Diffusion of Biomass Pyrolysis Products in H-ZSM ... - ACS Publications

Subscriber access provided by University of Colorado Boulder

Article

Diffusion of Biomass Pyrolysis Products in HZSM-5 by Molecular Dynamics Simulations Lintao Bu, Mark R Nimlos, David J Robichaud, and Seonah Kim J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b10871 • Publication Date (Web): 13 Dec 2016 Downloaded from http://pubs.acs.org on December 16, 2016

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

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

Page 1 of 32

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

Diffusion of Biomass Pyrolysis Products in H-ZSM-5 by Molecular Dynamics Simulations Lintao Bu, Mark R. Nimlos, David J. Robichaud, Seonah Kim*

National Bioenergy Center, National Renewable Energy Laboratory, 15013 Denver West Pkwy, Golden, CO, 80401

Abstract

Diffusion of biomass pyrolysis vapors and their upgraded products is an

essential catalytic property of zeolites during catalytic fast pyrolysis and likely plays a critical role in the selectivity of these catalysts. Characterizing the diffusivities of representative biofuel molecules is critical to understand shape selectivity and interpret product distribution. Yet, experimental measurements on the diffusivities of oxygenated biofuel molecules at pyrolysis temperatures are very limited in the literature. As an alternative approach, we conducted MD simulations to measure the diffusion coefficients of several selected molecules that are representative of biomass pyrolysis vapors, namely water, methanol, glycolaldehyde, and toluene in H-ZSM-5 zeolite. The results show the diffusion coefficients calculated via MD simulations are consistent with available NMR measurements at room temperature.1 The effect of molecular weight and molecular critical diameter on the diffusivity among the chosen model compounds is also examined. Furthermore, we have characterized the diffusivities of representative biofuel molecules, namely xylene isomers, in H-ZSM-5. Our calculations determined that the ratio of the diffusion coefficients for xylene isomers is p-xylene : o-xylene : m-xylene ≈ 83 : 3 : 1 at 700 K. Additionally, our results also demonstrate the different diffusivity between p-

ACS Paragon Plus Environment

1


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

Page 2 of 32

xylene and toluene is due to the molecular orientations when the molecules diffuse along the channels in H-ZSM-5 and provides deep insight into the effect of molecular orientation on its diffusivity. Introduction Lignocellulosic biomass is a promising sustainable and renewable feedstock that could potentially substitute petroleum produced transportation fuels in the near term.2 Fast pyrolysis of lignocellulosic biomass is an encouraging thermal-chemical conversion technology to produce pyrolysis oil, and further selectivity can be obtained by upgrading the vapors before condensation. This is commonly referred to as catalytic fast pyrolysis (CFP) and zeolite catalysts can be used to decrease the oxygen contents and produce aromatic hydrocarbons.3-5 During this conversion process, diffusion of pyrolysis vapors and their upgraded products in zeolites is an essential phenomenon.5-7 Diffusion can be a primary contributor to coke formation,8 longevity of the catalyst,9 and product selectivity10 and separations.11 Therefore, a deeper understanding of the effects of pore size and shape on diffusivity could facilitate the design of new zeolite catalysts to enhance the product selectivity and separation. While correlations and heuristic models of diffusion exist for larger scales (e.g. Knudsen12), the relative size of the microporous zeolite structure and the pyrolysis vapors places the system well within the configurational regime,13 where such correlations do not exist. Therefore, investigating this phenomenon is limited to experimental and advanced computational methods.

Unfortunately, experimental measurements on diffusivities of pyrolysis vapors


2

Page 3 of 32

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


and their upgraded products in zeolites are very limited in the literature and the diffusion coefficients measured by different experimental techniques display significant discrepancies.1,14-16 For instance, Mirth et al. used time-resolved in situ FTIR spectroscopy to investigate the diffusion of xylene isomers over H-ZSM-5 and reported that the ratio of the diffusion coefficients for the three xylene isomers is p-xylene : oxylene : m-xylene ≈ 1000 : 10 : 1 at 373 K using macroscopic techniques.17 Specifically, the diffusion coefficient for p-xylene was reported to be 6 × 10-16 m2 s-1. Choudhary et al. also studied the diffusion of xylene isomers in H-ZSM-5 via volumetric measurements and found the diffusion coefficient for p-xylene to be 5.13 × 10-16 m2 s-1 at 308 K,14 in good agreement with the work of Mirth et al. Han et al. used the zero length column method to measure the intracrystalline diffusivity of p-xylene in H-ZSM-518 and reported the diffusion coefficient as 4000 × 10-16 m2 s-1 at 373 K – a 3 order of magnitude difference. The same is true for other important biomass pyrolysis species, such as nhexane,19-22 benzene,14,22-24 and ethylbenzene.14,24 In summary, the measured diffusion coefficients for biomass relevant species in H-ZSM-5 can vary by orders of magnitude. The discrepancies among the experimentally measured diffusion coefficients are partially caused by the different experimental technologies.7,25 The macroscopic methods, e.g., gravimetric uptake rate technique, probe the mass transfer between the gas phase and the zeolite, then solves Fick’s law of diffusion to acquire the diffusion coefficients.26 Alternatively, microscopic techniques, e.g., pulse field gradient NMR (PFG-NMR)1,27,28 and quasi-elastic neutron scattering (QENS)20,29,30, track the mean-square displacement of a labeled molecule in a zeolite, then use Einstein’s equation to compute the self-diffusion coefficient. A limitation of the experimental technology is that these measurements


3


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

Page 4 of 32

usually are conducted at low temperatures, since many pyrolysis vapors will react with activated zeolites at

the pyrolysis relevant temperatures, making diffusivity

measurements under these conditions difficult. As an alternate approach, molecular dynamics (MD) simulations are able to investigate the dynamics of molecules in zeolites spanning the range of temperatures31 relevant to thermochemical conversion of biomass. Thus these simulations provide essential understanding of diffusivities of pyrolysis vapors and their upgraded products at pyrolysis temperature.32-36 For instance, the diffusion of xylene isomers in CIT-137, HZSM-5

38,39

, H-BEA 38-40, and NaY zeolites

41,42

have previously been investigated using

MD simulations. Llopis and co-workers reported the self-diffusion coefficients of pxylene, o-xylene, and m-xylene in H-ZSM-5 at 650 K are 40.76, 5.54, and 1.38 × 10-10 m2 s-1, respectively, while the self-diffusion coefficients of these isomers in H-BEA are 87.02, 41.29, and 53.32 × 10-10 m2 s-1, respectively.38,39 Schrimpf et al. studied the diffusion of p-xylene in zeolite Na-Y and reported a diffusion coefficient value of 4.50 × 10-10 m2 s-1 at 650 K.42 These calculations are controversial since the pore size and channel size of Na-Y is much larger than that of H-ZSM-5 and H-BEA and therefore it would be expected to observe a much faster diffusion rate in Na-Y. The discrepancies could be due to the use of different force fields and varying simulation times, which range from a few hundred ps38,39 to 100 ns42. Quite recently, Toda et al. investigated the effect of force fields on the diffusion of p-xylene and o-xylene in several zeolites including H-ZSM-5. Their results for the self-diffusion coefficient of p-xylene in HZSM5 varied from 1.30 to 27.47 × 10-10 m2 s-1, demonstrating the sensitivity of these simulation results on force fields.43 However, choosing an accurate force field for a


4

Page 5 of 32

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


specific zeolite system is not straightforward and, in general, the force fields developed for one zeolite are not transferable to other zeolites.44 Thus, to validate whether a force field is suitable for a selected zeolite is to compare the measured diffusion coefficients of benchmark molecules via MD simulations at relevant temperatures to reliable experimental measurements (e.g. PFG-NMR). In this work, we have conducted MD simulations to investigate the diffusivities of few chosen substrates relevant to catalytic fast pyrolysis. We first validate our approach and the chosen force fields on water, methanol, glycolaldehyde, and toluene in H-ZSM5 zeolite to compare with available PFG-NMR results.1 We then use this approach to investigate the diffusion behaviors of representative biofuel molecules, xylene isomers, including p-xylene, o-xylene and m-xylene to interpret the product distribution at the molecular level. Specifically, p-xylene is an important and more desirable product than the other two xylene isomers, since p-xylene can be used to produce terephthalic acid and polyethylene

terephthalate

(PET).45

Additionally,

experimental

studies

have

demonstrated that p-xylene molecule can be selectively produced over other xylene isomers using H-ZSM-5 zeolite during CFP of biomass46 or catalytic upgrading of biomass derived furans.47 Methods Models. H-ZSM-5 is a medium pore zeolite, consisting of straight channels (5.3 × 5.6 Å) along the b direction and zigzag channels (5.1 × 5.5 Å) along the a direction.48 The super cell of H-ZSM-5 (1 × 1 × 2 unit cell) includes 576 atoms. Six Si atoms at T12 sites are replaced by Al atoms, resulting in a Si/Al ratio of 31. Replacing of Si atoms by Al atoms and adding proton atoms to balance the charge and create Brønsted acid sties in the


5


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

Page 6 of 32

zeolite, as shown in Figure 1. We note H-ZSM-5 zeolite with a typical Si/Al ratio of 30 has been used in previous experimental studies to deconstruct biomass feedstock49 and upgrade biomass-derived oxygenates.50,51 A H-ZSM-5 zeolite with a Si/Al ratio of 35.5 (equivalent to 2.6 Al atoms/unit cell) was also used by Mirth et al. in their time-resolved in situ FTIR spectroscopy study as mentioned previously.15 The substrate loadings vary depending on the size of the molecules. Specifically, the loadings are 24 for water, 12 for methanol, ethanol, and glycolaldehyde, and 6 for toluene and xylene isomers.

Figure 1. The unit cell used to simulate H-ZSM-5 shown along the straight channel (A) and the zigzag channel (B). The six T12 sites (Brønsted acid sites) are shown in spheres (iceblue: Al atoms, red: O atoms, white: H atoms). The thick stick model shows the 10member ring in the straight and zigzag channel. Simulation protocols. The MD simulations were performed using the LAMMPS program.52 Each system undergoes 500,000 steps of equilibration, followed by 40 ns of production runs in the canonical ensemble, using a Nosé -Hoover thermostat.53 All the atoms, including the zeolite framework, are flexible. A cutoff radius of 10 Å was used for Lenard-Jones potentials. Ewald summation was applied in the electrostatics calculations.


6

Page 7 of 32

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


Awati et al. used a time step of 0.1 fs and 1 fs for a flexible and a rigid framework, respectively. Accordingly, a time step of 0.1 fs was used in this work to simulate the flexible zeolite framework. Three independent MD simulations are performed for each system for better statistics. Measurement of diffusion coefficient. The mean square displacement (MSD) of substrate molecules was recorded every 1000 steps (100 fs). These data were utilized to compute self-diffusion coefficient (DS) using Einstein relation

= lim →

〈| − 0| 〉

(1)

where r(t) indicates the coordinates of the substrate molecule at time t. At low temperatures, we made sure that the minimum MSD accomplished is at least the square of the distance between two adjacent intersections, i.e. 100 Å2 in H-ZSM-5. To achieve this minimal requirement, we found 40 ns timescale is necessary for o-xylene and mxylene molecules. Molecular descriptor for diffusivity. Two different molecular descriptors, the kinetic diameter (Km)1 and the critical diameter,14 which have been used in previous studies to describe the diffusion, were compared in this work. The kinetic diameter (Km) was calculated using the following equation: " = 1.234

!

(2)

where MW denotes the molecular weight of the substrate. The critical diameter is defined as the diameter of the minimum cylinder that can enclose the molecule at its most stable geometry,14 which is calculated using Gaussian 09 programs54 with M06-2X functional55 and 6-311++g(d,p) basis set. The optimized molecule is oriented with the three principal axes aligned in the x-, y-, and z-directions using the VMD program.56 The minimum


7


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 8 of 32

distances along these three directions are measured and a correction of the vdW radius of H-atom was added to these distances to take into account the molecular surface. The widely used vdW radius of H-atom is 1.2 Å, thus a value of 2.4 Å was added to each measured distance. Since the longest principal axis can be aligned parallel to the direction of the channel, the second longest principal axis is chosen as the critical diameter of the molecule. Force fields.

Various force fields have been used in the previous studies on

HZSM-5.34,44 The simulation protocols employed in the current work closely follows that of Awati and coworkers.57 The Hill-Sauer force field has been demonstrated to be able to reproduce the lattice parameters and geometries of H-ZSM-558,59and thus was used for the zeolite in this work. We note that the force field employed in Awati and coworkers57 is only for pure silica zeolites. To simulate aluminosilicates in this work, we have incorporated additional parameters from the published Hill-Sauer force field58,59 to describe the Brønsted acid sites which include bond, angle, torsion and all associated cross terms. All potential parameters for zeolite used in this work are summarized in Table S1 (Supporting Information). The TraPPE force field60, which has been shown to work well when simulating the adsorption and diffusivity of alcohol molecules in zeolites,61-63 was used for the methanol, ethanol, and glycolaldehyde molecules. The OPLS force field64, which has been applied to study the diffusion of alkanes65-67 and cyclic hydrocarbons,68-70 was used for the aromatic hydrocarbons. The calculated self-diffusion coefficients (Ds) of benchmark molecules, namely water, methanol, ethanol, glycolaldehyde, and toluene via MD simulations are shown in Table 1. The MD results show good agreement with previous studies in the literature.1,71


8

Page 9 of 32

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


Specifically, the calculated diffusion coefficients of water, methanol, and toluene molecules via MD simulations are in very good agreement with NMR measurements of Cheah et al.,1 which lends strong support to the validation of the force fields used in this work. As shown in Table 1, the only exception between MD simulations and NMR measurement is glycolaldehyde, which will be discussed in detail below. Table 1. Comparison of calculated DS with previous studies. DS (10-10 m2 s-1, 300K)

a

Literature

This work

Water

6.3a

5.5 ± 0.5

Methanol

1.6a

2.1 ± 0.1

Ethanol

1.2b

1.7 ± 0.2

Glycolaldehyde

0.12a

0.97c / 0.066d

Toluene

0.051a

0.052 ± 0.013

Cheah et al..1 bYang et al.71 cglycolaldehyde monomer dglycolaldehyde dimer

Glycolaldehyde is commercially available as a solid dimer, which can produce monomers, linear and cyclic dimers, and trimers upon dissolution in water at room temperature.72 Thus, the measured diffusion coefficient for glycolaldehyde (0.12 × 10-10 m2 s-1) at 300 K by Cheah et al. using the PFR-NMR technique is an overall average of the various species. Indeed, they noted that the slow diffusivity identified for glycolaldehyde could be caused by the larger glycolaldehyde dimer. However, at the pyrolysis temperatures the glycolaldehyde monomer is the thermodynamically preferred species and therefore the diffusion should be much faster than indicated by the NMR


9


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 32

measurement at room-temperature. To characterize the realistic diffusional behaviors of glycolaldehyde molecule, we have measured the diffusion coefficients for glycolaldehyde monomer and a cyclic glycolaldehyde dimer (1,4-Dioxane-2,5-diol). As shown in Table 1, the calculated diffusion coefficient for the monomer and the dimer is 0.97 and 0.066 × 10-10 m2 s-1 at 300 K, respectively. The NMR measured diffusion coefficient of glycolaldehyde species is slower than that of a glycolaldehyde monomer, but faster than that of a glycolaldehyde dimer. Since glycolaldehyde will dominate as monomers at pyrolysis temperature, we argue our measurement represents true diffusivity of glycolaldehyde molecule, which is beyond the ability of NMR measurement. Results and Discussions Comparison of molecular descriptors

As shown in Figure 2, there is a

robust correlation of the measured diffusion coefficients of chosen molecules on their kinetic diameters. The diffusion coefficients of water, methanol, ethanol, glycolaldehyde monomer, and toluene reduce linearly with increasing the kinetic diameters. The results also indicate diffusivity increases significantly with increasing temperature. The diffusion coefficients can increase by ~20-30 times by increasing temperature from 300 K to 700 K. Thus, NMR measured diffusivities at room temperature will not represent the realistic diffusion behaviors of pyrolysis vapors at fast pyrolysis temperatures, varying from 673 to 823 K.73 The precise determination of a molecular descriptor that can be used to describe diffusivity is essential, but challenging. We note the kinetic diameter defined in equation 2 only takes into account the effect of molecular weight, while ignoring other important


10

Page 11 of 32

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


properties of a molecule that should affect diffusivity, such as shape and polarity. For instance, kinetic diameter is not able to distinguish the enormous difference of the diffusivities among xylene isomers, since their molecular weight is the same.

Figure 2. Diffusion coefficients (Ds) vs kinetic diameter for benchmark molecules. GA = glycolaldehyde monomer. Additionally, the effect of critical diameter on the diffusivities of chosen model compounds is also investigated to compare with the kinetic diameter Km. As shown in Figure 3, the critical diameter as the molecule descriptor moderately improved the linear relationship between the diffusion and the molecule descriptor. The value of R2 increases from 0.76 to 0.88, indicating the molecular size plays essential larger role than the molecular weight in determining the diffusivity of these chosen molecules in H-ZSM-5. However, the critical diameter has its own weakness. For instance, critical diameter still does not distinguish the o-xylene and m-xylene correctly (although it does discriminate pxylene better). Furthermore, critical diameter will not be able to distinguish the different diffusivities between toluene and p-xylene molecules, which share the same critical diameter. Considering the benefits and shortcomings of both kinetic diameter and critical


11


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 32

diameter, it would be worthwhile to develop a novel molecular descriptor that combines the kinetic diameter and the critical diameter which might improve its prediction ability to characterize diffusivity. For instance, the diffusion of a molecule in the gas phase can be described as:

(")"

= #$ % & ' $

(3)

where k is the Boltzmann constant, P is the pressure, T is the temperature, d is the diameter of the molecule, and m is the molecular weight. We found a simple molecular descriptor with the form of * √, can slightly improve the linear relationship to a R2 value of 0.91 (shown in Figure S1). However, we note that molecular weight and critical diameters are not the only factors that can influence the diffusion. For instance, both factors can not explain the different diffusivities between toluene and p-xylene molecules. In that case, the molecular orientations need to be taken into account, which will be discussed in detail later. Furthermore, we note that developing such a correlation will require more data points than what are available in Figure 3. As a summary, developing an accurate descriptor is very challenging, and is beyond the scope of current study.


12

Page 13 of 32

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. Diffusion coefficients (Ds) as a function of kinetic diameter and critical diameter.

Diffusion of xylene isomers

It has been demonstrated previously that the

intersection regions between the straight channels and the zigzag channels are the most stable adsorption sites for benzene, toluene, and xylene molecules in H-ZSM-5.74 Figure 4 illustrates the positions, i.e., the x and y coordinates of xylene molecules during the 40 ns MD simulation. The trajectories demonstrate that p-xylene molecules have visited all the four intersections in a given time and spent the majority of time at the intersection region compared to within the channels. Alternatively, m-xylene and o-xylene molecules have only sampled two intersection regions and show little probability of being found within the channels between intersections. Similarly, Figure S2 demonstrates the positions of water molecules diffusing in H-ZSM-5. As expected, water molecules have


13


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 14 of 32

also visited all four intersection regions but also spends considerably more time in the channels of the zeolite.

Figure 4. The positions of p-xylene (A), o-xylene (B), and m-xylene (C) molecule diffusing in H-ZSM-5 at 700 K. The green circles indicate the intersection regions between the straight channels and the zigzag channels.

As shown in Figure 5, a single p-xylene molecule can diffuse along both the straight and the zigzag channels during the simulation, while the o-xylene and the m-xylene molecules mainly diffuse along the original channels in which the molecule has been placed initially and does not have a chance to switch between different channels due to their slower diffusivity. The MSD plots for xylene isomers and benchmark molecules diffusing in H-ZSM-5 are shown in Figures S6, S7, and S8. Our results demonstrate that the ratio of the diffusion coefficients for the three xylene isomers at 700 K is p-xylene : o-xylene : m-xylene ≈ 83 : 3 : 1(note experimentally this ratio was determined to be 1000 : 10 : 1 at 373 K15). Specifically, the ratio for p-xylene : o-xylene ≈ 28 is in good consistent with Roque-Malherbe’s study, where they reported the ratio is ~25-70 at temperatures ranging from 375 K to 425 K.75 The calculated diffusion coefficients for the xylene isomers can aid in interpreting the product distribution observed during CFP


14

Page 15 of 32

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


qualitatively. For instance, Iliopoulou et al. conducted catalytic upgrading of lignocellulosic biomass pyrolysis vapors over H-ZSM-5 using a pilot scale circulating fluidized bed reactor and demonstrated the yield of p-xylene and o-xylene in the aromatic phase is 16% and 4%, respectively.76 Jae et al. studied CFP of pine wood over H-ZSM-5 using a bubbling fluidized bed reactor and reported the yield of the sum of p-xylene and m-xylene is 9.2% in the aromatic phase, while the yield of o-xylene is 2.0%.77 Our calculated diffusion coefficients in the order of p-xylene > o-xylene > m-xylene are also consistent with the trend of product intensity observed by Luo et al. who studied the methanol-to-gasoline process in a fixed-bed microreactor using GC-MS and HPLC.78 However, we note that diffusion is not the only factor to determine the product distribution, which can also be affected by thermodynamics and kinetics of chemical reactions inside the zeolites as well as on the zeolite surfaces.


15


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 16 of 32

Figure 5. The trajectories of xylene isomers are diffusing in the straight channels and the zigzag channels in H-ZSM-5 at 700 K. The blue surface model represents the trace of the xylene molecules diffusing along the channels.

The molecular critical diameters of toluene and xylene isomers are summarized in Table 2. The critical molecular diameter is 6.7 Å, 7.4 Å, and 7.3 Å for p-xylene, oxylene, and m-xylene, respectively. Thus, the different diffusivities of xylene isomers in HZSM-5 are mainly caused by the steric hindrance because of the small pore size and comparable molecular size. Especially, if the critical diameter of a molecule is


16

Page 17 of 32

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


comparable to the size of a zeolite pore, a small increase in the critical diameter could lead to a very large decrease in the measured diffusion coefficient. Table 2. Molecular critical diameters and length of the longest axis of toluene and xylene isomers. critical diameter (Å)

longest axis (Å)

DFT

MD

DFT

MD

benzene

6.7

NA

7.4

NA

toluene

6.7

6.7 ± 0.1

8.2

8.2 ± 0.2

p-xylene

6.7

6.7 ± 0.1

9.1

9.2 ± 0.2

o-xylene

7.4

7.4 ± 0.1

7.9

8.0 ± 0.2

m-xylene

7.3

7.3 ± 0.1

9.1

8.8 ± 0.2

As shown in Table 3, the self-diffusion coefficient of p-xylene and toluene measured via MD simulations is 2.5 and 1.1 × 10-10 m2 s-1 at 700 K, respectively, indicating p-xylene diffuses faster than toluene in H-ZSM-5. This observation does not obey the rule of the effect of the kinetic diameter on the diffusivity, since p-xylene has a larger molecular weight than toluene. However, our results are consistent with previous experimental studies in the literature. For instance, Han et al. used zero length column method to measure diffusivity of toluene and p-xylene and they reported the ratio of the diffusion coefficients of p-xylene versus toluene is also around 2:1 at various temperatures ranging from 373 K to 473 K.18 Since the critical diameter is the same for toluene and p-xylene (6.7 Å), the different diffusivities of toluene and p-xylene molecules in H-ZSM-5 must be due to


17


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 18 of 32

factors other than the size or molecular weight. It has been speculated that this observation is due to the different orientations of the toluene and p-xylene molecules when they diffuse along the channels in H-ZSM-5.74 The proposed hypothesis is that the dimension of the longest principal axis of p-xylene molecule (9.1 Å) is larger than that of toluene molecule (8.2 Å) and the pore size of H-ZSM-5 is comparatively small, the pxylene molecule will be aligned with its longest axis parallel to the channel. On the contrary, toluene molecules can rotate more easily relative to p-xylene at the intersection regions. Thus, toluene molecule has to align itself before it can diffuse into the adjacent intersection through a channel leading to a slower diffusion. Table 3. Comparison of calculated Ds with previous work. DS (10-12 m2 s-1) Substrate

This work (700 K) Cormaa (650 K)

Toluene

110 ± 30

p-xylene

250 ± 30

4076

o-xylene

9±4

554

m-xylene

3±1

138

a

(373,423,473K)b

0.20, 0.25, 0.35

0.40, 0.52, 0.68

Llopis et al.39. bHan et al. 18


18

Page 19 of 32

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


To test this hypothesis, we have measured the angle between the longest principal axis of toluene/p-xylene molecule and the direction of the straight channel in H-ZSM-5 as a function of time. As shown in Figure 6, the most populated angle for p-xylene molecules is around ~0 degree, indicating the longest axis of p-xylene is parallel to the direction of the straight channel during most of the simulation time. Figure 6 also illustrates two broad peaks, i.e., around ~0 and ~70 degrees, for toluene molecules diffusing in H-ZSM-5. The population of the angles for each individual molecule of toluene and p-xylene is shown in Figure S3. On the contrary, no preference was observed on the similar angle in water molecules, as shown in Figure S4, suggesting small molecules can orient completely freely in H-ZSM-5. The results clearly demonstrate that the p-xylene molecules primarily diffuse parallel to the straight channels, while the toluene molecules can rotate more frequently relative to the p-xylene molecule. Figure 7 illustrates a typical pathway when p-xylene and toluene molecule diffuses along the straight channel in H-ZSM5. Thus, our simulations lend direct and strong support to the hypothesis regarding the effect of molecule orientation on the diffusivity.


19


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 20 of 32

Figure 6. The population of the angles between the longest principal axis of toluene/pxylene molecule and the direction of the straight channel in H-ZSM-5 as a function of time.

Figure 7. The superposition of positions of a p-xylene (A) and toluene (B) molecule crossing a 10-member ring via a straight channel. Red indicates the first frame and blue is for the last frame. The total time period is 5 ps. As previously demonstrated in Figure 5, p-xylene molecule can diffuse along both the straight and the zigzag channels. To further clarify the orientations of the p-xylene molecule diffusing along the channels, we mapped the positions of p-xylene molecule with its orientation as a function of time, as illustrated in Figure 8. The data clearly


20

Page 21 of 32

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


demonstrate the angle is around 0° when p-xylene diffuses along the straight channels, while this angle is close to 90° when p-xylene diffuses along the zigzag channels. To switch from a straight channel to a zigzag channel, p-xylene molecule needs to orient itself at the intersection regions.

Figure 8. The population of the angles between the longest principal axis of p-xylene molecule and the direction of the straight channel when p-xylene diffuses along the straight (A) and zigzag (B) channels. (C) The positions of xylene molecule as a function of time (red indicates the position along the straight channel and blue shows the position along the zigzag channels).


21


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 22 of 32

We note m-xylene and o-xylene molecules diffuse in H-ZSM-5 with different orientations from p-xylene molecule, as demonstrated in Figure S5. The length of the longest principal axis of m-xylene (8.8 ± 0.2 Å) and o-xylene (8.0 ± 0.2 Å) as measured in MD simulations is shorter than that of p-xylene (9.2 ± 0.2 Å), resulting in relatively more unrestricted rotations of m-xylene and o-xylene molecules than p-xylene in the intersection regions. Though DFT optimized geometries illustrate the same length of the longest principal axis for p-xylene and m-xylene molecules, the rotation of the –CH3 groups can have different effects on this length. As shown in Figure 9, the rotation of – CH3 groups does not change the length of the longest principal axis of p-xylene, while it can decrease the length of m-xylene from 9.1 Å to 8.6 Å, making m-xylene rotate more freely than p-xylene. Additionally, the orientations of m-xylene and o-xylene molecules are also different from that of toluene in H-ZSM-5, possibly due to the difference of their critical diameters. The slower diffusivities of m-xylene and o-xylene molecules result in longer residence time in the intersection regions, therefore they may explore other orientations. Thus, our results suggest that the diffusion behavior of a molecule can be affected by both the critical diameter and the length of the longest principal axis.


22

Page 23 of 32

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 9. The longest axes of p-xylene (A) and m-xylene with different conformations (B and C). The rotation of –CH3 group affects the length of the longest axis in m-xylene, while it has no effect on p-xylene. Diffusion along different channels

Since the size of the straight

channels (5.3 × 5.6 Å) in H-ZSM-5 is slightly larger than that of the zigzag channels (5.1 × 5.5),48 the substrate molecule is expected to diffuse faster in the straight channels. Additionally, the steric hindrance along the zigzag channels is anticipated to be more severe than that along the straight channels, which will also make diffusion along the straight channels faster than that along the zigzag channels. To investigate our hypothesis, we calculated the components of the diffusion coefficients of p-xylene and toluene molecules along the straight channels and the zigzag channels in H-ZSM-5. Table 4 shows the components of the diffusion coefficients calculated during the 40 ns MD simulation. The results indicate that diffusions along the straight channels and the zigzag channels are dominant in the overall diffusivity. The results also show that the diffusion along the straight channel is ~2 times faster than that along the zigzag channel for both toluene and p-xylene molecules. Therefore, we speculate the diffusivities in H-ZSM-11, which is analogical to H-ZSM-5 but only composed of the straight channels,79 might be enhanced due to the absence of the zigzag channels, making H-ZSM-11 a potential candidate for further investigation to improve diffusivities. Table 4. Calculated self-diffusion coefficients (Ds) along the zigzag and the straight channel in H-ZSM-5 at 700 K. DS (10-10 m2 s-1)

Zigzag

Straight


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

Page 24 of 32

toluene

0.32 ± 0.07

0.62 ± 0.32

p-xylene

0.96 ± 0.06

1.70 ± 0.40

Diffusion activation energy

The

diffusivities

of

substrate

molecules

significantly depend on the various pyrolysis temperatures. Thus, predicting the diffusivities at different temperatures will benefit interpreting the product distributions under various experimental conditions. The activation process of diffusion in zeolite can be described by: 01

= - . / 23

(3)

where Ds denotes the diffusion coefficient, D0 stands for pre-exponential factor, and ED is the diffusion activation energy. The calculated activation energies can then be used to predict the diffusivities of these compounds at any desired temperature. The Arrhenius graphs of water, methanol, glycolaldehyde monomer, p-xylene, and toluene molecule diffusion in H-ZSM-5 are shown in Figure 10. The calculated activation energies based on Equation 3 are in good agreement with the previous studies, as shown in Table 5. Table 5. Diffusion activation energies for selected molecules. ED (kcal mol-1)

Literature

This work

water

3.2a

3.5

methanol

2.9a

3.4

(monomer)

NA

3.9

toluene

4.9b

4.0

Glycoaldehyde


24

Page 25 of 32

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


4.5b

p-xylene

a

3.8

Yang et al.80 bSong et al.74

Figure 10. The Arrhenius plot of water, methanol, glycolaldehyde, p-xylene, and toluene molecule diffusion in H-ZSM-5.

Conclusions In this work, we have conducted molecular dynamics simulations to investigate the diffusivities of important bio-fuel molecules in H-ZSM-5. Our results demonstrate very good agreement with available NMR measures on the diffusivities of water, ethanol, and toluene molecules, thus proving the Hill-Sauer force field used in this work is appropriate force field to simulate the diffusion in H-ZSM-5. We have characterized the diffusivities of xylene isomers in H-ZSM-5 to interpret the product distribution observed at CFP of biomass.


25


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 26 of 32

Due to the complex nature and speciation of pyrolysis vapors, it is imperative that a correlation be developed to relate molecular properties to diffusivity within zeolite pores. This work has shown that such a correlation needs to take into account both molecular size and weight. Additionally, this work also demonstrates the important effect of molecular orientation on the diffusivity at the configuration regime and both the critical diameter and the length of the longest principal axis should be considered simultaneously, however, the latter factor has usually obtained less attention in the previous literatures. It is also likely that polarizability (e.g. hydrogen bonding to Brønsted acid sites) will also play a reduced, but important role. The scope of this challenge will almost certainly be intractable by experimental methods due to the reactivity of species involved and subsequent rapid catalyst deactivation and likely pore blocking, while MD simulations are a convenient and powerful method for addressing this problem.

Supporting Information. All supplementary figures associated with this article. This material is available free of charge via the Internet http://pubs.acs.org. AUTHOR INFORMATION Corresponding Author *E-mail and Phone: [email protected], 1-303-384-7323 Author Contributions The manuscript was written by all authors. All authors have approved the final version of this manuscript. Acknowledgements


26

Page 27 of 32

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


This work was supported by the Computational Pyrolysis Consortium funded by the U.S. Department of Energy’s Bioenergy Technologies Office (DE-AC36-08GO28308). Computational resources were provided by the Computational Sciences Center at National Renewable Energy Laboratory. We thank professor David Sholl, Dr. Ambarish Kulkarni, and Dr. Rohan Awati for providing their codes to conduct the MD simulations. We thank Dr. Singfoong Cheah and Dr. Erica Gjersing for helpful discussions.

Supporting Information

Potential parameters for zeolite framework (Table S1); Diffusion

coefficients as a function of various molecular descriptors (Figure S1); The positions of water molecules diffusing in H-ZSM-5 (Figure S2); The orientation of toluene/xylene/water molecules diffusing in H-ZSM-5 (Figure S3-S5); The MSD plots for the molecules (water, methanol, ethanol, glycolaldehyde, toluene, and xylene isomers) diffusing in H-ZSM-5 (Figure S6-S8).

Reference (1) Cheah, S.; Starace, A. K.; Gjersing, E.; Bernier, S.; Deutch, S. Reactions of Mixture of Oxygenates Found in Pyrolysis Vapors: Deoxygenation of Hydroxyacetaldehyde and Guaiacol Catalyzed by HZSM-5. Top. Catl. 2016, 59, 109-123. (2) Ragauskas, A. J.; Williams, C. K.; Davison, B. H.; Britovsek, G.; Cairney, J.; Eckert, C. A.; Frederick, W. J.; Hallett, J. P.; Leak, D. J.; Liotta, C. L. The Path Forward for Biofuels and Biomaterials. Science 2006, 311, 484-489. (3) Mullen, C. A.; Boateng, A. A. Chemical Composition of Bio-Oils Produced by Fast Pyrolysis of Two Energy Crops. Energy Fuels 2008, 22, 2104-2109. (4) Taarning, E.; Osmundsen, C. M.; Yang, X.; Voss, B.; Andersen, S. I.; Christensen, C. H. Zeolite-Catalyzed Biomass Conversion to Fuels and Chemicals. Energy Environ. Sci. 2011, 4, 793-804. (5) Zhou, C.-H.; Xia, X.; Lin, C.-X.; Tong, D.-S.; Beltramini, J. Catalytic Conversion of Lignocellulosic Biomass to Fine Chemicals and Fuels. Chem. Soc. Rev. 2011, 40, 5588-5617. (6) Chmelik, C.; Kärger, J. In Situ Study on Molecular Diffusion Phenomena in Nanoporous Catalytic Solids. Chem. Soc. Rev. 2010, 39, 4864-4884. (7) Kärger, J.; Valiullin, R. Mass Transfer in Mesoporous Materials: The Benefit of Microscopic Diffusion Measurement. Chem. Soc. Rev. 2013, 42, 4172-4197. (8) Javaid, R.; Urata, K.; Furukawa, S.; Komatsu, T. Factors Affecting Coke Formation on H-ZSM-5 in Naphtha Cracking. Appl. Catal., A. 2015, 491, 100-105. (9) Konno, H.; Okamura, T.; Kawahara, T.; Nakasaka, Y.; Tago, T.; Masuda, T. Kinetics of n-Hexane Cracking over ZSM-5 Zeolites–Effect of Crystal Size on Effectiveness Factor and Catalyst Lifetime. Chem. Eng. J. 2012, 207, 490-496. (10) Al-Khattaf, S.; De Lasa, H. The Role of Diffusion in Alkyl-Benzenes Catalytic Cracking. Appl. Catal., A. 2002, 226, 139-153.


27


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 28 of 32

(11) Yuan, W.; Lin, Y.; Yang, W. Molecular Sieving MFI-Type Zeolite Membranes for Pervaporation Separation of Xylene Isomers. J. Am. Chem. Soc. 2004, 126, 4776-4777. (12) Liu, J.; Wei, J. Knudsen Diffusion in Channels and Networks. Chem. Eng. Sci. 2014, 111, 1-14. (13) Xiao, J.; Wei, J. Diffusion Mechanism of Hydrocarbons in Zeolites-I. Theory. Chem. Eng. Sci. 1992, 47, 1123-1141. (14) Choudhary, V. R.; Nayak, V. S.; Choudhary, T. V. Single-Component Sorption/Diffusion of Cyclic Compounds from Their Bulk Liquid Phase in H-ZSM-5 Zeolite. Ind. Eng. Chem. Res. 1997, 36, 1812-1818. (15) Mirth, G.; Cejka, J.; Lercher, J. Transport and Isomerization of Xylenes over HZSM-5 Zeolites. J. Catal. 1993, 139, 24-33. (16) Brandani, S. Challenges in Macroscopic Measurement of Diffusion in Zeolites. Diffusion Fund. 2007, 6, 1-10. (17) Mirth, G.; Lercher, J. A.; Anderson, M. W.; Klinowski, J. Adsorption Complexes of Methanol on Zeolite ZSM-5. J. Chem. Soc., Faraday Trans. 1990, 86, 3039-3044. (18) Han, M.; Yin, X.; Jin, Y.; Chen, S. Diffusion of Aromatic Hydrocarbon in ZSM5 Studied by the Improved Zero Length Column Method. Ind. Eng. Chem. Res. 1999, 38, 31723175. (19) Caro, J.; Noack, M.; Richter-Mendau, J.; Marlow, F.; Petersohn, D.; Griepentrog, M.; Kornatowski, J. Selective Sorption Uptake Kinetics of n-Hexane on ZSM 5-A New Method for Measuring Anisotropic Diffusivities. J. Phys. Chem. 1993, 97, 13685-13690. (20) Jobic, H. Diffusion of Linear and Branched Alkanes in Zsm-5. A Quasi-Elastic Neutron Scattering Study. J. Mol. Catal. A: Chem. 2000, 158, 135-142. (21) Voogd, P.; van Bekkum, H.; Shavit, D.; Kouwenhoven, H. W. Effect of Zeolite Structure and Morphology on Intracrystalline n-Hexane Diffusion in Pentasil Zeolites Studied by the Zero-Length Column Method. J. Chem. Soc., Faraday Trans. 1991, 87, 3575-3580. (22) Jobic, H.; Bée, M.; Pouget, S. Diffusion of Benzene in ZSM-5 Measured by the Neutron Spin-Echo Technique. J. Phys. Chem. B 2000, 104, 7130-7133. (23) Baertsch, C. D.; Funke, H. H.; Falconer, J. L.; Noble, R. D. Permeation of Aromatic Hydrocarbon Vapors through Silicalite-Zeolite Membranes. J. Phys. Chem. 1996, 100, 7676-7679. (24) Wu, P.; Debebe, A.; Ma, Y. H. Adsorption and Diffusion of C 6 and C 8 Hydrocarbons in Silicalite. Zeolites 1983, 3, 118-122. (25) Kärger, J.; Ruthven, D. M. Diffusion in Nanoporous Materials: Fundamental Principles, Insights and Challenges. New J. Chem. 2016, 40, 4027-4048. (26) Haag, W. O.; Lago, R. M.; Weisz, P. B. Transport and Reactivity of Hydrocarbon Molecules in a Shape-Selective Zeolite. Faraday Discuss. Chem. Soc. 1981, 72, 317-330. (27) Jobic, H.; Schmidt, W.; Krause, C. B.; Kärger, J. PFG NMR and QENS Diffusion Study of n-Alkane Homologues in MFI-Type Zeolites. Micropor. Mesopor. Mat. 2006, 90, 299-306. (28) Heink, W.; Kaerger, J.; Pfeifer, H.; Stallmach, F. Measurement of the Intracrystalline Self-Diffusion of Xenon in Zeolites by the NMR Pulsed Field Gradient Technique. J. Am. Chem. Soc. 1990, 112, 2175-2178. (29) Jobic, H.; Bée, M.; Caro, J.; Bülow, M.; Kärger, J. Molecular Self-Diffusion of Methane in Zeolite ZSM-5 by Quasi-Elastic Neutron Scattering and Nuclear Magnetic Resonance Pulsed Field Gradient Technique. J. Chem. Soc. Faraday Trans. 1 1989, 85, 4201-4209. (30) Takahara, S.; Sumiyama, N.; Kittaka, S.; Yamaguchi, T.; Bellissent-Funel, M.-C. Neutron Scattering Study on Dynamics of Water Molecules in MCM-41. 2. Determination of Translational Diffusion Coefficient. J. Phys. Chem. B 2005, 109, 11231-11239.


28

Page 29 of 32

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


(31) Ghysels, A.; Moors, S. L.; Hemelsoet, K.; De Wispelaere, K.; Waroquier, M.; Sastre, G.; Van Speybroeck, V. Shape-Selective Diffusion of Olefins in 8-Ring Solid Acid Microporous Zeolites. J. Phys. Chem. C 2015, 119, 23721-23734. (32) Beerdsen, E.; Smit, B. Diffusion in Confinement: Agreement between Experiments Better Than Expected. J. Phys. Chem. B 2006, 110, 14529-14530. (33) Smit, B.; Maesen, T. L. Molecular Simulations of Zeolites: Adsorption, Diffusion, and Shape Selectivity. Chem. Rev. 2008, 108, 4125-4184. (34) Combariza, A. F.; Gomez, D. A.; Sastre, G. Simulating the Properties of Small Pore Silica Zeolites Using Interatomic Potentials. Chem. Soc. Rev. 2013, 42, 114-127. (35) Jiang, J.; Babarao, R.; Hu, Z. Molecular Simulations for Energy, Environmental and Pharmaceutical Applications of Nanoporous Materials: From Zeolites, Metal-Organic Frameworks to Protfin Crystals. Chem. Soc. Rev. 2011, 40, 3599-3612. (36) Keil, F. J.; Krishna, R.; Coppens, M.-O. Modeling of Diffusion in Zeolites. Rev. Chem. Eng. 2000, 16, 71-197. (37) Corma, A. Sorption, Diffusion and Catalytic Properties of Zeolites Containing 10-and 12-Member Ring Pores in the Same Structure. Micropor. Mesopor. Mat. 1998, 21, 487495. (38) Llopis, F. J.; Sastre, G.; Corma, A. Xylene Isomerization and Aromatic Alkylation in Zeolites NU-87, SSZ-33, β, and ZSM-5: Molecular Dynamics and Catalytic Studies. J. Catal. 2004, 227, 227-241. (39) Llopis, F. J.; Sastre, G.; Corma, A. Isomerization and Disproportionation of mXylene in a Zeolite with 9-and 10-Membered Ring Pores: Molecular Dynamics and Catalytic Studies. J. Catal. 2006, 242, 195-206. (40) Song, A.; Ma, J.; Xu, D.; Li, R. Adsorption and Diffusion of Xylene Isomers on Mesoporous Beta Zeolite. Catalysts 2015, 5, 2098-2114. (41) Klein, H.; Fuess, H.; Schrimpf, G. Mobility of Aromatic Molecules in Zeolite NaY by Molecular Dynamics Simulation. J. Phys. Chem. 1996, 100, 11101-11112. (42) Schrimpf, G.; Tavitian, B.; Espinat, D. Computer Simulation of the Structure, Energetics, and Diffusion Properties of p-Xylene in Zeolite Na-Y. J. Phys. Chem. 1995, 99, 10932-10941. (43) Toda, J.; Corma, A.; Abudawoud, R. H.; Elanany, M. S.; Al-Zahrani, I. M.; Sastre, G. Influence of Force Fields on the Selective Diffusion of para-Xylene over ortho-Xylene in 10-Ring Zeolites. Mol. Simul. 2015, 41, 1438-1448. (44) Bueno-Perez, R.; Calero, S.; Dubbeldam, D.; Ania, C.; Parra, J.; Zaderenko, A.; Merkling, P. Zeolite Force Fields and Experimental Siliceous Frameworks in a Comparative Infrared Study. J. Phys. Chem. C 2012, 116, 25797-25805. (45) Tomás, R. r. A.; Bordado, J. o. C.; Gomes, J. F. p-Xylene Oxidation To Terephthalic Acid: A Literature Review Oriented toward Process Optimization and Development. Chem. Rev. 2013, 113, 7421-7469. (46) Cheng, Y. T.; Wang, Z.; Gilbert, C. J.; Fan, W.; Huber, G. W. Production of pXylene from Biomass by Catalytic Fast Pyrolysis Using ZSM-5 Catalysts with Reduced Pore Openings. Angew. Chem. Int. Ed. 2012, 51, 11097-11100. (47) Williams, C. L.; Chang, C.-C.; Do, P.; Nikbin, N.; Caratzoulas, S.; Vlachos, D. G.; Lobo, R. F.; Fan, W.; Dauenhauer, P. J. Cycloaddition of Biomass-Derived Furans for Catalytic Production of Renewable p-Xylene. ACS Catal. 2012, 2, 935-939. (48) Baerlocher, C.; Olson, D.; Meier, W.: Atlas of Zeolite Framework Types (formerly: Atlas of Zeolite Structure Types); Elsevier, 2001. (49) Milovanovic, J.; Rajic, N.; Romero, A. A.; Li, H.; Shih, K.; Tschentscher, R.; Luque, R. Insights into the Microwave-Assisted Mild Deconstruction of Lignin Feedstocks Using NiO-Containing ZSM-5 Zeolites. ACS Sustain. Chem. Eng. 2016.


29


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 30 of 32

(50) Liu, C.; Evans, T. J.; Cheng, L.; Nimlos, M. R.; Mukarakate, C.; Robichaud, D. J.; Assary, R. S.; Curtiss, L. A. Catalytic Upgrading of Biomass-Derived Compounds via C–C Coupling Reactions: Computational and Experimental Studies of Acetaldehyde and Furan Reactions in HZSM-5. J. Phys. Chem. C 2015, 119, 24025-24035. (51) Viswanadham, N.; Saxena, S. K. Enhanced Performance of Nano-Crystalline ZSM-5 in Acetone to Gasoline (ATG) Reaction. Fuel 2013, 105, 490-495. (52) Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput. Phys. 1995, 117, 1-19. (53) Nose, S. Constant-Temperature Molecular Dynamics. J. Phys.: Condens. Matter 1990, 2, SA115. (54) Woon, D. E.; Dunning Jr, T. H. Gaussian Basis Sets for Use in Correlated Molecular Calculations. III. The Atoms Aluminum through Argon. J. Phys. Chem. 1993, 98, 1358-1371. (55) Zhao, Y.; Truhlar, D. G. The M06 Suite of Density Functionals for Main Group Thermochemistry, Thermochemical Kinetics, Noncovalent Interactions, Excited States, and Transition Elements: Two New Functionals and Systematic Testing of Four M06-Class Functionals and 12 Other Functionals. Theor. Chem. Acc. 2008, 120, 215-241. (56) Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33-38. (57) Awati, R. V.; Ravikovitch, P. I.; Sholl, D. S. Efficient and Accurate Methods for Characterizing Effects of Framework Flexibility on Molecular Diffusion in Zeolites: CH4 Diffusion in Eight Member Ring Zeolites. J. Phys. Chem. C 2013, 117, 13462-13473. (58) Hill, J.-R.; Sauer, J. Molecular Mechanics Potential for Silica and Zeolite Catalysts Based on ab Initio Calculations. 2. Aluminosilicates. J. Phys. Chem. 1995, 99, 95369550. (59) Hill, J. R.; Sauer, J. Molecular Mechanics Potential for Silica and Zeolite Catalysts Based on ab Initio Calculations. 1. Dense And Microporous Silica. J. Phys. Chem. 1994, 98, 1238-1244. (60) Chen, B.; Potoff, J. J.; Siepmann, J. I. Monte Carlo Calculations for Alcohols and Their Mixtures with Alkanes. Transferable Potentials for Phase Equilibria. 5. United-Atom Description of Primary, Secondary, and Tertiary Alcohols. J. Phys. Chem. B 2001, 105, 30933104. (61) Kuhn, J.; Castillo-Sanchez, J. M.; Gascon, J.; Calero, S.; Dubbeldam, D.; Vlugt, T. J.; Kapteijn, F.; Gross, J. Adsorption and Diffusion of Water, Methanol, and Ethanol in AllSilica DD3R: Experiments and Simulation. J. Phys. Chem. C 2009, 113, 14290-14301. (62) Maerzke, K. A.; Schultz, N. E.; Ross, R. B.; Siepmann, J. I. TraPPE-UA Force Field for Acrylates and Monte Carlo Simulations for Their Mixtures with Alkanes and Alcohols. J. Phys. Chem. B 2009, 113, 6415-6425. (63) Wang, C.-H.; Bai, P.; Siepmann, J. I.; Clark, A. E. Deconstructing HydrogenBond Networks in Confined Nanoporous Materials: Implications for Alcohol–Water Separation. J. Phys. Chem. C 2014, 118, 19723-19732. (64) Jorgensen, W. L.; Maxwell, D. S.; Tirado-Rives, J. Development and Testing of the OPLS All-Atom Force Field on Conformational Energetics and Properties of Organic Liquids. J. Am. Chem. Soc. 1996, 118, 11225-11236. (65) Borah, B. J.; Jobic, H.; Yashonath, S. Levitation Effect in Zeolites: Quasielastic Neutron Scattering and Molecular Dynamics Study of Pentane Isomers in Zeolite NaY. J. Chem. Phys. 2010, 132, 144507. (66) Huang, S.; Finsy, V.; Persoons, J.; Telling, M. T.; Baron, G. V.; Denayer, J. F. Rotation Dynamics of 2-Methyl Butane and n-Pentane in MCM-22 Zeolite: A Molecular Dynamics Simulation Study. Phys. Chem. Chem. Phys. 2009, 11, 2869-2875.


30

Page 31 of 32

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


(67) Rajappa, C.; Krause, C.; Borah, B.; Adem, Z.; Galvosas, P.; Kärger, J.; Subramanian, Y. Diffusion of Pentane Isomers in Faujasite-Type Zeolites: NMR and Molecular Dynamics Study. Micropor. Mesopor. Mat. 2013, 171, 58-64. (68) Castillo, J.; Vlugt, T.; Calero, S. Molecular Simulation Study on the Separation of Xylene Isomers in MIL-47 Metal− Organic Frameworks. J. Phys. Chem. C 2009, 113, 2086920874. (69) Lucena, S.; Pereira, J.; Cavalcante Jr, C. L. Sensitivity to Guest–Host Force Fields in Adsorption Equilibrium of Cyclic Hydrocarbons in One-Dimensional Molecular Sieve. Mol. Simul. 2007, 33, 437-448. (70) Snurr, R. Q.; Bell, A. T.; Theodorou, D. N. Prediction of Adsorption of Aromatic Hydrocarbons in Silicalite from Grand Canonical Monte Carlo Simulations with Biased Insertions. J. Phys. Chem. 1993, 97, 13742-13752. (71) Yang, J. Z.; Chen, Y.; Zhu, A. M.; Liu, Q. L.; Wu, J. Y. Analyzing Diffusion Behaviors of Methanol/Water through MFI Membranes by Molecular Simulation. J. Membr. Sci. 2008, 318, 327-333. (72) Kua, J.; Galloway, M. M.; Millage, K. D.; Avila, J. E.; De Haan, D. O. Glycolaldehyde Monomer and Oligomer Equilibria in Aqueous Solution: Comparing Computational Chemistry and NMR Data. J. Phys. Chem. A 2013, 117, 2997-3008. (73) Kersten, S.; Garcia-Perez, M. Recent Developments in Fast Pyrolysis of LignoCellulosic Materials. Curr. Opin. Biotechnol. 2013, 24, 414-420. (74) Song, L.; Rees, L. V. Adsorption and Diffusion of Cyclic Hydrocarbon in MFIType Zeolites Studied by Gravimetric and Frequency-Response Techniques. Micropor. Mesopor. Mat. 2000, 35, 301-314. (75) Roque-Malherbe, R.; Wendelbo, R.; Mifsud, A.; Corma, A. Diffusion of Aromatic Hydrocarbons in H-ZSM-5, H-Beta, and H-MCM-22 Zeolites. J. Phys. Chem. 1995, 99, 14064-14071. (76) Iliopoulou, E.; Stefanidis, S.; Kalogiannis, K.; Psarras, A.; Delimitis, A.; Triantafyllidis, K.; Lappas, A. Pilot-Scale Validation of Co-ZSM-5 Catalyst Performance in the Catalytic Upgrading of Biomass Pyrolysis Vapours. Green Chem. 2014, 16, 662-674. (77) Jae, J.; Coolman, R.; Mountziaris, T.; Huber, G. W. Catalytic Fast Pyrolysis of Lignocellulosic Biomass in a Process Development Unit with Continual Catalyst Addition and Removal. Chem. Eng. Sci. 2014, 108, 33-46. (78) Luo, G.; McDonald, A. G. Conversion of Methanol and Glycerol into Gasoline via ZSM-5 Catalysis. Energy & Fuels 2013, 28, 600-606. (79) Ahn, J. H.; Kolvenbach, R.; Al-Khattaf, S. S.; Jentys, A.; Lercher, J. A. Methanol Usage in Toluene Methylation with Medium and Large Pore Zeolites. ACS Catal. 2013, 3, 817825. (80) Yang, J. Z.; Liu, Q. L.; Wang, H. T. Analyzing Adsorption and Diffusion Behaviors of Ethanol/Water through Silicalite Membranes by Molecular Simulation. J. Membr. Sci. 2007, 291, 1-9. Insert Table of Contents Graphic. (maximum size: 3.25 x 1.75 inch)


31


Page 32 of 32

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 Paragon Plus Environment

32

Diffusion of Biomass Pyrolysis Products in H-ZSM ... - ACS Publications

Recommend Documents