ACS Publications
[Journal Home Page] [Search the Journals] [Table of Contents] [PDF version of this article] [Download to Citation Manager]

Langmuir, 23 (17), 8681 -8685, 2007. 10.1021/la701292d S0743-7463(70)01292-1
Web Release Date: July 11, 2007

Copyright © 2007 American Chemical Society

Nanomagnetic Sponges for the Cleaning of Works of Art

Massimo Bonini, Sebastian Lenz, Rodorico Giorgi, and Piero Baglioni*

Department of Chemistry and CSGI, University of Florence, via della Lastruccia 3 - 50019 Sesto Fiorentino, Florence, Italy

Received May 3, 2007

In Final Form: June 15, 2007

Abstract:

This letter reports the synthesis and characterization of functionalized magnetic nanoparticles associated with chemical gels and their application to the conservation of cultural heritage. Magnetic nanoparticles, which are associated with acrylamide ethylene oxide polymers, produce a sponge that can be loaded with oil-in-water microemulsions, forming a magnetically responsive gel-like system and acting as a permanent hydrogel. The magnetic gel-like system can be used for specific applications in detergents or in the release of the loaded material. The system can be magnetically manipulated and cleaned from the loaded materials and then dried and reused for a different application. We report an important application of this new nanomagnetic responsive material in the field of cultural heritage conservation.


Gel systems have been used in several applications since the middle of the last century. Nowadays, they pervade our world and are very popular in cosmetics, detergents, food, and biomedical applications and in producing numerous products as nanoparticles, dyes, and advanced ceramics using the sol-gel method. Gels can be physical, where the fluid state can usually be recovered by changing the temperature, or chemical, where the gel state is obtained through a polymerization process.

An interesting application of physical gels in the field of conservation science was first proposed by Wolbers for the cleaning of artifacts (i.e., oil paintings). The use of solvents in their gelated states partially overcomes the drawbacks in using pure organic solvents to clean painted surfaces that deeply penetrate the painted layer, possibly causing damage. Moreover, the selectivity and the control of the cleaning procedure with neat solvents are difficult to achieve,1-4 making problematic, in many cases, the selective removal of undesired layers of material from easel and wall paintings. Gels decrease the capillary penetration of the solvent into the artifact and solvent evaporation because it is immobilized within the gel network.5-7 The gel cleaning technique is very versatile because several organic solvents and many selective cleaning reagents (such as enzymes and chelating agents)6 can be "gelated". Therefore, it is possible to achieve good selectivity in the removal of dirt, varnishes, and synthetic polymers largely used in the past for the protection of paintings and nowadays recognized as being deleterious to the preservation of artwork.

Unfortunately, present gel technology is not without drawbacks. The removal of gellant residues from the surfaces of works of art usually requires the application of solvents,6 but very often gellant residues remain on and beneath the painted surface because of the high viscosity of gels, which makes them difficult to remove by simple washing. Moreover, they are often inappropriate for porous substrates such as easel and wall paintings. These are layered structures with variable composition and a painted outer layer that is usually less than 1 to 2 mm thick. The large porosity favors the entrapping of the solvent inside the pores, and cleaning is a very difficult task. In this case, the gel formulation does not represent a major advantage compared to pure organic solvents that are still used to clean painted surfaces and to remove undesired layers of material from easel and wall paintings.

Recently, the synthesis and characterization of new chemically responsive organogels was reported.8 A polyallylamine (PAA) gellant was used to prepare a new class of gels9 that behaves as a "smart" system that can switch from solution-type to gel-type rheological behavior by simple chemical action. PAA gels were used for the cleaning of easel paintings with a satisfactory removal of gel residues. However, these systems use weak acids to destroy the gels' network, and although the chemical action is moderate, it must be controlled very carefully, especially if applied to carbonatic material as wall paintings.

We pioneered the use of nanosystems in the conservation of cultural heritage, and in particular, we devised micellar solutions, microemulsions, and recently the use of responsive gels for cleaning artifacts.9-12 Oil-in-water microemulsions (E) and micellar solutions13 were used to solubilize and remove aged Paraloid B72 resin (acrylic and methacrylic copolymer) from wall paintings that could not be removed by using traditional cleaning methods.10-12

In this letter, we report on a new magnetically responsive compartmentalized nanosystem. Functionalized magnetic nanoparticles have been chemically incorporated into a polyacrylamide gel structure to obtain a chemical sponge that can be loaded with microemulsions or micelles solutions (sketched in Figure 1, hereinafter called a nanomagnetic gel) and can be used for several different applications (i.e., biotechnology, cosmetics, detergents, etc.). An important application to the conservation of cultural heritage is highlighted here. We show how oil-in-water (o/w) microemulsions can be associated with nanomagnetic sponges to obtain a gel-like system for the cleaning of historical stones or painted surfaces without undesired residuals on the works of art. The nanomagnetic gel (i.e., the sponge loaded with a microemulsion or micellar solution) can be shaped as desired and applied to a specific area with fine spatial control of the area. In addition, the nanomagnetic gel adhesion to the artifact can be modulated by controlling the polymer cross linking during the chemical synthesis of the sponge.


Figure 1 Schematic representation of the process of loading the microemulsion into the nanomagnetic sponge structure. In the inset is a representation of cross-linked particles (black spheres) bonded to MA residues (blue rectangles) and a PEG chain (red line); arrows represent the binding group to the polymer network of acrylamide and bis-acrylamide.

To the best of our knowledge, the nanomagnetic sponge loaded with a micelle or microemulsion system to give a responsive chemical gel represents one of the most advanced, versatile systems for cleaning works of art, avoiding any side effects. These responsive compartmentalized nanosystems are a real breakthrough in the conservation of cultural heritage and represent one of the few real applications of nanotechnology. They will have a dramatic impact on the conventional methods used in the conservation field and in several other fields where fine tuning of the release or uptake of confined material is required.

The nanomagnetic sponge was obtained by cross linking magnetic nanoparticles through a polymer network based on polyethylene glycol (PEG) and acrylamide. A ferrofluid consisting of positively charged CoFe2O4 nanoparticles in water (0.1 g/mL, 8 nm diameter) was obtained with minor modifications, according to the method developed by Massart.14-17 To cross link the particles, a PEG-based polymer (MA-PEG) was prepared through the esterification of polyethylene glycol with maleic anhydride (MA). The resulting MA-PEG molecules consist of carboxylic groups at both ends of the PEG chain. It has been previously shown that the binding reaction between magnetic nanoparticles and carboxylic acids take the complete coupling of the carboxylic headgroup over the surface of the particles.18 Therefore, MA-PEG was directly reacted with the magnetic nanoparticles, forming a slightly viscous magnetic fluid. Acrylamide and N,N'-methylene bisacrylamide solutions were then added, and the polymerization reaction was carried out at 42 C for 4 h, with ammonium persulfate as the radical initiator. Magnetic nanoparticles are embedded in the gel structure via reacting the double bonds of acrylamide or N,N'-methylene bisacrylamide with the double bond resulting from the esterification of polyethylene glycol with maleic anhydride. (For further information, see Supporting Information.)


Chart 1. Reaction Scheme of the PEG-Based Polymer (MA-PEG) Used to Cross Link Particles (Top)

At the end of the polymerization process, two phases were obtained: an upper transparent liquid phase and a magnetic black phase formed on the bottom of the vial. The magnetic phase was simply collected by decanting the liquid phase with the aid of a permanent magnet. The formed nanomagnetic sponge was found to have a pH of ~1. It was washed with distilled water until pH 5.5 was reached. To highlight the effect of the particles on the structure of the polymeric sponge, a reference sponge without magnetic nanoparticles was prepared using the same synthetic procedure. The nanomagnetic sponge prepared as described shows behavior that is typical of permanent hydrogels.

The gel formation is reversible. In fact, the nanomagnetic gel can be freeze dried to obtain a magnetic powder that can be rehydrated to reform the gel, as for permanent hydrogels. The maximum swelling of the sponge (maximum amount of loaded water) can be simply achieved by adding the sponge to a large amount of water. Under these conditions, the nanomagnetic gel sinks to the bottom of the container, but it does not break or dissolve and it can be easily collected using a permanent magnet or plastic tweezers. Because the sponge is a tightly bonded structure, the consistency of the nanomagnetic gel is hard enough to be handled with tweezers and cut with a knife or scissors. This behavior makes the gel suitable for several applications.


Figure 2 SAXS curves of the microemulsion before being loaded into the nanomagnetic gel structure and after being recovered with the aid of a permanent magnet.

For specific application in the field of conservation science, we loaded the sponge with a o/w microemulsion prepared according to the literature procedure.10,13 The co-surfactant (1-pentanol) was added dropwise to an aqueous solution of sodium dodecyl sulfate at room temperature and stirred until a clear solution was formed. An oil mixture of a nitrodiluent (commercial mixtures of xylenes and chloro derivatives, which are solvents that are used by restorers for paraloid removal), and p-xylene was finally added at room temperature until a stable system was formed. The microemulsion is constituted of nanometric droplets (about 4 nm in radius, as obtained from small-angle X-ray scattering) dispersed in the continuous water phase. As in the case of water, the loading of the microemulsion into the sponge structure was performed by simply adding a piece of nanomagnetic sponge to a beaker containing the microemulsion. After 10 min, the gel was collected with plastic tweezers and used to clean the artwork. It is worth noting that the microemulsion droplets are small enough to diffuse easily through the gel network. In fact, polyacrylamide gels are known to present pores with sizes on the order of hundreds of nanometers.19 Moreover, it has been previously shown20 that polyacrylamide gels with a chemical composition very similar to the one we discuss here are characterized by the presence of inhomogeneities with a large range of sizes, as estimated from 10 to 1000 Å. (For details of the chemical composition, see Supporting Information.) Therefore, the cleaning process mechanism of micellar solutions or microemulsions, themselves or when they are loaded into the magnetic gel, is the same. The inhomogeneities and the pores of the polyacrylamide gel allow the microemulsion or micelle droplets to migrate to the surface of the gel that is in contact with the artwork, solubilize the material to be removed, compartmentalize it in the droplet, and transfer it into the gel structure.

Because the microemulsion could damage the gel structure over time, the gel was always stored as dried powder or a hydrogel. For this purpose, the microemulsion can be removed from the nanomagnetic sponge after extensive washing with water or simply by magnetic squeezing. Once cleaned, the nanomagnetic gel can be dried and reused.

Interestingly, when the nanomagnetic sponge is loaded with the microemulsion, both the gel and microemulsion retain their structure. To show this, we have studied by SAXS the microemulsion before being loaded into the nanomagnetic sponge and the microemulsion recovered from a microemulsion-loaded nanomagnetic gel. This procedure was imposed by the dominating scattering of the magnetic nanoparticles embedded into the polymeric matrix of the gel. The microemulsion was recovered by applying a 1.4 T magnetic field to the nanomagnetic gel, which causes the shrinkage of the gel and the release of the microemulsion. SAXS results (Figure 2) demonstrated that the structure of the microemulsion remains unchanged. In fact, SAXS curves are almost identical, with a moderate shift in the position of the interaction peak in the scattering vector that can be ascribed to a slight change in the microemulsion concentration once the microemulsion is recovered from the gel. (The Q-vector shift reported in the Figure corresponds to about 3 Å in the correlation distance.)

To evaluate the maximum amount of water that could be loaded into the gel structure before detecting any phase separation, a known amount of powder obtained from the freeze drying of the nanomagnetic gel was fully hydrated. In both the reference (the gel without magnetic nanoparticles in the framework) and the magnetic gel, the amount of water in the fully hydrated gel was slightly higher than 90% by weight. No significant differences were found for either the reference or the magnetic gels loaded with the microemulsion. The capability of the gel to retain the solvent in its structure was studied by mixing 90 wt % solvent (either water or a microemulsion) and 10 wt % sponge obtained from the freeze-drying of the gel. The samples were stored in a humidity-controlled chamber (relative humidity 50%), and their weight was checked during time. In this way, dehydration curves were obtained for the hydrated reference gel, the hydrated magnetic gel, the microemulsion-loaded reference gel, and the microemulsion-loaded magnetic gel. In Figure 3, the percentage by weight of the solvent against storage time is shown. No differences between the reference and the magnetic gel, loaded with water or the microemulsion, have been found, suggesting that the presence of magnetic particles does not affect the water-retention properties of the gel. However, the similarity in the dehydration behavior indicates the active role of magnetic particles in solvent adsorption. In fact, the composition of the sponge is different between the reference and the magnetic gel. The inorganic content of the magnetic powder (the dried nanoparticle sponge) is around 28%, whereas the reference powder is entirely organic. Interestingly, for loaded nanomagnetic gels we should expect a very fast decrease in the solvent content corresponding to the evaporation of the volatile organic component in the microemulsions (i.e., p-xylene and nitrodiluent). This effect is indeed reported in Figure 3, but it is much slower than expected for a pure solvent or a microemulsion: the solvent content of microemulsion-loaded gels and water-loaded gels equalizes after about 10 days, indicating that the evaporation of the volatile fraction is consistently slowed down by the confinement effect played by the gel structure. In terms of the usability of the microemulsion-loaded nanomagnetic gel, these results show that the system is stable for several days (i.e., much longer than the typical application time of such systems in cultural heritage conservation, usually from a few minutes to hours).


Figure 3 Dehydration curves of both magnetic and reference gels, loaded either with water and a microemulsion.
Figure 4 Sequence illustrating the removal of the microemulsion-loaded nanomagnetic gel from the surface of marble.
Figure 5 Microreflectance FTIR spectra: (a) quarry marble sample without any treatment on the surface; (b) the same surface treated with a very thick layer of paraloid B72; (c) reference sample of paraloid B72 as a pellet; and (d) marble surface after complete removal of the thick paraloid B72 layer using the nanomagnetic gel. The thick layer of paraloid B72 completely hides signals coming from calcium carbonate (calcite) (image b). The efficacy of the novel gel is highlighted by the complete disappearance of polymer and the recovery of the marble signals (image d).
Figure 6 (A) SEM picture of a marble surface treated with paraloid B72 before cleaning and (B) after cleaning. (C) Mapping of calcium (green spot) through EDS (Cameo software) before cleaning; (D) 30 min after nanomagnetic gel application; and (E) after the complete removal of the polymer.
Figure 7 Application of the nanomagnetic gel to a fresco painting realized in the laboratory for testing the gel efficacy for paraloid B72 removal. Left panel: white glaze shows the area treated with Paraloid, and the circle shows the area that will be used for gel treatment (grazing angle light). Center panel: gel treatment (normal light). Right panel: area cleaned by the removal of paraloid B72 (grazing angle light).

The nanomagnetic gel loaded with the microemulsion was used to clean the surface of a marble sample (5 × 5 × 2 cm3) that was selected because of the white color that emphasizes the possible presence of black magnetic nanoparticle residuals on the surface. Paraloid B72 is the most widely used resin for the consolidation and protection of paintings and stones. Unfortunately, Paraloid B72 tends to yellow and lose its chemical and mechanical properties after natural aging, producing consistent damage to artifacts.21-23 The marble sample selected for the experiment was treated 8 years ago by brushing with a p-xylene solution of paraloid B72 and was stored at ambient conditions. The surface before the nanomagnetic gel application appeared glassy and yellowed. The removal of the resin was performed by direct application of the microemulsion-loaded magnetic gel onto the area to be cleaned. After the desired application time (variable from 10 min to 2 h), the gel was removed with the aid of a permanent magnet (see the sequence in Figure 4). It is worth noting that no contact between the magnet and the artifact is necessary to remove the gel, making the cleaning process particularly appropriate for the cleaning of precious artifacts. The efficiency of the removal process was investigated by means of microreflectance FTIR. A comparison of spectra (Figure 5) collected before and after the treatment showed that signals from the acrylate at 1751 cm-1 completely disappeared after the cleaning. In addition to FTIR, the SEM analysis and, in particular, the mapping of X-ray emission collected by an EDS system (energy-dispersive X-ray spectrometry) provided clear evidence that no residuals of the magnetic nanoparticles were present on the cleaned surface and that the polymer was completely removed (Figure 6). The nanomagnetic sponge was also used to clean the surface of a damaged fresco, showing that this formulation can be applied to this kind of artwork as well (Figure 7).

In conclusion, a new magnetic responsive sponge has been synthesized. The sponge can be loaded with common solvents or with more sophisticated dispersed systems, such as microemulsions or micellar systems, and can be easily manipulated, cut with a knife or scissors to the desired shape, and magnetically or mechanically removed from the region of application. The loaded phase (microemulsion or micelle) retains its original structure and properties. The overall system is particularly efficient in the uptake and release of the material contained in the loaded phase. We highlighted here the application to cultural heritage conservation, but it can be used in a large number of practical applications. The nanomagnetic gel represents the most advanced and versatile system for cleaning and will have a dramatic impact on the conventional methods used in the conservation field and in several other fields where fine tuning of the release or uptake of confined material is required.

Acknowledgment

Thanks are due to David Chelazzi for the FTIR analysis, Emiliano Fratini for SAXS experiments, and Michele Baglioni and Giacomo Pizzorusso for assistance with the application tests. Financial support from MIUR (PRIN-2006) and CSGI is acknowledged.

Supporting Information Available

Preparation of the magnetic nanosponge. This material is available free of charge via the Internet at http://pubs.acs.org.

* Corresponding author. E-mail: baglioni@csgi.unifi.it.

Present address: Max Planck Institute for Polymer Research, Ackermannweg 10, D-55128 Mainz, Germany.

1. Clark, A. H.; Ross-Murphy, S. B. Adv. Polym. Sci. 1987, 83, 57. [ChemPort]

2. Guenet, J. M. Thermoreversible Gelation of Polymers and Biopolymers; Academic Press: London, 1993.

3. Te Nijenhuis, K. Thermoreversible Networks: Viscoelastic Properties and Structure of Gels; Springer-Verlag: Berlin, 1997.

4. Michalski, S. Cleaning, Retouching and Coatings: Technology and Practice for Easel Paintings and Polychrome Sculpture; International Institute for Conservation of Historic and Artistic Works: Brussels, 1990.

5. Wolbers, R. C. Restoration '92: Conservation, Training, Materials and Techniques, Latest Developments; International Institute for Conservation of Historic and Artistic Works: London, 1992.

6. Wolbers, R. C. Cleaning Painted Surfaces: Aqueous Methods; Archetype Publications: London, 2000.

7. Dorge, V. Conservation; Getty Conservation Institute Newsletter; 2000; Vol. 15.3, pp 16-19.

8. Terech, P.; Weiss, R. G. Chem. Rev. 1997, 97, 3133-3159.[Full text - ACS] [ChemPort] [Medline]

9. Carretti, E.; Dei, L.; Baglioni, P.; Weiss, R. G. J. Am. Chem. Soc. 2003, 125, 5121-5129.[Full text - ACS] [ChemPort] [Medline]

10. Carretti, E.; Dei, L.; Baglioni, P. Langmuir 2003, 19, 7867-7872.[Full text - ACS] [ChemPort]

11. Carretti, E.; Salvadori, B.; Baglioni, P.; Dei, L. Stud. Conserv. 2005, 50, 1-8.

12. Carretti, E.; Giorgi, R.; Berti, D.; Baglioni, P. Langmuir 2007, 23, 6396-6403.[Full text - ACS] [ChemPort] [Medline]

13. Rance, D.; Friberg, S. J. Colloid Interface Sci. 1977, 60, 207

14. Massart, R. C. R. Seances Acad. Sci., Ser. C 1980, 291, 1-3. [ChemPort]

15. Massart, R. IEEE Trans. Magn. 1981, 17, 1247-1248.

16. Bonini, M.; Wiedenmann, A.; Baglioni, P. J. Phys. Chem. B 2004, 108, 14901-14906.[Full text - ACS] [ChemPort]

17. Bonini, M.; Wiedenmann, A.; Baglioni, P. Physica A 2004, 339, 86-91. [ChemPort]

18. Sartoratto, P. P. C.; Neto, A. V. S.; Lima, E. C. D.; Rodrigues de Sa, A. L. C.; Morais, P. C. J. Appl. Phys. 2005, 97, 10Q917.

19. Patras, G.; Qiao, G. G.; Solomon, D. H. Macromolecules 2001, 34, 6396-6401.[Full text - ACS] [ChemPort]

20. Benguigui, L.; Boué, F. Eur. Phys. J. B 1999, 11, 439-444. [ChemPort] [CrossRef]

21. Price, C. A. Stone Conservation: An Overview of Current Research; Getty Conservation Institute: Los Angeles, 1996.

22. Horie, C. Materials for Conservation: Organic Consolidants, Adhesives and Coatings; Butterworth-Heinemann: London, 1987.

23. Feller, R. L. Accelerated Aging: Photochemical and Thermal Aspects; The Getty Conservation Institute: Los Angeles, 1994.