This is the unspecified version of the paper.This version of the publication may differ from the final published version. Permanent repository link AbstractIn this work, the design and characteristics of a wavelength-dependent pH optical sensor have been studied. To create the sensor itself, brilliant yellow (BY) as a pH indicator and poly (allylamine hydrochloride) [PAH] as a cross-linker have been deposited on the end of a bare silica core of an optical fibre by use of a 'layer-by-layer' technique. In the experiments carried out to characterize the sensor, it was observed that the value of pK a (the dissociation constant) of the thin film is dependent both on the outer layer and the number of bilayers. A heat treatment process was applied to the sensor to reduce the effect on the deposited layers during the testing of the probe. As a result of these series of experiments, it could be concluded that the probe design on which were deposited structured layers comprising six double layers of (PAH/BY) showed the best sensitivity for a pH range from 6.80 to 9.00 (with an accuracy of ±0.20) and showing an average wavelength shift of 4.65 nm per 0.2 pH units, while the concentration of the BY and the PAH solutions was maintained as 0.25mM and 2.5mM respectively.

This is the accepted version of the paper.This version of the publication may differ from the final published version. The results obtained from a series of evaluations show that the sensitivity was enhanced by reducing the concentration of the indicator solution used and by designing a U-bend configuration sensor probe with a sharply bent fibre. However, when making an overall comparison, the straight (unbent) fibre probe resulted in a more sensitive probe when compared to the use of a high radius bend. Further, utilizing a small core diameter of the fibre allows a wide pH range to be measured and with high sensitivity. Additionally, the performance was shown to be improved for measurements over a narrower range of pH, by using a fibre with a larger core diameter. Considering the effect of the number of layers, work carried out has shown that probes with 5-6 bilayers presented the best performance. The sensitivity has been shown to diminish when more than 6 layers were used and the sensing range shifts towards higher pH values. When monitored, the value of pKa (the dissociation constant) of the thin film showed the smallest change of any of the design factors considered. In summary, using a larger core diameter, employing a larger curve radius, a higher number of bilayers, a higher concentration of the indicator solution and applying PAH as an outer layer, all cause a higher pKa value and consequently the probe sensitivity moves towards alkaline region. Permanent repository link

A mechanistic model is developed to simulate ethanol purification using membrane technology. In the considered process, a feed solution containing 10 wt% water + 90 wt% ethanol is contacted with a polymeric dense membrane in a pervaporation system. The membrane selectively separates water from solution in order to purify the ethanol. In the development of the model, it is assumed that the water is the main penetrant through the membrane due to the hydrophilicity of membrane material. The mass fraction of water molecules in the feed solution, as well as membrane, is estimated using Maxwell–Stefan approach. The governing equations are then solved using finite element method in order to predict mass fraction, mass transfer flux, and velocity of the solution in the membrane module. The results indicate that the model can predict the formation of concentration and velocity boundary layer in the feed solution near the membrane/feed interface. Moreover, the developed model is robust and reliable in the understanding of membrane separation processes applicable for dehydration of alcohols.

A comprehensive computational fluid dynamics simulation was developed for the rational design of a bioethanol purification system using a pervaporation process by tailoring the hydrodynamics of the process. The process involves the removal of water from a water/ethanol liquid mixture using a dense polymeric membrane. The model domain was divided into two compartments comprising the feed and the membrane. To describe water transport in the feed solution, the Maxwell-Stefan approach was used, whereas for mass transfer inside the membrane the molecular diffusion mechanism was adopted. The governing equations were solved numerically by using a finite element method. The model was capable of predicting mass transfer along with momentum transfer in the feed and membrane compartments.