Photodissociation and theoretical studies of the
Au
1–(C
5H
5N) complex
Hsu-Chen Hsu,aaFang-Wei Lin,aaChun-Chia Lai,bbPo-Hua Suaaand Chen-Sheng Yeh*aa
a
Department of Chemistry, National Cheng Kung University, Tainan, Taiwan 701, R.O.C. E-mail:csyeh@mail.ncku.edu.tw
b
Department of Chemistry, National Changhua University of Education, Changhua, Taiwan 50058, R.O.C.
Receivved (in Montpellier, France) 8th October 2001, Accepted 20th Novvember 2001 First published as an Advvance Article on the web
Laser vaporization combined with a supersonic molecular beam was employed to generate and study the Au(C5H5N)þcomplex for the first time. On the basis of the ionization energies (IE) between gold and pyridine,
the Au(C5H5N)þspecies is viewed as being a Auþ–C5H5N species. Photodissociative charge transfer was
observed with exclusive C5H5Nþ(pyridine) formation. The photofragmentation spectrum of Auþ–(C5H5N)
was scanned by monitoring the pyridine fragments. A structureless continuum spectrum was observed and the onset ofC5H5Nþappearance indicates the upper limit on the Auþ–C5H5N bond strength to be 59.6 kcal mol1.
Ab initiocalculations at the MP2 level were employed to optimize the geometries ofthe gold complexes and binding energies were obtained using CCSD(T) single point calculations. Besides from the C2vstructure
observed in Cuþand Agþcomplexes, the theoretical results yielded a second isomer with C1symmetry which is
24 kcal mol1less stable in energy than the C2visomer.
Metal ion–molecule complexes are especially ofinterest in the areas ofcatalysis, organometallic chemistry and biological sciences. There is no doubt that the bond energy is ofcentral importance for understanding the thermodynamic aspects ofmetal–molecule interactions. The electrostatic or covalent nature ofthe bonding has been described for a variety of complexes.1–3 In our continued efforts to investigate the che-mical and physical processes involving metal cations and het-erocyclic molecules in the gas phase, we have now studied the isolated Au(C5H5N)þcomplex.
In metal ion-molecule complexes, the third-row transition metals have received much less attention as compared to the large amount ofdata available for first- and second-row transition metals, in both experimental and theoretical studies. The heavy Auþ is ofparticular interest because ofthe pro-nounced relativistic effects that it may exhibit.4These result in electron transfer from ligands to Auþ and increases the covalent binding between Auþand the ligands. In this respect, Schwarz and co-workers have studied the complexation of Auþwith a series ofligands, such as F, CO, H2O, NH3, C2H4,
C3H6, C6H6and C6F6, quite extensively. 2,5
Experimental and theoretical results revealed that the binding ofAuþto several ligands is distinctly different from the other transition metals. In particular, an exceptionally large relativistic stabilization was observed in both Auþ–NH3 and Auþ–C2H4 complexes.5
Using the radiative association kinetics method, Dunbar and co-workers investigated the reactions ofAuþand Au with C6H6and C6F6as well.6
The photodissociation technique has proven to be an effective way to probe bond strengths ofion species.7 Recently, we
obtained an upper limit for the binding energies of the Mþ– pyridine complexes (Mþ¼ Cuþ, Agþ) with C
2v structures by
observing a photodissociative charge transfer (CT) fragment (the pyridine cation).8,9In this report, the Au(C5H5N)þcomplex was
generated for the first time and laser excitation was employed to inspect its photochemical behavior and ground-state binding energy. The structures and binding energies ofthe Au complexes
were also obtained from ab initio calculations, which showed there were two isomers with C2vand C1symmetries.
Experimental
The details ofthe experimental setup were described pre-viously.9The Au(C5H5N)þcomplex was generated using laser
vaporization combined with a supersonic molecular beam source. A gold foil (99.9%) was wrapped around an Al rod and the target rod was suspended in a cutaway holder, which is a rod holder without a growth channel to produce primarily ion complexes containing one and=or two ligands. The rotating and translating rod was irradiated using the 532 nm wavelength ofthe second harmonic output from a Nd:YAG laser operated at 10 Hz with 1–2 mJ pulse1. The vaporization laser beam was focused 1.5 cm in front of the metal rod. Prior to seeding the pyridine vapor, pure He gas (6 atm) was used to inspect the ion products from our source. It was found that only the Auþ atomic signal appeared, and that no other impurities, such as Alþand metal oxides, were detected. Ion complexes containing pyridine expanded adiabatically and were skimmed into a reflectron time-of-flight mass spectrometer, followed by mass selection ofAu(C5H5N)þfor carrying out the
photodissocia-tion experiments. An unfocused dissociaphotodissocia-tion laser (pulsed tunable dye laser pumped by an Nd:YAG laser) crossed the ion beam at the turning point in the reflector. The wavelength-dependent spectrum was measured by recording the intensity of the fragment as a function of the dissociation laser wavelength. The laser fluence was kept low (1.2 mJ cm2) to minimize
multiphoton absorption. The fragment power dependence revealed a linear relation with fluence up to at least 7 mJ cm2.
Theoretical computations
The calculations were performed using second-order M ller– Plesset perturbation theory (MP2). The double-zeta valence
DOI: 10.1039/b109107g New J. Chem., 2002, 26, 481–484 481
basis sets and the effective core potentials (ECPs) ofHay and Wadt were used for the Au atom10and the 6-31G(d,p) set was selected for C, H, and O. The relativistic effects were included in the Hay–Wadt ECPs ofAu. We also added fpolarization functions to increase the flexibility of the ECP basis set. The f exponent was 1.050 for Au.11The MP2=6-31G(d,p) level was employed to fully optimize the geometries of the ligand, transition state and ion complexes and the resulting harmonic vibrational frequencies were used to evaluate the zero-point energy corrections. CCSD(T) single point calculations with the corresponding MP2=6-31G(d,p) zero-point energies were derived from CCSD(T)=6-31G(d,p)==MP2=6-31G(d,p) calcu-lations. All calculations were performed using the Gaussian 98 suite ofprograms.12
Results and discussion
Fig. 1(a) shows both the Auþand Au(C5H5N)þmass peaks,
where the experimental conditions were adjusted to generate ion species in the low mass range. From the viewpoint oftheir ionization energies (IE), the positive charge should be localized on the Au atom, which has an IE of9.22 eV, lower than that of the pyridine molecule (IE¼ 9.26 eV).13,14The associated
pyri-dine adduct is viewed as being Auþ–pyridine. However, the IE difference between Au and pyridine is only 0.92 kcal mol1 (0.04 eV). It is possible that the charge transfer reaction of Auþ with pyridine gives rise to C5H5Nþ, followed by formation of
the Au–C5H5Nþ condensation product in the ion source. In
this case the C5H5Nþpeak should have a high probability of
being observed in addition to Auþ and Au(C5H5N)þ in the
mass spectrum, ifthe pyridine cation is formed in the beam
source. We saw no ions other than the product channels assigned in Fig. 1(a), that is no C5H5Nþ. It is worthy ofmention
that in our studies on the Au(furan)þsystem, where furan has a lower IE (8.89 eV) than that ofthe Au atom, we readily observed the molecule cation signal, furanþ.15 Although it is most likely that the positive charge resides on the Au atom, Au–pyridineþ complex formation cannot be excluded com-pletely. It is interesting to note that Dunbar and co-workers used the radiative association methodology to study the reac-tion ofAuþwith C6H6.6Benzene has a similar IE (9.25 eV) to
that ofpyridine.14Their results indicated that the Auþcharge
transfer reaction accompanied by C6H6þformation has a rate
constant close to zero. Such an experimental approach will be valuable to understand the possibility ofcharge transfer between Auþand pyridine.
Photodissociation was performed on the mass-selected Au(C5H5N)þ complex, as shown in Fig. 1(b). The resulting
fragment ion appeared as C5H5Nþ. Since the Auþ–C5H5N
complex is the parent signal, the photodissociation induces a ligand-to-metal CT. The same experimental results were obtained when the laser excited pyridine complexes containing Cuþ and Agþ.8,9 The photofragment spectrum over the 435–485 nm wavelength region is given in Fig. 2. Throughout the experiments, the parent ion intensity remained approxi-mately constant. Due to the limitation ofour dye laser, the spectrum could not be recorded in the 366–430 nm region. The pyridine daughter ion was the only fragment observed during the photodissociation process. No C5H5Nþchannel could be
detected between 472.7 and 485 nm, as seen in Fig. 2. The threshold ofphotofragment appearance was determined to be 472.3 nm, at which wavelength the C5H5Nþwas barely seen,
and the pyridine intensity continuously increased as the exci-tation wavelength was scanned towards the blue. As men-tioned above, a photo-induced dissociative CT has taken place from the Auþ–C5H5N complex. On the basis ofthe energetic
diagram shown in Fig. 3, the lowest CT state correlating to Au(2S)þ C
5H5Nþis only 0.92 kcal mol1(0.04 eV) above the
Auþ(1S)þ C5H5N(1A1) states corresponding to the Auþ–
C5H5N ground state. Excitation ofAuþ–C5H5N into the
dis-sociation limit ofthe lowest CT electronic state is a direct step yielding C5H5Nþproduct, whereas dissociative CT may also
involve a transition to a higher electronic surface, followed by electronic curve crossing to the Au(2S)þ C
5H5Nþasymptote.
It should be noted that the interpretation ofthe dissociative mechanism has been restricted to a one-photon absorption. A two-photon excitation cannot be ruled out ifthe absorption cross-section is significantly different for the two steps, while the power dependence measurement exhibits a linear relation.
Fig. 1 (a) Mass spectrum ofAuþ and Au(C5H5N)þ. (b) Photo-fragmentation mass spectrum of mass-selected Au(C5H5N)þat 355 nm with 5 mJ cm2light fluence.
Fig. 2 Photofragmention spectrum of Au(C5H5N)þ obtained by monitoring the pyridine cation signal as a function of laser wavelength. The arrow indicates the threshold ofpyridine appearance. The error bars represent the fragment intensity deviations. Each data point is an average oftwo or three measurements.
Taking the CT threshold (472.3 nm) as a reference, the upper bound ofthe binding energy (BDE) in the ground state can be derived from the following relation:
hn5D000þ DIE
DIE represents the IE difference (0.92 kcal mol1) between Au and pyridine. The upper limit on the Auþ–C5H5N bond
strength was calculated to be 59.6 kcal mol1. As reported previously, the same experimental method was employed to study pyridine complexes containing Cuþ and Agþ.8,9 The
observed upper limit BDEs are 65.5 and 45.2 kcal mol1for Cuþ–C5H5N and Agþ–C5H5N, respectively. Auþ–C5H5N has
a slightly smaller BDE than Cuþ–C5H5N and a greater one
than Agþ–C5H5N. For the complexation ofAuþwith other
nitrogen base ligands, ion-molecule reactions using the bracket methodology performed by Schwarz and co-workers led to a binding energy of71 7 kcal mol1 for the Auþ–NH3
complex.5c
Theoretical approaches as well were employed to obtain the complex geometry and binding energies. MP2=6-31G(d,p) was used to fully optimize the complex structures. Earlier studies showed that the optimizations ofthe pyridine complexes containing Cu and Ag ended up with the C2vminimum.8,9The
metal atoms lie off ofthe nitrogen atom and effect a slight distortion on the pyridine structure. The Au complex was revealed to also have C2vsymmetry (Fig. 4) with BDEs of69.9
and 67.0 kcal mol1in the MP2 and CCSD(T) calculations, respectively. The complexation ofAuþwith a series ofligands investigated by Schwarz and co-workers indicated that partial electron transfer occurred from the ligands to the gold metal ion, resulting in increased covalent interactions between Auþ and the ligands due to the relativistic effects in the gold atom.2,5Considering the p orbital electrons ofthe pyridine as
possible candidates to undergo electron transfer, one isomeric structure with C1 symmetry was observed in which the Au
atom resided over the C1–C2 bond and was slightly shifted
outside the perimeter ofthe ring, as depicted in Fig. 4(a). Such
a binding site in C1geometry could also be seen in the
inter-action ofAuþwith the benzene molecule, studied by Koch and co-workers.16 This intermediate state is bound by 43.0 kcal mol1[CCSD(T)] relative to the separated Auþand C5H5N.
Such a stable intermediate was not found in the pyridine complexes containing Cuþand Agþ.8,9As shown in Fig. 5, a and c are connected by a transition structure b, which has Auþ lying over the C1–N bond, and the activation barrier was
estimated to be 5.8 kcal mol1for the C1! C2visomerization.
Although a stable intermediate state a was observed, it remains to clarify whether or not the relativistic effects in Au play a key role for such an observation. A detailed theoretical analysis is required to resolve this question.
Table 1 lists the binding energies ofthe Mþ–pyridine com-plexes (Mþ¼ Cuþ, Agþ, and Auþ). The calculated binding energies from either MP2 or CCSD(T) show that the gold complex with a C2v structure has the strongest binding as
compared to the copper and silver complexes. As mentioned earlier and from the results shown for the Cuþ and Agþ complexes, the employed photodissociation technique result-ing in CT would provide an upper limit for the bindresult-ing energy in the ground state. Our determined upper binding energy is
Fig. 4 Geometrical parameters ofthe stationary points ofthe [Auþ, C5H5N] potential energy surface. The parameters were calculated at the MP2=6-31G(d, p) level oftheory. Selected lengths in angstroms and bond angles in degrees.
Fig. 3 Energy level diagram displaying the asymptotic states corres-ponding to gold and pyridine. The Auþ–pyridine ground state lies 59.6 kcal mol1below the Auþ(1S)þ pyridine(1
A1) reference level, as de-termined from the photodissociation experiments.
located between the a and c complex values. It is possible that the a isomer was synthesized in our cluster source. However, c formation cannot be ruled out, since two possible factors, internally hot ions and a two-photon excitation mechanism, need to be considered as possibly leading to an artificially low experimental threshold. Using Ar as the carrier gas provides a way to cool vibrationally excited ions. The results indicated that the onset ofpyridine appearance remained constant regardless ofthe gas properties. Another possible contribution is that two-photon absorption might take place at the photo-dissociation threshold. Dunbar and Honovich showed that one-photon behavior could occur ifa small fraction ofone-photon ions together with two-ofone-photon species were present in the threshold region.17The photofragment kinetic energy is the other uncertainty in the course ofcomplex dissociation, with consideration ofthe kinetic shift resulting in the true onset appearing shifted towards the red. In a survey of previous studies using the same experimental strategy as employed here, photodissociation performed on the Agþ–benzene complex by Duncan et al. resulted in an upper binding energy of30 kcal mol1,186.5 kcal mol1less than the theoretical prediction.19 It also should be mentioned that further theoretical calcula-tions using better basis sets should give a more confident binding energy for the Au ion complex. At this stage, our experiments cannot justify the prepared Auþ–C5H5N identity
from the cluster source. On the basis of the theoretical pre-diction at the CCSD(T) level oftheory, the CT dissociation threshold ofthe C2visomer is determined to lie at 421 nm. It
would be helpful to conduct the photodissociation signal search in the 366–430 nm region, which is inaccessible to our dye laser.
Acknowledgement
We thank the National Science Council ofthe Republic of China for its financial support for this work.
References
1 C. W. Bauschlicher and H. Partridge, J. Phys. Chem., 1991, 95, 3946.
2 J. Hrus˘a´k, D. Schro¨der and H. Schwarz, Chem. Phys. L ett., 1994, 225, 416.
3 C. N. Yang and S. J. Klippenstein, J. Phys. Chem. A, 1999, 103, 1094.
4 P. Pyykko¨, Chem. Rev., 1988, 88, 563.
5 See, for example: (a) P. Schwerdtfeger, J. S. McFeaters, M. J. Liddell, J. Hrus˘a´k and H. Schwarz, J. Chem. Phys., 1995, 103, 245; (b) J. Hrus˘a´k, R. H. Hertwig, D. Schro¨der, P. Schwerdtfeger, W. Koch and H. Schwarz, Organometallics, 1995, 14, 1284; (c) D. Schro¨der, H. Schwarz, J. Hrus˘a´k and P. Pyykko¨, Inorg. Chem., 1998, 37, 624; (d) R. H. Hertwig, W. Koch, D. Schro¨der, H. Schwarz, J. Hrus˘a´k and P. Schwerdtfeger, J. Phys. Chem., 1996, 100, 12 253.
6 Y.-P. Ho and R. C. Dunbar, Int. J. Mass Spectrom. Ion Processes, 1999, 182=183, 175.
7 S. W. Buckner and B. S. Freiser, Polyhedron, 1988, 7, 1583. 8 Y. S. Yang, W. Y. Hsu, H. F. Lee, Y. C. Huang, C. S. Yeh and
C. H. Hu, J. Phys. Chem. A, 1999, 103, 11 287.
9 Y. S. Yang and C. S. Yeh, Chem. Phys. L ett., 1999, 305, 395. 10 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299. 11 A. W. Ehlers, M. Bo¨hme, S. Dapprich, A. Gobbi, A. Ho¨llwarth,
V. Jonas, K. F. Ko¨hler, R. Stegmann, A. Veldkamp and G. Frenking, Chem. Phys. L ett., 1993, 208, 111.
12 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., 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. Ragha-vachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, 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. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian 98, rev. A.7, Gaussian, Inc., Pittsburgh, PA, 1998. 13 E. Moore, Atomic Energy L evels, National Bureau ofStandards,
Washington DC, 1971.
14 G. Herzberg, Electronic Spectra and Electronic Structure of Polyatomic Molecules, Van Nostrand Reinhold, New York, 1966. 15 P. H. Su, F. W. Lin and C. S. Yeh, J. Phys. Chem. A, 2001, 105,
9643.
16 T. K. Dargel, R. H. Hertwig and W. Koch, Mol. Phys., 1999, 96, 583.
17 R. C. Dunbar and J. P. Honovich, Int. J. Mass Spectrom. Ion Processes, 1984, 58, 25.
18 K. F. Willey, P. Y. Cheng, M. B. Bishop and M. A. Duncan, J. Am. Chem. Soc., 1991, 113, 4721.
19 C. W. Bauschlicher, H. Partridge and S. R. Langhoff, J. Phys. Chem., 1992, 96, 3273.
Fig. 5 Potential energy surface for the interaction of Auþ with C5H5N. The energies were obtained with the CCSD(T)=6-31G(d,p)== MP2=6-31G(d,p) method.
Table 1 Binding energies (D0in kcal mol1) for metal–pyridine com-plexes
Complex Symmetry Method D0
Cuþ-pyridinea C2v MP2=6-31G(d, p) 57.8 Exptal. 65.6 Agþ-pyridinea C2v MP2=6-31G(d, p) 39.3 Exptal. 45.2 Auþ-pyridine C2v MP2=6-31G(d, p) 69.9 C2v CCSD(T)=6-31G(d, p)== MP2=6-31G(d, p) 67.0 C1 MP2=6-31G(d, p) 47.5 C1 CCSD(T)=6-31G(d, p)== MP2=6-31G(d, p) 43.0 Exptal. 59.7
a Taken from ref. 8.
