142 Analyst, March, 1968, Vol. 93, $9. 142-147 The Spectrophotometric Determination of Vanadium(V) with 3,3’=Dimethylnaphthidine BY LORNA G. BANNARD AND J. D. BURTON (Department of Oceanography, The University, Southampton) An investigation has been made of the factors concerned in the reaction of vanadium(V) with 3,3’-dimethylnaphthidine. Some probable causes of previous difficulties in the application of the reaction in spectrophotometric analysis have been elucidated. A simple procedure is given for the spectro- photometric determination of vanadium in the range from 0.08 to 2 p.p.m.; with 0.32 p.p.m. the coefficient of variation ranged from 0.60 per cent. for analyses at one time to 1.8 per cent. for analyses over a period of 3 months. Small, but significant, deviations from adherence to Beer’s law occur.The limit of detection is about 0.004 p.p.m. of vanadium. The method is selective; the effects of numerous foreign ions have been examined, and interfering effects were found only with chromium(VI), iron(II), iron(III), cerium(II1) and cerium(1V). THERE is conflicting evidence concerning the value of 3,3’-dimethylnaphthidine (3,3‘-dimethyl- 4,4‘-diamino-1 ,l’-binaphthyl) as a reagent for the spectrophotometric determination of vanadium(V). Introduced as a spot-test reagent by Belcher, Nutten and Stephen,l it was later applied to the analysis of vanadium in alloys by Milner and Nall,2 who found that in a solution containing sulphuric and orthophosphoric acids, the reaction was rapid, and the colour of the oxidation product was stable for more than 30 minutes.Under apparently identical conditions, Scholes3 found that the colour was unstable. Although the stability could be improved by varying the acid concentrations, the reaction was still erratic and calibration graphs prepared at intervals were widely divergent; he rejected the use of the reagent for precise determinations. Similarly conflicting reports have been made of the reaction in solutions containing perchloric and orthophosphoric acids. Forrester and Jones4 described conditions under which the colour was stable for about 1 hour, but their findings were not confirmed by Macmillan and SamueL6 These contradictory reports suggested that the influences of various conditions on the reaction were inadequately understood. As the reagent has considerable advantages in its selectivity and high sensitivity for vanadium (it is, for example, about five times as sensitive as 3,3’-diaminobenzidine6), it seemed worthwhile to examine its use more fully.The results of an investigation into the factors concerned in the reaction are presented in this paper. EXPERIMENTAL Preliminary experiments were commenced by using the conditions recommended by Forrester and Jones4 To a solution containing 5 ml of dilute orthophosphoric acid (1 + 1, with 90 per cent. w/w H,PO,), 5 ml of dilute perchloric acid (1 + 1, with 72 per cent. w/w HC10,) and 8 pg of vanadium(V), 2.5 ml of a 0.1 per cent. w/v solution of 3,3‘-dimethyl- naphthidine in glacial acetic acid were added. The solution was diluted to 25 ml and measure- ments of optical density were made at 550 mp in a 4-cm cell, at intervals.The results (Fig. 1) confirmed the findings of Macmillan and Samuel6 that the colour was not stable. The experiment was repeated with reagent solutions of various other concentrations (viz., 0.003, 0.01 and 0-03 per cent. w/v). The results (Fig. 1) showed that colour development was markedly retarded at the lowest concentration, but that with the 0.01 per cent. w/v reagent 0 SAC and the authors.BANNARD AND BURTON 143 there was a period of about 25 minutes during which the colour was stable; this reagent concentration, which was used by Milner and Nal12 in solutions containing sulphuric and orthophosphoric acids, was used in all subsequent work. Investigations in which sulphuric acid was used, instead of perchloric acid, were not pursued because the rate of colour develop- ment was found to be slow under these conditions.10 20 30 40 50 60' Time, minutes Fig. 1. Effect of concentration of 3,3'-dimethylnaphthidine on the rate of colour development in solutions containing perchloric and orthophosphoric acids. Each solution contained 8 pg of vanadjum(V) and 2.5 ml of reagent solution in glacial acetic acid. Concentrations of 3,3'-dimethylnaph- thidine solution: curve A, 0.1 per cent.; curve B, 0.03 per cent.; curve C, 0.01 per cent.; curve D, 0.003 per cent. Optical densities are uncorrected for reagent blanks (about 0.010) The effects of varying the amounts of acid were studied next, but erratic results were frequently obtained. It was found that these were attributable to three causes.Firstly, the reaction is sensitive to sunlight; the optical density of a solution exposed directly to natural light during colour development was about 30 per cent. lower than that of an identical solution that was kept in darkness or subdued light during the development period. Secondly, there can be variations in sensitivity and rate of colour development between different batches of solid reagent. This effect was noted with two batches (from one supplier), one of which was noticeably darker than the other in the solid state. The response of both batches to changes in conditions was similar, despite the difference in rate of reaction. This question is discussed further below. The third cause of variable results was that, in some experiments, solutions containing vanadium, orthophosphoric and perchloric acids were allowed to stand for appreciable periods before colour development.Under these circum- stances low results are obtained, presumably because some reduction of vanadium takes place. A reduction in optical density of 5 per cent. was found to occur when the delay in adding the reagent was 1 hour, and when the period was extended to 3 hours the reduction was 9 per cent.; allowing the solution to stand for longer periods had little further effect. In further experiments, the procedure was modified to take account of these factors, and the effect of varying the concentration of each component in turn was examined, while keeping the concentration of the other components constant.These determinations were144 BANNARD AND BURTON : SPECTROPHOTOMETRIC DETERMINATION [Analyst, vol. 93 made with 8 pg of vanadium(V) in a final volume of 25 ml; optical densities were measured at 550 mp in 4-cm cells and corrected for reagent blanks, which were generally about 0.010. The effect of varying the amount of acetic acid added with the 3,3’-dimethylnaphthidine reagent was examined first; the solutions contained 5 ml of dilute orthophosphoric acid and 5ml of dilute perchloric acid. The results are given below. Volume of glacial acetic acid, ml . . 0.25 0.625 1.25 2.5 5.0 7.5 Optical density (less blank) . . . . 0.402 0.400 0.405 0.410 0.406 0.404 Variations in the concentration of acetic acid in this range thus had no considerable effect on the reaction.In subsequent experiments, 2.5ml of a reagent solution in 25 per cent. v/v acetic acid were added. This reagent solution is oxidised only slowly and shows no loss of sensitivity for 36 hours. The effect of varying the amount of dilute perchloric acid was studied next ; each solution contained 5 ml of dilute orthophosphoric acid. The following results were obtained. acid, ml . . 0 1 2 3 4 5 7.5 10 With solutions containing 5 ml of dilute perchloric acid, the effect of varying the amount Volume of dilute perchloric Optical density (1,s-blankj 0.353 0.374 0.387 0.404 0.404 0.401 0.392 0-381 of dilute orthophosphoric acid was investigated, with the following results. Volume of dilute orthophosphoric acid, ml . . 0 2-5 5 7.5 10 Optical density (less blank) .. .. . . 0.165 0.358 0.390 0.400 0.393 As the omission of perchloric acid had a comparatively small effect on the reaction, the possibility of using orthophosphoric acid alone was examined. The rate of colour development was measured first, 10 ml of dilute orthophosphoric acid being used. In Fig. 2, 10 20 30 40 50 60 Time, minutes Fig. 2. Rate of colour development in orthophosphoric acid solution with different batches of solid 3,3’-dimethylnaphthidine. Each solution contained 8 pg of vanad- ium(V) ; the reagent was added as an 0.01 per cent. solution in 25 per cent. v/v acetic acid. Curve A, with reagent used for results shown in Fig. 1; curve B, with different batch of 3.3’-dimethylnaphthidine under identical con- ditions. Optical densities are uncorrected for reagent blanks (about 0-010) the results are shown for the two batches of 3,3’-dimethylnaphthidine referred to previously.Curve A was obtained with the lighter-coloured reagent that was used to obtain the results shown in Fig. 1. Comparing this curve with curve C in Fig. 1, it is seen that the reaction occurs rather more rapidly in orthophosphoric acid solution than in the mixed solution, and a slight increase in sensitivity is obtained; the period of stability of the complex is somewhatMarch, 19681 OF VANADIUM(V) WITH 3,3'-DIMETHYLNAPHTHIDINE 145 reduced. With the second batch of reagent, the colour was more fugitive, but there was still an adequate period of stability of 10 minutes (curve B, Fig. 2) ; there was some reduction in sensitivity. A similar change in the rate of colour development occurred when the same reagent was used in solutions containing perchloric and orthophosphoric acids.Except for the results in Fig. 1 and curve A, Fig. 2, the values reported in this paper were obtained with the less satisfactory batch of reagent, measurements being made within the period of stability of the colour, according to the solution used. The effect of varying the concentration of orthophosphoric acid, while omitting perchloric acid, was investigated next, and the results are shown below. Volume of dilute orthophosphoric acid, rnl 0 1 2-5 5 7.5 10 15 Optical density (less blank) . . . . 0.009 0.090 0.157 0-291 0.388 0.425 0.456 In the conditions finally adopted for the determination, perchloric acid was omitted and 10 ml of dilute orthophosphoric acid were added.The absorption curve of the oxidation product was unchanged. There was no significant effect of temperature on the reaction in the range 10" to 25" C. REPRODUCIBILITY, BEER'S LAW AND LIMIT OF DETECTION- The reproducibility of the method was tested by making twelve measurements under the conditions of the recommended procedure and using 8 pg of vanadium(V). A coefficient of variation of 0.60 per cent. was obtained. Similar measurements were made in groups of up to four determinations, at intervals over 3 months; for 135 such determinations, the coefficient of variation was 1.8 per cent. Replicate determinations of various amounts of vanadium(V), in the range 0.25 to 50 pg, were made to investigate whether Beer's law was obeyed.The results (Table I) show that there was significant deviation from linearity, the response being reduced with lower concentrations ; the effect was especially pronounced at concentrations below 0.08 p.p.m., which may be taken as the lower end of the useful working range of the method. TABLE I DETERMINATION OF VANADIUM(V) Amount of vanadium(V) , pg per 25 ml 0-25 1 2 4 6 8 10 25 50 Mean optical density (less blank) in 4-cm cell 0.010 0.047 0.101 0.208 0.315 0.424 0.527 1-344* 2-708* Optical density per pg present 0*0400 0.0470 0.0505 0.0520 0,0525 0.0530 0.0527 0.0538 0.0542 Deviation from linearity, relative to highest concentration, per cent. 26.2 13.3 6-8 4- 1 3.1 2.2 2-8 0.7 * Measured in l-cm cell and calculated for 4-cm cell. The average value for the optical density of the blank in a 4-cm cell was 0.008, with a standard deviation of less than 0.001.If the response was linear below a concentration of 0.01 p.p.m., then, by using Wilson's definition' of the limit of detection, there would be a high probability of detection with 0-002 p.p.m. of vanadium. Because there is deviation from linearity, a somewhat higher limit of detection must apply. It was found experimentally that 0-004 p.p.m. of vanadium could be'detected with a high degree of certainty. EFFECTS OF FOREIGN IONS- The effects of various foreign ions were examined by carrying out determinations of 8 pg of vanadium(V) in the presence of each ion; measurements were also made in the absence of added vanadium. The results (Table 11) show the reagent to be selective for vanadium(V) .Cerium( IV) and chromium(V1) interfere by oxidising 3,3'-dimethylnaphthi- dine; cerium(II1) has a suppressing effect on the reaction with vanadium. The most important interfering ion likely to be encountered in significant amounts is iron. Iron(I1) interferes strongly, presumably by reducing vanadium ; iron(II1) also interferes, although to a lesser extent.146 BANNARD AND BURTON : SPECTROPHOTOMETRIC DETERMINATION [Analyst, Vol. 93 Amount, Ion added CLg Sodium . . . . 10,000 Magnesium . . 1000 Aluminium . . 100 Potassium .. 1000 Calcium .. 1000 Titanium(1V) .. 100 Chromium(II1) .. 100 Chromium(V1) .. 40 Manganese(I1) . . 100 Iron(I1) . . .. 5 10 Iron(II1) . . . . 60 100 Nickel . . . . 100 Copper .. .. 100 Zinc ... . 100 Arsenic(V) . . .. 100 Selenium(V1) . . 100 Niobium . . . . 100 Molybdenum(V1) 100 TABLE I1 EFFECTS OF Recovery Optical of 8 p g of density in vana- absence of dium(V), vanadium* per cent. 0.002 101.5 0-000 102.0 0.000 99.5 0.001 100.0 0.00 1 100.6 0.000 100.5 0.001 101.0 0-316 >400 - 0.001 99-8 -0-001 48-9 0~000 < 1 0.000 56-0 0-000 12-5 0-000 99.5 0.000 100-5 - 0.002 100.8 0~000 100.0 0.00 1 99.8 0.001 100.2 - 0.002 100.5 FOREIGN IONS Ion added Cadmium .. Tin(I1) . . Tin(1V) . . Antimony( I1 I) Barium Cerium( IV) Tantalum . . Tungsten (VI) Lead . . Thorium . . Uranium(V1) Chloride . . Nitrate . . Sulphate . . Phosphate Fluoride . . Oxalate . . Citrate . . Ceriurn(II1 j * Amount, CCg .. 100 . . 100 .. 100 . . 100 .. 100 .. 100 . . 50 .. 100 . . 100 .. 100 . . 100 .. 100 . . 1000 . . 1000 . . 1000 . . 1000 . . 1000 .. 1000 .. 1000 Recovery Optical of 8 p g of density in vana- absence of dium(V), vanadium* per cent. 0.001 100.0 0.006 99.3 99.8 0.002 0.000 100.7 99.8 0.000 0.00 1 81.8 0.691 107.6 99.6 0.001 100.6 0.00 1 O*OOO 100-5 0.002 100.5 100.6 0.001 0.000 98.8 0.000 101.5 0.001 98.8 - 0.002 99-1 0.00 1 98-8 0.000 99.5 -0.004 101.0 * Optical densities measured in 4-cm cell and corrected for reagent blank. Experiments in this laboratory on the application of the reagent to the analysis of vanadium in natural waters have shown that small amounts of vanadium are readily separated from iron( 111) by anion exchange in concentrated hydrochloric acid solution. After any preliminary separation, the vanadium must finally be converted into the quinqui- positive oxidation state.This may be achieved by evaporation with perchloric acid. During the present work it has been found that when solutions containing microgram amounts of vanadium are evaporated alone, with concentrated nitric or perchloric acids, or with dilute ammonia solution, the vanadium is not subsequently dissolved quantitatively in water or in dilute perchloric or orthophosphoric acids. The residues are completely dissolved, however, in a 1 per cent. w/v solution of sodium hydroxide. METHOD APPARATUS- against water in the compensator cell. Beckman DB spectrophotometer. Measurements of optical density were made with a Unicam SP500 spectrophotometer Absorption curves were plotted directly from a REAGENTS- 3,3’-Dimethylnaphthidine reagent-Dissolve 0.01 g of 3,3’-dimethylnaphthidine in 25 ml of glacial acetic acid and 5 ml of water, and dilute the solution to 100 ml with water.Discard the solution after 36 hours. Dilute orthophosphoric acid (1 + 1)-Mix equal volumes of analytical-reagent grade 90 per cent. w/w orthophosphoric acid and water. Standard vanadium ( V ) stock solutiort-Dissolve 0.4592 g of analytical-reagent grade am- monium metavanadate in water and dilute to 100ml. 1 ml of solution = 2 mg of vanadium(V). Standard vanadi.um( V ) dilute solution-Prepare, as needed, by appropriate dilution of the standard stock solution. 1 ml of solution = 2 pg of vanadium(V).March, 19681 OF VANADIUM(V) WITH 3,3’-DIMETHYLNAPHTHIDINE 147 RECOMMENDED PROCEDURE- Ensure that the vanadium is present in the quinquipositive oxidation state.To the test solution, containing 2 to 50 pg of vanadium, add 10 ml of dilute orthophosphoric acid and sufficient water to dilute the solution to about 20ml. Without delay, add 2 6 m l of 3,3’-dimethylnaphthidine reagent and dilute the solution to 25 ml. Keep the solution in the dark for 15 minutes and measure its optical density at 550 mp. Determine the reagent blank in a similar way, omitting the sample. Prepare a calibration graph by using the standard vanadium(V) dilute solution. Run duplicate standards of one concentration at least daily. Check the rate of colour development whenever a new batch of solid 3,3’-di- methylnaphthidine is used. CONCLUSIONS The reaction of vanadium(V) with 3,3’-dimethylnaphthidine provides a sensitive and selective method for the spectrophotometric determination of the element.Previous difficul- ties experienced with the reagent may have been caused by failure to take account of the effects of sunlight on the reaction, variations in the behaviour of different batches of the solid reagent and changes in the reactivity of vanadium on allowing it to stand in strongly acidic solutions. The use of too high a concentration of reagent has also probably been a cause of unsatisfactory results in some work. Colour development in orthophosphoric acid is as satisfactory as that in a mixture of orthophosphoric and perchloric acids. The colour forms rapidly and is stable for an adequate period. The reagent solution is stable for 36 hours and has a low extinction at the relevant wavelength. Under the conditions given in the recommended procedure the reagent may be used for precise determinations in the range from 0-08 to 2 p.p.m. of vanadium. With 0.32 p.p.m., the coefficient of variation was 0.60 per cent. for replicates analysed at one time; over a period of 3 months the corresponding value was 1.8 per cent. There are small, but significant, deviations from adherence to Beer’s law, but accurate values are obtained by the use of a calibration graph. Of the numerous foreign ions investigated, only chromium(VI), iron(II), iron(III), cerium(II1) and cerium(1V) had appreciable interfering effects. The authors are grateful to the Natural Environment Research Council for a grant in support of this work. The limit of detection is about 0.004 p.p.m. of vanadium. REFERENCES 1. 2. 3. 4. 5. 6. 7. Belcher, R., Nutten, A. J., and Stephen, W. I., Analyst, 1951, 76, 430. Milner, G. W. C., and Nall, W. R., Analytica Chim. Acta, 1952, 6, 420. Scholes, P. H., Analyst, 1957, 82, 525. Forrester, J. S., and Jones, J . L., Analyt. Claem., 1960, 32, 1443. Macmillan, E., and Samuel, B. W., Ibid., 1966, 38, 250. Cheng, K. L., Talanta, 1961, 8, 658. Wilson, A. L., Analyst, 1961, 86, 72. Received Octobav Wh, 1967