Science Page 5
Zinc
Bristol Glazes
The image is an early ad for ceramics using Bristol glazes.
Turning my attention now to Zinc, starting with a bit of background history into the use of zinc in glazes.
In 1835 the Bristol glaze was developed in Bristol (then a significant centre for ceramics) by William Powell, for use in his factory (Temple Gate Pottery) on stoneware, and then spread out to other potteries with time. The common stories about it being developed as an alternative for lead glazes seem to be a post hoc justification - awareness of lead poisoning of glazers was increasing, so any glaze developed around then or later must have been a lead replacement - as stoneware would not be lead glazed. He basically replaced much of the alkali and alkali earth fluxes in a standard feldspathic porcelain glaze with zinc oxide, resulting in a smooth, opaque white glaze that was very stiff when fired, and was resistant to chemical attack. It is likely that it caught on because the white colour was more attractive than the browns and greys of salt glazed stoneware of the day, and it matured from as low as 1150C compared to 1280C for salt glazing, so firing was quicker and cheaper. Of course this meant that the clay was not vitrified but, contrary to what many studio potters today seem to believe, glazes are essentially waterproof.
Liquids contained in Bristol glazed bottles were exported to the USA, which gave Americans the first sight of the glaze. It seems likely that American copies of Bristol glazes started in the 1870s, due to British potters emigrating and taking their knowledge with them, but did not begin to be widespread until the 1890s, as a replacement for Albany slip. Claims that Sherwood Brothers (in Pennsylvania) exhibited the glaze at the New Orleans World’s Industrial and Cotton Centennial Exposition in 1884-5 seem spurious, as they are not listed in the exhibition catalogues.
As for the glaze itself, it is a simple mixture of feldspar, whiting, zinc oxide, kaolin and silica. In Watts' study of the glaze, he found a eutectic at Cone 03 (1086 - 1100C) with a UMF of 0.4 KNaO, 0.3 CaO, 0.3 ZnO, 0.6 Al2O3 and 3.55 SiO2. But more of the chemistry next week.
Zinc is not an Alkaline Earth
Because ZnO looks similar to the alkali earths - CaO etc. - and because it dissolves into the glaze at a fairly low temperature, it is put into the fluxes in the UMF. But is this the whole story? (Note that I'm leaving crystals to another day).
If you speak to scientists, they will tell you that ZnO is an intermediate. And that means that with enough alkalis or alkali earths acting as compensating charges it will be able to form tetrahedra and act as a glass former. The image shows a simple Si (blue) and Zn (grey) glaze, with O (red) and Na (yellow).
As usual, Al has precedence in grabbing the fluxes, but then Zn will grab what is left and form the necessary tetrahedra - though its priority with others such as Fe isn't clear. It pairs most successfully with potassium. Also, interestingly, it can act as a balancing charge to itself - so one Zn is in the middle of the tetrahedron, and another is hovering nearby as an ion giving the necessary 2+ balancing charge. When part of the glaze matrix, Zn increases the strength, elasticity and chemical resistance of the glaze. Some say that the Zn binds with Ca to form structures similar to hardystonite (Ca2ZnSi2O7), whereas others say that the Zn tetrahedra bond to the Al tetrahedra in preference to Si.
Increasing the quantity of Zn and we will find phase separation, one phase being predominantly Si and Al, and the other is clusters of Si/Al-O-Zn, similar to willemite (Zn2SiO4). Here the glaze becomes more durable, and the zinc-poor phase leaches out first. More work is needed here.
About 10% of the Zn never acts as a glass former but acts as a network modifier, regardless of how much flux is available. And if there aren't enough fluxes available for charge balancing, the Zn will act as a network modifier. Here it acts similarly to MgO, and will be found on the boundary between the channels of flux and the islands of glass. In this case it is destabilizing the glaze, and increasing the propensity for leaching.
And as for the white Bristol glazes, these need upwards of 5% molar of ZnO, but there is quite a lot of variability between glaze recipes.
Zinc and Boron
If we add zinc to a boro-silicate glaze then things get a bit more complex, as both the zinc and the boron want to have fluxes to pair up with, to form tetrahedra and join in with the Si glaze matrix - in addition to the aluminium. Aluminium has the greatest precedence in grabbing fluxes, then zinc, and boron is at the bottom of the heap.
If there are enough fluxes, then all 3 can act as glass formers in the glaze. But analyzing glazes in Glazy shows that about 20% of these glazes have a deficit of fluxes, so not all of the B and Zn is acting as as a glass former, and for about 13% none of the B is forming tetrahedra so we can expect phase separation and/or devitrification.
Won't these flux poor glazes to unstable? In one experiment, in a glaze with 10% molar Ca the Ca was replaced with Zn in a number of steps, the glaze structure analysed, and leaching tests performed. As the Zn increased, less B formed tetrahedra to join in with the silica network, as there was less Ca available in the glaze, and more of it was required to charge balance the Zn. So this led to phase separation - go back to my write-up on boron for details of this. It also results in Willemite (Zn2SiO4) like structures forming at high Zn levels, and the weaker Hardystonite (Ca2ZnSi2O7) at low Zn levels.
Leach testing showed that leaching was more severe in the original glaze without Zn, and the rate of leaching decreased with time, suggesting ion exchange. But where Zn had been added, so there were no spare fluxes or a deficit, and we were getting phase separation of the boron, the rate of leaching was much lower (the minimum was around where the fluxes met the needs of the intermediates exactly). Also, the rate of leaching was linear, so the whole of the glaze was dissolving, as opposed to ion exchange. Interestingly, in low zinc glazes the Hardystonite leached out readily, creating channels inside the glaze and increasing the surface area, which increase the leaching rate.
So, in short, swapping out the alkaline earth with zinc will make the glaze more durable, especially if there is just enough flux to prevent phase separation of the boron.
Crystalline Glaze Recap
Photo: Willemite crystal with cobalt. A.R.Jamaludin et al.
Before going on to the specifics of zinc crystals, I thought a general recap on crystalline glazes (or glass ceramics if you're a scientist) would be useful.
As the glaze cools, the natural tendency is to grow into a crystal, but with non-crystalline glasses and glazes the rate of crystal growth is so slow that they end up as glasses. But, with the right chemistry, part of the glaze can grow into crystals.
In many cases, crystals grow most easily on the surface of the glaze, and don't necessarily develop through the thickness of the glaze - the spots in oil spot glazes are a good example of this. Next down, they will grow on pre-existing nucleii, which may be imperfections on the pot surface, or particles with a similar structure to the that of the crystal (e.g. Ti, P or Zr). Crystals growing in the middle of the glaze thickness on their own is the least common scenario.
Crystals develop between the melting point, Tm, above which the glass is liquid, down to the transition point, Tg, below which the glass cannot deform. Overall, the temperature range from the coolest growth of nucleii to the hottest crystal growth is typically less than 200C.
When the crystal starts to develop, it first forms or finds a nucleus. This needs a certain amount of energy, or heat, to do so, and so a higher temperature increases the rate of growth. But the higher the temperature, the larger the nucleus needs to be to be stable. So there is an optimal temperature for the growth of the nucleii - about 30-50C above Tg. But the temperature depends on the glaze composition - alkali metals or fluorine make the glaze less viscous and reduce the temperature, encouraging the growth of nucleii, whereas aluminium has the opposite effect.
Now that we have a stable nucleii, the crystals can grow on them. Generally this happens most efficiently at a slightly higher temperature than the optimal nucleation temperature, but still below Tm.
Often multiple types of crystal form, but not all are large enough to be visible, and some are transient. All of this depends on both the glaze chemistry and the firing schedule.
Zinc Crystalline Glazes
After looking at crystal formation and growth in general, this week we'll look at the most popular one for those who use crystalline glazes (I like Fara Shimbo's name of crystallieri for them). This is Willemite, or zinc orthosilicate. Willemite crystals are pretty rare in the wild but, given the right conditions, grow easily in glazes.
Willemite has the formula Zn2SiO4, and grows pencil-shaped crystals - a hexagonal cross-section with the edges taken off. The crystals we see in glazes are in fact many small individual crystals, growing next to each other. Lower hold temperatures encourage longer, thinner crystals and the formation of fans; higher temperatures, whilst higher temperatures give shorter, stouter crystals. But Zinc is quite reactive, so secondary crystals are also likely to form, such as gahnhite (ZnAl2O4), and other compounds using the fluxes and colourants.
When making a crystalline glaze, particularly if aiming at a lot of crystals, you need to see it as 2 glazes in 1. There's a non-crystalline base glaze, to which you add zinc oxide and silica for the crystal glaze that will separate out. The greater the proportion of the crystalline glaze you add, the more crystals you are likely to get.
For those wanting to seed their glazes to control where crystals form, Zincite (ZnO) has been used most often by studio potters, though Zincblende and other zinc minerals will work, as will... The seed needs to be about 0.3-0.5mm diameter.
For optimal crystal formation, we want a fluid glaze so that the molecules can move around easily. This means that the base glaze should be: low Al; low alkali earths; sodium in preference to potassium; and the addition of low amounts of fluorine, e.g. as CaF2.
As an aside, Simon Coote of Waterport Pottery drew my attention on the anti-bacterial effects of some Zinc glazes. I'd heard of this in tiles, but didn't know how it worked. Apparently Zn ions leaching from the glaze destroy the cell walls of bacteria. In the paper, the Zn concentration is increased from 5 to 25% molar. Above 15%, Willemite crystals formed on the surface, zinc leaching increased (though not other elements), and also anti-bacterial activity.