The physics of restoration

Initially focused on repairing decaying pieces of art, restoration has evolved with advances in technology, and has led to more safe and effective approaches for repairing and preserving work. The process is now a fundamental part of preserving cultural collections and used in museums and collections worldwide. Yet, new research from the Institut Laue-Langevin (ILL) – Europe’s flagship neutron science facility – reveals that some of the new methods proposed for art restoration may cause more damage to priceless works of art than the naked eye suggests.

Art plays a significant part in modern society, and the art of restoration is crucial to safeguarding our cultural heritage. However, due to accidental damage or deterioration over time, much of these works have been lost. Art restorers and conservators work tirelessly to repair and preserve these pieces of work against further damage, and bring it back to its original appearance.

Works of art can be restored multiple times. For example, Michelangelo’s frescos in the Sistine Chapel has been subject to a number of restorations since it was originally painted in the 16^th century. Its most recent restoration in 1994 has both praised for revealing never before seen colours and details, and criticised for damaging the original work from over-restoration. Such merits for art restoration has been debated by critics, artists, and scientists alike.

Traditionally, to prevent damage artists cover their works with a layer of varnish to preserve the layer of paint. This clear, soluble material acts as a barrier between the paint and the outside world, protecting it from airborne dust and soot, oxidation, and attack by free radicles. In turn, however, the varnish layer of a painting darkens as a result of these outside forces, and the artwork no longer looks as bright as it once did. In fact, the protective layer of varnish itself can crack and deteriorate over 100 years, exposing the paint to air and irreversibly damage the artwork.

There is still hope. To restore works of art back into its original form and mitigate further risk of decay, art restorers work to repair any damage to the painting by removing the top-layer of damaged varnish by hand, and replacing it with a fresh, transparent layer. Typically, this critical step is done with a swab soaked with solvent to dissolve the varnish, using minimal shear.

If done correctly, this process can reveal surprising features, such as the alarmingly humanoid face of a lamb in the recently restored Ghent Altarpiece that became a media frenzy. However, more often than not, details which were originally covered by a protective layer, such of the eyes of the Jesse spandrel in the Sistine Chapel, have been lost due to aggressive cleaning which stripped away both the layer of protection and parts of the paint itself.

The removal of varnish is a point of high risk in the process. Deliberate restoration attempts in the last century – not war or natural disasters – have been the source of damage to the largest number of paintings. This is because the application of excess solvent can enter cracks and damage the original paint, create capillary bridges that exert stress on the paint, and reduce the overall cohesion of the paint layer. A major improvement in recent years has been the introduction of hydrogels, which provides better control over the amount of solvent transferred to the varnish. Yet, the final distribution of solvent may still be too concentrated, allowing solvent molecules to diffused all the way through the varnish layer and enter the paint.

To study the physics behind this critical step, researchers from the ILL used neutron reflectometry to determine how solvent infiltrates the varnish, using the same kind of organic solvents that have been used in the restoration of easel paintings. Like telescopes for the night sky, neutrons help researchers see deep into the paint and varnish layer, and differentiate the structure of both mediums.

This research also marks the first time this type of varnish used in almost all art restoration since the 1960s has been studied at an atomic level. Using the FIGARO neutron reflectometer at the ILL, the researchers determined the depth profile of solvent in a ~100nm film of varnish sandwiched between a bulk layer of water and solvent and a single crystal silicon substrate.

Through this, they were able to model exactly how solvents infiltrate the varnish layer of a painting. Instead of a uniform layer of solvent, the researchers found that damage to the painting is less likely if the concentration profile of the varnish is a step function. This profile involves little solvent in the inner part of the varnish layer and a high concentration of solvent in the outer part, separated by a “front” of macromolecules. This reduces the concentration of molecules that are able to penetrate across the varnish and into the paint, which causes damage.

In addition to the model, the researchers found that the common proposal by art restorers to mix solvents with water is counter-productive. This concept has been thought to dilute the concentration of solvent and minimise the amount of dissolution, protecting the pigmented layer of the painting. Due to the presence of water in this concoction, the researchers observed the “de-wetting” of the varnish – where water molecules were attracted to the paint molecules, forming holes penetrating the whole varnish layer, thus exposing the paint layer to the cleaning liquid, creating irreversible damage.

The researchers hope that studying this interaction at the atomic level will prevent the further destruction of priceless artwork, and open the doors to automating art restoration using machines. In the future, a machine could analyse the varnish and perform a well-controlled restoration much better than a human could.

There is no denying that art restoration has been a fundamental part of preserving cultural collections. However, there is a lack of fundamental knowledge about the physical process of dissolution. While many art restorers are aware of the risk, there is currently no other method of performing the process better more safe and effectively. Art restorers and scientists should work more closely together to propose new methods and develop new solvents that can keep the integrity of such priceless works of art.