Welcome to Terry Moxon's Agate Website  
line decor
  HOME  ::  
line decor

Agate: microstructure and possible origin. Terry Moxon

  Agate: microstructure and possible origin

Publisher: Terra Publications. ISBN 0 9528512 0 2

Price £ 8-00 for UK delivery (EU £10-00 or Rest of World £12-00 including airmail delivery).

Agate is the first detailed review of the problems of agate genesis in English  since 1927. The fully referenced text considers the major hypotheses on agate genesis from 1776. Techniques that are within the scope of the amateur are described for the preparation of rock thin sections. Optical and electron microscopy reveal much about the microtexture and the cause of some colours in agate. The reader will realise that while some advances have been made over the past seventy years there are many problems to be solved in this fascinating story of the origin of agate.
The book is intended for the interested general reader in addition to the geologist  and mineral collector.

Click on the link to contact me by email: Contact Me


With Shipping Options


Chapter 1. The agate-bearing lava in the Midland Valley, Scotland.
Chapter 2. Rocks in thin section, equipment and techniques.
Chapter 3. The microstructure of agate.
Chapter 4. Colour in agate.
Chapter 5. Rhythmic banding in simulation experiments.
Chapter 6. Crystallisation of silica from glass, gels, solution and the amorphous state.
Chapter 7. The development of hypotheses on agate genesis since the eighteenth century.
Chapter 8. Agate genesis - a discussion .
Chapter 9. From vesicle to agate amygdale

Chapter 5 from  Agate: microstructure and possible origin.

Chapter 5 Rhythmic banding in simulation experiments

Fig5.1.jpg (45810 bytes)
Fig. 5.1 Orbicular diorite. Courtesy Dept. Earth Sciences, Cambridge University.

Rhythmic banding in agate is just one example of repetitive formations that can be observed in nature. Moulds and annular rings in trees are clear examples where ring structures are added to existing formations through further growth; in other cases, the pattern develops within an already defined medium eg orbicular diorite (Fig. 5.1). However, the manner of the formation of agate banding is not obvious. One of the earliestsimulation experiments that attempted to link the observed banding to agate genesis was carried out by the French experimental mineralogist  Daubrée, who in 1880, subjected bottle glass to a high temperature  hydrothermal attack. Similar experiments were carried out by Nacken (1948)  and White and Corwin (1961); although it was Liesegang's 1896 experiments with gels that lead, in 1915, to  the simulation of most agate types.

5.1 Liesegang  Rings

Fig5.2.JPG (29290 bytes)
Fig. 5.2 Iron silicate growths produced from iron (II) sulphate in sodium silicate.

Raphael Liesegang (1869-1947) was an outstanding colloid chemist whose scientific interests included early developments in photography, bacteriology, origins of silicosis and the study of crystal growth in gels. Periodic ring formation in a variety of organic and inorganic compounds had been known since 1850. By 1913, Liesegang's contributions resulted in the phenomenon being known as the >Liesegang Rings.= Scientific interest in rhythmic banding has continued and hardly any rhythmic pattern has escaped a possible link with Liesegang Rings: solidification of melted aspirin, patterns on butterfly wings, synthetic mother- of- pearl are just three of the diverse topics involving rhythmic banding to be studied. By 1965, more than 800 scientific papers had been published on this phenomenon.  
Liesegang wrote over thirty papers on rhythmic precipitation, although it is his contributions on agate genesis that are the best known. The present day popular supplement to children's Chemistry Kits,  "The Chemical Garden," was used by Liesegang to suggest a mechanism for the formation of moss agate. When coloured metal compounds are added to a concentrated solution of sodium silicate (water glass) the outer layer of metal ion dissolves and forms the metal silicate. This silicate acts as a semi-permeable membrane; osmosis creates sufficient pressure for the membrane to burst and new metal ions are exposed to fresh silicate solution. Density differences allow the thread-like growths to rise. If iron (II) (ferrous) sulphate is used, then the resulting pattern is very similar to moss agate (Fig. 5.2).
Horizontally banded agate (German: Uruguay-Bänderung) is the second major agate type and in many ways the banded gel produces the most superficial likeness between a simulated pattern and genuine agate. Liesegang's original experiment described a layer of concentrated hydrochloric acid on a silica gel producing a white rhythmic deposit of silica. If the banded gel is now surrounded by a 20% solution of iron (III) (ferric) chloride, then the Fe3+ works its way between the silica and gives an imitation of horizontally banded agate. Several hypotheses have been proposed in an attempt to explain the banding phenomenon. Stern (1954) believed that Ostwald's (1897) original supersaturation theory was adequate. Adsorption of ions plays a part but the rhythmic banding occurs even with large crystals over long distances leaving diffusion as the major mechanism. 

Fig5.3.JPG (18158 bytes)     
a) copper chromate   b) magnesium hydroxide     c) silver chromate            
Fig. 5.3 Rhythmic banding in silica gel.
Many media can allow the production of rhythmic banding and air plays this role in the reaction between the gases ammonia and hydrogen chloride. Under normal circumstances, the mixing of these gases produces a white smoke of solid ammonium chloride. However, if two cotton wool plugs soaked in concentrated hydrochloric acid and ammonium hydroxide are placed at either end of a long glass tube, then the diffusing ammonia and hydrogen chloride meet and produce a single white ring near to the heavier hydrochloric acid end. Single rings continue to be formed although the solid ammonium chloride does tend to sink to the bottom of the tube.
Nacken (1948) and Schlossmacher (1950) used earlier experiments involving crystallisation from a silica glass as a model for agate genesis that was based upon immiscible melt droplets. Rhythmic banding is often produced from melts and is readily obtained from liquid sulphur. When sulphur is gently heated, it melts at 113oC forming an amber melt of sulphur. If the liquid is rapidly poured out of the tube, the thin film remaining allows rhythmic banding to develop from the residual melt (Figs. 5.4,5.5). Any attempt at  reproduction of the sulphur experiments will often be frustrated because of differences in the heating rate and the subsequent behaviour of the S8 molecules. The initial heating results in the collapse of the sulphur crystal structure and the S8 rings enter the molten state. Further heating allows some S8 rings to break open and link with others to form very long sulphur chains. Some chains will break open and form simpler  sulphurmolecules of the S2, S4 or S 6 type. In essence, the melt will be a variable mixture that allows the  formation of a regular rhythmic banding (Fig. 5.4) or spherulitic growth (Fig. 5.5). Both types of structures can be observed in agate and the mechanism of crystallisation from a multi-component polymeric sol of  silica has been used to argue agate genesis (Chapter 8). 
The author thought that these observations on the rhythmic banding from molten sulphur that he made in the  mid 1970's were original; they had in fact been recorded in the scientific literature in 1915! 

Fig5.4.JPG (11637 bytes) Fig5.5.JPG (17299 bytes)
Fig. 5.4                                                Fig. 5.5
Fig. 5.4 Liesegang rhythmic banding produced when sulphur is rapidly poured out of a test tube (x 40, plane polarised light).
Fig. 5.5 On a separate occasion small spherulites are produced from the sulphur. (x 40, plane polarised light). Top edge represents 1.25 mm.

Fig5.6.JPG (40335 bytes)
Fig. 5.6 Agate-type banding that has etched onto the back of a steel clock. Scale - hole diameter = 1 cm. Courtesy, J Raeburn

One of the best examples of simulated agate pattern that the author has seen is shown in Fig. 5.6. The pattern   totally mimics a fortification agate but unfortunately the development of the banding is the only part that could be linked to the story of agate genesis. This pattern has formed on a steel plate that was originally a support backing for a quartz clock. The spindle that passed through the hole was brass and the back plate had clearly been handled as faint ghosts of finger prints can be observed. This selective etching was caused by the brass spindle in contact with the iron plate and together they behaved as electrodes for a simple electrolytic cell. If moisture and salt had been deposited during handling, then the surface of the iron would go into solution and the clock battery, although not essential for corrosion, may have been an added driving force in the etching process.

5.2 Mixing immiscible gels

Sodium silicate (sold under the trade name of water glass) is a thick syrupy, colourless liquid that behaves as a weak base when reacted with an acid:
Na2SiO3 + H2O + 2HCl ------> Si(OH)4 + 2NaCl
A supersaturated solution of silicic acid is formed and this unstable solution results in a condensation polymerisation reaction taking place. This actual polymerisation reaction is complex (the interested reader should consult Iler, 1979) but the final product at acid pH values is a three-dimensional gel. Silica particles at pH values >2 carry a negative charge and this allows the surface adsorption of positively charged cations or dyes. These particles act as a bridging effect of the silica that in turn would accelerate polymerisation.   However, the dyes are sufficiently strongly adsorbed so that different dyed silica sols retain their separate colours when mixed. The overall effect is to produce a parallel banding in addition to rhythmic fortification patterns.
Similar effects can be achieved with polyester resins. When two differently coloured, liquid resins are mixed then diffusion occurs and the result is a single combined colour. If a thixotropic resin gel (silica base) is diluted with neat polyester resin, separately coloured and mixed, then the final product is a mixture of two immiscible coloured components. Many different effects can be obtained when coloured silicified resins are simultaneously poured into a mould.

Fig5.7.JPG (30118 bytes)
Fig. 5.7 Flow pattern produced when treated coloured resins are poured into a mould. Actual size.

Fig. 5.7 shows part of a 15 x 9 cm block of coloured polyester resin that has formed from a single pouring point of brown, orange and white silicified resin. The colours have not mixed and the natural flow and later interference are recorded in the cured block. A close-up of a black and white block is shown in Fig. 5.8 and here, the rising air bubbles have produced the fortification pattern at A and B. Unless great care is taken during the mixing process, air becomes trapped and the air bubbles slowly rise up through the gel. Many bubbles escape but a number are trapped and the pathway is revealed by the ripple effect when the block is sliced.
A direct comparison between these simulated fortification patterns and agate is not intended. However, there are examples where this mixing mechanism could account for complex agate patterns. The formation of some types of vein agate could be explained as a mixing of an iron -rich and an iron-poor silica gel. Once within the vein, the crystallisation could continue as explained in Chapter 8.
'Blow holes' are not a common pattern but they can be observed in a number of agates. These features appear as neat circles within the general texture of the agate. It has been suggested that their formation is due to air bubbles rising and creating holes in the gel; at a later date, a silica sol/ solution diffuses from the  surrounding gel to crystallise in a different form (Moxon, 1996).

Fig5.8.JPG (21161 bytes)
Fig. 5.8 A fortification pattern has been produced and rising air bubbles have caused the ripple effects at A and B. Actual size

Fig5.9.JPG (43109 bytes)
Fig. 5.9 a) 'Blow holes' in agate with the spherulites virtually free of iron oxide. Burn Anne agate, Kilmarnock. (x 40, plane polarised light).Scale = 0.5 mm. b) as a) with crossed polars. Side edge represents 1.3 mm.