Design of Graft Block Polymer Thermoplastics - Macromolecules (ACS

Dec 1, 2016 - For a more comprehensive list of citations to this article, users are ... Domain Bridging in Thermoplastic Elastomers of Star Block Copo...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/Macromolecules

Design of Graft Block Polymer Thermoplastics Jiuyang Zhang,†,‡ Deborah K. Schneiderman,§ Tuoqi Li,‡ Marc A. Hillmyer,§ and Frank S. Bates*,‡ †

School of Chemistry and Chemical Engineering, Southeast University, Nanjing 211189, China Department of Chemical Engineering and Materials Science and §Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455-0431, United States



S Supporting Information *

ABSTRACT: Graft block polymers are defined by several architectural parameters, including backbone flexibility, graft density, backbone length, side-chain composition, and sidechain length. In this work we probe the impacts of each of these parameters on the phase behavior, rheological properties, and mechanical performance of these materials. Specifically, we examine two sets of materials prepared from backbones of different inherent flexibility. One set was prepared from poly[(n-butyl acrylate)-co-(2-hydroxyethyl acrylate)] (BxEy) copolymers; the other was prepared from hydroxypropyl methyl cellulose (HPMC) samples. Sequential ring-opening transesterification polymerization from these hydroxyl-functionalized macroinitiatiors yielded a diblock graft architecture containing a rubbery interior block and semicrystalline exterior blocks tethered to a flexible (BxEy) or rigid (HPMC) backbone. Good control over side-chain molar mass and composition and judicious choice of the graft block segments enabled the preparation of materials that were either ordered or disordered in the melt state. In the former case, crystallization destroys existing order in the material; in the latter case crystallization induces new microphase separation in the bulk. Many of the structure−mechanical property relationships observed for graft block copolymers with rigid backbones are maintained for graft block polymers with semiflexible backbones, including the tendency for samples to remain transparent when stretched. However, interestingly, the effects of graft density and backbone length are quite different depending on the rigidity of the backbone.



INTRODUCTION Thermoplastics currently comprise the majority (∼2/3) of the synthetic polymer industry. These versatile, melt-processable materials are used in a wide range of commercial products including packaging, medical devices, textiles, and elastomers.1−3 While low cost commodity homopolymers (e.g., PE, PP, PVC, and PS) dominate the thermoplastics market by volume, for high performance applications, higher value engineering plastics are required. This class of thermoplastics includes high impact polystyrene (HIPS), polyurethanes (PUR), acrylonitrile−butadiene−styrene (ABS) terpolymers, and styrenic block polymers (Kratons). In each of these commercial examples, mechanical performance is intimately linked to the underlying graft or block copolymer architecture. Linear block polymers, for example, the ABA triblock copolymers which comprise Kraton, achieve desired mechanical properties by anchoring a rubbery midblock between the two hard (semicrystalline or glassy) end blocks. The thermodynamic incompatibility of the disparate blocks drives microphase separation to form nanoscopic structures, the morphology of which is largely determined by the volume fraction of the respective components.4,5 Analogous linear multiblock copolymers can afford impressive toughness by bridging individual microdomains that further facilitate stress transfer within the matrix.6,7 One inherent limitation for the commercial utilization of linear block copolymers arises from their poor melt processability. That is usually due to a high order−disorder © XXXX American Chemical Society

transition temperature and high melt viscosity of these typically high molar mass materials.8−10 Furthermore, interior segments in linear multiblock copolymers have limited mobility, which can thwart crystallization of confined crystallizable components.6 To address these issues, polymers with more complex architectures may be utilized.11−14 Branched polymers, such as star polymers or graft polymers, exhibit unique topological, rheological, and mechanical behaviors.12,15−17 Rubber toughened HIPS and ABS, for example, are synthesized by the free radical polymerization of styrene (and in the case of ABS, acrylonitrile) in the presence of polybutadiene. In both cases, the resulting graft structure is critical to the mechanical performance of the material. Moreover, compared with linear polymers, branched polymers exhibit lower melt viscosity at high shear rates, making these materials easily processed using high shear melt processing techniques.14,18 Controlled polymerization strategies have facilitated the synthesis of a number of interesting complex branched polymers for thermoplastic applications. Early work by the Hadjichristidis group utilized a hydrosilylation/anionic polymerization approach to synthesize tetrafuntional multigraft copolymers with rubbery polyisoprene backbones and wellReceived: September 17, 2016 Revised: November 5, 2016

A

DOI: 10.1021/acs.macromol.6b02033 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

and (2) those with rigid (lp ≈ 15 nm) HPMC backbones.25,28 Whereas an array of HPMC polymers, differing in molar mass and hydroxy functionality are commercially available, the BxEy polymers were synthesized using controlled radical polymerization, namely, activator regenerated by electron transfer atom transfer radical polymerization (ARGET ATRP).29 In both cases, we used a simple sequential ring-opening transesterification (ROTEP) strategy to grow polyester diblock, grafts from the hydroxy-functional backbones.29,30

defined glassy polystyrene side chains; notably, these branched materials exhibited improved mechanical performance relative to linear triblock polymers (Kraton and Stryoflex).19 Subsequent to that study, Matyjaszewski and co-workers prepared three-arm star poly(n-butyl acrylate)-b-polyacrylonitrile block copolymer thermoplastic elastomers using atom transfer radical polymerization (ATRP).20 More recently, Tang et al. and Guan et al. have utilized ATRP to synthesize thermoplastic graft polymer elastomers with rigid cellulosic backbones and soft side chains.21,22 These previous studies reveal enormous opportunities embedded in strategically designed branched structures. In particular, we are interested in exploring the potential of graft block copolymers as high performance thermoplastic materials due to their intrinsic chemical tunability. We recently investigated poly[HPMC-graf t-(PMVL-blockPLLA) graft block copolymer thermoplastics, derived from hydroxypropyl methylcellulose (HPMC), poly(β-methyl-δvalerolactone) (PMVL), and poly(L-lactide) (PLLA). In these materials, rubbery PMVL functions as a bridge between the semirigid HPMC backbone and the hard PLLA end blocks. We found that this specific arrangement facilitates PLLA crystallization, which induces microphase separation and physical cross-linking. By changing the backbone molar mass or side chain composition, we were able to tailor the mechanical performance of these thermoplastics and access either hard plastics or tough elastomers.18 The overall performance of a graft block copolymer is influenced by backbone flexibility and length, side-chain length and composition, and graft density.14,18,23,24 Optimization of these parameters to attain tough and melt-processable thermoplastic materials often presents challenges. Until now, a detailed map for designing thermoplastics from graft block copolymers still remains elusive. In this work we investigate the impacts of graft block architecture, including backbone length, graft density, sidechain length, and side-chain composition, on the phase behavior, rheological properties, and mechanical performance of graft bock copolymers. As shown in Figure 1, the graft block polymers studied can be broadly classified into two categories based on the persistence length (lp) of the graft polymer backbone: (1) those with flexible (lp ∼ 2 nm) poly[(n-butyl acrylate)-co-(2-hydroxyethyl acrylate)] (BxEy) backbones25−27



RESULTS AND DISCUSSION Synthesis of Graft Block Copolymer BxEy-g(a-b). BxEyg(a-b) graft block polymers were synthesized in three steps as shown in Scheme 1. The flexible backbone, BxEy, was first Scheme 1. Synthesis of Graft Block Polymers BxEy-g(a-b) with Flexible Backbonesa

In BxEy-g(a-b), “x” and “y” denote the number of repeat units of nBA (B) and HEA (E) in the backbone, respectively. “A” and “B” represent the molar masses of the first block (poly(ε-decalactone), PDL) and the second block (poly(L-lactide), PLLA) in the grafts, respectively. Samples were prepared with between 62 and 105 graft blocks per molecule. a

prepared by copolymerization of HEM and nBA via ARGETATRP.29 The feed ratio was altered to adjust the density of grafting sites (hydroxyl groups) along the backbone, while the molar mass was controlled by adjusting the ratio of nBA and the initiator (EBiB). In each case the conversion of each monomer (∼40% in most cases) was determined using proton nuclear magnetic resonance (1H NMR spectroscopy) of crude aliquots taken prior to purification of the polymers by precipitation. The compositions of the feed in the crude samples at these modest conversions were similar to the starting compositions, in agreement with the reported reactivity ratios (rnBA = 0.894 and rHEA = 1.505) of this monomer pair under similar conditions.31

Figure 1. Graft block polymers with flexible (BxEy) and rigid backbone (HPMC). Poly(ε-decalactone) (PDL) and poly(β-methyl-δ-valerolactone) (PMVL) are grafted from flexible and rigid backbones, respectively. These simplified illustrations to not represent the actual number of graft blocks per molecule, which range from approximately 40 to 105. B

DOI: 10.1021/acs.macromol.6b02033 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 1. Molecular Characterization of Flexible Graft Block Copolymersa samples

Mnb (kg mol−1)

Mnc (kg mol−1)

Mwc (kg mol−1)

Đc

B357E103 B434E62 B728E105 B357E103-g(13.5) B434E62-g(6.5) B434E62-g(14) B728E105-g(13.5) B357E103-g(13.5-18) B357E103-g(13.5-37) B434E62-g(6.5-9) B434E62-g(14-20) B434E62-g(14-34) B728E105-g(13.5-20) B728E105-g(13.5-40)f LDL (19-28-19)

58 63 105 1450 460 930 1520 3300 5300 1020 2170 3040 3620 5720 66

67 49 120 600 520 530 840 1430 1550 1170 1300 1530 1350

100 73 190 1070 860 940 1450 1810 1930 1650 1720 1960 1900

1.50 1.50 1.54 1.75 1.67 1.74 1.73 1.26 1.24 1.41 1.32 1.28 1.40

110

140

1.22

Mn,PDLd (kg mol−1)

Mn,PLLAd (kg mol−1)

f Le

13.5 6.5 14.0 13.5 13.5 14.0 6.5 14.0 14.0 13.5 13.5 28

18 37 9 20 34 20 40 19

0.51 0.68 0.52 0.54 0.67 0.54 0.70 0.56

In BxEy-g(a-b), “x” and “y” denote the number of repeat units of nBA (B) and HEA (E) in the backbone, respectively. “a” and “b” represent the molar masses of the first block poly(ε-decalactone) (PDL) and the second block poly(L-lactide) (PLLA) in the grafts, respectively. bTheoretical molar mass based on added ratio of HEA and nBA to EBiB and monomer conversion, determined by 1H NMR analysis. cMolar mass (Mn and Mw) and dispersity (Đ) were determined using gel permeation chromatography (GPC) at 30 °C with THF as the mobile phase and calibrated with polystyrene standards. dMn,PDL and Mn,PLLA are number-average molecular for PDL block and PLLA block, respectively, determined by 1H NMR analysis of linear (ratio of graft end groups to repeat units). eVolume fraction of PLLA was determined by 1H NMR and calculated based on published bulk homopolymer room temperature densities.7 fB728E105-g(13.5-40) is soluble in THF but could not pass through the SEC columns at concentrations appropriate for detection by RI due to its extremely high Mw. a

(BxEy-g(a-0)) and poly[(n-butyl acrylate)-co-(2-hydroxyethyl acrylate)]-graf t-(poly(ε-decalactone)-block-poly(lactide)) (BxEy-g(a-b)), respectively.32 Compositions of BxEy-g(a-b) were determined by 1H NMR spectroscopy using the endgroup analysis of side chains (Figure 4, Figures S2 and S3). The PLLA fraction (wt %) was controlled at either 50% or 70%, and in all cases the backbone comprises less than 4% of the total mass of the graft polymer. As demonstrated in Figures S4−S6, the relative molar mass of the polymer increased with each chain extension step, yet remained relatively low (Đ ∼ 1.5). Shoulders in the SEC data were observed with the BxEy-graft PDL polymers in the high molar mass region, most likely attributable to slow initiation relative to ROTEP propagation. It is possible that if a similar grafting from approach is used in future studies, the temperature, concentration, and catalyst loading could be adjusted to avoid this effect. Several control samples were also prepared. The graft block copolymer B434E62-g(6.5-10) has shorter side chains (6.5 kDa of PDL block) and ∼50 vol % PLLA (9 kDa). A linear poly(lactide)-block-poly(ε-decalactone)−poly(lactide) triblock copolymer, LDL (19-28-19), has a 28 kDa PDL midblock and 19 kDa of two PLLA blocks (total molar mass of 38 kDa). Rigid HPMC-graft block polymers were also prepared (Tables S1, S2 and Figure S8) according to our previous report.18 Like the BxEy statistical copolymers that generated flexible graft block copolymers, the set of HPMC macroinitiators used to prepare rigid graft block polymers vary in both molar mass and degree of hydroxy substitution. Three different polymers, H1/3(15), H1/6(15), and H1/3(40), were used. Among these, both H1/3(15) and H1/3(40) polymers have on average about one unsubstituted hydroxyl group for each sugar moiety. H1/6(15) has a lower degree of hydroxyl functionality with one unsubstituted hydroxyl group for every two sugar moieties. Whereas H1/3(15) and H1/6(15) are of similar molar mass (∼15 kg mol−1), H1/3(40) is higher molar mass (∼40 kg mol−1). As described previously, HPMC polymers (Scheme S1) were used as macroinitiatiors for ring-opening polymerization of β-

Molecular characteristics of the BxEy statistical copolymers are summarized in Table 1. The compositions of purified polymer samples were determined from the ratio of the integrals corresponding to the methylene protons adjacent to the ester (for nBA) or alcohol groups (for HEA) in the 1H NMR spectra of the copolymers (Figure 2, Figures S2 and S3).

Figure 2. 1H NMR spectra for (A) flexible backbone, B434E62, and (B) graft block copolymer, B434E62-g(14-34). Inserted picture was a zoomed part (circled) in Figure 4B, indicating the end parts of grafted side-chains (B434E62, 4.1−4.3 ppm, and PLLA, 4.4 ppm).

As expected, the relative molar masses of the samples, determined using size exclusion chromatography (SEC) with polystyrene standards, increased with the theoretical molar mass (Figures S4−S6). Three different copolymers, denoted B434E62, B357E103, and B728E105, were synthesized. B434E62 and B357E103 have similar molar mass but differ in composition; B728E105 has a higher molar mass but is compositionally similar to B434E62. The BxEy statistical copolymers were used as macroinitiatiors for ring-opening polymerization of ε-decalactone (DL) and, subsequently, L-lactide (LLA) to form poly[(n-butyl acrylate)co-(2-hydroxyethyl acrylate)]-graf t-(poly(ε-decalactone)) C

DOI: 10.1021/acs.macromol.6b02033 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

transition temperature of PDL homopolymer (−50 °C) and the other near the expected glass transition temperature of the PLLA homopolymer (near 55 °C). These separate Tg values are consistent with microphase separation of the PLLA and PDL domains. Similar to the case of HPMC-graft block polymers, the PLLA blocks of the grafts retain the ability to crystallize and melt in flexible BxEy-g(a-b). The melting temperature of the BxEy-g(a-b) polymers increased with PLLA content, yet the crystallization temperature (Tc) decreased. This is likely due to the fact that longer PLLA end blocks have a better chance to interact and nucleate with other PLLA chains, facilitating the crystallization in the polymers.33,34 As expected, the relative crystallinity also increased with the PLLA content, with the higher PLLA content samples ( f LA = 0.70) consistently have crystallinity roughly double those with a lower PLLA volume fraction ( f LA = 0.50). Similar behavior has previously been reported for graft polymers with PLLA side chains.33,34 Rheological Behaviors of Flexible Graft Block Polymers. To explore the influence of the graft architecture on the rheological behavior and processability, all polymer samples were investigated in the melt state at 190 °C. This temperature is above the melting temperature of the PLLA block in every case but is below the degradation temperature.35 At this temperature, all of the BxEy graft block copolymers exhibit nonterminal scaling (G″ ∼ ω0.28, G′ ∼ ω0.17) consistent with microphase separation of the PDL and PLLA components of the polymer (Figures S10 and S11). The behavior of the control LDL (19-28-19) triblock copolymer is only slightly different with G″ ∼ ω1/3 and G′ ∼ ω1/3. This may be indicative of a different morphology between the ABA triblock and the graft block copolymer at the same composition. The complex viscosities (η*) of all polymers are high and display a shear thinning response at all measured frequencies, also consistent with nanoscopic ordering in these materials. The η* values of BxEy-graft block copolymers are consistently greater than that of the triblock, which can be attributed to the extremely large overall molar masses of graft polymers. The qualitative behavior of the BxEy-graft block copolymers were similar regardless of PLLA content, yet higher G′ and G″ values were obtained in the samples containing 70% PLLA than in those with 50% PLLA, likely because these polymers contain longer PLLA blocks. In contrast, the rheological behavior of the HPMC graft block polymers was highly dependent on composition.18 In that system, at 170 °C H1/3(15)-g(8.5-9) ( f LA = 0.5) exhibited liquid-like behavior with terminal scaling, while H1/3(15)-g(8.5-20) (f LA = 0.7) demonstrated nonterminal scaling (G″ ∼ ω1/3, G′ ∼ ω1/3) at the same temperature. Additionally, whereas H1/3(15)-g(8.5-20) exhibited a frequency-independent zero-shear viscosity at low frequency, H1/3(15)-g(8.5-20) was shear thinning, similar to the BxEy-graft block copolymers. We attributed this variation to a state of disorder in the f LA = 0.5 material. SAXS of Graft Block Copolymers. Small-angle X-ray scattering (SAXS) experiments provide further evidence that suggests the PDL and PLLA segments of the BxEy-graft block polymers are immiscible in all of the samples prepared.36 As shown in Figure 4, the samples are characterized by a broad principal reflection (q*) along with two scattering reflections at higher q. For these flexible graft block polymers, the position of the primary peak is more affected by the composition and molar mass of the diblock pendants than the backbone length, as expected. The compositionally symmetric samples ( f LA =

methyl-δ-valerolactone (MVL) and, subsequently, L-lactide (LLA) to form HPMC-graf t-poly(β-methyl-δ-valerolactone) (e.g., H1/340-g(a-0)) and HPMC-graf t-poly(β-methyl-δ-valerolactone)-block-poly(L-lactide) (e.g., H1/340-g(a-b)), respectively. The molecular information on HPMC-graft block copolymers are listed in Tables S1 and S2. Here we note that this work compares graft block polymers containing different soft blocks, i.e., PMVL and PDL, which introduces quantitative differences in the thermodynamic, melt viscoelastic, and solid state properties. Nevertheless, we are able to draw several general conclusions regarding the consequences of certain molecular architectural variables as described below. Thermal Properties of Graft Block Copolymers. The thermal properties of BxEy-g(a-b) graft block copolymers were studied using differential scanning calorimetry (DSC) and are summarized in Table 2. Figure 3 and Figure S9 show Table 2. Thermal Characteristics of BxEy-Graft Block Polymers and Control Triblock Copolymer samples B357E103-g (13.5-18) B357E103-g (13.5-37) B434E62-g(6.5-9) B434E62-g(14-20) B434E62-g(14-34) B728E105-g (13.5-20) B728E105-g (13.5-40) LDL (19-28-19)

Tg,Da (°C)

Tg,La (°C)

Tcb (°C)

T mc (°C)

Xd (%)

De (nm)

−48

60

112

169

4

44

−49

62

104

174

10

66

−44 −48 −47 −48

62 61 62 60

108 119 106 119

164 168 174 167

7 4 10 7

24 44 60 42

−47

60

105

172

11

64

−49

60

119

168

3

45

a

Glass transition temperatures and melting temperatures were obtained from DSC during second heating. Tg,D is the glass transition temperature of PDL domain, and Tg,L is for the PLLA domain. bPLLA crystallization temperature. The onset of the exothermic crystallization peak on the second heating curve was taken. cPLLA melting temperature. The apex of the endothermic melting peak on the second heating curve was taken. dPLLA crystallinity determined on the second heating curve. PLLA crystallinity (%) was calculated based on the area of the melting and cold crystallization peaks (subtract the area of cold crystallization peaks from melting ones) using 93.0 J/g as the enthalpy of fusion for PLLA crystals and normalized by the weight fraction of PLLA.6 eDomain spacing were calculated by using D = 2π/ q* based on SAXS data (Figure 4).

Figure 3. DSC curves of graft block polymers and linear control sample, LDL (19-28-19). Arrows indicate Tg values of PDL and PLLA domains (Tg,D and Tg,L). The second heating curves with a heating rate of 10 °C min−1 in DSC were used for all analyses.

representative DSC traces. For all samples, two glass transition temperatures (Tg) were observed: one near the expected glass D

DOI: 10.1021/acs.macromol.6b02033 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

revisited here. In fact, poly(β-methyl-δ-valerolactone) and poly(L-lactide) (χMVL‑LA = 0.046 at 90 °C and χMVL‑LA = 0.025 at 190 °C) are much more compatible.43 The pendant diblocks of the HPMC graft block polymers are also relatively low molar mass. For some samples, e.g., H1/3(15)-g(8.5-9) and H1/3(15)-g(8.5-20) cold crystallization actually induces microphase separation in the material.18 Mechanical Properties of Flexible BxEy and Rigid HPMC Graft Block Copolymers. Representative engineering stress versus strain curves obtained in uniaxial tension for each sample are shown in Figures 5−9. All polymer samples in this Figure 4. Small-angle X-ray scattering (SAXS) obtained at 190 °C for BxEy-graft block polymers with (A) 50% (volume fraction) and (B) 70% (volume fraction) PLLA. Arrows denote the position of the principal peak (q*) for samples.

0.5) exhibit apparent domain spacing values of about 43 nm, which is similar to our estimated values (52 nm) (see Supporting Information about calculations); samples with higher PLLA content (f LA = 0.70) are characterized by slightly larger domain spacing values around 57 nm. The broad peaks in the SAXS spectra indicate relatively poor long-range order, perhaps due to the short annealing time. We note that although the pendant diblocks are relatively low molar mass, the overall molar mass of the graft polymer is quite high. This may slow ordering and prohibit the polymer from adopting the well-ordered morphologies typical of linear diblocks at these compositions ( f LA = 0.5 and f LA = 0.70 are anticipated to be lamellar and cylindrical, respectively, based on previous work).37 Although the higher order peaks are broad and somewhat ill defined, it is interesting to note that these peaks are not observed in the same samples annealed at 90 °C (Figure S12). Additionally, in every case the intensity and position of the primary peak shift significantly upon cooling. These changes suggest that crystallization of the PLLA segments upon cooling disrupts the melt morphology of the BxEy-graft block polymers.38−40 For linear amorphous−semicrystalline block polymers, this type of crystalline breakout would be expected for ordered samples allowed to crystallize at elevated temperatures (Tg < Tc < TODT), provided the segregation strength, a product of the reference volume averaged degree of polymerization (N) and the Flory−Huggins interaction parameter (χ), is low.41 For the semiflexible BxEy-graft block polymers, the melt microstructures are largely controlled by the side-chain composition, while the graft density (or rigidity) has less of an impact on the phase separation. The large molar mass of these graft block copolymers may kinetically trap the material in a microphase-separated structure that lacks long-range order. This is supported by the SAXS results (Figure 4A), BxEy-graft block polymers with similar side-chain chemistry form similar structures in the melt although they have different backbone rigidity or graft density. Because the higher order peaks are not well-defined in these samples, it is not possible for SAXS alone to say whether the samples actually have the same morphology in the melt or not. The Flory−Huggins interaction parameter (χ) for the polymer pair which comprises the graft pendants is large enough (χDL‑LA = 0.12 at 90 °C and χDL‑LA = 0.08 at 190 °C) that one would expect linear diblocks of the same composition and molar mass to be ordered at experimentally relevant temperatures.36,37,42 This is qualitatively different than the rigid HPMC graft block polymers previously explored and

Figure 5. Investigation of the mechanical behavior of samples with different side-chain lengths with fixed backbone length and side-chain composition. Representative examples of room temperature uniaxial extension tests for selected samples from (A) BxEy-graft block polymers, B434E62-g(14-20),and B434E62-g(6.5-9) and (B) HPMCgrafted block polymers, H1/3(15)-g(8.5-9) and H1/3(15)-g(6.0-7). Pictures in black and red boxes are schematic illustrations of associated polymers.

study show typical plastic deformation with differences in elastic modulus (E), strain at break (εb), stress at break (σb), and toughness (defined here as the integrated area of stress strain curve). Average values (determined for at least five specimens) of E, εb, σb, σy, and toughness are summarized in Table S3 and are compared to the properties of HPMC-graft block polymers previously reported.18 Interestingly, unlike linear block polymers, graft block polymer thermoplastics from both flexible BxEy and rigid HPMC backbones remain transparent after being stretched (Figure S13). This architecture is particularly promising for applications where optical transparency is essential. Impact of Side-Chain Composition and Length. The mechanical performance of graft block polymers is highly dependent on the composition and length of the side chains. As shown in Figure 5, at fixed backbone length and PLLA volume fraction, elongation and stress at break both increase with the overall side chain length regardless of backbone rigidity. The result is that the tensile toughness of thermoplastics with longer side chains is about 4−5 times higher than that of polymers E

DOI: 10.1021/acs.macromol.6b02033 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules with shorter ones (Table S3). For samples with fixed PLLA content and a rigid (HPMC) backbone, the elastic modulus is similar regardless of side chain length. This trend is also observed in linear block copolymers.13,44 However, the modulus of B434E62-g(14-20) is nearly double that of B434E62g(6.5-9) (Table S3). Intuitively, one might anticipate longer side chains would be more likely than shorter side chains to form entanglements. Such physical cross-links should increase the mechanical strength of the bulk sample by coupling the elastic soft blocks and/or semicrystalline hard blocks.13 Nevertheless, it is interesting that a similar effect is observed in the flexible BxEy graft block polymers as the rigid HPMC-graft block polymers; these samples contain poly(β-methyl-δ-valerolactone)-blockpoly(lactide) and poly(ε-decalactone)-block-poly(lactide) as the as rubbery-block-glassy diblock pendants, respectively. Based on the entanglement molar masses (Me) of PLA, PDL, and PMVL, it is likely that (of the samples shown in Figure 5) only the pendants which comprise B434E62-g(14-20) are significantly entangled. As shown in Figure 6, at fixed backbone length and variable side-chain composition, Young’s modulus and stress at break

state, the HPMC materials crystallize from a disordered state during processing. This may translate to a difference in morphology and may explain why the strain at break only slightly decreases with PLA content for the BxEy graft block copolymers, contrary to the behavior observed for linear block polymers and the HPMC materials.13,44 Impact of Graft Density on Mechanical Properties. Graft density, the space between two neighboring side chains, is another key parameter for graft block polymers. For the flexible backbone system (BxEy) we investigated three macroinitiatiors (B357E103, B434E62, and B728E105). Samples B357E103 and B434E62 have similar molar mass (∼58 and 64 kg mol−1) but differ in graft density. On average, B357E103 has one hydroxyl group (grafting site) every 3.4 repeat units while B434E62 has one hydroxyl group every 7 repeat units, making the molar mass between grafts ∼450 and ∼900 g mol−1, respectively. Similarly, for the rigid backbone system (HPMC), we also investigated three macroinitiatiors (H1/3(15), H1/6(15), and H1/3(40)) with two different graft densities. Samples H1/3(15) and H1/6(15) have similar molar mass (both ∼15 kg mol−1). On average, H1/3(15) has one hydroxyl group every sugar moiety while H1/6(15) has one hydroxyl group every two sugar moieties, making the molar mass between grafts ∼230 and ∼460 g mol−1, respectively. As shown in Figure 7A, graft density has a significant impact on stress and strain at break. At fixed side chain molar mass and

Figure 6. Investigation of the mechanical behavior of samples with different side-chain compositions and lengths at fixed backbone length. Representative examples of room temperature uniaxial tensile testing of selected samples from (A) BxEy-graft block polymers: B434E62-g(1420) and B434E62-g(14-34); (B) HPMC-grafted block polymers: H1/3(15)-g(8.5-9) and H1/3(15)-g(8.5-20). Pictures in black and red boxes are schematic illustrations of associated polymers.

Figure 7. Mechanical behavior of samples with different graft densities. Representative examples of room temperature uniaxial tensile testing of selected samples from (A) BxEy backbones and (B) HPMC backbones with different graft densities (red curves: low graft density; black curves: high graft density). Pictures in black and red boxes are schematic illustrations of associated polymers with 50% PLLA in sidegrafts.

both increase with PLA content, while strain at break decreases slightly. As before, the same qualitative trends are observed for polymers with rigid HPMC as flexible BxEy backbones. Although toughness also increases with PLA content for the samples with BxEy backbones, the opposite trend is observed for the samples with HPMC backbones. As mentioned previously, while the BxEy samples crystallize from an ordered

composition, the flexible BxEy graft block polymers with a lower graft density have drastically higher tensile toughness relative to their counterparts with a high graft density. Specifically we observed toughness values of ∼2 and ∼60 MJ m−3 for B357E103g(13.5-18) and B434E62-g(14-20), respectively. The same trend is also preserved when the PLA content is increased. We F

DOI: 10.1021/acs.macromol.6b02033 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules postulate that the difference in tensile toughness may be due to a change in the free space occupied by the side chains of the graft block polymers. We speculate that for graft block polymers with flexible backbones, a high graft density may limit the room for individual side chains and entire macromolecules to interact, entangle, and cocrystallize with each other. We note that reducing the graft density at constant graft block composition (f L) changes the overall PLA content slightly (ca.