by Yun Liu
Paper heritage has long been threatened by the degradation caused by iron gall inks. Research has confirmed that the damage is a result of complex overlapping of different processes which are mainly generated by two degradation mechanisms of cellulose—acid catalysed hydrolysis and metal-catalysed oxidation, both being strongly influenced by ink properties. As part of the course work, we recently performed experiments of model ink synthesis in the Heritage Science Laboratory (HSL) at UCL Institute for Sustainable Heritage to get a good understanding of historic ink recipes as well as the qualities and behaviours of the reconstructed inks. We would like to share the procedures and results, as well as a few challenges we encountered during the reconstruction experiments of iron gall inks.
Materials and methods
Historically, there was a large variation in the composition and manufacturing method of iron gall inks, and both contributed to the variations in ink properties. To achieve simplification and representation, only four essential components were used in our experiments: gallic acid, iron (II) sulfate heptahydrate (FeSO4·7H2O), gum Arabic, and water. Due to sample availability, gallic acid was sourced in the form of Aleppo gall nuts (HSL, unspecified origin), pure tannic acid extracted from Chinese gall nuts (SIGMA-ALDRICH), and pure gallic acid chemicals (SIGMA life science) (Figure 1).
We adopted the methods for the preparation of historically representative model iron gall inks described by Stijnman in the book ‘Iron Gall Inks: On Manufacture, Characterisation, Degradation and Stabilisation’, edited by J. Kolar and M. Strlič (2006). Four model inks were reconstructed with varying proportions of FeSO4·7H2O and gallic acid, the two essential colour forming ingredients. A comparison of the recipes that were used for ink synthesis is summarized in Table 1. The proportions of ingredients were expressed in molar ratio under certain assumptions as necessary.
Table 1. Comparison of recipes used for the synthesis of model iron gall inks
|Ink 1||Ink 2||Ink 3||Ink 4|
|Molar ratio of FeSO4·7H2O to gallic acid||18 : 13||18:13||18:13||3 : 1|
|Gallic acid source||Aleppo gall nuts||Tannic acid extracted from Chinese gall nuts||Gallic acid||Gallic acid|
|Gallic acid source (g)||5.00||2.50||2.21||0.10|
|gum Arabic (g)||2.50||2.50||2.50||2.50|
The experiments started with a typical proportion in historical recipes which has a 1:1 proportion (w/w) of Aleppo gall nuts to FeSO4·7H2O (Ink 1). This gave an approximately 18:13 molar ratio of FeSO4·7H2O to gallic acid, under the assumption that tannic acid content of Aleppo gall nuts was 50% by weight, and the structure of Aleppo tannic acid was based on glucose, which was esterified with five gallic acids, an average of 2-6 range. Following the calculations, 5 g of coarsely crushed Aleppo gall nuts were boiled in 60 mL deionized water for 30 minutes, while the suspension being stirred at regular intervals. FeSO4·7H2O and gum Arabic were dissolved in 20 mL of deionized water respectively and were added to the cooked Aleppo gall nuts suspension in sequence (Figure 2).
Three more inks were synthesized by dissolving and mixing chemicals in a similar manner as the first one (Figure 2). Ink 2 and 3 had the same FeSO4·7H2O to gallic acid ratio as ink 1 for a comparison of different gallic acid sources, whereas the ink 4 had a 3:1 molar ratio of FeSO4·7H2O to gallic acid, which has been known to have the most chemical stability. The weight of tannic acid extracted from Chinese gall nuts was calculated based on the specification provided by the chemical manufacturer that each tannin molecule was formed by units of glucose bonded with ten gallic acid. Pure chemicals were used for the preparation of the most stable ink to minimize uncertainties.
Due to sample availability, we found it difficult to draw solid conclusions on ink properties in relation to proportions of FeSO4·7H2O and gallic acid. But there were still some interesting observations. As for ink colours, all the model ink solutions looked dark blue in capped glass bottles (Figure 2-d). However, the different behaviours of inks of various compositions started to show up when we applied the inks on paper (Figure 3). The hues of freshly prepared inks differed from blue-black (Ink 3 and 4) to purplish black (Ink 2) to brownish black (Ink 1) on Whatman filter paper. Although we tried to keep the ratio of FeSO4·7H2O to gallic acid constant for Ink 1 – 3, the behaviours of these inks still varied due to different gallic acid sources.
As for ink pH, we measured the pH of ink solutions as well as ink applications on Whatman filter paper No. 1 using surface method. The results of pH measurements are presented in Table 2. Apart from the systematic errors embedded in each measuring methods, the difference between the pH of the ink solutions and the pH of the same ink applications on paper was likely to be a result of the neutralisation by the paper support. We observed that inks with the same molar ratio of FeSO4·7H2O to gallic acid tended to have the same acidity, regardless of the variations in gallic acid sources. We also found it interesting that the most acidic ink was the one resembled the ‘most stable ink’, which seemed to be contradictory to the principles of acid-catalysed hydrolysis in cellulose degradation.
Table 2. Comparison of pH values of model ink solutions and surface pH values of model inks applied to Whatman filter paper No. 1.
|Ink 1||Ink 2||Ink 3||Ink4|
|Surface pH on Whatman filter paper No. 1||3.5||3.4||3.5||3.0|
|Model ink solution pH||2.9||2.8||3.1||2.5|
A few things to clarify
When we started the experiments, we consulted various literatures on iron gall ink recipes and reconstruction methods. We simplified our experiments to focus on the essential components only, however, we still found it difficult to either repeat historical ink recipes or reconstruct inks comparable to the model inks that had been used for other researches, because the terminologies were usually not well defined.
The most confusing term we found was tannin, a common term that was used to refer to gallic acid. It is ambiguous to us because tannin is a broad term which includes hydrolysable tannins and condensed tannins. Precisely, the tannin we refer to when we talk about iron gall ink production is tannic acid, a specific hydrolysable tannin that is extracted from gall nuts. Tannic acid molecule contains gallic acid, which is the component that reacts with iron sulfate to form ink colour.
Tannic acid can be extracted from different natural sources, Aleppo gall nuts and Chinese gall nuts being the two major ones for iron gall ink production. Different natural sources leads to variations in molecular structure of tannic acid which contains different amount of gallic acid units per molecule. For example, one Aleppo tannic acid molecule has been proved to be esterified with 2-6 gallic acid, whereas one Chinese tannic acid molecule is usually bonded with 10 gallic acid. Therefore, the use of tannin in literature without further specifications may cause confusion about the exact molar ratio of gallic acid to other ingredients in both historic and model ink recipes used in studies.
Similarly, clarification may also be needed to specify iron content of iron gall inks. There have been discussions about the possible forms of vitriol that might have been used in historic ink manufacturing since the hydration state of iron(II) sulfate was not specified in historic recipes. For model inks, the distinctions among iron (Fe), iron(II) sulfate (FeSO4), and iron(II) sulfate heptahydrate (FeSO4·7H2O) sometimes are not well defined in literature. Iron or Fe are usually the general terms that have been used to refer to either of these chemicals, which may not be clear enough especially when comparable reproduction is considered.
Proportion of ingredients was another crucial aspect for the reconstruction experiment. However, in literature, either being in weight or molar ratio is not always clearly explained in both historic and model ink recipes. It seems to be a convention that the ratio of gall nuts to vitriol is usually expressed in weight ratio in historic recipes, whereas the molar ratio is used in model ink recipes. But detailed explanations may be helpful especially when talking about historic and model ink recipes at the same time.
We acknowledge that clarification in these aspects does not necessarily contribute to research data interpretation. For example, differentiation among iron, iron (II) sulfate, and iron (II) sulfate heptahydrate regarding molar ratios would not make a difference. But clear definitions would be greatly appreciated in order to communicate with a broader audience and increase the opportunity for other researchers to make use of the data in the future.