1: \documentclass[11pt]{jacs}
2: \title{A Comparative Study of Structural, Acidic and Hydrophobic properties
3: of Sn-BEA with Ti--BEA using Periodic Density Functional Theory}
4: \author{\large{Sharan Shetty},$^{\dagger\S}$
5: Dilip G. Kanhere,$^{\dagger}$
6: Annick Goursot,$^{\ddagger}$ Sourav Pal$^{\S *}$ \\
7: \it{Contributions from
8: the Centre for Modeling and Simulation,} \\
9: \it{Department of Physics, University of Pune, Pune 411007,} \\
10: \it{National Chemical Laboratory, Pune 411008 and} \\
11: \it{Ecole de Chimie, Montpellier, Cedex 5 France} \\
12: s.pal@ncl.res.in}
13: %\usepackage[T1]{fontenc}
14: \usepackage[super,sort&compress]{natbib}
15: %\usepackage{epsf}
16: \usepackage[dvips]{graphicx}
17: \oddsidemargin = 2pt
18: %\evensidemargin = 2pt
19: %\marginparwidth = 0pt
20: %\marginparpush = 0pt
21: %\voffset = 0pt
22: \textwidth = 400pt
23: %\linespread{1.4}
24: %\usepackage{hyperref}\usepackage{setspace}
25: %\usepackage{multicol}
26: %\usepackage[normalem]{ulem}
27: %\usepackage{color}
28:
29:
30:
31:
32: \begin{document}
33:
34: \maketitle
35:
36: \begin{abstract}
37: Periodic density functional theory has been employed to characterize the
38: differences in the structural, Lewis acidic and hydrophobic properties
39: of Sn--BEA and Ti--BEA. We show that the incorporation of Sn increases
40: the Lewis acidity of BEA compared to the incorporation of Ti. Hence, the
41: present work gives an insight into
42: the role of Sn in increasing the efficiency of the oxidation
43: reactions.
44: The results also justify that the percentage of Sn substituted in BEA
45: is less than Ti. The structural analysis shows that the first coordination
46: shell of Sn is larger than that of Ti. However, the second coordination of
47: both sites remains the same. Moreover, the
48: water resistant properties of these substituted zeolites are quantified.
49: \end{abstract}
50:
51: \section{Introduction}
52: Zeolite beta (BEA) has been used as one of the active
53: catalysts for carrying out
54: several organic reactions such as epoxidation of olefins, \cite{bea1}
55: aromatic and aliphatic alkylation,\cite{bea2} acid catalyzed
56: reactions,\cite{bea3} etc.
57: Some of the important
58: reactions which can be catalyzed by BEA include, the Baeyer--Villiger
59: oxidation (BVO) reaction and the Meerwein--Ponndorf--Verley reduction of
60: aldehydes and Oppenauer's oxidation of alchols (MPVO) reaction.\cite{bea_bvo}
61: The reasons for using BEA as an efficient catalyst are its relatively large
62: pore size, its flexible framework and high acidity.\cite{appl_bea}
63: It has been well established that the
64: acidity of BEA can be finely tuned by the
65: incorporation of various atoms such as B, Al, Ti, Zr, Fe etc.
66: \cite{t_bea}$^-$\cite{ti_bea_hydro}
67: These sites substituted in the BEA framework
68: act as active Bronsted or Lewis acid sites depending upon their
69: valence states.\cite{vansanten}
70: Among these atoms,
71: Ti--substitution in BEA framework has proven to be an active
72: catalyst for the epoxidation of
73: olefins in the presence of H$_2$O$_2$.\cite{ti_bea}$^,$\cite{ti_epo}
74: The other Ti--zeolites, which have been
75: succesfully used for the oxidation of small organic molecules, are the
76: titanium silicates (TS--1, TS--2).\cite{ts-1} Several studies have been
77: reported to understand the differences of the activity and selectivity
78: between these two zeolites.\cite{bea1}$^,$\cite{diff_ts}
79: Corma et al have shown that these differences
80: are due to the hydrophilic/hydrophobic nature of the Ti sites.\cite{bea1}$^,$
81: \cite{ti_bea_hydro}
82: They showed that the Ti--sites in TS are more hydrophobic
83: than the Al--Ti--BEA. Hence, TS was prefered over Al--Ti--BEA when
84: the solvent used in the reaction is prepared in aqueous medium.
85:
86: One of the challenges in this field is to increase the efficiency of a zeolite
87: by substitution with other elements.
88: Such an attempt has been made recently by
89: incorporating Sn in BEA. Mal and Ramaswamy succesfully synthesized the
90: Al--free Sn--BEA.\cite{rama_bea}
91: In an interesting experimental work, Corma {\it et al} showed
92: that the incorporation of Sn in the BEA framework results into a more
93: efficient catalyst for the BVO reaction in the presence
94: of H$_2$O$_2$.\cite{corma_nature}
95: In their study, a new mechanism was proposed for the
96: oxidation of ketones. They showed that the Sn site in BEA activates the
97: carbonyl group of the cyclohexanone followed by the attack of H$_2$O$_2$,
98: unlike the Ti sites which initially activate the H$_2$O$_2$. This result
99: was attributed to the higher Lewis acidity of the Sn site with respect to
100: the Ti site.
101: Hence, incorporation of Sn in BEA leads to a high selectivity towards
102: the formation of lactones in the BVO reaction.\cite{corma_nature}$^,$
103: \cite{renz_cej}
104: On this background, highly selective MPVO reactions were carried out more
105: efficiently with Sn--BEA than Ti--BEA.\cite{corma_mpvo} In these studies,
106: it was shown that the Sn site is situated within the framework and no
107: extraframework Sn was detected.
108: Although much of the experimental studies have
109: focused on the efficiency of the Sn--BEA, the higher Lewis acidity of the
110: Sn site compared to the Ti site in BEA is still not known. Recently,
111: Sever and Root used the M(OH)$_4$ (M = Sn, Ti) cluster models to
112: investigate the reaction pathways for the BVO reaction.\cite{sever}
113:
114: The activity and selectivity of the zeolite mainly depend on the nature of
115: the active sites, such as local coordination, interaction
116: with the incoming molecules, percentage of substitution of T atoms
117: in the framework, etc.
118: One of the important issues concerning the activity and selectivity
119: of the zeolite is its hydrophobic/hydrophilic nature.\cite{corma_hydro}$^,$
120: \cite{weitkamp}
121: It is known that if the zeolite is hydrophilic in nature, the water present in
122: the solvent poisons the active sites. This hinders the kinetics
123: of the reaction and decreases the activity
124: of the zeolite. Corma {\it et al} have bypassed this problem
125: by modifying the catalyst design, which allows the use of Sn--BEA in the
126: presence of aqueous media.\cite{sn_hydro} Very recently, Boronat {\it et al}
127: have done theoretical calculations using a Sn(OSiH$_3$)$_3$OH cluster model
128: to understand the effect of H$_2$O during the BVO reaction. Their results
129: show that one water molecule is permanently attached to the Sn active site.
130: Interestingly,
131: Fois {\it et al} have studied the interaction of water molecules with the
132: Ti sites in Ti--Offretite using Car--Parrinello molecular dynamics.\cite{fois}
133: They found that at higher loading of water molecules, the Ti atom expands its
134: coordination number.
135:
136: In the last decade, several experimental and theoretical studies
137: have been employed
138: to characterize the role of Ti sites at a microscopic level in various
139: Ti-zeolite systems.\cite{bea1}$^,$\cite{ti_bea}$^,$\cite{dovesi}
140: It has been revealed that due to high crystallinity, low Ti content
141: and large quadrupolar moment
142: of Ti, accurate information on the Ti sites in BEA is not possible through
143: experimental techniques.\cite{notary}
144: Hence, it is necessary to use theoretical
145: methods to explore the local behavior for eg. structure, electronic and
146: bonding properties of these sites.
147: Sastre and Corma have used {\it ab initio} calculations to discuss
148: the role of the Ti sites in Ti--BEA and TS--1.\cite{sastre}
149: The energies of the lowest
150: unoccupied molecular orbital (LUMO)
151: of Ti--BEA and TS--1 with one Ti substituted in turn at every T site,
152: were shown to be different. Furthermore,
153: the Ti--sites in Ti--BEA were found to be more acidic than in TS--1 and
154: this acidity varies among all the Ti--sites in both
155: zeolites.\cite{sastre}
156: This proves that not only different Ti--containing zeolites have
157: different acidity, but also different T--sites within a particular
158: zeolite would have varying acidity. Very recently, Bare et al have used
159: EXAFS technique to investigate the Sn--site in Sn--BEA.\cite{bare}
160: They showed that
161: Sn was not randomly distributed in BEA, and takes specific crystallographic
162: sites, i.e. T5/T6 sites in their nomenclature, which corresponds to T1 and T2
163: in our nomenclature, following Newsam {\it et al}.\cite{newsam}
164: Surprisingly, they found that this substitution
165: takes place through pairing of these sites, within the six membered ring i.e.
166: two T1 or two T2.
167: However, no explanation was given
168: for this distribution.
169: At the same time, in a theoretical work using a periodic approach based on
170: density functional theory (DFT) we characterized the Sn--sites
171: in BEA.\cite{shetty} We showed that the T2--site
172: would be the most probable site for the Sn substitution
173: based on thermodynamics consideration of the largest
174: cohesive energy in a dehydrated BEA zeolite.
175: Moreover, we found
176: that the substitution of 2 Sn atoms per unit cell was thermodynamically
177: unfavorable. This was consistent with the earlier experimental results.
178: Parallel to this, Boronat {\it et al} carried out a cluster calculation
179: of the Sn--BEA interaction with cyclohexanone and H$_2$O$_2$.\cite{boronat}
180: They showed that the BVO reaction in Sn--BEA is due to the activation of
181: cyclohexanone at the Sn site.
182:
183: As it can be seen from the above description,
184: the incorporation of Sn in BEA proves it to be a
185: better catalytic site than Ti. Hence, a detailed information
186: on the differences in the properties of Sn and Ti sites in BEA, such as
187: the quantification of the Lewis acidity, number of T atoms to be substituted
188: in the unit cell and hydrophobicity, are of fundamental
189: importance and are still to be resolved.
190: The aim of the present theoretical study is to bring out the differences in
191: these substituted BEA zeolites by analyzing their structural,
192: electronic and hydrophobic properties.
193: Moreover, it is also important
194: to understand why the framework Sn site activates the carbonyl group of
195: cyclohexanone and not H$_2$O$_2$ in the BVO reaction, while Ti
196: behaves the other way.
197: We investigate this issue on the basis of hard--soft acid--base (HSAB)
198: principle.
199:
200: \section{Methodology and Computational Details}
201: Several theoretical studies based on a classical as well as
202: quantum potential have been proposed to study the properties of
203: zeolites.\cite{sastre}$^,$\cite{jentys}$^{--}$\cite{rozanska}
204: It has been a practise to adopt cluster models cut from the
205: zeolite crystals to study these properties. One of the obvious reasons
206: to use cluster model is that it is computationally cheap.
207: Sauer {\it et al} have done an extensive study of zeolites using cluster
208: models.\cite{deman}$^,$\cite{sauer1}
209: However, the active site represented in the
210: cluster model is in a different electronic environment
211: than in which it would be in
212: a crystal.\cite{sauer1}$^,$\cite{nicholas}$^,$\cite{rozanska}
213: A periodic approach provides a more realistic
214: description to study the properties
215: of a crystal.\cite{nicholas}$^,$\cite{dovesi}
216: Although zeolite catalysts are neither crystals nor periodic solids,
217: it is more convenient
218: to use periodic boundary conditions, when there are very few substituted
219: sites per unit cell.
220:
221: \begin{figure}
222: \caption{Crystallographically defined 9 T sites of BEA. The
223: grey spheres represent the Si sites.}
224: \begin{center}
225: \includegraphics[width=0.5\textwidth]{bea_black.ps}
226: \end{center}
227: \label{Fig. 1}
228: \end{figure}
229:
230:
231: Earlier experimental studies have indeed proved
232: that Sn and Ti sites in BEA are very few, that they are
233: situated within the framework and during
234: the BVO or MPVO reaction these sites do not dissociate from the
235: framework.\cite{bea_bvo}$^{(a)}$$^,$\cite{renz_cej}
236: In the present work we have employed the periodic DFT to investigate the
237: properties of Sn--BEA and Ti--BEA. Advantage of using periodic boundary
238: conditions is that the long range electrostatic interactions are included
239: within Ewald summations.
240: The instantaneous stationary electronic ground state is
241: calculated by solving the Kohn-Sham equation based on DFT. The valence
242: electrons have been represented by the plane waves in
243: conjunction with the Vanderbilt's ultra--soft pseudopotential for
244: the.\cite{vanderbilt}
245: It is worth mentioning that during the interaction between two systems the
246: complete plane wave avoids the basis set superposition error.
247: The exchange correlation functional is expressed by the
248: the generalized gradient approximation (GGA) with
249: the Perdew--Wang 91 functional.\cite{pw91}
250: The calculations were restricted to
251: the gamma point in the Brillouin zone sampling. All the calculations
252: have been performed by the VASP code.\cite{vasp}
253:
254: BEA is a high silica zeolite and consists of two different ordered
255: polytypic series {\it viz.} polymorph A and polymorph B.\cite{newsam}
256: It has two
257: mutually perpendicular straight channels with a cross section of
258: 0.76*0.64 nm which run along a and b directions.
259: Intersecting to these, at right angles,
260: a helical channel of 0.55 x 0.55 nm also
261: exists along the c-axis. This gives rise to a three dimensional
262: pore system of 12--membered ring aperture. The unit cell of an ideal fully
263: siliceous BEA
264: consists of 192 atoms with 64 Si and 128 O atoms distributed within
265: the tetragonal lattice of dimension 12.6 x 12.6 x 26.2 {\AA}.
266: There are 9 distinct crystallographically defined T sites as shown in Fig. 1.
267: We adopt the experimental structure as defined by Newsam {\it et al} and
268: accordingly define the 9 T sites in the unit cell of BEA.\cite{newsam}
269:
270: The structural optimization of the Si, Sn and Ti--BEA have been carried out
271: in two steps. In the first step, conjugate gradient method was used to
272: optimize the unit cell of BEA. The optimization was considered to
273: achieve when the forces on the atoms were less than 0.1 eV/{\AA}. In the
274: second step, these optimized geometries were re--optimized with
275: quasi--Newton method unless the forces on the atoms were less than
276: 0.06 eV/{\AA}. One should note that during the optimization the
277: cell shape of the unit cell has been fully relaxed, while keeping its
278: volume constant. In the case of Sn and Ti--BEA each of the 9 distinct
279: T sites were substituted by Sn and Ti atoms (i.e.
280: Si/(Sn or Ti) = 63/1, respectively, and were optimized. Once the active
281: site in Sn and Ti--BEA was confirmed, one water molecule was introduced near
282: to these active sites and the same optimization procedure was followed as
283: discussed above. The
284: structural data for the Sn--BEA has been taken from a recent publication by
285: us.\cite{shetty}
286: In addition to this, we have also carried out DFT calculations of
287: cyclohexanone and H$_2$O$_2$ molecules using a supercell. These systems
288: were optimized by the conjugate gradient methods only, until the
289: forces on the atoms were less than 0.005 eV/{\AA}.
290:
291: \section{Results and Discussion}
292: \subsection{Structure of Sn--BEA and Ti--BEA}
293: We have already discussed the structure and energetics of Sn--BEA in a
294: recent publication.\cite{shetty}
295: In the present work we briefly recall this discussion
296: which is necessary for comparing it with the structure and energetics of
297: Ti--BEA and also to study its hydrophobic characteristic.
298:
299: Table 1 and Table 2 present the optimized
300: structural details of all the 9 T sites of Sn--BEA and Ti--BEA, respectively.
301: It should be noted that only the average bond distances and bond angles
302: are presented. It can be seen from Table 1 that the Sn--O bond distance
303: range between 1.908 to 1.917 {\AA}, the Sn--O--Si bond angle range
304: from 136 to 144.2 deg and the Sn--Si distance is around 3.241$\pm$0.100 {\AA}.
305: Very recently, Bare {\it et al}, with the help of EXAFS
306: technique showed that the Sn--O bond distances and the Sn--Si distances
307: in Sn--BEA were around
308: 1.906 {\AA} and 3.5 {\AA}, respectively.\cite{bare}
309: This clearly shows that the theoretical results presented
310: by us are in good agreement with the experimental results. However, the
311: theoretical results of Bare {\it et al}
312: were not consistent with their experimental
313: data. This may be due to keeping the shape of the unit cell fixed during
314: the optimization and using the local density approximation
315: exchange--correlation potential in their study.\cite{bare}
316: On the other hand, we have relaxed the
317: lattice vectors of the unit cell during the optimization and also
318: used the GGA exchange--correlation potential, as explained in the
319: earlier section. The change in the local coordination of the
320: T site in Sn--BEA compared to the Si--BEA has been illustrated in
321: the earlier study. As already mentioned above, sites
322: T5 and T6 of ref. 25 correspond to sites T1 and T2 in our work, in which we use
323: the nomenclature of Newsam et al.\cite{newsam}
324:
325: \begin{table}
326: \begin{center}
327: \caption{Optimized structural parameters of Sn--BEA.
328: Average Sn--O bond lengths, Sn--O--Si bond angles and Sn--Si distances of
329: all the 9 T sites of Sn--BEA.}
330: \begin{tabular}{|l|l|l|l|}
331: \hline
332: T--sites & Sn--O (\AA) & Sn--O--Si (deg) & Sn--Si (\AA)\\
333: \hline
334: T1 & 1.911 & 143.5 & 3.336 \\
335: \hline
336: T2 & 1.909 & 144.2 & 3.341 \\
337: \hline
338: T3 & 1.910 & 140.6 & 3.241 \\
339: \hline
340: T4 & 1.917 & 136.0 & 3.281 \\
341: \hline
342: T5 & 1.913 & 142.2 & 3.297 \\
343: \hline
344: T6 & 1.910 & 141.2 & 3.297 \\
345: \hline
346: T7 & 1.911 & 140.6 & 3.282 \\
347: \hline
348: T8 & 1.908 & 140.0 & 3.282 \\
349: \hline
350: T9 & 1.912 & 137.8 & 3.270 \\
351: \hline
352: \end{tabular}
353: \end{center}
354: \end{table}
355:
356: \begin{table}
357: \begin{center}
358: \caption{Optimized structural parameters of Ti--BEA.
359: Average Ti--O bond lengths, Ti--O--Si bond angles, Ti--Si distances of all the
360: 9 T sites of Ti--BEA.}
361: \begin{tabular}{|c|c|c|c|}
362: \hline
363: T--sites & Ti--O (\AA) & Ti--O--Si (deg) & Ti--Si (\AA) \\
364: \hline
365: T1 & 1.799 & 151.7 & 3.302 \\
366: \hline
367: T2 & 1.797 & 152.4 & 3.304 \\
368: \hline
369: T3 & 1.794 & 145.0 & 3.220 \\
370: \hline
371: T4 & 1.797 & 145.4 & 3.233 \\
372: \hline
373: T5 & 1.799 & 148.1 & 3.257 \\
374: \hline
375: T6 & 1.799 & 148.5 & 3.263 \\
376: \hline
377: T7 & 1.794 & 149.0 & 3.269 \\
378: \hline
379: T8 & 1.795 & 147.4 & 3.249 \\
380: \hline
381: T9 & 1.798 & 144.0 & 3.225 \\
382: \hline
383: \end{tabular}
384: \end{center}
385: \end{table}
386: Table 2 shows that the average Ti--O bond distances of the 9 T sites in BEA
387: vary from 1.794 to 1.799 {\AA}. These values are in good agreement with the
388: earlier works on Ti--BEA.\cite{ti_bea}
389: Compared to Sn--O bond distances, the Ti--O distances
390: are smaller. This is due to the larger atomic size of Sn with respect to Ti.
391: From the data of Tables 1 and 2, it can be noticed that the average Sn-O
392: and Ti-O bond lengths are very similar for all T sites, whereas
393: the corresponding bond
394: angles have a large range of variation. Moreover, in both Sn and Ti--BEA
395: models, the largest average angles belong to the T1 and T2 sites. The average
396: experimental values of T1--O--T and T2--O--T
397: angles in the unsubstituted Si-BEA are
398: 155.3 and 155.9 deg respectively, and they also correspond to the largest
399: T--O--T angles in the framework.
400: If we compare Sn and Ti substituted in the framework with Si, we get the
401: expected order for average T-O bond lengths Sn-O$>$Ti-O$>$Si-O,
402: with around 0.12 to 0.15 {\AA} difference at each replacement.
403: The average T1--O--T or T2--O--T
404: bond angles vary as Sn--O--Si$<$Ti--O--Si$<$Si--O--Si.
405:
406: This Ti--O--Si bond angles which range between 144 to 152 deg,
407: are larger than the Sn--O--Si bond angles with a range between
408: 136.0--143.5 deg.
409: Due to the angular flexibility, the Ti--Si distance differ
410: only by $\sim$0.04 {\AA} from the Sn--Si
411: distance. Although the first coordination shell radius of Ti is
412: smaller than that of Sn, the second coordination shells are at similar
413: distances.
414: The adaptation of the BEA framework to Sn and Ti substitution
415: results thus into a quite localized deformation of the siliceous framework.
416: Hence, we can
417: infer that the difference in adsorption properties between Sn and Ti-BEA
418: should be mainly due to the electronic differences of these sites.
419:
420:
421: \subsection{Energetics of Sn--BEA and Ti--BEA}
422: In this subsection, we discuss the thermodynamic stability of Sn--BEA and
423: Ti--BEA. This is done by calculating the cohesive energies for each of the
424: 9 T--sites in Sn--BEA and Ti--BEA. Cohesive energy of a solid is
425: defined as the difference between the energy of the bulk (solid) at
426: equilibrium and the energy of the constituent atoms in their
427: ground state.\cite{shetty}
428: Cohesive energy does not account for the kinetic formation of the system,
429: neither for the different nature of the synthesis intermediates
430: generated in aqueous solution, which can generate different routes for the
431: solid growth.
432:
433: The cohesive energies of all the 9 substituted T--sites of Sn--BEA and Ti--BEA
434: are given in Table 3. In our earlier investigation on the energetics
435: of Sn--BEA, we showed that the substitution of Sn in the BEA framework
436: decreases the cohesive energy.\cite{shetty}
437: Hence, the incorporation of Sn in BEA was shown to be
438: thermodynamically less stable than the Si--BEA. On this basis, we
439: explained the fact that the incorporation of Sn in the BEA framework is
440: restricted. Interestingly, Bare {\it et al} predicted the formation of
441: Sn pairs as the active sites, where the two Sn atoms were shown to
442: be on the opposite sides of a six membered ring.\cite{bare}
443: They showed that one of these pairs is present per
444: 8 u.c of BEA. Unfortunately, at present, it is out of
445: scope to consider
446: 8 u.c of BEA. Nevertheless, we have carried out the calculations placing
447: 2 Sn atoms per u.c. at T1 and T2 (T5 and T6 according to Bare et al)
448: positions which are situated in the six membered ring and
449: are on the opposite side of each other (Fig. 1). We found that this does not
450: increase the cohesive energy.
451:
452: \begin{table}
453: \begin{center}
454: \caption{Cohesive energies of all the 9 T--sites of Sn--BEA and Ti--BEA}
455: \begin{tabular}{|l|l|l|}
456: \hline
457: &\multicolumn{2}{l|}{Cohesive Energies (eV)}\\
458: \cline{2-3}
459: T--Sites & Sn--BEA & Ti--BEA \\
460: \hline
461: T1 & -1521.387 & -1530.797 \\
462: \hline
463: T2 & -1521.681 & -1530.767 \\
464: \hline
465: T3 & -1521.468 & -1530.210 \\
466: \hline
467: T4 & -1521.523 & -1530.045 \\
468: \hline
469: T5 & -1521.405 & -1530.014 \\
470: \hline
471: T6 & -1521.431 & -1530.570 \\
472: \hline
473: T7 & -1521.457 & -1530.359 \\
474: \hline
475: T8 & -1521.621 & -1530.415 \\
476: \hline
477: T9 & -1521.323 & -1530.282 \\
478: \hline
479: \end{tabular}
480: \end{center}
481: \end{table}
482:
483: The cohesive energy of Si--BEA is -1527.902 eV.\cite{shetty}
484: From Table 3 we see that the
485: cohesive energy of Ti--BEA is about 3 eV higher than that of Si--BEA. This
486: indicates that the incorporation of Ti in BEA is thermodynamically
487: more favorable than that of Sn.
488: Among the 9 T--sites of Ti--BEA we found that the T1 and T2 sites
489: have the highest stability, and that T5 is the least stable.
490: We have also calculated the cohesive energy
491: with two Ti/u.c (i.e. Ti/Si = 2/62 per u.c). The two Ti atoms
492: were incorporated at two different T2 positions at a distance of 9 {\AA}.
493: This showed an increase in the cohesive energy of about 3 eV compared to
494: one Ti/u.c. This reveals that more Ti could be incorporated in BEA than
495: Sn.
496: We want to stress that these calculations are carried out on a dehydrated
497: solid resulting from a thermodynamically driven synthesis, ignoring the
498: effects of the various ingredients and formation conditions, i.e. the nature
499: and energies of the synthesis intermediates.
500: Nevertheless,
501: these results are consistent with the earlier experimental works, where it
502: has been shown that the amount of incorporated Ti is larger than that of Sn in
503: BEA.\cite{ti_bea}$^,$\cite{renz_cej}
504:
505: \subsection{Lewis acidity of Sn--BEA and Ti--BEA}
506: Earlier experimental studies have conjectured that Sn acts as a better Lewis
507: acidic site than Ti in BEA.\cite{corma_nature}$^{--}$\cite{corma_mpvo}
508: Hence, Sn--BEA acts as a more active catalyst for the
509: oxidation reactions. This motivated us to compare the Lewis acidity of
510: Sn and Ti--BEA.
511: First, one must recall that the Lewis acidity, being related
512: with an electron acceptor character, can be correlated with the global
513: electron affinity of the solid. Qualitatively, LUMO energies can be used for a
514: comparison between the electron affinities of Sn and Ti-BEA.\cite{sastre}
515: The HOMO and the LUMO energies, and their corresponding HOMO--LUMO gaps of
516: Sn--BEA and Ti--BEA have been reported in Table 4.
517: Globally, the average LUMO energy among the Sn substituted models is lower
518: than that for the Ti ones.
519: In our earlier
520: results on Sn--BEA, we have shown that out of the 9 T--sites the
521: T1 and the T2 sites have low LUMO energies compared to the other T--sites,
522: and would be the
523: probable sites for the reaction.\cite{shetty}
524: Interestingly, T1 and T2 have been
525: proposed as the most probable sites for Sn substitution from EXAFS
526: experiments.\cite{bare}
527: The two corresponding LUMO's have similar low energies, making
528: these two models good candidates as Lewis acids. Both sites have also the
529: smallest HOMO-LUMO gap. A smaller gap, in a solid, correlates with a larger
530: global softness. The most probable Sn-BEA solids would thus correspond to the
531: most Lewis acidic and the more ``soft" models.
532:
533: \begin{table}
534: \begin{center}
535: \caption{Energies of the HOMO, LUMO and the HOMO--LUMO gaps
536: of the 9 T--sites of Sn--BEA and Ti--BEA}
537: \begin{tabular}{|l|l|l|l|l|l|l|}
538: \hline
539: \multicolumn{1}{|c|}{T-Sites}&\multicolumn{3}{c|}{Sn--BEA}&\multicolumn{3}{c|}{Ti--BEA} \\
540: \cline{2-7}
541: & HOMO (eV) & LUMO (eV) & Gap (eV) & HOMO (eV) & LUMO (eV) & Gap (eV) \\
542: \hline
543: T1 & -3.124 & 1.333 & 4.457 & -3.135 & 1.417 & 4.552 \\
544: \hline
545: T2 & -3.125 & 1.366 & 4.491 & -3.133 & 1.469 & 4.602 \\
546: \hline
547: T3 & -3.131 & 1.557 & 4.688 & -3.121 & 1.548 & 4.669 \\
548: \hline
549: T4 & -3.117 & 1.421 & 4.538 & -3.120 & 1.492 & 4.612 \\
550: \hline
551: T5 & -3.131 & 1.450 & 4.581 & -3.152 & 1.500 & 4.652 \\
552: \hline
553: T6 & -3.120 & 1.426 & 4.546 & -3.145 & 1.453 & 4.598 \\
554: \hline
555: T7 & -3.121 & 1.419 & 4.540 & -3.156 & 1.486 & 4.642 \\
556: \hline
557: T8 & -3.117 & 1.497 & 4.614 & -3.144 & 1.470 & 4.620 \\
558: \hline
559: T9 & -3.114 & 1.506 & 4.620 & -3.121 & 1.454 & 4.575 \\
560: \hline
561: \end{tabular}
562: \end{center}
563: \end{table}
564:
565: In the case of Ti--BEA, we can see from Table 4 that the T1 site has the
566: lowest LUMO energy, whereas T3 has the highest. We can also notice that
567: T1 and T2, which have the largest cohesive energies, have also low HOMO-LUMO
568: gaps, T1 having the smallest.
569: Considering these two factors together, we propose that these sites would
570: be also the most favorable sites for the substitution by Ti and
571: also for the reaction to take place.
572: We propose thus that, in both cases, Sn and Ti would be more probably
573: substituted at the T1 and T2 sites. Considering their LUMO energies, about 0.1
574: eV lower for Sn-BEA, we can infer that Sn--BEA is more Lewis acidic
575: than Ti--BEA. Moreover, the corresponding HOMO-LUMO gaps being lower for Sn-
576: BEA than for Ti-BEA, this also suggests that Sn-BEA is a softer acid. This
577: conclusion is also supported by the following trend: whereas the cohesive
578: energies of the T1-T9 substituted Sn-BEA solids spreads on 0.36 eV, those of
579: the Ti-BEA solids spread over 0.66 eV. Despite its smaller radius, Ti has thus
580: less ability to adapt to the various geometric environments, showing the
581: behaviour of a "harder" species.
582:
583: \subsection{Hydrophobicity of Sn--BEA and Ti--BEA}
584: One of the important issues concerning the selectivity towards the
585: organic molecules in zeolites, is the hydrophobic character of these
586: catalysts.\cite{corma_hydro}
587: Indeed, for reactions such as BVO and MPVO in the
588: presence of aqueous solvents,
589: zeolites containing both, Lewis acidity and hydrophobicity would
590: be the most appropriate.\cite{sn_hydro}$^,$\cite{boronat}
591: In fact, being a product of reaction, water is always present in the catalyst
592: pores. However, this presence is not desirable, because its adsorption is
593: competitive with that of reactants and also due to the product hydrolysis. On
594: a perfect silicate surface, water is physisorbed, i.e., its interaction energy
595: is weak, mainly due to van der Waals forces. As soon as defects are present,
596: water may bind to the silanols or dissociate and react with the surface
597: \cite{ma}
598: In order to be
599: hydrophobic, zeolites must thus present less or no defects.
600: If this is achieved, i.e.
601: for highly hydrophobic samples, experimental results show that substituted Ti-
602: BEA is much more hydrophobic than Sn-BEA.\cite{sn_hydro}
603: Although it is hardly possible to compare Ti-BEA and Sn-BEA with a high
604: loading of water, it is
605: of particular interest to investigate, at the microscopic level,
606: the coordination of
607: Sn and Ti sites in presence and absence of one water molecule.
608: For this comparison, Sn and Ti have been located at
609: sites T2 and T1, respectively.
610: The full systems have then been optimized.
611:
612:
613: Table 5 gives the averaged optimized T--O(BEA), T--OH$_2$
614: bond lenghths, T--O--Si
615: bond angles and the T--Si distances, where T = Sn and Ti. We can see that
616: after hydration, the Sn--O distance has been increased by 0.014 {\AA} and the
617: Sn--O--Si angle is also increased by about 2.3 degs
618: with respect to the dehydrated Sn--BEA.
619: The bond distance between the Sn site and the H$_2$O
620: is 2.41 {\AA}.
621: The hydrated Ti--BEA shows a similar trend with a
622: Ti--O bond length and the Ti--O--Si bond angle which have been increased
623: by 0.019 {\AA} and 2.9 degs, respectively. The Ti--OH$_2$ bond distance is
624: 2.35 {\AA}. We see that the Sn--OH$_2$ distance is longer than Ti--OH$_2$.
625: In order to understand the adsorption of the H$_2$O molecule to the T sites,
626: we have calculated the binding energy (B.E.)(Table 5). This
627: is done as follows
628:
629: \begin{displaymath}
630: B.E. = E_{complex}(BEA+H_{2}O)-\{E(BEA)+E(H_{2}O)\}
631: \end{displaymath}
632:
633: As can be seen from Table 5, the B.E. is positive for both systems.
634: This shows that the
635: formation of the complex is less stable
636: than the separate entities (endothermic), and that water
637: molecules do not like to form a stable complex with either of the sites
638: {\it viz.} Sn and Ti in BEA.
639: Furthermore, the Ti--BEA and H$_2$O complex is $\sim$3 kJ/mol
640: less stable than the Sn--BEA and H$_2$O complex.
641:
642: \begin{table}
643: \begin{center}
644: \caption{Structural parameters and the
645: binding energies (B.E.) of Sn--BEA and Ti--BEA in the presence of
646: H$_2$O}
647: \begin{tabular}{|l|l|l|}
648: \hline
649: & Sn--BEA + H$_2$O & Ti--BEA + H$_2$O \\
650: \hline
651: T--O(BEA) (\AA) & 1.923 & 1.818 \\
652: \hline
653: T--O--Si (deg) & 146.50 & 154.66 \\
654: \hline
655: T--Si (\AA) & 3.369 & 3.330 \\
656: \hline
657: T--OH$_2$ (\AA) & 2.412 & 2.350 \\
658: \hline
659: B. E. (kJ/mol) & 4.87 & 7.82 \\
660: \hline
661: \end{tabular}
662: \end{center}
663: \end{table}
664:
665: These results are in qualitative agreement with the experimental findings that
666: Sn- and Ti-BEA are hydrophobic Lewis acid catalysts. However, they may also
667: appear surprising since adsorption of one water molecule has been reported on
668: Ti-zeolites with low but exothermic interaction energies.\cite{fois}
669: This difference may be
670: due to the nature of the zeolite framework (beta versus offretite) or to the
671: type of calculations (cluster versus periodic).
672: It must be recalled that interaction energies calculated with DFT based
673: methods do not include van der Waals attractive contributions. In recent work,
674: these dispersion terms have been added empirically\cite{wu}
675: or using
676: adequate correlation functionals.\cite{dion}
677: It is easy to
678: give an a posteriori estimate of the van der Waals stabilization of water
679: bound to the Sn or Ti sites in BEA, using an empirical correction. Using our
680: optimized Sn and Ti structures, the van der Waals stabilization energy of the
681: bound water molecule has been calculated using the universal force
682: field.\cite{rappe}
683: The following energies have been found: -3.3kcal/mol for Sn-BEA
684: and -2.4 kcal/mol for Ti-BEA. Since these empirical van der Waals terms are
685: additive, one can infer that a water dimer would form a very low exothermic
686: complex with the Sn-BEA model, but would still be non bonding with the Ti
687: model.
688: Hence, these results show that
689: Ti--BEA is more hydrophobic than Sn--BEA. This confirms the earlier
690: experimental findings.\cite{sn_hydro}$^{(a)}$
691:
692: \subsection{Reactivity of Sn--BEA and Ti--BEA towards cyclohexanone
693: and H$_2$O$_2$}
694: We have applied the HSAB principle to understand the reactivity of Sn-- and
695: Ti--BEA towards cyclohexanone and H$_2$O$_2$. Pearson formulated the
696: concept of HSAB principle for understanding the reactivity of
697: chemical systems, and their interactions.\cite{pearson}
698: This gave a new insight in
699: interpreting the reactivity of chemical systems on the
700: basis of their HOMO and LUMO energies.\cite{klopman}
701: The systems can be categorized as
702: soft acid (SA) with low lying LUMO, soft base (SB) with high lying HOMO,
703: hard acid (HA) with high lying LUMO and hard base (HB) with low lying HOMO.
704: It has been well established that the interactions between SA--SB are
705: covalent, HA--HB are ionic and SA--HB or HA--SB are mostly weak electrostatic
706: and form Lewis adducts.
707:
708: \begin{table}
709: \begin{center}
710: \caption{HOMO and LUMO energies of cyclohexanone and H$_2$O$_2$}
711: \begin{tabular}{|l|l|l|}
712: \hline
713: & HOMO & LUMO \\
714: \hline
715: Cyclohexanone & -5.066 & -1.333 \\
716: \hline
717: H$_2$O$_2$ & -5.730 & -1.546 \\
718: \hline
719: \end{tabular}
720: \end{center}
721: \end{table}
722:
723: In earlier experimental studies it has been proposed that the Sn site in BEA
724: polarizes the carbonyl oxygen atom of cyclohexanone
725: and forms a Lewis adduct.\cite{corma_nature}$^,$\cite{renz_cej}
726: From above, we
727: have shown that Sn-BEA behaves as a SA, compared with Ti-BEA.
728: Table 6
729: presents the HOMO and LUMO energies of cyclohexanone and H$_2$O$_2$. We
730: observe that the HOMO energy of cyclohexanone is $\sim$60 kJ/mol
731: lower in energy than the HOMO of H$_2$O$_2$. Hence, cyclohexanone and
732: H$_2$O$_2$ are HB and SB, respectively. From the HSAB prnciple we infer that
733: the Sn--BEA interacts with the cyclohexanone molecule to form a SA--HB
734: complex or a Lewis adduct.
735:
736: \section{Conclusions}
737: The present theoretical investigation reveals the differences between the
738: Sn--BEA and Ti--BEA based on their structural, Lewis acidic and hydrophobic
739: properties. Our analysis shows that the Sn and Ti atoms may occupy T2 and/or
740: T1 crystallographic positions in BEA. Although the first
741: coordination shell of Sn is larger than Ti, the second coordination shell
742: in both model zeolites is similar. This explains the relaxation of the local
743: environment of the substituted site.
744: The structural data on Sn--BEA and Ti--BEA
745: presented in this work are in good agreement with the earlier experimental
746: studies. The cohesive energy results demonstrate that the incorporation
747: of Ti is more favorable than Sn in BEA. Nevertheless, we
748: show that Sn--BEA is more Lewis acidic than Ti, and hence proves to be a
749: more efficient catalyst for the oxidation reactions than Ti--BEA. One of
750: the important aspects concerning the activity and selectivty of the zeolite
751: which we have addressed in the present work, is the water resistant
752: property of the Sn--BEA and Ti--BEA. A stable
753: interaction of H$_2$O with
754: the active sites of Sn--BEA and Ti--BEA is more favorable with Sn--BEA. Hence,
755: the hydrophobic property of the Sn--BEA and Ti--BEA
756: zeolites is predicted, as well as their comparison.
757: This clearly justifies the water resistant Lewis acidic sites in
758: Sn--BEA. We also extend our analysis to explain that the interaction of the
759: water molecule with the Ti and the Sn sites in BEA is a mere
760: physisorption rather than chemisorption.
761: Furthermore, we use the HSAB principle to interpret the
762: formation of Lewis adduct between the Sn site and the cyclohexanone.
763:
764: The present work gives thus an insight into the microscopic properties
765: of the active sites in Sn--BEA and Ti--BEA and the differences between them,
766: which would have been otherwise difficult to
767: understand through experimental methods.
768:
769: \section{Acknowledgements}
770: We partially acknowledge the Indo-French Centre for Promotion of Advanced Research (IFCPAR) for providing the computational facilities. \\
771: $^{\dagger}$University of Pune.\\$^{\ddagger}$Ecole de Chimie.\\
772: $^{\S}$National Chemical Laboratory.\\
773: $^*$Corresponding Author
774:
775: \begin{thebibliography}{99}
776: \bibitem{bea1}(a)Corma, A.; Esteve, P.;
777: Martinez, A. {\it J. Catal.} {\bf 1996}, {\it 161}, 11.
778: \bibitem{bea2} Bellussi, G.; Pazzuconi, G.; Perego, C.; Girotti, G.;
779: Terzoni, G. {\it J. Catal.} {\bf 1995}, {\it 157}, 227.
780: \bibitem{bea3} Creyghton, E. J.; Ganeshie, S. D.;
781: Downing, R. S.; van Bekkum, H. {\it J. Mol. Catal.} {\bf 1997}, {\it 115},
782: 457
783: \bibitem{bea_bvo}
784: (a) Jansen, J. C.; Creyghton, E. J.; Njo, S. L.; van Koningsveld, H., van Bekkum
785: H. {\it Catal. Today.} {\bf 1997}, {\it 38}, 205.
786: (b) Kunkeler, P. J.;
787: Zuurdeeg, B. J.; van der Waal, J. C.; van Bokhoven, J. A.; Koningsberger, D.
788: C.; van Bekkum, H. {\it J. Catal.}, {\bf 1998}, {\it 180}, 234.
789: \bibitem{appl_bea}
790: (a)Wadlinger, R. L.; Kerr, G. T.; Rosinski, E. J. US Pat. {\bf 1967}, 3308069.
791: (b)Tuan, V.A.; Li, S.; Noble, R. D.; Falconer, J. L. {\it Environmental
792: Sci. Technol.} {\bf 2003}, {\it 37}, 4007.
793: \bibitem{t_bea}(a)Sen, T.; Chatterjec, M.; Sivasanker, S.{\bf 1995}, 207.
794: (b)de M{\'e}norval, L. C.; Buckermann, W.; Figueras, F.; Fajula, F.
795: {\it J. Phys. Chem.}{\bf 1996}, {\it 100}, 465
796: (c)Juttu, G. C.; Lobo, R. F. {\it Catal. Lett.} {\bf 1999},
797: {\it 62}, 99.
798: (d)Dimitrova, R.; Neinska, Y.; Mih{\'a}lyi, M.; Pal-Borb{\'e}ly,
799: G.; Spassova, M. {\it Appl. Catal. A:} {\bf 2004}, {\it 266}, 123.
800: (e)P{\'e}rez-Ramirez, J.; Groen, J. C.; Br{\"u}ckner, A.; Kumar, M. S.;
801: Bentrup, U.; Debbagh, M. N.; Villaescusa, L. A.
802: {\it J. Catal.}{\bf 2005}, {\it 232}, 318.
803: \bibitem{ti_bea}(a)Blasco, T.; Camblor, M. A.; Corma, A.; P{\'e}rez-Pariente
804: {\it J. Am. Chem. Soc.}{\bf 1993}, {\it 115}, 11806
805: (b) van der Waal,
806: J. C.; van Bekkum, H. {\it J. Mol. Catal.} {\bf 1997}, {\it 124}, 137.
807: (c)Carati, A.; Flego, C.; Massara, P.; Millini, R.; Carluccio, L.; Parker Jr,
808: W. O.; Bellussi, G.{\it Microporous and Mesoporous Mater.}{\bf 1999},{\it 30},
809: 137.
810: \bibitem{vansanten}van Santen, R. A.; Kramer, G. J. {Chem. Rev.} {\bf 1995},
811: {\it 95}, 637.
812: \bibitem{ti_bea_hydro} Blasco, T.;Camblor, M. A.; Corma, A.; Esteve, P.;
813: Guil, J. M.; Martinez, A.; Perdig{\'o}n-Mel{\'o}n, J. A.; Valencia, S.
814: {\it J. Phys. Chem. B} {\bf 1998}, {\it 102}, 75.
815: \bibitem{ti_epo}(a)Saxton R.J.;
816: Zajacek J.G.; Crocco G.L.{\it Zeolites}, {\bf 1996},
817: {\it 17}, 315
818: (b)Corma, A.; Domine, M. E.; Gaona, J. A.; Navarro, M. T.; Rey, F.; Valencia,
819: S. {\it Stud. Surf. Sci. Catal.}{\bf 2001}, {\it 135}, 1812.
820: \bibitem{ts-1}Taramasso, M.; Perego, G.; Notari, B. US Patent {\bf 1983},
821: 4,410,501, (b)Reddy, J. S.; Sivasanker, S. {\it Catal. Lett.} {\bf 1991},
822: {\it 11}, 241 (c)Huybrechts, D. R. C.; De Bruycker, L.;
823: Jacobs, P. A. {\it Nature}, {\bf 1990} {\it 345} 240.
824: \bibitem{diff_ts}(a) van der Waal, J. C.; Lin, P.;
825: Rigutto, M. S.; van Bekkum, H.
826: {\it Stud. Surf. Sci. Catal.} {\bf 1996}, {\it 105B}, 1093. (b) Corma, A.;
827: Camblor, M. A.; Esteve, P.; Mart{\'i}nez, A.; P{\'e}rez-Pariente, J.
828: {\it J. Catal.} {\bf 1994}, {\it 145}, 151.
829: \bibitem{rama_bea} Mal, N. K.; Ramaswamy, A. V. {\it Chem. Commun.}
830: {\bf 1997}, 425.
831: \bibitem{corma_nature} Corma, A.; Nemeth, L. T.; Renz, M.; Valencia, S.
832: {\it Nature}, {\bf 2001}, {\it 412}, 423
833: \bibitem{renz_cej}Renz, M.; Blasco, T.; Corma, A.; Forn{\'e}s, V.; Jensen, R.;
834: Nemeth, L. {\it Chem. Eur. J.} {\bf 2002}, {\it 8}, 4708
835: \bibitem{corma_mpvo} Corma, A.; Domine, M. E.; Nemeth, L.; Valencia, S.
836: {\it J. Am. Chem. Soc.} {\bf 2002}, {\it 124}, 3194.
837: \bibitem{sever}(a)Sever, R. R.; Root, R. W. {\it J. Phys. Chem. B} {\bf 2003},
838: {\it 107}, 10521. (b)Sever, R. R.; Root, R. W. {\it J. Phys. Chem. B}
839: {\bf 2003}, {\it 107}, 10848.
840: \bibitem{corma_hydro}Corma, A. {\it J. Catal.} {\bf 2003}, {\it 216}, 298.
841: \bibitem{weitkamp}Stelzer, J.; Paulus, M.; Hunger, M.; Weitkamp, J.
842: {\it Micropor. Mesopor. Mater.} {\bf 1998}, {\it 22}, 1.
843: \bibitem{sn_hydro}(a)Corma, A.; Domine, M. E.; Valencia, S. {\it J. Catal.}
844: {\bf 2003}, {\it 215}, 294. (b)Corma, A.; Renz, M. {\it Chem. Commun.}
845: {\bf 2004}, 550. (c)Corma, A.; Forn{\'e}s, V.; Iborra, S.; Mifsud, M.;
846: Renz, M. {\it J. Catal.} {\bf 2004}, {\it 221}, 67.
847: \bibitem{fois}Fois, E.; Gamba, A.; Span{\'o}, E. {\it J. Phys. Chem.}
848: {\bf 2004}, {\it 108}, 154
849: \bibitem{dovesi}Zicovich-Wilson, C. M.; Dovesi, R.{\it J. Phys. Chem. B}
850: {\bf 1998}, {\it 102}, 1411
851: \bibitem{notary}(a) Notary, B. {\it Catal. Today.} {\bf 1993}, {\it 18}, 163.
852: (b) Berger, S.; Bock, W.; Marth, C.; Raguse, B.; Reetz, M. {\it Magn. Reson.
853: Chem.} {\bf 1990}, {\it 28}, 559.
854: \bibitem{sastre}Sastre, G.; Corma, A. {\it Chem. Phys. Lett.} {\bf 1999},
855: {\it 302}, 447.
856: \bibitem{bare}Bare, S. R.; Kelly, S. D.; Sinkler, W.; Low, J. J.;
857: Modica, F. S.; Valencia, S.; Corma, A.; Nemeth, L. T.
858: {\it J. Am. Chem. Soc.} {\bf 2005} {\it 127}, 12924.
859: \bibitem{newsam}Newsam, J.M.; Treacy, M.M.J.; Koetsier, W.T.;
860: de Gruyter, C.B. {\it Proc. R. Soc. Lond. A} {\bf 1988} {\it 420}, 375.
861: \bibitem{shetty}Shetty, S.; Pal, S.; Kanhere, D. G.; Goursot, A.
862: {Chem. Eur. J} {\bf 2006}, {\it 12}, 518.
863: \bibitem{boronat}Boronat, M.; Corma, A.; Renz, M.; Sastre, G.; Viruela, P. M.
864: {\it Chem. Eur. J.}{\bf 2005}, {\it 11}, 6905.
865: \bibitem{ma}
866: (a)Ma, Y.; Foster, A. S.; Nieminen, R. M. {\it J. Chem. Phys.}
867: {\bf 2005} {\it 122}, 144709.
868: (b)Mischler, C.; Horbach, J.; Kob, W.;
869: Binder, K. {\it J. Phys.: Condens. Matter} {\bf 2005}, {\it 17}, 4005.
870: \bibitem{jentys} Jentys, A.; Catlow, C. {\it Catal. Lett.} {\bf 1993},
871: {\it 22}, 251.
872: \bibitem{neurock}
873: Neurok, M.; Manzer, L. E. J. Chem. Soc., Chem. Commun. 1996, 1133.
874: \bibitem{deman}de Man, A. J. M.; Sauer, J. {\it J. Phys. Chem.} {\bf 1996},
875: {\it 100}, 5025.
876: \bibitem{sauer1}Sauer, J. {\it Chem. Rev.} {\bf 1989}, {\it 89}, 199.
877: \bibitem{annick}Valerio, G.; Goursot, A.; Vetrivel, R.; Malkina, O.;
878: Malkin, V.; Salahub, D. R. {\it J. Am. Chem. Soc} {\bf 1998}, {\it 120},
879: 11426.
880: \bibitem{maurin}Maurin, G.; Bell, R. G.; Devautour, S.; Henn, F.;
881: Giuntini, J. C. {\it Phys. Chem. Chem. Phys.} {\bf 2004}, {\it 6}, 182.
882: \bibitem{nicholas} Nicholas, J. B.; Hess, A. C. {\it J. Am. Chem. Soc.}
883: {\bf 1994}, {\it 116}, 5428.
884: \bibitem{rozanska}Rozanska, X.; Demuth, T.; Hutschka, F.; Hafner, J.;
885: van Santen, R. A. {\it J. Phys. Chem. B} {\bf 2002}, {\it 106}, 3248.
886: \bibitem{vanderbilt}Vanderbilt, D. {\it Phys. Rev. B} {\bf 1990}, {\it 41},
887: 7892.
888: \bibitem{pw91}Perdew, J. P.; Wang, Y.{\it Phys. Rev. B} {\bf 1992}, {\it 45},
889: 13244.
890: \bibitem{vasp} (a) Kresse, G.;
891: Hafner, J. {\it Phys. Rev. B}{\bf 1994}, {\bf 49}, 14251.
892: (b) Kresse, G.; Furthmuller. {\it Comput. Mater. Sci.} {\bf 1996},
893: {\it 6}, 15.
894: \bibitem{wu}(a)Wu, Q.; Yang, W. J. {\it J. Chem. Phys.} {\bf 2002}, {\it 116},
895: 515. (b) Zimmerli, U.; Parrinello, M.; Koumoutsakos, P. {\it J. Chem. Phys.}
896: {\bf 2004}, {\it 120}, 2693. Grimme, S. {\it J. Comput. Chem.} {\bf 2004},
897: {\it 25}, 1463.
898: \bibitem{dion} Dion, M.; Rydberg, H.; Schröder, E.;
899: Langreth, D. C.; Lundqvist, B. I. {\it Phys. Rev. Lett.}
900: {\bf 2004}, {\it 92}, 246401.
901: \bibitem{rappe} Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; GoddardIII, W. W.
902: Skiff, M. M. {\it J. Am. Chem. Soc.} {\bf 1992}, {\it 114}, 10024.
903: \bibitem{pearson} Pearson, R. G. {\it J. Am. Chem. Soc.}
904: {\bf 1963}, {\it 85}, 3533.
905: \bibitem{klopman}Klopman, G. {\it J. Am. Chem. Soc.} {\bf 1968}, {\it 90}, 223.
906: \end{thebibliography}
907:
908:
909:
910: \pagebreak
911:
912: \begin{figure}[p]
913: \begin{center}
914: Table of Content\\
915: Sharan Shetty, Dilip G. Kanhere, Annick Goursot, Sourav Pal
916: \end{center}
917: \begin{center}
918: \includegraphics[width=0.5\textwidth]{figure_1_bea.ps}
919: \end{center}
920: \end{figure}
921: \end{document}
922: