Search

This study used artificial neural network modeling to predict carbon dioxide adsorption capacity on two zeolite adsorbents, 13X (MS 544HP) and 5A (MS 522). Temperature and pressure were used as the system inputs and adsorption capacity as the output. To determine the network training algorithm, the optimal transfer functions in the hidden and output layers and the optimal number of neurons, the coefficient of the determination, and the root mean square error were calculated. To find the best network training algorithm, eight different algorithms, including TRAINBFG, TRAINRP, TRAINCGP, TRAINCGF, TRAINR, TRAINCBG, TRAINBR, and TRAINLM were compared. After modeling the experimental data, the Levenberg–Marquardt backpropagation (BP) algorithm was used to train the network for both zeolites. The optimal number of neurons for both 13X and 5A zeolites was obtained as 10. Finally, the results of artificial neural network modeling and the Toth model, obtained by Wang, were compared. Coefficients of determination for artificial neural network and Toth have been obtained for 13X adsorbent, as 0.9974, and 0.9918 and they were determined as 0.9941 and 0.9923 for 5A adsorbent, respectively. These R2 values show the high accuracy of the artificial neural network compared with the Toth model.

The pure titania (TiO2) and the heterogeneous ternary magnetic nanocomposite of copper ferrite/ferrite oxide (CuFe2O4/Fe2O3) deposited by titanium dioxide (TiO2) were fabricated using a facile one-pot hydrothermal synthesis for the photocatalytic decomposition of methylene blue (MB) dye, under visible light. The nanocomposite was encoded as TCF in this work, where T stands for TiO2, C for CuFe2O4 and F for Fe2O3. Various techniques such as powder X-ray diffraction (PXRD), field emission scanning electron microscopy (FESEM), energy-dispersive X-ray spectroscopy, diffuse reflectance spectroscopy, nitrogen physisorption, and vibrational sample magnetometry (VSM) were used to characterize the prepared samples. The PXRD data showed that the samples had pure anatase structure and the average crystal size of anatase TiO2 in the pure titania and ternary nanocomposite were calculated 147 Å and 135 Å, respectively. The nitrogen physisorption analysis data showed that the pore diameter was increased from 10.6 nm in pure titania to 16.0 nm in TCF. The pore volume was also increased from 0.316 in titania to 0.383 cm3/g in TCF sample. It also increased the typical magnitude of the mesopores’ diameter and volume per weight but it reduced the specific surface area of the samples. The VSM analysis of the ternary nanocomposite showed a considerable magnetic property of the sample (1.99 emu/g), qualifying it as a paramagnetic material. The photocatalytic decomposition efficiency of MB reached 77% and 68% in the presence of pure titania and TCF ternary nanocomposite, after 240-min exposure by the visible light. Active species trapping experiments showed that the major active species responsible for the photodecomposition of MB in the presence of TCF are $${\text{O}}_{2}^{ \cdot - }$$O2·- radicals and holes (h+).