Next Article in Journal
Effect of Al2O3 Content on High-Temperature Oxidation Resistance of Ti3SiC2/Al2O3
Next Article in Special Issue
Characterization of the Materials and Techniques of Red Lacquer Painting of a Horizontal Plaque Inscribed by General Feng Yü-hsiang
Previous Article in Journal
Effect of Phaeodactylum Tricornutum in Seawater on the Hydration of Blended Cement Pastes
Previous Article in Special Issue
Application of Advanced Analytical Techniques in Organic Cultural Heritage: A Case Study of Ancient Architecture Relics in the Palace Museum (Beijing)
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Coatings
Volume 12
Issue 11
10.3390/coatings12111640
coatings-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Multi-Analytical Protocol for Decision Making to Study Copper Alloy Artefacts from Underwater Excavations and Plan Their Conservation

by
Francesca Gherardi
1,* and
Heather Stewart
2
1
Investigative Science, Fort Cumberland Laboratories, Historic England, Portsmouth PO4 9LD, UK
2
MSDS Marine Ltd., Portsmouth PO4 9LD, UK
*
Author to whom correspondence should be addressed.
Coatings 2022, 12(11), 1640; https://doi.org/10.3390/coatings12111640
Submission received: 16 September 2022 / Revised: 20 October 2022 / Accepted: 24 October 2022 / Published: 29 October 2022
(This article belongs to the Special Issue Coatings and Interfaces in Cultural Heritage Conservation and Preservation)
Browse Figures
Versions Notes

Abstract

:
The multi-analytical protocol currently in use at Historic England for the investigation of copper alloy artefacts recovered during underwater excavations aims to determine their manufacturing processes, identify repairs, and assess their state of preservation. Each step of the scientific analysis is described, and the results obtained from the study of a selection of copper alloy objects recovered from the Dutch East India Company (VOC) Rooswijk shipwreck (1740) are used as examples of the application of the protocol to gain archaeological, metallurgical, and conservation data. This information is crucial to plan the most appropriate procedures and determine treatment steps for the study and conservation of copper alloy artefacts from the marine environment.

1. Introduction

The Rooswijk was a Dutch East India Company (VOC) vessel, which sank in Goodwin Sands (at the northeast end of the Kellet Gut), off Kent, in 1740 during her second journey from Texel (Netherlands) to Batavia (today Jakarta). The vessel and its cargo of trade goods are owned by the Dutch Government and are managed as a protected wreck site by Historic England. During excavations (2017 and 2018), funded by the Cultural Heritage Agency of the Netherlands (Ministry of Education, Science and Culture) in the framework of the #Rooswijk1740 project, many artefacts (silver coins, glass beads, bottles and vessels, copper alloy and pewter objects) were recovered and transported to Historic England, Fort Cumberland laboratories (FCL) (Portsmouth, UK) for the post-excavation phase, to be conserved and studied [1,2,3]. During the excavation of the shipwreck several copper alloy artefacts were recovered, including small (beads, rings, wires and thimbles) and large objects (trumpet, cauldron, pan, teapot, candlesticks, and candlesnuffers). Their scientific analysis is essential to determine the manufacturing processes and their alloy composition, identify repairs and corrosion products, and assess their state of preservation.
Many analytical techniques and protocols are available to study metal artefacts and their corrosion [4,5,6]. In particular, for archaeological copper alloy artefacts from underwater sites, the analytical protocol includes optical and 3D microscopy, X-ray radiography, scanning electron microscopy, spectroscopies (X-ray fluorescence, Raman and Fourier transform infrared spectroscopies), X-ray powder diffraction and more [7,8,9,10,11,12,13,14,15]. However, there are no defined protocols that standardize the different steps to investigate them prior to conservation. This paper explores the use of the multi-analytical protocol in use at Historic England for the study of copper alloy artefacts, with a focus on the objects recovered from the Rooswijk, which are examined in the framework of #Rooswijk1740 project. This protocol does not cover every technique that can be used for the technical examination and scientific analysis of maritime archaeological artefacts, but it includes the most representative and commonly used ones. In addition, the sequence of the steps is flexible, and can be easily adapted according to the individual artefact.
The scientific data gained by different analytical techniques are crucial to develop best practice in the study and conservation of copper alloy archaeological artefacts.

2. Materials and Methods

2.1. The Protocol

The first step of the protocol includes the visual and X-ray radiography investigations of the artefacts (Figure 1). Many of the small copper alloy artefacts (rings and beads) were found encased in dense concretions formed on the seabed around iron objects. X-ray radiography is used to record the condition of artefacts and to visualise what is inside concretions. X-ray images guide the conservators in the safe removal of the artefacts. This non-destructive technique allows for the study of the internal structure and defects of the objects, providing information about their state of preservation. Once the objects are removed from the concretion, visual observations can contribute to identify manufacturing processes and features such as hammer or touch marks.
Prior to carrying out desalination and surface cleaning, large objects are analysed with portable X-ray fluorescence (pXRF) spectroscopy to detect specific types of copper alloys and allow for the rapid grouping of artefacts from the same class of material (Figure 1).
The analysis of the elemental composition by pXRF can help in reconstructing the integrity of an object with the preliminary attribution of small pieces to a main object, and with the identification of different repairs. In addition, pXRF can support heritage scientists in the selection of the most suitable areas for collecting the samples for subsequent quantitative analysis [16].
Despite the advantages of being fast, non-invasive, and non-destructive, the results obtained by pXRF can be affected if the material is not sufficiently thick to absorb all of the X-rays from the XRF beam, as the required thickness depends on the material, and it is greater for low-density materials [16]. The penetration of X-ray photons depends on the energy, and is variable depending on the considered element and the density/composition of the material [17]. The presence of corrosion layers can generate misleading results about the bulk alloy composition. In addition, a reduction in the collected signal can result from the study of objects with a complex shape and surface roughness, such as a cauldron, teapot, or trumpet.
To overcome these limitations, analyses by a benchtop micro-XRF (μXRF) are performed on metal artefacts that fit in the machine (the sample chamber size of the instrument in our laboratory is W × D × H: 600 × 350 × 260 mm) (Figure 1). It is possible to carry out the measurements on a small area, from which the corrosion patina can be scratched from the surface to access the bare metal. This allows for the detection of the alloy in bulk, as well as the trace elements, which are necessary to gather information about specific manufacturing technology and provenance. In addition, μXRF can be used to map the distribution of chemical elements of the artefacts, allowing for the elemental characterisation of surface patina and corrosion.
To study the morphology of the alloys, their composition, and the corrosion layers on their surface, small samples (or surface powder) can be collected to prepare cross-sections. Samples are taken on areas which are representative of the object, and where there is an existing loss or fracture, using small tweezers or a scalpel. If multiple objects of the same typology (beads, rings, thimbles, etc.) are recovered, samples are collected only from a selection of artefacts that display common features. If the object exhibits surface corrosion or heterogeneity, powders can be gently scratched from selected areas and representative samples can be taken to study the stratigraphy. The use of a wide range of microanalytical methods that are nondestructive for the samples is essential to preserve them for future investigations. The cross-sections are observed by optical microscopy to identify the stratigraphy and the presence of corrosion layers underneath the surface. To study the samples at higher magnifications and, to characterise their composition, the cross-sections are analysed by scanning electron microscopy combined with an energy-dispersive spectrometer (SEM-EDS) (Figure 1). Compositional and morphological data can offer insight to the state of preservation of the objects and the presence of any defects and fractures. It can unravel the nature of eventual corrosion layers developed on the artefacts, together with information about their thickness, porosity, and distribution on the original substrate. SEM-EDS mapping can show the distribution of chemical elements in each layer, identifying leaching of ions from the bulk and the formation of corrosion layers. This important information can guide the conservators in the selection of the most appropriate cleaning procedures for the removal of all deposits and reactive compounds that are potentially harmful to the substrate. In addition, it can be used to monitor the effectiveness and the selectivity of conservation treatments.
The protocol includes the use of Fourier transform infrared (FTIR) spectroscopy and X-ray diffraction (XRD) on powders collected from the copper alloy objects to characterise the surface corrosion layers (Figure 1). Compared to FTIR and Raman spectroscopy, a relatively high amount of powder is required for XRD. Some inorganic compounds (oxides, sulfides, etc.) have no vibrational modes in the mid-infrared region (the most common region to study cultural heritage materials), and their identification is only possible if the FTIR spectrometer is equipped with a far-infrared detector (650 to 50 cm−1) [18,19]. This step helps in the set-up and selection of the most appropriate desalination treatments to remove harmful salts from the artefacts and monitor the effectiveness of the cleaning procedures by studying the composition of the surface patina developed after the conservation treatment.
During the excavation, conservation and investigation phases, the changes in the physicochemical parameters surrounding the object can induce modification in the stratigraphy and in the chemical nature of the surface corrosion layers. In some case, the corrosion products can modify the natural patina and lead to the development of new active corrosion during desalination; therefore, this is of paramount importance the monitoring of the artefacts under study.
By following the proposed analytical protocol and combining the data collected using different techniques, significant results were gathered regarding the manufacturing, composition and state of preservation of the copper alloy artefacts recovered from the Rooswijk shipwreck.

2.2. Experimental Conditions

The techniques used for the investigation of the case-studies presented in this study are as follows.
X-ray radiography took place using a Comet MXR320/23 X-ray machine (Comet Group, Flamatt, Switzerland) and Computed Radiography (CR, Comet Group, Flamatt, Switzerland) resulting in a digital archive. A Kodak Industrex HPX-1 Plus scanner, Kodak Industrex XL Blue Digital Imaging Plates with Copper (DIP) and Carestream Industrex Digital Viewing Software Version 5.4 (Carestream Health, Rochester, MN, USA) were used. Differently sized imaging plates were used and scanned at 25–35 μm resolution. Quality was ensured using an Image Quality Indicator (IQI): Duplex wire type EN462-5.
Portable XRF measurements were carried out using a Niton XL3t XRF (Thermo Fisher Scientific, Waltham, MA, USA), following Cu/Zn mining mode and a count time of 60 s. The results are averages of at least the three analyses and are normalised.
To study the bulk composition of the artefacts, a small area from their surface was gently scratched with a scalpel to remove the corrosion layers and expose the bare metal, which was analysed by μXRF, using a Bruker M4 Tornado μ-XRF spectrometer (Bruker, Billerica, MA, USA). The data were collected at 50 kV and 400 µA with a vacuum. The tabulated results are averages of at least the three analyses and are normalised. The distribution of major and trace elements in small artefacts was mapped by μXRF. The measurements were made under vacuum at 50 kV and 400µA and elemental mapping was measured with a step size of 50 µm. Following data acquisition, X-ray peaks were manually checked prior to deconvolution and element distribution maps were obtained, showing normalized count rates.
Some fragments were collected and embedded in epoxy resin, ground and polished to a 0.25 µm finish. The cross-sections were observed by stereomicroscopy (Leica M420 stereomicroscope, Leica Microsystems, Milton Keynes, UK) and analysed by SEM, using an FEI-Inspect F (FEI, Hillsboro, OR, USA) combined with an EDS INCA X-Act (Oxford Instruments, High Wycombe, UK). The samples were coated with 15 nm of carbon and the images were collected using a back-scattered electron (BSE,) detector. The EDS data were collected at 25 KeV and quantified using Oxford Instruments INCA software.
To characterise the patinas of the artefacts, samples were collected from the surface and analysed by XRD by using an X-Ray Diffractometer D8 Advance (Bruker, Billerica, MA, USA), operating at 40 mA and 40 kV, and by FTIR spectroscopy, using a Spectrum 100 spectrometer (Perkin Elmer, Waltham, MA, USA) equipped with a DTGS detector, fitted with Attenuated Total Reflection (ATR) diamond-ZnSe crystal accessory. Spectra were recorded over the range 650–4000 cm−1, with a resolution of 4.00 cm−1, and were averaged over 32 accumulations.

3. Results

3.1. The State of Preservation and Conservation of the Copper Alloy Artefacts

The Rooswijk wreck lies on the Goodwin Sands (fine sand sediment resting on an Upper Chalk platform), covering an area of about 800 m2 with stratified deposits surviving to a depth of at least 1.5 m [20]. The multi-beam echo sounder (MBES) survey, carried out in 2018, has pointed out a significant change to the overall character of the site, due to the movement of the sediment around the wreck, periodically covering and uncovering the wreck remains [21]. Burial conditions fluctuate between near-anaerobic conditions when the site is covered by sand, and exposed conditions when the sandbanks move away. Noticeable bed level loss on and around the wreck mound has exposed many more archaeological features, which have become vulnerable to seabed erosion, mechanical degradation and biological decay [21].
The majority of copper alloy artefacts from the Rooswijk are stable, with a substantial copper core and a thin overlying mineral patina from corrosion within the burial environment. On occasion accretions consisting of seabed sediment, stones and shells cover the surface of cupreous artefacts. Some finds are deformed by the compressive forces of burial or are brittle due to the corrosion of a specific alloy composition. The aim of conservation is to ensure the long-term stability of the copper alloys by removing chlorides from within the mineral layers. Chlorides can be present as residual ions or as insoluble cupreous chloride minerals. If the artefact is left untreated and dried, these minerals can hydrolyse in the presence of oxygen and water in the atmosphere to produce hydrochloric acid. The acid will attack the metal surface, releasing copper ions, which bond with chloride to create further copper chlorides, which catalyse the corrosion process [22]. Post-excavation decay can be considerably reduced by desalinating the artefact. This can be achieved by frequent water changes with either distilled or tap water or treatment in sodium sesquicarbonate. When using distilled or tap water, the desalination progress is monitored by conductivity readings.
The choice of solution is based on a combination of observations, X-ray interpretation and the composition of the metals. Copper alloyed with zinc or lead will corrode in distilled water, most notably due to the formation of white deposits [23]. Analysis of the composition as part of the treatment protocol is, therefore, invaluable. A 1% (w/v) sodium sesquicarbonate solution in distilled water (pH 10) inhibits the corrosion of copper alloys containing zinc or lead. The use of sodium sesquicarbonate neutralises any acidic corrosion products within the metal surface, while also converting copper chlorides to harmless copper oxides. The process releases chloride ions into solution to form sodium chloride, which can be washed away with successive baths. The process is monitored with chloride test strips. The artefact is rinsed in tap water to remove any remaining solution.
There are some drawbacks to the process. The conversion of copper chlorides to copper oxide can alter the patina of the metal; this includes the production of blue-green malachite Cu2CO3(OH)2 or crystalline deposits of chalconatronite Na2Cu(CO3)2 [22]. The volume ratio for efficient desalination is of 1:5, for every kg of an artefact, 5 litres of solution is required. With frequent solution changes, this can be a resource-heavy process when treating large artefacts. There are also environmental considerations in terms of the disposal of the solution. The process is slow, which can be problematic for the timeframe of a project. On average, desalination using sodium sesquicarbonate can take up to 6 months.
It is important to inspect the artefact prior to placement in sodium sesquicarbonate for any signs that it may be a composite object; this can include checking for repairs with other alloys or organic components, which can be adversely affected by alkaline solutions. With any solution, artefacts are monitored carefully to ensure the artefact is in the correct desalination procedure.
Following desalination, artefacts are either air- or solvent-dried and stored in a desiccated environment.

3.2. Application of the Protocol in Case-Studies

3.2.1. Visual and X-ray Radiography Investigations

The analytical results collected from a selection of copper alloy objects recovered from the Rooswijk are used as examples of the application of the multi-analytical protocol to address questions about their manufacture and state of preservation and gather information about the effectiveness of the conservation treatments.
Visual and X-ray radiography investigations were able to identify features, tool marks and defects in the material, providing information on the condition of the artefacts. X-ray radiographs were used as reference tool to lead the conservation intervention and micro-excavate the small objects from the concretions (Figure 2). The abundancy of the objects in the concretion and their composition (density) guided the conservators in the selection of the right tool (air-scriber or ultrasonic scaler) and the best operative condition and working pressures for mechanical cleaning.

3.2.2. Evaluation of the Elemental Composition by pXRF, µXRF and SEM-EDS

Prior to desalination, analysis with pXRF was carried out on the artefacts as a first screening to group them based on their alloy composition [16]. The identification of the elemental composition of the objects and the presence of any repairs is important to choose the most appropriate type of treatment for chloride removal, since copper alloys with lead and zinc corrode at a higher rate when soaked in deionised water than in a sodium sesquicarbonate solution [23].
Visual examination of parts of a trumpet (rim, bell and pipe) (Figure S1) from the Rooswijk indicates that they are made of copper alloy sheets. This was further confirmed by optical and electron microscopy investigation of a sample collected from the bell (discussed later). Point analyses by pXRF mainly show copper (about 77 g/100 g) and zinc (about 22 g/100 g), which are related to a copper alloy with zinc (brass), with traces of lead, iron, nickel and arsenic (Table S1). The beak of the trumpet is richer in zinc (about 30 g/100 g) and lead (about 3 g/100 g) (leaded brass) (Table S1). Since the 16th century brass has been used for the manufacture of wind instruments due to its malleability, durability and acoustic properties [24]. The mechanical and physical properties are influenced by the amount of zinc and lead in the alloy. The brass used in early instruments has inhomogeneous composition and less zinc and more lead, while, starting from the 19th century, the zinc content is more than 30 g/100 g [24,25]. Lead can make the alloy less durable and more brittle, resulting in cracks in the alloy. The results indicate that the trumpet was manufactured using high-quality brass sheets for the main parts (rim, bell and pipe), while the beak was probably casted. The compositional data were used by the conservators to select the desalination treatments.
The following case study showcases the issues that can be encountered when X-ray radiography and compositional investigations are not carried out prior to desalinating an object. A copper alloy teapot from the Rooswijk was desalinated in distilled water, but, soon, white corrosion products started to grow on the repairs of the body, which were not previously identified because the object was not analysed by X-ray radiography. Therefore, different areas of the teapot were investigated (body, handle, flange, rivet, spout), together with the surface repairs of the body and the internal solder used to attach the spout to the body (Figure 3). Where possible, small, selected areas were cleaned to expose unaltered metal prior to the analysis. The body, flange, rivet and spout are leaded copper with a composition of about 96 g/100 g copper, 1.9 g/100 g lead and traces of tin, arsenic, antimony, nickel and iron, while the handle is mainly copper (Figure 3). The internal solder that links the spout to the body is a tin–lead alloy with a high tin content (about 58 g/100 g), lead (about 30 g/100 g), iron (6 g/100 g) and copper (4 g/100 g) and traces of zinc (1.6 g/100 g). Cracks visible on the surface of the body were repaired using a solder based on high tin (43 g/100 g), lead (31 g/100 g) and copper (26 g/100 g) (Figure 3). Lead tends to decay in highly alkaline environments, resulting in the formation of a white patina. For this reason, the identification of lead and iron solders and repairs in the teapot resulted in the use of a sodium sesquicarbonate water-based solution, instead of tap water, during the desalination process.
Most of the copper alloy artefacts are characterised by surface corrosion layers, which cannot be easily removed to reveal the bare metal. The bulk composition and trace elements of these objects were investigated by µXRF after scratching a small area of the surface (about 0.5 mm2) to expose the bare metal. For this reason, several objects from the same class (beads, rings, wires, candle holders, candle snuffers, and thimbles), that could fit inside the machine, were analysed by µXRF to obtain data about specific manufacturing technology. Except for the beads and wires, the artefacts are made of leaded brass with a composition of about 70 g/100 g copper, 16–26 g/100 g zinc, 2–7 g/100 g lead and they were produced using the traditional cementation method (Table 1) [26,27]. The beads are made of leaded copper with a high content of copper (85 g/100 g) and lead (13 g/100 g), while the wires contain about 94 g/100 g copper and 5 g/100 g iron, the latter probably accumulated on the surface as a result of being embedded in a concretion (Table 1). These leaded copper alloy artefacts were most probably manufactured by casting. For casting objects, copper and copper alloys are often prepared adding lead contents of 5 g/100 g or more. The addition of lead reduces the viscosity of the melt, improving the ability to fill the mould, and making the alloys of copper easier to cast [28]. Unlike zinc and tin, lead does not alloy with copper as the alloy cools down, and the lead forms a series of globules. This makes the leaded copper alloy more prone to crack when hammered; for this reason, wires almost never contain more than 1 g/100 g lead [29].
To evaluate the distribution of the elements along the surface of the artefacts and characterise the surface patina, µXRF mapping was carried out. Copper and zinc are homogenously distributed along the rings (Figure S2), while areas with relatively high iron correspond to dark brown patina and silicon content is associated with seabed sediment (Figure 4).
Similar to pXRF, these results guided the next conservation steps, with the selection of the most appropriate desalination and cleaning treatments.
When sampling is a possible option, polished cross-sections can be prepared. SEM-EDS is employed for morphological and compositional studies of the alloys and the corrosion layers when high-resolution information is required. The cross-section of the bell of a trumpet (Figure S1) was analysed by SEM-EDS. The sample shows a complex stratigraphy, in which corrosion layers and marine sediment are found between brass sheets, which have a thickness varying between 40 and 120 µm (Figure 5). SEM-EDS analyses (Figure S3) show that the bulk of the corrosion layers (dark grey layers) is rich in Cu and S, while Zn accumulates on their external surfaces as a result of the dissolution of the brass sheets and leaching of metal ions [30]. These corrosion layers are embedded in a matrix of calcium, silicon and iron from seabed sediment (Figure 5 and Figure S3).
The cross-section of a sample taken from the rim of a copper (with traces of ZnO and PbO) lid after conservation was investigated by SEM-EDS, to assess the efficacy and selectivity of the desalination and cleaning intervention. Elemental maps proved that salts, such as chlorides, were completely removed from the surface of the object and that the treatments lead to the formation of a stable and compact patina, with a thickness of about 10 µm and with a composition of 99 g/100 g CuO (Figure 6) [22,31].
Elemental maps obtained by SEM-EDS enable the study of samples with a complex stratigraphy, such as copper wires. The analysis of cross-sections of the wires indicates that they have a copper core, which is characterised by irregular borders due to the leaching of copper ions from the surface, and several corrosion layers deposited on top (Figure 7). The bulky deposition layer is mainly composed of iron and a thin surface layer made of silver sulphide was also detected above this. These crystals were probably formed due to the reaction of metal silver with sulphide ions, suggesting that the wires are probably made of silvered copper. Similar results were obtained from the analyses of the embroidered metalwork on a 17th century toilet service recovered from the Burgzand North 17 shipwreck [32]. The threads have a similar composition to 17th–19th South American metal threads found on colonial ecclesiastic objects [33]. The exact function of the copper wires from the Rooswijk is still unclear, but the analytical results suggest that they could be metal threads used for decorative purposes, perhaps as part of the cargo. It is important to highlight that, due to the difficulties in the removal of the corrosion products from the surface of the wires during the cleaning operation, this information would have stayed undisclosed without sampling.

3.2.3. Characterisation of Corrosion Layers and Surface Patina by FTIR Spectroscopy and XRD

The examination of the corrosion layers on the surface of the copper alloy artefacts was carried out by ATR-FTIR spectroscopy and XRD. The analyses were performed on the powder collected from the surface patina before and after the desalination and cleaning of the artefacts to monitor the effectiveness of conservation treatments. Prior to conservation, some leaded brass thimbles were analysed, and several chlorine-based compounds (Cu2(OH)3Cl, Cu3Zn(OH)6Cl) were detected (Figure S4). Desalination of the objects in sodium sesquicarbonate solution was used to promote their hydrolysis and to release chlorine ions. The XRD spectrum obtained from the dark brown patina found after conservation on the thimble exhibits diffraction peaks that can be related to the presence of cuprite (Cu2O) and a small amount of quartz (SiO2) from the seabed sediment (Figure 8a) [22,31].
The FTIR spectrum obtained from the green patina on the thimble shows peaks at about 3400 and 3300 cm−1 and 1040 cm−1 related to hydroxyl-stretching and bending bands, at 1490 and 1380 cm−1 due to C-O antisymmetric stretching modes, at 1090 cm−1 assigned to the C-O symmetric stretching vibrations, and at about 820 and 747 cm−1 due to out of phase and in phase bending modes of a carbonate [34,35] (Figure 8b). The spectrum shows significant similarities with the reference spectrum of malachite (Cu2(CO3)(OH)2), proving the formation of this basic carbonate on the brass. Chloride-bearing compounds such as nantokite, clinoatacamite and atacamite were not detected. These results prove the efficacy of the desalination treatment in the removal of the salts from the thimbles and in promoting the formation of stable patinas based on cuprite and malachite, again demonstrating the important role of materials investigation before and after conservation.

4. Conclusions

The scientific analysis of copper alloy artefacts from underwater excavations is used to determine their composition, manufacture, and assess their condition. The technical examination of the artefacts before, during and after conservation treatments is crucial in the design and selection of the most appropriate conservation treatments and to monitor their effectiveness. The multi-analytical protocol proposed in this paper provides a guide to the information that can be gathered by each analytical step to answer the questions posed by archaeologists, conservators and scientists, with examples from copper alloy artefacts recovered from the Rooswijk shipwreck. Conservation problems were explored when there were deviations from the outlined protocol. The obtained results following the proposed analytical protocol provide essential information to guide and support conservators in the conservation decision-making process.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/coatings12111640/s1, Figure S1: Brass trumpet from the Rooswijk shipwreck: (a) rim and bell; (b) pipe; and (c) beak.; Figure S2: Copper alloy rings from the Rooswijk shipwreck (top left) and µXRF elemental maps showing the main elements of the alloy; Figure S3: SEM image (BSE detector) (on the left) and elemental maps of a cross-section from a sample collected from the bell of a trumpet; Figure S4: XRD spectra of green patina on a copper alloy thimble from the Rooswijk before desalination, with diffraction peaks indicating the presence of several corrosion products. Table S1: Chemical composition (pXRF, g/100 g, normalised) of different parts of a trumpet (bd = below detection).

Author Contributions

Conceptualization, F.G.; methodology, F.G. and H.S.; Investigation, F.G. and H.S.; data curation, F.G.; writing, F.G. and H.S.; review and editing, F.G. and H.S. All authors have read and agreed to the published version of the manuscript.

Funding

The #Rooswijk1740 project is funded and led by the Cultural Heritage Agency of the Netherlands (Ministry of Education, Science and Culture), in collaboration with project partner Historic England and UK-contractor MSDS Marine.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All data that support the findings of this study are included within the article.

Acknowledgments

We are grateful to the work previous and current project staff have undertaken: Alex Bliss, Alison James, Elizabeth Kuiper, Martijn Manders, Angela Middleton, Sarah Paynter, Kim Roche and Nicole Schout.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Rooswijk1740: Project Overview. Available online: https://english.cultureelerfgoed.nl/topics/maritime-heritage/rooswijk1740 (accessed on 10 October 2022).
  2. Rooswijk1740: The Post Excavation Phase. Available online: https://historicengland.org.uk/whats-new/research/rooswijk-shipwreck-excavation-the-post-excavation-phase/ (accessed on 10 October 2022).
  3. Rooswijk1740: Virtual Dive Trail. Available online: https://www.cloudtour.tv/Rooswijk (accessed on 10 October 2022).
  4. Robbiola, L.; Portier, R. A global approach to the authentication of ancient bronzes based on the characterization of the alloy–patina–environment system. J. Cult. Herit. 2006, 7, 1–12. [Google Scholar] [CrossRef]
  5. Aucouturier, M.; Darque-Ceretti, E. The surface of cultural heritage artefacts: Physicochemical investigations for their knowledge and their conservation. Chem. Soc. Rev. 2007, 36, 1605–1621. [Google Scholar] [CrossRef] [PubMed]
  6. Neff, D.; Reguer, S.; Dillmann, P. 4—Analytical techniques for the study of corrosion of metallic heritage artefacts: From micrometer to nanometer scales. In Corrosion and Conservation of Cultural Heritage Metallic Artefacts; Dillmann, P., Watkinson, D., Angelini, E., Adriaens, A., Eds.; Woodhead Publishing: Sawston, UK, 2013; pp. 55–81. [Google Scholar]
  7. Ingo, G.M.; Riccucci, C.; Faraldi, F.; Casaletto, M.P.; Guida, G. Micro-chemical and micro-structural investigation of the corrosion products on “The Dancing Satyr” (Mazara del Vallo, Sicily, Italy). Appl. Phys. A 2010, 100, 785–792. [Google Scholar] [CrossRef]
  8. Angelini, E.; Grassini, S.; Tusa, S. 12—Underwater corrosion of metallic heritage artefacts. In Corrosion and Conservation of Cultural Heritage Metallic Artefacts; Dillmann, P., Watkinson, D., Angelini, E., Adriaens, A., Eds.; Woodhead Publishing: Sawston, UK, 2013; pp. 236–259. [Google Scholar]
  9. Ashkenazi, D.; Cvikel, D.; Stern, A.; Klein, S.; Kahanov, Y. Metallurgical characterization of brass objects from the Akko 1 shipwreck, Israel. Mater. Charact. 2014, 92, 49–63. [Google Scholar] [CrossRef]
  10. Ciarlo, N.C.; Maxia, G.; Rañi, M.; De Rosa, H.; Geli Mauri, R.; Vivar Lombarte, G. Craft production of large quantities of metal artifacts at the beginnings of industrialization: Application of SEM–EDS and multivariate analysis on sheathing tacks from a British transport sunk in 1813. J. Archaeol. Sci. Rep. 2016, 5, 263–275. [Google Scholar] [CrossRef]
  11. Żmuda-Trzebiatowska, I.; Schaefer, K.; Sokołowska, A.; Rodzik, I.; Sobczyk, A.T.; Karczewski, J.; Śliwiński, G. Raman investigation of the patina layers on Hungarian copper ingots from a fifteenth century shipwreck. J. Raman Spectrosc. 2016, 47, 1528–1533. [Google Scholar] [CrossRef]
  12. Inberg, A.; Ashkenazi, D.; Cohen, M.; Iddan, N.; Cvikel, D. Corrosion products and microstructure of copper alloy coins from the Byzantine-period Ma’agan Mikhael B shipwreck, Israel. Microchem. J. 2018, 143, 400–409. [Google Scholar] [CrossRef]
  13. Ingo, G.M.; Riccucci, C.; Guida, G.; Pascucci, M.; Giuliani, C.; Messina, E.; Fierro, G.; Di Carlo, G. Micro-chemical investigation of corrosion products naturally grown on archaeological Cu-based artefacts retrieved from the Mediterranean sea. Appl. Surf. Sci. 2019, 470, 695–706. [Google Scholar] [CrossRef]
  14. van der Stok-Nienhuis, J.; Kuiper, E.; Beentjes, T.; Joosten, I.; van Eijck, L.; Zhou, Z.; van Bommel, M. A case study for scientific research prior to conservation of marine metal artefacts. J. Archaeol. Sci. Rep. 2021, 37, 102909. [Google Scholar] [CrossRef]
  15. Armetta, F.; Saladino, M.L.; Scherillo, A.; Caponetti, E. Microstructure and phase composition of bronze Montefortino helmets discovered Mediterranean seabed to explain an unusual corrosion. Sci. Rep. 2021, 11, 23022. [Google Scholar] [CrossRef] [PubMed]
  16. Analytical Methods Committee, A.N. Hand-held X-ray fluorescence analysis of archaeological artefacts: Challenges, advantages and limitations. Anal. Methods 2021, 13, 3731–3734. [Google Scholar] [CrossRef]
  17. Colomban, P.; Gironda, M.; Simsek Franci, G.; d’Abrigeon, P. Distinguishing Genuine Imperial Qing Dynasty Porcelain from Ancient Replicas by On-Site Non-Invasive XRF and Raman Spectroscopy. Materials 2022, 15, 5747. [Google Scholar] [CrossRef] [PubMed]
  18. Derrick, M.; Stulik, D.; Landry, J. Infrared Spectroscopy in Conservation Science; Getty Publications: Los Angeles, CA, USA, 1999. [Google Scholar]
  19. Kendix, E.; Moscardi, G.; Mazzeo, R.; Baraldi, P.; Prati, S.; Joseph, E.; Capelli, S. Far infrared and Raman spectroscopy analysis of inorganic pigments. J. Raman Spectrosc. 2008, 39, 1104–1112. [Google Scholar] [CrossRef]
  20. Dunkley, M. Rooswijk Goodwin Sands, off Kent: Conservation statement & management plan. In English Heritage Report; English Heritage: Swindon, UK, 2009. [Google Scholar]
  21. Pascoe, D. Multi-Beam Surveys of the Designated Wrecks on the Goodwin Sands and the Downs, 2018 report. Hist. Engl. Res. Rep. 2020, 243, 2020. [Google Scholar]
  22. Scott, D.A.; Institute, G.C. Copper and Bronze in Art: Corrosion, Colorants, Conservation; Getty Conservation Institute: Los Angeles, CA, USA, 2002. [Google Scholar]
  23. Pearson, C. Conservation of Marine Archaeological Objects; Elsevier Science: Amsterdam, The Netherlands, 1988. [Google Scholar]
  24. Vereecke, H.; Frühmann, B.; Schreiner, M. The Chemical Composition of Brass in Nuremberg Trombones of the Sixteenth Century. Hist. Brass Soc. J. 2012, 24, 61–77. [Google Scholar] [CrossRef]
  25. Elsener, B.; Alter, M.; Lombardo, T.; Ledergerber, M.; Wörle, M.; Cocco, F.; Fantauzzi, M.; Palomba, S.; Rossi, A. A non-destructive in-situ approach to monitor corrosion inside historical brass wind instruments. Microchem. J. 2016, 124, 757–764. [Google Scholar] [CrossRef]
  26. Craddock, P.T.; Eckstein, K. Production of brass in antiquity by direct reduction. In Mining and Metal Production through the Ages; Craddock, P.T., Lang, J., Eds.; British Museum Press: London, UK, 2003. [Google Scholar]
  27. Bourgarit, D.; Bauchau, F. The ancient brass cementation processes revisited by extensive experimental simulation. JOM 2010, 62, 27–33. [Google Scholar] [CrossRef]
  28. Scott, D.A.; Schwab, R. The Metallurgy of Pre-industrial Metals and Alloys. In Metallography in Archaeology and Art; Springer International Publishing: Cham, Switzerland, 2019; pp. 133–206. [Google Scholar]
  29. Dungworth, D. Roman Copper Alloys: Analysis of Artefacts from Northern Britain. J. Archaeol. Sci. 1997, 24, 901–910. [Google Scholar] [CrossRef]
  30. Santos, C.I.S.; Mendonça, M.H.; Fonseca, I.T.E. Corrosion of brass in natural and artificial seawater. J. Appl. Electrochem. 2006, 36, 1353–1359. [Google Scholar] [CrossRef]
  31. Zhang, X.; Odnevall Wallinder, I.; Leygraf, C. Mechanistic studies of corrosion product flaking on copper and copper-based alloys in marine environments. Corros. Sci. 2014, 85, 15–25. [Google Scholar] [CrossRef] [Green Version]
  32. Savelli, S.; van Bork, E.; Meijer, S.; Joosten, I. Metal Threads in 17th Century Textiles: Technological Research of the Findings of Burgzand North 17 Shipwreck. In Proceedings of the Metal 2019 Proceedings of the Interim Meeting of the ICOM-CC Metals Working Group, Neuchâtel, Switzerland, 2–6 September 2019; pp. 41–44. [Google Scholar]
  33. Muros, V.; Wärmländer, S.K.T.S.; Scott, D.A.; Theile, J.M. Characterization of 17Th-19Th Century Metal Threads from the Colonial Andes. J. Am. Inst. Conserv. 2007, 46, 229–244. [Google Scholar] [CrossRef]
  34. Frost, R.L.; Martens, W.N.; Rintoul, L.; Mahmutagic, E.; Kloprogge, J.T. Raman spectroscopic study of azurite and malachite at 298 and 77 K. J. Raman Spectrosc. 2002, 33, 252–259. [Google Scholar] [CrossRef] [Green Version]
  35. Zaffino, C.; Guglielmi, V.; Faraone, S.; Vinaccia, A.; Bruni, S. Exploiting external reflection FTIR spectroscopy for the in-situ identification of pigments and binders in illuminated manuscripts. Brochantite and posnjakite as a case study. Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 2015, 136, 1076–1085. [Google Scholar] [CrossRef] [PubMed]
Coatings 12 01640 g001 550
Figure 1. Multi-analytical protocol in use at FCL for the examination of copper alloy artefacts from marine sites.
Figure 1. Multi-analytical protocol in use at FCL for the examination of copper alloy artefacts from marine sites.
Coatings 12 01640 g001
Coatings 12 01640 g002 550
Figure 2. One of the concretions (top) with its corresponding X-ray image (bottom), where rings, a coin, beads and a candle snuffer can be observed.
Figure 2. One of the concretions (top) with its corresponding X-ray image (bottom), where rings, a coin, beads and a candle snuffer can be observed.
Coatings 12 01640 g002
Coatings 12 01640 g003 550
Figure 3. (a,b) Copper alloy teapot from the Rooswijk shipwreck (arrows indicate, on the left, a repair in the body and, on the right, a solder in the spout); (c,d) concentration of the elements in the alloys of the body, handle, flange, rivet, spout and in repairs of the teapot.
Figure 3. (a,b) Copper alloy teapot from the Rooswijk shipwreck (arrows indicate, on the left, a repair in the body and, on the right, a solder in the spout); (c,d) concentration of the elements in the alloys of the body, handle, flange, rivet, spout and in repairs of the teapot.
Coatings 12 01640 g003
Coatings 12 01640 g004 550
Figure 4. Copper alloy rings (top) and coloured µXRF maps of the elements accumulating on the surface patina and on concretion residues (Fe and Si, respectively) (bottom).
Figure 4. Copper alloy rings (top) and coloured µXRF maps of the elements accumulating on the surface patina and on concretion residues (Fe and Si, respectively) (bottom).
Coatings 12 01640 g004
Coatings 12 01640 g005 550
Figure 5. SEM images (BSE detector) of a cross-section from a sample collected from the bell of a brass trumpet, showing the full stratigraphy (a), the corrosion layer (dark grey) developed between the brass sheets (light grey) (b) and the corroded brass sheet (c).
Figure 5. SEM images (BSE detector) of a cross-section from a sample collected from the bell of a brass trumpet, showing the full stratigraphy (a), the corrosion layer (dark grey) developed between the brass sheets (light grey) (b) and the corroded brass sheet (c).
Coatings 12 01640 g005
Coatings 12 01640 g006 550
Figure 6. SEM image (BSE detector) and elemental map of copper (Cu) of a cross-section from a sample collected from the rim of a copper lid (on the left).
Figure 6. SEM image (BSE detector) and elemental map of copper (Cu) of a cross-section from a sample collected from the rim of a copper lid (on the left).
Coatings 12 01640 g006
Coatings 12 01640 g007 550
Figure 7. Optical microscopy image of a copper wire (a), SEM images (BSE detector) (b,c) and coloured maps of the main elements (Cu, Ag and Fe) (d) of a cross-section of the copper wire.
Figure 7. Optical microscopy image of a copper wire (a), SEM images (BSE detector) (b,c) and coloured maps of the main elements (Cu, Ag and Fe) (d) of a cross-section of the copper wire.
Coatings 12 01640 g007
Coatings 12 01640 g008 550
Figure 8. (a) XRD spectrum of dark brown patina on a copper alloy thimble from the Rooswijk after conservation, with diffraction peaks indicating the presence of cuprite (C) and quartz (Q); (b) ATR-FTIR spectrum collected from the green patina from the thimble after conservation (green spectrum) and the FTIR spectrum of a reference sample of malachite (black spectrum).
Figure 8. (a) XRD spectrum of dark brown patina on a copper alloy thimble from the Rooswijk after conservation, with diffraction peaks indicating the presence of cuprite (C) and quartz (Q); (b) ATR-FTIR spectrum collected from the green patina from the thimble after conservation (green spectrum) and the FTIR spectrum of a reference sample of malachite (black spectrum).
Coatings 12 01640 g008
Table
Table 1. Average chemical composition (µXRF, g/100 g, normalised) of the copper alloy artefacts from the Rooswijk (bd = below detection). N. indicates the number of analysed artefacts per category.
Table 1. Average chemical composition (µXRF, g/100 g, normalised) of the copper alloy artefacts from the Rooswijk (bd = below detection). N. indicates the number of analysed artefacts per category.
ObjectCuZnPbFeNiAsAgSnSbMaterialN.
Beads85 ± 110.36 ± 0.2813 ± 100.58 ± 0.510.18 ± 0.070.20 ± 0.120.045 ± 0.0350.54 ± 0.180.55 ± 0.14Leaded copper30
Rings72 ± 322 ± 34.0 ± 1.20.67 ± 0.280.12 ± 0.030.27 ± 0.150.049 ± 0.0440.35 ± 0.420.23 ± 0.21Leaded brass41
Wires94 ± 80.065 ± 0.0530.017 ± 0.0075.7 ± 7.70.058 ± 0.0130.080 ± 0.110.34 ± 0.540.067 ± 0.0640.065 ± 0.072Copper3
Candle holders71 ± 416 ± 47.6 ± 3.60.77 ± 0.410.37 ± 0.070.49 ± 0.120.062 ± 0.0273.0 ± 0.80.77 ± 0.54Leaded brass10
Candle snufffers73 ± 524 ± 52.2 ± 0.60.56 ± 0.210.16 ± 0.100.22 ± 0.080.12 ± 0.200.043 ± 0.0550.015 ± 0.016Leaded brass7
Thimbles 73 ± 623 ± 52.9 ± 0.41.3 ± 2.00.057 ± 0.0150.12 ± 0.040.046 ± 0.0250.11 ± 0.15bdLeaded brass26
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Gherardi, F.; Stewart, H. A Multi-Analytical Protocol for Decision Making to Study Copper Alloy Artefacts from Underwater Excavations and Plan Their Conservation. Coatings 2022, 12, 1640. https://doi.org/10.3390/coatings12111640

AMA Style

Gherardi F, Stewart H. A Multi-Analytical Protocol for Decision Making to Study Copper Alloy Artefacts from Underwater Excavations and Plan Their Conservation. Coatings. 2022; 12(11):1640. https://doi.org/10.3390/coatings12111640

Chicago/Turabian Style

Gherardi, Francesca, and Heather Stewart. 2022. "A Multi-Analytical Protocol for Decision Making to Study Copper Alloy Artefacts from Underwater Excavations and Plan Their Conservation" Coatings 12, no. 11: 1640. https://doi.org/10.3390/coatings12111640

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop