Device for Continuous Liquid-Liquid Extraction Adaptation for the Determination of Morphine JOHN R. MATCHETT AND JOSEPH LEVINE U. S. Bureau of Narcotics Laboratory, Washington, D. C.
C
OKTINUOUS liquid-liquid extraction devices for use with immiscible solvents lighter than water have been described in numerous publications. The most useful of these for the quantitative extraction of small samples are patterned after the general design of Palkin, Murray, and Watkins (I), in which solvent is conveyed from a condenser to the bottom of the solution being extracted, from whence it rises through the solution and returns t o the boiling flask. Sintered-glass plates, spirals, small orifices, and the like are used for distributing the solvent throughout the solution. These distributing devices are generally separate from the body of the extractor and, being fragile, are consequently subject to breakage; but more serious disadvantages are that solvent is distributed from a relatively small area and a space exists at the bottom of the extractor where solvent and solution do not come into contact. Under these conditions, much time is required for complete extraction and, in this laboratory, it has been found
necessary to disconnect the apparatus from time to time to stir the solution being extracted, in order to obtain quantitative results. It is necessary also to disconnect the apparatus for the purpose of withdrawing samples of solvent in order to determine when the extraction is complete. These operations are annoying and time-consuming, especially since one must wait until the solvent stops boiling before removing the condenser. The device described herein has been designed to obviate these difficulties. It has been found both quick and convenient for quantitative extraction of alkaloids, and is equally well suited for other substances. The sintered-glass solvent-distributing plate is permanently sealed in the extraction tube where i t covers the entire cross section and is protected from breakage. Solvent is distributed throughout the entire volume of the solution being extracted rather than in a narrow stream of droplets. There is no space under the distributing plate where solvent and solution do not come into contact. The apparatus is filled and emptied through a side arm closed b y a ground joint which also provides a n opening through which to pipet occasional samples of solvent to test for completeness of extraction. The opening is sufficiently small to permit removing samples without interrupting the boiling. A drawn-out glass tube provided with a rubber bulb is convenient for this purpose.
Apparatus The diagram is self-ex lanatory. In operation a boiling flask for the solvent is attacheg by the ground joint at the bottom and a ood condenser is attached in similar fashion at the top. The sofvent vapors distilling from the boiling flask are condensed, passed down the inner funnel, and distributed throughout the aqueous solution by means of the sintered-glass late. Solvent returns from the up er layer in the extractor toe!t boiling flask through the same,u!t which carries the vapors to the condenser. The device is char ed and emptied through the side opening. The dimensions can%e modified t o almost any desired extent. For most satisfactory operation it is important that the sintered plate be as coarse as possible, that the funnel tube extend as nearly as possible to the bottom of the extraction tube, and that the space between the plate and the bottom be small.
Determination of Morphine The device has been found particularly useful for the determination of morphine, where i t presents a number of advantages over the generally used method of extracting the alkaloid from a n ammoniacal solution by a mixture of chloroform and isopropanol. This operation is not readily adaptable to a continuous process where the solvent is recycled by boiling and condensation, because chloroform hydrolyzes under these conditions and the resulting acids unite with the extracted morphine, thus vitiating the results. The hand operation, employing separatory funnels, requires much solvent, is extremely tedious, and permits extraction of only small samples. A mixture of benzene and isopropanol (2 volumes to 1 volume) serves admirably as extracting solvent in the device 264
April 15, 1941
265
ANALYTICAL EDITION
described. Only small volumes are required and relatively large samples (about 0.20 gram of alkaloid) are readily extracted. Place sufficient solvent to cover the sintered plate in the device, introducing it through the inner funnel. Accurately weigh a sample e uivalent to approximately 0.20 gram of morphine alkaloid and 8issolve it in approximately 50 cc. of dilute hydrochloric or sulfuric acid. Quantitatively transfer the solution to the extractor through the side opening, using approximately 25 cc. of wash water. Place 50 to 75 cc. of solvent and a few Carborundum chips in the boiling flask and connect flask to the extractor. Drop sufficient ammonia to render the solution faintly alkaline through the inner funnel and connect the condenser. Distill the solvent rapidly through the solution until a sample of solvent removed (through the side tube by means of a pipet) from the upper layer in the extractor is found by a suitable test (Marquis reagent) to be free of morphine. The presence of sufficient ammonia in the solution may be confirmed at the same time by holding a piece of wet litmus paper in the vapor escaping from the side tube, When extraction is complete (0.5 to 1 hour) disconnect the boiling flask, evaporate the solvent, and titrate the alkaloid in the usual way.
Under the conditions described there is a volume increase of 20 per cent in the aqueous layer due to solubility of isopropanol. The addition of ammonia in the prescribed manner renders the solution alkaline at the same time that solvent is available to extract the liberated alkaloid. This is of consequence in morphine extraction and presumably also in other cases. T h e simple analytical procedure outlined is, of course, useful only where interfering substances are absent. When this is not the case, the extractor becomes even more useful. Morphine can be separated from many other alkaloids, for example, by dissolving the mixture in dilute acid, rendering the solution distinctly alkaline with sodium hydroxide, and extracting with benzene to exhaustion. T h e solution is acidified, then made alkaline with ammonia, and the morphine removed in fresh mixed solvent (benzene-isopropanol). A method for opium assay based on this procedure is now under consideration in this laboratory.
Literature Cited (1) Palkin, Murray, and Watkins, IND.ENQ.CHEM.,17, 612 (1925).
Perforated Plate Columns for Analvtical Batch Distillations J
C. F. OLDERSHAW, Shell Development Company, Emeryville, Calif.
A
CCURATE analytical batch distillations have become increasingly important during the past few years. Of the many different columns tested in these laboratories for this purpose, the perforated plate column described in this paper has been found to be of especial value for the analysis of hydrocarbon mixtures. It combines a low operating holdup per theoretical plate and a negligible static, or nondrainable holdup, with a high capacity or reflux rate. It has a low H. E. T. P., is easy to operate, and since it is constructed entirely of glass, may be used for corrosive liquids.
Description of Column The column (Figures 1 and 2) consists of a series of perforated glass plates sealed into a tube. Each plate is equip ed with a baffle to direct the flow of liquid, a weir to maintain a {quid level on the plate, and a drain pipe. The first plate in a series serves as a small reservoir which is necessary in order to maintain a liquid seal for the drain pipe from the first regular plate.
Construction
FIGURE 1. PERFORATED PLATE COLUMN
The plates are constructed in the following manner: A tube which later will form the baffle is placed in a lathe and rapidly rotated. The protruding end of the tube is heated in an oxy en flame and flared to form the disklike ortion of the plate. +he perforations are then drilled with a r e l h o t tungsten wire. This operation, which is ordinarily somewhat time-consuming when done by hand, is now entirely performed by an automatic drilling device designed in these laboratories for the purpose. The hot tungsten wire forms a small burr on the plate which is removed with an ordinary file. After the drilling operation the plates are examined and only those having perforations of uniform diameter and spacing are selected for use. The selected plates are ground to size; if they have been correctly flared, the grinding is a minor operation. The drain pipes are sealed in place and shaped as shown in Figure 3, and a bead 0.5 mm. high is made on the lower end of each to regulate the distance between the drain and the plate below. The baffle tube is cut off a t the desired height above the plate, and two longitudinal cuts are made with small high-speed
