 0.2 P -10 0.21 0.04 4 Pb 7.61 0.08 1.1 11 P -31 8.92 0.62 4 11 C -31 Th 2.41 0.09 3.8 2.4 P02.77 0.35 43C-20 U 0.94 0.04 4.3 1.2 P -22 1.08 0.17 4

Comparison with the available INAA data on the bauxite reference materials BX-N, NIST-SRM 69b, NIST-SRM 696 and NIST-SRM 698 yields an ambiguous picture (Tables 3.3-3.6). For virtually all elements our results show excellent agreement with the data for BX-N and NIST-SRM 69b

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 58 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 59

reported by Bédard and Barnes (2002) and Korotev (1996), respectively. In contrast, relative to the results for NIST-SRM 696 and NIST-SRM 698 of Grant et al. (2005), many of our concentrations deviate substantially (%dev. values are > 20 for some 30% of the elements). In general, our results compare well with solution ICP-MS data that cover a more complete set of trace elements than the INAA data. With some exceptions, there is excellent correspondence with the solution ICP-MS data for NIST-SRM 69b and NIST-SRM 696 (Zhang et al., 2016) and for FeR-2 (Yu et al., 2001; Sampaio and Enzweiler, 2015). Only for NIST-SRM 698 (Zhang et al., 2016), our results are consistently ca. 17% lower, comparable to the average difference with the INAA data of Grant et al., (2005). We tentatively attribute the systematic offsets for this reference to its

low Si content (0.68% m/m SiO2, Govindaraju 1994), which may affect the performance of this element as internal standard, either due to analytical uncertainty in the recommended value for this bauxite standard, or to the low count-rate ratio of 1.8 for the pre-ablation blank and the analyte signal recorded in our measurement. When adopting Ce as internal standard for the NIST-SRM 698 measurement and using the compilation value of 300 ppm of Govindaraju (1994), the agreement with the solution ICP-MS results of Zhang et al. (2016) becomes excellent for some 90% of the trace elements analyzed (Table 3.6 and Figure 3.2).

3.3.2 | Sources of error Sources of analytical error that potentially affected the results are either inherent to the ICP- MS technique or specifically associated with the use of borate glasses. Explanations for the 3 frequent, relatively large deviations of measured concentrations of some elements (e.g., Sc, V, Cr, Zn, La, Pb) compared to reported values may include contamination or loss during sample preparation, impurities in the borate flux, and polyatomic spectral interferences. The persistently overestimated concentrations of Sc, V and Cr are likely attributable to polyatomic interferences, which is a well-known problem in ICP-MS analysis of these elements (Evans and Giglio, 1993; May and Wiedmeyer, 1994; Reed et al., 1994). Relative to the working values of Govindaraju (1994, 1995) for the analyzed reference materials, the measured Sc concentrations are 47–93% too high at low concentrations (< 10 ppm), whereas they are much closer (≤ 20%) at elevated concentrations (> 50 ppm). Numerous polyatomic ions with a mass over charge ratio (m/z) of 45 could interfere on 45Sc (e.g., Whitty-Léveillé et al., 2016) and may thus contribute to this overestimation. They include silicon-based ions (29Si16O+, 28Si17O+, 28Si1H16O+) and ions that may be produced from lithium- 7 38 + 1 16 11 + 1 16 10 + borate constituents (e.g., Li Ar , H2 O2 B , H3 O2 B ). In addition, doubly charged Zr atoms (90Zr2+) may add to the interference problem on 45Sc as well. The high Zr concentrations in the bauxite standards (427–2882 ppm) and a clear correlation between Zr and %dev. values for Sc in our results suggests that this might be the case. Whitty-Léveillé et al., (2016) used mass filtering techniques in tandem quadrupole ICP-MS to reduce Si, B and Zr-based interferences on 45Sc. Similar to our results, they measured a high apparent Sc concentration of 76 ± 5 ppm for lithium-metaborate-fused BX-N by conventional quadrupole solution ICP-MS, but obtained an accurate and precise result (60 ± 2 ppm) relative to the expected value of Govindaraju (1994) after interference removal. Largest deviations from expected values at lowest concentrations are also observed for V (%dev. > 30 at < 50 ppm) and Cr (%dev. > 30 at < 70 ppm), both in FeR-2 (Table 3.7). Possible interference effects include those of11 B40Ar+ on 51V+ and Ar-based species (e.g., 36Ar16O+) on 52Cr.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 59 60 | Chapter 3

Figure 3.2 | Comparison of LA-ICP-MS results with reference values, INAA data and solution ICP-MS data for international reference bauxites (ANRT BX-N, NIST-SRM 69b, NIST-SRM 696, and NIST-SRM 698) and iron formation (FeR-2) (see Tables 3.3–3.7, and text for discussion). Labeled open symbols indicate measurement results with relatively large deviations (> 20%) from comparison values. Plotted results for measurements on NIST-SRM 698 are based on Ce as internal standard (value from Govindaraju, 1994) instead of Si.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 60 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 61

Figure 3.3 | Trace-element abundances of HFSE and REE, normalized to upper continental crust (UCC, McLennan, 2001) in selected samples from four bauxite deposits in Suriname (see Monsels, 2016 for locations) to illustrate the applicability of LA-ICP-MS on lithium borate glass beads for bauxite analysis. (a) Samples from the top (squares) and bottom parts (circles) of bauxite profiles on Proterozoic crystalline parent rocks (see text), data from Monsels and Van Bergen (2017); (b) Samples from the top (squares) and bottom parts (circles) of bauxite profiles on Tertiary sediment. Depth between brackets. Data from Monsels and Van Bergen (in prep.).* Note that La concentrations in the Nassau samples are overestimated due to impurity of the borate flux.

The results for As agree well with literature data except for a ca. 50% overestimation relative to the provisionally recommended low concentration of 2 ppm in FeR-2. Other well- known interferences (e.g., oxides of Ba and light REE on middle and heavy REE) were probably insignificant, given the overall good correspondence between measured and expected values for REE that are sensitive to this problem. Substantial overestimation of La at low expected concentrations (e.g., 60–90% at 12 ppm for FeR-2) is probably attributable to impure flux material. We noticed that batches of borate flux may contain detectible amounts of lanthanum despite its absence on the list of contaminants provided by the manufacturer. Test measurements on borate flux blanks confirmed the presence of small amounts of La. Similar problems from flux impurities were reported by Eggins (2003)

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 61 62 | Chapter 3

and Yu et al., (2003). Low Pb concentrations compared to expected values (%dev. values from -4 to -31) are possibly due to volatilization during fusion. Petrelli et al., (2007) suggested that the

amount of Pb loss could be related to the SiO2 content of the sample.

3.3.3 | Application to Surinamese bauxites Lateritic bauxite deposits in Suriname formed on a large diversity of precursor rocks, ranging from Proterozoic metamorphic crystalline basement rocks on plateaus in the country’s interior to Tertiary sediments in coastal lowlands (Bárdossy and Aleva, 1990; Monsels, 2016) (Chapters 4 and 5). Trace- element abundances (Monsels and Van Bergen, 2017, in prep.) tend to be unique for each deposit as is illustrated for REE and HSFE distributions normalized to upper continental crust values (McLennan, 2001) in Figure 3.3 (Chapters 4 and 5). Samples taken at two different depth levels in each of the vertical bauxite profiles display largely coherent patterns but different concentrations. The patterns of the two bauxites that developed on the Proterozoic parent rocks are clearly different (Figure 3.3a). The Nassau profile on a low-grade metabasalt/ andesite is marked by strong enrichments of HFSE relative to REE and heavy REE (HREE) relative to light REE (LREE). In contrast, the samples from the Snesie profile, developed on a high-grade metamorphic gneissic precursor, show elevated LREE/HREE ratios, a strong enrichment of HREE over HFSE and a conspicuous positive Eu anomaly. Likewise, bauxites from the two profiles on Tertiary sediments are compositionally distinct (Figure 3.3b). The REE part of the patterns shows a concave upward trend in the Kaaimangrasie samples and a concave downward trend in the Klaverblad samples, whereas in both cases the HFSE are enriched and concentration differences between samples from different depths are much larger for REE than for HFSE (Chapters 4 and 5). A combination of factors is inferred to be responsible for distinct trace-element signatures observed, including the lithology of the parent rock, the history and controls of the weathering process, mobility of the REE and the nature of mineral hosts (Monsels and Van Bergen, 2017). These examples demonstrate the versatility of the routine described here for trace-element analysis in bauxite studies.

3.4 | Conclusions

Laser ablation ICP-MS analysis of lithium borate glass beads has been applied on international reference materials (ANRT-BX-N, NIST-SRM 69b, NIST-SRM 696, NIST-SRM 698 and CCRMP-FeR-2) to test the reliability of this method for measuring trace-element concentrations in bauxite. Based on a comparison of the obtained results with reference values and literature data, we infer that the used technique enables a fast, accurate and precise determination of trace elements in bauxite samples, in particular the HSFE and REE. The lithium borate fusion procedure avoids sample decomposition problems associated with acid dissolution of refractory minerals that are commonly present in bauxites. Since no additional sample preparation is needed, the method is an efficient means to obtain a complete set of major and trace-element data on the same sample in conjunction with XRF, which can also serve to provide concentration data for internal standards. Application of methods to reduce polyatomic interference effects may be advisable to improve measurement accuracy

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 62 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 63

for elements that are sensitive to this problem such as Sc, V and Cr. The possible presence of impurities in lithium borate flux also requires attention. Future analytical work on trace elements in bauxites would benefit from the availability of more and better characterized reference materials.

Acknowledgements The authors would like to thank Helen de Waard for help with the LA-ICP-MS work at Utrecht University. This research was funded by a grant from the Suriname Environmental and Mining Foundation (SEMIF).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 63 64 | Chapter 3

References

– Abendini, A. and Calagari, A. (2014). REE geochemical characteristics of titanium-rich bauxites; the Permian Kanigorgeh horizon, NW Iran. Turkish Journal of Earth Sciences, 23, 513-532. – Ahmadnejad, F., Zamanian, H., Taghipour, B., Zarasvandi, A., Buccione, R. and Ellahi, S. (2017). Mineralogical and geochemical evolution of the Bidgol bauxite deposit, Zagros Mountain Belt, Iran: Implications for ore genesis, rare earth elements fractionation and parental affinity. Ore Geology Reviews 86, 755-783. – Amosova, A., Panteeva, S., Chubarov, V. and Finkelshtein, A. (2016). Determination of major elements by wavelength-dispersive X-ray fluorescence spectrometry and trace elements by inductively coupled plasma spectrometry igneous rocks from the same fused sample (110mg). Spectrochimica Acta Part B: Atomic Spectrometry 122, 62-68. – Andrade, S., Ulbrich, H., de Barros Gomes, C., and Martins, L. (2014). Methodology for the determination of trace and minor elements in minerals and fused rock glasses with laser ablation associated with quadrupole inductively coupled plasma mass spectrometry (LA-Q-ICPMS). American Journal of Analytical Chemistry 5 (11), 701. – Awaji, S., Nakamura, K., Nozaki, T. and Kato, Y. (2006). A simple method for precise determination of 23 trace elements in granitic rocks by ICP-MS after Lithium Tetraborate fusion. Resource Geology 56 (4), 471-478. – Bédard, L. and Barnes, S. (2002). A comparison o f N-type semi-planar and coaxial INAA detectors for 33 geochemical reference samples. Journal of Radioanalytical and Nuclear Chemistry 254 (3), 485-497. – Borra, C., Pontikes, Y., Binnemans, K. and Van Gerven, T. (2015). Leaching of rare earths from bauxite residue (red mud). Minerals Engineering 76, 20-27. – Butt, C. (1986). Vertical distribution of trace elements in laterite soil (Suriname-Discussion). Chemical Geology 56, 159-163. – Chen, S., Wang, X., Niu, Y., Sun, P., Duan, M., Xiao, Y., Guo, P., Gong, H., Wang, G. and Xue, Q. (2017). Simple and cost-effective methods for precise analysis of trace element abundances in geological materials with ICP-MS. Science Bulletin 62, 277-289. – Cotta, A. and Enzweiler, J. (2011). Classical and new procedures of Whole Rock dissolution for trace element determination by ICP-MS. Geostandards and Geoanalytical Research 36 (1), 27-50. – De Madinabeita, G., Lorda, S. and Ibarguchi, G. (2008). Simultaneous determination of major to ultratrace elements in geological samples by fusion-dissolution and inductively coupled plasma mass spectrometry techniques. Analytica Chimica Acta 625 (2), 117-130. – Eggins, S. (2003). Laser Ablation ICP-MS Analysis of Geological Material prepared as lithium borate glasses. Geostandards and Geoanalytical Research 27 (2), 147-162. – Evans, E. and Giglio, J. (1993). Interferences in inductively coupled plasma mass spectrometry. A review. Journal of Analytical Atomic Spectrometry 8 (1), 1-18. – Gao, S., Lui, X., Hattendorf, B., Gunter, D. and Hu, S. (2002). Determination of forty two major elements in USGS and NIST SRM glasses by laser ablation inductively coupled plasma mass spectrometry. Geostandards Newsletter 26, 181-196. – Govindaraju, K. (1994). 1994 compilation of working values and sample description for 383 geostandards. Geostandards Newsletter 18, 158 pp. – Govindaraju, K. (1995). 1995 Working values with confidence limits for twenty-six CRPG, ANRT and IWG-GIT geostandards. Geostandards Newsletter 19 (1), 1-32. – Grant, C., Lalor, C. and Vutchkov, M. (2005). Comparison of bauxites from Jamaica, the Dominican Republic and Suriname. Journal of Radioanalytical and Nuclear Chemistry 266 (3), 385-388. – Gu, J., Huang, Z., Fan, H., Ye, L. and Jin, Z. (2013). Provenance of lateritic bauxite deposits in the Wuchuan- Zheng’an-Daozhen area, Northern Guizho Province, China: LA-ICP-MS and Sims U-Pb dating of detrital zircons. Journal of Asian Earth Sciences 70, 265-282.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 64 Trace-element analysis of bauxite with LA-ICP-MS on lithium borate glass beads | 65

– Günter, D., v.Quadt, A, Wirz, R., Cousin, H. and Volker, D. (2001). Elemental analyses using Laser Ablation-

Inductively Coupled Plasma- Mass Spectrometry (LA-ICP-MS) of geological samples fused with Li2B4O7 and calibrated without matrix matched standards. Mikrochimica Acta 136, 101-107. – Jochum, K., Willbold, M., Stoll, B. and Herwig, K. (2005). Chemical characterization of the USGS reference glasses GSA-1G, GSC-1G, GSD-1G, GSE-1G, BCR-2G, BHVO-2G and BIR-1G using EPMA, ID-TIMS, ID-ICP-MS and LA-ICP-MS. Geostandards and Geoanalytical Research 29 (3), 285-302. – Jochum, K., Stoll, B., Herwig, K. and Willbold, M. (2007). Validation of LA-ICP-MS trace element analysis of geological glasses using a new solid state 193 nm Nd: YAG laser and matrix-matched calibration. Journal of Analytical Atomic Spectrometry, 22, 112-121. – Jochum, K., Weis, U., Schwager, B., Stoll, B., Wilson, S., Haug, G., Andreae, M. and Enzweiler, J. (2016). Reference values following ISO guidelines for frequently requested rock reference materials. Geostandards and Geoanalytical Research 40 (3), 333-350. – Korotev, R., (1996). A Self- consistent compilation of elemental concentration data for 93 geochemical reference samples. Geostandards Newsletter 20 (20), 217-245. – Leite, T, Escalfoni, R., da Fonseca, T. and Miekely, N. (2011). Determination of major, minor and trace elements in rock samples by laser ablation inductively coupled plasma spectrometry: Progress in the utilization of borate glasses as targets. Spectrochimica Acta Part B: Atomic Spectroscopy 66 (5), 314-320. – Longerich, H., Jenner, G., Fryer, B. and Jackson, S. (1990). Inductively coupled plasma mass spectrometric analysis of geological samples: A critical evaluation based on case studies. Chemical Geology 83, 105-118. – Mason, P. and Kraan, W. (2002). Attenuation of spectral interferences during laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) using an rf only collision and reaction cell. Journal of Analytical Atomic Spectrometry 17 (8), 858-867. 3 – McLennan, S. (2001). Relationships between trace element composition of sedimentary rocks and upper continental crust. Geochemistry, Geophysics, Geosystems 2 (4). – Miliszkiewicz, N., Walas, S. and Tobiasz, A. (2015). Current approaches to calibration of LA-ICP-MS analysis. Journal of Analytical Atomic Spectrometry 30 (2), 327-338. – Mongelli, G., Boni, M., Buccione, R. and Sinisi, R. (2014). Geochemistry of the Apulian karst bauxites (southern Italy). Chemical fractionation and parental affinities. Ore Geology Reviews 63, 9-21. – Monsels, D.A. (2016). Bauxite deposits in Suriname: geological context and resource development. Netherlands Journal of Geosciences. Geologie en Mijnbouw 95 (4), 405-418. – Monsels, D.A and Van Bergen, M.J. (2017). Bauxite formation on Proterozoic bedrock in Suriname. Journal of Geochemical Exploration 180, 71-90. – Monsels, D.A and Van Bergen, M.J. (in prep.). Bauxite formation on Tertiary sedimentary parent rocks in Suriname. – Nesbitt, R., Hirata, T., Butler, I. and Milton, J. (1997). UV laser ablation ICP-MS: Some allocations in the earth sciences. Geostandards Newsletter 20, 231-243. – Ochsenkühn-Petropoulou, M., Ochsenkühn, K. and Luck, J. (1990). Comparison of inductively coupled plasma mass spectrometry with inductively coupled plasma atomic emission spectrometry and instrumental neutron activation analysis for the determination of rare earth elements in Greek bauxites. Spectrochimica Acta, 46 (1), 51-65. – Ochsenkühn-Petropoulou, M., Lyberopulu, T. and Parissakis, G. (1994). Direct determination of Lanthanides, yttrium and scandium in bauxites and red mud from alumina production. Analytica Chimica Acta 296 (3), 305-313. – Orihashi, Y. and Hirata, T. (2003). Rapid quantitative analysis of Y and REE abundances in XRF glass bead for selected GSJ reference rock standards using Nd-YAG 266 nm UV laser ablation ICP-MS. Geochemical Journal 37, 401-412. – Panteeva, S., Gladkochoub, D., Donskaya, T., Markova, V. and Sandimirova, G. (2003). Determination of 24 trace elements in felsic rocks by inductively coupled plasma mass spectrometry after lithium metaborate fusion. Spectrochimica Acta Part B: Atomic Spectroscopy 58 (2), 341-350.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 65 66 | Chapter 3

– Paramguru, R., Rath, P. and Misra, V. (2004). Trends in red mud utilization- a review. Mineral Processing and Extractive Metallurgy Review 26 (1), 1-29. – Park, J., Al-Mutairi, A., Yoon, S., Mochida, I. and Ma, X. (2016). The characterization of metal complexes in typical Kuwait atmospheric residues using both GPC coupled with ICP-MS and HT GC-AED. Journal of Industrial and Engineering Chemistry 34, 204-212. – Petrelli, M., Perugini, D., Poli, G. and Peccerillo, A. (2007). Graphite electrode lithium tetraborate fusion for trace element determination in bulk geological samples by laser ablation ICP-MS. Microchimica Acta 158, 275-282. – Reed, N., Cairns, R., Hutton, R. and Takaku, Y. (1994). Characterization of polyatomic ion interferences in inductively coupled plasma mass spectrometry using a high resolution mass spectrometry. Journal of Analytical Spectrometry 9 (8), 881-896. – Regnery, J., Stoll, B. and Jochum, K. (2009). High-resolution LA-ICP-MS for accurate determination of low abundance of K, Sc and other trace elements in geological samples. Geostandards and Geoanalytical Research 34, 19-38. – Shazzo, Y. and Karpov, Y. (2016). Laser sampling in inductively coupled plasma mass spectrometry in the inorganic analysis of solid sampled: Elemental fractionation as the main source of errors. Journal of Analytical Chemistry 71 (11), 1069-1080. – Topp, S., Salbu, B., Roaldset, E. and Jørgensen, P. (1984). Vertical distribution of trace elements in laterite soil (Suriname). Chemical Geology, Elsevier Science Publishers B.V., Amsterdam 47, 159-174. – Totland, M. and Jarvis, K. (1993). Determination of the platinum-group elements and gold in solid samples by slurry nebulisation ICP-MS. Chemical Geology 104 (1-4), 175-188. – Ujaczki, E., Zimmermann, Y., Gasser, C., Molnár, M., Feigl, V. and Lenz, M. (2017). Red mud as secondary source for critical raw materials-extraction study. Journal of Chemical Technology and Biotechnology. – Vukotić, P. (1983). Determination of rare earth elements in bauxites by instrumental neutron activation analysis. Journal of Radioanalytical and Nuclear Chemistry 78 (1), 105-115. – Whitty-Léveillé, L., Drouin, E., Constatin, M., Bazin, C. and Larivière, D. (2016). Scandium analysis in silicon- containing minerals by inductively coupled plasma tandem mass spectrometry. Spectrochimica Acta Part B: Atomic Spectroscopy 188, 112-118. – Yu, Z., Norman, M. and Robinson, P. (2003). Major and trace element analysis of silicate rocks by XRF and laser ablation ICP-MS using lithium borate fused glasses: matrix effects, instrument response and results for international reference materials. Geostandards Newsletter 27, 67-89. – Zhang, J., Deng, Z. and Xu, T. (2005). Experimental investigations on leaching metals from red mud. Light Metals 2, 13-15. – Zhang, W., Qi, L., Hu, Z., Zheng, C., Lui, Y., Chen, H., Gao, S. and Hu, S. (2016). An investigation of digestion methods for trace elements in bauxite and their determination in ten bauxite reference materials using inductively coupled plasma mass spectrometry. Geostandards and Geoanalytical Research 40 (2), 195-216.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 66 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 67 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 68 Chapter 4 Bauxite formation on Proterozoic bedrock of Suriname

This chapter has been published as: Monsels, D.A. and Van Bergen, M.J. (2017). Bauxite formation on Proterozoic bedrock in Suriname. Journal of Geochemical Exploration 180, 71-90.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 69 70 | Chapter 4

Abstract

Lateritic bauxite deposits in Suriname rest on a variety of metamorphic, igneous and sedimentary parent rocks. Remnants of multiple planation surfaces with duricrusts that mark the tropical landscape are associated with recurrent episodes of bauxite formation since Late Cretaceous times. Plateau-type bauxites at the highest topographic levels developed on a range of Proterozoic crystalline bedrocks on the northern edge of the Guiana Shield in the country’s interior. Lateritic weathering profiles of the Bakhuis Mountains, Nassau Mountains, Lely Mountains and Brownsberg largely correspond to the classical sequence of an iron-rich cap on top of a bauxite layer that covers a clay-rich saprolite-interval grading into weathered and fresh bedrock. All of the investigated profiles are consistent with in-situ formation of the bauxite and are marked by Si-Al-Fe relationships indicative of medium to strong lateritization, with fresh bedrock being poorly exposed. The bauxite deposits contain gibbsite as the dominant Al- bearing phase, whereas boehmite is locally present in subordinate quantities. Their ferruginous character is expressed by relatively abundant goethite and hematite in the top layers. Kaolinite is the main mineral in the saprolite. Anatase and zircon are the most detected minor phases.

The investigated bauxite deposits are generally of a medium-grade (average Al2O3 contents 33–49%), but have variable chemical compositions according to exploration drilling results. Despite this overall conformity of the deposits, their thicknesses, textures, mineralogy and geochemistry are distinct in detail, reflecting contrasts in the nature of the parent rock and

weathering history. Inter-element relationships show conspicuous differences between SiO2-

poor (< 5%) upper parts and SiO2-richer (> 5%) lower parts of profiles in the Bakhuis Mountains, which developed on high-grade metamorphic pyroxene amphibolites and gneissic granulites. The bauxites of the Nassau Mountains and the other areas in eastern Suriname are marked by

higher TiO2 contents (average > 3.9%) and dissimilar profiles that reflect their development on variety of a low-grade metamorphic volcanic parent rocks. Weathering-induced redistribution of rare earth and other trace elements affected even the least mobile elements. Differences in distribution patterns between individual profiles can be attributed to a combination of primary compositional differences of parent rocks, the nature and content of accessory mineral phases, and unequal responses to multiple bauxitization cycles.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 70 Bauxite formation on Proterozoic bedrock of Suriname | 71

4.1 | Introduction

Lateritic bauxites of economic interest have formed on a range of different sedimentary, igneous and metamorphic parent rocks worldwide. A sufficiently high aluminium content together with favourable physical rock properties and weathering conditions determine the potential of a bauxite deposit as raw material for the aluminium industry (Valeton, 1972; Schellmann, 1983; Patterson et al., 1986; Bárdossy and Aleva, 1990; Aleva, 1994; Tardy, 1997). Widespread lateritic bauxite deposits on the Amazonian Craton and bordering coastal regions of South America, formed by extensive weathering in Late Cretaceous–Early Tertiary times (Bárdossy and Aleva, 1990). Because a warm and humid climate is a critical requirement for their origin, the lateritic weathering profiles of these South American bauxites show similarities with equivalent regolith profiles in other low-latitude (paleo-) tropical or (paleo-) equatorial regions in Africa, South Asia and Australia (Bárdossy and Aleva, 1990; Mutakyahwa and Valeton, 1995, and references therein; Tardy and Roquin, 1998; Bogatyrev et al., 2009; Horbe and Anand, 2011). All these deposits are distributed on tectonically stable post-Gondwanan continental landmasses. In Suriname, bauxite developed on a variety of parent rocks ranging from Tertiary sedimentary rocks to Precambrian igneous and metamorphic basement lithologies. Mining activities commenced about a century ago (Aleva and Wong, 1998) and have been restricted to deposits in the coastal lowlands where bauxite was formed from sedimentary parent rock (Aleva, 1965, 1979). Together with a refinery industry, the exploitation of these bauxite resources made Suriname one of the world’s leading exporters of bauxite and alumina around the 1950s (Patterson et al., 1986; Bárdossy and Aleva, 1990; Monsels, 2016) (Chapter 2). In contrast, bauxites on plateaus in Suriname’s interior, which originated on Proterozoic crystalline parent rocks of the Guiana Shield, have not been productive so far, although they have been comprehensively explored for their economic potential (e.g., Doeve, 1955; Van Kersen, 1956; 4 Krook and De Roever, 1975; Pollack, 1981, 1983; Aleva and Hilversum, 1984). Studies targeting vertical distributions of major and trace elements together with mineralogy in weathering profiles of bauxite deposits developed on Precambrian igneous rocks of the Guiana Shield are scarce. Some examples concern the Los Pijiguaos bauxite deposit, Venezuela, which developed on granite (Meyer et al., 2002), and deposits on granitic and volcanic bedrocks in Brazilian Amazonia (Horbe and Da Costa, 1999; Horbe and Anand, 2011). Similar work on Suriname’s plateau bauxites is limited to early studies of Pollack (1981, 1983) and Topp et al. (1984). The objective of this paper is to provide new insights into the weathering processes that created the plateau bauxites in Suriname’s interior from a variety of Precambrian crystalline bedrocks. Vertical mineralogical, textural and geochemical variations in lateritic weathering profiles are presented to discuss critical processes in the origin of the bauxite duricrusts and to assess the effect of parent rock properties.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 71 72 | Chapter 4

4.2 | Geological setting

4.2.1 | Bauxites in Suriname Suriname is part of the Guiana Shield in the north-eastern section of South America. Proterozoic rocks make up about 80% of the country (crystalline basement) and the remaining 20% consists of Cretaceous to Recent sediments that were deposited along the northern fringe of the craton (coastal area) (Figure 4.1a). The Proterozoic basement of Suriname consists of three metamorphic belts: the low-grade Marowijne Greenstone Belt in the NE and the high-grade Bakhuis Granulite Belt and the Coeroeni Gneiss Belt in the NW and SW, respectively (Kroonenberg et al., 2016). These belts are separated by a large area consisting of various types of granitoid rocks and felsic volcanic rocks in the central part of the country. The basement is overlain by a remnant of the Proterozoic Roraima Formation in the Tafelberg Sandstone Plateau, and transected by Proterozoic and Early Jurassic dolerite dikes (Figure 4.1a). Surinamese bauxite deposits have formed on two different types of parent rock (Figure 4.1a; see also Monsels, 2016 and references therein): 1. Crystalline rocks in the hinterland (plateau or highland bauxites). As part of the Guiana Shield Subprovince (Van Kersen, 1956; Janssen, 1979; Bárdossy and Aleva, 1990), these bauxites are mostly developed on intermediate to mafic Precambrian crystalline igneous and metamorphic rocks. 2. Sedimentary rocks in the coastal area (coastal-plain or lowland bauxites). These bauxites belong to the Guiana Coastal plain Subprovince and they are formed on an accumulation of continental sediments from Early Cenozoic times (Van der Hammen and Wijmstra, 1964; Valeton, 1983; Aleva and Wong, 1998; Wong et al., 1998).

Both groups of deposits are part of a sequence of five bauxitic or Fe-lateritic surface levels that have been distinguished in Suriname and neighbouring territories (King et al., 1964; Pollack, 1983; Aleva, 1984; Bárdossy and Aleva, 1990; Tardy and Roquin, 1998; cf. Figure 4.1b and Chapter 1). The succession of planation surfaces developed in different episodes between Late Cretaceous and Quaternary times, with the Early Tertiary as the most prominent interval of bauxite formation. The bauxite deposits studied here are the typical plateau bauxites of the Bakhuis Mountains, Nassau Mountains, Lely Mountains and Brownsberg (Figure 4.1a), which formed on a variety of metamorphosed crystalline parent rocks in this part of the Precambrian Guiana Shield, including (ultra)mafic to intermediate metavolcanics and greenschists.

4.2.2 | Local geology of plateau bauxites 4.2.2.1 | Bakhuis Mountains bauxite district The Bakhuis Mountains (04000’–05000’N, 56030’–57030’W), which form a chain of strongly dissected plateaus, are an expression of a 25 km wide and 95 km long NE-SW striking horst, one of the many structural features of the Bakhuis-Kanuku zone in the heart of the Guiana Shield (Kroonenberg and De Roever, 1975; Kroonenberg, 1976; Kroonenberg and Melitz, 1983; Kroonenberg et al., 2016). Hilltops reach heights of approximately + 480 m a.s.l. The climate is tropical humid with an average temperature of 27.50C and 1700–2200 mm of rainfall, divided

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 72 Bauxite formation on Proterozoic bedrock of Suriname | 73

Figure 4.1 | (a) Geological map of Suriname (modified after Kroonenberg et al., 2016) and bauxite districts. Areas 1 and 2 represent groups of plateau bauxites, which originated on various metamorphosed crystalline rocks of Precambrian age; those studied here are indicated. Areas 3 and 4 represent groups of Coastal-plain bauxite deposits, formed on Early Cenozoic sedimentary parent rocks; most of them are former mining areas; (b) Locations and types of bauxite deposits of the Guiana Shield (modified after Bárdossy and Aleva, 1990).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 73 74 | Chapter 4

Figure 4.2 | (a) Geological map of the Bakhuis Mountains horst (modified after Klaver et al., 2015). The study locations Bakhuis Base Camp (exploration Area 10.1) and Snesie and Macousi (Area 10.2) are indicated on the right panel, which represents a magnification of the hyphened rectangle on the left map; (b) Elevation map of the Nassau Mountains (modified after Van den Bergh, 2011). Black hyphened rectangle indicates the study area on Plateau C; (c) Laterite- and bauxite-capped mountains and hills in NE Suriname (Nassau bauxite district) (modified after Van Lissa, 1975; SPS and OAS, 1988); Ter Steege et al., 2006).

over a long and a short rainy season (Chapter 2). The basement of the Bakhuis Mountains consists of the high-grade metamorphic Falawatra Group, which includes banded rocks from a charnockite suite, sillimanite gneisses and clinopyroxene amphibolites (Figures 4.1a, 4.2a) (De Roever et al., 1976, 2003; Klaver et al., 2015; Kroonenberg et al., 2016). Zircon ages of the charnockites range from 1984.4 to 1992.5 Ma, while a zircon from a leucosome, formed under ultra-high temperature (UHT) conditions, revealed an age of 2172.6 ± 7.3 Ma (Klaver et al., 2015).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 74 Bauxite formation on Proterozoic bedrock of Suriname | 75

Variable compositions of bauxite weathering profiles reflect the large diversity of parent rocks (Aleva and Hilversum, 1984), which, together with differences in drainage conditions is also responsible for variability in thickness of the bauxite bodies. Locally, lenses of kaolinite- rich material occur within the bauxite horizon, and lenses and boulders of bauxite material within the kaolinitic saprolite (Pollack, 1981; Aleva and Hilversum, 1984) (Chapter 2). Different topographic levels of the Bakhuis duricrusts suggest two distinct events of bauxite formation. Bárdossy and Aleva (1990) attributed an Eocene age to the top surface (which includes the investigated Area 10 lateritic weathering profiles) and a Miocene age to the lower surface, while Pollack (1983) assigned a Late Cretaceous and Eocene age, respectively. A paleomagnetic age of 60 ± 20 Ma (Théveniaut and Freyssinet, 2002) is grossly consistent with the interpretation of Pollack (1983) and an earlier proposed age of 70 Ma (Billiton, 1979). The combined proven

bauxite reserves for Area 10 and 5 are 70 Mt at 43.8% Al2O3 and 2% SiO2 (Janssen, 1963), while the resources of the entire Bakhuis Mountain are estimated to be larger than 500 Mt at an average of 34% available alumina and 2% reactive silica (Bauxite Institute Suriname, 2009).

4.2.2.2 | Nassau Mountains District The Nassau bauxite district in eastern Suriname includes the bauxite- and/or laterite-covered high plateaus of the Nassau Mountains, Lely Mountains, and Brownsberg around the Van Blommenstein Lake (Bárdossy and Aleva 1990, and references therein) (Figures 4.1a, 4.2b-c). The mountains belong to the NW-SE striking Marowijne Greenstone Belt for which zircon ages between 2.26 and 2.10 Ga have been reported (Delor et al., 2003; Klaver et al., 2015 Kroonenberg et al., 2016) (Figure 4.1a) (Chapter 2). The belt predominantly consists of low- grade metamorphic mafic to intermediate volcanics, sedimentary rocks and granitoids (Priem et al., 1971, 1980; Bosma et al., 1984; De Vletter, 1984; De Vletter et al., 1998; Kroonenberg et al., 2016). All of the bauxite-capped mountains of this group are made up of rocks from the 4 Paramaka Formation, which mainly consists of greenschist-facies metabasalts and also contains meta-gabbros, meta-andesites, meta-cherts and other intermediate and felsic meta-volcanic rocks (Figure 4.1a) (Janssen, 1977; Kroonenberg and Melitz, 1983; De Vletter, 1984; De Vletter et al., 1998; Bárdossy and Aleva, 1990; Kroonenberg et al., 2016). The plateaus have all been tentatively associated to a Late Cretaceous or Early Tertiary planation surface (King et al., 1964). The Nassau Mountains (04046’–04056’N, 54030’–54038’W) form an isolated, U-shaped mountain ridge (20 x 20 km2), which is bordered by the Van Blommenstein Lake to the west and the Marowijne River to the east (Figures 4.1b, 4.2b-c). The ridge comprises four steep-sided, laterite-capped plateaus (A-D) at elevations between 500 and 564 m a.s.l. and they receive 2750–3000 mm rainfall per year. (Bárdossy and Aleva, 1990; Ter Steege et al., 2006; Bánki et al., 2008) (Chapter 2). The ridge is covered with high dry-land forest and small areas of wet palm- swamp forest in depressions on the plateaus, and with “mountain savanna” on the northeastern edge of the plateau (Ter Steege et al., 2006; Bánki et al., 2008). Swamps forms in local depressions on Plateau C when rainfall exceeds the infiltration capacity of the local unsaturated zone during wet seasons (Doeve, 1955). A shallow, unconfined aquifer system covers an area of 6.5 km2 in the porous bauxite (Ouboter et al., 2007). A Late Cretaceous to Early Tertiary age was assigned to the Nassau bauxite deposit by Bárdossy and Aleva (1990), whereas Aleva and Wong (1998) favoured an Eocene–Oligocene age. Doeve (1955) and Van Kersen (1956) characterized the deposit as a multicolored, up to 10 m thick bauxite with variable hardness, showing aphanitic,

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 75 76 | Chapter 4

ooïdal, pisolitic, gravelly and breccia-like textures. The bauxite beds are covered by a dark, up to 5 m thick Fe-rich duricrust, which is underlain by an up to 15 m thick saprolite composed of soft clayey material. The contact with the saprolite is gradational or sharp, whereas in certain areas, bauxite can also be absent so that locally the duricrust directly overlies the saprolite (Doeve, 1955; Bárdossy and Aleva, 1990). It has been estimated that the Nassau deposit contains between 32–60 Mt of bauxite resources (Bauxite Institute Suriname, 2009). The most promising areas for mining are Plateaus A and C (Figure 4.2b). The Brownsberg or Browns Mountain (04046’–05000’N, 55008’–55015’W), situated on the western border of the Van Blommenstein Lake (Figure 4.2c), is approximately 34 km long and has a maximum width of roughly 13.5 km. The rugged surrounding terrain is also known as the “Brokolonko landscape” (Kroonenberg and Meltiz, 1983). The Brownsberg bauxite predominantly occurs on the main plateau as part of a significantly larger laterite cap that covers a narrow plateau, at approximately 500 m a.s.l. Its resources range between 8.7 and 14.8 Mt (Bauxite Institute Suriname, 2009). The Lely Mountains (04013’–04029’N, 54035’–54046’W), located near the confluence of the Tapanahony River and Lawa River, are composed of a series of 600 to 700 m high plateaus arranged in an arcuate shape (Figure 4.2c). The vegetation mainly consists of high dryland tropical rainforest on the plateaus and slopes, while mountain savannah is only observed on the plateaus. The laterite-capped plateaus are dissected by incisions of several surrounding

creeks. The bauxite reserves are between 7–15 Mt at contents of 46–50% Al2O3 and 1.4–1.5%

reactive SiO2, while estimated resources range between 15 and 27 Mt (Janssen, 1963, 1979; Bauxite Institute Suriname, 2009). The bauxite deposit of the Lely Mountains has long been considered unattractive for processing due to its remote location, relatively high boehmite content and the scattered distribution of small pockets of higher-quality bauxite over the up to 40 km long plateaus (Janssen, 1963, 1979).

4.3 | Materials and methods

4.3.1 | Sample locations and studied materials The studied sites in the Bakhuis Mountains are the BHP Billiton/Suralco L.L.C. Bakhuis Base Camp (Area 10.1), the Snesie and the Macousi exploration trenches (Area 10.2), while those in the Nassau Mountains are situated on Plateau C (Figures 4.2a-c). Lateritic weathering profiles of the Snesi and Macousi trenches were sampled down to a depth of 7.5 m at vertical intervals of 50–100 cm, based on variations in the lithology (Figure 4.3). Ten drill cutting samples were collected from Snesie drill hole SNE-10-04 and from Macousi drill hole MAC-10-01. The deepest interval at Snesi (7.0–7.5 m) had not been recovered during drilling. The Bakhuis Base Camp samples were taken from an outcrop of a high-grade metamorphic rock considered to be the potential parent material of laterite/bauxite in the surrounding area. Sample BAK-02 is a relatively fresh piece of crystalline rock, whereas sample BAK-01 is a strongly weathered equivalent. Samples from Nassau were drill cuttings from drill hole PC-10-0131 on Plateau C, which were kindly provided by Suralco L.L.C. Major element XRF data for deposits from Brownsberg, Lely and Nassau Mountains were provided by the Bauxite Institute Suriname.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 76 Bauxite formation on Proterozoic bedrock of Suriname | 77

Figure 4.3 | Bauxite-bearing lateritic weathering profile of (a) the Snesie drill hole (SNE-10-04), (b) the Macousi drill hole (MAC-10-01) and (c) the Nassau drill hole (PC-10-0131).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 77 78 | Chapter 4

4.3.2 | Analytical methods Mineralogy and microstructures were investigated on polished thin sections of drill cuttings, and rock samples with an optical microscope and an electron microprobe (JEOL JXA-8600 Superprobe) using both energy dispersive (EDS) and wavelength-dispersive (WDS) analytical techniques. Back-scatter electron imaging (BSE) was used to identify mineral phases and to study textural relationships. Quantitative compositions of mineral phases were determined in representative textural domains of selected samples from the Snesie Trench, Bakhuis Base Camp and Nassau. X-ray diffraction (XRD) patterns were collected from randomly oriented powder samples using a Bruker D2 Phaser X-ray diffractometer, operated in a step-scan mode, with Co-Kα radiation (1.78897 Å). The counting time was 66 sec/step, the step size 0.050 and the range 5–850. Total acquisition time per sample was approximately 15 minutes. Major-element compositions were determined by X-ray fluorescence (XRF) on fused glass beads (lithium borate) with a Thermo ARL 9400 sequential XRF (Utrecht University) and a Panalytical MagiXPro XRF (VU University Amsterdam). Loss on ignition (LOI) data were obtained either by measuring weight loss upon heating of a powdered sample in an oven at > 1000 0C or by thermogravimetric analysis (TGA) during which, weight loss was continuously monitored over a temperature range between room temperature and 1000 0C. The major element data provided by Suralco L.L.C. and Bauxite Institute Suriname include values for loss on ignition

(LOI), total alumina (TAl2O3), available alumina (AA143), total silica (TSiO2) and reactive silica

(RSiO2). Trace elements were determined by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) on the fused glass beads prepared for XRF, using a ThermoFischer Scientific Element 2 magnetic sector instrument, integrated with a Lambda Physik excimer laser (193 nm) with GeoLas optics (Chapter 3). Main parameters for the ablation spot setup were: 5 mJ laser energy, 10Hz pulse repetition rate and 120µm spot diameter. The ICP-MS operating conditions were plasma power: 1300W; gas flow rates (L/min): cool 16.0 Ar, auxiliary 1.0 Ar, carrier: 0.685 He, 0.696 He; peak-jump scanning mode; time-resolved acquisition mode; 60 seconds total ablation time. Si was employed as internal standard. NIST-SRM 612 was used during the measurements to correct for background and drift with double-standard measurements bracketing each six samples. Reported compositions are averages of three measurements for each sample. Accuracy of the results was monitored by analyzing USGS standard BCR-2G after each six samples. The percentage of deviation from recommended values, determined in multiple sessions, was generally ≤ 10% for all reported trace elements. Results obtained on the ARNT bauxite standard BX-N are given in Appendix 4.1 (Chapter 3).

4.4 | Results

4.4.1 | Textures and mineralogy 4.4.1.1 | Bakhuis Mountains: Snesie and Macousi Several different textural/lithological zones were observed in the drill cuttings of the Snesie bauxite-bearing lateritic profile (Figure 4.3a). The greyish-red massive hard bauxite top layer passes downwards into a pale-red friable earthy layer consisting of bauxite fragments/ concretions and clayey bauxite. This zone then gradually grades into a kaolinite-rich saprolitic

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 78 Bauxite formation on Proterozoic bedrock of Suriname | 79

horizon with a relic parent rock structure of alternating light- and dark-colored bands (Figures 4.3, 4.4a, c). The bauxite of the Macousi trench is generally soft with concretions at the top of the lateritic profile, which passes downward into bauxite fragments in a clayey bauxite matrix and finally into a clayey saprolite with a similar gneissic foliation as observed at Snesie (Figure 4.3b). The most important difference between the bauxite from Macousi and Snesi is its degree of consolidation, as the bauxite from Macousi is mainly soft, while that of Snesie has a very hard top section and becomes softer with depth. In the Snesie and Macousi lateritic weathering profiles, gibbsite is the dominant mineral at the top and kaolinite at the bottom of the profiles. Bauxite samples also contain goethite, hematite, ilmenite, and traces of rutile/anatase and zircon (Figures 4.4a-f, 4.6 and Table 4.1). The Al-rich matrix is generally fine-grained and porous, while the voids are occasionally lined with coarse-grained secondary Al-hydroxides (gibbsite) or iron-(hydr)oxides (hematite and/or

goethite) (Figures 4.4b, f). EDS analysis revealed that the average Al2O3 content of gibbsite in the Snesie samples is 63.7 ± 1.24% (n=37). The middle friable earthy zone of the profile consists of gibbsite-rich fragments and Fe-rich concretions, and becomes more kaolinite-rich towards the bottom. This change in mineralogy is consistent with the XRD and XRF data from both weathering profiles (Figures 4.6a, b, 4.7). A clayey texture appears at the transition zone between the bauxite and the saprolite, around 5 m depth, where sample SNE-3142 was taken, which contains substantially more clay minerals (e.g., kaolinite) than the samples at the top of the profile. Further down, the texture in the samples becomes pelithomorphic (uniform clayey matrix) with a relic foliation of alternating light- and dark-colored layers and relic blasts from an apparently gneissic parent rock (Figures 4.4a, c). These blasts are surrounded by pressure shadows and are probably weathered sillimanite blasts that were almost completely converted into gibbsite (Figures 4.4c, d). The altered blasts also contain clusters of kaolinite flakes (Figure 4.4d). Vermicular kaolinite stacks 4 were observed in the saprolite of the Snesie area. Kaolinite peaks in XRD spectra occur in samples from approximately 4 m depth, consistent with the appearance of the first clusters of kaolinite flakes in sample SNE-3141 and MAC-759. It is possible that small quantities of kaolinite (< 0.5%) are present in the top samples but remained undetected by XRD. The coarse-grained Ti-Fe-oxides such as ilmenite generally have a subhedral to euhedral shape (Figures 4.4e-f). They are occasionally clustered and aligned, showing a foliated texture, which could be a relic texture (Figure 4.4e).

4.4.1.2 | Bakhuis Base Camp Sample BAK-01 is a severely weathered equivalent of BAK-02, which is a relatively unweathered piece of amphibolite with a characteristic foliation. Quartz, gibbsite hematite, goethite with minor amounts of Ti-oxides (rutile/anatase, ilmenite) and zircon are the main phases in BAK-01 (Figures 4.4g-l). Quartz grains in BAK-01 show signs of dynamic re-crystallization (Figure 4h). The presence of a fresh core with a weathered rim in BAK-02 provided an opportunity to analyze the weathering products and microtextures within and around the altered minerals. Parallel alignment of amphibole (magnesiohastingsite), plagioclase (64% anorthite), orthopyroxene (hypersthene), and quartz constitutes the foliation in BAK-02 (Figure 4.4g). Weathering in plagioclase (andesine) and amphibole, started along cleavages and fractures where gibbsite

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 79 80 | Chapter 4

Figure 4.4 | (a) Optical microscope image of gneissic banding in bottom sample SNE-3144 at 7 m depth (plane polarized light); (b) BSE image of void/geode filled with (Al)-goethite and hematite layers with a botryoidal texture in its center in SNE-3136 at 1.5 m depth; (c) Optical microscope image (plane-polarized light) of an altered Al-rich relic blast (with yellow outline) and foliation in saprolite sample SNE-3143 at 6 m depth;

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 80 Bauxite formation on Proterozoic bedrock of Suriname | 81

(d) Enlarged BSE image of an altered blast at 6 m depth where the areas with green outlines represent gibbsite (Gbs), while the pink outlines represent sillimanite (Sill) and the blue outline a kaolinite cluster (Kln); (e) BSE image showing aligned Fe-Ti-oxides (ilmenite) in the Al-rich matrix of SNE-3136; (f) Euhedral to subhedral Fe-Ti oxide crystals from the upper part of the weathering profile. The geode in the upper part of the image is filled with iron oxides (SNE3136); (g) BSE image of the unweathered core of outcrop sample BAK-02 consisting of amphibole (Am), orthopyroxene (Opx), plagioclase (Pl) and quartz (Qtz); (h) Optical microscope image (plane polarized light) of a fine- grained gibbsite boxwork in outcrop sample BAK-01 growing along brown Fe-rich relict fractures or cleavage planes of plagioclase. Note the quartz grain in the right corner; (i) Partially altered plagioclase (andesine) grain with secondary gibbsite along its fractures in BAK-02; (j) BSE image of indirect weathering of plagioclase into gibbsite via kaolinite; (k) Gibbsite in a weathered amphibole grain; (l) Goethite boxwork formed after hornblende weathering in BAK-01.

Table 4.1 | Compilation of the mineralogy for selected samples from the study areas. Estimated abundances are based on XRD, electron microprobe and microscopic analysis. Dominant (+++); Abundant (++); Detected (+); Absent (---).

SNE- SNE- SNE- BAK- BAK- NAS- NAS- NAS- MAC- MAC- MAC- 3135 3140 3144 01 02 01 03 08 1756 1760 1763 Gibbsite +++ +++ ++--- +++ +++ + +++ +++ + Boehmite ------++ + ------Hematite ++ ++ ++--- ++ + --- ++ + --- Goethite ++------++--- ++--- Kaolinite --- + +++ + ------++ --- + +++ Quartz ------++ ++++------Anatase + ------+++ + + --- 4 Amphibole ------+++ ------Pyroxene ------+++ ------Plagioclase ------+++ ------Zircon ++++++++ + + +

was formed (Figures 4.4i, j). Gibbsite is the dominant secondary mineral and its crystals are clustered along septa (threads) forming a boxwork texture (Figures 4.4h-k). The thin brown lineaments along the median plane of the septa mark the position of the initial fractures in the precursor feldspar grain (Figure 4.4h). Clusters of kaolinite are also present in the weathered rims of the samples (Figure 4.4j). The secondary Fe-hydroxides (goethite) that formed after hornblende show a regular boxwork texture or crossed septo-alteromorph (Figure 4.4l). Ilmenite, pyrite, zircon, magnetite, apatite and Ni-bearing iron sulfides were present as accessory phases in BAK-02. XRD spectra confirm the presence of the major phases (Figure 4.6).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 81 82 | Chapter 4

Figure 4.5 | (a) BSE image of pisoliths in bauxite with alternating Al-hematite and/or Al-goethite layers (NAS-03); (b) BSE image of an gibbsite-rich pisolith and cyclically precipitated iron-oxide layers (NAS-03); (c) Microscope image of an iron-stained gibbsite-rich pisolith with concentric layering (NAS-03); (d) Microscope image of double gibbsite lining (NAS-grey); (e) BSE image of a void with euhedral gibbsite crystal lining (grey), coated with goethite (white); (f) BSE image of geode filled with goethite crystals showing a botryoidal texture (NAS-08); (g) Microscope image of multiple concentric Fe-oxide linings (PPL); (h) BSE image of boehmite (Bhm) crystals with a parallel orientation determined by two cleavage planes of a precursor mineral, resulting in a cross pattern. The surrounding dark grey material consists of gibbsite (Gbs) (NAS-11); (i) Microscope image of illuvial textures in NAS-03 (PPL).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 82 Bauxite formation on Proterozoic bedrock of Suriname | 83

Figure 4.6 | X-ray diffraction (XRD) patterns of a select group of samples from the Bakhuis and Nassau study locations. The spectra reveal the changing mineralogy in the lateritic weathering profiles with depth.

4.4.1.3 | Nassau Mountains (Plateau C) The cores from the lateritic profile in drill hole PC-10-0131 were investigated down to a depth of 4.5 m. The upper part of the bauxite layer (0–3m) is massive and contains gibbsite, Al-goethite, hematite, and anatase. The massive pisolitic bauxite gradually grades into a bauxitic clay with a relic parent rock texture and large amounts of kaolinite in the lowermost 1.5 m (Figures 4.5a-d). The pisoliths occasionally have concentric layering and contain variable amounts (0–30%) of ilmenite inclusions. In sample NAS-08 a gneissic foliation with alternating dark- and light- colored layers was observed. The most dominant mineral in the Nassau bauxite samples is gibbsite, followed by Al-rich iron-(hydr-)oxides (goethite and hematite) and minor amounts

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 83 84 | Chapter 4

of Ti-Fe-oxides (ilmenite), boehmite, and quartz (Figure 4.6 and Table 4.1). Coarse euhedral and subhedral authigenic gibbsite crystals with numerous twins line the voids and constitute the matrix (Figures 4.5d, e). Some voids are marked by a primary lining of euhedral gibbsite crystals which is coated with a layer of goethite (Figure 4.5e). In some samples (e.g., NAS-03) yellow/reddish brown goethite and hematite banding was observed lining voids, surrounding concretions or forming a botryoidal texture within voids (Figures 4.5a, b, e-g). EDS analysis revealed Al-substitution (1–16%) in goethite and hematite in these layers. Sample NAS-11 contains an intergrowth of gibbsite and minor boehmite, with the latter forming a cross-linear pattern (Figure 4.5h). Several illuviation cutans are present, which result from absolute chemical, mechanical, or biological accumulation of mobile substances transported from the corresponding eluvial domain at the top of the profile (Figure 4.5i). These illuviation cutans provide evidence for the movement of significant matter at a microscopic scale.

4.4.2 | Geochemistry Major-element data from the Snesie, Macousi, Nassau, Lely and Brownsberg exploration drill holes are summarized in Table 4.2, and complete major and trace-element data sets for the lateritic weathering profiles that were investigated in detail are reported in Tables 4.3a and b. Examples of concentration trends in vertical profiles are shown in Figure 4.7. The average compositions of drill hole material from these areas are quite coherent, with

total Al2O3 ranging between 40 and 49 wt%, except for the Lely Mountains where average Al2O3

is only 33%. Average SiO2 contents are highest at Brownsberg (13%) and lowest at Lely Mts.

(6%), whereas Fe2O3 contents are highest at Lely Mts. (34%) and lowest at Macousi (13%). The

Nassau, Lely and Brownsberg areas stand out in having considerably higher TiO2 contents (ca. 4–4.5%) than the Snesie and Macousi deposits (ca. 1.6%). Vertical chemical profiles have many characteristics in common but show differences and

considerable local variation in detail (Figure 4.7). In the more complete profiles, the Al2O3 contents often increase first from the saprolitic bottom upwards until reaching a maximum in the bauxite horizon, and then decrease in the duricrust towards the top, as is best illustrated in the Macousi profiles (Figure 4.7). In the Snesie (-3 to -5 m), Macousi (-3 to -4 m) and Lely Mts. (-3

to -4 m) profiles, this inflexion point, where SiO2 reaches lowest concentrations, generally lies deeper than in the Nassau (< 1 to 4 m) and Brownsberg (ca. -2 m) profiles. At Snesie, the depth

of the Al2O3 maxima roughly coincides with a transition from friable bauxite to massive bauxite, whereas at Macousi it corresponds to a transition from clayey bauxite to friable bauxite. The uppermost part of the Snesie profiles is relatively complex in showing more fluctuations, in agreement with the presence of an additional top layer of massive bauxite (Figure 4.3).

The LOI values mirror those of Al2O3, independent of the color change, whereas Fe2O3 shows

opposite behavior, as it increases towards the top of the weathering profiles. The TiO2 contents tend to increase towards the higher parts throughout, which is most obvious in the Nassau and

Brownsberg profiles. There is also a steady increase in P2O5 content. The oxides of Ca, Mg, Na and K are depleted along the entire length of the profiles.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 84 Bauxite formation on Proterozoic bedrock of Suriname | 85

Table 4.2 | Concentration averages and standard deviations for Loss on ignition (LOI) and main oxides determined in samples from exploration cores from the Bakhuis area (Snesie and Macousi), Nassau Mountains (Plateau C), Lely Mountains and Brownsberg. n = number of samples. Data source: Suralco L.L.C. and Bauxite Institute Suriname.

Snesie Macousi Nassau Mountains Lely Mountains Brownsberg (Plateau C) LOI (%) 21.7 ± 4.3 25.6 ± 4.5 22.3 ± 5.9 17.2 ± 5.18 22.4 ± 5.6

SiO2 (%) 9.7 ± 11.5 10.4 ± 10.9 8.4 ± 11.6 6.1 ± 7.71 13.1 ± 12.4

Al2O3 (%) 40.5 ± 5.0 48.7 ± 5.9 43.5 ± 12.2 33.2 ± 9.67 43.3 ± 8.2

Fe2O3 (%) 25.3 ± 9.8 12.8 ± 5.53 21.2 ± 18.9 34.1 ± 13.6 16.5 ± 9.4

TiO2 (%) 1.65 ± 0.5 1.5 ± 0.75 3.9 ± 1.8 4.6 ± 2.1 4.0 ± 1.4 n 55 44 1368 3132 488

Trace-element trends in the investigated Snesie (SNE-10-04) and Macousi (MAC-10-01) lateritic profiles display considerable concentration variations and also show distinct patterns in the intervals above and below the transition zone (Figure 4.8). Virtually all of the trace elements analyzed are 2–10 times enriched in the top layer relative to the lower parts. Although the chemical variations in the two profiles are not identical, some systematics can be observed in relationships between trace elements and more abundant constituents. Many trace elements show conspicuous increases from the saprolite towards the top of the weathering profiles, followed by a modest decline in the top layer. The most prominent examples are Nb, Ta, REE, Y, Sr, Th, Pb, whereas Zr, Hf, U and Cr show an upward increase throughout. The trends for V and As are similar to those of Fe O and P O . The Snesie profile is more complex 2 3 2 5 4 than the Macousi profile, in particular since TiO2 and associated elements (Nb, Ta, REE, Y, Sr) show a double peak. In the relatively short Nassau profile (ca. 4 m) analyzed for major and trace

elements, Sc and V decrease along with SiO2 and Fe2O3 towards the top, while Cr increases

together with Al2O3. The pattern for TiO2 is somewhat irregular and is largely mimicked by Zr, Nb, Ta, REE, Y and Sr. The results for the two samples from the Bakhuis Base Camp location trace the compositional change between a fairly ‘fresh’ metamorphic parent rock (BAK-02) and a strongly weathered lateritic equivalent (BAK-01). Compared to the parent rock, the weathered sample is enriched in

Al2O3, Fe2O3, and TiO2, has lower SiO2 and is strongly depleted in Na2O, K2O, CaO, MgO and MnO. BAK-01 contains higher concentrations for Zr, Hf, Nb, Ta, Pb, Th, U, V, As, whereas other trace elements are depleted relative to BAK-02.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 85 86 | Chapter 4 Bakhuis Base Camp Snesie 10 1535 15 58 13 30 14 23 20 23 17 29 13 35 17 31 14 41 21 36 34 33 109 30 3812 44 13 26 15 29 13 28 12 36 12 28 12 11 7.4 19 2.0 10 4.2 265 13 2.5 9.20.3 12 0.2 11 n.d. 12 n.d. 8.8 n.d. 12 n.d. 0.2 4.0 0.1 2.2 0.3 3.1 0.4 3.5 0.8 8.2 0.6 0.8 3.0 4.7 5.0 3.0 3.7 4.0 6.4 4.3 1.4 2.7 2.9 13.2 100805 100256 840 100 329 1037 100 370 746 100 377 646 100 345 1005 100 342 523 100 267 263 100 173 546 100 94 374 100 169 526 100 456 184 627 142723 159 79 178 63 146 167 46 162 39 158 33 82 40 32 36 50 34 634 28 49 14 140 0.50 0.56 1.17 1.74 1.33 3.27 16.34 23.73 30.67 29.19 24.87 51.81 2.69 3.17 3.520.26 2.70 0.27 2.81 0.22 2.58 0.17 3.49 0.13 3.01 0.14 2.04 0.09 2.16 1.36 0.08 0.53 0.07 0.07 0.10 0.13 0.0310.010 0.030 0.004 0.028 0.006 0.025 0.007 0.025 0.006 0.017 0.006 0.033 0.010 0.017 0.006 0.029 0.009 0.020 0.008 0.043 0.015 3.04 0.23 41.8631.45 37.560.0170.037 37.18 39.100.007 0.021 0.038 33.86 45.11 0.033 0.020 0.042 25.20 51.2223.14 0.042 0.012 0.041 17.03 43.06 21.13 0.021 0.009 0.038 27.71 47.75 21.99 0.010 0.010 0.036 47.32 9.30 24.97 0.020 0.013 0.049 40.26 4.48 27.40 0.029 0.011 0.049 42.87 9.74 23.16 0.017 0.014 29.37 0.047 22.89 0.036 0.010 7.04 16.20 0.049 21.29 0.013 0.039 26.93 0.046 0.125 9.19 0.026 17.08 9.97 8.27 18.54 17.23 0.52 SNE-3135 SNE-3136 SNE-3137 SNE-3138 SNE-3139 SNE-3140 SNE-3141 SNE-3142 SNE-3143 SNE-3144 BAK-01 BAK-02 Major-oxide (XRF) and trace-element (LA-ICP-MS) contents of samples from drill hole SNE-10-04 (Snesie) and Bakhuis Base Camp. n.d.= not detected. n.d.= drill hole SNE-10-04 (Snesie) and Bakhuis Base Camp. Major-oxide of samples from (XRF) and trace-element contents (LA-ICP-MS) 3 3 (%) 5 O 2 2 O O 2 2 O 2 O 2 2 Al K P Table 4.3a | Table TiO Fe MnO CaO MgO Na LOI Sum Sc (ppm) V Cr Zn As Rb SiO Sr Y Zr Nb Ba

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 86 Bauxite formation on Proterozoic bedrock of Suriname | 87 Bakhuis Base Camp

Snesie 4 2339 3818 61 31 32 60 19 27 39 24 15 39 24 19 40 34 19 65 27 31 57 10 25 19 17 10 35 5.9 16 9.4 18 3.6 21 10 4.00.77.6 4.47.7 0.80.7 9.7 4.8 10 0.9 0.8 10.7 3.9 13 0.8 1.0 9.6 4.9 14 0.8 1.1 9.1 5.1 16 0.8 0.8 10.3 4.7 16 0.7 0.9 9.6 2.5 14 0.5 0.6 6.8 1.0 0.1 7.3 0.4 3.2 1.4 0.3 1.7 0.3 5.0 14.6 0.7 3.8 5.5 0.3 1.3 0.1 1.1 6.7 1.4 0.05 0.0 4.9 8.2 7.0 4.3 4.9 5.0 7.9 6.8 2.5 4.4 1.1 2.4 2.820.741.77 5.350.22 1.481.10 3.19 4.690.15 0.36 1.300.41 1.88 2.860.38 2.31 0.28 0.340.05 0.73 0.61 1.76 1.59 0.56 3.10 0.26 0.18 0.07 0.89 0.60 0.96 1.94 0.44 3.10 0.16 0.23 0.07 0.93 0.40 1.30 2.01 0.36 5.05 0.20 0.25 0.05 1.84 0.48 1.27 3.64 0.46 3.79 0.20 0.41 0.07 1.28 0.45 2.09 2.20 0.42 1.37 0.33 0.24 0.06 0.40 0.65 1.15 0.67 0.54 2.33 0.18 0.08 0.07 0.66 0.40 0.48 1.25 0.65 0.33 0.07 0.16 0.12 0.04 0.20 0.91 2.12 0.41 0.17 0.11 0.75 0.09 0.02 0.24 2.12 0.58 0.23 0.31 0.12 0.03 2.16 0.47 0.43 0.63 1.32 0.10 1.40 0.22 SNE-3135 SNE-3136 SNE-3137 SNE-3138 SNE-3139 SNE-3140 SNE-3141 SNE-3142 SNE-3143 SNE-3144 BAK-01 BAK-02 Continued Table 4.3a | Table Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U La Ce Pr Nd Sm

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 87 88 | Chapter 4 Nassau 19 35 14 13 55 72 38 32 5 5 10 22 30 43 51 8 15 11 8 9 7 9 Macousi 14 11 10 10 26 24 22 21 22 28 25 17 49 58 26 32 23 12 9.3 8.8 8.7 4.6 4.7 3.7 26 29 20 16 23 30 31 30 22 21 26 26 25 17 13 16 17 17 13 14 12 12 9.2 23 7.5 8.3 0.4 0.3 0.3 0.2 0.3 0.4 0.3 0.4 0.3 1.5 0.2 0.2 7.0 8.4 8.1 7.6 5.2 2.2 3.4 1.5 12 21 7.6 7.5 n.d. n.d. 0.001 0.001 n.d. n.d. n.d. n.d. n.d. 0.000 n.d. n.d. 100 100 100 100 100 100 100 100 100 100 100 100 749 676 666 651 555 391 389 323 954 895 569 440 299 284 273 239 154 64 105 42 863 1082 498 479 2.60 3.09 2.51 3.82 11.46 29.06 22.39 34.26 1.29 6.28 9.43 17.62 2.18 2.46 2.39 2.180.08 1.50 0.04 0.63 0.04 0.98 0.03 0.45 0.03 7.24 0.01 9.20 0.02 5.29 0.01 5.19 0.09 0.08 0.04 0.04 1343 606 427 423 385 242 220 198 1240 1207 1916 1601 0.0070.011 0.004 0.012 0.010 0.012 0.006 0.011 0.006 0.009 0.007 0.009 0.006 0.009 0.005 0.008 0.011 0.008 0.008 0.028 0.004 0.005 0.007 0.007 46.5722.15 53.76 11.36 56.24 56.30 8.33 52.31 7.57 44.01 8.26 48.35 7.08 41.49 5.95 52.74 6.47 45.92 12.10 41.57 14.27 35.13 22.75 24.75 26.37 29.26 30.45 30.08 26.42 19.19 22.28 17.30 26.46 24.11 20.88 17.20 0.015 0.015 0.013 0.011 0.007 0.002 0.004 0.002 0.060 0.083 0.035 0.041 0.014 0.006 0.003 0.001 n.d. 0.000 0.000 0.002 n.d. 0.020 n.d. 0.003 Major-oxide (XRF) and trace-element (LA-ICP-MS) contents of samples from drill holes MAC-10-01 (Macousi) and PC-10-0131 (Nassau Plateau C). n.d.= not C). not and PC-10-0131 (Nassau Plateau (Macousi) n.d.= drill holes MAC-10-01 Major-oxide of samples from (XRF) and trace-element contents (LA-ICP-MS) MAC-1756 MAC-1757 MAC-1758 MAC-1759 MAC-1760 MAC-1761 MAC-1762 MAC-1763 NAS-03 NAS-05 NAS-07 NAS-08 3 3 (%) 5 O 2 2 O O 2 2 O 2 O 2 2 Table 4.3b | Table detected. Al TiO K P Fe LOI MnO SiO Sum CaO Sc (ppm) MgO V Na Cr Zn As Rb Sr Y Zr Nb Ba

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 88 Bauxite formation on Proterozoic bedrock of Suriname | 89 Nassau 27 47 21 19 8

4 Macousi 25 30 31 31 24 14 21 21 22 34 20 19 29 38 40 39 28 12 19 12 16 17 17 12 4.8 7.6 3.1 10 18 8.0 7.1 14 16 17 18 14 5.9 9.5 4.0 23 23 12 13 3.5 4.6 4.7 4.6 3.3 1.4 2.2 0.9 2.9 5.0 2.3 2.1 6.9 6.8 6.7 6.0 4.0 1.7 2.7 1.1 20 25 12 12 1.2 1.5 1.6 1.5 1.0 0.4 0.7 0.3 3.2 4.3 2.3 1.9 9.2 10.1 10.4 10.9 8.2 4.4 6.0 3.3 10.5 18.3 9.9 9.0 1.4 1.4 1.3 1.3 0.9 0.5 0.7 0.4 1.9 3.0 1.5 1.5 2.07 2.69 2.70 2.60 1.83 0.75 1.22 0.43 1.55 2.97 1.22 1.24 0.47 0.59 0.62 0.61 0.39 0.18 0.29 0.11 0.31 0.62 0.28 0.26 1.37 1.83 1.74 1.72 1.26 0.50 0.85 0.33 1.29 2.44 0.86 0.94 0.20 0.25 0.25 0.25 0.15 0.07 0.11 0.04 0.20 0.41 0.14 0.14 1.31 1.64 1.58 1.56 1.01 0.45 0.70 0.29 1.62 3.08 1.10 1.12 0.25 0.31 0.30 0.28 0.19 0.09 0.14 0.06 0.40 0.69 0.27 0.27 0.77 0.93 0.88 0.86 0.60 0.26 0.39 0.18 1.57 2.57 1.02 1.04 0.87 1.01 1.03 0.97 0.66 0.31 0.48 0.21 2.78 4.44 1.90 1.84 0.14 0.17 0.16 0.14 0.10 0.04 0.06 0.04 0.61 0.93 0.43 0.42 Continued MAC-1756 MAC-1757 MAC-1758 MAC-1759 MAC-1760 MAC-1761 MAC-1762 MAC-1763 NAS-03 NAS-05 NAS-07 NAS-08 Table 4.3b | Table La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 89 90 | Chapter 4

Figure 4.7 | Examples of vertical variations of major element contents in lateritic weathering profiles from drill holes in the Bakhuis area (Snesie and Macousi) and Nassau Mountains (Plateau C).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 90 Bauxite formation on Proterozoic bedrock of Suriname | 91

4 Figure 4.7 | Continued

4.5 | Discussion

4.5.1 | Formation of weathering products and textural features 4.5.1.1 | Main weathering minerals Gibbsite and boehmite This study confirms previous observations that gibbsite is the main Al-hydroxide in the Bakhuis and Nassau bauxite deposits (Pollack, 1984; Billiton, 1979; Bárdossy and Aleva, 1990; Janssen, 1963, 1977, 1979). Based on morphological textures (cf. Delvigne, 1998), two different generations of gibbsite can be distinguished in the samples. The first generation is represented by: a. Aphanitic to coarse-grained gibbsite-rich matrix (crypto-alteromorphs) after kaolinite (Figure 4.4j); b. Coarse-grained septa forming a glomero-septo-alteromorph after feldspar (Figures 4.4h, i), while the second-generation gibbsite appears as: Coarse-grained void linings/coatings (Figures 4.5d, e).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 91 92 | Chapter 4

Figure 4.8 | Vertical concentration changes for selected trace elements in Snesie, Macousi and Nassau lateritic weathering profiles. X-axes at the top of the panels are valid for the full lines and those at the bottom for the dashed lines.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 92 Bauxite formation on Proterozoic bedrock of Suriname | 93

This second generation of gibbsite originated from the precipitation and aging of colloidal Al- rich solutions migrating in the bauxite horizon and in the duricrust. Multiple gibbsite linings in voids and the perpendicular orientation of grains against the walls of voids and pores are consistent with an allochthonous origin of the mineral, also known as allogenic deposit (Bárdossy and Aleva, 1990; Delvigne, 1998). Boehmite is only a minor phase in the Nassau and Bakhuis deposits, and its presence has also been reported for the bauxite deposits of Brownsberg and the Lely Mountains (Van Kersen, 1956; Janssen, 1963, 1979). The microtextures in the Bakhuis and Nassau samples thus suggest that gibbsite and boehmite are products of one or more of the following reactions: 1. Direct conversion of feldspar (plagioclase) into gibbsite or boehmite Gibbsite septa and aligned boehmite following fossil cleavage patterns or irregular transmineral fractures signal this mechanism (Figures 4.4h-l, 4.5h). Next to gibbsite, it is conceivable that boehmite locally originated as a primary weathering product as well. Direct formation from the dissolution of feldspar (Helgeson, 1971; Pollack, 1981; Bárdossy and Aleva, 1990; Kawano and Tomita, 1995 and references therein) would be generally consistent with the nature of the parent rocks at Nassau and Bakhuis. 2. Gibbsite as desilication product of kaolinite. Transformation of kaolinite into gibbsite (following the conversion of plagioclase into kaolinite; Figure 4.5e) is described by the reaction:

Al2Si2O5(OH)4 + H2O→ 2Al(OH)3 + 2SiO2(aq) Continuous plagioclase weathering into kaolinite can also give rise to the formation of boehmite by incongruent dissolution:

Al2Si2O5(OH)4 2AlO(OH) + 2SiO2(aq) + H2O These desilication reactions are a function of silica and water activity, humidity, depth to the water table and the presence of other ions (Tardy, 1997; Zhu et al., 2006; Wei, 2014). 4 3. Dehydration of gibbsite into boehmite Gibbsite, usually the primary Al-hydroxide mineral to form during bauxitization of water- rich, iron-poor systems (Saalfeld, 1958; Zhu, et al., 2010), may later dehydrate into boehmite (Trolard and Tardy, 1987; Gong et al., 2002). A decrease in water activity, high surface temperatures and dry conditions during the dry seasons either facilitates this dehydration reaction or establishes thermodynamic conditions where boehmite is the main crystallizing phase instead of gibbsite (Trolard and Tardy, 1987; Eggleton and Taylor, 2008 with references therein). The possible presence of mixed crystals, resulting from isomorphous replacement or intergrowth (Gong et al., 2002), could not be ascertained because of the small grain sizes of the Al-oxides in the matrix. 4. Direct conversion of pyroxene/amphibole into gibbsite This reaction represents a separate type of weathering mechanism, involving direct transformation of framework silicate minerals into the layered Al-OH structure of gibbsite (Valeton, 1972; Aleva, 1994; Dos Muchangos, 2000).

Hematite and goethite Hematite and goethite are both present as mottles, nodules, botryoidal void fillings, finely dispersed in kaolinitic matrix or as alternating rims enveloping pisoliths. The broad goethite and hematite peaks in the XRD spectra (Figure 4.6) suggest either that not a single, uniform

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 93 94 | Chapter 4

phase is present in the samples but a range of compositions resulting from solid solution, as was also reported for other Surinamese bauxites (Feret et al., 1997), or that crystallinity is poor. The goethite linings and voids fillings (Figures 4.5a, b, e, f, g), reflect cyclic iron dissolution and precipitation by redox or pH alternations that drive iron away from reducing toward oxidizing areas (Topp et al., 1984; Scott and Pain, 2009). The redox changes are probably linked to fluctuating groundwater in the mottled horizon of add e.g. Nassau Plateau C, where aquifers are strongly affected by the alternating dry and wet seasons.

The available EDS results point to substantial amounts of Al2O3 in goethite (ca. 16%) and hematite (ca. 2%) in the Nassau and Bakhuis samples, in accord with the generally lower Al- substitution in hematite than in goethite (Fitzpatrick and Schwertmann, 1982; Tardy and Nahon, 1985; Trolard and Tardy, 1987; Majzlan and Navrotsky, 2003). Earlier work reported

20–25% Al2O3 substitution into goethite from the Bakhuis bauxite deposit (Billiton, 1979).

Commonly, the Al2O3 contents in hematite and goethite increase towards the top of weathering profiles (Trolard and Tardy, 1987), apparently reflecting the amount of available Al in solution. Hence, high Al-substitution in goethite is found when the mineral is associated with gibbsite or boehmite.

Kaolinite Kaolinite was mainly observed in the bottom parts of the weathering profiles. Its absence in the bauxite layer can be attributed to a slow crystallization reaction of kaolinite relative to gibbsite

formation (Schellmann, 1994). This facilitates removal of freshly formed SiO2-rich colloids by draining solutions, essential for gibbsite stability (Schellmann, 1994 and references therein). The large kaolinite quantities at the bottom of the profiles have a negative impact on the drainage and leaching conditions at depth. Early-formed kaolinite in laterite and bauxite, produced from the weathering of primary silicates in parent rock, is often dissolved and replaced by a second generation of kaolinite with lower crystallinity and higher iron content as was observed in the studied lateritic weathering profiles (Schellmann, 1994; Nahon and Merino, 1997). The dissolution of kaolinite could be promoted by the release of protons by ferrolysis (Nahon and Merino, 1997). The quantity of secondary kaolinite generally increases during lateritization (Nahon and Merino, 1997). The observed textures such as kaolinite booklets and amorphous kaolinite-rich matrix point to a complex history for kaolinite in the studied weathering profiles. In all of the samples, neo-formed iron-stained kaolinite has a smaller grain size and fills voids in the matrix.

4.5.1.2 | Origin and evolution of secondary textures Origin of void fillings The gibbsite and goethite crystallaria and clay illuvia in veins and voids of the Bakhuis and Nassau samples were formed by different mechanisms. The crystallaria nucleated on the walls of voids in the matrix, and grew perpendicular to these from passing Al- and Fe-rich solutions, whereas transport of detrital material (illuvium) produced void linings that tend to be layered or microstratified parallel to the walls (Delvigne, 1998; Velde and Meunier, 2008). Goethite and gibbsite crystallaria in voids can also be a result of in-situ degradation of ferro-argillans previously deposited from suspensions of clay microparticles, but then the crystals should be lined parallel to the walls as well (Delvigne, 1998). The gibbsite-filled void with goethite lining in the Nassau sample (Figure 4.5h), demonstrates that both Al and Fe were mobile at some point

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 94 Bauxite formation on Proterozoic bedrock of Suriname | 95

and that initial formation of gibbsite crystals was followed by a goethite lining. Both elements were probably transported through the soil system by oscillating groundwater, which could also have induced fluctuations in chemical conditions favourable for dissolution-deposition sequences. The percolating water removes soluble substances, while evaporation brings dissolved components to the surface by capillarity (Scott and Pain, 2009). The ongoing wetting and drying cycle in the Nassau deposit from oscillating groundwater creates cracks during dry seasons and their re-filling with gibbsite, goethite or clay minerals during wet seasons. Illuviation textures (cutans) in the Bakhuis and Nassau samples are made up of amorphous

Figure 4.9 | Triangular plots depicting the degree of lateritization in all study areas (diagram based on Schellmann, 1983, 1986).

clay minerals (kaolinite-group minerals), which points to clay migration within the profiles at these locations (Figure 4.5i). This filling of cracks and voids with clay minerals has influenced the composition of the resulting bauxite, and it also reduced its porosity and permeability.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 95 96 | Chapter 4

Evolution of pisoliths and concretions The most prominent textural difference between the Bakhuis and Nassau bauxite is the abundance of hematite, goethite and gibbsite pisoliths in the latter, some of which showing concentric zonation (Figure 4.5c). Pisoliths can form under various specific conditions. Precipitation of Al and Fe-hydroxides tends to form rounded pedological features, driven by the material’s tendency to form shapes with minimum specific surfaces, as is observed in other Surinamese bauxite deposits (e.g., Klaverblad) and reported from other studies (Tardy and Nahon, 1985; Bárdossy and Aleva, 1990; Taylor and Eggleton, 2008; Wei et al., 2014). The segregation of Al and Fe in pisoliths and concretions can also result from outward migration of the iron (Boulangé, 1984; Trolard and Tardy, 1987). Fe-rich coatings around concretions are produced by absorption or precipitation of migrating iron. Delvigne (1998) attributed the formation of pisolitic bauxite to degradation under fluctuating chemical conditions, accompanied by the formation and evolution of glaebules at the expense of either isalteritic or alloteritic bauxite. The fluctuating groundwater table and plateau-shaped landscape of the Nassau Mountains have a significant influence on hydration control, since alternations of hydration, drainage and redox conditions are favourable for pisolith formation (Boulangé, 1984; Trolard and Tardy, 1987).

4.5.2 | Geochemical effects of weathering The petrographic observations and chemical compositions indicate that aluminium is mostly present as gibbsite, with minor amounts in kaolinite, boehmite, and Al-bearing goethite and hematite. Iron is mainly hosted in goethite and hematite, titanium in anatase/rutile and ilmenite, while silicon mostly resides in kaolinite, and occasionally in minor amounts of residual quartz, feldspar and other silicates from the original parent rocks. Chemical indicators confirm that the investigated bauxite deposits represent advanced stages of weathering. Values for

the Chemical Index of Alteration [CIA=100 x Al2O3 / (Al2O3 + CaO + Na2O + K2O); Nesbitt and Young, 1982; Wei et al., 2014] are >99 for all samples from the Snesie, Macousi and Nassau profiles, consistent with efficient removal of Ca, Na and K due to lateritization. According to the

Al2O3-Fe2O3-SiO2 classification of Schellman (1986), the degree of lateritization ranges between moderate in the bottom parts to strong in the top parts of the bauxite profiles (Figure 4.9).

The diagrams of Figure 4.9 further reveal that each area has its own Al2O3-Fe2O3-SiO2 signature, apparently reflecting differences in parent-rock controls and/or weathering histories. The effect of weathering can also be inferred from normalized major and trace element patterns. In absence of data for pristine parent rocks and in view of the chemical variability of metamorphic rocks in the Bakhuis area (Klaver et al., 2015), the concentrations were normalized against upper continental crust (Rudnick and Gao, 2004) in order to illustrate first-order chemical weathering effects (Figure 4.10). The patterns show a consistent depletion of fluid- mobile elements (Si, Mn, Na, K, Sr, Ba) and a relative enrichment of “immobile” elements (Al, Ti,

Sc, V, Cr, As, Zr, Nb, Hf, Ta, Th). SiO2-poor (< 5%) top samples of the profiles tend to be relatively enriched in most elements, with the exception of the highly mobile 1+ and 2+ charged ones (Figures 4.7, 4.8 and 4.10). A comparison of the patterns for the practically unweathered sample (BAK-02) and altered equivalent (BAK-01) from Bakhuis Base Camp demonstrates the relative enrichment of Al, Ti, Fe, V, Zr, Nb, Hf, Ta, Pb, Th, U in the weathering residue.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 96 Bauxite formation on Proterozoic bedrock of Suriname | 97

4.5.2.1 | Mobilization and redistribution of major elements

Inter-element relationships show clear differences between SiO2-poor (< 5%) upper parts and

SiO2-richer (> 5%) lower parts of the Snesie and Macousi profiles, which points to differences

in weathering regime and mineralogical control. In both cases Al2O3 correlates positively

with LOI, albeit along different trends (Figure 4.11). For the SiO2-richer intervals, where Al2O3

correlates negatively with SiO2, the Al2O3-LOI (and Al2O3-SiO2) trends are largely determined by

the proportions of gibbsite and kaolinite. In these intervals, Al2O3 correlates relatively strongly with almost all other elements with a 3+, 4+, or 5+ valence, except for Fe. These systematics are consistent with residual enrichment of relatively immobile elements following the breakdown of silicate minerals and removal of Si and associated soluble components.

In contrast, in the SiO2-poor parts where Al2O3 is almost exclusively hosted in gibbsite, there

is no correlation with SiO2. Instead, there is a strong inverse relationship between Al2O3 and

Fe2O3, reflecting the increasing amount of Fe-rich phases towards the top of the profiles. In the Macousi profile, where vertical concentration trends are relatively smooth and undisturbed 2 (Figure 4.11), there are also clear positive correlations (R = 0.79–0.98) between Al2O3 and Nb, Ta, Pb, Th, REE, Sr, Ba, and negative correlations (R2= 0.84–0.99) with Sc, V, Cr, Zn, As, a group of elements that tend to be associated with Fe. In the top parts of Snesie and Macousi profiles, 2 Fe2O3 correlates positively with P2O5 (R = 0.74 and 1.0, respectively), and with Sc, V, Cr and As 2 (all R > 0.97) only at Macousi. In the bottom parts, correlations with Fe2O3 are poor, except for Sc (R2= 0.94) and V (R2= 0.99) at Snesie.

The most conspicuous feature in the Nassau profile is that TiO2 strongly correlates with many trace elements (e.g., R2 > 0.9 for Zn, Sr, Y, Zr, Nb, Ce, Yb, Hf, Ta), which is comparable to

systematics in the SiO2-richer part of the Macousi profile that can be largely attributed to residual enrichment effects. 4 4.5.2.2 | Mobilization and redistribution of rare earth elements Numerous studies of other lateritic covers of metamorphic and igneous rocks in humid tropical regions have demonstrated the mobility of REE during weathering and differences in the behavior of the light rare earth elements (LREE) and heavy rare earth elements (HREE). Often, upper Fe-rich layers in lateritic weathering profiles are depleted in REE, whereas accumulation occurred in deeper layers such as the basal saprolite rock (Braun et al., 1993, 1998; Kamgang Kabeyene Beyala et al., 2009; Sanematsu et al., 2013; Berger and Frei, 2014). In other cases, net mass gains in REE have been observed in uppermost bauxite levels relative to underlying layers (e.g., Boulangé and Colin, 1994). Frequently, vertical variations point to differences in behavior among the REE, such as preferential leaching and removal of HREE or LREE from the supergene zone, the development of Ce anomalies, or preferential fixation through adsorption or incorporation in secondary minerals. The UCC-normalized REE concentrations in the Snesie, Macousi and Nassau profiles show different patterns (Figure 4.10). In the Snesie deposits, the LREE are enriched relative to the HREE, whereas the Macousi trends show little fractionation and the Nassau profile a flat pattern for the LREE and a gradual enrichment towards the HREE. Overall, the REE seem to be residually

enriched in the SiO2-poor ferruginous top parts of the sequences (Figure 4.10) relative to strongly leached elements (although this does not necessarily exclude their mobility), and there is considerable vertical variability in detail, in particular in the Snesie profile. There is no evidence for a significant Ce anomaly in any of the settings.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 97 98 | Chapter 4

Figure 4.10 | Upper continental crust (UCC) normalized element distributions in the studied lateritic weathering

profiles. Samples indicated in red have < 5% iS O2, while those in blue contain > 5% SiO2. Ca and Mg are not indicated as their concentrations are extremely low and often remained below detection limits for XRF analysis. Note the strong depletion of fluid-mobile elements and the contrasts in the REE part of the patterns for the different areas.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 98 Bauxite formation on Proterozoic bedrock of Suriname | 99

Figure 4.11 | The SiO2-Al2O3, Al2O3-Fe2O3, LOI-Al2O3 and Sr-TiO2 plots for analyzed bauxite profiles. Note the

differences in inter-element relationships between SiO2-poor (top) and SiO2-rich (bottom) parts of the profiles.

The REE trends show little variation within each individual bauxite deposit. This might be the result of insignificant fractionation among the REE in the weathering history, so that the different pattern shapes largely reflect variations of the parent rocks. Alternatively, redistribution of the 4 REE was accompanied by substantial fractionation but the effects from mobility differences during incipient leaching have been obliterated due to the advanced stage of weathering that the current sequences have reached. In absence of samples from pristine parent rocks, the extent to which weathering has re-shaped the REE patterns is difficult to test for Snesie, Macousi and Nassau but some insight into REE mobility and fractionation can be obtained by comparing the patterns of the amphibolite parent rock (BAK-02) and its strongly weathered equivalent (BAK-01) from Bakhuis Base Camp (Figure 4.10). The differences in UCC-normalized distributions suggest that REE were mobile during weathering resulting in lower overall concentrations BAK-01. Also, trends are not parallel, and MREE tend to be preferentially depleted. Finally, the positive Eu anomaly seen in the proposed parent rock is virtually absent in the weathered rock, apparently due to a stronger removal of Eu2+ upon breakdown of plagioclase and other primary minerals. Because the REE pattern for BAK-01 is comparable in shape to that of the Nassau samples, similar REE mobility and preferential loss of MREE may also have affected the bauxite in this area. Topp et al. (1984) also inferred differences in behavior of light and intermediate REE from enrichments at different levels in a 12.5 m laterite profile on Proterozoic charnockitic rocks in the same region as the Bakhuis profiles studied here. They reported an enrichment of Sm and Eu in a clay-rich zone near the base, presumably due to accumulation of species leached out higher up in the profile.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 99 100 | Chapter 4

4.5.2.3 | Possible mineral controls of trace-element distributions Multiple studies have provided evidence for mineralogical controls of trace-element behavior during lateritic weathering, either through the presence of particular (accessory) host minerals in the parent rock, their stability in the course of alteration, uptake by secondary minerals or sorption onto surfaces (Nesbitt, 1979; Banfield and Eggleton, 1989; Braun et al., 1993, 1998). Primary accessory phases that typically host REE and many other high-valence elements in parent rocks of lateritic bauxite include zircon, phosphates (monazite, apatite, xenotime), Ti-bearing minerals (rutile/anatase, titanite, ilmenite) and allanite. Phosphate weathering will usually release a significant amount of the REE budget, which may then become fixed in secondary phosphates such as rhabdophane and florencite, unless advanced weathering ultimately leads to their breakdown as well (Laveuf and Cornu, 2009). The observed trace-element distributions may thus be partly caused by the nature and quantity of original host minerals, differences in their resistance against weathering and/or different behavior after element release (e.g., storage in newly formed minerals or adsorption onto mineral surfaces). Of the most relevant primary accessory minerals, zircon is a ubiquitous phase in many samples of the profiles studied, whereas monazite, apatite, anatase/rutile were occasionally detected especially in the Bakhuis Base Camp samples. Secondary phosphates were not identified in any of the samples but their presence can certainly not be excluded. Because REE and actinide-bearing accessory minerals in Precambrian high-grade gneissic terrains commonly include monazite, apatite, zircon, titanite, allanite and xenotime (Harlov, 2011), it is reasonable to suppose that all of these phases potentially played a role in element re-distribution during bauxitization of the Bakhuis deposits, given the high-grade character (up to granulite facies) of parent rocks in this area. Inspected samples of the meta-volcanic parent rocks in the Nassau area contain zircon, anatase/rutile, while titanite may be present as well.

The often excellent correlation between TiO2 and Sr (Figure 4.11) in the studied areas is remarkable, given the expected mobile behavior of the latter, which suggests that a significant part of the Sr budget is stored in titanite, as this mineral is also capable of accommodating many of the high-valence elements (Tiepolo et al., 2002). If this is the case, then the inferred MREE depletion at Bakhuis Base Camp and Nassau profiles could reflect breakdown of titanite, as this is generally enriched in the MREE (Green and Pearson, 1986; Tiepolo et al., 2002). The LREE enrichment in the Snesie profile (Figure 4.10) could be due to monazite, apatite, or their breakdown products, as these phosphates preferentially incorporate LREE (Boulangé et al., 1990; Horbe and Da Costa, 1999; Laveuf and Cornu, 2009; Babechuk et al., 2015). From the above considerations we infer that weathering effects on mobility and redistribution of REE and other trace elements were significant in all of the areas studied, and affected even the least mobile elements. Differences in element distribution patterns between the individual profiles likely reflects a combination of primary compositional differences of parent rocks, the nature and content of accessory phases and their unequal behavior during weathering.

4.5.3 | Parent rocks and their influences Our results provide new insights into the diversity of crystalline parent rocks of the Surinamese highland bauxites and their influence on properties of the ultimate weathering products. The variability of the original high-grade metamorphic rocks in the Bakhuis Mountains is apparent in the differences between the Bakhuis Base Camp laterite that formed at the cost of pyroxene

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 100 Bauxite formation on Proterozoic bedrock of Suriname | 101

amphibolite and the bauxite-bearing profiles of Snesie and Macousi where pristine parent

rocks were not recovered. The higher Al2O3/Fe2O3 ratios and less complex vertical variability in chemistry in the Macousi profiles relative to the Snesie profiles (Figures 4.7, 4.11), as well as differences in trace element patterns (Figure 4.10), reflect distinct properties of the parent rocks. The compositional banding and altered relic blasts in Snesie saprolite samples point to a sillimanite-bearing gneissic parent rock. The LREE enrichment is consistent with a metasedimentary source, assuming that the apparent REE mobility did not seriously modify the original tendency of the pattern. The parent rock of the Macousi profiles was probably different

in view of the chemistry, but the petrography is inconclusive. The higher Al2O3/Fe2O3 ratios may be indicative of a more felsic parent than that of the Snesie profiles. The Macousi parent rock may have been similar to the pyroxene amphibolitic parent rock at Bakhuis Base Camp on an adjacent plateau (Area 10.1), considering the similar shapes of the trace-element patterns of the weathered rocks (Figure 4.10). In turn, the pattern of the unweathered pyroxene amphibolite (BAK-02) is grossly similar to that of fresh granulites of the Bakhuis Mountains (Klaver et al., 2015), which suggests that Macousi and Bakhuis Base Camp parents form part of the granulite suite, as is also consistent with the depletion in Th and U. The chemical and petrographic data also indicate that the bauxite occurrences in eastern Suriname originated on various parent rocks, in agreement with the variability of rock types such as greenschist-facies metabasalts and other meta-igneous rocks that make up the

Paramaka Formation (Aleva, 1994; Kroonenberg et al., 2016). Average Al2O3/Fe2O3 ratios are highest at Nassau and lowest at Lely, suggesting that the bauxite formed on igneous parent rocks with different degrees of differentiation. The parent rock of Nassau Plateau C is probably a meta-andesite or meta-basalt, based on the relict texture of a piece of saprock-saprolite. A fresh rock from Brownsberg with actinolite, clinozoisite and chlorite is consistent with a relatively mafic extrusive rock, metamorphosed in greenschist facies. 4 The variation in parent rock composition and texture, together with differences in drainage and climate conditions, age and tectonic stability of surfaces have affected the extent of lateritization, the resulting weathering mineralogy and thicknesses of bauxite bodies in the study areas. Original stratigraphic heterogeneity will have added to the diversity and may explain the local presence of lenses of kaolinite-rich material within the bauxite horizon, and lenses of bauxite material within the kaolinitic saprolite (Doeve, 1955; Coutinho, 1967; Janssen, 1979; Aleva and Hilversum, 1984). The diversity in rock types and weathering conditions explains the general lens-shaped morphology of the plateau-type bauxite bodies, the absence of bauxite in certain areas of the Bakhuis and Nassau Mountains, and local exposures of fresh bedrock and boulders within the main bauxite body of Bakhuis (Coutinho, 1967).

4.6 | Summary and conclusions

Textural, mineralogical and geochemical information on lateritic plateau bauxites of Suriname has been investigated to assess the influence of bedrock lithology on the composition of weathering profiles on a variety of Proterozoic crystalline basement rocks of the Guiana Shield. Studied bauxite deposits of the Bakhuis Mountains, Nassau Mountains, Lely Mountains and Brownsberg are distributed at relatively high topographic elevations in the country’s interior

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 101 102 | Chapter 4

as relics of a succession of stepped planation surfaces with a regional extent, which formed as a result of long-term tectonic stability and periods of continental uplift. The parent rocks range from (ultra)high-temperature metamorphic gneissic and amphibolitic rocks in the Bakhuis Mountains (Granulite Belt in west Suriname) to greenschist-facies metabasalts and other meta- igneous rocks in the other areas (Greenstone Belt in east Suriname). All of the investigated profiles show a strong mineralogical and chemical zoning according to the degree of lateritization, which ranges between moderate in the bottom parts to strong in the top layers. The bauxites contain gibbsite as principal Al-hosting mineral, have a ferruginous surface and grade downward into kaolinite-rich saprolite and altered bedrock. Different

signatures of the parent rocks of the deposits are expressed by variable Al2O3/Fe2O3 ratios and

TiO2 concentrations being markedly higher in the bauxites formed on top of metavolcanics of the Greenstone Belt than in those covering the Granulite Belt. The Nassau deposit stands out by an abundance of pisoliths consisting of hematite, goethite and gibbsite, which supposedly reflect fluctuating hydration, drainage and redox conditions. The presence of minor amounts of boehmite further supports a hydration control. First-order effects of chemical weathering are a consistent depletion of fluid-mobile elements (Si, Mn, Na, K, Sr, Ba) and a relative enrichment of “immobile” elements (Al, Ti, Sc, V,

Cr, As, Zr, Nb, Hf, Ta, Th). The SiO2-poor top parts of the profiles tend to be most enriched in the latter group. Normalized element abundances are distinct for different areas, but show uniform patterns in individual profiles. Frequent absence of pristine bedrock equivalents precludes quantitative assessments of mass changes. Despite this internal homogeneity of chemical patterns, weathering processes have induced significant deviations from original signatures and generated fractionation among REE and other “immobile” trace elements. Differences in element distribution patterns between the individual profiles likely reflect a combination of primary compositional differences of parent rocks, the nature and content of accessory phases and their unequal behavior during weathering. Field appearances, textural relationships between secondary minerals and locally complex chemical profiles signal a polygenetic character of the bauxites, which is inferred to be associated with multiple cycles of weathering since Late Cretaceous times.

Acknowledgements The authors would like to thank Tilly Bouten, Helen de Waard and Anita van Leeuwen-Tolboom for help with analytical work at Utrecht University, and Pieter Vroon for XRF data produced with the facility at the Free University of Amsterdam. The Bauxite Institute of Suriname and Suralco L.L.C. kindly provided major-element data and samples from exploration drilling campaigns. This research was funded by a grant from the Suriname Environmental and Mining Foundation (SEMIF).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 102 Bauxite formation on Proterozoic bedrock of Suriname | 103

References

– Aleva, G. (1965). The buried bauxite deposits of Onverdacht, Suriname, South America. Geologie en Mijnbouw 44, 45-58. – Aleva, G. (1979). Bauxitic and other duricrusts in Suriname. A review. Geologie en Mijnbouw 58, 321-336. – Aleva, G. (1984). Lateritization, bauxitization and cyclic landscape development in the Guiana Shield., In: Jabob Jr., L. (ed.), Bauxite [Bauxite Symposium, Los Angeles, California, 1984, Proceedings]; New York. American Institute of Mining, Metallurgical, and Petroleum Engineers, 297-318. – Aleva, G. (1994). Laterites. Concepts, geology, morphology and chemistry. ISRIC, Wageningen, 169 pp. – Aleva, G. and Hilversum, A. (1984). West Suriname: known deposits and potential. In: Jacob Jr., L. (ed.): Bauxite: Proceedings of 1984 Bauxite symposium, Los Angeles. American Institute of Mining Metallurgical and Petroleum Engineers, New York, 319-348. – Aleva, G. and Wong, Th. (1998). The history of bauxite exploration and mining in Suriname. In: Wong, Th., de Vletter, D., Krook, L., Zonneveld, I. and Van Loon, A. (eds.) The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences, and Netherlands Institute of Applied Geoscience. TNO, 275-310. – Babechuk, M., Widdowson, M., Murphy, M. and Kamber, B. (2015). A combined Y/Ho, high field strength element (HFSE) and Nd isotope perspective on basalt weathering, Deccan traps, India. Chemical Geology 369, 25-41. – Banfield, J. and Eggleton, R. (1989). Apatite replacement and rare earth mobilization, fractionation, and fixation during weathering. Clays and Clay Minerals 37 (2), 113-127. – Bánki, O., Ter Steege, H., Jansen-Jacobs, M. and Raghoenandan, U. (2008). Plant diversity of the Nassau Mountains, Suriname: Report of the 2003 Expedition. Internal report. NHN-Utrecht, BBS-Paramaribo. Utrecht, Netherlands. Paramaribo, Suriname, 1-52. – Bárdossy G. and Aleva G. (1990). Lateritic bauxites. Developments in Economic Beology. Elsevier Science Publishing 27, 624 pp. – Bauxite Institute Suriname (2009). Suriname as a source of bauxite and alumina. Prospects for growth? 15th Bauxite and Alumina Seminar, February 2009, Montega Bay, Jamaica. – Berger, A. and Frei, R. (2014). The fate of chromium during tropical weathering: A laterite profile from Central 4 Madagascar. Geoderma 213, 512-532. – Billiton (1979). Feasibility study of the Bakhuis Mountains, Internal report, 1-8. – Bogatyrev, B., Zhukov, V. and Tsekhovsky, G. (2009). Formation conditions and regularities of the distribution of large and superlarge deposits. Lithology and Mineral Resources 44 (2), 135-151. – Bosma, W., Kroonenberg, S.B., Van Lissa, R.V., Maas, K. and De Roever, E. (1984). An explanation to the geology of Suriname. In: Contributions to the geology of Suriname. Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo, 1-8. – Boulangé, B. (1984). Les formations bauxitiques latéritiques de Côte d’Ivoire. Travaux et documents d’ORSTOM, Paris, 175 pp. – Boulangé, B., and Colin, F. (1994). Rare earth element mobility during conversion of nepheline syenite into laterite bauxite at Passa Quatro, Minais Gerais, Brazil. Applied Geochemistry 9, 701-711. – Boulangé, B., Muller, J.and Sigolo, J. (1990). Behaviour of the rare earth elements in a lateritic bauxite from syenite (Brazil). Geochemistry of the Earth’s surface and of mineral formation. Second international symposium, July 1990, Aix en Provence, France, 350-351. – Braun, J., Pagel, M., Herbilln, A. and Rosin, C. (1993). Mobilization and redistribution of REEs and thorium in a syenitic lateritic profile: A mass balance study. Geochimica et Cosmochimica Acta 57 (18), 4419-4434. – Braun, J., Viers, J., Dupre, B., Polve, M., Ndam, J. and Muller, J. (1998). Solid/Liquid REE fractionation in the lateritic system of Goyoum, East Cameroon: The implications for the present dynamics of the soil covers of the humid tropical regions. Geochimica et Cosmochimica Acta 62 (2), 273-299. – Coutinho, H. (1967). Suralco final report Area 10.1, Suralco L.L.C. internal report, Paramaribo, 4-6.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 103 104 | Chapter 4

– De Roever, E., Lafon, J., Delor, C., Cocherie, A., Rossi, P., Guerrot, C. and Potrel, A. (2003). The Bakhuis ultrahigh- temperature Granulite Belt (Suriname); I. Petrological and geochronological evidence for a counterclock wise P-T path at 2.07-2.05 GA. Géologie de la France (2-3-4), 175-205. – De Roever, E., Kieft, C., Murray, E., Klein, E. and Drucker, W. (1976). Surinamite, a new Mg-Al silicate from the Bakhuis Mountains, western Surinam: I. Description, occurrence and conditions of formation. American Mineralogist 61,193-199. – De Vletter, D. (1984). Economic geology and mineral potential of Suriname. In: De Vletter, D.R (ed.). Geology of Suriname 8. Mededelingen Geologisch Mijnbouwkundige Dienst Suriname 27, 91-129. – De Vletter, D., Aleva, G. and Kroonenberg, S. (1998). Research into the Precambirian of Suriname. In: The history of earth sciences in Suriname. Royal Netherlands Academy of Science, Netherlands Institute of Applied Geoscience, TNO, Amsterdam, 15-63. – Delor, C., De Roever E., Lafon J., Lahondere D., Rossi P., Cocherie A., Guerrot C. and Potrel A. (2003). The Bakhuis ultrahigh-temperature granulite belt (Suriname): II. Implication for late Transamazonian crustal stretching in a revised Guiana Shield framework. Géology de la France (2-3-4), 207-231. – Delvigne, J. (1998). Atlas of micromorphology of mineral alteration and weathering, Mineral Association of Canada, 496 pp. – Doeve, G. (1955). De bauxiet exploratie op het Nassaugebergte. Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo, 1- 68. – Dos Muchangos, A. (2000). Mineralogy and geochemistry of bauxite and bentonite deposits from Mozambique. Dissertation at Utrecht University, Faculty of Earth Sciences, 145 pp. – Eggleton, R. and Taylor, G. (2008). Impact of fire on the Weipa Bauxite, Northern Australia. Australian Journal of Earth Sciences 55 (1), 83-86. – Feret, F., Authier-Martin, M. and Sajó, I. (1997). Quantitative phase analysis of Bidi Koum Bauxites (Guinea). Clays and Clay Minerals 45 (3), 418-427. – Fitzpatrick, R. and Schwertmann, U. (1982). Al-substituted goethite: an indicator of pedogenic and other weathering environments in South Africa. Geoderma 27, 335-347. – Gong, X., Nie, Z., Qian, M., Lui, J., Pederson, D., Hobbs, D. and McDuffie, N. (2002). Gibbsite to boehmite transformation in strongly caustic and nitrate environments. Document nr. WSRC-MS-2002-00850, US Department of Energy, 1-3. – Govindaraju, K. (1995). Working values with confidence limits for twenty-six CRPG, ANRT and IWG-GIT Geostandards. Geostandards Newsletter 19, 1-32. – Green, T. and Pearson, N. (1986). Ti-rich accessory phase saturation in hydrous mafic-felsic compositions at high P, T. Chemical Geology 54 (3-4), 185-201. – Harlov, D. (2011). Petrological and experimental application of REE- and actinide-bearing accessory minerals to the study of Precambrian high-grade gneiss terranes. GSA Memoirs 207, 13-24. – Helgeson, H. (1971). Kinetics of mass transfer among silicates and aqueous solutions. Geochimica et Cosmochimica Acta 35, 421-469. – Horbe, A. and Da Costa, M. (1999). Geochemical evolution of a lateritic Sn-Zr-Th-Nb-Y-Ree-bearing ore body derived from apogranite: the case of Pitinga, Amazonas-Brazil. Journal of Geochemical Exploration 66, 339- 351. – Horbe, A. and Anand, R. (2011). Bauxite on igneous rocks from Amazonia and Western of Australia: Implication for weathering process. Journal of Geochemical Exploration 111 (1-2), 1-12. – Janssen, J. (1963). Reconnaissance exploration for bauxite in the Adampada-Kabalebo Area 1961-1963. Internal Grassalco report, Paramaribo, 1-5. – Janssen, J. (1977). Bauxite. In: Encyclopedia of Suriname, Bruijining C.F.A. and Voorhoeve, Elsevier, Amsterdam, 227-229. – Janssen, J. (1979). Bauxite and laterite hard caps in Suriname. Internal Grassalco report, Paramaribo, 1-12.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 104 Bauxite formation on Proterozoic bedrock of Suriname | 105

– Kamgang Kabeyene Beyala, V., Onana, V., Priso, E., Parisot, J. and Ekodeck, G. (2009). Behaviour of REE and mass balance calculations in a lateritic profile over chlorite schists in South Cameroon. Chemie der Erde 69, 61-73. – Kawano, M., and Tomita, K. (1995). Experimental study on the formation of clay minerals from obsidian by interaction with acid solution at 150 0C and 200 0C. Clay and Clay Minerals 43 (2), 212-222. – King, L., Hobday, D. and Mellody, M. (1964). Cyclic denudation in Suriname. Geologische Mijnbouwkundige Dienst Suriname, Paramaribo, 12 pp. – Klaver, M., De Roever, E., Nanne, J., Mason, P. and Davies, G. (2015). Charnockites and UHT metamorphism in the Bakhuis Granulite Belt, western Suriname: Evidence for two separate UHT events. Precambrian Research 262, 1-19. – Krook, L. and De Roever, E. (1975). Some aspects of bauxite formation in the Bakhuis Mountains, western Suriname. Anais 10a Conferencia Geologica Interguianas, Bélem 1, 686-695. – Kroonenberg, S. and De Roever, E. (1975). Dumortierite in cordierite pseudomorphs and in shear zones in high grade metamorphic rocks from western Suriname, In: Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo 23, 255-259 – Kroonenberg, S. (1976). Amphibolite facies and granulite facies metamorphism in the Coeroenie Lucie Area, SW Suriname. In: Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo 25, 109-209. – Kroonenberg, S. and Melitz, P. (1983). Summit levels, bedrock control and the etchplain concept in the basement of Suriname. In: Van den Berg, M. and Felix, R. (eds): Special issue in the honor of J. De Jong, Geologie en Mijnbouw 62, 389-399. – Kroonenberg, S., De Roever, E., Fraga, L., Reis, N., Faraco, M., Lafon, J., Cordani, U. and Wong, Th. (2016). Paleoproterozoic evolution of the Guiana Shield in Suriname: A revised model. Netherlands Journal of Geoscience, Geologie en Mijnbouw 95 (4), 491-522. – Laveuf, C., and Cornu, S. (2009). A review on the potentiality of rare earth elements to trace pedogenetic processes. Geoderma 154 (1-2), 1-12. – Majzlan, J. and Navrotsky, A. (2003). Thermodynamics of the goethite-diaspore solid solution. European Journal of Mineralogy 15, 495-501. – Meyer, M., Happel, U., Hausberg, J. and Wiechowski, A. (2002). Geometry and anatomy of the Los Pijiguaos bauxite deposit, Venezuela. Ore Geology Reviews 20 (1-2), 27-54. 4 – Monsels, D.A. (2016). Bauxite deposits in Suriname: Geological context and resource development. Netherlands Journal of Geosciences, Geologie en Mijnbouw 95 (4), 405-418. – Mutakyahwa, M. and Valeton, I. (1995). Late Cretaceous–Lower Tertiary weathering event and its laterite- bauxite formation in Tanzania. Mitteilungen aus dem Geologisch-Paläontologischen Institut der Universität Hamburg 78, 1-66. – Nahon, D. and Merino, E. (1997). Pseudomorphic replacement in tropical weathering: Evidence, geochemical consequences and kinetic-rheological origin. American Journal of Science 29, 393-417. – Nesbitt, H. (1979). Mobility and fractionation of rare earth elements during weathering of granodiorite. Nature 279, 206-210. – Nesbitt, H. and Young, G. (1982). Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature 299, 715-717. – Ouboter, P., Jairam, R. and Wan Tong You, K. (2007). Additional records of amphibians from the Nassau Mountains, Suriname. In: A rapid Biological assessment of the Lely and Nassau Plateaus, Suriname (with additional information on the Brownsberg Plateau) (Alonso, L. and Mol, J., (eds.), RAP bulletin of Biological Assessment 43. Conservation international, Arlington, USA, 128-129. – Patterson, S., Kurtz, H., Olson, J. and Neeley, C. (1986). World Bauxite Resources; Geology and resources of aluminum. U.S. Geological Survey professional paper, 1076-B, United States Government Printing Office, Washinghton.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 105 106 | Chapter 4

– Pollack, H. (1981). Bauxites and laterites of the Bakhuis Mountain Zone, western Suriname; a general description with emphasis on geomorphology and chemistry. In: Lateritization processes, Proceedings of International seminar on lateritization processes, December 1979, Oxford and IBH Publishing Company, New Dehli, 270-268. – Pollack, H. (1983). Land surfaces and lateritization in Suriname. In: Melfi, A. and, Carvalho, A. (eds.). Proceedings of International seminar on lateritization processes, 1982, São Paulo, Brazil, 295-308. – Priem, H., Boelrijk, N., Hebeda E., Verdurmen E. and Verschure R. (1971). Isotopic ages of the Trans-Amazonian acidic magmatism and the Nickerie metamorphic episode in the Precambrian basement of Suriname, South America. GSA Bulletin 82 (6), 1667-1680. – Priem, H., De Roever, E. and Bosma, W. (1980). A note on the age of the Paramaka metavolcanics in northern Suriname. Geologie en Mijnbouw 59, 171-173. – Rudnick, R. and Gao, S. (2004). Composition of the Continental crust. In: Treatise on Geochemistry. Holland, H. and Turekian, K. (eds.), Elsevier, Amsterdam 3, 1-64.

– Saalfeld, H. (1958). The dehydration of gibbsite and the structure of a tetragonal γ-Al2O3. Clay Minerals, The Mineralogical Society 3 (19) 249-257. – Sanematsu, K., Kon, Y. and Imai, A. (2013). Geochemical and mineralogical characteristics of ion-adsorption type REE mineralization in Phuket, Thailand. Mineralium Deposita, 48 (4), 437-451. – Schellmann, W. (1983). A new definition of laterite. Natural resources and development. Hannover/Tubingen 18, 7-12. – Schellmann, W. (1986). A new definition of laterite. On the Geochemistry of laterites. Chemie der Erde 45, 39-43. – Schellmann, W. (1994). Geochemical differentiation in laterite and bauxite formation. Catena 21, 131-143. – Scott, K. and Pain, C. (2009). Regolith Science. Csiro publishing, Australia and Springer, The Netherlands, 472 pp. – SPS and OAS (Stichting Planbureau Suriname and Organization of American States) (1988). Suriname Planatlas. SPS/OAS. Washington D.C., 48 pp. – Taylor, G. and Eggleton, R. (2008). Genesis of pisoliths and of the Weipa Bauxite deposit, northern Australia. Australian Journal of Earth Sciences 55, 87-103. – Tardy, Y. (1997). Petrology of laterites and tropical soils. Balkema, Rotterdam/Brookfield, 408 pp. – Tardy, Y. and Nahon, H. (1985). Geochemistry of laterites, stability of Al-goethite, Al-hematite, and Fe3+- kaolinite in bauxites and ferricretes: an approach to the mechanism of concretion formation. American Journal of Science 285, 865-903. – Tardy, Y. and Roquin, C. (1998). Dérive des continents, paléoclimats et alterations tropicales, BRGM Ed, 473 pp. – Ter Steege, H., Bánki, O. and Haripersaud, P. (2006). Plant diversity of the bauxite plateaus of North east Suriname. In: Alonso, L and Mol, H. (eds.), A rapid Biological assessment of the Lely and Nassau Plateaus, Suriname (with additional information on the Brownsberg Plateau). RAP Bulletin, 76-85. – Théveniaut, H. and Freyssinet, Ph. (2002). Timing of lateritization on the Guiana Shield: synthesis of paleomagnetic results from French Guiana and Suriname. Palaeography, Palaeoclimatology, Palaeoecology 178, 91-117. – Tiepolo, M., Oberti, R. and Vannucci, R. (2002). Trace-element incorporation in titanite: constrains from experimentally determined solid/liquid partition coefficients. Chemical Geology 1, 105-119. – Topp, S., Salbu, B., Roaldset, E and Jørgensen, P. (1984). Vertical distribution of trace elements in laterite soil (Suriname). Chemical Geology 47, 159-174. – Trolard, F. and Tardy, Y. (1987). The stabilities of gibbsite, boehmite, alumious goethite and aluminous hematite in bauxites, ferricretes and laterites as a function of water activity, temperature and particle size. Geochimica et Cosmochimica Acta 51, 945-957. – Valeton, I. (1972). Bauxites. Elsevier, Amsterdam, 244 pp.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 106 Bauxite formation on Proterozoic bedrock of Suriname | 107

– Valeton, I. (1983). Palaeoenvironment of lateritic bauxites with vertical and lateral differentiation. Geological Society, London, Special Publications 11 (1), 77-90. – Van den Bergh, J. (2011). Executive summary for the Nassau Plateau bauxite project. Internal Suralco L.L.C. report, 1-14. – Van der Hammen, T. and Wijmstra, T. (1964). A palynological study on the Tertiary and the Upper Cretaceous of British Guiana. Leidse Geologische Mededelingen 30, 183-241. – Van Kersen, J. (1956). Bauxite deposits in Suriname and Demerara (British Guiana), Thesis Leiden; also published In: Leidse Geologische Mededelingen 21, 247-375. – Van Lissa, R. (1975) Review of bauxite exploration in the coastal plain of Suriname. In: Contributions to the geology of Suriname 4: Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo 23, 250- 259. – Velde, B. and Meunier, A. (2008). Origin of clay minerals in soils. Springer-Verlag, Berlin Heidelberg. – Wei, X., Ji, H., Wang, S., Chu, H. and Song, C. (2014). The formation of representative lateritic weathering covers in south–central Guangxi (Southern China). Catena 118, 55-72. – Wong, Th., Krook, L., Zonneveld, J. (1998). Investigations in the coastal plain and offshore area of Suriname. In: Wong, Th., De Vletter, D., Krook, L., Zonneveld, I. and Van Loon, A. (Eds.), The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences, Netherlands Institute of Applied Geoscience (TNO), 73-100. – Zhu, C., Veblen, D., Blum, A. and Chipera, S. (2006). Naturally weathered feldspar surfaces in the Navajo Sandstone aquifer Black Mesa, Arizona; Electron Microscopic characterization. Geochimica et Cosmochimica Acta 70, 4600-4616. – Zhu, B., Fang, B. and Li., X. (2010). Dehydration reactions and kinetic parameters of gibbsite. Ceramics International 36, 2493-2498.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 107 108 | Chapter 4

Appendix 4.1 | Comparison of trace-element contents obtained for the BX-N bauxite standard by LA-ICP-MS on lithium borate glass beads with compilation values (Govindaraju, 1995) (Chapter 3).

Compilation value Measured value Reliability (%dev.) As 115 131 14 Ba 30 33 11 Ce 520 558 7.4 Cr 280 346 24 Dy 18.5 18.6 0.7 Er 11 11 0.2 Eu 4.4 4 -9.5 Gd 20 17 -13 Hf 15.2 13.9 -8.8 Ho 4.1 3.7 -10 La 355 n.d. n.d. Lu 1.8 1.7 -7.2 Nb 52 55 4.9 Nd 163 156 -4.2 Pb 135 129 -4.2 Pr 54 51 -5.2 Rb 3.6 2.9 -19 Sc 60 72 20 Sm 22 21 -4.5 Sr 110 109 -1.3 Ta 4.6 3.7 -18 Tb 3 2.7 -11 Th 50 46 -7.2 U 8.8 7.8 -11 V 350 435 24 Y 114 118 3.4 Yb 11.6 11.9 2.7 Zn 80 82 2.2 Zr 550 564 2.5

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 108 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 109 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 110 Chapter 5 Bauxite formation on Tertiary sediments in the coastal plain of Suriname

An adapted version of this chapter has been submitted for publication in Journal of South American Earth Sciences as: Monsels, D.A and van Bergen, M.J. – Bauxite formation on Tertiary sediments in the coastal plain of Suriname.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 111 112 | Chapter 5

Abstract

The lateritic bauxite deposits in Suriname formed on a range of sedimentary, metamorphic and igneous parent rocks. The coastal-plain bauxite deposits of Suriname originated on Tertiary sediments in the lowlands of the coastal plain, where they are mostly buried under a layer of young sediments. Weathering profiles of the former Successor Mines, Lelydorp-1, Kankantrie Noord, Para Noord and Coermotibo deposits reveal similarities and differences in nature and contents of major and trace elements, accessory minerals, lithology and provenance of the precursor sediments. The coastal-plain deposits are mostly topped with an iron-rich layer consisting of hematite and goethite, while gibbsite is the dominant mineral in the bauxite horizon, followed by kaolinite in the bottom section with minor quantities of anatase, quartz and zircon as accessory minerals. Geochemical analysis of selected profiles revealed significant depletion of Si, K, Na, Mg and Ca and strong, primarily residual relative enrichment of Al, Ti, Zr, Nb, Hf, Ta and Th. The enrichment of high field-strength elements and heavy rare earth elements is largely attributable to the accumulation of heavy minerals like zircon in the terrigenous precursor sediments. There is no direct genetic relationship between bauxite and the underlying saprolitic clays of the coastal-plain deposits, based on petrological evidence and trace element signatures. The complex petrological characteristics, trace-element contents and signatures of the individual coastal-plain deposits can essentially be explained by fractionation, primarily through leaching, relative and absolute enrichment, erosion, and reworking during two-stage, polycyclic bauxitization of a heterogeneous precursor.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 112 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 113

5.1 | Introduction

Bauxite, the world’s primary source of aluminium, is defined by a specific enrichment of aluminium hydroxide minerals, such as gibbsite, boehmite or diaspore. It is formed after

intense chemical weathering of different suitable porous parent rocks (Al2O3-content ≥ 10 wt.%, 0 low SiO2) under tropical and subtropical conditions (T ≥ 20 C and rainfall ≥ 1200 mm/yr) in a tectonically stable area with favourable pH and Eh (5.5.< pH < 8.0) (Aleva, 1984; Bardossy and Aleva, 1990; Tardy, 1997; Bogatyrev et al., 2009 and references therein). The most productive period of bauxitization in the Earth’s history was the Cretaceous-Eocene era when bauxite formed on Gondwanan and Laurasian platforms. The lateritic bauxite deposits of the South American Platform province have been subdivided into the Guiana Shield, Coastal Plain and Amazon Basin subprovinces (Bárdossy and Aleva, 1990). The bauxite deposits of economic interest are located on the “Main Bauxite Level” or “Sul-Americano” planation level (Paleocene– Eocene), on either Precambrian crystalline basement rocks or Late Cretaceous–Early Tertiary sedimentary precursor rocks that are often concealed by younger sediments as a result of sea- level changes or tectonic activity (Bardossy and Aleva, 1990; Théveniaut and Freyssinet, 2002). An extensive overview of the planation levels and bauxite deposits of the South American Platform province is presented in Chapter 1. The lateritic bauxites in Suriname are categorized into two major geographic areas with different precursors, properties and exploitation histories. The coastal-plain bauxites formed on sedimentary parent rocks in the coastal zone and were mined since the early 20th century, while the plateau bauxites originated on metamorphic, metasedimentary and igneous rocks in the hinterland and have not been productive to date (Chapter 2). Bauxite deposits of economic interest occur in four bauxite districts (Bárdossy and Aleva, 1990) (Figures 5.1a, b): 1. Paranam-Onverdacht-Lelydorp District (coastal-plain) which includes the Lelydorp-1, Kankantrie Noord, and Para Noord bauxite deposits. 2. Moengo-Ricanau-Jones District (coastal-plain) with the Coermotibo bauxite deposit. 3. Bakhuis District (plateau) with the Bakhuis bauxite deposit. 4. Nassau District (plateau), which includes the Brownsberg, Lely and Nassau bauxite deposits. The trace element signatures of the Suriname coastal-plain bauxites have not been 5 explored extensively to date, despite their practical applications in determining various aspects of formation and evolution of bauxite deposits, such as the roles of diagenetic, epigenetic and supergene processes in relation to ore-formation, lithology and provenance of precursor materials, physicochemical conditions during bauxitization, and mineral controls on distribution, mobilization, and fractionation of trace and rare earth elements during bauxitization (e.g., Patterson et al., 1990; Boulangé et al., 1990; Bárdossy and Aleva, 1990; Abedini et al., 2014; Wei et al., 2014b). The objective of this study is to answer key questions surrounding the coastal-plain bauxite deposits of Suriname such as the nature and content of trace elements, the precursor lithology and provenance, possible relations with underlying sediments, heavy mineral assemblages and overall bauxitization history.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 113 114 | Chapter 5

5.2 | Geological background of Surinamese bauxites

Suriname is located in the north-eastern region of South America and is part of the Guiana Shield. The Proterozoic basement occupies a large part of the country and consists of igneous, metamorphic, and metasedimentary rocks, while the overlying unconsolidated sediments form the coastal plain (Figure 5.1a). The coastal plain is subdivided into two geomorphological units: the Young Coastal Plain in the north, and the Old Coastal Plain in the south, which is an accumulation of continental sediments that were deposited along a paleo-coastline in Early Cenozoic times (Wong et al., 2009) (Chapter 1 and 2). The Proterozoic basement of Suriname is overlain by a single remnant of the Proterozoic Roraima Formation in the Tafelberg Sandstone Plateau, and crosscut by Proterozoic and Early Jurassic dolerite dikes (Figure 5.1a). A comprehensive overview of the Proterozoic basement can be found in Kroonenberg et al. (2016). Surinamese bauxite deposits have formed on two different types of parent rock (Figure 5.1a): 1. Sedimentary precursor in the coastal area (coastal-plain or low land bauxites). These bauxites belong to the Coastal Plain subprovince (Bárdossy and Aleva, 1990). They are located in an elongate "bauxite belt" that continues into Guyana and runs parallel to a paleo-coastline, which is also reported from India and other parts of South America (Valeton, 1983; Aleva and Wong, 1998). Sedimentary formations that have been deposited during the Oligocene, Miocene and Pliocene, such as the Burnside, Coesewijne and Zanderij Formation have been deposited with an unconformable contact (Figure 5.1c). These formations consist of medium to coarse, angular kaolinitic quartz sands with interbedded clay (Wong et al., 2009). 2. Crystalline rocks in the hinterland (plateau or high land bauxites). As part of the Guiana Shield subprovince these bauxites are mostly developed on intermediate to mafic Precambrian crystalline igneous or metamorphic rocks (Van Kersen, 1956; Janssen, 1979; Bárdossy and Aleva, 1990).

The coastal-plain bauxite deposits in Suriname are represented by the Onverdacht Formation, which is a time equivalent of the deltaic Saramacca Formation and the marine Alliance Formation (Figure 5.1c) (Leonard, 1984; Wong et al., 2009). The Onverdacht Formation is a 10 km wide belt, approximately 40 km inland, which stretches over ca. 40 km between the west side of the Suriname River and French Guiana, and exclusively crops out in the Moengo area (Wong et al., 2009). The Surinamese bauxites are related to the well-developed and widespread Main Aluminous Laterite Level or Main Bauxite Level (Bardossy and Aleva, 1990; Théveniaut and Freyssinet, 2002) (Figure 5.1c). Bauxitization of the top section of the Onverdacht Formation occurred during the Bauxite Hiatus (Late Eocene- Oligocene) (Figure 5.1c) (Chapter 1 and 2).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 114 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 115

Figure 5.1 | (a) Geological map of Suriname (modified after Kroonenberg et al., 2016) with bauxite districts and studied locations. Areas 1 and 2 represent groups of the coastal-plain bauxite deposits, while areas 3 and 4 represent the plateau bauxite deposits; (b) Overview of the most important bauxite deposits of the Guiana Shield (Guiana Shield subprovince) with its Phanerozoic cover (Coastal Plain subprovince) (modified after Bárdossy and Aleva, 1990).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 115 116 | Chapter 5

Figure 5.1 | (c) The planation levels, stratigraphy and pollen zones of Suriname (modified after Van der Hammen and Wijmstra, 1964; Wong, 1989; Bárdossy and Aleva, 1990; Wong et al., 2009).

5.2.2 | Local geology 5.2.2.1 | Paranam-Onverdacht-Lelydorp District The Paranam-Onverdacht-Lelydorp (POL) District (5004’-5040’N, 55005’-55011’W) originally consisted of thirteen deposits, the majority of which has been mined out in open pits. Only four of the POL deposits had outcrops (nrs. 1, 3, 5 and 6 in Figure 5.2), whereas the others were covered by a thick (up to 14 m) packet of Pleistocene clastic sediments (Bárdossy and Aleva, 1990; cf. Wong et al., 2009). Most of these bauxite-capped hills are underlain by Early–Eocene and Paleocene sediments. It is unclear whether they are remnants of individual bauxite deposits or originally belonged to one extensive Eocene sandy flat that was bauxitized (Bárdossy and Aleva, 1990). The bauxite horizon usually caps a 10–40 m high dome-shaped residual hill above the Late Cretaceous planation surface, which now slopes 2°N. The deposits have the shape of a concave lens, with the largest thickness (up to 12 m) in the center and thinner (< 1.0 m) lower-lying edges. The bauxite produced from the POL District was a high-grade trihydrate (gibbsite) type containing very little boehmite and silica (Patterson et al., 1986). The last active mine, Lelydorp-1, was closed in 2015.

5.2.2.2 | Successor Mines The Successor Mines, better known as the Successor deposits, were discovered during the Brokopondo bauxite exploration campaigns in the early 1960s and 70s. They are located in the geographic district of Para (5036’–5040’N, 54048’–55003’W), adjacent to the POL District, east of the Suriname River: Klaverblad (KLB) is located alongside the river bank, Kaaimangrasie (KMG) 16 km further east, and Caramacca (CRM) at 22 km on the watershed with the Commewijne River (Figure 5.2). The deposits have the shape of a concave lens. The overburden at KLB is on

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 116 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 117

Figure 5.2 | Sketch map of the original bauxite deposits of the Paranam-Onverdacht-Lelydorp bauxite district (1–13) and the adjacent Successor deposits (14–16). Deposits 11–12 are known as the Onverdacht deposits. The current surface area is reduced since most of the deposits have been mined out. Study areas (black outlines) comprise the three Successor deposits, the Lelydorp-1, Kankantrie Noord and Para Noord deposits (11–16) (modified after Bárdossy and Aleva, 1990).

Figure 5.3 | Schematic relative position of the Successor deposits in a south-north section. 5

Figure 5.4 | Contour maps of the top of the bauxite ore at (a) Klaverblad (KLB), (b) Kaaimangrasie (KMG) and (c) Caramacca (CRM), showing that the deposits are located on hills, each dipping in a different direction (modified after Kisoensingh, 2009). Note differences in gray scales.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 117 118 | Chapter 5

average approximately 15 m thick and consists of Miocene–Holocene sediments, while the KMG and CRM deposits are covered by Pleistocene–Holocene sediments with average thicknesses of 3 and 9 m, respectively (BHP Billiton feasibility study, 2004) (Figure 5.3). The overburden in the northern section of CRM is very thin and consists of dark-grey organic-rich clay of the Holocene Mara Formation. The thickness increases towards the N-NE (BHP Billiton feasibility study, 2004). Contour maps of the top of the ore reveal that the three deposits are located on relatively steep hills. The KLB and KMG deposit both dip ca. 1° towards the North and the CRM deposit ca. 0.6° towards the NE (Figure 5.4).

5.2.2.3 | Moengo-Ricanau-Jones District The Moengo-Ricanau-Jones (MRJ) District, also known as the Moengo Group of deposits, is located in the eastern part of the coastal plain (5029’–5041’N, 54008’–54029’W) (Figures 5.5a, b). With a surface area of approximately 20 km2, it was originally split over 25 hills, varying in size between 0.03 and 7 km2, and with elevations between +75 m in the south to -40 m in the north (Bárdossy and Aleva, 1990). The bauxite deposits of the MRJ District dip northward as the N-S cross section of Figure 5.5b shows. The bauxite-capped hills are erosion remnants of one or more relatively large peneplains incised by a drainage pattern (Aleva, 1979, 1984; Wong, et al., 1998; Bárdossy and Aleva, 1990). The deposits in this district are also mined out, with the exception of the Moengo Hill and the Coermotibo deposit (Figure 5.5a). The bauxite layer of the mined deposits was generally 3–6 m thick, and the original geological reserve was approximately 127 Mt (Bárdossy and Aleva, 1990).

Figure 5.5 | Bauxite deposits in the Moengo-Ricanau-Jones district, also known as the Moengo bauxite district (modified after Bárdossy and Aleva, 1990). (a) Original distribution of the deposits. All deposits have been mined out except for the buried Coermotibo deposit; (b) north-south section showing the northward dip of the bauxite layer and the buried position of the Coermotibo deposit.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 118 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 119

5.2.2.3.1 | Coermotibo deposit The Coermotibo deposit, which is named after the nearby Coermotibo River, is the only buried MJR deposit with a ca. 40 m thick overburden and a reserve/resource of 18/37 Mt (Bárdossy and Aleva, 1990). (Figures 5.5a, b). The buried deposit was discovered in 1959 during reconnaissance

core drilling, but despite a high average aluminium grade (51% Al2O3) and low average Fe2O3 content (4%) (Bárdossy and Aleva, 1990) it has not been exploited due to a high sulfide content

(average 7%, maximum 60%), linked to large quantities of marcasite (FeS2) and re-silicication (Chapter 2).

5.3 | Materials and methods

5.3.1 | Sample locations and studied materials The studied sites are the Successor deposits (KLB, CRM, KMG), Lelydorp-1, Para Noord and Kankantrie Noord from the POL District , while the Coermotibo deposit and the former East Group of Hills from the MJR District were investigated as well. The Successor and Lelydorp-1 deposits are currently mined out, but sampling was performed when they were still actively mined. Lateritic weathering profiles of the Klaverblad, Kaaimangrasie and Caramacca deposit were sampled down to a depth of 7.0, 5.0 and 4.5 m respectively at various vertical intervals of 25, 50 or 100 cm, based on changes in lithology (Figure 5.3). Major element XRF data for Lelydorp-1 and Coermotibo were provided by Suralco L.L.C., while those from Kankantrie Noord and Para Noord were provided by the Bauxite Institute Suriname. Data from the East Group of Hills is from Diko et al., 2001. The XRF data from the Suralco L.L.C, the Bauxite Institute Suriname and Diko et al. (2001) are all based on drill cuttings, while those of the Successor deposits are grab and chip samples from fresh mine faces.

5.3.2 | Analytical methods Mineralogy and microstructures of the Successor deposits were investigated on polished thin sections of samples with an optical microscope and an electron microprobe (JEOL JXA-8600 Superprobe) using both energy dispersive (EDS) and wavelength-dispersive (WDS) analytical 5 techniques. Back-scatter electron imaging (BSE) was used to identify mineral phases and to study textural relationships. Quantitative compositions of mineral phases were determined in representative textural domains of selected samples from the Klaverblad, Kaaimangrasie and Caramacca deposit. X-ray diffraction (XRD) patterns were collected from randomly oriented powder samples using a Bruker D2 Phaser X-ray diffractometer, operated in a step-scan mode, with Co-Kα radiation (1.78897 Å). The counting time was 66 sec/step, the step size 0.050 and the range 5–850. Total acquisition time per sample was approximately 15 minutes. Major element compositions were determined by X-ray fluorescence (XRF) on fused glass beads (lithium borate) with a Thermo ARL9400 sequential XRF (Utrecht University) and a Panalytical MagiXPro XRF (VU University Amsterdam). Loss-on-ignition data were obtained either by measuring weight loss upon heating of a powdered sample in an oven at > 1000 oC or by thermogravimetric analysis (TGA) during which weight loss was continuously monitored over a temperature range between room temperature and 1000 oC. The major element data provided by Suralco L.L.C.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 119 120 | Chapter 5

and Bauxite Institute Suriname include values for loss on ignition (LOI), total alumina (TAl2O3),

available alumina (AA143), total silica (TSiO2) and reactive silica (RSiO2). Trace-element concentrations were determined by laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) on the fused glass beads prepared for XRF, using a ThermoFischer Scientific Element 2 magnetic sector instrument, integrated with a Lambda Physik excimer laser (193 nm) with GeoLas optics (Chapter 3). Main parameters for the ablation spot setup were: 5 mJ laser energy, 10 Hz pulse repetition rate and 120 µm spot diameter. The ICP-MS operating conditions were plasma power: 1300 W; gas flow rates (L/min): cool 16.0 Ar, auxiliary 1.0 Ar, carrier: 0.685 He, 0.696 He; peak-jump scanning mode; time-resolved acquisition mode; 60 seconds total ablation time. Si was employed as internal standard. SRM- NIST 612 was used during the measurements to correct for background and drift with double- standard measurements bracketing each six samples. Reported compositions are averages of three measurements for each sample. Accuracy of the results was monitored by analyzing USGS standard BCR-2G after each six samples. The percentage of deviation from recommended values, determined in multiple sessions, was generally ≤ 10% for all reported trace elements.

5.4 | Results

5.4.1 | Field appearances, textures and mineralogy 5.4.1.1 | Successor deposits (Klaverblad, Kaaimangrasie and Caramacca) The originally, 3–15 m thick sedimentary overburden of the Successor deposits was topped by an organic-rich layer in areas overlain by swamps. The lateritic weathering profiles (Figures 5.6, 5.7) are marked by a strong heterogeneity within individual layers. They can be subdivided into three zones, from bottom to top: kaolinite-rich saprolitic clays also known as “kaolin”, followed by a Al-rich (gibbsite) and Fe-poor bauxite zone, which is topped by an upper Fe-rich zone (hematite and goethite). The term “saprolite” is used here loosely because the relation with the underlying sediments is uncertain. The plastic underlying saprolitic clays (kaolin) vary in color from pure white to yellow and purple (Figure 5.8a). The bauxite ore in the three deposits is 4–6 m thick. The most dominant ore type is massive bauxite, followed by a breccia-type bauxite, which frequently contains remnants of what possibly was a former lateritic duricrust responsible for a high iron content. Other ore types are clay-like, cellular and nodular (soft or firm), and earthy bauxite. The clay-like (or clayey) and

earthy bauxites usually have a high Al2O3 content, which also marks red- and crème-colored banded bauxite ore that occurs in the transition zone of the Caramacca deposit (Kisoensingh, 2009) and the Lelydorp-1 deposit. Lateral variations in the bauxite horizons of the sequences may show abrupt transitions in color and/or texture (Figures 5.8a, b). The color changes are primarily linked to the iron content in the different lithologies as the red and yellow bauxite are respectively hematite- and goethite-rich, while the white bauxite or kaolin contains 2–3% iron. Hematite is finely dispersed in the duricrust, bauxite and saprolite, as stains, mottles (glaebules), nodules and concretions, while it formed euhedral to subhedral crystals in voids (Figures 5.8a-e). Goethite mainly formed coats around silica-rich lenses, gibbsite- and hematite-rich nodules. These grey

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 120 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 121

Figure 5.6 | Lateritic weathering profile of the Klaverblad deposit. Note the hard microcrystalline Al-rich layer in the bauxite zone (arrows). Similar horizontal and verticale scale.

silica-rich lenses contained pink Al-rich concretions and were only observed in the KMG deposit (Figure 5.8e). Quartz sand was frequently associated with these lenses, while organic matter was occasionally present as well. The nodules and concretions have variable sizes and shapes, and are generally embedded in a fine-grained to clay-like bauxite matrix. They are generally made up of cryptocrystalline gibbsite or hematite with traces of kaolinite, with occasionally a hematite, goethite or gibbsite coat. The mottles in the bauxite zone vary in size, shape (irregular or oval) and color (red, yellow, white, crème, black), and sometimes appear to be tubular or vermicular in cross-section (Figure 5.8d). Gibbsite is the dominant mineral at the top, and kaolinite at the bottom of the weathering profiles of the Successor mines. The gibbsite-rich matrix of the bauxite varies from aphanitic (clay-size) to coarse-grained, while voids, veins and burrows are lined or filled with coarse- 5 grained secondary gibbsite crystals (Figures 5.8f-h). The BSE image of Figure 5.8f shows texturally different gibbsite-bearing domains: a dark-grey aphanitic matrix, a medium-grey fine-grained part and a light-grey coarse domain (filled burrow). The color differences in this picture mainly reflect relative abundances in Fe- and Ti-bearing minerals. A white, up to a few cm thin microcrystalline Al-rich (gibbsite) layer was also observed in certain areas of the KLB deposit (Figures 5.6, 5.8i). Various fossil textures such as tubules, burrows, root- and rodshaped concretions with different sizes (1.0–12.0 cm long, 0.5–2.0 cm thick), colors (white to brown), shapes and orientations were frequently observed near the transition zone between the bauxite and the kaolin in the studied sequences (Figure 5.8j). They occasionally form clusters and have the same color as the host matrix. They mostly consist of illuvial clay or Al-rich material. The root-shaped concretions at KMG are significantly larger than in the other two deposits. Similar concretions and tubules have been reported from the Paragominas bauxite district in Brazil (Carvalho et al., 1997).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 121 122 | Chapter 5

Figure 5.7 | Lateritic weathering profiles of the Successor deposits (KLB, KMG, and CRM). “Saprolite” is the underlying kaolinite-rich saprolitic clay layer also known as kaolin (see text).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 122 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 123

Figure 5.8 | (a) Sharp contact between nodular/clayey bauxite and purple kaolin (saprolite) with an iron-rich layer at the boundary in KMG; (b) Abrupt transition between the red and white bauxite in CRM; (c) Euhedral to subhedral hematite crystals lining a void in the bauxite zone of CRM ; (d) Side view of some mottled bauxite in KMG; (e) Grey SiO2-rich lens with pink Al-rich concretions with a yellow goethite coat in KMG; (f) BSE microphotograph of three different gibbsite textures in KMG-11. See text for explanation; (g) Microscope image of prismatic secondary gibbsite (Gbs) cluster filling void in sample KLB-05 (XPL); (h) Microscope image of voids filled with multiple coarse-grained gibbsite (Gbs) linings in sample KLB-01 (XPL); (i) 2–5 cm thick white porcelanous Al-rich layer in KLB; (j) Root-shaped concretions of the KMG deposit; (k) Microphotograph BSE image of a burrow with “in-fill” illuviation texture in sample KLB-16; (l) Burrow (highlighed with red line) filled with crème cryptocrystalline Al-rich material in CRM.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 123 124 | Chapter 5

Figure 5.9 | (a) BSE microphotograph of two euhedral zircon crystals (Zir) and a weathered anatase crystal (An) with Ti-Al-rich dark-grey areas and light-grey Fe-Ti-rich areas in KLB-11; (b) BSE microphotograph of zircon with regrowth textures in KMG-11; (c) Rounded quartz grains in KLB-12 (PPL); (d) Accumulation of coarse- and fine-grained Fe-Ti-(Mn)rich minerals in KLB-16; (e) Peculiar boxwork texture of a weathered Ti-oxide; (f) Cluster of kaolinite vermiculite stack (Kln) and muscovite flakes (Mu) in KMG-31; (g) EDS spectrum of Ti-oxide in Figure 5.9e.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 124 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 125

Figure 5.10 | (a) Repesentative XRD spectra of bauxite sample KLB-01 from the Klaverblad deposit, which is identical to the spectra of the KMG and CRM bauxite samples; (b) XRD spectra from a quartz-rich grey lens (KMG-07); (c) XRD spectra of the white kaolin sample from the KLB deposit (KLB-16); (d) XRD spectra of the purple hematite-rich saprolite sample of the CRM deposit (CRM-08A).

Investigated bauxite samples also contain zircon, ilmenite, and traces of anatase, rutile and tourmaline. The quantity of these heavy minerals increases towards the bottom of the weathering profiles where they usually display a vague sedimentary layering. They generally have a sub-angular, sub-rounded to (sub-)euhedral shape (Figures 5.9a, b). In all three Successor deposits sub- to euhedral zircon grains show zonation and re-growth textures, suggesting complex crystallization histories (Figures 5.9a, b). 5 The rounded shapes of the quartz grains may be attributable to transport (Figure 5.9c). Several Ti-phases such as anatase, rutile and ilmenite were identified by EDS and XRD analysis (Figure 5.10). The Ti-phases frequently form clusters or line certain domains such as the clay-like bauxite matrix (Figure 5.9d). The triangular-shaped Ti-rich mineral in Figure 5.9a is a weathered

rutile with Ti-rich grey areas (85% TiO2, 2% Fe2O3) and Fe-rich white areas (61% TiO2, 21% Fe2O3). In supplementary sample KLB-17 a very peculiar boxwork texture of a Ti-rich mineral with some Fe-Al-substitution was observed (Figure 5.9e, g). The underlying saprolitic clay (kaolin) mainly consist of kaolinite (flakes and vermicular stacks), with variable amounts of mica, quartz and hematite (Figures 5.9f, 5.10c, d). The XRD spectra of the CRM kaolin samples (CRM-08A and CRM-08B) revealed their hematite content (responsible for its purple color), while the KLB (KLB-18) and KMG (KMG-16) kaolin samples are Fe-poor and white (Figure 5.10a, b). Some of the analyzed samples (e.g., KMG-07) contain sufficient quartz to be detected by XRD (Figure 5.10b) despite the problem of overlapping peaks in the presence of abundant gibbsite (cf. Feret and Roy, 2002).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 125 126 | Chapter 5

Figure 5.11 | (a) Location of the exploration drill holes from the 2004 exploration campaign at Coermotibo; (b) WE and SN cross-sections of the drill-holes, showing the different lithologies with red (hematite) and grey/black (marcasite) staining (not to scale).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 126 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 127

Figure 5.12 | (a) Drill cores from the Coermotibo deposit showing the yellow clay from the overburden, the black peat, and the grey bauxite; (b) The transition from light-gray bauxite into the red groundwater laterite layer; (c) Cluster of marcasite crystals from the Coermotibo bauxite; (d) Residual bottom-layers of the lateritic weathering profile in the mined out Madoekas Hill;(e) Close up of the laminated sandstone in Figure 5.12d showing the laminated sandstone; (f) Cluster of mica flakes in the laminated sandstone of Figure 5.12e; (g) Microprobe image of a muscovite (Mu) cluster filling a void in quartz grain (Qtz); (h) Quartz pebbles in the bottom layer of the abandoned Begi Gado mine. Hammer for scale.

5.4.1.2 | Coermotibo deposit The ca. 40 m thick overburden of the Coermotibo bauxite deposit consists of alternating unconsolidated sand and clay layers of Miocene age, locally with intercalated lenses (Figure 5.11b). Some of the sand layers are aquifers. A humus-rich topsoil is thicker beneath swamps than in more elevated areas (Figure 5.12a). The thickness of the bauxite layer ranges between 0.2 m and 16 m, with a mean value of 6 m. Five texturally different types of bauxite can be distinguished: massive, slurry-like, clayey, clastic or concretion-bearing, and kaolinitic bauxite (Figures 5.11b, 5.12a, b). 5 More than 12 different bauxite sequences have been documented in this deposit (cf. NS-EW cross-section in Figures 5.11a, b; Monsels, 2004). They alternate without a specific stratigraphic order, but each of the lithologies shows a fining-upward texture. The bauxite is heterogranular and contains gibbsitic concretions. Root-shaped concretions are generally present in the transition zone below the bauxite horizon. The presence of a swamp on much of the Coermotibo

deposit probably created the reducing conditions favourable for marcasite (FeS2) formation, which produces a grey hue of the bauxite (Figures 5.12a, c). The presence of boehmite in the Coermotibo deposit has also been reported (Van der Laan, 1998). A conspicuous red layer within the grey bauxite zone in the northern part of the deposit (Figure 5.12b) is identified as groundwater laterite, as its texture is identical to the clayey or clastic nature of the surrounding bauxite. Kaolinitic bauxite is relatively Si-rich and is usually present in the transition zone but lacks in certain areas. The underlying white- to grey-colored saprolite mainly consists of kaolinite, with other clay minerals and mica flakes. Stratigraphic equivalents of these sediments

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 127 128 | Chapter 5

Figure 5.13a1 | Schematic representation of the stratigraphy, lithology and structures in the Lelydorp-1 deposit.

cropping out in adjacent bauxite deposits of Madoekas and Begi Gado, are a muscovite- and kaolinite-rich sandstone and a quartz-pebble conglomerate layer, respectively (Figures 5.12d-h).

5.4.1.3 | Lelydorp-1 deposit The thickness of the main bauxite layer in the Lelydorp-1 deposit ranged between 0.5 m to 12 m. Some specific features of this deposit are (Figure 5.13a1–5.13g): – Iron- and clay-rich lenses in the bauxite (Figure 5.13b) – Bauxite dikes in the underlying saprolite (Figure 5.13c) – Kaolinite-rich diapir-like structures which penetrate the bauxite horizon from below – A secondary bauxite layer in the underlying saprolitic clays – Up to 30 cm large spheroids in the transition zone (Figures 5.13d–f). Their cores are either empty or filled with gibbsite or kaolinite, whereas their 1–4 cm thick walls consist of gibbsite and or kaolinite, with parallel bands of alternating colors (crème: gibbsite or kaolinite, pink: Fe-rich) (Figure 5.13f). – Depending on the location within the deposit, the underlying kaolinite-rich saprolitic clay had different appearances, ranging from pure white to a layered texture (white: Fe-poor, purple: Fe-rich (Figures 5.13e, 5.13g). It also contains clusters of coarse-grained colorless micas in certain areas.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 128 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 129

Figure 5.13 | (a2) Overburden of the Lelydorp-1 deposit; (b) Iron-rich lens in the bauxite layer (highlighted with red line); (c) Bauxite (BXT) dike surrounded by underlying saprolitic clays (KLN). Pen for scale; (d) Spheroid highlighted with red box. Hammer for scale; (e) Coarse-grained mica-rich underlying saprolite with small spheroid highlighted in red box; (f) Cut slab of a spheroid displaying horizontal layering with perpendicular fractures; (g) Banded underlying saprolitic clay (kaolin).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 129 130 | Chapter 5

Table 5.1 | Concentration averages and standard deviations for main oxides and loss on ignition (LOI) (in weight percentages), determined in grab samples from studied profiles (Klaverblad, Caramacca, Kaaimangrasie) and exploration and/or grade control cores from the Paranam-Onverdacht-Lelydorp (POL) District (Lelydorp-1, Kankantrie Noord, Para Noord) and the Moengo-Ricanau-Jones (MRJ) District (Coermotibo, East Hills). * Data from Diko et al. (2001).

Area Deposit SiO2 (%) Al2O3 (%) TiO2 (%) Fe2O3 (%) LOI (%) n Successor Klaverblad 6.0 ± 4.1 47.6 ± 9.4 1.8 ± 0.4 20.0 ± 16.5 24.3 ± 4.6 13 Deposits Caramacca 11.3 ± 16.0 47.3 ± 10.4 2.2 ± 0.6 15.2 ±14.8 23.8 ± 6.6 10 Kaaimangrasie 4.6 ± 10.3 45.7 ± 12.1 2.0 ± 0.6 27.2 ± 12.7 20.2 ± 6.4 9 POL Lelydorp-1 22.6 ± 16.8 49.7 ± 10.3 2.3 ± 0.7 1.5 ± 0.8 23.8 ± 6.5 36 District Kankantrie Noord 17.4 ± 20.3 51.4 ± 12.7 2.1 ± 0.7 1.4 ± 1.5 26.5 ± 8.1 243 Para Noord 12 .6 ± 10.4 53.9 ± 6.5 2.3 ± 0.8 2.1 ± 1.8 26.7± 4.9 287 MJR Coermotibo 6.9 ± 9.3 57.9 ± 5.3 3.3 ± 0.4 2.1 ± 2.9 28.7 ± 4.3 44 District East Hills* 10.2 ± 11.7 49.4 ± 8.3 2.3 ± 0.6 12.1 ± 10.2 25.3 ± 5.6 ca. 9250

5.4.5 | Geochemistry 5.4.5.1 | Major elements Major element data for investigated profiles (Successor deposits) and exploration drill cores (selected deposits in the Paranam-Onverdacht-Lelydorp and Moengo-Ricanau-Jones Districts) are summarized in Table 5.1, and complete major and trace element data sets for the Successor laterititic weathering profiles and supplementary mine samples are reported in Table 5.2a-c. Vertical concentration trends are shown in Figure 5.14.

All of the investigated bauxites are high-grade, with average Al2O3 concentrations ranging

between 45 and 58%. The highest average Al2O3 content was found in the Coermotibo deposit

(57.9 ± 5.3 %), which is further marked by significant amounts of sulfur (average SO3=7.1%).

Silica contents are relatively high in the POL District (average SiO2=12–23 %). The high average silica concentration of 22.6% in Lelydorp-1 probably reflects the large quantity of intercalated sand and kaolinite-rich layers and lenses. Iron contents are generally much higher in the

Successor deposits (average Fe2O3=15–27%) than in the POL and MRJ Districts. Overall, Ca, Mg,

Na and K are strongly depleted in all of the weathering profiles. Average TiO2 contents show only little variation between the different deposits. The principal chemical differences between the bauxite deposits are also visible in the

ternary Al2O3-Fe2O3-SiO2 diagram of Schellman (1986), used to determine the degree of lateritisation (Figure 5.15). The plots illustrate the iron-rich nature and overall high degree of lateritization of the Successor samples relative to the deposits of the POL District (Lelydorp-1, Kankantrie Noord and the Para Noord) and the Coermotibo deposit in the MRJ District, all of which show uniform signatures and a wider spread in lateritization degree. In general, the degree of lateritization becomes higher towards the top of the weathering profiles.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 130 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 131 19 19 88 1.2 3.0 9.1 197 181 100 2.31 0.84 0.17 6.75 0.081 0.011 0.002 27.13 0.101 0.031 55.26 14.07 6 16 14 0.3 2.0 8.5 201 210 107 100 2.40 0.77 0.17 9.15 0.063 0.002 0.001 29.45 0.056 0.007 57.94 18 17 11 ea. + mass. bxt.ea. + mass. Al-rich hard + concr. bxt. cl. 4.75 5.5 4 Klaverblad lateritic weathering profile weathering lateritic Klaverblad

mass.+ brec.+ cell. bxt. cell. brec.+ mass.+ 5 38 83 34 59 68 45 30 28 122 147 20 24 13 36 30 33 15 15 5.0 1.8 16 12 12 19 21 20 12 11 12 13 13 10 16 18 16 8.3 9.0 13 8.8 13 11 10 10 7.6 6.4 10 17 0.4 0.7 0.4 0.8 0.6 0.3 0.1 0.1 0.3 0.6 155 176 124 248 237 222 129 96 189 144 405 364 249 568 624 771 357 233 194 76 100 100 100 100 100 100 100 100 100 100 1.55 2.03 1.97 1.520.16 1.47 0.18 1.23 0.12 1.08 0.20 1.49 0.24 1.83 0.23 1.96 0.12 0.12 0.17 0.13 3.48 3.15 2.01 6.45 4.30 3.24 2.61 3.88 11.35 11.71 0.25 0.75 1.25 1.75 2.25 2.75 3.25 0.002 0.005 n.d. 0.027 0.022 0.010 n.d. 0.003 n.d. 0.004 0.041 0.045 0.038 0.044 0.051 0.039 0.038 0.038 0.054 0.059 0.003 0.003 0.002 0.00922.37 0.006 24.39 0.007 26.03 0.002 16.92 0.003 20.02 0.003 17.19 0.001 21.17 24.42 26.63 28.22 0.0390.006 0.036 0.012 0.035 0.007 0.023 0.012 0.023 0.008 0.036 0.007 0.034 0.003 0.096 0.005 0.027 0.008 0.109 0.018 44.68 47.6427.68 50.94 22.50 33.28 18.84 38.03 41.51 32.83 35.83 40.19 45.16 46.88 34.74 53.16 23.06 6.78 56.44 1.36 Major-oxide (XRF) and trace-element (LA-ICP-MS) contents of samples from the Klaverblad laterite profile. Abbreviations: brec.= breccia-like, bxt.= bauxite, bxt.= bauxite, breccia-like, brec.= Abbreviations: profile. the Klaverblad laterite Major-oxide of samples from (XRF) and trace-element contents (LA-ICP-MS) 3 (%) 3 5 O 2 2 O O 2 2 O 2 O 2 2 Y Sr Rb As Zn Cr V MgO Na Sc (ppm) CaO Total Lith. SiO Depth (m) K P TiO Fe MnO LOI Al Sample KLB-01 KLB-02 KLB-03 KLB-04 KLB-05A KLB-06 KLB-07 KLB-08 KLB-09 KLB-10 KLB-11 KLB-12 Table 5.2a | Table cell.= cellular, cl.= clayey, concr.= concretions, ea. = earthy, lith.= lithology, mass.= massive, n.d.= not detected. n.d.= massive, mass.= lith.= lithology, ea. = earthy, concretions, concr.= clayey, cl.= cellular, cell.=

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 131 132 | Chapter 5 69 20 28 11 56 16 69 90 49 6.3 3.1 2.4 2.6 6.0 1.1 8.8 2.4 171 0.38 0.97 6.75 1143 6 17 22 14 71 18 71 95 49 9.0 3.6 2.4 2.3 6.1 1.2 9.2 2.9 123 174 785 0.37 0.85 11 33 51 13 60 60 43 ea. + mass. bxt.ea. + mass. Al-rich hard + concr. bxt. cl. 4.75 5.5 4 Klaverblad lateritic weathering profile weathering lateritic Klaverblad mass.+ brec.+ cell. bxt. cell. brec.+ mass.+ 85 73 64 106 123 111 58 63 116 88 10 9.1 8.4 17 16 13 13 8.1 14 16 24 22 16 15 11 9.2 19 33 15 35 13 23 21 14 11 11 39 30 77 29 48 45 31 28 25 90 123 17 38 20 24 23 16 15 18 45 26 47 23 38 51 26 21 19 51 30 41 33 33 30 28 25 34 50 6.5 7.3 5.1 7.1 7.6 5.7 3.5 4.8 7.9 7.3 2.1 2.8 2.5 2.2 2.1 1.8 1.5 2.3 2.8 3.2 1.7 2.4 2.2 1.5 1.5 1.2 1.0 1.8 2.9 2.8 1.3 2.0 1.6 1.4 1.3 0.94 0.97 1.3 2.5 2.5 2.2 3.7 2.3 2.6 2.8 1.9 1.4 1.8 4.2 5.0 2.4 4.7 2.2 3.3 3.6 2.3 1.6 1.6 4.8 6.9 3.2 7.3 2.5 4.7 5.0 3.3 2.2 2.1 7.5 9.3 3.7 9.1 3.5 5.8 5.3 3.6 3.2 3.0 10 638 933 842 612 582 449 357 824 1363 1221 0.24 0.37 0.32 0.26 0.21 0.17 0.16 0.29 0.46 0.47 0.42 0.65 0.47 0.46 0.47 0.32 0.30 0.40 0.78 0.87 0.37 0.63 0.34 0.44 0.49 0.33 0.26 0.26 0.72 0.93 0.70 1.3 0.56 0.91 0.99 0.73 0.50 0.40 1.4 1.9 0.25 0.75 1.25 1.75 2.25 2.75 3.25 U Th Pb Ta Hf Lu Yb Er Ho Dy Tb Gd Eu Sm Nd Pr Ce La Ba Nb Lith. Zr Depth (m) Sample KLB-01 KLB-02 KLB-03 KLB-04 KLB-05A KLB-06 KLB-07 KLB-08 KLB-09 KLB-10 KLB-11 KLB-12 Table 5.2a | Continued Table

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 132 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 133 14 78 48 52 31 2.0 2.4 100 144 2.12 0.81 0.07 1769 38.51 43.82 0.132 0.066 0.006 14.32 0.099 0.040 ±10.5 roots. concr. roots. 30 26 10 82 2.6 n.d. 100 209 272 277 ±10 5.79 2.68 0.19 4.59 3636 57.35 0.030 0.069 0.018 29.21 0.069 saprolite 31 21 82 8.3 2.1 n.d. 100 231 267 612 5.42 2.19 0.29 3.80 4016 ± 9.5 58.35 0.049 0.076 0.013 29.75 0.073 cell. bxt cell. 12 24 32 7.2 0.8 100 182 152 124 2.44 4.20 0.14 1745 54.84 10.88 0.048 0.021 0.003 27.36 0.064 0.003 ± 9.25 kaolin* 57 39 11 4.7 2.1 1.3 100 109 130 514 1.98 0.53 0.10 3.98 61.41 0.163 0.038 0.001 31.72 0.060 0.009 ±8.25 Supplementary samples Klaverblad 12 14 0.8 7.3 6.4 5.2 1.5 6.9 1.7 ± 8 100 2.22 0.76 0.12 3.79 60.67 0.504 0.065 0.002 31.68 0.077 0.111 0.6 5.3 5.2 4.3 4.9 7.1 1.1 9.1 7.5 100 2.38 0.87 0.13 2.64 0.94 61.29 0.377 0.062 0.001 32.08 0.083 0.081 ea. bxt ea. bxt. bxt. cl. 5 10 14 26 59 0.3 8.5 100 338 144 623 1.90 0.15 7.30 2.25 47.22 18.64 0.074 0.016 0.002 24.64 0.048 0.013 mass. bxt mass. 71 95 13 8.5 8.4 2.2 0.3 n.d. 100 151 870 2.33 0.78 0.11 2.64 2.25 61.58 0.031 0.006 0.001 32.47 0.050 3 3 (%) 5 O 2 2 O O 2 2 O 2 O 2 2 Table 5.2a | Continued Table Lith. SiO Sample KLB-05B(C) KLB-05B(R) KLB-13 KLB-14 KLB-15 KLB-16 KLB-17 KLB-18 KLB-19 Depth (m) TiO Fe MnO LOI Al K P CaO Total MgO Sc (ppm) Na V Cr Zn As Rb Sr Y Zr

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 133 134 | Chapter 5 50 54 29 53 19 50 18 56 5.7 3.6 0.8 3.5 0.6 5.0 1.2 4.1 5.7 3.9 7.4 0.99 ± 10.5 roots. concr. roots. 24 93 18 16 11 15 97 10 49 16 3.5 2.2 3.4 2.4 136 129 138 209 181 13.4 ± 10 saprolite 49 37 17 10 14 70 15 7.3 2.8 3.2 2.3 8.5 123 304 311 435 196 100 187 21.7 ± 9.5 cell. bxt cell. 44 92 57 13 46 50 21 74 8.3 1.7 6.2 6.2 1.3 4.5 5.7 0.9 3.3 6.6 124 0.90 kaolin* 37 48 32 79 33 13 42 8.5 6.6 1.4 4.8 3.3 1.4 1.6 2.4 8.1 4.2 0.58 0.53 0.23 ± 8.25 ± 9.25 Supplementary samples Klaverblad 20 2.2 4.1 1.9 2.1 ± 8 0.87 0.48 0.39 0.07 0.34 0.05 0.33 0.06 0.19 0.19 0.03 0.41 0.07 0.72 0.16 13 1.5 3.0 1.4 1.6 7.5 0.62 0.35 0.27 0.05 0.24 0.03 0.23 0.04 0.13 0.12 0.02 0.26 0.05 0.49 0.11 ea. bxt ea. bxt. bxt. cl. 36 40 26 51 22 16 59 5.8 4.0 2.9 2.1 1.1 1.4 2.4 9.4 5.7 0.79 0.36 0.39 0.23 2.25 mass. bxt mass. 43 55 32 60 27 23 43 7.0 6.0 1.3 4.6 3.8 2.0 2.2 3.0 5.4 5.7 0.64 0.67 0.34 2.25 Table 5.2a | Continued Table Lith. Nb Sample KLB-05B(C) KLB-05B(R) KLB-13 KLB-14 KLB-15 KLB-16 KLB-17 KLB-18 KLB-19 Ba Depth (m) La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 134 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 135 56 56 15 75 6.2 0.3 8.4 7.0 100 samples Supplementary black Fe-rich lens. kaolin* 6.75 6.75 6 ea. + cl. bxt. ea. + cl. 4.5 5.25 4 ea. bxt 3 Caramacca lateritic weathering profile weathering lateritic Caramacca 3 5 2 52 7.4 17 4.0 3.7 30 26 42 4.3 22 17 18 26 102 78 82 37 13 46 53 10 11 brec. + cell. bxt. + cell. brec. 9.0 5.5 13 7.8 10 10 6.3 9.2 8.7 12 12 9.5 10 20 45 18 12 12 9.3 15 15 28 0.2 0.2 7.8 2.9 2.1 0.4 0.1 0.4 2.7 1.8 1.9 5.8 6.1 14 10 13 10 16 8.2 11 15 8.4 100 100 100 100 100 100 100 100 100 100 100 417 165 779 136 136 456 172 291 72 139 173 369 183 453 470 514 324 159 308 126 83 116 0.71 0.93 15.23 6.04 5.38 1.47 0.73 1.46 42.59 38.05 33.03 0.68 1.67 2.24 2.50 2.81 2.93 2.93 1.66 2.19 1.53 1.39 1.08 0.63 0.33 0.12 0.22 0.11 0.11 0.19 0.09 0.24 0.07 0.06 0.08 0.16 0.25 0.75 34.82 60.84 49.32 59.21 59.86 50.18 42.78 44.22 38.25 33.11 29.71 21.43 43.07 3.85 7.92 1.07 0.84 18.57 31.62 28.40 2.55 14.22 23.39 63.02 0.034 0.017 0.050 0.029 0.012 0.021 0.022 0.027 0.041 0.049 0.042 0.046 0.006 0.008 0.158 0.061 0.044 0.013 0.002 0.010 0.052 0.036 0.028 0.011 0.0000.039 0.001 0.041 0.002 0.061 0.002 0.075 0.002 0.047 0.001 0.045 0.006 0.039 0.001 0.046 0.006 0.058 0.007 0.057 0.010 0.051 n.d. 0.037 19.31 31.95 24.51 30.58 30.78 26.59 23.03 23.40 14.81 12.97 12.54 13.95 0.003 n.d. 0.036 0.011 n.d. n.d. 0.010 0.008 0.026 0.057 0.038 0.027 Major-oxide (XRF) and trace-element (LA-ICP-MS) contents of samples from the Caramacca lateritic weathering profile. * kaolin= underlying saprolitic clay. clay. saprolitic kaolin= underlying * profile. weathering lateritic Caramacca the from samples Major-oxideof (XRF) trace-elementand contents (LA-ICP-MS) 3 (%) 3 5 O 2 2 O O 2 2 O 2 O 2 2 Abbreviations: brec.= breccia-like, bxt.= bauxite, cell.= cellular, cl.= clayey, concr.= concretions, ea.= earthy, lith.= lithology, mass.= massive, n.d.= not detected. n.d.= massive, mass.= lith.= lithology, ea.= earthy, concretions, concr.= clayey, cl.= cellular, cell.= bxt.= bauxite, breccia-like, brec.= Abbreviations: Table 5.2b | Table Lith. SiO Al TiO Fe MnO CaO LOI K P SampleDepth (m) CRM-01 CRM-01-02 CRM-02 CRM-03A CRM-03B CRM-04 CRM-05 CRM-06 CRM-07 CRM-08A CRM-08B CRM-32 MgO Total Sc (ppm) V Na Cr Zn As Rb Sr Y

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 135 136 | Chapter 5 10 13 17 9.2 5.9 8.3 2.7 1.1 1.0 1.4 4.5 2.0 495 0.93 0.48 0.68 samples Supplementary black Fe-rich lens. kaolin* 6.75 6.75 6 ea. + cl. bxt. ea. + cl. 4.5 5.25 4 ea. bxt 3 Caramacca lateritic weathering profile weathering lateritic Caramacca 3 2 23 29 41 48 56 47 21 37 32 35 25 18 15 77 52 42 21 10 24 39 21 20 16 23 77 52 63 31 17 42 30 12 13 26 32 150 107 106 49 19 64 55 14 14 13 10 18 14 18 15 5.7 14 15 21 10 10 10 40 33 26 12 7.3 22 21 10 17 brec. + cell. bxt. + cell. brec. 2.5 3.3 14 9.4 10 4.7 2.1 5.7 5.1 1.4 1.7 7.2 10 41 27 31 15 6.7 16 16 4.3 5.5 1.1 1.5 5.4 3.4 3.8 1.9 1.3 2.0 2.6 0.94 1.0 1.0 1.1 2.7 1.9 2.5 1.7 1.5 1.4 1.9 2.1 1.5 1.1 1.1 2.3 1.6 2.3 1.7 1.0 1.4 2.1 2.8 1.5 1.7 2.3 3.5 3.5 4.2 3.5 1.6 2.8 2.2 2.3 1.8 3.4 3.0 4.8 7.1 4.5 4.4 2.3 3.7 3.5 4.3 3.2 506 355 605 506 643 576 205 503 561 846 342 107 65 117 106 137 104 39 99 46 36 40 0.20 0.25 0.83 0.64 0.66 0.41 0.31 0.31 0.46 0.16 0.19 0.08 0.83 0.95 3.1 1.7 2.3 1.4 1.5 1.3 1.6 1.2 0.93 0.48 0.13 0.15 0.40 0.26 0.32 0.20 0.21 0.19 0.26 0.24 0.18 0.12 0.22 0.25 0.58 0.36 0.53 0.37 0.32 0.31 0.43 0.56 0.34 0.26 0.81 0.82 2.0 1.3 1.8 1.2 1.0 1.1 1.5 2.0 1.1 0.17 0.17 0.36 0.26 0.37 0.26 0.16 0.23 0.35 0.47 0.26 0.22 0.25 0.75 Continued Table 5.2b | Table Lith. Zr Nb SampleDepth (m) CRM-01 CRM-01-02 CRM-02 CRM-03A CRM-03B CRM-04 CRM-05 CRM-06 CRM-07 CRM-08A CRM-08B CRM-32 Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Yb Lu Hf Ta Pb Th U

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 136 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 137 5 18 48 21 92 11 0.8 2.5 n.d. 246 100 4.95 2.97 0.98 0.10 1153 0.046 30.89 0.001 60.02 0.023 0.019 23 19 26 7.2 2.7 8.5 4.5 372 267 517 100 3.07 1.90 0.20 0.003 0.037 26.39 0.001 48.39 19.92 0.045 0.047 4 14 22 13 10 5.5 0.9 326 n.d. 460 679 100 1.13 1.54 0.16 0.008 0.035 19.37 36.08 41.63 0.033 0.016 18 62 51 25 63 16 3.5 356 217 460 100 1.75 0.10 0.260 0.057 0.003 17.03 32.05 36.30 11.53 0.113 0.801 Si-rich pocket bxt mass. + concr. bxt cl. bxt. ea.+ cl. 3 22 30 49 7.3 1.2 7.3 n.d. 349 159 483 100 1.46 1.76 0.17 0.098 0.001 26.08 47.75 22.62 0.036 0.025 12 11 13 8.4 0.2 6.3 2.5 212 132 120 100 0.97 1.21 0.10 0.007 0.038 0.003 18.91 34.10 44.62 0.039 0.006 Kaaimangrasie lateritic weathering profile weathering Kaaimangrasie lateritic 2 12 17 13 4.8 0.3 5.6 n.d. 265 191 207 100 0.45 2.86 0.17 0.039 0.001 26.64 49.80 20.02 0.021 0.006 5 mass. + brec. + cell. bxt. + cell. + brec. mass. 1 14 3.7 9.3 0.3 5.4 6.4 n.d. 190 n.d. 206 225 100 0.48 1.86 0.17 0.036 15.75 38.97 42.71 0.022 0.008 11 5.5 0.2 6.8 4.6 5.3 0.5 n.d. 264 n.d. 102 115 100 0.49 2.08 0.08 0.035 24.24 45.69 27.35 0.023 0.007 Major-oxide (XRF) and trace-element (LA-ICP-MS) contents of samples from the Kaaimangrasie lateritic weathering profile. Abbreviations: brec.= breccia- brec.= Abbreviations: profile. weathering lateritic the Kaaimangrasie Major-oxide of samples from (XRF) and trace-element contents (LA-ICP-MS) 3 (%) 3 5 O 2 2 O O 2 2 O 2 O 2 2 Zr Y Sr Rb As Zn Cr V Na Sc (ppm) Lith. SiO MgO Total CaO Al Depth (m) TiO Fe MnO LOI K P like, bxt.= bauxite, cell.= cellular, cl.= clayey, concr.= concretions, ea.= earthy, lith.= lithology, mass.= massive, n.d.= not detected. n.d.= massive, mass.= lith.= lithology, ea.= earthy, concretions, concr.= clayey, cl.= cellular, cell.= bxt.= bauxite, like, Table 5.2c | Table Sample KMG-01 KMG-02 KMG-04 KMG-05 KMG-06 KMG-07 KMG-08 KMG-09 KMG-10

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 137 138 | Chapter 5 5 26 32 20 71 51 29 55 6.1 4.2 3.3 2.5 3.0 2.3 3.4 6.8 107 0.55 0.63 0.36 0.48 78 21 10 10 39 25 38 37 3.1 2.6 1.2 1.2 1.1 1.5 3.6 4.5 0.19 0.90 0.28 0.16 0.28 4 15 10 21 17 16 26 2.4 2.0 1.1 1.0 1.0 7.0 2.3 101 0.15 0.76 0.22 0.13 0.72 0.22 51 42 11 28 79 50 34 4.4 2.5 2.5 2.3 3.4 3.3 4.5 8.3 3.5 199 0.37 0.68 0.48 0.78 Si-rich pocket bxt mass. + concr. bxt cl. bxt. ea.+ cl. 3 49 14 10 10 27 23 50 34 3.5 2.4 1.1 1.4 1.1 1.5 3.1 0.19 0.96 0.27 0.18 0.26 35 20 14 11 19 1.5 6.9 1.4 5.7 1.4 1.2 1.2 7.0 2.1 2.5 0.15 0.90 0.93 0.32 0.18 0.29 Kaaimangrasie lateritic weathering profile weathering Kaaimangrasie lateritic 2 54 22 18 13 33 2.4 8.0 2.5 7.6 6.4 2.3 0.14 0.89 0.67 0.18 0.87 0.12 0.69 0.33 0.95 mass. + brec. + cell. bxt. + cell. + brec. mass. 1 62 15 13 10 22 1.7 8.7 1.6 5.5 5.2 1.7 0.10 0.65 0.51 0.14 0.70 0.10 0.55 0.13 0.81 44 20 15 10 27 1.9 6.4 2.1 7.9 1.0 1.0 6.6 2.0 0.5 0.17 0.95 0.77 0.22 0.14 0.80 0.21 U Th Pb Ta Hf Lu Yb Er Ho Dy Tb Gd Eu Sm Nd Pr Ce La Ba Lith. Nb Depth (m) Table 5.2c | Continued Table Sample KMG-01 KMG-02 KMG-04 KMG-05 KMG-06 KMG-07 KMG-08 KMG-09 KMG-10

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 138 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 139 19 16 11 13 2.2 5.8 114 296 100 ___ 1.59 7.62 0.08 1140 0.018 0.051 15.70 0.003 0.046 0.038 38.52 36.35 42 60 23 21 8.1 5.1 189 187 100 ___ 2.37 4.47 0.15 2898 0.040 0.049 23.37 0.008 0.054 0.145 49.00 20.34 kaolin* 14 21 30 66 3.2 9.1 n.d. 818 110 100 1.37 1.13 0.06 ± 10 0.027 0.059 16.25 0.004 0.042 0.052 40.52 40.50 20 40 13 99 11 0.6 2.0 n.d. 254 100 3.63 0.91 0.11 2.32 ± 10 1295 0.043 31.98 0.002 0.023 0.014 60.96 26 72 7.9 0.2 1.8 8.3 5.7 ± 9 n.d. 455 n.d. 193 100 3.06 0.55 0.09 0.87 0.041 32.94 0.019 0.006 62.41 Supplementary Kaaimangrasie samples 12 41 10 0.9 2.7 9.2 ± 8 n.d. 666 308 168 100 3.53 1.09 0.13 1.35 0.044 0.002 32.54 0.024 0.018 61.28 5 22 47 26 14 2.2 3.7 ± 7 n.d. 313 173 100 3.27 2.25 0.12 3.78 1481 0.043 0.002 31.17 0.020 0.042 59.30 25 43 22 94 15 1.4 3.4 211 100 2.76 1.07 0.08 1841 0.010 0.052 0.002 23.71 0.027 0.024 50.17 22.10 ea. bxt. bxt mass. ea. bxt. bxt. mass. bxt. cl. e.a.+ 3 (%) 3 5 O 2 2 O O 2 2 O 2 O 2 2 Zr Y Sr Rb As Zn Cr V Sc (ppm) Na MgO SiO Total Lith. CaO K P Depth (m) ± 6.5 TiO Fe MnO LOI Al Sample KMG-11 KMG-12 KMG-13 KMG-14 KMG-15 KMG-16 KMG-32 KMG-33 Table 5.2c | Continued Table

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 139 140 | Chapter 5 57 13 29 26 22 22 39 4.1 2.6 3.1 2.4 2.9 1.3 1.4 7.4 2.4 ___ 0.51 0.62 0.33 0.27 50 72 20 81 56 56 54 7.7 3.8 1.3 7.6 5.4 1.5 6.1 3.3 3.7 6.8 113 ___ 0.73 0.71 kaolin* 44 14 20 26 20 29 33 3.8 2.2 2.5 1.8 2.0 1.2 1.4 7.1 2.5 0.40 0.51 0.24 0.29 ± 10 97 23 32 15 59 41 27 70 6.4 4.9 3.8 2.8 3.1 2.1 2.4 5.5 0.58 0.73 0.36 0.45 ± 10 60 19 11 11 45 32 23 48 3.6 3.3 1.3 1.0 1.3 1.5 4.1 ± 9 0.22 0.29 0.18 0.94 0.32 Supplementary Kaaimangrasie samples 36 19 21 76 53 29 60 5.9 4.4 2.2 1.6 2.3 1.9 3.1 7.3 ± 8 120 0.33 0.48 0.32 0.44 25 38 22 74 51 38 65 7.2 4.6 3.8 2.9 3.4 2.6 3.3 7.1 ± 7 127 0.65 0.80 0.43 0.64 26 50 18 57 38 27 57 7.2 4.3 4.3 3.2 3.7 2.3 2.7 7.6 105 0.78 0.87 0.50 0.49 ea. bxt. bxt mass. ea. bxt. bxt. mass. bxt. cl. e.a.+ U Th Pb Ta Hf Lu Yb Er Ho Dy Tb Gd Eu Sm Nd Pr Ce La Ba Nb Lith. Depth (m) ± 6.5 Sample KMG-11 KMG-12 KMG-13 KMG-14 KMG-15 KMG-16 KMG-32 KMG-33 Table 5.2c | Continued Table

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 140 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 141

The vertical major-element trends in the Successor deposits are irregular and largely follow color and textural variations in the layers of the profiles. For example, the relatively high iron- contents (45–18%) in the top part of the KLB deposit reflects brecciated massive duricrust,

presumably the infill of a paleochannel. The SiO2-peak at 3.5 m in the KMG profile is due to the presence of the grey Si-rich pocket in the bauxite layer (Figures 5.8e and 5.14a). It should be noted that not all of the POL and MRJ profiles reached the underlying saprolitic clays, as the mining companies stopped drilling and excavating when they reached this level. Vertical trends in the POL and MRJ districts (Figure 5.14a, b) are smoother, partly because of a less complex stratigraphic and partly due to a different sampling strategy, since individual samples from the

drill cores covered relatively large intervals. In most of the POL profiles, the Al2O3 concentrations

tend to decrease at the top and bottom, where SiO2 starts to increase, corresponding to re-

silicication and the increasing saprolitic character of the deposit. A few SiO2-poor samples near the bottom of some of the drill cores of the Kankantrie Noord and Para Noord deposits might correspond to a second bauxite layer or a bauxite “dike”, which has also been observed in the Lelydorp-1 deposit. Inter-element relationships are different for different levels in the Successor profiles. In the 2 SiO2-poor upper parts, Al2O3 shows an excellent inverse correlation with Fe2O3 (R ≥ 0.99) and

no correlation with SiO2, except for the KLB profile. Conversely, in the lower parts of the KMG

and CRM profiles, where silica decreases upward, Al2O3 correlates less conspicuously with Fe2O3 2 2 (R =0.47 and 0.66, respectively) and shows an inverse relationship with SiO2 (R =0.68 and 0.85).

In the POL and MRJ districts, where Fe2O3 is much lower, Al2O3 tends to anticorrelate with SiO2.

5.4.5.2 | Trace elements Trace-element distributions are fairly coherent for a given deposit of the Successor group, but concentrations show considerable variation within the profiles studied. For example, Cr concentrations range between 56 and 514 ppm in the CRM deposit, Zr between 9 and 4016 ppm in the KLB deposit, and V between 66 and 679 ppm in the KMG deposit. Concentration profiles for trace elements are irregular and generally do not display consistent trends with depth (Figure 5.15). Only the KLB profile, where deepest levels were reached, shows a tendency of decreasing concentrations for many trace elements (REE, Y, HFSE, Sr), whereas V increases towards to the top. 5 The patterns of UCC normalized trace-element concentrations (Rudnick and Gao, 2004) point to a significant depletion of Rb and Ba, and to a lesser extent Sr and Zn (Figure 5.16), in line with the loss of Si, K, Na and Ca that accompanies weathering of silicate minerals. All three deposits are marked by peak enrichments of HFSE with a +4 or +5 valence, notably Zr, Nb, Hf, Ta, Th and U, at approximately the same degree as Al and Ti. Elevated levels are also apparent for As, V, Sc but their concentrations are more variable and follow Fe contents. The REE part of the UCC-normalized patterns is not uniform, as it is upward concave for CRM and KMG, and downward concave for KLB (Figure 5.16, see Discussion). The KLB patterns illustrate the existence of systematic variations in trace-element concentrations as a function of

Si contents. Virtually all trace elements are enriched in the layers with SiO2 > 5 %, whereas the

Fe-rich top part, wherein SiO2 < 5%, has highest contents of As, V, Sc. These systematics are less clear or absent in the (shorter) CRM and KMG profiles.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 141 142 | Chapter 5

Figure 5.14 | Vertical concentration changes of major elements in laterite profiles of the Successor deposits (KLB, KMG and CRM), Lelydorp-1, Para Noord, Kankantrie Noord, Coermotibo and a compilation of the *Moengo data. (*data from Diko et al., 2001).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 142 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 143

Figure 5.15 | Vertical concentration changes of trace elements in laterite profiles of the Klaverblad, Kaaimangrasie and Caramacca deposits.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 143 144 | Chapter 5

Figure 5.16 | Upper-continental-crust (UCC) normalized element distributions in the studied weathering

profiles. Samples indicated in red have < 5 % SiO2, while those in blue contain > 5 % SiO2.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 144 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 145

In KLB, earthy bauxites (KLB-13 and KLB-14) are strongly depleted in virtually all trace elements

except for Ba, Rb and K2O, whereas TiO2 is relatively enriched. REE patterns of KLB-09, KLB-10, KLB-11 and KLB-12 (all more siliceous) are distinct, as their LREE are less depleted. Supplementary samples KLB-16 (clayey bauxite) and KLB-19 (root-shaped concretion), both with significant

SiO2, show identical HREE enrichments but LREE are depleted in the latter. Samples from the CRM deposit are marked by higher Cr contents. Supplementary black

Fe-rich lens sample CRM-32 is depleted in Nb, Ta, TiO2, REE and Th. REE patterns show a V-shape except for relatively iron-rich kaolin samples CRM-08A and CRM-08B, as well as CRM-32, which are relatively depleted in LREE. The KMG samples tend to have lower contents of Th, U, Sc, Sr, Y, REE, Zr, Hf, Ba than those

of the other locations. The SiO2-rich pocket (KMG-07) is less depleted in Rb, Ba, K2O than other samples. It also has a flatter REE pattern with a relative enrichment of LREE. Kaolin sample KMG-32 is enriched in REE compared to the other samples, whereas all kaolin samples (KMG- 16, KMG-32, KMG-33) are relatively enriched in HREE. Conversely, the massive bauxite sample (KMG-14) is comparatively enriched in LREE.

5.5 | Discussion

5.5.1 | Mineralogy and textures 5.5.1.1 | Minerals Optical microscopy, XRD, and EDS analysis of the Successor bauxite samples detected the presence of gibbsite, hematite, goethite, Ti-oxides, anatase, and minor quantities of quartz and zircon, in line with observations elsewhere in the coastal-plain deposits (e.g., Van der Hammen, 1969; Bárdossy and Aleva, 1990). In contrast to the plateau bauxites (Chapter 4), boehmite was not encountered. Morphological textures (cf. Delvigne, 1998) point to at least two different generations of gibbsite. The first consists of a fine- to coarse-grained gibbsite-rich matrix (crypto-alteromorphs) after kaolinite (Figure 5.8f) produced by the following desillication reaction:

Al2Si2O5(OH)4 + H2O 2Al(OH)3 + 2SiO2 (aq) This desilication reaction is a function of several factors such as silica and water activity, 5 humidity, depth to the water table, and the presence of other ions (Carvalho et al., 1997; Tardy, 1997; Zhu et al., 2006, 2010; Wei et al., 2014b). There are no indications for direct conversion of feldspar into gibbsite, as observed in the plateau bauxites (Chapter 4). The second generation of gibbsite appears as coarse-grained void linings (Figures 5.8g, h). The void linings originated from the precipitation and aging of colloidal Al-rich solutions migrating in the bauxite horizon and duricrust, while the rootshaped concretions are practically tabular relict burrows that are filled with illuvial clay and gibbsite. Multiple gibbsite linings in voids and perpendicular orientation of grains against the walls of voids are consistent with an “allochthonous origin” of the mineral, referred to as allogenic deposit (Bárdossy and Aleva, 1990; Delvigne, 1998). The tabular gibbsite crystals in Figure 5.8g show spiral growth, which has also been reported from bauxites in India (Valeton, 1972). This phenomenon reflects the degree of supersaturation that determines growth rates of individual crystal faces (cf. Lee and Parkinson, 1999).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 145 146 | Chapter 5

Figure 5.17 | Triangular plots depicting the degree of lateritization in all study areas (diagram based on Schellmann, 1983, 1986).

Hematite and goethite are observed as void filling, iron mottles, nodules and stains (Figures 5.8c, e) pointing to the mobility of iron. The shape and broadening of the goethite and hematite peaks in XRD-spectra (Figure 5.10) can be attributed to variable compositions (e.g., different Al contents) and/or poor crystallinity. Iron mottles are often associated with kaolinite in mottled zones (Tardy, 1997) where they may evolve into Al-hematite-rich concretions that are subsequently dehydrated into Al-goethite along the rims (Tardy and Nahon, 1985), similar to those observed in the Successor deposits. Although several different Ti-rich phases (rutile/anatase and ilmenite) were identified in the samples, only one distinct anatase peak was detected during XRD-analysis, presumably because overlapping peaks of more abundant kaolinite and other major phases could have obscured the minor phases (Figures 5.9a, d and 5.10a). The presence of multiple Ti-oxides in the Successor deposits points to formation as alteration product of the primary oxides (cf. Grey and Reid, 1975; Anand and Gilkes, 1984; Cornu et al., 1999) or inheritance from the sediment source. In all of the study areas kaolinite is the dominant silica-bearing phase in the transition zone and underlying saprolitic clays. These zones have variable heavy mineral contents and display various sedimentary textures such as layering or bedding planes inherited from the sedimentary precursor (Figure 5.13g). The kaolin “diapirs” of the Onverdacht and Lelydorp-1 deposits (Figure 5.13c) were presumably formed during the Oligocene when bauxite-capped hills dried out down to considerable depth. The resulting fissures enhanced the permeability

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 146 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 147

of originally impervious sediments and led to re-silicication of these fractures (Bárdossy and Aleva, 1990). Broad peaks in XRD spectra point to a relatively large fraction of poorly crystalline or amorphous kaolinite in the investigated samples (Figure 5.10d).

5.5.1.2 | Textures Microscopic analysis of the tubular burrows and rootshaped concretions revealed their internal “Fill-in” or “Stopfgefüge” texture (Figure 5.8k). They were observed in the bauxite and near the bottom of Successor and Lelydorp-1 deposits, similar to that described for the nearby Kankantrie and Onverdacht deposits (Valeton, 1971; Aleva, 1994) (Figures 5.8j, l) and the Paragominas bauxite district in Brazil (Carvalho et al., 1997). Both textures are probably suggestive of a tubular void that was filled from one end with material showing a distinct layering, usually perpendicular to the length direction of the tube and flatly concave as seen from the presumed entering direction of the fill. This texture presumably formed as a result of bioturbation and mechanical illuviation where percolating rain water reached a drier soil horizon, and water from the suspension was removed by absorption, evaporation or capillary action, leaving fine deposits (cutans) oriented along percolation macrochannels (cf. Velde and Meunier, 2008). These tubular burrows were formed in-situ under synsedimentary to early diagenetic conditions, and thus represent autochthonous relic textures of the original sedimentary parent material. The bioturbation and root horizons indicate an intertidal environment associated with mangrove vegetation, tidal flats and tidal channel sediments (Valeton, 1983). Spheroids were exclusively and more frequently observed near the transition zone or kaolintic bottom layer of the Lelydorp-1 deposit. Not much is known about its formation but the absence of concentric shells in these spheroids indicates that they were probably not formed due to cyclic processes that produce pisoliths (e.g., seasonal fluctuating of the groundwater table (Valeton, 1972; Tardy and Nahon, 1985; Taylor and Eggleton, 2008), as the horizontal layering in the thick walls points to spheroidal weathering of a pre-existing layered bauxite, followed by gibbsite precipitation into the empty core of these spheroids.

5.5.2 | Major element distribution in the weathering profiles

The distribution of SiO2, Al2O3 and Fe2O3 in the investigated profiles of the Successor deposits follow the relative abundances of the main minerals that host these oxides (gibbsite, kaolinite, 5 hematite and goethite) (Figure 5.14).

According to the Al2O3-Fe2O3-SiO2 classification of Schellman (1986), the degree of lateritization ranges between moderate in the bottom parts to strong in the top parts of the bauxite profiles (Figure 5.17). The diagrams of Figure 5.17 further reveal that each deposit has

its own Al2O3-Fe2O3-SiO2 signature, apparently reflecting differences in parent rock controls and/or weathering histories.

In the SiO2-poor top parts, where silica and other mobile elements were almost completely

leached out, the inverse correlation between Al2O3 and Fe2O3 largely reflects the proportions

of gibbsite and goethite. In lower parts, where SiO2 concentrations are higher, the presence of

kaolinite is associated with lower Al2O3 and Fe2O3 contents, in agreement with less extensive

removal of soluble components. Apart from kaolinite and other clay minerals, SiO2 is also

hosted in minor amounts of quartz. There is no clear trend in the profiles for TiO2, which is predominantly stored in anatase/rutile and ilmenite.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 147 148 | Chapter 5

All of these elements were mobile to a certain extent. This is obvious for SiO2, given its susceptibility to leaching and the concentration gradients observed. Also, increased concentrations in the top parts of POL profiles (e.g., Lelydorp-1, Kankantrie Noord and Para Noord) are evidence for re-silicication via infiltration of silica-bearing fluids derived from the overburden (Valeton, 1972; Bárdossy and Aleva, 1990) (Figure 5.14). The mobility of iron is evidenced by its stronger accumulation in the duricrust relative to aluminium, and by numerous void fillings by goethite and hematite, as well as locally abundant iron mottles, nodules and stains (Figures 5.8c, e). Observed staining of the generally white iron- poor kaolin in lower parts of the profiles can also be attributed to Fe enrichment, as kaolinite may incorporate this element in its structure when poorly crystallized (Tardy, 1997; Wei et al., 2014a) (Figues 5.8a, 5.13c, 5.13g). The mobility of Fe (and Al) in bauxite is influenced by organic complexing and effective leaching processes of percolating pore waters where pH-Eh conditions play a significant role (Huang and Keller, 1972; Boulangé, 1984; Tardy and Nahon, 1985; Bárdossy and Aleva, 1990; Schwertmann, 1991; Van der Laan, 1998; Laskou, 2007). Despite the generally assumed immobility of titanium, based on the resistance of Ti- bearing minerals to weathering and low solubility of Ti in water, there are indications of titanium mobility in the Successor samples as various intergrowths of Ti,Fe-oxides rimmed distinct microscopic domains, pointing to (small-scale) mobility of all these elements (Figure 5.9d). Cornu et al. (1999) documented evidence for significant titanium mobility in soil profiles in the tropical environment of central Amazonia, which was likely linked to the complexing capacity of organic compounds and acidic conditions. Valeton (1972) suggested that Ti can also be mobilized as colloidal hydrated titanium oxides, to be ultimately dehydrated to form fine-grained anatase, which requires some re-distribution at least on subcentimeter scale. The presence of marcasite in the Coermotibo and Onverdacht deposits results from secondary sulphur enrichment, as has also been documented for other buried deposits where marshy sediments are in lateral contact with the bauxite (Valeton, 1972; Bárdossy and Aleva, 1990; Van der Laan, 1998). The marcasite is of post-diagenetic origin and presumably formed under fluctuating pH and Eh (Valeton, 1972). There are also reports of authigenetic pyrite in the Lelydorp-1 deposit (Van der Laan, 1998). The predominance of one of these iron bisulfide polymorphs is often related to kinetically controlled formation mechanisms rather than thermodynamic stability, with increased acidity (pH < 5) tending to favour marcasite as predominant phase (Murowchick and Barnes, 1986; Schoonen and Barnes, 1991). Variations in acidity and other chemical or biological factors influencing precipitation kinetics may therefore explain the presence of different sulphide phases in the bauxite profiles.

5.5.3 | Trace-element behavior in the weathering profiles Inspection of the vertical profiles (Figure 5.15) and UCC-normalized trends (Figure 5.16) provides important clues concerning parameters that controlled the behaviour of trace elements in the Successor deposits. The trace elements are liberated from the dissolving primary minerals during weathering and moved by runoff and percolating meteoric water. Since the saprolite acts as an ionic exchange medium, repeated eluviations and ion-exchange processes will induce fractionation among elements which explains the substantial substitution of trace elements, especially Zr, Nb, Hf, Ta, Th and U, into weathering resistant minerals such as zircon

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 148 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 149

or goethite, with lesser amounts absorbed on clay particles and amorphous oxides (Topp et al., 1984; Aleva, 1994; Laveuf and Cornu, 2009). The UCC-normalized plots of the three Successor deposits are fairly similar with positive LREE, MREE and HFSE anomalies (except Pb and U). The REE part of the UCC-normalized patterns of the KLB profile emphasizes the correlation between trace element concentrations and Si

contents, as practically all trace elements (REE, Y, HFSE, Sr) are enriched in the SiO2-rich layers

(SiO2 > 5 %). This association can be ascribed to the presence of heavy mineral placers in the bottom layers of the profile and, to a lesser extent, absorption of these trace elements onto clay minerals. The placer heavy minerals that are important hosts of trace elements include zircon (Zr, Nb, Hf, Ta, Th and U), anatase/rutile (Ti, V and Zr), ilmenite (Cr, V, Ti), sphene (Ti, Nb, V), and xenotime for LREE and MREE (Laveuf and Cornu, 2009). Xenotime has also been reported from the other bauxite deposits from the POL District (Aleva et al., 1969; Krook, 1969a).

5.5.4 | Field evidence for the origin of the parent material 5.5.4.1 | Parent material – arkosic sandstone versus kaolinitic clayey sediment Despite the generally accepted sedimentary nature of the precursor of the coastal-plain bauxites, there is no consensus on its composition, partly because it has often been completely transformed. According to one group of researchers the bauxites in Suriname and Guyana originated from arkosic sandstone interlayered with clayey sediments (Aleva, 1965; Aleva et al., 1969; Krook, 1969b, 1979; Wong et al., 1998, 2009 and references therein). The discovery of Paleocene sediments with 20–30% feldspar in onshore (T-28) and offshore (Co-1) drill holes has been used to support the hypothesis. A 40 m thick kaolinite-rich zone above the feldspar-rich interval in the T-28 drill hole supposedly formed during the Bauxite Hiatus (Krook, 1979a and references therein). Erosion of basement rocks in the southern parts of Suriname, promoted by vertical movements and a wet climate, supplied the arkosic material to the coastal area (Aleva, 1965; Krook, 1969b, 1979; Wong et al., 2009). In contrast, other workers favoured kaolinitic clays as parent material of the coastal-plain bauxite deposits (Bakker et al., 1953; Van Kersen, 1956; Moses and Michell, 1963; Doeve and Groeneveld Meijer, 1963). Despite the ubiquitous presence of kaolinite in the bauxite sequences, field relationships do not support a uniform genetic relationship, since the kaolinite did not 5 necessarily share the same weathering history but could also have been present already, either because of earlier formation in-situ or through deposition as sediment (Van Kersen, 1956). In an allochtonous scenario, the kaolin originated as detrital material of intensely weathered crystalline basement rocks of the Guiana Shield, and was, after fluvial transport, deposited in swamps and local lagoons (Doeve and Groeneveld Meijer, 1963; Moses and Mitchell, 1963; Bleackly, 1964). Krook (1969b) argued for a marine pathway and proposed that most of the Tertiary clays derived from the sea where they had been supplied by the Amazon River. Despite the evidence for marine incursions, the origin of Tertiary sediments in the coastal plain is dominated by terrigenous supply. Irrespective of the provenance of allochtonous forms of kaolin, its bauxitization should have commenced after regression or slight epeirogenic uplift of the coastal plain.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 149 150 | Chapter 5

5.5.4.2 | Underlying material – saprolite versus weathered clay The banded kaolin of the Lelydorp-1 deposit (Figure 5.13g) is clear evidence that the saprolite horizon is not the precursor rock, as it has a completely different sedimentary texture than the overlying bauxite deposit. This is also relevant in the Onverdacht deposit as Bárdossy and Aleva (1990) concluded that “The saprolite horizon is not the bauxite parent rock, but represents a weathered clay horizon in the Paleocene–Early Eocene regional sedimentary cover”. Other observations for the Onverdacht deposit are: (a) presence of coarse sands beneath the bauxite in drill hole P1 where a 20 m thick interval of alternating clay and sands layers was passed before basement was touched (Krook, 1969a); while Valeton (1972) observed: (b) a sharp contact between bauxite and underlying kaolin; (c) different grain sizes of heavy minerals in the bauxite and underlying kaolinitic clay and sediments; (d) an abundance of banded bauxite and kaolinitic lenses in the lower segment of the bauxite layer. Based on the stratigraphy of the other deposits of the MRJ District, it can be inferred that the Coermotibo deposit is underlain by sedimentary sand layers (Aleva and Bardossy, 1990; Wong et al., 1998). Stratigraphic equivalents of these sediments crop out in adjacent bauxite deposits of Madoekas and Begi Gado where muscovite- and kaolinite-rich coarse sands and a well-rounded to subangular quartz-pebble conglomerate layer were observed (Figure 5.12d-h). The conglomerate layer can reach a thickness of up 2 m (O’Herne, 1961).

5.5.4.3 | Significance of heavy minerals – terrigenous provenance versus weathering of local basement The association of heavy minerals in the bauxite and underlying kaolin of the Successor and Onverdacht deposits tends to support an origin from crystalline rocks in Suriname’s hinterland or as weathering product of local Precambrian basement rock. Basement was encountered near the southern margin of the Onverdacht deposit (Onoribo II deposit), where saprolite of Precambrian gneisses, schists, granites and pegmatitic veins is directly overlain by bauxite, and in several drill holes in the coastal area where it was often strongly weathered (up to 10 m thick top part) and sometimes only recognizable as bedrock because of its mineralogical composition such as the abundance of mica (Aleva et al., 1969; Van der Hammen, 1969; Bárdossy and Aleva, 1990). This source might explain the observed muscovite flakes within and beneath the kaolin of the Successor, Lelydorp-1 and Moengo deposits (Figures 5.9f, 5.12f-g, 5.13e). The heavy mineral fraction of the Lelydorp and Onverwacht deposits consists for up to 75% of zircon and contains lesser amounts of staurolite, tourmaline and other minerals such as rutile, kyanite, sillimanite and andalusite (Krook, 1969a), pointing to derivation from medium to high-grade aluminous metamorphic rock types. At Moengo Hill (MRJ area), heavy mineral assemblages in bauxite and underlying layers of sedimentary kaolinite clay and sand on top of weathered bedrock, consist chiefly of staurolite, together with tourmaline, andalusite, kyanite, rutile, anatase and zircon (Van Kersen, 1956). Petrographic observations indicate the presence of multiple types of zircon in the lateritic weathering profiles of the Successor deposits. Small rounded zircons coexist with euhedral grains containing portions with indications for regrowth (Figures 5.9a, b). Since zircon growth during bauxite formation is unlikely, the textural diversity could be due to sediment provenance, either from different sources or from a single lithology containing a heterogeneous zircon population. From the presence of zircon outgrowths on detrital zircon grains in many prehnite-

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 150 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 151

pumpellyite and greenschist facies metasedimentary rocks of all ages worldwide, it has been inferred that zircon can grow under temperatures as low as 250oC (Rasmussen, 2005). Hence, it is conceivable that various zircon types were derived from one of the low-grade metamorphic Proterozoic sediment formations in the nearby Marowijne Greenstone Belt (Figure 5.1a).

5.5.5 | Geochemical evidence for the origin of the parent material 5.5.5.1 | Relation bauxite – underlying saprolitic clay Trace-element signatures can also be used to explore possible genetic relationships between bauxite and underlying saprolitic clay intervals (kaolin) in weathering profiles. Figure 5.18 shows a comparison of UCC-normalized REE patterns of bauxite and kaolin (red dotted lines) for the Successor deposits. Parallel REE patterns would argue in favour of a common genetic history, but only if fractionation during the weathering process can be excluded. In all cases, the kaolin tends to deviate from the bauxite, either in the shape of the trends or in concentration levels of the REE. The KLB kaolin (KLB-18) is marked by REE enrichment and a relatively flat pattern. While the HREE are parallel, the LREE do not show the “bulge” around Sm, seen in the majority of the bauxite samples. Also, the kaolin composition seems to follow an overall increase of REE contents towards the bottom of the KLB profile (Figure 5.18). These observations are difficult to reconcile with weathering-induced modification of an originally homogeneous package of sedimentary parent rock. Accumulation of REE, leached from overlying layers, is conceivable but is unlikely to have affected the HREE, because this group was probably immobile as it is largely stored in zircon (Section 5.3). Downward increasing Zr and Hf contents are consistent with higher amounts of zircon near the bottom of the profile. The geochemical signatures thus point to compositional variation of sediment layers, which argues against a direct genetic relationship between bauxite and the underlying saprolitic clay layers, either in a uniform weathering profile or in a scenario wherein bauxite formed on a parent of kaolinite clay. This interpretation for KLB is tentative, because the kaolin sample (KLB-18) was not taken from the same location as that of the vertical weathering profile, and lateral variation within the deposit cannot be excluded. In the KMG and CRM profiles the REE trends of the underlying kaolin and bauxite samples are also not entirely parallel. The shapes of the HREE parts largely coincide, but the LREE deviate and are relatively depleted in the CRM profile. Although the LREE are apparently more affected 5 by mobility (leaching) during weathering than the HREE, it is difficult to envisage why they would have been preferentially removed from the kaolin, whereas typically mobile elements such as Si, Ca, Na, K, Rb are relatively enriched compared to the bauxite samples (Figure 5.16). Again, for the KMG the comparison should be considered with care since the kaolin sample (KMG-16), comes from a different location than the sampled lateritic profile. Other kaolin samples from the same deposit (KMG-32, KMG-33, not shown) display the same REE pattern albeit at different locations. The divergent geochemical trend of sample KMG-07 (Figures 5.18, 5.8e) can be explained by its different lithology as this sample was taken form a coarse-grained grey lens (Figure 5.8e) that contained numerous quartz grains and probably accommodated a relatively large amount of weathering-resistant heavy minerals. Because all of the CRM samples were collected in the same continuous vertical lateritic weathering profile, the dissimilar REE patterns suggests that the kaolin in the bottom part

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 151 152 | Chapter 5

Figure 5.18 | Comparison of UCC-normalized REE patterns of bauxite and underlying saprolitic clay layers (red dotted lines) for the Successor deposits.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 152 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 153

Figure 5.19 | Ce/Sm vs. Tb/Lu ratios of the Successor bauxites, illustrating a chemically distinct trace-element signature of the Klaverblad (KLB) deposit. See text for discussion.

Figure 5.20 | Zr/Sc vs. Th/Sc ratios of the Successor bauxites, compared with average ratios of various Proterozoic rock types (Condie, 1993). See text for discussion.

(CRM-08A and CRM-08B) is not the parent rock of the overlying bauxite, nor can they be seen as products of a single weathering process that acted on a single precursor sediment. Instead, the REE signatures likely reflect heterogeneity in the original sequence of sediments (which may have experienced a different weathering history). Deviating ratios of immobile elements in the CRM kaolin (e.g., relatively low Th/Nb, Cr/Hf) confirm this.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 153 154 | Chapter 5

5.5.5.2 | Provenance of the sediment precursor The trace-element signatures provide indications for the provenance of the sediments on which the Successor bauxites developed. Ratios of elements that remain immobile during the weathering process may serve to identify chemical properties of the sediment precursor, which may in turn be used to identify the rock type in the hinterland from which it was derived by erosion. Complicating factors are possible effects from sorting/fractionation of the various possible mineral hosts of the elements during riverine transport and deposition. For example, preferential accumulation of zircon in a sediment would enhance the concentration of elements such as Zr, Hf, HREE and Th relative to immobile elements that are incorporated in other mineral hosts (e.g., Nb, Ta). Despite obvious mobility of many of the elements analyzed and other limitations, several observations can be made. The KLB precursor is chemically distinct from that of the two other Successor deposits, as is illustrated by the Ce/Sm–Tb/Lu plot of Figure 5.19. As this difference is difficult to explain by weathering effects, the KLB bauxite must have formed on a different sediment. In view of its separate location (next to the Suriname River), it was probably derived from another source than the sediments of the more remote KMG and CRM deposits (closer to the Commewijne River), which may have had a separate, common source (Figure 5.2). Figure 5.20 compares Zr/Sc and Th/Sc ratios of the Successor bauxites with the chemical signatures of various lithologies, based on average compositions of Proterozoic rock types (Condie, 1993). The bauxites plot at the high end of the trend for Proterozoic rocks, close to the compositions of granite and cratonic sandstone, which are present in the hinterland. However, the source rock of the precursor sediment may have had lower Zr/Sc and Th/Sc ratios and coincide with another rock type further down the trend. Firstly, because immobility of Sc is questionable, and secondly because sediment recycling may have promoted the accumulation of Zr and/or Th-bearing minerals, yielding higher Zr/Sc and Th/Sc ratios of the bulk sediment. Interestingly, two KLB samples (earthy bauxites; KLB-13, KLB-14) plot close to average Proterozoic shale, felsic volcanics, TTG and greywacke, suggesting that these samples possibly had a precursor with a different lithology. The bauxites show a horizontal trend (most obvious in the KLB samples), which is consistent with zircon accumulation due to sediment recycling (cf., McLennan et al., 1993), in agreement with the abundance of this mineral in the Successor deposits.

5.6 | History of bauxitization

5.6.1 | Age – single or multiple bauxitization phases According to King (1962), Wijmstra and van der Hammen (1964), and McConell (1968), all major bauxites in Suriname and Guyana of economic importance formed during a single bauxitization phase in Eocene–Oligocene times, based on morphological correlation of paleosurfaces and analysis of pollen from underlying Paleocene–Early Eocene sediments and a Miocene overburden (see also Aleva, 1965; Pollack, 1983; Bárdossy and Aleva, 1990). Pollack (1983) and Bárdossy and Aleva (1990) attributed the Onverdacht deposit to the ‘Early Tertiary Surface’ and ‘Main Bauxite Level’, respectively (see Chapter 1), which in both cases would correspond to a Paleocene–Eocene age. Théveniaut and Freyssinet (2002) found a palaeomagnetic age of 10 Ma for the Moengo coastal-plain bauxite deposit, with an error

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 154 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 155

margin intersecting at 5 and 12 Ma, which would correspond to the Miocene ‘Late Velhas surface’ or ‘Foothill Level’. This interval coincides with one of the favourable periods for lateritic bauxitization worldwide (Bárdossy and Aleva, 1990) and is much later than the previously accepted Eocene age for the coastal-plain bauxite deposits of Suriname (Wijmstra and Van der Hammen, 1964; Aleva, 1965; Pollack, 1983; Bárdossy and Aleva, 1990). A possible explanation is the absence of a significant overburden on the Moengo deposit, as the Pleistocene transgression did not reach much further south than the foot of the outcropping bauxite-capped hills near Moengo (Bárdossy and Aleva, 1990). These exposed hills were thus susceptible to polycyclic bauxitization, thereby altering/resetting their initial age. A much older palaeomagnetic age of 60 ± 20 Ma found for the Bakhuis plateau bauxite (Théveniaut and Freyssinet, 2002) testifies that multiple bauxitization events occurred in Suriname. This finding argues against Janssen’s (1979) conclusion that all bauxite laterite caps of Suriname belong to a single old pediplain, correlated with the ‘Kopinang level”.

5.6.2 | One-step versus two-step origin Théveniaut and Freyssinet (2002) inferred that, during a major change in weathering conditions in Oligocene times, between two main lateritization and or bauxitization phases (Paleocene– Eocene and Miocene), erosion stripped 50–80 m from the Guiana Shield landscape. Associated unconformities at the Eocene–Oligocene and Oligocene–Miocene boundaries record a tectonic uplift during the Oligocene, which raised parts of Suriname’s coastal area up to 40 m (Wong et al., 1989; Krook, 1994). Changes in heavy mineral composition, clay minerals and pollen populations in coastal drill holes provide evidence for a Middle–Upper Eocene hiatus (Van der Hammen, 1969). Several authors (Moses and Michell, 1963; Bleachley, 1964; Krook, 1969) proposed a two- step origin for the coastal-plain bauxites. According to Krook (1969b) bauxitization in the coastal plain occurred in two separate stages: initial kaolinitization of feldspars was followed by desilification and ultimate weathering to bauxite (i.e., formation of gibbsite) under favourable drainage conditions. Such a two-step process is conceivable for bauxitization of parent sediment in the coastal plain, where permeability, drainage and leaching conditions are less favourable for direct bauxite formation. Single-step bauxitization, where feldspar is directly converted into gibbsite has only been observed in the plateau bauxites that formed on crystalline parent 5 rock (Chapter 4). The apparent recurrence of bauxitization phases produced new deposits (e.g., Moengo) and reactivated existing ones, resulting in an increased thickness (e.g., at the Bakhuis plateau deposit). Periods of marine regressions, crustal warping or epeirogenic uplift led to the development of dendritic drainage patterns with narrow winding valleys. These dendritic drainage patterns were noticed in the upper iron-rich layer of all coastal-plain deposits. This points to physical reworking and transport of eroded bauxite, as evidenced by the presence of dark reddish- brown iron-rich breccia-like bauxite in paleo-fluvial channels in the Successor deposits. These valleys later filled up with sediments which eventually covered the flat-topped bauxite hills as well. Features as these are consistent with formation in two stages (e.g., in the KLB deposit), implying (1) mechanical erosion and alluvial transport of an earlier lateritic bauxite formed at a higher elevation, and (2) re-bauxitization of the accumulated alluvial bauxite, similar

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 155 156 | Chapter 5

to that of the Sangaredi district in West Africa (Gow and Lozej, 1993; Chardon, 2006). Post- burial bauxitization as proposed by de Vletter (1963) and Aleva (1965) is conceivable as well, but improbable in view of bad drainage conditions under the thick overburden and the high groundwater level.

5.7 | Conclusions

– Vertical compositional profiles, studied in selected coastal-plain bauxite deposits, revealed that major-element variations within the bauxite sequences mainly reflect distributions of secondary minerals that originated through weathering of terrigenous clastic sediments. Apart from showing classical relationships between mineralogy and chemistry of lateritic

bauxites (high Fe2O3 contents in goethite and hematite-rich duricrust, maximum Al2O3

contents in gibbsite-dominated bauxite layers and highest SiO2 contents in kaolinite-rich saprolite), there is considerable compositional variability beyond the effects of leaching and residual accumulation, due to heterogeneity in the original sediment stratigraphy. Bauxite reworking and polycyclic bauxitization created additional lithological complexity in the deposits. The upper sections of the POL and Coermotibo deposits show evidence for re-silicication by downward percolating silica-bearing fluids, derived from the sedimentary overburden, which is probably a widespread feature in bauxites of the coastal plain. – Evidence from the investigated profiles of the Successor deposits indicates that (re-) distributions of many trace elements, particularly HFSE and REE, are mainly controlled by the nature and abundances of heavy minerals in the precursor sediments, and their stability during weathering. Patterns of relatively immobile elements largely reflect abundances of zircon and other weathering-resistant heavy minerals. In contrast, some of the trends also suggest mobility, specifically for LREE, presumably due to breakdown of their original mineral host(s). – The trace-element signatures of the Successor deposits point to heterogeneity of precursor sediments on a small spatial scale, reflecting variations in provenance or the depositional regime of the local fluviatile environment. The evidence from trace elements and heavy minerals suggests that the coastal-plain bauxites mostly developed on Tertiary terrigenous sediments that are strongly heterogeneous, largely due to their predominant origin as erosion products from a diversity of Proterozoic igneous and metamorphic rocks in Suriname’s interior. Minor contributions from basement rocks may locally have played a role as well. Conspicuous differences in trace-element signatures (e.g., Ce/Sm and Tb/Lu ratios) between Klaverblad and the two other Successor deposits demonstrate the potential variability of precursor sediments and their riverine supply in the coastal-plain bauxites. – According to field relationships and the trace-element data of the Successor deposits, there is no obvious genetic relationship between bauxite and underlying kaolinitic clay, neither in a uniform weathering profile nor in a scenario wherein bauxite formed on a parent of kaolinite clay. The presence of kaolinite layers in coastal-plain deposits is probably largely controlled by original intercalations of clayey material in the original stratigraphy, although a secondary origin, associated with bauxite formation, cannot always be excluded and should be explored in each individual case.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 156 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 157

Acknowledgements The authors would like to thank Tilly Bouten, Helen de Waard and Anita van Leeuwen-Tolboom for help with analytical work at Utrecht University. The Bauxite Institute of Suriname and Suralco L.L.C. kindly provided major element data and samples from exploration drilling campaigns. This research was funded by a grant from the Suriname Environmental and Mining Foundation (SEMIF).

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 157 158 | Chapter 5

References

– Abendini, A. and Calagari, A. (2014). REE geochemical characteristics of titanium-rich bauxites; The Permian Kanigorgeh horizon, NW Iran. Turkish Journal of Earth Sciences 23, 513-5132. – Aleva, G. (1965). The buried bauxite deposits of Onverdacht, Suriname, South America. Geologie en Mijnbouw 44, 45-58. – Aleva, G., Coutinho, H., Haug, G., Janssen, J. and Krook, L. (1969). Some sections through the bauxite belt of the Zanderij-Onverdacht-Lelydorp Area, Suriname. Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootschap 27, 85-88. – Aleva, G. (1979). Bauxitic and other duricrusts in Suriname. A review. Geologie en Mijnbouw 58, 321-336. – Aleva, G. (1981). Essential differences between the bauxite deposits along the southern and northern edges of the Guiana Shield, South America. Economic Geology 76, 1142-1152. – Aleva, G. (1994). Laterites. Concepts, geology, morphology and chemistry. ISRIC, Wageningen, 169 pp. – Aleva, G. and Wong, Th. (1998). The history of bauxite exploration and mining in Suriname. In: The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences, and Netherlands Institute of Applied Geoscience, TNO, 275-310. – Allen, V. (1952). Petrographic relations in some typical bauxite and diaspore deposits. Bulletin of the Geological Society of America 63, 649-688. – Aller, R. (1988). Benthic fauna and biogeochemical processes in marine sediments: The role of burrow structures: Nitrogen cycling in coastal marine environments. Scope, 1-38. – Anand, R. and Gilkes, R. (1984). Weathering of ilmenite in a lateritic palid zone. Clays and Clay Minerals 32 (5), 363-374. – Bárdossy, G. (1982). Karst bauxites, Bauxite deposits on carbonate rocks. Elsevier Scientific Publication co., Amsterdam-Oxford-New York, 441 pp. – Bárdossy G. and Aleva G. (1990). Lateritic bauxites. Developments in Economic Geology, Elsevier Science Publishing, 27: 569 pp. – Bauxite Institute Suriname (2009). Suriname as a source of bauxite and alumina. In: Prospects for growth? 15th Bauxite and Alumina Seminar, February 2009, Montega Bay, Jamaica. – BHP Billiton Suriname (2004). Feasibility study Successor Mines project. Volume 2 of 3, section 5, 1-7. – Bleackly, D. (1964). Bauxites and laterites of British Guiana. Geological Survey of British Guiana Bulletin 34, 156 pp. – Bird, M., Chivas, A., and Andrew, A. (1989). A stable isotope study of lateritic bauxites. Geochimica et Cosmochimica Acta 53, 1411-1420. – Bogatyrev, B., Zhukov, V. and Tsekhovsky, Y. (2009). Formation conditions and regularities of the distribution of large and superlarge bauxite deposits. Lithology and Mineral Resources 44 (2), 135-151. – Boulangé, B. (1984). Les formations bauxitiques latéritiques de Côte d”Ivoire. ORSTOM, Paris. Trav. Doc. 175, 314 pp. – Boulangé, B., Muller, J. and Sigolo, J. (1990). Behavior of the rare earth elements in a lateritic bauxite from syenite (Brazil). Geochemistry of the Earth’s surface and mineral formation, 2nd International Symposium, Aix en Provence, France, 350-351. – Boski, T. and Paepe, R. (1988). Quantitative mineralogy of bauxite profiles in SE Guinea Bissau. Catena 15, 417-432. – Brimhall, G., Christopher, L., Ford, C., Bratt, J., Taylor, G. and Warin, O. (1991). Quantitative geochemical approach to pedogenesis: importance of parent material reduction, volumetric expansion, and eolian influx in lateritisation. Geoderma 51, 51-91. – Brindley, G. and Brown, G. (1980). Crystal structures of clay minerals and their X-ray identification. Mineralogical Society, London, 495 pp.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 158 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 159

– Brown, J., Clark, D. and Elliott, J. (1953). The thermal decomposition of the alumina trihydrate, gibbsite. Journal of the Chemical Society 13, 84-88. – Carvalho, A., Boulangé, B., Melfi, A. and Lucas, Y. (eds.) (1997). Brazilian Bauxites, USP, FAPESP, Paris, ORSTOM, 331 pp. – Chardon, D., Chevillote, V., Beauvais, A., Grandin, G. and Boulangé, B. (2006). Planation, bauxites and epeirogeny: One or two paleosurfaces on the West African Margin. Geomorphology 82, 273-282. – Choubert, B. (1957). Essai sur la morphologie de la Guyana. Mémior Carte Géologique détaillée de la France, Département de la Guyana Française, Paris, 48. – Cornu, S., Lucas, Y., Lebon, E., Ambrosi, J., Luizão, Rouiller, Bonnay, M. and Neal, C. (1999). Evidence of titanium mobility in soil profiles. Geoderma 91, (3-4), 281-295 – Deer, W., Howie, R. and Zussman, J. (1997). Rockforming minerals: Volume 1A: Orthosilicates. Geological Society, 932 pp. – Delor, C., de Roever, E., Lafon, J., Lahondere, D., Rossi, P., Cocherie, A., Guerrot, C. and Potrel, A. (2003). The Bakhuis ultrahigh-temperature Granulite Belt (Suriname): II. Implication for late Transamazonian crustal stretching in a revised Guiana Shield framework. Géology de la France (2-3-4), 207-231. – Delvigne, J. (1998). Atlas of micromorphology of mineral alteration and weathering, Mineral Association of Canada, 509 pp. – Diko, L., Vervoort, A. and Vergauwen, I. (2001). Geostatistical modelling of lateritic bauxite orebodies in Suriname: effect of the vertical dimension, Journal of Geochemical Exploration, 73, 131-153. – Doeve, G. (1955). De bauxietexploratie op het Nassaugebergte. Internal report Geologische Mijnbouwkundige Dienst Suriname, 68 pp. – Doeve, G. and Groeneveld Meijer, W. (1963). Bauxite deposits of British Guiana and Suriname in relation to underlying unconsolidated sediments suggesting two-step origin. Economic Geology 58, 1060-1062. – Feret, F. and Roy, D. (2002). Determination of quartz in bauxite by a combined X-ray diffraction and X-ray fluorescence method. Spectrochimica Acta Part B: Atomic Spectroscopy, 57(3), 551-559. – Gow, N. and Lozej, G. (1993). Bauxite. In: Sheaham, P. and Cherry, M. (eds.) Ore deposit models. Journal of the Geological Association of Canada, Geoscience Canada reprint 20 (1), 9-16. – Grey, I. and Reid, A. (1975). The structure of pseudorutile and its role in the natural alteration of ilmenite. The American Mineralogist 60, 898-906. – Grubbs, D., Rodenburg, J. and Wefers, K. (1981). Origin of aluminous goethite in Suriname bauxite. In: Bell, G. (ed.) Light Metals Proc. Tech. Sess. TMS Light Metal Com., 110th AIME Annual Meeting, Chicago, Metall. Soc. AIME. – Huang, W. and Keller, W. (1972). Geochemical mechanisms for the dissolution, transport, and deposition of 5 aluminum in the zone of weathering. Clays and Clay Minerals 20, 69-74. – Janssen, J. (1970). Preliminary report on the Coermotibo Exploration. Internal Suralco L.L.C. report, Paramaribo, Suriname, 1-8. – Janssen, J., (1979). Bauxite and laterite hard caps in Suriname. Unpublished internal Grassalco report, 1-12. – King, L.C., Hobday, D.K. and Mellody, M. (1964). Cyclic denudation in Suriname. Internal report Geologische Mijnbouwkundige Dienst Suriname, Paramaribo, 1-12. – Kisoensingh, M. (2009). Description and classification of the Klaverblad, Kaaimangrasie, Caramacca deposit, district of Commewijne, Suriname. Unpublished BSc. thesis, Anton de Kom University of Suriname, 1-60. – Klaver, M., De Roever, E., Nanne, J., Mason, P. and Davies, G. (2015). Charnockites and UHT metamorphism in the Bakhuis Granulite Belt, western Suriname; Evidence for two separate UHT events. Precambrian Research 262, 1-19. – Kogel, J., Trivedi, N. and Barker, J. (2006). Industrial minerals and rocks: Commodities, Markets and Uses, 1548 pp.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 159 160 | Chapter 5

– Krook, L. (1969a). Investigations on the mineralogical composition of the Tertiairy and Quaternary sands in northern Suriname. In: Proc.7th Guiana. Geol. Conf., Paramaribo, 1966. Verhandelingen van het Koninklijk Nederlands Geologisch Mijnbouwkundig Genootschap 27, 89-100. – Krook, L. (1969b). The origin of bauxite in the coastal plain of Suriname and Guyana. Mededelingen Geologisch Mijnbouwkundige Dienst 20, 173-180. – Krook, L. (1979). Sediment petrographical studies in northern Suriname, Doctoral thesis, Free University of Amsterdam, 155 pp. – Krook, L. (1994). De geologische en geomorfologische ontwikkeling van Noord Suriname. In: Van der Steen, L. (ed.), Recente Geologische en Mijnbouwkundige Ontwikkelingen in Suriname. Publ. Found. Sci. Res.,Caribbean Region, 23-40. – Kroonenberg, S. and De Roever E. (1975). Dumortierite in cordierite pseudomorphs and in shear zones in high- grade metamorphic rocks from western Suriname. In: Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo 23, 255-259 – Kroonenberg, S. (1976). Amphibolite facies and granulite facies metamorphism in the Coeroenie Lucie Area, SW Suriname. In: Mededelingen Geologisch Mijnbouwkundige Dienst Suriname Paramaribo 25, 109-209. – Kroonenberg, S. and Melitz, P. (1983). Summit levels, bedrock control and the etchplain concept in the basement of Suriname. In: Van den Berg, M. and Felix, R. (eds): Special issue in the honor of J. De Jong, Geologie en Mijnbouw 62, 389-399. – Kroonenberg, S., De Roever, E., Fraga, L., Reis, N., Faraco, M., Lafon, J., Cordani, U. and Wong, Th. (2016). Paleoproterozoic evolution of the Guiana Shield in Suriname: A revised model. Netherlands Journal of Geoscience, Geologie en Mijnbouw 95 (4), 491-522. – Laskou, M., Margomenou-Leonidopoulou, G. and Balek, V. (2006). Thermal characterization of bauxite samples. Journal of Thermal Analysis and Calorimetry 84 (1), 141-145 – Laskou, M. and Economou-Eliopoulos, M. (2007). The role of microorganisms on the mineralogical and geochemical characteristics of the Parnassos-Ghiona bauxite deposits, Greece. Journal of Geochemical Exploration 93, 67-77. – Laveuf, C. and Cornu, S. (2009). A review on the potentially rare earth elements to trace pedogenetic processes. Geoderma 154, 1-12. – Lee, M. and Parkinson, G. (1999). Growth rates of gibbsite single crystals determined using in situ optical microscopy. Journal of Crystal Growth 198, 270-274. – Leonard, J. (1984). Bauxite, Proceedings of the 1984 Bauxite Symposium, Los Angeles, California, February 27-March 1: Society of Mining Engineers AIME, 387-407. – Malengreau, N., Mullier, J. and Calas, G. (1995). Spectroscopic approach for investigating the status and mobility of Ti in kaolinitic materials. Clays and Clay minerals 43 (5), 615-621. – Mange, M. and Maurer, H. (2012). Heavy minerals in colour. Springer, 147 pp. – McConnel, R. (1968). Planation surfaces in Guyana. Geographic Journal 143, (4), 506-520. – McLennan, S., Hemming, S., McDaniel, D. and Hanson, G. (1993). Geochemical approach to sedimentation, provenance, tectonics. In: Johnsson, M. and Basu, A. (eds.) Processes controlling the composition of clastic sediments: Boulder, Coldorado. Geolocial Society of America Special Paper 284, 21-40. – Monsels, D.A. (2004). New data related to the origin of the Coermotibo Deepseated deposit, district of Marowijne. Unpublished BSc thesis, Anton de Kom University of Suriname, 1-38. – Moses, J. and Michell, W. (1963). Bauxite deposits of British Guiana and Suriname in relation to underlying unconsolidated sediments suggesting two-step origin. Economic Geology 58 (2), 250-262. – Mordberg, L. (1996). Geochemistry of trace elements in Paleozoic bauxite profiles in northern Russia. Journal of Geochemical Exploration, Elsevier 57, 187-199. – Mordberg, L., Stanley, C. and Germann, K. (2001). Mineralogy and geochemistry of trace elements in bauxites: the Devonian Schugorsk deposit, Russia. Mineralogical Magazine, 65, (1), 81-101.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 160 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 161

– Murad, E. (2005). Characterization of a standard bauxite and its deferration products by Mӧssbauer spectroscopy. Minerals Engineering 18, 984-986. – Murowchick, J. and Barnes, H. (1986). Marcasite precipitation from hydrothermal solutions. Geochimica et Cosmochimica Acta 50, 2615-2629. – Neiva, A. (1995). Distribution of trace elements in feldspars of granitic aplites and pegmatites from Alijo- Sanfins, northern Portugal. Mineralogical Magazine 59, 35-45. – Nesbitt, H. (1979). Mobility and fractionation of rare earth elements during weathering of granodiorite. Nature, 279, 206-210. – O’Connel, R. (1968). Planation surfaces in Guyana. The Geographic Journal 134 (4), 506-520. – Okada, K., Nagashima, T., Kameshima, Y., Yasumori, A. and Tsukada, T. (2002). Relationship between formation conditions, properties and crystallite size of boehmite. Journal of Colloid and Interface Science 253 (2), 308- 314. – Özlü, N. (1983). Trace-element content of karst bauxites and their parent rocks in the Mediterranean belt. Mineralium Deposita 18, 469-476. – Patterson, S., Kurtz, H., Olson, J., Neeley, C. (1986). World Bauxite Resources; Geology and resources of aluminum. U.S. Geological Survey professional paper, 1076-B. United States Government Printing Office, Washinghton, 158 pp. – Pollack, H. (1981). Bauxites and laterites of the Bakhuis Mountain Zone, western Suriname; a general description with emphasis on geomorphology and chemistry. In: Lateritization processes, Proc. Int. seminar on lateritization processes, December 1979, Oxford and IBH Publishing Company,New Dehli, 270-268. – Pollack, H. (1983). Land surfaces and lateritization in Suriname. In: Melfi, A. and Carvalho, A. (eds.), Proc. 2nd Int. Sem. on Lateritisation Processes, 1982, São Paulo, Brazil, 295-308. – Poldervaart, A. (1955). Zircons in rocks: I. Sedimentary rocks, American Journal of Science 253, 433-461. – Rivas Mercury, J., Cabral, A., Paiva, A., Angélica, R., Neves, R. and Scheller, T. (2011). Thermal behavior and evolution of the mineral phases of Brazilian red mud. Journal of Thermal Analysis and Calorimetry 104, 635- 643. – Salminen, R. (Chief-editor), Batista, M. J., Bidovec, M., Demetriades, A., De Vivo, B., De Vos, W., Duris, M., Gilucis, A., Gregorauskiene, V., Halamic, J., Heitzmann, P., Lima, A., Jordan, G., Klaver, G., Klein, P., Lis, J., Locutura, J., Marsina, K., Mazreku, A., O’Connor, P., Olsson, S., Ottesen, R., Petersell, V., Plant, J., Reeder, S., Salpeteur, I., Sandström, H., Siewers, U., Steenfelt, A. and Tarvainen, T. (2005). Geochemical Atlas of Europe. Part 1: Background Information, Methodology and Maps. Espoo, Geological Survey of Finland, 526 pp. – Schellmann, W. (1983). A new definition of laterite. Natural resources and development, Hannover/Tubingen 18, 7-12. 5 – Schellman, W. (1986). A new definition of laterite. On the Geochemistry of laterites. Chemie der Erde 45, 39-43. – Schoonen, M. and Barnes, H. (1991). Reactions forming pyrite and marcasite from solution: I. Nucleation of 0 FeS2 below 100 C. Geochemica et Cosmochima Acta 55, 1495-1504. – Schwarz, T. (1997). Lateritic bauxite in central Germany and implications for Miocene palaeoclimate. Palaeogeography, Palaeoclimatology, Palaeoecology 129 (1-2), 37-50. – Schwertmann, U. (1991). Solubility and dissolution of iron oxides. Plant and Soil 130, 1-25. – Sglavo, V., Campostrini R., Maurina, S., Carturan G., Monagheddu M., Budroni G. and Cocco, G. (2000). Bauxite “red mud” in the ceramic industry. Part 1: Thermal behavior. Journal of the European Ceramic Society 20 (3), 235-244. – Silva, F., Santos, R., Sampaio, J., Garrido, F. and Medeiros, M. (2010). Study on ore dressing and characterization of different granulometric fractions that compound bauxite from Pará/Brazil. In: Light metals 2010, Johnson J. (ed.) The Minerals, Metals and Materials Society, CT2010-013-00, Comunicação Técnica, Washington State Convention Center, 68-74. – Rasmussen, B. (2005). Zircon growth in very low grade metasedimentary rock: evidence for zirconium mobility at 250 0C. Contributions to Mineralogy and Petrology 150, 146-155.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 161 162 | Chapter 5

– Rudnick, R. and Gao, S. (2004). Composition of the Continental crust. In: Treatise on Geochemistry. Holland, H. and Turekian, K. (eds.), Elsevier, Amsterdam 3, 1-64 pp. – Tardy, Y. (1997). Petrology of laterites and tropical soils, A.A. Balkema publishers, Rotterdam, Brookfield, 377 pp. – Tardy, Y., and Nahon, H. (1985). Geochemistry of laterites, stability of Al-goethite, Al-hematite, and Fe3+- kaolinite in bauxites and ferricretes: an approach to the mechanism of concretion formation. American Journal of Science 285, 865-903. – Taylor, G. and Eggleton, R. (2008). Genesis of pisolites and of the Weipa Bauxite deposit, northern Australia. Australian Journal of Earth Sciences 55, 87-103. – Ter Meulen, J. (1948). Enkele waarnemingen bij de kaolien van Paranam. Mededelingen Geologisch Mijnbouwkundige Dienst Suriname, Paramaribo (1), 1-18. – Tilley, D. and Eggleton, R. (1996). The natural occurrence of eta-alumina (ӧ-Al2O3) in bauxite. Clay and Clay Minerals 44 (5), 658-664. – Tettenhorst, R. and Corbató, C. (1998). Comparison of experimental and calculated x-ray diffraction data for boehmite. Clays and Clay Minerals 36 (2), 181-183. – Théveniaut, H. and Freyssinet, Ph. (2002). Timing of lateritization on the Guiana Shield: synthesis of paleomagnetic results from French Guiana and Suriname. Palaeography, Palaeoclimatology, Palaeoecology, Elsevier 178, 91-117. – Topp, S., Salbu, B., Roaldset, E., Jørgensen, P, (1984/1985). Vertical distribution of trace elements in laterite soil (Suriname). Chemical Geology 4 (15), 9-174. – Valeton, I. (1971). Tubular fossils in the bauxites and the underlying sediments of Suriname and Guyana. Geologie en Mijnbouw 50 (6), 733-741. – Valeton, I. (1972). Bauxites. Elsevier, Development in Soil Sciences 1, 213 pp. – Valeton, I. (1983). Paleoenvironment of lateritic bauxites with vertical and lateral differentiation. Geological Society, London, Special Publications 11, 77-90. – Valeton, I. (1994). Element concentration and formation of ore deposits by weathering. Catena 21, 99-129. – Van der Hammen, T. (1969). Introduction and short outline of the history of the ‘younger’ areas of the Guiana’s. Verhandelingen van het Koninklijk Nederlands Geologisch, Mijnbouwkundig Genootschap 27, 9-12. – Van der Hammen, T. and Wijmstra, T. (1964). A palynological study on the Tertiary and the Upper Cretaceous of British Guiana. Leidse Geologische Mededelingen 30, 183-241. – Van Kersen, J. (1956). Bauxite deposits in Suriname and Demerara (British Guiana), Thesis Leiden; also published in Leidse Geologische Mededelingen 21, 247-375. – Van der Laan, S. (1998). Mineralogy and textures of Suriname bauxites: Report prepared for Billiton. Hoogovens Research and Development, IJmuiden, 1-15. – Velde, B. and Meunier, A. (2008). The origin of clay minerals in soils and weathered rocks. Springer-Verlag Berlin Heidelberg, 378 pp. – Wei, S., Tan, W., Lui,F., Zhao, W. and Weng, L. (2014a). Surface properties and phosphate adsorption of binary systems containing goethite and kaolinite. Geoderma 213, 478-484. – Wei, X., Ji, H., Wang, S., Chu, H. and Song, C. (2014b). The formation of representative lateritic weathering covers in south –central Guangxi (Southern China). Catena 118, 55-72. – Wong, Th. (1992). Quaternary stratigraphy of Suriname. In: Prost, M. and Charron, C. (Eds.), Evolution des littoraux de Guyane et de la zone caraïbe méridionale pendant le Quaternaire. Symposium PICG 247/ Orstom, Cayenne (Guyane), 9-14 November 1990. Orstom, Paris, 559-578. – Wong, Th., Krook, L. and Zonneveld, J. (1998). Investigations in the coastal plain and offshore area of Suriname. In: Wong, Th., de Vletter, D., Krook, L., Zonneveld, I. and Van Loon, A. (Eds.) The history of Earth Sciences in Suriname. Royal Netherlands Academy of Arts and Sciences, and Netherlands Institute of Applied Geoscience, TNO, 73-100.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 162 Bauxite formation on Tertiary sediments in the coastal plain of Suriname | 163

– Wong, Th., De Kramer, R., De Boer, P., Langereis, C., Sew-A-Tjon, J. (2009). The influence of sea-level changes on tropical coastal lowlands; the Pleistocene Coropina Formation, Suriname. Sedimentary Geology 216, 125-137. – Wijmstra, T. and Van der Hammen T. (1964). Palynological data on the age of the bauxite in British Guiana and Suriname. Geologie en Mijnbouw 44, 143. – Zack, T., Kronz, A., Foley S. and Rivers, T. (2002). Trace element abundances in rutiles from eclogites and associated garnet mica schists. Chemical Geology 184, 97-122. – Zhu, C., Veblen, D., Blum, A., Chipera, S. (2006). Naturally weathered feldspar surfaces in the Navajo Sandstone aquifer Black Mesa, Arizona: Electron microscopic characterization. Geochimica et Cosmochimica Acta 70, 4600-4616. – Zhu, B., Fang, B. and Li., X. (2010). Dehydration reactions and kinetic parameters of gibbsite. Ceramics International 36, 2493-2498.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 163 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 164 Chapter 6 Synopsis

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 165 166 | Chapter 6

This thesis is devoted to the bauxites of Suriname. Within the framework of the country’s geology, it investigates relationships between the large diversity of bauxite-bearing weathering profiles across the country and the wide spectrum of bedrock lithologies on which they formed. Petrological and geochemical approaches have been used to explore key properties that mark the variability of bauxite deposits, the underlying geological controls, and the conditions that governed the re-distribution of chemical components upon the weathering process. Suriname’s bauxite deposits are traditionally distinguished into plateau bauxites and coastal-plain bauxites, a subdivision that is primarily based on their geographic and topographic position. The first group is located on relatively high plateaus in the country’s interior, while the second group is found in the lowlands of the coastal plain, where the deposits are predominantly buried under a layer of younger sediments. Suriname became one of the world’s leading producers of bauxite, alumina and aluminium in the mid-20th century, when the bauxite industry was the anchor of the national economy for many decades. The more accessible bauxite deposits in the coastal lowlands are now virtually mined out, whereas significant remaining reserves are restricted to plateau bauxite deposits in the more remote highlands. Although these occurrences have been extensively explored, they have not been brought into production to date (Chapters 1 and 2). The bauxites of Suriname developed in different episodes over a period that roughly started in the Late Cretaceous and probably continues until the present day. Their formation required surface exposure of the precursor rock in a tectonically stable environment and strong weathering, factors that were favored by their setting on a passive continental margin and by a hot and humid climate over long periods of time. These conditions prevailed in Late Cretaceous– Early Tertiary times, when bauxite formation peaked. The deposits of Suriname present a unique opportunity to study the influence of parent rocks on bauxite forming processes and properties of the final products. In this restricted area, straddling the northern boundary of the Guiana Shield, a large variety could develop, as the coastal-plain bauxites originated on Cenozoic sediments, and the plateau bauxites on a diverse collection of Proterozoic metamorphic sediments and igneous rocks. The results of this research have exposed conspicuous contrasts between the two bauxite groups. On the other hand, they also have a number of characteristics in common: – All of the investigated deposits classify as lateritic bauxites and are marked by intermediate to high degrees of lateritization (Figure 6.1). In general, the deposits display the classical sequence of a lateritic weathering profile: an iron-rich duricrust cap on top of a bauxite layer that covers a clay-rich saprolite interval, which in turn grades into weathered or fresh bedrock (often unexposed). This sequence is more consistently developed in the plateau bauxites than in the coastal-plain deposits, where irregularities are common, largely due to original heterogeneity in the stratigraphy. – Gibbsite is ubiquitous as dominant aluminium-bearing phase in all of the deposits, while kaolinite prevails in the saprolite, and hematite and goethite are prominent as iron-hosting minerals in top parts of the profiles. Boehmite is of subordinate importance as aluminium- bearing phase. This mineral may be present in the plateau bauxites in small amounts, and only sporadically in the coastal-plain deposits. This distinction is presumably attributable to differences in climate and/or drainage conditions.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 166 Synopsis | 167

Figure 6.1 | Plots depicting the degree of lateritization in coastal-plain and plateau bauxite deposits (Chapters

4 and 5). Note the differences in Fe2O3 contents, which largely reflect diversity in the nature of the parent rocks.

Figure 6.2 | Selection of vertical major-element profiles through coastal-plain and plateau bauxite deposits (Chapters 4 and 5). Trends in the plateau bauxites were largely created by in-situ weathering of Precambrian precursor rocks, whereas those in the coastal plain were also influenced by stratigraphic heterogeneity within the original Tertiary precursor sediments, and by infiltrations from the sedimentary overburden.

– Major element contents reflect the proportions of principal mineral phases that formed in

a given weathering profile. Highest SiO2 contents were usually observed in bottom parts

(saprolite), maximum Al2O3 contents in the bauxite zone, and maximum Fe2O3 contents in duricrusts at the top (examples in Figure 6.2). The observed major-element distributions 6 are consistent with the weathering processes that produce lateritic bauxites: leaching of soluble elements from unstable minerals in the precursor rock, relative accumulation of poorly soluble or immobile components, and iron enrichment regulated by fluctuating redox conditions and groundwater levels. – Overall chemical effects of immobile-versus-mobile behavior of major and trace elements during formation of the plateau and coastal-plain bauxites are similar: strong depletion of Si, K, Na, Mg and Ca as well as Rb, Ba, Sr, Zn and to a lesser extent Pb, and strong, largely residual relative enrichment of Al, Ti, Zr, Nb, Hf, Ta, Th and U. The enrichment of high field-

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 167 168 | Chapter 6

strength elements (HFSE) and heavy rare earth elements (HREE) can be directly linked to the resistance of zircon and other mineral hosts against weathering.

In this research, special attention has been paid to trace elements because of their diagnostic significance in exploring the processes of bauxite formation. Because analytical results from solution-based techniques could be negatively influenced by incomplete digestion of zircon and other resistant mineral hosts of HFSE and REE, a method was developed to analyze bauxite samples with laser ablation ICPMS on lithium borate glass beads, which avoids these problems. This technique yielded excellent results, as could be validated by measurements on a set of international standard reference materials for bauxite. Virtually all trace elements of interest were measured to within 20% of reference values with an external reproducibility of < 20% relative standard deviation (RSD), with many elements showing much better levels of accuracy and precision. Based on a new geochemical inventory, petrological evidence and field relationship, this thesis documents a number of important findings (Chapters 4 and 5): 1. The bauxite sequences of Suriname show considerable differences in architecture and composition, which are to a significant extent attributable to variations in the nature of parent rocks. Differences not only exist between plateau and coastal-plain bauxites but also between individual deposits within these groups. 2. The investigated plateau bauxites contain significantly more iron than the coastal-plain deposits, which can be directly linked to the higher iron contents of the Precambrian parent rocks. These vary from ultrahigh-temperature (UHT) gneisses and amphibolites in the Bakhuis Mountains (Granulite Belt in west Suriname) to greenschist-facies metabasalts in the Greenstone Belt in east Suriname (Nassau Mts., Lely Mts., Brownsberg). 3. Chemical and mineralogical variability within a given bauxite sequence is generally small in the plateau bauxites, because weathering profiles developed on relatively homogeneous crystalline parent rocks (metamorphic volcanics, intrusives and sediments). Compositional diversity of the bauxites is mainly caused by the lithological variability of rock units in the Precambrian Shield that were exposed to weathering. 4. In contrast, the internal variability within bauxite sequences of the coastal plain is relatively large because of original lithological inhomogeneity in the stratigraphy of the Tertiary sediments that largely consist of sandy and clayey erosion products of Precambrian rocks, supplied by riverine transport from the country’s interior. Vertical compositional variations in the lateritic bauxite profiles are therefore not only caused by the weathering process but also due to changes in the nature and provenance of the sedimentary material. Sediment provenance is also an important explanation for the existence of compositional variations between the different coastal-plain bauxites. Higher iron contents in the studied profiles of the Successor deposits, compared to those of the nearby Paranam-Onverdacht-Lelydorp bauxite district, illustrate this. 5. The bauxite composition may locally have been influenced by external factors long after formation of the deposit. The coastal-plain bauxites are mostly buried under a packet of

younger sediments with thicknesses up to 40 m. Infiltration of SiO2, derived from these overlying sediments, re-silicified the top parts of the deposits. The origin of sulphur-bearing,

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 168 Synopsis | 169

marcasite-rich bauxite in the Coermotibo coastal-plain deposit (Moengo area) is probably associated with the presence of swamps. 6. Trace-element signatures of the deposits appeared to be an important indication for the nature of the parent rock and the chemical processes that modified the profiles during weathering. For example, the Successor deposits probably had distinct sedimentary precursors despite their geographical proximity, which can be explained by changes in the provenance of the terrigenous sediments in a dynamically evolving river landscape. Similar effects also apply to deposits elsewhere in the coastal plain. Changes in trace-element concentrations within weathering profiles are not only caused by leaching and accumulation, but also reflect original stratigraphic heterogeneity, especially in the bauxites of the coastal plain. Trace-element mobility was largely determined by the host mineral and their stability during weathering. The most resistant population contains the majority of relatively immobile elements (HFSE and REE), and has been preserved in the deposits as heavy minerals. Accumulations of zircon and other heavy minerals are responsible for locally strong enrichments of trace elements. (See Highlights of trace-element results and interpretation key for more details below).

Highlights of trace-element results and interpretation key The trace-element data, obtained by the LA-ICPMS analytical technique, provided compelling evidence for chemical heterogeneity of the bauxites. Some illustrative examples are shown in Figure 6.3, where measured concentrations, normalized to average concentrations in upper continental crust (UCC), are given for a comprehensive set of trace elements, ordered according to increasing mass. Downward pointing troughs denote elements that are susceptible to leaching, whereas peaks represent elements that are poorly soluble in percolating water and are predominantly hosted in weathering-resistant minerals. The strongly parallel trends seen in the plateau bauxites mark an in-situ origin on a single homogeneous rock type. The different concentration levels reflect different degrees of residual enrichment upon the removal of easily leachable components such as silica. The remaining shape of the parallel trends is largely inherited from the original parent rock. It appears that the parent rocks are distinct for the three profiles of plateau bauxites shown. The coastal-plain bauxites (Successor deposits) display some additional features. Their patterns also point to precursor heterogeneity, considering the rare earth elements (stretch between Ce and Lu) and unequal Zr-Nb and Hf-Ta relationships. Non-parallel trends for samples within a deposit are due to internal stratigraphic heterogeneity or leaching/accumulation effects. Narrowing of the bundles between Dy and Hf is probably due to a prominent share of zircon among potential heavy mineral hosts of immobile trace elements. These interpretations of geochemical signatures are not exhaustive (see 6 Chapters 4 and 5) but serve to emphasize their effectiveness for studies of Suriname’s bauxites.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 169 170 | Chapter 6

Figure 6.3 | Upper-continental-crust (UCC) normalized contents of trace elements in selected weathering

profiles of coastal-plain and plateau bauxites. Samples indicated in red have < 5 wt% SiO2, while those in blue

contain > 5 wt% SiO2. Except for the group of easily leached elements (e.g., Zn, Sr, Ba), the trends largely reflect the precursor rock on which the bauxite developed. Note the existence of differences between the coastal-plain and plateau bauxites, as well as within these groups. Deviations from parallelism within a deposit are due to internal stratigraphic heterogeneity or mobility and weathering-induced redistribution.

7. Heavy minerals in the plateau bauxites were derived from the crystalline parent rock, whereas those in the coastal-plain bauxites originated from Precambrian source areas of the sediments, from where they were supplied by riverine transport. In exceptional cases, heavy minerals were derived from underlying basement rock. Regional as well as local variations in heavy mineral populations of the coastal-plain bauxites can be attributed to heterogeneity in provenance areas of the sediments and changing routes of supply. 8. Apart from lithological and mineralogical differences, the geochemical characteristics, aluminium contents and reserves of the bauxites were also influenced by physical properties of the parent rock, local hydrological conditions and drainage. These factors are also responsible for texture differences between the bauxites. 9. This research has demonstrated that trace-element data provide important information on the genetic history of the bauxites of Suriname, both for identifying parent rocks and for unravelling weathering processes. Apart from providing new clues to the diversity of Suriname’s bauxites, trace-element data may also be of great value for more efficient exploration and exploitation strategies in the future.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 170 Nederlandse samenvatting | 171

Nederlandse samenvatting

Dit proefschrift is gewijd aan de bauxieten van Suriname. In het kader van de Surinaamse geologie werden de relaties tussen verschillende bauxiethoudende verweringsprofielen en de lithologie van hun bijbehorende moedergesteente onderzocht. Petrologische en geochemische methodieken werden toegepast om de belangrijkste factoren te onderzoeken die de variabiliteit van de bauxietafzettingen, de geologische eigenschappen en elementfractionatie tijdens de bauxietvorming hebben bepaald. De Surinaamse bauxietafzettingen worden van oudsher verdeeld in de plateau- of hoogland- bauxieten en de kustvlakte- of laagland-bauxieten. Deze onderverdeling is voornamelijk gebaseerd op hun topografische en geografische positie. De eerste groep ligt op de relatief hoge plateaus in het binnenland van Suriname, terwijl de tweede groep in de laaglanden van de kustvlakte voorkomt, waar de afzettingen partieel bedolven zijn onder een dik pakket jongere sedimenten. Suriname was in de vijftiger jaren van de vorige eeuw één van ‘s werelds meest toonaangevende producenten van bauxiet. In de jaren zestig werden er ook aluinaarde en aluminium geproduceerd. De bauxietindustrie is decennia lang de pijler van de Surinaamse economie geweest. De meest toegankelijke bauxietafzettingen in de kustvlakte zijn reeds uitgemijnd, terwijl er nog aanzienlijke reserves zijn op de plateaus van het binnenland. Deze plateau-bauxietafzettingen zijn, ondanks het vele exploratie-onderzoek dat in het verleden werd uitgevoerd, nog niet in productie gebracht (Hoofdstukken 1 en 2). De vorming van de Surinaamse bauxieten begon in het Laat Krijt en is waarschijnlijk nog steeds gaande. Er zijn duidelijke voorwaarden waaraan voldaan moet worden voordat bauxietvorming kan geschieden: een langdurige blootstelling van het moedergesteente aan een warm en vochtig klimaat en een tektonisch stabiele omgeving waarbij er intensieve chemische verwering plaatsvindt. Deze omstandigheden golden gedurende het Laat Krijt – Vroeg Tertiair, toen er wereldwijd grote hoeveelheden bauxiet gevormd werden (Hoofdstuk 1). De bauxietafzettingen in Suriname bieden een unieke kans om de invloed van het moedergesteente op de vormingsprocessen en de eigenschappen van de resulterende bauxiet te bestuderen. Langs de noordelijke rand van het Guiana Schild is een grote verscheidenheid aan moedergesteenten aanwezig: de kustvlakte-bauxieten zijn gevormd op Cenozoïsche sedimenten en de plateau-bauxieten op een uiteenlopende verzameling van Proterozoïsche metamorfe en stollingsgesteenten. De resultaten van dit onderzoek hebben opvallende contrasten tussen de beide bauxietgroepen aan het licht gebracht, maar er zijn ook specifieke overeenkomsten: – Alle bestudeerde afzettingen behoren tot de lateritische bauxieten en worden gekenmerkt door een gemiddelde tot zeer hoge graad van lateritisatie. De afzettingen vertonen in het algemeen de klassieke opeenvolging van een lateritisch verweringsprofiel met de volgende lithologie van de bodem tot de top: een vers of sterk verweerd moedergesteente dat naar boven toe overgaat in een kleirijke saproliet, gevolgd door een aluminiumrijke bauxietlaag die bedekt wordt door een ijzerrijke toplaag. Deze opeenvolging werd vaker in de plateau- bauxieten waargenomen dan in de kustvlakte-bauxieten, waarin onregelmatigheden

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 171 172 | Nederlandse samenvatting

veelvuldig voorkomen, veelal als gevolg van heterogeniteit in de stratigrafie van het moedergesteente. – Gibbsiet is het meest voorkomende aluminiumrijke mineraal in de bauxietlaag, terwijl kaoliniet overheerst in de saproliet. Hematiet en goethiet zijn de belangrijkste mineralen in de ijzerrijke toplaag van de profielen. Kleine hoeveelheden boehmiet zijn plaatselijk aanwezig in de plateau-bauxieten, terwijl het in de kustvlakte-bauxieten alleen sporadisch voorkomt. Dit onderscheid kan vermoedelijk worden toegeschreven aan verschillen in klimatologische en/of afwateringscondities. – De concentraties van de hoofdelementen zijn een rechtstreekse weerspiegeling van de relatieve hoeveelheden van de belangrijkste mineralen die in de verweringsprofielen

gevormd zijn. De hoogste SiO2-concentratie werd over het algemeen waargenomen in

de onderste delen van de profielen (de saproliet), de maximale Al2O3-hoeveelheid in de

bauxietlaag en de grootste hoeveelheid Fe2O3 in de toplaag. De waargenomen verdeling van de hoofdelementen zijn consistent met de verweringsprocessen die lateritische bauxieten produceren: uitloging van onstabiele elementen in het moedergesteente, relatieve accumulatie van slecht oplosbare of immobiele componenten, en relatieve aanrijking van ijzer die gereguleerd wordt door fluctuerende redox-omstandigheden en grondwaterstanden. – De chemische effecten van immobiel versus mobiel gedrag van de hoofd- en spoorelementen tijdens de bauxietvorming kunnen als volgt worden geïnterpreteerd: sterke uitloging van Si, K, Na, Mg en Ca, alsmede Rb, Ba, Sr, Zn en in mindere mate Pb, en intensieve relatieve aanrijking van de relatief immobiele elementen Al, Ti, Zr, Nb, Hf, Ta en Th. De aanrijking van deze hoge-veldsterkte-elementen (HFSE) en de zware zeldzame-aardmetalen (HREE) kan direct gekoppeld worden aan de verweringsresistentie van zirkoon en andere zware mineralen.

Vanwege hun diagnostische betekenis voor bauxietgenese is in dit onderzoek speciale aandacht besteed aan spoorelementen. Voor bauxiet zijn gebruikelijke technieken voor chemische analyse van gesteenten in opgeloste vorm vaak problematisch vanwege de slechte oplosbaarheid van zirkoon en andere resistente zware mineralen die rijk zijn aan spoorelementen. Om deze reden is een alternatieve methode ontwikkeld om bauxietmonsters nauwkeurig te analyseren, namelijk met behulp van laser ablatie –inductief gekoppeld plasma- massaspectrometrie (LA-ICP-MS) op glasparels van lithiumboraat. De toepassing van deze techniek op de bauxietmonsters leverde uitstekende resulaten op, gevalideerd door metingen aan internationale standaard-referentiematerialen voor bauxiet. Verschillen tussen de gemeten waarden en de referentiewaarden waren minder dan 20%, met een externe reproduceerbaarheid van minder dan 20% relatieve standaardafwijking voor vrijwel alle essentiële spoorelementen. Deze analytische methode is gedetailleerd beschreven in Hoofdstuk 3. Op basis van een nieuwe geochemische inventaris, de petrologische bevindingen en veldgegevens konden in dit proefschrift een aantal belangrijke conclusies worden getrokken (Hoofdstukken 4 en 5): 1. De bauxietsequenties van Suriname vertonen aanzienlijke verschillen in opbouw en samenstelling, die voor een belangrijk deel kunnen worden toegeschreven aan grote variaties in de aard van het moedergesteente. De verschillen bestaan niet alleen tussen de

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 172 Nederlandse samenvatting | 173

plateau- en kustvlakte-bauxieten, maar ook tussen de verschillende voorkomens binnen deze groepen. 2. De bestudeerde plateau-bauxieten bevatten significant meer ijzer dan de kustvlakte- bauxieten, wat rechtstreeks kan worden toegeschreven aan het hogere ijzergehalte van de Precambrische moedergesteenten. Deze variëren van ultrahoge temperatuur (UHT) gneissen en amfibolieten in het Bakhuisgebergte (Granulite Belt in west Suriname) tot groenschist-facies metabasalten in de Greenstone Belt in oost Suriname (Nassaugebergte, Lelygebergte en Brownsberg). 3. De chemische en mineralogische variabiliteit binnen een bauxietsequentie is doorgaans betrekkelijk gering in de plateau-bauxieten doordat verweringsprofielen ontwikkeld zijn op relatief homogene kristallijne moedergesteenten (gemetamorfoseerde vulkanieten, intrusiva en sedimenten). De samenstellingsdiversiteit van de bauxieten wordt vooral veroorzaakt door de lithologische variatie van de gesteenten in het Precambrische Schild die aan verwering zijn blootgesteld. 4. Daarentegen is de interne variabiliteit binnen bauxietsequenties van de kustvlakte relatief groot vanwege oorspronkelijke lithologische inhomogeniteit in de stratigrafie van de Tertiaire sedimenten. Deze sedimenten bestaan grotendeels uit zandige en kleiïge afbraakproducten van Precambrische gesteenten, die vanuit het binnenland door rivieren zijn aangevoerd. Vertikale samenstellingvariaties in de bauxietprofielen zijn daarom niet uitsluitend toe te schrijven aan het verweringsproces maar ook aan wisselingen in de aard en herkomst van het sedimentaire materiaal. De herkomst van de sedimenten is tevens een belangrijke verklaring voor samenstellingsverschillen tussen de afzonderlijke bauxietafzettingen in de kustvlakte. Zo bevatten de bestudeerde bauxietprofielen van de Successor-afzettingen beduidend meer ijzer dan de afzettingen in het nabijgelegen Paranam-Onverdacht-Lelydorp bauxietdistrict. 5. De bauxietsamenstelling kan plaatselijk, lang na de vorming, ook beïnvloed zijn door externe factoren. De onder een pakket sedimenten (op sommige locaties tot wel 40 m dik) begraven

kustvlakte-bauxieten vertonen door infiltratie van SiO2, afkomstig van deze bovenliggende sedimenten, resilicatie in de top van de bauxiet. Het ontstaan van zwavelhoudende marcasiet-rijke bauxiet in de kustvlakte-afzetting van Coermotibo (Moengo-gebied) kan worden geassocieerd met de aanwezigheid van moerassen. 6. De spoorelementsignaturen van de afzettingen zijn een belangrijke indicatie voor de aard van het moedergesteente en de chemische processen die zich in verweringsprofielen hebben afgespeeld. Zo blijken de drie Successor-afzettingen, ondanks hun onderlinge geografische nabijheid, verschillende sedimentaire voorlopers te hebben gehad, wat verklaard kan worden door wisselingen in de herkomst van het terrigene sediment in een dynamisch evoluerend rivierenlandschap. Soortgelijke effecten hebben zich elders in de kustvlakte ook voorgedaan. Concentratieveranderingen van spoorelementen binnen vertikale verweringsprofielen zijn niet alleen veroorzaakt door uitloging en accumulatie, maar zijn, vooral in de kustvlakte-bauxieten, ook het gevolg van oorspronkelijke stratigrafische heterogeniteit. De mobiliteit van spoorelementen werd vooral bepaald door hun gastheermineralen en de stabiliteit daarvan tijdens het verweringsproces. De meest resistente populatie bevat het merendeel van de relatief immobiele elementen (HFSE en HREE) en is in de afzettingen bewaard gebleven als zware mineralen. Accumulaties van

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 173 174 | Nederlandse samenvatting

zirkoon en andere zware mineralen hebben plaatselijk gezorgd voor sterke aanrijkingen van spoorelementen. 7. De zware mineralen in de plateau-bauxieten zijn uitsluitend afkomstig van het onderliggende kristallijne moedergesteente, terwijl die van de kustvlakte-bauxieten door riviertransport uit de brongebieden van de terrigene sedimenten in het Precambrische achterland zijn aangevoerd of, in uitzonderlijke gevallen, afkomstig zijn van onderliggende kristallijne gesteenten. Zowel regionale als lokale variaties in de populaties zware mineralen van de kustvlakte-bauxieten kunnen toegeschreven worden aan heterogeniteit van het herkomstgebied van het sediment en veranderende aanvoerroutes. 8. Afgezien van de lithologieverschillen en mineraalinhoud, worden de geochemische kenmerken, aluminiumgehaltes en reserves van de plateau-bauxieten beïnvloed door fysieke eigenschappen van het moedergesteente, lokale hydrologische omstandigheden en drainage. Deze factoren zijn ook verantwoordelijk voor textuurverschillen tussen de bauxieten. 9. Dit onderzoek heeft aangetoond dat spoorelementdata belangrijke informatie geven over de ontstaansgeschiedenis van de Surinaamse bauxieten, zowel voor het identificeren van moedergesteenten als voor het ontrafelen van de verweringsprocessen. Naast hun betekenis voor nieuwe inzichten in de bauxietgenese in Suriname kunnen spoorelementdata van grote waarde zijn voor efficiëntere exploratie- en exploitatiestrategieën in de toekomst.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 174 Acknowledgements | 175

Acknowledgements

This milestone and life-changing experience would not have been possible without the contribution of several people. I would like to express my sincere gratitude to dr. Manfred van Bergen, prof. dr. Rinus Wortel and prof. dr. Theo Wong for their guidance and insightful comments during this research. I would also like to thank the members of the thesis assessment committee for their valuable input. This research would not have been possible without the funding of the Suriname Environmental and Mining Foundation (SEMIF) and I would like to extend my special appreciation to Nathalie Pahalwankhan-Emanuels BSc., drs. Glenn Gemerts, dr. Shanti Venetiaan and Sergio Akiemboto. I am particularly grateful for the assistance given by the technical staff at Utrecht University and the Free University of Amsterdam, especially Otto Stiekema (for manufacturing the finest thin sections), Tilly Bouten (for her instructions on using the electron microprobe), ing. Helen de Waard (for all the chemical analyses in Utrecht), Anita van Leeuwen-Tolboom (for her assistance with XRD analyses), dr. Pieter Vroon and his assistants (for XRF analyses at the Free University of Amsterdam). I would like to thank the Bauxite Institute Suriname (drs. Rita Vaseur-Madhoeban) and Suralco L.L.C. (Aroen Gangaram-Panday MSc.) for providing supplementary data. My special thanks are extended to prof. dr. Jeannot Trampert, drs. Jan-Willem de Blok and drs. Franca Geerdes of Utrecht University for their support during a rough patch of my project. My discussions with dr. Emond de Roever and the late dr. Harold Pollack were also very valuable. I want to show my appreciation to my colleagues at the Anton de Kom University of Suriname and Utrecht University for their moral support, and my former students and my former students Kisoensingh and Taroenoredjo for their help during fieldwork. I most certainly want to thank my family, especially my sisters Fahrida and Suleta, my grandmother Alice, my aunts Hermien, Loes, and my cousins Jennifer, Nel and Mechthild for their spiritual support. And most importantly, I want to express my gratitude to my parents drs. Eduard Monsels and dr. Lilian Monsels-Thompson for their love, support and encouragement. I also want to thank them for giving me the opportunity to study abroad, which opened many doors for me. Last, but certainly not least, I want to dedicate this thesis to my daughter Kyarah, whose bubbly personality and my faith in the Lord guided me to the finish line.

Thank you!

Dewany Alice Monsels

Paramaribo, Suriname January 2018

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 175 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 176 About the author | 177

About the Author

Dewany Alice Monsels was born in Paramaribo, Suriname in 1979. She obtained her BSc degree in Geology at the Anton de Kom University of Suriname in 2004. The title of her thesis was: New data related to the origin of the Coermotibo deep-seated bauxite deposit, district Marowijne, Suriname. In 2005, she started an MSc study in Earth Materials at Utrecht University, where she graduated in 2007 upon finishing her research project entitled: The evolution, processes and timing of metamorphism in the Variscan rocks of the Pyrenees, Spain. After returning to Suriname she worked as an exploration geologist at State Oil Company. Since 2008 she is employed as a full-time lecturer at the Faculty of Technological Sciences of the Anton de Kom University of Suriname, where she is responsible for courses in petrology.

516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 177 516008-L-bw-monsels Processed on: 21-12-2017 PDF page: 178