Supporting information Adsorptive Denitrogenation of Model Fuel with CuCl-loaded Adsorbents: Contribution of Π- complexation and Direct Interaction between Adsorbates and Cuprous Ions Nazmul Abedin Khan, Nizam Uddin, Cheol Ho Choi, and Sung Hwa Jhung * Department of Chemistry and Green Nanomaterials Research Center, Kyungpook National University, Daegu 41566, Korea S1
Characterization of the adsorbents The sample phases were determined using X-ray diffraction (XRD; D2 Phaser, Bruker, Cu Kα radiation). Nitrogen adsorption isotherms were recorded at 196 C with a surface area and porosity analyzer (Micromeritics, Tristar II 3020) after evacuation of the adsorbents at 150 C for 12 h. The surface areas were calculated using the Brunauer-Emmett- Teller equation. The weight % of copper and elemental mapping of the samples were examined with FE-SEM and EDS (field emission scanning electron microscope and energydispersive X-ray spectroscopy, Hitachi, SU-8220). X-ray photoelectron spectroscopy (XPS) was performed using a Quantera SXM X-ray photoelectron spectrometer (ULVAC-PHI) with a dual-beam charge neutralizer. Adsorption experiments Stock solutions of the adsorbates (10000 µg/g) were prepared by dissolving QUI, tetra-qui and deca-qui in n-octane. Solutions of different concentrations (5000 50 µg/g) were prepared by successive dilutions of the stock solution with n-octane. The concentrations of the adsorbates were determined using a gas chromatograph (DS 6200, DS Science Inc.) equipped with a flame ionization detector. Before adsorption, the adsorbents were dried overnight under vacuum at 100 C and were kept in a desiccator. The adsorbents (~ 5.0 mg) were added to the quinolines solutions (5.0 ml) with fixed concentrations. The solutions containing the adsorbents were mixed well with magnetic stirring and maintained for a fixed period of time (1 to 12 h) at a constant temperature of 25 C. After adsorption for a predetermined time, the solution was separated from the adsorbent with a syringe filter (polytetrafluoroethylene, hydrophobic, 0.5 μm), and the concentrations of the quinolines were measured using gas chromatography. Calculation of adsorption capacities S2
All the adsorption capacities (mg/g) were calculated from the difference between the final and initial concentrations of the adsorbate using the following equation: ( Ci C f qt m ) V Where, q t is the amount adsorbed in time t (mg/g) C i is the initial concentration of the adsorbate (mg/ml) C f is the final concentration of the adsorbate after adsorption (mg/ml) V is the volume of the solution subjected to a single adsorption (ml) m is the mass of the adsorbent taken during a single adsorption (g). Calculation of maximum adsorption capacity (Q 0 ) The maximum adsorption capacity (Q o ) was calculated based on the Langmuir adsorption isotherm after adsorption for 12 h under various conditions. The adsorption isotherms for different adsorbents were plotted according to the Langmuir equation. 1 C q e e C Q e o 1 Q b o Where, C e is the equilibrium concentration of the adsorbate (mg/l) q e is the amount adsorbed at equilibrium (mg/g) Q o is the Langmuir constant (maximum adsorption capacity, mg/g) b is the Langmuir constant (L/mg). Therefore, the maximum adsorption capacity, Q o, can be obtained from the reciprocal of the slope of a plot of C e /q e against C e. Theoretical calculations of bond energy (BE) and pkb of deca-qui S3
The GAMESS 2 program was used to optimize the stable geometry of compounds and their relative energies are in gaseous state and are given at 298 K. Dunning-type Correlation Consistent systematic complete basis set cc-pvtz was employed with Second-order Møller Plesset (MP2) theory as the most economical way to treat electron correlation in a nonparameterized fashion. MacMolPlt 3 software package was used to view the molecular structure and molecular orbitals. The geometries of quinolone, tet-qui, deca-qui and also there complexes with Cu(I) were optimized at the MP2/cc-pVTZ level of theory. Each energy minimized structure was tested by a harmonic vibrational analysis. The bond energy (BE) is the measure of bond strength in a chemical bond that defines the amount of energy needed to break/make a bond between a covalently bound gases at a specified temperature, here at 298 K. The pk b of the deca-qui was predicted by the QM/MM-MD 4,5 simulation. A spherical system of QM molecules surrounded by 290 TIP5P water molecules was prepared for the QM/MM-MD simulations. The QM molecules are deca-qui and one quantum water, which are treated with HF/6-31G(d) level of theory. In order to prevent evaporation of waters during long time simulations, we applied a harmonic restraint potential with a force constant of 2.0 kcal/mol/å 2 for the boundary solvent molecules. The simulations on 18 umbrella windows were performed to cover the whole reaction path from hydrogen bonded structure to deprotonated structure. The one-dimensional potentials of mean force (PMF) from the umbrella samplings were obtained using the weighted histogram analysis method (WHAM). 6 Initially, the structure in each US window was equilibrated for 50 ps at 300 K. Then, QM/MM-MD production runs for the NVT ensemble at 300 K were continued for 100 ps after equilibrations. Details of the MD simulation was described in reference 5. The essential quantity required for the p prediction of a given molecule in aqueous solution is the Helmholtz free energy difference for hydrogen abstraction by water, S4
AH H O A H O (1) N H where, AH= and A = N After that, the p can be directly computed from PMF using the fundamental statistical thermodynamics formalism. For the acid dissociation reaction of eq. (1), the thermodynamic equilibration constant is defined as / / / /, (2) where, is the concentration of species A, and is the standard concentration in the same units as, which is 1 M (mole/liter) for solute and 55.5 M for water solvent. The is dimensionless and generally written without the water concentration, because the ratio of / is equal to 1. From classical statistical mechanics, given an arbitrary reaction coordinate, the probability distribution is related with the potential of mean force (PMF) as 4 e, where 1/ and is the free energy (PMF) along the reaction coordinate. The fraction of the configuration in the bound state can be computed as, (3) where, denotes transition state or the dividing surface between reactant and product. The concentration of the species are expressed with this fraction and the total volume ( ) as /, and 1 /. Thus is then given by. (4) S5
In the limit of the infinitely dilute solution, the fraction, 0 and. Since the simulation are calculated in particle/, the acid dissociation constant is obtained by the following: 1660 4 e, (5) where the unit of distance is and the that of free energy is RT. The number, 1660 is a factor converting particle/ to mol/l, which is calculated as follows: 1 particle 1/6.022 10 1.66 10 mol and 1 1 10 m 1 10 L, so 1 particle/ 1.66 10 mol/1.00 10 L 1660 mol/l. p p 14 (6) A modified version of GAMESS 2 was used to run QM/MM-MD simulation. The recent distribution contains those modifications. REFERENCE 1. Hasan, Z.; Jeon, J.; Jhung, S. H., Adsorptive Removal of Naproxen and Clofibric Acid From Water Using Metal-Organic Frameworks. J. Hazard. Mater. 2012, 209, 151-157. 2. Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S.; Windus, T. L.; Duplis, M.; Montgomery Jr, J. A., General Atomic and Molecular Electronic Structure System. J. Comput. Chem. 1993, 14, 1347 1363. 3. Bode, B., MacMolPlt, Gordon, M. S., MacMolPlt: A Graphical User Interface for GAMESS. J. Mol. Graphics Modell. 1998, 16, 133-138. 4. Choi, C. H.; Re, S.; Feig, M.; Sugita, Y. Quantum Mechanical/Effective Fragment Potential Molecular Dynamics (QM/EFP-MD) Study on Intra-Molecular Proton S6
Transfer of Glycine in Water. Chem.Phys. Lett. 2012, 539 540, 218 221. 5. Uddin, N.; Choi, T. H.; Choi, C. H., Direct Absolute pk a Predictions and Proton Transfer Mechanisms of Small Molecules in Aqueous Solution by QM/MM-MD. J. Phys. Chem. B 2013, 117, 6269 6275. 6. Grossfield, A., "WHAM: The Weighted Histogram Analysis Method", Version XXXX, http://membrane.urmc.rochester.edu/content/wham 7. Nguyen, M. T.; Tayakout-Fayolle, M.; Pirngruber, G. D.; Chainet, F.; Geantet, C., Kinetic Modeling of Quinoline Hydrodenitrogenation over a NiMo(P)/Al 2 O 3 Catalyst in a Batch Reactor. Ind. Eng. Chem. Res. 2015, 54, 9278-9288. 8. Ho, T. C., Hydrodenitrogenation Catalysis. Catal. Rev.: Sci. Eng. 1988, 30, 117-160. S7
(a) Simulated MIL-100(Cr) MIL-100(Cr) CuCl(10.0)/MIL-100(Cr) (b) Intensity (a.u) Intensity (a.u.) 5 10 15 20 25 30 2 theta (deg) AC CuCl(10.0)/AC 10 20 30 40 50 2 theta (deg.) Figure S1. XRD patterns of (a) MIL-100(Cr)s and (b) ACs. S8
600 (a) Quantity adsorbed, cm 3 /g 400 200 MIL-100(Cr) CuCl(2.5)/MIL-100(Cr) CuCl(5.0)/MIL-100(Cr) CuCl(10.0)/MIL-100(Cr) 0 0.0 0.2 0.4 0.6 0.8 1.0 Relative pressure, P/P o (b) Quantity adsorbed, cm 3 /g 400 200 AC CuCl(5.0)/AC 0 0.0 0.2 0.4 0.6 0.8 1.0 Relative pressure, P/P o Figure S2. Nitrogen adsorption isotherms of (a) MIL-100(Cr)s and (b) ACs. S9
C O Cr Cu Cl Figure S3. EDS mapping of CuCl(5.0)/MIL-100(Cr). S10
(a) CuCl 2 CuCl CuCl(10.0)/MIL-100(Cr) CuCl 2 CuCl CuCl(10.0)/AC Intensity (a.u) Intensity (a.u) 960 940 Binding Eenergy (ev) (b) 970 960 950 940 930 Binding energy (ev) Figure S4. XPS spectra of CuCl supported (a) MIL-100(Cr) and (b) AC. XPS spectra of CuCl and CuCl 2 are shown as references. S11
5 (a) 4 C e /q e, mg/l 3 2 1 MIL100-Cr CuCl(5.0)/MIL-100(Cr) 10 8 (b) 0 0 250 500 750 1000 1250 1500 C e, ppm 14 12 (c) C e /q e, mg/l 6 4 C e /q e, mg/l 10 8 6 2 MIL100-Cr CuCl(5.0)/MIL-100(Cr) 4 2 MIL100-Cr CuCl(5.0)/MIL-100(Cr) 0 0 250 500 750 1000 1250 1500 C e, ppm 0 0 250 500 750 1000 1250 1500 C e, ppm Figure S5. Langmuir plots for the adsorption of (a) QUI, (b) tetra-qui and (c) deca-qui over the virgin and CuCl-supported MIL-100(Cr)s. S12
QUI tetra-qui deca-qui Figure S6. Optimized structure of the QUI-, tetra-qui- and deca-qui-cu(i) complex by MP2/cc-pVTZ level of theory. Ash: carbon; grey: hydrogen; blue: nitrogen and orange: copper. S13
QUI tetra-qui deca-qui LUMO LUMO LUMO HOMO HOMO HOMO Figure S7. Calculated molecular orbitals HOMO-LUMO transitions of the QUI-, tetra-qui- or deca-qui-cu(i) by MP2/cc-pVTZ level of theory. Grey: carbon; white: hydrogen; blue: nitrogen and orange: copper. S14
20 BE, KJ/mol pk b 15 10 5 0 QUI tetra-qui deca-qui Figure S8. Cu(I)-quinolines bond energies and pk b values of the quinolines. The pk b values of QUI and tetra-qui were retrieved from reference 7,8.The pk b value of deca-qui was calculated by the QM/MM-MD 4,5 simulation by HF/6-31G* level which are implemented in GAMESS 2 (General Atomic and Molecular Electronic Structure System) software package. S15