Polymerization autoacceleration

Autoacceleration, Glass and Zutty (S) and Burnett and Melville 9) reported an increase in the rate and average degree of polymerization with increasing solution viscosity, heterogeneous conditions and chain coiling for free radical, vinyl polymerizations. Autoacceleration is also called Trommsdorff. (10) effect. 

Let us consider the kinetics of DMDAACh initial polymerization in more details. The autoacceleration effect in radical polymerization (the so-called gel-effect) plays an important role in polymers synthesis processes and during years number was a topic of intensive studies [20,29-32]. At present it is assumed, that the autoacceleration realization is due to a stmctural changes in polymer solution [31, 32]. The detailed interpretation of this point of view is adduced in the indicated above works. However, it should be noted, that polymerization autoacceleration beginning is accompanied by a number of other important effects, which get much less attention. Let us indicate the some from them. The attention is paid to very high molecular weights MM of polymer, which are realized on polymerization initial section at low reaction rates, small conversion degrees (up to 5%) and small reaction durations (of order of 5 min) [32], Secondly, as it... 

The Tromsdorf effect, also called the Norrish effect or gel effect, is associated with exothermic reactions during bulk polymerization. Autoacceleration of the polymerization rate can occur with medium to high polymerization conversions. This phenomenon inhibits termination. Strength of the Tromsdorf effect is calculated as the gel effect index [12]... 

Autoacceleration in the polymerization of MA poses a serious problem [21-23]. Saini et al. [24] attempted to polymerize MA by using /3-PCPY as the initiator with a view to minimize the difficulties experienced due to this phenomenon. The findings led to the conclusion that -PCPY can be used to obtain 19.5% conversion of MA without gelation due to autoacceleration, which is nearly double the conversion obtained by using the conventional free radical initiator (AIBN) in the same experimental conditions. 

The predominant mode of polymerization is in the interior of the particles and this leads to a reduction of macroradical mobility, usually referred to as radical occlusion, and a marked autoacceleration of the polymerization rate. 

Some typical examples of this autoacceleration are (Figure 5) Norrish and Smith ( 2) polymerized methyl methacrylate in bulk and in the presence of various precipitants and measured the polymerization rates dilatometrically. They determined that autoacceleration of the precipitation polymerizations was larger than that observed for the Trommsdorf effect in bulk polymerization. 

Finally, similar autoacceleration in the polymerization rate was reported by Crosato-Arnaldi, Gasparini and Talamini (18) for the bulk polymerization of vinyl chloride. 

Figure 5. Effect of autoacceleration on the precipitation polymerization of methyl methacrylate (2). The curves, from left to right, are for the diluents cyclohexane t-hutylsterate heptane and bulk.

The precipitation polymerization literature is reviewed with particular attention to the influence of particle formation and growth, autoaccelerating polymerization rates, and copolymer composition drift on polymer reactor design. 

Autoacceleration in the rate of polymerization occurs also with other monomers. It is far more marked with methyl acrylate or acrylic... 

In quest of an explanation for this phenomenon, one is led to conclude either that the combination of constants occurring in the rate equation (12) must undergo a large increase when autoacceleration occurs or that a totally different mechanism of polymerization must take over. We should obviously prefer the former alternative if it will lead to a satisfactory explanation of the facts. An increase in kdj seems unlikely autoacceleration is not a function of the initiator. This leaves us with the ratio which will be required to increase by... 

The above explanation of autoacceleration phenomena is supported by the manifold increase in the initial polymerization rate for methyl methacrylate which may be brought about by the addition of poly-(methyl methacrylate) or other polymers to the monomer.It finds further support in the suppression, or virtual elimination, of autoacceleration which has been observed when the molecular weight of the polymer is reduced by incorporating a chain transfer agent (see Sec. 2f), such as butyl mercaptan, with the monomer.Not only are the much shorter radical chains intrinsically more mobile, but the lower molecular weight of the polymer formed results in a viscosity at a given conversion which is lower by as much as several orders of magnitude. Both factors facilitate diffusion of the active centers and, hence, tend to eliminate the autoacceleration. Final and conclusive proof of the correctness of this explanation comes from measurements of the absolute values of individual rate constants (see p. 160), which show that the termination constant does indeed decrease a hundredfold or more in the autoacceleration phase of the polymerization, whereas kp remains constant within experimental error. 

The susceptibility of the polymerization of a given monomer to autoacceleration seems to depend primarily on the size of the polymer molecules produced. The high propagation and low termination constants for methyl acrylate as compared to those for other common monomers lead to an unusually large average degree of polymerization (>10 ), and this fact alone seems to account for the incidence of the decrease in A f at very low conversions in this case. 

Most polymerizations proceed with autoacceleration as the increasing viscosity of the reaction system results in an increase in the kp/kj.1/2 ratio. Both kp and kt decrease with increasing viscosity but k(- decreases more than does kp since termination involves reaction between two large-sized species while propagation involves reaction between one large and one small species. 

During conventional polymerizations of both HEMA and DEGDMA, complications resulting from diffusion limitations to termination and propagation are observed. Features such as autoacceleration, autodeceleration and incomplete conversion of double bonds characterize the rate behavior of these polymerizations. As TED is added to the reacting system, the carbon-DTC radical termination reaction is introduced. Diffusion limitations to carbon-DTC radical combination are lower than those to carbon-carbon radical termination as the DTC radical is smaller and much more mobile than a typical polymeric carbon radical. As a result, the cross-... 

The model captures the features seen in the experiment very well. The initial rate of polymerization and the conversion at the onset of autoacceleration are nearly identical with the experimentally generated values as is the rate of autodeceleration. The conversion and value of maximum rate are within 5%. The difference in maximum rate can be ascribed to the high sensitivity of autoacceleration on the At parameter a small change in can greatly influence the rate of polymerization during autoacceleration. The conversion at which the maximum rate occurs is dependent upon fcp, i.e., the free volume (and conversion) at which autodeceleration sets in. The simulated rate also shows a tail around 80% conversion, which can be ascribed to the DSC not capturing polymerization at high conversion and low rate. 

As in the case of HEMA polymerizations, the model predicts the experimentally observed rate successfully. The rates of autoacceleration and autodeceleration are predicted well by the model, and the initial rate and maximum rate are within 5%. A more pronounced shoulder is seen in the model predictions than in the experiments. This can be attributed to DSC startup, going from zero rate to a high rate as the photoaccessory is turned on. 

It can be observed that the initial rate of polymerization decreases and the autoacceleration peak is suppressed as the TED concentration is increased. The TED molecules generate dithiocarbamyl (DTC) radicals upon initiation. As a result, termination may occur by carbon-carbon combination which leads to a dead polymer and by carbon-DTC radical reaction which produces a reinitiatable ( living ) polymer. The cross-termination of carbon-DTC radicals occurs early in the reaction (with the carbon-carbon radical termination), and this feature is observed by the suppression of the initial rate of polymerization. As the conversion increases, the viscosity of the system poses mass transfer limitations to the bimolecular termination of carbon radicals. As has been observed in Figure 3, this effect results in a decrease in the ktCC. However, as the DTC radicals are small and mobile, the crosstermination does not become diffusion limited, i.e., the kinetic constant for termination of carbon-DTC radicals, ktCS, does not decrease. Therefore, the crosstermination becomes the dominant reaction pathway. This leads to a suppression of the autoacceleration peak as the carbon-DTC radical termination limits the carbon radical concentration to a low value, thus limiting the rate of polymerization. This observation is in accordance with results of previous studies (10) with XDT and TED, where it was found that when there was an excess of DTC radicals, the carbon radical concentration was lower and the cross-termination reaction was the dominant termination pathway. 

Figure 5 shows similar experimental rate data for the DEGDMA/DMPA/TED polymerization. As seen in the case of HEMA, TED addition decreases both the initial rate and the maximum rate of polymerization of DEGDMA. As described earlier, polymerization of DEGDMA results in a highly crosslinked polymer. The autoacceleration effect is characterisitc of highly crosslinked systems as the diffusional limitations reduce the carbon-carbon radical termination kinetic constant... 

From these experimental and modeling studies, the mechanism of the living free radical polymerizations initiated by a combination of TED and DMPA have been elucidated. The TED produces DTC radicals that preferentially cross-terminate with the propagating carbon radicals. By this cross-termination reaction, the carbon radical concentration is kept low (as was shown in figure 6) and the rate of polymerization is decreased, as is the autoacceleration effect. This suppression of the autoacceleration peak in HEM A polymerizations and, interestingly, in DEGDMA polymerization has been observed to increase as the TED concentrations are increased. This behavior has been predicted successfully by the model as well. 

The bulk polymerization of acrylonitrile in this range of temperatures exhibits kinetic features very similar to those observed with acrylic acid (cf. Table I). The very low over-all activation energies (11.3 and 12.5 Kj.mole-l) found in both systems suggest a high temperature coefficient for the termination step such as would be expected for a diffusion controlled bimolecular reaction involving two polymeric radicals. It follows that for these systems, in which radicals disappear rapidly and where the post-polymerization is strongly reduced, the concepts of nonsteady-state and of occluded polymer chains can hardly explain the observed auto-acceleration. Hence the auto-acceleration of acrylonitrile which persists above 60°C and exhibits the same "autoacceleration index" as at lower temperatures has to be accounted for by another cause. 

However, DMF is a solvent for polyacrylonitrile and the polymerization occurs in a homogeneous medium for solutions containing 30 per cent monomer or less. This reduces the value of these experiments as an argument to show the influence of a matrix effect. Indeed the fact that auto-acceleration disappears when DMF is added to acrylonitrile was considered as a proof for the fact that precipitation of the polymer was the cause of autoacceleration. 

These results conclusively demonstrate that precipitation of polyacrylonitrile as a fine powder and occlusion of growing chains resulting in post-polymerization do not bring about autoacceleration if a highly polar solvent is present in the system. 

---