Solid Phase Synthesis of Polymacromer and Copolymacromer Brushes

We report a novel solid phase method for the sequential coupling of heterobifunctional macromonomers to form a new class of polymeric materials that we refer to as polymacromers. Starting from an azide functional substrate, αazido, ω-protected-alkyne macromonomers are added sequentially by thermally initiated click reactions to form polymacromer brushes wherein macromonomers are linked via triazole groups. After each addition step, the terminal alkyne group can be deprotected to allow addition of the next macromonomer. The use of highly chemoselective click chemistry for the coupling reactions allows virtually any macromonomer to be employed, regardless of its chemical nature. The method is illustrated by formation of two different homopolymacromer brushes: one made by sequentially coupling four polystyrene macromonomers and another by sequentially coupling eight poly(tert-butyl acrylate) macromonomers. Two different types of copolymacromer brushes are also prepared: one by alternating sequential coupling of polystyrene macromonomers with poly(methyl methacrylate) macromonomers and another by alternating sequential coupling of polystyrene macromonomers with poly(tert-butyl acrylate) macromonomers. The polymacromer brushes are characterized by ellipsometry, contact angle analysis, X-ray photoelectron spectroscopy, and X-ray reflectivity measurements. In all cases the coupling reactions are found to be highly efficient. Analysis of thickness data for the poly(tert-butyl acrylate) homopolymacromer brushes indicates that a conversion within the range of 97.5−100% is achieved for each cycle. This solid phase synthesis method is extremely versatile because it can be used to prepare complex sequenced block copolymers from almost any polymers and because the brushes formed can be made to span the entire region of brush behavior between the “grafting from” and grafting to” limits by varying the macromonomer molecular weight. T development of precise techniques for preparing polymers of well-controlled architecture and narrow molecular weight distribution is a major goal of polymer chemistry. We describe herein the general principles of a new modular technique for polymer synthesis wherein a wide variety of preformed molecular building blocks can be simply linked together much like molecular Tinker Toys or Lego building blocks to form complex polymer and copolymer architectures. The new technique is based upon solid phase synthesis (SPS) concepts embodied in the Merrifield synthesis of oligopeptides. Solid phase synthesis techniques are modular in nature, generally involving stepwise addition of heterobifunctional building blocks to an initial reactant that is bound to a solid support. The buildup of specific sequences of the building blocks is controlled by the order of reactant addition mediated by functional group protection−deprotection strategies. Compared to conventional solution synthesis, SPS allows for the construction of nonsymmetric, sequenced structures as well as more facile removal of excess reactants and byproducts because the desired product is covalently bound to the solid substrate. SPS has been used recently in combinatorial chemistry and has become an essential route for the preparation of many important biological molecules including peptides, DNA, and other biopolymers for which a certain nonsymmetric predetermined sequence is desirable. Relatively fewer examples have appeared where solid phase techniques have been used in synthetic polymer synthesis and the majority of these efforts have been restricted to the synthesis of low molecular weight oligomers. Solid phase methods have been used to prepare oligothiophenes, poly(amidoamines), peptide−polymer hybrids, and dendrimers as well as triazole oligomers and their cyclics, the latter from peptide monomers modified to have terminal alkyne and azide groups for coupling by copper(I)-catalyzed azide− alkyne cycloaddition (CuAAC). Stepwise protection− deprotection processes have also been applied to prepare a Received: February 28, 2012 Revised: April 6, 2012 Published: April 25, 2012 Article pubs.acs.org/Macromolecules © 2012 American Chemical Society 3866 dx.doi.org/10.1021/ma3004168 | Macromolecules 2012, 45, 3866−3873 number of π-conjugated oligomers. An interesting recent variation on the theme of solid phase synthesis is the use of soluble polymer supports (liquid supported synthesis) prepared from initiators that contain the same cleavable p-alkoxybenzyl ester linkers used in solid phase Wang resins. “Molecularly defined” oligomers containing up to 64 ε-caprolactone monomers were synthesized using another liquid supported method (referred to as an iterative divergent−convergent approach), in which a substrate was avoided by clever use of two separate, orthogonal protection strategies to promote nonsymmetric monomer addition. A similar liquid phase dual protection approach was coupled with CuAAC to prepare monodisperse oligomers and macrocycles of ethylene oxide, its dimers, and its trimers. The few investigations that have used solid phase and similar liquid-supported methods for polymer synthesis have produced inspiring results but have only begun to tap the tremendous potential of supported polymer synthesis methods. In this first paper, we apply SPS principles to develop a modular method for the synthesis of a new class of polymers we term “polymacromers”. Polymacromers (PM) are multiblock polymers of a defined sequence produced by coupling heterobifunctional macromonomer building blocks. As prepared, they take the form of a polymer brush because they are bound to a solid substrate but can in principle be isolated by incorporating a cleavable linker at the substrate interface. In conventional polymer synthesis, multiple monomers are joined to form a “polymer”, where the term “polymer” is derived by taking the word “monomer” that describes the repeating unit, adding to it the syllable poly, meaning many, and dropping the syllable mono. If we apply the same logic to our new class of materials, that is, by adding poly and dropping mono from the word “macromonomer”, the result is the term polymacromer, which we believe is a suitable and accurate descriptor for the new materials. It follows that homopolymacromers are prepared by multiple addition cycles using a single macromonomer while segmented block copolymacromers of a desired sequence are prepared using different macromonomers in each cycle. Azide−alkyne click chemistry via the Hüisgen 1,3-dipolar cycloaddition mechanism is used as the coupling chemistry because the alkyne group can be readily protected and subsequently deprotected using relatively mild conditions. This click reaction can be initiated at room temperature with a copper catalyst (CuAAC) or thermally at temperatures as low as 70 °C without catalyst. The high chemoselectivity of the reaction enables the coupling of virtually any macromonomer, regardless of its chemical nature and the required heterobifunctional macromonomers, comprising one azide end group and a second protected alkyne end group, are readily prepared by atom transfer radical polymerization (ATRP). ■ RESULTS AND DISCUSSION The procedure for preparing polymacromer brushes by SPS is described schematically in Figure 1. The critical requirement that enables this strategy is an ability to synthesize heterobifunctional building blocks, heterobifunctional macromonomers (HetBi), that are terminated at one end with an azide group (N3) and on the other end with a trimethylsilane protected alkyne group (TMS−≡) as depicted in the figure. HetBi functional macromonomers of this nature can be prepared by atom transfer radical polymerization (ATRP), a controlled radical polymerization technique, using a functional initiator as described in the Supporting Information. The SPS method begins with functionalization of the substrate with surface alkyne groups (≡), accomplished by self-assembly of alkyne-functional phosphonate esters (metal oxide substrates), alkyne-functional thiols (gold substrates), alkynefunctional silanes (glass substrates and silicon wafers), or alkyne-functional block copolymers (glass and polymeric substrates). Covalent attachment of the first macromonomer (red chain in Figure 1) involves a “click” reaction (i.e., 1,3-cyclopolar addition) between the azide terminus on the red macromonomer and an alkyne group on the functionalized substrate surface, coupling the macromonomer to the surface via a triazole linkage. The result is a substrate coated with a covalently bound brush of red polymer that presents trimethylsilane-protected alkyne groups (TMS−≡) at the surface. An alkyne functional surface (≡) is subsequently regenerated by deprotection of the TMS-protected alkyne groups. After surface alkyne groups are regenerated, a second macromonomer (blue chain in Figure 1), not necessarily the same as the first macromonomer, is attached by another click reaction. The coupling/deprotection process may be applied multiple times to prepare a desired copolymacromer sequence (the preparation of three distinct segments is illustrated in Figure 1). Because click coupling reactions are highly chemoselective, the chemical nature of the macromonomer blocks is only limited by our ability to synthesize appropriate heterobifunctional building blocks. While the coupling of macromonomers is illustrated herein, our SPS method is quite flexible, and a versatile toolkit of molecular building blocks can be imagined including monomers, branching units, or molecules that furnish specific functions such as sites for cleavage or for subsequent attachment of receptors or ligands. Two simple steps are required to add each building block: click coupling of the building block and subsequent deprotection to regenerate surface alkyne groups. Three HetBi macromonomers were synthesized to illustrate the new method for preparing polymacromer brushes: αalkyne-trimethylsilane-ω-azide-poly(styrene) (TMS-alkyne-PSN3) and α-alkyne-trimethylsilane-ω-azide-poly(tert-butyl acrylate) (TMS-alkyne-PtBA-N3) and α-alkyne-trimethylsilane-ωazide-poly(methyl methacrylate) (TMS-alkyne-PMMA-N3). These polymers were synthesized by ATRP using a trimethylsilane protected alkyne-functional ATRP initiator, followed by conversion of the resultant terminal bromine groups to azides by the addition of sodium azide (see Supporting Information for details). Molecular characteristics of the macromonomers are presented in Table 1, and their chemical structures are shown in Scheme 1. Figure 1. Schematic representation of the SPS process for preparing sequenced copolymacromer brushes. The green circles denote triazole linkages. Macromolecules Article dx.doi.org/10.1021/ma3004168 | Macromolecules 2012, 45, 3866−3873 3867 The initial step in SPS brush synthesis was to functionalize the substrate with alkyne groups. For glass and silicon wafers, the substrates used herein, this was accomplished by forming a self-assembled monolayer (SAM) of an alkyne functional silane on the substrate surface. The thickness of the SAM, determined by angle-dependent X-ray photoelectron spectroscopy analysis, was 1.8 ± 0.3 nm, while that measured by ellipsometry was 1.7 ± 0.2 nm. From the structure of the silane, the thickness is expected to be about 1.1 nm. Similar silanes with 4−7 methylene units have a reported thickness of 1.5−2.3 nm. The finding that these SAMs were thicker than expected is consistent with previous studies of other silanes that reported their polymerization to form multilayers. The water contact angle of the alkyne-functionalized substrate was 61.8 ± 1°, compared to <10 ± 2° for the bare glass substrate, as expected for the more hydrophobic silane monolayer. After functionalizing the substrate with alkyne groups, HetBi macromonomers were coupled to the surface by a click reaction between terminal azide groups of the macromonomers and surface alkynes. Each HetBi macromonomer was dissolved in toluene and spin-coated onto the surface, after which the coated substrates were placed in a vacuum oven and heated to 100−115 °C for 3−12 h to effect a thermally initiated “click” reaction (in the melt state) between substrate-bound alkyne groups and the azide termini of the polymers. After the reaction period, excess polymer was removed by extensive washing with solvent (DCM) for 1−24 h. Click reactions proceed to very high conversion under mild conditions with no side reactions or byproducts, and the resulting aromatic triazole is extremely stable. In addition, click reactions are highly chemoselective such that virtually any polymer backbone may be used in the SPS process without interfering with the click reactions used to bond adjacent macromonomers. The thermal stability of the TMS protecting groups was verified by a control experiment in which we attempted to couple a macromonomer onto a polymer-modified surface that was not subject to the deprotection step. The thickness did not change when the deprotection step was omitted, indicating that no reaction occurred and that the TMS protecting groups are stable under the conditions used for the thermal click reaction. Substrate bound alkyne groups were regenerated after macromonomer addition by deprotection of the terminal TMS−alkyne groups, accomplished by dipping the substrate into a KCO3-saturated solution of 10:1 DCM/MeOH. The regenerated alkyne surface was then used to couple additional macromonomers by repeating the two-step addition cycle with other HetBi macromonomers. The first validation of the SPS method was homopolymacromer formation by multiple addition of TMS-alkyne-PSN3. Successful sequential addition of multiple macromonomers is indicated by the thickness data in Figure 2. (Details of the ellipsometry analysis are provided in the Supporting Information.) When the first macromonomer is added to the alkyne−silane functionalized substrate, surface alkyne groups are in excess and a PS brush with a thickness of about 4 nm is formed. If each macromonomer is assumed to form a cube with the density of bulk PS (note, the coupling is performed in the melt), the expected thickness of a monomolecular layer would be about 3.24 nm. Polystyrene chains in the first layer therefore assume somewhat extended configurations, consistent with the results of previous studies of polystyrene brushes prepared by end-grafting from the melt. The thicknesses for the second through fourth addition cycles are linearly dependent on the number of macromonomers added, with each addition cycle adding about 2 nm to the overall film thickness. The fact that the thickness increase is the same for each macromonomer addition cycle is an important result because it suggests that the conversion of the interfacial click reactions is effectively complete for each macromonomer addition cycle. In Table 1. Number-Average Molecular Weights (Mn), WeightAverage Molecular Weights (Mw), and Polydispersity Indices (PDI) of the Polystyrene, Poly(tert-butyl acrylate), and Poly(methyl methacrylate) HetBi Macromonomers Determined by Gel Permeation Chromatography (GPC) polymer code Mn [Da] Mw [Da] PDI Mn,adjusted [Da] TMS-alkyne-PS-N3 21 500 24 000 1.12 21 500 TMS-alkyne-PtBA-N3 17 000 20 000 1.17 22 170 TMS-alkyne-PMMA-N3 12 000 20 000 1.67 14 600 Adjusted Mn values employ a universal calibration based upon literature values of Mark−Houwink−Sakurada parameters to correct the GPC molecular weight for hydrodynamic volume effects. Scheme 1. Chemical Structures of HetBi Macromonomers: TMS-Alkyne-PS-N3 (top); TMS-Alkyne-PtBA-N3 (middle); TMS-Alkyne-PMMA-N3 (bottom) Figure 2. Ellipsometric thicknesses for SPS films prepared from TMSalkyne-PS-N3. The dashed line shows predictions for complete conversion based upon the areal density of the brush comprising one macromonomer layer, and the dotted line shows predictions based upon the areal density of brushes comprising two macromonomers. Macromolecules Article dx.doi.org/10.1021/ma3004168 | Macromolecules 2012, 45, 3866−3873 3868 other words, when a macromonomer adds to the growing brush on the substrate, the areal density of peripheral alkyne groups remains constant, each chain adding exactly one macromonomer in true step-growth fashion. If the conversion were less than complete, the areal density of alkyne groups and consequently the areal density and thickness of PS chains added would decrease each time a macromonomer was added. It is quite remarkable that complete conversion is apparently achieved for brushes prepared in a sequential “grafting to” fashion as is done herein. Certain attributes of the method are conducive to this result. First of all, the alkyne group is one of few reactive functional groups that has a low surface tension (estimated to be 26 mN/m by group contribution methods) and is therefore expected to segregate preferentially to the surface of most polymers. Second, the molecular weight of macromonomer added is identical in each cycle so that the occupied volume and functional group density of each are also the same. Third, because each cycle comprises the same PS macromonomer, there can be significant interpenetration between the brush and the macromonomer reacting to it. Interpenetration across the interface increases the effective volume for the reaction since complementary functional groups can only meet within the zone of interpenetration. Completion of the reaction for each layer can be verified by calculating the areal density of functional groups at the surface of the first PS layer and using this value to predict the thickness of subsequent layers. The areal density of functional groups for a tethered polymer brush layer can be calculated from the measured thickness, molecular weight, and density according to