Nano-motion imaging reveals cleaning solvent impact on The Night Watch

By Jesse Buijs

News in Conservation, April-May 2025, Issue 107, p. 16-21

The conservation and restoration of art has adopted many new technologies in recent years. In this article, I will introduce you to one of these new technologies: nano-motion imaging. Upon application of cleaning solvents to oil paint, nanoscopic motion, like diffusion and swelling, can cause chemical and physical degradation. With the ability to detect this nano-motion, it becomes possible to minimize the risk of accelerated aging of paintings during conservation efforts. Two years ago, I began the start-up NanoMoi that has developed a small and user-friendly instrument for this purpose. Since then, this instrument has already been used to measure the varnish removal from the iconic Night Watch in the Rijksmuseum in Amsterdam.

My nano-motion adventure began in 2018 by joining the laboratory of physical chemistry and soft matter at Wageningen University (NL) for a PhD. There was no connection to art within the research project. I worked on further developing a video imaging technique that could visualize motion inside non-transparent materials. The lab previously used it to study the drying of water-based paints to improve their final film quality. The measured nano-motion showed significant movement where the paint was wet and a lack of motion where the paint had dried. These high-resolution videos showed how heterogeneous the drying process actually is. They could not have been obtained with other microscopic techniques, as the color and contrast of the paint does not change during drying. The paper in which these results were published was picked up by researchers from the Rijksmuseum. They contacted us to ask if it would be possible to use the method to study the aging and restoration of paint that had already fully dried. We were intrigued by this potential application idea and started some trials as ”Friday afternoon experiments”.

From the beginning, the combination with art conservation was very promising. Our first test was just a simple drop of solvent on a dry sample paint layer. It showed lots of nano-motion where the solvent went into the paint. This nano-motion is two-sided: the solvent swells the oil paint like a sponge, which causes crack formation. And it also plasticizes the oil matrix which increases the freedom of movement due to thermal motion, which accelerates chemical reactions and degradation processes. In the nano-motion video we could see how far the solvent would spread from the original application area and how long it stayed in the paint.

From this test, we saw that the practical application of this technique to actual paintings looked promising. The nano-motion technique is based on a medical imaging technique that uses laser speckle interferometry to visualize the motion of blood flow. The laser can be set to low-power to ensure it is non-invasive; a requirement used on patients as well as on paintings. The scattered laser light penetrates the material—up to 100 microns for paint—so the measured nano-motion doesn’t just provide information on the surface-level solvent, but also provides data about solvents present deeper inside the paint. I was also already working on an algorithm to perform all data analysis automatically in real time so that the results would be immediately available during the restoration of objects.

With these insights, we moved on to systematic tests to lay the scientific foundation and prove our theories. The Rijksmuseum supplied artificially aged paint samples, cleaning supplies and a conservator to perform the cleaning tests while we performed the nano-motion measurements. In these measurements, all nano-motion is induced by solvents. Therefore, our instrument essentially became a solvent monitoring tool.

We decided to describe our results according to ‘solvent retention time’: the time the cleaning solvent remains inside the paint layer. Our tests showed that a drop of ethanol left on bare oil paint for one minute has a retention time of roughly 15 minutes. Running the same experiment with water, the resulting retention time was half a minute, and for acetone, the retention time exceeded more than an hour. Solvent retention depends on the solvent-paint interaction, mostly dominated by the diffusion rate and contact time. The retention time also increases for older paints as they have a higher porosity. When the paint was aged to the point at which craquelure had formed, we observed that these cracks enabled solvent spreading over distances of more than a centimeter—a huge difference when compared to solvent on a smooth surface, which typically stays within a millimeter of the applied area.

The real potential uses of this new technique were even more evident when we added varnish to the experiments. We were able to identify a solvent treatment that fully removed the varnish layer with almost no ethanol penetrating the paint layer below. Since the varnish layer is transparent, we cannot detect nano-motion within it; we exclusively detect solvent that has reached the paint layer. If the cleaning is too short, all solvent is blocked by the varnish layer and the varnish removal is incomplete. If the cleaning is too long, the varnish will be completely removed while also seriously impacting the paint. With nano-motion imaging we found the sweet spot where the varnish was removed without significant solvent impact to the paint layer.

This success gave us more than enough assurance to move on to the next phase of experimentation: adapting the tool to be used on real artworks. We developed a workflow so that our measurements could be incorporated into the standard workflow of conservators during initial cleaning tests on oil paintings. These tests are often done on small areas of the pain-ting to check which methods work best. The nano-motion measurement starts after each solvent test to determine the solvent retention time, gathering extra information from such tests.

We designed a new instrument, making it portable by shrinking it to the size of a shoebox. On the software side, we made huge improvements by automating all data analysis and designing a user-friendly interface. To enable other conservators to access this technique, I founded the start-up company NanoMoi, making the new instrument commercially available. With the new NanoMoi instrument, we measured ongoing restoration treatments on artworks scheduled for display in the Rijksmuseum galleries. The results were much more complex than our controlled tests during the scientific research, but they represented the situation well; the paintings were very complex in comparison to the model paintings we had previously tested. The collection paintings contain a mix of different oil paints with unique properties, and the paint films and varnish layers are not as smooth. There can even be differences within a single color area depending on its restoration history. Therefore, it is not always possible to find one perfect cleaning method for a whole painting. But with this new data, a well supported decision can be made on how to best perform a varnish removal. By taking into account the solvent retention time variation within a painting, it is more possible to create a cleaning strategy with the maximum varnish removal and minimum solvent impact. This new approach sometimes led to counter-intuitive results; using a more aggressive solvent could result in a lower total solvent impact if the required application time was significantly lower. Understanding solvent retention helps the conservator in making choices that can be scientifically supported.

We are proud to have completed the journey from fundamental research to full application within seven years. This was possible because of our collaboration that combined interdisciplinary knowledge and practical know-how with relevant case studies. The timing of this project couldn’t have been better, as the ongoing restoration of The Night Watch was heading into the varnish removal phase. The conservation and restoration team was eager to have the solvent impact data to support their decision-making process for this historic painting, and they had seen how useful the results could be on the previously tested paintings. The two weeks of measurements on The Night Watch were unforgettable, being so close to this masterpiece and performing the measurements while the painting was still on display for all museum visitors—it was a huge milestone for my young start-up company.

While working together with conservators, many other applications for nano-motion imaging came up. It is wonderful to work in such a multi-disciplinary field. For example, the instrument can be used to detect small gel residues, gels being another common method to deliver solvent during varnish removal. Our instrument could check if all gel has been completely removed from the paint surface. We have also performed tests on other artistic materials including crack activity detection in metal and glass to determine whether or not preventive restoration is required. We can also detect salt crystallization, which eventually leads to erosion on stone structures and sculptures. Outside of art conservation, many other potential applications await in materials, food, and coatings science. With this outlook, I hope that a future application of the technique might come close to matching NanoMoi’s milestone of measuring Rembrandt’s famous Night Watch.

Author bio

Jesse Buijs is the founder of start-up NanoMoi. He obtained his PhD in materials science and imaging at Wageningen University. His main drive is the wish to apply advanced technology in practice and make it available for a broader audience. He enjoys being the bridge between fundamental science and practical engineering, supporting conservators in their vital work.

Follow IIC for more News in Conservation content

You can sign up for free (scroll to the bottom) to receive each issue of NiC

Become an IIC member for more great content, membership programmes and benefits