>>click to enlarge<<
fibrous crystal
development of cristobalite
SCIENTIFIC DISCOVERIES

Atomic Structure of Opal

In this section we will look at the following areas:

Introduction
Opal-A
Opal-CT
Opal-C

Introduction

Most solid opal, both precious and common, has a conchoidal (glassy) fracture, and appears isotropic in thin sections when seen in the polarising microscope. Hence, opal was recorded in most textbooks as being amorphous. In the late nineteenth century, Mallard (R1589) noticed that some opals showed a distinct, though weak, birefringence, indicating that there was an underlying crystalline structure in these samples. The appearance was that of a mass of very fine, fibrous crystals. Mallard gave the name 'lussatite' to this type of opal. His observations were largely ignored for many years.

With the advent of X-ray diffraction, samples of opal were inevitably examined by this technique. Some did, indeed, appear to be amorphous, but others gave a rather ill defined pattern of broadened peaks, as was shown by Levin and Ott (R1585). Such results were interpreted by Dwyer and Mellor (R0439) as being due to the presence of beta-cristobalite crystallites, the development of which was dependent on the thermal history of the opal. While this interpretation was, at least to some extent, accepted for a long time, it was not consistent with the formation of opal at low temperatures.

Opaline silicas, especially those derived from biogenic sources such as marine micro organisms and plant silica, were found to be amorphous, but it was not until more recently that it was noted that some earthy forms also gave distinct XRD patterns, albeit with broadened peaks.

In 1955, Flörke (R1574) made a major contribution to the elucidation of the structure of these poorly crystalline silicas when he postulated that the opal structure consisted of disordered intergrowths of cristobalite and tridymite layers.

In 1963, Jones and Segnit (R0371), after examining several hundred opals and opaline silicas, showed that, with rare exceptions, natural opaline materials could be placed into one of three groups by their XRD patterns. These were called opal-A, opal-CT and opal-C. With some modifications, this classification has since been widely adopted.

Opal-A

Many opals and opaline silicas give XRD patterns indicating an amorphous or near amorphous structure. These comprise, in the main, opal, either precious or potch, from the sedimentary opal fields; biogenic silicas from plants or primitive animal life such as diatoms and sponges; and hyalite, the glass-like opal found mainly in association with volcanic rocks. These were originally thought to have a similar random network of silicon-oxygen tetrahedra. Some were found to yield several weak, diffuse bands in long exposure powder photographs, suggesting a degree of very short range order in the networks.

Flörke and Graetsch, however, showed that there was a difference in the underlying structure of hyalite and other amorphous silicas. Hyalite has a low water content (usually about 3%) and is formed at somewhat higher temperatures, generally in a volcanic environment, and is probably deposited from an aqueous vapour phase. It therefore has a structure more closely related to that of silica glass. They therefore suggested that this structure be termed opal-AG. On the other hand, most other amorphous silicas contained much more water (5-10%) and had a basic structure more closely allied to that of synthetic silica gels. They therefore recommended that this type be termed opal-AN.

Opal-CT

Most common opals have been found to give a strong XRD pattern with broad peaks. It was realised early that the positions of the peaks were related to the major spacings of the layer structures of cristobalite and/or tridymite, with the result that the patterns were interpreted as being from very small crystallites of, usually, high-cristobalite occurring in the disordered opal structure. It was not explained how high-cristobalite could form and be stable at low temperatures, however. Jones and the author (R1678) showed that most of the opals they examined gave a strong though diffuse pattern which could be interpreted as being due to a disordered interlayering of the cristobalite (ABC-ABC- stacking of silicon oxygen tetrahedra) and tridymite (AB-AB-) structures combined with small crystallite size. Such a structure is also consistent with the variations found from one sample to another of opal-CT, and their formation at ambient temperatures.

More recently, Guthrie et al (R1574), postulating ordered and disordered intergrowths of such planar units, were able to calculate theoretical XRD patterns for opal-CT. These patterns coincided closely with those obtained from naturally occurring opal-CT.

Latterly, Graetsch and Topalovic (R1771) using NMR and IR techniques as well as XRD have by and large confirmed these results. In addition, Elzea and Rice (R1770) examined opal-CT from earthy sources (but, unfortunately, not glassy types) and have confirmed, by some elegant high resolution transmission electron microscopy, that these opaline materials do consist of disordered interlayering of cristobalite and tridymite sequences.

Opal-C

A few common opal samples gave an XRD pattern closely resembling that of low *cristobalite (pj09). The few samples encountered were usually associated with volcanic rocks, with the suggestion that they had been formed at higher temperatures, or by re-heating of opal-A or opal-CT. Both of the latter forms, when heated to temperatures above C, recrystallise to a form similar to opal-C. Jones and Segnit concluded, therefore, that the structure of opal-C was close to that of low cristobalite.