Synthesis of Amphiphilic Statistical Copolymers Bearing Methoxyethyl and Phosphorylcholine Groups and Their Self-Association Behavior in Water

Biocompatible amphiphilic statistical copolymers P(MEA/MPCm) composed of 2-methoxyethyl acrylate (MEA) and 2-methacryloyloxyethyl phosphorylcholine (MPC) were prepared with three different mol% of the hydrophilic unit MPC (m = 6, 12 and 46 mol%). The monomer reactivity ratios of MEA (rMEA) and MPC (rMPC) were 0.53 and 2.21, respectively. The rMEA × rMPC value of 1.17 demonstrated that statistical copolymerization was successful. P(MEA/MPC12) and P(MEA/MPC46) copolymers did not undergo aggregation in water, whereas the P(MEA/MPC6) copolymer formed micelles in water with a hydrodynamic radius (Rh) of 96.9 nm and a critical aggregation concentration, which was determined using pyrene fluorescence, at 0.0082 g/L. The restricted motion of the protons in the hydrophobic MEA units in the micelles’ cores provided additional evidence of self-association in P(MEA/MPC6).


Introduction
The self-association of random copolymers has become increasingly important in drug delivery systems [1][2][3], imaging techniques [4][5][6][7] and many other fields [8]. The self-association behavior of amphiphilic random copolymers is more complicated than that of block polymers because of their ill-defined, randomly arranged monomer sequence. As such, the association between the hydrophobic units of a single polymer chain can form unimer micelles, and is independent of the polymer's concentration or the presence of interpolymer micelles, which are generated via the interaction of multiple polymer chains above the critical aggregation concentration (CAC). Morishima et al. [9] investigated the self-organization of random copolymers of sodium 2-(acrylamido)-2-methylpropanesylfonate (AMPS) and methacrylamides bearing bulky hydrophobic groups, including n-dodecyl (LA), cyclododecyl (CD) and 1-adamantyl (AD) groups. Here, the aggregates formed from these random copolymers in aqueous solutions were due to intra and/or intermolecular self-association. The copolymers of AMPS and CD or AD tended to form unimer micelles up to a concentration limit of ca. 7 wt.%; conversely, the LA-containing copolymers self-associated to form unimers at concentrations below 0.2 wt.% and interpolymer aggregates at higher concentrations. These results showed that the chemical structure of the hydrophobic monomer in random copolymers was a critical factor in the self-association process. Additionally, the self-association of random copolymers depends on other factors, such as the type of solvent, the hydrophobic content in the copolymer [10,11], the distribution of the

Monomer Reactivity Ratio and Polymerization Kinetics
The monomer reactivity ratio was determined using the Fineman-Ross method [34]. Here, MEA, MPC and V-70 were dissolved in methanol with feed ratios of MPC ranging from 10% to 90% (([MEA] + [MPC])/[V-70] = 100/0.4). A small amount of methanol-d4 was added to the solution, which was transferred to NMR tubes and purged with argon gas for 30 min. Polymerization was performed in an oil bath at 40 °C. Quenching procedures were conducted via rapid cooling in an ice bath when the monomer conversion, which was monitored via 1 H NMR, was less than 20%. The reaction mixture was dialyzed (MWCO:500 ~ 1000) using pure water for one night. Next, the solvent was removed via evaporation, and then the residue obtained was subsequently dissolved in 300 µL of methanol-d4 before 1 H NMR analysis was conducted. The MEA and MPC contents in the copolymer were determined by comparing the 1 H NMR integral intensities of the peaks attributed to the pendant methylene protons in MEA (3.62 ppm) and MPC (3.72 ppm) ( Figure S2 and Table S1).
The ratio (mMEA/mMPC = f) of the MEA and MPC contents in the copolymer was determined using the following equation:

Monomer Reactivity Ratio and Polymerization Kinetics
The monomer reactivity ratio was determined using the Fineman-Ross method [34]. Here, MEA, MPC and V-70 were dissolved in methanol with feed ratios of MPC ranging from 10% to 90% (([MEA] + [MPC])/[V-70] = 100/0.4). A small amount of methanol-d 4 was added to the solution, which was transferred to NMR tubes and purged with argon gas for 30 min. Polymerization was performed in an oil bath at 40 • C. Quenching procedures were conducted via rapid cooling in an ice bath when the monomer conversion, which was monitored via 1 H NMR, was less than 20%. The reaction mixture was dialyzed (MWCO:500~1000) using pure water for one night. Next, the solvent was removed via evaporation, and then the residue obtained was subsequently dissolved in 300 µL of methanol-d 4 before 1 H NMR analysis was conducted. The MEA and MPC contents in the copolymer were determined by comparing the 1 H NMR integral intensities of the peaks attributed to the pendant methylene protons in MEA (3.62 ppm) and MPC (3.72 ppm) ( Figure S2 and Table S1).
The ratio (m MEA /m MPC = f ) of the MEA and MPC contents in the copolymer was determined using the following equation:  The relationship between the extent of the monomer conversion and the polymerization time was also studied. Here, MEA (65.5 mg, 0.503 mmol), MPC (148.5 mg, 0.503 mmol) and V-70 (1.28 mg, 0.00412 mmol) were dissolved in 0.5 mL of methanol-d 4 ([MEA]/[MPC]/[V-70] = 50/50/0.4) before the solution was transferred to an NMR tube and purged with argon gas for 30 min. The polymerization reaction was performed at 40 • C, and NMR analysis was used to monitor the process. The resulting NMR spectra were recorded at different polymerization times to determine the extent of conversion of the MEA and MPC monomers by comparing the integral intensity ratios of the vinyl protons observed at 6.37 and 5.62 ppm, respectively, before and after polymerization.

Preparation of the MEA Homopolymer
PMEA was synthesized via the conventional free-radical polymerization reaction using the following procedure. First, MEA (0.260 g, 2.0 mmol) and V-70 (2.45 mg, 0.008 mmol) were dissolved in methanol (2.0 mL), and the mixture was purged with Ar gas in 30 min before polymerization was conducted at 40 • C for 18 h. We noted that MEA conversion was 66.7%. After polymerization, the solution was dialyzed using methanol for one day and pure water for an additional 24 h. Mild precipitation was observed after dialysis. The solution was transferred to a 20-mL glass bottle ( Figure S3a) and was subjected to dynamic light scattering (DLS) as a means of monitoring the formation of PMEA polymer micelles. After a portion of the aqueous solution was evaporated under vacuum, the residue obtained was dissolved in methanol-d 4 or the mobile phase to prepare samples for 1 H NMR analysis or size exclusion chromatography (SEC), respectively.

Preparation of the P(MEA/MPC m ) Copolymer
Statistical copolymers (i.e., P(MEA/MPC m )) were synthesized via conventional free-radical polymerization in methanol using three MPC feed mol% values, namely, 5, 10 and 40 mol% (Figure 1). A typical procedure for the synthesis of P(MEA/MPC m ), with m in the feed of 10 mol%, was as follows: first, MEA (0.240 g, 1.84 mmol), MPC (0.0594 g, 0.201 mmol) and V-70 (25.2 mg, 0.082 mmol) were dissolved in methanol (2.0 mL). The mixture was then purged with Ar gas for 30 min and subsequently stirred at 40 • C for 18 h. After polymerization, the reaction mixture was dialyzed using methanol for one day and then pure water for an additional 24 h. After dialysis, the P(MEA/MPC 6 ) solution obtained was cloudy, indicating the formation of polymer micelles. Clear liquids were obtained in other cases, i.e., for P(MEA/MPC 12 ) and P(MEA/MPC 46 ) ( Figure S3). The polymer solutions after dialysis were used for experiments focused on determining the association behavior of the copolymers in water. 1 H NMR spectroscopy was conducted after purification to estimate the MEA and MPC contents of the copolymers obtained. A portion of the solution was freeze-dried, and the residue was dissolved in methanol or methanol-d 4 to prepare a solution for further experimentation.

Measurements
1 H NMR spectroscopy was performed on a JNM-ECZ 400 MHz spectrometer (JEOL, Tokyo, Japan) using D 2 O or methanol-d 4 as the deuterated solvents. Spin-spin relaxation times (T 2 ) were measured using the Car-Purcell-Meiboom-Gill method. Echoes were observed at the 180 • pulse, and the amplitude of the successive echoes decayed exponentially with a time constant equal to T 2 [35]. Array parameters of the delay list, which included 16 points, were set for conducting the measurements. The data were analyzed via the Weight Linear Spin Lock method using Delta v5.3.1 software (JEOL, Tokyo, Japan). SEC was performed using an instrument equipped with a 7.0-µm bead GF-7M HQ column from Shodex (Tokyo, Japan) and a Shodex RI-101 refractive index (RI) detector operating at 40 • C. The elution phase was methanol containing 0.1 M lithium perchlorate at a flow rate of 0.6 mL/min. Poly(ethylene oxide) was used to prepare the universal standard curve to determine the number-average molecular weight (M n ) and the molecular weight distribution (Ð). Sample solutions were filtered using a 0.2-µm pore size membrane conducting the relevant measurements. The hydrodynamic radius (R h ) and light scattering intensity (LSI) of P(MEA/MPC m ) in water were obtained using a Malvern (Worcestershire, UK) Zetasizer 7.11 equipped with a 4-mW He-Ne laser at 25 • C. The wavelength of the light source was 632.8 nm. The data obtained were analyzed using a Malvern (Worcestershire, UK) Zetasizer 7.11. The polymer concentration (C p ) of P(MEA/MPC m ) in aqueous solution was fixed at 1.0 g/L. The sample solutions were filtered with a 0.45-µm pore size membrane before conducting the analysis. Transmission electron microscopy (TEM) was performed using a JEOL JEM-2100F (Tokyo, Japan) with an acceleration voltage of 160 kV. The samples were prepared by adding the respective aqueous polymer solution dropwise onto a JEOL (Tokyo, Japan) 150-mesh copper TEM grid, and subsequently staining it with 0.1 wt.% phosphotungstic acid aqueous solution. Next, the samples were dried under vacuum conditions at room temperature. Static light scattering (SLS) measurements were conducted using a DLS-7000 Otsuka Electronics Photal TM (Osaka, Japan) at 25 • C, with a He-Ne laser (10.0 mW at 632.8 nm) as the light source. The weight-average molecular weight (M w ), the radius of gyration (R g ) and the second virial coefficient (A 2 ) of the P(MEA/MPC 6 ) polymer micelles in water were estimated from Zimm plots constructed using data derived from aqueous polymer solutions at two different concentrations (i.e., 0.25 and 0.5 g/L). The M w and R g values of all copolymers in methanol in the random-coil state were calculated from the Zimm plot at 10 g/L. The RI increment (dn/dC p ) values were determined using a DRM-3000 differential refractometer (Otsuka Electronics Co., Osaka, Japan) at 25 • C. The dn/dC p values were 0.0619, 0.0768 and 0.104 mL/g for P(MEA/MPC m ) in methanol, with m values of 5, 10 and 40 mol%, respectively, in the feed. These values were used for conducting SLS analysis of the corresponding samples. The CAC of the polymer aqueous solution was determined using pyrene as the fluorescent probe. The fluorescence spectra of the pyrene/polymer aqueous solutions were recorded with an F-2500 fluorescence spectrophotometer from Hitachi (Tokyo, Japan).

Determination of the Monomer Reactivity Ratio
Conventional free-radical polymerizations of equimolar concentrations of MEA and MPC were conducted in the presence of the V-70 initiator at 40 • C. Here, the monomer concentration was estimated based on the observed decrease in the integral intensity of the vinyl protons in the 1 H NMR spectra, at 6.37 and 5.62 ppm for MEA and MPC, respectively. The time conversion plots (Figure S1a) indicated that MPC conversion rapidly reached 90% after 120 min and continued to increase to 99% after 300 min. Conversely, the MEA reaction was notably slower, as MEA conversion after 300 min was 82%. These results were consistent with the reported monomer reactivity findings obtained using the Fineman-Ross method ( Figure 2b). The first-order kinetic plots showed that both plots were linearly related to the polymerization time during the early stages of the polymerization reaction until the 100-min time mark, indicating that the propagating radical concentration was constant ( Figure S1b). The late stages of the polymerization reaction (i.e., after 100 min) produced plots with a downward curvature, suggesting that the propagating radical concentration had decreased. Thus, a straight fitted line was obtained from the Fineman-Ross plots, in which the slope and intercept represented the monomer reactivity ratios of MEA and MPC, respectively. The monomer reactivity ratios of MEA and MPC were 0.53 and 2.21, respectively. Since the reactivity of MPC was almost four times higher than that of MEA, the probability of MPC being incorporated into the copolymer was much higher. Figure 2a also showed that the MPC mol% in the copolymer was always higher than the percentage of MPC in the feed ratio. The reactivity ratios of MEA and MPC were in good agreement with the values observed during the copolymerization of methyl acrylate (MA) and MMA, both of which possessed similar structures. Grassie et al. [36] reported that the reactivity ratios of MA and MMA were 0.35 and 1.8, respectively, during copolymerization at 65 • C. For MA and MMA, these values were 0.47 and 2.3, respectively, during copolymerization at 130 • C, indicating that the reactivity ratio of MMA was almost four to five times higher than MA, and was temperature-independent. From the results, r MEA × r MPC was 1.17, which demonstrated that an almost random monomer sequence was the main feature of the copolymerization process in these compounds. Note, however, that the repeating units in the copolymer were composed of more MPC, since MPC was significantly more active than MEA.
Polymers 2020, 12, x FOR PEER REVIEW 6 of 14 however, that the repeating units in the copolymer were composed of more MPC, since MPC was significantly more active than MEA.

Preparation of PMEA and P(MEA/MPCm)
PMEA was prepared via conventional free-radical polymerization reactions. The conversion of MEA, which was estimated from the corresponding NMR data, was 66.7%. The Mn and Đ values were 1.32 × 10 4 g/mol and 2.54, respectively. The P(MEA/MPCm) copolymers were prepared with three different compositions (i.e., m = 6, 12 and 46 mol%) via conventional free-radical polymerization reactions. The conversions of the MEA and MPC monomers were in the ranges of 72-89% and 98-100%, respectively. The MPC content in the copolymer was estimated from 1 H NMR spectra obtained in methanol-d4, using the integral intensity ratio of the peaks attributed to the pendant methylene protons in the MEA and MPC units; these peaks were observed at 3.62 and 3.72 ppm for MEA and MPC, respectively ( Figure 3 and Table S2). SEC measurements were conducted for all polymers, and unimodal curves were observed for all samples with Đ values between 2.0 and 2.6 ( Figure S4). We theorized that this was due to the occurrence of uncontrolled polymerization. The SEC charts indicated that P(MEA/MPC46) had the longest retention time ( Figure S4), which was possibly due to unexpected interactions between the SEC column and the copolymer. The Mw values of a single polymer chain were determined using SLS measurements in methanol. Unimodal distribution was observed in the DLS results for the methanolic solutions of the copolymer, with small Rh values between 9 and 10 nm ( Figure S5). The Rg and Mw values of the copolymers were estimated via conducting SLS measurements in methanol ( Figure S6). Here, we noted that the Rg/Rh values for all the copolymers were less than or equal to 1.6, indicating that the polymers were dissolved in methanol as with a large polydispersity index (PDI) [37]. In particular, the Rg/Rh value for P(MEA/MPC46) was associated with a large degree of error due to the bimodal Rh distribution ( Figure  S5). These features made it possible to determine the molecular weight of the individual polymer chains using SLS measurements in methanol. The characteristics of all samples are summarized in Table 1.

Preparation of PMEA and P(MEA/MPC m )
PMEA was prepared via conventional free-radical polymerization reactions. The conversion of MEA, which was estimated from the corresponding NMR data, was 66.7%. The M n and Ð values were 1.32 × 10 4 g/mol and 2.54, respectively. The P(MEA/MPC m ) copolymers were prepared with three different compositions (i.e., m = 6, 12 and 46 mol%) via conventional free-radical polymerization reactions. The conversions of the MEA and MPC monomers were in the ranges of 72-89% and 98-100%, respectively. The MPC content in the copolymer was estimated from 1 H NMR spectra obtained in methanol-d 4 , using the integral intensity ratio of the peaks attributed to the pendant methylene protons in the MEA and MPC units; these peaks were observed at 3.62 and 3.72 ppm for MEA and MPC, respectively ( Figure 3 and Table S2). SEC measurements were conducted for all polymers, and unimodal curves were observed for all samples with Ð values between 2.0 and 2.6 ( Figure S4). We theorized that this was due to the occurrence of uncontrolled polymerization. The SEC charts indicated that P(MEA/MPC 46 ) had the longest retention time ( Figure S4), which was possibly due to unexpected interactions between the SEC column and the copolymer. The M w values of a single polymer chain were determined using SLS measurements in methanol. Unimodal distribution was observed in the DLS results for the methanolic solutions of the copolymer, with small R h values between 9 and 10 nm ( Figure S5). The R g and M w values of the copolymers were estimated via conducting SLS measurements in methanol ( Figure S6). Here, we noted that the R g /R h values for all the copolymers were less than or equal to 1.6, indicating that the polymers were dissolved in methanol as with a large polydispersity index (PDI) [37]. In particular, the R g /R h value for P(MEA/MPC 46 ) was associated with a large degree of error due to the bimodal R h distribution ( Figure S5). These features made it possible to determine the molecular weight of the individual polymer chains using SLS measurements in methanol. The characteristics of all samples are summarized in Table 1.

Self-association Behavior of P(MEA/MPCm) in Water
NMR measurements were recorded for all copolymers in D2O ( Figure 4). Here, the pendant methylene proton signals of the MEA units were observed at 3.68 ppm in D2O, which was shifted downfield from their position at 3.62 ppm in methanol-d4. As a result, the signals of the MEA and MPC units overlapped at 3.72 ppm.  a m indicates the MPC content (mol%) in the copolymer, as estimated using NMR data obtained in methanol-d 4 . b The number-average molecular weight (M n ) and molecular weight distribution (Ð) were estimated via SEC analysis. c The apparent weight-average molecular weight (M w ) and the radius of gyration (R g ) were estimated from SLS measurements in methanol. d The hydrodynamic radius (R h ) was obtained from DLS measurements in the methanolic solutions of the copolymers.

Self-Association Behavior of P(MEA/MPC m ) in Water
NMR measurements were recorded for all copolymers in D 2 O (Figure 4). Here, the pendant methylene proton signals of the MEA units were observed at 3.68 ppm in D 2 O, which was shifted downfield from their position at 3.62 ppm in methanol-d 4 . As a result, the signals of the MEA and MPC units overlapped at 3.72 ppm.
The dynamic motion of individual molecular segments could be estimated by measuring the 1 H NMR spin-spin relaxation time (T 2 ) [38,39]. We noted that when the polymer micelles were formed, the motion of the protons relative to the hydrophobic MEA unit was restricted to the hydrophobic domain of the polymer micelle. As a result, restrictions in the motion of the MEA units decreased the T 2 values. Here, the T 2 values were estimated for the methyl protons at 3.39 and 3.23 ppm in the MEA and MPC units of the copolymers, respectively, in D 2 O ( Figure 5). We noted that the T 2 values of the pendant methyl protons in the MEA units decreased with decreasing m, as exemplified by the smallest value of 332 ms, which was obtained when m was 6 mol%. For P(MEA/MPC 12 ) and Polymers 2020, 12, 1808 8 of 14 P(MEA/MPC 46 ), the T 2 values (namely, 528 and 561 ms, respectively) were greater than those observed for P(MEA/MPC 6 ). The T 2 values did not significantly change when m increased from 12 to 46 mol%. These observations indicated that the motion of the hydrophobic MEA units in P(MEA/MPC 6 ) was restricted due to the formation of the hydrophobic domain. In contrast, the motion of the MEA units in P(MEA/MPC 12 ) and P(MEA/MPC 46 ) was free, since these copolymers were dissolved as unimers in water. The T 2 values of the pendant methyl protons in the MPC units were almost the same, regardless of the m value; for P(MEA/MPC 6 ), P(MEA/MPC 12 ), and P(MEA/MPC 46 ), these values were 393, 414, and 412 ms, respectively. As noted earlier, the hydrophilic MPC units were arranged on the surface of the micelles when P(MEA/MPC 6 ) formed polymer micelles, whereas P(MEA/MPC 12 ) and P(MEA/MPC 46 ) were dissolved as unimers. Since the MPC units were always exposed to water, their motion remained almost the same for all samples. The  The dynamic motion of individual molecular segments could be estimated by measuring the 1 H NMR spin-spin relaxation time (T2) [38,39]. We noted that when the polymer micelles were formed, the motion of the protons relative to the hydrophobic MEA unit was restricted to the hydrophobic domain of the polymer micelle. As a result, restrictions in the motion of the MEA units decreased the T2 values. Here, the T2 values were estimated for the methyl protons at 3.39 and 3.23 ppm in the MEA and MPC units of the copolymers, respectively, in D2O ( Figure 5). We noted that the T2 values of the pendant methyl protons in the MEA units decreased with decreasing m, as exemplified by the smallest value of 332 ms, which was obtained when m was 6 mol%. For P(MEA/MPC12) and P(MEA/MPC46), the T2 values (namely, 528 and 561 ms, respectively) were greater than those observed for P(MEA/MPC6). The T2 values did not significantly change when m increased from 12 to 46 mol%. These observations indicated that the motion of the hydrophobic MEA units in P(MEA/MPC6) was restricted due to the formation of the hydrophobic domain. In contrast, the motion of the MEA units in P(MEA/MPC12) and P(MEA/MPC46) was free, since these copolymers were dissolved as unimers in water. The T2 values of the pendant methyl protons in the MPC units were almost the same, regardless of the m value; for P(MEA/MPC6), P(MEA/MPC12), and P(MEA/MPC46), these values were 393, 414, and 412 ms, respectively. As noted earlier, the The self-association behavior of P(MEA/MPC m ) in water was studied by conducting DLS measurements on the copolymer aqueous solutions at C p = 1.0 g/L ( Figure 6). Here, the R h distribution values for PMEA were established as a reference sample. Small amounts of precipitate were obtained after the purification of PMEA via dialysis, which contained high molecular weight chains (i.e., large aggregates) that were removed via filtration. The PMEA chains with lower molecular weight values did not precipitate, as they were dispersed in water during dialysis. For the aqueous solution of P(MEA/MPC 46 ), unimodal distribution was observed with a small R h value of 9.0 nm, suggesting that P(MEA/MPC 46 ) was dissolved in water as the unimer. Interestingly, bimodal distribution was observed in P(MEA/MPC 12 ), which contained one peak with the same R h value (i.e., 9.0 nm) as P(MEA/MPC 46 ) in water and another peak with a higher R h (i.e., 222 nm) value. In this case, the formation of interpolymer associates composed of some of the hydrophobic MEA units in the copolymer was identified as the reason for these observations. The self-association behavior of P(MEA/MPCm) in water was studied by conducting DLS measurements on the copolymer aqueous solutions at Cp = 1.0 g/L ( Figure 6). Here, the Rh distribution values for PMEA were established as a reference sample. Small amounts of precipitate were obtained after the purification of PMEA via dialysis, which contained high molecular weight chains (i.e., large aggregates) that were removed via filtration. The PMEA chains with lower molecular weight values did not precipitate, as they were dispersed in water during dialysis. For the aqueous solution of P(MEA/MPC46), unimodal distribution was observed with a small Rh value of 9.0 nm, suggesting that P(MEA/MPC46) was dissolved in water as the unimer. Interestingly, bimodal distribution was observed in P(MEA/MPC12), which contained one peak with the same Rh value (i.e., 9.0 nm) as P(MEA/MPC46) in water and another peak with a higher Rh (i.e., 222 nm) value. In this case, the formation of interpolymer associates composed of some of the hydrophobic MEA units in the copolymer was identified as the reason for these observations. Generally, weak hydrophobic interactions were noted when the hydrophobic content in the polymer was insufficient, resulting in the formation of only a few micelles. For PMEA and P(MEA/MPC6), unimodal distribution was noted with Rh values of 290 and 96.9 nm, respectively, indicating that P(MEA/MPC6), with its 94 mol% of the hydrophobic MEA unit, easily formed interpolymer aggregates in water. The radius of the polymer aggregates increased when PMEA was employed, resulting in self-association to generate larger aggregates in water. All of the above mentioned findings were supported by the results obtained via LSI analysis of these samples. Here, the precipitation observed after dialysis using pure water was evidence that PMEA could form very large aggregates. Only the supernatant was used for DLS analysis, and the estimated concentration of the PMEA aqueous solution was believed to be lower than the true value. This was proposed as the possible reason for the lower LSI values observed for PMEA relative to P(MEA/MPC6) in water at the same concentration. By extension, the LSI values of both PMEA and P(MEA/MPC6) were greater than those obtained for P(MEA/MPC12) and P(MEA/MPC46) due to the formation of polymer micelles. Since P(MEA/MPC46) was dissolved as random coils in water, the smallest recorded LSI  The self-association behavior of P(MEA/MPCm) in water was studied by conducting DLS measurements on the copolymer aqueous solutions at Cp = 1.0 g/L ( Figure 6). Here, the Rh distribution values for PMEA were established as a reference sample. Small amounts of precipitate were obtained after the purification of PMEA via dialysis, which contained high molecular weight chains (i.e., large aggregates) that were removed via filtration. The PMEA chains with lower molecular weight values did not precipitate, as they were dispersed in water during dialysis. For the aqueous solution of P(MEA/MPC46), unimodal distribution was observed with a small Rh value of 9.0 nm, suggesting that P(MEA/MPC46) was dissolved in water as the unimer. Interestingly, bimodal distribution was observed in P(MEA/MPC12), which contained one peak with the same Rh value (i.e., 9.0 nm) as P(MEA/MPC46) in water and another peak with a higher Rh (i.e., 222 nm) value. In this case, the formation of interpolymer associates composed of some of the hydrophobic MEA units in the copolymer was identified as the reason for these observations. Generally, weak hydrophobic interactions were noted when the hydrophobic content in the polymer was insufficient, resulting in the formation of only a few micelles. For PMEA and P(MEA/MPC6), unimodal distribution was noted with Rh values of 290 and 96.9 nm, respectively, indicating that P(MEA/MPC6), with its 94 mol% of the hydrophobic MEA unit, easily formed interpolymer aggregates in water. The radius of the polymer aggregates increased when PMEA was employed, resulting in self-association to generate larger aggregates in water. All of the above mentioned findings were supported by the results obtained via LSI analysis of these samples. Here, the precipitation observed after dialysis using pure water was evidence that PMEA could form very large aggregates. Only the supernatant was used for DLS analysis, and the estimated concentration of the PMEA aqueous solution was believed to be lower than the true value. This was proposed as the possible reason for the lower LSI values observed for PMEA relative to P(MEA/MPC6) in water at the same concentration. By extension, the LSI values of both PMEA and P(MEA/MPC6) were greater than those obtained for P(MEA/MPC12) and P(MEA/MPC46) due to the formation of polymer micelles. Since P(MEA/MPC46) was dissolved as random coils in water, the smallest recorded LSI Generally, weak hydrophobic interactions were noted when the hydrophobic content in the polymer was insufficient, resulting in the formation of only a few micelles. For PMEA and P(MEA/MPC 6 ), unimodal distribution was noted with R h values of 290 and 96.9 nm, respectively, indicating that P(MEA/MPC 6 ), with its 94 mol% of the hydrophobic MEA unit, easily formed interpolymer aggregates in water. The radius of the polymer aggregates increased when PMEA was employed, resulting in self-association to generate larger aggregates in water. All of the above mentioned findings were supported by the results obtained via LSI analysis of these samples. Here, the precipitation observed after dialysis using pure water was evidence that PMEA could form very large aggregates. Only the supernatant was used for DLS analysis, and the estimated concentration of the PMEA aqueous solution was believed to be lower than the true value. This was proposed as the possible reason for the lower LSI values observed for PMEA relative to P(MEA/MPC 6 ) in water at the same concentration. By extension, the LSI values of both PMEA and P(MEA/MPC 6 ) were greater than those obtained for P(MEA/MPC 12 ) and P(MEA/MPC 46 ) due to the formation of polymer micelles. Since P(MEA/MPC 46 ) was dissolved as random coils in water, the smallest recorded LSI value in our study was 0.087 Mcps. As noted, the aqueous solution of PMEA became cloudier after a few days, whereas the P(MEA/MPC 6 ) aggregates exhibited increased stability over the same observation time. The stability of the P(MEA/MPC 6 ) aggregates during dilution was confirmed via DLS measurements. Here, the R h values of P(MEA/MPC 6 ) were nearly constant and independent of C p in the 0.05-5 g/L region ( Figure S7). Note, however, that lower concentrations were difficult to measure due to the low LSI values associated.
On another note, spherical objects were observed for PMEA and P(MEA/MPC 6 ) copolymers using TEM (Figure 7), with average radii of 240 and 103 nm for PMEA and P(MEA/MPC 6 ), respectively. Additional TEM observations for P(MEA/MPC 6 ) are shown in Figure S8. value in our study was 0.087 Mcps. As noted, the aqueous solution of PMEA became cloudier after a few days, whereas the P(MEA/MPC6) aggregates exhibited increased stability over the same observation time. The stability of the P(MEA/MPC6) aggregates during dilution was confirmed via DLS measurements. Here, the Rh values of P(MEA/MPC6) were nearly constant and independent of Cp in the 0.05-5 g/L region ( Figure S7). Note, however, that lower concentrations were difficult to measure due to the low LSI values associated.
On another note, spherical objects were observed for PMEA and P(MEA/MPC6) copolymers using TEM (Figure 7), with average radii of 240 and 103 nm for PMEA and P(MEA/MPC6), respectively. Additional TEM observations for P(MEA/MPC6) are shown in Figure S8. The characterization of the P(MEA/MPC6) micelles in water was conducted via SLS measurements ( Figure 8). Here, the dn/dCp value for P(MEA/MPC6) in water was 0.0651 mL/g, and was subsequently used during SLS analysis to estimate the apparent Mw of the P(MEA/MPC6) micelles. The aggregation number (Nagg), which is the number of individual polymer chains present in one micelle, was calculated from the Mw values of the polymer micelle in water and the individual polymer chain in methanol obtained from SLS measurements. The Nagg of P(MEA/MPC6) micelle was 143 in water, indicating the formation of interpolymer micelles. It was speculated that the hydrophobic interactions between the MEA units in the individual polymer chains were too weak to facilitate the self-folding of a single chain. For P(MEA/MPC6) with 94 mol% of the relatively hydrophobic MEA units in the copolymer's composition, hydrophobic interactions between multiple chains resulted in interpolymer aggregates. This phenomenon was also seen in the aqueous solution of random copolymers of sodium 2-(acrylamido)-2-methylpropanesulfonate and n-dodecyl methacrylamide, and in the copolymers of poly(ethylene glycol) methacrylate and n-dodecyl methacrylate. Here, interpolymer aggregates were formed when the hydrophobic content of the copolymer exceeded 50 mol% [10,11]. Theoretically, the shape and size distribution of the aggregates can be estimated using Rg/Rh. For instance, a value of 0.778 represents a rigid sphere with a narrow distribution, whereas a value of ~1.0 is associated with spherical aggregates, and values higher than 2 are often seen in rodlike structures [37,40]. In our study, the Rg/Rh value for P(MEA/MPC6) in water was 0.95, indicating that The characterization of the P(MEA/MPC 6 ) micelles in water was conducted via SLS measurements ( Figure 8). Here, the d n /d Cp value for P(MEA/MPC 6 ) in water was 0.0651 mL/g, and was subsequently used during SLS analysis to estimate the apparent M w of the P(MEA/MPC 6 ) micelles. The aggregation number (N agg ), which is the number of individual polymer chains present in one micelle, was calculated from the M w values of the polymer micelle in water and the individual polymer chain in methanol obtained from SLS measurements. The N agg of P(MEA/MPC 6 ) micelle was 143 in water, indicating the formation of interpolymer micelles. It was speculated that the hydrophobic interactions between the MEA units in the individual polymer chains were too weak to facilitate the self-folding of a single chain. For P(MEA/MPC 6 ) with 94 mol% of the relatively hydrophobic MEA units in the copolymer's composition, hydrophobic interactions between multiple chains resulted in interpolymer aggregates. This phenomenon was also seen in the aqueous solution of random copolymers of sodium 2-(acrylamido)-2-methylpropanesulfonate and n-dodecyl methacrylamide, and in the copolymers of poly(ethylene glycol) methacrylate and n-dodecyl methacrylate. Here, interpolymer aggregates were formed when the hydrophobic content of the copolymer exceeded 50 mol% [10,11]. few days, whereas the P(MEA/MPC6) aggregates exhibited increased stability over the same observation time. The stability of the P(MEA/MPC6) aggregates during dilution was confirmed via DLS measurements. Here, the Rh values of P(MEA/MPC6) were nearly constant and independent of Cp in the 0.05-5 g/L region ( Figure S7). Note, however, that lower concentrations were difficult to measure due to the low LSI values associated.
On another note, spherical objects were observed for PMEA and P(MEA/MPC6) copolymers using TEM (Figure 7), with average radii of 240 and 103 nm for PMEA and P(MEA/MPC6), respectively. Additional TEM observations for P(MEA/MPC6) are shown in Figure S8. The characterization of the P(MEA/MPC6) micelles in water was conducted via SLS measurements ( Figure 8). Here, the dn/dCp value for P(MEA/MPC6) in water was 0.0651 mL/g, and was subsequently used during SLS analysis to estimate the apparent Mw of the P(MEA/MPC6) micelles. The aggregation number (Nagg), which is the number of individual polymer chains present in one micelle, was calculated from the Mw values of the polymer micelle in water and the individual polymer chain in methanol obtained from SLS measurements. The Nagg of P(MEA/MPC6) micelle was 143 in water, indicating the formation of interpolymer micelles. It was speculated that the hydrophobic interactions between the MEA units in the individual polymer chains were too weak to facilitate the self-folding of a single chain. For P(MEA/MPC6) with 94 mol% of the relatively hydrophobic MEA units in the copolymer's composition, hydrophobic interactions between multiple chains resulted in interpolymer aggregates. This phenomenon was also seen in the aqueous solution of random copolymers of sodium 2-(acrylamido)-2-methylpropanesulfonate and n-dodecyl methacrylamide, and in the copolymers of poly(ethylene glycol) methacrylate and n-dodecyl methacrylate. Here, interpolymer aggregates were formed when the hydrophobic content of the copolymer exceeded 50 mol% [10,11]. Theoretically, the shape and size distribution of the aggregates can be estimated using Rg/Rh. For instance, a value of 0.778 represents a rigid sphere with a narrow distribution, whereas a value of ~1.0 is associated with spherical aggregates, and values higher than 2 are often seen in rodlike structures [37,40]. In our study, the Rg/Rh value for P(MEA/MPC6) in water was 0.95, indicating that Theoretically, the shape and size distribution of the aggregates can be estimated using R g /R h . For instance, a value of 0.778 represents a rigid sphere with a narrow distribution, whereas a value of 1.0 is associated with spherical aggregates, and values higher than 2 are often seen in rodlike structures [37,40]. In our study, the R g /R h value for P(MEA/MPC 6 ) in water was 0.95, indicating that the polymer micelles were spherical ( Table 2). The A 2 value provides information about the affinity of solute molecules with the solvent of interest [41]. For our case, the A 2 value obtained from the SLS analysis of P(MEA/MPC 6 ) in water was 2.5 × 10 −5 cm 3 g −2 mol. This positive value was proof that the hydrophilic MPC unit had completely covered P(MEA/MPC 6 ), thereby facilitating its solubility in water. The A 2 value for the MPC homopolymer was reported as 2.5 × 10 −4 cm 3 g −2 mol in aqueous solution [42]. The A 2 value for P(MEA/MPC 6 ) was 10 times lower than that of the MPC homopolymer, which was attributed to the composition of the copolymer; here, the copolymer consisted of 94 mol% of the hydrophobic MEA unit. The CAC value of P(MEA/MPC 6 ) was determined via fluorescence spectroscopy using pyrene as the probe molecule. Pyrene is hydrophobic and its fluorescence spectrum is heavily dependent on the polarity of the surrounding environment. The intensity ratio (I 3 /I 1 ) between the third and the first vibronic peaks in the pyrene emission spectra can be used as a measure of the environmental polarity [43]. Therefore, the fluorescence spectra of pyrene in the presence of P(MEA/MPC 6 ) at varying C p were recorded at the excitation wavelength of 334 nm ( Figure S9). The changes in the I 3 /I 1 ratio in the pyrene emission spectra were plotted as a function of C p to determine CAC ( Figure 9). We noted that the copolymers did not form micelles below the CAC. The solubility of pyrene in water ensured that the resulting fluorescence spectrum remained almost unchanged (I 3 /I 1 ≈ 0.572). This calculated value was nearly the same as that obtained for pyrene in the absence of the copolymer (I 3 /I 1 ≈ 0.562), demonstrating that there was no discernible difference in the polarity of the environment surrounding pyrene as the P(MEA/MPC 6 ) copolymer was locked in random coils at low concentrations in water. However, the formation of copolymer micelles above the CAC meant that the pyrene molecules were entrapped in the hydrophobic domains. As such, the pyrene molecules were surrounded by a non-polar environment. The number of pyrene molecules in the hydrophobic domain increased when C p rose. Consequently, the I 3 /I 1 value gradually increased to a maximum of 0.602 at C p = 0.08 g/L. The CAC value of P(MEA/MPC 6 ) in water was determined at the inflection point as 0.0082 g/L. Unlike the formation of unimer micelles independent of C p , P(MEA/MPC 6 ) was incapable of aggregation in dilute aqueous solutions. Interpolymer association occurred when the P(MEA/MPC 6 ) concentration increased to a certain level. Therefore, the CAC value was determined as the C p at which the interpolymer micelles started to form. analysis of P(MEA/MPC6) in water was 2.5 × 10 cm g mol. This positive value was proof that the hydrophilic MPC unit had completely covered P(MEA/MPC6), thereby facilitating its solubility in water. The A2 value for the MPC homopolymer was reported as 2.5 × 10 −4 cm 3 g −2 mol in aqueous solution [42]. The A2 value for P(MEA/MPC6) was 10 times lower than that of the MPC homopolymer, which was attributed to the composition of the copolymer; here, the copolymer consisted of 94 mol% of the hydrophobic MEA unit. The CAC value of P(MEA/MPC6) was determined via fluorescence spectroscopy using pyrene as the probe molecule. Pyrene is hydrophobic and its fluorescence spectrum is heavily dependent on the polarity of the surrounding environment. The intensity ratio (I3/I1) between the third and the first vibronic peaks in the pyrene emission spectra can be used as a measure of the environmental polarity [43]. Therefore, the fluorescence spectra of pyrene in the presence of P(MEA/MPC6) at varying Cp were recorded at the excitation wavelength of 334 nm ( Figure S9). The changes in the I3/I1 ratio in the pyrene emission spectra were plotted as a function of Cp to determine CAC ( Figure 9). We noted that the copolymers did not form micelles below the CAC. The solubility of pyrene in water ensured that the resulting fluorescence spectrum remained almost unchanged (I3/I1 ≈ 0.572). This calculated value was nearly the same as that obtained for pyrene in the absence of the copolymer (I3/I1 ≈ 0.562), demonstrating that there was no discernible difference in the polarity of the environment surrounding pyrene as the P(MEA/MPC6) copolymer was locked in random coils at low concentrations in water. However, the formation of copolymer micelles above the CAC meant that the pyrene molecules were entrapped in the hydrophobic domains. As such, the pyrene molecules were surrounded by a non-polar environment. The number of pyrene molecules in the hydrophobic domain increased when Cp rose. Consequently, the I3/I1 value gradually increased to a maximum of 0.602 at Cp = 0.08 g/L. The CAC value of P(MEA/MPC6) in water was determined at the inflection point as 0.0082 g/L. Unlike the formation of unimer micelles independent of Cp, P(MEA/MPC6) was incapable of aggregation in dilute aqueous solutions. Interpolymer association occurred when the P(MEA/MPC6) concentration increased to a certain level. Therefore, the CAC value was determined as the Cp at which the interpolymer micelles started to form.

Conclusions
Biocompatible, amphiphilic, statistical copolymers, P(MEA/MPC m ), which were composed of hydrophobic MEA and hydrophilic MPC units, were prepared via conventional free-radical polymerization reactions. P(MEA/MPC 12 ) and P(MEA/MPC 46 ) possessed a high MPC content, were dissolved as unimers, and did not aggregate in water. On the other hand, P(MEA/MPC 6 ) formed uniform micelles composed of a PMEA core and a PMPC shell, with R h and N agg values of 96.9 nm and 133, respectively, due to the hydrophobic interactions of the MEA units. The mobility of the MEA units in P(MEA/MPC 6 ) was restricted, as confirmed via the observed 1 H NMR spin-spin relaxation times. A slight decrease in the motion of the MPC units was also observed and was attributed to the entrapment of some MPC units in the hydrophobic domain; the associated interpolymer aggregates were covered by phosphorylcholine groups. From these results, we concluded that P(MEA/MPC m ) copolymers with m ≤ 6 mol% formed micelles, and, as such, were well suited for biomedical applications.  Figure S7. Hydrodynamic radius (R h ) of P(MEA/MPC 6 ) as a function of polymer concentration (C p ) in water; Figure S8. Transmission electron microscopy (TEM) images for P(MEA/MPC 6 ) at C p = 1.0 g/L in water with different magnification; Figure S9. Fluorescence spectra of pyrene excited at 334 nm in water in the presence of P(MEA/MPC 6 ) at C p = 0.08 (solid line) and 0.0012 g/L (dashed line); Table S1. The Fineman-Ross parameters of the copolymers, as determined using 1 H NMR measurements in methanol-d 4 at room temperature; and Table S2. The composition of the copolymer, as estimated using 1 H NMR measurements in methanol-d 4 at room temperature.