
The value of petrography in assessing dimension stoneby Geoff Quick CSIRO Manufacturing & Infrastructure Technology In this invited paper, Geoff Quick outlines the value of an accurate petrographic description, not only in classifying stone, but also in describing the mineralogical and textural features that may influence its chemical, physical and mechanical performance in service. Two case studies are used to demonstrate some of the potential benefits of petrography. More accurate descriptions of stone, as now required in Europe, should assist in improved specifications.
In the original geological nomenclature, petrography is "the description and systematic classification of rocks by means of microscopic examination of thin sections" (Glossary of Geology, AGI, 1974); a technique widely used by geologists since the mid-19th century through the use of the petrographic (polarising) microscope. However, petrography is now also applied to a full range of building materials, including dimension stone, reconstituted stone, plaster, concrete and mortars; and can include macroscopic, image analysis and other microscopic techniques such as stereomicroscopy, electron probe microanalysis, and scanning electron microscopy (SEM) augmented by energy-dispersive X-ray microanalysis (EDX). If required, a petrographic examination can also be complemented by other analytical procedures, such as:
A routine petrographic examination usually involves an escalating series of magnification steps. A macroscopic examination is first undertaken to note features such as cracks, stylolites, veins, voids, fillings, bedding, cleavage, fossils, grain size, colour range, weathering and alteration, staining, inclusions, fabrication issues, etc. The macroscopic examination may include observations aided by a hand lens or a low-powered stereomicroscope. At least five polished thin sections (slices of stone carefully ground down to a thickness of 0.03 mm and then mounted on microscope slides) should then be prepared from representative samples. These are then viewed under the petrographic (polarising) microscope in both transmitted and reflected light modes in order to determine the detailed mineralogy and texture. More powerful microscopic techniques, such as the SEM (and EDX), can then be used to further differentiate potentially deleterious minerals (if required). An accurate and detailed petrographic assessment should only be undertaken by a qualified geologist/petrographer with experience in the dimension stone industry. Provided sampling is adequately representative, petrography can be used in assessing the potential durability of newly worked stone, as well as in the forensic analysis of existing stone structures. It is important to note that a petrographic examination should always complement (and not replace) other standard physical test methods, such as those described by the author in a previous paper [1].
The marketing of dimension stone has often led to the use of "descriptive" names that are of misleading non-technical nomenclature. For example, many of the stones typically referred to as granite are not granite at all. Geologically, granite is a visible granular, light coloured plutonic igneous rock consisting of quartz, alkali feldspar and mica minerals. In the trade, however, the term usually encompasses nearly all coarse-grained crystalline igneous and metamorphic rocks despite their origin, and includes such varieties as gabbro (black granite), diorite, syenite and gneiss. Similarly, the geological definition of a marble is a calcareous rock (usually a limestone) that has undergone recrystallisation due to heat and/or pressure (metamorphism). In the building trade many marbles are actually dense limestones, a sedimentary rock composed of calcium carbonate (often fossiliferous) capable of taking a polish. In Victoria, Australia, bluestone is a dark basaltic volcanic rock, while in South Australia and the USA the term is reserved for blue-coloured siltstones or fine-grained sandstones (sedimentary rocks). Stones that have been cleft (i.e. split) are sometimes all referred to as slates (e.g. some split-faced sandstones, limestones). However, a true slate is a fine-grained metamorphic rock, derived from sedimentary shales, and is composed mainly of planar micas, chlorite and quartz. True slates can be easily split into thin sheets because of the planar orientation of the minerals. To add to the confusion, some stones are named after their place of origin (e.g. Sydney Blue) or special visual characteristics of the stone (e.g. Desert Lilac) and bear no relationship to any geological names. Accurate petrographic definitions are now required in European Standard (CEN) natural stone test methods (BS EN 12407 and prEN 12670).
An accurate petrographic name and description should convey to the stone specifier information in terms of its mineralogy, grain size and texture. Where possible, the characterisation of the material should be in quantitative terms. An accurate petrographic description will assist the specifier in:
Table 1 provides a summary of the type of information that can be provided by a detailed petrographic examination of dimension stone.
The durability and visual appearance of dimension stone is controlled by the constituent minerals present and by their textural and structural relationships. This paper has shown that petrographic techniques can be effectively used, in conjunction with other standard physical test methods, to assess the potential performance in service of natural dimension stone. Australian producers should be aware that accurate petrographic descriptions and names are now required in specifying dimension stone in Europe. TABLE 1 - SUMMARY OF PETROGRAPHIC FEATURES AND OBJECTIVES
CASE STUDY 1 - Oh what a (yellow) feeling!A source of frequent complaints about some stones is that they are susceptible to staining problems. Recently CSIRO was asked to examine polished marble tiles on the floors and walls of three bathrooms at a new dwelling. The polished whitish-grey Italian marble had shown a progressive yellowing since it was installed some four months before our inspection. The problem was more evident on surfaces adjacent to the baths than on the walls (Figure 2). The tiles had been laid on a white cementitious adhesive and the grouting was white in colour. None of the tiles had been sealed. Microscopy was used to determine the petrography of the stone. Reflected-light optical microscopy and scanning electron microscopy coupled with an energy-dispersive X-ray spectrometer revealed that the stone contained many very small inclusions of a yellow metallic mineral; identified as the mineral pyrite (iron sulfide) (Figure 2 and 3). These very small inclusions were common and evenly dispersed throughout the tiles. Pyrite is potentially unstable in humid environments where it commonly oxidises to hydrated iron compounds. In this case, the marble, being a porous medium, was very susceptible to water penetration since it was unsealed. In addition, pyrite will also begin to oxidise when exposed to other oxidising agents such as some acids and household bleach. The rate at which oxidisation occurs can vary, depending on the amount of pyrite present, its size, and the environmental conditions the stone is subjected to (i.e. moisture from daily cleaning, the setting bed, or simply by humid conditions). In some cases it may take years before yellowing occurs. However, in water-saturated conditions the process would be greatly accelerated. It has been reported in some cases that when a marble (containing pyrite) has been subjected to a flood, the yellowing can occur overnight. If this particular stone had been examined petrographically, there is no doubt that this particular marble would have been rejected in the selection process. The oxidation of pyrite (and resultant iron staining) would not have occurred if oxygenated water had been excluded from the reaction. In other words, if the tiles (and grout joints) had been effectively sealed with a suitable deep penetrating sealer, the staining may not have occurred. This assumption is based on the fact that all water available to the system is absorbed from the top surface, and grout joints, by way of cleaning procedures and water spills. Marble, however, is a porous medium, which is capable of absorbing water, not only from the proper surface but from other sources. Since the base of the shower has not been sealed, it is possible that the tiles and grout joints in the shower base could act as a sink for downward percolating solutions. The water could then travel along the interface between the floor tiles and the mortar base (or membrane) and saturate the adjacent floor tiles from below. Whatever way, the oxidation of the pyrite will proceed if water is present in the system. There is a great debate in the stone industry on whether all (top and bottom) surfaces should be pre-sealed before installation. In theory, this is the most appropriate method to undertake. However, at CSIRO we have found that in some situations there are bonding problems between some adhesives and sealed tiles. It is thoroughly recommended that the adhesive bond between the sealed back of the marble tile and the chosen adhesive be tested before adopting this technique.
CASE STUDY 2 - A case of identity!CSIRO was asked by a property manager to examine samples of polished black granite tiles, which had been removed from an internal floor of an apartment block. It was reported by our client that after several cleaning regimes, a third of the laid tiles had faded and showed two toned features. As Figure 4 shows, this turned out to be black (phenocrystal) rounded crystals set in a fine-grained grey-coloured matrix. The remaining black granite tiles showed no signs of discolouration after using the same cleaning regime. Although the exact cleaning procedure was not disclosed, our client suspected it was weakly acidic. CSIRO was commissioned to examine supplied samples in order to identify why some of the tiles turned grey while others remained truly black. In the building industry, "black granites" are dark-coloured igneous rocks that geologists would classify as either basalt, basanite, dolerite, gabbro, diorite or anorthosite. To a geologist this is a pretty loose definition, since it incorporates both igneous rocks that have formed in lava flows (volcanic, e.g. basalt or basanite) as well as those formed in bodies deep within the earth's crust (plutonic, e.g. gabbro or monzonite). A macroscopic examination on freshly broken edges revealed that both affected and unaffected stones looked (at first glance) petrographically similar. Petrographic polished thin sections were then made from all representative samples and examined under the petrographic (polarising) microscope. It was found that the affected (faded) grey sample was porphyritic in texture. It consisted of large euhedral-shaped phenocrysts of titaniferous augite set in fine-grained matrix of augite, olivine, calcic plagioclase feldspar and minor nepheline, ilmenite and magnetite. However, interstitial to these crystals is a prominent dark brown volcanic glass (Fig. 5). Volcanic glass only occurs in igneous rocks formed in lava flows or rarely in very shallow intrusions. Either way, it is formed close to or on the earth's surface. The glass occurs as a consequence of rapid chilling of the lava. A geologist would classify this rock as a porphyritic basanite, an extrusive (volcanic) igneous rock. In contrast, the unaffected black granite sample was medium-grained with an equigranular texture. The minerals consisted of, in order of abundance, aegirine augite (clinopyroxene), olivine, calcic plagioclase feldspar, alkali feldspar, chloritised biotite and minor quartz and magnetite (Figure 6). A geologist would classify this rock as a monzonite, which is an intrusive (plutonic) rock formed deep in the earth's crust. One can instantly see from the thin section study that the two samples examined are definitely not of the same petrographic type and, as such, definitely do not come from the same geological source. Not all volcanic rocks contain glass, but where present, it is in a meta-stable state. It is this glassy region that has apparently reacted with the cleaning fluids. A small sample of the affected stone was then repolished using a petrographic polishing lap. After 60 minutes of lubricated polishing with 6 then 1µm diamond pastes, the specimen was re-examined. Although the polishing produced a very fine gloss to the specimen, the grey colour only darkened slightly. Therefore, it is concluded that normal industrial polishing techniques will not restore this stone to its original condition. The origin of the black granite(s) was not supplied. However, it is worth noting in an article by Hueston [2], that a common problem with black granites that have been fabricated in Asia (especially India), is that they have been doctored by using wax, linseed oil, etc. It is possible that as the stone is cleaned, the waxes and oils are removed, and permanent fading (bleaching) of the colour has occurred. Although this is a probable cause of the problem in question, we did not have the appropriate samples to either prove or disprove this theory. In conclusion, whatever the cause of the fading, petrography has revealed, unknown to the property developer at that time, that the so-called polished black granite tiles installed as internal flooring in the apartment block is made of two but very distinct petrographic stone types that in turn have reacted differently to the same cleaning regime. Geoff Quick is a geologist and microscopist at CSIRO Manufacturing & Infrastructure Technology. He is a member of the Standards Australia committees on natural and reconstituted stone tiles, and masonry units and segmented pavers. He is also a member of the ASTM C-18 Committee on Dimension Stone, and an office bearer of the Geological Society of Australia (Victorian Education Committee), the Australian Microbeam Analysis Society (AMAS), and the Victorian Electron Microscopy and Analysis Society (VEMAS). REFERENCES
|
Stone of the day DiscoveringStone Issue #19 Download |