Curation history and mineralisation of highly degraded pyrite fossil collection

A collection of fossil specimens in a wooden tray in the palaeontological collections at Keyworth was found in a catastrophically degraded state due to suspected pyrite decay. 56 of the 65 specimens had undergone total destruction and had been replaced by a large volume of grey dusts. The remaining nine specimens were extensively damaged. The specimens (mainly goniatites) were collected from the Westphaliam Coal Measures at Stubben Edge Hall in the 1940s. They were stored at the Leeds Kippax store until 1985 when they were trensferred to Keyworth. Most of the damage appears to have occurred since the specimens were transferred. However, pyrite oxidation may have been initiated much earlier than this and once begun can be self sustaining. Samples of the grey dust, and some of the surviving specimens, were examined by Scanning Electron Microscopy (SEM) and X-Ray Diffraction (XRD) in order to characterise the nature of the reactions and the mineralogy of the reaction products. These observations confirmed that the pyrite had been oxidised and larger volume hydrated Fe sulphates had subsequently formed. The main reaction product is römerite (Fe2+Fe3+ 2(SO4)4.14H2O) with lesser amounts of coquimbite (Fe3+ 2(SO4)3.9H2O). Occasional cyrstals of elemental sulphur and possible jarosite (ideal formula KFe3+ 3(SO4)2(OH)6) are also present. Elemental sulphur has not previously been recorded as a pyrite decay product in museum specimens. It is indicative of extremely acidic, oxygen-poor conditions. Previous work suggests that pyrite oxidation can generally be prevented or inhibited by storing samples at relative humidities of less than 60%. If specimens contain large amounts of organic carbon it may be necessary to store them at less than 30% to inhibit pyrite decay. Environmental monitoring of temperature and relative humidity was carried out in the area where the tray was stored, and inside an adjacent tray, for a week in each location. The relative humidity inside the tray was generally close to 40% even though extremes of 25% to 75% were recorded elsewhere in the store. This implies that catastrophic damage can happen to pyite specimens even when the %RH can be kept close to 40%. However, as the monitoring was only carried out for one week in each location; a full annual record should ideally be made to assess the long term conditions of storage. The main reaction products identified by SEM and XRD are probably stable at relative humidities of less than 60%, and so could have formed in the environmental conditions that seem to be present in the Keyworth store. A comparison with similar material from similar geological units was made. No other material was as badly damaged as the Stubben Edge Hall specimens and the possible factors contributing to their more extreme damage could include: exact geological horizon (potentially affecting amount of pyrite and organic carbon present); weathering conditions at the collection site; storage conditions immediately after sampling and prior to final storage in the BGS collection; and fossil specimen Genus (potentially affecting pyrite microstructure). Any correlation between fossil Genus and degree of damage was not investigated and it is recommended that this be carried out for both the Stubben Edge Hall specimens and all the comparator specimens. The nine surviving specimens will be treated with ammonia vapour to prevent further deterioration, and monitored at 6 monthly intervals.