Under the Surface of a Painting

Hello again!

Due to our collaboration with Shelley Svoboda at the Dewitt Wallace Collections and Conservation building, my lab has access to historical oil paintings to sample and study.  Currently, the piece of interest in my project is the portrait of Mrs. Nelson by Robert Feke, an American artist.  Despite the fact that the subject of the portrait is wearing a blue dress, yellows may be hiding among the paint layers.

Knowing that the painting should be naturalistic, some questions were raised regarding the bluish look of the leaves and stem of the rose in her hand.  After examining a cross section of the paint layer, sampled by Shelley Svoboda, we hypothesized that some organic yellows may have been used in this section of the painting.  The artist probably mixed the green paint of the leaves and stem by combining an inorganic blue pigment with an organic yellow pigment.  Therefore, the blue has not faded over time and the yellow has.  This presents an interesting opportunity for me to attempt to identify an organic yellow in an oil painting.

The first step in identifying this colorant was to sample the painting.  This was my first experience sampling anything, let alone something so valuable.  Under the microscope, I gently shaved off a layer of paint with a razor blade, depositing the near invisible paint samples onto a glass microscope slide.  A good way to tentatively identify an artistic material by process of elimination is called polarized light microscopy.  In order to prepare a sample, the paint is mounted with a resin-like solution onto a glass slide.  Once the mounting solution has dried, the sample can be viewed under a microscope to examine the size, shape, and visual behavior of the particles.

Once the size, opacity, and shape of the particle have been observed, the refractive index of the sample can be determined by focusing in and out on the particle.  Because the mounting solution has a known refractive index, the refractive index of the particle can be determined based on whether light is pulled into or out of the sample as it is moved in and out of focus on the stage.  As the refractive index of most dyes is known, a simple analysis of whether the refractive index is higher or lower than the mount is helpful.

An indicative property of a material can be determined based on its behavior under plane polarized light.  If a material has more than one refractive index, it transmits light along more than one vibrational direction.  When you shine polarized light, or light only traveling in one vibrational direction, at the particle, you will either view darkness or a bright, sometimes multicolored, sample.  Materials that have more than one refractive index are called birefringent or anisotropic materials, and will exhibit brightness under polarized light.

The shard-like yellow particles have been tentatively identified as orpiment, a commonly used inorganic yellow.  Orpiment is a sulfide of arsenic and, therefore, a rather toxic substance.  It was often coarsely ground to yield a brilliant yellow composed of waxy, glasslike particles under the microscope.

The inorganic blue used in the painting was probably smalt or ultramarine, considering the appearance of the blue paint samples under the microscope.  Smalt was a much less valuable pigment than ultramarine, as it was made by complexing cobalt with ground up glass to create a similar hue to the expensive ultramarine.  This only accounts for a tentative examination of the inorganics used in the painting.  As my project focuses on the organic components of paint, there is much work to be done.  However, due to the difficulty isolating and identifying faded organic colorants, this proves how useful SERS is to the world of art conservation.


  1. Ben Kirby says:

    I am glad to see that you have found a way to integrate your interest in art history with science. I’ve tried to do the same and hope you’ve had the same positive experience!