外文文献

Desalination 225 (2008) 301–311 0011-9164/08/$– See front matter © 2008 Elsevier B.V. All rights reserved *Corresponding author. Decolorization and COD removal of secondary yeast wastewater effluents by coagulation using aluminum sulfate Yu Zhou, Zhen Liang, Yanxin Wang* School of Environmental Studies and MOE Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences, Wuhan, 430074, PR China Tel. +86 (27) 62879198; Fax +86 (27) 87481030; email: yx.wang1108@gmail.com Received 14 February 2007; accepted 8 July 2007 Abstract Coagulation process using aluminum sulfate was employed as an advanced treatment method for biologically treated yeast wastewater in this paper. Factors such as solution pHcoagulant dosage, coagulant aids (flocculant), and mixing conditions that influence COD and color removal efficiency were experimentally tested. Coagulation mechanism using aluminum sulfate was studied as well. The results show that the optimum dosage of the coagulant increases with the increase in initial pH. In the pH range tested, the appropriate initial pH should be higher than 7 to ensure effective removal of colorants. Cationic polyacrylamide (PAM) significantly reduced the effluent turbidity, hence enhancing removal efficiency. Mixing conditions, particularly the mixing rate, have some effect on coagulation efficiency. The appropriate rate was determined to be between 300 and 500 rpm. Under optimal conditions, the removal efficiency of COD and color reached 72% and 90% respectively. The coagulation mechanism was discussed as well. Basically, charge neutralization plays the predominant role. Keywords: Yeast wastewater; Advanced treatment; Coagulation; Aluminum sulfate 1. Introduction The industrial production of yeast by fermen- tation that generally uses molasses as the raw materials includes operations and processes such as molasses preparation, fermentation, and sepa- ration and drying of yeast and produces a large quantity of high-strength liquid wastes called yeast wastewater [1]. These wastes are strongly acidic (pH 4–5) and contain a large amount of colorants that give dark brown color and high chemical oxygen demand (COD) in the range of 50–100 g/L. After a multi- stage biological treatment, most of the organic load is removed. However, due to the nature of the colorants known as melanoidins [2,3], which doi:10.1016/j.desal.2007.07.010 302 Y. Zhou et al. / Desalination 225 (2008) 301–311 are resistant to biodegradation, the brown color does not disappear and it can even increase due to repolymerization of colorants. Melanoidins are high molecular weight polymers. The formation of melanoidins comprises a set of consecutive and parallel chemical reactions taking place between amino compounds and carbohydrates during a Maillard reaction. Conventional anaerobic-aero- bic treatment processes can accomplish the deg- radation of melanoidins only up to 6% or 7%. As a result, the COD and color of the effluents after aerobic treatment can still highly exceed the na- tional discharge regulation. Therefore, it is nec- essary to further remove these recalcitrant organic matters by polishing treatment. Several treatment processes have been used in decolorization of yeast wastewater. Among them are chemical oxidation using hypochlorite and hydrogen peroxide [4,5] and physicochemical treatment, such as adsorption using activated car- bon [6], membrane processes such as ultrafiltra- tion (UF) and nanofiltraton (NF) [7–9]. Many of them are not cost-effective, limiting their wide- spread application in biologically treated yeast wastewater. Therefore, relatively simple and cost- effective alternative technologies are urgently needed in decolorization of yeast effluent. Coagu- lation is a process for combining particles (col- loidal and suspended) and/or dissolved organic matter into large aggregates, thereby facilitating their removal in subsequent sedimentation/flota- tion and filtration stages. Chemical coagulants can destabilize particulates by four distinct mecha- nisms: double layer compression; charge neutral- ization; enmeshment in a metal hydroxide pre- cipitate; and inter-particle bridging. Aluminum sulfate (hereafter abbreviated as alum) is the most widely used coagulant in waste- water treatment industry. Using alum, particle destabilization is believed to be brought about by Al polymers which are kinetic intermediates in the eventual precipitation of a metal hydroxide [10,11]. These polymers are adsorbed on colloi- dal particles. The amount of polymer adsorbed and consequently the dosage of alum coagulant necessary to accomplish destabilization of colloi- dal particles depend on the concentration of col- loids. If pH is below the zero point of charge (z.p.c) of the Al hydroxide, positively charged polymers will prevail and adsorption of these posi- tive polymers can destabilize negatively charged colloids by charge neutralization. At high doses of alum, a sufficient degree of oversaturation in- duces rapid precipitation of a large quantity of aluminum hydroxide, to form “sweep floc”. How- ever, few studies have been made on the use of alum for the removal of soluble and colloidal or- ganics from yeast effluents. This paper reports the results of our work on applying coagulation of effluent after biological process as an alternative process for removing COD and color from yeast wastewater. The efficiency of wastewater decolorization depends on removal of organic colorants such as pigments and natural organic matters (NOM). Wang et al. [12] reported their experimental re- sults of color removal from dyeing wastewater using cellular iron. The viability of coagulation for decolorization of colorants has also been tested [13–15]. On the basis of the above discussion, the main objective of this research was to examine the co- agulation performance of alum to removal color and COD under different pH conditions and vari- ous dosages and to determine the optimal param- eters in COD and color removal for the advanced treatment of yeast wastewater. 2. Materials and methods 2.1. Wastewater Yeast wastewater samples were collected from a yeast-manufacturing factory in Hubei province (central China) after being subjected to an anaero- bic–aerobic biological treatment. Upon collection of the samples, their characteristics were investi- gated. All samples were stored in 25 L plastic car- boys and kept in refrigerator at 5°C before use. Y. Zhou et al. / Desalination 225 (2008) 301–311 303 2.2. Materials and coagulation test procedures Alum (A12(SO4)3·18H2O) used for the experi- mental procedure was of analytic grade. Cationic polyacrylamide (MW>107 Dalton) was commer- cially available and was prepared as 1 g/L solu- tion with distilled water. The coagulation experiments were conducted by a series of standard six-beaker jar test appara- tus (TA-6, China). Each beaker contained 1 L of raw water samples. If necessary, the initial pH values were adjusted with concentrated sulfuric acid or NaOH solution. After a designated dos- age of alum was added to the water samples, co- agulation experiments proceeded with rapid agi- tation at 500 rpm for 1 min to guarantee thorough mixing between effluent and coagulant. Approxi- mately 5 ml of samples were taken at the end of rapid agitation for zeta potential measurement. The mixtures were then allowed by slow mixing at 60 rpm for 10 min and settling time of 1 h was used. After settling, 200 ml of supernatant fluid was taken out at a fixed distance of 2 cm below air-liquid interface with a syringe for pH, zeta potential, and turbidity measurements. All the experiments were carried out at room tempera- ture (23–25°C). 2.3. Analytical methods pH measurements were made using PS-2 pH meter which was calibrated with pH 6.86 and pH 4.00 standard buffers before use. Chemical oxygen demand (COD) concentrations were measured according to standard methods [16] and expressed as CODcr (potassium dichromate as oxidant). True color was determined at 475 nm, a wavelength that characterizes brown color [17], using a spec- trophotometer (722S, China) after pH adjustment to pH 5 followed by centrifugation at 4000 s for 20 min. Distilled water was used as the blank. The turbidity of the supernatant was measured in nephelometric turbidity unit, NTU, by a turbi- dimeter (TDT-2, China). Zeta potential was de- termined using a zeta meter (Nano-2S90 Zeta Sizer). 3. Results and discussion 3.1. Wastewater characterization After aerobic treatment the wastewater is still dark-brown colored and has high organic load. Some characteristics of the effluent are given in Table 1. Among them, samples were diluted 5 times before color measurement. Table 1 Water quality parameters of biologically treated yeast wastewater Parameters Range pH 8.0±0.2 CODcr, mg/L 1200±200 Color, abs/cm, 475 nm 2.2–2.5 Turbidity, NTU 30±10 3.2. Effect of coagulant dose and initial pH on COD and color removal The results of jar test are shown in Fig. 1. It can be seen clearly that COD and color removal increased with the increase in aluminum dose until the maximum values were obtained. For example, when initial pH value was 8, the removal rate of COD and color increased slightly when the co- agulant dose was increased from 2.5 g/L to 3.0 g/L; and that further addition of alum to 3.5 g/L re- sulted in significant improvement in removal ef- ficiency, giving total COD and color removal of 62 and 82%, respectively. In the coagulant dos- age range of 4.0–4.5 g/L, COD and color reduc- tion were at the highest level, ranging from 68– 69% and 87–89%, respectively. With further ad- dition of alum up to 5.0 g/L, coagulation rate be- gan to decrease, with COD removal less than 64%. At the initial pH 7 and 9, the curves of removal rate show similar trends. The only difference is that the optimal alum dose is in the range of 3.5– 4.0 g/L for initial pH 7 and 4.5–5.0 g/L for pH 9. Under the best condition, the COD removal effi- 304 Y. Zhou et al. / Desalination 225 (2008) 301–311 Fig. 1. COD and color removal as a function of alum dose and initial pH with 500 rpm agitation after one minute of alum addition. ciency is between 65–66% at pH 7 and 63–65% at pH 9, and color removal between 85% and 88% at pH 7 and between 85% and 87% at pH 9. When solution pH was adjusted to 6.0, the optimal alum dose was only 1.75 g/L, remark- ably lower than that at pH above neutral. On the other hand, the removal of COD and color were only 58% and 69%, respectively, much lower than those in the pH range of 7–9. Therefore it can be concluded that best efficiency was obtained at pH 8 and alum dose of 4 g/L. In addition, it can be seen that the color removal efficiency has similar trend as that of the COD, suggesting that colorants (melanoidins) contribute the major portion to the organic compounds in biologically treated yeast wastewater. Initial pH=6 0 10 20 30 40 50 60 70 80 90 100 1 1.25 1.5 1.75 2 Aluminum sulfate dose (g/L) Pe rc en t r em ov al (% ) Color COD Initial pH=7 0 10 20 30 40 50 60 70 80 90 100 2 2.5 3 3.5 4 4.5 Aluminum sulfate dose (g/L) Pe rc en t r em ov al (% ) Color COD Initial pH=8 0 10 20 30 40 50 60 70 80 90 100 2.5 3 3.5 4 4.5 5 Aluminum sulfate dose (g/L) Pe rc en t r em ov al (% ) Color COD Initial pH=9 0 10 20 30 40 50 60 70 80 90 100 3 3.5 4 4.5 5 5.5 Aluminum sulfate dose (g/L) Pe rc en t r em ov al (% ) Color COD 3.3. Effect of coagulant dose and initial pH on residual turbidity Residual turbidity in supernatant as a function of coagulant dose and initial pH is shown in Fig. 2. Unlike the COD and color removal curves, for all the initial pH tested, turbidity gradually decreased with increase in alum dose, and a point of inflec- tion did not appear in the curves. Visual inspec- tion also showed that with increase in alum dos- age, the supernatant became clearer. Although coagulant application is effective in COD and color removal, alum addition resulted in high residual turbidity. For all the pH values, the lowest turbidity was still greater than that of raw wastewater (without coagulation). For in- Y. Zhou et al. / Desalination 225 (2008) 301–311 305 Fig. 2. Residual turbidity vs. alum dosage and initial pH with 500 rpm agitation after 1 min of alum addition. Initial pH=6 0 20 40 60 80 100 120 140 160 180 200 1 1.25 1.5 1.75 2 Aluminum sulfate dose (g/L) Tu rb id ity (N TU ) Initial pH=7 0 20 40 60 80 100 120 140 160 180 200 2 2.5 3 3.5 4 4.5 Aluminum sulfate dose (g/L) Tu rb id ity (N TU ) Initial pH=8 0 20 40 60 80 100 120 140 160 180 200 2.5 3 3.5 4 4.5 5 Aluminum sulfate dose (g/L) Tu rb id ity (N TU ) Initial pH=9 0 20 40 60 80 100 120 140 160 180 200 3 3.5 4 4.5 5 5.5 Aluminum sulfate dose (g/L) Tu rb id ity (N TU ) stance, when 2.5 g/L of alum was added at the pH 8, the turbidity was more than 190 NTU, much higher than that of raw water. As the coagulant dose increased, the turbidity reduced gradually, until the lowest level (70 NTU) at the maximum alum dosage of 5 g/L. Similar results were ob- tained for other pH levels. The lowest residual turbidity attained at pH between 7 and 9 was in the range of 70–85 NTU, whereas the lowest turbidity in the finished water was 106 NTU at initial pH 6, much higher than those in the pH range of 7–9. This phenomenon can be attributed to accumulation of suspended and colloidal particles from complexation between organics (colorants) of the effluent and hydrolyz- ing alum species. It can be concluded that a dos- age adequate to destabilize dissolved organic matters but insufficient to effectively remove the dispersed cells results in high levels of turbidity, due to the presence of dispersed cells and adsorbed alum species. When alum is underdosed, the par- ticles are still strongly negatively charged and repel each other, leading to high residual turbid- ity. As more coagulant is added, electrostatic forces began to minimize and more flocs were formed, resulting in decrease in turbidity. Al- though the optimum coagulant dose required at initial pH 6 is much lower than those at other tested pH values, the coagulant performance is appar- ently poorer compared with neutral and weakly alkaline pH levels. Optimum conditions for COD or color removal are not always the same as those for turbidity re- moval, as can been from Fig. 2. In terms of efflu- ent turbidity, for each pH (6–9), turbidity was greater than that of raw wastewater and decreased 306 Y. Zhou et al. / Desalination 225 (2008) 301–311 with increase in coagulant dose. Within the other pH range being tested, the turbidity curves show downward trend similar to that of pH 8. In many previous publications, optimum coagulation pro- cesses were evaluated by color and turbidity re- moval after a fixed period of settling. In this work, optimum coagulation conditions are determined by a compromise between maximum COD (color) removal and minimum coagulant dose. Accord- ingly, the treatment process should be optimized for COD (color) removal when organic matter (COD) is the major contaminant. 3.4. Final pH and zeta potential as function of coagulant dose and initial pH The change in final pH and zeta potential with coagulant dosage is plotted in Fig. 3. All pH val- Fig. 3. Plots of zeta potential and final pH vs. alum dose and initial pH with 500 rpm agitation after 1 min of alum addition. Initial pH=6 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 1 1.25 1.5 1.75 2 Aluminum sulfate dose (g/L) Ze ta p ot en tia l ( m v) 4.2 4.4 4.6 4.8 5 5.2 5.4 5.6 Fi na l p H Zeta potential Final pH Initial pH=7 -30 -25 -20 -15 -10 -5 0 5 10 15 2 2.5 3 3.5 4 4.5 Aluminum sulfate dose (g/L) Ze ta p ot en tia l ( m v) 0 1 2 3 4 5 6 7 8 Fi na l p H Zeta potential Final pH Initial pH=8 -20 -15 -10 -5 0 5 10 15 2.5 3 3.5 4 4.5 5 Aluminum sulfate dose (g/L) Ze ta p ot en tia l ( m v) 0 1 2 3 4 5 6 7 8 Fi na l p H Zeta potential Final pH Initial pH=9 -25 -20 -15 -10 -5 0 5 3 3.5 4 4.5 5 5.5 Aluminum sulfate dose (g/L) Ze ta p ot en tia l ( m v) 0 1 2 3 4 5 6 7 Fi na l p H Zeta potential Final pH ues gradually decreased with increase of alum addition, which is in agreement with the gener- ally accepted concept that hydrolysis of metal salts results in pH reduction. Despite the difference in initial pH, the final pH at which the best COD removal was achieved falls within a narrow region, between 4.7 and 5.1. This means that the effect of alum coagulation is predominantly controlled by final pH, or the pH in the treated effluent. The zeta potential curves revealed the typical relationship between zeta potential and coagulant dose. Results of zeta potential measurements also supported adsorption and charge neutralization mechanisms. The optimum coagulation dosage corresponds to a zeta potential of the particles around zero, indicating that charge neutralization is responsible for the destabilization of the col- Y. Zhou et al. / Desalination 225 (2008) 301–311 307 loid particles. As seen from Fig. 3, in most cases, at optimum dose for each pH (i.e. 3.5–4 g/L for pH 7, 4–4.5 g/L for pH 8, 4.5–5 g/L for pH 9), zeta potential values are close to zero. At the maxi- mum alum dosages (i.e., 5 g/L at pH 8), the zeta potential increases rapidly to 11.9 mV, which is higher than the most effective zeta potential range (–5 – +5 mV) for charge neutralization, indicat- ing that charge reversal causes restabilization of the particles. When the alum is underdosed, the organics is still negatively charged after interaction with the coagulant, represented by the negative value of zeta potential. In the optimal dosage range, the zeta potential is at the best region of charge neu- tralization. After that, though the zeta potential continually increased with the coagulant, the re- moval efficiency started to decrease. The only exception is at initial pH 6, when zeta potential remained negative for all alum doses. Even at the optimal dose (1.75 g/L), the corresponding zeta potential was only –9 mV. There is no doubt that, at appropriate dosage, charge neutralization by adsorbed hydrolysis products or hydroxide precipitate can cause nega- tively charged particles to become destabilized and hence to coagulate. Melanoidins of the efflu- ent are high molecular weight polymers which are generally negatively charged because of depro- tonation of the surface groups [18,19]. Charge neutralization involves reducing the diffusive layer of suspended particles by lowering the en- ergy required for other suspended particles to con- tact with each other through adsorption of the coagulant onto suspended particle surfaces. Alum has a strong tendency to form insoluble complexes with a number of ligands, especially with polar molecules and with oxygen containing functional groups such as hydroxyl or carboxyl groups [20]. As the result of coordination reaction, the colorants are destabilized and gradually develop into settleable flocs [21–23]. At the alum dosage of 4.5 g/L, the net charge of organic compounds involved in the reaction is minimal, and conse- quently, the removal rates are the best. When the dosage is further increased to 5 g/L, a small por- tion of Al-organic substances (melanoidins) com- plexes reversed their charges, preventing the com- plexes from being removed. Excessive dosage can give charge reversal and restabilization of colloids. Sweep floc conditions for alum are character- ized by large amorphous floc formation at pH from 7 to 8 [24]. Besides, even at lower concentration (10–4 M), Al3+ ion starts to hydrolyze at very low pH and aluminum hydroxide precipitates at about pH 4.5. This is clearly seen in the plots of specia- tion of Al vs. pH (Fig. 4) [25]. It has been commonly accepted that Al3+ first hydrolyzes and then reacts to form monomeric, poly