Platinum Metals Rev., 1966, 10, (3), 78
Platinum in the Glass Industry
The Design Of Protective Sheathing
Glass technologists are well aware of the advantages of cladding furnace refractories with platinum in order to avoid contamination. This article emphasises the even greater advantages to be obtained by giving careful thought to the design of the refractories to ensure that they are in the most suitable form for platinum cladding.
The principal object of using platinum linings or claddings in the glass industry is not so much to prolong furnace life—modern refractories wisely employed will give long enough lives—but to protect the glass from contamination. This can result of course from attack by the molten glass upon the refractories, which have become increasingly difficult to dissolve in the glass by reason of their steady improvement so that the major risk arises from stones, cord and rejected ware. To ensure a uniformly high quality of glass, therefore, many different types of refractory elements are now clad with sheet platinum before being put into service. At the same time, difficulties seem to arise in some cases in achieving the full benefits of platinum cladding, often because of the conventional shapes of the refractory materials employed. Some thoughts on this subject of the appropriate design of refractories for platinum cladding form the basis of this article.
It has to be appreciated that platinum, and the 10 per cent rhodium-platinum alloy more generally used in the glass industry, have limitations in tensile strength at high temperatures. Thus, while there is ample strength for the platinum to support itself, there is frequently some doubt as to whether or not as a lining or as a cladding it will in fact support the viscous drag of the molten glass over its surface. It is this doubt that sometimes causes glass makers to hesitate to introduce platinum cladding into their furnaces. This difficulty is really one of design. When it is considered that from a simple flow block clad with platinum we may progress to a complete platinum-lined furnace operating at 1450°C it will be appreciated that the problem is not insuperable.
Where to use Platinum Cladding
Let us take a modern tank furnace and consider where platinum or its alloys may be used with advantage.
Most modern tank furnaces operate at temperatures in excess of 1450°C and it would seem at first sight that there is not much hope of employing platinum in the melting end of such a furnace. But no glass maker is going to pretend that the whole of the melting end is everywhere at the indicated melting temperature, particularly in the region of the charging point or “dog house” as it is more familiarly known. Here the temperature is far below that of the general melting temperature within the furnace, and it is here that some of the heaviest corrosion of the refractory materials takes place, be they fusion-cast or otherwise. This heavy corrosion is due to the melting of alkali in particular at a comparatively low temperature so that molten sodium carbonate is able to attack the refractories long before it reacts with the silica of the batch. Other low melting point constituents of the batch such as fluorides will have the same effect.
Now, if there is one thing that platinum and its alloys are resistant to it is attack by molten alkalis—witness the analytical chemist who daily carries out fusions of almost pure sodium carbonate in platinum crucibles for the purpose of glass analyses. Thus, platinum cladding of the “dog house” is certainly a possible application.
There would appear to be little possibility of the use of platinum cladding in the remainder of the melting end, except when we come to the throat. Here is the next point of most severe attack, to such an extent that the throat is frequently enlarged to some three times its original size during the course of the furnace campaign. Can anything be done here? The answer depends on the throat temperature. If the average throat temperature is below 1450°C the answer is “yes”.
No fear need be felt of the platinum falling away from the refractory; modern techniques of pinning the metal to the refractory will take care of that, and even if the refractory were not so pinned the flow of glass through the throat would be quite sufficient to maintain the platinum sheet in close contact with the refractory, it being assumed that the platinum sheet was restrained from movement by gripping it in some way at its upper edge.
As we pass from the melting end to the working or conditioning end, and the temperatures become lower, so do the opportunities to employ platinum coatings increase. Especially is this the case in the feeder channels.
One difficulty, arising from the reduction in temperature that must inevitably take place as we proceed from the throat to the feeder orifice, is that it becomes increasingly more difficult for the glass to dissolve any stones it may pick up, or to disseminate any cord arising from the attack of the glass on the working end refractories and becoming entrained in the main flow of glass to the feeder. In the feeder itself any stone picked up in this region will surely not have the chance to dissolve before it emerges from the feeder in the gob of glass fed to the forming machine, often at the rate of 30 tons per day per feeder from a relatively narrow and shallow channel.
The Question of Design
Just as the early railway carriages looked like Brougham coaches, and just as the engine of a motor-car is generally at the front because that is where the horse used to be, so has the design of platinum claddings been dictated in the past by the conventional shapes of refractory materials.
Let us illustrate this by reference to a classic case, which contains all the elements of design it is necessary to consider here. A furnace was desired for the melting of a special glass; this had to be a tilting furnace so that the glass could be poured out of it. For several years the furnace was built with the conventional rectangular bricks, resulting in a box-like structure with a sloping side leading to the rectangular pouring spout. The result was that the rhodium-platinum lining to this furnace, shown in Fig. 1, had to follow the same contours and contained, apart from the numerous sharp angles and welds, some fourteen right-angle bends and some six welding lines. This was quite difficult to produce by the platinum fabricator, while when the furnace was tilted the rear wall had nothing to support it and was in danger of “bellying” away from the refractory backing. In addition the edges of the refractories themselves formed ideal situations for the tearing of the platinum coating.
How different is the present-day lining designed for the same purpose and illustrated in Fig. 2. It has innumerable advantages; there are no sharp corners in the body of the lining; this being hemispherical it contains the maximum volume for the minimum surface area (and therefore the minimum amount of platinum) and there is only one weld where the spout is attached. Of course to do this the refractories had to be suitably shaped, but there is really no difficulty in this, whether they are fusion cast refractories or hand-made sillimanite blocks. In addition the shape of the lining alone is intrinsically stronger than its flat-sided forerunner, and one begins to wonder why it was not thought of in the first place. That it was not is due only to the fact that in the first place a furnace was required and it was built out of the conventional materials to hand—rectangular bricks. Then came the request to line it.
In the use of platinum cladding today one must think first of what the lining has to do and then see if by some modification of existing conventional refractories the job cannot be made simpler and more certain of a long life. Thus, there is no reason why a skimmer block should be of rectangular section, as in Fig. 3a, when if it is to be protected from attack by molten glass by a platinum coating it could not be made of the section shown in Fig. 3b, with no sharp corners and allowing the weld of the coating to be made clear of the glass. It can then be supported from above, or rest in suitable channels in the furnace walls.
A “gate” for controlling the flow of glass, as in the Danner trough for tube drawing, need not be rectangular; it could be round with advantage and even have a dome shaped end, as in Fig. 4, if only the refractories received a little attention first. Indeed, one design of gate uses a round refractory piece so that it can be “turned as the wear on the gate becomes too pronounced”; it needs only the bottom of the Danner trough (which in any case is hand made) to be lowered a little in the region of the gate and any sticking troubles when the gate is fully lowered and the flat surfaces come together would be eliminated, as with the domed bottom it would come into contact only with the base of the trough at a point, or at most over a very small area.
A feeder bowl is almost an ideal shape to line with platinum, as also is the round and pointed needle-like plunger. Why not then line the whole of the feeder channel? It does not have to be a rectangular shape, and the top of the channel does not have to have a sharp edge. A suitable design is shown in Fig. 5.
Again, floaters for indicating glass level by means of a lever arrangement do not have to be flat rectangular-sectioned discs, they can be made more like two spherical surfaces, as in Fig. 6.
Even tank blocks, where they can be covered with platinum, do not have to have a sharp upper edge; they are going to be protected in any case and their shape above the glass surface is immaterial, so why not make it as easy as possible for the expensive lining which is to be attached at little additional cost for the refractory?
This matter of design for the job has now come to be appreciated in the use of platinum coatings, just as developments in other spheres have led to improved products and performance. To use a material to its best advantage one must consider how much and to what extent the conventional equipment needs modification.
Tank furnace operators are not alone in their use of conventional refractory shapes; the user of pots is equally conventionally minded. There are many pot furnaces in the optical glass industry using conventional pots lined with platinum with the old flat bottom. These might be much better if this bottom were part of a spherical surface, as shown in Fig. 7. In continuous platinum-lined optical glass furnaces this principle is so well realised that the various sections of the furnace are in fact cylinders of platinum lining a cylindrical shell. In this case pure platinum is employed because any solution of rhodium in the glass, however slight, is held to be deleterious to the optical perfection of the glass.
Mention can be made also of the hand gatherer. Platinum-covered gathering rings were in use almost twenty years ago. Their use then was restricted to lead glass, as the rings would be too heavy for soda glass and would sink. However, with the introduction of lightweight refractories, or even rings made hollow, platinum-clad gathering rings may still be employed and the writer is currently engaged on their introduction into a Continental glass factory. Being protected from glass attack, the rings can be of any shape desired and the most convenient shape is a round section.
Similarly, floaters in a sheet glass tank, operating as they do at a temperature somewhat lower than 1400°C, can be made, and are made, tubular in section. The “lip” of a plate glass tank is again almost an ideal shape for covering with platinum; it is only necessary to ensure that the drag of the glass over the platinum surface is not likely to tear the platinum cladding. In the Fourcault and Pittsburgh processes of sheet glass drawing it is not unlikely that the surfaces of the debiteuse will be platinum-clad in the future and so prevent many of the draw-marks which result from a damaged edge of the debiteuse causing it to be necessary to replace it.
The technique of coating refractories with a film of platinum produced by painting with a suspension or solution of platinum or one of its compounds in an organic vehicle and firing on is often spoken of but rarely tried, and for very good reasons. Such metallising preparations have their proper place in yielding a film of platinum or other metals on substrates such as glass, ceramic or mica for use at low temperatures in the electronics industry, but it is extremely difficult to produce on a refractory surface a layer of platinum in close adherence to the refractory and of continuous sheet-like condition that can withstand the high temperatures involved in the handling of molten glass. The differences in thermal expansion alone will sooner or later introduce cracks into the surface (if they are not there already) and there is always the problem of a cracked refractory piece which might render the whole unit useless. Attractive though they might sound, platinised refractories are best avoided in the glass industry, at least in the present state of the art, in favour of platinum cladding. Platinising must inevitably leave some porosity and this defeats the object of covering the refractory completely with a layer of platinum.
In concluding this short survey of the benefits to be obtained from a little redesign of furnace refractories to enable full advantage to be taken of platinum cladding, the writer would repeat a statement that he made in this journal in 1960 and that has never been challenged. It is this. “It is a significant fact that the use of these so-called precious metals, admittedly high in initial cost, has always resulted in the production of the particular types of glass to which they have been applied at a more economic price. This is due entirely to the production of a purer glass, or a glass with greater freedom from defects from whatever cause.”