2017 Volume 58 Issue 6 Pages 914-920
Studies pertaining to kinetic aspects of the direct electrochemical reduction process of titanium preparation from TiO2 have not been substantial for determining the conversion time precisely. The solid state reduction process does not permit measuring either the rate of formation of product or the rate of depletion of the feed stock. In this work an attempt has been made to study the applicability of variation of CO2 (the product gas) in the vent gases of the electrolytic cell to approximate the bulk process kinetics of the process. From the electrolysis current trend, it might be possible to estimate CO2 profile during the electrolysis time so as to utilize the same for inferring optimal time of the electrochemical reduction process.
Extraction metallurgy of titanium has been very complicated and cumbersome mainly because of its high melting point (≈ 1663℃) and high chemical reactivity. Especially preparation of titanium metal directly from its widely occurring oxide (TiO2) by employing all the conventional reduction techniques proved futile because of high stability of the oxide and affinity of the metal towards oxygen. The Kroll process, in which titanium tetrachloride is reduced by magnesium metal, has been the predominant method of titanium extraction. However, since the emergence of the electrochemical reduction process popularly known as FFC Cambridge process1–3) wherein the metal oxide (TiO2) is converted into metal through cathodic action, expectations have been high for realizing an alternate cost and energy effective method of preparing titanium metal of required purity. Hence the process has been widely pursued for greater understanding of the process and for scale up to produce the metal in larger batches.
In the electrochemical reduction of TiO2, the overall conversion of the oxide to metal is described simply by removal of oxygen in ionic form, from the oxide which passes through the conducting electrolyte (molten calcium chloride at about 950℃) to a graphite anode to form CO/CO2 as per the following:
| \[ {\rm TiO}_{2} + 4 {\rm e} = {\rm Ti} + 2 {\rm O}^{2-} ({\rm at}\ {\rm cathode}) \] | (i) |
| \[ {\rm C} + 2{\rm O}^{2-} - 4{\rm e} = {\rm CO}_{2} ({\rm at}\ {\rm anode}) \] | (ii) |
However, the actual mechanism of conversion of TiO2 to titanium metal appears to be more complex with formation of several sub-oxides of titanium and calcium titanate as intermediate reaction species4,5). It has been convincingly explained that the calcium metal deposited (due to the electrolysis of CaO dissolved in the CaCl2 electrolyte), on the cathode (TiO2) can reduce the oxide into metal through various intermediate titanium suboxides. Lowering of activity of CaO, (generated due to calciothermic reduction of oxide) as it dissolves in molten CaCl2 circumvents the thermodynamic restrictions prevailing in the conventional calico-thermic reduction process in obtaining titanium sufficiently free from oxygen. Alternatively, the reduction process is also explained to be taking place through oxygen ion (generated due to applied DC voltage) transportation from the oxide through the electrolyte to form O/CO/CO2 at the graphite anode, which is let out through the vent gases of the cell. Under experimental conditions, carbon dioxide is however, the predominant product gas of the overall electrochemical reduction process. Current exercise, aims at understanding the bulk process kinetics in terms of CO2 release in the electrochemical reduction process. Based on experimental measurements and theoretical predictions on CO2 content in the vent gases of cell, it is possible to understand the kinetic aspects of the conversion process in various stages of the experiment. Before discussing the results of these efforts, a review of current understanding on various kinetic aspects of the process is brought out in the following.
Several investigators have reported works related to the mechanism of the reduction process and tried to relate the influence of various experimental conditions such as type of pre form, applied voltage, current densities, temperature etc. on the mechanism and pathway of metal formation6–16). According to Schwandt and Fray9) the reduction of TiO2 to Ti take place through formation and decomposition of calcium titanate and the conversion process is associated with substantial changes in the microstructure. Similar findings were also reported by Jayashree Mohanthy et al10). Lebdev et al11,12) discussed about possible mechanism and kinetics of processes occurring at TiO2 cathode and graphite anode independently in CaCl2-CaO melt. According to which the CaO concentration in the CaCl2 melt does play a role on the kinetic aspects of the process. Studies by Alexander et al13), indicated that the conversion of TiO2 to titanium metal takes place through formation of several sub-oxides of titanium and calcium titanate though formed, can decompose into titanium sub oxide and calcium oxide. It is also said that the formation of calcium titanate can be minimized by the use of highly sintered/less porous TiO2 pre-forms. Dring et al14) dwelled into thermodynamic aspects of Ca-Ti-O-C-Cl system to propose predominance area diagrams for electrochemical reduction of titanium dioxide in molten calcium chloride. Interesting results/suggestions reported by this work include favourable conditions for producing low oxygen titanium metal. According to these studies higher pO2- values in the melt are favourable for rapid reduction in the initial stages and towards end of the reduction, relatively a lower temperature and reduced oxide activity will support for high purity titanium formation. In an investigation into the kinetics of the electrochemical reduction process reported by Songli Liu15), it is stated that while the reduction process is associated with formation of sub-oxides of titanium and calcium titanate, hydrolysis and subsidiary reactions occurring at anode may result in reduced current efficiency of the process. According to these works, in the first stage, TiO2 is reduced to Ti3O5 and Ti2O3 and form CaTiO3. In the second stage CaTiO3 and Ti2O3 are converted to TiO and finally TiO is reduced to Ti. A few other works16,17) report on influence of various additives to TiO2 pre-cursers on the rate of the electrochemical reduction process. Bhagat et al18) conducted in situ synchrotron diffraction to propose electrochemical pathway of TiO2 reduction and discussed the kinetic aspects of oxygen ion generation at cathode and formation of CaO and other compounds. Hualin Chen et al19) studied influence of graphite anode area on the oxidation kinetics of O2- at the anode and suggested that a low anode current density is beneficial for reducing energy consumption. An elaborate review by Mohandas20) brings out summary of various developments that have been taken place over the years on understanding of the electrochemical conversion of oxides to metal in general.
From the above description, it is pertinent to state that simultaneous occurrence of various cathodic and anodic cell reactions and a set of chemical reactions associated with the species TiO2, Ti, Ca, C CaO and CO/CO2 would determine the bulk kinetics of the electrochemical reduction process of TiO2 to titanium metal. As already presented earlier21), the bulk kinetics of the reduction process can be determined from the rate constants of all the associated kinetic steps such as (i) oxygen ion generation at cathode and its transportation to cathode/electrolyte interface, (ii) transportation of oxygen ion through electrolyte to anode/electrolyte interface and (iii) evolution of O/CO/CO2 at the anode. These kinetic steps however, additionally might depend on aspects such as dissolution kinetics of CaO in CaCl2, solubility of calcium in CaCl2, viscosity related surface flows, parameters related to gas evolution at the anode, solid state phase/crystalline changes in the cathode feed stock etc. In the earlier work21) an attempt has also been made to determine time of conversion of the oxide granule under an assumption that oxygen ion mobility from the bulk oxide to oxide/electrolyte interface as the predominant and rate controlling kinetic step.
As reported earlier22), the overall reaction of conversion of TiO2 to Ti can be expressed in terms of a set of thermodynamically possible chemical reactions involving sub-oxides of titanium as given below:
| \[ \begin{split} & 8{\rm TiO}_{2} + {\rm C} = 2{\rm Ti}_{4} {\rm O}_{7} + {\rm CO}_{2}\\ & \qquad \Delta {\rm G}^{\circ}_{1200{\rm K}} = - 863178\,J/mol \end{split} \] | (iii) |
| \[ \begin{split} & 16{\rm Ti}_{4} {\rm O}_{7} + {\rm C} + 2{\rm Ti} = 22{\rm Ti}_{3}{\rm O}_{5} + {\rm CO}_{2}\\ & \qquad \Delta {\rm G}^{\circ}_{1200{\rm K}} = - 1124570\,J/mol \end{split} \] | (iv) |
| \[ \begin{split} & 3{\rm Ti}_{2} {\rm O}_{3} + {\rm C} + {\rm Ti} = 7{\rm TiO} + {\rm CO}_{2}\\ & \qquad \Delta {\rm G}^{\circ}_{1200{\rm K}} = - 843871\,J/mol \end{split} \] | (v) |
| \[ \begin{split} & 2{\rm TiO} + {\rm C} + {\rm Ti} = 3{\rm Ti} + {\rm CO}_{2}\\ & \qquad \Delta {\rm G}^{\circ}_{1200{\rm K}} = - 525832\,J/mol \end{split} \] | (vi) |
Chemical reactions associated with calcium titanate might not be considered in representing the overall conversion process as it can exist only as reaction intermediate and is also expected to dissociate into titanium sub-oxide and CaO.
Thus the overall electrochemical reduction process can be represented as
| \[ \begin{split} & {\rm TiO}_{2({\rm in\ contact\ with\ CaCl2})} + {\rm C}_{({\rm graphite\ anode\ in\ contact\ with\ CaCl2})} \\ & \quad = {\rm Ti}_{({\rm left\ behind\ in\ the\ starting\ material})} + {\rm CO}_{2({\rm anode\ gas})} \end{split} \] |
Since it is practically difficult to monitor the chemical changes taking place in the oxide during the process so as to correlate the experimental conditions with bulk process kinetics, study of CO2 generation during the course of reduction process shall be useful in understanding the kinetic aspects of the overall conversion process. In this work an attempt has been made to understand kinetic aspects of the process with experimentally measured values of CO2 content in the vent gases of the electrolytic cell during the course of reduction process. Measurements of CO2 in the out gasses could be made at selected time intervals of the process in two different scales of operation viz 500–1000 g cell and 5–8 kg cell.
Experimental work on electrochemical reduction of titanium dioxide to titanium metal has been pursued on different scales of operation so as to develop improved understanding of the process. After establishing the feasibility of the process on 500–1000 g scale, scaled up experimentation on 5–8 kg per batch is continuing. Detailed description of experimental procedures pertaining to establishment of feasibility of the process from gram scale to 1000 gram scale is available elsewhere23–25). In all the experimental works Merck make high purity titanium dioxide is used. Experimental set up employed for both the scales of operation are similar and comprise a cathode basket made of AISI 304 perforated sheet in which sintered TiO2 granules are loaded. Electrolytic grade graphite plates (rectangular/curved shaped) are used as anode. The electrode lead rods are used both for holding the cathode basket and graphite anode plates in specified positions in the cell as well as for connecting to DC power source. Typical experimental procedure of 5–8 kg/batch experimentation includes preparation of titanium oxide granules in the size range of 8–12 mm which are sintered at 1000℃ for 24 hours before taken into the cathodic basket which takes the dimensions of 250 × 450 × 100 mm. As shown in Fig. 1 (a), graphite plates of 250 × 400 × 15 mm thick placed on either side of the cathode basket form anode and the cathode- anode assembly is placed inside a stainless steel reactor (450 mm ID × 1300 mm height × 12 mm thick). with the help of stainless steel lead rods. These are connected to a DC power source for applying DC voltage during the electrolysis. A 40 kW three zone electrical resistance furnace is employed to provide heating requirements to heat the electrolyte and electrodes assembly along with stainless steel reactor to a temperature of 970℃ and maintain the temperature all through the experimental period. Initially about 500 kg of laboratory grade anhydrous calcium chloride is taken into the reactor and melted under argon gas atmosphere and then cooled to room temperature. At room temperature, after placing the electrodes assembly, the reactor is closed with a stainless steel lid that has necessary openings and nozzles for (i) taking out the electrode lead rods, (ii) for evacuation of the reactor system, (iii) argon gas supply and (iv) for the exit of vent gases. Pressure tightness of the assembly is achieved with the use of appropriate high temperature gaskets between the reactor flange and the lid flange (Fig. 1 (b)). The annular gap between the electrode lead rods and nozzles are filled with high temperature seals so as to ensure pressure tightness even at the higher temperatures experienced during the experimentation. The system is evacuated with the help of an oil rotary vacuum pump to 10–15 mm of Hg and filled back with argon gas to about 2 psig. The evacuation and filling back with argon gas is repeated to ensure withdrawal of residual air (up to a lowest limit) that initially present in the system.
(a) Schematic of experimental set up (5–8 kg scale), (b) Schematic of nozzles and vent line arrangement on the cell lid (5–8 kg scale).
Before each experiment, samples of calcium chloride are analyzed for CaO content by titration with dilute acid and the data is recorded. As a general practice voltage drops associated with the given cell configuration are experimentally determined under ‘no load’ conditions (basket without TiO2 granules) prior to the actual reduction experiment. Accordingly the experimental procedure pertaining to DC voltage application is evolved. This helps in understanding the applicable voltage window (avoiding CaCl2 decomposition and chlorine generation) for the given experiment.
The experiment begins with the heating of the set up to a temperature of 970℃ as measured at the reactor outer wall, under argon gas cover. After ensuring an electrolyte temperature of 930℃, application of DC voltage is initiated and increased stepwise from 2.0 to 4.5 V (Table 1). During the entire electrolysis period argon gas purge into the cell system is continued at a rate of 100–200 cc/min. The exit gases let out during the experiment are led through a copper/rubber tube to a CO2 detection and monitoring system. A carbon dioxide transmitter cum infrared sensor, ADT-D3 1164, Intertek, German make of (i) 0–5000 and (ii) 0–50000 ppm range were employed for measuring CO2 content in the exit gases at selected points of time during the experimentation (The higher range detector was used in the 5–8 kg experimentation whereas the lower range one is used in the experiments on 5000–1000 g scale set up). Otherwise the vent gases pass through a set of baryta (barium hydroxide solution) bubblers which helps in detecting CO2 in the gases qualitatively and enables monitoring of the same as formation of white barium carbonate precipitate indicates the presence of CO2 in the gases. Table 2 presents typical operating conditions and important process parameters recorded during the course of electrolysis time. Figure 2 shows a photograph of the cathode basket loaded with the oxide granules employed in the experimentation.
| DC voltage (V) | Time (hrs) |
|---|---|
| 2.0 | 2–4 |
| 2.6 | 4–6 |
| 2.8 | 4–6 |
| 3.2 | 4–6 |
| 3.8 | 6–10 |
| 4.5 | Remaining period of experiment |
| S. No. | Parameter | Value |
|---|---|---|
| 1 | Weight of TiO2 | 10–14 kg |
| 2 | Level of molten electrolyte | ≈ 650 mm |
| 3. | DC voltage | 2.5–5.2 V |
| 4 | Electrolysis temperature | 960℃ |
| 5. | Inert gas pressure in the cell | 1.1 atm |
| 6. | Free (gas) volume in the cell | 0.1 m3 |
A photograph of cathode basket loaded with sintered TiO2 granules.
After the electrolysis experiment, the entire assembly is cooled to room temperature under argon gas cover. Then the cathode basket is carefully taken out, examined and after visual inspection of the same, it is taken up for further handling. The granules are taken out from the cathode basket and are subjected to a series of washings with water, dilute acetic acid and dilute HCl solutions for completely removing the adhered electrolyte and other species. The anode after visual inspection is separately taken away and water washed. In every experiment new graphite plates are used as anode.
As per the understanding of the mechanism of conversion of TiO2 to titanium metal, the overall reduction process is described (as mentioned before) by eqs. (i) and (ii) and practically CO2 can be considered as the predominant products gas.
The essential requirement for CO2 evolution at anode is oxygen ion transportation from cathode through molten calcium chloride which is governed by applied voltage and transport properties of the associated species. Electrolysis current in a way represents the rate of transportation of oxygen ions to anode and hence the DC current profile, recorded during the electrolysis time can be applied to infer the rate of CO2 formation at the anode. However, CO2 formation cannot be in proportion to the magnitude of electrolysis current as there can be many electron transfer subsidiary reactions that are not associated with the conversion of the oxide to metal/CO2 formation. Still, the electrolysis current profile during the reaction time is expected to describe the trend of CO2 formation. In a simple and preliminary approach, the electrolysis current variation is correlated to rate of formation of CO2 and attempts have been made to predict CO2 variation in the vent gases and compare with the experimental measurements.
It is generally observed that during the electrochemical reduction process, as the voltage is applied and increased say from 2.0 to 4.5 volts, DC current raises from a lower value to a higher value which continues predominantly over a period of time and then the current gradually decreases. As can be seen in Fig. 3 the DC current – time profile broadly consists of three phases viz initial phase during which current increases with voltage and in the second phase higher current persists at a constant applied higher voltage and this is followed by final phase during which current decreases to a lower value even at the highest applied voltage. This kind of current variation can be understood to represent the pattern of oxygen ion transportation from the cathode to graphite anode through the electrolyte resulting in the formation of CO2 which is detected in the vent gases. In Fig 4 (a) and (b), measured values of CO2 content in the exit gases at various selected points of electrolysis time are presented. In the first part of the experiment after reaching the applied higher cell voltage, CO2 content above 5000 ppm in the case of 500–1000 g experiment and above 50000 ppm in the case of 5–8 kg experiment were observed. In the second half of the experiment, the CO2 content is seen gradually reducing from 5000 ppm to as low as 300 ppm in the case of small scale experiment and from 50000 ppm to about 500 ppm in the case of 5–8 kg experiment. In both the scales of operation the last/terminal phase of electrolysis is denoted by a lower and constant CO2 value. It is interesting to note that the variation of CO2 content in the exit gases, with time during the experiment follows the trend similar to the DC current profile as can be seen in Fig. 5. At a given time in the midst of the experiment, it is seen that the CO2 content is increasing with the applied voltage as can be seen from Table 3. This indicates that the overall conversion rate could be increasing with increased voltage within the applicable voltage range.
Variation of electrolysis current with time (5–8 kg scale).
(a): Measured CO2 content in the vent gases of 1 kg experiment, (b): Measured CO2 content in the vent gases of 5–8 kg experiment.
Variation of Electrolysis current and CO2 in vent gases with time (5–8 kg batch).
| Applied DC voltage (V) | Measured CO2 content in the exit gases (ppm) |
|---|---|
| 3.0 | 2630 |
| 3.5 | 4130 |
| 3.8 | 5120 |
| 4.0 | 5310 |
| 4.2 | 6260 |
The observations made on the conditions of electrodes after the experiment are highly informative. In Fig. 6 (a) and (b) photographs of graphite anode plates before and after the experiment are shown. The area of the anode immersed in the electrolyte is seen to be significantly corroded. The regions of the anode that are at the gas-electrolyte interface are generally seen to be highly corroded. The gas- electrolyte interface at the anode is apparently experiencing erosion due to intense gas evolution during the electrolysis time as was also observed and reported by Srimaha Vishnu et al26).
Photographs of graphite anode (a) before electrolysis (b) after the use in electrolysis.
As already brought out, the similar co-relative trends observed with the CO2 – time profile and DC current – time profile, could be used for estimating CO2 in the vent gases by simpler linear best fit methods. From the experimental data on DC current variation with time collected from different experiments, the average rate of CO2 generation could be determined. In Fig. 7, predicted CO2 content in the exit gases during the electrolysis period is compared with experimentally measured values. The predictions in general are found to be reasonably well though at times, the predicted values differ from the measured values significantly. The significant difference in predicted value and actual value is understandable because often the electrolysis current is masked by current spent in the side reactions such as calcium titanate formation as discussed in the literature27,28) and increased conductivity of the melt due to dissolved calcium metal, possible carbon fines floating on the electrolyte surface15,29) etc. Similarly there can be reactions that consume or generate CO2 gas as Ca/CaO dissolved in the melt may react with CO2 dissolved in the electrolyte to form calcium carbonate or decomposition of calcium titanate to form CO2. Consequently more efforts are required for fine tuning the predictions. Then it can be applied for determining the time required for the electrolysis through estimation of time at which CO2 in the exit gases is minimum represent completion of the reduction process. This is because the lowest values of CO2 content in the vent gases is expected to describe the completion of the reduction process.
Predicted CO2 profile – comparison with experimental profile (5–8 kg scale).
Some kinetic aspects of electrochemical reduction of TiO2 to titanium metal have been discussed. Based on the experimental work carried out on two different scales of operation, it could be seen that variation of electrolysis current during the course of reduction process has a correlation with CO2 content in the vent gases of the electrolytic system. It has been noticed that variation of CO2 content in the exit gases follows a trend similar to DC current – time profile. Understanding of CO2-electroysis time profile shall be useful for optimizing the process cycle time.
Authors wish to express their deep sense of gratitude to Dr. S.V. Kamat, Outstanding Scientist & Director, DMRL for the constant support and encouragement extended to this developmental activity and also for according permission to publish this technical paper. Authors also wish to acknowledge DRDO for providing funds to this R& D work on development of the electrochemical reduction process for titanium extraction.
