HIGH DENSITY POLYETHYLENE BOTTLES 201 home. Wide mouth bottles and jars can now be produced at reasonable weights with neck finishes stiff enough to give good sealing characteristics with standard closures. This stiffness also alters some of the functional properties which we have been accustomed to finding in plastic bottles. A high density bottle blown at the same weight as a normal polyethylene spray bottle is difficult to squeeze. If we make the bottle light weight to improve the spray prop- erties, we may find that we have problems in providing a good fit between the spray plug and the bottle neck and, in addition, that the recovery time of the bottle after squeezing is completely different. It is possible to pro- duce good spray packages with high density polyethylene but the bot- tle shape and weight must be carefully selected to insure proper function- ing. Most of the high density bottles produced now are intended for non- spray end uses. They are carrier-type containers for liquid detergents and shampoos, for household cleaning products and for basic chemicals. They serve as dispensers for lotions and creams and as drop-by-drop ap- plicators of pharmaceuticals. Some are used as laboratory ware and mil- lions have been sold as baby nurser bottles. The nurser bottles, laboratory ware and pharmaceutical packages illustrate another important property of high density polyethylenes. All of these bottles can be steam sterilized by regular autoclave techniques at 15 pounds steam pressure. One word of caution though autoclaving should be carried out with the containers unsealed or empty. The strength of the polyethylene is considerably decreased at high temperatures and any vola- tile materials sealed inside may cause rupture of the bottles. The question of bottle clarity is often an important one with users of polyethylene containers. The higher crystallinity of the high density resins results in a lower degree of clarity but product fill levels are visible through unpigmented high density bottles. All of the polyethylenes transmit ultraviolet light. For light sensitive products or where color is desired for appear.ance considerations, pigmented bottles can be provided in a wide variety of colors. To this point we have stopped for only a few comments on each of the bottle properties mentioned, but there are three more items which I think you may be interested in looking at a little more closely. These three are resistance to impact, to stress corrosion cracking and to permeation. Let us consider them in that order. BOTTLE IMPACT PROPERTIES We have already seen that the high density polyethylenes give lower Izod impact values than regular polyethylene. But what can we do to determine the impact resistance of a particular bottle? We have found

202 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS that a simple drop test of filled bottles can give valuable information. For preliminary evaluation, a random selection of bottles are filled with 70øF. water to nominal capacity and sealed. Drops are made onto an asphalt tile over concrete floor beginning at 3 feet. Each bottle is dropped from this height first on the bottom corner, second on the side and finally on the fiat bottom. The same bottles then go through the same three drops at 4 feet, 5 feet and 6 feet or until failure. Thus, any bottle which passes the full test has been submitted to 12 drops. Bottles are examined carefully at each drop level by squeezing and checking for pinholes, splits or obvious leaks. Any water leakage from the bottle is considered a fail- ure. This test, of course, is a tedious chore but it has demonstrated the importance of three critical points--bottle design, machine blowing condi- tions and selection of proper polyethylene resins. As would be expected, small bottles or bottles with enough polyethylene to give a fairly stiff squeeze usually do not present problems of impact resistance. Problems can occur with larger bottles such as 16 ounce or 32 ounce sizes which have been lightweighted for reasons of economy. We have found that in de- signing these lightweight bottles it is best to round off any sharp corners and to avoid designs which will cause extreme variations in wall thickness. With poor designs, adjacent thick and thin sections cool at different rates thus giving slightly different crystalline development and setting up internal stresses in the bottle. These areas then fail under strain. This effect must also be considered in determining blowing conditions. Extrusion temperatures and cooling rates 'must be closely controlled to achieve the optimum results. RESISTANCE TO STRESS CORROSION CRACKING The second of the three properties we are to examine is resistance to stress corrosion cracking. The first significant work on this phenomenon was done by Bell Laboratories several years ago in connection with cable insulation. A test procedure for determining crack resistance of various polyethylenes was developed and is still used in one form or another throughout the plastics industry. We have converted the test slightly, using a standard polyethylene to test the products which go into the bottles for activity as cracking agents. With high density polyethylene, an addi- tional test of the blown bottle has been devised to evaluate the finished con- tainer for crack resistance. This consists of filling the bottles with a 33 per cent solution of Igepal CO which is a wetting agent and an active cracking agent. Bottles are then sealed, placed in a pan of Igepal and con- ditioned at 140øF., for seven days, with daily examination for evidence of cracking. Failures of poor quality containers will normally occur within the first two or three days but good bottles will pass with ease. As