Heat and solvent responsive polytriazole: shape recovery properties in different solvents

Abstract

Shape memory polytriazole actuatable by both heat and solvents is presented. Hydrogen bonding solvents exhibit strong influence on shape recovery. In this work, a dual-stimuli responsive (both temperature and solvents) polymer is synthesized from propargylated bisphenol-F (PF) and a tris azide (TA) monomer via an azide–alkyne polycycloaddition reaction in the presence of a copper catalyst. The polytriazole exhibits a trigger temperature at 103 °C and displays cyclic shape memory properties. The SMP shows a storage modulus >2 GPa with a cross-link density of 6.9 × 10³ mol m⁻³. The SMP possesses shape memory features in different solvents. Highly polar solvents lead to shape recovery with cracks whereas hydrogen bonding solvents like ethanol supported crack-free shape recovery. A two-step kinetic profile is seen in the shape recovery sequence of SMP in hydrogen bonding solvents. The SMP is hydrophobic and no recovery in water is observed.

---

1. Introduction

Shape memory polymers (SMPs) are known for their exceptional capability to remember their original mechanical shape by erasing the temporary structure on exposure to external stimuli such as heat, light, current, magnetic field, pH and moisture.^1–4 Heat-driven SMPs are the most widely investigated class and they are used as heat shrinkable tubes, self-deployable antennae and mechanical hinges in aerospace; catheters, sutures and for drug delivery in the biomedical field; and self-adjusting shoes or apparel in the textile field.^5–10 Chemical/physical networks containing a switchable segment is the fundamental morphology essential for a polymer to exhibit shape memory properties.^11–13 The network structure enables retention of permanent shape meanwhile switching segment gets activated (or deactivated) by external stimulus. The SMPs are actuated at a certain temperature which is known as trigger temperature (T_trig) and they can be either glass transition temperature (T_g), melting temperature (T_m) or hydrogen bonds.^14–18 The mechanism of shape memory phenomenon relies on the stored strain (during shape fixing) releasing potential of SMP upon application of stimulus. To find alternative stimuli sources other than heat, recently thrust is being shifted to chemo-responsive SMPs i.e. solvent-driven SMPs, they can find application in drug delivery, chemo-sensing techniques and many other futuristic applications such as immobilized biocatalysts, solvent-responsive chromatography and chemo-responsive surfaces.^19–27

Solvent-responsive SMPs were studied by different groups and mostly water/moisture was employed as the triggering medium.^28–34 Weder et al. reported a water-driven shape memory system based on a blend of polyurethane and cellulose nanowhiskers where the recovery was highly dependent on water absorbing capacity of cellulose. Shape recovery ranging from 55% to 99% was achieved with different contents of cellulose.^28,29 Two hydrolytically degradable polymer networks were combined (synthesized by UV curing-oligo [(ε-caprolactone)-co-glycolide]-dimethacrylates) to form an SMP which exhibited potential applications in drug delivery.³⁰ Chen et al. prepared pyridine (Py) containing shape memory polyurethane from N,N-bis(2-hydroxylethyl) isonicotinamide and hexamethylene diisocyanate, which was highly responsive to relative humidity conditions. The mechanism for the moisture sensitivity of this polymer is attributed to the replacement of N–H⋯N–Py hydrogen bonding (in dry SMP) by Py–N–H⋯O–H–O⋯H–N–Py bonding upon exposure to water/moisture.³¹ In another work, poly(acrylamide-co-acrylic acid) hydrogels actuatable by both heat and solvents were investigated. At lower moisture content, the hydrogel exhibited moisture induced and heat assisted shape memory behavior whereas at higher moisture content, the gel displayed a mechano-responsive shape changing behavior.³² A plastic sponge filled with a hydrogel exhibited a low temperature triggering in water (at 20 °C).³³ A shape memory hybrid based on silicone–sodium acetate trihydrate took about 50 h for water actuated shape memory which may be due to the high water repellency of silicones.³⁴

Besides water, other solvents were also employed for actuation of SMPs. Previously, a styrene based commercially available SMP (Veriflex® S, VF62) was exposed to highly polar DMF solvent and this SMP recovered 95% of its original shape after a duration of 140 min.³⁵ Later, the physical swelling effect during the recovery of styrene based SMP was studied by Chen et al. and found that lowering of glass transition temperature is the rationale for shape recovery (T_g was lowered from 56 °C to 34 °C).³⁶ Thermosetting styrene based SMP was also reported to have lowering of T_g upon interaction with DMF. A reduction in modulus from 1475 to 604 MPa was also observed and the hydrogen bonding between polymer and solvent played a key role in improving the flexibility of polymer.³⁷ Natural rubber (T_trig ∼ 33 ± 0.5 °C) under constrained state generated a mechanical stress when exposed to toluene vapor and resulted in shape recovery. When the vapor was removed, the rubber reprogrammed itself and this shape memory was repeatable for several cycles.³⁸ Poly(vinyl butyral) based polymer network with dual-stimuli responsive shape memory behavior (heat and solvents like methanol, ethanol, propanol, dimethylsulphoxide, dimethyl formamide, acetone, dichloromethane, ethyl acetate, tetrahydrofuran) was prepared by Bai and coworkers.³⁹ About 99.5% shape recovery was noticed in different solvents with recovery time between 1 min to 2 h. By combining the Gordon Taylor and free volume theory, Lu et al. explained that the plasticization of shape memory polymer is the working mechanism of chemo-responsive shape memory systems.⁴⁰ However, studies on solvent-driven SMPs are limited hence this category of SMP is to be explored in detail.

Of late, click polymerizations are pursued as a new synthetic route to realize novel SMPs. Heat and light actuatable click-SMPs were reported widely. Diels-alder, azide–alkyne (copper catalyzed azide–alkyne cycloaddition, CuAAC) and thiol–ene are the commonly employed click polymerizations in the field of SMPs.^41–60 To cite a few, poly(lactic acid) with terminal furanyl moieties on cross-linking with tris(2-maleimidoethyl) amine exhibited shape memory properties with switching temperature in the range of 50–65 °C.⁴³ Semi crystalline poly(ε-caprolactone) copolymer network containing cinnamoyl moiety with shape memory behavior was synthesized by 2 + 2 polycycloadditions. They displayed shape recovery in <50 s with extent of shape recovery ∼99% (T_trig tunable between 54 °C and body temperature).⁴⁶ By exploiting the photo-reversibility of cinnamic acid molecule, cross-linked SMPs responsive to UV light (above and below 260 nm) were synthesized by Lendlein et al.⁴² In thiol–ene based SMPs, for example, trimethylol propane tris 3-mercaptopropionate and triallyl-1,3,5-triazine-2,4,6-trione were synthesized by thiyl radical–vinyl addition photo polymerization. They possessed T_trig of about 30–40 °C with shape recovery of 100% in the constrained condition.⁵⁰

Azide–alkyne polycycloaddition is an important route used for synthesizing new SMPs. We synthesized an SMP by cross-linking propargyl ether novolac oligomer and bisphenol-A (bisazidohydroxy propyl) ether which featured a high elastic ratio [modulus ratio between T_trig −20 °C to T_trig +20 °C] of 34 with T_trig of 73 °C. Shape retention and recovery of >90% were achieved.⁴⁷ The azide–alkyne chemistry was also explored for obtaining shape memory properties in the surface level.⁴⁸ Recently, we employed copper catalysed azide–alkyne reaction for synthesizing a shape memory polytriazole using a tris azide and bi/tri functional alkyne monomer which manifested high flexibility of triazole bridges with hydrogen bonding interactions.⁵⁵

From the discussion, it is evident that click based SMPs are in the growing stage and chemo-responsive SMPs in particular are rarely investigated. Out of different solvents, water is the most common solvent used to activate SMPs. Though water-responsive SMPs are easy to function, the entry of moisture during storage of temporary shape is a big concern in such systems. This eventually can result in premature shape recovery on storage. Hence, an SMP which is responsive to selective solvents and insensitive to water is always interesting for specific applications. Further, an SMP which can be actuatable by both heat (aerospace applications) and a green solvent like ethanol (biomedical use) can be highly impressive because a single SMP can serve applications in both fields. In this work, we synthesized an SMP using copper catalyzed azide–alkyne polycycloaddition reactions. The SMP is responsive to both heat and selective solvents but insensitive to water. Heat-triggered shape recovery is investigated in detail followed by its response towards different kind of solvents. The weak interactions between SMP and solvents are addressed and correlated with shape recovery behaviors observed in different solvents.

2. Experimental

2.1 Materials

The reagents, sodium azide (99%, Spectrochem, India), [tris(2-hydroxy 3-azido propyl) phenyl ether] methane (99%, Sigma Aldrich, USA), ammonium chloride (99.8%, High Purity laboratories, India), sodium hydroxide (98%, SD-Fine Chemicals, India), propargyl bromide (97%, 80% w/w in toluene, Alfa Aesar, UK), benzyl triethyl ammonium chloride (98%, Spectrochem, India), acetonitrile (99.5%, Sisco Chemicals, India), cuprous iodide (99.9%, Sigma Aldrich, USA), dimethyl formamide (99.5%, SRL, India), acetone AR (Finar, India), methanol AR (SRL, India), ethyl methyl ketone (99%, Alfa Aesar, UK), dichloromehane (SRL, India) and ethanol (95%, Chittur sugar refineries, India) were used as received.

2.2 Instrumental methods

FTIR spectra were recorded with Perkin Elmer spectrum GXA spectrophotometer at a scan rate of 4 cm⁻¹. ¹H and ¹³C NMR spectroscopies of monomers were carried out on Bruker Avance spectrometer (400 MHz) using either CDCl₃ or D-acetone as solvents. Differential Scanning Calorimetric studies were conducted using TA instrument model 2920 modulated DSC at a heating rate of 10 °C min⁻¹ under N₂. Dynamic mechanical analysis (DMA) was carried on DMA Q800 (TA Instruments, USA) in a 35 mm dual cantilever mode at a heating rate of 3 °C min⁻¹ at 1 Hz. Thermogravimetric analysis (TGA) was carried out on a TA instrument, model SDT-2960 simultaneous DTA (differential thermal analysis)-TGA at a heating rate of 10 °C min⁻¹ under N₂. Shape memory properties were evaluated by fold-deploy test (heating above and below the T_trig of SMPs) by measuring the folding and recovery angles. The rectangular sample was kept in a temperature controlled oven at T_trig +20 °C and deformed for fixing the shape. The angle of shape fixity (θ_f), maximum folded angle (θ_m), and shape recovery angle (θ_r) were measured. Shape retention and shape recovery ratios were calculated using the following equations;

Extent of shape fixity (S_f, %) = [θ_f/θ_m] × 100 (1)

Extent of shape recovery (R_r, %) = [θ_f − θ_r/θ_f] × 100 (2)

The static water contact angle (SWCA) of the polymer was measured in Data physics OCA 15EC instrument using SCA20 software by Laplace–Young fitting method. Water droplets of 5 μL were placed over the SMP surface (at five different locations and averaged) to measure the SWCA. Surface energy of the SMP was calculated by Owens, Wendt, Rabel and Kaelble method (OWRK) method.

2.3 Synthesis of shape memory polymer (TAPF)

Syntheses of clickable monomers viz.: [tris(4-(1-azido 3-oxy propan-2-ol) phenyl)] methane (TA) and [4,4′-(perfluoro propane-2,2-diyl) bis(prop-2-yn-1-yloxy) benzene (PF) were carried out as per our previous reports⁵¹ (ESI 1† for synthesis and characterization). The cross-linked polymer (TAPF) was synthesized via azide–alkyne click polymerization. The resinous monomers (TA and PF) were taken in 1 : 1 molar ratio and were magnetically stirred. To this clear mixture, cuprous iodide catalyst (0.1 wt%) was added, the reaction mixture was poured into a mold and degassed for one hour in a degassing unit. Subsequently, the mixture was subjected to a progressive heating up to 140 °C and maintained at this temperature for a duration of 5 h to complete the polytriazole formation. The cross-linked polymer was reddish brown in color and semi-transparent in nature. The as-prepared polymer with thickness of 0.3–0.4 mm was used for all studies.

3. Results and discussions

3.1 Characterization of shape memory polytriazole (TAPF)

The polytriazole SMP (TAPF) was synthesized by copper catalyzed azide–alkyne polycycloaddition reaction between TA and PF monomers in the presence of cuprous iodide catalyst. The completion of “click polymerization” was confirmed from the disappearance of azide groups (at 2105 cm⁻¹) and appearance of triazole peak at 1381 cm⁻¹ in FTIR. The heat of enthalpy of this click polymerization was 150 kJ/triazole formation (single exotherm in DSC indicates absence of other secondary reactions also) which is being quite high when compared to well-known thermosets like epoxy or polybenzoxazine.⁶¹ This implies also the high propensity of both the monomers for polycycloaddition reaction via triazole linkages. Synthesis scheme of TAPF, differential scanning calorimetry of click polycycloaddition and FTIR spectra of monomer mixture before and after click polymerization are shown in Fig. 1.

Fig. 1 (a) Schematic representation of synthesis of shape memory polytriazole via copper catalyzed azide–alkyne polycycloadditon (b) differential scanning calorimetric profile shows the cross-linking pattern of SMP (c) FTIR spectra of monomer mixture after and prior to click cross-linking.

3.2 Heat-induced shape memory properties of TAPF

The TAPF exhibits trigger temperature (T_trig) at 103 °C with a high damping factor >0.5 (dissipation of energy in a material under cyclic load) as seen in dynamic mechanical analysis (Fig. 2a). This high tan delta value is attributed to the cross-linked structure of TAPF polymer. The glassy modulus at 30 °C was 2.1 GPa which is a high value reported until now for SMPs derived by click polymerization method.^59,62,63 In literature, an SMP network formed from a diazide and a tetrayne exhibited a storage modulus of 1 GPa.⁵⁹ Similarly, a hyper branched polyurethane–tetrazole system displayed storage modulus in the regime of 1 GPa only.⁶² The previously reported shape memory polymers achieved storage moduli in the range of 1–1.5 GPa.⁶³

Fig. 2 (a) Dynamical mechanical analysis features the storage modulus as a function of temperature (b) demonstration of heat actuated shape recovery of TAPF polymer (c) extent of shape retention and recovery of TAPF on heat-actuated cyclic shape memory tests.

The storage moduli at T_trig −20 °C (E_g) and T_trig +20 °C (E_r) of TAPF were 1020 MPa and 23 MPa, respectively. This much large difference between E_g and E_r moduli is noteworthy, which provides a high elastic ratio (E_g/E_r ratio) of 44.3. The high ratio is an index of the good stability of the temporary shape and its easy shape recoverability. A high modulus ratio was noticed for click based SMPs in earlier reports.⁵⁵ Further, the recovery strength (strain produced per unit stress) was calculated from DMA (ESI 2†). A high recovery strength of 1 GPa was observed at E_g temperature. To elaborate more, the calculated cross-link density was 6.9 × 10³ mol m⁻³ which is higher than previously reported click based SMPs^51,55 (ESI 3†). The SMP is thermally stable up to 250 °C (T_10%, ESI 3b†). The extent of shape recovery, shape retention ratio and recovery time were determined by the universally accepted fold-deploy test (bending test, rectangular bar of size 70 × 7 × 0.4 mm³ was used, demonstration is shown with a 1.5 mm thick SMP bar). The average shape retention and recovery ratios were about 98% with a shape recovery time of 60 s. The demonstration of shape recovery sequence is shown in Fig. 2b. The fold-deploy test was conducted ten times repeatedly and determined the shape memory properties of TAPF bar (Fig. 2c). Surprisingly, the recovery time showed no much variation with cycles (varied between 60–63 s only). It is to be highlighted that, click based SMPs with elastic ratio of 28.6 and 8.0 have shown an increased recovery time from 40 s to 72 s with increase in number of cycles as per our previous reports.⁵⁵ In this work, the retention and recovery of shape in cyclic tests are attributed to the contribution of high elastic ratio of TAPF. This cyclic thermo-responsive performance of SMP demonstrates that they are very good candidates for the cutting-edge areas where a very precise control is highly demanded. For understanding heat transfer in TAPF, specific heat of TAPF was determined as 1.76 J g⁻¹ °C⁻¹. By heating TAPF bar (used for the studies of shape recovery, weight = 2.49 g) from ambient temperature (30 °C) to T_trans +20 °C (124 °C), heat energy of 4.4 J was required for 98% shape recovery. Further, thermal conductivity of TAPF was determined as 0.11 W m⁻¹ K⁻¹.

3.3 Solvent-induced shape memory properties of TAPF

The chemo-responsive/solvent-sensing behavior arises due to chemical or physical interaction between macromolecules and solvent molecules. Generally, polymers will swell when immersed in particular solvents. The absorption-cum-swelling of polymers is influenced by several factors, like molar volume, polarity, solubility parameter of solvents and so on. These physical/chemical interactions between solvent and polymer offer volume change or stress release and this feature can be capitalized for shape recovery of polymers. Here, we tried to investigate the shape recovery of TAPF polymer in (a) polar aprotic solvents (dimethyl formamide, acetone, dichloromethane and ethyl methyl ketone; type I solvents), (b) protic solvents (ethanol, methanol, n-butanol; type II solvents) and (c) water. Table 1 lists swelling and shape recovery parameters of TAPF in type I and type II solvents (ESI 4† for molecular level properties of solvents).

Table 1 Swelling and shape recovery parameters of TAPF in type I and type II solvents^a

	Solvents	ESr (%)	Rr (%)	Sr (%)	Rt (min)
a ESr – equilibrium swelling ratio, Rr – shape recovery ratio, Sr – swelling ratio at Rr, Rt – time for total recovery, S–F – swelled and fragmented.
Type I	DMF	95	S–F	—	—
	Acetone	51	98	21.8	345
	DCM	36	97	22.9	705
	MEK	35	96	23.4	720
Type II	EtOH	49	96	24.3	600
	MeOH	33	95	17.4	180
	n-BuOH	27	95	15.3	15 840
Water		08	—	—	—

---

3.3.1 Shape recovery in different solvents. Firstly, the equilibrium swelling of TAPF polymer in each solvent was gravimetrically determined by a 24 h immersion test in respective solvents. The highest swelling was observed for DMF (95%) while the lowest was noted for water (8%). To see the shape recovery of SMP in various solvents, a rectangular bar of 50 × 7 × 0.4 (mm³) was immersed in 100 ml of individual solvent under closed condition. The SMP bar was immediately got swollen in DMF and completely fragmented within 1 h. But, in acetone, dichloromethane and ethyl methyl ketone, the shape recovery was completed in ∼6 h, 12 h and 12 h, respectively (the specimen was flat after shape recovery in all cases i.e. near to 100% shape recovery, data is shown in Fig. 3a). However, accompanied with shape recovery, the samples got fractured (broken) and cracks were clearly observed right from the beginning of shape recovery. This behavior was observed in type I solvents. Apparently, the mechanism of shape recovery can be correlated to the extent of swelling since it follows the same trend of equilibrium swelling index, viz.: DMF > acetone > DCM ≈ MEK. Secondly, the shape recovery of TAPF in type II (hydrogen bonding solvents) was investigated. Initially, the green solvent, ethanol was studied for actuation. The SMP exhibited near to 100% shape recovery in ethanol (duration-10 h) without any crack on the SMP bar. Here, the shape recovery is tuned by ethanol. The reason for this unusual behavior cannot be attributed to swelling power as it has a swelling index of 49% which is higher than that of acetone.

Fig. 3 Swelling (black) and shape recovery (blue) ratios of TAPF in (a) acetone, DCM and MEK solvents (b) ethanol and (c) methanol as a function of time.

By carefully observing the shape recovery and swelling as a function of time, a radical difference was noticed in the rate of swelling and extent of shape recovery in ethanol (Fig. 3b). If physical swelling (due to imbibed solvent) was the reason for shape recovery, the rates of both swelling and shape recovery should have matched. Generally, the solvents fill the free volume of the polymer first, later solvents penetrate through the polymer chains and lead to significant swelling. Here, in the first phase of shape recovery/swelling (up to about 6 h), a 23% swelling was observed whereas the shape recovery was below 20%. Later, in the second phase (6–8 h), a quantum jump in the recovery was observed that reached 85% in a time span of about 2 h. It is to be noted here that the process of absorption of solvent in the second phase is almost saturated (about 2% absorption only occurred). So it can be concluded that, shape recovery observed in the first phase is due to physical swelling and the quantum jump in the shape recovery in the second phase is owing to weak secondary interactions.

To study this effect further, two similar kind of solvents namely, methanol (less hydrophobic than ethanol) and n-butanol (more hydrophobic than ethanol, results are not shown) were studied. In methanol, similar to the trend observed in ethanol, shape recovery process was split into two regimes, one governed by physical swelling and the second regime driven by weak interactions (Fig. 3b).

A complete shape recovery with no fragmentation/cracking was observed in methanol (3 h). This quick shape recovery compared to ethanol is attributed to the small molecular size of methanol. The shape memory ascribed to physical swelling was complete in about 90 min with a shape recovery of <20%. The second phase of shape memory was quite fast and reached about 95% in 1 h time. The shape recovery time in phase II is shorter when compared to ethanol (1 vs. 2 h) and recovery percentages are also higher (95% vs. 85%). In n-butanol, shape recovery was observed for longer duration of 7 days but without developing any crack. To the kinetic aspect of shape recovery in type I solvents, a steady kinetics with no appreciable delay (no induction period) was seen. The shape recovery pattern followed almost a single step profile. The rates of recovery in acetone, DCM and MEK were 22.3, 9.07 and 9.04% min⁻¹ respectively. However, in type II solvents, a slow kinetics followed by a fast phase is seen as discussed above. The rate of recovery of 1^st phase was 0.18 and 0.06% min⁻¹ in methanol and ethanol respectively. This difference in kinetics and shape memory behavior of TAPF in type I and type II solvents are to be addressed here. Apart from solubility parameter, other factors such as polar, non-polar and hydrogen bonding interactions must be playing crucial roles in determining the kinetics of shape recovery. Though the phase II kinetics is fast, crack-free shape recovery was observed in type II solvents, whereas type I showed uncontrolled strain release with propagation of cracks. Shape recovery of TAPF in acetone, methanol and ethanol are shown in Fig. 4.

Fig. 4 Demonstration of chemo-responsiveness of TAPF in (a) acetone (b) methanol (c) ethanol (d) cracked TAPF (in acetone) and crack-free TAPF (actuated in ethanol).

A series of experiments were carried out for understanding swelling/recovery kinetics (transport phenomenon) of solvents in TAPF. Towards that, swelling ratio of TAPF was determined in type I and type II solvents with respect to time. The most common and semi-empirical Fickian model was used for studying the mechanism of solvent transport (ESI 5†).⁶⁴

Q_t/Q_E = ktⁿ (3)

log(Q_t/Q_E) = log k + n log t (4)

where Q_t is the swelling ratio at time t, Q_E is the equilibrium swelling ratio, n and k are constants. The value of ‘n’ will provide the type of mechanism (‘n’ can be obtained by plotting log Q_t/Q_E vs. log t). The ‘n’ value of ethanol, DCM and MEK are 1.18, 1.08 and 1.17 respectively. This implies that, shape recovery and swelling follows ‘non-fickian super case’. Even though these three solvents follow same mechanism of solvent transport, ethanol resulted in no cracks but other two generated cracks on shape recovery. Meanwhile, diffusion in methanol registered a lowest ‘n’ value of 0.26, indicates that kinetics follows Fickian diffusion where relaxation is more predominant than swelling. The inference is, formation of cracks on shape recovery is independent of diffusion kinetics. Hence, the shape recovery depends on the secondary forces between polymer and solvents. Further, the diffusion constant ‘k’ provides information on the solvent–polymer interaction. Type I solvents gave ‘k’ value in the order of 10⁻⁴ whereas type II solvents furnished higher order of 10⁻², this very evidently support that type II solvents undergo strong interaction with TAPF (though their equilibrium diffusion is low compared to type I solvents, refer Table 1).

To describe the different shape recovery of TAPF in type I and type II solvents, interactions between polymer and solvent are investigated in detail. If solubility parameter (here Hansen Solubility Parameter, HSP) matches with that of solvent, swelling of SMP occurs. First, we calculated the HSP of TAPF by group contribution method as 25 (cal cm⁻³)^1/2.⁶⁵ But, all the type I and type II solvents with different HSP values are capable of inducing significant swelling and subsequent shape recovery of TAPF (vide supra) (Table 1). So, beyond the influence of HSP, independent forces such as polar, dispersive and hydrogen bonding interactions are to be considered. To explain the influence of different factors, parameters like δ_H/δ_P, δ_H/δ_d and δ_H/(δ_H + δ_p + δ_d) were calculated (Table 2). The hydrogen bonding factor in type II solvents is higher than in type I solvents as evidenced from the higher δ_H/δ_P, δ_H/δ_d and δ_H/(δ_H + δ_p + δ_d) values. In acetone actuated shape recovery, polarity plays a major role which leads to fast shape recovery compared to MEK (lower hydrogen bonding potential). Meanwhile, MEK and DCM driven shape recovery occurs at more or less same time due to the balance of polarity and hydrogen bonding factors. In ethanol and other type II solvents, high hydrogen bonding governs over moderate polarity. In a nutshell, when polymer–solvent interaction is governed by polarity, shape recovery with cracks occurs. If the interaction is governed by H-bonding, a constructive shape recovery is resulted.

Table 2 H-bonding factors of different solvents derived from Hansen solubility parameters

	Solvent	δH/δP	δH/δd	δH/(δH + δp + δd)
Type I	DMF	0.82	0.65	0.27
	DCM	0.97	0.41	0.23
	MEK	0.56	0.32	0.17
	Acetone	0.67	0.45	0.21
Type II	MeOH	1.81	1.52	0.45
	EtOH	2.20	1.22	0.44
	n-BuOH	2.77	0.98	0.42
Water		2.64	2.72	0.57

---

FTIR was used to collect the evidence for hydrogen bonding interaction between SMP and solvents. Certainly, the TAPF polymer can undergo hydrogen bonding interactions through –OH groups, and less sterically hindered –N– and –CH of triazole rings.⁴⁶ The three solvents (acetone, DCM and MEK) contain hydrogen acceptors (–CO and –Cl) and can interact with hydrogen bond donors such as –OH and –CH (triazole). In type II solvents, –OH of ethanol/methanol can form hydrogen bond with –OH of TAPF, less sterically hindered –N– and triazole –CH. The –OH absorption of untreated SMP was observed at 3436 cm⁻¹ which got lowered to 3389 cm⁻¹ in acetone and methanol actuated SMPs; whereas ethanol actuated SMP showed absorption at 3420 cm⁻¹ (the samples were subjected to ATR-IR study, immediately after shape recovery). The triazole absorption observed at 1381 cm⁻¹ in dry SMP was also lowered to 1361 cm⁻¹ in all cases after shape memory in solvents (Fig. 5). To support the play of hydrogen bonding in solvent-driven TAPF, the SMP sample after shape recovery in ethanol was heated at 100 °C for 1 h to remove imbibed solvent and to break the hydrogen bond between triazole and ethanol. As expected, FTIR showed vital evidence of loss of hydrogen bonding viz.: the triazole peak shifted from 1361 cm⁻¹ to 1381 cm⁻¹ (as in dry TAPF) after heating. These spectroscopic evidences (FT-IR) clearly imply hydrogen bonding interactions of solvents with TAPF polymer via –CH (triazole)/OH group and H-bonding groups (–CO in acetone, –OH in MeOH/EtOH) in solvents (Fig. 6). In toto, the collective interactions and physical swelling (due to high polarity and high swelling ratios) in acetone, DCM and MEK provided a catalytic drive for TAPF polymer to release energy in an excessive manner. The initial push of solvents produces anisotropic stress release and generate cracks. In type II solvents, diminished polarity, high hydrophobicity and hydrogen bonding control the stress release to result crack-free shape recovery.

Fig. 5 (a) FTIR spectra of TAPF polymer after shape recovery in acetone, ethanol and methanol solvents (b) enlarged view of spectra between 500–1500 cm⁻¹ absorption range to observe the shifts of H-bonding in –CH (c) loss of hydrogen bonding (triazole –CH) on heating (100 °C for 1 h.) of TAPF after shape recovery in ethanol.

Fig. 6 (a) Shifts of hydrogen bonding in dry and wet (ethanol-driven) TAPF (b) schematic representation of mechanism of chemo-responsive shape recovery in TAPF thermoset.

To evaluate the shape memory properties in water, TAPF was immersed in water for rather a long duration (30 days). But, no noticeable shape memory was observed for the SMP. A question naturally arises why water cannot trigger the shape recovery of TAPF. Despite all favorable features such as high polarity (good number of polar groups in SMP too), high hydrogen bonding potential and smaller size, water is insensitive to TAPF. To explore this, the surface energy, dispersive and polar forces of TAPF were experimentally determined by using contact angle measurement (via OWRK method using three liquids viz.: water, ethylene glycol and diiodomethane). The surface energy was found as 40.9 mJ m⁻² (ESI 6†). For any solvent to create an interaction with polymer, the surface force of solvent should be sufficient to overcome the surface energy of polymer. If the surface tension of solvent is lower when compared to polymer, it will overcome the surface force of the polymer and the polymer gets wet. This is the first step of interaction between polymer and solvents. Here, water with high surface energy (72.8 mN m⁻¹) cannot break the surface energy of polymer (or vice versa). It can be observed that the surface tension of all the type I and type II solvents are sufficiently lower than that of TAPF polymer (ESI 7†). This may be the reason for insensitiveness of SMP to water. Also, TAPF forms contact angle 80° with water which implies water repellant nature of TAPF.

It is necessary to discuss the sequence of shape recovery of TAPF in type I and type II solvents. The entry of solvent into polymer (if surface tension is lower than TAPF) matrix creates a pressure which will release the stored energy (pressure stored in polymer chains) of SMP. If the energy releasing rate is very fast, anisotropic expansion of shape occurs and subsequently crack may generate. This behavior is prominent in DMF and resulted in destruction of polymer chains. When it is slightly controlled, shape memory with crack occurs as seen in remaining three type I solvents. In type II solvents, weak hydrogen bonding plays the key role which dominates and controls the shape recovery with its feeble but multiple interactions with TAPF. The role of hydrogen bonding solvents in crack-free shape recovery is described here. On swelling, gap between polymer chains will increase. If this widening occurs in uncontrollable way due to the diffusion of solvent, crack will generate. If there is an opposing force such as hydrogen bonds between solvent and SMP, they will act as physical cross-link. The pseudo cross-linking will retard the movement of solvents to the core of SMP. This condition will facilitates more time for polymer chains to relax i.e. more control can be obtained. Hence, methanol and ethanol can give good hydrogen bonding effect (molecular size is also smaller compared to DCM, acetone and MEK). In type I solvents, polar force is more than hydrogen bonding. Hence, this kind of physical cross-links cannot be stable (though it forms) compared to ethanol/methanol systems.

For understanding the effect of temperature on solvent triggered shape recovery, recovery of TAPF in acetone and ethanol were carried out at a higher temperature (50 °C). It is observed that, recovery time got reduced to 235 min for acetone vis-a-vis 345 min at ambient temperature (30 °C). Similarly, in ethanol, recovery time lowered to 510 min from 600 min at 30 °C.

Table 3 gives an overview of solvent-induced shape memory polymers reported hitherto where water induced shape recovery dominates in most cases. It can be noted that a work on poly(vinyl butyral) only reported to have shape memory in ethanol. The importance of the present work is that the TAPF polymer can be actuatable by water (problem-free storage of temporary shape) and can be activated by solvents in two different modes. One will result in shape recovery with cracks and the other will show crack-free shape recovery. The most important feature of TAPF is, it can be actuatable in a more controlled way in ethanol (green solvent) which enables bio-applications too. Ethanol is a benign solvent for human body, hence the shape recovery can be used for drug delivery applications (recovery time is 10 hours which is reasonable for drug release). Moreover, the recovered SMPs got their strength and modulus after a few hours (ESI 8†). It implies the possibility of solvent-driven SMP for structural bio-applications too.

Table 3 An overview of solvent-triggered shape memory polymers and their shape memory parameters^a

No.	SMP	Solvent	Recovery time (min)	Recovery ratio (%)	References
a NA-not available.
1	Cotton cellulose nano whiskers in polyurethane matrix	Water	NA	55.2	19
2	Polymer sponge/cupric sulphate pentahydrate hybrid	Water	6	NA	21
3	Silicone/sodium acetate trihydrate hybrid	Water	3000	NA	21
4	Plastic sponge filled with poloxamer 407 gel	Water	60 (at 20 °C water) >1 min (at 0 °C water)	NA	23
5	Styrene based SM resin (Veriflex® S)	DMF (at 45 °C)	After 140 min	95	30
6	Cellulose nano whiskers with elastomeric PU matrix	Water	NA	99	20
7	Poly(vinyl butyral) based polymer	Methanol, ethanol, DMF, acetone, toluene etc.	8–43	98	27
8	Chemically cross-linked poly(vinyl alcohol)	Water, methanol, DMF, acetic acid and ethylene glycol	45–5760	95% in all solvents except methanol (55%)	66
9	Graphene oxide reinforced polyvinyl alcohol nano composites	Water	NA	NA	67
10	Sodium dodecyl sulphate/epoxy composite	Water	NA	NA	68
11	Thermoplastic polyurethane	Water	∼5	42	69
12	Hyper branched polyurethane	Toluene, ethyl acetate, THF and DMF	10–17	98	70
13	Styrene based thermosetting SMP	DMF	80	NA	71
14	Polytriazole	Methanol, ethanol and butanol	180, 600 and 15 840	95, 96 and 95	This work
15	Polytriazole	Acetone, DCM and MEK	705, 720 and 600	97, 96 and 96	This work
16	Polytriazole	Water	No recovery	—	This work

---

4. Conclusions

In this work, a cross-linked polytriazole is synthesized via CuAAC and the SMP exhibits a dual stimuli-responsive behavior viz.: by heat and various solvents. The SMP possesses T_trig at 103 °C and shows high storage modulus >2 GPa with high shape recovery features (>98%). The polytriazole SMP is actuatable by multiple solvents (including green solvent ethanol) irrespective of their solubility parameters. Depending upon polymer–solvent interaction, the SMP can lead to shape recovery with and without cracks. In hydrogen bonded solvents, the shape recovery of SMP follows a two stage kinetic profile, meanwhile highly polar solvents permit a single stage pattern. The most important observation is that, the SMP is not actuatable by water (problem-free storage of temporary shape) but by many other common solvents. This kind of SMP i.e. not triggered by water and actuatable by green solvent is rare. The polytriazole thermoset synthesized in this work may find applications in both aerospace and biomedical fields due to its dual-stimuli responsive capability. This work further proves the effectiveness of CuAAC cross-linking as an emerging tool for synthesizing shape memory polymers.

Acknowledgements

The authors acknowledge Director, VSSC for permission to publish this work. Analytical support from Analytical and Spectroscopy Division is also acknowledged. One of the authors (RR) thanks to University Grants Commission, India for providing a Junior Research Fellowship.

References

1. J. Leng, H. Lv, Y. Liu and S. Du, Appl. Phys. Lett., 2007, 91, 144105.
2. J. W. Cho, J. W. Kim, Y. C. Jung and N. S. Goo, Macromol. Rapid Commun., 2005, 26, 412.
3. H. Jiang, S. Kelch and A. Lendlein, Adv. Mater., 2006, 18, 1471.
4. S. Dai, P. Ravib and K. Tam, Soft Matter, 2008, 4, 435.
5. W. M. Huang, C. W. Lee and H. P. Teo, J. Intell. Mater. Syst. Struct., 2005, 17, 754.
6. A. Lendlein and S. Kelch, Angew. Chem., Int. Ed., 2002, 41, 2034.
7. R. Biju, C. Gouri and C. P. Reghunadhan Nair, Eur. Polym. J., 2012, 48, 499.
8. H. Deka, N. Karak, R. D. Kalita and A. K. Buragohain, Carbon, 2010, 48, 2013.
9. L. Xue, S. Dai and Z. Li, Biomaterials, 2010, 31, 8132.
10. A. Lendlein and R. Langer, Science, 2002, 296, 1673.
11. M. Behl and A. Lendlein, Mater. Today, 2007, 10, 20.
12. A. Lendlein, J. Mater. Chem., 2010, 20, 3332.
13. C. Liu, H. Qin and P. T. Mather, J. Mater. Chem., 2007, 17, 1543.
14. N. Zheng, G. Fang, Z. Cao, Q. Zhao and T. Xie, Polym. Chem., 2015, 6, 3046.
15. Y. Wang, Y. Li, Y. Luo, M. Huang and Z. Liang, Mater. Lett., 2009, 63, 347.
16. T. Ware, K. Hearon, A. Lonnecker, K. L. Wooley, D. J. Maitland and W. Voit, Macromolecules, 2012, 45, 1062.
17. Y. Shi, M. Yoonessi and R. A. Weiss, Macromolecules, 2013, 46, 4160.
18. X. L. Wu, W. M. Huang, Z. Ding, H. X. Tan, W. G. Yang and K. Y. Sun, J. Appl. Polym. Sci., 2014, 131, 39844.
19. A. Lendlein, J. Zotzmann, Y. Feng, A. Alteheld and S. Kelch, Biomacromolecules, 2009, 10, 975.
20. C. Yang, H. Wu, J. Sun, H. Hsiao and T. Wang, ACS Appl. Mater. Interfaces, 2013, 5, 10985.
21. C. Wischke, A. T. Neffe and A. Lendlein, Adv. Polym. Sci., 2009, 226, 177.
22. J. Zhou, A. M. Schmidt and H. Ritter, Macromolecules, 2010, 43, 939.
23. T. Govindarajan and R. Shandas, Polymer, 2014, 6, 2309.
24. S. Neuser, V. Michaud and S. R. White, Polymer, 2012, 53, 370.
25. M. Behl, K. Kratz, U. Noechel, T. Sauter and A. Lendlein, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 12555.
26. A. J. Leite, S. G. Caridade and J. F. Mano, J. Non-Cryst. Solids, 2016, 432, 158.
27. I. Y. Galaev and B. Mattiason, Trends Biotechnol., 1999, 17, 335.
28. J. Mendez, P. K. Annamalai, S. J. Eichhorn, R. Rusli, S. J. Rowan, E. J. Foster and C. Weder, Macromolecules, 2011, 44, 6827.
29. K. Fan, W. M. Huang, C. C. Wang, Z. Ding, Y. Zhao, H. Purnawali, K. C. Liew and L. X. Zheng, eXPRESS Polym. Lett., 2011, 5, 409.
30. A. T. Neffe, B. D. Hanh, S. Steuer and A. Lendlein, Adv. Mater., 2009, 21, 3394–3398.
31. S. Chen, J. Hu, C. Yuen and L. Chan, Polymer, 2009, 50, 4424.
32. C. C. Wang, W. M. Huang, Z. Ding, Y. Zhao and H. Purnawali, Compos. Sci. Technol., 2012, 72, 1178.
33. Y. Zhu, J. Hu, H. Luo, R. J. Young, L. Deng, S. Zhang, Y. Fanc and G. Ye, Soft Matter, 2012, 8, 2509.
34. J. L. Zhang, W. M. Huang, H. B. Lu and L. Sun, Mater. Des., 2014, 53, 1077.
35. H. Lv, Y. Liu, D. zhang, J. Leng and S. Du, Adv. Mater. Res., 2008, 50, 258.
36. S. Chen, J. Hu, C. Yuen and L. Chan, Polymer, 2009, 50, 4424.
37. H. Lu, Y. Liu, J. Leng and S. Du, Eur. Polym. J., 2010, 46, 1908.
38. H. Lu, J. Appl. Polym. Sci., 2013, 127, 2896.
39. Y. Bai, Y. Chen, Q. Wang and T. Wang, J. Mater. Chem. A, 2014, 2, 9169.
40. H. Lu, J. Leng and S. Du, Soft Matter, 2013, 9, 3851.
41. J. M. Lobez and T. M. Swager, Macromolecules, 2010, 43, 10422.
42. A. Lendlein, H. Jiang, O. Junger and R. Langer, Nature, 2005, 434, 879.
43. K. Inoue, M. Yamashiro and M. Iji, J. Appl. Polym. Sci., 2009, 112, 876.
44. T. Defize, R. Riva, J. Raquez, P. Dubois, C. Jerome and M. Alexandre, Macromol. Rapid Commun., 2011, 32, 1264.
45. T. Defize, R. Riva, C. Jerome and M. Alexandre, Macromol. Chem. Phys., 2012, 213, 187.
46. A. Garle, S. Kong, U. Ojha and B. M. Budhlall, ACS Appl. Mater. Interfaces, 2012, 4, 645.
47. J. Zhang, Y. Niu, C. Huang, L. Xiao, Z. Chen, K. Yang and Y. Wang, Polym. Chem., 2012, 3, 1390.
48. K. Inoue, M. Yamashiro and M. Iji, Proc. IEEE, 2005, 118.
49. C. Ninh and C. J. Bettinger, Biomacromolecules, 2013, 14, 1.
50. D. P. Nair, N. B. Cramer, T. F. Scott, C. N. Bowman and R. Shandas, Polymer, 2010, 51, 4383.
51. K. Sunitha, K. S. Santhosh Kumar, D. Mathew and C. P. Reghunadhan Nair, Mater. Lett., 2013, 99, 101.
52. S. K. Yadav, H. J. Yoo and J. W. Cho, J. Polym. Sci., Part B: Polym. Phys., 2013, 51, 39.
53. S. M. Brosnan, A. H. Brown and V. S. Ashby, J. Am. Chem. Soc., 2013, 135, 3067.
54. J. Schuetz, P. Wentao and P. Vana, Polym. Chem., 2015, 6, 1714.
55. (a) M. Ragin Ramdas, K. S. Santhosh Kumar and C. P. Reghunadhan Nair, J. Mater. Chem. A, 2015, 3, 11596; (b) S. Chatani, C. Wang, M. Podgorski and C. N. Bowman, Macromolecules, 2014, 47, 4949.
56. E. D. Rodriguez, X. Luo and P. T. Mather, ACS Appl. Mater. Interfaces, 2011, 3, 152.
57. D. P. Nair, N. B. Cramer, J. C. Gaipa, M. K. McBride, E. M. Matherly, R. R. McLeod, R. Shandas and C. N. Bowman, Adv. Funct. Mater., 2012, 22, 1502.
58. T. Ware, D. Simon, K. Hearon, C. Liu, S. Shah, J. Reeder, N. Khodaparast, M. P. Kilgard, D. J. Maitland, R. L. Rennaker II and W. E. Voit, Macromol. Mater. Eng., 2012, 297, 1193.
59. M. K. McBride, T. Gong, D. P. Nair and C. N. Bowman, Polymer, 2014, 55, 5880.
60. S. Chatani, C. Wang, M. Podgorski and C. N. Bowman, Macromolecules, 2014, 47, 4949.
61. K. S. Santhosh Kumar, C. P. Reghunadhan Nair, R. Sadhana and K. N. Ninan, Eur. Polym. J., 2007, 43, 5084.
62. Q. Ligen, A. Anilaand and S. WenFang, Sci. China: Chem., 2011, 54, 1461.
63. I. M. Pereira and R. L. Orefice, Polymer, 2010, 51, 1744.
64. (a) S. Gopakumar and M. R. Gopinathan Nair, Polymer, 2005, 46, 10419; (b) Y. Liu, W. Liu, W. Chen, L. Sun and G. Zhang, Polymer, 2007, 48, 2665.
65. J. Brandrup and E. H. Immergut, Polymer Handbook, Wiley, edn 2, 1975.
66. H. Du and J. Zhang, Soft Matter, 2010, 6, 3370.
67. X. Qi, X. Yao, S. Deng, T. Zhou and Q. Fu, J. Mater. Chem. A, 2014, 2, 2240.
68. W. Wang, H. Lu, Y. Liub and J. Leng, J. Mater. Chem. A, 2014, 2, 5441.
69. X. Gu and P. T. Mather, RSC Adv., 2013, 3, 15783.
70. H. Kalita, M. Mandal and N. Karak, J. Polym. Res., 2012, 19, 1.
71. H. Lv, J. Leng, Y. Liu and S. Du, Adv. Eng. Mater., 2008, 10, 592.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra11776g

---