High-Temperature Shape Memory Behavior of Semicrystalline Polyamide Thermosets

Transcription

1 This is n open ccess rticle pulished under Cretive Commons Non-Commercil No Derivtive Works (CC-BY-NC-ND) Attriution License, which permits copying nd redistriution of the rticle, nd cretion of dpttions, ll for non-commercil purposes. Cite This: Downloded vi on Decemer 31, 2018 t 13:44:33 (UTC). See for options on how to legitimtely shre pulished rticles. High-Temperture Shpe Memory Behvior of Semicrystlline Polymide Thermosets Ming Li,,, Qingo Gun, nd Theo J. Dingemns*,, Ntionl Engineering Reserch Center for Biotechnology, Nnjing Tech University, Nnjing , Chin Fculty of Aerospce Engineering, Delft University of Technology, Kluyverweg 1, 2629 HS Delft, The Netherlnds Dutch Polymer Institute (DPI), P.O. Box 902, 5600 AX Eindhoven, The Netherlnds Deprtment of Mteril Science nd Engineering, Soochow University, Suzhou , Chin Deprtment of Applied Physicl Sciences, University of North Crolin t Chpel Hill, 1113 Murry Hll, Chpel Hill, North Crolin , United Sttes *S Supporting Informtion ABSTRACT: We hve explored semicrystlline poly- (decmethylene terephthlmide) (PA 10T) sed thermosets s single-component high-temperture (>200 C) shpe memory polymers (SMPs). The PA 10T thermosets were prepred from rective thermoplstic precursors. Rective phenylethynyl (PE) functionlities were either ttched t the chin termini or plced s side groups long the polymer min chin. The shpe fixtion nd recovery performnce of the thermoset films were investigted using rheometer in torsion mode. By controlling the M n of the rective oligomers, or the PE concentrtion of the PE side-group functionlized copolymides, we were le to design dul-shpe memory PA 10T thermosets with rod recovery temperture rnge of C. The thermosets sed on the 1000 g mol 1 rective PE precursor nd the copolymide with 15 mol % PE side groups show the highest fixtion rte (99%) nd recovery rte ( 90%). High temperture triple-shpe memory ehvior cn e chieved s well when we use the melt trnsition (T m 200 C) nd the glss trnsition (T g = 125 C) s the two switches. The recovery rte of the two recovery steps re highly dependent on the crystllinity of the thermosets nd vry within wide rnge of 74% 139% nd 40 82% for the two steps, respectively. Reversile shpe memory events could lso e demonstrted when we perform forwrd nd ckwrd deformtion in triple shpe memory cycle. We lso studied the ngulr recovery velocity s function of temperture, which provides thermokinemtic picture of the shpe recovery process nd helps to progrm for desired shpe memory ehvior. KEYWORDS: semiromtic polymides, semicrystlline thermosets, high-temperture shpe memory, dul-shpe memory, triple-shpe memory, recovery velocity 1. INTRODUCTION Interest in shpe memory polymers (SMPs) hs grown rpidly since the 1980s. 1 5 Typicl pplictions re het-shrink tuing, temperture sensors nd ctutors, iomedicl nd surgicl mterils, nd erospce devices. 6,7 Thermoresponsive SMPs re the most investigted systems tht polymers re le to dopt temporry shpe upon deformtion nd to revert ck to the permnent shpe upon exposure to het. 7,8 They re generlly composed of polymer network to mintin the permnent shpe nd reversile switch responsile for the shpe fixtion nd recovery The glss trnsition temperture (T g ) nd melting temperture (T m ) re the two most importnt therml trnsitions for thermoresponsive SMPs. Polyurethnes, polyesters, nd (methyl)crylte-sed polymer networks hve een investigted s SMPs sed on the glss trnsition. Cross-linked semicrystlline networks or (multi)- lock copolymer systems hve een developed to design T m - sed SMPs, such s polyolefins, polyethers, nd polyesters. 6 However, most SMPs exhiit switching temperture lower thn 100 C, which my not meet the requirements for high temperture erospce, utomotive, or electronic pplictions. Recently, severl exmples of high-temperture (>200 C) SMPs hve een reported Vi nd co-workers synthesized n morphous fluorinted polyimide with shpe recovery temperture of 220 C. 20 This ws the first exmple of single-component high-temperture SMP. However, this fluorinted polyimide cn only e otined s thin films, which excludes the possiility to produce complex shpes using melt processing methods. Weiss et l. hve introduced ionic moieties into poly(ether ether ketone) (PEEK) leding to T g -sed SMP with switching temperture close to 200 C, where the exct switching temperture depends on the metl counterions Received: Mrch 4, 2018 Accepted: April 27, 2018 Pulished: My 9, Americn Chemicl Society 19106

2 Scheme 1. Structures of Precursors vi (A) Rective End-Group Approch nd (B) Rective Side-Group Approch used. 14 In lter puliction, this system ws further developed into T m -sed SMP y incorportion of sodium olete, which displys higher switching temperture of C due to the melting of sodium olete. 15 In contrst to the dul SMPs mentioned ove, triple SMPs feturing two independent temporry shpes were first reported y Lendlein s group. 22 They require two progrmming steps nd show two recovery steps. Triple-shpe memory effects cn e relized in polymer mteril possessing two distinct therml trnsitions 15,22 24 or one single rod trnsition such s rod glss trnsition. 16,25 High temperture triple SMPs sed on single-component thermotropic liquid crystlline poly(esterimide) thermosets hve recently een reported y Gun et l. The polymer shows two glss trnsition tempertures t 120 nd 200 C, oth of which cn e used s switches for triple SMPs. 21 In previous pulictions we reported on the synthesis nd (thermo)mechnicl properties of semiromtic polymide thermosets sed on poly(decmethylene terephthlmide) (PA 10T). 26,27 We showed tht this clss of polymers cn e processed into semicrystlline thermoset films where the degree of crystllinity cn e controlled. The melting/crystlliztion of the crystlline phse nd the covlent network cn e used s the high-temperture switch nd permnent scffold, respectively, for dul shpe memory effect (SME). More sophisticted triple nd one-wy reversile SME cn e designed with the glss trnsition s the second switch. To the est of our knowledge, this is the first demonstrtion of high-temperture SMP sed on single-component semicrystlline polymide thermoset. 2. EXPERIMENTAL SECTION 2.1. Mterils. The syntheses of phenylethynyl (PE) end-cpped PA 10T oligomers (Scheme 1A) nd PE side-group-functionlized copolymides (Scheme 1B) hve een descried in detil elsewhere. 26,27 In order to otin covlently cross-linked polymide network, the rective precursors were thermlly cured. A stndrd melt compression technique ws used to prepre the thermoset films ccording to the following procedure: The precursor polymers were ground into fine powder using mortr nd pestle. The powder ws plced etween two metl pltes lined with Kpton film, nd this stck ws consolidted in Joos hot-press using predetermined temperture progrm nd 5 kn force. The temperture progrm ws set to het to 350 C t5 C min 1, hold for 15 min, nd cooled to 50 C t20 C min 1. The otined cured films were used for further chrcteriztion. The film smples prepred from PA 10T rective oligomers (M n of 1000 nd 3000 g mol 1 ) re denoted s PE-1K nd PE-3K. TPE-5, TPE-10, nd TPE-15 represent the film smples prepred from rective TPE-copolymides with 5, 10, nd 15 mol % of rective TPE comonomer, respectively. A PA 10T thermoplstic smple (M n = 7500 g mol 1 ) ws used s the reference in this pper nd denoted s Ref. The 1 H NMR spectr of the synthesized precursors (Supporting Informtion, Figure S1) confirms tht the concentrtion of rective functionlities, either s end-cps or s side groups, is consistent with the feed rtio of the monomers Chrcteriztion. 1 H NMR spectr were recorded on 400 MHz Bruker WM-400 t 25 C using trifluorocetic cid-d s solvent. DSC ws conducted on PerkinElmer Spphire DSC under nitrogen tmosphere t heting/cooling rte of 20 C min 1. DMTA ws performed on PerkinElmer Dimond DMTA with film smples ( mm thick) t heting rte of 2 C min 1 under nitrogen tmosphere. Dt were collected t frequency of 1 Hz. The SME ws chrcterized in cyclic torsion mode s illustrted in Figure 1. Compred to the trditionl extension or ending tests, 28 torsion tests involve nonhomogeneous strins nd stresses in the cross section of rectngulr r nd enle the SMPs to rech lrge deformtions with moderte strins, which re elieved to e more representtive of prcticl pplictions of SMPs The shpe progrmming nd recovery cycles were performed using Thermofisher Hke MARS III rheometer equipped with solid clmp geometry under N 2 tmosphere. Rectngulr thin films with width of 2.5 mm nd thickness of 0.25 mm were loded into the setup with length of 15 mm etween the clmps. The smples were deformed in torsion mode t constnt strin rte of 0.1% s 1 equivlent to rottion speed of 3.4 s 1. The procedure of one cycle for dul SME test includes the following steps: (1) het the smple to the progrmming temperture 19107

3 Figure 2. DSC heting scns (first het) of (A) reference polymer (Ref), PE-3K, nd PE-1K nd (B) reference polymer (Ref), TPE-5, TPE-10, nd TPE-15. The sterisk refers to melting peks (N 2 tmosphere nd heting rte of 20 C min 1 ). Tle 1. Therml Properties of the Thermoplstic Reference Polymer (Ref) nd the Cured Thermoset Smples Figure 1. Illustrtion of shpe deformtion, fixtion, nd recovery cycle of the dul SME in torsion mode. (T prog ); (2) rotte the smple to the predetermined ngle (ϕ d ); (3) keep the ngle constnt, cool to the fixtion temperture (T f )t10 C min 1 nd stilize for 10 min; (4) remove the stress; (5) het the smple to T prog t 10 C min 1 in stress-free condition followed y n isotherml hold t T prog for 10 min to stilize. This cycle ws repeted multiple times to chrcterize the reproduciility of the shpe memory performnce. The rottion ngle of the smple ws monitored nd recorded during the whole test. For triple SME mesurements, the smple ws consecutively deformed in two steps in one progrmming cycle. Two different T prog s were used, one eing ove T m nd the other etween T g nd T m. 3. RESULTS AND DISCUSSION 3.1. Melting nd Thermomechnicl Properties. In generl, either the T g or the T m trnsition of polymer cn e pplied to trigger dul-shpe memory event; however, to chieve high-temperture triple-shpe memory ehvior in single-component polymer is chllenging, s two high-temperture temporry networks with distinct ruery plteus re required. 15,22 24 One solution to relize this type of structure is to introduce moderte cross-links into high-t g semicrystlline polymer. In our previous work, we hve explored semiromtic polymide, PA 10T (T g = 125 C, T m = 315 C), s the se polymer to prepre semicrystlline polymide thermosets. Phenylethynyl (PE) shows high cure-temperture of C nd thus is le to provide proper processing window for the PA 10T-sed precursors. Therefore, we hve synthesized rective PE-sed end-cp nd rective PEsed comonomer, which were copolymerized with terephthlic cid nd 1,10-diminodecne to yield curle PE end-cpped oligomers nd PE side-group functionlized copolymers. The polymide thermosets were otined fter susequent therml cure t temperture of 350 C for 15 min. PE-1K nd PE-3K refer to the cured end-cpped oligomers with M n = 1000 nd 3000 g mol 1, respectively. TPE-5, TPE-10, nd TPE- 15 represent the cured side-group functionlized copolymides with 5, 10, nd 15 mol % PE side groups, respectively. The melting curves nd properties of thermoplstic PA 10T reference polymer (Ref) nd the resultnt polymide thermosets re shown in Figure 2 nd Tle 1. The thermoset smples fter cure re semicrystlline showing T m of smple T m ΔH m (J g 1 ) T g cross-linking density c (kmol m 3 ) Ref PE-3K PE-1K TPE TPE TPE T m refers to the mx of the melting pek s oserved in DSC experiments. T g refers to the mx of E s oserved in DMTA experiments. c Cross-linking density (υ) ws clculted using υ =, RT where E is the storge modulus of cured films t 350 C (T = 623 K) nd R the universl gs constnt (8.314 J K 1 mol 1 ). C nd ΔH m of 8 33 J g 1. These vlues re much lower thn the T m nd the ΔH m of the reference polymer (Ref) (318 C nd 82 J g 1 ), which mens the crystllizility of the polymer chins in the thermosets is strongly suppressed. Figure 3 shows the thermomechnicl ehvior of the reference polymer (Ref) nd the thermosets. The Ref film exhiits T g t 127 C, ut this film fils t 297 C ecuse it hs reched the melting point (T m ). In contrst to the reference polymer (Ref), the thermoset smples show two plteu regions (T g T m nd >T m ) in the DMTA profiles. The second plteu of the thermosets is stle up to 400 C, which confirms the presence of network structure. The T g s of oth thermoset smples remin virtully unchnged ( C) when compred to tht of the reference polymer (Ref). The moleculr structure of the resultnt semicrystlline thermosets is depicted in Figure 4. PE groups comine t the cure temperture resulting in cross-links with multiple ftercure chemicl structures depending on cure temperture, time, nd PE concentrtion. We ttempted to investigte the cure chemistry using Rmn nd FTIR spectroscopy. 33,34 However, Rmn spectroscopy filed due to strong fluorescence ckground, nd FTIR cnnot detect the cetylene ond of PE ecuse of IR insensitivity. Hence, direct evidence is not ville to confirm the chemicl chnge of the PE functionlities. The DMTA results of the cured smples show ruery plteu ove T m, nd this plteu is stle up to 400 C, which confirms tht cross-linking hs tken plce during cure. The two-plteu therml ehvior enles tripleshpe memory properties y tking oth T g nd T m trnsitions s the switches. The T m trnsition cn e solely used s switch 19108

4 Figure 3. DMTA of (A) reference polymer (Ref), PE-1K, nd PE-3K films nd (B) reference polymer (Ref), TPE-5, TPE-10, nd TPE-15 films. Heting rte 2 C min 1 under N 2 tmosphere nd frequency of 1 Hz. The insets show E t the glss trnsition temperture. Figure 4. Moleculr representtion of the semicrystlline PA 10T thermosets. Crystlline polymer (drk lue) is emedded in n morphous mtrix (light lue), nd the red dots represent covlent cross-link points. to design high-temperture (>200 C) dul-shpe memory polymer Dul-Shpe Memory Behvior. Becuse of the semicrystlline nture of the synthesized polymide thermosets, their melting nd crystlliztion processes cn e used s the thermoresponsive switch for dul-shpe memory ehvior. Three consecutive deformtion, fixtion, nd recovery cycles in torsion mode were conducted for ech smple to test the shpe memory performnce over multiple cycles. Shpe fixtion rte ( ) nd shpe recovery rte ( ) re the most importnt prmeters to chrcterize the shpe memory performnce. 35 descries how ccurtely the temporry shpe cn e fixed, nd quntifies the ility of the polymer to memorize its permnent shpe. When performing the mesurements in torsion mode, nd cn e clculted using eqs 1 nd 2. φf = 100% φd (1) φ φ d r = 100% φd (2) where φ d, φ f, nd φ r denote the rottionl ngle fter deformtion, t the fixed temporry shpe t T f, nd fter Figure 5. Three consecutive dul-shpe memory cycles for (A) reference polymer (Ref), (B) PE-3K, nd (C) PE-1K. Shpe recovery velocity s function of temperture in the second cycle for (D) reference polymer (Ref), (E) PE-3K, nd (F) PE-1K. Cooling/heting rte 10 C min 1 nd N 2 tmosphere

5 recovery, respectively. The instntneous recovery velocity V r cn e clculted s the time derivtive of the ngle s shown in eq 3. = φ Vr t Plotting V r ginst temperture revels the temperture rnge corresponding to the shpe recovery process. 21 This provides cler thermokinemtic view of the shpe recovery nd helps to progrm for desired SME. The thermoplstic reference film (ref) cnnot e deformed ove the melting temperture ecuse of the sence of ruery plteu, thus the glss trnsition ws used s the switch for the SME. As shown in Figure 5A, the smple ws deformed t 220 C, which lies etween T g (127 C) nd T m (318 C), nd the shpe ws fixed t 100 C where the smple is in glssy stte. Therefore, the crystlline domins ct s the scffold, nd the deformtion of the polymer tkes plce in the morphous stte. In contrst to the reference polymer (Ref), PE-3K nd PE- 1K thermoset films re stle up to 400 C in DMTA experiments nd exhiit two plteu regions (T g T m nd >T m ), s shown in Figure 3. This llows for deforming the film smples in the second plteu region (>T m ). Figures 5B nd 5C show tht PE-3K nd PE-1K were deformed t 315 nd 275 C, respectively, which is out 30 C higher thn their T m s (283 nd 244 C). At this temperture the covlent network cts s the scffold, nd the temporry shpe is fixed y crystlliztion nd vitrifiction t the fixtion step. At the recovery step, the ngle shows slight decrese when the smple psses through the glss trnsition, which is ssocited with the relese of the stress trpped in the moile morphous region. A susequent slow increse in the ngle is oserved, which cn e explined y therml expnsion. 36 When the smple strts melting, the ngle recovers rpidly due to the fst relese of the trpped stress. The shpe-memory performnce of the first few cycles is usully not representtive, s cn e seen in Figure 5B,C. This is generlly ttriuted to residul strin from the processing history of the smple. 13,35 Tle 2 shows the fixtion nd Tle 2. Dul-Shpe Fixtion nd Recovery Results of the Reference Thermoplstic Polymer (Ref), PE-3K, nd PE-1K Thermoset Films cycle 1 cycle 2 cycle 3 smple T prog T r Ref PE-3K PE-1K , T prog refers to the progrmming temperture. T r refers to the temperture t the mximum recovery velocity. recovery results of the first three cycles of the reference thermoplstic polymer (Ref), the PE-3K, nd PE-1K thermoset films. The PE-3K film shows very low of 36% in the first cycle compred to the following cycles ( = 72% nd 76%), wheres the of the PE-1K film shows medium chnge over cycles ( = 81%, 90%, nd 88%). Such significnt difference cn originte from the different cross-linking densities of these two smples. PE-3K hs lower cross-linking density nd (3) therefore longer polymer chins etween cross-links, which is more likely to store residul strin. The recovery of oth smples cnnot rech 100% ecuse the irreversile chinsegment orienttion nd the relxtion effect in the polymer network prtly dissipte the stored entropy. 20 The shpe recovery velocity of the second cycle is plotted s function of temperture s shown in Figure 5D F. The reference polymer (Ref) shows low velocity (<20 min 1 ) through the recovery process ecuse the shpe recovery is triggered y the ctivtion of the morphous chin segments, which re strongly restricted y the crystlline domins. A low recovery velocity is generlly oserved in T g -sed shpememory polymers. 8 In contrst, PE-3K shows quick recovery in the temperture rnge of C nd reches mximum V r of 60 min 1 t 285 C, which is close to its T m (283 C). The shpe recovery is triggered y melting of the crystlline domins, which is reltively fst process. It is worthy to note tht the PE-1K film ws deformed t 275 C in one step; however, this polymer shows two distinct recovery steps with mximum recovery velocity of min 1 t round 144 nd 240 C, respectively. The first recovery tkes plce when the smple psses through the glss trnsition nd reches n of 42% t 200 C. The second recovery is triggered y the melting of the crystlline domins. These two recovery steps indicte tht the crystlline domins in PE-1K re not le to form penetrting scffold throughout the smple nd thus cnnot completely lock the morphous chins. This llows the smple to prtilly recover t the T g T m rnge. These results demonstrte tht the degree of crystllinity, which is reflected in the ΔH m vlues in Tle 1, strongly ffects the shpe memory ehvior of our semicrystlline PAs. The dul-shpe memory ehvior of the TPE-sed thermosets were investigted using the sme method. Tle 3 Tle 3. Dul-Shpe Fixtion nd Recovery Results of the TPE Thermoset Films cycle 1 cycle 2 cycle 3 smple T prog T r TPE TPE TPE , T prog refers to the progrmming temperture. T r refers to the temperture t the mximum recovery velocity. shows the shpe fixtion nd recovery results for these smples. All smples revel good shpe fixtion with vlues of 93 99%. The TPE-15 smples with the highest cross-linking density shown in Tle 1 exhiit excellent recovery efficiency ( > 90%), wheres the other smples show vlues of 62 66% in the first cycle nd 78 82% in the second nd third cycles. Figure 6 shows the dul-shpe memory cycles nd the recovery velocity of the TPE thermoset films. Similr to the PE- 3K smple, TPE-5 nd TPE-10 exhiit one recovery step in the temperture rnges of nd C, respectively. Interestingly, the TPE-15 film lso displys two recovery steps, which is similr to tht of the PE-1K film. The two recovery steps originte from the low degree of crystllinity in TPE-15 nd PE-1K, where the morphous chin segments cnnot e 19110

6 Figure 6. Three consecutive dul-shpe memory cycles for the TPE series (A) TPE-5, (B) TPE-10, nd (C) TPE-15. Shpe recovery velocity s function of temperture in the second cycle for the TPE series (D) TPE-5, (E) TPE-10, nd (F) TPE-15. Cooling nd heting rte 10 C min 1 nd N 2 tmosphere. Figure 7. Three consecutive triple-shpe memory cycles for the PE series (A) PE-3K nd (B) PE-1K. Shpe recovery velocity s function of temperture in the second cycle for the PE series (C) PE-3K nd (D) PE-1K. Cooling nd heting rte 10 C min 1 nd N 2 tmosphere. completely fixed when the temperture is ove T g. TPE-5, TPE-10, nd PE-3K, on the other hnd, show higher degrees of crystllinity, nd this prevents the T g -induced shpe recovery. The difference etween these smples clerly indictes tht the shpe memory ehvior strongly depends on the degree of crystllinity of the semicrystlline thermosets. By chnging the concentrtion of the PE side groups in the rective copolymides, the cross-linking density of the thermosets cn e controlled, consequently leding to djustle crystllinity nd shpe memory ehvior Triple-Shpe Memory Behvior. Bsed on our study ove, two reversile processes, the glss trnsition nd the melting process, cn oth ct s the switches to trigger shpe recovery. Unlike the trditionl high-temperture triple-smp composed of multiple components, we were le to design single-component triple-smp using oth switches to chieve two distinct recovery processes. Figure 7A shows the three consecutive triple-shpe memory cycles for PE-3K. The film smple ws first heted up to 315 C, which is ove the T m (283 C) nd rotted y 90 from the originl shpe A (φ A ) to temporry shpe B (φ B ). The smple ws cooled to 200 C, temperture etween T m nd T g,tofix the shpe B. A second rottion of 90 ws then pplied to rech the temporry shpe C (φ C ). The smple ws susequently cooled to 60 C, which is elow the T g of 127 C, to fix the shpe C. The externl stress ws then removed, 19111

7 Tle 4. Triple-Shpe Fixtion nd Recovery Results of the PE- nd TPE-Series Thermoset Films smple T prog1 T prog2 T r1 T r2 cycle 1 cycle 2 cycle 3 (C B) PE-3K PE-1K TPE TPE TPE T prog1 nd T prog2 refer to the first nd second progrmming temperture, respectively. T r1 nd T r2 refer to the temperture t the first nd second mximum recovery velocity, respectively. (B A) (C B) (B A) (C B) (B A) Figure 8. Three consecutive triple-shpe memory cycles for (A) TPE-5, (B) TPE-10, nd (C) TPE-15. Angulr velocity of shpe recovery s function of temperture in the second cycle for (D) TPE-5, (E) TPE-10, nd (F) TPE-15. Cooling nd heting rte 10 C min 1 nd N 2 tmosphere. leding to the finl fixed temporry shpe (φ f ). A shpe fixtion rte ( )of 90% ws clculted using eq 4. φf = 100% φc (4) In the susequent recovery step, the smple ws heted up to 315 C in stress-free stte to monitor the recovery of the rottionl ngle. The shpe recovery ws ccomplished in two distinct steps, s shown in Figure 7A. This mens the thermoset cn memorize two temporry shpes in one single shpe memory cycle. The shpe recovery rte (C B) nd (B A) were clculted sed on eqs 5 nd 6, respectively. 21 φ φ f B/rec (C B) = 100% φ φ C B (5) φ φ B/rec A/rec (B A) = 100% φ φ B A (6) where φ A, φ B, nd φ C denote the rottionl ngle of shpe A (φ A =0 ), shpe B (φ B =90 ), nd shpe C (φ C = 180 ); φ B/rec nd φ A/rec re the rottionl ngles of the first nd second recovered shpes, respectively. Figure 7B shows the triple-shpe memory cycles of PE-1K. This polymer exhiits lower T m (244 C) compred to PE-3K (283 C); therefore, lower progrmming tempertures (275 nd 150 C) were used. PE-3K shows moderte (C B) vlues of 67 80% in the first recovery step nd (B A) vlues of 60 73% in the second recovery step, s listed in Tle 4. Compred to PE-3K, PE-1K shows much higher (C B) vlues of % nd lower (B A) vlues of 40 49%. The first recovery is triggered y the glss trnsition nd is due to the relese of stress in the morphous phse. The mjor recovery tkes plce in the first step, indicting tht the crystlline phse cnnot completely lock the stress in the morphous phse. The stress, which is supposed to relese in the second recovery step, is ctully prtilly relesed in the first recovery step. This, together with the stress induced y the second deformtion, drives the smple to rech high recovery rte (>100%) in the first step. This result is consistent with the result we otined from the dulshpe memory experiments nd origintes from the low degree of crystllinity of the PE-1K thermoset. Figures 7C nd 7D show the shpe recovery velocity of PE- 3K nd PE-1K smples s function of temperture. The first recovery step of PE-3K shows lower V r (13 min 1 ) thn the second step (28 min 1 ) ecuse of the limited moility of the morphous chin segments restricted y the crystlline domins. In contrst, the PE-1K smple shows much higher V r (54 min 1 ) in the first step, which is due to its low degree of crystllinity. The triple-shpe memory cycles nd recovery velocity of the TPE thermosets re shown in Figure 8. The TPE-5 nd TPE

8 Figure 9. Three consecutive triple-shpe memory cycles for TPE-10 nd shpe recovery velocity s function of temperture in the second cycle. (A) nd (B) deformtion of 180 nd 180 in two steps, respectively; (C) nd (D) deformtion of 180 nd 360 in two steps, respectively. films show moderte (C B) nd (B A) (64 91%) nd higher V r vlues in the second recovery step, which is similr to the ehvior of PE-3K, wheres the TPE-15 smple, similr to PE- 1K, exhiits high (C B) (97 136%) nd V r vlues (38 min 1 ) in the first recovery step, ecuse this polymer hs the lowest degree of crystllinity. The tempertures t the first mximum recovery velocity (T r1 ) of ll smples re within nrrow rnge of C ecuse the T g of ll smples is round 125 C. However, the tempertures of the second mximum recovery velocity (T r2 ), which depend on the T m of the smples, disply rod vrition etween 236 nd 289 C. Therefore, the triple-shpe memory ehvior of the thermosets is highly tunle over wide temperture regime. Besides deforming the smple in one direction only, more interestingly, we cn involve oth forwrd nd ckwrd deformtions in triple-shpe memory cycle; therefore, the corresponding temporry shpes would recover in reverse direction. To demonstrte this, the TPE-10 thermoset smple, which shows high (>90%), (C B) ( 90%), nd (B A) ( 80%), ws selected s representtive exmple. The results of test I re shown in Figure 9A,B. The smple ws progrmmed forwrd nd ckwrd y 180 in two steps nd reched finl temporry shpe, which is identicl to the originl shpe. The susequent heting resulted in n increse in ngle in the first recovery step chieving mximum recovery velocity t 142 C nd n (C B) of 79 91% t 213 C. By incresing the temperture further, the smple recovered in the reverse direction reching mximum recovery velocity t 243 C nd n (B A) of 81 87% (Tle 5). These results demonstrte tht our semicrystlline PA thermosets show onewy reversile shpe memory ehvior. The reversile shpe chnges in this experiment occurred without ppliction of ny Tle 5. Triple-Shpe Fixtion nd Recovery Results of TPE-10 Deformed in Two Opposite Directions test cycle no. Δφ d1 (deg) Δφ d2 (deg) (C B) (B A) T mx I II Δφ d1 nd Δφ d2 refer to the deformtion ngles in the first nd second deformtion steps. T mx refers to the temperture t the mximum ngle. externl force. The shpe chnges were driven y internl stress of the oppositely strined networks formed in the first nd second progrmming steps. We cn djust the second progrmming step from Δφ d2 = 180 to Δφ d2 = 360 in test II, which results in two opposite temporry shpes, s shown in Figure 9C,D. The smple shows of 89 91%, (C B) of 82 86%, nd (B A) of 54 61% (Tle 5). The mximum recovery velocity occurs t 142 nd 244 C in the first nd second recovery steps, respectively, which is consistent with the results of Test I. Our results clerly demonstrte tht high-temperture SME with tunle recovery directions nd mplitudes cn e designed sed on our single-component semicrystlline polymide thermosets. 4. CONCLUSIONS We hve investigted two novel series semicrystlline PA thermosets nd demonstrted tht they cn e s singlecomponent high-temperture (>200 C) shpe memory 19113

9 polymers (SMPs). Two moleculr design pproches sed on rective phenylethynyl (PE) functionlities hve een explored: cross-linked semicrystlline PA films were prepred y curing rective thermoplstic PA 10T oligomers, nd films were prepred y curing rective thermoplstic side-group functionlized copolymides. Compred to the thermoplstic PA 10T reference polymer, the PE-3K, TPE-5, nd TPE-10 thermoset films show high-temperture dul-shpe memory ehvior (>200 C) when T m is used s the switching temperture. The densely cross-linked PE-1K nd TPE-15 films show the highest fixtion rte (99%) nd recovery rte ( 90%) nd two distinct recovery steps ecuse the degree of crystllinity is too low nd cnnot provide scffold tht locks the shpe t tempertures etween T g nd T m. Triple-shpe memory ehvior cn e demonstrted when the T g ( 125 C) is used s the second switching temperture. The recovery rte of the two recovery steps is highly dependent on the degree of crystllinity of the thermosets nd vry within rod rnge of 74% 139% nd 40 82% for the first nd second step, respectively. One-wy reversile shpe memory events cn lso e designed when we perform forwrd nd ckwrd deformtion in triple shpe memory cycle. We lso studied the recovery velocity s function of temperture to elorte the thermokinemtics of the shpe recovery process. The use of phenylethynyl rective functionlities llows us to control the degree of crystllinity in the finl polymide thermosets nd in turn tune their shpe memory ehvior in terms of recovery temperture, velocity, nd efficiency. We elieve tht the design rules presented herein will help in designing new shpe memory polymers sed on well-known semicrystlline polymide chemistries. ASSOCIATED CONTENT *S Supporting Informtion The Supporting Informtion is ville free of chrge on the ACS Pulictions wesite t. Figure S1 (PDF) AUTHOR INFORMATION Corresponding Author *E-mil: tjd@unc.edu (T.J.D.). ORCID Qingo Gun: Theo J. Dingemns: Present Address T.J.D.: Deprtment of Applied Physicl Sciences, University of North Crolin t Chpel Hill, 1113 Murry Hll, 121 South Rod, Chpel Hill, NC Notes The uthors declre no competing finncil interest. ACKNOWLEDGMENTS This reserch forms prt of the reserch progrm of the Dutch Polymer Institute (DPI), Project #743. The uthors thnk the DPI for finncil support. The uthors thnk Prof. S. Sheiko, Dr. J. Bijleveld, Dr. R. Bose, nd Dr. R. Rulkens for fruitful discussions. REFERENCES (1) Xie, T. Recent dvnces in polymer shpe memory. Polymer 2011, 52 (22), (2) Hu, J.; Zhu, Y.; Hung, H.; Lu, J. Recent dvnces in shpe memory polymers: Structure, mechnism, functionlity, modeling nd pplictions. Prog. Polym. Sci. 2012, 37 (12), (3) Rousseu, I. A. Chllenges of shpe memory polymers: A review of the progress towrd overcoming SMP s limittions. Polym. Eng. Sci. 2008, 48 (11), (4) Behl, M.; Rzzq, M. Y.; Lendlein, A. Multifunctionl Shpe- Memory Polymers. Adv. Mter. 2010, 22 (31), (5) Pilte, F.; Tonchev, A.; Duois, P.; Rquez, J.-M. Shpe-memory polymers for multiple pplictions in the mterils world. Eur. Polym. J. 2016, 80, (6) Hger, M. D.; Bode, S.; Weer, C.; Schuert, U. S. Shpe memory polymers: Pst, present nd future developments. Prog. Polym. Sci. 2015, 49-50, (7) Leng, J.; Ln, X.; Liu, Y.; Du, S. Shpe-memory polymers nd their composites: stimulus methods nd pplictions. Prog. Mter. Sci. 2011, 56 (7), (8) Zho, Q.; Qi, H. J.; Xie, T. Recent progress in shpe memory polymer: New ehvior, enling mterils, nd mechnistic understnding. Prog. Polym. Sci. 2015, 49-50, (9) Wu, X.; Hung, W. M.; Zho, Y.; Ding, Z.; Tng, C.; Zhng, J. Mechnisms of the shpe memory effect in polymeric mterils. Polymers 2013, 5 (4), (10) Meng, H.; Li, G. A review of stimuli-responsive shpe memory polymer composites. Polymer 2013, 54 (9), (11) Lewis, C. L.; Dell, E. M. A review of shpe memory polymers ering reversile inding groups. J. Polym. Sci., Prt B: Polym. Phys. 2016, 54 (14), (12) Xio, X.; Kong, D.; Qiu, X.; Zhng, W.; Liu, Y.; Zhng, S.; Zhng, F.; Hu, Y.; Leng, J. Shpe memory polymers with high nd low temperture resistnt properties. Sci. Rep. 2015, 5, (13) Xio, X.; Kong, D.; Qiu, X.; Zhng, W.; Zhng, F.; Liu, L.; Liu, Y.; Zhng, S.; Hu, Y.; Leng, J. Shpe-memory polymers with djustle high glss trnsition tempertures. Mcromolecules 2015, 48 (11), (14) Shi, Y.; Weiss, R. Sulfonted Poly (ether ether ketone) Ionomers nd Their High Temperture Shpe Memory Behvior. Mcromolecules 2014, 47 (5), (15) Shi, Y.; Yoonessi, M.; Weiss, R. High temperture shpe memory polymers. Mcromolecules 2013, 46 (10), (16) Yng, Z.; Chen, Y.; Wng, Q.; Wng, T. High performnce multiple-shpe memory ehviors of Poly (enzoxzole-co-imide) s. Polymer 2016, 88, (17) Xio, X.; Qiu, X.; Kong, D.; Zhng, W.; Liu, Y.; Leng, J. Opticlly trnsprent high temperture shpe memory polymers. Soft Mtter 2016, 12 (11), (18) Bi, Y.; Mo, L.; Liu, Y. High temperture shpe memory polyimide ionomer. J. Appl. Polym. Sci. 2016, 133 (30), (19) Wng, Q.; Bi, Y.; Chen, Y.; Ju, J.; Zheng, F.; Wng, T. High performnce shpe memory polyimides sed on π π interctions. J. Mter. Chem. A 2015, 3 (1), (20) Koerner, H.; Strong, R. J.; Smith, M. L.; Wng, D. H.; Tn, L.- S.; Lee, K. M.; White, T. J.; Vi, R. A. Polymer design for high temperture shpe memory: Low crosslink density polyimides. Polymer 2013, 54 (1), (21) Gun, Q.; Picken, S. J.; Sheiko, S. S.; Dingemns, T. J. High- Temperture Shpe Memory Behvior of Novel All-Aromtic (AB)n- Multilock Copoly(ester imide)s. Mcromolecules 2017, 50 (10), (22) Bellin, I.; Kelch, S.; Lnger, R.; Lendlein, A. Polymeric tripleshpe mterils. Proc. Ntl. Acd. Sci. U. S. A. 2006, 103 (48), (23) Behl, M.; Lendlein, A. Triple-shpe polymers. J. Mter. Chem. 2010, 20 (17), (24) Xie, T.; Xio, X.; Cheng, Y. T. Reveling triple-shpe memory effect y polymer ilyers. Mcromol. Rpid Commun. 2009, 30 (21), (25) Xie, T. Tunle polymer multi-shpe memory effect. Nture 2010, 464 (7286),

10 (26) Li, M.; Dingemns, T. J. Synthesis nd chrcteriztion of semicrystlline poly (decmethylene terephthlmide) thermosets. Polymer 2017, 108, (27) Li, M.; Bijleveld, J.; Dingemns, T. J. Synthesis nd Properties of Semi-crystlline Poly(decmethylene terephthlmide) Thermosets from Rective Side-group Copolymides. Eur. Polym. J. 2018, 98, (28) Wgermier, W.; Krtz, K.; Heuchel, M.; Lendlein, A. Chrcteriztion methods for shpe-memory polymers. In Shpe- Memory Polymers; Springer: 2009; pp (29) Dini, J.; Fredy, C.; Gilormini, P.; Merckel, Y.; Reǵnier, G.; Rousseu, I. A torsion test for the study of the lrge deformtion recovery of shpe memory polymers. Polym. Test. 2011, 30 (3), (30) Bghni, M. Anlyticl study on torsion of shpe-memorypolymer prismtic rs with rectngulr cross-sections. Int. J. Eng. Sci. 2014, 76, (31) Bghni, M.; Nghddi, R.; Arghvni, J.; Sohrpour, S. A constitutive model for shpe memory polymers with ppliction to torsion of prismtic rs. J. Intell. Mter. Syst. Struct. 2012, 23 (2), (32) Dini, J.; Gilormini, P.; Fre dy, C.; Rousseu, I. Predicting therml shpe memory of crosslinked polymer networks from liner viscoelsticity. Int. J. Solids Struct. 2012, 49 (5), (33) Roerts, C. C.; Apple, T. M.; Wnek, G. E. Curing chemistry of phenylethynyl-terminted imide oligomers: Synthesis of 13C-leled oligomers nd solid-stte NMR studies. J. Polym. Sci., Prt A: Polym. Chem. 2000, 38 (19), (34) Iql, M.; Norder, B.; Mendes, E.; Dingemns, T. J. All-romtic liquid crystlline thermosets with high glss trnsition tempertures. J. Polym. Sci., Prt A: Polym. Chem. 2009, 47 (5), (35) Lendlein, A.; Kelch, S. Shpe-memory polymers. Angew. Chem., Int. Ed. 2002, 41 (12), (36) Većhmre, C.; Buleón, A.; Chunier, L.; Guthier, C.; Lourdin, D. Understnding the mechnisms involved in shpe memory strch: mcromoleculr orienttion, stress recovery nd moleculr moility. Mcromolecules 2011, 44 (23),