Of Microspheres In Dispersion Ringopening Polymerization Of eCaprolactone And Lactides

Understanding how particles are formed and how they grow is essential for comprehensive description of dispersion polymerization of cyclic esters. In principle, the following basic mechanisms of particle formation and growth should be taken into account.

1. According to the first mechanism, initiation is slow. New chains are formed in solution throughout the polymerization process. In solution they grow fast, nucleating new particles or becoming adsorbed onto the already existing ones. Coalescence of particles is absent. In polymer particles due to the low local monomer concentrations propagation stops or proceeds very slowly. Because according to this model the new polymer chains and new particles are continuously formed in solution, particle diameters and polyester molecular weight distributions should be broad. For polymerizations proceeding according to this model, the concentration of initiator and/or propagating centers in the continuous phase should decrease slowly throughout the polymerization process.

2. The second mechanism is similar to the first, but with efficient coalescence of small, not properly stabilized particles until their total surface becomes small enough to be sufficiently saturated with surfactant, providing the necessary stabilization.

3. According to the third mechanism, all chains are initiated in the initial very short period. When growing chains reach the critical length they aggregate into nuclei of microspheres. The microspheres are swollen with monomer and further polymerization proceeds within these particles. Propagation stops when equilibrium monomer concentration is reached. In this type of polymerization all species suitable for reaction with monomer disappear from the liquid phase at a very early stage of polymerization. In an absence of particle coalescence all formed microspheres grow in the same manner. Their number should be constant and one could expect a narrow diameter distribution.

4. The last conceivable mechanism would be similar to the one described in point 3 but allowing for coalescence of particles. In the case of such polymerization, the diameter distribution should be broad.

The mechanism of particle formation and growth not only should affect particle diameter distribution but also, due to the different local concentrations of monomer and active species, should have an influence on the kinetics of polymerization and on molecular weights and molecular weight distributions in mi-crospheres.

The question of whether initiator and propagating active centers are present for a long time in the continuous phase (as for polymerizations according to mechanisms l and 2) or are quickly transferred to microspheres has been answered on the basis of studies of the dispersion polymerization of e-caprolactone initiated with diethylaluminum ethoxide [53]. During polymerization (initial monomer and initiator concentrations = 4.1 x 10-1 and 1.66 x 10-1 mol/L, respectively; reaction medium 1,4-dioxane/heptane 1:9 v/v, room temperature), samples of reaction mixture (known volume) were withdrawn at various moments. Each sample was added to heptane-containing acetic acid terminating the propagation. Microspheres were isolated by centrifugation and content of aluminum-containing compound in supernatant was determined by 8-hydroxy-quinoline method [53]. Isolated microspheres were dissolved in THF and molecular weight of polymer that constituted these particles was determined by GPC [calibration with poly(e-caprolactone) samples with narrow molecular weight distribution].

Figure 8 illustrates the dependence of concentration of aluminum-containing centers in supernatant on time of polymerization. The relation between molecular weight of poly(e-caprolactone) in microspheres and fraction of active centers in microspheres (assuming that fraction of active centers in microspheres equals ([AC(S)]0 - [AC(S)])/[AC(S)]0, where [AC(S)]0 and [AC(S)] denote the initial and actual concentration of active centers in liquid phase, respectively, is shown in Fig. 9.

Following are discussed characteristic features of relations illustrated by Figs. 8 and 9. After 150 s from the beginning of initiation, the fraction of active

FIG. 8 Polymerization of e-caprolactone. Dependence of concentration of propagating species in solution and time of polymerization. Polymerization conditions: [e-caprolac-tone]0 = 4.1 X 10-1_mol/L, [diethylaluminumalkoxide]0 = 1.66 X 10-1, mol/L, P(DA-CL) 1.6 g/L [Mn(CL)/Mn(DA-CL) = 0.125]. (From Ref. 53.)

FIG. 8 Polymerization of e-caprolactone. Dependence of concentration of propagating species in solution and time of polymerization. Polymerization conditions: [e-caprolac-tone]0 = 4.1 X 10-1_mol/L, [diethylaluminumalkoxide]0 = 1.66 X 10-1, mol/L, P(DA-CL) 1.6 g/L [Mn(CL)/Mn(DA-CL) = 0.125]. (From Ref. 53.)

centers in microspheres exceeds 96%. Moreover, number average molecular weight of poly(e-caprolactone) at this stage constitutes only 18% of molecular weight of polymer in microspheres after complete monomer conversion. These findings indicate that in dispersion polymerization of e-caprolactone propagation of all chains is initiated very early and that essentially all growing chains shortly after initiation, when their molecular weight is in the range 500-1000, become incorporated in microspheres. Subsequently, further propagation proceeds in the already formed microspheres. Thus, the above-described experiments conform to mechanisms 3 and/or 4, describing formation and growth of microspheres, and definitely exclude mechanisms 1 and 2.

Discrimination between mechanisms 3 and 4 was possible after monitoring changes of the concentration of microspheres with time. For dispersion polymerizations of e-caprolactone and l,l-lactide initiated with anionic and pseudoan-ionic initiators, the number of microspheres in a given volume of polymerizing mixtures was determined at various monomer conversions [54-57].

Microspheres were counted on Burker's plate using an optical microscope. Burker's plate is a device commonly used in medicine and biology for counting blood cells and other cells suspended in a liquid environment. It consists of a

FIG. 9 Polymerization of e-caprolactone. Relation between molecular weight of poly-(e-caprolactone) in microspheres and fraction of active centers in microspheres. Polymerization conditions the same as for Fig. 8. (From Ref. 53.)

quartz or glass plate with wells with perpendicular walls and of a cover slide that when placed on the plate closes tops of the wells, thus limiting their volumes. A drop of polymerizing mixture placed on the plate and covered with the cover glass is divided into several wells, allowing for several parallel determinations and thus for determination of the average particle numbers with better precision.

In spite of limitations in using optical microscopy for studies of microspheres with diameters smaller than a few micrometers (due to diffraction of light at the edges of light-scattering objects, spatial dimensions of particles with diameters comparable or only slightly larger than the light wavelength cannot be determined with a reasonable precision), this method can be used for detection of particles with diameters as small as 0.1 pm. Thus, analysis of optical microscopy pictures of wells in the Burker's plate filled with samples of polymerizing mixture allowed determinations of particle concentration. Examples of results of such determinations for polymerizations of e-caprolactone and l,l-lactide are shown in Fig. 10. Plots in the figure revealed that for monomer conversions higher than about 20% of monomer conversion the number of particles did not change. Thus, any coalescence of particles that occurred in dispersion polymerization of e-caprolactone and/or lactide would have happened only at the initial

FIG. 10 Dependence of concentration of microspheres on normalized monomer conversion. A, [£-caprolactone]0 = 4.3 X 10-1 mol/L, [(CH3CH2)2AlOCH2CH3]0 = 5.60 X 10-3 mol/L; B, [e-caprolactone]0 = 4.2 X 10-1 mol/L, [(CH3)3SiONa]0 = 1.83 X 10-3 mol/L; C, [l,l-lactide]0 = 4.4 X 10-1 mol/L, [Tin(II) 2-ethylhexanoate]0 = 7.16 X 10-4 mol/L. (From Ref. 56.)

FIG. 10 Dependence of concentration of microspheres on normalized monomer conversion. A, [£-caprolactone]0 = 4.3 X 10-1 mol/L, [(CH3CH2)2AlOCH2CH3]0 = 5.60 X 10-3 mol/L; B, [e-caprolactone]0 = 4.2 X 10-1 mol/L, [(CH3)3SiONa]0 = 1.83 X 10-3 mol/L; C, [l,l-lactide]0 = 4.4 X 10-1 mol/L, [Tin(II) 2-ethylhexanoate]0 = 7.16 X 10-4 mol/L. (From Ref. 56.)

stage. Therefore, all results of the above-mentioned studies conformed to the third mechanism listed at the beginning of this section.

Comprehensive description of particle formation and growth requires finding a law to determine concentrations of microspheres in a polymerizing system. For polymerizations of l,l-lactide carried out at various ratios of monomer and initiator concentrations, a linear relationship has been observed between the average mass of microsphere (Mm,) and mass of monomer molecules per mole of propagating chains [57]. An example of such dependence for polymerization of l,l-lactide initiated with 2,2-dibutyl-2-stanna-1,3-dioxepane is shown in Fig. 11 in which Mmn has been plotted as a function of ([l,l-Lc]0 - [l,l-Lc]e)144.13/ [T]0. It must be remembered that ([l,l-Lc]0 and [l,l-Lc]e denote the initial and equilibrium monomer concentrations, [T]0 denotes the initial concentration of initiator which for quantitative initiation is equal to concentration of growing chains, and 144.13 is the molar mass of l,l-lactide. The ratio of the average mass of microsphere and mass of momomer molecules per mole of growing

FIG. 11 Number-averaged mass of microsphere as a function of mass of monomer converted to polymer per mole of growing chains (([l,l-Lc]0 - [l,l-Lc]e)144.13/[/]0). Dispersion polymerization of l,l-lactide initiated with 2,2-dibutyl-2-stanna-l,3-dioxe-pane. (From Ref. 57.)

chains equals the number of moles of growing chains per microsphere. Thus, the average number of growing chains per microsphere (NC) equals:

[l,l-Lc]0 - [l,l-Lc]e where NA is Avogadro's number.

From the plot in Fig. 11 in which slope equals 3 x 10-16 (144.13[/]0Mmn/ ([l,l-Lc]0 - [l,l-Lc]£)) the average number of growing chains per microsphere in all polymerizations was 1.8 x 108. The linear dependence of Mmn on ([l,l-Lc]0 - [l,l-Lc]e)144.13/[/]o suggests that regardless of concentration of growing chains (i.e., initiator concentration) the number of propagating chains per micro-sphere was similar. Indeed, a plot of NC as a function of the initial initiator concentration indicated that for initial concentrations of 2,2-dibutyl-2-stanna-1,3-dioxepane varied from 8.0 x 10-4 mol/L to 1.02 x 10-2 mol/L the average number of growing chains in microsphere varied in a narrow range from 1.61 x 108 to 1.99 x 108 (cf. Fig. 12, [57]).

FIG. 12 Relation between anumber of propagating chains per microsphere (NC) and concentration of initiator (2,2-dibutyl-2-stanna-1,3-dioxepane) in dispersion polymerization of l,l-lactide. (From Ref. 57.)

At first it seems strange that at the beginning of polymerization, in spite of various concentrations of initiator, always the same very large number of propagating chains (about 2 X 108) form the nucleus of a microsphere. However, it could be shown that such a characteristic feature conforms to a rather simple model describing nucleation of microspheres. According to this model, at the stage at which growing chains reach their critical length (for polymerization of e-capro-lactone when the molecular weight of macromolecules is between 500 and 1000) they disappear from solution in one of the following two processes: nucleation of new microspheres or adsorption onto existing ones. Nucleation of new micro-spheres could be described by a differential equation corresponding to a process in which collision between two macromolecules creates nuclei of a microsphere:

dt 2

In Eq. (5) dN/dt denotes the rate of formation of new microspheres (N represents the concentration of microspheres expressed as number of microspheres in a unit volume), n the concentration of growing chains (also expressed as number of chains in a unit volume), and h1 the rate constant of microsphere nucleation. The coefficient / reflects creation of a new microsphere by collision of two chains.

The rate at which growing chains become incorporated into microspheres (dn/dt) could be described as a sum of the rate at which chains disappear from solution due to nucleation of microspheres (Rj) and the rate at which chains are removed from solution due to adsorption onto the existing particles (R2).

In Eq. (7), kA denotes rate of adsorption and S the average surface of one microsphere (i.e., R2 is proportional to the total surface of microspheres in a unit volume of propagating mixture). The relation between surface and volume of spherical particle is as follows:

Based on the simplifying assumption that formation of particles is so fast that during completion of this process propagating chains do not considerably change in length (molecular weight of aggregating chains was found to be in the range from 500 to 1000; in further estimations it was taken equal 750), it was possible to express the average volume of nucleated microsphere as:

In Eq. (9), NA denoted Avogadro's number, d density of polymer in microspheres [e.g., for poly(lactide) 1.25 g/cm3], N concentration of microspheres, n and n0 concentration of aggregating chains at time t and at the beginning of aggregation, respectively.

Introducing k2 = 400kA/VNAd2, R2 could be expressed as: nV(n - n0)2

and the total rate with which polymer chains disappear from solution (dn/dt = R1 + R2) would be described by an equation:

Numerical solution of a set of Eqs. (5) and (12), for any given initial concentration of chains (n2) and values of k and k2, allows calculations of changes of a number of microspheres during a process of particle formation. Results of such calculations (with k = 300 L/mol-s and k2/k = 7.36 x 104) for initial concentration of chains (n0) varied from 10-3 mol/L to 5 x 10-2 mol/L are shown in Fig. 13. Plots in the figure indicate that in the chosen range of propagating chains concentrations (corresponding to a typical range of initiator concentrations) the final concentrations of microspheres are the same and equal 1.8 x 108 (as in experimental studies of the above-described polymerization of l,l-lactide).

Was this article helpful?

0 0

Post a comment