Book Reviews

Abstract

The long-term stabilization of marine archaeological iron, whether cast or wrought, continues to challenge conservators responsible for treatment of this material. Results and observations obtained from past treatments highlight the daunting, prolonged, and laborious efforts required to desalinate large and complex ferrous artifacts recovered from the ocean. In general, the higher an artifact's chloride level, the less stable it is. Consequently, any stabilization treatment must involve the removal of as much Cl^?1 as possible without affecting the integrity of the corroded artifact. This problem is particularly acute with corroding cast iron objects that have formed thick, fragile, and highly unstable corrosion layers. Over the course of the twentieth century, conservators have used a variety of techniques in an attempt to mitigate the negative effects of chloride ions on iron artifacts and prevent disintegration. In spite of early promise, each of these stabilization techniques has significant disadvantages, particularly with regard to treatment efficiency, duration of treatment, and/or unacceptable risks to the artifact during treatment. For these reasons, conservators and conservation scientists at the Warren Lasch Conservation Center in Charleston, South Carolina, decided to look at the possibility of using subcritical fluids to stabilize archaeological iron. This paper compares the efficiency and effectiveness of traditional stabilization techniques (i.e. alkaline soaking and cathodic polarization) to subcritical fluids on wrought iron rivets and metal shavings from the H.L. Hunley submarine as well as Civil War era cast iron artillery shells recovered from a marine environment.     

Models for Chloride Ion Diffusion in Archaeological Iron

Summary

The rate at which chloride ions diffuse from archaeological iron into a treatment solution depends on how the chloride ions are initially distributed in the corrosion layer. This paper compares solutions of the diffusion equation for two limiting cases: (1) where the chloride ions are initially spread uniformly through the corrosion layer; and (2) where the chloride ions are initially concentrated at the interface between the iron and the corrosion layer. Although the first model has been used in the past to describe chloride ions diffusing from marine iron, the second is more appropriate in cases where corrosion has drawn chloride ions toward the iron surface. Because diffusion processes in archaeological iron are complicated, the limitations of both these models are discussed.     

Post-treatment Study of Iron/Sulfur-containing Compounds in the Wreck of Lyon Saint-Georges 4 (Second Century ACE)

ABSTRACT

A pre-restoration diagnosis revealed a high amount of pyrite in the wood of the Lyon Saint-Georges 4 shipwreck (end of the second century). The occurrence of this phase is supposed to result from the microbiologically influenced corrosion of the iron fasteners. So, all the nails and metallic elements were removed from the remains before treatment and the wreck was consolidated by polyethylene glycol impregnation coupled to a specific desalination process. Treated and non-treated samples extracted from the wreck were studied in order to identify the iron/sulfur-containing compounds present in the wood before and after treatment and figure out its effect. Sample analyses relied on an original approach combining magnetic characterization methods and more common elemental and structural analysis methods. The results showed that the treatment was effective in removing soluble salts. However, a large amount of unstable iron sulfides remained inside the wood.     

An Investigation of Deterioration of Archaeological Iron

Abstract

A small but significant proportion of the archaeological iron objects in the British Museum collection have been problematic in that some of them have required repeated treatment. The deterioration of iron objects during storage was studied using microscopy, Raman spectroscopy, X-ray diffraction, scanning electron microscopy/energy dispersive X-ray analysis, X-ray fluorescence spectroscopy and ion chromatography. A total of 125 iron objects, including 56 from two British Roman sites and 69 from three Anglo-Saxon sites, were studied. Both surface corrosion and corrosion layers on polished cross-sections were investigated. The study revealed that the present condition of the iron objects varies; some are in a stable condition and others have been deteriorating, with pitting and weeping present on the surface. No treatment method had stopped corrosion for every object, although the alkaline sulphite treatment seemed to be more effective than the other methods. The results also showed that chloride ions can be present at the interface of the metal and inner corrosion layers. It is recommended that iron objects which have been mechanically cleaned without desalination treatments are kept in a dry environment to prevent further deterioration.     

Cellulose degradation in the Vasa: The role of acids and rust

Abstract

The oak timbers of the Swedish warship Vasa are deteriorating. High amounts of oxalic acid have been found along with a low pH and low molecular weight cellulose deep in the wood timbers. The iron-rich surface wood differs from the interior wood in that it displays higher pH and cellulose with higher molecular weight. The objective of this study was to determine why there is a difference in cellulose degradation, pH, and oxalic acid amount between the surface region and the interior of the Vasa timbers. Analysis of cellulose weight average molecular weight by size exclusion chromatography was performed, as well as quantification of oxalic acid and iron by high-performance anion exchange chromatography and atomic emission spectroscopy, respectively. It was found that a decrease in iron content coincides with an increase in oxalic acid concentration and a drop in pH at a certain depth from the wood surface. When iron-rich surface wood samples from the Vasa were mixed with an aqueous solution of oxalic acid, a fast increase of pH over time was observed. Neither interior wood poor in iron nor the fresh oak reference showed the same neutralizing effect during the time of measurement. This indicates that the presence of iron (rust) causes a neutralization of the wood, through the formation of iron(III) oxalato complexes, thus protecting the wood from oxalic acid hydrolysis. This effect was not observed to the same extent for other acids observed in Vasa wood (sulfuric, formic, glycolic, and acetic acids).     

The impact of chloride desalination on the corrosion rate of archaeological iron

Abstract

Although desalination of archaeological iron reduces its chloride concentration and enhances object stability, the reduction in corrosion rate that this produces has never been quantified. This study measures post-treatment corrosion rates in accelerated corrosion environments to identify the impact of removing chloride ions on corrosion rate. Thirty-five archaeological iron nails, treated individually in either alkaline sulphite or nitrogen-deoxygenated sodium hydroxide, were exposed to 75°C and 75% relative humidity together with 31 untreated objects from the same archaeological sites. Object weight change and visual examination of physical change before and after the test period were used to monitor corrosion. 77% of treated objects showed no weight gain and no visible signs of corrosion, while 90% of untreated objects did corrode. The impact of chloride on corrosion of untreated objects was clearly established by a significant linear correlation between chloride content and weight gain. Treated objects with <400 ppm chloride content showed no corrosion behaviour. Corrosion of treated objects was attributed to incomplete treatment: 93% of objects treated to <5 mg/l Cl^? in the final solution bath displayed no corrosion behaviour. Based on these results, desalination of iron objects to enhance their stability offers a valuable option for reducing corrosion rates of archaeological iron, which should increase object lifespan. The results also raise the question of whether low levels of post-treatment residual chloride produce corrosion of any significance. Answering this will be an important step forward for managing the preservation of archaeological iron.     

The iron in black weathering crusts on Saxonian sandstones investigated by Mössbauer spectroscopy

Abstract

Black weathering layers on stone often contain iron. This is particularly true for the Turonian sandstones which were used extensively in architecture and sculpture in Saxony. In order to obtain more knowledge about the oxidic iron phases, samples were collected in the city of Dresden. Mineralogical and analytical characterization was provided for the black layers and the supporting stone. Material originating from Cotta and Posta stone was submitted to Mössbauer spectroscopy. After ample comparisons of Mössbauer parameters and evaluation of infrared spectra it was concluded that the predominant neogenic iron phase is not iron(III) oxide hydrite (ferrihydrite) Fe₂O₃·1·8H₂O or jarosite KFe₃(SO₄)₂(OH)₆ or goethite α-FeO(OH) but a defectively ordered precursor of this last compound, characterized by different environments of the iron(III) ions. By the reducing action of organic matter, magnetite Fe₄O₄ is formed and this contains up to 45% of the iron present in a black layer.     

Hydrogen plasma reactions in a d.c. mode for the conservation of iron meteorites and antiquities

Abstract

Four iron meteorite fragments and 13 archaeological iron objects were exposed to a lowpressure hydrogen plasma for up to two weeks in a pilot apparatus. In the experiment the objects acted as cathodes in a 600–1600V d.c. electrical field. While the chlorine-containing β-FeOOH, akaganeite, was present in all objects before the test, no akaganeite and no chlorides were present afterwards. The objects, including one meteorite which had been exposed for more than 5000 years to seawater, were found to be stable after the glow discharge treatment. The major mineral after treatment was identified as magnetite, and the objects generally assumed an iron-grey surface colour. No sputtering and no reduction to free iron took place under the test conditions, operating at or below 120°C and at pressures of O·4–1mb hydrogen.     

The efficiency of chloride extraction from archaeological iron objects using deoxygenated alkaline solutions

Abstract

Chloride-contaminated archaeological iron is unstable and problematic to store and display within museum collections. Reducing its chloride ion content using aqueous desalination followed by storage in controlled relative humidity offers one treatment option. This study reports a quantitative assessment of chloride extraction by aqueous deoxygenated alkaline desalination solutions from 120 individual archaeological iron nails. The three treatment methods comprised alkaline sulphite solution (0.1 M NaOH/0.05 M Na₂SO₃) at room temperature and at 60°C and sodium hydroxide solution (0.1 M) deoxygenated using a nitrogen gas positive pressure system at room temperature. Chloride extraction was monitored using a specific ion meter. The nails were digested after treatment to measure their residual chloride content. A wide range of extraction patterns emerged, with the majority of individual treatments extracting 60–99% of the chloride present. Residual chloride levels for 87% of the objects fell below 1000 ppm and 42% were below 200 ppm. Although no treatment extracted 100% of the chloride in the object, alkaline desalination produced very significant reductions in chloride content. The impact of this on future corrosion of the objects is discussed. This quantitative and statistically viable assessment of deoxygenated desalination treatments provides evidence to support their use in conservation practice, which will impact on procedures for the preservation and management of archaeological heritage.     

Iron removal from waterlogged wood: Extraction by electrophoresis and chemical treatments

Abstract

The presence of iron oxides (lepidocrocite, goethite) in archeological wood may result in a degradation of the wood matrix. Extraction of these iron oxides is largely dependent on their solubility. In this study, balsa wood samples were impregnated with iron oxides to test extraction treatments. Additionally, archeological wood samples were also examined to determine treatment efficiency. Electrophoresis and simple immersion treatments were performed using various chemical solutions: a neutral and a conductive substance (potassium nitrate), an acid (acetic acid), three alkaline chelating agents (tri-ammonium and tri-sodium citrate and sodium oxalate), three acidic and slightly acidic chelating agents (ethylenediaminetetraacetic acid (EDTA), citric acid, and oxalic acid), and a reductant (sodium dithionite). Potassium nitrate did not extract sufficient amounts of iron, irrespective of whether the treatment was conducted by electrophoresis or simple immersion; any observable dissolution was attributed to protonation because of the acidic pH around the anode (as low as 3). Dissolution in acetic acid did not extract iron with either treatment. Strong chelating agents improved extraction, and these compounds gave the best results for simple immersion, particularly EDTA. This chemical is well adapted for use on archeological objects because of its chemical properties (stability constant, speciation based on pH). The addition of sodium dithionite to the solution improved dissolution. Even though electrophoresis improved extraction (in particular for tri-ammonium citrate), none of the tested chelating chemicals were suitable for electrophoresis because of a significant increase in temperature as well as high anode corrosion. The presence of iron sulfide in the archeological wood limited the effectiveness of the tested chemicals. A pre-treatment in sodium persulfate was expanded to include oxidized iron sulfide in oxy/hydroxide iron, which improved the extraction rate.     

Optimisation d'un traitement de dechloruration d'objets ferreux par plasmad'hydrogene

Abstract

The method of removing chlorides from archaeological iron objects using hydrogen plasma at low pressure has been studied and improved. To carry out the initial experiments and to limit the use of ancient material, synthetic samples were made from iron powder and corrosion products consisting mainly of akaganeite. The time and the temperature required for the complete removal of the chlorides were determined. It was also possible to demonstrate that magnetite, and iron above 400°C, are the final reaction products of the reduction of akaganeite.     

A study of the effects of conservation treatment on the Fanshawe Archive

Abstract

Valence House Museum (Dagenham, England) has, since 1963, housed not only a large collection of Fanshawe family portraits but also a very rich family archive which comprises estate and private papers from the sixteenth to the twentieth centuries of at least two branches of the family. Of national and international interest are the papers of Sir Richard Fanshawe (1608-66) relating to his three embassies to the Iberian peninsula in the 1660s. Three documents from this collection were taken for conservation and treated using both aqueous and non-aqueous methods of deacidification. Prior to treatment, the papers were characterized using the techniques of scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), electron spectroscopy for chemical analysis (ESCA or XPS), thermogravimetry (TGA) and thermomechanical analysis (TMA and DLTMA). Measurements showed that the three papers were similar in composition and were basically cellulosic materials with varying trace levels of Fe and Cu impurities. All three documents contained iron gall ink; the ink of the Dublin letter had an additional feature in the form of small bright specks which were characterized as a form of mica. After conservation treatment the papers were examined using the above techniques in an attempt both to determine and to quantify any changes which had occurred during the treatment process. Scanning electron micrographs showed that pronounced changes had occurred in the surfaces of the treated papers. With XPS it was possible to measure the change in the surface composition of the papers which occurred on treatlp.ent. It was also found that the treatment reduced the thermal stability of the papers in all three cases.     

A METALLURGICAL EXAMINATION OF TWO EARLY IRON SWORDS FROM LURISTAN

Abstract

Two iron swords from Luristan were examined by X-rays and using metallographic sections. The first sword was shown to have a hilt made of five pieces of metal joined by four rivets, while the pommel was secured by a dowel on the end of the tang. Hilt and blade were sectioned for metallographic study. This showed that while the core of each part was pearlite, decarburization had taken place at the surfaces to give ferritic structures. The second sword was shown to have been made of seven pieces. The pommel was secured as in the first sword, but all other pieces were joined by fitting into prepared slots, the edges of which were burred. A section was cut from the hilt at a point where a decorative band had been applied. This showed large-grained ferrite at the surface and fine-grained ferrite with cementite at the boundaries in the core. It is argued that work previously carried out by other metallographers was inconsistent due to inadequate sampling, and that swords of this type were essentially of wrought iron, the presence of pearlite being accidental rather than intentional.     

Influence of crucial parameters on the dechlorination treatments of ferrous objects from seawater

Abstract

This article compares chemical dechlorination treatments (immersion in sodium hydroxide or alkaline sulphite) and electrochemical treatments of iron bars from the Gallo-Roman period excavated from a marine environment. Some important parameters, such as storage before treatment, temperature, solution composition, and drying after treatment, were varied during the study to assess their influence on the chloride extraction process. The kinetics of these treatments depend mainly on chloride diffusion through the corrosion layers. The kinetics are promoted by high temperatures and, in the case of electrolysis, by the electric field effect. The reduction of corrosion products during electrolysis occurs only for objects previously stored in air. In fact, the manner in which the objects are stored before treatment is critical in the dechlorination processes. The sooner the objects are treated after excavation (with water storage), the better the removal of chloride ions, in both chemical immersion and electrolysis treatments. But if the object is stored in air, material losses occur, and only electrolysis results in complete extraction of the chlorides. These differences are due to modifications in the corrosion products during storage. Drying after treatment also has a significant impact on the composition of the corrosion layers. If the objects are dried too quickly, Fe(OH)₂ oxidizes into FeOOH, which thus forms a layer with low cohesion.     

Les préparations colorées des tableaux de l'Ecole Française des dix-septième et dix-huitième siècles

Abstract

Study by X-ray microanalysis of 155 coloured grounds from French paintings of the seventeenth and eighteenth centuries allowed us to establish the nature of their constituents. The colour was introduced by three varieties of pigment, brown earths, ochres and iron oxides which were used in combination with calcium carbonate, lead white and minium. This confirms the recipes mentioned in the old treatises. The presence of barium sulphate, unexpected in this period, is connected with the nature of the coloured pigments and the place of execution of the works.     

A note on the analysis of metal threads

Abstract

The present work demonstrates by means of two practical examples the extreme importance of examination of cross sections of metal threads to obtain an overview of their structure and of the different materials employed in their manufacture. Problems that might be caused by disturbance of surface features during polishing were prevented by covering the sample before embedding with a layer of nickel, using electroless deposition. Results from scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDX) showed that, although the mean chemical composition of both metal wires is the same, their metallurgical structure is completely different: one consisting of pure silver with inclusions of copper particles and the other a silver–copper alloy.     

Metal corrosion rates in aqueous treatments for waterlogged wood-metal composites

Abstract

Corrosion rates are presented for metals in the aqueous solutions intended for treating both the wood and the metal components of a waterlogged wood-metal composite object. The metals include lead, mild steel, cast iron and the copper alloys of bronze, cast brass and wrought brass. The corrosion rates were determined using a standard weight-loss method in which clean metal surfaces were immersed in the solutions for two months at room temperature. The treatment solutions included Acrysol® G-11O for lead, Witcamine® RAD 1100 for copper alloys and Pluracol® 824 or the corrosion inhibitor Hostacor® KS 1 in PEG 400 for iron. These treatment solutions were prepared in distilled water as well as in a synthetic seawater. For reference, metal corrosion rates were also determined in distilled water, synthetic seawater and PEG 400. The results demonstrate that, in general, clean metal surfaces corrode more slowly in the treatment solutions than in PEG 400 solutions. However, before recommending these solutions for the treatment of composite objects, further testing is needed to determine how effective they are in reducing the corrosion rates of metals covered with a thick corrosion crust, and to see if Acrysol G-110 is capable of bulking and consolidating waterlogged wood.     

A note on burnt yellow earth pigments: documentary sources and scientific analysis

Abstract

The red iron oxide pigment known as burnt yellow earth has been used as a substitute for naturally occurring red earth since earliest times. Documentary evidence indicates that this pigment was often produced by heating yellow earth for anywhere from a few minutes up to several hours at temperatures below 8000 C. When yellow earth is burnt under these conditions, its constituent goethite is transformed to red iron oxide with an anomalous diffraction pattern corresponding to a disordered haematite crystal structure. Using pigments of known origin and processing, it is shown that X-ray microdiffractometry can be used to classify very small samples of red iron oxide pigments and that, in some cases, the diffraction pattern can provide evidence to support or refute the use of burnt yellow earth.