Dynamic Covalent Self‐Assembly of Chloride‐ and Ion‐Pair‐Templated Cryptates

Abstract While supramolecular hosts capable of binding and transporting anions and ion pairs are now widely available, self‐assembled architectures are still rare, even though they offer an inherent mechanism for the release of the guest ion(s). In this work, we report the dynamic covalent self‐assembly of tripodal, urea‐based anion cryptates that are held together by two orthoester bridgeheads. These hosts exhibit affinity for anions such as Cl−, Br− or I− in the moderate range that is typically advantageous for applications in membrane transport. In unprecedented experiments, we were able to dissociate the Cs⋅Cl ion pair by simultaneously assembling suitably sized orthoester hosts around the Cs+ and the Cl− ion.


General Experimental Section
All commercially available chemicals were purchased from Sigma Aldrich, ABCR GmbH, TCI Deutschland GmbH, Acros Organics and Alfa Aesar and used without further purification. CDCl3 and DMSO-d6 were stored over molecular sieves (3 Å). Molecular sieves and aluminum oxide were dried at 150 °C under reduced pressure (10 -2 mbar). Anhydrous solvents were dried prior to use in a MBraun SPS-800 instrument. All solvents used in exchange reactions were stored over molecular sieves (3 Å) for at least 24 hours prior to usage. All orthoester exchange reactions were carried out under nitrogen atmosphere.
High resolution mass spectra were recorded on Bruker solariX (Hybrid 7T FT-ICR) or Agilent QTOF 6546 using electrospray ionization (ESI). Acetonitrile was used as solvent.
ATR-IR spectra were recorded on a Bruker Alpha II with an ATR platinum Diamond using 60 scans in a range of 400 to 4000 cm -1 .
The normal-phase flash column chromatography was performed using silica 60 with a particle size of 0.04 -0.063 mm from Macherey-Nagel.
Molecular dynamics (MD) simulations were carried out with Amber 20 software package using the GPU accelerated pmemd.cuda module. [1] The General Amber Force Field (GAFF) parameters were used to define bonded and non-bonded parameters of the cryptand. [2] The partial charges have been derived applying the restrained electrostatic potential (RESP) [3] point charge fitting procedure based on the gas phase geometry optimizations at the HF/6-31G(d) level of theory, followed by ESP charge generation from single point calculations with the HF/6-31G(d) method and RESP fitting using antechamber program of the Amber tools. [4] All cryptand structures have been solvated in periodic truncated octahedral boxes of chloroform or DMSO. [5] The final solvated systems contained around 3000 and 1500 molecules of chloroform or DMSO, respectively. The monovalent ion parameters used for simulations with encapsulated chloride ion are based on the non-bonded model parameterisations by Li et al. [6] Following a steepest descent and conjugate gradient minimisation for 10000 steps, all simulations were carried out for at least 1 µs using integration time step intervals of 2 fs and applying constant pressure (NPT) molecular dynamics at 1 atm and 300 K using a Langevin dynamics. All bonds including hydrogens were constrained applying SHAKE algorithm. Electrostatic long-range interactions were treated with the Particle Mesh Ewald (PME) method and a 12 Å cut-off for non-bonded interactions.

Synthesis and Characterization of Diol Building Blocks
Overview of the synthesis of diols (1)

Synthesis of (3-aminophenyl)methanol (S6)
3-Aminobenzoic acid (S5) (5.04 g, 36.8 mmol, 1 equiv.) was suspended in 30 mL anhydrous THF under argon and cooled to 0 °C. LiAlH4 solution (1 M in THF, 44 mL, 44.0 mmol, 1.2 equiv.) was added dropwise. The mixture was warmed up to room temperature and left stirring for 1 h. Afterwards, the mixture was heated to 60 °C for 3 h. After cooling to room temperature, it was quenched with 110 mL cooled water. The aqueous layer was extracted with chloroform and the combined organic layers were washed with brine and dried over MgSO4. Removal of the solvent under reduced pressure yielded the product as a beige solid (2.55 g, 20.7 mmol, 56%). The observed 13 C NMR data corresponds to the literature [8] .

Synthesis of 1,3-bis(3-(4-hydroxybut-1-yn-1-yl)phenyl)urea (S13)
Compound (S13) was synthesized according to the same procedure as compound (S12). After 18 h of stirring at room temperature, the product precipitated. The product was filtered off and the precipitate was washed with THF and Et2O. The crude product was recrystallized from THF to yield the pure product as a colorless solid (943 mg, 2.56 mmol, 40%).

Drying of starting materials
Diols (1, 2, 3, S9), CsCl and salts used for titration experiments (TPPCl, TPPBr, TPPI, TMANO3) were filled in screw threaded glass vials and dried at 80 °C under vacuum over phosphorus pentoxide for three days. Tetraphenylphosphonium chloride (TPPCl) used as template in self-assembly experiments was dissolved in anhydrous acetonitrile, and trimethyl orthoacetate (7 equiv.) and TFA (1.2 equiv.) were added. After stirring for 1 hour, the volatiles were removed under reduced pressure to obtain the dry salt. Triethylene glycol, trimethyl orthoacetate and trimethyl orthofomate were stored over 3 Å MS and aluminum oxide. 2,3,4,5,6-pentafluorobenzenethiol was used as purchased. The solvents were dried over 3 Å MS for at least 24 hours.

General procedure (D) for the synthesis of Cryptands (chloride removal)
About 30 mg of corresponding cryptate were loaded into a screw capped vial. 100 µL TEA (to prevent undesired hydrolysis) and 4 mL dry MeOH were added and the suspension was stirred for 7 hours. The solution was removed and the TEA/MeOH mixture was added a second time. After stirring overnight, the solution was removed and the product was dried in vacuum. The cryptands were obtained in quantitative yields.    Each attempted recording of a 13 C NMR spectrum showed too much of the hydrolysed compound. For related 13 C NMR data, please refer to Figure S10 (o-Me2-ur-C3).

Synthesis of o-Me2-ur-C2 o-Me2-2.2.2 (CsCl removal)
To a solution of about 6 mg of [Cl -⊂o-Me2-ur-C2][Cs + ⊂o-Me2-2.2.2] in 600 µL DMSO, 500 µL D2O was added. The resulting precipitate was isolated by centrifugation. 1 mL D2O was added and the suspension was ultrasonicated and subsequently centrifuged. This step was repeated 2 times. The product was dried in vacuum. A 133 Cs NMR spectrum was recorded and some residual Cs + could be detected. Further washing with D2O did not lead to complete removal of the Cs + signal.

Evidence for template effect I: Attempting synthesis of o-Me2-ur-C3 without PPh4Cl
General procedure (A) was followed. Without addition of TPPCl. Molecular sieves were renewed after 2.5 days.

Me2-2.2.2] in pure DMSO
General procedure (A) was followed. Pure DMSO-d6 was used as solvent.    The precipitate is not the clean product, due to co-precipitation of both residual diols (these have to be present in this type of a competition experiment). Due to overlapping (broad) signals in the 1 H NMR spectrum ( Figure S41), the ratio between both cryptates was determined using quantitative 13 C NMR spectroscopy using suppressed NOE coupling. The integration of such 13 C NMR spectra can be used for a rough estimation of compound ratios. A mixture of both cryptates in the ratio of 2:1 [Cl -⊂o-Me2-ur-C2]/[Cs + ⊂o-Me2-2.2.2] was observed (in the isolated reference compound, as expected, a 1:1 ratio was observed).

Using CsCl as a template: Synthesis of [Cl -⊂o-Me2-ur-C2]Cs +
General procedure (A) was followed. TPPCl was substituted by CsCl. The reaction progress was monitored by 1 H NMR spectroscopy. After one day, disappearance of all signals corresponding to -OMe groups indicated completion of reaction. After workup (precipitation using diethyl ether), the clean product [Cl -⊂o-Me2-ur-C2]Cs + was isolated in 42% yields.

Using CsCl as a template: Unsuccessful synthesis of [Cs + ⊂o-Me2-2.2.2]Cl -
General procedure (A) was followed. Triethylene glycol was used as diol and TPPCl was substituted by CsCl. To support cation binding to the glycol chains, the amount of chloroform was increased (55:1 CDCl3/DMSO-d6). Due to the low solubility of CsCl in chloroform, DMSO could not completely be omitted. Reaction control by 1 H NMR spectroscopy and HRMS provides evidence of cryptate formation, but the isolation of the cryptate was unsuccessful. Hence, only traces of cryptate were formed during the reaction.

General Procedure (E) for 1 H NMR titrations of cryptands with anions
Stock solutions of corresponding cryptand and salt in DMSO-d6 or in CDCl3/DMSO-d6 (5:1) mixture were prepared and the precise concentration of the solutions was determined by 1 H NMR spectroscopy using 1,4-dinitrobenzene or 1,3,5-trimethoxybenzene as internal standard. 600 µL of cryptand stock solution were loaded into an NMR tube with a screw cap containing a septum. 3 µL TEA were added to avoid hydrolysis. During the titration, varying amounts of the salt stock solution and additional cryptand stock solution to keep the concentration of the cryptand constant were added through the septum using Hamilton syringes. The titration was monitored by 1 H NMR spectroscopy (400 MHz, 298 K). The data were fitted using Bindfit [10,11] (Fit method: Nelder-Mead, Binding model: 1:1). One representative salt (TPPCl) was titrated in triplicate to obtain a more complete picture of experimental error(s).

High resolution ESI-MS/MS measurements
Collision energy is given as center-of-mass frame (Ecom) in dependence on masses of collision gas (N2) and precursor ion, and the measured laboratory collision energy (Elab) according to Eqn.