Multi-stimuli responsive and multi-functional oligoaniline-modified vitrimers

Introducing oligoaniline into a vitrimer resulted in a smart material that simultaneously responds to six different stimuli and performs six different functions.


Electrochemical experiments of ACAT-Vitrimer
The electrochemical measurements were operated on CHI 760D potentiostatgalvanostat (CH Instruments Inc.) using typically threeelectrode electrochemical cell. The cyclic voltammetry data was measured in a 0.5 M H 2 SO 4 mixture solution of dimethyl sulfoxide (DMSO) and deionized water with a scan range of 0 V to 1 V and scan rate of 50 mV s -1 . A platinum electrode and a saturated calomel electrode acted as the counter electrode and the reference electrode, respectively. The Vol. (1M H 2 SO 4 )/Vol. (DMSO) ratio was 1:1.
The working electrode was obtained by spin-casting the aforementioned pre-crosslinking ACAT-Vitrimer mixture on the fluorine doped tin oxide (FTO) conductive glass with the spinning rate of 500 r min -1 for 15 s and 1000 r min -1 for 30 s then curing 2 h at 120ºC, 1 h at 140ºC and 0.5 h at 160ºC.

Characterization of amino-capped aniline trimer (ACAT)
To confirm the chemical structure of ACAT, liquid phase 1 H nuclear magnetic resonance ( 1 H NMR) spectroscopy was performed on a JEOL-ECX400 spectrometer. ACAT was dissolved in deuterated dimethyl sulfoxide (10 mg ml -1 ), and the data were acquired at 25ºC.
Signals appeared at 5.4 ppm is attributed to the terminal amine protons (-NH 2 ) of ACAT, and the rest characteristic peaks at around 7.0-6.5 ppm are in accordance with the aromatic proton of ACAT (Fig. S1a). The attenuated total reflectance FTIR spectra were conducted on a Perkin Elemer spectrum 100 at room temperature ranged from 4000 to 650 cm -1 . The characteristic absorption peaks at 3302 cm −1 and 3194 cm −1 are resulted from the terminal -NH 2 . The characteristic absorption bands located at 1596 cm −1 and 1499 cm −1 belong to the vibrational bands of quinoid rings and benzenoid rings, respectively (Fig. S1b). The mass spectrum was run on a Shimadzu LCMS-IT-TOF high resolution mass spectrometry (HRMS). From Fig. S1c, we can see a molecular ion peak at 289.1443, which is corresponding to the molecular weight of ACAT.

Characterization of ACAT-Vitrimer
The swelling experiments were done using tricholorbenzene as the solvent. The volume of the film increased by 61% after 1 h at 100ºC and increased slowly after swelling 1 h at 140ºC, 180ºC and 20 h at 180ºC, resulting in a final volume increase of 115%. (Fig.   S2) The thermal property was measured on a differential scanning calorimetry (DSC) instrument TA Q2000 in conjunction with a controller and associated software to make up a thermal analysis system. The measure procedure consisted of two scanning cycle, where both the heating and cooling rate were 10ºC min -1 , and the first scanning was in order to eliminate the thermal history of the sample. As is shown in Fig. S3, the glass transition T g is about 40ºC (the midpoint of the transition) upon heating.
The mechanical properties were studied with a TA Q800 DMA instrument. In the stress-strain experiments at 25ºC, the force ramping rate was 0.4 N min -1 . Dilatometry test was performed under a constant load of 10 kPa in a temperature ramping rate of 3ºC min -1 to 250ºC. T v is calculated to be about 160ºC (Fig. S4).
Optical microscopy experiments were tested on a Nikon Eclipse LV100POL polarized light optical microscope, equipped with a Nikon Digital Sight DS-U3 digital camera and hot-stage.
The thermal stability of ACAT-Vitrimer was measured on a TA Q50 TGA under air and nitrogen atmosphere. (Fig. S5)

Investigation on the catalytic effect of ACAT
To examine whether the transesterification reaction can be catalyzed by oligoaniline, we conducted the stress relaxation experiments on different samples. As is shown in the stress relaxation experiments conducted at 180ºC (Fig. S6), the stress of the non-vitrimer type epoxy without ACAT only relax about 20% after 10 4 s, which is probably caused by the side reactions such as hydrolysis and alcoholysis of the ester groups at elevated temperature and pressure (Science, 2011, 334, 965-968). The relaxation time τ* of ACAT-Vitrimer with 1 mol% ACAT and ACAT-Vitrimer with 10 mol% ACAT is about 415 s and 724 s, respectively. The relaxation time τ* of the non-vitrimer type epoxy with 1 mol% ACAT and non-vitrimer type epoxy with 10 mol% ACAT is 4744 s and 3926 s, respectively. The molar ratio of TBD was 10 mol% to suberic acid, which was close to the molar ratio of ACAT (11 mol% to suberic acid) in non-vitrimer with 10 mol% ACAT. Therefore, ACAT indeed can catalyze the transesterification reaction, but the catalytic efficiency is much weaker than that of TBD.
We also conducted the recycling experiments to qualitatively justify the catalytic effect of ACAT. As is shown in Fig. S7, the non-vitrimer type epoxy without ACAT was hot-pressed for 30 min at 200ºC with a pressure of 5 MPa. The pieces can not be joined together. On the contrary, pieces of ACAT-Vitrimer with 1 mol% ACAT and ACAT-Vitrimer with 10 mol% ACAT could be easily reprocessed into intact films within only 10 min. Pieces of the non-vitrimer type epoxy with 1 mol% ACAT and non-vitrimer type epoxy with 10 mol% ACAT could also be combined together by hot-pressing but the surface of the recycled films are rough and has many visible cracks. The cracks on the recycled non-vitrimer type epoxy with 1 mol% ACAT are not only more but also larger than that of sample with 10 mol% ACAT, which means that the more ACAT, the better reprocessability.

TGA curve of the polymer coated silica gel
The polymer coated silica gel was obtained by curing the network on the solid support silica gel in one pot using the aforementioned method, where the mass ratio of polymer and silica gel was 1:2. As shown in Fig. S8, the mass percentage of polymer was 31.1%.

Fig. S8
TGA curve of the polymer coated silica gel in air (heating rate: 20ºC min -1 ).

The quantitative retention ability of ACAT-Vitrimer for copper ions
UV-Vis spectra were utilized to determine the retention ability of copper ions. The UV-Vis adsorption spectra of the copper (II) acetate monohydrate/THF solutions showed a characteristic peak at 674 nm and their intensities increased with the copper ions concentration linearly (Fig. S9a), based on which the standard curve was established (Fig. S9b). The polymer coated silica gel (0.5 g) was introduced to 10 ml of the prepared 5 mmol L -1 Cu (II)/THF solution. The mixture was stirred at 650 r min -1 at 25ºC for different time ranging from 2 min to 20 min. Then the mixture was filter and the filtrate was analysed with UV-Vis spectra. The numbers of moles adsorbed (N f ) per gram of polymer was calculated from the standard curve. The results showed that the saturation reached about just 10 min and the maximum adsorption N f was 0.10 mmol g -1 (Fig. S9c).

Shape recovery of different samples
The shape memory effect of the non-vitrimer type epoxy with 1 mol% ACAT was checked. As shown in Fig. S11a, the temporarily spiraled (80ºC in oven) film could quickly recover its permanent shape when heated at 80ºC again. Moreover, the stretched film could also revert to its original shape via irradiation with light (0.22 W cm -2 ) for just 10 s (Fig. S11b).

Healing experiments by direct heating and light irradiation
To figure out whether the difference between direct heating and optical healing in our system is expansion, we designed another experiment. Three films were cut deeply with a knife for cracks with comparable size. One film was irradiated and consequentially photothermally heated in a uniform fashion (Fig. S12a-I and II), leading to a concerted expansion of the film as a whole (as would happen inside an oven, Fig. S12a-III and IV), while the third film was only irradiated with a narrow beam at the site of the inflicted defect, causing only this area to be heated significantly, and thus resulting in a local expansion that pushes or squeezes the borders of the defect more closer together, giving a much more efficient healing (Fig. S12b).

Investigation of the effect of the copper ion absorption on the properties of ACAT-Vitrimer
To investigate the effect of copper ion absorption on the properties of ACAT-Vitrimer, we prepared the following materials: ACAT-Vitrimer with 1 mol% ACAT before and after copper ion absorption and ACAT-Vitrimer with 10 mol% ACAT before and after the copper ion absorption (swelled in 3 mmol L -1 Cu (II)/THF solution for 20 min). The glass transition temperature T g and transesterification temperature T v of these samples were individually studied. As shown in Fig. S14 to S17, the copper ions absorption lowers T g and the transesterification reaction rate of ACAT-Vitrimer. T g on heating and the relaxation time τ* were summarized in Table 1.

Fig. S14
DSC curves of ACAT-Vitrimer with 1 mol% ACAT before and after copper ions absorption.

Fig. S15
DSC curves of ACAT-Vitrimer with 10 mol% ACAT before and after copper ion absorption.

Fig. S16
Stress relaxation curves of ACAT-Vitrimer with 1 mol% ACAT before and after copper ion absorption.

Fig. S17
Stress relaxation curves of ACAT-Vitrimer with 10 mol% ACAT before and after copper ion absorption.

PH response experiments
To check whether ACAT-Vitrimer could maintain the vitrimer property after treated with base, we conducted the recycling experiments. As shown in Fig. S18, pieces of the sample could still be combined together by hot-pressing but the recycling efficiency was not as excellent as the untreated one (Fig. S7e). The reason for this is that the transesterification catalyst (TBD) is a stronger base than TEA so, in contrast to ACAT, TBD would probably not be fully deprotonated. Thus some of the processability would lose (for ACAT still maintain its catalytic effect). We also studied the pH response of the non-vitrimer type epoxy with the same contents of ACAT. The sample was first swelled in tetrahydrofuran (THF) to reach swelling equilibrium then successively placed in acid (1 M ptoluene sulfonic acid, PTSA) and base (1 M triethylamine, TEA) (Fig. S19). From THF to acid, the length change was about 17% and from acid to base, the length change was about 18%. The experiments were repeated for several times, which reveals that the nonvitrimer type epoxy also has excellent pH responsive property.

Fig. S18
Recycling experiments of ACAT-Vitrimer with 10 mol% ACAT after treated with acid then followed with base.

Swelling behaviors of the vitrimer with and without ACAT
The vitrimer films with and without ACAT were individually swelled in THF at room temperature. The equilibrium swelling ratios in length were 137% and 160%, respectively (Fig. S20).

Films used in the multiple shape memory construct
The three films A, B and C were representative of ACAT-Vitrimer, ACAT-Vitrimer swelled in 1 mmol L -1 Cu (II)/THF solution for 20 min and ACAT-Vitrimer swelled in 3 mmol L -1 Cu (II)/THF solution for 20 min, respectively. The temperature of the three films at the different light intensities, in the cold environment about -20ºC (placed in the liquid nitrogen vapor as show in Fig. S22), was measured using an infrared thermal imager (Fig. S23).