202 RUSSELL AND HART: THE DETERMINATION OF COPPER I N GEL-4TIN [Vol. 83 The Determination of Copper in Gelatin BY G. RUSSELL AN:D P. J. HART (Chemical Research Laboratory, Ilford Limiteal., Woodman Road, Brentwood, Essex) Some newer methods for the determination of traces of copper have been applied to gelatin and results are compared with those obtained by present procedures ; the advantages over the use of substituted dithiocarba- mate-type reagents are discussed. The :preferred procedure is with 2: 2’- diquinolyl as reagent. TRACES of impurities in raw materials may have profound effects on photographic emulsions. It is desirable to have a method capable of determining the copper content of various gelatins in the range 0 to 15 p.p.m. The method to be selected had to fulfil the following conditions- (i) the reagents to be used should be readily available in a reasonably pure state; (ii) the procedure should be specific, yet simple and sensitive;April, 19581 RUSSELL AND HART: THE DETERMINATION OF COPPER I N GELATIS 203 (iii) it should preferably involve an extraction stage; the partition coefficient should (iu) the coloured complex produced should be stable, especially to light.The reagents used in well known methods and their shortcomings may be summarked Dithizone is non-specific,l unstable and has an intense colour of its own. Sodium diethyldithiocarbamate is used in the British Standards method.2 It is less sensitive than dithizonel and is not specific. An extraction stage into chloroform or carbon tetrachloride is often used, but the partition coefficient is small.The complex with copper is unstable to light.3 This type of reagent has been used in conjunction with ethylene- diaminetetra-acetic acid to increase specificity,4p5 but interference still occurs. Zinc dibenxyldithiocarbamnte also suffers from non-specificity6 and instability of the complex with copper to light. It has the advantage over sodium diethyldithiocarbamate that it can be used at a lower pH. More recently, other methods have become available, some of which have been examined. These include methods in which 2 : 2'-diquinolyl, biscyclohexanone oxalyldihydrazone or 2 : 0-dirnetliyl-1 : 10-phenaiithroline is used. Polarographic methods have also been examined. be high in favour of the organic phase; as follows- &THOD OF DESTROYING ORGANIC MATTER REAGENTS- &411 reagents should be of recogiiised analytical grade.Nitric mid, concentvnted. Perchloi.ic acid, 73 pev cent. Sulphziric acid, concentvated. PROCEDURE- Heat 2 g of gelatin and 10 ml of nitric acid in a 100-m1 conical flask on a hot-plate until vigorous evolution of brown fumes O C C U ~ S . ~ I t is advisable to renove the flask from the hot-plate at this stage. Add 2 ml of sulphuric acid and continue heating until no more brown fumes are evolved and charring begins. Add 4 m l of perchloric acid and continue heating until the liquid is colourless or very pale yellow and then maintain the solution at the same temperature for a further 3 to 4 hours to ensure the removal of excess of perchloric acid. Little attention is required at this stage and several digestions can be conducted simultaneously.After the solution has cooled, dilute with about 10ml of water, boil for a few minutes and then cool. The solution is now ready for the determination. COLORIMETRIC METHODS All colorimetric measurements were made with a Gallenkamp photo-electric colorirneter. 2 : 2'-Diquinolyl gives a magenta-coloured complex with cuprous copper, which can be extracted into isoamyl alcohol. isoAmy1 alcohol is said8 to give a slightly greater colour intensity than n-amyl alcohol. The complex of 2 : 2'-diquinolyl with cuprous copper is remarkably stable to atmospheric oxi&dtion.9 The partition coefficient is largelo and is relatively slightly affected by high salt concentrations. The solubility of isoamyl alcohol in water at room temperature is appreciable, hence temperature control is desirable at the extraction stage.The reagent is sensitive to oxidising agents that may occur in the isoamyl alcohol or remain after the digestion with perchloric acid.ll The alcohol is subjected to the pre-treatment described later and the digestion is prolonged to overcome these effects. It is claimed12 that the extraction and reduction stages can be combined by adding the solvent cliniethylformamide to the solution, but we have preferred to retain the advantages of CN traction. FVITH Z : 2'-DIQUINOLYL AS REAGENT- Reagents- 2 : 2'-DiqzainolyZ solution-A 0.02 per cent. w/v solution of analytical-reagent grade The commercial grade reagent, m.p. 186" According to Dr.J. Hoste the solution 2 : 2'-diquinolyl, m.p. 196" C,13 in isoamyl alcohol. to 192" C was also used without recrystallisstion. is stable for several months if stored in a brown bottle.204 RUSSELL AND HART: THE DETERMINATION OF COPPER IN GELATIN [VOl. 83 isonmy1 alcohol-Analytical-reagent grade, b.p. 128" to 132" C. This was treated before use in accordance with a communication from Dr. J. Hoste, as follows. An 800-ml portion was shaken with 100ml of a 10 per cent. solution of sodium metabisulphite, the layers were separated and the alcohol layer was, dried overnight in contact with anhydrous magnesium sulphate. After filtration, the alcohol was distilled and the fraction boiling over the range 128" to 132" C was collected and stored in a brown bottle.Tartaric acid solution, 50 per cent. w/v. Iiydroxylamine hydrochloride solution, 15 per cent. w/v-Prepare freshly each week.8 Sodium hydroxide solution, 30 per ccnt. wlv. Proccdure- After digestion to destroy the organic matter, treat the solution with 2 ml of tartaric acid solution and 2 ml of hydroxylamine hydrochloride solution. Adjust the pH to between 4 and 7 with sodium hydroxide solution (test by spotting micro drops on pH papers). Dilute the solution to 50 ml and transfer it to a thermostatically controlled water bath at 25" & 0.5" C for 10 minutes. Add 10 ml of 2 : 2'-diquinolyl solution, also at 25" C, and shake the mixture for 3 minutes. Separate the organic layer and measure its optical density, using an Ilford No. 625 filter. ReszLlts-Calibration curves with simple aqueous copper solutions and with known additions of copper to a de-ionised copper-free gelatin are shown in Fig.1. The displacement of one curve from the other is due to copper in the digesting acids and the sodium hydroxide used for neutralisation. 40 II Copper added, yg Fig. 1. Calibration curves for copper by 2: 2'-diqujnolyl method: curve A, digested with gelati~l; curve B, no digestion WITH BISC~C~OHEXANONE OXALYLDIHYDRAZONE AS REAGENT- known, it is not extractable. mining copper in paper and pulp15 and in plants.l6 of the metal hydroxide. Reagents- ethanol, warmed to dissolve the solute and filtered before use. freshly prepared each day. This reagent gives a blue colour with copper in the pH range 7 to 9 and, so far as is I t is highly specific for copper14 and has been used for deter- Citrate is added to prevent precipitation Biscyclohexanone oxalyldihydrazone solution--A 0.5 per cent.w/v solution in 50 per cent. This solution should be Ammoniwn citrate solution, 10 per cent. w/v. Ammonia solution, sp.gr. 0.880-Analytical-reagent grade. Neutral red indicator solution, 0.05 per cent. wlv. Sodium hydroxide solution, 50 per cent. w/:Y. Hydrochloric acid, concentrated-Analytical-reagent grade. Procedwc- Prolonged digestion is not necessary and, after a clear solution is obtained, heating for 1 hour is sufficient. After it has cooled, transfer the solution to a 25-ml calibrated flask, add 0.5 ml of ammonium citrate solution and 4 drops-about 0.15 ml--of ammonia solutionApril, 19581 205 and then 1 drop of indicator.Adjust the pH with sodium hydroxide or hydrochloric acid until the solution is just yellow (pH 7 to 9). Cool the solution frequently to minimise loss of ammonia. Add 1 ml of biscyclohexanone oxalyldihydrazone solution and dilute to the mark. Occasion- ally, cloudy solutions are obtained, which require filtration into the colorirneter cell. The calibration curves obtained in this way resemble those shown in Fig. 1. WITH 2 : 9-DIMETHYL-1 : 10-PHENANTHROLINE (NEOCUPROINE) AS REAGENT- used for 2:Z'-diquinolyl. tungsten.lg extracted by the same solvent. RUSSELL AND HART: THE DETERMINATION OF COPPER I N GELATIN Measure the optical density after 15 minutes, using an Ilford No. 626 filter. This reagent gives a yellow colour with cuprous copper under similar conditions to those It has been used for determinations of copper in paper1' and It is subject to the same interference from oxidising agents and the complex is I t is advisable17 to store the reagent solution in a refrigerator.Reagent s- Buffer solution, p H 5-Dissolve 57 g of anhydrous sodium acetate and 17.0 ml of glacial acetic acid in water and dilute almost to 1 litre. Adjust the pH to 5.0 and then dilute to 1 litre. Neocuproine solzction-Dissolve 75 mg of 2 : 9-dimethyl-l : 10-phenanthroline in 100 ml of buffer solution with vigorous shaking. *4 SCOY bic acid. Tartaric acid solution, 50 per cent. w/v. Procedure- After digestion, add 2 ml of tartaric acid solution. Adjust the pH to approximately 6, and add 10 ml of buffer solution. Treat the solution with 3 ml of neocuproine reagent and about 50mg of solid ascorbic acid, dilute to about 50ml and then place the solution in a thermostatically controlled water bath at 25" 0.5" C for 10 minutes.Add 10 ml of isoamyl alcohol (purified as for use with 2: 2'-diquinolyl), also at 25" C, and shake the mixture Store the reagent in a refrigerator. for 3 minutes. No. 622 filter. Separate the alcohol laye; and measure its optical density, using an Ilford POLAROGRAPHIC METHOD A polarographic method in which dry oxidation is used has been described.19 However, dry oxidation is not suitable for gelatin. The material froths considerably, and, on burning, leaves a mass of porous carbon, which is only slowly oxidised. The operation is tedious and requires considerable attention from the analyst, apart from the known risks of loss in such a procedure. The wet-oxidation procedure described is suitable for the purpose.Prolonged digestion is necessary and decomposition products of the perchloric acid may give spuriously high diffusion currents for the copper wave. Such effects have been noted by other workers.18 APPARATUS- A Tinsley pen-recording polarograph was used. TABLE I DETERMINATION OF COPPER IN GELATIN BY VARIOUS METHODS Copper found by- A - I._.--- _____ 7- ~ biscy clohexanone zinc dibenzyldithio- 2 : 2'-diquinolyl oxalyldihydrazone neocuproine polarographic carbamate method, method, method, p.p.m. p. p. in. p.p.m. 14.5 11.7 12.2 3.5 1.9 3.5 8.0 7.8 8.1 3.0 1-4 3.4 3.0 1.8 3.3 1.0 0.0 0.7 1.0 0.0 1.0 After digestion, transfer the cooled solution to mark with distilled water.This solution, which Pi<OCEDL?RE-- - metlhod, p.p.m. 13-4 3-3 8-6 2.4 2-8 0.5 1.0 method, p.p.m. 14.5 2.4 8.1 1.5 1-4 0-0 0.0 a 10-ml calibrated flask and dilute to the is approximately 5 N in sulphuric acid,206 RUSSELL AND HART: THE DETERMINATION OF COPPER IK GELATIN [VOl. 83 is used directly for the polarographic determination. No maximum suppressor is required. I t is convenient to use a mercury pool as anode; the half-wave potential of the copper wave is at -0.34 volt against the pool. Preliminary experiments showed that variations in sulphuric acid concentration from 3 to 8 N did not seriously affect the wave heights. RESULTS The results by these four methods on seven commercial gelatins are shown in Table I.Figures obtained with the zinc dibenzyldithiocarbamate procedure of AndruslS with special precautions to minimise light fading are included for comparison. In a series of recovery experiments, known amounts of copper were added to the gelatins before digestion. The average recoveries were as follows- 2 : 2’-Diquinolyl method . . .. .. . . . . 101 & 5 per cent. Biscyclohexanone oxalyldihydrazone method . . 96 5 per cent. Neocuproine method . . .. .. . . . . 99 i 5 pcr cent. Zinc dibenzyldithiocarbamate method. . . . . . 96 & 5 per cent. The results shown in Table I are reasonably satisfactory, considering the low coimn- trations involved and the simplicity of the instruments used for measuring. COMPARISON OF THE METHODS S E NSI TI v ITY - According to the values given in the literature, zinc dibenzyldithiocarbamate6 has about the same sensitivity as biscyclohexanone oxalyldihydrazone,20 when both complexes are measured with a spectrophotometer.In this work, with the best filter available, the zinc reagent was only about one-third as sensitive. On the other hand, biscyclohexanone oxalyldihydraz- one was found to be about 2i-times as sensitive as 2:2’-diquinolyl. This figure is in agreement with those of other However, this advantage is offset by the “concen- trating” effect of the extraction stage in the latter case. Like other workers,17 we have found 2 : 2’-diquinolyl and neocuproine to be very similar in sensitivity. Caution must be exercised in comparing sensitivities with filter iiist ruments.I ” I I ‘ I I OLI-.--L 0 5 10 50 100 500 IOOC Time, rnhutes Fig. 2. Colour stability: curvc ~1, sodium diethyldithiocarbamate in fluorescent light; curve B, sodium diethyldithiocarba mate i n diffuse day- light; curve C, 2 : 2’-diquinolyl and neocuproine in all lighting : curve D, biscyclohexanone oxalyldi- hydrazone in all lighting; curve E, sodium diethyl- dithiocarbamate in sunlight INTERFERENCE- The only known interference when the 2 : 2’-diquinolyl and neocuproine methods are used is from ferric iron, and this can be masked by tartaric acid. The zinc dibenzyldithio- carbamate method is affected by the presence of bismuth, cobalt and nickel, and the polaro- graphic method may be affected by stannous tin. There is no known interference with the biscyclohexanone oxalyldihydrazone method.April, 19581 RVSSELL AND HART: THE DETERMINATION OF COPPER IN GELA4TIN 207 STABILITT OF THE COMPLEX TO LIGHT- The relative stabilities of the various complexes to light of different kinds are illustrated in Fig.2 (note the logarithmic time scale). A curve is shown for sodium diethyldithio- carbamate ; the zinc dibenzyldithiocarbamate reagent complex fades almost as quickly in bright light. As has been reported,21 the biscyclohexanone oxalyldihydrazone complex fades slowly; this may be partly due to high salt ~0ncentration.l~ The 2:T-diquinolyl and neo- cuproine complexes are stable to light for at least several days. STABILITY OF THE REAGENT- Carbamate-type reagents deteriorate on keeping.22 J~~ 2 : 2’-Diquinolyl solution, when prepared as described, is stable for a t least 3 months.Biscyclohexanone oxalyldihydrazone solution is less stable and becomes yellow after a few days. A freshly prepared solution of this reagent, with the addition of a standard amount of copper, gave a reading of 25.2 colori- meter units; when 4 days old, with the same amount of copper, it gave a reading of 20.6 units. The need to keep neocuproine solution in a refrigerator has been mentioned. When the solution was kept at laboratory temperature, the reading with a standard amount of copper fell from 25.7 units when fresh to 22.0 units at 4 days old. PKICE OF THE REAGENT- the difference in the concentrations used, they are about the same cost. oxalyldihydi-zone is about the same price as carbamate reagents. 2 : 2’-Diquinolyl and neocuproine are fairly expensive reagents; by taking into account Biscyclohexanone CONCLUSIONS 2 : 2’-Diquinolyl and neocuproine are equivalent in most respects except for the stability of the reagent.The stabilities of the complexes to light favour these reagents; simplicity, sensitivity and economy favour biscyclohexanone oxalyldihydrazone. All these reagents have advantages over sodium diethyldithiocarbamate. The polarographic method is simple and convenient , although possibly not so generally attractive as a colorimetric method. By taking into account all the above factors, the preferred reagent is 2:2’-diquinolyl. The procedure described, used in place of the British Standard method, has given satisfactory results during 6 months. We thank the Directors of Ilford Limited for permission to publish this paper, and Miss M.E. Bell, Miss P. T. King and Mrs. G. A. Nichols for technical assistance. The value of the 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. communication frGm DF, J. Hoste of Ghent University is greatly appreciated. REFERENCES Sanclell, E. B., “Colorimetric Determination of Traces of Metals,” Second Edition, Interscience “Sampling and Testing of Gelatin,” British Standard 757 : 1944. Ovenston, T. C. J., and Parker, C. A., Anal. Chim. Acta, 1950, 4, 135. Sedivek, V., and VaSAk, V., Coll. Czech. Chem. Compz., 1950, 15, 260. Forster, W. A,, Analyst, 1953, 78, 614. Johnson, m’. C., Editor, “Organic Reagents for Metals,” Fifth Edition, Hopkin and Williams Ltd., Chadwell Heath, Essex, 1955, p. 188. Reed, J. l?., and Cummings, R. W., I n d . Rng. Chem., A~zal. E d . , 1041, 13, 124. Guest, R. J., Anal. Chenz., 1953, 25, 1484. Gahler, A. R., Ibid., 1954, 26, 577. Hoste, J., Eeckhout, J., and Gillis, J., Anal. Chim. A d a , 1953, 9, 263. Ferrett, D. J., and Milner, G. W. C., Analyst, 1956, 81, 193. Pflaum, R. T., Popov, A. I., and Goodspeed, N. C., Anal. Chenz., 1955, 27, 253. Iloste, J., Anal. Claim. Acta, 1950, 4, 24. Nilsson, G., Acta CJzem. Scand., 1950, 4, 205. Wetlesen, C. U., and Gran, G., Sveizsk Papperstidn., 1952, 55, 212. Williams, T. R., and Morgan, R. R. T., Chetn. G. I n d . , 1954, 461. Zak, B., and Ressler, N., Anal. Ckenz., 1956, 28, 1158. Crawley, R. H. A., Anal. Chim. A d a , 1955, 13, 373. Rlichel, G., and Maron, N., Ibid., 1960, 4, 542. Peterson, K. E., and Bollier, M. E., Anal. Chew,., 1955, 27, 1195. Somers, E., and Garraway, J. L., Chenz. & I n d . , 195‘7, 395. Johnson, W. C., Editor, op cit., p. 172. Martens, R. I., and Githens, R. E., 4naZ. Chein., 1952, 24, 991. Publishers Inc., New York, 1950, p. 300. Received November 1 lth, 1957