V O L U M E 21, NO. 7, J U L Y 1 9 4 9 Tors incurred in these separations and determinations are within the same range as those found in the determination alone. Values obtained by this method for some citrus juices are compared with values obtained by the use of the pentabromoacetone method of the Association of Official Agricultural Chemists ( I ) in given is an average Of four ali17. Each fluorometric quots.
813 LITERATURE CITED
(1) Assoc. Offic. AD. Chemists, “Official and Tentative Methods of Analysis,” 6th ed., pp. 391-4, 1945. (2) Feigl, F., Anger, v., and Frehden, O., ,Tfilzrochemie, 17, 35 (1935). RECErVED September 20, 1948. Abstracted f r o m a portion oE a thesis submitted by Sidney Kats in partial fulfillment of the requirements for the Ph.D. degree.
Diphenylamine Test for Nitrates in Mixtures of Cellulose Esters A. G. ROBERTS, National Bureau of Standards, Washington, D . C . Diphenylamine in sulfuric acid is a sensitive reagent for indicating nitrates by producing a blue coloration. Earlier investigations were limited to nitrates in very dilute aqueous solution. Uncertainty exists concerning its use in the higher nitrate range and upon solid materials. An investigation was made of the speed and strength of color development when solutions containing diphenylamine and water were brought into contact with fiIms cast from mixtures of cellulose esters covering
D
IPHENYLAMINE in sulfuric acid has long been known ( 6 ) as a sensitive reagent for indicating the presence of nitrates by the production of a blue coloration. Because the blue color results from the oxidation of the diphenylamine, strong oxidizing agents such as nitrite, chromate, ferric salts, etc., interfere (6). The work of earlier investigators (3, 7, 10, 11) indicated that the sensitivity of the test depends upon the concentrations of diphenylamine and sulfuric acid. They employed reagents containing from 0.008 to 0.67 gram of diphenylamine and from 0 to 150 ml. of water per 100 ml. of concentrated sulfuric acid. These investigations were limited to the detection of very small amounts of nitrate in highly dilute aqueous solution, of the order of 1 mg. of nitrate nitrogen per liter (0.0001%). No investigation appears to have been made of the broad higher nitrate range (from 0.001 to 12.0% nitrogen) or of solid materials such as mixtures of cellulose esters, nor has the time for production of color been emphasized as an effective quantitative criterion. Standard reference works such as the “Modern Plastics Encyclopedia” (8), the “Handbook of Chemistry and Physics” (4), and others (2, IS) are a t variance with regard to the indicator reagent compositions recommended. It was the purpose of this investigation to determine the optimum concentrations of diphenylamine and sulfuric acid in the diphenylamine indicator solution for general use in the detection and estimation of nitrates in cellulosic films having a wide range of nitrate nitrogen compositions. EXPERIMENTAL
A series of diphenylamine indicator solutions was prepared covering the range of 0.01 to 1.0 gram of diphenylamine per 100 ml. of concentrated (96y0) sulfuric acid, and a range of 0 to 150 ml. of Kater per 100 ml. of concentrated sulfuric acid. These were evaluated with films of various nitrate contents, ranging from 12% nitrogen (as determined by a Kjeldahl titration) in.a film composed wholly of cellulose nitrate, to 0.001% nitrogen in a film consisting of a 1 to 12,000 mixture of cellulose nitrate and cellulose acetate butyrate. The cellulose nitrate employed was a dope grade complying with Army-Savy specifications (1). The mixed nitrate-butyrate compositions were obtained by casting films from homogeneous acetone solutions of cellulose
a wide range of nitrate compositions. The color development time is a minimum in the region of 2 to 6% nitrogen and increases at higher and lower concentrations. Water content is important; diphenylamine concentration exerts a minor influence. A suitable reagent for general use over a wide nitrate range contains 0.1 gram of diphenylamine, 100 ml. of concentrated sulfuric acid, and 30 ml. of water. Quantitative estimates are possible when films of known nitrate composition are used for comparison. nitrate and cellulose acetate butyrate in the desired ratios. The latter was selected as the diluent because it is the standard material for doping fabric surfaces of naval aircraft (9) and is widely used in the aircraft industry. The diphenylamine solutions were prepared by suspending the diphenylamine in the water to be added, then adding the concentrated acid. The heat of mixing was thus utilized to effect rapid solution of the diphenylamine. (Maximum sensitivity is achieved only with freshly prepared reagents. The solution gradually deteriorates and should be discarded when it no longer produces an adequate color with a standard nitrate sample.) The compositions investigated are indicated in Table I. The time required for a particular color to develop after a single drop of the indicator solution was placed on a test film was found to depend in a consistent manner upon the nitrate content of the film. In order to standardize the time measurement, so as to permit visual resolution of nitrate composition differences among the samples, it was found convenient to use two arbitrarily selected reference colors-viz., a light blue (about 2.5PB 5 / 8 Munsell) and a deep blue (about 5.OPB 3/8 Munsell). I n the extremes of the nitrate compositions tested, where relatively little color is produced, the time required for the appearance of the first observable blue coloration was a useful criterion. Figure 1 shons the relation between nitrate nitrogen content and the color development time, using the several reference colors, with indicator reagents containing 30 and 50 ml., respectively, of water per 100 ml. of concentrated sulfuric acid. -4semilogarithmic plot has been employed solely for convenience in treating the data. Although the data presented are not inTable I. Compositions Investigated to Determine Sensitivity of Diphenylamine Test for Nitrate Radical Water added t o sulfuric acid, ml. per 100 ml. pf H+04 (96%) Sulfuric acid in reagent, Z ’ by weight Diphenylamine in reagent, grams per 100 ml. of &SO4 (96%) Nitrate nitrogen in film, %
0,10,30,50,80, 150 96,91,82.5, 75.5, 67, 53 0.01,0.05, 0.10, 0.50, 1.00 0.000, 0.001, 0.005,0.015, 0.025, 0.050, 0.100, 0.185, 0.36, 0.71, 1.3, 2.4, 4.8, 7.2, 9.6, 10.6, 12.0
ANALYTICAL CHEMISTRY
814
t,eiided to be an absolute basis for analysis, they are typical and serve to demonstrate how markedly the time required for color development depends upon the amount of nitrate present. Through use of a suitable reference color-i.e., one for which there is a relatively large change in the time required to develop the reference color for small changes in the nitrate content, as exemplified by t’he curves in Figure 1 with the steepest slopesit is possible to resolve dxerences in nitrat,e content over a wide range of compositions with a single general-purpose indicator solution; estimation of the amount of nitrate present then becomes practical by comparing the unknown sample with films of known nitrate content. The usefulness of a particular reference color is determined by the range in which it may be applied. Thus it is evident. from Figure 1 that although the time to develop a deep blue color is the most satisfactory criterion for resolution of differences in nitrate composition in the broad middle nitrate range, its use is Limited by the insufficiency of color developed in the extremes of nit,rate content, where the time for the first, observable color to devdop becomes practical as a criterion. DISCUSSION OF RESULTS
,
Effect of Water Content. Indicator solutions of several diphenylamine concentrations and covering the range of water contents indicated in Table I were test,ed over the nitrate range from 2 to 11% nitrogen. The concentrated sulfuric acid solution produced charring of the film rather than a blue coloration. The “10-ml.” solution-i.e., containing 10 ml. of water per 100 ml. of concentrated sulfuric acid-gave a dirty violet color and erratic results. B pure blue color and consistent results were obtained with the “3O-ml.” solution. Good results were also obtained with the “50-ml.” solution, but strength of color and sensitivity were Lower than with the “30-ml.” reagent (Figure l), and the useful range was not so great. S o color whatever was obtained with either the “80-ml.” or “150-ml.” reagents. On the basis of the above data, 30 ml. of water per 100 ml. of concentrated sulfuric acid were selected as a suitable amount of water to use in the diphenylamine indicator reagent. Such a reagent contains approximately 82.5% sulfuric acid by weight. Effect of Diphenylamine Content. Indicator solutions having the concentrations of diphenylamine given in Table I and containing the amount of m t e r as determined above were investigated with films in the 2 to 11% nitrogen range. Saturation of color and the rate a t which it developed increased with increasing diphenylamine content within this ranyr. .I similar series of tests conducted in the low nitrogen range revealed a somewhat greater sensitivity and a lessening of the color development time with decreasing diphenylamine content for compositions beloa0.02% nitrogen. With the higher diphenylamine concentrations investigated the reference color was attained so rapidly as to make difficult the resolution of differences in nitrate composition. On the othei hand, the solution containing 0.01 gram of diphenylamine per 100 ml. of concentrated sulfuric acid was low in sensitivity. The intermediate reagent containing 0.1 gram of diphenylamine per 100 ml. of Concentrated sulfuric acid plus 30 ml. of water was selected as a general-purpose indicator which would he suitahlc for use over a wide range of nitrate content. Effect of Nitrate Content. The general-purpose indicator was evaluated with films covering the entire nitrogen range listed in Table 1. The results are shown graphically in Figure 1. Color was found to develop faster with increasing nitrate content up to about 2% nitrogen. There was no appreciable change of color development time with nitrate content in the 2 to 6% nitrogexi range. Above approximately 6% nitrogen, color developed more slowly with increasing nitrate content. Because for any given color development time there are two possible values of the nitrate content, some idea of theapproximate range in which the unknown composition lies is needed in order
y:
-
0001
0 005 001
0 0 5 010
0 50
I 00
5 0
100
L
NITRATE
I-
NITROGEN,
PERCENT
Figure 1. Relation between Nitrate Nitrogen Content and Time Required for Development of a Reference Color with Diphenylamine Indicator Solutions Differing in Water Content a reagent containing 0.1 gram of diphenylamine per 100 ml. of concentrated rrulfuric acid
(:urvro shown are for
Reference color. A.
B. C.
First observable blue Light blue Deep blur
to select the applicable value. Such information, if not already known, can be ascertained by noting whether the time for color development increases or decreases when the specimen is diluted by the casting technique previously described. If, for example, the time for color development increases upon dilution, obviously the sample must’ be in the lower nitrate range. Resolution of composition differences in the 2 to 6% nitrogen range can h e accomplished by quantitatively diluting the specimen to the low>r nitrate region before comparison with k n o m standards. The test is not so sensitive with the solid materials herein described as in the nitrate solutions studied by earlier investigators. The practical lower limit of sensitivity with films of mixed cellulose esters is about 0.003% nitrate nitrogen. Values below 0.005% nitrogen have not been plotted in Figure 1 because of the poor reproducibility of results in this region. Estimation of Nitrate Content. The principles which have been described make practipable a simple and rapid procednrr for the estimation of nitrate nitrogen in a cellulosic film.
Prepare a series of film standards of known nitrate conteiit covering the nitrate range likely to be encountered. Such standards can be prepared by the diluting and casting technique already described. Once prepared, the films may be cut int,tr small test pieces and stored indefinitely for use as needed. Place a single drop of the diphenylamine indicator solutioii on the unknown and on the standard samples and note to the nearest second the time required for each sample t,o develop the reference color. The nitrate content, of the unknorl-n sample may t,hen be taken as approximately that of the standard sample whose color development time it most nearly matches. Or, if the diphenylamine reagent is applied practically simultaneously t,o the unknown and t,o t’he standard samples, the nitrate cont,ent may be estimated merely by comparing the colors among them at a given time and noting which standard is most closely matched in color by the unknown. If the latter method is employed, thr t,ime interval allowed before comparison must be long enough to develop an intensity of color sufficient t o enhance differences among the samples, and yet not so long as to obliterate such differences by the production of colors of too high a degree of‘ saturation for visual resolution. Where two values are cbtained (as the color development time has been shown to be a double-valued function of the nitrate content), the correct value can be recognized from such factors as knowledge of the manufacturing process xvhich produced the sample, its texture or appearance, it,s flammability, or its solubility in certain solvents. In t,he absence of any such information, the applicable value can be ascertained by noting whether the time for color development inrreases or decreases when t h e
V O L U M E 2 1 , NO. 7, J U L Y 1 9 4 9
815
qecimen is diluted by the techniques mentioned. When the composition lies in the less distinguishable 2 t o 6% nitrate nitrogen range, the composition may be more closely approximated by quantitatively diluting the unknown specimen t o the lower nitrate region before comparing it with the known standards
this work was performed, and to acknowledge the assistance of Honora A. Mattare in the experimental work. LITERATURE CITED
Army-Navy Aeronautical Specification AN-LC-181, Assoc. Offic. Agr. Chemists, “Official and Tentative MettlodP of Analysis,” 6th ed., p. 136, 1945. Harvey, E. M., J . Am. Chem. Soc., 42, 1245 (1920). Hodgman, C. D., ed., “Handbook of Chemistry and Physics,” p. 1343, Cleveland, Ohio, Chemical Rubber Publishing Co..
SUMMARY
Speed and strength oi color development were used ab criteria to evaluate the utility of a variety of diphenylamine indicator bolutions with a wide range of nitrate compositions in the form of films cast from mixtures of cellulose esters. The plot of time for color development versus nitrogen content has a minimum value in the 2 to 6% nitrogen region. Water content is shown to be a critical factor, and diphenylamine concentration is found to exert a minor influence. Although no single indicator solution is best for all nitrate contents, a suitable reagent for general applicability in qualitative tests over a wide nitrate range consists of 0.1 gram of diphenylamine, 100 ml. of concentrated sulfuric acid, and 30 ml. of water. The method can be used for estimating nitrate nitrogen content by comparing samples with a series ol film standards of known nitrate composition.
1947.
Kolthoff, I. M., and Noponen. G. E. disillation and oxidized to formaldehyde, after n hich it is determined colorimetrically, using a sensitive fuchsin-sulfurous acid reagent. The method is applicable to the determination of small amounts of glycosidic methox5l. If ether-linked methoxql is present, as in highly methylated glycosides and cellulose derivatives, it is necessary to plot yield of methoxyl against time and extrapolate to zero time to obtain glycosidic methoxyl values. The method is not applicable in the presence of ester-linked methoxyl or if the _. sample steam-distills.
HEX the glucose-glucose bonds in cellulose are ruptured by methyl alcohol in the presence of hydrochlorir acid, tnethoxyl groups are introduced into the cellulose ( 7 ) . In applying tho conventional microadaptation of the Zeisel methosyl method ( 2 ) to the analysis of such modified celluloses it Kas noted that glucose or purified cotton cellulosr gave blanks that amount to approximately o.3y0. Similar result,s have been observed with polyhydric alcohols ( 1 ) . These blanks would account for more than half of t,he met’hoxyl found by the Zeisel procedure in fully methanolyzed native cotton cellulose and over a third of that on niethanolyzed mercerized cellulose; hence the Zeisel method is not suited to the accurate determination of small amounts of methoxyl in these materials. Frcudenberg and Soff (4)have suggested a modification of the Zeisel apparatus for measuring acid-labile, glycosidic methoxyl. hut their method is not easily adapted to the determination of the very small amounts encountered in methanolyzed cellulose. Inasmuch as the methoxyl in methanolyzed cellulose is hydrolyzed by aqueous acid, consideration was given to the possibility of determining the methyl alcohol so produced. The technique of von Fellenberg (8) for the determination of methoxyl in pectin appeared more promising. By substituting acid for alkaline h>-drolysis and refining the Schiff colorimet,ric procedure it was surressfully adapted t o the present purpose.
.
The methanol is separated by distillation, arid oxidized to formaldehyde by means of an acid solution of potassium permanganate ( S ) , and the formaldehyde is determined by a modified colorimetric method ( 6 , 6 ) , After drastic methanolysis native cotton celluloses yielded 0.20 to 0.25% and mercerized celluloses about O.5y0 methoxyl by the present method. Closely agreeing values are also obtained by the Zeisel procedure, if they are corrected for the large blank. With simple methyl glycosides the two methods give values which are in agreement. But with methylated methyl glycosides, such as methyl 2,3,4,6-tetramethyl-@-~-glucoside and methylheptamethyl-P-cellobioside, the Zeisel method measured all the methoxyl groups whereas the new procedure gave erratic results Tvhich in some cases were slightly greater than one methoxyl per molecule. In the latter cases it was found that, in addition to the acid-labile glycosidic methoxyl, a trace of etherlinked methoxyl was reacting. Because the two types of linkages are cleaved at vastly different rates, a modified method was developed for glycosidic methoxyl in the presence of ether methoxy1 groups. The methods described here are not intended to replace the Zeisel method for the determination of total methoxyl in materials containing appreciable methoxyl, but have proved useful in measuring small amounts of acid-labile methoxyl in materials surh as methanolyzed nelliilose.
