By Roger Williams
Part I: Analysis
Currently at my bench is a mid-18th-century manuscript chart of a section of the Bahamas: "A Plan of the Harbour of Providence and of the islands adjacent survey'd by Capt. John Gascoigne." There is no production date. The only date mentioned in its cartouche is 1729, though this is probably the year of the original survey. This chart is more likely a somewhat later reproduction.
John Gascoigne was a captain in the Royal Navy, and in 1727 he was commissioned by the Crown to take surveys of several colonies on the HMS Aldborough. He covered South Carolina, Florida, Cuba and the Bahamas. He used the most advanced methods available to make highly detailed hydrographic charts, as made evident by the dense series of numbers in the areas on the chart representing water.
The chart is quite striking, largely due to the wash of green pigment used to colour the areas representing land. What is immediately noticeable, though, is how brittle the chart's paper is. There are cracks and tears along its entire surface, mostly vertically-oriented-probably due to the chart having been stored rolled-up for most of its existence, which is quite likely-and areas of loss (~3%) along its edges. Old repairs of two varieties are present, probably from two different periods of conservation. When I first encountered the chart, I found several loose shards that had apparently been recently broken, and I needed to take great care to move the chart to my bench without causing further damage to the delicate paper.
To someone with a modest grasp of book history, this may be a bit surprising. The paper produced in the 18th century was of mostly high quality, still made by hand out of rag fibres. Paper from this period is often still in good condition-typically better than industrially-produced paper from a century later, which contains impurities from the chemical breakdown of wood fibres and from the machinery used in producing mass quantities.
The chart is in need of stabilization to allow usage without further damage. But in order to properly treat this item, I first need to understand the roots of its problems. What causes the paper's brittleness? What does that mean for its future, and for my treatment?
The obvious first step is to identify the pigments used to colour the chart. Many pigments contain materials that can be damaging to cellulose, much like the impurities in 19th century papers. From a simple visual and historical standpoint, one can deduce that the green colour is likely to be a copper-based pigment known as verdigris.
Verdigris is one of the oldest manufactured pigments, having been produced in ancient Greece (thus its name, "green of Greece"). It is an umbrella term of sorts used to describe the colour obtained from the corrosion of copper, which may be caused by exposure to a variety of substances. In Greece, the copper was stored with vinegar; in later centuries, verdigris was produced alongside the wine industry, with copper plates exposed to the fermenting grapes. One of the primary elements of vinegar is acetic acid-a simple acid.This, combined with the copper carbonate patina that forms naturally when copper is exposed to moisture and carbon dioxide in the atmosphere, creates the bluish-green of verdigris.
To observe this reaction, I set up a copper acetate system at my bench. Several plates of copper were placed in an open glass beaker. This beaker was placed in a larger glass container, which contained a small amount of acetic acid. The larger container was sealed, creating a system where the copper was exposed to the acetic acid vapours, but not the acid itself, much like in the traditional winery system. After a month of exposure, a blue-green patina had formed covering the copper's entire surface.
I then scraped the corrosion off of the copper plates. Half of the blue powder was then dissolved in acetic acid and allowed to dry, creating what is called "neutral verdigris." The two types of powder were then mixed with water and gum arabic. I brushed both types of verdigris onto white paper, photographed them, and stored them to observe how they age.
Analysis:
Before taking any action on treating the chart, I needed to confirm the presence of verdigris. Though it is nearly impossible to determine the definite pigments used by the maker of the chart, it is entirely possible to detect the presence of key elements. In the case of verdigris, that key element is copper.
X-ray fluorescence (XRF) analysis is a relatively quick, non-interventive method of identifying the inorganic components of an object. It is a surprisingly simple-to-use system that takes advantage of the precise atomic structures of the elements. Each element contains numerous electrons, which have a precise range of energies. The range of electron energies associated with each element is unique. Therefore the range of gaps between the electron energies of each element is also unique to each element.
XRF spectrometers shoot a series of x-rays-short electromagnetic waves, an intense energy source related to visible light-into the object under analysis. These x-rays act on the electrons whirling around at specific energy levels. When hit by an x-ray, an electron may be shot out of its energy level and ejected from the atom. If there are any electrons present in an energy level higher than the lost electron, then one of those higher-energy electrons will automatically drop down to replace it. When doing so, it loses that energy difference in the levels. This energy is emitted as fluorescence.
When you consider the fact that each element has a precise number of electrons, then you will know that each element also contains a precise amount of electron energy, which acts as a sort of fingerprint. When x-rays are shot at the atoms, then, we can use the amount of energy being released to identify what element on the period table we are looking at. This is how XRF scanning works-by analysing this fluorescence that follows the x-rays.
All pigments and inks on the chart underwent XRF analysis, as did an area of paper that had no pigment. By comparing the spectrum of each pigment to the spectrum of the blank paper, I could make out what elements are present in the pigments. As you can see on the following spectrum graph, the spectrum of the green area showed a far greater presence of copper than the blank paper.
It is easy to conclude, then, that it is a copper-based pigment, and, due to its appearance, historical context, and degrading effects, it is most likely verdigris.
The areas with the brown wash, which I assumed were iron gall ink and therefore expected to display a large presence of iron, actually did not. The iron peak on the spectrum is very similar to that of the blank paper. Instead, the brown area showed a similar copper peak to the green area, though not as high. It is likely, then, that the wash is itself a copper-based pigment, rather than iron gall. As the verdigris colour range expands into blues, it is possible that this brown was once a blue, representing water, and that it has faded over its 280-odd years. We can safely assume, then, that the chart is 90% covered with copper-based pigments.
So, why does verdigris cause the paper to be brittle?
Cellulose, the primary component of paper, is an organic polymer-simply put, a long chain-like molecule made up of many subunits. It is this chain-like structure that creates the strong but flexible fibres of paper. Essentially all characteristics of paper are rooted in this shape.
Over time, cellulose undergoes oxidation, the inevitable cleavage of its units through the introduction of water to its structure. The atoms of a water molecule split and join themselves to the ends of the cellulose subunits-an OH group on one end, and a single hydrogen atom on the other-causing them to break off from one another, weakening the once-strong cellulose.
oxidation does not occur quickly in cellulose under normal conditions. The bonds between the polymer units are much stronger than those with water. The water is given a great deal of help, though, when certain foreign materials are present. Metals like copper act as catalysts to the oxidation process, rapidly increasing the rate of breakdown.This is what is occurring all over the chart, which we now know is absolutely covered in copper.
What does this mean for my treatment?
The most obvious, efficient and effective treatment for a flat object with tears is to use Japanese tissue patches adhered with wheat starch paste. Following this analysis, though, I know that I am dealing with a reactive, copper-covered surface. Introducing more moisture to the system would only quicken the cellulose oxidation. As wheat starch paste is a very wet adhesive, I need to avoid using it.
There are treatments for verdigris damage that have been successfully performed by advanced conservation workshops, in which materials are introduced to the system to bond with the copper ions, leaving them less able to cause oxidation.
My treatment will be far less interventive, as I won't be changing anything in chart at the chemical level. I will be using a material known as "re-moistenable tissue." In this technique, Japanese tissue is coated with water-based adhesive, which is allowed to dry. The adhesive is then reactivated with a minimal amount of moisture just before application. Though there is inevitably some moisture introduced, this is a much drier process than typical Japanese tissue repair using fresh water-based adhesive.
This process and its results will be discussed on Part Two of this post.
