Influence of Chain Primary Structure and Topology (Branching) on

Article Cite This: Macromolecules XXXX, XXX, XXX−XXX

pubs.acs.org/Macromolecules

Influence of Chain Primary Structure and Topology (Branching) on Crystallization and Thermal Properties: The Case of Polysulfides Ricardo A. Peŕ ez-Camargo,† Richard d’Arcy,‡,§ Amaia Iturrospe,∥ Arantxa Arbe,∥ Nicola Tirelli,*,‡,§ and Alejandro J. Müller*,†,⊥

Downloaded via WEBSTER UNIV on February 21, 2019 at 05:17:41 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

POLYMAT and Polymer Science and Technology Department, Faculty of Chemistry, University of the Basque Country UPV/EHU, Paseo Manuel de Lardizabal 3, 20018 Donostia-San Sebastián, Spain ‡ Division of Pharmacy & Optometry, University of Manchester, Manchester, M13 9PT, U.K. § Laboratory of Polymers and Biomaterials, Fondazione Istituto Italiano di Tecnologia, 16163 Genova, Italy ∥ Centro de Física de Materiales (CFM) (CSIC-UPV/EHU) - Materials Physics Center (MPC), Paseo Manuel de Lardizabal 5, 20018 Donostia-San Sebastián, Spain ⊥ IKERBASQUE, Basque Foundation for Science, 48013 Bilbao, Spain S Supporting Information *

ABSTRACT: We have investigated how variations in primary structure (distribution of monomeric units, chain length) and degree of branching affect the crystallization behavior of polysulfides. In particular, we have used copolymers of crystallizable (ethylene sulfide, ES) and noncrystallizable (propylene sulfide, PS) units, prepared via anionic ringopening mechanism. PS units interrupt the crystallization of ES sequences; therefore, strong differences in thermal properties are expected between copolymers with a gradient (single addition of ES/ PS mixtures) or a semirandom structure (sequential additions, which yield a more constant composition along the chain). Additionally, we prepared different chain topologies: linear, 4- and 8-armed stars, and combs with 10, 15, and 20 arms, in a comparison where each arm had a gradient primary structure and degree of polymerization (DP) of 10, 20, and 30. The influence of these variables (topology, number of arms, and DP) on the thermal properties was studied by polarized light optical microscopy (PLOM), standard and advanced DSC techniques (i.e., self-nucleation and successive self-nucleation and annealing (SSA)), and small and wide angle X-ray scattering (WAXS and SAXS). First, we have confirmed the much higher order obtained in gradient polymers, in comparison to the semirandom ones. Second, we have seen that the type of crowding of polymer chains affected the level of order achievable. In star polymers, an increasing number of arms increased topological restrictions, which in turn decreased crystallization and melting temperatures, crystallinity, and lamellar thickness. For combs, which are characterized by a more parallel than convergent crowding of the chains, the increase in arms number did not produce a significant decrease in crystalline order; on the contrary, the higher density of chains caused moderate increases in crystallization and melting temperatures, crystallinity, and lamellar thickness.

1. INTRODUCTION Polysulfides have been extensively studied as oxidationresponsive materials, owing to the biological relevance of oxidation phenomena. Indeed, oxidants (reactive oxygen species, ROS) play a key role in a variety of cellular phenomena, including mitosis, angiogenesis, tumorigenesis, and above all inflammatory pathologies, and a number of materials have been developed to perform ROS-responsive actions with a therapeutic value.1−4 Some of these materials are based on organic sulfides, where the oxidation of hydrophobic sulfur(II) atoms to more hydrophilic sulfoxide or sulfone determines a phase transition (e.g., a solubilization) of the material,5 which can be used for example as a circulating nanocarrier6 for, e.g., site-specific drug release in a ROSresponsive fashion7 or to transduce a chemical signal such as glucose (via glucose oxidase) into a measurable output signal.8 © XXXX American Chemical Society

Importantly, the responsive behavior depends not only on the nature of the oxidant,9−11 but also on details of the polysulfide architecture, in particular the primary structure12 and the degree of branching.13 However, in addition to evaluating the final influence of the macromolecular architecture on the oxidative response, it is important to separate the direct effects (e.g., short-range hindrance by groups neighboring sulfur atoms) from those that are mediated by the capacity of producing ordered structures and by the presence of topological restrictions. In this study, we aimed to clarify the links between macromolecular architecture and long-range order capacity in Received: December 13, 2018 Revised: January 18, 2019

A

DOI: 10.1021/acs.macromol.8b02659 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Scheme 1. Polymer Architectures Investigated in This Studya

a

In linear polymers, the polysulfide chains were produced through a single addition of the monomer mixture (protocol a; single, longer gradient chains) or through repeated additions (protocol b shorter, double gradient; closer to a random distribution); in the scheme, we color-code ES in white and PS in blue to show the different gradient structures. The side chains of the branched polymers were prepared only with single gradient structure. Please note that for simplicity of presentation, the polysulfide chains (ES/PS copolymers) are replaced by the symbol R in the branched structures.

homosequences (ES-rich blocks) formed due to the higher reactivity of ES (rES = 0.58, rPS = 3.11) allow for the formation of crystalline phases.12 The crystallization of A-B random, tapered, or graft copolymers where one of the co-units cannot crystallize is very interesting, as the non-crystallizable units will interrupt the crystallizable sequences, thereby causing dramatic reductions in crystallization and melting temperatures and crystallinity degrees.20−23 On the other hand, depending on the distributions of the interrupting non-crystallizable sequences along the chain, complex distribution of lamellar thicknesses will be generated. In this case, such copolymers are ideal candidates to study their thermal behavior with thermal fractionation techniques, such as successive self-nucleation and annealing (SSA) as any type of defect that interrupts the crystallizable linear sequences will strongly promote molecular fractionation during crystallization.24−26 In the present work, P(PS-co-ES) were successfully prepared by employing two different protocols of anionic ring-opening copolymerization that provide different primary structures: (1) protocol a, a one-shot addition of all monomers to the initiators, which leads to gradient (blocky) structures (due to

polysulfides; specifically, we focused on how the latter are affected by primary structure (order of repeating units) and by branching (single or multiple branching points) in copolymers comprising a crystallizable with a non-crystallizable repeating unit (Scheme 1). The former is propylene sulfide (PS); atactic poly(propylene sulfide) (PPS, obtained from racemic PS) is an amorphous polymer with a low Tg,14 soluble in most organic solvents.12 The crystallizable unit is closely related to PS, but devoid of the methyl group: ethylene sulfide (ES). The regularity of the ES repeating units makes its homopolymer, poly(ethylene sulfide) (PES), highly crystalline, with a high melting point (i.e., 205−210 °C)15,16 and very poor solubility (only in drastic conditions such as nitrobenzene, >170 °C). If present in sufficiently long sequences, ES units can crystallize in copolymeric structures. PES, for example, confers crystallinity to block copolymers with isoprene, butadiene, or styrene.17 Due to the high tendency to crystallize, an ordered assembly can be obtained also with very short blocks; for example, ES sequences as short as three (oligo(ethylene sulfide), OES) allow association and gelation of PEG-OES copolymers,18,19 and it has been shown that when ES is copolymerized with other episulfides, including PS, the B


Article

Macromolecules Table 1. Characterization Data for the Polysulfides Prepared in This Worka degree of polymerization per armb linear

star

comb

samplea

arms

PS

L_(PS10)2 L_(PS20)2 L_(PS30)2 L_(PS7.5-ES2.5)2 L_(PS15-ES5)2 L_(PS22.5-ES7.5)2 La_(PS5-ES5)2 La_(PS10-ES10)2 La_(PS15-ES15)2 Lb_(PS10-ES10)2 Lb_(PS15-ES15)2 S_(PS10)4 S_(PS20)4 S_(PS30)4 S_(PS7.5-ES2.5)4 S_(PS15-ES5)4 S_(PS22.5-ES7.5)4 S_(PS5-ES5)4 S_(PS10-ES10)4 S_(PS15-ES15)4 S_(PS5-ES5)8 S_(PS10-ES10)8 S_(PS15-ES15)8 C_(PS15-ES15)10 C_(PS5-ES5)15 C_(PS10-ES10)15 C_(PS15-ES15)15 C_(PS15-ES15)20

2 2 2 2 2 2 2 2 2 2 2 4 4 4 4 4 4 4 4 4 8 8 8 10 15 15 15 20

9.5 18.4 28.9 7.1 15.2 21.8 5.1 9.5 14.0 9.7 14.5 9.4 20.6 28.5 7.5 15.2 22.8 5.0 10.3 14.8 5.1 9.6 14.3 16 5 10.3 14 15.7

Mn (g/mol) theor

exp

Đ

ac

end-capping yield (%)d

1700 3200 4700 1700 3100 4500 1600 2900 4300 2900 4300 3500 6500 9500 3400 6200 9000 3300 6000 8600 6800 1100 17000 24000 15000 25000 35000 47000

1450 2870 4440 1540 3100 3900 1650 2240 3730 2770 3810 3230 6610 9340 3390 6090 9110 3160 6320 8210 6670 13180 20630 26050 13120 27880 33900 45590

1.18 1.13 1.15 1.06 1.08 1.08 1.32 1.19 1.09 1.12 1.15 1.17 1.15 1.08 1.19 1.16 1.09 1.08 1.05 1.07 1.07 1.09 1.26 1.24 1.24 1.19 1.21 1.25

0.97 1.04 1.08 0.96 1.03 1.00 0.93 0.97 0.99 0.96 1.01 0.75 0.79 0.86 0.73 0.76 0.85 0.82 0.85 0.91 0.51 0.55 0.59 0.48 0.34 0.37 0.42 0.41

97 100 99 96 100 100 100 99 98 100 96 98 100 99 98 100 100 95 97 96 99 100 95 100 87 100 100 100

ES

2.3 4.9 7.2 5.0 9.7 14.2 9.8 14.3

2.3 5.1 7.3 4.7 10.1 14.6 4.9 9.8 14 14.5 4.4 9.7 14.5 15.1

The nomenclature reports first the chain topology (L for linear, S for stars, C for combs), then the primary structure (La, gradient-addition method a; Lb, semirandom, addition method b), the composition of the arms (PS and ES units) and finally the number of arms. In black we list the polymers used for the characterization studies reported below. The PS homo and copolymers listed in bold PS are those that exhibited too low crystallinity and were therefore excluded from further analysis. bCalculated as the integral ratio of main chain protons from PPS methyl peak (1.33−1.46 ppm) or PES repeat unit (−S-CH2−CH2−, 2.73−2.81 ppm) and the initiator peak (linear: −S−CH2−CH2−O−CH2−CH2−O− CH2−CH2−S− at 3.60−3.69 ppm, star: C−CH2−O−CH2−CH2−CH2−S− at 1.70−1.85 ppm or comb: triazole proton at 7.55−7.77 ppm). ca (the Mark−Houwink parameter) was calculated from the slope of the log−log MW vs [η] graph measured from the triple detection method in GPC using the Cirrus MultiOffline GPC/SEC (v3.4) software. dEnd-capping yield was calculated from 1H NMR by comparing the integral of the CH3CH2−O−C(O)− peak of ethyl 2-bromoacetate at 4.12−4.24 ppm to the initiator peaks (as described in footnote a). a

the higher reactivity ratio of ES to PS12); and (2) protocol b, repeated additions of the monomer mixture, which produce shorter gradients along the chain; this determines a local ES/ PS ratio that deviates less from that of the feed, and due to the presence of shorter homosequences of the crystallizable unit, the copolymers should exhibit a lower tendency to crystallize. Further, the copolymers obtained through the one-shot addition protocols were also produced in different chain topologies following procedures described in a recent paper:13 linear macromolecules obtained from a bifunctional initiator (theoretically a 2-armed star topology) were compared to “real” branched polymers with a 4- and 8-armed star structure produced from multifunctional initiators, and to comb polymers (10, 15, and 20 arms); the latter were obtained through the polymerization of propargyl episulfide, followed by click reaction with an azide-containing thioacetate, which was finally used to initiate the side-chain copolymerization of ES/ PS. In all cases, the degree of polymerization (DP) was also varied, using DP = 10, 20, or 30 per arm, which for linear

polymers corresponds to a total DP of 20, 40, or 60 monomeric units The objectives of the paper are to study the influence of (a) the monomer addition protocol employed during the polymerization, which influences the distribution of comonomeric sequences/sequence lengths and (b) the chain topology (type and number of branches) on the morphology, nucleation, and crystallization of these complex P(PS-co-ES) copolymers. Several experimental techniques are used to accomplish these objectives including: polarized light optical microscopy (PLOM), differential scanning calorimetry (DSC), SSA and X-ray diffraction techniques, such as SAXS and WAXS.

2. MATERIALS AND METHODS 2.1. Initiator and Polymer Synthesis and Molecular Characterization. All synthetic operations are described in the Supporting Information, section 1.2SI. The physicochemical characterization is described in section 1.3SI and Table 1. 2.2. Polarized Light Optical Microscopy (PLOM). A polarized light optical microscope, Olympus BX51, was employed, incorporating a λ plate in between polarizers at 45° to facilitate observation. The C


Article

Macromolecules

Figure 1. (A) Persistence length of chains with identical composition (1:1 ES/PS) is inversely proportional to the slope of the plot of

Mw2 [η ]

1/3

( )

vs

1/2

M w . Our analysis stops at the quantification of the slopes without explicitly calculating the persistence lengths, because of the likely large uncertainties in (a) the intrinsic viscosity values obtained from GPC measurements (nonzero shear) and (b) the values of mass per unit length of the polymers (used to calculate the persistence length from the slopes), which in gradient polymers inherently vary along the chain. (B) Left: Plots of Rg and RΗ vs molecular weight for all polymers with 1:1 ES/PS composition. RΗ is obtained as viscometric radius (Rη = (3[η]Mw̅ )1/336); the R approximation RH ≈ Rη is assumed valid in the melt,37 and to at least a qualitative extent in solution too ( η ≈ 1−1.2).38 Right: plots of the shape RH

parameter

Rg RH

vs molecular weight and of the Mark−Houwink parameter a vs degree of branching; in both cases, it is apparent that already 8-armed Rg

stars are rather compact structures and combs are increasingly similar to hard spheres ( R ≤ 1, a ≤ 0.4 ). H

2.6. Wide and Small Angle X-ray Scattering (WAXS and SAXS). A Bruker D8 Advance diffractometer working in parallel beam geometry with Cu Kα transition photons of wavelength λ = 1.54 Å was employed for WAXS measurements. A linear detector by LYNXEYE was used with an active area of 14.4 mm × 16 mm. The experiments were performed in reflection mode (θ − 2θ configuration), varying the scattering angle 2θ from 10° to 30° with steps of 0.05°. The measuring time was 10 s/point. SAXS experiments were performed with a Rigaku three-pinhole PSAXS-L instrument operating at 45 kV and 0.88 mA. A MicroMax002+ X-ray generator system was employed. This system is composed of a microfocus sealed tube source module and an integrated X-ray generator unit which also produces Cu Kα transition photons. Both flight path and sample chamber are under vacuum. A two-dimensional multiwire X-ray detector (Gabriel design, 2D-2000X) was used. The azimuthally averaged scattered intensities were measured as a function of wave vector q, and q = 4πλ−1 sin θ. Reciprocal space calibration was performed using silver behenate as standard. The samples were films that were placed in a Linkam Scientific Instruments THMS 600 temperature controller (range −196 to +600 °C, stability < 0.1 °C) in transmission geometry, with sample−detector distances of 2 m and 50 cm. Measuring times of 20 min were employed. In some cases, when the material was in the rubbery state, the films could not be measured, as they flow out of the measuring spot. Measurements were only performed on samples that had enough dimensional stability during measurements. In temperature-dependent experiments, patterns were taken at room temperature and during heating.

microscope was equipped with an Olympus SC50 digital camera. A Mettler Toledo FP82 hot stage was coupled to the microscope. Film samples with 10 μm thickness were sandwiched between two glass slides and they were first heated to 140 °C in order to erase their thermal history and then crystallized from the melt by cooling to 30 °C (this Tc was selected for comparison purposes) at 20 °C/min. 2.3. Differential Scanning Calorimetry (DSC). A PerkinElmer DSC PYRIS 1 instrument was employed. All measurements were performed under ultrahigh purity nitrogen as inert atmosphere. Sample masses of 8−11 mg were used, and the instrument was calibrated with indium and zinc standards. The samples were first heated to 140 °C and kept at that temperature for 3 min in order to erase all crystalline thermal history. Then the samples were cooled at 20 °C/min and the corresponding cooling scans were recorded. Finally, the samples were heated at the same rate to register the subsequent heating scans. 2.4. Self-Nucleation (SN) Studies. Self-nucleation is a thermal technique designed to produce self-nuclei within a polymer melt, so that its nucleation density can be greatly enhanced or controlled. The best nucleating agent for any polymer should be its own crystallographically ideal crystal fragments or chain segments with residual crystal memory.25−29 Fillon et al. adapted self-nucleation techniques to DSC.27 Self-nucleation was performed in order to determine the ideal self-nucleation temperature (Ts,ideal) needed to design the SSA protocol (see below). For that purpose, the samples with DP = 30 were employed (since they possess the highest melting points). The SN procedure applied here follows that used in the literature,25,27 in which Fillon et al.27 defined the so-called “domains” of self-nucleation (see the SN section in the Supporting Information for a detailed description of the SN domains (section 1.4SI)). After analyzing the SN results (see Figure S1), it was decided to employ the highest Ts,ideal (i.e., 111 °C, corresponding to the linear DP30 sample) as the first Ts temperature for the SSA experiments. 2.5. Thermal Fractionation by Successive Self-Nucleation and Annealing (SSA). Thermal fractionation applies a carefully designed temperature program (i.e., a series of heating and cooling steps) to a crystallizable polymer in order to produce a distribution of lamellar crystals or thermal fractions based on the molecular fractionation that occurs during crystallization. The SSA technique was developed by Müller et al. and has been recently reviewed.24−26,30−32 After thermal conditioning of the sample, a final DSC heating scan reveals the distribution of melting points induced by SSA. The details of the SSA procedure employed here can be found in the Supporting Information (section 1.4SI)

3. RESULTS AND DISCUSSION 3.1. Polymer Synthesis. All polymers prepared in this work exhibited negligible deviations from the theoretical composition and DP (Table 1), and were characterized by narrow molecular weight distribution (generally Đ = 1.05− 1.2), except for combs (Đ = 1.19−1.25). The higher dispersity of the latter is due to (A) the cumulative effect of the two polymerization steps (first main chain, then side chains), and (B) possible disulfide formation during/at the end of the side chain polymerization, due to the close proximity of the thiolate propagating species. It is worth mentioning that the values of the Mark−Houwink a parameter ranged from around 1 for linear polymers to less than 0.5 for combs, indicating that the D


Article

Macromolecules

Figure 2. Polarized light optical micrographs showing granular birefringent crystalline texture for 8-armed stars with DP per arm of 10 (A, S_(PS5ES5)8), 20 (B, S_(PS10-ES10)8), and 30 (C, S_(PS15-ES15)8), taken at 30 °C.

polymer coils clearly became more compact with increasing degree of branching. PS was used as a racemic mixture, leading to polymers with an atactic distribution of the PS units; since the latter are therefore incapable of crystallization, in the copolymers they will disrupt that of ES sequences, and this will occur in a fashion proportional to degree of ES:PS interdispersion. The synthesis of polymers with ES:PS ratios > 0.5 resulted in precipitation during polymerization, whereas an excess PS caused a lack of any significant crystallization (not shown), therefore we limited our crystallization studies to the 50:50 copolymers. It is worth mentioning that in solution the introduction of branching dramatically changed the macromolecular structure. For example, it is apparent that independently of the nature or the degree of branching, all stars and combs have the same persistence length, (from molecular weight and intrinsic viscosity data through the Bushin−Bohdanecky method,33−35 Figure 1A), which is considerably smaller than that of the linear polymers. This effect is likely to be unrelated to chain flexibility, and stems from a purely geometrical cause, i.e., the “kink” in the chains in correspondence to branching points. 3.2. Morphology by PLOM. Isothermal crystallization experiments were performed by PLOM at 30 °C, after cooling from the melt (i.e., 140 °C). For all samples, nucleation took place almost instantaneously, and a granular crystalline superstructure was observed. Therefore, the spherulitic growth rate was impossible to determine. The number of nuclei was also found to increase with increasing DP, as judged by the size of the birefringent grains (Figure 2; a typical crystalline texture for 8-armed P(PS-co-ES)). This behavior is attributed to a molecular weight effect: at low DP (i.e., DP = 10 or DP = 20), the barrier to primary nucleation will be relatively high, as chain mobility induces attachment and detachment of the chains at the nucleation front until a critical nuclei size is eventually produced,23,39,40 thus higher DP values increase the nucleation rate (at constant Tc as in Figure 2). For the other samples, a high nucleation rate was also observed which prevented the growth of superstructures (see Figures S2 and S3). 3.3. Nonisothermal DSC. Thermal analysis of the copolymers described thus far (i.e., linear obtained by both protocols of addition, stars and combs with different DP) was carried out by DSC. The results for linear and star copolymers can be seen, respectively, in Figures 3 and 4, whereas the DSC scans results for comb copolymers, which display a similar trend, are shown in Figure S4 (see the Supporting

Figure 3. DSC cooling (A) and subsequent heating (B) scans obtained for La and Lb copolymers with different DPs (i.e., 10, 20, and 30). Please refer to Table 1 for their full nomenclature.

Figure 4. DSC (A) cooling and (B) heating scans for La and star (S) copolymers with different number of arms (i.e., 2, 4, and 8) and variable DP (i.e., DP of 10, 20, and 30). Please refer to Table 1 for their full nomenclature.

Information). The data obtained from the DSC traces of all the samples are summarized in Table 2. 3.3.1. Longer vs Shorter Gradients (Protocol a vs Protocol b). Cooling scans after erasing thermal history and subsequent heating scans were measured with DSC (Figure 3) with most samples exhibiting a bimodal crystallization and melting. Even though reorganization during the heating scan cannot be totally ruled out, DSC experiments performed at different cooling rates and simultaneous in situ SAXS/WAXS experiE


Article

Macromolecules Table 2. Relevant Calorimetric Quantities Obtained From DSC Scans in Figures 3, 4, and S5a,b,c,d sample

Tc,onset (°C)

Tc,peak (°C)

ΔHnc (J/g)

Tm,peak (°C)

Tm,end (°C)

ΔHnm (J/g)

La_(PS5-ES5)2 La_(PS10-ES10)2 Lb_(PS10-ES10)2 La_(PS15-ES15)2 Lb_(PS15-ES15)2 S_(PS5-ES5)4 S_(PS10-ES10)4 S_(PS15-ES15)4 S_(PS5-ES5)8 S_(PS10-ES10)8 S_(PS15-ES15)8 C_(PS15-ES15)10 C_(PS5-ES5)15 C_(PS10-ES10)15 C_(PS15-ES15)15 C_(PS15-ES15)20

67.4 78.5 49.5 82.2 53.0 42.5 65.3 69.1 22.4 48.1 50.4

(12.3) 42.2 (38.8 )59.3 23.6 (45.5) 66.0 31.8 (−3.3) 19.5 (18.5) 48.2 (25.1) 53.3 4.3 33.2 32.7

88 96 28 88 38 54 82 82 24 38 44

30.0 (98.0) 35.7 (100.4) 4.7 (50.7) 37.4 (98.4) 8.6 (57.2) 10.2 (44.4) 36.7 (70.1) 44.1 (76.8) 3.85 (47.8) 27.1 (60.7) 26.2 (60.4) 31.7d 31.9 23.7 (47.9) 47.8 42.5

112.9 115.9 83.8 117.2 85.0 76.1 94.7 102.4 61.8 79.8 82.9 60.6 87.0 85.5 99.2 95.1

104 98 34 106 44 68 92 96 30 38 44 4 28 28 42 34

55.2 46.9 65.3 64.5

25.7 24.2 18.4 18.1

24 24 42 32

Xc,WAXS (%)

34 13

30 15 24 28

16 20 19

a

Onset (Tc,onset) and peak (Tc,peak) crystallization temperatures, peak (Tm,peak) and end (Tm,end) melting temperatures, crystallization (ΔHnc) and melting (ΔHnm) normalized enthalpies, calculated mass of crystals (Xc,WAXS) from WAXS patterns. bOne-shot monomer addition or protocol (a). c Sequential monomer addition or protocol (b). dBefore the melting, a cold crystallization occurs at Tcc,onset = −2.9 °C, Tcc,peak = 6.3 °C, Tcc,end = 18.4 °C and a normalized cold crystallization enthalpy of 3 J/g.

In general, both crystallization and melting occurred in a wide range of temperatures with bimodal exothermal and endothermal transitions, respectively, while the heat flow values were also very small (see y-axis scale bars), indicating low levels of crystallinity (see later WAXS measurements). It was found that the onset crystallization temperature (Tc,onset), as well as the end melting temperature (Tm,end), could better characterize the changes experienced by the material in comparison to peak values, as a result of the presence of two or more peaks and the extreme temperature width of the exothermal and endothermal first order transitions. Similar behavior was found for comb copolymers (see Figure S4 in the Supporting Information). As shown in Figure 5, peak crystallization and melting temperatures appeared to show three regions in their

ments (results not shown) demonstrated that if reorganization during heating is present, neither DSC nor WAXS/SAXS techniques are able to clearly show it. This means that the bimodal distributions of both crystallization exotherms and melting endotherms are a result of the distribution of noncrystallizable units along the chains, a fact corroborated by SSA results shown below. The longer gradient copolymers (La in Figure 3) are characterized by a much broader crystallization and melting range than those obtained by the sequential addition method (Lb in Figure 3). The longer gradient copolymers contain on average longer crystallizable sequences which can form thicker lamellae that melt at higher temperatures. However, they also contain shorter crystallizable units which will crystallize and melt at lower temperatures. Therefore, the distribution of crystal sizes produced in the longer gradient copolymer is much broader than in those copolymers prepared with shorter gradients. In the case of shorter gradients, the average crystallizable sequence lengths are shorter. Therefore, the calorimetric differences observed between copolymers prepared by the two addition methods reflect the difference in monomer distribution throughout the copolymer chains. In P(PS-co-ES), an increase in the DP per arm resulted in an increase of both the crystallization (Tc) and melting (Tm) temperatures, as expected for samples of increasing molecular weight (Figure 3, and quantitatively summarized in Table 2). The effect of the molecular weight on first-order transition temperatures usually saturates as chain length reaches a critical value (DP 20−30) as the differences between their Tc and Tm values are negligible (Figure 3B). 3.3.2. Chain Topology and Molecular Weight Influence. Due to their higher crystallinity and higher Tc and Tm of the longer gradient copolymers, all branched macromolecules were prepared through protocol a, investigating the effect of DP and number of arms on DSC cooling scans (after erasing thermal history) and subsequent heating scans (performed at 20 °C/ min), as shown in Figures 4 and S4. The data calculated from the DSC scans (Figures 3, 4, and S4) are also summarized in Table 2.

Figure 5. (A) Crystallization onset temperatures (Tc,onset) and end melting temperatures (Tm,end) and (B) crystallization (ΔHnc) and melt (ΔHnm) normalized enthalpies as a function of number of arms, for linear, star, and comb samples with different DPs. Solid lines represent guides to the eye. Dashed lines separate the behavior of linear, star, and comb samples. Please refer to Table 1 for the full polymer nomenclature. F


Article

Macromolecules

undergo radial stretching in the melt (and show “excess” anisotropy).36 This situation could facilitate intramolecular crystallization, whereas the “convergent crowding” of the stars, whose arms are likely to be less tightly packed (higher values of

dependency on the number of arms. At any given DP, the linear polymers have higher values for thermal transitions and enthalpies, stars decreased their values with increasing number of arms, and combs in most cases reached a plateau or even increased. Such behavior is attributed to the topological restrictions induced by the increasing crowding, which hinders crystallization. Another general trend that can be appreciated in Figure 5 is that as DP increases there are small increases in melting temperature and even in crystallinity, especially for samples with a higher number of arms. A more detailed analysis of each specific case is given below. 3.3.2.1. Linear Copolymers. Since linear copolymers are free of topological restrictions, their size is the main factor affecting crystallization. However, an increase in DP only modestly changed their thermal properties, since their values tend to saturate at DP = 20 (almost identical values at DP 20 and 30 were obtained). It is also worth noting that while the ES gradient structure of the copolymer chains with DP 10 is identical to that of DP 20 (both are prepared with protocol a, La), Tc,onset and Tm,onset are lower for the copolymer with lower molecular weight (i.e., DP 10). This effect is probably due to differences in chain length, or to the fact that the ES sequences in the center of the copolymer are less likely to enter the crystals. 3.3.2.2. Star Copolymers. Crowding around branching points typically limits the conformational freedom of polymer chains, reducing their flexibility and diffusion,41 and hence decrease their ability to crystallize in comparison to linear analogues. This effect should decrease with increasing length of the arms, therefore higher DP would correspond to higher crystallinity. Tc,onset and Tm,end, as well as crystallization and melting enthalpies increased as DP increased in the 4 and 8 arm stars. This is consistent with the nucleation observations presented above (see Figure 2) and shows that the topological restrictions of the central branching point are at least partially compensated when the ES homosequences extend further from it. In the case of the lowest DP (i.e., 10), the topological restrictions for the stars to crystallize are the strongest and increase as the number of arms increases, as judged by the dramatic decrease in crystallization and melting temperatures (as well as associated enthalpies) that can be observed in Figure 5. 3.3.2.3. Comb Copolymers. The combs with 15 and 20 arms behaved differently from star copolymers (see comb region on Figure 5): the values of Tc,onset and Tm,end increased (instead of decreasing) as the number of arms increased, and the crystallization and melting enthalpies stopped decreasing; this would appear to indicate a higher crystallizability of combs vs stars. As for the star polymers, also for the combs, crystallization can only take place in the side chains: their main chain is atactic. Therefore, any comb/star difference should be ascribed to the different packing density of the side chains. In the melt, flexible polymer chains are typically assumed to adopt a conformation analogous to a theta state; however, chains that are already contracted/globular in solution, are likely to maintain this state as also demonstrated by the almost globular structure of the combs (Mark−Houwink parameter a < 0.5; see Table 1). It could be argued that the side chains in the combs experience a sort of “parallel crowding” (= high packing density; please refer also to Scheme 1), not dissimilarly to chains of hyperbranched polystyrene that are supposed to

Rg RH

and a; Figure 1B), may only rely on the increasingly difficult

intermolecular crystallization; this would rationalize why combs with 15 and 20 arms (high coil compaction already in solution: parameter a = 0.42 or 0.41, respectively) are significantly more crystalline than the corresponding stars. The behavior of combs with DP = 10 is different, and dominated by the problems of crystallization of these short sequences. DSC cooling scans provide clear evidence of their retarded crystallization (see Supporting Information, Figure S4A): the ES sequences within the comb copolymer with 15 arms of DP = 10 are not able to crystallize during cooling and they undergo cold-crystallization during the subsequent heating scan followed by melting (see Supporting Information, Figure S4B). In summary, using short chains with identical composition and primary structure, their crystallinity very strongly decreases with increasing branching, and this does not appear to be due to the more compact morphology of combs. This can be an important indication of: sensitivity to the actual length of the ES sequences and could indicate that comb crystallization possibly proceeds in a predominantly intramolecular fashion. 3.4. Successive Self-Nucleation and Annealing (SSA). The thermal fractionation by SSA yields multiple endothermic peaks that correspond to the melting of the different fractions produced by the applied thermal protocol. Each peak represents the melting of a population of lamellae with specific mean lamellar thickness, i.e., a thermal fraction. Typically, the average lamellar thickness values increase as Tm increases and the largest lamellar thickness corresponds to the longest uninterrupted ES sequence. Melting peaks are identified with numbers: the highest melting temperature (melting peak 1) corresponds to thermal fraction 1, i.e., the annealed population produced mainly during a 5 min holding time at Ts,1, although the successive steps might have also some limited influence on the size of the fraction. Thermal fraction 2 is produced at Ts,2 and so on. The frequent interruption of ES sequences facilitates their segregation, and this caused excellent fractionation in all samples, as shown by a peak separation where the DSC baseline is typically reached. This clear separation is essentially the deconvolution of a standard DSC curve into thermal fractions, which can therefore provide much more detailed information than standard DSC measurements. 3.4.1. Linear Copolymers. In Figure 6, solid lines indicate the Ts values employed for the fractionation, while the dashed line corresponds to the Ts,ideal for the linear a copolymer (La), which is also the first Ts temperature employed for all samples (i.e., linear, stars, and combs). SSA fractionation profiles are a necessary reflection of the chain primary structure,24−26,30−32 due to the impossibility for PS to crystallize, and indeed a clear difference between protocols a and b can be seen. The copolymers with longer ES sequences possess the highest melting point fractions, (fractions 1−3), whereas the melting range of those with shorter ES sequences shifted to lower temperatures, and Ts,1− Ts,3 did not produce any detectable thermal fractions. The influence of DP is small for both linear copolymers (La) and (Lb). Even though as DP increases, chain length also G


Article

Macromolecules

Figure 6. Final heating runs after SSA thermal fractionation for P(PSco-ES) linear copolymers obtained by protocols a (La) and b (Lb), with DP = 10, 20, and 30. Please refer to Table 1 for their full nomenclature.

Figure 7. Final heating runs after SSA thermal fractionation for P(PSco-ES) copolymers with La and S topologies and variable DP (i.e., DP = 10, 20 and 30). Please refer to Table 1 for their full nomenclature.

increases, there are only small changes in the fractionation capacity. The La and Lb copolymers with the highest and the lowest DP show the same number of thermal fractions; however, as DP increases, the areas under the highest temperature fractions increase. This result indicates that as DP increases, the longest uninterrupted linear ES sequences that constitute the lamellar crystals melt at the highest temperature peaks (i.e., fractions 1, 2, and 3) are more abundant. It is worth noting that the maximum melting peak (corresponding to the melting of fraction 1) after SSA treatment is around 114 °C, which is much lower in comparison with the reported value for the equilibrium melting point of ES homopolymer (i.e., 215.6 °C).16 This is due to the short length of the crystallizable sequences of ES, which on the other hand is also responsible for the remarkable fractionation of the samples. 3.4.2. Star Copolymers. Both the star and linear polymers presented in Figure 7 in principle have the same primary structure (longer gradients, protocol a), but fraction 1 was only present in the latter. If a comparison is made of the SSA fractionation profile at constant DP, it can be clearly seen that the topological restrictions, which increase with arm number, provoke the gradual disappearance of several of the highest melting point thermal fractions. These topological restrictions should be maximum near the junction points (at the center of the stars), where in fact the longest uninterrupted PES linear sequences are located; hence as the number of arms increases, confinement increases and these segments cannot crystallize. On the other hand, upon increasing arm length (i.e., DP), the topological constraints are gradually released, and Figure 7 shows how some higher melting point fractions are indeed recovered in the stars. As fractionation profiles suffer important changes due to variations in chain topology (linear chains versus star chains), the interpretation of these profiles in terms of only comonomer distribution is no longer possible. 3.4.3. Comb Copolymers. In this analysis, we have varied both the number of arms (10, 15, and 20) and the DP (10, 20, and 30), again keeping fixed the primary structure and using a linear polymer as a reference (Figure 8).

Figure 8. Final heating runs after SSA thermal fractionation for P(PSco-ES) La and combs copolymers with different number of arms (i.e., 10, 15, and 20) at a fixed DP of 30 and variable DP (i.e., DP = 10, 20 and 30) for the 15 arms sample. Please refer to Table 1 for their full nomenclature.

First, the trend toward increasing crystallinity with increasing DP is similar to what is observed for stars: in this respect, the behavior of the 15-armed comb is analogous to that of, e.g., 8-armed star. With low DP (DP 10) arms, combs also continued the trend of decreasing crystallinity with increasing branching: the first detectable fraction is that of peak 1 for the linear reference, peaks 5 and 6, respectively for the 4- and 8-armed stars, and only peak 7 for the 15-armed comb. A significant difference, on the other hand, is seen at high DP (DP 30): the first detectable fraction is peak 1 for the linear polymer, peaks 2 and 4 respectively for the 4- and 8armed stars, peak 4, and peak 3 (slight recovery of crystallinity), for the 10-, 15-, and 20-armed combs. Here, the remarkable resolution of SSA has confirmed the decreasing crystallinity with decreasing ES gradient and increased branching. For side chains with very short ES gradients (low DP), branching has a monotonous and detrimental effect on the presence of thicker lamellar crystals; for longer ES gradients, on the contrary, the parallel crowding of combs allow the chains to regain some of the order. These H


Article

Macromolecules

at low conversion, and therefore differing from ours in the gradient structure), Roggero et al.44 reported an increase of the crystalline PES unit cell volume to accommodate the PS methyl groups in the crystal lattice, albeit with no X-ray evidence supporting this conclusion. In our case, the diffraction spacings were not significantly different among all different copolymers examined to draw the same conclusion (Table 3). Differently from their location, the reflection intensity depended both on topology and DP. For instance, the intensity of both reflections increased with DP (Figure 9A), indicating that the crystal fraction is increasing. The (101) reflection in the sample with DP = 30 is the most intense of the three samples in Figure 9A, while in the sample with DP = 10 the shoulder is barely visible; similar changes were found for all samples (see Supporting Information, Figures S4 and S5). The software Peakfit was employed to deconvolve WAXS patterns into amorphous and crystalline contributions, obtaining the degree of crystallinity by dividing the area under a crystalline peak by the total area under the diffractogram (column 3 (Xc,WAXS), Table 3). This analysis broadly confirmed the calorimetric results: a) longer gradients yield larger crystalline fractions than shorter gradients, both when they were obtained varying primary structure or DP; b) decreasing crystallinity with increasing branching, with a slight gain with longer combs. 3.6. Small Angle X-ray Scattering (SAXS). SAXS patterns were measured at 25 °C for all the samples (previously cooled from the melt at 20 °C/min to reproduce the same thermal history as that applied by DSC). The Lorentz representation was chosen to analyze the SAXS results, by plotting the product of the intensity of diffracted X-rays and the square of the scattering vector q as a function of q. A clear maximum that represent the scattering from lamellar stacks was observed in the SAXS patterns of all samples (see Figure 9B for DP 30). This is associated with the average distance between adjacent lamellae, or long period d* (column 3 in 2π Table 4), which can be estimated as d* = q (peaks in Figure

observations support an intramolecular nature of the crystallization of the branched materials. 3.5. Wide Angle X-ray Scattering (WAXS). WAXS patterns were measured at RT (25 °C) for all the samples (previously cooled from the melt at 20 °C/min to reproduce the same thermal history applied by DSC) with a fixed DP of 30 (see Figures S5 and S6 in Supporting Information). In WAXS patterns (Figure 9A, Table 3), it is easy to see the presence of a main peak and a shoulder/secondary peak for all

Figure 9. WAXS patterns for (A) 8-arm star P(PS-co-ES) copolymers with variable DP (10, 20, and 30: S_(PS5-ES5)8, S_(PS10-ES10)8, and S_(PS15-ES15)8), and Lorentz corrected SAXS patterns (B) for P(PSco-ES) copolymers with different chain topologies (linear, star, and comb) and fixed DP (30) at 25 °C.

Table 3. Calculated Diffraction Spacings (d) According to Bragg’s Law, 2θ Angles, and Calculated Mass Fraction of Crystals (Xc,WAXS) from WAXS Patterns sample

2θ (deg)

d (nm)

La_(PS15-ES15)2 Lb_(PS15-ES15)2 S_(PS15-ES15)4 S_(PS5-ES5)8 S_(PS10-ES10)8 S_(PS15-ES15)8 C_(PS15-ES15)10 C_(PS15-ES15)15 C_(PS15-ES15)20

20.2/24.0 20.3/23.8 20.0/23.6 19.5/23.5 20.2/23.9 20.1/24.2 20.0/23.4 20.3/24.2 20.3/24.1

0.439/0.370 0.437/0.373 0.443/0.377 0.454/0.379 0.439/0.372 0.441/0.367 0.443/0.380 0.437/0.367 0.437/0.369

Xc,WAXS (%) 34 13 30 15 24 28 16 20 19

± ± ± ± ± ± ± ± ±

5.1 2.0 4.5 2.3 3.6 4.2 2.4 3.0 2.9

max

Table 4. Long Period (d*) Values Obtained at RT for PPSco-PES Copolymers with Different Topologies and a Fixed DP (30) and Calculated from the qmaxa

samples, which indicates that crystals have the same unit cell regardless of the chain topology and the DP of the arms. It is noteworthy that the shape of these diffraction maxima does not change significantly with temperature until melting occurs (temperature-dependent WAXS and SAXS patterns are included in the Supporting Information, see Figures S6−S9 and Table S1). The reflections appear at 2θ angles of 20 and 24° (diffraction spacings d = λ/2sin θ, of 0.439 and 0.370 nm, respectively); they would therefore correspond to the (100) and (101) planes of the orthorhombic PES unit cell (a = 8.50 Å, b = 4.95 Å, and c (fiber axis) = 6.70 Å),42 which is reasonable since ES is the only crystallizable component. It must be said, however, that only few works have applied X-ray diffraction to PES homopolymer,42 and even its structure was determined by electron diffraction.42,43 To our knowledge, in the only report about P(ES-co-PS) copolymers (but prepared

sample

qmax (A−1)

d* (nm)

La_PS15-ES15 Lb_PS15-ES15 S4_PS15-ES15 S8_PS15-ES15 C10_P15-ES15 C15_P15-ES15 C20_P15-ES15

0.05512 0.04821 0.06821 0.07609 0.06840 0.06485 0.06485

11.41 12.31 9.21 8.26 9.19 9.69 9.69

Xc,WAXS (%)

l (nm)

± ± ± ± ± ± ±

3.84 1.56 2.75 2.31 1.51 1.89 1.81

34 13 30 28 16 20 19

5.1 2.0 4.5 2.0 2.4 3.0 2.9

a

The percentages of crystallinity calculated by WAXS (Xc,WAXS), as well as the lamellar thickness estimated from it (l), are also reported.

9B are labeled according to their d* values). Since the peak was generally well pronounced, a fair amount of lamellar stacking must have always been present, also for lowcrystallinity samples. We have also calculated a probably more informative parameter: the lamellar thickness l, l = Xvd*, where Xv is the crystalline volume fraction. Here, since accurate density measurements would be needed to obtain the crystalline volume fraction, we have replaced it with the crystalline mass I


Article

Macromolecules

possibly to the build-up (in combs) of some forms of intramolecular order.

fraction determined from WAXS (third and fourth column in Table 4). In short, d* and l followed qualitatively the same pattern previously seen for DSC results, especially SSA, and for the crystal fractions calculated from WAXS. All decreased with increasing branching in stars and then slightly increased for longer combs, and they also further confirmed the loss of order in the linear polymers prepared via repeated monomer addition (method b). 3.7. General Comparison. We have graphically summarized (Figure 10) the influence of branching on the various

4. CONCLUSIONS The degree of crystallinity of a P(PS-co-ES) copolymer containing both a crystallizable (ES) and noncrystallizable monomer (PS) strongly depends on (1) the primary structure (length of ES:PS gradients) and on (2) the presence of topological restrictions introduced by branching. SSA generated thermal fractions based on molecular segregation of the crystallizable sequences (due to the interruptions of the non-crystallizable monomer units) allow us to probe polymers with a broad melting temperature in significantly greater depth than standard DSC alone. This confirmed that polymers with longer ES sequences/gradients (addition protocol a) had higher temperature thermal fractions (reflecting their much broader melting range) and were confirmed to have a higher crystalline mass fraction by WAXS. Chain topology (i.e., branching) on the other hand conferred a strong steric barrier on the capacity of ES sequences to crystallize; in general, increasing branching resulted in a reduced capacity to crystallize, with crystallization and melting temperatures, relevant enthalpies, annealing capacity and lamellar thickness all strongly decreasing. It is intriguing that this trend mirrors that of the polymer coil R compactness (Mark−Houwink parameter a and g ratio), RH

which indicates the decreasing availability of the chain for topological interactions (such as entanglements). Stars and combs showed clear differences. For example, in combs, the key melting/crystallization temperatures/enthalpies did not further decrease with increasing number of arms. This is probably due to combs showing a chain crowding that allows for easier, although possibly intramolecular, packing (“parallel” vs convergent) or the topological constrains for crystallization being released as the number of arms (or chain length) surpasses a saturation value. Here we have therefore demonstrated that branching allows for fine-tuning of the local order in polysulfides, which also likely includes a balance between intra- and intermolecular assembly. This can pave the way to detailed control over, e.g., solubility and diffusion of low MW compounds (drugs) in these matrices, with clear applications in drug delivery (loading and release kinetics).

Figure 10. Summary of key parameters measured/calculated in this study on polymers with DP 30 per arm and a 1:1 PS:ES ratio. All parameters are normalized against the values presented by the linear (formally 2-armed) polymer obtained with method a. The data for the linear polymer obtained via repeated monomer addition (method b) are presented in red. The Mark−Houwink parameter a and the shape parameter

Rg RH

are reported in blue to distinguish these dilute solution

(hence single molecule) parameters from those referring to bulk materials, which are represented by black empty symbols.

parameters investigated in this study, namely, ΔHc, ΔHm, Xc, d, l and the temperature of the highest peak in SSA. These parameters offer different perspectives over the capacity of these polymers to assemble in an ordered fashion; this is confirmed by the very low values of most parameters (except d) for the linear polymers with irregular primary structure (obtained through method b). All parameters coherently show a loss of order with increasing branching for stars, and a stabilization/slight regain for combs. Interestingly, a similar branching-dependent trend is also seen in shape/compactness parameters measured in solution, i.e., the Mark−Houwink a parameter and the shape parameter Rg ; this trend indicates that the macromolecules grow

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.8b02659.

RH

■

increasingly compact and are unlikely to form entanglements in solution, to the limit of behaving as impenetrable spheres. Clearly, these parameters reflect dimension and conformation of the polymers in dilute solution, which, differently from the bulk, are typically dominated by excluded volume effects. Yet, it seems logical to see a parallel in that the same cause (branching) in solution overcomes the excluded volumedependent swelling, determining the coil collapse into increasingly compact globular structures, and in the bulk leads to the progressive loss of intermolecular order and

Materials synthesis, morphology, additional DSC scans, self-nucleation, and X-ray results (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Arantxa Arbe: 0000-0002-5137-4649 Nicola Tirelli: 0000-0002-4879-3949 Alejandro J. Müller: 0000-0001-7009-7715 J


Article

Macromolecules Notes

(14) Nicol, E.; Nicolai, T.; Durand, D. Dynamics of Poly(propylene sulfide) Studied by Dynamic Mechanical Measurements and Dielectric Spectroscopy. Macromolecules 1999, 32 (22), 7530−7536. (15) Catsiff, E. H.; Gillis, M. N.; Gobran, R. H. Poly(ethylene sulfide). II. Thermal degradation and stabilization. J. Polym. Sci., Part A-1: Polym. Chem. 1971, 9 (5), 1271−1292. (16) de Chirico, A.; Zotteri, L. Crystallization and glass transitions of polyethylene sulphide and polyisobutylene sulphide. Eur. Polym. J. 1975, 11 (7), 487−490. (17) Cooper, W.; Hale, P. T.; Walker, J. S. Elastomeric block polymers from ethylene sulphide. Polymer 1974, 15 (3), 175−186. (18) Sevgen, E.; Dolejsi, M.; Nealey, P. F.; Hubbell, J. A.; de Pablo, J. J. Nanocrystalline Oligo(ethylene sulfide)-b-poly(ethylene glycol) Micelles: Structure and Stability. Macromolecules 2018, 51 (23), 9538−9546. (19) Brubaker, C. E.; Velluto, D.; Demurtas, D.; Phelps, E. A.; Hubbell, J. A. Crystalline Oligo(ethylene sulfide) Domains Define Highly Stable Supramolecular Block Copolymer Assemblies. ACS Nano 2015, 9 (7), 6872−6881. (20) Hu, W.; Mathot, V. B. F.; Alamo, R. G.; Gao, H.; Chen, X. Crystallization of Statistical Copolymers. In Polymer Crystallization I: From Chain Microstructure to Processing; Auriemma, F., Alfonso, G. C., de Rosa, C., Eds.; Springer International Publishing: Cham, 2017; pp 1−43. (21) Auriemma, F.; De Rosa, C.; Di Girolamo, R.; Malafronte, A.; Scoti, M.; Cioce, C. Molecular View of Properties of Random Copolymers of Isotactic Polypropylene. In Polymer Crystallization I: From Chain Microstructure to Processing, Auriemma, F.; Alfonso, G. C.; de Rosa, C., Eds. Springer International Publishing: Cham, 2017; pp 45−92. (22) Mandelkern, L. Crystallization of Polymers: Vol. 1, Equilibrium Concepts; Cambridge University Press: 2002. (23) Mandelkern, L. Crystallization of Polymers: Vol. 2, Kinetics and Mechanisms; Cambridge University Press: 2004. (24) Müller, A. J.; Hernández, Z. H.; Arnal, M. L.; Sánchez, J. J. Successive self-nucleation/annealing (SSA): A novel technique to study molecular segregation during crystallization. Polym. Bull. 1997, 39 (4), 465−472. (25) Müller, A. J.; Arnal, M. L. Thermal fractionation of polymers. Prog. Polym. Sci. 2005, 30 (5), 559−603. (26) Müller, A. J.; Michell, R. M.; Pérez, R. A.; Lorenzo, A. T. Successive Self-nucleation and Annealing (SSA): Correct design of thermal protocol and applications. Eur. Polym. J. 2015, 65, 132−154. (27) Fillon, B.; Lotz, B.; Thierry, A.; Wittmann, J. C. Self-nucleation and enhanced nucleation of polymers. Definition of a convenient calorimetric “efficiency scale” and evaluation of nucleating additives in isotactic polypropylene (α phase). J. Polym. Sci., Part B: Polym. Phys. 1993, 31 (10), 1395−1405. (28) Lorenzo, A. T.; Arnal, M. L.; Sánchez, J. J.; Müller, A. J. Effect of annealing time on the self-nucleation behavior of semicrystalline polymers. J. Polym. Sci., Part B: Polym. Phys. 2006, 44 (12), 1738− 1750. (29) Blundell, D. J.; Keller, A.; Kovacs, A. J. A new self-nucleation phenomenon and its application to the growing of polymer crystals from solution. J. Polym. Sci., Part B: Polym. Lett. 1966, 4 (7), 481− 486. (30) Lorenzo, A. T.; Arnal, M. L.; Müller, A. J.; De Fierro, A. B.; Abetz, V. High speed SSA thermal fractionation and limitations to the determination of lamellar sizes and their distributions. Macromol. Chem. Phys. 2006, 207 (1), 39−49. (31) Müller, A. J.; Lorenzo, A. T.; Arnal, M. L. Recent Advances and Applications of “Successive Self-Nucleation and Annealing” (SSA) High Speed Thermal Fractionation. Macromol. Symp. 2009, 277 (1), 207−214. (32) Lorenzo, A. T.; Arnal, M. L.; Müller, A. J.; Lin, M. C.; Chen, H. L. SAXS/DSC analysis of the lamellar thickness distribution on a SSA thermally fractionated model polyethylene. Macromol. Chem. Phys. 2011, 212 (18), 2009−2016.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The POLYMAT/UPV/EHU team would like to acknowledge funding from the following projects: UPV/EHU Infrastructure: INF 14/38; Mineco/FEDER: SINF 130I001726XV1/Ref: UNPV13-4E-1726; and Mineco MAT2017-83014-C2-1-P*. R.A.P.-C. gratefully acknowledges the award of a Ph.D. fellowship by POLYMAT Basque Center for Macromolecular Design and Engineering. A.A. and A.I. acknowledge financial support from IT-654-13 and MAT2015-63704-P (MINECO/ FEDER, UE). We gratefully acknowledge the ALBA synchrotron facility (Proposal Number: 2018022683) for the funding and help to perform experiments at BL11-NCDSWEET beamline with the collaboration of ALBA staff.

■

REFERENCES

(1) Lee, S. H.; Gupta, M. K.; Bang, J. B.; Bae, H.; Sung, H.-J. Current Progress in Reactive Oxygen Species (ROS)-Responsive Materials for Biomedical Applications. Adv. Healthcare Mater. 2013, 2 (6), 908−915. (2) Vo, C. D.; Kilcher, G.; Tirelli, N. Polymers and Sulfur: what are Organic Polysulfides Good For? Preparative Strategies and Biological Applications. Macromol. Rapid Commun. 2009, 30 (4−5), 299−315. (3) Lallana, E.; Tirelli, N. Oxidation-Responsive Polymers: Which Groups to Use, How to Make Them, What to Expect From Them (Biomedical Applications). Macromol. Chem. Phys. 2013, 214 (2), 143−158. (4) d’Arcy, R.; Tirelli, N. Fishing for fire: strategies for biological targeting and criteria for material design in anti-inflammatory therapies. Polym. Adv. Technol. 2014, 25 (5), 478−498. (5) Napoli, A.; Valentini, M.; Tirelli, N.; Muller, M.; Hubbell, J. A. Oxidation-responsive polymeric vesicles. Nat. Mater. 2004, 3 (3), 183−189. (6) Rehor, A.; Schmoekel, H.; Tirelli, N.; Hubbell, J. A. Functionalization of polysulfide nanoparticles and their performance as circulating carriers. Biomaterials 2008, 29 (12), 1958−1966. (7) Missirlis, D.; Kawamura, R.; Tirelli, N.; Hubbell, J. A. Doxorubicin encapsulation and diffusional release from stable, polymeric, hydrogel nanoparticles. Eur. J. Pharm. Sci. 2006, 29 (2), 120−129. (8) Rehor, A.; Botterhuis, N. E.; Hubbell, J. A.; Sommerdijk, N. A. J. M.; Tirelli, N. Glucose sensitivity through oxidation responsiveness. An example of cascade-responsive nano-sensors. J. Mater. Chem. 2005, 15 (37), 4006−4009. (9) Hu, P.; Tirelli, N. Scavenging ROS: Superoxide Dismutase/ Catalase Mimetics by the Use of an Oxidation-Sensitive Nanocarrier/ Enzyme Conjugate. Bioconjugate Chem. 2012, 23 (3), 438−449. (10) Jeanmaire, D.; Laliturai, J.; Almalik, A.; Carampin, P.; d'Arcy, R.; Lallana, E.; Evans, R.; Winpenny, R. E. P.; Tirelli, N. Chemical specificity in REDOX-responsive materials: the diverse effects of different Reactive Oxygen Species (ROS) on polysulfide nanoparticles. Polym. Chem. 2014, 5 (4), 1393−1404. (11) Carampin, P.; Lallana, E.; Laliturai, J.; Carroccio, S. C.; Puglisi, C.; Tirelli, N. Oxidant-Dependent REDOX Responsiveness of Polysulfides. Macromol. Chem. Phys. 2012, 213 (19), 2052−2061. (12) d’Arcy, R.; Siani, A.; Lallana, E.; Tirelli, N. Influence of Primary Structure on Responsiveness. Oxidative, Thermal, and ThermoOxidative Responses in Polysulfides. Macromolecules 2015, 48 (22), 8108−8120. (13) d’Arcy, R.; Gennari, A.; Donno, R.; Tirelli, N. Linear, Star, and Comb Oxidation-Responsive Polymers: Effect of Branching Degree and Topology on Aggregation and Responsiveness. Macromol. Rapid Commun. 2016, 37 (23), 1918−1925. K


Article

Macromolecules (33) Bohdanecky, M. New method for estimating the parameters of the wormlike chain model from the intrinsic viscosity of stiff-chain polymers. Macromolecules 1983, 16 (9), 1483−1492. (34) Kök, M. S.; Abdelhameed, A. S.; Ang, S.; Morris, G. A.; Harding, S. E. A novel global hydrodynamic analysis of the molecular flexibility of the dietary fibre polysaccharide konjac glucomannan. Food Hydrocolloids 2009, 23 (7), 1910−1917. (35) Seo, K. S.; Caflisch, G. B. Molecular dimensions and melt rheology of an all-aromatic liquid crystalline polymer. J. Polym. Sci., Part B: Polym. Phys. 2001, 39 (20), 2378−2389. (36) Kharchenko, S. B.; Kannan, R. M. Role of Architecture on the Conformation, Rheology, and Orientation Behavior of Linear, Star, and Hyperbranched Polymer Melts. 2. Linear Viscoelasticity and Flow Birefringence. Macromolecules 2003, 36 (2), 407−415. (37) Chremos, A.; Jeong, C.; Douglas, J. F. Influence of polymer architectures on diffusion in unentangled polymer melts. Soft Matter 2017, 13 (34), 5778−5784. (38) Mays, J. W.; Nan, S.; Lewis, M. E. Hydrodynamic and thermodynamic properties of poly(α-methylstyrene) in dilute n-butyl chloride solution. Macromolecules 1991, 24 (17), 4857−4860. (39) Fatou, J. G.; Marco, C.; Mandelkern, L. The crystallization kinetics of low-molecular-weight polyethylene fractions. Polymer 1990, 31 (5), 890−898. (40) Ergoz, E.; Fatou, J. G.; Mandelkern, L. Molecular Weight Dependence of the Crystallization Kinetics of Linear Polyethylene. I. Experimental Results. Macromolecules 1972, 5 (2), 147−157. (41) Tezuka, Y. Topological Polymer Chemistry: Progress of Cyclic Polymers in Syntheses, Properties, and Functions; World Scientific: 2013. (42) Takahashi, Y.; Tadokoro, H.; Chatani, Y. Structure of polyethylene sulfide. J. Macromol. Sci., Part B: Phys. 1968, 2 (2), 361−367. (43) Moss, B.; Dorset, D. L. Poly(ethylene sulfide): Reevaluation of the electron diffraction structure analysis. J. Macromol. Sci., Part B: Phys. 1983, 22 (1), 69−77. (44) Roggero, A.; Zotteri, L.; Proni, A.; Gandini, A.; Mazzei, A. Syntheses and properties of ethylene sulphide-isobutylene sulphide copolymers and their unsaturated terpolymers. Eur. Polym. J. 1976, 12 (12), 837−842.

L


Influence of Chain Primary Structure and Topology (Branching) on

Recommend Documents