94
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y .
MacMillan & Co., 1899-says that liquid CO, is soluble in all proportions in alcohol, ether, naphtha, turpentine and CS,. We believe that D'Arsonval,' a French investigator writing a t the request of M. Henri Moissan, first called general attention to the efficiency for experimental refrigeration of solutions of CO, or C,H, in ace tone. Liquid acetylene forms snow which also dissolves freely in acetone, and yields a solution which reaches -112' C. Solutions of CO, snow in acetone can be made to reach this temperature if placed under vacuum, or if evaporation is hastened b y the passage of refrigerated air. Similarly the temperature of acetylene snow in acetone can be further reduced. As a n illustration of the rapidity of the action of the CO, acetone, i t caused 430 grams of mercury to freeze solid in 2 min. and 40 seconds. We placed a n open beaker containing i t in a larger open beaker of the refrigerant. H g freezes a t -39' C. The refrigerant weighed 193 grams and contained about 35 per cent. CO,. I t s temperature rose from -63' to -43' during the freezing of the mercury. The approximate amounts of CO, snow and acetone necessary to produce certain temperatures are as follows: 150 gm. acetone a t 23' C. 6 gm. CO, = - 5' C. 11 gm. CO, = -20' C. 24 gm. CO, = -38' C. 45 gm. CO, = -54O C. 7 7 gm. CO, = -6" C. CO, in excess = -78' C.
+ + + + +
I 14
gm. petrolic ether of 86 ' B. at 2 2 O C. : 5 gm. CO, = -22' IO gm. CO, = -36' 15 gm. CO, = -55' 30 gm. CO, = -69'
+ + + +
C. C. C. C. It appears from this that petrolic ether reaches a low temperature more quickly. This is probably because i t is more volatile. It foams more when the snow is added to it. Acetone, however, is, as D'Arsonval says, the best of the solvents for general use. Most of the others give deposits which would clog cooling jackets. If absolute freedom from fire risk is desired, CC1, or CHCl, and CO, can be used for a limited range. I wish to give credit to my assistants, Messrs. Craver, Johnson, Thompson and Rurris, for their work with me in this matter. SUMMARY, ( I ) CO, snow dissolved in acetone or other volatile solvent will quickly chill substances to -75' C. or to any intermediate temperature (D'Arsonval). (2) The snow is prepared by allowing the liquid 1 "Production and Maintenance of Low Temperatures," D'Arsonval in Comgfes rendus, Vol. 133, 980 (1901) ~
Mar., 1910
CO, to blow off through canvas or flannel bags, according to the well-known method. (3) This convenient method of refrigeration is of immediate use for many investigations, and is likely to be of service for purification of chemical products and for analytical separations. ASSAY O F MEDICINAL PLASTERS. FREDERICK B KILNER. Received July 31, 1909.
By
Comparatively little literature has appeared in scientific publications in respect to the analysis of assay of medicinal plasters made with a n India rubber base. The enactment of the Federal Food and Drugs Act and the enactment of similar laws in the various states has brought the subject to the attention of the pharmaceutical chemist, and a t the present time the subject is one of moment to both the medical and pharmaceutical professions. Prior to the issuance of the last revision of the pharmacopoeia no authoritative process of assay of any medicinal plaster appeared in that work. I n this revision a process was given for the assay for mydriatic alkaloids in belladonna plasters made with rubber base. Those who have given attention to the subject will no doubt agree with the statement that has been made, that it is a very difficult matter to assay some kinds of medicinal plasters made with a rubber base. I t is a well-known fact that for some such plasters no method is known for the assay of the drugs contained therein, for the reason that the drug used has no alkaloid or no inorgaP ic substance capable of definite measurement. I n other varieties of these plasters the presence in the rubber base of resinous or other matters is such as to confuse the results and make them of doubtful value. I n the laboratory of the writer considerable work has been done in the assay of this class of preparations, and in our work we possess some advantages over any outside laboratory, in that a s the laboratory is connected with the manufacturing department, i t is a t once known exactly what has been put into a plaster, and it is only necessary to provide methods of assay that will give the results looked for. The methods and processes which I herewith present are simply those which have been worked out in the laboratory, namely as a check to manufacturing processes. They are presented for what they may be worth, with the hope that other workers may have a n opportunity to try them and thus bring out information of value to those interested. For belladonna plasters we used the assay method of the eighth revision of the United States Pharmacopoeia. This process, with some slight modifications, has been entirely satisfactory in our laboratory, where assays, running perhaps into thousands, have been performed. I t has also had the benefit of an
K I L M E R ON ASSAY OF MEDICINAL PLASTERS.
.
extended check test, in that in this laboratory i t is our practice to first assay the.drug used in the manufacture of the plaster and then the solid extract made from the drug; the mass itself is made up of the assayed extract taken in a n amount which will give in the finished product the desired percentage of alkaloid required in the plaster, and by this method of plaster assay the results have shown that in mixing several hundred pounds the assayed plaster corresponds with the calculated amount often to the second and third decimal. Salicylic Acid Plaster.-The method which we use for the analysis of salicylic acid plaster is a colorimetric method with ferric chloride. The depth of color imparted to a ferric chloride solution by a measured amount of a solution of the plaster is compared with that of a measured quantity of a standard solution of salicylic acid of known strength. Colorimetric methods are always subject to some variations, chiefly, on account of errors of eyesight, as well as to the presence of resinous materials in rubber plaster masses; no other method has been found to be satisfactory and we have found that by careful manipulation, the color comparison method has given sufficiently satisfactory results. Plaster Solution.-LVeigh out accurately about j grams of the plaster cut into rather small strips. Place on the table two beakers of about I j o cc. capacity each. Designate them as No. I and KO. 2. Place the weighed strips of plaster in No. I . Add to it 50 cc. chloroform. Stir gently until all compound is removed from the plaster-cloth and in solution. Pour the sirupy solution into beaker No. 2 . Add to this in No. 2 beaker 40 cc. ordinary 94 per cent. alcohol, stir thoroughly to precipitate and coagulate the rubber and allow i t to stand. Pour off all possible of the liquid into a glass-stoppered, graduated 2 j o cc. flask. The rubber should be worked up into a compact mass so that no particles are carried over when the liquid is poured off, and all possible liquid should be pressed out of the mass with a glass rod. To the plaster-cloth in beaker No. I add 2 5 cc. chloroform. Stir carefully and thoroughly until all remaining plaster-mass is dissolved from the cloth and sides of the beaker. Pour off again into beaker No. 2 , which contains the precipitated rubber. m’ork up with a glass rod until all of the rubber mass is again in solution in the chloroform. Now reprecipitate the rubber from this solution with 2 0 cc. alcohol, working up with rod and pouring off as before, mixing the fluid with the first portion in the flask. Once again wash the cloth in beaker No. I with 2 j cc. chloroform. Pour off into No. 2 beaker, dissolving again the rubber mass in it. Reprecipitate the rubber from it with 2 0 cc. alcohol a s before, pour off and mix the decanted fluid with the other two portions in the flask which now contains all of the sali-
95
cylic acid in solution. Fill the flask up to the 2 5 0 cc. mark with alcohol. Remove the cloth, which should now be white and clean, from beaker No. I , allow to dry spontaneously and weigh. Subtract its weight from the total weight of plaster used, thus ascertaining the weight of plaster compound taken for assay. Standard Salicylic Acid Solution.-Weigh out exactly 0 . j gram pure salicylic acid and dissolve i t in j o per cent. alcohol. Transfer to a 500-cc. glassstoppered, graduated flask, rinse out the vessel, in which the acid was dissolved, with repeated portions of j o per cent. alcohol, adding each portion to the solution in the flask. Make up to the 500 cc. mark with j o per cent. alcohol, shake thoroughly. I cc. of this solution contains o OOI gram salicylic acid. Analisis by Color Co?izparison.-For this work we use two large test tubes of similar internal diame~ by 6 inches long). Any pair of glass ter ( I ~ / ’inches cylinders or tubes will suffice, but small diameter test tubes do not give sufficient thickness of solution to secure enough depth of color with transmitted light. Place these tubes side by side. In a beaker place IOO cc. distilled water, to which add one drop ferric chloride solution, U. S. P. Stir and pour into each of the test tubes j o cc. of this. Designate them No. I and No. 2. Now add to No. I tube from burette sufficient of the standard salicylic acid solution to give a strong, clear, wine color. Stir with a glass rod after each addition of the acid solution. Multiply the number of cc. used by 0.001which will give the weight in grams of salicylic acid used in tube No. Ithe standard. Now add to tube No. 2 from another burette sufficient of the plaster solution exactly to match the color obtained in tube No. I . This plaster solution must be added a little at a time and the solution in the test tube well stirred with a glass rod after each addition. When the matching point is nearly reached i t may be necessary to filter off the contents of test tube No. 2 . Clean the tube and replace the fluid-proceeding thereafter to add the plaster solution, a drop at a time. The reason for doing this is on account of the small amount of resinous matter separated from the solution, which may cloud the mixture in No. 2 test tube and interfere with the color judgment. By closing one eye and observing the colors while holding the tubes side b y side between the eye and a window, the colors can be matched very closely. It is obvious that the quantity of plaster solution used contains the same amount of salicylic acid as was used in tube No. I . T h a t quantity is already determined by means of the standard solution. Therefore 2 j o (total amount of plaster solution) divided by the number of cc. plaster solution used in tube No. 2 and the result multiplied by grams acid found to have been used in tube No. I gives the weight of acid in the plaster compound.
.
96
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y .
Multiply the weight of the salicylic acid found by IOO and divide the result by the weight of the plaster compound used. The result is the per cent. salicylic acid present. This method has been used chiefly with plasters containing 2 0 , ~ j and , 30 per cent. of salicylic acid, When very small percentages are present i t will of course necessitate the addition of a much larger quantity of plaster solution to tube No. 2 in order to secure the same color as No. I tube and, consequently, the color of tube No. z will be unduly diluted. If by trial this is found to be the case, it will be necessary to prepare a weaker standard solution or else a more concentrated plaster solution. This matter can readily be determined by trial. In making these assays it is well to make three of each and take the average. By doing this we have been able to get satisfactory results, the natural variations of color judgment balancing one another and the result being nearly correct in the average of the three determinations from one lot of plaster s o h tion. Mercurial Plaster.-The requirement of the U. S. P. is that it shall contain 30 per cent. metallic mercury. Method.-Weigh out about j grams of the plaster, cut in rather small strips. Place in a beaker and add about j o cc. benzol. Stir well for some time to soften and dissolve the compound. Pour off into a beaker of about 300 cc. capacity, allowing the cloth to remain in the first beaker. To the cloth in the first beaker add repeated portions of about 50 cc. benzol, stirring well and pouring off each time into the second beaker until the mixed solutions amount to about 2 0 0 cc. The clean cloth is now allowed to dry in the air, weighed, and its weight subtracted from the weight of the plaster taken. This gives the weight of the compound used. The mixed solutions of the compound, measuring about 2 0 0 cc., are now well stirred and then allowed to stand covered, in a tall beaker, until the gray metallic mercury has settled to the bottom. This usually is accomplished in about twenty-four hours. Chloroform can be substituted for benzol in this dissolving operation, but benzol is used in our work on account of its relative lightness, because of which the mercury settles more quickly. When the gray mercury has settled out pour off carefully the supernatant benzol, which contains the rubber and resins in solution. On account of the high gravity of the mercury the benzol can be decanted until nothing remains but a slime of mercury powder. To this add a t once one or two cc. of aqua regia, warm it, stir and let stand. If all gray color is not removed in about a n hour add another cc. of aqua regia, stir, w a r d and let stand again, repeating, if necessary, to dissolve all of the mercury. Use,
Mar., 1910 1-' however, as little acid as possible so as to avoid all but a slight excess, for much excess will interfere with the subsequent precipitation of the mercury by means of H,S. When the acid solution has lost all gray color, indicating the complete solution of the mercury, add about 50 cc. water. Stir, and filter t h r y g h a paper filter con.taining a tuft of absorbent cotton. The cotton catches the flocculent particles of resinous matter and prevents stoppage of the filtration by clogging. Rinse out the beaker with repeated portions of water, which is poured through the filter. Continue until about zoo cc. filtrate are secured. Place this filtrate in a n Erlenmeyer flask of ample capacity to pass H,S through i t until the mercury is all precipitated as black HgS. Let i t settle a few minutes and filter a t once through a weighed filter paper of fine texture, wash the precipitate on the filter with a little water, and dry a t 100' C., cool and weigh. Subtract weight of filter paper and calculate the resultant weight of the mercury sulphide to mercury, viz. : 232 : 200 = wt. HgS : x wt. mercury. The weight of mercury found multiplied by IOO and divided by the weight of the compound used gives per cent. mercury in the compound. A m m o n i a c and Mercury Plaster.-The requirement of the U. S. P. is that it shall contain 18 per cent. metallic mercury. Method of assay for mercury-same as with mercurial plaster. Strengthening Plaster.-Assay for iron : weigh out accurately two pieces of plaster of about j grams each, cutting them side by side from the same piece in order that the .relatiye amount of compound to cloth will be the same in each piece. Dissolve off the compound from one of the pieces with repeated portions of chloroform until the cloth is clean. Allow the cloth to dry spontaneously and weigh it. illultiply its weight by IOO and divide by the total weight of plaster used. This gives the per cent. of cloth in the plaster, the figure so obtained to be used later. The second piece of plaster of known exact weight is placed in small clippings in a porcelain crucible and ignited. Dissolve the residue in warm concentrated hydrochloric acid, dilute about 2 0 0 cc. with water and filter. Place in a porcelain dish, heat nearly to boiling and add ammonia in excess to precipitate the iron as Fe,OH,. Let settle, decant on to a paper filter of known low ash value, wash the precipitate with hot water and dry. This is now ignited in a weighed porcelain crucible with lid, the iron hydroxide being first carefully scraped from the filter and ignited alone in the crucible, and the paper containing only a little adhereflt hydroxide is ignited separately on the lid of the crucible. When the paper is completely reduced to ashes, the lid is placed
RICHARDSON ON M E T H O D S FOR T E S T I N G A M M O N I A . on the crucible and both are transferred to a desiccator to cool. Li’hen cooled, weigh and subtract the weight of the crucible and lid from the weight of the same with contents. This gives the weight of iron oxide (Fe,O,) contained in the crucible. This is calculated to metallic iron, viz. : 160 : 1 1 2 = wt. Fe,O, : x wt. Fe. The percentage of cloth in the plaster has been determined as above in other piece of plaster, therefore the weight of the piece of plaster ignited is multiplied by the per cent. figure for cloth, determined in the other piece of plaster. The weight so found is subtracted from the weight of plaster ignited, which result is the weight of compound ignited. The weight of Fe found, multiplied by IOO and divided by the weight of compound ignited, gives the per cent. Fe in the compound. To Arthur LT. Clark and Powell Hampton acknowledgment is due for work in the elaboration of these processes. I.ABORATORY OF
JOHNSON
& JOHNSON, J.
NEW BRUKSWICR, N.
METHODS FOR TESTING COMMERCIAL, ANHYDROUS, LIQUID AMMONIA. B y W.D. RICHARDSON. Received J a n u a r y 1, 1910.
The average refrigerating engineer is very critical of the quality of the anhydrous ammonia which he uses in his refrigerating system. If anything goes wrong with his plant he is very a p t to blame the ammonia, which he is using, for the trouble. These facts must be taken into consideration when an engineer criticizes any given brand of liquid ammonia. I t may be that the ammonia is not a t fault at all, but that bad management, bad operation or leaks are responsible for the trouble. Nevertheless; there is a certain amount of agreement among engineers all over the country as to the quality of certain brands of liquid ammonia sold for refrigerating purposes. This agreement is so general among engineers a t widely isolated points, that it must be given a good deal of consideration in forming a n opinion as to the quality of any given brands. Criticism of certain makes of ammonia were so persistent that it became necessary to investigate a number of different brands on the market with the idea of determining, if possible, what brand was best to use and what one worst for refrigerating purposes. The ammonias investigated mill be referred to by letters .4, B, C, D, and E, and the quality of the different brands was in the order given, A being the best and E the worst. To determine the residue after evaporation, USF was made of the apparatus described by Dr. F. W-. Frerichs. From one to five lots of 100 cc. each were evaporated slowly either in the air or in a calcium chloride bath, the * Tms JOURNAL, 1, 368 (1909)
97
latter being used to avoid the accumulation of snow and ice on the bulb. The evaporation was conducted until the ammonia was boiled off, but no longer than this, when the volume of the ammonia was read off. The results by this method were as follows from 500 cc. of the liquid ammonia: “A,” 0 . 0 5 cc.; “ B , ” 0 . 0 5 cc.; “C,” 0 . 0 5 cc.; “D,” 0.07j CC. ; “E,” 0 . I O CC. These residues all consisted of dark oily-looking matter containing some iron oxide. The residue was darkest in the case of sample ‘ ‘ E ” and lightest in the case of “A,” and the colors were pretty well graduated from “ A ” to “E.” The amount of residue obtained indicates that the ammonia now sold on the market is nearly free from substances non-volatile at the temperature of boiling ammonia; but a most suggestive point in connection with these results is that many engineers a t different points from their experience with different makes of ammonia grade them just as they would be graded from the results of these evaporative tests, that is, engineers who have never made or seen any chemical tests in their experience would grade the various makes of ammonia from their experience just as they would be graded by these evaporative tests. I t was thought that some information as to the quality of the ammonia might be obtained by investigation of the non-condensing gases, or rather of the non-basic gases dissolved in the liquid ammonia. Among the non-basic gases would be included practically all of the non-condensing gases likely to be present and also carbon dioxide. For this purpose the drum was supported in the ordinary position for drawing off liquid, Frerich’s apparatus was attached for convenience in drawing off and the following apparatus was made use of: a 50-cc. gas measuring burette graduated in 1 / 1 0 cc., with stop-cock in the upper end, was sealed by its open end to the bottom of a flask of convenient size, say 1,000 or 2,000 cc. capacity approximately. This apparatus was filled with standard sulphuric acid of convenient strength, say -V/2, N or 2 AV, and inverted in the same acid contained in 1,000 or 2 , 0 0 0 cc. evaporating dish. The total volume of acid used is known; the acid is colored by means of cochineal indicator. A long narrow delivery tube of glass with end bent upwards is connected with Frerich’s apparatus by means of a strong piece of rubber tubing. The valves are now opened and a small quantity of ammonia allowed to pass through the delivery tube in order to clean i t out. While the ammonia is still issuing from the tube in a slow, steady stream i t is suddenly thrust under the gas-collecting apparatus ; the ammonia is allowed to pass into the standard acid until this is saturated, which is shown by the change of color of the cochineal indicator. The acid must, of course, be suitably agitated. A t the saturation point the tube is quickly withdrawn and the gases afterwards drawn off into a
