Monomer conversion monitoring

Monitoring monomer conversion during emulsion polymerization... 

Chabot et al. at Atofina Chemicals (King of Prussia, PA, USA) used in-line NIR to monitor monomer conversion in real time in a batch emulsion polymerization process. The business value of this monitoring... 

The available data from emulsion polymerization systems have been obtained almost exclusively through manual, off-line analysis of monomer conversion, emulsifier concentration, particle size, molecular weight, etc. For batch systems this results in a large expenditure of time in order to sample with sufficient frequency to accurately observe the system kinetics. In continuous systems a large number of samples are required to observe interesting system dynamics such as multiple steady states or limit cycles. In addition, feedback control of any process variable other than temperature or pressure is impossible without specialized on-line sensors. This note describes the initial stages of development of two such sensors, (one for the monitoring of reactor conversion and the other for the continuous measurement of surface tension), and their implementation as part of a computer data acquisition system for the emulsion polymerization of methyl methacrylate. 

Monomer conversion has traditionally been determined gravitimetrically by drying emulsion samples to constant weight. The procedure is slow, requiring several hours for analysis, and precludes automated data acquisition. A new method has been developed based on the DMA-series digital densitometers manufactured by Anton Paar of Austria, and marketed in the United States by Mettier Instrument Corporation. (Very recently Dr. Kirk Abbey made us aware of his parallel work in these directions and of some initial data reported from his laboratory [1, 2]). This instrument is capable of immediate determination of the density of any test fluid, and, if equipped with a flow cell, can continuously monitor the density of a process stream. Results are displayed locally and can be transmitted digitally to a data acquisition computer. 

For the purposes of conversion monitoring of emulsion polymerization, we have found the DMA40D with a precision of +1 x 10 g/cm capable of resolving monomer conversion to +0.2% in the absence of thermostating and sampling errors. On-line, in the presence of such possible errors, a resolution of at least +0.5% can be expected. Care must be taken to ensure a representative sample and good temperature control of the sample stream before introducing it into the instrument. Some example results with this instrument are presented below. 

Contrary to viscosity measurements GPC provides number average molar mass data (Mn). For a few studies on the polymerization of BD with Nd-carboxylate-based catalyst systems GPC was systematically applied for the monitoring of Mn as a function of monomer conversion. In these studies three catalyst systems were used (1) NdV/DIBAH/EASC [178], (2) NdV/TIBA/EASC [179] and (3) NdO/TIBA/DEAC [188]. In the first two studies linear Mn-conversion plots were obtained at various molar ratios Al/ Ndv- In these studies, however, molar mass data at low monomer conversions (<20%) are lacking and positive intercepts on the Mn-axis were found. For the ternary catalyst system NdO/TIBA/DEAC used in the third of these studies the concentrations of Nd and TIBA were varied. A linear increase of Mw and Mn on monomer conversion was found. Deviations from linearity were also observed for low monomer conversions. 

In summary, it has to be mentioned that in many studies intrinsic viscosity, inherent viscosity and dilute solution viscosity (DSV) were used in order to monitor the increase of molar mass on monomer conversion. Unfortunately, only a few studies use GPC rather than viscosity measurements. For a few Nd-carboxylate-based catalyst systems linear dependencies of Mn on monomer conversion were established and proof in favor of requirement No. 2 linear increase of Mn with monomer conversion (no irreversible chain transfer) was provided. A more detailed analysis of the data, however, reveals deviations from linearity particularly at low monomer conversions (< 20%). These deviations are particularly pronounced for polymerizations with induction periods. Also the extrapolation of the straight lines to zero monomer conversion reveals intercepts on the Mn-axis. 

The activity of CALB immobilized on resins 1 to 4 (see Table 1) was assessed by s-CL ring-opening polymerizations. All immobilized resins have similar CALB loading (5.1 to 5.7 %-by-wt) and water content (1.4 to 1.9%). As shown previously , different chemical shifts are observed for methylene protons (-OC//2-) of e-CL monomer, PCL internal repeat units and chain terminal -012-OH moities. Hence, by in situ NMR monitoring, monomer conversion and polymer number average molecular weight (A/ ) were determined. 

It is sometimes possible to estimate those properties of a polymerization reactor which are not measurable. The most common application of these techniques is the use of a heat balance to estimate the rate of reaction. If the coolant duty necessary to remove the heat of polymerization is carefully monitored, the rate of polymerization can be calculated if the heat of reaction (propagation) is known. Unlike most situations where scale-up is detrimental, this is one case where the use of a large reactor is beneficial. A large reactor has a low surface-to-volume ratio, and therefore heat losses tend to be smaller (relatively) and more constant (and hence, characterizable) than in a smaller kettle. If the rate of polymerization is known as a function of time, monomer conversion can be calculated by integrating the rate of polymerization over time. This technique is inexpensive to implement and is applied in many latex reactors with surprisingly good results. Heat balance estimates of conversion can be combined with other, less frequent conversion measurements within the framework of a state estimator. 

The need for continuous monitoring of monomer conversion in batch, semi-continuous and continuous emulsion polymerization is growing because the requested performance of polymer materials has caused higher demands on the reproducibility and fine tuning of the production processes. 

The knowledge of monomer conversion not only allows delicate adjustment of the process conditions but also intelligoit monomer addition strategies in copolymerizations strongly rely on the monitoring of monomer conversions. 

Bon and coworkers carried out a study on the fate of the nanoparticles throughout solids-stabilized emulsion polymerization [119], A quantitative method based on disk centrifugation was developed to monitor the amount of nanoparticles present in the water phase in solids-stabilized emulsion polymerizations of vinyl acetate, methyl methacrylate, and butyl acrylate. The concentration profile of nanoparticles in the water phase as a function of monomer conversion agreed with theoretical models developed for the packing densities in these systems [120]. Noteworthy was that in the case of silica-nanoparticle-stabilized emulsion polymerization of vinyl acetate, the event of late-stage limited coalescence, leading to small armored non-spherical clusters, could be predicted and explained on the basis of the concentration profiles and particle size measurements. Adjusting the amount of silica nanoparticles prevented this phenomenon. 

These studies inspired a third report where Hoogenboom et al7 used automated methods to perform detailed copolymerization studies and establish structure-property relationships of these copolymers. Twenty-seven polymerizations were performed by taking combinations of MeOx, EtOx, and NonOx in monomer ratios of 0-100%. Monomer conversion was monitored by GG and linear kinetics confirmed the... 

The NIR in situ process also allowed for the determination of intermediate sequence distribution in styrene/isoprene copolymers, poly(diene) stereochemistry quantification, and identification of complete monomer conversion. The classic one-step, anionic, tapered block copolymerization of isoprene and styrene in hydrocarbon solvents is shown in Figure 4. The ultimate sequence distribution is defined using four rate constants involving the two monomers. NIR was successfully utilized to monitor monomer conversion during conventional, anionic solution polymerization. The conversion of the vinyl protons in the monomer to methylene protons in the polymer was easily monitored under conventional (10-20% solids) solution polymerization conditions. Despite the presence of the NIR probe, the living nature of the polymerizations was maintained in... 

The FTIR conversion (77%) was somewhat lower than the gravimetric conversion (82%), most likely due to the overlap of the 1486 cm band with another deformation vibration at 1448 cm. According to Figure 10, final conversion was reached in 10 minutes. Monomer consumption was also evidenced by the decreasing intensity of the 1655 cm olefinic band (C=C). However, this band could not be used for conversion monitoring due to the interference of the polybutenamer olefinic signal. 

Previous work of one of the authors dealt with mid IR monitoring of IB polymerization using fiber-optic equipment. Fourier transform near infrared spectroscopy can be accomplished without expensive hardware. It therefore seemed to be a desirable goal to combine the advantage of fiber optics with low-cost fibers available for measurements in the NIR range. The NIR spectrum of IB obtained after solvent subtraction (Figure 2) reveals at least three signals, which should be suitable for the determination of monomer conversion. 

Monomer conversions were monitored using NMR spectroscopy the vinyl signals from the monomers appear at 5.3 and 6.0 ppm and decrease in intensity as they are consumed in the production of polymer. As the polymerization proceeds, signals of the methacrylate backbone increase between 0.9-1.4 ppm (Fig. 2). 

The polymerization reactors are equipped with two sapphire windows [8], which allow for initiation of the polymerization by UV laser pulses, for in-line FT-NIR spectroscopic monitoring of the degree of monomer conversion, and for visual and NIR-spectroscopic inspection of homogeneity of the polymerizing system [10]. 

Other non-invasive (e.g., calorimetry and ultrasounds) and invasive (e.g., densimetry) techniques can also be used for monitoring of monomer concentration in homopolymerization reactions. However, in multimonomer formulations the individual monomer concentrations cannot be obtained with these techniques, meaning that a state estimator is required [66, 67, 69]. The use of reaction calorimetry is appealing because the hardware is very cheap and, when coupled with a state estimator, provides good estimation of the monomer concentration. The performance of calorimetry was compared with that of the Raman spectroscopy in emulsion polymerization to monitor overall and individual monomer conversions [68]. Calorimetry was as good as FT-Raman spectroscopy when monomer concentrations in the reactor were relatively high, but the performance of calorimetry was poorer when monomer concentrations were low. 

---