E. VARADI, V. I. TSVETKOVA, N. M. CHIRKOV

ON CATALYTICALLY ACTIVE PARTICLES IN THE SYSTEM TiCl$_4$—Et$_2$AlCl IN THE PROCESS OF ETHYLENE POLYMERIZATION

(Presented by Academician N. N. Semenov, June 12, 1963)

Deep polymerization of alpha-olefins is readily carried out on systems that include titanium trichloride or tetrachloride in combination with R$_3$Al or R$_2$AlCl. In the case of systems containing TiCl$_3$, the system is catalytically active from the moment the monomer is introduced; in the case of TiCl$_4$, however, the appearance of catalytically active particles is associated with the development of oxidation–reduction processes occurring over time. It has been established that in the reduction of TiCl$_4$ not only TiCl$_3$ is formed, but also solid products of more complex composition containing aluminum ($^1$, $^2$). The sharp difference in the molecular weights of the polymers obtained ($^3$) also characterizes the different nature of the catalytically active particles in these systems. The aim of the present investigation was to establish the nature of the catalytically active particles in the TiCl$_4$—Et$_2$AlCl system, present both in the solid phase and in solution. The most detailed study of the composition was carried out for two special cases: when the activity of the system reached its maximum value and when the activity practically did not change with time.

Fig. 1. Dependence of the initial rate of polymerization of C$_2$H$_4$ on the aging time of the TiCl$_4$—Et$_2$AlCl system without monomer

The experiments were carried out at a TiCl$_4$ concentration of 10 g/l (5.28·10$^{-2}$ mol/l), a molar ratio TiCl$_4$ : Et$_2$AlCl = 1 : 1, and 30°. The solvent was dry, spectroscopically pure n-heptane. The precipitates formed in the interaction of TiCl$_4$ and Et$_2$AlCl were isolated by filtration or centrifugation, washed repeatedly with n-heptane, and dried under vacuum.

The composition of the precipitate was determined by volumetric methods: chlorine by Folgard, titanium colorimetrically with H$_2$O$_2$ ($^4$), aluminum qualitatively with alizarin and quantitatively by complexometric titration ($^5$), and alkyl groups by the volume of gas evolved upon decomposition with alcohol. The composition of the gas was determined chromatographically. Colorimetric determination of titanium in the filtrates and of trivalent titanium in the hydrolysis products of the precipitates and filtrates was also carried out with the aid of FeCl$_3$ ($^6$). Polymerization experiments were performed in a metal apparatus. The reaction vessel, of volume 0.25 l, was equipped with a stuffing-box stirrer ($^7$). During the experiment, constant temperature and C$_2$H$_4$ pressure were maintained.

In order to establish how the activity of the system under study changes with time, experiments were carried out in which the monomer was introduced at different time intervals (from 0 to 6 hr) after mixing TiCl$_4$ and Et$_2$AlCl. The dependence of the initial rate of polymerization on the aging time of the system without monomer is presented in Fig. 1. The curve shown characterizes the change in activity of the system with time in the absence of monomer. Under the conditions studied, the maximum rate of polymerization (500 g C$_2$H$_4$/hr·l·atm) is reached 30–40 min after mixing the catalyst components. By 4 hr the activity decreases to 300 g C$_2$H$_4$/hr·l·atm and then remains practically constant. The intrinsic viscosity [$\eta$] of the polyethylene formed, when the aging time is increased from 0.5 to 4 hr, increases from 1.1 to 1.5·100 ml/g.

The amount and composition of the precipitate also change with time. In 30 min, about 1.5 g of precipitate (1) with a specific surface area of 28 m²/g is formed in 1 liter (determined by the BET method (^8)). By 4 h, the concentration of the precipitate reaches 6.4 g/l. The surface area of this precipitate (2) is 90 m²/g.

Table 1

Composition of precipitates formed after 0.5 h (precip. 1) and after 4 h (precip. 2) after mixing TiCl₄ and Et₂AlCl (molar ratio 1:1; TiCl₄ concentration 10 g/l; 30°C)

* After decomposition of the precipitate, all Ti is titrated as trivalent.
** The analysis was carried out by staff member of the Institute of Chemical Physics V. N. Likhashertova.

The data on the composition of the isolated precipitates are given in Table 1. Aluminum is absent from precipitate (1). The chlorine and titanium contents in this precipitate and the amount of gas evolved upon its decomposition (in the case of alcohol this is mainly ethane) correspond to EtTiCl₃. This is also confirmed by the results obtained in elemental analysis for C, H, and Cl. A qualitative reaction with Michler’s ketone (^9) indicates the presence of a Ti—C bond. Precipitate (1) and the filtrate after its separation give no EPR signal at room temperature. It is interesting that after decomposition of precipitate (1) with alcohol or with an aqueous solution of H₂SO₄, the titanium contained in this precipitate (which can be titrated with FeCl₃) as trivalent (^2) has a more complex composition, described by the general formula Ti₃Cl₆·AlEt (see Table 1), and gives an EPR signal at room temperature. Both precipitates are poorly soluble in hydrocarbons, but in the presence of TiCl₄ their solubility increases sharply (the solutions are colorless). For example, at 20°C the solubility of precipitate (1) increases from 8.5·10⁻⁵ to 4.5·10⁻² mol/l upon introduction into n-heptane of TiCl₄ in an amount of 8.4·10⁻² mol/l. This explains the presence in the filtrates of compounds, evidently EtTiCl₃, which hydrolyze with formation of Ti³⁺. Under the conditions studied, the content of these compounds in solution reaches 1.56·10⁻² mol/l by 0.5 h (with a total amount of Ti in solution of 4.56·10⁻² mol/l); by 2 h it decreases to 0.52·10⁻², and by 4 h to 0.15·10⁻² mol/l (the total amount of Ti in the filtrate then decreases correspondingly to 2.38·10⁻² and 0.18·10⁻² mol/l). These substances, with organoaluminum compounds, give colored complex compounds in the presence of TiCl₄ (red-yellow color of the filtrates).

Fig. 2. Kinetic curves of C₂H₄ polymerization in the presence of various catalysts and precipitate (1). \(P_{C_2H_4}=3.5\) atm.; 30°C. Solvent: n-heptane. 1 — filtrate after separation of (1), diluted 11.7-fold with n-heptane (Ti content 1.77 g/l); 1a — the same + (1) (1.2 g/l), 2 — precipitate (1) (1.2 g/l) without cocatalyst, 3 — Et₂AlCl (0.5 g/l) and (1) (1.2 g/l), 4 — Et₂AlCl (0.45 g/l), TiCl₄ (0.7 g/l) + (1) (1.2 g/l).

The filtrate remaining after separation of precipitate (1) and containing a significant amount of these compounds was diluted with n-heptane (11.7-fold) in order to slow further transformations and was tested in the polymerization process. It turned out (see Fig. 2, 1) that at the initial moment the filtrate is inactive, but its activity increases with time. The time during which

the maximum rate is reached depends on the pressure of $\mathrm{C_2H_4}$ and the temperature ($[\eta]$ of the polyethylene obtained is $0.4 \cdot 100$ ml/g). The precipitate (1) suspended in n-heptane, in the absence of cocatalyst, is active only for several minutes after the introduction of ethylene; the polymerization rate is low (Fig. 2, 2). Precipitate (1) is practically inactive also in the presence of $\mathrm{TiCl_4}$.

Fig. 3. Kinetic curves for the polymerization of $\mathrm{C_2H_4}$ in the presence of various cocatalysts and precipitate (2), $P_{\mathrm{C_2H_4}} = 3.5$ atm; 30 °C. Solvent: n-heptane. 1—$\mathrm{Et_2AlCl}$ (0.5 g/l) and (2) (1.5 g/l); 2—filtrate after separation of (1), diluted 11.7-fold with n-heptane, and precipitate (2) (precipitate concentration 1.5 g/l); 3—$\mathrm{Et_2AlCl}$ (0.45 g); $\mathrm{TiCl_4}$ (0.7 g/l) and (2) (1.5 g/l); 3a—the same without precipitate.

For the system precipitate (1)—$\mathrm{Et_2AlCl}$, the rate (see Fig. 2, 3) is at least 100 times lower than the value that could have been expected on the basis of the curve shown in Fig. 1, and the intrinsic viscosity is $[\eta] = 20 \cdot 100$ ml/g. Consequently, $\mathrm{Et_2AlCl}$, as well as $\mathrm{EtAlCl_2}$, which is still less active, is not the cocatalyst determining the rate of the process and the properties of the product in the $\mathrm{TiCl_4}$—$\mathrm{Et_2AlCl}$ system. Such cocatalysts are evidently complex compounds formed in solutions containing $\mathrm{EtTiCl_3}$, $\mathrm{TiCl_4}$, and $\mathrm{Et_2AlCl}$.

Indeed, if, instead of $\mathrm{Et_2AlCl}$, the filtrate after separation of precipitate (1), diluted with n-heptane, is used as cocatalyst, the polymerization rate increases sharply (Fig. 2, 1a), while $[\eta]$ of the polyethylene decreases to $1.7 \cdot 100$ ml/g. When this same filtrate is added to precipitate (2), which in the absence of cocatalyst is inactive, but with $\mathrm{Et_2AlCl}$ polymerizes $\mathrm{C_2H_4}$ at a rate of $14$ g/hour·atm·g, the rate increases approximately threefold (Fig. 3, 1 and 2). (With $\mathrm{Et_2AlCl}$, $[\eta]$—14.6; with filtrate—$2.4 \cdot 100$ ml/g.) From these data it is seen that the decrease in the activity of the $\mathrm{TiCl_4}$—$\mathrm{Et_2AlCl}$ system with time (Fig. 1) is due to a decrease in the concentration of active cocatalysts, which exist only in the presence of $\mathrm{TiCl_4}$ in solution. The addition of small amounts of $\mathrm{TiCl_4}$ to (1)—$\mathrm{Et_2AlCl}$ (Fig. 2, 4) and to (2)—$\mathrm{Et_2AlCl}$ (Fig. 3, 3) leads to a significant increase in the rate of the process. At the same time, $[\eta]$ of the product decreases, respectively, to 1.7 and $2.4 \cdot 100$ ml/g.

Evidently, in the presence of $\mathrm{TiCl_4}$, part of the precipitate dissolves and active complex cocatalysts are formed, in the presence of which lower-molecular-weight polyethylene is obtained. Probably, the sharp dependence of the rate of the process and of the molecular weight of the product on the concentration of cocatalysts and on the ratio between catalytic components, characteristic of systems containing $\mathrm{TiCl_4}$, $\mathrm{VCl_4}$ ($^{10,11,12}$), as well as the change in the molecular weight of the product formed with time ($^{3,13}$), are due to analogous phenomena.

Institute of Chemical Physics
Academy of Sciences of the USSR

Received
12 VI 1963

REFERENCES

1. G. Natta, J. Inorg. and Nucl. Chem., 8, No. 6 (1958).
2. A. Simon, L. Kovacs, L. Kollar, G. Dezsenyi, Monatshefte für Chem., 90, H. 4, 443 (1959).
3. H. Gaylord, G. Mark, Linear and Stereoregular Polymers, IL, 1962.
4. I. M. Kolthoff, E. B. Sandell, Quantitative Analysis, Moscow, 1948.
5. S. J. Ostvay, Magyar Kemiai Folyoirat, 60, No. 11331 (1954).
6. R. Khim., No. 23, 55347 (1955).
7. H. Martin, J. Stedefeder, Ann. Chem., B618, 1–3 (1958).
8. I. N. Meshkova, G. P. Belov, V. I. Tsvetkova, N. M. Chirkov, Plast. mater., No. 7 (1963).
9. S. Brunauer, Adsorption of Gases and Vapors, IL, 1948.
10. H. N. Fridlander, K. Oita, Ind. Eng. Chem., 49, 1885 (1957).
11. G. Natta, P. Pino, G. Mazzanti, P. Longi, Gaz. Chim. Ital., 87, 549 (1957).
12. A. G. Pozamantir, A. A. Korotkov, I. S. Lishanskii, High-Molecular Compounds, No. 8, 1207 (1959).
13. I. N. Meshkova, V. I. Tsvetkova, N. M. Chirkov, Izv. AN SSSR, OKhN, 1963.
14. E. J. Badin, J. Am. Chem. Soc., 80, 6545 (1958).