Facile Mechanochemical Preparation of Polyamide-derivatives via Solid-state Benzoxazine-isocyanide Chemistry

With the exploration of novel sustainable protocol for functional polyamides’ (PAs) construction as the starting point, herein, the small molecular model compound (M1-ssBIC) was prepared firstly by manual grinding of monofunctional benzoxazine (1a) and isocyanide (1b) via solid-state benzoxazine-isocyanide chemistry (ssBIC) to evaluate the feasibility of ssBIC. Linear PAs (P1-series polymers) were subsequently synthesized from biunctional benzoxazine (2a) and isocyanide (2b), and the influence of the loading of catalyst (octylphosphonic acid) (OPA) on the polymerization was investigated. Afterwards, two kinds of cross-linked PAs were successfully constructed via ssBIC by using trifunctional benzoxazine (3a) and cross-linked polybenzoxazine (4a) as reaction substrates, respectively, thus verifying the adaptability of ssBIC. Structural characterization indicates that amide, phenolic hydroxyl and tertiary amine substructures, with metal-complexing capability, have been successfully integrated into the obtained PAs. A type of representative PA/silver composite (P3-AgNPs) was prepared subsequently via in situ reduction treatment, and its application as recyclable reduction catalyst for organic pollutant p-nitrophenol (4-NP) was preliminarily investigated here to provide the example for possible downstream application of ssBIC. We think that this current work could provide a new pathway for the construction of functional PAs through facile and sustainable ssBIC protocol.


INTRODUCTION
Compared with the liquid-state reaction, the solid-state reaction (SSR) (usually categorised as typical mechanochemical reaction (MCR)) shows the advantages of being independent of the solubility of reaction substrates/intermediates, no need to use solvents, easy scaling up and high reaction speed, etc. Therefore, SSR has attracted intensive attention in recent years to act as a sustainable synthetic approach. [1−4] Up to now, SSR has been successfully utilized in reactions of Scholl, [5] Sonogashira, [6] Michael, [7] Click, [8] and Ritter, [9] etc. Despite the high efficiency of SSR in the preparation of small molecular compounds, there are much fewer reports regarding the preparation of polymers by SSR. In general, mechanical force has been considered as the destruction power for polymer structure. [10−12] The utilization of mechanochemical activation to achieve the growth of polymer chain in solid-state has limited success. [13−19] The exploration of novel solid-state polymerization protocol will provide effective supplement for the development of polymer chemistry.
As one of the typical engineering plastics, the synthesis and application of polyamides (PAs) have been extensively studied. PAs can be used as structural materials with excellent mechanical performance, and have certain application prospect in the fields of functional materials. [20−22] However, traditional syntheses of PAs are mainly carried out by solution/interfacial polymerization in liquid state, and there are only scarce reports on the preparation of PAs by SSR. [23−25] It can be seen from these reports that to realize the preparation of PAs via SSR, it is inevitably to adopt highly reactive, multifunctional aromatic acyl chloride as monomers to propagate with primary amines. It is well known that the preparation, purification and storage of acyl chloride are difficult to be implemented, and sometimes strong alkali solutions have to be introduced as acid-binding agent in SSR preparation of PAs. [23−25] These aspects have brought significant limitations to structural design and subsequent performance modulation of PAs via SSR. Thus, the exploitation of new SSR for the preparation of PAs, which can be carried out under mild condition and with good substrate designing flexibility, is desirable. In 2016, Juaristi et al. reported the preparation of a series of small-molecular amides through mechanical forceactivated Ugi and Passerini multicomponent reactions. [26] This work provided a new idea for the preparation of amides by MCR, but their work did not involve the construction of polymeric amides. Recently, our group exploited novel benzoxazine-isocyanide chemistry (BIC) to act as a powerful tool to construct PAs, and this reaction system displayed some attractive advantages, such as high flexibility in substrates designing, mild reaction condition, practical simplicity, and insensitivity to water/oxygen. [27−29] However, the above-mentioned BIC reactions were implemented in solution state, and the environmentally unfriendly solvent, chloroform, was used as the solvent. [27−29] The characteristics of BIC made it possible to be an ideal candidate to realize SSR preparation of PAs. To the best of knowledge, the relevant research has not been reported yet.
Based on the above research background, as a continuation of our recent work regarding BIC, in this study the feasibility of solid-state BIC (ssBIC) was testified firstly by the synthesis of model compound (M1-ssBIC). Subsequently, linear PA (P1) was successfully prepared via ssBIC to verify its adaptability in polymer construction, and the feed loading of catalyst (OPA) in ssBIC process was optimized. On this basis, the corresponding cross-linked polymer (P2) was prepared by ss-BIC [3+2] polymerization between trifunctional benzoxazine (3a) and bifunctional isocyanide (3b) monomers. In addition, an insoluble cross-linked polybenzoxazine (4a) was used as the precursor to obtain corresponding PA (P3) via ssBIC postpolymerization, proving the excellent universality of ssBIC. Structural characterization implies that some functional segments (phenolic hydroxyl, tertiary amine and amide), which possess good metal complexing capability, have been simultaneously introduced into PA's structure after ssBIC process. Based on this, P3 was selected as the representative polymer to act as the platform for the in situ preparation of polymer/Ag nanoparticles composite (P3-AgNPs). Further study revealed that P3-AgNPs can act as active agent to catalyse the chemical reduction of organic pollutant p-nitrophenol (4-NP).

Materials and Instruments
Organic solvents, such as 1,4-dioxane, THF and toluene were dried and distilled prior to use. 4,4′,4″-Trihydroxytriphenylmethane, aniline, 4,4′-diamino diphenyl, paraformaldehyde, octylphosphonic acid (OPA), p-nitrophenol (4-NP) and other reagents were purchased from Adamas or Aladdin and used as received. Benzoxazines (1a and 2a) and isocyanides (1b−3b) were prepared according to the reported procedures. [28] 1 H-and 13 C-NMR spectra were recorded on a Bruker AVANCE III HD NMR spectrometer at 400 MHz and 101 MHz in CDCl 3 using tetramethylsilane (TMS, δ=0) as internal standard. Solid-state 13 C-NMR (cross polarization/magic angle spinning, CP/MAS) spectra were obtained using a JNM-ECZR 400 MHz NMR spectrometer. FTIR spectra were recorded on a WQF-520 FTIR spectrometer with samples as thin films on KBr pellets. High resolution mass spectra (HRMS) were measured on a Waters Q-TOF Premier mass spectrometer. Relative weightaverage and number-average molecular weights (M w and M n ) and polydispersity indexes (M w /M n ) of polymers were estim-ated by an Aglient 1260 gel permeation chromatography (GPC) system equipped with a UV detector (eluent is THF, at a flow rate of 1.0 mL/min, with polystyrene as reference). Morphologies were studied by scanning electron microscopy (SEM) using Quanta 650FEG, and high-resolution transmission electron microscopy (HR-TEM) using JEM-2100F instrument. X-ray photoelectron spectroscopy (XPS) results were recorded with Thermo Scientific Escalab250Xi photoelectron spectrometer. Inductively coupled plasma mass spectroscopy (ICP-MS) was conducted on Agilent ICPOES730. Nitrogen physisosorption experiments were performed on an TriStarII 3020. Specific surface areas were calculated using the equation from Brunauer-Emmett-Teller (BET) in a relative pressure range. Pore size distributions were calculated using the density functional theory (DFT) method. UV-visible absorption spectra were recorded on SHIMADZU UV-1800 spectrophotometer. Dynamic light scattering (DLS) characterization was obtained by Brookhaven BI-200SM.

Preparation of P3-AgNPs
P3 (80 mg) and 2 mL of methanol were added to the reaction tube, and sonicated for 10 min. Subsequently, 10 mL of deionized water was added to the mixture, and sonicated again for 10 min. The powder was uniformly dispersed in the mixed solvent to form a suspension. AgNO 3 (79 mg, 0.46 mmol) was dissolved in water (4 mL) and was added dropwise to the mixture and stirred for 12 h. The product was centrifuged and washed respectively three times with deionized water and ethanol, and dried under vacuum at 50 °C overnight to give an intermediate (P3-Ag + ).
P3-Ag + and 2 mL of methanol were added to the reaction tube, and sonicated for 10 min. Subsequently, 10 mL of deionized water was added in the mixture, and sonicated again for 10 min. A fresh solution of NaBH 4 (1 mol·L −1 , 10 mL) was added dropwise to the mixture and stirred at room temperature for 24 h. The product was centrifuged and washed respectively three times with deionized water and ethanol, and dried under vacuum at 50 °C overnight to obtain grey-black powder (P3-AgNPs).

Feasibility Assessment of ssBIC
The main aim of this study is to evaluate the feasibility of ssBIC under ambient conditions (room temperature and without additional water/oxygen isolation). Based on this, the simplest and the most convenient mechanochemical protocol, manual grinding, was selected to implement ssBIC due to its operational convenience and can be implemented with relatively small amount of reaction substrate (as compared to ball milling). In order to verify the feasibility of ssBIC, model molecule (M1-ssBIC) was prepared firstly via ssBIC by using monofunctional benzoxazine (1a) and isocyanide (1b) as reaction substrates with OPA (20 mol%) as catalyst (Scheme 1). Thin-layer chromatography (TLC) monitoring indicated that ssBIC occurred rapidly at room temperature. After 15 min of manual grinding, the yield (purified) of M1-ssBIC (76%) was comparable to that of liquidstate BIC by using chloroform as solvent (reaction time is 6 h), [28] and further extension of grinding time endowed very slight influence to the yield. NMR (Fig. S1 in the electronic supplementary information, ESI) and TOF-MS ( Fig. S2  analyses of M1-ssBIC confirmed that it is consistent with the product obtained by solution method. To further verify the feasibility of ssBIC in the construction of polymers, bifunctional benzoxazine (2a) and isocyanide (2b) were used as monomers (with functionality ratio of 1:1), and OPA (addition of 20 mol%) as catalyst for ssBIC polymerization. After 15 min of manual grinding, the product was dissolved in THF and reprecipitated with ethyl ether. Unfortunately, the dispersion looked like a turbid emulsion, and the corresponding polymer was not obtained after filtration, suggesting that only oligomers were formed at this stage. We speculate that the relatively short reaction time did not provide sufficient contacting opportunities for active oligomer chain ends. The target polymer (P1-20) was successfully obtained after extending the reaction time to 1 h, while the yield was unsatisfactory (~16%). As reflected by Fig. 1, 1 H-NMR spectrum of P1-20 is consistent with reported result of polymer obtained by solution method. [28] By carefully analyzing Fig. 1 one can find that specific methylene signals, which correspond to N-CH 2 -O (circle) and N-CH 2 -Ar (square) in benzoxazine ring, appear at δ~5.30 and ~4.57 ppm, respectively, suggesting the presence of unreacted benzoxazine end groups in P1-20. We presume that this might be due to the insufficient encountering between catalyst (OPA) and benzoxazine, which stems from the difficulty in diffusion and migration of OPA in solid state. According to previous literatures, compared with traditional solution method, the solid-state protocol sometimes required much more catalyst to promote the extent of reaction. [14,17] Based on this, the influence brought by the feed ratios of OPA (from 20 mol% to 100 mol%) was investigated. Experimental results (Table S1 in ESI) show that with the increase of OPA, the yields of P1 gradually raise. When the feed of OPA is increased to 100 mol%, the yield of P1-100 reaches ~60% (Table S1 in ESI), which is slightly higher than that of the liquid phase polymerization system (50%). [28] 1 H-NMR spectrum of P1-100 ( Fig. 1) reveals that the benzoxazine methylene signals almost vanish, suggesting that the participation of monomers in ssBIC has been effectively improved by increasing the dosage of OPA.
The effect of OPA addition on the molecular weight of P1 was also explored. The molecular weights of corresponding polymers were analysed by GPC, and the results are summarized in Table S1 (in ESI). GPC results reveal that all polymers possess satisfactory molecular weight (with M n >8000). Molecular weight dispersity index (PDI) values of P1-20 and P1-40 are 4.73 and 3.66, respectively, which are much greater than PDIs of other three polymers. This indicates that there are molecular chains with considerable discrete chain lengths in P1-20 and P1-40, which might be brought by the insufficient OPA dosage. When the OPA dosages surpass 60%, PDI values tend to be stable. M n (1.27×10 4 ) of P1-100 was comparable to the corresponding value of the product prepared in liquid phase. [28] Based on this, 100 mol% addition of OPA was selected as a catalyst dosage for subsequent ssBIC experiments.

Preparation of Cross-linked PAs from ssBIC ssBIC [3+2] to construct cross-linked PA (P2)
After verifying the feasibility of ssBIC in the construction of linear PAs, we subsequently explored ssBIC's application in the preparation of cross-linked PAs. In order to construct cross-linked PAs via ssBIC, we need firstly to prepare multifunctional (with functionality>2) benzoxazine or isocyanide monomers. Here 4,4′,4″-trihydroxytriphenylmethane (1), aniline (2) and paraformaldehyde were used as reaction substrates to prepare a novel trifunctional benzoxazine (3a) (Scheme 2), and the bifucntional isocyanide (3b) was chosen as comonomer. The structure of 3a was analysed by 1 H-and 13 C-NMR (Figs. 2A and 2B). As shown in Fig. 2(A), methylene proton signals of benzoxazine ring (N-CH 2 -O (a) and N-CH 2 -Ar (b)) appear at δ~5.32 and ~4.53 ppm, [30] respectively, and the proton signal (c) of methine in triphenylmethane appears at 5.22 ppm. There are six groups of aromatic proton signals in the low field (7.25−6.66 ppm), and their respective assignments are shown in Fig. 2(A). Further analysis reveals that the integral of each proton signal is consistent with its structure. As shown in Fig. 2 Cross-linked PA (P2) was synthesized via ssBIC [3+2] reaction (Scheme 2). Solubility test revealed that P2 was insoluble in common solvents (toluene, CHCl 3 , DMSO, THF, etc.), which is reasonable due to its cross-linked molecular skeleton. Chemical structure of P2 was analyzed by FTIR spectrum subsequently (Fig. 3a). As shown in Fig. 3(a), distinct from 3a and 3b, FTIR spectrum of P2 displays distinctive bands at ~1673 and ~3279 cm −1 , which belong to the stretching vibration of amide (O＝C-NH) and phenolic hydroxyl (-OH) groups, respectively. In addition, as compared to that of 3a, the characteristic absorption of benzoxazine ring (~940 cm −1 ) [31] in P2 significantly weakens. Besides, there is no appearance of -NC band at ~2120 cm −1 in the FTIR spectrum of P2, [32,33] indicating that benzoxazine/isocyanide groups sufficiently participated in ssBIC. Parallel experiment of the preparation of P2 in solution state (P2-solution) was also carried out, and one can note that the FTIR spectra (Fig. S4 in ESI) of P2-ssBIC and P2-solution highly resemble each other, suggesting the feasibility of ssBIC in the construction of cross-linked PA. By carefully analysing the FTIR spectrum of P2-solution, one can find  the weak band of residual -NC at ~2120 cm −1 (Fig. S4 in ESI). In contrast to this, the -NC signal is hard to be detected in the FTIR spectrum of P2-ssBIC, suggesting that ssBIC displays superiority to liquid-reaction in the preparation of crosslinked polymer, owing to the ignorance of solubility issue during solid-state process. Solid-state 13 C-NMR (CP/MAS) spectrum of P2 suggests the presence of C＝O, C-O&C-N, aromatic carbons and methylene/methine carbons (Fig. S5 in ESI). XPS analysis was also carried out to investigate the chemical composition and state of elements contained in P2. As shown in Fig. 3(b), the full scan spectrum of P2 shows the presence of C, N and O ele-ments. High-resolution C 1s XPS spectrum (Fig. 3c) displays three obvious peaks at 284.4 eV (C-C/C-H), 285.6 eV (C-OH/C-N) and 287.7 eV (C＝O). [34,35] High-resolution N 1s spectrum (Fig. 3d) splits into two peaks at 400.3 and 399.7 eV, which can be assigned to amide (O＝C-N) and tertiary amine (C-N) groups, respectively. O 1s spectrum in Fig. 3(e) displays two obvious peaks at 532.8 eV (O＝C-N) and 531.4 eV (C-OH). The composition of each element is summarized in Table S2 (in ESI).
The microstructure characteristic of P2 was analysed by SEM and TEM. SEM (Fig. 4a) shows that P2 has a loose block structure, with rough surface and many wrinkles. As shown in  Fig. 4(b), TEM image of P2 exhibits agglomerated structure with bright and dark contrast, suggesting there might be preliminary pore structure inside. Figs. 4(c) and 4(d) display the isothermal adsorption-desorption and pore size distribution curves of P2. The BET specific surface area (SSA) of P2 was 11.5 m 2 ·g −1 , and the pore was dominated by mesopores (DFT method) (Fig. 4d). The small SSA of P2 might be due to the presence of flexible methine and amide substructures, which caused partial pore structure collapse during polymerization. [23] Constructing cross-linked polyamide (P3) by ssBIC post-polymerization In order to further verify the adaptability of ssBIC, an insoluble, cross-linked polybenzoxazine precursor (4a) was selected as the reaction substrate to investigate the feasibility of ssBIC in post-polymerization modification. 4,4′,4″-Trihydroxytriphenylmethane (1), 4,4′-diaminobiphenyl (3) and paraformaldehyde were used as staring materials to prepare a cross-linked polybenzoxazine (4a). Subsequently, 4a and 3b were reacted to construct the cross-linked PA (P3) by ssBIC (Scheme 2). FTIR analyses reflect the presence of amide (~1667 cm −1 ) and phenolic hydroxyl (~3222 cm −1 ) in the structure of P3, and the significant decrease of benzoxazine (~935 cm −1 ) [31] and -NC (~2120 cm −1 ) [32,33] groups relative to that of 4a and 3b, respectively. Solid-state 13 C-NMR (CP/MAS) spectrum of P3 (Fig.  S5 in ESI) resembles that of P2, which is resonable due to the simliar structural skeletons of both polymers. As the discussion for P2, XPS analysis (Figs. 5b−5e) of P3 indicates that P3 also contains amide, phenolic hydroxyl and tertiary amine groups. SEM image reflects that the surface of P3 (Fig. 6a) is rough and with many gullies and wrinkles. TEM image (Fig. 6b) of P3 exhibits a contrast between light and dark regions, indicating that P3 has preliminary pore structure. Figs. 6(c) and 6(d) are the isothermal adsorption-desorption and pore size distribution curves of P3. The BET results (Fig. 6c) show that the SSA of P3 is 23.7 m 2 ·g −1 , and the pore size distribution (Fig. 6d) is also dominated by mesopores. As can be seen from Table S2 (in ESI), P3 displays higher SSA and overall pore volume than those of P2, which might be due to the stabilization effect brought by the cross-linked skeleton structure of ssBIC precursor (4a).
The above analyses reveal that ssBIC can work smoothly even when using insoluble cross-linked polybenzoxazine as substrate. The summarization of mechanochemical preparation of polyamide-derivatives by us and other groups has been carried out here (Table S3 in ESI). As revealed by Table  S3 (in ESI), compared to available reports about mechanochemical preparation of PAs, the ssBIC polymerization protocol utilized in this work displayed advantages in wide range of monomers, high structural designing flexibility and no requirement for basic additives, etc.

Preparation and Characterization of P3-AgNPs
According to previous literatures, thanks to the containing of lone pair electrons on O/N atoms, molecular fragments such as amide, [36] phenolic hydroxyl [37,38] and tertiary amine [38,39]  display strong complexation with metal ions. These functional substructures have been successfully integrated into the target PAs by ssBIC (Scheme 2), suggesting that theses PAs might complex with metal ions and can be used as substrates to load metal nanomaterials by further in situ reduction treatment. Since P3 has rough surface (Fig. 6a) and relatively large SSA than P2, it was selected here to act as matrix for the loading of metal nanomaterials. Compared with other metal nanoparticles such as Au, Pt and Pd, Ag nanoparticles are relatively inexpensive and possess special chemical activity. Therefore, the preparation of P3/Ag nanocomposite (P3-AgNPs) was attempted. Firstly, Ag + and P3 were stirred in a mixed solvent (methanol:H 2 O=1:5, V:V) overnight to uniformly load Ag + on P3 (P3-Ag + ). The reaction mixture was centrifuged to remove free Ag + , and the corresponding P3-AgNPs were obtained by subsequent in situ reduction of P3-Ag + by NaBH 4 . As shown in FTIR spectra of P3 and P3-AgNPs (Fig. 7a), relative to that of pristine P3 (~1662 cm −1 ), the absorption of amide group in P3-AgNPs (~1685 cm −1 ) displays ~23 cm −1 shift to the higher wavenumber region, [25,27] hinting that the amide groups in P3 play an important role in the loading of AgNPs. Further XPS analysis provides more direct evidence for the successful loading of AgNPs. Full scan XPS spectrum of P3-AgNPs displays an obvious signal at 368.11 eV (Fig. 7b), which belongs to the   signal of Ag 3d. [39−41] High-resolution XPS spectrum of Ag 3d (Fig. 7c) clearly shows peaks at 368.4 and 374.4 eV, which can be assigned to Ag 3d 3/2 and Ag 3d 5/2 . The spin energy difference between these two peaks is ~6.0 eV, indicating that Ag exists in the zero-valent state here. [39−41] Moreover, XRD pattern of P3-AgNPs (Fig. 7d)  As shown in the SEM image of P3-AgNPs (Fig. 8a), P3-Ag-NPs were formed by the accumulation of microplates, and with rough surface. DLS characterization tells that the average diameter of P3-AgNPs is ~1062 nm (Fig. S6 in ESI). In TEM image (Fig. 8b) one can clearly observe that AgNPs have been loaded in polymer matrix. No detached free AgNPs were observed in TEM image (Fig. 8b), suggesting the strong affinity between P3 and AgNPs, which might be due to the presence of metal-complexing amide, hydroxyl and ternary amine segments in P3's structure. Locally enlarged TEM image (Fig. 8c) shows that AgNPs with diameters of 8 ± 3 nm (evaluated by one hundred randomly selected points) uniformly distribute in polymer platform. As shown in HR-TEM image of P3-AgNPs (Fig. 8d), the lattice spacing of AgNP equals 0.233 and 0.209 nm, corresponding to (111) and (200) plane of silver, respectively. [39] The corresponding selected area electron diffraction (SAED) pattern is shown in Fig. 8(e), which reflects the polycrystalline nature of AgNPs. [39] ICP-MS analysis tells that the composition ratio of silver in P3-AgNPs is ~5.78%.

Catalytic Reduction of 4-NP by P3-AgNPs
AgNPs have excellent catalytic reduction activity and have been applied to realize the chemical degradation of dyeing wastewater and nitro compounds. [37,39,40] It can be seen from above analyses that AgNPs have been uniformly loaded in P3 platform, and the presence of inert polymer substrate might be helpful to improve AgNPs' catalytic stability and thus can afford recyclable catalyst. Here we selected the toxic p-nitrophenol (4-NP) as the degradation target, and used NaBH 4 as the reducing agent to investigate the catalytic reduction activity of P3-AgNPs. With the introduction of aqueous NaBH 4 solution, 4-NP solution shows yellow colour due to the formation of p-nitrophenolate ion, and its concentration can be monitored by the absorption at ~400 nm. [30,42] As shown in Fig. 9(a), after the addition of P3-AgNPs, with the extension of reaction time, the absorption at ~400 nm gradually decreases. At the same time, the absorption of reduction product, 4-aminophenol (4-AP), appears at 300 nm, [30] and its intensity increases continuously. The apparent color of reaction system gradually changes from yellow to colourless (inset of Fig. 9a). After 8 min of reaction, the absorption at ~400 nm almost disappears, indicating that 4-NP has been completely reduced. Parallel experiment tells that NaBH 4 alone cannot reduce 4-NP and the apparent color retains after stirring at room temperature overnight, indicating that P3-AgNPs have good catalytic reduction activity for 4-NP.
The catalytic kinetic of 4-NP by P3-AgNPs was evaluated by a pseudo first-order kinetic equation as: where k is the rate constant (min −1 ), t is the reaction time (min), c 0 and c t are the initial and t time concentrations of 4-NP, respectively. As shown in the ln(c t /c 0 )-t relationship curve (Fig. 9b), the rate constant is evaluated to be 0.816 min −1 , indicating that P3-AgNPs have high catalytic activity. The catalytic stability of P3-AgNPs was evaluated by catalyst recovery experiments. As shown in Fig. 9(c), the catalytic efficiency maintains ~92% of the initial value after 5 cycles of catalyzing/recovery process, indicating that the presence of polymer substrate effectively improved the chemical stability of AgNPs.

CONCLUSIONS
In summary, in this study the small molecular model amide compound (M1-ssBIC) was synthesized firstly by ssBIC, verifying the high efficiency of ssBIC (yield 76% in 15 min). Subsequently, the preparation of linear PA (P1) was attempted via ssBIC between bifunctional benzoxazine/isocyanide comonomers. With 100 mol% OPA as catalyst, P1 was successfully obtained after grinding for 1 h in 60% yield, and with satisfactory molecular weight (M n =1.27×10 4 ) and PDI (2.02), proving that ssBIC was equally applicable to the construction of PA.
Multifunctional benzoxazines were used subsequently as reaction substrates to construct cross-linked PAs (P2 and P3) by ssBIC, further testifying the universality of ssBIC to construct PAs. To exploit the possible application of the synthesized PAs, P3 was selected as the substrate to prepare polymer/AgNPs composites (P3-AgNPs) by in situ reduction treatment. We found that P3-AgNPs can act as efficient and recyclable catalyst for the reduction of 4-NP. The ssBIC reported here displayed advantages of simple operation, mild condition, high efficiency, high substrate designing flexibility and with no solubility limitation, which can act as prospective pathway for sustainable preparation of functional PAs.

Electronic Supplementary Information
Electronic supplementary information (ESI) is available free of charge in the online version of this article at https://doi.org/10.1007/s10118-021-2510-6.