INDUSTRIAL AND ENGINEERING CHEMISTRY
412
ether solution, and absence of any observable ferric chloride test. Literature Cited , (1) Adams, Roger, Science, 92, 115-19 (1940). (2) Adams, Roger, Hunt, Madison, and Clark, J. H., J . Am. Chem. SOC.,62, 196 (1940). (3) Adams, Roger, Loewe, S., Pease, D. C., Cain, C. K., Wearn, R. B.. Baker. R. B., and Worn, Hans. Ibid., 62, 2566-7 (1940). (4) Adams, Roger, Pease, D. C., Cain, C. K., Baker, B. R., 'Clark, J. H., Worn, Hans, and Wearn, R. B., Ibid., 62, 2245-6 (1940). (5) Adams, Roger, Pease, D. C., Cain, C. K., and Clark, J. H., Ibid., 62, 2402-05 (1940). (6) Adams, Roger, Pease, D. C., and Clark, J. H., Ibid., 62, 21946 (1940). (7) Beam, W., Wellcome Tropical Research Lab., Khartoum, 4th Rept., (B) 25-6 (1911). (8) Blackie, W. J., IND. ENG.CHEM.,ANAL.E D . , 13, 96-7 (1941). (9) Blatt, A. H., J . Wash. Acud. AS&, 28, 465-77 (1938). (10) Bouquet, J., "Indian Hashish", League of Nations Document 0. C. 1542 (0)(1937); Bull. sci. pharrnacol., 45, 107-22, 16173 (1938).
Vol. 14, No. 5
Cahn, R. S., J . Chem. SOC.,1931, 630-8. Ibid., 1933, 1400-5. C a s ~ a r i s P.. . Phatm. Acta Helv.. 1, 210-17 (1926). Duquenois, P., and Moustapha; H. N., J . Egypt. Med. Assoc., 21, 224-7 (1938). Hoffmann-LaRoche & Co., F., German Patent 285,829 (December 5 , 1913). Ghamrawy, M. A., J . Egypt. Med. Assoc., 20, 193-208 (1937). Jacob, A,, and Todd, A. R., Nature, 145, 350 (1940). Johns, I. B., letters to the author (Oct. 20 and Dec. 19, 1938). Myttenaere, F. de, League of Nations Document 0. C. 1724 (a) and J . phaTm. Belg., 20, 341-4, 357-9 (1938). Myttenaere, F. de, League of Nations Document 0. C. 1724 (a) Addendum and J . phann Belg., 20,683-6,702-7,723-8 (1938). Myttenaere, F. de, League of Nations Document 0. C./Cannab i s l l 3 (1939) and J. pharm Belg., 22, 163-8 (1940). Nickolls, L. C., Analyst, 61, 604 (1936). Wollner, H. J., Matchett, J. R., Levine, Joseph, and Valaer, Peter, J . Am. Phurm. Assoc., 27, 29-36 (1938). Wood, T. B., Spivey, W. T. N., and Easterfield, T. H., J. C h m . SOC.,69, 539-46 (1896). Ibid., 75, 20-36 (1899).
Amperometric Titrations Amperometric Determinatian of Phosphate with Uranyl Acetate I. M. KOLTHOFF AND GUNTHER COHN, University of Minnesota, Minneapolis, Minn.
S
INCE phosphate is not reducible a t the dropping mercury
electrode, its polarographic determination or amperometric titration has t o be carried out indirectly. A reagent which forms a precipitate with phosphate and yields a welldefined diffusion current must be used for its determination with the dropping mercury electrode. A polarographic determination of phosphate by precipitation with an excess of molybdate has been described by Uhl ( B I ) , and an amperometric titration by means of precipitation with bismuth o x y h l o r a t e was attempted by Neuberger (16). Both met ods have the advantage that the precipitation is performed in strongly acid medium in which interference by coprecipitation of alkaline earth metals, etc., does not occur. The recipitation with molybdate has also the very favorable ratio o r 1 to 12 between phosphate and molybdate. Both methods, however, involve several disadvantages. The molybdate wave is badly deb e d and is much affected by changes in the acidity and composition of the solution. The precipitation procedure is complicated and requires heating to boiling, standin overnight, and filtering. Chloride, if present, has to be remove$ The titration with bismuth oxyperchlorate as described by Neuber er yields useful results only with relatively concentrated phospfate solutions (lower limit 0.035 M in phosphate). The end point does not correspond to the stoichiometric composition of the precipitate and is not easily located. Moreover, all anions except erchlorate must be absent because they form complexes with \ismuthyl ions (17).
In this paper a simple amperometric titration of phosphate with uranyl acetate is described'. Attempts were made t o titrate phosphate with lead solutions in weakly acid medium from which alkaline earth phosphates are not precipitated. Although quantitative precipitation could be obtained, the method had serious limitations and cannot be recommended. Uranyl acetate is a well-known reagent for the gravimetric (7, 14) and volumetric (potentiometric) (1, 2,9,13,18) determination of phosphate. According to Chrhtien and Kraft (3) uranyl acetate forms with solutions of orthophosphates precipitates of the composition UO1 MP04 (M = Na, K, NH4, 1/2 Ca) which are insoluble in aceAfter this paper had been submitted for publication, an important paper on the polarographic determination of phosphate was published by Stern ($0). 1
tic acid but freely soluble in mineral acids. Analogous compounds are formed with arsenate (16) and vanadate ( 1 4 , which also have been recommended for analytical purposes.
Polarographic Behavior of Uranyl Acetate The reduction wave of uranyl ion at the dropping mercury electrode was found to be greatly affected by the composition of the medium (11). Since the phosphate titration had t o be carried out in weakly acid solution, the effect of acetic acid and buffer solutions upon the current voltage curve of uranyl acetate was studied first. These measurements do not involve a n exhaustive study of the polarography of uranyl solutions, which is being carried out at present in this laboratory. The experiments described in this paper were carried out with the purpose of finding the potential and concentration range at which uranyl ion gives well-defined diffusion currents in the most suitable titration medium. Experimental The manual apparatus, described in previous publications (IO), was used. The tem erature was 25".C., and the solutions were free from air. For tge sake of convenience a pool of mercury was usually employed as the anode. Analytically pure chemicals were used. The stock solution was 0.1 M in uranyl acetate, and 0.1 to 0.18 M in acetic acid, the latter being necessary to prevent hydrolysis. In the literature different statements are found concerning the stability of uranyl acetate solutions. Courtois (4) found that both concentrated and dilute solutions were stable in the dark, but that hydrolysis occurred in the presence of light. Although according to Dworaak and Reich-Rohrwig (6) a precipitate soon appears in a 0.05 M solution, the uranyl-ion concentration is not appreciably diminished, even after several months. Sin h and Ahmad (19) found uranyl acetate to be 2.13 per cent hYf rolyaed in 0.00078 M solution at 30" C. In the present work it was found that the solutions were stable only in presence of some free acetic acid. The minimum concentration of acetic acid in 0.1 M uranyl acetate solution which yielded quick dissolution of the solid uranyl acetate was 0.1 M . CURRENT-VOLTAQE CURVES OF URANYL ACETATE IN WEAKLYACIDAKD IN BUFFEREDSOLUTIONS. In order t o get constant current readings within a reasonably short time it
May 15, 1942
ANALYTICAL EDITION
FIGURE 1. POLAROGRAMS OF 0.001 M URANYL ACETATE 1. I n 0.05 M potassium nitrate, 0.04 M acetic acid and 0.005 M ammonium acetate. 2. Same in 20% ethanol. Value; not corrected for residual current.
was found desirable to carry out the titrations in a medium of 20 per cent ethanol. Therefore, most of the current-voltage curves of u+anyl acetate were determined in this medium. The presence of alcohol also aids in suppressing or eliminating the maximum of the first uranyl wave, as is evident from Figure 1. The diffusion current of the first wave was constant over an applied voltage range between -0.8 and -1.1 volts. Solutions which were 0.002 M in uranyl acetate, 0.002 M in acetic acid, and 0.1 M in potassium chloride and 20 per cent in ethanol were not stable. The diffusion current was found to decrease with increasing time of standing. The decrease was noticeable even before a precipitate was visible (Figure 2). This hydrolysis does not interfere in titrations because i t takes place slowly, and moreover does not occur in the more strongly acid medium which usually was obtained after the precipitation of phosphate. The first wave of the uranyl ion, which is the only one obtained in the above solution, showed a slight flat maximum and a poorly defined diffusion current. At this uranium concentration the maximum was eliminated when the solution was also made 0.1 M in potassium nitrate. However, hydrolysis occurred more quickly in the presence of nitrate (Figure 2). With increasing acetic acid concentration the wave in 20 per cent ethanol and 0.1 M potassium chloride became better defined. At higher acidity the first wave did not show a maximum and the diffusion current was constant between an applied e. m. f. of 0.6 and 0.85 volt. A rise in the current started a t about -0.9 volt, the curve showing a peculiar maximum a t -1.05 volts. The shape of the curve near the maximum was dependent on the amount of acetic acid present. At the higher acidity no hydrolysis occurred and the current-voltage curves were found unchanged after various periods of standing. Figure 3 shows the current-voltage curves obtained in 0.042 M and 0.022 M acetic acid. After the peculiar maximum the current became nearly constant, but a new wave started a t about -1.25 volts. This wave was better defined in 0.042 M than in 0.022 M acetic acid. The total height a t -1.45 volts was about twice the height of the first wave a t -0.7 volt. I n the amperometric titration of phosphate we are mainly interested in the diffusion current of the f i s t wave. I n 0.042 M acetic acid and 0.1 M potassium chloride solution the diffu-
413
E oppl.
FIGURE 2 . POLAROGRAMS OF 0.002 M URANYL ACETATEIS 0.002 M ACETICACID AND 20 PERCENTETHANOL 1 2 3. In 0.1 M potassium chloride. 4, 5. In 0.1 M potassium chliride and 0.1 M potassium nitrate. Curve 1 measured 30 minutes after mixing; curve 2, 30 minutes later; curve 3. 2 hours after curve 2. Curve 4 measured 25 minutes after mixing; curve 5, 25 minutes later. Values not corrected for residual current. Cathode Dotential refers to saturated calomel electrode.
LappI.
FIGURE 3. POLAROGRAMS OF 0.002 M URANYL ACETATE 1. In 0.042 M acetic acid 0.1 M potassium chloride and 2 0 7 ethanol. 2. In 0.022 M ace& acid, 0.1 M,potassium chioride, and 20% ethanol. Values not corrected for residual current. Cathode potential refers to saturated calomel electrode.
sion current of the first wave was hardly affected by the addition of potassium nitrate or sodium sulfate. I n a medium which was 20 per cent in ethanol, 0.1 M in potassium chloride, and 0.02 to 0.05 M in acetic acid the first diffusion current was found proportional to the uranyl concentration up to concentrations of 0.0025 M . At higher concentrations the currents at various potentials were larger than expected. This, in part at least, is due to the occurrence of a maximum, as is evident from Figure 4. Addition of 0.01 to 0.2 per cent of gelatin suppressed the maximum (Figure 5), but the solutions apparently were not stable in the presence of gelatin. The diffusion current increased gradually on standing. Even if the measurements were made in the presence of
INDUSTRIAL AND ENGINEERING CHEMISTRY
414
gelatin, the diffusion current increased more than in proportion to the concentration, when the uranyl concentration was greater than 0.0025 M. This is clearly brought out by Figure 6. On the left are given the diffusion currents a t different concentrations of uranium, measured a t an applied voltage of -0.8 volt in a medium which was 20 per cent in ethanol, 0.02 M in acetic acid, and 0.1 M in potassium chloride, upon successive additions of 0.1 M uranyl acetate solution in 0.1 M acetic acid. The right-hand side gives the same plot when the solution was also made 0.02 per cent in gelatin. The diffusion current in both cases is proportional to the uranyl concentration up to 0.0025 M . When the concentration becomes larger, the diffusion current becomes disproportionately greater. The addition of gelatin in an amperometric titration of phosphate, therefore] is of no advantage. I n order to get a straight line after passing the end point the concentration of the excess uranyl acetate should not exceed 0.0025 M . On the basis of this work it is concluded that in a medium of 20 per cent ethanol, 0.1 M potassium chloride] and 0.02 to 0.04 M acetic acid an applied e. m. f. of 0.7 to 0.8 volt is suitable for the titrations. If potassium nitrate is used in place of potassium chloride, an e. m. f. of 0.9 to 1.1 volts should be applied because of the change of the anode potential (pool of mercury).
Vol. 14, No. 5
8
E appl.
FIGURE5. POLAROGRAMS OF 0.002 M (CURVE1) AND 0.004 M (CURVE 2) URANYL ACETATEIN 0.1 M POTASSIUM AND 20 PERCENTETHANOL CHLORIDE in 0.01% gelatin, 2 in 0.02% gelatin. Values not corrected for residual current. Cathode potential refers to saturated calomel electrode.
Amperometric Titration of Phosphate After many preliminary experiments, it was decided to carry out the titration in a medium containing 20 per cent ethanol. Alcohol decreases the solubility of the precipitate U02MPOdl and before the end point it considerably reduces the time
E,
I 0
0.5
0.93
I
1.0
I
I
I
I
I
_
1.5
E appl.
FIGURE4. POLAROGRAMS OF 0.002 (CURVE 1) AND 0.00373 M (CURVE2) CRANYL ACETATE IN 0.1 M POTASSIUM CHLORIDE, 0.055 M ACETICACID,AKD 20 PERCENTETHANOL Values not corrected for residual current. Cathode potential refers to saturated oalomel electrode.
The titration of phosphate has also been investigated in buffers of acetic acid and acetate in a medium of 20 per cent ethanol. Therefore] current-voltage curves of uranyl acetate were also investigated in such buffer solutions. A solution of 0.002 M uranyl acetate in 0.2 M sodium acetate and 0.022 M acetic acid and 0.1 M potassium chloride] which had a pH of 5.6, gave a current-voltage curve shown in Figure 7. The curve is similar to the one obtained in 0.002 M acetic acid (see Figure 2) except that it was displaced 0.1 volt to more negative potentials. The solution in dilute acetic acid was not stable but on standing deposited a precipitate as a result of hydrolysis, I n spite of the much higher pH, the uranyl solution in the buffer was stable and did not form a precipitate on standing. This is explained by complex formation between uranyl and acetate. Dittrich (6) found that in more concentrated solutions of sodium acetate the uranium was transported to the anode while in dilute solution it moved to the cathode. The complex formation also accounts for the fact that the wave in the buffer solution occurred a t more negative potentials than in dilute acetic acid alone.
FIGURE6. DIFFUSIONCURRENTS OF INCREASING URAXYL CONCENTRATIONS Successive addition of 0.1 M uran 1 acetate t o 50 ml. of 0.1 M potassium chloride andl 0.02 M acetio acid in ? O p e:hanol. (a) Without gelatin, ESP I. 0.8 volt b) with 0.02% gelatin, &.PPI. 0.7 voft. Correctibn applied for dilution.
ANALYTICAL EDITION
May 15, 1942
415
LappI.
FIGURE7. POLAROGRAM OF 0.002 X URANYL ACETATE IN ACETATEBUFFERpH 5.6 (0.2 M SODIUM ACETATE,0.022 iM ACETIC ACID), 0.1 iM IN POTASSICM CHLORIDE, AKD 20 PER CENTIX ETHANOL Values not corrected for residual current. Cathode potential refers to saturated calomel electrode.
necessary for attaining a constant reading of the current after each addition of uranyl acetate. I n the following the authors usually titrated 50 ml. of a solution of monopotassium phosphate in 20 per cent ethanol under the conditions specified in the tables. A layer of mercury was introduced into the cell to serve as an anode. With the pool of mercury as anode the applied e. m. f. was 0.7 to 0.8 volt when the solution was 0.05 to 0.2 M in chloride. If chloride was absent and potassium nitrate was used as the s u p porting electrolyte, the applied e. m. f. was 1.05 * 0.05 volt. When an outside saturated calomel electrode is used as an anode, the potential applied t o the dropping electrode should be -0.7 * 0.05 volt. The titrations were carried out at room temperature. Nitrogen (or hydrogen) was passed through the solution until the oxygen had been removed. The uranyl acetate was added in successive portions from a microburet graded in 0.01 ml., nitrogen being passed through for 2 t o 3 minutes after each addition. The current readings, after correcting for the dilution, were
FIGURE8. TITRATIOX OF 50 ML. OF 0.005 M PHOSPOTASSIUM CHLORIDE AND PHhTE SOLUTION I N 0.1 WITH 0.1 M URANYL ACETATE 20 PERCEKTETHANOL E s p p ~ .0.8
volt. Correction applied for dilution.
plotted a ainst the volume of reagent added and the end point was fount graphically ( l a ) .
Results obtained in the titration of monopotassium phosphate solutions are given in Table I. The 0.1 M uranyl acetate solution was 0.1 or 0.14 M in acetic acid; the more dilute solutions of uranyl acetate were obtained by diluting the 0.1 M solution with water; hence, the pH was relatively low during the titrations (3.5). The current values measured in a titration of 0.005 M phosphate in 0.1 M potassium chloride are d o t t e d in Figure 8. It is seen that the residual current remains small and constant (of the order of 0.1 microampere) until the end point. OF MONOPOTASSIUM PHOSPHATE WITH URANYL h x T a 4 T E TABLE I. TITRATION Hence, it is only necessary to determine the Concencurrent after two or three additions of reagent tration of Uranyl before the end point in order to find the practiIndifferent Uranyl Acetate Used cally horizontal precipitation line. Thus it is hlolaritv Electrolyte Acetate CalcuNo. of KHzP.04 Salt Molarity M lated Found Error possible to finish the titrations within half a n MZ. MZ. % hour. When the phosphate concentration is -0.8 2.50 2.48 0.2 0.1 1 0.005 KNO: greater than 0.004 M , the galvanometer can be -0.8 2.50 2.48 0.2 0.1 2 0.005 KC1 3 0.005 KC1 0.1 0.1 2.50 2.51 2.48 4-0.4 -0.8 used simply as a nullpoint instrument. Uranyl 4 0.005 KC1 0.1 5 0.005 KC1 0.1 0.1 2.50 2.495 -0.2 acetate is added with exclusion of air until the O*Oo5 {%O: ::65 0.1 2.50 2.49 -0.4 deflection of the galvanometer suddenly increases 7 0.002086 KC1 0.1 0.05 2.086 2.075 -0.5 (Table I). The accuracy of the titrations is 1.00 0.99 -1.0 0.02 0.01 8 0.0002 KC1 9 0.0001 KCl 0.02 0.01 0.50 0.47 -6.0 very satisfactory. Concentrations of phosphate 1.009 1.01 0.0 10 0.00010093 KNO: 0.02 0.005 greater than 0.0003 M can be determined with 0.7845 0.84 11 0.00007845 KNOi 0.02 0.005 $;:;}+4.6 12 0.00007845 KNO: 0.02 0.005 0.7845 0.80 a n accuracy of 1 per cent or 0.505 13 0.00005097 KNO: 0.02 0.005 0.56 4-11 better. When the concentration gets smaller, the accuracy UPON TITRATION decreases because these conTABLE11. EFFECTOF ACETICACID CONCENTRATION Indifferent Molarity of Uranyl Molarity of Uranyl Acetate centrations are near the lower Acetic Acid Used Electrolyte Acetate Solution Molarity limit of ordinary polarographic A t beAt end CalcuM0Used of No. KHzPO4 Salt larity UOz(CzH:Oz)z CrHiOt ginning point lated Found Error work. Titrations 11and 12 in M2. M1. % Table I show, however, that in 1 0.005 KC1 0.1 0.1 0.18 0.002 0.02 2.50 2.48 -0.8 2.50 2.46 - 1.6 0.00008 M phosphate solution 0.1 0.1 0.1 0.02 0.0333 2 0.005 KC1 3 0.005 KC1 0.1 0.1 0.18 0.02 0.0371 : E: 1;:; anaccuracyof 5 per cent could 0.1 0.1 0.1 0.04 0.0524 4 0.005 KCI be obtained with two titrations. 5 0.005 {g$bl 0.1 0.1 0.04 0.0524 2.50 2.44 -2.4 2.50 2.43 -2.8 The addition of ammonium 0.18 0.04 0.0562 0.1 0.1 6 0.006 KC1 0.005 7 0.0001009 KNO: 0.1 0.008 f 0.04 0.0401 + 1.01 1.05 +4 Salts, usually recommended in 0.001 0.00005 ammoammoorder to obtain well crvstalline nium nium u02NHaP04,had no &hence acetate acetate upon the titrations.
i::;
;:i5
416
INDUSTRIAL A N D ENGINEERING CHEMISTRY
The effect of varying amounts of acetic acid added to the phosphate solution was investigated next. According to the literature the precipitated uranyl phosphate, UOZMPOI, is not soluble in acetic acid. Nevertheless, the amount of acetic acid present during the titration has an influence on the result, as shown in Table 11. I n the course of a titration the concentration of acetic acid changes according to
Vol. 14, No. 5
CALCIUM. Calcium in larger concentrations caused low results due to coprecipitation of calcium phosphate with uranyl phosphate. However, when the concentration of calcium is smaller than 0.02 M and that of phosphate is 0.005 M or less, the error due to coprecipitation becomes negligibly small (Table IV). Therefore, in most practical cases calcium does not interfere. A more serious error arises when both calcium and sulfate are present in appreciable concentrations. The solubility of calcium sulfate in 20 Der cent ethanol is markedly less than in I n Table 11, therefore, the initial concentrations of acetic water. When the caicium sulfate precipitates upon addition of ethanol considerable amounts of phosphate are coprecipiacid and the concentrations at the end point are given. The table shows that under given conditions a concentration of tated and low results are found (experiments 13 and 14 in Table IV). Addition of acetic acid decreased the coprecipitaacetic acid of 0.02 M a t the end point is about the upper limit tion only slightly. Even when the precipitation of calcium for obtaining satisfactory results. If 50 ml. of 0.005 M phossulfate took place from about 0.04 to 0.08 M hydrochloric phate solution are titrated with 0.1 M uranyl acetate in acid, there was a marked coprecipitation of phosphate, al0.1 M acetic acid, the allowed initial concentration of acetic though it was less than when the precipitation took place in acid in the phosphate solution is 0.006 M . more weakly acid medium (experiments 17 to 24 in Table IV). Before the addition of alcohol, acid was added to make enough TABLE111. EFFECTOF MAGNESIUM AND BARIUM IN 20 PERCENTETHANOL the final concentration of chloride in the titration (Total volume titrated, 50 ml. of 0.005 M KHlPOk solution; concentration of uranyl mixture 0.04 to 0.08 M . After addition of the acetate, 0.1 M ) ethanol the solution was neutralized with sodium Acetic 1. MAcid 1 M MeCh Molar Added Indifferent Uranyl hydroxide, taking the green color of bromocresol Solution Ratio before Electrolyte Acetate Used No. Added Me:PO4 Titrating Salt Molarity Calculated Found Error green as the end point. MI. M1. M1. ML % The error due to coprecipitation of phosphate 0.11 MgClr MgCL .. KC1 KC1 0 .. 1 2.50 2.49 -- 0 with calcium sulfate depends upon the calcimn, 21 41 :: 2 l .5 .. 0 1 2.50 2.49 0 .. 4 4 3 1 BaClr 4:l 2 KC1 0.1 2.50 2.48 -0.8 sulfate, and phosphate concentrations, and it is 4 2 BaClr 8:l 1 KC1 0.1 2.50 2.48 -0.8 not possible to give minimum concentrations of the three separately at which no error occurs. I n a mixture of 25 A. of 0.01 M monopotassium The titration was also performed in buffered solutions. I n phosphate, 1 ml. of 1 M calcium chloride, and 0.5 ml. of 1 M citrate and in oxalate solutions uranyl ion obviously forms sodium sulfate diluted to 50 ml., the results were still correct. complexes which delay the precipitation of phosphate and When 1 ml. instead of 0.5 ml. of sodium sulfate was added, the interfere with the titration. Correct results were obtained in error was - 1.4 per cent and with 2 ml. of sodium sulfate -5.2 acetate buffer of pH 4.7 but the precipitation was slow and per cent. I n general, the error caused by the presence of both therefore impracticable. At a p H between 5 and 6 the titracalcium and sulfate is negligibly small when upon addition of tion could be carried out in the usual way, but the precipitaalcohol little or no precipitate is formed. If the amount of tion remained slow near the end point. I n a medium which precipitated calcium sulfate is appreciable, a separation has to was 0.2 M in sodium acetate and 0.02 to 0.06 M in acetic acid be made. When organic anions which form complexes with and 20 per cent in ethanol, good results were obtained when uranyl (citrate, etc.) are absent, the easiest method is by prethe phosphate concentration was 0.005 M or greater. I n cipitation of the phosphate in ammoniacal solution as calcium more dilute phosphate solutions the precipitation was too phosphate. The precipitate, which need not be washed, is slow for practical purposes. I n biphthalate buffers with a pH dissolved in hydrochloric acid and the solution is neutralized between 5.6 and 3.8 the precipitation was rapid. After the with sodium hydroxide and bromocresol green as indicator. end point, however, the current did not increase linearly with I n this way phosphate determinations were carried out in the the concentration of uranyl acetate. Still, the end point can presence of a hundredfold molar excess of calcium with an be found satisfactorily when the galvanometer is used as a accuracy of 0.5 per cent. When organic matter which has to nullpoint instrument and the titration is carried out until be decomposed is present, the interference of calcium can be eliminated a t the same time. From the data just given rethe deflection suddenly increases. garding the amounts of calcium and sulfate which can be presEffect of Other Substpees on Titration ent in 0.005 M phosphate solution without causing interferORGANICACIDS. Because of the interference of organic ence, it can be seen that the concentration limit corresponds anions like citrate, oxalate, much acetate, and probably many approximately to a saturated calcium sulfate solution (about 0.015 A!). However, even when higher concentrations of others (tartrate, formate, etc.) which form complexes with uranyl ion, it is recommended that organic matters be decalcium and sulfate were present, a precipitate was not formed until alcohol was added. When the presence of organic composed by ashing or by treating with sulfuric acid (and nitric acid) similar to the decomposition according to Kjehlmatter makes necessary a decomposition by treating with strong sulfuric acid (and nitric acid) the remaining concendahl. MAGNESIUM AND BARIUM. Magnesium, when present in tration of calcium sulfate after the treatment with acid does not interfere in the phosphate determination. Such a treatnot too large amounts, did not interfere with the titration. Relatively large amounts of barium also can be present if the ment is usually also used for the dissolution of insoluble phosprecipitation of barium phosphate in the presence of 20 per phates. For the present purposes the removal of calcium in cent ethanol is prevented by addition of acetic acid (Table 111). the presence of organic acids can also be accomplished by I n experiment 3 with an initial concentration of 0.04 M acetic ashing with a n excess of sodium carbonate. acid in the solution a correct result was obtained, while in IRON. Iron, both in the ferrous and ferric states, intercorresponding experiments without barium an error of about feres with the titration. The authors found that the simplest -2.5 per cent was found (Table 11). way to remove the iron was by precipitation with cupferron.
ANALYTICAL EDITION
May 15, 1942
417
TABLE IV. EFFECTOF CALCIUM IN 20 PERCEST ETHANOL (Total volume titrated, 50 ml.; concentration of uranyl acetate, 0.1 M )
Found M1. 0.05 0.05 0.05
M1. 25 25 10 25 25 25 25 25 25 10
1 2 1 1 1 1 4 1 2
1
4:l 8:l 1O:l 4:l 4:l 4:l 16:l 4:l s:l 10:1
12 13 14 15 16
25 25 25 25 25 25
1 1 1 1 1 1
4:l 4:l 4:l 4:l 4:l 4:l
0.05 0.05 0.05 0.05 2 2
17 18
25 25
0.5 0.5
2:l 2:l
....
5 5
e.
5
2
5 5 5 2 10
2 2 4 2 2
1
2 3 4 5 6 7 8
9
10 11
19
0
25
pH
-
5.7. b pH
0.5
-
2:l 4:l 4:l 4:l 8:l 2O:l
..
1 1 1 2 2 2
M1.
... ... ... ... ... ... ... ...
... ...
.. .. o:i
..
0.5 1 2 5 5 5
5.3.
The Fo!ution is made 10 per cent in hydrochloric acid and cooled in ice, and the iron is precipitated with about 0.5 M cupferron solution. The free henyl nitrosohydroxylamine and its decomposition products, w&ch are quickly formed in acid medium, yield polarographic’waves. Therefore, the iron precipitate with the excess of cupferron and its decomposition products were removed by shaking out with ether. The resulting aqueous phase was freed from ether (by gently heating or bubbling nitrogen through), neutralized t o bromocresol green, and titrated after addition of alcohol. As an example, the following analysis is reported. To a mixture of 25 ml. of 0.01 M monopotassium phosphate solution and 1 ml. of 1 M ferric chloride solution were added 3 ml. of concentrated hydrochloric acid, and after cooling in ice about 7 ml. of 0.5 M cupferron solution. After separation as described above, the aqueous solution was diluted to exactly 100 ml., and 50 ml. were then added to 10 ml. of ethanol and t!trated with 0.1 M uranyl acetate solution. The error of the titration was -0.8 per cent.
OTHERINTERFERING SUBSTANCES. Interference will occur with all metals which at a p H of about 3.5 precipitate phosphate-e. g., lead, aluminum, and trivalent chromium. Chemical separations might be possible in these cases. Pyrophosphate and other anions, such as arsenate and vanadate, which precipitate with uranyl acetate interfere with the phosphate titration. According t o Dworeak and Reich-Rohrwig (6), pyrophosphate forms soluble complexes with uranyl: [U02(P207)2]------. Obviously these complexes are not very stable, because when working in a medium of 20 per cent ethanol, pyrophosphate was immediately precipitated with uranyl. It seems possible t o eliminate pyrophosphate by precipitation with cadmium according to Hull (8). The precipitation of uranyl by arsenate can also be used for an amperometric titration of arsenate. Preliminary experiments performed by titrating a 0.005 M arsenate solution in a medium of 20 per cent ethanol under the same conditions as used in the titration of phosphate gave results accurate to about 1 per cent. Procedure On the basis of the various experiments the following general procedure is given. REAGENTS. Uranyl Acetate. A 0.1 M solution is prepared by dissolving 42.422 grams of chemically pure uranyl acetate, UO2-
M1.
Ml.
... ... ... ... ... ... ...
... *..
... ... ... ... ... ... ... 2 2
KC1 KC1 KCI KC1 KCI KCI KC1 KC1 KC1 KC1 0.02% gelatin KC1 KCI KCI KCI KCI 1 ml. KC1 1 M BaCh
+
Ml. 2.47 2.46 1.00 2.48 2.45 2.46 2.435 2.43 2.42 1.01
Error
%
-- 1.2 1.6 0.0 -
0.8 2.0 1.6 2.6 2.8 3.2
-- 1.0 i-
+
0.1 0.1 0.1 0.1 0.1 0.1
2.50 2.50 2.50 2.50 2.50 2.50
2.50 2.465 2.37 2.35 2.40 2.26
NaAc
0:i.
2.50 2.50
2.46 2.48
-- 0.8 1.6
2.50
2.51
4- 0.4
2.50 2.50 2.50 2.50 2.50
2.46 2.45 2.44 2.30 2.20
1.6 --- 2.0 2.4 --12.0 8.0
HAC NaAc HAC
.. .. .... NaAc
HAC
0.04 0.4’ 0.04
.. .. ..
O:2b
0.06
-- 0.0 1.4 5.2 - 6.0 -
4.0 9.6
(C2HaO2).2H*0, in a 1-liter volumetric flask by shaking with about 300 ml. of water and 6 to 10 ml. of glacial acetic acid. After complete dissolution the flask is filled up to volume. The dissolution can be accelerated by careful heating. The solution can be standardized against standard phosphate solution. For titrations of low phosphate concentrations the uranyl acetate solution is diluted with water. Standard Phosphate, hlerck’s preparation “Sorensen” of monopotassium phosphate is dried a t 110’ C. for 1 hour and a 0.01 M solution is prepared by dissolving 1.3614 grams in 1 liter. Potassium Chloride. A 1 M or 2 M solution is prepared by dissolving 74.56 or 149.12 grams of potassium chloride of Merck’s purestquality in 1 liter. The dissolved Sam le is diluted, so.that in the PROCEDURE. final volume (all additions included; the concentration of Dhosphate is not greater than approximately 0.01 M , or if cafcium (and sulfate) are present, not greater than 0.005 M . The concentration of calcium and sulfate must not be greater than 0.02 M and 0.01 M ,respectively. If a previous neutralization is necessary, a few drops of bromocresol green are added to the solution and sodium hydroxide or hydrochloric acid is added until the indicator has a green color. About 0.5 ml. of 0.1 iM acetic acid is added and enough potassium chloride to make the chloride concentration approximately 0.1 M . The sample is introduced into the titration cell and enough ethanol added to make its concentration 20 per cent. Air is removed by passing nitrogen through the solution for about 15 minutes (the removal is complete when the residual current has become constant), and the titration is carried out a t an applied potential of -0.7 to -0.8 volt. NOTES.If a precipitate of barium phosphate is formed by the addition of ethanol, it is dissolved by adding 1 M acetic acid d r o p wise. In the presence of large amounts of both calcium and sulfate the phosphate is first Precipitated in ammoniacal medium and then treated as described above. Iron if present is removed with cupferron as described above. In the presence of organic substances like citrate, tartrate, oxalate, etc., which form complexes with uranyl and in the analysis of organic phosphorus, a destruction of the organic matter is necessary. In the latter instance the total content of phosphorus is determined as orthophosphate.
Summary The polarographic waves of uranyl acetate in weakly acid media were studied. A procedure has been given for the amperometric titration of phosphate with uranyl acetate at room temperature. The accuracy was 1 per cent or better with 0.01 to 0.0003 M con-
418
Vol. 14, No. 5
INDUSTRIAL A N D ENGINEERING CHEMISTRY
centrations of orthophosphate. In 0.0001 M phosphate the accuracy was of the order of 4 per cent. Alkaline earth phosphates can be titrated by the standard procedure. Calcium in large amounts, iron, and organic anions interfere. Methods are described to eliminate the interference.
Dittrich, C., Z. phys. Chem., 29, 448 (1899). Dworzak, R., and Reich-Rohrwig, W., Z . anal. Chem., 77, 14 (1929).
Gmelins Handbuch der anorganischen Chemie, “Uranium”, 8th ed., pp. 67 ff., Berlin, Verlag Chemie, 1936. Hull, D. E., J . Am. Chem. SOC., 63, 1269 (1941). Kolthoff, I. M., “Massanalyse”, 2nd ed., Vol. 11,p. 269, Berlin, Julius Springer, 1930. Kolthoff, I. M., and Lingane, J. J., Chem. Reus., 24, 1 (1939). Kolthoff, I. M., and Lingane, J. J., “Polarography”, p. 296, New York, Interscience Publishers, 1941. Ibid., pp. 447 ff. Langer, A., J. Phus. Chem., 45, 639 (1941). Lewis, D. T., Andust, 65, 256 (1940). Lewis, D. T., and Davis, V. E.,‘J. Chem. SOC.,1939,285. Neuberger, A., 2.anal. Chem., 116, 1 (1939). Rathje, W., Angew. Chem., 51, 256 (1938). Repman, B. R., Lab. Prakt. U.S. S. R., 16,27 (1941). Singh, B., and Ahmad, G . , J . chim. phys., 34, 351 (1937). Stern, A., IXD.ENO.CHEM.,ANAL. ED., 14, 74 (1942). Uhl. F., Z. anal. Chem., 110, 102 (1937).
Acknowledgment Acknowledgment is made to the Carnegie Corporation of New York for a grant which enabled the authors to carry out the present investigation.
Literature Cited Atanasiu, I. A,, and Velculescu, A. I., Z. anal. Chem., 102, 344
(1)
(1935). (2) Bodfors, S., Svensk Kem. Tid., 37, 296 (1925). (3) Chrbtien, A., and Kraft, J., BUZZ.SOC. chim., (5) 5, 372 (1938). (4) Courtois, G . , Ibid., 33, 1773 (1923).
Quantitative Determination of 2-Methyl-
1,4=naphthoquinone AMEL R. MENOTTI, Chemical Laboratory, American Medical Association, Chicago, 111.
T
HE increased use of 2-methyl-l,4-naphthoquinoneas a
therapeutic agent necessitated the development of a n accurate and convenient method for its quantitative estimation in marketed preparations, The procedure described in this paper has been found reliable for the determination of 2methyl-l,4naphthoquinonein quantities as low as 0.05 mg., and has been employed successfully in the assay of oil solutions and of alcoholic extracts containing the drug. I n a study of the reactions involved in this determination, the 2,4dinitrophenylhydrazone of 2-methyl-1,4-naphthoquinonewas isolated and characterized, and its absorption spectrum examined in both alkaline and neutral solvents.
more stable yellow colored derivative in solution. The yellow solution is then compared with a suitable standard. Results of analysis obtained by the method of Pinder and Singer (3) were found, in this laboratory, to be su%ciently accurate, if the naphthoquinone could first be isolated by alcoholic extraction. The method appeared t o be inadequate when applied to vegetable oil solutions from which the naphthoquinone could not be quantitatively extracted. Partial saponification of the vegetable oil, which occurred on the addition of potassium hydroxide, produced turbid solutions that required considerable treatment to effect clarification. It was observed also that the presence of oil complicated the formation and subsequent hydrolysis of the intermediate, unsta6le purpie reaction product. As an alternative, it was considered that the direct spectrophotometric determination of 2-methyl-l,4naphthoquinonein oil solution, employing the absorption maximum at 3360 A., might provide a convenient method of assay. However, direct spectrophotometric examination (1) was found t o be limited in application to those solutions in which the oil used as a vehicle exhibited low absorption in the desired spectral region. The method was of little value when applied to certain commercial products which contained vegetable oil that showed complete absorption at 3360 A.
Novelli (2) described a sensitive color test for 2-methyl-1,4naphthoquinone and related substances which forms the basis for the method finally adopted in this laboratory for the quantitative estimation of 2-methyl-l,4naphthoquinone. The procedure detailed below depends on ( a ) formation of the 2,4dinitrophenylhydrazone of 2-methyl-l,4-naphthoquinone,(b) interaction of this dinitrophenylhydrazone with ethanolic ammonia to yield a green t o blue-green colored solution, and ( c ) comparison of the intensity of this color with that produced in a control with known quantities of the naphthoquinone. The 2,4-dinitrophenylhydrazoneof 2-methyl-1,4-naphthoquinone in alcoholic ammonia solutions exhibits a bright blue color. The Preen color described bv Novelli (2) was found to
a
i-
80 -
I L
- 0.4 ii) 2.4-DINIIROPHENYLHYDRAZONE Of 2-METHYL -I.4-NAPHTHOOUINONE (2) 24-OINITROPHENYLHYDRAZINE
-
-0.3
c k
-
I‘
8 c
LE*
240-
-0.2
R
cp
%
s
,
IO
2103
7000
FIGURE
6000
ANGSTROMS
5000
4000
1. LIQHT ABSORPTION OF SOLVTIONS OF THE 2,4-DINITROPHENYLHYDRAZONE OF 2-h’fETHYL-1,4-NAPHTHOQUlNONEAND OF 2,4-DINITROPHENYL HYDRAZINE
