Maria Felice explores the history of sheet glass, also known as architectural glass or window glass, including some recent high-tech developments such as self-cleaning glass.
Ubiquitous and ancient, sheet glass was first manufactured by the Romans. Its quality and extent of use have increased over the centuries, in part due to technology that has enabled the purification and precise measuring of its constituents. More significantly, however, it is thanks to improved manufacturing techniques that enable large panes of sheet glass to be mass-manufactured in a repeatable and efficient manner. A milestone in this was the development of the Pilkington float process in the 1950s.
Today, more than half of the global ceramics market is glass products, and 30% of this is sheet glass. Most sheet glass is soda-lime silicate, which typically consists of silica sand (72% weight), limestone, soda ash, alumina hydrate and burnt dolomite. These components are mixed in appropriate amounts to make the batch, which is then melted.
In Roman times, casting was employed to make window glass. Molten glass was poured into a bed of sand, allowed to cool slightly and then pulled at the edges to form a rectangular sheet. As one would expect, this type of glass was rather thick and had a roughened surface where the glass was in contact with the sand. From the early 17th Century, a variation of casting was used where glass was cast in large sheets using stone or metal troughs. The glass was then polished on both surfaces using a succession of finer abrasives.
During the medieval period, crown glass manufacture was introduced (not to be confused with optical crown glass used in lenses), which enabled larger sheets to be made. This required the glassmaker to place a small amount of molten glass on the end of a blowing iron, inflate it into a large bubble and then flatten it to form a circle. A large, thin disc could be formed by rotating the glass rapidly and horizontally. Another popular method was the cylinder technique, in which an elongated bubble of molten glass was formed, allowed to cool, cut down the centre and opened into a sheet.
The Great Exhibition of 1851 renewed interest in many crafts and industries, including that of glass. The Crystal Palace, in London, where the Exhibition was held, was a record-breaking example of the use of sheet glass. The pre-fabricated building was designed by Joseph Paxton, who had built greenhouses for the Duke of Devonshire at Chatsworth House in Derbyshire. The Crystal Palace consisted of about 300,000 panes of glass and at that point it was the largest ever commission of glass for a specific building project. The glass was supplied by Chance Brothers of Birmingham – the recent removal of a glass tax had encouraged the company’s development of plate glass. The sheets were each 130x25cm –the biggest made at the time, although today the record for the largest piece of sheet glass is held by an 800cm diameter mirror made by Schott for a telescope.
In the first half of the 20th Century, sheet glass was made by rolling or drawing. Rolling involves passing molten glass from a tank furnace over a refractory lip and between rollers. In contrast, drawing involves dipping a metal plate into a bath of molten glass and slowly withdrawing it – the glass is passed between two coolers to stop it necking down to a ribbon. There were three types of drawing that, by the mid-1900s, were responsible for the entire global production of flat glass – Fourcault, Pittsburgh and Libbey-Owens. After rolling or drawing, the flat glass was converted into what was referred to as plate glass by a series of grinding and polishing, so that it had two perfectly plane and parallel faces. These processes can be mechanised, but remain expensive and time consuming.
The revolutionary float process was developed in 1959 by Sir Alastair Pilkington and Kenneth Bickerstaff, of Pilkington Brothers, in the UK. It remains the method of choice to this day. The process consists of feeding molten glass between rollers onto a bath of molten tin (melting temperature is 232°C) at about 1,000°C. The resulting glass sheet is of uniform thickness because of the equilibrium between gravitational forces and surface tension. At the exit of the bath, the temperature is reduced to 600°C, at which point the tin is still fluid but the glass is rigid enough to be moved to the annealing stage, where it is slowly cooled and stresses are relieved. The optical quality of glass made by the float process is almost as good as that of plate glass, without the need to grind and polish, while the production rate is 5–10 times higher than that of drawing. Interestingly, there are only about 170 float glass processing plants in the world, and their combined daily output is almost 5,000km long and approximately two metres wide.
There is a popular misconception that old windows have bulges at the bottom because when window panes have been upright for many years, the glass flows downwards under the force of gravity. Of course, this is false, and one possibility of how it came about is that the German physicist Gustav Tammann, who was among the first to study the thermodynamics of glass, described it as a ‘frozen supercooled liquid’, but ‘frozen’ was omitted by later authors. In most cases, glass does start off as a liquid, and it is structurally the same as a liquid in that it is not crystalline. However, thermodynamically, liquid and glass are very different. Glass, even with the many impurities that medieval glass would have contained, would need to be heated to more than 300°C for changes due to ‘flowing’ to be observed. So what causes these bulges at the bottom of windows? Some windows bulge out at the sides and top rather than the bottom, indicating that it is not gravity that is the cause but poor manufacturing techniques. The bulging is often seen at the bottom of the pane because, since it occurred during manufacturing, it made sense to fit the windows so that the thickest section would be at the base.
In the 20th Century, the automotive revolution meant increased business for sheet glass manufacturers. Windscreens were originally made from window glass but this was very dangerous because it shattered into sharp pieces. It was replaced with toughened or tempered glass, which is made by heating flat glass to above its transition temperature (the temperature at which glass becomes softer) and then blasting it with cold air. The outside of the glass cools more quickly than the inside, resulting in the outer layer being subject to compression and the inside to tension. This changes the way the glass breaks – it crumbles into fragments with few sharp edges instead of breaking into shards. The problem with this type of glass was that it was a nuisance when it broke due to the number of small pieces to be cleaned up, and it broke quite easily. The third type of glass, which is in use today, is laminated glass, made by joining two or more layers of float glass (themselves tempered) with a layer of optically transparent polymer, such as polyvinyl butyral, using pressure and heat. When broken, the glass remains stuck to the polymer and the sheet remains transparent.
A change for modern times
A more recent innovation by Pilkington is a self-cleaning glass. A 15nm coating of titanium dioxide is applied during manufacture of the glass using chemical vapour deposition. The coating has two interesting properties that contribute to the cleaning process. It is photocatalytic and UV light from natural sunlight activates it so that it breaks down organic dirt. Secondly it is hydrophilic, so when it rains the rainwater spreads evenly on the glass instead of forming droplets, and runs off in a sheet taking loose dirt with it. In addition, the glass dries quickly so the water does not leave streaks.
The key property of smart glass is that its light transmission properties can be changed. One way of doing this is by using suspended particle technology, whereby a thin film of rod-like particles is suspended in a fluid that is placed between two sheets of glass. The particles are randomly oriented and do not let light or heat pass through, however when a voltage is applied they align themselves and become transparent. The National Research Council in Canada is developing micro-blind technology, which has advantages over other smart technologies, including better durability and higher switching speed. Very small metallic foil blinds that are invisible to the human eye are applied to the glass using optical lithography (transfer of pattern using light). The position of the foil pieces is changed by applying a voltage and, just like normal blinds, when rolled up they let light and heat through and when flattened out onto the glass they block them.
Since the revolution of the Pilkington float process less than a century ago, scientists and engineers have been improving window glass year on year. As more and more shiny, high rise buildings are constructed, it is encouraging that thermally efficient glass is becoming ubiquitous in the form of multiple glazing and higher tech assemblies, such as smart glass.
In brief: a potted history of The Crystal Palace
- 1851 Created by Joseph Paxton to house the Exhibition of the Industry of all Nations staged in London’s Hyde Park
- 1852 Redesigned and rehomed to Sydenham Hill in south east London and reconstruction commenced
- 1936 Destroyed by fire