Hydrolysis Reaction Mechanism in Atrazine Metabolism and

Article pubs.acs.org/JAFC

Hydrolysis Reaction Mechanism in Atrazine Metabolism and Prediction of Its Metabolites’ Toxicities Jia Li, Jing Hu, Wenli Xu, Min Ling, and Jianhua Yao* Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, Shanghai 200032, China S Supporting Information *

ABSTRACT: Atrazine (ATR) is a widely used herbicide. There are several types of reactions in its metabolism. Herein, the mechanism of three paths of hydrolysis reactions in its metabolism and predictions of toxicities of its metabolites in the three paths will be presented. The calculation results by B3LYP (Becke, 3-parameter, Lee−Yang−Parr), one of the approaches in density functional theory, indicated that (1) there were three models in the three hydrolysis paths of ATR. The dissociation mechanisms of C(9/11)−N(8/10), C(4/6)−N(8/10), and C−Cl were dealkylation, deamination, and Cl substitution, respectively. (2) The energy barrier of C−Cl dissociation was lower. The dissociation was advantageous in dynamics and the primary reaction in the three hydrolysis paths. In these hydrolysis reactions, the different intermediates had different concentrations because of the impact of the reaction rate. (3) In addition, it was necessary to consider the solvent effect to investigate hydrolysis reaction. The conductor-like polarizable continuum model (CPCM) was used to simulate the hydrolysis reaction in bond length and energy barrier because of the solvent effect. Experimental or predictive results showed that atrazine and its metabolites in the three hydrolysis paths were carcinogenic. KEYWORDS: atrazine, density functional theory, hydrolysis reaction, solvent effect, prediction of toxicity

■

INTRODUCTION Atrazine (ATR; IUPAC name, 2-chloro-4-ethylamino-6-isopropylamine-s-triazine; CAS Registry No. 1912-24-9) is one of the triazine herbicides widely used in the United States.1 Its chemical structure is shown in Figure 1. Its acute toxicity was at

rate of ATR and (2) metabolic processes in various environmental fates. Dr. Wen and his colleagues’ showed that the salinity, pH, and temperature affected the degradation rate of ATR; the degradation of ATR in subsurface flow constructed wetland (SFCW) followed first-order kinetics, with a half-life of approximately 17.5 days.13 Drs. Chu and Lin studied the relationship between four fates (pH, temperature, ATR concentration, and sunlight) and ATR degradation rate. The results showed that ATR’s decomposition rate constant outdoors was slower in neutral pH and faster at extreme pH levels; the higher the ATR concentration was, the lower was the decomposition rate constant.14 Dr. Lin studied ATR degradation in selected plant species. The results showed that main product was ammeline (AMN) by N-dealkylation and hydrolysis in the metabolic process.10 Dr. Cheney studied the degradation reaction of ATR on synthetic birnessite (δ-MnO2). The result showed that hydrolysis of the C−Cl bond was primary in the metabolic process, and the main product was 2-hydroxy-4-ethylamino-6isopropylamine-s-triazine (HA).12 In other works about ATR degradation, HA is also the main product, and the hydrolysis of ATR was accelerated obviously in alkaline solution15 or ultraviolet light.16 Dr. Pelizzetti studied ATR degradation over TiO2 particles under simulated solar light. The result showed that the final product, 2,4,6-trihydroxy-s-triazine (CA) was generated by photocatalytic hydrolysis via the intermediate AMN.17

Figure 1. Structure of ATR.

a low level (rat, oral, LD50 = 2000 mg/kg), and AMES was negative,2 Muntagenic (sperm from human) was positive.3 It was carcinogenic by RTECS.4 Its half-life depended on environmental fates. In the document reported by the U.S. EPA,5 “Decision Documents for Atrazine”, atrazine’s overall half-life, water half-life, and sediment half-life were given as 608, 578, and 330 days, respectively. Atrazine may act as an endocrine disrupter at sensitive stages in the developmental process and may subsequently significantly reduce the reproductive capacity of amphibians.6 Therefore, the degradation rate and the properties of compounds generated in the degradation process should be attract attention. Many works about ATR’s metabolism have been reported: photolysis,7 microbial degradation,8,9 degradation in plants,10 hydrolysis,11−19 etc. All of these works focused on (1) the relationship between environmental fates and the metabolic © 2014 American Chemical Society

Received: Revised: Accepted: Published: 4852

March 7, 2014 April 29, 2014 May 7, 2014 May 7, 2014 dx.doi.org/10.1021/jf501101q | J. Agric. Food Chem. 2014, 62, 4852−4863

Journal of Agricultural and Food Chemistry

Article

Figure 2. Metabolism pathways of ATR.

Figure 3. Part of the hydrolysis reactions in ATR metabolism (bold in Figure 2).

metabolism (shown in Figure 3) was studied by B3LYP (Becke, 3-parameter, Lee−Yang−Parr)21/6-31++G(d, p),22 one of the approaches in density functional theory (DFT). The calculation results explained three points related to the hydrolysis reactions in the three paths: (1) the mechanism of each hydrolysis

ATR’s metabolism paths were summarized on the basis of publications7−20 and are listed in Figure 2. All reported studies of ATR’s metabolism focused on its metabolites and metabolic conditions. Herein, the mechanism of hydrolysis reactions in three hydrolysis paths in ATR’s 4853

dx.doi.org/10.1021/jf501101q | J. Agric. Food Chem. 2014, 62, 4852−4863


Article

Figure 4. Hydrolysis reaction processes of ATR in the three hydrolysis paths.

in Figure 4. The structure parameters of the stationary points of the reactions are shown in Figure 5. Information about atomic charge of these first transition states in the three paths and the dissociation transition state of the amino group connecting to C(4) is shown in Figure 6. In hydrolysis degradation shown in Figures 4 and 5, there were several steps in path I: (1) HA was formed. When one H2O closed to N(1), the hydrogen bond complex COM1-1 was formed at first. The bond length of the hydrogen bond is 203.2 pm. (2) The O atom of H2O attacked the C(2) to form the transition state TS1-1 with a four-membered ring (C(2)− O−H−Cl). In TS1-1, the bond length of C(2)−Cl increased from 176.1 to 227.5 pm and that of O−H in H2O is elongated to 99.5 pm. The O and H atoms all closed to C(2)−Cl, and bond lengths of C(2)−O and Cl−H were 179.4 and 206.6 pm, respectively. (3) HA was generated via the transition state TS1-1. Meanwhile, one H2O was consumed and one HCl was produced in the reaction. (4) The bond of C(11)−N(10) was hydrolyzed by H2O. The complex COM2A-1 was formed by H2O approaching the N(5) with the hydrogen bond. The hydrogen bond lengths of N(5)−H and O−H were 188.2 and 211.8 pm, respectively. In the transition state TS2A-1, there was a four-membered ring (N(10)−H−O−C(11)), and the C(11)−N(10) bond was elongated to 277.4 pm. The O−H bond in H2O is slightly elongated to 103.5 pm, and C(11)−O and N(10)−H were generated in this reaction. DIHA was generated by the end of this reaction. On the other hand, the hydrolysis of C(9)−N(8) also occurred. DEHA was generated via COM2B-1 and TS2B-1. (5) DIHA was hydrolyzed to AMN. The complex COM3A-1 was formed when one H2O closed to the C(9)−N(8); the transition state TS3A-1 was generated when a four-membered ring was formed (N(8)− H−O−C(9)). There was a difference between bond lengths of the four-membered ring in TS2A-1 and those in TS3A-1. By the end of this reaction, AMN, HCl, alcohol, and isopropanol were generated by three H2 O molecules consumed. (6) AMN was hydrolyzed to DHAA. The complex COM4-1 was formed when one H2O closed to the amino groups connected to C(6), the transition state TS4-1 was generated when a four-membered ring was formed (N(10)− H−O−C(6)). By the end of this reaction, DHAA was generated. (7) DHAA was hydrolyzed to CA. The complex COM5-1 was formed when one H2O closed to the amino groups connected to C(4), and the transition state TS5-1 was generated when a four-membered ring was formed (N(8)− H−O−C(4)). By the end of this reaction, CA was generated. The bond lengths of the four-membered ring in TS4-1 were a little different from those in TS5-1. We propose that the

reaction in the three paths; (2) the reaction path with the advantage of dynamics in the three paths and the influence of the advantage of dynamics on intermediate concentrations; and (3) in the calculation, the influence of the conductor-like polarizable continuum model (CPCM),23 which was used to simulate the aqueous environments on the geometry and potential energy surface (PES) of hydrolysis reactions. In this work, acute, mutagenic (AMES test)24 and carcinogenic toxicities of ATR and its metabolites in Figure 3 were also predicted by three computer programs, CISOCPSAT,25 CISOC-PSMT,26 and CISOC-PSCT,27 respectively.

■

COMPUTATIONAL DETAILS In this work, DFT calculations were performed with the Gaussian 09 program28 using the B3LYP method and the 6-31+ +G(d,p) basis set. Equilibrium and transition state structures were fully optimized. For all calculated structures, harmonic vibration frequencies were calculated by numerical differentiation of the energy gradients. All structures were shown to be transition states (with one imaginary frequency) or stationary points (with no imaginary frequency). The vibrational zero point energy (ZPE) corrections were based on the corresponding frequency calculation. To analyze reaction path characters, the minimum energy path was followed in both directions (forward and backward) using the intrinsic reaction coordinate (IRC).29 For all structures, natural bond orbital (NBO) analysis30 was performed for further elucidation. In addition, CPCM was used to simulate the aqueous environments for interpreting the influence of the solvent effect. To gain insight into the various bond-breaking or bond-making processes in this reaction, an analysis of bond order31−33 was carried out. CISOC-PSAT is a program used to predict acute toxicity of an organic compound. The prediction results correspond to rat, oral and range of LD50. CISOC-PSMT is a program to predict mutagenic toxicity (AMES test) of an organic compound. The prediction results correspond to mutagenicity in the test model: Salmonella typhimurium. CISOC-PSCT is a program to predict carcinogenic toxicity of an organic compound. The prediction results correspond to carcinogenicity in the test model: rat, oral.

■

RESULTS AND DISCUSSION Atom Charge and Bond Length in This Hydrolysis Reaction. In principle, hydrolysis reactions of these groups connecting to C(2), N(8), and N(10) were able to occur. Therefore, three hydrolysis paths were observed (in Figure 3), and the reaction continued to hydrolyze the amino group connecting to C(4) and C(6). Their procedures are proposed 4854



Article

Figure 5. continued

4855



Article

Figure 5. continued

4856



Article

Figure 5. Optimized structures of the reactant, intermediates, transition states, and product in the hydrolysis reaction of ATR. The values in parentheses were calculated by CPCM. (Distance in pm.)

Figure 6. NBO analysis of ATR and transition states in paths. The values in parentheses were calculated by CPCM.

255.3 and 157.9 pm, respectively. DIA was generated via the transition state TS1-2. (3) DIA was degraded to DIHA or DDA. If C(2) was attacked by H2O, a four-membered ring (C(2)−O−H−Cl) generated in the transition state TS2A-2. The bond length of C(2)−Cl was increased from 175.7 to 225.8 pm, and that of O−H in H2O was elongated to 99.4 pm. DIHA was generated when H in H2O was transferred. If C(9) was attacked by H2O, a four-membered ring (C(9)− O−H−N(8)) generated in the transition state TS2B-2 via complex COM2B-2. The bond length of C(9)−N(8) was increased from 146.2 to 254.9 pm. DDA was generated from DIA.

amino group and hydroxyl affected TS4-1 and TS5-1 slightly. In path I, the pH value of the reaction system was decreased. Reactions in path II (shown in Figures 4 and 5) also included several steps: (1) DIA was formed. When one H2O closed to N(10), the two hydrogen bond complex COM1-2 was formed at first. The bond lengths of the hydrogen bonds were 190.8 and 208.5 pm, respectively. (2) The O atom of H2O attacked the C(11) to form the transition state TS1-2 with a four-membered ring (N(10)−H−O−C(11)). In TS1-2, the bond length of N(10)−C(11) increased from 147.6 to 287.6 pm, and that of O−H in H2O was elongated to 103.5 pm. Bond lengths of C(11)−O and N(10)−H were 4857



Article

(C(9)−O−H−N(8)). (2) DEA was degraded into DEHA or DDA. If C(2) was attacked by H2O, a four-membered ring (C(2)−O−H−Cl) generated in the transition state TS2A-3, and DEHA was generated. If N(10) was attacked by H2O, a four-membered ring (C(11)−O−H−N(10)) was generated in the transition state TS2B-2 via complex COM2B-2. DDA was generated from DEA. (3) DDA was degraded into AMN or AMD. If N(1) was attacked by H2O, a four-membered ring (C(2)−O−H−Cl) was generated in TS3A-3 via complex COM3A-3. AMN generated from DDA. If N(10) was attacked by H2O, a four-membered ring (C(6)−O−H−Cl) was generated in TS3B-3 via complex COM3B-3. AMD generated from DDA. (4) AMD was degraded into DHAA or DHA. The complex COM4A-3 was formed when one H2O closed to N(1), the transition state TS4A-3 was generated when a fourmembered ring was formed (C(2)−O−H−Cl). By the end of this reaction, DHAA was generated. The complex COM4B-3 was formed when one H2O closed to N(8), and the transition state TS4B-3 was generated when a four-membered ring was formed (C(4)−O−H−N(8)). By the end of this reaction, DHA was generated. (5) DHA was hydrolyzed to CA. All results mentioned in the front were calculated in gaseous environments. In this work, the actual reaction process was in an aqueous environment. Therefore, the reactant, intermediates, transition states, and products were calculated by simulating the aqueous environment by CPCM. Some works reported that the solvent effect little influences compound structures.34,35 From information shown in Figure 5, normal chemical bond lengths were slightly affected, about 1 pm, but the bond lengths in the center of the transition state were effected obviously, about 30 pm. Therefore, the solvent effect must necessarily be considered in the optimization of the geometry structure, especially for transition states in reactions. In these hydrolysis reactions, the nucleophile H2O attacked C(2), C(4), C(9), and C(11), respectively, to form corresponding transition states. As shown in Figure 6, the O atom in H2O did not attack these C atoms directly, but formed complexes by the interaction of the hydrogen bond first. In the process of forming transition states, C(2), C(4), C(9), and C(11) lost electrons. In these corresponding transition states, the dissociated C atoms all had positive charges (+0.674/ (+0.714) C(2), +0.648/(+0.648) C(4), +0.217/(+0.313) C(9), +0.002/(+0.063) C(11)), which were good for the nucleophilic reaction. The corresponding information in detail (calculated in gaseous) is in the Supporting Information. Potential Energy Surface of Reaction. The imaginary frequencies of transition states in the reactions and relative energies corrected by ZPVE of reactant, intermediates, transition states, and product are listed in Table1. The potential energy surfaces of the reactions are shown in Figure 7. We can discern relative energies corrected by ZPVE of various compounds and transition states in reaction coordinates according to the information in Figure 7. Information given in Table 1 and Figure 7 shows that system energy decreased about 20 kJ/mol with formation of complexes because the influence of the hydrogen bond was weak. In the reaction process, the energy barrier of the transition state was the key factor of the reaction rate. As shown in Figure 7, the energy barriers of the hydrolysis of the C(2)−Cl bond in three paths were 185.6, 187.8, 203.7, 187.7, 186.2, and 202.0 kJ/mol, respectively, and their vibrational frequencies were 380.3i, 380.3i, 378.0i, 397.6i, 342.8i, and 338.8i, respectively. The vibration mode was the swing of the Cl between the C(2) and

Table 1. Imaginary Frequencies (v) of Transition States and Relative Energies (ΔE) Corrected by ZPVE of Reactants, Intermediates, Transition States, and Productsa compound ATR

ΔEb (kJ/mol)

vb (cm−1)

0

ΔEc (kJ/mol)

v (cm−1)c

0 Path I

COM1-1 TS1-1 HA COM2A-1 TS2A-1 DIHA COM3A-1 TS3A-1 COM2B-1 TS2B-1 DEHA COM3B-1 TS3B-1 AMN COM4-1 TS4-1 DHAA COM5-1 TS5-1

−18.2 167.4 −42.7 −59.1 239.8 −30.7 −57.5 281.5 −70.8 269.7 −26.1 −40.5 257.4 −14.1 −45.6 211.7 9.8 −21.6 232.4

COM1-2 TS1-2 DIA COM2A-2 TS2A-2 COM2B-2 TS2B-2

−14.8 277.2 12.3 −6.2 181.6 −12.8 319.9

COM1-3 TS1-3 DEA COM2A-3 TS2A-3 COM2B-3 TS2B-3 DDA COM3A-3 TS3A-3 COM3B-3 TS3B-3 AMD COM4A-3 TS4A-3 COM4B-3 TS4B-3 DHA COM5-3 TS5-3 CA

−26.4 307.9 18.4 0.9 186.8 4.5 298.1 31.7 1.9 201.8 1.9 254.1 60.8 40.1 226.3 30.9 280.8 93.5 50.9 257.9 35.7

−380.3i

−334.6i

−464.9i −469.2i

−324.2i

−762.1i

−847.1i Path II −320.5i

−380.3i −435.3i Path III −441.0i

−397.6i −317.2i

−378.0i −836.5i

−342.8i −939.7i

−338.8i

−6.1 157.2 −36.2 −38.3 220.8 −23.2 −36.6 276.9 −50.3 263.6 −16.4 −16.9 240.2 −3.2 −15.3 209.4 20.1 8.1 230.9 −0.9 250.0 14.2 7.5 174.0 2.6 308.8 −12.6 295.3 21.0 15.6 180.6 21.0 269.4 36.8 25.5 197.5 25.5 246.7 63.2 53.4 218.7 52.8 312.2 95.1 69.8 252.4 47.2

−360.9i

−215.8i

−408.3i −413.9i

−208.3i

−419.6i

−466.2i

−195.9i

−374.8i −400.6i

−408.3i

−373.4i −185.4i

−382.9i −448.3i

−301.6i −466.2i

−296.7i

a

The other molecules in the reaction were considered. bThe results in gas. cThe results by CPCM.

Reactions in path III (shown in Figures 4 and 5) were similar to those in path II and included several steps: (1) C(9)−N(8) was hydrolyzed. DEA was generated via complex COM1-3 and transition state TS1-3 including a four-membered ring 4858



Article

Figure 7. Potential energy profiles of ATR and its metabolites in the three paths.

C(6)−N(10), those dissociation energy barriers were 256.3, 254.0, 256.0, and 257.2 kJ/mol, respectively, so the reaction energy barriers of the C(2)−Cl bond were obviously lower than those of the C−N bond. In this reaction, the hydrolysis reaction rate of ATR was slow because of higher reaction energy barriers. In general, nucleophilic reactions are an interaction between the HOMO of a nucleophile and the LUMO of an electrophile. In this reaction system, there was an interaction between the HOMO of the nucleophile H2O and the LUMO of ATR. The LUMO of ATR (the antibonding orbital of the C(2)−Cl bond) in Figure 8 shows that the bond of C(2)−Cl was elongated and dissociated by the interaction between the lone pair electrons of the nucleophile H2O and the LUMO of ATR, so the bond of C(2)−Cl was dissociated previously by the nucleophile H2O because of its lower energy barrier. In principle, an energy barrier has an exponential relationship with the rate constant. The reaction energy barrier of the C(2)−Cl bond was about 110−150 kJ/mol lower than that of the C(9)−N(8) or C(11)−N(10) bond, and its reaction rate constant was obviously higher than that of the C(9)−N(8) or C(11)−N(10) bond. Therefore, there were obvious kinetics differences in the three paths. The hydrolysis of the C(2)−Cl

Figure 8. Schematic diagrams of (a) LUMO of ATR and (b) HOMO of H2O.

H atom. The Cl was substituted in this step. The energy barriers of the hydrolysis of the bond of C(9)−N(8) were 340.5, 339.0, 332.7, and 334.3 kJ/mol. The energy barriers of the hydrolysis of the C(11)−N(10) were 298.9, 297.9, 292.0, and 293.6 kJ/mol. The vibrational frequencies of the C(9)− N(8) and C(11)−N(10) bonds were about 460i and 320i, respectively. Those vibration modes were all the swing of the H atom between N(8)/N(10) and the O atom, and the H atom was transferred in this step. For bonds of C(4)−N(8) and 4859



Article

Table 2. Wiberg Bond Orders (Bi), Percentage of Evolution of Bond Orders (Ev), and Synchronicity (Sy) BiRCa

BiTS1b

BiINTc

Ev

Sy

I

H−O H−Cl(7) C(2)Cl(7) C(2)O

0.6961 0.0022 1.0290 0.0011

0.6599 0.0457 0.4062 0.3445

0 0.9113 0 1.0520

5.20 4.78 39.48 32.68

0.495

II

H−O H−N(10) C(11)N(10) C(11)O

0.6757 0.0007 0.9378 0.0005

0.6006 0.1423 0.0865 0.0565

0 0.7853 0 0.9118

11.11 18.05 90.77 6.15

0.373

III

H−O H−N(8) C(9)N(8) C(9)O

0.6761 0.0006 0.9524 0.0005

0.5699 0.1686 0.1999 0.2064

0 0.7840 0 0.9176

15.71 21.44 79.01 22.45

0.575

path

bond

a The value for hydrogen bond complexes of the first step. bThe value for transition states of the first step. cThe value for intermediates of the first step.

bond was the easiest one in the three reactions: from ATR to HA, DIA, and DEA. In this hydrolysis reaction system (shown in Figure 3), the intermediate concentration in path I was the highest, such as HA in path I. The result corresponds to some experimental works.12,15,16 The prediction will be beneficial to expand the study of pesticide effect on the environment from the pesticide compound to the possible intermediates. The CPCM was used to simulate these reactions in aqueous state because of the solvent effect. The results showed that the dissociation energy barriers of the C(2)−Cl bond were 163.3, 166.5, 172.0, 159.6, 165.3, and 182.6 kJ/mol and lower than those in the gaseous state. Dissociation energy barriers of the C(9)−N(8) bond were 313.9, 313.2, 306.2, and 307.9 kJ/mol, respectively, and the dissociation energy barriers of the C(11)− N(10) bond were 259.1, 257.1, 250.9, and 248.4 kJ/mol. For the C(4)−N(8) and C(6)−N(10) bonds, the dissociation energy barriers were 224.7, 239.0, 221.2, and 259.4 kJ/mol. They were all lower than those in the gaseous state. It is necessary to consider the solvent effect in the study of potential energy surface in hydrolysis reactions. Bond Order Analysis. To gain insight into the various transition states along with the reaction coordinate, an analysis of bond order was studied. The value of Wiberg bond order was calculated by NBO calculation. With regard to the reaction mechanism, synchronicity (Sy), proposed by Moyano et al.,33 was used to represent the global nature of bond breaking/ forming processes in the decomposition reaction and expressed by eq 1.

The information about Wiberg bond orders (Bi) and percentage of evolution of bond orders (Ev) of several bonds of the first step in paths I, II, and III are listed in Table 2. The synchronicity(Sy) of the first step in paths I, II, and III was calculated by eqs 1, 2 and 3, the values being 0.495, 0.373 and 0.575, respectively. The corresponding reactions can be viewed as very asynchronous processes according to Sy. Toxicity of the Metabolites. The toxicities (experimental and prediction results) of ATR and a part of its metabolites are listed in Table 3. The information listed in Table 3 shows that prediction results corresponded to experimental results if the experimental data were public. The prediction results showed that (1) carcinogenic toxicity of ATR and its metabolites was positive, (2) mutagenic toxicity (AMES test) of DHHA and CA was positive, and (3) acute toxicity of ATR and its metabolites was at middle or low toxic levels. These results explained ATR’s harmful impact on the environment. In this work, the mechanism of hydrolysis reactions in the metabolism of ATR was studied by the DFT B3LYP/6-31++g (d, p) method. It is necessary to use CPCM to simulate its aqueous system. The results showed the following: (1) The three degradation mechanisms corresponded to the three hydrolysis paths, dealkylation of C(9)−N(8) and C(11)− N(10) bonds by H transferred, deamination of C(4)−N(8) and C(6)−N(10) bond by H transferred, and dehalogenation of C(2)−Cl(7) bond by Cl substituted. (2) The hydrolysis reaction of ATR was prior to path I, and the C−Cl was dissociated first because its dissociation activation energy was lower than that of the C(9)−N(8) or C(11)−N(10) bond. (3) The concentrations of intermediates in the hydrolysis depended on reaction rate. The results of toxicity prediction and experimental reports of ATR showed that it is carcinogenic. The prediction of its metabolites showed that they are carcinogenic. The results also showed that not only ATR but also it metabolites are harmful for the environment. We propose that we should evaluate the toxicity of not only candidates of pesticides but also their metabolites by in silico methods, which also could be employed before the candidates were synthesized or isolated and in the period of design.

n

Sy = 1 − [∑ |(Ev )i − (Ev )av | /(Ev )av ]/(2n − 2) i=1

(1)

In eq 1, n denotes the number of bonds directly involved in the reaction, and the percentage of evolution of bond orders (Ev)i is given by eq 2. (Ev )i = 100 × [BiTS − BiR ]/[BiP − BiR ]

(2)

Superscipts TS, R, and P refer to the transition state, reactant, and product, respectively. The average (Ev)av is calculated by eq 3: (Ev )av =

1 n

n

∑ (Ev )i i=1

(3) 4860



Article

Table 3. Toxicities of ATR and Its Metabolites

4861



Article

Table 3. continued

a Rat, oral,

Hydrolysis Reaction Mechanism in Atrazine Metabolism and

Recommend Documents