ANALYST, JANUARY 1992, VOL. 117 35 Catalytic-Adsorptive Stripping Voltammetric Determination of Ultratrace Levels of Molybdenum in the Presence of Organic Hydroxy Acids Joseph Wang, Jianmin Lu and Ziad Taha Department of Chemistry, New Mexico State University, Las Cruces, NM 88003, USA The adsorptive collection of the molybdenum complex with 3-methoxy-4-hydroxymandelic acid was coupled with the catalytic current of the adsorbed complex to yield an ultrasensitive voltammetric procedure for measuring picomolar levels of molybdenum. Optimum solution conditions (particularly the solution composition) were established to give a detection limit of 4 ng dm-3 (4 x 10-11 rnol dm-1) of molybdenum (following a 2 min preconcentration). The relative standard deviation (at 0.5 pg dm-3) was 3.7%.Possible interferences were investigated and the applicability to assays of tap water is illustrated. Such coupling of catalytic and adsorptive collection processes holds great promise for the development of ultratrace voltammetric procedures for other analytes. Keywords: Molybdenum determination; adsorptive stripping voltammetry; catalytic current; 3-methoxy-4-h ydrox ymandelic acid Catalytic and adsorptive processes have greatly enhanced the sensitivity of voltammetric procedures. For example, catalytic polarographic waves 100-1000-fold larger than ordinary diffu- sion-controlled currents can offer convenient pulse polaro- graphic measurements down to the nanomolar level.' Even lower detection limits (usually down to 10-1(' rnol dm-3) can be achieved through adsorptive stripping voltammetric pro- cedures (based on the interfacial accumulation of analytes').It has been shown recently that the coupling of catalytic and adsorption processes, via controlled adsorptive accumulation of a catalyst, yields remarkable sensitivity and detectability down to the picomolar (ppt) The determination of platinum and titanium, in the presence of formazone and mandelic acid, respectively, has thus been reported. This paper describes an extremely sensitive catalytic- adsorptive stripping voltammetric procedure for the determi- nation of ultratrace amounts of molybdenum. Because of the importance of molybdenum and its extremely low levels in various matrices, an ultrasensitive method is required for its determination. Voltammetry has been shown to be useful for the measurement of trace amounts of molybdenum.Catalytic currents of molybdenum (e.g., in nitrate of perchlorate media) have permitted convenient polarographic determinations of nanomolar concentrations.s.6 The adsorptive collection of molybdenum complexes with quinolin-8-ol7 or phosphates has been exploited for stripping measurements down to the 1 x 10-l0 rnol dm-? level (10 min accumulation). The catalytic- adsorptive stripping procedure reported here lowers the detectability further to the picomolar level (i.e., 4 x 10-11 rnol dm-3 with 2 min collection). The procedure is based on the catalytic response of the accumulated molybdenum complex with 3-methoxy-4-hydroxymandelic acid (VMA). Hasebe et al.9 reported o n the catalytic polarographic determi- nation of VMA in urine in the presence of molybdenum and bromate ions. By reversing this scheme and controlling the accumulation of the complex, an ultrasensitive procedure for monitoring molybdenum is obtained, the characteristics and advantages of which are discussed below.Experimental Apparatus and Reagents The equipment used to obtain the voltammogram, a PAR 264 A voltammetric analyser with a PAR 303 static mercury drop electrode, has been described in detail elsewhere. A medium-sized hanging mercury drop electrode (HMDE) with a 0.016 cm2 surface area was employed. All solutions were prepared with doubly distilled water. A 1000 ppm stock molybdenum solution (atomic absorption standard, Aldrich) was diluted as required. Stock solutions ( 5 X 10-2 rnol dm-3) of VMA and homovanillic acid (HVA) (Aldrich) were prepared in 0.1 rnol dm-3 formic acid.The blank solution consisted of 2 x 10-6 mol dm-3 VMA, 5 x 10-4 mol dm-3 KBr03, 5 X 10-3 rnol dm-3 K2S04 and 2 X 10-3 rnol dm-3 H2SO4. Drinking water samples were collected at this laboratory. Procedure The blank solution (10 ml) was pipetted into the cell and purged with nitrogen for 8 min. The preconcentration potentizl (usually 0.0 V) was applied to a fresh mercury drop while the solution was stirred. Following the accumulation period, the stirring was stopped and after 15 s the voltammo- gram was recorded by applying a differential-pulse scan (at 10 mV s-1) in the negative direction; the scan was terminated at -0.7 V. After background voltammograms had been re- corded, an aliquot of the diluted molybdenum standard solution was introduced.Throughout this operation, a stream of nitrogen was passed over the surface. All data were obtained at room temperature. Results and Discussion Fig. 1 shows cyclic voltammograms for ( a ) 1 and (b) 5 vg dm-3 (ppb) molybdenum in the presence of 2 x 10-6 rnol dm-3 VMA and HVA, respectively (and also 5 x 10-3 rnol dm-3 K2S04, 2 x 10-3 rnol dm-3 H$04 and 5 x 10-3 rnol dm-3 KBr03), obtained after (A) 0 and (B) 60 s accumulation. A cathodic peak is observed (at about -0.48 V) during the scan in the negative direction. Scanning in the reverse direction also produces a cathodic peak, indicative of a catalytic process. The response increased dramatically when an ac- cumulation period preceded the potential scan (B versus A), indicating an interfacial accumulation of the molybdenum- hydroxy acid complex.Subsequent scans resulted in smaller, but stable, cathodic peaks (not shown) that indicate desorp- tion of the complex from the surface. Higher sensitivity was observed in the presence of VMA, which was used in all subsequent work. The fact that a defined and intense (microamps) response is obtained in cyclic voltammetry for pg dm-3 concentrations indicates the remarkable sensitivity associated with the coupling of the catalytic and interfacialANALYST, JANUARY 1992, VOL. 117 36 t A - I I I I I I I 1 I 0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 PotentialN Fig. 1 Cyclic voltammograms for ( a ) 1 and (6) 5 pg dm--3 of Mo in the presence of 2 x mol dm-3 VMA and HVA, respectively.after A. 0 and B, 60 s accumulation at 0.0 V. The solutions also contained 2 X lo-' rnol dm-3 H2SO4 and 5 x 10-3 rnol dm--3 K2S0,; scan rate, 10 mV s-I t c.' 2 3 0 0.5 0.3 L 3 u ' 0.1 -0.3 -0.4 -0.5 -0.6 Poten ti a IN 0 1 2 3 4 5 6 Time/m in Fig. 2 ( a ) Differential pulse voltammograms for 0.1 yg dm-3 of Mo after preconcentration periods of A. 0; B , 30; C, 60; and D. 120 s (at 0.0 V and 400 rev min-l stirring) and (b) the resulting current versus time plot. Cell solution. 5 x 10-3 rnol dm--3 K2S04. 2 x 10-3 rnol dm--3 H2S04, 2 X mol dm-3 VMA and 5 x 10-4 mol dm--3 KBr03; scan rate, 10 mV s-l; and amplitude. 25 mV accumulation processes. Additional gains in sensitivity can be obtained in a differential-pulse stripping operation, as illus- trated below.Fig. 2(a) shows differential-pulse voltammograms for 0.1 pg dm-3 (1.05 x 10-9 rnol dm-3) molybdenum after different preconcentration periods (0-120 s, A-D). Despite the extremely low (sub-ppb) concentration and the short accumu- lation times, well defined peaks are observed. The peak I , 2 4 6 10 20 [VMA]/pmol dm-3 2 4 6 8 tH7SOal/rnmol dmP3 20 - I I I I 4 8 12 16 [KBr031/10-4 rnol dm-3 on the catalytic adsorptive stripping current of = 2 pg dm-3 and preconcentration time = 30 s); = 1 yg dm-3 and preconcentration time = 15 s); and = 5 pg dm-3 preconcentration time = 15 s). Other conditions as in Fig. 2 increases rapidly with increasing preconcentration time, indicating (again) an enhancement of the concentration of the complex on the mercury surface.For example, a 1 min accumulation yielded about a six-fold enhancement of the peak (relative to that obtained without preconcentration; compare A and C). In Fig. 2(b) the resulting current versus preconcentration time plot is shown. The rapid increase of the current observed for preconcentration times shorter than 3 min is followed by a levelling off for longer periods. A detection limit of 4 ng dm-3 (4 X 10-11 mol dm-3) can be calculated based on the signal-to-background characteristics of the response shown in Fig. 2(a), D. Such an extremely low detection limit, obtained with a 2 min accumulation, compares favourably with the 5 x 10-9 and 1 x 10-10 rnol dm-3 detection limits reported for conventional adsorptive stripping voltammetry, following 2 and 10 min preconcentration, respectively.7 3 The catalytic-adsorptive stripping response is strongly dependent on the solution conditions (Fig. 3). The peak increases with increasing VMA concentration until it levels off at 4 X 10-6 rnol dm-3 [Fig. 3(a)]. In addition to the complexing hydroxy acid, it is essential to have a bromate ion oxidant and acidic medium to produce the catalytic response (through regeneration of the MoV1-VMA species from the MoV speciesg). The largest catalytic-adsorptive stripping peakANALYST, JANUARY 1992, VOL. 117 37 & A -0.3 -0.4 -0.5 -0.6 PotentialN 2 L 1.5 u 1 .o 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Concentration (ppb) Fig. 4 (a) Catalytic-adsorptive stripping voltammograms for solu- tions of increasing Mo concentrations; A, 0; B.0.2; C, 0.4; D, 0.6; E, 0.8; F. 1.0; G, 1.25; and H, 1.5 pg dm-3; preconcentration time, 15 s. Other conditions as in Fig. 2. ( h ) Calibration graphs for A , 15 and B 30 s preconcentration times was obtained in the presence of 1.5 X mol dm-3 H?SO4 [Fig. 3(6)]. The response increased linearly with increasing bromate ion concentration up to 8 x mol dm-3 and then more slowly [Fig. 3(c)]. The optimum solution conditions (with 2 x 10-6 mol dm-3 VMA and 5 x 10-4 mol dm-3 Br03-) were selected also for minimizing the background peak (due to the bromate ion reduction) which appeared at the same potential as that of the target Mo-VMA species. The effect of the accumulation potential was evaluated over the range from 0.2 to -0.3 V (2 pg dm-3 molybdenum with 10 s preconcentration; not shown).Only slight variations in the response were observed, with 0.0 V yielding the largest peak. Some enhancement (about 40%) of the response to 0.5 pg dm-3 molybdenum was observed by stirring the solution (during the 30 s preconcentration), compared with analogous measurements with a quiescent solution. The differential- pulse stripping mode yielded slightly better signal-to-back- ground characteristics than linear scan measurements and was used throughout. Fig. 4(a) shows voltammograms obtained after increasing the molybdenum concentration in 200 (B-F) and 250 (G, H) ng dm-3 steps. Well defined peaks are obtained following the very short (15 s) preconcentration period. These measure- ments are part of a calibration experiment up to 1.8 pg dm-3. The resulting calibration graphs (at 15 and 30 s preconcentra- tion) are shown in Fig.4(h). High linearity (Y = 0.999) prevailed up to 0.8 pg dm-3, with slower current increases at higher levels (slopes of the initial linear portions = 1.83 and 1.55 pA dm-3 pg-I). The exact nature of the discontinuity at about 0.8 pg dm-3 is not clear. I t does not appear to relate solely to surface saturation, but rather to the multitude of processes involved in the catalytic-adsorption process. A prolonged series of 30 repetitive measurements of 0.5 pg dm-3 molybdenum was used to establish the reproducibility of the data (30 s accumulation). The mean peak current found was 1.02 pA, with a range of 0.91-1.06 PA and relative standard deviation of 3.7%. In addition to high sensitivity and -0.3 -0.4 -0.5 -0.6 PotentialN Fig.5 Voltammograms for a drinking water sample (A), as well as for subsequent concentration increments of 0.4 pg dm-3 of Mo (B-D): preconcentration, 30 s; and sample, 1 ml of drinking water plus 9 ml of the blank/electrolyte solution. Other conditions as in Fig. 2 precision, the catalytic-adsorptive stripping procedure offers high selectivity toward molybdenum. Ions tested at the 5 pg dm-3 level and found not to interfere with the determina- tion of 0.5 pg dm-3 molybdenum were Cd", Pb", Bill1, Cull, Ni", V", Zn", TI', Crlll, Mn" and Co" (30 s preconcentration). Surface-active materials, in contrast, could compete with the surface adsorption sites of the catalyst. For example, albumin at 1 and 2 pg dm-3 resulted in 9 and 27% depressions, respectively, of the 0.5 pg dm-3 molybdenum response (30 s preconcentration). Depending on the complexity of the sample, destruction of organic surfactants, e .g . , through ultraviolet irradiation, may therefore be required. The applicability of the method to the analysis of drinking water is demonstrated in Fig. 5 . With a short preconcentration time of 30 s and ten-fold sample dilution, the method yields a well defined molybdenum peak for the sample (A), and subsequent standard additions of 0.4 pg dm-3 (B-D). A molybdenum level of 4.1 pg dm-3 was therefore calculated for the sample. In conclusion, this study demonstrates again that the coupling of catalytic and adsorption processes can constitute the basis for an ultrasensitive voltammetric procedure. Anal- ogous catalytic systems involving other metals may be utilized for their detection at picomolar levels. The power and utility of catalytic-adsorptive stripping voltammetry will continue to expand in the near future. This work was supported by grants from SANDIA NL and Battelle PNL. 1 2 3 4 5 6 7 8 9 10 References Mairanovski, S. G., Catalytic and Kinetic Waves in Polar- ography, Plenum Press, New York, 1968. Wang. J., Am. Lab (Fairfield CT), 1985. 17, No. 5 , 41. Wang. J., Zadeii, J.. and Lin, M. S . , J. Electroanal. Chem., 1987. 237, 281. Yokoi. K . , and van den Berg, C. M. C., Anal. Chim. Acta, 1991, 245. 167. Hidalgo, J . L., Caballero. G. M., Ccla, R., and Pcrez-Busta- mente, A. P.. Talanta, 1988, 35, 301. Navratilova. Z . , and Kopanica. M.. Anal. Chim. Acta, 1991. 244, 193. van den Berg. C. M. G., Anal. Chem., 1985, 57, 1532. Fogg, A. G., and Alonso, R. M., Analyst, 1988, 113, 361. Hasebe. K., Kakizaki, T., and Yoshida, H., Anal. Chem., 1987, 59, 373. Wang. J . . Farias. P. A. M., and Mahmoud, J . S . , Anal. Chim. Acta, 1985, 172. 57. Paper 1 I039376 Received July 30, 1991 Accepted August 28, 1991