碩士 / 國立清華大學 / 化學工程學系 / 89 / Our research emphasizes on developing new methods to simulate the effective conductivities of the composites possessive of contact resistance. We concern composites with inclusions dispersed randomly in the matrix. The inclusions can be spherical particles, or infinitely long fibers. They are named particulate-reinforced composite and long-fiber-reinforced composites , respectively. There is an abundant knowledge about how to obtain the effective conductivities of composites. The critical Biot number, , gives “neutral effect.” That is, we can neglect interface properties while we treat composites’ effective conductivities. We focus on construction of a simulation model, that can be used to calculate the effective conductivities of the composites with any size or any irregular type inclusions dispersed randomly in the matrix. We image that thermal molecules do random walks among inclusions and matrix. Kim and Torquato (1990,1991) brought up the ideas whether the test thermal molecules will walk across the interface or not and how much time it will spend on crossing the interface or coming back. They obtained the effective conductivity of two phase composites. However, they considered single-sized inclusions, and a perfect contact with the matrix. (Temperature fields are continuous across the matrix-inclusion, so are the normal heat fluxes). They didn’t work on problems involving poly-dispersed inclusions with contact resistance at the matrix-inclusion interface. (Temperature fields are not continuous, but the normal heat fluxes are continuous.) However, it is common that contact resistance exists between matrix and inclusions. When the Biot number is infinite, contact resistance will disappear. Then, the matrix contacts the inclusions perfectly. Therefore, the contact resistance problem is a limiting case of the contact resistance problems. The new method is a breakthrough in the area of composites’ effective conductivities. We obtained satisfactory results of the effective conductivities of the fibers and particles arranged regularly in the matrix. Moreover, the effective conductivities of the particulate-reinforced and the fiber-reinforced composites with inclusions randomly distributed are also computed. And we confirm that Biot number is an important parameter again. When Bi is less than the critical Biot number, the combined effect of interface and inclusion is enhancing to the matrix. When Bi is greater than the critical Biot number, the combined effect of interface and inclusion is enhancing to the matrix. We also verify that the inclusions generated are dispersed randomly enough.

碩士 / 國立清華大學 / 化學工程學系 / 89 / Processes involving diffusion of small entities onto the surface of much larger inclusions and incorporation of these small entities into the much larger inclusions, are commonly found in nature and industry. There were many researches by computer simulation for diffusion-limited processes, but the researches for the non-diffusion-limited were a few. In this research, we suppose that there are many randomly distributed particles in space. By using boundary collocation method and Monte-Carlo simulation I compute the normalized overall rate constants of reaction when the surface reaction rate is finite. We can control the surface reaction rate to be finite or not by a dimensionless parameter P. As P=0, the process is diffusion limited. And as P→∞, the process is reaction-limited. When particles are randomly distributed in space, we see that normalized overall rate constants increase as increasing the volume fraction of particles, and normalized overall rate constants decrease as increasing the P value. If we compare normalized overall rate constants obtained from particles in random or in regular arrays, we see when the volume fraction of particles is smaller than 0.6, normalized overall rate constants of particles of random arrays are smaller. But when particles in are a very dilute situation, normalized overall rate constants increase with increasing the P value. And when particles are arranged in a rod (1D), square (2D) or cube (3D), normalized overall rate constant and the number of particle (N) will have the following relation ( D: fractal number ) : as D>1 as D=1 If we use the slopes from the equations and P value to make a plot, the plot can be separated into three regions：P=0~0.1, P=0.1-10, P=10-infinite. And the slopes decrease dramatically when P value change 0.1 from 10.

碩士 / 國立清華大學 / 化學工程學系 / 89 / In this research we form a novel intramolecular charge transfer complex --- Cyanine-TCNQ dye through grafting. Through the UV absorption of dye solution, fluorescence emission spectrum measurement and the electrochemical properties that is measured under the condition with or without exciting light source, we discover that the optics and electrochemical properties of Cyanine-TCNQ dye, which form intramolecular charge transfer complex are significantly different from Cyanine-I with the same main structure. Through the spectrum properties and redox potential measured of dye solution, molecular orbital of Cyanine-TCNQ is calculated and we further submit Two-Step Photoinduced Electron Transfer to explain the number of oxidation / reduction electron of Cyanine-TCNQ dye solution, which change according to time. It is almost the same reason with or without exciting light.

碩士 / 國立清華大學 / 化學工程學系 / 89 /

碩士 / 國立清華大學 / 化學工程學系 / 89 / The preparation of porous aluminum by die casting was investigated in this work, including both the 〝lost foam 〞method and the〝ceramic ball preform 〞method. In the lost foam die casting method, we utilized the swelling phenomenon caused by soaking the polymer foam in a proper organic solution to increase the ligament thickness and its porosity. As a result, the ligament thickness and porosity of finished products can be easily tailored. Using different polymer foams from various suppliers, we can make porous aluminums whose ligament thickness is in the range of 0.117∼0.458mm and porosity about 90∼95%. In the ceramic ball preform method, we found a special formula（weight ratio of Al2O3：bentonite：HPMC：water：PVA is 100：1：2：48.3：2.4） for making ceramic balls with sufficient strength to survive the casting procedures. These ceramic balls of several mm in size, can be easily removed from the finished products by high pressure water after casting. One of the ingenuity of this procedure lies in the fact that these ceramic balls are soft and deformable at room temperature. For this reason, we can apply pressure to increase the compactness of the packed bed made of these ceramic balls. As a result, the porosity the finished products can be substantially increased. On the other hand, the uniformity of the product can be improved if regular packing technique was employed rather than the random packing method. On the average, the porosity of the product is in the range of 80∼89%.

碩士 / 國立清華大學 / 化學工程學系 / 89 / 英文摘要 Monascus pigments are traditional natural food colorants in oriental countries. They are a mixture of yellow (monascin, ankaflavin), orange-red (rubropunctatin, monascorubrin) and red-purple (rubropunctamine, monascorubramine) compounds. Traditionally, solid-state culture is employed for Monascus species. The microorganisms are normally cultivated on steamed rice grains and proper temperature and humidity have to be maintained throughout the whole process. Hence, it is difficult to control the process parameters and to keep the cultivation process from contamination in a large-scale production. Consequently, the quality of the products is difficult to maintain. On the contrary, even though the capital investment involved for the submerged cultivation process is greater than that for the solid state cultivation, one can better control the process parameters and take advantage of the more flexible processing conditions. The products from the submerged cultivation process are expected to be purer and more uniform. However, the amount of products produced by solid-state fermentation is greater than that produced by submerged fermentation (Lin, 1982). This might be related to the composition and microstructure of the rice (Johns and Stuart, 1991). In this research, a combined process utilizing the advantages of both submerged and solid state cultivation was developed. The solid state cultivation was proceeded in a 5 L fermentor, while rice was employed not only as the substrate but also as the support for Monascus. To integrate the fermentation and separation of the red pigments, one intermittently rinsed the rice with monosodium glutamate solution every 12 hours to wash out the resulting red pigments. An adsorption column was employed to adsorb the red pigments in the MSG solution and the MSG solution was recycled. The result showed that, the amount of Monascus red pigments produced by the integration system is 20% more than that produced by the fixed-bed cultivation.

碩士 / 國立清華大學 / 化學工程學系 / 89 / In this paper, the dispensing process for flip chip underfill is discussed by theoretical analysis. Above all the simulation results have considered the properties of chemorheology for the thermosetting polymer, especially the effect of temperature and conversion. Using the collocated cell-centered finite volume method based on three-dimensional calculates the flow field during underfilling process, and predicts the shape and movement of the melt front with time with the method of computation of free surface which solves the volume fraction equation. In regard to the calculation of surface tension, we use continuum surface force (CSF) model. Castro-Macosko Model which considers the effect of conversion and shear rate is induced for providing a described of the resin viscosity change. Because the gap between bumps is slightly larger or smaller than the gap between chip and substrate, Hele-Shaw approximation is not suitable. In the work, we simulation the flow during underfilling process by using the three-dimensional technology and study the effect of temperature , dispensing pattern and bumps on the process further.

碩士 / 國立清華大學 / 化學工程學系 / 89 / The project intends to extend phthalocyanine’s applications to non-linear optical areas as optical limiting materials（OL）. The latter meets the need of eye and/or instrument protection for laser practice. When the intensity of incident laser beam exceeding certain limit of safety standard, OL keeps a constant level of input energy in order to avoid damage. In this project, we shall study the OL mechanism via a five-energy-state diagram by UV and fluorescence spectrometry. Phthalocyanine derivatives, especially those absorption peaks beyond 800nm exceeding the visible range（400-800nm）, shall be synthesized. Its performance shall be evaluated through Z-scan by laser.

碩士 / 國立清華大學 / 化學工程學系 / 89 /

碩士 / 國立清華大學 / 化學工程學系 / 89 /