Supporting Information for Facile Synthesis of Highly Crystalline, Covalently Linked Porous Boronate Network R. William Tilford, William R. Gemmill, Hans-Conrad zur Loye and John J. Lavigne* FTIR Analysis Full Spectra 0 0.1 absorbance 0.2 0.3 0.4 3550 2550 1550 550 wavenumbers (cm -1 ) Figure 1. FT-IR analysis of 1,2,4,5-tetrahydroxybenzene.
0 absorbance 0.5 1 1.5 3550 2550 wavenumbers (cm -1 ) 1550 550 Figure 2. FT-IR analysis of benzene-1,3,5-triboronic acid.
0 0.5 absorbance 1 1.5 2 2.5 3550 2550 1550 550 wavenumbers (cm -1 ) Figure 3. FT-IR analysis of COF-18Å.
110 105 100 95 90 % Mass 85 80 75 70 65 60 0 100 200 300 400 500 600 700 800 Temperature ( o C ) Figure 4. Thermogravimetric Analysis of COF-18Å.
Gas Adsorption Analysis Prior to analysis, samples were suspended in dry THF for 24 h. Solvent was decanted and the sample was once again suspended in THF for another 24 h and decanted. After drying under vacuum (1 Torr) for 24 h, the sample was additionally dried at 400 o C under high vacuum (~10-4 Torr) for 2.5 h. 600 cm 3 / g 400 200 0 0 0.2 0.4 0.6 0.8 1 1.2 P/P o Figure 5. Nitrogen gas adsorption isotherm for COF-18Å. Solid-black data points correspond to adsorption data. White data points correspond to desorption data.
600 Uptake (cm 3 /g) 300 0 0 0.5 1.0 P/P o Figure 6. Nitrogen gas adsorption isotherm for COF-18Å. Solid-blue data points correspond to adsorption data. Desorption data was not collected on this analysis. However, this plot provides more data points for surface area calculations (Figure 7).
P/V m (P o -P) 0.03 0.03 0.025 0.02 P / V m (P o - P) 0.02 0.015 0.01 0.01 P/V(PP / V m - P o -P) ) = 0.0842(P =0.0035 / P o ) - 0.0013(P/P o ) 5E-05 R 2 = 0.9962 V m = 1 / (m + b) = 12.06 S A = 1346 m 2 / g R 2 = 0.9983 V m = 0.00345 S A = 1263 m 2 /g 0.005 0 0 0.1 0.2 0.3 P/P o 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 P / P o Figure 7. The BET equation was applied to data points on the isotherm in Figure 6, from 0.05 to 0.3 P/P o in order to calculate the surface area.
Powder X-ray Diffraction Analysis: Powder X-ray diffraction data were collected on a Rigaku DMax 2200 using Cu Kα radiation. The sample was mounted in a deep well glass slide. Data were collected from 2-70 2θ with steps of 0.02 with a count time of 12 s per step. There were no observable diffraction lines beyond 50 2θ so this data was not used in the data analysis. Unit cell dimensions were first determined in the JADE 7 program for X-ray diffraction pattern processing, identification, and quantification. The analysis suggested a hexagonal crystal system with a unit cell of approximately a = 20.805 Å and c = 3.461 Å. The observed diffraction lines were then compiled and used for lattice parameter indexing in the Crysfire suite 1. The indexing program TREOR 2 was used, and a satisfactory indexing was achieved that indicated a primitive hexagonal unit cell, with unit cell dimensions of a = 20.832 Å, c = 3.456 Å. As a consequence of the low number of observed lines (seven) in the diffraction pattern, all attempts to determine an appropriate space group inevitably yielded a large number of possible primitive hexagonal space groups. Structural models were developed in order to identify the stacking sequence in the phase using the DIAMOND 3 3 software package and, thus, to determine the probable space group. There are two likely structural models that differ in the layer stacking sequence: an AA arrangement, where each layer of the network lies directly on top of one another, forming large cavities (Figure 8) and an AB sequence, where the layers are staggered in a graphite-type fashion, forming smaller cavities (Figure 9). The AA model was built in space group P6/mmm and the AB model was built in space group P6 3 /mmc, consistent with similar structural models in the work of Yaghi 4. The atoms were placed on special positions (Table 1) to form the network while maintaining reasonable bond lengths and angles, based on information obtained from single crystal data of the network s structural components.
From these models, the powder diffraction patterns were generated within the software package Diamond 3, and compared to the diffraction pattern of the sample. The calculated diffraction patterns (calculated from Powdercell 5 ) for both models along with the experimental diffraction pattern of the sample are shown in Figure 10. It is clear that the calculated pattern for the AA model most closely resembles the experimental pattern, and the calculated pattern for the AB model is quite different that the experimental data. In particular, the mismatch in the relative intensity of the 110 and 200 peaks and the presence of calculated intensity in the range of 14-19 2θ tend to rule out the AB model. Figure 8. Structure of COF-18Å as proposed by the AA model
Figure 9. Structure of COF-18Å as proposed by the AB model (110) (200) Exp. 10 20 30 Degrees 2θ Figure 10. X-ray diffraction patterns obtained experimentally for COF-18Å and calculated for the two proposed models, AA and AB. AA AB
Table 1. Fractional atomic coordinates for AA model obtained from Diamond 3. Atom ID Wyckoff site x y z C1 6k 0.557 0 1/2 C2 12q 0.569 0.072 1/2 C3 6m 0.2966 0.5932 1/2 C4 6m 0.6265 0.253 1/2 B1 6m 0.59 0.18 1/2 O1 12q 0.63 0.142 1/2 Diffraction data analysis was performed using the EXPGUI interface of GSAS 6,7. The model based LeBail method was used to refine the lattice parameters and refine the profile function. The background was graphically fit by applying a linear extrapolation function with 3 terms. The model used was obtained from the structural framework previously built in Diamond 3. The profile was refined with both Gaussian and Lorentzian peak shape. The refined lattice parameters were determined to be a = 20.812(3) Å and c = 3.385(1) Å. This refinement rapidly converged with good statistics, Table 2. The observed, calculated, and difference plots along with the allowed reflection markers are shown in Figure 11. Attempts at refining atomic positions all diverged due to the small scattering power of the light atoms in the structure. Consequently, the atomic positions were fixed during the structural analysis. A reasonable fit to the data was achieved which supported the proposed model. Finally, it should be noted that when the AB structural model was used for the LeBail method the fit was poor throughout the data range, and efforts for further refinement were abandoned.
Table 2. Refinement statistics from LeBail fit of AA model in GSAS 6,7 Refinement Statistic Value R p 0.0598 wr p 0.0422 χ 2 2.286 14000 12000 Intensity (Arbitrary Units) 10000 8000 6000 4000 2000 0-2000 10 20 30 40 50 Degrees 2θ Figure 10. LeBail method fit to powder X-ray diffraction data for COF-18Å. Black crosshatches are collected data, red solid lines are fitted. Differences between observed and calculated intensities are shown in blue below. Vertical marks indicate allowed peak positions.
References 1 Shirley, R. The Crysfire 2002 System for Automatic Powder Indexing: User's Manual: The Lattice Press: 41 Guildford Park Avenue, Guildford, Surrey GU2 7NL, England, 2002. 2 Werner, P.-E., Eriksson, L.; Westdahl, M. TREOR, a Semi-Exhaustive Trial-and-Error Powder Indexing Program for All Symmetries. J. Appl. Cryst. 1985, 18, 367-370. 3 Bradenburg, K. Diamond, version 3.1a; Crystal Impact GbR: Bonn, Germany, 2005. 4 Côté, A. P.; Benin, A. I.; Ockwig, N. W.; O Keeffe, M.; Matzger, A. J.; Yaghi, O. M. Science 2005, 310, 1166-1170.. 5 PowderCell for Windows, version 2.4; Kraus, W.; Notze, G.; Federal Institute for Materials Research and Testing: Berlin, Germany, 2000. 6 Larson, A. C.; von Dreele, R. B. General Structure Analysis System (GSAS) Los Alamos National Laboratories: 1990. 7 Toby, B. H. J. Appl. Crystallogr. 2001, 34, 210.