Mechanistic study on the gas-phase generation of "rollover"- Cyclometalated [M(bipy - H)]+ (M = Ni, Pd, Pt)

Article abstract of DOI:10.1021/om100757e

"Rollover"-cyclometalated [Pt(bipy - H)]⁺ (bipy = 2,2′-bipyridine) can be easily generated in the gas phase via HX elimination brought about by collision-induced dissociation (CID) of the cationic complexes [Pt(X)(bipy)]⁺ with X = CH₃, Cl; the latter as well as other [M(X)(bipy)]⁺ complexes (M = Ni, Pd; X = CH₃, F, Cl, Br, I, OAc) are accessible by electrospray ionization mass spectrometry. For the nickel and palladium methyl complexes [M(CH ₃)(bipy)]⁺, upon CID, no cyclometalation occurs; rather, homolytic cleavage of the M-CH₃ bond takes place. The related chloro complexes [M(Cl)(bipy)]⁺ (M = Ni, Pd) undergo competitive eliminations of HCl and Cl upon CID, and the branching ratios depend strongly on the collision energy. On the basis of DFT calculations, this metal- and ligand-controlled behavior is a consequence of the rather different energetic requirements for the direct loss of X versus elimination of HX (X = CH ₃, Cl). Deuterium-labeling experiments reveal that formation of CH₄ and HCl is only for the platinum complexes due to a genuine "rollover" cyclometalation process, i.e., selective abstraction of a hydrogen atom from the C(3)-position of bipy. In the series of halo-substituted complexes [Ni(X)(bipy)]⁺ (X = F, Cl, Br, I), the Ni-X bond strength decreases in the sequence F > Cl > Br > I. For X = F, one observes the elimination of HF, which benefits from the particular stability of this molecule; the hydrogen atom in HF is mostly (>90%) abstracted from position C(6) with <10% originating from C(3) of the bipy ligand; thus, for this system "rollover" cyclometalation occurs only to a small amount. [Ni(Cl)(bipy)]⁺ undergoes competitive losses of HCl and Cl upon collision with Xe, and for X = Br and I only homolytic Ni-X bond cleavage takes place. In the elimination of HCl from [Ni(Cl)(bipy)]⁺, >60% of the hydrogen atoms originate from C(3), and the remaining from C(4,5,6), as inferred from deuterium-labeling experiments. The acetate complexes [Ni(OAc)(bipy)]⁺ and [Pd(OAc)(bipy)]⁺ exhibit eliminations of neutral AcO^?, and at elevated collision energies, decarboxylation occurs. C-H bond activation resulting in the formation of HOAc is absent at the detection limit.

Full text of DOI:10.1021/om100757e

6002 Organometallics 2010, 29, 6002–6011
DOI: 10.1021/om100757e
Mechanistic Study on the Gas-Phase Generation
of “Rollover”-Cyclometalated [M(bipy - H)]þ (M = Ni, Pd, Pt)†
Burkhard Butschke and Helmut Schwarz*
€
Institut fu€r Chemie, Technische Universitat Berlin, Strasse des 17. Juni 135, D-10623 Berlin, Germany
Received August 3, 2010
“Rollover”-cyclometalated [Pt(bipy-H)]þ (bipy=2,20-bipyridine) can be easily generated in the
gas phase via HX elimination brought about by collision-induced dissociation (CID) of the cationic com-
plexes[Pt(X)(bipy)]þ with X = CH3, Cl; thelatter as well as other [M(X)(bipy)]þ complexes (M = Ni, Pd;
X=CH3, F, Cl, Br, I, OAc) are accessible by electrospray ionization mass spectrometry. For the nickel and
palladium methyl complexes [M(CH3)(bipy)]þ, upon CID, no cyclometalation occurs; rather, homolytic
cleavage of the M-CH3 bond takes place. The related chloro complexes [M(Cl)(bipy)]þ (M = Ni, Pd)
undergo competitive eliminations of HCl and Cl upon CID, and the branching ratios depend strongly on
the collision energy. On the basis of DFT calculations, this metal- and ligand-controlled behavior is a
consequence of the rather different energetic requirements for the direct loss of X versus elimination of HX
(X = CH3, Cl). Deuterium-labeling experiments reveal that formation of CH4 and HCl is only for the
platinum complexes due to a genuine “rollover” cyclometalation process, i.e., selective abstraction of a
hydrogen atom from the C(3)-position of bipy. In the series of halo-substituted complexes [Ni(X)(bipy)]þ
(X = F, Cl, Br, I), the Ni-X bond strength decreases in the sequence F>Cl>Br>I. For X=F, one
observes the elimination of HF, which benefits from the particular stability of this molecule; the hydrogen
atom in HF is mostly (>90%) abstracted from position C(6) with <10% originating from C(3) of the bipy
ligand; thus, for this system “rollover” cyclometalation occurs only to a small amount. [Ni(Cl)(bipy)]þ
undergoes competitive losses of HCl and Cl upon collision with Xe, and for X = Br and I only homolytic
Ni-X bond cleavage takes place. In the elimination of HCl from [Ni(Cl)(bipy)]þ, >60% of the hydrogen
atoms originate from C(3), and the remaining from C(4,5,6), as inferred from deuterium-labeling
experiments. The acetate complexes [Ni(OAc)(bipy)]þ and [Pd(OAc)(bipy)]þ exhibit eliminations of
neutral AcO•, and at elevated collision energies, decarboxylation occurs. C-H bond activation resulting in
the formation of HOAc is absent at the detection limit.
Introduction
not only because of their photoluminescent and electronic
properties.2 In fact, they are also used as metallomesogens3 or
as building blocks for supramolecular architectures;4 further,
they are encountered as intermediates in different metal-
mediated organic transformations, such as the Heck reaction
or the Suzuki coupling,5 and can be employed in the synthesis of
Among the plethora of organometallic compounds, cyclom-
etalated transition-metal complexes1 are extremely versatile
† Dedicated to Professor H.-J. Freund on the occasion of his 60th birthday.
*To whom correspondence should be addressed. Fax: (þ49) 30-314-
21102. E-mail: Helmut.Schwarz@mail.chem.tu-berlin.de.
(1) (a) Cope, A. C.; Siekman, R. W. J. Am. Chem. Soc. 1965, 87, 3272.
(b) Dehand, J.; Pfeffer, M. Coord. Chem. Rev. 1976, 18, 327. (c) Bruce, M. I.
Angew. Chem., Int. Ed. Engl. 1977, 16, 73. (d) Omae, I. Chem. Rev. 1979,
79, 287. (e) Omae, I. Coord. Chem. Rev. 1980, 32, 235. (f) Constable, E. C.
Polyhedron 1984, 3, 1037. (g) Rothwell, I. P. Polyhedron 1985, 4, 177. (h)
Ryabov, A. D. Chem. Rev. 1990, 90, 403. (i) Albrecht, M.; van Koten, G.
Angew. Chem., Int. Ed. 2001, 40, 3750. (j) van der Boom, M. E.; Milstein, D.
Chem. Rev. 2003, 103, 1759. (k) Omae, I. Coord. Chem. Rev. 2004, 248,
(3) (a) Pucci, D.; Barberio, G.; Bellusci, A.; Crispini, A.; Ghedini, M.
J. Organomet. Chem. 2006, 691, 1138. (b) Marcos, M. In Metallomesogens;
ꢀ
Jose Luis, S., Ed.; Wiley-VCH: Germany, 2007; p 235.
(4) (a) Huck, W. T. S.; Hulst, R.; Timmerman, P.; van Veggel, F. C. J.
M.; Reinhoudt, D. N. Angew. Chem., Int. Ed. 1997, 36, 1006. (b) Hall,
J. R.; Loeb, S. J.; Shimizu, G. K. H.; Yap, G. P. A. Angew. Chem., Int. Ed.
1998, 37, 121. (c) Huck, W. T. S.; Prins, L. J.; Fokkens, R. H.; Nibbering,
N. M. M.; van Veggel, F. C. J. M.; Reinhoudt, D. N. J. Am. Chem. Soc. 1998,
120, 6240. (d) Carina, R. F.; Williams, A. F.; Bernardinelli, G. Inorg. Chem.
2001, 40, 1826. (e) Lopez-Torres, M.; Fernandez, A.; Fernandez, J. J.; Suarez,
A.; Castro-Juiz, S.; Vila, J. M.; Pereira, M. T. Organometallics 2001, 20,
1350. (f) Slagt, M. Q.; van Zwieten, D. A. P.; Moerkerk, A. J. C. M.; Gebbink,
R. J. M. K.; van Koten, G. Coord. Chem. Rev. 2004, 248, 2275. (g) Lu, W.;
Roy, V. A. L.; Che, C.-M. Chem. Commun. 2006, 3972. (h) Gagliardo, M.;
Rodríguez, G.; Dam, H. H.; Lutz, M.; Spek, A. L.; Havenith, R. W. A.;
Coppo, P.; De Cola, L.; Hartl, F.; van Klink, G. P. M.; van Koten, G. Inorg.
Chem. 2006, 45, 2143.
ꢀ
995. (l) Mohr, F.; Priver, S. H.; Bhargava, S. K.; Bennett, M. A. Coord.
Chem. Rev. 2006, 250, 1851. (m) Albrecht, M. Chem. Rev. 2010, 110, 576.
(2) (a) Lai, S.-W.; Cheung, T.-C.; Chan, M. C. W.; Cheung, K.-K.; Peng,
S.-M.; Che, C.-M. Inorg. Chem. 2000, 39, 255. (b) Ma, B.; Djurovich, P. I.;
Thompson, M. E. Coord. Chem. Rev. 2005,249, 1501. (c) Ghedini, M.; Aiello, I.;
Crispini,A.;Golemme,A.;LaDeda,M.;Pucci,D.Coord. Chem. Rev. 2006,250,
1373. (d) Lowry, M. S.; Bernhard, S. Chem.—Eur. J. 2006, 12, 7970. (e) Wu,
L.-L.; Yang, C.-H.; Sun, I.-W.; Chu, S.-Y.; Kao, P.-C.; Huang, H.-H. Organome-
tallics 2007, 26, 2017. (f) Kui, S. C. F.; Sham, I. H. T.; Cheung, C. C. C.; Ma,
C.-W.; Yan, B.; Zhu, N.; Che, C.-M.; Fu, W.-F. Chem.—Eur. J. 2007, 13, 417.
(5) (a) Xiong, Z.; Wang, N.; Dai, M.; Li, A.; Chen, J.; Yang, Z. Org.
Lett. 2004, 6, 3337. (b) Frey, G. D.; Schutz, J.; Herdtweck, E.; Herrmann,
W. A. Organometallics 2005, 24, 4416. (c) Mulcahy, S. P.; Carroll, P. J.;
Meggers, E. Tetrahedron Lett. 2006, 47, 8877.
ꢀ
(g) Zucchi, G.; Maury, O.; Thuery, P.; Gumy, F.; B€unzli, J.-C. G.; Ephritikhine, M.
Chem.—Eur. J. 2009, 15, 9686.
r
pubs.acs.org/Organometallics
Published on Web 10/12/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 22, 2010 6003
various heterocyclic products.6 The majority of these organo-
metallic compounds contain a group 10 metal, with palladium
being the most versatile one,7 followed by platinum and then
nickel. Also Mn, Fe, Co, Ru, Rh, Os, and Ir are popular transi-
tion metals for cyclometalation reactions, and recently the
synthesis of the first “rollover”-cyclometalated gold complex
was reported.8 Quite a few mechanistic studies, dealing with
the formation of cyclometalated compounds, were conducted
during the last decades, addressing electronic aspects,9 the
influence of the transition metal,10 or the effect of the donor
atom that is coordinated to the metal center.11 However, mecha-
nistic studies of the so-called “rollover” cyclometalation,12
where decomplexation and rotation of a heteroaryl ring
constitute prerequisites for a metal-mediated activation of
a C-H bond, are quite rare, and there are only very few
comparative studies for cyclometalation reactions of the group
10 metals Ni, Pd, and Pt.13 Gas-phase experiments are ideally
suited to probe the intrinsic features of a chemical reaction in
contrast to solution-phase experiments, which—by definition—
are obscured by often ill-defined solvent and aggregation
effects as well as the influence of counterions. Earlier, we
have reported in quite some detail on the gas-phase generation
of “rollover”-cyclometalated [Pt(bipy - H)]þ (2-Pt; Chart 1)
starting from the corresponding platinum-methyl complexes
[Pt(CH3)(bipy)((CH3)2S)]þ and [Pt(CH3)(bipy)]þ (1-PtMe).14
In the present study, we extend this work to the nickel and
palladium complexes, 2-Ni and 2-Pd, respectively; moreover,
we were interested to probe whether or not halogens, especially
chlorine, are also adequate ligands for the formation of 2 via an
intramolecular C-H bond activation. In addition, mechanistic
Chart 1. Overview of the Relevant Structural Units
[M(X)(bipy)]þ (1 = 1-MX), [M(bipy - H)]þ (2 = 2-M), and
[M(bipy)]þ (3 = 3-M) with M = Ni, Pd, Pt and X = CH3, F, Cl,
Br, I, OAc As Well As of the Deuterated 2,20-Bipyridine Ligands
[3,30-D2]-bipy and [6,60-D2]-bipy Used in This Study
Table 1. Overview of the m/z Values That Were Chosen for Mass
Selection in Order to Produce a [D2]-Enriched [M(X)(bipy)]þ Ion
Beam (M = Ni, Pd, Pt; X = CH3, Cl, F) Together with the
Resulting [D0], [D1], and [D2] Fractions (given in %)
m/z
[D0]
[D1]
[D2]
[3,30-D2]-1-NiCl
[6,60-D2]-1-NiCl
[3,30-D2]-1-PdCl
[6,60-D2]-1-PdCl
[3,30-D2]-1-PtCl
[3,30-D2]-1-PtMe
[3,30-D2]-1-NiF
[6,60-D2]-1-NiF
251a
251a
301b
301b
391c
369d
235e
235e
2
3
1
2
0
1
1
2
3
1
3
1
3
3
0
1
95
96
96
97
97
96
99
97
a The [D2] fraction of the signal contains exclusively [D2]-1-58Ni35Cl.
b The [D2] fraction of the signal contains [D2]-1-108Pd35Cl (74%), [D2]-
1-106Pd37Cl (24%), and a small amount of 13C-containing [D2]-1-105Pd37Cl
(2%). c The [D2] fraction of the signal contains [D2]-1-198Pt35Cl (44%),
[D2]-1-196Pd37Cl (49%), and 13C-containing [D2]-1-195Pt37Cl (7%). d The
[D2] fraction of the signal contains [D2]-1-196PtMe (87%) and 13C contain-
ing [D2]-1-195PtMe (13%) (but only 1.3% [D2]-1-195Pt13CH3). e The [D2]
fraction of the signal contains exclusively [D2]-1-58NiF.
(6) (a) Pfeffer, M. Pure Appl. Chem. 1992, 64, 335. (b) Vicente, J.; Abad,
J.-A.; Bergs, R.; Jones, P. G.; Ramirez de Arellano, M. C. Organometallics
1996, 15, 1422. (c) Ferstl, W.; Sakodinskaya, I. K.; Beydoun-Sutter, N.; Le
Borgne, G.; Pfeffer, M.; Ryabov, A. D. Organometallics 1997, 16, 411. (d)
Hwang, S. J.; Cho, S. H.; Chang, S. J. Am. Chem. Soc. 2008, 130, 16158. (e)
Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc. 2009, 131, 9651.
(7) (a) Ryabov, A. D. Synthesis 1985, 233. (b) Dupont, J.; Pfeffer, M.;
Spencer, J. Eur. J. Inorg. Chem. 2001, 2001, 1917.
aspects are addressed; for example, are we dealing with an OA/
RE (oxidative addition/reductive elimination) or rather a
metathesis scenario in the “rollover” cyclometalation to gen-
erate 2 and HX for the different metals M = Ni, Pd, Pt and the
ligands X = CH3, Cl?15 Furthermore, as many synthetic proce-
dures for the generation of cyclometalated palladium com-
plexes are based on the use of palladium acetate,1h,7a,9c,16
1-PdOAc has been included in our investigations and compared
with 1-NiOAc in order to study if X = OAc can also serve
as a suitable precursor ligand for the formation of “rollover”-
cyclometalated transition-metal complexes in the gas phase.
Here, we present the results for the gas-phase fragmenta-
tions of [M(CH3)(bipy)]þ and [M(Cl)(bipy)]þ (M=Ni, Pd,
Pt; bipy = 2,20-bipyridine), [Ni(X)(bipy)]þ (X=F, Cl, Br, I),
and [M(OAc)(bipy)]þ (M = Ni, Pd) generated via electro-
spray ionization (ESI).17 2,20-Bipyridine has been chosen not
only because bipy and related heterocyclic systems represent
versatile ligands in the coordination chemistry of transition-
metal complexes18 but bipy possesses the potential to bring
about “rollover” cyclometalation.12,14 Labeling experiments
have been conducted to trace the origin of the hydrogen atom
(8) Cocco, F.; Cinellu, M. A.; Minghetti, G.; Zucca, A.; Stoccoro, S.;
Maiore, L.; Manassero, M. Organometallics 2010, 29, 1064.
(9) (a) Bruce, M. I.; Gardner, R. C. F.; Stone, F. G. A. J. Chem. Soc.,
Dalton Trans. 1976, 81. (b) Takahashi, H.; Tsuji, J. J. Organomet. Chem.
1967, 10, 511. (c) Canty, A. J.; van Koten, G. Acc. Chem. Res. 1995, 28, 406.
(d) Owen, J. S.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2004, 126,
8247. (e) Li, L.; Brennessel, W. W.; Jones, W. D. Organometallics 2009, 28,
3492.
(10) Bruce, M. I.; Goodall, B. L.; Stone, F. G. A. J. Chem. Soc.,
Chem. Commun. 1973, 558.
(11) Parshall, G. W. Acc. Chem. Res. 1970, 3, 139.
(12) (a) Wickramasinghe, W. A.; Bird, P. H.; Serpone, N. J. Chem. Soc.,
Chem. Commun. 1981, 1284. (b) Constable, E. C.; Seddon, K. R. J. Chem. Soc.,
Chem. Commun. 1982, 34. (c) Nord, G.; Hazell, A. C.; Hazell, R. G.; Farver, O.
Inorg. Chem. 1983, 22, 3429. (d) Spellane, P. J.; Watts, R. J.; Curtis, C. J. Inorg.
Chem. 1983, 22, 4060. (e) Dholakia, S.; Gillard, R. D.; Wimmer, F. L. Inorg.
Chim. Acta 1983, 69, 179. (f) Braterman, P. S.; Heath, G. A.; MacKenzie, A. J.;
Noble, B. C.; Peacock, R. D.; Yellowlees, L. J. Inorg. Chem. 1984, 23, 3425. (g)
Skapski, A. C.; Sutcliffe, V. F.; Young, G. B. J. Chem. Soc., Chem. Commun.
1985, 609. (h) Minghetti, G.; Doppiu, A.; Zucca, A.; Stoccoro, S.; Cinellu, M.;
Manassero, M.; Sansoni, M. Chem. Heterocycl. Compd. 1999, 35, 992. (i)
Zucca, A.; Doppiu, A.; Cinellu, M. A.; Stoccoro, S.; Minghetti, G.; Manassero, M.
Organometallics 2002, 21, 783. (j) Minghetti, G.; Stoccoro, S.; Cinellu, M. A.;
Soro, B.; Zucca, A. Organometallics 2003, 22, 4770. (k) Zucca, A.; Cinellu,
M. A.; Minghetti, G.; Stoccoro, S.; Manassero, C. Eur. J. Inorg. Chem. 2004,
4484. (l) Stoccoro, S.; Zucca, A.; Petretto, G. L.; Cinellu, M. A.; Minghetti, G.;
Manassero, M. J. Organomet. Chem. 2006, 691, 4135. (m) Minghetti, G.;
Stoccoro, S.; Cinellu, M. A.; Petretto, G. L.; Zucca, A. Organometallics 2008, 27,
3415. (n) Zucca, A.; Petretto, G. L.; Stoccoro, S.; Cinellu, M. A.; Manassero, M.;
Manassero, C.; Minghetti, G. Organometallics 2009, 28, 2150.
(15) For a detailed discussion of various mechanistic scenarios of gas-
ꢀ
phase organometallic transformations, see: Armelin, M.; Schlangen,
M.; Schwarz, H. Chem.—Eur. J. 2008, 14, 5229.
(16) (a) Gomez, M.; Granell, J.; Martinez, M. Organometallics 1997,
ꢀ
16, 2539. (b) Gomez, M.; Granell, J.; Martinez, M. J. Chem. Soc., Dalton
Trans. 1998, 37. (c) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. J. Am.
Chem. Soc. 2005, 127, 13754.
(17) http://nobelprize.org/chemistry/laureates/2002/.
(18) (a) Togni, A.; Venanzi, L. M. Angew. Chem., Int. Ed. Engl. 1994,
33, 497. (b) Himeda, Y.; Onozawa-Komatsuzaki, N.; Miyazawa, S.; Sugi-
hara, H.; Hirose, T.; Kasuga, K. Chem.—Eur. J. 2008, 14, 11076.
(13) Piro, N. A.; Owen, J. S.; Bercaw, J. E. Polyhedron 2004, 23, 2797.
€
(14) (a) Butschke, B.; Schlangen, M.; Schroder, D.; Schwarz, H.
€
Chem.—Eur. J. 2008, 14, 11050. (b) Butschke, B.; Schroder, D.; Schwarz,
H. Organometallics 2009, 28, 4340.

6004 Organometallics, Vol. 29, No. 22, 2010
Butschke and Schwarz
Table 2. Major Fragments in the CID of Mass-Selected [M(CH3)(bipy)]þ (1-MMe) Complexes (M = Ni, Pd, Pt) at Various Collision
Energies (Elab in eV)a
Elab
1-MMe
- CH3
- CH4
- bipy
- CH3/bipy
- MCH2
- [M,C,H3]
- [M,C,H4]
1-NiMe
1-PdMe
1-PtMe
0
5
10
15
20
25
100
24
24
41
48
72
87
100
100
100
100
100
3
14
45
8
27
0
5
10
15
20
25
100
52
16
22
41
52
22
100
100
100
100
100
2
4
3
1
>0
9
41
80
3
24
69
3
10
12
2
9
17
0
5
10
15
20
25
100
68
58
29
17
19
44
100
100
81
36
22
1
15
49
100
100
100
a Relative intensities are normalized to the base peak = 100. Numbers given in bold involve intramolecular C-H bond activation processes. Numbers
in italics indicate simple ligand-loss processes.
involved in the formation of HX. Finally, DFT-based calcu-
lations were performed to provide insight in those mechan-
istic aspects of the C-H bond activation that are not directly
accessible from the experiments.
tively, and 2,20-bipyridine in methanol were used. As [Ni(CH3)2-
(bipy)], prepared according to ref 25 starting from commercially
available [Ni(CH3)2(tmeda)], is extremely sensitive to protic solvents,
the green solid was dissolved in dry acetonitrile to give a nearly
colorless solution; ESI of this solution produced [Ni(CH3)(bipy)]þ.
For the preparation of [Pd(Cl)2(bipy)], equimolar amounts of PdCl2
and bipy were stirred at room temperature in methanol for 6 h and
the beige precipitate was filtered; [Pt(Cl)2(bipy)] was synthesized
according to ref 26. Both chloride complexes are nearly insoluble in
methanol and only slightly better in acetonitrile. However, [Pd(Cl)-
(bipy)]þ and [Pt(Cl)(bipy)]þ cannot be produced via ESI of the
acetonitrile solutions. Therefore, equimolar acetonitrile solutions of
the chloride complexes and of AgOTf were mixed in order to substi-
tute one chloride anion for the weakly coordinating triflate counter-
ion. Electrospray ionization of these solutions yielded the desired
species. [Ni(X)(bipy)]þ (X=Cl, Br, I) cations can be easily gener-
ated from methanolic solutions of [Ni(X)2(bipy)], which preci-
pitates from a mixture of equimolar ethanolic solutions of the
corresponding nickel halide and 2,20-bipyridine at room tempera-
ture. [Ni(F)2(bipy)] cannot be generated in this way; rather, [Ni(F)-
(bipy)]þ was generated via ESI from a diluted solution of NiF2 and
bipy in ethanol. ESI of millimolar solutions of [Ni(OAc)2] or [Pd-
(OAc)2] together with 2,20-bipyridine in methanol produced
[M(OAc)(bipy)]þ (M = Ni, Pd). For the deuterium-labeling experi-
ments [3,30-D2]-2,20-bipyridine ([3,30-D2]-bipy) and [6,60-D2]-2,20-
bipyridine ([6,60-D2]-bipy) were employed. [3,30-D2]-bipy was
prepared through palladium-mediated coupling (see ref 27) of
2-chloro-3-deuteropyridine synthesized according to ref 28. Based
on a mass spectrometric analysis, a labeling distribution of
[D0]:[D1]:[D2] =2:19:79 was obtained. [6,60-D2]-bipy was synthe-
sized via lithiation of 6,60-dibromo-2,20-bipyridine with n-BuLi in
THF29 and subsequent quenching of the reaction mixture with D2O,
resulting in a deuterium distribution of [D0]:[D1]:[D2] = 4:8:88. On
the basis of these isotope distributions, [D2]-[M(X)(bipy)]þ com-
plexes (M = Ni, Pd, Pt; X = CH3, Cl, F) could be produced in at
least 95% purity when the appropriate peak was mass-selected. A
summary of the precursor ions that were used for mass selection
and the deuterium incorporation for the various complexes is given
in Table 1. In the case where the mass-selected signal consisted of
Experimental and Computational Details
The present experiments were performed with a VG BIO-Q
mass spectrometer of QHQ configuration (Q: quadrupole, H:
hexapole) equipped with an ESI source as described in detail
previously.19 In brief, millimolar solutions of various precursors in
puremethanol, ethanol, or acetonitrile (see below for moredetails)
were introduced through a fused-silica capillary to the ESI source
via a syringe pump (ca. 3 μL min-1) in order to produce the ligated
metal cations under investigation. Nitrogen was used as a nebuliz-
ing and drying gas at a source temperature of 80 °C. Maximal
yields of the desired complexes were achieved by adjusting the
cone voltage (Uc) between 50 and 70 V; Uc determines the degree
of collisional activation of the incident ions in the transfer from the
ESI source tothe mass spectrometer.20 The identity of the ions was
confirmed by comparison with the expected isotope patterns.21
The isotope pattern also assisted in the choice of the adequate
precursor ion in order to avoid coincidental mass overlaps of
isobaric species in the mass-selected ion beam.22 Collision-induced
dissociation experiments were probed at different collision ener-
gies Elab between 0 and 25 eV, and the intensities of all ions were
normalized to the base-peak intensity, which was set to 100%.
For the generations of [Pd(CH3)(bipy)]þ and [Pt(CH3)(bipy)]þ
millimolar solutions of [Pd(CH3)2(tmeda)]23 (tmeda=N,N,N0,N0-
tetramethylethylenediamine) and [Pt(CH3)2(μ-(CH3)2S)]2,24 respec-
€
(19) (a) Schroder, D.; Weiske, T.; Schwarz, H. Int. J. Mass Spectrom.
€
2002, 219, 729. (b) Trage, C.; Schroder, D.; Schwarz, H. Chem.—Eur. J.
2005, 11, 619.
(20) Cech, N. B.; Enke, C. G. Mass Spectrom. Rev. 2001, 20, 362.
(21) http://www.sisweb.com/mstools/isotope.htm.
€
(22) Schroder, D.; Schwarz, H. Can. J. Chem. 2005, 83, 1936.
(23) Biscoe, M. R.; Fors, B. P.; Buchwald, S. L. J. Am. Chem. Soc.
2008, 130, 6686.
(24) Hill, G. S.; Irwin, M. J.; Levy, C. J.; Rendina, L. M.; Puddephatt,
R. J.; Andersen, R. A.; Mclean, L. In Inorganic Syntheses; Marcetta,
Y. D., Eds.; Wiley: New York, 1998; p 149.
(27) Bamfield, P.; Quan, P. M. Synthesis 1978, 537.
€
ꢀ
^
ꢀ
ꢀ
(28) Berillon, L.; Lepretre, A.; Turck, A.; Ple, N.; Queguiner, G.;
(25) Kaschube, W.; Porschke, K. R.; Wilke, G. J. Organomet. Chem.
1988, 355, 525.
(26) Egan, T. J.; Koch, K. R.; Swan, P. L.; Clarkson, C.; Van
Schalkwyk, D. A.; Smith, P. J. J. Med. Chem. 2004, 47, 2926.
Cahiezc, G.; Knochel, P. Synlett 1998, 1359.
(29) Parks, J. E.; Wagner, B. E.; Holm, R. H. J. Organomet. Chem.
1973, 56, 53.

Article
Organometallics, Vol. 29, No. 22, 2010 6005
Figure 2. Schematic PES for the fragmentation of [Pt(CH3)-
(bipy)]þ accordingto[Pt(CH3)(bipy)]þ (1-PtMe) f(bipyþ H)þ þ
PtCH2 and [Pt(CH3)(bipy)]þ (1-PtMe) f [M(bipy)]þ (3) þ CH3
based on DFT calculations using the B3LYP functional.
In the computational studies we have employed the Gaussian09
program package30 using basis sets of approximately triple-ξ
quality. For H, C, N, and Cl atoms these were the triple-ξ plus
polarization basis sets (TZVP) of Ahlrichs and co-workers.31 For
complexes containing Ni, Pd, and Pt, the Stuttgart-Dresden scalar
relativistic pseudopotential was employed in conjunction with the
corresponding basis sets.32 The fragmentation pathways for the
processes defined in eqs 1 and 2 (M=Ni, Pd, Pt; X=CH3, Cl) were
calculated with the B3LYP33 functional using the basis sets given
above. All energies (given in kJ mol-1) are corrected for (unscaled)
zero-point vibrational energy contributions. Intrinsic reaction coor-
dinate (IRC) calculations were performed to link transition-state
structures with the respective intermediates.34 The discussion of the
computational findings will be confined to the singlet state of the
various platinum and palladium cations, because exploratory
calculations revealed that their triplet states are generally much
higher in energy for all species investigated and, therefore, are not
likely to be involved in the experiments conducted at ambient
temperature. However, for the nickel complexes both the singlet
and the triplet states must be taken into account.
½MðXÞðbipyÞꢀþ ð1Þ f ½Mðbipy - HÞꢀþ ð2Þ þ HX ð1Þ
½MðXÞðbipyÞꢀþ ð1Þ f ½MðbipyÞꢀþ ð3Þ þ X
ð2Þ
(30) 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.;Nakatsuji, H.;Caricato, M.;Li, X.;Hratchian, H. P.;Izmaylov, A. F.;
Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.;
Nakai, H.; Vreven, T.; Montgomery, J. A.; , Jr., Peralta, J. E.; Ogliaro, F.;
Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.;
Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.;
Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.;
Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth,
€
G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; O.
Farkas, Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
Figure 1. Schematic PESs for the fragmentations of [M(CH3)-
(bipy)]þ (M = Ni, Pd, Pt) according to [M(CH3)(bipy)]þ
(1-MMe) f [M(bipy - H)]þ (2) þ CH4 and [M(CH3)(bipy)]þ
(1-MMe) f [M(bipy)]þ (3) þ CH3 based on DFT calculations using
the B3LYP functional. For further explanations, see Scheme 1 and text.
Revision A.1; Gaussian, Inc.: Wallingford, CT, 2009.
€
(31) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829.
€
(32) (a) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chem. Acc. 1990, 77, 123. (b) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer,
R. M. J. Phys. Chem. 1993, 97, 5852. (c) Peterson, K. A.; Figgen, D.; Dolg,
M.; Stoll, H. J. Chem. Phys. 2007, 126, 124101.
35Cl- and 37Cl-containing species, only the signals for the losses of
H37Cl and D37Cl were considered in the evaluation of the CID
results, as only for these signals can an overlap with isobaric
fragmentation products be excluded.
(33) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785. (b)
Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(34) (a) Fukui, K. J. Phys. Chem. 1970, 74, 4161. (b) Fukui, K. Acc.
Chem. Res. 1981, 14, 363. (c) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem.
1990, 94, 5523.

6006 Organometallics, Vol. 29, No. 22, 2010
Butschke and Schwarz
Figure 3. Geometry-optimized minimum structures of the species mentioned in Figures 1 and 2 together with relevant bond lengths
˚
(in A) around the metal cores as calculated with the B3LYP functional.
Results and Discussion
the fragmentation of 1-PtMe, i.e., the abstraction of the hydro-
gen atom from position C(3) of the bipy ligand after initial
decomplexation and rotation of a pyridyl ring. Upon CID of
[Ni(CH3)(bipy)]þ (1-NiMe) and of [Pd(CH3)(bipy)]þ (1-PdMe),
the most dominant process corresponds to the homolytic
cleavage of the M-CH3 bond to liberate the methyl group
(Table 2). At elevated collision energies, elimination of the bipy
ligand as well as the consecutive eliminations of CH3 and bipy
are observed. For 1-NiMe, at any collision energies employed,
CH4 is not formed at all; however, for 1-PdMe at low collision
energies a rather small amount of CH4 is lost. At higher Elab
values, the elimination of neutral [Pd,C,H4] occurs; this is also
indicative for an intramolecular hydrogen-abstraction process.
DFT calculations were performed for the processes described
in eqs 1 and 2 for the complexes [M(CH3)(bipy)]þ (M = Ni, Pd,
Pt) aimed at explaining the different fragmentation behavior,
i.e., why at low energies is CH3 preferentially lost for M=Ni and
Pd but CH4 for M=Pt? The corresponding potential-energy
Fragmentation of [M(CH3)(bipy)]þ (M = Ni, Pd, Pt). The
fragmentation behavior of cationic [Pt(CH3)(L)]þ with L =
2,20-bipyridine, 2,20-bipyrimidine (bipyrm), and 1,10-phenan-
throline (phen) was already discussed elsewhere.14b Upon CID
of [Pt(CH3)(bipy)]þ (1-PtMe) with xenon, elimination of neu-
tral CH4 occurs even at the lowest collision energies; at elevated
collision energies, PtCH2 loss gains in importance (Table 2).
In contrast, for [Pt(CH3)(bipyrm)]þ and [Pt(CH3)(phen)]þ no
liberation of CH4 takes place at the detection limit of the
instrument; these findings already indicate a genuine “rollover”
cyclometalation operative in the reaction [Pt(CH3)(bipy)]þ
f
[Pt(bipy-H)]þ þCH4 because for, L=bipyrm, no hydrogen
atoms are available at either the C(3)- or C(30)-positions, and for
L=phen, rotation around a C-C single bond is not possible.
When [3,30-D2]-1-PtMe (m/z=369) is used, the methane elimi-
nated consists of 97% CH3D, thus demonstrating the (near
exclusive) occurrence of a “rollover”-cyclometalation process in

Article
Organometallics, Vol. 29, No. 22, 2010 6007
surfaces (PESs) are given in Figure 1, and selected structures are
shown in Figure 3 (the full set of structures is available in the
Supporting Information). Starting from 1, after decomplexa-
tion, rotation of a pyridyl ring constitutes a prerequisite for the
eventual activation of the C(3)-H bond concomitant with the
formation of 2 and HX. As in the singlet-state structures of
[M(CH3)(bipy)]þ (1-MMe), for M = Ni, Pd, and Pt, the metal
center has a nearly square-planar coordination sphere (with one
vacant coordination site), there are two options: rotation of the
pyridyl ring oriented cis or trans to the methyl group (path a-cis
or path a-trans, respectively, in Scheme 1; note that in all
structural representations the upper pyridyl ring is displayed
as rotated). Due to a trans effect of the methyl group, the M-N
interaction trans to the CH3 ligand is weakened. This becomes
obvious upon inspection of the starting structures 11-NiMe, 1-
PdMe, and 1-PtMe, where the M-N bond lengths trans to the
˚
methyl group are 0.108, 0.114, and 0.183 A longer than those
M-N bonds that are alligned cis to the CH3 ligand. Therefore,
the rotation of the ring trans to the methyl group (11-NiMe f 14,
1-PdMe f 9, and 1-PtMe f 12 in Figures 1 and 3; 1 f 10-trans
in Scheme 1) is facilitated by 39, 31, and 62 kJ mol-1, respec-
tively, compared with the rotation of the ring cis to the methyl
group (leading to the 10-cis structures 15, 10, and 13, re-
spectively). For [Ni(CH3)(bipy)]þ, the triplet-state structure
31-NiMe is 34 kJ mol-1 more favorable than the singlet-state
structure 11-NiMe; due to the Cs symmetry of 31-NiMe, rota-
tions of both rings (via 3TS 1/7, see SI 1) are equivalent. After
rotation of the pyridyl ring, the metal center interacts with the
C(3)-H bond and CH4 can be formed either in a σ-bond
metathesis scenario for M = Ni and Pd or via an OA/RE
(oxidative addition/reductive elimination) mechanism for M =
Pt.15 Also in the C-H bond activation step, the trans effect of
the methyl group affects the energetic requirements. This is
further reflected in the geometric features of the 10-trans struc-
tures 14, 9, and12 as compared to the 10-cis structures 15, 10, and
13: in the 10-trans structures, the M-C(3) distance is increased
Scheme 1. Generalized Mechanistic Scheme for the Fragmenta-
tions of Singlet [M(X)(bipy)]þ (M = Ni, Pd, Pt; X = CH3, Cl)
˚
by 0.271, 0.334, and 0.488 A and, more importantly, the
C(3)-H bond distance is decreased by 0.013, 0.006, and 0.007
˚
A for the Ni-, Pd-, and Pt-containing structures, when com-
pared with the respective 10-cis structures. Thus, the interaction
of the metal center with the C(3)-H bond is weaker in the 10-
trans complexes, and C-H bond activation via 1TS 4/6, TS 9/
11, and TS 12/14 is energetically more demanding by 171, 173,
and 119 kJ mol-1 for M = Ni, Pd, and Pt, respectively, as
compared with the corresponding 10-cis structures. For the
C-H bond activation step in the fragmentation of the triplet-
state species 31-NiMe, only one transition state, 3TS 7/8 (SI 1),
could be located. However, for the fragmentation of [Ni(CH3)-
(bipy)]þ all pathways (including singlet and triplet states) that
lead to the elimination of methane are at least 65 kJ mol-1
higher in energy than the direct homolytic cleavage of the
Table 3. Major FragmentsintheCID ofMass-Selected[M(Cl)(bipy)]þ (1-MCl) (M = Ni, Pd, Pt) at VariousCollision Energies (Elab ineV)a
Elab
1-MCl
- Cl
- HCl
- HCl/HCN
- Cl/bipy
- [M,Cl]
- [M,H,Cl]
- [M,H,Cl]/HCN
- C5H4NCl
1-NiCl
1-PdCl
1-PtCl
0
5
10
15
20
25
100
100
100
100
100
100
>0
2
28
53
62
51
> 0
5
25
27
19
14
2
7
14
24
2
10
21
28
6
19
0
5
10
15
20
25
100
92
42
53
84
73
23
100
100
100
87
6
19
48
62
60
1
10
52
100
100
5
17
29
6
17
25
72
0
5
10
15
20
25
100
100
73
42
30
13
99
100
100
100
100
5
11
17
3
8
25
4
26
55
a Relative intensities are normalized to the base peak = 100. Numbers given in bold involve intramolecular C-H bond activation processes. Numbers
in italics indicate simple ligand-loss processes.

6008 Organometallics, Vol. 29, No. 22, 2010
Butschke and Schwarz
Figure 4. Relative intensities for the losses of HCl and Cl in the CIDs of mass-selected (a) [Ni(Cl)(bipy)]þ and (b) [Pd(Cl)(bipy)]þ at
variable collision energies (Elab in eV).
Ni-CH3 bond, producing CH3 and doublet 23-Ni (the quartet-
Table 4. Positions of the Hydrogen Atoms in bipy That Are
Involved in the Formation of HCl upon CID of 1-NiCl and 1-PdCl
state structure 43-Ni is even 131 kJ mol-1 higher in energy than
23-Ni). These computational findings are in good agreement
at Various Collision Energies Elab Based on the Data Shown in SI
with the breakdown behavior of 1-NiMe, as only CH3 (but no
CH4) is lost upon CID. Also for [Pt(CH3)(bipy)]þ, the experi-
8 without Accounting for Possible Kinetic Isotope Effectsa
position
5 eV
10 eV
15 eV
20 eV
mental results are in keeping with the DFT results: elimination
of CH4 (via path a-cis, b-cis; Scheme 1) is predicted to be
58 kJ mol-1 more favored than direct cleavage of the Pt-CH3
bond. The fact that even at higher collision energies no elimina-
tion of CH3 is observed is due to the existence of an efficient
channel for the competitive formation of neutral PtCH2, as
shown in Figure 2 (selected structures for this path are presented
in Figure 3).35 For [Pd(CH3)(bipy)]þ, the C-H bond activation
via TS 10/11 (SI 2) corresponds to the rate-determining step in
the formation of CH4, which, according to the DFT results, is
located only slightly above the methyl-elimination products
3-Pd and CH3. Within error bars, this is also reflected in the
experimental results, as (minor) CH4 elimination is observed
only at low collision energies; due to a kinetic hindrance in the
elimination of CH4, this channel decreases when the collision
energy is increased.36
1-NiCl
1-PdCl
3
4, 5
6
67
18
15
64
18
18
60
19
21
3
4, 5
6
85
10
5
79
15
6
75
17
8
68
19
13
P
a The distributions are normalized to
= 100%.
(bipy)]þ undergo intramolecular C-H bond activation con-
comitant with the elimination of HCl upon collision with
xenon. In addition, homolytic cleavage of the M-Cl bond
(M = Ni, Pd) takes place, and the ratio for the losses of HCl
and Cl decreases with increasing collision energy for both
complexes (Figure 4). This is a consequence of the fact that
the formation of HCl is connected with several (tight) transi-
tion states (and is therefore kinetically hindered at higher
collision energies), whereas scission of the M-Cl bond
proceeds through a loose transition state.36 On the basis of
the branching ratios for HCl versus Cl formation, C-H
bond activation is more efficient for the palladium than for
the nickel complex, as clearly shown in Figure 4; this
becomes also obvious from the fact that the signal for HCl
loss in the CID spectrum of 1-PdCl amounts to the base peak
at Elab = 5, 10, and 15 eV (Table 3). As to the regioselectivi-
ties of C-H bond activation associated with HCl loss, in the
fragmentations of 1-NiCl and 1-PdCl the selectivity is not as
high as for 1-PtCl. This becomes obvious upon inspection of
Table 4, which shows the C-H positions of bipy that are
involved in HCl formation upon CID of 1-NiCl and 1-PdCl
(the data given in the table are based on the ratios for the
losses of HCl and DCl in the CID spectra of mass-selected
[3,30-D2]-1-NiCl, [6,60-D2]-1-NiCl, [3,30-D2]-1-PdCl, and
[6,60-D2]-1-PdCl at various collision energies Elab; the com-
plete set of data is given in the Supporting Information). For
1-NiCl, abstraction of a hydrogen atom from position C(3),
to form HCl, is below 70%; for 1-PdCl it is below 90%, and
upon increasing the collision energy, the selectivity for the
abstraction of a hydrogen atom from a particular position of
the bipy ligand decreases even further.
Fragmentation of [M(Cl)(bipy)]þ (M = Ni, Pd, Pt). Similar
to the fragmentation of the methyl complex [Pt(CH3)-
(bipy)]þ, upon CID of [Pt(Cl)(bipy)]þ only neutral fragments
are lost that contain a hydrogen atom from the bipy ligand
(data in bold in Table 3). The most dominant channel
corresponds to the formation of HCl, but at higher collision
energies also the consecutive eliminations of HCl and HCN
(due to the fragmentation of the (bipy - H) ligand), the
generation of neutral [Pt,H,Cl], and the consecutive elimina-
tions of [Pt,H,Cl] and HCN take place. No homolytic
cleavage of the Pt-Cl bond is observed. In the fragmentation
of [3,30-D2]-1-PtCl, HCl and DCl are formed in a ratio of
3:97 (m/z = 391 f [D0]:[D1]:[D2] = 0:3:97); this demon-
strates that a hydrogen atom bound to C(3) or C(30) serves as
the exclusive source for the formation of hydrogen chloride
(“rollover” cyclometalation).
In contrast to [Ni(CH3)(bipy)]þ and [Pd(CH3)(bipy)]þ, the
analogous chloride complexes [Ni(Cl)(bipy)]þ and [Pd(Cl)-
(35) The analogous processes for the complexes 1-MMe (M = Ni,
Pd), e.g., the formations of 3NiCH2 þ (bipy þ H)þ and PdCH2 þ (bipy þ
H)þ, require 373 and 279 kJ/mol, respectively; therefore, no competition
exists with the losses of a methyl group from these complexes.
€
ꢀ
(36) Schroder, D.; Roithova, J.; Alikhani, E.; Kwapien, K.; Sauer, J.
Chem.—Eur. J. 2010, 16, 4110.

Article
Apart from the losses of HCl and Cl from 1-NiCl and
1-PdCl, elimination of neutral [M,H,Cl] as well as the com-
bined losses of Cl and the bipy ligand is observed at increased
collision energies. For 1-PdCl, at higher collision energies
neutral [Pd,Cl] is formed, and for 1-NiCl the elimination of
C5H4NCl points to cleavage of the C(2)-C(20) bond of the
heterocyclic ligand.
Organometallics, Vol. 29, No. 22, 2010 6009
Pt into the C(3)-H bond is energetically rather easy (compare
TS 13/14 and TS 28/29); especially, the barrier for 13 f 14
amounts to only 2 kJ mol-1. In sharp contrast, barriers of at
We have also performed DFT calculations for the processes
defined in eqs 1 and 2 for the chloro complexes [M(Cl)(bipy)]þ
(M = Ni, Pd, Pt). The potential energy surfaces are shown in
Figure 5, and selected structures are displayed in Figure 6 (the
full set of structures is given in the Supporting Information). For
the singlet states of 1-NiCl and 1-PdCl, due to the trans effect of
the Cl ligand, initial rotation of the pyridyl ring trans to M-Cl
1
(producing 18 and 23) is facilitated by 15 and 12 kJ mol-1
,
respectively, compared with the rotation of the ring cis to the
M-Cl bond; the latter leads to 119 and 24 (the M-Nbondcis to
Cl in 1-MCl is 0.035 and 0.042 A shorter than that trans to Cl for
˚
M = Ni and Pd, respectively). However, the trans effect of a Cl
ligand is weaker than that of a methyl group (Figures 1 and 3).
For 1-PtCl, only a transition state for the rotation of the pyridyl
ring trans to the chloro ligand (TS 1/27) could be located.
Interestingly, and in contrast with the methyl complexes
1-MMe (M=Ni,Pd,Pt),the10-trans structures (Scheme 1) 118,
23, and 27 can easily isomerize to the 10-cis structures 119, 24,
and 28 (for the methyl complexes such a transition state could be
located only for M = Pt; TS 12/13; 219 kJ mol-1); only these
10-cis intermediates give rise to the formations of the HCl com-
plexes 120, 26, and 30. For the transformation 119 f 120 (M =
Ni), on the basis of the DFT findings a metathesis mechanism
applies and the sequence 28 f 29 f 30 (M = Pt) proceeds via
an OA/RE scenario. For M = Pd, transition states for both
mechanisms could be located and the σ-bond metathesis path-
way (24 f26) is24kJmol-1 higher in energy than the transition
state for an oxidative addition (24 f 25). The triplet-state
structure 31-NiCl is located 63 kJ mol-1 below the singlet-state
1
3
species 1-NiCl, and similar to 1-NiMe, only one transition
state (3TS 1/21) for the pyridyl-ring rotation and one (3TS 21/
22) for the formation of the HCl complex 322 via a metathesis
pathway could be located.
However, the demands for the formation of HCl via “roll-
over” cyclometalation in both spin states are comparable, and
HCl formation is predicted to be ca. 30 kJ mol-1 less
favored than the formation of doublet 23-Ni concomitant with
homolytic cleavage of the Ni-Cl bond; within error bars, this is
reflected by the experimental results as both processes compete,
with loss of the Cl ligand becoming dominant at higher Elab
values (Figure 4). The calculations on the fragmentation of
[Pd(Cl)(bipy)]þ predict HCl formation, when generated via a
“rollover”-cyclometalation process, to be ca. 50 kJ mol-1 more
favorable than direct Cl loss; moreover, the overall energy
demand to bring about fragmentation is more than 100 kJ mol-1
lower for M = Pd than for M = Ni. These differences explain
why intramolecular C-H bond activation is more efficient for 1-
PdCl than for 1-NiCl and why 1-PdCl already loses HCl at quite
low collision energies (with only a small amount of Cl being for-
med); in contrast, 1-NiCl eliminates HCl and Cl only at more ele-
vated Elab values. Also the fact that upon CID of [Pt(Cl)(bipy)]þ
no loss of Cl occurs is clearly reflected in the calculated PES, as
the bond-dissociation energy of the Pt-Cl bond is calculated to
be more than 120 kJ mol-1 higher than the energy required for
the formation of HCl via “rollover” cyclometalation.
Figure 5. Schematic PESs for the fragmentations of [M(Cl)-
(bipy)]þ (M = Ni, Pd, Pt) according to [M(Cl)(bipy)]þ (1-MCl) f
[M(bipy - H)]þ (2) þ HCl and [M(Cl)(bipy)]þ (1-MCl) f
[M(bipy)]þ (3) þ Cl based on DFT calculations using the B3LYP
functional. For further explanations, see Scheme 1 and text.
There is one interesting feature that deserves to be mentioned
in passing. For both 1-PtMe and 1-PtCl, oxidative addition of

6010 Organometallics, Vol. 29, No. 22, 2010
Butschke and Schwarz
˚
Figure 6. Geometry-optimized minimum structures of the species mentioned in Figure 5 together with relevant bond lengths (in A)
around the metal cores as calculated with the B3LYP functional.
least 54 kJ mol-1 have to be overcome in the C(3)-H bond
activation steps for the systems 1-NiMe and 1-NiCl as well as for
1-PdMe and 1-PdCl, respectively.
formation of HF: [3,30-D2]-1-NiF loses HF and DF in a ratio of
96:4; in contrast, CID of [6,60-D2]-1-NiF at 10 eV results in the
eliminations of HF and DF in a ratio of 8:92. This translates
into abstraction of hydrogen atoms from positions 3, 4 or 5, and
6 in a ratio of 4:4:92. Obviously, near to no “rollover” cyclom-
etalation occurs for the Ni complex, but the C-H bond atom
next to the nitrogen atom of bipy is attacked preferentially. We
have initiated computational investigations on mechanisms
that could possibly explain the highly selective abstraction of
a hydrogen atom from position C(6) of bipyridine in the
fragmentation of 1-NiF; however, we did not yet succeed in
finding a completely satisfying potential-energy surface that
accounts for all experimental observations.
Fragmentation of [Ni(X)(bipy)]þ (X = F, Cl, Br, I). As
already discussed above, [Ni(Cl)(bipy)]þ eliminates HCl upon
collision with Xe, while in the CID experiments with [Ni(CH3)-
(bipy)]þ no intramolecular C-H bond activation is observed at
all. Therefore, we were interested in the fragmentation patterns
of analogous nickel-halogen complexes subjected to similar
conditions. The results are summarized in Table 5. Only 1-NiF
and 1-NiCl lose HX upon CID. This process as well as the
elimination of neutral [Ni,H,X] is slightly more efficient for
1-NiF than for 1-NiCl. Moreover, in contrast to 1-NiCl, 1-NiF
does not undergo homolytic cleavage of the Ni-F bond upon
collision. Common to both complexes is the formation of
C5H4NX (X = F, Cl), which, however, is more pronounced
for X = F. Interestingly, 1-NiF exhibits a high selectivity in the
Upon CID of 1-NiBr and 1-NiI, no products due to C-H
bond activation are observed. Obviously the Ni-X bonds
(X = Br, I) are so weak that the sequence of decomplexation,
rotation of a pyridyl ring, and C-H bond activation cannot

Article
Organometallics, Vol. 29, No. 22, 2010 6011
Table 5. Major Fragments in the CID of Mass-Selected
[Ni(X)(bipy)]þ (1-NiX) Complexes (X = F, Cl, Br, I) at Various
Collision Energies (Elab in eV)a
appropriate ligand to bring about “rollover” cyclometala-
tion in the gas phase. However, as shown in Table 6, acetate
is too weakly bound to the metal center in 1-NiOAc and
1-PdOAc, and upon CID, homolytic cleavage of the
M-OAc bond prevails. No C-H bond activation processes
are observed. At slightly higher collision energies, the acetate
complexes lose neutral CO2 and, quite likely, the corresponding
methyl complexes 1-NiMe and 1-PdMe are concomitantly
formed (for analogous examples see ref 37).
Elab 1-NiX - X - HX - X/bipy - [Ni,H,X] - C5H4NX
1-NiF
1-NiCl
1-NiBr
1-NiI
0
5
100
100
100
100
100
100
1
13
55
50
31
18
10
15
20
25
8
18
25
28
6
38
76
83
Conclusions
0
5
100
100
100
100
100
100
>0
2
>0
5
The branching ratios and the mechþanistic details of the gas-
phase fragmentations of [M(X)(bipy)] (bipy=2,20-bipyridine)
complexes depend strongly on the nature of the metal M (M =
Ni, Pd, Pt) as well as that of the ligand X (X=CH3, F, Cl, Br, I,
OAc). Only the [Pt(CH3)(bipy)]þ and [Pt(Cl)(bipy)]þ com-
plexes selectively and exclusively undergo losses of HX (X =
CH3, Cl), resulting in the formation of “rollover”-cyclometa-
lated [Pt(bipy - H)]þ. Upon low-energy collisions with xenon,
[Ni(CH3)(bipy)]þ and [Pd(CH3)(bipy)]þ undergo homolytic
cleavage of the M-CH3 bond, and also from [Ni(Br)(bipy)]þ,
[Ni(I)(bipy)]þ, [Ni(OAc)(bipy)]þ, and [Pd(OAc)(bipy)]þ the
neutral nonheterocyclic ligand is eliminated as a radical X
(X = Br, I, OAc); any C-H bond activation processes are
absent. However, the scenarios for the fragmentations of
[Pd(Cl)(bipy)]þ, [Ni(Cl)(bipy)]þ, and [Ni(F)(bipy)]þ are more
subtle. For all three substrates, in addition to the homolytic
cleavage of the M-X bond observed for [Pd(Cl)(bipy)]þ and
[Ni(Cl)(bipy)]þ, HX (X = F, Cl) is formed upon CID; however,
the site selectivity for the hydrogen-atom abstraction from the
bipy ligand varies: for [Pd(Cl)(bipy)]þ and [Ni(Cl)(bipy)]þ the
hydrogen atom originates predominantly from position C(3),
while for [Ni(F)(bipy)]þ position C(6) serves as the main
hydrogen source. The product distributions in the fragmenta-
tions of [M(X)(bipy)]þ with M=Ni, Pd, Pt and X=CH3, Cl are
quite straightforwardly explained by DFT calculations. The
trans effect of the nonheterocyclic ligand (especially for X =
CH3) has a strong influence on the preferred pathway in the
course of the “rollover”-cyclometalation scenario. A metathesis-
like mechanism is predicted to be operative for M=Ni, and an
OA/RE pathway for M = Pt. For palladium, the preferred
mechanism depends on the nature of the ligand X, and for X=
Cl, an OA/RE pathway is slightly preferred energetically.
10
15
20
25
28
53
62
51
25
27
19
14
2
7
2
10
21
28
6
19
14
24
0
5
100
100
72
9
35
10
15
20
25
100
100
100
100
30
36
6
24
52
0
5
100
74
30
11
12
20
9
100
100
100
100
100
10
15
20
25
2
8
a Relative intensities are normalized to the base peak =100. Numbers
given in bold involve intramolecular C-H bond activation processes.
Numbers in italics indicate simple ligand-loss processes.
Table 6. Major Fragments in the CID of Mass-Selected
[M(OAc)(bipy)]þ (1-MOAc) (M = Ni, Pd) at Various Collision
Energies (Elab in eV)a
Elab
1-MOAc
- AcO•
- AcO•/bipy
- CO2
1-NiOAc
0
5
10
15
20
25
100
100
93
38
43
>0
11
100
100
100
100
>0
6
12
4
2
2
40
1-PdOAc
0
5
10
15
20
25
100
100
74
28
32
1
33
100
100
100
100
5
29
29
11
8
Acknowledgment. We thank Dr. Xinhao Zhang for
enlightening discussions and are grateful for financial
support to the Fonds der Chemischen Industrie as well
as the Cluster of Excellence “Unifying Concepts in Cata-
€
lysis” coordinated by the Technische Universitat Berlin
and funded by the Deutsche Forschungsgemeinschaft.
8
20
39
8
a Relative intensities are normalized to the base peak = 100. Numbers
in italics indicate simple ligand-loss processes.
compete with the direct ligand loss, being the only event
observed at moderate collision energies. At higher collision
energies, the consecutive eliminations of X and bipy take place.
As expected, at a given energy, loss of X from [Ni(X)(bipy)]þ
is more dominant for X = I than for X = Br.
Supporting Information Available: A summary of all com-
puted structures is available free of charge via the Internet at
http://pubs.acs.org.
Fragmentation of [M(OAc)(bipy)]þ (M = Ni, Pd). Palla-
dium acetate is frequently used in synthetic protocols to
generate cyclometalated palladium complexes. Therefore,
we were interested to see if acetate could also serve as an
(37) (a) Rijs, N.; Khairallah, G. N.; Waters, T.; O’Hair, R. A. J.
J. Am. Chem. Soc. 2008, 130, 1069. (b) Rijs, N. J.; O'Hair, R. A. J.
Organometallics 2009, 28, 2684. (c) Rijs, N. J.; Sanvido, G. B.; Khairallah,
G. N.; O'Hair, R. A. J. Dalton Trans. 2010, 39, 8655.