Electrostatic Potential Field Effects on Amine Macrocyclizations within Yoctoliter Spaces: Supramolecular Electron Withdrawing/Donating Groups

The central role of Coulombic interactions in enzyme catalysis has inspired multiple approaches to sculpting electrostatic potential fields (EPFs) for controlling chemical reactivity, including ion gradients in water microdroplets, the tips of STMs, and precisely engineered crystals. These are powerful tools because EPFs can affect all reactions, even those whose mechanisms do not involve formal charges. For some time now, supramolecular chemists have become increasingly proficient in using encapsulation to control stoichiometric and catalytic reactions. However, the field has not taken advantage of the broad range of nanocontainers available to systematically explore how EPFs can affect reactions within their inner-spaces. With that idea in mind, previously, we reported on how positively and negatively charged supramolecular capsules can modulate the acidity and reactivity of thiol guests bound within their inner, yoctoliter spaces (Cai, X.; Kataria, R.; Gibb, B. C. J. Am. Chem. Soc. 2020, 142, 8291–8298; Wang, K.; Cai, X.; Yao, W.; Tang, D.; Kataria, R.; Ashbaugh, H. S.; Byers, L. D.; Gibb, B. C. J. Am. Chem. Soc.2019, 141, 6740–6747). Building on this, we report here on the cyclization of 14-bromotetradecan-1-amine inside these yoctoliter containers. We examine the rate and activation thermodynamics of cyclization (Eyring analysis), both in the absence and presence of exogenous salts whose complementary ion can bind to the outside of the capsule and hence attenuate its EPF. We find the cyclization rates and activation thermodynamics in the two capsules to be similar, but that for either capsule attenuation of the EPF slows the reaction down considerably. We conclude the capsules behave in a manner akin to covalently attached electron donating/withdrawing groups in a substrate, with each capsule enforcing their own deviations from the idealized SN2 mechanism by moving electron density and charge in the activated complex and TS, and that the idealized SN2 mechanism inside the theoretical neutral host is relatively difficult because of the lack of solvation of the TS.


3) Encapsulation of 3 within hosts 12 and 22
1 mM D2O solutions of hosts 1 and 2 were prepared in the presence of 10 equivalents of NaOD. Subsequently, the bromide salt of guest 3 (14-bromotetradecan-1-ammonium bromide, 3·Br) was added and the complex formed by vortexing for 10 minutes at 25 ˚C. Figure S5-S12 show the 1 H NMR, DOSY NMR, and COSY NMR spectra, as well as the calculated Δδ values for each set of protons of the guest, for both the complex with host 1 and 2. This data confirmed that both complexes possessed a 2:1 stoichiometry.   Figure S1: 1 H-1 H COSY NMR spectrum of the bound guest region and peak assignment of complex 3@12. Figure S7: 1 H-1 H COSY NMR spectrum of the bound guest region and peak assignment of complex 3@12.    Dd (ppm) S12

4) Eyring analysis for the cyclization of 3 within 12 and 22
In order to compare the kinetics of macrocyclization inside hosts 1 and 2, i.e., we determined the logarithm of the reaction rate constant (lnk) as a function of temperature. The Eyring equation (Eq. 1) relates ln(k/T) to the reciprocal of the temperature (1/T): Where k is the reaction rate constant at temperature T (Kelvin), R is the gas constant (8.314 J·K -1 ·mol -1 ), ΔH ‡ is the activation enthalpy (kJ. mol -1 ) and ΔS ‡ is the activation entropy (J. mol -1 . K -1 ), KB is the Boltzmann constant (1.380649 ×10 -23 J.K -1 ) and h is the Planck constant (6.63×10 −34 J⋅s). The experiments were performed between 339 K and 351 K for host 1, and between 325 K to 338 K for host 2.
The individual rate constants at each temperature were themselves determined using a first order kinetic model (Eq. S2): Where y = the percentage conversion of the reactants, A1 = amplitude fitting constant, t1 = lifetime, x = time, y0 = y-offset at t = 0), and the first-order rate constant k is the inverse of the determined lifetime t1. Thus, monitoring the appearance of the new 1 H NMR signal from the Hd proton ( Figure  S1) in the spectrum of both hosts gave the extent of reaction as a function of time. Individual rate constants were determined at least in duplicate, with an obtained error of <10%. Figures S13 -S24 show: 1) the stacked 1 H NMR spectroscopy of each reaction as a function of time at different temperatures; 2) the change in integration of the Hd signal and the fit of this data using Eq. S2 (all experiments were duplicated or triplicated, with the average or standard deviation shown as error bars), and 3) the Eyring plots for the cyclization processes within 12 and 22. Table 1 shows the obtained Eyring data for cyclization. The error in the gradient was calculated (Excel, LINEST function) to be 5%.

5) Cyclization of 3 within 12 in the presence of various salts
Previous studies have shown that anions can bind to both the pendent groups or feet of host 1, and the non-polar pocket. 7 Additionally, studies showed that the electrostatic potential field (EPF) of either host had a significant effect on reactions carried out in the inner-space of the capsule, and that this EPF was modulated by counter-ion binding to the outside of the host. We therefore investigated the cyclization rate of 3 in the presence of anions known to bind to 1. To study this, the effect of nine different anions was investigated:

S29
The extent of reaction of each experiment was determined by tracking and integrating the 1 H NMR signal from the Hd proton (Mnova) and fitting this data as a function of time to give the rate constant data (Origin Pro 2018. Figure S39 shows the cumulative data for all the salts. Each experiment was carried out in duplicate (error bars shown), with the average calculated used to give the rate constant (Table 2). Errors were <10%.

5.2) Anion effect: Individual data
The reaction rate in each experiment was determined by tracking and integrating the "d" peak in the 1 H NMR spectroscopy using Mnova, then fitted the integrations as a function of time to give the kinetic data with Origin Pro 2018. Lines fitted by the first order kinetic model are demonstrated in red and the standard deviations are shown as error bars. Each experiment was carried out in duplicates, with the average calculated to give the corresponding rate constants.

6) Cyclization of 3 within 22 in the presence of various salts
We also studied how the presence of cations impacted the cyclization rate of 3 within the capsule formed by 2. Six cations were selected: Na + , Li + , K + , Rb + , Cs + and tetramethylammonium (TMA + ). Figures S60 -S76 show the corresponding data for reactions carried out with the same protocol but in the presence of 10 mM of the cations (as their chloride salts). The experiments were performed with 0.5 mL samples of 1.0 mM host 2 in D2O at 336 K. In each case, a small volume of 400 mM salt solution was added to make a 10 mM salt solution of the complex. Figures S60 and S61 show the 1 H NMR spectra of the complex with 22 in the presence of 10 mM salt.

S44
The extent of reaction of each experiment was determined by tracking and integrating the 1 H NMR signal from the Hd proton (Mnova), and fitting this data as a function of time to give the rate constant data (Origin Pro 2018. Figure S39 shows the cumulative data for all the salts. Each experiment was carried out in duplicate (error bars shown), with the average calculated used to give the rate constant (Table 3). Errors were <10%.

6.2) Cation effect: Individual data
The reaction rate in each experiment was determined by tracking and integrating the "d" peak in the 1 H NMR spectroscopy using Mnova, then fitted the integrations as a function of time to give the kinetic data with Origin Pro 2018. Lines fitted by the first order kinetic model are demonstrated in red and the standard deviations are shown as error bars. Each experiment was carried out in duplicates, with the average calculated to give the corresponding rate constants. Errors were <10%.             The reaction rate in each experiment was determined by tracking and integrating the "d" peak in the 1 H NMR spectroscopy using Mnova, then fitted the integrations as a function of time to give the kinetic data with Origin Pro 2018. Lines fitted by the first order kinetic model are demonstrated in red and the standard deviations are shown as error bars. Each experiment was carried out in duplicates, with the average calculated to give the corresponding rate constants. Errors in individual rate constant were <10%. Gradient error in the Eyring analysis was 5% (Excel LINEST function). Figures S77 -S88 show the data for the cyclization of 3 in 12 in the presence of NaClO4, whilst Figures S89 -S100 show the data for the cyclization of 3 in 22 in the presence of CsCl.     Figure S82: Reaction kinetics for guest 3 inside of host 12 in presence of 10 mM NaClO4 at 345 K.   Figure S84: Reaction kinetics for guest 3 inside of host 12 in presence of 10 mM NaClO4 at 348 K.   Figure S86: Reaction kinetics for guest 3 inside of host 12 in presence of 10 mM NaClO4 at 351 K.