Porosity and distribution of water in perlite from the island of Milos, Greece
© Kaufhold et al.; licensee Springer. 2014
Received: 1 October 2014
Accepted: 6 October 2014
Published: 12 October 2014
A perlite sample representative of an operating mine in Milos was investigated with respect to the type and spatial distribution of water. A set of different methods was used which finally provided a consistent view on the water at least in this perlite. Infrared spectroscopy showed the presence of different water species (molecular water and hydroxyl groups / strongly bound water). The presence of more than 0.5 mass% smectite, however, could be excluded considering the cation exchange capacity results. The dehydration measured by thermal analysis occurred over a wide range of temperatures hence confirming the infrared spectroscopical results. Both methods point to the existence of a continuous spectrum of water binding energies. The spatial distribution of water and/or pores was investigated using different methods (CT: computer tomography, FIB: scanning electron microscopy including focused ion beam technology, IRM: infrared microscopy). Computer tomography (CT) showed large macropores (20 – 100 μm) and additionally revealed a mottled microstructure of the silicate matrix with low density areas up to a few μm in diameter. Scanning electron microscopy (FIB) confirmed the presence of μm sized pores and IRM showed the filling of these pores with water. In summary, two types of pores were found. Airfilled 20 – 100 μm pores and μm-sized pores disseminated in the glass matrix containing at least some water. Porosity measurements indicate a total porosity of 26 Vol%, 11 Vol% corresponding to the μm-sized pores. It remains unsolved wether the water in the μm-sized pores entered after or throughout perlite formation. However, the pores are sealed and no indications of cracks were found which indicated a primary source of the water, i.e. water was probably entrapped by quenching of the lava. The water in these pores may be the main reason for the thermal expandability which results in the extraordinarily porous expanded perlite building materials.
KeywordsPerlite Porosity Water distribution Volcanic glass FIB CT-analysis
Perlite is a hydrous volcanic material dominated by alumosilicate glass. As an industrial material, perlite is mostly used in its expanded form, i.e. after heat treatment resulting in a light-weight macroporous product. Most of the expanded perlite is used for building construction in plasters, mortars, and tiles. Minor components of perlite are phenocrysts or microlites which formed before eruption of the magma (e.g. feldspar or biotite). Technically, the term perlite is used for glassy volcanic rock which can be thermally expanded to about 20 times of its volume (Koukouzas et al. 2000). Scientifically the term perlite is used for hydrated volcanic glass. The most common volcanic glass is obsidian. The water content of typical obsidian is about 0.1 mass%. Larger water contents of obsidians mostly result from post-emplacement secondary hydration. Perlites, in contrast, contain up to 5 mass% water. They are believed to form upon hydration of volcanic glass (Ross and Smith 1955). Classical perlites (round particles with an onion like appearance) (Lorenz and Gwosdz 2000) are distinguished from banded perlites (Allen 1988). The water required for perlite formation (glass hydration) is supposed to enter through small cracks present in the volcanic glass. Diffusion of water into the glass may cause the cracks or crack formation facilitates water diffusion (Marshall 1961; Friedmann et al. 1966; Denton et al. 2009). At the wall of the cracks the glass dissolves and smectite crystallizes as alteration proceeds (Denton et al. 2009). The typical hydration shells around the primary particles are about 20 μm in size (Friedmann et al. 1966). The smectite at the walls would explain the presence of both, hydroxyls and molecular water in the perlite. Keller & Picket (1954) detected hydroxyls as well as hydrogen bound water using infrared (IR) spectroscopy. According to Friedmann et al. (1966) both primary magmatic and meteoric water causing the post-formational hydration of the glass can be found in perlites. This water, in contrast to the hydration shell, was supposed to be located within the Al-Si-framework. The difficulty of distinguishing different types of perlite water also results from the fact that the primary water of the volcanic glass varies in both amount and composition, with variable mixtures of hydroxyls and molecular water (Stolper 1982; Eckert et al. 1988; Dobson et al. 1989; Silver et al. 1990; Pandya et al. 1992).
The water in perlites can be measured by IR spectroscopy. Dobson et al. (1989) found different binding energies which could be correlated with IR stretching vibration. The amount of water is measured e.g. considering the 3550 cm-1 vibration (e.g. (Nichols et al. 2002)). Also NIR spectroscopy can be used (Stolper 1982). Differential thermal analysis (DTA) is commonly used for distinguishing hydroxyls and molecular water in clay minerals. However, few studies use DTA for hydrated volcanic glasses. One DTA curve was published by Tazaki et al. (1992) but the almost continuous mass decrease was difficult to interpret.
Few studies about the spatial distribution of water in perlites were published. Tazaki et al. (1992) studied freshly hydrated volcanic glasses with TEM and described spherical structures containing hydroxyls and concluded that these spherical structures could be precursors to the formation of clay minerals. Wysoczanski and Tani (2006) published IR image analysis based on the 3550 cm-1 vibration. They found that the glass particles all contained some water whereas low water domains corresponded to phenocrysts. However, they could not reveal the water distribution within the glass.
The present study, therefore, was conducted to improve the understanding of the spatial distribution of water in perlites. Knowledge about the spatial distribution would improve the understanding of the thermal expansion of technically produced expanded perlite.
Both, bulk methods to characterize the binding of the water (infrared spectroscopy, IR, and differential thermal analysis combined with a mass spectrometer, DTA-MS) were used along with three different 2D or 3D methods (computer tomography, CT, infrared microscopy, IRM, and focused ion beam technology in combination with a scanning electron microscope, FIB).
The perlite investigated in the present study was from the most important European perlite mines located in Milos, Greece. Most of the European perlite is derived from the Greek islands of Milos, Kimolos, and Kos (Koukouzas et al. 2000). From the middle part of the Upper Pliocene to the late Quarternary, Milos was more or less continuously affected by extensive volcanism. The perlites are supposed to result from the youngest volcanic activity in Milos which in contrast to the older ones is supposed to result from relatively shallow magma chambers (Fytikas et al. 1986). This perlite is characterized as calc-alkaline rhyolite with commonly more than 85 mass% glass and some phenocrysts. According to Koukouzas and Dunham (1994) different textural types can be distinguished: pumiceous perlite (light-weight and frothy features, only present in mines), hard perlite (dense and reddish and only appearing in the Kerdari Cap), and classical perlite (dense and mostly in contact with the hard and the pumiceous perlite in the mines). Much is known about the geological history of Milos and some information about compositional variation of Milos perlites and their applicability is available (Koukouzas et al. 2000) but none about the spatial distribution of the water within the perlite.
2 Materials and methods
The perlite investigated in the present study was collected in a running perlite mine in Milos (Greece). Three different specimens representing three different regions in the outcrop were collected. They differed with respect to their color (reddish, cream, white). However, no difference in water content or type of water (measured with DTA-MS and IR) was detected. Therefore, each of them would have been suitable for studying the spatial distribution of water of perlite of this deposit. One of the cream white particles (about 3 cm3) was selected for producing a polished thin section. This particle (the part not used for the production of the thin section) was further used for the investigation of the spatial distribution of water. For the computer tomography (CT) investigation a small piece (ca. 1 mm3) was cut off. In addition an expanded 1 mm3 sized perlite specimen (again about 1 × 1 × 1 mm = 1 mm3) that was taken from an industrially produced product (thermally expanded perlite) was investigated with CT.
For the characterization of the bulk material, three particles with the different colors were ground together to represent the entire outcrop.
The methods used were selected to i) characterize the bulk material and ii) to gather information about the spatial distribution of pores and or water. The latter was thought to be difficult because of expectedly small pores and/or water domains. Therefore, apart from light microscopy, methods with higher resolution and different strengths were used.
Bulk material characterisation
X-ray diffraction (XRD) patterns were recorded using a PANalytical X’Pert PRO MPD Θ-Θ diffractometer (Cu-Kα radiation generated at 40 kV and 30 mA), equipped with a variable divergence slit (20 mm irradiated length), primary and secondary soller slits, a Scientific X’Celerator detector (active length 0.59°) and a sample changer (sample diameter 28 mm). The samples were investigated from 2° to 85° 2Θ with a step size of 0.0167° 2Θ and a measuring time of 10 sec per step. For specimen preparation, the top loading technique was used and quantification performed based on Kaufhold et al. (2010).
The chemical composition of powdered samples was determined using a PANalytical Axios. Samples were prepared by mixing with a flux material (Lithiummetaborate Spectroflux, Flux No. 100A, Alfa Aesar) and melting into glass beads. The beads were analyzed by wavelength dispersive X-ray fluorescence spectrometry (WD-XRF). To determine loss on ignition (LOI) 1000 mg of sample material was heated to 1030°C for 10 min.
For measuring mid (MIR) infrared spectra the KBr pellet technique (1 mg sample/200 mg KBr) was applied. Spectra were collected on a Thermo Nicolet Nexus FTIR spectrometer (MIR beam splitter: KBr, detector DTGS TEC). The resolution was adjusted to 2 cm-1.
Thermoanalytical investigations were performed using a Netzsch 449 F3 Jupiter thermobalance equipped with a DSC/TG sample holder linked to a Netzsch QMS 403 C Aeolus mass spectrometer (MS). 100 mg of powdered material previously equilibrated at 53% relative humidity (RH) was heated from 25–1100°C with a heating rate of 10 K/min.
The cation exchange capacity (CEC) was measured using the Cu-Triethylenetetramine method (Meier and Kahr 1999; Kaufhold and Dohrmann 2003). A sample mass of 0.3 and 0.4 g was used to increase the smectite detection limit up to about 0.5 mass%.
The porosity was determined based on measuring the particle or specific density (AccuPyc 1330 of micromeritics using He) and bulk or envelope density (micromeritics GeoPyc 1360 using a free-flowing, finely divided, dry powder (DryFlow®) as the fluid medium instead of a liquid with a lower diameter limit of 50 μm). Different types of samples were investigated: powder, 1–2 mm particles, >2 mm, and after melting the sample (1150°C for 6 hours).
Methods to gather spatial information
Water can be detected by its characteristic infrared vibrations and the spatial distribution can be probed by IR-microscopy. In the present study a Thermo Nicolet Continuum FT-IR microscope was used. A freshly broken even surface of the perlite was fixed on the x-y stage. The following experimental conditions were selected: beam splitter CaF2, detector MCT/A, aperture 5, spectroscopic range of each spectrum 1000 – 4000 cm-1 with a resolution of 4 cm-1 and 16 scans each, and the step width was 3 μm. Because of the typically low intensity at large wave length the water deformation band at about 1635 cm-1 was selected to produce a 2-D plot.
The micro-computed tomography (μ-CT) imaging was performed with an “nanotom s 180” device, developed by GE Sensing & Inspection Technologies and using the product line of phoenix x-ray. This CT has a special high power nanofocus tube (180 kV/15 W) with an adjustable focal spot size down to 1 μm in diameter, which enables very sharp imaging data sets. After the scanning process, the 3D data sets have been evaluated with VG Studio Max 2.0. Phase segmentation has been performed by using quantification tools, such as edge detection and phase contrast filter operations, to ensure high accuracy phase thresholding and volume determination. Afterwards, 2D as well as 3D visualization of regions of interest within the samples took place. The scanning parameters for the investigated samples were voltage 55 kV, current 80 μA, projections 1500, average 7, skip 3, timing 1000 ms and voxel size 1.06 μm.
For the high-resolution SEM focused ion beam (FIB-SEM) investigation a Zeiss Auriga field equipped with a field emission cathode and extra-large charge coupled device (CCD) 80 mm2 CCD detectors for energy dispersive x-ray (EDX) analyses was used. Before starting the milling process the sample was sputtered with Pd. A “slice-and –view” procedure was run by milling 25 nm thick cuts in the form of a cross-section 20 × 20 μm in size. A full description of the methodology is described in Warr and Grathoff (2011).
3.1 Basic characterization of the perlite
The cation exchange capacity was practically 0 meq/100 g. In none of the 6 separate measurements any systematic reduction of the extinction at 578 nm was detected. The smectite content, therefore, is less than 0.5 mass%.
3.2 Characterization of the water binding
3.3 Spatial information
Computer tomography (CT)
CT allows investigation of the 3D structure of a sample and provides 2D sections. The grey values of the image shown in Figure 7 (1 μm thick 2D section) correspond to the density of the investigated particles. A mottled structure was observed within the glass particles. Hence even the interior of the volcanic glass is not as homogeneous as expected. Disseminated dark spots were observed which indicated a possibly existing porosity. The diameter of these spots was 1–4 μm. The spots were dark because they were transparent in the 1 μm section due to water or gas filling.
The porosity was determined by He- and Dryflow®-pycnometry. The density of the glass without any pores (after melting) was about 2.6 g/cm3 a typical value of silicate glasses. Without extensive heat treatment and grinding the μm-sized matrix porosity was thought to be intact but He could of course - only - enter the larger macropores. The specific density of this material was about 2.3 g/cm3. This difference, therefore, can be explained by the existence of the μm-sized matrix-porosity accounting for 11 Vol%. The Dryflow®-density of the larger grain fraction was 1.8 g/cm3 which results in a total porosity of about 26 Vol%. This value is supposed to include the macropores. The porosity of the silicate matrix is about 10%.
Infrared spectroscopy showed extinction over a wide range of wavenumbers covering both the OH-stretching region of clay minerals as well as the spectral region typical of free water. This indicated the presence of a range of water binding forms which is in accordance with Dobson et al. (1989). As a further tool to study water binding, differential thermal analysis was applied which confirmed the IR results. The MS-H2O curve (evolved gas analysis measured by mass spectrometry) showed dehydration over an unusually broad temperature range (from 100 to 400°C) which was already observed by Tazaki et al. (1992). Desorption of free water is commonly observed between 100 and 200°C well separated from dehydroxylation (500–750°C). However, no separate second peak which could be assigned to dehydroxylation was observed. Instead one broad peak was found which may correspond to both dehydration and dehydroxylation. This conclusion is supported by the IR results which also indicated the presence of a range of different water bindings, ranging from the typical hydroxyl stretching vibrations to adsorbed water. The hydroxyls do not result from smectite because the smectite content as determined by the cation exchange capacity (CEC) method was ≤0.5 mass%. This content was too low to conclude that the hydroxyls were associated to smectite. One explanation would be that the walls of an extensive pore system are covered with hydroxyls or at least strongly bound water which in terms of binding energy are similar to clay mineral hydroxlys but in this case only cover the very surface.
Using the thermal gravimetry curve the water liberated from the perlite could be quantified (about 2.7 mass%). This value was slightly larger than the loss on ignition (LOI = 2.5 mass%) which can be explained by the fact that the sample was stored at 53% r.H. prior to thermal analysis but dried at 60°C prior to LOI determination. Both values, however, are in good agreement.
The total porosity was 26 Vol% and the μm-sized matrix-porosity accounted for 11 Vol%. IR microscopy showed that the pores are at least partially filled with water. A 3D-FIB visualization of one pore showed columnar structures in the pore. These could be precipitations or newly formed phases. Further discussion about them would be highly speculative.
Most of the perlites are supposed to have formed upon post-emplacement hydration. The μm-sized finely distributed pores identified in the present study, however, are ubipresent in the volcanic glass and the pores are not supposed to be connected. The pores observed are all closed. One option to explain the abundance of this water is that it was trapped when the magma was quickly cooled. This had to be investigated further, e.g. using isotope methods.
Summary and conclusions
Perlites are believed to form upon post-formational hydration often leading to smectite formation. The presence of smectite in the investigated sample, however, can be excluded. CEC measurements with high sample masses are accurate enough to prove that less than 0.5 mass% smectite was present. Infrared spectroscopy and thermal analysis showed the presence of a continuous range of water binding, ranging from hydroxyls/strongly bound water to molecular water. Using CT and FIB an extensive pore system of closed pores was found with pore diameters in the range of 1–5 μm. In the same range, infrared microscopy revealed domains with significant extinction in the water deformation region. Because of similar size and distribution these signals were believed to represent the filling of the pores. According to the consistent picture gained from applying a set of different methods, the glass particles of at least the investigated material contain appreciable small water filled pores. It remains unsolved wether the water in these pores entered after or throughout the emplacement. However, the pores are sealed and no indications of cracks were found which indicates a primary source of the water, i.e. water was probably entrapped by quenching of the lava. The water in these pores may be important for the possible formation of clay minerals out of perlites and may have implications for the formation of bentonites in Milos. The water in the μm-sized pores may be the main reason for the thermal expandability of this perlite.
- Allen RL: False pyroclastic textures in altered silicic lavas, with implications for volcanic associated mineralization. Econ Geol 1988, 83: 1424-1446. 10.2113/gsecongeo.83.7.1424View ArticleGoogle Scholar
- Bishop JL, Pieters CM, Edwards JO: Infrared spectroscopic analyses on the nature of water in montmorillonite. Clay Clay Miner 1994, 42(6):702-716. 10.1346/CCMN.1994.0420606View ArticleGoogle Scholar
- Denton JS, Tuffen H, Gilbert JS, Olding N: The hydration and alteration of perlite and rhyolite. J Geol Soc Lond 2009, 166: 895-904. 10.1144/0016-76492008-007View ArticleGoogle Scholar
- Dobson PF, Epstein S, Stolper EM: Hydrogen isotope fractionation between coexisting vapor and silicate glasses and melts at low pressure. Geochim Cosmochim Acta 1989, 53: 2723-2730. 10.1016/0016-7037(89)90143-9View ArticleGoogle Scholar
- Eckert H, Yesinowski JP, Silver LA, Stolper EM: Water in silicate glasses: quantitation and structural studies by 1H solid echo and MAS-NMR methods. J Phys Chem 1988, 92: 2055-2064. 10.1021/j100318a070View ArticleGoogle Scholar
- Friedmann I, Smith RL, Long WD: Hydration of natural glass and formation of perlite. Geol Soc Am Bull 1966, 77: 323-328. 10.1130/0016-7606(1966)77[323:HONGAF]2.0.CO;2View ArticleGoogle Scholar
- Fytikas M, Innocenti F, Kolios N, Manetti P, Mazzouli R, Poli G, Rita F, Villai L: Volcanology and petrology of volcanic products from the island of Milos and neighbouring islets. J Volcanol Geotherm Res 1986, 28: 297-317. 10.1016/0377-0273(86)90028-4View ArticleGoogle Scholar
- Kaufhold S, Dohrmann R: Beyond the Methylene Blue method: determination of the smectite content using the Cu-triene method. Z Angew Geol 2003, 49(2):13-18.Google Scholar
- Kaufhold S, Ufer K, Kaufhold A, Stucki J, Anastácio A, Jahn R, Dohrmann R: Quantification of allophane clay from Ecuador. Clay Clay Miner 2010, 58(5):707-716. 10.1346/CCMN.2010.0580509View ArticleGoogle Scholar
- Keller WD, Picket EE: Hydroxyl and water in perlite from superior, Arizona. Am J Sci 1954, 252: 87-98. 10.2475/ajs.252.2.87View ArticleGoogle Scholar
- Koukouzas N, Dunham A: Genesis of a Volcanic Industrial Rock. Trachilas perlite deposit, Milos island, Greece; 1994.Google Scholar
- Koukouzas NK, Dunham AC, Scott PW: Suitability of Greek perlite for industrial applications. Appl Earth Sci: Trans Inst Min Metall B 2000, 109: 105-111. 10.1179/aes.2000.109.2.105View ArticleGoogle Scholar
- Lorenz W, Gwosdz W: Bewertungskriterien für Industrieminerale, Steine und Erden, Teil 4: Vulkanische Gesteine und Leichtzuschläge, Bims, Bimsasche, andere vulkanische Aschen und Schlacken, Tuffgesteine, Perlit, Basalt, andere vulkanische Gesteine. Geol. Jb. 2000, ᅟ(Reihe H7):3-111.Google Scholar
- Marshall RR: Devitrification of natural glass. Geol Soc Am Bull 1961, 72: 1493-1520. 10.1130/0016-7606(1961)72[1493:DONG]2.0.CO;2View ArticleGoogle Scholar
- Meier LP, Kahr G: Determination of the cation exchange capacity (CEC) of clay minerals using the complexes of Copper (II) ion with Triethylenetetramine and Tretraethylenepentamine. Clay Clay Miner 1999, 47: 386-388. 10.1346/CCMN.1999.0470315View ArticleGoogle Scholar
- Nichols ARL, Carroll MR, Höskuldsson Á: Is the Iceland hot spot also wet? Evidence from the water contents of undegassed submarine and subglacial pillow basalts. Earth Planet Sci Lett 2002, 202: 77-87. 10.1016/S0012-821X(02)00758-6View ArticleGoogle Scholar
- Pandya N, Muenow DW, Sharma SK: The effect of bulk composition on the speciation of water in submarine volcanic glasses. Geochim Cosmochim Acta 1992, 56: 1875-1883. 10.1016/0016-7037(92)90317-CView ArticleGoogle Scholar
- Ross CS, Smith RL: Water and other volatiles in volcanic glasses. Am Mineral 1955, 40: 1071-1089.Google Scholar
- Silver LA, Lhinger PD, Stolper E: The influence of bulk composition on the speciation of water in silicate glasses. Contrib Mineral Petrol 1990, 104: 142-162. 10.1007/BF00306439View ArticleGoogle Scholar
- Stolper E: Water in silicate glasses: an infrared spectroscopic study. Contrib Mineral Petrol 1982, 81: 1-17. 10.1007/BF00371154View ArticleGoogle Scholar
- Tazaki K, Tiba T, Aratani M, Miyachi M: Structural water in volcanic glass. Clay Clay Miner 1992, 40: 122-127. 10.1346/CCMN.1992.0400113View ArticleGoogle Scholar
- Warr LN, Grathoff G: Geoscientific applications of particle detection and imaging techniques with special focus on the monitoring clay mineral reactions. In Handbook of Particle Detection Edited by: Grupen C, Buvat I. 2011. Springer, ISBN 978-3-642-13270-4, 667 – 685Google Scholar
- Wysoczanski R, Tani K: Spectroscopic FTIR imaging of water species in silicic volcanic glasses and melt inclusions: An example from the Izu-Bonin arc. J Volcanol Geotherm Res 2006, 156: 302-314. 10.1016/j.jvolgeores.2006.03.024View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.