вход по аккаунту


Mass Spectrometric Investigations of the Dissociation of [C@W6Cl17] Ions in the Gas Phase.

код для вставкиСкачать
Cluster Compounds
Mass Spectrometric Investigations of the
Dissociation of [C@W6Cl17] Ions in the Gas
Ralf Burgert, Katharina Koch, Hansgeorg Schnckel,*
Martina Weisser, Hans-Jrgen Meyer,* and
Hans Georg von Schnering
Dedicated to Professor Martin Jansen
on the occasion of his 60th birthday
The first steps into the field of metal-cluster compounds with
endohedral hetero atoms were the planned synthesis and
structural characterization of [HNb6I11] by Simon[1] as well as
the identification of [Ru6C(CO)17].[2–4] Since then, the [M6X12]
clusters of the transition elements and their substituted
variants have been discussed in particular detail.[5–7] In
recently published experiments on the synthesis of octahedral
tungsten halogenide clusters, the carbon-centered compounds
[C@W6Cl16] (1) and [C@W6Cl18] (2), formed as by-products,
have been detected for the first time; their formation can be
put down to the fact that the commercially available educt
WCl4, besides containing WCl5 und WOCl4, was also soiled
with 1,2,4,5 tetrachlorobenzene which provided a source of
carbon atoms.[8] Based on this knowledge, the synthesis of 1
[*] R. Burgert, K. Koch, Prof. Dr. H. Schnckel
Institut fr Anorganische Chemie
Universitt Karlsruhe (TH)
Engesserstrasse 15, 76128 Karlsruhe (Germany)
Fax: (+ 49) 721-608-4854
M. Weisser, Prof. Dr. H.-J. Meyer
Institut fr Anorganische Chemie
Universitt Tbingen
Auf der Morgenstelle 18, 72076 Tbingen (Germany)
Fax: (+ 49) 7071-29-5702
Prof. Dr. H. G. von Schnering
Max-Planck-Institut fr Festkrperforschung
Heisenbergstrasse 1, 70569 Stuttgart (Germany)
[**] This publication was financially supported by the Deutsche
Forschungsgemeinschaft and the Fonds der Chemischen Industrie.
Part of the results was presented on March 29, 2004 at the EFTMS
2004 (Seventh European Workshop On Fourier Transform Ion
Cyclotron Resonance Mass Spectrometry) in Konstanz.
Supporting information for this article is available on the WWW
under or from the author.
Angew. Chem. Int. Ed. 2005, 44, 265 –269
DOI: 10.1002/anie.200460972
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
und 2 was optimized and the bonding properties were
described in detail based on the crystal-structure examination.[8, 9] Now we have developed an improved synthesis, in
which the lithium as well as the calcium salts of the ion
[C@W6Cl18]2 (22) are generated.[10, 11] Independent from our
findings, the generation of the 22 during the synthesis of 1
from WCl6 and CCl4 with elemental bismuth as the reducing
agent was described recently.[12] Cyclovoltametric experiments in addition to structural data helped to examine the
stability of the various ions [C@W6Cl18]n (2n n = 1, 2, 3) and
formed the basis for the discussion of the bonding properties
of these clusters.[12]
Herein we describe for the first time investigations on a
structurally characterized endohedral cluster species by FTICR-MS (ion cyclotron resonance mass spectrometry). In
these experiments 22 and the cluster ion [C@W6Cl17] (3 ;
which is formed after loss of Cl) are transferred into the gas
phase entirely, without any initial decomposition. Thus, the
decomposition of 3 can be observed step by step in different
isolation and collisionally activated dissociation (CAD)
First, a polycrystalline [Li2(22)] sample was analyzed by
laser desorption ionization (LDI). A multitude of tungsten
chloride clusters (many of them C-bearing) could be identified,
however none of them had the composition of 22 or 3 .[14]
Electrospray ionization (ESI) proved to be more efficient.[15–17] In accordance with earlier experiments[12] even
sensitive compounds, such as heavy cluster ions,[18] can be
transferred into the gas phase without disintegration and
subsequently investigated. Figure 1 a shows the anion spectrum of a freshly prepared green solution of [Li2(22)] in
methanol. Along with small traces of [C@W6Cl18] (2) only
the isotope pattern of 3 can be detected. The ESI mass
spectrum of the same solution about 30 min later, shows new
signals (Figure 1 b) which can be assigned to the ion 22 as well
as to the dissociation products 4 and 5 described below.[19]
To examine the stepwise fragmentation of this cluster in
the gas phase, several isolation and dissociation experiments
were performed. First, the most intensive signal (m/z 1718)
out of the great number of isotopomers of 3 could be isolated
(Figure 2). The fragmentation of 3 takes place after collisionally activated dissociation by radio frequency (RF)
irradiation and admission of argon (SORI-CAD).[20] Three
new signal groups are observed which can be assigned to the
compounds [C@W5Cl13] (4), [C@W4Cl9] (5), and [C@
W3Cl6] (6). To confirm experimentally that the dissociation
of 3 proceeds by a consecutive mechanism in which two
WCl4 and one WCl3 fragment are sequentially split from the
parent cluster, each of the species 4 , 5 , and 6 was isolated
once more and dissociated by collisional activation. The
resulting fragments proved to be both 5 and 6 (from 4) and
6 (from 5 ; Figure 2). During the fragmentation of 6 only
the appearance of Cl is observed. Based on these experimental findings the reaction channel shown in Equation (1) is
½C@W6 Cl17 ð3 Þ ƒƒƒ! ½C@W5 Cl13 ð4 Þ ƒƒƒ!
½C@W4 Cl9 ð5 Þ ƒƒƒ! ½C@W3 Cl6 ð6 Þ
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 1. a) Mass spectrum (by electrospray ionization, ESI) of a
freshly prepared solution of [Li2(22)] in methanol. b) Mass spectrum
(by electrospray ionization, ESI) of an aged solution of [Li2(22)] in
methanol (the experiment was performed under identical conditions to
those used to obtain Figure 1 a, in particular the skimmer voltages
(20 V) were similar).
Figure 2. SORI-CAD experiments: The products after the collisionally
activated dissociation of 3 , 4 , and 5 . The marked signals (*) were
previously isolated isotope selectively.
These fragmentation processes are of note as both WCl4
and WCl3 are used as starting compounds for the synthesis of
However, determination of the dissociation pathway of 6
by mass spectrometric methods is not possible, because no
further fragmentation of 6 into smaller ionic [C@WxCly]
Angew. Chem. Int. Ed. 2005, 44, 265 –269
units is detected. Moreover, no conclusions can be made
regarding the stabilizing function of carbon or about the
structures of the fragments involved in the dissociation
process. In experiments with our apparatus configuration
the energy amounts DE1, DE2, DE3, which are necessary for
the dissociation, can not be determined by CAD.
To find answers to these questions concerning the
structure of the intermediates as well as the energy balance
during the dissociation processes, we performed density
functional theory (DFT) calculations.[22, 23] To check the
reliability of the energetic aspect of these calculations we
examined the isodesmic disproportionation reaction given in
Equation (2).
2 WCl4ðgÞ ! WCl6ðgÞ þ WCl2ðgÞ
Using highly correlated CCSD(T)/TZVPP methods[24] a
DE value of 337 kJ mol1 is obtained which is close to the
values calculated by the DFT methods mentioned above
(334 kJ mol1).[25, 26] Therefore, the dissociation energies determined by DFT methods in equation (1) should be an
adequate basis for the discussion of the experimentally
found results: DE1 = 196 kJ mol1, DE2 = 63 kJ mol1, and
DE3 = 256 kJ mol1.
In Figure 3 a–d the calculated structures of 3 , 4 , 5 , and
6 are displayed. Detailed information and interatomic
distances of all the ions calculated by this method can be
found in the Supporting Information.
The mass spectrometric methods used here cannot answer
the question of the fragmentation of [C@W3Cl6] (6), since
no further fragmentation to smaller [C@WxCly] units is
observed. A plausible reaction channel could be a simple
electron loss from 6 , since for [C@W3Cl6] an electron affinity
of 332 kJ mol1 was calculated. However, the exclusive
detection of Cl ions from the dissociation 6 suggests a
different reaction channel. Under preparative conditions, as
in the synthesis of 22 (see above) a reduction could be
possible where, apart from Cl ions, only the neutral starting
compounds emerge [Eq. (3)]. For this hypothetical mode of
dissociation we calculated an energy amount of
+ 306 kJ mol1.[27]
½C@W3 Cl6 ! Cl þ 1=4 WCl4 þ CCl4 þ 11=4 WðsÞ
The hypothesis of this reaction path is supported by
approximate calculations, for alternative dissociation pathways giving products such as [C@W2], [C@W2Cl2], and
[ClW3], they predict a significantly higher energy requirement. Thus, the reaction in Equation (3) seems to be a
plausible suggestion, especially after consideration of retrosynthesis. The results obtained experimentally and by quantum chemical methods for the dissociation of 3 via 6
demonstrate that the smallest carbon-containing tungsten
cluster [C@W3Cl6] detected in experiments is possibly one of
the first clusters generated in syntheses that start from solid
tungsten, WCl4, and CCl4 [Eq. (3)].
The results reported herein raise the principle questions,
to what extent the insertion of a carbon atom in these clusters
is reflected in the energy balance and how strongly a carbon
Angew. Chem. Int. Ed. 2005, 44, 265 –269
Figure 3. Calculated DFT structures of the fragment ions 3 (a), 4 (b),
5 (c), 6 (d), and of the hypothetical [W4Cl8] molecule 8 (e).
atom is bound? Since with the neutral cluster [C@W4Cl8] (7),
which is derived from [C@W4Cl9] (5), our experiments
provide a good model compound with which to try and
answer these questions, we investigated the following DFTsupported idea: the removal of a carbon atom from 7 to
generate the hypothetical [W4Cl8] molecule (8; Figure 3 e)
requires 699 kJ mol1.[28] This calculated energy gain for the
insertion of a carbon atom into a molecular tungsten-atom
framework corresponds to the experimentally determined
energy gain which results from the formation of solid tungsten
carbide WC from tungsten metal and carbon
(752 kJ mol1).[29] Thus, carbon-centered metal clusters such
as 1 are suitable molecular models for the topological and
energetic situation in the appropriate solid-state compounds.[30]
In addition to the results for the bonding energy of carbon
in the endohedral tungsten clusters reported herein, the
dissociation experiments in the gas phase of 3 , 4 , and 5 as
well as the accompanying quantum-chemical calculations
make it seem plausible that the recently presented synthesis
of 2 and 22 also proceeds over the observed intermediates, in
particular because the educts WCl4/WCl3 of the synthesis
were released. This conclusion is also based on the fact that
the intermediate products 4 and 5 can be detected in a
methanol solution of [Li2(22)] after ageing.
Since, to date, only 2 and 2n (n = 1,2,3) can be obtained in
the crystalline state, it is a challenge for future preparative
investigations to isolate and characterize the cluster frag-
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ments presented herein. Furthermore, through the comparison of the experimental data with quantum-chemical calculations (see Supporting Information) the accuracy of the
simple bonding descriptions[31, 32] of the mostly open-shelled
structures presented could be verified.
Experimental Section
The experiments presented here were performed on an Ionspec
ULTIMA FT-ICR mass spectrometer with a 7.0 T magnet, equipped
with both a MALDI and an electrospray (Analytica) ion source.
For the electrospray ionization a solution of [Li2(22)] in
methanol (approximately 10 mg mL1) was sprayed slowly
(2 mL min1) through a stainless steel capillary (with an internal
diameter of 100 mm) into an inert gas atmosphere (nitrogen). The ions
lose their solvent and arrive over several intermediate stages in an
ultrahigh vacuum area, which is generated with a differential pump
system (consisting of a rotary vane pump, a turbo molecular pump,
and two cryo pumps). The ions are transferred into a hexapole
through a glass capillary (with voltages of 3900 V, and 140 V on their
nickel coated ends) and a skimmer (20 V), wherein they are stored
for 1500 ms. Within the hexapole the pressure is approximately
106 mtorr. The ions finally arrive at the ICR cell (at 1010 mtorr)
through a quadrupole.
For the SORI-CAD experiments ions were selected by SWIFTisolation (SWIFT = stored wave inverse Fourier transformation) and
accelerated by a RF-pulse in the presence of a collision gas (argon) at
108 mtorr.[33] The excitation frequency is about 200 Hz higher than
the cyclotron frequency, with a amplitude of 5–15 V and a duration of
1000 ms.
Received: June 15, 2004
Revised: July 23, 2004
Keywords: carbon · cluster compounds · density functional
calculations · mass spectrometry · tungsten
[1] A. Simon, Z. Anorg. Allg. Chem. 1967, 355, 295.
[2] B. F. G. Johnson, R. D. Johnston, J. Lewis, Chem. Commun.
1967, 1057.
[3] A. Sirigu, M. Bianchi, E. Benedetti, J. Chem. Soc. D 1969, 596.
[4] “Clusters with Interstitial Atoms from the p-Block: How do
Wades Rules Handle Them ?”: C. E. Housecroft in Structural
and Electronic Paradigms in Cluster Chemistry (Ed.: D. M. P.
Mingos), Springer, Heidelberg, 1997, p. 138.
[5] H. Schaefer, H. G. von Schnering, Angew. Chem. 1964, 76, 833.
[6] F. Brian, G. Johnson, C. M. Martin in Metal Clusters in
Chemistry, Vol. II (Ed.: P. Braunstein), Wiley-VCH, Weinheim,
1998, p. 877.
[7] J. Robert, J. Goudsmit, D. H. Farrar in Metal Clusters (Ed.: M.
Moskovits), Wiley, New York, 1986, p. 29.
[8] Y.-Q. Zheng, H. G. von Schnering, J.-H. Chang, Y. Grin, G.
Engelhardt, G. Heckmann, Z. Anorg. Allg. Chem. 2003, 629,
[9] In particular the structures of the cluster units with and without
endohedral C-atoms, as well as the dramatic changes in the
symmetry (from quasi-octahedral to trigonal prismatic) were
described in detail and discussed therein (see ref [8]). In
agreement with the observation, our own DFT calculations
show that [W6Cl18] with octahedral W6-framework is favored
over the trigonal-prismatic arrangement by around 70 kJ mol1.
In addition we found that for the C-centered cluster [C@W6Cl18]
the trigonal-prismatic arrangement is more favored than the
octahedral by around 260 kJ mol1.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
[10] M. Weisser, H.-J. Meyer, unpublished results.
[11] M. Weisser, Diploma thesis, Tbingen, 2003.
[12] E. J. Welch, N. R. M. Crawford, R. G. Bergman, J. R. Long, J.
Am. Chem. Soc. 2003, 125, 11 464.
[13] The stabilizing function of centred carbon atoms in metal
clusters could also be demonstrated recently by reactions at
naked metal clusters with CH4 by FT-ICR-MS: C. P. G. Butcher,
A. Dinca, P. J. Dyson, B. F. G. Johnson, P. R. R. LangridgeSmith, J. S. McIndoe, Angew. Chem. 2003, 115, 5930; Angew.
Chem. Int. Ed. 2003, 42, 5752.
[14] The disadvantage of this kind of ionization is that numerous
fragmentations and side reactions are induced by the UV laser
(337 nm) energy, and thus no more parent ions can be detected.
[15] J. B. Fenn, Angew. Chem. 2003, 115, 3999; Angew. Chem. Int. Ed.
2003, 42, 3871.
[16] The compound is dissolved in a polar solvent and atomized in a
strong electrical field. Depending on the sign of their charge,
ions accumulate on the surface of the solvent droplets. A heated
nitrogen counter current lets the solvent droplets shrink until
they burst through the electrostatic repulsion of the ions. In this
way no “new” ions are produced, on the contrary solvated ions
are released from their solvent.
[17] D. Schroeder, Angew. Chem. 2004, 116, 1351; Angew. Chem. Int.
Ed. 2004, 43, 1329.
[18] K. Weiss, H. Schnckel, Z. Anorg. Allg. Chem. 2003, 629, 1175.
[19] In all the experiments the skimmer potential was kept at a
constant 20 V. The comparison of Figure 1 a and Figure 1 b
shows that fragmenting reactions caused by the difference of
potential between capillary exit and skimmer can be excluded.
These findings can be interpreted that 3 (and Cl) arises during
the dissolving of 22, see Figure 1 a. We explain the presence of
22 in an aged solution by the following equilibrium reaction: at
the beginning the equilibrium is exclusively on the side of 3 and
the Cl ions. By a substantial increase of the Cl concentration
(also experimentally observed) the equilibrium is shifted in favor
of 22. These additionally needed Cl ions are released during
the aging processes, which also yields the fragments 4 and 5 .
This conclusion could be confirmed by the fact that after
addition of Cl (as tetrabutylammonium chloride), 3 was no
longer detected while 22 could still be observed.
[20] SORI-CAD (sustained off-resonance irradiation collision activated dissociation). A small shift in the excitation frequency
compared to the cyclotron frequency ensures that excitation
only occurs with the flank of the RF pulse. The kinetic energy of
the system is raised moderately over 1000 ms to give an energy
distribution of the particles according to Boltzmann statistics. By
collision-induced conversion into internal energy the weakest
bonds in the cluster molecule finally break. J. W. Gauthier, T. R.
Trautman, D. B. Jacobson, Anal. Chim. Acta 1991, 246, 211.
[21] The appearance of 4 and 5 (however not 6) in an aged
methanol of [Li2(22)] solution shows clearly that they represent
intermediate stages in the decomposition of 3 , and probably
also in the assembly. The absence of 6 in this context
demonstrates that the spectrum in Figure 1 b actually shows
the composition of the aged solution, and the decomposition
products 4 and 5 are not the results of fragmenting reactions
caused by differences of potential between glass capillary and
skimmer within the medium pressure range.
[22] a) R. Ahlrichs, M. Br, M. Hser, H. Horn, C. Klmel, Chem.
Phys. Lett. 1989, 162, 165 – 169; b) O. Treutler, R. Ahlrichs, J.
Chem. Phys. 1995, 102, 346 – 354; c) J. C. Slater, Phys. Rev. 1951,
81, 385; d) S. H. Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980,
58, 1200; e) A. D. Becke, Phys. Rev. A 1988, 38, 3098; f) J. P.
Perdew, Phys. Rev. B 1986 33, 8822; g) A. D. Becke, J. Chem.
Phys. 1993, 98, 5648; h) A. Schfer, C. Huber, R. Ahlrichs, J.
Chem. Phys. 1994, 100, 346; h) W: E. D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss, Theor. Chim. Acta 1990, 77,
Angew. Chem. Int. Ed. 2005, 44, 265 –269
123; i) Cl: A. Bergner, M. Dolg, W. Kuechle, H. Stoll, H. Preuss,
Mol. Phys. 1993, 80, 1431.
[23] Although compounds of heavy transition metals can be
described only to a limited extent correctly in this way,
Schimmelpfennig et al. recently reported on DFT calculations
for different tungsten chlorides, which are in agreement with the
experimental data.[26] We therefore oriented ours approach
towards these methods (B3LYP-functional, with quasirelativistic
core potential (ecp) in combination with the TZVPP basic set),
and thus confirmed bond lengths from the crystallographic data
of [C@W6Cl18].
[24] Gaussian 98 (Revision A.7), M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, R. E. Stratmann, J. C. Burant, S.
Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain,
O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A.
Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J.
V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W.
Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle, J. A.
Pople, Gaussian, Inc., Pittsburgh, PA, 2001..
[25] Unfortunately, there exists only imprecise data for the chemical
energy of our experimental reaction: In ref. [29], DHR is quoted
with a value of 166 kJ mol1 at a tolerance of 190 kJ mol1 (!)
and therefore can not be used to judge the quality of the
calculation methods. All energy values calculated herein relate
Angew. Chem. Int. Ed. 2005, 44, 265 –269
to the respective molecular structures at 0 K. We abstained from
a correction with respect to the zero-point vibrational energy,
this contribution for the educts and products is almost identical
so that according to our approximate calculations the energy
difference is only about (5 5) kJ mol1.
B. Schimmelpfennig, U. Wahlgren, O. Gropen, A. Haaland, J.
Chem. Soc. Dalton Trans. 2001, 1616.
For W(s) in Equation (3) we have used the evaporation heat
(850 kJ mol1) from ref. [29], and thus corrected the calculated
reaction energy for Equation (3) with W(g).
In cooperation with H.-J. Himmel a study of the cluster
molecules [W2Cl4], [W4Cl8], and [W6Cl12] is in preparation
where the possibility of multiple bonding is discussed in
comparison to organic compounds with aromatic bonding
M. Binnewies, E. Milke, Thermochemical Data of Elements and
Compounds, Vol. 2, Wiley-VCH, Weinheim, 2002; M. W.
Chase, Jr., C. A. Davies, J. R. Downey, Jr., D. J. Frurip, R. A.
McDonal, A. N. Syverend, JANNAF-Thermodynamical Tables,
3rd ed., American Chemical Society, American Institute of
Physics, US National Bureau of Standards, Midland, MI, 1985.
As demonstrated by our experiments and calculations, the
significant energy gain from the insertion of carbon atoms is
already obtained with the generation of the smallest clusters (e.g.
C@W3 units in 6).
D. M. P. Mingos, Structural And Electronic Paradigms In Cluster
Chemistry, Springer, Heidelberg, 1997.
D. M. P. Mingos, Acc. Chem. Res. 1984, 311.
A. G. Marshall, T. C. L. Wang, T. L. Ricca, J. Am. Chem. Soc.
1985, 107, 7893.
2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Без категории
Размер файла
159 Кб
investigation, mass, gas, dissociation, ions, w6cl17, phase, spectrometry
Пожаловаться на содержимое документа