A replicated investigation of nitroxide-mediated radical polymerization of styrene over a range of reaction conditions.
Controlled radical polymerization (CRP) allows the production of well-defined polymeric materials with controlled architecture. It is anticipated that new products via CRP will be introduced in the market within the next several years (Rutsch and Cech, 2007). Nitroxide-mediated radical polymerization (NMRP) is one of the three currently most popular approaches towards controlled radical polymerization. The literature on NMRP is extensive and growing. Table 1 summarizes some of the most important contributions in this area. Despite the heightened academic interest in the last 10-15 years or so and although the industrial production of polymeric materials made by NMRP has started (Auschra et al., 2002; Pirrung and Auschra, 2005), there are still issues that remain to be resolved before the production of materials created using this method can be compared to the manufacturing scales of regular radical polymerization.
The bimolecular NMRP of styrene using 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) as controller and dibenzoyl peroxide (BPO) as initiator was first studied by the group of Georges et al. (1993) and subsequently, a lot of work has been accomplished (see the first entry in Table 1 and related references). Although at first glance the system seems relatively well studied, no detailed/reliable experimental results (over the whole range of polymerization) are available in the literature. For instance, the conversion versus time data points reported in the past only capture the first few hours of the experiment and low to medium conversion (see, e.g., Veregin et al., 1996a,b; Dollin et al., 2007). Average molecular weight data are usually not readily available (Veregin et al., 1996a,b) and, if available, are presented as number-average molecular weight points evaluated at a limited number of conversion levels (usually, low-medium-high conversion; see, e.g., MacLeod et al. 1997; Dollin et al., 2007). Weight-average molecular weight data are hardly ever reported (with the usual argument that the polydispersity index, PDI, is usually low and close to 1.1-1.2). In addition, checks for reproducibility and independent replication are almost non-existent, an exception being the work by Roa-Lung et al. (2007). Overall, sound experimental studies for bimolecular NMRP of styrene with TEMPO and BPO with detailed experimental data over the full conversion range are still rare, and usually carried out in non-systematic ways. This makes the subsequent mathematical modelling stage even more complicated (with added uncertainties with respect to important parameter values).
These observations motivated the study presented herein. The objectives were, (a) to investigate the effect of controller to initiator molar ratio ([TEMPO]/[BPO]) and temperature on polymerization rate and molecular weights and thus, find a region in which conditions are optimal with respect to productivity (rate) and quality (molecular weights), and (b) to generate a source of reliable experimental data for validation and improvement of a mechanistic mathematical model already generated in our group (Bonilla et al., 2002; Roa-Lung et al., 2007; Ximenes et al., 2007). This mathematical model with the enhanced experimental information included via updated parameter values, can subsequently guide mechanistic model-based non-linear experimental design schemes, which can further shed light on the more uncertain parts of our process understanding.
Styrene (Aldrich Canada Ltd., Oakville, ON) was washed three times with a 10% (w/v) sodium hydroxide solution, washed three times with distilled water, dried over calcium chloride and distilled under vacuum. Solvents such as tetrahydrofuran, ethanol, dichloromethane, and acetone used during the course of the experimental analysis were used as received. Both BPO and TEMPO were used as received from suppliers (ATOFINA Chemicals King of Prussia, PA, and Aldrich, respectively) without further purification.
Polymerizations were completed in borosilicate glass ampoules (capacity ~4 mL). Reagents were weighed, mixed and pipetted into ampoules. Ampoules were then degassed by several vacuum-freeze-thaw cycles, sealed under vacuum with a gas/oxygen torch and immersed in a silicone oil bath having a temperature control of [+ or -] 0.1[degrees]C. Ampoules were removed at selected time intervals to ensure a well-defined conversion versus time plot. Once removed from the bath, the ampoules were placed in liquid nitrogen to stop the polymerization. Ampoules were then thawed, weighed, and opened. The contents were dissolved in dichloromethane, and poured into a flask containing ethanol to precipitate the polymer. The polymer samples were air-dried to remove the solvent and vacuum-dried for 3 days at approximately 60[degrees]C until a constant weight was reached.
Conversion levels were determined by gravimetry. Average molecular weights and polydispersity (PDI) were obtained by size exclusion chromatography (SEC) (interchangeably used as gel permeation chromatography, GPC). The equipment consisted of an isocratic Waters 515 HPLC pump, a "Waters 717 plus" injector, three PLgel 10 [micro]m MIXED-B columns (Polymer Laboratories Inc., Amherst, MA), and a Viscotek TDA 302 set of four detectors: refractive index (RI), ultra-violet (UV), low angle laser light scattering (LALLS) and right angle laser light scattering (RALLS), as well as an intrinsic viscosity detector (IV-DP viscometer differential pressure).
Summary of Experiments
A series of styrene polymerizations were performed varying the TEMPO to BPO ratio from 0.9 to 1.5. The details of experiments conducted are summarized in Table 2. Experiments were carried out at two different temperatures (120, 130[degrees]C). The initial initiator concentration (BPO) was kept constant at 0.036 M (1 wt% with respect to monomer), while TEMPO concentration was changed to give different TEMPO/BPO molar ratios.
It is important to make sure that the experimental data obtained are reliable and error in each section of the experiment is at a minimum. To do so, individual ampoule replicates were taken out of the oil bath at specific times to check the sampling error. In addition, in order to check the reproducibility of data, completely independent replicates were conducted for runs 2, 6, and 7.
Reliability of molecular weight measurements was checked by running GPC replicates at different times. In addition, during GPC analysis, two independent injections were done for every sample. Molecular weight values plotted in subsequent figures are averages from these two injections.
[FIGURE 1 OMITTED]
RESULTS AND DISCUSSION
Figure 1 shows the basic steps during the bimolecular nitroxide-mediated radical polymerization of styrene in the presence of TEMPO as the nitroxide radical and BPO as the radical initiator. The first step is thermal decomposition of BPO into benzoyloxy primary radicals with high reactivity (Eq. 1), which initiates the polymerization of styrene by attacking the carbon-carbon double bond (Eq. 2). Radicals with chain length unity will then propagate (Eq. 3), until they are trapped by TEMPO radicals. The TEMPO radical forms a labile bond (C-O) with the radical chain, leading to the formation of alkoxyamines in situ. The C-O bond is relatively weak at temperatures greater than 100[degrees]C so it reversibly dissociates, establishing the activation-deactivation equilibrium between dormant and active chains (Eq. 4).
The core reaction in NMRP is the equilibrium between active and dormant species (Eq. 4). [TEMPO]/[BPO] ratio and temperature are the factors which influence the equilibrium. [BPO] dictates the concentration of active radicals early in the reaction while TEMPO influences the concentration of dormant species; so obviously the [TEMPO]/[BPO] ratio is one of the leading factors in guiding the equilibrium. Ximenes et al. (2007) and Bonilla et al. (2002) cite the rate constants, [k.sub.a] and [k.sub.da], as functions of temperature; as a result, temperature influences the equilibrium by affecting the individual rate constants of the equilibrium reaction.
Effect of [TEMPO]/[BPO] Ratio
Figure 2 shows the effect of decreasing the [TEMPO]/[BPO] ratio (R) on rate of polymerization, illustrated as In[[M].sub.0]/[M] versus time. As can be seen, decreasing the ratio from 1.1 to 0.9, increases the rate of polymerization. Figure 3a shows that decreasing the ratio results in higher initial concentration of active (living) radicals. Since the TEMPO concentration is lower at [TEMPO]/[BPO] = 0.9, free radicals cannot all be trapped by TEMPO leading to a decrease in dormant species concentration in comparison to ratio 1.1 (see Figure 3c). Propagation occurs mainly as in regular radical polymerization for R = 0.9, resulting in an increase in dead polymer concentration compared to R = 1.1 (see Figure 3b). Given that for R = 0.9, relative to R = 1.1, the dead polymer concentration (Figure 3b) is higher, the dormant species concentration (Figure 3c) is lower, and the active radical concentration (Figure 3a) is much higher, the result is a higher overall polymerization rate for R = 0.9.
[FIGURE 2 OMITTED]
Let us now digress for a moment and try to anticipate the corresponding number- and weight-average molecular weight trends. In order to do this, we will make use in our arguments of the moments of the molecular weight distribution. There are three populations in the system: living (active) radicals, dormant species and dead polymer molecules. The moments for these species are defined as [[lambda].sub.i], [[delta].sub.i], and [[mu].sub.i], respectively. Based on the moments of the different species, number- and weight-average molecular weights can be calculated using Equations (5) and (G)
[M.sub.n] = [MW.sub.M] ([[mu].sub.1] + [[lambda].sub.1] + [[delta].sub.1]/[[mu].sub.0] + [[lambda].sub.0] + [[delta].sub.0] (5)
[M.sub.w] = [MW.sub.M] ([[mu].sub.2] + [[lambda].sub.2] + [[delta].sub.2]/[[mu].sub.1] + [[lambda].sub.1] + [[delta].sub.1] (6)
It is evident from Equation (G) that if we show that the numerator of the equation is higher for R = 0.9, then we will expect the weight-average molecular weight for R = 0.9 to be higher (than for R = 1.1). Careful scrutiny of Figure 3 along with Equation (G) shows that the dominant terms are [[delta].sub.2] and [[mu].sub.2]. Equations (7) and (8) below give the expressions for the second moments of the dormant species ([[delta].sub.2]) and dead polymer molecules ([[mu].sub.2]), respectively (for the sake of brevity, no details are given here about the derivation of these equations, however derivation details can be found in Bonilla et al. (2002) and Ximenes et al. (2007)). From Equation (7), which gives the rate of change of [[delta].sub.2] with time, [[delta].sub.2] for R = 0.9 will be higher than for R = 1.1 (observe the right-hand side of Eq. 7 along with Figure 3c). In an analogous way, based on Equation (8), the second moment of the dead polymer molecule population ([[mu].sub.2]) will have the tendency to be similar or higher (at least initially, due to larger [[lambda]] values; see Figure 3a) for R = 0.9 compared to R = 1.1. Hence, the weight-average molecular weight (and similarly, the number-average molecular weight) will be anticipated to be higher for R = 0.9 than for R = 1.1
[FIGURE 3 OMITTED]
d/([[delta].sub.2])/dt = +[k.sub.da][N[O.sup.*.sub.x]][[[lambda].sub.2]] - [k.sub.a][[[delta].sub.2]] (7)
d/([[mu].sub.2])/dt = +[k.sub.tc]([[[lambda].sub.0]][[[lambda].sub.2]] + [[[[lambda].sub.1]].sup.2]) + [k.sub.td] [[[lambda].sub.0]][[[lambda].sub.2]] + [k.sub.fM][[[lambda].sub.2]][M] + [k.sub.fD][[[lambda].sub.2]][D] (8)
This is exactly what is observed from the experimental data of Figures 4 and 5: both weight- and number-average molecular weights are higher at [TEMPO]/[BPO] = 0.9 (compared to R = 1.1). Both number- and weight-average molecular weights increase linearly at [TEMPO]/[BPO] = 0.9, showing that polymerization is still controlled at this condition after the initial reaction period. Selective GPC replicates were carried out to check the accuracy of molecular weight measurements at low and high conversions. As can be seen, the results are relatively reproducible. However, the error in GPC measurements is lower at higher molecular weights, as one would have expected.
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Reducing the [TEMPO]/[BPO] ratio increases the polydispersity (PDI) at equivalent conversions, for example, from 1.12 at TEMPO/BPO = 1.1 to 1.35 at R = 0.9 at [approximately equal to] 40% conversion (see Figure G). Due to the decreased TEMPO concentration at R = 0.9 and occurrence of "uncontrolled" polymerization in the initial reaction period, more growing polymer radicals are produced, thus more termination occurs at the same conversion level and a relatively larger fraction of polystyrene (PS) is produced by uncontrolled polymerization (see again Figure 3a and b). However, as conversion increases, polydispersity values become almost identical. For example, around 90% conversion, polydispersities are 1.14 and 1.13 for [TEMPO] /[BPO] = 1.1 and 0.9, respectively. The reason is that as polymerization proceeds and more polymer molecules are formed by the controlled process (after the initial phase), the fraction of PS produced by the uncontrolled process drops and hence over time the cumulative PDI decreases. Since the dormant species dominates (see Figure 3c), eventually PDI values become almost identical for both values of R.
[FIGURE 7 OMITTED]
Figure 7 shows the effect of increasing [TEMPO]/[BPO] ratio on rate of polymerization, shown this time as conversion versus time. Increasing the initial TEMPO concentration decreases the rate of polymerization dramatically. After 10 h, monomer conversion was around 20% for [TEMPO]/[BPO] = 1.5 while it was 50% for [TEMPO] /[BPO] = 1.1. When R = 1.5, the concentration of radicals produced by BPO decomposition is lower than the actual concentration of free TEMPO. An excess of free TEMPO remains in the system and it follows that the resulting concentration of active radicals is too small (especially at the beginning) to give significant polymerization rates. Therefore, an induction period arises. This induction period (about 3 h as seen in Figure 7) can also be confirmed by looking at the simulated results of Figure 8. Specifically, one can clearly see from Figure 8b that with R = 1.5 the living radical concentration profile behaves very differently from that of R =1.1 (with the concentration being almost zero during the initial phase). Eventually, the active (living) radical concentration for R = 1.5 starts to build up (see Figure 8b), so polymerization proceeds but with a slower rate in comparison to ratio R = 1.1 due to the higher concentration of dormant species present (see Figure 8c).
[FIGURE 8 OMITTED]
Figures 9 and 10 show the corresponding profiles for number-and weight-average molecular weights, respectively. It is observed that slightly lower average molecular weights ([M.sub.n] and [M.sub.w]) are obtained when the [TEMPO]/[BPO] ratio is increased (the explanation is the same as discussed earlier for Figures 4 and 5 for ratios 0.9 and 1.1). As expected for any CRP process, molecular weights increase linearly with conversion (the linear increase of molecular weights with conversion indicates that the proportion of chains that are self-initiated and terminated is low). Again, selective independent GPC replicates for ratio 1.5 show good reproducibility of average molecular weight measurements.
Figure 11 shows polydispersity versus conversion. Frequent sampling for the [TEMPO]/[BPO] ratio of 1.5, carried out at the beginning of the reaction, captured the initial variation in polydispersity values. Rather large changes in PDI values had been predicted in earlier modelling work (Veregin et al., 1996a; Roa-Lung et al., 2007) but never captured experimentally. This is the first time that such corroborating observations have been made. As can be seen in the insert of Figure 11, PDI values do not show much difference for different ratios after 20% conversion and very low values, below 1.2, are obtained.
Figure 12 shows the full picture of the effect of [TEMPO]/[BPO] molar ratio on polymerization rate (expressed as conversion vs. time) at 120[degrees]C. As expected, the larger the ratio (the more TEMPO in the recipe), the slower the polymerization proceeds. Polymerization rates are quite similar for TEMPO/BPO = 1.1 and 1.2.
[FIGURE 9 OMITTED]
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
The experimental data for conversion versus time, [M.sub.n] and [M.sub.w] versus conversion, and PDI versus conversion for the NMRP of styrene at 130[degrees]C and [TEMPO]/[BPO] = 0.9, 1.1, and 1.3 are shown in Figures 13-16. The trends are similar to the ones observed at 120[degrees]C, accounting of course for the temperature effect. However, the effect of [TEMPO]/[BPO] ratio on polymerization rate and molecular weights at 130[degrees]C is not as pronounced as it is at 120[degrees]C. More on the effect of temperature will be discussed later.
The results obtained are in agreement with previous studies. Veregin et al. (1996a,b) had conducted NMRP of styrene at 125[degrees]C, [TEMPO] /[BPO] = 1.1 and 1.3, as part of their kinetic studies. They had captured the linear trend of rate of polymerization with time and also shown that as [TEMPO]/[BPO] increased from 1.1 to 1.3, rate of polymerization decreased. However, their study only captured the first 8 h of the experiment and low to medium conversions (up to 33% and 50% for [TEMPO]/[BPO] = 1.3 and 1.1, respectively). They had shown that molecular weights increase linearly with conversion and slightly higher values are obtained at [TEMPO]/[BPO] = 1.1. Polydispersity values were higher at [TEMPO]/[BPO] = 1.1 and that is in agreement with our studies whereby as [TEMPO]/[BPO] ratio increases, lower polydispersities are obtained.
[FIGURE 12 OMITTED]
[FIGURE 13 OMITTED]
[FIGURE 14 OMITTED]
[FIGURE 15 OMITTED]
[FIGURE 16 OMITTED]
Subsequently, MacLeod et al. (1997) published a report showing a series of styrene polymerizations performed with varying ratios of [TEMPO]/[BPO], from 1.35 to 0.5. Results were presented in a table showing number-average molecular weights, polydispersity and conversion after 5 h at 135[degrees]C. It was shown that as the ratio was lowered from 1.35 to 0.5, the polydispersity increased, and molecular weights and conversion level increased. Results from MacLeod et al. (1997) have been plotted as polydispersity values versus [TEMPO]/[BPO] ratio (Figure 17a). As can be seen, the minimum value is around ratio 1.2, which is in agreement with our data at 120 and 130[degrees]C (see Figure 17b and c). In order to get a better feel about how narrow the polydispersity values are in NMRP, values from thermal polymerization of styrene (in the absence of initiator and nitroxide), as well as thermal NMRP (nitroxide-mediated polymerization of styrene in the absence of initiator) are also presented in Figure 17b and c. As can be seen from these bar graphs, the optimal operating range to achieve the lowest polydispersity is around [TEMPO]/[BPO] = 1.1-1.2 for both temperatures.
[FIGURE 17 OMITTED]
Polydispersity versus [TEMPO]/[BPO] ratio plots for different conditions (after 10 and 20 h for styrene polymerization at 120[degrees]C, and after 8 and 30 h for styrene polymerization at 130[degrees]C) are shown is Figures 18 and 19. The general trends are the same, namely, the optimal operating range to achieve the lowest polydispersity is around [TEMPO]/[BPO] = 1.1-1.2.
[FIGURE 18 OMITTED]
[FIGURE 19 OMITTED]
Figures 20 and 21 give a "bird's eye-view" of our experimental observations complementary to Figures 17-19. These summary snapshots can give a useful overview of the process data if one is interested in, say, an optimization problem. One can easily establish from these summary figures a "corridor of operation" for the process in question, in addition to a summary of the sensitivity of the process (process outputs and hence properties) to specific important factors (inputs). For instance, if polydispersity (PDI) were the sole variable in an optimization performance index, and one were interested in minimizing PDI, then information such as in Figures 17-19 would suggest choosing a range for R values between 1.1 and 1.2. Figures 20 and 21 subsequently would allow one to consider additional variables in a multi-objective optimization framework (Figure 20 shows different measurements corresponding to a polymerization duration of 10 h; Figure 21 shows measured properties corresponding to 50% conversion). For instance, if one would like to minimize time to reach 50% conversion at 120[degrees]C, at the same time keeping PDI at low values and, for example, achieving an average molecular weight target of not more than 20000, then Figure 21 would guide one to pick a value of R ~ 1. In this way, one can easily appreciate potentially conflicting trends and make practical compromises.
[FIGURE 20 OMITTED]
[FIGURE 21 OMITTED]
Effect of Temperature
Figure 22 shows the effect of temperature on polymerization rate as indicated by conversion versus time, for [TEMPO] /[BPO] = 0.9 and 1.1. As expected, polymerization proceeds faster at 130[degrees]C than at 120[degrees]C. The effect of temperature is more pronounced at [TEMPO] /[BPO] = 1.1 (Figure 22b). Veregin et al. (199Gb) have illustrated the effect of temperature on rate of polymerization in NMRP of styrene at [TEMPO]/[BPO] = 1.1 with T = 115, 125, 135[degrees]C. Their results (in good agreement with our experimental data) only capture the first 8 h of the experiment while our data points give a far more detailed picture (up to 72 h).
Figure 23 shows profiles for number- and weight-average molecular weights at [TEMPO]/[BPO] = 0.9. A slight reduction in the values of [M.sub.n] and [M.sub.w] with respect to the profile obtained at 120[degrees]C, is observed at 130[degrees]C. In conventional free radical polymerization molecular weight decreases due to the increased rate of bimolecular termination when temperature increases, which dominates over the increase of propagation rate. In any CRP process, the effect of bimolecular termination is suppressed to a great extent; so increasing temperature promotes the controlled growth of the polymer chains, with the increase in termination step not being significant. That is why there is no significant change in the molecular weight versus conversion picture when temperature increases from 120 to 130[degrees]C. Figure 24 shows the corresponding plots for [TEMPO]/[BPO] = 1.1. The general trends are reasonable and similar to the ones observed at [TEMPO]/[BPO] = 0.9. However, the effect of temperature is more prominent at [TEMPO]/[BPO] = 0.9.
[FIGURE 22 OMITTED]
This can be explained based on the polymerization degree (run length; see also Wang et al., 2005). As shown in Equation (9), run length per activation cycle (RLPAC) is calculated as the ratio of propagation rate over deactivation rate
run length per activation cycle (RLPAC) = [R.sub.p]/[R.sub.d] [k.sub.p][M][R.]/[k.sub.da][X.][R.] = [k.sub.p][M]/[k.sub.da][X.] (9)
Figure 25 illustrates RLPAC at R = 0.9 for two different temperatures. Comparing this figure with its counterpart at R = 1.1 (Figure 2G) shows that the effect of temperature on RLPAC is more pronounced at R = 0.9. According to Wang et al. (2005), RLPAC represents the mean polymerization degree increase of a chain for each activation/deactivation cycle. So higher RLPAC results in higher molecular weights (given the fact that the number of cycles is almost the same for both cases). This could be the explanation for a more pronounced temperature effect at R = 0.9 (contrast again Figures 23 and 24).
[FIGURE 23 OMITTED]
Figure 27 shows the effect of temperature on polydispersity for [TEMPO]/[BPO] = 0.9 and 1.1. Although slightly higher polydispersity values are obtained at the lower temperature (120[degrees]), in general the effect of temperature on polydispersity values is slight.
Veregin et al. (1996a) conducted experiments for [TEMPO]/[BPO] = 1.1 at 115, 125, and 135[degrees]C. Their experimental data showed higher polydispersity values in comparison to our experimental data. The reason could be related to the different polymerization and/or measurement methods used. It has been shown that using different polymerization methods can affect polymerization rate, molecular weights and polydispersity experimental data (Roa-Lung et al., 2007). Faster polymerization rates, lower molecular weights and more scattered polydispersity values are obtained using Schlenk techniques for polymerization (which is similar to the method employed by Veregin et al. (1996a)) in comparison to ampoule polymerizations (Roa-Lung et al., 2007), which was used by our group.
[FIGURE 24 OMITTED]
[FIGURE 25 OMITTED]
[FIGURE 26 OMITTED]
[FIGURE 27 OMITTED]
Experimental data (including polymerization rates, molecular weights and polydispersities) for NMRP of styrene at various operating conditions were collected over the full conversion range. The extensive database collected is appropriate as a back-up of future modelling and parameter estimation studies, and includes confirmations from independent replication runs, probably for the first time in the history of such investigations (along with Roa-Luna et al., 2007). The experimental work showed that increasing temperature increases the rate of polymerization while slightly decreasing molecular weight averages. It was also observed that increasing the ratio of controller to initiator decreases both the rate of polymerization and molecular weights. This is also the first time that polydispersity data were collected more systematically even at the early stages of conversion, thus giving a full picture and at the same time confirming (unsubstantiated so far) mathematical model predictions.
The authors wish to acknowledge financial support from the Natural Sciences and Engineering Research Council (NSERC) of Canada, the Canada Research Chair (CRC) program, Consejo Nacional de Ciencia y Tecnologia (CONACYT) (Project CIAM U40259-Y) (Mexico), and CNPq (Brazil), through a special Inter American Materials Collaboration (IAMC or CIAM) grant. Finally, many thanks to Martha Roa-Lung (UNAM, Mexico), and Juliana Belicanta Ximenes (UNICAMP, Brazil), for some independent experimental confirmations.
NOMENCLATURE [degrees]C denotes degrees Celsius [D] dimer concentration K equilibrium constant ([k.sub.da]/[k.sub.a]) [k.sub.a] rate constant for activation [k.sub.da] rate constant for deactivation [k.sub.fD] rate constant for transfer to dimer [k.sub.fM] rate constant for transfer to monomer [k.sub.f] rate constant for initiation [k.sub.p] rate constant for propagation [k.sub.tc] rate constant for termination by combination [k.sub.td] rate constant for termination by disproportionation ln([M.sub.0]/M) logarithmic conversion [M.sub.n] number-average molecular weight [M.sub.w] weight-average molecular weight M[W.sub.M] monomer molecular weight [M] monomer concentration [N[O.sup.*.sub.x]] concentration of nitroxyl radicals R [TEMPO]/[BPO] molar ratio [[R.sup.*]] total concentration of active (living) radicals [R.sup.*] active (living) radical [R.sub.d] deactivation rate [R.sub.p] propagation rate t time T temperature [[X.sup.*]] total TEMPO concentration [symbol] denotes concentration (of symbol) [[symbol].sub.0] denotes initial concentration (of symbol) Greek Symbols [[delta].sub.i] ith-order moment for dormant species [[lambda].sub.i] ith-order moment for active (living) radicals [[mu].sub.i] ith-order moment for dead polymer molecules [DELTA] denotes heating in Equation (1)
Manuscript received February 29, 2008; revised Manuscript received May 28, 2008; accepted for publication Jiame 4, 2008.
Auschra, C., E. Eckstein, A. Muhlebach, M. O. Zink and F. Rime, "Design of New Pigment Dispersants by Controlled Radical Polymerization," Prog. Org. Coat. 45, 83-93 (2002).
Bonilla, J., E. Saldivar, A. Flores-Tlacuahuac, E. Vivaldo-Lima, R. Pfaendner and F. Tiscareno-Lechuga, "Detailed Modeling, Simulation, and Parameter Estimation of Nitroxide Mediated Living Free Radical Polymerization of Styrene," Polym. React. Eng. J. 10, 227-263 (2002).
Braunecker, W A. and K. Matyjaszewski, "Controlled/Living Radical Polymerization: Features, Developments, and Perspectives," Prog. Polym. Sci. 32, 93-146 (2007).
Cunningham, M. F., K. Tortosa, M. Lin, B. Keoshkerian and M. K. Georges, "Influence of Camphorsulfonic acid in Nitroxide-Mediated Styrene Miniemulsion Polymerization," J. Polym. Sci. Part A: Polym. Chem. 40, 2828-2841 (2002a).
Cunningham, M. F., K. Tortosa, J. W Ma, K. B. McAuley, B. Keoshkerian and M. K. Georges, "Nitroxide Mediated Living Radical Polymerization in Miniemulsion," Macromol. Symp. 182, 273-282 (2002b).
Dollin, D., A. R. Szkurhan and M. K. Georges, "Rapid Additive-Free TEMPO-Mediated Stable Free Radical Polymerizations of Styrene," J. Polym. Sci. Part A: Polym. Chem. 45, 5487-5493 (2007).
Fischer, H., "The Persistent Radical Effect in "Living" Radical Polymerization," Macromolecules 30, 5666-5672 (1997).
Fischer, H., "The Persistent Radical Effect: A Principle for Selective Radical Reactions and Living Radical Polymerization," Chem. Rev. 101, 3581-3610 (2001).
Fischer, H., "Criteria for Livingness and Control in Nitroxide Mediated and Related Radical Polymerizations," in "Advances in Controlled/Living Radical Polymerization," K. Matyjaszewski, Ed., American Chemical Society, Washington, DC (2003), pp. 10-23.
Fischer, H. and L. Radom, "Factors Controlling the Addition of Carbon-Centered Radicals to Alkenes," Macromol. Symp. 182, 1-14 (2002).
Fukuda, T., "Fundamental Kinetic Aspects of Living Radical Polymerization and the Use of Gel Permeation Chromatography to Shed Light on them," J. Polym. Sci. Part A: Polym. Chem. 42, 4743-4755 (2004).
Fukuda, T. and A. Goto, "Gel Permeation Chromatographic Determination of Activation Rate Constants in Nitroxide-Controlled Free Radical Polymerization. 2. Analysis of Evolution of Polydispersities," Macromol. Rapid Commun. 18, 683-688 (1997).
Fukuda, T., Y Tsujii and T. Miyamoto, "Mechanism and Kinetics of Nitroxide-Controlled Free Radical Polymerization," Polym. Preprints 38, 723-724 (1997).
Georges, M. K., "An overview of the Nitroxide-Mediated Living-Radical Polymerization Process," Polym. Mater. Sci. Eng. 80, 283-284 (1999).
Georges, M., R. Veregin, R. Kazmaier and G. Hamer, "Narrow Molecular Weight Resins by a Free-Radical Polymerization Process," Macromolecules 26, 2987-2988 (1993).
Georges, M. K., R. A. Kee, R. P. N. Veregin, G. K. Hamer and P. M. Kazmaier, "Nitroxide Mediated Free-Radical Polymerization Process--Autopolymerization," J. Phys. Org. Chem. 8, 301-305 (1995).
Goto, A. and T. Fukuda, "Effects of Radical Initiator on Polymerization Rate and Polydispersity in Nitroxide-Controlled Free Radical Polymerization," Macromolecules 30, 4272-4277 (1997).
Goto, A. and T. Fukuda, "Kinetics of Living Radical Polymerization", Prog. Polym. Sci. 29, 329-385 (2004).
Goto, A., T. Terauchi, T. Fukuda and T. Miyamoto, "Gel Permeation Chromatographic Determination of Activation Rate constants in nitroxide-controlled free radical polymerization .1. Direct Analysis by Peak Resolution," Macromol. Rapid Commun. 18, 673-681 (1997).
Greszta, D. and K. Matyjaszewski, "Mechanism of Controlled/"Living" Radical Polymerization of Styrene in the Presence of Nitroxyl Radicals. Kinetics and Simulations," Macromolecules 29, 7661-7670 (1996).
Hawker, C. J., G. G. Barclay, A. Orellana, J. Dao and W Devonport, "Initiating Systems for Nitroxide-Mediated "Living" Free Radical Polymerizations: Synthesis and Evaluation," Macromolecules 29, 5245-5254 (1996).
Hawker, C. J., D. Mecerreyes, E. Elce, J. L. Dao, J. L. Hedrick, I. Barakat, P. Dubois, R. Jerome and W Volksen. ""Living" Free Radical Polymerization of Macromonomers: Preparation of Well Defined Graft Copolymers," Macromol. Chem. Phys. 198, 155-166 (1997).
Hawker, C. J., D. Benoit, F. Rivera, M. Piotti, I. Rees, J. L. Hedrick, C. Zech, G. Maier, B. Voit, R. Braslau and J. M. J. Frechets, "Versatile Route to Functionalized Block Copolymers by Nitroxide-Mediated "Living" Free-Radical Polymerization," Abstr. Papers Am. Chem. Soc. 218, U411-U412 (1999).
Ide, N. and T. Fukuda, "Nitroxide-Controlled Free-Radical Copolymerization of Vinyl and Divinyl Monomers. Evaluation of Pendant-Vinyl Reactivity," Macromolecules 30, 4268-4271 (1997).
Le Mercier, C., S. B. Acerbis, D. Bertin, F. Chauvin, D. Gigmes, O. Guerret, M. Lansalot, S. Marque, F. Le Moigne, H. Fischer and P. Tordo, "Design and Use of Beta-Phosphorus Nitroxides and Alkoxyamines in Controlled/"Living" Free Radical Polymerizations," Macromol. Symp. 182, 225-247 (2002).
Leduc, M. R., C. J. Hawker, J. Dao and J. M. J. Frechet, "Dendritic Initiators for "Living" Radical Polymerizations: A Versatile Approach to the Synthesis of Dendritic-Linear Block Copolymers," J. Am. Chem. Soc. 118, 11111-11118 (1996).
MacLeod, P. J., R. P. N. Veregin, P. G. Odell and M. K. Georges, "Stable Free Radical Polymerization of Styrene: Controlling the Process with Low Levels of Nitroxide," Macromolecules 30, 2207-2208 (1997).
Malmstrom, E. E. and C. J. Hawker, "Macromoleculer Engineering via 'Living' Free Radical Polymerizations," Macromol. Chem. Phys. 199, 923-935 (1998).
Malmstrom, E., R. D. Miller and C. J. Hawker, "Development of a New Class of Rate-Accelerating Additives for Nitroxide-Mediated 'Living' Free Radical Polymerization," Tetrahedron 53, 15225-15236 (1997).
Mardare, D. and K. Matyjaszewski, "Thermal Polymerization of Styrene in the Presence of Stable Radicals and Inhibiters," Polym. Preprints 35, 778-779 (1994).
Matyjaszewski, K., "Controlled/Living Radical Polymerization: State of the Art in 2002," in "Advances in Controlled/Living Radical Polymerization," K. Matyjaszewski, Ed., American Chemical Society, Washington, DC (2003), pp. 2-9.
Odell, P. G., R. P. N. Veregin, L. M. Michalak, D. Brousmiche and M. K. Georges, "Rate Enhancement of Living Free-Radical Polymerizations by an Organic-Acid Salt," Macromolecules 28, 8453-8455 (1995).
Pirrung, F. O. H. and C. Auschra, "Nitroxide-Mediated Synthesis of Acrylic Block Copolymers- and their Use as Coating Additives," Polym. Preprints 46, 316-317 (2005).
Roa-Lung, M., A. Nabifar, M. P. Diaz-Barber, N. T. McManus, E. Vivaldo-Lima, L. M. F. Lona and A. Penlidis, "Another Perspective on the Nitroxide Mediated Radical Polymerization (NMRP) of Styrene Using 2,2,6,6-Tetramethyl-1-Piperidinyloxy (TEMPO) and Dibenzoyl Peroxide (BPO)," J. Macromol. Sci. Part A: Pure Appl. Chem. 44, 337-349 (2007).
Rutsch, W and M. A. Cech, "Effects to Improve the Quality of Life: Colour, Performance and Protection from Ciba Specialty Chemicals," Chimia 61, 33-41 (2007).
Sobek, J., R. Martschke and H. Fischer, "Entropy Control of the Cross-Reaction between Carbon-Centered and Nitroxide Radicals," J. Am. Chem. Soc. 123, 849-2857 (2001).
Tang, C. B., T. Kowalewski and K. Matyjaszewski, "Preparation of Polyacrylonitrile-Block-Poly(n-Butyl Acrylate) Copolymers Using Atom Transfer Radical Polymerization and Nitroxide Mediated Polymerization Processes," Macromolecules 36, 1465-1473 (2003).
Tang, C., B. Dufour, L. Bombalski, K. Matyjaszewski and T. Kowalewski, "Well-Defined Polyacrylonitrile Copolymers Prepared by Various CRP and their Applications as Nanostructured Carbons," Polym. Preprint 46, 497-498 (2005).
Tsujii, Y, T. Fukuda and T. Miyamoto, "Computer Simulation on the Kinetics of Nitroxide-Controlled Free Radical Polymerization," Polym. Preprints 38, 657-658 (1997).
Veregin, R. P. N., M. K. Georges, G. K. Hamer and P. M. Kazmaier, "Mechanism of Living Free-Radical Polymerizations with Narrow Polydispersity--Electron-Spin-Resonance and Kinetic-Studies," Macromolecules 28, 4391-4398 (1995).
Veregin, R. P. N., P. G. Odell, L. M. Michalak and M. K. Georges, "Molecular Weight Distributions in Nitroxide-Mediated Living Free Radical Polymerization: Kinetics of the Slow Equilibria between Growing and Dormant Chains," Macromolecules 29, 3346-3352 (1996a).
Veregin, R. P. N., P. G. Odell, L. M. Michalak and M. K. Georges, "The Pivotal Role of Excess Nitroxide Radical in Living Free Radical Polymerizations with Narrow Polydispersity," Macromolecules 29, 2746-2754 (1996b).
Veregin, R. P. N., P. M. Kazmaier, P. G. Odell and M. K. Georges, "The Role of Thermal Initiation in Nitroxide Mediated Living Free Radical Polymerizations," Chem. Lett. 5, 467-468 (1997).
Wang, Y, R. A. Hutchinson and M. F. Cunningham, "A Semi-Batch Process for Nitroxide Mediated Radical Polymerization," Macromol. Mater. Eng. 290, 230-241 (2005).
Ximenes, B. J., P. V. R. Mesa, L. M. F. Lona, E. Vivaldo-Lima, N. T. McManus and A. Penlidis, "Simulation of Styrene Polymerization by Monomolecular and Bimolecular Nitroxide-Mediated Radical Processes Over a Range of Reaction Conditions," Macromol. Theory Simulat. 16, 194-208 (2007).
([dagger]) A. Penlidis dedicates this paper to Professor John F. MacGregor, who taught him how to appreciate process data. Thank you, JFM
Afsaneh Nabifar, (1) Neil T. McManus, (1) Eduardo Vivaldo-Lima, (2) Liliane M.F. Lona (3) and Alexander Penlidis (1) *
(1.) Department of Chemical Engineering, Institute for Polymer Research (IPR), University of Waterloo, Waterloo, Ontario, Canada N2L 3G1
(2.) Facultad de Quimica, Departamento de Ingenieria Quimica, Universidad Nacional Autonoma de Mexico (UNAM), Conjunto E, Ciudad Universitaria, Mexico CP 04510, D.F., Mexico
(3.) Department of Chemical Processes, School of Chemical Engineering, State University of Campinas (UNICAMP), P.O. Box 6066, Campinas, Sao Paulo, Brazil
* Author to whom correspondence may be addressed. E-mail address: perlidis @a cape.uwaterloo.ca
Table 1. Selective significant contributions to NMRP Year (a) Main topics Brief remarks 1993 Kinetics and fundamentals of First introduction of bimolecular initiation of the concept of NMRP for styrene with BPO and TEMPO producing low (role of excess nitroxide polydispersity (Veregin et al., 1996b), role polystyrene; Bimolecular of thermal initiation (Georges approach et al., 1995; Veregin et al., 1997), mechanism and kinetics (Veregin et al., 1995, 1996a; Georges, 1999), rate in NMRP (Odell et al., 1995; Cunningham et al., 2002a)) NMRP in aqueous phase (emulsion, miniemulsion, etc.) (Cunningham et al., 2002b) 1994 Mechanism and other Kinetic and mechanistic fundamental aspects (Mardare studies and Matyjaszewski, 1994; Greszta and Matyjaszewski, 1996; Matyjaszewski, 2003; Braunecker and Matyjaszewski, 2007) Synthesis of well defined block and graft copolymers (Tang et al., 2003, 2005) 1996 Fundamental kinetic aspects, Mainly using model simulations for NMRP unimolecular initiators (Fukuda et al., 1997; Goto and as mediators; Kinetic Fukuda, 1997, 2004; Tsujii et studies al., 1997) Determination of kinetic rate constants in NMRP by gel permeation chromatography (Fukuda and Goto, 1997; Goto et al., 1997; Fukuda, 2004) Copolymerization of styrene and divinyl-biphenyl (Ide and Fukuda, 1997) 1996 Development of variety of Chemistry and synthetic TEMPO-based unimolecular aspects initiators to examine the effect of structural variation on the efficiency of these derivatives (Hawker et al., 1996) Examining the effect of acylating agents as rate-accelerating additives (Malmstrom et al., 1997) Synthesis of complex macromolecular architectures, star, hyperbranched and dendritic polymers as well as block and graft copolymers (Leduc et al., 1996; Hawker et al., 1997, 1999; Malmstrom and Hawker, 1998) 1997 Introducing the concept of Chemistry aspects and persistent radical effect in kinetic studies NMRP (Fischer, 1997, 2001) Defining criteria for livingness and control in NMRP (Fischer, 2003) Design and synthesis of [beta]-phosphorus nitroxides and alkoxyamines to be used instead of TEMPO nitroxide (Le Mercier et al., 2002) Kinetic investigations on cross-reaction between carbon-centered and nitroxide radicals (Sobek et al., 2001; Fischer and Radom, 2002) (a) Year (approximate) that the specific group published their first paper on controlled radical polymerization. Table 2. Summary of experimental runs Experiment Temperature [[BPO].sub.0] M no. ([degrees]C) 1 120 0.036 2 0.036 3 0.036 4 0.036 5 130 0.036 6 0.036 7 0.036 Experiment [TEMPO]/ Remarks no. [BPO] 1 0.9 2 1.1 + Replicate 3 1.2 4 1.5 5 0.9 6 1.1 + Replicate 7 1.3 + Replicate