Hydrophobic Treatments for Stone Conservation

Hydrophobic Treatments for Stone Conservation

INFLUENCE OF THE APPLICATION METHOD ON PENETRATION, DISTRIBUTION AND EFFICIENCY

Giulio Cesare Borgia, Mara Camaiti, Fanny Cerri, Paola Fantazzini and Franco Piacenti

The performance of a water-repellent treatment applied to a stone surface is generally evaluated without considering the depth of penetration and uniformity of distribution of the active product in the treated porous material. Magnetic resonance imaging (MRI), a non-destructive methodology, has recently made it possible to visualize the presence and distribution of different hydwphobic polymers in rocks, thanks to their water-repellent properties. The authors show, through the use of this technique, how the penetration depth and the distribution of a water-repellent material in a stone arc affected, and therefore may to some extent be directed, by the application method. Two commercial Hydrophobic polymers (Paraloid B-12 and Silirain 50) frequently used in the restoration/conservation of stone artifacts have been tested on a biocalcarenite (pietra di Lecce).

INTRODUCTION

The protection of stone can be performed by application of water-repellent polymers to the stone surface [1-5]. The efficacy of the treatment applied is commonly evaluated on the basis of water-repellence determinations such as contact angle and capillary water absorption of the upper layers of the treated rock. These techniques provide an indication of the efficacy of the treatment in terms of reduction of water uptake through the surface examined, but they do not allow any insight into the penetration of the polymer into the structure, which is relevant from the point of view of both durability and efficacy [6-10]. It is generally thought that the higher the amount of protective material absorbed, the higher the efficacy and durability of the treatment. On the other hand, if the polymer accumulates inside the pores of the superficial layers, a significant reduction in the permeability of the rock to water vapour will take place and/or formation of mechanical inhomogeneitycan cause further deterioration upon environmental stress [11-13],

The amount of product absorbed and its distribution in a stone strongly depend on the structure of both the product and the substrate and on the application technique [14]. The authors therefore thought it interesting to determine the depth of penetration and the distribution of products applied to a stone surface in protection and/or consolidation treatments, when the application method was varied.

The techniques traditionally in use to get information on the penetration depth of hydrophobic products, such as the visual identification of the water-repellence on an internal surface of a stone after cutting [6], the analysis of the consistency of the material through drilling tests [15, 16], the iodine vapour technique [17] or the 'Karsten-tube' method [18], are destructive and/or not sufficiently accurate, showing low reproducibility.

'H magnetic resonance imaging (1H-MRI) allows the visualization of the presence and distribution of liquid water inside a porous opaque medium [19-22]. This new methodology was used in the

present work to follow the water uptake by capillarity into a highly porous biocalcarenite, pietra di Lecce, and to compare the results obtained before and after treatment with hydrophobic products when different application techniques were used. These determinations were performed by applying the polymeric materials, having different structures and molecular size, from solutions of different concentrations, and using alternatively a brush or the capillary absorption technique. The final performance of the treatment was evaluated by both MRI and traditional analytical determinations, such as gravimetric tests for capillary water absorption and water vapour permeability. The aim of this work is to verify the great influence of the application method of a product on the final performance of a protective and/or consolidation treatment, and to relate it to the different polymer distribution achieved.

MATERIALS AND METHODS

Two hydrophobic products were employed tor treatments, both widely used in the conservation of stone artifacts: Paraloid D-72 (poly ethylmethacrylate/ co-methylacrylate 70/30) from Röhm & Haas, and Silirain 50 (polymethylsiloxanic oligomer 5% in white spirit) from Rhône-Poulenc. The treatments were applied on prismatic samples (5x5x2 cm) of pietra di Lecceonto a 5 X 5 cm surface. Paraloid B-72 was applied as chloroform solutions, while Silirain 50 was applied as supplied by the producer (Table 1). Each treatment was applied on three samples with the same operative conditions. The performance of the treatments was evaluated by both traditional methods and MRI (see Appendix).

Evaluation of performance by traditional methods

The main parameters to be considered in order to evaluate the efficacy and compatibility of a hydrophobic treatment on stone are the water-repellence achieved by the external surface of the rock and its permeability to water vapour after application of the product and evaporation of the solvent [5].

In order to evaluate the protective efficacy of the treatment against liquid water absorption, the following traditional test was performed:

a Capillary water absorption, according to UNI 10859 15, 23]. The protective efficacy (PE%) of the hydrophobic treatment was calculated from the amounts of liquid water absorbed by the unit surface of the rock in 20 minutes, PE% (20 min), or in three hours, PE% (3 hrs), before (A0) and after (A1) the treatment:

PE% = 100 (A0-A1)/A0

The reproducibility of the determination on each sample was within 4%, while the representativity of samples, due to non-homogeneity of the rock, tested on three different samples, was ±14%.

b Kinetics of liquid water absorption by imbibition through the treated face of the samples. The test was canned out following the same procedure used in (a), but the samples were kept in contact with liquid water for a longer time (up to seven days).

The residual permeability to water vapour of the treated rock was measured by the 'bicchierino' method [24]. The water vapour permeability of rock samples

Table 1 Treatments with Paraloid B-72: application method and performance evaluated by traditional techniquesa

aEach value is calculated as the mean value of three samples and the error associated was expressed as ½(max value - min value).

bPE% (20 min) = protective efficacy percent of the treatment calculated after 20 minutes of capillary water absorption.

cPE% (3 lus) = protective efficacy percent of the treatment calculated after three hours of capillary water absorption.

dRP% — percent residual permeability

(5x5x2 cm) was evaluated by measuring the mass of water vapour passing through the unit surface in 24 hours, at 30°C, under controlled conditions. The mass of water evaporated in 24 hours was measured when the system had achieved equilibrium. In a typical 'permeability test' the mass change of the bicchierino/stone sample/water system is recorded every 24 hours for 8-10 days, and equilibrium is considered to be achieved when the mass of water evaporated in 24 hours remains constant or differs by less than 5% in two consecutive weighings.

The parameter RP% (percent residual permeability) was calculated from the sample permeability, after (P1) and before (P0) the treatment, by the formula:

RP%= 100 (P1/P0)

The maximum standard deviation achieved on each sample for RP% was 1 %. The representativity of the samples was ±8%.

Image acquisition

The 1H magnetic resonance imaging technique allows the visualization of the presence and the spatial distribution of the 'H nuclei of liquid water inside porous materials. In order to get indirect information on the penetration depth and distribution of a hydro-phobic polymer inside a stone, the samples must be exposed to a water source and checked for 1H nuclei after different periods of water absorption. The complete absence or a different intensity and distribution of the 1H signal from water in the stone after treatment, compared to the results obtained before the hydro-phobic treatment, provides evidence of the presence and efficacy of the polymer inside the stone. As reported in [20] and [21], the MR-images were acquired after the rock samples were left in contact with a water source for 20 minutes, one, two, three and 24 hours, and seven days, respectively. For the same sample, a series of images was obtained for each of the following stages:

1 Before treatment, in order to check the free diffusion of the liquid water into the selected lithotype.

2 After treatment, through the face of the sample opposite to the treated one {untreated face). In many cases, the untreated face is not (or scarcely) water repellent, because the hydrophobic product did not reach it (low penetration depth). In such conditions, liquid water is able to penetrate into the sample and be detected.

3 After treatment, through the treated face, which is usually highly water-repellent after short times of contact with the water source, but can allow water to penetrate during long exposure times.

Only one sample for each treatment was analysed by MRI.

RESULTS AND DISCUSSION

Treatments with Paraloid B-72

Paraloid B-72 was applied in two different concentrations (1% and 3% by weight chloroform solutions), by brush or by capillary absorption (Table 1). The data reported in Table 1 show that the amount of polymer absorbed by the rock, when applied by brush, is very low (P1) if compared with the amount absorbed by capillarity from a solution at the same concentration (P2). For the treatment by capillary absorption, an increase in the concentration causes only a modest increase in the mass of product absorbed by the rock (P2, P3).

The protective efficacy of the treatments (PE%) seems to be highly dependent on the method of application of the product, rather than on the amount of polymer applied. The treatment by brush, in fact, shows the best value of PE%, both in short (20 minutes) and long (up to seven days) tests, despite the low amount of product applied (Table 1 and Figure 1). The treatments performed by capillary absorption, P2, P3 and P4, in which comparable masses of polymer were applied, show in general lower PE% values (Table 1). This behaviour is probably due to a different distribution of the polymer in the rock material. In the case of the

Figure 1 Kinetics of capillary water absorption of samples treated with Paraloid B-72. 'Here and in the following figures, the mass of water was absorbed through a surface of 25 cm2.

application by brush, the high efficacy may be related to the formation of a thin layer of the water-repellent polymer on the surface, while in the case of the application by capillary absorption, the product is very likely distributed over a wider region inside the stone.

The residual permeability of the treated stone to water vapour (RP%) shows high values in all the cases observed, except for treatment P4, indicating in this case a different distribution of the product with respect to that obtained for treatment P3 (Table 1).

The MR-images of the samples, acquired after water capillary absorption through the untreated face, show different behaviours:

• The diffusion of water in the samples treated by brush with a dilute solution (treatment PI) is similar to that observed for the untreated stone (Figure 2).

• The water uptake is slower in the samples treated by capillary absorption with the 1% solution (treatment P2), if compared with the untreated rock. After three hours, the upper portion of the sample still appears dark, indicating the presence and activity of the polymer in that region of the rock. After 24 hours, water has diffused into the whole sample but the penetration depth and the profile distribution of the polymer are still visible, through different intensities of the 'H signal (Figure 3b).

• In the case of treatment P3, only after seven days is it possible to detect the presence of water in the whole sample (Figure 4b). In 24 hours, water diffuses into only part of the rock, indicating, by difference, the presence and activity of the hydrophobic product in the residual part. The depth of penetration of the polymer is particularly evident in this case, thanks to a local accumulation of water along the distribution profile of the product.

• When treatment P3 is followed by a solvent uptake (treatment P4), no water is absorbed through the face opposite to the treated one, even after seven days of contact with a water source, showing a displacement of the polymer towards the untreated face (images not shown). Such displacement may have caused a partial pore blockage, as indicated by the low value of residual permeability.

The MR-images of the same samples, when water is absorbed through the treated face, confirm the data obtained by gravimetric methods (Table 1 and Figure 1).

For treatments P1 and P3, no MR signal was visible at anv time. The mass of water absorbed through the

Figure 2 MR-images and quantitative mass determination of the capillary water absorption in a sample of pietra di Lecce treated by brush with Paraloid B-72 (treatment PI): (a) before treatment, (b) through the face opposite to the treatment.

Figure 3 MR-images and quantitative mass determination of the capillary water absorption in a sample of pietra di Lecce treated by capillarity with Paraloid B-72 (treatment P2): (a) before treatment, (b) through the face opposite to the treatment, (c) through the treated face.

Figure 4 MR-images and quantitative mass determination of the capillary water absorption in a sample of pietra di Lecce treated by capillarity with Paraloid B-72 (treatment P3): (a) before treatment, (b) through the face opposite to the treatment.

Figure 5 MR-images and quantitative mass determination of the capillary water absorption in a sample of pietra di Lecce treated by capillarity with Paraloid B-72 (treatment P4): (a) before treatment, (b) through the treated face.

treated face at short times, in fact, is very low. After seven days, on the other hand, the small amount of water absorbed is probably distributed within the whole volume ot the samples, in capillary pores of small dimensions, causing a very low proton density and short relaxation times, that MRI can hardly detect. In these cases the images are completely dark and they are not shown.

When the protective efficacy of the treatment is low (treatments P2 and P4), MR-images show the uptake of water and its distribution into the rock (Figures 3c and 5b).

The data reported above might lead one to believe that the application of dilute solutions ot B-72 by brush

Figure 6 Kinetics of capillary water absorption in a sample treated by brush with Paraloid B-72 (treatment P1). Determinations made after (A) no wetting-drying cycles, (B) one wetting-drying cycle, (C) three wetting-drying cycles.

is a valid protective treatment for stone surfaces. However, other observations must be considered. As already proved and discussed in [21], the performance of treatments with B-72 is lowered by the action of wetting-drying cycles. When the samples are made to absorb water by capillarity and then allowed to air-dry to constant weight, they show a kind of behaviour that is particularly evident in the case of the treatment by brush (Figure 6).

Treatments with Silirain 50

The low penetration observed for the treatments with Paraloid B-72 can be partially ascribed to its high molecular dimensions. Treatments carried out with the same application techniques (brush and capillary absorption) with Silirain 50, a low molecular weight product, show a lower influence of the application method on the penetration depth of the product and on its performance. The protective efficacy of the treatment and the residual permeability to water vapour of the treated rock do not seem to be highly affected by the amount of product absorbed, except in the case of treatment S4, where the long time of application of the product determines the absorption of a high quantity of product and the residual permeability to water vapour is significantly reduced (Table 2 and Figure 7). In the case of treatment S2, a higher water uptake is observed only at long times of absorption (Figure 7), owing to a very low amount of active product applied (80 g.m-2).

The penetration depth of the product is comparable for treatments SI, S2 and S3, as shown by the MR-images (Figures 8b, 9b, 10b). The capillary rise of water through the face opposite to the treatment is slower than

Table 2 Treatments with Silirain 50: application method and performances evaluated by traditional techniquesa

aEach value is calculated as the mean value of three samples and the error associated was expressed as 'A(max value - min value).

bPE% (20 min) = protective efficacy percent of the treatment calculated after 20 minutes of capillary water absorption.

cPE% (3 hrs) = protective efficacy percent of the treatment calculated after three hours of capillary water absorption.

dRP% = percent residual permeability

Figure 7 Kinetics of capillary water absorption of samples treated with Silirain 50.

Figure 8 MR-images and quantitative mass determination of the capillary water absorption in a sample of pietra di Lecce treated by brush with Silirain 50 (treatment S1): (a) before treatment, (b) through the face opposite to the treatment.

in the untreated rock. Unlike Paraloid B-72, Silirain 50 provides very regular front profiles of the capillary water rise. In the case of treatment S4, the amount of product applied and its penetration depth are such as to determine a very high water-repellence of the untreated face, and MRI analysis did not show any 'H signal inside the volume of the treated rock even after seven days of contact with the water source (images not shown).

As can be predicted from PE% data, the low amounts of water diffusing through the treated face of the samples in the case of treatments SI, S3 and S4 are not detectable by MRI. The images obtained for treatment S2 are shown in Figure 9c.