Weak functional group interactions revealed through metal-free active template rotaxane synthesis

Modest functional group interactions can play important roles in molecular recognition, catalysis and self-assembly. However, weakly associated binding motifs are often difficult to characterize. Here, we report on the metal-free active template synthesis of [2]rotaxanes in one step, up to 95% yield and >100:1 rotaxane:axle selectivity, from primary amines, crown ethers and a range of C=O, C=S, S(=O)2 and P=O electrophiles. In addition to being a simple and effective route to a broad range of rotaxanes, the strategy enables 1:1 interactions of crown ethers with various functional groups to be characterized in solution and the solid state, several of which are too weak — or are disfavored compared to other binding modes — to be observed in typical host–guest complexes. The approach may be broadly applicable to the kinetic stabilization and characterization of other weak functional group interactions.

T he bulky axle end-groups of rotaxanes mechanically lock rings onto threads, preventing the dissociation of the components even if the interactions between them are not strong and attractive [1][2][3] . In principle the enforced high local concentration of convergent functional groups brought about by such mechanical bonding can stabilize weak non-covalent interactions 4 . In practice such outcomes are rarely observed [4][5][6][7][8] because most rotaxane syntheses rely upon strong attractive interactions between the building blocks 2,3,[9][10][11][12][13] to promote the rotaxane assembly process. Strong binding modes generally 'live on' in the interlocked product, an outcome useful for the design of artificial molecular machinery 2,3,14-16 , whether intended to operate in solution 17 or when organized on surfaces 18,19 or within metal-organic frameworks 20,21 , but one that tends to override alternative weaker binding modes that could occur between the components. It is sometimes possible to remove strong template interactions by post-assembly modification, for example by deprotonation of an ammonium unit 22,23 , but this is often not straightforward and can require forcing conditions 23 .
Active template synthesis [24][25][26][27][28][29][30][31][32][33][34][35] , in which a macrocycle accelerates a strand-forming reaction through the ring cavity, does not require strong pre-association of the starting materials. Although most active template syntheses have been developed from transition metal catalyzed reactions [24][25][26][27][28][29][30][31][32][33][34][35] , a metal-free active template system was recently discovered 36,37 in which the addition of primary amines to electrophiles can be significantly accelerated through crown ethers 37 and related macrocycles 36 by stabilization of the reaction transition state [38][39][40][41][42][43] . The reaction of a primary amine and an electrophile in the presence of a crown ether was found 37 to form [2]rotaxanes by metal-free active template Nalkylation, aza-Michael addition or N-acylation. In these reactions the crown ether stabilizes developing partial charges in the transition state causing initial rate accelerations of up to 26× through the macrocycle compared with the reaction exo-to the cavity that forms the non-interlocked axle. The N-acylation reaction is particularly effective: simply mixing together 1.0 equivalents of each of 24-crown-8 1, amine 2, and activated ester 3 in toluene at room temperature spontaneously assembles amide-axle [2]rotaxane 4 in 56% yield, without the need for any other reagents or excess building blocks (Fig. 1). This potentially offers access to kinetically locked systems with unusual combinations of functional groups on the different components forced into close proximity and a 1:1 stoichiometry. The interaction of the groups on different components might further be enhanced by the tendency of interlocked architectures to have poorly solvated inner surfaces.
To explore the scope of this unexpected method of rotaxane synthesis, here we carry out a study of the reaction with a series of related electrophiles. After developing an optimized set of reaction conditions, rotaxanes were accessed by crown etherstabilized formation of (thio)urea, carbamate, sulfonamide, and phosphoramidate/phosphinamide-containing axles. The stabilization of S N Ar reactions between primary amines and electrondeficient aryl halides led to rotaxanes with aniline threads. Singlecrystal X-ray diffraction of the rotaxanes enabled weak interactions between the crown ether and the newly formed functional groups in the axles to be studied.

Results
Optimization of metal-free active template rotaxane synthesis by N-acylation. Complete consumption of crown ether 1 does not occur with the experimental protocol originally used for the active template N-acylation rotaxane-forming reaction (Fig. 1), even with a fivefold excess of amine 2. Proton nuclear magnetic resonance ( 1 H NMR) showed that rotaxane 4 is initially formed rapidly, but over time its rate of formation slows relative to the background reaction of amine and ester, resulting in increasing amounts of non-interlocked axle. The color change that occurs during the early stages of that reaction suggested that liberation of the yellow 4-nitrophenolate anion 44 might be inhibiting the formation of rotaxane 4. We reasoned that 4-nitrophenol, formally the other product of the N-acylation reaction, would be deprotonated by 2 and the resulting primary ammonium cation (2H + ) would bind strongly to the crown ether preventing it from participating in the active template reaction. Accordingly, we investigated whether the yield of 4 could be improved by the addition of tertiary amines, which when protonated bind more weakly to crown ethers than primary ammonium salts 45 (Supplementary Table 1). Pleasingly, addition of 10 equivalents (equiv.) of triethylamine (Et 3 N) led to the formation of rotaxane 4 in 68% yield after 1 h and 92% yield after 24 h. Under these conditions the ratio of rotaxane 4 to non-interlocked axle improved from 8:1 to 17:1 after 24 h, indicating that Et 3 N does not promote aminolysis of the building blocks in the absence of the crown ether. In contrast, the use of a stronger base, 1,8diazabicyclo [5.4.0]undec-7-ene (DBU), significantly reduced the formation of 4 (10% yield after 1 h) while increasing the amount of non-interlocked axle formed, suggesting that DBU accelerates the reaction of 2 and 3 at the expense of the active template reaction [46][47][48] .  We next investigated the efficacy of rotaxane formation with less nucleophilic benzylic amines (Supplementary Table 2). Commercially available amine 5, bearing two CF 3 substituents, proved the most effective amine tested, with [2]rotaxane 6 formed in 84% yield after 24 h without the need for Et 3 N (Fig. 1), with a rotaxane:non-interlocked axle ratio >100:1 (determined by 1 H NMR). This remarkable selectivity for acylation through the cavity appears to be a consequence of the background acylation reaction (to form the non-interlocked axle) having an activation energy in the 'sweet spot' for active template synthesis: too high for acylation to occur quickly with the less nucleophilic amine (5) but low enough that a few kcal mol −1 stabilization of the transition state by the crown ether brings about a very significant rate enhancement.
It also proved possible to use more reactive electrophiles with amine 5 (Supplementary Table 3). Rotaxane 6 was obtained in 54% yield from the corresponding acid chloride and in 40% yield when using the 1-hydroxybenzotriazole ester as the electrophile.
C=O/C=S/SO 2 /P=O electrophile scope. With improved conditions for active template ester aminolysis in hand we investigated whether the type of rotaxanes accessible could be expanded upon using electrophiles based on different, but structurally related, chemical functionality (7-14, Table 1). The aminolysis of carbamates 49 follows a similar mechanistic pathway to ester aminolysis: nucleophilic attack at the carbonyl forms a tetrahedral intermediate followed by loss of the leaving group to form urea [50][51][52] . Accordingly we tested whether carbamate 7 was a suitable electrophile for the metal-free active template reaction. Reaction of 7, amine 5 and 24-crown-8 1 in a 1:1:1 ratio, under the standard reaction conditions (without Et 3 N), afforded urea [2]rotaxane 15 in 73% yield ( Table 1, entry 1). Urea rotaxane formation was also possible without generating a leaving group byproduct through the use of isocyanate 8, which gave rotaxane 16 in 55% yield (Table 1, entry 2). The reaction between 5 and 8 proceeded extremely quickly; full conversion of 5 was achieved within 1 min. The corresponding thiourea rotaxane 17 was prepared in an analogous manner from isothiocyanate 9 in 54% yield ( Table 1, entry 3).
Carbamate rotaxanes were accessible using common commercially available electrophiles. Activated carbonate 10, used to form carbamates that can be readily decomposed with fluoride 53 , gave 18 in 83% yield (Table 1, entry 4), while chloroformate 11 (Fmoc-Cl) generated 19 in 70% yield with Et 3 N added to neutralize the HCl product (Table 1, entry 5) 54 . The ability to release the macrocycle from these types of rotaxanes in response to a specific chemical stimulus (stoichiometric fluoride for 18; catalytic base for 19) may prove useful for future applications.
Electrophiles containing a heteroatom at the site of nucleophilic attack also proved effective for rotaxane formation. Sulfonyl chloride 12, a bulky analog of tosyl chloride, reacted with 5 and 1 to give sulfonamide rotaxane 20 in 95% yield ( Metal-free active template synthesis by N-arylation. To further expand on the general applicability of active template rotaxane synthesis with crown ethers, we explored other potential reaction modes. Prompted by a recent report 55 of crown ether catalysis of   S N Ar reactions between aryl halides and primary amines, we investigated rotaxane formation by N-arylation. This proved effective using different electrophiles: combining amine 5 and 24crown-8 1 with aryl fluoride 23 in the presence of Et 3 N produced aniline rotaxane 24 in 85% yield (Fig. 2a), while combining 5 and 1 with aryl chloride 25 formed aniline rotaxane 26 in 75% yield (Fig. 2b). Both rotaxanes were isolated as neutral amines rather than as the corresponding ammonium salts. As the pK a values of protonated anilines are readily modulated by changing the aromatic substitution 56 , rotaxanes such as 24 and 26 have the potential to be used as tunable pH-sensitive molecular switches of basicity lower than that of commonly used dibenzylammoniumcrown ether systems 3 .
Crown ether-functional group interactions. Complexes between crown ethers and neutral molecules were first reported by Pedersen nearly 50 years ago 57 , with the majority of examples described to date involving relatively small macrocycles such as 18-crown-6 (refs. [58][59][60][61]. Neutral molecules cannot be fully encapsulated within such small cavities and so the complexes tend to be of a 'perch' type. With such binding modes the crown ethers often bind to more than one guest to maximize favorable host-guest interactions and to balance the dipole moments of polar guests. In the solid state discrete 1:1 crown ether-neutral molecule complexes are rare and a range of different binding modes and ratios can sometimes be observed with only minor variations in structure arising from different crystallization conditions [59][60][61] . In contrast to such host-guest complexes, the interlocked components of rotaxanes have a strictly defined stoichiometry (usually 1:1), are held in close proximity, and possess limited coconformational 62 degrees of freedom [63][64][65][66][67][68][69][70][71] . In the absence of strong binding between the components weak interactions that are seldom observable in supramolecular complexes can form and significantly influence co-conformation 4

Discussion
Metal-free active template synthesis is a simple and versatile method through which to access crown ether rotaxanes with a discrete but diverse set of functionalities in the axle. The rotaxane assembly procedure is exceptionally simple, requiring only mixing of a crown ether, amine, and electrophile in toluene. All of the building blocks used in this paper are either currently commercially available or, in the case of amine 2, ester 3, carbamate 7, isothiocyanate 9, and sulfonyl chloride 12, accessible in a single synthetic step. The rotaxane-forming reactions can be performed using a 1:1:1 stoichiometry of the three building blocks, in some cases generating rotaxanes in yields as high as 95% with >100:1 rotaxane: axle selectivity. In addition to being a simple and effective route to a   84 : "…intramolecular hydrogen bonds are widespread in biological molecules and are crucial in the design of new drugs and materials, including supramolecular machines… Unfortunately, the characterization of intramolecular H-bonds (and their utilization) is still limited, most likely as a consequence of the complexities that hamper the interpretation of experimental data obtained for large and flexible entities." The ability to kinetically stabilize modest strength hydrogen bonding modes is another noteworthy consequence of the mechanical bond 3 .
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/.