Anodic Stripping Tin Titration: A Method for the Voltammetric

Jun 9, 2014 - *Phone: +39 0312386450. Fax:+39 ... Cloud-Point Extraction for Preconcentration and Platinum Determination Using Spectrophotometry in ...
2 downloads 3 Views 425KB Size
Article pubs.acs.org/ac

Anodic Stripping Tin Titration: A Method for the Voltammetric Determination of Platinum at Trace Levels Barbara Giussani,† Simone Roncoroni,† Anna Nemenyi,† Vladimiro Dal Santo,‡ Damiano Monticelli,† and Sandro Recchia*,† †

Dipartimento di Scienza e Alta Tecnologia, Università degli studi dell’Insubria, Via Valleggio, 11-22100 Como, Italy CNRIstituto di Scienze e Tecnologie Molecolari, Via Golgi, 19-20133 Milano, Italy



ABSTRACT: We propose here a novel voltammetric method for the determination of platinum at trace levels. The method is based on the interference that platinum generates on the anodic stripping signal of tin acidic solutions: in appropriate conditions platinum uses the intermediate formation of tin(II) ions, taking place during the tin cathodic reduction, to reduce itself and to form mixed Pt(II)−Sn(II) chloro-complexes. From the analysis of the anodic stripping plots obtained after subsequent additions of tin in a Pt-containing solution, it is possible to quantify accurately and precisely the Pt concentration from 3 ppb to more than 10 ppm. This novel method is validated for the analysis of Pt in heterogeneous catalysts, but in principle could be extended to other matrixes. field of catalysts characterization. Actually, the determination of metal loading in fresh and exhaust catalysts and of metal leaching are of primary importance, whereas dilutions to the ng/L level require clean conditions which cannot be achieved in a catalysis lab that every day deals with significant amounts of platinum. The voltammetric method that we would like to present here for the first time is certainly more applicable to studies related to catalysts’ characterization as platinum can be easily determined from the ppb to the ppm level in solution. The following is a short discussion to explain the unusual and curious route that drove us to the development of this method. Some years ago some of us were involved in the study of tin-promoted platinum catalysts.17,18 During this period it was decided to develop an easy and robust voltammetric method for the determination of tin, as it was experienced that other techniques failed due to the high ionic strength of digested catalyst solutions. Surprisingly, we observed a strong interference of Pt on the appearance of the profile of tin in differential pulse voltammetry (DPV) that caused a strong and not-constant bias on the quantification of tin.19 To get rid of this interference we decided to deeply study this issue, and in the end we were able to set up an interference-free method, together with a rationale description of the nature of the interference.19 Later, we started to work around the idea that such an interference may be potentially exploited to set up a method for the indirect voltammetric determination of platinum. The first collected data told us that this intuition was correct, and therefore we decided to set up a validated

I

n the last 25 years the exponential growth of the utilization of platinum group metals (PGMs) in automotive catalytic converters has caused a widespread dispersion of such elements into various environmental compartments.1,2 Among these elements, platinum is mostly used, due to its peculiar properties toward the activation of a large variety of catalytic reactions. The increase in the environmental concentration of platinum with respect to its natural background level, together with potential concerns about its toxicity and ecotoxicity, represents an open issue which is actively investigated.3−5 From an analytical chemist’s point of view, these are probably the reasons why we assisted in the development of various methods for the determination of platinum on different samples in the last two decades: depending on the matrix and on the concentration level to be determined, several spectrometric techniques such as XRF, AAS, ICP-OES, and ICP-MS may be successfully used to investigate samples like exhausted catalysts, road dust, particulate matter, vegetation, soil, water, and biological materials.6−10 Voltammetric methods play a major role in PGM determination, as they can reach extreme sensitivities (see ref 11 and references therein): limits of detections (LOD) below 0.1 μg/L are reported for adsorptive stripping voltammetry (AdSV),12 whereas catalytic adsorptive stripping voltammetry (CAdSV) techniques show LOD ranging from 0.03 to 70 ng/ L.13−15 Nevertheless, these extremely low LOD values (close to the ones achievable with sector field ICP-MS) are partially counterbalanced by limited linear ranges, as known from the literature11 and directly experienced by some of us.16 Therefore, these voltammetric techniques (which are surely suitable to study environmental samples) reveal some limitations in the © 2014 American Chemical Society

Received: April 14, 2014 Accepted: June 9, 2014 Published: June 9, 2014 6654

dx.doi.org/10.1021/ac501349y | Anal. Chem. 2014, 86, 6654−6659

Analytical Chemistry

Article

Reference Catalyst Preparation and Dissolution. Two catalysts with a platinum content of 2% and 5% approximately (2%Pt catalyst and 5%Pt catalyst hereafter) were used as reference materials. 5%Pt is a Pt/Al2O3 catalyst with a nominal Pt loading equal to 5% wt, commercially available from BASF Italy catalysts division (sample date 21/07/95, sample code 7004 (ESCAT24), lot EM5129, type dry/reduced); 2%Pt catalyst was a Pt/MgO Pt 2% wt and was prepared according to a procedure reported elsewhere.20 Briefly, a solution of Pt(acac)2 in dry toluene (dried on activated 5 Å molecular sieves and stored under Ar) was poured on activated MgO under Ar atmosphere and stirred for 3 h; the impregnated MgO was dried in vacuo overnight; the resulting powder was reduced at 500 °C for 1 h under H2 flow to obtain metallic Pt nanoparticles dispersed on MgO. Before wet digestion, both catalysts were carefully grinded in an agate mortar, homogenized, and dried to constant weight. Both 2%Pt and 5%Pt catalysts (about 100 mg) were then accurately weighted and dissolved. The acid dissolution was carried out with an optimized procedure reported elsewhere.21 Briefly, a microwave-assisted digestion (MLS-1200 Mega, Milestone) with sealed PTFE vessels was used. The digestion was performed with 4 mL of aqua regia according to the following power program: 250 W for 2 min, 2 min thermalization step, 250 W for 2 min, 0 W for 1 min, 400 W for 2 min, 0 W for 1 min, 500 W for 5 min, 700 W for 5 min.

analytical protocol for the determination of platinum, which was named anodic stripping tin titration (ASTT). As we will show later, the focus of this novel method relies on the evidence that the intermediate formation of Sn(II) (which happens during the Sn(IV) cathodic discharge) is used to furnish an alternative way for Pt(IV) reduction, with respect to its very difficult cathodic discharge. Finally, we would like to highlight that the principle at the basis of the ASTT technique may have a wider application: the intermediate formation of Sn(II) could be used to catalyze the reduction of substances that cannot be easily reduced to the mercury drop, thus producing a signal for their indirect determination.



EXPERIMENTAL SECTION Reagent and Sample Preparation. Ultrapure water from a Millipore Milli-Q system (18 MΩ × cm resistivity, 0.05). ANOVA table was also computed and studied. As the result of the model and of the ANOVA table, the deposition potential was the only factor affecting the response, showing the highest positive regression coefficient (model equation: Sn-tc = 32.41 + 3.3*deposition time + 24.3*deposition potential − 1.8*scan rate + 1.6*deposition time*deposition potential − 0.3*deposition time*scan rate + 0.3*deposition potential*scan rate) and a p-value less than 0.05 (a = 0.05). It is important to note that highly negative deposition potentials cause a noticeable decrease of Sn-tc. This evidence may be explained suggesting that at these potentials the rate of the electrochemical tin discharge to Sn(0) increases against the rate of formation of the mixed Pt−Sn complex. Finally, the wide variability of Sn-tc at fixed Pt concentration here reported clearly indicates that the Sn-tc/Pt ratio cannot be used to guess information about the nature of the Pt−Sn chloro-complex being formed. Figures of Merit of the Proposed Method. Following the optimization of the instrumental and chemical parameters,

The experimental observation of the existence of such a linear relationship is of primary importance, as it opens the chance to use the interfered Sn ASV signal for the determination of Pt. It is in fact sufficient to collect the Sn-ASV signals after subsequent Sn additions and to plot the recorded peak height against the added tin concentration in order to determine Sn-tc. The unknown Pt concentration is then determined by interpolation on a calibration curve like the one shown in Figure 5. Since the procedure so far presented resembles a titration in which a standard tin solution is used as the titrant, we may suggest naming this technique anodic stripping tin titration (ASTT). Additionally, at sufficiently high Pt concentrations, namely above 50 μg/L, we found that the utilization of the intercept of the second segment with the X axis lead to results which are not significantly different from the ones obtained using the method proposed in Figure 4 (i.e., using the intersection between the two segments). However, for lower concentrations the utilization of this alternative method produces non-negligible effects on the calibration curve, as its intercept becomes significantly negative. Finally, it must be highlighted that the presence of tin in real samples obviously causes a bias on Pt determination. However, the presence of Sn can be easily checked via a simple DPV scan, using the conditions we reported in our previous work.19 DoE Optimization of Instrumental Parameters. All data so far presented were obtained after a careful optimization of all relevant parameters that could play a significant role in the performances of this novel method. The effect of the concentration of HCl, which is used as the supporting electrolyte, was first studied and optimized. In general higher HCl concentrations favor a steepest raise of peak heights once Sn-tc is overcome, which, in turns, allows a better estimation of Sn-tc. On the contrary, with a fixed Pt concentration, the higher the HCl concentration, the lower the value of Sn-tc: this causes a substantial decrease of the method sensitivity. A good compromise was found for an HCl concentration equal to 0.25 6657

dx.doi.org/10.1021/ac501349y | Anal. Chem. 2014, 86, 6654−6659

Analytical Chemistry



the figures of merit of the proposed method were analyzed. The limit of detection was determined through the evaluation of the standard deviation obtained over 10 independent replicated analyses of a solution containing 10 μg/L of Pt. According to the IUPAC definition (xD = 3.29 sB, assuming that the standard deviation of the 10 μg/L of Pt is a good estimation of sB),27 the estimated limit of detection is equal to 3 μg/L. A slightly lower LOD (2 μg/L) was obtained analyzing 10 blank Pt solutions and using the intercepts with the x-axis to assess blank Sn-tc values. However, the LOD value determined on the 10 μg/L Pt solutions should be preferred, because the first flat segment is totally absent on Pt blanks and uncertainties connected to the determination of this first segment are not accounted for with blank solutions. The linearity of the curve Sn-tc vs Pt concentration was tested up to 10 mg/L. In this range the response is linear, and this is probably true even for higher concentrations. Higher Pt concentrations were not tested for two reasons. First of all, if Pt concentrations higher than 10 mg/L are expected, it is preferable to dilute in order to avoid unwanted memory effects which necessarily implies extensive postanalysis cell washings. Second, a 3 orders of magnitude linear range is fully satisfactory for the proposed method. It should be noticed that the intercept value reported in Figure 5 (that represents an extended-range calibration curve) is quite high if low Pt concentrations have to be determined. In such cases it is preferable to use short-range calibrations (0− 100 μg/L, as an example) that normally show intercept values closer to zero. Validation of the Proposed Method. The proposed method was initially intended to carry out platinum determination on heterogeneous catalysts. Therefore, validation was performed on two different heterogeneous Pt catalysts, namely a 2% w/w Pt supported on magnesia (Pt/MgO) and a 5% w/w Pt supported on α-alumina (Pt/α-Al2O3). These two catalysts were chosen because of their different behavior under digestion. In fact, on Pt/α-Al2O3 platinum is simply extracted (and alumina remains substantially undissolved), while on Pt/ MgO magnesia dissolved, too. As neither Pt/α-Al2O3 nor Pt/MgO can be considered as standard reference materials (and no commercially available SRMs of this kind of materials can be purchased), we decided to perform in parallel the same determination with two other completely independent techniques such as ETAAS and ICPMS. Table 2 reports the determined Pt concentrations in the two real catalysts obtained with the proposed ASTT method and with the two control methods. It can be immediately seen that the obtained values are in good agreement between each other, meaning that accurate Pt determinations can be obtained with the proposed method.

CONCLUSION We have demonstrated here that the voltammetric determination of Pt is feasible through the interference that platinum itself generates on the electrochemical reduction of tin(IV). The ASTT method enables the determination of platinum from the μg/L up to the mg/L level. We have seen that the method here proposed is surely suitable for the voltammetric determination of platinum in heterogeneous catalysts. The wide linear range of the ASTT method could in principle allow the determination of Pt at trace levels also in other matrixes. Indeed, a careful validation is necessary to extend the applicability of this method to other matrixes: significant effects on the analytical signal due to changes in the electrolyte composition (i.e., high concentration of salts), and to the presence of complexing agents (as it happens in natural waters), is expected. The possibility to perform speciation analysis with this method should also be exploited, since it selectively determines free platinum in solution: free vs complexed species may possibly be determined, according to the reaction mechanism involved. The principle of the ASTT method (i.e., the reduction of the analyte performed by the intermediate formation of Sn(II)) may be successfully extended to other analytes that are difficult to be reduced at the mercury drop. The utilization of mercury as working electrode represents an issue that, according to the principles of green chemistry for avoiding the usage of toxic compounds, must be taken into great consideration in terms of proper disposal of wastes.



*Phone: +39 0312386450. Fax:+39 0312386449. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Financial support from the Italian Ministry of Education, University and research (MIUR) through the FIRB project “Oxides at the nanoscale: multifunctionality and applications” (RBAP115AYN) is gratefully acknowledged.



ETAAS

ICP-MS

ASTT proposed

2.10 ± 0.09 5.6 ± 0.3

2.04 ± 0.02 5.86 ± 0.08

1.92 ± 0.08 5.9 ± 0.1

REFERENCES

(1) Gebel, T. In Anthropogenic Platinum Group Element Emissions: Their Impact on Man and Environment; Zereini, F., Alt, F., Eds; Springer: Berlin, Germany, 1999. (2) Ek, K. H.; Morrison, G. M.; Rauch, S. Sci. Total Environ. 2004, 334−335, 21−38. (3) Hoppstock, K.; Sures, B. In Elements and Their Compounds in the Environment, 2nd ed.; Merian, E., Anke, M., Ihnat, M., Stoeppler, M., Eds; Wiley-VCH: Weinheim, Germany, 2004; pp 1047−1074. (4) Osterauer, R.; Faßbender, C.; Braunbeck, T.; Köhler, H.-R. Sci. Total Environ. 2011, 409, 2114−2119. (5) Colombo, C.; Monhemius, A. J.; Plant, J. A. Ecotoxicol. Environ. Saf. 2008, 71, 722−730. (6) Barefoot, R. R.; Van Loon, J. C. Talanta 1999, 49, 1−14. (7) Soro, L. J.; Oleron-Hamdous, A.; Béchet, B.; Legret, M. J. Soils Sediments 2013, 13, 569−574. (8) Barefoot, R. R. Environ. Sci. Technol. 1997, 31, 309−314. (9) Balcerzak, M. Crit. Rev. Anal. Chem. 2011, 41, 214−235. (10) Dal’nova, O. A.; Shiryaeva, O. A.; Karpov, Yu. A.; Alekseeva, T. Yu.; Shiryaev, A. A.; Kulikauskas, V. S.; Filatova, D. G. Inorg. Mater. 2010, 46, 1599−1604.

% w/w platinum loadings (mean ± s)a 2% Pt 5% Pt

AUTHOR INFORMATION

Corresponding Author

Table 2. Comparison of the Pt Concentration Determined in the Two Different Reference Catalysts by the Proposed Method and Reference Ones (ETAAS and ICP-MS) sample

Article

Data are reported as the mean and standard deviation of five independent determinations. a

6658

dx.doi.org/10.1021/ac501349y | Anal. Chem. 2014, 86, 6654−6659

Analytical Chemistry

Article

(11) Locatelli, C. Electroanalysis 2007, 19, 2167−2175. (12) Locatelli, C. Electroanalysis 2007, 19, 445−452. (13) Wang, J.; Czae, M.-Z.; Lu, J.; Vuki, M. Microchem. J. 1999, 62, 121−127. (14) Huszal, S.; Kowalska, J.; Krzeminska, M.; Golimowski, J. Electroanalysis 2005, 17, 299−304. (15) Huszal, S.; Kowalska, J.; Krzeminska, M.; Golimowski, J. Electroanalysis 2005, 17, 1841−1846. (16) Pozzi, A.; Recchia, S.; Monticelli, D.; Dossi, C.; Rampazzi, L.; Curi, P. Ann. Chim. (Rome, Italy) 2003, 93, 181−186. (17) Recchia, S.; Dossi, C.; Poli, N.; Fusi, A.; Sordelli, L.; Psaro, R. J. Catal. 1999, 184, 1−4. (18) Stievano, L.; Wagner, F. E.; Calogero, S.; Recchia, S.; Dossi, C.; Psaro, R. Stud. Surf. Sci. Catal. 2000, 130, 3903−3908. (19) Monticelli, D.; Psaro, R.; Pozzi, A.; Dossi, C.; Recchia, S. Anal. Bioanal. Chem. 2005, 383, 115−121. (20) Dossi, C.; Pozzi, A.; Recchia, S.; Fusi, A.; Psaro, R.; Dal Santo, V. J. Mol. Catal. A: Chem. 2003, 204, 465−472. (21) Recchia, S.; Monticelli, D.; Pozzi, A.; Rampazzi, L.; Dossi, C. Fresenius’ J. Anal. Chem. 2001, 369, 403−406. (22) Bard, A. J.; Parson, R.; Jordan, J. Standard potentials in aqueous solution; Dekker: New York and Basel, 1985. (23) Bishop, E.; Hitchcock, P. H. Analyst 1973, 98, 635−646. (24) Pérez-Herraz, V.; García-Gabaldón, M.; Guiñoń , J. L.; García, A. Anal. Chim. Acta 2003, 484, 243−251. (25) Young, J. F.; Gillard, R. D.; Wilkinson, G. J. Chem. Soc. 1964, 5176−5189. (26) Moodley, K. G.; Nicol, M. J. J. Chem. Soc., Dalton Trans. 1977, 3, 239−243. (27) Currie, L. A. Pure Appl. Chem. 1995, 67, 1699−1723.

6659

dx.doi.org/10.1021/ac501349y | Anal. Chem. 2014, 86, 6654−6659