Supramolecular Engineering with Macromolecules: An Alternative Concept for Block Copolymers

Abstract

Metallo-supramolecular AB block copolymers are the first examples of macromolecular engineeering. A new concept for the formation of block copolymers by a supramolecular approach is presented. For this purpose the authors utilized an asymmetrical bis(terpyridine)–ruthenium complex as the supramolecular linkage between the two blocks (see figure). Block copolymers represent an important class of materials, which have received widespread attention because of their remarkable micro- and nanophase morphology. This morphology leads to unique properties compared to homopolymers or their blends. Classical examples of block-copolymer morphology are lamellae, hexagonal-packed cylinders and body-centered-cubic arrays of spheres.1 During the last decades, important advances have been made in the synthesis, characterization, and application of block copolymers. In particular, anionic polymerization has been successfully applied in their controlled synthesis.2 Several other routes have been realized as well, for example, controlled radical polymerization,3 cationic polymerization,4 group transfer,5 and metathesis,6 or combinations of such techniques. Nevertheless, block-copolymer synthesis remains a challenge for certain materials and several interesting combinations could not be realized to date. On the other hand, recent developments in the field of supramolecular chemistry have shown that small (self-)complementary building blocks can lead, through self-organization processes, to large, well-defined structures, which are held together by noncovalent interactions such as hydrogen bonds7 and metal-to-ligand coordination.8 Herein we present a new highly controlled and well-defined construction principle for block copolymers that utilize supramolecular interactions, in this case metal-to-ligand coordination. By this method new combinations of block copolymers can be prepared, which are not accessible, or have been very difficult to access to date. This allows a comparison of the new metallo-supramolecular compounds with classical well-documented covalent block copolymers. For this purpose we chose the terpyridine ligand as the central building unit, which is well-known for its outstanding ability to form stable bis complexes with a large variety of transition-metal ions (Figure 1).9 The main advantage of having a metal complex as the bridging unit is the possibility of cleavage at this junction point. Therefore new “smart materials” are accessible. Moreover, the metal ion being at the interface between the blocks may cause additional interesting features regarding morphology, thermal and mechanical as well as photophysical properties. Bis(terpyridine) metal complex (left: schematic representation, right: results of the molecular modeling study). The construction of the supramolecular block copolymers works along the same principle as their covalent counterparts. First, defined oligomers of narrow polydispersities (<1.15) and with functional end groups (telomers) are formed by controlled polymerization methods. The terpyridine ligand is introduced at the chain end(s) of the telomers by simple organic reactions. Combining the different terpyridine-functionalized oligomers by self-organization processes directly leads to the desired block copolymers (Figure 2). An extremely powerful tool in the construction of well-defined supramolecular block copolymers is the ability of the terpyridine ligand to bind selectively only to a single transition-metal ion and form a “mono”-complex under specific conditions (1:1 ratio of ligand:metal ion).10 In particular, ruthenium ions have been used successfully in supramolecular chemistry to build up asymmetric complexes (Figure 2).11 Modular engineering of An, AB-, and ABA-metallo-supramolecular block copolymers. Herein we have used terpyridine-terminated polymers to engineer supramolecular AB, (AB)n, and ABA block copolymers. To show the feasibility of this concept we chose well-known systems which also have covalent counterparts. Recently we demonstrated that the terpyridine ligand could be successfully introduced into hydroxy-functionalized oligomers of poly(ethyleneoxide) (PEO) and poly(tetramethyleneoxide) (PTMO) in a suspension of KOH in DMSO.12 This strategy has now been extended to hydroxy-terminated polystyrene (PS)13 and poly(ethylene-co-butylene) (PEB; Scheme 1), by utilizing toluene as the solvent and using a crown ether to improve the base solubility. The terpyridine ligand could react in high yields through a Williamson type ether formation reaction. For this purpose, different reaction conditions were optimized for each class of polymers. The complete derivatization of the prepolymers with terpyridine-moieties was demonstrated by 1H NMR, 13C NMR, UV/Vis, gel-permeation chromotography (GPC) and MALDI-TOF MS. To form AB-type polymer structures selectively, the poly(ethylenoxide) 1 was treated with Ru(III)Cl3, which resulted exclusively in the desired monocomplex 4 (simply isolated by precipitation). This compound was then further treated, in a self-organization step, with the terpyridine-functionalized polymers 1, 2, and 3, respectively. In this approach, ethanol is utilized as the reducing agent and N-ethylmorpholine as the catalyst, to reduce the RuIII ions to RuII ions. Subsequently, a bis complex was formed between the ruthenium-filled terpyridine unit and the uncomplexed terpyridine unit, which resulted in unsymmetrical dimers. After exchange of the counterions by addition of an excess of NH4PF6, the supramolecular AB block copolymers 5, 6, and 7 were isolated by precipitation or extraction (Scheme 2). As an additional purification step, size-exclusion chromatography (BioBeads SX-1 in CH2Cl2) was applied. The low yields (35–50 %) could be explained by the extensive purification procedures in conjunction with the small amounts we were working with to date. The corresponding model complexes were isolated in yields up to 97 %.9b Low selectivity was not an issue, because homopolymers (AA or BB) could not be isolated. Terpyridine-functionalized building blocks, poly(ethyleneoxide) 1, poly(ethylene-co-butylene) 2, polystyrene 3, and poly(ethyleneoxide) treated with the RuIIICl3 compound 4. Isolated metallo-supramolecular block copolymers: PEO70-[Ru]-PEO70 (5), PEO70-[Ru]-PEB70 (6), and PS20-[Ru]-PEO70 (7). Characterization by UV/Vis spectroscopy revealed the characteristic metal–ligand charge-transfer (MLCT) band of the bis(terpyridine) ruthenium(II) complex at 490 nm, while the band at 390 nm—indicative of the mono(terpyridine) ruthenium(III) complex—has completely disappeared (Figure 3). This result is already unambiguous evidence for the successful formation of the block copolymers. In addition, because of the complexation of the ruthenium(II) ions, the signals arising from the terpyridine unit in both the 1H and 13C NMR spectra have shifted when compared to the uncomplexed terpyridine-functionalized prepolymers. All signals can be assigned by a comparison with the corresponding low-molar-mass model complexes. The integral ratio of the two polymer blocks was in good agreement with the expected ratio in all three cases. Also, this finding excludes the existence of any “homodimers”. MALDI-TOF MS revealed the presence of the different blocks in the supramolecular block copolymers, although the supramolecular connection seems to be partly broken during the MALDI process (as is known for this type of metal complex). UV/Vis spectra of supramolecular block copolymers showing the metal-to-ligand charge-transfer band (MLCT) at 490 nm (in CH3CN): — 5, – – – 6, •••• 7. Size-exclusion chromatography (GPC), which applied a UV/Vis as well as a refractive-index (RI) detector, revealed a single signal. This clearly shows that the metallo-supramolecular block copolymer was formed and that no homopolymers are present (Figure 4). The similar molar masses of 5, 6, and 7 result in a similar elution volume. The absence of further peaks also shows that the compounds are stable on the GPC column. Moreover, by varying the pH value from 0 to 14 in water, no changes in the UV/Vis spectrum could be observed, which indicates the unusual stability of the ruthenium complex. However, by utilizing redox processes or a large excess of a competitive chelator at higher temperatures (such as hydroxyethylethylenediaminotriacetic acid sodium salt, HEEDTA), the metal complex can be disassembled. Because the described A-[Ru]-B block copolymers represent amphiphilic materials, the formation of defined micelles is possible (see ref. 14 for the first examples). This oppportunity will certainly be of interest for applications in supported catalysis and nanotechnology. GPC traces of a) 5, b) 6, and c) 7 measured using a UV detector (eluant CHCl3, polystyrene calibration standards). In conclusion, we have demonstrated that a supramolecular approach towards block copolymers is feasible. In particular, combinations of blocks that to date have been inaccessible, or accessible by established routes only with great difficulties, can be engineered utilizing the described approach. By introducing the terpyridine ligand at the chain end(s) of different polymers of various lengths, a large number of building blocks for block copolymers can be obtained (several hundreds of potential functional prepolymers and telechelic molecules are already commercially available). Applying the strategy employed herein and using different metal centers and counterions, the possibilities for assembling supramolecular block copolymers of the type described are virtually unlimited. This approach is highly suitable for the development of libraries of block copolymers by combinatorial methods, which vary the block properties systematically. Initial experiments of this type are currently in progress. In addition, the reversibility of the supramolecular connections opens up avenues for the construction of “switchable” materials. This can lead to new applications, for example, as intelligent glues or release-on-demand systems. General preparation of 1–3: Powdered KOH and the prepolymer (3:1 ratio) were stirred under Ar in dry DMSO at 70 °C or dry toluene with [18]-crown-6, heated under reflux. After 30 min, 4′-chloro-2,2′:6′,2′′-terpyridine was added. The mixture was stirred for 24 h–48 h, then poured into cold water and extracted with CH2Cl2. The combined organic layers were dried over Na2SO4 and the solvent removed in vacuo. All compounds were purified by preparative size-exclusion chromatography (BioBeads SX-1, CH2Cl2) and the compounds were subsequently precipitated by addition of a solvent in which they are insoluble. Compound 2 was further purified by column chromatography (silica, CH2Cl2 followed by CH2Cl2:Et3N 9:1). Yields were 9.75 g of 1 (95 %), 3.9 g of 2 (46 %), and 2.10 g of 3 (75 %). Selected analytical data: 1: 1H NMR (400 MHz, CDCl3, 25 °C): δ=8.68 (dt, J=4.8, 2.0 Hz, 2 H; H6, H6′′), 8.61 (dt, J=8.0, 2.0 Hz, 2 H; H3, H3′′), 8.04 (s, 2 H; H3′, H5′), 7.85 (td, J=8.0, 2.0 Hz, 2 H; H4, H4′′), 7.34 (ddd, J=8.0, 4.8, 2.0 Hz 2 H; H5, H5′′), 4.40 (t, J=5.2 Hz, 2 H; tpyOCH2), 3.93 (t, J=5.2 Hz, 2 H; tpyOCH2CH2) 3.83–3.45 (m, 280 H; PEO backbone), 3.38 ppm (s, 3 H; OCH3); UV/Vis (H2O): λmax=278, 234 nm; MALDI-TOF MS: Mn=3436 g mol−1; GPC (RI): Mn=2360, polydispersity index (PDI)=1.21. 2: 1H NMR (400 MHz, CDCl3, 25 °C): δ=8.69 (dt, J=4.8, 2.0 Hz, 2 H; H6, H6′′), 8.62 (dt, J=8.0, 2.0 Hz, 2 H; H3, H3′′), 8.01 (s, 2 H; H3′, H5′), 7.85 (td, J=8.0, 2.0 Hz, 2 H; H4, H4′′), 7.32 (ddd, J=8.0, 4.8, 2.0 Hz, 2 H; H5, H5′′), 4.25 (t, J=6.4 Hz, 2 H; tpyOCH2), 1.58–0.68 ppm (m, 560 H; PEB backbone); UV/Vis (CHCl3): λmax=278, 243 nm; MALDI-TOF MS: Mn=4151 g mol−1; GPC (UV): Mn=4450, PDI=1.19. 3: 1H NMR (400 MHz, CDCl3, 25 °C): signals are broad δ=8.67 (bm, 2 H; H6, H6′′), 8.60 (bm, 2 H; H3, H3′′), 7.91 (s, 2 H; H3′, H5′), 7.83 (bm, 2 H; H4, H4′′), 7.31–6.39 (bm, 102 H; C6H5 PS backbone; H5, H5′′), 4.11–3.96 (bm, 2 H; tpyOCH2), 2.66, 2.18–1.10, 0.78 ppm (bm, 49 H; CH2, CH PS backbone); UV/Vis (CHCl3): λmax=278, 243 nm; MALDI-TOF MS: Mn=2104 g mol−1; GPC (UV): Mn=1850, PDI=1.10. Preparation of 4: A solution of 2 (1.00 g, 0.29 mmol) in methanol (25 mL) was stirred at 60 °C. Then an equimolar amount of RuCl3⋅n H2O (0.076 g, 0.29 mmol) was added. Stirring was continued overnight. Subsequently, the reaction mixture was cooled to −20 °C. The resulting dark orange precipitate was collected by filtration and washed twice with ice-cold water, followed by diethyl ether, yielding 1.05 g of product (99 %). 1H NMR (400 MHz, CDCl3, 25 °C): δ=3.92–3.46 (m, 280 H; PEO backbone), 3.38 ppm (s, 3 H OCH3); UV/Vis (H2O): λmax=272, 375 nm; MALDI-TOF MS: Mn=3447 g mol−1. General procedure for the preparation of 5–7: A solution of 1, 2, or 3 (1:1 ratio), respectively, and 4 in MeOH (5 mL) was stirred for 30 min, heating under reflux. A few drops of N-ethylmorpholine was added and the solution turned from orange to red. Heating under reflux was continued overnight, after which an excess of NH4PF6 (45 mg, 0.27 mmol) was added. The crude products were collected by filtration of the precipitate or extraction with CH2Cl2 after pouring them into water. Further purification was carried out by preparative size-exclusion chromatography (BioBeads SX-1, THF and/or CH2Cl2). Excess NH4PF6 was washed out with water by dissolving 5, 6, or 7 in CHCl3. Yields were 97 mg of 5 (50 %), 76 mg of 6 (36 %), and 40 mg of 7 (35 %). Selected analytical data: 5: 1H NMR (400 MHz, CD3CN, 25 °C): δ=8.49 (dt, J=8.0, 1.2 Hz, 4 H; H3, H3′′), 8.35 (s, 4 H; H3′, H5′), 7.91 (td, J=8.0, 1.2 Hz, 4 H; H4, H4′′), 7.37 (dt, J=5.6, 1.2 Hz, 4 H; H6, H6′′) 7.16 (ddd, J=8.0, 5.6, 1.2 Hz, 4 H; H5, H5′′), 4.67 (t, J=4.4 Hz, 4 H; tpyOCH2), 4.05 (t, J=4.4 Hz, 4 H; tpyOCH2CH2), 3.80–3.38 (m, 570 H, PEO backbone), 3.31 ppm (s, 6 H; OCH3); UV/Vis (CH3CN): λmax: 486, 304, 267 nm; GPC (UV): Mn=16 780, PDI=1.13; MALDI-TOF MS: Mn=6627 g mol−1. 6: 1H NMR (400 MHz, CDCl3, 25 °C): δ=8.40 (dt, J=7.2, 1.2 Hz, 2 H; H3, H3′′ PEO), 8.36 (dt, J=8.0, 1.2 Hz, 2 H; H3, H3′′, PEB), 8.29 (s, 2 H; H3′, H5′, PEO), 8.17 (s, 2 H; H3′, H5′, PEB), 7.81 (m, 4 H; H4, H4′′, PEB + PEO), 7.37 (m, 4 H; H6, H6′′, PEB + PEO), 7.18 (m, 4 H; H5, H5′′, PEB + PEO), 4.76 (t, J=4.0 Hz, 2 H; tpyOCH2, PEO), 4.58 (t, J=4.8 Hz, 2 H; tpyOCH2, PEB), 4.09 (t, J=4.0 Hz, 2 H; tpyOCH2CH2, PEO), 3.87–3.41 (m, 290 H, PEO backbone), 3.38 (s, 3 H; OCH3), 1.68–0.80 ppm (m, 580 H, PEB backbone); UV/Vis (CH3CN) λmax: 486, 304, 267 nm; GPC (UV): Mn=12 410, PDI=1.10; MALDI-TOF MS: Mn=7502 g mol−1. 7: 1H NMR (400 MHz, CDCl3, 25 °C): δ=8.36 (td, J=8.0, 1.2 Hz, 2 H; H3, H3′′, PEO), 8.26 (s, 2 H; H3′, H5′, PS), 8.17 (bm, 2 H; H3, H3′′, PS), 7.88–7.73 (bm, 6 H; H3, H3′′(PS), H3′, H5′ (PS), H4, H4′′ (PEO)), 7.33–6.32 (m, 110 H; C6H5 PS backbone and tpy signals), 4.74 (t, J=3.6 Hz, 2 H; tpyOCH2, PEO), 4.28–4.04 (m, 4 H, tpyOCH2CH2 (PEO), tpyOCH2 (PS)), 3.92–3.42 (m, 290 H; PEO backbone), 3.38 (s, 3 H; OCH3), 1.72–0.60 ppm (m, 50 H; CH2, CH PS backbone); UV/Vis (CH3CN): λmax: 486, 305, 268 nm; GPC (UV): Mn=7800, PDI=1.08; MALDI-TOF MS: Mn=5348 g mol−1.