Environ. Sci. Technol. 2001, 35, 4868-4873
Determination of Arsenic Speciation in Poultry Wastes by IC-ICP-MS BRIAN P. JACKSON* AND PAUL M. BERTSCH Advanced Analytical Center for Environmental Sciences, Savannah River Ecology Laboratory, University of Georgia, Aiken, South Carolina 29802
The aromatic organoarsenic compounds 4-hydroxy 3-nitrobenzenearsenic acid (Roxarsone, ROX) or 4-aminobenzenearsenic acid (p-arsanilic acid, p-ASA) are used as feed additives in the poultry industry for disease control and enhanced feed efficiency. While federal regulations govern acceptable As concentrations in edible tissue, elevated As concentrations occur in poultry litter, which raises concerns over repeated land application of poultry litter in agriculture. As a precursor to studying the fate of these organoarsenic compounds in soils, three speciation methodologies were developed to separate ROX and p-ASA from the more common and more toxic As species arsenate, arsenite, dimethyl arsenic acid (DMA), and monomethyl arsenic acid (MMA). The six arsenic species were separated on a Dionex AS14 column using a PO4 eluant, an AS16 column using a OH- eluant, and an AS7 column using a HNO3 eluant. While all three methods provided detection limits below 0.5 µg L-1 for all species, detection limits were lowest for the AS16 and AS7 columns, where all detection limits were generally < 0.05 µg L-1. The major arsenic species in a water extract of a poultry litter sample was identified as ROX by all three methods with trace concentrations of DMA and As(V) also detected. The AS14 and AS16 separations also revealed a number of unidentified As species present at low concentrations, presumably metabolites of ROX. This methodology should prove useful in identifying organoarsenic compounds and the more toxic inorganic species in soils subject to poultry litter application.
Introduction The biogeochemistry of As continues to be an area of active research, and the current As poisoning disasters in Bangladesh and India are salient reminders of the consequences of As contamination of potable waters (1, 2). The recent proposal for reduction in the U.S. drinking water maximum contaminant level for As to 0.01 mg L-1 (66, FR, 6976) and the subsequent political debate have focused yet more attention on this element. The biogeochemical behavior and toxicity of As is a function of the prevalent As species, which in soils is determined by chemical and microbial conditions. Arsenic can occur as a number of different species in soils and aqueous environments, with the most prevalent being the inorganic oxyanions arsenate (H3AsO4) and, under reducing conditions, arsenite (H3AsO3) (3-5). The aliphatic organoarsenic compounds mono- and dimethylarsenic acid (MMA and DMA) are still used in agriculture as herbicides, * Corresponding author phone: (803)725-0854; fax: (803)725-3309; e-mail:
[email protected]. 4868
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001
and biogeochemical transformations of these compounds in soils may include further methylation followed by volatilization (6) and mineralization to the more toxic inorganic species (7). The organoarsenic compounds 4-aminobenzenearsenic acid (p-arsanilic acid, p-ASA) and 4-hydroxy-3-nitrobenzenearsenic acid (Roxarsone, ROX) are not naturally occurring in soils, but, due to the use of these compounds as feed additives in the poultry industry, they are anthropogenically introduced into soils through current management practices for disposal of poultry litter. These two organoarsenic compounds are regulated for use as feed additives for poultry (21 CFR 558.62, 558.530), where either compound can be used as the sole As source for the feed, and their beneficial properties are stated as control of coccidiosis, enhanced weight gain, and improved feed efficiency. However, it has been reported that the animal actually adsorbs very little of these compounds, with most being excreted unchanged in chemical form (8). Current USFDA regulations require that the feeding of poultry with As additives cease 5 days prior to slaughter in order to maintain muscle tissue As concentrations below 0.5 mg kg-1 and edible organ As concentrations below 2 mg kg-1 (21 CFR 556.60). However, the practice of feeding organoarsenic compounds to poultry results in elevated As concentrations in the manure, with concentrations in the range of 20-40 mg kg-1 reported (9-11). Because >90% of poultry manure is subsequently land applied (12) the question arises as to the fate of these aromatic arsenic compounds upon land application. In contrast to biosolids application, where land application is limited based on allowable trace element loadings (including As) by EPA 503 regulations (13), regulations governing poultry litter (if any) are generally governed at the state level and based on total N and P soil loading. Land application of poultry litter is accompanied by a flux of dissolved organic carbon (11, 14), which may lead to enhanced microbial activity and a reduced redox potential in the soil. Under these conditions it is likely that biotransformations of the aromatic arsenic compounds could occur, which in turn may result in the production of the more toxic inorganic species. Additionally, it is possible that the As-C bond may be broken by UV radiation; a result that has been reported for other aromatic arsenic compounds (15). The purpose of this study was to develop methodology for the separation and quantitation of p-ASA and ROX in poultry litter and in amended soil extracts. The method must resolve the aromatic arsenicals from the inorganic and simple aliphatic As compounds so that biogeochemical transformations of these aromatic arsenicals can be identified. Ion chromatography coupled to inductively coupled plasma mass spectrometry (IC-ICP-MS) has become the technique of choice for speciation analysis of As (16-18). In a previous study, IC-ICP-MS was used to quantify As(III), As(V), MMA, and DMA in soil solutions from fly ash-, poultry litter-, and sewage sludge-amended soils (19) but that chromatographic method was not optimized for separation and quantitation of the aromatic-arsenic compounds. A comprehensive study of chromatographic techniques for the separation of a number of aromatic arsenicals used as feed additives (including p-ASA and ROX) with detection by ICP-MS has been reported (20). This study reported that the nitroaromatic arsenicals were strongly retained by the stationary phase of a C18 column during reversed phase chromatography and were only eluted if methanol was a component of the mobile phase, but under these conditions As(III) and As(V) were not retained and coeluted in the void volume. Other studies have 10.1021/es0107172 CCC: $20.00
2001 American Chemical Society Published on Web 11/10/2001
focused on identification of ROX in animal tissue, with reversed phase HPLC-ICP-MS used to determine residual ROX in enzyme digests of chicken tissue (21) and spectrophotometric methodology exists for the determination of ROX in feeds (22). As a prerequisite to examining the biogeochemical dynamics of organoarsenic compounds in amended soils, this report details the development of IC-ICP-MS methodology that will allow the quantification of As(III), As(V), DMA, MMA, p-ASA, and ROX in a single analysis. Extracts of poultry litter are analyzed using this methodology to assess any matrix effects on the separation efficiency.
Materials and Methods Arsenic Standard Solutions. Sodium arsenite (GFS Chemicals, Columbus, OH), sodium cacodylate, (CH3)2As(O)ONa (Sigma, St. Louis, MI), sodium arsenate (Fisher Scientific, Fairlawn, NJ), 4-hydroxy-3-nitrobenzenearsenic acid (Aldrich, Milwaukee, WI), and p-aminobenzenearsenic acid (Sigma, St. Louis, MI) were used to prepare stock solutions of each As species at 500 mg L-1 as As in deionized water. Monomethyl arsenic acid (Crescent Chemicals Haupage, NY) was obtained as 100 mg L-1 as MMA in methanol. Chromatography standards were prepared on a daily basis using 18 MΩ-cm deionized water as the diluent. The concentration of each individual As species was determined by ICP-MS. Ion Chromatography-Inductively Coupled Plasma Mass Spectrometry. A Dionex (Sunnyvale, CA) GP50 gradient pump was used in conjunction with a Thermo Separations AS3500 autosampler. In our analytical configuration, timed event outputs on the autosampler were used to initiate the method clock on the pump and data collection by the ICPMS software upon sample injection and to reset the method clock (and, hence, restart the method under initial conditions) at the end of a run. The separation of As species was attempted on three columns that differed in either functional group or degree of cross-linking: the AS14 column has medium-high hydrophobicity, 55% cross-linking, and alkyl quaternary ammonium functional group; the AS16 column has ultralow hydrophobicity, 55% cross-linking, and an alkanol quaternary ammonium functional group; the AS7 column has mediumhigh hydrophobicity, 2% cross-linking, and an alkyl quaternary ammonium functional group. The ion exchange columns were supplied by Dionex Corp. (Sunnyvale, CA); all columns were 4 × 250 mm and were used with their appropriate guard column. Eluants used for the separations were 10 mM NaH2PO4 (pH 7.2), 50 mM NaOH, and 2.5 and 50 mM HNO3. These eluants were prepared from reagent grade salts or solutions and 18 MΩ-cm deionized water. The end of the analytical column was connected to the nebulizer of the ICP-MS (Perkin-Elmer, Elan 6000) by a 93 mm length of PEEK (polyetheretherketone) tubing (0.25 mm i. d.). Prior to chromatographic analysis, ICP-MS conditions, notably static lens voltage and nebulizer gas flow were optimized for maximum signal intensity at mass 75 by aspirating a 50 µg L-1 As standard under normal sample introduction conditions. The ICP-MS was at RF power of 1300 W, and a pneumatic nebulizer and cyclonic spray chamber were used. For IC-ICP-MS, the injection volume was 100 µL for all standards and samples. The calibration curves were generated for each separation method by running a three-point calibration with mixed standard solutions in the range 1-100 µg L-1 (as As) for each As species. All calibrations were linear with R2 > 0.995. Method detection limits were calculated based on 3σ of the baseline noise, and this value was converted to a concentration based on a three-point peak height calibration. Arsenic Concentration and Speciation of Poultry Litter. An air-dried poultry litter sample was kindly supplied by Dr. Miguel Cabrera (University of Georgia). Total arsenic in this
TABLE 1. Gradient Elution Program for AS14 Columna time
(D.I. H2O) A
(10 mM PO4, pH 7.2) B
flow rate
0.0 2.99 3.00 10.00 10.01
80% 80% 0 0 80%
20% 20% 100% 100% 20%
1 mL min-1 1 mL min-1 2 mL min-1 2 mL min-1 2 mL min-1
a One percent methanol was added to the mobile phase to enhance As signal intensity for ICP-MS.
poultry litter sample was determined by HNO3/H2O2 digestion of triplicate samples with analysis by ICP-MS. The water-soluble constituents of the poultry litter were determined on an extraction of 1 g of poultry litter with 10 mL of deionized water. After shaking for 2 h the extracts were centrifuged, filtered (0.22 µm), and diluted 20× prior to analysis of soluble trace elements by ICP-MS. The poultry litter extracts were high in dissolved organic carbon, and, because hydrophobic organic compounds have a high affinity for the stationary phase of ion exchange columns and can reduce the effective exchange capacity of the column, a cleanup step was performed on the poultry litter extract prior to IC-ICP-MS. An aliquot of the diluted poultry litter extract was passed through a solid-phase extraction (SPE) Sepak C18 cartridge to remove hydrophobic organic compounds. Soluble trace elements were determined on this fraction by ICP-MS and compared with the filtered sample prior to SPE to assess any trace element retention by the C18 cartridge; an aliquot of the filtered SPE sample was also used for IC-ICPMS As speciation analysis. The six standard As species were also passed through a C18 cartridge and were found not to be retained. A 5 mL aliquot of the diluted sample (after filtration and SPE) was spiked with 50 µL of a 5 mg L-1 mixed As standard, containing all six species and prepared from the primary standards referred to earlier, to assess any matrix effects during IC-ICP-MS analysis of the poultry litter extracts.
Results and Discussion The inorganic As species As(III) and As(V) and the simple aliphatic compounds MMA and DMA can be well separated on an Ionpac AS4 column using a carbonate eluant (23). Using this eluant and a similar ion exchange column as a starting point, the separation of As(III), As(V), MMA, DMA, p-ASA, and ROX was attempted on an AS14 column, which is a general purpose ion exchange column commonly used for the separation of inorganic anions using a carbonate eluant. However, ROX was strongly retained on the column even with increased carbonate eluant concentrations (data not shown). To elute ROX from the AS14 column it was necessary to use a stronger anion exchanger in the eluant, and by using PO4 as the eluant a satisfactory separation of all six As species was obtained. An example chromatogram of a mix of As standards (10 µg L-1 as As) is shown in Figure 1, and the gradient conditions used for the separation are given in Table 1. The most retained As species is ROX, which elutes with a retention time (RT) of 8.3 min. At pH 7.2, arsenite (pKa 9.2) is undissociated and elutes with the void volume. Excessive tailing of the As(III) peak is observed, indicating some interaction between As(III) and the stationary phase; in fact all the As species exhibit tailing to some extent. Also the two aliphatic organic arsenic compounds, MMA and DMA, are not baseline resolved. Previous studies have shown that addition of methanol to the eluant increases signal intensity for As, which has been ascribed to an increased ionization efficiency in the plasma (24); additions of methanol to the mobile phase used for this separation increased the maximum peak height signal intensity for all As species; the effect being most evident on VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4869
FIGURE 1. Separation of six As species on AS14 column with gradient elution program given in Table 1. Each As species is 10 µg L-1 as As.
FIGURE 3. Separation of six As species on AS16 column with gradient elution program given in Table 2. Each As species is 10 µg L-1 as As.
TABLE 2. Gradient Elution Program for AS16 Columna time
(D.I. H2O) A
(50 mM NaOH) B
flow rate
0.0 2.99 3.00 9.99 10.00
60% 60% 0 0 60%
40% 40% 100% 100% 40%
1 mL min-1 1 mL min-1 1.5 mL min-1 1.5 mL min-1 1.5 mL min-1
a One percent methanol was added to the mobile phase to enhance As signal intensity for ICP-MS.
FIGURE 2. Effect of additions of methanol to mobile phases for AS14 separation on (A) As signal intensity for each As species and (B) detection limits. going from 0 to 1% methanol in the mobile phase but also with further signal enhancement at 2.5% methanol (Figure 2A). An increase in the baseline signal was also observed as the methanol concentration of the mobile phase increased, and, more importantly, the baseline noise also increased. This increase in baseline noise meant that, despite the increase in signal intensity, detection limits were essentially unchanged by addition of methanol to the mobile phases (Figure 2B). The poor resolution of the early eluting peaks and the broad tailing peak shape for all As species exhibited under the above conditions on the AS14 column are not ideal for trace element work. It is also possible that for samples high in soluble Fe, a separation methodology at pH 7 may lead to within the column precipitation of Fe oxides, which may affect the elution of As oxyanions. Thus, further method development was carried out on ion-exchange columns more suited to separation of oxyanions. The AS16 column is a high capacity hydroxide selective column primarily designed for separation of polarizable anions. The six As species were effectively separated on this column (Figure 3) using a hydroxide step gradient program (Table 2). At the high pH used in this method As(III) is dissociated (pKa 9.2) and therefore elutes after the void volume; in fact As(III) elutes after DMA under these conditions. Roxarsone is the most retained component with an RT of approximately 9 min. Methanol addition to the mobile phase increased As signal intensity for all species and led to an improvement in detection limits for all species (Figure 4). Again the increase in As ICP-MS signal was greatest on going from 0 to 1% methanol with a further slight improvement on going to 2.5% 4870
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001
FIGURE 4. Effect of additions of methanol to mobile phases for AS16 separation on (A) As signal intensity for each As species and (B) detection limits. methanol, although in this case baseline noise for 1% and 2.5% methanol was about the same. The peaks for all the As species have high peak height:peak area ratios, do not exhibit excessive peak tailing, and thus give very low detection limits for all species. Using 50 mM NaOH as the eluant may lead to salt build-up on the ICP-MS cones and possible signal suppression during long run times. In these situations using a base where all the constituent elemental components are volatile in the plasma will alleviate concerns over salt deposition. We have recently obtained very similar chromatograms on the AS16 column using 50 mM tetramethylammonium hydroxide (TMAH) instead of NaOH with an identical gradient method to that given in Table 2 (data not shown). The AS7 column is a strong anion exchanger with additional capacity for hydrophobic interactions. This column has previously been used to separate the inorganic As(III), As(V), DMA, MMA, and arsenobetaine using HNO3 as the
TABLE 3. Gradient Elution Program for AS7 Columna time
(2.5 mM HNO3) A
(50 mM HNO3) B
flow rate
0.0 1.00 1.01 8.00 8.01
100% 100% 0 0 100%
0% 0% 100% 100% 0%
1 mL min-1 1 mL min-1 1 mL min-1 1 mL min-1 1 mL min-1
a One percent methanol was added to the mobile phase to elute ROX.
FIGURE 5. Separation of six As species on AS7 column with gradient elution program given in Table 3. Each As species is 10 µg L-1 as As. eluant (18) and As(III), As(V), DMA, MMA, and benzenearsonic acid using NH4BO3, NH4OH, and NH4H2PO4 in a gradient elution procedure (Dionex AS7 column manual). Given that the latter separation was effective for an aromatic arsenic compound, we attempted the separation of p-ASA and ROX using these eluants. However, ROX exhibited a high selectivity for the AS7 column and a suitable set of eluant conditions based on PO4 as the anion exchanger could not be found for this column. The recommended eluant for this column is HNO3, and separation of inorganic and simple aliphatic As compounds has been demonstrated using this eluant (18). Using the HNO3 gradient conditions of Mattsuch and Wennrich (18) ROX had a long RT, and both p-ASA and ROX gave broad peaks. Peragantis et al. (20) noted that ROX would not elute from a reverse-phase column unless methanol was present in the mobile phase; while the AS7 column is not fully compatible with organic solvents, methanol can be added to the mobile phase up to concentrations of 5%. When methanol was added to the eluants at 1% concentration ROX eluted from the column with a RT of 5 min, and with the gradient elution conditions specified in Table 3, all components were fully resolved with a run time of 7 min (Figure 5). No further signal enhancement was observed at 2.5% methanol concentration in the mobile phase; indeed, at this higher methanol concentration reproducible chromatograms were not obtained. The elution order of the As species is different on the AS7 column than the AS14 or the AS16 which is in part due to both the hydrophobicity of the column and the pH of the eluants. Dimethylarsenic acid, which is poorly retained from the void peak under previous anion exchange separations, elutes after As(V) with an RT of 4.2 min on the AS7 column. This observation is in contrast to the results obtained by Mattusch and Wennrich (18), using an AS7 column with an HNO3 eluant, where DMA eluted before As(V). This discrepancy is presumably due to difference in the injection pH of the eluant, which in our separation was nominally pH 2.6 (2.5 mM HNO3) compared with an injection pH of 3.3 (0.5 mM HNO3). The lower initial pH of our separation is nearer to the first pKa of As(V) (2.2), hence As(V) should elute earlier
TABLE 4. Arsenic Species Detection Limits for Separation on Three Different Anion Exchange Columns µg As L-1 species
As14
As16
As7
As(III) DMA MMA As(V) p-ASA ROX
0.112 0.044 0.061 0.079 0.076 0.254
0.015 0.011 0.014 0.029 0.018 0.061
0.024 0.019 0.006 0.008 0.053 0.027
under our conditions, while the later RT of DMA is presumably due to hydrophobic interactions with the column. Roxarsone, which is typically the most retained of any of the six As species in the other methods, elutes before p-ASA under these separation conditions. The early elution of ROX may be due to protonation of the phenol group on ROX at the low eluant pH. All species give sharp peaks, although all species also exhibit some tailing, with As(III) again eluting in the void volume and tailing significantly. The HNO3 eluant used for this separation is ideal for interfacing with ICP-MS as HNO3 is generally used during normal sample analysis by ICP-MS as the rinse solution and also in preserving samples prior to analysis. As can be seen in Table 4, detection limits for all As species were in the low ng L-1 concentration range on all three columns. The AS7 and AS16 columns gave comparable detection limits ca. 0.05 µg L-1 or less, and for any particular species detection limits were lower on either of these columns than on the AS14 column, with markedly better detection limits for As(III) and ROX being attainable with the AS16 or AS7 columns in comparison to the AS14 column. A drawback to the AS16 method utilizing the OH- eluant was the on-column oxidation of As(III) resulting in a small but detectable concentration of As(V) when running an ostensibly pure solution of As(III). Although thermodynamically favorable, oxidation of As(III) by O2 is kinetically slow (25); however, oxidation of As(III) has been reported in solutions of pH >9 (26, 27). For the AS16 separation the level of oxidation was reproducible and equated to 3.05 ( 0.13% (n ) 3) for 100 µ L-1 As(III). In terms of their applicability to analysis of soil extracts for As speciation, all three methods have some relative merit. Both OH- and PO4 have been used as chemical extracts of As from soils; OH- is generally a more effective extractant for all the species but leads to partial oxidation of As(III) (27). Hence, the AS14 or AS16 methods described above may be appropriate for soil extracts when OH- or PO4 has been used as the extractant. The AS7 method is ideally suited for interfacing with ICP-MS because the use of HNO3 as the eluant is very compatible with introduction into the ICPMS, and problems of signal drift or signal suppression should be minimal with this eluant. The constant flow rate of 1 mL min-1 for the AS7 method is also the most compatible for direct interfacing with the sample introduction system of ICP-MS and combined with the excellent detection limits and short run-time for elution of all six species this may be the most generally applicable of the three methods. However, it is also useful to use any two of these methods in tandem to more certainly identify the presence of a species as opposed to an unexpected species eluting with an identical RT to an analyte in one of the methods. Arsenic Speciation in Poultry Litter. Arsenic concentrations in poultry litter as high as 40 mg kg-1 have been reported. Arsenic occurs in the litter because of the practice of using the p-ASA and ROX as feed additives for poultry. These organic arsenic compounds are of low toxicity, but, mineralization of these compounds to the more toxic inorganic As comVOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4871
TABLE 5. Total (HNO3/H2O2), Water Soluble (0.22 µm Filtered), and Water Soluble + Nonretained by a C18 Solid Phase Extraction Column mg kg-1
V Cr Mn Fe Co Ni Cu Zn As Se Cd Ba Pb U
total elemental
water soluble
C18-water soluble
3.63 2.59 274.52 497.06 0.68 7.97 656.09 246.88 16.79 0.95 0.10 17.69 0.74 2.17
0.65 0.42 6.46 51.78 0.31 5.53 314.12 18.24 15.45 0.38 0.02 1.82 0.02 0.03
0.59 0.43 5.47 39.50 0.21 4.97 202.96 35.26 14.54 0.29 0.03 1.77 0.40 0.02
pounds could occur either during handling and storage of the litter or after the poultry litter is land-applied. This is of particular concern because areas of land receiving repeated poultry litter application may accumulate elevated As concentrations over time. Unlike other biosolids, land application of poultry litter is not regulated on the basis of trace element concentrations and loading. However, poultry litter can contain As concentrations that are comparable with sewage sludges, and for sewage sludges annual soil loading rates for As are limited to 2 kg ha-1 yr-1 and sludges with total As concentration > 40 mg kg-1 are prohibited from being land applied (13). Total and water-soluble concentrations of trace elements in the poultry litter sample used in this study are given in Table 5. Total As in the poultry litter was 16.7 mg kg-1, and 92% of total As was water soluble under the above conditions; also there was no significant difference in As concentration before and after the extract was passed through the C18 cartridge. Clearly, in comparison to other trace elements As from poultry litter is extremely water-soluble, and, unlike Cu, soluble As species are not retained during SPE. The speciation methodology developed above was applied to a water extract of poultry litter. Roxarsone was identified as the major As species in the poultry litter water extract using each of the above separation procedures (Figure 6), with trace concentrations of DMA and As(V) also present in each chromatogram. Under current U.S. federal regulations, only one of the aromatic organoarsenicals can be used as the As feed additive, hence, as it is most likely that all the poultry at any particular farm would be fed the same feed mix, one should expect to detect either ROX or p-ASA in a particular poultry litter extract. Clearly in this case ROX was used as the feed additive, and, in accordance with previous reports, it is mostly excreted unchanged in chemical form (8). A number of other unidentified As species, possible metabolites of ROX, were identified in the AS16 and AS14 separations but not in the AS7 separation (see insets to Figure 6A-C), where instead, it appeared that these unidentified As species eluted in illdefined peaks with similar RT to As(V) and DMA. The concentrations of As species in a 20× dilution of a 10:1water extraction of 1 g of poultry litter are shown in Table 6. Each method gives essentially identical concentrations for ROX, which equates to approximately 61% of the total water soluble As from the poultry litter. There are some discrepancies between concentrations reported for DMA and As(V) between the methods, and in the case of the AS7 separation it appears that unknown As species coelute, causing an increase in apparent DMA and As(V) concentra4872
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001
FIGURE 6. As species chromatograms for poultry litter water soluble extract on (A) AS14 column, (B) AS16 column, and (C) AS7 column. The inset to each figure shows the smaller eluting peaks in greater detail. One percent methanol was added to all eluants.
TABLE 6. Concentrations of Known Arsenic Species Detected in Poultry Litter Water Soluble Extractiona As species, µg L-1 column
As(III)
DMA
As(V)
ROX
As16 As7 As14
0.19 (0.05) n.d.* 0.31 (0.03)
0.92 (0.04) 1.64 (0.05) 0.64 (0.05)
1.94 (0.10) 3.90 (0.07) 1.70 (0.12)
40.33 (1.23) 40.77 (1.08) 38.49 (0.97)
a Three replicates of 1 g of poultry litter were extracted in 10 mL of D.I., diluted a further 20×, filtered (0.22 µm), and passed through a C18 solid-phase extraction cartridge. Concentrations shown are the average of the three extraction replicates with one SD in parentheses.
tions. The sum of the As species for any of the separations totaled only 64-71% of the total water soluble As determined by conventional ICP-MS analysis. One reason for this discrepancy is that there are clearly a number of unidentified As species present in the poultry litter extract (albeit at low concentrations). However, when corrections are made to the sum of the species based on the apparent As concentration of each of these unknown peaks, recovery for each of the methods increases to approximately 75-80%. The reason for the remaining discrepancy between sum of the species and total water soluble extracts is not known; spike recoveries for ROX in the poultry litter extracts were 103 ( 3.3% (n ) 3) indicating that the matrix did not affect the quatitation of this species. The discrepancy in mass balance may relate to an unknown As species with a high affinity for the stationary phase of the ion-exchange resins or, more simply, cumulative errors in either the total water soluble As analysis or the speciation analysis. In summary, we have identified three methodologies for the separation and quantitation of six As species of importance in environmental studies of poultry litter and litteramended soils. The major As species in poultry litter was identified as ROX, the original feed additive, but As(V), DMA, and other unidentified As species were also shown to be present. Further studies on the fate of ROX and p-ASA in soils are ongoing.
Acknowledgments This research was partially supported by financial Assistance Award Number DE-FC09-96SR18546 from the U.S. Department of Energy to the University of Georgia Research Foundation and by USDA National Research Initiative Program award number 2001-35107-09942. Thanks to Dr. P. E. Jackson of Dionex Corp. for supplying the AS7 and AS16 columns.
Literature Cited (1) Nickson, R. T.; McArthur, J. M.; Burgess, W. G.; Ahmed, K. M.; Ravenscroft, P.; Rahman, M. Nature 1998, 395, 338. (2) Masud Karim, M. D. Water Res. 2000, 34, 304. (3) Masscheleyn, P. H.; Delaune, R. D.; Patrick, W. H., Jr. Environ. Sci. Technol. 1991, 25, 1414. (4) Bowell, R. J.; Morley, N. H.; Din, V. K. Appl. Geochem. 1994, 9, 15. (5) Onken, B. M.; Hossner, L. R. Soil. Sci. Soc. Am. J. 1996, 60, 1385. (6) Woolson, E. A.; Kearney, P. C. Environ. Sci. Technol. 1973, 7, 47. (7) Gao, S.; Burau, R. G. J. Environ. Qual. 1997, 26, 753. (8) Andreae, M. O. In Organometallic compounds in the environment. Principles and reactions; Craig, P. J., Ed.; Longman: Harlow, U.K., 1986; p198. (9) Alloway, B. J. In Heavy Metals in Soils; Blackie: London, 1995; p 38 (10) Kingery, W. L.; Wood, C. W.; Delaney, D. P.; Williams, J. C.; Mullins, G. L. J. Environ. Qual. 1994, 23, 139. (11) Jackson, B. P.; Miller, W. P.; Sumner, M. E.; Schumann A. W. J. Environ. Qual. 1999, 28, 639. (12) Moore, P. A., Jr.; Daniel, T. C.; Sharply, A. N.; Wood, C. W. In Agricultural uses of municipal, animal, and industrial byprod-
(13) (14) (15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27)
ucts; USDA-ARS Conservation research report number 44; Wright, R. J., Kemper, W. D., Millner, P. D., Power, J. F., Korcak, R. F., Eds.; 1998. U.S. Environmental Protection Agency 1993. Fed. Reg. 58, 93879404. Jackson, B. P.; Miller, W. P. J. Environ. Qual. 2000, 29, 430. Stringer, C. E.; Attrep, M., Jr. Anal. Chem. 1979, 51, 731. Sheppard, B. S.; Caruso, J. A.; Heitkemper D. T.; Wolnik, K. A. Analyst 1992, 117, 971. Inoue, Y.; Kawabata, K.; Takahashi, H.; Endo, G. J. Chromatogr. A 1994, 675, 149. Mattusch, J.; Wenrich, R. Anal. Chem. 1998, 70, 3649. Jackson, B. P.; Miller, W. P. Environ. Sci. Technol. 1999, 33, 270. Pergantis, S. A.; Heithmar, E. M.; Hinners, T. A. Analyst 1997, 122, 1063. Dean, J. R.; Ebdon, L.; Foulkes, M. E.; Crews, H. M.; Massey, R. C. J. Anal. Atom. Spec. 1994, 9, 615. Webster, G. K.; Mandzij, R.; Drong, L. A.; Williams, W. H.; Mess, B. G.; Wodke, N. J. AOAC Intl. 1996, 79, 1012. Jackson, B. P.; Miller, W. P. J. Anal. Atom. Spectrosc. 1998, 13, 1107. Larsen, E. H.; Sturup, S. J. J. Anal. Atom. Spectrosc. 1994, 10991105. Cherry, J. A.; Shaikh, A. U.; Tallman, D. E.; Nicholson, R. V. J. Hydrol. 1979, 43, 373. Manning, B. A.; Goldberg, S. Environ. Sci. Technol. 1997, 31, 2005. Jackson, B. P.; Miller, W. P. Soil. Sci. Soc. Am. J. 2000, 64, 1616.
Received for review March 7, 2001. Revised manuscript received September 10, 2001. Accepted September 18, 2001. ES0107172
VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4873
