Biomagnification profiles of tributyltin (TBT) and triphenyltin (TPT) in Japanese coastal food webs elucidated by stable nitrogen isotope ratios

The measurement of organotins in the various biotas of coastal food webs with stable nitrogen isotope ratios (δ¹⁵N), which increase 3.4‰ per trophic level, can provide a biomagnification profile of organotins through food web. In this study, various biological samples were collected from three localities in Western Japan between 2002 and 2003 for analyses. Tributyltin (TBT) and triphenyltin (TPT) were still detected with a maximum of 99.5 and 8.7 ng wet weight g⁻¹, respectively. Unlike TBT, significant biomagnification of TPT through the food web (expressed by δ¹⁵N) was found in all three localities. The log transformed octanol–water partition coefficient (log K_ow) of TPT of 2.11–3.43 was overlapped by, but was slightly lower than, that of TBT of 3.70–4.70. Thus, this study demonstrates that although these chemicals have a log K_ow lower than 5, at least TPT undergoes significant biomagnification through the food web.     

Biomagnification of polybrominated diphenyl ethers (PBDEs) and polychlorinated biphenyls in a highly contaminated freshwater food web from South China

To evaluate the biomagnification extent of polybrominated diphenyls ethers (PBDEs) and polychlorinated biphenyls (PCBs) in a highly contaminated freshwater food web from South China, trophic magnification factors (TMFs) for 18 PBDE congeners and 53 PCB congeners were calculated. The TMF values ranged 0.26-4.47 for PBDEs and 0.75-5.10 for PCBs. Forty-five of 53 PCBs and BDEs 47, 100 and 154 had TMFs greater than one, suggesting their biomagnification in the present food web. The TMFs for PBDEs were generally smaller than those for PCBs with the same degree of halogenation, indicating a lower biomagnification potential for PBDEs compared to PCBs. For PCBs, it followed a parabolic relationship between TMFs and log K_OW (octanol-water partition coefficient). However, this relationship was not significant for PBDEs, possibly due to the more complex behaviors of PBDEs in the food web (e.g., metabolism), compared to that of PCBs.     

Uptake and translocation of organophosphates and other emerging contaminants in food and forage crops

Emerging contaminants in wastewater and sewage sludge spread on agricultural soil can be transferred to the human food web directly by uptake into food crops or indirectly following uptake into forage crops. This study determined uptake and translocation of the organophosphates tris(1-chloro-2-propyl) phosphate (TCPP) (log K _ow 2.59), triethyl-chloro-phosphate (TCEP) (log K _ow 1.44), tributyl phosphate (TBP) (log K _ow 4.0), the insect repellent N,N-diethyl toluamide (DEET) (log K _ow 2.18) and the plasticiser N-butyl benzenesulfonamide (NBBS) (log K _ow 2.31) in barley, wheat, oilseed rape, meadow fescue and four cultivars of carrot. All species were grown in pots of agricultural soil, freshly amended contaminants in the range of 0.6–1.0 mg/kg dry weight, in the greenhouse. The bioconcentration factors for root (RCF), leaf (LCF) and seed (SCF) were calculated as plant concentration in root, leaf or seed over measured initial soil concentration, both in dry weight. The chlorinated flame retardants (TCEP and TCPP) displayed the highest bioconcentration factors for leaf and seed but did not show the same pattern for all crop species tested. For TCEP, which has been phased out due to toxicity but is still found in sewage sludge and wastewater, LCF was 3.9 in meadow fescue and 42.3 in carrot. For TCPP, which has replaced TCEP in many products and also occurs in higher residual levels in sewage sludge and wastewater, LCF was high for meadow fescue and carrot (25.9 and 17.5, respectively). For the four cultivars of carrot tested, the RCF range for TCPP and TCEP was 10–20 and 1.7–4.6, respectively. TCPP was detected in all three types of seeds tested (SCF, 0.015–0.110). Despite that DEET and NBBS have log K _ow in same range as TCPP and TCEP, generally lower bioconcentration factors were measured. Based on the high translocation of TCPP and TCEP to leaves, especially TCPP, into meadow fescue (a forage crop for livestock animals), ongoing risk assessments should be conducted to investigate the potential effects of these compounds in the food web.     

Estimation of octanol/water partition coefficients for organic pollutants using reverse-phase HPLC

Octanol/water partition coefficients (K_ow's) of organic solutes are estimated with reasonable accuracy (standard deviation ± 0.25 log K_ow units) by elution from a C-18 column with 75:25 (v/v) MeOH-H₂O, based on results with 37 test compounds. Changes in solute activity coefficients from water to 75% MeOH account for the slope of the log K_ow-log k' plot, where k' is the HPLC capacity factor. The method is used to estimate K_ow's for 25 additional organic compounds, and some disparities between the results and those calculated using group additivity-constitutive factors are noted.     

Compound dependence of the relationship log Kow and log BCFL

Linear relatinships between log bioconcentration factor (BCF) and log K^ow for a variety of compounds have been reported many times in the literature. Analysis of the thermodynamics of the two partition processes has, however, shown that they are not analogous and that linear relationships can be expected to have different slopes for structurally unrelated compounds. In this study a set of literature lipid normalized BCF (BCF_L) values of chlorbenzenes (CBs) for rainbow trout and polycyclic aromatic hydrocarbons (PAHs) forDaphnia was put together with literature K_ow values. The slopes of the regression lines for log BCF_L versus log K_ow for the two groups of compounds proved to differ significantly in a statistical test using analysis of variance (ANOVA). The difference, which is of significance for estimates of BCFs in environmental modelling of these types of compounds, is explained by the differences in chemical structure of the two groups of compounds.     

Evaluation of reverse phase liquid chromatography/mass spectrometry for estimation of n-octanol/water partition coefficients for organic chemicals

A reverse-phase high pressure liquid chromatography/mass spectrometry (HPLC/MS method was developed for estimating n-octanol/water partition coefficients (K_ow) of anthropogenic molecules in complex chemical mixtures (e.g., complex effluents and solid waste leachates). The average error for an estimated log K_ow was ca. 0.5 and this error was similar for both aliphatic and aromatic compounds. The minimum level of detection using the total ion current profile generally decreased with increasing molecular weight between 100 and 600 daltons. Results obtained demonstrate that the HPLC/MS method is a viable technique for estimating log K_ow's of anthropogenic chemicals in complex environmental samples.     

Octanol–air partition coefficients of polybrominated biphenyls

The octanol–air partition coefficients (K_OA) for PBB15, PBB26, PBB31, PBB49, PBB103 and PBB153 were determined as a function of temperature using a gas chromatographic retention time technique with 1,1,1-trichloro-2,2-bis (4-chlorophenyl) ethane (p,p′-DDT) as a reference substance. The internal energies of phase change from octanol to air (Δ_OAU) were calculated for the six compounds and were in the range from 74 to 116 kJ mol⁻¹. Simple regression equations of log K_OA versus relative retention times (RRTs) on gas chromatography (GC), and log K_OA versus molecular connectivity indexes (MCI) were obtained, for which the correlation coefficients (r²) were greater than 0.985 at 283.15 K and 298.15 K. Thus the K_OA values of the remaining PBBs can be predicted by using their RRTs and MCI according to these relationships.     

Organotin concentrations in three intertidal neogastropods from the coastal waters of Taiwan

This study was aimed to determine organotin concentrations in the muricids Thais clavigera, Thais tuberosa and Morula granulata on an island wide scale and to find an adequate bioindicator species for long-term monitoring purpose. Samples were collected from the coast of Taiwan and vicinity islands with organotin analysis by graphite furnace atomic absorption spectrophotometry. Because triphenyltin (TPT) could not be discriminated from tributyltin (TBT) by the extraction method, organotins were measured as TBT + TPT and total hexane extractable organotins. Severity of imposex in the muricid population samples expressed as relative penis size (RPS) indices were calculated and compared to determine a suitable bioindicator species. In T. clavigera, T. tuberosa and M. granulata, TBT + TPT were 17-157, 1-44 and 117-1343 ng Sn g(-1) dry wt, respectively. And total organotins varied in the range of 181 to 1125, 23 to 44 and 229 to 1402 ng Sn g(-1) dry wt, respectively. A positive correlation was found between TBT + TPT and total organotins in M. granulata (TBT + TPT = Total organotins x 1.01-110.79; R2=0.97; p<0.001). At the site of Dapaisha, total organotins in M. granulata was 10 times higher than in T. tuberosa (i.e. 242 versus 23 ng Sn g(-1) dry wt). No signs of imposex were found in T. tuberosa (a single site in two subsequent years) and in M. granulata (seven collection sites). The degree of imposex (RPS) of T. clavigera varied from 0.2 to 38.1%. The RPS indices from fishing harbors were not higher than from rocky shores and oyster culture sites. No correlation was found between RPS indices and concentrations of TBT + TPT or total organotins. M. granulata seemed to be insensitive to organotin pollution although organotins were also detected. Our results indicate that organotin pollution is wide spread in coastal waters of Taiwan and pollutants at least include butyltins and phenyltins. As a bioindicator species, T. clavigera is more appropriate than T. tuberosa and M. granulata to monitor organotin contamination.     

Neptunium(V) adsorption to calcite

The migration behavior of the actinyl ions U(VI)O₂²⁺, Np(V)O₂⁺ and Pu(V,VI)O₂^(+,2+) in the geosphere is to a large extend controlled by sorption reactions (inner- or outer-sphere adsorption, ion-exchange, coprecipitation/structural incorporation) with minerals. Here NpO₂⁺ adsorption onto calcite is studied in batch type experiments over a wide range of pH (6.0–9.4) and concentration (0.4 μM–40 μM) conditions. pH is adjusted by variation of CO₂ partial pressure. Adsorption is found to be pH dependent with maximal adsorption at pH 8.3 decreasing with increasing and decreasing pH. pH dependence of adsorption decreases with increasing Np(V) concentration. EXAFS data of neptunyl adsorbed to calcite and neptunyl in the supernatant shows differences in the Np(V)-O-yl distance, 1.85 ± 0.01 Å for the adsorbed and 1.82 ± 0.01 Å for the solution species. The equatorial environment of the neptunyl in solution shows about 5 oxygen neighbours at 2.45 ± 0.02 Å. For adsorbed neptunyl there are also about 5 oxygen neighbours at 2.46 ± 0.01 Å. An additional feature in the adsorbed species' R-space spectrum can be related to carbonate neighbours, 3 to 6 carbon backscatterers (C-eq) at 3.05 ± 0.03 Å and 3 to 6 oxygen backscatterers (O-eq2) at 3.31 ± 0.02 Å. The differences in the Np(V)-O-yl distance and the C-eq and O-eq2 backscatterers which are only present for the adsorbed species indicate inner-sphere bonding of the adsorbed neptunyl species to the calcite surface. Experiments on adsorption kinetics indicate that after a fast surface adsorption process a continuous slow uptake occurs which may be explained by incorporation via surface dissolution and reprecipitation processes. This is also indicated by the part irreversibility of the adsorption as shown by increased K_D values after desorption compared to adsorption.     

Sorption of alpha and beta hydrophobic endosulfan in a Vertisol from southeast region of Turkey

Endosulfan has been applied to control numerous insects in a variety of food and non-food crops. Limited information is available on dynamics of this pesticide in the soil. The objective of this research was to determine the adsorption–desorption behavior of the alpha (α) and beta (β) endosulfan in a Vertisol from the southeast region of Turkey, where cotton is the main crop in the large irrigated lowlands. The α and β endosulfan were adsorbed considerably and Freundlich adsorption–desorption isotherms fitted the α and β endosulfan data (R² > 0.98). Freundlich adsorption coefficients (K_f) for the α endosulfan ranged between 21.63 and 16.33 while for the β endosulfan they were between 14.01 and 17.98 for the Ap and Bw2 horizons. The difference of K_f values of α and β endosulfan for two horizons were explained with the slight difference in the amount of organic matter and clay, but considerable difference in Fe contents of the two horizons. Alpha and β endosulfan K_fd values were 118.03 and 45.81 for the Ap and 48.08 and 68.71 for the Bw2 horizons. Higher adsorption and desorption behavior of the endosulfan isomers for the same horizon was attributed to poor physical bonding between the endosulfan molecule and the surfaces of fundamental soil particles. This fact is thought to increase the effective use of endosulfan in agriculture with a possibility of its movement to the surface and groundwater in the Vertisol studied.     

Monoterpene emissions and carbonyl compound air concentrations during the blooming period of rape (Brassica napus)

An increasing percentage of agricultural land in Germany is used for oil seed plants. Hence, rape has become an important agricultural plant (in Saxony 1998: 12% of the farmland) in the recent years. During flowering of rape along with intensive radiation and high temperatures, a higher production and emission of biogenic VOC was observed. The emissions of terpenes were determined and more importantly, high concentrations of organic carbonyl compounds were observed during this field experiment. All measurements of interest have been carried out during two selected days with optimal weather conditions. It is found that the origin or the mechanism of formation of different group of compounds had strong influence on the day to day variation of their concentrations. The emission flux of terpenes from flowering rape plants was determined to be 16–32 μg h⁻¹ m⁻² (30–60 ng h⁻¹ per g dry plant––540–1080 ng h⁻¹ per plant), in total. Limonene, α-thujene and sabinene were the most important compounds (about 60% of total terpenes). For limonene and sabinene reference emission rates (M_S) and temperature coefficients were determined: β_limonene=0.108 K⁻¹ and M_S=14.57 μg h⁻¹ m⁻²; β_sabinene=0.095 K⁻¹ and M_S=5.39 μg h⁻¹ m⁻².The detected carbonyl compound concentrations were unexpectedly high (maximum formaldehyde concentration was 18.1 ppbv and 3.4 ppbv for butyraldehyde) for an open field. Possible reasons for these concentrations are the combination of primary emission from the plants induced by high temperature and high ozone stress, the secondary formation from biogenically and advected anthropogenically emitted VOC at high radiation intensities and furthered by the low wind speeds at this time.     

Bioaccumulation Factors (BAFs) and Biota to Sediment Accumulation Factors (BSAFs) for PCBs in pike and eels

“Freely-dissolved” aqueous concentrations of 9 trichlorothrough heptachlorobiphenyls are reported, alongside those in sediments and fish from the R. Severn. For most congeners, BSAFs and lipid-normalised BAFs for pike exceed those for eels. Whilst R. Severn BSAFs are comparable with those for L. Ontario trout and New Bedford Harbour flounder, R. Severn BAFs are 1–2 orders of magnitude lower. This discrepancy may be due to inter-species variability, as well as inter-laboratory differences between operational definitions of “freely-dissolved” aqueous PCB, underlining that the same operational definition must be employed if R. Severn BAFs are extrapolated elsewhere. For eels, correlation of Log K^ow with Log BAF is better (R² = 0.66) than with BSAF (R² = 0.13), whilst similar correlation coefficients (R² = 0.81 and 0.82) were observed for pike. When Log K^ow is plotted against BSAF and Log BAF for both species combined, better correlation is observed for Log BAF (R² = 0.65), than BSAF (R² = 0.36). For both species combined, the observed relationship between Log BAF and Log K^ow for trichloro-through heptachlorobiphenyls is: Log BAF = 0.96 * Log K^ow −0.24.     

Heavy-metal ion complexation by particulate matter in the leachate of solid waste: a multi-method approach

The metal ion binding characteristics of particulate matter obtained from column experiments on the anaerobic digestion of solid waste were studied using a titrimetric approach. The experimental set-up allowed us to study the dynamics of particle bound ligand concentrations during digestion processes typically found in landfills.We developed a continuous titration method by simultaneously using a Cd-sensitive and pH electrode and combining metal and acid/base titrations. This technique allows for a more precise determination of pK_a-log K_M pairs for each ligand than metal titrations alone. The results were compared with titration methods using differential pulse anodic stripping voltammetry (DPASV) and atomic absorption spectroscopy (AAS) with longer equilibration times in order to further characterize ligand properties such as reaction kinetics, the electrochemical lability of the respective complex during DPASV, the distinction between metal adsorption to particulate matter and metal complexation by soluble ligands adhered to particles, reversibility of the binding process by competition studies, and resistance against purging with nitrogen gas.The properties of seven major metal binding ligands were identified and assignments to the most likely functional groups were made. The most important ligand properties are for ligand A: pK_a ≈ 9.2, log K_cd ≈ 7.0 fast reaction kinetics (mercapto groups); ligand B: pK_a = 4.8, log K_Cd ≈ 6.0, slow reaction kinetics (chelates with 3 or 4 carboxylic groups); ligand C: pK_a ≈ 6.0, log K_Cd ≈ 13.0, irreversible metal binding at basic pH-values (uptake inside bacterial cells); ligand D: pK_a = 7.7, log K_Cd = 4.0, runs parallel to N content of particulate matter with digestion time (primary amines neighboring oxo groups); ligand E: pK_a ≈ 12.0, log K_Cd = 9.0, runs parallel to P content of particulate matter (phosphate); ligand F:pK_a > 9.0, log K_{Cdf = pKa + 0.4}, runs parallel to N content of particulate matter (primary amines neighboring SH groups); and ligand G: pK_a ≤ 4.8, log K_Pb ≈ 4.3, strong Pb²⁺ ligand, even at low pH-values.Metal ions were found to be irreversibly bound by ligand C at low heavy-metal concentratins, whereas at higher concentrations the binding is reversible and can be predicted using the mass of the digestion process (methanogenic phase). All other ligands have their concentration maximum in the transition phase between acetogenic and methanogenic phase.     

Sorption of polar and nonpolar organic contaminants by oil-contaminated soil

Sorption of nonpolar (phenanthrene and butylate) and polar (atrazine and diuron) organic chemicals to oil-contaminated soil was examined to investigate oil effects on sorption of organic chemicals and to derive oil–water distribution coefficients (K_oil). The resulting oil-contaminated soil–water distribution coefficients (K_d) for phenanthrene demonstrated sorption-enhancing effects at both lower and higher oil concentrations (C_oil) but sorption-reducing (competitive) effects at intermediate C_oil (approximately 1 g kg⁻¹). Rationalization of the different dominant effects was attempted in terms of the relative aliphatic carbon content which determines the accessibility of the aromatic cores to phenanthrene. Little or no competitive effect occurred for butylate because its sorption was dominated by partitioning. For atrazine and diuron, the changes in K_d at C_oil above approximately 1 g kg⁻¹ were negligible, indicating that the presently investigated oil has little or no effect on the two tested compounds even though the polarity of the oil is much less than soil organic matter (SOM). Therefore, specific interactions with the active groups (aromatic and polar domains) are dominantly responsible for the sorption of polar sorbates, and thus their sorption is controlled by available sorption sites. This study showed that the oil has the potential to be a dominant sorptive phase for nonpolar pollutants when compared to SOM, but hardly so for polar compounds. The results may aid in a better understanding of the role of the aliphatic and aromatic domains in sorption of nonpolar and polar organic pollutants.     

Atmospheric deposition and canopy exchange processes in alpine forest ecosystems (northern Italy)

Throughfall and bulk precipitation chemistry were studied for five years (June 1994–May 1999) at two high elevation forest sites (Val Gerola and Val Masino) which were known to differ in terms of tree health, as assessed by live crown condition. The ion concentration of bulk precipitation samples did not differ significantly between sites, except for Mg²⁺, while the throughfall concentrations differed in the measured values of H⁺, N-NO₃⁻, Cl⁻, Na⁺, K⁺, DOC and weak organic acids. The results of the application of the canopy exchange model indicated a higher contribution from the dry deposition of N-NO₃⁻, N-NH₄⁺ and H⁺ at Val Gerola, where the damage symptoms were more evident. In addition, the canopy leaching of Ca²⁺, K⁺ and weak organic acids were 47%, 21% and 27% higher at Val Gerola than at Val Masino. Annual SO₄²⁻ deposition fluxes (21.3 kg ha⁻¹ yr⁻¹ at Val Masino and 23.6 kg ha⁻¹ yr⁻¹ at Val Gerola) were similar to those reported for moderately polluted European and U.S. sites. Annual N loads were 13.6 and 13.1 kg ha⁻¹ yr⁻¹ in the bulk input, and 15.0 and 18.0 kg ha⁻¹ yr⁻¹ in throughfall inputs, at Val Masino and Val Gerola, respectively. The contribution of the organic fraction to the total N atmospheric deposition load is significant, constituting 17% of the bulk flux and 40% of the throughfall flux. Measured nitrogen loads exceed the critical nutrient loads by several kg N ha⁻¹ at both stations. In particular the nitrogen throughfall load at Val Gerola was about 3 times higher than the critical values.     

Comparative studies of the Fe–UV, H2O2–UV, TiO2–UV/vis systems for the decolorization of a textile dye X-3B in water

The degradation of a common textile dye, Reactive-brilliant red X-3B, by several advanced oxidation technologies was studied in an air-saturated aqueous solution. The dye was resistant to the UV illumination (wavelength λ  320 nm), but was decolorized when one of Fe³⁺, H₂O₂ and TiO₂ components was present. The decolorization rate was observed to be quite different for each system, and the relative order evaluated under comparable conditions followed the order of Fe²⁺–H₂O₂–UV  Fe²⁺–H₂O₂ > Fe³⁺–H₂O₂–UV > Fe³⁺–H₂O₂ > Fe³⁺–TiO₂–UV > TiO₂–UV > Fe³⁺–UV > TiO₂–visible light (λ  450 nm) > H₂O₂–UV > Fe²⁺–UV. The mechanism for each process is discussed, and linked together for understanding the observed differences in reactivity.     

Removal of organic contaminants in bioretention medium amended with activated carbon from sewage sludge

Bioretention, also known as rain garden, allows stormwater to soak into the ground through a soil-based medium, leading to removal of particulate and dissolved pollutants and reduced peak flows. Although soil organic matter (SOM) is efficient at sorbing many pollutants, amending the bioretention medium with highly effective adsorbents has been proposed to optimize pollutant removal and extend bioretention lifetime. The aim of this research was to investigate whether soil amended with activated carbon produced from sewage sludge increases the efficiency to remove hydrophobic organic compounds frequently detected in stormwater, compared to non-amended soil. Three lab-scale columns (520 cm³) were packed with soil (bulk density 1.22 g/cm³); activated carbon (0.5% w/w) was added to two of the columns. During 28 days, synthetic stormwater—ultrapure water spiked with seven hydrophobic organic pollutants and dissolved organic matter in the form of humic acids—was passed through the column beds using upward flow (45 mm/h). Pollutant concentrations in effluent water (collected every 12 h) and polluted soils, as well as desorbed amounts of pollutants from soils were determined using GC-MS. Compared to SOM, the activated carbon exhibited a significantly higher adsorption capacity for tested pollutants. The amended soil was most efficient for removing moderately hydrophobic compounds (log K _ow 4.0–4.4): as little as 0.5% (w/w), carbon addition may extend bioretention medium lifetime by approximately 10–20 years before saturation of these pollutants occurs. The column tests also indicated that released SOM sorb onto activated carbon, which may lead to early saturation of sorption sites on the carbon surface. The desorption test revealed that the pollutants are generally strongly sorbed to the soil particles, indicating low bioavailability and limited biodegradation.

     

N-octanol/water partition coefficients by reverse phase liquid chromatography/mass spectrometry for eight tetrachlorinated planar molecules

N-octanol/water partition coefficients (K_ow's) were determined using reverse phase liquid chromatography for eight tetrachlorinated planar molecules. The log K_ow's are as follows: 2,3,7,8-, 1,2,3,4-, 1,3,7,9-, and 1,3,6,8-tetrachlorodibenzo-p-dioxins, 7.02, 7.18, 7.06, and 7.20, respectively; 2,3,7,8-tetrachlorodibenzofuran, 5.82; 1,3,6,8-tetrachloro-9H-carbazole, 5.75; 2,3,6,7-tetrachlorobiphenylene, 6.14; and 3,3′,4,4′-tetrachlorobiphenyl, 5.81. The accuracy of these values was estimated to be ca. ± 0.50.Reevaluation of log K_ow's previously determined using reverse phase liquid chromatography for chlorinated dibenzofurans and dibenzo-p-dioxins yielded values of ca. 7.0 for the tetrachlorodibenzo-p-dioxins, also. This reevaluation was performed by using experimental rather than estimated log K_ow's in the correlative relationship. In addition, these values suggest that the log K_ow's for dibenzo-p-dioxins and dibenzofurans range from ca. 4.0 for the unsubstituted parent molecule to ca. 8.6 for the octachlorinated compounds; much lower than previously reported.     

Removal of toxic ions (chromate, arsenate, and perchlorate) using reverse osmosis, nanofiltration, and ultrafiltration membranes

Rejection characteristics of chromate, arsenate, and perchlorate were examined for one reverse osmosis (RO, LFC-1), two nanofiltration (NF, ESNA, and MX07), and one ultrafiltration (UF and GM) membranes that are commercially available. A bench-scale cross-flow flat-sheet filtration system was employed to determine the toxic ion rejection and the membrane flux. Both model and natural waters were used to prepare chromate, arsenate, and perchlorate solutions (approximately 100 μg L⁻¹ for each anion) in mixtures in the presence of other salts (KCl, K₂SO₄, and CaCl₂); and at varying pH conditions (4, 6, 8, and 10) and solution conductivities (30, 60, and 115 mS m⁻¹). The rejection of target ions by the membranes increases with increasing solution pH due to the increasingly negative membrane charge with synthetic model waters. Cr(VI), As(V), and rejection follows the order LFC-1 (>90%) > MX07 (25–95%)  ESNA (30–90%) > GM (3–47%) at all pH conditions. In contrast, the rejection of target ions by the membranes decreases with increasing solution conductivity due to the decreasingly negative membrane charge. Cr(VI), As(V), and rejection follows the order CaCl₂ < KCl  K₂SO₄ at constant pH and conductivity conditions for the NF and UF membranes tested. For natural waters the LFC-1 RO membrane with a small pore size (0.34 nm) had a significantly greater rejection for those target anions (>90%) excluding (71–74%) than the ESNA NF membrane (11–56%) with a relatively large pore size (0.44 nm), indicating that size exclusion is at least partially responsible for the rejection. The ratio of solute radius (r_i,s) to effective membrane pore radius (r_p) was employed to compare ion rejection. For all of the ions, the rejection is higher than 70% when the r_i,s/r_p ratio is greater than 0.4 for the LFC-1 membrane, while for di-valent ions (, , and ) the rejection (38–56%) is fairly proportional to the r_i,s/r_p ratio (0.32–0.62) for the ESNA membrane.