An anaerobic-aerobic sequencing batch reactor (SBR) with a working volume of 3.2 L was used for the cultivation of activated sludge. The reactor was operated with a cycle time of 12 h, consisted of 7 min of influent filling, 6 h of anaerobic, 5 h of aeration, 40 min of settling, 7 min of effluent withdrawal and 6 min of standing idle. The hydraulic retention time (HRT) and the sludge retention time (SRT) were set at 24 h and 12 days, respectively. Nitrogen gas was purged into the reactor with mechanical mixing for anaerobic conditions. Air was introduced through an air diffuser by an air pump for aeration. The temperature was kept at 25˚C with water jacket. The azo dye of MR was introduced to activated sludge gradually from 5 to 40 mg/L.

The seed sludge was obtained from a secondary sedimentation tank of the sewage treatment station in Han Bai Textile Company, Nanjing, China. The seed sludge had a mixed liquor suspended solids (MLSS) concentration of about 3000 mg/L. Synthetic wastewater used in this work comprised (mg/L): sodium acetate 1000, NH₄Cl 170, CaCl₂・2H₂O 7.5, NaH₂PO₄ 48, MgCl₂ 11, and KCl 25. In addition, 2.5 mL microelement solution was added, which contained (mg/L): H₃BO₃ 2, FeCl₂・4H₂O 2, EDTA 2, ZnCl₂・4H₂O 0.4, MnCl₂・4H₂O 0.8, CuCl₂・2H₂O 0.2, (NH₄)₆MoO₇・4H₂O 1.1 and NiCl₂・6H₂O 1.

The batch experiments were carried out in a 2.0 L batch reactor. When the reactor reached the steady state, the activated sludge with a mixed liquor volatile suspended solids (MLVSS) of 2700 - 2800 mg/L was withdrawn from the SBR to batch reactor after starved for several hours and then washed twice with distilled water. A required amount of MR solution was added to batch reactors at the beginning of cycle to provide a constant MR concentration of 5, 10, 20, and 40 mg/L for each treatment cycle, respectively. The reactor without MR was used as the control. The influent chemical oxygen demand (COD) concentration was 400 mg/L and the medium pH was adjusted at around 7.0 through the addition of 1M HCl or NaOH. The argon gas was filled into the synthetic wastewater for 20 min to remove the oxygen before into the reactor. In the aerobic condition, the SBR was aerated continuously to keep the dissolved oxygen (DO) concentration around 6 mg/L. Samples of the activated sludge mixture were withdrawn from the reactor at set intervals for the analysis of COD, dye, phosphorus, and intracellular storage products of PHA and glycogen.

The measurements of COD, phosphorus (PO₄³⁻-P), MLSS, and MLVSS were conducted according to the Standard Methods [16] . The concentration of azo dye was determined using the modified method of Wong and Yuen [17] . Aliquots from the test solution were centrifuged at 5200 rpm for 5 min. The dye solution was measured at 430 nm, which is the absorbance maximum for the orange color of methyl red, using a spectrophotometer (ALPHA-1506, Pu Yuan, China).

The PHA was determined according to Fang et al. [18] , and glycogen was measured according to Smolders et al. [19] with minor modification. Glycogen was extracted and hydrolyzed from activated sludge by heating a sample of a known amount of freeze-dried activated sludge (typically 50 mg) in a known volume of 0.6M HCl (typically 5 mL) for 5 h at 100˚C. The glycogen content in the supernatant was determined by anthrone method [20] with glucose as the standard.

The influences of dye on COD and dye removal efficiencies are shown in Figure 1. At the MR concentration of 5 mg/L, the COD removal efficiency was almost the same as that in control experiment. However, with the increase of MR concentration, COD removal efficiency decreased slightly. The increase of MR concentration from 0 to 40 mg/L led to the decrease of COD removal efficiency from 97.9% to 84.0% (Figure 1(a)). The inhibition of PAOs and other heterotrophic microorganisms might be the reason for this phenomenon.

The dye removal efficiency was also decreased from 99.7% to 86.9% with the increase of MR concentration from 5 to 40 mg/L (Figure 1(b)), and the result was consistent with that reported by Hakimelahi et al. [21] . The decrease of dye removal efficiency was probably due to the lessening of the COD removal caused a lower production of reducing equivalents, negatively affecting the color removal [22] . Furthermore, main color was removed at the anaerobic condition and the contribution of aerobic condition to decolorization was low. The results were in good agreement with other results in the sequencing anaerobic-aerobic conditions [13] [23] .

To further confirm the biodegradation of MR by activated sludge, the UV-Vis spectra of the samples taken at different time are illustrated in Figure 2. The peak in the visible region at 430 nm which represents the azo linkage of MR was decreased at anaerobic condition. Correspondingly, the appearance of the absorbance peak at 241 nm and 305 nm is probably due to the production of intermediate products of N, N’ dimethyl-p-phenyle-nediamine (DMPD) and 2-aminobenzoic acid (2-ABA) [24] . At aerobic condition, the two absorbance peaks were decreased. This indicated that the ability of the activated sludge for the biodegradation of MR.

The P release at anaerobic condition and P uptake at aerobic condition at different MR concentrations are shown in Figure 3. It was found that P removal efficiency was almost 100% in the control experiment. At low MR concentration of 5 mg/L, the P release and uptake were almost the same as those in control experiments, which indicated that low MR concentration had no inhibitory effect on P removal in the EBPR system. At MR concentration of 10 mg/L, P removal efficiency decreased slightly to 93.6%. However, when MR concentration was further increased to 20 and 40 mg/L, P removal efficiency decreased to 59.9% and 19.9%, respectively. Under anaerobic conditions, P release decreased significantly at MR concentration of 20 and 40 mg/L, suggesting that polyphosphate degradation was inhibited at high MR concentration [3] . Correspondingly, P uptake at aerobic condition was inhibited due to the shortage of energy provided by PHA degradation [25] . Thus, the increase of MR concentration severely decreased P removal capacity and had a toxic effect on the activity of PAOs.

The changes of intracellular storage product of PHA at anaerobic condition at different MR concentrations are illustrated in Figure 4. Under the control experiment, the content of PHA stored at anaerobic condition was 152.4 mg/gVSS. With the increase of MR concentration from 5 to 40 mg/L, the content of PHA stored at anaerobic condition decreased from 147.2 to 79.0 mg/gVSS. The linear regression analysis indicated that the stored PHA content at anaerobic condition was linearly decreased with the increase in MR concentration (R² = 0.978). The results indicated that high dye concentration could inhibit the PHA synthesis at anaerobic condition in EBPR system, especially at high MR concentration. Due to the inhibitions of high concentration of intermediate aromatic amide on substrate uptake and poly-phosphate degradation at anaerobic condition, insufficient energy available can be used for PHA synthesis, leading to less PHA stored at higher MR concentration [1] [5] .

Figure 4 also gives the content of PHA degradation at aerobic condition with different concentrations of MR. The content of PHA consumption at aerobic condition was 146.8 mg/gVSS for the control experiment, and similar PHA degradation

was displayed at MR concentration of 5 mg/L. However, with the increase of MR concentration from 10 to 40 mg/L, PHA degradation content decreased sharply, indicating that high concentration of MR had severely inhibition effect on PHA degradation at aerobic condition. The results demonstrated that the mechanism of aerobic PHA degradation was more sensitive to MR dosing than that of anaerobic PHA synthesis, which was in agreement with the findings of Tsai and Chen [4] and Wang et al. [5] .

Figure 5 depicts the variation of intracellular storage product of glycogen at different MR concentrations. Under the control experiment, the hydrolyzed-glycogen was 46.7 mg/g VSS at anaerobic condition. At low MR concentrations (5 and 10 mg/L), the amount of glycogen hydrolysis at anaerobic condition was similar with that in the control experiment, indicating that low concentration of MR had no effect on the hydrolysis of glycogen at anaerobic condition in EBPR system. However, glycogen contents at anaerobic condition increased when dye concentrations were increased from 20 and 40 mg/L. The higher glycogen content at anaerobic condition was probably due to the maintenance adenosine triphosphate required by the cell [5] .

At aerobic condition, the content of glycogen was 54.9 mg/g VSS for the control experiment (Figure 5). At low MR concentrations (5 and 10 mg/L), the content of glycogen at aerobic condition was similar with that in the control experiment, indicating that low concentration of MR had no effect on glycogen replenishment at aerobic condition in EBPR system. However, the glycogen content decreased sharply at MR concentrations of 20 and 40 mg/L. Due to the accumulated PHA at anaerobic condition decreased at high MR concentration, less available PHA could be used for glycogen reproduction, resulting in the decrease in glycogen content [3] .