Corrosion Science 51 (2009) 2387–2392
Contents lists available at ScienceDirect
Corrosion Science journal homepage: www.elsevier.com/locate/corsci
Passivation techniques to prevent corrosion of iron sulphides in rooﬁng slates V. Cárdenes Van den Eynde *, R. Paradelo, C. Monterroso Dpto. Edafología y Química Agrícola, Universidad de Santiago de Compostela, 15782 Santiago de Compostela, Spain
a r t i c l e
i n f o
Article history: Received 24 March 2009 Accepted 13 June 2009 Available online 18 June 2009 Keywords: A. Stonework A. Rooﬁng slate C. Image analysis C. Oxide coatings C. Passivity
a b s t r a c t The aim of the present study was to evaluate the efﬁcacy of different methods of sulphide microencapsulation in terms of inhibiting sulphide oxidation in rooﬁng slate. For this, a preliminary test was carried out with ground pyrite and the methods that provided the best results were applied to samples of rooﬁng slate. Protection against oxidation was measured by the amount of iron released after enforced oxidation with H2O2 in the tests with pyrite, and by thermal cycle alteration, SO2 and saline spray tests (EN 123262:2000 and EN 14147:2004) in the tests with the slate samples. The results indicate that treatments with potassium phosphate and potassium silicate, proposed for controlling acid mine drainage, were the most effective at protecting pyrite against oxidation, and that these methods are also effective when applied to slate. Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction Spain is the main producer of rooﬁng slate in the world, and the principal areas of exploitation are in the Hercynian areas in the northwest of the peninsula [1,2]. The presence of iron sulphides in different types of slate is currently one of the main problems facing the sector as oxidation of these compounds produces colour changes and rust marks on the slate. This defect is purely aesthetic and the slate rarely becomes deteriorated as a result of oxidation. However, even though the other characteristics (splitability, resistance to ﬂexion, water absorption, etc.) may be optimal, slate containing an abundance of sulphides cannot be categorised as top quality . Oxidation of iron sulphide has been studied in relation to the problem of acid mine drainage (AMD) in mine dumps [4,5]. The reactions are initiated by the oxidation of sulphide by atmospheric O2, which results in release of SO42 and Fe2+, and a decrease in pH, thereby generating an environment that favours development of iron-oxidising bacteria such as Thiobacillus ferrooxidans and Leptospirillum ferrooxidans . Under highly acidic conditions, such bacteria can catalyze the oxidation of Fe2+ to Fe3+, and oxidation of pyrite by Fe3+ then becomes the main mechanism whereby acid is produced . Three basic strategies have been used to prevent oxidation of iron sulphides: application of hydrofugants, complexation of Fe3+ and microencapsulation or passivation. Treatment with hydrofugants inhibits oxidation by preventing contact between the sul* Corresponding author. Tel.: +34 985228625/+34 981563100x13288; fax: +34 985228625. E-mail address: [email protected]
(V. Cárdenes Van den Eynde). 0010-938X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.corsci.2009.06.025
phide and atmospheric O2. For example, Lalvani et al.  tested the inhibitory effect of various compounds, and found that siloxanes provided the best results. Jiang et al.  used sodium oleate to create a hydrophobic layer on the sulphide surface. Similarly, Zhang et al.  and Kargbo et al.  proposed the use of a phospholipid (23:2 Diyne PC, commercialized by Avanti Polar Lipids) as a protective agent. Treatment of slate with hydrofugants is effective but very expensive. Compounds used to bind Fe3+ include triethylentetramine (TETA) and diethylentetramine (DETA) . Backes et al.  proposed the use of 1:10 phenanthroline (ligand) with Panacide (a biocide) to inhibit oxidation of Fe2+ to Fe3+. Finally, passivation or microencapsulation can be achieved by the formation of precipitates on the surface of the sulphides, which then act as a protective barrier. Lalvani et al.  proposed cationic coating of the pyrite by electrolysis. Nicholson et al.  observed that during the oxidation of pyrite under alkaline conditions, iron oxide accumulated and acted as a coating that inhibited further oxidation. Treatment of pyrite-containing coal mine spoils with solutions of sodium silicate (Na2SiO3) or potassium phosphate (KH2PO4) have generated great expectations in terms of the control of acid mine drainage (AMD) [16–18]. Evangelou, professor of Soil/ Water Physical chemistry at Kentucky University, summarizes the pyrite oxidation processes and passivation methods developed by himself in the book ‘‘Pyrite oxidation and its control” . These treatments induce the formation of a protective coating on the pyrite, by incorporating the Fe3+ released during oxidation. Potassium phosphate (KH2PO4) was also used by Georgopoulou et al.  to passivate pyrrhotite, with good results. The utilization of silicate or phosphate solutions to induce the formation of protective coatings is also a successful method to prevent corrosion of steel [20– 22]. More recently, Satur et al.  proposed passivation of iron
V. Cárdenes Van den Eynde et al. / Corrosion Science 51 (2009) 2387–2392
sulphides with Ti oxide/hydroxide obtained from rutile and anatase (TiO2), with catechol as a carrier. In recent years various chemical products have been proposed for application to rooﬁng slate , including polyurethanes, siloxanes and acrylates. Costagliola et al.  applied three different types of synthetic resins (acrylic, ﬂuorinated and siliconized) to protect museum samples of pyrite against the action of ultraviolet light; the best protection – as evaluated by tests in saline spray chambers including exposure of the samples to ultraviolet light and immersion in an oxidising solution of H2O2 was obtained with the ﬂuorinated resins (NH + Kynar), followed by the acrylics. However, these products altered the external appearance of the samples, modifying their lustre and natural colour and conferring them an artiﬁcial appearance. Moreover, the durability of these products on slate building material has not been tested in the ﬁeld. Passivation of iron sulphides is a possible alternative to the application of chemical products to the slate surface for preventing iron sulphides corrosion. The advantage of this method is that it only acts on the iron sulphide, and not on the other minerals present in slate (phyllosilicates, quartz and feldspars). Also, the utilized compounds do not present a risk for the environment. Given the simplicity, efﬁcacy and low cost of the method, its application should be evaluated on an industrial scale as an alternative to other protective methods more expensive and harmful for the environment. The data obtained by the different techniques used by these authors show a reduction in the rate of oxidation of iron sulphides. However, none of these methods have resulted in an industriallyfeasible method of inhibiting oxidation of iron sulphides. In the present study the efﬁcacy of two methods of passivation proposed in [16–18,23,26] were evaluated and their applicability to rooﬁng slate was assessed.
2. Material and methods 2.1. Materials Two passivation experiments were carried out in this study, the ﬁrst with ground pyrite crystals and the second with samples of iron sulphide-rich slate. The pyrite crystals were obtained from sedimentary deposits in Navajún (La Rioja, España) formed under hydrothermal conditions of low supersaturation .The sample of crystals used was composed of 99.1% FeS2 and 0.9% SiO2. Before the experiment the pyrite was ground in an agate ball mill to <100 lm and washed with 10% HCl, to remove any possible products of prior oxidation, then with distilled water and acetone. For the second experiment, 27 rectangular samples of slate (30 cm 20 cm, 3–4 mm thick) were used. The slate was obtained from the upper layers of the Luarca slate formation in the mining areas of the Casaio valley (Ourense, España). The Luarca slate formation is the largest slate deposit in the Iberian Peninsula , and is exploited at several locations in northwest Spain. It consists of a monotonous succession of grey and black slates, with some levels of interlayed sandstone, particularly on the footwall of the deposit and vulcano-sedimentary levels of variable thickness towards the roof, and occasionally on the footwall of the deposit. The grain is ﬁne to medium, with a great homogeneity that confers the slate a high degree of cleavability. The total thickness of the deposit is estimated at 200 m . There is a characteristic existence towards the roof, of abundant pyritized brachiopods and cubic pyrite crystals. The formation is thought to be from the Llanvirniense era . The slate formations of Sinclinal de Truchas (Casaio Formation, Rozadais Formation and Losadilla Formation) are situated above this formation, and are also exploited for rooﬁng slate on a regional scale. The iron sulphides in these slates are generally
a mixture of mono- and disulphides of iron (Clemente Recio, University of Salamanca, pers. comm.). 2.2. Passivation of pyrite crystals The ground pyrite crystals were subjected to one of three different passivation treatments: with Na2SiO3, with KH2PO4 , or with catechol–Ti . For the ﬁrst two treatments, the protective solutions used were: (a) 0.145 M H2O2, 0.1 M NaCl, 0.1 M NaOAc, 0.1 M Na2SiO3, and (b) 0.145 M H2O2, 0.1 M NaCl, 0.1 M NaOAc, 0.1 M KH2PO4. About 500 mg of the pyrite powder was placed in 1 L of each protective solution and the resulting suspension was shaken for 6 h at a constant temperature of 22 (±2 °C), after adjusting pH at 5–6 with 10% HCl, since low pH values increases the oxidation reactions of iron sulphides , pp. 215–235. At the end of this period, the treated pyrite powder was ﬁltered, washed in acetone and stored in a vacuum desiccator. For passivation with Ti, a preliminary reaction was carried out with 1 g of rutile and 10 mL of 0.1 M catechol, shaken for 24 h at room temperature to produce a catechol–Ti complex. The solution was ﬁltered (0.45 lm) through a cellulose acetate ﬁlter. Six grams of pyrite were then added to 10 mL of this solution, with shaking for 24 h at room temperature. The passivated pyrite was ﬁltered (0.45 lm), washed in acetone and stored in a vacuum desiccator for later testing. To evaluate the efﬁcacy of the passivation treatments with pyrite, 500 mg of the passivated pyrite (Py-Si, Py-P and Py-Ti) were placed in a glass beaker with 1000 mL of 0.145 M H2O2 (three replicates for each treatment). Aliquots of 20 mL of the solution were removed after the following times: 5, 30, 60, 90, 120, 150, 180, 210, 240, 300, 360, 420, 480, 540 and 1440 min, the extracts were ﬁltered by cellulose acetate ﬁlters (0.45 lm) and the concentration of dissolved iron in each was determined by ﬂame atomic absorption spectrometry (Perkin-Elmer 1100B). To evaluate the capacity of the pyrite for autopassivation, derived from in situ precipitation of the iron released during oxidation, the same experiment was carried out with previously oxidised pyrite (Py-Fe), which consisted in pyrite passivated after reaction with 0.145 M H2O2 during 30 min. Finally, a sample of untreated pyrite was used as a control (Py-Con) under the same test conditions. Development of the pyrite oxidation was monitored by measuring the concentration of Fe in the extracts. The data were transformed to percentage of oxidised Fe and ﬁtted to an exponential model of type (Eq. (1)):
%Fe ¼ C ð1 ekt Þ
where %Fe is the percent of Fe oxidised (on a weight basis), C is the maximum quantity of oxidisable Fe, k is the ﬁrst-order rate constant, and t is the time expressed in minutes. These parameters enable better description of the oxidation reaction than a single measurement of the iron released by oxidation; the rate constant is also fundamental for the study of chemical reactions as amongst other things it enables prediction of the rate of oxidation at different temperatures. The differences between the different treatments were evaluated by ANOVA and Duncan’s test (p < 0.05) by use of SPSS software (version 15.0) for Windows. 2.3. Passivation of slate samples The treatments that provided the best results in the previous study were applied to slate samples. The samples were divided into 3 groups of 9 units each. One group was treated with Na2SiO3 (PzSi), another with KH2PO4 (Pz-P), and the third was left untreated as a control (Pz-Con). The slate samples were stacked horizontally in a
V. Cárdenes Van den Eynde et al. / Corrosion Science 51 (2009) 2387–2392
Fig. 1. System used for passivation of slate samples.
water bath (Microterm P. Selecta 30000423, Fig. 1) and separated by fragments of slate (3–4 mm) placed at each corner of the samples. Each group of samples was submerged in the corresponding solution for 6 h at a constant temperature of 22 (±2 °C), after adjusting pH at 5–6 with 10% HCl. At the end of the treatment period, the slate samples were washed with distilled water and left to dry at room temperature. The surface of each of the treated samples was then examined under a scanning electron microscope (LEO-435VP). To evaluate the efﬁcacy of the treatments, the protocol for assessing protective products for rooﬁng slate described in  was followed. The protocol consists of a ﬁrst stage in which the appearance of the treated slate is evaluated qualitatively, and a second part in which the efﬁcacy of the product or treatment is evaluated by thermal cycle and SO2 exposure tests , and the test of resistance to aging by saline spray . In the thermal cycle test, the slate samples were subjected to 20 consecutive cycles, each of 24 h duration and consisting of the
following stages: (a) heating to 110 °C for 17 h in a ventilated oven, (b) cooling for 1 h and (c) immersion in distilled water at 23 °C (±5 °C) for 6 h. At the end of each of the 20 cycles, the degree of oxidation of each slate slab was categorised as follows: T1, (change in colour of the iron sulphides) T2, (change in colour and oxidation plus diffusion) and T3 (change in colour, diffusion and perforation). For the present study, another category, T0 (absence of changes), was also considered and a numerical value was assigned to each (T0 = 0, T1 = 1, T2 = 2 and T3 = 3). The arithmetic mean value of the scores for each group was used as an index of the daily changes due to oxidation. The SO2 exposure test was used to assess the alteration that slate may undergo in an urban environment. The test consisted of placing the slate samples in hermetically sealed containers (60 L) with 0.60 L of an aqueous solution containing 5–6% SO2 and 0.18 L of distilled water, which creates a sulphur dioxide atmosphere. After 21 days, the slate samples were examined for ﬁssures and changes in colour that may have occurred as the result of the formation of new mineral phases, e.g. gypsum. For the purposes of the present study, only the alterations generated in the iron sulphides present in the slates were considered, without taking into account possible losses in volume or ﬂaking of the slate. The test of resistance to aging by saline spray was carried out in a sealed chamber in which the slate samples were sprayed with a vaporised solution of 100(±10) g/L of NaCl for 4 h, then left in the chamber to dry for 8 h; these two stages constituted one cycle. After 60 such cycles, the slate samples were washed with distilled water and were examined visually to detect any possible alterations. The surface of each of the slate samples was digitally scanned (Canon Pixma MP150) before and after each test and the area occupied by sulphides and by the colour changes produced by their
Fig. 2. Example of the application of image analysis to the slate samples before (A) and after (B) the SO2 test. The area occupied by the sulphides and the surrounding area affected by the oxidation are shown in white (A1 before the test and B1 after the test). Escale measure is 5 cm.
V. Cárdenes Van den Eynde et al. / Corrosion Science 51 (2009) 2387–2392
oxidation were quantiﬁed by analysis of the digital images with the ImageJ image processing programme  (Fig. 2). 3. Results and discussion 3.1. Efﬁcacy of passivation of pyrite crystals The results of the oxidation tests on ground pyrite subjected to the different treatments are shown in Fig. 3. Oxidation of the raw untreated pyrite used as a control (Py-Con) showed a ﬁrst, rapid stage of oxidation, in which 30% of the iron was oxidised (mean value) in 90 min, followed by slow oxidation, in which 40% of the Fe was oxidised by the end of the experiment (1440 min). Autopassivation and passivation with catechol–Ti (Py-Fe and Py-Ti) produced a signiﬁcant decrease in oxidation, with maximum values between 18% and 20% of the Fe oxidised. By far the greatest reduction was produced by the treatments with phosphate and silicate (Py-P and Py-Si), which resulted in less than 10% of the iron being oxidised. The most effective protective treatments not only modiﬁed the absolute values of the percentage of oxidation, but also greatly modiﬁed the form of the corresponding curve. Treatments Py-Fe and Py-Ti produced almost identical curves, similar to that corresponding to Py-Con, whereas treatments Py-P and Py-Si produced curves with a much lower slope in the ﬁrst stage, which indicates a long delay in the start of oxidation and a decrease in the rate of the reaction. The parameters of the changes in pyrite oxidation were ﬁtted to Eq. (1). Data derived from the model ﬁts (C, k, and the product between them, which also provides a measure of the rate of oxidation), as well as the percentage of Fe weight oxidised in 24 h and the regression coefﬁcient R2, are shown in Table 1.The values of C, k, and the product C k are mean values from the ﬁts derived from each replicate, whereas the values of
R2 correspond to mean values from the experimental and estimated data. The data indicate that all of the protective treatments reduced pyrite oxidation with respect to the control, but that the treatments with silicate and phosphate were much more effective than the treatment with the catechol–Ti complex. There were no signiﬁcant differences in the results produced by the silicate and phosphate treatments, and both were equally effective. Treatment with catechol–Ti induces precipitation of Ti on the surface of the pyrite in the form of Ti(OH)4 or TiO2, which acts as a protective coating against oxidation  and its effect appears similar to that obtained after the in situ precipitation of the iron derived from the oxidation of the pyrite (Py-Fe). The treatments with Py-P and Py-Si induced formation of a protective coating of ferric hydroxide and ferric silicate or phosphate, depending on the protective agent used – on the surface of the pyrite. These coatings signiﬁcantly reduce the oxidation by both their tendency to trap Fe3+ and by acting as a physical barrier that inhibits oxygen diffusion . Generally speaking, trapping dissolving cations is not sufﬁcient to insure passivation, but the protection given by this method can be good enough to prevent this speciﬁc problem. Rooﬁng slate may have up to 1% volume of iron sulphides, and its corrosion may lead to an aesthetic problem due to the oxidation stains, but not the collapse of the tile. This kind of protection delays the beginning of the corrosion enough time to form other natural coatings, such as iron oxides , which assure the protection, or even the disappearance of the iron sulphides. However, coating formation is optimum in circumneutral conditions, decreasing its effectiveness as acidity increases. The normal conditions of use for rooﬁng slate are slightly acid, due to rainwater , so it is expected to create effective protection against the oxidation by the coating. The H2O2 in the protective solution oxidises and releases Fe3+ necessary to form the coating, the NaCl provides a background electrolyte that favours ion exchange and the NaOAc acts as a buffer, inhibiting the activity of the protons produced during the oxidation. The reactions by which the passivation is produced are as follows : þ FeS2 þ 7:5H2 O2 þ Na2 SiO3 ¼ FeHSiO4 þ 2SO2 4 þ 2H þ 6H2 O
þ 2Naþ þ þ FeS2 þ 7:5H2 O2 þ KH2 PO4 ¼ FePO4 þ 2SO2 4 þ 3H þ 7H2 O þ K
3.2. Efﬁcacy of passivation of slate samples
Fig. 3. Oxidation of ground pyrite subjected to the different treatments: Py-Con: control, raw, untreated pyrite; Py-Fe: previously oxidised pyrite; Py-Si: pyrite passivated with sodium silicate; Py-P: pyrite passivated with potassium phosphate; Py-Ti: pyrite passivated with titanium. Bars represent two standard deviations.
Table 1 Parameters of the changes in pyrite oxidation for 24 h. Different letters within the same column indicate signiﬁcant differences between treatments, according to Duncan’s test (p < 0.05). Treatment
C k (% min1)
Py-Con Py-Ox Py-Si Py-P Py-Ti
45.1a 23.8b 7.9d 9.0d 18.6c
32.9ª 20.7b 8.6c 8.7c 18.9b
0.019a 0.016a 0.003c 0.007bc 0.011b
0.620a 0.330b 0.023d 0.054d 0.200c
0.823 0.934 0.956 0.986 0.990
The slate samples were subjected to the passivation methods that produced the best results in the pyrite oxidation test, i.e. with potassium phosphate and sodium silicate. After the application of these treatments the ﬁnal appearance of the slate samples was not altered, and the lustre and colour of the surface was maintained. Examination of the scanning electron microscope images conﬁrmed that both passivation treatments induced the formation of a coating (10–20 lm) on the surface of the iron sulphides (Fig. 4), with no alteration of the rest of the surface of the slate. The results of the thermal cycle tests carried out to evaluate the efﬁcacy of these treatments are shown in Fig. 5. During this test, the cycles of ventilated heating and moistening ensured a continuous input of oxygen, an environment suitable for release of Fe3+ and a temperature that increased the kinetics of the oxidation reaction. In the untreated slate samples (Pz-Con) there were no signs of oxidation during the ﬁrst 6 cycles, after which there was a continuous increase in the index until the end of the test. In these samples, the oxidation index reached a value of 1 in the 10th cycle and a value of 2 after the 16th cycle; a value of 3 was never reached. All of the passivated slate samples developed a greater resistance to oxidation during the thermal cycle test, as indicated by the oxidation indices obtained, which only reached a value
V. Cárdenes Van den Eynde et al. / Corrosion Science 51 (2009) 2387–2392
slightly higher than 1 at the end of the test. Oxidation marks (with no diffusion) on the slate surface (characteristic of an oxidation index of 1 in the thermal cycle) are generally due to alteration of the cubic pyrite crystals, whereas the oxide streaks (grade 2 in the thermal cycle test) are mainly due to pyrrhotite. Perforation of the slate (grade 3 in the thermal cycle test) is uncommon as the
iron sulphide crystals must be very large for this and slates with such crystals are usually removed during processing. Analysis of the images taken before and after the test carried out to evaluate the efﬁcacy of the passivation treatments enabled accurate quantiﬁcation of the pyrite oxidation by calculation of the size of the affected area. The results obtained after the thermal
Fig. 4. Scanning electron microscope image of the coating formed on the slate samples after passivation with Na2SiO3 and with KH2PO4.
Fig. 5. Oxidation index calculated after each cycle in the thermal cycle test, in the three groups of slate samples: untreated (Pz-Con), treated with sodium silicate (Pz-Si) and treated with potassium phosphate (Pz-P).
Fig. 6. Mean values (and standard deviations) of the percentage area of altered iron sulphides on the slate samples, obtained by image analysis, at the end of the thermal cycle, SO2 and saline spray tests in the three groups of slate samples: untreated (Pz-Con), treated with sodium silicate (Pz-Si) and treated with potassium phosphate (Pz-P).
V. Cárdenes Van den Eynde et al. / Corrosion Science 51 (2009) 2387–2392
cycle test are consistent with the oxidation indices calculated (Fig. 6). In the untreated slates (Pz-Con) the area affected by sulphide oxidation was 3.5–4 times greater after the thermal cycle test, and the increase in the area affected was signiﬁcantly lower in the passivated slates (Pz-Si and Pz-P). The area affected by pyrite oxidation after the SO2 test was 5 times greater than before the test, and was between 1.5 and 2 times larger than the area affected by the thermal cycle test (significant difference, p < 0.001), which indicates that the conditions generated in this test are more favourable to oxidation (Fig. 5). During this test an acidic environment is created that may be partially neutralized by mineral weathering, mainly of carbonates , however extremely acidic conditions may be produced on the surface of the slate, thus favouring the presence of Fe3+ and an increase in the velocity of oxidation. Despite this, the results of this test also demonstrated the protective effect induced by the treatments with sodium silicate and potassium phosphate. Despite the conditions being hugely favourable to oxidation, signiﬁcantly less of the surface was affected in the treated slates than in the untreated slates (p < 0.001). After the SO2 test, the surface affected by sulphide oxidation after the Pz-P and Pz-Si treatments was respectively 45% and 40% less than after the Pz-Con treatment. Finally, analysis of the image also showed that under the conditions generated during the saline spray test, pyrite oxidation was less notable than in the SO2 and thermal cycle tests. This test creates an environment with a high background level of electrolytes that should favour oxidation of Fe2+ to Fe3+. The precipitation of iron hydroxide further contributes to inhibiting oxidation. After exposure in the salt spray chamber, the area affected by sulphide oxidation only increased by a factor of 2.5 and there were no signiﬁcant differences among the three groups of slates (with/without treatment). 4. Conclusions Passivation of iron sulphides in rooﬁng slate, with potassium phosphate or sodium silicate, is shown here to be an effective method of inhibiting their oxidation. The method may reduce the aesthetic defects that usually develop in rooﬁng slate in situ. As a result, the commercial classiﬁcation of slates containing sulphides may be improved, which would increase the commercial yield and greatly reduce the environmental impact generated by the waste material dumped in mine spoil heaps. The main advantages with respect to other protective methods are the efﬁcacy, the simplicity of application, the low cost and the lack of any effect on the appearance of the slate. These results support the proposed testing of the method on an industrial scale. Acknowledgement This study was ﬁnanced by the Fundación Centro Tecnológico de la Pizarra de Galicia. References  J. García-Guinea, M. Lombardero, B. Roberts, J. Taboada, Spanish rooﬁng slate deposits, Transactions of the Institute of Mineral Metallurgy, Section B 106 (1997) 205–214.  M. Lombardero, J. Garcia-Guinea, V. Cárdenes, The geology of rooﬁng slate, in: C. Bristow, B. Ganis (Eds.), Industrial Minerals and the Extractive Industry Geology, Geological Society Publishing House, Bath, 2002, pp. 59–66.  V. Cárdenes, M. Lombardero, J. García-Guinea, J.C. Barros, Factores de calidad en la elaboración de placas de pizarra para cubiertas en Galicia y León, Roc Máquina 67 (2001) 90–98.  C. Monterroso, F. Macías, Drainage waters affected by pyrite oxidation in a coal mine in Galicia (NW Spain): composition and mineral stability, Science of the Total Environment 216 (1998) 121–132.
 M. Urrutia, J. Graña, R. García-Rodeja, F. Macias, Procesos de oxidación de pirita en medios superﬁciales: potencial acidiﬁcante e interés para la recuperación de suelos de mina, Cuadernos do Laboratorio Xeolóxico de Laxe 11 (1987) 131– 145.  K. Edwards, M. Schrenk, R. Hamers, J. Banﬁeld, Microbial oxidation of piryte: experiments using microorganisms from extreme acidic environment, American Mineralogist 83 (1998) 1444–1453.  W. Stumm, J.J. Morgan, Aquatic Chemistry: An Introduction Emphasizing Chemical Equilibria in Natural Waters, John Wiley & Sons, New York, 1981.  S.B. Lalvani, B.A. DeNeve, A. Weston, Passivation of pyrite due to surface treatment, Fuel 69 (1990) 1567–1569.  C.L. Jiang, X.H. Wang, B.K. Parekh, Effect of sodium oleate on inhibiting pyrite oxidation, International Journal of Mineral Processing 48 (2000) 305–318.  X. Zhang, M.J. Borda, M.A.A. Schoonen, D.R. Strongin, Pyrite oxidation inhibition by a cross-linked lipid coating, Geochemical Transactions 4 (2) (2003) 8–11.  D. Kargbo, G. Atallah, S. Chatterjee, Inhibition of pyrite oxidation by a phospholipid in the presence of silicate, Environmental Science Technology 38 (2004) 3432–3441.  Y. Chen, Y. Li, M. Cai, N. Belzile, Z. Dang, Preventing oxidation of iron sulﬁde minerals by polyethylene polyamines, Minerals Engineering 19 (2005) 19–27.  C.A. Backes, I.D. Pulford, H.J. Duncan, Studies on the oxidation of pyrite in colliery spoil. The inhibition pathway and inhibition of the ferrous–ferric oxidation, Reclamation and Revegetation Research 4 (1986) 279–291.  S.B. Lalvani, M. Shami, Passivation of pyrite oxidation with metal cations, Journal of Materials Science 22 (1987) 3503–3507.  R.V. Nicholson, R.W. Gillham, E.J. Reardon, Pyrite oxidation in carbonatebuffered solution 2. Rate control by oxide coatings, Geochimica et Cosmochimica Acta 54 (2) (1990) 395–402.  V.P. Evangelou, Potential microencapsulation of pyrite by artiﬁcial inducement of ferric phosphate coatings, Journal of Environmental Quality 24 (3) (1995) 535–542.  V.P. Evangelou, Pyrite Oxidation and its Control: Solution Chemistry, Surface Chemistry, Acid Mine Drainage, CRC Press, 1995.  V.P. Evangelou, Pyrite microencapsulation technologies: principles and potential ﬁeld application, Ecological Engineering 17 (2001) 165–178.  Z.J. Georgopoulou, K. Fytas, H. Soto, V.P. Evangelou, Feasibility and cost of creating an iron–phosphate coating on pyrrhotite to prevent oxidation, Environmental Geology 28 (2) (1996) 61–69.  V. Deodeshmukh, A. Venugopal, D. Chandra, A. Yilmaz, J. Daemen, D.A. Jones, S. Lea, M. Engelhard, X-ray photoelectron spectroscopic analyses of corrosion products formed on rock bolt carbon steel in chloride media with bicarbonate and silicate ions, Corrosion Science 46 (2004) 2629–2649.  J. Flis, The passivation of iron–carbon alloys in acidic phosphate solution and its relation to stress corrosion cracking, Corrosion Science 25 (1985) 317–330.  M. Holloway, J.M. Sykes, Studies of the corrosion of mild steel in alkaliactivated slag cement mortars with sodium chloride admixtures by a galvanostatic pulse method, Corrosion Science 47 (2005) 3097–3110.  J. Satur, N. Hiroyoshi, M. Tsunekawa, I. Mayumi, H. Okamoto, Carrier – microencapsulation for preventing pyrite oxidation, International Journal of Mineral Processing 83 (2007) 116–124.  V. Cárdenes, J. García-Guinea, C. Monterroso, R. de la Horra, Protocol for assessing the effectiveness of protective coatings for rooﬁng slate, Materiales de Construcción 58 (2008) 263–279.  P. Costagliola, C. Cipriani, C.M. Delfa, Pyrite oxidation: protection using synthetic resins, European Journal of Mineralogy 9 (1) (1997) 167–174.  Y.L. Zhang, V.P. Evangelou, Formation of ferric hydroxide–silica coatings on pyrite and its oxidation behavior, Soil Science 163 (1) (1998) 53–62.  J. Alonso-Azcárate, M. Rodas, L. Fernández-Díez, S.H. Bottrell, J.R. Mas, S. López-Andrés, Causes of variation in cristal morphology in metamorphogenic pyrite depositsof the Cameros Basin (Spain), Geological Journal 36 (2001) 159–170.  V. Cárdenes, R. de la Horra, C. Monterroso, J. García-Guinea, V. Pais, Depósitos de pizarras para cubiertas en la Península Ibérica. in: VII Congreso Geológico de España, Las Palmas de Gran Canaria, Geotemas, vol. 10, 2008, p. 453.  J.C. Barros, Nuevos datos geológicos y cartográﬁcos sobre el ﬂanco Sur del Sinclinorio de Truchas, Cuadernos do Laboratorio Xeolóxico de Laxe 14 (1989) 93–116.  J.C. Gutiérrez-Marco, C. Aramburu, M. Arbizu, E. Bernárdez, M. Hacar, I. Méndez-Bedia, R. Montesinos, I. Rábano, J. Truyols, E. Villa, Revisión Bioestratigráﬁca de las pizarras del Ordovícico Medio en el Noroeste de España (zonas Cantábrica, Asturoccidental – Leonesa y Centroibérica septentrional), Acta Geológica Hispana 34 (1999) 3–87.  CEN/TC_128, EN 12326-2:2000, Slate and Stone Products for Discontinuous Rooﬁng and Cladding – Part 2: Methods of Test, 2000.  CEN/TC_128, EN 14147:2004, Natural Stone Test Methods – Determination of Resistance to Aging by Salt Mist, 2004.  M.D. Abramoff, P.J. Magelhaes, S.J. Ram, Image processing with ImageJ, Biophotonics International 11 (7) (2004) 36–42.  R. Bogan, S. Ohde, T. Arakaki, I. Mori, C. McLeod, Changes in rainwater pH associated with increasing atmospheric carbon dioxide after the industrial revolution water, Air and Soil Pollution 196 (2009) 263–271.