Empirical examination and scientific analysis are fundamental to conservation research and treatment; conservators frequently collaborate with scientists in order to clarify specific questions: to identify materials, elucidate degradation mechanisms, or test the efficacy of conservation methods.

In our last post we examined the archival record of One: Number 31, 1950 to better understand its history and current condition. The documentary evidence of letters, reports, and a key photo, makes it clear that One was restored some time after 1962, and it definitively identifies the incongruous areas we observed to be a restorer’s overpaint rather than Pollock’s work. This clarification presents us with a choice regarding our current conservation treatment:

Do we remove the overpaint or leave it in place?

To arrive at an informed decision on this matter requires characterizing the paint surface as fully as possible. This part of our research into the painting’s condition makes use of a variety of scientific examination and analytical techniques.

We first examined the painting using ultraviolet (UV) illumination, a type of light that causes many materials to fluoresce in characteristic ways. If you’ve ever been to a dance club or funhouse where clothing glows under “black lights,” that’s UV fluorescence. Based on the hue and intensity of the fluorescence, we can differentiate  materials that appear very similar in visible light.

We observe that the areas we had visually isolated as overpaint do indeed fluoresce differently from Pollock’s original. UV illumination, therefore, provides one way of mapping the areas of overpaint.

We also sought to better understand the state of the original paint beneath the overpaint. The photograph from 1962 shows Pollock’s paint, despite some cracking, largely intact. If the restorer’s paint covers Pollock’s original as generously as it does in the area seen above, then a powerful argument exists for removing the overpaint to expose the long-buried details of Pollock’s composition.

Just as X-radiography is useful in medicine for “seeing” bone beneath tissue layers, conservators use X-rays to reveal the internal structure and layers of an art object. We chose several locations of overpaint on One and took a series of X-radiographs to image the paint layer beneath the restorer’s work.

The X-radiographs reveal a consistent pattern. In each case, the underlying original paint appears to have suffered cracking. The paint added during the 1960s restoration extends well beyond, covering the cracks as well as original, undamaged paint.

We know, then, that Pollock’s paint hasn’t been lost, that it is intact beneath the overpaint. The next step is to identify the paint across the surface, both Pollock’s original paint and the restoration paints. The chemical difference between the two will determine whether or not it’s possible to remove the overpaint without causing any damage to Pollock’s paint beneath.

Paint can contain a long list of ingredients, but, in its simplest form, it is made up of two things: pigment (i.e. the colorant) and binder (the resin, or similar material, into which the pigment is mixed). Thus it may be possible to differentiate these components between Pollock’s and the restorer’s paints. Using a form of elemental analysis called X-ray fluorescence (XRF), conservation scientist Chris McGlinchey was able to identify many pigment differences between the original paint and overpaint. Though these two blues look almost identical in hue, for example, only the overpaint contains the element cobalt, demonstrating that the pigment cobalt blue is present in the restoration paint.

Whenever possible, we make use of the many non-invasive techniques, such as UV, X-radiography, and XRF, at our disposal—non-invasive meaning that no material is removed from the painting or changed in any way during the procedure. To characterize the paint binders, however, requires a small sample be taken from the painting. These samples are so tiny that one needs a microscope to study them, but, even so, conservators and scientists do not make the decision to remove original material from a work of art lightly. The goal of this binder analysis is twofold. First it will add to our understanding of the types of paints Pollock, in fact, used to create his paintings, and, two, it will clarify whether or not the overpaint can be safely separated from the original.

McGlinchey analyzed representative samples of both binders using Fourier Transform Infrared Spectroscopy (FTIR), a technique commonly used by conservation scientists to identify organic compounds. His work confirms the popular understanding that Pollock painted One using industrially manufactured house paints, specifically modified oil paints called alkyds. Moreover, the restorer’s paint binder is completely different. It’s a synthetic resin, poly(vinyl) acetate, that was commonly used by restorers in the 1960s. Because the two paints are chemically different, it is likely that it will be possible to separate the two with an appropriate solvent.

With all of this information, both archival and analytical, our next steps will involve testing solvents that will dissolve the overpaint but not affect Pollock’s paint.

This is done by using a cotton-tipped swab to introduce solvent to the area of interest, gently rolling the swab across the surface until the overpaint starts to be absorbed by the cotton.

Removing overpaint from a painting is, obviously, a significant and potentially risky decision. It is for these reasons that we researched the history and composition of the overpaint so extensively, to make sure that when it is removed we will be left with a painting that more fully reflects Pollock’s original painting and intentions.