Computational Chemistry Analysis of Hydrodesulfurization Reactions Catalyzed by Molybdenum Disulfide Nanoparticles

Page 1 of 37 The Journal of Physical Chemistry
1
2
3
4 Computational Chemistry Analysis of Hydrodesulfurization Reactions Catalyzed
5
6 by Molybdenum Disulfide Nanoparticles
7
8
9
10 Narendra Kumar,1 and Jorge M. Seminario1,2,3,*
11
12
13 1
Department of Chemical Engineering
14
2
15 Department of Electrical and Computer Engineering
16 3
17 Department of Materials Science and Engineering
18
19
Texas A&M University
20 College Station, TX 77843, USA
21
22
23
24 ABSTRACT
25
26 Molybdenum disulfide, MoS2, is a very versatile material used in several applications of science
27 and engineering. In particular, when supported on alumina is widely used catalyst for
28
29 hydrotreating processes in petroleum refineries. Stringent environmental norms and uncertainty
30
31 in import of low sulfur crude oil put pressure on refiners worldwide to develop more efficient
32
33 catalysts to meet the demand of clean fuel. The hydrodesulfurization of thiophene and
34
35
dibenzothiophene on both unpromoted and nickel-promoted MoS2 catalyst is analyzed using a
36 multiscale approach including classical molecular dynamics (MD) and density functional theory
37
38 (DFT). MD simulations are performed on a hydrocarbon mixture containing sulfur compounds
39
40 in order to determine relative positions of thiophene and dibenzothiophene molecules with
41
42 respect to the MoS2 catalytic surface previous to possible reactions that are then studied with
43 DFT. The sample box contains a total of 57,515 atoms is able to represent a realistic size of the
44
45 nanocatalyst process. Under typical hydrodesulfurization conditions, vacancies are needed for
46
47 adsorption of sulfur compounds and therefore activation of hydrogen becomes an important step.
48
49 In addition, complexes of MoS2 with graphene and boron nitride, BN, surfaces are also analyzed.
50
MoS2-graphene complexes show improvement in efficiency of adsorption of thiophene on the
51
52 catalyst edge. Both hydrogenation and direct desulfurization pathways are presented on pristine
53
54 MoS2 clusters as well as under typical hydrodesulfurization conditions.
55
56
57
58 *Corresponding author: seminario@tamu.edu, Telephone: (979) 845-3301
59
60 1
ACS Paragon Plus Environment
The Journal of Physical Chemistry Page 2 of 37
1
2
3
Keywords: hydrodesulphurization; heavy oils; graphene; DFT; Molecular Dynamics;
4
5 nanotechnology
6
7
8
9 1. INTRODUCTION
10
11
Molybdenum disulfide, MoS2, is a versatile material belonging to layered transition-metal
12 dichalcogenide family and composed of vertically stacked, weakly interacting layers held
13
14 together by van der Waals forces. The wide range of electronic, optical, mechanical, chemical
15
16 and thermal properties of MoS2 has been the focus of many review articles in the past.1-3 It has a
17
18 sizable bandgap that changes from indirect to direct bandgap on exfoliation of single layer which
19 causes changes in photoconductivity, absorption spectra and photoluminescence.2 As a result,
20
21 MoS2 has great potential applications in electronic devices,2,4 photodetectors, and optoelectronic
22
23 devices5. The edge modification of MoS2 clusters is responsible for the enhanced catalytic
24
25 activity in HDS catalysis6 and electrocatalysis for hydrogen evolution7-11. Its high surface-to-
26
27
volume ratio can be exploited in molecular sensing of NO12, NO213 as well as of biomolecules14.
28 The application of MoS2 for energy storage application1,15 has also generated significant interest
29
30 among researchers. Due to high reversible capacity, easier intercalation and extraction of lithium
31
32 ions and low average voltage, MoS2 has potential applications as both anode11,16 and cathode17
33
34 materials for secondary lithium ion batteries.1
35
36
37 In particular, with the ever growing demand and consumption of transportation fuels and stricter
38
39 environmental norms, refiners across the globe are looking for ways to produce cleaner fuels.
40
41 The hydrodesulphurization (HDS) is one the most important processes in a refinery to remove
42
43
sulfur containing compounds from petroleum fractions (Scheme 1). Stricter legislations now
44 require removing even the most refractory sulfur compounds such as alkyl dibenzothiophenes.
45
46 Although hydrodesulfurization processes have been investigated for several decades,
47
48 considerable research is underway to understand these catalytic mechanisms and to develop new
49
50 chemistries that may be able to improve efficiency and reduce costs needed for more efficient
51 processes and more active catalysts. Common catalysts used for removal of sulfur from heavy
52
53 oil in refineries are mainly based on molybdenum disulfide (MoS2), usually promoted with
54
55 cobalt or nickel, which modify the electronic properties of the MoS2 matrix, thus enhancing its
56
57 desulfurization catalytic activity.
58
59
60 2
1
2
3
4
5
6 S
7
8
9 (a) (b) (c) (d) (e)
10
11 Scheme 1. Common sulfur-containing compounds present in petroleum fractions: (a) mercaptan, (b)
12 thiophene, (c) benzothiophene, (d) dibezothiophene, and (e) alkyl-substituted dibenzothiophene
13
14
15
16 Catalyst design from first principles is a challenge that is becoming reality with current advances
17
18
in computational software and hardware. Over the last decade, significant effort has been made
19 to elucidate the reaction mechanisms and to identify reaction intermediates using both periodic
20
21 and finite cluster approaches via density functional theory. The advantage of using periodic
22
23 calculations using plane wave basis sets is the lower computational cost to study large MoSx
24
25 clusters; however the drawback is the high coverage and defects are difficult to model.18
26 However, industrial catalysts are highly dispersed and have predominant sizes within 10-30 ,
27
28 i.e., one to four Mo atoms on a corresponding edge assuming regular hexagonal clusters as most
29
30 stable structure.19 Therefore, the use of MoS2 clusters with sizes close to real MoS2 particles
31
32 instead of periodic structures should be more realistic.20 The cluster approach provides efficient
33
34
analysis tools for single molecule adsorption.18
35
36
37 Theoretical study of MoS2 surface showed that depending on the reaction conditions, the
38
39 catalytic edges are covered by sulfur and hydrogen and there are not many actives sites or
40
41 coordinately unsaturated sites available for HDS reaction.21-23 Hydrogen undergoes dissociative
42 adsorption forming MoH and SH pairs as most favorable case. Under working conditions, the
43
44 most stable surface obtained was Mo-edge with 50% sulfur coverage with S atoms in bridging
45
46 positions.22,24-26 The promoter atoms Ni and Co were located predominantly at the edges with Ni
47
48 substituting Mo atoms at the Mo-edge while Co atoms substituting Mo atoms at the S-
49
edge.21,24,26-27 The Ni atoms at the Mo-edge resulted in square planar structures and were not
50
51 covered by S atoms under HDS conditions while Co and Mo atoms at the S-edge were 50%
52
53 sulfide with sulfur atoms in bridge positions. The bridging sulfur atoms on Mo atoms were
54
55 located in zig-zag configuration while those on Co atoms occupied regular bridge positions
56
57
between metal atoms.28 Hydrogen can also react with S creating vacancies and it was shown that
58
59
60 3
1
2
3
the barrier to create a vacancy on Mo-edge was 0.67 eV, i.e., it was lower than the barrier for S-
4
5 edge, which was 1.34 eV. The promoters Co and Ni decrease the barrier to create vacancies on
6
7 Mo- and S-edges.22 The H2 to H2S molar ratio decreases at the end of a catalytic cycle or at the
8
9 bottom of an industrial reactor and the most stable surface is the one with three sulfur atoms on
10
11
Mo-edge and S-edge containing six sulfur atoms. The adsorption mechanism of thiophene and
12 DBT on MoS2 catalyst have been studied in detailed.29 Two common modes of adsorption are
13
14 the 1 and 5. The HDS of DBT can follow either direct desulfurization (DDS) forming biphenyl
15
16 or a hydrogenation (HYD) pathway forming cyclohexyl benzene as final product.29 There is also
17
18 considerable research done on hydrogen activation on Mo-S, Co-Mo-S and Ni-Mo-S under
19 working conditions. The adsorption of hydrogen on Mo-S was exothermic with high activation
20
21 barrier while on Co-Mo-S adsorption was exothermic. Ni-S sites were not observed due to their
22
23 instability. Thus, substitution of promoters helps in creating stable coordinately unsaturated
24
25 sites.30 Molecular hydrogen undergoes heterolytic dissociation on Mo-edge to form one Mo-H
26
27
and S-H pair with Ea = 0.50 eV and two SH on S-edge with an activation barrier of 1.25 eV.23
28 Cobalt was more stable on S-edge while Ni can be present on both Mo-edge and S-edge. A
29
30 systematic evaluation of the number of S atoms and then of the number of vacancies on the
31
32 edges for different promoters has not been done yet. Adding cobalt increases the amount of
33
34 hydrogen present on a catalytic surface.18
35
36
37 Sun et al.26 studied MoS2 modified with Ni and Co partially or fully under sulfiding conditions.
38
39 Ni was mostly present on Mo-edge while Co was preferred on the S-edge.26 For unpromoted
40
41 MoS2 at molar ratio of H2S to H2 >> 1, fully sulfided Mo-edge and S-edge was obtained. For Ni-
42
43
Mo-S catalyst, sulfur was only present on Mo atoms and NiS was unstable. On the other hand,
44 for Co-Mo-S, CoS bonds formed if molar ratio of H2S to H2 > 0.05 and the most stable
45
46 structure is bridge S between Mo and Co.26 Regarding the dissociation of hydrogen on the
47
48 Ni(Co)Mo catalyst, it was found that heterolytic dissociation on unpromoted MoS2 leads to the
49
50 formation of MoH and SH pairs and Ni-promoted in Mo-edge requires a slightly lower
51 activation energy of 0.04 eV while Co-promoted in an S-edge requires an activation energy of
52
53 0.6 eV; however, fully promoted Ni has no sulfur on Mo-edge while fully promoted Co has
54
55 bridge sulfur on the S-edge.31-32 On an unpromoted Mo-edge containing bridge S, homolytic
56
57 dissociation to form two SH groups is preferred over the Mo-H and S-H pairs and over two
58
59
60 4
1
2
3
MoH pairs; however, on unpromoted S-edge, hemolytic dissociation becomes endothermic.
4
5 For Ni-promoted Mo-edge, the molecular hydrogen dissociation is the rate determining step;
6
7 however, once dissociated, NiH and SH pairs are very mobile. For Co-promoted on S-edge,
8
9 dissociation is highly exothermic forming two SH groups. For Ni-Mo-S, the ability to generate
10
11
H-species with higher reactivity on the surface, favors the HYD pathway. For Co-Mo-S, cobalt
12 reduces the dissociation energy of hydrogen on the S-edge but dissociated H-species are less
13
14 mobile and therefore favors HDS over HYD.31-32
15
16
17
18 Adsorption of thiophene and DBT on catalytically active MoS2 has been studied in great detail.
19 Thiophene was shown to adsorb in the 5 configuration with thiophenic ring parallel to the Mo-
20
21 edge. This configuration resulted in the highest adsorption energy and helped in the CS bond
22
23 scission.33 Orita et al.34 studied adsorption of thiophene on MoS2 cluster for several adsorption
24
25 modes using a finite cluster approach, showing that simplification to a Mo16S32 cluster is
26
27
reasonable for examining the adsorption of thiophene due to its correct stoichiometry and
28 electroneutrality without any saturating hydrogen atoms. Thiophene remains flat in the 1 mode
29
30 but becomes bent in other configuration. The adsorption of thiophene required coordinately
31
32 unsaturated sites on Mo-edge.34 Faye et al.35 used a Mo12S24 cluster as a model for active phase
33
34 for adsorption of thiophene.35
35
36
37 Yang et al.36 used a Mo10S18 cluster to study adsorption of DBT and alkyl substituted DBT as
38
39 well as their hydrogenated derivatives and showed that flat adsorption is favored over
40
41 perpendicular adsorption. The adsorption energy increased for a flat adsorption on
42
43
hydrogenation of phenyl rings while it decreased for perpendicular adsorption due to fewer
44 available sites. Their results showed that DBT and alkyl substituted DBT show similar
45
46 interaction in flat mode while steric hindrance plays an important role in perpendicular
47
48 adsorption.36 Wen et al.20 studied structure and stability of several MoSx clusters and showed
49
50 that Mo16Sx can model S coverage on Mo-edge effectively and the addition of S to Mo-edge was
51 always exothermic. They obtained two stable structures, one with 33% S-coverage and one with
52
53 50% coverage on Mo-edge with 100% S on the S-edge. They also obtained one stable structure
54
55 with 67% S on the S-edge with no S on the Mo-edge.20 Borges et al.37 used distributed multipole
56
57
58
59
60 5
1
2
3
analysis to study the electronic structure and microwave effects on adsorption of thiophene over
4
5 a Mo16S32 cluster.37
6
7
8
9 The HYD and DDS reaction pathways for HDS of thiophene over MoS2 at both Mo-edge and S-
10
11
edge under typical reaction conditions were reported using density functional theory and finding
12 that under HDS conditions, the most stable edge configurations correspond to the Mo-edge with
13
14 50% S coverage and 50% H coverage and a S-edge with 100% S coverage and 100% H
15
16 coverage.38-39 Their work showed that HYD can proceed at Mo-edge even in the absence of
17
18 coordinately unsaturated Mo sites. The HYD pathways follow hydrogenation of thiophene to
19 form 2-hydrothiophene followed by hydrogenation to form 2,5-dihydrothiophene and subsequent
20
21 SC bond scission while DDS involves the first initial hydrogenation followed directly by SC
22
23 scission.38 Thus, it was predicted the interplay of Mo-edge and S-edge in catalyzing HDS
24
25 reactions where adsorption and hydrogenation occur at Mo-edge and SC scission occurs at S-
26
27
edge. However, vacancies were required at the S-edge for reactions to occur.38 DBT and alkyl
28 substituted DBT also underwent through two parallel reaction pathways: DDS yielding biphenyl-
29
30 type compound, or HYD yielding tertrahydro-DBT and cyclohexyl benzene-type products.40 For
31
32 DDS pathway, hydrogenation of one double bond in the aromatic ring next to sulfur atom
33
34 occurred followed by SC bond scission by elimination process while for HYD successive
35 hydrogenation of one of the aromatic ring occurred followed by SC bond scission.40 The role
36
37 of Ni and Co promoters on HDS of dibenzothiophenes in the presence of H2S and amines was
38
39 studied experimentally and the results showed that Ni-promoted MoS2 catalyst was less sensitive
40
41 to amine than unpromoted as well as Co-promoted MoS2 catalyst.28 The rate of DDS pathway
42
43
for DBT was dependent on the nature of the catalytic surfaces and increased as Mo << CoMo <<
44 NiMo. The higher DDS rate in CoMo was a result of the higher number of S-vacancies available
45
46 on the S-edge while in Ni-promoted catalyst, the rate of DDS increased significantly due to
47
48 presence of one free coordination site which can easily adsorb DBT. On the other hand, the
49
50 HYD pathway of DBT can occur on both S-covered and uncovered Mo, CoMo and NiMo edges,
51 but the reaction was strongly hindered in the presence of amines.28
52
53
54
55 In this work we focus on the removal of thiophene and dibenzothiophene over both, unpromoted
56
57 and Ni-promoted MoS2 catalyst using a finite cluster approach. Efforts have been made to
58
59
60 6
1
2
3
analyze HDS reactions over both pristine MoS2 as well as under industrial conditions. Since
4
5 HDS of thiophene has been widely investigated, we use it as control calculations and we focus
6
7 on the HDS of DBT.
8
9
10
11 2. METHODOLOGY
12 The molecular dynamics simulations of sulfur-containing hydrocarbon oil over MoS2 catalyst are
13
14 performed using the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS)
15
16 program.41 This program has been used successfully in several applications related to
17
18 nanotechnology42-47 The hydrocarbon sample is a mixture of 217 n-C15, 369 n-
19 nonylcyclohexane, 130 n-nonylbenzene, 152 naphthalene and 217 thiophene or
20
21 dibenzothiophene molecules and it corresponds to a specific gravity of ~0.68-0.76 (diesel
22
23 fraction), sandwiched between two parallel MoS2 clusters each of 111.8914.8595.53 3. The
24
25 simulation box also contains H2 molecules distributed randomly in the oil mixture. The initial
26
27
atomic coordinate file is generated with the Packing Optimization for Molecular Dynamics
28 Simulations (Packmol) program.48 Periodic boundary conditions are applied in all three
29
30 directions and the system is first minimized at 0 K and then heated to 600 K under the
31
32 microcanonical (NVE) ensemble. A cut-off of 10 is used for both van der Waals and
33
34 Coulombic interactions. All parameters for bonded and non-bonded interactions are taken from
35 the universal force field, UFF.49-50 The charge equilibration method is used to determine charges
36
37 on each atom based on the electronegativity equalization for each atom as described by Rappe
38
39 and Goddard.51 Although the use of lump charges to represent the electron density is not quite
40
41 perfect yet as there are some issues hardly difficult to describe with atomic lump charges, see for
42
43
instance Murray et al.52 and references there in; therefore use of ab initio methods to determine
44 the actual arrangements and interactions taking place at interfaces with the catalysts is strongly
45
46 needed. The ionization potentials, electron affinities and atomic radii required for the charge
47
48 equilibration procedure are obtained from the values given by Rappe et al.51 Atom velocities are
49
50 rescaled at every time-step with the Berendsen thermostat53 within the microcanonical ensemble
51 to control the temperature. This is followed by equilibration at 600 K for 50 ps with a time-step
52
53 of 0.05 fs (0.025 fs in case of thiophene) under the canonical ensemble (NVT). Such a short
54
55 time step is required for the convergence of the charge equilibration procedure. Visualization of
56
57 trajectories is performed using the graphics software program Visualization Molecular Dynamics
58
59
60 7
1
2
3
or VMD.54 A number density plot of S atoms from thiophene and DBT as well as of H atoms
4
5 from H2 molecules are then generated using the Density Profile Tool55 available for the VMD
6
7 program.
8
9
10
11
All quantum chemistry calculations are performed using the GAUSSIAN-09 Program.56 Density
12 functional theory57-58 with the hybrid functional B3PW9159-64 is used for all calculations; such a
13
14 functional has been tested in several type of applications.46,65-72 The LANL2DZ basis-set with
15
16 effective core potentials73 is applied for the transition elements Mo and Ni while the 6-31G(d,p)
17
18 basis-set74-75 is used for the remaining atoms. Geometry optimization is performed using the
19 Berny method.76-77 A self-consistent convergence threshold of 10-6 is used for the density matrix
20
21 and 10-8 for the root-mean-square and maximum density matrix error. A nanocluster for the
22
23 MoS2 catalyst, Mo16S32, is employed as suggested by other researchers34. The binding energies
24
25 for adsorption of hydrogen, thiophene, and dibenzothiophene are then calculated from the total
26
27
energies. Calculations are performed for both unpromoted and Ni-promoted MoS catalysts.
28
29
30 3. RESULTS AND DISCUSSION
31
32 Figure 1 provides the initial snapshot of a simulation box containing a hydrocarbon mixture
33
34 containing thiophene and gaseous hydrogen between two layers of MoS2 catalyst layers. The
35 catalyst edges exposing Mo-edge without any sulfur and S-edges without any defects are tested
36
37 in this simulation. The objective of these MD simulations is to identify the possible structures
38
39 and orientations of thiophene molecules with respect to the catalyst edges as well as their
40
41 locations before the HDS reactions progress.
42
43
44 Following equilibration, we observe that most of the thiophene molecules diffuse from the bulk
45
46 hydrocarbon oil to the catalyst edges. The density plot of thiophene molecules is shown in Figure
47
48 4(a) after 50 ps of equilibration at 600 K. The density plot shows two peaks near 2 and 5
49
50 from the catalyst edges indicating reactive sites for thiophene molecules. The thiophene
51 molecules show multiple orientations close to the catalyst edge as observed in Figure 2. The
52
53 temperature and total energy plots during the equilibration are shown in Figure 3. The density
54
55 plot of gaseous hydrogen molecules is shown in Figure 4(b). The density maxima are located at 1
56
57 from the edges. Since we have not considered dissociative adsorption of hydrogen in the
58
59
60 8
1
2
3
molecular dynamics, hydrogen molecules after the first layer are repelled away from the edges as
4
5 seen from a wider peak near the middle of the simulation box.
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
Figure 1. Unit cell simulation box for periodic in all three dimensions MD simulations. The unit cell
35
36 contains 217 n-C15, 369 n-nonylhexane, 130 n-nonylbenzene, 152 naphthalene and 217 DBT, with 108 H2
37
38 molecules sandwiched between two parallel MoS2 slabs; each MoS2 slab contains 3,456 Mo and 6912 S
39 atoms, making a total of 57,515 atoms for the full unit cell.
40
41
42
43 The molecular dynamics of DBT containing hydrocarbon mixture is run along similar procedure
44
45
as done for thiophene containing oil. Following equilibration at 600 K, we observe that density of
46 the DBT molecules is largest near the edges of MoS2 clusters with two peaks at 2 and 5
47
48 respectively (Figure 5). By visual inspection of the orientations of dibenzothiophenes in Figure 5,
49
50 we observe that the DBT molecules closest to the edges are parallel to the surface. However, the
51
52 peaks are narrower when compared to thiophene as seen in Figure 6. However, as the reaction
53 progresses, the catalytic surfaces will be covered by sulfur and hydrogen atoms making parallel
54
55 adsorption less likely.
56
57
58
59
60 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 Figure 2. Snapshot of equilibrated system showing thiophene and H2 between the two layers of MoS2
31
32 catalyst using the universal force field, UFF, with charge equilibration. For clarity, HC oil is not shown.
33
34
35 620
36
37 615
38
39 610
40
605
41
T (K)
42 600
43
44 595
45
46 590
47 585
48
49 580
50 0 10 20 30 40 50
51
52 t (ps)
53
54
55 (a)
56
57
58
59
60 10
1
2
3
4 -130
5
6
7 -135
E (Mcal/mol)
8
9
10
-140
11
12
13
14 -145
15
16
17 -150
18 0 10 20 30 40 50
19
20 t (ps)
21
22
23 (b)
24 Figure 3. (a) Temperature (T) and (b) total energy (E) versus time (t) plots of the hydrocarbon mixture
25
26 between the two MoS2 layers as decribed in Figure 1.
27
28
29
30 0.0025 Thiophene_S 0.06
31
32 MoS2_Mo
33 0.002
0.045
34
35
0.0015
nMo/3
36
nS/3
37 0.03
38
0.001
39
40
41 0.015
0.0005
42
43
44 0 0
45 0 20 40 60 80
46 y ()
47
48 (a)
49
50
51
52
53
54
55
56
57
58
59
60 11
1
2
3
4 0.002 0.08
5
H2_H
6 MoS2_S
7 0.0016
8
0.06
9
10 0.0012
nH/3
nS/3
11 0.04
12
13 0.0008
14
15 0.02
16
0.0004
17
18 0 0
19
20 0 20 40 60 80
21 y ()
22
23 (b)
24
25 Figure 4. Density n (atoms/3) of (a) S from thiophene and Mo from MoS2, and (b) H from H2 and S
26 from MoS2 along the y axis.
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 12
1
2
3 Figure 5. Snapshot of equilibrated system showing DBT and H2 between two layers of MoS2 catalyst
4
5 using the universal force field, UFF, with charge equilibration. HC oil is not shown for the sake of
6
7 clarity.
8
9
nS/3
0.0030 0.045
nMo/3
10
11
DB
12 0.0025 Mo
13
14
0.0020 0.03
15 0.0015
16
17 0.0010 0.015
18
19 0.0005
20
21
0.0000 0
0 20 40 60 y () 80
22
23
24 (a)
25
0.0020 H2_H 0.08
nH/3
nS/3
26
27
28 0.0016
0.06
29
30 0.0012
31 0.04
32 0.0008
33
34 0.02
35 0.0004
36
37 0.0000 0
38 0 20 40 60 y () 80
39
40
(b)
41
42 Figure 6. Density n (atoms/3) of S from DBT and Mo from MoS2 (a) and H from H2 and S from MoS2
43
44 (b) along the y axis ().
45
46
47
48
In order to study specific reaction sites of thiophene with MoS2 and some of their doped
49 variations, we calculate several of their cluster conformations (Figure 7). Most of them are
50
51 based on variations of basic cluster of Mo16S32 to which we have doped with Ni atoms and
52
53 approached or attached to H2 or thiophene molecules. As reported in the literature, Ni and Co
54
55 act as promoters and predominantly present on the edge sites.78 We obtain the optimized
56 geometries of Ni-promoted MoS2 clusters for both partially promoted and fully promoted cases.
57
58
59
60 13
1
2
3
Pristine MoS2 yields the strongest binding to thiophene when interacting with two Mo in parallel
4
5 adsorption mode (Table 1). This bonding strongly reduces ~60% when the interaction is to only
6
7 one Mo in perpendicular adsorption mode. Adding promoters further weakens these interactions
8
9 as the calculated binding energies decreases further as shown in Table 1. However, it has
10
11
already been shown by other researchers78 that pristine forms of MoS2 clusters are less likely to
12 be present under reaction conditions. Therefore we also obtain the optimized structures of MoS2
13
14 clusters with sulfur atoms present on the Mo-edge to study the adsorption of thiophene. Our
15
16 calculations show that S-vacancies are necessary on Mo-edge to create coordinately unsaturated
17
18 sites for adsorption of thiophene as can be seen from close to zero value of binding energy.
19
Since the reactions occur in an atmosphere of H2 and H2S, these vacancies will be continuously
20
21 created and destroyed. The most stable Mo-edge under reaction conditions was shown to contain
22
23 three S atoms in bridging positions.22,24-26 In order to create a single S-vacancy, we remove one
24
25 S atom from the Mo-edge and re-optimize the structure for both unpromoted and Ni-promoted
26
27
clusters. It is also interesting to note that when promoted with Ni, S atoms are only present on
28 the Mo atoms which is understandable since Ni-S bonds are less stable as reported earlier.30 The
29
30 binding energies of adsorption of thiophene on sulfur-containing Mo-edges are much smaller
31
32 than those of pristine Mo-edges.
33
34
35
36
37
38
39
40
41
42
43
44
T Mo16S32 NiMo15S32
45
46
47
48
49
50
51
52
53
54
55
56
57 Ni3Mo13S32 Mo16S35 Mo16S34
58
59
60 14
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 Ni1Mo15S34 Ni1Mo15S33 T-Mo16S32 on eMo
17
18
(-2.77 eV)
19
20
21
22
23
24
25
26
27
28
29
30
31
32
T-Mo16S32 on eMo T-Ni1Mo15S32 on pNi eMo T-Ni3Mo13S32 on pNi eMo
33
34 (-1.07 eV) (-0.97 eV) (-0.69 eV)
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50 T--- Mo16S35 on eMo T--- Ni1Mo15S34 on pNi eMo T-Mo16S34 on eMo
51
52 (-0.01 eV) (0.0 eV) (-0.57 eV)
53
54
55
56
57
58
59
60 15
1
2
3
4
5
6
7
8
9
10
11
12
13
14 T-Ni1Mo15S33 on pNi eMo
15
16 (-1.96 eV)
17
Figure. 7. Adsorption of thiophene (T) on unpromoted and Ni-promoted Mo16S32 clusters. Binding
18
19 energies with thiophene inside parentheses. Mo(cyan), C(grey), S(yellow), Ni(blue), and H(white).
20
21
22
23
24 Table 1. Total (E) and Binding Energies (D) of Thiophene and MoS2 Clusters at the
25
26 B3PW91/LANL2DZ/6-31G(d,p) Level of Theory
27 Complex E (Hartree) D (eV)
28
29 Thiophene (T) -552.89741
30
31
Mo16S32 -13823.38206
32 NiMo15S32 -13925.15425
33
34 Ni3Mo13S32 -14128.68166
35
Mo16S35 -15018.01421
36
37 Mo16S34 -14619.85984
38
39 Ni1Mo15S34 -14721.59832
40 Ni1Mo15S33 -14323.37740
41
42 T-Mo16S32 on eMo -14376.38136 -2.77
43
44 T-Mo16S32 on eMo -14376.31897 -1.07
45 T-Ni1Mo15S32 on pNi eMo -14478.08712 -0.97
46
47 T-Ni3Mo13S32 on pNi eMo -14681.60426 -0.69
48
T--- Mo16S35 on eMo -15570.91201 -0.01
49
50 T--- Ni1Mo15S34 on pNi eMo -15274.49576 0.0
51
52 T-Mo16S34 on eMo -15172.77806 -0.57
53 T-Ni1Mo15S33 on pNi eMo -14876.34697 -1.96
54
55
56
57
58
59
60 16
1
2
3
Since the HDS reactions are carried out in the presence of hydrogen, it is equally important to
4
5 study the activation of hydrogen on MoS2 clusters and the role of promoters on the nature of
6
7 catalytic surface present under reaction conditions. Gaseous hydrogen can get adsorbed on Mo-
8
9 edge activating the HH bond as indicated by increases in bond length from 0.73 to 0.86 .
10
11
Hydrogen atoms once dissociated can adsorb on Mo-edge and S-edge as well as the corner sites as
12 shown in Figure 8. Our calculations show that hydrogen undergoes dissociative adsorption on
13
14 both Mo-edge and S-edge forming a variety of surfaces. The adsorption is most exothermic in
15
16 case of Mo-edge and least on S-edge as shown in Table 2. When Mo-edge is partially promoted
17
18 by nickel atoms, the exothermicity of the reaction decreases and becomes endothermic when the
19
Mo-edge is fully promoted by nickel atoms as there are no Mo atoms available for MoH
20
21 formation. Hydrogen plays a key role in creation of S-vacancies in addition to directly participate
22
23 in HDS reactions. Similarly, H2S can also be adsorbed on Mo-edges when sulfur vacancies are
24
25 present. The reaction changes from exothermic to endothermic when nickel atoms are present as
26
27
promoters.
28
29
30
31
32
33
34
35
36
37
38
39
40
41 H2-Mo16S32 on eMo H2---Mo16S32 on c H2---Mo16S32 on eS
42
43 (-0.69 eV) (0.00 eV) (0.00 eV)
44
45
46
47
48
49
50
51
52
53
54
55
56
57 2H-Mo16S32 on eMo 2H- Mo16S32 on c 2H-Mo16S32 on eS
58
59
60 17
1
2
3 (-1.06 eV) (-0.71 eV) (-0.44 eV)
4
5
6
7
8
9
10
11
12
13
14
15
16 2H-Ni1Mo15S32 on pNi eMo 2H-Ni3Mo13S32 on pNi eMo H2S- Mo16S34 on eMo
17 (-0.61 eV) (0.32 eV) (-1.03 eV)
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32 2SH- Mo16S33 on eMo H2S-Ni1Mo15S33 on pNi eMo 2SH-Ni1Mo15S32 on pNi eMo
33
34 (-0.62 eV) (0.48 eV) (0.46 eV)
35 Figure. 8. Adsorption of H2 and H2S on unpromoted and Ni-promoted Mo16S32 clusters. Catalytic sites are
36
37 indicated by e = edge, c= corner, p = promoted, and indicates bonded and --- indicates nonbonded.
38
39
Color coding: Mo(cyan), C(grey), S(yellow), Ni(blue), and H(white).
40
41
42
43
Table 2. Energy (E) and Binding Energy of H2 and H2S Adsorption (D) on Mo16S32 Clusters at the
44
45 B3PW91/LANL2DZ/6-31G(d,p) Level of Theory.a
46
47 Complex E(Hartree) D(eV)
48 H2-Mo16S32 on eMo -13824.58490 -0.69
49
50 H2---Mo16S32 on c -13824.55957 0.00
51
52 H2---Mo16S32 on eS -13824.55957 0.0
53 2H-Mo16S32 on eMo -13824.59853 -1.06
54
55 2H-Mo16S32_H2 on c -13824.58580 -0.71
56
57
2H-Mo16S32_H2 on eS -13824.57564 -0.44
58
59
60 18
1
2
3
Complex E(Hartree) D(eV)
4
5 2H-Ni1Mo15S32 on pNi eMo -13926.35418 -0.61
6
7 2H-Ni3Mo13S32 on pNi eMo -13824.57564 0.32
8 H2S-Mo16S34 on eMo -15019.22969 -1.03
9
10 2SH-Mo16S33 on eMo -15019.21452 -0.62
11
12 H2S-Ni1Mo15S33 on pNi eMo -14722.75808 0.48
13 2SH-Ni1Mo15S32 on pNi eMo -14722.75898 0.46
14
a
15 For the cases of dissociative H2, a total energy of 1.17752 Ha and a bonding energy of -4.71 eV
16 for H2 was found and considered in the calculations of D.
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33 C4H5S-Mo16S32 on eMo C4H6S-Mo16S32 on eMo C4H7SH-Mo16S32 on eMo
34 (-2.12 eV) (-1.66 eV) (-1.68 eV)
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
C4H6S-Mo16S32 on eMo C4H7SH-Mo16S32 on eMo C4H5SH-Mo16S32 on eMo
50
51 (-2.94 eV) (-3.00 eV) (-3.53 eV)
52
53
54
55
56
57
58
59
60 19
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
T-Mo16S33(SH)2 on eMo C4H5SH-Mo16S32 on eMo C4H5S-Mo16S32H on eMo
16
17 (-1.04 eV) (-3.11 eV) (-0.99 eV)
18
19
20
21
22
23
24
25
26
27
28
29
30
31 C4H5SH-Mo16S32 on eMo C4H6S-Ni1Mo15S33 on pNi eMo
32
33 (-1.99 eV) (2.68 eV)
34
35 Figure. 9. Reaction intermediates formed during hydrogenation and direct desulfurization pathways of
36
37 thiophene after adsorption on Mo16S32 clusters. Mo(cyan), C(grey), S(yellow), B (pink), Ni(blue), and
38
39
H(white).
40
41
42 We study next the HDS reactions of thiophene over MoS2 clusters. Thiophene can undergo both
43
44 HYD as well as DDS pathways for the HDS reaction. We focus on both parallel and
45 perpendicular adsorption modes of thiophene. Thiophene once adsorbed on MoS2 cluster
46
47 undergoes successive hydrogenation to form 2-hydrothiophene and 2,5-dihydrothiophene. We
48
49 obtained the binding energy for each case on pristine MoS2 cluster (Table 3). The HYD
50
51 mechanism follows SC bond scission following the formation of 2,5-dihydrothiophene by
52
53
hydride-shift while DDS follows SC scission immediately after formation of 2-hydrothiophene.
54 The binding energy of final products are also calculated and presented in Table 3. The binding
55
56
57
58
59
60 20
1
2
3
energy of adsorption of reaction intermediates of thiophene HDS are higher for parallel adsorption
4
5 and increases as the reaction progresses.
6
7
8
9
Table 3. Energy (E) and Binding Energy (D) of Reaction Intermediates Formed during HYD/DDS of
10 Thiophene on Mo16S32 Clusters at the B3PW91/LANL2DZ/6-31G(d,p) Level of Theory
11
12 Complex E(Hartree) D (eV)
13
C4H5S-Mo16S32 on eMo -14376.92649 -2.12
14
15 C4H6S-Mo16S32 on eMo -14377.54524 -1.66
16
17 C4H7SH-Mo16S32 on eMo -14378.74467 -1.68
18 C4H6S-Mo16S32 on eMo -14377.59205 -2.94
19
20 C4H7SH-Mo16S32 on eMo -14378.79312 -3.00
21
22 C4H5SH-Mo16S32 on eMo -14377.58055 -3.53
23 T-Mo16S33(SH)2 on eMo -15173.97291 -1.04a
24
25 C4H5SH-Mo16S32 on eMo -14377.57111 -3.11a
26
27
C4H5S-Mo16S32H on eMo -14377.49337 -0.99a
28 C4H5SH-Mo16S32 on eMo -14377.43415 -1.99a
29
30 C4H6S-Ni1Mo15S33 on pNi eMo
31
-14877.42608 2.68a
32 a
The binding energy is the difference between total energy of products and sum of total energy of each
33
34 reactant molecule in these cases.
35
36
37
For perpendicular adsorption mode, MoS bond length decreases initially from 2.37 to 2.31
38
39 in case of 2-hydrothiophene but again increases to 2.38 . The MoS bond length further
40
41 increases to 2.43 during CS bond scission. However in case of parallel adsorption, MoS
42
43 bond length decreases from 2.48 to 2.39 on CS bond scission. In some cases (Figure 9), we
44
45
observe adsorption of H atoms directly to the Mo-edge. However, since hydrogen is continuously
46 supplied to the reaction system, other hydrogen atoms can initiate the HDS reactions via HYD or
47
48 DDS pathways. The role of promoters and nature of catalytic edges in HDS of thiophene has
49
50 been studied already in great detail; we focus on these effects on HDS of DBT molecules later on
51
52 in the manuscript.
53
54
55 In order to reduce costs of desulfurization processes, it is imperative to develop new chemistries
56
57 to improve efficiency of the catalyst. A favorable effect was observed when MoS2 and graphene
58
59
60 21
1
2
3
were combined together in the electrocatalytic process for generation of hydrogen in fuel cells.10
4
5 In HDS process, -alumina is the most commonly used support material for MoS2 catalysts. The
6
7 -alumina support tends to form MoOAl linkages which then require higher sulfiding
8
9 temperature that can cause sintering effects leading to loss of catalytic activity. Graphene on the
10
11
other hand does not form such linkages and the plasmonic properties of graphene can be
12 exploited to improve the selectivity and sensing properties of MoS2 clusters. Boron nitride (BN)
13
14 also form hexagonal monolayer like graphite and can be used to tune the electronic properties of
15
16 the MoS2 cluster. Hence, we study the effect of graphene and BN on the catalytic activities of
17
18 MoS2 cluster by calculating the change in adsorption binding energy of thiophene. The
19
MoS2/graphene and MoS2/BN complexes with and without adsorbed thiophene are shown in
20
21 Figure 10. The MoS2 cluster undergo physisorption on both BN and graphene monolayers. The
22
23 interaction between MoS2 cluster and graphene is stronger than that between MoS2 cluster and
24
25 BN monolayer as observed from higher binding energy (Table 4) as well as the inter-planar
26
27
distance (Figure 10) between the clusters and individual BN and graphene monolayers. We also
28 calculated the adsorption energy of thiophene as a test molecule on MoS2/graphene and
29
30 MoS2/BN complexes. The binding energy of thiophene changes from -2.77 eV for parallel
31
32 adsorption and -1.07 eV for perpendicular adsorption in the pristine MoS2 cluster to -2.41 eV and
33
34 -1.02 eV, respectively, in the MoS2-graphene complex. The MoS bond length also increases
35 from 2.58 for parallel adsorption and 2.37 for perpendicular adsorption for pristine MoS2
36
37 cluster to 2.66 and 2.39 , respectively, in the case of MoS2-graphene complex. On the other
38
39 hand, BN has no such effect on adsorption energetics of thiophene. The binding energy for
40
41 perpendicular adsorption of thiophene however increases in the MoS2-BN complex, but the Mo
42
43
S bond length (2.38) remains more or less unchanged.
44
45
46
47
48
49
50
51
52
53
54
55 B29N29H29 (BN) C58H18 (G)
56
57
58
59
60 22
1
2
3
4
5
6
7
8
9
10
11
12 Mo16S32---G Mo16S32---BN T-Mo16S32---BN
13
14 (-0.33 eV) (-0.13 eV) (-2.76 eV)
15
16
17
18
19
20
21
22
23
24
25
26 T-Mo16S32---BN T-Mo16S32---G T-Mo16S32---G
27
28 (-1.13 eV) (-2.41 eV) (-1.02 eV)
29 Figure. 10. Adsorption of thiophene (T) on BN promoted Mo16S32 clusters. Mo(cyan), C(grey), S(yellow),
30
31 B (pink), N(dark blue), and H(white).
32
33
34
35
36 Table 4. Total (E) and Binding (D) Energy of MoS2 Clusters at the B3PW91/LANL2DZ/6-31G(d,p)
37
38
Complex E (Hartree) D (eV)
39 B29N29H29 (BN) -2322.81462
40
41 C58H18 (G) -1839.49751
42
Mo16S32---BN -16146.20110 -0.13
43
44 Mo16S32---G -15662.89156 -0.33
45
46 T-Mo16S32---BN -16699.20030 -2.76
47 T-Mo16S32---BN -16699.14020 -1.14
48
49 T-Mo16S32---G -16215.87753 -2.41
50
51 T-Mo16S32---G -16215.82658 -1.02
52
53
54 We now focus on the the adsorption and HDS mechanism of DBT on both pristine and Ni-
55
56 promoted MoS2 clusters. Attempts are also made to represent the reactions occuring under
57
58 reaction conditions. DBT can adsorb on Mo-edge using both thiophinic ring and phenyl ring
59
60 23
1
2
3
depending on available sites as shown in Figure 11. The binding energy of adsorption is higher
4
5 when phenyl ring also interacts with Mo atoms present on the Mo-edge as indicated in Table 5.
6
7 The calculated MoS bond length for pristine, sulfur-covered and Ni-promoted MoS2 clusters are
8
9 2.45 , 2.63 and 2.42 , respectively. The binding energy is the largest in the case of pristine
10
11
MoS2 cluster and the shortest when Mo-edge is covered with sulfur atoms.
12
13
14
15
16
17
18
19
20
21
22
23
24
25 DBT DBT-Mo16S32 (A) DBT-Mo16S32 (B)
26 (-1.39 eV) (-1.79 eV)
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
DBT-Mo16S34 DBT-Ni1Mo15S33 on pNi eMo
42 (-0.72 eV) (-1.22 eV)
43
44 Figure 11. Adsorption of dibenzothiophene (DBT) on unpromoted and Ni-promoted Mo16S32 clusters.
45 Binding energies with thiophene inside parentheses. Mo(cyan), C(grey), S(yellow), Ni(blue), and
46
47 H(white).
48
49
50
51
52 Table 5. Total (E) and Binding Energies (D) of Dibenzothiophene (DBT) and MoS2 Clusters at the
53
54
B3PW91/LANL2DZ/6-31G(d,p) Level of Theory
55 Complex E (Hartree) D (eV)
56
57 DBT -860.08575
58
59
60 24
1
2
3
Complex E (Hartree) D (eV)
4
5 Mo16S32 -13823.38206
6
7 DBT-Mo16S32 (A) -14683.51890 -1.39
8 DBT-Mo16S32 (B) -14683.53376 -1.79
9
10 DBT-Mo16S34 -15479.97198 -0.72
11
12 DBT-Ni1Mo15S33 -15183.50790 -1.22
13
14
15 In order to study the two reaction pathways, HYD and DDS, we chose three types of MoS2
16
17 clusters representing different reaction conditions and nature of the surface. When S atoms are
18
19 present on the Mo-edge, we created a single S-vacancy to create coordinately unsaturated sites for
20
adsorption and successive reaction steps of the HDS of DBT. Like thiophene, DBT undergoes
21
22 hydrogenation to form 2,3-dihydrodibenzothiophene. The MoS bond length remains unchanged
23
24 within 0.01 for pristine and Ni-promoted MoS2 clusters on hydrogenation. For the sulfur-
25
26 covered Mo-edge, the MoS bond length reduces by 0.03 on hydrogenation. Once
27
28
hydrogenated, DBT can undergo successive hydrogenation along HYD pathway or undergo SC
29 scission along DDS pathway. The binding energy increases as the phenyl ring undergoes
30
31 successive hydrogenation for all three types of clusters as inferred from Table 6. During HYD
32
33 pathway, DBT can undergo further hydrogenation of phenyl ring or undergo SC bond scission
34
35 leading to mixture of products. On the other hand, DDS of DBT leads to biphenyl as final
36 product.40
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52 C12H10S-Mo16S32 C12H10S-Mo16S34 C12H10S-Ni1Mo15S33
53
(-2.07 eV) (-1.04 eV) (-1.51 eV)
54
55
56
57
58
59
60 25
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15 C12H12S-Mo16S32 C12H12S-Mo16S34 C12H12S-Ni1Mo15S33
16
(-2.14 eV) (-1.08 eV) (-1.57 eV)
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32 C12H11SH-Mo16S32 C12H11SH-Mo16S34 C12H11SH-Ni1Mo15S33
33
34 (-2.10 eV) (-0.89 eV) (-1.19 eV)
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50 C12H9SH-Mo16S32 C12H9SH-Mo16S34 C12H9SH-Ni1Mo15S33
51 (-2.04 eV) (-0.94 eV) (-1.22 eV)
52
53
54
55
56
57
58
59
60 26
1
2
3
4
5
6
7
8
9
10
11
12
13
14 C12H9S-Mo16S32H C12H9S-Mo16S32H C12H8SH-Mo16S34H
15
16 (-1.88 eV) (-3.07 eV) (0.28 eV)
17
18
19
20
21
22
23
24
25
26
27
28
29
30 C12H8-Mo16S34H2 C12H8SH-Ni1Mo15S33H
31
32 0.20 0.01
33 Figure. 12. Hydrogenation and direct desulfurization pathways of DBT after adsorption on
34
35 Mo16S32 clusters: prehydrogenation of phenyl ring (a-c), second hydrogenation step (d-f), C-S
36
37 bond scission in HYD (g-i), CS bond scission in DDS pathway following prehydrogenation step
38
39 (jl), and other secondary products (mq). Mo(cyan), C(grey), S(yellow), B (pink), Ni(blue),
40
41
and H(white).
42
43
44
45
46 Table 6. Energy (E) and Adsorption Binding Energy (D) of Reaction Intermediates Formed During
47 HYD/DDS of DBT on Mo16S32 Clusters at the B3PW91/LANL2DZ/6-31G(d,p) Level of Theory
48
50
51 C12H10S-Mo16S32 -14684.71150 -2.07
52
53 C12H10S-Mo16S34 -15481.15141 -1.04
54
55 C12H10S-Ni1Mo15S33 -15184.68600 -1.51
56
57
58
59
60 27
1
2
4
5 C12H12S-Mo16S32 -14685.94496 -2.14
6
7
C12H12S-Mo16S34 -15482.38362 -1.08
8
9
10 C12H12S-Ni1Mo15S33 -15185.91935 -1.57
11
12 C12H11SH-Mo16S32 -14685.90314 -2.10
13
14 C12H11SH-Mo16S34 -15482.34015 -0.99
15
16 C12H11SH-Ni1Mo15S33 -15185.87351 -1.42
17
18 C12H9SH-Mo16S32 -14684.72455 -2.04
19
20 C12H9SH-Mo16S34 -15481.16172 -0.94
21
22
C12H9SH-Ni1Mo15S33 -15184.68948 -1.22
23
24
25 C12H9S-Mo16S32H -14684.71430 -1.88a
26
27 C12H9S-Mo16S32H -14684.75806 -3.07a
28
29 C12H8SH-Mo16S34H -15481.11273 0.28a
30
31 C12H8-Mo16S34H2 -15481.11577 0.20a
32
33 C12H8SH-Ni1Mo15S33H -15184.64031 0.01a
34
a
35 The binding energy is the difference between total energy of products and sum of total energy of each
36
37 reactant molecule.
38
39
40 4. CONCLUSIONS
41
42 Molecular dynamics performed on hydrocarbon mixture containing thiophenes and
43
44 dibenzothiophenes show that these sulfur-compunds diffuse from bulk to the MoS2 catalytic
45
46
edges. MoS2/graphene complex shows improvement over pristine MoS2 cluster for adsorption of
47 thiophene. A comprehensive analysis of HDS reaction mechanism has been presented using the
48
49 Mo16S32 nanocluster starting from activation of hydrogen and adsorption of sulfur compunds.
50
51 Under typical HDS conditions, Mo-edges are covered with sulfur atoms and therfore vacancies
52
53 are needed for adsorption of sulfur compounds. Both hydrogenation and direct desulfurization
54 pathways proceed first by hydrogenation step. The adsorption or binding energies of thiophene,
55
56 DBT and the reaction intermediates reduce when the Mo-edges are promoted with nickel atoms.
57
58
59
60 28
1
2
3
4
5 ACKNOWLEDGEMENTS
6
7 We would like to acknowledge the support from the TAMU-CONACYT Research Grant Program
8
9 and the Lannater and Herb Fox Professorship as well as the high-performance computing support
10
11
provided by the Texas A&M Supercomputer Facility and the Texas Advanced Computing Center
12 (TACC).
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 29
1
2
3
References
4
5
6
7 1. Stephenson, T.; Li, Z.; Olsen, B.; Mitlin, D., Lithium Ion Battery Applications of
8
9
Molybdenum Disulfide (MoS2) Nanocomposites. Energy Environ. Sci. 2014, 7, 209-231.
10 2. Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S., Electronics and
11
12 Optoelectronics of Two-Dimensional Transition Metal Dichalcogenides. Nat. Nanotechnol.
13
14 2012, 7, 699-712.
15
16 3. Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L.-J.; Loh, K. P.; Zhang, H., The Chemistry of Two-
17 Dimensional Layered Transition Metal Dichalcogenide Nanosheets. Nat. Chem. 2013, 5,
18
19 263-275.
20
21 4. Radisavljevic B; Radenovic A; Brivio J; Giacometti V; Kis A, Single-Layer MoS2
22
23 Transistors. Nat Nano 2011, 6, 147-150.
24
5. Wilson, J. A.; Yoffe, A. D., The Transition Metal Dichalcogenides Discussion and
25
26 Interpretation of the Observed Optical, Electrical and Structural Properties. Adv. Phys. 1969,
27
28 18, 193-335.
29
30 6. Lauritsen, J. V.; Kibsgaard, J.; Olesen, G. H.; Moses, P. G.; Hinnemann, B.; Helveg, S.;
31
32
Nrskov, J. K.; Clausen, B. S.; Topse, H.; Lgsgaard, E., Location and Coordination of
33 Promoter Atoms in Co-and Ni-Promoted MoS2-Based Hydrotreating Catalysts. J. Catal.
34
35 2007, 249, 220-233.
36
37 7. Greeley, J.; Jaramillo, T. F.; Bonde, J.; Chorkendorff, I.; Norskov, J. K., Computational
38
39 High-Throughput Screening of Electrocatalytic Materials for Hydrogen Evolution. Nat.
40 Mater. 2006, 5, 909-913.
41
42 8. Kibsgaard, J.; Chen, Z.; Reinecke, B.; Jaramillo, T., Engineering the Surface Structure of
43
44 MoS2 To preferentially Expose Active Edge Sites for Electrocatalysis. Nat. Mater. 2012, 11,
45
46 963-969.
47
48
9. Laursen, A. B.; Kegnaes, S.; Dahl, S.; Chorkendorff, I., Molybdenum Sulfides - Efficient and
49 Viable Materials for Electro- and Photoelectrocatalytic Hydrogen Evolution. Energy
50
51 Environ. Sci. 2012, 5, 5577-5591.
52
53 10. Li, Y.; Wang, H.; Xie, L.; Liang, Y.; Hong, G.; Dai, H., MoS2 Nanoparticles Grown on
54
55 Graphene: An Advanced Catalyst for the Hydrogen Evolution Reaction. J. Am. Chem. Soc.
56 2011, 133, 7296-7299.
57
58
59
60 30
1
2
3
11. Chang, K.; Chen, W., In Situ Synthesis of MoS2/Graphene Nanosheet Composites with
4
5 Extraordinarily High Electrochemical Performance for Lithium Ion Batteries. Chem.
6
7 Commun. 2011, 47, 4252-4254.
8
9 12. Li, H.; Yin, Z.; He, Q.; Li, H.; Huang, X.; Lu, G.; Fam, D. W. H.; Tok, A. I. Y.; Zhang, Q.;
10
11
Zhang, H., Fabrication of Single- and Multilayer MoS2 Film-Based Field-Effect Transistors
12 for Sensing No at Room Temperature. Small 2012, 8, 63-67.
13
14 13. He, Q.; Zeng, Z.; Yin, Z.; Li, H.; Wu, S.; Huang, X.; Zhang, H., Fabrication of Flexible
15
16 MoS2 Thin-Film Transistor Arrays for Practical Gas-Sensing Applications. Small 2012, 8,
17
18 2994-2999.
19 14. Zhu, C.; Zeng, Z.; Li, H.; Li, F.; Fan, C.; Zhang, H., Single-Layer MoS2-Based Nanoprobes
20
21 for Homogeneous Detection of Biomolecules. J. Am. Chem. Soc. 2013, 135, 5998-6001.
22
23 15. Shembel, E.; Apostolova, R.; Kirsanova, I.; Tysyachny, V., Electrolytic Molybdenum
24
25 Sulfides for Thin-Layer Lithium Power Sources. J. Solid State Electrochem. 2008, 12, 1151-
26
27
1157.
28 16. Bhandavat, R.; David, L.; Singh, G., Synthesis of Surface-Functionalized Ws2 Nanosheets
29
30 and Performance as Li-Ion Battery Anodes. J. Phys. Chem. Lett. 2012, 3, 1523-1530.
31
32 17. Julien, C.; Saikh, S. I.; Nazri, G. A., Electrochemical Studies of Disordered MoS2 as Cathode
33
34 Material in Lithium Batteries. Mater. Sci. Eng., B 1992, 15, 73-77.
35 18. Paul, J.-F.; Cristol, S.; Payen, E., Computational Studies of (Mixed) Sulfide Hydrotreating
36
37 Catalysts. Catal. Today 2008, 130, 139-148.
38
39 19. Ma, X.; Schobert, H. H., Molecular Simulation on Hydrodesulfurization of Thiophenic
40
41 Compounds over MoS2 Using Zindo. J. Mol. Catal. A: Chem. 2000, 160, 409-427.
42
43
20. Wen, X.-D.; Zeng, T.; Li, Y.-W.; Wang, J.; Jiao, H., Surface Structure and Stability of Mosx
44 Model Clusters. J. Phys. Chem. B 2005, 109, 18491-18499.
45
46 21. Byskov, L. S.; Nrskov, J. K.; Clausen, B. S.; Topse, H., DFT Calculations of Unpromoted
47
48 and Promoted MoS2-Based Hydrodesulfurization Catalysts. J. Catal. 1999, 187, 109-122.
49
50 22. Raybaud, P.; Hafner, J.; Kresse, G.; Kasztelan, S.; Toulhoat, H., Structure, Energetics, and
51 Electronic Properties of the Surface of a Promoted MoS2 Catalyst: An Ab Initio Local
52
53 Density Functional Study. J. Catal. 2000, 190, 128-143.
54
55 23. Paul, J.-F.; Payen, E., Vacancy Formation on MoS2 Hydrodesulfurization Catalyst: DFT
56
57 Study of the Mechanism. J. Phys. Chem. B 2003, 107, 4057-4064.
58
59
60 31
1
2
3
24. Raybaud, P.; Hafner, J.; Kresse, G.; Kasztelan, S.; Toulhoat, H., Ab Initio Study of the H2
4
5 H2S/MoS2 GasSolid Interface: The Nature of the Catalytically Active Sites. J. Catal. 2000,
6
7 189, 129-146.
8
9 25. Cristol, S.; Paul, J.; Payen, E.; Bougeard, D.; Clmendot, S.; Hutschka, F., Theoretical Study
10
11
of the MoS2 (100) Surface: A Chemical Potential Analysis of Sulfur and Hydrogen
12 Coverage. J. Phys. Chem. B 2000, 104, 11220-11229.
13
14 26. Sun, M.; Nelson, A. E.; Adjaye, J., On the Incorporation of Nickel and Cobalt into MoS2-
15
16 Edge Structures. J. Catal. 2004, 226, 32-40.
17
18 27. Schweiger, H.; Raybaud, P.; Toulhoat, H., Promoter Sensitive Shapes of Co(Ni)Mos
19 Nanocatalysts in Sulfo-Reductive Conditions. J. Catal. 2002, 212, 33-38.
20
21 28. Egorova, M.; Prins, R., The Role of Ni and Co Promoters in the Simultaneous HDS of
22
23 Dibenzothiophene and Hdn of Amines over Mo/-Al2o3 Catalysts. J. Catal. 2006, 241, 162-
24
25 172.
26
27
29. Cristol, S.; Paul, J.-F.; Payen, E.; Bougeard, D.; Hutschka, F.; Clmendot, S., DBT
28 Derivatives Adsorption over Molybdenum Sulfide Catalysts: A Theoretical Study. J. Catal.
29
30 2004, 224, 138-147.
31
32 30. Travert, A.; Nakamura, H.; van Santen, R. A.; Cristol, S.; Paul, J.-F.; Payen, E., Hydrogen
33
34 Activation on Mo-Based Sulfide Catalysts, a Periodic DFT Study. J. Am. Chem. Soc. 2002,
35 124, 7084-7095.
36
37 31. Sun, M.; Nelson, A. E.; Adjaye, J., Ab Initio DFT Study of Hydrogen Dissociation on MoS2,
38
39 Nimos, and Comos: Mechanism, Kinetics, and Vibrational Frequencies. J. Catal. 2005, 233,
40
41 411-421.
42
43
32. Sun, M.; Nelson, A. E.; Adjaye, J., Adsorption and Dissociation of H2 and H2S on MoS2 and
44 Nimos Catalysts. Catal. Today 2005, 105, 36-43.
45
46 33. Raybaud, P.; Hafner, J.; Kresse, G.; Toulhoat, H., Adsorption of Thiophene on the
47
48 Catalytically Active Surface of MoS2: An Ab Initio Local-Density-Functional Study. Phys.
49
50 Rev. Lett. 1998, 80, 1481.
51 34. Orita, H.; Uchida, K.; Itoh, N., Adsorption of Thiophene on an MoS2 Cluster Model Catalyst:
52
53 Ab Initio Density Functional Study. J. Mol. Catal. A: Chem. 2003, 193, 197-205.
54
55 35. Faye, P.; Payen, E.; Bougeard, D., Cluster Approach of Active Sites in an MoS2 Catalyst. J.
56
57 Mol. Model. 1999, 5, 63-71.
58
59
60 32
1
2
3
36. Yang, H.; Fairbridge, C.; Ring, Z., Adsorption of Dibenzothiophene Derivatives over a MoS2
4
5 Nanoclustera Density Functional Theory Study of StructureReactivity Relations. Energy &
6
7 Fuels 2003, 17, 387-398.
8
9 37. Borges, I.; Silva, A. M.; Aguiar, A. P.; Borges, L. E.; Santos, J. C. A.; Dias, M. H., Density
10
11
Functional Theory Molecular Simulation of Thiophene Adsorption on MoS2 Including
12 Microwave Effects. THEOCHEM 2007, 822, 80-88.
13
14 38. Moses, P. G.; Hinnemann, B.; Topse, H.; Nrskov, J. K., The Hydrogenation and Direct
15
16 Desulfurization Reaction Pathway in Thiophene Hydrodesulfurization over MoS2 Catalysts
17
18 at Realistic Conditions: A Density Functional Study. J. Catal. 2007, 248, 188-203.
19 39. Berit, H.; Poul Georg, M.; Jens, K. N., Recent Density Functional Studies of
20
21 Hydrodesulfurization Catalysts: Insight into Structure and Mechanism. J. Phys.: Condens.
22
23 Matter 2008, 20, 064236.
24
25 40. Bataille, F.; Lemberton, J.-L.; Michaud, P.; Prot, G.; Vrinat, M.; Lemaire, M.; Schulz, E.;
26
27
Breysse, M.; Kasztelan, S., Alkyldibenzothiophenes Hydrodesulfurization-Promoter Effect,
28 Reactivity, and Reaction Mechanism. J. Catal. 2000, 191, 409-422.
29
30 41. Plimpton, S.; Crozier, P.; Thompson, A., Lammps-Large-Scale Atomic/Molecular Massively
31
32 Parallel Simulator. Sandia National Laboratories 2007, 18.
33
34 42. Bobadilla, A. D.; Seminario, J. M., Argon-Beam Induced Defects on Silica-Supported Single
35 Walled Carbon Nanotube. J. Phys. Chem. C 2014, 118, 2829928307.
36
37 43. Bellido, E. P.; Seminario, J. M., Molecular Dynamics Simulations of Ion Bombarded
38
39 Graphene. J. Phys. Chem. C 2012, 116, 4044-4049.
40
41 44. Bobadilla, A. D.; Samuel, E. L. G.; Tour, J. M.; Seminario, J. M., Calculating the
42
43
Hydrodynamic Volume of Poly(Ethylene Oxylated) Single-Walled Carbon Nanotubes and
44 Hydrophilic Carbon Clusters. J. Phys. Chem. B 2013, 117, 343354.
45
46 45. Bobadilla, A. D.; Seminario, J. M., Assembly of a Noncovalent DNA Junction on Graphene
47
48 Sheets and Electron Transport Characteristics. J. Phys. Chem. C 2013, 117, 26441-26453.
49
50 46. Rodrguez-Jeangros, N.; Seminario, J. M., Density Functional Theory and Molecular
51 Dynamics Study of the Uranyl Ion (UO2)2+. J. Mol. Mod. 2014, 20, 2150_1-12.
52
53 47. Bellido, E. P.; Seminario, J. M., Graphene Based Vibronic Devices. J. Phys. Chem. C 2012,
54
55 116, 8409-8416.
56
57
58
59
60 33
1
2
3
48. Martnez, L.; Andrade, R.; Birgin, E. G.; Martnez, J. M., Packmol: A Package for Building
4
5 Initial Configurations for Molecular Dynamics Simulations. J. Comput. Chem. 2009, 30,
6
7 2157-2164.
8
9 49. Rapp, A. K.; Casewit, C. J.; Colwell, K.; Goddard III, W.; Skiff, W., UFF, a Full Periodic
10
11
Table Force Field for Molecular Mechanics and Molecular Dynamics Simulations. J. Am.
12 Chem. Soc. 1992, 114, 10024-10035.
13
14 50. Rappe, A.; Colwell, K.; Casewit, C., Application of a Universal Force Field to Metal
15
16 Complexes. Inorg. Chem. 1993, 32, 3438-3450.
17
18 51. Rappe, A. K.; Goddard III, W. A., Charge Equilibration for Molecular Dynamics
19 Simulations. J. Phys. Chem. 1991, 95, 3358-3363.
20
21 52. Murray, J. S.; Lane, P.; Politzer, P., Simultaneous -hole and Hydrogen Bonding by Sulfur-
22
23 and Selenium-Containing Heterocycles. Int. J. Quantum Chem 2008, 108, 2770-2781.
24
25 53. Berendsen, H. J.; Postma, J. P. M.; van Gunsteren, W. F.; DiNola, A.; Haak, J., Molecular
26
27
Dynamics with Coupling to an External Bath. J. Chem. Phys. 1984, 81, 3684-3690.
28 54. Humphrey, W.; Dalke, A.; Schulten, K., VMD: Visual Molecular Dynamics. J. Mol. Graph.
29
30 1996, 14, 33-38.
31
32 55. Giorgino, T., Computing 1-D Atomic Densities in Macromolecular Simulations: The Density
33
34 Profile Tool for VMD. Comput. Phys. Commun. 2014, 185, 317-322.
35 56. Frisch, M.; Trucks, G.; Schlegel, H.; Scuseria, G.; Robb, M.; Cheeseman, J.; Scalmani, G.;
36
37 Barone, V.; Mennucci, B.; Petersson, G., Gaussian 09, Revision B. 01. Gaussian Inc.,
38
39 Wallingford, CT 2010.
40
41 57. Hohenberg, P.; Kohn, W., Inhomogeneous Electron Gas. Phys. Rev. 1964, 136, B864.
42
43
58. Kohn, W.; Sham, L. J., Self-Consistent Equations Including Exchange and Correlation
44 Effects. Phys. Rev. 1965, 140, A1133.
45
46 59. Becke, A. D., Density Functional Thermochemistry. III. The Role of Exact Exchange. J.
47
48 Chem. Phys. 1993, 98, 5648-5652.
49
50 60. Perdew, J. P.; Burke, K.; Wang, Y., Generalized Gradient Approximation for the Exchange-
51 Correlation Hole of a Many-Electron System. Phys. Rev. B 1996, 54, 16533.
52
53 61. Perdew, J. P.; Chevary, J.; Vosko, S.; Jackson, K. A.; Pederson, M. R.; Singh, D.; Fiolhais,
54
55 C., Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient
56
57 Approximation for Exchange and Correlation. Phys. Rev. B 1992, 46, 6671.
58
59
60 34
1
2
3
62. Perdew, J. P.; Chevary, J.; Vosko, S.; Jackson, K. A.; Pederson, M. R.; Singh, D.; Fiolhais,
4
5 C., Erratum: Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized
6
7 Gradient Approximation for Exchange and Correlation. Phys. Rev. B 1993, 48, 4978.
8
9 63. Perdew, J. P.; Ziesche, P.; Eschrig, H., Electronic Structure of Solids 91. Akademie Verlag,
10
11
Berlin: 1991; Vol. 11.
12 64. Burke, K.; Perdew, J.; Wang, Y.; Dobson, J.; Vignale, G.; Das, M., Electronic Density
13
14 Functional Theory: Recent Progress and New Directions. Dobson, JF 1998, 81-112.
15
16 65. Ma, Y.; Martinez de la Hoz, J. M.; Angarita, I.; Berrio-Sanchez, J. M.; Benitez, L.;
17
18 Seminario, J. M.; Son, S.-B.; Lee, S.-H.; George, S. M.; Ban, C. M., et al., Structure and
19 Reactivity of Alucone-Coated Films on Si and LixSiy Surfaces. ACS Applied Materials &
20
21 Interfaces. 2015, 7, 11948-11955.
22
23 66. Kumar, N.; Seminario, J. M., Recent Advancements in Detection and Disposal of
24
25 Radionuclides Using Graphene Oxide. Ann. Materials Sci. Eng. 2015, 2, 1018 (1-3).
26
27
67. Kumar, N.; Seminario, J. M., Solvation of Actinide Salts in Water Using a Polarizable
28 Continuum Model. J. Phys. Chem. A 2015, 119, 689-703.
29
30 68. Kumar, N.; Sandi, G.; Kaminski, M.; Bobadilla, A.; Mertz, C.; Seminario, J. M., Electron
31
32 Transport in Graphene-Based Nanosensors for Eu(III) Detection. J. Phys. Chem. C 2015,
33
34 119, 12037-12046.
35 69. Salazar-Salinas, K.; Baldera-Aguayo, P. A.; Encomendero-Risco, J. J.; Orihuela, M.; Sheen,
36
37 P.; Seminario, J. M.; Zimic, M., Metal-Ion Effects on the Polarization of Metal-Bound Water
38
39 and Infrared Vibrational Modes of the Coordinated Metal Center of Mycobacterium
40
41 Tuberculosis Pyrazinamidase Via Quantum Mechanical Calculations J. Phys. Chem. B 2014,
42
43
118, 10065-10075.
44 70. Kumar, N.; Seminario, J. M., A Quantum Chemistry Approach for the Design and Analysis
45
46 of Nanosensors for Fissile Materials. In Design and Applications of Nano Materials for
47
48 Sensors, Seminario, J. M., Ed. Springer: 2014.
49
50 71. Crdenas-Jirn, G. I.; Len-Plata, P.; Seminario, J. M., Functionalized Graphene and Cobalt
51 Phthalocyanine Based Materials with Potential Use for Electrical Conduction. In Design and
52
53 Applications of Nanomaterials for Sensors, Seminario, J. M., Ed. Springer: 2014.
54
55
56
57
58
59
60 35
1
2
3
72. Bobadilla, A. D.; Ocola, L. E.; Sumant, A. V.; Kaminski, M.; Kumar, N.; Seminario, J. M.,
4
5 Europium Effect on the Electron Transport in Graphene Ribbons. J. Phys. Chem. C 2015,
6
7 119, 22486-22495.
8
9 73. Hay, P. J.; Wadt, W. R., Ab Initio Effective Core Potentials for Molecular Calculations.
10
11
Potentials for the Transition Metal Atoms Sc to Hg. J. Chem. Phys. 1985, 82, 270-283.
12 74. Hehre, W. J.; Ditchfield, R.; Pople, J. A., SelfConsistent Molecular Orbital Methods. Xii.
13
14 Further Extensions of GaussianType Basis Sets for Use in Molecular Orbital Studies of
15
16 Organic Molecules. J. Chem. Phys. 1972, 56, 2257-2261.
17
18 75. Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; DeFrees, D. J.;
19
20
Pople, J. A., SelfConsistent Molecular Orbital Methods. Xxiii. A PolarizationType Basis
21
22
Set for SecondRow Elements. J. Chem. Phys. 1982, 77, 3654-3665.
23 76. Li, X.; Frisch, M. J., Energy-Represented Diis within a Hybrid Geometry Optimization
24
25 Method. J. Chem. Theory Comput. 2006, 2, 835 - 839.
26
27 77. Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J., Using Redundant Internal Coordinates
28
29 to Optimize Equilibrium Geometries and Transition States. J. Comp. Chem. 1996, 17, 49-56.
30
31
78. Lauritsen, J. V.; Kibsgaard, J.; Olesen, G. H.; Besenbacher, F.; Hinnemann, B.; Helveg, S.;
32 Norskov, J. K.; Clausen, B. S.; Topsoe, H.; Lgsgaard, E., et al., Location and Coordination
33
34 of Promoter Atoms in Co- and Ni-Promoted MoS2-Based Hydrotreating Catalysts. J. Catal.
35
36 2007, 249, 220-233.
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 36
1
2
3
4 TOC
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 37

Computational Chemistry Analysis of Hydrodesulfurization Reactions Catalyzed by Molybdenum Disulfide Nanoparticles

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Computational Chemistry Analysis of Hydrodesulfurization Reactions Catalyzed by Molybdenum Disulfide Nanoparticles

Hochgeladen von

Copyright:

Verfügbare Formate

Page 1 of 37 The Journal of Physical Chemistry

Das könnte Ihnen auch gefallen