Metal-Organic Framework Nodes as Nearly Ideal Supports for Molecular Catalysts: NU-1000- and UiO-66-Supported Iridium Complexes

Article abstract of DOI:10.1021/jacs.5b02956

Metal-organic frameworks with Zr₆ nodes, UiO-66 and NU-1000, were investigated as supports for Ir(CO)₂ and Ir(C₂H₄)₂ complexes. A single bonding site for the iridium is identified on the nodes of NU-1000, whereas two sites are identified on UiO-66, although at low iridium loadings only one site is occupied. Density functional theory calculations provide structural results that are in good agreement with infrared and X-ray absorption fine-structure spectra. The reactivity of node-supported Ir(CO)₂ with C₂H₄ and the catalytic activity and selectivity of the species initially present as Ir(C₂H₄)₂ for ethylene hydrogenation and dimerization were investigated both experimentally and computationally and shown to be strongly influenced by the node.

Full text of DOI:10.1021/jacs.5b02956

Article
pubs.acs.org/JACS
Metal−Organic Framework Nodes as Nearly Ideal Supports for
Molecular Catalysts: NU-1000- and UiO-66-Supported Iridium
Complexes
Dong Yang,† Samuel O. Odoh,‡ Timothy C. Wang,§ Omar K. Farha,§,∥ Joseph T. Hupp,§
Christopher J. Cramer,‡ Laura Gagliardi,*,‡ and Bruce C. Gates*,†
†Department of Chemical Engineering & Materials Science, University of California, Davis, California 95616, United States
‡Department of Chemistry, Chemical Theory Center, and Supercomputing Institute, University of Minnesota, Minneapolis,
Minnesota 55455-0431, United States
§Department of Chemistry, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States
∥Department of Chemistry, Faculty of Science, King Abdulaziz University, Jeddah 22254, Saudi Arabia
S
* Supporting Information
ABSTRACT: Metal−organic frameworks with Zr6 nodes, UiO-66 and NU-1000, were
investigated as supports for Ir(CO)2 and Ir(C2H4)2 complexes. A single bonding site for
the iridium is identified on the nodes of NU-1000, whereas two sites are identified on UiO-
66, although at low iridium loadings only one site is occupied. Density functional theory
calculations provide structural results that are in good agreement with infrared and X-ray
absorption fine-structure spectra. The reactivity of node-supported Ir(CO)2 with C2H4 and
the catalytic activity and selectivity of the species initially present as Ir(C2H4)2 for ethylene
hydrogenation and dimerization were investigated both experimentally and computation-
ally and shown to be strongly influenced by the node.
acetylacetonate to convert cyclohexene to benzene by oxidative
dehydrogenation; this catalyst was used at temperatures up to
350 °C, pointing to stability advantages of this class of MOF.
Mondloch et al.16 also reported the use of such MOF nodes as
supports for aluminum and zinc catalysts made by atomic layer
deposition of gas-phase catalyst precursors.
1. INTRODUCTION
Metal−organic frameworks (MOFs) are a class of crystalline
and porous materials formed by coordination bonds between
metal-containing nodes and organic linkers.1−3 MOFs that are
stable at high temperatures are appealing but still little-
investigated as catalyst supports,4−10 especially for trans-
formations of gas-phase reactants. Several investigators have
used the organic linkers of MOFs to anchor metal complexes
used as catalysts. For example, Manna et al.11 prepared MOFs
with Fe(II) and Co(II) salicylaldimine complexes as linkers,
finding them to be highly active for 1-octene hydrogenation. Fei
and Cohen12 modified the linkers of MOFs to form
thiocatechol-functionalized sites for palladium centers that
selectively catalyze the oxidative functionalization of aromatic
C−H bonds. Gonzalez et al.13 used an MOF with bipyridine
sites in the linkers that reacted with Ir(COD)2BF4 (COD is 1,5-
cyclooctadiene) to give a catalyst for arene C−H borylation.
These catalysts are all comparable to organic polymer-
supported catalysts.
Supported metal catalysts that are essentially molecular and
structurally well-defined offer numerous advantages for
fundamental understanding; such catalysts have been synthe-
sized on the surfaces of zeolites and in prospect can be
synthesized similarly on MOF nodes.17 We now report such
catalysts: MOF node-supported metal complexes that are
structurally well-defined, from the viewpoints of both the metal
species and the support. We selected Zr6-based UiO-66 and
NU-1000 frameworks because of their exceptional thermal
stability and well-defined structures.16,18 UiO-66 is composed
of [Zr6(μ3-O)4(μ3-OH)4]12+ nodes, each linked to 12
carboxylates of terephthalate ligands to form tetrahedral and
octahedral cages; however, there are many defects from missing
linkers in this MOF, with the resulting otherwise open metal
sites occupied by hydroxo ligands.18−20 The node topology of
NU-1000 has been determined to be [Zr6(μ3-O)4(μ3-
OH)4(OH)4(OH2)4]8+,21 and it is coordinated to eight
Alternatively, MOF nodes that resemble small pieces of
metal oxides have been used as catalyst supports that are
comparable to metal oxides and zeolites. Larabi and Quadrelli14
pointed out that the OH groups on such nodes provide sites for
binding of organometallic catalyst precursors. For example,
Nguyen et al.15 used defect sites in the zirconium oxide nodes
of UiO-66 as a support for a catalyst synthesized from vanadyl
Received: March 20, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.5b02956
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. IR spectra in (A) υOH region, (B) H-bonded H2O and −OH region and (C) υCO region characterizing bare NU-1000, a, and bare UiO-66,
b, (black traces) and the samples formed by adsorption of Ir(CO)2(acac) on them (red traces).
activated under vacuum (10−7 Torr) at 423 or 393 K, respectively, for
24 h before use) in a Schlenk flask and slurried in 40 mL of dried n-
tetratopic linkers of 1,3,6,8-tetrakis(p-benzoic-acid)pyrene
(H4TBAPy) to form triangular and hexagonal pores. We used
pentane (Fisher, 99%) at room temperature. After 4 h, the solvent was
Ir(CO)2(acac) (acac = acetylacetonate) and Ir(C2H4)2(acac) as
removed by evacuation for a day so that all the iridium remained in the
precursors to react with functional groups such as OH on the
MOF. The resultant solids, containing 10 or 1.0 wt % iridium, were
nodes of NU-1000 and UiO-66, accompanied by removal of the
stored in the argon-filled glovebox. The iridium loadings were inferred
acac ligands.17 The samples were characterized by infrared (IR)
from the conditions of the syntheses, whereby all of the added iridium
and extended X-ray absorption fine-structure (EXAFS)
remained in the MOF.
spectroscopies as well as density functional theory (DFT)
2.4. Infrared Spectroscopy. A Bruker IFS 66v/S spectrometer
with a spectral resolution of 2 cm−1 was used to collect transmission IR
spectra of power samples. Approximately 30 mg of solid sample inside
a glovebox was pressed into a thin wafer and loaded into a cell that
served as a flow reactor (In-situ Research Institute, Inc., South Bend,
IN). The cell was sealed and connected to a flow system that allowed
recording of spectra while the reactant gases flowed through the cell at
reaction temperature. Each spectrum is the average of 64 scans.
2.5. X-ray Absorption Spectroscopy. X-ray absorption spectra
were recorded at X-ray beamline 4-1 at the Stanford Synchrotron
Radiation Lightsource (SSRL). The storage ring electron energy and
ring current were 3 GeV and 300 mA, respectively. A double-crystal
Si(220) monochromator was detuned by 20−25% at the Ir LIII edge to
minimize the effects of higher harmonics in the X-ray beam.
calculations. Our results demonstrate the presence of bonding
between the Ir species and node sites and determine their
catalytic properties for conversion of ethylene in the presence
of H2.
2. EXPERIMENTAL SECTION
2.1. Synthesis of UiO-66. ZrCl4 (0.080 g, 0.343 mmol) and 0.7
mL of acetic acid (modulator) were dissolved in 20 mL of DMF in a
100 mL Teflon-capped glass jar by using ultrasound for 5 min. The
linker precursor benzene-1,4-dicarboxylic acid (0.057 g, 0.343 mmol)
was then added into the solution and dissolved by ultrasound applied
for about 5 min. The jars were kept in an oven at 393 K under static
conditions for 24 h. White precipitates were produced, and they were
isolated by centrifugation after cooling to room temperature. The
solids were washed with DMF three times to remove unreacted
precursors and with methanol six times to remove DMF. Then, the
powder was dried at room temperature.
2.2. Synthesis of NU-1000. A sample consisting of 97 mg of
ZrOCl2·8H2O (0.30 mmol) and 2700 mg (22 mmol) of benzoic acid
was mixed in 8 mL of DMF (in a 6 dram vial) and ultrasonically
dissolved. The clear solution was incubated in an oven at 353 K for 1
h. After cooling to room temperature, 40 mg (0.06 mmol) of
H4TBAPy was added to this solution, and the mixture was sonicated
for 20 min. The yellow suspension was held in an oven at 353 K for 24
h. After cooling to room temperature, yellow polycrystalline material
was isolated by filtration (35 mg of activated material, 54% yield) and
washed with DMF and subsequently activated with HCl. Then, the
solid was washed twice with DMF and six times with acetone. The
powder was dried at room temperature.
2.3. Synthesis of Ir(CO)2 and Ir(C2H4)2 Supported on MOFs.
Sample syntheses and handling were performed with the exclusion of
moisture and air by use of a double-mainfold Schlenk vacuum line and
an argon-atmosphere glovebox.
The precursor Ir(CO)2(acac), with 98% purity, was purchased from
Strem Chemicals, Inc. The precursor Ir(C2H4)2(acac) was synthesized
as described elsewhere;22 it has been characterized by X-ray diffraction
crystallography and 1H and 13C NMR, Raman, and IR spectroscopies.
The supported iridium complex was prepared by bringing the
precursor Ir(CO)2(acac) or Ir(C2H4)2(acac) (36.2 mg for a 10 wt %
iridium loading or 3.6 mg for a 1.0 wt % iridium loading) in contact
with 200 mg of activated MOF powder (UiO-66 or NU-1000 was
2.6. Catalytic Reaction Rates. Ethylene hydrogenation catalysis
was carried out in a conventional laboratory once-through tubular
plug-flow reactor at 298 K and 1 bar. The catalyst (10−50 mg) was
mixed with 10 g of inert, nonporous α-Al2O3 and loaded into the
reactor in the argon-filled glovebox. The feed partial pressures were
100 mbar of C2H4, 200 mbar of H2, and 700 mbar of He, with a total
flow rate of 100 mL (NTP)/min. Products were analyzed with an
online Agilent 6890 gas chromatograph. The ethylene conversions
were <5%, and the reactor was well approximated as differential,
determining reaction rates directly.
2.7. Computational Methods. Density functional theory (DFT)
calculations were performed on neutral cluster models with the M06-
L23 density functional as implemented in the Gaussian 09 software
package. In this software package, we used ultrafine grids (99 radial
nodes and 590 angular nodes for C, H, and O atoms, but 199 radial
nodes and 974 angular nodes for Zr and Ir atoms) for numerical
integrations. The def2-svp basis set was used for the C, H, and O
atoms, whereas the def2-tzvpp24−26 basis set and associated effective
core potential were employed for Zr and Ir atoms. The positions of all
atoms, with the exceptions of the atoms in the aryl organic linkers,
were optimized. Vibrational frequency analyses were carried out to
confirm the nature of the stationary points found after each geometry
optimization. The geometry optimizations and vibrational frequency
analyses were carried out for the samples in the gas phase as well as in
n-pentane. The polarizable continuum model (PCM27) was used to
account for solvation effects for the calculations carried out with the
samples in n-pentane.
B
DOI: 10.1021/jacs.5b02956
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The cluster models were constructed from periodic unit cells of
NU-1000 and UiO-66 optimized with the generalized-gradient
approximation PBE28 density functional while using plane-wave basis
sets within the projector-augmented wave approach as implemented in
the Vienna Ab Initio Simulation Package (VASP29−32). In these
periodic calculations, we used a plane-wave kinetic energy cutoff of 520
eV, a convergence criterion of 0.05 eV/Å for atomic positions, and an
automatically generated gamma-centered Monkhorst−Pack k-point
grid.
The procedure for generating the cluster models from the
optimized periodic structure of NU-1000 has been described.21
Briefly, the first coordination sphere of the Zr6-metalate node was
extracted from the periodic structure while truncating the TPAPy4−
linkers to benzoate groups. A two-step protocol was then used for
optimizing the resulting clusters. First, we optimized the positions of
the aromatic H atoms while fixing all other atoms. Second, the C6H5
units of the benzoate groups were held fixed, whereas the positions of
all other atoms were optimized. A similar procedure was used for UiO-
66. However, the linkers in its periodic model were truncated to
acetate groups to speed up the computations.
Table 1. EXAFS and Calculated DFT Structural Parameters
of Ir(CO)2 Supported on NU-1000 Zr6 Nodes
a
EXAFS
103 × Δσ2 (Å2)
DFT
shell
ΔE0 (eV)
N
R (Å)
R (Å)
Ir−Ot
5.4
14.6
14.5
6.8
−5.2
−5.9
7.4
2.0
2.1
2.1
1.9
2.0
2.05
1.97
2.99
3.13
3.55
2.11
1.83
2.98
3.39
3.72
Ir−CCO
Ir−OCO
Ir−Ob
Ir−Zr
−5.6
−0.7
14.2
a
Notation: Ot, O atoms of terminal −OH groups on Zr6 node; CCO
,
carbonyl carbon; OCO, carbonyl oxygen; Ob, O in bridging μ3-OH or
μ3−O group on Zr6 node; N, coordination number; R, distance
between absorber and backscatterer atoms; Δσ2, disorder term; ΔE0,
inner potential correction. Estimated EXAFS error bounds: N, 20%;
R, 0.02 Å; Δσ2, 20%; ΔE0, 20% (Errors characterizing the Ir−Zr
contribution are greater.); fit range: 3.87 < k (wave vector) < 13.61
Å−1; 0.5 < R < 4 Å; goodness of fit value = 3.5.
3. RESULTS AND DISCUSSION
At the high loading of 10 wt % iridium, we also detected
physisorbed Ir(CO)2(acac) on each MOF, as shown by the IR
bands at 2085 and 2010 cm−1 (Figure 1C), which are close in
frequency to those characterizing Ir(CO)2(acac) in n-hexane
solution.37 Because the pore aperture of UiO-66 (∼6 Å)18 is
barely greater than the narrowest dimension of the precursor
(∼4 × 6 Å2), we infer that the supported iridium species were
present preferentially near the external UiO-66 particle
surfaces.38,39 The 2010 cm−1 band is also evident in the
spectrum of the NU-1000-supported species, and it disappeared
after 8 h of contact of the precursor with this MOF, during
which time two sharp bands at 2082 and 2005 cm−1 (also
assigned to physisorbed Ir(CO)2(acac) Figure 1C) decreased in
intensity and then disappeared after 20 h (Figure S5). Because
these disappeared slowly, we suggest that they were indicative
of iridium-containing species on the internal NU-1000 surface
and that their removal was slowed by their diffusion in the
narrow pores.
3.1. Ir(CO)2 Complexes Supported on Nodes of NU-
1000 and of UiO-66. Ir(CO)2(acac) in n-pentane solution
reacted with NU-1000 nodes, removing the protons in
hydrogen-bonded H2O/OH groups,21 as shown by the
decreased intensities of the 2745 cm−1 band in the IR spectra
(Figure 1B) and forming chemisorbed species assigned on the
basis of new IR bands at 2066 and 1990 cm−1 as iridium gem-
dicarbonyls bonded to two ligands inferred to be oxygen atoms
also coordinated to the Zr6 node (Figure 2). The iridium
To determine the structures of the iridium-supporting sites,
we performed DFT calculations characterizing the structure of
iridium complexes on neutral cluster models of NU-1000. The
calculations for NU-1000 show that the most stable iridium
complexes are situated in the MOF channels, with the metal
atom replacing the proton in a terminal OH2 group that
otherwise bridges to an adjacent terminal OH group on the
node (Figure 2). This result is consistent with the experimental
observation showing a decreased intensity of the H-bonded
H2O and −OH band at 2745 cm−1 after reaction of
Ir(CO)2(acac) with NU-1000 (Figure 1B). To confirm the
assignment of the IR band at 2745 cm−1, we note that DFT
predicts that this vibration will be found at approximately 2851
and 2875 cm−1. Application of a scaling factor of 0.956 changes
these calculated values to 2726 and 2749 cm−1, in good
agreement with the experiment. The choice of this scaling
factor is discussed below.
The calculated structural parameters for this binding mode,
Figure 2, are compared with EXAFS data in Table 1.
Agreement is generally good, although the distances between
the Ir atoms and bridging O or Zr atoms of the node are
overestimated by 0.2−0.3 Å, consistent with results of
analogous combined DFT/EXAFS40,41 investigations. The
calculated υCO bands of this binding mode are 2157 and
2085 cm−1. Alecu et al.42 determined scale factors of 0.951−
0.958 for fundamental vibrational frequencies obtained with
Figure 2. Ir(CO)2 deposited onto the Zr6 nodes of NU-1000. (A)
Location of the iridium complex in a periodic crystal of NU-1000. (B)
Expanded view of the iridium complex at the Zr6 node. (C) Same site
in pristine NU-1000 prior to Ir(CO)2 complexation.
loading was 10.0 wt %, corresponding to nearly one Ir atom per
node. The υCO bands shifted to 2017 and 1943 cm−1 when the
sample was brought in contact with 13CO at room temperature,
confirming the identification (Figure S2 and Table S1,
Supporting Information). EXAFS parameters measured at the
Ir LIII edge confirm the structure, with the coordination
numbers and interatomic distances shown in Table 1. This
structure closely matches that of Ir(CO)2 on a zeolite.33−36
Ir(CO)2(acac) in n-pentane solution reacted similarly with
UiO-66 nodes, but, in contrast to NU-1000, UiO-66 reacted
with Ir(CO)2(acac) to give two chemisorbed species, identified
on the basis of IR spectroscopy, with a pair of υCO bands at
2074 and 1996 cm−1, and a band at 2066 cm−1. Both bands
characterizing non-hydrogen-bonded node OH groups, at 3781
and 3692 cm−1 (Figure 1A, Table S2), and hydrogen-bonded
node H2O/OH species, at 2745 cm−1 (Figure 1B), decreased in
intensity as a result of the chemisorption reaction, indicating
the participation of these groups, in this case associated with
site defects (see below). Correspondingly, the pore volumes of
the MOFs decreased (section S4, Supporting Information).
C
DOI: 10.1021/jacs.5b02956
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
various M06-L model chemistries; a scaling factor of 0.956
brings the calculated frequencies of the two υCO bands to 2062
and 1993 cm−1, both within 4 cm−1 of experiment, Figure 1C.
The calculated separation of the bands (69 cm−1 after scaling)
matches experiment (76 cm−1) to within 10%.
On UiO-66 (Figure 3A−C), there are three possible types of
defects that can be created by the loss of an organic linker
computed values (giving 2070 and 1998 cm−1 for site 1 and
2077 and 2011 cm−1 for site 2), the mean unsigned error
compared to experimental values is 8 cm−1, offering good
support for these structural assignments. We note that the
higher carbonyl stretching frequencies associated with site 2
suggest less back-donation from Ir into acceptor π*CO orbitals.
Consistent with this analysis, calculated partial atomic charges
on the Ir center are more positive in site 2 (Hirshfeld:43 0.29;
Mulliken:44 0.51) than in site 1 (Hirshfeld: 0.17; Mulliken:
0.36).
We emphasize that Ir(CO)2(acac) did not react with the
bridging μ3-OH groups on the Zr6 nodes, characterized by the
IR band at 3673 cm−1, as indicated by the lack of change of this
band, either on NU-1000 or UiO-66, after anchoring of the
iridium precursor (Figure 1A). DFT calculations confirm that
reaction of Ir(CO)2 groups with the bridging μ3-OH groups
found at defects in UiO-66 is endoergic by about 28 kcal/mol, a
value that is in marked contrast to the calculated reaction
energies for the adsorption of Ir(CO)2 on site 1 (−4.3 kcal/
mol) and site 2 (1.2 kcal/mol).
Ir(CO)2 supported on the MOFs, when treated with flowing
C2H4, underwent a limited ligand exchange reaction, sometimes
leading to replacement of one CO by C2H4 but only for site 2
in the UiO-66 complex. In that case, a new band at 2020 cm−1
was observed and assigned to Ir(CO)(C2H4) (Figure 4).
Figure 3. (A) Periodic structure showing the pore channels of UiO-66.
(B) Model cluster representing the Zr6 node. (C) Same model cluster
rotated 90 deg about the horizontal axis. (D) Ligand defect with
terminal −OH and OH2 groups replacing a missing linker. (E) Site 1
complexation of Ir(CO)2. (F) Site 2 complexation of Ir(CO)2 (with
“extra” μ3-OH group circled in black).
(section 8, Supporting Information). The most stable defect is
shown in Figure 3D and offers a binding site (site 1) similar to
that of NU-1000 (Figure 2C). The decrease of the hydrogen-
bonded H2O and −OH band at 2745 cm−1 (Figure 1B)
indicates the reaction of Ir(CO)2(acac) with site 1 on UiO-66.
After bonding of Ir(CO)2 to this defect, Figure 3E, we calculate
unscaled υCO bands of 2165 and 2090 cm−1, very similar to
those obtained for NU-1000 and corresponding to the lower-
frequency pair observed experimentally, Figure 1C. The
structural parameters of this complex are very similar to those
in Table 1 for NU-1000. At the M06-L level, formation of
Ir(CO)2 supported at site 1 from Ir(CO)2(acac) is calculated to
be exoergic by −4.3 kcal/mol.
After consideration of a number of possibilities and an
assessment of their energies, we assign the second site (site 2)
for adsorption of Ir(CO)2 on UiO-66 as shown in Figure 3F. In
this case, a proton has migrated from one coordinating OH
group to a μ3-O oxo group (creating an additional μ3-OH
group, Figure 3F). The above-mentioned non-hydrogen-
bonded OH bands at 3781 and 3692 cm−1 are calculated to
be the hydroxyl groups of site 2 (Table S2). They disappeared
after reaction of Ir(CO)2(acac) with these sites. These bands
were calculated as being at approximately 3947 and 3875 cm−1,
respectively, from our DFT calculations. Again, application of
the scaling factor brings good agreement with empirical data
(3773 and 3704 cm−1, respectively).
Figure 4. IR spectra in the υCO region characterizing in the sample
initially formed by the chemisorption of Ir(CO)2(acac) on NU-1000
(a) and UiO-66 (b) as initially present (dashed black traces) and after
exposure to flowing ethylene at 298 K and 1 bar for 20 min (red
traces).
3.2. Ir(C2H4)2 Complexes Supported on the Nodes of
NU-1000 and UiO-66. Supported Ir(C2H4)2 complexes were
prepared alternatively from reactions with pristine MOFs
employing precursor Ir(C2H4)2(acac) in a fashion analogous to
that for Ir(CO)2(acac). IR data (Figures S6 and S7, Supporting
Information) indicate that Ir(C2H4)2 also adsorbs onto the
nodes, again with single iridium complexes on NU-1000 and a
mixture on UiO-66. We sought to prepare supported catalysts
dominated by the species analogous to those characterized
above, rather than more complex mixtures, by using low
loadings of iridium on the supports (1 wt %). At this loading,
IR spectra (Figure 5) are consistent with the supports being
nearly ideal, and these well-defined samples were tested as
The unscaled calculated υCO bands for the resulting Ir(CO)2
complex at site 2 are 2173 and 2104 cm−1, corresponding to the
higher-frequency pair observed experimentally, Figure 1C.
When the scaling factor of 0.956 is applied to the four
D
DOI: 10.1021/jacs.5b02956
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 6. Calculated pathway for the dimerization of ethylene to 1-
butene at sites 1 and 2 for Ir(C2H4)2 on UiO-66 and NU-1000 at
M06-L level. (MOF nodes as counteranions are implicit.) More details
are available in section S8 of the Supporting Information.
Figure 5. IR spectra in the υCO region characterizing the sample
initially formed by the adsorption of Ir(C2H4)2(acac) (with iridium
loading 1%) on NU-1000 (a, red trace) and UiO-66 (b, black trace)
after exposure to flowing CO 10%/He at 298 K and 1 bar for 2 min.
hydrogenation was also investigated, and the results show that
hydrogenation is favored over dimerization because of the
lower calculated barriers (2−12 kcal/mol for hydrogenation,
section S8, Supporting Information, versus 8−19 kcal/mol for
dimerization, Figure 6) and the weak adsorption of ethane on
the iridium sites (relative to alkenes), which favors desorption
of this product, Table 2.
We probed the used catalysts (after 1 h of continuous
ethylene hydrogenation) with a pulse of CO introduced into
the feed stream. The results (Figure S8, Supporting
Information) indicate the formation of the iridium gem-
dicarbonyl bands essentially matching those of the fresh
catalyst, thereby demonstrating the stability of the mononuclear
iridium complexes under catalytic reaction conditions.
catalysts for ethylene conversion in a once-through flow reactor
in the presence of H2 at 300 K and 1 bar.
The supported Ir(C2H4)2 complexes were found to be
precatalysts for ethylene hydrogenation accompanied by slight
ethylene dimerization (Table 2), with the activity and selectivity
Table 2. Catalytic Activities of MOF Node-Supported
Iridium Complexes for Hydrogenation and Dimerization of
a
Ethylene in a Flow Reactor at 298 K and 1 bar
selectivity under differential conversion conditions
(mol %)
butene isomers
4. CONCLUSIONS
TOF
trans-2-
1-
cis-2-
The nodes of NU-1000 and UiO-66 MOFs are essentially
molecular catalyst supports upon which highly uniform catalytic
metal complexes can be chemisorbed. These are among the
simplest and best-defined of any supported metal species,
constituting, we infer, an emerging class of materials that may
offer broad opportunities for discovery of new catalysts guided
by theory. We may expect high selectivities associated with the
uniformity of the catalytic species and high stabilities when
combinations of metals and MOFs are appropriately chosen.
b
c
support
(s−1
)
ethane butane butene butene butene butene
NU-1000
UiO-66
0.010
0.017
99.5
98.5
0
0.5
1.1
0
0.5
0.8
0
0.4
0.2
0.1
a
There was no detectable reaction in the absence of the catalyst under
our conditions, and the MOF alone was catalytically inactive. Support
for catalyst initially incorporating node-anchored Ir(C2H4)2. Turnover
frequency determined from C2H4 conversions <5%; catalyst mass =
10−50 mg; feed partial pressures = 100 mbar C2H4, 200 mbar H2, and
700 mbar He; total flow rate = 100 mL (NTP)/min.
b
c
ASSOCIATED CONTENT
* Supporting Information
Experimental and computational details as reported in the text.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.5b02956.
■
both depending on which MOF was used as the support. As
noted above, the iridium center of site 2 is more electropositive
(or electrophilic) than that of site 1 in the Ir(CO)2 complexes.
This pattern extends to the corresponding Ir(C2H4)2, Ir(CO)-
(C2H4), Ir(CO)(C2H4)(H2), and Ir(C2H4)2(H2) species as
well. Thus, the nature of the MOF support affects the transition
states and activation energies for the various catalytic reactions.
In Figure 6, for instance, the calculated barriers for ethylene
dimerization catalysis via an intermediate metallacycle are lower
for site 2 of UiO-66 than for site 1. Insofar as the NU-1000
catalytic site is analogous to the UiO-66 site 1, Figure 6, this
result rationalizes the increased dimerization selectivity
observed for UiO-66 and is also consistent with IR data
suggesting that site 2 dominates the catalytic reactivity of UiO-
66.
S
AUTHOR INFORMATION
■
Corresponding Authors
*gagliard@umn.edu
*bcgates@ucdavis.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported as part of the Inorganometallic
Catalyst Design Center, an Energy Frontier Research Center
funded by the U.S. Department of Energy, Office of Science,
Basic Energy Sciences, under Award DE-SC0012702.
Competing reaction mechanisms for ethylene dimerization
were explored; the one reported in Figure 6 was found to be
energetically the most favorable. The pathway for ethylene
E
DOI: 10.1021/jacs.5b02956
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(39) Ogino, I.; Chen, C. Y.; Gates, B. C. Dalton Trans. 2010, 39,
8423.
(40) Delgado, M.; Santini, C. C.; Delbecq, F.; Baudouin, A.; De
Mallmann, A.; Prestipino, C.; Norsic, S.; Sautet, P.; Basset, J.-M. J.
Phys. Chem. C 2011, 115, 6757.
(41) Williams, L. A.; Guo, N.; Motta, A.; Delferro, M.; Fragala, I. L.;
Miller, J. T.; Marks, T. J. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 413.
(42) Alecu, I. M.; Zheng, J. J.; Zhao, Y.; Truhlar, D. G. J. Chem.
Theory Comput. 2010, 6, 2872.
(43) Hirshfeld, F. L. Theor. Chem. Acc. 1977, 44, 129.
(44) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833.
REFERENCES
■
(1) Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. Science
2013, 341, 974.
(2) Ferey, G. Chem. Soc. Rev. 2008, 37, 191.
(3) Horike, S.; Shimomura, S.; Kitagawa, S. Nat. Chem. 2009, 1, 695.
(4) Corma, A.; Garcia, H.; Llabres i Xamena, F. X. Chem. Rev. 2010,
110, 4606.
(5) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.;
Hupp, J. T. Chem. Soc. Rev. 2009, 38, 1450.
(6) Gascon, J.; Corma, A.; Kapteijn, F.; Llabres i Xamena, F. X. ACS
Catal. 2014, 4, 361.
(7) Farrusseng, D.; Aguado, S.; Pinel, C. Angew. Chem., Int. Ed. 2009,
48, 7502.
(8) Dhakshinamoorthy, A.; Alvaro, M.; Garcia, H. Catal. Sci. Technol.
2011, 1, 856.
(9) Ma, L. Q.; Abney, C.; Lin, W. B. Chem. Soc. Rev. 2009, 38, 1248.
(10) Yoon, M.; Srirambalaji, R.; Kim, K. Chem. Rev. 2012, 112, 1196.
(11) Manna, K.; Zhang, T.; Carboni, M.; Abney, C. W.; Lin, W. B. J.
Am. Chem. Soc. 2014, 136, 13182.
(12) Fei, H. H.; Cohen, S. M. J. Am. Chem. Soc. 2015, 137, 2191.
(13) Gonzalez, M. I.; Bloch, E. D.; Mason, J. A.; Teat, S. J.; Long, J.
R. Inorg. Chem. 2015, 54, 2995.
(14) Larabi, C.; Quadrelli, E. A. Eur. J. Inorg. Chem. 2012, 3014.
(15) Nguyen, H. G. T.; Schweitzer, N. M.; Chang, C. Y.; Drake, T.
L.; So, M. C.; Stair, P. C.; Farha, O. K.; Hupp, J. T.; Nguyen, S. T. ACS
Catal. 2014, 4, 2496.
(16) Mondloch, J. E.; Bury, W.; Fairen-Jimenez, D.; Kwon, S.;
DeMarco, E. J.; Weston, M. H.; Sarjeant, A. A.; Nguyen, S. T.; Stair, P.
C.; Snurr, R. Q.; Farha, O. K.; Hupp, J. T. J. Am. Chem. Soc. 2013, 135,
10294.
(17) Serna, P.; Gates, B. C. Acc. Chem. Res. 2014, 47, 2612.
(18) Cavka, J. H.; Jakobsen, S.; Olsbye, U.; Guillou, N.; Lamberti, C.;
Bordiga, S.; Lillerud, K. P. J. Am. Chem. Soc. 2008, 130, 13850.
(19) Cliffe, M. J.; Wan, W.; Zou, X. D.; Chater, P. A.; Kleppe, A. K.;
Tucker, M. G.; Wilhelm, H.; Funnell, N. P.; Coudert, F.-X.; Goodwin,
A. L. Nat. Commun. 2014, 5, 4176.
(20) Vermoortele, F.; Bueken, B.; Le Bars, G.; Van de Voorde, B.;
Vandichel, M.; Houthoofd, K.; Vimont, A.; Daturi, M.; Waroquier, M.;
Van Speybroeck, V.; Kirschhock, C.; De Vos, D. E. J. Am. Chem. Soc.
2013, 135, 11465.
(21) Planas, N.; Mondloch, J. E.; Tussupbayev, S.; Borycz, J.;
Gagliardi, L.; Hupp, J. T.; Farha, O. K.; Cramer, C. J. J. Phys. Chem.
Lett. 2014, 5, 3716.
(22) Bhirud, V. A.; Uzun, A.; Kletnieks, P. W.; Craciun, R.; Haw, J.
F.; Dixon, D. A.; Olmstead, M. M.; Gates, B. C. J. Organomet. Chem.
2007, 692, 2107.
(23) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101.
(24) Metz, B.; Stoll, H.; Dolg, M. J. Chem. Phys. 2000, 113, 2563.
(25) Weigend, F. Phys. Chem. Chem. Phys. 2006, 8, 1057.
(26) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297.
(27) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem.
2003, 24, 669.
(28) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77,
3865.
(29) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169.
(30) Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49, 14251.
(31) Kresse, G.; Hafner, J. Phys. Rev. B 1993, 47, 558.
(32) Kresse, G.; Furthmuller, J. Comp. Mater. Sci. 1996, 6, 15.
(33) Lu, J.; Serna, P.; Aydin, C.; Browning, N. D.; Gates, B. C. J. Am.
Chem. Soc. 2011, 133, 16186.
(34) Miessner, H. J. Am. Chem. Soc. 1994, 116, 11522.
(35) Uzun, A.; Bhirud, V. A.; Kletnieks, P. W.; Haw, J. F.; Gates, B. C.
J. Phys. Chem. C 2007, 111, 15064.
(36) Wu, H.; Chua, Y. S.; Krungleviciute, V.; Tyagi, M.; Chen, P.;
Yildirim, T.; Zhou, W. J. Am. Chem. Soc. 2013, 135, 10525.
(37) Bonati, F.; Ugo, R. J. Organomet. Chem. 1968, 11, 341.
(38) Lu, J.; Aydin, C.; Browning, N. D.; Gates, B. C. Langmuir 2012,
28, 12806.
F
DOI: 10.1021/jacs.5b02956
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX