Supplementary Information
Organocopper Complexes during Roxarsone Degradation in Wastewater Lagoons
Materials and Methods
Details on swine samples collection and characterization was reported earlier (Makris et al., 2008a). In brief- samples 1 - 8 were collected from lagoons surface and denoted as ‘supernatant’ samples, samples 9 - 16 were collected from lagoons bottom and denoted as ‘sludge’, and samples 17-19 were collected from swine housing facilities and denoted as ‘manure’. Solids content was determined by drying the samples at 105 °C for 24h. Following USEPA 3050B acid digestion procedure (USEPA. 2000), total recoverable As (Makris et al., 2008a), as well transition metals such as Cd, Cu, Cr, Fe, and Mn concentration were determined using ICP-MS. Liquid/liquid extraction procedure was developed to separate hydrophilic compounds from hydrophobic ones in PL water extracts. Study samples, toluene, and sodium perchlorate (100 mmol L-1) were mixed in 1:2:1 ratio (by volume), shaken (85 g) for 2 h (23°C), and frozen for 2 h. Frozen aqueous phase was allowed to thaw after discarding the toluene supernatant. The aqueous phase was filtered using 0.2 μm Whatman filters and used for As speciation analysis using LC/ICP-MS (Makris et al., 2008a). Based on the occurrence of seven known As species in animal waste, we chose to quantify As (III), DMA, MMA, 3-amino-4-hydroxyphenylarsonic acid (3-AHPAA), p-arsanilic acid (PSA), As (V), and roxarsone. A Dionex column IonPac AS14 (4 mm x 250 mm), HPLC (Prostar 210, Varian, Inc., Palo Alto, CA, USA), and ICP-MS (Elan 9000, PerkinElmer, Waltham, MA, USA) were used. Mobile phase for As speciation study was composed of 10 mM NaH2PO4 in 1% CH3OH (pH 7.2) and the gradient protocol was: 0-2.59 min, 20% of 10 mM NaH2PO4 in 1% CH3OH and 80% d-H2O in 1% CH3OH at a flow rate of 1 mL min-1; 3-10 min, 100% 10 mM NaH2PO4 in 1% CH3OH at a flow rate of 1.7 mL min-1.
For this study the two swine samples with the highest (sample # 13) and lowest (sample # 8) solids content from a previous work (Makris et al., 2008a) were selected for characterizing organoarsenicals and possible organo-metallic arsenical complexes. The samples were incubated for at room temperature for 16 d under aerobic (open to air atmosphere) and 12 h light/12 h dark conditions. Sixteen days incubation was selected based on significant degradation of roxarsone and As speciation findings from a previous study (Makris et al., 2008a). Following liquid/liquid extraction of the samples on 16th day of incubation, the filtered aqueous phase was diluted twice with the ES-MS solvent mixture. The solvent mixture consisted of 1:20 (v/v) of chloroform and acetonitrile (70%).
Infusion experiments were performed as follows. A quadrupole ion trap mass spectrometer (Finnigan LCQ Duo, Thermo Finnigan, San Jose, CA) with an on-axis electrospray source was used. Samples were infused continuously at 5 μL/min with a Hamilton (model 1750) syringe. All data were collected and analyzed using the Xcalibur software (Thermo Finnigan) in full scan and MS/MS mode. The LCQ Duo was tuned with a 5.0 ppb solution of roxarsone using the singly charged peak at m/z 262.0. The ESI-MS response of roxarsone was explored in both positive and negative ionization modes. Best results, detection and sensitivity were obtained in the negative ion-mode. Blanks and performance evaluation samples were analyzed with each set of samples. Nitrogen (80 PSI) and high-purity helium were used as the nebulizing gas and collision gas, respectively. The relative collision energy used for CID was uncalibrated and in arbitrary units. The collision energy was optimized to generate essentially mono-collisions. Normalized collision energies (NCE) between 5 and 70 were used. The MS conditions routinely used were: source voltage, 2.07 kV; capillary temperature, 181.20° C; capillary voltage, -11.64 V; sheath gas flow rate, 57.71 (arbitrary units), and the auxiliary gas flow rate, 48.07 (arbitrary units). The scan range of the mass spectrometer was m/z 50–600. The isolation width was optimized in order to isolate selectively the mono- and di-isotopic parent ion, arsenic and copper, with a minimum loss of sensitivity. Generally, an isolation width of m/z 10 was used for the samples. Simulations of the stable isotope patterns were made using Isotope Viewer in Xcalibur. For the simulated mass spectra, the default settings were used unless otherwise noted: resolution, 1 Da at 5% height and a -1 charge. To ensure optimal detection limits and reproducibility, the sample cone of the ES-MS was cleaned daily according to the manufacturer’s instructions.
5
Table S1. Collision-induced dissociation (CID) analyses for the selected molecular ion peaks observed in the full MS spectra (Fig. 2) of the Swine lagoon sample with high (# 13) and low (# 8) solid contents. The information in the parenthesis of each chemical consists of m/z of the fragment ion, normalized collision energy, and relative abundance (%), respectively.
Sample # 13 / Sample # 8[3-AHPAA]- / [3-AHPAA : N2O]- / [Rox]- / [Cu : 3NO3-]-
MS2 / [3-AHPAA]- (231.9,35,25%), [3-AHPAA - H2O]- (213.9,35,100%) / [3-AHPAA : N2O]- (275.9,35,25%), [3-AHPAA : N2O - H2O]- (257.9,35,100%), [3-AHPAA : N2O - H2O - N2O]- (214.1,35,24%) / [Rox]- (261.8,40,22%), [Rox - H2O]- (243.8,40,100%) / [Cu : 3NO3-]- (248.9,65,25%), [Cu : 2NO3]- (186.9,65,100%)
MS3 / [3-AHPAA - H2O]- (213.9,38,28%), [3-AHPAA - 2H2O]- (195.9,38,100%), [AsO3]- (122.9,38,38%), [AsO2]- (107.1,38,4%) / [3-AHPAA : N2O - H2O]- (257.8,40,100%), [3-AHPAA : N2O - 2H2O]- (239.8,40,100%), [3-AHPAA : N2O - H2O - N2O]- (213.8,40,100%), [3-AHPAA - 2H2O]- (196.0,40,5%), [AsO3]- (123.0,40,2%), [AsO2]- (106.9,40,8%) / [Rox - H2O]- (243.8,45,30%), [Rox - H2O - As(V)-O]- (152.8,45,55%), [Rox-H2O–As(V)-O-NO2]- (106.8,100%) / [Cu : 2NO3-]- (186.9,70,30%), [Cu : 2NO3- - N2O]- (142.9,70,100%), [Cu : 2NO3 - NO2]- (140.9,65,40%), [Cu : 2NO3 - Cu -NO3]- (61.9,65,10%)
MS4 / [3-AHPAA : N2O - 2H2O]- (239.8,42.5,38%), [3-AHPAA : N2O - 2H2O - N2O -N - 2O]- (149.0,42.5,100%)
5
Figure S1. MS/MS analyses of the possible organo-copper complex in sample # 8 using ES-MS.
5