Towards the Synthesis of Helical Multisubstituted Oligoimidazole Foldamers

Towards the Synthesis of Helical Multisubstituted Oligoimidazole Foldamers

Eric L. Spitler, Noel M. Paul, and Jonathan R. Parquette*

Abstract: A method towards the synthesis of an oligomer of imidazole substituted with two phenyl ring-containing groups is presented, with the premise that the resulting molecule will have a structure that inherently affords it well-defined conformational (specifically, helical) properties.

Recent studies in organic chemistry have focused on the synthesis and characterization of large molecules with clearly defined conformational properties.1 One incarnation of these studies is the concept of rigidly arranged chains of molecules with conformational properties inherent in their structure that cause them to fold into a certain, predictable shape. Such molecules are known as foldamers.2,3 Peptides, proteins, and nucleic acids are a few examples of natural foldamers. Most foldamers synthesized up to now have been based on these natural structures. Helical foldamers have been synthesized, for example, by Günther and Hofmann from hydrazino peptides.4 In this paper, the synthesis of an oligomer of 5-membered nitrogen-containing aromatic rings, substituted with phenyl-containing groups at, as well as coupled through, the 2- and 5- positions, has been undertaken. Molecular mechanics calculations done by supercomputer have indicated that this manner of structure should, due to the π-stacking nature of aromatic rings, as well as the considerable steric hindrance associated with this kind of system, orient itself in a helical conformation along the axis of coupling. Such a structure could prove useful in such areas as enantioselective catalysis.

Fig. 1: Structure of intended helical foldamer – repeating sterically hindered groups rotate along the axis to the lowest energy conformation (i.e., a helical conformation)

The calculations also indicate that pyrrole would be an excellent candidate for oligomerization. The intended structure requires 5-membered aromatic rings joined through the 2- and 5- positions along an axis (perpendicular to the ring faces) that contain repeating, sterically bulky groups such as phenyl groups. Pyrrole can be added to at the 1- position, and can undergo electrophilic aromatic substitution at the 2- and 5-positions. Pyrrole, an electron-rich ring, is very reactive, and would therefore require also that a strongly electron-withdrawing group be placed at the 1-position to prevent oxidative and undesired polymerization reactions from taking place. A withdrawing group with additional steric effects would be preferred. Although results indicate that synthesis of the monomer from pyrrole is unfeasible, a similar route using imidazole, which is much less reactive due to inductive electron density withdrawal by the basic nitrogen, has been attempted and shows promise.

Fig. 2: Helical conformation of theoretical oligopyrrole (trimer ) and supercomputer calculation of 3-dimensional structure

The monomer from which the oligopyrrole is obtained is pyrrole substituted by aniline at the 2- and 5-positions, and a strongly electron-withdrawing group at the 1-position. Tert-butoxycarbonyl, a good withdrawing group that also has steric effects, was chosen for the intended synthesis:

Fig. 3: Intended monomer for oligoyrrole

The substituent “-NH-Ph” in needed for two reasons. The extra “vertex” allows the phenyl ring the freedom to orient itself in the desired conformation. Also, the intended oligomerization requires that the chain be linked through this intermediate group. The most facile method of accomplishing these goals is to use this arylamine, and couple the monomers through amide bonds.

The attempt to synthesize such a molecule began with a trisubstituted pyrrole ring, with tert-butoxycarbonyl as the withdrawing group:

Fig. 4: Synthesis scheme for monomer from pyrrole

This route was exhaustively pursued first by bromination5, then chlorination6 at the second step. However, it was concluded finally that the product was in the first case too unstable to be isolated, resulting only in a black sludge, probably a mixture of oxidation and polymerization products. In the second case the reaction proceeded too slowly to be useful and a red oil was obtained that did contain some of the desired product, but it was unable to be isolated and purified. Focus then turned to imidazole. It has been shown that 2,5-dibromoimidazole cannot be synthesized by the NBS method7:

Fig. 5: Dihalogenation of imidazole unfeasible

It became clear that a new approach was needed.

One promising method lies in the halogenation of imidazole rings that have been synthesized already substituted at the 1- and 4- positions, leaving only 2,5-dihalogenation possible. A slight modification of the work of Rapoport, et. al.8 could prove fruitful. Synthesis of 4,5-dicyanoimidazole from diaminomaleonitrile and triethyl orthoformate, followed by methylation at the 1- position, hydrolysis, and selective decarboxylation would afford 1-methyl-4-imidazolecarboxylic acid. Esterification, dihalogenation, and aniline displacement affords a modified tetrasubstituted target monomer:

Fig. 6: New synthesis scheme for monomer from imidazole

Although a more circuitous route, this synthesis is not only more likely to succeed, but for the most part has already been done. Synthesis by the Rapoport group of the diacid, followed by decarboxylation at the 4- position, esterification of the 5-acid, and bromination with NBS afforded the 2,4-dibromo product. 2,5-dibromination by the proposed route should be even more favored. The tetrasubstituted monomer may even prove preferable, since the added functionality at the 4-position allows for a great deal of manipulation of physical properties. Hydrolysis of the ester after oligomerization will make the foldamer more water-soluble, and we can even control the polarity by adding organic chains of any length through transesterification. Addition of many other types of groups is also possible:

Fig. 7: Selective functionalization of the oligomer.

For the proposed synthesis, diaminomaleonitrile and triethyl orthoformate were heated to give 4,5-dicyanoimidazole (DCI) and ethanol. Addition of a sub-catalytic amount of sodium methoxide caused the product to partially precipitate. Filtration and recrystallization yielded the DCI as fine, light brown-to-white crystals. The reaction proceeded cleanly, however workup was difficult and impurities resulted. When enough relatively pure DCI was obtained, deprotonation with sodium bicarbonate, then methylation with dimethyl sulfate took place in water with relative ease under modest conditions. The yellow-brown solid obtained by ethyl acetate extraction was hydrolyzed in 6N NaOH, then acidified with concentrated HCl to give the diacid as a white-to-off- white solid. The diacid was selectively decarboxylated at the 5- position by heating to 180oC in dimethyl acetamide, triturating with benzene, and recrystallizing from ethanol. The monoacid was esterified by refluxing in methanol with a catalytic amount of acetyl chloride. The white crystals were heated in CCl4 with N-bromosuccinimide and a catalytic amount of AIBN

Unlike in the case of pyrrole, an excess can be used, since 2,5-dibromaination is the only reaction possible. Still to be accomplished is displacement of the bromides with aniline to afford the monomer.

The next step after synthesis of the monomer is coupling. This can be carried out by first selectively protecting at the 2- aniline position with TMS-ethyl-chloroformate, then isolating half of the resulting compound and protecting the 5- aniline position with allyl-chloroformate, then cleaving the TMS-ethyl ester with HF in pyridine. The two portions are then combined in the presence of phosgene, or a coupling agent that acts in the same manner as phosgene, to make the dimer:

Fig. 8: Coupling the monomer.

Also, utilizing the regioselectivity of the monomer, a more direct approach would be simply to carefully combine the monomer and phosgene:

Fig. 9: A more direct (and risky) approach to coupling.

Appropriate modification of the coupling procedure can lead to an oligomer of any desired length. Synthesis will produce a racemic mixture of M and P type helices, to which chiral groups can be added and enantioselectively separated. Upon synthesis of the oligomer, conformation analysis will be undertaken to understand the actual folding properties versus computational values.

Experimental

1-tert-butoxycarbonylpyrrole9: 1.41g (21.06 mmol) pyrrole was put into a flame-dried 25 mL round bottomed flask and stirred with 45 mL dichloromethane. 0.27 g (2.29 mmol) DMAP was added with stirring. Upon dissolution of DMAP 5.51 g (25.27 mmol) di-tert-butyldicarbonate was injected under argon. The solution was stirred at rt 12 h under argon covered in aluminum foil to protect from light decomposition, with a balloon attached to a septum cap to allow for evolution of gas. 50 mL anhydrous diethyl ether was added to the resulting solution and washed with 1M NaHSO4 (25 mL) in a 125 mL seperatory funnel. Organic phase was washed with additional NaHSO4 (5 x 12.5 mL), venting frequently. Solution was then washed with 12.5 mL distilled water, 12.5 mL saturated NaHCO3 and 12.5 mL brine. Organic phase was dried with MgSO4 and solvent was distilled at atmospheric pressure and 40oC to remove ether, then at 10 mm Hg to remove dichloromethane. Further distillation at 110oC/ 10 mm Hg gave 3.18 g (90%) clear liquid solution. H1NMR (400 MHz, CDCl3 ):  = 1.60 (s, 9 H); 6.21 (t, 2 H, J = 2.36 Hz); 7.23 (t, 2H, J = 2.31 Hz).

4,5-dicyanoimidazole8: 10.81g (100.03 mmol) diaminomaleonitrile was put into a flame-dried 500 mL round bottomed flask with 120 mL anisole, and 18.20g (122.81 mmol) was added with stirring, then heated in a bath at 135oC while distilling ethanol. After 2 h 0.32g (5.85 mmol) NaOMe was added and heating continued until no more ethanol distilled. The solution was allowed to cool, the anisole was filtered off, and the residue was recrystallized in portions from water to yield 7.31g (62 %) fine, white-to-tan crystals. H1NMR (400 MHz, CDCl3 ):  = 7.89 (s, 1H).

1-methyl-4,5-dicyanoimidazole8: 0.63g (5.29 mmol) 4,5-dicyanoimidazole was put into a 50 mL round bottomed flask with 8 mL distilled H2O. 0.76g NaHCO3 in 1.4 mL H2O was carefully added in portions. The flask was placed in a 65 oC bath and 1.8 mL (7.93 mmol) dimethyl sulfate was added dropwise over 1 h with stirring. Heating continued another h, then the solution was cooled in a 125 mL seperatory funnel. The mixture was extracted into ethyl acetate (4 x 10 mL) and washed with 0.5M NaHCO3 (3 x 10 mL). The combined organic layer was dried (MgSO4) and evaporated to give 0.52g (74%) yellow-brown solid. H1NMR (400 MHz, CDCl3 ):  = 3.90 (s, 3H); 7.69 (s, 1H).

1-methyl-4,5-imidazoledicarboxylic acid8: 3.76g (28.43 mmol) 1-methyl-4,5-dicyanoimidazole was put into a 250 mL round bottomed flask. 70 mL 6N NaOH was added with stirring, and the solution was heated to reflux for 2 h. The hot solution was carefully acidified with concentrated HCl to pH = 2, allowed to cool, and filtered. The solid was dried overnight at 80oC to afford 4.83 (100%) off-white powder. H1NMR (400 MHz, D2O):  = 3.98 (s, 3H); 7.96 (s, 1H).

1-methyl-4-imidazolecarboxylic acid8: 1.10g (6.46 mmol) 1-methyl-4,5-imidazoledicarboxylic acid was put into a flame-dried 250 mL round bottomed flask with 30 mL dimethylacetamide and heated to a bath temperature of 180oC with a reflux condenser for 4 h. The dimethylacetamide was distilled out under 0.1 mm Hg pressure, then vacuum evaporated to remove remaining traces. The solid was triturated with 20 mL benzene, then recrystallized from ethanol to give 0.57g (70%) white powder. H1NMR (400 MHz, D2O):  = 3.94 (s, 3H); 7.73 (s, 1H); 8.60 (s, 1H).

methyl-1-methyl-4-imidazolecarboxylate8: 10 mL anhydrous methanol was cooled to 0oC in a 50 mL flame-dried teardrop flask and 1 mL acetyl chloride was carefully added with stirring under argon. The mixture was stirred 10 min and warmed to rt. 0.2451g (1.94 mmol) 1-methyl-4-imidazolecarboxylic acid was added with stirring and stirred at rt 10 min, then heated to reflux for 48 h. Solid NaHCO3 was added until pH neutral, and the methanol was evaporated, removing last traces at 0.1 mm Hg. The residue was partially dissolved in 20 mL chloroform and fully dissolved upon addition of 20 mL water. Organic phase was extracted (3 x 20 mL) in chloroform, with small amounts of ethyl acetate and brine solution to break up emulsions. Combined organic phase was washed with 0.1 M NaOH (15 mL) and dried (MgSO4) and evaporated to afford 0.05g (18 %) white crystals. H1NMR (400 MHz, CDCl3 ):  = 3.74 (s, 3H); 3.89 (s, 3H); 7.46 (s, 1H); 7.58 (s, 1H).

methyl-1-methyl-2,5-dibromo-4-imiadazolecarboxylate: 0.05g (0.35 mmol) methyl-1-methyl-4-imidazolecarboxylate was put into a 50 mL flame-dried teardrop flask. 0.16g (0.87 mmol) N-bromosuccinimide was added with 0.01g (0.03 mmol) AIBN, 6 mL CCl4 and 3 mL chloroform and heated to bath temperature of 65oC with a reflux condenser. The orange solution was stirred 20 h and the solvent was evaporated. The crude was purified by column chromatography ( 1: 19 diethyl ether: dichloromethane) and recrystallized from heptane to afford 0.09g (9 %) yellow-white solid. H1NMR (400 MHz, CDCl3 ):  = 3.71 (s, 3H); 3.88 (s, 3H).

References

1. B. Huang, J. Parquette. Organic Letters2000, 2, 239
2. A. Sudhakar, T. Katz, B. Yang. J. Am. Chem. Soc.1986, 108, 2790
3. D. Appella, L. Christianson, I. Karle, D. Powell, S. Gellman. J. Am. Chem. Soc.1996, 118 (51), 13071
4. R. Günther, H. Hofmann, J. Am. Chem. Soc. 2001, 123 (2), 254
5. S. Martina, V. Enkelmann, G. Wegner, A. Schlüter. Synthesis, 1991, 8, 613
6. K. Tatsuka, M. Takahashi, N. Tanaka. Tet. Lett. 1999, 40, 1929
7. B. Palmer, W. Denny. J. Chem. Soc. Perkin Trans. 1989, 1, 95
8. J. O’Connell, J. Parquette, W. Yelle, W. Wang, H. Rapoport. Synthesis 1988, 10, 767

9. L. Grehn, U. Ragnarsson, Angew. Chem. Int. Ed. Eng.1984, 23 (4), 296