SUSAN BRADLEY and DAVID THICKETT

The British Museum

THE POLLUTANT PROBLEM IN PERSPECTIVE

The British Museum has a long involvement in studying the causes and effects of indoor pollutants, which started in the 18th century. In 1966-70 Werner defined the problem of the materials used to make showcases and published a short paper on this problem. In 1972 Oddy published a standard corrosion test to investigate the adverse effects of display materials on the metals copper, silver and lead. In 1973 the routine screening of fabrics and other materials such as adhesives, sealants, and paints, used inside showcases was started and has continued. The standard corrosion test is a 'catch all' test.

Organic acids have long been known to cause the corrosion of lead; but they can also have an adverse effect on other metals such as copper and its alloys, and zinc, and react with materials such as calcium carbonate and soluble salts to complex inorganic/organic mixtures. Organic acids are not present in the ambient air at concentrations great enough to cause corrosion within collections; but they are given off by wood and wood based products, as is formaldehyde. In 1993 there were 160 stores in the Museum and 6-7 million objects. Many of the storage units within the stores are constructed of wood, and replacement of these would be expensive in both time and money. The approach taken to deal with the problem of outgassing in storage has been to identify objects at risk, eg lead medals, and move these to metal storage units. Sometimes identifying at risk objects is not as easy as it might be. For instance it is only by analysis that lead is identified as a component in a 'bronze' object.

When considering showcases, organic acids can be totally eliminated by testing all internal construction materials. However we have found that the working properties of wood cannot be duplicated by using other inert materials. Plywood or MDF are used to construct case inserts, and composite boards such as Sundeala laminated to MDF are used for back boards onto which objects are pinned. Observation suggested that although all of the showcases contained some wood or wood based product, most objects made from or containing lead were not corroding. Using laboratory experiments, pollutant monitoring with metal coupons and commercially available monitoring methods, and by observation of the collection we have been able to put the pollution problem within the Museum into perspective.

Experiments
In the 1980s there were two instances of the corrosion of copper alloy objects to copper acetate, one in a showcase and the other in a storage area. Both occurred at a period of high RH and high temperature in summer, and it appeared that this combination had promoted outgassing causing the corrosion.

Recently the effect of RH and temperature on the rate of outgassing of Oak and Spruce has been investigated. Increasing RH or temperature increased both outgassing and the extent of corrosion of lead test coupons. Although acetic acid was detected at 30% RH, at all temperatures, no corrosion of the lead coupons occurred.

In other experiments the effect of formaldehyde on lead was investigated at ambient temperature. At 100% RH after 120 days acetic acid had caused complete corrosion, formic acid had caused extensive corrosion, and formaldehyde caused least corrosion. At 50% RH acetic acid and formic acid were less corrosive than at 100% RH, and formaldehyde did not cause corrosion.

These experiments suggest that, as we had thought, acetic and formic acids are major problems for the collection, and formaldehyde is of little significance at ambient RH and temperature.

Showcase design
In 1988 a metal and glass showcase design was introduced into the museum. It was not possible to remove wood entirely because no stable alternative which was easy to work, and to pin into, could be found, as discussed above. A method of reducing the emissions from the wood was sought. The method chosen was to wrap the boards in Melinex and then the dressing fabric. This was expected to reduce emissions by presenting a barrier to movement into the showcase. More work on sealing which will be published in the Autumn 98 edition of The Conservator has shown that Melinex is not the best sealant, and also that pinning reduces the level of protection obtained. However it is still worth covering the boards.

All other showcase materials have to be considered. Although all materials used in the construction have been tested, not all of the materials have passed all of the tests, but this has not presented substantial problems. We have also tested materials used for wall coverings and flooring in galleries, and have concluded that the outgassing from these materials is of minor importance to the air quality inside a showcase. The low levels of carbonyl pollutants in the galleries, particularly where there are wood floors, has been confirmed by monitoring.

Monitoring inside showcases
We decided to monitor inside the showcases for organic acids and formaldehyde.
Several galleries were chosen and the earliest opened in 1989. None of the galleries are air conditioned; and all of the showcases chosen contain objects made from either lead or alloys containing lead. No corrosion of the objects has occurred. Several periods of monitoring were carried out and these showed a seasonal variation in pollutant levels (higher in summer and lower in winter). For brevity, typical values are given below. In the gallery opened in 1989 the levels of acetic and formic acid and formaldehyde were determined in one showcase containing leaded bronzes. The objects had been on exhibition in the showcase since 1989. The level of acetic acid was 303 mg m^-3 and formic acid was below the level of detection; the level of formaldehyde was 24 ug m^-3. In a gallery opened in 1994 the pollutant levels in a showcase containing objects made from lead, copper alloy, pewter and silver were 262 ug m^-3 of acetic acid, formic acid was below the level of detection and the formaldehyde level was 171 ug m^-3. In a gallery opened in 1997, because of the number of objects containing lead, extra effort was made to reduce the wood content at an additional cost of £14,000. In one of the showcases the acetic acid level was 317 ug m^-3, formic acid was below the level of detection, and the formaldehyde level 470 ug m^-3.

The data gathered within showcases suggests that levels of acetic acid of around 350 ug m^-3 can be present in a showcase, or other container without lead or copper alloy objects corroding. However the RH and temperature have to be taken into account, and the experiences of high RH and high temperature leading to corrosion observed in a wood showcase, and wood storage drawers; and the experimental findings that corrosion is greatly increased at high RH compared with that at 50% RH should not be ignored.

Egyptian bronze artefacts stored in wood drawers in wood cupboards since the 1930s had an unusual blue corrosion product. This was known to have been present in the early 1970s. In the collection of 2875 objects 6.4% (184) had the blue corrosion. (About 50% of the objects had, or had had bronze disease.) A new curator was concerned; was the corrosion stable or was it growing? The blue corrosion was analysed and found to be a complex mixture containing acetate. The environment in the cupboards was monitored. The RH achieved a maximum of 45%, which is typical of such storage installations in the Museum. Acetic acid was the most abundant pollutant in the cupboards with levels of approximately 1000 ug m^-3 in two cupboards and of >2500 ug m-3 in a third cupboard. There was a direct relationship between the level of acetic acid and the percentage of objects with the turquoise corrosion in the cupboards. There was no evidence that the corrosion was on going; and lead coupons put into the cupboards have still not corroded after two years exposure. It is considered likely that as heating in the Museum has improved and the cupboards have seasoned over time the RH has fallen to a level at which the corrosion process does not proceed.

Conclusions
The authors have so far concluded that three factors are important in corrosion of objects by carbonyl pollutants. These are; the level of pollutant present in the closed environment, the RH, and the temperature. Higher RH and temperature lead to corrosion, but of these two factors RH appears to be most important. At high temperature and low RH corrosion does not occur, as outgassing is retarded, and corrosion rates decrease at low RH. The nature of the artefact and its history also has a role in the process.

The evidence shows that of the carbonyl pollutant gases monitored acetic acid is the most aggressive and formaldehyde the least agressive.

To put the pollution problem in perspective in the Museum collection there are 6-7 million objects. If 1000 object were affected by carbonyl pollutants this would be 0.016% of all objects; if 10,000 were affected it would be 0.16% of all objects. Despite many years of on-going research we have not identified acetate or formate corrosion products on anywhere near these numbers of objects. We conclude that the problem is small; but of course it can be damaging for objects which are affected, and are not identified as undergoing change. The most important thing we do has not changed in the 245 years since the Museum was founded, it is to look at the objects.

ADDITIONAL NOTES

Of the 60 million objects, housed in 160 stores, many with wooden cabinets, only a few odd incidences of corrosion have been observed. Several studies have been undertaken to assess the importance of carbonyl pollution :

1) In the 1980's 2 instances of copper acetate were detected on bronzes. In one instance the acetic acid evolved from paint, and in the second study from the oak storage cabinets during periods of high temperature and humidity. It was noted that the corrosion appeared during the summer months.

2) In the laboratory a number of experiments have been undertaken to assess the susceptibility of metals to acids and aldehydes. They found that at 5 and 0.5 ppm, acetic acid, formic acid and formaldehyde were all corrosive to lead at 100 % RH. At 50 % RH, acetic acid and formic acid were found to be corrosive, but formaldehyde was not. They believe that the formaldehyde is not easily oxidised to the acid. For copper, the corrosive activity of acetic acid, formic acid and formaldehyde reduces significantly when the relative humidity is lowered from 100 to 50 % RH, even with 5 and 0.5 ppm. Negligible corrosive effects were observed for formaldehyde with copper even at 5 ppm and 100 % RH.

3) In the galleries the BM use MDF boards that are covered with Melinex to provide a barrier to outgassing. Pining, however, will increase outgassing. They have investigated a number of different sealant methods and this work will be published soon in The Conservator.

4) They monitored in permanent galleries (since 1989) with lead, copper and silver coupons as well as with acid and aldehyde tubes. Spent £14,000 to remove wood from the newest cases and used sealents. However, they still found significant levels of both organic acids (200 - 400 ug m^-3 acetic acid) and aldehydes (20 - 600 ug m^-3 formaldehyde). No associated corrosion has been observed to date.

5) Enamels stored with an acetic acid concentration higher than 2000 ug m^-3 showed evidence of acetate salts.

6) MDF fibreboard in a display case was painted. After only 4 months leaded letters inside the case corroded to form lead formate. The leads are now isolated in a Perspex box.

7) Corrosion was identified on items in the Egyptian bronze collection. Turquoise salt was identified as a NaCu(CO₃)(CH₃COO), a pale white that was sodium acetate trihydrate and a pale green Cu₂Cl(OH)₃ - atacamite. The corrosion found did correlate with the concentration of acetic acid in the storage environment.

The risk to objects must be considered. Factors which contribute to the risk include the environment (pollutant, RH and temp) and the artefact (alloy type, associated materials, previous burial and conservation history). They have seen no deterioration with their Egyptian limestones. The BM concludes that there is no significant risk with formaldehyde. Formic acid is a risk but is rarely present and acetic acid is the most aggressive pollutant. However the corrosion observed is a very small problem compared to the others faced by conservators. Carbonyl pollutants can be present at low concentrations (400 ug m^-3 acetic acid and 600 ug m^-3 formaldehyde), but not cause corrosion unless the RH and temperature are sufficiently high.