Article pubs.acs.org/est
Bioregeneration of Spent Anion Exchange Resin for Treatment of Nitrate in Water Xiaoyang Meng,† David A. Vaccari,† Jianfeng Zhang,‡ Antonio Fiume,† and Xiaoguang Meng*,† †
Center for Environmental Systems, Stevens Institute of Technology Hoboken, New Jersey 07030, United States School of Environmental and Municipal Engineering, Xi’an University of Architecture and Technology, Xi’an 710055, China
‡
S Supporting Information *
ABSTRACT: Anion exchange resin treatment is a commonly used technique for removal of nitrate from water. However, spent anion exchange resins are themselves regenerated using brine solution, which produces spent solution containing a high concentration of nitrate and salt. The present study developed a bioregeneration technique for conversion of nitrate on the spent resins to nitrogen gas while eliminating the use of brine solutions. Batch experiments were conducted to investigate the effect of biomass content, pH, salinity, and molar ratio of exogenous organic carbon to nitrate on the kinetics of bioregeneration. The bioregeneration rate decreased when pH increased from 7 to 10. It increased with increasing microbial concentration from 8.3 to 13.8 g/L as volatile suspended solid (VSS) and with decreasing conductivity of the regeneration suspension from 31 to 9 mS/cm. Spent exchange resins were effectively regenerated within 5 h under the optimal conditions and the regenerated resins could be used repeatedly for filtration removal of nitrate from water. A desorption−denitrification model was developed to describe bioregeneration kinetics. Modeling results indicated that the bioregeneration was through desorption of nitrate from the spent resin and subsequent denitrification of the soluble nitrate. Denitrification was the rate-limiting process. This research demonstrated the feasibility of using a biological process to regenerate nitrate-saturated resins.
■
INTRODUCTION
Biological denitrification process converts nitrate to nitrogen gas and can be used to treat wastewater with high nitrate concentration.9−11 However, the process requires a long reaction time and thus large reactors with large footprint area, compared to anion exchange resin filtration. A few hours of contact time is required for denitrification treatment while only a few minutes of empty bed contact time is needed in the filtration process. It cannot be used for treatment of wastewater containing chemicals that are toxic to denitrifying bacteria. In addition, biological treatment requires the supply of exogenous organic carbon and nutrients, which are seldom completely consumed. These nutrients remain in the treated water, causing secondary pollution. The bioregeneration technique developed in this research combined the advantages and eliminated some problems associated with exchange resin filtration and denitrification processes. Nitrate-selective resin was used for rapid filtration treatment of water and for concentrating nitrate from a large volume of water into a small volume of resin. Then denitrifying bacteria were used to convert nitrate on the spent resin to nitrogen gas. Because the volume of the spent resin was a few
Nitrate is one of the most common contaminants in groundwater and surface water worldwide.1 Ingestion of nitrate-contaminated water can cause serious health problems such as blue baby syndrome, methemoglobinemia, hypertension, increased infant mortality, goiter, stomach cancer, thyroid disorder, cytogenetic defects, and birth defects.2 Nitrate also contributes to lake eutrophication. To protect human health, the U.S. Environmental Protection Agency has set the maximum contaminant level goal (MCLG) of 10 mg-N/L for nitrate in drinking water. Several techniques have been developed and tested for treatment of nitrate in water, including biological treatment,3 anion exchange resin filtration,4 reverse osmosis,5 chemical denitrification,6 catalytic reduction, 7 and electrodialysis.8 Among them, anion exchange resin filtration and biological treatment are the most commonly used processes. Anion exchange resin filtration systems require short contact time and small footprint area and are easy to operate. However, nitrate is not destroyed in this process. When spent resins are regenerated with brine solution (typically 10−15% NaCl solution), nitrate is released from the resins into the brine solution. Discharge of the spent brine solution into the environment increases the nitrate concentration and salinity of the receiving water, which is increasingly restricted. © 2014 American Chemical Society
Received: Revised: Accepted: Published: 1541
June 6, 2013 December 28, 2013 January 13, 2014 January 13, 2014 dx.doi.org/10.1021/es4043534 | Environ. Sci. Technol. 2014, 48, 1541−1548
Environmental Science & Technology
Article
hundred times smaller than the volume of the filtered water, the size of the bioregeneration reactor would be much smaller than that of denitrification reactor for treatment of water directly. In addition, the biomass suspension could be reused in the bioregeneration process to increase the utilization and reduce the discharge of the exogenous organic carbon and nutrients. The materials would be discharged together with the treated water in the denitrification process. In the present study, suitable biomass content, pH, and salinity for bioregeneration were determined. The effect of molar ratio of exogenous organic carbon to nitrate on bioregeneration rate was evaluated for reduction of residual organic carbon and salt concentrations in the used biomass suspension and for its repeated reuse. The anion exchange resin was bioregenerated and reused repeatedly to evaluate the nitrate removal capacity of the bioregenerated resin. The major soluble nitrogen compounds in the bioregeneration suspension were analyzed to determine the denitrification products. Kinetic equations for nitrate desorption from the spent resin and denitrification of soluble nitrate were determined separately and then integrated to develop a desorption−denitrification model to describe the bioregeneration processes and to determine the rate-limiting step in the bioregeneration processes.
supernatant was decanted and replaced by fresh culture medium. The batch cultivation was repeated continuously for more than 4 weeks until rapid denitrification was achieved. Denitrification Experiments. Denitrification kinetic tests were conducted by adding certain amounts of nitrate and acetate into 500 mL of cultivated biomass suspension. The suspension was controlled at desired pH values using sodium hydroxide and hydrochloric acid solutions. Aliquots of suspension samples were taken at different periods of mixing time, and were centrifuged immediately at 10 000 rpm for 5 min to separate the solution from the solids. Concentrations of NO3−-N, NO2−-N, NH4+-N, and total N in the solution were measured using nitrate test kits (HACH TNT 835/836, 839/ 840, 831/832, 827) and a spectrometer (HACH DR 2800). Bioregeneration of Spent Resin. After the exchange resin (20 mL) was saturated by nitrate in the column filtration tests, it was transferred into a container. Five hundred mL of the cultivated biomass suspension was transferred into the container. After the addition of acetate to reach a CH3COO− to NO3− (on the resin) molar ratio of 2.1:1 and 1.5:1, the container was capped and mixed for bioregeneration of the spent resin. The amount of nitrate on the spent resin was calculated based on the volume of water filtered. The pH of the suspension was controlled at desired values using hydrochloric acid and sodium hydroxide solutions after the samples were taken. During the bioregeneration, 1.5 mL of suspension samples was taken and immediately centrifuged for analysis of residual nitrate, nitrite, ammonium, and total nitrogen in the solution. Biomass content, conductivity, and pH in the regeneration suspension were varied to evaluate their effect on the kinetics of bioregeneration. Biomass content was determined by measuring the volatile suspended solids (VSS). The acetate concentration in the bioregeneration suspensions was analyzed by using a spectrophotometric method at a wavelength of 210 nm.13 Standard addition method was used for the analysis to eliminate the matrix effect. Duplicate or triplicate experiments were conducted to confirm that the general trends of effect of pH, salinity, biomass content, and acetate to nitrate ratio on bioregeneration rate were reproducible. One set of representative experimental results was reported for each affecting parameter. Nitrate Desorption Experiments. Stepwise nitrate desorption kinetic experiments were carried out by mixing 5.0 mL of spent resin with 125 mL of biomass supernatant collected after the biomass suspension was washed and used for the first bioregeneration test. The supernatant was used so that the solution composition and concentration in desorption system were the same as those in the bioregeneration suspension. After 20 min of mixing to reach the first batch desorption equilibrium, 1.5 mL of the solution sample was taken for analysis of the equilibrium nitrate concentration. The second batch desorption test was carried by replacing the equilibrated supernatant with a new supernatant. The procedures were repeated 13 more times.
■
MATERIALS AND METHODS Filtration Tests Using Anion Exchange Resin. SBG I anion exchange resin was obtained from Resin Tech Inc. and was used in this study because of its high selectivity for nitrate.12 A small column with an inside diameter of 0.5 in. (1.25 cm) was packed with 20 mL of the resin. A nitrate solution was prepared by adding potassium nitrate into tap water to reach a NO3−-N concentration of 45 mg/L. Filtration tests were carried out by passing the water through the resin column in a downward flow and at a flow rate of 5 mL/min. The empty bed contact time (EBCT = bed volume of resin/ flow rate) was 4 min. Filtered water samples were collected during the tests for analysis of nitrate. Cultivation of a Mixed Culture of Nitrate-Reducing Bacteria. Anaerobic sludge collected from the digester of a wastewater treatment plant in north New Jersey was washed with tap water to remove the coarse solid particles in the sludge. The gravity-settled sludge was subsequently collected and used as the seed of heterotrophic nitrate-reducing bacteria. The seed was then cultivated for the enrichment of nitratereducing bacteria in a covered container. Fresh culture medium containing (per liter) 1.44 g of NaH2PO4, 0.1 g of (NH4)2SO4, 0.1 g of MgSO4, 4.0 mg of FeSO4, 0.6 mg of Na2MoO4·2H2O, 1.0 mg of NaSeO3, and 0.6 mg of H3BO3 was prepared and adjusted to pH 7.0 using 1.0 M NaOH and was purged with oxygen-free nitrogen. During each cycle of batch cultivation, the seed was mixed with the culture medium, which was spiked with 1600 mg/L acetate (27.1 mM of CH3COO−) and 180 mg/L NO3−-N (12.9 mM of NO3−). The molar ratio of CH3COO− to NO3− was 2.1:1, which was much higher than the stoichiometric ratio of 1.25:1 as described in the following denitrification reaction using sodium acetate as a carbon source (and neglecting cell yield):
■
RESULTS AND DISSCUSSION Filtration Removal of Nitrate. Figure 1 shows filtration breakthrough curves for the virgin resin and bioregenerated SBG I resin. Because nitrate exchange with chloride ions on the anion exchange resin occurred rapidly, a short EBCT of 4 min for water in column was used. Nitrate was completely removed by the virgin resin column in the first 360 bed volumes of the water treated. After the resin was saturated by nitrate at about
5CH3COO− + 4NO3− + 4H+ → 2N2 + 5HCO3− + 7H 2O
After 1−3 days of mixing in each cycle, the biomass was allowed to settle for about an hour. About two-thirds of the 1542
dx.doi.org/10.1021/es4043534 | Environ. Sci. Technol. 2014, 48, 1541−1548
Environmental Science & Technology
Article
suspension. The amount of the adsorbed nitrate was calculated based on filtration results from Figure 1. The soluble nitrate concentration decreased to zero in the pH 8 and 9 suspensions after 330 min of bioregeneration. However, nitrate concentration was only reduced from 95 to 77 mg/L at pH 10 after 540 min of reaction, which suggested that denitrification was inhibited at high pH. Supporting Information (SI) Figure S1 shows that denitrification rate decreased with increasing pH from 7 to 10, which was consistent with the reported inhibition effect of high pH on denitrification.9,10,16−22 The optimal pH for denitrification treatment of wastewater was reported to be in a range between 7.5 and 8.6.9,23 Based on the results in Figures 2 and SI S1 and the reported optimal pH range, the suspension pH should be controlled below about 9 to ensure rapid bioregeneration of the spent resin. When nitrate concentration in the solution was below the HACH method detection limit of 1 mg/L, little nitrate should have remained on the resin. Therefore, after about 400 min of bioregeneration at pH ≤ 9, the resin was separated from the suspension by pouring the suspension out of the container. Because the specific gravity of the resin was higher than the biomass, it remained at the bottom of the container. The separated biomass was used for regeneration tests again. The effect of salinity on the bioregeneration rate was evaluated because salt concentration increased during bioregeneration and high salt concentration could inhibit denitrification.22 During bioregeneration and cultivation, the conductivity of the biomass suspension increased gradually mainly due to the accumulation of sodium and bicarbonate in it. Sodium was added into the suspension with sodium acetate and acetate was converted to bicarbonate by the denitrification reaction. The biomass suspension was cultivated several times without washing after each cultivation cycle to allow the salt to accumulate in it for evaluation of salinity effect on bioregeneration. A CH3COO− to NO3− molar ratio of 2.1:1 was used in most tests to ensure that the carbon concentration was not the rate-limiting factor for the denitrification process.18,23−26 Figure 3 shows the effect of salinity on bioregeneration process. It is interesting to note that the peak nitrate
Figure 1. Nitrate breakthrough curves for column filtration of nitratespiked tap water using virgin and bioregenerated resins. Initial NO3−N concentration C0 = 45 mg/L, resin volume = 20 mL, filtration flow rate = 5 mL/min.
400 bed volumes, effluent nitrate concentration increased dramatically. The completely saturated resin was used in the bioregeneration tests. Effect of pH, Conductivity, Biomass Content, and Carbon Content on Bioregeneration Kinetics. During microbial facilitated reduction of nitrate, it is ultimately converted to nitrogen gas through a series of intermediate products, including nitrite (NO2−), and gaseous nitric oxide (NO) and nitrous oxide (N2O).9,14,15 Because the biological suspension pH could be increased to greater than 9.5 as the results of proton consumption,9,10 the effect of pH on bioregeneration rate was evaluated. Figure 2 shows the soluble nitrate concentration in the bioregeneration suspensions as a function of reaction time. The
Figure 2. Effect of pH on bioregeneration: CH3COO− to NO3− molar ratio = 2.1:1, room temperature, conductivity ≈ 20 mS/cm, biomass content ≈ 11.4 g/L.
pH of three suspensions was controlled at 8, 9, and 10, respectively. When the spent resin was mixed with the biomass suspension, soluble NO3−-N concentration increased from zero to greater than 90 mg/L within 10 min followed by gradual decrease in nitrate concentration. The results suggested nitrate was released rapidly from the resin to form soluble nitrate and denitrification occurred relatively slowly during the bioregeneration. A mass balance calculation indicated that NO3−-N concentration would reach about 677 mg/L if all nitrate on the 20 mL of spent resin was released into 500 mL of the
Figure 3. Effect of conductivity on bioregeneration: pH = 7.5−8.5, biomass content ≈ 11.4 g/L.
concentration increased with increasing conductivity, which indicated that more nitrate was released from the spent resin by high concentration of solutes through anion exchange. Nitrate was adsorbed on the anion exchange resin through electrostatic attraction to the quaternary amine (R-N-(CH3)3+) groups on 1543
dx.doi.org/10.1021/es4043534 | Environ. Sci. Technol. 2014, 48, 1541−1548
Environmental Science & Technology
Article
The bioregeneration was completed in 480 min, which was 120 min longer than the time required for complete bioregeneration in the system with a molar ratio of 2.1:1 and with a similar VSS content of 11.4 g/L (Figure 4). It should be noted that the conductivity of the bioregeneration suspension was about 5 mS/cm when the ratio was 1.5:1, which was much lower than the conductivity (about 9 mS/cm) in the suspension with a ratio of 2.1:1. When the initial acetate to nitrate ratio was 1.5:1, acetate concentration was reduced from 1.6 g/L in the initial biomass suspension to 0.32 g/L after the bioregeneration. The acetate concentration was only reduced from 2.66 to 0.97 g/L during the bioregeneration when the ratio was 2.1:1. The results indicated that the residual acetate concentration could be reduced significantly at an acetate to nitrate ratio of 1.5:1. Because the conductivity of the used biomass suspension was only about 5 mS/cm, it could be reused repeatedly several times to reduce the discharge of acetate. Only enough amounts of acetate needed to be added into the used bioregeneration suspension containing the residual acetate to make the acetate to nitrate ratio to 1.5:1 before subsequent bioregeneration. The concentrations of major nitrogen compounds in the solution were measured to determine the bioregeneration products and completion of denitrification. Figure 5a shows
the resin. It was released from the spent resin through ion exchange with bicarbonate, carbonate, and other anions in the solution during the bioregeneration. The total carbonate concentration in the biomass suspension increased from approximately 500 to 2000 mg/L after the bioregeneration due to the oxidation of acetate. Rokicki and Boyer 27 observed that NaHCO3 solution had similar efficiency for regeneration of nitrate-saturated anion exchange resins as NaCl solution. Nitrate concentration decreased to less than 5 mg/L after 250 min of bioregeneration when the conductivity was 9 and 15 mS/cm. Denitrification rate decreased dramatically when the conductivity increased from 15 to 31 mS/cm. Salinity affected both nitrate release from the resin and activity of denitrifying bacteria. Whereas high soluble nitrate should favor the denitrification, high salinity would inhibit the biological activity. The results indicated that the salinity inhibition effect outweighed the enhanced nitrate release effect when the conductivity was higher than 15 mS/cm. When the conductivity was 15 and 31 mS/cm, salt concentration was approximately 9 and 20 g/L, respectively. Denitrification can be inhibited significantly when salt content is greater than 10 g/L or about 1%, although acclimated activated sludge culture has been used to achieve efficient denitrification at higher salt content. 10,22 The relatively rapid denitrification at the conductivity of 21 mS/cm (Figure 3) suggests that the denitrifying bacteria were still very active at relatively high salt concentration because the biomass was cultivated repeatedly at relatively high salt conditions. The salt concentration in the biomass suspension was reduced by removing the supernatant in the settled suspension and adding aged tap water into the settled biomass in this research. The tap water was aged for a day to remove the free chlorine. Figure 4 shows that biomass content had a significant effect on the bioregeneration rate. When the VSS was 8.3 g/L, it took
Figure 5. Concentration variation of (a) measured and (b) calculated other nitrogen compounds during bioregeneration.
that both nitrate and total nitrogen concentrations increased rapidly to reach the maximum at the same time and then decreased gradually. The total nitrogen concentration was higher than nitrate-N concentration, indicating the formation of nitrogen products other than N2 gas. A small amount of nitriteN was formed as an intermediate product. It reached a maximum concentration of 0.8 mg-N/L at 180 min and decreased to 0.2 mg-N/L at the end of the bioregeneration. Ammonium-N was detected mainly because the culture medium in the biomass suspension contained 0.1 g/L of (NH4)2SO4 or 21 mg/L of NH4+-N. The ammonium-N concentration decreased gradually to 7 mg/L at the end of the bioregeneration, which could be attributed to ammonium incorporation in biomass. Some ammonium could also be converted to dinitrogen gas with nitrite as an electron acceptor under anoxic conditions (NH4+ + NO2− → N2 + 2H2O) through anaerobic ammonium oxidation (anammox).
Figure 4. Effect of biomass content and molar ratio of acetate to nitrate on bioregeneration: molar ratio = 2.1:1 and conductivity ≈ 20 mS/cm in the first three systems; VSS ≈ 11 g/L and conductivity ≈ 5 mS/cm in the last system; pH = 7.5−8.5.
660 min to reduce the soluble nitrate concentration to zero in the bioregeneration suspension. The time required for complete bioregeneration of the spent resins was reduced to about 240 and 340 min when the VSS content was increased to 11.4 and 13.8, respectively. The high bioregeneration rate was attributed to the improved denitrification at high biomass content.17 To reduce the discharge of acetate, bioregeneration tests were conducted at a low acetate to nitrate molar ratio of 1.5:1. 1544
dx.doi.org/10.1021/es4043534 | Environ. Sci. Technol. 2014, 48, 1541−1548
Environmental Science & Technology
Article
To determine the amounts of other denitrification products in the solution, the difference between the concentrations of total N concentration and total concentration of the three measured nitrogen compounds was calculated. The results in Figure 5b show that the total N concentration was higher than the sum of nitrate-N, nitrite-N, and ammonium-N concentrations, which indicated the formation of other denitrification products, such as nitric oxide (NO) and nitrous oxide (N2O).9,14,15 When the bioregeneration was completed at 360 min, no NO and N2O existed in the solution (Figure 5b). The results in Figure 5 indicate that majority of the nitrate was converted to N2 gas by the bioregeneration since some of the NO and N2O might be released into the headspace of the reactor. After evaluation of the bioregeneration conditions, the spent resins were regenerated at pH 7.5−8.5, conductivity under 20 mS/cm, biomass content of 11.4 g/L, and acetate to nitrate molar ratio of 2.1:1 and 1.5:1. The regenerated resins were reused for column filtration tests. The results in Figure 1 demonstrate that the bioregenerated resins could be reused repeatedly. The amount of water treated decreased from about 400 to 350 bed volumes after the resin was regenerated and reused 5 times, which could be attributed to the accumulation of some solute and biomass on it. It might be cleaned by soaking the resin in dilute HCl solution for several h. Fluidization tests presented in the SI indicated that the spent resin in the column could be bioregenerated by pumping the biomass suspension through the column in an upward flow and the biomass could be separated from the resin at a flow velocity of 19 cm/min. Modeling Bioregeneration Process. Various denitrification kinetics models have been developed to describe the removal of soluble nitrate in water.16−19,28 However, nitrate ions associated with the binding sites on the spent resin, especially those in the pores of the resin beads, are not readily available to bacteria. Based on the observation of initial rapid increase and then gradual decrease of soluble nitrate during the bioregeneration (Figures 2−4), it was assumed that the bioregeneration process consisted of two steps, including (a) nitration desorption from the spent resin, and (b) denitrification of the soluble nitrate. A desorption−denitrification model was developed to describe the bioregeneration process. Independent modeling parameters were determined from the nitrate desorption kinetics and isotherm data and denitrification kinetics data obtained under conditions similar to those used in the bioregeneration tests for the desorption−denitrification modeling calculations. Desorption kinetics results in Figure 6A showed that nitrate was released rapidly from the spent resin and reached equilibrium within 10 min. The NO3−-N concentration in equilibrium with the saturated resin (Se) was 159 mg/L. Though the same resin to liquid volume ratio was used in both desorption and bioregeneration systems, the desorption equilibrium nitrate concentration was higher than the peak nitrate concentration in the bioregeneration suspensions in Figures 2−4, which was mainly attributed to the denitrification in the bioregeneration suspension. A first order desorption kinetic equation (eq 1) was used to fit the batch desorption data in Figure 6A.29 rdesorption = Kdes(Se − S)
Figure 6. Nitrate desorption and the best model fitting: (A) kinetics; (B) isotherm. Experimental conditions: resin to biomass supernatant volume ratio = 5:125 (g/L), pH = 7−9, conductivity
