Make Your Own Oil Paint!

Hi Everyone,

The end goal of my summer research was to prepare a comprehensive library of SERS spectra for the many possible plant sources of organic yellow dyes and pigments.  Once I tested the water soluble, cooperative dyes and extracted the insoluble pigments, the next step was to test oil paints.  As the pigments and dyes would be found on a painting mixed with an oil binder, the best simulation of an art sample would be to take SERS spectra of oil paints made from the yellows I have been testing.

In order to know exactly what is in the paint samples, I mixed my own oil paints!  As a long time artist, this is something I have always wanted to do and I was surprised by how easy it was.  The more heavily processed pigments were ground finely enough to just mix with an oil binder.  After depositing a small pile of the pigment on a stone workspace, I slowly added a few drops of linseed oil.  From there, additional amounts of pigment and oil were mixed together with a glass rod until the appropriate consistency was reached.  These paints were spread onto glass slides and allowed to dry.  Surprisingly, few of these “yellows” truly appeared yellow in their raw state.  Many were greenish and some were even a dark brown hue.  Therefore, these dyes were most likely mixed with lead white or another pigment to create an ideal yellow color in actual paintings.  Also, many of these yellow colorants were used primarily in the pursuit of the perfect green color when mixed with an inorganic blue.

The dyes, or basic plant components I worked with had to be further ground up in order to accurately simulate paint.  Using a mortar and pestle, I ground up the more coarsely ground dyes and then proceeded to mix them with linseed oil.  These samples were stored for a few weeks until the paint was mostly dry.  Then, I sampled them the same way I would a real painting and performed the corresponding extraction or basic SERS to acquire matching spectra.  Once I could confirm that the chromophores, dye, and paint of one material all yielded the same peaks in the SERS spectrum, I could hypothetically identify that colorant in a real painting.