Michael Leist Varian
Analytical Instruments Mulgrave, Victoria 3170, Australia E-mail: firstname.lastname@example.org
As toxicity, mobility, and bioavailability can differ greatly between the various chemical species in which an element occurs, reporting only the total concentrations can often be misleading. Arsenic is one such example where the various species differ; the inorganic trivalent form (As III) is the most toxic, followed by the inorganic pentavalent form (As V). Other common forms of arsenic include monomethyl arsenic (MMA), dimethyl arsenic (DMA) and arsenobetaine (AsB), which have signifi cantly reduced toxicities. Therefore, separation and detection of these species can greatly assist risk-based toxicity assessments.
When Liquid Chromatography (LC) is interfaced with Inductively Coupled Plasma Mass Spectrometry (ICP-MS), species elute one by one from the LC column directly to the ICP-MS for detection by elemental speciation. The coupling of LC to ICP-MS is a straight forward task; the LC column is connected to the nebulizer of the ICP-MS by a piece of PEEK tubing, and no hardware changes are required to either the LC or the ICP-MS. Coupling an LC to a Varian ICP-MS has the added advantage of offering high sensitivity, with over 80% of the analyte ions passing through the skimmer cone being transferred to the quadrupole1. The Varian ICP-MS, with its patented 90 degree ion mirror2, is the worlds fi rst ICP-MS with tunable gigahertz sensitivity.
The following experiment shows the sensitivity and detection capabilities of the high sensitivity Varian ICP-MS for element speciation.
A schematic of the overall system is shown in Figure 1.
The LC system used was a Varian ProStar 230 Tertiary Solvent Delivery Module, a Rheodyne injector with a 50 μL sample loop, and a Hamilton PRP-X100 (4.1 mm x 250 mm, 10 μm) anion exchange column.
The Varian ICP-MS was used for all arsenic measurements. ICP-MS setup was fully controlled by Varians ICP-MS Expert software, which provides one-step instrument setup, optimization and method development. Prior to connecting the ICP-MS to LC, it was tuned to high sensitivity mode automatically by using the auto optimization routine included in the software. The ICP-MS nebulizer was then connected to the LC column using a 20 cm length of PEEK tubing (0.010 I.D.). See Tables 1-3 for the ICP-MS and LC conditions used.
Reagents and samples
Deionized water (18 MΩ cm, Millipore MilliQ, Billerica, MA, USA) was used for all solution preparations (calibration solutions and mobile phase).
Arsenic pentoxide and arsenic trioxide were obtained from SPEX CertiPrep (Metuchen, NJ USA). Arsenobetaine (AsB, C5H11AsO2) was obtained from Sigma Aldrich (Castle Hill, NSW Australia). Cacodylic acid (dimethylarsenic acid, DMA) and monosodium acid methane arsonate (MMA) were purchased from Chem Service (West Chester, PA USA). All calibration solutions were prepared daily.
Mobile phase LC
Ammonium carbonate (Merck, Kilsyth, Victoria, Australia) and either Suprapur ammonia solution obtained from Merck (Kilsyth, Victoria, Australia) or AR Select Plus nitric acid, (Mallinckrodt Baker, Phillipsburg, NJ, USA) were used for pH adjustments of the mobile phase. The mobile phase was prepared daily.
Standard Reference Material 1640, Trace Elements in Natural Water was obtained from the National Institute of Standards and Technology (Gaothersburg, MD, USA). Trace Metals in Fish (HPS CRM-TMF Lot #123921), Trace Metals in Drinking Water (HPS CRM-TMDW Lot#222518) and Oyster Tissue (HPS CRM-OT Lot #209510) were all obtained from High-Purity Standards (Charleston, SC, USA). The urine reference material (Seronorm Lot#No2525) was obtained from Sero As, (Billingstad, Norway).
Results and Discussions
The time resolved data from Varian ICP-MS Expert was exported as a .CDF fi le and opened in the Galaxie software, where program data analysis was performed. For reviewing acquired data, Galaxie provides an easy way to view single or multiple chromatograms, and compare results in a single screen.
Speciation of the fi ve inorganic and organic arsenic species can be performed using either an isocratic or gradient LC method (see Figures 2 and 3). Using either method, excellent calibrations for each of the arsenic species were obtained with calibration coeffi cients ≥ 0.999 achieved when calibrating on solutions ranging from 0.01 to 50 μg/L. An example calibration curve for MMA is shown in Figure 4. Note the low calibration solution in this calibration is 0.01 μg/L (10 ng/L).
While the gradient method increases the analysis time by approximately 9 minutes (4 minutes increased analysis time and 5 minutes column re-equilibration), it does provide better separation of As(III) from DMA, and improved separation from potential interferences. For example, samples with a high Cl content can result in an additional peak (75ArCl). As shown in Figure 5, the gradient method adequately separates the Cl peak from the neighbouring MMA and AsV peaks.
Typical detection limits
Table 4 shows the typical detection limits (DLs) for the fi ve common organic and inorganic forms of arsenic. All the measurements were made under routine analytical laboratory conditions, not clean-room conditions. This work indicates typ ical DL values that can be routinely achieved outside a clean-room in a clean laboratory.
Although previous experiments have reported signal improvements when 13% methanol is aded to the mobile phase4, this application note demonstrates that the Varian ICP-MS can produce low detection limits without this step. The addition of methanol can also cause problems if other elements such as Cr (VI) are measured, as the formation of 40Ar13C leads to substantial degradation of the signal-to-noise ratio, and higher limits of detection. Additions above 1% methanol can also cause problems, as the chromatographic separation between AsB and As(III) may be degraded to such an extent that base line resolution is not achievable.
Analysis of reference materials
Given the lack of reference materials that are certifi ed for all fi ve arsenic species, a number of reference materials were spiked with known quantities of each species and the % recovery recorded, to test for robustness of the Gradient method.
The reference materials analyzed included two water reference materials, NIST 1640 and HPS TMDW, two biological tissue reference materials, HPS OT and HPS TMF, and a urine reference material. Each reference material was spiked with 11 μg/L of each As compound and analyzed. The HPS TMF and the Seronorm urine reference material were diluted by a factor of 100 prior to injection. The percentage spike recoveries for each reference material can be viewed in Table 5. Very good recoveries were achieved, especially considering that no internal standard was used.
This work has demonstrated that the Varian ICP-MS is an excellent detector when coupled with liquid chromatography for elemental speciation. Low ng/L detection limits can be routinely obtained and very good recoveries were obtained on a variety of spiked referenc e materials.
1. S. Elliott, M. Knowles and I., Kalinitchenko, A New Direction in ICP-MS, Spectroscopy , 19(1), 30 (2004)
2. I. Kalinitchenko, Ion Optical System for a Mass Spectrometer , Australian Patent 750860, 14 November 2002, US Patent 6,614,021B1, 2 September 2003.
3. K. Ackley, C. BHymer, K. Sutton, J. Caruso, Speciation of Arsenic in Fish Tissue Using Mircowave-Assited Extraction Followed by HPLC-ICP-MS, Journal of Analytical Atomic Spectrometry , 14, 845 (1999)
4. Y. Martinez-Bravo, A.F. Roig-Navarro, F.J. Lopez, F. Hernandez, Multielemental Determination of Arsenic, Selenium and Chromium(VI) Species in Water by High-Performance Liquid Chromatography Inductively Coupled Plasma Mass Spectrometry, Journal of Chromatography A , 926, 265 (2001).