ReaxFF Simulations of Lignin Fragmentation on a Palladium-Based

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Surfaces, Interfaces, and Catalysis; Physical Properties of Nanomaterials and Materials

Experimental and Theoretical Studies of Lignin Fragmentation on a Palladium-Based Heterogeneous Catalyst in Methanol--Water Solution Susanna Monti, Pemikar Srifa, Ivan Kumaniaev, and Joseph S. M. Samec J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b02275 • Publication Date (Web): 21 Aug 2018 Downloaded from http://pubs.acs.org on August 21, 2018

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 22 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 Letters

ReaxFF Simulations of Lignin Fragmentation on a Palladium-Based Heterogeneous Catalyst in Methanol-Water Solution Susanna Monti,∗,† Pemikar Srifa,‡ Ivan Kumaniaev,‡ and Joseph S. M. Samec∗,‡ †CNR-ICCOM , Institute of Chemistry of Organometallic Compounds, via G. Moruzzi 1, I-56124 Pisa, Italy ‡Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, S-106 91 Stockholm, Sweden E-mail: [email protected]; [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 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

Abstract The interaction of fragments derived from lignin depolymerization with a heterogeneous palladium catalyst in methanol–water solution is studied by means of experimental and theoretical methodologies. Quantum chemistry calculations and molecular dynamics simulations based on the ReaxFF approach are combined effectively to obtain an atomic level characterization of the crucial steps of the adsorption of the molecules on the catalyst, their fragmentation, reactions and desorption. The main products are identified and the most important routes to obtain them explained through extensive computational procedures. The simulation results are in excellent agreement with the experiments and suggest that the mechanisms comprise a fast chemisorption of identified fragments from lignin on the metal interface accompanied by bond breaking, release of some of their hydrogens and oxygens to the support, eventual desorption depending on the local environment. The strongest connections are those involving the aromatic rings, as confirmed by the binding energies of selected representative structures, estimated at the quantum chemistry level. The satisfactory agreement with the literature, quantum chemistry data and experiments confirms the reliability of the multilevel computational procedure to study complex reaction mixtures and its potential application in the design of high performance catalytic devices.

Keywords: Lignin depolymerization, solvent effects, green chemistry, catalytic fractionation of biomass, ReaxFF, heterogenous catalysis


Page 2 of 22

Page 3 of 22 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


Graphical TOC Entry



Biopolymers are fundamental to support plants, to transport water and nutrients through the plant tissue, to protect the cell walls from the attack of chemical and microbial agents, and to act as cross-linkers between the cells. Lignin is one of these biopolymers, it is heterogeneous and do not have a definite primary structure. 1 Essentially, it is a complex racemic aromatic system interlinked by means of different motifs that largely depend on the type of wood. 2 The phenylpropane units are organized as three–dimensional amorphous polymeric networks with different degrees of aggregation. This vegetal biopolymer is an abundant form of biomass and, as such, represents a promising alternative to fossil fuels. Therefore, its conversion into valuable aromatic chemical species has recently attracted much attention as a means for reducing the fossil fuel dependence. 3 From a practical point of view, depolymerization of lignin is a very difficult process that requires specific reaction conditions and selected catalysts. The most common strategy adopted for this purpose is to break the chains acting on the weakest and more frequently appearing motif, namely the β-O-4’ alkyl–aryl ether linkage. 4,5 Reaction strategies using metal-based catalysts have been proposed but the experimental characterization and optimization of these reactions are very challenging6,7 and still need to be demonstrated. To this aim, computational studies has a great potential.8 Many theoretical investigations have focused on the characterization of the dynamics of long lignin chains at high temperature or complexed with other molecular species, 9–11 whereas other studies have been devoted to disclose the behavior of representative portions of this molecule in the gas phase, in specific environments (such as surrounded by oxygen species) or in contact with catalytic supports (in the gas phase).6,8,12–16 In these works the authors tried to identify important conditions, such as global and local perturbations, in elaborate environments devoid of solvent species. Unlike these aspects, the conformational variability of representative lignin fragments in solution or in contact with catalytic layers in the presence of solvent molecules has received relatively little attention. It is well known that solvent plays very often important roles in molecule self–assembly, adsorption on specific supports, 4 ACS Paragon Plus Environment

Page 4 of 22

Page 5 of 22 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


reaction mechanisms, etc., and as a consequence its involvement should be explored more deeply to design more efficiently tuned catalytic procedures. For example, the structure of technical lignins from pulping was hardly transformed and only a few changes by condensation reactions (by breaking the weak C-O bonds and forming the strong C-C bonds) were reported in the literature.17 Recently it has been shown that lignin fragments can be released and reductively transformed to less reactive monophenolic compounds by adding a transition metal to an organosolv pulping. 18–21 Furthermore, it was also observed that organosolv and hydrogenolysis transformations can be separated in time and space by applying a continuous flow where pulping occurs in the first reactor (percolator) and hydrogenolysis takes place in a second chamber 22 (see Figure 1). Through the continuous flow both processes can be optimized and studied separately: pulping gave both reactive monomeric and oligomeric species and the transition metal reduced the allylic double bond and reductively cleaved the β-O-4’ bond in the oligomers. However, the experimental characterization alone was not sufficient to disclose the reaction mechanisms at the molecular level because a heterogeneous catalyst is involved in these processes. The catalytic procedure we used for fractionating lignocellulose–based materials23 consists of a continuous flow–reaction system with two sequential reactors: a percolation chamber filled with woody biomass and a reactor chamber containing a palladium on carbon (Pd/C) catalyst, working at high temperature (453–493 K). The reaction is performed in methanolwater (7:3, v/v) solution. First, a partial depolymerization occurs in the percolation chamber without the action of the catalyst and then the products, which are phenolic monomers, dimers, and short oligomers24,25 are transported by the solvent to the second chamber, containing the palladium catalyst, and further converted. The main reactions observed are the cleavage of the β-O-4’ bond, the reduction of the double bond in the allylic alcohol, the dehydrogenation of the alcohols to generate hydrogen. These processes are highly dynamic and the final yield is heavily influenced by the solvent action, temperature, flow rate, and 5 ACS Paragon Plus Environment


Figure 1: Schematic depiction of the experimental process.

operation time. All these parameters should be tuned appropriately for efficient functioning and optimal performance. This could be realized through a deeper investigation of the reaction steps inside the cartridges by complementing the experimental tests with theoretical chemistry simulations. The power of the multilevel computational tools we have defined over the last few years, consisting of ReaxFF molecular dynamics and quantum chemistry (QC) calculations, has been demonstrated in a number of papers 26–28 and here it is used to mimic as closely as possible the catalytic environment inside the Pd/C cartridge. The computational models employed for this investigation are specifically designed to reproduce realistically key transformations of experimentally verified monophenolic compounds and also representative dimeric species comprising the β-O-4’ bond on a palladium interface in a mixed methanol-water solution at the experimental conditions (high temperature and pressure). ReaxFF parameters for these systems were not available, thus it was necessary to carry out several preliminary checks and 6 ACS Paragon Plus Environment

Page 6 of 22

Page 7 of 22 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


repeated parametrizations (using the fast and efficient Monte Carlo Force Field - MCFF parameter optimizer available in the Amsterdam Density Functional (ADF)29 package) in order to obtain a good candidate reactive force field (RFF) for the planned simulations. In summary, a series of conformational searches, using a molecular mechanics FF coupled with a genetic algorithm, were performed and then combined with quantum chemistry (QC) calculations (energy minimization at the M06-2X/6-31G(d) level) to identify all the conformations of different lignin–like species. These data plus other useful models, describing conformational transitions and simple reactions at the QC level, were collected (training set) and used to parametrize a RFF for depicting the behavior of these molecules in the gas phase. Subsequently, the conformational studies performed in the gas phase were complemented with MD simulations in a mixed methanol-water solution to explore solute-solvent interactions and solvation effects on the molecules. These studies were useful to slightly re-tune the RFF parameters. Then, the generated RFF was enriched with parameters able to describe the interactions of solute and solvent molecules with palladium substrates (low index surfaces). The initial values of all these parameters were taken from well tested ReaxFF force fields available in the literature.30,31 Essentially, the RFF was re–parametrized by adding other configurations to the training set (periodic QC calculations of the adsorption of representative fragments on small palladium slabs). The final RFF was validated against QC data not included in the parametrization. Geometries, binding energies and relative stability were satisfactorily reproduced even though some of the energy differences are overestimated by the force field (all the data used to parametrize the RFF are included in the Supporting Information). Closer inspection of those structures revealed that this was due to the presence of various intramolecular hydrogen bonds. The hydrogen bond term was not modified in the parametrization procedure because it was tuned for describing water structure. Then, this RFF was used for all the planned RMD simulations. In this first study, in line with the experimental needs, we have focused on the interaction 7 ACS Paragon Plus Environment


of two representative species from catalytic fractionation with a heterogeneous palladium surface: (i) dimeric species MOL1 comprising the β-O-4’ bond; and (ii) highly reactive monophenolic compound MOL2 (Figure 2). In (i) we have studied the Pd-catalyzed C–O bond cleavage and in (ii) we have studied the reduction of the double bond. These transformations are key processes in the catalytic fractionation of lignocellulose. The model reactor was built by combining the equilibrated solvent boxes containing the solutes (eight molecules in a 82 × 82 × 76 ˚ A3 solvent box), which were extracted from the RMD trajectories of MOL1 and MOL2 in methanol-water solution, with a Pd(111) (10 layers of 90 atoms) or Pd(100) (15 layers of 50 atoms) slab (box sizes: 25.9 × 25.9 × 70 ˚ A3 or 19.4 × 25.9 × 70 ˚ A3 ) by placing the solvent+solute boxes on top of the slabs in such a way that one of the solute molecules was accommodated close to the metal (at a distance of around 3.5 ˚ A). This was required to avoid long sampling times. Then, the resulting boxes were appropriately cut and reshaped to match the metal slab dimensions. RMD simulations were carried out in the NV T ensemble at T=453 K by means of the ReaxFF/ADF program. 29 Configurations were saved every 0.05 ps for a maximum simulation time of about 200 ps. Temperature was controlled through the Berendsen’s thermostat with a relaxation constant of 0.1 ps and the time step was set to 0.25 fs. More than two hundreds simulations were carried out for describing the adsorption of MOL1 and MOL2 on both slabs in methanol-water solution and only the most profitable results are reported and commented in this manuscript. In all cases, after an equilibration time (around 25 ps) the molecules reached the surface by extending, first, the side chains towards the Pd atoms and then, if the surrounding environment was favorable (i.e. the surface was not occupied by the solvent), adsorbing the aromatic rings, which were promptly chemisorbed. Indeed, the dominant interactions were those related to bond terms between the carbon atoms and the Pd sites. Visual inspection of the trajectories revealed that all the solutes, i.e. MOL1 and MOL2, had to compete with the solvent to reach

the surface. This is because both water and 8

ACS Paragon Plus Environment

Page 8 of 22

Page 9 of 22 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


Figure 2: Schematic depiction of the catalyst cartridge used in experiments. On the left-hand side: original molecules. On the right-hand side: identified products.

methanol molecules were quickly adsorbed on the interface, through their hydrogen and oxygen atoms. Such a behavior is apparent in Figure S1 where atom–atom radial distribution functions of Pd and MOL1 oxygens with solvent oxygens are shown together with the typical solvation of the two surfaces (blue patches). It is apparent that methanol molecules are adsorbed readily at shorter separations and can be both physisorbed and chemisorbed.32 This confirms the importance of the solvent influence on solute adsorption and fragmentation on the catalytic support and thus the need of an appropriate choice of the methanol-water ratio for obtaining specific compounds in high yields (see Ref.23 and references therein). Due to the presence of the adsorbates (MOL1 or MOL2), solvent coverage (calculated considering solvent molecules in close contact with the surface and the projection of their solvent accessible surface area on top of the Pd average plane) was found in the range 82-95% (for both Pd(111) and Pd(100) faces) and the adsorbates occupied at most 18% of the surface. Another common effect is the release of the solute/solvent hydrogen atoms to the metal sup9 ACS Paragon Plus Environment


ports and their migration to the second and third metal layers. This phenomenon was more pronounced for the Pd(111) surface where the population of the migrated atoms was around 5%, whereas for the Pd(100) slab the subsurface hydrogen population was around 2%. Such a behavior is known and depends on the characteristics of the surfaces and has widely been studied both experimentally33–35 and theoretically.36,37 Essentially, the simulation results confirm the easy permeation of these atoms into the subsurfaces but also reveal their ability to re-emerge to the interface, as it will be shown in the case of MOL2 hydrogenation. As far as the adsorption of MOL1 is concerned, it was noticed that the starting orientation was not fundamental for determining the fragmentation of the molecule. Indeed, after a few ps the molecule became chemisorbed on the surface and these tighter contacts were sufficient to induce a further elongation of the C–O bond, which was already slightly extended due to the effect of the high temperature. It is apparent from Figure 3a that the topology of the interface could have an effect on the bond breaking time, which was fractions of picosecond for both surfaces but longer in the the case of the Pd(111) (around 0.6 ps) one. These time differences are not sufficient to discriminate different catalytic behaviors but could be ascribed to the topology of the Pd(100) facet which is more open and binds intermediate species more strongly. 38 In Figure 4 the potential energy of the Pd(111)+MOL1+solvent system as a function of the simulation time is shown together with selected structures relevant for disclosing the various steps of the mechanism. Inspection of the initial configurations (Figures 4A, B, C) and the potential energy trend reveals that, first, the physisorbed structure tries to find a more convenient arrangement on the surface reorienting the Cα hydroxyl group (lone pairs pointing on top of a palladium atom - decreasing energy), then this non-bonding interaction is transformed in a chemisorption by the release of the hydrogen atom (Figures 4D) to a nearby atom of the surface. Such an action is accompanied by a potential energy increase, which is essentially due to the addition of important energy bonding terms. This is con10 ACS Paragon Plus Environment

Page 10 of 22

Page 11 of 22 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


Figure 3: Evolution of the C–O bond length during the dynamics of adsorption of MOL1 on Pd(111) (black dots) and Pd(100)a,b (red and blue dots) surfaces (the results of two different simulations shown in the SI movies MOL1onPd100a.mpg and . MOL1onPd100a.mpg are displayed).



firmed by the examination of the trend of the energy components displayed in Figure S2. The whole process is very fast (about 10 ps) and is followed by a further slow equilibration that comprises the breakage of the lignin ether linkage (β-O-4’ linkage), visible in Figures 4I, and the relocation of the resulting fragments, namely FRAG1, which detaches from the surface and goes into the solvent, and FRAG3, which maintains the original strong adsorption on the support. In Figures 3b and 4 it is also apparent that after bond breaking the two fragments could remain quite close to each other in a stacking orientation, before separating into longer distances (> 10 ˚ A - Figure 3b black and blue dots) (the three movies depicting the bond breaking dynamics on both Pd(111) – MOL1onPd111.mpg – and Pd(100) – MOL1onPd100a.mpg, MOL1onPd100b.mpg – are available in the Supporting Information). The adsorption mechanism of MOL1 on top of the Pd(100) surface and the relative potential energy trends are very similar to the dynamics on Pd(111). The final products depend on local environmental condition and solvent effects, they can be found in solution or adsorbed on the surfaces in various arrangements characterized by a variable number of contacts. The study of MOL2 was focused on the possible routes to reduce the double bond (and reproduce the protonation of the carbon atoms without imposing constraints on the dynamics). We carried out several simulations on both slabs starting from a great variety of orientations of the molecule in relation to the palladium surfaces. By finding correct alignment of the molecule on the Pd(100) slab, we managed to observe the whole process of hydrogen transfer to the double bond. This was a result of a strong coordination of the whole molecule (both aryl and propenyl chain) to the palladium surface. The elongation of the C–C double bond as a function of the simulation time is displayed in Figure 5. The process started when the molecule became chemisorbed through the ring. These interactions forced a readaptation of the other atoms of the chain to the topology of the substrate in search of the best attachment compatible with the local environment. Rotation of the 12 ACS Paragon Plus Environment

Page 12 of 22

Page 13 of 22 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


Figure 4: Structures extracted from the simulation of the adsorption of MOL1 on Pd(111). Only the atoms/molecules within 3.5 ˚ A of MOL1 are displayed. Evolution of the potential energy of the system during the dynamics.



Figure 5: Elongation of the C-C bond as a function of the simulation time.


Page 14 of 22

Page 15 of 22 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


terminal oxygen in between two rows of palladium atoms joint with the chemisorption of the connected C–C bond induced the pyramidalization of the nearby carbon. Then, a subsurface hydrogen became bonded to this site (movie MOL2onPd100.mpg in the Supporting Information). Once started, the whole mechanism was quite fast, around 30 ps (Figure 5). In order to give an estimate of the relative strength of the molecule–surface associations we extracted the stabilized structures of the adsorbates and carried out periodic QC calculations after reducing the system size to render the calculations computationally feasible. Essentially, we removed all the solvent molecules and replaced the slab with the one used in the parametrization process (the details can be found in the Supporting Information). The comparison of the binding energies of the initial (double bond) and final (single bond +hydrogen) species, calculated at the QC level, showed that the hydrogenation of the molecule produced a less strongly adsorbed monophenolic compound (by about 54 kcal/mol) (the geometries used for the calculations are available in the Supporting information). Such a finding suggests that these types of products could have a higher probability to be found in solution. However, the calculated BEs that are around -101 and -47 kcal/mol, for the original molecule and the hydrogenated species, respectively, are rough estimates and could be greatly overestimated because they do not take into account short range intermolecular interactions between the adsorbate and the various species close by. The BEs calculated by means of the ReaxFF approach, for reactant and product, are -109.4 and -99.8 kcal/mol, respectively. These values well compare with the QC data even though the BE of the product seems greatly overestimated. Most probably this is due to product species that cannot easily defined for the QC calculations. Furthermore, molecular self–aggregation on the surface, interactions with the adsorbed solvent molecules and the action of the surrounding solution are all important perturbations that contribute to weaken the Pd–molecule attachment.39 As a consequence, we could consider the calculated values as an upper limit of the ”real ” BE, which cannot be estimated unequivocally also from an experimental point of view, given the 15 ACS Paragon Plus Environment


complexity of the whole processes. Besides, the continuous flow conditions represent another variable that should be taken into account. Regarding the BEs of other adsorbed fragments it was found that they were spread on a large range of values (-150 to -24 kcal/mol) that depended on very slight variations of the structures and the surrounding system components. From a computational point of view these data are in line with the energies of the original parametization confirming the correct behavior of the force field to reproduce the adsorption on Pd surfaces. In conclusion, ReaxFF MD simulations based on the new parametrized FF for describing lignin-based compounds in solution in contact with a heterogenous palladium catalyst were capable of reproducing the experimental observations by disclosing, at the atomic level, fundamental details of the adsorption dynamics, C–O bond cleaving events of a model oligomer (MOL1), the reduction of a reactive monomer (MOL2) and the desorption dynamics of the generated fragments from both Pd(111) and Pd(110) surfaces in a solvent mixture. The adopted computational strategy successfully simulated the extended reaction network and identified several of the main products collected through experiments.

Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: Materials related to the force field parametrization, additional procedures and supporting data (PDF).

Acknowledgments S.M. is grateful to Adri C. T. van Duin for the stand-alone version of ReaxFF, for his support and collaboration. JSMS thanks the Energy Agency for financial support.


Page 16 of 22

Page 17 of 22 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


References (1) Sabu, T., Visakh, P. M., Mathew, A. P., Eds. Advances in Natural Polymers, Composites and Nanocomposites ; Advanced Structured Materials; Springer–Verlag Berlin Heidelberg, 2013; Vol. 18. (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. 2011, 110, 3552–3599. (3) Corma, A.; Iborra, S.; Velty, A. Chemical Routes for the Transformation of Biomass into Chemicals. Chem. Rev. 2007, 107, 2411–2502. (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) Rahimi, A.; Azarpira, A.; Kim, H.; Ralph, J.; Stahl, S. S. Chemoselective Metal–Free Aerobic Alcohol Oxidation in Lignin. J. Am. Chem. Soc. 2013, 135, 6415–6418. (6) Galkin, M. V.; Sawadjoon, S.; Rohde, V.; Dawange, M.; Samec, J. S. M. Mild Heterogeneous Palladium-Catalyzed Cleavage of β-O-4’-Ether Linkages of Lignin Model Compounds and Native Lignin in Air. ChemCatChem 2014, 6, 179–184. (7) Galkin, M. V.; Dahlstrand, C.; Samec, J. S. M. Mild and Robust Redox–Neutral Pd/C–Catalyzed Lignol β-O-4’ Bond Cleavage Through a Low–Energy–Barrier Pathway. ChemSusChem 2015, 8, 2187–2192. (8) Lu, J.; Wang, M.; Zhang, X.; Heyden, A.; Wang, F. β-O-4’ Bond Cleavage Mechanism for Lignin Model Compounds over Pd Catalysts Identified by Combination of First– Principles Calculations and Experiments. ACS Catal. 2016, 6, 5589–5598.



(9) Lo´pez-Albarrańa, P.; Pizzi, A.; Navarro-Santos, P.; Hernández-Esparza, R.; Garza, J. Oligolignols within Lignin–Adhesive Formulations Drive their Young’s Modulus: A ReaxFF–MD Study. Int. J. Adhesion and Adhesives 2017, 78, 227–233. (10) Beste, A. ReaxFF Study of the Oxidation of Softwood Lignin in View of Carbon Fiber Production. Energy Fuels 2014, 28, 7007–7013. (11) Beste, A.; Buchanan, A. C. Role of Carbon–Carbon Phenyl Migration in the Pyrolysis Mechanism of β–O–4’ Lignin Model Compounds: Phenethyl Phenyl Ether and α– Hydroxy Phenethyl Phenyl Ether. J. Phys. Chem. A 2012, 116, 12242–1224. (12) Yamaguchi, A.; Mimura, N.; Shirai, M.; Sato, O. Bond cleavage of lignin model compounds into aromatic monomers using supported metal catalysts in supercritical water. Scientific Reports 2017, 7, 46172. (13) Beste, A.; Buchanan, A. C. Computational Investigation of the Pyrolysis Product Selectivity for α–Hydroxy Phenethyl Phenyl Ether and Phenethyl Phenyl Ether: Analysis of Substituent Effects and Reactant Conformer Selection. J. Phys. Chem. A 2013, 117, 3235–3242. (14) Beste, A.; Buchanan, A. C. Kinetic Analysis of the Phenyl-Shift Reaction in β-O-4’ Lignin Model Compounds: A Computational Study. J. Org. Chem. 2011, 76, 2195– 2203. (15) Beste, A. ReaxFF Study of the Oxidation of Lignin Model Compounds for the Most Common Linkages in Softwood in View of Carbon Fiber Production. J. Phys. Chem. A 2014, 118, 803–814. (16) Elder, T.; Beste, A. Density Functional Theory Study of the Concerted Pyrolysis Mechanism for Lignin Models. Energy Fuels 2014, 28, 5229–5235. 18 ACS Paragon Plus Environment

Page 18 of 22

Page 19 of 22 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


(17) Löfstedt, J.; Dahlstrand, C.; Orebom, A.; Meuzelaar, G.; Sawadjoon, S.; Galkin, M. V.; Agback, P.; Wimby, M.; Corresa, E.; Mathieu, Y. et al. Green Diesel from Kraft Lignin in Three Steps. 2016, 9, 1392–1396. (18) Rinaldi, R.; Jastrzebski, R.; Clough, M. T.; Ralph, J.; Kennema, M.; Bruijnincx, P. C.; Weckhuysen, B. M. Paving the Way for lLignin Valorisation: Recent Advances in Bioengineering, Biorefining and Catalysis. Angew. Chem. Int. Ed. 2016, 55, 8164–8215. (19) Galkin, M. V.; Samec, J. S. M. Lignin Valorization through Catalytic Lignocellu- lose Fractionation: a Fundamental Platform for the Future Biorefinery. ChemSusChem 2016, 9, 1544–1558. (20) 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. (21) Sun, Z.; Fridrich, B.; de Santi, A.; Elangovan, S.; Barta, K. Bright Side of Lignin Depolymerization: Toward New Platform Chemicals. Chem. Rev. 2018, 118, 614–678. (22) Anderson, E. M.; L.Stone, M.; Katahira, R.; Reed, M.; Beckham, G. T.; RomanLeshkov, Y. Flowthrough Reductive Catalytic Fractionation of Biomass. Joule 2017, 1, 613–622. (23) Kumaniaev, I.;

Subbotina, E.;

Sa¨vmarker, J.;

Larhed, M.;

Galkin, M. V.;

Samec, J. S. M. Lignin Depolymerization to Monophenolic Compounds in a Flow– Through System. Green Chem. 2017, 19, 5767–5771. (24) Huang, X.; Ouyang, X.; Hendriks, B.; Morales, O.; Zhu, J.; Korańyi, T. I.; Boot, M.; Hensen, E. J. M. Selective Production of Mono-Aromatics from Lignocellulose over Pd/C Catalyst: the Influence of Acid Co-Catalysts. Faraday Discuss. 2017, 202, 141– 156. 19 ACS Paragon Plus Environment


(25) Renders, T.; Van den Bosch, S.; Vangeel, T.; Ennaert, T.; Koelewijn, S. F.; Van den Bossche, G.; Courtin, C. M.; Schutyser, W.; Sels, B. F. Synergetic Effects of Alcohol/Water Mixing on the Catalytic Reductive Fractionation of Poplar Wood. ACS Sustainable Chem. Eng. 2016, 4, 6894–6904. (26) Monti, S.; Carravetta, V.; ˚ Agren, H. Simulation of Gold Functionalization with Cysteine by Reactive Molecular Dynamics. J. Phys. Chem. Lett. 2016, 7, 272–276. (27) Monti, S.; Carravetta, V.; ˚ Agren, H. Theoretical Study of the Adsorption Mechanism of Cystine on Au(110) in Aqueous Solution. Small 12, 2016, 6134–6143. (28) Zhu, C.; Monti, S.; Mathew, A. P. Cellulose Nanofiber–Graphene Oxide Biohybrids: Disclosing the Self–Assembly and Copper–Ion Adsorption Using Advanced Microscopy and ReaxFF Simulations. ACS Nano 2018, 12, 7028–7038. (29) van Duin, A. C. T.; Goddard, W. A.; Islam, M. M.; van Schoot, H.; Yakovlev, A. L. ReaxFF 2017, SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam, The Netherlands. http://www.scm.com 2017, (30) Senftle, T. P.; Meyer, R. J.; Janik, M. J.; van Duin, A. C. T. Development of a ReaxFF Potential for Pd/O and Application to Palladium Oxide Formation. J. Chem. Phys. 2013, 139, 044109. (31) Senftle, T. P.; Janik, M. J.; van Duin, A. C. T. A ReaxFF Investigation of Hydride Formation in Palladium Nanoclusters via Monte Carlo and Molecular Dynamics Simulations. J. Phys. Chem. 2014, 118, 4967–4981. (32) Lichtenberger, J.; Lee, D.; Iglesia, E. Catalytic Oxidation of Methanol on Pd Metal and Oxide Clusters at Near-Ambient Temperatures. Phys. Chem. Chem. Phys. 2007, 9, 4902–4906. 2 ACS Paragon Plus Environment

Page 20 of 22

Page 21 of 22 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


(33) Monguchi, Y.; Ichikawa, T.; Sajiki, H. Recent Development of Palladium-Supported Catalysts for Chemoselective Hydrogenation. Chemical and Pharmaceutical Bulletin 2017, 65, 2–9. (34) Ohno, S.; Wilde, M.; Mukai, K.; Yoshinobu, J.; Fukutani, K. Mechanism of Olefin Hydrogenation Catalysis Driven by Palladium-Dissolved Hydrogen. J. Phys. Chem. C 2016, 120, 11481–11489. (35) Wilde, M.; Fukutani, K. Penetration mechanisms of surface-adsorbed hydrogen atoms into bulk metals: Experiment and model. Phys. Rev. B 2008, 78, 115411. (36) Gross, A. Ab initio Molecular Dynamics Simulations of the Adsorption of H2 on Palladium Surfaces. Chemphyschem 2010, 17, 1374–1381. (37) Verma, A. M. Molecular Simulations of Palladium Catalysed Hydrodeoxygenation of 2-Hydroxybenzaldehyde using Density Functional Theory. Phys. Chem. Chem. Phys. 2017, 19, 25582–25597. (38) Plaucka, A.; Stanglandb, E. E.; Dumesica, J. A.; Mavrikakisa, M. Active Aites and Mechanisms for H2O2 Decomposition over Pd Catalysts. PNAS 2016, 113, E1973– E1982. (39) Froment, G. F.; Waugh, K. C. In Reaction Kinetics and the Development and Operation of Catalytic Processes ; Elsevier,, Ed.; Elsevier Science, 2001; Vol. 133.



TOC. Representation of lignin depolymerization

ACS Paragon Plus Environment

Page 22 of 22

ReaxFF Simulations of Lignin Fragmentation on a Palladium-Based

Recommend Documents