Proline Functionalization of the Mesoporous Metal-

Transcription

1 Supporting Information Proline Functionalization of the Mesoporous Metal- Organic Framework DUT-32 Christel Kutzscher, Herbert C. Hoffmann, Simon Krause, Ulrich Stoeck,, Irena Senkovska, Eike Brunner, Stefan Kaskel. Department of Inorganic Chemistry, Technische Universität Dresden, Bergstraße 66, D Dresden, Germany. Department of Bioanalytical Chemistry, Technische Universität Dresden, Bergstraße 66, D Dresden, Germany. Current address: IFW Dresden, Helmholtzstraße 20, Dresden, Germany. S1

2 1. General methods, equipment and used parameters: Powder X-ray diffraction (PXRD) measurements were performed in transmission geometry on a Stoe StadiP X-ray diffractometer with an imaging plate position sensitive detector. The used X- ray source CuKα 1 provided a wavelength of Å. Measured materials were soaked in solvent (mostly DMF) and prepared as flat base or packed in a glass capillary. Nitrogen physisorption measurements were performed at 77 K using BELSORP MAX instrument. The sample was pretreated in vacuum at room temperature for 1 h directly after supercritical drying (SCD). Thermogravimetric (TG) analysis was performed on air using the STA 409 instrument of Netzsch Company with a heating rate of 5 K/min starting from room temperature to mass constancy. Infrared (IR) spectra were obtained using Vertex 70 instrument from Bruker with a resolution of 2 cm -1. Samples were grounded with potassium bromide and pressed to tablets. Following abbreviations were applied: s-strong, m-medium, w-weak, br-broad. For HPLC investigations the system Elite LaChrom from VWR/Hitachi was used with an evaporative light scattering detector (ELSD 80). Separation of the chiral ligand H 2 L was achieved with enantioselective column Chiracel IA from Daicel. For further analysis the software EZChrom Elite (Agilent Technologies) was used. For ee-value determination percentage peak areas were compared. For analysis of the ligand after MOF synthesis remaining synthesis solution was filtered over celite and washed with methanol. Solvent was evaporated under reduced pressure. Liquid-phase adsorption experiments were carried out under argon at 298 K. The ethyl cinnamte solution in n-heptane (c 0 = 0.1 mol L -1 ) was used. About 50 mg of MOF material were contacted S2

3 with 5 ml of solution and stirred overnight. The amount of adsorbed ethyl cinnamate was determined by GC/MS (Gas Chromatography/Mass Spectrometry). The sample was filtered using a syringe filter (0.45 µm, PTFE). To 0.2 ml of the sample 0.4 ml of a standard solution (hexadecane in n-heptane, M) was added. The concentrations were determined using a calibration curve. The amount of adsorbed substrate (m ads ) was derived from the difference of initial concentration and effective concentration after adsorption. 2. Dye adsorption experiments: To investigate dye adsorption of DUT-32-NHProBoc the solvent was exchanged by dichloromethane (DCM) several times. 5 mg of corresponding dye was dissolved in 10 ml DCM. MOF crystals were dipped into the dye solution and after 60 min washed by fresh solvent. Figure S1. Dye adsorption experiments: (I) DUT-32-NHProBoc in DCM, (II) DUT-32-NHProBoc/Nile blue composite, (III) DUT-32-NHProBoc/Reichardt s dye composite. S3

4 3. X-ray diffraction pattern of powder sample: Figure S2. Powder X-ray diffraction pattern of DUT-32-NHProBoc: (a) calculated from crystal structure, (b) measured after synthesis in DMF. 4. Calculations of total pore volume (V P ): Calculation of total pore volume was performed using the software POREBLAZER 3.0. Calculated total pore volume (V P ) was used to evaluate results from liquid phase adsorption. S4

5 5. Nitrogen-Physisorption: Figure S3. Nitrogen physisorption isotherm at 77 K of DUT-32-NHProBoc after supercritical CO 2 drying: (a) exchange from acetone, (b) exchange from amyl acetate (filled circles: adsorption, empty circles: desorption). 6. Thermal analysis: Figure S4. Thermal analysis of DUT-32-NHProBoc after supercritical drying from CO 2 (exchange from acetone). S5

6 7. Pycnometer mass determination: Due to high porosity, DUT-32-NHProBoc collapses after solvent removal. Therefore the material cannot be dried and weighted for catalytic experiments. To determine the mass of the sample without drying, pycnometer was used. On basis of known crystallographic density of solvent-free crystal (ρ cryst ) and void volume (V void ) of the material, framework density (ρ skeleton ) can be calculated using following equation: ρ skeleton = ρ cryst 100 V void 100% Mass of material (m MOF ) can be calculated by following formula: m MOF = m total ρ solvent V pycno 1 ρ solvent ρ skeleton m total = m MOF+sovent m pycno In this context m total is defined as masses amount of pycnometer filled with material and solvent (m MOF+solvent ) reduced by mass of dry pycnometer (m pycno ). The volume of pycnometer is referred to as V pycno. 8. Stereoselective HPLC measurements of chiral linker H 2 L: For investigation of the enantiomeric purity of the H 2 L linker, HPLC method was elaborated to separate both enantiomers on chiral column Chiralpak IA (DAICEL). Separation could be achieved in methanol / water ( % TFA) mixture as mobile phase with a flow rate of 0.2 ml min -1 (Figure S5). The ee-value was determined by comparison of peak areas of both enantiomers using following equation: ee = Area max Area min Area max + Area min 100% S6

7 The mother liquor after MOF synthesis of DUT-32-NHProBoc was used to investigate the eevalue of H 2 L after MOF synthesis. For sample preparation, the mother liquor was filtered over Celite bed and washed with methanol. Furthermore sample solution was concentrated by evaporation. Figure S5. HPLC chromatogram of the (S)-H 2 L and (R)-H 2 L enantiomers mixture. Systematic study on the thermally catalyzed racemization of H 2 L linker: Further investigations on racemization were done by thermal treatment of ligand H 2 L in DMF (c = mol l -1 ) at 100, 120 and 140 C (Figure 2). HPLC measurements of the solutions at 100 C and 120 C after defined treatment times show only low variations in ee-value. Temperature of 140 C causes distinctive decrease of ee-value within one day. These results indicate a stable ee-value of the incorporated ligand H 2 L under synthesis conditions of DUT-32-NHProBoc (c = mol l -1, 100 C, DEF). S7

8 9. Deprotection studies of DUT-32-NHProBoc: Table S1. Deprotection tests with variation in solvent, temperature and heating method. Degree in deprotection was determined by liquid 1 H-NMR after deconstruction of the material. Structural change was proofed by powder X-ray crystallography. trial solvent method deprotection 1 DMF 2 DMF 3 DMF 4 DMF 5 mesitylene 6 mesitylene 7 mesitylene oven, 165 C, 18 h microwave, 170 C, 4 h oven, 120 C, 7 d oven, 140 C, 7 d oven, 175 C, 12 h oven, 170 C, 24 h oven, 170 C, 72 h structural change 40 % yes 30 % yes (0) - (0) yes 75 % - ~90 % Figure S6. Powder X-ray diffraction pattern of DUT-32-NHProBoc: (a) after MOF synthesis, (b) after deprotection in mesitylene at 170 C for 72 h. S8

9 Figure S7. 1 H-NMR spectra of DUT-32-NHProBoc dissolved in DMSO-d 6 and DCl: (a) after MOF synthesis, (b) after deprotection in mesitylene at 170 C for 72 h. 10. Asymmetric aldol addition and qualitative HPLC investigations: Catalysis was performed using approximately 20 mol-% (with regard to proline groups) of deprotected DUT-32-NHPro. A solution of 4-nitrobenzaldehyde (150 mg, 0.99 mmol) in 5 ml DCM was added to the MOF and the mixture was cooled down to 0 C. Cyclohexanone addition (0.51 ml, 4.96 mmol, 5 eq.) starts the reaction. The suspenison was allowed to warm up to room temperature. Reaction process was monitored by qualitative HPLC measurements. The mixture was sampled every day and filtered with a PTFE syringe filter (0.45 µm). 30 µl of filtered sample was mixed with 500 µl n-hexane. HPLC was performed using the chiral column Chiralpak IA (DAICEL) (mobile phase: n-hexane/2-propanol 80/20) and flow rate of 1 ml min -1. S9

10 The retention times (t R ) of the four possible isomers were identified previosly. Identification of all four isomers in first analysis after reaction start pointed towards no stereoselective catalysis. Figure S8. HPLC chromatogram of products from catalysed asymmetric aldol addition with four possible isomers: syn I, syn II, anti I and anti II. 11. Solvent exchange by TFPE: For solvent exchange with TFPE the as-made microcrystalline phase of DUT-32- NH( 13 C 5, 15 N)ProBoc synthesis was scraped off the pyrex tubes and homogenized. Two aliquots (of ca mg each) were transferred to filter papers for about 10 s in order to remove adhesive DMF from the outer surface of the crystallites. The obtained powders were transfered into 1 ml glass flasks with a conical compartment. Solvent exchange was performed by the addition of µl TFPE to each sample with a subsequent equilibration period of ca. 24 h. After each exchange step, the content of residual DMF in the supernatant solution was determined. For this purpose, an efficient monitoring method has been developed. A small amount of the supernatant solution (ca. 5 µl) was taken by dint of a glass pipette tip. This solution could be directly applied onto the diamond ATR (attenuated total reflection) unit surface S10

11 of the FTIR (fourier transform infrared) spectrometer Bruker FS 88 (Bruker, Karlsruhe, Germany), which has been extended by a Golden Gate ATR unit (Specac Inc., USA). A cap prevented the solvent from evaporation through the formation of a micro atmosphere scans were accumulated for each spectrum in the range of cm -1 with a resolution of 2 cm -1. After the ATR correction, background correction has been obtained by subtraction of a background spectrum. The amount of DMF has been determined by integrating the amide I band in the range cm -1 after a multipoint baseline correction. No IR signals due to decomposition or dissolution products of the MOF samples were observed which shows that the MOF is stable in TFPE. Validation of the monitoring method yields a level of detection of 0.03 vol.-% DMF in TFPE and a level of quantification of 0.09 vol.-% DMF in TFPE. Due to the chemical structure of TFPE and DMF, it can be assumed that the DMF concentration in DUT- 32-NHProBoc is even lower than in the supernatant solution since TFPE should be expected to exhibit a higher affinity to DUT-32-NHProBoc than DMF due to the TFPE's possibility of forming H bonds as well as its aromatic character. The exchange of DMF by TFPE was considered as complete if the DMF content in the supernatant solution dropped below 1 vol.-%. To reach this exchange degree, the DMF-TFPE-mixture was removed after 24 h by a glass pipette. Adhesive liquid was absorbed by a filter paper. To validate this exchange and monitoring method, a blank test has been performed using two aliquots of unlabeled DUT-32-NHProBoc loaded with a racemic mixture of TFPE. After 3 exchange cycles, the DMF content dropped below 0.5 vol.-% for both samples. Both samples were transferred into 2.5 mm ZrO 2 rotors for solid-state NMR measurements. The resulting 13 C CP MAS NMR spectra for the blank test are practically identical. Minor differences in the intensities are due to slightly different bulk densities of the DUT-32-NHProBoc in the rotors. Furthermore, no 13 C NMR signals due to DMF S11

12 were observed which supports the complete removal of DMF from the samples. The method is illustrated in Figure S.I.6 (top) in more detail. At the bottom, we demonstrate the reproducibility of this procedure. Two samples were solvent-exchanged following this strategy. Afterwards, the 13 C CP MAS NMR spectra of these samples are found to be identical. Figure S9. Top: Schematic illustration of the solvent exchange by TFPE. The numbers in vol.-% denote the concentration of DMF detected in the supernatant solution after 24 hours in TFPE. The exchange is considered to be complete after the amount of DMF in the supernatant drops below 1 vol.-%. Bottom: 13 C CP MAS NMR spectra of two different samples after independent solvent-exchange following the described procedure. The relevant measurement parameters (cross-polarization contact time tcp and sample spinning rate fmas are given). 12. Isotope-labeling of the proline side groups and solid-state NMR experiments: Solid-state NMR experiments were carried out on a Bruker Avance 800 spectrometer (Bruker, Karlsruhe, Germany) using a 2.5 mm triple resonance probe. All solid-state NMR spectra were S12

13 acquired using a sample spinning rate of 16 khz and at resonance frequencies of MHz and MHz for 13 C and 1 H, respectively. The 1 H 90 pulse length was 4 µs. For the 13 C- 13 C proton driven spin diffusion experiments, 1 ms 1 H- 13 C CP contact time, 10 ms mixing time, and a recycle delay of 5 s were used. Figure S10. Illustration of the influence of using isotope-labeled proline upon the 13 C CP MAS NMR spectra of as-synthesized ChirDUT-32 containing the solvent DMF (i.e., before solvent exchange by TFPE). Signal assignments to corresponding carbon positions are explained in the inserts. Signals due to DMF are denoted by D. S13