Polarographic Determination of Amino Acids DANIEL R. NORTON', The Georgewashington University, Washington 6, D. C., and N. HOWELL FURMAN, Frick Chemical Laboratory, Princeton University, Princeton, N. 1. The reaction between amino acids and phthalaldehyde was used to determine milligram quantities of glycine, alanine, tr?ptophan, aspartic acid, lyaine, and histidine. In a borate or carbonate bnffer adjusted to pH 10.5 the reaction between aniino acids and phthalaldehyde reached completion in approximately 2 hours at reactant concentrations on the order of lO-?iM. Z'nder these conditions the suppression of the reduction current of the second phthalaldehj de wave was directly proportional to the concentration of the amino acid. Since the reaction is a general one for the amino group, the indiridual amino acids must be separated from each other prior to their determination. The amino acid must also be separated from amnionia, gelatin, and other compounds that are known to react with phthalaldelirde under these conditions. A preliminary study of the reaction between gelatin and plithalaldehyde is B i b e n . The accuracj , precision, and sensitibitj of the n~ethoclarr discussed.
B
ECAVSE niosT of the aminoacidsare not reduced at the drop-
ping merrur y electrode, a direct polarographic determination is generally riot possible. However, by reacting acids with a reagent that is reducible or that gives a reducible product, an iridirect polarographic. determination can be achieved. Wiesner ( 7 ) made a polarographic study of the reaction between aniino acids and quiriorie and found that in slight,ly alkaline aqueOUJ solutions products were formed that were reversibly reduced a t potentials considerably more negative than the potentials of the original quihoriir. -4mino acids h a w been determined by the polarographic estimation of t,heir copper complexes. Jones ( 2 ) found that the half-n.ave potentials of amino acid-copper ronipleses do not difler significantly from encsh other. Martin and Mittelniann ( 4 )applied the method to the determination of amino ncicls separated by paper chroniatoyraphj~. Zimnerman (8) described a reaction between glycine and plitlialaldehyde that could beused for thedetermination of glycine. Klein and Linser (3)modified this method to make it specific for glycine in the absence of tryptophan and ammonia. The dark green color and precipitat.e formed by adding phthalaldehyde t o n slightly alkalirie solution of glycine can be quantitatively estracted with chloroform. Colors are also produced by ammonia, tryptoplign, cystine, arginine, alanine, and asparagine, but only the complexes formed by the first two are estracted with chloroform. I'atton arid Foreman ( 5 )applied thi3 reaction to the identifirntion of glycine and t o the detection of histidine, tryptophan, 2nd ammonia ori paper chromatograni3. This paper reports a polarographic study of this reaction that led to the developmerit of a general method for the polarographic detelminatiori of milligram quantities of some typical :imino acids-n:imely, glycine, nlnnine, tryptopli:in, 3sp:irtic acid, lysine, and histidine.
lary constants were determined at a mercury column height of 400 mm. with an open circuit. The value of ni (rate of flow of mercury, mg. sec.-l) in air wa9 1.536 mg. sec.-l while the drop time in 0.1.V potassium chloride wa3 3.68 seconds. The cell assembly used in this work has been described ( 1 ) . ,4 Beckman Model G p H meter was used to measure the pH of the buffer solutions. The temperature of the solutions polarographed was controlled to 25.0' =t0.1" C. Chemicals. A pure sample of phthalaldehyde was prepared by the method of Thiele and Gunther ( 6 ) . It had a melting range of 55.0" t o 55.4" C. The amino acids were obtained from Merck & Co., Inc., and had maximum total inorganic impurities of 0.2%. The ethyl alcohol was checked polarographically and found to be substantially free from reducible Impurities. Buffer Solutions. A boric acid-borate buffer and a bicarbonate-carbonate buffer were used in these experiments. Buffer solutions approximately 0.2-44 in each buffer component were prepared by neutralizing 800 ml. of a solution containing 0.4 mole of boric acid or sodium bicarbonate with sodium hydroxide until the desired p H was obtained. The solutions were subsequently diluted t o 1 liter. Phthalaldehyde Solution. A stock solution was prepared by dissolving phthalaldehyde in ethyl alcohol and diluting to volume with distilled water t o give a solution which was 2.00 X 10-zJ4 phthalaldehyde and 15% by volume ethyl alcohol. Procedure. I n all the following experiments the concentrations of the buffer components and the ethyl alcohol were held constant. The composition of the background medium was 0.1M in each of the buffer components and 1.5%by volume ethyl alcohol. Buffer solutions containing phthalaldehyde and amino acids were deaerated and polarographed after predetermined intervals. Using the borate buffer the pH of the solutions polarographed was 10.50, while using the carbonate buffer the p H of the solutions polarographed was 10.65. Correction for the residual current of the background electrolyte was applied to the measurement of the reduction current. POL4ROGR APIIY OF PHTAA LA LDEAYDE
The polarographv of plithalaldehyde has been described in detail ( 1 ) . As the suppression in wuve height of the qecondphthalaldehyde wave is tahen as a quantit:itive measure of the amino acids in the present investigation, it is important t o review the polarographic characteristics of this reagent. Phthalaldehyde is reduced a t the dropping mercury elertrode giving two characteristic waves. The half-wave potentials become more negative with increasing pH. The magnitudes of the reduction currents are controlled by r:Lteu of reaction as well as rates of diffusion. Since
0
> 2.0 c D:
4 i 0 I 1
W
0
1.2
I W 0 -8
4 -I
*I
EXPERI.MEYT4 L
c I
Apparatus. The polarograms were recorded with a Smgent 3lodel XI1 polarograph. The sensitivity of the galvanometer wns 4.30 X pa, per millimeter. Half-wave potentials were niensured against a saturated (dome1 reference elertrode. Capil-
0 4
2 0 REDUCTION
Preaent address. U. S. Geological Survey. Department of the Interior, Washington 2 5 , D. (:.
4.0 CURRENT
6 0
.
8 0
( MICROAMPERES )
1
Figure 1. Calibration Curve for Phthalaldehyde
1116
1117
V O L U M E 26, NO. 7, J U L Y 1 9 5 4 the reactions are base-catalyzed, it is important to hold the composition of the background electrolyte constant. The half-wave potentials of phthalaldehyde in the boric acidborate buffer a t p H 10.50 are - 1.05 volts versus saturated calomel electrode for the first wave and - 1.47 volts versus saturated clrloniel electrode for the second wave. The reduction current for the first wave was 1.61 pa. per millimole per liter while the value for the second wave was 4.00 pa. per millimole per liter. The reduction currents of both waves are directly proport,ional to t lie concentration of phthalaldehyde. A calibration curve for the detcrrnination of the aldehyde in the borate buffer is shown in Figure 1. The polarographic characteristics of phthalaldehyde in the carbonate buffer are substantially the same. However, a separate calibration curve is required for each background electrolyte.
Table I.
Data from Amino Acid-Phthalaldehyde Reactions
Amino Acid
Type
Buffer
pH
Glycine Alanine Tryptophan Aspartic acid Lysine Histidine
Neutral Neutral Neutral Acidic Basic Basic
Borate Borate Borate Carbonate Carbonate Borate
10.50 10.50 10.50 10.65 10.6.5 IO, 50
Timen, Hr.
Ratiob
2.0 2.0 1.8 1.0
0.3
0.9 0.9 1.1 0.8 1.4
1.5
1.5
Time for equilibrium t o be established in, a solution of 1.00 millimolar phthalaldehyde and 0.5 millimolar amino acld as measured by reduction current of second aldehyde wave. b Ratio of moles of phthalaldehyde consumed t o moles of amino acid added in reactions under conditions described. Values calculated from slopes of calibration curves shown in Figures 4 a n d 6.
AMIhO ACID REACTIONS
Reaction with Excess of Amino Acid.
In order to determirie
tile effect of amino acids on the two phthalaldehyde waves, n
tn ofold molar excess of alanine, tryptophan, aspartic acid, lysine, nnd histidine was polarographed a t intervals in the presence of 0 001;M phthalaldehyde. In each case the first wave of the aldehyde diminished and became extended as shown by the slight change in half-wave potentials and slopes of the waves as the rcaction progressed. The second wave was completely suppressed 111 2 hours. The results of a typical experiment with alanine are shown in Figure 2. In all experiments the solutions turned yellow. In the case of aspartic acid the color was pale yellow, M hile in the case of tryptophan the color was brownish yellow. The intensity of the colors increased as the reaction progressed I n the case of lysine, a third wave, attributable to one of the prodIICts, appeared a t a half-waTTe potential between the two aldehyde wcives.
I
I
I
I
3 0
6 0
9 0
I 2 0
MINUTES
Figure 3.
Time Curve for Phthalaldehyde-Alanine Reaction
Reduction current of second phthalaldehyde wave is recorded as a function of time lapsed after mixing Conditions employed: borate buffer; phthalaldehyde, 1.00 X ,10-3 M ; alanine, 5.00 X 10-4 M
Figure 2.
Polarograms of Phthalaldehyde in Presence of Alanine
First curve represents phthalaldehyde alone while succeeding curves represent phthalaldehyde in the presence of alanine as a function of time elapsed after mixing Conditions employed. Borate buffer: phthalaldehyde, 0.85 X 1 0 - 3 M; alanine, 2.35 X 10-8 M
1. 2. 3.
0 minutes 3 minutes 13 minutes
f. 7.
33minutes 61minutes
6. 106 minutes
Reaction with an Excess of Phthalaldehyde. The time required for completion of the reaction or for equilibrium conditions to he established was determined by polarographing a t intervals n-ith a tn-ofold molar excess of the aldehyde in the presence of 5.00 X 10-4Alfamino acid. The time varied between 18 and 120 minutes for the amino acids as shown in Table I. The time curve for histidine is shonm in Figure 3. Calibration Curves for Amino Acids. Calibration curves were
obtained by determining the amount of phthalaldehyde consumed by the reaction with the amino acid. The excess aldehyde was determined by measuring the height of the second wave and referring to calibration data for the second phthalaldehyde wave. It was then possible to calculate the amount of aldehyde consumed. The data for glycine, alanine, and tryptophan are shown in Figure 4,while the data for aspartic acid, lysine, and histidine are shown in Figure 5 . The calibration curves are essentially straight lines and their failure to cross the axis a t zero indicates a constant error. This error may be the result of the suppression of the phthalaldehyde maximum by the condensation product. Condensation Products. Table I gives the ratio of the moles of aldehyde consumed t o the moles of amino acids originally present in the reaction mixture. I t is postulated that the predominant reaction in the case of neutral and acidic amino acids is an equimolar one, while the predominant reaction for basic amino acids is one involving 3 moles of phthalaldehyde for every 2 moles of amino acid. In an earlier study of the polarography of phthalaldeh7.de it x a s concluded that the concentration of free aldehyde in aqueous solution is very small compared to the hydrated forms ( 1 ) . If it is assumed that the amino acids react nith the monohydrated form of the aldehyde, an equimolar ratio for neutral and acidic amino acids would be expected. The reaction between the aldehyde and basic amino acids is more complicated but not entirely without precedence. Thiele and Giinther (6) reported that a compound was formed between isophthal-
ANALYTICAL CHEMISTRY
1118 aldehyde and ammonia in which the molar ratio of phthalaldehyde to ammonia was 3 to 2. Accuracy, Precision, and Sensitivity. The accuracy of the determination depends upon the separation of other substances that react with phthalaldehyde or that interfere with the polarographic measurements. Thus, the solutions must be free from oxygen, certain aldehydes, ammonia, and amino compounds other than the one being determined.
- 1 I. 2
D
1
I
0.4
0.4
Figure 4.
1.2
0.8 ( M O L A R I T Y I 10'1
PHTHALALDEHYDE
Calibration Curves for Amino Acids in Borate Buffer 1. 2. 3.
though the same level of accuracy and precision might be obtained in the lo-* and l O - ~ M range, it would first be necessary to study the time required for the completion of the reaction under these conditions. At reactant Concentrations on the order of 10-6.11 the difficulties of removing last traves of ammonia and the increased time for completion of the reaction would introduce difficulties. At reactant concentrations on the order of 10-2M the maximum of the second aldehyde wave would probably interfere with the accurate determination of amino acids. shorvn in an earlier study (I), most common maximum suppressors do not affect the aldehyde maximum while lowering the aldehyde concentration suppresses the maximum considerably. Gelatin is effective in suppressing the aldehyde maximum but it appears to react nith the aldehyde and therefore cannot he used. Reaction of Phthalaldehyde with Gelatin. The technique used for amino acids mas applied to a study of the reaction of phthalaldehyde with gelatin. A solution of gelatin was prepared by adding 0.1 gram of powdeied gelatin to 100 ml. of distilled Tvater and heating to 52" c. for 5 minutes. The solution n a s used immediately for the experiments described. A4t a phthalaldehyde concentration of 10-3.11 and a gelatin concentration of lO-3% a reaction was observed by the suppression of height of the second aldehyde wave and by the yellow color produced. The maximum of the second wave was completely suppressed. Further study showed that the reaction reached completion within 15 minutes. Khile the amount of aldehyde consumed was small, it could be measured accurately. As shown in Figure 6, it is directly proportional to the concentration of gelatin. The intensity of the yellow color increased with the concentration of gelatin.
Glycine Alanine Tryptophan
The method could be used for the functional group analysis of the amino group. This would be possible only when the calibration curves are superimposable. For example, it would be possible to determine the total amino group content of a solution containing glycine and aspartic acid, since their calibration curves are practically identical. I n most cases, however, the method requires the separation of individual amino acids because the slopes of their calibration curves are not generally the same. The precision of the method is comparable to other polarographic determinations, being about 2% in the 10-3.W range. Al-
I
I
I
/i
d
z I-
4 J
1
5.0
PHTHALALDEHYCE
Figure 6.
j MOLARITY
I
I OI
)
Calibration Curve for Gelatin i n Carbonate Buffer
/
c 7.
-
LL
J 4
0
1 0 6
-
"0
4
0
Z o 2 4
I
I
I
0.4
0 8
12
PHTHALALDE~YDE
i MOLARITY
a
IO'
I
Figure 5. Calibration Curves for Amino Acids 1. 2. 3.
Amino acid Aspartic acid Lysine Histidine
Buffer Carbonate Carbonate Borate
It is worth while to examine the errors that might be introduced in the interpretation of these results. While gelatin is known to suppress reduction current without a reaction taking place, it is generally conceded that this occurs only a t a gelatin concentration greater than lo-*%. I n the experiments described the concentration of gelatin was less than this value. The production of a yellow color in the solutions was similar to that found for amino acids and ammonia. While it is possible that the gelatin might have been denatured or decomposed to give these products, it is calculated that 25% of the gelatin would have had to undergo denaturation or decomposition to account for the amount of phthalaldehyde consumed and this is notbkely under the conditions employed. Assuming that the aldehyde reacted directly with the gelatin, it was calculated that one molecule of phthalaldehyde reacte with approximately three amino acid units of the gelatin molecule.
V O L U M E 26, NO. 7, J U L Y 1 9 5 4 ACKNOWLEDGMENT
Grateful acknowledgment is extended to the Research Corp. for a grant made in support of this work.
1119 (6) Patton, A. R.,and Foreman, E. hl., Science, 109,339 (1949). (6) Thiele, J., and Gunther, O., Ann., 347, 106 (1906). (7) Wiesner, K.,Chem. Zisty, 36, 313 (1942). (8) Zimmerrnan, W., 2. physiol. Chem., 189,4 (1930).
LITERATURE CITED
Sorton, D. R., ANAL. CHEM.,26, 1111 (1953). ( 2 ) Jones, T. 8. G., Biochem. J . , 42, ix (1947). (3) Klein, G.,and Linser, H . , 2. physiol. Chem., 205, 261 (1932). (4) Martin, A. J. P., and Nittelmann, R., Biochem. J., 43, 353 (1948). (1) Furman, N. H., and
RECEIVED for review December 14, 1953. Accepted .4pril 8, 1954. Presented before t h e Analytical Chemistry Section of the X I I t h International Congress of Pure a n d Applied Chemistry, New York, N. Y., 1951. Taken in part from t h e thesis presented by Daniel R. Korton t o the Faculty of Princeton University in partial fulfillment of the requirement for the Ph.D. degree, J u n e 1948.
Polarographic Study of lead in a Potassium Thiocyanate Supporting Electrolyte JAMES 0. HIBBITS and STANCIL S. COOPER St. Louis University, St. Louis,
Mo.
Because the diffusion coefficient, hence the diffusion current, of an ion being reduced at the dropping mercury electrode is a function of the medium in which the reduction is taking place, certain fundamental data must be determined with a wide variety of supporting electrolytes in order to extend the usefulness of the polarographic method. The purpose of this investigation was to examine the polarographic behavior of lead in potassium thiocj-anate. The half-wave potential, reduction step, diffusion coefficient, and diffusion current constant of lead in this medium were determined. As the diffusion current was found to be pro) 0.50 and portional to concentration ( i ~ 0 . 6 7 ~between 2.00 m M lead in 0.1Jrl thiocj-anate, the use of a thiocj-anate supporting electrolyte for the determination of lead should be of analytical interest. The solubility product constant of lead thiocyanate was determined hy amperometric titration of lead with potassium dichromate.
I
(4)that when a metallic ion is reduced a t the dropping mercury electrode, the diffusion current observed is proportional to the molar concentration of the ion, the characteristics of the capillary used, the number of electrons involved in the electrode reaction, and the square root of the diffusion coefficient of the reducible substance. Since the diffusion copfficient of the ion being reduced is dependent upon the medium in which the reduction is taking place, it is necessary that certain fundamental data be determined with a wide variety of supporting electrolytes in order to extend the usefulness of polarography as an analytical method. One of the first major steps toward supplying these data was taken by Lingane (a), who reported the behavior of a number of different ions in various supporting electrolytes. However, an extensive literature search disclosed that relatively little attention has been devoted to polarography in thiocyanate media. The present investigation was undertaken in order to extend the work begun by Lingane and supplemented by others in order that the polarographic technique moy he more widely applied to analytical problems. T WAS first demonstrated by IlkoviE
1 Present address, Carbide and Carbon Chemicals Co., Y-12, Oak Ridge, Tenn.
REAGENTS
The materials used in the experimental work were of analytical reagent grade. Stock solutions of lead were prepared from weighed amounts of dried lead nitrate. A few drops of nitric acid were added to prevent hydrolysis. The potassium dichromate solutions used for the amperometric titrations were prepared by dissolving a weighed amount of dried reagent. Potassium thiocyanate solutions were prepared by appropriate dilution of a stock solution, approximately 2 M , which had been standardized by the Volhard method (11). APPAR 4TU S
Data for the polarograms were obtained with a Fisher Elecdropode. The galvanometer and potential scale were calibrated according to the procedure given by Kolthoff and Lingane (6); the maximum sensitivity of the galvanometer was found to be 2.00 X 10-2 Fa. per division. The arrangement of the H-type cell used was similar to that shown by Kolthoff and Lingane ( 7 ) , in which a saturated calomel electrode served as the reference anode. The cell was placed in a water bath maintained a t a temperature of 25' f 0.1 O C. I n all cases, oxygen was eliminated by passing a stream of nitrogen through the cell for 10 to 15 minutes. Calibrated microburets were used in the amperometric titrations. All potentials presented are with respect to the saturated calomel electrode. EXPERIMENTAL RESULTS
Polarographic waves obtained with lead in potassium thiocyanate solution exhibited maxima when methyl red was used as a suppressor, but were well defined when 0.01% gelatin was used; consequently this concentration of gelatin was used in all cases. When solutions of lead nitrate, varying in concentration from 1 to 1 0 m X were prepared in 0.1M potassium thiocyanate, precip itation of lead thiocyanate occurred in those solutions in which the lead concentration was above 2mM. -2s there is disagreement in the literature ( 1 , 3, 10) regarding the solubility of lead thiocyanate, it seemed advisable to determine its solubility in order to find the maximum concentration of lead which may be determined polarographically, with potassium thiocyanate as the supporting electrolyte. K,, of Lead Thiocyanate. The solutions of lead nitrate and potassium thiocyanate listed in Table I were prepared, shaken for 38 hours (25' f 1' C.) to establish equilibrium, and filtered, and the diffusion currents of the filtrates were determined a t -1.0
