Energy‐Efficient Iodine Uptake by a Molecular Host⋅Guest Crystal

Abstract Recently, porous organic crystals (POC) based on macrocycles have shown exceptional sorption and separation properties. Yet, the impact of guest presence inside a macrocycle prior to adsorption has not been studied. Here we show that the inclusion of trimethoxybenzyl‐azaphosphatrane in the macrocycle cucurbit[8]uril (CB[8]) affords molecular porous host⋅guest crystals (PHGC‐1) with radically new properties. Unactivated hydrated PHGC‐1 adsorbed iodine spontaneously and selectively at room temperature and atmospheric pressure. The absence of (i) heat for material synthesis, (ii) moisture sensitivity, and (iii) energy‐intensive steps for pore activation are attractive attributes for decreasing the energy costs. 1H NMR and DOSY were instrumental for monitoring the H2O/I2 exchange. PHGC‐1 crystals are non‐centrosymmetric and I2‐doped crystals showed markedly different second harmonic generation (SHG), which suggests that iodine doping could be used to modulate the non‐linear optical properties of porous organic crystals.


S2
-Experimental procedures 1/ Chemical compounds. Distilled water was used to grow single crystals and iodine, D2O and Nylon membrane filters (450 nm) were purchased from commercial sources. CB [8] was prepared according to a previous paper. [1] TMB-AZAP was prepared according to the literature. [2] 2/ NMR Measurements. NMR spectra were recorded on BRUKER Avance III nanobay -300 or 400 spectrometers ( 1 H-NMR frequencies 300.13, and 400.13 MHz) at 300 K using D2O as the solvent and a watergate sequence (water suppress) when necessary (potentially affecting signals and integrals near the signal suppressed). Acetone was also used when necessary, as a reference. Splitting patterns are indicated as follows: s, singlet; d, doublet; t, triplet; m, multiplet. The spectra of the water inside the crystal were recorded at 300 K on a BRUKER Avance III -600 MHZ spectrometer, using the standard one pulse sequence with 16 scans and a sum of the acquisition time and the relaxation delay equal to 3 seconds. NMR self-diffusion experiments were recorded at 300 K, by a conventional pulse sequence, based on the stimulated echo and incorporating bipolar gradient pulses and an Eddy current delay (BPP-LED). [3] The shape of all gradient pulses was Smoothed Square and the LED delay was of 5 ms. The diffusion delay was set at 100 ms, the gradient strength, g, was linearly incremented in 32 steps from 2% to 98% of its maximum value with a constant duration of 4 ms, and 32 scans. The sum of the acquisition time and the relaxation delay was equal to 3 seconds. The diffusion coefficients were obtained using the dynamic center available on TopSpin Software.

3/ Molecular modelling.
The molecular electrostatic potential (MEP) was depicted on van der Waals surface with the Jmol software. [4] A structure containing 24 CB [8]•TMB-AZAP 1+ complexes, portion of a channel, was constructed based on the single-crystal X-ray structure without water. The atomic charges were calculated with Ampac 11 [5] at the PM6 [6] level of theory.

4/ Single-crystal X-ray Diffraction.
Suitable crystals for CB [8]•TMB-AZAP and CB [8]•TMB-AZAP•I2 were measured on a Rigaku Oxford Diffraction SuperNova diffractometer at 295 K at the CuKα radiation (λ=1.54184 Å). Data collection reduction and multiscan ABSPACK correction were performed with CrysAlisPro (Rigaku Oxford Diffraction). Using Olex2, [7] the structures were solved with ShelXT [8] using intrinsic phasing and ShelXL [9] was used for full matrix least squares refinements. For CB [8]•TMB-AZAP 6.5 mostly disordered water molecules were determined experimentally and were refined with partial occupation factors when necessary. A mask of solvent accounting more or less for 4300 Å 3 per unit cell was applied in the final stages of the refinement. For CB [8]•TMB-AZAP•I2 5 iodine atoms were found in the asymmetric unit, extremely disordered over several sites and refined with partial occupation factors, as well as one full structural water molecule. Most of the H-atoms were found experimentally for CB [8]•TMB-AZAP except some of the azaphosphatrane and those of the water molecules that were introduced at geometrical positions. All H-atoms were introduced at geometrical positions for CB [8]•TMB-AZAP•I2. In both cases their coordinates and Uiso parameters were constraint to 1.2Ueq (parent atoms) for the CH and CH2 groups and to 1.5Ueq (parent atoms) for all the other groups.

5/ Thermogravimetric analysis. Thermogravimetric analysis was executed in a TGA 8000TM
(PerkinElmer) apparatus. The tests were conducted at a rate of 5°C/min from 30°C to 800°C under inert atmosphere (Nitrogen flow rate: 20 mL/min) followed by a rate of 20°C/min from 800°C to 1000°C under air and holding for 7 minutes at 1000°C.

7/ Solid-state UV-vis spectroscopy.
Diffuse reflectance measurements were performed in the 400-800 nm range, using a Varian 300 spectrophotometer equipped with an integrating sphere DRA-CA-30I. The crystals were placed in a Teflon sample holder. Several measurements were done to account for problems arising from a possible surface inhomogeneity and a mean value of the reflectance was used that was then converted in absorbance (A). ) and the emitted light was collected in the back reflected direction by the same objective. The SHG signal at the wavelength 400 nm was recorded with photomultiplier tubes (PMTs, Hamamatsu R9110), and filtered using 400/10nm band-pass filters. Scanning and data acquisition was performed using an in-house LabVIEW (National Instruments Corp.) program. The image acquisition was performed with a pixel dwell time of 20 s, over 100×100 pixels covering a field of view of about 50 to 100 m size. Typical average powers at the focal plane were about 10 to 50 mW. To perform the polarized SHG measurements, a polarizing beamsplitter (PBS252, Thorlabs Inc.) was placed after the scanning mirrors to ensure that the input polarization was linearly polarized when reflected on the incident reflection dichroic mirror along its p or s polarization direction. An achromatic half-wave plate HWP (AQWP10M-980, Thorlabs Inc.) was placed after this dichroic mirror, mounted on a motorized rotational mount (PR50CC, Newport Corp.) to rotate the incident linear polarization by an increasing angle  with respect to the X axis (horizontal axis in the sample image). We recorded the image for each polarization state over the range of 0° to 180° with an  angle step of 10°, chosen to provide a signal to noise of the total signal above 5 to ensure sufficient precision (~ a few degrees) on the retrieved parameters (anisotropy and orientation, see below). The emitted nonlinear signal was detected in the reflected direction without any analyzer in the detection path.
Polarized SHG data analysis. The quantification of the polarized SHG responses follows the principle derived in Reference 10. [10] Briefly, each pixel of an image is a recorded stack of images, from which we retrieve the modulation of the signal as a function of the input polarizations angle . At each pixel of the SHG image, the modulation can be expressed ( ) ∝ 1 + 2 cos 2( − 2 ) + 4 cos 4( − 4 ) . The anisotropy parameter 2 is retrieved by circular projection of the intensity modulation onto the function (cos 2 ). The fourth order parameter 4 is the signature of high order symmetry in the crystal. Both parameters depend strongly on both the symmetry of the crystal and its orientation in the measured sample. In this work we focused on 2 which is a signature of the onedimensional symmetry of the crystal projected in the sample plane. For each crystal region measured (typically 50x50 m 2 ) a map of 2 and intensity values is reported. Mean and standard deviation values were recorded over 6 crystal regions (taken on 4 different crystals) as plotted in Figure S8. The deviation visible from crystal to crystal is likely due to the different crystals off-plane orientations in the sample, nevertheless clear differences can be seen in the two crystals. In PHGC-1, 2 = 0.52 ± 0.07 and intensity values (in V) are of the order of 0.32 ± 0.07 , while in PHGC-1-I2 , 2 = 0.37 ± 0.05 and intensity values are of the order of 3.62 ± 3.2 . In comparison, values reported in a KTP crystal are 2 = 1.04 ± 0.13 and intensity (for an identical incident intensity) around 764 . At last, the angle 2 corresponding to the phase of the measured intensity modulation was also retrieved, which corresponds to the in-plane orientation of the anisotropy response of the crystal. In both PHGC-1 and PHGC-1-I2, this axis is seen to lie along the long axis of the crystals ( Figure S8).

9/ Preparation, NMR spectra and binding constant for the CB[8]•TMB-AZAP complex.
Preparation: to 0.63 mg of solid CB [8]  The exchange was slow with respect to the NMR timescale and unexpectedly, two equivalents of CB [8] were necessary to displace equilibria exclusively toward the CB[8]•TMB-AZAP complex leaving a fraction of insoluble CB [8].

S6
Evaluation of the binding constant for TMB-AZAP•CB [8] complex formation by competitive NMR.
The TMB-AZAP name was contracted to AZAP in the following for clarity. The AZAP•CB [8] binding constant was evaluated using a competitive 1 H NMR binding method, following a procedure of Macartney and co-workers. [11] NMR spectra were collected at 298 K on a Bruker AC400 (64 scans) from 1 mM solutions of AZAP•CB [8].
The competitor guest was dimethylviologen dichloride (DMV). The binding constant of DMV•CB [8] (1.1 × 10 5 M -1 ) has been reported in the literature. [12] The chemical shifts of free DMV and of the DMV•CB [8] complex were determined in D2O from 1 mM solutions and using acetone as internal reference (2.220 ppm) ( Table S1). The Δδlim value was determined from the subtraction of the chemical shift of free DMV and of DMV•CB [8] (to ensure the full inclusion of DMV in CB [8], a CB [8]:DMV ratio of 2:1 was applied). Since the signal of proton H3 of DMV presents the highest Δδlim, calculations of the binding constant were based using this characteristic signal. Then, 1 H NMR spectra of a mixture of 1 equiv of DMV (1 mM), 1 equiv of AZAP (1 mM) and 1 equiv of CB [8] (1 mM) were recorded. According to the method of Macartney et al, [11] the binding constant of AZAP was calculated from the chemical shifts of proton H3 of DMV•CB [8] in the competitive spectra ( Figures S3) and     CB [8] interacts with 4 other neighboring macrocycles, 1 guest inside its cavity and 2 guests outside and 6 water molecules (2 outside, and two with each CB [8] Figure S5. Comparison between the shape of channels [7] of the previously reported CB [8] hydrate structure (top line) and of the shape of channels of PHGC-1 (bottom line). Figure S6. TGA analysis of (hydrated or as-prepared) CB [8]•TMB-AZAP (PHGC-1) single crystals.

12/ Elemental analysis of PHGC-1 and PHGC-1•I2.
1CB  S12 Table S3. Selected parameters from the crystal structure of CB [8]•TMB-AZAP•I2. Figure S8. Evolution of the diffusion coefficient of water as a function of time.  in (e,f)). The sticks orientation relative to the horizontal direction correspond to the retrieved 2 angle, while the color of the sticks corresponds to the measured 2 at the pixel location.