COMPRESSED GASES AS PROPELLANTS FOR COSMETIC PRODUCTS 227 the history of these propellants and refers to a patent issued as long ago as 18628. More recently, the use of nitrogen for dispensing viscous food pro- ducts was recommended by Pyenson• and other publications •,',7 also foreshadowed the use of such propellants. The salient difference between compressed and the hitherto convention- ally used liquefied propellants is the nature of the pressure characteristics. With the latter type of propellant, the pressure within the dispenser remains constant until all the product is exhausted whereas with the former, the pressure decreases as the product is dispensed. The reasons for the pressure decrease with compressed gases have been described in other publications 8 9 and it is therefore not proposed to deal with this aspect in any detail. Suffice it to state that immediately after charging a dispenser with gaseous propel- lant, said propellant will exert a unit pressure and occupy a pre-determined ullage*. By the time sufficient product has been dispensed to make the headspace* equal to twice the ullage, the same quantity of gas will exert only half of its original pressure. This gradual drop in pressure, more marked with gases that are insoluble in the product is the chief disadvantage encountered. Often, the pressure remaining when the dispenser is nearly empty will be insufficient to expel the whole of the product, leaving a product residue of ten or more per cent. This pressure drop is the reason why lique- fied propellants have been preferred hitherto to compressed propellants for dispensing liquid products such as room deodorants, hair lacquer, etc. Although the particle size of sprays propelled by compressed gases is coarser, this can be largely overcome by resorting to break-up spray valves which, as is implied by their name, cause a mechanical break-up of the spray particles. Until the announcement already referred to •, compressed gases wer e only employed for dispensing food products such as whipped cream •ø, where a nitrous oxide/carbon dioxide mixture is often employed, flavoured drink con- centrates • and for fire extinguishers based on carbon tetrachloride •. For the last two products, carbon dioxide is employed. In all these examples, the propellant is soluble in the product and the gradual drop in pressure is not quite so marked as with the insoluble compressed propellants. (Incidentally, none of the food products just mentioned is at present marketed in Great Britain.) Reference was made on a previous occasion •8, in the light of the knowledge then prevailing, to the manner in which aqueous products could be packed when employing liquefied propellants. The possibility of packing toothpaste as a foam pack was mentioned, having in mind a product then marketed in Switzerland. This product does, however, bear little resemblance to a con- ventional toothpaste. In order to dispense a toothpaste in its original, * Ullage •?: Volume in dispenser not filled with liquid content immediately after packing. Headspace •?: Volume in dispenser not filled with liquid content.

228 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS non-aerated form it is necessary to employ a propellant which, to all intent and purposes, is insoluble in the product. The only liquefied propellant which appears to fulfil this requirement is octafluorocyclobutane (Freon C318) •, which is, however, not yet approved by the U.S. Food and Drug Administration for use in edible products. Nitrogen has been found to be suitable •,5,6,7 for dispensing toothpaste in non-aerated form and, although it is generally referred to as an "insoluble" propellant, one must •ot lose sight of the fact that even the "insoluble" propellants are often soluble to a small degree. For example, the solubility coefficient of nitrogen in water, at a pressure of one atmosphere and at a temperature of 25 ø C. is 0.0143 •5 and in cottonseed oil, at 24-25 ø C., it is 0.061 •5 The marketing of toothpaste and similar products, pressurized in this manner, did not become a practical proposition until a valve capable of permitting the flow of viscous products became commercially available •. Quality Factors A number of factors must be determined for each product in order to ensure the ultimate success of the marketed package. Amongst these are viscosity, ullage and initial pressure. Viscosity is undoubtedly first and foremost, but is at the same time closely linked with the other two mentioned factors. In spite of any thoughts to the contrary, the viscosity of the product must be decreased when changing from the tube to the pressurized dispenser. In general terms, the viscosity of the product should be such that it is too thin to be squeezed from a tube and too thick to be poured from a bottle. With a toothpaste, for instance, it is naturally necessary to ensure that the paste does not run off the brush. The viscosity of the product has a direct bearing on the amount of residue left in the dispenser. If the viscosity is too high, then too much paste may adhere to the walls of the dispenser. The flow from the walls to the lower end of the dip tube will also be sluggish and should the dispenser have been stored on its side or upside down prior to use, then propellant rather than product might be expelled. This phenomenon is often referred to as cavitation. It has been suggested •6 that the interior of the dispenser might be coated with a lubricant in order to avoid cavitation. Should this be resorted to, then it is necessary to ensure that the lubricant is unaffected by any of the ingredients of the product. The suggestion of using flanged dip tubes to avoid cavitation does not appear to have been successful, as experience has shown this phenomenon to be rather a function of the concentrate formulation itself. The drop in pressure as the dispenser is emptied, as already referred to, will effect the amount of product residue. Ullage and initial gas pressure must therefore be very carefully correlated. The pressure required to expel