Cocatalysts behavior

Ethylene Polymerization Behavior of FI Catalysts with Cocatalysts Other than MAO... 

The ethylene polymerization behavior of FI catalysts has been described in previous sections. It is often observed that the cocatalyst that is employed has an influence on the catalytic behavior of a transition metal-based olefin polymerization catalyst. FI catalysts can exhibit unique catalytic behavior depending on the cocatalyst that is used for polymerization. 

It is well known that in conventional catalyst systems a chemical interaction between the catalyst and the metal-alkyl takes place, which essentially leads to a variation of the transition metal oxidation state. This is likewise true with MgCl2 catalysts however, in this case there are many more possible reactions, given the greater complexity of the system. Thus, besides modifying the Ti valence, the metal-alkyl may interact with the Lewis base incorporated in the catalyst. The Lewis base added to the cocatalyst can, in turn, interact both with the support and with the TiCl4, as can the byproducts originating from the reaction between Al-alkyl and Lewis base. The situation appears to be quite complex. However, detailed knowledge about these processes is absolutely necessary for any attempt to rationalize the polymerization behavior of these catalytic systems. 

The effect of increasing amounts of water, added as cocatalyst, is shown in Figure 1 for the catalyst with highest Ca content (V). The reaction rate increased sharply with small water additions and then leveled off at about 10 H20/Ca (2H20/cage). Similar behavior was observed with the other 4 catalysts of this series. Separate experiments showed that an increasing fraction of this H2O was in the gas phase with increasing addition under reaction conditions. 

Type (e) behavior is attributed to almost instantaneous breakdown of porous catalyst particles on treatment with the cocatalyst so that the acceleration or settlement period is practically eliminated. This behavior is shown by ether-treated highly porous catalysts in propylene polymerization [7j. The polymerization rate decreases very gradually with time and the catalyst system shows good stability. 

The behavior shown in Figure 28, a leveling of catalyst activity with increased chromium loading, could be viewed as a consequence of mass transport limitations. However, the activity can still be increased or decreased according to other preparation variables and reaction conditions. For example, activity is improved by increasing the activation temperature, by the addition of cocatalyst, by the reduction of Cr(VI) in CO, or by increasing the ethylene concentration in the reactor. The activity can also be lowered by poisons. 

The addition of small amounts of metal alkyls to the reactor can significantly enhance the activity of Phillips catalysts, and modify the polymers they produce as well. This behavior was discovered during the 1950s and has been used commercially ever since, at least for some polymer grades. When used in this way, metal alkyls are usually called "cocatalysts", with the term borrowed from Ziegler catalysis. But in contrast to Ziegler catalysts, Phillips catalysts do not require cocatalysts. They are, at most, used to enhance the activity, not develop it. 

SCHEME 45 Six ways in which cocatalysts might influence active-site behavior. In each class a plausible illustration is offered, although the details of such reactions are not necessarily known. 

Although there are many similarities between the behavior with cocatalysts and the behavior with organochromium catalysts, including the sometimes intense activity enhancement, use of a cocatalyst with Cr(VI) / silica is not quite equivalent to the use of organochromium catalysts, which usually become active immediately, with full rate instantaneously developed. The organochromium catalysts are usually already reduced and alkylated before entry into the reactor. 

Most cocatalysts continue to influence the polymer MW distribution even after the optimum ratio of cocatalyst to chromium (maximum activity) has been reached. This behavior indicates that some of the lower pathways in Scheme 45 also contribute. Either sites are changed, or even destroyed, by attack on the Cr-O-Si bonds, or the alkyl exchange reaction occurs. This latter pathway tends to enhance the low-MW side of the MW distribution by introducing another chain transfer mechanism. Only some very specific, perhaps the most acidic, sites are probably involved, such as those sites producing low-MW polymer on Cr/silica-titania or Cr/AIPO4. These sites already have a tendency to produce low-MW polymer [52,332,681],... 

The first polymer grade had a specification HLMI of 10 and a specification density of 0.949 g mL 1. A Cr/silica catalyst, activated at 670 °C, was run with and without 4 ppm of triethylboron in the reactor. Upon addition of the cocatalyst, the activity of the catalyst increased nearly two fold. The polymer HLMI/MI ratio (response to shear stress) increased, which reflects the broader MW distribution of the polymer made in the presence of cocatalyst. The reaction temperature was maintained at 90.6 °C, but when BEt3 was added, H2 also had to be added to the reactor to raise the HLMI back to the target value of 10. This behavior indicates that the cocatalyst increased the polymer MW by the addition of a high-MW tail. 

An example illustrating this behavior is shown in Table 62 Cr(VI) and Cr(II) catalysts, otherwise identical in composition, were tested for ethylene polymerization in the absence and presence of BEt3 cocatalyst. The second column in the table shows the degree of branching found in the resultant polymer, and the third column shows the density of that polymer. The response by Cr(VI) to the cocatalyst was only slight, but the response by Cr(II) was remarkable. With the addition of BEt3, the density of the polymer dropped so severely that the product type changed from the class of HDPE (homopolymer) to LLDPE. 

The initial temperature of catalyst activation can also influence the amount of in situ branching obtained in the polymer. This is in agreement with the olefin-generating behavior of the organochromium catalysts (Figures 185 and 192, Table 55). Table 67 shows an experiment in which Cr/silica-titania was activated at 800 °C or at 650 °C, and then it was reduced and tested for polymerization activity with 5 ppm triethylboron cocatalyst. The 800 °C catalyst resulted in significantly lower polymer density than the 650 °C catalyst. This derives from two causes. The 800 °C... 

Four examples illustrating this point are shown in Figure 219, where data represent experiments in which various cocatalysts were used at a concentration of 8 ppm with a Cr(II) /silica-titania catalyst. Each cocatalyst produced a different a-olefin profile. Triethylboron produced a prominent 1-hexene spike with a much smaller Schulz-Flory background. The addition of a small amount of 02 with the triethylboron cocatalyst increased the relative amount of 1-hexene while decreasing the overall yield of a-olefins. At the other extreme, diethylsilane produced a sharp Schulz-Flory distribution with little additional 1-hexene. The behavior of diethylaluminum ethoxide was intermediate, as a large amount of 1-hexene was produced, superimposed onto a significant Schulz-Flory... 

Data shown in Table 63 indicate that H2 itself was ineffective as a reducing agent for Cr(VI), yielding mostly Cr(III), poor activity, and little in situ branching when tested with a cocatalyst. This behavior has been attributed to the fact that water is formed as the by-product, which hydrolyzes the Cr-O-Si bond, allowing Cr(III) formation. Treatment of the Cr(VI) catalyst with CO before H2, however, is different. When added to the reactor with Cr(II)/silica, H2 produces a slight lowering of density, as shown in Table 73. 

This behavior is reversed if even a small amount of H2 is added to the reactor. Figure 243 shows an example of 0.2 MI polymers made with titanated Cr/silica in a pilot plant with triethylboron cocatalyst. The expected rise in MI during extrusion was almost completely curtailed... 

Type (a) behavior is observed for many first generation catalyst systems, e.g., or-TiCls, VQ3, etc. with diaUcylaluminum hahdes as cocatalysts in the polymerization of propylene in hydrocarbon media. During an initial acceleration period, which is of 20-60 minutes duration for many propylene polymerizations at 1 atm pressure in the temperature range 50-70°C, the rate increases from the beginning to reach a more or less steady value. Natta and Pasquon (1959) attributed this behavior to the breakdown of the or-TiCls matrix to smaller crystallites due to the pressure of the growing polymer chains in the initial stages, leading to exposure of fresh Ti atoms and creation of new active centers with consequent increase in... 

---