to avoid difficulty with nitric acid in the later reduction of vanadium with zinc. Following the reduction with zinc and pH adjustment, the organic extract should have only an amber to slight red coloration. A deep red color indicates incomplete reduction of ferric iron and, therefore, the need of zinc. Although efforts were made to evaporate the organic extracts completely to dryness on a hot plate, the vanadium results were lorn. A literature search indicated that metal acetylacetonates have relatively low boiling points. The evaporation of the final 10 ml. from a water bath eliminated the loss of vanadium as the volatile metal chelate. The analytical results in Table I incorporate the factor 1.07 to compensate for a single extraction a t pH 2.0 (93% extraction from Figure 1).
Double extraction or extractions a t higher p H values were avoided to reduce the simultaneous extraction of interfering elements. Aluminum and titanium are partially extracted with vanadium; aluminum does not interfere in the peroxide color procedure for vanadium, but titanium does. Although steel samples rarely contain both titanium and vanadium, the titanium peroxide interference can be destroyed by adding ammonium fluoride. lIolybdenum, an element which ordinarily accompanics vanadium and titanium in a cupferron separation and forms a yellow peroxy complex, is not extracted by acetylacetone from a reduced solution. The method should find utility in the analysis of steels containing low amounts of vanadium
which are not suitable for potentiometric titration or to high vanadium steels when the available sample is limited. LITERATURE CITED
(1) McKaveney, J. P., Freiser, Henry, ANAL.CHEM.29, 290 (1957). (2) Morgan, G. T., Moss, H.W., J . Chem. Soc. 103, 78 (1913).
(3) Sidgwick, N. V.,
"The Chemic:! Elements and Their Compounds, Vol. I. D. 811, Oxford Univ. Press, London; 1952.' (4) Steinbach, J. F., Ph.D. thesis, University of Pittsburgh, Pittsburgh, Pa., 1953. (5) Taube, H., Chem. Revs. 50, 83 (1952). RECEIVEDfor review July 24, 1957. Accepted October 31, 1957.
Colorimetric Estimation of Tertiary and Quaternary Amines SAMUEL SASS, JOYCE J. KAUFMAN', ARTURO A. CARDENAS, and JOHN J. MARTIN Chemical Research Division, Chemical Warfare laborafories, Army Chemical Center, Md.
)Two methods are presented for the colorimetric estimation of tertiary aliphatic amines using, in the one case, the reaction of amine with aconitic anhydride and, in the other, the reaction o f amine with chloranil. When used in conjunction with one another, the methods can b e used for quantitatively differentiating between tertiary amines and amine salts or quaternary amines. The respective sensitivities for tertiary aliphatic amines are on the order o f 3 y per ml. of solution for aconitic anhydride and 50 y per ml. for chloranil.
M
were required for the estimation of and differentiation between small quantities of aliphatic tertiary amines, their amine salts, and quaternary amines a t a time when very little information was available on either micro or macroprocedures. The problem was one of determining the quantity of free aliphatic tertiary amines and amine salts in mixtures containing them in small amounts, Palumbo (3) found that cis-aconitic anhydride could be used for detecting tertiary aliphatic amines in the presence of primary and secondary amines. This qualitative procedure was adapted by Cromwell (f) to the quantitative colorimetric determination of trimethylETHODS
Present address, Johns Hopkins University, Baltimore, Md.
amine. Richter and associates (4) and Dyer (2) estimated trimethylamine from aqueous solution as a yellow picrate formed in picric acid-toluene solution. Sivadjian ( 5 ) found that chlorani1 in epichlorohydrin could be used t o distinguish qualitatively among primary (red), secondary (Tiolet), and tertiary amines (emerald green). The above methods, or modifications thereof, appeared to be possible approaches to the solution of the analytical problem. EXPERIMENTAL
The test approach followed the prescribed procedures, when existent, and adapted the available qualitative tests to quantitative techniques. As the basic problem consisted of analysis for the presence of both free tertiary amines and salts of tertiary amines, the latter as combined strong and weak salts, the methods would of necessity have to fulfill this requirement. Picric Acid System. Salts of trimethylamine as the hydrochloride, hydrogen phosphate, hydrogen borate, hydrogen sulfate, and others were found to produce varying results in the picrate-toluene system. Strong salts, such as the hydrochloride and the hydrogen sulfate, gave lorn results as amine. The others, or weaker salts, gave quantitative results. cis-Aconitic Anhydride System. The problem of interferences is character-
istic in the determination of quaternary amines or salts of tertiary amines. Possible interferences may be due to the presence of Lewis bases, such as sodium acetate and sodium phosphate, as impurities in solvents and reagents or to contamination of glassware. Most of these can be precluded by purifying the reagents and by using borosilicate glass burets or pipets and quartz absorption cells or colorimeter tubes that have been tested previously. Cromwell (1) found heating necessary to obtain maximum color. Work done in this study indicated that maximum reproducible color could be attained if the reagent was aged for 24 hours prior to use and if a 30-minute reaction time was allowed for free tertiary amine. For tertiary amine salts,- a 15-second heating time was found to be sufficient. This treatment also produced a linear relationship for amine concentrations below 5 y us. absorbance. Chloranil System. The qualitative test of Sivadjian, in which the amine is heated with chloranil in epichlorohydrin, was tested as a quantitative procedure. Known quantities of amine and amine salts were treated under varying conditions of dilution, temperature, reagent strength, and heating time. Optimum conditions were found for obtaining a linear function for concentration us. absorption. However, the tertiary amine and tertiary amine salts both appeared to VOL. 30, NO. 4, APRIL 1958
529
participate in the color reaction. The method, as a quantitative system for tertiary amines, was not a s sensitive as the cis-aconitic anhydride or picrate methods, nor would it allow for the differentiation between amines and amine salts. Other solvents for chloranil were substituted for the epichlorhydrin until one was found that showed color formation with tertiary amines but little or no effect with amine salts. Toluene was the best of this group and is the solvent used for the procedure that follows.
Klett tube, add 1 ml. of chloranil in toluene, and heat on a boiling water bath for 15 minutes. Cool for 5 minutes and measure the green color on a colorimeter using Klett filter KO,69 or on a spectrophotometer at 610 mp, Construct a calibration curve by treating 3-ml. aliquots in the range of 100 t o 800 y of amine using the procedure described above. Aconitic Anhydride Method.
Table I. Analytical Recovery of Amine from Prepared Mixtures of Amine and Amine Hydrochloride in Toluene
Probable Reactions.
H
I / /
H
O
H-C-C
1
0
I / /
Mixtures Trimethylamine and hydrochloride
H-C-C
Triethylamine and hydrochloride
I
H-C=C
Tributylamine and hydrochloride
/
REA-
GEKTS. Aconitic anhydride (Smith Organic Chemicals, Kew York), recrystallized from hot toluene solution. Prepare by dissolving 0.25 gram of reagent in 40 ml. of acetic anhydride and dilute t o 100 ml. with toluene. Age reagent 24 hours before using. Acetic anhydride, c.P., redistilled Toluene, c.P., pyrrole- and thiophene-free APPARATUS.Klett-Summerson color-
\
bNRI
cis-Aconitic Anhydride Method, Total Amine, y Added Found 1050 1045 745 740 450 447 230 234 980 970 375 373 561 564 185 183 1127 1125 877 868 251 248 750 748 627" 629
Chloranil Method, Free Amine, y Added Found 410 402 0 3 450 455 100 96 650 643 0 2 560 556 0
0
500 250 0
503 253 2 755 4
650
645
275 255 485
272 257 48 1 2 651 288
0
750
Aconitic Anhydride in Acetic Anhydride Primary and secondary amines do not show this color reaction, presumably because amide formation removes the amines.
iV-Methyldiisopropylamine
cl-o-cl
2-Diethylaminoethanol and hydrochloride
0 I1
Cl-1
I/
1-a
+ KR3-c
8
and hydrochloride
a
650 570 478 255 1042 755 648 285
655 574 475 254 1035 751 643 287
0
0
648 285
0
Tributylamine hydrofluoride.
Table II.
Analytical Results for Mixtures of Quaternary and Tertiary Amines
cis-Aconitic Anhydride Method, Total Amine, yAdded Found Mixtures 500 502 500 501 Tetramethylammonium 400 403 1000 997 _. .. chloride and trimethyl0 2 500 502 amine 250 252 250 252 700 705 Tetraethylammonium Chloranil in Toluene 920 915 chloride and triethyl350 353 amine RECOMMENDED PROCEDURES 150 151 495 490 495 482 Chloranil Method. REAGEXTS. Acetylcholine chloride and 999 995 999 993 trimethylamine Chloranil, 1% solution in toluene 500 498 1003 1010 Toluene, C.P. Remove pyrrole and 0 2 499 505 thiophene by washing with hydro500 496 496 494 Trimethylphenylammonium chloric acid, sulfuric acid, and water. 490 500 998 994 chloride and trimethylDistill Fashed material out of cal248 250 502 496 amine cium chloride. 0 3 500 495 APPARATUS.Klett-Summerson color478 Tetrabutylammonium 480 476 480 imeter with filter KO. 69 (660 to 740 424 iodides and tributylamine 984 882 465 WL) 10 495 500 0 Beckman DU spectrophotometer 476 479 476 500 Tetra-n-prop lammonium Klett tubes 940 960 454 411 iodidea a d tri-n-propylPROCEDURE. Weigh and dilute in 520 490 0 15 amine toluene, or pipet as bubbler samples in toluene, proportionate quantities of Tetraalkylammonium iodides showed erratic results. N o attempt was made to determine the mechanism of interference. The obtained results were better than semitertiary amine such that 3-ml. aliquots quantitative. represent a range of amine between 100 and 800 y. Pipet 3 ml. of sample into a 0
530
ANALYTICAL CHEMISTRY
inieter with filter Xo. 54 (520 to 580 mp) or a spectrophotometer rrith measurements made a t 500 nip Klett tubes Burets, borosilicate glass PROCEDURE. Dissolve specimens containing tertiary amine or amine salts in toluene such that 2-ml. aliquots will represent from 20 t o ‘70 y of amine. Add 1 nil. of aconitic anhydride reagent, heat for 15 seconds on a boiling m t e r bath, and allolv to stand for 15 minutes. Add 5 ml. of toluene, allow t o stand for 15 minutes, and measure the developed color on a Klett-Suminerson colorimeter using filter No. 54 or on a spectrophotometer a t 500 mp. Prepare a calibration curve by making dilutions of amine in toluene such that 2-ml. aliquots represent from 5 to 70 y of amine. PROCEDURAL KOTES. The sensitivity of the method can be increased by decreasing the quantity of toluene as a diluent. When estimating or detecting tetraalkylphosphonium salts, dissolve the specimen in acetic anhydride and dilute 6 i t h toluene. Khen determinine difficultlv soluble compounds, such & choline “chloride, tetra-n-propylammonium chloride, tetrabutylamnionium iodide, etc., use a 1 to 1 mixture of acetic anhydride and toluene for initial dilutions. Dissolve the sample in warm acetic anhydride and dilute with toluene until a 1 t o 1 ratio is reached. Perform all calibrations on this basis. SPECTRAL CHARACTERISTICS. The spectral characteristics of the color produced by the reaction of cis-aconitic anhydride and chloranil with tertiary amine were determined by means of a Cary recording spectrophotometer in the range of 400 to i o 0 mp. Figure 1 shows the absorption spectra for cisaconitic anhydride and chloranil with 2-diethylaminoethanol. The other tertiary aliphatic amines produced similar curves with these reagents. Colorinietric measurements were made by a Klett-Sumnierson photoelectric colorimeter and a Beckman Model DU spectrophotometer. In both the cis-aconitic anhydride and chloranil methods straight lines were obtained when absorbance was plotted against concentration of amine. The chloranilamine method was conducted with Klett filter No. 69 (660 to i 4 0 mp) or a t a w v e length of 610 mp on a Beckman Model DU spectrophotometer. The cis-aconitic anhydride-amine method
I
I
I 40 100
Figure 1 . a. b.
560 600 WAVE LE NGTH, mp
700
Absorption spectra
Chioranii and 2-diethylaminoethanol Aconitic anhydride and l-diethylaminoethanol
rras conducted rvith Klett filter S o 54 (520 to 580 mp) or a t a ware length of 500 mp on a Beckman Model DU spectrophotometer.
dride, but produced no color with chloranil. The similarity of tetraalkylphosphonium salts to quaternary amines suggested the use of aconitic anhydride as a means for detecting or determining the phosphonium salts. A sensitivity on the order of 10 y per ml. of solution was found for tetramethylphosphonium chloride. KO study was made on the quantitative aspects of the phosphinephosphonium salt system. CONCLUSIONS
Two improved methods are presented for the colorimetric estimation of tertiary amines when used as neutralizers or for measuring tertiary amine in air. The cis-aconitic anhydride method can be used for either amines, amine salts, or quarternary amines, with a sensitivity on the order of 3 y per ml. of solution. The chloranil method, while not as sensitive (50 7 per ml. of solution), is useful for determining free tertiary amines in the presence of amine salts.
RESULTS
Known mixtures of tertiary amines and their salts were dissolved in toluene and analyzed by the aconitic anhydride and chloranil methods. I n one case, two extremes of salt stability n-ere used. The hydrochloride of tributylaniine was used to represent a strong salt and the hydrofluoride a weak salt. Recoveries as total amine on the order of 99% were obtained in all cases as shown in Table I. Results for quaternary amines are shown in Table 11. These methods showed poor sensitivity for tertiary aromatic amines and no measurable color with amides (acetanilide, dimethylformamide, etc.). Sensitivities found with aconitic anhydride for some aromatic tertiary amines \)-ere 70 y per ml. for pyridine, 600 y per ml. for quinoline, 300 y per ml. for dimethyl aniline, and no color with triphenylamine. Chloranil showed an even lower sensitivity. The straight-chain, tertiary aliphatic amines from trimethyl through trianiylamine, including 2-diethylaminoethanol and n’-methyl diisopropylamine, showed the same order of sensitivity. Triethanolamine showed poor sensitivity with both aconitic anhydride and chloranil. Triisopropanolamine showed a sensitivity approximately 267, of that for tributylamine with aconitic anhy-
LITERATURE CITED
romwell, B. T., Biochem. J. 46, 578-
RECEIVED for review hIav 9, 1956. Accepted Sovember 4, 1957. Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., February 1956.
Constant Rate Flow Device for Electrolyte Eluents in Column Chromatography Correction
-
The circuit diagram for the electronic relay, in the article “Constant Rate Flow Device for Electrolyte Eluents in Column Chromatography” by Main, Cole, Bryant, and Morris [ANAL CHEM.29, 1558-60 (1957)] is in error. Pin 2 or pin i of the 2D21 tube should be connected to the 115-volt line a t the bottom of the transformer primary as shown in the published circuit. R. K. L ~ A I N
VOL. 30, N O . 4, APRK 1958
531
