For over 13 years, Van Gogh’s drawing inks have been studied at our institute, the State Heritage Laboratory/Cultural Heritage Agency, in close collaboration with the Van Gogh and the Kröller-Müller Museum. Together, these museums possess the world’s most extensive collection of Van Gogh drawings. Due to light exposure, the inks on some of these drawings have faded considerably. Also, inks on some of his letters, present at the Van Gogh Museum and the Pierpont Morgan Library in New York, have been studied. As these have less often or never been exhibited, it was expected that the inks have retained their original appearance. Long term objective is a better understanding of the making and the change in appearance over time of the drawings. During these years, our capability to characterise inks and interpret the results has developed significantly.
In-situ analysis of the metals present in the brown inks on 73 drawings and 20 letters by X-Ray Fluorescence Spectrometry, amended by in-situ Fibre Optics Spectroscopic analysis showed that, during the Dutch period (1880-1885), he mainly used iron-containing inks, presumably iron gall ink, for writing and drawing. During the French period (1885–1890), he mainly used chromium-containing inks, presumably chrome-logwood ink. The violet ink used in several drawings he made during a short period in France (1888) is containing aniline dyes, mainly methyl violet, as initially was shown by Surface Enhanced Raman Spectrometry (SERS).
UHPLC with PDA detection
The organic compounds in these inks, as well as in the chrome-logwood inks were identified by Ultra High Performance Liquid Chromatography (UHPLC) with PDA detection. For this, samples were taken with a newly developed micro-sampling method, using magnesium-oxide rods. This sampling method also allowed other techniques to be used, e.g. SEM-EDX to determine the elemental composition.
Van Gogh used commercially available writing inks, therefore, to investigate their composition and know what was available, during the last 5 years, over 150 bottles of writing inks from the period 1890 – 1950 have been collected as a reference collection. Several of the violet inks of this collection have been analysed by UHPLC-PDA and the results were compared to those obtained from the violet inks on the drawings and letters, as well to those obtained from our dye reference library, constructed by chromatographic analyses of our collection of reference dyes. These comparisons were conducted by automated data visualisation. Most remarkable, besides methyl violet, benzyl violet was present in the violet inks on the drawings.
The ink colours were calculated from the relative ratios of the peak areas of their coloured constituents, as obtained from the UHPLC-PDA data. Using this method and assuming a distinct relationship between the non-coloured products of the fading of methyl violet and their coloured precursors, as derived from its fading mechanism, it is expected that the original ink colours can be calculated and the drawing be virtually reconstructed in its original colours. For this, also the changed paper colour has to be taken into account.
UHPLC-analysis also showed the presence of methyl violet in chrome-logwood ink on a drawing. According to an early 20th C. description of the extraction process, the ink manufacturers added this to compensate for the decrease in hematein content of the logwood extract during steam extraction. Light-induced fading of methyl violet has caused these originally nearly black inks to turn brown.
This is an abstract of the presentation 'The quest for Van Gogh's inks' at 'Dyes in History & Archaeology' at Lisbon, 25-26 Oct 2018 (this research is part of the ReViGo Project), by Birgit Reissland, Frank Ligterink, Art Proãno Gaibor and Han Neevel, researchers at the Cultural Heritage Agency in the Netherlands.