Spectrophotometric Determination of Primary Amines in Aqueous Solution with Copper(Et hylenedinitri lo)tetraace tic Acid IRVIN M. CITRON and ALLAN MILLS Department of Chemisfry, Fairleigh Dickinson Universify, Rutherford, N. I.
b The absorption of the copper-EDTA complex in an aqueous medium at 720 mp is considerably diminished by the addition of small quantities of primary amines. This phenomenon is used as the basis of a simple spectrophotometric scheme for the quantitative determination of primary amines in aqueous solution. A properly diluted portion of the primary amine is added to a known quantity of the copperEDTA complex, and the absorption at 720 mp is measured and located upon a previously or simultaneously prepared plot of absorbance vs. ratio of molar concentrations of primary amine to copper-EDTA, or simply absorbance vs. molar concentration of primary amine (since the concentration of copper-EDTA is constant). The molar quantities of primary amines can be accurately determined from the plot.
A
SPECTROPHOTOMETRIC
STUDY Of
the copper-EDTA complex was carried out by Kirson and Citron (6), who discovered that its intense absorption at 720 mp is considerably diminished by adding small quantities of primary amine. The above authors postulated that the complex between Cu+* and the tetrasodium salt of EDTA, Xa4Y (where Y represents EDTA), exists as a polymer in solution. This complex is stable within the p H range 4.0 to 11.0 as proved by the addition of HC1 or XaOH to change the p H of the aqueous medium. However, when primary amines are added, two carboxyl linkages to the central copper ion of each cell of the polymer are disrupted as the amine enters the complex t o form the monomer. Kirson and Citron also found that secondary amines disrupt the polymeric Cu-EDT4 complex to a far lesser extent than do primary amines, and that tertiary amines have very little effect upon the absorption at 720 mp. Steric factors, it seems, prevent the approach of secondary, and especially tertiary, amines to the central copper ion. Citron continued the investigation (1) with a study of the copper-EDTA 208
ANALYTICAL CHEMISTRY
complex in the infrared, and proved that the primary, secondary, and tertiary amines do coordinate to the central copper ion in diminishing extent as listed above. The coordination strongly affects the infrared spectrum of the copper-EDTA in the 2.5- to 15-micron range. The present investigators decided to study the analytical possibilities of these effects. The first step was to perfect a simple but accurate spectrophotometric method for the quantitative determination of primary amines in aqueous solution without the presence of other kinds of amines. This analytical procedure has been completed and is described in this paper. EXPERIMENTAL
Apparatus a n d Reagents. The spectrophotometer was a Beckman Model D B with recorder and scale expansion accessory. The absorption cells were Beckman standard silica U cells, 1.0 em. in diameter. The reagents employed came from several sources Cupric sulfate, anhydrous, Baker
Figure 1 . Spectra of a series of standard solutions Ratio
Solution
[n-CsHNHzl l(CuYI-~l
A
0.000
6
0.125 0.250
C D E
F G
0.375
0.500 0.600 0.700
H
0.750
I
0.800
analyzed reagent (assay CuSO4, 99.i7,). (Ethylenedinitrilo) tetr a a c e t i c a c i d , tetrasodium salt. Natheson Coleman and Bell. N-Propylamine, M'atheson Coleman and Bell. b.u. 48-49.5' C. Ethylamine (anhydrous), Eastman, m.w. 45.09. n-Butylamine, Eastman, b p. i 6 78" C., m.m. 73.14. Benzylamine, Eastman, b.p. TO71' C./lO mm., m.w. 107.16. 2-Aminoethanol, Eastman, m.p. c310" C., m.m. 61.09. 3-Amino-1-propanol, Natheson Coleman and Bell, b.p. 60-62' C., 2 mm., m.w. 75.11. Procedure. Froin 2 buret or nith a pipet place 4.0 ml. each of R C C U rately prepared solutions of 0.1M CuS04 and 0 . l M EDTA-tetrasodium salt into a series of six or seven 25-ml. volumetric flasks. Prepare accurately a 0.1M aqueous solution of primary amine, preferably the amine to be determined quantitatively. Into the above series of 25-ml. volumetric flasks enter 0.5-ml. increments of the standard primary amine solution, starting with 0.0 ml. and ending w t h 2.5 or 3.0 ml. For each unknown amine solution to he determined, prepare a 25-ml. volumetric flask as above, putting 4.0 ml. of 0.1X CuSO1, 4.0 ml. of 0.1M Nary, and a selected amount of the unknown amine into the flask. Dilute all the flasks to the mark. Make sure that the final concentrations of unknown primary amine (in the solutions to be examined) do not exceed 0.012M. Examine all the solutions in the above series with a colorimeter or spectrophotometer at 720 mp. If a recorder is used, record only a small segment of the visible spectrum, say from 760 to 680 mp. Plot the value of absorbance us. the molar ratio [primary amine]/[(C~Y)-~] or absorbance us. [primary amine] for the known concentrations in the above series. A straight-line plot should result up to a ratio of [primary amine]/ [ (CuY)-*] = 0.75. Locate absorbance values for the unknowns in the series on the straight line, and read off the molar concentrations of primary amine directly, or the molar ratio of primary amine to [(CuY)-*] in the solutions examined. Take into account dilution factors employed, and calculate the molar concentrations of primary amines in the original solutions of the unknowns.
Table 1. Precision of Method in Presence of Foreign Anions
t
[.4nionl, 0.0080. [(CuY)-21, 0.0160. L~-C3H7NHz], 0.0080. [~-CSH~NH / I] [(CUY)-~],0.500. S o . of samples, 5.0. Mean absorbance, no foreign anion present, 1.300. Mean valuea dnion of absorbance, Deviation present A in A , % CzHaOz-2
1.01
t
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0
1.6
3.2
4.8
6.4
8.0
9.6
11.2
12.8
14.4-Cn.
Figure 2.
Plot of absorbance vs. ratio [n-C3H,NH2] and absorbance vs. [n-CzH,NH2]
I n the foregoing procedure it is suggested that the primary amine under analysis be used in making up the series of standard solutions Primary diamines and polyamines such as ethylenediamine and diethylenetriamine not only loner the absorption of the Cu-EDTA2 complex but also shift the wavelength of m:tsinium absorption from 720 mp to shorter wavelengths. The quantitative analysis of diamines and polyamines will be investigated and discussed in a ,separate paper, as will the analysis of important biological amines (3). Substituted groups such as -0Hand -C00Hmay or may not have an effect upon the coordination of the amine with the Cu-E DT.2. This would depend, of course, upon the location of the substituent group in the amine molecule. Substituents separated by a t least one -CH2-- group from the primary amine group in the molecule should not interfere. In these cases, for best results the m a l j s t should use the amine being detumined in making iip the series of standard solutions. The spectra obtained with a recording spectrophotomett,r within the range SO0 to 680 mp for part of a series of standard solutions ucing n-propylamine are shown in Figure 1. Figure 2 shows the straight -line relationship obtained, and its limits, when the absorbance is p1ottl:d us. the molar ratio [n-propylamine ] / [(CuY)-*]or us. [n-propylamine]. A series of tests on replicate samples was run to determine the precision of this analytical method. Again n-propylamine was used. The concentration of [(CUY)-~]was held a t 0.016M in all four series, but the ratio [n-GH~NH’2]/[(CuY)-*] was different for each.
Summary of Results [n-C3Hill;H2]/ [(CuY)+] = 0.100. For six replicate samples, the mean concentration of aC3H7KH2was 1.6 X lO-3-U; the range, 0.40 x 10-3-lf; the standard deviation and the relative standard 1.38 x deviation, 8.67,. [n-C3H7PiH2]/ [ (CUY)- 2 ] = 0.250. For 10 replicate samples, the mean concentration of n-C3HiSH2 was 4.00 X lO-3Jf; the range, 0.60 x 10-3Jf; the and the standard deviation, 1.92 X relative standard deviation, 4.8%. [n-C3HiSHz]/[(CuY)-2] = 0.375 For 10 replicate samples, the mean concentration of n-C3Hih-Hnv a s 6.00 X lO-3-V; the range, 0.16 X 10-3LlI; the standard deviation, 0.57 X and the relative standard deviation, 0 . Q 5 ~ o .
i
0
Figure 3.
0.17
[n-C3H,NH2]/ [(GUY)-*] = 0.500. For 10 replicate samples, the mean concentration of n-CaH7NHr was 8.02 X IO-3-U; the range, 0.44 x 10-3JI; the and the standard deviation, 1.48 X relative standard deviation, 1.8%.
[ (cuY)-21 RESULTS AND DISCUSSION
0.8’’
1.205 0.41 a .4t absorbance value 1.200 a 1.070 error in determining A would mean a 6.0% error in determining [n-CaH,NHJ.
-’- ’ ‘‘cuyl’
0.17
1.202 1.210 1.202
co,
The above results compare favorably with the commonly used nonaqueous titration method for determining primary amine. Moreover, this spectrophotometric procedure is much less tedious and time-consuming than the nonaqueous titration procedure (2, 4). Sample preparation is brief, and an entire set of unknowns can be examined in one series of brief spectral runs. The presence of most foreign inorganic and some organic anions in the amine solution does not interfere in the analysis. Data to verify This are
I
I
I
I
I
I
I
I
I
I
I 0.1
I
1
I
I
0.3
I 0.5
I
0.2
I 0.4
0.6
0.7
0.8
I 0.9
I 1.0
Effect of various primary amines on absorbance of
1
t I
[(CuY)-*]
CzHsNHa 2. o-CIHBNHZ 3. CsHsCHzNHz 4 . HiN(CH2)sOH 5. HzN(CH2)zOH [(CuY)-*] = 0.016 1.
VOL. 36, NO. 1, JANUARY 1964
209
shown in Table I, the anions tested being NOa-, COS+, C1-, and CzHsOn-. Cations that complex with EDTA would interfere, and would have to be separated from the solution (by ion exchange or other techniques) before amine analysis, if present in sufficient amount to affect the copper-EDTA complex. Tolerances for these cations might be large in some cases and small in others, depending upon the strength of the foreign metal-EDTA complex. For most foreign metals, no trouble should arise if the foreign metal concentration in the solution examined is less than 0.0016MJ or one-tenth the molar concentration of the copperEDTA complex in the solution examined. In developing this analytical method the authors have purposely used equimolar amounts of C U + and ~ EDTA to ensure the highest degree of reproducibility. Small excesses of C U + ~ or EDTA could most likely be tolerated without seriously affecting the results. However, a considerable excess of C U + ~ would tend to give low results because of coordination between Cu+z and primary amine. On the other hand, a large excess of EDTA would be expected to yield high results because
of its tendency to decrease the degree of polymerization of the Cu-EDTA polymer. The use of a controlled excess of EDTA to eliminate certain cationic interferences is a definite possibility which will be investigated. Data are also included (Figure 3) to show that this method is generally applicable to all primary amines. However, except in cases of very similar primary amines, such as CaHsNHz and n-CsH,NHn which have the same slope for the curve of absorbance vs. [primary amine]/ [(CuY) the amine being analyzed should be used in making up the standard solutions. Figure 3 shows that the limits of linearity for the curves of absorbance vs. [primary amine]/ [(CuY) -21 for the various amines tested are about the same. For 0.016M [(CUY)-~]the ratio [primary amine/ [(CUY)-~]should not exceed 0.75 to ensure results within the limits of reliability presented here. Among the primary amines tested were two amine alcohols, 2-aminoethanol, and 3-aminoI-propanol. From the slopes of the curves for these two compounds in Figure 3 it is evident that the -OH functional group in HgN(CH2)zOH interferes with the coordination of the amine to the Cu-EDTA complex to a greater
extent than in HzrYT(CH2)30H where it is further removed from the amino group. Other correlations between the kinetics of amine coordinations to CuEDTA and the structures of the amines might be possible by thorough study of more absorbance curves of this type. ACKNOWLEDGMENT
The authors are indebted to Harold Weinberger, Chairman of the Department of Chemistry, to Kathleen Hillers, and to the faculty of the Department of Chemistry a t Fairleigh Dickinson University for their advice and encouragement during this research. LITERATURE CITED
(1) Citron, I., Anal. Chim. Acta 26, 44657 (1962). ( 2 ) Day,. R . A,, Underwood, A. L., “Quantitative Analysis,” pp. 88-9, Prentice-Hall, Englewood Cliffs, N. J., 1958. (3) Erspamer, V., Progr. Drug Res. 3,152367 (1961). (4) Fritz, J. S., “Acid-Base Titrations in Non-Aqueous Solvents,” G. F. Smith Chemical Co., Columbus, Ohio, 1952. (5) Kirson, B., Citron, I., Bull. SOC. Chzm. France 1959,365-9.
RECEIVED for review July 9, 1963. Accepted October 24, 1963.
Quantitative Isolation and Improved Nephelometric Microdetermination of Lupine Alkaloids from Plant Tissues DONALD M. GRAHAM and MARY SPENCER Departments o f Biochemistry and Plant Science, Universify o f Alberta, Edmonton, Canada
b A simple method is described for quantitative isolation from plants of submicromole quantities of lupine alkaloids and for their purification, separation, and estimation. The method involves extraction with water, removal of the alkaloids from the extract with a phosphonic acid cation exchanger, followed by elution with acid, pH adjustment, solvent extraction, and low-pressure removal of solvent to yield the alkaloid salts in crystalline form. Separation is achieved by descending paper chromatography using a solvent system based on acetic acid, tertiary amyl alcohol, and water. The alkaloids are determined nephelometrically after elution from the paper. This procedure can be used successfully when the tissue investigated contains as little as lo-* mole of any one alkaloid.
0
NE OF THE CHIEF DIFFICULTIEB
in biochemical analysis of plant alkaloids is that most of the classical extraction procedures, even those de210
0
ANALYTICAL CHEMISTRY
scribed as quantitative, do not ensure complete extraction of alkaloids from the tissue; purification procedures are usually tedious and require large quantities of materials; transfer losses occur readily; and conventional analytical methods are often difficult to use with minute quantities of material (4). Some attempts, notably by Lee (6), Mattocks (7), and Tompsett (IO), have been made to use ion exchange resins for isolation of alkaloids from plant tissues and to a lesser extent for separation of alkaloids from one another, but these methodfi are of limited usefulness, and in particular, none is suitable for lupine alkaloids. Lee’s method, originally designed for the fractionation of opium, necessitates beginning with fairly large (gram) quantities of alkaloid that have already been isolated in crude form; Mattocks’ procedure is useful only for very large quantities and all three employ Dowex 50, which we have found to bind lupine alkaloids irreversibly. Acid cannot be used for elution, even a t negative p H
values, because this resin absorbs alkaloid from such solutions, and basic solutions are unsuitable since several of the alkaloids are unstable or insoluble, or both, in the unionized form. We discovered also that eshaustive extraction of the plant tissue with various solvents in common use for this purpose was usually not in fact quantitative. Furthermore, solvent partition methods, when used alone, are unsatisfactory for the quantitative isolation of alkaloids from plant tissues because they tend to be tedious, and involve large transfer losses as well as difficulties due to the formation of troublesome emulsions. The present method, on the other hand, permits of high extraction efficiency, gives a crystalline product, allows clean separations, and can be used with minute quantities of alkaloida (small fractions of a micromole). WPERIMENTAL
Apparatus. Beckman Model B Spectrophotometer with 1-cm. matched Pyrocell quartz cells.