Combining Crystallization-Driven Self-Assembly with Reverse Sequence Polymerization-Induced Self-Assembly Enables the Efficient Synthesis of Hydrolytically Degradable Anisotropic Block Copolymer Nano-objects Directly in Concentrated Aqueous Media

Herein we combine the well-known processing advantages conferred by polymerization-induced self-assembly (PISA) with crystallization-driven self-assembly (CDSA) to achieve the efficient synthesis of hydrolytically degradable, highly anisotropic block copolymer nano-objects directly in aqueous solution at 30% w/w solids. This new strategy involves a so-called reverse sequence PISA protocol that employs poly(l-lactide) (PLLA) as the crystallizable core-forming block and poly(N,N′-dimethylacrylamide) (PDMAC) as the water-soluble non-ionic coronal block. Such syntheses result in PDMAC-rich anisotropic nanoparticles. Depending on the target diblock copolymer composition, either rod-like nanoparticles or diamond-like platelets can be obtained. Furthermore, N-Acryloylmorpholine is briefly evaluated as an alternative hydrophilic vinyl monomer to DMAC. Given that the PLLA block can undergo either hydrolytic or enzymatic degradation, such nanoparticles are expected to offer potential applications in various fields, including next-generation sustainable Pickering emulsifiers.


■ INTRODUCTION
−11 CDSA typically utilizes an insoluble crystalline block and a soluble steric stabilizer block.−21 Depending on the target diblock copolymer composition, this approach typically yields spheres, worms, or vesicles. 22Recently, we reported a counterintuitive reverse sequence aqueous PISA formulation in which a hydrophobic precursor is solubilized in concentrated aqueous media using a water-miscible vinyl monomer as a cosolvent. 23Polymerization of this monomer gradually worsens the solvency for the growing amphiphilic diblock copolymer chains, which subsequently undergo in situ selfassembly to form spheres. Herein, we combine the processing advantages offered by reverse sequence PISA with CDSA to prepare highly anisotropic hydrolytically degradable block copolymer nano-objects directly in aqueous media at 30% w/w solids. 24This new strategy involves a hydroxy-functional trithiocarbonate reagent, 7,10,25,26 employs poly(L-lactide) (PLLA) as the crystallizable core-forming block, and uses poly(N,N′-dimethylacrylamide) (PDMAC) as the watersoluble non-ionic coronal block (see Scheme 1).

■ RESULTS AND DISCUSSION
Recently, we reported a reverse sequence PISA formulation based on a hydrophobic poly(ε-caprolactone) (PCL) precursor, which exhibits a melting transition, T m , at approximately 50 °C.In this prior study, the in situ DMAC polymerization was performed at 80 °C, which results in the formation of spherical PCL−PDMAC nanoparticles with amorphous cores. 23In contrast, PLLA has a T m of 114−153 °C (see Figure S1).Hence reverse sequence PISA syntheses performed at 70 °C should lead to the formation of anisotropic PLLA-PDMAC nanoparticles with semicrystalline cores via CDSA (see Scheme 1).
In the present study, the anionic ring-opening polymerization of L-lactide was initiated using a hydroxy-functional reversible addition−fragmentation chain transfer (RAFT) agent in the presence of a 4-(dimethylamino)pyridine (DMAP) catalyst, as previously reported. 25Anhydrous conditions ensured controlled polymerization to produce a hydrophobic semicrystalline poly(L-lactide) PLLA precursor with a mean degree of polymerization (DP) of 14 as determined by end-group analysis using 1 H NMR spectroscopy.More specifically, the integrated oxymethine PLLA signal at 5.21 ppm was compared to the integrated aromatic proton signals assigned to the benzyl end-group at 7.29−7.41ppm (see Figures S2 and S3).Furthermore, UV GPC analysis (λ = 305 nm) confirmed the absence of any unreacted RAFT agent after purification of PLLA 14 -TTC (see Figure S4).
This PLLA 14 -TTC precursor was then dissolved in N,N′dimethylacrylamide (DMAC) and RAFT polymerization of this vinyl monomer was conducted in the bulk 27,28 at 70 °C.Once a significant increase in solution viscosity was observed (corresponding to a DMAC conversion of 13−65%), degassed deionized water (preheated to 70 °C) was added to the reaction mixture to target a final solids content of 30% w/w (see Scheme 1).The DMAC polymerization was allowed to proceed for 16 h at 70 °C before quenching by cooling the reaction mixture to 20 °C with concomitant exposure to air.For the synthesis of a PLLA 14 -PDMAC 120 diblock copolymer, the reaction mixture was periodically sampled, and aliquots were analyzed by GPC and 1 H NMR spectroscopy to study the polymerization kinetics.After 100 min at 70 °C, 99% DMAC conversion was achieved (Figure 1a).
Remarkably, there was no discernible reduction in the rate of polymerization after dilution of the initial bulk reaction mixture to 30% w/w solids.This is presumably because acrylamides polymerize much faster in aqueous media than in Scheme 1. Synthesis of a 30% w/w Aqueous Dispersion of PLLA 14 -PDMAC x Diblock Copolymer Nanoparticles by the Judicious Combination of Reverse Sequence PISA with CDSA a a A PLLA 14 -TTC precursor is first prepared via anionic ring-opening polymerization of L-lactide at 35 °C using a hydroxy-functional RAFT agent as an initiator.The RAFT polymerization of DMAC is then conducted in the bulk at 70 °C.At a suitable intermediate DMAC conversion, the reaction mixture is diluted with deionized water to induce self-assembly of the growing amphiphilic PLLA 14 -PDMAC x diblock copolymer chains.the bulk. 29,30GPC analysis indicated a linear increase in molecular weight (M n ) with conversion, suggesting a wellcontrolled RAFT polymerization.Dispersities always remained less than 1.40 (M w /M n = 1.15 to 1.36) but were nevertheless relatively high for a RAFT polymerization (see Figure 1b).This is attributed to the chemical structure of the RAFT agent, which is not optimized for the well-controlled polymerization of DMAC.
Indeed, a control experiment conducted using the same hydroxyl-functional RAFT agent to target a PDMAC 100 homopolymer via RAFT polymerization (initially in the bulk followed by dilution with water to 30% w/w solids at an intermediate conversion of 45%) afforded more than 99% conversion after 16 h at 70 °C.GPC analysis indicated an M n of 10.3 kg mol −1 and an M w /M n of 1.26, which is somewhat higher than those reported in the literature for such syntheses. 31,32The same synthetic protocol was then used to target a series of PLLA 14 -PDMAC 70−300 diblock copolymers.GPC analysis indicated a linear increase in molecular weight when targeting higher PDMAC DPs with final dispersities as low as 1.28 (Table 1 and Figure 2).UV GPC (λ = 305 nm) studies indicated reasonably high chain extension efficiencies when targeting PDMAC DPs up to 210, suggesting minimal contamination by the PLLA 14 -TTC precursor (<10% residual precursor, see Figure S6).
However, when targeting a PDMAC DP of either 230 or 300, a low molecular weight species corresponding to 15−17% of the total signal was observed.Interestingly, such contamination was not discernible when using a refractive index detector, see Figure 2a.Moreover, 1 H NMR spectroscopy analysis indicated that more than 99% DMAC conversion was achieved for all syntheses, see Figure S2.
When targeting lower PDMAC DPs of either 50 or 60, problems were encountered when attempting similar syntheses at 70 °C.No precipitation was observed on dilution with water (after 45 or 37 min, respectively).However, DMF GPC analysis using a refractive index detector indicated a bimodal GPC trace in each case, suggesting poor chain extension efficiency (see Figure S7).Furthermore, no high molecular weight species were detected by UV GPC analysis.This indicates that such DMAC polymerizations are poorly controlled because these longer chains do not possess RAFT end-groups (see Figure S7).Moreover, targeting PDMAC DPs below 50 resulted in immediate macroscopic precipitation after dilution with water.Presumably, this is simply because the PDMAC chains that are present when water is added to the reaction mixture are too short to confer effective steric stabilization on the nascent nanoparticles.In contrast, if such syntheses were conducted at 90 °C, then PDMAC DPs as low as 40 could be targeted (see Figure S8).
TEM analysis of the series of PLLA 14 -PDMAC 70−300 nanoparticles confirmed that the final copolymer morphology depended on the target PDMAC DP.Targeting the highest PDMAC DP of 300 led to the formation of diamond-like platelets (see Figure 3a).In contrast, short rod-like nanoparticles were obtained when targeting the lowest PDMAC DP of 70 (see Figure 3i).Hence our new approach to CDSA enables the efficient formation of highly concentrated aqueous dispersions of anisotropic nanoparticles.−35 However, certain applications such as Pickering emulsifiers and foam stabilizers do not require particularly uniform nanoparticles.In such cases, the ability to prepare anisotropic  nanoparticles at high solids concentrations directly in water is likely to be a decisive advantage.
Various binary mixtures of these two morphologies were observed for intermediate PDMAC DPs (Figure 3b−g).These observations are in good agreement with literature reports for the self-assembly of PLLA-PDMAC diblock copolymers via CDSA in dilute solution using organic solvents such as methanol or ethanol. 11,33It is perhaps worth emphasizing that the design rules for PISA differ significantly from those of CDSA.Our prior reverse sequence PISA syntheses invariably yielded a spherical morphology. 23,36This is because such formulations always require a relatively large volume of hydrophilic monomer to solubilize the hydrophobic precursor, which inevitably leads to a relatively long steric stabilizer block.Such diblock copolymer compositions are known to favor the formation of spheres, rather than worms or vesicles. 22−35 Hence the judicious combination of reverse sequence PISA with CDSA is an important advance because it provides access to a significantly wider range of copolymer morphologies.The mean % degree of crystallinity, D c , of the PLLA 14 -PDMAC 300 platelets and PLLA 14 -PDMAC 70 rods was determined by X-ray diffraction (XRD), see Figure 4.The diffraction pattern recorded for the PLLA 14 -TTC precursor has a Bragg peak at 17°that corresponds well to that reported in the literature. 37he D c for this reference sample was 41%.Similarly, D c values of 16% and 2% were calculated for PLLA 14 -PDMAC 70 and PLLA 14 -PDMAC 300 , respectively.
Aqueous electrophoresis studies confirmed that both PLLA 14 -PDMAC 300 and PLLA 14 -PDMAC 70 nanoparticles exhibited essentially zero zeta potentials from pH 4 to 9 (see Figure S9), which is consistent with the non-ionic nature of the PDMAC steric stabilizer chains.Differential scanning calorimetry (DSC) studies were performed to examine whether the PLLA 14 -TTC precursor exhibited crystallinity (see Figure S1).As expected, a well-defined glass transition temperature (T g ), crystallization temperature (T c ), and melting temperature (T m ) were observed. 38−11 Given that (i) the DMAC polymerization is initially performed in the bulk and (ii) the DMAC monomer is a good solvent for both PLLA and PDMAC, no in situ selfassembly should occur prior to addition of water at a suitable intermediate DMAC conversion.Since water is a bad solvent for PLLA and a good solvent for PDMAC, its addition should result in immediate self-assembly of the growing diblock copolymer chains to form nascent PLLA-core nanoparticles.More specifically, PLLA 14 -PDMAC 70 and PLLA 14 -PDMAC 300 reaction mixtures were diluted with water after 32 and 29 min, respectively.The corresponding instantaneous DMAC conversions were 58 and 13%, which correspond to unreacted DMAC/water mass ratios of 1:3 and 1:7, respectively.These concentrated aqueous dispersions were then immediately further diluted to 0.1% w/w for TEM studies, which indicated the formation of nascent spherical aggregates in each case, see Figure 5.However, the final copolymer morphology was either rods or platelets after annealing for 16 h (>99% DMAC conversion).In traditional CDSA syntheses, thermal annealing is important for the growth of the initial copolymer seeds to form the final anisotropic nanoparticles. 39,40Accordingly, we examined the effect of annealing PLLA 14 -PDMAC 70−300 nanoparticles at 70 °C.When targeting PLLA 14 -PDMAC 70 nanoparticles, 1 H NMR studies indicated more than 99% DMAC conversion within 2 h at 70 °C.TEM analysis indicated the presence of rod-like nanoparticles at this time point, but some aggregates were also observed, see Figure 5. Annealing at 70 °C for 6 h leads to the disappearance of these aggregates, with no further change in copolymer morphology being observed up to 16 h.The corresponding DLS experiments corroborate the TEM studies: z-average diameters of 224 and 194 nm were obtained after 2 and 6 h, respectively.After 16 h, the z-average diameter remained almost unchanged at 190 nm.
Similarly, more than 99% DMAC conversion was achieved within 2 h when targeting PLLA 14 -PDMAC 300 platelets.At this time point, diamond platelets can be observed with a mean long axis of up to 1.3 μm, see Figure 5. Annealing at 70 °C led to the formation of progressively larger diamond platelets: the mean long axis of 1.7 μm observed after 4 h increased to 1.9 μm after 6 h and 2.3 μm after 16 h.Again, DLS studies are consistent with these observations: the apparent z-average diameter increased from 74 nm after 2 h to 225 nm after 16 h.It is perhaps worth emphasizing that these DLS values differ significantly from the mean long axes observed by TEM because the Stokes−Einstein equation used to calculate the zaverage diameter assumes a spherical morphology. 41Comparable results were reported by O'Reilly and co-workers when annealing PLLA 48 -PDMAC 1000 diamond platelets during conventional CDSA syntheses conducted in dilute ethanol at 90 °C. 33n another experiment, rod-like nanoparticles were targeted while varying the PLLA DP.Accordingly, PLLA 34 -TTC and PLLA 48 -TTC precursors were prepared via anionic ROP using the same hydroxy-functional RAFT agent and characterized by NMR and GPC analysis (see Figures S10−S12).Subsequently, each precursor was chain-extended in turn with DMAC, initially via bulk polymerization followed by dilution with water at intermediate conversion.Unfortunately, the PLLA 48 -TTC precursor could not be molecularly dissolved in the DMAC monomer: the initial reaction mixture remained slightly turbid even at 70 °C.Subsequent chain extension when targeting a PDMAC DP of 400 produced an ill-defined copolymer with a dispersity above 1.80.Moreover, increasing the reaction temperature up to 90 °C did not alleviate this problem: the initial reaction mixture was always turbid rather than transparent.Thus the new approach reported herein may be limited to relatively short PLLA DPs, at least when using DMAC monomer.Nevertheless, this is sufficient to provide access to higher order morphologies.
In contrast, the PLLA 34 -TTC precursor proved to be soluble in DMAC monomer at 70 °C when targeting a DP of 150, and the ensuing polymerization resulted in reasonably well-defined diblock copolymers (M w /M n = 1.20) at a final concentration of 30% w/w solids, see Table 1 and Figure 6.Furthermore, UV GPC analysis confirmed no significant contamination from the PLLA 34 precursor (see Figure S13).In this case, the final reaction mixture formed a thick paste.XRD analysis indicated D c values of 31% and 12% for the PLLA 34 -TTC precursor and PLLA 34 -PDMAC 150 rod-like nanoparticles, respectively (see Figure S14).TEM analysis of dilute aqueous dispersions of the PLLA 34 -PDMAC 150 nanoparticles confirmed that a rod-like morphology was formed when targeting a PDMAC DP of 150, see Figure S15.
The effect of varying the final nanoparticle concentration was also examined.Hence PLLA 34 -PDMAC 150 nanoparticles were prepared at 20 and 40% w/w solids to compare with syntheses targeting 30% w/w solids.In both cases, reasonably well-defined diblock copolymer chains were obtained with dispersities of 1.26 and 1.29 respectively, see Figure S16.Furthermore, the same rod-like morphology was obtained when targeting 40% w/w solids, see Figure 7.In contrast, only ill-defined aggregates were obtained when targeting 20% w/w solids, see Figure S15.
For the same target PLLA 34 -PDMAC 150 nanoparticles prepared at 30% w/w solids, the time at which water was added to the reaction mixture was systematically varied.No macroscopic precipitation was observed regardless of whether such dilution occurred after 10, 15 or 20 min (which corresponds to intermediate DMAC conversions of 19, 38 and 43%, respectively).Essentially the same diblock copolymer chains were produced when water was added after either 15 or 20 min (see Figure S17) and a rod-like morphology was obtained in each case (see Figures S15b and S18).In contrast, a significantly lower copolymer molecular weight and a higher dispersity were obtained when water was added after 10 min.This suggests the loss of RAFT control under the latter conditions, presumably because the nascent PDMAC chains are too short to confer effective steric stabilization.Water addition was also attempted after 25 min.However, the reaction mixture had solidified at this time point and could not be diluted.Nevertheless, these experiments suggest that our new reverse sequence aqueous PISA plus CDSA protocol can be used to target a range of final copolymer concentrations and is reasonably tolerant of the precise time chosen for water addition.
Finally, a second hydrophilic vinyl monomer was briefly examined.Accordingly, N-acryloylmorpholine (NAM) was used as an alternative hydrophilic monomer to DMAC when targeting a DP of either 100 or 400 using the PLLA 14 -TTC precursor.Essentially full NAM conversion was achieved (see Figure S19), but copolymer dispersities were broader than those achieved with the DMAC monomer (see Figure S20).More specifically, copolymer dispersities of 1.54 and 1.69 were obtained when targeting PLLA 14 -PNAM 400 and PLLA 14 -PNAM 100 , respectively.Furthermore, when targeting a PNAM DP of 400, the chain extension efficiency was estimated to be only around 75%. Targeting a mean DP of 100 resulted in rod-like nanoparticles, see Figure 8.However, no welldefined aggregates could be obtained when targeting PLLA 14 -PNAM 400 .
PLLA is widely recognized to be hydrolytically degradable owing to the cleavable ester bonds within its backbone. 42ence full degradation of this hydrophobic polyester block should yield water-soluble PDMAC chains.Accordingly, degradation studies were undertaken at 37 °C by preparing 1.0% w/w dispersions of PLLA 14 -PDMAC 40 nanoparticles via dilution using an acidic, basic or neutral aqueous buffer.The rod-like morphology of the PLLA 14 -PDMAC 40 nanoparticles was confirmed by TEM studies (see Figure S21).GPC was used to assess the extent of the hydrolytic degradation over time.As expected, significant degradation of the diblock copolymer chains was observed within one week for PLLA 14 -PDMAC 40 nanoparticles stored at pH 10.8 (see Figure 9a).Hydrolytic degradation was also observed in an acidic buffer solution (pH 2.9) and in the presence of a PBS buffer (pH 7.4), albeit at a somewhat slower rate (Figure 9b,c).This is because the base-catalyzed hydrolysis of ester bonds is known to be faster than acid-catalyzed hydrolysis. 43However, aging a 30% w/w aqueous dispersion of the same nanoparticles in deionized water (pH 6.7) at 20 °C led to minimal discernible degradation over four weeks (see Figure 9d).Furthermore, these PLLA 14 -PDMAC 40 nanoparticles retained their colloidal stability over the same time period (see Figure S22).
DLS studies of these rod-like PLLA 14 -PDMAC 40 nanoparticles in mildly basic solution (pH 10.8) confirmed a significant reduction in light scattering count rate from 56,000 to 5000 kcps within four weeks.This suggests that the original nanoparticles are converted into water-soluble PDMAC 40 chains.Similarly, GPC analysis confirmed that the diblock copolymer M n was reduced from 9.6 to 5.7 kg mol −1 within 24 h during an accelerated aging experiment performed at 60 °C in the presence of 5.0% w/w aqueous KOH, indicating complete degradation of the PLLA block.Concomitant DLS studies indicated a substantial reduction in the light scattering count rate from 60,000 to 500 kcps, while the number-average particle diameter was reduced from 160 to 2.8 nm.This is consistent with nanoparticle dissolution to form water-soluble PDMAC chains.One reviewer of this manuscript has pointed out that this non-degradable component comprises the majority of the mass of the original nanoparticles.

■ CONCLUSIONS
In summary, reverse sequence PISA has been combined with CDSA to enable the efficient preparation of 30% w/w aqueous dispersions of highly anisotropic hydrolytically degradable PLLA 14 -PDMAC x diblock copolymer nanoparticles.The crystalline nature of the hydrophobic PLLA block produces either diamond-like platelets (e.g., PLLA 14 -PDMAC 300 ) or short rod-like particles (e.g., PLLA 14 -PDMAC 70 ). 1 H NMR spectroscopy analysis confirms that approximately 99% DMAC conversion is achieved within 100 min at 70 °C when targeting PLLA 14 -PDMAC 120.A linear increase in molecular weight with increasing conversion is observed, but relatively broad molecular weight distributions are observed owing to the use of a suboptimal RAFT agent.Nevertheless, M w /M n values do not exceed 1.44 for syntheses conducted at 70 °C, and this minor technical problem should be readily addressable in the future.Given that these anisotropic nanoparticles are prepared directly in concentrated aqueous media, this is the first truly viable route for their industrial manufacture.Furthermore, preliminary data suggest that a PLLA 34 precursor and an alternative hydrophilic vinyl monomer (NAM) can be employed for such syntheses.Importantly, such nanoparticles are susceptible to hydrolytic degradation.We anticipate that this highly convenient new synthetic protocol should aid the evaluation of these anisotropic nanoparticles as next-generation sustainable Pickering emulsifiers 34 and foam stabilizers. 44

Figure 1 .
Figure 1.Kinetics of polymerization for the synthesis of PLLA 14 -PDMAC 120 nanoparticles, where polymerization is initiated in the bulk at 70 °C followed by dilution with degassed deionized water after 23 min (DMAC conversion = 57%) to target 30% w/w solids.Conditions: [PLLA 14 -TTC]/[ACVA] molar ratio = 10.(a) Conversion vs time curve (black data) obtained by 1 H NMR spectroscopy studies and (b) corresponding molecular weight (M n , blue points) and dispersity (M w /M n , red points) vs conversion data obtained by DMF GPC analysis (expressed relative to a series of PMMA calibration standards).Selected GPC traces are shown in Figure S5.

Figure 2 .
Figure 2. (a) DMF GPC curves (refractive index detector) recorded for a series of PLLA 14 -PDMAC 70−300 diblock copolymers and the corresponding PLLA 14 -TTC homopolymer.Each copolymer was prepared by reverse sequence aqueous PISA at 70 °C.(b) Corresponding number-average molecular weight (M n ) and dispersity (M w /M n ) data plotted against the target PDMAC DP.

Table 1 .
Summary of Dilution Times, Intermediate DMAC Conversions, and Molecular Weight Data Obtained for a Series of PLLA 14-34 -PDMAC 70-300 Nanoparticles and PLLA 34 -PDMAC 150 Nanoparticles a a More than 99% DMAC conversion was achieved for all diblock copolymer syntheses.