Nanosized Clusters of Molybdenum Chlorides

ISSN 0012-5016, Doklady Physical Chemistry, 2009, Vol. 427, Part 2, pp. 150–154. © Pleiades Publishing, Ltd., 2009. Original Russian Text © E.G. Il’in, A.S. Parshakov, A.K. Buryak, D.I. Kochubei, D.V. Drobot, V.I. Nefedov, 2009, published in Doklady Akademii Nauk, 2009, Vol. 427, No. 5, pp. 641–645.

PHYSICAL CHEMISTRY

Nanosized Clusters of Molybdenum Chlorides—Active Sites in Catalytic Acetylene Oligomerization E. G. Il’ina, A. S. Parshakova, A. K. Buryakb, D. I. Kochubeic, D. V. Drobotd, and Academician V. I. Nefedov† Presented by Academician I.I. Moiseev April 2, 2009 Received April 15, 2009

DOI: 10.1134/S0012501609080053 †

Formation of nanosized molybdenum clusters in the reaction of molybdenum pentachloride with acetylene in benzene and toluene was demonstrated for the first time, and their composition and structure was suggested. As is known, highest halides of Group V and VI transition metals, as well as some halide complexes of metals in low oxidation states, are catalysts of acetylene polymerization. In particular, niobium and tantalum pentahalides are efficient catalysts of stereoselective cyclotrimerization of acetylene [1] and monosubstituted acetylenes HC≡CR [2], whereas molybdenum and tungsten halides and their complexes are catalysts of linear polymerization and form mixed products [3]. As regard the mechanism of catalytic reactions, the best studied reactions of molybdenum and tungsten halides with mono- and disubstituted acetylene hydrocarbons are those with propyne, 2-butyne, and diphenylacetylene [4]. It was found that the reaction with a propyne excess yielded paramagnetic compounds åïm(C3H4)n, where m = 3–4 and n ≤ 7. Only the reaction of MoCl5 with a sixfold propyne excess at room temperature yielded a product of definite composition, MoCl4(C3H4)3. According to IR spectroscopy, this product has a system of conjugated nonaromatic double bonds. The major product of the reaction of Moël5 with 2-butyne was found to be MoCl4(C2R2). The IR spectral data for it best fit the formation of a three-membered † Deceased. a

Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences, Leninskii pr. 31, Moscow, 119991 Russia b Frumkin Institute of Physical Chemistry and Electrochemistry, Russian Academy of Sciences, Leninskii pr. 31, Moscow, 119991 Russia c Boreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, pr. Akademika Lavrent’eva 5, Novosibirsk, 630090 Russia d Lomonosov State Academy of Fine Chemical Technology, pr. Vernadskogo 86, Moscow, 117571 Russia

ring rather than a π complex by the ligand. We are not aware of any studies that address the composition and structure of the products of the reaction of molybdenum pentachloride with acetylene. Initial solutions were saturated solutions of Moël5 in benzene and toluene bubbled with purified and dried acetylene. As acetylene passed through the solution, its color changed from dark green to black, the solution warmed up and transformed into a black gel, and after a time, solution temperature decreased to ambient. The reaction was accompanied by release of HCl. After 4– 6 h, the gas bubbling was terminated, and a black gellike precipitate was deposited, which was filtered off in an inert atmosphere, washed with a dry solvent, and dried under vacuum. According to 1H NMR, the filtrate was the pure solvent; i.e., the molybdenum compounds and the products of acetylene oligomerization were completely coprecipitated. The resulting substances were fine air-unstable black powders insoluble in water and nonpolar organic solvents and partially soluble in DMSO and ëHCl3 on heating. Depending on the reaction conditions, the composition of reaction products somewhat varied (Table 1), but the Cl : Mo ratio was close to two and the H : C ratio, to unity. The samples are stable in an inert atmosphere and under high vacuum at temperatures up to 350°C and, according to X-ray diffraction data, represent an unknown phase. The presence of a broad maximum in the X-ray diffraction pattern at small angles suggested an amorphous or nanocrystalline structure. The EPR spectrum shows a strong broad signal typical of exchange coupling of metal atoms, which can be evidence of the formation of molybdenum clusters. The IR spectra of the products lack a band of the coordinated triple bond ë≡ë (2040–2050 cm–1). The absorption bands at 2923 and 1289 cm–1 were assigned to C–H vibrations and the band at 1400 cm–1, to vibrations of conjugated double bonds C=C. The bands at 344 and 268 cm–1 were assigned to molybdenum–chlo-

150

NANOSIZED CLUSTERS OF MOLYBDENUM CHLORIDES—ACTIVE SITES

151

Table 1. Chemical analysis data (%) for the products of the reaction of MoCl5 with acetylene Solvent

C

H

Cl

Mo

Empirical formula

Benzene Toluene

65.94 66.23

5.38 5.24

11.70 11.87

16.98 16.66

MoCl1.86C31.02H30.10 MoCl1.93C31.75H29.88

rine and molybdenum–molybdenum vibrations, respectively. The reaction product was studied by matrix assisted laser desorption/ionization time-of-flight (MALDITOF-MS) mass spectrometry in the negative-ion mode on a Bruker Ultraflex mass spectrometer. The accuracy of mass measurements was 0.002%. Figure 1 shows the mass spectrum of MoCl1.9 ± 0.1(C30 ± 1H30 ± 1). The strongest lines in the mass spectrum of MoCl1.9 ± 0.1(C30 ± 1H30 ± 1) (Fig. 1) corresponded to negative ions with the ratio m/z = 2157.730 and 2270.690. Symmetric isotope composition of the lines was evidence of a symmetric structure of the corresponding fragments. Taking into account the observed Cl : Mo ratio close to two and the ability of lower molybdenum chlorides to form clusters, we can assume that the inorganic part of the major observed fragment is a (MoCl2)n cluster. The strongest line in the mass spectrum corresponded to the anion with m/z = 2270.690. According to the MoCl2 mass of 167, the number of Mo atoms in the cluster can be 12 or 13. In the former case, the organic part of the anionic fragment accounts

for 267 mass units and in the latter case, for 171 mass units. Thus, the formulas [Mo12Cl24(C20H21)] or [Mo13Cl24(C13H14)], respectively, can be assigned to these clusters. The second variant with Mo13 seems to be more probable since the theoretical composition of the fragment [Mo13Cl24(C13H14)]– is better consistent with the experimental composition (Fig. 2). Analogously, for the anionic fragment with m/z = 2157.73, which is more than twice lower in intensity, the number of molybdenum atoms in the inorganic core of the cluster anion can also be taken as 13 or 12. As for the composition of the hydrocarbon part of the fragment, taking into account the ability of Mo and W compounds to catalyze alkene metathesis reactions through the intermediate formation of metal–carbyne or metal–carbene fragments [5], we assume that the coordination sphere of Mo atoms of the catalytically active cluster can contain both acetylene molecules and carbyne åÓ≡ëç, carbene åÓ=ëç2, or vinylidene åÓ=ë=ëç2 fragments, which is, in our opinion, more probable.

J × 108 2271 8

6

4

865

2 432 576

0 0

718

1006

500

1000

1500

Fig. 1. Negative ion MALDI-TOF mass spectrum. DOKLADY PHYSICAL CHEMISTRY

Vol. 427

Part 2

2009

2000

2500

m/z

152

IL’IN et al. J × 108 (‡) 8

6

4

2

0 2220 J 100

2240 2260 [Mo13Cl24C13H14]–

2280

2300 J 100

(b)

90

2320

80

70 60

70 60

50

50

40

40

30 20

30 20

10 0

10 0

2300

(c)

90

80

2250

m/z [Mo12Cl24C20H21]–

m/z

2250

2300 m/z

Fig. 2. (a) Experimental and (b, c) theoretical isotope compositions of the fragment with m/z = 2270.690.

The cluster structure of molybdenum complexes was confirmed by EXAFS spectroscopy. The molybdenum K-edge absorption spectra were recorded in the routine transmission mode [6] at the EXAFS station of the VEPP-3 storage ring (electron beam energy, 2GeV; beam current, 80 mA) at the Siberian Synchrotron Radiation Center (Novosibirsk). The oscillating part of the EXAFS spectra (χ(k)) was analyzed as k3χ(k) in the wavenumber range k = 2.5–15.0 Å–1. The extraction of the oscillating part of the absorption coefficient and simulation of spectra for determining structural data were performed with the VIPER program [7]. Quantum-chemical calculations and curve fitting were performed with the FEFF7 code [8]. In simulation, in order not to exceed the number of parameters, the same ionization potential E0 was taken for all coordination spheres. For the same purpose, the Debye–Waller factors for the Mo–C and Mo–Cl distances and two Mo– Mo distances were equated. The true coordination

number for the Mo–Cl distance could be underestimated because of the existence of several close distances, which were modeled by one coordination sphere. Farther coordination spheres were not modeled. Figure 3 shows the EXAFS spectrum of the reaction products. A major contribution to the experimental spectrum (Fig. 3) is made by two distances molybdenum–carbon and molybdenum–chlorine. Two distinct Mo–Mo distances are observed for the cluster core (Table 2). If the cluster contains 13 molybdenum atoms, the central atom is at a distance of 2.58 Å from the other 12 atoms and the peripheral atoms are at a distance of 2.83 Å from each other. The coordination sphere of the peripheral molybdenum atoms contains each a carbon atom (CN is about unity) at a distance of 1.71 Å, which is evidence of the formation of a multiple bond by this atom and supports the probable formation of carbyne frag-

DOKLADY PHYSICAL CHEMISTRY

Vol. 427

Part 2

2009

NANOSIZED CLUSTERS OF MOLYBDENUM CHLORIDES—ACTIVE SITES J 10

153

Mo–Cl

8 Mo–C

6

Mo–Mo

4 2 0

2

4

6

8 R, Å

Fig. 3. EXAFS spectrum of MoCl1.9 ± 0.1(C30 ± 1H30 ± 1).

ments åÓ≡ëç, and chlorine atoms at a distance of 2.37 Å (CN, 1.32). The observed number of chlorine atoms per molybdenum atom is lower than the number following from the chemical analysis of the product, which can be explained assuming that the Mo–Cl distances differ within 0.1 Å and their signals interfere and cancel each other out to some extent. As for the possible configuration of 13-atom molybdenum clusters, there are theoretical calculations of the åÓ13 cluster with an icosahedral or lower symmetry biplanar configuration [9, 10]. The two-plane 13-atom Mo(+3) cluster with an oxygen environment of åÓ13 is known, which was characterized by X-ray crystallography in a single crystal of Pr4Mo9O18 [11]. The Mo–Mo distances in this 13-atom cluster vary from 2.52 to 2.96 Å, i.e., close to the observed bond lengths (Table 2). In symmetric clusters with a closed shell, the most frequently encountered magic number of metal atoms is 13 [12], for example, Au13(diphos)6(NO3)4 has an icosahedral structure [13] and [Rh13(CO)24H5 – x]x–, a cuboctahedral structure [14]. Although the [ëu12S8]4– cubooctahedron with a missing central atom is known [15], the variant with Mo13 seems to be more probable. As for the structure of the supposed catalytically active 13-atom cluster of molybdenum chloride, the symmet-

ric line shape for the fragment suggests a cuboctahedron or icosahedron (Fig. 4). According to Nefedov and Kustov who considered possible coordination polyhedra of nanosized clusters, or the structure of [Mo13Cl24(C13H14)]– – [Mo12Cl24(C20H21)] is completely explained on the basis of both an icosahedron and a cuboctahedron. For the most frequently encountered cuboctahedron, if all 12 peripheral molybdenum atoms of the åÓ13 cluster are equivalent and each of them is bound to one carbon atom (the CN of Mo with respect to carbon is 1), four geometric isomers are possible depending on the arrangement of 24 chloride ions. The chloride ions can be located over 24 edges, acting as bridges (the CN of Mo with respect chlorine is 4). Another possible arrangement can be the variant found in [Rh13(CO)24H5 – x]x– for 24 CO groups [24]: 12 bridging chloride ions located over edges and 12 terminal chloride ions each bonded to one peripheral molybdenum atom (the CN of Mo with respect to chloride is 3); two such variants are possible where the chlorine atoms occupy vertical or horizontal sides of squares. The fourth cuboctahedron isomer can be a cluster in which all 24 chloride ions are terminal and each peripheral molybdenum atom is bonded to two chloride ions. At the same time, in a regular cuboctahedral 13atom cluster, the Mo–Mo distances are equal, whereas in an icosahedral cluster, they can be different. An

Table 2. Results of studying MoCl1.9 ± 0.1(C30 ± 1H30 ± 1) by EXAFS spectroscopy Coordination sphere

Distance R, Å

Coordination number

Debye–Waller factor, ×104 Å2

E0 , eV

R, %

Mo–C Mo–Cl Mo–Mo Mo–Mo

1.71 2.37 2.58 2.83

1.14 1.32 2.74 1.49

24.1 24.1 12.1 12.1

1.29

30.2


Vol. 427

Part 2

2009

154

IL’IN et al. (a)

(b)

Cuboctahedron

Hexagonal cuboctahedron

(c)

(d)

Icosahedron

Biplanar structure

Fig. 4. Probable structures of the metal core of the 13-atom molybdenum cluster: (a, b, c) hypothetical structures, and (d) the observed structure for the Mo13 cluster with an oxygen environment [11].

icosahedron is less chemically stable, i.e., more catalytically active. Quantum-chemical calculations of possible isomers of cuboctahedral, icosahedral, and biplanar configurations are currently in progress. ACKNOWLEDGMENTS We are grateful to Prof. O.N. Temkin for valuable comments. This work was supported by the Presidium of the RAS (program “Development of Methods of Synthesis of Chemical Compounds and New Materials,” subprogram “Design of New Construction and Functional Materials on the Basis of Nanotechnologies”). REFERENCES 1. Dandliker, G., Helv. Chim. Acta, 1969, vol. 52, p. 1482. 2. Masuda, T., Deng, Y.X., and Higashimura, T., Bull. Chem. Soc. Jpn., 1983, vol. 56, p. 2798. 3. Masuda, T., Hasegawa, K., and Higashimura, T., Macromolecules, 1974, vol. 7, p. 728. 4. Greco, A., Pirinoli, F., and Dall’asta, G., J. Organomet. Chem., 1973, vol. 60, p. 115.

5. Temkin, O.N., Soros. Obraz. Zh., 2001, vol. 7, no. 6, p. 728. 6. Kochubei, D.I., EXAFS-spektroskopiya katalizatorov (EXAFS Spectroscopy of Catalysts), Novosibirsk: Nauka, 1992. 7. Klementev, K.V., J. Phys. D: Appl. Phys., 2001, vol. 34, p. 209. 8. Rehr, J.J. and Ankudinov, A.L., Radiat. Phys. Chem., 2004, vol. 70, p. 453. 9. Chang, C.M. and Chou, M.Y., Phys. Rev. Lett., 2004, vol. 93, no. 13, p. 133401. 10. Yang, J., Deng, K., Xiao, Ch., and Wang, K., Phys. Lett., 1996, vol. 54, no. 17, p. 11907. 11. Tortelier, J. and Gougeon, P., Inorg. Chem., 1998, vol. 37, p. 6229. 12. Sakurai, M., Watanabe, K., Sumiyama, K., and Suzuki, K., J. Chem. Phys., 1999, vol. 111, no. 1, p. 235. 13. Van der Velden, J.W.A., Volenbrock, F.A., Bour, J.J., et al., Rec. Trav. Chim., 1981, vol. 100, no. 4, p. 148. 14. Albano, V.G., Geriotti, A., Chini, P., et al., J. Chem. Soc., 1975, no. 20, p. 859. 15. Betz, P., Krebs, B., and Henkel, G., Angew. Chem. Int. Ed., 1984, vol. 23, p. 311.


Vol. 427

Part 2

2009