STABILITY TESTING OF LOW-PRESSURE AEROSOLS 319 be evaluated in terms of the feasi- bility of filling, using special equip- ment. In many instances it is possible to reformulate to obtain a concentrate which will meet both the compatibility and the low- temperature stability requirements. Although compatibility is of prime importance in formulation it is imperative that actual con- ditions in the can be simulated for corrosion studies. In order to accomplish this a pressure flask is set up containing the contents and all parts of the can or container that come in contact with the formula. These tests are made both in the liquid and the vapor phase. The flasks are incubated at 130øF and observations made on the conditions of the can parts and liquid components over a period of at least three months if negative results are not indicated sooner. Similar tests are run at room tem- perature. Proper precaution must be taken in performing these ex- periments since cans and glassware may be shattered due to improper compounding or apparatus failure. Safety goggles or protective shields provide suitable protection and do not hamper the efforts of the tech- nician. Let us assume now that our prod- uct is satisfactory in so far as compatibility and freezing tests are concerned. The aging tests are under way. To complete stability testing of this product we must make up the complete formulation in the container. According to I.C.C. regulations, cans shall not be liquid full at 130øF. Knowing the water overflow capacity of our container and the specific gravity of the concentrate and propellant at room temperature and at 130øF, we can then determine the practical safe net contents. In determining this figure it is advisable to add 4 per cent to the safety margin set up by the I.C.C. since we may encounter a 4 per cent accumulative air during our filling operation. The greatest danger in overfilling is encountered when the container is liquid full at 130øF. When this condition arises the normal vapor pressure of the system no longer exists and we have to contend with the hydrostatic pressure produced by the expansion of the liquid. Two sets of cans are prepared based on the compatibility test and the content calculations. One set is incubated at 130øF along with the pressure flasks and the other set is retained at room temperature. Both sets of cans are subjected to an intermittent-use test so as to determine the effect of contact of the material with the working parts of the valve. The cans are prepared in the' laboratory by weighing the con- centrate into the can and placing it in the deepfreeze or similar cooling device. Special apparatus is re- quired to handle the propellant. This consists of a copper coil emersed in a bath of acetone and dry ice. One end of the coil is tapped and threaded so that it may be attached to the inverted cylinder. The other end of the

320 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS coil coming out of the bath has a needle valve attachment so that the flow of propellant out of the cylinder through the cooling coils may be controlled. When the can and contents have been cooled to below 0øF the propellant is added until the desired weight is obtained. In this manner the propellants may be handled with a minimum loss. The can is then seamed and placed in hot water bath to be tested. Visual observations are made weekly on the pressure flasks to determine how the component parts are holding up. Any adverse action can be readily determined and noted. In the regular cans corro- sion is usually determined by spray- ing and noting the change in odor or pressure. These cans are also cooled in the deepfreeze and opened te determine the extent cf the damage of the can and valve parts or any change that may have taken place in the contents. In the case of corrosion or decomposition there is usually a change in color of the contenl•'s or a definite mal odor of the contents coupled with very serious attack on the can parts or the valve parts. The 130 ø test, like any other accelerated test, is not completely reliable. Before a final decision is made, therefore, the results at room temperature must be taken into consideration. Where the 130 ø test has been satisfactory the room temperature tests have also been satisfactory. In some cases where the 130 ø test has been unsatis- factory after a period of one month, we have found that room tempera- ture samples for a period of one year have been completely satis- factory. In these cases an 18- month test at 100øF must be satisfactory unless formula revision can be made. It is also true that actual produc- tion conditions cannot be fulfilled in the laboratory. It is therefore a nec- essary part of stability testing to make a production run of between 1000 and 5000 cans on the pro- duction line before a product can be given a go signal for the market. The problem of perfuming pres- sure products is particularly im- portant. It has been found that perfume compounds will stand up well in the concentrate in contact with glass but will show marked evidence of decomposition and sub- sequent odor change when used in the container. Many of the per- fume compounds themselves are a cause of corrosion. A great deal of work has been done on this problem by the essential oil suppliers and pre-tested materials are now available. However, it is abso- lutely necessary to test every per- fume compound under the exact conditions that it is to be used. To complete the picture for a successful product the evaluation of stability tests is not final. It is necessary that the concentrate formulas be controlled very accu- rately for quality and composition. In waterless products it is specially important to check moisture con- tent, free-acidity and specific grav- ity. In water base products, pH and specific gravity play an impor-