I thank the CSGI consortium - University of Florence and in particular Dr. Piero Baglioni, Dr. Luigi Dei, Dr. Rodorico Giorgi for making available facilities, labor and time for the realization of this research. Dr. Andreas Küng and Dr. Cristina Mosca (SUPSI, Lugano) for their valuable collaboration in this research and Mr. Ezio Pesenti (SUPSI, Lugano) for constant help in the laboratory. We thank the restorer Marco Somaini (SUPSI, Lugano) and the company Habilis Restauro s.n.c. for providing sites for case studies. We also thank Dr. Filippo Gambinossi, Dr. Michele Baglioni, Dr. Lorenza Bernini, Dr. Marcia Carolina Arroyo (CSGI, Florence) for assistance in the laboratory.
1The presence of water is one of the main causes that influence the degradation of porous structures, such as murals, together with all the preparatory layers associated with them. The water has a very important role in the channeling, solubilization and crystallization of salts, because of its mobility in the porous system. The mechanisms which engage have a strong dependence on temperature and humidity conditions and are determined by phenomena of infiltration or capillary rise within the walls, condensation and percolation in the paint surface.
- 1 Nanorestore® is a registered trademark CSGI protecting the technology developed there and protecte (...)
2Nanorestore® - based on the mechanisms of nanolime dispersion in isopropyl alcohol1- represents an innovative solution for the consolidation of wall paintings. The treatment can restore some of the material lost due to degradation, replacing it with a substance identical to the binding medium of the artifact; therefore this treatment is entirely compatible with the mural paint.
3If, for example, the wall painting has problems of decohesion and does not have sufficient resistance to withstand cleaning, the consolidating material needs to fulfill the role of pre-consolidating. It is important to understand, therefore, whether the material used for consolidation has good results even in a system not yet stabilized. In particular, questions can be raised in cases of decohesion, whether the consolidating material does not exert an inhibitory effect in conditions of high salt contamination.
4When pre-consolidation tests were performed on a study site used during the course of Master of Conservation and Restoration SUPSI, it was observed that in a situation of high concentration of nitrate salts on the surface, the use of Nanorestore® gave very positive responses. It is assumed, therefore, that the presence of hygroscopic salts could affect the kinetics of crystallization of calcium carbonate formed from hydroxide, slowing the process of carbonation, giving the finished product the best characteristics.
5To test this hypothesis, we tried to reproduce similar situations in the laboratory, providing a standard system to work on. Samples were made to simulate a state of crumbling paint; they were contaminated with different salt solutions and consolidated with Nanorestore ®. Additionally, a series of specimens were not treated in order to have materials allowing a comparison between the treated systems and the untreated samples.
6Given the results obtained in the laboratory, we subsequently sought to verify the results on real cases. Surface heterogeneity and the impossibility of obtaining a significant amount of specimen material because of the destructive nature of the sampling was probably the reason for not achieving comparable results in situ with proven laboratory research.
- 2 Piero Baglioni, Luigi Dei, Rodorico Giorni Consorzio CSGI-Dipartimento di Chimica, Università degl (...)
7The analytical phases of this study were carried out in the laboratories of the consortium CSGI2, where the product Nanorestore® was developed.
8Standard system preparation
9In order to study the influences determined by the different temperature and humidity conditions and the presence of salts on the consolidation operated by Nanorestore®, a standard model to work on was prepared on the base of previous experience. The system was designed with the aim to reproduce the conditions of a mural painting that is crumbling.
- 3 Samples : dimensions 5cm x 5cm x 2cm, 0.5cm thick plaster and 1.5cm arriccio.
- 4 Slaked lime Candor Company Adriatic Timbers Ltd - S.S.16 Fasano Brindisi (BR) - Italy.
- 5 Geniobeton SA - 6532 Castione - Canton Ticino – Switzerland
- 6 Bresciani S.r.l. – via Breda 142 – 20126 Milano – Italy
10To obtain a system of crumbling samples3, mortars4 were prepared by increasing the volume of aggregate5 in comparison to the usual proportions and a considerable amount of water was added to the mix. The layers of plaster were not compacted with manually to facilitate drying of the mortar and ensure faster carbonation. A layer of burnt umber paint6 was applied to the dry plaster two days after application of the plaster.
- 7 Type road marking paint “Strassensignierfarbe” RUCO Company Rupf Eichstr & Co AG. 42-8152 Glattbru (...)
11After 45 days, a waterproof coating7 was applied on all sides of the samples , except on the painted surface and some of the surfaces contaminated saline solutions.
- 8 VWR International AG - Lerzenstrasse 16/18 - 8953 Dietikon – Canton Zurich – Switzerland.
12As shown in (ill. 1), five sets of samples contaminated with salts8.
13We used salt solutions selected from those most commonly found in walls and characterized by relative humidity values of medium-high equilibrium (75-80%): sodium nitrate and sodium chloride. In addition to these single salt solutions, we set up a mixture of the two salts with the addition of selected bi-hydrate calcium sulfate (gypsum), which is mainly responsible for the degradation of salt-contaminated murals.
Fig. 1 Overview of the types of samples studied
Credits : S.di Gregorio
14The solutions prepared for this study were:
- 10 Arnold Andreas, Zehnder Konrad, Crystal growth in salt efflorescence, Journal of Crystal Growth, 1 (...)
15The amount of solution applied was equivalent to 5% of the total weight of the sample. In order to ensure that the distribution of salts occurred consistently over the actual painted surfaces affected by salt damage, we adopted the system followed by Konrad Zehnder and Andreas Arnold10to study the growth of salt efflorescence in samples of mortar. In an environment maintained at a constant temperature of 20°C, a climate chamber was constructed, in which samples were placed after “salting”. Samples were kept in the chambers for a week at a constant relative humidity of 85%.
- 11 CTS S.r.l. –Via Piave 20/22 - 36077 Altavilla Vicentina – Vicenza - Italy.
16The consolidation was performed about 100 days after the completion of the samples. Nanorestore® diluted in isopropyl alcohol11 in a 1:2 ratio was applied with a brush. In total, the specimens absorbed about 3 ml of suspension, with an average weight gain of 65 grams per sample. The process of carbonation of calcium hydroxide nanoparticles was followed under two different conditions of temperature and humidity: a series of samples was subjected to a relative humidity (RH) of 50%, and another series to a RH of 85%. Thus, the consolidation process took place in different conditions: those when deliquescence of hygroscopic salts occurred and those when salts did not exhibit deliquescence. The specimens treated in this manner were kept in constant temperature and humidity conditions for 45 days according to the values shown in (ill. 1).
17To ensure reproducibility, three different samples from each system under consideration were subjected to a series of laboratory tests. The following aspects were measured:
18 In addition the study of surface morphology of the samples by electron microscopy SEM was conducted in order to make comparisons in behavior among different types of samples due to the effect of surface consolidation by Nanorestore®.
- 15 The trend is described by an exponential equation y = a [1 - exp (-b • √ t)], where the term 'a' i (...)
19The extent of absorption of water by capillary action was performed on the same sample before the "salting" treatment and later, after treatment with Nanorestore®. The extent of absorption is related to capillary porosity of the material. The more porous it is, the more water it will rapidly absorb. The continued absorption in a porous system involves filling the entire system according to kinetics related to the size and distribution of pores, which then influence the phenomenon. The impregnation of the porous structure by capillary pore water occurs first in larger pores, initially as a sort of coating of the interior walls, and finally completly filling them. The rate of absorption decreases with time and reaches an equilibrium value where the amount of water that impregnates the material is matched by gravity and evaporation, determining the values of capillary absorption, until achieving an asymptote value. The curves that are obtained in these cases have an exponential trend15. The experimental data shown in ill. 2 express the amount of water absorbed by the sample before and after treatment with Nanorestore® and was collected until the asymptotic value was reached.
Fig. 2 Capillary absorption curves of samples
Comparison of capillary absorption curves of samples contaminated with sodium nitrate before and after consolidation, maintained at a relative humidity of 85%.
Credits : S.di Gregorio
- 16 AC = A/ t*1/2, and expressed in g/(cm2 • s1/2), where A indicates the asymptotic value of water am (...)
20The final results are expressed in terms of a capillary absorption coefficient CA16. They were derived from measurements of the curves of the development of capillary water absorption and from values of the capillary absorption coefficient. The curves on all analyzed samples pointed to an apparent slowing of water absorption by capillarity and a reduction in the amount of liquid absorbed by the samples after treatment with Nanorestore®.
Fig. 3 Capillary absorption curves of samples contaminated with a mixture of salts
Before and after treatment with Nanorestore ®, maintained at a relative humidity of 85%
Credits : S.di Gregorio
21Analyzing the trend of the curves for the different types of samples before and after consolidation with Nanorestore® (ill. 3), it was found that the water absorbed by the specimens after treatment was reduced by about 1-2 grams ( about 15-20% less) and that the asymptote of the curves was achieved about 15-20 minutes later than the curves obtained before treatment. This suggests that the surface re-cohesion happened and changed the surface porosity of the specimens.
Fig. 4 Highlighted, average changes in the slope
Average expressed as a percentage of the samples after Nanorestore®
Credits : S.di Gregorio
22Average changes in the capillary absorption coefficient were determined from values obtained on each specimen before and after Nanorestore® treatment; a comparison with a baseline was performed for each investigated sample. The data obtained are represented in ill. 4; they show a greater reduction in the value of the coefficient of capillary absorption for higher values of relative humidity because of Nanorestore® treatment.
23This phenomenon is evident in all specimens except those from the gypsum polluted mixture of salts, where ratios are reversed. It can be assumed that increasing the value of relative humidity affects the stage of carbonation as to slow the process enough to ensure the formation of a very regular and compact crystalline structure of calcium carbonate; thereby the quality of consolidation is improved. A greater effect should occur (according this hypothesis) on samples contaminated by hygroscopic salts, whose deliquescence occurs in environments with relative humidity of 85%.
24This hypothesis seems to be confirmed by analyzing the behavior of these samples regarding the phenomenon of capillary rise.
25Data from samples contaminated with salt mixtures, show that Nanorestore® treatment produced a significantly smaller effect because the change in the values of capillary absorption coefficient before and after treatment was even lower than the salt-free sample.
26With this figure it is assumed that the cause of this phenomenon is due to the presence of bi-hydrated calcium sulfate. Indeed, if we consider the values of solubility, we realize that the hydroxide and calcium sulfate have close enough values. It may be assumed therefore that ion exchange reactions occur, and the subsequent co-precipitation removes hydroxide in the process of carbonation. That determines a much weakened re-cohesion effect. One can argue that consolidation with Nanorestore® is not minimally affected by the presence of saline solutions with RH equilibriums of less than 85%. At such humidity values chlorides and nitrates deliquesce, and then the wall has a high moisture content. The bi-hydrate calcium sulfate, however, imparts a partial inhibitory effect, at least with the Nanorestore® concentrations and quantities, which are usually used.
27The contact angle measurement allows us to evaluate the wettability of a surface and also provides guidance on specimen surface tension and capillary phenomenon dependent on surface features. Considering the action of Nanorestore®, recording data values from this analysis led to the search, using one of two methods that could assess changes of compactness and regularity of the surface determined by the consolidation and, in practice, give feedback on the recohesion of the surface material.
Fig. 5 Highlighted, averaged contact angle of spécimens
Specimens were consolidated with Nanorestore®
Credits : S.di Gregorio
28The collected data (ill.5) show some correspondence, in trend, with the results obtained in studying the kinetics of capillary absorption.
29As evidence that the treatment affects the paint surface, the samples that were not treated with Nanorestore® have an increased surface wettability comparing to all other tested systems (possibly because the samples have higher surface porosity). The value of the contact angle of samples that were not treated is not so different from samples contaminated by mixtures of salts under conditions of 85% RH. In general, a lower wettability of the surface (and therefore a better strengthening effect) was obtained in samples not contaminated by salts, and in samples with sodium nitrate, where the treatment was conducted in an environment of 85% RH.
30Intermediate results were obtained on samples treated with Nanorestore® and contaminated by nitrate and sodium chloride, and by mixtures of salts, exposed to 50% RH. In general, the treatment seems to have compacted the surface by reducing the surface tension, making it less wettable as shown by the increased value of angle for almost all samples.
31It should be noted, regarding the influence of hygrothermal factors, that an increasing value of the high humidity conditions (for which is assumed to obtain a better surface carbonation) was observed only for specimens contaminated with sodium nitrate. An opposite trend is observed in the presence of mixtures of salts (consisting of nitrates, chlorides and sulfates).
Fig. 6 Evolution of water loss
Vapor permeability for different types of treated samples.
Credits : S.di Gregorio
32The data derived from the vapor permeability measurements gave indications about the characteristics of the samples’ porosity. The trend of water loss through permeability was determined by measurements for each studied sample. We obtained by these data a trend that is representative of the average values of three samples for each considered type. Ill. 6 shows appreciable differences between samples consolidated in environments of 50% RH compared to those consolidated at 85% RH. In particular, a smaller water loss by permeability in samples consolidated at 85% RH indicates a greater compactness.
33These results are particularly interesting for the samples contaminated by sodium nitrate, whose consolidation showed characteristics indicating particularly low water loss because of permeability. An exception is found for samples contaminated by sodium chloride, where the water loss by permeability is particularly significant in relation to results obtained by other measurements.
34A morphological analysis was also carried out on treated surfaces. We tried to highlight the morphological characteristics of painted surfaces by SEM microscopy for each type of specimen.
35A scanning electron microscope Stereoscan 360 (Cambridge-Oxford) was used for this purpose. It was possible to observe the irregular surface of the non-contaminated samples and a lack of consistency of the paint layer on the microscope images (ill. 7). The grains of pigment used in painting and fillers are visible in some areas because of the lack of binding carbonate, due to the use of lean mortar.
Fig. 7 Scanning Electron Microscope (SEM)
Surface morphology observed at 1200 magnification. Uncontaminated sample not consolidated with Nanorestore®.
Credit : S.di Gregorio
Fig. 8 Scanning Electron Microscope (SEM)
Surface morphology observed at 1200 magnification. Uncontaminated sample consolidated with Nanorestore® (50% RH)
Credits : S.di Gregorio
Fig. 9 Scanning Electron Microscope (SEM)
Surface morphology observed at 1200 magnification. Uncontaminated sample consolidated with Nanorestore® (85% RH)
Credits : S.di Gregorio
Fig. 10 Scanning Electron Microscope (SEM)
Morphology of surfaces observed at 1200 magnification. Sample contaminated with sodium nitrate and consolidated with Nanorestore® (85% RH)
Credits : S.di Gregorio
Fig. 11 Scanning Electron Microscope (SEM)
Surface morphology observed at 1200 magnification. Sample contaminated with sodium chloride and consolidated with Nanorestore® (85% RH)
Credits : S.di Gregorio
Fig. 12 Scanning Electron Microscope (SEM)
Surface morphology observed at 1200 magnification. Sample contaminated with mixtures of salts and consolidated with Nanorestore® (85% RH)
Credits : S.di Gregorio
36When images of uncontaminated samples are compared, we can find after consolidation (ill. 8) a substantial "filling" of the gaps between the grains with a certain strengthening effect that should be attributed to the carbonation of calcium nanoparticles on the surface .The strengthening effect seems to be greater in conditions of relative humidity of 85% (ill. 9); in fact, the surface shows a greater continuity in the paint layer. This result was found regardless of temperature and humidity conditions in the two samples contaminated with sodium nitrate (ill. 10). A similar result was not observed on the samples previously contaminated by sodium chloride (ill. 11). The effect is even less evident for the samples "salted" by mixtures (ill. 12) containing sulfate salts.
37Comparing all the images, we can state that, in general, the values of relative humidity of 85% (being kept constant during the process of the Nanorestore® reaction) lead to carbonation of nanoparticles, which produces a more compact layer of paint because of the neoformation of calcium carbonate.
38The results of adhesion tests performed by the Scotch tape test did not permit to characterise the treatment regarding the cohesion of the paint film. This is due to the fact that such specimens were not suitable for sensitivity testing.
39The work showed some problems about the effectiveness of Nanorestore® consolidation in samples contaminated by a mixture of salts. An investigation was performed, testing whether the inhibitory effect on consolidation was due to the presence of calcium sulfate. In particular, it was conjectured that the value of solubility of calcium sulfate, very close to that of calcium hydroxide, might lead to an exchange reaction between the salt and the Nanorestore® product.
- 17 Nanorestore : 2 parts by volume, suspension of nanoparticles of Strontium : 1 part by volume, isop (...)
- 18 CTS S.r.l. –Via Piave 20/22 - 36077 Altavilla Vicentina – Vicenza
- 19 Concentration 8gr /l
40It was decided for that purpose to select 6 samples among those already prepared in the laboratory with a 5% weight of salt in the sample. They were contaminated by a saturated solution of calcium sulfate and were consolidated with a dispersion17 of Nanorestore® that was diluted in isopropyl alcohol18 and mixed with a dispersion of strontium nanoparticles19. These nanoparticles are engaged in an exchange reaction with calcium sulfate, thereby permitting calcium nanoparticles to lead to consolidation without any other interference.
41After consolidation, the specimens were kept in an environment of 85% RH and 20° C for about 45 days.
42The samples have been studied at the end of the consolidation process by analyzing capillary absorption, measurement of the angle contact and vapor permeability. The results actually show a positive response to the consolidation with Nanorestore®, due to the action of strontium nanoparticles on the exchange reaction with the bi-hydrate calcium sulfate.
43If the obtained values are compared with those of previously analyzed samples, one may highlight a response to consolidation similar to that found on the untreated samples, always considering the margin of experimental error. One can then associate such a result to what has been achieved in a situation that is not influenced by salts.
44The research shows that the influence of salts in porous materials results in some inhibition in the consolidation process with the exception of hygroscopic salts, when the consolidation takes place in a high relative humidity. Some doubts exist regarding the behaviour of sodium chloride, due to discrepancies between the results of different analysis. This positive behavior of salts on consolidation can be explained by the fact that, under conditions of high relative humidity, the hygroscopic salts that contaminate specimens deliquesce, subsequently causing a high moisture content not only on the surface (as in the untreated samples) but in the internal system. One can speculate whether this affects the stage of carbonation to the extent that the process turns to be slow enough to ensure the formation of a very regular and compact crystalline structure of calcium carbonate, thereby improving the quality of consolidation. This process would lead to benefits that would counterbalance problems related to the presence of salts as well as, in some cases, it would probably provide an advantage in the kinetic process of consolidation.
- 20 Solubility in water at 20 ° C : Ca(OH)2 = 1,7 g /l CaSO4.2H2O = 2,4 g /l
45The problems affecting the effectiveness of consolidation, resulting from the presence of mixtures of salts, have to be attributed to bi-hydrate calcium sulfate salt. The solubility values for calcium hydroxide and calcium sulfate are so close20that one can assume ion exchange reactions take place as well as subsequent co-precipitation, which would remove the calcium hydroxide from the process of carbonation.
46A solution to this problem was given by the use of alcoholic dispersions of strontium hydroxide. This, being more reactive to the bi-hydrate calcium sulfate than the calcium hydroxide, allows Nanorestore® to perform consolidation without the process being affected by plaster.
47These data show that Nanorestore® treatment on murals that are affected by the presence of hygroscopic salts has no contraindications in conditions when relative humidity is above 85%. The water that is absorbed by condensation of the salt species seems to have a positive influence on the process of carbonation of the involved particles. In the case of plaster, the product Nanorestore® shows a lower efficiency, and then a more consistent application in terms of quantity of material or of a support of alcoholic dispersion of strontium hydroxide is necessary.