TESTING DEODORANTS WITH CHLOROPHYLL AND DERIVATIVES 47 measurements of breath odors from alliaceous substances. As applied to the objectives of the investigation under discussion, the method includes the following con- ' secutive steps: (1) a stream of atmospheric air is drawn through a preliminary chain of washing bottles in order to remove interfering vola- tile substances, (2) after passing through subsequent drying columns, the air is bubbled through the sources of the odors in gas-washing bottles heated on steam baths, (3) then it passes over the layer of dried, pulverized chlorophyllins, (4) through a tower of pellets of potas- sium hydroxide, (5) then, through a U tube containing iodine pentoxide immersed in an oil bath at 180øC., and finally (6) into a solution of potassium iodide. The total vol- ume of air drawn through the sys- tem is recorded by a gasometer at the terminal end of the system. The three fundamental reactions involved in this procedure are: (a) the diallyl sulfide liberates iodine from the iodine pentoxide, (b) the liberated iodine is carried over into the solution of potassium iodide in which it dissolves, and (c) the amount of iodine set free from the iodine pentoxide is determined by titration of the potassium iodide solution with a standard solution of sodium thiosulfate, using soluble starch as an indicator. The neces- sity of running both repeated series of preliminary clearances of the apparatus and of controls with sources of odor but no chlorophyll eliminate this method as a routine procedure for evaluations of deo- dorant actions but, in the hands of well trained and experienced an- alysts, it has been found to be a satisfactory procedure for exploring some of the chemical reactions con- tributing to deodorizations. In tests carried out on each source of odor, the series of experi- TABLE 5--THE COMPARATIVE EFFECTS ON VOLATILE• ODOROUS CONSTITUENTS OF GARLIC RE- SULTING FROM CONTACTS WITH DRIED CHLOROPHYLLINS OK WITH CHLOROPHYLLINS IN SOLU- TION Average Per- Method of Use Average Titer t centage Re- Series of Ex- Source of of Dried Chloro- with Na2S2Oa duction of Di- periments Odors phyllins c.c. of 0.005 N allyl Sulfide I Aqueous ex- tract of garlic II Diallyl sulfide 0 21.0 .. In contact* with 8.1 61 vapors Added to source of 12.2 42 odors 0 24.5 In contact with 14.7 3•(5 vapors Added to source of 10.3 58 odors * 300 mg. of dried water-soluble chlorophyllins. t All titers reported are average'results of three experiments.

48 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS ments was divided into two groups. In one group of experiments, 300 rag. of the dried preparation of the chlorophyllins were spread over the bottom of the gas-washing bottle through which the odorous vapors were drawn. In this respect, the technique was similar to that adopted in the experiments with butyric acid which are presented in Table 4. However, in the second group of experiments, 300 mg. of the preparation of chlorophyllins were added directly to the source of the odor. Air was drawn through the system at a uniform rate of 200 c.c. per minute for thirty consecutive minutes. Diallyl sulfide, or as it is some- times called allyl sulfide, is an un- saturated compound containing two double bonds. It reacts readily with iodine pentoxide and an end result of this reaction is the libera- tion of free iodine. Table 5 presents a summary of this series of experiments. In the tests with the aqueous extract of garlic, the dried chlorophyllins in con- tact with the vapor phase effected greater reductions in the volatile odorous compound than did an equivalent amount of the same preparation added to the liquid phase of the source of the odor. This result is in agreement with the findings of several comparable series of experiments all of which indicate that dried preparations of either natural chlorophyll or chlorophyl- lins derived from it exhibit greater capacities to deodorize gases than their abilities to eradicate the odors of liquids in which they were sus- pended or dissolved. However, the results of the experi- ments with diallyl sulfide appear 'to be exceptions to this general state- ment. In tests carried out on this source of odor, the chlorophyll preparation added to the liquid phase was more effective in reducing intensities of odor than it was as a solid in contact with the vapor phase. The apparent explanation of this exceptional result is the fact that rapid volatilization of di- allyl sulfide in the stream of air during the first five to ten minutes of the test period (thirty minutes) resulted in the delivery of practi- cally all of the odorous vapors into the bottle containing the chloro- phyll preparation within a period which was less than the time requisite for deodorization by the chlorophyllins. On the other hand, addition of the chlorophyll prepara- tion to the liquid phase of the source of the odor provided a longer interval for deodorizing actions of the chlorophyllins. The deodorant capacities of chlo- rophyll may be influenced to a marked degree by the balance between bases and acids in liquid sources of odors. The group of experiments which are summarized in Table 6 has been selected as an instructive illustration of the com- parative effects of varying levels of pH upon the efficiency of the prep- aration of sodium-copper chloro- phyllins as a deodorizer. A 0.20 per cent mixture of benzyt mercaptan with water was utilized