Lactic acid recovery from dilute fermentation broths is a growing requirement due to the increasing demand for pure lactic acid. Reactive extraction is proposed as an alternative to conventional methods of recovery, since the selectivity of separation is remarkably enhanced in reactive extraction. The aim of this study is to perform the equilibrium studies for the recovery of lactic acid from its synthetic aqueous solutions (not from real fermentation broths) by reactive extraction and investigate the effects of various parameters such as initial lactic acid concentration in the aqueous phase (0.25 - 1.3 M), initial pH of the aqueous phase (2 &ndash / 6), organic phase extractant concentration (0.1 &ndash / 0.5 M), type of the extractant (chloride, hydrogensulphate and hydroxide salts of tri-n-octylmethylammonium) and the type of diluent (oleyl alcohol or octanol). The results of the experiments showed that the degrees of extraction decreased with increasing use of diluent with the extractant and increasing initial lactic acid concentration of the aqueous phase. Highest degrees of extraction were achieved for undiluted extractants. The performance of the diluents were investigated by performing extraction experiments with solutions of TOMAC in oleyl alcohol or octanol at different pH values and it was observed that octanol had a higher solvating power than oleyl alcohol especially at lower aqueous phase pH values. Higher extraction efficiencies were obtained for TOMAC dissolved in octanol rather than oleyl alcohol. Initial aqueous pH of 6 was identified as the optimum pH for the extraction, also due to its being equal the average fermentation pH for the extractions with Lactobacillus species. Among the different salts of tri-n-octylmethylammonium, hydroxide salt exhibited the highest degrees of extraction (83% with undiluted TOMA(OH) and 78% with 0.5 M TOMA(OH) in octanol for the extraction of 0.316 M lactic acid solutions). The present work is the first step in the design of an industrial reactive extraction process that is going to attempt forward and backward extraction of lactic acid simultaneously in a hollow fiber membrane module that is going to be attached to the lactic acid fermentor to achieve continuous product recovery. The equilibrium data obtained from this study can be combined with the kinetic studies as the next step of designing a separation module.

Neuro-Fuzzy systems are the systems that neural networks (NN) are incorporated in fuzzy systems, which can use knowledge automatically by learning algorithms of NNs. They can be viewed as a mixture of local experts. Adaptive Neuro-Fuzzy inference system (ANFIS) is one of the examples of Neuro Fuzzy systems in which a fuzzy system is implemented in the framework of adaptive networks. ANFIS constructs an input-output mapping based both on human knowledge (in the form of fuzzy rules) and on generated input-output data pairs. Effective control for distillation systems, which are one of the important unit operations for chemical industries, can be easily designed with the known composition values. Online measurements of the compositions can be done using direct composition analyzers. However, online composition measurement is not feasible, since, these analyzers, like gas chromatographs, involve large measurement delays. As an alternative, compositions can be estimated from temperature measurements. Thus, an online estimator that utilizes temperature measurements can be used to infer the produced compositions. In this study, ANFIS estimators are designed to infer the top and bottom product compositions in a continuous distillation column and to infer the reflux drum compositions in a batch distillation column from the measurable tray temperatures. Designed estimator performances are further compared with the other types of estimators such as NN and Extended Kalman Filter (EKF). In this study, ANFIS performance is also investigated in the adaptive Neuro-Fuzzy control of a pH system. ANFIS is used in specialized learning algorithm as a controller. Simple ANFIS structure is designed and implemented in adaptive closed loop control scheme. The performance of ANFIS controller is also compared with that of NN for the case under study.

In this study, the performance of the Model Predictive Controller (MPC) algorithm is investigated in two different reaction systems. The first case is a saponification reaction system where ethyl acetate reacts with sodium hydroxide to produce sodium acetate and ethanol in a CSTR. In the reactor, temperature and sodium acetate concentration are controlled by manipulating the flow rates of ethyl acetate and cooling water. The model of the reactor is developed considering first principal models. The experiments are done to obtain steady state data from the reaction system and these are compared with the model outputs to find the unknown parameters of the model. Then, the developed model is used for designing SISO and MIMO-MPC considering Singular Value Decomposition (SVD) technique for coupling. The second case is the reaction system used for the production of boric acid by the reaction of colemanite and sulfuric acid in four CSTR&rsquo / s connected in series. In the reactor, the boric acid concentration in the fourth reactor is controlled by manipulating the sulfuric acid flow rate fed to the reactor. The transfer functions of the process and disturbance (colemanite flow rate) are obtained experimentally by giving step changes to the manipulated variable and to the disturbance. A model-based and constrained SISO-MPC is designed utilizing linear step response coefficients. The designed controllers are tested for performance in set point tracking, disturbance rejection and robustness issues for the two case studies. It is found that, they are satisfactory except in robustness issues for disturbance rejection in boric acid system.

In this research, molybdophosphoric acid (PMo) was blended with polysulfone polymer (PSF) and form a film catalyst by using a common solvent dimethylformamide (DMF). Kinetic and mass transfer parameters were evaluated for catalytic films in ethanol lactic acid esterification reaction as film surface area, film thickness and catalyst loading were varied at 50&amp / #61616 / C, 1 atm and 1:1 ethyl alcohol to lactic acid mole ratio conditions. Also prepared films were characterized by DSC, TGA, FTIR, X-ray and SEM analysis. It was observed that the catalytic films showed higher activities with respect to the unloaded form of PSF and activities were increased with the increasing loading levels. The stabilities of the loaded catalysts were tested by means of deactivation experiments. A decrease was observed after 5th trial for 10wt% PMo loaded PSF film and at 4th trial for 15wt% PMo loaded PSF film. However, activities of the loaded films gave still higher conversion results than the unloaded PSF film. Also it was proved that with increasing film thickness conversion was decreased and increasing surface area conversion was increased. It was observed from the characterization studies that PMo catalysts have no chemical interaction with the PSF polymer and there was trace amount of DMF solvent was observed in the PMo-PSF catalytic film. Particulate structure of the PMo catalyst was observed and there is no catalyst agglomeration in the membrane network. However, along the thickness, catalyst particles were not homogeneously but finely dispersed in amorphous film structure.

In the control of batch and continuous distillation columns, one of the most challenging problem is the difficulty in measuring compositions. This problem can be handled by estimating the compositions from readily available online temperature measurements using a state observer. The aim of this study is to design a state observer that estimates the product composition in a multicomponent batch distillation column (MBDC) from the temperature measurements and to test this observer using a batch column simulation. To achieve this, first a model for MBDC is prepared and compared with the data from literature where a case column is utilized. After checking the validity of the simulation package, it is used as a fictitious process for the performance evaluations. In the second phase of the study, an extended Kalman Filter (EKF) is designed by utilizing a simplified model of MBDC and it is implemented for performance investigation on the case column with 8 trays separating the mixture of cyclohexane, n-heptane and toluene. The simplified model utilized in EKF results in response, which have some deviation with rigorous model, mainly due to the simplification of vapor-liquid equilibrium relationship. In the performance evaluation, the tuning parameters of EKF / the diagonal terms of process noise covariance matrix and the diagonal terms of measurement model noise covariance matrix are changed in the range of 50&iexcl / 1x10&iexcl / 7 and 0:5&iexcl / 5x108 and the optimum values are found as 0:00001 and 5000, respectively. The effect of number of measurement points is also investigated with a result of number of component measurements. The effect of measurement period value is also studied and found that it has a major effect on the performance which has to be determined by the available computational facilities. The control of the column is done by utilizing the designed EKF estimator and the estimator is successfully used in controlling the product purities in MBDC under variable reflux-ratio operation.

Discovery of mesoporous materials by Mobil researchers in 1992 opened a new field in catalytic applications. The materials designated as M41S family are MCM-41 with one-dimensional hexagonal structure, MCM-48 with three-dimensional cubic structure and MCM-51 with unstable lamellar structure. This family of materials have high surface areas up to 1500 m2/g, narrow pore size distributions with pore sizes varying from 20 to 100 &Aring / . These materials can be activated by incorporation of metals or active compounds into their structures. In this study, copper and nickel incorporated MCM-41 type catalytic materials were synthesized via different methods namely, impregnation, high temperature and low temperature direct synthesis methods. The Cu-MCM-41, and Ni-MCM-41, as well as synthesized MCM-41 were characterized by using XRD, TEM, N2 sorption, SEM, XRF, EDS, AAS and TPR. MCM-41 was synthesized with high temperature direct synthesis method. High surface area values up to 1400 m2/g of MCM-41 mesoporous materials were obtained with high pore volumes up to 1.17 cc/g. Cu-MCM-41 type catalytic materials were synthesized with three different methods. Impregnation and high temperature direct synthesis methods gave better results than those of low temperature direct synthesis method. In impregnation, relatively high surface area values (730 m2/g) were obtained with Cu/Si mole ratio as high as 0.3 in the product. For the case of high temperature direct synthesis products, Cu/Si mole ratios as high as 0.26 were obtained with somewhat smaller surface areas (400 m2/g). Low temperature direct synthesis method is the least favorable method in metal loading. Ni-MCM-41 type of catalytic materials were synthesized by impregnation and high temperature direct syntheses methods. Ni incorporation by high temperature direct synthesis method gave high surface area values (560-930 m2/g) having Ni/Si mole ratios of 0.12-0.28.

The aim of this thesis work was to extract lipids and antioxidants from wheat germ using an ultrasonic bath. Alternative solvents: Ethanol, isopropanol and acetone were used for the extraction purposes and for the fat content determination hexane was used. Alternative solvents dissolve wax, phospholipids and some other proteins because of their polar nature which increased the yield. Since waxes and some proteins can be added to the soaps and creams which were made from carrier oils, obtaining these substances together with the oil in the extract was very useful. For isopropanol a linear relation was found between the extract ratio and ultrasonication time. The highest extract ratios were achieved with ethanol. For 30 minutes of ultrasonication 0.10 g extract / g germ was obtained. After the extraction in order to enhance the phase separation between the solid phase (germ and other solids) and the extract phase, combinations of centrifugation, storage in the refrigerator and decantation were tested. The technique involving storage the extracts in the refrigerator for 24 hours and centrifugation at 2800 rpm for 20 minutes gave the second highest yield but the best phase separation. Ethanol extracts from both roasted and nonroasted wheat germs were characterized in terms of their total polyphenol contents. For non roasted wheat germs total polyphenol contents of 200 mg gallic acid / L solution and for roasted wheat germs an average of 170 mg gallic acid / L solution were obtained for 30 minutes of ultrasonic extraction. For nonroasted wheat germs weak but significant linear relations were found between total polyphenol contents of the extracts and ultrasonication time.

Zinc borate is a flame retardant additive used in polymers, wood applications and textile products. There are different types of zinc borate having different chemical composition and structure. In this study, the production of zinc borate that had the molecular formula of 2ZnO.3B2O3.3,5H2O was studied. The aim of this study was to investigate the effects of reaction parameters on the properties of zinc borate that had been synthesized by the reaction of boric acid and zinc oxide at the existence of the seed crystals and to determine the optimum experimental conditions for zinc borate production reaction. Reaction kinetics was also investigated to find a suitable kinetics model. The effect of boric acid to zinc oxide ratio -H3BO3:ZnO ratio- (3:1, 3.5:1, 5:1 and 7:1), the particle size of zinc oxide (10&micro / m and 25&micro / m), stirring rate (275 rpm, 400 rpm, 800 rpm and 1600 rpm), temperature (75&deg / , 85&deg / and 95&deg / ) and size of seed crystals (10&micro / m and smaller size) on reaction rate, reaction completion time, composition and particle size distribution of zinc borate were investigated. Experiments were performed in a continuously stirring, temperature controlled batch reactor with a volume of 1.5L. During the experiments samples were taken to be analyzed in regular time intervals. The analyses of the samples gave the concentration change of zinc oxide and boron oxide in the solid as well as the conversion of zinc oxide to zinc borate with respect to time and the rate of reaction was calculated. The products were also analyzed for particle size distribution. The experimental results showed that the reaction rate increased with the increasing H3BO3:ZnO ratio, particle size of zinc oxide, stirring rate and temperature. The reaction completion time was also decreased by increasing H3BO3:ZnO ratio, stirring rate and temperature. The particle size of final product, zinc borate, decreased with increasing stirring rate and size of zinc borate used as seed and increased with increasing particle size of zinc oxide used as reactant. The average particle sizes of the final product zinc borates synthesized at the end of the experiments were ranged between 4.3 &micro / m and 16.6 &micro / m. The zinc borate production reaction was mainly fitted the unreacted core model for the case of diffusion through product layer controls.

In this research, oxidative dehydrogenation of ethane to ethylene was studied over Cr-O and Cr-V-O mixed oxide catalysts through steady-state and dynamic experiments. The catalysts were prepared by the complexation method. By XRD, presence of Cr2O3 phase in Cr-O / and the small Cr2O3 and V2O4 phases of Cr-V-O were revealed. In H2-TPR, both catalysts showed reduction behaviour. From XPS the likely presence of Cr+6 on fresh Cr-O was found. On Cr-V-O, the possible reduction of V+5 and Cr+6 forms of the fresh sample to V+4, V+3 and Cr+3 states by TPR was discovered through XPS. With an O2/C2H6 feed ratio of 0.17, Cr-O exhibited the highest total conversion value of about 0.20 at 447&deg / C with an ethylene selectivity of 0.82. Maximum ethylene selectivity with Cr-O was obtained as 0.91 at 250&deg / C. An ethylene selectivity of 0.93 was reached with the Cr-V-O at 400&deg / C. In the experiments performed by using CO2 as the mild oxidant, a yield value of 0.15 was achieved at 449&deg / C on Cr-O catalyst. In dynamic experiments performed over Cr-O, with C2H6 pulses injected into O2-He flow, the possible occurrence of two reaction sites for the formation of CO2 and H2O was detected. By Gaussian fits to H2O curves, the presence of at least three production ways was thought to be probable. Different from Cr-O, no CO2 formation was observed on Cr-V-O during pulsing C2H6 to O2-He flow. In the runs performed by O2 pulses into C2H6-He flow over Cr-V-O, formation of CO rather than C2H4 was favored.

In this study, the effect of Pt particle size on the reaction rate and selectivity of preferential oxidation of CO (PROX) reaction was investigated on Pt/Al2O3. 2% Pt/&amp / #947 / -Al2O3 catalysts were prepared by incipient wetness method / the particle size of the catalysts was modified by calcination temperature and duration. Therefore, the relative amounts of low and high coordination atoms on the metal particle surface can be changed. Over these catalysts, first, the CO oxidation reaction was studied in the absence of hydrogen. The catalyst having the highest dispersion, i.e., lowest metal particle sizes, had the highest activity as indicated by its lowest light-off temperature. On the other hand, the turnover frequencies (TOF) of the catalysts were increasing with decreasing dispersion. The activation energy of the catalysts were also compared and examined: as the particle size increased, the activation energy decreased. In the second part, preferential oxidation of CO reaction in the presence of hydrogen was studied. Both CO conversion and selectivity first increased with increasing reaction temperature, then exhibited a maximum, and finally decreased. Both CO conversion and selectivity did not show any trend for different dispersed catalysts for &amp / #955 / (2PO2/PCO) was 1. In order to reach a definite conclusion about the structure sensitivity of selective CO oxidation, the experiments with different &amp / #955 / &rsquo / s and space times over the same catalysts should be performed.