Chlorobenzene

Nonspecific DAP metabo-lites (nanomoles per liter) were summed and transformed to the log₁₀ scale. We created “pregnancy” DEs, DMs, and total DAP values by averaging the two log-transformed pregnancy measures. The two pregnancy total DAP measurements were correlated (r = 0.14, p = 0.005) and did not significantly differ (paired t-test = −0.28, p = 0.78). For 29 women, only one DAP measurement was available. Prenatal and postnatal DAP measures were uncorre-lated, so we placed both exposures into a single model for each outcome. Coefficients were similar to those in models containing either prenatal or postnatal exposures alone. Because a large proportion of women had nonde-tectable levels of MDA and TCPy, we categorized levels into three groups for each metabolite: < LOD for both pregnancy measurements, and for those with at least one detectable level, subdivided below and above the median of the average pregnancy level.
To assess the relationship between metabo-lite levels and Bayley performance, we constructed separate multiple regression models for MDI and PDI at each of the three time points: 6, 12, and 24 months. We evaluated MDI and PDI continuously using linear regression. We included the same covariates in all Bayley models. Covariates were selected for these analyses if they were related to conditions of testing [i.e., psychometrician (n = 4), location (office or RV), exact age at assessment]; related to neurodevelopment in the literature and associated (p < 0.10) with most outcomes [i.e. sex, breast-feeding duration (months), HOME score (continuous), and household income]; or consistently related to neurodevel-opment in the literature even if not in our data [i.e., parity and maternal PPVT (continuous)]. We classified household income as above or below poverty by comparing total household income to the federal poverty threshold for a household of that size (U.S. Census Bureau 2000 ). In addition to the variables we included, we examined the potential confounding effects of several other variables suggested by the literature (i.e., maternal age, education, depressive symptoms, active/passive smoking exposure during pregnancy, regular alcohol use during pregnancy, marital status, father’s presence in home, housing density, maternal work status, ≥ 15 hours out-of-home childcare/week), but they did not markedly alter the observed associations. For simplicity, the same set of covariates was used for CBCL models with three exceptions: maternal depression, found to be important (p < 0.10), was added, and psychometrician and assessment location were dropped, because scores were based on maternal report. Covariates in final models were categorized as noted in Table 1, unless otherwise specified above. To preserve the size of the analytic population, each missing covariate value was imputed by randomly selecting a value from participants with non-missing values. Maternal depression had the largest percentage of values requiring imputation (5%). Of remaining covariates, between 0% and 1.8% of values were imputed.
In secondary analyses, we controlled for some factors potentially on the causal pathway (birth weight, gestational age, abnormal reflexes) and re-ran models excluding low birth weight and preterm infants. We also considered whether controlling for other suspected neuro-toxicants (i.e., PCBs, lead, and DDT) and other high-level exposures in our population (i.e., β-hexachlorocyclohexane and hexa-chlorobenzene) (Fenster et al. 2006 (link)) altered our results for DAPs in the subsample with both DAPs and the other chemical. Furthermore, we examined interactions between child DAPs and child sex and, because we previously observed an association with maternal DDT and 24-month MDI (Eskenazi et al. 2006 (link)), between maternal DAPs and DDT. Finally, we re-ran models using log-transformed creatinine-adjusted metabolites for comparative purposes.
In addition, we performed longitudinal data analyses [generalized estimating equations (GEE)] of the relationship between DAPs and Bayley scores, which produced similar findings. These GEE models included indicators for age at assessment and interaction terms for most independent variables and age. We obtained a minor increase in precision when the effects of some potential confounders were assumed constant over time. In addition, assuming a single effect of DAPs over time also produced small increases in precision of the relevant regression coefficient estimators, but again the gains were slight. Thus, we only present the results from the cross-sectional analyses for ease of understanding.

Set up of the system: Our model set‐up starts from the crystal structure coordinates of substrate‐ and α‐KG‐bound TauD as deposited by the 1OS7 pdb file, which is an enzymatic monomer.^[9] Hydrogen atoms were added to the structure based on pH 7 conditions with the H++ webserver.^[26] The MCPB.py tool as implemented in the AMBER‐18 software package was used to generate the force field parameters for a penta‐coordinated iron(II) centre, while for the description of the protein atoms the ff14SB4 force field parameters were used.[²⁷, ²⁸] Substrate taurine and α‐KG were described with the second generation General Amber Force Field (GAFF2) using AM1‐BCC charges.^[29] These GAFF2 parameters were generated with the antechamber and parmchk2 modules of the AMBER‐18 software package. Finally, the substrate‐bound enzyme structure was solvated in a cubic box of TIP3P water molecules and subsequently neutralized by adding Na⁺ ions. Thereafter, a molecular dynamics (MD) simulation was carried out using the Particle Mesh Ewald Molecular Dynamics (PMEMD) module as implemented in the AMBER‐18 software package.^[30] The system was minimized using 2000 steps of steepest decent minimization with a starting restraint potential of 2.0 kcal mol⁻¹ on all heavy atoms of the protein, which was gradually released. After that the system was heated for 100 ps from a temperature of 0 K to 298.15 K under NVT ensemble conditions with the Langevin thermostate and subsequently equilibrated for 1 ns under NPT conditions using the Berendsen barostate without constraints to the energy and the structure.^[31] Thereafter, a 20 ns MD simulation under NPT conditions was carried out in 2 fs time‐steps, using the SHAKE protocol on hydrogen atoms and a 10 Å non‐bonded cut‐off with periodic boundary conditions. The MD simulation (Supporting Information, Figures S1 and S2) shows that the system is highly rigid and little changes with respect to the crystal structure coordinates are seen.
QM/MM approaches: QM/MM calculations follow procedures as reported previously.^[32] In general, the QM/MM calculations were carried out using ONIOM scheme as implemented in the Gaussian software package.[³³, ³⁴] The QM/MM calculations were setup using molUP, a VMD plugin, from the equilibrated final structure of the MD simulation.^[35] The QM/MM model D consist of the complete protein, substrate and water molecules within 15 Å of the protein and has a total of 9,551 atoms and has His₇₀ in its singly protonated form. The QM region consist of the first‐coordination sphere around the iron(IV)‐oxo group and substrate and the link‐atom procedure was used to bridge the QM and MM regions.^[36] QM region A has 53 atoms, while QM region B contains the atoms as shown in Figure 2a. The QM region is treated using DFT with the unrestricted B3LYP hybrid functional and a basis set with LANL2DZ (with electron core potential) on iron and 6‐31G* on the rest of the atoms: basis set BS1,[³⁷, ³⁸] while the MM region was treated with the Amber force field.[²⁸, ³⁰] We initially ran QM/MM geometry optimizations using QM region A and followed these up with a set of single point calculations whereby the QM region was expanded to the atoms shown in Figure 2 under model B. The electrostatic interactions between the QM and MM regions were described with the electronic embedding scheme. The geometry optimizations were performed with UB3LYP/BS1 level of DFT and followed by an analytical frequency calculation at 298.15 K and 1 atm. The transition states were located using potential energy scan calculations followed by full geometry optimization with the Berny algorithm. The energy values were corrected through single point calculations at the ONIOM(UB3LYP/BS2:Amber) level of theory with BS2 a basis set representing LACV3P+ (with electron core potential) on iron and 6‐311+G* on the rest of the atoms.
The M06 density functional method was also tested for the QM region in the QM/MM calculations,^[39] however this gave essentially similar energetics than those found with B3LYP, see Supporting Information.
DFT cluster models: Cluster models focus on the substrate and oxidant binding environment and the second‐coordination sphere and consider all atoms in the model with a quantum chemical approach.^[40] As the MD simulation gives little variety between the active site orientation of the various snapshots, we created five cluster models (A, B, B2, C and C2) based on the last step of the MD simulation, see Figure 2. We truncated the active site model by selecting the residues of amino acids and second coordination sphere groups, which determine substrate and oxidant binding and positioning and particularly include all polar (hydrogen bonded and π‐stacking interactions) close to the metal and substrate. In the active site model we replaced the iron(II) ion by iron(IV)‐oxo species, while α‐KG was replaced by succinate. A truncated model (minimal) cluster model was considered, namely a minimal cluster model A that contains only the first coordination sphere of residues to iron and the substrate and had 72 atoms. Thereafter, larger cluster models B and C were created that incorporate the environments around substrate and oxidant and their hydrogen bonding interactions. The active site cluster model B incorporates the metal and its first‐coordination sphere as well as the substrate and a number of second‐coordination sphere residues that determine the size and shape of the substrate binding pocket and incur hydrogen bonding interactions. Thus, the model includes a short protein chain of amino acids that links to the equatorial ligands of the iron, namely His₉₉−Thr₁₀₀−Asp₁₀₁−Val₁₀₂. In addition, two short chains for Ser₁₅₈−Phe₁₅₉ and Asp₉₄−Asn₉₅ were included. The axial histidine ligand (His₂₅₅) of iron was abbreviated to methylimidazole and α‐KG replaced by acetate. Finally, the side chains of the residues His₇₀, Tyr₇₃, Ile₈₃, Asn₉₇, Phe₁₀₄, Phe₂₀₆ and Arg₂₇₀ as well as substrate taurine were included. Overall our cluster model contains 244 atoms and has a neutral charge. To prevent the model from changing dramatically in structure during the geometry optimizations we placed some constraints on some of the C^α‐protein atoms as identified with a star in Figure 2. Cluster Model C contains 279 atoms and is based on the QM/MM optimized geometry and had the atoms of cluster model B expanded with the full peptide chain Asp₉₄−Asn₉₅−Asp₉₆−Asn₉₇−Trp₉₈−His₉₉−Thr₁₀₀−Asp₁₀₁−Val₁₀₂−Thr₁₀₃−Phe₁₀₄ with Asp₉₆, Trp₉₈ and Thr₁₀₀ truncated to a Gly residue. In addition, model C has the His₇₀ residue doubly protonated. To test the protein environment on the kinetics and selectivity of the reactions, we created two further models based on models B and C, namely B2 and C2. Model C2 is similar to model C but has His₇₀ singly protonated, while model B2 is model B with the chains of Asp₉₄, Asn₉₅, Ser₁₅₈ and Phe₁₅₉ removed.
DFT procedures: The Gaussian‐09 software package was used for all quantum chemical calculations discussed here.^[33] Following previous experience with cluster models of nonheme iron complexes,^[41] we utilized the unrestricted B3LYP density functional method in combination with a LANL2DZ (with electron core potential) on iron and 6‐31G* on the rest of the atoms: basis set BS1.[³⁷, ³⁸] This method was shown to reproduce experimental spin‐state assignments and rate‐constants well.^[42] All geometry optimizations were performed with these basis sets in which all local minima were verified by the absence of negative eigenvalues in the vibrational frequency analysis while all the transition state structures were found using the Berny algorithm, and verified by vibrational analysis and visualized by animating the imaginary frequency. For key transition states also intrinsic reaction coordinate calculations were done, which confirmed the transition states to connect with the local minima as identified. In order to improve the energetics, single point energies on the optimized geometries were calculated at UB3LYP/BS2, whereby BS2 represents the LACV3P+ (with electron core potential) on iron and 6‐311+G* on the rest of the atoms. The latter set of calculations included a continuum polarized conductor model (CPCM) with a dielectric constant mimicking chlorobenzene and the dispersion correction (D3) developed by Grimme.^[43] For a selection of structures, we ran full geometry optimizations at UB3LYP‐D3/BS1 level of theory, but these studies gave the same trends as those obtained at UB3LYP/BS1, see Supporting Information for details. To obtain the free energies at 298.15 K and 1 atm, the zero‐point energy (ZPE), thermal corrections and entropy contribution evaluated from the unscaled vibrational frequencies at the UB3LYP/BS1 level of theory were then added to the electronic energies calculated from the same level of DFT.
The primary kinetic isotopic effect (KIE′s) were calculated using the classical equations due to Eyring (Eq. (1)) and with tunnelling corrections included due to Wigner (Eq. (2) and 3).^[44] In the Eyring KIE, the activation free energy (ΔG^≠) of hydrogen and deuterium substituted reaction was considered in the gas phase at room temperature (T=298.15 K) with R being the gas constant. By contrast, the Wigner tunnelling factor incorporates a factor that considers the change in the magnitude of the imaginary vibrational frequency in the transition state for the hydrogen and deuterium substituted systems. In Equation (3), h is Planck's constant and k_B is the Boltzmann constant, The KIE values for the various hydrogen atom abstraction pathways were calculated by replacing the both pro‐R and pro‐S hydrogen atoms at the C¹ and C² positions of taurine substrate. $${\displaystyle \mathrm{KIE}{}_{\mathrm{Eyring}}=\exp [(\Delta \mathrm{G}{}^{\ne }{}_{\mathrm{D}}-\Delta \mathrm{G}{}^{\ne }{}_{\mathrm{H}})/\mathrm{RT}]}$$
$${\displaystyle \mathrm{KIE}{}_{\mathrm{Wigner}}=\mathrm{KIE}{}_{\mathrm{Eyring}}\times \mathrm{Q}{}_{\mathrm{t},\mathrm{H}}/\mathrm{Q}{}_{\mathrm{t},\mathrm{D}}}$$
$${\displaystyle \mathrm{Q}{}_{\mathrm{t}}=1+(\mathrm{h}\nu /\mathrm{k}{}_{\mathrm{B}}\mathrm{T})){}^2/24}$$
Finally, we performed the electric field effect (EFE) calculations as implemented in Gaussian software package using the “field” keyword.^[33] The EFE is calculated by performing a single point energy calculation on various optimized geometries with an electric field perturbation along the molecular x, y or z‐axis with the positive direction as defined in Gaussian. These electric field perturbations were done with various field magnitudes in either positive or negative field directions along each principal axis.

A scanning electron microscope (SEM) (JSM-6010LV, JEOL) was used to observe the shape of micro and nano CdO particles and cross section morphologies of CdO/HDPE composites. To prepare the samples for SEM observation, the samples were coated with an ultrathin gold coating using a low-vacuum sputtering coating device (JEOL-JFC-1100E). The SEM images were obtained at magnification order of 5,000x at 20 KV.
The γ-ray spectrometric measurements were performed using 100 cm³ well calibrated Hyper pure germanium cylindrical detector (HPGe) from Canberra (Model GC1520) in conjunction with multichannel analyzer (MCA). The detector has a resolution of 1.85 keV at 1.33 MeV gamma ray peak ⁶⁰Co and relative efficiency of 15% in the energy range from 50 keV to 10 MeV²⁰ . The control of acquisition parameters and analysis of the collected spectra was carried out using ISO 9001 Genie 2000 data acquisition and analysis software fabricated by Canberra. The detector was housed in a lead shielding of 15 cm thickness to diminish the background radiations. The radiation measurements were done by using five radioactive sources of ²⁴¹Am, ¹³³Ba, ¹³⁷Cs, ⁶⁰Co and ¹⁵²Eu purchased from Physikalisch-Technische Bundesanstalt PTB in Braunschweig and Berlin. The emitted energies corresponding to these radioactive sources are listed in Table 1. The radioactive sources were placed at 508.67 mm from the detector surface to get very narrow beam and also to ignore the effect of detector dead time^{21 (link)}. The produced composite of 2.5 mm thickness was placed on a holder between the standard gamma point source and detector. The setup and geometry of the measurement system is displayed in Fig. 1.Table 1

Radioactive source and its corresponding photon energy.

source	Photon energy (keV)
241Am	59.53
133Ba	80.99
	356.01
137Cs	661.66
60Co	1173.23
	1332.5
152Eu	121.78
	244.69
	344.28
	778.9
	964.13
	1408.01

Figure 1

The experimental setup for examining γ-ray shielding property.

During measurement, the photon beam generated from the radioactive sources reacted with the sample and detected by HPGe crystal detector. Electrical signal generated by the detector was amplified and then analyzed using Genie 2000 software by choosing a narrow region symmetric with respect to the centroid of photon peak. The data acquisition time was high enough to give <1% count error. The net area under the photo peak was determined and then the count rate (N) was calculated.
The linear attenuation coefficient µ (cm⁻¹) of each composite material can be obtained for γ-ray of appropriate energy according to Lambert- Beer law^{22 (link)} given by Eq. (1) $$\mu =\frac{1}{x}\ln \left[\frac{N_{(0)}}{N_{(x)}}\right]$$ where x is the thickness of the sample, N₍₀₎ and N_(x) are the detector count without and with the composite target. The linear attenuation coefficient µ can be evaluated as the slope of the best fitted line of a linear relation between $\ln \left[\frac{N_{(0)}}{N_{(x)}}\right]$ versus sample thickness x.
The mass attenuation coefficient µ_m (cm²/g), which is an important parameter for characterizing the interactions of γ-rays with matter, can be determined by dividing µ by the measured density (ρ) of the sample²³ .
The average density of each composite sample was determined accurately by applying Archimedes technique according to ASTM D 792-91²⁴ . For this purpose, a calibrated single pan electrical balance with accuracy 0.0001 g and three organic liquids such as ethanol, toluene and chlorobenzene were used.

The gaseous concentration of CB was evaluated using GC (Agilent 6890, USA). The temperatures of the inlet, column (HP-Innowax), and hydrogen flame ionisation detector were 250 °C, 200 °C, and 280 °C, respectively. The liquid concentration of CB was calculated using the gas–liquid partition coefficient from the gas concentration of CB according to our previous work (Chen et al. 2018 (link)). An Agilent 6890 GC (Santa Clara, CA, USA) was used to measure the concentrations of CO₂, and the temperatures of the inlet, column (HP-Plot-Q, 30 m × 320 µm × 20 µm), and thermal conductivity detector were 100 °C, 40 °C, and 180 °C, respectively.

The performance of catalytic combustion for chlorobenzene was evaluated in a fixed-bed reactor consisting of a quartz tube (i.d. = 4 mm) under standard atmospheric pressure. In each experiment, 100 mg of catalysts (40–60 mesh) was immobilized using quartz wool, resulting in a gas hourly space velocity (GHSV) of 20,000 mL·g⁻¹ h⁻¹ and a total flow rate of 33.33 mL·min⁻¹. The liquid reactant (chlorobenzene) was introduced into the feed stream by passing dried air through a saturator maintained at a temperature of 6 °C. The feed stream was then diluted with dried air generating a feeding flow containing 1000 ppm reactant and 21% O₂/79% N₂. Flow rates were regulated using online mass flowmeters. Concentrations of CB and chlorinated byproducts were determined via analysis on an online gas chromatograph (GC3420A, Beifen Ruili Co., Ltd., Beijing, China) equipped with both flame ionization detector (FID) and electron capture detector (ECD) for quantitative assessment of organic compounds. Furthermore, the outlet CO₂ underwent reduction by hydrogen within a methanation furnace before being detected by the same chromatograph system. A K-type thermocouple was inserted into the catalyst to monitor its temperature during the catalytic performance test. Data of catalytic oxidation tests were collected after achieving steady state conditions for at least thirty minutes. The detailed structure of the catalytic test system is shown in Figure 1.
The CB conversion rate ( $X_{CB}$ , %), CO₂yield ( ${CO}_{2}yield$ , %) and organic byproduct yield ( $Y_b$ , %) were calculated as follows: $$X_{CB}=\frac{{\left(\mathrm{C}\mathrm{B}\right)}_{in}-{\left(\mathrm{C}\mathrm{B}\right)}_{out}}{{\left(\mathrm{C}\mathrm{B}\right)}_{in}}\times 100\mathrm{\%},$$
$${CO}_{2}yield=\frac{({{\mathrm{C}\mathrm{O}}_2)}_{out}}{6\times {\left(\mathrm{C}\mathrm{B}\right)}_{in}}\times 100\mathrm{\%},$$
$$Y_b=X_{CB}-{CO}_{2}yield,$$
where ${\left(\mathrm{C}\mathrm{B}\right)}_{in}$ and ${\left(\mathrm{C}\mathrm{B}\right)}_{out}$ are the inlet and outlet concentrations of chlorobenzene, respectively. $({{\mathrm{C}\mathrm{O}}_2)}_{out}$ is the outlet concentrations of CO₂.
The reaction rate of 1,2-dichloroethane was calculated as follows: $$\mathrm{r}=\frac{Q\times X_{CB}}{m}$$
where r is the reaction rate (mol·g⁻¹·s⁻¹), Q is the molar flow of chlorobenzene (mol·s⁻¹), and m is the mass of the catalyst (g).
The stability test for chlorobenzene was performed continuously at 290–320 °C for 100 h under the same conditions as the catalytic combustion test, unless otherwise indicated.

Manganese (II) nitrate solution (A.R. grade, Mn(NO₃)₂, 50 wt%), Chromium (III) nitrate (A.R. grade, Cr(NO₃)₃·9H₂O), Terephthalic acid (A.R. grade, C₈H₆O₄), and N,N-dimethylformamide (DMF) (A.R. grade, C₃H₇NO) were purchased from Aladdin Biochemical Technology Co.,Ltd., Shanghai, China. Ethanol (A.R. grade, C₂H₆O) was purchased from Nanshi Chemical Reagent Co. Ltd., Nanjing, China. Chlorobenzene (A.R. grade, C₆H₅Cl), Zirconium tetrachloride (A.R. grade, ZrCl₄), and Glacial acetic acid (A.R. grade) were purchased from Merrill Chemical Technology Co. Ltd., Shanghai, China. All chemicals were used without further treatment.