Composition drift

It was realized by Staudinger, as early as 1930, that when two monomers copolymerize, the tendency of each monomer to enter the chain can differ markedly. He found that if an equimolar mixture of vinyl acetate and vinyl chloride was copolymerized, the chemical composition of the product varied throughout the reaction, and that the ratio of chloride to acetate in the copolymers changed from 9 3 to 7 3 to 5 3 to 5 7. 

This phenomenon, known as composition drift, is a feature of many copolymerizations and has been attributed to the greater reactivity of one of the monomers in the mixture. Consequently, in a copolymetization, it is necessary to distinguish between the composition of a copolymer being formed at any one time in the reaction and the overall composition of the polymer formed at a given degree of conversion. 

Two major questions arise that must be answered if the criteria controlling copolymerizations are to be formulated  

Can the composition of the copolymer be predicted when it is prepared from the restricted conversion of a mixture of two monomers  

Can one predict the behavior of two monomers that have never reacted 

---

Monomer compositional drifts may also occur due to preferential solution of the styrene in the mbber phase or solution of the acrylonitrile in the aqueous phase (72). In emulsion systems, mbber particle size may also influence graft stmcture so that the number of graft chains per unit of mbber particle surface area tends to remain constant (73). Factors affecting the distribution (eg, core-sheU vs "wart-like" morphologies) of the grafted copolymer on the mbber particle surface have been studied in emulsion systems (74). Effects due to preferential solvation of the initiator by the polybutadiene have been described (75,76). 

Copolymers are typically manufactured using weU-mixed continuous-stirred tank reactor (cstr) processes, where the lack of composition drift does not cause loss of transparency. SAN copolymers prepared in batch or continuous plug-flow processes, on the other hand, are typically hazy on account of composition drift. SAN copolymers with as Httle as 4% by wt difference in acrylonitrile composition are immiscible (44). SAN is extremely incompatible with PS as Httle as 50 ppm of PS contamination in SAN causes haze. Copolymers with over 30 wt % acrylonitrile are available and have good barrier properties. If the acrylonitrile content of the copolymer is increased to >40 wt %, the copolymer becomes ductile. These copolymers also constitute the rigid matrix phase of the ABS engineering plastics. 

Styrene—maleic anhydride (SMA) copolymers are used where improved resistance to heat is required. Processes similar to those used for SAN copolymers are used. Because of the tendency of maleic anhydride to form alternating copolymers with styrene, composition drift is extremely severe unless the polymerization is carried out in CSTR reactors having high degrees of back-mixing. 

Vinyhdene chloride copolymerizes randomly with methyl acrylate and nearly so with other acrylates. Very severe composition drift occurs, however, in copolymerizations with vinyl chloride or methacrylates. Several methods have been developed to produce homogeneous copolymers regardless of the reactivity ratio (43). These methods are appHcable mainly to emulsion and suspension processes where adequate stirring can be maintained. Copolymerization rates of VDC with small amounts of a second monomer are normally lower than its rate of homopolymerization. The kinetics of the copolymerization of VDC and VC have been studied (45—48). 

The effect of different types of comonomers on varies. VDC—MA copolymers mote closely obey Flory s melting-point depression theory than do copolymers with VC or AN. Studies have shown that, for the copolymers of VDC with MA, Flory s theory needs modification to include both lamella thickness and surface free energy (69). The VDC—VC and VDC—AN copolymers typically have severe composition drift, therefore most of the comonomer units do not belong to crystallizing chains. Hence, they neither enter the crystal as defects nor cause lamellar thickness to decrease, so the depression of the melting temperature is less than expected. 

The batch-suspension process does not compensate for composition drift, whereas constant-composition processes have been designed for emulsion or suspension reactions. It is more difficult to design controUed-composition processes by suspension methods. In one approach (155), the less reactive component is removed continuously from the reaction to keep the unreacted monomer composition constant. This method has been used effectively in VT)C-VC copolymerization, where the slower reacting component is a volatile and can be released during the reaction to maintain constant pressure. In many other cases, no practical way is known for removing the slower reacting component. 

Benkoski, J.J., Fredrickson, G.H. and Kramer, E.J., The effect of composition drift on the effectiveness of random copolymer reinforcement at polymer-polymer interfaces. Macromolecules (2001, in press). 

Compositional drift in continuous reactor trains can be altered by introducing feed streams of the more reactive monomer between reactors. This procedure is equivalent to programmed addition of the more reactive monomer in a semi-continuous system. 

The major unresolved questions for these systems are the coalescence and flocculation of particles under mechanical agitation and the parameters which influence copolymer composition drift. 

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. 

Because the dependence of probability P Uk on x should be established by means of the theory of Markov chains, in order to make such an averaging it is necessary to know how the monomer mixture composition drifts with conversion. This kind of information is available [2,27] from the solution of the following set of differential equations ... 

Block copolymer synthesis from living polymerization is typically carried out in batch or semi-batch processes. In the simplest case, one monomer is added, and polymerization is carried out to complete conversion, then the process is repeated with a second monomer. In batch copolymerizations, simultaneous polymerization of two or more monomers is often complicated by the different reactivities of the two monomers. This preferential monomer consumption can create a composition drift during chain growth and therefore a tapered copolymer composition. 

The alternating tendency of the copolymers is advantageous in that the polymerizations can be carried out to high conversions with little or no compositional drift. For random copolymerizations in which there is preferential incorporation of one monomer due to a mismatch in reactivity ratios, the compositional variations with conversion can be substantial. Such compositional heterogeneities in resist materials can lead to severe problems during image development. 

The polymers are isolated and their composition is determined. This gives the value of x, i.e. instantaneous polymer composition ratio. Since X and x are different as polymerization progresses, the unreacted polymer ratio x will change resulting in a continuous change in the composition of the polymer formed. This is known as the composition drift. 

In the early 1930s, Nobel laureate Staundinger analyzed the product obtained from the copolymerization of equimolar quantities of VC and VAc. He found that the first product produced was high in VC, but as the composition of the reactant mixture changed because of a preferential depletion of VC, the product was becoming higher in VAc. This phenomenon is called the composition drift. 

Wall studied the composition drift and derived what is now called the Wall equation, where n was equal to rx when the reactivity ratio r was equal to the ratio of the propagation rate constants. Thus, r was the slope of the line obtained when the ratio of monomers in the copolymer (M1/M2) was plotted against the ratio of monomers in the feed (mi/ma). The Wall equation is not general ... 

The batch-suspension process does not compensate for composition drift, whereas constant-composition processes have been designed for emulsion or suspension reactions. It is more difficult to design controlled-composition processes by suspension methods. 

Except in very special cases (azeotropic copolymerizations), copolymerization via radical mechanism shows a drift in the composition of the copolymers produced through the polymerization process. Emulsion copolymerization obeys this rule too, although the special features of its mechanism can change the drift process. The most common way to obviate that composition drift is to use the semi-continuous process where, after polymerization has been initiated with a small percent of the total charge (say 10 to 20 %) like in the batch process, most of the charge is added continuously at a much smaller rate (Ra) than the rate (Rp) at the end of the batch period, so that the added charge is polymerized quite instantaneously (J, 2). Then,the composition drift is limited to the initial period and most of the product does possess actually a constant composition. 

But, as shown in figure 4, composition of the monomer mixture in the particles does not change very much when droplets are still present. After that point, however, the composition drift is very large, although styrene remains preferentially in the particles. 

One way to achieve this result relies on the change in the relative monomer reactivity following composition drifts. Thus, in a combination ofhigh and low reactivity monomers, the former will preferentially react first, leaving a considerable proportion of the latter for copolymerization when the supply of the high reactive monomer is depleted. This has been confirmed in the terpolymerization of methyl methacrylate/butyl acrylate/vinyl acetate in the presence of the maleate Surfmer reported in Figure 6.49. 

Under the copolymerization of more than two monomers Eqs. (5.3) cannot be integrated explicitly, and in order to determine the system trajectories one should need the numerical calculations. Examples of such calculations of the conversional change of composition and structure characteristics of the terpolymers have been reported in Refs. [195-200]. One should pay special attention to Ref. [200] where the programs for the computer realization of such calculations are presented. Under the copolymerization of four or more monomers, the composition drift with the conversion was calculated [7,8] only within the framework of the simplified terminal model described above in Sect. 4.6. 

Unfortunately, as far as the author knows, there are only a few publications where the problem of the validity of this or that model over a wide range of conversions and initial monomer feed compositions was discussed carefully enough. Here one might mention the works listed as Refs. [310,201] on the bulk copolymerization of styrene and heptyl acrylate, where the adequacy of the terminal model was undoubtedly proved, and its parameters rj = 0.87 and r2 = 0.27 were estimated. Really, the calculated copolymer composition and monomer feed composition drift with conversion are in full agreement with both NMR (Table 6.9) and UV (Fig. 18) data. 

For copolymers, in particular random copolymers, instead of discrete functionality fractions a continuous drift in composition is present (see Fig. 3). To determine this chemical composition drift in interrelation with the molar mass distribution, a number of classical methods have been used, including precipitation, partition, and cross-fractionation [2]. The aim of these very laborious techniques is to obtain fractions of narrow composition and/or molar mass distribution which are then analyzed by spectroscopy and SEC. 

We ll discuss the measurement of reactivity ratios shortly. First, we ll have a look at what types of copolymers we would expect to get for certain limiting values of the reactivity ratios and then have an initial look at the problem of composition drift. 

---