Spongy hydrogels clean textured paintings

Pastes, poultices, thickeners, and absorbent materials have been used for thousands of years to apply fluids to surfaces. You’d probably clean up a mess using a wet rag or sponge rather that dousing it with soapy water. Confining a liquid in a solid or semisolid matrix prevents it from immediately inundating the surface and flowing away.

The use of gels for cleaning art and artifacts, which dates back to the 1980s, is based on a similar principle. A sheet of gel, imbued with water, surfactant solution, or other solvent, is gently placed onto the surface to be cleaned and left there for a minute or two. During that time, tiny amounts of fluid seep out of the gel and loosen the dirt particles. The gel is then peeled away, and with it, one hopes, comes the dirt.

Conventional gels—many of which were adapted from food additives such as agar and gellan gum—don’t solve all cleaning problems, and they sometimes introduce their own. A gel can lose its structural integrity and leave bits of itself behind. It may deposit a chemical residue that alters the finish and appearance of the paint. Or it might just be ineffective at removing dirt.

But gels aren’t all created equal; their diverse chemical, mechanical, and structural properties affect their cleaning performance. Baglioni and colleagues’ idea is to use the tools of soft-condensed-matter physics to design new materials tailored to the needs of art conservation. They’ve already concocted a polymer hydrogel that can remove the residue of adhesive tape from centuries-old sheets of paper.² (See Physics Today, July 2018, page 21.) That gel works well on flat surfaces, but as figure 1b shows, it’s far too stiff to conform to the irregularities of a contemporary textured painting.

In search of a more mechanically compliant gel, the Florence researchers turned to polyvinyl alcohol (PVA), a material long known in the biomedical field for being soft yet sturdy, resilient, and chemically benign. Some of its advantageous properties stem from its gelation mechanism. Typically, polymers are transformed from viscous goop into elastic solid through chemical cross-linking, the addition of a curing agent that binds nearby polymer chains together. Any excess curing agent, however, can leach out of the gel and cause problems later.

A PVA hydrogel, in contrast, can be solidified simply by freezing and thawing a solution of PVA in water.³ As ice crystals in the mixture grow and expand, they press the PVA molecules closer together. And by some mechanism that’s not totally clear—maybe through a combination of hydrogen bonding and the formation of tiny PVA crystallites—that compression is enough to lock adjacent PVA chains together permanently.

As figure 1c shows, a PVA hydrogel is soft enough to drape over the peaks and into the troughs of a rough painted surface. It is not, however, effective at cleaning. The problem, Baglioni and colleagues hypothesized, is the gel’s pore structure. Ice crystals in PVA grow long, thin, and straight, so the hydrogel is thus honeycombed with narrow, parallel pores, as shown in figure 2a—hardly ideal for fluid mobility and dirt pickup.

But how does one modify that structure without compromising the gel’s softness? It’s known that a PVA gel’s properties can be tuned by repeating the freeze–thaw cycle more than once. Subsequent cycles widen the pores while retaining their shape. But repeated cycling also makes the PVA walls a bit thicker and thus more rigid—exactly the opposite of what the researchers wanted.

Baglioni’s pivotal idea was to try making a hydrogel out of a mixture of PVA molecules of two different lengths. PVA’s temperature-dependent solubility in water also depends on the length of the polymer chain: As a watery mixture of long- and short-chain PVA is cooled, the short polymers become insoluble before the long ones do. The difference in miscibility would push the short- and long-chain molecules to phase separate, but their sluggish motion would keep them at least partially intertwined. The resulting tangle, Baglioni reasoned, must have some effect on the size and shape of the ice crystals and thus on the gel’s pore structure.

That effect turned out to be surprisingly dramatic. Instead of having long, thin pores like a homogeneous PVA hydrogel, the twin-chain PVA gel, as it’s come to be known, looks more like a sponge. As figure 2b shows, its pores are larger, rounder, more irregular, and more interconnected. When tested on a mock painting, the twin-chain gel proved excellent for cleaning.

To see how that pore structure formed, the Florence researchers tagged their PVA molecules with contrasting fluorescent dyes: green for long-chain molecules and red for short-chain ones. Sure enough, as the solution cooled, the short-chain molecules clumped together into blobs several microns in diameter. The blobs disrupted the growth of the ice crystals and formed the basis for the gel’s pores.

What is it about the sponge-like structure that makes it so good for cleaning? Baglioni and colleagues aren’t sure, but they suspect it has to do with the ease with which the gel both releases fluid and reabsorbs it. As the gel rests on the soiled painting, water gradually evaporates from its upper surface. To compensate, water from the lower surface gets pulled through the interconnected pores into the gel bulk—and the dirt from the painting get pulled with it. Dirt particles are more reliably removed when they’re lodged in the gel’s pores rather than clinging to its surface. “But this is all still a hypothesis,” says Baglioni. “We were working on testing it when the coronavirus hit.”

Beyond demonstrations on mock-ups, the twin-chain PVA gels have already been used to clean real works of art. In collaboration with conservators at the Peggy Guggenheim Collection, a modern-art museum on Venice’s Grand Canal, the Florence researchers used their gels to restore two Jackson Pollock paintings, Two and Eyes in the Heat, to their original 1940s glory. Guggenheim used her inherited fortune to boost the careers of young artists of her day, Pollock among them, and her collection includes several of his works that predate his most famous poured paintings.⁴ Whereas Two was painted with traditional brushstrokes, Eyes in the Heat was created by squeezing thick paints onto the canvas and smearing them with blunt instruments. Both paintings have rough surfaces that have been, until now, hard to clean.

Baglioni’s group has also collaborated⁵ with Ormsby and others at Tate to clean Whaam!, a large two-panel painting by Roy Lichtenstein that the gallery bought in 1966. As with the rest of Lichtenstein’s comic-book-inspired pop art, the surface of Whaam! isn’t especially rough, although its cotton canvas has a texture to it. The main cleaning challenge it poses is Lichtenstein’s use of three kinds of paint, each with its own chemical properties and distinct finish to maintain. A version of Baglioni’s new PVA gel—frozen and thawed a few more times to give it different mechanical properties—proved the best tool for the task. “Each work of art degrades in a different way and needs different conservation,” says Baglioni. “With different tiny modifications, our versatile family of gels can address many needs.”