The Microscope in Art Conservation and Authentication Studies

Part 1: Questions of Authenticity

Some of a microscopist’s most interesting projects involve investigations into the authentication of works of artistic, historical or archeological interest. These are objects that, if proven to be created by or associated with a particular artist or historical person, can have enormous financial value, and the possibility of acquiring great wealth is always strong motivation to find (in the attic or a yard sale) that long lost Rembrandt; or, for the less scrupulous, to create it themselves.

By far the most common objects studied are works of art, and especially paintings. Experience in this area has led to the development of a number of protocols that help to focus our efforts quite efficiently; nevertheless, each work of art is received with a completely open mind and with the attention to detail required in these cases. This article, and a few to follow, will discuss the authentication process as it applies to paintings in particular, but the general principles apply to drawings, pastels, sculpture, terra cotta, flags and other textiles, historical documents, and the like.

Authenticity: Three Factors

How do we decide that something is authentic? The process takes into account three important factors:

1. Style. Is the object possible in terms of style, technique and subject matter? Does the object look like it was done by Van Gogh? This is the responsibility of the art historian or appraiser, and we generally recommend that our clients first have it looked at by a professional acquainted with the artist in question.
2. Provenance. Does the object have a reasonable and well-documented history? An excellent provenance for a painting would be a bill of sale from a gallery or a serious collector, a photo of the work in an exhibition catalog, or, best of all, publication in the catalog raisonné, showing that the work has been completely accepted by the art historians as authentic. The collector contemplating a purchase should, however, be aware that determined forgers may – and often do – invent documentation supporting their objects’ history. Provenance is of greatest concern to the person making the final decisions as to a piece’s authenticity or market value. The auction houses pay close attention to provenance as a major criterion in determining the salability of an object.
3. Scientific Analysis. No matter how well painted, no matter how detailed the documented provenance, if the object consists of materials not available in the artist’s lifetime, it cannot be considered authentic. Authenticity is also ruled out if evidence suggests deliberate alteration of the piece.

Authenticity: Who Decides?

Who decides that an object is authentic? In the case of works of art, there is frequently a single person, an individual that is considered the authority on the works of a particular artist; for some artists, a committee performs that role. Historically, the criteria they employ in determining authenticity are largely style and provenance; too often, technical studies have not been performed. Sometimes, to the deep embarrassment of the experts, subsequent analysis proves works that have received judgments of authenticity by the experts are shown to be recent forgeries. Nowadays, the experts generally require a scientific analysis of a work before pronouncing it to be authentic.

Variations in Authenticity and its Absence

Ideally, the role of the scientific investigator should be the identification of the materials and methods used to create the work, without regard to any judgement of ‘authenticity’. From the client’s point of view, however, ‘authenticity’ is precisely the question they need answered. We have come to learn that there are many ways in which a work might be considered “authentic” or not. Here is a breakdown of seven of the possibilities.

Authentic. All of the materials found in the work and all of the techniques displayed point to a product created by the artist to which the work is attributed. For the analyst, these are truly exciting finds, for all the evidence, as it accumulates, points more and more directly to the attributed artist. However, it is not the role of the scientist to declare the work to be authentic – only to determine whether the materials in the work are consistent with authenticity. When an item is truly authentic, we find that the more deeply we probe, the more evidence we find supporting its validity. Here’s an example:

Case 1: The Old Master

The last pieces of the puzzle fell neatly into place. Velázquez was known to use every pigment found in the piece, an elegant and dramatic portrait of King Phillip IV of Spain. The palette included azurite and lead tin yellow, pigments unused after 1750 and 1800 respectively, so the painting was clearly of sufficient age. These specific pigments were not all that distinctive, though, as the 17th century painter had only a limited number of blue pigments available to him. The ground layer, however, included a layer of a light red ochre with a number of accessory minerals including quartz, clays and mica—and this distinctive ground was known to have been used during his early Madrid period! Another of his blue pigments was smalt, a cobalt-containing ground glass. The cobalt ore, which was the source for his smalt, included trace amounts of arsenic, nickel and iron—and we found this precise mix in the portrait of Phillip IV (Figure 1).

Mistaken identity. This is a large class with many variations. The provenance is generally that of the ‘found object’: an estate, yard or garage sale, flea market or antique shop purchase; a family heirloom.

Reproductions. If inspection with a magnifying glass reveals that the image consists of halftone dots, we can be sure the work is not a painting; usually, this is quite obvious, but not always:

I once received the ‘paint’ samples as cotton swab rubbings. At first, I found the paints odd in that they were mostly soluble dyes – until I realized that there were pure (unmixed) colors present: cyan, magenta, yellow and black. Slap the forehead with the palm of the hand: CMYK, the printer’s ink colors! Problem solved (Figure 2A and 2B).

Stylistic similarities. The work is a real painting, made somewhat in the style of a famous painter and may or may not have the real artist’s signature present. These generally are easily solved by the presence of materials inconsistent with the era of the attributed artist.

Copies. During the Old Master period, copies of particularly effective works were sometimes painted, often within the master’s workshop, as a way to fill the great demand for art that could be used for teaching. The missionary Church was an especially eager market for religious subject matter. Furthermore, many artists copy the creations of other artists as an integral part of their training. Their works may enter the marketplace as by the copied artist, but, naturally, lack a reasonable provenance. For the analyst, determining the work’s true nature may represent significant challenges.

I missed a great buy in a Salvation Army store a while back, a ‘self portrait by Van Gogh’, done rather poorly in watercolor. Nevertheless, it typifies the efforts that student artists go through in learning their craft.

Misrepresentations and altered originals. A perfectly authentic object may become more valuable by association with a famous person. A painting by an unknown artist may become a lost work of Rembrandt or Renoir by scraping away an original signature and replacing it with the more famous one.

My family’s library includes a Bryant and Stratten Commercial Arithmetic from 1860, inherited from my great grandfather, who was a businessman. My sister – she was about nine at the time – decided to “increase its value” by writing in Abraham Lincoln’s signature – in her own handwriting, with ballpoint pen. Ironically, this has actually added to its value, but as a family memento, not monetarily. And it would be a pretty easy analysis – the ball point pen was not commercially available until about 1943.

Forgeries. Finally, the forger, generally a fine artist in their own right, constructs a work very much in the style of the famous artist. It may or may not include a signature – the style is the signature, especially when the style is highly distinctive. Clearly, misrepresentations or altered originals and outright forgeries reflect intentional deceit, and, when sufficient evidence is available, the perpetrators can be prosecuted as criminals.

Case 2:  Who Forged the Cubist?

At 400 times magnification, the lovely, light blue green paint looked suspicious. I crossed the polarizers in my microscope, and I knew immediately that the painting had a serious problem. The sea of a mixture of blue green and rust colored particles could only be one pigment, manganese blue, a pigment completely unavailable to artists until 1935, and more likely not until the 1950s (Figures 3A – 3B). The painting was dated, under the signature, 1915. However, the famous artist did not die until the mid 1960s. The subject matter and style were only used by the artist for a few years and would not have been appropriate during the latter part of his life; however, the art historian studying the work felt that it represented the authentic hand of the artist. Did the master forge a work in a manner reminiscent of an early period in his career? After all, in his later years, those works brought higher prices than his more recent creations. Or was a different hand at work?

Part 2: Protocols for Painting Analysis

Our analyses are directed to determining the earliest possible date of creation of the painting, based on our ability to identify materials constituting the work and our knowledge about their dates of discovery or invention.

The opaque white pigments provide a clear example of the workings of the protocol. If a painting contains lead white, a pigment in use from Roman times and even before, it could be Roman, Medieval, Renaissance, Impressionist, Modern or Contemporary–it is still manufactured as an artist’s pigment, although its use has declined considerably in recent years due to the pigment’s toxicity and the availability of suitable non-toxic substitutes. If the paint contains zinc white, it was created some time after 1825, when the pigment was first introduced, and if the paint includes titanium white, we know that the painting was made sometime after about 1919 or so. In actuality, the titanium whites have a complex history and provide us with about a half dozen dates ranging from 1916 to 1943, depending on purity and the crystal form of the pigment.

The Analytical Process

The analysis of a painting progresses in a number of discreet stages:

Examination. All studies begin with a careful examination of the painting using a variety of light sources, from white light to ultraviolet, both with and without supplementary magnification. A general assessment is made as to the condition, and characteristics such as size, substrate, conservation history and so forth. If a signature is present, careful attention is given to the surrounding region, looking for signs of alteration or manipulation. Conservation history is particularly important, as a characterization of a 19th century restoration on a 16th century painting could be sadly misleading.

Sampling. Small samples of every important color (ignoring obvious mixes) are taken using an extremely fine-pointed tungsten needle; the cosmetic result after such sampling is excellent (Figure 4).

Polarized light microscopy (PLM). In spite of a world-class array of analytical instrumentation, PLM remains our first and most important analytical technique. The analyst can see the individual particles through the microscope, even in complex mixtures, whereas bulk analysis can completely miss them. Even a little titanium white pigment, scattered throughout a work of art, points to post 1920s creation. The pigment tends to agglomerate, which is handy for analysis confirmation.

Elemental analysis. We routinely run all samples through the scanning electron microscope (SEM) for energy dispersive x-ray spectrometry (EDS). This technique provides an elemental profile of each paint sample and is the perfect complement to PLM. It can confirm or deny the presence of certain pigments, and the printed spectra make great, easy-to-interpret figures for the report (Figure 5).

Infrared spectroscopy (IR) is used routinely for analysis of the medium. Anna Teetsov, our prize-winning specimen-preparation microscopist, has developed a number of methods to extract the different components from a paint mixture for individualized analysis. IR provides information based on the molecular vibrations of the molecule, and is the most important tool in identifying organic materials.

Raman spectroscopy (Raman), another technique based on molecular vibrations, complements IR very nicely. Materials unidentifiable with IR can often be characterized with Raman, and vice versa. It also complements both PLM and EDS as well: a Raman spectrum can be obtained from a single particle on a PLM microscope slide, in mounting medium and under a cover slip. Also, it can further elucidate the nature of, say, titanium white by showing the crystal form of the material, anatase or rutile, each of which has different dates of availability in the 20th century (Figures 6A and 6B).