Role of Al13 species in removal of natural organic matter from low specific UV absorbance surface water and the aggregates characterization
Weiying Xu,Baoyu Gao,Yan Wang,Zhonglian Yang,Xiaowen Bo
DOI: https://doi.org/10.1016/j.cej.2011.04.042
IF: 15.1
2011-01-01
Chemical Engineering Journal
Abstract:The effect of coagulant dose and pH on the coagulation behaviors of nano-Al 13 polymer (Al 13 O 4 (OH) 24 7+ ) was studied with respect to the treatment of low specific UV absorbance (SUVA) surface water. The removal of fine particles and natural organic matter (NOM) was investigated under different coagulation conditions. Polyaluminum chloride (PACl) and traditional aluminum coagulant, alum (Al 2 (SO 4 ) 3 ) were used for purpose of comparison. Floc size, strength and fractal structure were investigated under different coagulation conditions. The results showed that fine particles and hydrophobic NOM with large molecular weight (MW) could be easily removed by the three coagulants in a wide dose range (5–17 mg/L); while the proportion of hydrophilic NOM in the removed dissolved organic matter (DOM) was larger in Al m and Al 13 species coagulation at low and moderate dose (lower than 11 mg/L). Neutral and alkaline pH facilitated the particle removal, while acidic pH contributed to better hydrophilic NOM removal. The floc strength investigation indicated that the floc strength formed by Al 13 under different coagulation conditions was in the following hierarchy: pH 6.5 (15 mg/L) > pH 4.0 (9 mg/L) > pH 8.0 (15 mg/L). For Al 2 (SO 4 ) 3 coagulation, the order was as following: pH 4.0 (9 mg/L) > pH 6.5 (15 mg/L) > pH 8.0 (15 mg/L). Additionally, the fractal dimension of flocs indicated that Al 13 species resulted in compact aggregates at low concentration and pH. Keywords Al 13 polymer Natural organic matter SUVA Floc characterization 1 Introduction Natural organic matter (NOM) in rivers and lakes is a complex mixture of molecules with varying molecular weight (MW) and chemical nature. It can cause odor, taste, color and bacterial re-growth in potable water. Even worse, it is a precursor for disinfection by-products, such as trihalomethanes (THMs) and haloaetic acids (HAAs), which are proven to be carcinogens [1] . Thus, effective removal of NOM is a major objective of modern drinking water treatment. At present, coagulation and flocculation is the most common process used for the removal of NOM in water treatment. Polyaluminum chloride (PACl) is widely used as the coagulant, which has replaced many traditional aluminous coagulants in recent years. Previous studies have shown that Al(III) species in PACl can be driven into various hydrolyzed Al species such as monomeric and dimmer Al species (Al(OH) 2+ , Al(OH) 2 + and Al(OH) 4 − , Al 2 (OH) 2 4+ , denoted as Al m ), tridecamer (Al 13 O 4 (OH) 24 7+ , Al 13 for short) and other unknown species including amorphous precipitate Al(OH) 3(am) [2] . Previous studies stated that the diameter of Al 13 molecule is a few nanometers; the lengths of their aggregates vary in the range of 300–400 nm [3–5] . Thus the Al 13 species is commonly called nano-Al 13 . Nano-Al 13 was regarded as the most effective coagulation species due to its strong charge neutralization capability, high structure stability [6–8] . Three common mechanisms, charge neutralization, bridging and adsorption, have been proposed to explain the interaction of NOM with hydrolyzed Al species [9,10] . Charge neutralization is thought to occur in the region where aluminum hydroxide precipitation is minimal and cationic aluminum interacts electrostatically with anionic NOM to form insoluble charge-neutral products; and destabilization by bridging happens when segment of a polymer chain is absorbed on more than one particle, thereby linking the particles together; while surface adsorption of NOM could play an important part in the regions where the concentration of coagulant was high enough to cause rapid precipitation of Al(OH) 3 . Additionally, the chemical interaction of soluble NOM with soluble Al species is also claimed to be an important mechanism. Yan et al. [11] stated that the Al m species could complex with HA to form octahedrally coordinated complexes and Hiradate and Yamaguchi [12] found that Al 13 species reacted with HA to form HA–Al 13 complexes. However, the complex structure of NOM such as high negative charge and wide MW range complicates the process of coagulation. Natural water contains a very wide variety of organic compounds, which are generally divided into two categories: (1) particulate/colloidal organic matter and (2) dissolved organic matter (DOM). Additionally, the DOM could be further divided into two categories: one is small (less than about 1000 Da), which is unlikely to provide nuclei for precipitate formation as particle; the other is larger and more complicated functional groups with the MW of several kDa to several hundred kDa, such as humic acid. Although there is increasing interest in the research of coagulation of NOM with Al 13 [11,13] , most of which focus on the treatment of turbidity-dominated water or NOM-rich water with hydrophobic NOM with high MW. The effect of Al 13 on the treatment of low specific UV absorbance (SUVA) water, which may be characterized by having NOM of hydrophilic character and less aromaticity, has not yet been fully explored. Moreover, the properties of flocs generated during Al 13 coagulation are not well understood. The operational parameters of flocs, including floc size, strength and structural property, could significantly affect the efficiency of separation and purification operation in treatment works. Boller and Blaser [37] claimed that small particles generally have lower removal efficiencies. Jarvis et al. [14] also demonstrated that smaller particles in raw water would pose a challenge to the solid/liquid separation processes if the pre-formed flocs tend to be broken by small increases in shear rate during water works unit processes. Additionally, previous studies showed that the floc shapes affected the behaviors of aggregated particles, particularly with regard to collision efficiency and settling rates [15,16] . A thorough understanding of the essential relationship between coagulation condition and floc properties is meaningful for evaluating water treatment works. The overall objective of this research was to determine the effect of Al 13 polycation on the removal of organic matter in the low SUVA surface water treatment. Moreover, the changes of floc size distribution under an enhanced shear were investigated. Floc strength and fractal properties were studied separately to characterize the flocs formed by Al 13 polymer. Alum and PACl with low content of Al 13 species were used in this study for the purpose of comparison. 2 Materials and methods 2.1 Raw water Water sample was collected every week during the experiment from Jinan section (Shandong, China) of the Yellow River in March 2009 and quiescently settled for 24 h to remove silt. The turbidity, ultraviolet light absorption at 254 nm (UV 254 ), dissolved organic carbon (DOC), SUVA, zeta potential and pH of raw water were measured after every sample collection and its major physicochemical characteristics are summarized in Table 1 . It generally indicated that natural waters with high SUVA values, e.g., ≥4 L/(mg m), have relatively high contents of hydrophobic, aromatic and high MW NOM fractions, such as humic substances. This portion of NOM is more easily removed by coagulation [17,18] . In this study, the SUVA of the raw water was less than 4 L/(mg m), which indicated that the DOM was non-humic and low in MW. 2.2 Coagulants preparation and characterization All the reagents used in preparing the coagulants were of analytical grade and deionized water was used in all solutions. Three Al-based coagulants, alum, PACl and Al 13 polymer were used in this study. Alum was prepared by directly dissolving Al 2 (SO 4 ) 3 into deionized water. The preparation of PACl and Al 13 polymer is briefly described as follows: a pre-determined amount of AlCl 3 ·6H 2 O and Na 2 CO 3 was continuously added into 100 mL deionized water under vigorous agitation within 30 min to synthesize PACl. The target basicity (OH/Al molar ratio) of the PACl was 2.0 and the temperature was kept at around 75 ± 5 °C using recycling water bath [5] . Al 13 polymer was separated and purified from the PACl by ethanol–acetone method based on the different organic solvent solubility of different Al species. Specifically, 10 mL of PACl solution was transferred into a 1 L glass beaker, and then a pre-determined volume of ethanol and acetone was introduced under gentle agitation. Samples were taken at different reaction time and filtered using 0.45 μm millipore membranes. The precipitates formed by 150–200 mL organic solvent were obtained and dried at room temperature. The Al 13 precipitate was diluted with deionized water to release Al 13 from solids into solution when it was used. All the pre-prepared coagulants were stored in refrigerator for later use and there was no apparent turbidity in the coagulant solutions during the study. The total aluminum concentration (Al T ) was determined by ICPAES (PerkinElmer, Optima 2000, UK). The Al m and Al 13 species in different Al coagulants were analyzed by 27 Al NMR spectroscopy with 27 Al NMR spectra obtained from a Varian UNITY INOVA (500 MHz). The undetectable species, denoted as Al other , was Al T minus Al m and Al 13 . The properties of coagulants are shown in Table 2 . 2.3 Jar tests Jar tests were performed on a programmable jar test apparatus (ZR4-6, Zhongrun Water Industry Technology Development Co. Ltd., China) with 1.0 L cylindrical plexiglass beakers. The test was started with a rapid mixing at 200 revolutions per minute (rpm) for 1.5 min, followed by a mixing at 40 rpm for 15 min and then 30 min of quiescent settling. After the settlement, samples were collected from about 2 cm below the solution surface for subsequent measurements, including residual turbidity, UV 254 , DOC and final pH. An unfiltered sample was obtained immediately after the rapid mixing phase for zeta potential measurement. All experiments were conducted at room temperature of 20 ± 2 °C and all coagulants dosages used in this study were in the unit of mg/L as Al 2 O 3 . 2.4 Floc breakage Gregory [19] stated that when comparing different flocs, the size for a given shear rate indicated floc strength. This is the case for the given shear condition where the flocs are formed. However, it dose not give an indication of how flocs will behave under exposure to an increasing shear rate, which usually occur at a WTW when flocs are transferred from flocculators or to a higher shear treatment process such as high rate filtration. One of the simplest quantitative evaluations of floc strength is the calculation of a floc strength factor, which is calculated as follows [20,21] : (1) Strength factor = d 2 d 1 × 100 where d 1 (μm) and d 2 (μm) are the average size of flocs at the steady phase before breakage and after the breakage period, respectively. A larger strength factor implies flocs that are better able to withstand shear and thus should be considered stronger than those with lower factors. In this paper, the jar tests were programmed similarly as aforementioned. After the slow stir phase (15 min of 40 rpm), the suspension was exposed to a high shear of 200 rpm for a further 5 min. A laser diffraction instrument (Malvern Mastersizer 2000, Malvern, U.K.) was used to measure dynamic floc size as the coagulation proceeded. The suspension was monitored by drawing water through the optical unit, which was transferred back into the jar by a peristaltic pump on the return tube using 5 mm internal diameter peristaltic pump tubing at a flow rate of 1.5 L/h. Size measurements were taken every 35 s during coagulation and logged onto a PC. 2.5 Shear-based floc strength Strong empirical evidence shows that the stable floc size is governed by the prevailing shear rate within the containing vessel [22,23] . Previous studies revealed the empirical relation on a log–log scale between the average velocity gradients G in the flocculation and the floc size of the suspension in equilibrium [22] : (2) lg d = lg C − γ lg G where d is the floc diameter; C is a constant that strongly depends on the method used for particle size measurement; G is the average velocity gradient and γ is the stable floc size exponent. In this paper, the suspension was exposed to a slow stir phase (15 min at 40 rpm) and then, an increased shear was applied. Separate experiments were carried out and repeated 3 times at increased shears of 40, 50, 75, 100, 150 and 200 rpm (the corresponding G values were 10.1, 13.6, 23.5, 34.6, 59.7 and 87.8 s −1 ) for 5 min. In this study, the rpm was used to substitute for the G and an adapted version of Eq. (2) was used as shown below: (3) lg d = lg C ′ − γ ′ lg rpm The slope of the line ( γ ′) gives an indication of the rate of degradation. A larger γ ′ value is indicative of floc that is more prone to break into smaller size under an increasing shear force. 2.6 Floc fractal structure analysis Floc mass fractal dimension was determined using the light scattering method and previous literatures have reported in detail about the mass fractal dimension [24,25] . The scattered light intensity I is measured as a function of the magnitude of the scattering wave vector, Q . Lin et al. [26] presented the relationship between the scattered intensity from the aggregate, I and Q as follows: (4) I ∝ Q − D f where (5) Q = 4 π n sin ( θ / 2 ) λ here, n is the refractive index of the fluid, λ is the wavelength in vacuum of the laser light used, and θ is the scattering angle. Based on Eq. (4) , the fractal dimension D f can be calculated from the absolute slope of log I versus log Q by fitting a straight through the scattered plot. The compact aggregates presented higher D f , while aggregates with loosely bound structures showed lower D f . 3 Results and discussion 3.1 Effect of dosage on coagulation performance Turbidity and NOM removal by the three kinds of Al coagulants in the dose range of 1–20 mg/L as Al 2 O 3 was studied and the results are shown in Fig. 1 . DOC and UV 254 removal were separately assessed and meanwhile, zeta potentials of flocs as well as SUVA and final pH of the effluents were investigated. At low dosages (1–5 mg/L), the residual turbidity decreased dramatically with dose and the zeta potentials remained negative but steadily approached zero. Al 13 species contributed to the most effective particle removal, which was due to its higher charge neutralization ability. This indicated that charge neutralization played an important role in the coagulation at low doses. As dose increased, the removal efficiency reached a plateau and further increase of dose would lead to a slight re-stabilization of particles, which was ascribed to the charge reverse of particles. It was noticed that monomeric Al species in dose range of 5–11 mg/L resulted in high removal efficiency though the zeta potentials were still negative. This was attributed to the polymerization of Al m into Al 13 and further into amorphous precipitates, which has been demonstrated by Zhao et al. [27] and Lin et al. [28] . The generated Al precipitates could destabilize the fine particles by adsorption and make it easier for the flocs to settle. According to the results as shown in Fig. 1 b, the DOC fraction was removed most efficiently by Al 13 polymer. Yan et al. [29] believed that the high neutralization ability of Al 13 species contributed to the effective removal of DOC. Yet, it was observed that the DOC removal kept increasing with dose although charge reversal occurred, proving that charge neutralization was not the sole dominant mechanism. The complexing action between Al species and NOM and adsorption mechanism also played important roles in DOC removal. Zhao et al. [27] claimed that the large size of Al 13 polymer produced larger Al 13 –NOM complexes even with the hydrophilic fraction in low MW, which could be filtered out by filtration through 0.45 μm membrane. Similarly, monomeric Al species (Al m ) also had high complexation ability even though the charge of Al a was not as high as Al b . Yet, the charge of per Al ion in Al a was higher than that in Al b . However, the Al m –NOM floc was not large enough to be removed by filtration through 0.45 μm membrane at low Al concentration. Thus, it could be found that in the dose range of 1–9 mg/L, the efficiency of DOC removal was in the order of Al 13 followed by PACl and then Al 2 (SO 4 ) 3 . As dose increased, the Al m –NOM floc could be adsorbed by further hydrolyzed and polymerized Al precipitates, and the elevated concentration would be in favor of collision frequency and efficiency. As a consequence, Al 2 (SO 4 ) 3 led to a more rapid increase and achieved the similar efficiency of Al 13 at dose of 13–14 mg/L, which did better than PACl. The removal of UV 254 , which was essentially attributed to the presence of aromatic compounds, presented a continuous increase with dose ( Fig. 1 c). These compounds were more easily removed by coagulation than DOC and this agreed well with the results obtained by Chang et al. [30] and Lu et al. [31] . Al 13 could remove these large organic matters more easily at low dose due to its higher positive charge per molecule; while as the dose increased, PACl displayed the best ability dependent on the adsorption of large Al other species and amorphous Al precipitates. As shown in Fig. 1 d, the SUVA slightly increased at low and moderate doses and then decreased obviously at high dose. This implied that the hydrophilic fractions of organic matters with low MW were more easily removed by charge neutralization and complexation at low and moderate doses. High dose was more favorable for the removal of large hydrophobic compounds. In view of the whole residual SUVA, Al 2 (SO 4 ) 3 and Al 13 species were inclined to result in larger SUVA effluent, indicating that they could remove larger proportion of low MW fractions of organic matters. 3.2 Effect of pH on coagulation performance It is claimed that the monomeric Al could hydrolyze to form Al(OH) 3 precipitate at neutral and slightly alkaline pH [5] . Additionally, Al 13 polymer favored aggregation at pH greater than 6 to form Al 13 aggregates, which had much more branched structure [32,33] . The effects of initial pH on coagulation behaviors by the three kinds of Al coagulants were investigated in the pH range of 4–9. Based on the results shown in Figs. 1 b and c, it was found that both DOC and UV 254 removal kept increasing with dose and in order to ensure an effective removal of organic matters, the coagulant dosage was fixed at 15 mg/L. The changes of residual turbidity and zeta potential are presented in Fig. 2 a . The residual turbidity decreased sharply with pH at acidic pH and reached the lowest values at around pH 6.5. Al 13 polymer led to the highest zeta potentials and destabilized the particles more efficiently at low pH, implying the dominant role of charge neutralization. Neutral and alkaline pH contributed to effective particle removal for the three coagulants except for a slight increase of the residual turbidity at pH above 7.5 for Al 13 and PACl and at pH above 8.0 for Al 2 (SO 4 ) 3 . However, the zeta potentials at high pH did not stay close to the isoelectric point in high pH range, which were either apparently above or below zero. This implied that the particles were destabilized by electrostatic patch of Al(OH) 3 due to adsorption of precipitation [34] . For Al 13 coagulation, the large size of Al 13 aggregates also facilitated the particle destabilization through bridging and adsorption mechanisms [28,5] . Fig. 2 b shows the changes of DOC removal with pH variation. It could be found that Al 13 induced to the best DOC removal except at pH 6.0, where Al 2 (SO 4 ) 3 led to the best efficiency. At low pH, the Al 13 molecule with larger size could act as a nuclear to form colloids which could be removed by filtration. As pH increased, DOC removal efficiency increased and reached the summit at around pH 6.0. The abrupt increase in DOC removal by Al 2 (SO 4 ) 3 at pH 5–6 could be ascribed to the adsorption of NOM onto hydrolysis products Al(OH) 3 . With further increase in pH, DOC removal efficiency decreased, especially for Al 2 (SO 4 ) 3 . The coagulant with a higher Al 13 proportion showed a less variation in DOC removal with pH, which was due to the stable structure of the pre-hydrolyzed Al 13 species. The response of UV 254 removal to the variation of pH was similar to that of DOC for PACl and Al 2 (SO 4 ) 3 , which increased first and then decreased with increasing pH. Al 13 polycation displayed a more stable UV 254 removal across the pH range investigated, with a relatively better removal at around pH 7.0. Based on Fig. 1 d, the SUVA values for Al 13 and PACl decreased with increasing pH, which implied that a larger proportion of hydrophilic DOM was removed at acidic pH than at alkaline pH. Additionally, the effluent for Al 13 was with high SUVA than those for PACl and Al 2 (SO 4 ) 3 . Then, it could be concluded that the hydrophilic NOM was most inclined to be removed through charge neutralization and complexation by Al 13 ; and the bridging and adsorption of large precipitates contributed little to their removal. Another point worth attention was the SUVA values of effluents in Al 2 (SO 4 ) 3 coagulation at slightly acidic pH were higher than those at alkaline pH, which meant the higher removal of hydrophilic fractions of organic matters. 3.3 Floc size distribution and strength factor Floc property investigation in Al 13 coagulation was conducted under several different coagulation conditions, which were dominated by different coagulation mechanisms based on the above analysis. The coagulant concentration at pH 4.0 was fixed at a lower value than pH 6.5 and 8.0 to confirm the dominant role of charge neutralization and complexation at acidic pH. The coagulation settings and the corresponding dominant mechanisms are summarized in Table 3 . Al 2 (SO 4 ) 3 was used to indicate the influences of Al m and Al(OH) 3 on floc properties for comparison. Floc size distribution before and after breakage by 5 min high shear of 200 rpm is shown in Fig. 3 . For the initial flocs, a shift in the major peak to a larger size when pH increased was observed, which appeared consistent in all the cases but more prominent in Al 2 (SO 4 ) 3 coagulation. The comparison of Figs. 3 a and b revealed that Al 13 induced larger flocs at pH 4.0 and 6.5 and smaller flocs at pH 8.0 when compared to Al 2 (SO 4 ) 3 . It implied that high charge neutralization ability of Al 13 facilitated the aggregates of particles at acidic pH and the bridging of Al 13 aggregates at pH 6.5 also promoted the fine particles to form larger aggregates. At alkaline pH, the largest flocs were produced by adsorption onto amorphous precipitates in both Al 13 and Al 2 (SO 4 ) 3 coagulation. As presented in Figs. 3 c and d, the major peak shifted to the smaller size after breakage for all the cases investigated. Additionally, flocs formed by Al 13 displayed a smaller decrease in size than those formed by Al 2 (SO 4 ) 3 at pH 6.5 and 8.0; while at pH 4.0, Al 13 induced a greater decrease in floc size. This was visualized by the evaluation of the floc strength factors as shown in Fig. 3 . The factor value for Al 13 was in the following hierarchy: pH 6.5 (15 mg/L) > pH 4.0 (9 mg/L) > pH 8.0 (15 mg/L), and for Al m , the order was: pH 4.0 (9 mg/L) > pH 6.5 (15 mg/L) > pH 8.0 (15 mg/L). Strength factors for Al 13 were larger than those for Al 2 (SO 4 ) 3 at pH 6.5 and 8.0 and the result at pH 4.0 was just the reverse. The differences of floc strength could be ascribed to the resulted floc size and mechanisms involved in the coagulation, which will be discussed in detail later. 3.4 Empirical shear-based floc strength The log–log curves of stable floc size decreasing with rotary rate under different coagulation conditions are shown in Fig. 4 . The decreasing slope ( γ ′) was calculated according to Eq. (3) . The γ ′ values of pre-hydrolyzed Al 13 under different conditions were in the following order: pH 8.0 (15 mg/L) > pH 4.0 (9 mg/L) > pH 6.5 (15 mg/L). This implied that the flocs for Al 13 were most resistant to exposure to elevated shear at pH 6.5, followed by the flocs at pH 4.0 and then those formed at pH 8.0. Considering coagulation mechanisms, the flocs formed by bridging were the strongest, followed by those generated by charge neutralization; and the flocs formed by adsorption were the weakest. Comparatively, the floc strength for Al 2 (SO 4 ) 3 declined with the increasing pH. The results were in agreement with the conclusions of strength factors analysis. It appeared that the results were inconsistent with the statement of Yan et al. [35] , which demonstrated that the flocs formed by charge neutralization through a suppression of the electric double layer, gave weak physical particle–particle bonds such as vander waals forces. However, only taking the coagulation mechanisms into consideration was one-sided. Floc size and the shear rate in the vessel should also be taken into account [36] . Larger flocs were more exposed to micro scale energy-dissipating eddies that gave rise to floc breakage, while smaller flocs were more likely to become entrained within these eddies rather than being broken by them [37] . Based on the results presented in Figs. 3 a and b, it could be concluded that the floc size increased with pH for both coagulants. As a consequence, the floc strength decreased with pH with the exception of flocs formed by Al 13 species at pH 6.5, which could be interpreted by the coagulation mechanisms. At pH 6.5, the bridging of particles onto Al 13 aggregates could markedly enhance the flocs strength even when the size was larger than those at pH 4.0. This also agreed with other previous research [5,36] . 3.5 Floc fractal structure analysis The floc structure property was studied in terms of fractal dimension under different coagulation conditions as shown in Table 3 . The D f of aggregates were investigated during floc formation process, which consisted of 1.5 min of rapid shear at 200 rpm and 15 min of slow shear at 40 rpm. The results are presented in Fig. 5 . For Al 13 species, it was observed that lower coagulant concentration and pH produced higher D f of flocs. That was because the high positive charge of Al 13 could effectively neutralize the negative charge on particles, thus significantly reducing the interior repulsion of flocs to produce compact aggregates. Comparatively, at higher pH (6.5 and 8.0) and coagulant concentration (15 mg/L), the D f of flocs evidently decreased as shown in Fig. 5 a. This means that both the bridging of Al 13 aggregates and adsorption of Al(OH) 3 precipitates led to loosely arranged aggregates. Similarly, Al 2 (SO 4 ) 3 also induced loosely structured flocs at pH 8.0 according to the results shown in Fig. 5 b. However, the D f of flocs at pH 6.5 were slightly higher than at pH 4.0, which could be attributed to the transformation of Al m into Al 13 species at pH 6.5 [27,28] . According to the analysis of the entire coagulation process, a continuous increase in D f followed by a maximum at the end of the slow stir stage could be observed in all the cases. That was possibly because the weak points of aggregates were ruptured under prolonged mixing and rearranged at more dense points and subsequently, the floc structure became more compact and stable by restructuring [38,39] . 4 Conclusions A series of tests were developed to investigate the coagulation behaviors of pre-hydrolyzed Al 13 as well as floc properties with respect to the treatment of low SUVA surface water. The main conclusions from this work were listed as follows: (1) The hydrophilic NOM with small MW was inclined to be removed by charge neutralization and complexation at low and moderate doses. Al 13 species could better remove low MW NOM at low concentration than alum and PACl coagulants. (2) Pre-hydrolyzed Al 13 displayed less variation in coagulation behavior with pH changes due to its stable structure. The hydrophilic NOM was more easily removed by Al 13 species at acidic pH than at alkaline pH. (3) Compared to Al 2 (SO 4 ) 3 , Al 13 polymer produced larger flocs at lower dose and acidic ambience. At alkaline pH, the flocs formed by Al 2 (SO 4 ) 3 were a little larger than those formed by Al 13 . (4) Both the results of floc strength factor and empirical floc size exponent for Al 13 under different coagulation conditions indicated that the floc strength was in the following hierarchy: pH 6.5 (15 mg/L) > pH 4.0 (9 mg/L) > pH 8.0 (15 mg/L). While, the order for Al 2 (SO 4 ) 3 was as follows: pH 4.0 (9 mg/L) > pH 6.5 (15 mg/L) > pH 8.0 (15 mg/L). (5) Al 13 resulted in compact and dense flocs at lower dose concentration and pH. The D f of flocs were evidently increased as coagulation proceeded. Acknowledgements The research was supported by the National Natural Sciences Foundation of China (Nos. 50678095 , 50808114 ), the Key Projects in the National Science and Technology Pillar Program in the Eleventh Five-year Plan Period (No. 2006BAJ08B05 ) and the National Major Special Technological Programs Concerning Water Pollution Control and Management in the Eleventh Five-year Plan Period (No. 2008ZX07422-003-02 ). References [1] USEPA Enhanced Coagulation and Enhanced Precipitative Softening Guidance Manual 1998 EPA, Office of Water and Drinking Ground Water Washington, DC pp. 20–50 [2] J.W. Akitt A. Farthing Aluminium-27 nuclear magnetic resonance studies of the hydrolysis of aluminium (III) Part V. Slow hydrolysis using aluminum metal J. Chem., Soc. Dalton Trans. 1981 1624 1626 [3] J.Y. Bottero D. Tchoubar M.A.V. Axelos P. Quienne F. Fiessinger Flocculation of silica colloids with hydroxy aluminum polycations, relation between floc structure and aggregation mechanisms Langmuir 6 1990 596 602 [4] D.S. Wang H.X. Tang F.C. Cao Particle speciation analysis of inorganic polymer flocculants: an examination by photon correlation spectroscopy Colloid Surf. A 166 2000 27 32 [5] J.L. Lin C.J.M. Chin C.P. Huang J.R. Pan D.S. Wang Coagulation behavior of Al 13 aggregates Water Res. 42 2008 4281 4290 [6] N. Pasrthasarathy J. Buffle Study of polymeric aluminum (III) hydroxide solutions for application in wastewater treatment: properties of the polymer and optimal conditions preparation Water Res. 19 1985 25 36 [7] C. Hu H. Liu J. Qu D. Wang J. Ru Coagulation behavior of aluminum salts in eutrophic water: significance of Al 13 species and pH control Environ. Sci. Technol. 40 2006 325 331 [8] B. Gao Z. Zhang J. Ma X. Cao Solid–solid mixed method to prepare polyaluminum chloride Environ. Chem. 24 2005 569 572 (in Chinese) [9] J.E. Gregor C.J. Nokes E. Fenton Optimising natural organic matter removal from low turbidity waters by controlled pH adjustment of aluminum coagulation Water Res. 31 1997 2949 2958 [10] J.K. Edzwald J.E. Tobaison Enhanced coagulation: us requirements and a broader view Water Sci. Technol. 40 1999 63 70 [11] M. Yan D. Wang J. Qu W. He C.W.K. Chow Relative importance of hydrolyzed Al(III) species (Ala, Alb, and Alc) during coagulation with polyaluminum chloride: a case study with the typical micro-polluted source waters J. Colloid Interface Sci. 316 2007 482 489 [12] S. Hiradate N.U. Yamaguchi Chemical species of Al reacting with soil humic acids J. Inorg. Biochem. 97 2003 26 31 [13] V. Kazpard B.S. Lartiges C. Frochot J.B. d’Espinose de la Caillerie M.L. Viriot J.M. Portal T. Görner J.L. Bersillon Fate of coagulant species and conformational effects during the aggregation of a model of a humic substance with Al 13 polycations Water Res. 40 2006 1965 1974 [14] P. Jarvis B. Jefferson S.A. Parsons How the natural organic matter to coagulant ratio impacts on floc structural properties Environ. Sci. Technol. 39 2005 8919 8924 [15] C.P. Johnson X. Li B.E. Logan Settling velocities of fractal aggregates Environ. Sci. Technol. 30 1996 1911 1918 [16] X. Li B.E. Logan Collision frequencies between fractal aggregates and small particles in a turbulently sheared fluid Environ. Sci. Technol. 31 1997 1237 1242 [17] J.K. Edzwald W.C. Becker K.L. Wattier Surrogate parameters for monitoring organic matter and THM precursor J. Am. Water Works Assoc. 77 1985 122 131 [18] J.L. Weishaar G.R. Aiken B.A. Bergamaschi M.S. Fram R. Fujii Evaluation of specific ultra-violet absorbance as an indicator of the chemical composition and reactivity of dissolved organic carbon Environ. Sci. Technol. 37 2003 4702 4708 [19] J. Gregory Monitoring floc formation and breakage Proceedings of the Nano and Micro Particles in Water and Wastewater Treatment Conference 2003 International Water Association Zurich [20] R.J. Francois Strength of aluminum hydroxide flocs Water Res. 21 1987 1023 1030 [21] M.A. Yukselen J. Gregory The reversibility of floc breakage Int. J. Miner. Process 73 2004 251 259 [22] D.S. Parker W.J. Kaufman D. Jenkins Floc breakup in turbulent flocculation processes J. Sanit. Eng. Div.: Proc. Am. Soc. Civ. Eng. SAI 1972 79 99 [23] P.T. Spicer S.E. Pratsinis J. Raper R. Amal G. Bushell G. Meesters Effect of shear schedule on particle size, density, and structure during flocculation in stirred tanks Powder Technol. 97 1998 26 34 [24] J. Guan T.D. Waite R. Amal Rapid structure characterization of bacterial aggregates Environ. Sci. Technol. 32 1998 3735 3742 [25] G.C. Bushell Y.D. Yan D. Woodfield J. Raper R. Amal On techniques for the measurement of the mass fractal dimension of aggregates Adv. Colloid Interface Sci. 95 2002 1 50 [26] M.Y. Lin H.M. Lindsay D.A. Weitz R.C. Ball R. Klein P. Meakin Universality in colloid aggregation Nature 339 1989 360 362 [27] H. Zhao C. Hu H. Liu X. Zhao J. Qu Role of aluminum speciation in the removal of disinfection byproduct precursors by a coagulation process Environ. Sci. Technol. 42 2008 5752 5758 [28] J.L. Lin C. Huang C.J.M. Chin J.R. Pan The origin of Al(OH) 3 -rich and Al 13 -aggregate flocs composition in PACl coagulation Water Res. 43 2009 4285 4295 [29] M. Yan D. Wang J. Nia J. Qu C.W.K. Chow H. Liu Mechanism of natural organic matter removal by polyaluminum chloride: effect of coagulant particle size and hydrolysis kinetics Water Res. 42 2008 3361 3370 [30] E.E. Chang P.C. Chiang W.Y. Tang S.H. Chao H.J. Hsing Effects of polyelectrolytes on reduction of model compounds via coagulation Chemosphere 58 2005 1141 1150 [31] X. Lu Z. Chen X. Yang Spectroscopic study of aluminium speciation in removing humic substances by Al coagulation Water Res. 33 1999 3271 3280 [32] W.E. Dubbin G. Sposito Copper-glyphosate sorption to microcrystalline gibbsite in the presence of soluble Keggin Al-13 polymer Environ. Sci. Technol. 39 2005 2509 2514 [33] G. Furrer C. Ludwig P.W. Schindler On the chemistry of the Keggin Al 13 polymer. 1. Acid–base properties J. Colloid Interface Sci. 149 1992 56 67 [34] S.K. Dentel Application of the precipitation-charge neutralization model of coagulation Environ. Sci. Technol. 22 1988 825 832 [35] Y. Yan J.L. Burns G.J. Jameson S. Biggs The structure and strength of depletion force induced particle aggregates Chem. Eng. J. 80 2000 23 30 [36] P. Jarvis B. Jefferson J. Gregory S.A. Parsons A review of floc strength and breakage Water Res. 39 2005 3121 3137 [37] M. Boller S. Blaser Particles under stress Water Sci. Technol. 37 1998 9 29 [38] D.C. Hopkins J.J. Ducoste Characterizing flocculation under heterogeneous turbulence J. Colloid Interface Sci. 264 2003 184 194 [39] C. Selomuly R. Amal G. Bushell T.D. Waite Evidence of shear rate dependence on restructuring and breakup of latex aggregates J. Colloid Interface Sci. 236 2001 66 77