Adsorption modeling and optimization of thorium (IV) ion from aqueous solution using chitosan/TiO2 nanocomposite: Application of artificial neural network and genetic algorithm

https://doi.org/10.1016/j.enmm.2020.100400Get rights and content

Highlights

  • A novel Ch/Ti-ONC nanocomposite adsorbent was prepared and modified.

  • Hydrothermal and chemical methods were combined to synthesize Ch/Ti-ONC.

  • The adsorption capacity of Ch/Ti-ONC was 455 mg  g−1 composite, leading to 99 % Th4+ removal.

  • All experiments were designed and carried out by ANN, and GA was utilized to determine the most practical condition in the adsorption of Th4+.

  • Comparing to the literature, Ch/Ti-ONC has a high capacity for Th4+ adsorption.

Abstract

In this study, novel chitosan/TiO2 nanocomposite (Ch/Ti-ONC) synthesized and evaluated as an adsorbent for removing thorium (IV) (Th4+) ion from aqueous solution. The Ch/Ti-ONC was characterized by X-ray diffraction (XRD), Brunauer–Emmett–Teller (BET), Fourier Transform Infra-Red (FT-IR), and Scanning Electron Microscopy (SEM) and the specific surface area of Ch/Ti-ONC was found to be 28.11 m2 g−1. Artificial Neural Network (ANN) was employed to model the time diversity of the Th4+ adsorption. Three operating condition parameters were considered, namely, adsorbent weight, contact time, and pH. Various ANN structures were built up, trained, validated, and tested using 144 experiments. Optimal ANN structure was specified matching three criteria: the mean absolute percentage error, normalized root mean squared error, and R. The prediction process inside the optimal ANN was derived in the form of simple and user-friendly mathematical formulas. Genetic algorithm (GA) was applied to determine the most practical Ch/Ti-ONC in the adsorption of Th4+. At a fixed exposure time, the optimization procedure was carried out utilizing a GA with the mathematical formulas' objective function from the optimal ANN model. The adsorption kinetics was well described by the pseudo-second-order equation when the Langmuir model better fit the adsorption isotherms. Ch/Ti-ONC's adsorption capacity was 455 mg g−1 composite, which leads to 99 % removal at 45 °C. Besides, the calculated thermodynamic parameters (i.e., standard enthalpy, entropy, and Gibbs free energy) indicates the spontaneous and endothermic nature of the adsorption process. Besides, the ultimate adsorption was received when adsorbent weight, contact time, and pH equaled to 200 mg L−1, 64 min, 6.5, respectively. The loaded Th4+ can be efficiently restored with HNO3, and the Ch/Ti-ONC could be utilized several times in the absence of a dramatic decline in its adsorption capability.

Introduction

Increment of industrial waste production due to industrial advancement has resulted in the entry of toxic radioactive ions to the environment (Balasubramaniana, 2014). These class of ions, are very toxic, even in deficient concentrations, and are remarkably detrimental to the environment (Huang et al., 2019). Th4+ is a valuable strategic element found in wastewaters of the production of nuclear fuels. Moreover, soil, rocks, sand, and water are its principal reservoirs (Humelnicu et al., 2011). This ion exists in the environment in a single isotopic form (232Th) and decays very deliberately. Conversion of Th4+ into uranium-233 (233U) leads to its usage as nuclear fuel (Jain et al., 2006; Nilchi et al., 2013). As it is available several times more than uranium, it counts as a valuable nuclear fuel (Rao et al., 2006; Vijayan et al., 2016). In the process of utilizing thorium as the primary fuel, less crude materials are consumed, resulting in less waste (Rao et al., 2006). Besides, other thorium compounds have a wide range of applications. These include a representative element for tetravalent actinides in their separation studies, using high-quality lenses, and the generation of ceramics at extraordinary temperatures (Höllriegl et al., 2007). In accordance, removing thorium ion from wastewaters is vital due to the extensive use of this ion in industries and its long-term permanency in the environment, besides its potential to threaten human health and the surroundings (Aydin and Soylak, 2007).

Heretofore, some methods have been employed for the removal of thorium from wastes or aqueous solutions, including solvent extraction, chemical precipitation, ion exchange, evaporation, bio adsorption, and adsorption (Brattain and Becker, 1933; Kraus et al., 1956; Sill and Willis, 1964; Tsezos and Volesky, 1981; Jyothi and Rao, 1990; Zhang et al., 2010; Shiri-Yekta et al., 2013). However, many of these methods are not efficient, economical, healthy, and produce hazardous waste (Pollard et al., 1992). Adsorption stands as one of the most frequent procedures in segregating heavy metals from aqueous solutions, owing to its specifications, including great capableness, low priced, and low sludge manufacture (Rahmati et al., 2012).

Chitosan (Ch) is one of the abundant polymeric adsorbents achieved from chitin’s deacetylation (Hien et al., 2005; Huang et al., 2017). Since there are active functional groups such as hydroxyl and amino in chitosan structure, it is the best adsorbent for metal ions compared with other biopolymers (Yi et al., 2005; Anirudhan et al., 2010). Nonetheless, Ch has some weaknesses that limit its application as an adsorbent, including softness, a tendency to agglomerate or form gels, poor porosity, inadequate surface area, hydrodynamic constraints in the adsorption column, and absence of reactive binding sites (Saifuddin and Kumaran, 2005; Mohan and Pittman, 2006). To overcome these weaknesses and modify its features, Ch was combined with other adsorbents in the conducted researches (Jordan et al., 2005; Gerente et al., 2007; Ngah et al., 2011). Hitherto, few papers have been published in the usage of Ch-based nanocomposites (Akkaya and Ulusoy, 2008; Anirudhan et al., 2010; Humelnicu et al., 2011; Muzzarelli, 2011). Ge et al. (2016) investigated the capacity of grafted and cross-linked chitosan nanoparticles prepared by using Pb2+ for lead adsorption (Ge et al., 2016). In another study, ethylenediamine-modified magnetic Ch particles were used as the adsorbent of radioactive uranyl ions (Wang et al., 2011). Hritcu et al. (2012), used unmodified magnetic Ch particles for the adsorption of Th4+ and U6+ ions (Hritcu et al., 2012).

Numerous parameters can affect the adsorption capacity, including pH, initial ion concentration, contact time with the adsorbent, sorbent concentration etc (Hosseini et al., 2017). In accordance, the determination of optimal parameters to achieve maximum efficiency is imperative (Khataee et al., 2013). Response Surface Methodology (RSM), factorial design, and Taguchi are some of the optimization methods that can establish a mathematical relationship between independent and response parameters with the lowest number of experiments (Montgomery, 2009). However, the application of these methods has some weaknesses, including inaccurate estimation of the approximations, validity only for limited ranges and regions, and erroneous optimization of the multiple non-linear objectives (Hassani et al., 2015; Anandharamakrishnan, 2017; Hassani et al., 2018). ANN is a modeling approach for approximating the experimental regions even in non-linear systems (Podstawczyk et al., 2015). In addition, the literature confirms that the combination of the ANN model with Genetic Algorithm (GA) is an effective method for modeling, optimization, and problem-solving (Mandal et al., 2015; Mohan, Singh, et al. 2015; Azadi et al., 2017). Khandanlou et al. (2016) applied ANN to increase the adsorption of heavy metals using nanocomposites of rice straw and Fe3O4 nanoparticles as the adsorbent (Khandanlou et al., 2016).

This study presents the application, adsorption properties, accurate modeling, and optimization of novel Ch/Ti-ONC as an adsorbent to remove Th4+ from aqueous solution. The composition of synthesized Ch/Ti-ONC was examined, and its adsorption capability for the removal of Th4+ under varied experimental conditions was investigated. Time diversity of Th4+ concentration throughout the adsorption process was modeled matching the ANN model, regarding three affective parameters (adsorbent weight, contact time, and initial pH). Several ANN structures were outstretched, validated, and tested using the data of 144 experiments, and its optimal condition was specified using the amalgamation of ANN-GA. From the compatibility of adsorption isotherms to the Langmuir and Freundlich models, adsorption parameters were obtained by which the adsorptive properties were evaluated. In addition, kinetics data have been modeled by the pseudo-first and second-order kinetics models. All the experiments and analyses were carried out in of Iran’s nuclear science and technology research institute laboratories.

Section snippets

Material and method

The same procedures as mentioned in our previous papers (Broujeni and Nilchi, 2018; Broujeni et al., 2018a, b), all the materials were acquired from Merck and utilized without additional purification. Further explanation is given in the supplementary file, Section 1 and 2. All the data obtained were the average of two replications with an average comparative error of less than 5%.

Synthesis of TiO2 nanoparticle (TiO2-NPs)

TiO2-NPs was provided by a hydrothermal synthesis method. 18.75 g titanium tetrachloride (TiCl4) and 6 g CH4N2O were

Characterization of Ch/Ti-ONC

Fig. 2 depicts the XRD schema of Ch, TiO2-NPs, and Ch/Ti-ONC. The XRD pattern of Ch signifies broad diffraction peaks at 2θ = 9.5° and 19.5°, which represent the mutual fingerprints of crystal Ch (Ogawa et al., 1984). Sharp diffraction peaks at 2θ of 25.41°, and 48.01° confirms the production of TiO2 in the anatase phase (Bian et al., 2019). The sharpness of the peaks and their broad diffraction shown in the sample’s XRD pattern imply that crystalline nanoparticles with extremely small size

Conclusion

In this research, a novel Ch/Ti-ONC adsorbent was prepared and characterized by a detailed examination of its capability for the removal of Th4+ ion from aqueous solution. One hundred forty experiments with three variables (adsorbent weight, contact time, and pH) were carried out in a distributed domain and different ANN trained structure. The ANN's optimal structure incorporates a tansig activation function, LM training algorithm, and 4 neurons intrant the hidden layer. The statistical

Authorship contributions

TermDefinition
MethodologyCombining artificial neural network via genetic algorithm Software MATLAB R2018b
ValidationAll data validate by Iran atomic energy organization laboratory
InvestigationAll experiments have been validated in Iran atomic energy organization laboratory
SupervisionThe research has neem supervising by Prof. A. Nilchi
Funding acquisitionThe Nuclear Science and Technology Research Institute of Iran funded this research, with project ID: PRI-C5-93-001.

Declaration of Competing Interest

I enclose a copy of the type written manuscript, titled, “Adsorption modeling and optimization of thorium (IV) ion from aqueous solution using chitosan/TiO2 nano composite: application of artificial neural network and genetic algorithm” as requested for publication in the “Environmental Nanotechnology, Monitoring and Management”. Furthermore, all authors are aware of the manuscript content and approve its submission. The work presented is original, is not being considered by any other journal

Acknowledgment

This research was funded by the Nuclear Science and Technology Research Institute of Iran, with project ID: PRI-C5-93-001. All authors would like to acknowledge the Institute for granting support to analyze samples carried out throughout this research.

References (73)

  • A. Hassani et al.

    Photocatalytic degradation of ciprofloxacin by synthesized TiO2 nanoparticles on montmorillonite: effect of operation parameters and artificial neural network modeling

    J. Mol. Catal. A Chem.

    (2015)
  • A. Hassani et al.

    Photocatalytic ozonation of ciprofloxacin from aqueous solution using TiO2/MMT nanocomposite: nonlinear modeling and optimization of the process via artificial neural network integrated genetic algorithm

    Process Saf. Environ.

    (2018)
  • N.Q. Hien et al.

    Radiation grafting of acrylic acid onto partially deacetylated chitin for metal ion adsorbent

    Nucl. Instrum. Methods Phys. Res. B

    (2005)
  • V. Höllriegl et al.

    Observation of changes in urinary excretion of thorium in humans following ingestion of a therapeutic soil

    J. Environ. Radioact.

    (2007)
  • D. Hritcu et al.

    Magnetic chitosan composite particles: evaluation of thorium and uranyl ion adsorption from aqueous solutions

    Carbohydr.

    (2012)
  • Z. Huang et al.

    Interaction mechanism of uranium (VI) with three-dimensional graphene oxide-chitosan composite: insights from batch experiments, IR, XPS, and EXAFS spectroscopy

    Chem. Eng. J.

    (2017)
  • Z.W. Huang et al.

    Adsorption of Eu (III) and Th (IV) on three-dimensional graphene-based macrostructure studied by spectroscopic investigation

    Environ. Pollut.

    (2019)
  • D. Humelnicu et al.

    Adsorption characteristics of UO22+ and Th4+ ions from simulated radioactive solutions onto chitosan/clinoptilolite sorbents

    J. Hazard. Mater.

    (2011)
  • P. Ilaiyaraja et al.

    Adsorption of uranium from aqueous solution by PAMAM dendron functionalized styrene divinylbenzene

    J. Hazard. Mater.

    (2013)
  • V.K. Jain et al.

    Simultaneous preconcentration of uranium (VI) and thorium (IV) from aqueous solutions using a chelating calix [4] arene anchored chloromethylated polystyrene solid phase

    Talanta

    (2006)
  • J. Jordan et al.

    Experimental trends in polymer nanocomposites—a review

    Mater. Sci. Eng. A

    (2005)
  • A. Jyothi et al.

    Solvent extraction behaviour of lanthanum (III), cerium (III), europium (III), thorium (IV) and uranium (VI) with 3-phenyl-4-benzoyl-5-isoxazolone

    Talanta

    (1990)
  • R. Khandanlou et al.

    Enhancement of heavy metals sorption via nanocomposites of rice straw and Fe3O4 nanoparticles using artificial neural network (ANN)

    Ecol. Eng.

    (2016)
  • D. Langmuir et al.

    The mobility of thorium in natural waters at low temperatures

    Geochim. Cosmochim. Acta

    (1980)
  • S. Li et al.

    Efficient thorium (IV) removal by two-dimensional Ti2CTx MXene from aqueous solution

    Chem. Eng. J.

    (2019)
  • N.M. Mahmoodi et al.

    Novel magnetic amine functionalized carbon nanotube/metal-organic framework nanocomposites: from green ultrasound-assisted synthesis to detailed selective pollutant removal modelling from binary systems

    J. Hazard. Mater.

    (2019)
  • S. Mandal et al.

    Enhanced removal of Cr (VI) by cerium oxide polyaniline composite: optimization and modeling approach using response surface methodology and artificial neural networks

    J. Environ. Chem. Eng.

    (2015)
  • D. Mohan et al.

    Activated carbons and low cost adsorbents for remediation of tri-and hexavalent chromium from water

    J. Hazard. Mater.

    (2006)
  • S. Mohan et al.

    Synthesis of CuO nanoparticles through green route using Citrus limon juice and its application as nanosorbent for Cr (VI) remediation: process optimization with RSM and ANN-GA based model

    Process Saf. Environ. Prot.

    (2015)
  • R.A. Muzzarelli

    Potential of chitin/chitosan-bearing materials for uranium recovery: an interdisciplinary review

    Carbohydr. Polym.

    (2011)
  • P. Naderi et al.

    Efficient removal of crystal violet from aqueous solutions with Centaurea stem as a novel biodegradable bioadsorbent using response surface methodology and simulated annealing: kinetic, isotherm and thermodynamic studies

    Ecotoxicol. Environ. Saf.

    (2018)
  • A. Nilchi et al.

    Kinetics, isotherm and thermodynamics for uranium and thorium ions adsorption from aqueous solutions by crystalline tin oxide nanoparticles

    Desalination

    (2013)
  • K. Parida et al.

    Adsorption of toxic metal ion Cr (VI) from aqueous state by TiO2-MCM-41: equilibrium and kinetic studies

    J. Hazard. Mater.

    (2012)
  • D. Podstawczyk et al.

    Biosorption of copper (II) ions by flax meal: empirical modeling and process optimization by response surface methodology (RSM) and artificial neural network (ANN) simulation

    Ecol. Eng.

    (2015)
  • S. Pollard et al.

    Low-cost adsorbents for waste and wastewater treatment: a review

    Sci. Total Environ.

    (1992)
  • A. Rahmati et al.

    Kinetic and thermodynamic studies of uranium (VI) adsorption using Amberlite IRA-910 resin

    Ann. Nucl. Energy

    (2012)
  • Cited by (19)

    • Development of green photocatalytic geopolymers for dye removal

      2022, Materials Chemistry and Physics
      Citation Excerpt :

      Therefore, traditional statistical methods and models are not suitable for simulating the adsorption process. The artificial neural network (ANN) can establish accurate replantation among variables, which is why it has the potential to forecast the adsorption efficiency [18]. In this project, MATLAB 2020 ANN Toolbox was used to model the adsorption efficiency and capacity.

    View all citing articles on Scopus
    View full text