V O L U M E 23, N 0.
OCTOBER 1951
in with the microscope drive, one after another, up to the maximum and then removed successively, the recorder drew the pattern shown in B , in which the ordinates rise and fall in uniform increments-Le., proportibnal to h. The range or proportionality factor in this network can be varied by changing the applied potential or by the choice of swamping resistors. If the latter method is chosen, the only practical requirement is to keep the magnitude of the swamping resistors a t least 1000 times greater than the maximal value of the squaring element. In this application, no elaborate precautions were taken to ensure the maximum attainable precision. The battery source was permanently connected to the network, wherein the drain was a constant 450 Fa. Temperature coefficients of resistance and source e.m.f. were ignored and under these circumstances the reproducibility over weeks a t a time was of the order of 0.09 chart divisions, which corresponds to an error signal a t the input of the recorder of 2 pv. Any doubt about the absolute response is easily checked by setting the microscope drive to an arbitrarily chosen height. If the indicated response differs detectably from the correct value, the operator may restore the original deflection electrically or use a small correction factor in computing the slope of the square-law plot. In anv case the quadratic response is rig-
1495 orously maintained-only the slope factor is subject to change. For anyone interested in precision beyond these high levels, or in very long-term stability, the conventional potentiometric methods of checking against a standard cell can be used. The literature ( 1 , 4, 6) on related computer elements iq extensive. Any instrumental development which eliminates tedium and fatigue, and which increases the speed and precision with which data can be accumulated, may well defeat its purpose if it does not include means for the rapid calculation and assimilation of the data. LITER4TURE CITED (1) Greenwood, I. A., H o l d a m , J. V., Jr., a n d MacRae, D., Jr., “ E l e c t r o n i c I n s t r u m e n t s , ” R a d i a t i o n L a b o r a t o r y Series, Vol. 21, p. 121, N e w Y o r k . M c G r a w - H i l l B o o k Co., 1948. (2) Ibid., p. 139. (3) Muller, R. H., a n d Clegg, D. L., AXAL. CHEM., 23, 396-411 (1951). (4) M u r r a y , F. J., “ T h e o r y of h l a t h e n i a t i c a l Machines." 2 n d ed., X e w Y o r k , King’s C r o w n P r e s s , 1948. ( 5 ) S v o b o d a , “ C o m p u t i n g M e c h a n i s m s a n d Linkages,” M.I.T. Radia t i o n L a b o r a t o r y Series, V o l u m e 27, K e w Tork. McGraw-Hi11 B o o k Co., 1947. RECEIVED February 6 , 1961.
lithium Aluminum Hydride as a Qualitative Test Reagent for Aromatic Nitro Compounds LLOYD S. NELSON’ AND DONALD E. LASKOWSKI Illinois Institute of Technology, Chicago 16, I l l . I X T R O M and Brown ( I O ) reported the reduction of nitrobenzene, p-nitrobromobenzene, and nitromesitylene directly to the corresponding symmetrical azo compounds in good yields with lithium aluminum hydride. Because of the distinct color changes which occurred, these workers suggested the possible applicability of lithium aluminum hydride as a reagent for determining aromatic nitro groups. This paper presents the results of a series of experiments designed to investigate the applicability and determine the sensitivity of lithium aluminum hydride as such a test reagent. -4fter its acceptance for publication a note on the same subject appeared (3). Several qualitative tests for the aromatic nitro group have heen reported. Reduction to the amine and subsequent identification appear in several laboratory texts ( 6 , 7 , 14.16). Reduction to the substituted hydroxylamine, .rvhich then gives a positive Tollens test, is also common practice ( 2 , 6,9,12,13,15). Bost and SichOlson ( 1 ) have reported the action of alkali on acetone solutions of aromatic nitro compounds as a color test. Olivier (11) states that, aromatic nitro compounds give a red color with aluminum bromide in benzene. Hearon and Gustavson (4)have explored the use of ferrous hydroxide as a color test reagent for nitro compounds in general. PROCEDURE
T w o materials are required. 1. Lithium aluminum hydride. The product of Metal Hydrides, Inc., was used. The same results are obtained whether the hydride is dark gray or practically white. 2. Absolute ethyl ether. Commercial anhydrous ethyl ether dried over sodium metal was used. Approximately 100 mg. of the unknown are dissolved in 5 ml. of anhydrous ethyl ether. About 10 mg. of lithium aluminum hydride are added to this solution. A change in (but not a disappearance of) the color of the solution, the formation of a colored precipitate, or both, within 5 minutes, is taken as a positive test. 1
Present address, General Elertric Co.. Waterford X Y
DISCUSSION AND CONCLUSIOY
Twenty-six aromatic nitro compounds containing a wide variety of functional groups were tested. These are listed in Table I
Table I Coinpound Solution Color 1-Nitronaphthalene Y tint I O Y normal tone Nitrobenzene 2,4-Dinitrophenylhydrazine Colorless VR shade 2 2-Nitrodiphenylamine YO tint 2 2,4,6-Trinitrotoluene m-Nitrobenzenesulfonic c acid Colorless 3,B-Dinitrobenzoic acid 2.4-Dinitro~henvlthio. ” R O - 0 shade 1 cyanate 0 shade 1 Picric acid o-Nitroiodobenaene c 2-Nitroresorcinol 4-Bromo-3-nitrobenzeneY tint 1-2 sulfonic acid 8-Nitroquinoline R-OR shade 2 o-Nitrobromobenzene YO normal tone e o-Kitroaniline c m-Xitrobenzaldehyde o-Nitrochlorobenzene VR normal tone p-Nitrobenzoic acid Y broken tone 2,4-Dinitrochlorobenzene e rn-Nitrobenzvl alcohol O Y tint 1 Green-brownd o-Xitroanisoie o-Ethylnitrobenzene OR-RO normdl tone 0 - Y O normal o-Nitrotoluene tone Y shade 2 n-Nitrochlorobenzene 0 tint 1 o-Nitrodiphenyl 2-Chloro-5-nitrobenzenesulfonic acid Colorless Abbreviations V. Violet R. Red 0. Orange RO. Red-orange Y. Yellow OR. Orange-red
Precipitate Color Y tint 2 YO shade 1
O Y shade 2 0 shade 1
T tint 2
Limit of Detection, G . / M . X 10 2 3 a
0.01 1
Brownd
2 0.7
R O - 0 shade 2 OR-RO shade 1 R O - 0 tint 1 V broken tone
0 2 8
I’ tint 1-2
R-OR shade 2 V broken tone OY tint 2 Y O shade 2 Y broken tone 0 - Y O shade 2 O Y tint 2
0 3 4 0 1 3 0 2 1 0.4
3
2
,
YO shade 1
0.4 3
YO normal tone
4
Y broken tone OY tint 1
8 3
White
e
b
OY. Orange-yellow YO. Yellow-orange VR. Violet-red
5 Only a saturated solution was tested because of insolubility of compound. b Color uncertain b u t similar to so1,ution color, C Suspended material made detection of solution color uncertain. d N o t adeauatelv described b y color standard. e Negative test.
ANALYTICAL CHEMISTRY
1496 together with the colors observed and the limits of detection. The colors reported correspond t o those found on the hlulliken color standard (8). The limits of detection were obtained by successive halving of the concentrations of the nitro compounds and are defined as lying between the concentration a t which the test was questionable and the next higher concentration. Of the compounds listed, only 2-chloro-5-nitrobenzenesulfonic acid failed t o give the test. A large number of substances other than aromatic nitro compounds were also tested. Of importance is the fact that hydroxylamine, amyl nitrate, amyl nitrite, benzoquinone, hydroquinone, nitroethane, and nitrocyc.lohexane all gave negative tests. Many of these compounds interfere with other test.s for aromatic nitro groups. Phenylacetonitrile gave a positive test, presumably because of a reaction involving its alpha-hydrogen, as octanenitrile and benzonitrile gave negative tests. p-Dimethylaminonitrosobenzene and azosybenzene, as was expected, gave positive tests. Carefully purified hydrazobenzene also gave a positive test. All substances, such as 8-quinolinol, which are capable of forming colored complexes with aluminum ion will also be expected to give positive tests. Michler’s ketone (p,p’-bisdimethylaminobenzophenone) gave a negative test. If, however, the reaction mixture of Michler’s ketone and lithium aluminum hydride is hydrolyzed and then oxidized, various shades of blue and green are obtained depending on the concentration of hydride (IO). I n contrast t o some other color tests, colored compounds can be tested by this procedure because a change (but not a disappearance) of color constitutes a positive test. It is, of course, advisable to compare the color of the tested solution with that of a solution having an equivalent concentration of the unknown. The test described here is both simple t o perform and sensitive. It is given by those compounds which can be reduced to
azo compounds by lithium aluminum hydride-Le., aromatic nitro, aromatic nitroso, and azoxy compounds. hliphat,ic nitro compounds, nitrates, and nitrites do notjnterfere. LITERATURE CITED
(1) Bost, R. W., and Nicholson, F., ISD. ESG. CHEM..AXIL. ED., 7, 190-1 (1935).
(2) Cheronis, N. D., and Entrikin, J. B., ”Semimicro Qualitative Organic Analysis,” p . 135, Xem York, Thomas Crowell Co., 1947. (3) Gilman, H., and Goreau, T. N., J . A m . Chertt. Yoc., 73,293940 (1951). (4) Hearon, W. M.,and Gustavson, R. G., IND.ESG. CHEDI., ~ A L ED., . 9, 352-2 (1937). ( 5 ) Kamm, O., “Qualitative Organic Analysis.” p. 69. New York, John Wiley & Sons, 1923. (6) Ihid., p. 72. (7) AlcElvain. S. AI.. ”Characterization of Oiaaiiic Comi,ounds.” p. 145, Sew York. Macmillan Co., 1945. (8) hlulliken, S. P., “Identification of Pure Organic Compounds,” Vol. I, Xew York, John Wiley & Sons, 1904. (9) Mulliken, S. P., and Huntress, E. H., “Identification of Organic Compounds,” p. 162, Cambridge, ~ ~ I s s s .Cummings , Co., 1937. (10) Systrom, R. F., and Brown, ST” G., J . Am. Chem. Soc., 70,373840 (1948). (11) Olivier, S. C. J., Rec. trac. chim.,37,241-1 (1918). (12) Porter, C. W., Stewart, T. D., and Branch, G. E. K., “Methods of Organic Chemistry,” p. 203, Boston, Ginn & Co., 1927. (13) Shriner, R. L., and Fuson, R. C., “Systematic Identification of Organic Compounds,” 2nd ed., p. 73, Sew York, John Wiley &Sons, 1940. (14) Ibid.,p. 75. (15) Vogel, -4.I., ”A Text-Book of Practical Organic Chemistry,” p. 930, London, Longmans, Green & Co., 1948. RECEIVED February 17, 1951. Presented before the Division of Organic Chemistry a t the 119th Jfeeting of the AXERICASCHEMICAL SOCIETY. Cleveland, Ohio.
Colorimetric Determination of Phosphorus in Steel U S 0 T. HILL Inland Steel Co., East Chicago, Ind. determination of phosphorus in steel bv COLORIMETRIC the molybdenum blue method generally requires either elimi-
nation of all nitric acid from the solution prior to reduction of the bound complex (1,S), or a separation by estraction with an organic solvent (4, 5 ) . This work m s undertaken t o eliminate these steps and if possible t o speed up the method for control purposes in steel manufacture where rapid determinations are a necessity. I n a previous paper ( 2 )it was reported that, by the addition of a fluoride to the silicomolybdate complex in a solution of ferric and ferrous salts, the bound molybdenum is reduced t o the blue, the color being proportional to the silicon present. Apparently after the ferric ions are removed from solution as the fluoride complex, the ferrous ions are capable of reducing the bound molybdenum under the described conditions. Further study of the complexing action of the fluoride ion has resulted in a rapid colorimetric method for the determination of phosphorus in steel based on the molybdenum blue color in a nitric acid medium. EQUIPMENT
The Coleman Model 11 spectrophotometer with 18-mm. matched cylindrical tubes \vas used. EXPERIMENTAL
I t was found that a stable molybdenum blue color proportional to the phosphorus present could be formpd in an oxidized nitric
acid solution of steel by complexing the ferric ions and auppressing the ionization of the nitric acid with a fluoride. Potassium permanganate as an oxidant and dtannous chloride as a reductant proved to be the most reliable reagents for the production of the blue color. Excess permanganate was removed with either sodium nitrite or sulfite. Where a colorimetric manganese determination is to be made on the same solution of sample, ammonium persulfate may be used as an oxidant, but the excess must be destroyed as in the permanganate ouidation. Hydrochloric acid in any appreciable amounts produced color instability and the use of stannous chloride dissolved in hydrochloric acid was abandoned, although a fairly accurate method based on its use was a t first developed. As stannous chloride is readily soluble in solutions of a fluoride, it became convenient t o incorporate it in the complexing reagent. The small amount of chloride introduced in this manner has little effect on the color stability. A spectral transmittance curve obtained with the Coleman instrument indicated a maximum absorption a t 780 mp. I n order to obtain a suitable range, a working curve was constructed a t 650 mp. Some interference has been noted a t this wave length iyith a filter instrument, but none Ivith the narrower band of the instrument used in this work. SOLUTIONS REQUIRED
Nitric acid, 1.20 specific gravity Potassium permanganate, 2.5% Sodium nitrite or sulfite, 10%