Anal. Chem. 1999, 71, 652-659
Liquid-Liquid Extraction of Cadmium by Sodium Diethyldithiocarbamate from Biological Matrixes for Isotope Dilution Inductively Coupled Plasma Mass Spectrometry Richard A. Vanderpool* and Wayne T. Buckley†
Grand Forks Human Nutrition Research Center, Agricultural Research Service, United States Department of Agriculture, Grand Forks, North Dakota 58203, and Brandon Research Centre, Agriculture and Agri-Food Canada, Brandon, Manitoba, Canada R7A 5Y3
A published procedure for the liquid-liquid extraction of Cd by sodium diethyldithiocarbamate (NaDDC) was modified and tested on 12 biological matrixes of plant and animal origin for use with isotope dilution (ID) inductively coupled plasma mass spectrometry (ICPMS). The tested matrixes were reference materials, certified for Cd, and a feces in-house standard. The digested and extracted standards were analyzed for Cd stable isotopes by ICPMS and the resulting isotope ratios examined for isobaric and polyatomic interferences. Cadmium recoveries, after extraction, ranged from 73% to 20% and apparently were inversely related to Cd concentration, even in the presence of excess chelator. For each reference material, the measured isotope ratios for Cd were corrected for instrumental bias and compared to natural abundance Cd isotope ratios. Nonextracted samples had large isotope ratio deviations for all but one or two ratios. Extraction improved all the isotope ratios measured (lowered % error relative to natural abundances), but interferences were noted for a few samples. The extracted human in-house feces standard was found to have Sn signals reduced by 300-fold, but residual Sn concentrations still interfered with 116Cd, though not 112Cd or 114Cd. Thus, evaluation of the NADDC-extracted in-house fecal standard and reference materials indicate the successful removal of interferences that otherwise prevented accurate determinations of Cd by IDICPMS and show that a number of Cd isotope ratios could be accurately measured (1.5% Error)a ratio
BMP
AL
RF
TD
CL
WF
U
DWF
BL
Feces
TL
SL
114/116 113/116 112/116 111/116 110/116 108/116 106/116 113/114 112/114 111/114 110/114 108/114 106/114 112/113 111/113 110/113 108/113 106/113 111/112 110/112 108/112 106/112 110/111 108/111 106/111 108/110 106/110 106/108
-/-/-/-/-/-/-/-/9 -/-/-/O/-/O/-/-/-/-/-/9 -/9 -/-/-/9 -/-/-/-/-/-
-/-/-/-/-/-/9 -/O -/-/-/-/-/-/-/-/-/-/-/9 -/9 -/-/-/9 -/-/-/-/-/-
-/-/-/-/-/-/-/-/9 -/9 -/9 -/9 -/-/9 -/9 -/9 O/9 -/-/9 -/9 -/9 -/-/9 O/9 -/-/9 -/-/9 -/-
-/-/-/-/-/O -/-/9 -/9 -/9 -/9 -/-/9 -/9 -/9 -/9 -/-/9 -/9 -/9 -/-/9 -/9 -/-/9 -/-/-/-
-/-/-/-/-/-/-/-/9 -/-/-/-/-/-/-/-/-/-/-/9 -/9 -/-/-/9 -/-/-/-/-/-
-/9 -/9 -/9 -/9 -/-/-/9 -/9 O/9 O/-/-/-/9 O/9 -/9 -/-/-/9 -/9 -/-/-/9 -/-/9 -/-/-/-/9
-/-/-/-/-/-/-/-/9 -/9 -/9 -/-/-/O/9 O/O/-/-/O/9 O/9 -/-/O/9 -/-/-/-/O/-
-/-/-/-/-/-/-/-/9 -/9 -/9 -/9 -/9 -/-/9 -/9 O/9 -/9 -/-/9 -/9 -/9 -/9 -/9 -/9 -/-/9 -/9 -/9
-/-/-/-/-/-/-/-/9 -/9 -/9 -/9 -/9 -/9 -/9 O/9 O/9 -/9 -/9 -/9 -/9 O/9 -/9 O/9 -/9 -/9 -/9 -/9 -/9
-/-/-/-/-/-/-/-/9 -/9 -/9 -/9 -/9 -/9 -/9 O/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9 -/9
-/9 -/9 -/9 -/9 -/9 -/-/9 O/9 O/9 O/9 -/9 -/-/9 O/9 O/9 -/9 -/-/9 O/9 -/9 -/-/9 -/9 -/-/9 -/9 -/9 -/9
-/9 -/9 -/9 -/9 -/9 9 -/9 O/9 O/9 O/9 O/9 -/9 O/9 O/9 O/9 O/9 -/O/9 O/9 O/9 O/O/9 -/9 O/O/9 O/O/-
a AL, Apple Leaves; BL, Bovine Liver; BMP, Bovine Muscle Powder; CL, Citrus Leaves; DWF, Durum Wheat Flour; IHS, in-house standard; NaDDC, sodium diethyldithiocarbamate; NIST, National Institute of Standards and Technology (U.S.), RF, Rice Four; SL, Spinach Leaves; TD, Total Diet; TL, Tomato Leaves; U, Toxic Metals in Freeze-Dried Urine; WF, Wheat Flour.
After extraction, the feces IHS data had e1.5% error for all isotope ratios except the 116Cd column, which implied an incomplete extraction of Sn. Unsatisfactory isotope ratios for 116Cd were also noted after extraction for Bovine Muscle Powder, Apple Leaves, Rice Flour, and Durum Wheat Flour. Because most of the matrixes studied do not have certified Sn determinations and only a few have noncertified Sn determinations, the potential for isobaric Sn interference in most of these samples is unknown. However, considering the use of Sn in the environment, such as in coatings for steel cans,62 the fecal IHS Sn concentration probability reflects anthropogenic enrichment. To further examine the Sn contamination, both nonextracted and extracted feces IHS samples were scanned from 104 to 124 m/z (Figure 1). The nonextracted feces revealed (solid line) a spectrum dominated by a Sn isotope pattern (solid bars) and no apparent Cd isotope pattern at the scale plotted. After extraction (dotted line), a Cd isotope pattern was apparent (open bars), and the plot scale was decreased by a factor of 300, which indicates a major reduction of Sn. However, a Sn isotope pattern was still apparent in the spectrum, though significantly reduced in intensity. Though present, Sn isotopes were not influencing the ratios with 112Cd or 114Cd, but the Sn contribution was apparent in the percentage errors for 116Cd-based ratios. Increased chelator concentrations did not change this result. We have not investigated the potential for using correction routines for isobaric interferences in the analysis of Cd isotope (62) Sherlock, J. C.; Smart, G. A. Food Addit. Contam. 1984, 1, 277-282.
658 Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
ratios. Such routines correct for isobaric interferences based on analysis of nonisobaric isotopes of interfering elements. Although routinely used in conventional, quantitative ICPMS analysis, such routines may cause an unacceptable propagation of errors in IDICPMS and should be carefully tested. Considering the need to measure small differences in isotope ratios in tracer experiments, the potential for significant error from isobaric correction routines seems much greater than that in conventional analysis. The ideal solution is to remove the interfering elements. Also, relative peak intensities in Figure 1 for the nonextracted feces sample indicated noncharacteristic abundances for Sn isotopes; this suggests that a simple isobaric correction for Sn would not be successful. After extraction, the Sn isotope abundances appear normal, and a correction routine may be possible. SUMMARY AND CONCLUSIONS As might have been expected for biological matrixes, we found no consistent interference-free Cd isotope ratio in the digested, nonextracted matrixes. In fact, there were so few interferencefree ratios in most of the nonextracted standards that translation of an interference-free isotope ratio between certified standards or from one particular standard to a matrix-matched sample (e.g., NIST Wheat Flour versus an uncharacterized wheat sample) would not be possible. A previously published NaDDC extraction procedure was modified to use small sample sizes and extraction volumes and tested on a feces IHS and several NIST SRMs and RMs. Upon extraction of digests of these materials with NaDDC, multiple isotope ratios were found with percentage errors e1.5%.
Figure 1. Upper line graph: ICPMS intensities for m/z 104-124 scans of a nonextracted feces in-house standard (IHS) digest (s) and a NaDDC-extracted feces IHS digest (---). Lower bar graph: Atom percentages20 for the natural abundance isotopes of Cd (0) and Sn (9).
In matrixes with low Cd concentrations, ratios containing 106Cd and 108Cd were found to have marginal to poor percentage errors. However, detection limits for the low natural abundance isotopes may have been less than adequate to generate percentage errors e1.5% for these determinations. In many cases, 116Cd ratios also showed a large percentage error in both low- and high-Cd concentration samples, which apparently was caused by a Sn interference that was not completely eliminated by extraction. Although the Sn concentration was high in the feces IHS, NaDDC extraction reduced the concentration so that interference from Sn appeared only in 116Cd ratios and not in 112Cd or 114Cd ratios. Thus, evaluation of IHS and NaDDC-extracted NIST standards showed that NaDDC extraction successfully removed interferences that otherwise prevented accurate determinations of Cd by IDICPMS and that a number of Cd isotope ratios could be accurately measured for multiple stable isotope tracer studies in a broad range of NaDDC-extracted biological matrixes. Note: Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products that may also be suitable.
U.S. Department of Agricultural, Agricultural Research Service, Northern Plains Area, is an equal opportunity/affirmative action employer and all agency services are available without discrimination. SUPPORTING INFORMATION AVAILABLE Supplemental Tables 1-22, comprising bias-corrected isotope ratios ((SD) and percentage errors relative to natural abundance isotope ratios from both digested and digested plus NaDDCextracted NIST standards for Apple Leaves (SRM 1515), Bovine Liver (SRM 1577b), Bovine Muscle Powder (RM 8414), Citrus Leaves (SRM 1572), Durum Wheat Flour (RM 8436), Rice Flour (SRM 1568a), Spinach Leaves (SRM 1570a), Tomato Leaves (SRM 1573a), Total Diet (SRM 1548), Toxic Metals in Freeze-Dried Urine (elevated level human Urine, SRM 2670), and Wheat Flour (SRM 1567) (22 pages). See any current masthead page for access information. Received for review May 14, 1998. Accepted November 9, 1998. AC980538B
Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
659
