Anal. Chem. 2010, 82, 5663–5668
Preconcentration of Trivalent Lanthanide Elements on a Mercury Film from Aqueous Solution Using Rotating Disk Electrode Voltammetry Paul D. Schumacher, Nicholas A. Woods, James O. Schenk, and Sue B. Clark* Department of Chemistry, Washington State University, Pullman, Washington 99164 An approach to concentrate trivalent lanthanide elements into or onto mercury film electrodes supported on rotating disk glassy carbon electrodes in small volumes (e1000 µL) is described. La3+ was used as a model for the trivalent f-element cations because its standard potential is the most negative. La3+ cathodically sorbed onto or into a preformed Hg film electrode from LiCl. After initial characterizations, Nd3+, Eu3+, Gd3+, Ho3+, and Lu3+ were tested individually and in a mixture. ICPMS analyses demonstrated that these elements could be accumulated cathodically from LiCl solution and stripped anodically in 2% HNO3, both individually and as a mixture. These cations can be electrochemically preconcentrated onto and stripped from the electrode in 5 min, a vast improvement over traditional approaches such as evaporation that require hours to days in some cases. Determination of lanthanide and actinide elemental and isotopic signatures is necessary for problems such as geochemical taphonomy;1 geochemical provenance;2-4 environmental transport pathways;5,6 nuclear forensic analysis;7-9 human health, nutrition, and bioassay,10-14 materials process monitoring;15-17 and nuclear reactor fuel burn-up,18,19 to name a few. These methods involve f-element preconcentration, if necessary, via methods such as evaporation, adsorption, or extraction, followed by their separation using chromatography,20-23 solid phase extraction,24,25 or electro* Corresponding author. Phone: 1-509-335-1411. Fax: 1-509-335-8867. E-mail:
[email protected]. (1) Iliopoulis, G.; Galanidou, N.; Pergantis, N. A.; Vamvakaki, V.; Chaniotakis, N. J. Archaeol. Sci. 2010, 37, 116–123. (2) Singh, P. Chem. Geol. 2009, 266, 251–264. (3) Song, Y. H.; Choi, M. S. Chem. Geol. 2009, 266, 337–351. (4) Maryutina, T. A.; Soin, A. V. Anal. Chem. 2009, 81, 5896–5901. (5) Novikov, A. P.; Kalmykov, S. N.; Utsunomiya, S.; Ewing, R. C.; Horreard, F.; Merkulov, A.; Clark, S. B.; Tkachev, V. V.; Myasoedov, B. F. Science 2006, 314, 638–641. (6) Kersting, A. B.; Efurd, D. W.; Finnegan, D. L.; Rokop, D. J.; Smith, D. K.; Thompson, J. L. Nature 1999, 397, 56–59. (7) Varga, Z.; Wallenius, M.; Mayer, K.; Keegan, E.; Millett, S. Anal. Chem. 2009, 81, 8327–8334. (8) Glasner, A. Nucl. Sci. Eng. 2009, 163, 26–33. (9) Grant, P. M.; Moody, K. J.; Hutcheon, I. D.; Phinney, D. L.; Haas, J. S.; Volpe, A. M.; Oldani, J. J.; Whipple, R. E.; Stoyer, N.; Alcaraz, A.; Andrews, J. E.; Russo, R. E.; Klunder, G. L.; Andresen, B. D.; Cantlin, S. J. Forensic Sci. 1998, 43, 680–688. (10) Modna, D. K.; Jerome, S. M.; White, M. A.; Woods, M. J. Appl. Radiat. Isot. 2000, 53, 265–271. 10.1021/ac101180w 2010 American Chemical Society Published on Web 06/09/2010
chemistry,26,27 and then detection by a variety of methods.24,28-31 These methods are similar to approaches used for the determination of transition metal cations in environmental samples.32-34 Although quite sensitive, accurate, and reproducible, methods for f-element isotopic signatures tend to be laborious, and reporting of final results is often slower than desired for emergency response situations.35 Recently, descriptions of methods providing rapid determination of metal cations in various sample matrixes have begun to appear, most of which have focused on speeding the separation steps.12,36,37 In particular, rapid solid phase extraction,36 ion chromatography,20,21 and capillary electrophoresis (CE)12,37–39 are attractive alternatives to traditional approaches because of the (11) Ngai, V.; Wimmer, M. A.; Kunze, J. Wear 2009, 267, 679–682, Part 1, Special Issue. (12) Chailapakul, O.; Korsrisakul, S.; Siangproh, W.; Grudpan, K. Talanta 2008, 74, 683–689. (13) Ulusoy, U.; Whitney, J. E. Br. J. Nutr. 2000, 84, 605–617. (14) Bandura, D. R.; Baranov, V. I.; Ornatsky, O. I.; Antonov, A. Anal. Chem. 2009, 81, 6813–6822. (15) Belova, V. V.; Voshkin, A. A.; Kholkin, A. I.; Payrtman, A. K. Hydrometallurgy 2009, 97, 198–203. (16) Zigalo, I.; Pavlenko, Y. Mater. Sci. Forum 1996, 215, 423–428. (17) Hayashibe, Y. Bunseki Kagaku 1996, 45, 971–972. (18) Portier, S.; Bremier, S.; Hasnaoui, R.; Bildstein, O.; Walker, C. T. Microchim. Acta 2008, 161, 479–483. (19) Sivaraman, N.; Subramaniam, S.; Srinivasan, T. G.; Rao, P. R. V. J. Radioanal. Nucl. Chem. 2002, 253, 35–40. (20) Verma, S. P.; Santoyo, E. Geostand. Geoanal. Res. 2007, 31, 161–184. (21) Nesterenko, P. N.; Jones, P. J. Sep. Sci. 2007, 30, 1773–1793. (22) Janos, P. Electrophoresis 2003, 24, 1982–1992. (23) Nash, K. L.; Jensen, M. P. Sep. Sci. Technol. 2001, 36, 1257–1282. (24) Qiao, J. X.; Hou, X. L.; Miro, M.; Roos, P. Anal. Chim. Acta 2009, 652, 66–84. (25) Jassin, L. E. J. Radioanal. Nucl. Chem. 2005, 263, 93–96. (26) Liezers, M.; Lehn, S. A.; Olsen, K. B.; Farmer, O. T.; Duckworth, D. C. J. Radioanal. Nucl. Chem. 2009, 282, 299–304. (27) Clark, W. J.; Park, S. H.; Bostick, D. A.; Duckworth, D. C.; van Berkel, G. J. Anal. Chem. 2006, 78, 8535–8542. (28) Pourjavid, M. R.; Norouzi, P.; Ganjali, M. R. Int. J. Electrochem. Sci. 2009, 4, 923–942. (29) Feng, X. G.; He, Q. C. Nucl. Instrum. Methods Phys. Res. A 2009, 609, 165–171. (30) Vajda, N.; Torvenya, A.; Kis-Benedek, G.; Kim, C. K. Radiochim. Acta 2009, 97, 9–16. (31) Kim, G.; Burnett, W. C.; Horwitz, E. P. Anal. Chem. 2000, 72, 4882–4887. (32) Puls, C.; Limbeck, A. J. Anal. At. Spectrom. 2009, 24, 1434–1440. (33) Chen, J. B.; Louvat, P.; Gaillardet, J. Chem. Geol. 2009, 259, 120–130. (34) Jiann, K. T.; Presley, B. J. Anal. Chem. 2002, 74, 4716–4724. (35) May, M.; Davis, J.; Jeanloz, R. Nature 2006, 443, 907–908. (36) Qiao, J. X.; Hou, X. L.; Roos, P.; Miro, M. Anal. Chem. 2009, 81, 8185– 8192. (37) Clark, S. B.; Friese, J. I. J. Radiat. Nucl. Chem. 2009, 282, 329–333. (38) Xu, Y. H.; Wang, E. K. J. Chromatogr., A 2009, 1216, 4817–4823. (39) Coufal, P.; Pacakova, V.; Stulik, K. Electrophoresis 2007, 28, 3379–3389.
Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
5663
faster separations they often provide and the opportunity to automate the separation steps. For example, we have demonstrated complete separation of all the rare earth elements (REEs) in less than 10 min using CE.37 Except for radiometric methods that benefit from increased sensitivity offered by integrating total counts as a function of time, detection is rarely the rate-limiting step in the overall analysis. Recent advances in the detection of metal cations, and in particular, the f-elements, suggest that new approaches may provide both increased sensitivity and the opportunity for miniaturization. Chailapakul et al.12 have demonstrated that amperometry coupled with CE on a microchip provides direct, rapid, and reproducible detection of heavy metal ions in food matrixes. Pourjavid et al.28 used ion exchange chromatography coupled with fast Fourier transform continuous cyclic voltammetry to separate and detect the light REEs. Using a scan rate of 30 V/s, they achieved a linear dynamic range between 250 and 21 000 ppb and a detection limit of 90 ppb. Electrochemical detection of metal cations is also described by Coufal et al.,39 and REEs have been applied to develop rapid electrochemical detection of common pharmaceuticals via oxidation when the REEs are incorporated into nanowire modified carbon paste electrodes.40 In many cases, though, preconcentration of the analyte of interest is required, and this is often the slowest step in the overall method, especially when compared to evaporation. Preconcentration of metal cations in geological samples (e.g., groundwater, seawater, soils, and rocks) has been demonstrated using various types of extraction or sorption approaches. Liquid-liquid extraction was employed by Chang et al.41 to preconcentrate hexavalent U from seawater using an ionic liquid and dimethylphenylazosalicylfluorone. Chen et al.42 used a combination of Chelex-100 and AG 1-X4 resins to preconcentrate Zn isotopes from samples of river water and rainwater, similar to the earlier work of Jiann and Presley.34 Waqar et al.43 developed a chelating resin based on fluorinated β-diketone extractants to preconcentrate REEs from seawater, whereas Jia et al.44 used a strongly basic cinnamene anion exchange resin to preconcentrate REEs from high-purity oxide matrixes. Adsorption onto nanometer-sized TiO2 particles was used by Hang et al.45 to preconcentrate REEs from a variety of geologic samples. Although these preconcentration methods are more rapid than evaporation, the required sample processing time for a rapid response is still long (e.g., hours), and they do not provide the necessary volume reduction to allow for interfacing with microvolumetric techniques, such as CE. As an alternative to the approaches summarized, we have explored using electroanalytical chemistry for preconcentration of the trivalent f-element cations prior to separation and detection. Although electrochemical detection cannot provide isotopic information, it can be miniaturized to as small as the micrometer scale for use in micro to nanoliter volumes to provide rapid (40) Norouzi, P.; Dousty, F.; Ganjali, M. R.; Daneshgar, P. Int. J. Electrochem. Sci. 2009, 4, 1373–1386. (41) Chang, J.; Li, Z. J.; Li, M. Int. J. Environ. Anal. Chem. 2008, 88, 583–590. (42) Chen, J. B.; Louvat, P.; Gaillardet, J.; Birck, J. L. Chem. Geol. 2009, 259, 120–130. (43) Waqar, F.; Jan, S.; Mohammad, B.; Hakim, M.; Alam, S.; Yawar, W. J. Chin. Chem. Soc. 2009, 56, 335–340. (44) Jia, Q.; Kong, X. F.; Zhou, W. H.; Bi, L. H. Microchem. J. 2008, 89, 82–87. (45) Hang, Y. P.; Qin, Y. C.; Shen, J. J. Sep. Sci. 2003, 26, 957–960.
5664
Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
preconcentration. Liezers et al.26 recently demonstrated that very low levels of Pu could be concentrated from a 2% HNO3 solution onto an anodized glassy carbon electrode (GCE), thereby achieving large volumetric concentration enhancements. Carbon nanotubes have been used to preconcentrate Eu(III), followed by detection via adsorption stripping voltammetry.46 Using a Nafion composite film-modified GCE, the Eu(III) was first adsorbed onto the carbon nanotube and then reduced to Eu(II) during a potential sweep from -0.1 to -1.0 V, providing detection that is selective for Eu over other dissolved inorganic ions and the other REEs. For application to nuclear forensic analysis, methods that significantly speed the processing of samples containing f-element isotopes are desired in preparation for analysis by capillary-based separations and mass spectrometry. As follow-on to our studies of rapid separation of the REEs by CE,37 we describe here our initial studies on the rapid preconcentration of the REEs using a mercury film electrode (MFE) while also providing significant volume reduction. A macro-scale MFE on a glassy carbon rotating disk electrode (GCRDE) support was used in preparation for transitioning to a MFE on glassy carbon microelectrode support. In this work, a strong signal associated with cathodic electrodeposition of the REE cations into or onto a Hg film is observed. Once cathodically deposited, the REE cations can be quantitatively removed by anodic stripping. EXPERIMENTAL SECTION Reagents. The reagents Hg(NO3)2, LiCl, and the nitrate salts of lanthanum, neodymium, europium, gadolinium, holmium, and lutetium used in the study were reagent grade, purchased from Fisher Scientific, (Waltham, MA, www.fishersci.com) and used as received. Procedures. Electrochemical Procedures. Electrochemical measurements were made on a model 100B Bioanalytical Systems, (West Lafayette, IN, www.basinc.com) potentiostat. The waterjacketed glass 1.5 mL electrochemical cell was constructed and set up similarly to Earles and Schenk.47 The working electrode was a 3 mm GCRDE, rotated at 2000 rpm (Pine Instruments, Grove City, PA, www.pineinst.com). The reference and auxiliary electrodes were Ag/AgCl and Pt wire, respectively. Solutions within the electrochemical cell were purged with purified Ar prior to conducting an experiment. The electrochemical procedures to make the MFE were similar to those of Florence,48 with modifications. In brief, a 1 mM solution of Hg2+ was placed in the electrochemical cell, background electrolyte was 0.1 M LiCl, and a deposition potential was applied via linear sweep voltammetry (LSV) (0.0 to -2.65 V) to form the mercury film on the GCRDE. The analyte was then added, and the LSV was restarted with the same parameters used for making the mercury film. For stripping the analyte, the GCRDE with the electrodeposited mercury film and adsorbed analyte was removed, and the cell was emptied and cleaned. A 2% HNO3 solution was added to the cell along with the GCRDE, and an oxidation potential was applied via LSV (0.0 to +2.65 V). For all electrochemical (46) Yuan, S. A.; He, Q. O.; Yao, S. J.; Hu, S. S. Anal. Lett. 2006, 39, 373–385. (47) Earles, C.; Schenk, J. O. Anal. Biochem. 1998, 264, 191. (48) Florence, T. M. J. Electroanal. Chem. 1970, 27, 273.
Figure 1. Outline for the procedures for electrodepositing Hg2+ and lanthanides to a GCRDE and collecting the ICPMS samples. Sample A was the total moles of analyte in solution, taken before any potential is applied. Sample B was moles remaining after deposition of the MFE and analyte. Sample C was moles on the electrode. Samples B and C added together should be equivalent to Sample A if the procedure was quantitative.
measurements, the scan rate for LSV experiments was 100 mV/ s. A typical experiment involving formation of the MFE, preconcentration of the trivalent f-element and oxidation into 2% HNO3 required 5 min. The cell and electrode were cleaned between experiments following the procedures of Earles and Schenk.47 ICPMS Procedures. For ICPMS measurements, aliquots removed from the electrochemical cell after applying an oxidation potential were diluted in a 2% HNO3 solution for subsequent analysis on a Hewlett-Packard 4500 ICPMS utilizing an internal indium standard and scanned in the positive mode. Prior to analyzing any samples, the ICPMS was calibrated with a set of prepared standards in 2% HNO3, and the same 2% HNO3 was used as the blank to correct for background. Analyses of electroanalytical and ICPMS data were performed using GraphPad Prism version 5.02 for Windows, (GraphPad Software, San Diego California USA, www.graphpad.com). To ensure that the preconcentration of lanthanides onto and subsequent oxidation off the GCRDE was quantitative, a series of measurements were made at different points in the procedures as depicted in Figure 1. Aliquots, identified as samples A, B, and C in Figure 1, were removed and analyzed by ICPMS, n ) 3 for each sample. Sample A reflected total moles of analyte in solution prior to any potential application. Samples B and C represent samples obtained during electrochemical sampling. Sample B reflects total moles remaining in solution after MFE formation and electrodeposition of the analyte of interest. The expectation
Figure 2. Panel A is a composite LSV scan for the deposition and oxidation of Hg2+ on a GCRDE. The composite was made from two independent experiments (the first one was a scan from 0.0 to -2.60 V, and the other was a scan from 0.0 to 2.60 V). For cathodic deposition, the background electrolyte was 0.1 M LiCl, and anodic oxidation was in 2% HNO3. Panel B is 10 averaged LSV scans with the errors ((SD) within the dimensions of the line starting at 0.0 V from independent experiments in 1 mM Hg2+. The 10 scans were randomly selected over a 5 month period. The inset is a comparison of moles accumulated on the electrode as analyzed by ICPMS and by integration of the charge accumulated (shaded area in the Figure) from Hg voltammograms.
was that sample B contained fewer moles of analyte because an unknown amount was electrodeposited onto the MFE. The MFE was first removed from the cell, and sample B was taken. Sample C was removed after the mercury film had been oxidized off the MFE. The samples were then analyzed by ICPMS to determine the amount of analyte in each sample. RESULTS AND DISCUSSION Figure 2A shows a composite voltammogram for cathodic deposition and anodic oxidation of Hg2+ onto and off a GCRDE. For cathodic deposition, Hg accumulation began around -400 mV and maintained a relatively constant current until -2000 mV, when H2O reduction began to dominate the signal. For anodic oxidation, the film remained adsorbed to the surface until +1500 mV was applied and the film was stripped off. Figure 2B shows only the cathodic deposition of Hg2+. Formation of the film was remarkably consistent; if the initial Hg2+ deposition scan did not follow the expected current-potential Analytical Chemistry, Vol. 82, No. 13, July 1, 2010
5665
voltammogram, the experiment was stopped, the electrode and cell were cleaned, and the experiment was restarted. The voltammetric behavior then returned to that shown in Figure 2B, and
