Determination of Iron in the Presence of Cobalt Two-Component Colorimetric Method ERNEST A. BROWN, McGean Chemical Company, Cleveland, O h i o A colorimetric thiocyanate method is given for the estimation of 0.07 to 0.5 mg. of iron in the presence of variable amounts of cobalt
If, for simplicity, we let L" and L E represent the log transmittancy obtained with reference to a standard light intensity
up to 90 mg. A filter-type photometer with two color Alters is used to circumvent the interference of the cobalt ion color. The method is rapid with an accuracy of *3%,
THE
quantitative separation of small amounts of iron from large amounts of cobalt that might be found in technical grade cobalt salts is long and difficult. The ammonia sepanttion gives the ferric hydroxide contaminated with relatively large amounts of cobaltous hydroxide. Even a double precipitation gives appreciable amounts of cobalt carried down with the ferric hydroxide ( 2 ) . The basic acetate separation, which offers a more careful control of the pH, gives an incomplete separation and usually requires a double precipitation. The electrodeposition of cobalt from an ammoniacal chloride or sulfate solution containing sodium bisulfite is open to question, since the cathode plate is contaminated with small amounts of iron (6). Other methods of separation of cobalt from iron, such as the use of a-nitroso-B-naphthol and phenylthiohydantoic acid (6), precipitate iron more or less completely. The ammonia separation appean to be acceptable in separating the iron from the major part of the cobalt, but leaves the ferric hydroxide contaminated with varying amounts of cobaltous hydroxide up to 100 mg. A method employing a single ammonia separation and estimation of the iron in the presence of highly colored cobalt ion is desirable. A number of colorimetric methods are available for the estimation of small amounts of iron. The thiocyanate method is possibly the oldest and best known. Peters and French (9) studied the effect of acidity, salts, and thiocyanate iron concentration on the color. Spectrophotometric absorption curves for the ferric thiocyanate system and interfering ions have been published (8). The sensitivity of the color reaction and fading of the colors have been improved through the use of acetone (8) and 2-methoxyethanol ( 7 ) . .4 number of advantages may be listed for the thiocyanate method. The color is developed in a relatively strong mineral acid solution which simplifies the adjustment of the acidity by addition of a fixed amount of mineral acid. The sensitivity of the color reaction is good and compares favorably with other colorimetric methods for iron. In the application of a photoelectric photometer the absorption bands for ferric thiocyanate and cobaltous ion overlap to such an extent that a single color filter is not selective enough to prevent the interference of the cobaltous ion. The color, due to the cobaltous ion, tends to vary somewhat in hue with the thiocyanate concentration and acidity of solution. Strong hydrochloric acid solutions (10to 12 N ) containing cobalt show the characteristic blue color, while the familiar red or pink colors are noted in more dilute acid solutions. The difficulties found with the single-color method often may be easily eliminated through the use of two color filters. This principle was applied by Knudson, Meloche, and Juday (1) to the determination of aluminum in the presence of iron by the hematoxylin method and is based on the validity of Beer's law for the two-component system and each component's behaving independently of the other. Simultaneous equations are obtained from the extinction coefficients from a pure solution of each component with color filters A and B. 228
Constants KIA, K I B ,K z A ,and K z B depend on the characteristics of particular color filter used, the cell thickness, and the color of the solution. The value of each constant may be easily determined by preparing a pure solution of each component at a known concentration and measuring the transmittancy with each color filter. The log transmittancy values may then be substituted in Equations 1 and 2. For example, with a pure solution of known concentration C1, Cz = 0,KoAC2 = 0, and KzBC2 = 0.
Kid
L"
In a similar manner constants K z Aand K15 may be determined with a pure solution of known concentration CZ. The values of constants KIA and K I B may be said to represent the slope of line
10 Figure 1. Tranrmittancy Curves for Ferric Thiocyanate and Cobalt Chloride. Transmittance Curves for Selected Color Filters
.
In 100 ml. (0.65 M ammonium thiocyanate, 0.59 N hydrochloric a c i j , 5-cm. cell) ms. In 100 ml. 0.65 M ammonium thiocyanate, 0.59 N hydrochloric acid, 5-cm. cell) 111. 425-mr blue color Rlter IV. 595-mr green color filter
I. Iron rolution,0.1 m
II. Cobalt solution, 50.7
ANALYTICAL EDITION
April, 1945
plotting concentration against log transmittancy for a pure component CI with color filters ,4and B. Equations 1 and 2 may be combined to eliminate CZ.
C? =
L“
- KIA c1 = Ki”
L”
- k‘,B CI R2B
froin \vhich it follows
Some recenr, makes of filter-type photoelectric photometers are equipped with a logarithmic scale in which the concentration ie a linear function of the photometer reading if Beer’s law is vrtiid. Ordinary metric graph paper may be used with such E. scale to check the validity of Beer’s law. The Fisher electrophotometer has a logarithmic scale in which the instrument read100 The use of this i c ing represents 1W iog % transmittance’ strument, scale makes it possible to determine the constants io Equation 3 in terms of instrument readings rather than log transmittancy values. I t may be easily demonstrated that the cor-stants so determined follow the same relationship developed in Equation 3 by adjusting the imtrument to zero scale reading wit.b a blank. The use of this instrument scale greatly simplifies the determination of the constants and calculation&. ,4n examination of the spectrophotometric transmittancy c u r v e for ferric thiocyanate and cobaltous chloride in Figure 1 shows a maximum absorption range in the vicinity of 475 mp for iron and 513 mp for cobalt. Marked overlapping occurs in this range 475 to 513 mp. A greater difference in photometer rea;ings may be obtained by selecting a color filter in the vicinity of 425 mp, where the ferric thiocyanate color gives a relatively great absorption difference, and ir. the vicinity of 525 mp where the cobaltous ion shows the maximum absorption difference. Accordingly filters .4 and B were selected. The over-all accuracy of this method is increased by such a selection of color filters rather than a t the points of maximum absorption.
Table
229
I. Analysis of Cobalt-Iron M i x e d Solutions
Cohalt Takei?
L1
LE
Xg.
11.4 28.4 28.4 28.4 47.4 56.8 71.5 83.8 0.0 0.0
Fe
co
11.2 23.8 40.5 36.8 25 7 41.6 40.7 37.6 10.7 54.8
0 0
13.3 28.6 65.5 40.2 25.5 48.0 44.2 36.2 15.7 79.0
02 25
Iron Taken Ma.
Iron Found
0,070 0.150 0.390 0.270 0.100 0.250 0.200 0.140 0.100 0.500
0.070 0.152 0.383
tO.OOO +O.OO? -0.007
0.276 0.103 0.243 0.206 0.137 0.100 0.497
+0.006 f0.003 -0,007 +0.006 -0,003
!IC
MR.
Error
Ma.
tO.OOO
-0.003
03
50 75 COYCENTFAT 0’.“,‘C/IOOi;rii
Figure 2.
Graph for Electrophotorneter at Various Concentrations of Iron and Cobalt
MATERIALS AND PROCEDURE
ELECTROPHOTOMIETER. The A(,’ Model Fisher electropho-
tonieter equipped with 2-cm. cylindricai absorption cells was used. This instrument utilizes barrier layer photocel!s aiid a balanced tiridge circuit. The galvanometer is used as a null i:lstrunieni. The 525- and 4?5-mp color filters furnished with the instrument were used. i a o s SOLCTION. Dissolve 0,0500 gram of electrolytic iron \Tire in a feu- milliliters of dilute hydrochloric acid, add 6 drops of 30% hydrogen peroxide, and boil to oxidize the iron. Add a sliglit excess of 0.1 A- potassium permanganate to ensure complete oxidation and remove tile excess by boiling. Cool and make to 1000 mi. C’UHALT SnLrTIox. l’repare a standud cobalt solution from re:i;eiit graue iron-free cobah chloride containing 1 to 2 mg. of c o l ~ a l tin 1 ml. Since ruost reagent grade cobalt salts may contain traces of iron, this may be purified by oxidation with perd e and ammonia. After filtration plate the cobalt a t 1 to 2 pcres on a platinum cathode, dissolve the plate in hydrocllloric, acid, and remove the excess acid by evaporation. The cobalt concentration may be most conveniently determined by no:irig the difference in cathode weight. THIOCI-.%NATE SOLCTIOS.Dissolve 10 grams of reagent grade amnionium thiocyanate in 100 ml. of iyater. HYURUCHLOKIC ACID, reagent grade, specific gravity 1.185. standards by taking known amounts P R O C E D ~ RPrepare E. of iron or cobalt standard in a 100-mi. volumetric flask. Add 5 ml. of hydrochloric acid and adjust the volume to approximately 90 nil. with water. Add 5 ml. of thiocyanate solution, make to volume, and mix thoroughly. hleasure the color immediately n.it1; tlie 42% and 525-mp color filters. A blank containing 5 m!. of hydrochloric acid, 5 nil. of thiocyanate solution, and water to 100 mi. is used as a reference standard for the photometer zero adjustment. The results are given in Figure 2 . CALCULITIOS. If we let C1represent iron concentration in mg. per 100 mi., and C2 the cohalt Concentration in mg. per 100 ml., the 523-mp color filter will be represented by -4 and 425 mp
color filter by R.
Kl-4
I
