The systematic study of the microstructure of crosslinked copolymers from siloxane macromonomers and methacrylates by changes in composition and components

Introduction

Polymerization-induced phase separation¹ is a very convenient method to create microheterophase structures from a homogeneous solution containing reactive monomeric compounds and polymeric compounds. The technique has been reported in many fields, including in the polymerizations of monomer/polymer mixtures² and hydrogels.³

While progress has been made in this field, many radically polymerizable silicone compounds have been developed for use in silicone polymer products, especially with respect to medical devices, such as contact lenses, intraocular lenses and artificial lungs. In the course of improving the limitations associated with silicone, such as fragility and hydrophobicity, the introduction of silicone compounds with long Si-O-Si chains (for example, polydimethylsiloxane (PDMS)) to the polymer backbone have been developed, which leads to microheterophase structure of polymer. One of these approaches is the introduction of PDMS chains by condensation reactions with rigid segments, such as polyamide and polyimide groups, to create multi-block copolymers.^{4, 5} Another approach is the polymerization-induced phase separation method that consists of radical copolymerization of methacrylates with macromers that contain radically polymerizable groups at the end of the PDMS chain and crosslinkers that have multiple radically polymerizable groups.

With the latter approach, an amphiphilic co-network created from the combination of PDMS macromonomer and hydrophilic vinyl monomer^{6, 7} has been reported, as has been the crosslinked copolymer from PDMS macromers and hydrophobic vinyl monomers.^{8, 9} Previous research on the amphiphilic network focused on analysis of the structure or dynamics for a specific polymer composition. Research on hydrophobic crosslinked copolymers focused on the properties of various polymerizable groups at the PDMS tail end. Furthermore, as far as we know, no articles have been published that investigated the microstructure of PDMS-bearing copolymers at the molecular level with solid-state NMR.

In the radical copolymerization approach, a systematic study is desired for the relationship between composition and physical properties as a function of the various monomers and PDMS macromonomers of various molecular weight (M_n). However, no articles using such an approach were found. We have been studying crosslinked copolymers produced from siloxane macromonomers with varied PDMS chain lengths, various methacrylate monomers and crosslinkers. As part of this study, we have already reported the high performance of products created using the polymerization-induced phase separation method that exhibit microheterophase structure. The materials showed very high oxygen permeability and durability when compared with a material with a low-M_n silicone constituent and that was as uniform a structure as contact lens material.¹⁰ In addition, it has been reported that the copolymerization of the perfluoroalkyl ester group-bearing monomer with PDMS-containing material caused an increase in oxygen permeability.¹¹ However, further analysis of microstructures of the copolymers is required at the molecular level to develop a copolymer material with superior characteristics.

In this study, we used ¹³C solid-state CP-MAS NMR measurements of the aforementioned copolymers because it provides fine-grain information with respect to the molecular mobility or relaxation time, which enables analysis of the internal structure of heterogeneous polymer composites.^{12, 13, 14} We chose methylmethacrylate (MMA) and trifluoroethylmethacrylate (TFEMA) as the methacrylates because they have been used as optical materials in medical materials, such as contact lenses, due to their high glass transition temperature (T_g) and transparency. TFEMA also has a low refractive index, similar to that of PDMS, and oxygen permeability was also considered. We believe that our approach to investigate the relationship between copolymer compositions and their physical properties with solid-state NMR will contribute greatly to the new field of the study of polymerization-induced phase separation products.

Experimental Procedure

Materials and detailed procedures for the preparation of the crosslinked copolymers from methacrylates/PDMS macromonomers and NMR measurements are provided in the Supplementary Information section. The molecular weights of the PDMS macromonomers are listed in Supplementary Table S1 in the Supplementary Information. Chemical structures of the monomers and crosslinker are shown in Figure 1. The compositions of the samples are shown in Table 1.

Figure 1

Chemical structures of the monomers and crosslinker.

Table 1 Composition^a of the specimens employed in this study

Results and discussion

CP-MAS spectra

Examples of the obtained CP-MAS spectra and assignments of the signals are shown in Figure 2. The peak patterns differ depending on the chemical structure of the monomers. In addition, the intensity at 1.5 p.p.m. (a) of specimen M147, with a PDMS M_n of 1700, was higher than those of specimens M442 and T446, with a PDMS M_n of 4700. It is reported that because the efficiency of cross-polarization inversely depends on mobility, a higher intensity signal corresponds to a lower mobility in the CP-MAS spectrum.¹⁴ Because the PDMS content is almost the same for all specimens (PDMS is 38–42 wt %), it was recognized from the signal at 1.5 p.p.m. (Figure 2a) that the larger PDMS segment (M_n=4700) is more mobile than the PDMS segment with an M_n of 1700. These results suggest that a higher PDMS M_n allows a higher diffusivity of gas molecules and results in a higher gas permeability. This shows good correspondence to our previous report that described a material with a microheterophase structure with a high M_n that showed a higher oxygen permeability than a material made from a low-molecular-weight silicone constituent.¹⁰

Figure 2

¹³C CP-MAS NMR spectra of crosslinked copolymers from methacrylates and the PDMS macromonomer. a, b, and c correspond to the following compositions. a: M442 (MMA/PDMS macromonomer with M_n=4700, PDMS=38 wt%), b: M147 (MMA/PDMS macromonomer with M_n=1700, PDMS=42 wt%), c: T446 (TFEMA/PDMS macromonomer with M_n=4700, PDMS=38 wt%).

Comparison of T₁ values from the copolymer and pristine polymer

The measured T₁^H, T_1ρ^H, T₁^C and T_1ρ^C values of pristine PDMS (I), PMMA (I) and PTFEMA (II) are shown with those of the copolymer in Figure 3a, b, c and d. The T₁^H, T_1ρ^H, T₁^C and T_1ρ^C values of pristine PDMS are clearly different from those of copolymerized PDMS. Furthermore, comparing the T₁ values of pristine PMMA and PTFEMA with those of copolymerized PMMA and PTFEMA, a composition dependency of the T₁ values can be seen. Thus, it was recognized that the copolymerization of the PDMS macromonomer with MMA or TFEMA resulted in a finely dispersed microstructure that is dependent on the composition and PDMS M_n value.

Figure 3

Measured T₁^H (a), T_1ρ^H (b), T₁^C (c) and T_1ρ^C (d) values. a, b, c, d, e, f, g, and h correspond to the peak signals in Figure 2. PDMS VF indicates the PDMS volume fraction. The dotted-line brackets in (d) show only the trends in the measured values and not the exact values. I and II indicate pristine PDMS (M_n=4700) combined with pristine PMMA, and pristine PTFEMA, respectively.

T₁^H vs PDMS M_n, PDMS volume fraction, and co-monomer

The proton spin-lattice relaxation time T₁^H can be used to elucidate the parts that have a relatively high mobility, on the order of MHz, and T₁^H measurements have been performed to obtain information with respect to the size of the domains in polymers.^{15, 16, 17, 18} The measured T₁^H values for all specimens, including pristine PDMS, PMMA and PTFEMA, are shown in Figure 3a. In addition, the T₁^H values for the PDMS macromonomer with M_n=1700 and 4700 were 1.16 s and 1.25 s, respectively.

Specimen M129 (MMA copolymer and 1700 M_n PDMS) had almost the same T₁^H values at all peaks (average value is 0.60 s), implying effective spin diffusion due to the low molecular mobility and implying a homogeneous dispersion structure. The diameter of the homogeneously distributed area (domain size) was calculated using the following formula¹⁶ and average T₁^H value:

where L represents the domain size, D is ∼10⁻¹² cm² s^–1 (spin diffusion coefficient), and t represents T₁^H.

Thus, this implies that in M129, PDMS and PMMA are dispersed homogeneously within regions that are 19 nm in diameter. Similarly, the specimens M147 and M432, which had different PDMS M_n values, showed T₁^H values with relatively small dispersion, and both were determined to have a similar diameter of homogeneously dispersed range.

Alternatively, T₁^H values at peak (a) for the specimens M442, T436 and T446, all with an M_n of 4700, showed higher values than those of M129, M147 and M432. Because longer T₁^H values correspond to higher degrees of molecular motion, this implies that a larger PDMS M_n results in more molecular motion, a higher diffusivity of gas molecules and a higher gas permeability, similar to what was observed using the CP-MAS spectra. A longer T₁^H value also corresponds to inefficient spin diffusion and results in phase separation of PDMS and PMMA or PTFEMA. Therefore, these results correspond well with our previous report.¹⁰

MMA copolymers with a higher PDMS content showed higher values of T₁^H at peak (a), as shown in Figure 3a. Alternatively, TFEMA copolymers with a higher PDMS content, T446, showed lower T₁^H values. At the same time, the dispersion of the T₁^H values of all peaks of T446 was smaller than that of T436, suggesting differences in the microstructure between T446 and other copolymers. Previously, different morphologies have been reported from copolymers with different ratios of constituents or different M_n values using TEM observations of the acrylate/PDMS macromonomer copolymer.⁹ Therefore, the aforementioned results suggest that a similar morphology in the TFEMA copolymer may appear at a lower PDMS content than that in the MMA copolymer. Furthermore, the smaller dispersion of T₁^H values of all peaks from T446 implies the finer structure or more continuous structure in T446 than that in T436 or M442. Thus, resulting differences in gas permeability for MMA copolymers and TFEMA copolymers can be expected. Therefore, the reported oxygen permeability increase¹¹ can be supported by the possible differences in morphology.

These observations with respect to the correlation between morphology and dispersion of T₁^H values could explain the T₁^H value at peak (a) in M432.

T_1ρ^H vs PDMS M_n, PDMS volume fraction, and co-monomer

The T_1ρ^H value reflects motion modes of the order of several kHz, which appear primarily in the main chain of the polymer. The relationship between the spin-diffusion phenomenon and molecular mobility is the same as with T₁^H.

As shown in Figure 3b, specimen M147, with a PDMS M_n of 1700 and a PDMS volume fraction of 0.47, showed minimum dispersion among all T_1ρ^H values. Because this implies effective spin diffusion, the diameter of the homogeneously distributed area (domain size) was calculated to be 2.6 nm using equation (1). This suggests that in M147, PDMS and PMMA are dispersed homogeneously in an area with a diameter of 2.6 nm. Similarly, the dispersion structure of M129 is considered to be the same as that of specimen M147, suggesting that the 1700 M_n PDMS macromonomer did not influence the microphase separation when the PDMS volume fraction ranged from 0.29 to 0.47. This suggests that a smaller M_n results in a transparent copolymer and could explain the results related to copolymer transparency shown in Table 1.

Comparison of the T_1ρ^H values of specimens with different PDMS M_n values suggested that a larger PDMS M_n value caused higher T_1ρ^H values at peak (a) from 13.8 and 12.4 ms to 23.6, 20.2, 25.1 and 18.1 ms. Specifically, the influence of -OSi(CH₃)₂- chain length on the molecular mobility and gas permeability can be observed using the T_1ρ^H value and the T₁^H value.

As shown in Figure 3b, a comparison of the T_1ρ^H values at peak (a) of the specimens with different PDMS volume fractions showed that higher PDMS volume fractions resulted in lower T_1ρ^H values. The magnitude of change in the T_1ρ^H value depended on the PDMS M_n and co-monomer type. It is interesting that the nature of the T_1ρ^H dependency on PDMS content is opposite to that of T₁^H.

Furthermore, T446 displayed a somewhat reduced dispersion of all T_1ρ^H values when compared with the distinct separation of T_1ρ^H values between peak (a) and the other peaks (b), (c), (e), (f), (g) and (h) in T436. In addition, it is interesting that the T_1ρ^H values at peak (g) (C in CF₃ group) and peak (h) (C in C=O group) are close in value to those of peak (a) in T446. A comparison of T_1ρ^H for T446 and M442 led to the same result. These results can be correlated to the possible existence of domain dispersion of the order of several nm or another morphology with finer structures or more continuous structures in T446 than those in T436 or M442. These differences in morphology could lead to the observed difference in gas permeability and support the previously reported results.¹¹

T₁^C Values

Results from the T₁^C measurements are shown in Figure 3c. Because the spin-diffusion phenomenon does not occur in the ¹³C nucleus, independent T₁^C values for each C atom were available. This relaxation time also reflects the molecular motion at a frequency of several MHz. Comparing the T₁^C data from M129 and M147, both with a PDMS M_n of 1700, it was determined that the T₁^C values of peaks (c), (d) and (h) decreased with an increase in PDMS content. Differences between M432 and M442, both with a PDMS M_n of 4700, could not be found. However, the decrease in T₁^C from peaks (c), (f) and (h) for specimens T436 and T446, both with a PDMS M_n of 4700, and the PTFEMA copolymer, was recognized with an increase in PDMS content. The decrease in T₁^C means an increase in the molecular mobility because these peaks result from PMMA or PTFEMA that have a high T_g value and the applicability of the relationship ω_cτ≫1, where ω_c is angular velocity of the ¹³C nucleus and τ is the correlation time. Therefore, these T₁^C values suggest a difference in the mobility of PMMA or PTFEMA when comparing the M129 and M147, as well as T436 and T446.

T_1ρ^C vs the PDMS M_n value, the PDMS volume fraction, and the co-monomer

T_1ρ^C yields information with respect to molecular motion on the order of several kHz, similar to T_1ρ^H, and independent T_1ρ^C values are available. Results from the measurements are shown in Figure 3d. T_1ρ^C values from peak (a) for pristine PDMS, peaks (c) and (h) for pristine PMMA and PTFEMA, peak (g) for pristine PTFEMA, and peaks (g) and (h) for T436 are plotted only to demonstrate the trends in the measured values and are not exact values because of possible measurement errors.

From Figure 3d, it was suggested that a higher T_1ρ^C value at peak (a) was obtained in specimens with a larger PDMS M_n when the PDMS volume fraction ranged from 0.29 to 0.36. In other words, it is suggested that the longer -OSi(CH₃)₂- chain resulted in higher molecular mobility and in higher gas molecule permeability, as was also observed in the T₁^H and T_1ρ^H measurements. Moreover, it is interesting that the T_1ρ^C value at peak (a) in the TFEMA/PDMS copolymer is larger than that of the MMA/PDMS copolymer and that the T_1ρ^C values at peak (g) and (h) are larger than that at peak (a). This implies amplified molecular motion in the -OSi(CH₃)₂- chain, -CF₃ group and -C=O group and results in a large contribution from the fluorine-containing group to the oxygen permeability, as has been previously reported.¹¹

Phase size and phase separation estimation

Specimens containing PDMS with an M_n of 1700 exhibited a homogeneously dispersed structure with a diameter of 2.6 nm. When PDMS with an M_n of 4700 was employed, some samples showed a homogeneously dispersed structure with a diameter of approximately 20 nm; therefore, from the viewpoint of a scale on the order of several nm, all specimens had a phase-separated structure.

Conclusion

We conducted ¹³C solid-state NMR measurements to investigate the molecular-level relationship between the physical properties and the composition of copolymers produced from methacrylates, PDMS macromonomers and crosslinkers. Using the data from T₁^H and T_1ρ^H, it was suggested that a larger PDMS M_n value resulted in a higher molecular mobility of -OSi(CH₃)₂-, and the high mobility could be correlated to a higher gas permeability. In addition, the copolymer with an M_n of PDMS of 1700 was found to have a homogeneously dispersed structure with a diameter of approximately 3 nm. PDMS with higher values of M_n, such as 4700 M_n PDMS, resulted in a phase-separated structure on the order of several nm and eventually resulted in a homogeneously dispersed structure with a diameter of tens of nm. This result could be correlated to the appearance of the copolymer.

Differences in the morphology of MMA copolymers and TFEMA copolymers were suggested from the T₁^H, T_1ρ^H and T_1ρ^C measurements. Copolymerization of CF₃-bearing trifluoroethyl methacrylate resulted in a larger T_1ρ^C value at peak (a), and the -CF₃ (g) and -C=O (h) groups showed higher T_1ρ^C values than that of the -OSi(CH₃)₂- chain. The described results implied that the amplified molecular motion of aforementioned groups is related to the mechanism of increased oxygen permeability in copolymers produced from fluorine-bearing monomers and the PDMS macromonomer.