PROTECTIVE LACQUER SYSTEMS i½0R ALUMINIUM CONTAINERS to hydroxyl groups 2e -+- O -+- H•O----•2(OH)- may, in the case of standard test pieces, mean very little. However, in the case of a container where the wall thickness rarely exceeds .015", and can be as low as .005", pitting may be synonymous with perforation. Even where this does not occur and the corrosion proceeds sluggishly, the cathodic areas become rich in OH groups, corresponding in certain cases to the formation of free alkali, until such time• as the product immediate to the cathode is alkaline enough to attack the metal directly, with rapid hydrogen evolution. These points should be borne in mind when consulting the literature concerning corrosion and corrosion resistance of aluminium, which is nearly always concerned with data derived from test pieces of solid metal. A vast range of products packed in aluminJure are "borderline" cases, and very minor changes in either pack or product can and do give rise to failures by corrosion. There are many examples where a product in unprotected aluminium has satisfied stringent packaging and test market trials, but which, in full scale distribution, give a percentage of failures by corrosion. The causes of this are difficult to detect, because examination very often shows that both the pack and the product are, on the face of it, unchanged. Minor changes have, however, taken place and as these are very often physical rather than chemical, it is not surprising that they are overlooked. Amongst the very many minor factors which can trigger off corrosion in the "borderline" products are changes associated with the surface of, and the oxide film present on, the aluminium container. For it is in fact this oxide film which mainly gives aluminium its corrosion resistance. The type and condition of the oxide film depend on the heat-treatment history of the container, and also to some extent on the period which the container has been standing empty. It follows, therefore, that changes in the heat-treat- ment--which are not readily detectable--profoundly affect the subsequent corrosion resistance. Again, maximum corrosion resistance presupposes minimum mechanical damage to the oxide film and the actual metal surface. Thus a deeply scratched container will corrode whereas a container identical in other respects will not corrode. Cases are known where deep scratch marks have given rise to failure of containers by corrosion. It is easy to demonstrate experimentally that scratched areas are anodic to surrounding smooth areas of metal. Both the points mentioned, which are two examples only, of a number of minor factors affecting the serviceability of a tube or can, have very direct practical significance. Nowadays, in the field of tube and can making, as in many industries, there is a big changeover to automatic or semi-automatic line production, which has meant that drastic reductions of processing times for every stage of manufacture have been made. In the tube industry, annealing times have been cut down from periods of 10 minutes to times of the order of 70 to

JOURNAL OF THE SOCIETY Oi•' COSMETIC (2HEMISTS .80 seconds, with consequent temperature variations, whilst the tubes them- selves are handled more violently from the mechanical point of view, i.e., speed of feeding on to and withdrawal from processing spindles, with a resultant increase in inner surface roughness. From this it can be seen that these minor factors are capable of ruining an otherwise satisfactory packaging line, and often internal lacquering is resorted to--the so-called "standard" internal lacquering of can and tubes--in order to cancel out these variables. Very many products, •vhich in the main would perform satisfactorily in unprotected metal if optimum conditions were satisfied, are packed in lacquered containers as a precaution only, but it is precisely this that has, from time to time, given the standard lacquered lines an evil name. Many instances have been experienced •vhere a customer has assumed that standard lacquering is a maximum protective system, and that products are marketed satisfactorily in this type of pack--which they are--and has then made the further assumption that his o•vn particular product, which is in fact quite corrosive to aluminium, would be satisfactory in a similar container, with disastrous results. Standard lacquered containers are not intended for such applications, and it may be of interest to know that until recently, it was standard practice in a number of continental factories to lacquer all aluminium tube pro- duction, not for the protection afforded, but because it is almost impossible to feed dry annealed aluminium on to steel printing or enamelling spindles, whereas lacquered aluminium is satisfactory in this respect. The protective lacquers in use today are invariably based on one of the following types of resins, these having largely superseded the older phenolic resins which were used extensively in the container field. TYPES OF LACQUER 1. Epoxy resin based, i.e. Epikote or Araldite resins cured by amines, phenols, etc. 2. Epoxy esters--esters produced by reacting epoxy resins with drying and non-drying oil fatty acids. 3. Vinyls--polymers or co-polymers based on vinyl chloride, e.g., vinyl chloride/vinyl acetate, vinyl chloride/vinyl acetate/vinyl alcohol, vinylidene chloride/vinyl chloride, etc. Lacquers based on (1) and (2) are generally referred to as high stoving lacquers, and the others as low stoving. All of these have excellent chemical resistance, mechanical flexibility and adhesion. Providing the inherent