Water recycling has become a global trend because of water scarcity and increased demand of water supply. Therefore, attentions to the improvement of reclaimed water quality have been paid. In the past decade various environmental hormones and PPCPs (pharmaceuticals and personal care products) have been detected in different aquatic environments. Even though their concentrations are in the range of ng/L to £gg/L, these emerging contaminants might cause harm to human health and the environment. Nanoscale contaminants are another type of emerging contaminants cannot be neglected because many nanomaterials have been used in household goods of our daily lives. Thus, how to effectively separate and/or recover those nanomaterials from aqueous solution to reduce their potential hazards is an important issue The first objective of this study was to assess the efficiency of nano-Fe3O4/S2O82- oxidation against selected environmental hormones (i.e., di(2-ethylhexyl)phthalate (DEHP) and perfluorooctane sulphonates (PFOS)) and pharmaceuticals (i.e., erythromycin (ERY) and sulfamethoxazole (SMX)) in aqueous solution. The optimal operating conditions obtained from the above-indicated oxidation process were then transferred to a simultaneous electrocoagulation and electrofiltration (EC/EF) treatment module into which a tubular TiO2/Al2O3 composite membrane was incorporated. The purpose of this practice was to evaluate whether the EC/EF process could further enhance the removal of target contaminants. In this work nanoscale magnetite (nano-Fe3O4) used for activation of S2O82- oxidation was prepared by chemical coprecipitation. Then, X-ray powder diffractometry was used to confirm the crystal structure of the prepared particles as magnetite. The employment of 3 wt% soluble starch was found to be sufficient to stabilize nano-Fe3O4 for later uses. Further, slurries of nano-Fe3O4 and S2O82- (sodium persulfate) were prepared with three dosage ratios, namely 1:2.5, 1:5 and 1:10. Nano-Fe3O4/S2O82- slurries thus prepared were used for evaluating their efficiencies in removing target contaminants (i.e., DEHP, PFOS, ERY, and SMX) of two concentration levels. In this study the high concentration level referred to 38 mg/L for DEHP and 10 mg/L each for the rest of target contaminants, whereas 10 £gg/L as the low concentration level for each of target contaminants. Batch experiments of nano-Fe3O4/S2O82- oxidation against target contaminants were first carried out in glass beakers. In the case of high concentration level with a nano-Fe3O4-to-S2O82- dosage ratio of 1:10, the respective removal efficiencies for all target contaminants were greater than 98%. Using the same dosage ratio for the case of low concentration level, however, the respective removal efficiencies for all target contaminants decreased to 78-91% except for ERY. When all target contaminants of low concentration level co-existed in the reaction vessel, the residual concentrations of environmental hormones were found to be greater than that of pharmaceuticals. Under the circumstances, the removal efficiency of DEHP dropped to 70% or so. The reaction pathways of nano-Fe3O4/S2O82- oxidation against each of target contaminants with a high concentration level were also investigated. The degradation intermediates detected for all target contaminants were all in line with the literature. Besides, the degradation intermediates were all close to their respective end products except those originated from DEHP. In other words, nano-Fe3O4/S2O82- per se had a phenomenal oxidation rate against each target contaminant. The performance of EC/EF-assisted nano-Fe3O4/S2O82- oxidation against target contaminants of low concentration level was also evaluated in this study. In each test every contaminated aqueous solution was physically preconditioned within the EC/EF treatment module for 20 min prior to the application of an electric field to enact electrocoagulation and electrofiltration. The optimal operating conditions obtained were given as follows: aluminum anode, electric field strength of 60 V/cm, transmembrane pressure of 98 kPa, and crossflow velocity of 3.33 cm/s. Under such conditions, the removal efficiencies for DEHP, PFOS, ERY, and SMX were determined to be 95%, 99%, 100%, and 99%, respectively. In the case of mixed environmental hormones and pharmaceuticals, the respective removal efficiencies slightly decreased to 85-99%. It is evident that the coupling of the EC/EF process with nano-Fe3O4/S2O82- oxidation yielded a substantial removal increase for selected target contaminants. Additionally, in all tests of EC/EF-assisted nano-Fe3O4/S2O82- oxidation against target contaminants, no residual nano-Fe3O4 was found in permeate. After a simple adjustment of pH, permeate thus treated would be ready for reuse in cooling towers.

This study addresses three major parts: (1) to establish the technology for the preparation of tubular ceramic membrane substrates; (2) to establish the technology for the preparation of tubular carbon/ceramic membranes; and (3) to reclaim water from chemical mechanical polishing (CMP) wastewaters by a combined treatment system of a novel simultaneous electrocoagulation/electrofiltration (EC/EF) process coupled with laboratory-prepared tubular composite membranes (TCMs) and evaluate its feasibility of water recycling and operating cost. First, in this work the green substrates of tubular porous ceramic membranes consisting of corn starch were prepared using the extrusion method, followed by curing, drying, and sintering processes. Experimental results have demonstrated that an addition of starch granules to the raw materials would increase the porosity, pore size, and permeability of the sintered matrices but accompanied by a decrease of the compressive strength. It revealed that the membrane substrates with desired pore sizes and permeability could be obtained by adding a proper amount of corn starch. The nominal pore sizes of the prepared membrane substrates were ranging from 1 to 2 £gm. The membrane substrates thus obtained are suitable for crossflow microfiltration applications. Second, the carbon/alumina TCMs and carbon fibers/carbon/alumina TCMs were obtained by the chemical vapor deposition (CVD) method resulting in a pore size distribution of 2 to 20 nm and a nominal pore size ranging from 3 to 4 nm. Besides, during the CVD process the reaction temperature was found to be the main factor for influencing the pore size of carbon fibers/carbon/alumina TCMs and the type of carbon fibers. When the reaction temperature was above or equal to 1000 ¢J, the pore size of TCMs increased due to the pyrolysis of thin carbon layers. The ¡§Tip-Growth¡¨ mechanism was found for tubular carbon fibers formation under such conditions. On the other hand, ¡§Base-Growth¡¨ (also known as ¡§Root-Growth¡¨) mechanism was found for curved and irregular carbon fibers formation when reaction temperature was under or equal to 950 ¢J. Third, for reclaiming water from CMP wastewaters, experimental results of laboratory-prepared carbon/alumina TCMs incorporated into the custom-made EC/EF treatment module used was found to be capable of treating oxide-CMP wastewater in a proper manner. Permeate thus obtained had a turbidity of below 0.5 NTU and the removal efficiencies of TS (total solids content) and Si were 80% and 93 %, respectively. Further, for understanding the applicability of fractional factorial design and Taguchi experimental design, two laboratory-prepared carbon fibers/carbon/alumina TCMs (i.e., Tube B and Tube E obtained from two different preparation conditions) incorporated into the EC/EF treatment module were chosen for evaluating the performance of CMP wastewaters treatment. Permeate obtained based on the fractional factorial design of experiments had a turbidity of below 1.0 NTU and the removal efficiencies of TOC (total organic carbon), Cu and Si were all above 80 % except for the TS (i.e., ranging from 72 to 74%). Permeate obtained based on the Taguchi experimental design had a turbidity of below 0.3 NTU and the removal efficiencies of TS, TOC, Cu and Si were ranging from 82 to 91%. Apparently, similar optimum operating conditions were obtained from the fractional factorial design and Taguchi experimental design. Permeate thus obtained could be reused as the make-up water of cooling towers. The operating cost of Cu-CMP wastewater treatment based on a total water reclaim of 600 m3 per day was determined to be NT$ 98 (i.e., US$ 3.22) and NT$ 35 (i.e., US$ 1.05) per m3 of permeate for Case 1 (i.e., the filtration area of 0.0189 m2 in one EC/EF module) and Case 2 (i.e., the filtration area of 0.0801 m2 in one EC/EF module), respectively.