AU619609B2 - Process for producing alumina from gibbsitic bauxites - Google Patents

Description

COMMONWEALTH OF AUSTRA Form PATENTS ACT 1952-69 COMPLETE SPECIFICATION

(ORIGINAL)

Clas I t. Class A pplication Number: Lodged'.

Complete Specification Lodged: Accepted: Published: PrigrIty: Re'lated Art Name of Applicant:. AAGYAR ALUMINIUMIPARI TROSZT Address of Applicant:. 56, Pozsonyi ut, Budapest, Xiii, Hungary.

Artual Inventor: Address for Service.

GYORGY BANVOLGYI, JOZSEF ZOLDI, PEI'ERSIKLOST, IBOR FERENCZI, ANNA OSORDAS, IVAN FEHER, IILDIKO TASSY and TSTVAN SAJO.

EDWD. WATERS SONS, 50 QUEEN STREET, MELBOURNE, AUSTRALIA, 3000.

Complete Specification for the invention entitled: PROCESS FOR PRODUCING ALUMINA FROM GIBBSITIC BAUXITES The following statement is a full description of this invention, including the best method of performing it known 1 to sodium aluminum hydrosilicate of 10 to 80%, and preferab.y to 50%* of the kaolinite. I hswytecmlt In this way the complete llj i i I

I

PROCESS FOR PRODUCING ALUMINA FROM GIBBSITIC BAUXITES Field of the Invention This invention relates to an improved and highly efficient process for the extraction of alumina from gibbsitic bauxites.

Backaround of the Invention In the practice of the well-known Bayer process, bauxite is digested in a sodium hydroxide-sodium aluminate liquor, unsaturated for sodium aluminate, at temperatures between 100°C and 30 0 depending on its mineralogical composition, and the soluble aluminum minerals are transformed to sodium aluminate. After digestion, the slurry is cooled to the atmospheric boiling point. After dilution and rerr. al of digestion 0 residue (red mud) aluminum hydroxide (alumina hydrate) is precipitated out of a solution supersaturated for dissolved alumina by further cooling followed by separating the alumina hydrate precipitate. The mother liquor possibly after concentrating a part of it by evaporation is recycled to the digestion. After washing, the alumina hydrate is calcined to alumina, technically pure A1203, at a temperature exceeding 10000C.

;r

,I

eer rr t rre rr

I

I1i

(II

.7 _ff; 9:A/ i o a> a iI i I rl r i 1 1 i l| 20 In addition to the digestible aluminum minerals (gibbsite, boehmite, diaspore) bauxites always contain some amounts of silicon-containing components. The SiO 2 content occurs mainly as kaolinite (A/203.2Si 2 .2H20) but also as quartz (SiO 2 chamosite, illite, halloysite in the bauxites.

Moreover Fe 2 03- and Tio 2 -containing components can also be found in bauxites. Bauxites containing at least 90% of the digestible aluminum minerals in the form of gibbsite are considered to be of the gibbsitic type.

It is well known from the technical literature that the silica-containing minerals found in the bauxites transform to sodium aluminum hydrosilicates during the Bayer process.

Various authors give values between 0.62 and 0.79 for the Na20 to SiO 2 mass ratio of the sodium aluminum hydrosilicates formed from reactive silica in solution of to 200 g/l caustic Na20 concentrations (135 to 350 g/l as Na 2 C0 3 and 2.5 to 3.5 caustic molar ratios (.27 to .38 A/C ratios) at 140 to 150"C temperatures. (Adamson, A.IT.: Alumina Production: Principles and Practice. The Chemical Engineer, June 1970, pp. 156-171; Sartowski Vargane Kiss Zs., Bulkai D El6kovasavtalanitas a Bayer-eljarasban.

Banyaszati es Kohaszati Lapok, Kohaszat, 114, (1981), 2, pp.

79-85; Yamada, K. et al.: Process for Extracting Alumina From

A

Aluminous Ores. US Pat. 4,426,363). This ratio makes it possible to calculate the chemical caustic soda loss, which is a significant cost component of the alumina manufacturing by Bayer process. Caustic Na 2 0 is defined, as the amount of +suim o 4ce 0- 10(15 O4,d off Na ions equivalent to theAr e act i 'c 0O1 ions present as complex aluminate anions e.g. [AJ.(OH) 4 in the liquid phase, expressed as Na 2 0. The caustic molar ratio is the ratio of the caustic Na 2 0 and the Al 2 0 3 mols present in the solution.

In the American terminology, the caustic concentration 0 0 0o "o understood similarly is expressed as Na 2

CO

3 Likewise, 00 o the ratio of the concentration of dissolved AI 2 0 3 a;i &.LE the So0 caustic concentration expressed as Na 2

CO

3 is also widely used 04000 in American terminology, and it is called "A/C ratio".

Gibbsitic bauxites are usually digested at 140 to 0000 15 150'C (in some exceptional cases at about 105°C) by the so- 04 called low-temperature digestion process. According to the t typical practice, the bauxite is slurried in about 10% to of the liquor recycled to the digestion, which is relatively poor in dissolved A1 2 0 3 (so-called spent liquor). The slurry having a temperature of about 80'C is mixed, after a holding time usually not exceeding 1 to 2 hours and usually without further preheating, with the main flow of the digestion liquor preheated in heat exchangers to about 160 0

C.

By introducing live steam into the mixed slurry the pre- 3 MT0 r1 _1 i o O o 0 0 0 000 o0 0 000 0 0 0 0 0 0 0 00 scribed digestion temperature of 140 to 150°C is maintained.

This briefly described process is called two-stream d.qestion.

Mother liquors with typical caustic Na20 concentrtions of 90 to 120 g/l (150 to 205 caustic as Na 2 CO3) and typical molar ratios of 2.6 to 2.8 (.34 to .37 A/C ratios), obtained after hydrate precipitation (called spent liquors in the technical literature dealing with alumina manufacturing), are used for the digestion of gibbsitic bauxites. The bauxite dosage into the digestion liquor (the bauxite to digestion liquor ratio) is contr .led so that the caustic molar ratio of the liquor phase of the slurry leaving the digestion reactor should be. 1.35 to 1,45 (its A/C ratio should be 0.66 to 0.71). A retention time of about 40 to 100 minutes is usually utilized in the low temperature digestion process. This is carried out in 3 to 5 series-connected digester vessels (autoclaves). Gibbsite is digested during the first 10 to 20 minutes of the digestion process.

Parallel to the reaction of the gibbsite, kaolinite also is dissolved into the liquid phase. The dissolved silica, probably present in the form of some complex ion, transforms into a solid-phase sodium aluminum hydrosilicate in a reaction leading to an equilibrium. A retention time longer than that required for the dissolving of gibbsite is necessary for the formation of nuclei and growth of the i\ crystals. Therefore, retention times-of 40 to 100 minutes are provided in the reactor (Carlos, Interalumina Bauxite Grinding and Digestion. Engineering and Mining Journal, May, 1983, pp.29-94; Kotte, Bayer Digestion and Predigestion Desilication Reactor System. Light Metals. Proc. of AIME Annual Conference, 1981, pp.45-79).

The liquid phase of the slurry, leaving the reactor K at a temperature of 140 to 150°C and with a molar ratio of 1.35 to ].45 (.66 to .71 A/C ratio), is supersaturated for boehmite. The ratio of the bauxite and the digestion liquor is controlled so that the difference between the actual A1 2 0 3 concentration of the liquor and the equilibrium value be not i less than 18 to 20 g/l. Should this difference be less, the secondary boehmite formation amounting to 0.5% to 2% Al203 during the usual retention time would increase and attain an uf\accef4abtle (eve-( ;tont unbear bl from the economic point of view because of the increase of the supersaturation of the liquor for |boehmite.

Li The purpose of the red mud washing is to minimize the caustic Na20 and A1 2 0 3 concentrations of the liquor accompanying the red mud. In the settler the liquid phase is supersaturated for Al 2 0 3 so the precipitation of the equilibrium phase i.e. gibbsite begins. Due to the lower caustic Na20 concentration and temperature in the first t 'Ile

T

j washing stage the supersaturation for-Al 2 0 3 is higher than in the settler, so the secondary gibbsite formation is more intensive. Due to the secondary gibbsite formation, a loss amounting to 2% to 3% of the Al 2 0 3 content of the processed bauxite can be observed during the settling and washing of the red mud.

US Pater- No. 4-650653 Eurcpcan PtCnt Applicatio- n NTo. 3032,73 deals with the O processing of low-Sio 2 gibbsitic bauxites. According to the o 40 o process of Lepetit and Mordini, the bauxite is ground in a liquor of 50 to 120 g/l caustic Na20 concentration (90 to 205 g/l as Na 2

CO

3 originating at least partially from the 4000 washing of red mud or alumina hydrate. The resulting slurry is treated at a temperature of 80 to 100"C for such a time that at least 85% of the kaolinite is uransformed into sodium i 15 aluminum hydrosilicate. The disadvantage of this process is that, as a result of the grinding in a wash water, the digestion is burdened with the feeding of excess water, the evaporation of which will increase the energy consumption of Sthe processing.

20 A procedure for the processing of high-kaolinitecontaining gibbsitic bauxites is proposed by Grubbs (US Patent No. 4,614,641). According to this invention the bauxite is crushed and subsequently classified at 105m (150 mesh). The fine fraction, richer in kaolinite, is digested 6 t I .,tt; lot to r t SIt tots I it So tsrt 1 a 4 Xt in a sodium aluminate liquor, having a-caustic Na20 concentration higher than 140 g/l (240 g/l as Na 2

CO

3 at a temperature of 80 to 130°C. The solid phase is removed by settling and the liquid phase, rich in aluminum and silica, is mixed with the coarse fraction digested in a separate stream at 140 to 150°C.

An unfavorable characteristic of the process widely used for the processing of gibbsitic bauxites is the chemical caustic soaa loss caused by the reactive Sio 2 fed into the alumina plant with an Na20/Sio 2 mass ratio of about 0.65 to 0.69. The dissolving potential of the digestion liquor is not utilized completely, so excess heat and power is used for heating and pumping part of the liquor flow. This also reduces the capacity of the digestion plant uiiit and increases the specific energy consumption. Another disadvantage of the typical low-temperature digestion process is the retention time of 0.5 to 2 hours, which requires the installation of expensive pressure vessels (autoclaves). It is difficult to solve the material handling and to prevent the settling of the solid phase in the autoclaves. Yet another disadvantage of the widely used low-temperature digestion process is the secondary boehmite formation, which often causes A1 2 0 3 losses amounting to 0.5% to The secondary gibbsite formation in the settling and washing system is also C I a 8 a disadvantage of the present processes, leading usually to Al203 losses of 2% to 3%.

Detailed Disclosure of the Invention A principal objective of the present invention is the reduction of the disadvantages of the known processes by a process improvement, which makes possible a reduction of the chemical caustic soda losses caused by reactive SiO 2 enables a more complete utilization of the dissolving potential of the digestion liquor and of the capacity of the equipment. and permits a significant reduction of the time required for the digestion reaction and of the energy consumption of the digestion process. Furthermore, the process of the invention enables the attaining of an SiO 2 concentration in the liquor .0 favorable for the manufacturing of alumina of a quality which corresponds to the S requirements of the market. A further aim of the invention is to reduce or prevent the secondary beohmite and gibbsite formations which cause Al g O 2 losses.

Predesilication tests, carried out at 100°C in liquors containing 100 to 180 g/l caustic Na20 (170 to 310 g/l as Na 2

CO

3 and having molar ratios of 3.1 to 3.4 1 5 (.28 to .31 A/C ratios) with slurries having various solids contents, have shown that, in the concentration range of 100 to 150 g/l caustic Na20 (170 to 260 g/l as Na 2 COs) the transformation of kaolinite practically stops after 4 to 6 hours. An increase of the retention time to even 15 hours will not lead to a greater transformation into sodium aluminum silicate that the 50 to 70% attained with 4 to 6 hours.

j;JL

I,

i$ *i

'I

'i *i 'j .1 'i i** c The connection between the dissolution of gibbsite and the transformation of kaolinite has not previously been investigated in a systematic way. We have investigated the mechanism of the reaction by means of kinetic modelling. Through this kinetic modelling, we have come to the conclusion that the corrmon driving force of dissolution of both the kaolinite and the gibbsite is the reactive OH- concentration in a complex reaction mechanism; the kaolinite and the gibbsite compete for the reactive OHions present in the system. The definition of the reactive OH- concentration is the following: AH 34[((c(eq)A203,gi- cA1 2 0 3 )/102) (cSiO 2 1 0 where: c(eq)AI 2

O

3 gi is the equilibrium solubility of Al 2 0 3 for gibbsite, in g/l.

*Vdo: ft ftt 9 a 0 tee t cA1 2 0 3 is the actual concentration of A 2 0 3 in g/ and cSiOg is the actual concentration of SiO 2 in g/i.

The 102 in the formula represents the molar mass of A1 2 0 3 the 60 the molar mass of SiO 2 and the 34 the molar mass of two OH- ions. There is no significant change if the concentration of the dissolved SiO 2 is neglected at the ME A4VS I C1 II

I:

calculation of AH since this is usually lower by 2 to 3 orders of magnitude than the concentration of A1 2 0 3 According to the technical literature the kaolinite is dissolved in a very short time in the course of the transformation process (Adamson, Alumina Production: Principles and Practice. The Chemical Engineer, June, 1970, pp. 156-171; Kotte, BFyer Digestion and Predigestion Desilication Reactor System. Light Metals. Proc. of AIME Annual Conference. 1981, pp. 45-79). Insofar as we are aware, the relative reaction rates of the parallel reactions of gibbsite and kaolinite in alkali hydroxide-alkali aluminate liquors, and the connection between these reactions, have not been investigated.

In the course of the kinetic modelling we have found (unexpectedly on the basis of the technical literature) that the dissolution rate of gibbsite is 2 to 8 times higher than that of kaolinite. This experience has led to the perception that if a sufficient amount of gibbsite is present, it will use up the available reactive OH- ions, and a certain (under given circumstances a significant) part of the kaolinite will "survive" digestion, even at 140 to 160"C. This perception has been supported by our tests. Where a sufficient amount of gibbsite was present, the Na20/SiO 2 ratios found in the digestion residue were significantly lower (0.4 to 0.5) than w4_ c- te- theoretical value of 0.69. Unreacted kaolinite could be identified by X-ray diffractometry even after a 150°C digestion lasting for 20 or even 60 minutes. These results were unexpected on the basis of previous knowledge.

The best kinetic model, describing the results of the i; tests carri' o i in sodium hydroxide-sodium aluminate solutions nf various concentrations but similar molar (A/C) ratios, suggest that the sodium aluminum hydrosilicate formed during the reaction slows down and finally stops the further 1i transformation of kaolinite. This pnenomenon can be explained by the perception that, during a proper predesilication, kaolinite is partly covered by a sodium aluminum hydrosilicate layer formed during the transformation of a part of the kaolinite and this coating layer prevents the further transformation of kaolinite. Recent research has proven the presence of sodium aluminum hydrosilicate on the i surface of partially transformed kaolinite (Roach G.I.D., White, Dissolution Kinetics of Kaolin in Caustic Liquors. Light Metals. Proc. of AIME Annual Conference, 1988, pp.41-47).

The basis of the present invention is the perception that complete transformation of the kaolinite into sodium aluminum hydrosilicate can be prevented during a 140 to 180*C digestion if the reactive OH- concentration of the liquid 11 I 1 o 0 0 o 0 00 00 00 0 0 0 oo 0 0 0000 I 0 0 oo oo0 0 .004 o 0* 4 0 a 00004 0o 4 0 0 phase of a slurry is kept below 6 g/l .(and preferably below 2 at the outlet of the digestion reactor, and the kaolinite in the bauxite fed into the reactor is not converted into sodium aluminum hydrosilicate during -he previous operations to an extent of more than 80 percent.

The present invention is directed to an improvement in an otherwise known process for producing alumina from gibbsitic bauxites by digesting the bauxite in a sodium hydroxide-sodium aluminate liquor at a temperature between 100 and 180°C, separating the red mud, precipitating alumina hydrate from the sodium aluminate liquor through cooling and agitation, and calcining the precipitated alumina hydr:ate, wherein the improvement is that the reactive OH- ion concentration as defined by the above formula, is kept below 6 g/l, and preferably between 0 and 2 g/l, in the liquid phase of the slurry leaving the digestion reactor. This is achieved by properly controlling the ratio of the bauxite to the digestion liquor.

The digestion equipment consist of three main partsf a preheating section for preheating the digestion liquor and/or the bauxite slurry, a digestion reactor, and equipment for reducing the temperature and pressure of the slurry leaving the reactor.

.qS 4" r~~ equilibrium solubility and actual concentration of Al 2 0 3 Therefore, in the process according to this invention the digestion is -rried out so that the actual A203 concentration, i the liquid phase of the slurry at the g outlet of the digestion reactor, is maintained within 18 g/l, preferably to 0 to 6 g/1 of the equilibrium solubility by controlling the ratio of bauxite and digestion liquor.

Th During the practical realization of this invention it is preferable to compute the controlled variable with the help of a mathematical model of the process. The process can -lso be easily controlled by measuring the electric conductivity of the digested slurry.

During the practice of the process of this invention it is preferable to hold the mixture of the bauxite, and at least 10% of the digestion liquor, for some time. This procedure (called predesilication) is carried out at a temperature between 80 and 120C, and preferably between and 100*C, while providing gentle agitation with an energy input of 0.03 to 0.2 kW/m 3 of slurry. The transformation of the kaolinite is interrupted after 0.5 to 20 hours, and preferably after 3 to 6 hours, after the transformation 13 13 Is i

;I

7 I

I,

A I 1 J20 to sodium aluminum hydrosilicate of 10 to 80%, and preferably to 50% of the kaolinite. In this way the complete transformation of the still-unreacted part of the kaolinite into sodium aluminum hydrosilicate, during the subsequent operations, can be prevented. At the same time, by forming sodium aluminum hydrosilicate nuclei in the solid phase fed into the digestion reactor, the precipitation of the SiO 2 dissolved into the liquid phase during the digestion operation in the form of sodium aluminum hydrosilicate is promoted.

The process of the present invention provides significantly shorter digestion retention times for gibbsitic bauxites than is customary with prior art processes, i.e., 1 to 60 minutes, and preferably 3 to 20 minutes. This process can also be economically carried out in a tube digestion reactor which is more advantageous from the point of view of operation than the so-called autoclave batteries typically employed.

Furthermore, it is preferable to feed a sodium hydroxide-sodium aluminate liquor weak in dissolved A1 2 0 3 to the dilution and/or red mud separation (settling or filtering) and/or red mud washing. In this way the supersaturation of the liquid phase for A1 2 0 3 in the washing operation can be reduced, and thereby the undesirable secondary gibbsite

*I

formation is also reduced. An alumina plant liquor is considered to be weak in dissolved Al 2 0 3 if its caustic molar ratio exceeds 2.5, and preferably 2.7 (A/C ratio is less than i 0.38, preferably 0.36). Such liquor, weak in dissolved A1 2 0 3 can be spent liquor and also a liquor originating froim red mud causticizing, carbonate salt causticization, alumina hydrate washing and hydrothermal treatment of red mud, or any mixture of them. The liquor(s) weak in dissolved A1 2 0 3 can be fed to the red mud slurry, or to the washing liquor, at one or more points in the process of red mud settling and/or washing.

It is preferable to determine the optimum distribution i| of the spent liquor by-passing the digestion between the settling operation ar,. the washing line ei the basis of a techno-economic calculation, also taking into consideration the process control aspec.;.

The optimum combina-ion of the process parameters has to be determined for bauxites of various origin and quality by optimizing experiments.

A simplified example is given for the determination of the acceptable SiO 2 concentration after digestion. Let us consider an alumina plant processing a good quality gibbsitic bauxite and giving a low amount of red mud. Let 7 -T ll i0 1 I ji !i 5 i i i 1 i i: l i r.; 115 4;1 us suppose that 75 kg A1 2 0 3 will be -precipitated from 1 m 3 of green (pregnant) liquor during the decomposition. The SiO 2 concentration of the product alumina is to be limited to 0.020%, and 25% of the SiO 2 content of the liquid phase leaving the digestion is to be precipitated during dilution and settling. It is known from the technical literature that some 2% of the dissolved SiO 2 fed into the decomposition will get into the product alumina (Teas, Kotte. the Effect of Impurities on Process Efficiency and Methods for Impurity Control and Removal. Proc. of JBI/JGS Symposium 1980 titled "Bauxite/Alumina Industry in the Americas"). Let us further suppose that the caustic Na 2 O0 concentration of the liquor leaving the digestion reactor, before the heat recovery, and that of the liquor fed into the decomposition, do not dtiffer significantly and that the whole amount z? spent liquor will pass through the digestion. In this case the amount of the liquid phase leaving the digestion reactor and that entering the decomposition will be similar. According to the assumptions, a 1 g/l SiO 2 concentration may be allow;ed in the liquid phase leaving the digestion reactor, calculated from the SiO 2 balance.

The main advantages of the process corresponding to the present invention are the following: The complete transformation of the silicacontaining minerals is prevented. As a rei i II suit, the chemical caustic soda losses are reduced to an Na2O/SiO 2 ratio of 0.4 to 0.55.

Compared to the previously used processes, more bauxite can be digested by a volume unit of digestion liquor. Accordingly, the energy of consumption -e the digestion process is less and the digestion equipment can be smaller or the capacity of existing equipment increased.

The digestion retention time can be reduced, o-f which enables a further capacity increase oexisting reactors or a significant saving can be made at the installation of new equipment.

By using the present process, the gibbsiteto-boehmite transformation, proceeding as a secondary reaction of the digestion, and re-uced resulting Al 2 0 3 losses, can be dereeaeCld or be practically eliminated.

Secondary gibbsite formation in the settling and red mud washing operation can be reduced significantly.

The following comparative Table shows the comparison of the widely used procedures for the low-temperature digestion, and ot the process according to the invention.

17 r ~3111 I r^ COMPARATIVE TABLE typical conventional process process according to this invention digestion liquor Caustic Na 2 O concentration g/l caustic as Na 2 CO3 g/

A

2 0 3 concentration g/1 A/C ratio SiO 2 concentratirn g/l molar ratio after digestion A/C ratio after digestion digestion temperature °C 1 5 retention time of the slurry in the reactor min reactive OH- ion concentration at the outlet of the digestion reactor g/l 2 ratio in the red mud secondary boehmite formation during digestion A1203% secondary gibbsite formation during red mud settling and washing A1 2 0 3 140 240 84 0.35 0.4 1.35 0.71 143 100 7 0.69 0.5-2 2-3 140 240 84 0.35 0.4 1.20-1.25 0.77-0.80 150-160 3-15 2 0.40-0.55 0-0.5 0-0.5 1.40 0.69 20-25 by means of the following pregnant liquor molar ratio 1.40 A/C ratio in pregnant liquor 0.69 spent liquor by-passing the digestion 0 The present invention will be further explained 1 examples. However, they are not intended to limit the claims.

MVS

;i i 1: I I Example 1 An "ultrafine" quality bauxite, originating from the -fo Ito yWvy Trombetas deposit in Brazil and having the filing composition, was used for the tests: i 0 I I i A1 2 0 3 in gibbsite boehmite Sio 2 as quartz in kaolinite Fe 2 0 3 as hematite in goethite Loss on ignition 51.0-51.4 0.4- 0.8 2.2 54.0 0.52 2.53 3.05 7.7 4.3 12.0 29.9 This bauxite was digested in alumina plant liquor containing 152.3 g/l caustic Na20 (260.4 g/l as Na 2

CO

3 79.7 g/1 A1 2 0 3 (A/C ratio 0.31) and 0.3 g/l SiO 2 235g dry bauxite was charged into the digester bombs for every liter of digestion liquor.

The testing time was measured from the moment when the bombs were immersed into the oil bath previously preheated to 150°C. As a consequence of the experimental technics applied (because of the time required for the heating of the digester bombs and of the slurry inside them) the actual retention time of the slurry at the temperature of thp bath was less than the measured test time. Subsequent to the test time periods the bombs were opened after cooling, the liquid phase was separated by centrifuging and analyzed. The solid phase was washed with distilled water three times, dried and analyzed. The results were the following liquid phase test time caustic A12CQ molar A/C SiO2. reactive as Na2CQO ratio ratio OHmin g/ g/I g/I g/ g/ 10 136.2 232.9 179.2 1.25 0.77 1.07 3.14 133.1 227.6 182.4 1.20 0.80 0.92 0.26 134.2 229.5 185.5 1.19 0.81 0.68 0.03 Test solid phase time Na2D A1~ 1 5 min SiO 2 Si02 1 0 0.137 6.90 0.319 1.80 20 0.327 1.52 SIt should be mentioned that, if the Al103/SiO2 mass ratio of the sodium aluminum hydrosilicate formed from the kaolinite content of the bauxite under the given test condition is taken as 0.85 (equal to that of the kaolinite), and the boehmite content is considered as undigestible, the A1 2 0 3 /SiO 2 mass ratio expected in the mud in the case of a complete digestion of the gibbsite content would be 1.01 or 1.17 depending on whether the boehmitic Al 2 0 3 content is calculated with 0.4 or 0.8%.

As the above table shows, the molar ratio of the liquor, the Na20/SiO 2 and the Al20 3 /SiO 2 ratios measured in the solid phase, and the SiO 2 concentration of the liquor are all favourable after 15 and 20 minutes. No secondary beohmite formation was experienced.

4J4 MV A'lqS- ^w s i? 'i i- c~- 21 Example 2 The composition of both the bauxite and the liquor used for the tests was the same as in Example 1. The digestion was carried out in the digester bombs at 1500C with 60 min reaction time. The results were the following: liquid phase Bauxite charge caustic A1 2 q0 molar A/C SiO 2 reactive as Na20 as Na2CQ 3 rai:- ratio g/l 235 221 207 Bauxite charge 9 g/l 135.4 134.9 132.3 solid phase Na 2 O A1 2 0 3 SiO2 SiO2 g/l 231.5 230.7 226.2 g/I 189.6 180.9 174.5 1.17 1.23 1.25 g/l 0.82 0.58 0.78 0.73 0.77 0.54 g/l 0.00 1.97 2.63 235 221 207 0.346 0.386 0.498 1.21 1.02 1.10 Despite the long reaction time, favorable molar ratios, SiO2 concentrations and Na20/SiO 2 and A1203/SiO 2 mass ratios were obtained. The presence of unreacted kaolinite could be detected in the solid phase after 60 min digestion time. No secondary boehmite formation was experienced.

jvlg* 22 Example 3 The composition of the bauxite was the same as in Example 1. The alumina Splant liquor used for this test contained 140 g/I caustic Na20 (239.4 g/I as Na 2

CO

3 74.1 g/I Al 2 0 3 (0.31 A/C ratio) and 0.4 g/l SiO 2 After charging 800 g dry bauxite to 1 liter liquor, a 2 hour predesilication was applied at 100°C under gentle mixing.

Subsequently the slurry was diluted with the digestion liquor to such an extent that a Ibauxite charge of 232 g/l liquor was obtained. The slurry was subjected to a 15 min digestion at 1500C in digester bombs. After the digestion, the following results were obtained: j 10 liquid phase Test time caustic A1 2 g 3 molar A/C SiQCI reactive as NaO2 as Na 2

CQ

3 ratio ratio OHmin g/l g/l g/l g/l g/l 15 127.2 217.5 165.4 1.27 0.76 0.81 2.49 Test solid phase I time Na20 A 2 lgQ min SiO 2 SiO 2 0.504 1.01 Favorable Na20/SiO 2 and Al 2 0 3 /SiO 2 mass ratios, digestion molar ratio and SiO 2 concentration were obtained. No secondary beohmite formation was experienced.

J 1 iA, A VS rl ",Mx ,j S -LC 1 r jC- 23 Example 4 The composition of the bauxite and the liquor and also the bauxite charge were he same as in Example 3. A 4 hour predesilication was applied at 1000C, the slurry was diluted to the same extent as in Example 3 and subsequently digested. The data measured after a 150°C digestion were the following: liquid phase Test time caustic A1 2 ql molar A/C SiO2 reactive as Na20 as Na2CCQ ratio ratio OH- 1 0 min g/l g/l g/l g/l g/l 126.4 216.1 164.2 1.27 0.76 0.71 3.08 Test solid phase time Na20 min SiO 2 SiO 2 15 0.522 1.02 The digestion molar ratio, the SiO 2 concentration and the Na20/SiO 2 and the A2lg0/SIO 2 mass ratios were all favorable. No secondary boehmite formation was experienced.

j*

Y

A4V

Claims (7)

1. In the process for producing alumina from gibbsitic bauxites by mixing the bauxites with a sodium hydroxide-sodium aluminate digestion liquor, digesting the bauxites in said liquor in a digestion reactor at a temperature in the range 100-1800C, diluting the resulting slurry, separating the red mud therefrom, precipitating the alumina hydrate from the thus-obtained sodium aluminate liquor through cooling and agitation, and calcining the precipitated alumina hydrate into alumina, and washing the separated red mud, the improvement characterized by so controlling the process that, in 1 0 the liquid phase of the slurry leaving the digestion reactor, reactive hydroxide ion concentration (AH) is maintained below 6 g/l, where such reactive hydroxide ion concentration is defined by the formula AH 34[((c(eq)A1 2 0 3 ,gi cAl 2 0 3 )/102) (cSiO 2 where: c(eq)Al 2 0 3 gi is the equilibrium solubility of Al 2 0 3 for gibbsite, in g/l. 1 5 cAl 2 C03 is the actual concentration of A1 2 0 3 in g/l and cSiO 2 is the actual concentration of SiO 2 in g/l.

2. A process according to claim 1, further characterized by the reactive hydroxide ion concentration being maintained at a level between 0 and 2 g/l.

3. A process according to claim 1, further characterized by the reactive hydroxide ion concentration being maintained by controlling the ratio of bauxite to the digestion liquor.

4. A process according to claim 1, further characterized by the bauxite being mixed with at least 10% of the digestion liquor, and the resulting bauxite slurry being held before digestion at a temperature of 80-1200C for 0.5 to 20 hours.

A process according to claim 4, further characterized by said resulting bauxite slurry, while being held before digestion, being agitated using agitating power of 0.03-0.2 kw per cubic meter of slurry.

6. A process according to claim 1, further characterized by a sodium hydroxide-sodium aluminate liquor weak in dissolved Al 2 0 3 as determined by its having a caustic molar ratio in excess of 2.5 (A/C ratio as hereinbefore defined being less than IijE r 0.38), being fed to the digested slurry in the course of its dilution and/or of the separation and/or washing of the red mud.

7. A process according to claim 1, further characterizer' the digestion procedure being carried out in such manner that the difference bu,, ,en the actual concentration of A2l g O and the equilibrium solubility thereof, in the liquid phase of the slurry leaving the digestion reactor is not greater than 18 g/l. DATED this 6th day of November, 1991. MiAGYAR ALUMINUMIPARI TROSZT WATERMARK PATENT TRADEMARK ATTORNEYS THE ATRIUM 1 5 290 BURWOOD ROAD SHAWTHORN VICTORIA 3122 AUSTRALIA I