Modeling slurry oxidation

A solid-liquid mass transfer coefficient of 0.015 cm/s was found by comparing the predictions of [S(IV)] to experimental results obtained under conditions in which the liquid phase kinetics were fast. The model was then applied to slurry oxidation under more general conditions by using liquid phase reaction rate kinetics obtained in clear solutions. The results of the model agree with experimental findings for the total rate of oxidation. 

Modeling of the Slurry Oxidation. In 1969 Ramachandran and Sharma(18) first proposed a film model for gas absorption accompanied by a fast chemical reaction in a slurry containing sparingly soluble, fine particles. A first case assumed that the solid dissolution in the liquid film next to the gas-liquid interface was unimportant. The second case assumed that it was important. Numerical solutions were given for the second case which indicated hat the specific rate of absorption of gas in the presence of fine particles could be considerably higher than in the absence of solids. 

Variable Catalyst Concentration Solutions. Using the mass transfer coefficient of 0.015 cm/sec, the model was then used to simulate the slurry oxidation with three concentrations of added Mn catalyst. Results are presented in Figure 8. The 1.5 order homogeneous reaction rate constants for 0, 6.66, and 200 ppm added Mn reactions were found from the model to be 0.35, 2.25, and 5.5 2,0 5 mol-0 5 sec-1 respectively. The corresponding values of 1.5 order rate constants from the comparable clear solution experiments are 0.162, 0.35, and 0.8 to 5... 

Slurry Density Variations. The model was also used to simulate the slurry oxidations with different initial conditions. Initial pH s of 4.5 and 5.5 were both tested. Results (total concentration and solution phase concentration curves) for the initial pH 4.5 are presented in Figure 9. By using the predicted mass transfer coefficient and rate constants, the computer curves can match these experimental results very well. 

By incorporating the film theory into the mathematical model for the batch slurry oxidation, a mass transfer coefficient of 0.015 cm/sec was obtained by matching the model to highly catalyzed (2000 ppm Mn added) slurry oxidation data. Saturation concentration of sulfite is most important in determining mass transfer coefficient(32). A correlation is given... 

Wang, C. C., "Experiments and Modeling of Calcium Sulfite Slurry Oxidation," M.S. Thesis, University of Virginia... 

The solids analysis described above can be taken to yet another level by correlating the color measurement to chemical properties. An excellent model system is vanadium pyrophosphate (VPO), which is a well-known catalyst for butane oxidation to maleic anhydride. During the synthesis of the catalyst precursor, solid V2O5 particles are dispersed in a mixture of benzyl alcohol and i-butanol. In this slurry phase, the vanadium is partly reduced. Addition of phosphoric acid leads to a further reduction and the formation of the VPO structure. With a diffuse reflectance (DR) UV-vis probe by Fiberguide Ind., the surface of the suspended solid particles could be monitored during this slurry reaction. Four points can be noted from Figure 4.4 ... 

The forward reaction is favored by the alkaline slurry solutions which result in breakage of the Si—O bonds. In metal CMP, oxidizing slurries are often used, resulting in faster removal rates. Since the contributions of the chemical and mechanical components are not well known, modeling efforts have focused on empirical approaches guided by physical intuition of process mechanisms. 

Promotion and deactivation of unsupported and alumina-supported platinum catalysts were studied in the selective oxidation of 1-phenyl-ethanol to acetophenone, as a model reaction. The oxidation was performed with atmospheric air in an aqueous alkaline solution. The oxidation state of the catalyst was followed by measuring the open circuit potential of the slurry during reaction. It is proposed that the primary reason for deactivation is the destructive adsorption of alcohol substrate on the platinum surface at the very beginning of the reaction, leading to irreversibly adsorbed species. Over-oxidation of Pt active sites occurs after a substantial reduction in the number of free sites. Deactivation could be efficiently suppressed by partial blocking of surface platinum atoms with a submonolayer of bismuth promoter. At optimum Bi/Ptj ratio the yield increased from 18 to 99 %. 

The air-oxidation of 1-phenylethanol to acetophenone in an aqueous alkaline solution has been chosen as a model reaction. The catalytic experiments were completed with the application of an in-situ electrochemical method for studying catalyst deactivation and the role of promoters. The potential of the catalyst, which was considered as a slurry electrode, was measured during the oxidation reaction. More details of the method can be found elsewhere 13,14). 

In the production of formic acid, a slurry of calcium formate in 50% aqueous formic acid containing urea is acidified with strong nitric acid to convert the calcium salt to free acid, and interaction of formic acid (reducant) with nitric acid (oxidant) is inhibited by the urea. When only 10% of the required amount of urea had been added (unwittingly, because of a blocked hopper), addition of the nitric acid caused a thermal runaway (redox) reaction to occur which burst the (vented) vessel. A small-scale repeat indicated that a pressure of 150—200 bar may have been attained. A mathematical model was developed which closely matched experimental data. 

In order to reveal the nature of deactivation, the potential of the catalyst slurry was continuously measured during the partial oxidation of alcohols. Cyclic voltammetric measurements [16] were also performed in the same aqueous alkaline solution with model (unsupported) catalysts for the interpretation of the potential values. The experiments revealed that the oxidation of alcohols may be divided into three groups. The basis of classifying is the oxidation state of proroot-ed catalyst and its surface coverage with hydrogen or oxygen (OH) during reaction. 

In the design of upflow, three phase bubble column reactors, it is important that the catalyst remains well distributed throughout the bed, or reactor space time yields will suffer. The solid concentration profiles of 2.5, 50 and 100 ym silica and iron oxide particles in water and organic solutions were measured in a 12.7 cm ID bubble column to determine what conditions gave satisfactory solids suspension. These results were compared against the theoretical mean solid settling velocity and the sedimentation diffusion models. Discrepancies between the data and models are discussed. The implications for the design of the reactors for the slurry phase Fischer-Tropsch synthesis are reviewed. 

The derivation of the model follows Langmuir-Hinshelwood style kinetics. The surface at any point in time is assumed to present an array of reaction sites to the slurry, of which a fraction 62 has reacted with some component of the slurry, such as an oxidizer, whereas the remaining fraction 61 = 1 — 62 has not yet reacted. For example, the unreacted fraction may represent a portion of the surface that has been recently uncovered by mechanical action. The rate ri (in MKS units of mol/m s) at which unreacted sites are converted to reacted sites is assumed to follow first-order kinetics that is, the rate is proportional to the product of the fraction 61 of available sites and the molar concentration c of the reactant in the slurry ... 

An efficient slurry health monitoring tool should be able to provide both chemical as well as abrasive particle information on a continuous basis. There have been some efforts in this direction using an NIR absorption spectrum based analyzer [19]. This unit can provide oxidizer concentration and abrasive particle information in CMP slurry and operates on the principles of chemometrics, which is a two-phase process. In the first calibration phase, samples with known property values are measured by the system. A mathematical procedure then determines the correlation between the measured spectra and the true property values. The output of this phase is a model that optimally calculates the parameter values from the measured spectra of the calibration samples. In the second measurement phase, unknown samples are measured by the system, employing a model to produce estimates of the property values. 

The model contends that material is removed from the surface primarily by mechanical abrasion. The abraded material is either dissolved into the slurry, swept away from near the surface as undissolved copper or copper oxides by the fluid motion of the slurry, or redeposited onto the surface. The polish rate is then the abrasion rate minus the redeposition rate. Material is removed from the surface by chemical etching secondarily only, if at all. 

From all observations in compendium, it is reasonable to model the CMP removal rate for this particular slurry as a temperature-activated, abrasion-assisted etch process, in which a by-product of a few monolayers forms to inhibit - but not arrest - the etch process. The abrasive serves the dual purpose of increasing the temperature and clearing the transformed surface to increase the efficacy of the etch. The oxidizer is really a controlled etcher/complexer without which the abrasion rate is low. Under normal polishing conditions, the removal rate is limited by the formation of the surface complex. This explains the failure to observe a distinct oxidation layer in TEM as well as the temperature dependence on rate. The model does not require copper redeposition to explain low rate. 

---