Modelling of n

May 11, 2023

Scientific Reports volume 12, Article number: 19662 (2022) Cite this article

656 Accesses

Metrics details

Diesel oil is known to be one of the major petroleum products that can pollute water and soil. Soil pollution caused by petroleum hydrocarbons has substantially impacted the environment, especially in the Middle East. In this study, modeling and optimization of hexadecane removal from soil was performed using two pure cultures of Acinetobacter and Acromobacter and consortium culture of both bacterial species using artificial neural network (ANN) method. Then the best ANN structure was proposed based on mean square error (MSE) as well as correlation coefficient (R) for pure cultures of Acinetobacter and Acromobacter as well as their consortium. The results showed that the correlations between the actual data and the data predicted by ANN (R2) in Acromobacter, Acinetobacter and consortium of both cultures were 0.50, 0.47 and 0.63, respectively. Despite the low correlation between the experimental data and the data predicted by the ANN, the correlation coefficient and the precision of ANN for the consortium was higher. As a result, ANN had desirable precision to predict hexadecan removal by the cobsertium culture of Ochromobater and Acintobacter.

Soil pollution caused by petroleum hydrocarbons is considered a major threat to the environment, especially in Middle Eastern countries. Failure in transmission lines, leakages from storage tanks and, oil tankers’ accidents are some of examples of soil pollution caused by petroleum hydrocarbons in Middle Eastern countries1. Diesel oil is an important crude oil product that has the tendency to pollute soil and water. It is produced during the oil refining process and consists of aromatic compounds, natural and, branched alkanes2. Among medium-chain alkanes, hexadecane (C16H34) has been studied as a model contaminant by many researchers3,4,5. Thus, treatment approaches to reduce the hazardous effects of hexadecane pollution are needed. Hexadecane is removed from soil and water by various methods, such as physical, thermal, chemical and, biological methods. Despite their low solubility in water, they are rapidly degraded by microorganisms6. Biological methods are one of the most common methods in decomposition and removal of these substances7,8. In addition to its simplicity, cost effectiveness and feasibility, it is also environmentally friendly recently, researchers have sought to optimize this process and utilize it 3,9,10,11. However, conventional physical–chemical treatments have high costs and can generate residues that are toxic to the biota12. Applying high efficiency and low cost bioremediation processes represent an extremely important way of recovering contaminated areas among several other clean up techniques. Treatment of soil in slurry bioreactors has gotten to be one of the leading alternatives for the bioremediation of soils contaminated by obstinate poisons beneath controlled natural conditions13. SBs are very often practical to determine the possibility and actual potential of a biological strategy in the final repair of a contaminated soil or site. In reality, beneath slurry conditions, the poison exhaustion rates depend primarily on the corruption action of the microorganisms accessible within the framework14 and the comes about gotten for the most part reflects the real natural depuration potential of the soil13. Modeling is an important tool for designing and operating a wastewater treatment process. To model wastewater treatment processes, various models such as Principal Component Analysis (PCA)15,16, Multiple Linear Regression (MLR)17, Random Forest (RF)18,19 and, Artificial Neural Networks (ANN)20,21 are implemented. Among these methods, ANN is a powerful method for modeling nonlinear systems particularly22.

Numerous researchers have examined the secondary effects of pollutants from simple assessable soil properties. Artificial neural network (ANN) models have been substituted in recent times for multiple linear regression (MLR) in developing prediction simulations for soil pollutants23. The major improvement of the ANN is that the models are trained to comprehend the non-linear and complex relationship between the input and output configurations and they do not put limitations on either the input or the output space24.

A major benefit of ANNs is their ability to detect trends in data that show significant unpredictable non-linearity. As a data-driven approach, The ANN may capture spatial features of the configuration at different scales describing linear and non-linear effects. Due to its simplicity in simulation, forecasting and, modeling, it is considered a promising tool25.

ANNs have been widely used in the last decade in forecast removing pollutants from the environment, thanks to their reliable, robust and salient characteristics of nonlinear relationships between input and output data. Nowadays, many researchers investigate these models and determine their ability to predict the bioremediation process, Bioremediation is a biological process that remediates the environment through processes alike adsorption, redox transformation, and precipitation reactions26. Due to the complex nature of the biodegradation process, it is difficult to modeled and simulated these processes by traditional mathematical models. Artificial neural networks (ANNs) are a promising and powerful modeling technique since the mathematical details of the phenomena related to the process are not required.

In this study 7 independent parameters (include the initial concentration of hexadecane, micronutrients concentration, C: N: P ratio, salt, seed, slurry and, exposure time of bacteria to the contaminant) were investigated on removal of hexadecane , and Since that the effects and relationship between parameters cannot be determined with linear relationships, In the present study, ANN models were developed to predict hexadecane removal efficiency from soil was performed using two pure cultures of Acinetobacter and Acromobacter and consortium culture of both bacterial species. , In the end, kinetic and linear equations have been used to check the factors on the removal efficiency.

In this study, experimental data used for the removal of hexadecane from soil by slurry bioreactors using a combination of bioaugmentation and biostimulation. The experiment designed according to Taguchi's method for optimizing the removal rate of hexadecane in the soil. The ratio depends on the quality characteristics of the product/process to be optimize. Taguchi's method uses the logarithmic functions to measure the signal-to-noise ratios (S/N). This technique is a powerful and high quality tool for the design of systems based on orthogonal array experiments that arrange the optimum setting of process control parameters. Moreover, it used as for the design of optimization step and output to serve as objective functions for optimization27. Besides, the designed processes and products by this method not influenced by external conditions28. Therefore, the noise decreased as the precision improved.

First, the removal efficiency of biological method by in the slurry reactor carried out on 54 samples by Acinetobacter, Acromobacter and consortium mass. Then it compare to control samples that there is any bacteria in slurry. Moreover, the chemical properties of raw and effluent of reactor estimated that included pH, C:N:P, as well as the nutrients, salt and seed concentration. Reactor designed with dimension of 1 litter that clean soil was passed through a 2 mm sieve and washed with hexane, and dried. Then, it was spiked with hexadecane at a concentration of 3000 mg/kg dry soil. Then it well mixed completely and stored at 50 °C. After this, the soil seed by bacteria mass and finally it added to reactor for bioremediation29. The schematic from this reactor was shown in Fig. 1.

The schematic from this reactor.

The pH of all bioreactors were adjusted between 7.4 and 7.8. It should be noted that all the experiments of this study were conducted in triplicate. The residual hexadecane was analyzed by GC-FID (Varian CP-3800, Palo Alto, CA, USA) after 20, 50 and 80 days.

In this study, the rate of hexadecane removal from the soil investigated using two pure cultures of Acinetobacter, Acromobacter, and, a consortium culture of both bacterial species. All of bacteria mass cultured in Mineral salts medium (MSM) and trace elements included NH4Cl (4 g/L), KH2PO4 (2.5 g/L), NaCl (0.5 g/L), MgSO4 (0.3 g/L), FeCl3.6H2O (0.3 g/L), CaCl2 (0.01 g/L) and MnCl2.4H2O (0.01 g/L). MSM supported by micronutrients included MnCl2.4H2O (40 mg/L), MoO3 (80 mg/L), CuSO4 (6 mg/L), ZnSO4 (60 mg/L) and H3BO3 (0.03 mg/L). Furthermore, the effective variables of the experiment were also studied. These variables include the initial concentration of hexadecane, micronutrients concentration, C: N: P ratio, salt (NaCl = 0–1%), seed (active bacteria = 3–10% V nutrient broth with active bacteria to volume of bioreactor), slurry (soil/water ratio) and, exposure time of bacteria to the contaminant.

The total number of experiments for each culture was 30 runs. The range of each parameter is given in Table 1.

ANN is a powerful modeling method for determining the nonlinear relationship between variables30. This study was performed by feed-forward backpropagation and Levenberg–Marquardt using MATLAB 201831,32. In the neural network, there are several inputs, hidden and output layers. The number of neurons in each hidden layer has very important in the response database. Hexadecane was predicted using two Multi-Layered Perceptrons (MLPs). Sigmoid and linear transfer functions were used for hidden and output layers, respectively. Model inputs included reaction time, initial hexadecane concentration, salt concentration, applied micronutrients concentration, reactor seeding rate, slurry percentage and, carbon to nitrogen to phosphorus ratio. In addition, the assigned weight determined the relationship between the layers.

Initially, the network was trained with laboratory data from cell viability, of which 70%, 15% and 15% of the data were used for training, validation and testing, respectively. Then, the weight and bias were determined based on network training to monitor the level of error validation. On the condition of an increase in validation error due to specific repetition, the network training was stopped to prevent overfitting. Therefore, it prevents from over fitting when the validation error increased with a specific epoch.

Due to the importance and accuracy of Mean Square Error (MSE) and correlation coefficient (R), the best ANN structure was selected based on MSE and R that were obtained by the number of neurons in the hidden layer. Therefore, the hexadecane removal percentage was predicted by pure cultures of both Acinetobacter and Acromobacter as well as their consortium.

The number of neurons in the hidden layer is determined based on the overlapping range of both Eqs. 1 and 233.

i = number of inputs, o = the number of outputs and n = number of hidden layer neuron.

According to this study, descriptive statistics such as mean, standard deviation, percentage and coefficient of variation have been used to indicate the concentration and hexadecane removal efficiency using the biostimulation method. Furthermore, analytical statistics, including t-test, was used to compare the hexadecane removal rate by different bacterial species under various conditions. The significance level was considered 0.05. Linear and nonlinear models, having different degrees and, weights were used to determine the effect of different variables on the hexadecane removal and the relationship between them. Finally, among the studied models included Polynominal, cubic interpolant, general model, Gussian, spline interpolant and quadratic regression, the best model was selected based on the correlation coefficient (R2) and the sum of squares errors (RMSE). In this study, MATLAB version 2018 software was used for modeling and Excel version 2013 software was used for graphs.

The hexadecane removal efficiency using slurry reactor by bacterial species including Acromobacter, Acinetobacter and, their consortium is shown in Table 4. The mean hexadecane removal by Acinetobacter, Acromobacter and, consortium of both species was 26.9 ± 9.9, 27.86 ± 11.98 and 27.94 ± 12.22 mg/L, respectively. According to Table 4, the maximum removal by Acinetobacter was 46.7% after 80 days of exposure. In these conditions, the concentration of nutrients, salt, inoculation, C: N: P ratio and slurry percentage to 2.5, 0.5, 10, 4.17 and 5%, respectively. Also, the maximum in vitro Acromobacter removal was 59.2%, which at the initial concentration of 30 mg / L, micronutrient equal to 2.5%, slurry and seed content equal to 20 and 2.5% respectively, and carbon to nitrogen ratio 4.17 without salt is obtained. The maximum removal percentage by bacterial consortium, having a concentration of 50 mg / l, resulted in the following percentages: 2.5% of micronutrients, 1% of salt, 2.5% of seed and 5% of slurry. Additionally, carbon to nitrogen ratio of 4.17 was obtained in a period of 50 days. Because the mineralization of the petroleum compounds was performed by consortium cultures at higher concentrations. So that metabolites of one species are used as substrate for other species34. In this condition, the inhibitory effect of metabolites and high concentrations of initial substrate was reduced.

Another concept studied in this research paper was the ability of the neural networks to predict hexadecane removal using Acinetobacter, Acromobacter and, consortium bacteria of both bacterial species. They were investigated separately using the several algorithms that the best results was associated to levenberg-marque (Table 2).

Based on Table 2, the best results associated to Levenberg–Marquardt that expressed in article.

Then the results from this algorithm with a number of different neurons in the hidden layer were estimated. Based on the overlapping range of Eqs. (1) and (2) and according to the number of input and output variables, 6–16 neurons were placed in the hidden layer. Based on the MSE rate and correlation coefficient, the best model was selected. The plot of the best model for all bacteria culture and its results are shown in Fig. 2 and Table 3.

The plot of optimum structure of ANN. (a) Acintobacter, (b) Acromobacter and (c) consortium.

According to Table 3, the best ANN topology for predicting the removal of hexadecanes for Acromobacter, Acinetobacter and, their consortium was 7: 9: 1, 7: 7: 1 and 7: 8: 1, respectively. In addition, according to Fig. 1, the minimum ANN error for predicting the removal percentage for all three cultivars: Acinetobacter, Acromobacter and consortium was 40.5, 32.9 and 9.4, at epochs of 1, 5 and 4 respectively. Despite that, the increase rate of validation error was such that for all three cultures, the algorithm was stopped in iterations 7, 11 and 10, respectively. Besides, Rall (the total regression coefficient of ANN structure) for the hexadecane removal by Acromobacter, Acinetobacter and, their consortium was 0.75, 0.89 and 0.89, respectively. As the R for all of them were higher than 0.75, the prediction by using this model was good35. In other studies, Rall for PAH removal was higher than 0.9 using ANN24,36. According to these results, the ANN has a high tendency to predict hexadecane removal by using Acromobacter species. Table 3 shows the mean N-hexadecane removal by all three cultures in the actual data and the values predicted by ANN.

According to Table 4, the average hexadecane removal by consortium, Acinetobacter and, Acromobacter was 26.9%, 27.86% and 27.92%, respectively. Moreover, the best removal efficiency for any culture in real data was expressed in Table 5. Because it is expected that in consortium cultures with the production of more surfactants, more ester linkage is formed between hexadecane and biosurfactants, which is effective in reducing toxicity and increasing bioremediation32,37. The average elimination predicted in ANN by the consortium, Acinetobacter and Acromobacter was 26.94, 28.9 and 27.66%, respectively. The standard deviation and coefficient of variation predicted in ANN was lower than the actual data.

According to Table 5, the percentage of hexadecane removal by Acinetobacter ranged between 6% up to 46%, while ANN predicted a removal range of 9–39%. Moreover, the lowest and highest percentages of hexadecane removal by Acinetobacter occurred in the 5th and 26th runs, respectively, whereas they related to runs 4 and 39 by ANN. Although during the actual conditions, the maximum Acinetobacter removal rate was in accordance with the ANN, the range of ANN changes in the consortium was more in line with the actual conditions. In other studies, it has been predicted that microbial populations with 4-9-1 and 3-25-1 structures were used for the removal of toluene and trichloroethylene. Therefore, the results of this study were confirmed by previous studies30,31.

The correlation values predicted by ANN and actual data for all three culture are shown in Fig. 3.

The performance of ANN. (a) Acintobacter, (b) Acromobacter and (c) consortium of Acintobacter and Acromobacter.

According to Fig. 4, the correlation between the actual data and the data predicted by ANN (R2) in Acromobacter, Acintobacter, and Consortium of both crops was 0.50, 0.47 and 0.63, respectively. Although there was no high correlation between the actual data and the ANN forecast, the forecast was higher for the consortium. According to these results, ANN had a higher ability to predict hexadecane removal using Acromobacter and Acinetobacter consortium. Similar results were also indicated by previous studies38,39.

The correlation coefficient of ANN predicted vs actual data. (a) Acintobacter, (b) Acromobacter and (c) consortium of Acintobacter and Acromobacter.

As the concentration of pollutants continuously changes in the ambient environment, the synergistic effect both exposure time and initial concentration of hexadecane on the removal efficiency (actual data) predicted by the regression model (Fig. 5). Also, to compare the conditions predicted by ANN, the concentrations obtained from ANN are shown in Fig. 4.

The effect of exposure time and initial hexane concentration on hexane removal perxent. (a) the presence of nutrient, (b) the presence of microbial salt, (c) the presence of seed and d) the presence of slurry, by excel version 2013 software.

As shown in Fig. 6, the maximum removal rate of N-hexadecane occurred at a low concentration (30 g / kg) with an exposure time of 80 days. According to Table 6, considering the factors affecting the growth of consortium bacteria, the percentage of hexadecane removal in the presence of different concentrations of nutrients has followed the first-order equation. This is because the compound has a linear structure, and its mineralization rate depends on bioavailability29.

The effect of exposure time and initial hexane concentration on hexane removal perxent by ANN modeling. (a) the presence of nutrient, (b) the presence of microbial salt, (c) the presence of seed and d) the presence of slurry, by excel version 2013 software.

While in the presence of different concentrations of salt, seed, and slurry; it fitted with the square of hexadecane concentration and the cube of exposure time. While the values predicted by ANN show that the removal percentage in the presence of different concentrations of nutrients and salt, it is better predicted by the first-order polynomial regression. However, in the presence of different concentrations of seed and slurry, the best fitting was provided with quadratic regression, as seen in Table 6.

In actual data, changes in micronutrient concentrations have been linearly associated with decreasing hexadecane concentrations and increasing exposure time. While for salt, seed and slurry only the removal percentage is optimized during a long reaction time and extremely low concentrations. In terms of cost–benefit assessments, these conditions are not appropriate.

Thus, it is better to achieve the desired conditions in studies or microbial decontamination by changing the concentration of micronutrients at lower costs. Meanwhile, the artificial neural network predicts similar nutrient conditions for salt. It seems that further studies on the salt parameter are necessary to determine the validity of the neural network.

In the modeling conducted by ANN, the best model was obtained by the simultaneous effect of slurry and salt with the polynomial regression model (SSE: 338.7, R-square: 0.8757, Adjusted R-square: 0.8665, RMSE: 3.542). This was due to the fact that both parameters in the actual deletion of hexadecane had less correlation with the polynomial regression model (SSE: 2363, R-square: 0.4551, Adjusted R-square: 0.3415, RMSE: 9.923) Figure 4.

On the other hand, the best suited model in terms of actual removal was related to the simultaneous effect of slurry and time by using the response surface (SSE: 1915, R-square: 0.5584, Adjusted R-square: 0.4664, RMSE: 8.933) (Table 6). However, according to Gosai et al. the neural network performed better than the linear model in predicting PAH removal40. This was due to hexadecane having a linear structure, thereby they are better predicted by linear models.

In this study, the removal of N-hexadecane from soil was modeled by Acromobacter, Acinetobacter and their consortium in the slurry reactor by the neural network. Subsequently, the effect of functional parameters on biological removal was investigated using linear regression models. The modeling results showed that the artificial neural network had a higher ability in predicting the biological removal of N-hexadecane from the soil using the microbial population of Acromobacter. So that a higher correlation coefficient was obtained that there was no overfitting until the fifth epoch. In addition, the greater number of neurons in the lattice layer indicated that there was more complex non-linear relationships for prediction of hexadecane biodegradation using Acromobater than Acintobacter. On the other hand, in both, real data and neural network prediction, N-hexadecane removal was associated with first-degree nutritional concentration while it was associated with second-degree seed and slurry regression. However, different results have been obtained regarding time and salt concentration, suggesting that more studies should be done in this field. Furthermore, only the removal percentage is optimised during prolonged shelf life and low concentrations for salt, seed, and slurry, which is not suitable for cost–benefit assessments.

All data generated or analysed during this study are included in this published article [and its supplementary information files].

Shokouhian, M. & Moradi, M. J. Bioremediation of diesel fuel contaminated soil using cow manure. J. Ferdowsi Civil Eng. 24(2), 14 (2014).

Google Scholar

Gallego, J. L. et al. Bioremediation of diesel-contaminated soils: evaluation of potential in situ techniques by study of bacterial degradation. Biodegradation 12(5), 325–335 (2001).

Article CAS PubMed Google Scholar

Bouchez Naïtali, M. et al. Diversity of bacterial strains degrading hexadecane in relation to the mode of substrate uptake. J. Appl. Microbiol. 86(3), 421–428 (1999).

Article PubMed Google Scholar

Östberg, T. L. et al. The effects of carbon sources and micronutrients in fermented whey on the biodegradation of n-hexadecane in diesel fuel contaminated soil. Int. Biodeterior. Biodegradation 60(4), 334–341 (2007).

Article Google Scholar

Nozari, M., Samaei, M. R. & Dehghani, M. The effect of co-metabolism on removal of hexadecane by microbial consortium from soil in a slurry sequencing batch reactor. J. Health Sci. Surveill. Syst. 2(3), 113–124 (2014).

Google Scholar

Volke-Sepulveda, T., Gutiérrez-Rojas, M. & Favela-Torres, E. Biodegradation of hexadecane in liquid and solid-state fermentations by Aspergillus niger. Biores. Technol. 87(1), 81–86 (2003).

Article CAS Google Scholar

Loehr, R. C. & Webster, M. T. Decreased release of PAHs from soils as a result of field bioremediation. Pract. Period. Hazardous Toxic Radioactive Waste Manag. 4(3), 118–125 (2000).

Article CAS Google Scholar

Kiamarsi, Z., et al., Evaluating the bio-removal of crude oil by vetiver grass (Vetiveria zizanioides L.) in interaction with bacterial consortium exposed to contaminated artificial soils. Int. J. Phytoremed. p. 1–10 (2021).

Zghair, F. S. & Jebar, M. S. Bioremoval of lead by some fungi isolated from soil. Plant Arch. 20, 6806–6808 (2020).

Google Scholar

Setti, L. et al. Further research into the aerobic degradation of n-alkanes in a heavy oil by a pure culture of a Pseudomonas sp. Chemosphere 26(6), 1151–1157 (1993).

Article ADS CAS Google Scholar

Younus, R. M., Aziz, E. M. T. & Mohammed, D. A. Degradation of hydrocarbon substances by some bacterial species isolated from contaminated soils with motor oil. EurAsian J. BioSci. 14(1), 1087–1095 (2020).

CAS Google Scholar

Bidoia, E., Montagnolli, R. & Lopes, P. Microbial biodegradation potential of hydrocarbons evaluated by colorimetric technique: a case study. Appl. Microbiol. Biotechnol. 7, 1277–1288 (2010).

Google Scholar

Robles-González, I. V., Fava, F. & Poggi-Varaldo, H. M. A review on slurry bioreactors for bioremediation of soils and sediments. Microb. Cell Fact. 7(1), 1–16 (2008).

Article Google Scholar

Fava, F., Di Gioia, D. & Marchetti, L. Role of the reactor configuration in the biological detoxification of a dump site-polychlorobiphenyl-contaminated soil in lab-scale slurry phase conditions. Appl. Microbiol. Biotechnol. 53(2), 243–248 (2000).

Article CAS PubMed Google Scholar

Wang, B. et al. A probabilistic principal component analysis-based approach in process monitoring and fault diagnosis with application in wastewater treatment plant. Appl. Soft Comput. 82, 105527 (2019).

Article Google Scholar

Cui, F. et al. Application of principal component analysis (PCA) to the assessment of parameter correlations in the partial-nitrification process using aerobic granular sludge. J. Environ. Manage. 288, 112408 (2021).

Article CAS PubMed Google Scholar

Beigzadeh, B. et al. A new approach in adsorption modeling using random forest regression, Bayesian multiple linear regression, and multiple linear regression: 2, 4-D adsorption by a green adsorbent. Water Sci. Technol. 82(8), 1586–1602 (2020).

Article CAS PubMed Google Scholar

Zhou, P. et al. A random forest model for inflow prediction at wastewater treatment plants. Stoch. Env. Res. Risk Assess. 33(10), 1781–1792 (2019).

Article Google Scholar

Mateo Pérez, V. et al. A random forest model for the prediction of FOG content in inlet wastewater from urban WWTPs. Water 13(9), 1237 (2021).

Article Google Scholar

Sakiewicz, P. et al. Innovative artificial neural network approach for integrated biogas–wastewater treatment system modelling: Effect of plant operating parameters on process intensification. Renew. Sustain. Energy Rev. 124, 109784 (2020).

Article CAS Google Scholar

Bekkari, N., & Zeddouri, A. Using artificial neural network for predicting and controlling the effluent chemical oxygen demand in wastewater treatment plant. Manag. Environ. Qual. Int. J. (2018).

Samadi-Maybodi, A. & Nikou, M. Removal of sarafloxacin from aqueous solution by a magnetized metal-organic framework; artificial neural network modeling. Polyhedron 179, 114342 (2020).

Article CAS Google Scholar

Annadurai, G. & Lee, J.-F. Application of artificial neural network model for the development of optimized complex medium for phenol degradation using Pseudomonas pictorum (NICM 2074). Biodegradation 18(3), 383–392 (2007).

Article CAS PubMed Google Scholar

Olawoyin, R. Application of backpropagation artificial neural network prediction model for the PAH bioremediation of polluted soil. Chemosphere 161, 145–150 (2016).

Article ADS CAS PubMed Google Scholar

Kabuba, J., The use of neural network for modeling of waste activated sludge for solubilization and biodegradability enhancement using ozonolysis process. Biochem. Eng. J. 84. (2014)

Ram Talib, N.S., et al., Artificial neural networks (ANNs) and response surface methodology (RSM) approach for modelling the optimization of chromium (VI) reduction by newly isolated Acinetobacter radioresistens strain NS-MIE from agricultural soil. BioMed Res. Int. (2019).

Sahoo, P. Tribological performance of electroless Ni-P coatings. In Materials and Surface Engineering 163–205 (Elsevier, 2012).

Chapter Google Scholar

Mohamed, S. et al. Optimization method of injection molding parameters for vinyl-based polymer composites. In Natural Fibre Reinforced Vinyl Ester and Vinyl Polymer Composites 97–108 (Elsevier, 2018).

Chapter Google Scholar

Samaei, M. R. et al. Isolation and kinetic modeling of new culture from compost with high capability of degrading n-hexadecane, focused on Ochrobactrum oryzae and Paenibacillus lautus. Soil Sedim. Contamin. Int. J. 29(4), 384–396 (2020).

Article CAS Google Scholar

Azizi, E. et al. 4-chlorophenol removal by air lift packed bed bioreactor and its modeling by kinetics and numerical model (artificial neural network). Sci. Rep. 11(1), 1–10 (2021).

Article Google Scholar

Blagojev, N. et al. Modelling and efficiency evaluation of the continuous biosorption of Cu (II) and Cr (VI) from water by agricultural waste materials. J. Environ. Manage. 281, 111876 (2021).

Article CAS PubMed Google Scholar

Shahzadi, I. et al. Arginine-based cationic surfactants: Biodegradable auxiliary agents for the formation of hydrophobic ion pairs with hydrophilic macromolecular drugs. J. Colloid Interface Sci. 552, 287–294 (2019).

Article ADS CAS PubMed Google Scholar

Abbasi, F. et al. The toxicity of SiO2 NPs on cell proliferation and cellular uptake of human lung fibroblastic cell line during the variation of calcination temperature and its modeling by artificial neural network. J. Environ. Health Sci. Eng. 19(1), 985–995 (2021).

Article CAS PubMed PubMed Central Google Scholar

Abbasi, F., A. Azhdarpoor, & M.R. Samaei, Kinetic modeling of Methyl Tert-Butyl Ether (MTBE) removal by bacterial isolated from contaminated soils. Int. J. Environ. Anal. Chem. 1–16 (2020).

Ostertagová, E. Modelling using polynomial regression. Procedia Eng. 48, 500–506 (2012).

Article Google Scholar

Bao, H. et al. Effects of corn straw on dissipation of polycyclic aromatic hydrocarbons and potential application of backpropagation artificial neural network prediction model for PAHs bioremediation. Ecotoxicol. Environ. Saf. 186, 109745 (2019).

Article CAS PubMed Google Scholar

Barin, R. et al. Fast bioremediation of petroleum-contaminated soils by a consortium of biosurfactant/bioemulsifier producing bacteria. Int. J. Environ. Sci. Technol. 11(6), 1701–1710 (2014).

Article CAS Google Scholar

Kumar, K. V. & Porkodi, K. Modelling the solid–liquid adsorption processes using artificial neural networks trained by pseudo second order kinetics. Chem. Eng. J. 148(1), 20–25 (2009).

Article CAS Google Scholar

Khataee, A. et al. Neural network modeling of biotreatment of triphenylmethane dye solution by a green macroalgae. Chem. Eng. Res. Des. 89(2), 172–178 (2011).

Article CAS Google Scholar

Gosai, H. B. et al. Bioengineering for multiple PAHs degradation using process centric and data centric approaches. Chemom. Intell. Lab. Syst. 179, 99–108 (2018).

Article CAS Google Scholar

Download references

The authors would like to thank the Research Vice-chancellor of Shiraz University of Medical Sciences for supporting the research (Proposal No).

Department of Environmental Health Engineering, School of Health, Shiraz University of Medical Sciences, Shiraz, Iran

Roya Morovati, Fariba Abbasi & Mohammad Reza Samaei

Research Center for Health Sciences, Department of Environmental Health Engineering, School of Health, Shiraz University of Medical Sciences, Shiraz, Iran

Mohammad Reza Samaei

Bachelor of Engineering (Honours) in Environmental Engineering, School of Civil and Environmental Engineering, University of New South Wales, Sydney, NSW, Australia

Hamid Mehrazmay

Bachelor of Engineering (Honours) in Civil Engineering, School of Civil and Environmental Engineering, University of Technology Sydney, Sydney, NSW, Australia

Ali Rasti Lari

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

You can also search for this author in PubMed Google Scholar

R.M. and M.R.S. write the articleF.A. Data AnalysisH.M. and A.L. edited the article.

Correspondence to Mohammad Reza Samaei.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

Morovati, R., Abbasi, F., Samaei, M.R. et al. Modelling of n-Hexadecane bioremediation from soil by slurry bioreactors using artificial neural network method. Sci Rep 12, 19662 (2022). https://doi.org/10.1038/s41598-022-21996-6

Download citation

Received: 29 May 2022

Accepted: 07 October 2022

Published: 16 November 2022

DOI: https://doi.org/10.1038/s41598-022-21996-6

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.