lustre pottery: the secret of early nanomaterials revealed by transmission electron microscopy
A. Larrea and J. Pérez-Arantegui
(Departamento de Química Analítica, Universidad de Zaragoza)
Lustre is a type of ceramic decoration, which results in a beautiful metallic shine and coloured iridescence on the surface of the ceramic object . Among the different ceramic techniques often termed as ‘lustre', the ‘reduced-pigment lustre' is made by applying some pigments (usually silver and copper compounds) mixed with clay or ochre to the surface of an already-fired glaze. Then the ceramic piece is re-fired to a low temperature in a reducing atmosphere. When the piece is cool and out the kiln, the surface is cleaned by rubbing to remove the clay that is on the design, and lustre appears with metallic sheen and different colours. Almost all the lustre made before 1800 belongs to this family and it includes all Islamic, Spanish and Italian lustre ware.
The earliest lustre was probably made in Iraq in the IXth century AD on tin-glazed ceramics. Lustre technology spread from the Middle East to Persia , Egypt and Spain , through Islamic cultural exchanges. In the Muslim Spain, the first examples date back to the 12th-13th c. AD, and they were followed by a very large and splendid Hispano-Moresque production during the 14th and 15th centuries in the area of Valencia and continued in the following centuries (Figure 1). From Spain , lustre pottery was introduced in the Italian peninsula where the famous ‘majolica' was decorated by this technique since the 15th and 16th centuries.
||Figure 1: Lustre ceramic from Manises, circa XV th century. Courtesy of Museu de Ceràmica, Barcelon , Spain .
Therefore, lustre pottery has been known for more than a thousand years but it has always been perceived as enigmatic. This special consideration becomes from the secrecy of its very difficult production technology, kept zealously during centuries by the potters: “many make them (the lustre pottery) on the floors of houses which are locked and under close guard, for they look on the manner of making the kiln as an important secret and say that in this consists the whole art” (Piccolpasso, 1557) [ 1 ]. Although an important amount of information is now available, the theory of this technology is not completely understood and the practice remains uncertain because of the number of variables.
The production of lustre was seldom described in detail before the late 19th century. However no written records exist for the first four centuries, therefore there is a risk of interpreting the ancient technology in terms of current techniques. This is why an extensive study of Medieval lustre pottery is carried out in order to know and to understand the pigments used, the way of production, the microstructure of the decoration and its optical behaviour and appearance [ 2 ] .
The knowledge of some of the compounds used as pigments for lustre, the reducing atmosphere currently used and the resulting metallic aspect helped to suppose that lustre was caused by a metallic film created during firing on the glaze surface, but at present only a few attempts had been published to verify the chemical and structural composition of lustre. These previous studies concluded that silver and copper were the main components of lustre. But many questions about metallic and crystalline character of these components and its complete microstructure remained still unanswered. Therefore an especial analytical approach, the Transmission Electron Microscopy (TEM), was needed to resolve the composition and microstructure of lustre. In order to study the complete in-depth lustre microstructure, a cross section was preferred for observing the samples under TEM. The preparation method used for preparing the thin cross-section specimens was first developed by Abrahams and Buiocchi for studying semiconductor epitaxial structures [ 3 ].
Two different types of samples were selected for studying lustre decoration, they were chosen because they belonged to the extremes of colour range appeared on lustre pottery. The first sample (A) is a fragment decorated with red-brown copper-like lustre from Paterna (Spain), 14th century AD, and the second one (B) is part of a Syrian ceramic, 12th century AD, with a green-olive yellowish lustre. In both cases, samples belonged to ceramics decorated with lustre designs on white tin-glazes.
In TEM, sample A appeared with a homogeneous surface microstructure formed by small quasi-spherical clusters, arranged in two different sizes and embedded in an amorphous matrix (the glaze) (Figure 2). The total thickness of this structure was between 200 and 500 nm. A first outer layer was formed by the biggest clusters, which had a diameter of about 50 nm. Diameter of the other smaller inner clusters was between 10 and 20 nm.
||Figure 2: TEM image of a red-brown, copper-like lustre. The inset corresponds to the CBED pattern of a copper particle.
About the composition of these clusters, TEM-EDS analyses indicated that the nanoclusters were particles of pure copper. Convergent-beam electron diffraction (CBED) in 50 nm diameter particles proved that they were crystals of metallic copper; inset in Figure 2 shows the face-centred cubic arrangement pattern, with a cell parameter of 1.80 Å corresponding to metallic copper.
Sample B showed a slightly different microstructure in TEM. Also in this case a nanostructure formed by small clusters appeared in the surface of the glaze (Figure 3). Nanoclusters were embedded in the matrix covering a thickness about 1 m m. Clusters seemed to be classified in three ranges of size; the outer layer was constituted by clusters of diameter between 7 and 10 nm; these first clusters were continued by a layer of very small particles of about 4 nm of diameter and the most inner clusters were the biggest with a diameter of 15-20 nm.
TEM-EDS analyses in the nanoparticles indicated that they were composed by silver, but in this case the smaller size of particles did not allow us to avoid the presence of glaze components by means only of the micro-analytical results. In this second sample, the electron diffraction was also key in resolving the nanoparticle nature. SAD patterns (inset of figure 3) of these particles showing its diffraction rings proved that nanoparticles were crystals of metallic silver.
||Figure 3: TEM image of a green-olive, yellowish lustre. The inset corresponds to the SAD pattern of silver particles.
Summarising, lustre is the first nanostructured metallic thin film made by
person, as revealed by Transmission Electron Microscopy. Considering cultural
heritage of this type of decoration, this is a remarkable point in the History
of the Technology, because nanocrystal films were produced empirically since
the medieval period and this means a high level of technological knowledge
of materials science to obtain and to reproduce lustre layers.
J. Pérez-Arantegui, A. Larrea, Trends in Analytical Chemistry, 2003, 22 , 327-329.
J. Pérez-Arantegui, J. Molera, A. Larrea, T Pradell, M Vendrell-Saz,
I Borgia, B. G. Brunetti, F. Cariati, P. Fermo, M. Mellini, A. Sgamellotti
and C. Viti, J. Am. Ceram. Soc. 2001 , 84
This study was supported by P047-2000 research project of the CONSID-DGA
(Aragonese Government), Spain .
 Piccolpasso, C. I Tre Libri dell'Arte del Vasaio, 1557 ; Edizioni all'Insegna del Giglio: Firenze, 1976.
 Molera, J.; Mesquida, M.; Pradell, T.; Pérez-Arantegui, J.; Vendrell-Saz, M. Archaeometry 2001, 43 , 455-460.
 Abrahams, M. S.; Buiocchi, C. J. J. Applied Physics 1974, 45 , 3315- 3316.