Conformational Reorganization Of Hyperbranched And Linear Polymers And Functionalized Porous Polymer Films

Abstract

The main focus of the research work presented in the thesis is the understanding of structural and conformational reorganizations in hyperbranched and linear polymers. The thesis includes three different investigations: a) the design, synthesis, conformational reorganizations and self-assembly of hyperbranched polymers (HBPs), b) the Raman spectroscopic studies of the melting of polyethylene glycol (PEG), and c) the preparation of functionalized porous polymer films. HBPs are structurally imperfect analogues of the defect-free branched polymers called dendrimers. Dendrimers prepared using a stepwise polymerization methodology will carry all the unreacted B groups at the periphery and therefore modification of these peripheral units with suitable linear segments (e.g. PEG) would naturally generate a core-shell structure. On the other hand, in the case of HBPs, that contain a large number of linear defects, the B groups are distributed throughout the highly branched backbone, and hence modification of these units may not generate a core-shell configuration. However, our earlier studies have demonstrated the unimolecular micelle type behavior (core-shell) of PEG-ylated aromatic HB polyethers in solution. Hence, even though the PEG segments are randomly placed throughout the HBP backbone, the polymer achieves a core-shell configuration through conformational reorganizations; this suggests that the core of HBPs may be flexible and adaptable. In order to further examine this issue, we have investigated the self-assembly of these PEG-ylated HBPs. These polymers were found to organize into uniform nano-aggregates (Figure 1). More interestingly, these aggregates were found to restructure on mica surface when exposed to solvent vapor. This study confirmed the formation of core-shell domains in the PEG-ylated HBPs and the ability of the HB core to reconfigure. Figure 1. Restructuring of nano-aggregates formed from PEG-ylated HBPs on mica surface. As mentioned earlier, the precise control of the placement of functional groups would be impossible in the case of hyperbranched polymers (HBPs); most approaches would result in random placement of the functional entities as the growth of the polymer occurs in a statistically governed manner. Hence, creating Janus hyperbranched structures appears a daunting task. Janus is the name given for facially amphiphilic systems, such as a sphere made by joining a hemisphere with a hydrophobic surface and one with a hydrophilic surface. Since we have demonstrated that the HBP backbone is adaptable, it would be interesting to test the possibility of generating a Janus configuration through reorganization of peripheral segments. Peripherally “clickable” HB polyesters were synthesized using the transesterification polymerization methodology. The peripheral groups were modified with two dissimilar segments, PEG and docosyl (C22 alkyl, DOCO) chains, by azide-yne click reaction using the corresponding azides. The immiscibility of the chains along with the tendency of alkyl chain to crystallize was expected to cause self-segregation of segments at the periphery aided by the reconfiguration of HB core. Melting transition associated with alkyl segments was evident in the DSC thermogram. In addition, the melting enthalpies, normalized with respect to the weight of alkyl segments, were close to that of the C22 alkyl chain for three HBPs with different compositions of DOCO and PEG. This suggests complete phase segregation of segments at the periphery. The behavior of these HBPs at the air-water interface was also investigated. In this experiment a single layer of molecules, deposited at the air-water interface, is compressed and decompressed at a predetermined rate. The surface pressure variation during the compression-decompression cycle reveals the molecular organization at the surface. Stable and reversible isotherms were obtained for these HBPs; which suggested their Janus nature with PEG chains solvated in water and alkyl chains projecting out into air. A change in slope was observed in the isotherm and we have ascribed this to the crystallization of docosyl segments. In order to verify the assignment, quasi-static compression (QSC) experiments were performed. In QSC experiment, the monolayer was rapidly compressed to a required surface area and the drop in surface pressure, as a function of time at this fixed surface area, was recorded. During the rapid compression the molecules may not have adequate time to organize optimally and therefore, holding the area constant often leads to a drop in surface pressure with time. The total pressure-drop after the compression reflects the extent of molecular reorganizations; this, for example, could be due to the crystallization of alkyl segments. In the present study, the extent of pressure drop scaled with the composition of alkyl chains; this proves the assignment and supports the self-segregation of the peripheral segments to generate a Janus configuration (Figure 2). AFM images of the transferred monolayer allowed the estimation of monolayer height, which was roughly in agreement with the expected single molecule dimension. Small angle X-ray scattering experiments (SAXS) on HBPs demonstrated existence of a lamellar morphology in the solid state. Janus structures may be expected to exist as bilayers in the solid state. As expected, the lamellar dimensions estimated were twice the monolayer height observed in AFM images. This confirms the bilayer structure and the Janus configuration of HBPs. We have also investigated the self-assembly of these Janus HBPs in solution and it was found that disk-like aggregates were formed in methanol, whereas vesicles were formed in water by systems that had the appropriate balance of hydrophilic and hydrophobic segments. Figure 2. a) The structure of Janus HBP. b) A cartoon depiction of reconfiguration of randomly functionalized HBP to Janus. c) DSC thermograms showing clear melting and crystallization peaks. d) Langmuir isotherms for Janus HBPs. The change in slope is indicated by orange arrow. e) Quasi-static compression isotherms. The extent of pressure drop for different HBPs (inset). f) The SAXS pattern suggesting lamellar structure in the solid state and a proposed model (inset). It was clear from the study that the incompatibility between segments is the primary reason for the reconfiguration. In order to test the generality of the approach, we have prepared two other types of HBPs: one with PEG and fluorocarbon (FC) and the other with DOCO and FC. These segments were chosen because of their mutual incompatibility. In both these systems, the formation of Janus configuration was similarly confirmed by DSC, SAXS and Langmuir isotherm studies. We have extended these studies to HBPs functionalized with three different immiscible segments (tripod HBPs) namely, PEG, DOCO and FC; in the case of PEG two different molecular weights were used: PEG350 and PEG1000; the latter exhibits a strong tendency to crystallize. Melting transition associated with each of the segments was observed in the DSC experiments. The normalized melting enthalpy of the docosyl domain was found to vary substantially depending on the composition and configuration of HBPs (Janus or tripod). Considering the variation of these values it was concluded that the presence of an immiscible amorphous segment, like PEG350, enhances the phase separation and crystallization of DOCO by providing more flexibility to the core. The presence of one or more immiscible but crystallizable segments (e.g. FC or PEG1000), on the other hand, lead to less effective crystallization. Interfacial behavior of the tripod HBPs, that carry the hydrophilic PEG units, were also studied using Langmuir trough. The tripod HBPs formed clear monolayer at the air-water interface at higher surface pressures (40 mN/m); where the hydrophobic segments, FC and DOCO, oriented away from water into air and PEG segments remained solvated in water. Due to the incompatibility between FC and DOCO, it may be anticipated that these segments also form phase segregated domains. While the AFM images allowed the estimation of monolayer thickness, these domains could not be resolved in the images. Interestingly, the transferred monolayer appeared smoother after thermal annealing. To investigate the reason, contact angle (CA) measurements were performed. It is widely accepted that the contact angle is very sensitive to the structure and chemistry of a few angstrom thick region of the top surface of a monolayer. The water CA on the monolayer before thermal annealing was 84˚ while it increased to 104˚ after annealing. This behavior was attributed to the effective phase segregation and crystallization of FC and DOCO domains. The crystallization causes CF3 end groups to orient normal to the surface and hence increases the water contact angle. The receding contact angle, on the other hand, was found to be substantially smaller (45˚); this large contact angle hysteresis could be indicative of chemical heterogeneity on the surface, and it has been suggested that the presence of small domains of low surface free energy could indeed result in such large contact angle hysteresis. Interestingly, when the film was aged under ambient conditions for 12 h, the contact angle dropped to the initial value of 84˚; re-annealing the film at 75˚C for 12 h again raised the contact angle back to 104˚. This reversible annealing-aging behavior reveals dynamic nature of the monolayer that permits gradual restructuring at the surface (Figure 3). Self-assembly of these tripod hybramers resulted in the formation of unique aggregated structures. Morphology of these aggregates, observed in AFM images, were different from those formed by Janus hybramers. These results provide a clear evidence for the segregation of all the three segments into individual domains -the hypothesized tripod hybramer configuration. Figure 3. AFM images of transferred monolayer of tripod HBP before and after thermal annealing. Contact angle (CA) measured on the corresponding monolayers are shown in the inset. A model proposed for explaining the variation in the appearance of the monolayer (image) and the CA is also provided in the figure. Effective phase segregation of FC and DOCO is proposed the reason for the change. Another interesting question we have investigated as a part of the thesis is the molecular mechanism of melting of PEG. Crystallization of polymer is hypothesized as polymer chains moving down the free energy landscape though sequence of conformational reorganizations into an ordered crystalline (lamellar) state. Such conformational reorganizations also occur during melting. We have attempted to probe conformational dynamics in PEG during its melting using Raman spectroscopy, polarized optical microscopy, DSC and DFT calculations. Single crystal X-ray and vibrational spectroscopic studies have demonstrated helical configuration of PEG chains in the solid state with a gauche O-C-C-O, trans C-C-O-C and trans C-O-C-C conformations. There are three ways in which the PEG-helix can change during heating; one where C-O single bond rotates, second where the C-C rotates and third where both rotate, but the barriers to these rotations are different, as was established by earlier NMR studies. The intensity ratio of two Raman bands, I2880/2850, was found to be sensitive to the configuration about O-C-C-O units along the polymer backbone. An invariant I2880/2850 ratio of ~0.78 was ascribed to the consistent C-C dihedral angle (gauche configuration) during melting. This agrees with the well-known “gauche effect” in ethylene glycol. Prominent spectral changes were also observed in the methylene rocking region (~800 cm-1). In order to gain insight into the new conformers formed during melting, we carried out Gaussian calculations using model conformers. Among helical models considered the heptamer and tetradecamer helix models were found to be most suitable for the solid state helical structure of PEG. The Gaussian calculations using different model conformers revealed that the Raman band at 810 cm-1 corresponds to CH2 rocking vibrations of the gauche C-C-O-C units along the polymer chain; while the band at ~1500 cm-1 region was characteristic of O-C-C-O trans conformer. This agrees with the earlier assignment of 1500 cm-1 band to CH2 scissoring vibrations of PEG in an all-trans configuration. Interestingly, no peak was observed in the 1500 cm-1 region of the Raman spectrum during melting of PEG, but new peaks appeared in the 810 cm-1 region (Figure 4). Hence, it was concluded that C-C bond rotation does not occur during melting of PEG. Thus, our study confirmed that C-O single bond rotation is the molecular mechanism of PEG melting. Figure 4. a) Raman spectra of PEG at different temperatures. b) helical model considered for the study. c) Comparison of the Raman spectrum of molten PEG (2) with calculated spectra of two different models. The dynamics of a system, where the successive changes occur as a function of external perturbation, can be studied using 2D correlation spectroscopy (2D-CoS). 2D-Raman Correlation spectroscopy was employed to study the molecular structural dynamics during the melting of PEG. Conventional rules of 2D-CoS were used to retrieve the order in which the vibrational bands respond to temperature. Vibration at 934 cm-1 corresponding to the amorphous domain of PEG solid state was found to respond to temperature first. As the temperature of the system rises, CH2 rocking (1280 cm-1), CH2 wagging (1472 cm-1) and CH2 scissoring (1124 cm-1) vibrations becomes active and this provides flexibility to the chain. As the polymer chain gains adequate energy (at particular temperature) bond rotation takes place resulting in the transformation of a few TGT segments to GGG segments. On increasing the temperature further more C-O bond rotations occur, leading to destruction of lamellar domains and eventually PEG melts. The final study presented in this thesis deals with the generation of functionalized porous polymer films. Condensation of water droplets during solvent evaporation from a polymer solution, under humid conditions, is known to generate uniformly porous polymer films. We have investigated the possibility of pore formation through water phase separation strategy (Figure 5). In the presence of added surfactants (SDS and CTAB in the present study), the interface of phase separated water droplets and the polymer would naturally become lined with the surfactants and consequently the internal walls of the pores generated, upon removal of the water, could become decorated with the hydrophilic head groups of the surfactant molecules. The size of the pores and their distribution were examined using AFM and IR imaging methods. We have demonstrated that the ATR-IR imaging is an efficient method for analyzing a few nanometer thick surface section of the polymer film (Figure 5e). It was observed that the presence of surfactant is important for the pore formation and irregular pore formation resulted in the absence of surfactant. In addition, both the surfactant concentration and the relative volume fraction of the surfactant solution were found to govern the size of the pores formed. Cloud point measurements suggested that the occurrence of surfactant facilitated the phase separation of water. Although IR imaging possessed inadequate resolution to confirm the presence of surfactants at the inner surface of pores, exchange of the inorganic counter-ion, such as the sodium-ion of sodium dodecyl sulphate (SDS), with suitable ionic organic dyes permitted the unequivocal demonstration of the presence of the surfactants at the interface with confocal fluorescence microscopy (Figure 5f). Figure 5. Schematic depiction of a homogeneous polymer solution (sky blue) in THF–water; yellow lines with red dots depicts dissolved surfactant molecules. (b) Initial stages of water droplet formation that are stabilized by surfactants. At this stage some polymer precipitation may also occur at the interface (depicted by the dark ring around the droplet). (c) Some droplet coalescence and continued formation of new water droplets leads to a slightly broad size distribution (coalescence and redistribution could be retarded due to the high solution viscosity). (d) Complete removal of THF followed by water generates the internally functionalized porous structures. e) IR spectral Image of the porous film. f) confocal fluorescence image of the porous film where the inner surface was functionalized with Rhodamine B. (For figures pl refer the pdf file)