G. II. B E N D IX a n d D O R IS GRABEN STETTER
Research D epartm ent, C o ntinental Can C om pany, In c ., Chicago, 111.
A m ethod is described for the rapid colorimetric estim ation o f copper w ith dithizone. Interference from all m etallic ions except p la tin u m , p a lla d iu m , gold, silver, mercury, b is m u th , and stannous tin is elim inated by extracting the copper as the dithi- zonale complex from a n aqueous solution a t pH 2.3. The reaction products o f silver, gold, mercury, b is m u th , and stannous tin w ith dithizone are destroyed by shaking the extract w ith acidic potassium iodide solution. Recoveries of better th a n =*=0.3 mierograin were obtained.
E
ST IM A T IN G the copper content of canned foods, tin plate, and solders requires a method capable of detecting quantities of copper as low as a few micrograms, yet the determinations must not be subject to interference by many times that quantity of other metallic ions. Various procedures have been recom
mended for each of these products, but no single method is well suited for all of them. Because no provision is made for the separation of copper from interfering metals in the colorimetric extraction methods of Fischer and Leopoldi (6) and Liebhafsky and Winslow (9) with dithizone, they are unsuitable in general for the analyses of the above-mentioned products. Likewise, the titrimetric extraction of copper with dithizone, a procedure described by Assaf and Hollibaugh (1), cannot be employed.
For a similar reason the use of organic reagents other than dithizone in procedures hitherto published is unsatisfactory.
Nickel, cobalt, and bismuth interfere with the determination by means of sodium diethyldithiocarbamate, and with the recently published determination by Benzo Fast Yellow (12). The Biazzo method (2) cannot be used in presence of relatively large amounts of iron, and determinations by this method and the carbamate method are frequently complicated by the precipitation of phos
phates from which copper cannot be quantitatively recovered.
Preliminary separations of copper from interfering metals by precipitation as the sulfide or by electrolysis are slow, and, in the former case, incomplete (4).
Several methods (11, 13) have been described in which copper is separated from nickel and cobalt and from part of the bismuth by a quantitative extraction of the copper from acid solution with
a carbon tetrachloride or chloroform solution of dithizone. At the same time, some of the bismuth will form the dithizonate com
plex and be extracted along with the copper dithizonate. The dithizonate complexes are destroyed by ashing and copper is determined colorimetrically with sodium diethyldithiocarbamate;
bismuth, if present, contributes to the color and is estimated as copper.
Greenleaf (7) eliminated interference from bismuth by ex
tracting the solution containing copper and bismuth dithizonates with acidified potassium iodide solution, removing bismuth as the iodide complex. The remainder of his procedure involves the oxidation of the copper dithizonate with bromine in 5 per cent sulfuric acid, and the extraction of the copper from the carbon tetrachloride layer to the aqueous layer. The aqueous solution is next digested with nitric and perchloric acids. Copper is determined on the resulting solution by the carbamate method.
From aqueous solutions having a pH value less than 3, gold, platinum, palladium, silver, mercury, bismuth, stannous tin, and copper are extracted by solutions of dithizone in chloroform or carbon tetrachloride; other metals do not react with dithizone at this pH (5). The first three metals mentioned may be left out of consideration because of their rarity, and only silver, mercury, bismuth, and stannous tin treated as possible interferences.
Fischer and Leopoldi recommended the addition of potassium iodide solution to the weakly acidic aqueous solution containing copper and mercury with the formation of the complex ion (H gl,)— as a means of eliminating interference by mercury.
However, it was found in this laboratory that the amount of potassium iodide recommended by Fischer was inadequate to shield the mercury from reaction with dithizone. When larger quantities were used, the dithizone was oxidized by the iodine released in the acidic solution.
On the other hand, as the data below show, copper may be extracted by means of dithizone solution along with the other metals which react at a pH of 2.3, and then separated from these metals by shaking the extract with 2 per cent potassium iodide solution acidified with hydrochloric acid and decolorized with sodium thiosulfate. The copper remains in the carbon tetra
chloride layer as the dithizonate, while the other metals are ex
tracted as iodide complexes. The thiosulfate present in the iodide solution prevents oxidation of the dithizone by free iodine.
650 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 15, No. 10 E xperim ental
De t e r m i n a t i o n o f Sh a k i n g Ti m e. Since copper reacts with dithizone at a rate visibly slower than that of lead, data were accumulated showing the relationship between the time of shak
ing the aqueous solution of copper ions with dithizone solution and the transmittancy of the solution of copper dithizonate. The points on the curves in Figure 1 were obtained by adding 10 ml.
of the dithizone solution to a 150-ml. separatory funnel contain
ing 7 micrograms of copper in 25 ml. of 10 per cent sulfuric acid, adjusting the pH to the desired value, shaking for a given time, and then removing excess dithizone from the carbon tetrachloride layer with 25 ml. of 1 to 100 ammonium hydroxide solution. A blank of 25 ml. of 10 per cent sulfuric acid was run along with each sample. The transmittancy of the copper dithizonate solution was determined at 520 millimicrons, setting the reading of the blank in the spectrophotometer at 100 per cent transmission.
F i g u r e 1. E x t r a c t i o n o f C o p p e r a t V a r i o u s
pH V a l u e s
Figure 1 shows that at a pH of 2.3 it requires at least 6 minutes of vigorous shaking (275 cycles per minute) to obtain equilibrium between the copper in aqueous solution and the copper dithizo
nate in carbon tetrachloride solution when 7 micrograms of copper were originally in the aqueous solution; and that equilibrium is attained more rapidly at pH 4.15 than at the lower values. How
ever, in practice it is advantageous to work at a pH between 2 and 3 in order to avoid extracting metals which interfere with the copper determination. Increasing the strength of the dithi
zone or making two extractions and combining the extracts would bo expected to decrease the shaking time.
Re c o v e r y o f Co p p e r f r o m So l u t i o n s Co n t a i n i n g Ot h e r Me t a l s. The desired volumes of a standard solution of copper and another metal whose effect as an interference was to be tested were run into a 150-ml. separatory funnel. The volumes were made up to 25 ml. with 10 per cent sulfuric acid. From this point the basic procedure for copper as described below was ap
plied. The transmittancy of the dithizonate solution was deter
mined in the spectrophotometer with the blank at 100 per cent transmission. The amount of copper in the sample was read from a calibration curve prepared with known amounts of copper.
The results of experiments with mercuric, bismuth, silver, stan
nous, and auric ions are described in Table I.
It is evident that substantial amounts of zinc, cadmium, and bismuth do not influence the quantitative extraction of copper.
Silver, mercury, and stannous tin do not interfere if the dithizone is present in a quantity more than sufficient to combine with both these ions and the copper. Zinc, cadmium, and bismuth are less readily extracted than copper and only partially accompany the copper. Silver, mercury, and stannous tin, however, are ex
tracted more readily than copper at a pH of 2.3, so that in the presence of these elements it is absolutely necessary that the quantity of dithizone be sufficient to combine with both these elements and the copper. As dithizone complexes with silver and mercury are yellow, it is usually not difficult to detect inter
ference from these two elements. Interference from stannous tin may be eliminated entirely by oxidizing it to the stannic state.
Interference from somewhat larger amounts of silver and mer
cury than are shown in the Table I could undoubtedly be elimi
nated by using a more concentrated solution of dithizone or by increasing the number of extractions.
Re m o v a l o f Ex c e s s Di t h i z o n e b y Di l u t e Am m o n i a Wa s h.
In the removal of excess dithizone from carbon tetrachloride solution by means of an ammonia wash, the dithizone is not quantitatively removed, but an equilibrium is reached. In order to determine the effect of the concentration of ammonia in the wash solution on the transmittancy of the carbon tetrachloride layer containing copper dithizonate, in a series of experiments the concentration of ammonia was varied from 1 part in 200 of redistilled water to 5 parts in 100. Aliquots of standard copper solution were analyzed according to the usual procedure up to the removal of excess dithizone. Then each sample was shaken for 2 minutes with 25 ml. of wash solution, and the transmittancy was read against a blank. Another group was shaken for 2 minutes with 15 ml. of wash solution and then for another 2-min
ute period with 10 ml.
MeLhod
Ap p a r a t u s. The only special apparatus used was a Coleman spectrophotometer. Readings were made on this instrument at
C u Added
October 15, 1943 A N A L Y T I C A L E D I T I O N 651
520 millimicrons, since a minimum transmittancy occurred at this wave length. A mechanical shaker is very desirable.
Re a g e n t s. Water redistilled from Pyrex was used for all dilutions, c. p. nitric acid was redistilled from Pyrex. All solu
tions were stored in Pyrex containers.
Glassware was cleaned with concentrated nitric acid and rinsed first with tap water, then with water redistilled from Pyrex.
Copper standard, 1 microgram of copper per ml. Accurately weigh a 0.5000-gram sample of electrolytic sheet copper and treat with 20 ml. of 6 iV nitric acid. Evaporate the solution almost to dryness and then add 2 to 3 drops of glacial acetic acid.
Transfer the solution quantitatively to a 500-ml. volumetric flask. From this stock solution a working standard of 1 micro
gram of copper per ml. should be prepared immediately before use.
Sulfuric acid, 10 per cent. Dilute 1 volume of the concentrated c. p. acid to 10 volumes.
Cresol red, 0.02 gram per 100 ml. of water.
Buffer solution, pH 2.3. Dissolve 38 grams of c. p. citric acid and 21 grams of c. p. disodium hydrogen phosphate dodecahydrate in redistilled water. Purify by shaking with a concentrated solu
tion of dithizone in carbon tetrachloride. Wash out the excess dithizone with clear carbon tetrachloride, discard the washings, and dilute the aqueous layer with redistilled water to 250 ml. solution, the solution is buffered at a pH of 2.3.
Dithizone solution. Dissolve 15 mg. of pure dithizone in 1 liter of c. p. carbon tetrachloride. Since solutions of dithizone are exceedingly unstable, certain precautions must be observed in their preservation (3). In addition, Greenleaf (8) has sug
gested the use of hydroxylamine in the sample solution itself to prevent oxidation of the dithizone.
Potassium iodide solution, 2 per cent. Dissolve 10 grams of c. p. potassium iodide in 450 ml. of water. Acidify with 5 ml. of 1 N hydrochloric acid. Discharge the color caused by the presence of free iodine with 0.1 N sodium thiosulfate solution added drop- wise. Shake in a separatory funnel with successive 10-ml. por
tions of dithizone solution until no discoloration of the dithizone occurs. Discard the carbon tetrachloride extract and wash the aqueous layer with clear carbon tetrachloride. Dilute the aque
ous solution to 500 ml. with redistilled water. Check the potas
sium iodide solution frequently (once a day) with dithizone solu
tion for the presence of free iodine, and if necessary add more sodium thiosulfate solution. Since the potassium iodide solution may not be checked often enough to detect traces of free iodine, Greenleaf (5) suggests adding some reducing agent such as hypo- phosphorous acid or sodium hypophosphite to the iodide solution.
Ammonium hydroxide, 1 to 200. Dilute 1 volume of concen
trated c. p. ammonium hydroxide to 2 0 0 volumes with redistilled water.
Pr o c e d u r e. Pipet a suitable aliquot of the sample solution into a 150-ml. separatory funnel. If the volume is less than 25 ml., make it up to that volume with 10 per cent sulfuric acid.
Add 2 drops of cresol red solution and bring the contents of the flask to the yellow color of the indicator range with concentrated ammonium hydroxide. Add 2 ml. of the buffer solution (pH 2.3) and exactly 10 ml. of dithizone solution. Shake for 10 minutes.
Transfer the carbon tetrachloride layer to a clean separatory funnel and add 10 ml. of acidic potassium iodide solution. Shake for 2 minutes. Transfer the carbon tetrachloride layer to another separatory funnel and add 25 ml. of 1 to 200 ammonium hydroxide solution. Shake for 2 minutes. Determine the transmittancy in the spectrophotometer at 520 millimicrons with the blank at 100 per cent transmission. From a calibration curve, such as shown in Figure 2, estimate the amount of copper present.
Ta b l e III. An a l y s e s o f Fo o d s
Food Product m ethod0 method M ethod
P . p. in. P . p. VI. P . p. in. P . p. m. in a Waring Blendor, and then wet-ashing a suitable sample (20 to 50 grams) with concentrated sulfuric and nitric acids. The well- mixed sample is weighed into a 500-ml. Erlenmeyer flask and heated until excess moisture has been driven off and the material has just begun to char. Then 10 ml. of concentrated sulfuric acid are added to the cooled contents of the flask, and the flask is returned to the burner. Heating is continued and, when white sulfur trioxide fumes appear, nitric acid is added dropwise. As the nitric acid is consumed, more is added dropwise until the contents of the flask are colorless or at most a straw yellow in color. The flask is removed from the flame and allowed to cool.
The contents are diluted with redistilled water to approximately 50 ml. Heating is resumed until sulfur trioxide fumes appear.
The cooling, dilution, and heating are repeated to ensure removal of all oxides of nitrogen. The contents of the flask are trans
ferred to a volumetric flask and diluted to volume.
652 I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y Vol. 15, No. 10
coveries of added copper were made from some samples, and in a few cases results were checked by the Greenleaf method (7').
Results of these analyses are given in Table I I I . I t is evident that the agreement between the dithizone method and the Greenleaf method is good and that copper recoveries from foods by the dithizone method are satisfactory.
St e e l. Since samples of steel are analyzed occasionally for copper in this laboratory, the dithizone method was applied to a set of six steel samples which the Weirton Steel Company had distributed previously for a collaborative study of the determina
tion of copper in tin plate base metal.
The sample was prepared for analysis by dissolving 1 gram of steel filings in 100 ml. of hot 1 to 9 sulfuric acid. Redistilled nitric acid was added dropwise and heating continued until the solution was a dear yellow color. After cooling it was diluted to 250 ml. in a volumetric flask. A suitable aliquot was then analyzed for coppcr.
In Table IV data obtained by the dithizone method are to be found, together with data furnished by Weirton Steel Company on the results of the collaborative study. I t is apparent that the dithizone method may be applied to steel without modification.
So l d e r. A 1.0-gram sample was dissolved in concentrated hydrochloric acid and bromine. After filtering off the lead chloride the solution was neutralized to a pH of 2 to 3. Copper was determined in a suitable aliquot by the regular procedure.
The copper content of Bureau of Standards Sample No. 127 was found to be 0.016 per cent as compared with 0.013 per cent deter
mined spectrochemically and 0.014 per cent chemically by Ruehle and Jaycox (10).
A cknow ledgm ent
The authors wish to thank G. C. Jenison of the Weirton Steel Company for supplying steel samples and the results given in
(5) Fischer, H., Angew. Chem., 47, 685 (1934).
(6) Fischer, H., and Leopoldi, G., Ibid., 47, 90-2 (1934).
(7) Greenleaf, C. A., J . Assoc. Official Agr. Chem., 25, 385 (1942).
(8) Greenleaf, C. A., private communication.
(9) Liebhafsky, H. A., and Winslow, E. H., J . Am. Chem. Soc., 59 ,
(13) Sylvester, N. D., and Lampitt, L. H., Analyst, 60, 980 (1942).