The principal problems of aluminum electrowinning: an update

The principal problems of aluminum electrowinning: an update

S. Prasad

Departamento de Engenharia Química, CCT, Universidade Federal da Paraíba,

CEP 58109-970, Phone: [phone]. Fax: [phone]

Campina Grande - Paraíba, Brazil.

E-mail: [e-mail]

(Received: April 6, 1999 ; Accepted: April 13, 2000)

Abstract - The aluminum electroreduction process is still faced with several problems such as high energy consumption, stability of cathode lining, development of nonconsumable anodes, current efficiency, alumina feeding and environmental problems with hazardous spent potliners and fluoride emissions. A short review of recent literature related to these problems has been provided. A critical analysis of these problems, based on practical experience with a modern plant, is also given.

Keywords: Aluminum electrowinning, Hall-Heroult process, electroreduction.

INTRODUCTION

Aluminum is now the most important nonferrous metal product. It is the third-most abundant element in the earth's crust, accounting for 7.3% of its total mass. Owing to its reactivity, aluminum is never found in nature in its elementary state, but rather in its oxidized form, most commonly in minerals such as aluminates and silicates. Other important naturally occurring compounds include the oxide, bauxite, and the fluoride, cryolite. The commercial production of metallic aluminum started in 1889 with electrolysis of cryolite-alumina melt by the Hall-Heroult process. Apart from the technological refinements made since, the electrolytic aluminum process remains basically the same as it was 110 years ago (Grjotheim et al., 1982). Over the years, numerous alternatives to the Hall-Heroult process have been considered in an attempt to reduce the cost of the metal. Carbothermal reduction processes have technological problems associated with their high temperature requirement which make them commercially inviability (Grjotheim and Welch, 1988). The Alcoa Smelting Process, based on electrolysis of AlCl₃ in molten chloride mixtures, seems to be a promising alternative, but the method appears to include several difficult technical problems which are costly to solve satisfactorily (Grjotheim and Welch, 1988).

A survey of the literature reveals that no new process is going to replace the classical Hall-Heroult process in the near future, as the latter has the advantage of a mature depth of understanding which enables the introduction of steady improvements. But with the changing economic conditions, development of a low-cost cell is of vital importance to industry. A potential improvement of the process that has received considerable attention in recent years is in relation to reduction of the production cost of the metal by intensifying research in areas such as increase in current efficiency, development of nonconsumable anodes and more resistant potlining, reduction in energy losses, compensation of electromagnetic effects and optimization of the bath composition. A brief description of recent developments in these areas is presented in this paper. This work is in continuation to the research on electrodeposition of metals (Prasad, 1996, 1998, 1999; Prasad and Marinho, 2000) and chemistry of bauxite (Prasad and Oliveira, 1997) reported from these laboratories. A substantial part of the work was realized by the author on the premises of an aluminum production company (Alumar).

THE BASICS OF THE HALL-HEROULT PROCESS

Electrolysis is performed in the pot (cell) of a steel box lined with a refractory, thermal insulator and carbon. In fact, the base of the pot is lined with prebaked carbon blocks and the sides with partially graphitized anthracite in coal-tar pitch. Carbon anodes are manufactured from a carbon source such as coke and a pitch binder. Two types of anodes are in use: prebaked and selfbaking (known as the Soderberg type) carbon blocks. The latter type of anodes were common in the 1940s and 1950s and are now being replaced by more efficient prebaked anodes. While the electrolyte is essentially molten cryolite, certain additions are made and a typical electrolytic medium also contains excess AlF₃ (10-12%) and CaF₂ (4-6%) along with a regular addition of alumina. The additives increase the conductivity of the medium and lower the melting point of the cryolite from 1011 to 920-970 ⁰C, resulting in a decrease in energy consumption. But the additives cause a decrease in solubility of alumina (from 15 wt% to about 6 wt%) and this limits their total concentration. Alumina is added to the cell periodically because during electrolysis its concentration drops, and if it is allowed to fall below about 2%, the electrolytic cell undergoes a sudden and major operational failure known as an "anode effect" (Fig. 1). The cell voltage decreases rapidly from -4.5 V to a value in the range of -40 to -60 V. Overfeeding the cell with alumina causes the formation of sludge under molten aluminum pad thus decreasing electrical conductivity and resulting in a "sick pot". The optimum current density is around 1 A cm²- with a total cell current of 150-300 kA and a cell voltage of -4.0 to -4.5 V. A typical cell house will contain about 200 cells, arranged in series on two lines. All cell houses have a strong magnetic field due to the large currents used. The field can produce turbulence at the aluminum/electrolyte interface if the bus bars are not arranged in a compensating manner. Aluminum is probably present as a mixture of several oxy-fluoride species in the electrolyte. Therefore, the exact chemistry of the system is not known. However, the cathode reaction is the reduction of the Al(III) species to the metal, forming a pool of liquid aluminum at the bottom of the pot and acting as a cathode. The anode reaction should be the oxidation of the oxide ion to oxygen, but it is difficult to find an anode material which is inert under the aggressive conditions of electrolysis. Hence electrolysis has always been conducted with a consumable carbon anode so that the overall cell reaction is

(1)

Figure 1:
Calculated cell voltage as a function of alumina concentration at two different anode-cathode gaps (curve 1: 4.0 cm, curve 2: 4.8 cm) for a cell such as that in Table 1. (Desired concentration range of alumina 2-5%. At Concentrations lower than this, there is the risk of the anode effect and at concentrations higher than this, there is the risk of sludge formation.)

ANALYSIS OF THE PROBLEMS

Energy Consumption

Great strides have already been made in the reduction in electrical energy consumption for the production of aluminum, as shown in Fig. 2. The specific energy consumption has decreased from about 50 kWh/Kg to about 14 kWh/Kg. Improvements to date have been largely due to an increase in pot size and thus increased thermal efficiency as well as a general tightening of operating conditions. The ultimate limit for a self-sustaining pot operating at 975^oC is 6.37 kWh/Kg Al, based on the enthalpy change for the overall reaction (Eq. 1).

Energy consumption of the aluminum electrolyser by the plant under observation was also found to be about 14.0 kWh/Kg Al. Voltage distribution observed in a typical Hall Heroult cell bearing prebaked anodes is shown in Table 1. The reversible potential calculated from thermodynamic data represents only a small fraction of the observed cell voltage. If the energy required to maintain the cell at 960 ⁰C is taken into account, the energy efficiency of the process is only 33%. The iR drops in the cell are large. Those in the electrodes arise because of their size and the relatively low conductivity of carbon. The substantial electrolyte iR drop is due to the need for a large interelectrode gap. The voltage drop throughout the bath includes about 0.15 to 0.25 V due to the gas bubbles formed at the anode. Carbon dust and/or undissolved alumina suspended in the bath also lower the electrical conductivity of the bath. The anode seems to be the main source of this voltage drop. Some research has been carried out in China to decrease the anodic overvoltage by doping the anode with lithium salt (Liu and Xiao, 1987; Liu et al., 1989, 1993, 1995). Researchers claim that several plants in China are now using lithium-doped anodes. The formation of carbon dust appears to be the result of the nonhomogeneous structure of the anode. The anodes are made of relatively low reactivity coke particles agglomerated in a more reactive pitch binder (Grjotheim et al., 1982). Research on a more appropriate binder, which may reduce the problem of uneven consumption of the anode and formation of carbon dust, should be carried out. The problem of undissolved suspended alumina particles can only be solved by more controlled and regular additions alumina (Kvande et al., 1994).

Current Efficiency

The amount of aluminum predicted by Faraday's law is never obtained. It is reported that the best-equipped smelters have hit a ceiling between 95-96% CE (Forberg, 1996). The principal loss mechanism is by aluminum reacting with NaF in the bath to form sodium that dissolves in the bath.

(2)

(3)

(4)

Therefore, the dissolved sodium must diffuse in the metal pad to decrease reactions (3) and (4). This explains the higher current efficiency in the cells with good magnetic compensation. Sodium dissolved in the electrolyte may also decrease the efficiency by imparting electronic conductivity to the bath (Haarberg et al., 1991). Impurities that show a variable valence such as phosphorus and vanadium, are particularly bad as they can be reduced at the cathode and then reoxidized at the anode, consuming current without producing the product. The plant under observation faced this problem, decreasing CE from 95% to 94% when the pots were fed with alumina from an altered refinery process. The presence of LiF in the bath lowers the liquidus and increases the electrical conductivity of the bath, which results in an increase in current efficiency (Tabereaux et al., 1993; Haupin, 1995).

Bath Composition

Molten cryolite, which has a high solubility for alumina, is the major component of the bath. The most common additives are AlF₃ and CaF₂, which serve to reduce the liquidus temperature. Haupin (1994) has done an excellent review of the effect of additives on the properties of baths. The objectives in changing bath composition are to lower temperature, raise current efficiency, lower voltage, maintain operating stability and decrease emissions while maintaining the proper ledge and not generating sludge under the metal pad. Even a slight change in bath composition may upset the proper functioning of the cell. In the plant under observation, it was noted that a little change in the bath ratio (Wt% ratio NaF/AlF₃ changed from 1.12 to 1.09) decreased bath stability, probably due to the decrease in solubility of alumina resulting in the formation of insoluble particles. Cells with high AlF₃ baths are required to maintain an almost constant bath composition (Richards, 1994). They require a careful feeding of alumina controlled by computer systems using sophisticated software, such as the Track Program (Silva, 1995) and other modifications which are undergoing tests at Alumar. The concentration of alumina is usually kept in the range of 2-5%. With lower concentrations there is a risk of the anode effect, while with higher concentrations there is a risk of sludge formation (Kvande et al., 1994). Production results have revealed that operating with low bath ratios contributes to high current efficiencies. Control of the granular distribution of the alumina added is also very important. As an example, in the test cells where a 90% fraction of +325 mesh alumina was used instead of 82%, the frequency of anode effects decreased. The addition of lithium within a specific composition range improves bath performance (Tabereaux et al., 1993).

The Cathode Lining

Both the cryolite-alumina melt and the liquid aluminum are extremely corrosive materials, and very few materials can withstand their combined corrosive action at 1000 ⁰C. Carbons, graphite and refractory borides (particularly TiB₂) are the only electronic conductor materials that are satisfactory. Preformed blocks of semigraphitized carbon are generally used for the lining. Carefully selected good quality materials are essential for the long life of the pots. In the observation plant, it was found that a slight difference in quality of the material made a lot of difference in terms of pot life. Normally cathode resistance increases with time and the lining is rebuilt after 1200 to 3000 days of operation. One of the main causes suggested for cathode failure is the penetration of sodium metal into the lining and the then the formation of C₆₀Na and C₆₈Na intercalation compounds (Asher, 1959) between the graphite layers causing swelling and disruption (Brilliot et al., 1993). Sidewall failure may be caused by Al₄C₃ formation. Pawlek (1995) has recently reviewed the use of SiC as sidewall material and concluded that SiC-refractories having a nonoxide bonding system are found to be corrosion resistant under the conditions of electrolysis. Micropyretically synthesized pitch-free carbon sidewall blocks are also reported to possess improved properties of oxidation and sodium resistance (Sekhar et al., 1995b). A TiB₂ cathode coating has been reported to be sodium resistant and aluminum wettable, resulting in increased current efficiency and cell life (Sekhar et al., 1995a). The spent potliner is a big problem as it is reported to contain hazardous toxic cyanide (Burkin, 1987). Keller et al. (1995) has reported that additives such as B₂O₃ and B₄C could suppress cyanide formation in the cathode lining. Some research on utilizing the spent potliner in building and road constructions has been carried out.

The Anode

The consumable carbon anode currently used has several disadvantages. It has an overvoltage of about 0.5 V and is an important source of impurities and of carbon particles introduced during electrolysis. The uneven consumption of the electrode creates a disturbance in interpolar distance resulting in an ohmic voltage drop (Fig. 1). A major part of the aluminum reduction cost (about 20%) is contributed by carbon anodes (Grotheim and Welch, 1988). These problems with carbon anodes have stimulated an extensive investigation of both the possibility of improving their functioning and of developing nonconsumable anodes. Several formulations of aggregate coke, binder pitch and the paste material for anodes have been reported (Wang and Liu, 1989; Liu and Xiao, 1989; Liu et al., 1993). Paste materials with lithium compounds have been found to decrease anodic overvoltage by reducing the activation energy of the C-O₂ reaction and increasing the wettability towards molten bath (Pawlek, 1998; Liu and Xiao, 1987; Liu et al., 1993). Qiu et al. (1994) and Watanabe (1983) believe that the presence of lithium decreases the anode effect. Liu et al. (1995)have studied the effect of several electrocatalytic dopants on anodic reaction and have reported that Ba-Fe and Mg-Al dopants can lower the anodic overvoltage by 200 mV and 170 mV, respectively. The optimization of space utilization in the pot by changing the dimensions of the anodes and their relative positions may lead to a significant improvement in the economics of the smelter (Forberg, 1996). It was recently observed at Alumar that the anodes located down-stream and on the pot ends had more burnoffs than the upstream ones. Raising the level of these anodes by 4 cm resulted in a decrease in the number of burnoffs.

Replacement of the carbon anodes by nonconsumable material has been the subject of extensive research since the invention of the process (Billehaug and Oye, 1981; Grotheim and Welch, 1988; Wilkening and Winkhaus, 1989; Nora, 1992). The consumable carbon anode presently used in the cells acts as a "depolarizer", lowering the reversible emf by about one volt, but this advantage is partly offset by a high anodic overvoltage on carbon, i.e., ~ 0.5 V as compared to ~ 1.0 V on some inert anode materials (Thonstad et al., 1987). The use of inert anodes will probably allow operation at a shorter anode-cathode distance so that an overall net energy saving can be envisaged (Jarrett et al., 1982). In addition, one eliminates the need for carbon materials and frequent replacement of anodes and the accompanying disturbance of cell operation. The use of nonconsumable anodes is expected to eliminate the emission of CO₂, CO, CF₄ and polycyclic aromatic hydrocarbons. For nonconsumable anode materials, research has mainly been focused on materials which could be chemically inert towards molten cryolite-alumina baths and to the oxygen produced at the anode. The literature shows that research has mainly concentrated on three classes of materials, namely, metals (Windisch and Marschman, 1987; Hryn and Sadoway, 1993), ceramic oxides (Nora et al., 1978; Yang et al., 1993) and ceramets (Ray, 1983, 1986; Tarcy, 1986). Unfortunately problems such as oxidation and corrosion, contamination of the aluminum produced, electrical conductivity and cost still remain to be solved (Pawlek, 1996).

Process Control

Most process control systems rely on measurement of cell voltage, since other parameters cannot be measured on a continuous basis in the aggressive medium of the cell. When fluctuations (noise) or a rapid change in voltage is detected, the control system must analyze its nature and take appropriate action (e.g., activate alumina feeders, the anode position control system, etc.).

The complex and interactive nature of the cell parameters makes system analysis difficult. For example, cell voltage can be influenced by alumina concentration, interelectrode spacing, cell temperature, electrolyte composition, metal pad depth, amount of sludge in the cell, evolution of undissolved material in the electrolyte, etc., which shows that one cannot attribute a change in voltage to a specific factor. It highlights the importance of an alert, trained and experienced operator, in spite of the use of highly sophisticated computer-controlled cells. A recent report (Caissy et al. 1998) shows that agreement between technical and management practices at Lauralco (Alumax) has led to new industry standards which approach 325 kA amperage and 97% current efficiency. This success could be achieved by using the latest generation pots and a selection of human resources based on their ability to work as part of a team, to be autonomous and to take initiatives. Operators were given theoretical and practical courses, and were then divided into small teams with a clear mandate and precise objectives to meet. A similar system has been adopted at the plant under observation (Prasad, 1999), where operators are offered training courses and are divided into "quality teams" with specific objectives. Frequent meetings are held to exchange ideas over problems encountered and their solutions. Incentives are offered to the teams for important achievements. The system has contributed to the search for new ways to progress and evolve.

CONCLUSION

In spite of intensive research, the aluminum electroreduction process is still faced with several problems such as high energy consumption, stability of cathode lining, development of nonconsumable anode, etc. In the opinion of the author, the long-range goal of the industry should be to combine nonconsumable anodes with inert cathodes to make bipolar electrodes, which would result in a more compact, efficient and economic cell design.

ACKNOWLEDGEMENTS

The author is indebted to UFPB-CCT-PRODENGE and ALUMAR for providing facilities for this work and the operation engineers for useful discussions.

• Asher, R.C., Lamellar Intercalation Compounds of Sodium with Graphite. J. Nucl. Inorg. Chem., 10, 238 (1959).
• Beck, T.R., New Directions in the Aluminum Industry, in: Landau, U., Yeager, E. and Kortan, D. (eds.) Electrochemistry in Industry New Directions, Plenum Press, New York, 331 (1982).
• Billehaug, K. and Oye, H.A., Inert Anodes for Aluminium Electrolysis in Hall-Heroult Cell (I). Aluminium, 5, 7146 (1981) and (II), ibid, 57, 228 (1981).
• Brilliot, P., Lossius, L.P. and Oye, H.A., Studies on Penetration of Metallic Sodium in Cathode Lining of Hall-Heroult Process. Matall. Trans. 34B, 75 (1993).
• Burkin, A.R. (ed.), Production of Aluminum and Alumina. John Wiley, New York (1987).
• Caissy, J., Dufour, G. and Lapointe, P., On the Road to 325KA. Light Metals, 215 (1998).
• Forberg, H.O., Untapped Opportunities for Improving Performance and Increasing Smelter Production. Light Metals, 313 (1996).
• Grjotheim, K., Krohn, C., Malinovsky, M., Matiasovsky, K. and Thonstad, J., Aluminium Electrolysis Fundamentals of Hall-Heroult Process, 2^nd ed. Aluminium-Verlag, Dusseldorf, p. (a) 6, (b) 394 (1982).
• Grjotheim, K. and Welch, B.J., Aluminium Smelter Technology, 2^nd ed. Aluminium-Verlag, Dusseldorf, p. (a) 5, (b) 158 (1988).
• Haarberg, G.M., Osen, K.S., Thonstad, J., Heus, R.J. and Egan, J.J., Light Metals, 283 (1991).
• Haupin, W.E., The Influence of Additives on Hall-Heroult Bath Properties. J. Metals, 28 (1994).
• Haupin, W.E., Principles of Aluminum Electrolysis. Light Metals, 195 (1995).
• Hryn, J.N. and Sadoway, D.R., Cell Testing of Metal Anodes for Aluminum Electrolysis. Light Metals, 475 (1993).
• Jarrett, N., Frank, W.B. and Keller, R., in Metallurgical Treatises, ed. Tien, J.K. and Elliott, J.F., TMS, AIME (1982).
• Keller, R., Stofesky, D.B. and Cochran, C.N., The Potential of Potliner to Suppress Cyanide Formation in Hall-Heroult Cells. Light Metals, 345 (1995).
• Kvande, H., Chen, J. and Haupin, W.E., Minimizing Energy Consumption through Optimizing Alumina Concentration in the Bath of Hall-Heroult Cells. Light Metals, 429 (1994).
• Liu, Y., Wang, X., Huang, Y. and Wang, H., A New Field to Reduce Energy in Hall-Heroult Process Research and Application of an Anode Paste Containing Lithium Salt. Light Metals, 599 (1993).
• Liu, Y., Wang, X., Huang, Y., Yang, J. and Wang, H., New Type of Electrocatalysts for Energy Saving in Aluminum Electrolysis. Light Metals, 247 (1995).
• Liu, Y. and Xiao, H., A Study of the Electrolytic Activity of Doped Carbon Anodes in Cryolite-Alumina Melts, Proc. of the Joint International Symp. on Molten Salts. The Electrochemical Society Inc., Honolulu, Hawaii, USA, 744 (1987).
• Liu, Y. and Xiao, H., A New Approach to Reduce the Anodic Overvoltage in the Hall-Heroult Process. Light Metals, 275 (1989).
• Nora, V., Non-Carbon Anodes for Aluminum Electrowinning. Dechema Monographs, Vol. 125, VCH, Germany, 377 (1992).
• Nora, V., Spaziante, V.P.M. and Nidola, A. (Diamond Shamrock Technologies S.A.): US Patent 4,098,669 (1978).
• Pawlek, R.P., SiC in Aluminium Electrolysis Cells. Light Metals, 527 (1995).
• Pawlek, R.P., Inert Anodes for the Primary Aluminum Industry: An Update. Light Metals, 243 (1996).
• Pawlek, R.P., Lithium in Anodes and Cathodes of Aluminium Electrolysis Cells. Light Metals, 547 (1998).
• Prasad, S., Monitoraçăo e Controle dos Banhos de Eletrodeposiçăo de Ligas Co-Ni. Tratamento de Superfície, 76, 18 (1996).
• Prasad, S., Eletrodeposiçăo de Ligas Amorfas de Tungstęnio. Tratamento de Superfície, 87, 32 (1998).
• Prasad, S., Studies on Hall-Heroult Aluminum Electrowinning Process. Anais do XI Simpo. Bras. Eletroquím. Eletroanal., April 5-9 (1999).
• Prasad, S. and Marinho, F.A., Optimization and Control of the Baths for Electrodepotion of Cobalt-Tungsten Amorphous Alloys. Metal Finishing, 98 (2000). (in press)
• Prasad, S. and Oliveira, L.G., Iron-free Aluminum Sulfate from Bauxite. J. Indian Chem. Soc. 74, 556 (1997).
• Qiu, Z., Yao, G. and Zang, Z., Wettability and Critical Current Density of Lithium-Containing Carbon Anode. Jinshu Xuebao, 30(10), B439 (1994).
• Ray, S.P., Inert Electrode Compositions. US Patent, 4,374,050, Feb. 15 (1983).
• Ray, S.P., Inert Anodes for Hall Cells. Light Metals, 287 (1986).
• Richards, N.E., Strategies for Decreasing the Unit Energy and Environmental Impact of Hall-Heroult Cells. Light Metals, 393 (1994).
• Sekhar, J.A., Bello, V.N., Liu, J. and Duruz, J.J., Cathodic Coating for Improved Cell Performance. Light Metals, 507 (1995a).
• Sekhar, J.A., Liu, J., Yan, J., Bello, V.N. and Duruz, J.J., Carbonaceous Micropyretic Sidewall Materials, 257 (1995b).
• Silva, A.F., Sistema de Controle de Alimentaçăo de Alumina das Linhas 1 e 2 da Alumar. An. V Simp. Tec. Ind. Alumínio, Săo Paulo, 265 (1995).
• Tabereaux, A.T., Alcorn, T.R. and Trembley, L., Lithium-modified Low Ratio Electrolyte Chemistry for Improved Performance in Modern Reduction Cells. Light Metals, 221 (1993).
• Tarcy, G.P., Corrosion and Passivation of Cermet Inert Anodes in Cryolite-type Electrolytes. Light Metals, 309 (1986).
• Thonstad, J., Liu, Y.X. and Jarek, S., Overvoltage on Inert Anode Materials in Cryolite Melts. Proc. 8^th Intl. Light Metal Meeting, Vienna, 150 (1987).
• Wang, P. and Liu, F., Composition of Anode for Electrolytic Cell for Aluminium Production. C.N. Patent 1,039,451 (1989).
• Watanabe, N., Anode Effect and Graphite Intercalation Compounds A New Elucidation of Addition Agent Effect in Carbon Anode Reaction of Molten Fluoride Cells. Proc. 1^st Intl. Symp. Molten Salt Chem. Techn., Kyoto, 21 (1983).
• Wilkening, S. and Winkhaus, G., Materials Problems in Electrowinning of Aluminium by the Hall-Heroult Process. J. Appl. Electrochem., 19, 596 (1989).
• Windisch, C.F. and Marschman, S.C, Electrochemical Polarization Studies on Cu and Cu-Containing Cermet Anodes for the Aluminum Industry. Light Metals, 351 (1987).
• Yang, J., Liu, Y. and Wang, H., The Behavior and Improvement of SnO₂-based Inert Anodes in Aluminium Electrolysis. Light Metals, 493 (1993).

Publication Dates

• Publication in this collection
  06 July 2000
• Date of issue
  June 2000

History

• Received
  06 Apr 1999
• Accepted
  13 Apr 2000