A new approach to the characterization of nanomaterials: Predicting Young’s modulus by correlation weighting of nanomaterials codes
Graphical abstract
General scheme of Young’s modulus modeling for nanomaterials.
Introduction
Nanomaterials are becoming an important component of modern life and are the subject of many investigations in various areas of natural sciences. However, theoretical modeling of physicochemical and biological activity of these species is very scarce. It is a well-known to predict the properties and/or activities of ‘classical’ substances via correlating with some molecular descriptors. These methods are often cited in the literature as quantitative structure–property/activity relationships (QSPR/QSAR). The prediction of properties/activities by QSPR/QSAR is based on information concerning the molecular structure of the molecules of interest. As a rule the molecular graph is an elucidation of molecular structure in the QSPR/QSAR analysis [1], [2], [3], [4]. As an alternative to the molecular graph in QSPR/QSAR analyses SMILES notation can also be used [5], [6]. In the case of nanomaterials even simple mathematical calculations revealing their architecture (similar to the molecular graph) is scarce. That is the reason that, in spite of an increase in the degree of influence of the nanomaterials in modern physical chemistry, industry, and biomedical disciplines, the concept of using QSAR to predict the properties of nanomaterials has not been yet developed.
The aim of the present study is to estimate the ability of a SMILES-like description of nanomaterials as a basis for predicting Young’s modulus of these materials. The SMILES-like nomenclature for a given nanomaterial contains data on atom composition and the technological conditions of it’s synthesis and is used as basis for calculating optimal descriptors.
However it should be noted that the nomenclature used in the present study is not analogical to the SMILES, since the function of the nomenclature used here nomenclature is restricted to encoding the available information on the genesis of the nanomaterials as commercial products. The SMILES characteristics reflect detailed (2D, 3D, and even quantum chemical) information on molecular architecture.
Data on Young’s modulus applied in this study has been taken from [7]. Nanomaterials in ceramic form are included in this data set. The differences between various substances include variations in atomic composition and in temperature of synthesis.
Section snippets
Method
The information on nano substances includes the following characteristics: (a) atom composition, (b) type of substances (bulk or not), and (c) temperature of synthesis. The data for 29 nanomaterials under consideration is presented in Table 1. Each nano structure contains some components which are also included in other nanostructures. The descriptor used for modeling Young’s modulus (YM) has been defined aswhere Ik is the component information on the nanostructure (e.g., Al, N,
Results
The separation of the considered nanostructures into training and test sets has been done randomly, but according to the following rules: first, all components of the considered species are included in the training set; second, diapasons of Young’s modulus values for the training and test sets are approximately the same. Using these rules 29 nanomaterials under consideration have been divided into a training set of 21 nanomaterials and a test set of eight nanomaterials.
The statistical
Conclusions
Twenty-nine different nanomaterials characterized by by experimental studies were selected to test a new theoretical methodology. The applied approach allows for the prediction of reasonable numerical values of the Young’s modulus of eight randomly selected nanomaterials for a test set from a training set of 21 nanomaterials.
Acknowledgement
The authors thank for support the High Performance Computational Design of Novel Materials (HPCDNM) Project funded by the US Department of Defense through the US Army Engineer Research and Development Center (Vicksburg, MS) Contract #W912HZ-06-C-0057.
References (8)
- et al.
Russ. J. Coord. Chem.
(1998) - et al.
J. Chem. Inf. Comput. Sci.
(1998) - et al.
J. Chem. Inf. Comput. Sci.
(1999) - et al.
J. Chem. Inf. Comput. Sci.
(2001)
Cited by (43)
Application of quasi-SMILES to the model of gold-nanoparticles uptake in A549 cells
2021, Computers in Biology and MedicineCitation Excerpt :One of the challenging aspects of this approach is the selection of appropriate descriptors. Attempts to applying the quantum mechanics descriptors [15–17], the analysis of energetic aspects [18], as well as some other approaches [19,20] are among the considered tactics used to combine QSAR techniques with the uniqueness of studied nanoparticles. So-called quasi-SMILES were suggested as a tool for building up nano-QSAR [21–23].
A dose-response-recovery clustering algorithm for categorizing carbon nanotube variants into toxicologically distinct groups
2019, Computational ToxicologyCitation Excerpt :Quantitative structure activity relationships (QSARs) are one of the modeling schemes utilized; these models hypothesize that the activity of any nanostructure is a function of its physicochemical properties [5]. QSARs have been employed to predict the solubility, Young’s modulus, and partition coefficient of nanomaterials, showcasing their applicability to NMs [6,7]. Puzyn et al. demonstrated the use of QSARs to predict the mechanism of cytotoxicity of 17 different metal oxide nanoparticles [4].
A further development of the QNAR model to predict the cellular uptake of nanoparticles by pancreatic cancer cells
2018, Food and Chemical ToxicologyCORAL and Nano-QFAR: Quantitative feature – Activity relationships (QFAR) for bioavailability of nanoparticles (ZnO, CuO, Co<inf>3</inf>O<inf>4</inf>, and TiO<inf>2</inf>)
2017, Ecotoxicology and Environmental SafetyFrameworks and tools for risk assessment of manufactured nanomaterials
2016, Environment International(Q)SAR modelling of nanomaterial toxicity: A critical review
2015, ParticuologyCitation Excerpt :This computational approach has many advantages in terms of cost, time-effectiveness, and ethical considerations. Although it has been satisfactorily used to predict the physicochemical properties of NMs, such as solubility (Gajewicz et al., 2012; Sivaraman, Srinivasan, Vasudeva Rao, & Natarajan, 2001; Toropov, Leszczynska, & Leszczynski, 2007; Toropov, Toropova, Benfenati, Leszczynska, & Leszczynski, 2009) and elasticity (Mohammadpour, Awang, & Abdullah, 2011; Toropov & Leszczynski, 2006), development of reliable (Q)SAR models becomes more complicated when the actual processes and the endpoints of interest are biologically complex. Despite all the challenges and open questions, there have been some pioneering studies investigating the use of (Q)SAR models to predict the toxicity of ENMs (Epa et al., 2012; Fourches et al., 2010; Liu, Rallo, et al., 2013; Puzyn et al., 2011; Sayes & Ivanov, 2010; Wang et al., 2014; Zhang et al., 2012).