A Silica Supported Molecular Palladium Catalyst for Selective

Subscriber access provided by Nottingham Trent University

Article

A Silica Supported Molecular Palladium Catalyst for Selective Hydrodeoxygenation of Aromatic Compounds Under Mild Conditions Nicholas A DeLucia, Amy M. Jystad, Katherine Vander Laan, John Meynard M Tengco, Marco Caricato, and Aaron K. Vannucci ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.9b02460 • Publication Date (Web): 23 Aug 2019 Downloaded from pubs.acs.org on August 24, 2019

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 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 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.

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 23 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

ACS Catalysis

A Silica Supported Molecular Palladium Catalyst for Selective Hydrodeoxygenation of Aromatic Compounds Under Mild Conditions Nicholas A. DeLucia,† Amy Jystad,‡ Katherine Vander Laan,‡ John Meynard M. Tengco,§ Marco Caricato,*‡ and Aaron K. Vannucci*† †University

of South Carolina, Department of Chemistry and Biochemistry, Columbia, SC USA 29208

§University

of South Carolina, Department of Chemical Engineering, Columbia, SC USA 29208

‡University

of Kansas, Department of Chemistry, Lawrence, KS USA 66045

*corresponding author email: [email protected], [email protected] Keywords: Biomass, lignin, heterogeneous, deoxygenation, DFT, interaction energy. Abstract The molecular complex, chloro(2,2':6',2''-terpyridine-4'-carboxylic acid)palladium(II) chloride, was synthesized and was attached to the surface of amorphous silicon dioxide to generate a molecular/heterogeneous catalyst motif. This catalytic system exhibited excellent selectivity (>99 %) for hydrodeoxygenation of oxygenated aromatics under mild reaction conditions. A kinetic analysis showed that this molecular/heterogeneous catalyst was an order of magnitude more active than analogous homogeneous catalysts. Characterization techniques such as XRD and solid-state NMR, in conjunction with ICP-MS, indicate that the molecular catalyst is present on the surface of SiO2 and the formation of unwanted metallic Pd nanoparticles can be avoided. Computational modeling shows the catalysts can adhere to the oxide surface through a hydrogen bonding interaction, via a Coulombic attraction between the charged molecule and the oxide surface, or through covalent bonding. Post reaction analysis of the surface-modified oxide catalysts confirmed prolonged molecular integrity of the catalyst and sustained binding of the catalyst to the oxide surface when nonpolar solvents were employed for reactions. These surface-attached molecular catalysts thus were recycled through multiple catalytic reactions. Introduction Biomass is the largest possible renewable carbon source available. Composed of cellulose, hemicellulose, and lignin, terrestrial lignocellulosic biomass is also highly oxygenated.1 The cellulosic component of biomass is both an agricultural feedstock, and an industrially relevant source of renewable fuel additives.2 Lignin, however, is both not an edible component of biomass and is typically treated as waste and burned for energy during cellulose processing.3 Yet, obtaining higher value from lignin, such as deriving chemical feedstocks and fuel, is being recognized as integral to the economic feasibility of biorefineries.4-5 Lignin is a polymer containing aromatic subunits, thus, could be a viable source for commodity chemicals and liquid fuels.3 The first step in upgrading lignin to fuel is the depolymerization of lignin to form oxygenated aromatic monomers. Recent advances, such as “lignin first” processes have greatly increased the efficiency of biomass processing and lignin depolymerization into stable, oxygenated monomeric units.6-7 There is, therefore an increased need for catalysts that target selective conversion of lignin-derived aromatics. Deoxygenation of the subsequent monomeric aromatic units would then increase the energy density of the resulting liquid fuel,8 or lead to the isolation of important industrial chemical feedstocks.5 Selectively deoxygenating lignin derived compounds without hydrogenation of the aromatic units is of specific interest because aromatics and alkenes are higher value chemicals compared to alkanes, oxygen would be removed as water, the hydrogen use efficiency would be maximized, and carbon loss would be minimized, which prevents char and catalyst deactivation.9

ACS Paragon Plus Environment

ACS Catalysis 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

Numerous catalytic studies have examined deoxygenation of aromatic compounds through the use of metallic nanoparticle heterogeneous catalytic systems.2 Heterogeneous catalysts are commonly employed in large scale industrial settings, such as steam reforming and hydrocarbon cracking, due to the ease of separation and robustness of the catalysts.10-11 Heterogeneous catalysts, however, tend to lack product selectivity when compared to homogenous catalysts.12-13 Many studies have focused on the deoxygenation of model substrates to examine the efficiency and selectivity of hydrodeoxygenation (HDO) catalysts. Addition of Zn to the reaction solution containing Pd nanoparticles on a carbon support has been shown to increase the selective deoxygenation of the benzylic position of vanillyl alcohol.14 Controlling the hydrophilicity of the catalyst support and surface can also increase the catalytic selectivity of benzylic deoxygenation.15-16 Ru/Nb particles supported on SiO2 have shown the ability to catalytically convert pcresol to toluene with selectivities as high as 85%.17-18 Direct deoxygenation of guaiacol to benzene and pcresol to toluene with good selectivities have also been observed for various nanoparticle catalysts.19-21 Carbon coated Pt(111) samples or rutile TiO2 (110), when properly prepared, exhibited high selectivities for the deoxygenation of benzyl alcohol to toluene. Atomically dispersed cobalt on MoS2 nanomaterials was recently able to achieve selective deoxygenation of phenolics without observed ring hydrogenation.22 Another approach to avoiding ring hydrogenation is through the use of molecular catalysts. Molecular catalysts can be highly selective, and product selectivity can be tuned through well-known synthetic modifications of the catalyst structure.23-24 Activities of molecular catalysts also tend to be high.2526 Utilizing molecular catalysts in large scale industrial reactions, however, is commonly limited by the robustness of the catalysts and the difficulty of post reaction separations of a homogeneous catalyst solution. The use of homogeneous catalysts for the deoxygenation of organic compounds has been explored, as well as probing the activation of C–O bonds in homogeneous systems.27-31 Recently, we have reported the homogeneous molecular (2,2':6',2''-terpyridine)palladium(II) catalyst is capable of low temperature, selective HDO of model lignin monomers, specifically benzylic substrates.32 The catalyst showed complete selectivity for deoxygenated products over ring hydrogenated products at 100 °C, however, with moderate kinetic activity. Catalyst recyclability was also difficult due to the need to separate the homogeneous catalyst from the reaction mixture through column chromatography.

Figure 1. Structure of chloro(2,2':6',2''terpyridine-4'-carboxylic acid)palladium(II) chloride (1), and a graphical representation of the molecular catalyst attached to a SiO2 support (1-SiO2). Combining the positive aspects of homogeneous selectivity and heterogeneous robustness and ease of post reaction separations could lead to the development of ideal catalysts for selective deoxygenation reactions. We have thus designed a molecular/heterogeneous catalyst for the selective, catalytic deoxygenation reactions. The catalyst is modeled after our reported molecular deoxygenation catalyst and


Page 2 of 23

Page 3 of 23 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

ACS Catalysis

utilizes an inexpensive and inert high surface area SiO2 support. The molecular catalyst utilized for this study is shown in Figure 1 along with a representation of the molecular/heterogeneous catalyst 1-SiO2. The attachment of molecular complexes to the surface of metal oxide supports affords a catalyst structure that combines the positive aspects of a homogeneous catalyst with the desirable components of heterogeneous catalytic systems, such as high stability, recyclability, and expanded solvent scope. The general approach for modifying oxide surfaces with molecular catalysts is through “binding groups”. These binding groups are typically acid functionalities, such as carboxylate, phosphonate, and hydroxymate, covalently bound to polypyridyl ligands.33-34 The binding-group-polypyridyl-ligand unit then bonds to transition metals to generate molecular catalysts that can be attached to the surface of metal oxide particles. This attachment generates a molecular/heterogeneous catalyst with the activity and selectivity of a homogeneous molecular catalyst. Multiple reports have shown that molecular/heterogeneous catalysts are stable under a variety of conditions, and that molecular catalysts can remain bound to oxide surfaces for extended periods of time.35-37 Two critical reviews, however, have shown that many catalysts of this motif can decompose to either homogeneous catalysts in solution or metal nanoparticles.38-39 As such, extensive and thorough characterizations of molecular/heterogeneous catalysts both pre- and post-reaction are needed to identify the active catalytic species. In this work, 1-SiO2 was examined for the selective HDO of oxygenated aromatics. Using a molecular/heterogeneous catalyst introduces a new catalytic motif to the field of catalytic hydrodeoxygenation. This molecular/heterogeneous catalyst motif has been characterized both pre- and post-reaction to show the molecular catalyst is attached to the oxide support and remains attached throughout multiple catalytic cycles. A full kinetic analysis of 1-SiO2 was performed to determine the rate law for the catalytic deoxygenation of benzyl alcohol to toluene. Kinetic analysis revealed first order catalysis with respect to Pd as well as benzyl alcohol and that the kinetic activity of 1-SiO2 is comparable to or greater than multiple reports for related HDO transformations. Furthermore, computational modeling has been used to explore the possible geometries of the catalyst on the surface and quantify the binding energies between the carboxylate linker groups and the SiO2 support. Results Catalyst Preparation The synthesis of the molecular catalyst 1 and attachment to SiO2 to generate the molecular/heterogeneous catalyst 1-SiO2 was supported by a variety of characteristic techniques. To attach the catalyst to the A300 SiO2 particles, a 1.0 mM solution of 1 in DMF was prepared. Two grams of the A300 SiO2 particles were added to the 1.0 mM solution and allowed to soak overnight. The amount of the molecular catalyst in the loading solution is in great excess compared to the amount of SiO2 particles to ensure the loading solution does not limit the amount of molecular catalyst that attaches to the solid support. After the overnight soak, the particles were filtered and rinsed with cold methanol. The resulting oxide solid was no longer white, but instead a faint shade of yellow, indicative of the molecular complex 1 attached to the SiO2 support (Figure S1). To determine the amount of palladium present in a sample of 1-SiO2, 0.01g of the solid catalyst was digested with 4 mL aqua regia at 180°C (heat block) for 5 hours. The digestate was brought to ~10 g before ICP-MS analysis. ICP-MS analysis showed that 1-SiO2 contained 2.1 wt% Pd. This equates to 6.6x10-7 mols of catalyst per m2 of silica support. The preparation of 1-SiO2 is direct and does not expose the catalyst to any increase in temperature or reductive/oxidative conditions, thus it is reasonable to assume all the palladium detected using ICP-MS is present in the molecular catalyst form shown as 1-SiO2 in Figure 1. To support this assumption, the catalyst was further characterized using solid-state NMR and powder XRD. Evidence for the tpy ligand


ACS Catalysis 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

being attached to the SiO2 surface can be seen via solid-state 13C NMR, in which peaks corresponding to the 2,2':6',2''-terpyridine-4'-carboxylic acid ligand can be observed (Figure S2). X-ray diffraction patterns of 1-SiO2 on A300, compared with fresh A300, showed no presence of metallic palladium in either of the samples (Figure S3), noting that the detection limit of the XRD instrument used is 1 nm crystalline particles.40-41 Combining the measured concentration of palladium as detected by ICP, along with the solidstate NMR and XRD results, it is concluded that the molecular catalysts is bound to the surface of the A300 support. Catalytic Testing With the success of derivatizing the surface of A300 with molecular catalyst 1, batch reactions were carried out to test the catalytic activity of 1-SiO2 towards the selective hydrodeoxygenation of benzyl alcohol. This model substrate was chosen for a direct comparison to previously reported homogenous catalysis results.32 In that previous report, homogeneous catalyst 1 was able to convert benzyl alcohol to toluene with complete selectivity at 100 °C at an average turn over frequency of 5.0 hr-1 over the course of 4 hours.32 Thus, the activity of the homogeneous molecular catalyst in methanol solvent was low, but the selectivity was exceptional. For the molecular/heterogeneous catalyst 1-SiO2, a variety of solvents were explored as solubility of the molecular catalyst is no longer a concern when it is attached to a solid oxide support. The solvents were chosen to complement work that was performed in lignin depolymerization studies.42 Additionally, dodecane was chosen as a nonpolar solvent. The results for the catalytic HDO of benzyl alcohol using 1-SiO2 in the test solvents is shown in Table 1. In the first row of Table 1, 1-SiO2 exhibited great HDO selectivity, but low activity in methanol solvent. While no aromatic ring hydrogenation of benzyl alcohol was observed, the 1-SiO2 catalyst only achieved 10 turnovers in 4 hours in methanol solvent. Low activity was also observed in the polar organic solvents tetrahydrofuran (THF) and ethyl acetate. In the case of ethyl acetate selectivity also dropped as benzyl acetate was observed in the product mixture. The catalytic HDO activity of 1-SiO2 in glacial acetic acid led to complete conversion of the benzyl alcohol in under 4 hours, however, the increase in conversion was concurrent with a decrease in selectivity for the HDO product as benzyl acetate was also observed in that product mixture. Testing aqueous acidic (pH = 4.7) as the solvent suppressed benzyl acetate formation, however, a sharp decrease in the substrate conversion was observed. Furthermore, little reactivity was observed in water at 100°C. This lack of reactivity can be rationalized by highly polar water both detaching the catalyst off the metal oxide support and decomposing the molecular catalyst to a Pd-aqua complex.43 Switching from polar solvents to non-polar dodecane provided compete conversion of benzyl alcohol and complete selectivity to the desired HDO product. Dodecane, or non-polar solvents, may be ideal solvents for this catalytic system as the non-polar nature of the solvent may stabilize catalyst binding to the oxide support and dodecane will not coordinate to the Pd metal center of 1-SiO2. Furthermore, the molecular complex 1 exhibits poor to no solubility in dodecane, thus detachment of 1 from the oxide support to in situ generate a homogeneous catalyst is not expected. Table 1. Solvent exploration for catalytic HDO of benzyl alcohol using 1-SiO2.a Rxn. P Convers Select Solvent b c (bar) (%) (%)d Methanol 27 35 >99 THF 27 20 >99 Ethyl Acetate 27 48 86e Water 25 99 f Acetic Acid 22 >99 84e


Page 4 of 23

Page 5 of 23 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

ACS Catalysis

Acetic Acidg Dodecane

24 23

20 >99

>99 >99

a8.8x10-5

mol 1 on SiO2 (450 mg total 1-SiO2) in 25 mL of solvent with 0.1 M benzyl alcohol. Reaction vessel was pressured to 20 bar with H2 at room temperature before heating. Reaction time 4 hrs. bTotal pressure inside the reactor during reaction. cPercent conversion of benzyl alcohol. dPercent selectivity for toluene formation. eOther major product detected was benzyl acetate. fGlacial acetic acid. gAqueous acetic acid at pH = 4.7

Table 1 shows an apparent increase in catalytic HDO activity for 1-SiO2 in dodecane solvent. Extensive characterization of molecular/heterogeneous catalysts is first needed, however, to ensure the identity of the active catalytic species.38 To further characterize the molecular/heterogeneous catalyst 1-SiO2 a series of control reactions were performed. Table 2 lists the results of these control reactions which examined the hydrodeoxygenation of benzyl alcohol in dodecane solvent. The first row of Table 2 shows that the molecular/heterogeneous catalyst 1-SiO2 exhibits high activity and selectivity for HDO of benzyl alcohol to toluene at 100°C in dodecane solvent. Conversely, no conversion of the benzyl alcohol substrate was observed in the presence of just SiO2 without the surface-attached molecular catalyst 1. No conversion was also observed without hydrogen gas, or with a complete lack of all catalyst components. Pre-made Pd nanoparticles on silica support SBA-15 (see experimental section) were also tested for comparison. The Pd nanoparticles were active for benzyl alcohol conversion, however, the particles were not nearly as selective towards the desired HDO product compared to the molecular/heterogeneous catalyst 1-SiO2. The traditional heterogeneous Pd nanoparticles instead formed a considerable amount of ring hydrogenated products including methyl cyclohexane. This result illustrates the clear catalytic advantage of molecular/heterogeneous catalysts, which lack extended metallic surfaces that can lead to ring hydrogenation products.44 Furthermore, mercury drop tests were conducted on both the prepared Pd nanoparticles as well as 1-SiO2. Approximately 2 g of mercury placed into the reaction solution containing the Pd nanoparticle catalyst completely shut off catalytic activity. Conversely, addition of mercury to the reaction containing 1-SiO2 had little effect on the catalytic activity and no effect on the complete selectivity. Table 2. Catalyst survey for catalytic HDO of benzyl alcohol in dodecane solvent.a Convers. Select. Catalyst (%)b (%)c 1-SiO2 >99 >99 Just SiO2 0 0 1-SiO2 no H2d 0 0 None 0 0 Pd particlese >99 24f Pd particles + Hg 0 0 1-SiO2 + Hg 95 >99 Reaction filtrateg 0 0 1h >99 1.0 nm Pd particles prepared on SBA-15. fOther major product detected was methyl cyclohexane in 65 % yield. gReaction performed after filtering 1SiO2 from a post-reaction solution. h1.1x10-4 mol 1 in dodecane without SiO2. iMajor products detected were methyl cyclohexane and cyclohexane methanol.

Control reactions were also performed to show that the observed catalytic activity arises from the molecular/heterogeneous catalyst 1-SiO2 and not from homogenous catalysts that may form in situ from catalyst detachment from the SiO2 support. First, the reaction filtrate was examined for catalytic activity. To perform this test, a standard reaction was performed as described in the experimental section. After the reaction, the heterogeneous 1-SiO2 catalyst was filtered from the reaction mixture. Additional benzyl alcohol was then added to the reaction mixture and a second reaction was performed. No consumption of the added benzyl alcohol was observed, and no additional toluene formation was detected (Table 2), indicating no catalytically active species are present in the reaction solution post-reaction. Second, molecular complex 1 in dodecane without SiO2 was tested for catalytic activity. It is worth noting that 1 is not soluble in dodecane, thus the reaction solution was a suspension of 1 in dodecane. Under standard conditions, 1 is active for benzyl alcohol conversion, however, with very low selectivity to the HDO product toluene. The major products of this reaction were ring hydrogenated cyclohexane methanol and methyl cyclohexane along with black, metallic-like particles. Lastly, chloro(2,2':6',2''-terpyridine)palladium(II) chloride (Pd-tpy), was examined as an analogous catalyst to 1 to examine the role of the carboxylic acid binding group. Pd-tpy does not contain a carboxylic acid moiety and thus is not expected to bind to the SiO2 support.45 The Pd-tpy catalyst was subjected to the identical loading procedure for 1 on SiO2, followed by performing a standard reaction. Table 2 shows that the SiO2 “loaded” with Pd-tpy yielded very little benzyl alcohol conversion, suggesting that Pd-tpy does not strongly bind to the SiO2 particles and likely was washed off of the surface of the particles prior to running the reaction. Catalyst/Amorphous Silica Interaction Characterization

Figure 2. Contour plots of a) electronic IE (kcal mol-1) of complex 1 in vacuo and b) free energy in solution IE of complex 1 in DMF solution. Sites labelled A-D are the same in vacuo and in solution.


Page 6 of 23

Page 7 of 23 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

ACS Catalysis

Computations were performed to gain further insight into the interaction and binding between molecular catalyst 1 and the amorphous silica support. Attachment of molecular catalysts to oxide supports has been previously reported,33,46 but little quantitative information is available about the interaction between the molecular units and SiO2 surfaces, particularly when the surface is irregular. Since the mode and the strength of binding of complexes onto the support surface cannot be easily probed experimentally, we utilized density functional theory (DFT) simulations. In particular, a model of amorphous silica was used to evaluate the interaction energy, IE, between the surface and the catalyst complex through equation 1. 𝐼𝐸 = 𝐸𝐶𝑜𝑚𝑝𝑙𝑒𝑥 ― 𝑆𝑖𝑡𝑒 ―(𝐸𝐶𝑜𝑚𝑝𝑙𝑒𝑥 + 𝐸𝑆𝑖𝑡𝑒)

(1)

where E indicates the electronic energy of the system in gas phase or in solution. Twenty-five sites on the model silica sample were evaluated for IEs in both gas phase and implicit DMF solvent, as this is the solvent used for catalyst loading onto the SiO2 support. As described more in detail in the Computational Protocol section, the sampling involved semi-rigid structures where only the carboxylic group and the surface OH groups were allowed to geometrically relax, thus providing an initial estimate of the interaction of the carboxylic acid with the SiO2 surface. The IE values for the sampled sites are depicted as contour plots in Figure 2 and reported in Table S2. The qualitative IE pattern on both the gas-phase and implicit DMF solvent contour plots are similar, showing that the interaction between the catalyst and the SiO2 surface is highly dependent on the local amorphous silica structure. In solution the IE is significantly smaller than in the gas phase, due to screening effects of the solvent. The carboxylic acid group interaction with the silica surface is via hydrogen bonding (HB) was first examined. In general, we found three HB schemes with the surface: 1a. COH—O(silanol), 1b. COH—O(siloxane), and 1c. C=O— H(silanol) shown in Figure 3. From this initial sampling of the binding on the amorphous surface, four sites with large IE values (-28 to -50 kcal mol-1) were located, indicated in Figure 2 with letters A-D. Thus, these sites were used as model sites to explore the mode of binding in more detail. The hydrogen bonding lengths and the corresponding interaction energy for complex 1 at sites A-D in the semi-rigid structures are shown in Table 3. Note that these sites do not bind via 1c in Figure 3, indicating that 1a and 1b likely contribute to the strongest interactions. The hydrogen bond lengths of the four sites increase by 0.02-0.08 Å in solution. When the catalyst is forced in the upright orientation, the IE is completely dependent on the nature of the HB scheme.

Figure 3. Hydrogen bonding schemes identified by the simulations. Scheme 1a: COH--O (silanol); Scheme 1b: COH--O (siloxane); Scheme 1c: C=O-H (silanol). Cat refers to the catalyst structure.


ACS Catalysis 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

Table 3. Interaction energies (kcal mol-1) of sites A-D in vacuo and in solvent. H-bond distance (Å) of Sites A-D between the carboxylate linker and the silica surface for the HB schemes in Figure 3. H-bond H-bond H-bond H-bond IE in IE in Scheme Scheme Scheme Scheme Site vacuo DMF 1a in 1a in 1b in 1b in vacuo DMF vacuo DMF A -47.0 -14.2 1.62 1.70 B -29.4 -9.6 2.37 2.39 C -28.4 -12.6 2.10 2.13 D -33.4 -13.9 1.58 1.66 Following the discovery of the strongest IE sites under a constrained catalyst geometry, the constraint on the orientation of the catalyst on the surface was relaxed. A Cl- was coordinated to the Pd, as in experiment, and while the internal coordinates of the Cl-Pd-tpy molecular framework were kept frozen, the catalyst itself was allowed to move as a partially rigid body (the COOH group was still allowed to move freely). The geometry of the catalyst on all four HB sites relaxed to an orientation parallel to the surface, as shown in Figure 4 for site D. Thus, the simulations predict that a complex surrounded by an implicit solvent tends to lay parallel to the surface regardless of the hydrogen bonding schemes, consistent with literature.46 The IE for the relaxed orientations are much larger than those in the rigid calculations, as shown by the dark blue bars in Figure 5 (and Table S3) In fact, the relaxed IE values become comparable to the rigid values in vacuo.

Figure 4. Example of a relaxed complex orientation (site D). The green sphere is Cl, teal is Pd, blue is N, red is O, aqua is F, light gray is C, dark gray is Si and white is H.


Page 8 of 23

Page 9 of 23 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

ACS Catalysis

Figure 5. Interaction energies, kcal mol-1, of complex 1 derivates over Sites A-D in DMF. The varieties of the protonated complex are in blue, the deprotonated in red. As the Cl- and Pd2+ are removed, the bar color is lightened. The IE values between the catalyst with a deprotonated carboxylate group and the silica surface were also examined. It is worth noting that it is unlikely that the carboxylic acid is deprotonated in the experimental conditions used in this work during catalyst loading onto the SiO2 surface. The IEs of the deprotonated complex, plotted as dark red bars in Figure 5 and reported in Table S3, are similar to those of the protonated complex, varying by 1-3 kcal mol-1 except for Site C, which shifts by 9 kcal mol-1. In the latter, the deprotonated complex overlaps more with the center of the silica surface than the protonated species, thus increasing the strength of the interaction. This indicates that the large shift in IE is not due to the protonation state of the complex, but on the different placement of the catalyst on the surface. In all computations involving the carboxylate group, the geometry of the catalyst relaxed to a parallel position to the surface. This further indicates that the protonation state of the carboxylate group does not have a large effect on the lowest energy computed interaction between the catalyst and the SiO2 surface. In is also worth noting, the computed structure of the molecular catalysts are reasonable based on simulations of 13C NMR spectra for the complex in solution and on the silica, see Figure S13. The data in the figure shows peaks in good agreement with the experimental spectra in Figures S2, S10, and S12 (the calculated peaks are shifted by about ~10 ppm compared to experiment). In particular, the calculations support the observation that the signature complex peaks do not shift when the complex passes from homogeneous solution to binding on the silica. The aforementioned IE values may be affected by three parameters: the interaction between the linker group and the surface, dispersion effects of the -conjugated ligand with the surface and the solvent, and Coulombic attraction between the positively charged metal and the surface/solvent. We discriminated between these possible effects by performing single-point energy calculations of the relaxed geometries for 1, the catalyst without the Cl ligand, and the ligand without the Pd metal. These calculations were performed for both the protonated and deprotonated forms of the carboxylic acid linker and the IEs are plotted in Figure 6 and reported in Table S3. The removal of Pd and Cl does not change the IE considerably, except for the case of tpy-COO-. For this species, the IE is smaller due to the negative charge on the complex repelling the oxygen-rich silica surface. The other alterations to 1 shifts the IE, but by smaller amounts than


ACS Catalysis

the actual IE values, suggesting that the interaction between the silica surface and the -conjugated terpyridine has the largest contribution to the IE under non-covalent bonding of the catalyst to the surface. Another possibility is that the complex covalently bonds to the surface through the linker group via a dehydration reaction of a silanol or by breaking a siloxane bond,47 either of which may prevent the parallel orientation. The formation of covalent bonds was investigated using only a small model silica system, due to difficulties in the determining transition states for chemical reactions on a large amorphous silica model. The reaction pathways for the catalyst covalently binding to a silanol and siloxane site are reported in Figure 6 and the transition state structures are reported in Figure S14. For both silica sites, the catalyst hydrogen bonded to the surface represents the lowest energy point on the reaction pathway shown in Figure 6c. The computations indicate that the hydrogen bonding interaction is more stable than the covalently bonded analog by ~6-7 kcal mol-1 and that the transition state energy barrier to form covalent bonds to the surface is roughly only 8 kcal mol-1. Both H-bonded and covalently bonded configurations are more stable than having the complex and surface infinitely apart, thus indicating that the complex is most likely on or near the silica surface under experimental conditions. The relative stability of the reactant and product are further confirmed with simulations at a higher level of theory, coupled cluster singles and doubles with perturbative triples (CCSD(T)), using the same silica models and formic acid instead of the full complex. The structures and total energies for the CCSD(T) calculations are reported in Figure S15. It is important to note that these values may need to be revised when a more complete silica surface model is used.

c)

SiOH (a) SiOSi (b) 2O

a)

-H

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 10 of 23

b)

Figure 6. a) Hydrogen bonded and covalently bonded structures at a silanol site. b) Hydrogen bonded and covalently bonded structures at a siloxane site. For both a) and b) the green sphere is Cl, teal is Pd, blue is N, red is O, aqua is F, light gray is C, dark gray is Si and white is H. c) Reaction energy pathway (kcal mol-1) of covalently bonding to a silanol through a dehydration, black trace, or a siloxane through Si-O bond breaking, red trace. All calculations were performed in DMF solution and represent electronic energies in solution. Catalytic Kinetics and Recyclability With an understanding of the binding, stability, and reactivity of the molecular/heterogeneous catalyst 1-SiO2, the HDO of additional oxygenated aromatic substrates was explored in dodecane solvent. Table 4 shows the catalytic HDO of benzaldehyde to form toluene was also performed with complete selectivity. A lack of ring hydrogenated products was also observed for the catalytic conversion of benzophenone to diphenylmethane. The conversion of benzophenone was slower as the reaction time was extended to 24 hours and the ketone hydrogenation to diphenylmethanol was also observed. This result


Page 11 of 23 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

ACS Catalysis

indicates that the 1-SiO2 catalyst likely deoxygenates ketones via a two-step process where the ketone is first hydrogenated to the alcohol, followed by hydrodeoxygenation of the alcohol. Catalyst 1-SiO2 was also tested for the hydrodeoxygenation of two common monomers derived from lignin samples, vanillin and vanillyl alcohol. The catalyst exhibited high activity and selectivity for the hydrodeoxygenation at the benzylic position without any observed ring hydrogenation. Deoxygenation of the phenolic or methoxy group directly on the aromatic ring of these substrates was not observed under the mild reaction conditions. Table 4. Catalytic activity of 1-SiO2 in dodecane for HDO of various oxygenated aromatic compounds in dodecane solvent.a Convers. Select. Substrate (%)b (%)c Benzaldehyde 88 >99d e Beznophenone >99 67f Vanillyl Alcohol >99 >99g Vanillin >99 >99g a8.8x10-5

mol 1 on SiO2 (450 mg total 1-SiO2) in 25 mL of dodecane at 100 °C. Reaction vessel was pressured to 20 bar with H2 at room temperature before heating, with reaction pressure of 23 bar. Reaction time 4 hrs. bPercent conversion of substrate. cPercent selectivity HDO product formation. dHDO product is toluene. e24 hr. reaction. fHDO product is diphenylmethane, other major product observed is diphenylmethanol. gHDO product observed is creosol.

The catalyst 1-SiO2 exhibits high selectivity towards HDO of benzylic alcohols, thus the kinetics of benzyl alcohol HDO with 1-SiO2 were examined to gain a greater understanding of the catalyst activity. A reaction progress plot (Figure 7) shows the consumption of benzyl alcohol and the formation of toluene as a function of time. The reaction conditions used to generate Figure 7 follow the standard procedure for stirring speed (900 rpm), H2 pressure (20 bar) and temperature (100 °C). Catalyst loading for these reactions was 225 mg of 1-SiO2, equating to 4.4 x 10-5 mol Pd with a catalyst surface concentration of 21 mg Pd per g of catalyst material or 6.5x10-7 mol Pd per m2 surface area of oxide particles. From the data in Figure 7, the initial turn-over frequency (TOF) for the catalytic HDO of benzyl alcohol from 1-SiO2 in dodecane solvent was determined to be 30 hr-1. This TOF is a marked increase from the 10 turnovers in 4 hours from 1-SiO2 in methanol solvent and nearly 10x greater than the TOF observed for the previously reported, analogous, molecular homogeneous catalyst.32


ACS Catalysis 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 12 of 23

Figure 7. Formation/consumption curves of toluene/benzyl alcohol for the selective HDO of benzyl alcohol catalyzed by 1-SiO2. The deoxygenation of benzyl alcohol is proposed to occur in two steps, as shown in equations 2 and 3. First, the Pd center of the molecular/heterogeneous catalyst 1-SiO2 is activated by H2 to form an intermediate species. Attempts to isolate the Pd intermediate were unsuccessful, but it is proposed to go through a reduced Pd or transient Pd-hydride species. The Pd intermediate then selectively reacts with benzyl alcohol to produce toluene in the second step. The rate determining step of under the chosen reaction conditions was found to be equation 3. The solubility of hydrogen in dodecane at 20 bar and 100 °C leads to a 1000-fold excess of H2 compared to moles of the catalyst.48 Furthermore, increasing the pressure of H2 to 40 bar or decreasing it to 15 bar did not have an effect on the observed TOFs. Additionally, reactivity is beyond mass transport limitations as stirring rpm does not play a role in enhancing the catalyzed reaction (rpms tested: 1300, 900, 500 rpm) Further, a 1-SiO2 sample was prepared with approximately half of the loading of Pd as used for the data shown in Figure 7. The half-loaded sample was tested for catalysis under the standard reaction conditions and found to yield the same initial TOF as the fully loaded catalyst (30 hr-1), which is an indication that mass-transport limitations do not exist for the reaction.49 With equation 3 as the rate-determining step, the concentrations of benzyl alcohol and 1-SiO2 were systematically varied. The data (see SI) was consistent with a first order rate dependence on both benzyl alcohol and 1-SiO2 concentration. Thus, the rate constant for this reaction was calculated using the relationship rate = k[Pd][BA]. The full rate derivation can be found in the SI. Following the full kinetic analysis, the second order rate constant for the catalytic HDO of benzyl alcohol from 1-SiO2 was found to be k = 7.3 x 10-2 M-1 s-1. 𝑃𝑑𝐼𝐼 + 𝐻2↔𝑃𝑑𝑖𝑛𝑡

(2)

𝑃𝑑𝑖𝑛𝑡 +𝐵𝑒𝑛𝑧𝑦𝑙𝐴𝑙𝑐𝑜ℎ𝑜𝑙→𝑇𝑜𝑙𝑢𝑒𝑛𝑒

(3)

Increased ease in catalyst separation from reaction mixtures and the possibility of easily recycling the catalyst are another advantages of molecular/heterogeneous catalysts over traditional homogeneous catalysts. Thus, 1-SiO2 recyclability studies were performed under limited substrate conversions, and the results are shown in Table 5. Throughout 10 consecutive reactions, 1-SiO2 maintained selectivity for HDO of benzyl alcohol. Furthermore, the mean turnover frequency for these reactions is 21 hr-1, indicating the activity of 1-SiO2 was also consistent over multiple recycled reactions. The recycled 1-SiO2 catalyst used was then examined with XRD to look for the possible presence of Pd nanoparticles. The XRD pattern (Figure S8) does not show evidence for Pd nanoparticles on the SiO2 supports. While the lack of observed XRD patterns for Pd nanoparticles does not definitely conclude the lack of Pd nanoparticles, the lack of the observed peaks in the XRD pattern does suggest that 1-SiO2 does maintain molecular integrity. To further examine the nature of 1-SiO2 post-reaction, the catalyst was characterized with solid-state NMR measurements. Solid-state 13C NMR of 1-SiO2 after a catalytic reaction still contained peaks with chemical shifts between 160 and 120 ppm indicating the presence of the 2,2':6',2''-terpyridine-4'-carboxylic acid ligand on the surface of the SiO2 support (Figure S10). These results show that the catalyst is still attached to the surface of SiO2 post recycle reactions. Using dodecane as a solvent likely assists this prolonged attachment as dodecane cannot hydrogen bond to the surface of support nearly as well as the carboxylic acid of the tpy ligand. Conversely, catalyst decomposition to metallic Pd particles can be observed under


Page 13 of 23 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

ACS Catalysis

full substrate conversion recycle experiments. Upon substrate depletion, the molecular catalyst can be reduced by H2 down to metallic Pd nanoparticles. These nanoparticles arise from an unwanted side reaction that produces a material which lacks the desired catalytic selectivity. Full conversion recyclability studies led to observed changes in product distribution following the first recycle, (Table S1) and the formation of Pd nanoparticles was observed with XRD following the full conversion recycle reactions (Figure S9). Table 5. Catalyst 1-SiO2 recycle studies for HDO of benzyl alcohol in dodecane solvent.a Select. Select. Select. Convers. Rxn. Tol. MCH CHM (%)b (%)c (%)d (%)e 1 17.5 >99 0 0 2 18.0 >99 0 0 3 17.7 >99 0 0 4 18.4 >99 0 0 5 17.6 >99 0 0 6 17.9 >99 tracef trace 7 21.0 >99 trace trace 8 19.4 >99 trace trace 9 21.1 >99 trace trace 10 19.8 >99 trace trace a4.4x10-5

mol 1 on SiO2 (225 mg total 1-SiO2), in 25 mL of dodecane at 100 °C. Reaction vessel was pressured to 20 bar with H2 at room temperature before heating, with reaction pressure of 23 bar. Reaction time 30 minutes. bPercent conversion of benzyl alcohol. cPercent selectivity for toluene formation. dPercent selectivity for methyl cyclohexane (MCH) formation. ePercent selectivity for cyclohexane methanol (CHM) formation. fTrace refers to signals that equate to less than 1% of the carbon balance.

Discussion The data in Table 1 indicated that dodecane is the optimal solvent for the catalytic HDO of benzyl alcohol from 1-SiO2. Dodecane likely stabilizes catalyst binding to the oxide support as the non-polar dodecane will not break up any of the hydrogen bonding or Coulombic interactions that calculations predicted exist between the molecular catalyst and the SiO2 support. The main result of the simulations is that strong binding of the catalyst to the surface is predicted. When allowed to relax, the catalyst seems to prefer an orientation parallel to the surface. However, these calculations are only a first step in understanding the interaction between this catalyst and the amorphous silica. For instance, the relaxation of a single catalyst on the surface in implicit solvent may not accurately reproduce the energy penalty for displacing solvent molecules also interacting with the surface. Additionally, multiple complex molecules may bind to the surface in close proximity, thus preventing the full interaction of the tpy ligand with the silica. Both of these possible interactions may lead to the catalyst favoring a perpendicular orientation with the surface under experimental conditions. Thermal effects were also not considered, which have been shown experimentally to promote covalent binding.50 Thermal effects may also help desorb the water, therefore stabilizing the COSi covalent bond. We are currently investigating all of these possibilities with more refined models, which will be the subject of a future report.


ACS Catalysis 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

Furthermore, the data in Table 2 strongly indicates that the molecular/heterogeneous 1-SiO2 maintains molecular integrity in dodecane solvent. The HDO selectivity of 1-SiO2 is vastly greater than the HDO selectivity of Pd nanoparticle catalysts. In addition, 1-SiO2 is unaffected by additions of Hg to the reaction mixture whereas Pd nanoparticle activity is completely suppressed by the addition of Hg. Table 2 also shows that the reactivity of 1-SiO2 is quite different from 1 in solution without an SiO2 support. These results show that the molecular/heterogeneous catalyst 1-SiO2 offers unique catalytic reactivity due to isolating molecular catalysts onto the surface of a solid oxide support. The molecular/heterogeneous catalyst 1-SiO2 exhibited complete selectivity for the HDO of benzylic positions over ring hydrogenation reactions. In addition, 1-SiO2 exhibited greater catalytic activity in dodecane solvent compared to polar solvents. A similar solvent trend has been observed and attributed to the interaction of the solvent with the catalytically active site.51 Polar solvents, such as methanol, can bind to the Pd-center of the molecular catalyst and compete with substrate binding. Non-polar dodecane will not strongly bind to the Pd-center, thus all active sites will be able to bind and activate the benzyl alcohol substrate, leading to greater observed kinetics. The TOFs determined for 1-SiO2 are an order of magnitude greater than previously reported homogeneous32 and heterogeneous16 catalysts for HDO of benzylic alcohols. The catalyst 1-SiO2 also exhibits comparable kinetics to Pd nanoparticles catalysts modified with phosphonate groups for the HDO of vanillyl alcohol.15 Recycle studies also showed that the molecular/heterogeneous catalyst 1-SiO2 is stable over multiple catalytic reactions. Catalyst activity and complete selectivity for HDO was maintained for at least 10 consecutive catalytic reactions. Post reaction analysis with solid state NMR showed evidence for 2,2':6',2''-terpyridine-4'-carboxylic acid ligand on the surface of the SiO2 support following the recycle reactions. In addition, XRD analysis did not show any peaks corresponding to metallic Pd nanoparticles, indicating that 1-SiO2 maintains molecular integrity and molecular binding to the SiO2 support over extended reaction times. Conclusions The molecular/heterogeneous catalyst 1-SiO2 was prepared, characterized, and utilized for the selective hydrodeoxygenation of oxygenated aromatics. The favorable binding of the carboxylic acid moiety of 1 to the SiO2 support led to a simple loading procedure for the synthesis of 1-SiO2. Computational studies indicate that the molecular catalyst can hydrogen bond, bond through Coulombic, π interactions, or covalently bind to the surface of SiO2. Further simulations that investigate, for instance, multiple complexes, thermal effects, and explicit solvent molecules are necessary to obtain a more complete picture of the catalyst-surface interaction. Overall, the calculated interaction energies are in agreement with the experimental finding that these molecular catalysts strongly bind onto the surface of amorphous silica for prolonged reactivity. The molecular catalyst attached to the silica support exhibited excellent selectivity towards HDO and did not result in any observed ring hydrogenation when an excess of substrate was present in the reaction mixture. Recyclability studies showed consistent product formation could be obtained over multiple catalytic runs. This catalytic selectivity can be attributed to ensuring a lack of the extended catalyst surfaces typically encountered with metallic nanoparticle catalysts, without the need for difficult control of nanoparticle synthesis typically encountered in traditional heterogeneous catalysts. Thus, the molecular/heterogeneous catalysts 1-SiO2 successfully combines of the desired aspects of molecular and heterogeneous catalysts. Not only does the molecular/heterogeneous catalytic motif present offer the combined advantages of each system, our system also presents the ability to access new solvents that would not be feasible for a molecular homogeneous catalyst to work. The solvent selection of dodecane led to


Page 14 of 23

Page 15 of 23 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

ACS Catalysis

enhanced kinetics of catalysis when compared to kinetics observed in polar solvents. With the ability to fine tune molecular catalysts through well-known synthetic modifications, this hybrid catalyst motif offers a new avenue for selective, catalytic HDO of oxygenated aromatic compounds important for lignin upgrading. Experimental Materials N,N-Dimethylformamide (DMF) (anhydrous, 99.9 %, VWR), methanol (MeOH) (99.8 %, VWR), dodecane (anhydrous, 99 %, Sigma-Aldrich), acetic acid (glacial, Fisher Scientific) were purchased and used without any further purification as solvents for catalyst synthesis, immobilization and catalytic testing. Dichloro(1,5-cyclooctadiene)palladium(II) (Oakwood Chemical) was used as the palladium precursor for catalyst synthesis. 2,2':6',2''-terpyridine-4'-carboxylic acid (tpy-COOH) (95 %, Alfa Aesar) was purchased and used as the ligand to synthesize complexes suitable for immobilization. Aerosil 300 (A300, Evonik) was used as the metal oxide support for the molecular catalyst. A300 is amorphous, hydrophilic fumed SiO2 with 300 m2/g surface area and an average particle size of 20 nm. Tetraamminepalladium(II) chloride (99.99%, Aldrich) was used as the palladium precursor for nanoparticle preparation. SBA-15 (ACS Material) was used as the silica support for the nanoparticles. SBA-15 is a mesoporous silica with 99.95 %, Praxair) was used as the hydrogen source during catalytic testing. Catalyst Preparation [PdCl([2,2':6',2''-terpyridine]-4'-carboxylic acid)]Cl (1) Dichloro(1,5-cyclooctadiene)palladium (0.285 g, 1 mmol) and tpy-COOH (0.277 g, 1 mmol) were added to a 9:1 mixture of MeOH and DMF (100mL) and stirred at reflux for 1 hour and allowed to cool to room temperature. The yellow-orange precipitate that formed was filtered and washed with cold MeOH and allowed to dry. Yield: 0.40 g (0.88 mmol, 88 %). 1H NMR (DMSO-d6, ppm): 9.005 (s, 2H), 8.900 (d, 2H), 8.760 (dd, 2H), 8.471 (dt, 2H), 7.935–7.899 (m, 2H). 13C NMR (DMSO-d , ppm): 164.68, 158.03, 155.91, 152.55, 144.87, 143.15, 129.70, 126.54, 124.27. 6 Elemental analysis for PdC16H11N3O2Cl2 (1) calculated: C 42.27 %, H 2.44 %, N 9.24 %. Found: C 41.47%, H 2.53%, N 8.91%. [PdCl([2,2':6',2''-terpyridine]-4'-carboxylic acid)]Cl on SiO2 (1-SiO2) Complex 1 (50 mg) was dissolved in 100 mL DMF and sonicated to ensure complete solubility. Immediately following, A300 (1.0 g) was added to the DMF solution and sonicated for 10 minutes to ensure a well-dispersed suspension. The resulting mixture was allowed to sit undisturbed for 12+ hours to allow the complex to interact with the surface of the silica particles. Following the soak period, the solution/suspension mixture is filtered and washed with cold MeOH to remove any excess DMF and allowed to air dry. Following this procedure, the resulting solids were characterized to ensure the molecular nature of the catalyst remained intact and attached to the surface of the A300 support. 1-SiO2 supported on A300 has a yellow tint to it (Figure S1), indicative of the color of the molecular complex 1. 13C SSNMR (500 MHz, ppm): 164.35, 155.81, 152.04, 142.22, 140.16, 129.93, 126.65, 124.07. Palladium nanoparticles supported on mesoporous silica (Pd/SBA-15) was prepared by the method of strong electrostatic adsorption with synthesis conditions based on previous work.40 The Pd on the support powder was calculated to be 1.3 wt% Pd/SBA-15. The XRD patterns for the support and final catalyst give a particle size estimation of small particles equal to or less than 2 nm.


ACS Catalysis 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

Instrumentation X-ray diffraction (XRD) analysis was carried out with a benchtop powder X-ray diffractometer (Rigaku Miniflex-II with a silicon strip detector, D/teX Ultra – capable of detecting nanoparticles down to 1.2 nm) with Cu Kα radiation (λ = 1.5406 Å), operated at 15 kV and 30 mA. Powder samples were loaded on a zero-background holder and scans were made from the 20° – 80° 2θ range, with a scan rate of 3.0° 2θ/min. Solution-phase NMR analysis was taken on a Bruker Avance III-HD (400MHz). Solid-state NMR analysis was taken on a Bruker Avance III-HD (500 MHz). Inductively Coupled Plasma-mass spectrometry (ICPMS, Finnigan ELEMENT XR double focusing magnetic field) analysis was used for the analysis of palladium present on silica with rhenium as internal standard. Quartz torch and injector (Thermo Fisher Scientific) and 0.2 mL/min Micromist U-series nebulizer (GE, Australia) were used for sample introduction. Post-reaction products were analyzed through GC-MS (Shimadzu QP-2010S). The separation column was a 30-meter-long Rxi-5ms (Restek) with a 0.25 mm id and the oven temperature program was 40°C for 0.5 minutes followed by a 10°C/min ramp to 280°C and held for 2 minutes. Mass spectrometer electron ionization was at 70 eV and the spectrometer was scanned from 500 to 50 m/z at low resolution. Catalytic Testing – General Procedure All reactions were carried out in a 130-mL capacity stainless steel batch reactor (Parker Autoclave Engineers). For a typical experiment, 260 µL (0.10 M) benzyl alcohol, and 450 mg (2.1 wt.% Pd, 8.8x10-5 mol) catalyst were mixed in 25 mL of the chosen solvent. The resulting mixture was then sealed in the batch reactor and purged with 25 bar H2 three times. Then, the reactor was filled to the desired pressure of H2 and kept sealed for the duration of all experiments. For reactions that required elevated temperatures, a heating jacket was used for heating, with a thermocouple for temperature control. The reaction was then stirred at 900 rpm for the duration of the experiment. Following the experiment, the reactor was allowed to cool to room temperature without stirring and the pressure released. Products of reactions were analyzed through GC-MS following filtration of the catalyst. Catalytic Testing – Recyclability Testing/Low Conversion All recyclability testing was carried out in a 130-mL capacity batch reactor following the general procedure above. Following each reaction, the solid, heterogeneous catalyst was filtered from the reaction solution, rinsed multiple times with hexanes, and dried. A new reaction solution was then prepared, according to the general procedure above. The dried, recycled catalyst was then added to the new reaction solution and the catalytic reaction was carried out under identical conditions to the first catalytic reaction. The procedure was chosen to ensure that product formation from multiple catalytic cycles could only arise from the heterogeneous catalyst system, and that catalysis was not occurring in homogeneous solution. Low conversion studies were carried out in similar fashion as described in the general procedure; however, the amount of catalyst was decreased to 225 mg (2.1 wt. % Pd, 4.4x10-5 mol) and the reaction time was decreased to 30 minutes. Low conversion recyclability was achieved through the same means as described above for the full conversion recycle tests. Catalytic Testing – Sampling/Kinetics All sampling and kinetic evaluations were carried out in a 130-mL capacity batch reactor following the general procedure above. To obtain reaction progress throughout the course of a 4-hour reaction, the reaction mixture was sampled by using a sampling valve with a sampling tube fitted with a frit to ensure no catalyst was removed from the reactor vessel. At each sampling time, approximately 100 µL of the reaction


Page 16 of 23

Page 17 of 23 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

ACS Catalysis

mixture was sampled and analyzed with GC-MS. The concentration of benzyl alcohol and catalyst loading were varied to obtain the required kinetic data. Computational Protocol As the binding of the catalyst on an amorphous surface depends on the local structure of the surface, it is difficult to determine a single structure of the silicate that is representative of the experimental conditions. Therefore, we started by systematically sampling possible binding sites of 1 on a large slab of amorphous silicate, see Figure 8. The slab of amorphous silicate was obtained with a procedure based on molecular dynamics simulations, in which a unit cell of β-cristabolite is heated to 8000K. The system was then cooled quickly to maintain the disordered structure. Further specifics regarding the generation of the amorphous silica can be found in our previous work.52 The sampling is performed by maintaining both the complex and the silicate surface at a fixed distance and almost completely rigid (see detailed description below), in order to quickly determine the most favorable sites for, and modes of binding. Once the strongest binding sites were located, the complex was relaxed on the surface while keeping the internal Pd-tpy structure rigid (the COOH group was still free to move) to limit the computational effort. In this way, the energetics and geometries of catalyst on the surface of the silica were determined.

Figure 8. a) The amorphous silica unit cell with the 5x5 grid. The catalyst is located at the center of the blue ring, i.e. the binding site. Each binding site is carved out of the slab for the binding energy calculations. b) The catalyst 1 is placed perpendicular to the silicate surface. The dashed blue line indicates the average height of the highest 10 atoms. Teal spheres are Pd, blue is N, light gray is C, dark gray is Si, red is O and white is H. Twenty-five sites on the silicate slab (approximately 5 x 5nm in size) were selected as the sample set, as shown in Figure 8a. We then proceeded with a cluster-model approach for the calculations, because the binding is local to each site, and the silica slab is too large (5103 atoms) for periodic boundary condition calculations (PBC). DFT-PBC has the advantage of considering the surface as a solid, albeit infinitely repeated, and it has been used successfully to study binding on amorphous surfaces, but on considerably smaller slabs.53-55 On the other hand, our approach allows us to sample many different amorphous local configurations of the silica, which would be extremely computationally demanding to do with DFT-PBC methods. Cylindrical silicate clusters with a radius of 7 Å and depth of 3-5 Å were carved out and capped with H to complete valency, an example shown in Figure 8b. The complex was placed perpendicularly to


ACS Catalysis 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 surface of the silicate cluster, at a fixed distance of about 3 Å. Since the cluster surface is irregular, the 10 highest atoms were selected, and the average position of the z coordinate perpendicular to the surface of the top 10 atoms was chosen for analysis. We then placed the C in the carboxylic group 3 Å above this average z-coordinate and at the center of the site; the principal axis of the catalyst is along the z axis. This procedure places the catalyst close enough to the surface to allow hydrogen bonding with nearby silanol groups. A constrained geometry optimization was then performed where only the carboxylate group and the pre-existing silanol groups are allowed to relax. The system is treated using the B3LYP hybrid functional and Grimme’s D3 dispersion correction.56 The metal, Pd, is treated with the Def2TZVP basis set and Stuttgart’s pseudopotentials.57 The groups that are most likely to contribute to the IE, the carboxylate group and pre-existing silanols at each site, are treated with 6-31++g(2d,p). This basis set includes diffuse functions to better describe the hydrogen bonding interactions between the silica and the linker group. The remaining atoms were treated with the 631g(d) basis set. The IEs were calculated both in vacuo and in the presence of DMF, using geometries that are relaxed in the corresponding medium. As the complex is immobilized on the silica in DMF, it is the only solvent utilized in these geometry optimizations. The solvent effect is described with the solvation model with density approach (SMD).58 The four sites with the largest IE are used for an additional geometry optimization in which the catalyst is allowed to relax on the silicate surface. This relaxation is performed in two steps. First, the catalyst is allowed to relax on the surface as a rigid body, also keeping the bulk silica fixed except for the top silanol groups. Then, almost all constraints are relaxed, and a full geometry optimization is performed; the only constraints left are on the capping -OH groups on the silica, which are kept frozen to maintain the amorphous nature of the cluster consistent with that of the original slab. A chloride ion is included in these calculations, treated with 6-31++g(2d,p). The relaxation is performed in DMF. Note that the chloride ion was not included in the initial sampling calculations in which 1 is perpendicular to the surface, as it is far away from the silica and not expected to affect the interaction energy. However, the Cl- may play a part when the complex is allowed to relax on to the surface and is therefore included in the relaxation calculations. The contribution from the Cl and Pd are examined by removing these centers from the relaxed geometries and obtaining the IE from a single-point calculation. Finally, we consider the binding of complex 1 with a deprotonated carboxylic group. The 13C NMR spectra are computed at B3LYP-D3/6311+g(2d,p) level for all elements except Pd, for which the def2tzvp basis set and Stuttgart ECP are used. All calculations are performed in implicit DMSO solvent and with gauge including atomic orbitals (GIAOs). The reference shielding value is TMS: σref=187.435 ppm. The mechanism to covalently bond complex 1 to silica via the carboxylate group is explored on small model systems. Covalent bonds are defined as atomic distances that are consistent with typically defined covalent bonds in experimental data. For example, Si-O atomic distance lengths that are shorter than 1.70 Å. Both a silanol and siloxane model system are considered, F3Si-OH and F3Si-O-SiF3, respectively. The transition state is optimized and confirmed with a frequency calculation, ensuring it has a single imaginary frequency. The structure is then moved along the TS normal mode in either direction and optimized to the closest minima to confirm that the TS goes to the expected reactant and product. Calculations are performed with the same protocol as those done to obtain the IE, with the exception of the capping F, which are treated with 6-31g. This deficient basis set was shown by Goldsmith and coworkers to best mimic the size of oxygen and the electronegativity of an extended silica network.59 These calculations were carried out in implicit DMF solvent. All calculations were performed with the GAUSSIAN suite of programs.60 ASSOCIATED CONTENT


Page 18 of 23

Page 19 of 23 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

ACS Catalysis

Supporting Information. XRD spectra, 1H and 13C NMR data, computational data, and kinetic analysis. This material is available free of charge via the Internet at http://pubs.acs.org. Conflicts of Interest There are no conflicts to declare. Acknowledgements The support of the National Science Foundation through the EPSCoR R-II Track-2 grant number OIA1539105 is gratefully acknowledged. References 1. De, S.; Saha, B.; Luque, R., Hydrodeoxygenation processes: Advances on Catalytic Transformations of Biomass-Derived Platform Chemicals into Hydrocarbon Fuels. Bioresour. Technol. 2015, 178, 108-118. 2. Zakzeski, J.; Bruijnincx, P. C. A.; Jongerius, A. L.; Weckhuysen, B. M., The Catalytic Valorization of Lignin for the Production of Renewable Chemicals. Chem. Rev. 2010, 110, 3552-3599. 3. Pandey, M. P.; Kim, C. S., Lignin Depolymerization and Conversion: A Review of Thermochemical Methods. Chem. Eng. Technol. 2010, 34, 29-41. 4. Rahimi, A.; Ulbrich, A.; Coon, J. J.; Stahl, S. S., Formic-Acid-Induced Depolymerization of Oxidized Lignin to Aromatics. Nature 2014, 515, 249-252. 5. Luo, H.; Abu-Omar, M. M., Lignin Extraction and Catalytic Upgrading from Genetically Modified Poplar. Green Chem. 2018, 20, 745-753. 6. Renders, T.; Van den Bosch, S.; Koelewijn, S. F.; Schutyser, W.; Sels, B. F., Lignin-First Biomass Fractionation: The Advent of Active Stabilisation Strategies. Energy Environ. Sci. 2017, 10, 1551-1557. 7. Das, A.; Rahimi, A.; Ulbrich, A.; Alherech, M.; Motagamwala, A. H.; Bhalla, A.; da Costa Sousa, L.; Balan, V.; Dumesic, J. A.; Hegg, E. L.; Dale, B. E.; Ralph, J.; Coon, J. J.; Stahl, S. S., Lignin Conversion to Low-Molecular-Weight Aromatics via an Aerobic Oxidation-Hydrolysis Sequence: Comparison of Different Lignin Sources. ACS Sustain. Chem. Eng. 2018, 6, 3367-3374. 8. Gollakota, A. R.; Reddy, M.; Subramanyam, M. D.; Kishore, N., A Review on the Upgradation Techniques of Pyrolysis Oil. Renew. Sustain. Energy Rev. 2016, 58, 1543-1568. 9. Nolte, M. W.; Shanks, B. H., A Perspective on Catalytic Strategies for Deoxygenation in Biomass Pyrolysis. Energy Technol. 2017, 5, 7-18. 10. Ertl, G., Molecules at Surfaces: 100 Years of Physical Chemistry in Berlin‐Dahlem. Angew. Chem. Int. Ed. 2012, 52, 52-60. 11. Subramaniam, B.; Helling, R. K.; Bode, C. J., Quantitative Sustainability Analysis: A Powerful Tool to Develop Resource-Efficient Catalytic Technologies. ACS Sustain. Chem. Eng. 2016, 4, 58595865. 12. Somorjai, G. A.; Kliewer, C. J., Reaction Selectivity in Heterogeneous Catalysis. React. Kinet. Catal. Lett. 2009, 96, 191-208. 13. Thomas, J. M., The Societal Significance of Catalysis and the Growing Practical Importance of Single-Site Heterogeneous Catalysts. Proc. Royal Soc. A 2012, 468, 1884-1903. 14. Parsell, T. H.; Owen, B. C.; Klein, I.; Jarrell, T. M.; Marcum, C. L.; Haupert, L. J.; Amundson, L. M.; Kenttämaa, H. I.; Ribeiro, F.; Miller, J. T.; Abu-Omar, M. M., Cleavage and Hydrodeoxygenation (HDO) of C–O Bonds Relevant to Lignin Conversion using Pd/Zn Synergistic Catalysis. Chem. Sci. 2013, 4, 806-813.


ACS Catalysis 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

15. Hao, P.; Schwartz, D. K.; Medlin, J. W., Effect of Surface Hydrophobicity of Pd/Al2O3 on Vanillin Hydrodeoxygenation in a Water/Oil System. ACS Catal. 2018, 8, 11165-11173. 16. Lien, C.-H.; Medlin, J. W., Promotion of Activity and Selectivity by Alkanethiol Monolayers for Pd-Catalyzed Benzyl Alcohol Hydrodeoxygenation. J. Phys. Chem. C 2014, 118, 23783-23789. 17. Shao, Y.; Xia, Q.; Dong, L.; Liu, X.; Han, X.; Parker, S. F.; Cheng, Y.; Daemen, L. L.; RamirezCuesta, A. J.; Yang, S.; Wang, Y., Selective Production of Arenes via Direct Lignin Upgrading over a Niobium-Based Catalyst. Nat. Commun. 2017, 8, 16104/1-16104/9. 18. Guo, T.; Xia, Q.; Shao, Y.; Liu, X.; Wang, Y., Direct Deoxygenation of Lignin Model Compounds into Aromatic Hydrocarbons Through Hydrogen Transfer Reaction. Appl. Catal. A: Gen. 2017, 547, 30-36. 19. Baddour, F. G.; Witte, V. A.; Nash, C. P.; Griffin, M. B.; Ruddy, D. A.; Schaidle, J. A., LateTransition-Metal-Modified β-Mo2C Catalysts for Enhanced Hydrogenation during Guaiacol Deoxygenation. ACS Sustain. Chem. Eng. 2017, 5, 11433-11439. 20. Hsu, P.-J.; Jiang, J.-W.; Lin, Y.-C., Does a Strong Oxophilic Promoter Enhance Direct Deoxygenation? A Study of NiFe, NiMo, and NiW Catalysts in p-Cresol Conversion. ACS Sustain. Chem. Eng. 2018, 6, 660-667. 21. Ju, C.; Li, M.; Fang, Y.; Tan, T., Efficient Hydro-Deoxygenation of Lignin Derived Phenolic Compounds Over Bifunctional Catalysts with Optimized Acid/Metal Interactions. Green Chem. 2018, 20, 4492-4499. 22. Song, W.; Zhou, S.; Hu, S.; Lai, W.; Lian, Y.; Wang, J.; Yang, W.; Wang, M.; Wang, P.; Jiang, X., Surface Engineering of CoMoS Nanosulfide for Hydrodeoxygenation of Lignin-Derived Phenols to Arenes. ACS Catal. 2018, 259-268. 23. Cornils, B.; Herrman, W. A., Concepts in Homogeneous Catalysis: The Industrial View. J. Catal. 2003, 216, 23-31. 24. Loch, J. A.; Crabtree, R. H., Rapid Screening and Combinatorial Methods in Homogeneous Organometallic Catalysis. Pure Appl. Chem. 2001, 73, 119-128. 25. McLain, S. J.; Wood, C. D.; Schrock, R. R., Preparation and Characterization of Tantalum(III) Olefin Complexes and Tantalum(V) Metallacyclopentane Complexes Made from Acyclic .Alpha. Olefins. J. Am. Chem. Soc. 1979, 101, 4558-4570. 26. Schwab, P.; France Marcia, B.; Ziller Joseph, W.; Grubbs Robert, H., A Series of Well‐Defined Metathesis Catalysts–Synthesis of [RuCl2(=CHR′)(PR3)2] and Its Reactions. Angew. Chem. Int. Ed. 1995, 34, 2039-2041. 27. Kusumoto, S.; Nozaki, K., Direct and Selective Hydrogenolysis of Arenols and Aryl Methyl Ethers. Nat. Commun. 2015, 6, 6296/1-6296/7. 28. Sergeev, A. G.; Hartwig, J. F., Selective, Nickel-Catalyzed Hydrogenolysis of Aryl Ethers. Science 2011, 332, 439-443. 29. Minard, T. A.; Oswin, C. T.; Waldie, F. D. C.; Howell, J. K.; Scott, B. M. T.; Mondo, D. D.; Sullivan, R. J.; Stein, B.; Jennings, M.; Schlaf, M., In Situ Generation of Water-Stable and -Soluble Ruthenium Complexes of Pyridine-Based Chelate-Ligands and their use for the Hydrodeoxygenation of Biomass-Related Substrates in Aqueous Acidic Medium. J. Mol. Catal. A: Chem. 2016, 422, 175-187. 30. Sullivan, R. J.; Latifi, E.; Chung, B. K. M.; Soldatov, D. V.; Schlaf, M., Hydrodeoxygenation of 2,5-Hexanedione and 2,5-Dimethylfuran by Water-, Air-, and Acid-Stable Homogeneous Ruthenium and Iridium Catalysts. ACS Catal. 2014, 4, 4116-4128. 31. Latifi, E.; Marchese, A. D.; Hulls, M. C. W.; Soldatov, D. V.; Schlaf, M., [Ru(triphos)(CH3CN)3](OTf)2 as a Homogeneous Catalyst for the Hydrogenation of Biomass Derived 2,5-Hexanedione and 2,5-Dimethyl-Furan in Aqueous Acidic Medium. Green Chem. 2017, 19, 46664679. 32. DeLucia, N. A.; Das, N.; Overa, S.; Paul, A.; Vannucci, A. K., Low Temperature Selective Hydrodeoxygenation of Model Lignin Monomers from a Homogeneous Palladium Catalyst. Catal. Today 2018, 302, 146-150.


Page 20 of 23

Page 21 of 23 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

ACS Catalysis

33. McNamara, W. R.; Snoeberger Iii, R. C.; Li, G.; Richter, C.; Allen, L. J.; Milot, R. L.; Schmuttenmaer, C. A.; Crabtree, R. H.; Brudvig, G. W.; Batista, V. S., Hydroxamate Anchors for WaterStable Attachment to TiO2 Nanoparticles. Energy Environ. Sci. 2009, 2, 1173-1175. 34. Ashford, D. L.; Gish, M. K.; Vannucci, A. K.; Brennaman, M. K.; Templeton, J. L.; Papanikolas, J. M.; Meyer, T. J., Molecular Chromophore–Catalyst Assemblies for Solar Fuel Applications. Chem. Rev. 2015, 115, 13006-13049. 35. Losego, M. D.; Hanson, K., Stabilizing Molecular Sensitizers in Aqueous Environs. Nano Energy 2013, 2, 1067-1069. 36. Vannucci, A. K.; Alibabaei, L.; Losego, M. D.; Concepcion, J. J.; Kalanyan, B.; Parsons, G. N.; Meyer, T. J., Crossing the Divide Between Homogeneous and Heterogeneous Catalysis in Water Oxidation. Proc. Natl. Acad. Sci. 2013, 110, 20918-20922. 37. Materna, K. L.; Crabtree, R. H.; Brudvig, G. W., Anchoring Groups for Photocatalytic Water Oxidation on Metal Oxide Surfaces. Chem. Soc. Rev. 2017, 46, 6099-6110. 38. Phan, N. T. S.; Van Der Sluys, M.; Jones, C. W., On the Nature of the Active Species in Palladium Catalyzed Mizoroki–Heck and Suzuki–Miyaura Couplings – Homogeneous or Heterogeneous Catalysis, A Critical Review. Adv. Synth. Catal. 2006, 348, 609-679. 39. Hübner, S.; de Vries, J. G.; Farina, V., Why Does Industry Not Use Immobilized Transition Metal Complexes as Catalysts? Adv. Synth. Catal. 2016, 358, 3-25. 40. O’Connell, K.; Regalbuto, J. R., High Sensitivity Silicon Slit Detectors for 1 nm Powder XRD Size Detection Limit. Catal. Lett. 2015, 145, 777-783. 41. Liu, Q.; Samad, J.; Copple, J. E.; Eskandari, S.; Satterwhite, C.; Regalbuto, J. R., A Pinch of Salt to Control Supported Pt Nanoparticle Size. Catal. Today 2017, 280, 246-252. 42. Nandiwale, K. Y.; Danby, A. M.; Ramanathan, A.; Chaudhari, R. V.; Subramaniam, B., Zirconium-Incorporated Mesoporous Silicates Show Remarkable Lignin Depolymerization Activity. ACS Sustain. Chem. Eng. 2017, 5, 7155-7164. 43. Vicente, J.; Arcas, A., Aqua Palladium Complexes: Synthesis, Properties and Applications. Coord. Chem. Rev. 2005, 249, 1135-1154. 44. Liu, H.; Jiang, T.; Han, B.; Liang, S.; Zhou, Y., Selective Phenol Hydrogenation to Cyclohexanone Over a Dual Supported Pd–Lewis Acid Catalyst. Science 2009, 326, 1250-1252. 45. Key, R. J.; Tengco, J. M. M.; Smith, M. D.; Vannucci, A. K., A Molecular/Heterogeneous Nickel Catalyst for Suzuki-Miyaura Coupling. Organometallics 2019, 38, 2007-2014. 46. Costa, D.; Tougerti, A.; Tielens, F.; Gervais, C.; Stievano, L.; Lambert, J. F., DFT Study of the Adsorption of Microsolvated Glycine on a Hydrophilic Amorphous Silica Surface. Phys. Chem. Chem. Phys. 2008, 10, 6360-6368. 47. Bauer, T.; Schmaltz, T.; Lenz, T.; Halik, M.; Meyer, B.; Clark, T., Phosphonate- and Carboxylate-Based Self-Assembled Monolayers for Organic Devices: A Theoretical Study of Surface Binding on Aluminum Oxide with Experimental Support. ACS Appl. Mater. Interfaces 2013, 5, 60736080. 48. Gao, W.; Robinson, R. L.; Gasem, K. A. M., High-Pressure Solubilities of Hydrogen, Nitrogen, and Carbon Monoxide in Dodecane from 344 to 410 K at Pressures to 13.2 MPa. J. Chem. Eng. Data 1999, 44, 130-132. 49. Madon, R. J.; Boudart, M., Experimental Criterion for the Absence of Artifacts in the Measurement of Rates of Heterogeneous Catalytic Reactions. Ind. Eng. Chem. Fundam. 1982, 21, 438447. 50. Pujari, S. P.; Scheres, L.; Marcelis, A. T. M.; Zuilhof, H., Covalent Surface Modification of Oxide Surfaces. Angew. Chem., Int. Ed. 2014, 53, 6322-6356. 51. Wan, H.; Vitter, A.; Chaudhari, R. V.; Subramaniam, B., Kinetic Investigations of Unusual Solvent Effects During Ru/C Catalyzed Hydrogenation of Model Oxygenates. J. Catal. 2014, 309, 174184.


ACS Catalysis 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

52. Jystad, A. M.; Biancardi, A.; Caricato, M., Simulations of Ammonia Adsorption for Characterization of Acid Sites in Metal-Doped Amorphous Silicates. J. Phys. Chem. C 2017, 121, 2225822267. 53. Tielens, F.; Folliet, N.; Bondaz, L.; Etemovic, S.; Babonneau, F.; Gervais, C.; Azaïs, T., Molecular Picture of the Adsorption of Ibuprofen and Benzoic Acid on Hydrated Amorphous Silica through DFT-D Calculations Combined with Solid-State NMR Experiments. J. Phys. Chem. C 2017, 121, 17339-17347. 54. Neouze, M. A.; Kronstein, M.; Litschauer, M.; Puchberger, M.; Coelho, C.; Bonhomme, C.; Gervais, C.; Tielens, F., Exploring the Molecular Structure of Imidazolium-Silica-Based Nanoparticle Networks by Combining Solid-State NMR Spectroscopy and First-Principles Calculations. Chem. Eur. J. 2014, 20, 15188-15196. 55. Folliet, N.; Roiland, C.; Bégu, S.; Aubert, A.; Mineva, T.; Goursot, A.; Selvaraj, K.; Duma, L.; Tielens, F.; Mauri, F.; Laurent, G.; Bonhomme, C.; Gervais, C.; Babonneau, F.; Azaïs, T., Investigation of the Interface in Silica-Encapsulated Liposomes by Combining Solid State NMR and First Principles Calculations. J. Am. Chem. Soc. 2011, 133, 16815-16827. 56. Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H., A Consistent and Accurate ab initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010, 132, 154104/1-154104/19. 57. Andrae, D.; Häußermann, U.; Dolg, M.; Stoll, H.; Preuß, H., Energy-Adjusted ab initio Pseudopotentials for the Second and Third Row Transition Elements. Theor. Chim. Acta 1990, 77, 123141. 58. Marenich, A. V.; Cramer, C. J.; Truhlar, D. G., Universal Solvation Model Based on Solute Electron Density and on a Continuum Model of the Solvent Defined by the Bulk Dielectric Constant and Atomic Surface Tensions. J. Phys. Chem. B 2009, 113, 6378-6396. 59. Goldsmith, B. R.; Sanderson, E. D.; Bean, D.; Peters, B., Isolated Catalyst Sites on Amorphous Supports: A Systematic Algorithm for Understanding Heterogeneities in Structure and Reactivity. J. Chem. Phys. 2013, 138, 204105/1-204105/11. 60. 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.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. . Gaussian 09, Revision A.02; Gaussian, Inc., Wallingford CT, 2016.


Page 22 of 23

Page 23 of 23 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

ACS Catalysis

TOC Graphic


A Silica Supported Molecular Palladium Catalyst for Selective

Recommend Documents