Inorganic modification of Palabora vermiculite

By Dissertation submitted in partial fulfilment of the requirements for the degree of In the Department of Chemistry, Faculty of Natural and Agricultural Sciences, University of Pretoria Pretoria February 2011 © University of Pretoria Student: Hermínio Francisco Muiambo Supervisor: Prof. Walter W. Focke Department: Chemistry University: University of Pretoria Degree: Master of Science Conventional processes for the manufacture of exfoliated natural vermiculite employ temperatures exceeding 800 °C and the onset temperature is above 450 °C. In many applications, notably fire retardants, it is desirable that exfoliation of vermiculite should occur at temperatures in the range of 200 °C to 350 °C. For this purpose, South African Palabora vermiculite was modified by ion exchange with ammonium and selected alkali metal and alkaline earth metal ions. These experiments were performed at room temperature using an orbital shaker for 15 days. Another set of experiments was performed by immersing vermiculite in saturated sodium chloride solution for periods up to six months. At the end of every month, sodium4exchanged vermiculite samples were taken for analysis. The thermal expansion and degradation of modified vermiculites were studied using thermo4mechanical analysis (TMA) and thermogravimetric analysis (TGA) respectively. Interlayer composition was studied using cation exchange with ammonium acetate. The leached cations were quantitatively determined using inductively coupled plasma optical emission spectroscopy (ICP4OES). Scanning electron microscopy (SEM) was applied to study the i samples’ morphology. X4ray diffraction (XRD), infrared (FT4IR) and Raman spectroscopy confirmed that the basic vermiculite crystal structure was left unchanged during inorganic modification in both sets of experiments. It was confirmed that the blowing agent in the neat vermiculite was the interlayer water. In the exchanged vermiculite with ammonium ions, it was found that ammonia and water were the blowing agents and they were simultaneously released during thermal degradation. However, ammonium4 vermiculite did not show a relatively better thermal expansion ratio, despite the presence of both ammonia and water in the interlayer. In the increasing ionic potential series: Na4vermiculite, Ba4vermiculite, Ca4vermiculite and Mg4vermiculite, it was found that the onset expansion temperature also increased dramatically. Exchange with relatively low ionic potential species, namely sodium, potassium and ammonium, lowered the exfoliation onset temperature of vermiculite to below 300 °C. ! : Vermiculite; Ion exchange; Thermal analysis. ii Exfoliation; Interstratification; "# Discente: Hermínio Francisco Muiambo Supervisor: Prof. Walter W. Focke Departmento: Química Universidade: Universidade de Pretória Grau académico: Mestrado $ % &' & $($) )* Processos convencionais de produção da vermiculita natural exfoliada empregam temperaturas que excedem 800 °C e a temperatura inicial de expansão está acima de 450 °C. Em muitas aplicações, particularmente em áreas onde se deseja extinção de fogo, é desejável que a exfoliação da vermiculite ocorra na região entre 200 °C e 350 °C. Assim, a vermiculita proveniente de Palabora, África do Sul, foi quimicamente modificada com cloretos de amónio, de metais alcalinos e de metais alcalino4terrosos previamente seleccionados. As experiências foram realizadas à temperatura ambiente sob agitação constante por 15 dias. Outro desenho experimental foi executado por imersão da vermiculita em solução saturada de cloreto de sódio durante 6 meses. Mensalmente foram retiradas alíquotas de vermiculita modificada para a devida análise. A expansão e degradação térmicas das amostras da vermiculita modificada foram estudadas usando análises termo4mecânica (TMA) e termogravimétrica (TGA). O método de extracção de catiões usando acetato de amónio foi aplicado em todas as amostras e os extractos foram analisados usando plasma inductivamente acoplado ligado ao espectroscópio óptico de emissão (ICP4OES). A morfologia das amostras foi comparada usando iii microscopia electronica (SEM). Difracção por raios4X (XRD), espectroscopia de infravermelho (FT4IR) e Raman, confirmaram a manutenção da estrutura cristalina da vermiculita durante a reacção de troca iónica em ambas experiências. Na vermiculita nativa foi confirmada a libertação de água antes e durante a exfoliação. Na amostra de vermiculita4NH4 foi detectado amoníaco, para além da água, e ambos foram libertos simultaneamente durante a degradação térmica. Entretanto, vermiculita4NH4 não exibe uma relativamente melhor expansão térmica apesar da presença de água e amónia entre as camadas da vermiculita. Na seguinte série crescente the potencial iónico: vermiculita4Na, vermiculita4Ba, vermiculita4Ca e vermiculita4Mg, a temperatura inicial de exfoliação também é proporcionalmente crescente. Trocas catiónicas da vermiculita natural pelos seguintes iões de baixo potencial iónico: sódio, potássio e amónio resultaram na redução da temperatura inicial de exfoliação para temperaturas abaixo de 300 °C. + , : Vermiculita; Troca iónica; Exfoliação; Interstratificação; Análise térmica. iv -)- &. To my wife Inês and daughters Keila, Flora and Inês. I highly appreciated their love, motivation and unconditional support, especially during those long periods of absence. v .&/0)-() ). I thank almighty God for his blessings and the inspiration that He has been giving me gracefully and in abundance. My profound gratitude goes to Prof. Walter W. Focke for his time, guidance, most valuable contributions, patience and encouragement during the execution of this work. I am particularly indebted to Maria Atanasova for her support and friendship, especially her inputs during the discussion of the XRD and SEM results and other related matters. Particular appreciation also goes to Isbe Van der Westhuizen for the TMA experimental design and analysis. My sincerest thanks to Chris van der Merwe, Alan Botha and Lowrens Tiedt for their help and other valuable inputs during the electron microscopy work. I would also like to express gratitude to my colleagues at the Institute of Applied Materials (IAM, UP): Dan Molefe, Kolela Ilunga, Lumbi Moyo, Nontete Nhlapo, Washington Mhike, Shepherd Tichapondwa and Mthokozisi Sibanda. Thank you for your friendship and companionship during the completion of this study. I extend my acknowledgements to my colleagues at the Chemistry Department of University Eduardo Mondlane (especially to Carvalho Madivate, Arão Manhique and Pedro Massinga Jr.) for their support and encouragement to further my studies. Also, special recognition goes to Suzette Seymore for her valuable administrative support. Financial support for this research from the Institutional Research Development Programme (IRDP), the South Africa/Mozambique collaboration programme of the National Research Foundation (NRF) and the Fundo vi Nacional de Investigação (FNI) is gratefully acknowledged, as well as the Canon Collins Trust. And, finally, I am grateful to my parents for the opportunities they have provided for my personal growth. vii -) 0 &. I, the undersigned, declare that the dissertation, which I hereby submit for the degree MSc at the University of Pretoria, is my own work and has not previously been submitted by me for the degree at this or any other tertiary institution. …………………………………… Hermínio Francisco Muiambo viii 0) & &. ). 1111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 2 ' - * 111111111111111111111111111111111111111111111111111111111111111111111111111111111111 11111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 3 ! 111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 - 11111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 0 111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111114 0 1111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111114 0 , ( 1111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111114 , 5 111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 4 6 111111111111111111111111111111111111111111111111111111111111111111111111111111 6 1.1 Background.......................................................................................... 2 1.2 Aims and objectives............................................................................. 4 1.3 Dissertation outline ............................................................................ 5 , 5 7 0 111111111111111111111111111111111111111111111111111111111111111111111 8 2.1 Clay minerals ...................................................................................... 6 2.2 Composition of trioctahedral vermiculite........................................... 7 2.3 Localisation and geological origin of Palabora vermiculite............... 8 2.4 Physical properties of Palabora vermiculite .................................... 12 2.5 Ion exchange in vermiculite.............................................................. 12 2.6 Kinetics and thermodynamics of ion exchange in vermiculite........ 16 2.7 Exfoliation of vermiculite.................................................................. 17 2.8 Interstratification in vermiculite...................................................... 19 2.9 Potential applications of raw vermiculite ........................................ 25 2.10 Potential applications of chemically modified vermiculite.............. 25 2.11 Potential applications of exfoliated vermiculite .............................. 26 2.12 Commercial value of vermiculite...................................................... 27 ix 2.13 Characterisation techniques............................................................. 29 2.13.1 X4ray Fluorescence (XRF) ............................................................. 29 2.13.2 Inductively Coupled Plasma4Optical Emission Spectroscopy (ICP4OES) and ammonium acetate method ................................. 30 2.13.3 Thermogravimetry (TG) ................................................................ 31 2.13.4 Thermomechanical Analysis (TMA) ............................................. 33 2.13.5 X4ray Diffraction (XRD) ................................................................ 33 2.13.6 Scanning Electron Microscopy (SEM) .......................................... 35 2.13.7 Other characterisation techniques ............................................... 36 , 5 9 )45 111111111111111111111111111111111111111111111111111111111111111111111111111 9: 3.1 Starting materials and suppliers ..................................................... 37 3.2 Ion exchange of vermiculite .............................................................. 37 3.3 Characterisation of chemically modified vermiculites .................... 39 3.3.1 X4ray Fluorescence (XRF) ............................................................. 39 3.3.2 Interlayer composition .................................................................. 39 3.3.3 Particle size.................................................................................... 40 3.3.4 Thermogravimetry (TG) ................................................................ 40 3.3.5 Thermomechanical Analysis (TMA) ............................................. 40 3.3.6 X4Ray Diffraction (XRD) ............................................................... 41 3.3.7 Scanning Electron Microscopy (SEM) .......................................... 42 3.3.8 Fourier Transform–Infrared (FT4IR) Spectroscopy ..................... 42 , 5 4.1 ; 1111111111111111111111111111111111111111111111111111111111 ;; X4ray Fluorescence (XRF) ................................................................. 44 4.1.1 Neat and long4term Na4exchanged vermiculites ......................... 44 4.1.2 Ammonium, alkali metal and alkaline earth metal ion exchanged vermiculites ................................................................. 46 4.2 Interlayer composition and extractable cations............................... 48 4.3 Thermogravimetry (TG).................................................................... 49 4.4 Thermo4Mechanical Analysis (TMA)................................................ 54 x 4.5 X4Ray Diffraction (XRD) ................................................................... 59 4.6 Scanning Electron Microscopy (SEM) .............................................. 64 4.7 Particle size distribution................................................................... 65 4.8 Fourier Transform–Infrared (FT4IR) and Raman Spectroscopy..... 67 , 5 < 111111111111111111111111111111111111111111111111111111111111111111111111111111 := , 5 8 11111111111111111111111111111111111111111111111111111111111111111111111111111111 :7 55 11111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111111 >: xi 0 & ($ ) 61 Simplified geological map of the Palabora igneous complex. (1) Open pit which produces phosphate; (2) Vermiculite open pit; (3) Open pit which produces basically copper and magnetite......... 11 71 Worldwide production of vermiculite, from 1994 to 2009 .......... 28 91 X4ray fluorescence after photoelectric effect with respective production of Kα and Kβ X4rays ................................................... 29 ;1 Basic TGA instrument................................................................. 32 <1 X4ray diffraction by a crystal according to Bragg’s law ............. 34 81 Determination of vermiculite content from d4spacing ............... 35 :1 SEM image of neat superfine vermiculite .................................. 44 >1 Change in interlayer composition of vermiculite during long4term treatment with NaCl .................................................. 48 ?1 TG plots of micron vermiculite and its Na4exchanged forms .... 50 6=1 Mass loss for neat vermiculite and its Na4exchanged forms ..... 51 661 TG plots of metal ion exchanged vermiculites............................ 52 671 Mass loss of ion exchanged vermiculites .................................... 53 691 TMA expansion of vermiculite and its Na4exchanged forms ..... 55 6;1 Effect of the TMA applied force on the expansion behaviour of vermiculite flakes reacted for 6 months with saturated brine .. 56 6<1 TMA characterisation of the isothermal expansion of long4term Na4exchanged vermiculite, at selected temperatures................ 57 681 TMA expansion of ion exchanged vermiculites .......................... 58 6:1 Evolution of the XRD spectra as a function of brine exposure time............................................................................................... 59 xii 6>1 XRD patterns of blank, alkali metal and alkaline metal treated vermiculites.................................................................................. 62 6?1 XRD patterns of neat vermiculite and blank vermiculite.......... 63 7=1 SEM image of vermiculite flakes reacted for 6 months with saturated brine ............................................................................ 64 761 SEM image of heat4expanded Na4vermiculite............................ 65 771 Particle size distribution of the neat and long4term Na4exchanged vermiculites ......................................................... 66 791 Particle size distribution of vermiculite and its ion exchanged forms............................................................................................. 67 7;1 FT4IR plots of neat vermiculite and all its Na4exchanged forms, obtained after immersion in saturated brine ............................. 68 7<1 FT4IR spectra of neat vermiculite, ammonium, alkali metal and alkaline metal ion exchanged vermiculites ................................ 69 xiii 0 & 0) 61 Physical properties of Palabora vermiculite............................... 12 71 CEC of some clay minerals.......................................................... 14 91 Dehydrated and hydrated ionic radii and respective calculated ionic potentials............................................................................. 15 ;1 Vermiculite WLHS properties..................................................... 24 <1 XRF composition analysis of long4term Na4exchanged samples 45 81 XRF composition analysis of ammonium, alkali metal and alkaline earth metal ion exchanged vermiculite samples ......... 47 :1 Mineralogical composition of blank vermiculite......................... 61 xiv 0 & ) ) Scheme I: General formula of dioctahedral clay minerals ........................ 7 Scheme II: General formula of trioctahedral clay minerals ....................... 7 Scheme III: Simplified representation of the partial conversion of biotite into vermiculite .......................................................................... 9 Scheme IV: Ion exchange reaction involving vermiculite and another electrolyte. ................................................................................ 16 Scheme V: Ion exchange equilibrium constant in vermiculite. ................ 16 Scheme VI: UG° of the ion exchange reaction on vermiculite..................... 16 Scheme VII: Van’t Hoff equation for determination of UH°. ........................ 17 Scheme VIII: US° that accompany the ion exchange on vermiculite............ 17 Scheme IX: Schematic drawing of interstratified vermiculite ................... 21 Scheme X: Planck’s energy quantization law............................................ 30 Scheme XI: Calculation of interlayer CEC from interlayer charge............ 31 Scheme XII: Bragg’s law................................................................................ 34 Scheme XIII: Schematic representation of ion exchange in interstratified vermiculite................................................................................ 38 Scheme XIV: Reaction of Ag+ and Cl4 ions with formation of a white precipitate of silver chloride. ................................................... 39 Scheme XV: Structural formula of neat4vermiculite consistent with the XRF data. .......................................................................................... 46 Scheme XVI: Determination of apparent total metal charge. ...................... 49 Scheme XVII: Thermal decomposition of ammonium ions in vermiculite. ... 53 Scheme XVIII: Dehydroxylation of vermiculite. .............................................. 53 Scheme XIX: Dehydroxylation of vermiculite aid by hydrogen ions. ........... 54 xv (0& @ CEC Cation exchange capacity cmol centimol DSC Differential scanning calorimeter DTG Derivative thermogravimetry FT4IR Fourier transform infrared spectroscopy ICP4OES Inductively coupled plasma optical emission spectrometer IR Infrared spectrometer IUPAC International Union of Pure and Applied Chemistry Ksp Constant of solubility product LOI Loss on ignition pH Potential hydrogen RF Radiofrequency SEM Scanning electron microscopy TGA Thermogravimetric analysis/analyser TMA Thermomechanical analysis/analyser XRD X4ray diffraction XRF X4ray fluorescence WLHS Water layer hydration state 04WLHS Zero water layers hydration state 14WLHS One water layer hydration state 24WLHS Two water layers hydration state xvi ) 6 . &-$ &. Clay minerals are attracting more and more interest in nanoscience and nanotechnology due to the huge effect that they can have on materials at low dosage levels. Applications include nanocomposites, catalysis, adsorbents, sensors, nuclear waste storage, antibacterials, pharmaceuticals, pesticide carriers and others (Grim, 1968; Liu , 2007; Bergaya ., 2006; Annabi4Bergaya ., 2008). The chemical stability and mineralogical behaviour of clay minerals in saline solutions has attracted interest owing to their ability to absorb and immobilise toxic ions derived from radioactive and chemical waste. Most studies have focused on the mineralogical and chemical transformation of bentonite in saline water at elevated temperatures (Herbert and Moog, 1999; Herbert ., 2004; Suzuki ., 2008). Due to the high cation exchange capacity (CEC) and surface area of vermiculites, research has been focused on the improvement of its surface area and porosity, the replaceability of the interlayer cations as well as its hydration/dehydration properties (Barshad 1950; Fripiat , 1992; Beyer and Reichenbach, 1998; de Haro According to some authors (Justo , 1960; Saehr , 2005). ., 1993; Marcos ., 2009), vermiculite4mica interstratification materials show larger expansion at relatively lower temperatures and the thermal effects are also observed at these temperatures. In addition, this modified and interstratified vermiculite is amenable to producing a durable exfoliated material (Frank and Edmond, 2001). In fire barrier applications, is desirable for exfoliation of vermiculite to occur at temperatures in the range of 200 °C to 350 °C. 1 3 616 According to many authors (Grim, 1968; Anthony 2006; Bergaya ., 2004b; Hindman, , 2006), vermiculite was first mentioned by Thomas Webb in 1824 while studying a mineral from Massachusetts (United States of America). Thomas Webb described it as a variety of talc mineral that during thermal expansion looks like small worms having the its name ‘vermiculite’ (from Latin motion, hence to breed worms). Vermiculite’s remarkable exfoliation when subjected to high temperatures attracted the attention of researchers. Available relevant literature on vermiculite dates back to the 1930s. Gruner (1934) and, later on, Hendricks and Jefferson (1938) were the first to clarify the vermiculite and hydrobiotite crystal structures. Early research on vermiculite ion exchange was done by Barshad (1950; 1954a; 1954b). Lopez4Gonzalez and Cano4Ruiz (1957) and Raman and Jackson (1963) focused on the surface morphology of both neat vermiculite and exchanged forms following exposure to alkaline salts and after thermal treatment. Recently, da Fonseca . (2005) concluded that vermiculite exhibits good adsorption capacity for divalent heavy metal ions from aqueous solutions. The application of clay minerals in oil production, petroleum engineering and catalysis is intrinsically related to their swelling behaviour (Bougeard and Smirnov, 2007). The swelling of vermiculite solutions is the result of the insertion of relatively large particles into the interlayer space or solvation of the interlayer cation due to lack of balance of attractive and repulsive forces between adjacent 2:1 layers (Newman, 1987; de Haro . 2005). Previous investigations dealt with the preparation of free4settling dispersions and swollen solutions of vermiculite by impregnating it with a variety of lithium salts (Ou and Yang, 1987; Ou, 1992) and found an increase on solution swelling from two to eight times in volume. 2 Chemical modification of vermiculite has also been done by treating it with organic cations. The majority of studies related to the intercalation of vermiculite were carried out using heat (higher than room temperature), apart from other mechanical stratagems. Alkylammonium vermiculite complexes were prepared at 70 °C in an oven (Serratosa e shaker (Abate (Martynková ., 1970); at 50 °C with a , 2006); and at 70 °C in a centrifuge and dried at 40 °C , 2007). Zhu (2008) stirred the suspensions at 50 °C and dried them at 80 °C. . (2005 and 2006) had attempted cation Nevertheless, da Fonseca exchange and intercalation at room temperature. Actually, da Fonseca . (2005 and 2006) exchanged vermiculite with heavy metals at 25 ± 1 °C and intercalated cyclic aliphatic amines as well as aromatic heterocyclic amines. Thomas and Bohor (1969), Couderc and Douillet (1973) and Justo (1989), in their studies on ion exchanged vermiculites, noticed that Mg4vermiculite retains coordinated water more powerfully at comparable elevated temperatures than other vermiculites. On the other hand, Reichenbach and Beyer (1994) concluded that some ion exchanged vermiculite, without well4defined water layer hydration states (WLHS), are less thermally stable than Mg4vermiculite. The Palabora mine in the Limpopo province of South Africa is a major source of commercial vermiculite. The exfoliated vermiculite is of industrial interest because it is incombustible, lightweight and shows excellent absorption and thermal insulation properties (Wada, 1973b, Schoeman, 1989; Tomanec ., 1997; Obut and Girgin, 2002; Bergaya ., 2006). In this interstratified vermiculite from Palabora, the individual flakes are made up of elementary layers of two types: vermiculite and biotite (MacEwan . 1961, Gast and Klobe, 1971; Newman, 1987). Unlike the potassium ions in biotite, the hydrated interlayer ions (usually magnesium ions) in vermiculite are easily exchanged (Newman, 1987). Justo (1989), Justo 3 . (1993), Ou and Bablouzian (1994) and Marcos . (2009) have shown that, compared with pure vermiculite, vermiculite4mica interstratified materials feature higher expansion ratios, with the thermal exfoliation also commencing at lower temperatures. In some applications, notably intumescent fire barriers, it is desirable that the exfoliation of vermiculite should occur at intermediate temperatures, i.e. in the range 200 °C to 350 °C. Flame4retardant compositions have also been proposed that rely on vermiculite expansion properties (Wada, 1973a). However, vermiculite expansion occurs at temperatures well above the melting or decomposition temperatures of common polymers. A 617 It was the aim of the present study to lower the expansion onset temperature of Palabora vermiculite to temperatures in the range of 200 °C to 350 °C. Since the surface morphology and the properties of vermiculites are closely related to the nature of the interlayer cation (Pérez4Maqueda , 2004), ion exchange of original interlayer cations was performed, through room4temperature treatment of Palabora vermiculite. The first approach consisted of long4term exposure of Palabora vermiculite to saturated brine (NaCl) at room temperature. Ion exchange was also carried out using chloride solutions of ammonium (NH4Cl), sodium (NaCl), potassium (KCl), magnesium (MgCl2), calcium (CaCl2) and barium (BaCl2). 4 619 This dissertation is structured in six chapters. The first chapter describes the history of vermiculite discovery and the first attempts to its characterisation as well as giving a brief summary of the relevant work done so far for the beneficiation of vermiculite. The aims and objectives of current research are also described. Chapter 2 presents an extensive literature review. In Section 2.1, attention is briefly paid to the definition of clay minerals (due to controversy on this topic) and to phyllosilicates, namely dioctahedral and trioctahedral. In Section 2.2, a brief discussion about the structure and composition of vermiculite is provided. The geological origin, composition and physical properties of Palabora vermiculite are described in the following sections. Previous work aimed at modifying the properties of vermiculite is also discussed in this chapter, with the main focus on ion exchange and exfoliation phenomena. Special attention is paid to a discussion of interstratifications, as well as the potential applications of vermiculite. This chapter closes with a description of each characterisation technique applied in this study. Chapter 3 reports the experimental procedures followed and briefly describes the equipment used in each analytical technique. Chapter 4 presents details of all the experimental results achieved in this study and discusses these results. The conclusions are summarised in Chapter 5 and the references are given in Chapter 6. 5 ) 7 0 ) $ ) $ B)@ 716 According to the charge of the interlayer cation, clays and clay minerals are classified into two broad classes: cationic clays and anionic clays. While the cationic clays are widespread in nature, the anionic clays or mixed metal hydroxides are rarely found but are relatively simple and inexpensive to synthesise (Reichle, 1986; Vaccari, 1998). Clay minerals are defined as natural or synthetic phyllosilicate minerals and minerals that impart plasticity to clay and that harden upon drying or firing (Bergaya , 2006). According to Bergaya (2006), the properties of clay minerals are usually associated with those of smectites and are characterised by: • a layered structure with one dimension in the nanometre range • anisotropy of the layers • different types of surface (basal, edge and interlayer) • cation and anion exchange capacity • relatively easy chemical, physical or thermal modification • interlayer swelling in appropriate solvents • plasticity. The structures of phyllosilicates are all based on tetrahedral (T, with four sites) and octahedral (O, with six sites) sheets that may condense anisotropically in a proportion of either 1:1 (TO) or 2:1 (TOT) (Newman, 1987; Vaccari, 1998; Bergaya , 2006). The unit cell of a 1:1 phyllosilicate layer structure is characterised by four tetrahedral sites and six octahedral sites, while a 2:1 layer unit cell is composed of eight tetrahedral sites and six octahedral sites. The oxygen atoms 6 are responsible for connecting the sheets (Newman, 1987; Frank, 2001; Bergaya , 2006; da Fonseca ( , ) ( , 2006). ) ( ) ( )/ General formula of dioctahedral clay minerals C In dioctahedral clay minerals (Scheme I) only four of the six octahedral sites are occupied, while in trioctahedral clay minerals all the sites – either the octahedral (six) or the tetrahedral (four) sites – are occupied, as shown in Scheme II (Newman, 1987; Vaccari, 1998; Bergaya ( , C ) ( ) ( ) , 2006). ( )/ General formula of trioctahedral clay minerals In clay minerals the thickness of the 1:1 layer is about 0.7 nm and that of the 2:1 layer is about 1 nm (Bergaya 717 5 , 2006). , Natural vermiculite consists mainly of macroscopic and microscopic types. Macroscopic vermiculite (large plate4like morphology crystal) is trioctahedral and has a relatively narrow range of CEC, whereas microscopic or clay vermiculite may be either trioctahedral or dioctahedral (Basset, 1963; Brindley and Brown, 1980; Newman, 1987). Vermiculite is a magnesium, iron, aluminium, silicate, hydroxide hydrate mineral. Structural water occurs between the silicate layers, depending on the interlayer cation, and non4structural water also occurs in an amount depending on the porosity and relative humidity (Brindley and Brown, 1980; Newman, 1987; Harben and Roberts, 1990; Frank, 2001; da Fonseca , 2006). In vermiculites, the majority of the octahedral cations are trivalent, such as Al3+ or Fe3+. The layers carry a net negative charge owing to isomorphous 7 substitutions of Si in the tetrahedral sites by trivalent ions such as Al3+ or Fe3+. However, this is partially compensated for substitution of Mg2+ in the octahedral sites by trivalent ions, e.g. Fe3+. The excess sheet charge is neutralised by alkali metal or alkaline earth interlayer cations, e.g. Mg2+, present in the interlayer space and lying in plane midway between adjacent 2:1 layers. These cations are hydrated and can be exchanged (Newman, 1987; Brindley and Brown, 1980; Pérez4Maqueda . 2003; Bergaya , 2006). According to Farmer (1974), the surface charge in vermiculite is concentrated on oxygens of Al4O4Si linkages and adjacent oxygens of Si4O4Si connection. Vermiculite shows similarities in its crystalline structure expansibility with smectites (Grim, 1968; Basset, 1963; de Haro and , 2005; Tjong, 2006). However, the electrostatic charge of the vermiculite layer is larger and more heterogenic than that of smectites due to the variations in charge density from layer to layer or within the individual layers (Grim, 1968; Tjong, 2006; de Haro Palabora , 2005; Xu vermiculite, like , 2005). other commercial types, also shows considerable variability in both composition and CEC. This is attributed to differences in the composition of the original parent mica and the degree of progress in the chemical changes induced by weathering (Basset, 1963; Frank, 2001). 719 0 The vermiculite used in this work is from Phalaborwa, Limpopo Province, South Africa. The Palabora mine is a major source of such commercial vermiculite. Actually, Palabora vermiculite is not pure vermiculite but rather a mixed4layer vermiculite4mica (Schwellnus, 1938; Basset, 1963; Schoeman, 1989). In fact, commerce vermiculite is a generic name for minerals with characteristics 8 similar to pure vermiculite (Justo, 1989; de Haro ., 2005). Schoeman (1989) proposed the name hydrophlogopite for the mineral considered presently. Palabora vermiculite is a product of the hydration of phlogopite/biotite under the influence of percolating meteoric water at the weathering surface. During the vermiculization process there was progressive leaching of potassium ions, an increase in the total water content, a change in colour and a simultaneous loss in elasticity and transparency on the part of the mica (Basset, 1963; Schoeman, 1989). 3+ + 2+ 3+ + (Mg 3- x Fe 2+ x )[Si 4- y - z Al y Fe z ]O 10 (OH) 2 K y + z → α (Mg 3- x Fe x )[Si 4- y - z Al y Fe z ]O 10 (OH) 2 K y + z 3+ 2+ + (1- α) (Mg 3- x Fe 3+ x )[Si 4- y - z Al y Fe z ]O 10 (OH) 2 Mg ( y + z - x ) / 2 .n H 2 O , C Simplified representation of the partial conversion of biotite into vermiculite (Newman, 1987). Scheme III illustrates the partial transformation of biotite to vermiculite in a decidedly simplified manner. It is assumed that Fe3+ is present in the tetrahedral sheets and that the Fe2+ in the octahedral sheets in vermiculite is completely oxidised during the vermiculisation process. The vermiculization process entails the replacement of the interlayer potassium ions by hydrated cations (usually magnesium and, less frequently, calcium ions) and the oxidation of the octahedral Fe2+ ions (Newman, 1987). Thus, the net negative charge is lower than that of the parental biotite and usually falls in the range 0.6 to 0.9 per O10(OH)2 unit. The decreased charge density facilitates the stripping of K+ ions and their replacement by (most commonly) hydrated Mg2+ ions (Brindley , 1983; Moore and Reynolds, 1989). The co4intercalation of water increases the distance between the 2:1 layers. The hydration energy of the cation (i.e. its ionic potential, which is a function of charge density and ionic radius) attracts water. A monovalent cation has lower ionic potential than an equivalent4sized divalent cation 9 (discussed in detail in Section 2.5). The interlayer distance is therefore dependent on the balance of forces necessary to dehydrate the cation and the forces of attraction between the layers and the cation (Brindley and Brown, 1980; Brindley During the , 1983). transformation from mica to vermiculite, numerous intermediate mixed4layer phases with various compositions and ordering are formed. In general, a more ordered structure results at a higher layer charge and also when the humidity is high enough to give stable water layer hydration states (WLHS) (Brindley 1983). The Palabora Proterozoic igneous complex resulted from an alkaline intrusive activity. The Palabora mine is a zoned deposit with ultramafic rocks (pyroxenite) intruded by alkalic rocks, most of which are syenitic in composition (Frank, 2001). Palabora vermiculite mining and concentration were started during 1946 by Dr Hans Merensky (Palabora Mining Company, 1976; Schoeman, 1989). At the Palabora complex there are three separate open pits (see Fig. 1). The vermiculite open4pit mine (represented by number 2 in Fig. 1) is operated by the Palabora Mining Company and since 1990 has become the world’s largest producer (Evans, 1987, 1993; Hindman, 2006). This vermiculite is mined from a coarse4grained zoned ultramafic body consisting of phlogopite4serpentine rock enveloped by phologopite4diopside rock (Palabora Mining Company, 1976; Evans, 1993). The crude vermiculite is transported by rail from the mine in Phalaborwa to the port of Richards Bay on the Indian Ocean in KwaZulu4Natal province, South Africa. From this port it is shipped in bulk to a diversity of destinations and is also stored in the Netherlands and United Kingdom (Eaves, 2006). 10 61 Simplified geological map of the Palabora igneous complex (Evans, 1993). (1) Open pit which produces phosphate; (2) Vermiculite open pit; (3) Open pit which produces basically copper and magnetite Environmentally, Palabora vermiculite is described as essentially free of asbestiform fibres and crystalline silica (Evans, 1993; Symons, 1999; Eaves, 2006) and, by the same token, as a potential raw material for application in environmentally friendly technologies. 11 71; , 5 5 Like other clay minerals, vermiculite can exhibit different properties according to its composition which is affected by the composition of the parent mica, changes during weathering and ion exchange (Basset, 1963; Frank and Edmond, 2001; Hindman, 2006). 61 Physical properties of Palabora vermiculite (Symons, 1999; Anthony 5 ., 2004a) , Melting point 1 315 °C Sinter temperature 1 260 °C Thermal conductivity 0.062–0.065 w/m°C Colour brownish, golden yellow Hardness 1.5– .5 (Mohs scale) Density 2.2–2.8 (g/cm3) Specific heat 0.2 J/(g—K) Bulk density 50–130 g/L Cleavage d001, perfect, typically relict Tenacity brittle to flexible Palabora vermiculite’s main properties are summarised in Table 1. It shows properties close to those of the parental biotite: odourless, non4corrosive, non4combustible, non4allergenic and harmless if swallowed (Symons, 1999). 71< 4 , Bergaya (2006) have identified and discussed in detail different ways of modifying 2:1 clay minerals, as follows: (i) Adsorption 12 (ii) Ion exchange with inorganic cations and cationic complexes (iii) Ion exchange with organic cations (iv) Binding of inorganic and organic cations (v) Grafting of organic compounds (vi) Reaction with acids (vii) Pillaring by different types of poly(hydroxo metal) cations (viii) Interlamellar or interparticle and interparticle polymerisation (ix) Dehydroxilation and calcination (x) Delamination and reaggregation. Ion exchange generally consists of the stoichiometrical and reversible replacement or swapping of ions between two electrolytes in different media (Grim, 1968, Harland, 1994). Many authors (Walker, 1959; Grim, 1968; Harland, 1994; Tan, 1996) also describe ion exchange, in clay minerals, as a diffusion process of substitution of the ions held electrostatically on the surface of the ion exchanger (solid phase) and a solution phase. According to Grim (1968) and Harland (1994), the ion exchangeability of phyllosilicates is frequently accompanied by a change in the d4spacing and is basically made by: • The exchangeable cations are mostly located on cleavage surfaces and in vermiculite this region contributes approximately 80% of the total CEC. This charge results from isomorphous substitutions within the lattice structure. • The availability of these exchange sites is seriously affected by the particle size, lattice distortions and crystallinity. Broken bands are responsible for up to 20% of total CEC in vermiculite. 13 • Obviously, the availability of these sites during the exchange reaction is highly influenced by the pH of the medium and is of particular significance for layered minerals like vermiculite. The ion exchange phenomenon is quantitatively measured by the CEC, which is defined as the quantity of cations available to be exchanged at a certain pH. It is traditionally expressed in (meq)/100 g of calcinated clay, which is equivalent to cmol(+)/kg, units recommended by IUPAC (Bergaya ., 2006). 71 CEC of some clay minerals (Bergaya ) Vermiculite ., 2006) ' DE6== * 130–210 Montmorillonite 70– 20 Illite 10–40 Kaolinite 3–15 The CEC values of selected clay minerals are presented in Table 2. It appears that the broad intervals given are due to the diverse factors that may influence the clay properties. The main aspects that affect CEC measurements are: pH, temperature, ionic strength, nature of the index cation, nature of interlayer cation(s), particle morphology and size and surface area (Raman and Jackson, 1963; Grim, 1968; Stone and Wild, 1978; Harland, 1994; Bergaya , 2006; Peralta, 2009). Due to its wide use in the literature, in this work the term ‘ion exchange’ is used to designate the exchange of interlayer inorganic ions adsorbed in vermiculite (Ca2+ and Mg2+) by the following cations: ammonium (NH4+), alkali metals (Na+ and K+) and alkaline earth metals (Mg2+, Ca2+ and Ba2+). Some of their relevant properties are summarised in Table 3. 14 Dehydrated and hydrated ionic radii and respective calculated ionic 91 potentials (Grim, 1968) - , ' ' * * 'FE * 0.143 0.537 7.0 0.133 0.380 7.5 + 0.098 0.560 10.2 2+ 0.078 1.080 25.6 2+ 0.106 0.960 18.9 2+ 0.143 0.880 14.0 NH4 K + 5 + Na Mg Ca Ba The ‘ionic potential’ is defined as the ratio of the ion’s electrical charge to its radius. It gives an idea of the strength of the ionic bond that can be formed by the cation and the extent of hydration. Many authors (Wild and Keay, 1964; Gier and Johns, 2000; Badreddine ., 2002a; Ferrage ., 2005; Vidal and Dubacq, 2009) have observed, in their work with clay minerals, that cations with high ionic potential possess high selectivity coefficients. As a result, they are able to create a richer interlayer hydration and consequent better enlargement of the interlayer space. Insightful was the work done by Thomas and Bohor (1969) in their study of the surface area of ion exchanged vermiculite. They stated that “Mg4 vermiculite would retain coordinated water more tenaciously at comparable elevated temperatures than would the other vermiculites studied.” According to them, this is due to three combined effects: (1) high ionic potential; (2) a stronger polarising effect; and (3) high covalent character. These results were later confirmed and augmented by Couderc and Douillet (1973), Stone and Wild (1978) and Justo (1989). 15 718 , 4 , The ion exchange reaction in vermiculite can be systematically represented by Scheme IV (Barrer and Klinowski, 1977; Inoue, 1984; Harland, 1994). Z Z , BC Ion4exchange Z Z ZA BVB Vermiculite) + ZB A.A Solution) ZB AVA Vermiculite) + ZA B. B Solution) reaction involving vermiculite and another electrolyte By applying mass action law to the exchange reaction (Scheme IV) 5 between vermiculite interlayer cations 34 6 and a solution containing 58 7 cations, the equilibrium constant (K) can be determined by Scheme V (Barrer and Klinowski, 1977; Inoue, 1984): K= , BC Z ZA Z Z Z ZB Z Z AVB ;mBS = ƒAB γBA BVA ;mAS = ƒBA γAB Ion exchange equilibrium constant in vermiculite Where AV and BV, and ZA and ZB are the equivalent fractions and respective charges of ions A and B in vermiculite; mBS , mAS represent the molalities in the solution; and ƒ? , ƒA , γ? and γA are the activity coefficients in vermiculite and in solution, respectively. The standard free energy change (UG°) for the reaction demonstrated by Scheme VI can be determined by the expression: ∆G° = − , B C RT lnK ZA ZB UG° of the ion exchange reaction on vermiculite Where R is the universal gas constant and T is the absolute temperature. 16 On the other hand, the standard enthalpy change (UH°) can be estimated by the Van’t Hoff equation (Scheme VII): ∆H° = − ∂(lnK) R ∙ ZA ZB ∂(1/T) B C Van’t Hoff equation for determination of UH° , And the standard entropy change (US°) can be calculated from the following relationship (Scheme VIII): ∆S° = , B 1 (∆H° − ∆G°) T C US° that accompanies the ion exchange on vermiculite The standard free energies (UG°) determined by Wild and Keay (1964) and Inoue (1984) indicate the following affinity of alkali and alkaline metal ions to vermiculite: Na+ < Ba2+ ~ Ca2+ < Mg2+ < K+. According to Wild and Keay (1964) and Inoue (1984), vermiculite selectivity, in the above series, is determined mostly by the entropy increase rather than the enthalpy change since both enthalpy and entropy are positive during ion exchange. For instance, Wild and Keay (1964) found that vermiculite’s preference for divalent cations is due mainly to an increase in the (configurational) entropy during the replacement of monovalent cations by corresponding half4divalent cations. 71: )4 Commercial vermiculite features the desirable property that it expands by more than eight times in volume when heated rapidly to elevated temperatures (Walker, 1961; Wada, 1973a; Wada, 1973b; Justo, 1989; Friedman Tomanec ., 1997; Hindman, 2006). 17 ., 1994; During exfoliation it is desirable to preserve the relevant properties for vermiculite applications. Besides its light weight, sound absorption, heat insulation and attractive appearance, properties such its ion exchangeability, humidity control, high water4holding capacity and elimination of odours are advantageous and should be conserved. Exfoliation of vermiculite can be achieved by means of various processes. The conventional processes for the manufacture of exfoliated vermiculite employ temperatures exceeding 800 °C. The procedure most commonly used is thermal exfoliation. This occurs as a sheet exfoliation process (i.e. normal to the basal cleavage) which is driven by an explosive release of interlayer water as steam (Baumeister and Hahn, 1976; Justo, 1989; Obut and Girgin, 2002; Pérez4Maqueda , 2003; Bergaya , 2006). The interlayer water is said to escape easily and explosively between the layers (pushing them apart), favouring the degree of exfoliation when the particle size is relatively small (Bergaya , 2006). For the application of vermiculite in intumescent fire barriers, exfoliation of vermiculite has been made to occur at relatively lower temperatures than the usual by impregnating it with a compound of urea and thiourea consisting chemical group or ammonium dihydrogen phosphate. The resulting composition is heated at temperatures (160 °C – 300 °C) not lower than the decomposition point of the urea or thiourea compound (Wada, 1973a; Wada, 1973b). Langer and Marlor (1981) also describe the exfoliation of vermiculite via the addition of inorganic ammonium salts, such as ammonium dihydrogen phosphate, ammonium hydroxide and ammonium carbonate, as well as urea. When this inorganically treated vermiculite is heated to 400 °C, it expands by more than 200%. 18 Exfoliation of vermiculite also occurs by sudden decomposition on heating to 100 °C after pre4treatment and the penetration of large quantities of H2O2 into the interlayer space at 4 °C (Baumeister and Hahn, 1976). Early research also focused on vermiculite expansion through irradiation of electromagnetic waves in a specific range of frequencies. This process was carried out for ore vermiculite and previously intercalated vermiculite (Wada, 1973a). 71> Interstratified minerals are compositions in which individual crystals are made up of elementary layers of two or more different types (MacEwan 1961; Moore and Reynolds, 1989; Bergaya . , 2006). These minerals are structurally and chemically more complex than minerals with a one4unit structure (Newman, 1987). Newman (1987) and Bergaya et al. (2006) classify interstratified minerals according to the stacking order or periodicity along the • direction as: Ordered or regular mixed4layer structures – if the layers arrange in an alternate order, like ...ABABAB...; ...AABAABAAB...; ...AAABAAAB...; etc. • Disordered or irregular mixed4layer structures – when the stacking order is random, like ...ABAABA...; ...ABBAABAB...; ...AABABBABBBBA...; etc. The literature reveals two types of interstratification in clay minerals. One form is due to the types of mineral forming the mixed layers (e.g. vermiculite/mica, vermiculite/chlorite, mica/montmorillonite) and the other is due to the difference in terms of the water layer hydration states (WLHS) of the interlayer cations (e.g. 14WLHS mixed with 24WLHS). 19 Apparently, mixed mineral layers occur naturally, while the WLHS interstratifications are usually observed during the dehydration and/or rehydration of clays (Gruner, 1934; Brindley Beyer, 1997; Frank and Edmond, 2001; Marcos Justo . (1993) and Marcos , 1983; Reichenbach and .,2009). . (2009) have shown that vermiculite4 mica interstratification materials have the highest expansion and the thermal effects occur at lower temperatures compared with pure vermiculite. Hydrobiotite is a regular 1:1 interstratification of vermiculite and biotite layers (Brindley , 1983; Newman, 1987). By the same token, hydrophlogopite is described as a regularly interstratified 1:1 mix of phlogopite/vermiculite. It has basal reflections with d = 2.43 nm, 1.223 nm, 0.827 nm, 0.490 nm and 0.349 nm (Gruner, 1934; Palabora Mining Company, 1976; Brindley , 1983; Frank and Edmond, 2001). However, the name ‘hydrobiotite’ is not internationally recognised as a valid mineral name, although it has long been used in the literature and is associated with interstratified biotite/vermiculite minerals found in commercial vermiculite deposits (Boss, 1967; Van der Marel and Beutelspacher, 1976; Brindley and Brown, 1980; Brindley , 1983; Hudson , 1999; Zhu , 2008). Mixed4layer clay minerals with any appreciable degree of randomness show a seemingly irrational series of reflections that make their identification and interpretation quite a complex exercise. However, computer programs are available for predicting the diffraction patterns of one4dimensional mixed4layer clays (Reynolds, 1985). The presence of mixed4layer vermiculite4mica (Scheme IX) in ore vermiculite makes it more valuable, taking into account the following facts: 20 , GC Schematic drawing of interstratified vermiculite (adapted from Zhu, 2008) • The interlayer spacing of hydrobiotite (2.5 nm), for instance, is largely higher than that of pure vermiculite (1.5 nm) and/or biotite (1.0 nm). In fact, this first4order reflection is close to the algebraic sum of the component interlayer spacing of the unit minerals (Ruiz4 Amil • ., 1992; Newman, 1987). Laboratory experiments by Saehr (1990 and 1992) also showed similar results while preparing biionic Na4 and K4vermiculites in 21 equilibrium with aqueous solutions of NaCl and KCl. Their X4ray diffraction (XRD) results showed interlayer spaces of K4vermiculite (1.08 nm), Na4vermiculite (1.41 nm) and the ordered interstratified Na/K4vermiculite (2.50 nm), and consequently the availability of two chemically different exchange sites owing to the existence of two different structural unities (Na4vermiculite and K4vermiculite). • The interlayer space created after intercalation of interstratified vermiculite by organic surfactants is also much higher than that from pure vermiculite and biotite, as well as their exfoliated forms (Ruiz4 Amil et al., 1992; Justo et al. 1993). • Furthermore, this interstratified vermiculite is amenable to producing a durable exfoliated material (Frank and Edmond, 2001). Water bound to the interlayer cations in clay minerals like vermiculite and montmorillonite is arranged in a series of layers, i.e. water layer hydration states (WLHS). These WLHS define the basal reflection and are intrinsically related to the interlayer cation(s), relative humidity and temperature (Barshad, 1950; Fripiat ., 1960; Marcos , 2003). According to Barshad (1948) and Ferrage . (2005), at room temperature and relative humidity around 35%, cations like Mg2+, Ca2+ and H+ will exhibit 24WLHS; Ba2+, Li+ and Na+ only 14WLHS; and NH4+, K+, Rb+ and Cs+ 04WLHS. In the case of the latter cations, exhibiting 04WLHS, their relatively smaller d4spacing is caused by the size of the dehydrated interlayer cation. In natural biotite (vermiculite parent mica) the interlayer ions are potassium and the interlayer space contains no water (04WLHS). By contrast, pure vermiculite (Mg4vermiculite) contains the equivalent of up to two sheets of 22 interlayer water (24WLHS) (Barshad, 1950; Harben and Roberts, 1990; Bergaya , 2006). During dehydration, initially there is release of water from external surfaces and meso4pores (unbound water), then the water of hydration (bound with the interlayer cation) and finally the water derived from dehydroxylation (combination of surface clay hydroxyl groups) (Bergaya ., 2006). The dehydration and/or rehydration of vermiculite is, to a certain extent, a reversible process (Grim, 1968; Newman, 1987). It progresses stepwise during the elimination and/or accumulation of one or several water layer structures, with some overlap, accompanied by a change in the d4spacing (Fripiat ., 1960; Kresten and Berggren, 1978; Bergaya , 2006). Reichenbach and Beyer also conducted several experiments with alkaline earth exchanged vermiculites and published an interesting series on dehydration and rehydration of vermiculites (Reichenbach and Beyer, 1994, 1995 and 1997; Beyer and Reichenbach, 1998). They applied thermo4analysis and in situ XRD to study the relationship between temperature and the WLHS of homionic vermiculites. Some of their results are depicted in Table 4. Furthermore, Reichenbach and Beyer (1994) concluded that Mg4 vermiculites are more thermally stable due to the well4defined water layer hydration states (24WLHS, 14WLHS and 04WLHS) compared with other ion exchanged vermiculites studied. Several WLHS were also found by Marcos work with natural commercial vermiculites supplementing preceding findings. 23 . (2003 and 2009) in their (including Palabora), ;1 Vermiculite WLHS properties (Grim, 1968; Hougardy et al., 1970; Reichenbach and Beyer, 1997) , 3 /0 + 5 ! ' ' * / * ' E * 2 0.51 1.38–1.44 4.00–4.71 1 N/A 1.15–1.16 1.57–1.88 0 N/A 0.93–1.00 0.00–0.51 2 0.58 1.47–1.59 4.59–5.39 1 N/A 1.18–1.19 1.23–1.83 0 N/A 095–0.98 0.00–0.21 1 0.34 1.24 2.36–2.53 0 N/A 0.99–1.00 0.00–0.50 2 N/A 1.48 5.6–6.0 1 0.56 1.18 2.0 0 N/A N/A N/A NH4+ 0 ~ 0.15 1.08 (diffuse) N/A K+ 0 ~ 0.13 1.06 (diffuse) N/A Mg2+ Ca2+ Ba2+ Na+ 1 N/A = data not available In previous work a vacuum was also applied to study the dehydration and rehydration of vermiculite (Okhotnikov ., 1989; Marcos , 2003). The results showed that during thermal dehydration of vermiculite there is no destruction of the crystals: water molecules are released by interlayer diffusion and escape through the cracks (Walker, 1959; Wild and Keay, 1964; Okhotnikov ., 1989). 24 Marcos (2003) found that exposing vermiculite to a vacuum causes rapid dehydration and that below 120 °C the vacuum dehydration is arrested at 14WLHS. They also observed that exfoliation under vacuum conditions is not dependent on the presence of water above 120 °C. 71? 55 ! Clays and soils containing clays are well known in agriculture in processes such as the control of swelling, ion transport and the retention of water. Small amounts of them can have a large impact on the desired properties (Annabi4Bergaya, 2008). The application of neat vermiculite is closely associated with its grade. However, it has been applied in lightweight insulation, aesthetic and decorative plastering, fire4resistant concrete and chimney insulation. Of some importance is its use by the breeders of chicken and some economic reptiles, and as a vitamin carrier (Mokhonoana, 1996; Horacek and Pieh, 2000). Vermiculite has been used effectively as an absorbent in water purification due to its capacity to exchange its interlayer cations with some toxic ions like Sr, Pb, Zn and Cd (Schoeman, 1989; Das and Bandyopadhyay, 1991; Konta, 1995; Mekhamer and Assaad, 1999). Many authors have mentioned the effectiveness of vermiculite as an adsorbent for the removal of natural and artificial organic matter, as well as for the controlled release of fertilisers and pesticides (Bailey and White, 1964; Schoeman, 1989; Abate, 2006). 716= Flame4retardant 55 compositions , have been proposed that rely on vermiculite’s expansion properties (Wada, 1973a). It is believed that vermiculite could impart fire resistance to polymers, provided the exfoliation temperature can be lowered sufficiently. Some patents claim that appropriately 25 intercalated vermiculites show lowered expansion onset temperatures (Langer and Marlor, 1981). Acid4activated vermiculite is used as a catalyst, especially during the cracking of heavy fuels to the highest gasoline portion in the petroleum industry (Suquet , 1994). Organomodified vermiculite has been used in the removal of herbicides from water and in herbicide formulations for the controlled release of active components in soils (Plachá 2008; Holešová , 2010; Yu 2010). Organovermiculite polymer nanocomposites with improved mechanical properties, as well as good fire4retardant and intumescent properties, are also discussed in the literature, although less frequently than other clay minerals (montmorillonite, hectorite and saponite) which are relatively less abundant and much more expensive (Connell 2005; Gomes 7166 , 2009; Qiu 55 , 1994; Connell , 2000; Liu , , 2009). 4 The lightweight (low bulk density) vermiculite produced by thermal expansion finds use in a variety of areas, according to the particle size and mechanical disintegration (Konta, 1995). Exfoliated vermiculite is of industrial interest because it is incombustible, lightweight and shows excellent absorption and thermal insulation properties (Wada, 1973b, Schoeman, 1989; Tomanec Bergaya ., 1997; Obut and Girgin, 2002; ., 2006). It is used as lightweight aggregate in concretes and plasters (Wada, 1973b; Konta, 1995; da Silva ., 2003). Expanded vermiculite is used in commercial and residential construction for fire protection to prevent warping, bending or collapse of the reinforced4 concrete or steel4beam framework. Due to its inertness with respect to heat and fire, expanded vermiculite is also used as a high4temperature insulator, as well 26 as for ambient4temperature refrigeration and sound dissipation (Schoeman, 1989; Evans, 1993; The Vermiculite Association, 2000; Bergaya , 2006). It is also used as thermal protection and shock4proof filling material for glass containers and vessels in the packaging industry (Konta, 1995; Bergaya , 2006). Since expanded vermiculite is able to retain liquids and gases better than natural soils, it is used in agriculture and animal feeds as an absorbent of water, air and other fluid plant and animal nutrients, and as a filler and carrier for powdered fertilisers and animal nutrients (Schoeman, 1989; Evans, 1993; Konta, 1995; The Vermiculite Association, 2000). A mixture of exfoliated vermiculite, modified bentonite and peat is currently applied as a deodorising agent in the protection of the environment. Exfoliated vermiculite is also recommended as a bio4remediate aid to clean contaminated soils and sites, as an absorbent or filter in water purification and beverages, and for the controlled release of fertilisers and pesticides (Schoeman, 1989; Konta, 1995; The Vermiculite Association, 2000). Exfoliated vermiculite is used in the friction brake linings market, for high4temperature insulation, loft insulation, various construction products, animal feeds, horticulture and many other industrial applications. Vermiculite, with its layered structure and surface characteristics, is utilised in products such as intumescent coatings and gaskets, for the treatment of toxic waste and in air4freight packaging of hazardous goods (Konta, 1995) 7167 Fig. 2 shows world production of vermiculite from 1994 to 2009. During this period, the average South African production was about 200 000 tons per year. 27 250 I 6=9 150 1 200 08 06 100 H I@ 04 02 00 50 98 96 0 94 71 Worldwide production of vermiculite, from 1994 to 2009 (Potter, 1996 to 2009; and Cordier, 2010) According to Potter (1996), from 1991 to 1995 South Africa was the main exporter of vermiculite to the USA (about 99%), while during the period 2005 to 2008 the country was responsible for around 59% of such exports (Potter, 2002; Cordier, 2010). This apparent decrease in exports is due to the emergence of new suppliers such as Australia, Brazil, China and Zimbabwe. Even with the appearance of these new suppliers, as shown above, the price of vermiculite 28 remained at approx. US$143 per ton, and production remained relatively stable during the last decade (Potter, 2005; Cordier, 2010). 7169 , , ! D "! The XRF technique is widely used in various areas since this method is fast, non4destructive and, usually, does not require laborious sample preparation. Absorbed energy from a primary X4ray source (X4ray tube or radioactive source) can eject sample electrons from inner shells, creating vacancies. So, to re4establish the stable condition the outer shell electrons have to move into the inner shells, producing characteristic XRF patterns (Fig. 3) (Jenkins, 1988; Wilson, 1994). 91 X4ray fluorescence after photoelectric effect with respective production of Kα and Kβ X4rays X4rays possess energy (E) that is a sinusoidal waveform with a typical wavelength (λ). These parameters and the frequency (ν), speed of light (c) and Planck’s constant (h) are related by the equation represented by Scheme X: 29 E = hν = , hc λ Planck’s energy quantisation law GC X4rays are labelled K, L, M or N, according to the destination shell. The suffixes alpha (α), beta (β) or gamma (γ) are added to distinguish X4rays originating from electrons from the different shell levels (Jenkins, 1988; Wilson, 1994). # $ % & '% ( % % #$& '( Several techniques are disclosed in the literature regarding the determination of CEC and exchangeable cations in clays and clay minerals. The most commonly applied techniques are based on the exchange and saturation of interlayer cations using an mono4ionic ‘index’ cation in solution (theoretically not present in the sample) in a known quantity (Bergaya 2006, Aran, , , 2008). Various analytical techniques (atomic absorption, spectrophotometry, ICP4MS, ICP4OES, UV4Vis, titration, etc.) are then used to quantify the index cation and the displaced cations. The basic ICP4OES instrument has the following components: (i) light path to optical system; (ii) excitation zone; (iii) emission zone; (iv) water4cooled plasma interface; (v) argon supply; (vi) plume; (vii) analytical zone; (viii) radiofrequency (RF) coil; (ix) plasma torch; and (x) a sample flow compartment (Ametek, 2009). In the excitation zone, the temperature of the air4acetylene flame reaches values between 1 700 °C and 2 700 °C. Inert argon in the plasma is heated to high temperatures which efficiently excite many elements and prevent interference from oxides and nitrides (Ametek, 2009). 30 The current ICP4OES instrument records the entire spectrum between 175 and 777 nm, enabling simultaneous determination of more than 70 elements (Ametek, 2009). The most relevant CEC determination methods are usually called by the name of the chemical compound used to extract the interlayer cations (index cation) and comprise: • Ammonium acetate (Schollenberger, and Simon, 1945; Barshad, 1954a; Barshad, 1954b; Tan, 1996) • Barium chloride (Bascomb, 1964; Hendershot and Duquette, 1986) • Cobaltihexamine chloride (Ciesielski • Methylene blue (MB) (Carter and Wilde, 1972; Kahr and Madsen, , 1997; Aran , 2008) 1995) • Silver4thiourea (Dohrmann, 2006a; Dohrmann, 2006b). The interlayer CEC of vermiculites and other 2:1 clay minerals can also be estimated from the layer charge density as shown in Scheme XI (Lagaly, 1979 and 1982; Ghabru , 1989; Mermut and Lagaly, 2001). ξ Interlayer CEC RcmolTKgU = 10W Y M , G C Calculation of interlayer CEC from interlayer charge Y is the mean relative molar mass and ξ is the layer charge, which Where M can be determined by measuring the interlayer spacing after intercalation with alkylammonium cations (Z [\ ] ) of varying carbon chain lengths. ) * )+ TG is an analytical technique that measures the mass change of the analyte as a function of temperature, independent of bonding mode. TG results 31 are evaluated and interpreted by thermogravimetric analysis (TGA) and/or by evaluating the respective derivative against temperature (dm/dT), known as derivative thermogravimetry (DTG). DTG also shows the rate at which the mass changes (Brown, 2001; Haines, 2002; Bergaya ;1 , 2006; Gabbott, 2008). Basic TGA instrument (adapted from Brown, 2001) Modern thermal instruments (TG, DT, TMA, DSC, etc.) comprise basically an electronic balance, furnace, atmosphere control, sample holder, thermocouples, temperature control and a data collection device (Brown, 2001; Haines, 2002; Bergaya , 2006). A schematic representation of a TG instrument is shown in Fig. 4. 32 , ) - ) - TMA or static force4TMA (sf4TMA) is a technique that derives from thermomechanometry. It is a very important tool for studying materials that are to be subjected to a wide variation in temperature during their usage. An illustrative example is food packaging which has to undergo thermal processes such as heating, pasteurisation, freezing, etc. (Brown, 2001; Haines, 2002). In TMA the sample’s dimension or length is measured as a function of temperature while it is under constant mechanical stress. Usually the results are evaluated by examining the changes in the thermal expansion coefficients with temperature and/or time (Haines, 2002; Gabbott, 2008). Modern TMA instruments are designed to measure: (i) penetration; (ii) extension; (iii) flexure; and (iv) torsion (Brown, 2001). For extension measurements, a flat4ended probe is inserted and rested on the top surface of the sample, and a static force is applied. A sensor then measures the movement of the probe (Haines, 2002). . / "/ The XRD technique has been extensively used to identify and distinguish vermiculite from other clay minerals (e.g. chlorite and smectites) and, in fact, the definition of vermiculite is based on its ca. 1.45 nm basal spacing for Mg4 vermiculite treated with glycerol and on the characteristic 1.0 nm spacing for the K4vermiculite form on heating at 300 °C (Brindley and Brown, 1980; Bergaya , 2006). XRD is also used to distinguish the dioctahedral (d060 = 0.149–0.150 nm) from the trioctahedral (d060 = 0.151–0.153 nm) clay minerals (Brindley and Brown, 1980). XRD is a versatile and also non4destructive technique that reveals detailed information about the phases contained, as well as crystallographic features (Cullity, 1978; Moore and Reynolds Jr., 1989). 33 their θ <1 X4ray diffraction by a crystal according to Bragg’s law X4rays are a radiation of short wavelength, from 1043 nm to 10 nm. They are generally produced during the collision of a high4energy electron with a metal target (Cullity, 1978; Moore and Reynolds, 1989). When the incident wavelength (λ) is of the same order of magnitude as the repeated distance (d) between scattering centres, diffraction can occur. This diffraction phenomenon, known as Bragg’s law and represented in Fig. 5, is mathematically described by Scheme XII. n λ = 2 d sinθ , G C Bragg’s law Where: n is an integer (n = 1, 2, ...) d is the interlayer spacing θ is the diffraction angle or Bragg’s angle λ is the wavelength of the incident X4rays 34 Vermiculite content , % 100 80 60 40 20 0 1.0 1.1 1.2 1.3 1.4 1.5 d-spacing, nm 81 Determination of vermiculite content from d4spacing (according to Moore and Reynolds, 1989) By knowing λ and measuring θ, d can be calculated using the relationship shown in Scheme XI. Furthermore, vermiculite content can be determined by applying the d4spacing determined using Bragg’s law, as shown in Fig. 6 (Moore and Reynolds, 1989). 0 *( % ( Nowadays SEM (along with TEM) is one of the most important characterisation techniques as a result of the huge technological improvements made since its invention and the relatively easy interpretation of data. This technique has also been used by some researchers to study the surface changes in vermiculite resulting from physical or chemical modification. SEM apparatus basically consists of (Lawes, 1987): (i) electron4optical column (principally an electron gun, lens and detector); (ii) vacuum system (to 35 prevent scattering of electrons by collision with gas molecules); and (iii) electronics and a display system (mainly an amplifier, cathode4ray tubes and a control system). In a SEM, the electron gun bombards the sample with electrons of a specific wavelength. Deflected electrons, commonly elastically scattered electrons, are collected as signals for imaging. Different topographies and different chemical compositions bring about different signal intensities, thereby giving different contrasts on the SEM screen (Lawes, 1987). ' 1 Other relevant techniques have been being applied to characterise vermiculite and other clay minerals. Techniques such as particle size analysis, IR spectroscopy (Farmer, 1971; Farmer, 1974; Beutelspacher, 1976; Raupach, 1988; Badreddine Raman spectroscopy (Okui , 1998; Arab Van der Marel and 2002c), as well as , 2002; Rinaudo , 2004), have been shown to be useful for the characterisation of vermiculite and its exchanged forms. 36 ) 9 )G ) ). 0 916 55 Mandoval Vermiculite supplied the superfine (1 mm) and micron (0.5 mm) samples of Palabora material. All metal chlorides and ammonium salts were obtained from Merck Chemicals. The orbital shaker (SPO4MP15) used to run the experiment was supplied by LABEX. For these experiments, a long4term approach and/or the use of a shaker were applied due to the high layer charge of vermiculite. Previous kinetic studies have already shown a very slow rate of exchange in similar vermiculites due to the high layer charge density (Inoue, 1984; Skipper , 1991; Kodama and Komarneni, 1999). All experiments were run at the pH of the salt solution and it was not controlled accurately since the exchange rate is not affected by the pH in the range of 4 to 9 (Wild and Keay, 1964). 917 4 , For the first set of reactions it was attempted to replace the magnesium ions with sodium ions using concentrated brine, since vermiculite selectivity rises with an increase in the preferred ion loading (Wild and Keay, 1964; Gast and Klobe, 1971) (Scheme XIII). The sodium exchange reaction was performed over a period of six months in closed containers, having the vermiculite submerged in a large volume of saturated sodium chloride solution. The suspension was agitated daily. Every fortnight approximately 1 000 ml of the supernatant liquid was exchanged with fresh saturated brine. At the end of every month, vermiculite samples were taken for analysis. They were rinsed with distilled water and allowed to air4dry at ambient conditions. 37 The exchange process was therefore controlled by monthly extraction of aliquots by taking out Na4exchanged vermiculite samples after first month up to the sixth month. The second approach of ion exchange by ammonium, alkali metal and alkaline earth metal ions was carried out using a shaker for exactly 360 hours (15 days) at room temperature. The frequency of the shaker was adjusted to 200 rpm (rotations per minute). Na+ (excess) Legend: , G Mg2+ C Ca2+ K+ Na+ Schematic representation of ion exchange in interstratified vermiculite. It is assumed that the biotite layers (containing K as interlayer cation) remain intact after treatment For these experiments, 50 g of vermiculite were weighed off for each sample and enough salt was added to prepare 250 ml of saturated solution (for the reasons mentioned above). For both sets of experiments, the ion exchanged vermiculites were then washed thoroughly with distilled water until all chloride anions had been 38 removed quantitatively. This test was performed by reaction of an aliquot of supernatant solution with 0.1M AgNO3 (Scheme XIV). + (aq) + Ag 4 → Cl (aq) Kps = 1.8x10410 AgCl(s) G BC Reaction of Ag+ and Cl4 ions with formation of a white , precipitate of silver chloride 919 , , ! The chemical "! composition of neat and long4term Na4exchanged vermiculites was determined by XRF. For major element analysis a milled sample (< 75 dm) of the neat or ion exchanged material was calcined at 1 000 °C for at least 3 h to oxidise Fe2+ and S and to determine the loss on ignition (LOI). Glass disks were prepared by fusing 1 g of roasted sample and 8 g of 12–22 flux consisting of 35% LiBO2 and 64.71% Li2B4O7 at 1 050 °C. The glass disks were analysed by a PANanalytical Axios X4ray fluorescence spectrometer equipped with a 4 kW Rh tube. Ammonium, alkali metal and alkaline earth metal ion exchanged vermiculites were analysed using an ARL9400XL+ spectrometer. One gram of sample was then placed together with 6 g of Li2B4O7 into a Pt/Au crucible and fused into a glass bead. The samples were roasted at 1 000 °C to determine the LOI. The analyses were done using Quantas software. The results were also monitored and filtered to eliminate the presence of some of the flux – the elements of the wetting and oxidising agents. # % With the ammonium acetate extraction method it was attempted to quantify the interlayer cations regardless of the compositional (interstitial or 39 structural) cations, which in some cases are definitely present in higher quantities. The extracted Na+, K+, Mg2+ and Ca2+ ions were quantitatively determined using an ICP4OES Spectro Genesis spectrometer, equipped with Smart Analyzer Vision software and a 27 MHz free4running generator for constant plasma power output. & 2 The particle size distribution (PSD) was obtained by laser diffraction using a Malvern Mastersizer 2000 fitted with a Hydro 2000G dispersion unit. Samples were prepared in duplicate and three runs were made per replicate. , ) * )+ Thermogravimetry (TG) was performed by the dynamic method on a Mettler Toledo A851 TGA/SDTA instrument. For TG analysis, a sample of about 15 mg was placed in an open 150 µl alumina pan. The temperature was scanned from 25 to 1 000 °C at a rate of 10 °C/min with air flowing at a rate of 50 ml/min. . ) - ) - Thermal expansion measurements were conducted with a TA Instruments Q400 Thermo Mechanical Analyzer. The thickness of the micron4grade flakes tested ranged from 91 em to 231 em and their cross4sectional surface area varied from ca. 1.66 mm2 to about 4.2 mm2. In each experiment, a single flake was sandwiched between the flat surface probe and the bottom surface of an alumina pan. The flake’s expansion behaviour was measured under a suitably chosen applied force. 40 In the dynamic scan experiments, the temperature was scanned from 30 °C to 1 000 °C in argon at a scan rate of 10 °C/min. Since the objective was to follow the free expansion of the flakes, the lowest possible force (0.001 N) was applied. The expansion relative to the original flake dimension was reported. Additional TMA experiments were limited to the sodium vermiculite obtained after six months of brine exposure. The effect of the magnitude of the applied force was investigated using settings of 0.05, 0.20 and 1.0 N. Finally, the expansion behaviour of the flakes subjected to a stepwise increasing temperature4change protocol was investigated. In these experiments the temperature was scanned at 10 °C/min to an isothermal hold temperature at which the sample was kept for 10 minutes before the temperature scan was continued. The first hold temperature was 270 °C and the subsequent hold temperatures were set at 10 °C increments, with the final isothermal halt at 340 °C. 0 " / "/ For X4ray diffraction (XRD) analysis, samples of neat and ion exchanged vermiculite were split and a sub4sample of each was homogenised and milled to a fine powder. Due to the layered structure of vermiculites, careful sample preparation was essential for meaningful and accurate results. Special precautions were taken to prevent preferred orientation when it was not needed. Samples were split and representative portions were milled and homogenised to a fine powder of approximately 15–20 µm in size. Prior to analysis, a sub4sample was lightly ground in a mortar and pestle, and pressed into a shallow sample holder against a rough filter paper in order to ensure random orientation. The evaluation was done on both random and oriented preparations. The neat vermiculite was saturated with Mg prior to the various treatments. All 41 oriented preparations were analysed in an air4dried state (AD), after ethylene glycol (EG) solvation and heating to 550 ºC, for additional information. X4ray diffraction data were obtained on a BRUKER D8 Advance diffractometer instrument with CuKα radiation (λ=1.5406), a Johansson crystal primary monochromator and a LynxEye detector with an active area of 3.7º. Scans from 2 to 70º 2θ for random powder preparations and 2 to 28º 2θ for the oriented slides were recorded in step4scan mode at a speed of 0.01º 2θ step size/4 s and generator settings of 40 kV and 40 mA. Mineral identification was done by means of the BRUKER DIFFRACPlus – EVA evaluation program using the International Centre for Diffraction Data (ICDD) Inorganic/Organic Database. Phase concentrations were determined by the Rietveld quantitative method using the DIFFRACPlus – TOPAS software at an estimated accuracy of ± 1%. Refinement of diffraction data was done using the same program. *( % ( Images of neat and Na4exchanged vermiculite were obtained on a JSM4840 scanning electron microscope equipped with Orion 6.60.4 software. For these studies the vermiculite flakes were mounted on adhesive carbon tape. The temperature4driven dynamic exfoliation process was also studied under a scanning electron microscope (FEI QUANTA 200 ESEM) fitted with a heating stage. The samples were placed inside a crucible and mounted in the heating stage. They were viewed at 200x magnification. The pressure was 0.5 kPa, voltage 20 kV, spot size 6–7 and a working distance of 16–20 mm was used. The temperature was ramped at 20 °C/min. 3 ! ) 4# !) #" % % Fourier Transform–Infrared Spectroscopy (FT4IR) spectra were recorded using a PerkinElmer Spectrum RX FT4IR, coupled with a computer using 42 Spectrum v5.0.1 software, with a scan resolution of 2.0 cm41. A total of 32 interferograms were collected for each sample, applying single4beam radiation. KBr compressed pellets were prepared from a fine powder that was obtained by grinding a mixture of 3–5 mg of sample with 100 mg of dehydrated KBr using a mortar and pestle. 43 ) ; ) $0 .- - $ &. The superfine (1 mm) and micron (0.5 mm) grades of vermiculite ostensibly differed in flake size only. A backscattered SEM image (Fig. 7) of the neat material showed the flaky nature of vermiculite. Granular impurities were observed in between the flakes and on the surface of the aggregates. SEM image of neat superfine vermiculite :1 ;16 , G+ 5 'G * 5 * 6 * The XRF4derived chemical compositions are presented in Table 5. The sodium content showed a step change from effectively zero (0.01 mass %) to an average value of ca. 1.67 ± 0.11 mass % after one month’s exposure and 44 thereafter. The presence of calcium and phosphorus is attributed to apatite and dolomite contamination. These XRF results also emphasise the fact that Palabora vermiculite is not pure because the content of potassium oxide (K2O) is higher than 0.35% (Justo . 1986). This is in accordance with previous findings (Inoue, 1984; Marcos ., 2009). <1 XRF (%) composition analysis of long4term Na4exchanged samples , = 6 7 9 ; < 8 SiO2 39.74 40.27 40.01 39.73 39.72 40.28 40.41 MgO 22.70 22.23 21.60 22.75 21.73 22.40 22.12 Al2O3 8.36 8.62 8.35 8.76 8.33 8.74 8.61 Fe2O3(t) 7.06 7.30 7.10 7.58 7.12 7.42 7.27 CaO 5.72 5.03 6.46 4.28 6.27 4.71 5.37 K2O 4.81 4.97 4.85 5.30 4.84 5.08 4.87 P 2 O5 2.81 2.40 3.29 1.95 3.15 2.32 2.66 Na2O 0.01 1.50 1.44 1.64 1.59 1.71 1.51 TiO2 0.88 0.90 0.86 0.92 0.87 0.90 0.90 MnO 0.05 0.05 0.05 0.05 0.05 0.05 0.05 Cr2O3 0.02 0.02 0.02 0.02 0.02 0.02 0.02 LOI 7.83 6.66 5.83 6.59 6.21 6.22 6.20 ??1?; ??1?< ??1>8 ??1<: ??1?6 ??1>8 ??1?? Cardille and Slade (1987), Schoeman (1989), Badreddine Marcos (2002a), . (2009) and Marcos and Rodríguez (2010) also found that Palabora vermiculite has, comparatively, an unusually high level of iron. These iron ions (and also aluminium ions) are the main magnesium isomorphous substitutes in 45 the octahedral sites, and there are also considerable silica replacements in the tetrahedral sites, in agreement with the findings of Cardille and Slade (1987). 2+ 3+ + 0.60 ( Mg 2.60 Fe0.60 )[ Si3.2 Al0.78 Fe0.02 ]O10 (OH ) 2 K 0.49 2+ 3+ 2+ + 0.40( Mg 2.60 Fe0.60 )[ Si3.2 Al0.78 Fe0.02 ]O10 (OH ) 2 Mg 0.20 .nH 2O , GBC Structural formula of neat vermiculite consistent with the XRF data The unknown quantities in Scheme XV were determined by fitting the formulae to the XRF data for the other elements using atom balances. In this calculation it was assumed that titanium and manganese occupied the tetrahedral and octahedral positions respectively. However, neglecting the existence of these impurities made little difference to the calculations. The composition estimated from the XRF data suggests about 60% biotite and 40 % vermiculite with a CEC value of about 98 meq/100 g. , 6 7 8 8 * The modified vermiculites showed an expected increase in the exchanged cations, confirming the success of the ion exchange process under the conditions considered. As can be seen from Table 6, barium vermiculite (BaVER) shows the most noticeable change in chemical composition – it increased by almost 4% (described in terms of BaO) compared with neat vermiculite. This is in agreement with previous findings that refer to the high affinity of barium ions for the surface of clay minerals (del Rey4Perez4Caballero and Poncelet, 2000; Shahwan and Erten, 2004). 46 81 XRF (%) composition analysis of ammonium, alkali metal and alkaline earth metal ion exchanged vermiculite samples 5 . B) B) B) B) B) . ;B) SiO2 38.76 35.64 39.54 39.45 39.38 40.29 MgO 23.42 23.81 24.21 22.44 22.49 22.74 Al2O3 9.47 9.24 10.16 9.60 9.69 10.11 Fe2O3 8.27 10.09 8.12 8.17 8.24 8.21 CaO 3.87 5.40 3.42 5.06 3.40 3.76 K 2O 5.10 6.77 5.07 5.36 4.89 5.18 P 2O 5 0.88 1.19 1.17 0.98 1.12 0.99 Na2O 1.69 <0.01 <0.01 <0.01 <0.01 <0.01 TiO2 0.94 1.10 1.05 1.00 1.02 1.01 MnO 0.04 0.06 0.06 0.06 0.07 0.06 Cr2O3 0.03 0.03 0.03 0.03 0.04 0.04 NiO 0.03 0.03 0.03 0.03 0.04 0.03 V 2 O5 <0.01 <0.01 0.01 0.02 <0.01 <0.01 SO3 <0.01 <0.01 <0.01 0.04 <0.01 <0.01 BaO <0.01 <0.01 <0.01 <0.01 4.00 <0.01 Cl <0.01 <0.01 <0.01 0.03 <0.01 <0.01 CuO 0.01 0.03 0.02 0.02 0.02 0.02 ZnO 0.02 <0.01 0.02 0.02 0.01 0.01 Rb2O 0.05 0.09 0.05 0.05 0.04 0.05 SrO 0.02 0.07 0.02 0.01 0.01 0.02 LOI 7.39 6.43 7.02 7.64 5.50 7.49 6==1== 6==1== 6==1== 6==1== 6==1== 6==1== 47 By contrast, magnesium vermiculite (MgVER) did not show substantial compositional changes. This was expected due to the fact that neat vermiculite is composed mainly of magnesium ions as exchangeable cations. ;17 5 4 To determine whether the exchangeable cations in the neat vermiculite have been successfully exchanged, the interlayer composition of the neat vermiculite and the ion exchanged vermiculites were determined by the ammonium acetate method. This procedure showed tangible results compared with the XRF results. For instance, the exchangeable Mg2+ ions are negligible in comparison with the structural Mg2+ and thus the difference after exchange is not easily detectable by XRF. However, after extraction with ammonium acetate it was possible to differentiate between them. 6=== Ca Na K Qapp E Mg I 6== 6= 6 = 6 7 9 I >1 Change in interlayer < 8 , composition of long4term treatment with NaCl 48 ; vermiculite during According to the ICP4OES results (Fig. 8), the vermiculite interlayer spaces initially contained both Mg2+ and Ca2+ ions, in addition to some K+ ions. The extraction data indicate almost quantitative replacement of the exchangeable Mg2+ and Ca2+ ions, in the vermiculite phase, by Na+ ions. The Ca2+ ions that were not replaced by Na+ ions but were extractable by ammonium acetate are attributed to the presence of the apatite [Ca3(PO4)2]. Baddredine (2002b) ascribed this to calcite impurities, although they could not detect it using XRD techniques. The majority of the K+ ions, i.e. those associated with the mica phase (located in the hexagonal cavities of the O24 ion sheets of the layer surface), were not exchanged by sodium, consistent with previous reports (Barshad, 1948; MacEwan . 1961, Newman, 1987). The apparent total metal charge adsorbed into the vermiculite (Qapp) was determined using Scheme XVI. According to Fig. 8, Qapp is not affected during long4term brine treatment. abcc = d ef gf GB C Determination of apparent total metal charge , Where: Qapp is the apparent total metal charge Z is the ionic charge n is the number of mol ;19 , , ' (* * 5 6 * The thermogravimetric (TG) curves for the unaltered materials and the Na4exchanged forms are shown in Fig. 9. Mass loss commenced below 100 °C 49 and occurred in several steps for all the samples. As shown in Fig. 10, the total mass loss was 7–9 mass % for all samples. This compares with a total expected mass loss of 10.5 mass %, with interlayer water contributing 6.5 mass %, for the structure indicated in Scheme XV with the assumption that n = 4. 100 IJ 99 98 97 96 95 94 Neat 1 month 2 months 3 months 4 months 5 months 6 months 93 92 0 200 400 600 5 ?1 800 1000 I° TG plots of micron vermiculite and its Na4exchanged forms Fig. 10 reports the overall mass loss and percentages for the temperature regions < 250 °C, 250–520 °C and 520–1 000 °C. Most of the mass loss occurred in the first and final temperature ranges. As shown below, most of the vermiculite exfoliation occurred in the temperature range 250–520 °C. Surprisingly, the explosive exfoliation corresponded to the release of less than 0.5 mass % water. 50 8 7 To 250°C 250 4 520°C 520 4 1000°C Total I! J 6 5 4 3 2 1 0 = 6 7 9 I < 8 , Mass loss for neat vermiculite and its Na4exchanged forms 6=1 , ; 6 7 8 8 * According to Fig. 11, ion exchanged vermiculites show a multi4stage decomposition. Mg4vermiculite (MgVER) exhibits relatively (well known) stable intermediates, while NH44vermiculite (NH4VER) shows extensive dehydration in the second stage. NH4VER shows a decomposition that resembles the recoil effect (Gabbott, 2008) and an overall total mass loss close to 18%, whereas the other ion exchanged vermiculites show losses below 10%. The higher mass loss in NH4VER is a result of the release of both water and ammonia (NH3(g)) (Stone and Wild, 1978; Pérez4Maqueda , 2004; Pérez4Rodríguez 51 , 2004). 100 98 IJ 96 94 92 90 NaVER KVER 88 MgVER CaVER 86 BaVER NH4VER 84 82 80 0 661 200 400 5 600 IK 800 1000 TG plots of metal ion exchanged vermiculites In the range of 250–520 °C, which is the expansion temperature of vermiculites, only around 1.0% of mass is lost for all vermiculites ion exchanged with metal ions. However, in the same interval, NH4VER shows mass losses of around 7.1%, for the same reasons mentioned above (Fig. 12). 52 12 To 250°C 2504520°C 52041000°C Total 10 I! J 8 6 4 2 0 . B) 671 B) B) B) B) . ;B) Mass loss of ion exchanged vermiculites Between 520 and 1 000 °C dehydroxylation of all samples occurs. Again NH4VER shows comparatively high mass losses. This behaviour is attributed to the release of both ammonia (Scheme XVII) and water due to expected vermiculite dehydroxylation (Scheme XVIII) (Stone and Wild, 1978; Pérez4 Maqueda , 2004; Pérez4Rodríguez hij)\ , → hij) (bk) + \ ](m) GB C Thermal decomposition of ammonium ions in vermiculite hij)2 , (bk) , 2004). GB C (bk) → hij) (bk) + (m) Dehydroxylation of vermiculite A supplementary water quantity, resulting from the combination of hydrogen ions released during ammonia liberation and hydroxyl groups from 53 the vermiculite surface (Scheme XIX), is also expected to contribute, as observed by Pérez4Maqueda (2004) and Pérez4Rodríguez (2004) during thermal decomposition of NH4VER. hij) (bk) + hij) (bk) → 2 hij) + (m) G GC Dehydroxylation of vermiculite aided by hydrogen ions , ;1; , + , ' * The free expansion curves, obtained at an initial force of 0.02 N and with constant temperature ramping (Fig. 14 and Fig. 16), showed a rather abrupt rise at the expansion onset temperature, followed by a more gradual expansion over a temperature range of several hundred degrees Celsius. The initial steep rise leads to linear expansions equivalent to three to six times the initial flake thickness at rates ranging from 3%/s to 16%/s. The maximum expansion was attained at around 700 °C. Above this temperature some sample contraction was observed. ,, * 5 6 * The exfoliation onset temperature of the neat material exceeded 450 °C. All the long4term sodium4exchanged samples had similar exfoliation onset temperatures between 260 °C and 300 °C (Fig. 13). Irregular variation in the initial stages of expansion of different flakes from the same sample was detected during the experiments. This could obviously be attributed to the variability of the properties of those individual flakes. 54 10 I '+* 8 6 Neat )45 1 month 4 2 months 3 months 4 months 2 5 months 6 months 0 100 200 300 400 5 691 500 600 700 800 I TMA expansion of vermiculite and its Na4exchanged forms Fig. 14 shows the effect of the applied force on the expansion behaviour of the six4month sodium4exchanged vermiculite. The observed variation in the expansion onset temperature was attributed to the variability of the properties of individual flakes rather than to the change in the applied force. The smallest flake tested had a cross4sectional area of 1.66 mm2. At an applied force of 1 N, this corresponds to a pressure of only 0.6 MPa. The lowest onset temperature exceeded 250 °C, and at this point the vapour pressure of water is already 3.92 MPa. Thus, it is unlikely that the TMA applied force would have affected the expansion onset temperature. 55 11 10 Expansion ratio, (-). 9 8 7 6 5 4 3 2 0.001 N 0.05 N 1 0.20 N 1.0 N 0 200 400 600 800 1000 Temperature, (°C) 6;1 Effect of the TMA applied force on the expansion behaviour of vermiculite flakes reacted for 6 months with saturated brine Fig. 15 shows the effect of the stepped4temperature4increase protocol on the expansion behaviour of six4month sodium4exchanged vermiculite. Again it reveals considerable variability in the behaviour of the individual flakes. Notably, significant expansion occurred during the isothermal periods. This is indicated by the staircase4like expansion as a function of time. The expansion of Flake A commenced at the first isothermal hold temperature of 270 °C. The temporal evolution of this isothermal expansion is shown in Fig. 16, together with the isothermal expansion results for other flakes proceeding at different hold temperatures. Remarkably, in at least two of these curves, the expansion was characterised by an induction time before a sigmoidal expansion occurred. It is assumed that water is continuously lost as time proceeds during the isothermal 56 period. The observed time delay then suggests that expansion was triggered when a critical water content level was reached inside the interlayer space. The expansion of Flake A, at the hold temperature of 270 °C, actually took place in two distinct steps. This can be explained by assuming that the composition of the flakes is not homogeneous along the c4axis, i.e. that the vermiculite content varies with distance along the flake thickness. TMA characterisation of the isothermal expansion of long4term 6<1 Na4exchanged vermiculite, at selected temperatures ,, 6 7 8 8 * It is evident from Fig. 16 that the sodium, potassium and ammonium ion exchanges lowered the exfoliation onset temperature of vermiculite to temperatures below 300 °C. Among these, sodium vermiculite (NaVER) shows a comparably better expansion ratio, while magnesium vermiculite (MgVER) shows a thermal behaviour similar to that of neat vermiculite (with the majority of magnesium ions as interlayer 57 cations). This latter fact demonstrates the great contribution of the interlayer cations to the expansion behaviour of vermiculites. In the increasing ionic potential series of Na4, Ba4, Ca4, Mg4exchanged vermiculites, the expansion onset temperature seems to increase dramatically. Hence these TMA results also provide evidence that cations with lower ionic potential (see Table 3) produce ion exchanged vermiculites with comparatively low expansion onset temperatures. 12 NeatVER NH4VER I '+* 10 NaVER KVER 8 MgVER CaVER 6 BaVER )45 4 2 0 0 100 200 300 400 5 681 500 600 700 800 I 'K * TMA expansion of ion exchanged vermiculites The second step of expansion of NH4VER occurs at slightly lower temperatures. This is also due to the release of ammonia and water resulting from the combination of hydrogen ions from ammonium ions and hydroxyl groups from the vermiculite surface. This is in agreement with the results of Pérez4Maqueda (2004) who also observed that the quantity of water released was more than expected from the dehydroxylation of vermiculite itself. 58 The stages involved in this process are demonstrated above in Schemes XVII and XVIII. ;1< G+ ,. * 'G -* 5 6 * The main mineral impurities found via XRD analysis were dolomite and apatite. Fig. 17 depicts the evolution of the XRD spectra with time of brine treatment. 6:1 Evolution of the XRD spectra as a function of brine exposure time The neat material featured a strong, broad reflection at 2θ = 7.43 ° (1.19 nm) and weaker reflections at 2θ = 6.17 ° (1.43 nm) and 2θ = 8.77 ° 59 (1.02 nm). The 1.43 nm reflection is consistent with the Mg4vermiculite phase with two (slightly incomplete) water bilayers (Mathieson, 1958; Pérez4Maqueda , 2003). The 1.02 nm reflection is from the mica (biotite/phlogopite). Perfectly alternating 50/50 mixed4layer vermiculite, i.e. ‘hydrobiotite’, features reflections at 2.44 nm (001) and 1.22 nm (002) (Newman, 1987; Ruiz4 Amil ., 1992). In the present sample the main reflection is located at higher 2θ angles. This and the extensive line broadening are indicative of a random distribution of the vermiculite layers. While the sodium uptake reached a plateau by the end of the first month, the XRD pattern (Fig. 17) suggested continued evolution up to month six. The intensity of the main reflection increased with time but had settled at its ultimate d4spacing of 1.12 nm already by the end of the first month. After six months of brine exposure, reflections were found at ca. 2.44 nm, 1.22 nm; 1.12 nm and 1.01 nm. The broad reflection at ca. 2.44 nm and the well4resolved 1.12 nm reflections are diagnostic for the formation of a more ordered structure, similar to ‘hydrobiotite’. However, pattern simulations using NEWMOD software indicated a composition of ca. 45 % vermiculite. ,. -8 8 6 * According to the results shown in Fig. 18, Mg4vermiculite (MgVER) showed similarities with the blank vermiculite, i.e. neat vermiculite subjected to the same physical treatment as the other samples. The main reflections were almost in the same positions and the main Mg4vermiculite peak was present in both samples at ca. 2θ = 6.17 ° (1.43 nm). However, the blank vermiculite showed more well4defined interstratification peaks, at 2θ = 3.26 ° (2.7 nm), 2θ = 7.13 ° (1.24 nm) and 2θ = 7.41 ° (1.19 nm), due to the presence of other interlayer cations (mainly K). This resemblance supports the fact that Palabora vermiculite is mainly of the Mg4type. 60 Rietveld quantitative analysis of blank vermiculite showed the following mineralogical composition (Table 7): Mineralogical composition of blank vermiculite :1 5, Vermiculite 'J* ' * 15.88 1.43 72.10 2.71, 1.24, 1.19 Mica (biotite/phlogopite) 6.13 1.01 Apatite 2.73 Dolomite 2.61 Mica/Vermiculite interstratifications In K4vermiculite (KVER), all typical vermiculite features had disappeared and the lattice had drastically collapsed (Gast and Klobe, 1971). By the same token, the typical well4ordered mica4like structure reflection at 2θ = 8.74° (10.10 nm) was mostly observed (Fig. 18). Ca4vermiculite (CaVER) and Ba4vermiculite (BaVER) showed similar characteristics. The expected regular interstratification reflections at lower 2θ angles (long4spacing reflections) were not clearly observed, although it was seen that K+ ions were still present in those samples. They both showed a distinctive and intense (002) reflection and a typical mica reflection at ca. 2θ = 8.7° (10.10 nm). 61 NH4VER I 1 1 Blank BaVER CaVER MgVER KVER NaVER 7 ; 8 > 6= 67 7θ, θ, °' 6>1 6; 68 6> 7= 77 α* XRD patterns of blank, alkali metal and alkaline metal treated vermiculites All samples, prepared with different salts, showed the d060 reflection at 1.53 nm, confirming the preservation of the trioctahedral mineral structure. A comparison of the XRD results of neat and blank vermiculites is presented in Fig. 19. It shows the effect of applying the shaker during vermiculite treatment. Blank vermiculite showed well4resolved reflection peaks, especially at small 2θ angles, and no reflection shift was observed. This result should be attributed to the effect of washing and the removal of soluble impurities and small powdered vermiculite produced during milling process, as well as the shaking process. 62 I 1 1 Blank Neat 0 10 20 30 7θ, θ, °' 6?1 40 50 60 70 α* XRD patterns of neat vermiculite and blank vermiculite (the latter subjected to similar physical treatment given to all ion exchanged samples) Characteristic interstratified vermiculite reflection peaks were noticeable in both samples at 2θ = 3.30° (2.68 nm) and 2θ = 7.41° (1.19 nm). Typical Mg4 vermiculite and mica reflection peaks were located at 2θ = 6.18° (1.43 nm) and 2θ = 8.78° (1.01 nm) respectively. 63 ;18 ) 7=1 5 ' ) * SEM image of vermiculite flakes reacted for 6 months with saturated brine Generally, the vermiculite particles are flake4like with striations, humps and folds on the surface. The brine treatment did not affect the macroscopic appearance of the Palabora material. Fig. 20 shows a SEM surface topographic image of flakes that were exposed to brine for six months. The damage caused by the milling process is manifested by the folded and lacerated edges. The expansion process of sodium4exchanged vermiculite was studied as a function of time and temperature in an ESEM (Appendix V). Careful study of the full video clearly shows heterogeneous exfoliation of a single flake. 64 761 SEM image of heat4expanded Na4vermiculite The expansions of sodium4exchanged vermiculite occurred in stages at different positions along the flake thickness (Fig. 21). It is remarkable that significant expansion still occurred despite the low pressure of only 0.5 kPa in the ESEM chamber. ;1: F The particle size distribution for the micron4grade vermiculite showed an apparent bimodal distribution (Fig. 22) and was not materially affected by exposure to the concentrated brine. The lower peaks range from ca. 5 em to just above 200 em. Since the material consisted essentially of high4aspect4ratio flakes, the lower hump is indicative of the variation in flake thickness. 65 Neat 10 1 month 2 months 8 3 months 4 months 6 5 5 months 6 months J 4 2 0 0 500 1000 Particle size 2000 F 'µ µ * 5 771 1500 distribution of the neat and long4term Na4exchanged vermiculites Fig. 23 shows the particle size distribution for blank vermiculite and its ion exchanged forms. After ion exchange, the particle size distribution of vermiculites is relatively broadened. All ion exchanged samples show also more than 2% of particles with dimensions greater than 2000 em. Alkaline earth ion exchanged vermiculites have a size of ca. 1 002 em, while alkali metals ion exchanged vermiculites have a size of ca. 893 em. 66 10 8 BaVER 5 6 CaVER MgVER 4 J KVER 2 NaVER 0 0 500 1000 1500 F 'µ µ * 5 2000 Particle size distribution of vermiculite and its ion exchanged 791 forms ;1> L 5 ,3 ' + * 5 * 5 6 * According to Fig. 24, the FT4IR spectra of the Na4exchanged forms were almost identical to that of the neat vermiculite. All the IR spectra showed a strong band at 3420–3446 cm41, ascribed to the characteristic O–H stretching vibration of the hydration water and bound silanol groups (Hougardy , 1970; Farmer, 1974). Other strong bands were observed at around 1641–1646 cm41. They are due to the O–H bending vibrations of the hydration water molecules (Hougardy , 1970; Farmer, 1974; Van der Marel and Beutelspacher, 1976). 67 At 995–999 cm41 there is a strong band attributed to the Si–O–Si and Si4O–Al stretching vibrations (Farmer, 1974; Van der Marel and Beutelspacher, 1976). 6 months 5 months I 1 1 4 months 3 months 2 months 1 month Neat 4000 3600 3200 2800 2400 2000 1600 1200 800 400 / I +6 FT4IR plots of neat vermiculite and all its Na4exchanged forms, 7;1 obtained after immersion in saturated brine ,3 6 7 8 8 * All FT4IR spectra of the neat, alkali and alkaline metal ion exchanged vermiculites showed the structural O–H group’s stretching vibrations at 3709 cm41. The ammonium vermiculite (NH4VER) O–H stretching vibrations were shifted to 3697 cm41 since the position of this peak maximum is a function of the nature of the interlayer cation (Bradley and Serratosa, 1960; Hougardy , 1970; Farmer, 1974). 68 NH 4VER BaVER I 1 1 CaVER MgVER KVER NaVER Neat 4000 3600 3200 2800 2400 2000 1600 1200 800 400 +6 / I 7<1 FT4IR spectra of neat vermiculite, ammonium, alkali metal and alkaline metal ion exchanged vermiculites Characteristic vermiculite absorption bands were also observed in the results depicted in Fig. 25. Strong hydration H2O bands are noticeable at 3409– 3424 cm41 (O–H stretching vibrations and also bound silanol groups) and at 1638–1646 cm41 (O–H bending vibrations) (Hougardy , 1970; Farmer, 1974; Van der Marel and Beutelspacher, 1976). Well4defined Si–O–Si and Si–O–Al stretching vibrations are present at 996–997 cm41, as well as a medium4intensity band at ca. 456 cm41 attributed to the Si–O–Mg structural bending vibration (Farmer, 1974; Van der Marel and Beutelspacher, 1976). In addition, NH4VER presents bands at 3529 cm41, 3283 cm41, 3064 cm41 and 2835 cm41, as well as a well4resolved band at 1430 cm41 attributed to NH4+ vibration (Stone and Wild, 1978; Pérez4Rodríguez 2006). 69 , 2004, Petit , ) < &. 0$ &. The effect of inorganic modification on the properties of Palabora vermiculite was studied by scanning electron microscopy (SEM), X4ray fluorescence (XRF) spectroscopy, Fourier transform–infrared (FT4IR) spectroscopy, inductively coupled plasma – optical emission spectroscopy (ICP4 OES), X4ray diffraction (XRD) analysis and thermo4mechanical analysis (TMA). Modification of Palabora vermiculite was achieved by ion exchanging with ammonium and selected alkali metal and alkaline earth metal chlorides and by submerging the macroscopic flakes in saturated brine for periods of several months. In the long4term treated vermiculite, the sodium content reached a plateau value by the end of month one, but the XRD pattern continued to change, reaching a final form only after an exposure period of six months. The final XRD pattern indicated the presence of at least two distinct interstratified phases containing different amounts of vermiculite. All samples, prepared with different salts and/or by different methods, showed the typical d060 clay mineral reflection at 1.53 nm, confirming the preservation of the trioctahedral unit cell structure. Barium ions showed relatively high affinity to the vermiculite surface and ammonium ions were able to exchange potassium ions from biotite layers. Thermal expansion of the neat vermiculite commenced at temperatures above 450 °C and reached levels exceeding an eight4fold expansion at ca. 700 °C. Beyond this temperature a slight contraction occurred. It was confirmed that the blowing agent was the interlayer water in the neat, alkali metal and alkaline earth metal ion exchanged vermiculite samples. In the ammonium ion exchanged vermiculite, it was found that ammonia and water were the blowing agents and that they were simultaneously released during thermal degradation. However, ammonium4vermiculite did not show a 70 relatively better thermal expansion ratio, despite the presence of both ammonia and water in the interlayer. In the increasing ionic potential series: sodium4vermiculite, barium4 vermiculite, calcium4vermiculite and magnesium4vermiculite, the expansion onset temperature also increased dramatically. So, sodium, potassium and ammonium ion exchange lowered the expansion onset temperature to below 300 °C. It was the aim of this study to lower the expansion onset temperature of Palabora vermiculite to temperatures in the range of 200 °C to 350 °C and therefore this aim has been achieved. The simple process of exchanging the interlayer magnesium ions with the aforementioned ions of relatively low ionic potential lowered the exfoliation onset temperature to a range that may have technological significance, for instance in applications like intumescent fire barriers. 71 ) 8 ) ) ). ) Abate, G., dos Santos, L.B.O., Colombo, S.M., Masini, J.C., 2006. Removal of fulvic acid from aqueous media by adsorption onto modified vermiculite, 32, 261–270. Ametek, 2010. Spectro Genesis. Accessed in July 2010 at: www.ametek.com/press/press4display/genesis4eop4sop.aspx Annabi4Bergaya, F., 2008. Layered clay minerals: basic research and innovative composite applications. ! ! ! 107, 141–148. Anthony, J.W., Bideaux, R.A., Bladh, K.W., Nicholis, M.C., 2004a. Hydrobiotite. " Mineral Data Publishing. Accessed in July 2009 ! at: www.handbookofmineralogy.org/pdfs/vermiculite.pdf. Anthony, J.W., Bideaux, R.A., Bladh, K.W., Nicholis, M.C., 2004b. # " . Mineral Data Publishing. Accessed in July 2009 ! at: www.handbookofmineralogy.org/pdfs/vermiculite.pdf. Arab, M., Bougeard, D., Smirnov, K.S., 2002. Experimental and computer simulation study of the vibrational spectra of vermiculite. $ $ 4, 1957–1963. Aran, D., Maul, A., Masfaraud, J., 2008. A spectrophotometric measurement of soil cation absorbance. exchange % capacity based on cobaltihexamine chloride . 340, 865–871. Badreddine, R., Vandormael, D., Fransolet, A.4M., Long, G.J., Stone, W.E.E., Grandjean, F., 2002a. A comparative X4ray diffraction, Mössbauer and NMR spectroscopic study of vermiculites from Béni Bousera, Morocco and Palabora, Republic of South Africa. ! . 37, 367–376. Badreddine, R., Le Dred, R., Prost, R. 2002b. Far infrared study of K+, Rb+, and Cs+ during their exchange with N+ and Ca2+ in vermiculite. 71–31. 72 ! . 37, Badreddine, R., Vandormael, D., Fransolet, A.4M., Long, G.J., Stone, W.E.E., Grandjean, F., 2002c. A far infrared study of K+ ions during K+ exchange in vermiculite. Ca2+ . 37, 59–70. ! Bailey, G.W., White, J.L., 1964. Review of adsorption and desorption of organic pesticides by bioactivity. soil colloids, with implications concerning pesticide 12, 324–332. & Barrer, R.M., Klinowski, J., 1977. Theory of isomorphous replacement in aluminosilicates. $ 285, 637–676. ' Barshad, I., 1948. Vermiculite and its relation to biotite as revealed by base exchange reactions, X4ray analyses, differential thermal curves, and water content. ! 33, 655–678. Barshad, I., 1950. The effect of the interlayer cations on the expansion of the mica type of crystal lattice. ! 35, 225–238. Barshad, I., 1954a. Cation exchange in micaceous minerals: I. Replaceability of the interlayer cations of vermiculite with ammonium and potassium ions. 77, 463–472. Barshad, I., 1954b. Cation exchange in micaceous minerals: II. Replaceability of ammonium and potassium from vermiculite, biotite and montmorillonite. 78, 57–76. Bascomb, C.L., 1964. Rapid method for the determination of the cation exchange capacity of calcareous and non4calcareous soils. ( & . 15, 821– 823. Basset, W.A., 1963. The geology of vermiculite occurrences. $ ! )* , New York: Pergamon Press, pp 61–96. Baumeister, W., Hahn, M., 1976. An improved method for preparing single crystal specimen supports: H2O2 exfoliation of vermiculite ! 7, 247– 251. Bergaya, F., Theng, B.K.G., Lagaly, G. (Eds), 2006. " Developments in Clay Science Series, Volume 1, Amsterdam: Elsevier. 73 . Beyer, J., Reichenbach, H.G.V., 1998. Dehydration and rehydration of vermiculites: IV. Arrangements of interlayer components in the 1.43 nm and 1.38 nm hydrates of Mg4vermiculite. + $ 207, 67–82. Boss, B.D., 1967. Differential thermal analysis of biotitic vermiculite to determine vermiculite content. ! . 52, 294–298. Bougeard, D., Smirnov, K., 2007. Modelling studies of water in crystalline nanoporous aluminosilicates. $ $ . 9, 2264245. Bradley, W.F., Serratosa, J.M., 1960. A discussion of the water content of vermiculite. ! 7, 260–270. Brindley, G.W., Brown, G. 1980. Crystal structures of clay minerals and their X4 ray identification. London: Mineralogical Society. Brindley, G.W., Zalba, P.E., Bethke, C.M., 1983. Hydrobiotite, a regular 1:1 interstratification of biotite and vermiculite layers. ! . 68, 420– 425. Brown, M.E., 2001. , ' ' - 2nd ed. Dordrecht: Kluwer Academic. Cardille, C.M., Slade, P.G., 1987. Structural study of a benzidine4vermiculite intercalate having a high tetrahedral4iron content by spectroscopy. ! 57Fe Mössbauer . 35, 203–207. Carter, R.C., Wilde, P., 1972. Cation exchange capacity of suspended material from coastal sea water off central California. ! % 13, 107–122. Ciesielski, H., Sterckeman, T., Santerne, M., Willery, J.P., 1997. Determination of cation exchange capacity and exchangeable cations in soils by means of cobalt hexamine trichloride. Effects of experimental conditions. 17, 1–7. Connell J.E, Kendrick D., Marks G., Parsonage J.R., Thomas M.J.K., Vidgeon E.A., 1994. Synthesis of QMD and QD polyorganosiloxanes from 74 . tetrakis(trimethylsiloxy)silane and Palabora vermiculite. ( ! 4, 3994406. Connell, J.E., Metcalfe, E., Thomas, M.J.K., 2000. Silicate–siloxane fire retardant composites. $ 49, 1092–1094. , Cordier, D.J., 2010. Vermiculite . Survey, 180–181. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Couderc, P., Douillet, Ph., 1973. Les vermiculites industrielles: exfoliation, caractéristiques minéralogiques et chimiques. / Cullity, B.D., 1978. 01 2 . 99, 51–59. & . Reading: Addison4Wesley. Da Fonseca, M.G., Cardoso, C.M., Wanderley, A.F., Arakaki, L.N.H., Airoldi, C., 2006. Vermiculite as a useful host for guest cyclic aliphatic amine intercalation, followed by cation adsorption. 280, 39–44. Da Fonseca, M.G., de Oliveira, M.M., Arakaki, L.N.H., Espinola, J.G.P., Airoldi, C.A., 2005. Natural vermiculite as an exchanger support for heavy cations in aqueous solution. ( . 285, 50–55. , Da Silva Jr., U.G., Melo, M.A.F., da Silva, A.F., de Farias, R.F., 2003. Adsorption of crude oil on anhydrous and hydrophobised vermiculite. ( 260, 302–304. , Das, N.C., Bandyopadhyay, M., 2001. Removal of lead by vermiculite medium. . 6, 221–231. De Haro, M.C.J., Pérez4Rodríguez, J.L., Poyato, J., Pérez4Maqueda, L.A., Ramírez4Valle, V., Justo, A., Lerf, A., Wagner, F.E., 2005. Effect of ultrasound on preparation of porous materials from vermiculite. 30, 11–20. Del Rey4Perez4Caballero, F.J., Poncelet, G., 2000. Microporous 18 Å Al4pillared vermiculites: preparation and characterisation. ! 37, 313–327. 75 ! ! . Dohrmann, R., 2006a. Cation exchange capacity methodology II: A modified silver4thiourea method. 34, 38–46. Dohrmann, R., 2006b. Cation exchange capacity methodology III: Correct exchangeable calcium determination of calcareous clays using a new silver– 34, 47–57. thiourea method. Eaves, D., 2006. Sampling and analysis of crude vermiculite samples for possible asbestiform fibre and quartz content. IOM Consulting Report 609401591, pp 1–6. Evans, A.M., 1987. , 3 , 2nd ed. Oxford: Blackwell % Scientific, pp 112–114. Evans, A.M., 1993. Ore geology and industrial minerals: An introduction, 3rd ed. Oxford: Blackwell Scientific, pp 114–115. Farmer, V.C., 1971. The characterization of adsorption bonds in clays by infrared spectroscopy. 112, 62–68. Farmer, V.C., 1974. The infrared spectra of minerals. Monograph 5, London: Mineralogical Society., Ferrage, E., Tournassat, C., Rinnert, E., Lanson, B., 2005. Influence of pH on the interlayer cationic composition and hydration state of Ca4montmorillonite: Analytical chemistry, chemistry, chemical modeling and XRD profile modeling study. % . 69, 2797–2812. Frank, D., Edmond, L., 2001. Feasibility for identifying mineralogical and geochemical tracers for vermiculite ore deposits. Seattle: United States Environmental Protection Agency. Friedman, S.D., McKinney, R.W., Ou, C., Spotnitz, R.M., Wu, S., 1994. Vermiculite composition with improved thermal expansion properties. . $ 5340558. Fripiat, J.J., Chaussidon, J., Touillaux, R., 1960. Study of dehydration of montmorillonite and vermiculite by infrared spectroscopy. ( $ . . 64, 1234–1241. 76 1 Gabbott, P. (Ed), 2008. $ . Oxford: ' Blackwell. Gast, R.G., Klobe, W.D., 1971. Sodium4lithium exchange equilibria on vermiculite at 25° and 50°C. 19, 311–319. ! Ghabru, S.K., Mermut, A.R., Arnaud, R.J., 1989. Layer4charge and cation4 exchange characteristics of vermiculite (weathered biotite) isolated from a gray luvisol in Northeastern Saskatchewan. ! 37, 164– 172. Gier, S., Johns, W.D., 2000. Heavy metal adsorption on micas and clay minerals studied by X4ray photoelectron spectroscopy. Gomes, E.V.D., Visconte, L.L.Y., Pacheco, 16, 289–299. E.B.A., 2009. Thermal characterization of polypropylene/vermiculite composites. ( ' 97, 571–575. Grim, R.E., 1968. 2nd ed. New York: McGraw4Hill. ! Gruner, J.W., 1934. The structures of vermiculites and their collapse by dehydration. ! . 19, 557–575. Haines, P.J. (Ed), 2002. $ . ' Cambridge: The Royal Society of Chemistry. Harben, P.W., Roberts, R.L., 1990. Industrial minerals geology and world deposits. London: Metal Bulletin Plc. Harland, C.E., 1994. , ' $ , 2nd ed. Cambridge: The Royal Society of Chemistry. Hendershot, W.H., Duquette, M., 1986. A simple barium chloride method for determining cation exchange capacity and exchangeable cations. (. 50, 605–608. Hendricks, S.B., Jefferson, M.E., 1938. Crystal structure of vermiculites and mixed vermiculite4chlorites. ! 77 . 23, 851–862. Herbert, H.J., Kasbohm, J., Moog, H.C., Henning, K.H., 2004. Long4term behaviour of the Wyoming bentonite MX480 in high saline solutions. 26, 275–291. Herbert, H.J., Moog, H.C., 1999. Cation exchange, interlayer spacing and water content of MX480 bentonite in high molar saline solutions. 54, % 55–65. Hindman, J.R., 2006. Vermiculite. In: Kogel, J.E., Trivedi, N.C., Krukowski, S.T. (Eds), , ! ! . Colorado: Society for Mining, Metallurgy, and Exploration, pp 1015–1027. Holešová, S., Valaškova, M., Plevová, E., Pazdziora, E., Matejová, K., 2010. Preparation of novel organovermiculites with antibacterial activity using chlorhexidine diacetate. ( 342, 593–597. , Horacek, H., Pieh, S., 2000. The importance of intumescent systems for fire protection of plastic materials. $ , 49, 1106–1114. Hougardy, J., Serratosa, J.M., Stone, W., Van Olphen, H., 1970. Interlayer water in vermiculite: Thermodynamic properties, packing density, nuclear pulse resonance, and infra4red absorption. 2 1, 187– & 193. Hudson, E.A., Terminello, L.J., Viani, B.E., Denecke, M., Reich, T., Allen, P., Bucher, J.J., Shuh, D.K., Edelstein, N.M., 1999. The structure of U6+ sorption complexes on vermiculite and hydrobiotite. ! . 47, 439–457. Inoue, A., 1984. Thermodynamic study of Na4K4Ca exchange reactions in vermiculite. Jenkins, R., 1988. 01 ! & . 32, 311–319. . New York: Wiley. Justo, A., 1989. Expansibility of some vermiculites. 4, 509–519. Justo, A., Pérez4Rodríguez, J.L., Sánchez4Soto, P.J., 1993. Thermal studies of vermiculites and mica4vermiculite interstratifications. ( ' 59–65. 78 40, Kahr, G., Madsen, F.T., 1995. Determination of the cation exchange capacity and the surface area of bentonite, illite and kaolinite by methylene blue adsorption. . 9, 327–339. Kodama, T., Komarneni, S., 1999. Alkali metal and alkaline earth metal ion exchange with Na444mica prepared by a new synthetic route from kaolinite. . 9, 2475–2480. ( ! Konta, J., 1995. Clay and man: Clay raw materials in the service of man. 10, 275–335. Kresten, P., Berggren, G., 1978. The thermal decomposition of vermiculite. 23, 171–182. ' Lagaly, G., 1979. The “layer charge” of regular interstratified 2:1 clay minerals. ! . 27, 1–10. Lagaly, G., 1982. Layer charge heterogeneity in vermiculites. ! . 30, 215–222. Langer, R.L., Marlor, A.J., 1981. Intumescent sheet material1 . $ 4305992. Lawes, G., 1987. ! 01 ! . Chichester: Wiley. Liu, B., Ding, Q., Zhang, J., Hu, B., Shen, J., 2005. Preparation and properties of new EPDM/vermiculite nanocomposites. $ . 26, 706–712. Liu, Y., Xiao, D., Li, H., 2007. Kinetics and thermodynamics of lead (II) adsorption on vermiculite. ' . 42, 185–202. Lopez4Gonzalez, J.D.D., Cano4Ruiz, J., 1957. Surface area changes of a vermiculite by acid and thermal treatment. ! . 6, 399–405. MacEwan, D.M.C., Ruiz4Amil, A., Brown, G., 1961. Interstratified clay minerals. In: Brown, G. (Ed.), ' ! 01 , , London: Mineralogical Society, pp 393–445, 79 Marcos, C., Arango, Y. C., Rodríguez, I., 2009. X4ray diffraction studies of the thermal behaviour of commercial vermiculites. 42, 368–378. Marcos, C., Argüelles, A., Ruiz4Conde, A., Sánchez4Soto, P.J., Blanco, J.A., 2003. Study of the dehydration process of vermiculites by applying a vacuum pressure: Formation of interstratified phases. ! . 67, 1253– ! 1268. Marcos, C., Rodríguez, I., 2010. Expansion behaviour of commercial vermiculites . 48, 492–498. at 1000 °C. Martynková, G.S., Valášková, M., Čapková, P., Matějka, V., 2007. Structural ordering of organovermiculite: Experiments and modeling. ( , 313, 281–287. Mathieson, A.M., 1958. Mg4vermiculite: A refinement of the crystal structure of the 14.36 Å phase. 43, 216–227. ! Mekhamer, W.K., Assaad, F.F., 1999. Thermodynamics of Sr4Mg vermiculite exchange and the effect of PVA on Mg release. ' 334, 33– 38. Mermut, A., Lagaly, G., 2001. Baseline studies of the Clay Minerals Society source clays: Layer4charge determination and characteristics of those minerals containing 2:1 layers. ! . 49, 393–397. Mokhonoana, M.P., 1996. A study of the surface and adsorption properties of Palabora vermiculite. MSc Dissertation, Johannesburg: University of the Witwatersrand. Moore, D.M., Reynolds Jr., R.C., 1989. 01 2 , . New York: Oxford University Press. ! Newman, A.C.D. (Ed.), 1987. ! London: Longman Scientific & Technical Mineralogical Society. Obut, A., Girgin, I., 2002. Hydrogen peroxide exfoliation of vermiculite and phlogopite. ! 15, 683–687. 80 Okhotnikov, V.B., Babicheva, I.P., Musicantov, A.V., Aleksandrova, T.N., 1989. Thermal decomposition of materials with layered structures: Isothermal dehydration of vermiculite single crystals in vacuum. 7, 273– 287. Okui, H., Omura, Y., Wada, N., Kamitakahara, W.A., Fujiwara, A., Suematsu, H., Murakami, Y., 1998. Structural, lattice4dynamical and magnetic properties of alkali4metal intercalated vermiculite. ! 4- 311, 339–344. Ou, C., 1992. Vermiculite dispersions and method of preparing same. . $ 5102464. Ou, C., Bablouzian, L., 1994. Intumescent sheet material. . $ 5340643. Ou, C., Yang, J.C., 1987. Vermiculite dispersions and method of preparing same. . $ 4655842. Palabora Mining Company Limited, Mine Geological and Mineralogical Staff, 1976. The geology and the economic deposits of copper, iron, and vermiculite in the Palabora Igneous Complex: a brief review. % 71, 177–192. Peralta, M.M.C., 2009. Tratamento químico de uma vermiculite visando seu uso em compósitos de polipropileno. MSc Dissertation, São Paulo: Universidade de São Paulo. Pérez4Maqueda, L.A., Balek, V., Poyato, J., Pérez4Rodriquez, J. L., Šubrt, J., Bountsewa, I. M., Beckman, I. N., Málek, Z., 2003. Study of natural and ion exchanged vermiculite by emanation thermal analysis, TG, DTA and XRD. ( ' 71, 715–726. Pérez4Maqueda, L.A., Poyato, J., Pérez4Rodriquez, J. L., 2004. Effect of sonication on the thermal decomposition of ammonium vermiculite as studied by TG4MS hyphenated technique. ( ' 383. 81 78, 375– Pérez4Rodríguez, J.L., Poyato, J., de Haro, M.C.J., Pérez4Maqueda, L.A., Lerf, A. 2004. Thermal decomposition of NH4+4vermiculite from Santa Olalla (Huelva, Spain) and its relation to the metal ion distribution in the octahedral sheet. $ 31, 415–420. ! Petit, S., Righi, D., Madejová, J., 2006. Infrared spectroscopy of NH4+4bearing and saturated clay minerals: A review of the study of layer charge. 34, 22–30. Plachá, D., Martynková, G.M., Rümmeli, M.H., 2008. Preparation of organovermiculites using HDTMA: structure and sorptive properties using naphthalene. ( 327, 341–347. , Potter, M.J., 1996. Vermiculite. . Survey, 184–185. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Potter, M.J., 1997. Vermiculite. . Survey, 186–187. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Potter, M.J., 1998. Vermiculite. . Survey, 186–187. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Potter, M.J., 2000. Vermiculite. . Survey, 186–187. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Potter, M.J., 2002. Vermiculite. . Survey, 184–185. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Potter, M.J., 2005. Vermiculite. . Survey, 184–185. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. 82 Potter, M.J., 2007. Vermiculite. . Survey, 186–187. , Mineral Commodity % Accessed in July 2010 at: http://minerals.usgs.gov/minerals/pubs/commodity/vermiculite/index.html. Qiu, Z., Zhang, J., Zhou, Y., Song, B., Chang, J., Yang, K., Wang, Y., 2009. Biodegradable poly(p4dioxanone) reinforced and toughened by organo4 modified vermiculite. $ DOI: 10.1002/pat.1606. ' Raman, K.V.V., Jackson, M.L., 1963. Vermiculite surface morphology. $ 6 ! Raupach, M., 1988. H.G., 121, 446–475. , Beyer, J., 1994. Dehydration vermiculites: I. Phlogopitic Mg4vermiculite. Reichenbach, H.G., Beyer, J., 1995. H.G., Beyer, J., 1997. ! Dehydration vermiculites: II. Phlogopitic Ca4vermiculite. Reichenbach, , Volume 2, 423–429. Polarized infrared study of aluminium4vermiculite intercalate. ( Reichenbach, )5 ! Dehydration vermiculites: III. Phlogopitic Sr4 and Ba4vermiculite and rehydration of 29, 327–340. and rehydration of 30, 273–286. and rehydration ! of 29, 327– 340. Reichle, W.T., 1986. Synthesis of anionic clay minerals (mixed metal hydroxides, hydrotalcite). 22, 135–141. , Reynolds Jr., R.C., 1985. NEWMOD©. A computer program for the calculation of one4dimensional diffraction patterns of mixed4layered clays. R.C. Reynolds Jr., 8 Brook Rd., Hanover, NH, USA. Rinaudo, C., Roz, M., Boero, V., Franchini4Angela, M., 2004. FT4Raman spectroscopy on several di4 and tri4octahedral T4O4T phyllosilicates. 6 ( ! ! 12, 537–554. Ruiz4Amil, A., de la Cruz, F.A., Ruiz4Conde, A., 1992. Study of a material from Libby, Montana, containing vermiculite and hydrobiotite: Intercalation with aliphatic amines. ! 27, 257–263. 83 Saehr, D., Le Dred, R., Baron, J., 1990. Ion exchange K+ Na+ and availability of two groups of exchange sites in a vermiculite: I. Providing evidence from two groups. % 86, 155–163. ! Saehr, D., Le Dred, R., Baron, J., 1992. K+ Na+ ion exchanges and availability of two groups of exchange sites in a vermiculite: II. Variability of the equivalent fraction of the two groups. % 45, 109–119. / Schoeman, J.J., 1989. Mica and vermiculite in South Africa. ( ! , ! 89, 1–12. Schollenberger, C.J., Simon, R.H., 1945. Determination of exchange capacity and 59, 13–24. exchangeable bases in soil4ammonium acetate method. Schwellnus, C.M., 1938. Vermiculite deposits in the Palabora area. 6 ' / . 11, 7–28. % Serratosa, J.M., Johns, W.D., Shimoyama, A., 1970. I.R. study of alkyl4 ammonium vermiculite complexes. ! 18, 107–113. Shahwan, T., Erten, H.N., 2004. Temperature effects in barium sorption on natural kaolinite and chlorite4illite clays. ( 6 260, 43– 48. Skipper, N.T., Soper, A.K., McConnell, D.C., 1991. The structure of interlayer water in vermiculite. ( 94, 5751–5760. $ Stone, M.H., Wild, A., 1978. The reaction of ammonia with vermiculite and hydrobiotite. ! 13, 337–350. Suquet, Y., Franck, R., Lambert, J.F., Elsass, F., Marcilly, C., Chevalier, S., 1994. Catalytic properties of two pre4cracking matrices: a leached vermiculite and a Al4pillared saponite. 8, 349–364. Suzuki, S., Sazarashi, M., Akimoto, T., Haginuma, M., Suzuki, K., 2008. A study of the mineralogical alteration of bentonite in saline water. 41, 190–198. 84 Symons, M.W., 1999. Method of preparing an exfoliated vermiculite for the manufacture of finished product. . $ Tan, K.H., 1996. Soil 5879600. . New York: Marcel $ Dekker, pp 114–126, 285–294, 308–330. The Vermiculite Association, environmental aspects. 2000. Vermiculite: Accessed in Health, safety July 2010 and at: http://www.vermiculite.org. Thomas Jr., J., Bohor, B.F., 1969. Surface area of vermiculite with nitrogen and carbon dioxide as adsorbates. Tjong, S.C., 2006. Structural nanocomposites. ! 17, 205–209. ! and mechanical properties of polymer R 53, 73–197. Tomanec, R., Popov, S., Vučinič, D., Lazič, P., 1997. Vermiculite from Kopaonik (Yugoslavia): Characterization and processing. & 7 7 $ 31, 247–254. ! Vaccari, A., 1998. Preparation and catalytic properties of cationic and anionic clays. ' 41, 53–71. Van der Marel, H.W., Beutelspacher, H., 1976. , . Amsterdam: Elsevier. ! Vidal, O., Dubacq, B., 2009. Thermodynamic modelling of clay dehydration, stability and compositional evolution with temperature, pressure and H2O activity. % 73, 6544–6564. Wada, T., 1973a. Manufacture of expanded vermiculite employing a urea compound and low temperatures. . $ 3753923. Wada, T., 1973b. Method for the expansion of vermiculite. . $ 3758415. Walker, G.F., 1959. Diffusion of exchangeable cations in vermiculite. 6 1392–1393. 85 184, Walker, G.F., 1961. Vermiculite minerals. In: Brown, G. (Ed.), ' , . ! 01 London: Mineralogical Society, pp 297–342. Wild and Keay, 1964. Cation exchange equilibria with vermiculite. ( 15, 1354144. Wilson, M.J., 1994. ! ! 2 . London: Chapman & Hall. Xu, J., Li, R.K.Y., Yu, Y., Li, L., Meng, Y.Z., 2005. Preparation of poly(propylene carbonate)/organo4vermiculite intercalation. $ nanocomposites via direct melt ( 41, 881–888. Yu, X., Wei, C.H., Ke, L., Hu, Y., Xie, X., Wu, H., 2010. Development of organovermiculite4based adsorbent for removing anionic dye from aqueous solution. ( " 7 ! 180, 499–507. Zhu, R., Zhu, L., Zhu, J., Xu, L, 2008. Structure of cetyltrimethylammonium intercalated hydrobiotite, 42, 224–231. 86 ).- 87 ) 55 5 4 1 5 4 , H ' E Sodium chloride * ' E0* ' * NaCl 58.44 35.9 53.85 Potassium chloride KCl 74.55 34.4 51.6 Magnesium chloride MgCl2 95.21 54.3 81.45 Calcium chloride CaCl2 110.98 74.5 111.75 Barium chloride BaCl2 208.23 35.8 53.7 Ammonium chloride NH4Cl 53.49 37.2 55.8 Saturated solutions of the respective chlorides were preferred and prepared since the vermiculite selectivity coefficient rises with an increase in the selected ion loading. All experiments were run at the pH of the salt solution since the rate of exchange is not affected by the pH in the range of 4 to 9 (Wild and Keay, 1964; Gast and Klobe, 1971). 88 55 4 1 B 7==? ' 5 , I 6??8L7==? 89 6??; I 7=6=* 55 4 5 1 G - . + I5 5 4 , 5 50 45 40 35 I 6=9 30 25 NaVER6 20 NaVER 15 10 5 0 2 4 6 8 10 2θ, °( 12 14 16 18 20 22 α) NaVER6 is the sample prepared by long4term treatment for up to six months. This reaction was performed in closed plastic boxes containing vermiculite submerged in a large volume of saturated sodium chloride solution. The solution was agitated daily and every fortnight approximately 1 000 ml of the supernatant liquid was exchanged with fresh saturated brine. NaVER was prepared in 14 days using an orbital shaker. The reaction was performed in a closed glass container with saturated sodium chloride solution. 90 5 4 B1 3 3 4 , BaVER I 1 1 55 CaVER MgVER KVER NaVER NeatVER 100 200 300 400 500 600 700 800 900 1000 1100 / I 91 +6 55 4 B1 ) ) 5 5, + , 45 4 ! , 55 4 B+ * Just before exfoliation, which commenced at 263 °C 92 , 55 4 B+ * During exfoliation at 345 °C 55 4 B+ * During exfoliation at 367 °C 93 55 4 B+ * After exfoliation at 432 °C 94 55 4B 1 B , 95 , 55 4B 1 & G - 3 3 4 , 0 ' * 40000 30000 20000 10000 0 2 10 Blank 4 Step: 0.009 ° 4 Step time: 744. s Operations: Smooth 0.050 d=27.05586 d=14.31312 d=12.38420 d=11.92897 d=10.07959 d=8.41666 Blank 7θ θI K ' α* d=4.91554 d=4.59902 20 d=3.47944 d=3.44143 55 4 B + * XRD pattern of blank vermiculite 96 0 ' * 30000 20000 10000 0 2 10 NaVER415S 4 Step: 0.009 ° 4 Step time: 744. s Operations: Smooth 0.050 d=26.85258 d=14.56414 d=11.92892 d=11.37619 d=10.11789 d=8.36558 NaVER 7θ θI K ' α* d=4.96566 d=4.59909 20 d=3.44143 d=3.34118 55 4 B + * XRD pattern of Na4vermiculite 97 0 ' * 30000 20000 10000 0 2 10 MgVER415S 4 Step: 0.009 ° 4 Step time: 744. s Operations: Smooth 0.050 d=28.35963 d=14.29710 d=12.46447 d=11.95378 d=10.09096 d=8.30542 MgVER 7θ θI K ' α* d=4.92205 d=4.59987 20 d=3.47970 d=3.44535 55 4 B + * XRD pattern of Mg4vermiculite 98 0 ' * 30000 20000 10000 0 2 KVER-15S - Step: 0.009 ° - Step time: 744. s Operations: Smooth 0.050 | Import d=11.02182 d=10.10470 10 KVER 7θ θI K ' α* d=4.99319 d=4.60248 20 d=4.25227 d=4.06457 d=3.43956 d=3.40759 55 4 B + * XRD pattern of K4vermiculite 99 CaVER d=4.59977 20 d=4.24965 d=3.44055 d=3.33983 40000 d=10.07634 CaVER415S 4 Step: 0.009 ° 4 Step time: 744. s d=11.99331 Operations: Smooth 0.050 10 α* 100 30000 20000 10000 0 2 7θ θI K ' 4 B + * XRD pattern of Ca4vermiculite 55 * ' 0 BaVER 15S - Step: 0.009 °- Step time: 744. s Operations: Smooth 0.050 | Import 0 2 10 20 7θ θI K ' α* 101 4 B + * XRD pattern of Ba4vermiculite 10000 55 d=4.60078 0 ' d=10.10981 20000 d=3.43861 BaVER d=11.16686 * 30000 55 4 B + * XRD pattern of NH44vermiculite 102 55 4B , 1 , Muiambo, H.F., Focke, W.W., Atanasova, M., Van der Westhuizen, I., Tiedt, L., 2010. Thermal properties of sodium4exchanged Palabora vermiculite. 50, 51–57. 103