Computationally Guided Discovery of a Catalytic ... - ACS Publications

Sep 15, 2016 - Aaron B. League,. †. Zhanyong Li,. ‡ ... Minnesota 55455-0431, United States. ‡. Department of ... Evanston, Illinois 60208, Unit...
2 downloads 0 Views 1MB Size
Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

Computationally-Guided Discovery of Catalytic CobaltDecorated Metal–Organic Framework for Ethylene Dimerization Varinia Bernales, Aaron B. League, Zhanyong Li, Neil M. Schweitzer, Aaron W. Peters, Rebecca K. Carlson, Joseph T. Hupp, Christopher J. Cramer, Omar K. Farha, and Laura Gagliardi J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b07362 • Publication Date (Web): 15 Sep 2016 Downloaded from http://pubs.acs.org on September 18, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Computationally-Guided Discovery of Catalytic CobaltDecorated Metal–Organic Framework for Ethylene Dimerization Varinia Bernales,1 Aaron B. League,1 Zhanyong Li,2 Neil M. Schweitzer,3 Aaron W. Peters,2 Rebecca K. Carlson,1 Joseph T. Hupp2, Christopher J. Cramer1, Omar K. Farha*2,4, and Laura Gagliardi*1 1. Department of Chemistry, Chemical Theory Center, and Supercomputing Institute, University of Minnesota, Minneapolis, Minnesota 55455-0431, United States 2. Department of Chemistry, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States 3. Department of Chemical and Biological Engineering, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States 4. Department of Chemistry, Faculty of Science, King Abdulaziz University, Jeddah 21589, Saudi Arabia

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 12

ABSTRACT: The catalytic performance of a cobalt(II) single-site catalyst supported on the zirconia-like nodes of the metal organic–framework (MOF) NU-1000 is herein characterized by quantum chemical methods and compared to an iso-structural analogue incorporating nickel(II) as the active transition metal. The mechanisms of atomic layer deposition in MOFs and of catalysis are examined using density functional theory. We compare the catalytic activity of Co and Ni installed on the zirconialike nodes for ethylene dimerization, considering three plausible pathways. Multiconfigurational wave function theory methods are employed to further characterize the electronic structures of key transition states and intermediates. Finally, we report confirmation of Co catalytic activity for ethylene dimerization from experiments that were prompted by the computational prediction.

2

ACS Paragon Plus Environment

Page 3 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

INTRODUCTION Metal–organic frameworks (MOFs) are porous materials with a well-defined crystalline structure formed by metal ions or clusters connected in three dimensions by multitopic organic linkers.1 Due to the wide variety of metal nodes and linkers, tunability and rational design can inform MOF synthesis.2-4 As a result, MOFs have found use in many processes, such as gas separation,5 gas storage,6-7 and catalysis,8-10 and recently they have emerged as highly promising heterogeneous supports for incorporating transition-metal-based catalysts.11-14 Heterogeneous catalysis in a MOF provides an opportunity to design single-site catalysts that remain isolated from one another (owing to the intervention of the MOF framework), and can resist loss of catalytic activity that might otherwise occur through sintering or rearrangement. In addition, detailed mechanistic studies are possible for single-site catalysts that are homogeneous in constitution.15-22 These well-defined catalytic sites can be obtained via post-synthetic modification (PSM)13, 23-25 or atomic layer deposition of metals in MOFs (AIM)19, 26-28. Recently, there has been substantial interest in the design of heterogeneous catalysts for selective ethylene dimerization.17, 23, 29 This is a key process in the formation of high purity 1-butene, a precursor in the production of linear low-density polyethylene which is the most common plastic in everyday life.30 Recently, two examples of solution-based post-synthetic MOF functionalization with Ni(II) as the metal have been shown to catalyze ethylene dimerization upon activation with diethyl aluminum chloride (Et 2 AlCl). The first case is the Fe-MIL-101 MOF, consisting of trimeric iron(III) octahedral clusters linked by 2-aminoterephthalate ligands,31 where the linkers have been functionalized to support the equivalent of a Ni(II) homogeneous catalyst reported by Canivet et al25, and the second example involves the anchoring of a small Ni(II) complex onto the inorganic Zr 6 node of the MOF NU-1000.11 NU-1000 node, [Zr 6 (µ 3 –O) 4 (µ 3 – The Zr 6 -based OH) 4 (OH) 4 (OH 2 ) 4 ]8+, is particularly suitable for the installation of transition metals using PSM and AIM methodologies because of the reactive –OH and –OH 2 groups that decorate the node (Figure 1). By linking such nodes with tetratopic 1,3,6,8-tetrakis(p-benzoate)pyrene linkers the resulting MOF present high pore volume, as well as high thermal and chemical stabilities—all important characteristics for a catalyst platform.26-28, 32-35 In Figure 1, we present the largest pore of NU-1000 associated with channels having a diameter of about 31 Å, facilitating the diffusion of reactants and products to and from reactive sites.11, 19 Motivated by the high stability, tunability, and large pores that the NU-1000 nodes offer, a series of metals have been installed including metals from group XIII, e.g., Al(III)27-28 and In(III),27 and transition metals (TMs) such as Fe(III),13 Co(II),26, 33, 36 Ni(II),11, 19, 33 Cu(II),33 Zn(II),28, 33 Zr(IV),24 and Ir(I)37-38 using both PSM and AIM techniques. Interestingly, Ni(II) ions deposited directly onto the NU1000 node by AIM, which we will hereafter designate as NiAIM+NU-1000, upon activation showed remarkable activity for both ethylene hydrogenation and dimerization.19 In a previous study, we reported a quantum chemical characterization of a possible mononuclear catalytic site and the details of its ethylene hydrogenation cycle.19 However, dimerization of ethylene catalyzed by Ni-AIM+NU-1000 has not yet been computationally studied.

Figure 1. Schematic representation of NU-1000 MOF topology and Zr 6 -based inorganic and organic building block (a) 1,3,6,8tetrakis(p-benzoic acid)pyrene linker, TBAPy4−; (b) 8+ [Zr 6 (μ 3 −O) 4 (μ 3 −OH) 4 (OH) 4 (OH 2 ) 4 ] node; (c) the Zr 6 -based framework of the NU-1000; Color code: Carbon is gray; hydrogen is white; zirconium is cyan; and oxygen is red.

In this work we report a detailed investigation of three plausible mechanisms for ethylene dimerization with a mononuclear Ni-AIM+NU-1000 catalyst based on density functional theory calculations. We compare this catalyst to an isostructural mononuclear Co-AIM+NU-1000 in order to understand how the coordination environment of the support influences the relative catalytic properties of the two transition metals.19, 36 Additionally, higher level multireference calculations on key transition states are presented to contribute to the rational design of new improved catalysts based on an understanding of key details of metal electronic structure. [As a point of nomenclature, note that depending on completeness of dosing the AIM experiment can create M-AIM+NU-1000 products that contain mixtures of mono-, bi-, and polynuclear metallated nodes. Theory unambiguously addresses specific nuclearities by construction of the computational model, and in this paper we will consider only mononuclear metallated nodes, noting that this does not preclude the possibility of additional catalytic activity from bi- and polynuclear sites in experimentally prepared samples.] MODELS AND COMPUTATIONAL PROCEDURE Cluster Model: All calculations were performed starting from the mix-S topology for the bare (undecorated) Zr 6 -NU1000 node, [Zr 6 (μ 3 –O) 4 (μ 3 –OH) 4 (OH) 4 (OH 2 ) 4 ]8+, as reported by Planas et al.34 The Zr 6 node of NU-1000 was modeled as a finite cluster extracted from an optimized periodic unit cell to allow a detailed investigation of the supported M2+ (M = Co and Ni) complexes. The organic linkers were modeled as benzoate (150-atom cluster) and formate (70-atom cluster) groups when studying deposition and dimerization mechanisms, respectively (Figure 2).34, 38 Additionally, we designed small clusters to perform high level theory calculations (for more details, see SI).

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 12

one nickel atom was deposited onto one face of the Zr 6 -node of NU-1000, thereby generating an isolated, single-site, mononuclear catalyst.33 The most energetically favorable products from the AIM process as determined by DFT are intermediates A and B (Figure 3; a similar computational procedure to that reported previously for following Ni(II) deposition was employed in this work for Co(II); See the SI for more details). These structures are pre-catalysts for ethylene dimerization. Figure 2. Schematic representation of (a) benzoate and (b) formate cluster model used to calculate AIM processes and dimerization mechanisms, respectively. Color code: Carbon is gray; hydrogen is white; zirconium is cyan; and oxygen is red.

Ethylene dimerization catalyzed via a single-site Co(II) and Ni(II) atom deposited at the NU-1000 node. Three plausible mechanisms for ethylene dimerization were investigated and are summarized in Figure 3. Cycle I involves the generation of an active catalyst by C–H bond activation of ethylene to form a vinyl-metal species (with proton transfer to a coordinated oxide or hydroxide). Subsequent generation of α-olefins by consecutive ethylene insertion steps is followed by a proteolytic release at some point (regenerating the catalyst, B). In cycle II, the potential formation of metallo-cyclic intermediates was also investigated, which begins with the binding of two ethylene molecules to B followed by generation of the new C–C bond through oxidative coupling.50-51 Finally, Cycle III, follows a Cossee-Arlman mechanism that involves the pre-activation of the M-AIM+NU-1000 catalyst by Et 2 AlCl, as previously reported for nickel functionalized MOFs acting as catalysts for ethylene dimerization.11, 19, 25, 52 Cycle I is a Philips-type reaction,53 and the calculated activation enthalpy, ∆H‡, for this step proceeding from intermediate A (i.e., with coordinated hydroxide that activates the ethylene C–H bond to form the vinyl group), is about 32 kcal/mol for Co-AIM+ and 27 kcal/mol for NiAIM+NU-1000.54 The alternative addition to intermediate B involves the µ 3 –O atom activating the ethylene C–H bond and activation enthalpies of 43 and 44 kcal/mol are computed for Co-AIM+NU-1000 and Ni-AIM+NU-1000, respectively. Although previously reported as a viable mechanism for Cr-atom catalysts deposited on silica,53 these ∆H‡ are too high to make Cycle I accessible under relevant reaction conditions. Cycle II involves a Ziegler-Natta metallocycle-like mechanism,50-51, 55 with two coordinated ethylene molecules reacting to form a metallocycle. However, the activation enthalpies for this oxidative cyclization are 35 and 42 kcal/mol for Co-AIM+ and Ni-AIM+NU-1000, respectively. As the metal oxidation state formally increases from 2 to 4 for this step, its overall poor kinetic favorability is unsurprising. Like Cycle I, without pre-activation of the catalyst, the reaction is unfavorable. More detailed descriptions of cycles I and II that include full characterization of all relevant stationary points are provided in the SI. In Cycle III (Scheme 1, Figure 4), alkylation of intermediate A with Et 2 AlCl generates tetracoordinate M(II) species 1, which readily coordinates ethylene in a π complex. The resulting intermediate 2 is lower in enthalpy than the separated reactants by 12 and 19 kcal/mol for Co and Ni, respectively.

Density Functional Calculations: Clusters were optimized using the M06-L39 density functional as implemented in the Gaussian 09 software package40, together with the def2-SVP basis set for C, H, and O atoms, and the def2-TZVPP41-42 basis set for Co, Ni, and Zr atoms. Associated effective core potentials were used on Zr atoms. In all cases, the positions of all atoms were optimized, except for the C and H atoms of the capping benzoate and formate groups. The nature of all stationary points was verified by analytical computation of vibrational frequencies, which were also used to compute 298.15 K thermochemical quantities. Multireference Calculations: The key reactive intermediates identified by density functional theory (DFT; vide infra) were further characterized at the complete active space selfconsistent field (CASSCF) level of theory, followed by second order perturbation theory (CASPT2) calculations.43 All CASSCF calculations were performed with the MOLCAS-7.8 package44 on the M06-L optimized geometries, using the all-electron ANO-RCC basis sets.45-46 In all of these calculations, the triple-zeta quality ANORCC-VTZP basis set was used for Co and Ni, the doublezeta quality ANO-RCC-VDZP basis set was used for O and Zr, and the minimal ANO-RCC-MB basis set was chosen for the C and H atoms. Scalar relativistic effects were included by using the Douglas–Kroll–Hess Hamiltonian.47-48 Twoelectron integral evaluation was simplified by employing the Cholesky decomposition technique.49 In the CASSCF/CASPT2 calculations, an active space of n electrons in m orbitals (n,m) was used for all species, where n is the number of 3d electrons for the mononuclear catalyst TM, together with the electrons in substrate C–C π bonds (two electrons per ethylene molecule). In order to reduce the computational cost, we designed a small cluster (See SI). These clusters included the first coordination sphere of the Co and Ni atom and two proximal Zr atoms connected as supports. Terminal hydrogen atoms were added to keep charge neutrality; the positions of the added hydrogen atoms were optimized using the same level of theory as for the larger cluster (for more details, see SI). RESULTS AND DISCUSSION Experimental evidence indicates that installation of Co(II) via AIM to produce Co-AIM+NU-1000 yields between one and four cobalt ions per Zr 6 -node, where the loading of Co can be modified by changing certain experimental conditions.26, 33, 36 In order to model the catalytic activity and compare with the nickel analogue, Ni-AIM+NU-1000, an idealized system was considered, where only one cobalt or

4

ACS Paragon Plus Environment

Page 5 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Insertion of ethylene into the existing metal-carbon bond forms 3 with an extended butyl chain. This process can repeat to oligomerize ethylene (vide infra). Alternatively, βhydride elimination can occur to generate 4, after which ethylene may substitute for the coordinated 1-butene (αolefin), leading to intermediate 5. Metal-hydride addition across the double bond regenerates 1, which may then reenter the dimerization cycle. Selected bond lengths of the first coordination sphere of the metal center in intermediate 1 are shown in Figure 5. These intermediates present similar geometries for the two metals. However, Co, compared to Ni, has shorter bonds with the terminal –OH groups, and longer bonds with the µ 3 –OH and C(ethyl) groups. The energetics associated with all of these steps are summarized in Figure 4 for both Ni-AIM+ and CoAIM+NU-1000. In each case, the rate-determining step is that for insertion, with activation enthalpies of 24.1 and 15.6 kcal/mol to go from 2 to 3 for Co and Ni, respectively. Unlike Cycles I and II, it is apparent that Ni-AIM+NU-1000 is predicted to be a considerably more effective catalyst than Co-AIM+NU-1000 following Cylcle III. [Note that the enthalpic separation between 3 and TS5-1 for Co-

Figure 3. Schematic representation of three mechanistic pathways for ethylene dimerization mediated by M-AIM+NU1000, where M = Co(II) or Ni(II) metal. Linkers of the MOF are not shown for clarity.

Scheme 1: Cycle III

1 2 3 4

Figure 4. Computed enthalpies (298.15 K, 1 atm) for stationary points along reaction coordinates for ethylene dimerization catalyzed by M-AIM+NU-1000 (M = Co and Ni). Note that the enthalpies recorded for 5 and TS5-1 include the liberated product (1-butene). Relative enthalpies and free energies calculated at 318 K, 1 atm are reported in the SI.56

5

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 12

Figure 5. Intermediate 1 for Co(II) (in blue) and Ni(II) (in green).

AIM+NU-1000 is 24.0 kcal/mol, and this would affect the turnover frequency for this catalyst to roughly the same extent as the C-C bond formation insertion step, but we will focus on the latter as the same electronic structure principles are at play in differentiating both insertion transition-state (TS) structures for Co vs Ni.] A decisive difference between the two catalysts is the electronic configuration of the transition metal: 3d7 for Co(II) and 3d8 for Ni(II). The Co(II) species can have either one or three unpaired electrons, giving rise to doublet and quartet states. The Ni(II) species, by contrast, must adopt singlet or triplet electronic states. This difference in electron structure has a profound influence on the geometries associated with the rate-determining TS structures, as shown in Figure 6. Co(II) has a penta-coordinated environment that is approximately trigonal bipyramidal: the metal interacts with three oxygen atoms of the node (one µ 3 –OH (apical) and two terminal –OH groups (equatorial)), the ethyl group (apical) and the reacting ethylene molecule (equatorial). By contrast, the Ni(II)-AIM+NU-1000 species is effectively square planar, lacking any significant interaction with the µ 3 –OH group and organizing the hydrocarbon groups trans to the remaining hydroxyls on the node.

Figure 6. Rate-determining TS structures TS2-3 for Co(II) (blue, quartet, trigonal bipyramidal) and Ni(II) (green, singlet, square planar).

structure through hybridization with the (C 2 H 4 -2p)2 MO (because a half-filled antibonding orbital combination is created in addition to a bonding interaction); the CASSCF frontier MO has only a 15% Co 3d orbital contribution. This, in conjunction with the coordination environment around the metal, are the key differences in the ability of the two metal centers to stabilize the C–C bond-forming TS structure explains the enhanced catalytic activity of the Ni (for further details of the overall electronic structures, see the SI).

Analysis of the Electronic Configuration by Multireference Methods. In order to rationalize the different reactivity between Co and Ni, the most relevant CASSCF molecular orbitals (MOs) for the rate-determining TS structure TS2-3 of cycle III are presented in Figure 7. The frontier (C 2 H 4 -2p)2 MO is a combination of the 2p orbitals of the reactive carbon atoms, i.e., the π bond of the ethylene, where one of the two p orbitals is being rehybridized into the forming C–C bond with the ethyl group. This orbital is stabilized through some degree of hybridization with the 3d orbitals of the coordinating metal. For the Ni case, there are four doubly occupied 3d orbitals. The fifth, formally empty d orbital, hybridizes strongly (48%) with the 2p orbitals of the reactive carbon atoms (52%) (Figure 7). By contrast, the high-spin configuration of Co leaves the appropriate d orbital halffilled and thus substantially less able to stabilize the TS

Figure 7. Electronic configuration of TS2-3 of Co(II)-AIM+ and Ni(II)-AIM+NU-1000, where Ni(II) better stabilizes the frontier MO through hybridization with an empty d orbital. The frontier MOs are shown with percentage d orbital amplitudes and CASSCF occupation numbers in parentheses.

Propagation versus Termination Steps. In order to assess the likely formation of longer α-olefin chains, we compared the relative energetics of propagation and termination. The propagation step involves coordination of an ethylene molecule to the butyl-metal intermediate 3 to generate 6 (Scheme 2), which is exothermic by 12.3 and 12.6 kcal/mol for Co and Ni, respectively. The subsequent insertion steps required for dimerization are computed to

6

ACS Paragon Plus Environment

Page 7 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

have activation enthalpies of 24.7 and 8.9 kcal/mol for Co and Ni, respectively (Table 1). These ∆H‡ values are similar to those for the competing β-hydride elimination reactions, 22.7 and 8.4 kcal/mol for Co and Ni, respectively. As propagation and termination are similar from a kinetic standpoint, it is clear that the distribution of α-olefins will be sensitive to the partial pressure of ethylene in the MOF interior. In addition, while we have not studied it here, it is likely that local accessible volume constraints imposed by the organic linkers in the MOF framework may significantly reduce the ability of ethylene to approach the open metal site as dimers of increasing length occupy that volume. Experimental Verification of Co-NU-1000 Catalytic Activity Motivated by the above computational results, which suggest that ethylene dimerization should be catalyzed by CoAIM+NU-1000, albeit with lower activity than that of NiAIM+NU-1000, we conducted experiments to assess its efficacy. As with Ni-AIM+NU-1000, Co-AIM+NU-1000 (both, in this case, prepared by atomic layer deposition (ALD) in MOFs (AIM)) can be transformed into its catalytically active form by the addition of Et 2 AlCl under air-

including 1-butene (~70%), trans-2-butene (~20%), and cis2-butene (~10%) (Figure 8). By varying the flow rates of ethylene, the TOF for this catalytic process was determined to be ~0.012 s-1 on a per-cobalt-atom basis, ~5 times lower than that with Ni-AIM+NU-1000 under the same experimental conditions (Figure 9.a). The apparent Arrhenius activation energy for this process is ~12.4 kcal/mol for Co-AIM+NU-1000 (Figure 9.b), which ~4 kcal/mol higher than that for Ni-AIM+NU-1000.19 The differences between the calculated (24.1 and 15.6 kcal/mol for Co and Ni, respectively) and experimental (11.8 and 7.9 kcal/mol, respectively) activation enthalpies (which differ from Arrhenius activation energies by a factor of RT) are likely attributable to the nature of the employed cluster models, which focus on a mononuclear reactive sites in order to understand relevant electronic structure details differentiating the metals. This model does not consider the influence of a second metal deposited on the same face of the NU-1000 node or on an adjacent face, although if such polynuclearity does not affect the local spin-state at each metal center, that same analysis as presented in Figure 7 will hold. We note that the effect of the alkyl-aluminum precursor (Et 2 AlCl) depositing directly onto the node, i.e., to generate Al(III)+NU-1000,27 has also been studied and been found not to affect the overall reactivity trend, i.e., NiAIM+NU-1000 remains more reactive than Co-AIM+NU1000 (see SI). Our mononuclear model successfully predicts, at low computational cost, the trend in catalytic ethylene dimerization activity subsequently confirmed by experiment. Theory further proved useful for rationalizing electronic structure effects on overall catalytic activity. As noted in the introduction, we have considered here only mononuclear catalytic sites/mechanisms, and we have determined that such sites should be competent for catalysis, within the reaction parameters described. Sites of higher nuclearity may also be catalytically competent — and potentially employ different mechanisms. We anticipate that ongoing work to compare the relative reactivities of metaloxo clusters of varying size supported on a common MOF will be informative and interesting..

Scheme 2. Propagation versus Termination Step.

Table 1. Computed enthalpies (298.15 K) for stationary points along propagation and termination step catalyzed by M-AIM+NU-1000 (M = Co and Ni). ∆Ha,b (kcal/mol) Catalyst: Co-AIM+NU-1000 Ni-AIM+NU-1000 Propagation Path -38.8 -41.8 6 -14.1 (24.7) -32.9 (8.9) TS6-7 -51.2 -51.7 7 Termination Path -26.5 -29.2 3 -3.8 (22.7) -20.8 (8.4) TS3-4 -14.5 -30.7 4 a 1 and separated reactants as zero of enthalpy (see Figure 4); b Activation enthalpies for elementary steps are shown in parentheses.

Figure 8. Product distribution of ethylene dimerization by activated Co-AIM+NU-1000 at different ethylene conversion.

CONCLUSIONS Ethylene dimerization can be catalyzed by Co-AIM+ and NiAIM+NU-1000, upon activation with Et 2 AlCl. In the Ni case, this reaction was originally suggested19 to proceed via the formation of an ethyl-nickel intermediate. We have now elucidated the full dimerization mechanism and compared it

and moisture-free conditions. (See SI for detailed experimental procedure.) During the first 10 hours on stream, there was a constant loss of the catalytic activity of the activated Co-AIM+NU-1000, similar to Ni-AIM+NU1000.19 The products were determined to be mainly butenes,

7

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

with non-activated Ni-AIM+NU-1000, which showed no catalytic activity. A Co-AIM+NU-1000 analogue was also examined computationally and predicted to be competent for the catalysis of ethylene dimerization, albeit with lower activity than Ni-AIM+NU-1000. This theoretical prediction was subsequently validated by experimental measurement. For both Co-AIM+ and Ni-AIM+NU-1000, ethylene insertion into the ethyl-metal bond is the rate-determining step, with calculated activation enthalpies of 24.1 and 15.6 kcal/mol for mononuclear Co-AIM+ and Ni-AIM+NU-1000, respectively. The greater activity of the Ni(II) active site is attributed to TS stabilization associated with an empty 3d orbital that hybridizes more readily with relevant carbon p orbitals in low-spin Ni-AIM+NU-1000 than in high-spin CoAIM+NU-1000. Nevertheless, based on computational and subsequent experimental results, we have demonstrated a first Co-AIM+NU-1000 supported catalyst that offers a promising route for the formation of linear products from olefins. While this new material has a higher activation enthalpy and lower TOF than that previously reported for Ni-AIM+NU-1000,19 it contributes to our growing understanding of the structure-function relationship between transition-metal electronic configuration and catalytic activity for such catalysts, and points to broader opportunities for theory-guided discovery of new catalysts.

Page 8 of 12

AUTHOR INFORMATION Corresponding Authors *E-mail: Omar K. Farha (O.K.F.): [email protected]. Phone: +1 847 467-4934 *E-mail: Laura Gagliardi (L. G.): [email protected]. Phone: +1 612 625-8299 Notes: The authors declare no competing financial interest

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

ACKNOWLEDGMENTS We thank Dr. Konstantinos D. Vogiatzis and Dr. Manuel A. Ortuño for helpful discussions. This work was supported as part of the Inorganometallic Catalyst Design Center, an EFRC funded by the DOE, Office of Basic Energy Sciences (DESC0012702).

REFERENCES 1.Zhou, H.-C.; Long, J. R.; Yaghi, O. M. Introduction to Metal– Organic Frameworks. Chem. Rev. 2012, 112, 673–674. 2.Wang, C.; Liu, D.; Lin, W. Metal–Organic Frameworks as A Tunable Platform for Designing Functional Molecular Materials. J. Am. Chem. Soc. 2013, 135, 13222–13234. 3.Wang, Z.; Cohen, S. M. Postsynthetic Modification of Metal– Organic Frameworks. Chem. Rev. 2009, 38, 1315–1329. 4.Cohen, S. M. Postsynthetic Methods for the Functionalization of Metal–Organic Frameworks. Chem. Rev. 2011, 112, 970– 1000. 5.Li, J.-R.; Sculley, J.; Zhou, H.-C. Metal–Organic Frameworks for Separations. Chem. Rev. 2012, 112, 869–932. 6.Sumida, K.; Rogow, D. L.; Mason, J. A.; McDonald, T. M.; Bloch, E. D.; Herm, Z. R.; Bae, T.; Long, J. R. Carbon Dioxide Capture in Metal–Organic Frameworks. Chem. Rev. 2012, 112, 724–781. 7.Rosi, N. L.; Eckert, J.; Eddaoudi, M.; Vodak, D. T.; Kim, J.; O'Keeffe, M.; Yaghi, O. M. Hydrogen Storage in Microporous Metal-Organic Frameworks. Science 2003, 300, 1127–1129. 8.Yoon, M.; Srirambalaji, R.; Kim, K. Homochiral Metal– Organic Frameworks for Asymmetric Heterogeneous Catalysis. Chem. Rev. 2012, 112, 1196–1231. 9.Xiao, D. J.; Bloch, E. D.; Mason, J. A.; Queen, W. L.; Hudson, M. R.; Planas, N.; Borycz, J.; Dzubak, A. L.; Verma, P.; Lee, K., et al. Oxidation of Ethane to Ethanol by N2O in a Metal–Organic Framework with Coordinatively Unsaturated Iron(II) Sites. Nature Chem. 2014, 6, 590–595. 10. Mlinar, A. N.; Keitz, B. K.; Gygi, D.; Bloch, E. D.; Long, J. R.; Bell, A. T. Selective Propene Oligomerization with Nickel(II)-Based Metal–Organic Frameworks. ACS Catal. 2014, 4, 717–721. 11. Madrahimov, S. T.; Gallagher, J. R.; Zhang, G.; Meinhart, Z.; Garibay, S. J.; Delferro, M.; Miller, J. T.; Farha, O. K.; Hupp, J. T.; Nguyen, S. T. Gas-Phase Dimerization of Ethylene under Mild Conditions Catalyzed by MOF Materials Containing (bpy)NiII Complexes. ACS Catal. 2015, 5, 6713–6718. 12. Pullen, S.; Fei, H.; Orthaber, A.; Cohen, S. M.; Ott, S. Enhanced Photochemical Hydrogen Production by a Molecular Diiron Catalyst Incorporated Into a Metal–Organic Framework. J. Am. Chem. Soc. 2013, 135, 16997–17003. 13. Beyzavi, M. H.; Vermeulen, N. A.; Howarth, A. J.; Tussupbayev, S.; League, A. B.; Schweitzer, N. M.; Gallagher,

ASSOCIATED CONTENT Supporting Information The Supporting Information, including full experimental and computational methods, details of Cycles I and II as characterized by theory, and cartesian coordinates and electronic energies for all computational species, is available free of charge on the ACS Publications website.

Figure 9. Catalytic conversion of ethylene dimerization by activated Co-AIM+NU-1000. a, TOF of ethylene dimerization. b, Arrhenius plot of ethylene dimerization reaction rates.

8

ACS Paragon Plus Environment

Page 9 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

J. R.; Platero-Prats, A. E.; Hafezi, N.; Sarjeant, A. A. et al., A Hafnium-Based Metal–Organic Framework as a Nature-Inspired Tandem Reaction Catalyst. J. Am. Chem. Soc. 2015, 137, 13624–13631. 14. Cohen, S. M.; Zhang, Z.; Boissonnault, J. A. Toward “metalloMOFzymes”: Metal–Organic Frameworks with SingleSite Metal Catalysts for Small-Molecule Transformations. Inorg. Chem. 2016, 55, 7281–7290. 15. Burgess, S. A.; Kassie, A.; Baranowski, S. A.; Fritzsching, K. J.; Schmidt-Rohr, K.; Brown, C. M.; Wade, C. R. Improved Catalytic Activity and Stability of a Palladium Pincer Complex by Incorporation into a Metal–Organic Framework. J. Am. Chem. Soc. 2016, 138, 1780–1783. 16. Dau, P. V.; Cohen, S. M. A Bifunctional, Site-Isolated Metal–Organic Framework-Based Tandem Catalyst. Inorg. Chem. 2015, 54, 3134–3138. 17. Metzger, E. D.; Brozek, C. K.; Comito, R. J.; Dincă, M. Selective Dimerization of Ethylene to 1-Butene with a Porous Catalyst. ACS Cent. Sci. 2016, 2, 148–153. 18. Dau, P. V.; Kim, M.; Cohen, S. M. Site-selective Cyclometalation of a Metal–Organic Framework. Chem. Sci. 2013, 4, 601–605. 19. Li, Z.; Schweitzer, N. M.; League, A. B.; Bernales, V.; Peters, A. W.; Getsoian, A. B.; Wang, T. C.; Miller, J. T.; Vjunov, A.; Fulton, J. L. et al. Sintering-Resistant Single-Site Nickel Catalyst Supported by Metal-Organic Framework. J. Am. Chem. Soc. 2016, 138, 1977-82. 20. Denysenko, D.; Jelic, J.; Reuter, K.; Volkmer, D., Postsynthetic Metal and Ligand Exchange in MFU-4l: A Screening Approach Toward Functional Metal–Organic Frameworks Comprising Single-Site Active Centers. Chem. Eur. J. 2015, 21, 8188–8199. 21. Fei, H.; Cohen, S. M., Metalation of a ThiocatecholFunctionalized Zr(IV)-Based Metal–Organic Framework for Selective C–H Functionalization. J. Am. Chem. Soc. 2015, 137, 2191–2194. 22. Fei, H.; Shin, J.; Meng, Y. S.; Adelhardt, M.; Sutter, J.; Meyer, K.; Cohen, S. M. Reusable Oxidation Catalysis Using Metal-Monocatecholato Species in a Robust Metal–Organic Framework. J. Am. Chem. Soc. 2014, 136, 4965–4973. 23. Liu, B.; Jie, S.; Bu, Z.; Li, B.-G. Postsynthetic Modification of Mixed-Linker Metal–Organic Frameworks for Ethylene Oligomerization. RSC Adv. 2014, 4, 62343–62346. 24. Klet, R. C.; Tussupbayev, S.; Borycz, J.; Gallagher, J. R.; Stalzer, M. M.; Miller, J. T.; Gagliardi, L.; Hupp, J. T.; Marks, T. J.; Cramer, C. J. et al. Single-Site Organozirconium Catalyst Embedded in a Metal–Organic Framework. J. Am. Chem. Soc. 2015, 137, 15680–15683. 25. Canivet, J.; Aguado, S.; Schuurman, Y.; Farrusseng, D. MOF-Supported Selective Ethylene Dimerization Single-Site Catalysts through One-Pot Postsynthetic Modification. J. Am. Chem. Soc. 2013, 135, 4195–4198. 26. Kung, C. W.; Mondloch, J. E.; Wang, T. C.; Bury, W.; Hoffeditz, W.; Klahr, B. M.; Klet, R. C.; Pellin, M. J.; Farha, O. K.; Hupp, J. T. Metal–Organic Framework Thin Films as Platforms for Atomic Layer Deposition of Cobalt Ions To Enable Electrocatalytic Water Oxidation. ACS Appl. Mater. Interfaces 2015, 7, 28223–28230. 27. Kim, I. S.; Borycz, J.; Platero-Prats, A. E.; Tussupbayev, S.; Wang, T. C.; Farha, O. K.; Hupp, J. T.; Gagliardi, L.; Chapman, K. W.; Cramer, C. J. et al. Targeted Single-Site MOF Node Modification: Trivalent Metal Loading via Atomic Layer Deposition. Chem. Mater. 2015, 27, 4772–4778. 28. 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. et al. Vapor-Phase Metalation by Atomic Layer Deposition in a Metal–Organic Framework. J. Am. Chem. Soc. 2013, 135, 10294–10297.

29. Finiels, A.; Fajulaa, F.; Hulea, V. Nickel-Based Solid Catalysts for Ethylene Oligomerization – a review. Catal. Sci. Technol. 2014, 4, 2412–2426. 30. Stürzel, M.; Mihan, S.; Mülhaupt, R. From Multisite Polymerization Catalysis to Sustainable Materials and AllPolyolefin Composites. Chem. Rev. 2016, 116, 1398–1433. 31. Bauer, S.; Serre, C.; Devic, T.; Horcajada, P.; Marrot, J.; Ferey, G.; Stock, N. High-Throughput Assisted Rationalization of the Formation of Metal–Organic Frameworks in the Iron(III) Aminoterephthalate Solvothermal System. Inorg. Chem. 2008, 47, 7568–7576. 32. Yuan, S.; Chen, Y.-P.; Qin, J.; Lu, W.; Wang, X.; Zhang, Q.; Bosch, M.; Liu, T.-F.; Lian, X.; Zhou, H.-C. Cooperative Cluster Metalationand Ligand Migration in Zirconium Metal– Organic Frameworks. Angew. Chem. Int. Ed. 2015, 54, 14696– 14700. 33. Klet, R. C.; Wang, T. C.; Fernandez, L. E.; Truhlar, D. G.; Hupp, J. T.; Farha, O. K. Synthetic Access to Atomically Dispersed Metals in Metal–Organic Frameworks via a Combined Atomic-Layer-Deposition-in-MOF and MetalExchange Approach. Chem. Mater. 2016, 28, 1213–1219. 34. Planas, N.; Mondloch, J. E.; Tussupbayev, S.; Borycz, J.; Gagliardi, L.; Hupp, J. T.; Farha, O. K.; Cramer, C. J. Defining the Proton Topology of the Zr6-Based Metal–Organic Framework NU-1000. J. Phys. Chem. Lett. 2014, 5, 3716–3723. 35. Howarth, A. J.; Liu, Y.; Li, P.; Li, Z.; Wang, T. C.; Hupp, J. T.; Farha, O. K. Chemical, Thermal and Mechanical Stabilities of Metal–organic Frameworks. Nature Reviews Materials 2016, 2016, 15018. 36. Peters, A. W.; Li, Z.; Farha, O. K.; Hupp, J. T. Atomically Precise Growth of Catalytically Active Cobalt Sulfide on Flat Surfaces and within a Metal–Organic Framework via Atomic Layer Deposition. ACS Nano, 2015, 9, 8484–8490. 37. Yang, D.; Odoh, S. O.; Wang, T. C.; Farha, O. K.; Hupp, J. T.; Cramer, C. J.; Gagliardi, L.; C., G. B. Metal–Organic Framework Nodes as Nearly Ideal Supports for Molecular Catalysts: NU-1000- and UiO-66-Supported Iridium Complexes. J. Am. Chem. Soc. 2015, 137, 7391–7396. 38. Yang, D.; Odoh, S. O.; Borycz, J.; Wang, T. C.; Farha, O. K.; Hupp, J. T.; Cramer, C. J.; Gagliardi, L.; Gates, B. C. Tuning Zr6 Metal–Organic Framework (MOF) Nodes as Catalyst Supports: Site Densities and Electron-Donor Properties Influence Molecular Iridium Complexes as Ethylene Conversion Catalysts. ACS Catal. 2015, 6, 235–247. 39. Zhao, Y.; Truhlar, D. G. A New Local Density Functional for Main-Group Thermochemistry, Transition Metal Bonding, Thermochemical Kinetics, and Noncovalent Interactions. J. Chem. Phys. 2006, 125. 40. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al. Gaussian, 09; Gaussian, Inc.: Wallingford CT, 2009. 41. Weigend, F. Accurate Coulomb-fitting basis sets for H to Rn. Phys. Chem. Chem. Phys. 2006, 8, , 1057– 1065. 42. Weigend, F.; Ahlrichs, R. Balanced Basis Sets of Split Valence, Triple Zeta Valence and Quadruple Zeta Valence Quality for H to Rn: Design and Assessment of Accuracy. Phys. Chem. Chem. Phys. 2005, 7, 3297– 3305. 43. Roos, B. O.; Taylor, P. R.; Siegbahn, P. E. M. Chem. Phys. Lett. 1980, 48, 157. 44. Aquilante, F.; De Vico, L.; Ferré, N.; Ghigo, G.; Malmqvist, P.-Å.; Neogrády, P.; Pedersen, T. B.; Pitoňák, M.; Reiher, M.; Roos, B. O. et al. MOLCAS 7: The Next Generation. J. Comput. Chem. 2010, 31, 224–247. 45. Roos, B. O.; Lindh, R.; Malmqvist, P.-Å.; Veryazov, V.; Widmark, P.-O. Main Group Atoms and Dimers Studied with a New Relativistic ANO Basis Set. J. Phys. Chem. A 2004, 108, 2851–2858.

9

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

46. Roos, B. O.; Lindh, R.; Malmqvist, P.-Å.; Veryazov, V.; Widmark, P.-O. New Relativistic ANO Basis Sets for Transition Metal Atoms. J. Phys. Chem. A 2005, 109, 6575–6579. 47. Hess, B. A. Relativistic Electronic-Structure Calculations Employing a Two-Component No-Pair Formalism with External-Field Projection Operators. Phys. Rev. A 1986, 33, 3742–3748. 48. Douglas, M.; Kroll, N. M. Quantum Electrodynamical Corrections to the Fine Structure of Helium. Ann. Phys. 1974, 82, 89–155. 49. Aquilante, F.; Pedersen, T. B.; Lindh, R. Low-cost Evaluation of the Exchange Fock Matrix from Cholesky and Density Fitting Representations of the Electron Repulsion Integrals. J. Chem. Phys. 2007, 126, 194106. 50. Agapie, T.; Schofer, S. J.; Labinger, J. A.; Bercaw, J. E. Mechanistic Studies of the Ethylene Trimerization Reaction with Chromium−Diphosphine Catalysts:  Experimental Evidence for a Mechanism Involving Metallacyclic Intermediates. J. Am. Chem. Soc. 2004, 126, 1304–1305. 51. Fan, L.; Krzywicki, A.; Somogyvari, A.; Ziegler, T., Theoretical Study of Ethylene Oligomerization by an Organometallic Nickel Catalyst. Inorg. Chem. 1996, 35, 4003– 4006. 52. Arlman, E. J.; Cossee, P. Ziegler-Natta Catalysis III. Stereospecific Polymerization of Propene with the Catalyst System TiCl3-AlEt3. J. Catal. 1964, 3, 99–104.

Page 10 of 12

53. Delley, M. F.; Núñez-Zarur, F.; Conley, M. P.; ComasVives, A.; Siddiqi, G.; Norsic, S.; Monteil, V.; Safonova, O. V.; Copéret, C. Proton Transfers Are Key Elementary Steps in Ethylene Polymerization on Isolated Chromium(III) Silicates. PNAS 2014, 111, 11624–11629. 54. Note that the we focus on enthalpies in most of our energetic discussion, as the corresponding free energies will depend on the partial pressures of reactant ethylene and product olefins. Our goal is to understand intrinsic bond-making/bond-breaking barriers, noting that entropic effects associated with reactant and product concentrations may be exploited under favorable conditions. 55. Zheng, F.; Sivaramakrishna, A.; Moss, J. R. Thermal Studies on Metallacycloalkanes. Coord. Chem. Rev. 2007, 215, 2056– 2071. 56. Note that the free energies and enthalpies calculated at 318 K (experimental conditions) changed due to thermal correction by less than 2.0 kcal/mol overall the reaction intermediates and transition states without changing the activation barriers.

10

ACS Paragon Plus Environment

Page 11 of 12

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table of Contents Entry:

11

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

M+N

hylene (2 t E ) d 3 ( 0 0 U -1 0

Page 12 of 12

Co(II)

p) 3d orbitals

2e2p sigma dona

tion

Co Ethylene

VS

Ni

ACS Paragon Plus Environment

1-Butene

Ni(II)