Search

To clarify the stereospecificity of Natta-type catalysts, the temperature dependence of the isotacticity of a polypropylene has been investigated by using H-type or A-type TiCl3 with Al(C2H5)3 or Al(C2H5)2Cl in a study of the kinetic behavior of the polymerization under the conditions T = 30–75°C., P ≈ 1 atm., and [Al]/[Ti] = 1. With the use of Al(C2H5)3 the isotacticity increases with a rise in polymerization temperature, and it approaches that of Al(C2H5)2Cl, which is almost independent of the polymerization temperature. On the other hand, with the use of Al(C2H5)3 the number of the polymerization centers decreases with a rise in temperature, while that with Al(C2H5)2Cl remains unchanged. To explain these results, two kinds of polymerization center have been proposed; one is stable at any temperature and produces a higher-isotacticity polymer, and the other is stable at the higher temperatures and produces a lowerisotacticity polymer. Some solvent effect of the isotacticity has been observed with the use of A-TiCl3. The isotacticity with A-TiCl3 decreases by several per cent when n-heptane is used instead of toluene. This solvent effect has been explained on the basis of the existence of a considerable amount of AlCl3 in A-TiCl3.

The kinetics of the propylene polymerization with TiCl3-AlEt3 has been investigated in the temperature range (30-70°C.), the pressure range of propylene (100-700 mm. Hg), and the concentration range of the catalysts (2 g. TiCl3/liter, [A]/[Ti] = 0.4-3.0). The polymerization rate R increases rapidly with the polymerization time t at first (stage I), which is followed by a gradual decrease to the stationary value Roo, (stage II). The kinetic behavior in stage I in the case when propylene is introduced after the addition of AlEt3 is expressed by and that in the case when propylene is introduced before the addition of AlEt3 is expressed by The kinetic behavior in stage II, which is independent of the above procedure, is expressed by where R∞ is the stationary rate represented by and Ro a constant. The activation energies are E1 = 15, E2 = 17, Et3 = 2-3, E3″ = 12, and EP = 10 kcal./mole, respectively. The heat of adsorption of AlEt3 was obtained as 12 kcal./ mole from the temperature dependence of K. Stage I is explained on the basis of the formation of polymerization centers from Al2Et6 and a propylene monomer on the surface of TiCl3. Stage II is attributed to an irreversible deactivation of the polymerization centers localized on unstable surface sites due mainly to a spontaneous process and partially to a propylene monomer (or polymer) attack.

An explanation of the time-dependent log-normal molecular weight distribution of polyolefins in heterogeneous Ziegler polymerization is proposed on the basis of the non-uniformity of surface sites concerning the activation energy of the propagation reaction. It is a theoretical derivation of the Wesslau equation from the usual kinetic model. The theory explains the behavior of the molecular weight distribution, i.e., time and temperature dependences of polydispersity as well as large values of polydispersity in the early stages of polymerization, which have been observed experimentally.

Thermal analysis in combination with infrared spectroscopy was applied to investigate the states of ethylbenzoate (EB) and TiCl4 in the simple type of the MgCl2-supported high-yield catalysts. A further attempt was made to evaluate kinetic parameters in the initial stage of propylene polymerization (< 1s) by a stopped flow method. The catalysts employed in the present study were prepared by co-grinding MgCl2 and TiCl4 · EB complex. Thermal analysis and infrared spectroscopy showed that TiCl4 · EB complex decomposes during co-grinding process. It was also indicated that EB, thus produced, has strong interaction with MgCl2. This was consistent with observations made by infrared spectroscopy. It was deduced that EB and TiCl4 are supported at separate sites of MgCl2 in the co-ground catalyst. Kinetic parameters including rate constants of propagation (kp) and transfer (ktr), and concentration of polymerization centers ([C*]) were determined. The values of kp and ktr remained almost constant throughout the range of co-grinding time, but [C*] increased with time of co-grinding. No significant changes in the value of ktr were observed. Accordingly, the improvement of catalyst activity was attributed to an increase in [C*] rather than kp. These results lead to the conclusion that the complex decomposes during the co-grinding to yield EB and TiCl4 which interact only with MgCl2.

The theory of slow monomer diffusion to catalyst site through polymer matrix explains broad molecular weight distribution(MWD) of polymer produced with heterogenous Ziegler catalysts as well as rate-decay during polymerization with hydrogen dominant transfer can not be applied for polymerization with monometer dominant transfer, which is an intrinsic defect of the theory. The theory should predict that the narrowest MWD(Mw/Mn =2) for the latter case and reduction of the monomer contribution by changing concentration of monomer or by hydrogen-addition broaden MWD, which gives the basis of useful method to evaluate the theory itself. The accumulated experimental results of propene polymerizations with traditional and supported catalysts refute the above prediction.

Many rival theories proposed for the origin of the broad MWDs of polymers produced with heterogeneous catalysts still persist. As an absolute test for them, the importance of the transitional behavior of MWDs during the quasi-living stage of polymerization is emphasized. Based upon the different mechanisms, all the rival theories have been constructed so as to explain the broad MWDs at the stationary state. Therefore, all competing rate processes are balanced in the stationary state, and any specification of the kinetic mechanism can not be tested by the stationary MWDs. However, for the transitional MWD from the living stage to the stationaiy state, all the rival theories predict respective behavior in response to their different mechanisms. The observed transitional MWDs show that MWDs in the living stages are always broad and then mostly narrow or remain unchanged for which all the rival theories are invalid. To understand this situation, a new theory of non-uniform sites not only with propagation but also slightly with transfer rate is proposed. The non-uniformity that conforms to a popular kinetic model gives a unified explanation about the abnormal decay kinetics recognized in many systems as well as that of the supported Ziegler catalyst.

Publisher Summary The kinetic behavior of the polymerization in a slurry system is so complex that the usual kinetic analysis cannot be applied. The kinetic behavior is supposed to be suffering from the rapid rate decay of the polymerization, the chapter attempts to develop some new method that is applicable for observing kinetic behavior of the polymerization free from any rate decay or those, at least, with a negligibly small rate decay. A stopped flow reactor was useful to carry out the polymerization for a short time, such as 0.03–1s at room temperature, where the polymerizations were those of quasi-polymerization with constant rates. From the results of the quasi-living stages of the polymerizations, one can determine precise values of rate constants of propagation and transfer reaction as well as concentrations of polymerization centers. In addition, the observation provides some important key points to understand the kinetic data of the slurry polymerization together with those reported by Giannini and Hsu et al., based on a unified kinetic model, the description of which is the purpose of this chapter.

Publisher Summary This chapter points out that the behavior of molecular weight distributions (MWD) of polymers produced at the initial stages of polymerization, plays an important role for the mechanistic discussion of the broadness of MWDs. Supposing initial stages of polymerization where the yield and molecular sizes of produced polymers are very small, the original catalyst surface may remain unchanged and the diffusion control theory should predict a narrow MWD, because of the absence of any room to take the “multi-grain model” or “core model” of the diffusion control theory. The chain-length theory predicts the same conclusion. The nonuniform centers theory, however, predicts broad MWDs even for the initial stages of polymerization. Such initial polymerization stages appropriated for this criterion must belong to the quasi-living polymerization stage of any heterogeneous polymerization, where the average molecular size increases from a small and time-proportional value to a large stationary value. A few MWDs during quasi-living stages of heterogeneous Ziegler–Natta polymerization have been reported. The time dependency of MWDs reported are not unique but scatter, from a broader to a narrow, or vice versa, and down and up to stationary. Complex changes are assumed on the basis of changes in the number of some kinds of active centers. Thus, meaningful studies on MWDs in the course of quasi-living polymerization must be carried out together with observation of the kinetic behaviors of polymerization themselves. In this aspect, the experimental results described in the chapter are useful.

Dependences of the molecular weight distribution and stereochemical regulation of the polypropylenes produced with VCl4–AlEt2Cl catalyst on the polymerization temperature were examined. The molecular weight distributions of the polymers obtained at temperatures below −40°C were unimodal and narrow (Mw/Mn ≤ 2). The molecular weight distributions obtained at higher temperatures (above −21°C) were bimodal with one narrow distribution and one wide one (Mw/Mn > 2), and the polymer fraction of the wide distribution increased with the polymerization temperature. The fractional amount of (CH2)2 groups in the polymers, which corresponds to tail-to-tail linkage of two propylene units, increased to a maximum at −21°C followed by a gradual decrease with the polymerization temperature. The production of isotactic polymers was confirmed at temperatures above −21°C. From these data, it is concluded that only the homogeneous form of the catalyst system is responsible for the polymerization at temperatures below about −21°C while the heterogeneous form appears and catalyzes the polymerization together with the homogeneous one at temperatures above −21°C.

Vanadium tetrachloride combined with alkylaluminium could be shown to produce two types of polypropylenes, the syndiotactic and the isotactic ones, depending on both the kind of alkylaluminium component and polymerization temperature. It was found that the syndiotactic polypropylenes always have narrow molecular weight distributions (1,5≤Mw/Mn≤2,5), whereas the isotactic polypropylenes always have wide distributions (16≤Mw/Mn≤19), suggesting that the isotactic polymerizations are not induced by homogeneous but by heterogeneous catalysis. With the aid of ESR analysis for catalyst systems it was concluded that the syndiotactic polymerizations take place in a soluble complex of trivalent vanadium, whereas the isotactic polymerizations proceed on the insoluble vanadium dichlorides formed from the catalyst mixture.

The changes in the polymerization rates induced by the elimination or replacement of soluble organometallic co-catalysts have been examined in the course of the polymerization of propene initiated with titanium trichloride and various types of organometallic co-catalysts. The elimination of an organometallic co-catalyst, A, in the liquid phase during the polymerization resulted in a slow deactivation of the catalysts by the attack of a propene monomer. In contrast, the replacement of an initial organometallic co-catalyst, A, by a different co-catalyst, B, during the polymerization caused a rapid change in the polymerization rate from the specific rate with TiCl3 and A to that with TiCl3 and B. These results have been interpreted in terms of the bimetallic mechanism that an organometallic co-catalyst coordinates to a titanium ion, in an exchangeable form with the co-catalysts in the liquid phase, and regulates the catalytic activity.

A new method, viz. stopped flow polymerization, has been employed to observe the quasi-living polymerization state of propene. It was found that the molecular weight distribution was maintained constant from the very start of the polymerization process ( i.e . after 0.1 s) and that the concentration of Al(C 2 H 5 ) 3 used as well as method of catalyst preparation influenced some of the kinetic parameters.

The effects of temperature and catalyst homogeneity on the molecular weight distribution (MWD) and stereochemical regulation of polypropylenes produced by Ti(OC4H9)4Al2(C2H5)3Cl3 system have been investigated. The MWD of polymers obtained at temperatures below 21°C were unimodal and narrow (MwMn⩽2.0), whereas those obtained at temperatures higher than 31°C were bimodal with one narrow distribution and the other broad one (MwMn=18) at higher molecular weights. The existence of two different types of catalyst, one soluble with homogeneous catalytic centres and the other insoluble with heterogeneous catalytic centres was found in the polymerization at 41°C. At temperatures below 21°C only soluble catalyst was present and produced isotactic polypropylenes with [m]=0.65. The isospecific nature of soluble titanium-based catalyst is greatly contrasted to the syndiospecific nature of soluble vanadium-based catalyst.

Tetrabenzylzirconium (ZrBz4) was synthesized and used as catalyst for the polymerization of propylene in toluene in the temperature range from 0°C to 65°C. The isotacticity of the produced polymer was analysed (by IR and NMR spectroscopy) as well as the molecular weight distribution (by gel permeation chromatography). No colloidal particles could be usually observed after the polymerization, but both the isotacticity and the Q value (Mw/Mn) calculated from the molecular weight distribution curve increased with increasing polymerization period. The Q value was estimated always far greater than 2, which would be obtained in the case of homogeneous polymerization. From these results, it was concluded that isotactic polymer derives from polymerization on small (invisible) colloidal particles formed by decomposition of ZrBz4 during the course of polymerization.

The soluble V(acac)3-Al(i-C4H9)2Cl system initiated living polymerization of propene at −78°C affording monodisperse polymers ( M w M n = 1.15 ± 0.10 ). A kinetic study (of the living polymerization) was carried out to evaluate the rate coefficients for propagation. The equilibrium constant KM for a propene monomer coordinated to an active vanadium and the rate constant kp for a subsequent insertion of coordinated monomer into a living polymer chain were determined and compared with the values for the polymerization of propene with other soluble vanadium-based catalyst systems. The relation between KM and kp revealed that a strong interaction between vanadium and propene is unfavourable for the insertion of the coordinated propene into a living polymer chain. The mechanism of an initiation reaction involving alkylation and complexing of V(acac)3 with Al(i-C4H9)2Cl has been proposed.

A kinetic study has been carried out on the low-temperature polymerization of propene with the soluble catalyst system, VCl4 and alkylaluminium (AlEt2Cl, AlEt2Br, or AlEt3). The rate coefficients of the elementary reactions were evaluated at −78 °C. The molecular weight distribution function was derived on the basis of the polymerization mechanism. The theoretical distribution curves and experimental distribution curves obtained by use of a gel permeation chromatography agree, supporting the mechanism proposed for the coordination polymerization.

The kinetic features of the molecular weight distributions (MWDs) of polypropylene produced with TiCl4/MgCl4/C6H5COOC2H5/Al(C2H5)3 as catalyst are investigated as functions of the polymerization time and the concentrations of Al(C2H5)3, C6H5COOC2H5 and hydrogen. The MWDs are dependent upon the concentrations of Al(C2H5)3 and C6H5COOC2H5, and are correlated with a change of the isotactic index of the produced polymers. In contrast, both MWD and the number-average molecular weight Mn remain unchanged during the polymerization from 5 s to 3 h. When hydrogen is added as a chain transfer reagent, Mn decreases as anticipated but the polydispersity index (Mw/Mn) is independent of the hydrogen concentration. The MWDs obtained with this supported catalyst are rather narrow (3≤ Mw /Mn ≤8)and are represented by a log-normal form or Wesslau equation, which is different from the MWDs (6≤ Mw /Mn ≤20) obtained with the traditional TiCl3 catalysts, which are represented by the Tung equation. Broadening of the MWD obtained with heterogeneous catalysts is also discussed. The constancy of Mw/Mn, independent of the hydrogen concentration leads to the important conclusion that non-uniform surface sites on heterogeneous catalysts are responsible for the broadening of the MWD. A surface heterogeneity of the propagation rate constant, kp, is demonstrated on the basis of the kinetic data, concerning the effect of CO-inhibition on the polymerization rate. A surface distribution function of kp, capable of accounting for the broad MWDs, is presented and discussed in terms of a distribution of the activation energy for the propagation step.

AA型TiCl3-Al(C2H5)3系触媒上でのプロピレン液相重合における酸素の添加効果を,速度論的挙動および生成ポリマーの立体規則性の両面から検討した。{O2+Al(C2H5)3}+TiCl3の順序で酸素を添加した場合には,酸素がAl(C2H5)3と不可逆的に反応し,重合に不活性な錯体を生成する。このために溶液中のAl(C2H5)3濃度が減少し,重合速度低下と生成ポリマーの立体規則性の向上が起こる。{O2+TiCl3}+Al(C2H5)3の順序で添加した場合は,酸素がTiCl3を酸化しTi4+を生成する。4価のチタンは,Al(C2H5)3により還元され低活性のβ型TiCl8とAl(C2H5)2Clを生成する。この結果,重合速度の低下が起こる。さらに,生成Al(C2H5)2ClとAA型TiCl3とによる立体規則性向上の効果と,生成β型TiCl3とアルキルアルミニウムとの系による立体規則性低下の効果とが重なり,立体規則性に極大が現われる。

Mass transfer which affects the rate of propylene polymerization with titanium trichloride–triethylaluminum, has been evaluated by use of a new method developed for this heterogeneous reaction. The polymerization was carried out with the usual flask reactor equipped with a paddle stirrer; the rate of gas absorption into the polymerization slurry was proportional to stirring speed and the reciprocal of the total amount of polymers produced. It has been confirmed that the polymerization rate separated from the absorption rate is purely kinetic (propagation), and an effective physical process, such as monomer transfer through a polymer film covering the catalyst surface, no longer exists.

Using catalyst systems, TiC13-A1Et3 and TiC13-AIEt2Cl, polymerizations of ethylene were carried out in gas phase. The effects of hydrogen on the polymerization rate and on the average molecular weight of produced polymers have been investigated in connection with the kinetic behaviors of polymerization. The kinetic behaviors of polymerization are similar to those of the propylene polymerizations with the same catalyst systems. The effects of hydrogen on the rate of polymerization with these catalysts were different; a small but clear enhancement of the rate in the region of small amounts of hydrogen added was observed in the case with TiCl3-AIEt2Cl, while a monotonic depression of the rate with increasing amount of hydrogen added was observed with TiC13-AIEt3. The stationary polymerization rate in the latter case is expressed by, R0RH=------1+ PnH2and the average molecular weight is expressed byMOVMHV=------1+ PmH2with a=a6exp(-11 kca1/RT), b=b0exp(-5kcal/RT), n=O.6 and m=0.6. In these equations, superscripts H and O designate the values in the presence and absence of hydrogen, respectively. The latter expression can be also applied to the polymerization with TiCl3-AIEt2Cl, providing that m=O.5.The above results have been understood on the basis of supposition that dissociative hydrogen atoms on the catalyst surface are participating in the transfer of growing chains.

The influence of SeOCl2 on the polymerization of propylene by TiCl3–Al(C2H5)3, and the temperature dependence of the stereospecificity of the catalyst, TiCl3–Al(C2H5)3, have been investigated. SeOCl2 decreases the rate of polymerization and increase the stereospecificity of the catalyst, which could be explained on the basis of a decrease of the concentration of Al(C2H5)3 accompanied by a reaction between Al(C2H5)3 and SeOCl2. On the other hand, the stereospecificity of the catalyst, TiCl3–Al(C2H5)3, increases gradually with a decrease in polymerization temperature from 40 to 0°C. From these results, we conclude that SeOCl2 exerts no essential influence on the polymerization of propylene by TiCl3–Al(C2H5)3, and that the stereospecificity of the catalyst is attributed mainly to the reducing ability of the organometallic compound.

Propylene polymerization over TiCl3-AlEt3 catalyst in the absence of solvent have been carried out at various values of propylene pressure, P, molar ratio of AlEt3 to TiCl3, (A)and polymerization temperature. The kinetic curves(polymerization rate vs. time)are lower than those in the case with solvent and stationary has been represented by R R (K(A)jp K(A)P), were R is the stati nary rate in the case with solvent. The rate constant and activation energy(12 kca1 mo1)of propagation and the concentration of polymerizatio centers(1 x 1 2mollmo1-TiCI8)are same with those in the case with solvent. B t the average lifetimes of the growing chains(24 min)were shorter than those(40 min)in the case with solvent.

The effets of the addition of methyl methacrylate (MMA), methyl acrylate (MA) and acrylonitrile (AN) on the catalytic activity of TiC13-AIEtC12 for propylene polymerization have been studied. TiCls-AIEtC12 was inactive without additives. On the addition of MMA or MA to the system highly stereoregular polypropylene (1.1.=94N96%e) was obtained, although the activity was lower and less stable than that ef TiCls-AIEt2Cl. The system with AN was inactive for propylene polymerization.On the basis of the kinetics of propylene polymerization with TiCls+{AIEtC12 MMA (or MA)}, it has been conc1uded that a complex formed from AIEtC12 and MMA (or MA), AIEtC12. MMA (or A12Et2Cl, MA), acts as an active cocatalyst for TiC13, while free MMA (or MA)acts as a poison for TiCls-surface. The active complexes evolved considerable amount of ethane and ethylene when they were mixed with TiCls at 830C, but inactive ones, AIEtC12 and AIEtC12, AN, did not. This result suggests that MMA and MA activate Al-Et bond. The polymerization activities of these complexes have been discussed in connection with NMR data of these additives.

前報で4 種の三塩化チタンのうち, アルミニウム還元によって製造された三塩化チタンを, 活性化したと称するもの(以下ARAと略称する)のみがプロピレン低分子重合活性を持つことを報告した。ARA中に少量含有されている塩化アルミニウム(塩化アルミニウムとして25mol%)の役割を明らかにするために,塩化アルミニウムに対するプロピレンの挙動,および塩化アルミニウムを含まない三塩化チタン,すなわち水素還元によって製造された三塩化チタンを活性化したと称するもの(以下HRAと略称する)への塩化アルミニウム添加効果をしらべた。その結果より,ARAの低分子重合活性は塩化アルミニウムの助触媒効果によるものであることを結論した。さらに,HRA,塩化アルミニウム,およびHRAと塩化アルミニウムの磨砕混合物について,電子スピン共鳴吸収を観測し,磨砕混合物はARAと同じ吸収シグナルを与えることを見出した。これにより上記助触媒効果は,塩化アルミニウムによるHRAのARA化であろうと推論した。

not available

Natta反応の機作を解明してゆく手始めとして,その固体触媒である三塩化チタンに対するプロピレンの挙動をしらべる実験を行なった。その結果,プロピレンは三塩化チタンに対して単に吸着しているだけでなく,10量体程度の低分子重合を起こすことを見い出した。なお,低分子重合活性はアルミニウム還元による三塩化チタンが特に大であった。初期反応速度は,小泉が酸化ニッケルケイソウ土上でのプロピレンの重合について提出した初期式にあてはまり,活性化エネルギーは0℃～40℃で約1kcal/mol,40℃～80℃で5～6kcal/molであった。

The isotacticities of the polypropylenes produced with TiCl3–AlEt3 catalyst without solvent were 6∼7% higher than those obtained in the case with n-heptane solvent. In the case of TiCl3–AlEt2Cl cat...

Electron donor compounds employed in Ziegler-Natta catalysts have various effects on the catalyst performance. For propylene polymerization, the most important role of an electron donor is to improve stereospecificity of the catalyst. Ever since a MgCl2-supported high yield catalyst was developed by Shell International Research (Hewett, et al. 1964) and improved to a great extent by Montedison (Giannini, et al. 1977) and Mitsui Petrochemical (Kashiwa, et al. 1977), many types of the high yield catalysts as well as the electron donor compounds have been proposed and studied in both academic and industrial field (for example, Chien, et al. 1976).

Although our present fundamental understanding of Ziegler-Natta catalysis is one of the most advanced among the commercial catalysts, it is still far from perfect. Here, the kinetic approaches are critically summarized on the basis of a three-stage methodological classification system: characterization, phenomenological formalism and mechanistic approach. Under characterization the present conception of stereoregularity is criticized. In the second stage the proposed rate equations are discussed in light of experimental precision and kinetic models. The contradictions of our models with the experimental data found in studies on the effect of hydrogen on polymerization are pointed out, and the theories proposed for molecular weight distributions are discussed. Under mechanistic approach, the confusion in defining polymerization centers are pointed out. The definition suffers from both the definition in radical polymerizations and experimental operations. Reconsideration reveals some problems which remain unsolved and these are listed herein. Finally, an attempt is made to explain the effect of hydrogen on both polymerization rate and molecular weight.

The most reasonable method for determining values of the rate constants of propagation, transfer and the concentration of polymerization centers, kp, ktr and C*, is that based upon the observation of both transitional changes of number average degree of polymerization \({{\bar{P}}_{n}}\) and polymer yield Y in the initial stages of polymerization. This method has been used for long time, however, the results reported before the development of GPC are not so precise because of the use of viscosity average molecular weight in place of number average molecular weight. In addition, the use of viscosity average molecular weight in place of number average molecular weight may be possible only in the case of constant polydispersity independent of polymerization time, as pointed out by the present author (Keii 1986). Furthermore some limitations in the application of the method should be noted. For example, the method can not be applied for polymerizations of rapid transfer reaction. The applicability of the method is determined by value of the mean lifetime of growing chain, τ, i.e. the reciprocal of transfer rate constant, 1/ktr, as below.

A soluble V (acac)3/Al(C2H5)2C1 system was found to polymerize propylene to give a syndiotactic “living” polypropylene having narrow molecular weight distribution \(({\bar M_w}/{\bar M_n} = 1.05 - 1.20)\) at temperatures below -65°C. The coupling reaction between iodine-terminated living polypropylene and monofunctional living polystyrene anion was examined at 50°C and shown to result in the quantitative formation of monodisperse AB block copolymer of propylene and styrene. In addition, sequential polymerization of ethylene and propylene was attempted in the presence of the soluble V (acac)3/Al (C2H5)2Cl system to synthesize block copolymers.

“Living” polymerization is of great utility both in understanding the mechanism of chain propagation reactions and in synthesizing well-defined block copolymers. In this review we deal with the development of soluble Ziegler-Natta catalysts for the living coordination polymerization of ethylene and α-olefins and discuss the possibility of synthesis of new block copolymers.

“Living” polypropylene of uniform chain length was prepared by low-temperature polymerization of propene with some soluble vanadium-based catalysts. The chain end structure of living polypropylene was studied by 1H NMR analysis of iodine-bonded polypropylene. 1H NMR analysis revealed that the majority of active centers exist as a secondary vanadium-carbon bond during the chain propagation of living polypropylene. The mechanism of the living coordination polymerization is discussed based on kinetic and stereochemical data.

THE kinetic behaviour in the initial stages of the polymerization of propylene with titanium trichloride–triethyl aluminium can be divided into two types: the build-up type (the rate increases to the stationary value) and the decay type (the rate approaches decreasingly to the value). Natta and Pasquon1 observed that the build-up type changed into the other type as grinding the titanium trichloride, whereas the stationary rate did not change, and supposed that the produced polymer adjusted the particle size of titanium trichloride into the stationary value appropriate for the stationary polymerization. This supposition is based on the experimental result that the same stationary rate per unit weight of titanium trichloride was obtained with the use of the titanium trichloride catalysts in various particle sizes.

TO control the effect of mass transfer and to distinguish the chemical reaction rate, from an overall rate it is often necessary to evaluate the mass transfer rate in a heterogeneous reaction. This may be difficult in complex heterogeneous reactions such as the heterogeneous Ziegler-Natta polymerization, which occurs in a three-phase system: gas (olefin)-liquid (solution of organometallic compound)-solid (titanium trichloride powder). The high polymers produced accumulate as precipitates and solutes in the solution1. Both the hydrodynamical state of this polymerization system and the activity of the catalyst itself change with time. For such a complex heterogeneous reaction the usual method of evaluation of mass transfer may not be suitable.

THE Ziegler–Natta polymerization of α-olefines has been widely investigated by many workers mainly from the practical point of view. On the other hand, the details of the reaction mechanism are not yet clear. There are, of course, some basic investigations of the reaction by Natta et al.1, one of which is considered here; that is, the kinetic behaviour of the polymerization of propylene in the initial stages (Natta's ‘the adjustment period’), where the polymerization-rate, r, changes with the polymerization time, t, from the initial rate, r0, at t = 0, to the stationary rate, r∞. The phenomenon can be divided into two types, that is, the build-up type and the decay type. The decay type was observed with the ground catalysts (sizes ⩽ 2µ), while the build-up type was observed with the unground ones (1∼100µ). Analysing the latter types of rate change, Natta et al. found that: where tN is the polymerization time at which the rate equals ¾ of the value of the stationary rate, r∞, and P the pressure of propylene. Natta suggested that the stationary reaction corresponds to a definite particle size of the catalyst, and any other catalyst with a different particle size would alter to this definite size.