Tunable Shape Memory Polymers from α‑Amino Acid-BasedPoly(ester urea)sGregory I. Peterson, Erin P. Childers, Hao Li, Andrey V. Dobrynin, and Matthew L. Becker*

Department of Polymer Science, The University of Akron, Akron, Ohio 44325-3909, United States

*S Supporting Information

ABSTRACT: The thermal shape memory behavior of poly(esterurea)s (PEUs) composed of varying α-amino acids and linear diolshas been explored. The thermal, mechanical, and shape memoryproperties of PEUs are shown to be controlled by changing theamino acid and diol components of the polymer, without negativelyaffecting the shape memory performance of the polymer in mostcases. These materials display triple-shape memory behavior andtemperature memory properties due to a broad glass transitiontemperature interval. The versatility of the shape memory behaviorof PEUs was explored by demonstrating shape transformations of thin films, salt leached scaffolds, and larger constructs. Overallgood shape memory behavior in combination with the tunable properties of PEU materials makes them prime candidates for useas shape memory materials in biomedical applications.

■ INTRODUCTION

Shape memory polymers (SMPs) provide materials that canchange shape upon application of a stimulus and have shownconsiderable promise for use in biomedical applications.1−6

Self-tightening sutures,7 deployable structures in medicaldevices,8 shape changing scaffolds and surfaces,9,10 and drugdelivery systems11 represent some of the potential uses ofSMPs. The advantages of using polymers, as opposed to shapememory alloys,12,13 include the potential for biodegradability,lower cost, larger achievable strains during shape actuation, andtunable material properties. Considering the broad range oftissue types and corresponding mechanical properties encoun-tered in the body, the ability to tune a material’s properties tomeet the demands of a particular application is vital.The simplest SMPs are dual-shape memory materials that

require, first, programming a temporary shape, followed byapplication of an appropriate stimulus (heat being the mostcommon) to trigger recovery of the permanent shape.14−18

Other stimuli can be used such as light, chemical impetus, orvarious methods of indirect heating (e.g., photo-, electro-, andmagneto-thermal transduction). The two basic requirementsfor a thermal SMP are possessing: 1) a reversible thermaltransition (i.e., glass or melt transition) to activate and suppresschain mobility and 2) a cross-linked structure to prevent chainslippage and set the permanent shape.18 In addition, importantdesign considerations for SMPs in biomedical applicationsinclude biodegradability, biocompatibility, compatible mechan-ical properties, and sterilizability.3 A wide range of thermalSMPs, including polyesters, polyurethanes, and polyacrylates,have been identified as viable candidates for biomedicalapplications,1,4,5,19,20 although the extent to which thesematerials meet the design considerations lacks a comprehensivestudy.

α-Amino acid-based poly(ester urea)s (PEUs) have recentlyemerged as an important class of tunable materials forbiomedical applications. These materials are biodegradable,sterilizable, nontoxic, have nontoxic degradation products, andlead to no inflammatory response during degradation in vivo.21

Their mechanical properties can be tuned for use in both hardand soft tissues, such as bone and blood vessels.22−24

Additionally, the materials can be prepared with variousfunctionalities for specific applications, such as peptides forbone growth,25 iodine for radiopacity,26 catechols foradhesion,27 fluorescent probes for visualization,28 and ther-apeutics for drug delivery.29 The main advantages of PEUs overmany other biodegradable polymers include simple scalablesynthesis, tunable degradation and mechanical properties, andmechanical properties derived from hydrogen bonding ratherthan crystallinity. This versatility, and the previously demon-strated examples of biocompatibility, makes PEUs viablecandidates for a wide range of biomedical applications.We recently demonstrated that L-valine (VAL)-based PEUs

exhibit thermal shape memory behavior that takes advantage ofa broad glass transition temperature (Tg), above whichsignificant chain mobility could be activated, and shapeprogramming and recovery were achieved.30 The materials donot have chemical cross-links but possess a strong hydrogenbonding network that form the physical cross-links required forshape imprinting. Excellent dual- and triple-shape memoryperformance was observed, and quadruple-shape memorybehavior could be achieved by blending VAL-based PEUswith different diol chain lengths incorporated into the polymer

Received: April 1, 2017Revised: May 9, 2017Published: May 18, 2017

Article

pubs.acs.org/Macromolecules

© 2017 American Chemical Society 4300 DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

backbone. In this article, we explore the effect of amino acidand diol lengths on the shape memory behavior of PEUsprepared from L-alanine (ALA), L-isoleucine (ILE), L-phenyl-alanine (PHE), and the unnatural amino acid L-2-aminobutyricacid (ABA).

■ EXPERIMENTAL SECTIONMaterials. Chloroform was either obtained from an Inert Pure Solv

solvent purification system or dried over calcium hydride overnightand then distilled. All other reagents and solvents were used asobtained from commercial sources.Characterization. NMR spectra were collected with Varian NMR

spectrometers (300 and 500 MHz). All chemical shifts were reportedin ppm (δ) and referenced to the chemical shifts of the residual solventresonances (1H NMR, dimethyl sulfoxide (DMSO)-d6: 2.50 ppm; 13CNMR DMSO-d6: 39.50 ppm). The following abbreviations were usedto explain the multiplicities: s = singlet, d = doublet, t = triplet, q =quartet, and m = multiplet. Number-average molecular mass (Mn) andpostprecipitation molecular mass distribution (ĐM) were determinedby size exclusion chromatography (SEC), and molecular mass valueswere determined relative to polystyrene standards. The SEC analyseswere performed using a TOSOH HLC-8320 gel permeationchromatograph instrument with dimethylformamide (DMF) (con-taining 0.01 M LiBr) as eluent (flow rate of 1 mL/min andtemperature of 50 °C) and a refractive index detector. The Tg ofpolymers was determined by differential scanning calorimetry (DSC,TA Q2000, scan rate of 20 °C/min) or dynamic mechanical analysis(DMA, TA Q800, 3 °C/min and a frequency of 1 Hz). X-raydiffraction (XRD) data were collected on a Rigaku Ultima IV X-raydiffractometer. IR spectra of monomers and polymers were collectedon a Nicolet i550 FT-IR (Thermo Scientific) after dissolution inchloroform and application to a KBr salt plate (32 scans, 8 cm−1

resolution).Synthesis of VAL- and PHE-Based PEUs. The VAL- and PHE-

based PEUs were prepared and characterized as previouslydescribed.22,24

General Procedure for Synthesis of PEU Monomers. Either1,6-hexanediol, 1,8-octanediol, 1,10-decanediol, or 1,12-dodecandiol(1.0 mol equiv), a L-amino acid (2.3 mol equiv), p-toluenesulfonic acidmonohydrate (TsOH) (2.4 mol equiv), and toluene (1 mL per gramof TsOH) were added to round-bottom flask equipped with Dean−Stark trap and condenser. The solution was heated to reflux (ca. 110°C) while stirring with a magnetic stir bar. After ca. 20 h, the reactionmixture was cooled to ambient temperature. The resulting precipitatewas collected by vacuum filtration. The solid product was dissolved inminimal hot water and decolored using a small amount of activatedcarbon black for 2−3 min. This solution was filtered to remove thecarbon black and was left to cool to room temperature. The precipitatewas then recrystallized three times using hot water to give the purifiedmonomer.Synthesis of m(1-ALA-6). The monomer was prepared by

following the general procedure described above, with the exceptionof the recrystallization procedure. The monomer was recrystallizedfour times from a 1:1 mixture (by volume) of ethanol and isopropanol.The monomer was prepared on a 145 mmol scale (based on the diol)and obtained with a 79% yield. 1H NMR (500 MHz, DMSO-d6, δ):8.27 (s, 6H; NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 7.8Hz, 4H; Ar−H), 4.16 (m, 4H; CH2), 4.10 (q, J = 7.2 Hz, 2H; CH),2.29 (s, 6H; CH3), 1.61 (m, 4H; CH2), 1.39 (d, J = 7.2 Hz, 6H; CH3),1.35 (m, 4H; CH2).

13C NMR (126 MHz, DMSO-d6, δ): 169.92,145.21, 137.89, 128.10, 125.46, 65.49, 47.93, 27.76, 24.71, 20.75,15.70. IR (cm−1): 1743 (−C−(CO)−O−).Synthesis of m(1-ALA-8). The monomer was prepared by

following the general procedure described above, with the exceptionof the recrystallization procedure. The monomer was recrystallizedfour times from a 1:1 mixture (by volume) of ethanol and isopropanol.The monomer was prepared on a 147 mmol scale (based on the diol)and obtained with a 79% yield. 1H NMR (500 MHz, DMSO-d6, δ):8.27 (s, 6H; NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 7.8

Hz, 4H; Ar−H), 4.13 (m, 6H; CH2 and CH), 2.29 (s, 6H; CH3), 1.59(m, 4H; CH2), 1.39 (d, J = 7.2 Hz, 6H; CH3), 1.32 (m, 8H; CH2).

13CNMR (126 MHz, DMSO-d6, δ): 169.92, 145.27, 137.83, 128.10,125.45, 65.56, 47.92, 28.43, 27.87, 25.05, 20.71, 15.69. IR (cm−1):1749 (−C−(CO)−O−).

Synthesis of m(1-ALA-10). The monomer was prepared byfollowing the general procedure described above, with the exception ofthe recrystallization procedure. The monomer was recrystallized fourtimes from a 1:1 mixture (by volume) of ethanol and isopropanol. Themonomer was prepared on a 132 mmol scale (based on the diol) andobtained with an 80% yield. 1H NMR (500 MHz, DMSO-d6, δ): 8.26(s, 6H; NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 8.0 Hz, 4H;Ar−H), 4.14 (m, 6H; CH2 and CH), 2.29 (s, 6H; CH3), 1.60 (m, 4H;CH2), 1.39 (d, J = 7.2 Hz, 6H; CH3), 1.29 (m, 12H; CH2).

13C NMR(126 MHz, DMSO-d6, δ): 169.92, 145.32, 137.78, 128.05, 125.45,65.58, 47.90, 28.81, 28.55, 27.89, 25.12, 20.73, 15.68. IR (cm−1): 1736(−C−(CO)−O−).

Synthesis of m(1-ALA-12). The monomer was prepared byfollowing the general procedure described above, with the exception ofthe recrystallization procedure. The monomer was recrystallized fourtimes from a 1:1 mixture (by volume) of ethanol and isopropanol. Themonomer was prepared on a 145 mmol scale (based on the diol) andobtained with an 80% yield. 1H NMR (500 MHz, DMSO-d6, δ): 8.27(s, 6H; NH3), 7.49 (d, J = 7.5 Hz, 4H; Ar−H), 7.12 (d, J = 7.5 Hz, 4H;Ar−H), 4.13 (m, 6H; CH2 and CH), 2.29 (s, 6H; CH3), 1.59 (m, 4H;CH2), 1.39 (d, J = 7.0 Hz, 6H; CH3), 1.27 (m, 16H; CH2).

13C NMR(126 MHz, DMSO-d6, δ): 169.90, 145.21, 137.85, 128.07, 125.45,65.57, 47.92, 28.93, 28.89, 28.58, 27.90, 25.13, 20.74, 15.67. IR (cm−1):1736 (−C−(CO)−O−).

Synthesis of m(1-ABA-6). The monomer was prepared byfollowing the general procedure described above, with the exception ofthe recrystallization procedure. The monomer was recrystallized threetimes from a 3:4 (by volume) of ethanol and ethyl acetate. Themonomer was prepared on a 46 mmol scale (based on the diol) andobtained with a 73% yield. 1H NMR (300 MHz, DMSO-d6, δ): 8.33(s, 6H, NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.13 (d, J = 7.8 Hz, 2H;Ar−H), 4.15 (m, 4H; CH2), 4.01 (m, 2H; CH), 2.29 (s, 6H; CH3),1.81 (m, 4H; CH2), 1.60 (m, 4H; CH2), 1.34 (m, 4H; CH2), 0.91 (t, J= 7.4 Hz, 6H; CH3).

13C NMR (75 MHz, DMSO-d6, δ): 169.46,145.08, 138.07, 128.21, 125.53, 65.51, 53.11, 27.84, 24.80, 23.46,20.83, 9.06. IR (cm−1): 1745 (−C−(CO)−O−).

Synthesis of m(1-ABA-8). The monomer was prepared byfollowing the general procedure described above, with the exception ofthe recrystallization procedure. The monomer was recrystallized threetimes from a 3:4 (by volume) of ethanol and ethyl acetate. Themonomer was prepared on a 46 mmol scale (based on the diol) andobtained with an 81% yield. 1H NMR (300 MHz, DMSO-d6, δ): 8.31(s, 6H, NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 7.9 Hz, 2H;Ar−H), 4.16 (m, 4H; CH2), 4.00 (m, 2H; CH), 2.29 (s, 6H; CH3),1.81 (m, 4H; CH2), 1.59 (m, 4H; CH2), 1.29 (m, 8H; CH2), 0.92 (t, J= 7.5 Hz, 6H; CH3).

13C NMR (75 MHz, DMSO-d6, δ): 169.47,144.92, 138.23, 128.27, 125.58, 65.61, 53.18, 28.54, 27.99, 25.21,23.49, 20.87, 9.09. IR (cm−1): 1745 (−C−(CO)−O−).

Synthesis of m(1-ABA-10). The monomer was synthesized asdescribed in the general procedure described above. The monomerwas prepared on a 46 mmol scale (based on the diol) and obtainedwith a 67% yield. 1H NMR (300 MHz, DMSO-d6, δ): 8.30 (s, 6H,NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 7.9 Hz, 2H; Ar−H), 4.16 (m, 4H; CH2), 4.00 (t, J = 6.0 Hz, 2H; CH), 2.29 (s, 6H;CH3), 1.81 (m, 4H; CH2), 1.60 (m, 4H; CH2), 1.27 (m, 12H; CH2),0.92 (t, J = 7.5 Hz, 6H; CH3).

13C NMR (75 MHz, DMSO-d6, δ):169.46, 145.05, 138.08, 128.20, 125.54, 65.60, 53.12, 28.90, 28.62,27.99, 25.25, 23.46, 20.83, 9.05. IR (cm−1): 1745 (−C−(CO)−O−).

Synthesis of m(1-ABA-12). The monomer was synthesized asdescribed in the general procedure described above. The monomerwas prepared on a 46 mmol scale (based on the diol) and obtainedwith an 83% yield. 1H NMR (300 MHz, DMSO-d6, δ): 8.30 (s, 6H,NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 7.7 Hz, 2H; Ar−H), 4.16 (m, 4H; CH2), 4.02 (m, 2H; CH), 2.29 (s, 6H; CH3), 1.81(m, 4H; CH2), 1.60 (m, 4H; CH2), 1.25 (m, 16H; CH2), 0.92 (t, J =

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4301

7.5 Hz, 6H; CH3).13C NMR (75 MHz, DMSO-d6, δ): 169.44, 144.95,

138.15, 128.23, 125.57, 65.59, 53.16, 29.05, 29.02, 28.69, 28.02, 25.29,23.47, 20.85, 9.07. IR (cm−1): 1742 (−C−(CO)−O−).Synthesis of m(1-ILE-6). The monomer was synthesized as

described in the general procedure described above. The monomerwas prepared on an 80 mmol scale (based on the diol) and obtainedwith an 85% yield. 1H NMR (500 MHz, DMSO-d6, δ): 8.30 (s, 6H;NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 8.1 Hz, 4H; Ar−H), 4.16 (m, 4H; CH2), 3.96 (s, 2H; CH), 2.29 (s, 6H; CH3), 1.87 (m,2H; CH), 1.60 (m, 4H; CH2), 1.36 (m, 8H; CH2), 0.89 (m, 12H;CH3).

13C NMR (126 MHz, DMSO-d6, δ): 168.69, 145.44, 137.69,128.01, 125.43, 65.40, 56.06, 35.91, 27.75, 25.23, 24.74, 20.71, 14.18,11.41. IR (cm−1): 1736 (−C−(CO)−O−).Synthesis of m(1-ILE-8). The monomer was synthesized as

described in the general described above. The monomer was preparedon a 70 mmol scale (based on the diol) and obtained with a 92% yield.1H NMR (500 MHz, DMSO-d6, δ): 8.29 (s, 6H, NH3), 7.49 (d, J = 8.0Hz, 4H; Ar−H), 7.12 (d, J = 8.2 Hz, 4H; Ar−H), 4.15 (m, 4H; CH2),3.96 (d, J = 3.9 Hz, 2H; CH), 2.29 (s, 6H; CH3), 1.88 (m, 2H; CH),1.59 (m, 4H; CH2), 1.44 (m, 2H; CH2) 1.28 (m, 10H; CH2), 0.88 (m,12H; CH3).

13C NMR (126 MHz, DMSO-d6, δ): 168.71, 145.42,137.70, 128.01, 125.43, 65.49, 56.07, 35.92, 28.35, 27.85, 25.23, 25.14,20.72, 14.17, 11.41. IR (cm−1): 1747 (−C−(CO)−O−).Synthesis of m(1-ILE-10). The monomer was synthesized as

described in the general procedure described above. The monomerwas prepared on a 60 mmol scale (based on the diol) and obtainedwith an 86% yield. 1H NMR (500 MHz, DMSO-d6, δ): 8.30 (s, 6H;NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 7.9 Hz, 4H; Ar−H), 4.15 (m, 4H; CH2), 3.96 (d, J = 3.6 Hz, 2H; CH), 2.29 (s, 6H;CH3), 1.87 (m, 2H; CH), 1.58 (m, 4H; CH2), 1.44 (m, 2H; CH2) 1.28(m, 10H; CH2), 0.90 (m, 12H; CH3).

13C NMR (126 MHz, DMSO-d6, δ): 168.71, 145.39, 137.74, 128.04, 125.45, 65.54, 56.09, 35.93,28.78, 28.47, 27.89, 25.25, 25.24, 20.73, 14.18, 11.43. IR (cm−1): 1745(−C−(CO)−O−).Synthesis of m(1-ILE-12). The monomer was synthesized as

described in the general procedure described above. The monomerwas prepared on a 40 mmol scale (based on the diol) and obtainedwith an 82% yield. 1H NMR (500 MHz, DMSO-d6, δ): 8.29 (s, 6H,NH3), 7.49 (d, J = 8.0 Hz, 4H; Ar−H), 7.12 (d, J = 8.0 Hz, 4H; Ar−H), 4.15 (m, 4H; CH2), 3.96 (d, J = 3.9 Hz, 2H; CH), 2.29 (s, 6H;CH3), 1.88 (m, 2H; CH), 1.59 (m, 4H; CH2), 1.45 (m, 2H; CH2) 1.27(m, 10H; CH2), 0.88 (m, 12H; CH3).

13C NMR (126 MHz, DMSO-d6, δ): 168.70, 145.41, 137.69, 128.01, 125.44, 65.52, 56.08, 35.92,28.86, 28.47, 27.88, 25.23, 25.22, 20.72, 14.16, 11.41. IR (cm−1): 1745(−C−(CO)−O−).General Procedure for the Synthesis of PEUs. Monomer (1.0

mol equiv), sodium carbonate anhydrate (2.1 mol equiv), anddeionized water (10 mL per mmol of monomer) were added into a3 L three-neck round-bottom flask. The solution was mechanicallystirred (400−450 rpm) in a 35 °C water bath for 0.5 h to dissolve themonomer. An ice bath was then used to cool the solution to 0 °C, andanother aliquot of sodium carbonate (1.05 mol equiv) in deionizedwater (4 mL per mmol of monomer) was added. Next, a solution oftriphosgene (0.35 mol equiv) dissolved in dry chloroform (2.5 mL permmol of monomer) was added to the round-bottom flask, all at once,with an addition funnel. After 0.5 h, an additional aliquot oftriphosgene (0.08 mol equiv) in chloroform (1 mL per mmol ofmonomer) was added dropwise via the addition funnel. Thepolymerization solution was stirred for 2−21 h, and the ice bath wasallowed to expire. After the reaction time, the solution was transferredto a separatory funnel and added dropwise into hot (>70 °C)deionized water. The polymer was collected and reprecipitated ifresidual monomer was detected by NMR. Polymers were dried underreduced pressure.Synthesis of p(1-ALA-6). The polymer was prepared by following

the general procedure described above with the exception the numberof triphosgene addition steps. To further increase the molecular massof the polymer, the amount of triphosgene in the second addition wasincreased to 0.16 mol equiv, and a third addition of triphosgene (0.16mol equiv, in chloroform, 1 mL per mmol of monomer) was added

after 2 h from the second addition. The polymer was prepared on a 33mmol scale (based on monomer), stirred for 17 h after the thirdtriphosgene addition, and obtained with a 70% yield. 1H NMR (300MHz, DMSO-d6, δ): 6.35 (d, J = 7.7 Hz, NH), 4.03 (m, CH2 andCH), 1.53 (m, CH2), 1.32 (m, CH2), 1.21 (d, J = 6.3 Hz, CH3). IR(cm−1): 1550, 1638 (−NH−(CO)−NH−), 1728 (−C−(CO)−O−),3356 (−NH−(CO)−NH−).

Synthesis of p(1-ALA-8). The polymer was prepared by followingthe general procedure described above with the exception the numberof triphosgene addition steps. To further increase the molecular massof the polymer, the amount of triphosgene in the second addition wasincreased to 0.16 mol equiv, and a third addition of triphosgene (0.16mol equiv, in chloroform, 1 mL per mmol of monomer) was addedafter 2 h from the second addition. The polymer was prepared on a 30mmol scale (based on monomer), stirred for 17 h after the thirdtriphosgene addition, and obtained with a 71% yield. 1H NMR (300MHz, DMSO-d6, δ): 6.35 (s, NH), 4.00 (m, CH2 and CH), 1.53 (m,CH2), 1.25 (m, CH2), 1.21 (d, J = 7.2 Hz, CH3). IR (cm−1): 1564,1634 (−NH−(CO)−NH−), 1738 (−C−(CO)−O−), 3323 (−NH−(CO)−NH−).

Synthesis of p(1-ALA-10). The polymer was prepared byfollowing the general procedure described above with the exceptionthe number of triphosgene addition steps. To further increase themolecular mass of the polymer, the amount of triphosgene in thesecond addition was increased to 0.16 mol equiv, and a third additionof triphosgene (0.16 mol equiv, in chloroform, 1 mL per mmol ofmonomer) was added after 2 h from the second addition. The polymerwas prepared on a 30 mmol scale (based on monomer), stirred for 17h after the third triphosgene addition, and obtained with an 89% yield.1H NMR (300 MHz, DMSO-d6, δ): 6.35 (d, J = 7.7 Hz, NH), 4.02 (m,CH2 and CH), 1.52 (m, CH2), 1.23 (m, CH2 and CH3). IR (cm−1):1562, 1634 (−NH−(CO)−NH−), 1736 (−C−(CO)−O−), 3350(−NH−(CO)−NH−).

Synthesis of p(1-ALA-12). The polymer was synthesized asdescribed in the general procedure described above. The polymer wasprepared on a 29 mmol scale (based on monomer), stirred for 12 hafter the second triphosgene addition, and obtained with an 84% yield.1H NMR (300 MHz, DMSO-d6, δ): 6.35 (d, J = 7.7 Hz, NH), 4.00 (m,CH2 and CH), 1.52 (m, CH2), 1.23 (m, CH2 and CH3). IR (cm−1):1562, 1634 (−NH−(CO)−NH−), 1736 (−C−(CO)−O−), 3339(−NH−(CO)−NH−).

Synthesis of p(1-ABA-6). The polymer was synthesized asdescribed in the general procedure described above. The polymerwas prepared on a 16 mmol scale (based on monomer), stirred for 21h after the second triphosgene addition, and obtained with a 95% yield.1H NMR (300 MHz, DMSO-d6, δ): 6.37 (d, J = 7.5 Hz, NH), 4.07 (m,CH2 and CH), 1.61 (m, CH2), 1.31 (m, CH2), 0.85 (t, J = 6.9 Hz,CH3). IR (cm−1): 1563, 1636 (−NH−(CO)−NH−), 1734 (−C−(CO)−O−), 3347 (−NH−(CO)−NH−).

Synthesis of p(1-ABA-8). The polymer was synthesized asdescribed in the general procedure described above. The polymerwas prepared on a 15 mmol scale (based on monomer), stirred for 21h after the second triphosgene addition, and obtained with a 97% yield.1H NMR (300 MHz, DMSO-d6, δ): 6.39 (d, J = 8.1 Hz, NH), 4.01 (m,CH2 and CH), 1.62 (m, CH2), 1.26 (m, CH2), 0.84 (t, J = 7.3 Hz,CH3). IR (cm−1): 1559, 1640 (−NH−(CO)−NH−), 1736 (−C−(CO)−O−), 3356 (−NH−(CO)−NH−).

Synthesis of p(1-ABA-10). The polymer was synthesized asdescribed in the general procedure described above. The polymer wasprepared on a 15 mmol scale (based on monomer), stirred for 4 h afterthe second triphosgene addition, and obtained with an 89% yield. 1HNMR (300 MHz, DMSO-d6, δ): 6.40 (d, J = 7.5 Hz, NH), 4.06 (m,CH2 and CH), 1.60 (m, CH2), 1.25 (m, CH2), 0.85 (m, CH3). IR(cm−1): 1561, 1638 (−NH−(CO)−NH−), 1738 (−C−(CO)−O−),3355 (−NH−(CO)−NH−).

Synthesis of p(1-ABA-12). The polymer was synthesized asdescribed in the general procedure described above. The polymer wasprepared on a 15 mmol scale (based on monomer), stirred for 5 h afterthe second triphosgene addition, and obtained with a 90% yield. 1H

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4302

NMR (300 MHz, DMSO-d6, δ): 6.37 (d, J = 8.0 Hz, NH), 4.04 (m,CH2 and CH), 1.60 (m, CH2), 1.24 (m, CH2), 0.84 (m, CH3). IR(cm−1): 1562, 1638 (−NH−(CO)−NH−), 1738 (−C−(CO)−O−),3352 (−NH−(CO)−NH−).Synthesis of p(1-ILE-6). The polymer was synthesized as

described in the general procedure described above. The polymerwas prepared on a 70 mmol scale (based on monomer), stirred for 2 hafter the second triphosgene addition, and obtained with an 85% yield.1H NMR (500 MHz, DMSO-d6, δ): 6.37 (d, J = 8.9 Hz, NH), 4.04 (m,CH2 and CH), 1.71 (m, CH), 1.55 (s, CH2), 1.34 (m, CH2), 1.12 (m,CH2), 0.84 (m, CH3). IR (cm−1): 1547, 1631 (−NH−(CO)−NH−),1732 (−C−(CO)−O−), 3356 (−NH−(CO)−NH−).Synthesis of p(1-ILE-8). The polymer was synthesized as

described in the general procedure described above. The polymerwas prepared on a 50 mmol scale (based on monomer), stirred for 2 hafter the second triphosgene addition, and obtained with a 92% yield.1H NMR (500 MHz, DMSO-d6, δ): 6.38 (d, J = 6.9 Hz, NH), 4.03 (m,CH2 and CH), 1.71 (m, CH), 1.53 (s, CH2), 1.33 (m, CH2), 1.12 (m,

CH2), 0.86 (m, CH3). IR (cm−1): 1547, 1629 (−NH−(CO) − NH−),1736 (−C−(CO)−O−), 3360 (−NH−(CO)−NH−).

Synthesis of p(1-ILE-10). The polymer was synthesized asdescribed in the general procedure described above. The polymerwas prepared on a 40 mmol scale (based on monomer), stirred for 20h after the second triphosgene addition, and obtained with a 91% yield.1H NMR (500 MHz, DMSO-d6, δ): 6.38 (d, J = 8.6 Hz, NH), 4.05 (m,CH2 and CH), 1.70 (m, CH), 1.49 (s, CH2), 1.41 (m, CH2), 1.13 (m,CH2), 0.83 (m, CH3). IR (cm−1): 1552, 1631 (−NH−(CO)−NH−),1738 (−C−(CO)−O−), 3356 (−NH−(CO)−NH−).

Synthesis of p(1-ILE-12). The polymer was synthesized asdescribed in the general procedure described above. The polymerwas prepared on a 30 mmol scale (based on monomer), stirred for 20h after the second triphosgene addition, and obtained with an 89%yield. 1H NMR (500 MHz, DMSO-d6, δ): 6.43 (d, J = 7.6 Hz, NH),4.06 (m, CH2 and CH), 1.71 (m, CH), 1.53 (s, CH2), 1.29 (m, CH2),0.86 (m, CH3). IR (cm−1): 1554, 1631 (−NH−(CO)−NH−), 1738(−C−(CO)−O−), 3356 (−NH−(CO)−NH−).

Figure 1. Synthesis and nomenclature of amino acid-based poly(ester urea)s. Diamine monomers are prepared by Fischer esterification betweenvarious amino acids and diols with varying carbon chain lengths. Monomers are polymerized with triphosgene in an interfacial polymerization.Nomenclature is based on the amino acid and diol from which its structure is derived.

Table 1. Molecular Mass, Thermal Properties, and Shape Memory Performance Characterization for PEUs

aTg from E″. bTtrans determined from the onset of shape recovery in the third cycle. cFixity and recovery parameters reported for the third cycle.dFrom the third cycle. eTg determined from DSC, shape memory performance was not able to be characterized due to sample failure upon anyappreciable sample extension. fTd = 50 °C. gFixity and recovery parameters reported for the second cycle, samples failed in third cycle. hTd = 70 °C.iTd = 81 °C. Values with errors are an average of three runs ± one standard deviation.

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4303

Shape Memory Characterization. Rectangular test specimens(25 × 5 × 0.5 mm) were prepared by compression molding. ABA-based PEU test specimens were annealed at their Tg for 72 h to ensurecomplete crystallization was achieved (only the ABA-based PEUsexhibited this behavior). Cyclic thermomechanical testing wasconducted using a DMA Q800 instrument. Testing was completedin controlled force mode with heating and cooling rates of 10 °Cmin−1. Dual-shape memory testing was completed with threesequential runs, repeated three times with independent test specimens.The deformation temperature (Td) was equal to the Tg determinedfrom the peak in loss modulus (unless otherwise noted). Triple-shapememory testing was completed with one cycle, repeated three timeswith independent test specimens. The fixity and recovery parametersfor dual- and triple-shape memory were calculated using the standardshape memory equations previously published.15 Temperaturememory effect testing was completed in duplicate. The plot of strainrecovery rate in Figure 5C was obtained by fitting the plot of recoveryparameter vs time with a sigmoidal function and taking its derivative.Preparation of Specimens for Visual Demonstrations. Thin

films were prepared by compression molding of powdered PEU usinga stainless steel mold sandwiched between Kapton film and metalplates. Salt leached scaffolds were prepared by first adding aconcentrated polymer solution (50% w/v of DMF) into a 0.5 cmthick layer of NaCl (sieved to a size of 250−400 μm) in a 6 cmaluminum dish. The samples were air-dried for 72 h and then vacuum-dried for 168 h to remove any residual solvent. The polymer−saltcomposites were then soaked in deionized water (the water waschanged every 12 h, and a AgNO3 solution was used to check for thepresence of NaCl until no white precipitation was found) to leach outthe salt and provide the desired scaffold. The tube construct wasprepared via melt extrusion using powdered PEU.

■ RESULTS AND DISCUSSIONPEUs composed of varying amino acids and diol chain lengthswere prepared in a two-step synthesis (Figure 1).22−24 Diaminemonomers were prepared via Fischer esterification of aminoacids and a diol with a chain length of 6, 8, 10, or 12 methyleneunits. The yields of the monomers ranged from 67 to 95%.Polymers were prepared via an interfacial step-growth polymer-ization between the monomers and triphosgene, with yieldsranging from 70 to 97%. The Mn and ĐM of PEUs ranged from12 to 70 kDa and 1.4 to 3.0, respectively (see Table 1).Dispersities significantly lower than 2.0 can likely be attributed

to fractionation during postpolymerization precipitation ofpolymers. The p(1-ALA-6) and p(1-ALA-8) PEUs wereobtained with lower molecular weights (12 and 14 kDa,respectively, likely due to poor solubility of the polymers duringpolymerization), prohibiting the preparation of sample speci-mens for shape memory characterization.The thermal properties of the PEUs were studied by DMA.

Of particular interest was the Tg, which indicates the minimumtemperature for shape programming and recovery, and thestorage modulus (E′) at room and physiological temperatures,which provides an indication of the stiffness of the material. Inaccord with other studies, the Tgs across all PEUs increasedwith decreasing diol chain length,22−24 with Tg values rangingfrom 21 to 73 °C, and providing a set of materials whose shapechanges could be triggered at or near physiological temper-atures. The modulus was shown to increase with decreasingdiol chain length (Figure 2). The PHE-based PEUs showed thehighest modulus values, likely due to π−π stacking between thearomatic side chains. The ABA-, VAL-, and ILE-based PEUsgenerally showed the same modulus trends, with the exceptions(including the ALA-based PEUs) being attributed to enhancedcrystallinity (see below). Overall, these results illustrate thebroad tunability of the thermomechanical response of thesematerials. By simply changing the amino acid and diol, the Tgcan be tuned over a range of nearly 55 °C and the modulusover a range of nearly 2.5 GPa.DMA curves are a good indicator of the thermal shape

memory behavior of a material.18 The drop in modulus at theTg provides an indication of the extent of chain mobility (with alarge modulus drop indicating that a material can be easilyprogrammed into a temporary shape), and a rubbery plateauabove the Tg indicates the presence of a network capable ofmaintaining shape recovery. The decreasing slope of theplateau, prior to melt/failure, is likely a result of the dynamicnature of the hydrogen bonding (i.e., some chain slippageoccurs). While all PEUs showed a rubbery plateau, themagnitude of the modulus drop did have some variancebetween polymers. For example, p(1-PHE-6) showed a smallmodulus drop (ca. 1 order of magnitude, Figure 3A) while theremaining polymers in the PHE series showed a large drop (at

Figure 2. Plots of storage modulus vs Tg at (A) 23 °C and (B) 37 °C for all PEUs in this study. The storage modulus values were determined fromthe DMA temperature sweep experiment (e.g., Figure 3A) and are equal to the modulus values at the indicated temperatures. The Tg and modulusgenerally increase with decreasing diol length length for a given amino acid. The plot demonstrates the broad tunability of PEUs by changing theamino acid or diol component of the polymer.

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4304

least 2 orders in magnitude). We attributed the behavior of p(1-PHE-6) to the increased crystallinity of the material (asapparent from the XRD curves of PHE-based PEUs, Figure3C). We found this trend to hold true for all other PEUs andthat only those with an increased degree of crystallinity (p(1-ALA-10), p(1-ALA-12), p(1-ABA-6), and p(1-ABA-8)) ex-hibited a small modulus drop (this is especially apparent whencomparing PEUs with a consistent diol length, Figure 3D−F).These results suggested that all of the PEUs, regardless ofamino acid or diol, should exhibit shape memory behavior.However, the materials with a small modulus drop are likely tohave diminished performance.The dual-shape memory properties of PEUs were studied by

cyclic thermomechanical testing with a DMA. A single cycleconsists of programming a temporary shape and recovering thepermanent shape. The temporary shape was programmed byelongating the sample at a Td equal to the Tg of the material (asdetermined from the peak in the plot of loss modulus vstemperature), cooling below the Tg, and removing the tensileload. The permanent shape was recovered by simply reheatingthe material above the Tg in the absence of a tensile load. Threeback-to-back cycles were completed to remove the effects of the

thermal processing and to demonstrate the fidelity of shapememory performance (Figure 4A). The strain fixity andrecovery parameters, Rf and Rr, respectively, are used tocharacterize the efficacy of the programming and recoverysteps.15 We observed that the VAL-, ILE-, and PHE-basedPEUs with a large drop in storage modulus had high Rf (>90%)and Rr (>95%) values. In these materials the incorporatedamino acid or diol had little effect on those parameters (Table1). The PEUs with a small modulus drop demonstrateddecreases in Rf (79−90%) and Rr (85−95%) values; however,the largest decrease in performance was the strain achievableduring the shape programming step. The maximum achievablestrain for these materials was less than 10%, even with increasedTd temperatures (applying higher strains led to failure of thematerials), in contrast to achievable strains of greater than 90%for PEUs with a large modulus drop. The ABA-based PEUswith a large modulus drop (p(1-ABA-10) and p(1-ABA-12) hadhigh Rf values but decreased Rr values. This could be attributedto the low storage modulus, manifested in the material storingless energy for shape actuation. The molecular weight of PEUsis not expected to have a significant effect on the shape memoryproperties of PEUs, unless it were to drop below the

Figure 3. (A) DMA single frequency (1 Hz) temperature sweep (3 °C/min) experiment: storage modulus plots for the PHE-based PEUs, showing adrop in modulus at the Tg, followed by a rubbery plateau. (B) Tan δ plots for PHE-based PEUs. (C) XRD plots for the PHE-based PEUs showingincreased crystallinity for p(1-PHE-6). (D) DMA storage modulus plots for the 1,10-decanediol-based PEUs showing the variability in Tg, anddecreased modulus drop above the Tg for the ALA-based PEU. (E) Tan δ plots for the 1,10-decanediol-based PEUs. (F) XRD plots for the 1,10-decananediol-based PEUs showing the increased crystallinity for p(1-ALA-10).

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4305

entanglement molecular weight (decreasing the degree ofphysical cross-linking). For instance, a 33 and 99 kDa p(1-PHE-10) were compared head-to-head and shown to have Rfvalues of 97 and 99%, respectively, and both had Rr values of99%. It is, however, expected that the molecular weight wouldhave a small impact on the Tg.

31,32 As shape fixing and recoveryare based on the Tg, we observed that the transitiontemperature (Ttrans, the onset of shape recovery; see Table 1)trended well with the Tg, as expected. Overall, PEUs showedgood to excellent shape memory performance across all aminoacids and diol combinations studied, except when the materialswere crystalline. This is a significant finding as being able totune the mechanical properties without significantly impactingthe shape memory performance is an advantageous feature ofthese materials.PEUs exhibit a broad Tg range (>20 °C), which is a common

feature of many multishape memory polymers.18 We previouslydemonstrated that p(1-VAL-8) had good triple-shape memoryperformance.30 We further subjected p(1-ILE-8) and p(1-PHE-8) to thermomechanical testing to determine how the aminoacid might affect the SMP performance. In triple-shape memorytesting, two temporary shapes are programmed at two differentTds above the Tg, followed by recovery of the first temporaryshape and permanent shape by heating at the two Tdsconsecutively (Figure 4B). Each polymer showed good Rfand Rr parameters (Table 2). During the recovery of the firstshape of p(1-PHE-8), a Rr > 100% was obtained due to partialrecovery of the second temporary shape. This is primarily due

to having used only a 20 °C difference between Tds (asopposed to 30 °C for VAL and ILE). This adjustment wasrequired as the width of the thermal transition for p(1-PHE-8)is narrower than the other two polymers (see Figure S1 in theSupporting Information).Another common feature of SMPs with broad thermal

transitions is that they often have temperature memoryproperties.15,18,33,34 The temperature memory effect (TME)was explored in p(1-VAL-8), a representative PEU. We firstlooked at the stress-based TME. In this test, multiple sampleswere heated at various Tds and then stretched to 100% strain.The samples were then held in an isostrain state and heated,and the recovery stress was monitored as a function oftemperature. We observed a steady increase in the peakrecovery stress (Figure 5), indicating the material had“memorized” the Td. With the exception of the tests completedusing a Td of 30 °C, which is below the Tg of the material, thepeak recovery stress was within ca. 5 °C from the Td, which isconsistent with other temperature memory materials.33 Toconfirm that the material would also show a strain-based TME,multiple samples were heated at various Tds and then stretchedto 100% strain; then the samples were heated in the absence ofa tensile load, and the recovery of the permanent shape wasmonitored as a function of time and temperature. We observedthat the recovery profile of the material shifted to highertemperatures with increasing Td. The peak in recovery rate alsotrended toward higher temperatures as higher Tds were used.This confirmed that p(1-VAL-8) exhibited both stress- andstrain-based temperature memory properties. Given that thediol and amino generally had minimal effect on the dual- andtriple-shape memory properties of PEUs, it would follow thatthe other PEUs would also exhibit the TME.To explore the shape memory behavior of PEUs in

constructs that might be relevant to biomedical applications,several specimens were prepared to explore the effect of shapechanges outside the realm of pure tensile elongation. Forinstance, we demonstrated multidirectional unfolding of a boxinto a thin film (Figure 6A). The same specimen could also bereprogrammed to fold from the flat film to the box. Shaperecovery, in both cases, occurred in 3 s or less. We next lookedat the surface topology of a thin film, which was altered andrecovered (Figure 6C). Shape changes of larger constructs suchas a salt-leached scaffold or a thick cylindrical specimen couldalso be achieved. These demonstrations show the versatility ofthe shape memory effect in PEUs.

■ CONCLUSIONSNew PEUs with differing amino acid and diol compositionswere prepared and shown to have unique and tunable

Figure 4. (A) Representative example of three sequential dual-shapememory cycles for p(1-PHE-8). The permanent shape was heated to58 °C and stretched by applying a tensile load, followed by cooling thesample and removing the load to program the temporary shape. Thepermanent shape was recovered by heating to 68 °C in the absence ofa tensile load. (B) Representative example of a triple-shape memorycycle for p(1-PHE-8). The first and second temporary shapes wereprogrammed at 70 and 50 °C, respectively. Recovery of the firsttemporary shape and the permanent shape was observed upon heatingto 50 and 70 °C, respectively, in the absence of a tensile load.

Table 2. Strain Fixity and Recovery from Triple-ShapeMemory Cycles of the 1,8-Octanediol-Based PEUs

polymerRf(I→II)a

(%)Rf(II→III)

(%)Rr(III→II)

(%)Rr(II→I)(%)

p(1-VAL-8)b 98.8 ± 0.8 96.9 ± 0.6 98.6 ± 1.3 93.3 ± 2.2p(1-ILE-8)b 98.8 ± 0.3 85.1 ± 7.2 91.9 ± 1.1 90.0 ± 1.9p(1-PHE-8)c 99.1 ± 0.3 97.8 ± 0.8 153 ± 28 94.6 ± 2.1aFixity and recovery parameters reported for the third cycle. bTd1 = 70°C, Td2 = 40 °C. cTd1 = 70 °C, Td2 = 50 °C. Roman numeral Icorresponds to the permanent shape, II to the first temporary shape,and III to the second temporary shape. Values are an average of threeruns ± one standard deviation.

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4306

thermomechanical properties. The Tgs of PEUs covered a rangeof nearly 55 °C, and the storage moduli, a range of nearly 2.5GPa. We found that most of the studied PEUs demonstratedshape memory behavior with excellent shape memorycharacteristics. The main inhibitor of SMP performance wasfound to be associated with the appearance of a crystallinemorphology, leading to a reduction in chain mobility and verylow achievable strains. Because of the broad glass transitiontemperature of these materials, we found them to exhibit triple-shape memory and temperature memory behavior. The overallshape memory performance together with tunable mechanicalproperties and previous demonstrations of biocompatibilitymake PEUs prime candidates for use as SMPs in variousbiomedical applications.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.macro-mol.7b00680

Figure S1, SEC plots, IR spectra, and 1H/13C NMRspectra for all new monomers and polymers (PDF)

■ AUTHOR INFORMATIONCorresponding Author*(M.L.B.) E-mail becker@uakron.edu.ORCIDAndrey V. Dobrynin: 0000-0002-6484-7409Matthew L. Becker: 0000-0003-4089-6916NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis work was supported by the Ohio Department ofDevelopment’s Innovation Platform Program and the Bio-materials Division of the National Science Foundation (DMR-1507420). M.L.B. acknowledges support from the W. Gerald

Figure 5. (A) Stress generated by p(1-VAL-8) when heated afterprogramming a temporary shape with 100% strain. The strain is fixed,and the temperature is increased from 10 °C at a rate of 5 °C/minuntil sample failure occurs. Each plot corresponds to the Td used toprogram the temporary shape. The peak in recovery stress increaseswith increasing Td, demonstrating stress-based TME properties. (B)Strain recovery profiles of p(1-VAL-8) specimens that wereprogrammed to 100% strain. Samples were heated at 5 °C/min to90 °C. (C) Strain recovery rates obtained from the recovery profiles in(B). The peak in recovery rate increases with increasing Td,demonstrating strain-based TME properties. The arrow on the x-axisindicates the temperature corresponding to the Tg of the material.

Figure 6. (A) Demonstration of dual-shape memory actuation in ap(1-ILE-8) thin film. A ca. 1 cm3 cube is programmed to open to a flatfilm upon heating. The shape change occurred over ca. 2 s. (B) Thesame p(1-ILE-8) thin film programmed to close into the shape of a 1cm3 cube upon heating. The shape change occurred over ca. 3 s. (C)Changing surface roughness of a p(1-ILE-6) thin film. The film in i isthe starting polymer, ii is after applying roughness to the sample, andiii is after heating the sample, showing recovery of the smooth surface.(D) Bending of a p(1-VAL-8) salt leached scaffold. The scaffold in i isprior to deformation, ii is after bending, and iii is after heating thescaffold. (E) Shape change of a large tube construct of p(1-PHE-6).The tube in i is the original shape, ii is after deformation, and iii is therecovered shape. Shape recovery was triggered by gentle heating with aheat gun. All scale bars represent 1 cm.

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4307

Austen Professor in Polymer Science and Polymer Engineeringendowed by the Knight Foundation.

■ ABBREVIATIONS

ABA, L-2-aminobutyric acid; ALA, L-alanine; ĐM, molecularmass distribution/dispersity; DMA, dynamic mechanicalanalysis; DMF, dimethylformamide; DMSO, dimethyl sulf-oxide; DSC, differential scanning calorimetry; Mn, number-average molecular weight; PEU, poly(ester urea); PHE, L-phenylalanine; ILE, L-isoleucine; Rf, fixity parameter; Rr,recovery parameter; SEC, size exclusion chromatography;SMP, shape memory polymer; Td, deformation temperature;Tg, glass transition temperature; TME, temperature memoryeffect; TsOH, p-toluenesulfonic acid; Ttrans, transition temper-ature; VAL, L-valine; XRD, X-ray diffraction.

■ REFERENCES(1) Hardy, J. G.; Palma, M.; Wind, S. J.; Biggs, M. J. ResponsiveBiomaterials: Advances in Materials Based on Shape-MemoryPolymers. Adv. Mater. 2016, 28, 5717−5724.(2) Fan, X.; Chung, J. Y.; Lim, Y. X.; Li, Z.; Loh, X. J. Review ofAdaptive Programmable Materials and Their Bioapplications. ACSAppl. Mater. Interfaces 2016, 8, 33351−33370.(3) Chan, B. Q. Y.; Low, Z. W. K.; Heng, S. J. W.; Chan, S. Y.; Owh,C.; Loh, X. J. Recent Advances in Shape Memory Soft Materials forBiomedical Applications. ACS Appl. Mater. Interfaces 2016, 8, 10070−10087.(4) Hager, M. D.; Bode, S.; Weber, C.; Schubert, U. S. ShapeMemory Polymers: Past, Present and Future Developments. Prog.Polym. Sci. 2015, 49−50, 3−33.(5) Ebara, M. Shape-Memory Surfaces for Cell Mechanobiology. Sci.Technol. Adv. Mater. 2015, 16, 014804.(6) Wong, Y.; Kong, J.; Widjaja, L. K.; Venkatraman, S. S. BiomedicalApplications of Shape-Memory Polymers: How Practically Useful areThey? Sci. China: Chem. 2014, 57, 476−489.(7) Lendlein, A.; Langer, R. Biodegradable, Elastic Shape-MemoryPolymers for Potential Biomedical Applications. Science 2002, 296,1673−1676.(8) Buckley, P. R.; McKinley, G. H.; Wilson, T. S.; Small, W., IV;Benett, W. J.; Bearinger, J. P.; McElfresh, M. W.; Maitland, D. J.Inductively Heated Shape Memory Polymer for the MagneticActuation of Medical Devices. IEEE Trans. Biomed. Eng. 2006, 53,2075−2083.(9) Bao, M.; Lou, X.; Zhou, Q.; Dong, W.; Yuan, H.; Zhang, Y.Electrospun Biomimetic Fibrous Scaffold from Shape MemoryPolymer of PDLLA-co-TMC for Bone Tissue Engineering. ACSAppl. Mater. Interfaces 2014, 6, 2611−2621.(10) Ebara, M.; Uto, K.; Idota, N.; Hoffman, J. M.; Aoyagi, T. Shape-Memory Surface with Dynamically Tunable Nano-Geometry Activatedby Body Heat. Adv. Mater. 2012, 24, 273−278.(11) Wischke, C.; Neffe, A. T.; Steuer, S.; Lendlein, A. Evaluation of aDegradable Shape-Memory Polymer Network as Matrix for ControlledDrug Release. J. Controlled Release 2009, 138, 243−250.(12) Lobo, P. S.; Almeida, J.; Guerreiro, L. Shape Memory AlloysBehaviour: A Review. Procedia Eng. 2015, 114, 776−783.(13) Otsuka, K.; Ren, X. Recent Developments in the Research ofShape Memory Alloys. Intermetallics 1999, 7, 511−528.(14) Pilate, F.; Toncheva, A.; Dubois, P.; Raquez, J.-M. Shape-Memory Polymers for Multiple Applications in the Materials World.Eur. Polym. J. 2016, 80, 268−294.(15) Zhao, Q.; Qi, H. J.; Xie, T. Recent Progress in Shape MemoryPolymer: New Behavior, Enabling Materials, and MechanisticUnderstanding. Prog. Polym. Sci. 2015, 49−50, 79−120.(16) Berg, G. J.; McBride, M. K.; Wang, C.; Bowman, C. N. NewDirections in the Chemistry of Shape Memory Polymers. Polymer2014, 55, 5849−5872.

(17) Huang, W. M.; Zhao, Y.; Wang, C. C.; Ding, Z.; Purnawali, H.;Tang, C.; Zhang, J. L. Thermo/chemo-Responsive Shape MemoryEffect in Polymers: A Sketch of Working Mechanisms, Fundamentalsand Optimization. J. Polym. Res. 2012, 19, 9952.(18) Xie, T. Recent Advances in Polymer Shape Memory. Polymer2011, 52, 4985−5000.(19) Balk, M.; Behl, M.; Wischke, C.; Zotzmann, J.; Lendlein, A.Recent Advances in Degradable Lactide-Based Shape-MemoryPolymers. Adv. Drug Delivery Rev. 2016, 107, 136−152.(20) Karger-Kocsis, J.; Keki, S. Biodegradable Polyester-Based ShapeMemory Polymers: Concepts of (Supra)Molecular Architecturing.eXPRESS Polym. Lett. 2014, 8, 397−412.(21) Stakleff, K. S.; Lin, F.; Smith Callahan, L. A.; Wade, M. B.;Esterle, A.; Miller, J.; Graham, M.; Becker, M. L. Resorbable, AminoAcid-Based Poly(ester urea)s Crosslinked with Osteogenic GrowthPeptide with Enhanced Mechanical Properties and Bioactivity. ActaBiomater. 2013, 9, 5132−5142.(22) Childers, E. P.; Peterson, G. I.; Ellenberger, A. B.; Domino, K.;Seifert, G. V.; Becker, M. L. Adhesion of Blood Plasma Proteins andPlatelet-rich Plasma on l-Valine-Based Poly(ester urea). Biomacromo-lecules 2016, 17, 3396−3403.(23) Gao, Y.; Childers, E. P.; Becker, M. L. L-Leucine-BasedPoly(ester urea)s for Vascular Tissue Engineering. ACS Biomater. Sci.Eng. 2015, 1, 795−804.(24) Yu, J.; Lin, F.; Lin, P.; Gao, Y.; Becker, M. L. Phenylalanine-Based Poly(ester urea): Synthesis, Characterization, and in vitroDegradation. Macromolecules 2014, 47, 121−129.(25) Policastro, G. M.; Lin, F.; Callahan, L. A.; Esterle, A.; Graham,M.; Stakleff, K. S.; Becker, M. L. OGP Functionalized Phenylalanine-Based Poly(ester urea) for Enhancing Osteoinductive Potential ofHuman Mesenchymal Stem Cells. Biomacromolecules 2015, 16, 1358−1371.(26) Li, S.; Yu, J.; Wade, M. B.; Policastro, G. M.; Becker, M. L.Radiopaque, Iodine Functionalized, Phenylalanine-Based Poly(esterurea)s. Biomacromolecules 2015, 16, 615−624.(27) Zhou, J.; Defante, A. P.; Lin, F.; Xu, Y.; Yu, J.; Gao, Y.; Childers,E.; Dhinojwala, A.; Becker, M. L. Adhesion Properties of Catechol-Based Biodegradable Amino Acid-Based Poly(ester urea) CopolymersInspired from Mussel Proteins. Biomacromolecules 2015, 16, 266−274.(28) Lin, F.; Yu, J.; Tang, W.; Zheng, J.; Xie, S.; Becker, M. L.Postelectrospinning “Click” Modification of Degradable Amino Acid-Based Poly(ester urea) Nanofibers. Macromolecules 2013, 46, 9515−9525.(29) Diaz, A.; del Valle, L. J.; Tugushi, D.; Katsarava, R. Puiggali,New Poly(ester urea) Derived from L-Leucine: Electrospun ScaffoldsLoaded with Antibacterial Drugs and Enzymes. J. Mater. Sci. Eng., C2015, 46, 450−462.(30) Peterson, G. I.; Dobrynin, A. V.; Becker, M. L. α-Amino Acid-Based Poly(Ester urea)s as Multishape Memory Polymers forBiomedical Applications. ACS Macro Lett. 2016, 5, 1176−1179.(31) Blanchard, L.-P.; Hesse, J.; Malhotra, S. L. Effect of MolecularWeight on Glass Transition by Differential Scanning Calorimetry. Can.J. Chem. 1974, 52, 3170−3175.(32) Steendam, R.; van Steenbergen, M. J.; Hennink, W. E.; Frijlink,H. W.; Lerk, C. F. Effect of Molecular Weight and Glass Transition onRelaxation and Release Behaviour of Poly(dl-lactic acid) Tablets. J.Controlled Release 2001, 70, 71−82.(33) Xie, T.; Page, K. A.; Eastman, S. A. Strain-Based TemperatureMemory Effect for Nafion and Its Molecular Origins. Adv. Funct.Mater. 2011, 21, 2057−2066.(34) Miaudet, P.; Derre, A.; Maugey, M.; Zakri, C.; Piccione, P. M.;Inoubli, R.; Poulin, P. Shape and Temperature Memory ofNanocomposites with Broadened Glass Transition. Science 2007,318, 1294−1296.

Macromolecules Article

DOI: 10.1021/acs.macromol.7b00680Macromolecules 2017, 50, 4300−4308

4308