246 Artalyst, April, 1973, Vol. 98, PP. 246-250 Bis( 6 - methyl-2 = pyridy1)glyoxal Dihydrazone as a Spectrophotometric Reagent for the Rapid Determination of Copper in Alkalis, Milk and Brine BY M. VALCARCEL AND F. PIN0 (Department of Analytical Chemistry, University of Seville, Seville, Spailz) The synthesis, characteristics and analytical applications of bis(6-methyl- 8-pyridyl) glyoxal dihydrazone are described. This compound produces coloured solutions selectively with copper(1) ions (Amax. = 440 nm, E = 8.7 x 103 1 mol-1 cm-l) that can be extracted into various organic solvents; it behaves as a cuproine-type reagent. The 1: 1 orange-yellow copper(1) complex has been used for the spectrophotometric determination of trace amounts of copper in saturated brine, alkalis and milk.The most important advantage of using this reagent is the total recovery of copper that is possible from ammoniacal solutions. THE work described in this paper forms part of an investigation into the use of hydrazones for trace-metal analysis. The ferroin-type reagents, 2,2'-bipyridylglyoxal dihydrazone, di- acetyl dihydrazone and phenyl 8-pyridyl ketone hydrazone, have been applied1 to the spectrophotometric determination of trace amounts of iron while 6-methylpicolinaldehyde hydrazone was used2 as a selective cuproine-type reagent for the determination of copper. In the present work, bis(6-methyl-2-pyridyl)glyoxal dihydrazone was used for the selec- tive determination of trace amounts of copper in different materials. EXPERIMENTAL SYNTHESIS OF REAGENT- Five millilitres of 99-9 per cent.m/m hydrazine hydrate were added to 6 g of bis(6-methyl- 8-pyridy1)glyoxal (Fluka) dissolved in 150ml of hot absolute ethanol and the mixture was refluxed for 24 hours. On cooling to 0 "C, the crystals that separated out were filtered off and recrystallised twice from hot ethanol. The crystals finally obtained, dried at 60 "C under pressure, were white and melted at 154 to 155 "C. On analysis their elemental content was found to be: C 62.5, H 5-9 and N 31.5 per cent.; the content calculated for Cl4HI6N6 was: C 62.67, H 5.96 and N 31.34 per cent. APPARATUS- meters, equipped with l-O-cm glass or quartz cells. SOLUTIONS- Spectrofikotometers-Unicam SP800, Unicam SP600 and Beckman DU spectrophoto- Digital fiH meter-Philips, PW9408, with glass - calomel electrodes.All solvents and reagents were of analytical-reagent grade. Bis(6-methyl-2-pyridyl)glyoxal dihydrazone reagent solutions-Solutions of concentration Standardised solutions of copper(II). Ascorbic acid, 2 per cent. m/V aqueous solution (as reducing agent). Acetic acid - sodium acetate bzc$er solution, pH 4.8. Q SAC and t h e authors. 0.1 per cent. m/V in ethanol and 0.05 per cent. m/V in nitrobenzene.VALCARCEL AND P I N 0 247 Extraction solution 1-Dissolve 5 g of sodium perchlorate monohydrate and 3 g of ascorbic Extvaction solution 2-Dissolve 5 g of sodium perchlorate monohydrate and 3 g of ascorbic Trichloroacetic acid, 50 per cent. m/V aqueous solution. acid in 250 ml of the above buffer solution. This solution remains stable for 1 week.acid in 250 ml of distilled water. This solution also remains stable for 1 week. RECOMMENDED PROCEDURES ALKALIS- To 10 ml of extraction solution 2 in a separating funnel add 50 ml of alkali, followed by 10 ml of 0.05 per cent. m/V reagent solution in nitrobenzene. Shake the mixture vigorously for 2 minutes, allow the phases to separate and transfer the lower, organic, layer into a flask containing anhydrous sodium sulphate. Measure the absorbance of this solution at 440 nm against the reagent solution in nitrobenzene. Prepare a calibration graph in a similar manner by adding appropriate amounts of copper to 2 M sodium hydroxide solution. SATURATED BRINE- Method A-To 25 ml of saturated brine in a 50-ml calibrated flask, add 10 ml of acetic acid - sodium acetate buffer solution, 2 ml of 2 per cent.m/V ascorbic acid solution and 5 ml of 0.1 per cent. m/V ethanolic reagent solution, and dilute to the mark with distilled water. Measure the absorbance of the solution at 440nm against a reagent blank prepared in a similar manner. Method B-To 10 ml of saturated brine in a separating funnel add 10 ml of extraction solution 1 and 10 ml of 0.05 per cent. m/V reagent solution in nitrobenzene. Shake the funnel vigorously for 1 minute, allow the phases to separate and transfer the lower layer into a flask containing anhydrous sodium sulphate. Measure the absorbance at 440 nm against the reagent solution in nitrobenzene. Calibration graphs are constructed from standard solutions treated in the same way. MILK- To 100 ml of milk in a 300-ml Erlenmeyer flask, add 25 ml of 50 per cent.m/V trichloro- acetic acid solution, slowly and with constant shaking. Shake the flask vigorously, place it in a boiling water bath for 16 minutes and cool it in ice - water to 10 "C. Transfer 25 ml of the supernatant liquid into a separating funnel. Add 7 ml of 2~ sodium hydroxide solution, 10ml of extraction solution 1 and 10ml of 0.05 per cent. m/V reagent solution in nitrobenzene. Shake the funnel vigorously for 1 minute, allow the phases to separate and transfer the lower layer into a flask containing anhydrous sodium sulphate. Measure the absorbance of this solution at 440 nm against the reagent solution in nitrobenzene. Obtain the content of copper, in micrograms, from a standard graph prepared by substituting the appropriate copper solutions and distilled water for the milk sample.RESULTS AND DISCUSSION BIS(6-METHYL-2-PYRIDYL)GLYOXAL DIHYDRAZONE REAGENT- The ultraviolet spectrum for the reagent shows a bathochromic shift in an acidic medium (Amax. 347 and 268 nm) compared with an alkaline medium (Am,,, 295nm, with a shoulder at 266 nm), with two isosbestic points at 277 and 313 nm. The Phillips and Merritt3 method is used for the determination of the ionisation constant; the average pK value is 4.82. This behaviour may be caused by protonation of the nitrogen atoms in the pyridine rings; the pK value is very similar t o that of pyridine and its derivatives. The reagent is resistant to hydrolysis in a strongly acidic medium (6 M hydrochloric acid).This property is usual with a-diimines, in contrast with the corresponding imines ; 6-methyl- picolinaldehyde hydrazone2 hydrolyses in 2 M hydrochloric acid at 20 "C in 30 minutes. The reaction of the reagent with thirty cations at various pH values was investigated; it reacts only with copper(1) and palladium, and the absorption spectra of solutions of these metals are shown in Fig. 1. It acts as a cuproine-type reagent in that methyl groups adjacent to the nitrogen atom in pyridine produce the well known blocking effect.248 VALCARCEL AND P I N 0 : BIS(6-METHYL-2-PYR1DYL)GLYOXAL DIHYDRAZONE [A?ZaZy!yst, VOl. 98 " 400 450 500 550 600 650 Wavelengthhm Fig. 1. Absorption spectra of solutions of COIII- plexes formed with reagent: 1, 10 p.p.m. of copper ( 1 ) ; 2, 10 p.p.m.of palladium; and 3, reagent alone (water blank) REACTION WITH COPPER(I)- Aqueous media-The orange - yellow 1 : 1 copper(1) complex (ratio found by the Job absorptiometric method, by isolation of the perchlorate complex, [Cu(C,,H,,N,)]ClO,, in the solid state and elemental analysis*) of the reagent is formed completely over the pH range from 4.5 to 11.2 in aqueous solution (see Fig. 2). The effect of other experimental variables was determined. The system conforms to Beer's law over this pH range, the molar absorptivity ( E ) being 8.7 x lo3 1 mol-l cm-l. The optimum concentration range, evaluated by Ringbom's method, is 2 to 5 p.p.m. of copper. The colour intensity of the solutions of the complex remains constant for several hours. The relative error (P = 0.05) of the method is &0-13 per cent.The colour reaction is selective for copper. Silver, mercury( 11), mercury(I), cadmium, zinc, iron(II), cobalt(II), thallium, lead, tin(II), uranium(VI), calcium, strontium and barium do not interfere at the 300 p.p.m. level in pH 4.8 buffer. Nickel does not interfere at the I 1 Fig. 2. Effect of pH on the formation of the copper (I) complex: 1, in water; 2, in pent- anol; 3, in nitrobenzene; and 4, in chloroform * The results of the analysis were C 38.7, H 3.5 and N 19.1 per cent. ; calculated values, C 38.96, H 3-71 and N 19-48 per cent.April, 19731 AS A SPECTROPHOTOMETRIC REAGENT FOR THE DETERMINATION OF coprm 249 50 p.p.m. level, while iron(III), bismuth, antimony(II1) and manganese( 11) precipitate at the 100 p.p.m.level. The most serious interferences are from palladium, gold(III), EDTA and oxalate. At pH 8.7 interferences are greater. If an excess of reagent instead of ascorbic acid is used as reducing agent, the same results are obtained over the pH range 4.5 to 9. Extraction-The coloured complex formed in aqueous solution can be extracted into various organic solvents, such as pentanol, chloroform and isobutyl methyl ketone, with no perchlorate, but in the pH range from 9 to 12 the absorbance of the organic layer is not stable with time (Fig. 2). Copper(1) ions are completely extracted with either nitrobenzene or chloroform, with perchlorate in the aqueous phase. The optimum pH range is 3.2 to 5.7 (nitrobenzene) and 4-5 to 5.7 (chloroform); from pH 9 to 12 the absorbance remains lower but stable with pH and time (Fig.2). Beer’s law is adhered to at pH 4.8 and 10.9 with both of these solvents. The coloured organic solutions are stable for several hours. The extraction has a high relative error with chloroform but the relative error (P = 0.05) of the method with nitrobenzene is &0.36 per cent. The optimum concentration range evaluated by Ringbom’s method is 2 to 6 p.p.m. of copper. After the extraction the effect of the ions mentioned above can be investigated. Those ions which did not interfere with the colour reaction do not interfere when using this technique at pH 4.8 or 11. Interference from nickel, iron(III), bismuth and manganese(I1) is now suppressed up to a level of 300 p.p.m. Palladium, gold(TIT), EDTA and oxalate also interfere in this method.DETERMINATION OF COPPER- Typical results for the determination of copper in sodium hydroxide, potassium hydroxide and ammonia solutions are given in Table I. The method is very useful because the recovery TABLE I DETERMINATION OF COPPER IN ALKALIS Each result is the average of three separate analyses Solution Copper added/pg g-1 Copper recovered/pg g-’ liecovery, per cent. Sodium hydroxide. . .. 10.0 10.26 102.5 Potassium hydroxide . . 14.3 14.6 102.1 Ammonia .. .. .. 24.1 24.1 100.0 from ammonia solutions is complete. The present reagent is more suitable than 6-methyl- 2-pyridylphenyl ketoxime, proposed by Pemberton and Dieh14 in 1969 as a reagent for the determination of copper in alkalis, because with the latter the recovery of copper from am- monia solutions is only partial (32 per cent.).Trace amounts of copper in saturated brine can be determined accurately in homo- geneous media (Method A) and with the extraction technique (Method R). Typical results for the determination of copper are given in Table 11. TABLE I1 DETERMINATION OF COPPER IN SATURATED BRINE Each result is the average of three separate analyses Copper r - Copper found/pg ml-1 Recovery, per cent. added/pg ml-l Method A Method €3 Method A Method B 2.0 4.0 6.3 6.8 2.0 4.0 6-2 6.8 1.95 100 97.6 4.0 100 100 6.3 98.6 100 6.7 100 99 The reagent is also suitable for the rapid determination of the copper content of milk. The results (Table 111) are comparable with those obtained by Smiths in 1967 when using the same method with a 0.05 per cent. m/V solution of zinc dibenzyldithiocarbamate in toluene.260 VAZCARCEL AND PIN0 TABLE I11 DETERMINATION OF COPPER IN MILK Number Average concentration of Confidence limits Commercial sample of determinations copperlpg per 100 ml (96 per cent.) 1 6 34.6 f3-6 2 6 8 f 2.3 3 6 12 f 2.6 REFERENCES 1. 2. 3. 4. 6. Graciani, E., Ph.D. Thesis, University of Seville, 1969 (and unpublished work). Valcarcel, M., and Pino, F., Infcidn. Quim. Analit. Pura APl. Ind., 1972, 26, 116. Phillips, J . P., and Merritt, L. L., J. Amer. Chem. SOC., 1948, 70, 410. Pemberton, J. R., and Diehl, H., Talanta, 1969, 16, 393. Smith, A, C., J. Dairy Sci., 1967, 50, 664. Received July 12th. 1972 Accepted November loth, 1972