Self-immolative Amphiphilic Diblock Copolymers with Individually Triggerable Blocks.

Self-immolative polymers are a growing class of degradable polymers that undergo end-to-end depolymerization after the stimuli-responsive cleavage of an end-cap or backbone unit. Their incorporation into amphiphilic block copolymers can lead to functions such as the disintegration of copolymer nanoassemblies when depolymerization is triggered. However, diblock copolymers have not yet been developed where both blocks are self-immolative. Described here is the synthesis, self-assembly, and triggered depolymerization of self-immolative block copolymers with individually triggerable hydrophilic and hydrophobic blocks. Neutral and cationic hydrophilic polyglyxoylamides (PGAm) with acid-responsive end caps were synthesized and coupled to an ultraviolet (UV) light-triggerable poly(ethyl glyoxylate) (PEtG) hydrophobic block. The resulting block copolymers self-assembled to form nanoparticles in aqueous solution, and their depolymerization in response to acid and UV light was studied by techniques including light scattering, NMR spectroscopy, and electron microscopy. Acid led to selective depolymerization of the PGAm blocks, leading to aggregation, while UV light led to selective depolymerization of the PEtG block, leading to disassembly. This self-immolative block copolymer system provides an enhanced level of control over smart copolymer assemblies and their degradation.

Introduction

Block copolymers are macromolecules comprising covalently linked, chemically distinct homopolymers. Often, the homopolymer blocks are immiscible and undergo microphase separation, leading to a wide array of ordered structures with interesting properties and applications.[1,2] For example, poly(styrene-block-butadiene-block-styrene) organizes into polystyrene domains arranged in a rubbery polybutadiene matrix, with the high glass transition temperature (Tg) styrene domains providing physical cross-links below the Tg.[3] The resulting thermoplastic elastomers are used in a variety of applications from packaging to footwear. In solution, amphiphilic block copolymers such as poly(ethylene oxide-block-propylene oxide-block-ethylene oxide) have been used as surfactants and thickening agents in the pharmaceutical and cosmetics industries.[4] More recently, block copolymer self-assembly in the solid state has been used in the preparation of mesoporous materials for energy storage and conversion applications.[5] Furthermore, there has been substantial interest in the development and study of block copolymers in solution for nanomedicine.[6]

With increasing demands on block copolymer assemblies to perform specific functions such as transformations in morphology[7,8] or triggered release of molecules,[9,10] there has been growing interest in the incorporation of stimuli-responsive polymers into block copolymers and their assemblies. For example, polymer blocks can be reversibly switched between hydrophilic and hydrophobic states through changes in pH,[11,12] temperature,[13,14] or CO2 concentration.[15,16] Chemical degradation induced by stimuli such as light[17,18] or acid[19,20] can also be used to degrade individual blocks or the linkages between blocks.

While typical stimuli-responsive polymers respond to stimuli by undergoing one or a small series of bond cleavage events at a time, self-immolative polymers (SIP) are a class of polymers that undergo an extended cascade of reactions leading to end-to-end depolymerization following a single stimulus-mediated cleavage of an end-cap or backbone unit.[21,22] Examples of SIP backbones include poly(benzyl carbamate/carbonate)s,[23−25] poly(benzyl ether)s,[26] polyphthalaldehydes,[27] polyglyoxylates,[28] and more recently polyenynes[29] and polydisulfides.[30,31] Thus far, there are several examples involving the incorporation of SIPs into block copolymers. For example, block copolymers composed of polycarbamates,[24,32−35] polyglyoxylates,[28,36−38] or polyglyoxylamides[39] as the hydrophobic blocks and poly(ethylene glycol) (PEG) or polyacrylamide as the hydrophilic blocks have been synthesized and self-assembled into micelles and vesicles. Triggering depolymerization of the SIP blocks led to disintegration of the assemblies and in some cases the release of encapsulated drugs and dye molecules. However, to the best of our knowledge, self-immolative block copolymers with both the hydrophilic and hydrophobic blocks being depolymerizable have not yet been reported. Such systems are of interest from a fundamental perspective to understand the influence of depolymerization on block copolymer self-assembly, but also for applications where it is desirable to trigger the release of cargo or control the aggregation state of assemblies.

We describe here the preparation and study of diblock copolymers where both blocks are self-immolative and can be independently triggered to depolymerize in response to orthogonal stimuli (Figure A). To achieve this, we used the poly(ethyl glyoxylate) (PEtG) SIP platform, as the pendent groups can be easily changed through postpolymerization amidation reactions to produce polyglyoxylamides (PGAms), thereby altering the solubility and charge of the polymer, while retaining the depolymerizable polyacetal backbone (Figure B).[40] An ultraviolet (UV) light-triggerable hydrophobic PEtG block and four different hydrophilic blocks having acid-cleavable end caps and either neutral hydroxyl or cationic pendent amino groups were prepared. These blocks were coupled to obtain a series of amphiphilic block copolymers, which were self-assembled to form nanoparticles under aqueous conditions. It was demonstrated that the hydrophilic and hydrophobic blocks could be triggered in an orthogonal manner, leading to either aggregation or disintegration of the assemblies, respectively.

Figure 1

(A) Schematic showing how triggering the depolymerization of the hydrophobic block leads to disintegration of fully self-immolative block copolymer assemblies, whereas depolymerization of the hydrophilic block leads to aggregation. (B) General depolymerization mechanism for PEtG/PGAms.

Experimental Section

General Materials

Additional synthesis procedures are included in the Supporting Information. All reactions were performed under a N2 atmosphere using Schlenk techniques. All glassware for PEtG synthesis were predried in a 165 °C oven for at least 16 h before use. The glassware was rapidly set up for the reactions when they were hot and immediately pumped for 5 min using an Edwards RV5 rotary vane pump and then filled with dried N2, which was repeated two more times. Ethyl glyoxylate (EtG) solution (50% in toluene) was obtained from Alfa Aesar and purified as previously described.[41] CaH2, Nile red, n-butyllithium (nBuLi) solution (2.5 M in hexanes), lithium bis(trimethylsilyl)amide ((TMS)2NLi), 2-nitrobenzyl alcohol, 2-aminoethanol, and N,N-dimethylethylenediamine were obtained from Sigma-Aldrich and used as received. Propargyl alcohol was obtained from Sigma-Aldrich and freshly distilled over 4 Å molecular sieves immediately before use. Toluene was obtained from Caledon Laboratories and refluxed/distilled over sodium using benzophenone as an indicator. NEt3 was purchased from EMD Millipore Corp., stored over CaH2, and then refluxed for 2 h before it was distilled prior to use. Dialyses were performed using a Spectra/Por 3500 Da molecular weight cutoff regenerated cellulose membranes purchased from Fisher Scientific.

General Procedures

1H and 13C{1H} NMR spectra were recorded on a 400 MHz Bruker AvIII HD 400 instrument or Varian INOVA 600 MHz spectrometer. 1H NMR spectra were referenced to residual CHCl3 (7.26 ppm), DMSO-d (2.50 ppm), or DOH (4.79 ppm), and 13C{1H} NMR spectra were referenced to CDCl3 (77.2 ppm) or DMSO-d (39.5 ppm). Size-exclusion chromatography (SEC) was performed at 85 °C in DMF containing 10 mM LiBr and 1% v/v NEt3 using a Waters 515 HPLC pump with a Waters In-Line Degasser AF, two PLgel mixed D 5 μm (300 × 1.5 mm) columns connected to a corresponding PLgel guard column, and a Wyatt Optilab Rex refractive index (RI) detector. Number-average molar mass (Mn), weight-average molar mass (Mw) and dispersity (Đ) of the samples were determined relative to poly(methyl methacrylate) (PMMA) standards. FT-IR spectra were obtained using a PerkinElmer FT-IR Spectrum Two instrument with attenuated total reflectance sampling. Characterization by DLS was performed on a Malvern Zetasizer Nano ZS instrument from Malvern Instruments at 25 °C at a polymer concentration of 1 mg/mL. TEM imaging was performed using a CM10 microscope from Phillips at an accelerating voltage of 80 kV. The nanoparticle suspension was diluted to 0.3 mg/mL, deposited on a copper grid, and then dried under air for 24 h prior to imaging. The diameters of the particles were measured relative to the scale bar and are indicated as the mean ± standard deviation. Fluorescence spectra were obtained using a QM-4 SE spectrometer from Photon Technology International (PTI) equipped with double excitation and emission monochromators.

Synthesis of NB-PEtG-N3

To a 250 mL flame-dried Schlenk flask under a N2 atmosphere, 2-nitrobenzyl alcohol (115 mg, 0.750 mmol, 1.0 equiv) and (TMS)2NLi (126 mg, 0.75 mmol, 1.0 equiv) were mixed in 20 mL of dry toluene for 5 s. Purified ethyl glyoxylate (7.0 mL, 75 mmol, 100 equiv) was added to this flask, and the resulting solution was stirred for 20 min. It was subsequently cooled to −20 °C and stirred for 20 min. Freshly distilled NEt3 (0.73 mL, 5.3 mmol, 7.1 equiv) was then added to the polymerization flask, and the solution was stirred for 20 min. The end-cap 4-(azidomethyl)benzoyl chloride[42] (1.47 g, 7.5 mmol, 10 equiv) was dissolved in 5.0 mL of CH2Cl2, cooled to −20 °C, and added to the reaction flask. The reaction mixture was stirred at −20 °C for 24 h, gradually warmed to room temperature over 2 h, and then concentrated in vacuo. The resulting concentrate was then precipitated into 1 L of 5:1 methanol/water. Then, the flask was sealed and transferred into a −20 °C freezer where it was kept for 16 h before the solvent was decanted, and the resulting residues were vacuum-dried, affording 4.70 g of a colorless tacky solid. Yield: 61%. 1H NMR (CDCl3, 400 MHz): δ 8.15–8.05 (m, 4H), 7.45–7.40 (m, 4H), 5.75–5.50 (m, 78H), 4.43 (s, 2H), 4.10–4.30 (m, 154H), 1.20–1.40 (m, 231H). 13C{1H} NMR (CDCl3, 100 MHz): δ 165.7, 92.8, 62.2, 14.0 (note that end-cap peaks cannot be observed by 13C NMR). FT-IR: 2980, 2100, 1750 cm–1. SEC (DMF, PMMA): Mn = 7030 g/mol, Mw = 8780 g/mol, Đ = 1.25.

Synthesis of P-PEtG-T

To a flame-dried Schlenk flask under a N2 atmosphere, 20 mL of dry toluene, propargyl alcohol (29 μL, 0.50 mmol, 1.0 equiv), and 2.5 M n-butyllithium solution (0.20 mL, 0.50 mmol, 1.0 equiv) were added at 0 °C, and the resulting solution was then stirred at room temperature for 3.5 h. Purified ethyl glyoxylate (4.7 mL, 50 mmol, 100 equiv) was then added to this flask, and the solution was stirred at room temperature for 2 h. It was subsequently cooled to −20 °C and stirred for 30 min. Freshly distilled NEt3 (0.42 mL, 3.0 mmol, 6.0 equiv) was then added to the polymerization flask, and it was stirred for 20 min. Meanwhile, trityl chloride (0.84 g, 3.0 mmol, 6.0 equiv) was mixed with silver triflate (0.77 g, 3.0 mmol, 6.0 equiv) in 5.0 mL of dry CH2Cl2 and cooled to −20 °C. This mixture was added to the reaction flask and stirred at −20 °C for 30 min, then kept in a −20 °C freezer for 48 h. The reaction mixture was then allowed to gradually warm to room temperature over 2 h and concentrated in vacuo. The resulting concentrate was precipitated into 600 mL of 5:1 methanol/water. Then, the flask was sealed and transferred into a −20 °C freezer where it was kept for 16 h before the solvent was decanted, and the resulting residues were vacuum-dried, affording 3.65 g of a colorless tacky solid. Yield: 72%. 1H NMR (CDCl3, 400 MHz): δ. 7.49–7.40 (m, 10H), 7.35–7.25 (m, 6H), 5.70–5.50 (m, 83H), 4.10–4.40 (m, 165H), 2.51 (m, 1H), 1.20–1.40 (m, 247H). 13C{1H} NMR (CDCl3, 100 MHz): δ 165.7, 129.1, 127.7, 92.4, 62.2, 14.0 (note that end-cap peaks cannot be observed by 13C NMR). FT-IR: 2985, 1750 cm–1. SEC (DMF, PMMA): Mn = 6630 g/mol, Mw = 10 490 g/mol, Đ = 1.58.

Synthesis of P-PEtOH-T and General Procedure for the Synthesis of All Polyglyoxyamides

P-PEtG-T (1.00 g, 9.8 mmol of ester, 1.0 equiv) was dissolved in 10 mL of dry 1,4-dioxane in a 25 mL round-bottom flask. To this flask, 2-aminoethanol (2.96 mL, 49 mmol, 5.0 equiv) was added, and the reaction mixture was stirred at room temperature for 24 h. All volatiles were removed under vacuum conditions to give the crude product. This product was subsequently purified by dissolution in minimal MeOH and precipitation in cold CH2Cl2. After the liquid was decanted, the precipitate was dried in vacuo to afford 1.04 g of an off-white powder. Yield: 91%.1H NMR (D2O, 400 MHz): δ. 5.75–5.50 (m, 1H), 3.67 (br s, 2H), 3.35 (br s, 2H). 13C{1H} NMR (D2O, 100 MHz): δ.167.4, 94.0–96.5, 59.6, 48.8, 41.6. FT-IR: 3250, 3094, 2939, 1661, 1538 cm–1. SEC (DMF, PMMA): Mn = 9020 g/mol, Mw = 11460 g/mol, Đ = 1.27.

Synthesis of NB-PEtG-PEtOH-T and General Procedure for the Coupling of the Hydrophobic and Hydrophilic Blocks

To a 10 mL round-bottom flask, NB-PEtG-N (0.200 g, 29 μmol, 1.0 equiv) and P-PEtOH-T (0.258 g, 29 μmol, 1.0 equiv) were dissolved in 1 mL of DMF while stirring at room temperature. This solution was degassed by bubbling N2 nitrogen gas for 30 min. To this solution, sodium ascorbate (28 mg, 0.14 mmol, 5.0 equiv) and CuSO4 (23 mg, 0.14 mmol, 5.0 equiv) were added quickly. The mixture was stirred under a N2 atmosphere for 16 h. The mixture was then filtered through Celite, and the filtrate was dialyzed against DMF for 16 h. The DMF was removed in vacuo, and 0.33 g of an off-white solid was obtained. Yield: 72%. 1H NMR (DMSO-d, 400 MHz): δ 8.35–8.10 (m, 0.4H), 5.80–5.25 (m, 1H), 4.76 (br s, 0.4H), 4.13 (br s, 1.3 H), 3.19 (br s, 1H), 1.40–1.15 (m, 1.8H). 13C{1H} NMR (DMSO-d, 100 MHz): δ 167.7–162.9, 91.6–96.5, 62.2, 59.8, 42.0, 36.2, 31.3, 14.0. FT-IR: 3338, 2985, 2939, 1750, 1668 cm–1. SEC (DMF, PMMA): Mn = 15460 g/mol, Mw = 19020 g/mol, Đ = 1.23.

Nanoparticle Preparation

A 10 mg/mL solution of the block copolymer in DMSO was prepared with stirring overnight. Then, 0.1 mL of this solution was rapidly injected into 0.90 mL of rapidly stirring deionized H2O, and the resulting suspension was stirred for 10 min. Each nanoparticle system was prepared in triplicate.

Nanoparticle Depolymerization Studied by NMR Spectroscopy

The polymer was dissolved in DMSO-d6 at a concentration of 40 mg/mL. For the degradation of the polymers in response to pH, 50 μL of the resulting solution was rapidly injected into 450 μL of 10 mM, pH 7.0 phosphate-buffered D2O, and 50 μL was injected into 450 μL of D2O. Then, 2.6 μL of acetic acid was added (final concentration of acetic acid = 100 mM, pH 3.3). After stirring for 30 s, the nanoparticle suspension was transferred into NMR tubes, and initial 1H NMR spectra were obtained. The samples were stored in the dark, and NMR spectra were obtained at selected time points. For the samples triggered by UV light, 50 μL of the DMSO polymer solution was rapidly injected into 450 μL of 10 mM, pH 7.0 phosphate-buffered D2O, and then initial 1H NMR spectra were obtained. Irradiation with UV light was then performed using a 10 W LED light source with 1 A constant current, 400 Hz pulse, and a wavelength range of 365–370 nm for 30 min, and then NMR spectra were obtained at selected time points. Control samples were stored in the dark. To assess the extent of degradation, the sum of the integrals of the polymer backbone peak at 5.25–5.80 ppm (Ipoly) and the peak corresponding to the hydrate of the monomer depolymerization product (Idegra) at 5.08 ppm was set to 1.0 in total. The extent of depolymerization was then determined as % depolymerization = Idegra × 100%.

Nanoparticle degradation studied by DLS and TEM

Nanoparticles were prepared as above for the NMR experiment at a final concentration of 1 mg/mL, and the corresponding stimuli were also applied as described for the NMR experiment except that H2O was used instead of D2O. The count rate and Z-average diameters were measured by DLS while fixing the attenuator at 7. The degradation of all polymers was studied in triplicate. For the TEM studies, the samples were prepared and treated in the same manner as for the DLS studies, except that for pH 7, the pH was adjusted to 7 using NaOH instead of buffer to minimize the presence of inorganic salts.

Results and Discussion

Block Copolymer Synthesis

A UV light-responsive end cap was selected for the hydrophobic block. The polymerization of ethyl glyoxylate (EtG) was initiated by o-nitrobenzyl alcohol (1) after its deprotonation using (TMS)2NLi (Scheme ).[43] The polymer was end-capped with 4-(azidomethyl)benzoyl chloride (2)[42] to provide NB-PEtG-N. The resulting polymer had an Mn of 7030 g/mol and a Đ of 1.25 based on SEC in DMF relative to PMMA standards (Table ). 1H NMR spectroscopy showed the presence of both end caps in a ∼1:1 ratio (Figure S1). The molar masses of the polymers based on end-group analysis were of similar magnitude compared to those determined by SEC. IR spectroscopy showed an azide stretching peak at 2100 cm–1 (Figure S24).

Scheme 1

Synthesis of a Hydrophobic PEtG Block with a UV Light-Responsive End Cap and a Terminal Azide

Table 1

Molar Mass Data for the Homopolymers and Corresponding Block Copolymers As Determined by SEC in DMF, Relative to PMMA Standardsa

polymer	Mn (g/mol)	Đ
NB-PEtG-N3	7030 (8020)	1.25
P-PEtG-T	6630 (8470)	1.58
P-PEtG-MT	5600 (9790)	1.32
P-PEtOH-T	9020	1.27
P-PEtOH-MT	9040	1.14
P-PDMAE-T	9650	1.28
P-PDMAE-MT	9610	1.25
NB-PEtG-PEtOH-T	15460	1.23
NB-PEtG-PEtOH-MT	13500	1.36
NB-PEtG-PDMAE-T	18450	1.31
NB-PEtG-PDMAE-MT	12550	1.39

Values shown in parentheses for the PEtGs are from end-group analysis based on 1H NMR spectroscopy.

To prepare the hydrophilic blocks, trityl and 4-methoxytrityl end-caps were selected as they can be cleaved using acid,[44] a trigger which is orthogonal to that of the o-nitrobenzyl end cap on the hydrophobic block. The polymerization of EtG was initiated with propargyl alcohol (3) after its deprotonation by n-butyllithium (Scheme ). Then, end-capping was performed using either trityl chloride (4a) or 4-methoxytrityl chloride (4b) in the presence of silver triflate to provide P-PEtG-T and P-PEtG-MT, respectively. SEC in DMF indicated that P-PEtG-T had an Mn of 6630 g/mol and a Đ of 1.58, while P-PEtG-MT had an Mn of 5600 and a Đ of 1.32. Peaks were observed in the 1H NMR spectra corresponding to the propargyl initiator and terminal trityl groups (Figures S2, S3). Next, the hydrophobic ethyl ester pendent groups were replaced with either hydroxyl or amine-containing hydrophilic groups by amidation reactions using either 2-aminoethanol or N,N-dimethylethylenediamine, respectively, to provide P-PEtOH-T, P-PEtOH-MT, P-PDMAE-T, and P-PDMAE-MT. In each case, 1H NMR spectroscopy showed full conversion of the ethyl ester groups (Figures S4–S7), and only minor changes in molar mass and Đ were indicated by SEC (Table ).

Scheme 2

Synthesis of Hydrophilic Blocks Having Either Trityl or 4-Methoxytrityl End Caps and Pendent Tertiary Amines or Hydroxyls

Next, the hydrophobic block was coupled to the four alkyne-terminated hydrophilic blocks using Cu-assisted azide–alkyne cycloaddition reactions. For the hydroxyl-containing PEtOH blocks, conditions involving the use of CuSO4 and sodium ascorbate in DMF were used (Scheme A). However, for the amine-containing PDMAE blocks, Cu(I)Br performed better, presumably due to binding of the amines to Cu(II) (Scheme B). For each coupling product, 1H NMR spectroscopy showed the presence of both blocks (Figures S8–S11). IR spectroscopy showed a loss of the azide peak from NB-PEtG-N upon conjugation (Figures S24 and S25). Furthermore, SEC showed an increase in Mn and no evidence of uncoupled homopolymer blocks (Figures , S23). Particularly for the 4-methoxytrityl end-capped block copolymers, the increase in Mn was somewhat lower than would be expected based on adding the two blocks. This may arise from the poor stability of the end cap, leading to partial depolymerization during SEC analysis at 85 °C.

Scheme 3

Synthesis of Self-Immolative Diblock Copolymers from a Hydrophobic NB-PEtG-N Block and Hydrophilic Blocks with Pendent (A) Hydroxyls or (B) Amines

Figure 2

SEC traces of NB-PEtG-N, P-PEtOH-T, and NB-PEtG-PEtOH-T showing the coupling of the two blocks to form the amphiphilic block copolymer.

Block Copolymer Self-Assembly

The block copolymers were self-assembled via nanoprecipitation of a DMSO solution of polymer into rapidly stirring water. It was noted that assemblies formed from the 4-methoxytrityl end-capped block polymers were too unstable to be characterized as the hydrophilic block rapidly depolymerized in water, resulting in aggregation and precipitation of the remaining PEtG blocks. Rapid degradation of these hydrophilic blocks was confirmed by 1H NMR spectroscopy (Figures S26 and S27) and can be attributed to the π-donor ability of the methoxy group, which stabilizes the trityl cation through resonance upon cleavage. Therefore, further self-assembly and depolymerization studies were performed on NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T. On the basis of DLS, assemblies formed from NB-PEtG-PEtOH-T had a Z-average diameter of 74 ± 1 nm and a polydispersity index (PDI) of 0.16 ± 0.01, while assemblies formed from NB-PEtG-PDMAE-T had a Z-average diameter of 67 ± 1 nm and a PDI of 0.18 ± 0.01 (Figure A). TEM indicated that the nanoparticles were primarily spherical, with the solid cores observed (Figure B,C). The diameters were in reasonable agreement with the DLS results, indicating a mean of 58 ± 17 nm for NB-PEtG-PEtOH-T and 59 ± 13 nm for NB-PEtG-PDMAE-T.

Figure 3

(A) Intensity distribution of diameters measured by DLS for nanoparticles prepared from NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T; TEM images of nanoparticles formed from (B) NB-PEtG-PEtOH-T and (C) NB-PEtG-PDMAE-T.

Copolymer Degradation Induced by Acid

To examine the effect of acid on the copolymer assemblies, the assemblies were prepared as described above except that acetic acid was added after nanoprecipitation to achieve a final concentration of 100 mM (pH 3.3), while control samples were nanoprecipitated into 10 mM, pH 7.0 phosphate buffer (D2O for NMR and H2O for DLS). The samples were examined by 1H NMR spectroscopy and DLS over a period of 19 days. Treatment with acid was expected to selectively degrade the hydrophilic blocks of the assemblies. At the initial time point (∼10 min), primarily broad peaks corresponding to the intact hydrophilic blocks were observed in the NMR spectra, although a small amount of depolymerization was visible already for NB-PEtG-PDMAE-T (Figure A). No peaks corresponding to the PEtG blocks were observed as these blocks were in the hydrophobic cores of the particles resulting in long proton relaxation times.[45,46] As depolymerization occurred, sharp peaks corresponding to the hydrate of the hydrophilic monomer emerged, with the characteristic methine proton peak at 5.25 ppm, enabling quantification of the extent of depolymerization. Under acidic conditions, NB-PEtG-PDMAE-T underwent 54% depolymerization of the hydrophilic block over 6 days and complete depolymerization over 19 days (Figure A, B). The depolymerization of NB-PEtG-PDMAE-T was slower at pH 7 than at acidic pH, with only 21% depolymerization of the hydrophilic block at 6 days and 38% over 19 days (Figure B, S30). In contrast, NB-PEtG-PEtOH-T underwent depolymerization at similar rates under acidic and neutral conditions, with about 27% depolymerization of the hydrophilic block over 6 days and 48% over 19 days (Figures B, S28, S29). We have previously observed that at neutral and moderately acidic pH values, the cleavage rate of the trityl end cap is relatively independent of the pH alone and instead depends strongly on the environment of the trityl group, and thus the pendent groups of the polymer.[39] The presence of pendent amines on the polycationic PDMAE block, which are more highly protonated at pH 3.3 compared to pH 7,[47] allow increased water access to the trityl group selectively at lower pH, thereby enabling pH dependence. In contrast, the neutral PEtOH blocks do not exhibit this phenomenon, and the cleavage of the trityl end cap does not exhibit pH-dependence in this range of pH values. Even after 19 days, in all cases only trace levels of ethyl glyoxylate hydrate were observed due to background degradation of the PEtG block, thereby demonstrating the highly selective depolymerization of the hydrophilic block.

Figure 4

(A) 1H NMR spectra over time (9/1 D2O/DMSO-d6 with 50 mM acetic acid, 400 MHz) of NB-PEtG-PDMAE-T showing selective depolymerization of the PDMAE block over 19 days. Note that the PEtG block remains insoluble and is therefore not visible in the spectrum. (B) Depolymerization of the hydrophilic block versus time for NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T as quantified by 1H NMR spectroscopy.

Visually, gradual macroscopic aggregation was observed for all samples, which can be rationalized by a loss of stability of the assemblies upon depolymerization of the stabilizing hydrophilic blocks. In DLS, the mean count rate remained quite stable for NB-PEtG-PEtOH-T under both neutral and acidic conditions (Figure A), with a small reduction in count rate for the acidic sample at later time points. The Z-average diameter remained relatively constant for the first 18 days and then increased after 20 days (Figure C). The count rate in DLS is an indication of the scattered light intensity, which depends on the number of scattering species in solution and their molar mass. Therefore, the reduction in count rate concomitant with an increase in diameter can be explained by sedimentation of macroscopic aggregates in the solution, traces of which can be seen in the DLS intensity distributions (Figures E, S32). TEM images of the NB-PEtG-PEtOH-T assemblies incubated under neutral or acidic conditions for 15 days showed the presence of remaining particles, along with debris presumably due to depolymerization, such as monomer hydrate (Figures A, S34A). NB-PEtG-PDMAE-T under acidic conditions exhibited a relatively stable count rate over the first couple of weeks, followed by a modest increase (Figure B). The Z-average diameter initially decreased after the pH was reduced, which may result from protonation of a high fraction of the tertiary amines on the PDMAE block upon addition of the acetic acid and thus stabilization of smaller assemblies (Figure D,F). However, at later time points, the diameter increased. Interestingly, TEM images obtained after 7 days suggested the transformation of particles into polymersomes (Figure B), but by 15 days only solid particles were observed again, presumably because the transient polymersomes were subsequently destabilized through further depolymerization (Figure C). These particles at day 15 were probably sufficiently stabilized by the ∼20% of remaining polymerized PDMAEA block. NB-PEtG-PDMAE-T assemblies under neutral conditions, either with or without phosphate buffer, exhibited poor stability and readily aggregated in less than 10 days, as indicated by a dramatic increase in the count rate, Z-average diameter, and the appearance of large aggregates in the intensity diameter distribution (Figure S32), as well as in TEM (Figure S34B). As the PDMAE blocks clearly underwent slower depolymerization under neutral compared to acidic conditions (Figure B), the aggregation was not dominated by the more rapid loss of the hydrophilic blocks. Instead, the observed differences between neutral and acidic conditions can be likely attributed to the pH-dependent protonation states of the amino groups. For example, the dimethylamino groups of poly(N,N-dimethylaminoethyl methacrylate) have been reported to under progressive protonation between pH 9 and 5,[47] with a pKa of ∼7.1.[48] Therefore, the NB-PEtG-PDMAE-T assemblies would be less cationic at pH 7 than at pH 3.3 and therefore less stable, as surface charge is known to enhance colloidal stability of nanoparticles and polymer assemblies.[49] The acidic conditions, while leading to more rapid depolymerization, also imparted increased charge to the assemblies, therefore enabling stabilization even with a lower percentage of the hydrophilic block intact compared to at pH 7.

Figure 5

(A and B) count rate; (C and D) Z-average diameter; (E and F) intensity diameter distribution from DLS for NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T particles incubated under neutral (9/1 10 mM, pH 7 phosphate buffer/DMSO) or acidic conditions (9/1 H2O/DMSO with 50 mM acetic acid). Error bars correspond to the standard deviations on triplicate measurements (smaller than the data points in A and B).

Figure 6

TEM images of (A) NB-PEtG-PEtOH-T assemblies after 15 days under acidic conditions; (B) NB-PEtG-PDMAE-T assemblies after 7 days under acidic conditions; (C) NB-PEtG-PDMAE-T assemblies after 15 days under acidic conditions.

Copolymer Degradation Induced by UV Light

To examine the effect of UV light on the assemblies, the NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T nanoassemblies were irradiated for 30 min using a 365–370 nm LED light source, and then they were monitored by DLS and NMR spectroscopy for 3.5 h. Prior to irradiation, only peaks corresponding to the hydrophilic blocks and not the PEtG could be observed by 1H NMR spectroscopy as during the pH-mediated depolymerization studies. After irradiation, the depolymerization of the hydrophobic PEtG blocks of both copolymers occurred rapidly, plateauing over a couple of hours, as indicated by the appearance of peaks corresponding to the depolymerization product ethyl glyoxylate hydrate (Figures , S31). The hydrophilic blocks remained intact, again demonstrating the selective depolymerization of one block. In addition, SEC analysis of NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T before and after irradiation with UV light was performed. A reduction in copolymer Mn and a substantial decrease in the RI signal were observed (Figure S35).

Figure 7

1H NMR spectra over time in 9:1 D2O/DMSO-d6 before and after 30 min of irradiation with a 450 W mercury light source showing the depolymerization of the PEtG block as the appearance of the soluble depolymerization product ethyl glyoxylate hydrate. Note that the small peak at 1.2 ppm corresponds to ethanol that is generated by some hydrolysis of ethyl glyoxylate hydrate after depolymerization.

The nanoassemblies remained well dispersed/soluble throughout the depolymerization process, which was expected as degradation of the hydrophobic blocks should leave water-soluble hydrophilic blocks remaining. Consistent with the NMR spectroscopy results, DLS showed a rapid reduction in count rate during the first hour, followed by a plateau for both NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T nanoassemblies, whereas the count rate remained constant over this time period in the absence of irradiation for both systems (Figure A,B). NB-PEtG-PDMAE-T nanoassemblies underwent an 8-fold reduction in count rate upon irradiation, consistent with the disassembly of the vast majority of particles. On the other hand, the reduction in count rate was smaller for the NB-PEtG-PEtOH-T nanoassemblies, suggesting that some aggregated species remained, possibly due to a small fraction of intact PEtG blocks that the hydrophilic PEtOH blocks were able to disperse less effectively compared to the PDMAE blocks, or perhaps due to some aggregation of the resulting PEtOH blocks. Both systems underwent only modest reductions in their Z-average diameters (Figures C,D, S33), although given the large reduction in count rate for NB-PEtG-PDMAE-T in particular, these diameters after irradiation may be less reliable. Nanoparticles could no longer be observed by TEM for either polymer.

Figure 8

(A and B) count rate; (C and D) Z-average diameter from DLS for NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T particles irradiated for 30 min with a 450 W mercury light source and then incubated in 9:1 H2O/DMSO. Error bars correspond to the standard deviations on triplicate measurements (smaller than the data points in A and B).

To further probe the depolymerization and examine the potential for the system to release a hydrophobic payload upon depolymerization, Nile red was incorporated into the assemblies. After UV light irradiation, both the NB-PEtG-PEtOH-T and NB-PEtG-PDMAE-T nanoassemblies exhibited ∼66% reduction in Nile red fluorescence over 6 h, consistent with its release into a more polar environment (Figure S36).[50] Overall, the triggering of the hydrophobic PEtG block depolymerization had the opposite effects compared to the triggering of the hydrophilic block. Whereas depolymerization of the hydrophilic block leads to aggregation, depolymerization of the hydrophobic block leads to a substantial degree of particle disassembly.

Conclusions

Using the versatile chemistry of PEtG as a starting point, a small library of self-immolative poly(ethyl glyoxylate)-polyglyoxylamide diblock copolymers was synthesized. Different end caps were incorporated on the hydrophilic and hydrophobic blocks enabling selective triggering of one block. The amphiphilic block copolymers were self-assembled by nanoprecipitation to form particles with diameters less than 100 nm. Copolymers with 4-methoxytrityl end caps were too hydrolytically unstable for further investigation, but the triggered depolymerization of assemblies formed from the trityl end-capped copolymers was studied. Overall, depolymerization of the hydrophilic blocks led to aggregation of the nanoassemblies. However, a cationic charge on the hydrophilic block impacted the behavior. While acidic pH led to more rapid depolymerization of the amino-functionalized polyglyoxylamide, the assemblies were less stable at neutral pH due to reduced charge-based stabilization. On the other hand, depolymerization of the hydrophobic PEtG block, triggered by UV light, led to disassembly. Overall, this work provides a platform to control and trigger the depolymerization of diblock copolymers. It is anticipated that this ability can be used to control the morphology and aggregation state of polymer assemblies, and also potentially for patterning or the generation of micro/nanoporous structures in the solid state.