UPM Institutional Repository

Artificial neural network assessment for predicting elastic and optical properties of binary tellurite and borate glass systems


Citation

Effendy, Nuraidayani (2021) Artificial neural network assessment for predicting elastic and optical properties of binary tellurite and borate glass systems. Doctoral thesis, Universiti Putra Malaysia.

Abstract

The introduction of artificial neural networks (ANNs) in the glass field has greatly improved this area to further increase fabrication productivity. ANNs are the systems that help the glass expert to estimate a few parameters such as density, ultrasonic velocity, elastic moduli and optical band gap in the glass composition. In this present works, the ANNs system was implemented in a series of zinctellurite (ZnO-TeO2), bismuth-tellurite (Bi2O3-TeO2), zinc-borate (ZnO-B2O3) and bismuth-borate (Bi2O3-B2O3) glass systems which have been successfully fabricated using conventional melting and quenching methods with the configuration of (ZnO)m (TeO2)100-m where m = 0, 5, 10, 15, 20, 25, 30 mol%, (Bi2O3)n (TeO2)100-n where n = 0, 5, 7, 10, 13, 15 mol%, (ZnO)p (B2O3)100-p where p = 0, 40, 45, 50, 55, 60 mol% and (Bi2O3)q (B2O3)100-q where q = 0, 40, 45, 50, 55, 60 mol%, respectively. The experimental measurements have been investigated on the physical, structural, elastic and optical properties of binary tellurite and borate glass systems containing various amounts of ZnO and Bi2O3 concentrations. The experimental density measurement on the effect of ZnO substitution in both tellurite and borate glass systems showed the increment values as the amount of ZnO content increases with the highest density value of 5.283 g/cm3 at 30 mol% of ZnO in tellurite glass systems. Meanwhile, the molar volume value displayed an opposite behavior with the glass density which the lowest value of the molar volume is 21.569 cm3/mol at 60 mol% of ZnO in borate glass systems. The dropping value of the molar volume is attributed to the changes in the glass network connectivity. For the Bi2O3 substitution in both tellurite and borate glasses, the experimental density and molar volume exhibited a similar behavior which is increasing with the increase of Bi2O3 content. The highest density and molar volume values are 6.550 g/cm3 and 46.935 cm3/mol at 60 mol% of Bi2O3 in borate glass systems, respectively. The glassy state and amorphous nature of all glass samples have been confirmed through the presence of a broad hump peak in the XRD analysis. FTIR transmission and Raman absorption spectra have discovered the existence of TeO4, TeO3, BO4 and BO3 structural units in the glass samples. The substitution of modifier ZnO and Bi2O3 into the tellurite and borate glass structure caused the glass structure to become more rigid and increase the elastic moduli values. This modification process affected the formation of bridging oxygen which leads to an increase in cross-link density and gives a better packing in the glass structure as calculated in bond compression and Makishima-Mackenzie theoretical model. The optical behavior revealed that the shifts of the absorption edge to the longer wavelength leading to the reduction in the optical band gap value. The minimum optical band gap value for the effect of ZnO and Bi2O3 substitution in both tellurite and borate glass systems is 2.557 eV at 30 mol% of ZnO in tellurite glass systems and 2.210 eV at 15 mol% of Bi2O3 in tellurite glass systems. Subsequently, the experimental values resulting from the composition of the glass series were compared with the values obtained from the prediction by ANNs. This study has concluded that the ANNs system was relevant to be used in the glass fields since the coefficient of R2 values showed by the prediction against the experimental graph were between 0.9941 to 1.000 which is considered to be very satisfactory.


Download File

[img] Text
FS 2022 35 IR.pdf

Download (1MB)

Additional Metadata

Item Type: Thesis (Doctoral)
Subject: Artificial intelligence
Subject: Elasticity
Subject: Glass - Optical properties
Call Number: FS 2022 35
Chairman Supervisor: Sidek Hj. Ab Aziz, PhD
Divisions: Faculty of Science
Depositing User: Editor
Date Deposited: 07 Aug 2023 07:26
Last Modified: 07 Aug 2023 07:26
URI: http://psasir.upm.edu.my/id/eprint/104200
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item