Swelling prediction of calcium alginate/cellulose nanocrystal hydrogels using response surface methodology and artificial neural network

• Published in: Industrial crops and products Vol. 192; p. 116094
• Main Authors: Soleimani, Soraya, Heydari, Amir, Fattahi, Moslem
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.02.2023
• Subjects: Artificial neural network; Calcium alginate; Cellulose nanocrystals; Hydrogels; Response surface methodology; Swelling; Response surface methodology; Artificial neural network; Swelling; Hydrogels; Calcium alginate; Cellulose nanocrystals
• ISSN: 0926-6690; 1872-633X
• DOI: 10.1016/j.indcrop.2022.116094

The calcium alginate (CA) hydrogels of the raw and reinforced with cellulose nanocrystals (CNC) were synthesized using the ion-exchange method. FT-IR, FESEM, EDX, XRD, and TGA were used to characterize the samples. Swelling as a crucial property of hydrogels was investigated. CNCs-reinforced hydrogels (CA/CNC) with a higher swelling ratio were selected for the subsequent tests. The results of the effect of independent variables, including concentrations of sodium alginate, calcium chloride, and CNC on CA/CNC swelling were analyzed using response surface method based on central composite design (RSM-CCD) and compared with the artificial neural network-salp swarm algorithm (ANN-SSA) hybrid. ANN-SSA codes were developed using the Python programming language. The best optimal topology for the ANN was 3:7:1, and the trained ANN-SSA was converted to matrix form to have a predictive equation based on the weights and biases. The validation of the models was performed using the correlation coefficient (R2), root mean square error (RMSE), and normal standard deviation (Δq (%)). The results of R2, RMSE, and Δq (%) for ANN-SSA are 1.000, 2.657, and 0.095, respectively, and for RSM are 0.984, 202.598, and 7.384, respectively, confirming that ANN-SSA was more precise than RSM. [Display omitted] •Calcium alginate hydrogels synthesized and reinforced with cellulose nanocrystals.•The reinforced hydrogels synthesized by ion-exchange method and characterized.•Swelling of hydrogels were investigated and at different operating conditions.•The swelling behavior of hydrogels modeled by hybrid of RSM and ANN-SSA.•ANN model converted to the matrix form to have a predictive equation.