The Effect of Topology on Block Copolymer Nanoparticles: Linear versus Star Block Copolymers in Toluene

Linear and star block copolymer (BCP) nanoparticles of (polystyrene-block-poly(4-vinylpyridine))n (PS-b-P4VP)n with arm numbers of 1, 2, 3, and 4 were prepared by two methods of polymerization-induced self-assembly (PISA) and general self-assembly of block copolymers in the low-polar organic solvent, toluene. The effect of the arm number on the size and/or morphology of the (PS-b-P4VP)n nanoassemblies synthesized by the two methods in toluene and on the polymerization kinetics was investigated in detail. Our results show that in toluene, a low-polar solvent, the topology not only affected the morphology of the BCP nanoparticles prepared by PISA, but also influenced the BCP nanoparticles synthesized through general self-assembly.

It is well known that the morphologies of amphiphilic BCPs are dependent on the solvent characteristics [20,21] and especially the intrinsic molecular architecture [22][23][24][25][26][27]. Star block copolymers include a single branch point from which chemically different building blocks spread out [28][29][30][31]. Owing to this exclusive structure, star block copolymers show many interesting characteristics and properties unattainable by linear polymers [22][23][24][25][26][27]. For example, Ma and coworkers successfully synthesized the novel dual-functional linear and star POSS-containing organic-inorganic hybrid block copolymer poly(glycidyl methacrylate)-block-poly(methacrylisobutyl polyhedral oligomeric silsesquioxane) 1,4,6 ((PGMA-b-PMAPOSS) 1,4,6 ) by the core-first ATRP method, then crosslinked the block copolymers in the presence of trimethylamine to form a 3-dimensional network. Furthermore, the self-assembly behavior of these block copolymers with similar volume fractions of insoluble block was investigated in detail in water. It was revealed that the linear block copolymers (BCPs) PGMA-b-PMAPOSS showed the formation of spherical micelles with a size of 140-200 nm, star four-arm BCPs (PGMA-b-PMAPOSS) 4 assembled as a multi-core core-shell morphology, and star six-arm BCPs (PGMA-b-PMAPOSS) 6 assembled as the dendritic feature [26]. Huh and his team found that self-assembling the amphiphilic star poly(ethylene glycol)-[poly(ε-caprolactone)] 2 and the corresponding linear counterparts with approximately the same chemical composition in selective solvent lead to cylindrical and spherical micelles, respectively [27].
The preparation of star block copolymer self-assemblies through the first strategy, that is, general self-assembly, has been widely reported [9,10]. In addition, there are also few reports on the preparation of star block copolymer assemblies via the PISA method, and most of them are carried out in a polar solvent, such as water or ethanol/water mixture solution [32][33][34][35][36][37]. As we all know, solvent properties or composition have an impact on the morphology of block copolymer nano-assemblies [20,21]. In addition, BCP nanoparticles prepared in low polar solvents could be easily used as lubrication and emulsifiers for water/oil emulsions [38,39]. However, rare reports on the effect of the arm number on block copolymer nano-assemblies, i.e., linear block copolymers versus star block copolymers in low polar organic solvent, have been reported.
The polystyrene-mediated RAFT dispersion polymerization of the 4-vinylpyridine (4VP) monomer is a typical example that uses the low-polar solvent, toluene [40,41]. In this paper, the effects of the topological structure on the morphology and evolution process of block copolymer nanoassemblies in toluene were studied using RAFT dispersion polymerization of 4VP regulated by linear or star PS macro-RAFT agents. Firstly, block copolymer nanoassemblies of (polystyrene-block-poly(4-vinylpyridine)) n ((PS-b-P4VP) n , n = 1, 2, 3, and 4) were formed through RAFT dispersion polymerization in toluene employing mono-and multi-functional macromolecular chain transfer agents. Then, the polymerization kinetics of these dispersion RAFT polymerizations were investigated, and the effect of the arm number or the DPs of the PS/P4VP chain on the block copolymer nanoassemblies prepared via PISA was explored by changing the length of PS or P4VP. Finally, 2-, 3-, and 4-arm star and linear block copolymers of (PS-b-P4VP) 1,2,3,4 with similar chemical compositions were selected to prepare the nanoassemblies in toluene by the general self-assembly method to further prove that the topology also has an important effect of on the morphology of block copolymer nanoassemblies prepared by general self-assembly.

Materials
Styrene (St, >98%, Shanghai chemical reagent, Shanghai, China) and 4-vinylpyridine (4VP, 96%, Aladdin) were purified under reduced pressure for later use. The synthesis of the RAFT reagent of four small molecules called I, II, III, IV as shown in Scheme S1 was carried out according to the steps in our previous paper [42,43] and the detailed synthesis process and structure characterization are described in the Supporting Information (Scheme S2) and Figure S1. 2,2 -Azobis(2-methylpropionitrile) (AIBN, >99%, Tianjin Chemical Company, Tianjin, China) was recrystallized from ethanol.

Synthesis of Linear and Star macro-CTAs
Linear and star macromolecular chain transfer agents (macro-CTAs) of (PS m -TTC) n , in which m and n represent the arm number of the PS arms and the DP of each PS arm, with TTC representing the RAFT terminal of trithiocarbonate, were synthesized by solution RAFT polymerization of St monomers employing I-IV as chain transfer agents (CTAs).
Here is a typical synthesis of 3-arm (PS 25 (Table 1). The styrene monomer, CTA and AIBN were dissolved in toluene to obtain a homogeneous solution with a solid content of 20 wt%, which was subjected to RAFT dispersion polymerization. Taking  To check the resultant block copolymer nanoassemblies, a small drop of the block copolymer dispersion was deposited onto a piece of copper grid, dried at room temperature under vacuum, and then observed by transmission electron microscope (TEM). To collect the block copolymer for GPC analysis and 1 H NMR analysis, the synthesized (PS-b-P4VP) 3 nanoassemblies were diluted with dichloromethane and precipitated into ethanol, and finally dried at 40 • C under vacuum.

Preparation of the Block Copolymer Micelles through Self-Assembly in Toluene
The above-synthesized linear or star (PS-b-P4VP) n (n = 1, 2, 3) block copolymer nanoassemblies were centrifuged and dissolved in dichloromethane (DCM), then precipitated in toluene, centrifuged and washed with ethanol/toluene (v/v = 1:8) three times, and finally dried at 40 • C under vacuum to obtain the (PS-b-P4VP) n (n = 1, 2, 3) block copolymers. The above processes were designed to eliminate the morphology of the block copolymer nanoassemblies obtained in toluene via the PISA method. Then, (PS-b-P4VP) n was dissolved in DCM at room temperature to prepare a 0.5 mg/mL solution, adding a given volume of toluene at a rate of 1 drop (1 drop was about 7 µL) every 10 s under stirring. With the addition of toluene, the solution became turbid, indicating the formation of nanoassemblies. Toluene was then added slowly until the concentration of block copolymer nanoassemblies was about 0.2 mg/mL. Finally, the DCM in the solution was removed under vacuum at 25 • C, and a dispersion solution of block copolymers with 0.2 mg/mL concentration was obtained. TEM was used to characterize the morphology of the (PS-b-P4VP) n (n = 1, 2, 3) block copolymer nanoassemblies.

Characterization
The molecular weight (M n ) and dispersity (Ð, Ð = M w /M n ) of the polymers were obtained using a Waters 1525 µ gel permeation chromatograph. The samples were passed through three columns using DMF solution containing 0.05 M lithium bromide as the eluent at a flow rate of 1.0 mL/min at 50 • C. The narrow-polydispersity samples of polystyrene were used as a calibration standard. The 1 H NMR analysis was performed on a Bruker Avance II 400 MHz NMR spectrometer using CDCl 3 as a solvent. The TEM observation was performed using FEI Tecnai 12 transmission electron microscopy (TEM) with 120 kV or JEM-2100 TEM with 200 kV, whereby a small drop of the block copolymer dispersion was deposited onto a piece of copper grid and then dried at room temperature under vacuum.

Synthesis of macro-CTAs with Linear and Star Structure
Due to the similar R and Z groups in the structure of I, II, III, IV CTAs (Scheme S1), each arm of the synthesized star macro-CTAs was assumed to have a similar degree of polymerization. The linear and star macro-CTAs of (PS) n -TTC (n = 1, 2, 3, 4) (Scheme 1) were synthesized by solution RAFT polymerization employing an initiator of AIBN and chain transfer agents (CTAs) of I, II, III or IV. By varying the molar ratio of St/CTAs/AIBN (Table 1), (PS) n with a suitable DP of the PS arms was prepared. The (PS-TTC) n (n = 1, 2, 3, 4) macro-RAFT agents were characterized by 1 H NMR analysis and GPC analysis, and the results for the typical (PS 24 -TTC) n (n = 1, 2, 3, 4) with a similar DP of the PS arms, around 25, are shown in Figures 1 and 2. The molecular weight, M n,NMR by 1 H NMR analysis was calculated by comparing the integration area of the peaks at 0.88 ppm and 6.20-7.20 ppm attributed to the RAFT terminal and benzene ring. Furthermore, the M n,NMR was approximately equal to the theoretical molecular weight M n,th calculated from the monomer conversion according to eqn S1 (Table 1). In Figure 2, unimodal GPC traces were observed for linear PS 24 -TTC, (PS 24 -TTC) 2 , and (PS 25 -TTC) 3 , while for the 4-arm star (PS 25 -TTC) 4 , a small acromion appeared on the highmolecular-weight side. In the synthesis of star polymers, a molecule usually contains more than one propagating radical, which is more likely to lead to a bimolecular termination reaction as discussed elsewhere [44]. However, the dispersity of star (PS-TTC) n was narrow, which can be verified from D < 1.3 (Table 1). It should be noted that the molecular weight obtained by GPC for the star-shaped (PS-TTC) n prepared in this study was close to the amount obtained by NMR characterization, which was not quite consistent with previous results recorded in the literature, in that M n,GPC was less than M n,NMR for star-shaped polymers [45,46], which will also be discussed in the following sections. (Note: herein the DP values of PS 24 -TTC, (PS 24 -TTC) 2 , (PS 25 -TTC) 3 , and (PS 25 -TTC) 4 were calculated

Effect of the Arm Number on BCP Nanoassemblies Prepared through PISA
It has been reported that the arm number of star polymers has an important effect on the morphology of block copolymer nanoassemblies in ethanol/water. However, the topology of star polymers on the morphology of nanoassemblies formed through PISA in toluene has not been studied. To achieve this, we synthesized a series of (PS-b-P4VP)n nanoparticles with different compositions by varying the length of the PS or P4VP chain segments according to scheme S3, and characterized their morphologies by TEM.

Effect of the Arm Number on BCP Nanoassemblies Prepared through PISA
It has been reported that the arm number of star polymers has an important effect on the morphology of block copolymer nanoassemblies in ethanol/water. However, the topology of star polymers on the morphology of nanoassemblies formed through PISA in toluene has not been studied. To achieve this, we synthesized a series of (PS-b-P4VP)n nanoparticles with different compositions by varying the length of the PS or P4VP chain segments according to scheme S3, and characterized their morphologies by TEM.

Effect of the Arm Number on BCP Nanoassemblies Prepared through PISA
It has been reported that the arm number of star polymers has an important effect on the morphology of block copolymer nanoassemblies in ethanol/water. However, the topology of star polymers on the morphology of nanoassemblies formed through PISA in toluene has not been studied. To achieve this, we synthesized a series of (PS-b-P4VP) n nanoparticles with different compositions by varying the length of the PS or P4VP chain segments according to Scheme S3, and characterized their morphologies by TEM.  4 nanospheres, respectively. When the polymerization degree of PS arms is short, such as at about 17 or 24, linear block copolymers still form discrete nanoassemblies, while star block copolymers n = 2, 3, and 4 all formed aggregates, especially the aggregates of the 3-and 4-armed block copolymers. The self-assembly of star block copolymers or linear BAB (B is insoluble chain segment) polymers can also form large aggregates, which may be due to the formation of bridging interactions between nanoparticles [47][48][49][50].
nm (PS60-b-P4VP274)4 nanospheres, respectively. When the polymerization degree of PS arms is short, such as at about 17 or 24, linear block copolymers still form discrete nanoassemblies, while star block copolymers n = 2, 3, and 4 all formed aggregates, especially the aggregates of the 3-and 4-armed block copolymers. The self-assembly of star block copolymers or linear BAB (B is insoluble chain segment) polymers can also form large aggregates, which may be due to the formation of bridging interactions between nanoparticles [47][48][49][50].   Figure 4 lists the different case of (PS-b-P4VP)n nanoparticles with a similar DP of the PS segments at about 25, while increasing the DP of P4VP segments. Correspondingly, (PS-b-P4VP)n (n = 1−4) with a short block chain of P4VP arms (DP < 200) formed discrete nanoassemblies, and star (PS-b-P4VP)n (n = 2, 3, 4) with long block chain of P4VP arms at about 285 formed bridged aggregates, while linear PS-b-P4VP with long block chain of P4VP still formed discrete nanoassemblies. It was found that both the linear and star (PSb-P4VP)n followed similar rules, i.e., the diameter of their nanoparticles increased with the DP of P4VP block increasing. This rule is consistent with the previous morphologic change in linear BCP nanoparticles [51][52][53][54]. In addition, the morphology of star (PS-b-P4VP)2-3  Figure 4 lists the different case of (PS-b-P4VP) n nanoparticles with a similar DP of the PS segments at about 25, while increasing the DP of P4VP segments. Correspondingly, (PS-b-P4VP) n (n = 1−4) with a short block chain of P4VP arms (DP < 200) formed discrete nanoassemblies, and star (PS-b-P4VP) n (n = 2, 3, 4) with long block chain of P4VP arms at about 285 formed bridged aggregates, while linear PS-b-P4VP with long block chain of P4VP still formed discrete nanoassemblies. It was found that both the linear and star (PS-b-P4VP) n followed similar rules, i.e., the diameter of their nanoparticles increased with the DP of P4VP block increasing. This rule is consistent with the previous morphologic change in linear BCP nanoparticles [51][52][53][54]. In addition, the morphology of star (PS-b-P4VP) 2-3 nanoparticles is much more complex than that of linear PS-b-P4VP.  Figure 4 lists the different case of (PS-b-P4VP)n nanoparticles with a similar DP of the PS segments at about 25, while increasing the DP of P4VP segments. Correspondingly, (PS-b-P4VP)n (n = 1−4) with a short block chain of P4VP arms (DP < 200) formed discrete nanoassemblies, and star (PS-b-P4VP)n (n = 2, 3, 4) with long block chain of P4VP arms at about 285 formed bridged aggregates, while linear PS-b-P4VP with long block chain of P4VP still formed discrete nanoassemblies. It was found that both the linear and star (PSb-P4VP)n followed similar rules, i.e., the diameter of their nanoparticles increased with the DP of P4VP block increasing. This rule is consistent with the previous morphologic change in linear BCP nanoparticles [51][52][53][54]. In addition, the morphology of star (PS-b-P4VP)2-3 nanoparticles is much more complex than that of linear PS-b-P4VP.

Effect of the Arm Number on Polymerization Kinetics and the Evolution of (PS-b-P4VP)n Nanoassemblies
To further study the effect of the arm number on the polymerization kinetics and the evolution of (PS-b-P4VP)n nanoassemblies, the (PS-TTC)n (n = 1-3) with a similar arm

Effect of the Arm Number on Polymerization Kinetics and the Evolution of (PS-b-P4VP) n Nanoassemblies
To further study the effect of the arm number on the polymerization kinetics and the evolution of (PS-b-P4VP) n nanoassemblies, the (PS-TTC) n (n = 1-3) with a similar arm length of about 25 mediated PISA of 4VP was studied in detail. In order to ensure that the polymerization conditions were similar, [4VP] 0 :[trithiocarbonate] 0 :[AIBN] 0 were designed as a constant in all PISA reactions.
As can be seen in Figure 5A, the monomer conversion-time in the three RAFT dispersion polymerization reactions was very similar, indicating that the three reactions had similar polymerization kinetics. This may be due to the fact that the (PS-TTC) n macromolecular chain transfer agents used had similar R and Z groups and close arm lengths. As shown in Figure 5B, the (PS-TTC) n (n = 1, 2, 3) mediated PISA of 4VP firstly underwent a slow polymerization process, followed by an accelerated polymerization after about 4 h reaction ( Figure 5B), which had similar polymerization kinetics as commonly reported for RAFT dispersion polymerization [55][56][57][58][59][60][61]. Generally, the synthesized 3-armed stars (PS 25b-P4VP) 3 were characterized by GPC ( Figure 5C) and NMR to obtain their M n, GPC and M n,NMR , respectively, and the results are listed in Figure 5D. A small acromion appeared when the monomer conversion was greater than 87% in the GPC curve of (PS 25 -b-P4VP) 3 , which was also reflected in the increase in the Ð value in Figure 5D. Moreover, the RAFT synthesis was controllable given that (PS 25 -b-P4VP) 3 had a star structure and Ð was around 1.3. It is worth mentioning that when monomer conversion was low, the molecular weight obtained by NMR was close to the molecular weight obtained by GPC, which is in agreement with that described above. When the monomer conversion was greater than 20%, the molecular weight obtained by GPC was smaller than that obtained by NMR, which is consistent with previous results [45,46]. In addition, the formation process of the 3-armed star-shaped (PS25-b-P4VP)3 nanoparticles was recorded by TEM as shown in Figure 6. With the increase in P4VP, (PS25-b-P4VP)3 nanoparticles changed from 16.3 ± 1.4 nm nanospheres at the beginning of 4h to 53.8 ± 3.9 nm nanospheres at 8h, and finally to worms of (PS25-b-P4VP190)3. However, In addition, the formation process of the 3-armed star-shaped (PS 25 -b-P4VP) 3 nanoparticles was recorded by TEM as shown in Figure 6. With the increase in P4VP, (PS 25 -b-P4VP) 3 nanoparticles changed from 16.3 ± 1.4 nm nanospheres at the beginning of 4 h to 53.8 ± 3.9 nm nanospheres at 8 h, and finally to worms of (PS 25 -b-P4VP 190 ) 3 . However, the linear PS 24 -b-P4VP was still a solid sphere when the length of P4VP was 290 (Figure 4). Figure 6F summarizes the average diameter (D) of the (PS 25 -b-P4VP) 3 and linear PS 24b-P4VP nanoparticles with the DP of P4VP arms increasing, showing that the average diameter of (PS 25 -b-P4VP) 3 changed very little after 10 h, which may be related to the change in morphology from nanospheres to worms. Moreover, the rate of change with the DP was different for the 3-armed (PS 25 -b-P4VP) 3

Effect of the Arm Number on BCP Nanoassemblies Prepared through General Self-Assembly in Toluene
As introduced above, the polymerization-induced self-assembly of block copolymers and the self-assembly of block copolymers in block-selective solvent represent two strategies to prepare block copolymer nano-objects. Herein, the effect of the arm number on BCP nanoassemblies prepared through general self-assembly in toluene was investigated by checking the morphology of the linear and star block copolymer nanoassemblies with similar chemical compositions. To depress the residual monomer effect, the block copolymer nanoassemblies prepared at high monomer conversion through the polymerizationinduced self-assembly were chosen. The preparation of the (PS-b-P4VP) n (n = 1, 2, 3) block copolymer nanoassemblies through general self-assembly in the block-selective solvent was achieved by initially dissolving the block copolymer in DCM, and then adding toluene slowly, finally removing the excess DCM under vacuum at room temperature as discussed elsewhere [42]. Whereas, differently from the highly concentrated block copolymer in the polymerization-induced self-assembly (~20 wt %), very diluted block copolymer (0.2 mg/mL) was employed in the present self-assembly strategy, since destabilization of the (PS-b-P4VP)n (n = 1, 2, 3) block copolymer dispersion was found when the block copolymer concentration was above 1 wt %. Under this diluted block copolymer concentration at 0.2 mg/mL, all the block copolymer nanoassemblies were of great stability. Figure 7 shows the TEM images of the linear or star (PS-b-P4VP) 3 block copolymer nanoassemblies with similar PS block chains at about 25 and similar P4VP block chains at about 140 (top in figure) and with similar PS block chains at about 25 and similar P4VP block chains at about 190 (bottom in figure) prepared through general self-assembly in toluene. It indicates that the PS 24 -b-P4VP 139 formed vesicles ( Figure 7A1), whereas the similar chemical composition of star block copolymers of (PS 24 -b-P4VP 144 ) 2 self-assembled into vesicles and bicontinuous nanospheres ( Figure 7B1). Similar bicontinuous nanospheres of the mixture of poly-(ethylene glycol)-b-polystyrene/polystyrene-b-poly(ethylene glycol)-bpolystyrene [62] and poly(ethylene oxide)-b-poly(octadecyl methacrylate) containing a long side octadecyl chain [63] and amphiphilic polynorbornene block copolymer [64] were also prepared, and star block copolymers of (PS 25 -b-P4VP 144 ) 3 self-assembled into aggregates ( Figure 7C1). The self-assembly of PS 24 -b-P4VP 188 in toluene resulted in large nanospheres ( Figure 7A2), and the self-assembly of star block copolymers of (PS 24 -b-P4VP 190 ) 2 and (PS 25b-P4VP 190 ) 3 with similar chemical composition lead to multilayered vesicles ( Figure 7B2), and bicontinuous nanospheres ( Figure 7C2). These results clearly demonstrate that the topology of block copolymers also influenced the nanoassemblies prepared through general self-assembly in toluene.  [62] and poly(ethylene oxide)-b-poly(octadecyl methacrylate) containing a long side octadecyl chain [63] and amphiphilic polynorbornene block copolymer [64] were also prepared, and star block copolymers of (PS25-b-P4VP144)3 self-assembled into aggregates ( Figure 7C1). The self-assembly of PS24-b-P4VP188 in toluene resulted in large nanospheres ( Figure 7A2), and the self-assembly of star block copolymers of (PS24-b-P4VP190)2 and (PS25-b-P4VP190)3 with similar chemical composition lead to multilayered vesicles ( Figure 7B2), and bicontinuous nanospheres ( Figure 7C2). These results clearly demonstrate that the topology of block copolymers also influenced the nanoassemblies prepared through general self-assembly in toluene.

Conclusions
In conclusion, linear and star BCP nanoparticles of (PS-b-P4VP)n with numbers of 1, 2, 3, and 4 were prepared by two methods of polymerization-induced self-assembly using (PS-TTC)n (n = 1, 2, 3, 4) macro-RAFT agents and self-assembly of block copolymers in the low-polar organic solvent, toluene. Furthermore, the effect of the topology on the size and/or morphology of the (PS-b-P4VP)n nanoassemblies synthesized by the two methods in toluene was investigated in detail, and star (PS-b-P4VP)n had a more complex morphology than its linear counterpart. In addition, the PISA process of linear or star

Conclusions
In conclusion, linear and star BCP nanoparticles of (PS-b-P4VP) n with numbers of 1, 2, 3, and 4 were prepared by two methods of polymerization-induced self-assembly using (PS-TTC) n (n = 1, 2, 3, 4) macro-RAFT agents and self-assembly of block copolymers in the low-polar organic solvent, toluene. Furthermore, the effect of the topology on the size and/or morphology of the (PS-b-P4VP) n nanoassemblies synthesized by the two methods in toluene was investigated in detail, and star (PS-b-P4VP) n had a more complex morphology than its linear counterpart. In addition, the PISA process of linear or star polymers with different topologies had similar polymerization kinetics. The possible reason is that all the macro-CTAs had similar R and Z groups and similar arm lengths. Our results show that in toluene, a low polar solvent, the topology not only influenced the morphology of the BCP nanoparticles synthesized by PISA, but also affected their nanoassemblies prepared through general self-assembly.

Conflicts of Interest:
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.