The invention is concerned with a method of regeneration of a hydrocarbon based liquid that contains titanium and chlorine compounds. Such hydrocarbon liquids are formed for example in the washing step in the preparation of Ziegler- Natta catalysts.

One of the steps in the preparation of Ziegler-Natta catalysts is often the treatment of MgCl *3EtOH carrier with a titanation reagent. Normally, a big excess of TiC is used as the titanation reagent. After the treatment, the excess of TiC is washed away. This washing is carried out with several successive washings with a suitable hydrocarbon as for example heptane. Big amounts of heptane have to be used to achieve a good washing result and several successive extractions. This will increase the volume of the hydrocarbon used in a normal synthesis of the catalysts to 200-300 litres per catalyst.

Because of the big amounts of the hydrocarbons a recirculation is necessary in all industrial applications. The purification of the hydrocarbon for recirculation is for the time being carried out by distillation. As the components are separated with a method that is based on the different boiling points of these, the boiling point of the hydrocarbon has to be 20°C lower than that of TiC , while the boiling point of the hydrocarbon should be as high as possible to achieve a washing that is as efficient as possible. In practice heptane is a compromise. The difference between the boiling points is 38°C.

Titanium compounds to be removed from the washing liquid is for example TiC and TiClgOEt. This removal requires extensive apparatuses. As even small amounts of TiC have to be removed, the capacity of the distillation unit is relatively low. It is difficult to get a hydrocarbon completely free from rests of TiCL. Also the energy consumption becomes high.

In spite of the sufficient difference between the boiling points of heptane and TiCL long distillation times, big devices and a low flowing rate are needed in the distillation. As the efficiency of the washing is proportional to the TiCL content of the circulated heptane, the distillation method has to be optimized to lower the resting TiCL content. This further decreases the capacity of the recirculation unit.

There is always formed a distillation rest in the distillation due to the slow condensation of the organic components in the boiling solutions. The rest, which still is very reactive, has to be removed, neutralized and separated. This distill¬ ation rest has become a serious waste problem due to its reactivity and consist¬ ence.

The object of this invention is therefore a better method of recirculation.

The method of the invention is mainly characterized in that titanium and chlorine compounds in a hydrocarbon solution are removed so that the hydrocar¬ bon solution is brought into contact with a silica that contains hydroxylic groups after which the silica is regenerated by means of water, for recovery.

The preferable embodiment forms of the invention have the characteristics of the subclaims.

The new recirculation technique is based on that the TiCL in the hydrocarbon solution is brought into contact with silica. The TiCL reacts with the hydroxylic groups on the silica surface and will thus be bound to the solid. By using elev¬ ated temperatures even three chlorine equivalents of TiCL can be reacted with the hydroxylic groups. The silica can be regenerated by adding water and be reused to purificate the hydrocarbon. The water amount needed is calculated on the basis of the amount of the formed HC1.

Reference is made to EP-publication 0 240 064 BI as being prior art wherein

there is presented a method for removing TiCL vapour from gaseous streams to a silica based metal oxide that contains hydroxylic groups (for example TiO _p ).

Reference is also made to US-patent 4 433 194 wherein silica has been used as a prestage of a purification process for cyclohexane carried out by distillation in the purification of liquid hydrocarbons. In this process the cyclohexane is purified from contaminants as isoparaffins, olefines, aromatics, compounds containing oxygen and sulphur and from water by adding 0,05-0,6% anhydrous titanium tetrachloride, whereas the titanium tetrachloride forms complexes with the contaminants. The solvent is filtered through silica to separate the complexes after which the filtrate is distillated in the presence of alkalimetal boron hydride and aluminium hydride.

Excellent results are achieved with the method of the invention. The TiCL value of the recovered hydrocarbon were below the observed limit. When an elevated temperature is used, the recovery is fast. The HC1 by-product is formed in pure form and can easily be absorbed as hydrous HC1. The solid SiOp-TiOCl mass is easily separated from the solution so that a clear liquid is left. The solid pro¬ duced is an inert free-flowing granular material.

This new circulation technique affords much advantages in view of both econ¬ omical considerations and device costs and service. The method does consider¬ ably lower the need of devices and a more pure end product is also achieved and at the same time an easily handled waste without any need of neutralization. The solid that is formed from Si0 ₂ *OTiCl*TiOp is completely inert and it can be handled in the open air. It does not react with water or oxygen in any way.

The essential feature in the circulation method of the invention is the use of a silica that contains hydroxylic groups as neutralization buffer for the TiCL in the hydrocarbon solvent and the regeneration of silica for recovery by adding water. The hydroxylic groups are achieved on the surface of the silica by hydrolysis:

Si0 ₂ + H ₂ 0 = OSi(OH) ₂ (1)

The hydrocarbon solution (for example a heptane solution) is brought into contact with hydrolized silica. As the hydroxy groups are somewhat protected in 5 the silica there exists a slow neutralization of TiCL takes place:

OSi(OH) ₂ + 2 TiCl ₄ = ClgTi-O-SiO-O-TiC^ + HC1 (2)

As the reaction product in equation (2) still is a reactive component, the reaction 10 can proceed:

-SiO-O-TiClg + HO-SiO- = -SiO-0-TiCl ₂ -0-SiO- + HC1 (3)

The following reaction step is probably slower than the two frist reactions: 15

(-OSi-0-)2-TiCl ₂ + HO-SiO- = (-OSi-0-)-2-TiCl-0-SiO- + HC1 (4)

The last stage of the neutralization reaction does probably not exist in a quanti¬ tative extent: 0

(-OSi-0-)3-TiCl + HO-SiO- = (-OSi-0-)3-Ti-0-SiO- + HC1 (5)

Hydrogen chloride is liberated from the system. This gaseous acid can either be neutralized or absorbed in the water solution: 5

Hα + 2H ₂ 0 = HC1*2H ₂ 0 (6)

In the latter case a 40% HC1 water solution is achieved as a by-product which is a reagent that easily can be used anywhere. (f

The above reactions can be carried out in temperatures of 20-90°C. However, elevated temperatures are preferably used. A suitable temperature is 60-90°C.

The reacted silica can easily be activated by repeating the hydrolysis (1). The water amount needed can easily be calculated from the amount of the HC1 formed in the neutralization method.

Such an amount of water is preferably added that is enough to substitute the hydroxylic groups consummed for the contact of the hydrocarbon solution and silica. The thus regenerated silica can be reused several times.

In the following the invention is further described by means of the following non- limiting examples.

EXAMPLES

The preparation of the rest solution

The washing solution from the synthethis of the catalyst contains in average 5- 6% TiCL. In these tests this solution was made by adding 20 ml TiCL to one litre heptane. This solution was used in all tests as a standard rest solution.

To study the repeated neutralization capacity of the silica a new TiCL dosage was added after each neutralization.

The neutralization agent

Crossfield's EP10 silica was used as washing agent. According to the product declaration this material contained 5% damp. This amount considered to be sufficient to carry out the first neutralization. After each neutralization the consumption of the hydroxylic groups was calculated from the amount of the formed HC1. Thereafter a stoichiometric amount of water was added to com- pensate for the consummed hydroxylic groups in the neutralization. 100 g silica was used as neutralization agent in these tests.

The test apparatus

The test apparatus used in all tests is presented in Fig. 1. A heptane-silica slurry (0) was placed in a 2 litre reactor (A) that was equipped with a thermometer (T), a mixing device (H), a theπnostated oil bath (B), a reflux unit (E) and an inlet for addition of TiCL (I). The whole unit was kept in inert conditions. The reactor was coupled through the refluxing condenser and oil seal (K) with a pipe line (P) to the neutralization unit (M). A water solution was used as adsorption medium (Z) that contained phenol-red as indicator. 1M natriumhydroxide (N) was used as titrator.

The used temperatures

The reaction was first carried out in room temperature. After this the reaction was continued in the temperatures of 40°C, 60°C and 80°C. The last tested temperature (90°C) was near the boiling point of heptane (98°C).

The circulation

The circulation was carried out in the following way: 20 ml TiCL was added to the silica slurry while agitating. The temperature was increased from room temperature to 90°C. The formed HC1 was continuously titrated by NaOH. When no development of HC1 gas could be observed any more, the reaction solution was let to cool to the temperature ca 40°C. The agitation was stopped and the silica let to set. When the sohds and the solution was separated, samples were taken from the solution for determination of the titanium and the humidity. The agitation was started again and an equivalent amount of water was added to compensate for the consummed hydroxylic groups of silica. The silica was hydrolized until the moisture content of the solvent was below 10 ppm. There- after the whole circulation could be started again.

The results and the evaluation of these

The hydrolysis of silica

Fig. 2 presents the hydrolysis rate of the used silica. 6,4 ml water has been added to the heptane-silica slurry. 85% of the water had been bound to the silica during the first hour of the agitation. During the next hour the resting 15% is consummed. Only moisture rests can be observed in the heptane solution after this. After six hours' agitation there is only 3 ppm moisture left.

The reaction depending on the temperature

The reaction between the hydrolysed silica and TiCL appeared to be dependent on the temperature. Fig. 3 presents the neutralization curve when the tempera- ture is increasing step wise. After each increase of the temperature a new stability value is observed, that corresponds to a given molar composition of the titanium compound. The intermediate titanium compound can be expressed with the formula Cl(4-x)*Ti*(-0-Si)x. The titration result in 90°C was not complete in the first test drive. Even consumptions of 537 ml NaOH was true in the pro- longed tests. In spite of this, the ratio between the reaction temperature and the amount of chlorine atoms included in the neutralized TiCL is linear. The results are collected in Table 1 and they are graphically presented in Fig. 4.

Table 1. The consumption of IM NaOH and the corresponding calculated amount of those chlorine atoms that are neutralized in 20 ml TiCL, (x) as a function of the reaction temperature.

Temperature (°C) ml NaOH

0 20 40 60 80 90

The titanium content of the heptane

Samples were taken in the end of each neutralization. The titanium content was determined from these samples to follow up the decrease of this contaminant. It appeared that alreadey in 20°C, more than 80% of the TiCL was absorbed in the solid mass even if it had not been reacted. In 60°C there was only 2,5% TiCL left in the solution. No amounts of TiCL could be observed in 80°C and 90°C. The results showed that at least two chlorine atoms of TiCL have to be neutralized so that they completely could pure the heptane solution from TiCL .

The results are collected in Table 2 and they are graphically presented in Fig. 5.

Table 2. The Ti-content of the heptane solutions and the corresponding amount of the total amount of TiCL calculated as a percentage which TiCl ₄ has been added as a function of the reaction temperature.

Neutralization rate

The neutralization rate of TiCL was high. In fact, it was depending on the rate of the increase of the temperature. In Fig. 6 there is shown a typical neutraliz- ation curve. This curve is almost identical with the temperature gradient curve of the test. This means that if a hot heptane-TiCL solution is added to a preheated silica slurry, an almost immediate neutralization reaction takes place. On the other hand, the rate of the reaction can completely be regulated by using a temperature gradient.

The purity of the recovered heptane

Fig. 7 presents a gas chromatogram of the recovered heptane solution. This solution has been recovered 6 times. The purity of the heptane corresponds to the purity of the original solution (Merck Art. 4365). New contaminant rests could not be observed.

Moisture in the heptane solution

A quantitative absorption of the water added in the recovery of the silica can be achieved if the operation is carried out in room temperature. This absorption is 5 also depending on the temperature, being less quantitative in higher tempera¬ tures. Much longer hydrolysis times are then needed. The resting moisture contents in Table 3 have been listed in five regeneration tests as a function of both the temperature and the hydrolysis time.

10 Table 3. The moisture content in the recovered heptane-silica slurry as a function of the temperature and the hydrolysis time.

Regneration Added Temperature Hydrolysis Resting free

15 H ₂ 0 (ml) (°C) time (min) HgO (ppm)

1 2 3 0 4 5

A complete absorption of the free water in the hydrolysis method had a great 5 importance, because a free water amount tends to exist in the final purified heptane, also after the reaction with TiCL . TiCL tends to react with the hydroxy groups adsorbed on silica by first leaving the non-reactive free water. This free water then goes together with the recovered heptane back to the process and loads the purification systems in the fed line. This phenomena that TiCL reacts (5 ^" intensively with the hydrolysed silica but leaves the free water unchanged gives the impression that silica reminds one of a heterogenic catalyst t-hat facilitates the neutralization reaction that otherwise would not proceed at all in low

consentrations.

The recovery test drives

Six recoveries were carried out in the laboratory. The results have been listed in Table 4. Depending on the fact that the work for better reaction conditions still continued during the test there is unregularities in the results both in the consumption of NaOH and the final moisture content of the heptane. More than three chlorine atoms in the TiCL could be neutralized in these conditions.

Table 4. Recovery of the heptane-TiCl ₄ solution. Consumption of NaOH, the calculated amount of the HC1 equivalents, the titanium content after the reaction and the remaining water amount.

Recovery

460 0.460 2.6

2 234 3 584 4 591 5 593 6 560