Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

1 downloads 0 Views 2MB Size Report
A laboratory-scale spout-fluid bed reactor with a dc plasma torch was used to study the solid-phase synthesis of calcium carbide. Calcium oxide powder with a ...
Plasma Chemistry and Plasma Processing, Vol. 18, No. 3, 1998

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor M. H. El-Naas,1 R. J. Munz,1 and F. Ajersch2 Received August 17, 1996; revised December 4, 1997

A laboratory-scale spout-fluid bed reactor with a dc plasma torch was used to study the solid-phase synthesis of calcium carbide. Calcium oxide powder with a mean particle size of 170 um was reacted with graphite powder (130 um). Argon was used to initiate the plasma and hydrogen gas was then added to increase power and raise the plasma jet enthalpy. Experimental results showed that the reaction took place in the vicinity of the plasma jet and that conversion to calcium carbide increased linearly with reaction time. The rate of conversion increased exponentially with plasma jet temperature, indicating that chemical reaction was the controlling mechanism. Microscopic analysis of the solid product showed that calcium carbide was formed around both reactants, and that the reaction followed a shrinking core model. Although melting and agglomeration of partially reacted particles occurred at high temperature, resulting in instability of the bed and impeding the reaction progress, high conversions are expected in a continuous process with optimized reactor design. KEY WORDS: Calcium carbide; solid-phase synthesis; plasma synthesis; plasma process; plasma spout-fluid bed; solid-phase reactions.

1. INTRODUCTION Calcium carbide is an important industrial commodity. Its growing industrial applications in the desulfurization of steel and cast iron and in the production of acetylene and cyanamide have given it great importance as a chemical and further enhanced commercial interest in its production. Calcium carbide is presently produced by reacting calcium oxide and carbon in large electric arc furnaces at about 2400 K. The molten product is tapped from the furnace at about 2073 K, cooled, cast, and then crushed and ground to the size required for marketing. The energy consumption for a typical electric arc furnace is of the order of 4 kWh/kg CaC2. Several methods, 1CRTP,

Department of Chemical Engineering, McGill University, 3610 University St., Montreal, Quebec H3A 2B2, Canada. 2Departement de Metallurgie et de Genie des materiaux, Ecole Polytechnique, C.P. 6079 Succ. Centre-ville, Montreal, Quebec H3C 3A7, Canada.

409 0272-4324/98/0900-0409$15.00/0 © 1998 Plenum Publishing Corporation

410

El-Nans, Munz, and Ajersch

aiming at developing a more efficient process, have been studied for the production of calcium carbide using different furnace designs and different heating mechanisms.(1-4) Baba and Shohata(1) and Eriksson(2) patented processes using high-frequency plasma for the production of calcium carbide. Both processes, however, have an energy consumption that is much higher than that of the present electric arc furnace. Calcium carbide is formed by reacting calcium oxide and carbon according to the following reaction

This reaction is believed to proceed according to the following two-step mechanism first proposed by Kameyama(12)

Tagawa and Sugawara(13) studied the solid-phase reaction of carbon and calcium oxide in pellets and found that the two solid reactants formed a solid interstate compound CaO • C at temperatures above 1273 K, where the diffusion of carbon into calcium oxide and vice versa was believed to be the controlling step. Mukaibo and Yamanka(14) studied the kinetics of reaction (2) by heating pellets of calcium oxide and carbon between 1473 and 1673 K under vacuum. They found the reaction to be zero-order and the reaction rate rather than diffusion to the rate-limiting step. Reaction (3) is an essential step in the formation of calcium carbide through the surface reaction of calcium vapor and carbon. However, the mechanism by which calcium vapor and carbon monoxide are formed in a solid-phase reaction is not well understood. Hellmold and Gordziel(15) and Muller(16) discussed the idea of ionic diffusion as a means of carbon transport into the CaO lattice and of reducing the oxide to Ca(g). According to Muller,(l7) it is the role of the carbon materials to provide the C3 molecules to initiate the ionization, form another "interstate" compound (CaC3O), and eventually form calcium vapor. Once formed, calcium vapor reacts with either the diffused carbon at the surface of the oxide or with the free carbon to form calcium carbide. It is this surface reaction (3) that is believed to control the overall rate of formation of calcium carbide. A new plasma process has been studied to replace the present electric arc furnace. In this new process, the reaction between calcium oxide and carbon takes place in the solid phase in a plasma fluid bed reactor. Plasma fluidized and spouted beds have been studied extensively in the past thirty years.(5-10) The combination of high plasma temperature and the good mixing associated with fluidization make plasma fluidized beds ideal for the

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

411

highly endothermic gas-solid and solid-solid reactions. A thermodynamic analysis of the formation of calcium carbide in a plasma fluidized bed predicted that the reaction would be completed at 2150 K and that the plasma process could lower the energy consumption by up to 40%.(11) The use of a solid-state reaction with fine particle reactants will produce a granular product and thus avoid the cost of grinding associated with the electric arc process. Also, the sensible heat of the products can be recovered to heat the reactants and hence lower the energy requirements. 2. EXPERIMENTAL METHODS The experimental apparatus consisted of a power supply, a dc plasma torch, a fluid bed, a CO analyzer, and a temperature measuring system as shown schematically in Fig. 1. The plasma torch was composed of a conical thoriated tungsten cathode and an annular copper anode, with a design power of the order of 20 kW. Power to the torch was controlled by varying the current and the plasma gas composition. The fluid bed system consisted of two parts: the reactor section and the disengaging section. The inner part of the reactor was lined with a graphite cylinder (8 cm in internal diameter and 35 cm in height) surrounded by 2 cm of graphite felt insulation. Both the graphite cylinder and the felt were contained within a stainless steel cylinder (13 cm in internal diameter) which was also insulated on the outside

Fig. 1. A schematic diagram of the experimental apparatus.

412

El-Naas, Munz, and Ajersch

Fig. 2. A detailed representation of the fluid bed reactor.

to minimize heat loss. The lower part of the reactor housed the distributor and the plasma torch. The stainless steel disengaging section measured 25 cm in internal diameter and 25 cm in height. The large diameter lowered the gas velocity and minimized elutriation of particles. A detailed representation of the fluid bed reactor is shown in Fig. 2. Argon at a flow rate of 40 L/min was used to initiate the arc, and hydrogen was subsequently added to increase the arc voltage and hence raise the power and plasma enthalpy. Three different levels of hydrogen content were used in the experiments (33,45, and 67 vol.%). Argon was also used to fluidize the bed at just above the minimum fluidization velocity. The bed temperature ranged from ambient to about 1573 K and was measured using Accufiber High Temperature Measurement and Control System Model

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

413

100C. Calcium oxide powder with a mean particle size of 170 um was reacted with graphite powder (130um). The reactants were charged into the batch reactor at a stoichiometric ratio of 3 C to 1 CaO with a total mass of 1 kg. The gaseous products leave the reactor with the plasma gas, where they pass through a heat exchanger and a baghouse filter before evacuation into a fume hood. The gases were continuously analyzed for carbon monoxide using a Horiba Mexa-201GE CO analyzer. 3. RESULTS AND DISCUSSION 3.1. Effect of Plasma Jet Temperature Argon at a flow rate of 40 L/min was used to initiate the arc and start the plasma torch. The maximum power that was achieved with argon plasma, however, was too low to provide a temperature high enough for good conversion to calcium carbide. Hydrogen was then added to the plasma gas to increase the arc voltage and hence raise the plasma power. For a fixed current (240 A), the arc voltage increased with hydrogen addition, resulting in a power increase from 5 kW with pure argon to 18 kW with 67% hydrogen. Hydrogen addition increases the power as well as the thermal conductivity and the heat capacity of the plasma gas. Both the enthalpy and the temperature of the plasma jet are increased as the hydrogen content is increased at constant current. The jet temperature was calculated by performing an energy balance with a torch efficiency of 60%, measured experimentally. While keeping the current fixed at 240 A, experiments were conducted for different plasma powers by varying the hydrogen concentration in the plasma gas. Figure 3 shows a plot of the integral conversion of calcium oxide as a function of run time for four plasma gas compositions (H 2 , vol.% in Ar): 0%, 33%, 45%, and 67%. Conversion was calculated from the cumulative total quantity of carbon monoxide measured experimentally. Thus, conversion at any time was calculated as the moles of carbon monoxide produced per mole of calcium oxide initially fed to the reactor. It can be clearly observed that the conversion increased linearly with time, and the rate of conversion increased with plasma power or hydrogen concentration in the plasma gas. The rate of conversion at any time represented the instantaneous global rate of reaction. This rate was found to be constant with time for a fixed power (fixed plasma jet temperature) and increased exponentially with increasing plasma jet temperature as is shown in Fig. 4. The reaction rates at different conditions are shown in Table I. One would expect the rate of conversion to increase with time, since the bed was heating up and the bed temperature was increasing with the run time as shown in Fig. 5. The fact

414

El-Naas, Munz, and Ajersch

Fig. 3. Conversion of calcium oxide to carbon monoxide as a function of run time for different plasma powers. Carbon to calcium oxide molar ratio was 3:1 and plasma gas flow rate was 40 L/min.

that the rate was constant indicated that the change in reaction temperature with time for the duration of the experimental run was negligible. A plot of conversion vs. bed temperature for different conditions is shown in Fig. 6. The highest conversion (30%) occurred at a bed temperature of 1573 K, which is less than the temperature required thermodynamically for the reaction to proceed. This clearly indicated that the reaction between calcium oxide and graphite did not take place in the bed, but in the vicinity of the plasma jet where the temperature was sufficiently high for the reaction to occur. Thus, the fluid bed reactor is assumed to consist of two different zones: a high-temperature reaction zone and a well-mixed isothermal bed zone. The size of the reaction zone depends on the plasma conditions and is proportional to the plasma jet enthalpy. The bed zone, which represents most of the reactor volume, acts as a mixing zone that feeds particles into

415

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

Fig. 4. Rate of conversion vs. calculated plasma jet temperature.

Table I. Reaction Rates at Different Conditions

H2 (vol.%)

Power (kW)

Calculated Tjet (K)

Reaction rate (mol/min)

0 33 45 67

5 12 14 18

8400 9400 9800 10200

0.006 0.026 0.042 0.091

the jet zone where they react. A model of the plasma reactor with the two zones is described in detail in a recent publication.(18) As particles enter the reaction zone, their temperature rises from the bed temperature to a maximum particle temperature. Except for the first 45 min of each run, the rate of increase in the bed temperature is relatively small. On the other hand, the increase of the particle temperature is very

416

El-Naas, Munz, and Ajersch

Fig. 5. Bed temperature vs. run time for different plasma powers. Plasma gas flow rate was 40 L/min and current was 240 A.

high as the particles pass through the reaction zone. Thus, the difference in the bed temperature for the entire run is much smaller than the particle temperature increase in the jet. As a result, the particle temperature, which determines the surface reaction temperature, is almost constant with time and explains the constant rate of reaction despite the continuous heating of the bed. 3.2. Effect of Hydrogen Addition

The plasma power could be varied by either changing the current while fixing the plasma gas composition, or by varying hydrogen concentration while fixing current. The overall effect of increasing power by either mechanism is increasing the plasma jet enthalpy and hence plasma jet temperature. It is expected that hydrogen addition affects the reaction rate by raising

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

417

Fig. 6. Conversion of calcium oxide to carbon monoxide vs. bed temperature at different plasma powers. Carbon to calcium oxide molar ratio was 3:1 and plasma gas flow rate was 40 L/min.

plasma power and enthalpy and not by involvement in the reaction. To verify this point, hydrogen concentration was kept constant at 33%, while power was increased by raising the current to equal the power of 45% hydrogen. A plot of conversion vs. run time for the three cases is shown in Fig. 7. It can be observed that for the same power input the rate of conversion was about the same for different hydrogen concentrations. This implies that hydrogen affects the rate only by raising the plasma power and plasma jet enthalpy. The influence of an instantaneous reaction zone temperature change on the rate of conversion was further investigated by injecting methane into the jet and by lowering the current. This was carried out for a plasma gas composition of 67% hydrogen. In the first run, methane was injected after 11 min into the jet through the spouting port at a flow rate of 4 L/min. In

418

El-Naas, Munz, and Ajersch

Fig, 7. The effect of hydrogen addition on the rate of reaction. Carbon to calcium oxide molar ratio was 3:1 and plasma gas flow rate was 40 L/min.

the second run, the current was lowered from 240 to 200 A at 17min, decreasing the plasma power from 18 to 15 kW. Both methods resulted in lowering the reaction zone temperature and consequently the rate of conversion as shown in Fig. 8 The effect immediately confirmed the fact that the reaction was taking place in this zone. Also, the sensitivity of the rate to the temperature change suggested that the global rate of reaction was controlled by chemical reaction. The jet temperature was lowered from 10,200 to 8600 K in the first case and to 9500 K in the second case. The rate of reaction was lowered from 0.091 mol/min to about the same level (0.032 mol/min) for both cases despite the difference in the predicted plasma jet temperature. This is mainly due to the presence of methane as another source of carbon that is more reactive than graphite. The reactivity of graphite and other sources of carbon was discussed in a previous work.(19)

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

419

Fig. 8. The influence of sudden change of plasma jet temperature on the rate of reaction. Carbon to calcium oxide molar ratio was 3:1 and plasma gas flow rate was 40 L/min.

3.3. Bed Stability The fluid bed reactor, as described earlier, was divided into two zones: the reaction zone at the bottom and the bed zone, where most particles were present in the fluidized state. The reaction zone was created by positioning the plasma torch at the bottom of the reactor. Although this arrangement was effective in transferring plasma enthalpy to the bed materials, it can result in a high-temperature zone where particles may melt and agglomerate. When the reaction zone temperature is sufficiently high, partially reacted particles soften and agglomerate when reaching a temperature near the melting point of calcium carbide. The large agglomerates tended to fall into the bottom of the jet, where they combined and formed a cylindrical mass around the jet. This occurred while the bed approached a temperature well below the reaction temperature based on thermodynamic calculations. Stable

420

EI-Naas, Munz, and Ajersch

fluidization and reaction were achieved, however, for most conditions (except for 67% hydrogen) for periods up to 40 min. The stability of a run was monitored by stable bed temperature and carbon monoxide concentration. The bed materials collected at the end of each stable experiment showed little agglomeration and no melting. The solids seemed to be well mixed and no segregation was observed. It is worth noting at this point that in a continuous plasma process, the reactor will be run in a steady state at a lower temperature. This eliminates the problems associated with melting and agglomeration and results in high conversions.

3.4. Effect of CaO Particle Size Calcium oxide powder had particle sizes ranging from 53 to 1100 um. This range was divided into two portions in order to investigate the effect of particle size on the rate of formation of calcium carbide. The first portion contained particles greater than 53 and less than 425 um and had a mean particle diameter of 150 um. Particles greater than 425 um and less than 1100 um were included in the second portion and had a mean particle size of 600 um. Each portion was reacted with the graphite at a plasma power of 16 kW. A plot for conversion vs. time for the two cases is shown in Fig. 9. Smaller particles exhibited a higher rate of reaction, due to the greater surface area. However, the difference in reaction rate was not significant compared to the difference in particle size (0.048 mol/min for 150 um and 0.039 mol/min for 60 um). For the reaction of spherical particles, the ratio of reaction times needed to achieve a given conversion can be related to the ratio of particle sizes as follows

where R1 and R2 are the radii of particles having the same conversion at different times t1 and t2 respectively. The value of the exponent n may be used to infer the controlling mechanism; it is equal to 2 for diffusion control, 1.5-2 for film diffusion control, and 1 for chemical reaction control.(20) For the calcium oxide particles used in this study, the value of n is less than unity as shown in Fig. 9. In a chemical reaction control mechanism, particle size affects the rate through surface area. Clearly, the smaller oxide particles have larger surface area for reaction. However, because of the high porosity of calcium oxide, the effect of lowering particle size on surface area is minor and, consequently, the effect on the rate is small.

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

421

Fig. 9. The effect of calcium oxide particle size on the reaction rate. Hydrogen concentration was 45% and plasma power was 16kW. Carbon to calcium oxide molar ratio was 3:1 and plasma gas flow rate was 40 L/min.

3.5. Product Analysis 3.5.7. Product Identification The bed materials were examined visually after every experimental run and samples were taken for analysis. In all experiments, except for pure argon, the solid materials were black in color and no white solids were observed. For the lowest power (0% H2), however, the bed contained some grayish, white solids. This was mainly due to the low conversion for this plasma condition. Samples of the solid product were reacted with water to determine the calcium carbide content of the solid product (conversion). Details of this procedure and other analytical techniques can be found elsewhere.(21) X-ray diffraction analyses of the product were compared with those of a standard calcium carbide sample. The samples were pulverized

422

El-Naas, Munz, and Ajersch

to very fine powder to facilitate the analysis. The grinding and the exposure of the powder to moisture in the atmosphere during the analysis resulted in converting the carbide into calcium hydroxide. Thus, X-ray diffraction for both the solid product and the standard sample showed only calcium oxide and calcium hydroxide. 3.5.2. Microscopic Analysis Microscopic analyses of the solid reactants and products were carried out to examine the product morphology. Cross sections of particles were examined and elemental spot analysis for calcium, carbon, and oxygen was determined. Also, representative particles of unreacted and partially reacted graphite and calcium oxide particles were mapped to determine the relative distribution of calcium, carbon, and oxygen. It should be noted here that for every product sample about one hundred particles were scanned, but only a few were chosen for mapping. All particles examined had similar morphology and indicated that all particles in a specific experiment reacted to about the same extent. Mapping of the different elements proved to be an effective tool in determining the distribution of calcium, carbon, and oxygen across the unreacted and the partially reacted particles. The relative intensity of each element detected is indicated by the brightness of the image as the particle is scanned. Due to the presence of carbon and oxygen in the mounting resin (polymer), the backgrounds also show both of these elements. Maps for a partially reacted graphite particle (Fig. 10) indicate that the particle is subjected to a topochemical, shrinking core-type reaction. In the carbon map (Fig. 10A), the rim is darker than the rest of the particle, indicating less carbon is present at the rim of the particle. On the other hand, in the calcium map (Fig. 10B) the rim is brighter, indicating that the outer edge of the particle has more calcium. The lack of carbon in the rim indicates that it is not calcium carbide. Spot analysis at different points in the rim showed that it contained oxygen and calcium in an atomic ratio of about 2 to 1. This implies that the outer edge of the particle has hydrated to calcium hydroxide, which is formed as a result of the following reaction

Thus, carbon in the calcium carbide layer around the particle was removed as acetylene. This was also observed for partially reacted calcium oxide particles as shown in Fig. 11. Calcium carbide, therefore, had formed at the surface of both reactants. The microscopic analysis indicated that the formation of calcium carbide preceded according to the shrinking core model. The formation of

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

423

Fig. 10A. Carbon map for a partially reacted graphite particle.

calcium carbide on the graphite particles occurs through the reaction of calcium vapor and carbon according to Eq. (3). This reaction is also believed to be responsible for the formation of calcium carbide on the calcium oxide particles by the reaction of calcium vapor and the carbon diffused into the oxide particles. Calcium carbide can also form on the surface of the oxide particles by the decomposition of the interstate compound (CaC3O) proposed by Muller.(17) It is difficult to state, however, which mechanism is more important for the formation of calcium carbide on the calcium oxide particles.

3.6. Product Decomposition Previous studies of the solid-phase formation of calcium carbide showed that calcium carbide decomposed according to the reverse of reaction (3), and that the rate was zero order and increased with temperature.(13) The decomposition occurred even at atmospheric pressure as reported by Mu and Hard(3) and was slowed down by lowering the partial pressure of carbon

424

El-Naas, Munz, and Ajersch

Fig. 10B. Calcium map for a partially reacted graphite particle.

monoxide. Brookes et al.(22) stated that the decomposition led to the formation of carbon and calcium vapor. The latter reacted with carbon monoxide (2), deposited calcium oxide, and carbon on cooler surfaces as a gray dust. The researchers also observed a carbide-free carbon ring on the outside of the product layer. In the present study, it was rather difficult to confirm the decomposition of calcium carbide. Although gray dust was observed on the walls of the upper part of the disengaging section and the filter, no carbon layers were observed on the partially reacted particles. Microscopic analysis, as discussed in the preceding section, would clearly detect the presence of any carbon around the particles, but no carbon was found around any partially reacted graphite or calcium oxide particles. It is possible, nevertheless, that decomposition has occurred and that the carbon ring around the particles was stripped off by attrition due to particle movement in the bed. The extent of the decomposition, however, is expected to be minor due to the low carbon monoxide partial pressure, which tends to hamper the decomposition of calcium carbide.

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

425

Fig. 11 A. Carbon map for a partially reacted calcium oxide particle.

4. CONCLUSIONS A new plasma spout-fluid bed process for the synthesis of calcium carbide was investigated. Calcium carbide can be produced in a granular form by the solid-phase reaction of carbon and calcium oxide. The spout-fluid bed reactor was found to have two different zones: a high-temperature plasma reaction zone and a well-mixed isothermal bed zone. The experimental results showed that the reaction took place in the reaction zone and that conversion to calcium carbide increased linearly with reaction time. The rate of conversion to carbon monoxide was constant with time for all conditions and increased exponentially with increasing plasma jet temperature, indicating that chemical reaction was the controlling mechanism. Microscopic analysis of the solid product showed that calcium carbide was formed around both graphite and calcium oxide particles, indicating

426

El-Naas, Munz, and Ajersch

Fig. 11B. Calcium map for a partially reacted calcium oxide particle.

that the carbon-calcium vapor reaction took place at the surface of both free carbon and the carbon diffused into the oxide. Limitations with the present experimental apparatus made it difficult to achieve conversions beyond 30% due to the melting and agglomeration of particles in the plasma jet zone. Higher conversions, however, can be expected in a continuous plasma process where the reactor will be run in a steady state and at a lower temperature thus eliminating the problems of melting.

ACKNOWLEDGMENTS The financial support of the Libyan Secretariat of Scientific Research in the form of a scholarship to M. H. El-Naas, the Natural Sciences and Engineering Research Council of Canada, and FCAR, Quebec, are gratefully acknowledged.

Solid-Phase Synthesis of Calcium Carbide in a Plasma Reactor

427

REFERENCES 1. K. Baba and N. Shohata, "Manufacture of calcium carbide micropowders by highfrequency plasma method," Japanese patent, No. 63112409 (1988). 2. S. Eriksson, "Calcium carbide from powdered limestone or lime," Belgian patent, No. 897179 (1983). 3. J. J. Mu and R. A. Hard, "A rotary kiln process for making calcium carbide," Ind. Eng. Chem. Res. 26, 2063-2069 (1987). 4. C. S. Kim, R. Baddour, J. Howard, and H. Meissner, "CaC2 production from coal or hydrocarbons in a rotating-arc reactor," Ind. Eng. Chem. Process Des. Dev. 18, No. 2, 323-328, 1979. 5. W. M. Goldberger and J. H. Oxley, "Quenching the plasma reaction by means of the fluidized bed," A.I.Ch.E.J. 9, 778-782 (1963). 6. J. Jurewicz, P. Proulx, and M. I. Boulos, "The plasma spouted bed reactor," Proceedings of ISPC-7, Eindhoven, July 1985, Vol. I, pp. 243-248. 7. C. R. Wierenga and T. J. Morin, "Characterization of a plasma fluidized bed reactor," A.I.Ch.E. J. 35, 1555-1560 (1989). 8. G. Flamant, "Hydrodynamics and heat transfer in a plasma spouted bed reactor," Plasma Chem. Plasma Process. 10, 71-85 (1990). 9. R. J. Munz and O. S. Mersereau, "A plasma spout-fluid bed for the recovery of vanadium from vanadium ore," Chem. Eng. Sci. 45, 2489-2495 (1990). 10. C. W. Zhu, G. Y. Zhao, and V. Hlavacek, "A dc plasma-fluidized bed reactor for the production of calcium carbide," J. Mater. Sci. 30, 2412-2419 (1995). 11. M. H. EL-Naas, R. J. Munz, and F. Ajersch, "Production of calcium carbide in a plasmajet fluid bed reactor," Proceedings of ISPC-12, August 1995, Minneapolis, Minnesota, U.S.A., Vol. II, pp. 613-618. 12. N. Kameyama, Electrochemistry: Theory and Application, Vol. III-2, 1956, pp. 134-141. 13. H. Tagawa and H. Sugawara, "The kinetics of the formation of calcium carbide in solidsolid reaction," Bull. Chem. Soc. Jpn. 35, 1276-1279 (1962). 14. T. Mukaibo and Y. Yamanka, J. Chem. Soc. Jpn., Ind. Chem. Sec. 58, 1955. [Cited in Tagawa and Sugawara, 1962.] 15. P. Hellmold and W. Gordziel, "Investigation of the particular reactions of CaC2. 2: Investigation of the formation of CaC2 from calcium oxide and carbon," Chem. Technol. 35, 297300 (1983). 16. M. B. Muller, "Structure, properties and reactions of CaO in burnt lime, Part I: Design of a new model of burnt lime," Scand. J. Metall. 19, 64-70 (1990). 17. M. B. Muller, "Structure, properties and reactions of CaO in burnt lime, Part III: Composite reactions of CaO and C in solid and liquid state," Scand. J. Metall. 19, 210-217 (1990). 18. M. H. EL-Naas, R. J. Munz, and F. Ajersch, "Modelling of a plasma reactor for the synthesis of calcium carbide," Can. Metall. Quart, (in press). 19. M. H. EL-Naas, F. Ajersch, and R. J. Munz, "Effect of carbon reactivity on plasma synthesis of calcium carbide," Proceedings of the 35th Annual Conference of Metallurgists, Montreal, Quebec, Canada, August, 1996. 20. C. Y. Wen, "Non-catalytic heterogeneous solid fluid reaction models", Ind. Eng. Chem. 60, No. 9, 34-52(1968). 21. M. H. EL-Naas, "Synthesis of calcium carbide in a plasma spout fluid bed," Ph.D. Thesis, McGill University, 1996. 22. C. Brookes, C. E. Gall, and R. R. Hudgins, "A model for the formation of calcium carbide in solid pellets," Can. J. Chem. Eng. 53, 527-535 (1975).