FR2458512A1 - CHLORINATION OF AN ALUMINA-LIKE MATERIAL AT PRESSURES HIGHER THAN ATMOSPHERIC PRESSURE - Google Patents

Abstract

PROCESS FOR CHLORINATING AN ALUMINOUS SUBSTANCE CONTAINING OXYGEN, CONSISTING OF CHLORINING THE ALUMINOUS SUBSTANCE IN THE PRESENCE OF A REDUCING AGENT, CHARACTERIZED BY THE FACT THAT THE CHLORINATION IS CARRIED OUT UNDER A PRESSURE GREATER THAN ABOUT 3BARS.

Description

CHLORINATION OF AN ALUMINUM MATERIAL

  AT PRESSURES HIGHER THAN ATMOSPHERIC PRESSURE

  The present invention relates to an improved process for chlorinating an aluminous substance containing oxygen,

in the presence of a reducing agent.

  The chlorination of aluminous substances containing oxygen, in the presence of a reducing agent, such as carbon, to prepare aluminum chloride, is well known. Various processes have been studied to improve the degree

or the quality of the chlorination, but the possibility of

  Serving pressures above atmospheric pressure is an approach whose exploration has never been pushed very far until now. Pressure was used to condense aluminum chloride vapors, to promote the melting of aluminum chloride (French patent no. 334,132), to force chlorine gas through molten aluminum (United States patent 'America No. 1,165,065 in the name of Brooks) and to form briquettes from coking coal and alumina or bauxite (United States Patent No. 1,217,471 in the name of Mc Afee). However, it has not been recognized, up to the present invention, that the chlorination of an aluminous substance in a reactor maintained at pressures higher than about 3.04 bars would at the same time have the advantage of increasing the mass capacity of the apparatus, of factor F, (in which F = gauge pressure in the reactor [barsll + 1.014 [bars] 1.014 [bars] and to increase the degree of approach of the complete conversion of the gas at the constant contact time between materials An explanation for this shortcoming to seriously consider the use of pressures higher than about 3 bars, lies in the fact that the thermodynamic studies conducted at atmospheric pressure indicated that the dependence between the reaction speed

  and the chlorine pressure was lower than first order.

As one would expect that a reaction which is inferior to the first order in relation to the chlorine pressure, should provide a product with a ratio between speed and pressure which decreases with increasing pressure, l those skilled in the art could not consider the use of pressures above atmospheric pressure as a valid means to improve the

production of aluminum chloride.

US Patent No. 3,142,163, in the name of Russell et al., Is an example which illustrates that some specialists in the technical field considered, had considered the use of pressures above atmospheric pressure without believing that the increased pressure in the reactor would improve the efficiency of the chlorination. Although it is mentioned in this last patent that it is possible to use pressures between approximately 0.10 and 10 bars, this patent indicates that chlorination is generally carried out at approximately 1 bar, pressures of 1 -3 bars being described as preferred. Although in the Russell et al. Patent, it is mentioned that higher pressures allow a greater flow rate, which normally has the effect of compensating for any reduction in yield, an operation at pressures higher than approximately 3 bars was not considered a sufficiently interesting technique for us to find

  in literature actual test descriptions at its

topic. According to US Patent No. 2,048,987 in the name of Atherholt, pressures of about 2 to 3 bars are used to promote the approach of the reaction towards equilibrium and consequently cause the formation of more aluminum chloride per kg of chlorine. However, the applicant for this latter patent escaped by increasing the mass flow in a reactor maintained at pressures above 3 bars,

  the speed: pressure ratio would be increased.

The object of the invention is essentially to prepare aluminum chloride by the chlorination of alpha alumina

under high pressure conditions.

  Alumina, or aluminum oxide, exists in various forms or phases. For example, the product resulting from the implementation of the Bayer process is the compound of formula Al (OH) 3, commonly known as alumina trihydrate. To avoid the formation of undesirable amounts of HCl during chlorination, these hydrated forms of alumina are normally dehydrated or calcined at elevated temperatures to remove a substantial portion of the water. However, in the subsequent atmospheric chlorination of these dehydrated or calcined aluminas, it has been found that most of the most stable alpha alumina phase present

  very low reactivity in the chlorination reaction.

  Thus, in the British patent n ° 668.620, one warns against the use of calcination temperatures higher than 1000 ° C to avoid or inhibit the production of alpha aluminum oxide. In the latter patent, it is mentioned that the use of aluminas produced below

1000'C "... makes it possible to obtain considerable yields

  significantly higher compared to chlorine and oxide

carbon, than it is allowed with alpha aluminum oxide. "

  In U.S. Patent No. 3,842,163

aforementioned, a process is described and claimed for the preparation

ration of aluminum chloride which is based on the chlorination in a fluidized bed of an alumina with which is closely associated a carbonaceous reducing agent, as a result of the cracking and coking beforehand of a liquid hydrocarbon in the presence of particles of alumina. In all of this latter patent, the applicants however emphasize the use of alumina having an alpha alumina content of less than 5% by weight. The lack of reactivity on the part of alpha alumina takes on particular importance when using a continuous process instead of a discontinuous process due to the gradual accumulation of non-reactive alpha alumina in the reactor, which requires possibly shutting down the reactor to remove the alumina

non-reactive alpha.

It is known to chlorinate an aluminous substance in a molten bath using a chlorine sauce with a liquid or solid reducing agent. Such a process is described in US Pat. No. 4,039,648 in the name of Haupin et al. In the above-mentioned molten bath process, various types of aluminas can be chlorinated, including alpha alumina. However, it remained advantageous to be able to have a fluidized bed process (non-molten) or a modification of the existing fluidized bed process in which it is possible to chlorinate non-reactive alpha alumina, thereby eliminating the need to update shutdown of fluidized bed reactors, and the need to remove non-reactive alpha alumina to get rid of them

accumulated.

  According to the invention, there is provided a method for chlorinating an aluminous substance containing oxygen, which consists in chlorinating the aluminous substance in the presence of a reducing agent under a pressure greater than about

3.04 bars.

In the context of the invention, the definitions are used.

following tions A metallurgical grade alumina (MGA) is an alumina suitable for the preparation of aluminum by

  Hall process, which has a higher alpha alumina content

less than 10%.

  A partially calcined alumina (PCA) is an alumina having a gamma or other transition phase alumina content greater than 99% and an alpha alumina content

less than 1%.

  The initials are B.E.T. a method for measuring the specific surface which is described by Brunauer et al., Journal of American Chemical Society, 60,

309-319 (1938).

  Kg-mole is a unit of measurement which, once multi-

  folded by the molecular weight of a compound, provides the

  weight of said compound in kilograms.

  The method of the invention comprises the characteristic

2 4 5 85 12

that a reactor pressure greater than about 3 bars is maintained. This pressure above atmospheric pressure has the effect of increasing the production rate of aluminum chloride to an unexpected degree. The rate of production of aluminum chloride is similar

  to a mass transfer coefficient and does not differ on its own-

  ment of the latter only by the fact that the production rate of aluminum chloride is measured by means of the volumes

rather than surfaces.

  In the accompanying drawings Figure 1 shows a comparative graph showing the effect of pressure on the production rate of aluminum chloride for a coked alumina system, a mixture of PCA-coke and a mixture of MGA-coke in a known chlorination process. We can see the graphical representation of the production rate of aluminum chloride in kg-moles of AlCl3 / hour.m.bar as a function of the pressure in bars Figure 2 is a graph showing the effect of

pressure on the average production rate of aluminum chloride

minium in a reactor operating discontinuously under high pressure; Figure 3 shows a reactor system for the

chlorination of an aluminous substance at higher pressures--

  res at atmospheric pressure; and Figure 4 is a reaction diagram illustrating the

method of the invention.

  In accordance with the invention, an aluminous substance containing oxygen is subjected to chlorination under a pressure greater than approximately 3 bars in the presence of a reducing agent, the preferred pressure range being between approximately 5.07 to 15 , 20 bars, and all in all

particularly preferable from 7.10 to 15.20 bars.

  The aluminous substance can be a refined alumina, such as that resulting from the Bayer process, a coked alumina, as described in the aforementioned US Patent No. 3,842,163, or a raw material such as bauxite or clay. used in conjunction with an agent

  separate reducer. Appropriate reducing agents include

  sulfur, coke, finely divided carbon, oxide

  carbon, COC12 and carbon tetrachloride.

In accordance with said embodiment of the invention, a source of aluminous substance containing alpha alumina is subjected to chlorination under a pressure greater than 10 bars in the presence of a carbonaceous reducing agent in a fluidized bed. The use of the expression 1.0 "fluidized bed" is in no way intended to be limited to any particular degree of fluidization, but is intended

  simply to be distinguished from reactions in the molten bath.

Although the source of the aluminous substance can vary significantly, it is more than likely that it is refined substances, namely alumina from the Bayer process, which are the starting substances, since these substances are commonly subjected dehydration by calcination at a high temperature which can lead to the production of alpha alumina if the temperature is high enough. Most likely, the source of the aluminous substance will be a product resulting from the chlorination process as described in the aforementioned US Patent No. 3,842,163. As described above, such low pressure chlorination reactions result,

  when implemented continuously, accumulating-

  tion of non reactive alpha alumina which must be disposed of in any way. In accordance with the implementation of the

process of the invention, this can be done conveniently-

ment by transferring the non-reactive alpha alumina into a reactor holding the pressure or by arranging the reactor currently operating at atmospheric pressure in such a way that it can be used differently under high pressures, namely pressures higher than

bars.

The source of the aluminous substance containing alpha alumina can be a coked alumina, as described in the patent of the United States of America in the name of Russell et al. cited above. Alternatively, the source of the aluminous substance may be a refined alumina or a raw material such as bauxite or clay used in conjunction with a separate reducing agent, such as coke, finely divided carbon, or agents gaseous reducers such as CO, COCi2 and CC040 The amount of alpha alumina in the source of the aluminous substance can vary considerably. However, in a continuous process, even very small amounts of alpha alumina in the aluminous substance could possibly accumulate if atmospheric pressure were used, i.e. amounts on the order of as little as 0.1-0 , 5% by weight. Thus, according to the invention, the amount of alpha alumina of the substance

feed can vary from 0.1 to 100% by weight.

  Consequently, when the chlorination reaction is carried out in accordance with the invention, under a pressure greater than, 13 bars, the differences in the mass transfer coefficient or in the rate of production of AlCl 3 as a function of the calcination temperature of the source of the charge of aluminous substance in the chlorination reactor,

  are insufficient to interfere with satisfactory production.

  Chlorination can be carried out under conditions

discontinuous or continuously.

  In accordance with said embodiment of the invention, a pressure-holding container must be used capable of withstanding pressures greater than bars and which can rise to values' reaching bars. Preferably, the reaction is carried out at approximately 15 bars in a pressure-holding container provided with an external envelope capable of withstanding these pressures and lined with materials capable of resisting corrosion.

at high temperature under the effect of reactants.

  According to said embodiment of the invention, the reaction pressure must be within a range ranging from a value greater than 10 bars up to approximately 20 bars. Preferably, the pressure range ranges from about 12 to 20 bars, - 2458fl2 the most preferred pressure value.

being approximately 15.20 bars.

The reaction is carried out at a temperature of about 550 ° C to 800 ° C. The temperature can be varied somewhat, inversely as a function of the pressure, so that lower temperatures can be used as the pressure is increased. Preferably, a temperature of approximately 700 ° C. is used under a pressure of approximately

15.20 bars.

The chlorination can be carried out under batch conditions or preferably carried out continuously. When chlorination is carried out continuously, suitable apparatus well known to those skilled in the art is used

  to supply the reagents under sufficient pressure.

  Thus, the chlorination reaction can be carried out using any suitable device which is capable of withstanding the pressures of the invention greater than atmospheric pressure. We will now describe the reactor system used to develop the results appearing in the examples below, with the aim of illustrating a system

possible given.

  The fluidized bed reactor 6, represented in FIG. 3, is an electric furnace which comprises an envelope

  made of chlorine-resistant metal (for example, the alloy containing

  nominally 80% Ni, 15% Cr and 5% Fe, and marketed under the brand name INCONEL) lined with graphite. The fluidizing grid is a disc of mullite, 0.64 cm thick, provided with 3 holes with a diameter of 0.32 cm for the distribution of chlorine gas, and a hole

central for the introduction of a thermocouple in the bed.

Buffer chambers are pressurized with nitrogen to introduce the solid charge into the reactor via line 10, for the return of dust via line 11 and to discharge the solid residues from the bed.

  in the solid residue tank 7 via line 12.

  A small continuous discharge of solids from the bed is passed through an overflow, to ensure a constant bed height. A quartz baffle is inserted in the graphite liner between the charge and discharge branches to prevent the charge of solids from

  either short-circuited out of the discharge branch.

A nitrogen purge is added to the bottom of the liner to prevent chlorine leakage behind the graphite liner. This nitrogen purge represents a slight dilution

however negligible, the charge of chlorine gas.

  The flow rate of the solids charge is adjusted by means of

of a feed device 1 with rotary paddles.

  A programmable regulator regulates the opening and closing of the chamber valves

  pads in cycles of one minute.

  The chlorine gas flow rate is regulated by a ball valve 8. The mass flow of chlorine is measured by a differential pressure transmitter. The mass flow signal is corrected for pressure and

temperature.

The reactor is contained in an electric furnace with five zones, the temperature of each zone of which is regulated

  separately. The temperature of the bed is measured by a thermo

internal torque extends upward through the central hole of the fluidization grid up to 15 cm above said grid. The temperatures of the furnace zones are manually adjusted to maintain the desired reaction conditions. The dust is filtered from the gaseous product of the reactor in two dust collectors 2 connected in parallel. The two collectors are each equipped with gravel filters and nitrogen back-blowing means

  hot for cleaning filters.

  A sample stream 13 of the filtered production gas is taken directly downstream of the dust collectors. The gas sample is passed through a quartz wool filter to remove the final traces of dust, then it is cooled to 50 ° C. in a water jacketed heat exchanger to condense and remove the metal chlorides before d 'send said sample to a gas chromatograph 9 for analysis of CO, CO2, C12' COC12 and N2 * Electric heating means are mounted on the gas exhaust pipe, dust collectors, the sampling line for gas samples and side branches of the reactor to prevent condensation of metal chlorides and resulting clogging. All the gas exhaust pipes are metallic Inconel pipes, including the gas sampling line. All actuation valves have nickel ball valves with metal or polytetrafluoroethylene seats depending on the operating temperature. Pressure taps are

distributed over the entire reactor, exhaust pipes

  gases and dust collectors to measure pressure drops. We purge all the taps with

nitrogen to prevent clogging.

The pressure from the reactor to the pipe is measured

  inlet 4 of chlorine gas, at the bottom of the reactor.

  The pressure adjustment valve is positioned between the

  dust collectors and desublimer.

  A desublimator 3 is used, as described in the patent of the United States of America No. 3,930,800 of January 6, 1976, with a heat exchanger cooled with water and fluidized with air, to condense and collect the aluminum chloride produced. We send the waste gas

of the desublimer in a caustic leaching installation 5

only for final processing.

The tests described in examples 1 to 18 correspond respectively to four to six hours of operation in steady state with a power supply

continuous and a total recycling of dust in the reactor.

  A constant bed height of 70 cm is maintained, as mentioned above, by discharging a small amount of solids from the reactor (1-2 kg / hour) from the side discharge branch of the reactor. The pressure drops through the grid and the fluidized bed typically amount, respectively, to 6 cm and 60 cm, measured in a column of water. The following examples are intended to illustrate the invention by way of indication without in any way limiting it for

  both in its setting and its spirit.

Examples 1-9

Charged alumina having a carbon content of 18% by weight, a hydrogen content of 0.27% by weight, an alpha alumina content of 0.6% by weight is charged to the reactor described above. and a specific surface of 8 m / g (BET), and it is subjected to chlorination. Sodium chloride is added to the reactor at the same time as the alumina feed, to react a certain amount of the aluminum chloride produced so as to ensure the presence of a high amount of NaAlCl 4 as catalyst in the reaction reactor. chlorination. We represented

  visually in Figure 1, the parameters and results

states of the process as can be seen in Table I. Table I Examples with coked alumina Experiment NI II III IV V VI VII VIII IX Temperature (C) Pressure (bars) 4.66 1.41 1.62 3.44 3.34 3.44 4.46 4.66 4.76 Surface speed (cm / sec) 12.1 13.8 16.1 6.5 13.2 6.1 8.3 6, 2 11 , 6 A1203 conversion rate (kg / hour) 6.3 1.7 2.1 1.8 6.2 2.6 3.1 2.2, 6 AlC13 production rate (kg-mole / hour. m3.bar) 13.43 8.30 8.20 7.40, 60 12.48, 37, 00 13.07 "J ru Ln co Qn no I We demonstrate the calculation of the production rate of aluminum chloride, depicted in the last column of the

Table I, using the results of Example 1.

  The relevant experimental conditions and the results of Example 1 are as follows: Reactor pressure 4.66 bars of bed height 0.7 m Bed diameter 0.0825 m C12 feed speed 9.8 kg / hour Speed N2 feed rate 0.14 kg / hour Solid matter feed rate 6.3 kg / hour H2 content of solid matter loading 0.27% Reactor temperature 635 C Analysis of reactor waste gases

CO2 11.2%

CO 0.0%

C12 1.1%

COC12. 0.8%

The percentage of the volume concentration of AlCl3 in the reactor production gas is calculated from the analysis of the waste gases: AlCl3 = 4/3 CO2 + 2/3 (CO + COC1l2) AlC13 = 15.5% On normalizes the concentrations of the gaseous product to 100% (without the N2 content from the pressure taps,

back-blowing on filters and purges).

Actual volume% Normalized volume% Co2 11.2 39.2

2,000.0

CO 0.0

c12 1.1 3.8

2 2.8

COC12 0.8 2.8

AlCi3 15.5 54.2

3,100.0%

  The moles of the production gas are calculated from the abovementioned standardized percentages and from a mass balance of C12. Moles of C12 from production gas = Moles of supply C12 (C12 + COC12 + (1.5AlC13)) (moles of gas) / 100 = (C12 rate) / 70.9 moles of production gas = 0.15725 kgmoles / hour. The dilution with N2 and the release of HCl are then added to the abovementioned results to determine the number

  total moles of reactor waste gases.

  Dilution with N2 = (0.14 -kg - / 28 = 0.005 kg-m / hour.

  hour To determine the release of HCl, it is assumed that the total hydrogen content of the solid charge reacts in

forming HC1.

  HC1 released = 0.0027 x 6.3 = 0.01701 kg-m / hour Total number of moles of residual gas from the reactor =

= 0.15725 + 0.005 + 0.01701

= 0.17926 kg-m / hour The production gas concentrations are renormalized,

including the values of N2 and HCl, at 100%.

Normalized volume%

CO2 34.4

CO 0.0

Cl2 3.3

COC12 2.5

AlC13 47.5

HC1 9.5

2 2.8

  0.0% The standard concentrations of production gases, the pressure of the reactor and the temperature of said reactor are introduced as input data into a

  calculator that assesses partial pressures at equilibrium

  bre of various gases, based on thermodynamic values.

Normalized volume% Co2 34.4

CO 0.0

C12 3.3

COC12 2.5

AlCl3 47.5

A12Ci6 -

Aid -

HCl 9.5

N2 2.8

Total 100.0% Partial pressures (bars) at actual temperatures and pressures 1.77 0.121 0.295

0.00517

  1.20 0.616 0.489 0.144 4.66 bars The logarithmic pressure recorded for chlorine gas is calculated. P.-P p 1 o lm -Pi i ln (PO) Pi (N2 corrected) = (rate of C12 / 70.9) / ((rate of C12 / 70.9) + (rate of N2 / 28)) pressure Rx P. = 4.50 bars i Po = Pc12 + PCOCl2 = 0.2958 bars + 0.00518 bar = 0.30098 bar

= 4.498 - 0.30098

P = m 4 in (4,498)

0.30098

  = 1.552 bar The rate of production of AlC13 is calculated from the rate of C12, and the conversion of C12 is corrected taking account

of the release of HC1.

  rate of 1.-1 feed rate H% 10) 345 AlC3 2 rate <(solid matter) (H% / 100) (35.45)

  , 9 (conversion of Cl2) (PAî î + 2PAî CJ.

3 2 6

1.5 PAîîî + 3PAl Cl

AlC3 rate = 0.079906 kg-m / hour.

The volume of the bed for the pilot reactor is determined. Volume of the bed - i (O, 0825) x 0.7 Volume of the bed 0.003742 m3 The rate of chloride production is then calculated.

aluminum (PRC) -

  PRC = (rate of AlCl3) / (bed volume x Plm) PRC = 13.7 kg / mole AlCl3 / hour.m 3.bar The significance of the rate of production of aluminum chloride can be better appreciated by means of the chlorine pressure and a simple speed equation. When the pressure effect of chlorine is less than the first order <1 (ie, speed = k [Pci] ..), as is the case with chlorination of aluminum at atmospheric pressure, the rate of Aluminum chloride production decreases as the pressure increases. If the reaction rate is first order compared to the chlorine pressure (ie, speed = k [Pci] 1), the rate of production of aluminum chloride remains constant for increasing pressures. When the dependence of the reaction rate is greater than the first order (ie, speed = k [PC1 cil as is the case in the chlorination of aluminum at pressures above atmospheric pressure, the rate of production of aluminum chloride increases as

the pressure increases.

The rate of aluminum chloride production is affected by the rate of fluidization, that is, higher production rates are obtained at higher rates of fluidization. Apparently, a more efficient fluidization is obtained in the experiments carried out at speeds of 12.1-13.8 cm / second compared to the tests carried out at 6.1-8.3 cm / second. The fact that the rate of production of aluminum chloride is always (in all tests except one case) significantly higher than that obtained at atmospheric pressure, in spite of lower fluidization rates, clearly indicates that increased pressures (i.e., pressures greater than.

  3.04 bars) have a notable effect on the rate of production of aluminum chloride as indicated by the rate of

  production of said aluminum chloride.

Example 10-14

  A mixture of PCA (79.4% by weight), petroleum coke (19.9% by weight) and sodium chloride (0.7% by weight) is subjected to chlorination in the reactor system shown in Figure 3. Coke has been calcined

  at 8250C for 30 minutes and classified at - 210+ 105 microns.

Said coke has the following contents carbon: 96.8% by weight, hydrogen 0.79% by weight, nitrogen: 0.52% by weight, sulfur: 0.99% by weight, ash: 0.25% by weight and a specific surface of 10 m 2 / g. The alumina has the following values on analysis: Humidity 1.30% by weight Inflammation losses (PPI) 1.29 Alpha-alumina 0.60 Specific surface 98 m / g (BET) Grain size analysis + 149 microns - 2 % + 74 microns -72% + 44 microns - 96% Sodium chloride is added to the alumina to ensure the presence of a high level of NaAlCl4 as catalyst. The experimental conditions and the rate of production of aluminum chloride are shown in Table II. Figure 1 which contains the results presented in Table II, graphically represents the effect of the

  pressure on the production speed of aluminum chloride.

  The temperature influences the rate of production of aluminum chloride when the PCA is subjected to chlorination at these pressures, this temperature effect not however being significant enough to have a serious effect on the chlorination. The fact that the rate of production of aluminum chloride decreases as the temperature increases from 615 ° C to 720 ° C should not be taken as an indication that at high temperatures and pressures, the rate of production of aluminum chloride is lower

  to that which is obtained at atmospheric pressure.

Indeed, the rate of production of aluminum chloride is significantly higher at pressures above the

atmospheric pressure despite this temperature effect.

  One possible explanation for this temperature effect is that, for a mixture of coke and alumina, one stage of the reaction could be the adsorption of chlorine on carbon with the formation of a chlorinated carbon radical which then reacts with alumina. If a sorption were involved, the reaction rate would decrease as the temperature increased. There is another possibility that the reaction proceeds with the formation of an intermediate compound, such as carbon tetrachloride, which is

  less stable at higher temperature.

  Table II - Examples with Coke + partially calcined alumina Height of the bed = 0.70 m Experiment NX XI XII XIII XIV Temperature Pressure (C) (bars) 3.55 3.55 4.46 4.56 1.82 Speed (cm / s) 6.4 11.3 6.4 11.5 9.9 Conversion of

C12 (%)

  89.2 62.3 98.7 82.5 23.4 Production rate of aluminum chloride (kgmole / hour.m3.bar) 6.41 4.84 11.45 8.79 1.29 d-0 4: - Lo 0f Ln ru

Examples 15-21

A mixture of petroleum coke (19.9% by weight), alumina (79.4% by weight) and NaCl (0.7% by weight) is subjected to chlorination according to the procedure of Examples 1- 9. Coke has the following analysis values: Particle size fraction - 210 + 105 microns Carbon 96.80% Hydrogen 0.79% - Nitrogen 0.52% Sulfur 0.99% Ash 0.28% Specific surface 10 m2 / g ( BET) Alumina (MGA) has the following values on analysis Humidity (25 C-300 C) 1.96% Losses on ignition (PPI) e:

(1100 C) 1.28%

  Alpha phase content 18.00% Specific surface 54 m / g (B.E.T.) Particle size analysis + 149 microns - 4% + 74 microns - 56% + 44 microns - 76% The results appearing below in Table III are

visually represented in Figure 1.

There is an increase of approximately three times in the rate of production of aluminum chloride, by a

increase in reaction pressure from 1.92 to 3.44 bar.

The slower rate of increase in the production rate between 3.44 and 4.56 bars that can be seen in Figure 1, may be due to variations in the proportions of reactive alumina

and / or NaAlCl4 in the reactor bed. Another explanation

tion of this decrease in improvement in

production rate is the influence of the speed of fluidi-

sation on the reaction. In the two experiments at 3.55 bars, the speeds are greater than 10 cm / second, while the two experiments are carried out at 4.56 bars with the lowest production rates of aluminum chloride, in the presence of fluidization speeds between 6 and 7 cm / second. A more efficient fluidization of the beds with higher concentrations of sodium at speeds higher than 10 cm / second or the production and recycling of larger quantities of reactive dust at higher speeds, could also explain this leveling.

Table III

  Experience N XV XVI XVII XVIII XIX XX XX XXI Temperature Pressure Speed Conversion of (C) (bars) (cm / s) C12 (%) 1.92 1.92 3.44 3.67 4.66 4.56 4.46 9.9 9.5 11.4 12.5 6.1 6.6, 7 28.0 21.1 62.3 47.5 77.3 77.1 67.0 Production rate of aluminum chloride ( kg-mole / hour.m3. bars) 1.38 3.96, 25 3.46 3.96 4.06, 33 t'J r Ln Co Ln tu The experiments of Examples 22 to 25 are carried out in a reactor in fluid bed quartz at the scale of

laboratory, following a discontinuous mode of production.

Example 22 Two experiments are carried out using coked alumina (at 17.9% carbon) with an initial bed height of 0.3 m and with a surface speed of chlorine.

  8 cm / second, one at 5.07 bars and the other at 15.20 bars.

  The coked alumina has the following values on analysis: Alpha alumina 0.3% Na2O 0.23%

PPI (0-11000C) 19.5%

  Specific surface 8 m2 / g (B.E.E.T) At 5.07 bars and 7000C, the chlorine conversion is 97.2%, providing a chloride production rate

  of aluminum 28.7 kg-moles of aluminum chloride per m3.

  hour. bar. At 15.20 bars and 7150C, the chlorine conversion is 99.7% providing an aluminum chloride production rate of 51.4 kgmol of aluminum chloride per m3.hour.bar. Example 23 Two experiments were carried out using partially calcined alumina (79.4% by weight) as well as separate petroleum coke (19.9% by weight) and sodium chloride (0.7% by weight ), with an initial bed height of 0.27 m and a surface chlorine speed of 3 cm / second, one at 5.07 bars and the other at 10.13 bars. The alumina has the following values on analysis: Alpha alumina 0.4% Na O 0.35% Humidity (0-3000C) 1.11%

PPI (300-12000C) 1.02%

The coke used was calcined at 1250C for 30 minutes and classified at 210 + 105 microns. At 5.07 bars, and 6950C, the conversion of chlorine is 59.6% providing an aluminum chloride production rate of 7.59 kg-moles of aluminum chloride per m3.hour.bar. At 10.13 bars and 690 ° C, the chlorine conversion is 69.6%, providing an aluminum chloride production rate of 10.6 kg-moles of

  aluminum chloride per m3.hour.bar.

  Example 24 Four experiments were carried out using alumina (MGA) (79.4% by weight), sodium chloride (0.7% by weight), and separated petroleum coke (19.9% by weight ) at a surface chlorine speed of 8 cm / second at the rate of 2 to 5.07 bars and 2 to 10.13 bars. The alumina has the following values on analysis: Specific surface 59 m2 / gm (B.E.T.) Alpha-alumina 17.0% Na20 0.54% Humidity (0-300 C) 1.68%

PPI (300-1200 C) - 0.88%

  The coke was calcined at 825 C for 30 minutes and has the following values on analysis: Particle size fraction - 210 + 105 microns Specific surface 10 m2 / g (BET) Carbon - 96.8% Hydrogen 0.79% Nitrogen 0 , 52% Sulfur 0.99% Ash 0.28% One of the two experiments is carried out on chlorinated samples at 5.07 bars, with an initial bed height of 0.53 m and at 650 C. The conversion of the chlorine is 96%, providing an aluminum chloride production rate

15.1 kg-moles of aluminum chloride per m3.hour.bar.

The other test is carried out at 5.07 bars with an initial bed height of 0.56 met at 650 C. The conversion of chlorine is 97.2%, providing a production rate of aluminum chloride of 15, 99 kg-moles of aluminum chloride per m3.hour.bar. One of the experiments is carried out on the two chlorine samples at 10.13 bars, with an initial bed height of 0.53 m and at 665 C. The conversion of chlorine is 99.6% while providing a production rate of aluminum chloride 26.36 kg- moles of aluminum chloride per 3. m.hour.bar. For the other test carried out at 10.13 bars, an initial bed height of 0.56 m and a temperature of 655 C. is used. The chlorine conversion is 99.6% providing a production rate of chloride of aluminum

  24,87 moles of aluminum chloride per m.hour.bar.

Example 25

Alumina (80% by weight) and petroleum coke (20% by weight) (- 210 + 105 microns) are reacted at a chlorine surface speed of 8 cm / second, at 700 C, with a height of an initial bed of 0.58 m, respectively, at 07, 7.09, 9.12 and 13.17 bars. The alumina has the following values on analysis: Specific surface area 0.3 m2 / g (BET) Alpha-alumina 85.0% Beta-alumina 15.0% Na20 0.61% Humidity (0-300 C) 0 .25%

PPI (300-1200 C) 0.06%

  The coke has the same values on analysis as that of Example 21. Each test is ended once 90% of the theoretical chlorine has been sent. The results are represented graphically in Figure 2, in which the curve of the average production rate of

aluminum chloride as a function of pressure in bars.

  The average production rate of aluminum chloride is the arithmetic mean of all instantaneous readings

  readings, calculated assuming a constant bed volume.

  These comparisons show that the dependence between the speed and the pressure of chlorine is greater than one, because both the conversions of chlorine and the levels of

  production of aluminum chloride increase appreciably-

* as the pressure increases.

EXAMPLE 26 640 g of alumina placed on the market under the name Alcoa of A3 quality, containing 70% by weight of alphaalumine, are loaded into a reactor with 160 g of petroleum coke placed on the market under the name of Great Lakes Carbon 8250C,

  having a particle size range of - 210 + 105 microns.

  The reactor, which has a bed 0.54 m high, is maintained at a temperature of 7000C and under a pressure of 15.20 bars while a stream of chlorine is sent through the bed at the speed of 3, 02 kg / hour. After 35 minutes, analysis of the residue indicates that 90% of the initial charge of alumina has been chlorinated, this reveals the chlorination of practically all the alpha-alumina, that is to say that the mass transfer coefficient is not seriously affected by the temperature at

which the aluminous substance has been previously calcined.

Claims (10)

  1. Process for chlorinating an aluminous substance containing oxygen, consisting in chlorinating the aluminous substance in the presence of a reducing agent, characterized in that the chlorination is carried out under a pressure greater than approximately 3 bars.   2. Method according to claim 1, characterized in that the chlorination of the substance is carried out   aluminous at a temperature between 550'C and 800'C. 3. Method according to claim 1 or 2, characterized in that the reducing agent consists of coke, carbon, carbon monoxide, COC12 or CCl4. 4. Method according to any one of the claims   1 to 3, characterized in that the aluminous substance is alumina.   5. Method according to any one of the claims   above, characterized in that the oxygen-containing substance is calcined before being subjected to the chlorination.   6. Method according to any one of the claims   - previous, characterized in that the pressure is at least 5.07 bars, in particular at least 7.09 bars,   minimum of 9.12 bars, minimum of 10.13 bars, minimum   11.14 bar minimum, 13.17 bar minimum or minimum , 20 bars.   7. Method according to any one of the claims   1 to 5, characterized in that the pressure is equal to a maximum of 15.20 bars, and that it is in particular between 3.34 and 4.76, between 5.07 and 15.20, between 5.7 and 10.13 or between 10.13 and 15.20 bars. 8. Method according to any one of the claims   1 to 5, characterized in that the aluminous substance contains alpha-alumina (in particular at least at the rate of 0.1%) and that it is subjected to chlorination under pressure greater than 10, bars. 9. Method according to claim 8, characterized in that the pressure is within a range extending from a pressure greater than 10.13 bars up to   , 26 bars, in particular from 12.16 to 20.26 bars, and any part especially at around 15.20 bars.   10. Method according to claim 8 or 9, characterized in that the mass transfer coefficient is not seriously influenced by the temperature at which it has previously calcined the aluminous substance.