Chiral and SHG-Active Metal–Organic Frameworks Formed in Solution and on Surfaces: Uniformity, Morphology Control, Oriented Growth, and Postassembly Functionalization

We demonstrate the formation of uniform and oriented metal–organic frameworks using a combination of anion effects and surface chemistry. Subtle but significant morphological changes result from the nature of the coordinative counteranion of the following metal salts: NiX2 with X = Br–, Cl–, NO3–, and OAc–. Crystals could be obtained in solution or by template surface growth. The latter results in truncated crystals that resemble a half structure of the solution-grown ones. The oriented surface-bound metal–organic frameworks (sMOFs) are obtained via a one-step solvothermal approach rather than in a layer-by-layer approach. The MOFs are grown on Si/SiOx substrates modified with an organic monolayer or on glass substrates covered with a transparent conductive oxide (TCO). Regardless of the different morphologies, the crystallographic packing is nearly identical and is not affected by the type of anion or by solution versus the surface chemistry. A propeller-type arrangement of the nonchiral ligands around the metal center affords a chiral structure with two geometrically different helical channels in a 2:1 ratio with the same handedness. To demonstrate the accessibility and porosity of the macroscopically oriented channels, a chromophore (resorufin sodium salt) was successfully embedded into the channels of the crystals by diffusion from solution, resulting in fluorescent crystals. These “colored” crystals displayed polarized emission (red) with a high polarization ratio because of the alignment of these dyes imposed by the crystallographic structure. A second-harmonic generation (SHG) study revealed Kleinman symmetry-forbidden nonlinear optical properties. These surface-bound and oriented SHG-active MOFs have the potential for use as single nonlinear optical (NLO) devices.


S3
kV and 200 mA and with a scintillation detector aligned at the diffracted beam after a bent Graphite monochromater. Next, 2θ/θ scans were performed under specular conditions in the Bragg-Brentano mode with variable slits and scanned from 1 to 35 degrees of 2θ with step sizes of 0.02 degrees and a scan speed of 0.5 degree/min. Samples of MOFs (obtained in solution) were scanned in reflection geometry using an Ultima III (Rigaku, Japan) diffractometer equipped with a sealed Cu anode Xray tube operating at 40 kV and 40 mA. A bent graphite monochromator and a scintillation detector were aligned in the diffracted beam. Then θ/2θ scans were performed under specular conditions in the Bragg-Brentano mode with variable slits. The 2θ scanning range was from 2 to 60 degrees using a step size of 0.025 degrees and a scan speed of 0.5 degree/min.

Characterization of surface-grown MOFs (sMOFs) by electron diffraction
Scanning electron diffraction data were obtained in a double aberration-corrected Themis Z microscope (Thermo Fisher Scientific Electron Microscopy Solutions, Hillsboro, USA) equipped with a high-brightness FEG at an acceleration voltage of 200 kV. For the diffraction recording, an electron probe with a convergence angle of 0.2 mrad was adjusted in scanning transmission electron microscopy (STEM) microprobe mode and further defocused by typically 5-10 µm to reduce the probe size and electron flux. A primary beam current of less than 10 pA was used. The electron microscope pixel array detector (EMPAD) S4 enabled a rapid data collection of the entire unsaturated diffraction pattern with a pixel dwell time of 1 ms for each pattern. Typically, diffraction patterns were acquired over a raster of 128 pixels × 128 pixels.

Atomic Force Microscopy (AFM)
The AFM measurements were made on a Bruker MMAFM system, with Nanoscope V control electronics. The PF-QNM mode was used, which allows obtaining topographical as well as nanomechanical maps. An AC160 probe (Olympus) was used with a calibrated spring constant and a tip radius of 35 N/m and 20 nm, respectively. Images were processed and analyzed using Gwyddion open source software (http://gwyddion.net/).

Fluorescent Microscopy
Confocal imaging was conducted by using an upright Leica TCS SP8, equipped with internal Hybrid (HyD) detectors and an Acusto Optical Tunable Filter (Leica microsystems CMS GmbH, Germany). Dye excitation was performed using a laser (l = 561 nm) at 8.0% power. The emission signal was collected using an internal HyD detector at a range of l = 571-649 nm. Images were acquired using the galvometric scanner with a 20× air objective (HC PL APO 20X/0.75 CS2) S4 providing images with a field of view of 123.1 mm and a pixel size of 0.152 mm. Z-stacks were acquired using the galvo stage, with 0.684 µm intervals. The acquired images were visualized using ImageJ.

Second-Harmonic Generation (SHG) Microscopy
SHG imaging was performed on 2PM; a Zeiss LSM 510 META NLO equipped with a broadband Mai Tai-HP-femtosecond single box tunable Ti-sapphire oscillator, with automated broadband wavelength tuning l = 700-1020 nm from Spectraphysics. Images were acquired using a laser wavelength of 940 nm and detection through a filter of l = 415-637 nm. All SHG microscopy images were false colored.

Single-Crystal SHG Microscopy Set-up
For single-particle SHG measurements, 100fs pulses at l = 1000 nm and at an 80 MHz repetition rate from a Ti:Sapphire laser (Coherent, Chameleon Ultra II), were passed through a half wave plate, coupled into an inverted microscope (Zeiss, Axiovert 200 inverted microscope) and focused using an objective (Zeiss, Plan-Apochromat 10×/0.45NA). The epi-detected signal was filtered using a dichroic mirror (Thorlabs, DMSP950R), a band pass filter (Semrock, FF01-500/24-25), and a color glass (Thorlabs, FGS900-KG3). A polarizer (analyzer) was placed right before a multimode fiber (Thorlabs, FG050LGA). The SHG signal was detected by a single-photon avalanche photodiode (ID Quantique, ID100) that was connected to a time-correlated single-photon counting (TCSPC) system (Picoquant HydraHarp 400). The laser trigger output was connected to the TSCPS for synchronization. Samples were drop cast on a #1 cover slip and left to dry. In all experiments the laser power was ~200 mW. For simplicity, single crystals were chosen, based on their orientation in the lab frame of reference such that their major axis was parallel to either the x or y coordinate in the lab. Mapping was done by scanning the sample position using a piezo stage (Mad City Labs, Nano-BioS100) and measuring the SHG intensity while the excitation and collection polarizations were kept fixed. Polarimetry was performed by measuring the SHG intensity from a single point on the microcrystal as a function of the half wave plate angle (changing the excitation polarization) for two analyzer positions (the x and y polarizations in the lab frame). In both the mapping and the polarimetry measurements the time-correlated single photon counting (TCSPC) was operated in histogram mode.

Single-crystal X-ray crystallography
All crystals were coated in Paratone oil (Hampton Research) and mounted on MiTeGen loops. They were flash frozen in the liquid nitrogen stream of the Oxford Cryostream. Diffraction data of MOF-NiBr2 (CCDC numbers: 1965786 and 1965784) and MOF-Ni(OAc)2 (CCDC numbers 1965788 and 1965787) were measured at a low temperature of 100(2) K using Cu Kα λ = 1.54184 Å on a Rigaku XtaLab Pro diffractometer equipped with a Dectris Pilatus 3R 200K-A detector or λ = 0.70 Å at the ESRF ID29 beamline with a Dectris Pilatus 6M detector. For ligand Ad-DB, the diffraction data were collected at 100K on a Bruker KappaApexII CCD system using MoKα λ = 0.71073 Å.
Data were processed and reduced using the Bruker Apex3 Suite of programs. The Rigaku data were  Table S1.

Concentration effects
The procedure for preparing MOF-NiBr2 under different concentrations is similar to that described above. The ligand-to-metal salt ratio was kept constant (mole ratio 1:2). The ligand (Ad-DB) concentration ranged from 0.47-5.64 µmol/mL ( Figure S1).

Preparation of the Surface-Bound MOFs (sMOFs)
Ad-DB (18.0 mg, 0.021 mmol) is dissolved in DMF (6.0 mL). NiBr2 (8.2 mg, 0.038 mmol) is dissolved in DMF (4.0 mL). The silicon or quartz substrates, functionalized with a TPEB-based monolayer (Scheme 1) or ITO/glass, S3 are placed inside a glass pressure tube with the monolayer or the metal-oxide layer facing the bottom. The solution of Ad-DB (2.0 mL), 1.5 mL DMF, and 0.5 mL CHCl3 is added to the pressure tube and sonicated for 1 min. A solution of NiBr2 (DMF, 1.0 mL) is added and the reaction mixture is sonicated for 1 min. The sealed pressure tube is transferred to an oven and kept at 105 o C for 48 h. Next, the thermostat of the oven is lowered 10 o C degrees every hour until reaching 25 o C. Finally, the substrate is washed with DMF, CHCl3, and dried in the air. Identical procedures for the preparation of MOFs with Ni(NO3)2 and Ni(OAc)2 are used.

Detachment of Surface-Bound MOFs (sMOFs)
A silicon substrate (1 cm × 1 cm) with surface-bound MOF-NiBr2 is placed into a vial with ethanol (1.0 mL) and sonicated for 20 min. The mixture is transferred to a centrifuge tube (4 mL) and the detached MOFs are separated from the reaction mixture by centrifugation (4000 rpm, 5 min). The ethanol is decanted, and the MOFs are washed with DMF, CHCl3, and EtOH.

Calculation of Second Harmonics
The polarization of light interacting with MOF-NiBr2 was calculated and plotted in MATLAB. In order to calculate the induced second harmonic generation polarization, the second order nonlinear susceptibility tensor of the MOF crystal is multiplied by the polarization vectors of the incoming field. ( Since the MOF has a symmetry of 622, the ̅ ̅ !"" matrix has only 2 components that are connected to each other, such as $% = − #& S11 (note that both components would vanish if the Kleinmann symmetry is valid, i.e., for far-detuned excitation); these components were assigned arbitrary units and marked as 1 and -1, respectively. Because the crystal in the microscope is not perfectly aligned with the laboratory axes, rotation matrices have been introduced. We used the Euler rotation convention of Z-Y'-Z'', where the first angle of rotation around the Z axis determines the angle of the crystal within the plane. The second rotation angle around the Y axis determines the out-ofplane angle between the surface and the crystal. The third angle defines rotation along the crystal main axis, which, due to the symmetry of the crystal (622), does not affect the second-harmonic polarization. The incident light in the lab axis was rotated to the crystal frame: Equation (1) was then used to obtain the second-harmonic polarization, first in the frame of the crystal: & ′(2 ) = 2 ̅ ̅ !"" ′ = ( ) # And then rotated back to the lab axis: In MATLAB, the light amplitude was arbitrarily defined to be 1. 360 angles between 0 and 279 degrees; it was used for the various angles of polarization. The calculations mentioned above were done for all the polarization angles of incident light and for all rotation angles for the first Z rotation and the second Y' rotation. The result was plotted in MATLAB's app designer on polar plots, with the intensity on the X axis, and on the Y axis separately, and with the ability to choose different rotation angles for the crystal orientation.

Scattering Configuration
For measured HRS with a constant wavelength and intensity: (2) where < # > is the rotational average for the second-harmonic intensity, ( ) # is the incident excitation field, and N is the number density of the particles in the solution. S12 Since a similar ( ) was used for measuring both the MOF and the ZnO (Fig. S21), it is possible to obtain < # > -./ from the ratio of the measured signals by: -./ To calculate < # > )0. we used the known ̿ !"" matrix for ZnO (of 6 mm symmetry). S13 The experiment was done with incident light polarized only in the Y direction; therefore, we can derive the excitation field.
where Rot is the rotation matrix for the Z-X'-Z'' convention.