S1

SUPPORTING INFORMATION

Exceptional Gravimetric and Volumetric CO2 Uptake in a Palladated NbO-type MOF Utilizing Cooperative Acidic and

Basic, Metal-CO2 Interactions

I. Spanopoulosa, I. Bratsosb, C. Tampaxisa,b, D. Vourloumisb, E. Klontzasa, G.E. Froudakisa, G. Charalambopouloub, T.A. Steriotisb and P.N. Trikalitisa,*

a Department of Chemistry University of Crete, Heraklion, Voutes, 71003, Greece. Email: ptrikal@uoc.gr. b National Center for Scientific Research “ Demokritos”, 15341 Agia Paraskevi Attikis, Greece.

Email: ptrikal@uoc.gr

Table of Contents

Methods and instrumentation S2

Synthesis and Characterization S3

Single crystal X-ray crystallography S7

Powder X-ray diffraction (PXRD) measurements S9

Gas sorption measurements and analyses S10

Thermal gravimetric analysis (TGA) S18

Computational Study S19

References S22

Electronic Supplementary Material (ESI) for ChemComm.This journal is © The Royal Society of Chemistry 2016

S2

Methods and instrumentation

Starting materials. All starting materials for synthesis were purchased commercially and were used

without further purification. CuCl2·2H2O (99%) and N,N-dimethylformamide (DMF, ≥ 99.8%) were

purchased from Merck, [Pd(PhCN)2Cl2] (99%) from Strem Chemicals, pyridine-3,5-dicarboxylic acid

(PDC, 98%) from Alfa Aesar, whereas tetrahydrofuran (THF, 99.9%) from Aldrich.

Powder X-ray diffraction patterns were collected on a Panalytical X’pert Pro MPD system (CuKα

radiation) operated at 45 kV and 40 mA. A typical scan rate was 1 sec/step with a step size of 0.02

deg. Thermogravimetric analyses (TGA) were performed using a Setaram Setsys Evolution 16

instrument. An amount of 10 mg of sample was placed inside an alumina crucible and heated up to

800 oC under Argon flow with a heating rate of 5 oC/min. Scanning electron microscopy (SEM) images

were collected using a JEOL JSM-6390LV instrument equipped with an Oxford Energy Dispersive

Spectroscopy (EDS) detector for elemental analysis.

Gas-sorption measurements. Low-pressure nitrogen, argon, carbon dioxide and methane sorption

measurements were carried at different temperatures using an Autosorb-1MP instrument from

Quantachrome equipped with a cryocooler system. Ultra-high purity grade N2 (99.999%), Ar

(99.999%), He (99.999%), CO2 (99.9995%) and CH4 (99.9995%) were used for all adsorption

measurements.

S3

Synthesis and Characterization

N

COOHHOOC

C N Pd N

Cl

Cl

C

N Pd N

Cl

Cl

HOOC

HOOC

COOH

COOH

THF

+

ArR.T.

H4L

PDC

Figure S1. Synthetic route for the synthesis of the palladated ligand H4L.

Synthesis of H4L ligand: Complex H4L was synthesized according to a slight modification of a

published procedure.1 An amount of PDC (334.2 mg, 2.0 mmol) was suspended in THF (150 mL) and

the mixture was degassed by three vacuum/Ar cycles. After the addition of [Pd(PhCN)2Cl2] (383.6 mg,

1.0 mmol) the mixture was degassed again ( 3), and then was stirred for 5 h under Ar atmosphere

at room temperature to yield a yellow solution. Rotary concentration under reduced pressure to ca.

1/3 of the initial volume and addition of hexane (250 mL) induced the formation of the product as a

pale yellow solid. It was collected by filtration, washed with THF/hexane and hexane, and vacuum-

dried. Yield = 501.1 mg, (98%). Characterization data agreed with those previously reported. FT-IR

(neat solid): 3570, 3501, 3067, 2494, 1714, 1595, 1450, 1375, 1319, 1290, 1216, 1161, 938, 791, 746,

666, 570, 497. 1H NMR (500 MHz, CD3OD): 9.55 (d, J = 1.7 Hz, 4H, C2H/C6H), 8.96 (t, J = 1.7 Hz, 2H,

C4H). 13C NMR (500 MHz, CD3OD): 165.0 (COOH), 158.3 (C2H/C6H), 141.6 (C4H), 130.0 (C3/C5).

S4

Figure S2. 1H NMR spectrum of H4L in CD3OD (500 MHz).

Synthesis of NbO-Pd-1

A solution of 10 mL DMF, 0.025 g (0.049 mmol) of H4L and 0.025 g (3 equiv., 0.147 mmol) of

CuCl2·2H2O were placed in a 20 mL glass scintillation vial. The vial was sealed and placed in an

isothermal oven at 85 °C for 72 hours. During this period, small light blue cubic crystals of [Cu2L],

suitable for X-ray diffraction analysis, were deposited. (40% yield based on H4L).

S5

Figure S3. A representative SEM image of NbO-Pd-1 crystals.

Figure S4. A representative EDS spectrum of NbO-Pd-1 single crystals. The observed Cu:Pd:Cl ratio is

very close to 2:1:2.

S6

4000 3500 3000 2500 2000 1500 1000 500

20

40

60

80

100

1375

1568

Tran

smitt

ance

(%)

Frequency, cm-1

H4L NbO-Pd-1

1715

1616

Figure S5. ATR-IR spectrum of as-made NbO-Pd-1. For comparison, the corresponding spectrum of

palladated ligand H4L is also shown.

Activation procedure: Prior to gas sorption measurements, the as-made sample was washed with

DMF four times per day for 2 days and then soaked in methanol over a period of 4 days, replenishing

the methanol 4 times per day. The sample was activated under dynamic vacuum at 60 °C for 12

hours. The activation of NbO-Pd-1 is visually evidenced by the change of the color of the sample from

green (as-synthesized) to deep blue-violet (activated), as shown in Fig. S6.

Figure S6. Visual color change of NbO-Pd-1 upon activation.

S7

Single crystal X-ray crystallography

Single-crystal diffraction data were collected at beamline I19, 90 Diamond Light Source, Didcot, UK

using a wavelength λ = 1.0402 Å at 120 K. Solvated single-crystals were mounted on MiTeGen loops.

Data collection, integration and reduction were performed using CrystalClear software from Rigaku.2

The structure was solved by direct methods using SHELXS and refined by full-matrix least squares on

F2 using SHELXL software.3 6 CCDC number 1482371.

Table S1. Crystal data and structure refinement for NbO-Pd-1.Empirical formula C12 H6 Cl2 Cu2 N2 O10 Pd

Formula weight 642.57

Temperature 120(2) K

Wavelength 0.6889 Å

Crystal system Hexagonal

Space group R-3m

Unit cell dimensionsa = 18.899(6) Å, α = 90°b = 18.899(6) Å, β = 90°c = 31.690(1) Å, γ = 120°

Volume 9802(7) Å3

Z 9

Density (calculated) 0.980 g/cm3

Absorption coefficient 1.523 mm-1

F(000) 2790

Crystal size 0.200 x 0.200 x 0.200 mm3

θ range for data collection 2.492 to 26.573°

Index ranges -24<=h<=24, -23<=k<=24, -41<=l<=37

Reflections collected 32466

Independent reflections 2722 [Rint = 0.0470]

Completeness to θ = 24.415° 99.4%

Refinement method Full-matrix least-squares on F2

Data / restraints / parameters 2722 / 0 / 83

Goodness-of-fit 2.415

Final R indices [I > 2σ(I)] Robs = 0.0877, wRobs = 0.2933

R indices [all data] Rall = 0.0899, wRall = 0.2946

Extinction coefficient .

Largest diff. peak and hole 1.650 and -0.660 e·Å-3

R = Σ||Fo|-|Fc|| / Σ|Fo|, wR = {Σ[w(|Fo|2 - |Fc|2)2] / Σ[w(|Fo|4)]}1/2 and w=1/[σ2(Fo2)+(0.1000P)2] where P=(Fo2+2Fc2)/3

S8

Figure S7. In NbO-Pd-1, the Cl- anions are found to be disordered over two positions.

S9

Powder X-ray diffraction (PXRD) measurements

5 10 15 20 25 30 35 40 45 50

Evacuated

As-madeInte

nsity

(arb

. uni

ts)

2, CuK

Calculated

Figure S8. Experimental PXRD patterns of the as-made (red) and activated (blue) NbO-Pd-1. For comparison, the PXRD pattern calculated from the single crystal structure (black) is also shown.

S10

Gas sorption measurements and analyses

Table S2. Gravimetric and volumetric CO2 uptake of the best performing MOFs at 273 K and 298 K, up to 1 bar.

273 K 298 KCompound

cm3 g-1 cm3 cm-3 cm3 g-1 cm3 cm-3

Mg-MOF-74 228.3 208.4 178.1 162.5NbO-Pd-1 201.8 187.8 124.8 116.2Cu-TDPAT(rht-7) 227.0 177.8 132.0 103.4CPM-200(Fe/Mg) 207.6 165.1 127.3 101.2

Cu-TPBTM 216.6 135.9 118.5 74.3

NJU-Bai21 206.5 121.2 115.1 72.2HNUST-3(NOTT-125a) 203.7 140.6 92.6 63.9

CPM-33b 173.9 168.7 126.4 122.6

Figure S9. Comparison of the gravimetric (blue) and volumetric (orange) CO2 uptake at 298 K and 1 bar, of NbO-Pd-1 with top performing MOFs.

S11

Figure S10. BET plot for NbO-Pd-1 calculated from the Ar adsorption isotherm at 87 K.

Figure S11. Langmuir plot for NbO-Pd-1 calculated from the Ar adsorption isotherm at 87 K.

S12

Figure S12. Argon adsorption isotherm of NbO-Pd-1 recorded at 87 K and the corresponding NLDFT

fitting.

Low pressure CO2, N2 and CH4 sorption isotherms, determination of heat of adsorption and selectivity (CO2/N2, CO2/CH4) calculations using IAST.

Heat of adsorption. To calculate heats of adsorptions, the corresponding adsorption isotherms at different temperatures were simultaneously fitted using the virial type4 Equation 1:

(1)

m

i

n

i

ii

ii NbNa

TNP

0 0

1lnln

The heat of adsorption at zero coverage was calculated from Equation 2, where as a function of surface coverage, from Equation 3:

(2)ost RaQ

(3)

m

i

iist NaRNQ

0)(

For the determination of the isosteric heat of adsorption using the Clausious Clapeyron equation a commercially available software, ASiQwin (version 3.01) purchased from Quantachrome, was used.

S13

Gas selectivity using IAST. The corresponding calculations were performed according to an established procedure.5 Specifically, the single-component adsorption isotherms were described by fitting the data with the following virial-type equation:

(4)𝑝=

𝑛𝐾𝑒𝑥𝑝(𝑐11𝑛+ 𝑐2𝑛

2 + 𝑐3𝑛3 + 𝑐4𝑛

4)

where p is the pressure in Torr, n is the adsorbed amount in mmol g-1, K is the Henry constant in mmol g-1 Torr-1 and ci are the constants of the virial equation.

The free energy of desorption at a given temperature and pressure of the gas is obtained from the analytical integration of eq. (4):

(5)

𝐺(𝑇,𝑝) = 𝑅𝑇𝑝

∫0

𝑛𝑝𝑑𝑝= 𝑅𝑇(𝑛+

12𝑐1𝑛

2 +23𝑐2𝑛

3 +34𝑐3𝑛

4 +45𝑐4𝑛

5)

The free energy of desorption is a function of temperature and pressure G(T,p) and describes the minimum work (Gibbs free energy) that required to completely degas the adsorbent surface.

For a binary mixture of component i and j eq. (5) yields the individual pure loadings and at the 𝑛0𝑖 𝑛0𝑗

same free energy of desorption:

(6)𝐺0𝑖(𝑛0𝑖) = 𝐺0𝑗(𝑛0𝑗)

The partial pressure of component i and j in an ideal adsorption mixture is given by the following equations:

(7)𝑝𝑦𝑖= 𝑝0𝑖(𝑛

0𝑖)𝑥𝑖

(8)𝑝𝑦𝑗= 𝑝0𝑗(𝑛

0𝑗)𝑥𝑗

where yi (=1-yj) and xi (=1-xj) is the molar fraction of component i in the gas phase and the adsorbed

phase respectively and , is the pure component pressure of i and j respectively. From eq. (6)-(8) 𝑝0𝑖 𝑝0𝑗

and (3), the selectivity for the adsorbates i and j (Si,j) and the total pressure (p) of the gas mixture were calculated from eq. (9) and eq. (10), respectively.

(9)

𝑆𝑖𝑗=𝑥𝑖/𝑦𝑖𝑥𝑗/𝑦𝑗

=𝑝0𝑗

𝑝𝑜𝑖

S14

(10)

𝑝=𝑗

∑𝑖

(𝑝𝑜𝑖𝑥𝑖)

S15

Figure S13. Virial type fitting of CO2 adsorption isotherms of NbO-Pd-1 at 273 K, 288 K, 293 K and 298 K according to equation 1.

Figure S14. CO2 isosteric heat of adsorption in NbO-Pd-1 as a function of surface coverage, extracted from the virial-type analysis. The corresponding Clausius-Clapeyron calculation is show with the solid line.

S16

Figure S15. CH4 sorption isotherms of NbO-Pd-1 recorded at the indicated temperatures.

Figure S16. Virial type fitting of CH4 adsorption isotherms of NbO-Pd-1 at the indicated temperatures, according to equation 1.

S17

Figure S17. CH4 isosteric heat of adsorption in NbO-Pd-1 as a function of surface coverage, extracted from the virial-type analysis. The corresponding Clausius-Clapeyron calculation is show with the solid line.

Figure S18. N2 sorption isotherms of NbO-Pd-1 recorded at the indicated temperatures.

S18

Figure S19. Selectivity of CO2 over N2 at the indicated temperatures for NbO-Pd-1 as predicted by IAST for a 5/95 CO2/N2 molar mixture.

Figure S20. Selectivity of CO2 over CH4 at the indicated temperatures for NbO-Pd-1 as predicted by IAST for a 5/95 CO2/CH4 molar mixture.

S19

Thermal gravimetric analysis (TGA)

Figure S21. TGA curve of the evacuated NbO-Pd-1, recorded under Argon flow with a heating rate of 5 deg/min.

S20

Computational Study

All calculations were performed by using Density Functional Theory (DFT) and the Generalized

Gradient approximation. The geometry optimizations of the molecular models and the calculation of

the binding energies were performed with Perdew-Burke-Ernzerhof (PBE)6 exchange-correlation

functional with the latest dispersion correction scheme proposed by Grimme (-D3)7 and the triple z

quality basis set def-TZVPP8. The molecular models used were taken from the corresponding

crystallographic information file and the dangling bonds were saturated with H atoms. During

calculations the position of the atoms was fixed except for H atoms and the CO2. Tight convergence

criteria were enforced on the SCF energy (10-8 au) and the Cartesian gradient (10-4 au). All

calculations were performed with a locally modified version of the Turbomole quantum chemistry

package9.

In order to have a better understanding of the CO2 adsorption behavior of Pd-Cu MOF,

quantum chemistry calculations were performed to analyze the interaction of CO2 with the binding

sites of the MOF. We focused on the uncoordinated metal sites which can be found in Pd-Cu-MOF,

since it has been proved that such binding sites mainly contribute to the CO2 adsorption in MOF. Two

distinct open metal sites can be identified in Pd-Cu MOF. The first is located on the Cu-PDW building

unit, where two Cu open metal sites are available to coordinate up to 1 CO2 per Cu. The second is

located on Pd linker, where 2 CO2 can be coordinated over Pd atoms on opposite side of the linker.

We investigated the structural characteristics and the corresponding interaction energies of CO2

adsorption with these sites by using various molecular models obtained from the available CIF. These

models are presented in Figure S22. In the case of Cu open metal site, CO2 can have five distinct

orientations with respect to the local environment of the Cu open site (Figure S22). Optimization of

the CO2 location showed that the energetically most favorable orientation is the one presented in

Figure S22a,b. CO2 is oriented towards an imaginary plane formed by the Pd linker (Figure S22a),

where in the same time the whole molecule stays almost parallel to the imaginary plane formed by

Cu and the surrounding oxygen atoms (O-Cu-O-C dihedral : 13.3 degrees, Figure S22b). As it has been

found in previous studies, CO2 acts as a Lewis base, acting as an electron density donor from the

nearest O atom towards Cu atom. The closest oxygen atom of the CO2 is located over the Cu atom at

a distance of 2.54 Å and the angle formed between the Cu atom and the oxygen and carbon atom of

the CO2 was measured at 117.4o. The carbon – oxygen bond nearest the Cu atom was found to be

slightly elongated while the other bond was slightly contracted and a slight distortion of the O-C-O

S21

angle (0.7 degrees) of CO2 was also noticed. The interaction energy for this specific CO2 orientation

was found to be 30.3 kJ/mol, which is a value similar to the interaction energies that have been

calculated for CO2 in Cu-PDW MOF. Two more configurations (not shown) of CO2 were found during

the optimization procedure for all CO2 orientations with lower interaction energies at 28.7 kJ/mol

and 25.5 kJ/mol respectively.

The interaction of CO2 with the Pd linker was also examined by using the model shown in

Figure S23. Various orientations of CO2 with respect to Pd were generated, but in all cases the

optimized configuration was the one presented in Figure S23. CO2 adopted a different orientation

with respect to the Pd metal site, which can be considered as an evidence of the different nature of

the interaction with respect to the Cu site. CO2 is oriented in a parallel fashion to an imaginary axis

formed by the N-Pd-N atoms of the linker, as can be seen in Figure S23. Moreover, the carbon atom

of the CO2 is located almost exactly in front of Pd (Cl-Pd-Cl-CCO2 dihedral: 167.7o) at a distance of

3.37Å. CO2 act as a Lewis acid through the carbon atom and Pd as a Lewis base, donating electron

density towards the electron deficient carbon atom. The interaction is further stabilized due to the

presence of chlorine atoms, as it is evident from the short CCO2-Cl distance (3.5Å). The interaction

resulted in a slightly bend of the CO2 angle at 178.1o and a slight elongation of both C-O bonds. The

interaction energy was calculated to be 24.1 kJ/mol. We also found that a second CO2 can

simultaneously interact with the Pd linker by adopting the same orientation in the opposite site of

the linker with respect to the first CO2. The average interaction energy for both CO2 was found to be

23.3 kJ/mol.

We also examined the effect on the complex geometry and the corresponding interaction

energy for the simultaneous occupation of a Cu site and a Pd site by CO2. We adopted the molecular

model of Figure S23 and we placed to CO2 molecules in the optimized geometries that we have

previously found for each individual site. We observe that both Cu and Pd sites can be simultaneously

occupied by CO2, where the orientation of CO2 on Cu stay unaltered. The position of CO2 on Pd site is

slightly altered with respect to geometry in Figure S23, with the major alternation to be the longer

Pd-C distance (3.95Å) and the similar distance between CCO2 and the neighboring Cl atoms (~3.4 Å).

The average interaction energy for both CO2 was found to be 31.4 kJ/mol, which dictates that the

simultaneous presence of two CO2 in neighboring Cu and Pd sites increase the overall CO2 interaction

energy due to CO2-CO2 interaction.

We also performed calculations in order to check the simultaneous occupation of open Cu

sites in the vicinity of the pore window. The molecular model that was used during calculations can

S22

be seen in Figure S24. Three open Cu sites can be found in this area which can be simultaneously

occupied by three CO2 molecules. We found that there is a cooperative effect between the CO2

molecules as we increase the occupation of the Cu sites, which lead to the increase of the average

CO2 binding energy. This effect is accompanied by a small distortion of the orientation of the CO2

molecule with respect to the orientation of the first CO2 molecule which was added in the molecular

model. The binding energy was found to be 26.7 kJ/mol for the first CO2 molecule, where the average

binding energy when considering three CO2 molecules was found to be 29.2 kJ/mol.

a)

b)Figure S22: Different views (a, b) of the CO2 geometry obtained after optimization of CO2 position

over Cu site. Pink, dark green, red, blue, gray, light green and white correspond to Cu, Pd, O, N, C, Cl and H atoms respectively.

S23

Figure S23. (Top) Different views (a, b) of the CO2 geometry obtained after optimization of CO2 position over Pd site. Pink, dark green, red, blue, gray, light green and white correspond to Cu, Pd, O, N, C, Cl and H atoms respectively. (Bottom) The case of two CO2 molecules at the opposite sites of the Pd(II) atom.

S24

Figure S24. Optimized geometries for the interaction of several CO2 molecules with the open Cu sites around pore window, (a) 1 CO2, (b) 2 CO2 και (c) 3 CO2. Two different views (top and bottom) of the model are considered for each case in order to get a better understanding of the orientation of the CO2 molecules. Pink, red, blue, gray and white correspond to Cu, O, N, C and H atoms respectively.

References

1 Qin, Z.; Jennings, M. C.; Puddephatt, R. J.; Muir, K. W. Inorg. Chem. 2002, 41, 5174.2 Rigaku Corporation, CrystalClear-SM Expert 2.0 r13, Tokyo, Japan, 2011.3 Sheldrick, G. M. SHELXL; University of Göttingen: Göttingen, Germany, 2002.4 (a) Czepirski L., Jagiello J., Chem. Eng. Sci., 1989, 44, 797-801, (b) Sumida K., Rogow D. L., Mason J.

A., McDonald T. M., Bloch E. D., Herm Z. R., Bae T.-H., Long J. R., Chem. Rev., 2012, 112, 724-781.5 (a) Myers A. L., Prausnitz J. M., AIChE J., 1965, 11, 121-127. (b) Bae Y., Mulfort K. L., Frost H., Ryan

P., Punnathanam S., Broadbelt, L. J., Hupp J. T., Snurr R. Q., Langmuir, 2008, 24, 8592-8598. (c)

Armatas G.S., Kanatzidis M.G., Nat. Mater, 2009, 8, 217-222.6 J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. ReV. Lett., 1996, 77, 3865.7 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010, 132, 154104.8 A. Schafer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994, 100, 5829.9 R. Ahlrichs, M. Bär, M. Häser, H. Horn, and C. Kölmel; Chem. Phys. Letters 1989, 162, 165.