This amount was recovered in the first 25 ml. of effluent. Only traces could be detected spectrophotometrically in the next 25 ml. of effluent, after which blank increases due to thorium were noted. KO loss of rare earth occurred in the steps after the ion exchange. By the procedure described here, 6 to 12 determinations may be conveniently carried out simultaneously in 12 to 14 hours. The coprecipitation and separation procedures of Lerner and Pinto (6) have been adapted to allow the use of both smaller samples and equipment. The method appears to be applicable to many other steels where interfering ions would have soluble fluorides and oxalates.
ACKNOWLEDGMENT
The authors thank the Riverside Foundry Division of Sivyer Steel Co., Bettendorf, Iowa, for supplying the base steels used in this study. Appreciation is extended t o the Lindsay Chemical Division, American Potash and Chemical Corp., West Chicago, Ill., for supplying the rare earth oxides used for calibration purposes. LITERATURE CITED
U. S. A., by the Israel Program for Scientific Translations. 1960. Pages (original) 162-75. Pages (translation) 177-90. (2) Banks, C. V., Thompson, J. A., O’LaughCn, J. W., ANAL. CHEM.30, 1792-5 (1958). (3) Faris, J. P., Appl. Spectroscopy 12, 157-61 (1958). (4) Fritz, J. S., Richard, Marlene J., Lane. W. J.. ANAL.CmM. 30. 1776-9 (1958). --, ( 5 ) Kuteinikov, A. F., Lanskoi, G. A., Zhur. Anal. Khim. 14,686-90 (1959). (6) Lerner, M. W., Pinto, L. J., ANAL. CHEM.31,549-51 (1959). > -
~
(1) Alimarin, I. P., Paviotskaya, F. I.,
Akademiia Nauk S.S.S.R. Inst. geokhimii i analitic9koi khimii. “Rare Earth Elements, Washington. Published for the National Science Foundation and the Department of Commerce,
BERNARD J. BORNONG JOHN L. MORIARTY Lunex Co. Box 196 Pleasant Valley, Iowa
Titrimetric Determination of Some Amine Oxides and Phosphine Oxides in Acetic Anhydride SIR.Amine oxides have been determined by reduction with stannous chloride (9) and with titanium(II1) ( 1 , 4 ) . Amine oxides, in general, are not sufficiently basic in glacial acetic acid to permit titration with perchloric acid. This is also trde of the phosphine oxides (6). Some heterocyclic N-oxides undergo irreversible reduction at the dropping mercury electrode (9); however, a direct approach to purity is more desirable. Formation of a chromophore with dimethylaniline has been proposed for the detection of amine oxides (2). No attempt has been made to quantitate this reaction. A red or reddish brown color is reportedly produced when amine oxides are treated 11-ith acetic anhydride (7’). Acetic anhydride has been demonstrated to be the solvent of choice for the titration of certain very weak bases (8, 10, 11). During the investigation of the basicity of some S-oxides, trimethylphosphine oxide was found to exhibit very sharp inflections in acetic anhydride when titrated with perchloric acid. This work has been extended to include other phosphine oxides and a number of heterocyclic amine oxides. This method is probably applicable to aliphatic, alicyclic, and aromatic amine oxides, also; however, pure representative samples of these mere not available. EXPERIMENTAL
Apparatus and Reagents. A Precision-Dom Recordomatic titrator, Model K-3-247, was used in all titrations. T h e modified calomel-glass electrodes and reagents used are described elsewhere (10, 11). For best
results the electrodes should be equilibrated continously in acetic anhydride when not in use rather than the 12-hour immersion period previously specified. Procedure and Results. An approximately 0.001 mole sample, accurately weighed, is dissolved i n 100 ml. of acetic anhydride with stirring
Table I. Titration of Amine Oxides and Phosphine Oxides in Acetic Anhydride
Sl,
Compound Purity TV-Methyl-bis( 2-hydroxyethy1)amine oxide 89 .Oa Nicotinamide 1-oxide 44. 8b 2,6-Dimethylpyridine N-oxide 99.6 99.0 Pyridine N-oxide 96.7 97.lC 100.4c 2-Methylpyridine N-oxide 100.6d 97.5d 4-Cyanopyridine N-oxide 100,4 99.8 4-Hydroxyquinazoline 3-oxide 95.8 97.1 4-Hydroxy-3-imino-l,2,4-
benzotriazene 1-oxide PNitropyridine N-oxide
99.4 101 .o 100.2 Trimethylphosphine oxide 100.1e Tri-n-octylphosphine oxide 96.6 96.4 Triphenylphosphine oxide 94.9 95.6 a Impure sample, exhibited two inflections indicating presence of parent amine. b Incompletely soluble in acetic anhydride. c Irregular titration curves. d Irregular titration curves, titration in acetonitrile gave a value of 100.2%. e Reference ( 1 1 ) .
and then titrated with 0.1 N perchloric acid i n dioxane. All samples were analyzed as received without further purification. The results are shown in Table I. DISCUSSION
Polonovski observed the exothermic reaction of aliphatic amine oxides with acetic anhydride to produce an aldehyde and a secondary amine (7‘). This reaction reportedly accompanied by a reddish brown color was not noted among the compounds reported here. Although the results for pyridine 1-oxide were quantitative, the titration curves were somewhat irregular, indicating the possibility of a side reaction. Acetic anhydride is said to convert 1-oxides (without a n alkyl group alpha to the K +- 0) into pyridones (6). No explanation can be given for the anomalous behavior of 2-methylpyridine 1-oxide. Triphenylphosphine oxide exhibited a distinct inflection; however, i t was much less sharp than the corresponding trimethyl compound. The decreased basicity observed is due to the electronegativity of the aromatic rings. The nature of the substituent as well as its position relative to the amine oxide group will affect the sharpness of the inflection. The polar nature of amine oxides favors solubility in acetic anhydride; however, nicotinamide 1-oxide was only partially soluble. All other amine oxides examined were freely soluble . ACKNOWLEDGMENT
T h e author thanks J. A. Carbon and VOL 34, NO. 7, JUNE 1962
873
R. B. Hasbrouck for submitting some of the compounds reported.
LITERATURE CITED
(1) Brooks, R. T., Sternglana, P. D., ANAL.CHEM.31,561-5 (1959). (2) Coats, N. A., Katritaky, A. R., J . Org. Chem. 24, 1836-7 (1959).
(3) Glynn, E., Analyst 72, 248-50 (1947). (4) Hjorth-Hansen, S., Anal. Chim. Acta 6,438 (1952). (5) Katritaky, A. R., Quart. Revs. 10, 403 (1956). ( 6 j Kennedy, J., Lane, E. S., Willans, J. L., J. Chem. SOC.1956,4670. (7) Polonovski, M., Bull. SOC. Chim. Belg. 39, 24 (1930). (8) Streuli, C. A., ANAL.CHEM.30, 9971000 (1958). (9) Varyukhina, L. V., Pushkareva, Z. V.,
J . Gen. Chem. (U.S.8.R.) 26, 1953-8 (1956). (10) Wimer, D. C., ANAL. CHEM.30, 7780 (1958). (11) Win&, D. C., Ibid., 30, 2060-1 (1958). DAVIDC. WIMER
Analytical Research Department Abbott Laboratories North Chicago, Ill.
An Improved Exhaust System for the Perkin-Elmer Vapor Fractometer R. J. DiCenzo, Ferris Institute, Big Rapids, Mich. SIMPLE MODIFICATIONof the Perkin-
A Elmer Model 154-C Vapor Fractometer is herein reported which gives superior results and eliminates contamination of fractions. During numerous attempts to use this instrument on a preparative scale, it was noticed that the effluent vapors had been assuming a gradually increasing odor indicative of decomposition within the exhaust network. It seemed plausible that such decomposition, if it existed, would account for the contamination of the fractions being collected. The entire flow system was removed, and the contamination was traced to the three-way solenoid valve, which upon disassembly revealed accumulations of tars throughout. The design of the exhaust system is such that the path traveled by the effluent vapors is lengthy and winding. Further, most of the exhaust system is located outside of the oven and is thus a t a considerably lower temperature than the incoming vapors. Therefore, the system tends to condense the hot vapors before they emerge from the instrument. This is especially true of the solenoid valve, which confronts the vapors with a large, cool mass. Finally, those vapors which do condense within the system are prevented from flowing out of the instrument by three sections of vertical tubing incorporated into the network. These problems and the contamination of the fractions were eliminated by the following simple modification. The complete exhaust network, including
Figure 1. A. 6. C. D.
E.
F.
874
the solenoid valve, was disconnected from the detector cell and removed. A hole approximately 35 mm. in diameter was drilled through the left outer wall of the cabinet and was extended through the wall of the heating chamber itself. The hole was located 55 mm. from the bottom of the cabinet and 135 mm. from its forward edge. A piece of stainless steel tubing, 3.2 X 255 mm., was used as the exhaust pipe and was bent in two places. The first bend was made a t an 80" angle and was located 25 mm. from one end. A suitable Swegelok connector was then attached to this end. The other bend was made at a n angle of 45" and was located 25 mm. from the other end. T o this end was attached a threeway, stainless steel, medical stopcock consisting of one female Luer, one male Luer, and one male Luer-Lok fitting. The stopcock was attached by soldering the end of the tube into the mouth of the female Luer fitting. The tube was then tightly wrapped between bends with heavy asbestos cord and then wound with several turns of suitable resistance wire, with the free ends of the wire located near the stopcock to s n e as power leads. The wire was itself nrnpped with asbestos cord for insulation. A thermometer was positioned along the axis of the tube, and the entire unit was encased in several layers of asbestos tape impregnated with plaster
Improved exhaust system
Thermometer Three-woy medical stopcock Stainless steel tubing Power leads for autotransformer Asbestos insulation Swegelok connector
ANALYTICAL CHEMISTRY
of Paris, and allowed to harden. Care was taken to ensure that the temperature range of the heating chamber was visible on the thermometer when the unit was attached to the detector cell. The complete unit is shown in Figure 1. The new exhaust system was then installed in the heating chamber by connecting it to the detector cell T-tube outlet, and the hole in the cabinet wall was firmly packed with glass wool to minimize heat loss. Finally, the wire leads were connected to a variable autotransformer having a n output of 0 to 120 volts. The two free male Luer fittings allow a variety of collection systems to be adapted according to need. The LuerLok fitting is especially suitable for attachment of long, large bore hypodermic needles which can be lead into small, easily changeable test tubes immersed in refrigerant baths or ,containing solvents or adsorbent traps. The stopcock allows rapid and convenient diversion of the effluent to either of the outlets so that fraction cuts can be made quite readily. Finally, this unit offers a short, direct exhaust path to the vapors and can be heated to any desired temperature along its entire length. The common problems of condensation and contamination of frxtions are thereby minimized or eliminated.
