Synthesis, High-Resolution Millimeter-Wave Spectroscopy, and Ab

May 15, 2012 - ABSTRACT: The millimeter-wave rotational spectrum of an organomercury ... observation of the six most abundant isotopes of mercury, ...
1 downloads 0 Views 1012KB Size
Article pubs.acs.org/JPCA

Synthesis, High-Resolution Millimeter-Wave Spectroscopy, and Ab Initio Calculations of Ethylmercury Hydride Manuel Goubet,*,† Roman A. Motiyenko,† Laurent Margulès,† and Jean-Claude Guillemin§ †

Laboratoire de Physique des Lasers, Atomes et Molécules, UMR CNRS 8523, Université de Lille 1, F-59655 Villeneuve d’Ascq, France § Ecole Nationale Supérieure de Chimie de Rennes, CNRS, UMR 6226, Avenue du Général Leclerc, CS 50837, F-35708 Rennes Cedex 7, France S Supporting Information *

ABSTRACT: The millimeter-wave rotational spectrum of an organomercury compound, ethylmercury hydride, has been recorded and assigned for the first time. The spectroscopic study is complemented by quantum chemical calculations taking into account relativistic effects on the mercury atom. The very good agreement between theoretical and experimental molecular parameters validates the chosen ab initio method, in particular its capability to predict accurate quartic centrifugal distortion constants related to this type of compound. Estimations of the nuclear quadrupole coupling constants have less predictive power than those of the structural parameters, but are good enough to satisfy the spectroscopic needs. In addition, the orientation of the axis of the H−Hg−C bonds deduced from the experimental nuclear quadrupole coupling constants compares well with the corresponding ab initio value. From the good agreement between experimental and theoretical results, together with the observation of the six most abundant isotopes of mercury, ethylmercury hydride is unambiguously identified as the product of the chemical reaction described here, and its calculated equilibrium geometry is confirmed.



INTRODUCTION The first characterization of organomercury hydrides RHgH (R = Me, Et, neophyl) was performed by their polarographic oxidation wave in a methanol solution in 1981.1 More recently, an extensive spectroscopic and ab initio study of gaseous HgH2 and HgD2 has been performed in the infrared region.2 However, their kinetic instability, toxicity, and repulsive smells account for the sparse literature on these compounds since this date. Thus, generating and trapping these compounds in situ3,4 has often been preferred to reactions starting from an isolated species.5 Nevertheless, alkyl-, aryl-, and vinyl-mercury hydrides have been isolated,6−9 and the approach reported for vinylic derivatives gave easy access to the low boiling derivatives.9 1H, 13 C, and 199Hg NMR spectra have been recorded, and, as a noteworthy feature, the chemical shift of the hydrogen connected to the mercury atom has been observed up to 17 ppm downfield.6,7,9,10 With compounds in the gas phase, the infrared spectra showed the νHg−H absorption around 1955 cm−1 1,11,12 and mass10 or photoelectron spectra9 were recorded. © 2012 American Chemical Society

In the present work, these investigations are extended to include the first study of the rotational spectrum of an organomercury hydride, namely, ethylmercury hydride, assisted by quantum chemical calculations. The capability of the chosen ab initio method to predict the spectroscopic constants is discussed, in particular the values of the quartic centrifugal distortion constants related to this type of system and the hyperfine nuclear quadrupole coupling constants. The calculated molecular geometry is confirmed by comparison with the experimental molecular constants deduced from the rotational spectrum.



EXPERIMENTAL SECTION Caution. Mercury compounds are potentially highly toxic materials which must be handled with great care, using a wellventilated hood. Received: March 19, 2012 Revised: May 14, 2012 Published: May 15, 2012 5405

dx.doi.org/10.1021/jp302617y | J. Phys. Chem. A 2012, 116, 5405−5409

The Journal of Physical Chemistry A

Article

Synthesis. The ethylmercury hydride was synthesized as previously reported for methylmercury hydride.9 A yield of 76% was obtained starting from 3 mmol of ethylmercury chloride as a precursor, with tributyltin hydride and duroquinone as the reducing agent, and tributyltin chloride as solvent. τ1/2 (5% in CDCl3): 7 min. 1H NMR (CDCl3, 400 MHz) δ 1.18 (qd, 2H, 3 JHH = 7.7 Hz, 3JHH = 3.7 Hz, CH2) ; 1.30 (td, 2H, 3JHH = 7.7 Hz, 4JHH = 3.3 Hz, CH3) ; 17.1 (tq, 1H, 3JHH = 3.7 Hz, 4JHH = 3.3 Hz, 1JHgH = 2339.8 Hz (d), HgH). 13C NMR (CDCl3, 100 MHz) δ 13.4 (t, 1JCH = 127.7 Hz, 2JHgC = 36.3 Hz (d), CH3) ; 32.6 (q, 1JCH = 131.3 Hz, 1JHgC = 831.2 Hz (d), CH2). Conventional Absorption Spectroscopy Experiment. Due to the high kinetic instability of ethylmercury hydride, experimental measurements of its rotational spectrum were performed using the Lille BWO-based fast scan spectrometer13 and in the so-called “flow mode”. While the absorption cell was kept at room temperature, the sample was evaporated outside the cell at a temperature of about −30 °C. Then the sample was continuously injected through a side opening at one end of the cell and pumped out through another side opening at the other end, and the optimum gas pressure in the cell was kept close to 25 Pa. In total, the spectral region 120−180 GHz was covered by the measurements. The fast scan spectra recorded in “flow mode” seemed to contain no significant impurities, at least those having a permanent dipole moment, since all the strongest lines observed were assigned to ethylmercury hydride.

Figure 1. The calculated structure of ethylmercury hydride at the MP2/cc-pVTZ-PP level.

each case, geometry optimizations went back to the linear conformation, showing that it is the only stable form. The transition state along the methyl internal rotation coordinate (C−C bond torsion) has also been characterized. A saddle point with an eclipsed orientation of the C−C bond groups, exhibiting only one imaginary frequency, has been found. Thus, the barrier height has been estimated to be 1080 cm−1 (1106 cm−1 including the zero point energy (ZPE) correction). These ab initio calculations were very helpful in analyzing the recorded rotational spectrum. The dipole moment components have been determined using the calculated Mulliken atomic charges and the Cartesian coordinates of the atoms in the principal inertial axis orientation, leading to values of μa = 0.43 D and μb = 0.13 D. In agreement with these results, the spectrum is dominated by series of aR0,1 bands separated by Δν ∼ 5.3 GHz (see Figure 2a), which is close to the theoretical value Δν = B + C ∼ 5.47 GHz.21 Almost every band consists of six subbands (see Figure 2b). The relative positions and intensities of these subbands correspond well to the isotopic shifts (the lighter species spectrum is shifted to higher frequencies) and the isotopic abundances of the mercury atom (198Hg: 9.97%, 199Hg: 16.87%, 200Hg: 23.10%, 201Hg: 13.18%, 202Hg: 29.86% and 204Hg: 6.87%,). The 196Hg isotope with a very low abundance of 0.15% is not observed. Another observed characteristic feature is the hyperfine structure of the 201 Hg isotope (with a nuclear spin I = 3/2) bands due to the nuclear quadrupole coupling (see Figure 2c). No additional line splitting is observed. Indeed, according to the relatively high barrier along the C−C bond torsion, no observable tunneling splitting of the rotational lines in the vibrational ground state, involving this internal rotation motion, is expected in the spectrum.20,22 From these observations, the Hamiltonian used for treating the spectrum was a standard asymmetric-top Watson’s Areduction Hamiltonian in Ir representation, including the centrifugal distortion corrections. The hyperfine structure of the 201Hg bands was modeled with the χii (i = a, b, or c) diagonal components of the nuclear quadrupole coupling tensor, taking into account that χaa + χbb + χcc = 0.21 The predictions and fits were undertaken using the SPCAT/SPFIT programs suite.23 Assignments were made by using a homemade graphical program interfacing the experimental spectrum with the predictions. The rotational parameters obtained as the result of the fits for all observed isotopes are listed in the Table 1. The initial guess for the main isotope (202Hg) spectrum was estimated from the ab initio set of rotational and quartic centrifugal distortion constants at the MP2/cc-pVTZ-PP level. Thanks to these predictive values,



THEORETICAL CALCULATIONS In this study, all the calculations were conducted using the Gaussian 09 software package.14 Basis sets that are not included in the software were obtained from the EMSL basis set library.15,16 The geometry was fully optimized, using the “tight” convergence option, and the vibrational frequencies were calculated at the MP2 level. In the case of atoms with a heavy nucleus like Mercury, relativistic effects due to the inner core electrons’ celerity close to the speed of light should be taken into account. Therefore, the small-core relativistic pseudopotential basis set (cc-pVTZ-PP) was employed to describe the Mercury atom,17 the cc-pCVTZ basis set with extra core/ valence functions was used for the carbon atoms,18,19 and the standard correlation-consistent polarized triple valence basis set (cc-pVTZ) of Dunning and co-workers18,19 was employed to describe the hydrogen atom’s electron. This combination was chosen because it has offered the best compromise between calculations time and accuracy of the results in a previous study of ethanetellurol.20 The transition state related to the methyl group internal rotation was characterized using the “QST2” procedure implemented in the Gaussian 09 software package. Finally, the nuclear quadrupole coupling constants for the 201 Hg isotope were estimated using various combinations of methods and basis sets, as discussed hereafter.



RESULTS AND ANALYSIS The ab initio geometry optimizations have led to an equilibrium structure with a staggered orientation of the C− C bond groups and a linear arrangement of the H−Hg−C bonds (see Figure 1). Geometrical parameters are given in Table S1, and calculated harmonic frequencies are listed in Table S2 in the Supporting Information. Calculations with other starting point geometries have been carried out, such as an antiperiplanar or synclinal orientation of the C−C−Hg-H dihedral angle (i.e., the stable conformations of ethanol). In 5406

dx.doi.org/10.1021/jp302617y | J. Phys. Chem. A 2012, 116, 5405−5409

The Journal of Physical Chemistry A

Article

Figure 2. A part of the experimental spectrum of ethylmercury hydride in the 152 − 170 GHz region showing (a) the intense series of aR0,1-type transitions, (b) the six subbands produced by different isotopic species of the Hg atom, and (c) the hyperfine structure for a transition of the 201Hg isotope.

Table 1. Spectroscopic Constants for the Six Observed Isotopes of Ethylmercury hydride 202

abund. /% A/MHz B/MHz C/MHz ΔJ /kHz ΔJK /kHz ΔK /kHz δJ /kHz δK /kHz HKJ /Hz LKKJ /mHz χaa /MHz χbb /MHz N lines Jmax; Kmax σ /MHz σwb

Hg

29.86 29302.2001(35) 29464.946a 2750.73634(10) 2817.149a 2593.295144(97) 2654.657a 1.349427(53) 1.372a −28.73258(74) −29.288a 514.23(20) 496.021a 0.148784(10) 0.154a 6.1780(76) 6.127a −2.0210(36) 0.1296(43)

561 52; 26 0.024 0.63

200

199

Hg

201

Hg

Hg

198

Hg

204

Hg

23.10 29302.823(13)

16.87 29303.129(11)

13.18 29302.57(19)

9.97 29303.509(44)

6.87 29301.593(56)

2753.73409(19)

2755.25383(17)

2752.22888(73)

2756.79080(44)

2747.79563(50)

2595.96453(19)

2597.31756(17)

2594.62014(57)

2598.68407(43)

2590.67525(49)

1.352251(62)

1.353555(61)

1.35144(20)

1.355191(96)

1.34668(11)

−28.76191(78)

−28.77643(82)

−28.7520(23)

−28.7924(11)

−28.7069(15)

513.07(94)

513.00(57)

514.23c

516(16)

514.23c

0.149193(44)

0.149477(45)

0.14952(16)

0.149911(88)

0.14848(12)

6.241(28)

6.200(32)

6.64(10)

6.392(65)

6.270(62)

−2.0265(38) 0.1334(48)

−2.0346(44) 0.1467(60)

−2.0342(56) 0.1419(76)

−2.0291(99) 0.143(16)

537 51; 26 0.023 0.63

483 43; 24 0.019 0.53

−1.996(11) −1169.50(67) 473.33(61) 559 34; 16 0.041 0.81

449 46; 25 0.027 0.66

395 34; 22 0.026 0.69

a

Ab initio values. bDimensionless rms error with most of the transitions assigned with a measurement uncertainty of 50 kHz for 201Hg and 30 kHz for all others isotopes. cFixed to the 202Hg isotope value.

of Ka < 10. Then, stepwise refinements of the fits were made by including increasingly higher Ka levels, less intense bR-type

quantum numbers were assigned to the experimental frequencies of the most intense aR-type transitions with values 5407

dx.doi.org/10.1021/jp302617y | J. Phys. Chem. A 2012, 116, 5405−5409

The Journal of Physical Chemistry A

Article

Table 2. Nuclear Quadrupole Coupling Constants χaa and χbb Calculated Using Various Combinations of Methods and Basis Setsa

transitions, and even less intense bQ-type transitions. To reach the experimental accuracy (root-mean-squared (rms) error of the fits less than 30 kHz), the HKJ sextic and LKKJ octic centrifugal distortion constants were included, especially for transitions involving levels with Ka > 20. Due to the lack of bRtype and bQ-type transitions for the 204Hg and 201Hg isotopes, the ΔK constant was fixed to the value of the main isotope. In addition, for the 201Hg isotope, the LKKJ constant was excluded from the fit because of the low maximum value of Ka. The consistency of the fits to the results over all observed isotopes gives confidence in the obtained sets of molecular constants. Their very good agreement with the ab initio values (see Table 1; less than 4% error most probably due to the comparison of fundamental state constants from the experiment with equilibrium constants from the calculations) makes unambiguous the identification of ethylmercury hydride as the product of the reaction described here. Estimations of the nuclear quadrupole coupling tensor χij for a molecule containing a heavy atom are less straightforward to obtain than the structural parameters. Since these constants involve properties of the nucleus (namely, the nuclear electric quadrupole moment eQ), a pseudopotential basis set should not be used. In this case, high-level all-electron calculations of the electric field gradient with atoms as heavy as mercury (i.e., 80 electrons) might become unreasonably time-consuming. However, Bailey has shown that low-level field gradient calculations (B1LYP/6-311G(df,p)) based on the geometry from high level calculations give very good results over a large collection of molecules with iodine as the heaviest quadrupolar atom.24 In the same manner, Kisiel et al. obtained very good agreement between experiments and calculations at the Hartree−Fock level with relatively small basis sets over 11 molecules containing quadrupolar nuclei up to iodine as well.25 For mercury, which is a post-transition metal of the next row down from iodine, obtaining predictive values from these methods is not obvious. Thus, a careful analysis of the nuclear quadrupole coupling constants was carried out using various combinations of methods and basis sets. The χii values for the 201 Hg isotope were calculated at the HF, B1LYP, and B3LYP levels with the all-electron ANO-RCC basis26 for the Hg atom and the cc-pVDZ, 6-311G(df,p), cc-pVTZ and ANO-RCC basis sets for the C and H atoms. First, for the consistency of the basis set over the molecule, the initial ANO basis was reduced to a double- or triple-ζ polarization (DZP or TZP) according to the set used for H and C. Second, the two-electron integrals were calculated using the second-order Douglas− Kroll−Hess (DKH) Hamiltonian to take into account the relativistic effect.27,28 Third, in each case, the input geometry was the atomic Cartesian coordinates calculated at the MP2/ccpVTZ-PP in the principal inertial axis orientation. Calculations, performed without taking care of the three important points cited above, have resulted in values changing signs and varying over an order of magnitude, depending on the basis set. Final results are displayed and compared with the experimental values in Table 2. Not surprisingly, values improve with an increasing level of the method and number of basis functions. This trend implies the validity of the results. The agreement is improved faster by increasing the level of the method rather than the quality of the basis set, so that if one has to compromise between accuracy and calculation time, the best choice would be a combination of a high level method and a small basis set. Moreover, at the HF level, the discontinuity in the results when using the cc-pVTZ basis set shows that this

a

χaa /MHz χbb /MHz

HF

B1LYP

B3LYP

exp.

cc-pVDZ (126) 6-311G(df,p) (154) ANO−DZP (180) cc-pVTZ (237) ANO-TZP (291)

−1962 786 −1956 784 −1953 782 −1919 778 −1942 778

−1545 617 −1532 612 −1513 603 −1467 585 −1415 562

−1517 606 −1503 600 −1481 590 −1435 572 −1376 546

−1170 473

The number of basis functions for each set is reported in parentheses.

level might be less reliable than the others. In the case of a relatively small molecule such as ethylmercury hydride, the highest combination (B3LYP/ANO-TZP) represents a calculation of only about 29 h of CPU time, which is very reasonable according to the actual computer’s capabilities. However, even at this highest level, results could not be considered as reliably predictive as for molecules containing atoms up to the fifth row, but as a good initial guess when starting the spectroscopic analysis. In particular, the sign is very valuable for the first assignments. Indeed, in this case of a half-integer nuclear spin, opposite signs of both χaa and χbb will lead to a hyperfine structure of a band with ΔF = ΔJ transitions almost symmetric compared to its center of gravity, so that a fit will converge with misassigned frequencies. Geometrical parameters involving the quadrupolar nucleus can be obtained from the nuclear quadrupole coupling tensor components,29 as briefly reviewed in the following. In the case of ethylmercury hydride, the observation of only a-type and btype transitions shows that the molecule is in the Cs point group with the symmetry plane being the principal inertial plane ab. If the internuclear axis is called z and assuming that the electronic density is cylindrical about this axis, the angle αaz between the principal inertial axis a, and the bond axis z can be calculated from the formula αaz =

⎛ −2χ ⎞ 1 ab ⎟⎟ tan−1⎜⎜ 2 ⎝ χaa − χbb ⎠

(1)

In the present case, only transitions with ΔF = ΔJ have been observed, the frequencies of which are insensitive to the offdiagonal term χab. However, this parameter can be estimated from the diagonal components in the assumption that the angular oscillation of the subunit containing the quadrupolar nucleus is two-dimensionally isotropic in the ab plane:29 χab2 = χaa χbb + 2χcc2

(2)

Although eqs 1 and 2 have been determined for a singly bonded terminal quadrupolar nucleus, these formulas are assumed to be valid for ethylmercury hydride according to the linear arrangement of the two bonds around the mercury atom. The αaz angle would then represent the angle between the molecular principal inertial axis a and the H−Hg−C axis. From the experimental values of the diagonal components of the nuclear quadrupole coupling tensor, χab = 644.78 MHz is estimated from eq 2 and used in eq 1, leading to a value of αaz = 5408

dx.doi.org/10.1021/jp302617y | J. Phys. Chem. A 2012, 116, 5405−5409

The Journal of Physical Chemistry A

Article

19.1°. For comparison, a value of αaz = 19.6° is calculated from the ab initio Cartesian coordinates in the principal inertial axis orientation of the H and C atoms. This very good agreement, together with the comparable values of the experimental and theoretical rotational constants (see Table 1), validates unambiguously the calculated equilibrium structure of ethylmercury hydride.

(5) Bellec, N.; Guillemin, J.-C. Tetrahedron Lett. 1995, 36, 6883− 6886. (6) Craig, P. J.; Mennie, D.; Needham, M.; Oshah, N.; Donard, O. F. X.; Martin, F. J. Organomet. Chem. 1993, 447, 5−8. (7) Craig, P. J.; Garraud, H.; Laurie, S. H.; Mennie, D.; Stojak, G. H. J. Organomet. Chem. 1994, 468, 7−11. (8) Kwetkat, K.; Kitching, W. J. Chem. Soc., Chem. Commun. 1994, 345−347. (9) Guillemin, J.-C.; Bellec, N.; Kiz-Szetsi, S.; Nyulaszi, L.; Veszpremi, T. Inorg. Chem. 1996, 35, 6586−6591. (10) Craig, P. J.; Needham, M. I.; Ostah, N.; Stojak; Stojak, G. H.; Symons, M.; TeesdaleSpittle, P. J. Chem. Soc., Dalton Trans. 1996, 153−156. (11) Filippelli, M.; Baldi, F.; Brinckman, F.; Olson, G. J. Environ. Sci. Technol. 1992, 26, 1457−1460. (12) Greene, T. M.; Andrews, L.; Downs, A. J. J. Am. Chem. Soc. 1995, 117, 8180−8187. (13) Alekseev, E. A.; Motiyenko, R. A.; Margulès, L. Radio Phys. Radio Astron. 2012, 3, 75−88. (14) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, revision B.01; Gaussian, Inc.: Wallingford, CT, 2009. (15) Feller, D. J. Comput. Chem. 1996, 17, 1571−1586. (16) Schuchardt, K. L.; Didier, B. T.; Elsethagen, T.; Sun, L.; Gurumoorthi, V.; Chase, J.; Li, J.; Windus, T. L. J. Chem. Inf. Model. 2007, 47, 1045−1052. (17) Peterson, K. A.; Figgen, D.; Goll, E.; Stoll, H.; Dolg, M. J. Chem. Phys. 2003, 119, 11113. (18) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007−1023. (19) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1995, 103, 4572−4585. (20) Motiyenko, R. A.; Margulès, L; Goubet, M.; Mollendal, H.; Konovalov, A.; Guillemin, J.-C. J. Phys. Chem. A 2010, 114, 2794− 2798. (21) Gordy, W.; Cook, R. L. Microwave Molecular Spectra, 3rd ed.; John Wiley & Sons, Inc.: New-York, 1984. (22) Demaison, J.; Margulès, L.; Mäder, H.; Sheng, M.; Rudolph, H. D. J. Mol. Spectrosc. 2008, 252, 169−175. (23) Pickett, H. M. J. Mol. Spectrosc. 1991, 148, 371−377 See also: http://spec.jpl.nasa.gov/. (24) Bailey, W. C. Calculation of Nuclear Quadrupole Coupling Constants in Gaseous State Molecules. http://web.mac.com/ wcbailey/nqcc/. (25) Kisiel, Z.; Bialkowska-Jaworska, E.; Pszczolkowski, L. J. Chem. Phys. 1998, 109, 10263−10272. (26) Roos, O.; Lindh, R.; Malmqvist, P.-A.; Veryazov, V.; Widmark, P.-O. J. Phys. Chem. A 2005, 109, 6575−6579. (27) Barysz, M.; Sadlej, A. J. J. Mol. Struct. (THEOCHEM) 2001, 573, 181−200. (28) deJong, W. A.; Harrison, R. J.; Dixon, D. A. J. Chem. Phys. 2001, 114, 48−53. (29) Legon, A. C. Faraday Discuss. 1994, 97, 19−33.



CONCLUSION This paper presents the results of studies of six ethylmercury hydride isotopes. The millimeter-wave rotational spectrum has been recorded and assigned for the first time. The assignment was supported by high-level ab initio calculations taking into account relativistic effects on the mercury atom. Calculated and experimental values of the rotational constants agree with about 2% accuracy. Together with the observation of the six most abundant isotopes of mercury, this very good agreement validates unambiguously the identification of ethylmercury hydride as the product of the chemical reaction described here. The diagonal components of the nuclear quadrupole coupling tensor χaa and χbb for the 201Hg isotope have been estimated by all-electron calculations using various combinations of methods and basis sets. The comparison with the experimental values shows that these estimations do not have the predictive power of those for molecules containing iodine, but the sign and the order of magnitude are good enough to satisfy the spectroscopic needs for initial fitting of parameters. Finally, the orientation of the H−Hg−C bond axis compared to the molecular principal inertial axis a has been calculated from the experimental nuclear quadrupole coupling constants. The good agreement with the theoretical value confirms the observation of ethylmercury hydride and the validity of the calculated equilibrium geometry.



ASSOCIATED CONTENT

S Supporting Information *

Calculated molecular structure and harmonic vibrational frequencies. Rotational line assignments, measured frequencies, experimental uncertainties, and deviations from the final fits for the six studied isotopes of ethylmercury hydride. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Telephone: +33-3-20434905. Fax: +33-3-20337020. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS D. Duflot, F. Réal, and V. Vallet are gratefully acknowledged for helpful discussions on ab initio calculations. The calculations were partially performed on the PhLAM cluster financed by the European Regional Development Fund (FEDER) through the “Contrat de Projets Etat Region” (CPER) 2007−2013.



REFERENCES

(1) Devaud, M. J. Organomet. Chem. 1981, 220, C27−C29. (2) Shayesteh, A.; Yu, S.; Bernath, P. F. J. Phys. Chem. A 2005, 109, 10280−10286. (3) Giese, B. Angew. Chem., Int. Ed. Engl. 1985, 24, 553−565. (4) Barluenga, J.; Yus, M. Chem. Rev. 1988, 88, 487−509. 5409

dx.doi.org/10.1021/jp302617y | J. Phys. Chem. A 2012, 116, 5405−5409