"Smart" polymers that degrade in response to external triggers have found applications in many elds, including drug delivery, transient electronics, encapsulation and sensing. [1][2][3][4][5][6] More recently, chain-shattering degradable polymers and selfimmolative degradable polymers attracted considerable attention because their backbones can be completely degraded into small fragments with high sensitivity to different types of triggers including pH, light, and redox agents. [7][8][9][10][11][12][13][14][15][16] However, both types of polymers require a stoichiometric amount of triggering agent and degradation rates are constant at best. Some selfimmolative polymers suffer from slow or incomplete breakdown as side reactions can occur as the degradation proceeds down the polymer backbone. An alternative approach uses autocatalysis degradation chemistry wherein a specic catalytic trigger generates additional triggers for acceleration of degradation. [17][18][19][20][21] Phillips and coworkers reported that ROMP polymers with appropriate pendant chains could exhibit dramatic changes in macroscopic properties through amplied, self-propagating side-chain reactions. 22,23 In particular, a global switch in hydro-phobicity and a change in the optical properties of a lm occurred with local stimulation. In an effort to develop polymeric materials that might degrade with accelerated rate proles and inspired by acid amplier small molecules, we recently reported poly(3-iodopropyl)acetals that breakdown liberating HI. 24 In essence such polymers carry the seeds of their own destruction, 25 with liberated acid catalyzing further cleavage in an autocatalytic loop. It is important to determine the generality of the autocatalytic polymer degradation strategy by developing breakdown pathways using other triggers such as base, light and redox agents. Herein we report a new polyurethane that undergoes self-amplied degradation mediated by base and further show that in analogous gels, a small localized addition of base leads to rapid long-range breakdo wn (Fig. 1).
As recently noted, base-degradable polymers are underdeveloped relative to acid-degradable polymers. 26,27 In designing auto-catalytic base-degradable polyurethanes, the base ampli-ers reported by Ichimura and others were considered. [28][29][30][31][32] Within this class of small molecules, the Fmoc protected carbamate offered a convenient aromatic scaffold for functionalization and the potential for conventional polyurethane synthesis. The actual polyurethanes studied, 1 and 1c, were prepared in six steps as shown in Scheme 1. Functionalization of the Fmoc aromatic ring was achieved through a Friedel-Cras acylation aer protection of alcohol with acetic anhydride, thus affording 4 or 5. Acidic deprotection and reduction with BH 3 $THF produced intermediate 6 and 7, which were further converted to diol monomer 8 and 9, respectively, by selectively reducing the benzylic alcohol functional group with Et 3 SiH. Traditional polycondensation was performed with a 1 : 1 ratio of diol monomer and hexylmethylene diisocyanate to afford polymer 1 and 1c.
As illustrated in Scheme 2, the addition of base can abstract the weakly acidic uorenyl methine proton on the polymer 1 backbone, followed by E1cB elimination and decarboxylation to generate a dibenzofulvene and stoichiometric amine that can catalyse additional cleavage reactions before or aer addition to the dibenzofulvene unit. The two polyurethanes 1 and 1c are identical structurally except that control polymer 1c is unable to undergo base-triggered degradation because the additional methylene group prevents the E1cB elimination from occurring. Both 1 and 1c were characterized by gel permeation chromatography (GPC) with DMF as the eluent (Fig. S6 andS7 †). Polymer 1 has a M n ¼ 22 kDa (Đ ¼ 2.1) and control polymer 1c has M n ¼ 11 kDa (Đ ¼ 2.6). The 1 H NMR was consistent with the expected structure of 1 (Fig. S3 †) and 1c (Fig. S4 †).
Thermal gravimetric analysis (TGA) of polymer 1 and polymer 1c revealed the onset of thermal degradation to occur around 120 C and 280 C respectively (Fig. S25 andS26 †) and the onset thermal temperature at 120 C of polymer 1 correlates well to what Simeunovic and his coworkers reported. 33 The T g of polymer 1 and 1c were determined to be 61 C and 45 C, respectively, the later value measured by differential scanning calorimetry (DSC) (Fig. S27 †).
Several bases were found to trigger the autocatalytic degradation of 1 (Fig. S11 †), with hexylamine chosen for further study because its basicity and steric hindrance is most similar to the amplied amine species. Thus, the base-triggered degradation of polymers 1 and 1c in DMF solution was initiated by the addition of hexylamine and monitored by gel permeation chromatography (GPC). When 1 was exposed to 5 mol% hexylamine (per repeat unit), it showed a progressive and signicant decrease in molecular weight over a 12 h period. As seen in Fig. 2a, the reduction in polymer size over time is nonlinear. Thus, the retention time of the 1 shis only 1 min during the rst 2 h but between 6 h and 9 h signicantly broadens and shis to longer times. In contrast, under the same conditions, the GPC of polymer 1c remained unchanged over 24 h (Fig. S8 †).
1 H NMR was used to monitor the molecular details of the degradation of polymers 1 and 1c in the presence of hexylamine in DMSO-d 6 solution. Consistent with the GPC study, no change in the NMR of 1c was observed over 24 h with 5 mol% hexylamine (Fig. S10 †). In the case of 1, addition of 5 mol% of hexylamine led to the simultaneous disappearance of the methine and methylene protons labelled a and b at d 4.33 and 4.16 ppm, respectively and the appearance of alkene protons at d 6.25 ppm from the dibenzofulvene elimination product (Fig. 2b and S9 †).
To determine how the concentration of the base trigger affects the rate of the degradation, quantitative 1 H NMRmonitored kinetics were carried out in the presence of 0.5 mol%, 1 mol%, 5 mol%, 20 mol% and 100 mol% hexylamine. As seen in Fig. 2c, a stoichiometric amount of hexylamine induced complete polymer degradation at room temperature within 1 h. The rate prole and time for complete degradation correlated with the amount of base trigger. Thus, with no added base the polymer was stable, whereas for 0.5 mol%, 1 mol%, 5 mol%, 20 mol% and 100 mol% hexylamine the degradation reached 90% at ca. 47 min, 2 h, 10 h, 12 h, and 15 h, respectively. Most excitingly was the observation that the three low hexylamine experiments (0.5 to 5 mol%) exhibited obvious induction periods and sigmoidal conversion curves indicative of autocatalytic degradation. Additional support for the autocatalytic, base amplication mechanism came from tting the degradation data of polymer 1 to an autocatalytic kinetic model (eqn (S3) †). 23,24,[34][35][36] In this model, rate constants k 1 and k 2 separately represent the nonautocatalytic and autocatalytic, amine-accelerated rate constants (see ESI † for details). Consistent with the mechanism shown in Scheme 2, tting the sigmoidal curves seen in Fig. 2c, led to k 2 values that were quite close and k 2 c 0 values that are larger than the k 1 values (Table 1). The latter is especially true for the 0.5 mol% hexylamine run, in which the k 2 c 0 (6.7 Â 10 À3 min À1 ) is 30 times larger than k 1 (2.1 Â 10 À4 min À1 ). This larger k 2 c 0 value is characteristic of an autocatalytic reaction (Table 1 and Fig. S14-S16 †).
To further characterize the degradation of 1, liquid chromatography coupled mass spectrometry (LC-MS) was utilized to identify the major degradation products, and further indicate the chemical structure of the polymer repeating units. Analysis of the degradation products from polymer 1 through LC-MS revealed two major peaks, degradation product 1 with higher intensity appearing at 6.4 min (m/z ¼ 393.4) and degradation product 2 with a lower intensity appearing at 9.3 min (m/z ¼ 669.4) (Fig. S12 †). These products are consistent with two types of repeating units in 1 (Fig. 3) and degradation product 1 Polyurethanes are important and widely used polymeric materials commonly found in plastics, adhesives and coatings. 37,38 Unlike polymer 1, these materials are usually prepared from a polyol that produces cross-linking. The combination of cross-linking and the stability of the urethane linkage makes poly-urethanes highly durable but also limits their end-of-life break-down. To examine whether the base-amplied degradation might be applicable to bulk materials, triol 6 was prepared (see ESI †) and polymerized with hexamethylene diisocyanate and dibutyltindilaurate (DBTDL) as catalyst in N-methylpyrrolidone (NMP) with bromothymol blue present to visualize the gel and provide a pH indicator (Fig. 4b and S17 †). The polymerization was performed at room temperature in a circular Teon mold for 24 h to give a polyurethane lm of 11 with a 500 mm thickness.
To characterize the polymer lm, it was immersed in additional NMP which induced signicant swelling, but did not dissolve the gel. This observation is consistent with a crosslinked gel. To demonstrate the urethane network, the polymer lm was dried under high vacuum and characterized by attenuated total reection infrared spectroscopy (ATR-FTIR). The absorption peak at 1694 cm À1 and 1252 cm À1 were assigned to the urethane structure and the absorption peak at 3326 cm À1 was assigned to unreacted hydroxyl groups in the polymer network (Fig. S18 †). 39 Degradation study of the polymer lm was performed with lm being swelled by NMP solution. In the degradation study, the centre of the polymer lm changed from yellow to blue aer 2 mL 180 mM hexylamine NMP solution was added in the centre and photographs were acquired over time (Fig. 4c). It was observed that the degradation area kept increasing, producing a deep blue colour, suggesting the formation of increasing numbers of terminal amino groups with conversion of the bromothymol blue pH indicator to its blue colour ring open form. Quantication of the degradation area with blue colour change for the polymer lm was assisted by the Image-Pro Plus (Fig. 4d). An increase in degradation area also simulates a sigmoidal curve, with a nonlinear increase from 10% at 100 min to 90% at 300 min, which is consistent with an autocatalytic degradation process for the crosslinked gel. The The degradation process was also monitored by rheology. The storage modulus of the gel was measured and no major rheological change was observed from the polymeric network without addition of the base trigger (blue curve, Fig. 4e and S23 †). However, the bulk polymeric network underwent a rapid decrease in storage modulus from about 5300 Pa to nearly 0 Pa upon addition of a very small amount of a dilute hexylamine solution in NMP at room temperature (red curve, Fig. 4e andS23 †). In this case autocatalytic equations (eqn (S4) and (S5) †) that relate the storage modulus to degradation time were utilized to quantify the gel breakdown kinetics as described in more detail in the ESI. † In particular, these equations relate the storage modulus decrease to the concentration of crosslinks, thus enabling inference of apparent chemical rate constants. The tting of triplicate runs (Fig. S24 †) gave k 2 c 0 ¼ 15.9 AE 5.3 min À1 , which is much larger than the k 1 ¼ 2.1 Â 10 À3 AE 1.1 Â 10 À3 min À1 . These observations are consistent with an autocatalytic degradation process.
In conclusion, we developed a new type of self-amplifying degradable polymer with self-accelerating degradation properties using the well-developed base-sensitive Fmoc protecting group used in peptide synthesis. The incorporation of Fmoc in every repeating unit provides extremely sensitive polymeric materials with a small amount of base leading to rapid and amplied degradation. The base amplication process may be useful in applications where rapid production of an amine base is desirable. The crosslinked gel provides a rare example where a tiny local stimulation generates long range, rapid macroscopic degradation. In principle such a degradation might propagate over very large distances. Our current efforts are focused on generalizing this self-amplied degradation process to other kinds of triggers such as light, ions, and ROX agents.