1 INTRODUCTION

Damage created by swelling clays commonly found as minor components in ornamental and building stone is a major problem in the conservation of cultural heritage (Kuhnel et al. 1994; Ruiz de Argando�a et al. 1995; Brattli and Broch 1995). It has been reported that structural damage to building and sculptural materials can occur even when the clay fraction in a particular stone is quite small (Dunn and Hudec 1966; Delgado-Rodrigues 1976; McGreevy and Smith 1984; Wendler et al. 1991; Snethlage et al. 1995). Water penetrating within the pore system of the stone can produce swelling of the clays, and the resulting swelling pressures can damage the internal stone structure. Repeated cycles of wetting and drying can lead to almost complete destruction of the stone.

Much of the limestone commonly used for building and sculptural purposes usually contains a small proportion of clay (Caner and Seeley 1978). However, the clay fraction in micritic and biomicritic limestones can reach values of more than 10% (Folk 1962). It has been reported that limestone used as building stone or for sculptures is prone to rapid decay when in contact with water if the stone contains more than 5% clay (Rodriguez-Navarro 1994). This rapid decay is associated with clay that is concentrated along bedding planes, because, as reported by Dunn and Hudec (1966), some limestones containing more than 30% clay homogeneously distributed within the stone are very sound and stable.

In the work reported here, the composition and textural arrangement of the clay minerals in samples of Egyptian limestone were studied. The samples were obtained from a micritic limestone sculpture that had sustained major structural damage while stored in a museum environment. Typically, the damage consisted of splitting and spalling, which resulted in loss of the surface details. The samples were taken from a stela that was excavated from a Second- to Fourth-Dynasty (2720–2150 B.C.) cemetery in Naga el-Deir in the early 20th century (Phoebe Hearst Museum of Anthropology, Berkeley, California).

In previous studies carried out on Thebes-Abydos limestone sculptures, it was concluded that most of the observed damage was due to the presence of salts (sodium chloride and sodium nitrate) (Oddy et al. 1976; Helms 1977; Charola et al. 1983; Hanna 1984; Bradley and Middleton 1988; Middleton and Bradley 1989; Miller 1992; Nunberg et al. 1996). In fact, many sculptures were routinely immersed in water shortly after excavation to extract salts (Winlock 1921; Oddy et al. 1976). However, in most cases the decay process was not stopped by this treatment (or other salt-extractive treatments such as poulticing), as evidenced by major loss of surface stone after storage for a period of years (Miller 1992). Barton and Blackshaw (1976) and Bradley and Middleton (1988) pointed out that the mineralogy, especially the clay content, should play an important role in making this stone susceptible to decay, but further studies on the contribution of clays to the decay of these ancient Egyptian sculptures were not performed. It should be mentioned, however, that salt damage can be enhanced in the presence of clays (Harvey et al. 1978; Fookes and Poole 1981). McGreevy and Smith (1984) proposed a model in which decay problems related to clays could be enhanced by the presence of soluble salts.

It has been observed that these Egyptian sculptures, as well as many others fabricated from Thebes-Abydos limestone and stored in different museums, developed the same pattern and degree of decay, even though they were not exposed to an outdoor environment (Charola et al. 1983; Bradley and Middleton 1988; Miller 1992). In fact, in a museum environment these limestone sculptures experienced a varied loss of surface relief while being subjected only to relative humidity and temperature changes in the storage areas. This damage, in many cases, is not explained solely by the action of salts.

To demonstrate the role of clay minerals in the decay of the Naga el-Deir Egyptian limestone stela, experiments were performed that included a thermomechanical analysis of clay swelling behavior in the presence of distilled water, cyclic wetting/drying (using distilled water and ethylene glycol), and relative humidity (RH) cycling decay tests.