Abstract. The adsorption of n-C4H10 on V4O10, SV4, SV5, SV6, SV7 catalysts
were investigated using the density functional theory (DFT). The obtained
calculations indicated that SV4 was the best adsorbent and the interaction between
the methylene group and V–O(1) was the most energetic and therefore favorable
pathway among the studied reactions.
7 trang |
Chia sẻ: thanhle95 | Lượt xem: 4 | Lượt tải: 0
Bạn đang xem nội dung tài liệu Theoretical study of the adsorption of n-C4H10 on V2O5/SBA-15 by density functional theory, để tải tài liệu về máy bạn click vào nút DOWNLOAD ở trên
JOURNAL OF SCIENCE OF HNUE
Chemical and Biological Sci., 2012, Vol. 57, No. 8, pp. 36-42
This paper is available online at
THEORETICAL STUDY OF THE ADSORPTION OF n-C4H10
ON V2O5/SBA-15 BY DENSITY FUNCTIONAL THEORY
Tran Thi Thoa and Nguyen Ngoc Ha
Faculty of Chemistry, Hanoi National University of Education
Abstract. The adsorption of n-C4H10 on V4O10, SV4, SV5, SV6, SV7 catalysts
were investigated using the density functional theory (DFT). The obtained
calculations indicated that SV4 was the best adsorbent and the interaction between
the methylene group and V–O(1) was the most energetic and therefore favorable
pathway among the studied reactions.
Keywords: DFT, ODH, adsorption, activation energy.
1. Introduction
The oxidative dehydrogenation (ODH) of light alkanes is one of the best routes
to produce alkenes. Experimental studies have proposed that the ODH reaction of light
alkanes over vanadium based catalysts occurs following a Mars-van Krevelen redox
mechanism [3, 4] in which the first step is a weak adsorption of light alkanes by interaction
with lattice oxygen. Theoretical and experimental studies of the adsorptions of n-butane
over a catalyst are still rare. This paper presents the calculations in the adsorptions of
n-butane over V4O10, SV4, SV5, SV6 and SV7 catalysts using the DFT method.
2. Content
2.1. Computational methodology
All quantum chemical calculations of the geometry and energies of ground states
and transition-state structures have been performed using the DFT method with nonlocal,
gradient-corrected DFT. The Perdew, Burke and Ernzerhof (PBE) gradient-corrected
functional [5] was employed in the calculation of the exchange-correlation energy. This
method is implemented in the SIESTA code [6].
The Double Zeta basis plus Polarization orbitals (DZP) for valence electrons
has been used for all calculations while core electrons are “frozen” in their atomic
Received June 4, 2012. Accepted October 1, 2012.
Contact Nguyen Ngoc Ha, e-mail address: hann@hnue.edu.vn
36
Theoretical study of the adsorption of n-C4H10 on V2O5/SBA-15 by density functional theory
state and replaced by norm-conserving pseudopotentials [7] in its fully nonlocal
(Kleinman-Bylander). Transition states have been found using the Nudged Elastic Band
(NEB) method [8, 9] and the forces acting on dynamic atoms are all smaller than 0.1
eV/A˚. All calculations are applied using the spin polarization model.
2.2. Result and discussion
2.2.1. The catalytic systems
Figure 1. Structures of some catalytic systems
V4O10 is one of the most stable clusters of V2O5, confirmed by many experimental
studies so it can be used to replace V2O5 crystal.
SV4 stands for V4O10 placed on SBA-15 surface.
SV5, SV6, SV7 catalysts obtained by substituting one, two, three silicon atoms in
the SBA-15 structure by one, two, three vanadium atoms.
All of the catalytic systems contain the vanadyl V–O(1) site and the V–O(2)–V site
(see Figure 1).
Some characters of the O(1) and O(2) sites, which influence the adsorption process
of n-butane, are given as follows:
- The possibility of losing/gaining electron of each site. In the ODH reaction, V+5 is
the oxidizing agent while alkanes act as reducing agents and the O atom acts as a bridge.
Electrons transfer from alkanes to V+5 [1, 13] through the O atoms.
- Steric hindrance at each site.
These factors decided a priority of each direction (O(1) and O(2)).
37
Tran Thi Thoa and Nguyen Ngoc Ha
Many studies have shown that vanadyl V–O(1) is the most active site in the ODH of
C4H10 [2], the most favorable step is an O(1) insertion into the C–H bond of the methylene
group [2 and 10-12].
2.2.2. The adsorption of n-C4H10 on the surface of SBA-15
The optimized result shows that the adsorption is only a physical adsorption with
adsorption energy (Eads)
Eads = Ef – Ei
Eads = -7228.397 - (-7227.897) = -0.500 (eV)
= -11.542 (kcal/mol)
in which Ei and Ef denote the energies of initial and final states.
2.2.3. The adsorption of n-C4H10 on V4O10
Because two different lattice oxygen particles are available on the V4O10 surface
(Figure 1), starting from the C4H10/V4O10 system, four possible interactions between
lattice oxygen particles and butane were considered: CH2–O (1), CH2–O (2), CH3–O (1)
and CH3–O (2).
The activation energies (Ea ) and adsorption energies (Eads ) of the adsorption
processes are shown in Table 1.
Table 1. Ea and Eads corresponded to the interactions
between lattice oxygen and n-butane in the adsorption process
Energies CH2 CH3
O (1) O (2) O (1) O (2)
Ea (kcal/ mol) 21.552 22.521 33.608 27.962
Eads (kcal/ mol) -17.182 -29.233 -15.727 -22.587
From the results presented in Table 1, it is clear that:
- For n-butane, the interaction of the CH2 group with the surface was more likely
than the CH3 group in all cases.
- Regarding the catalyst, it was found that the O(1) was better than the O(2) site in
comparison to the CH2 direction and in contrast to the CH3 direction.
- The CH2–O (1) interaction was the most convenient path of the four reaction
pathways studied because of the lowest energy barrier in relation to other ways. This
conclusion is in good agreement with results from [1].
* Reaction at the O(1) site
The CH2–O (1) interaction involved the insertion of the vanadyl oxygen into a
methylen group to form the transition state denoted as TS.1.1 (see Figure 2).
38
Theoretical study of the adsorption of n-C4H10 on V2O5/SBA-15 by density functional theory
At this TS.1.1, the distance between the C atom and the O(1) atom decreased from
3.036 to 2.580 A˚ and the O–H distance was shortened to 0.975 A˚ indicating that the OH
group was nearly formed (0.960 A˚ in the experimental H–O bond length). In structure 3
(the related intermediate structure), the distances of the C–O (1) bond and H–O (1) bonds
were 1.478 and 0.978 A˚, respectively.
Using the NEB method, an activation barrier of 21.552 kcal/mol was calculated,
indicating that this is the most energetically favorable path of the four reaction paths
studied.
Figure 2. Some structures in the CH2–O (1) direction
The CH3–O (1) interaction started with the insertion of the O(1) into a methyl group
to form the transition state. The calculation resulted in an activation barrier of 33.608
kcal/mol for this reaction, which is 12.056 kcal/mol higher than that of the CH2–O (1)
interaction. Thus, this interaction is less favorable compared with that of CH2–O (1).
This is understandable because the bonding energy of C–OH in the methylene group
is 98.5 kcal/mol, which is weaker than it is in the methyl group (101.0 kcal/mol).
39
Tran Thi Thoa and Nguyen Ngoc Ha
Furthermore, structures in the CH2–O (1) direction are more stable than those in
CH3–O (1) due to the stronger hyperconjugations (called H) from CH3 and CH2 species.
Structures in the CH2–O (1) reaction pathway were stabilized by 5 H, but those in CH3–O
(1) were only stabilized by 2 H.
* Reaction at the O(2) site
The pathways of this site that were studied were similar to those of the O(1) site.
The calculated results revealed that the process of forming transition states TS.2.1 and
TS.2.2 have activation energies of 22.521 and 27.962 kcal/mol, respectively. So, the
CH3 direction was less favorable compared to the CH2 direction at O(2) with regards
to adsorption of n-butane.
For the CH2–O (2) reaction pathway, CH2–O (2) interaction involved the breaking
of a C–H bond in a methylene and the formation of an H–O (2) bond to form the transition
state TS.2.1. This state was stabilized by the hyperconjugations from the CH3 and CH2
groups (5 H).
Similarly, for the CH3–O (2) direction, CH3–O (2) interaction included the breaking
of the C–H bond in a methyl and the formation of an H–O (2) bond to form the transition
state TS.2.2. This TS was stabilized by the hyperconjugations from a CH2 group (2 H).
In addition, when comparing the O(1) and O(2) pathways, it can be seen that the
attaction of the CH2 group at the O(1) site was more convenient than at the O(2) site.
The explanation for this is based on atoms’ ability to lose/ gain electrons and states in the
electron transfer process of the adsorption step of n-butane on the catalyst [2].
Conversely, the O(2) site is more convenient than the O(1) site in the CH3 pathway.
This is due to a steric effect.
2.2.4. The adsorption of n-butane on SV4, SV5, SV6 and SV7 catalysts
The calculation results are revealed in Table 2.
Table 2. Eads, Ea (kcal/ mol) corresponded to adsorption processes
on SV4, SV5, SV6 and SV7
SV5 SV6 SV7 SV4
Systems Ea Eads Ea Eads Ea Eads Ea Eads
CH2–O (1) 21.509 -14.495 21.168 -14.211 22.250 -12.856 19.331 -19.489
CH2–O (2) 26.523 -27.335 25.219 -31.586 28.289 -29.889 28.347 -25.568
CH3–O (1) 35.064 -9.424 32.561 -10.560 35.893 11.321 35.498 -9.186
CH3–O (2) 29.352 -21.831 31.163 -19.367 27.517 -20.605 30.241 -21.423
Table 2 indicated that in the adsorption process:
- In each of the catalytic systems, the CH2–O (1) interaction was the most favorable
pathway.
- The difference between the adsorption and activation energies between pathways
is very little. The CH2–O (2) and CH3–O (1) directions in adsorption on the SV6 catalyst,
40
Theoretical study of the adsorption of n-C4H10 on V2O5/SBA-15 by density functional theory
the CH2–O (1) direction on SV4 and the CH3–O (2) direction on SV7 are most convenient
in terms of thermodynamics because of the low value of activation energies (Ea).
2.2.5. Comparing the adsorption of n-butane on the surface of V4O10, SV4, SV5,
SV6 and SV7 catalytic systems
Table 3. Ea and Eads of adsorption steps of n-C4H10
on V4O10, SV4, SV5, SV6 and SV7 in the CH2 and CH3 direction
Ea (kcal/ mol) Eads (kcal/ mol)
Systems O(1) O(2) O(1) O(2)
CH2 CH3 CH2 CH3 CH2 CH3 CH2 CH3
V4O10 21.552 33.608 22.521 27.962 -17.182 -15.727 -29.233 -22.587
SV4 19.331 35.498 28.347 30.241 -19.489 -9.186 -25.568 -21.423
SV5 21.509 35.064 26.523 29.352 -14.495 -9.424 -27.335 -21.831
SV6 21.168 32.561 25.219 31.163 -14.211 -10.560 -31.586 -19.367
SV7 22.250 35.893 28.289 27.517 -12.856 11.321 -29.889 -20.605
By comparing the activation energies of the adsorptions (see Tables 1 and 2), one
can come to the following conclusions:
- For each catalyst, the CH2–O (1) interaction is the most favorable path.
- In pure V4O10, the interaction between CH2–O (2) is the most favorable.
- In SV4, the CH2–O (1) and CH3–O (2) directions are the most favorable.
- In SV6, the CH3–O (1) direction is pronounced.
Thus, when V4O10 was placed on the SBA-15 forming SV4, SV5, SV6 and SV7
systems, the CH2–O (1) direction is still the most favorable in the adsorption step. For
this direction, the SV4 system is the best catalyst in both the kinetic and thermodynamic
aspects because of the low activation and adsorption energies.
3. Conclusion
This study is a theoretical study on the mechanism of the adsorption step in the
oxidative dehydrogenation of n-C4H10. The catalysts under investigation are based on
V4O10 and V4O10/SBA-15. Some conclusions have been reached:
- SV4 has the best catalytic activity for the adsorption process of n-C4H10 because
of the small activation energy.
- V4O10, SV5, SV6, SV7 can be used as catalysts for the adsorption step, although
the activation energies are slightly higher than when using SV4.
- The CH2–O (1) direction is the most convenient for the adsorption step in the
ODH n-butane on V4O10, SV4, SV5, SV6 and SV7.
41
Tran Thi Thoa and Nguyen Ngoc Ha
REFERENCES
[1] Nguyen Ngoc Ha, Ngo Duc Huyen, Le Minh Cam, 2011. Ab-initio study of effect
of basic MgO to V2O5 catalyst on oxidative dehydrogenation of C3H8 and n-C4H10.
Applied Catalysis A: General 407, pp. 106-111.
[2] Ngoc Ha Nguyen, Tran Thanh Hue, Minh Tho Nguyen, Minh Cam Le, 2009. Theory
and Applications of Computational Chemistry Conference. AIP Conf. Proc. 1102 287,
doi: 10.1063/1.3108387.
[3] Dejoz, A.; Lopez Nieto, J. M.; Melo, F.; Va’zquez, 1997. I. Ind Eng, Chem Res, 36,
2588.
[4] Cortes, I.; Rubio, O.; Herguido, J.; Menendez, 2001. M. Catal To-day, 281, pp. 91-92.
[5] J.P. Perdew, K. Burke, M. Ernzerhof, 1996. Phys. Rev. Lett. 77, 3865.
[6] José.M. Soler, Artacho Emilio, Julian.D. Gale, Alberto García, Javier Junquera, Pablo
Ordejón, Daniel Sánchez-Portal, 2002. J. Phys. Condens. Matter. 14, 2745.
[7] D.R. Hamann, M. Schluter, C. Chiang, 1979. Phys. Rev. Lett. 43, pp. 1494-1497.
[8] H. Jonsson, G. Mills, K.W. Jacobsen, in: B.J. Berne, G. Cicotti, D.F. Coker
(Eds.), 1998. Classical and Quantum Dynamics in Condensed Phase Systems. World
Scientific.
[9] Henkelman Graeme, Blas. P. Uberuaga, Hannes Jonsson, 2000. J. Chem. Phys. 113,
9901.
[10] Zhong xue Wang, Daxi Wang, Zhen Zhao, Yu Chen Jie Lan, 2011. A DFT study
of the structural units SBA-15 messoporous molecular sieve. Computational and
Theoretical Chemistry, 963 pp. 403-411.
[11] Cheng, M. J.; Chenoweth, K.; Oxgaard, J.; Van Duin, A.; Goddard, W. A., III, 2007.
J Phys Chem C, 111, 5115.
[12] Hua-Jin Zhai, Jens Dobler, Joachim Sauer, Lai-Sheng Wang, 2007. J. Am. Chem.
Soc. 129, 13270.
[13] Ngoc Ha Nguyen, Tran Thanh Hue, Minh Tho Nguyen, Minh Cam Le. 2010. J.
Quant. Chem. 110, 2653.
[14] F. Gilardoni, A.T. Bell, A. Chakraborty, P. Boulet, 2000. J. Phys. Chem. B104,
12250.
[15] S.Q. Niu, M.B. Hall, 2000. Chem. Rev., 100, 353.
42