Journal of Hazardous Materials 167 (2009) 9951001

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journal homepage: www.elsevier.com/locate/jhazmat

Understanding effects of chemical structure on azo dye decolorization characteristics by Aeromonas hydrophila
Chung-Chuan Hsueh, Bor-Yann Chen , Chia-Yi Yen
Department of Chemical and Materials Engineering, National I-Lan University, I-Lan 260, Taiwan

a r t i c l e

i n f o

a b s t r a c t
This novel comparative study tended to disclose how the molecular structures present in seven azo dyes including two types of azo dyes (i.e., naphthol type azo dyes Reactive Black 5 (RB 5), Reactive Blue 171 (RB 171), Reactive Green 19 (RG19), Reactive Red 198 (RR198), Reactive Red 141 (RR141) and non-naphthol type azo dyes Direct Yellow 86 (DY86), Reactive Yellow 84 (RY84)) affected color removal capability of Aeromonas hydrophila. Generally speaking, the decolorization rate of naphthol type azo dye with hydroxyl group at ortho to azo bond was faster than that of non-naphthol type azo dye without hydroxyl group, except of RG19. The azo dyes with electron-withdrawing groups (e.g., sulfo group in RR198, RB5 and RR141) would be easier to be decolorized than the azo dyes with the electron-releasing groups (e.g., NH-triazine in RB171 and RG19). In addition, the azo dyes containing more electron-withdrawing groups (e.g., RR198, RB5 and RR141) showed signicantly faster rate of decolorization. The azo dyes with electronwithdrawing groups (e.g., sulfo group) at para and ortho to azo bond (e.g., RR198, RB5 and RR141) could be more preferred for color removal than those at meta (e.g., DY86 and RY84). The former azo dyes with para and ortho sulfo group provided more effective resonance effects to withdraw electrons from azo bond, causing azo dyes to be highly electrophilic for faster rates of reductive biodecolorization. However, since the ortho substituent caused steric hindrance near azo linkage(s), azo dyes with para substituent could be more favorable (e.g., SO2 (CH2 )2 SO4 in RR198 and RB5) than those with ortho substituent (e.g., sulfo group at RR141) for decolorization. Thus, the ranking of the position for the electron-withdrawing substituent in azo dyes to escalate decolorization was para > ortho > meta. This study suggested that both the positions of substituents on the aromatic ring and the electronic characteristics of substituents in azo dyes all signicantly affected the performance of biodecolorization of A. hydrophila. 2009 Elsevier B.V. All rights reserved.

Article history: Received 1 December 2008 Received in revised form 20 January 2009 Accepted 21 January 2009 Available online 30 January 2009 Keywords: Biodecolorization Aeromonas hydrophila Azo dye Chemical structure

1. Introduction Azo dyes are the largest chemical class of synthetic dyes, and widely used as colorants in textile dyeing, leather, plastics, food, cosmetics, paper printing, with the textile industry as the largest consumer [13]. Inevitably, the residual azo dyes in wastewater from the dyestuff or textile industry would be a signicant threat to public and environmental health. Azo dyes are originally designed to be recalcitrant for long-term use and thus resistant to aerobic wastewater treatment [25]. Moreover, azo dyes are electron-decient xenobiotics and thus are capable to be degradable via azo reduction [26]. However, due to diverse structures present in the synthetic dyes, changes in the chemical structures (e.g., isomers or the presence of different substituents) would signicantly affect the decolorization capability (e.g., reductive biodecolorization). Most of researchers have related the capacity

Corresponding author. Fax: +886 39357025. E-mail addresses: bychen@niu.edu.tw, boryannchen@yahoo.com.tw (B.-Y. Chen). 0304-3894/$ see front matter 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jhazmat.2009.01.077

of color removal to the dye class rather than the molecular feature [7,8]. However, the characteristics of substituents and their relative positions to azo bond all played crucial roles for the performance of azo dye decolorization. Moreover, Zimmermann et al. [9] used puried Orange II azoreductase from a Pseudomonas strain KF46 to assess decolorization efciency of various Orange dyes. Evidently, the specicity of Orange II azoreductase toward azo linkage for decolorization is strongly dependent upon the properties (e.g., electron-withdrawing capability) of substituents in the proximity of azo linkage(s) and thus strongly determines whether the dye is susceptible to biodecolorization. On the other hand, the hydroxy group on the 2-position of the naphthol ring of the azo dye (e.g., 1(4 -sulfophenylazo)-2-naphthol) provides a benecial prerequisite to assist dye decolorization. In contrast, charged groups near azo bond (e.g., 1-(2 -sulfophenylazo)-2-naphthol) signicantly hinder the decolorization efciency. Zimmermann et al. [9] also quantitatively revealed that the slope of the correlation between Hammetts substituent constant, (an experimental value of the electronegativity of a substituent on a phenyl ring [10]) of various substrates and the logarithm of the decolorization rate of these azo dyes was pos-

996

C.-C. Hsueh et al. / Journal of Hazardous Materials 167 (2009) 9951001

itive (i.e., the electron-withdrawing groups present on the phenyl ring escalate this color removal). To have conclusive remarks, Suzuki et al. [11] also provided a correlation of aerobic biodegradability of 25 sulfonated azo dyes with their chemical structures. Although there are many structure-based methods to reveal biodegradation as well as biodegradability, they still need an abundant supply of database from experiments to obtain highly adequate model predictions [12]. Recently, our prior studies presented preliminary comparisons to propose the plausible reasons for reaction selectivity of azo dye decolorization by Pseudomonas luteola [13,14]. These results disclosed that the sulfo group produced the strongest electron-withdrawing effect to result in the fastest rate of color removal compared to the carboxyl and hydroxyl group. In addition, the sulfo or carboxyl group was ortho to azo bond, the decolorization rate signicantly decreased compared to the para substituent to azo bond (e.g., p-MO > o-MO or p-MR o-MR; refer to [14] for details) likely due to steric hindrance near azo linkage(s). However, detailed gures of chemical structures related to their reaction selectivity in a systematic analysis were still remained open for discussion. Although thousands of organic compounds (e.g., azo dyes) have been developed for myriads of uses, most of the biodegradability and persistence of these compounds in environment are still remained unexplored. These compounds in particular recalcitrant or cytotoxic ones might be invisible threats to the globe [15]. Thus, this study tended to provide a systematized comparison, in general terms, to uncover these detailed gures. The model bacterium Aeromonas hydrophila which expressed a promising capability for color removal was predominantly isolated from fountain springs in Northeast Taiwan [16]. Chen et al. [16] mentioned that A. hydrophila could exhibit 10-fold color removal performance of azo dyes as much to P. luteola. However, the mysteries to have such a high performance of A. hydrophila for decolorization still remained to be unsolved for possible applications. In addition, A. hydrophila is known to have high productivity for biodegradable polymer (e.g., polyhydroxy-alkanoates (PHAs)) [17]. As known, polyhydroxy-alkanoates can be accumulated by various microorganisms (e.g., A. hydrophila) as carbon and energy storage materials under unbalanced growth conditions in the presence of excess carbon source. However, Aeromonas species is probably more appropriate to be cultivated in closed-vessel bioreactor due to its possible pathogenic characteristics [18]. Recently, Chen et al. [19] presented a comparative assessment to quantitatively disclose interactive toxicities of some aromatic amines (AAs; intermediates of decolorization) to A. hydrophila and P. luteola, indicating that A. hydrophila had a higher tolerance to AAs for a more promising performance of color removal. This follow-up study was to compare reductive decolorization performance of some model azo dyes using A. hydrophila, suggesting how the steric hindrance of substituents at different position (e.g., ortho, meta and para) to azo bond and electronic effect (e.g., electron-withdrawing or releasing substituent) in azo dyes affect biodegradability. Therefore, we could associate higher color removal efciency of A. hydrophila with higher tolerance to decolorization intermediates (AAs) and different reactivities of chemical structures in azo dyes. 2. Materials and method 2.1. Chemicals Seven azo dyes (Reactive Black 5 (RB5), Reactive Blue 171 (RB171), Reactive Green 19 (RG19), Reactive Red 198 (RR198), Reactive Red 141 (RR141), Direct Yellow 86 (DY86), Reactive Yellow 84 (RY84); Everlight Chemical Ltd., Taipei, Taiwan) were classied into two categories. Category A dyes, naphthol type azo dyes (e.g., RR198 (mono azo dye), RB5, RR141, RB171 and RG19; Fig. 1A), were

synthesized from H acid (8-amino-1-naphthol-3,6-disulfonic acid). Category B dyes were non-naphthol type dyes (RY84 and DY86; Fig. 1B). 2.2. Culture conditions A. hydrophila predominantly isolated from fountain springs near Chiao-Hsi in Northeast Taiwan was used as a reporter bacterium for decolorization performance [16]. To guarantee the consistent growth characteristics of cultures for study, a loopful of A. hydrophila seed taken from an isolated colony on a LB-streak plate was precultured in 50-mL LB broth, Miller (in g L1 : casein peptone 10.0, yeast extract 5.0, sodium chloride 10.0) for 24 h at 30 C, 125 rpm using a water bath shaker (SHINKWANG, SKW-12). Then, precultured broths in appropriate volume ratios were inoculated into fresh YE medium (in g L1 : yeast extract 5.0, sodium chloride 10.0) with ca. 0.3 g L1 dye for shake ask cultures. Once cells had grown to late exponential or early stationary phase (ca. 6 h), shaking was switched off and the culture was kept in static incubation (i.e., t 0) for reductive decolorization. 2.3. Analytical methods The model azo-dyes used for decolorization were RB 5 ( max = 600 nm), RB171 ( max = 609 nm), RG19 ( max = 631 nm), RR198 ( max = 522 nm), RR141 ( max = 544 nm), DY86 ( max = 393 nm), RY84 ( max = 411 nm). Dye solutions were sterilized by ltration (Millipore Millex -GS 0.22 mm lter unit), since azo dyes may be unstable in moist-heat sterilization. With appropriate calibrations at specic wavelengths, concentrations of biomass and dyes were determined using an UVvis spectrophotometer (HITACHI Spectrophotometer, model UV-2001). The concentration of dye was primarily determined by measuring the optical density (OD) of the supernatant of the sample after centrifugation for 5 min at 700 g (HSIANGTAI Centrifuge MCD-2000). In addition, a sterile cell-free medium was chosen as the control for spectrophotometric measurement. As all samples contained biomass and dye, concentrations of biomass ((a) and (b)) and dye (c) were evaluated as follows: (a) OD600nm of sample mixtures without centrifugation: OD600nm = ODX 600nm (b) OD600nm of sample supernatant (sup) after centrifugation: sup Dye OD600nm = OD600nm , (c) OD OD
X
max max

X +Dye

Dye + OD600nm ,

of sample supernatant after centrifugation: OD + OD

Dye
max

X +Dye
max

Samples were diluted to an optical density of less than 0.6 when absorbance was not within the linear arrange (ca. 0.10.7). The relation ship between the biomass concentration and OD600nm is 1.0 ODU = 0.373 g L1 dry cell weight = (1.7 0.2) 108 cfu mL1 . As the function of cell density (X) and dye concentration ([Dye]) are continuous, strictly monotonic and differentiable over time, their differential terms dX and d[Dye] could be denoted by forwarddifference formulae (e.g., dX = = X t + t X t = X, d [Dye]

[Dye] t + t [Dye] t = [Dye]) for specic rate determinations. To ensure the step size t sufciently small enough for convergence, numerical differentiations were compared with differentiations by reducing step size as t/2 i.e., (d f
t

df

t/2

) /

df

Only the error less than 1% was dened as achieving of convergent criteria within the calculated tolerance. Otherwise, the step size t was continuously reduced by half for approximations until convergence was reached. Therefore, specic growth rate (SGR), and

C.-C. Hsueh et al. / Journal of Hazardous Materials 167 (2009) 9951001

997

Fig. 1. (A) Naphthol type azo dyes (Category A dyes RR198, RB5, RR141, RB171 and RG19) used for the comparative study. (B) Non-naphthol type azo dyes (Category B dyes RY84 and DY86) used for the comparative study.

998

C.-C. Hsueh et al. / Journal of Hazardous Materials 167 (2009) 9951001

specic decolorization rate qp (SDR) could be evaluated as SDR = qp = 1 X d [Dye] dt 1 X = X t 1 X , [Dye] t and

SGR =

dlnX = dt

respectively; where X, [Dye], and t denoted for cell concentration, dye concentration, and time, respectively. 2.4. Construction of phase-plane portraits According to Perlmutter [20], the (bio)chemical reactor models could be in terms of the form: SDR = qp = SGR = = 1 d [Dye] = f1 (X, [Dye] , S ) , X dt (1a) (1b)

1 dX = f2 (X, [Dye] , S ) , X dt

where X, [Dye] and S denoted the concentration of biomass, dye and limiting nutrient substrate, respectively. Such models lend themselves to a graphical analysis in which the dependent variables (or state variables) are plotted against one another on what is called a phase-plane. Moreover, for phase-plane analysis it is postulated that the contribution of other factors (e.g., nutrient substrate, dissolved oxygen) was relatively insignicant compared to two major state variables (i.e., dye and biomass concentration). As the system evolves continuously over time through a series of states from its initial to nal condition, its mapping on the (X, [Dye]) plane will trace a trajectory to be a biological ngerprint specic to the culture system. To combine this perspective with Gadens classication scheme of fermentation [21], the phase-plane was modied to disclose on the ( , qp ) plane since the instantaneous slope of the curve in the (X, [Dye]) plane and in the ( , qp ) plane are theoretically identical. For example, the instantaneous slope m on the ( , qp ) plane can be obtained by the formula qp = 1/X 1/X d [Dye] /dt dX/dt d [Dye] f1 (X, [Dye] , S ) = = =m. dX f2 (X, [Dye] , S ) (2)

Fig. 2. Comparison on time courses of bacterial growth and decolorization of A. hydrophila cultures with various azo dyes (e.g., naphthol and non-naphthol type azo dyes) at initial concentrations of 300 mg L1 (arrows denoted the maximal specic decolorization rates (in mg L1 h1 ODU1 ): RR198(260.2); RB5(80.9); RR141(66.5); RB171(36.0); RY84(33.7); DY86(30.9); RG19(22.5).

And the entire trajectory can be drawn as a sequence of small steps using the computed slope at each step to move on to the next (i.e., exploited in the trajectory direction to show time frame on the phase-plane). This is so-called a set of auxiliary isoclines in the (X, [Dye]) plane: f1 = mf2 , which is the loci of points where all of their trajectories have slope m. As Perlmutter [20] indicated, for systems of the form (1), the trajectory from any point will be unique when the f 1 and f 2 functions have continuous rst partial derivatives (e.g., the cases of biochemical reactor model shown herein). In light of these, no trajectories can ever cross, except at singular points where the derivatives dX/dt and d[Dye]/dt are both zero (i.e., lag phases in growth and phenol degradation curves should be excluded) to prevent different bases for comparison. Thus, it is feasible to apply this phase-plane analysis to reveal growth-dependent phenomena of fermentation. 3. Result and discussion 3.1. Naphthol type azo dyes The ranking of maximal specic decolorization rate (mg L1 h1 ODU1 ) of A. hydrophila NIU01, in decreasing order, was monoazo dye RR198(260.2) > RB5(80.9) > RR141(66.5)

> RB171(36.0) > RY84(33.7) > DY86(30.9) > RG19(23.2) (Fig. 2 and Table 1). As indicated in Fig. 3, the trajectories in phase-plane plots were all nearly located in +SDR and +SGR axes (except RB5), indicating that azo dye decolorization was non-growth associated. That is, the maximal biodecolorization was only taking place when cell growth was repressed. Although Chen et al. [22] showed the relative order of initial specic decolorization rates of various dyes (RR198 > RB5 > DY86 > RY84 RR141 RB171 RG19; Table 1), both relative rankings were almost in parallel. These consistent results seemed to support that azo dye decolorization was a genotypically constitutive characteristics of A. hydrophila regardless of origins of strain isolation. Note that A. hydrophila NIU01 [16] and DEC
Table 1 Comparison upon decolorization performance of Aeromonas hydrophila isolated from different sources. Type of dyes (C.I. number) Decolorization ratea , c NIU01 Reactive Red 198 (18221) Reactive Black 5 (20505) Reactive Red 141 () Reactive Blue 171 () Reactive Yellow 84 () Direct Yellow 86 () Reactive Green 19 ()
a b c d e f e

Decolorizationb , d DEC2f 85 67 17 15 19 28 14 2 3 2 2 2 2 2 DEC3f 86 62 13 13 15 20 10 3 1 1 1 2 1 3 DEC5f 82 66 10 15 14 20 7 3 3 2 2 2 2 2

260.2 80.9 66.5 36.0 33.7 30.9 23.2

Unit: mg L1 h1 ODU1 . Unit: %. Initial concentration 300 mg L1 . Initial concentration 100 mg L1 . This study. Data adopted from Chen et al. [22].

C.-C. Hsueh et al. / Journal of Hazardous Materials 167 (2009) 9951001

999

Fig. 3. Phase-plane proles of SGR versus SDR for the batch cultures with various azo dyes. The arrows indicated the orientation of all trajectories of proles as time went by.

strains [22] were isolated from a unpolluted fountain spring and dye-laden wastewater treatment sludges, respectively. In general, the decolorization rates of naphthol type azo dyes (RR198, RB5, RR141, RB171 and RG19) except RG19 were apparently higher than those of non-naphthol type azo dyes (RY84, DY86). The reasons to cause such an order in decolorization rates are straightforward. All of naphthol type azo dyes contained a hydroxyl group at ortho to azo bond, as they were synthesized from H acid (Fig. 1A). In contrast, non-naphthol type azo dyes were free of hydroxyl group(s) in their chemical structures (Fig. 1B). The phenomena of fast decolorization rates of naphthol type azo dyes could also be conrmed in the ndings of Zimmermann et al. [9]. They pointed out that the hydroxyl group on the 2-position of the naphthol ring in the azo dye (e.g., 1-(4 -sulfophenylazo)-2naphthol) is very likely prerequisite for dye decolorization. Due to the hydroxyl group on the 2-position of naphthol ring in azo dyes, azo-hydrazone tautomerism (i.e., N N NNH) (refer to Fig. 4 for an example of RB5; [23,24]) could take place and thus the hydroxyl group (i.e., an electron-releasing group) would become carbonyl group (i.e., an electron-withdrawing group). Due to the presence of an electron-withdrawing group (carbonyl group) dyes could be reductive to be more favorable for biodecolorization and could thus stabilize the reduced forms of azo dyes by A. hydrophila (Fig. 5).

Although the sulfo group at ortho to azo bond in those naphthol type azo dyes (e.g., RR198 as shown in Fig. 1A) introduced steric hindrance to reductive biodegradation of azo dyes, electrons were still withdrawn by inductive and resonance effect due to high electronegativity characteristics of the sulfo group. As the redox potential of azo substrate became more positive, this molecule is more readily to be reduced. That is, the azo bond became more electrophilic and was thus more favorable to be reduced, as the sulfo group stabilized the negative charge generated in reduced dyes (Fig. 5). In addition, even the high electronegativity group was present at para to azo bond, it still could withdraw electron by the resonance effect as shown in Fig. 5. This led to azo dye more preferred for reductive biodecolorization compared to the group at ortho, due to lack of steric hindrance in the proximity of the azo bond [14]. In addition, Pearcea et al. [24] also mentioned that the substitution of electron-withdrawing groups (SO3 H, SO3 NH2 ) in the para position of the phenyl ring, relative to the azo bond, causes an increase in the reduction. In contrast, compared to ortho and para in naphthol type azo dyes, a sulfo group at meta to azo bond in nonnaphthol type azo dyes resulted in the least electron-withdrawing capability through the weak inductive effect. Apparently, the relative position (e.g., ortho, meta, para) of the electron-withdrawing substituent to azo bond considerably inuenced the capability of biodecolorization of A. hydrophila. In contrast, azo dyes without any electron-withdrawing substituent on aromatic ring would decrease the efciency of azo dye decolorization [14]. Among ve naphthol type azo dyes, the decolorization rate of RR198 was ranked the top one due to the sulfo group at ortho and SO2 (CH2 )2 SO4 at para to azo bond. Both high electronegativity groups would withdraw electron from azo bond via resonance, leading to the azo bond more electrophilic and reductive for color removal. Plus, they could also stabilize the negative charge on reductive azo dye (Fig. 5). Moreover, a monoazo dye (i.e., RR198) could be more favorable to be decolorized than polyazo dyes. That is, the more the azo bonds, the less readily the molecule is reduced [13,22,25]. RB5 was ranked the second fast for biodecolorization in the naphthol type azo dyes. RB5 was synthesized from H acid to be a symmetric diazo dye (Fig. 1A). Two group (SO2 (CH2 )2 SO3 Na) at para to azo bond on aromatic ring withdrew electrons from azo bond via resonance, causing azo dye more electrophilic for reductive decolorization. Thus, the negative charge was stabilized when azo dye was reduced as indicated previously in RR198. However, the biodecolorization rate of RB5 was less than that of RR198 since diazo dye was less reductive for color removal than monoazo dye.

Fig. 4. Example demonstration of azo-hydrazone tautomerism of RB5.

1000

C.-C. Hsueh et al. / Journal of Hazardous Materials 167 (2009) 9951001

Fig. 5. Example presentation of resonance effect of RR198. The sulfo group at ortho and SO2 (CH2 )2 SO4 at para to azo bond in RR198 are both high electronegativity substituents. These substituents may stabilize the negative charge present in the reduced intermediates, since azo dyes can be reduced through the inductive effect and resonance effect.

Regarding RR141, it was a symmetric diazo dye (i.e., symmetric center at the benzene ring as shown in Fig. 1A) and had a sulfo group at ortho to azo bond on each of four naphthalene rings. Although these introduced more considerable steric hindrance for azo bond breakage, they still could withdraw electron from azo bond by resonance, allowing azo dye more electrophilic to be reduced. Moreover, the sulfo group could stabilize the negative charge of reduced azo dye as mentioned before (RR198, RB5). Plus, due to steric hindrance effect from sulfo group at ortho to azo bond, the biodecolorization rate of RR141 was thus less than that of RB5. RB171 and RG19 are isomers of a symmetric diazo dye and both were synthesized from H acid with an electron-releasing group (NH-triazine). This electron-releasing group was present at meta to azo bond in RB171, but at para to azo bond in RG19. In the effect of resonance, para position was more inuenced to azo bond breakage than meta position. That is, the azo bond of RG19 was less electrophilic because of an electron-releasing group (NH-triazine) at para to azo bond. This appreciably changed the azo bond less feasible to be reduced, causing the negative charge to be less stable at reduced azo dye. Thus, its biodecolorization rate was ranked the last in seven azo dyes. In comparison with RB171 and RR141, the decolorization rate of RB171 was slower than that of RR141 because of an electronreleasing group (NH-triazine) existing in RB171. As RR141 was more electrophilic than RB171, RR141 was more feasible to be reduced for color removal [24]. In general, the less electron density in the proximity of azo bond induced this azo bond more electrophilic for reductive decolorization. That is, substituent(s) with high electronegativity near azo bond could stabilize the negative charge formed at reduced azo dye, considerably escalating the decolorization rate of azo dye. 3.2. Non-naphthol type azo dyes Although there existed a methyl group at ortho to azo bond in non-naphthol type azo dyes RY84 and DY86 (Fig. 1B), the decol-

orization rates of both dyes were slower than those of naphthol type azo dyes (except RG19) with hydroxyl group at ortho to azo bond. Nigam et al. [26] mentioned that azo compounds with a hydroxyl group or an amino group are more likely to be degraded than those with a methyl, methoxy, sulfo or nitro group. This was due to the azo-hydrazone tautomerism about naphthol type azo dyes with hydroxyl group at ortho to azo bond as shown previously. Even RY84 and DY86 were both synthesized from sulfonaphthalene, the decolorization rate of RY84 was relatively faster than that of DY86. This was because there was four sulfo groups (electronwithdrawing group) conjugate with azo bond in RY84. However, there is an electron-releasing group (i.e., NH(CH2 )2 OH) conjugate with azo bond in DY86. The sulfo group could withdraw electron from azo bond by resonance, and the environment in the proximity of azo bond became more electrophilic for higher reductive decolorization. 4. Conclusion Considering chemical structures and decolorization rates of seven model azo dyes by A. hydrophila, we could reach a conclusive remark. For similar structures of azo dyes, monoazo dyes were easily biodecolorized than polyazo dyes [13,24]. Azo dyes with hydroxyl group at ortho to azo bond were benecial to biodecolorization by A. hydrophila [9]. The azo bond was thus in a less electron density to become more electrophilic for reductive biodecolorization. Literature [9,14] also conrmed that azo dyes with electron-withdrawing groups would be more feasible for reductive decolorization than those with electron-releasing groups. Thus, azo dyes with more electron-withdrawing groups near azo bond would more feasible to be reduced for decolorization. Plus, as the position of electron-withdrawing groups to azo bond signicantly inuenced the reductive decolorization, their ranking of the decolorization rate was para > ortho > meta. Since para and ortho position could more efciently withdraw electron from azo bond through resonance, azo bond became more electrophilic to be reduced for

C.-C. Hsueh et al. / Journal of Hazardous Materials 167 (2009) 9951001

1001

color removal [6,14]. However, compared to substituents at para the steric hindrance at ortho would become more signicantly to result in a slower decolorization. This study explored more systematized details of chemical structure effect in azo dye decolorization by A. hydrophila. This nding was almost consistent with the nding for P. luteola (i.e., membrane-enclosed biocatalyst) [14] and puried azo reductase (i.e., naked biocatalyst) [9]. Follow-up studies would focus on effects of other chemical structures in azo dyes using different biodecolorizers to grasp a general perspective for biodecolorization. Acknowledgements Financial supports (NSC 95-2221-E-197-005, NSC 96-2221-E197-012, NSC 97-2221-E-197-019) from National Science Council, Taiwan, R.O.C. for this research and seeding grants for Biochemical Engineering Laboratorysdg of National I-Lan University (NIU) from the Ministry of Education, Taiwan, R.O.C. are very much appreciated. The authors also extend sincere appreciation to Distinguished Professor Jo-Shu Chang (Department of Chemical Engineering, National Cheng Kung University) for valuable suggestions. References
[1] S.V. Kulkarni, C.D. Blackwell, A.L. Blackard, C.W. Stackhocese, M.W. Alexander, Textile Dyes and Dyeing Equipment, Classication, Properties and Environmental Aspects, U.S. Environmental Protection Agency, Research Triangle Park, N.C., 1985, EPA-600/2-85/010. [2] A. Stolz, Basic and applied aspects in the microbial degradation of azo dyes, Appl. Microbiol. Biotechnol. 56 (2001) 6980. [3] A. Pandey, P. Singh, L. Iyengar, Bacterial decolorization and degradation of azo dyes, Int. Biodeterior. Biodegrad. 59 (2007) 7384. [4] K.T. Chung, S.E. Stevens Jr., Degradation of azodyes by environmental microorganisms and helminths, Environ. Toxicol. Chem. 12 (1993) 21212132. [5] P.G. Rieger, H.M. Meier, M. Gerle, U. Vogt, T. Groth, H.J. Knackmuss, Xenobiotics in the environment: present and future strategies to obviate the problem of biological persistence, J. Biotechnol. 94 (2002) 101123. [6] H.J. Knackmuss, Basic knowledge and perspectives of bioelimination of xenobiotic compounds, J. Biotechnol. 51 (1996) 287295. [7] A.J. Greaves, D.A.S. Phillips, J.A. Taylor, Correlation between the bioelimination of anionic dyes by an activated sewage with molecular structure. Part 1. Literature review, J. Soc. Dyers Colorists 115 (1999) 363365.

[8] H.R. Hitz, W. Huber, R.H. Reed, The absorption of dyes on activated sludge, J. Soc. Dyers Colorists 94 (2) (1978) 7176. [9] T. Zimmermann, H.G. Kulla, T. Leisinger, Properties of puried orange II azoreductase, the enzyme initiating azo dye degradation by Pseudomonas KF46, Eur. J. Biochem. 129 (1982) 197203. [10] L.P. Hammett, The effect of structure upon the reactions of organic compounds benzene derivatives, J. Am. Chem. Soc. 59 (1937) 96103. [11] T. Suzuki, S. Timofei, L. Kurunczi, U. Dietze, G. Schuurmann, Correlation of aerobic biodegradability of sulfonated azo dyes with the chemical structure, Chemosphere 45 (2001) 19. [12] J.W. Raymond, T.N. Rogers, D.R. Shonnard, A.A. Kline, A review of structurebased biodegradation estimation methods, J. Hazard. Mater. B84 (2001) 189215. [13] C.C. Hsueh, B.Y. Chen, Comparative study on reaction selectivity of azo dye decolorization by Pseudomonas luteola, J. Hazard. Mater. 141 (2007) 842849. [14] C.C. Hsueh, B.Y. Chen, Exploring effects of chemical structure on azo dye decolorization characteristics by Pseudomonas luteola, J. Hazard. Mater. 154 (2008) 703710. [15] L.P. Wackett, L.B.M. Ellis, Predicting biodegradation, Environ. Microbiol. 1 (2) (1999) 119124. [16] B.Y. Chen, W.M. Chen, F.L. Wu, P.K. Chen, C.Y. Yen, Revealing azo-dye decolorization of indigenous Aeromonas hydrophila from fountain spring in Northeast Taiwan, J. Chin. Inst. Chem. Eng. 39 (2008) 495501. [17] S.Y. Lee, Bacterial polyhydroxyalkanoates, Biotechnol. Bioeng. Rev. 49 (1996) 114. [18] B.Y. Chen, W.M. Chen, H.Y. Kuo, C.C. Hsueh, Comparative assessment upon dye removal capability of indigenous bacterial strains from Lanyang Plain in northeast Taiwan, J. Hazard. Mater. 161 (1) (2009) 526533. [19] B.Y. Chen, K.W. Lin, Y.M. Wang, C.Y. Yen, Revealing interactive toxicity of aromatic amines to azo-dye decolorizer Aeromonas hydrophila. J. Hazard. Mater. (2009), in press. [20] D.D. Perlmutter, Geometry of dynamics, in: Stability of Chemical Reactors, Prentice-Hall, U.S.A., 1972, p. 51, Chapter 3. [21] E.L. Gaden Jr., Fermentation process kinetics, J. Biochem. Microbiol. Technol. 1 (4) (1959) 413429 (Reprinted Biotechnol. Bioeng. 67(6) (2000) 629635). [22] K.C. Chen, J.Y. Wu, D.J. Liou, S.C.J. Hwang, Decolorization of the textile dyes by newly isolated bacterial strains, J. Biotechnol. 101 (2003) 5768. [23] J.A. Libra, M. Borchert, L. Vigelahn, T. Storm, Two stage biological treatment of a diazo reactive textile dye and the fate of the dye metabolites, Chemosphere 56 (2004) 167180. [24] C.I. Pearcea, J.R. Lloydb, J.T. Guthriea, The removal of colour from textile wastewater using whole bacterial cells: a review, Dyes Pigments 58 (2003) 179196. [25] T.L. Hu, Kinetics of azoreductase and assessment of toxicity of metabolic products from azo dyes by Pseudomonas luteola, Water Sci. Technol. 43 (2) (2001) 261269. [26] P. Nigam, I.M. Banat, D. Singh, R. Marchant, Microbial process for the decolorization of textile efuent containing azo, diazo and reactive dyes, Process Biochem. 31 (5) (1996) 435442.