Grantee Research Project Results
Final Report: Advanced Analytical Methods for the Direct Quantification and Characterization of Ambient Metal Species in Natural Waters
EPA Grant Number: R826187Title: Advanced Analytical Methods for the Direct Quantification and Characterization of Ambient Metal Species in Natural Waters
Investigators: Hering, Janet G.
Institution: California Institute of Technology
EPA Project Officer: Aja, Hayley
Project Period: January 1, 1998 through June 30, 1999 (Extended to March 31, 2000)
Project Amount: $116,985
RFA: Exploratory Research - Environmental Chemistry (1997) RFA Text | Recipients Lists
Research Category: Water , Land and Waste Management , Air , Safer Chemicals
Objective:
The objective of this research project was to develop the application of electrospray mass spectrometry (ESMS) for the direct determination of metal speciation in model systems and natural water samples. The applicability of the ESMS method was examined in model systems (i.e., well-defined mixtures of metals and complexing agents of known metal affinities). Competitive effects (both kinetic and equilibrium effects) in multi-metal systems were investigated. Limitations of the ESMS method for application to whole water samples were determined.
Summary/Accomplishments (Outputs/Outcomes):
Advantages and Prior Applications of ESMS
In ESMS, nebulization of liquid samples in a voltage gradient generates charged droplets. This process is compatible with non-volatilizable analytes. Common solvents for ESMS are aqueous methanol and acetonitrile. The electrospray ionization is less aggressive than previous ionization methods, such as fast atom bombardment (FAB) or matrix-assisted laser desorption (MALDI), and thus allows compounds to be analyzed with minimal fragmentation. Either positive or negative ions can be generated by controlling the voltage gradient. The droplets then are desolvated by adiabatic cooling in a free jet expansion, which delivers the ions through a capillary to a series of differential pumping stages, and ultimately, to the high vacuum inlet of the mass spectrometer (MS). The mechanism of formation of gas phase ions is not well understood, but occurs as the charged droplets shrink through either or both the evaporation of solvent from the droplets, and fissioning of the droplets.
Prior ESMS studies of ethylenediaminetetraacetic acid (EDTA) and its metal complexes were conducted in our laboratory using a Hewlett-Packard (HP) 59,987A electrospray, coupled with an HP 5,989B quadrupole mass spectrometer. Both uncomplexed EDTA and metal-EDTA complexes were detected in the positive ion mode as protonated species with a single positive charge. Protonation (and, in the case of the uncomplexed EDTA, formation of the Na adduct) was the only perturbation of initial metal and ligand speciation. Molecular ions were detected for EDTA, and its complexes with Cu, Pb, Cd, Al, and Fe(III). The characteristic isotopic signature of Pb was detected for the Pb-EDTA complex.
Experimental Approaches in This Study
Early in this study, the HP 59,987A ES-5,989B MS became unavailable for further use. Therefore, further studies were conducted primarily using the replacement instrument, an HP series 1,100 LC/MSD. This instrument allows, in addition to the electrospray positive (ES+) and electrospray negative (ES-) modes of ionization, the use of atmospheric pressure chemical ionization (APCI) in both positive and negative modes (i.e., generating positive and negative ions, respectively). In the APCI process, samples are evaporated prior to ionization of the solvent (and, by charge transfer, the analytes) in the gas phase by corona discharge; thus the application of APCI may be limited by sample volatility. Some additional studies were performed with a Finnigan MAT LCQ, a quadrupole ion trap mass spectrometer with MS/MS capability.
Studies were performed with a series of model ligands that varied in their hydrophobicity, affinity for metals, and stoichiometry of their metal complexes. The model ligands were: EDTA, nitrilotriacetic acid (NTA), 8-hydroxyquinoline-5-sulfonate (HQS), 8-hydroxyquinoline (HQ), diethyldithiocarbamate (DETC), and 1,10-phenanthroline (Phen).
A preliminary survey of the model ligands was conducted using different ionization modes (ES± and APCI±). Operating parameters and conditions (e.g., solvent composition) then were optimized for individual ionization modes. The concentration range for quantification of the ligands and their metal complexes then was determined. Performance of the HP 1,100 and Finnigan LCQ were compared.
Observations With Model Systems
Detection of the model ligands significantly varied with the different ionization methods. Response in the APCI- mode was poor for all compounds. DETC was not detected using any ionization method. All other ligands were detected in the ES+ mode, with variable results in other modes. In the ES+ mode, signals for both the protonated ligands and their Na adducts were observed at characteristic mass-to-charge ratios (m/z) in the mass spectrum. The sensitivity for a given compound with the different ionization methods could be rationalized based on analyte structure. For example, strong signals were obtained for Phen in both ES+ and APCI+, but the compound was not detected in the ES- mode, whereas EDTA gave a stronger signal in the ES- rather than the ES+ mode. Instrumental response with any given ionization method was dependent on the specific compound tested and, therefore, quantifiable only for a specific analyte.
Effects of operating parameters and conditions were examined to determine optimum conditions. The HP 1,100, fragmentor voltage (applied between the end of the capillary and the skimmer) was found to affect analyte detection. A setting of 30 V was selected to avoid decreased transfer of ions through the skimmer region and into the detector at lower voltages, while increasing analyte fragmentation at higher voltages. Addition of 0.3 percent v/v acetic acid to the aqueous methanol spray solution significantly improved sensitivity in ES+ mode for HQ, EDTA, NTA, and Phen. Problems of loss of analyte and cross-contamination between samples occurred using the HP 1,000 autosampler, but were avoided by using a dedicated syringe pump for sample delivery. The instrument signal was integrated during this continuous sample delivery to reduce signal variability. The observed sensitivity and linearity of the HP 1,100 response were quite different in the scan mode (where data are acquired over a range in m/z) and single ion mode (SIM, which targets specific m/z values). Quantitation required use of SIM detection.
Preliminary characterization of metal-ligand complexes was performed using the HP 1,100 in ES+ mode with Cu(II) at a concentration half that of the ligand (L) concentration ([L]T ≈ 10-4.3 M). Signals were detected for 1:1 complexes of Cu with EDTA (m/z 354) and NTA (m/z 313), 1:2 complexes with HQ (m/z 374) and both 1:1 and 1:2 complexes with Phen (m/z 302 and 485). With the Finnigan LCQ, both 1:1 and 1:2 complexes were observed for HQ, indicating that the ambient speciation in the spray solution is altered in the electrospray interface. Although multiple peaks appeared in these mass spectra, peaks corresponding to Cu complexes could be identified by the isotopic signature of Cu-63 (69 percent abundance) and Cu-65 (31 percent abundance). Observation of this characteristic ratio at an m/z separation of 2 thus indicates a Cu complex. With HQS in the presence of Cu, the mass spectrum was indicative of substantial fragmentation, and no major attributable peaks could be identified.
Using the HP 1,100 in ES+ mode (with SIM detection), ESMS response was calibrated against concentration for both free ligands and metal-ligand complexes in the range of 10-6 to 10-5 M. For metal-ligand complexes, both the metal and ligand concentrations were varied, keeping a fixed 1:2 ratio of the metal-to-ligand. Excess ligand was maintained to prevent precipitation of the metal in the sample injector or ES interface. Both Cu and Pb complexes were quantified for EDTA and NTA, but Pb complexes were not detected for either HQ or Phen.
Competition experiments were performed with binary ligand (EDTA and NTA) and binary metal (Cu and Pb) systems. The distribution of species detected corresponded to the expected speciation based on the relative affinities of the ligands for the metal(s). For the conditions studied, no kinetic effects were observed (i.e., the order of addition or time of preequilibration had no effect on the observed spectra). However, even in such relatively simple systems, the appearance of multiple peaks in the mass spectra (corresponding to various adducts, complexes, and, in some cases, fragments) complicates interpretation. Interpretation is feasible when the parent molecular ion for the ligand is known (or identifiable in the mass spectrum) but may be difficult for unknown or ill-characterized analytes.
Limitations on Applications of ESMS to Whole Water Samples
The studies conducted with model ligands clearly indicated that application of ESMS to natural ligands and their metal complexes were extremely problematic. Natural organic matter (NOM) has known metal-binding properties, but is a heterogeneous and ill-defined mixture of partly degraded and/or polymerized biogenic compounds. ESMS studies of humic and fulvic have produced very complex mass spectra with evidence of aggregation in some solvents. In studies where metal complexation in natural waters has been interpreted as a single ligand (type), the strong ligand concentrations only are slightly in excess of the metal concentrations (i.e., in the nanomolar range). Copper complexes of tannins, a component of NOM, have been detected by ESMS only at Cu concentrations of about 10-4 M. Even the micromolar concentration range examined in the present study is too high for uncontaminated natural waters. The combination of low concentrations and complex composition poses severe challenges for ESMS.
A significant issue in quantitative applications of ESMS is the observed variation in response for different model ligands (or even for different adducts or complexes of the same ligand). Thus, each signal (obtained for at a specific m/z) must be individually calibrated. Again, while this is feasible in simple systems, the application to more complex (and more dilute) natural samples is questionable.
There are a number of difficulties with ESMS that have been identified by other investigators and are of particular concern for the possible application of ESMS to whole water samples. These include: matrix effects and sensitivity of instrument response to conditions in the electrospray interface, occurrence of multiple analyte peaks, changes in sample pH and metal oxidation state because of electrolytic processes occurring at the charged capillary, cross-contamination of samples, and metal contamination derived from stainless steel capillaries. While good agreement between observed and predicted metal speciation has been found for transition metal complexes with strong ligands, observed discrepancies suggested that changes in speciation can occur in the electrospray interface.
Future Prospects for ESMS as Applied to Ambient Metal Species
The use of ESMS for quantitation of ambient metal species severely is compromised by the lack of sensitivity and the inherent variability in response. It is more likely that ESMS could be used in a qualitative manner to assess variations between (or "fingerprint") natural samples rather than quantify individual components in such samples.
For such applications, some strategies may be adopted to minimize some of the problems to which ESMS is subject. With metal complexes as the primary target analytes, metal contamination can be minimized by substituting other materials for the stainless steel in the capillary needle; both glass and Al-coated fused silica have been used for this purpose. In general, cross-contamination can be minimized by using dedicated sample delivery systems (e.g., the syringe pump used in this study, rather than an autosampler) and by careful oversight of multiple instrument users. The use of preconcentration or separation techniques can minimize matrix effects by removing incompatible co-solutes and increasing the concentration of the target analyte(s); the compatibility of liquid chromatography with ESMS offers a range of possible avenues for such improvements.
References:
Baron D, Hering JG. Journal of Environmental Quality 1998;27:844-850.
Brown TL, Rice JA. Analytical Chemistry 2000;72:384-390.
Colton R, D'Agostino A, Traeger JC. Journal of Mass Spectrometry 1995;14:79-106.
Constantopoulos TL, Jackson GS, Enke CG. Analytical Chemistry Acta 2000;406:37-52.
Fievre A, Solouki T, Marshall AG, Cooper WT. Energy Fuels 1997;11:554-560.
Henderson W, Nicholson BK, McCaffrey LJ. Polyhedron 1998;17:4,291-4,313.
Kebarle P, Tang L. Analytical Chemistry 1993;65:972A-986A.
Lloyd T. Ph.D. Dissertation. California Institute of Technology, Environmental Engineering Science Department, Pasadena, CA, 1999.
McIntyre C, Batts BD, Jardine DR. Journal of Mass Spectrometry 1997;32:328-330.
Mollah S, Pris AD, Johnson SK, Gwizdala AB, Houk RS. Analytical Chemistry 2000;72:985-991.
Morel FMM, Hering JG. Principles and Applications of Aquatic Chemistry, Wiley-Interscience, New York, NY, 1993.
Olesik JW, Thaxton KK, Olesik SV. Journal of Analytic Atomic Spectrometry 1997;12:507-515.
Persson L, Alsberg T, Kiss G, Odham G. Rapid Commercial Mass Spectrometry 2000;14:286-292.
Ross ARS, Ikonomou MG, Orians KJ. Analytical Chemistry Acta 2000;411:91-102.
Ross ARS, Ikonomou MG, Thompson JAJ, Orians KJ. Analytical Chemistry 1998;70:2225-2235.
Shahgholi M, Callahan JH, Rappoli BJ, Rowley DA. Journal of Mass Spectrometry 1997;32:1080-1093.
Solouki T, Freitas MA, Alomary A. Analytical Chemistry 1999;71:4719-4726.
Stewart II. Spectrochim. Acta Pt. B-Atomic Spectroscopy 1999;54:1649-1695.
Van Berkel GJ, Zhou FM, Aronson JT. International Journal of Mass Spectrocopy Ion Proceedings 1997;162:55-67.
Voyksner RD. Environmental Science and Technology 1994;28:118A-127A.
Wang HJ, Agnes GR. Analytical Chemistry 1999;71:4166-4172.
Journal Articles:
No journal articles submitted with this report: View all 1 publications for this projectSupplemental Keywords:
environmental chemistry, heavy metals, water quality., Scientific Discipline, Air, Ecology, Environmental Chemistry, Engineering, Chemistry, & Physics, metal affinities, natural waters, electrospray mass spectrometry, equilibrium speciation modeling, metal speciation, chemical kinetics, quantification, water quality, physicochemical, ambient metal speciesProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.