A N A L Y T I C A L CHEMISTRY
1580
Table I1 shorn that accurate determinations can be made with mandelic acid in the presence of large amounts of aluminum, lron, and titanium. The phosphate method, however, gave rather erratic results in the presence of these impurities. When less than 3 mg. of airconium oxide w&s present, no precipitate was obtained even upon the addition of 20 grams of ammonium phosphate. Obviously, the mandelic method is the superior method. Kumins gave no data concerning separations from cobalt, magnesium, manganese, mercury, nickel, uranium, and zinc Separations from these elements were studied and the results shown in Table 111. These data show that successful separations can he made from cobalt, magnesium, manganese, mercury, niekel, uranium, and zinc. The precipitates formed in the presence of cobalt and manganese were discolored slightly, showing some contamination Solutions of hafnium chloride, when treated with mandelip acid.. gave a. precipitate of the same nature as those of zirconium . mrtndelate. These precipitates were ignited and weighed, and compared with those obtained by Inecipitation with smmoniurn hydroxide:
...
.
. . ..
Weight of HfO, by precipitation wicn ammonium hyaroxme(average), 0.0410 gram Weight of Hi01 by precipitation with 1mandelic acid (average ), 0.0407 gram
Table 111. Effect of Impuriti %PO1
htr (0.250 G.) c0-a
8
M g + +LSiMzCh ~I ~
~
.
Mn++as MnCl Kg * 8s HgCl. Ni + + 8 8 NiCh ^"
0.11766 0.11153
0.0771
0.0153
0.0156 0.0772
0.11766
0.0766 0.0153 0.0766
+
r,n.++
Foun.. 0.
G.
COCI.
,,n.,x,n.>.
zro
Tiken.
G.
0,0156 0.0768
0.0153 0.0766
0.0157 0.0768 0.0154 0,0768
lmmz
$ m &
+0.0005
+0.0003 +0.0002
+0.0003 +O.OCO6 +0.0004
+0.0002 +O.OOCI f0.0002
y.:;:;
hianoeiic acla, ime a11 ocner reagents Gnat precipnare z m a nium, also precipitates hafnium quantitatively. Hence, in using mandelic acid for the determination of zirconium, one actuallv . . . d . (A,
lLllleULLlll"
" I ' *
YULLUSll,
"pp"""
'""lg'LL1u
""iuylua.
L"sr
York, John Wiley i (2) Kumins. C. A., ANAL. (3) Lundell, Hoffman. an Steel," New York,
.
3
RECWEDFebruary 23, 1841I ,
2-Anthraquinone Sulfonate Derivatives of Morphine and Codeine MILTON FELDSTEIN, NIELS C. KLENDSHOJ, AND ALICE SPRAGUE School of Medicine, University of Buffalo, Buffalo General Hospital, Buffalo, N . Y . teristie crystals with many of the organic bases, but actual dis. tinction between morphine and codeine is difficult. The authors fmd that a solution of sodium 2-anthraquinone sulfonate forms characteristic crystals with morphine and eadeine, and that the melting points of these derivatives w e sufficiently faapart to afford a means of specific identification for either. Morphine and codeine are the only alkaloids of 19 tested which give a crystalline precipitate under the conditions outlined Cocaine. Dontocaine. procaine. metvcaine. iltronine. homatronine.
N TOXICOLOUICAL analyses of blood or tissue specimens it is often necessary to confirm the presence of small amounts of morphine or codeine. The alkaloids are usually isolated througl .one of the modifications of the Stass-Otto process (6) m d thi purified residue obtained after evaporation of the ether or chloro b r m extract is examined for the presence of basic substances. Among the many color reactions which will detect the presenci .of morphine and codeine are Marquis' (4, Meeke's (5), an< fioehde's ( I ) reagents. However, these color reactions do no differentiate between morphine or codeine as isolated from tissuq specimens. There are also several crystal reagents, such a 3 Kraut's (8), Wagner's (7). and Marme's ($), which form charac-
Figure 2.
Codeine 2-Anthraquinone Sulfonate (X130)
V O L U M E 21, NO. 1 2 , D E C E M B E R 1 9 4 9
1581
not form insoluble 2-anthraquinone sulfonate salts in 6 N sulfuric acid a t room temperature.
in.an oven at 100' C. and the melting point determined under the mcroscope.
To make the reagent, 1 gram of sodium %anthraquinone sulfonate is added to 20 ml. of water containing 2 ml. of 3 A' hydrochloric acid. The mixture is stirred for a few minutes, filtered, and stored in a brown bottle. Reagents are added to the preparations with a dropping pipet with the tip drawn out to a capillary 1 mm. in diameter. About 1 to 2 mg. of the dry purified residue obtained from the extract of tissue are placed on a slid? and dissolved in one drop of 6 AT sulfuric acid. One drop of the reagent is added directly to the drop on the slide, and the preparation is allowed to stand for 10 to 15 minutes for crystallization t o occur. The crystals formrd with morphine and codeine isolated from tissue samples are shown in Figures 1 and 2. Excess solution is decanted from the preparation by means of a capillary pipet of 0.5-mm. bore and about 6 cm. in length. The pipet is placed vertically in the drop, and then slowly inclined until it is almost horizontal. Liquid rises in the tube by capillary action. The tube is then carefully removed without disturbing the crystals and a drop of water is added to the residue. The decantation and washing are repeated once more. The slide is dried
The morphine derivative melts a t 198' to 199' C., and its solubility in water a t 20 O C. is 0.85 mg. per ml. The codeine derivative melts a t 175" to 176' C. and its solubility in water a t 20' is 0.87 mg. per ml. Both derivatives are deep yellow in color when prepared in large amounts but appear colorless in minute amounts on a slide.
e.
LITERATURE CITED
(1) (2) (3) (4) (5) (6) (7)
Froehde, A., 2. anal. Chem., 5, 214 (1866). Kraut, K., Ann., 210, 311 (1881). Marme, W., Z.a n d Chem., 24, 642 (1885). Marquis, E., Pharm. Zentralhalle, 1896, 814. Mecke, A., 2. d f e n t l . Chem., 1899, 351. Panaer, Th., Z . anal. Chem., 47, 572 (1908). Wagner, R., Arch. Pharm., 1863, 260.
RECEIVED February 10, 1949.
Apparatus for Fractional Crystallization in Vacuum HAROLD A. SCHERAGA'
AND
MILTON MANES2, Duke University, Durham, A'. C.
7EVERAL different designs have been reported for accomplishing the purification of liquids by fractional crystallization (1-4). However, they did not contain all the features that were desired for the purification of benzalchloride which was to be used for kinetic studies ( 5 ) . In particular, an apparatus was desired in which all operations could be carried out in vacuum with the liquid being stirred during cooling. In addition, it was desirable to be able to determine cooling curves and pour off the rejected and accepted fractions without breaking the vacuum. The apparatus described below accomplishes this purpose without contamination of the liquid by the lubricant on the joints, and may be used for liquids that freeze within a very wide range of temperatures. Two aspects of the crystallization apparatus are shown in Figures 1and 2.
with the liquid. All rejected fractions were poured through outlet D on the cow to a receiver. The raised ring-sealed tube, H ,prevented any drippings on the bottom of the cow from running into the "accepted" receiver. When the accepted fraction was ready to be poured to the accepted receiver, the crystals were melted and the cow was rotated on joint G to bring outlet C into place. In many cases, the sintered-glass disk may not be necessary, m the crystals tend to remain behind during the pouring operation. In such cases, the absence of a disk facilitates pouring. A bulge in the back of the vessel near the top of the thermocouple well, shown in Figure 2 and indicated by the dotted oval in Figure 1, altered the geometry of the apparatus in a manner which aided the pouring operation.
Figure 1 is a profile of the entire apparatus; Figure 2 represents a cross section of part of the apparatus, cut by a plane which is perpendicular to the plane of Figure 1 and passes through the inlet tube, sealed to the vessel a t A . A is a point common to both drawings. The tubular vessel, in which the liquid was frozen, was constructed from 45-mm. outside diameter tubing and had a capacity of approximately 130 ml. Graduations on this vessel served to indicate the volume of liquid after successive pourings of unfrozen material. The vessel was charged through an inlet tube, A (Figure 2 ) , which was inclined about 20" upward. The inlet tube was then closed with a cap made from a ground-glass joint, J, to which a stopcock, K , was sealed. The system was then immediately evacuated, with stirring, through stopcock B (Figure 1) to outgas the liquid a t a pressure of 1 mm. of mercury. The stirrer was made from glass rod and had a glass-enclosed iron rod sealed to its top for magnetic operation. If it were desired to open the system without atmospheric contamination-e.g., to change receivers-a positive pressure of nitrogen could be maintained by admitting the gas through L and releasing it through C or D. Cooling curves could be determined by means of a thermocouple inserted in the thermocouple well which contained a small pool of mercury a t the bottom for good thermal contact-for example, with a three-junction copper-constantan thermocouple, used a t about -20" C. in conjunction with a Leeds & h-orthrup type K potentiometer a;d a type R galvanometer, temperatures could be read to 0.01 , using suitable cooling baths for the crystallization. Pouring off of unfrozen material as a "rejected" fraction was accomplished by rotating the vessel on joint E, a sintered-glass disk serving to catch any crystals that might be carried through 1 Preaent addreas, Department of Chemistry, Cornell University, Ithacs, N. Y. Present address, U. S. Bureau of Mines, Pittsburgh, Pa.
Figure 1. Crystallization Apparatus