Dr. LORRAINE GIBSON

University of Strathclyde

STANDARDISATION : IS IT ESSENTIAL ?

To protect museum acquisitions from ingressed outdoor pollutants, the public, thieves and fluctuating relative humidity, objects have traditionally been stored or displayed in wooden cabinets or drawers. However, it is now well established that woods, and wood-related materials, adhesives, varnishes and paints, all off-gas carbonyl pollutants (primarily aldehydes and organic acid vapours) for significantly long periods of time (> 10's years) after the cabinet or drawer has been manufactured. As the cabinets are normally well sealed and have low air exchange rates with gallery air, carbonyl pollutants can accumulate. High concentrations ( 100 - 1000's ppbv) of formaldehyde, acetic acid and formic acid vapours have been measured in wooden facilities currently in use¹. This poses a great risk to susceptible artefacts, such as lead² and calcareous items³, which are often irreversibly damaged after pollution attack.

For example, an efflorescent salt produced by pollution attack was identified on a limestone relief of King Amasis II at the Burrell Collection in Glasgow. The extent of crystal growth is so dramatic that it has almost obscured the 19th c Greek carving. The efflorescent salt, identified previously as a calcium acetate nitrate chloride hydrate⁴ (known as thecotrichite), is ubiquitous on excavated limestones held in similar contaminated enclosures⁵. Its mode of formation is complex, however, it is known that the presence of acetic acid vapour in the surrounding environment is a prerequisite for its formation. After simulating the mineral's growth in laboratory conditions, it was shown that to prevent formation of this, and similar, efflorescent salts it is necessary to accurately and quantitatively detect carbonyl pollutants in museum air. If high levels of acid or aldehydes are detected, mitigation measures should be implemented to reduce pollution levels. At these lower concentrations, the necessary saturated salt solutions will not form in the pores of the limestone, and so crystallisation of the salt on the surface of the artefact will not occur. In light of this problem, the Burrell Collection removed their wooden cabinets (and by association the acetic acid problem) from the storage areas and replaced them with inert metals cases.

It is difficult to estimate an 'acceptable pollution threshold' whereby the measured concentration is so low that it is assumed to be inconsequential to the artefacts in store. Moreover, before this question can be addressed, the first step is to ensure that the sampling device used to assess the environment is accurate, precise and reliable. At present there are no standard operating sampling protocols for measuring carbonyl pollutants in museum air. Different museums use different methods depending on their experience, access to equipment and availability of finances. In general, as long as the results obtained by the different techniques currently in use are comparable, it would make no difference what sampling device the user chooses.

To assess the performance of commercially available acid and aldehyde passive sampling devices, an interlaboratory comparison was established between laboratories at the University of Strathclyde, the Netherlands Institute for Cultural Heritage, the Getty Conservation Institute, the University of Oxford-Brookes and the University of East Anglia. Passive sampling devices were prepared by each laboratory and deployed simultaneously in a number of cabinets in eight museums in Southern California. After sampling, half of the sampling devices were returned to the original laboratory for analysis, the remaining half were analysed at one site; the laboratory at the Getty Conservation Institute.

The results obtained gave some interesting results. In some situations, the concentrations measured by different sampling devices, and by different laboratories were comparable. However, large discepancies were observed for a number of sampling situations. For example, the badge devices were easily saturated if exposed for long periods of time in areas with high formaldehyde concentration. This can be amended by repeating the experiment for shorter periods of time, however, the commerical supplier of the badges did not indicate oversampling had occurred and the results were given as 'accurate' despite the fact that the measured concentrations were underestimated by factors of 2-100. It was also shown that sampling preparation and methodologies adopted must be meticulously adhered to, otherwise discrepant results will be obtained even by trained laboratory staff.

In conclusion, the study has shown that in order to obtain accurate and precise results, standardisation of sampling protocols is essential. Unless standard operating procedures are devised it is unlikely that the user will obtain accurate results. The different methods are currently under review and the participating groups named above are in the process of fully validating different sampling procedures. It is our intention to prepare, and publish, standard operating protocols so that museums can use their preferred sampling method with confidence.

References:

1. Grzywacz CM and Tennent NH, in Preventative Conservation, Pracitice, Theory and Research, Roy A and Smith P (Eds.), The International Institute for Conservation of Historic and Artistic Works, London, p164, 1994.
2. FitzHugh EW and Gettens RJ, in Science and Archaeology, Brill R (Ed.), MIT Press, Cambridge, p91, 1971.
3. Tennent NH and Baird T, Studies in Conservation, 30, p73, 1985.
4. Gibson LT, Cooksey BG, Littlejohn D and Tennent NH, Anal. Chim. Acta, 337 p151, 1997.
5. Gibson LT, Cooksey BG, Littlejohn D and Tennent NH, Anal. Chim. Acta, 337 p253, 1997.

ADDITIONAL NOTES

Currently no standard methods of analysis exist, however, an interlab comparison has been set up to look at the methods currently in use. Differences in methodologies were detected, especially when the badge-type devices were used in areas with a high aldehyde concentration. One problem of using a commercial badge is that the result is sent to the user with no indication of whether or not saturation has occurred. This means that the user thinks the low concentration reported is accurate, whereas it simply has trapped too much aldehyde in the first few hours of exposure to provide an accurate result. The weight of trapped formaldehyde is assumed to have accumulated over the 4 day period of exposure, i.e. the weight is divided by 192 h, rather than its active exposure time of 2 or 3 hours. This results in a gross under-estimation of the pollutant concentration.

Conclusion : STANDARDISATION IS ESSENTIAL