Copper (Cu) and cuprous chloride (CuCl) powders were used to establish the critical RH value that CuCl transforms into copper trihydroxychlorides, the corrosion products of the so-called “Bronze Disease”. XRD analysis of the tested samples showed that the rate of transformation is fast above the deliquescence point of CuCl (68.4% RH at 19.4º C) but very slow below it. The critical RH value for CuCl transformation was found to be at 63% RH. However, subtle variables such as air movement, composition of samples and type of substrate may result in the depression of the RH value that this transformation occurs. Nevertheless, the results of this study suggest that copper artefacts would be safe from the occurrence of “Bronze Disease” in the ambient museum environment (45-60% RH), provided that the upper limit is not exceeded. More experimental data are needed to clarify the above suggestions.
Corrosion of metallic artefacts in the museum environment
Archaeological metallic artefacts are susceptible to accelerated corrosion reactions once they are excavated and exposed to adverse environmental conditions. High Relative Humidity (RH) levels in a museum environment combined with high pollutant concentrations increase the corrosion rate of metals . Archaeological metals can also become heavily contaminated with salts from the burial environment and the most well known examples are chloride contamination of iron and copper (see equation 1). Iron artefacts contaminated with ferrous chloride may be subject to physical and chemical damage from cyclic corrosion reactions  and a similar phenomenon is noted from the oxidation and hydrolysis of cuprous chloride (CuCl) contaminated copper artefacts .
The term “bronze disease” is used in conservation literature to describe the oxidation and hydrolysis of CuCl into copper trihydroxychlorides:
Cu + Cl- = CuCl + e- (1)
4CuCl(s) + O2(g)+ 4H2O = 2Cu2(OH)3Cl(s) + 2H+(aq) + 2Cl-(aq) (2)
It is generally accepted that formation of HCl is an important parameter of the corrosion reactions and that HCl in turn attacks copper to form more CuCl :
2HCl + 2Cu = 2CuCl + H2 (3)
Although the copper corrosion stratigraphy can be complex, there is a general pattern which applies to burial environments and is shown in figure 1 along with the relative thickness of the corrosion products. First there is the metal; then the CuCl layer, followed by the cuprous oxide (Cu2O) layer and on top cupric salts are formed .
“Bronze disease” ensues when the corrosion layers overlaying the CuCl are disturbed with oxygen and moisture oxidizing and hydrolyzing the CuCl according to equation (2). This corrosion mechanism is so quick that the CuCl layer transforms into loose and powdery copper trihydroxychlorides that cause major mechanical disruption to the stable Cu2O patina (figure 2).
Historical background of “bronze disease” – Suggestions for environmental control
Rosenberg  was the first to recognize that oxygen and moisture is required for “bronze disease” to develop. He experimented with a series of saturated salt solutions giving a range of different RHs and he found that the transformation of cuprous chloride in bronzes occurred above 71% RH. Some years later, it was proposed  that a RH in the 40-50% range should be specified by museums for safe displaying of copper artefacts, but without presenting experimental data that would justify this suggestion.
It seems that the lack of experimental evidence coupled with the daunting name of the reaction has brought confusion concerning the RH values that CuCl remains stable. This is exemplified by the different suggestions brought up by different authors. A common advice is that chloride-contaminated archaeological copper alloys should stay in a dry environment . Indeed this seems to be the conservation approach in museums; copper alloy artefacts, like iron artefacts, are being stored and displayed under desiccated conditions.
Some  have recommended that the RH should be kept in the 40-50% range, whereas others  have stated that CuCl is unstable above 40% RH. Scott , on the other hand, reported that CuCl reacted at 70% RH within a day. He also suggested that a RH between 42-46% is adequate for the storage or display of untreated artefacts. The specific suggestion was based on experimental results of compressed tablets of cuprous chloride, powdered cuprous chloride and copper powder mixtures kept in a humidity cabinet for two years inside which the RH fluctuated between 42-46%. After the end of the experiment no change was observed to the tested samples. However, Scott did not expose the samples to a range of RH’s. This was done by Tsatsouli  who found that CuCl powder does not react at 45% RH and below, but when in contact with copper powder it transforms into Paratacamite (6% of the mixture), which is a sign for “bronze disease”, at 40% RH.
Aims and objectives
Having said the above, the present study concentrates on the Cu-CuCl system to fill gaps in the current knowledge on the critical environmental conditions that lead to the manifestation of “bronze disease”. Specifically, the aims and objectives of the study are:
- To suggest the critical RH value that CuCl transforms into copper trihydroxychlorides.
- To investigate the aggressiveness of the corrosion reaction leading to “bronze disease”.
- To test the theory developed during the project that corrosion reactions leading to “bronze disease” are apprehended below the deliquescence point of CuCl.
- To use the results of this study to suggest safe RH levels for copper and copper alloy objects and thus contribute to the optimum management of the museum’s environmental and financial resources.
Materials and Methods
Powdered samples for investigating “bronze disease” corrosion reactions have been used in previous studies [3,10] and seem to produce reliable experimental results. They provide a large surface area which produces faster corrosion reactions. Indeed a few days are enough to determine if “bronze disease” has occurred, due to the rapidity and spontaneous nature of the corrosion reaction. Analar grade CuCl and Cu powders were used to make up the samples for the experiments. The tested samples were:
- CuCl in contact with Cu powder. The ratio for the CuCl/Cu mixtures was 1:1;
- CuCl on its own as a control;
- Cu on its own as a control.
The experimental tests were carried out in a Vötsch Industrietechnik VC 4018 model environmental chamber. The test space has humidity and temperature sensors which are controlled via S!MPATI (Simulation Package for Test System Integration) computer software. Deionized water was used for the operation of the climatic chamber.
For meeting the aims and objectives set out for this project the following experiments were carried out (table 1). A flexible approach was required for identifying the critical RH value that CuCl transforms into copper trihydroxychlorides. Therefore, experimental conditions for each set of experiments depended on the analytical results obtained from the previous one. This order also reflects the unpredictability of the project. The amount of variables was controlled by keeping the temperature stable at 20 ºC.
The above experiments were carried out to suggest the critical RH value of CuCl transformation. In addition, experiment 4 was carried out to investigate the aggressiveness of the corrosion reactions leading to “bronze disease”, as the samples were taken out of the environmental chamber at specified time intervals.
X-ray diffraction analysis
The specimens were analyzed by XRD on an XPERT-PRO diffractometer system. The system’s scan step size was set at 0.0170 [º2Theta], the scan step counting time at 21.3216 [sec] and the scan range at 5.0084-74.9634 [º2Theta]. The X-ray sources used were Cu Ka and Cu Kß and spectra were acquired at 30 mA and 40 kV. The mineral phase identification and semi-quantitative analysis was performed by X’Pert HighScore software. The computer software searched the reference database, with over 150,000 XRD patterns, for matches to the specimen’s spectrum. It suggested the mineral phases that may be present based on the existence and position of peaks as well as the quality of the peak intensities of the specimen.
The obtained powder diffraction peak lists were always double-checked by visual inspection against a number of mineral phases that were expected to occur in the specimens. The d-spacing and relative intensity of peaks were used as identification criteria and the powder diffraction files used for the visual checking belonged to the following crystalline phases:
- Cuprous chloride (CuCl)
- Copper (Cu)
- Atacamite – Botallackite – Clinoatacamite – Paratacamite [Cu2(OH)3Cl]
- Cupric hydroxychloride [Cu(OH)Cl]
- Eriochalcite [CuCl2.2(H2O)]
- Cupric chloride (CuCl2)
- Cuprite (Cu2O)
Evaluation of the results of the experimental procedures was based on a combination of the visual inspection of the colour of the samples and the XRD analysis performed on them. Reference diffraction patterns of CuCl, Cu and Cu+CuCl from the bottle were obtained to act as a standard and for comparison with the tested samples.
XRD results on the Cu+CuCl and CuCl samples exposed at 38% for 10 days, at 42% for 12.7 days and at 62% RH for 5.8 days showed no measurable change in their mineral composition, and therefore no signs of “Bronze Disease”, when compared with the reference samples from the bottle.
Samples at 70% RH and 20ºC
Eleven Cu+CuCl samples, 4 CuCl samples and 1 Cu sample were prepared and apart from the Cu sample the rest were taken out from the environmental chamber at specified time intervals to evaluate the rate of transformation of CuCl into copper trihydro-xychlorides. CuCl samples developed a dark green colour and had the consistency of thick slurry so they had to be dried in desiccated storage before XRD analysis. The Cu+CuCl samples were covered with the characteristic pale green Paratacamite powder, which is a sign for “Bronze Disease”, less in the beginning (e.g. 3 days) and more by the end of the experimental period (9 days).
XRD analysis results for the Cu+CuCl samples are displayed in Table 2. The XRD spectra from three different time periods of the experiment are compared in Figure 3. For the Cu+CuCl samples no change was observed during the first six hours, but at 9 hours Cupric Hydroxide Chloride (CuOHCl), which is a metastable corrosion product, and Paratacamite started forming. This mineral formation was slow for the first two days and by the beginning of the 2nd day CuOHCl disappeared. From the 3rd day onwards the concentration of Paratacamite increased at a faster rate, to reach 22% on the 9th day of the experiment.
XRD analysis of the CuCl samples showed a different picture (table 3) from that of the Cu+CuCl samples. Although after 1 day CuCl concentration was very high (86%), after four days there was no CuCl left in the sample and instead CuOHCl was the major mineral phase (66%) followed by Paratacamite (19%) and Eriochalcite (15%). The mineral composition of the samples stayed more or less the same after 12 days. The Cu sample remained unchanged.
Samples at 65% RH and 20ºC
Two Cu+CuCl samples, one CuCl sample and one Cu sample were prepared for this experiment and left in the environmental chamber for 7.9 days. The CuCl sample developed a dark green colour, whereas the Cu+CuCl samples had a few spots of light green Paratacamite powder on their surface. The XRD analysis results are shown in Table 4 and the XRD spectra of the tested Cu+CuCl and of the Cu+CuCl from the bottle are compared in Figure 4. The Cu+CuCl sample contained 2% Paratacamite. The CuCl sample had 1% Paratacamite and 3% CuOHCl. The Cu sample remained unchanged.
Samples at 63% RH and 20ºC
Two Cu+CuCl samples, one CuCl sample and one Cu sample were prepared for this experiment and left in the environmental chamber for 6.5 days. The CuCl sample showed no colour alteration, whereas the Cu+CuCl samples had very few tiny spots of light green Paratacamite powder on their surface. The XRD analysis results are shown in Table 5 and the XRD spectra of the tested Cu+CuCl and of the Cu+CuCl from the bottle are compared in Figure 5. The Cu+CuCl sample contained 1% Paratacamite whereas the CuCl sample had 1% CuOHCl. The Cu sample remained unchanged.
Figure 3. Comparison of the XRD spectra between 3 Cu+CuCl specimens exposed for 3 hours, 1 day and 9 days at 70% RH. The major peaks for Paratacamite are noted at 16.254, 32.437 and 39.835 2θ(º).Discussion
Figure 4. Comparison of the XRD spectra of Cu+CuCl (65% RH, 7.9 days) and Cu+CuCl (bottle). The major peaks for Paratacamite are noted at 16.254, 32.437 and 39.835 2θ(º).
Figure 5. Comparison of the XRD spectra of Cu+CuCl (63% RH, 7.9 days) and Cu+CuCl (bottle). The slight shift in the peak positions denotes the presence of Paratacamite. The major peaks for Paratacamite are noted at 16.254, 32.437 and 39.835 2θ(º).
Figure 6. The almost identical XRD spectra of Cu+CuCl exposed at 65% RH for 7.9 days and Cu+CuCl exposed at 70% RH for 2 days. The major peaks for Paratacamite are noted at 16.254, 32.437 and 39.835 2θ(º).
The critical RH value for CuCl transformation to copper trihydroxychlorides
XRD results on the Cu+CuCl and CuCl samples exposed at 38% for 10 days, at 42% for 12.7 days and at 62% RH for 5.8 days showed no measurable change in their mineral composition when compared with the reference samples from the bottle. Similarly, the copper metal on its own did not change during the tested RH’s probably because rapid copper corrosion occurs only at around 95-98% RH . However, the Cu+CuCl exposed at 63% RH for 6.5 days showed the appearance of Paratacamite at 1% concentration whereas the CuCl sample had 1% CuOHCl.
Furthermore, the tested samples exposed at higher RH’s showed that the rate of transformation of CuCl into the copper trihydroxychlorides is RH-dependent. Figure 6 clearly demonstrates that samples left for short periods at high RH’s had almost identical diffractograms with samples exposed to lower RH’s for longer periods.
The deliquescence point of CuCl
The XRD results of the tested samples at 63% and 65% RH show clearly that the manifestation of “bronze disease” is much slower compared to that at 70% RH. The increased transformation rate of CuCl at 70% RH is probably due to the deliquescence point of CuCl which occurs at 68.4% RH when exposed to 19.4ºC . Above the deliquescence point of CuCl, at 70% RH, liquid water is absorbed into the crystal structure of CuCl and because water is polar it gets attracted to the electrostatic charge of the ions derived from the salt and therefore the escape of water molecules is hindered .
Therefore, when the deliquescence point of CuCl is reached, the water is incorporated into the crystal structure of the mineral which results in the quick formation of “bronze disease”. At 70% RH, water and oxygen are freely available to be consumed via the oxidation and hydrolysis of CuCl:
4CuCl + O2 + 4H2O = 2Cu2(OH)3Cl + 2H+ + 2Cl- (4)
The presence of HCl in the CuCl samples as a result of equation (4) was verified by the measurement of their pH which was found to be as low as 2.5.
Why “bronze disease” occurs at 63% RH? Not as aggressive as is currently thought?
If water is freely available for corrosion reactions only above the deliquescence point of CuCl then why Paratacamite occurs at 65% and 63% RH? Since water is required for the occurrence of “bronze disease”, it means that CuCl incorporates water into its crystal structure before the deliquescence point is reached. It seems that there are other factors, apart from RH and temperature, affecting the kinetics of water absorption in a salt such as the composition and type of substrate, and air movement .
Overall, the results from this study show that whereas the rate of CuCl transformation into copper tri-hydroxychlorides is slow at 65% and 63% RH, it dramatically increases above its deliquescence point. In addition, the present study suggests that the copper and copper alloy artefacts should remain stable at a RH below 60% and that the current belief on the critical RH value of “bronze disease” occurrence is overstated and exaggerated. However, this recommendation needs clarification with longer experimental periods.
From the early part of the 20th century “bronze disease” caused great concern among museum curators . Since then, several authors have recommended that safe RH values for copper and copper alloy artefacts should be below 40% RH. These suggestions, apart from two exceptions [3,10], are not backed up by scientific evidence.
In this study, it was found that corrosion reactions leading to “bronze disease” are very fast above the deliquescence point of CuCl, whereas they are much slower below it. Furthermore, a Cu+CuCl sample after 6.5 days exposure at 63% RH formed Paratacamite but no change could be identified in Cu+CuCl samples after 5.8 days exposure at 62% RH. This finding does not necessarily mean that 63% RH is the critical point of CuCl transformation.
In fact there may be no critical RH for CuCl transformation because the chemisorption of water into the crystal structure of CuCl probably depends on many subtle variables such as sample composition and air movement. However, the results of the present study and taking into account the fact that the experimental conditions were very aggressive and would not occur in real conditions, suggest that the copper and copper alloy artefacts would remain stable at a RH below 60%. This means that chloride-contaminated copper alloy artefacts would not require any special attention and would probably be safe in ambient museum environmental conditions (45-60%) as long as the upper limit is not exceeded. Longer experimental periods and variable experimental conditions, especially in the 50-60 RH range, are required in order to test this suggestion.
The present study does not aspire to provide definitive answers on the optimum storage of copper and its alloys. The word optimum also refers to energy efficiency which is very important nowadays. There has to be adequate justification for the specification of environmental conditions for cultural objects, especially when air-handling units (e.g. air-conditioners) with high operating costs are employed. These costs will be higher if desiccated conditions are specified.
Therefore, more scientific data are required to elucidate our understanding of the mechanism of CuCl transformation into copper trihydroxychlorides so as to avoid basing environmental specifications on largely empirical observations. Based on the results of the present study it is suggested that museums become bolder in their environmental approach of storing and displaying copper alloy artefacts and quit the “better safe than sorry” policy if they ever hope to be cost-effective.
The following people are thanked for their valuable help in completing this article: David Watkinson, my supervisor, for his continuous support and useful discussions and comments that much improved the scientific quality of the project; Mark Lewis of the Cardiff University conservation department for demonstrating the operation of the climatic chamber; Louise Joyner and Ian Freestone of the conservation department for their kindness on showing me how to operate the XRD and for the useful discussions; Michael Lambert, Tom Cottrell and Amanda Valentine of the NMGW for their kindness on showing me how to operate the XRD.
 J. H. Payer, “Bronze Corrosion: Rates and Chemical Processes”, in T. Drayman-Weisser (ed.), Dialogue/89 - The Conservation of Bronze Sculpture in the Outdoor Environment: A Dialogue Among Conservators, Curators, Environmental Scientists and Corrosion Engineers, National association of Corrosion Engineers, Houston Texas, 1992, pp. 103-122
 B. Knight, “A Review of the Corrosion of Iron From Terrestrial Sites and the Problem of Post-Excavation Corrosion”, The Conservator 14, 1990, pp. 37-43
 D. A. Scott, “Bronze Disease: A Review of Some Chemical Problems and the Role of Relative Humidity”, Journal of the American Institute for Conservation 29, no. 2, 1990, pp. 193-206
 D. A. Scott, “A Review of Copper Chlorides and Related Salts in Bronze Corrosion and as Painting Pigments”, Studies in Conservation 45, 2000, pp. 39-53
 J. H. Payer et al., “Role of Transport Properties in Corrosion Product Growth”, Materials Science and Engineering A198, 1995, pp. 91-102
 G. A. Rosenberg, “Antiquities and Humidity”, The Museums Journal 33, 1933, pp. 307-314
 R. M. Organ, “Aspects of Bronze Patina and its Treatment”, Studies in Conservation 8, 1963, pp. 1-9
 W. T. Chase, “Bronze Disease and its Treatments”, The Department of Fine Arts: Bangkok National Museum, 1975
 S. Turgoose and S. J. Duncan, “Techniques for Metal Sculpture Corrosion Inhibition”, in T. Drayman-Weisser (ed.), Dialogue/89 - The Conservation of Bronze Sculpture in the Outdoor Environment: A Dialogue Among Conservators, Curators, Environmental Scientists and Corrosion Engineers, National association of Corrosion Engineers, Houston Texas, 1992, pp. 275-287
 N. Tsatsouli, “Bronze Disease: The Influence of Relative Humidity on Cuprous Chloride”, Unpublished MSc in Conservation Thesis, Cardiff University, 2004
 F. M. Howie, “Elements, Alloys and Miscellaneous Minerals”, in The Care and Conservation of Geological Material. Minerals, Rocks, Meteorites and Lunar Finds, F. M. Howie (ed.), Butterworth-Heinemann, Oxford, 1992, pp. 51-55
 C. P. Hedlin and F. N. Trofimenkoff, “Relative humidities over saturated solutions of nine salts in the temperature range from 0 to 90º F”, International Symposium on Humidity and Moisture, Proceedings, Washington, D.C., USA, Vol. 3, chapter 31, 1963, pp. 519-520. Available at: URL [pdf]
 C. Price, “Salt Damage in Porous Materials”, in An Expert Chemical Model in Determining the Environmental Conditions Needed to Prevent Salt Damage in Porous Materials, European Commission Research Report No. 11 (Protection and Conservation of European Cultural Heritage), Archetype Publications, London, 2000, pp. 3-12
 T. Drayman-Weisser, “A Perspective on the History of the Conservation of Archeological Copper Alloys in the United States”, Journal of the American Institute for Conservation 33, No 2, 1994, pp. 141-152
About the author
Alexios Papapelekanos has studied in the U.K at Durham University BA Archaeology (1999-2002) and MSc Palaeopathology (2002-2003). Subsequently, he studied at Cardiff University MSc Care of Collections (2003-2004) and BSc Conservation of Objects in Archaeology and Museums (2004-2006). Throughout the years of his study and subsequent years he has gained expertise in the field of metals conservation research and in the preventive conservation profession. On 2008 he became the co-founder and Head Scientist of the Collections Care Company (CCco) in Thessaloniki, Greece. The company promotes the proper care and preservation of collections housed in institutions or privately owned by collectors.
This html version may not include all features of the article.
To read this article with images at full quality please
Download this article
Permanent link to this resource: http://www.e-conservationline.com/content/view/863