The basis of our experiments will be the RAFT synthesis of amphiphilic block copolymers under PISA conditions in an aqueous medium. This will lead to supramolecular structures emerging from a Dissipative Self-Assembly (DSA) process preceded by the out-of-equilibrium self-organization of the synthesized amphiphilic block copolymers (ABCs). Both the RAFT process and the associated PISA scenario take place under well controlled reaction conditions. We investigate the impact of oxygen and illumination on the resulting autonomously self-assembled supramolecular structures that are produced.
- Synthesis of PEG-b-PHPMA amphiphiles and resulting morphology
To study the above, a photo-induced electron transfer polymerization-induced self-assembly (PET-PISA) process was implemented to chain-extend a polyethylene glycol macro-reversible addition-fragmentation chain transfer (m-RAFT) agent with hydroxypropyl methacrylate (HPMA) as monomer and photocatalyzed by a Ru(bpy)32+ salt29,30 in an oxygen-poor aqueous medium in a temperature controlled 1.5 mL quartz cuvette that can be closed to reduce oxygen concentration, Fig. 1. After 16-hours of blue light irradiation at 25 oC in the reactor, the PISA reaction generated highly defined core-shell nano-structures (DP = 20, PDI = 1.14, average diameters = 10.5 nm, Fig. S1b-d) which have polyethylene glycols as their hydrophilic stabilizers and PHPMA as the core-forming blocks. Using TEM, Fig. S1a, their morphology was characterized as micelles.
After the first 16 hours, small aliquots of the resulting reaction solution were transferred to an optical microscope slide for their observation while the photo PISA reaction continued under illumination by the microscope light. Each of the scenarios we studied will now be presented and the pertinent results discussed.
- Morphological dynamics of an oxygen-poor PISA specimen subject to light irradiation while being observed using optical microscopy.
Prior to the direct optical microscopy observation of the time evolution of the morphological dynamics of these objects in the microscope, small aliquots of the PISA solution were stained with rhodamine 6G (4 uM in the specimen) and then subject to 15-minute nitrogen bubbling in order to prepare oxygen-poor microscopic specimens. After this, the aliquots were transferred to a blue plastic frame sealed chamber on a standard microscope glass slide and a cover slip was used to de facto seal the sample on the slide. Then the slide was mounted on the microscope, where the PISA specimens were subject to blue light (470 nm and 6.65 mW power as measured on the slide) irradiation from the microscope light source and in-field fluorescence images were taken every 5 seconds. In the absence of irradiation, no observable polymer objects were detected in the fluorescence images, even after 16 hours, due to their sizes being below the resolution limit of the optical microscope. A fluorescence image taken at 0 min is shown in Fig. 2a. As seen in Fig. 2a and supplementary movie 1, upon irradiation, a phase rich in the dye, R6G, with bright fluorescence emission gradually separated from the water phase and ultimately occupied the entire image field. Given the affinity between the dye and the PHPMA-blocks31,32, the observed phase separation can be argued to result from gelation which eventually filled the image field with the hydrophobic phase containing PHPMA blocks and HPMA.
This gelation is associated with previously reported temperature-dependent gelation as an inherent property of PEG-b-PHPMA co-polymers33. In fact, due to the presence of a large number of hydroxy groups, the core-forming block, PHPMA, in spite of its hydrophobic nature, is a highly hydrated polymer. Chain extension of PHPMA coupled with an increase in temperature produces a limited increase in hydration level in the PHPMA blocks34 which can swell the self-assembled micelles35 and/or induce micelle-to-worm evolution36,37 and eventually result in gelation. (We point out that during irradiation with blue light in the optical microscope, we observed that the environmental temperature near the specimen increased to 37 oC from the initial 25 oC at which the PISA process in the quartz cuvette was carried out.) Therefore, gelation is expected to take place at the blue light irradiated spot of the oxygen-poor specimen due to efficient chain-extension and the subsequent increase in hydration of the polymer cores.
- Morphological dynamics of oxygen-rich PISA specimens subject to light irradiation under the microscope.
If instead of following the procedure described in the previous sub-section, we remove the low-oxygen restriction and we air-bubble the PISA solution prior to blue light irradiation at the microscope, once under the microscope, we observe a morphological dynamics which is distinctly different from the one reported in the previous section. The illuminated observed spot in the optical microscope slide became gradually populated by giant polymer objects of various sizes with the simultaneous presence of hollow structures which were characterized as polymer vesicles, Fig. 2b-e, supplementary movie 2 and Fig. S2. Detailed observation showed that these vesicular structures had either emerged within the field of view of the microscope lens or migrated from outside of the imaged area and gathered towards a spot with the highest light intensity, Fig. S7d. This observed behavior is of the same type that was reported by Albertsen et al..28 Beside the emerging vesicles, in a few of the observed cases, the large nascent micelle aggregates appeared first and were followed by consecutive stages of morphological evolution. First a morphological transition from a micelle aggregate to a vesicle, accompanied by a slight outward budding and internal multi-compartmentalization during the transition. After the onset of vesicle formation, the supramolecular polymer structures started exhibiting what would eventually become cyclic episodes of size-growth accompanied by thinning membranes which, at some maximum sustainable surface area, imploded and became smaller vesicles with proportionately thicker membranes, Fig. 2c. The collapsed vesicles repeated the same process of growth and collapse for a number of times, which in some cases reached about 25 cycles. Interestingly, during such cyclic growth-collapse dynamics, the giant vesicles clearly increased in number (Fig. S3) and gradually filled the entire imaged area (Fig. 2b) of the PISA specimen. (As in reference Albertsen et al. we will refer to this morphological dynamical evolution as “Phoenix” dynamics.)28
The Phoenix dynamics of course must result from a mechanism different from the previously described gelation observed in oxygen-poor specimens. Indeed it is known38,39 that in the presence of oxygen, reactive oxygen species (ROS) are generated in PISA systems by the photosensitive species, which in our PISA system includes the photocatalyst, Ru(bpy)32+, and the staining dye, rhodamine 6G.40 Therefore, radical polymerization and its contribution to hydrophobic block elongation can be expected to play a limited role in the observed Phoenix dynamics due to its deactivation by the generated ROS.
We know that due to the ongoing polymerization and its many structural and energetic consequences, an increase in temperature (for example due to the reaction or external illumination) is another factor which can potentially promote morphological transitions.41 Therefore, in order to understand if thermal effects play a role in our observed Phoenix dynamics, an oxygen-rich PISA sample was incubated at 40 oC in darkness. Under these conditions no Phoenix dynamics were observed, although a few micron-sized objects without Phoenix behavior were observed in the fluorescence images, Fig. S4.
In sharp contrast to the above, oxygen-rich samples incubated at 25°C and exposed to the microscope’s blue light irradiation (wavelength of 470 nm and 6.65 mW power measured on the slide) exhibited Phoenix dynamics similar to the samples without external temperature control. Therefore, given that the contributions from polymerization and temperature are relatively minor, we infer that the observed experimental behavior indicates that the presence of oxygen plays an important role for Phoenix morphological dynamics to occur when our PISA specimens undergo irradiation in the microscope with the above mentioned blue light. We are then led to interpret the Phoenix dynamics as primarily being the result of water influx into the polymer core of the vesicles driven by osmolarity mismatch between the core of the assembled system and the surrounding solution containing unreacted PISA material. This mismatch originates in an increase of water-soluble species in the polymer core as a consequence of photo-induced oxidative reactions within the self-assembled supramolecular polymer objects. Moreover, under irradiation with blue light in oxygen-rich conditions, the chemical degradation not only produces water soluble species through photosensitization of the Ru(bpy)32+ photocatalyst and the R6G dye within the polymeric cores of the vesicles, but also leads to the oxidation of the core forming blocks which results in their increased hydrophilicity.40 This enhances the osmotic water influx into the vesicles and their subsequent Phoenix dynamics.
- Photo-induced chemical degradation tests.
Our microscope specimens can be considered as closed systems with respect to the transfer of matter. Thus, the oxidative products mentioned above must have originated from the chemicals already present in the PISA solution aliquot that was deposited on the microscope slide. Of course, during controlled radical polymerization, the macro-RAFT agents are the key substances controlling polymer chain extension which, as the reaction proceeds and the degree of polymerization (DP) changes, modifies the packing parameter21 and leads to a potential sequence of polymer morphologies. It has been reported that many RAFT agents undergo degradation in organic solvents by UV or blue light irradiation.42–45 For example, in some cases, a nanoscale morphological transition from worms to vesicles can be generated by prolonged exposure to UV irradiation.46
To examine if our macro-RAFT agent underwent a similar degradation process, we first prepared an oxygen-poor aqueous macro-RAFT solution which was then subject to blue light irradiation and for which we monitored its characteristic absorption peak via UV-VIS spectroscopy. As seen in Fig. S5, an aqueous solution of our macro-RAFT agent exhibits a characteristic absorption with a maximum at 505 nm which corresponds to an n to π* transition.45 After five-hours of blue light irradiation, the absorbance peak intensity went down by approximately 6%, which indicates a good stability for our macro-RAFT agent in an oxygen poor environment.45
However, a progressive decrease in the absorbance signature appeared when such photo-induced degradation experiments were performed in oxygen-rich m-RAFT solutions. Up to 23% of the m-RAFT agent undergoing degradation points to the presence of some oxidative reaction taking place due to the presence of oxygen, Fig. 3a. In addition to m-RAFT agent degradation, Ru(bpy)32+ and rhodamine 6G (present in our PISA systems) are two photocatalysts well-known to be sensitive to photobleaching.47,48 As expected, in Fig. 3b and c, we show that both photocatalysts are indeed vulnerable to degradation when irradiated with blue light in an oxygen-rich environment, and show reductions in their absorbance peak intensities of 24% and 22% respectively. Furthermore, the hydrophilic oxidative products formed by the degradation of m-RAFT, Ru(bpy)32+ and rhodamine 6G are capable of rapidly dissolving into the water phase and thereby increase the osmotic solute concentrations.44,49,50
In order to induce osmotic water influx from the surrounding bath into the polymer cores of the vesicles, some chemicals need to degrade within the polymer cores to increase its osmotic solute concentrations. In our control experiments we found that the addition of pre-degraded macro-RAFT agents into an oxygen-poor PISA specimen prior to irradiation with microscope light, resulted in gelation instead of Phoenix dynamics which support the above degradation hypothesis. This emphasizes the importance of osmolarity mismatch induced by ongoing in-core chemical degradation.
- Effects of different monomers on Phoenix dynamics.
Next we investigated the effects that different monomers can have on Phoenix dynamics. Given a micelle with a highly hydrated core, the oxidative products produced by an on-going in-core chemical degradation tend to dissolve in the nearby internal water-rich domains rather than in the surrounding bath. But the self-assembled micelles in our system consist of amphiphilic diblock copolymers with PHPMA as their hydrophobic blocks. The cores of the micelles contain both HPMA monomer and PHPMA blocks which are in a highly hydrated state due to the presence of a large number of hydroxyl groups capable of capturing water molecules which then induce the formation of many tiny hydrophilic domains within the hydrophobic phase and eventually dissolve the oxidative species. Therefore, it is natural to conclude that the osmolarity mismatch induced by in-core chemical degradation drives a water inflow from the surrounding bath into the polymer core to ultimately induce the hydrophilic domains to coalesce into a single internal lumen, and lead on to Phoenix dynamics.51
The relevance of core hydration to Phoenix dynamics becomes more explicit in our PISA experiments (cf. below) in which three other polymers with cores of different hydrophobicities were prepared using different monomers. We used hydroxybutyl acrylate (HBA), butyl acrylate (BA) and styrene as monomers. Of these three monomers, only self-assembled polymer superstructures synthesized using HBA, Fig. 4b, exhibit Phoenix dynamics close to the one with HPMA, Fig. 4a. (Note that the glass transition temperature, Tg, for fully dried PHPMA200 is 95oC but Tg for fully hydrated PHPMA200 is 47oC.52 Higher hydration increases the plasticity of PHPMA.53 The Tg for the other three monomers is reported to be PBA (-24oC), PS (100oC), and PHBA (-40oC). HBA is a water-soluble monomer with one hydroxyl group just like HPMA. Their similar molecular structure allows their polymer forms to be hydrated to the same extent. With photo-induced degradation occurring in an oxygen rich environment the polymer cores consisting of PHBA and HBA monomers contain water-rich domains which, as already discussed, can accommodate oxidative products and then experience osmotic water influx. As a result, the resulting self-assembled nanoscale polymer structures evolved and grew in size to form collective micron-scale structures with a mixture of morphologies that included vesicles and vesicle-like objects with outward budding or incipient multi-compartmentalization. We note that the PHBA cores tend to form larger objects than the cores formed with PHPMA, which can be attributed to the higher flexibility of PHBA compared to that of HPMA. The lower Tg of the PHBA molecule allows higher flexibility of the PHBA chains in the cores which can then adapt to a larger lumen expansion by minimizing energy and, eventually, leading to continuous size-growth during hours of irradiation with the light from the microscope.
However, a non-polar monomer such as styrene forms micron-scale emulsions in which the cores have the lowest degree of hydration compared to the other monomers that we studied. Despite the presence of a few micron-scale emulsion droplets, as expected, most self-assembled objects showed only slight swelling and a negligible number of Phoenix dynamics with no observable formation of stable of vesicles. On the other hand, BA has a polarity between that of HBA and styrene. It is interesting that with PBA only a few objects exhibited Phoenix dynamics together with the formation of some vesicular structures. However, after hours of irradiation with the microscope’s blue light, no large-scale presence of Phoenix dynamics like the ones observed for PHPMA or PHBA cores was observed, although we saw the precipitation of tiny bright objects. This indicates that the majority of the polymer objects tend to grow at nanoscale.
From the results obtained with the use of these selected monomers, we conclude that the hydrated polymer cores are critically important for the presence of Phoenix dynamics. As the chemicals within a polymer-core degrade, the oxidation products disperse into the nearest water phases such as those associated with the hydrated water regions in the core. A water rich polymer-core consisting of the hydroxyl groups in the monomers and their polymer forms is at a lower energy state, and therefore more stable, than the water-poor polymer cores in capturing and dissolving the oxidation products which is what happens in the larger scale of Phoenix dynamic.
- Application of photoinduced chemical degradation to PET-PISA in reactors.
As already discussed, optical microscopy observations show that Phoenix dynamics takes place and generates giant vesicles when nanoscale polymer objects containing degradable chemicals and water rich cores are exposed to blue light in the presence of oxygen. We then asked ourselves if by applying a similar degradation protocol it would be possible to obtain giant vesicles in a reactor running a conventional (oxygen poor, or run with nitrogen bubbling to remove oxygen) PET-PISA reaction which normally does not generate giant vesicles.54 To understand whether this would be possible or not, we first conducted a conventional PET-PISA reaction using HPMA as monomer in an oxygen-poor reactor.
As seen in Fig. 5a, most polymer objects reach only nanometer scale sizes and therefore escape detection in the optical fluorescence microscopy imaging even after ten-hours running the PET-PISA process. However, as oxygen (through air bubbling) and rhodamine 6G were introduced into a reactor with an already ongoing conventional PET-PISA reaction for four hours, we observed that giant objects with hollow or internal multi-compartmental structures gradually forming after hours of exposure to a more intense blue light irradiation, Fig. 5 b and c. In contrast, such in-situ chemical degradation procedure did not generate similar hollow structures when styrene was used to form the polymer cores. As the results from optical microscopy observations show, the majority of the polymer objects formed with PS cores tend to remain in the form of emulsions and very few or no Phoenix events were observed in experiments performed under these conditions.