ANALYTICAL EDITION
April 15, 1943
Literature Cited (1) Carolus, R. L., Va. Truck Expt. Sta., Bull. 98 (1938). (2) Emmert, E. M., Ky. Agr. Expt. Sta., Circ. 43 (1934). (3) Emmert, E.M., J. A& SOC.Agron., 27 ,1-7 (1935). (4) Hance, F. E., Hawaiian Sugar Planters Assoc. Expt. Sta., Bull. 50 (1936); 51 (1937); 53 (1941). ( 5 ) Hester, J. B., Commercial Fertilizer Yearbook, pp. 31-9, 1941. (6) Hester, J. B.,Va. Truck Expt. Sta., Bull. 82 (1934). (7) Merkle, F. G.,Penn. State Coll., Bull. 398 (1941).
251
(8) Morgan, M. F.,Conn. Agr. Expt. Sta., Bull. 450 (1941). (9) Scarseth, G.D.,Soil Sci., 55, 113-20 (1943). (10) Snell, F. D.,and Snell, C. T., "Colorimetric Methods of Analysis", Vol. 1, New York, D.Van Nostrand Co., 1941. (11) Spurway, C.H., Mich. Agr. Expt. Sta., Tech. Bull. 132 (1933). (12) Thornton, S. F.,Canner, S. D., and Fraser, R. R., Purdue Univ. Agr. Expt. Sta., Circ. 204 (rev.) (1939). (13) Ulrich, Albert, Soil Sci., 55, 101-12 (1943). (14) Wolf, B.,J . Am. SOC.Agron., 34, 646-50 (1940). (15) Zinsadse, Ch., IND.ENG.CHEM., . ~ N A L . ED.,7, 227-30 (1935).
Photometric Determination of Benzene, Toluene, and Their Nitro Derivatives H. D. B-iERNSTEIN, Division of Industrial Hygiene, National Institute of Health, Bethesda, &Id.
Certain factors relating to the production of maximum color from nitro compounds with butanone and alkali have been studied and modifications of existing procedures are recommended. The analysis of mixtures of benzene and toluene, such as occur in some new commercial solvents, has been improved by introducing a differential oxidation of the nitro derivatives.
T
HE fundamental reaction between certain nitro compounds and ketones i n the presence of concentrated alkali has been studied by many workers ( 1 , 8, 4-7, 9, 11). A summary of their findings leads to the following conclusions: Mononitro derivatives of benzene and its homologs do not give the reaction. Among the dinitro derivatives tested, only those of the meta series (nitro groups meta to each other) respond. When further substitution in the ring occurs, only those compounds retaining ketone solubility give color. At least one position para to one of the nitro groups must be unsubstituted. Blues or reds signify the presence of aromatic polynitro derivatives of this class. Certain subs ,itutions, such as hydroxyl and amino groups, change these colors to yellows and oranges and indicate a different reaction. The concentration of alkali used determines the intensity of the color obtained and the rapidity of its production. The evidence indicates a condensation between the enol form of the ketone and the aci-nitro form of the nitro compound with t h e elimination of a molecule of water: R I
k=CH2
/
H
d
v
R-
c = CH* + KOH + AH
nN2
+ HzO
VNO,- Y
O=N--O-K
T h e colors gradually fade, and brown insoluble compounds are produced (11). In some cases, one of the nitro groups is replaced ( I S ) . Quantitative procedures have been described b y Pearce, Schrenk, and Yant (IO, 12,14,15) and by K a y (8) but are not
entirely satisfactory regarding the strength of alkali used or the development and measurement of the colored compounds produced. T h e analysis of mixtures of benzene and toluene was only approximate (14). T h e present report describes some improvements which were made in these procedures.
Experimental REAGENTS.Fuming nitric acid, specific gravity 1.50; chromic acid (CrOJ, saturated aqueous solution; potassium hydroxide, 70 per cent (one pound of pellets, dissolved and diluted to 650 ml.); and butanone. APPARATUS.Nitration of hydrocarbons is carried out in a small U-tube filled with beads. The sample is transferred to the tube from a gas buret by means of mercury if vapors are being analyzed, or, if the hydrocarbons are in aqueous solution, they may be blown over by means of a current of air. Rubber connections and greased stopcocks may not be used. A suitable lubricant for stopcocks may be made from starch and glycerol ( 3 ) . A Bausch & Lomb visual spectrophotometer with 5- and 10mm. cells was used, alt,hough any other suitable instrument could be substituted. PROCEDURE.The sample containing 10 to 100 micrograms of a mixture of hydrocarbons is collected in 2.0 ml. of fuming nitric acid in the nitration tube filled with beads. The tube and contents are heated in an oil bath a t 90" C. for 15 minutes. If benzene only is to be determined, 1 drop of a saturated chromic acid solution is added and the sample is mixed by bubbling a little air through it. If benzene, toluene, and xylene are to be determined, the chromic acid is omitted. The tube and sample are returned t o the oil bath for another period of 15 minutes. Beads and sample are transferred to a 60-ml. glass-stoppered bottle with the aid of three 2.0-ml. portions of water. The mixture is cooled in ice water and carefully neutralized with 70 per cent potassium hydroxide until the dichromate orange changes to chromate yellow-green. I n the toluene determination, where no oxidation is required, 1 drop of chromic acid solution may be added after dilution and cooling without loss of toluene. Ten milliliters of butanone are added to the neutralized sample, and the bottle is firmly stoppered and shaken vigorously. I t is placed in a water bath a t 60" C. to prevent crystallization of potassium nitrate and shaken occasionally during 5 to 10 minutes. The sample is poured into a jacketed buret kept a t about 60" C. and allowed t o separate. The beads are caught in a funnel which is fitted with a wire to prevent closure of the stem by the beads. The lower aqueous layer is carefully withdrawn and discarded and the upper butanone layer containing the nitrated sample is drained into a 25-ml. glass-stoppered cylinder. Ten milliliters of 70 per cent potassium hydroxide are added and the mixture is shaken vigorously for 2 minutes. A motor-driven shaker has been found convenient if a large number of analyses are required. The colored butanone layer is allowed to separate and is transferred to the cell of the photometer. If the color is too dark for satisfactory measurement, the butanone may be quantitatively diluted or a shorter cell may be
252
1.40
1.20
1.00 2.
2c
80
W
0
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 15, No, 4
or potassium nitrate (14); therefore, the nitro compounds formed during nitration of aromatic compounds must be extracted from the nitration mixture. Butanone is fairly soluble in acid solutions, and nitro compounds are only partially extracted from alkaline solutions of pH greater than 10. The most convenient indicator to use for adjusting the pH for extraction is the chromic acid solution used as oxidant in differentiating benzene from toluene. -4 sharp change from orange to yellow-green occurs a t about pH 7.
.6 0
Factors Affecting Color Development According to the mechanism of the re.40 action proposed, the alkali serves several functions: converts ketone to enol form, converts nitro compound to aci-nitro salt, .2 0 and removes water formed by condensation. Figure 2 shows the effect of increasing .o0 concentration of alkali on density of color 500 520 540 560 580 600 620 640 660 680 obtained from nitrated benzene and toluene, MILLI-MICRONS Owing to losses of butanone during extraction without corresponding losses of nitro COMPOUNDS FIGURE 1. SPECTRAL ABSORPTIOXCURVES O F BUTAXONE-XITRO compound, the densities were too high. A T N B . 1 3 5-Trinitrobenzene D N B . m-Dinitrobenzene correction was therefore made and these values T N T . 2,’4:6-Trinitrotoluene D N T . 2,4-Dinitrotoluene were plotted. The increase in density obtained by increasing concentrations of alkali is best explained as an increase in efficiency of substituted. Owing to fading of the colors, not more than 5 to 10 the alkali as a dehydrating agent. Almost the maximum minutes should elapse between the shaking of the mixture and density is obtained with 50 per cent alkali and dinitrobenzene, the reading of the instrument. whereas only about one third of the maximum density is A calibration curve should be pre ared relating milligrams of obtained with dinitrotoluene and this concentration of alkali. hydrocarbon to color density for eac: substance determined, and since losses occur in the several steps, standards must be treated Further confirmation of the dehydrating action of the exactly s! the samples are treated. Color density = alkali is shown in Figure 3, where the larger volumes of 70 per 1 cent alkali solution give greater color densities than the log transmission’ smaller volumes. Since the density-millieram curve is linear. it is convenient to Potassium nitrate inhibits color development, as shown in express the relatioi by a Pactor. Figure 4. Seventy per cent potassium hydroxide solution Table I gives the factors calculated when nitro compounds was saturated with dry potassium nitrate and the resulting and hydrocarbons are analyzed by the procedures described. They are expressed as density per milligram of hydrocarbon for a 10-mm. cell. The corrected values are obtained by I I I I making allowances for loss of butanone in extraction and loss of water in dehydration. Comparison of these corrected 1.6 0 values with those obtained with dry butanone standards gives a measure of the losses of nitro compound sustained in nitraD DN NB 8 tion and extraction. The small increase shown upon heating 1. 4 0 nitric acid solutions of dinitrobenzene indicates that traces of COR. mononitrobenzene are present in the preparation. The dinitrotoluene sample is purer, as judged by the identity of the 1.20 values for heated and unheated samples. Some interesting comparisons regarding the nitration of > DNT c benzene and toluene can be made. The values obtained indiv) 1.00 cate that 86 per cent of benzene and 76 per cent of toluene are z W converted to m-dinitro compounds by fuming nitric acid when 0 DNT COF; liquid hydrocarbons are used. The corresponding figures for .80 the vapors are 77 and 78. The losses may be due to the formation of ortho and para isomers and to oxidation. That no appreciable loss was due to inefficient trapping of vapors by .60 the nitration tube was shown by introducing a second tube in series with the first. .40 ABSORPTION CURVES. Figure 1 shows the absorption 20 40 60 80 curves for several of the compounds studied. The wave length selected for quantitative comparison was 560 milliPER C E N T POTASSIUM HYDROXIDE microns. PREPARATION OF BUTAXONE EXTRACTS.It is not possible FIGURE2. COLORDENSITYAT VARIOUSCoxCENTRATIONS OF POTASSIUM HYDROXIDE to develop these colors quantitatively in the presence of water
-
i-c
A N A L Y T I C A L EDITION
April 15, 1943 TABLEI.
COLORDENSITYPER MILLIGRAM OF HYDROCARBON -Benzene-
Standards
++ + ++ +++ ++
Butanone m-dinitro m-dinitro (unheated) "Os m-dinitro (corrected) "0s m-dinitro (heated) "Os m-dinitro (corrected) "Or "0; liquid (heated) "0s liquid (corrected) vapor (heated) HNOs "0s v a p x (corrected) trinitro Butanone
180
30.3 *0.7 32.1 * 0 . 8 26.3 * 0 . 7 35.2 * 1.4 28.8 * 1.1 30.1 *0.5 24.7 * 0.4 27.0 * 1.6 22.1 1 1 . 3 8.8 1 0 . 3
After CrOa
...
...
33.4'2 1.6 27.4 1 1.3 28.8 * 1.2 23.6 * 1.0 27.6 *2.0 22.6 1 1.6
...
--Toluene-
20.9 1 0 . 4 24.0 1 0 . 9 19.7 1 0 . 7 23.8 10.9 19.5 1 0 . 7 18.2 1 1.4 14.9 * 1.2 17.4 1 0 . 5 14.3 a 0 . 4 9.6 * 0 . 4
... ...
...
0.0 0.0 0.0 0.0 0.0 0.0
...
nitration, only the color of the benzene derivative is obtained. A paired series of mixtures was prepared by dissolving various amounts of standard benzene and toluene solutions in 2.0 ml. of fuming nitric acid. One set of mixtures received 1 drop of a saturated aqueous solution of chromic acid in each sample; the other set served as controls. Both sets were placed in an oil bath at 90" C. for 15 minutes and then diluted, neutralized, and extracted as usual. The results in Table I1 show an average error of k5.0 per cent.
Analyses of benzene and toluene vapors and of mixed solvents have shown the methods to be equally satisfactory. The differentiation probably depends upon the loss of butanone solubility of the oxidized products. Experiments with 2,4-dinitrobenzoic acid indicate that dinitrotoluene is oxidized to this substance. m-Dinitrobenzene is not oxidized under the conditions selected but is lost with permanganate oxidation and with longer heating periods with t h e oxidant chosen.
-
>
k
til 170-
z a W
160
After CrOa
253
-
Summary T h e colored compounds formed by shaking mdinitrobenzene and its homologs with 0 2 4 6 ketones and alkali are probably ML. POTASSIUM HYDROXIDE 0 0 5 10 15 quinoids formed by condenSATURATED WITH KNOj sation of the aci-nitro and enol ML. POTASSIUM HYDROXIDE F I G ~ E4. EFFECT OF POTASSIUM forms with the eliminetion of FIGURE3. COLOR DENSITYAND VOLUME water. OF 70 PER CENT POTASSIUX HYDROXIDE NITRATEON COLORDENSITY Various factors concerned with the production of these colors have been studied, and certain modifications recommended for their quantitative solution mixed in various proportions with 70 per cent potasdetermination. The use of 70 instead of 50 per cent alkali sium hydroxide solution. Ten milliliters of these mixtures greatly hastens and increases color production. were shaken with the butanone extract of nitrated toluene Mixtures of benzene and toluene have been satisfactorily and the densities determined. analyzed by oxidation of dinitrotoluene, presumably to diThis experiment emphasizes the importance of making a nitrobenzoic acid, which gives no color under the conditions clean separation of the butanone extract from the aqueous chosen. m-Dinitrobenzene is not attacked and gives its phase in the determinations. It also supports the need for usual color with undiminished intensity. larger volumes of potassium hydroxide solution used to develop the colors than were recommended by Pearce et al. (IO). Literature Cited Differentiation of Benzene and Toluene (1) Bitto, B.von, Ann., 269,377 (1892). (2) Boat, R.W., and Nicholson, F., IND.ENG.CHEM..ANAL.ED., 7, When mixtures of benzene and toluene are nitrated with 190 (1935). (3) Herrington, B.L.,and Starr, M. P., Ibid., 14,62 (1942). fuming nitric acid, heated, diluted, neutralized, and extracted (4) Jackson, C. L.,and Earle, R. B., Am. Chem. J.,29,98 (1903). with butanone, the density obtained is equal to the sum of (5) Jackson, C. L.,and Gazzolo, F. H., Ibid., 23,376 (1900). those obtained with the individual substances separately. (6) Janovsky, J. V., Ber., 24,971 (1891). If, however, the mixture is oxidized with chromic acid after (7) Janovsky, J. V., and Erb, L., Ibid., 19,2158 (1886). 150 -
Lu
ASD TOLUENE EIIXTURES TABLE11. ANALYSISOF BENZENE
Sample
Benzene
Taken
Toluene
Micrograms
Found Benzene Toluene Micrograms
(8) Kay, Kingsley, Can. J. Research, 19,86 (1941). (9) Meisenheimer, J., Ann., 323,224 (1902). (10) Pearce, S. J., Schrenk, H. H., and Yant, W. P., U. S. Bur. Mines, Repts. Investigations 3302 (1936). (11) Reitsenstein, F.,and Stamm, G., J.prakt. Chem., 81,168 (1910). (12) Schrenk, H. H.,Pearce, S. J., and Yant, W. P., U. S. Bur. Mines, Repts. Investigations 3287 (1935). (13) Whitmore, F. C.,"Organic Chemistry", p. 731, New York, D. Van Nostrand Co., 1937. (14) Yant, W.P.,Pearce, S. J., and Schrenk, H. H., U. S. Bur. Mines, Repts. Investigations 3323 (1936). (15) Yant, W. P., Schrenk, H. H., and Manta, P. H., Ibid., 3282 (1935).
