Haifeng Gao's Research: Polymer Chemistry
|
Link
|
Research Interests and Experience:
Recent Achievements on Research
1. Synthesis of Star and Miktoarm Star Copolymers via Atom Transfer Radical Polymerization (ATRP)
As a branched nanoscale polymeric material, star polymers have many interesting properties, arising from their compact structure and globular shape. By using various living polymerization techniques, such as anionic polymerization and controlled/living radical polymerizations (CRPs), star polymers are synthesized by one of three strategies: "core-first", "coupling-onto" and "arm-first".
The "core-first" method involves the use of a multifunctional initiator (core). The number of arms per star polymer is determined by the number of initiating functionalities on each initiator. The initiating sites preserved on the star polymers can be further used for chain extension with a second monomer to form star block copolymers (Scheme A).
In the "coupling-onto" method, a linear polymeric chain (arm) containing a reactive chain-end group reacts with a multifunctional coupling agent (core) (Scheme B). Theoretically, the functionalities on the coupling agents predetermine the arm numbers per star, although this method suffers from low coupling efficiency due to the slow reaction between the polymer chain end and the multifunctional core. In order to improve the coupling efficiency, excess amount of linear chains are used and the final star product is purified by repeated fractionation to remove the unreacted linear chains. Recently, the Cu(I)-catalyzed 2 + 3 cycloaddition reactions between an azide and an alkyne, i.e. "click reactions", have gained a great deal of attention due to their high specificity and nearly quantitative yields in the presence of many functional groups. By using "click reactions", several kinds of star and miktoarm star polymers have been synthesized with predetermined structure and high star yield.
In the "arm-first" method, the linear arms of the star polymers are synthesized first followed by binding of the arms to form the core, usually achieved by using a divinyl cross-linker (Scheme C). The resulting star polymers generally have a statistical distribution of the number of arms and the preformed arms can be either linear macroinitiator (MI) or macromonomer (MM). It is worth noting that when linear MM was used to copolymerize with divinyl cross-linker by using small molar mass initiator, the number of initiating sites (derived from the initiator) and number of arms (derived from MM) per star molecule were independently controlled. A low molar ratio of initiator to MM decreases the number of initiating sites in the star core, which effectively limits the extent of star-star coupling reactions and results in star polymers with low polydispersity. The preserved initiating sites in the star cores can be further employed to initiate the polymerization of another monomer to form miktoarm star copolymers by the "in-out" method.
(1). Synthesis of star polymers by "coupling-onto" method
We reported the first synthesis of star polymers by a coupling procedure using a combination of ATRP and click chemistry. Polystyrene (PS) linear chains with high azido chain-end functionality coupled with multifunctional alkyne-containing coupling agents under a mild condition to produce linear and star polymers with different arm numbers. The click reaction between an azido-terminated PS (PS-N3) and an alkyne-containing multifunctional compound proved to be fast and efficient. All coupling reactions were finished within 3 hours, proven by the disappearance of signals from the azido groups in NMR spectra and the high yields of the coupled products by GPC analysis. For the model coupling reaction between a PS-N3 polymer and a dialkyne-containing compound, the final yield of the coupled PS-PS polymer was ca. 95 %. When a PS-N3 polymer was reacted with a trialkyne-containing or tetraalkyne-containing compound, the yields of 3-arm star and 4-arm star polymers were around 90 % and 83 %, respectively. The influence of several parameters on the efficiency of the click coupling reaction was studied, including the molecular weight of the PS-N3 polymer, the presence of an added reducing agent, Cu(0), and the stoichiometry between the azido and alkynyl groups. The results indicated that the yield of the coupled product was higher when a lower molecular weight PS-N3 was employed in conjunction with a small amount of reducing agent and the molar ratio of azido and alkynyl groups was close to 1.
Gao, H.; Matyjaszewski, K. Macromolecules 2006, 39, 4960
(2). Synthesis of star polymers with controlled structures by "arm-first" method
By using different kinds of functoinal initiators, monomers and cross-linkers, star polymers with controlled structures and various functionalities at either star core or star surface, are synthesized by the economic "arm-first" method. Several experimental parameters have significant influence on the star structures, including the arm length, the amount of cross-linker, the monomer conversion at which cross-linker was added, and the chemical composition of the star core (fraction of monomer units in the star core). Employing shorter arm lengths and using more cross-linker produced star polymers in higher yield, with higher molecular weight, with more arms per star, and with a more compact structure. The addition of cross-linker at lower monomer conversion caused the incorporation of more monovinyl monomer into the star core, which decreased the cross-link density of the core, facilitated incorporation of more arms into each star molecule and increased the star size and star yield.
A. "Arm-first" method by cross-linking linear macroinitiator (MI)
Gao, H.; Matyjaszewski, K. Macromolecules 2006, 39, 3154
B. "Arm-first" method by cross-linking linear macromonomer (MM)
Recently, we developed a new method for synthesis of star polymers with high molecular weight, high star yield and low polydispersity via copolymerization of linear macromonomers (MM) with a divinyl cross-linker using low molar mass ATRP initiators (MM method). In this method, the number of initiating sites (derived from the initiator) and number of arms (derived from MM) per star molecule are independently controlled. A low molar ratio of initiator to MM decreases the number of initiating sites in the star core and reduces the possibility of star-star coupling reactions and decreases star polydispersity. Moreover, MMs, initiators and cross-linkers can be added in several steps, increasing the flexibility of star synthesis. An extra advantage of this new MM method for star synthesis is that additional functional groups can be easily introduced into the star core via functional initiators. The synthesis and application of functional ATRP initiators is much easier and more efficient than the displacement of the embedded alkyl halide chain ends in the preformed star core.
Gao, H.; Ohno, S.; Matyjaszewski, K. J. Am. Chem. Soc. 2006, 128, 15111
Gao, H.; Matyjaszewski, K. Macromolecules 2007, 40, 399
(3). Preparation of degradable miktoarm star copolymers via ATRP by "in-out" method
Miktoarm (or heteroarm) star copolymers are defined as star molecules containing two or more than two types of arms with different chemical compositons and/or molecular weights connecting to one central core. Miktoarm star copolymers have attracted considerable attention due to their branched architectures, globular shapes and segmented block structures. The different chemical compositions of the arms in miktoarm star molecules lead to their interesting microphase separations in bulk, in solution and at different interfaces. The segregated compartments in the aggregates, such as micelles, could in turn provide distinct chemical environments to store various kinds of drug molecules, fragrance compounds and gene therapy agents.
The "in-out" method is an important strategy for synthesis of miktoarm star copolymers containing two kinds of arms with different chemical compositions and multiple arm numbers. A multifunctional star polymer, synthesized by "arm-first" method, is used as a macroinitiator (MI) to initiate the polymerization of another monomer and form a miktoarm star copolymer, (polyA)n-polyX-(polyB)p, where polyX represents the cross-linked core of the miktoarm star polymer; n and p are the average number of polyA and polyB arms per star molecule, respectively. Based on the synthetic procedure used for the synthesis, this method is termed the "in-out" method, in which the word "in" refers to the arm-first method for formation of the star polymer MI and the word "out" represents the subsequent growth of the second generation of arms from the multifunctional star core. Due to the congested environment around the cross-linked core in the star MI, not all of the initiating sites participated in the formation of the second generation of arms and the initiation efficiency (IE) of the star MI was less than 100%.
Two methods have been reported to determine the IE of the star MI during the synthesis of miktoarm star copolymers. One is synthesis of miktoarm star polymers containing a cleavable core. The IE value of the star MI was determined via GPC analysis of the cleaved product of the miktoarm star polymers, which was a mixture of linear triblock and diblock polymer chains.
Gao, H.; Tsarevsky, N. V.; Matyjaszewski, K. Macromolecules 2005, 38, 5995
The second method for quantitative determination of the IE value of star MI is a kinetic method, which compared the relative rates of monomer consumption by using star MI to a control ATRP reaction under the same experimental conditions as those during the miktoarm star synthesis except for the use of small alkyl halide molecule as initiator.The mole of added small alkyl halide in the control reaction was the same as that of the bromine initiating sites in the miktoarm star synthesis. Therefore, the IE of star MI equaled to the ratio of the apparent kinetic constants in the two reactions.
Gao, H.; Matyjaszewski,
K. Macromolecules 2006, 39, 7216
(4). Synthesis of miktoarm star copolymers with controlled structures by using a simple and general "arm-first" method
As discussed above, the "in-out" method is an important methodology for synthesis of miktoarm star copolymers with varioius arm numbers and arm compositions. However, the miktoarm star copolymers synthesized by this mehtod also ave an intrinsic problem, the low IE of star MI, which results in an unequal number of two kinds of arms in the miktoarm star copolymers. Furthermore, it is strategically impossible to synthesize a miktoarm star copolymer containing more than two kinds of arms by using the "in-out" method. Both problems motivated us to develop new and easier methodology for synthesis of miktoarm star copolymers with various molar ratios and species of the arms connecting to one star core.
Herein, we for the first time reported the synthesis of miktoarm star copolymers containing two or more arm species using simple and general "arm-first" method, i.e., one-pot cross-linking a mixture of different linear macroinitiator (MI) species by divinyl cross-linker, such as divinylbenzene. Using linear MIs with high degree of bromine chain-end functionality, including polyacrylate, polystyrene, polymethacrylate and poly(ethylene oxide), resulted in high-yield star polymers (> 90% analyzed by GPC). Characterized by liquid adsorption chromatography techniques, which separated star polymers based on the chemical composition of arms, the obtained star product was proved to be miktoarm star copolymers containing two or more arm species in one molecule, instead of mixture of different homoarm star polymers. Within our investigation, the molar ratios of the arms in the miktoarm star copolymers were always in agreement with the composition of the initial MI mixture, indicating the powerful capacity of this "arm-first" method for synthesis of miktoarm star copolymers with potentially any molar ratios and species of the arms. By using a mixture containing five types of linear MIs with different chemical compositions, miktoarm star copolymers containing five kinds of arms were synthesized for the first time, which significantly expanded the methodologies for synthesis of miktoarm star copolymers by living polymerization techniques.
Gao, H.; Matyjaszewski, K. J. Am. Chem. Soc. 2007, 129, 11828
2. Synthesis of Polymer Brushes by "Grafting-Onto" Method
Due to their high specificity and nearly quantitative yields in the presence of many functional groups, the Cu(I)-catalyzed 2 + 3 cycloaddition reactions between an azide and an alkyne, i.e. "click reactions", were applied here for synthesis of polymer brushes via "grafting-onto" method. By using click reactions, molecular brushes can be synthesized by two strategies. One is to use alkyne-containing polymeric backbone to react with azido-terminated polymeric SCs. The opposite approach is to react azido-functionalized backbone with alkyne-terminated SCs. Here, the first strategy was applied to prove the feasibility of click reactions for synthesis of molecular brushes. The functional alkynyl side groups were introduced into the polymeric backbones by the esterification reactions between pentynoic acid and the hydroxyl side groups on poly(2-hydroxyethyl methacrylate) (PHEMA). Then, the azido-terminated linear SCs with various chemical compositions and molecular weights were used to click coupling with the functional PHEMA-alkyne backbones and to form grafted copolymers, PHEMA-g-PB.
Five kinds of azido-terminated polymeric side chains (SCs) with different chemical compositions and molecular weights were used, including poly(ethylene glycol)-N3 (PEO-N3), polystyrene-N3, poly(n-butyl acrylate)-N3 and poly(n-butyl acrylate)-b-polystyrene-N3. All click coupling reactions between alkyne-containing polymeric backbones (PHEMA-alkyne) and azido-terminated polymeric SCs were completed within 3 h. The grafting density of the obtained molecular brushes was affected by several factors, including the molecular weights and the chemical structures of the linear SCs, as well as the initial molar ratio of linear chains to alkynyl groups. When linear polymers with "thinner" structure and lower molecular weight, e.g., PEO-N3 with Mn = 775 g/mol, were reacted with PHEMA-alkyne (degree of polymerization = 210) at a high molar ratio of linear chains to alkynyl groups in the backbone, the brush copolymers with the highest grafting density were obtained (Ygrafting = 88%). This result indicates that the average number of SCs was ca. 186 per brush molecule and the average molecular weight of the brush molecules was ca. 190 kg/mol.
Gao, H.; Matyjaszewski, K. J. Am. Chem. Soc. 2007, 129, 6633
3. Synthesis of Polymer Network by Atom Transfer Radical Copolymerization with Cross-linker
Radical copolymerization of monovinyl monomers with a small amount of divinyl cross-linkers has been widely used for synthesis of branched polymers and cross-linked gels. The branching points are generated in the polymer chains via reaction of pendant vinyl groups with the propagating radicals either inter-molecularly or intra-molecularly. With the progress of inter-molecular reactions, the molecular weight and/or size of the branched polymers increases and finally reaches an "infinite" value with the formation of insoluble gel. Based on Flory-Stockmayer's mean-field theory, if all vinyl groups are assumed to have the same reactivity and no intra-molecular cyclization reactions are considered, gelation occurs in a system whenever the averaged number of cross-linking units (defined as cross-linkers with both vinyl groups reacted) per primary chain is larger than unity ([reacted M2P]/[PX]t > 1).
Conventional free radical polymerization (FRP) method has been extensively used for copolymerization of monomers and cross-linkers. However, highly branched polymers, including gels, are formed at very early stage of polymerization (low monomer conversions) and the polymerization behavior largely deviates from the prediction by Flory-Stockmayer theory due to several inherent features of FRP, including slow initiation, fast chain propagation, and termination reactions. In contrast to FRP, recently developed controlled radical polymerization (CRP) techniques are based on fast activation/deactivation equilibrium. The fast initiation reactions, relative to bimolecular termination, result in that all polymer chains are initiated at approximately the same time and the number of primary growing chains is nearly constant throughout the polymerization. The dynamic equilibrium, established between a low concentration of active propagating chains and a large number of "dormant" chains, ensures that only a few monomers are incorporated into the polymer chains in every activation/deactivation cycle. During the long "dormant" period, the polymer chains cannot propagate, but can diffuse and relax, which results in that the reaction possibility of each vinyl species (from monomer, cross-linker and pendant vinyl group) is statistically determined by their concentrations. Therefore, the sols or gels synthesized by CRP methods have a more homogeneous distribution of branching points than the polymers synthesized by FRP methods.
Using ATRP technique, series of poly(MA-co-EGDA) cross-linked copolymers were successfully synthesized by copolymerization of methyl acrylate (MA) and ethylene glycol diacrylate (EGDA) with various initial molar ratios of EGDA to ethyl 2-bromopropionate (EBrP) (X = [EGDA]0/[EBrP]0 = 1.1 ~ 10.0). During copolymerization, all initiators are consumed quickly and converted into short polymer chains containing pendant vinyl groups and halogen chain-end functionality. The fast activation/deactivation equilibrium between a low concentration of active propagating chains and a large number of "dormant" chains provides each chain a sufficiently long "dormant" period for diffusion and relaxation and, additionally, suppresses the irreversible termination. The propagating chain ends can react with monovinyl monomers and cross-linkers, but also can react with the pendant vinyl groups in polymer chains and form branched polymers. The branched polymers formed by inter-molecular reactions, represent a higher generation of polymer molecules with enhanced possibility to react with each other than the linear primary chains, because they contain more pendant vinyl groups and chain-end active centers than their linear counterparts. Therefore, branched polymers grow from smaller, low-generation molecules to larger and high-generation molecules and finally form a polymer network when the number of reacted pendant vinyl groups per primary chain reaches a critical value (Scheme).
The experimental gelation points were determined based on MA and/or EGDA conversions
when the reaction fluid lost its mobility at an upside down position for 10
seconds. The experimental gel points were determined for series of poly(MA-co-EGDA)
cross-linked copolymers by using various molar ratios of X = [EGDA]0/[EBrP]0.
In contrast, the theoretical gel points based on the assumptions of Flory-Stockmayer's mean-field theory were determined by Predici simulation. When the concentration of reacted pendant vinyl groups ([reacted M2P]) exceeds the amount of primary chains ([PX]t) in the system, statistically an averaged primary chain contains at least one reacted cross-linking unit. Thus, the theoretical gel point is reached and the gelation occurs, if no cyclization reaction is considered.
Series of copolymerizations of MA and EGDA with fixed concentration of MA but various molar ratios of X = [EGDA]0/[EBrP]0 were simulated by Predici. The theoretical gel points based on MA conversions were obtained. The results in Figure 1 show the experimental gel points are very close to, but always slightly higher than the theoretical values, indicating a homogeneous structure of the gels synthesized by ATRP technique. The higher experimental gel points than theoretical values were caused by the inevitable cyclization reactions during experiments. As discussed above, the intra-molecular cyclization consumed pendant vinyl groups but made no contribution to the increase of molecular weights. Since probability of cyclization reactions increases with a decrease of monomer concentration, it should enhance the difference between the experimental gel points and the theoretical values. When [MA]0 = 5.98 M, the difference between the theoretical values and experimental values was small, indicating that the contribution of cyclization reactions under these conditions was insignificant. This is a fundamental difference from the gel synthesized by FRP techniques.
Gao, H.; Min, K.; Matyjaszewski, K. Macromolecules 2007, 40, 7763
4. Synthesis of Telechelic Polymers by Chain-End Modification Using "Click Reactions"
Alpha, Omiga-dibromo-terminated polystyrene (PS) was synthesized by atom transfer radical polymerization (ATRP) using CuBr/N,N,N',N'',N''-pentamethyldiethylenetriamine (PMDETA) as catalyst and dimethyl 2,6-dibromoheptanedioate as initiator. The two bromo chain-end groups were then transformed to azido groups by nucleophilic substitution with sodium azide in N,N-dimethylformamide (DMF). Subsequently, the reaction between the azido groups and an excess of propargyl alcohol in DMF with CuBr/PMDETA as the catalyst produced a,w-dihydroxy-terminated PS polymers via "click reactions". The results from 1H NMR spectroscopy indicated fast and complete transformation from bromo to azido chain-end groups, and after click reactions with propargyl alcohol, nearly quantitative conversion of the end groups was observed. Gradient polymer elution chromatography (GPEC) and GPEC¡äsize exclusion chromatography (GPEC¡äSEC) two-dimensional (2D) liquid chromatography techniques were employed to characterize the resulting polymers from click reactions. GPEC analysis successfully separated the dihydroxy-containing PS from the monohydroxy- and nonhydroxy-containing PS species and quantified the fraction of each species at various reaction times.
Gao, H.; Louche, G.; Sumerlin, B. S.; Jahed, N.; Golas, P.; Matyjaszewski, K. Macromolecules 2005, 38, 8979
Gao, H.; Min, K.; Matyjaszewski, K. Macromol. Chem. Phys. 2007, 208, 1370
5. Characterization of Complex (Co)polymers via HPLC
Major Techniques:
A. Liquid Chromatography at Critical Conditions (LCCC)
Since the 1950's, HPLC has emerged as a powerful technique to analyze various molecular distributions in synthetic (co)polymers. Size exclusion chromatography (SEC) is the most prevalent example of the use of HPLC for polymer characterization. Because of the simple relationship between hydrodynamic volume and molecular weight for linear homopolymers, SEC has become the established method to determine the molecular weight and MWD of the synthetic polymers. However, two intrinsic reasons hinder SEC from being an effective tool in fully characterizing block copolymers. The first reason is the low resolution of SEC, which in most cases cannot fully separate block copolymer from its precursor MI. The second reason is that the hydrodynamic volume of a copolymer is influenced by both molecular weight and chemical composition. Specifically, SEC cannot provide information on the MWD of each individual block in the whole block copolymer. Therefore, new HPLC methods, such as liquid chromatography at critical conditions (LCCC) and Gradient Polymer Elution Chromatography (GPEC), were developed, which consider the contribution of the enthalpic interaction between the analytes and the stationary phase in the column as a factor for polymer separation.
In LCCC analysis of a homopolymer polyA, the entropic size exclusion effect and enthalpic interaction effect between polyA chains and the stationary column compensate for each other. The Gibbs free energy of the macromolecule does not change when the polymer enters the stationary column. Thus, the elution volume of polyA in the column is equal to the total void volume of the column and is independent of its molecular weight under a certain scale of investigation. At this point, polyA can be regarded chromatographically "invisible" and the retention of the diblock copolymer is solely determined by the "visible" block(s).
An interesting application of LCCC is the characterization of block copolymers since it is important to evaluate the efficiency of initiation from the MI during the synthesis of block copolymers. A theoretical model shows that under the critical condition for polyA homopolymer, the elution behavior of polyA-polyB block copolymer (polyA is defined as MI) is entirely dependent on the length of polyB block; while under the critical condition for polyB homopolymer, the elution behavior of the block copolymer depends only on the length of polyA block. Therefore, for a polyA-polyB block copolymer with polyA as MI, the LCCC for polyA can be used to characterize the initiation efficiency of the polyA MI; while the LCCC for polyB can be used to determine the existence of any polyB homopolymer in the final product. Furthermore, the molecular weight and its distribution of the "visible" block can be determined by a standard calibration curve, as commonly used in SEC. It is worth noting that if the theoretical assumption in LCCC analysis were valid, LCCC technique would undoubtedly be a powerful method for the characterization of selected blocks within block copolymers.
Gao, H.; Min, K.; Matyjaszewski, K. Macromol. Chem. Phys. 2006, 207, 1709
B. Gradient Polymer Elution Chromatography (GPEC)
GPEC separates polymers according to their interaction with the stationary column. As the mobile phase gradually changes from a poor solvent to a good solvent, the chains with weaker column interactions are expected to elute first. An example shown below illustrates the GPEC separation of telechelic polystyrnee (PS) linear polymers with distribution of hydroxyl groups. A normal phase bare-silica column was used as the stationary phase. When the eluent composition changed from hexane to THF, the linear PS chains containing fewer hydroxy end-groups should elute earlier than the chains containing more hydroxy end-groups. It is worth noting that the enthalpic interaction energy between the PS chains and the stationary phase is determined by both the molecular weight and the hydroxyl functionality. The GPEC retention behavior of nonhydroxy-PS standards was determined in adsorption mode (supporting information). However, with the existence of hydroxy groups at the chain end, the elution behavior of the PS chains was mainly determined by the hydroxyl functionality.
C. Two Dimensional Liquid Chromatography (2D-LC)
Min, K.; Gao, H.; Matyjaszewski, K. J. Am. Chem. Soc. 2005, 127, 3825
6. Synthesis of Thermo-sensitive Poly(N-isopropyl acrylamide) (PNIPAM) Nanocapsules
Gao, H.; Yang, W.; Min, K.; Zha, L.; Wang, C.; Fu, S. Polymer 2005, 46, 1087