V O L U M E 2 8 , N O . 11, N O V E M B E R 1 9 5 6
1655
for magnesium after magnesium had been added. I n other words, all analyses were made but once, but in effect each mixture was analyzed in triplicate. These iesults are shown in Table IV.
Diehl, H., Goeta. C. A,, Hach, C. C., J . A m . Water W o r k s Assoc. 42, 40 (1950). Eldjarn, L., NygaiErd, D., Sveinsson, S. L., Scand. J . Clin. Lab. Inrest. 7, 92 (1955). Fales, F. W., J . Bid. Chem. 204, 577 (1953). Fortuin, J. R.1. H., Karsten, P., Kies, H. L., Anal. Chim. Acta 10, 356 (1954). Friedman, H. S., Rubin, AI. A,, Clin. Chem. 1, 125 (1955). Gehrke, C. W., Affsprung, H. E., Lee, R . C., ASAL.CHEM.26, 1710 (1954). Goddu, R.F., Hume, D. N., Ibid., 26, 1740 (1954). Greenbiatt, I. J., Hartman, S., Ibid., 23, 1708 (1951). K a r s t e n , P., Kies, H. L., Van Engelen, H. T. J . , DeHoag, P., A n a l . Chim. Acta 12, 64 (1955). Kenny, A. D., Tovernd. S.U., A x . 4 ~ C . H E ~ 26, I . 1059 (1951). Kibrick, A. C., Ross, AI., Rogers, H. E., Proc. Sac. Exptl. B i d . and X e d . 81, 353 (1952). Lehman, J., Scand. J . Clin. Lab. Incest. 5 , 203 (1953). llartell, A. E., Calvin, lI.,“Chemistry of the Metal Chelate Compounds,” Prentice-Hall, New York, 1952. Pribel, R., “Complexometrie,” Chemapol, Prague, Czechoslovakia, 1954. Shaviro. R.. Brannock. W.W.. AKAL.CHEM.27. 725 (1955). Sobel, A. E., Xedoff, S., Abstracts, 127th Meeting, A\IERICAN CHEMICAL SOCIETY, Cincinnati, 1955, p. 21C. Wilson, A. A , , J . Comp. Pathol. Therap. 63, 294 (1953); 65, 285 (1955).
K i t h the exception of an occasional determination, the analytical results for the recoveries using calcium or magnesium appear to be eqiially good, regardless of xhether pure calcium and magnesium mixtures alone or other mixtures were analyzed. Figure G s h o w the titrimetric curves obtained a t 660 mp for varying concentrations of calcium between 1.25 and 5.0 mey per liter of spinal fluid. The intersection of the tangent to the steep ascending slope of the titration curve to a line parallel to the abscissa and tangent to the upper portion of the cuive indicates the end point. A >imilar study ip shoxn in Figure 7 , where increasing concentrations of concentration of magnesium in the range of 1.05 to 4 20 meq. per liter of spinal fluid are titrated a t 660 mp. The end point is graphically determined in the same manner a.; io1 calciiim. LITERATURE CITED
(1) Banexvirs, J. L.. Iiriiiier, c‘. T., ANAL.CHEM.24, 1186 (1952). ( 2 ) Betz, J. D., Sol!, C . .I..J . Am. Wuter W o r k s Assoc. 42,49 (1950). ( 3 ) Buckner. B., Shively, J. -4., Am. J . Med. Technol. 21, 269 (1955). (1) C’heng, K. L., Kurtz, T.. Bray, R. H., ANAL.CHEX 24, 1040
(1952).
RECEIVED for review January 7 , 1956. Accepted July 18, 1956. Division of Biological Chemistry, 128th Meeting, ACS, Minneapolis, Minn., September 1955. Supported in part by a Grant-in-Aid from the Receiving Hospital Research Corp.
Photometric Determination of Chlorides in Water DAVID M. ZALL, DONALD FISHER, and MARY Q. GARNER U. S.
N a v a l Engineering Experiment Station, Annapolis,
A method has been devised for the photometric determination of small amounts of chloride in water. The method is based on the displacement of thiocyanate f r o m mercuric thiocy-anate by chloride ion and the subsequent reaction of the liberated thiocyanate with ferric iron to form the colored complex [Fe(SCN)]++, which is measured either visually or in a spectrophotometer. Concentrations of chloride as low as 0.05 p.p.m. can be determined.
T
HE literature on the determination of chloridesisvoluminous.
\Yhether present as a required constitumt or as an impuiity, the chloride ion is usually determined by either gravimetric or volumetric methods. The oldest and the classical method is the gravimetric, in which the chloride ion is evaluated as silver chloride. ..inother method frequently employed is the volumetiic, several variations of which are available. The Volhard method, oiiginated by Carpentier ( 7 ) , described by T‘olhard (38), and later improved by Lundbak (65) and others (26), is more accurate than the Mohr method ( 2 7 ) . A comparatively recent method is the mercurometric method, which was developed in 1933 by Dubskj. and Trtilek (9, 10). Diphenplcarbohydrazide v a s used as an indicator in the titration n-ith mercuric nitrate. Other 13-orkers (1, 5 , 8, 20, 62, 29, 32, 54) later adopted this method with some modifications. ill1 these methods, homeever, are not always suitable for the detei mination of micro quantities of chloride. The present investigation was the result of a need for a simple colorimetric method for the determination of less than microgram quantities of chloride in condensate. Luce, Denice, and ilkerlund ( 6 4 ) determined small amounts of chloride turbidimet-
Md.
1 ically. This method, however, lacked the required precision. Other methods (2, 4, 6, 14, 15, 17, 81, 28, 55, S 9 ) for the determination of small amounts of chloride either required special apparatus or lacked the desired simplicity. The method presented here is a modification of that proposed b!. Vtsumi (36, 37) and followed u p by Imasaki (16). This modified procedure has been greatly improved and broadened in ita application. The use of ferric perchlorate instead of ferric ammonium sulfate eliminates a variable inherent in the latter iragent. Use of an aqueous instead of an alcoholic solution of inei curie thiocyanate minimizes the glaring blank when visual color comparisons are made. The improved sensitivity thus obtained and the adaptability to either visual or spectrophotometric comparison made it useful for a more varied application. Although designed for the determination of chloride in condensate, it can also be applied in other fields.
REAGENTS
Mercuric thiocyanate, saturated water solution (0.07%). Ferric perchlorate. Dissolve 6 grams in 100 ml. of LV perchloric acid. This reagent may also be prepared by dissolving 14.0 grams of pure iron wire in dilute nitric acid. Upon dissolution of the iron, add 120 ml. of perchloric acid and heat the solution until it fumes. Continue heating until the solution turns purple, then cool and dilute t o 1 liter. PROCEDURE
Place 10 ml. of sample in a 50-ml. volumetric flask, add 5 ml. of (30% perchloric acid, 1 ml. of mercuric thiocyanate, then 2 ml. of ferric perchlorate. Make up to volume and mix well. Allow t o stand for 10 minutes, then read the transmittancy on a spectrophotometer a t 460 mp. The color can also be compared visually with suitably prepared standards.
1666
ANALYTICAL CHEMISTRY
Preparation of Standards. Weigh 1 gram of dried C.P. sodium chloride t o the nearest milligram. Dissolve in distilled water, transfer t o a 1-liter volumetric flask, and dilute to volume. Each milliliter contains 1 mg. of sodium chloride. Pipet 10 ml. of this solution into a 1-liter volumetric flask and dilute to volume. This master solution, containing 10 p.p.m. of sodium chloride, is used to prepare standards containing chloride concentrations from 0.05 to 5 p.p.m. DISCUSSIOS
The existence of a distinctive color reaction between F e + r + and SCN-ions has been known for over a century. This reaction has been employed in the colorimetric determination of iron.
0 u
Z C ii 5 0
c 0 (0 4
0 0
100
200
FeSCN++. When thiocyanate is in excess the appearance of Fe(SCN)2+becomes evident (Figure 1). I n addition to the two complexes indicated, others are possible when thiocyanate reacts with iron. A series of complexes, represented by Fe(SCN),+3-", where n = 1,. . .6, can be obtained (19). Equilibria data for six complexes formed by the interaction of iron and thiocyanate are given by Lewin and Wagnei ( 2 3 ) All these complexes are red. There is not much difference in hue, although there is a shift in maximum absorbance Kith various acid concentrations. I n perchloric acid medium the maximum absorbance is obtained a t 460 mp (Figure 2 ) . A shift to 495 mp is obtained when sulfuric acid is used instead of perchloric acid. The addition of acetone does not affect the position of the peak materially. Results of the determination of ferric iron in perchloric acid with excess thiocyanate are summarized in Figure 1. Fifty per cent acetone solutions were much more stable than v-ater solutions. The per cent transmittance of 50% acetone solution containing 306 y of ferric iron had changed in 1 hour from 13.2 to 13.8%, introducing an error of less than 5%. Solutions containing nitric acid darkened on standing. The visible absorption spectra of sulfuric acid-nitric acid solutions with and without acetone are markedly similar, with absorption maxima a t 490 mp. The maximum absorption spectrum of a solution of ferric iron with excess thiocyanate in l M
300
-r Fe ~n 50 0 m l
Table I.
0 500
Effect of Perchloric Acid Concentrations
' ( 2 nil. of 0 2 5 M ferric iron and 294 y of thiocyanate in 50-ml. final volume)
hll. GOY PerchlorPo Acid 1.0 2.0 3.0 4.0 5.0 G O
w
z u
a c, 2 5 0 m c 0
s
Absorbance3 0,2822 0.2512 0.2440 0.2408 0.2406 0,2401
II C 0
0
200
I00 IF e
Figure 1.
300
1-cm. cell.
350
'Table 11. Visual Estimation of Chlorides
in 50 0 ml.
(All values are in parts per million) Concn. of Standard Chloride Concn. Solution Found" 0.0 0 2 0.05 0.2 0.1 0.2t 0.2 0.3 0.3 0.4 0.5 0 5
Influence of various thicqyanate-perchloric acid concentrations on absorbance
The first investigation into the aspects of the iron-thiocyanate reaction appeared in 1931 ( S 3 ) ,xhen Schlesinger arid 1-an T-allienburgh (33) made a spectral st,udy of the iron-thiocyanate complex. They measured the light absorption of aqueous solutions of ferric thiocyanate and Saa[Fe(SC?;)6] and of anhydrous ether solutions of ferric thiocyanate. Tliey concluded from the similar spectra obtained that the same absorbing species is probably responsible for the color in every instance. Kiss, .kbr:ihiini, and Hegedus ( 1 8 ) confirmed this work and concluded that, 77-ith excess thiocyanate ion, the prinripal absorbing species is [Fe-
a
Identical results were obtained from 3 operators.
Table 111. Influence of Other Ions on Chloride Determination Amount Taken,
(SCN)6] Bent and French (5)and Edmorids and Birnbauni ( l a ) proved the existence of FeSCS++ in ferric thiocyanate solut,ions. They presented evidence that when ferric iron is in excess the iron thiocyanate is present as FeSCS++ and is the only absorbing species. On the other hand, Frank and Oswalt ( I S ) showed that, a t total concentrations greater than about 0.004M, an excess of thiocyanate leads to a higher absorbance than the same excess of ferric iron. Thus, when ferric iron is in excess there is only one equilibrium involved: the one leading to the formation of
Contaminant FHPOl- HPOa - HPOa-NaNOa
Mg. 0,0125
2.5 5.0 12.5 0.05 Grams 1.25 0.1 0.5 1.25 0.1 0.5 1.25
Chloride,
y
Taken 125 125 125 125 40
Found 125 118 105 74 39.5
125 25 125 12.5 23 125 125
75
15 14 0 23 12 5
V O L U M E 28, NO. 11, N O V E M B E R 1 9 5 6 0 000
1667
100
0 10 M POTASSIUM
90
DILUTED TO lOOml ( 5 c m CELLS)
0 097
EO W u
w
z
70
0
m
$
2 t
f
L
0222
60
4
2
+ 8
50
0 390 4 00
on the colored complex. Ordinarily, the acidity of the solution does not affect the color intensity materially, provided the acid does not form a complex with the ferric iron. For instance, when nitric acid is used for acidification, the color intensity shows little change in the range from 0.05- to 0.08N (SI). However, when larger concentrations of nitric acid are used, the color intensity changes somewhat with the acid concentration. Table I shows the effect of perchloric acid concentrations. Large excess of acid offers no interference; however, the acidity cannot be decreased below a certain lower limit. Erratic results are obtained with low acid concentrations. Less concentrated perchloric acid can be tolerated when spectrophotometric measurements are made than \Then visual comparisons are made.
40
500
600
700
CHLORIDE DETERMINATIONS
W A V E LENGTH, Mp
Figure 2. Absorption spectrum in perchloric acid
Figure 3.
Visual (Table 11)and spectrophotometric determinations were made on prepared samples of known chloride content. The absorption spectrum without additional perchloric acid shows a secondary peak a t 420 mp (Figure 6 ) due to low acid concentration. I n Figure 5 curve 2 is the most suitable for visual comparison, because a low blank value is evident. The system (Figure 5 ) does not follow Beer's law. However, up to 40 p.p.m. of chloride ion, a straight line is obtained and the intensity (chloride concentration) is proportional to the thiocyanate concentration. For the photometric evaluation, curve 3, Figure 5, is most sensitive, even though the blank is rather high.
Effect of varying iron concentration
perchloric acid is a t 460 mp. It shows relatively less absorption a t 400 mp than the same solutions containing an equivalent amount of sulfuric acid. The color intensity of the iron thiocyanate complex depends upon several variables: excess of thiocyanate (18), kind of acid, and time of exposure. I n the application of this reaction to the determination of chlorides, the thiocyanate ion is measured and no excess is possible. The iron is in excess and its effect on the iron-thiocyanate complex x a s studied. Veasured quantities of thiocyanate were allowed to react rrith various amounts of excess iron. Results of these experiments are shown in Figure 3. Acetone, having a low dielectric constant, intensifies the color (Figure 4). The visible absorption spectrum of thiocyanate-excess ferric iron shows a pronounced absorption band between 400 and 130 mp which the thiocyanate-excess ferric iron-acetone combination lacks. The absorption maximum for. both solutions is a t about 480 mp. Figures 4 and 5 show the effect of various acid concentrations
x F P tn I 0 0 nil
Figure 4.
Calibration curves for iron in various acid media and influence of acetone
IXTERFERENCES
The bromide ion interferes in any quantity. It displaces thiocyanate from mercuric thiocyanate in the same manner as the chloride ion does. Small amounts of fluorides, nitrates, nitrites, sulfates, and phosphates do not interfere. Higher concentrations of sulfates and phosphates bleach the color. High concentrntions of ethyl and isopropyl alcohols, tartaric acid, and acetone impart a yellowish brown to the ferric thiocyanate color. The
ANALYTICAL CHEMISTRY
1668 0 .SO0
0 W
5‘U.50 a VI 0
m 4
0
0
y
Figure 5.
100
50
I50
CI - in 50.0 ml.
Effect of acid concentration on absorbance
2 ml. 0.25M F e + + - in perchloric acid; 5 ml. s i t u r a t e d aqueous mercuric thiocyanate 1. p H 1.7 2 . 1 M perchloric acid 3. 1.5M perchloric acid (approx.)
blanks are also colored yellowish brown. data are summarized in Table 111.
Spectrophotometric
EXTRACTION
The iron-thiocyanate complex in the presence of excess iron is not extracted by ether or other organic solvents. It was pointed out by Durand and Bailey ( 2 1 ) in 1923 that the red color of the iron-thiocyanate complex is not extractable unless there is an excess of thiocyanate. This n-as confirmed by Peters and French (SO). Molecular weight determinations by Schlesinger and Van Valkenburgh (33) in anhydrous ether and benzene showed the formula to be Fe2(SCN)6. LITERATURE CITED
hsper, S.P., Jr., Schales, Otto, Schales, Selma, J . B i d . Chem. 168, 779 (1947).
Avaliani, K., Zavodskaya Lab. 12, 179-82 (1946). Bent, H. E., French, C. L., J . Am. Chem. Soc. 63, 568 (1941). Binkley, Francis, J . Bid Chem. 173, 403-5 (1948). Bohm, E., Sturz, O., Chemie D i e , 55, 319 (1942). Bruggemann, Joh., 2. anal. Chem. 126,297 (1943). Carpentier, P., Bull. soc. ing. cia. France 1870, p. 325. Clarke, F. E., ANAL.CHEM.22, 553 (1950). Dubskf, J. V., Trtilek, J., Mikrochemie 12, 315-20 (1933). Ibid., 15, 95-8 (1934). Durand, J. F., Bailey, K. C., Bull. soc. chim. 33, 654 (1923). Edmonds, S. hl., Birnbaum, N., J . Am. Chem. SOC.63, 1471 (1941).
Frank, H. S.,Oswalt, R. L., Ibid., 69, 1321 (1947). Hach. A.. Franke. H. W.. Mikrochemie ter. Mikrochim. Acta 33, 13.5-6 (1947). (15) Hettche, O., 2. anal. chem. 124, 270 (1942). (16) Imasaki, Iwaji, Utsumi, Sartori, Ozawa, Takejiro, Bull. Chem. Soc. J a p a n 25, 226 (1952).
W A V E LENGTH, Mp
Figure 6. Absorption spectra of iron-thiocyanate complex in strong and weak acid media 80 y chloride, 4 ml. of 0.25.V F e + + + in perchloric acid, 10 ml. saturated aqueous mercuric thiocyanate; final volume, 100 ml.; 5-cm. cells
(17) Kellogg, R. H., Burack, W.R., Isselbacher, I