Available online at www.sciencedirect.com Environmental Pollution 153 (2008) 37e43 www.elsevier.com/locate/envpol Formation of aerobic granules in the presence of a synthetic chelating agent Yarlagadda V. Nancharaiah*, Hiren M. Joshi, Tulsi V. Krishna Mohan, Vayalam P. Venugopalan, Sevilimedu V. Narasimhan Water and Steam Chemistry Division, Chemistry Group, Bhabha Atomic Research Centre, BARC Facilities, Kalpakkam 603 102, India Received 9 November 2007; accepted 11 November 2007 Synthetic chelating agent enhances aerobic microbial granulation. Abstract This paper examines the development of aerobic granular sludge in the presence of a synthetic chelating agent, nitrilotriacetic acid (NTA), in sequencing batch reactors (SBR). The growth of seed sludge at 0.26 mM, 0.52 mM and 1.05 mM of NTA was found to be significantly lower as compared to that in the absence of NTA. Aerobic granulation was significantly enhanced in the three SBRs (R2, R3 and R4), which were fed with 0.26 mM, 0.52 mM and 1.05 mM of NTA as a co-substrate, in comparison to the acetate-alone fed SBR (R1). After 2 months of operation, the mean diameter of the biomass stabilized at 0.35 mm in R1 (acetate alone), as compared to 2.18 mm in R4 (1.05 mM NTA þ acetate). NTA degradation was established in SBRs, with almost complete removal during the SBR cycle. Batch experiments also showed efficient degradation of NTA by the aerobic granules. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Aerobic granules; Aerobic granular sludge; Enhanced microbial granulation; NTA; Synthetic complexing agent 1. Introduction Immobilization of microorganisms into biofilms or granules, a process of cell-to-substratum or cell-to-cell attachment, is extensively employed in biotechnological applications (Liu and Tay, 2004; Nancharaiah et al., 2006a). Formation of microbial granules from microbial flocs under aerobic conditions is currently an active area of investigation for developing new generation wastewater treatment plants (Nicolella et al., 2000; Morgenroth et al., 1997; de Kreuk et al., 2005). Microbial granules cultivated under aerobic conditions are generally less stable as compared to anaerobic granules (Morgenroth et al., 1997; Liu et al., 2004). The poor stability of microbial granules can be a limiting factor in their application in real wastewater treatment practice (Liu et al., 2004). A possible * Corresponding author. Tel.: þ91 44 27480 203; fax: þ91 44 27480 097. E-mail address: yvn@igcar.ernet.in (Y.V. Nancharaiah). 0269-7491/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2007.11.017 reason for the poor stability of aerobic granules is the fast growth rate of heterotrophic bacteria, which dominate the granules (Liu et al., 2004). Earlier work on aerobic granulation indicated that the selection of slow growing bacteria would lead to the formation of stable microbial granules (de Kreuk and van Loosdrecht, 2004; Liu et al., 2004). The presence of slowly degrading compounds would slow down the growth of bacteria and may ultimately result in selection of slow growing strains. Therefore, it was hypothesized that the presence of a carbon source that is relatively difficult to degrade might facilitate better granulation due to the selection of slow growing bacteria, resulting in the formation of stable granules. For the purpose of the study, acetate and a synthetic chelant were chosen as a model labile and refractory carbon sources, respectively. Synthetic chelating agents are used in wide range of applications including detergent, food, pharmaceutical, cosmetic, metal-finishing, photographic, textile, paper and nuclear power 38 Y.V. Nancharaiah et al. / Environmental Pollution 153 (2008) 37e43 industries. For example, chemical decontamination process in nuclear power reactors uses one or a mixture of chelating agents such as ethylenediaminetetraacetic acid (EDTA), nitrilotriacetic acid (NTA), picolinic acid and citric acid (White and Knowles, 2000). Co-disposal of heavy metals or radionuclides along with synthetic organic chelating agents creates environmental problems because the latter may promote undesirable displacement of toxic heavy metals/radionuclides away from the primary disposal site (Bolton et al., 1996). It is desirable to remove the chelating agents from the wastes prior to disposal. As a result, there is significant research interest in developing effective biological treatment process for degrading synthetic chelating agents (White and Knowles, 2000; Bucheli-Witschel and Egli, 2001). Therefore, a typical synthetic chelating agent such as NTA was used as substrate in this study. Accordingly, the major objective of this work was to test the hypothesis that presence of NTA as substrate would favour slower growth of microorganisms in a bioreactor and lead to rapid formation of stable granular sludge. Concurrently, biodegradation of NTA using mixed species granular sludge was studied in laboratory SBRs and batch experiments. 2. Materials and methods 2.1. Operation of SBRs Four (R1, R2, R3 and R4) laboratory scale column type SBRs made of glass and with a working volume of 3 L each were used for the cultivation of aerobic microbial granules. The dimensions of the SBRs used were: height 1200 mm and diameter 62 mm and height to diameter ratio 19.4. The SBRs were operated with a 6-h cycle time and 66% volumetric exchange ratio. Vigorous aeration was provided in the reactors from the bottom, using air dispensers and an airflow rate of 3 L min1 (superficial upflow air velocity of 1.7 cm s1). The chemical composition of synthetic wastewater (SWW) prepared using tap water and used as feed in the SBR is given in Table 1. The acetate concentration was varied in the four reactors in order to maintain uniform total influent chemical oxygen demand (COD) in all reactors. Nitrilotriacetic acid (Loba Chemie, India) was prepared as a stock solution (1 g l1) and its pH was adjusted to 7.5 with NaOH. Addition of the influent (SWW with or without NTA) and the withdrawal of the effluent were accomplished with the help of peristaltic pumps, activated through preset electronic timers. The SBRs were operated at room temperature (30e31  C) in sequencing mode, with 6 h cycle time consisting of 60 min fill, 282 min aeration, 3 min settling, 10 min effluent draw and 5 min idling. The dissolved oxygen (DO) concentration in the reactors during the aeration phase was observed to be about 7.5 mg l1. The pH was not controlled but was observed to vary from 7.6 to 8.2 during the course of a cycle. Changes in biomass content and sludge volume index (SVI) in the SBRs were monitored during granule development, following standard methods (APHA, 1995). SVI after 10 min or 30 min was used for describing the settleability of flocs and granules. Table 1 Composition of synthetic wastewater (SWW) used in the four sequencing batch reactors Chemical R1 R2 R3 R4 Acetate (mM) Nitrilotriacetate (mM) MgSO4$7H2O (mM) KCl (mM) NH4Cl (mM) K2HPO4 (mM) KH2PO4 (mM) 90 0 3.6 4.7 35.4 4.2 2.1 84 0.26 3.6 4.7 35.4 4.2 2.1 78 0.52 3.6 4.7 35.4 4.2 2.1 66 1.05 3.6 4.7 35.4 4.2 2.1 2.2. Seed sludge The SBRs were slug-inoculated with 800 ml of wastewater containing activated sludge flocs collected from a municipal wastewater treatment plant at Kalpakkam (South India). The seed sludge characteristics were: mixed liquor suspended solids (MLSS) 840 mg l1 and sludge volume index (SVI)30 min 278 ml g1. For growth experiment, freshly collected activated sludge was grown in the laboratory overnight in the SWW containing acetate as the sole carbon source. 2.3. Growth of seed sludge in presence of NTA Growth experiments were performed in graduated glass cylinders (diameter: 7.6 mm; height: 460 mm) having 1 L working volume. SWW was prepared in ultrapure water as per the details given in Table 1. Trace elements were introduced by adding 0.1 ml stock solution per liter medium. Trace element stock solution consisted of ZnSO4$6H2O (5 mM), MnCl2$4H2O (5 mM), CoCl2$4H2O (8 mM), CuCl2$2H2O (1 mM), NiCl2$6H2O (1 mM), and NaMoO4$2H2O (1.5 mM). The SWW was inoculated with activated sludge (5% v/v) and incubated at room temperature (30e31  C). The glass cylinders were aerated by introducing air at the bottom at a flow rate of 2 L min1 with the help of air dispensers. Bacterial growth was monitored by measuring absorbance at 600 nm and colony forming units (CFU) at different time points. Samples were serially diluted in phosphate buffered-saline, spread plated on to R2A agar (Difco). The colonies were counted after 2 days of incubation at room temperature. Three independent experiments were carried out and percentage reduction in growth in presence of NTA was calculated. Data were plotted as mean of the three experiments. The microbial growth in the presence of NTA was compared to the growth in the absence of NTA. Growth studies were also carried out using media containing equal amount of acetate (90 mM) but different concentrations of NTA or providing variable amounts of NTA as the sole carbon source. 2.4. Biodegradation experiments Biodegradation experiments using NTA were performed in graduated glass cylinders having 1 L working volume. Fresh granules collected from SBR were used in experiments dealing with different initial concentrations of free NTA biodegradation. The degradation studies were carried out using synthetic wastewater prepared in sterile ultrapure water, with NTA added as the sole source of carbon, nitrogen and energy. Pre-cultivated (see above) microbial granules (40 ml) were introduced into 1 L for the degradation studies. Reactors containing NTA without granules were used as control. Oxygen was introduced at the bottom of experimental glass cylinders at a flow rate of 2 L min1 by using an air dispenser. Aliquots were withdrawn periodically, filtered through 0.22 mm filters (Millex GS, Millipore) to remove suspended solids and analysed for NTA. Control reactors containing no biomass were operated in parallel in order to exclude NTA-removal mechanisms other than biodegradation. 2.5. Microscopy and image analysis Morphogenesis of the microbial sludge was periodically monitored with the help of a stereozoom microscope (Nikon SMZ1000). The images acquired using a digital camera (Olympus DP70) were processed and analysed using the freeware ImageJ 1.33x (downloadable from the site http://rsb.info.nih.gov/ij), for calculating mean granule size and circularity. Prior to quantification, the images were interactively thresholded and binarised. A minimum detection limit of 60 mm was set for calculating the mean granule size (that is, granules less than 60 mm were ignored by the programme). The biomass collected from the SBRs was imaged directly after sampling without separating flocs and granules. Microstructure of the seed sludge and granules was visualized using a confocal laser scanning microscope (CLSM) (model Leica TCS SP2 AOBS, equipped with an inverted microscope DMIRE2). For confocal imaging, the microbial granules were stained with 0.01% acridine orange for 15 min and thoroughly rinsed with phosphate buffered-saline for 15 min. The stained individual microbial granule was mounted on a cover slip and imaged using 39 a 63  1.2 NA water immersion objective. The 488-nm line from an argon laser was used for excitation, the emission was collected by setting the detection bandwidth between 510 nm and 550 nm. a 2.6. NTA analysis Optical Density at 600 nm Y.V. Nancharaiah et al. / Environmental Pollution 153 (2008) 37e43 NTA was analysed by spectrophotometry as described previously (Nancharaiah et al., 2006b). Briefly, the samples were incubated with 10 mM copper sulphate resulting in the formation of coppereNTA complex. The samples were filtered through 0.22 mm filter and the absorbance of the filtrate was measured at 305 nm using a UVevisible spectrophotometer (Shimadzu, Japan). The absorption of coppereNTA complex showed a linear relationship between 0.1 mM and 3.0 mM of NTA. The minimum detection limit of the photometric assay was 0.1 mM. 0.18 0 mM NTA 0.16 0.26 mM NTA 0.52 mM NTA 0.14 1.05 mM NTA 0.12 0.10 0.08 0.06 0.04 0.02 2.7. Statistical analysis 0.00 Growth rates of seed sludge under different concentrations of NTA were compared using one way analysis of variance (ANOVA), followed by StudenteNewmaneKeuls (SNK) post-test. Differences were considered significant at P < 0.05. 0 1 2 3 4 5 6 Time (h) b 0.18 0.16 3.1. Effect of NTA on growth rate of seed sludge The experiment was performed in 1 L glass cylinders with upflow aeration from the bottom in order to simulate SBR condition. The experimental duration was fixed as 6 h (after inoculation) as the SBRs used for granulation had 6 h cycle period. The growth curves of the seed sludge in the presence of different concentrations of NTA are shown in Fig. 1a. The growth of seed sludge at 0.26 mM, 0.52 mM and 1.05 mM of NTA after 6 h was significantly lower as compared to that in the absence of NTA (P < 0.0001). Fig. 1b shows the growth of seed sludge (after 6 h of inoculation) in the presence of (1) different concentration of NTA as well as acetate, but with uniform COD levels, (2) fixed amount (90 mM) of acetate but different levels of NTA and (3) only acetate or NTA in the medium. Presence of NTA significantly reduced the growth on acetate, while in the absence of acetate, NTA supported no observable growth of the seed sludge during the 6 h experiment. No significant differences in the growth of the seed sludge were observed when the NTA concentration in the medium was varied. Moreover, the concentration of NTA in the medium did not change during the course of the growth experiment (data not shown). The growth rate of seed sludge was found to be 1.9 h1on acetate alone. Reduced growth rate of 1.76 h1, 1.73 h1 and 1.73 h1 were observed at 0.26 mM, 0.52 mM and 1.05 mM of NTA, respectively. Speciation of the NTA present in the media was predicted with the help of CHEAQS, a freeware (provided by Wilko Verweij and downloadable from http://home. tiscali.nl/cheaqs/). It showed that at pH 7.6, it mostly exists as magnesiumeNTA complex (50%) and HNTA (45%). 3.2. Microbial granulation in the presence of NTA The seed sludge used for reactor inoculation had a mean floc size of 60 mm. The seed sludge primarily consisted of Optical Density at 600 nm 3. Results 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 2 1 Acetate 0.26 mM NTA 3 0.52 mM NTA 1.05 mM NTA Fig. 1. Growth of seed sludge in the presence of acetate and different concentrations of nitrilotriacetic acid. (a) Growth curves in media as shown in Table 1. (b) Growth of seed sludge in media with (1) different concentrations of acetate and NTA (as mentioned in Table 1), (2) 90 mM acetate and different concentrations of NTA, (3) acetate alone or NTA alone as sole carbon source. Data obtained 6 h after inoculation are shown. Each data point is the mean of three independent experiments. Error bar represents one standard deviation. activated sludge flocs, which had fluffy and irregular threedimensional structure. The average SVI10 min and SVI30 min of the seed sludge were 278 ml g1 and 183 ml g1, respectively. Confocal microscopic observation showed that the seed sludge was dominated by filamentous bacteria (Fig. 2), along with rod and cocci shaped bacteria. During reactor operation, the seed sludge slowly transformed into granular form, as evidenced by changes in morphology and SVI. The colour of the seed sludge also changed from light black to yellow to brown during the granulation process. Filamentous microorganisms were observed up to 10 days of reactor operation and were not seen thereafter. Microbial granules appeared in the second week of operation in all the four reactors. Fig. 3a, b shows the changes 40 Y.V. Nancharaiah et al. / Environmental Pollution 153 (2008) 37e43 Fig. 2. Photomicrograph of seed sludge used in this study. Image is a maximum intensity projection consisting of 14 xy-confocal slices obtained at 2 mm zinterval. Scale bar ¼ 20 mm. in the mean size of bioflocs and biomass content in R1 and R4 over 2 months of operation. The mean size of biomass in R1 steadily increased and stabilized at about 0.35 mm (Fig. 3a). The biomass concentration decreased during the initial period a Biomass mean size (mm) 2.0 R1 R4 1.5 1.0 0.5 0.0 0 10 20 30 40 50 60 70 Reactor operation(days) Biomass concentration (g SS l-1) b 2.5 of operation (possibly due to washout) and thereafter increased steadily and reached a value of 1.2 g SS l1 after 60 days of operation (Fig. 3b). In the case of R4, the biomass concentration stabilized at about 2.3 g SS l1 and the increase in biomass coincided with an increase in the mean size of granules (Fig. 3a). In reactors run with NTA (R2, R3 and R4), the mean size of granules increased steadily and stabilized at about 1.1 mm, 1.25 mm and 1.6 mm, respectively, while in R1 the mean size of granules was only 0.35 mm. In R1, the biomass was mostly dominated by flocculent sludge with granules forming a minor component, while in R2, R3 and R4, granules became the dominant form of biomass and flocs constituted a minor fraction (Fig. 4). Morphological features of the granular sludge in the four reactors after 27 days of operation are shown in Fig. 4. It is clear that flocs were coexistent with granules in all the four reactors. Nevertheless, the microscopic images show that there are remarkable differences in the morphological features of the biomass among the four reactors. Fig. 5 shows the size distribution of biomass in the four reactors after 27 days of operation. Approximately 50% of the biomass in R1 was in the range of 0.02e0.1 mm, representing mostly flocculent sludge and the rest was more than 0.1 mm in size, representing small granules (also see Fig. 4). In contrast, the biomass size distribution in R4 was completely different with about just 10% of the sludge having size less than 0.1 mm, representing flocculent sludge and 90% having size above 0.1 mm, representing granules (Fig. 5). SVI values showed that the settleability and compactness of biomass were excellent in the case of granules formed in presence of NTA. After 2 months of operation, the average SVI10 min and SVI30 min of biomass in R1 were 176 ml g1 and 168 ml g1, respectively. In R4, the average SVI10 min and SVI30 min of biomass were the same e 43 ml g1, indicating maturity of the granules. Confocal microscope images of the granules from the four reactors showed that the granules consisted mostly of rod and coccoid shaped bacteria (Fig. 6). Filamentous microorganisms present during the initial stages of the granulation process (Fig. 2) were absent in the mature granules. Optical sectioning revealed several cell clusters (microcolonies) of nearly spherical shape consisting of tightly packed rod or cocci shaped bacteria within the microbial granules (Fig. 6b). R1 R4 2.0 3.3. Biodegradation of chelating agent by microbial granules 1.5 1.0 0.5 0.0 0 10 20 30 40 50 60 70 Reactor operation(days) Fig. 3. Changes in mean biomass size and biomass concentration as a function of run time of reactors R1 (with no NTA) and R4 (with 1.05 mM NTA). (a) Mean biomass size and (b) Biomass concentration. Biodegradation of NTA by aerobic granules is presented in Fig. 7. It is clear from the data that removal of influent NTA is almost complete during the 6 h SBR cycle period (Fig. 7a). In batch experiments, the biodegradation of NTA was almost complete in about 16 h (Fig. 7b). An initial lag phase was quite evident during the first cycle of NTA degradation, which disappeared in the second cycle, indicating microbial adaptation to NTA (Fig. 7b). Moreover, degradation was faster during the second cycle of operation with same granules, indicating adaptation. Y.V. Nancharaiah et al. / Environmental Pollution 153 (2008) 37e43 41 Fig. 4. Morphology of the granular sludge in the four reactors R1, R2, R3 and R4. The biomass (sampled on 27th day) sampled from reactors was imaged directly without separating granules from flocs (scale bar ¼ 1 mm). 4. Discussion R1 R2 R3 R4 Percentage frequency 60 50 40 30 20 10 0 0 1 2 3 4 5 6 Diameter (mm) Fig. 5. Patterns of biomass size distribution in the four sequencing batch reactors. Biomass was collected from the reactors on the 27th day of operation and used for determining size distribution. Better microbial granulation observed in presence of NTA may possibly be explained in the following way. First, it is possible that NTA present in the SWW used for cultivation of the granules acted as selective pressure for strain enrichment. Apart from NTA, only acetate was present as a carbon source. As a carbon source, acetate is readily degradable and would be degraded in the first 30 min of the aeration phase (Beun et al., 2000). This would mean that for the major part of the aeration period, only NTA would be available as carbon source. This might lead to the enrichment of slow growing microorganisms in NTA-fed reactors. Therefore, NTA could have provided the selective pressure needed to reduce the overall growth rate, which could have ultimately led to the formation and predominance of stable granular sludge in the NTA-fed SBRs. The degradation of NTA in SBRs indicates the enrichment of NTA utilizing bacteria. Speciation is of crucial importance in discussing the behaviour of complexing agents in the environment (Nowack, 2002). Analysis showed that the majority of the NTA in the SWW in the three SBRs (R2, R3 and R4) was present as magnesiumeNTA or HNTA. NTA was not 42 Y.V. Nancharaiah et al. / Environmental Pollution 153 (2008) 37e43 Fig. 6. Microstructure of microbial granules as revealed by confocal laser scanning microscopy. (a) A maximum intensity projection made from 17 xy-optical sections imaged at 2.0 mm z-interval. (b) A 2 mm thick xy-confocal slice collected at a depth of 13 mm from the edge of an individual aerobic granule showing microcolonies consisting of rod shaped bacteria. Scale bar ¼ 10 mm. a 1.1 R3 R4 1.0 NTA concentration (mM) 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1 0 3 2 5 4 6 SBR Cycle period (hours) b 2.0 Cycle 1 Cycle 2 NTA concentration (mM) 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 2 4 6 8 10 12 14 16 18 Time (hours) Fig. 7. Biodegradation of NTA by aerobic microbial granules. (a) NTA degradation during SBR cycle period in laboratory scale SBRs observed on day 20. (b) NTA degradation in 1 L batch experiments. Microbial granules collected from R4 were used in batch experiments. Each data point is a mean of two measurements. Error bars represent standard deviation. removed during the growth experiment when supplied as sole carbon source or co-substrate along with acetate. However, its presence in the medium significantly decelerated the growth of bacteria and overall biomass yield (Fig. 1a). The characteristic smooth surface of aerobically grown granules is achieved when the outgrowth of biofilm surface is properly balanced by detachment forces (de Kreuk and van Loosdrecht, 2004). In granular reactors, these forces are mainly contributed by hydrodynamic shear and erosion due to granuleegranule collision (Gjaltema et al., 1997). de Kreuk and van Loosdrecht (2004) pointed out that extra operating conditions must be created in granular sludge reactors to reduce the growth rate of microorganisms and to obtain stable microbial granules. Chen and Stewart (2000) reported significant biofilm detachment from a substratum by metal-complexing agents. It is possible that NTA, being a strong metal chelating agent, could have exerted strong detachment force on granule surface. This would lead to the removal of outgrowth on granule surface leading to smoother and dense granule formation as seen in reactors R2, R3 and R4 (Fig. 3). In earlier experiments, we had observed that enhanced microbial granulation was feasible in the presence of EDTA in the medium, despite the fact that EDTA was not degraded by the biomass during SBR operation (Nancharaiah, unpublished data). Although chelating agents have been shown to dislodge biofilms at higher concentrations (Banin et al., 2006), it is likely that they behave differently at lower concentrations, probably enhancing biogranulation. The study underlines the need for investigations on the possible role of chelating agents at relatively lower concentrations (as opposed to those used in biofilm disruption studies) on bacterial aggregation and granulation. Aminopolycarboxylic acid (APCA) type of chelating agents (e.g. NTA and EDTA) and their metal complexes are usually recalcitrant and resist microbial degradation (Bucheli-Witschel and Egli, 2001; White and Knowles, 2000). Nevertheless, NTA biodegradation has been reported in environments including soil, river water and activated sludge (Bucheli-Witschel and Egli, 2001). In this study, NTA present in the synthetic wastewater used for cultivation of the granules Y.V. Nancharaiah et al. / Environmental Pollution 153 (2008) 37e43 very likely acted as selective pressure for strain enrichment. The multispecies aerobic granules formed in this study were capable of degrading different initial concentrations of free NTA in SBRs and in batch experiments. Biodegradation of NTA was observed in all the three SBRs. In the batch experiments, granules were used to degrade higher initial concentrations (1.8 mM) of NTA. Successful microbial granulation concomitant with the degradation of NTA in sequencing batch reactors suggests the possibility of developing granule-based SBR system for treatment of wastes-containing NTA. 5. Conclusions Presence of NTA in the medium significantly reduced the growth of seed sludge used as seed sludge for sequencing batch reactors. This has led to the early predominance of granular sludge in SBRs, which received NTA as co-substrate in the feed. Granules formed in the presence of synthetic chelating agent were smooth, denser and compact and showed better settling characteristics than those formed in the absence of it. Efficient degradation of different initial concentrations of NTA in SBRs and batch experiments suggests their potential application in the treatment of NTA containing wastewaters or effluents. References APHA, 1995. Standard Methods for the Examination of Water and Wastewater, 19th ed. American Public Health Association, Washington, DC. Banin, E., Brady, K.M., Greenberg, E.P., 2006. Chelator-induced dispersal and killing of Pseudomonas aeruginosa cells in a biofilm. Appl. Environ. Microbiol. 72, 2064e2069. Beun, J.J., van Loosdrecht, M.C., Heijnen, J.J., 2000. Aerobic granulation. Water Sci. Technol. 41, 41e48. 43 Bolton Jr., H., Girvin, D.C., Plymale, A.E., Harvey, S.D., Workman, D.J., 1996. Degradation of metalenitrilotriacetate complexes by Chelatobacter heintzii. Environ. Sci. Technol. 30, 931e938. Bucheli-Witschel, M., Egli, T., 2001. Environmental fate and microbial degradation of aminopolycarboxylic acids. FEMS Microbiol. Rev. 25, 69e106. Chen, X., Stewart, P.S., 2000. Biofilm removal caused by various chemical treatments. Water Res. 34, 4229e4233. Gjaltema, A., Vinke, J.L., van Loosdrecht, M.C.M., Heijnen, J.J., 1997. Abrasion of suspended biofilm pellets in airlift reactors: importance of shape, structure, and particle concentrations. Biotechnol. Bioeng. 53, 88e99. de Kreuk, M.K., van Loosdrecht, M.C., 2004. Selection of slow growing organisms as means for improving aerobic granular sludge stability. Water Sci. Technol. 49, 9e17. de Kreuk, M.K., Heijnen, J.J., van Loosdrecht, M.C., 2005. Simultaneous COD, nitrogen, and phosphate removal by aerobic granular sludge. Biotechnol. Bioeng. 90, 761e769. Liu, Y., Tay, J.H., 2004. State of the art of biogranulation technology for wastewater treatment. Biotechnol. Adv. 22, 533e563. Liu, Y., Yang, S.F., Tay, J.H., 2004. Improved stability of aerobic granules by selecting slow-growing nitrifying bacteria. J. Biotechnol. 108, 161e 169. Morgenroth, E., Sherden, T., van Loosdrecht, M.C.M., Heijnen, J.J., Wilderer, P.A., 1997. Aerobic granular sludge in a sequencing batch reactor. Water Res. 31, 3191e3194. Nancharaiah, Y.V., Joshi, H.M., Mohan, T.V.K., Venugopalan, V.P., Narasimhan, S.V., 2006a. Aerobic granular biomass: a novel biomaterial for uranium removal. Curr. Sci. 91, 503e509. Nancharaiah, Y.V., Schwarzenbeck, N., Mohan, T.V.K., Narasimhan, S.V., Wilderer, P.A., Venugopalan, V.P., 2006b. Biodegradation of nitrilotriacetic acid (NTA) and ferriceNTA complex by aerobic microbial granules. Water Res. 40, 1539e1546. Nicolella, C., van Loosdrecht, M.C., Heijnen, J.J., 2000. Particle-based biofilm reactor technology. Trends Biotechnol. 18, 312e320. Nowack, B., 2002. Environmental chemistry of aminopolycarboxylate chelating agents. Environ. Sci. Technol. 36, 4009e4016. White, V.E., Knowles, C.J., 2000. Effect of metal complexation on the bioavailability of nitrilotriacetic acid to Chelatobacter heintzii ATCC 29600. Arch. Microbiol. 173 (5e6), 373e382.