Quantitative structure activity relationship (QSAR) modelling of the degradability rate constant of volatile organic compounds (VOCs) by OH radicals in atmosphere

https://doi.org/10.1016/j.scitotenv.2020.138871Get rights and content

Highlights

  • A new QSAR for predicting the rate constant of OH radical degradation of VOCs

  • EHOMO, f (0)n and BOx have great influence on the degradation of VOCs.

  • The proposed model has been verified internally and externally.

  • Recommended model has broader applicability domains.

Abstract

The reaction with hydroxyl radicals (•OH) is an important way to remove the most volatile organic compounds (VOCs) in atmospheric environment. Thus, the reaction rate constant (kOH) is important for assessing the persistence and exposure risk of VOCs, and is of great significance in evaluating the ecological risk of volatile organic chemicals. Fukui indices and bond order have a large effect on the degradation of VOCs, but so far, quantitative structure activity relationship (QSAR) models for VOCs degradation have rarely been considered these two factors. In this study, these two momentous factors will be considered along with other relevant quantitative parameters. A total of 180 substances are divided into training set (144 substances) and test set (36 substances), which are used to build and validate quantitative structure activity relationship (QSAR) models, respectively. Internal, external verification and y-randomization tests showed that the established model had excellent stability and reliability. The energy of the highest occupied molecular orbital (EHOMO), the possibility of being attacked by radicals (f (0)n) and the breaking of chemical bonds (BOx) are the main factors affecting VOCs removal. Finally, the scope of the application domain was determined and the robustness of the model was further verified.

Introduction

Volatile organic compounds (VOCs) refer to organic compounds with a melting point below room temperature and a boiling point between 50 and 260 °C. VOCs from coal chemical, petrochemical and other processes are considered to be important precursors for air pollution such as urban haze and photochemical smog (Guo et al., 2019; He et al., 2015; Villanueva et al., 2015). The sources of VOCs in the atmosphere mainly include anthropogenic sources and natural sources (Gupta et al., 2015; Sahu, 2012). On a worldwide basis, the contribution of natural sources to VOCs exceeds that of human sources. However, in urban areas, anthropogenic VOCs often dominate (Atkinson and Arey, 2003). VOCs participate in the formation of ozone and secondary aerosols in the atmospheric environment, which have an important impact on regional atmospheric ozone pollution and particulate matter pollution. Most VOCs have unpleasant special odors and are toxic, irritating, teratogenic and carcinogenic, especially benzene, toluene and formaldehyde, which cause great harm to human health (Luan et al., 2006; Nurmatov et al., 2015; Tsai, 2016). Therefore, it is imperative to control the emission of VOCs, and this action has been effectively implemented in many countries and regions (Everaert and Baeyens, 2004; Parmar and Rao, 2008).

In addition to the physical processes (wet and dry depositions), chemical transformations constitute the main degradation pathway and mechanism of VOCs in troposphere (Atkinson, 2000; Atkinson and Arey, 2003). The reaction of hydroxyl radicals (•OH) and ozone during daytime and nitrate radicals (NO3) at night constitutes the main degradation process of VOCs (Atkinson, 1986; Atkinson, 2000; Khan et al., 2015). It can be seen that the reaction between VOCs and free radicals is of direct importance to various chemical subdisciplines including atmospheric chemistry, and the relevant reaction rate is of great significance for evaluating the behavior and fate of VOCs in the atmosphere (Chen et al., 2014; McGillen et al., 2007). Since the reaction rate of hydrocarbon and hydroxyl radical is about 30 times that of nitrate radical (Schindler, 2016). Therefore, the reaction of VOCs with hydroxyl radicals in the atmosphere has been concerned for a long time (Atkinson, 1987; Bakken and Jurs, 1999; Luo et al., 2017; Markelj and Pompe, 2016; Pompe et al., 2004). The photolysis of O3 forms excited oxygen and reacts with water vapor to form OH radicals in the troposphere, and the global mean concentration of OH radicals has been estimated at 1.5 × 106 molecules·cm−3 (Mao et al., 2009). Since the concentration of OH radical is much lower than most VOCs, it can be considered that the degradation kinetics of organic compounds conforms to the pseudo-first-order kinetics (Gusten, 1999; Öberg, 2005). As the experimental derivation of VOCs reaction rate is very demanding, they can only be used for a limited number of chemicals (Schindler, 2016). Moreover, the experimental method is time-consuming and laborious, and may also cause a large error in results (Basant and Gupta, 2018; Chen et al., 2014; Gupta et al., 2016). In order to reduce unnecessary effort, it is worthwhile to find a reliable way to determine VOCs degradation rate instead of an experimental method.

The theoretical prediction is considered to be a fast and economic evaluation method and among them, the quantitative structure activity relationship (QSAR) method has been widely used to predict the gas phase kinetic rate constants of different chemicals and •OH reactions. The Atkinson's group/fragment contribution method seems to be the most commonly used estimation method (Kwok and Atkinson, 1995). Meylan and Howard (2003) predicted the kOH value of 720 organics based on this method and the predicted value of kOH for about 90% of the compounds is within 2 times of the experimental value. However, due to the fact that the experimental kOH value of some compounds was lower than a specific value, the group contribution method could not predict them accurately, and the model correlation was not good, so about 30 compounds were excluded from the data set. Although the group contribution method can accurately predict the kOH value of organics with similar structures to the data set, the prediction results of organics with significantly different structures from the data set lack reliability, so it is not recommended (Bakken and Jurs, 1999; Gramatica et al., 2004b; Kwok and Atkinson, 1995; Öberg, 2005). In addition, rigorous validation and well-defined application areas are key factors for the success of QSAR models, and these steps have often been lacking in past model development (Atkinson, 1987; Peeters et al., 2007; Pompe et al., 2004; Xu et al., 2013).

Partial least square regression (PLSR) (Long and Niu, 2007; Öberg, 2005; Wang et al., 2009) and multiple linear regression (MLR) (Fatemi and Baher, 2009; Gramatica et al., 2002; Gramatica et al., 2004a; Gramatica et al., 2004b; Gusten et al., 1995; Huang et al., 2012; Li et al., 2014; Roy et al., 2011; Xu et al., 2013) have a wider application. In addition, there are other regression models used by researchers, such as artificial neural networks (ANNs) (Gramatica et al., 1999), support vector machines (SVMs) (Yu et al., 2015) and genetic algorithms (GAs) (Fatemi and Baher, 2009; Yu et al., 2015). These studies investigated the relationship between VOCs degradation rates and their quantum chemical, constitutional, electrical, physicochemical, and topological descriptors in the atmosphere. Based on the molecular structure of a compound, quantum chemical calculations can not only predict its reactivity, reaction kinetics and thermodynamic data, but also give the corresponding reaction rate constant by combining with the QSAR method (Han et al., 2014; Luo et al., 2011; Zhou et al., 2011). Therefore, the relationship between pollutant structure and degradation behavior estimated by quantum chemical parameters has long been sought after by researchers (Huang et al., 2012; Li et al., 2014; Roy et al., 2011; Xu et al., 2013). Long and Niu (Long and Niu, 2007) have pointed out that when the highest occupied molecular orbital energy (EHOMO), the lowest unoccupied molecular orbital energy (ELUMO) and ELUMO + EHOMO are higher, the corresponding alkyl naphthalene reaction rate constants would be higher. Schindler (2016) built a prediction model for the reaction rate constant between VOCs and nitrate radicals, which further verified the importance of EHOMO to the degradation of pollutants. Besides, the ionization potential is also an important descriptor. The model established by Yu et al. (2015) also introduced frontier orbital energy, and introduced other quantitative parameters such as dipole moment and atomic charge. Luo et al. (2017) have found that polarizability is highly correlated to Hammett constant based on meta–substituents, and a new model established based on this had good predictability. Li et al. (2014) also established a new model and found GATS1e (Geary autocorrelation of lag 1 weighted by Sanderson electronegativity) as a main descriptor. A positive coefficient of 0.5870 confirms that •OH tends to attack molecules with high electronegativity. However, Fukui indices as key indicators of affinity for the attack in organic compounds degradation reactions are seldom taken into considerations. As important descriptors of density functional theory, they have been used to predict the success of certain chemical processes (Roos et al., 2009). Bond order, as another important quantitative parameter, in general, when all bond orders value of a molecule are <4, the stability of the molecule will stronger with the increase of the bond order value (Cheng et al., 2018a; Su et al., 2016).

The objective of this work was to establish a new QSAR model by involving Fukui indices, bond orders and other appropriate quantum chemical parameters, and then to find a general VOCs degradation rule and provide a reliable way to prediction for reaction rate constants other VOCs and OH radical. In this study, a total of 180 pollutants were selected and divided into two groups, of which 144 substances in the training set were used to build the model, and 36 substances in the test set were used to check the stability of the model. The new QSAR model was strictly validated using the internal and external validation procedures. In addition, the new model was also subjected to the Y-randomization test and visualization of the applicability domain (APD).

Section snippets

Experimental data sets

Various chemical groups such as alkane, alkene, aldehyde, aromatic hydrocarbon, ketone, alcohol, ether, dicarbonyl and unsaturated carbonyl, and alkyl nitrate were selected to make data values more representative. Reaction rate constants (kOH) about experimental values (cm3 molecule−1 s−1) in this study were collected from the literature and this data set contained values of 180 chemicals at room temperature (298 K) (Atkinson and Arey, 2003; Li et al., 2014; Markelj and Pompe, 2016; Wang et

Computation results

Tables S1 to S4 show the 18 quantized parameters for all considered organic compounds. As shown in tables, the average value of total energy (E(B3LYP)) is −281.741 kcal/mol, but the largest value (−40.543) is >14 times larger than the smallest value (−586.113), suggesting a high structural diversity of the compounds in two sets. It is worth noting that 22 of the 180 organics have a dipole moment (μ) value of 0, and the dipole moment (μ) value of other organics range from 0.001 Debye

Conclusions

A QSAR model with -logkOH as the dependent variable and 18 quantization parameters as independent variables was established using the MLR method. Relevant indices show that the model has good performance (R2 = 0.785, q2 = 0.754, Qext2 = 0.642). Three quantization descriptors, including EHOMO, f (0)n and BOx have the greatest influence on the degradation of VOCs, among which EHOMO and BOx are negatively positively correlated with -logkOH, while.

f (0)n is positively correlated with -logkOH.

CRediT authorship contribution statement

Yawei Liu: Conceptualization, Methodology, Software, Investigation, Writing - original draft. Zhiwen Cheng: Validation, Software. Shiqiang Liu: Writing - review & editing, Supervision. Yujia Tan: Writing - review & editing, Supervision. Tao Yuan: Supervision, Writing - review & editing. Xiaodan Yu: Supervision, Writing - review & editing. Zhemin Shen: Supervision, Writing - review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported by National water pollution control key project, China (2017ZX07202005-005) and the Medicine & Engineering Collaborative Research Fund of Shanghai Jiao Tong University (No. YG2017ZD15), China.

References (70)

  • D. Han et al.

    Theoretical investigation on the mechanisms and kinetics of OH-initiated photooxidation of dimethyl phthalate (DMP) in atmosphere

    Chemosphere

    (2014)
  • Z. He et al.

    Pollution characteristics and health risk assessment of volatile organic compounds emitted from different plastic solid waste recycling workshops

    Environ. Int.

    (2015)
  • L. Jia et al.

    QSAR models for oxidative degradation of organic pollutants in the Fenton process

    J. Taiwan Inst. Chem. E.

    (2015)
  • M.A.H. Khan et al.

    Global modeling of the nitrate radical (NO3) for present and pre-industrial scenarios

    Atmos. Res.

    (2015)
  • K. Khan et al.

    Ecotoxicological QSAR modeling of endocrine disruptor chemicals

    J. Hazard. Mater.

    (2019)
  • E.S.C. Kwok et al.

    Estimation of hydroxyl radical reaction rate constants for gas-phase organic compounds using a structure-reactivity relationship: An update

    Atmos. Environ.

    (1995)
  • C. Li et al.

    Development of a model for predicting hydroxyl radical reaction rate constants of organic chemicals at different temperatures

    Chemosphere

    (2014)
  • X. Long et al.

    Estimation of gas-phase reaction rate constants of alkylnaphthalenes with chlorine, hydroxyl and nitrate radicals

    Chemosphere

    (2007)
  • F. Luan et al.

    Quantitative structure-activity relationship models for prediction of sensory irritants (logRD50) of volatile organic chemicals

    Chemosphere

    (2006)
  • S. Luo et al.

    A novel model to predict gas-phase hydroxyl radical oxidation kinetics of polychlorinated compounds

    Chemosphere

    (2017)
  • J. Markelj et al.

    Modeling of atmospheric OH reaction rates using newly developed variable distance weighted zero order connectivity index

    Atmos. Environ.

    (2016)
  • T. Öberg

    A QSAR for the hydroxyl radical reaction rate constant: validation, domain of application, and prediction

    Atmos. Environ.

    (2005)
  • M. Schindler

    A QSAR for the prediction of rate constants for the reaction of VOCs with nitrate radicals

    Chemosphere

    (2016)
  • K. Takahashi et al.

    Selective synthesis of unsymmetric dibenzo[a,e]pentalenes by a rhodium-catalysed stitching reaction

    Chem. Sci.

    (2017)
  • F. Villanueva et al.

    Levels and sources of volatile organic compounds including carbonyls in indoor air of homes of Puertollano, the most industrialized city in Central Iberian Peninsula. Estimation of health risk

    Int. J. Hyg. Environ. Health

    (2015)
  • Y-n. Wang et al.

    Predicting rate constants of hydroxyl radical reactions with organic pollutants: algorithm, validation, applicability domain, and mechanistic interpretation

    Atmos. Environ.

    (2009)
  • Z. Yang et al.

    Rate constants of hydroxyl radical oxidation of polychlorinated biphenyls in the gas phase: a single-descriptor based QSAR and DFT study

    Environ. Pollut.

    (2016)
  • X.L. Yu et al.

    Predicting reaction rate constants of ozone with organic compounds from radical structures

    Atmos. Environ.

    (2012)
  • H. Zhu et al.

    Degradation mechanism study of organic pollutants in ozonation process by QSAR analysis

    Chem. Eng. J.

    (2014)
  • H. Zhu et al.

    QSAR models for degradation of organic pollutants in ozonation process under acidic condition

    Chemosphere

    (2015)
  • S.A. Amin et al.

    An integrated QSAR modeling approach to explore the structure–property and selectivity relationships of N-benzoyl-l-biphenylalanines as integrin antagonists

    Mol. Divers.

    (2018)
  • R. Atkinson

    Kinetics and mechanisms of the gas-phase reactions of the hydroxyl radical with organic compounds under atmospheric conditions

    Chem. Rev.

    (1986)
  • R. Atkinson

    A structure-activity relationships for the estimation of rate constants for the gas-phase reactions of OH radicals with organic compounds

    Int. J. Chem. Kinet.

    (1987)
  • R. Atkinson et al.

    Atmospheric degradation of volatile organic compounds

    Chem. Rev.

    (2003)
  • G.A. Bakken et al.

    Prediction of hydroxyl radical rate constants from molecular structure

    J. Chem. Inf. Comp. Sci.

    (1999)
  • Cited by (24)

    • QSAR models on degradation rate constants of atmospheric pollutants

      2023, QSAR in Safety Evaluation and Risk Assessment
    • A quantitative structure activity relationship (QSAR) model for predicting the rate constant of the reaction between VOCs and NO<inf>3</inf> radicals

      2022, Chemical Engineering Journal
      Citation Excerpt :

      Fukui indices can evaluate the affinity between molecules and oxidants, which has a guiding significance for predicting the degradation path of a compound [37]. In previous studies, we have applied this parameter to the prediction model of VOCs degradation by OH radicals and ozone, and found that the Fukui indices show a high correlation with the degradation rate [28,38]. In addition, bond order, another important parameter describing the strength of chemical bonds, is also an important parameter affecting the degradation of VOCs by OH radicals [38].

    View all citing articles on Scopus
    View full text