Bicarbonate-form anion exchange: affinity, regeneration, and stoichiometry

Magnetic ion exchange (MIEX) is an effective process for removing dissolved organic carbon (DOC) from natural waters, but its implementation has been limited due to production of waste sodium chloride solution (i.e., brine) from the regeneration process. Chloride is of concern because elevated concentrations can have adverse effects on engineered and natural systems. The goal of this research was to explore the efficacy of using anion exchange resin with bicarbonate as the mobile counter ion, which would produce a non-chloride regeneration solution. It was found that bicarbonate-form MIEX resin had a similar affinity as chloride-form MIEX resin for sulfate, nitrate, DOC, and ultraviolet-absorbing substances. Both bicarbonate-form and chloride-form MIEX resins showed the greatest removal efficiencies as fresh resin, and removal efficiency decreased with multiple regeneration cycles. Nevertheless, sodium bicarbonate solution was as effective as sodium chloride solution at regenerating MIEX resin. Regeneration of the bicarbonate-form MIEX resin was illustrated by sparging carbon dioxide gas in a water/resin slurry. This regeneration process would eliminate the need for the addition of salts such as sodium chloride or sodium bicarbonate. The stoichiometry of the bicarbonate-form resin revealed that the bicarbonate was deprotonating within the resin matrix leading to a mixture of both carbonate and bicarbonate mobile counter ions. This work makes an important contribution to ion exchange applications for water treatment by evaluating the affinity, regeneration, and stoichiometry of bicarbonate-form anion exchange.     

Removal of bromide and natural organic matter by anion exchange

Bromide removal by anion exchange was explored for various water qualities, process configurations, and resin characteristics. Simulated natural waters containing different amounts of natural organic matter (NOM), bicarbonate, chloride, and bromide were treated with a polyacrylate-based magnetic ion exchange (MIEX) resin on a batch basis to evaluate the effectiveness of the resin for removal of bromide. While bromide removal was achieved to some degree, alkalinity (bicarbonate), dissolved organic carbon (DOC), and chloride were shown to inhibit bromide removal in waters with bromide concentrations of 100 and 300 μg/L. Water was also treated using a two-stage batch MIEX process. Two-stage treatment resulted in only a slight improvement in bromide removal compared to single-stage treatment, presumably due to competition with the high concentration of chloride which is present along with bromide in natural waters. In view of the relatively poor bromide removal results for the MIEX resin, a limited set of experiments was performed using polystyrene resins. DOC and bromide removal were compared by treating model waters with MIEX and two polystyrene resins, Ionac A-641 and Amberlite IRA910. The two polystyrene resins were seen to be more effective for bromide removal, while the MIEX resin was more effective at removing DOC.     

Effects of blending of desalinated water with treated surface drinking water on copper and lead release

This study examined effects of desalinated water on the corrosion of and metal release from copper and lead-containing materials. A jar test protocol was employed to examine metal release from copper and lead-tin coupons exposed to water chemistries with varying blending ratios of desalinated water, alkalinities, pHs and orthophosphate levels. Increasing fractions of desalinated water in the blends resulted in non-monotonic changes of copper and lead release, with generally lower metal concentrations in the presence of desalinated water, especially when its contribution increased from 80% to 100%. SEM examination showed that the increased fractions of desalinated water were associated with pronounced changes of the morphology of the corrosion scales, likely due to the influence of natural organic matter. This hypothesis was corroborated by the existence of correlations between changes of the ζ-potential of representative minerals (malachite and hydrocerussite) and metal release. For practical applications, maintaining pH at 7.8 and adding 1 mg/L orthophosphate as PO₄ were concluded to be adequate to decrease copper and lead release. Lower alkalinity of desalinated water was beneficial for blends containing 50% or more desalinated water.     

Effects of magnetic ion exchange pretreatment on low pressure membrane filtration of natural surface water

Magnetic ion exchange (MIEX) pretreatment has been increasingly employed by water treatment plants for removal of dissolved organic carbon (DOC). In this study, the effects of MIEX pretreatment on low pressure membrane filtration of natural surface water were investigated under different feedwater qualities, membrane properties, and MIEX dosing conditions. Regardless of feedwater DOC, moderate decrease in the total and hydraulically irreversible fouling was observed for a polyvinylidene fluoride (PVDF) microfiltration membrane and a polyethersulfone ultrafiltration (UF) membrane after MIEX pretreatment, which was coincident with moderate removals of high molecular weight DOC in the feedwaters. Comparatively, the fouling of a PVDF UF membrane did not decrease after MIEX pretreatment, revealing the impact of membrane properties on membrane fouling in the presence of MIEX pretreatment. Reuse of virgin or regenerated MIEX resulted in similar membrane fouling as observed with single use of the virgin MIEX. The level of DOC removal by MIEX was similar to the removal of MS2 bacteriophage spiked in the feedwater, suggesting a potential similarity in the removal of organic and microbial particles. In conclusion, MIEX pretreatment was effective for DOC removal, but less effective in controlling short-term membrane fouling or removing viruses.     

Bench-scale testing of a magnetic ion exchange resin for removal of disinfection by-product precursors

The objective of this research was to compare enhanced coagulation with anion exchange for removal of disinfection by-product (DBP) precursors (i.e. natural organic matter (NOM) and bromide). Treatment with a magnetic ion exchange resin (MIEX((R))) was the primary focus of this study. Raw waters from four utilities in California were evaluated. The waters had low turbidity, low to moderate organic carbon concentrations, a wide range of alkalinities, and moderate to high bromide ion concentrations. The treated waters were compared based on removal of ultraviolet (UV) absorbance, dissolved organic carbon (DOC), trihalomethane formation potential (THMFP), and haloacetic acid formation potential (HAAFP). The results indicated that treatment with MIEX is more effective than coagulation at removing UV-absorbing substances and DOC. Treatment with MIEX and treatment with MIEX followed by coagulation yielded similar results, suggesting that coagulation of MIEX-treated water does not provide additional removal of organic carbon. MIEX treatment reduced the THMFP and HAAFP in all waters, and did so to a greater extent than coagulation. Treatment with MIEX was most effective in raw waters having a high specific UV absorbance and a low anionic strength. Following MIEX treatment, subsequent chlorination resulted in a shift to the more brominated THM and HAA species as compared to chlorination of the raw water. MIEX also removed bromide to varying degrees, depending on the raw water alkalinity and initial bromide ion concentration.     

Combined ion exchange treatment for removal of dissolved organic matter and hardness

Dissolved organic matter (DOM) and hardness cations are two common constituents of natural waters that substantially impact water treatment processes. Anion exchange treatment, and in particular magnetic ion exchange (MIEX), has been shown to effectively remove DOM from natural waters. An important advantage of the MIEX process is that it is used as a slurry in a completely mixed flow reactor at the beginning of the treatment train. Hardness ions can be removed with cation exchange resins, although typically using a fixed bed reactor at the end of a treatment train. In this research, the feasibility of combining anion and cation exchange treatment in a single completely mixed reactor for treatment of raw water was investigated. The sequence of anion and cation exchange treatment, the number of regeneration cycles, and the chemistry of the regeneration solution were systematically explored. Simultaneous removal of DOM (70% as dissolved organic carbon) and hardness (>55% as total hardness) was achieved by combined ion exchange treatment. Combined ion exchange is expected to be useful as a pre-treatment for membrane systems because both DOM and divalent cations are major foulants of membranes.     

Effects of chloride,sulfate and natural organic matter (NOM) on the accumulation and release of trace-level inorganic contaminants from corroding iron

This study examined effects of varying levels of anions (chloride and sulfate) and natural organic matter (NOM) on iron release from and accumulation of inorganic contaminants in corrosion scales formed on iron coupons exposed to drinking water. Changes of concentrations of sulfate and chloride were observed to affect iron release and, in lesser extent, the retention of representative inorganic contaminants (vanadium, chromium, nickel, copper, zinc, arsenic, cadmium, lead and uranium); but, effects of NOM were more pronounced. DOC concentration of 1 mg/L caused iron release to increase, with average soluble and total iron concentrations being four and two times, respectively, higher than those in the absence of NOM. In the presence of NOM, the retention of inorganic contaminants by corrosion scales was reduced. This was especially prominent for lead, vanadium, chromium and copper whose retention by the scales decreased from >80% in the absence of NOM to <30% in its presence. Some of the contaminants, notably copper, chromium, zinc and nickel retained on the surface of iron coupons in the presence of DOC largely retained their mobility and were released readily when ambient water chemistry changed. Vanadium, arsenic, cadmium, lead and uranium retained by the scales were largely unsusceptible to changes of NOM and chloride levels. Modeling indicated that the observed effects were associated with the formation of metal–NOM complexes and effects of NOM on the sorption of the inorganic contaminants on solid phases that are typical for iron corrosion in drinking water.     

Influence of natural organic matter on the morphology of corroding lead surfaces and behavior of lead-containing particles

Influence of natural organic matter (NOM) on the morphology of lead surfaces exposed to drinking water and on the properties of lead-containing colloidal particles was explored based on the data of scanning electron microscopy, sequential filtrations, measurements of particle size distributions and electrophoretic potential. It was demonstrated that NOM prevented the formation of cerussite and hindered the growth of hydrocerussite crystals. Measurements of zeta-potential showed that the surface activity was highest for unaltered NOM, while ozonation and chlorination decreased it. The concentrations of soluble lead and tin increased several fold in the presence of NOM, while large colloidal particles of lead and solder corrosion products tended to break down to form smaller fragments. It is suggested that these phenomena are important for understanding of lead release mechanisms in drinking water.     

A pilot-scale evaluation of magnetic ion exchange treatment for removal of natural organic material and inorganic anions

The objective of this research was to evaluate a magnetic ion exchange process (MIEX) for the removal of natural organic material (NOM) and bromide on a continuous-flow pilot-scale basis under different operating conditions and raw water characteristics. The most important operating variable was the effective resin dose (ERD), which is the product of the steady-state resin concentration in the contactor and the regeneration ratio. The raw water employed in this study had a moderate concentration of ultraviolet (UV)-absorbing substances and dissolved organic carbon (DOC), and a low turbidity, alkalinity, and concentration of competing anionic species. Experiments were conducted using the ambient raw water and raw water spiked with bromide, chloride, and sulfate. Substantial removal of UV-absorbing substances and DOC was achieved at ERDs as low as 0.16mL/L. Moderate bromide removal was achieved, depending on the ERD. Increasing the sulfate concentration resulted in decreased removal of UV-absorbing substances, DOC, and bromide. Consistent results were observed between the continuous-flow pilot plant tests and batch equilibrium studies.     

Long-term performance of bicarbonate-form anion exchange: removal of dissolved organic matter and bromide from the St. Johns River, FL, USA

The goal of this research was to evaluate the long-term performance of magnetic ion exchange (MIEX) treatment using bicarbonate as the mobile counter ion (i.e., MIEX-HCO₃) and sodium bicarbonate for regeneration. This work is important because there are many unknowns concerning the affinity and regeneration efficiency of bicarbonate-form anion exchange, whereas chloride-form anion exchange (i.e., MIEX-Cl resin) is well-studied. Raw water samples were collected approximately two times per month for one year from a single location on the St. Johns River (SJR), FL, USA. The SJR is characterized by high concentrations of dissolved organic carbon (DOC; 12-26 mg C/L) and bromide (550-1100 μg/L), and is being considered as an alternative drinking water supply. Jar tests were conducted using MIEX-HCO₃ resin, and MIEX-Cl resin was used as a baseline for comparison. The same batch of MIEX-HCO₃ and MIEX-Cl resin was used for the entire study, which was accomplished by regenerating the resins after each jar test in concentrated solutions of sodium bicarbonate and sodium chloride, respectively, and resulted in 21 regeneration cycles. Maximum removal efficiency was achieved with fresh MIEX-HCO₃ resin and virgin MIEX-Cl resin. Both forms of fresh/virgin MIEX resin also had the same affinity sequence with sulfate ≈ UV-absorbing substance > DOC > bromide. The removal efficiency of both forms of MIEX resin decreased as the number of regeneration cycles increased, with MIEX-HCO₃ resin showing 7-18% lower removals than MIEX-Cl resin after 21 regeneration cycles. The affinity sequence of regenerated MIEX-HCO₃ and MIEX-Cl resins differed from fresh resin with UV-absorbing substances > DOC > sulfate > bromide. Scanning electron microscopy and simulated MIEX-HCO₃ treatment under rapidly changing water quality were also used to improve the understanding of bicarbonate-form anion exchange. The major contribution of this research is a systematic study of the extended use of bicarbonate-form anion exchange resin in the context of affinity, regeneration efficiency, and changing water quality.     

Effects of NOM properties on copper release from model solid phases

This study examined impacts of concentrations and properties of natural organic matter (NOM) on copper release from characteristic copper solid model phases such as tenorite CuO and malachite Cu₂(OH)₂CO₃. Unaltered Aldrich humic acid (AHA) and standard Suwannee River fulvic acid (SRFA) strongly increased copper release from the model phases but NOM alteration by chlorination or ozonation gradually suppressed or, at higher oxidant doses, eliminated these effects. The nature of NOM changes induced by chlorination and ozonation was examined using differential absorbance spectroscopy (DAS) and high-performance size-exclusion chromatography (HPSEC). The data of these methods show that NOM molecules with higher apparent molecular weight (AMW), higher aromaticities and contributions of protonation-active phenolic and carboxylic groups play a key role in adsorption and colloidal dispersion of the model solids. The data also show that metal release from model phases was well correlated with a number of spectroscopic parameters characterizing NOM properties, notably SUVA₂₅₄, spectral slopes of NOM absorbance, and differential absorbance at wavelength of 280 nm and 350 nm that is indicative of the contributions of carboxylic and phenolic functional groups. Changes of ζ-potential of the model solid phases were the strongest predictor of the enhancement of copper release especially in the system controlled by malachite. While effects of NOM on the ζ-potential of tenorite and malachite were prominent for unaltered NOM, its oxidation by chlorine and ozone was accompanied by a gradual decrease and ultimately disappearance of its surface activity.     

Application of the Nernst-Planck approach to lead ion exchange in Ca-loaded Pelvetia canaliculata

Ca-loaded Pelvetia canaliculata biomass was used to remove Pb²⁺ in aqueous solution from batch and continuous systems. The physicochemical characterization of algae Pelvetia particles by potentiometric titration and FTIR analysis has shown a gel structure with two major binding groups - carboxylic (2.8 mmol g⁻¹) and hydroxyl (0.8 mmol g⁻¹), with an affinity constant distribution for hydrogen ions well described by a Quasi-Gaussian distribution. Equilibrium adsorption (pH 3 and 5) and desorption (eluents: HNO₃ and CaCl₂) experiments were performed, showing that the biosorption mechanism was attributed to ion exchange among calcium, lead and hydrogen ions with stoichiometry 1:1 (Ca:Pb) and 1:2 (Ca:H and Pb:H). The uptake capacity of lead ions decreased with pH, suggesting that there is a competition between H⁺ and Pb²⁺ for the same binding sites. A mass action law for the ternary mixture was able to predict the equilibrium data, with the selectivity constants α_Ca^H = 9 ± 1 and α_Ca^Pb = 44 ± 5, revealing a higher affinity of the biomass towards lead ions. Adsorption (initial solution pH 4.5 and 2.5) and desorption (0.3 M HNO₃) kinetics were performed in batch and continuous systems. A mass transfer model using the Nernst-Planck approximation for the ionic flux of each counter-ion was used for the prediction of the ions profiles in batch systems and packed bed columns. The intraparticle effective diffusion constants were determined as 3.73 × 10⁻⁷ cm² s⁻¹ for H⁺, 7.56 × 10⁻⁸ cm² s⁻¹ for Pb²⁺ and 6.37 × 10⁻⁸ cm² s⁻¹ for Ca²⁺.     

Impacts of blending ground, surface, and saline waters on lead release in drinking water distribution systems

The impacts of distribution water quality changes caused by blending different source waters on lead release from corrosion loops containing small lead coupons were investigated in a pilot distribution study. The 1-year pilot study demonstrated that lead release to drinking water increased as chlorides increased and sulfates decreased. Silica and calcium inhibited lead release to a lesser degree than sulfates. An additional 3-month field study isolated and verified the effects of chlorides and sulfates on lead release. Lead release decreased with increasing pH and increasing alkalinity during the 1-year pilot study; however, the effects of pH and alkalinity on lead release, were not clearly elucidated due to confounding effects. A statistical model was developed using nonlinear regression, which showed that lead release increased with increasing chlorides, alkalinity and temperature, and decreased with increasing pH and sulfates. The model indicated that primary treatment processes such as enhanced coagulation and RO (reverse osmosis membrane) were related to lead release by water quality. Chlorides are high in RO-finished water and increase lead release, while sulfates are high following enhanced coagulation and decrease lead release.     

Adverse health effects of lead exposure on children and exploration to internal lead indicator

Our research on adverse effects of lead exposures on physical and neurobehavioral health of children aged 6–12 years in 4 villages, labeled as K, M, L, and X, in rural China, was reported in this article. Lead in blood (PbB), urine (PbU), hairs (PbH), and nails (PbN) were measured by graphite furnace atomic absorption spectrometry. Abbreviated Symptom Questionnaire of Conner's instruments and Revised Raven's Standard Progressive Matrices were applied to evaluate childhood attention deficit/hyperactivity disorders (ADHD) and intelligences. Geometric means (SD) of PbB, PbU, PbH and PbN concentrations were 71.2 μg/L (1.56), 11.7 μg/g (1.75), 12.5 μg/g (2.82), and 25.3 μg/g (2.79), respectively. 54 (17.0%) children had PbB levels of ≥ 100 μg/L. Boys, the 6–10 years old, and living in village K were 2.11, 2.48, and 9.16 times, respectively, more likely to be poisoned by lead than girls, aged 11–12 years, and residing in X. 18 (5.7%) and 37 (11.7%) subjects had ADHD and mental retardations, respectively. Inverse relationships between intelligences and natural log transformed PbU and PbH levels were observed with respective odds ratios (95%CI) of 1.79 (1.00–3.22) and 1.46 (1.06–2.03) or 1.28 (1.04–1.58) and 1.73 (1.18–2.52) by binary or ordinal logistic regression modeling. ADHD prevalence was different by gender and age of subjects. PbU, PbH, and PbN related to PbB positively with respective correlation coefficients of 0.530, 0.477, and 0.181. Receiver operating characteristic (ROC) curves of the three measurements reveled areas under curves (AUCs) being 0.829, 0.758, and 0.687, respectively. In conclusion, children had moderate levels of lead exposures in this rural area. Intelligence declines were associated with internal lead levels among children. ROC analysis suggests PbU an internal lead indicator close to PbB.     

Combination of coagulation and ion exchange for the reduction of UF fouling properties of a high DOC content surface water

The treatment of a high DOC content surface water (about 6mg DOC/L) using anion exchange resins (MIEX resin from Orica or IRA958 resin from Rohm and Haas) can remove up to 80% of DOC in less than 45min. The combination of coagulation prior to or after resin treatment only slightly improves the removal of DOC (0.2-0.3mg/L) but eliminates the high MW organic compounds (MW >20kDa) attributed to biopolymers (proteins and polysaccharides) that were not removed using anion exchange resins alone and that were found to be responsible for reversible fouling of UF membranes (YM 100 UF membrane from Millipore with MW cut-off of 100kDa). The combination of treatments then significantly improves the permeability of the UF membrane. Also, the combination of both treatments allows a reduction of the coagulant doses by a factor of 6 with no impact on the DOC removal and the filterability of produced waters.     

Combined influence of membrane surface properties and feed water qualities on RO/NF mass transfer, a pilot study

The impact of membrane surface characteristics and NOM on membrane performance has been investigated for varying pretreatment and membranes in a field study. Surface charge, hydrophobicity and roughness varied significantly among the four membranes used in the study. The membranes were tested in parallel following two different pretreatment processes, an enhanced Zenon ultrafiltration process (ZN) and a compact CSF process (Superpulsator (SP)) prior to RO membrane treatment for a total of eight integrated membrane systems. All membrane systems were exposed to the similar temperature, recovery and flux as well as chemical dosage. The membrane feed water qualities were statistically equivalent following ZN pretreatment and SP pretreatment except for NOM and SUVA. Membrane surface characteristics, NOM and SUVA measurements were used to describe mass transfer in a low-pressure RO integrated membrane system. Solute and water mass transfer coefficients (MTCs) were investigated for dependence on membrane surface properties and NOM mass loading. Inorganic MTCs were accurately described by a Gaussian distribution curve. Water productivity decreased with NOM loading and increased with contact angle and roughness. The negative effects of NOM loading on productivity were reduced as the negative charge on the membrane surface increased. Inorganic MTCs were also correlated to surface hydrophobicity and surface roughness. The permeability change of identical membranes was related to NOM loading, hydrophobicity and roughness. Organic fouling as measured by water, organic and inorganic mass transfer was less for membranes with higher hydrophilicity and roughness.     

高温长距离供热系统腐蚀结垢控制研究

在2018-2019年采暖季供热期间,发现大温差供热管网换热设备腐蚀结垢情况严重.经过检验和分析,垢主要来源于循环水结垢和腐蚀产物的沉积;腐蚀产物来源于管道内壁腐蚀.同时热网循环水氯离子含量过高,板式换热器不锈钢发生了点蚀.研究试验结果表明:采用阻垢剂、缓蚀剂、调节pH和控制氯离子含量等措施可以保证循环水不结垢,也可以...     

Impact of a magnetic ion exchange resin on ozone demand and bromate formation during drinking water treatment

The objective of this research was to examine the impact of a magnetic ion exchange resin (MIEX) on ozone demand and bromate formation in two different ozonated waters at bench scale. The first raw water had a high bromide ion concentration, a high ozone demand, and was highly colored. Based on experimental findings from the first water, the second water was selected as a model water in which more controlled experiments were performed. The waters were treated with the MIEX resin using jar test procedures to find the optimal MIEX dosage based upon the removal of ultraviolet (UV)-absorbing substances, dissolved organic carbon (DOC), and bromide. The optimal resin dosage was chosen for bulk MIEX treatment and subsequent ozonation in a semi-batch reactor. The ozone demand and formation of bromate were analyzed as a function of ozone dosage and dissolved ozone concentration for the MIEX pre-treated water, and compared to the results obtained by ozonating the water without MIEX pre-treatment. The results indicate that pre-treatment of the water with the MIEX resin significantly reduces total organic carbon, DOC, UV absorbance, color, and to some extent, bromide. MIEX pre-treatment of the water prior to ozonation substantially lowered the ozone demand and formation of bromate during subsequent ozonation.     

Treatment of nanofiltration and reverse osmosis concentrates: comparison of precipitative softening, coagulation, and anion exchange

Disposal and treatment of concentrate from nanofiltration (NF) and reverse osmosis (RO) are major challenges to implementing membrane treatment processes. Intermediate treatment of membrane concentrate, between primary and secondary membrane stages, has the potential to increase membrane recovery rates and decrease the volume of concentrate produced. To achieve this, however, there is a need to better understand treatment of membrane concentrate. As a result, this work systematically evaluated lime softening, ferric sulfate coagulation, and magnetic ion exchange (MIEX) as individual, intermediate treatment processes for membrane concentrate. Six membrane concentrates, from NF and RO, with varying concentrations of calcium, dissolved organic matter (DOM), and sulfate were chosen for this study. Maximum removal of calcium was achieved by lime softening, whereas maximum removals of DOM and sulfate were achieved by MIEX. The results of this work show that intermediate treatment of NF/RO concentrate is capable of producing treated concentrate with water quality approximately equal to the initial source water.     

Mechanism of natural organic matter removal by polyaluminum chloride: effect of coagulant particle size and hydrolysis kinetics

The mechanism of natural organic matter (NOM) removal by AlCl(3) and polyaluminum chloride (PACl) was investigated through bench-scale tests. The fraction distributions of NOM and residual Al after coagulation in solution, colloid and sediment were analyzed as changes of coagulant dosage and pH. The influence of NOM, coagulant dose and pH on coagulation kinetics of AlCl(3) was investigated using photometric dispersion analyzer compared with PACl. Monomeric Al species (Al(a)) shows high ability to satisfy some unsaturated coordinate bonds of NOM to facilitate particle and NOM removal, while most of the flocs formed by Al(a) are small and difficult to settle. Medium polymerized Al species (Al(b)) can destabilize particle and NOM efficiently, while some flocs formed by Al(b) are not large and not easy to precipitate as compared to those formed by colloidal or solid Al species (Al(c)). Thus, Al(c) could adsorb and remove NOM efficiently. The removal of contaminant by species of Al(a), Al(b) and Al(c) follows mechanisms of complexation, neutralization and adsorption, respectively. Unlike preformed Al(b) in PACl, in-situ-formed Al(b) can remove NOM and particle more efficiently via the mechanism of further hydrolysis and transfer into Al(c) during coagulation. While the presence of NOM would reduce Al(b) formed in-situ due to the complexation of NOM and Al(a).