Janet Hamer outlines the new glazes of Avril Farley and
describes how these sculptural crystal shapes are formed.

We are all keen to make a mixture of minerals, take it to a temperature and wonder at transformations –or we should be. It is like pushing seeds into a garden and, with a bit of nurture, flowers open and spread their colours. Making crystals is like that: a lot of science, patience and beauty. Large crystals grown in a glaze even look like flowers. They often resemble lichens and three-dimensional fans and feathers.

Crystalline glazes are those in which the oxides in the melt reform in new associations as the glaze cools to give a glass with crystals visible at the surface. A mass of crystals too small to be seen individually can give opacity and matt surfaces. Larger crystals can be grown upwards of 15 cm, appearing to float in a glassy matrix. These wonderful shapes, of distinctly different colour from their background are the result of manipulating the glaze formula and the cooling rate of the kiln. This is the science and the patience. The beauty manifests itself magically from these processes.

Typical crystal shapes are soft flat rounds which may impinge on each other in clusters, modifying the symmetry. A central star or bundle of needles can be seen in a smooth area before a fibrous ring fans out as a halo. Further haloes edge the shape where it meets the background glaze or matrix. The whole of this shape is coloured by the penetration of a colouring oxide into the crystal and is clearly seen against the background. There is often a delicate fringe of a slightly different colour where crystal meets matrix. The haloes can be deliberately placed centrally or around the border of the crystal. Where the glaze is thicker, as in the well of a bowl, the three-dimensional forms can be seen as fibrous fans filling the depth. Time is needed during the cooling of the glaze for crystals to form.

In the early stages of cooling, if the temperature is held around 1090ºC for approximately two hours, the crystals begin as simple needle shapes. These can fan out at each end into ‘double-axehead shapes’. These attractive and intriguing crystals can be retained, frozen at this stage by cooling the kiln rapidly after this growing period. The fuller rounder shapes develop when the temperature is subsequently maintained for further crystal growing periods. These periods, or pauses, are programmed into the cooling graph of the kiln controller and may last from three to eight hours.

Left: This complex crystal shows the fan-like growth and three-dimensional appearance in the depth of glaze. The blue staining ofthe crystal is incomplete due to the small percentage of cobalt oxidein the recipe. The pot was glazed first with a Ferro Frit-based glazeand over this a High Alkaline Frit 2275 base glaze. Each onecontained 0.5% cobalt oxide and 3% manganese carbonate.

Crystal formation is a selective process. As some constituents are precipitated, the remaining matrix is changed. The isolation of some of the constituents upsets the previous balance. Some of the remaining oxides can no longer remain unattached. They combine as larger molecules and stiffen the matrix. The matrix then sets quickly and crystals can no longer develop.

There are orders and preferences for how the molecules which jostle freely in the melt will re-bond into new lattice structures. Each element is characterised by a value number or valency. This number is based on the number of electrons in each atom and establishes its combining power. Valencies are balanced to match. For example, hydrogen is 1, oxygen is 2, therefore two hydrogen atoms are needed to match one oxygen, giving the familiar chemical symbol H2O (water).
The elements which are frequently used to colour the zinc silicate crystals are cobalt, manganese and copper. They have valencies of cobalt 2 and 3, manganese 2, 3 and 4, and copper 1 and 2. They have 2 in common with zinc and therefore compete for the same sites when new combinations are being formed. In the Periodic Table the colouring elements are grouped together as ‘transition’ elements. Other elements in this grouping have similar properties and are likely to be useful in a similar way.
There are further rules which govern the selective process by which crystals are positively coloured, why blue on an ochre ground predominates whereas green can be subtly combined. The electrons of the atoms, which are negatively charged, exert forces of attraction or repulsion on others which are in close proximity. This activity is referred to as an energy field. Colour separation is explained in the following extract from the section on crystalline glazes in The Potter’s Dictionary:

In order to colour the precipitating zinc-silicate crystals, the colouring oxides must be able to fit into the lattice structure. To enter the crystal, the metal colouring atom must be able to occupy one of the six sites otherwise held by zinc in the zinc-silicate lattice. Cobalt, nickel, copper, iron and manganese are transition metals and are adjacent to zinc in the periodic table. They are polyvalent and can be divalent to match zinc. Their atom sizes are also similar to that of zinc. Therefore all these metals can enter and colour the crystals.

Right: The colours here are from ilmenite (FeTiO3, iron and titanium oxide) andcerium oxide (CeO2) which has properties similar to tin oxide.

The reason for the order of precedence is that they have different liquid-to-crystal partition co-efficients, or field strengths. Cobalt oxide and nickel oxide have high field strengths. Manganese oxide is intermediate and copper oxide is low. Zinc oxide has a higher field strength than does copper oxide and, therefore, copper oxide tends not to partition strongly but will stain both the matrix and the crystals at the same time.

Avril Farley has been making crystal-glazed ceramics for four years. Her workshop is a small, neatly organised outbuilding. The workroom and its surrounding work areas are next to the cottage where Avril lives with her husband, Ken. The setting is rural Gloucestershire in the Royal Forest of Dean, England. The cottage and workshop stand above a delightful sunny garden which slopes steeply down to a stream. The slope and footbridge are the route up and down which all clay, equipment and finished work must be portered. In the late 1780s this cottage was a pub humorously named ‘The Sow with Three Tits’. Here cider was made and served to the iron workers who toiled immediately opposite, across the stream.

Avril Farley’s production consists of thrown plates, bottles and bowls. She prefers Limoges porcelain body, from Potterycrafts. She uses a Mervyn Fitzwilliam Craftsman wheel. She does a small amount of turning to make neat footrims which accommodate the inevitable glaze run and grinding. The two 4 1/2 cu ft electric kilns are fired at night for economy with the use of two Cambridge 401+ controllers. She usually achieves two biscuit and one glaze per week, alternating with one biscuit and two glaze firings.

This fringe-edged crystal is approximately 4 cm across. Itappears in a glaze with 1.5% vanadium oxide (V205) and1.5% ilmenite in a Ferro Frit base glaze.	This glaze shows how copper oxide can give a green stain toboth crystal and matrix. The glaze has 3% copper oxide and 3% barium carbonate (BaCO3). The barium carbonate shiftsthe colour towards turquoise.

Glazes are brushed on, thicker above than below to allow for considerable glaze movement. Calcining the zinc oxide removes water and helps to avoid flaking of the glaze. A binder, ‘CMC’, is used to make the glaze less friable. Every pot is fired on a ‘catcher’ made to measure. Surplus run off glaze is contained by the ‘catcher’ which must be separated after firing. Glaze and foot are then ground smooth. This is a demanding process requiring specialised grinders for the particular shapes and a skilled operator who is efficiently clad in protective clothing, goggles and helmet. Every firing contains tests. New shapes show their effect on glaze run and positioning of crystals. The permutations seem infinite. Variations of amount and combinations of colouring oxides are studied in glazes with The diagram shows how felspar, Alkaline Frit 2275 or Ferro Frit 3110, sometimes layered cobalt oxide can replace zinc together. Firing is usually to between 1245ºC and 1265ºC with many crystal-oxide, being the same size, growing pauses in the cooling cycle. The first at 1080ºC for 30 minutes may be and fit into the lattice followed by a pause at 1060ºC for 30 minutes, then a rise to 1080ºC again. structure of a growing crystal There may be as many as six pauses but Farley also likes to vary the timing of of zinc silicate. these periods to give more interestingly placed haloes.

Farley declares her approach to ceramic chemistry was instinctive. From a nonscientific early career she is now developing a firm understanding of chemistry through methodical practice and application. She enjoys the discipline which the creation of crystalline glazes demands. In a detailed directive to herself, ‘The Learning Curve’, she stipulates every practical rule steering a controlled course through trial and error progress.

Record keeping is unquestionable. Glaze making is precise and meticulous with thorough attention to care of equipment. Firing records include identification of all tests, positions in the kiln, weather (for its influence on cooling rates), temperature readings and digital pictures are stored. Suggestions for variations on every aspect of the making follow with practical guidance for control of glaze fusion, and ending with warnings against impatient kiln opening. Avril Farley is currently experimenting with more materials based on elements from the lanthanoid series sometimes called ‘rare earths’.