Freeze concentration is a thermal separation process that is used to purify aqueous solutions. One application of recent interest is seawater desalination. For freeze concentration to be an effective desalination method, a high ice growth rate and product purity must be achieved with energy usage comparable to that of competing technologies. The purpose of this thesis is to develop a coupled heat and mass transfer model to predict the growth rate and purity of the solid phase for ice grown about a horizontal, immersed tube. By simultaneously solving the heat and mass transfer problems, this model improves upon previous attempts found in the literature. In addition, an experimental apparatus was constructed and a series of ten experiments was run, considering a range of cooling rates, process times, and saltwater concentrations. Average ice growth velocities ranged from 3.1-13.1 mm/h and the observed partition coefficient ranged from 0.42-0.71. The model was calibrated using experimental data, and the coefficients of variation for the fitted model's prediction of ice mass and capture concentration were 15.4% and 7.6% respectively. Based on insights from modeling and experimentation, a series of suggestions are made regarding future modeling and process design. / Master of Science / Freeze concentration is a thermal process that is used to purify a liquid containing dissolved solids. One application of recent interest is seawater desalination. For freeze concentration to effectively purify seawater, a high ice growth rate and product purity must be achieved with energy usage comparable to that of competing technologies. The purpose of this thesis is to develop a coupled heat and mass transfer model to predict the growth rate and purity of the solid phase for ice grown about a horizontal, immersed tube. By simultaneously solving the heat and mass transfer problems, this model improves upon previous attempts found in the literature. In addition, an experimental apparatus was constructed and a series of ten experiments was run, considering a range of cooling rates, process times, and saltwater concentrations. Average ice growth velocities ranged from 3.1-13.1 mm/h and the salinity of the ice ranged from 0.42-0.71% of the original concentration. Based on insights from modeling and experimentation, a series of suggestions are made regarding future modeling and process design.

Thermal separation is defined as a mass transfer process driven by molecular forces. The process involves the heat transfer between two phases with different composition. In general, thermal separation occurs when heat is generated in the system additionally to the already existing phases. In a second phase the mass is transferred in the system (adsorption) and at the end of this step the separation is completed. The thermal separation can be achieved in temperature or concentration gradient function of system configuration [1]. Thermo-chromatography is a process in which the separation occurs in gase-ous phase. By passing a heated gas through a column a thermal gradient is created with a continuously decreasing temperature along the column. The separation occurs based on the different volatilization temperatures, the less volatile species will condense on the column walls at the higher temperatures and the highly volatile compounds will condense at lower temperatures. Parameters like temperature, carrier flow rate, column geometry and length have impact on the absorption of the compound on the column material affecting the separation efficiency. The thermal separation has been used for separation of Molybdenum (Mo) and Technetium (Tc) by either sublimation in the case of 94mTc {2,3,4] or dry distillation in the case of 99mTc from neutron irradiated MoO3 [5]. The thermal separation process has been used in the development of a new type of Mo/Tc generators starting from the MoO3 as target material for production of 99mTc in linear accelerators [6]. Dry distillation has become a standard procedure for separation of radioiodine from tellurium targets [7]. The present paper describes the thermal separation of a three component system (Cu/Mo/Tc) used as a target in the production of 99mTc through the 100Mo(p,2n) reaction. Material and Methods The separation method involves the use of oxygen as a carrier gas and oxidation agent. The method is based on the different volatilization temperatures of Tc formed oxides and the MoO3 formed in the system during the oxidation. In the presence of oxygen the existing Tc is oxidized to its anhydride as Tc2O7 (b.p. 319 ⁰C; m.p. 110.9 ⁰C) following the reaction: 4Tc + 7O2 →2Tc2O7 The T2O7 has a saturated vapor pressure of 310 ⁰C whilst Mo is completely oxidized to MoO3 having a sublimation temperature at 750 ⁰C. The initial experimental setup comprised a quartz tube (6 mm internal diameter, 40 cm long) which is introduced into a horizontal tube furnace (model 55035A, Lindberg). The left end of the quartz tube is connected to a pure oxygen supply which flows through the separation tube at a rate of 10 mL/min. The other end of the tube is opened to the atmosphere and protected with quartz wool. The quartz tube is heated over a length of 23 cm at a temperature of 850 ⁰C. The heated carrier gas is flowing on the tube length and the temperature gradient is created along the tube from 850 ⁰C to room temperature. During the process, the oxygen carries out the Tc oxides to a lower temperature and Tc2O7 is deposited in the cooler region of the tube in a similar manner as described by Tachimory [5]. The temperature gradient is calibrated by meas-uring the temperature inside the tube at each centimeter along its length (FIG. 1). The radioactivity counting is performed by scan-ning the tube along its length every 2 centimeters by using a detection system shown in figure 2. The system comprises a GM tube coupled to a computer controlled linear actuator (Velmex Unislide). The tube is placed at a distance of approximately 25 mm from the collimator of GM. Preliminary testing using Mo powder Prior to testing the three component separation, a reference test was performed by using 120 mg of natural Mo powder (Alpha Aesar, 99.9 %) soaked with 50 MBq NaTcO4 (Cardinal Health, radiochemical purity >95 %). After evaporation the dried powder was introduced into a quartz tube (6 mm ID, 40 mm long) and heated up to 850 ⁰C in the presence of oxygen flowing at a rate of 10 mL/min. Three component separation The targets prepared for the production of 99mTc by a cyclotron were comprised of copper (Cu) (C101, oxygen free) support having a Mo layer deposited on the surface in an elliptical form as described in literature [8,9]. About 60 to 250 mg of Mo (99.9%, Alpha Aesar) was deposited on the target surface. 70 MBq of Tc (Cardinal Health) as NaTcO4 (> 99 % radiochemical purity) was deposited on the Mo insert to mimic the conditions created during proton irradiation. The Tc spike was evaporated to dryness and the Cu/Mo/Tc target was then introduced into the experimental setup. The process was allowed to continue for 20 min. The experiment was carried out by inserting the target plates in a quartz tube (CanSci, Canada) of similar design to those described by Fonslet for the separation of radio-iodine from TeO2 targets [7]. The quartz tube can be seen in FIG. 2 and illustrated with dimensions indicated in FIG. 3. Separation of in-situ cyclotron produced Tc by irradiation of Mo targets with a proton beam. A third set of experiments have been performed for in-situ generated Tc by irradiation of circular targets containing approximately 60 mg Mo deposited on a copper support. The targets were irradiated for 30 min with a proton beam with the energy of 15 MeV and a current of 50 µA. The separation was performed using similar experimental conditions as previously described. The quartz tube was scanned in length by using a RadioTLC scanning system calibrated for 99mTc and 99Mo isotopes. After the thermal separation was completed 99mTc was recovered as NaTcO4 by selectively washing the quartz tube with 1 M NaOH (Fisher) solution. The presence of Mo in the NaTcO4 solution was verified by a colorimetric strip test (EM-Quant Mo test kit, Millipore). The presence of copper was qualitatively analyzed by adding a few drops of concentrated NH4OH (Fisher) solution and checking the formation of Schweitzer reagent. Results Thermal separation of Tc-Mo powder After 20 min the deposition of MoO3 was ob-served as yellow crystals in the region of tem-perature of 770 ⁰C, which is in accordance with the results reported in the literature [5]. The activity of 99mTc was detected at about 5 cm from the exit of the tube furnace in a temperature range starting with 310 ⁰C and ending at 46 ⁰C (FIG. 4).

Thermal separation is defined as a mass transfer process driven by molecular forces. The process involves the heat transfer between two phases with different composition. In general, thermal separation occurs when heat is generated in the system additionally to the already existing phases. In a second phase the mass is transferred in the system (adsorption) and at the end of this step the separation is completed. The thermal separation can be achieved in temperature or concentration gradient function of system configuration [1]. Thermo-chromatography is a process in which the separation occurs in gase-ous phase. By passing a heated gas through a column a thermal gradient is created with a continuously decreasing temperature along the column. The separation occurs based on the different volatilization temperatures, the less volatile species will condense on the column walls at the higher temperatures and the highly volatile compounds will condense at lower temperatures. Parameters like temperature, carrier flow rate, column geometry and length have impact on the absorption of the compound on the column material affecting the separation efficiency. The thermal separation has been used for separation of Molybdenum (Mo) and Technetium (Tc) by either sublimation in the case of 94mTc {2,3,4] or dry distillation in the case of 99mTc from neutron irradiated MoO3 [5]. The thermal separation process has been used in the development of a new type of Mo/Tc generators starting from the MoO3 as target material for production of 99mTc in linear accelerators [6]. Dry distillation has become a standard procedure for separation of radioiodine from tellurium targets [7]. The present paper describes the thermal separation of a three component system (Cu/Mo/Tc) used as a target in the production of 99mTc through the 100Mo(p,2n) reaction. Material and Methods The separation method involves the use of oxygen as a carrier gas and oxidation agent. The method is based on the different volatilization temperatures of Tc formed oxides and the MoO3 formed in the system during the oxidation. In the presence of oxygen the existing Tc is oxidized to its anhydride as Tc2O7 (b.p. 319 ⁰C; m.p. 110.9 ⁰C) following the reaction: 4Tc + 7O2 →2Tc2O7 The T2O7 has a saturated vapor pressure of 310 ⁰C whilst Mo is completely oxidized to MoO3 having a sublimation temperature at 750 ⁰C. The initial experimental setup comprised a quartz tube (6 mm internal diameter, 40 cm long) which is introduced into a horizontal tube furnace (model 55035A, Lindberg). The left end of the quartz tube is connected to a pure oxygen supply which flows through the separation tube at a rate of 10 mL/min. The other end of the tube is opened to the atmosphere and protected with quartz wool. The quartz tube is heated over a length of 23 cm at a temperature of 850 ⁰C. The heated carrier gas is flowing on the tube length and the temperature gradient is created along the tube from 850 ⁰C to room temperature. During the process, the oxygen carries out the Tc oxides to a lower temperature and Tc2O7 is deposited in the cooler region of the tube in a similar manner as described by Tachimory [5]. The temperature gradient is calibrated by meas-uring the temperature inside the tube at each centimeter along its length (FIG. 1). The radioactivity counting is performed by scan-ning the tube along its length every 2 centimeters by using a detection system shown in figure 2. The system comprises a GM tube coupled to a computer controlled linear actuator (Velmex Unislide). The tube is placed at a distance of approximately 25 mm from the collimator of GM. Preliminary testing using Mo powder Prior to testing the three component separation, a reference test was performed by using 120 mg of natural Mo powder (Alpha Aesar, 99.9 %) soaked with 50 MBq NaTcO4 (Cardinal Health, radiochemical purity >95 %). After evaporation the dried powder was introduced into a quartz tube (6 mm ID, 40 mm long) and heated up to 850 ⁰C in the presence of oxygen flowing at a rate of 10 mL/min. Three component separation The targets prepared for the production of 99mTc by a cyclotron were comprised of copper (Cu) (C101, oxygen free) support having a Mo layer deposited on the surface in an elliptical form as described in literature [8,9]. About 60 to 250 mg of Mo (99.9%, Alpha Aesar) was deposited on the target surface. 70 MBq of Tc (Cardinal Health) as NaTcO4 (> 99 % radiochemical purity) was deposited on the Mo insert to mimic the conditions created during proton irradiation. The Tc spike was evaporated to dryness and the Cu/Mo/Tc target was then introduced into the experimental setup. The process was allowed to continue for 20 min. The experiment was carried out by inserting the target plates in a quartz tube (CanSci, Canada) of similar design to those described by Fonslet for the separation of radio-iodine from TeO2 targets [7]. The quartz tube can be seen in FIG. 2 and illustrated with dimensions indicated in FIG. 3. Separation of in-situ cyclotron produced Tc by irradiation of Mo targets with a proton beam. A third set of experiments have been performed for in-situ generated Tc by irradiation of circular targets containing approximately 60 mg Mo deposited on a copper support. The targets were irradiated for 30 min with a proton beam with the energy of 15 MeV and a current of 50 µA. The separation was performed using similar experimental conditions as previously described. The quartz tube was scanned in length by using a RadioTLC scanning system calibrated for 99mTc and 99Mo isotopes. After the thermal separation was completed 99mTc was recovered as NaTcO4 by selectively washing the quartz tube with 1 M NaOH (Fisher) solution. The presence of Mo in the NaTcO4 solution was verified by a colorimetric strip test (EM-Quant Mo test kit, Millipore). The presence of copper was qualitatively analyzed by adding a few drops of concentrated NH4OH (Fisher) solution and checking the formation of Schweitzer reagent. Results Thermal separation of Tc-Mo powder After 20 min the deposition of MoO3 was ob-served as yellow crystals in the region of tem-perature of 770 ⁰C, which is in accordance with the results reported in the literature [5]. The activity of 99mTc was detected at about 5 cm from the exit of the tube furnace in a temperature range starting with 310 ⁰C and ending at 46 ⁰C (FIG. 4).