THE CHOICE OF PERFUMES FOR PRESSURIZED PRODUCTS 391 They may be acting as catalysts or inhibitors, and their own action may be influenced by another ingredient of the product 7. Equally, the effect of perfumes on plastic and rubber components cannot be ignored. CHEMICAL COMPATIBILITY Familiar reactions of perfumes with products under normal circumstances may be modified under aerosol conditions. Proneness to oxidation and the effects of light, for example, may be reduced, but on the other hand certain perfumes may react with propellants or with the metallic, and other, products of interactions taking place in the aerosol. The commonly used propellants are comparatively stable, the most reactive being trichlorofluoromethane. If water is present, the latter can be hydrolysed catalytically with the liberation of acid. Furthermore, it has been claimed that under certain conditions ethyl alcohol will react with this propellant TM. A variety of isomerisations, cyclisations, etc., of perfumery materials can take place when the pH is thus reduced. Witjens •ø has shown ho•v ethyl alcohol is dehydrogenated to acetaldehyde under the influence of acid and metal or metal halides, especially if water and oxygen are present. The acetaldehyde would be capable of entering into a variety of reactions with perfumery chemicals under these conditions. It is even conceivable that certain perfume ingredients could facilitate the primary reactions, for example by acting as a free radical initiator or a hydrogen acceptor. The ideal is to formulate and pack aerosol products in such a way that the drastic initial changes are avoided. In this case, the questions to be considered when selecting a perfume are less complex. However, as a compromise is often the nearest approach to the ideal, the perfumery chemist must take into account all relevant factors when making recommendations if odour changes, discolouration and formation o[ insoluble residues are to be minimised. PHYSICAL COMPATIBILITY The effects of perfumes on viscosity and emulsion stability of some non- pressurized products are well reco•ised. They are equally important in the field of aerosols. Detrimental effects could be the precipitation o[ wax from pressurized polish or the poor foaming characteristics of a shaving cream. TESTING It follows that adequate testing of perfumes in pressurized products is essential. Countless evaluations of the behaviour of individual ingredients in various aerosol systems have been made TM, and continue to be carried out. The results of these are extremely valuable from both the practical and theoretical aspects but they do not obviate the necessity for testing

392 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS finished compounds, in which each ingredient can modify the behaviour of others. For example, in the first of the papers just cited n, the conclusion is reached that the odour of hydroxycitronellal is changed in a typical aerosol cologne, but we know that this chemical is used with great satis- faction in many perfume compounds for this application. Similarly, solu- bility determinations on single ingredients may have no bearing on their behaviour in particular concentrations in compounds. In the second paper •2, it is shown that coumarin is insoluble at 1% in propellants, but this by no means precludes the effective use of this material in practice. Nothh•g can replace the adequate testing of a perfume compound in the precise formu- lation and pack under normal storage conditions. However, where time is limited, accelerated tests at elevated temperatures can be of service if interpreted on the basis of considerable previous experience. A typical testing procedure in the case of a perfumed pressure pack would involve the preparation of twelve packs exactly as proposed for marketing. Six would be placed in the incubator at 130øF, and six stored on the shelf at room temperature. At suitable intervals covering a period of one year, one shelf container and one container from the incubator would be examined for odour change, for corrosive effects on all the internal metallic parts and for swelling effects on the plastic components, also for any evidence of deposition of insoluble matter and for changes in the colour, pH, etc., of the solution. Odour comparisons would be made with control unpressurized solutions kept under the same conditions and at appropriate intervals with freshly prepared pressure packs. In order to supplement these generalisations, some specific points of importance in the perfuming of individual groups of aerosol products are described below-. AIR FRESI•4ERS The function of an aerosol air freshener is to eliminate undesirable odours from the atmosphere to be treated. Some of the theories of space odour removal have been described in detail by Gee •3 and Johnsen •4. In brief, this can be achieved by masking the unpleasant odour with one which is more acceptable, by blending another odour with it to give a neutral effect, by temporarily anaesthetizing the sense of smell, or bv chemical combination with the malodorous molecules to give less offensive complexes. The part played by perfumes is obviously very important. The design of perfume compounds to participate in these various mechanisms of odour removal, including the entering into polymerisation and condensation reactions with a variety of active groupings, is a subject of constant research and experiment. Systems have been designed for assessing the effect of