Interaction of organoplatinum(II) complexes with monovalent coinage metal triflates

Article abstract of DOI:10.1021/ja900449y

The organoplatinum(II) complexes [(NN)PtMe₂] and [(NN)PtPh ₂] (NN ) ArNC(Me)C(Me)NAr, Ar = 2,6-dichlorophenyl) can act as donor ligands for copper(I) and silver(I) triflates, affording a series of homo- and heteroleptic complexes which were characterized by X-ray diffraction. [(NN)PtMe₂] binds to the coinage metals through short, ligand-unsupported d-d¹⁰ contacts that are best described as Pt→M dative bonds (M ) Cu, Ag), in which the d_z² orbital of the square-planar Pt(II) center donates electron density to the Lewis-acidic metal. Spectroscopic studies in solution and DFT calculations corroborate this description. [(NN)PtPh₂] binds preferentially by η¹ or η² complexation of the ipso carbon atoms of the phenyl groups to the coinage metal, affording homoleptic complexes {[(NN)PtPh₂]₂M}⁺{TfO}^- in the solid state. The 1:1 adducts of formula {[(NN)PtPh₂]M(OTf)}_n (M ) Cu, n ) 1; M ) Ag, n ) 2) are observed in solution, and a 1:2 adduct of formula {[(NN)PtPh₂]Ag₂(OTf)₂(C ₆H₆)}_n was characterized in the solid state, showing that unsupported Pt→M bonds are also accessible for [(NN)PtPh ₂]. The thermolyses of the complexes [(NN)PtMe₂]MOTf in benzene affords moderate yields of [(NN)PtPh₂] through an oxidatively induced double C-H activation process.

Full text of DOI:10.1021/ja900449y

Published on Web 03/30/2009
Interaction of Organoplatinum(II) Complexes with Monovalent
Coinage Metal Triflates
Marc-Etienne Moret and Peter Chen*
Laboratorium fu¨r Organische Chemie, Eidgeno¨ssische Technische Hochschule (ETH) Zu¨rich,
CH-8093 Zu¨rich, Switzerland
Received January 20, 2009; E-mail: peter.chen@org.chem.ethz.ch
Abstract: The organoplatinum(II) complexes [(NN)PtMe2] and [(NN)PtPh2] (NN ) ArNC(Me)C(Me)NAr, Ar
) 2,6-dichlorophenyl) can act as donor ligands for copper(I) and silver(I) triflates, affording a series of
homo- and heteroleptic complexes which were characterized by X-ray diffraction. [(NN)PtMe2] binds to the
coinage metals through short, ligand-unsupported d-d10 contacts that are best described as PtfM dative
2
bonds (M ) Cu, Ag), in which the dz orbital of the square-planar Pt(II) center donates electron density to
the Lewis-acidic metal. Spectroscopic studies in solution and DFT calculations corroborate this description.
[(NN)PtPh2] binds preferentially by η1 or η2 complexation of the ipso carbon atoms of the phenyl groups to
the coinage metal, affording homoleptic complexes {[(NN)PtPh2]2M}+{TfO}- in the solid state. The 1:1 adducts
of formula {[(NN)PtPh2]M(OTf)}n (M ) Cu, n ) 1; M ) Ag, n ) 2) are observed in solution, and a 1:2
adduct of formula {[(NN)PtPh2]Ag2(OTf)2(C6H6)}n was characterized in the solid state, showing that
unsupported PtfM bonds are also accessible for [(NN)PtPh2]. The thermolyses of the complexes
[(NN)PtMe2]MOTf in benzene affords moderate yields of [(NN)PtPh2] through an oxidatively induced double
C-H activation process.
A particular case of d8-d10 interactions is constituted by the
Introduction
bonds formed between an electron-rich platinum(II) center and
a Lewis-acidic coinage metal, which are thought to have a
significant donor-acceptor character and thus are often referred
to as PtfM dative bonds. This kind of bonding has been found
in a broad range of compounds for M ) Ag(I),9,19-34 but is
also known for M ) Au(I).9,33-36 For example, Yamagushi et
al. 29 reported an interesting one-dimensional metal chain built
from PtfAg bonds, and compounds featuring PtfAg bonds
Metallophilic bonds, i.e. attractive interactions between closed
shell transition metals, are a topic of increasing interest, due
both to the theoretical challenge of accurately describing
them,1-6 and to their practical utility. The latter includes their
use as a tool for crystal engineering7,8 and for the design of
luminescent compounds.9-11 Square-planar d8 complexes can
be regarded as closed-shell species due to the strong ligand-
field splitting of the d orbitals,1 and thus can be involved in
metallophilic bonds. In fact, d8-d10 interactions have been found
in a variety of compounds.12-18
(15) Balch, A. L.; Catalano, V. J.; Olmstead, M. M. Inorg. Chem. 1990,
29, 585–586.
(16) Kim, M.; Taylor, T. J.; Gabbai, F. P. J. Am. Chem. Soc. 2008, 130,
6332–6333.
(1) Pyykko¨, P. Chem. ReV. 1997, 97, 597–636.
(2) Hermann, H. L.; Boche, G.; Schwerdtfeger, P. Chem.-Eur. J. 2001,
7, 5333–5342.
(17) Kickham, J. E.; Loeb, S. J. Organometallics 1995, 14, 3584–3587.
(18) Yamaguchi, T.; Yamazaki, F.; Ito, T J. Am. Chem. Soc. 1999, 121,
7405–7406.
(3) Xia, B.-H.; Zhang, H.-X.; Jiao, Y.-Q.; Pan, Q.-J.; Li, Z.-S.; Sun, C.-
C. J. Chem. Phys. 2004, 120, 11487–11492.
(19) Baudron, S. A.; Hosseini, M. W. Chem. Commun. 2008, 4558–4560.
(20) Janzen, D. E.; Mehne, L. F.; vanDerveer, D. G.; Grant, G. J. Inorg.
Chem. 2005, 44, 8182–8184.
(4) Pyykko¨, P.; Patzschke, M. Faraday Discuss. 2003, 124, 41–51.
(5) Zhang, J.-P.; Wang, Y.-B.; Huang, X.-C.; Lin, Y.-Y.; Chen, X.-M.
Chem.-Eur. J. 2005, 11, 552–561.
(21) Song, H. B.; Zhang, Z.-Z.; Hui, Z.; Che, C.-M.; Mak, T. C. W. Inorg.
Chem. 2002, 41, 3146–3154.
(6) Poblet, J.-M.; Be´nard, M. Chem. Commun. 1998, 1179–1180.
(7) Katz, M. J.; Sakai, K.; Leznoff, D. B. Chem. Soc. ReV. 2008, 37, 1884–
1895.
(22) Kampf, G.; Miguel, P. J. S.; Willermann, M.; Schneider, A.; Lippert,
B. Chem.-Eur. J. 2008, 14, 6882–6891.
(8) Zheng, X.-D.; Jiang, L.; Feng, X.-L.; Lu, T.-B. Inorg. Chem. 2008,
47.
(23) Fornie´s, J.; Mart´ınez, F.; Navarro, R.; Urriolabeitia, E. P. Organo-
metallics 1996, 15, 1813–1819.
(9) Yip, H.-K.; Lin, H.-M.; Cheung, K.-K.; Che, C.-M.; Wang, Y. Inorg.
Chem. 1994, 33, 1644–1651.
(24) Uso´n, R.; Fornie´s, J.; Menjo´n, B.; Cotton, F. A.; Falvello, L. R.; Toma´s,
M. Inorg. Chem. 1985, 24, 4651–4656.
(10) Gil, B.; Fornie´s, J.; Go´mez, J.; Mart´ın, A.; Moreno, M. T. Inorg. Chem.
2006, 45, 7788–7798.
(25) Uson, R.; Fornie´s, J.; Toma´s, M.; Cotton, F. A.; Falvello, L. R. J. Am.
Chem. Soc. 1984, 106, 2482–2483.
(11) Ferna´ndez, E. J.; Laguna, A.; de Luzuriaga, J. M. L. Dalton Trans.
2007, 1969–1981.
(26) Fornie´s, J.; Navarro, R.; Toma´s, M.; Urriolabeitia, E. P. Organome-
tallics 1993, 12, 940–943.
(12) Balch, A. L.; Catalano, V. J. Inorg. Chem. 1991, 30, 1302–1308.
(13) Heckenroth, M.; Kluser, E.; Neels, A.; Albrecht, M. Angew. Chem.,
Int. Ed. 2007, 46, 6293–6296.
(27) Fornie´s, J.; Fortun˜o, C.; Iba´n˜ez, S.; Mart´ın, A. Inorg. Chem. 2008,
47, 5978–5987.
(28) Fornie´s, J.; Iba´n˜ez, S.; Mart´ın, A.; Sanz, M.; Berenguer, J. R.; Lalinde,
E.; Torroba, J. Organometallics 2006, 25, 4331–4340.
(29) Yamagushi, T.; Yamazaki, F.; Ito, T. J. Am. Chem. Soc. 2001, 123,
743–744.
(14) Crespo, O.; Laguna, A.; Ferna´ndez, E. J.; de Luzuriaga, J. M. L.; Jones,
P. G.; Teichert, M.; Monge, M.; Pyykko¨, P.; Runeberg, N.; Schu¨tz,
M.; Werner, H. J. Inorg. Chem. 2000, 39, 4786–4792.
9
10.1021/ja900449y CCC: $40.75
2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 5675–5690 5675

A R T I C L E S
Moret and Chen
Scheme 1
have been found to be intermediates for halogen abstraction
reactions on platinum(II).27 Of particular relevance to this work
is the observation by the group of Puddephatt34 of the sandwich
complex {[(bpy)PtMe2]2Ag}+{BF4}- (bpy ) 2,2′-bipyridyl) at
low temperature. This complex was found to undergo reversible
Pt-Me bond cleavage at -40 °C, but decomposes with
precipitation of Ag(0) when allowed to warm to room temper-
ature. Only a few compounds are known that incorporate
PtfCu(I) bonds,19,21,33,35 and to our knowledge, the metal-metal
interaction is supported by bridging ligands in all cases. We
recently reported compounds featuring a Pt2Cu2 core in which
the central copper(I) atom is held only by a cuprophilic
interaction and two PtfCu bonds.37
accessible through the reaction of 1 with suitable precursors
(Scheme 1). We set out to study the interaction of 2 and 3 with
coinage metal triflates with a two-fold objective: (1) to access
structurally simple compounds featuring PtfAg and PtfCu
bonds in order to allow a better characterization of this kind of
binding, and (2) to probe whether these interactions would alter
the chemical properties of the Pt-R bonds and give access to
novel reactivity in relation with C-H bond activation. We
reasoned that the ortho-dichlorophenyl groups of ligand 1 would
at the same time offer some steric protection against aggregation
and, by withdrawing electron density from the diimine moiety,
render the complexes more robust toward the above-mentioned
oxidative decomposition reported by Puddephatt.34
In this contribution, we first report the solid-state character-
ization of a series of compounds featuring ligand-unsupported
PtfCu and PtfAg bonds. Then, we present the study of these
interactions in solution by a range of spectroscopic techniques,
and discuss their nature on the basis of density functional theory
(DFT) calculations. Finally, we describe an oxidatively induced
double C-H activation reaction which transforms 2 into 3 in
the presence of CuOTf or AgOTf in benzene.
Compounds of general formula [(NN)PtR2] (NN ) R-diimine
ligand, R ) Me, Ph) have attracted much attention because of
their use as precursors in C-H activation reactions.38-51 Upon
activation with a Brønsted acid, they eliminate one equivalent
of methane or benzene and generate active species with a weakly
bound solvent molecule or counteranion that can be exchanged
for an alkane or arene in the first step of the C-H activation
process. A subsequent oxidative addition/reductive elimination
sequence then generates methyl and phenyl complexes from
methane and benzene, respectively. A slightly more complex
reaction yields dehydrogenation products from higher alkanes.52
The initial ligand exchange is often rate determining,46 and thus
the absence of strong ligands in the reaction mixture is critical.
The R-diimine ligand 1 (Scheme 1) has previously been used
in our group to stabilize reactive intermediates relevant to C-H
bond activation in mass spectrometric studies.53-56 The dimeth-
yl-53 and diphenylplatinum(II)56 complexes 2 and 3 are readily
(30) Uson, R.; Fornies, J.; Tomas, M.; Casas, J. M.; Cotton, F. A.; Falvello,
L. R.; Llusar, R. Organometallics 1988, 7, 2279–2285.
(31) Uso´n, R.; Fornie´s, J.; Toma´s, M.; Ara, I.; Casas, J. M.; Martin, A.
J. Chem. Soc., Dalton Trans. 1991, 2253–2264.
Results and Discussion
Solid-State Studies. The interaction of the dimethylplatinu-
m(II) complex 2 and its diphenyl analogue 3 with monovalent
coinage metal triflates was first studied by preparing crystals
of the corresponding adducts and analyzing them by X-ray
diffraction (XRD). The compounds involving 2 are discussed
first.
(32) Uso´n, R.; Fornie´s, J. Inorg. Chim. Acta 1992, 198-200, 165–177.
(33) Xia, B.-H.; Zhang, H.-X.; Che, C.-M.; Leung, K.-H.; Phillips, D.-L.;
Zhu, N.; Zhou, Z.-Y. J. Am. Chem. Soc. 2003, 125, 10362–10374.
(34) Arsenault, G. J.; Anderson, C. M.; Puddephatt, R. J. Organometallics
1988, 7, 2094–2097.
(35) Yin, G.-Q.; Wei, Q.-H.; Zhang, L.-Y.; Chen, Z.-N. Organometallics
2006, 25, 580–587.
(36) Yip, H.-K.; Lin, H.-M.; Wang, Y.; Che, C.-M. J. Chem. Soc., Dalton
Trans. 1993, 2939–2944.
The addition of CuOTf ·1/2C7H8 to a dark-green solution of
2 in toluene yielded a clear, dark-red solution from which the
1:1 adduct 4a could be isolated in 46% yield (Scheme 2).
Compound 4a slowly decomposes at room temperature in
aromatic solvents (Vide infra), but crystals suitable for XRD
could be obtained from a benzene/hexane mixture at -35 °C.
The most striking feature in the crystal structure of 4a (Figure
1) is the ligand-unsupported Pt-Cu bond. With 2.3992(16) Å,
it is markedly shorter than sum of the covalent radii (1.36(Pt)
+ 1.32(Cu) ) 2.68 Å),57 and, to the best of our knowledge, it
is the shortest Pt-Cu contact reported to date. The Pt-Cu bond
makes an angle of 81° with the coordination plane of the
platinum center, which supports its description as a PtfCu
dative bond involving donation of electron density from the filled
(37) Moret, M.-E.; Chen, P. Organometallics 2008, 27, 4903–4916.
(38) Johannsson, L.; Ryan, O. B.; Tilset, M. J. Am. Chem. Soc. 1999, 121,
1974–1975.
(39) Wik, B. J.; Lersch, M.; Krivokapic, A.; Tilset, M. J. Am. Chem. Soc.
2006, 128, 2682–2696.
(40) Heiberg, H.; Johansson, L.; Gropen, O.; Ryan, O. B.; Swang, O.; Tilset,
M. J. Am. Chem. Soc. 2000, 122, 10831–10845.
(41) Lersch, M.; Tilset, M. Chem. ReV. 2005, 105, 2471–2526.
(42) Johansson, L.; Ryan, O. B.; Rømming, C.; Tilset, M. J. Am. Chem.
Soc. 2001, 123, 6579–6590.
(43) Gerdes, G.; Chen, P. Organometallics 2004, 23, 3031–3036.
(44) Gerdes, G.; Chen, P. Organometallics 2006, 25, 809–811.
(45) Labinger, J. A.; Bercaw, J. E.; Tilset, M. Organometallics 2006, 25,
808–808.
(46) Johansson, L.; Tilset, M.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem.
Soc. 2000, 122, 10846–10855.
(47) Zhong, H. A.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2002,
124, 1378–1399.
2
dz orbital of Pt(II) to the Lewis-acidic Cu(I) ion. Thus, the Pt(II)
(48) Driver, T. G.; Day, M. W.; Labinger, J. A.; Bercaw, J. E. Organo-
metallics 2005, 24, 3644–3654.
(53) Hammad, L. A.; Gerdes, G.; Chen, P. Organometallics 2005, 24, 1907–
1913.
(49) Driver, T. G.; Williams, T. J.; Labinger, J. A.; Bercaw, J. E.
Organometallics 2007, 26, 294–301.
(54) Gerdes, G.; Chen, P. Organometallics 2003, 22, 2217–2225.
(55) Moret, M.-E.; Chen, P. Organometallics 2007, 26, 1523–1530.
(56) Gerdes, G. Thesis, Catalytic C-H activation of benzene by platinum(II):
a mechanistic study, ETHZ, 2004.
(50) Williams, T. J.; Caffyn, A. J. M.; Hazari, N.; Oblad, P. F.; Labinger,
J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2008, 130, 2418–2419.
(51) Owen, J. S.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2006,
128, 2005–2016.
(57) Cordero, B.; Go´mez, V.; Platero-Prats, A. E.; Reve´s, M.; Echeverr´ıa,
J.; Cremades, E.; Barraga´n, F.; Alvarez, S. Dalton Trans. 2008, 2832–
2838.
(52) Chen, G. S.; Labinger, J. A.; Bercaw, J. E Proc. Natl. Acad. Sci. U.S.A.
2007, 104, 6915–6920.
9
5676 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
Scheme 2
Figure 1. ORTEP representation of the X-ray crystal structure of 4a.
Ellipsoids are drawn at 50% probability. Selected distances [Å] and angles
[deg]: Pt1-Cu1 2.3992(16), Cu1-O1 1.907(8), Pt1-C1 2.032(11), Pt1-C2
2.039(13), Pt1-N1 2.110(10), Pt1-N2 2.100(9), O1-Cu1-Pt1 175.4(3),
C1-Pt1-C2 87.4(6), C1-Pt1-N198.3(5), C2-Pt1-N2 98.6(5), N1-Pt1-N2
75.6(4), C1-Pt1-Cu1 82.3(4), C2-Pt1-Cu1 84.8(4), N1-Pt1-Cu1
95.1(3), N2-Pt1-Cu1 98.4(3), Cu1-O1-S1 121.7(5).
complex can be considered as a σ-donor ligand. The coordina-
tion around the Cu(I) center is then best described as a linear,
two-coordinate environment involving the Pt(II) complex and
the η1 triflate anion as ligands. Comparison with the structure
of 253 reveals a slight lengthening of the Pt-C (2.032(11) and
2.039(13) vs 2.007(7) Å) and Pt-N bonds (2.100(9) and
2.110(10) vs 2.076(5) Å) upon coordination to Cu(I), suggesting
that these bonds are somewhat weakened.
The silver analogue of 4a was prepared in 64% yield by
reaction of 2 with AgOTf in benzene (Scheme 2). Analysis of
the dark-red crystals by XRD revealed the formation of the
centrosymmetric 2:2 adduct 5b (Figure 2). The crystal structure
of 5b exhibits a planar Pt2Ag2 core built from four unsupported
PtfAg bonds (2.8290(6) and 2.9097(5) Å) and an argentophilic
interaction (2.6972(2) Å). This core structure is similar to that
of the known compound (NBu4)2[Pt2Ag2Cl4(C6F5)4], which is
however additionally stabilized by four bridging chloride anions
and weak ArF-Ag contacts.24,25 The influence of two silver(I)
ions on the coordination environment of platinum(II) is stronger
than that of a single copper(I) ion, as indicated by somewhat
longer Pt-C (2.039(4) and 2.043(4) Å) and Pt-N bonds
(2.117(3) and 2.118(3) Å) in 5b than in 4a, and by a slight
displacement of the Pt atom out of the plane of the ligands in
5b (0.17 Å from the best C2N2 plane). The fact that a dimeric
structure is observed for 5b and not for 4a supports the idea
that argentophilic interactions are stronger than cuprophilic
interactions.2,5
The bonding situation becomes different when the complexes
of the diphenyl complex 3 are considered. The reaction of 3
with 0.55 equiv of CuOTf·1/2C7H8 in dichloromethane yielded
a dark-purple solution, from which the 2:1 adduct 6a could be
crystallized as black needles (52%) by addition of toluene
followed by slow concentration. The XRD study of 6a (Figure
3) reveals an ionic structure containing a free triflate anion and
a homoleptic 2:1 complex cation, which possesses a C2 rotation
axis. While the Pt-Cu distance (2.7254(3) Å) is relatively short,
the very acute angle (25°) between the coordination plane of
the Pt center and the Pt-Cu axis presumably does not allow
Figure 2. ORTEP representation of the X-ray crystal structure of 5b.
Ellipsoids are drawn at 50% probability. For clarity, only the silver-bound
-
oxygen atom of the CF3SO3 units is plotted, and an unbound benzene
molecule is omitted. Selected distances [Å], angles [deg], and torsion angles
[deg]: Ag1-Ag1′ 2.6972(2), Ag1-O1 2.354(3), Ag1-Pt1 2.8290(6),
Ag1-Pt1 2.9097(5), Pt1-C1 2.039(4), Pt1-C2 2.043(4), Pt1-N1 2.117(3),
Pt1-N2 2.118(3), Ag1′-Ag1-Pt1 63.489(15), Ag1-Pt1-Ag1′ 56.047(15),
O1-Ag1′-Ag1 164.82(8), C1-Pt1-C2 89.28(17), C1-Pt1-N1 97.19(14),
C2-Pt1-N2 96.28(14), N1-Pt1-N2 75.59(11), Pt1-Ag1-Ag1′-Pt1′
180.0.
2
for a significant donation of electron density from the dz orbital
of Pt(II) to the Cu(I) cation. Thus, the Pt-Cu interaction is
expected to consist mainly of dispersion forces and, conse-
quently, to be weaker than in 4a. However, this fact is
compensated by a close contact (2.138(5) Å) and a more distant
one (2.336(5) Å) between the Cu(I) center and the ipso carbon
atoms of the phenyl ligands. These interactions are also reflected
by the tilting of the phenyl rings, shown by the nonzero values
of the Pt-C3-C4-C1 (7.9 °) and Pt-C5-C6-C2 (-7.8 °)
dihedral angles. Additionally, the Pt-C bonds are somewhat
longer than those in the free platinum complex 356 (2.022(5)
and 2.033(5) vs 2.003(3) and 2.004(4) Å).
The similar 2:1 silver adduct 7a could be obtained in 61%
yield from the reaction of 3 with AgOTf. However, all the
conditions we applied yielded severely disordered crystals. The
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5677

A R T I C L E S
Moret and Chen
Figure 3. ORTEP representation of the X-ray crystal structure of the cation
of 6a. Ellipsoids are drawn at 50% probability. The disordered, unbound
CF3SO3 counterion is omitted for clarity. Selected distances [Å], angles
[deg], and torsion angles [deg]: Cu1-Pt1 2.7254(3), Cu1-C1 2.138(5),
Cu1-C2 2.336(5), Pt1-C1 2.033(5), Pt1-C2 2.022(5), Pt1-N1 2.106(5),
Pt1-N2 2.082(4), Pt1-Cu1-Pt1′ 144.86(4), Pt1-Cu1-C1 47.55(13),
Pt1-Cu1-C2 46.31(12), C1-Cu1-C2 82.60(18), C1-Cu1-C1′ 126.3(3),
C2-Cu1-C2′ 130.4(3), C1-Cu1-C2′ 120.50(17), C1-Pt1-C2 93.64(18),
C1-Pt1-N1 95.47(17), C2-Pt1-N2 95.07(17), N1-Pt1-N2 75.37(15),
Pt1-C3-C4-C1 7.9, Pt1-C5-C6-C2 -7.8.
Figure 4. ORTEP representation of the centrosymmetric molecular unit
of 12. Ellipsoids are drawn at 50% probability. The aromatic groups of the
diimine ligand, the fluorine atoms of the CF3SO3 units, and six CH2Cl2
solvent molecules are omitted for clarity. Selected distances [Å], angles
[deg], and torsion angles [deg]: Cu1-Pt1 2.5995(8), Pt1-C1 2.048 (5),
Pt1-C2 2.028(6), Pt1-N1 2.097(5), Pt1-N2 2.097(5), Cu1-C1 2.160(5),
Cu1-C2 2.146(6), Cu1-O1 2.083(5), Cu1-O2 2.123(5), Cu2-C3 2.112(6),
Cu2-C4 2.148(7), Cu2-O3 2.034(5), Cu2-O4 2.077(5), Cu3-C5 2.484(7),
Cu3-C6 2.069(7), Cu3-C7 2.412(9), Cu3-O5 2.079(5), Cu-O6 2.148(5),
Cu3-O72.177(5), C1-Pt1-C293.8(2), Pt1-C3-C8-C110.1, Pt1-C5-C9-C2
9.8.
-
XRD data (see Supporting Information) is good enough, though,
to confirm that 7a is isostructural with 6a and to measure a
Pt-Ag distance of 3.28 Å, which indicates a weak (if any)
d8-d10 interaction, the bonding consisting mainly of the
coordination of the phenyl rings to the silver cation.
As solution-phase UV-vis and 1H NMR spectroscopic data
(Vide infra) indicated the existence of species with a different
stoichiometry than that found in 6a and 7a, we explored
crystallization conditions with a higher CuOTf:3 or AgOTf:3
ratio. At a 1:1 ratio, only crystals of 6a and 7a, respectively,
were obtained.
to the PtfCu bond found in structure 4b. The coordination of
the Ag1 cation is completed by two oxygen atoms from different
triflate anions and is roughly trigonal planar. The Ag2 atom is
bound to 3 in a fashion similar to the one found in 6a and 7a:
the Pt-Ag2 distance of 3.1694(9) indicates at most a weak
d8-d10 interaction, and the bonding mainly consists of
Ag(I)-Phenyl contacts. The small difference between the
Ag2-C1 and Ag2-C2 contacts (2.500(6) and 2.638(8) Å)
entices us to describe these interactions as a distorted η2
coordination, in contrast with the η1 mode found for the
Cu-Phenyl bonds in 6a. Another η2 interaction with a benzene
ligand and the bond to the bridging triflate anion complete the
distorted tetrahedral coordination environment of Ag2.
In most crystallization attempts involving higher CuOTf:3
ratios, the low solubility of CuOTf caused it to precipitate
together with the platinum containing compound(s), and the
obtained residues were unsuitable for both XRD and elemental
analysis. However, saturation of a solution of 3 in dichlo-
romethane with CuOTf ·C6H6 followed by slow cooling afforded
a small amount of the 2:6 adduct 12 as dark-purple needles.
XRD analysis (Figure 4) shows that the molecular unit of 12
consists of a centrosymmetric, hexanuclear Cu(I) triflate cluster
stabilized by interaction with two (NN)PtPh2 units. One of the
three crystallographically independent Cu(I) centers (Cu1) is
bound to the (NN)PtPh2 moiety in a similar fashion to that found
in compound 6a, with slightly shorter contacts with the ipso
carbon atoms of the phenyl groups (2.160(5) and 2.146(6) vs
2.138(5) and 2.336(5) Å) and with the platinum center (2.5995(8)
vs 2.7254(3)). The two other Cu(I) atoms are bound to each of
the phenyl groups in an η2 (Cu2) or η3 (Cu3) fashion.
A structurally different 1:2 adduct, 11b, was obtained from
a solution containing 3 and a large excess of AgOTf. XRD
analysis of 11b (Figure 5) revealed the presence of one-
dimensional chains consisting of Ag-3-Ag units connected
by a triflate anion, the Ag2Pt plane being a crystallographic
symmetry plane. Both triflate anions in the structure are
disordered and have been refined on two positions, which
precludes an accurate estimation of the corresponding bonding
parameters. The Ag1 atom is bound to the platinum center Via
a short (2.7616(8) Å), unsupported PtfAg bond which makes
an angle of 86° with the ligand plane, and is thus very similar
In summary, XRD studies allowed us to identify two distinct
binding modes of 2 and 3 to monovalent coinage metals. The
first one, available for both 2 and 3, consists of a single dative
PtfM bond, whose geometry is consistent with donation of
2
electron density from the dz of platinum(II) to the Lewis-acidic
metal. The second mode, which is preferred for 3, involves
mainly coordination to the ipso carbon atoms of the platinum-
bound phenyl groups, in an η1 or η2 fashion depending on the
ionic radius of the coinage metal.
Solution-Phase Studies. The interaction of 2 and 3 with Cu(I)
and Ag(I) was further characterized in solution by means of
1
UV-vis and H NMR spectroscopies.
The UV-vis spectrum of 2 in benzene (Figures 6 and 7, black
line) exhibits one band at 416 nm and one band at 644 nm with
two high-energy shoulders and one weak low-energy shoulder,
58
in agreement with what was found by Scollard et al. for a
series of closely related compounds. Both bands exhibit a
strongly negative solvatochromic effect of similar magnitude
(see Supporting Information), indicating that the corresponding
electronic excited states are less polar than the ground state.
Following Scollard et al.,58 we assign the lower energy band to
1
a set of (d(Pt)fπ*) singlet metal-to-ligand charge-transfer
(58) Scollard, J. D.; Day, M.; Labinger, J. A.; Bercaw, J. E. HelV. Chim.
Acta 2001, 84, 3247–3268.
9
5678 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
Figure 7. UV-vis absorption of 1.97 × 10-4 M solutions of 2 in benzene
containing 0 (black) to 1.53 (red) equiv of AgOTf. (Inset) Absorbance at
644 nm as a function of the Ag:Pt ratio.
amounts of CuOTf ·1/2C7H8 are plotted in Figure 6. The bands
at 644 and 416 nm are progressively depleted upon addition of
CuOTf·1/2C7H8, and a band at 507 nm simultaneously appears.
A band at ∼350 nm also appears as a shoulder of the growing
1
band at ∼320 nm, which is due to a (d(Cu)fπ*(benzene))
MLCT transition of free CuOTf ·1/2C6H6.59 The presence of
three clean isosbestic points indicates that a single equilibrium
is involved, which we attribute to the formation of the
crystallographically characterized 1:1 adduct 4a (Scheme 3).
The blue shift of the 1MLCT bands is consistent with the
description of the Pt-Cu bond as a donor-acceptor bond, the
2
Figure 5. ORTEP representation of the X-ray crystal structure of 11b.
Ellipsoids are drawn at 50% probability. The CF3SO3- units are disordered
and were refined isotropically over two symmetry-related positions with
50% occupancy each. Two unbound benzene molecules are omitted for
clarity. Top: view of one [PtAg2] unit, including only the coordinated oxygen
platinum dz orbital being stabilized by its interaction with the
Cu(I) cation. However, the considerable modification of the fine
structure of the first band makes it difficult to accurately quantify
the energy shift.
-
atoms of the disordered CF3SO3 units. Bottom: Side view of one one-
The behavior of 2 in the presence of silver(I) triflate is similar,
as seen in the UV-vis titration data plotted in Figure 7. The
dimensional chain. Selected distances [Å], angles [deg], and torsion angles
[deg]: Pt1-Ag1 2.7616(8), Pt1-Ag 2 3.1694(9), Pt1-C1 2.009(7), Pt-N1
2.104(6), C1-Ag2 2.500(6), C2-Ag2 2.638(8), C3-Ag2 2.534(8),
Ag1-Pt-Ag2 122.30(3), C1-Pt1-Ag1 93.18(18), N1-Pt1-Ag1 87.20(14),
C1-Pt1-C1′ 92.0(4), C1-Pt1-N1 96.7(3), N1-Pt1-N1′ 74.6(3),
Pt1-C1-Ag 2 88.6(2), C1-Ag1-C1′ 70.6(3), C1-Pt1-C1′-C2′ 113.6(6).
1
first MLCT band (λmax ) 644 nm in 2) undergoes a more
pronounced blue shift upon coordination to Ag(I) (λmax ) 475
nm) compared to Cu(I) (λmax ) 507 nm). As in the case of
CuOTf, the observation of clean isosbestic points establishes
the formation of a single complex in solution, which is assigned
to the monomeric species 5a (Scheme 3) on the basis of pulsed
magnetic field gradient NMR (PFG-NMR) diffusion measure-
ments (Vide infra). The lack of knowledge of the aggregation
state of CuOTf and AgOTf in benzene solution precludes the
establishment of appropriate equilibrium equations, and thus no
quantitative determination of association constants was at-
tempted. However, the comparison of the traces obtained at λ
) 644 nm for both coinage metals qualitatively shows that the
association constant is higher for 5a than for 4a.
The electronic consequences of the formation of PtfM bonds
1
were also investigated by H NMR spectroscopy. Since the
UV-vis data show that there is only one adduct involved in
1
both cases, the H NMR spectra of 4a and 5a (Figure 8) were
Figure 6. UV-vis absorption of 1.96 × 10-4 M solutions of 2 in benzene
containing 0 -1.83 equiv of CuOTf. The black spectrum corresponds to
pure 2 and the red one to 2 in the presence of a large excess of CuOTf.
(Inset) Absorbance at 644 nm as a function of the Cu:Pt ratio. The dashed
line corresponds to the absorbance in the presence of excess CuOTf.
recorded in the presence of an excess of the corresponding
triflate to ensure the absence of uncomplexed 2. The resonance
corresponding to the methyl groups of the R-diimine ligand
shifts downfield by 0.76 and 0.77 ppm upon complexation to
CuOTf and AgOTf, respectively. This shift most likely origi-
nates from inductive effects, the electron density depletion on
the Pt(II) center translating into a net deshielding of the methyl
protons. The effect of complexation on the platinum-bound
(1MLCT) transitions. However, these authors’ assignment of the
higher energy band to a 1(πfπ*) ligand-centered (1LL) transi-
tion seems questionable in the case of complex 2 because of
the important solvent sensitivity of this band. Thus, we propose
1
1
methyl H resonances seems to be more complex, since they
that the band is at least partially due to (d(Pt)fπ2*) MLCT
are shifted by only a small amount and toward higher field.
transitions.
The UV-vis spectra of benzene solutions containing ∼2 ×
10-4 M of the dimethylplatinum(II) complex 2 and various
(59) Kunkely, H.; Vogler, A. Chem. Phys. Lett. 2003, 368, 49–52.
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5679

A R T I C L E S
Moret and Chen
Scheme 3a
a Evidence for the existence of the 2:1 complexes 10a is found by 1H NMR at ∼2.5 × 10-2 M, but not at concentrations relevant for UV-vis spectroscopy
(∼2 × 10-4 M). No evidence for the formation of 11a in solution was found.
Figure 9. UV-vis absorption of 1.96 × 10-4 M solutions of 3 in benzene/
dichloromethane 1:1 containing 0-1.43 equiv of CuOTf ·1/2C7H8. The black
spectrum corresponds to pure 3 and the red one to 3 in the presence of
excess CuOTf ·1/2C7H8. (Inset) Absorbance at 622 nm as a function of the
Cu:Pt ratio. The dashed line corresponds to the absorbance in the presence
of excess CuOTf ·1/2C7H8.
Figure 8. 1H NMR spectra of 2 pure (bottom), and in the presence of an
excess CuOTf (middle) or AgOTf (top). Peak assignment: NAr3,5, ∆; NAr4,
0; CMe, O PtMe, ]. The peak of benzene coming from CuOTf ·1/2C6H6
is marked with *.
Table 1. Diffusion Coefficients (D) and Hydrodynamic Volumes
(VH) Measured by PFG NMR.
sample
D (× 10-10 m2 s-1
)
VH (nm3)
2
13.4
11.5
12.7
12.7
9.7
9.5
10.4
8.7
0.41
0.59
0.47
0.47
0.90
0.94
0.76
1.19
2+CuOTfa (4ab)
Figure 10. UV-vis absorption of 1.96 × 10-4 M solutions of 3 in benzene/
dichloromethane 1:1 containing 0 (black) to 2.51 (red) equivalents of AgOTf.
(Inset) Absorbance at 622 nm as a function of the Ag:Pt ratio. The dashed
line corresponds to the absorbance in the presence of excess AgOTf.
2+AgOTfa (5ab)
3
6a
7a
3+CuOTfa (8ab)
3+AgOTfa (9bb)
the UV-vis titration data shown in Figures 9 and 10. This data
was obtained in a 1:1 mixture of dichloromethane and benzene
to avoid precipitation problems. The electronic spectrum of 3
resembles the one of 2: a first band at 622 nm (646 nm in pure
benzene) with two high energy shoulders, and a second, more
intense band at 380 nm (387 nm in benzene) are observed. Both
bands exhibit a strong negative solvatochromic effect (see
Supporting Information). As for 2, we assign the first band to
a set of 1(d(Pt)fπ*) 1MLCT transitions, and the second one to
a In equilibrium with solid MOTf. b Main species in solution.
The most striking effect on this signal is the substantial decrease
of the 2J(195Pt-1H) coupling constant from 87.9 Hz in 2 to 56.5
and 63.8 Hz in 4a and 5a, respectively.
The aggregation state of species 4a and 5a in solution was
investigated by estimation of their hydrodynamic volume (VH)
using PFG NMR translational diffusion measurements (Table
1). Addition of an excess of CuOTf to a solution of 2a in CD2Cl2
results in an increase of VH from 0.41 to 0.59 nm3, in accord
with the formation of monomeric 4a. The hydrodynamic volume
obtained for 5a (0.47 nm3) is smaller than that of 4a, showing
that the solid-state dimer 5b does not form in DCM solution.
The difference between 4a and 5a may be due to partial
dissociation of the Ag-OTf bond in solution.
1
1
overlapping LL and MLCT transitions. Both bands undergo
a blue shift upon addition of CuOTf or AgOTf, which likely
originates from the stabilization of the HOMO Via its interaction
with the Lewis-acidic metal. Surprisingly, the addition of CuOTf
to 3 also causes an increase in the absorbance between ∼700
and 800 nm, resulting in a marked low energy shoulder on the
1
first MLCT band. Since this shoulder is not observed for the
1
silver complexes, we tentatively assign it to a (d(Cu)fπ*)
The reaction of the diphenyl complex 3 with coinage metal
triflates in solution is more complex than for 2, as appears from
1MLCT transition, in accord with the higher stability of the
Cu(II) oxidation state as compared to Ag(II).
9
5680 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
2:1 complex 6a. As seen in the crystal structure of 6a (Figure
3), the formation of the complex brings the phenyl groups of
the two [(NN)PtPh2] units in close proximity. Thus shielding
of the ortho protons of the phenyl groups by phenyl groups of
the second [(NN)PtPh2] moiety accounts for the shift of the
corresponding peak toward high field. When more CuOTf ·1/
2C6H6 is added, the concentration of the 1:1 adduct 8a increases,
and purely inductive effects cause the downfield shift of the
whole spectrum. The substantial downfield shifts observed from
the spectrum with two equivalents to that with a large excess
of CuOTf·1/2C6H6 may indicate the formation of a small
amount of the 1:2 adduct 10a (Scheme 3). This contrast with
the UV-vis spectroscopic observations is supposedly due to
the higher concentration (2.5 × 10-2 vs 2 × 10-4 M) and solvent
polarity (pure dichloromethane vs 1:1 benzene/dichloromethane).
The 1H NMR resonances of 3 also shift upon addition of AgOTf,
but strong overlap in the aromatic region precludes a detailed
interpretation of the obtained spectra (see Supporting Informa-
1
tion). However, we observed that the H NMR spectra of 3 in
the presence of 1 equiv or an excess of AgOTf are identical,
showing that the 1:2 adduct 11a is not formed in dichlo-
romethane solution.
More information about the solution-phase structure of the
adducts of 3 with CuOTf and AgOTf is gained by PFG NMR
experiments (Table 1). The hydrodynamic volume of 6a (0.90
nm3) and 7a (0.94 nm3) are almost identical and about twice as
large as that of 3 (0.47 nm3), in accord with their formulation
as isostructural 2:1 adducts. When an excess of CuOTf is added
to a solution of 6a, the measured hydrodynamic volume
decreases to 0.76 nm3, indicating the formation of the smaller,
mononuclear 1:1 complex 8a and ruling out the existence of
high-nuclearity clusters like 12 in solution. In contrast, addition
of AgOTf to a solution of 7a results in an increase of VH to
1.19 nm3, which strongly suggests that the main species is the
2:2 adduct 9b (Scheme 3). The detailed structure of 9b is
unknown.
Figure 11. 1H NMR spectra of ∼25 mM solutions of 3 in CD2Cl2
9; PtPh4, b; NAr3,5, ∆; NAr4, 0; CMe, O. The peak of benzene coming
from CuOTf ·1/2C6H6 is marked with *. It is absent in the spectrum with
0.5 equiv of CuOTf because the latter was recorded from crystallized 6a.
containing various amounts of CuOTf. Peak assignment: PtPh2,6, 2; PtPh3,5
,
In contrast to what is found for 2, no isosbestic points are
observed, and there are regions where the absorbance succes-
sively increases and decreases upon progressive addition of
CuOTf·1/2C7H8 or AgOTf. This indicates that several equilibria
are involved. The observation that, in both cases, the spectra
do not evolve significantly for M:Pt (M ) Cu, Ag) ratios above
ca. 1.5, allows us to exclude the formation of complexes
incorporating more than one MOTf unit per molecule of 3 in
solution. Thus, we propose that the two leftmost equilibria
depicted in Scheme 3 are taking place at these concentrations,
involving the existence of both the crystallographically char-
acterized 2:1 complexes 6/7a and complexes with a M:Pt ratio
of 1:1 (8a and 9b, Vide infra).
UV-vis and 1H NMR data consistently show that the PtfM
bonds observed in the solid-state structures 4a and 5b also exist
in solution. With complex 3 as the ligand, the 2:1 complexes
6a and 7a, the 1:1 complexes 8a, and the 2:2 complex 9b are
formed, and evidence for the existence of the 1:2 adducts of
CuOTf 10a is found at high overall concentrations. Coordination
of 2 and 3 to Cu(I) and Ag(I) results in a blue shift of absorption
The existence of the two successive equilibria described in
the previous paragraph was confirmed in the case of Cu(I) by
1
1
bands corresponding to MLCT transitions in all cases. This
recording H NMR spectra of ca. 25 mM solutions of 3 in
observation parallels the one by Fornie´s et al.28 of a similar
blue shift of the MLCT band of the [Pt(bzq)(C6F5)2]- (bzq )
7,8-benzoquinolate) anion upon coordination to the (Ph3P)Ag+
cationic fragment through a PtfAg dative bond. The interaction
of the platinum complexes 2 and 3 with coinage metal triflates
generally results in a shift of most proton resonances toward
low field in the 1H NMR spectrum, which is most likely due to
inductive effects. Additionally, in the case of the dimethyl
CD2Cl2 in the presence of various amounts of CuOTf ·1/2C6H6
(Figure 11). Upon progressive addition of CuOTf, the resonances
corresponding to the methyl groups (O) and the para hydrogen
(b) atoms of the phenyl groups shift monotonically toward low
field, which is easily explained by remote inductive effects.
However, other peaks in the aromatic region of the spectrum
follow a more complex trajectory. The most striking effect
concerns the ortho hydrogen atoms of the phenyl groups (2),
whose resonance is easily identified by its double doublet (dd)
structure and the satellite signals originating from 195Pt-1H
coupling. This signal appears at δ 6.95 ppm for pure 3. In the
presence of 0.5 equivalent of CuOTf, this peak broadens and
shifts toward high field to δ 6.47 ppm. At higher Cu:Pt ratios,
it first becomes very broad and starts moving back to lower
field, and is then found as a relatively sharp resonance at δ 7.08
ppm at a Pt:Cu ratio of 1:2, and finally at δ 7.30 ppm when the
solution is saturated with CuOTf.
2
complex 2, a substantial decrease of the J(195Pt-H) coupling
constant upon complexation is observed. Similar reductions of
the 2J(195Pt-H) coupling constant in the [(bpy)PtMe2] complex
from 86 to 64 Hz and 75 Hz upon coordination to the (Ph3P)Au+
and (Ph3P)Ag+ fragments, respectively, were observed by
Puddephatt.34 This effect thus appears to be diagnostic for the
presence of PtfM bonds in solution.
Gas-Phase Study. The interaction of 2 and 3 with coinage
metal monocations was also studied in the gas phase by means
of electrospray ionization mass spectrometry (ESI-MS). Elec-
trospraying dilute acetonitrile solutions containing either 2 or
We rationalize these observations as follows: when only 0.5
equiv of CuOTf is present in solution, the major species is the
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5681

A R T I C L E S
Moret and Chen
Scheme 4a
a Dashed bonds represent PtfM dative bonds for R ) Me and the interaction of M with the π electrons of phenyl groups for R ) Ph (see Figure 13).
Table 2. CID Fragmentation Patterns for Cations 4c, 5c, 8c, 9c,
13, 14, 6b, 7b
ion
m/z
CID fragments
4c
704 663 (-MeCN), 681 (-MeCN + H2O), 645 (-MeCN -
CH4), 547 (-MeCN - CH4 - CuCla), 437 (-MeCN -
PtMe2)
13 1261 1035 (-PtMe2), 663 (-(NN)PtMe2), 645 (-(NN)PtMe2 -
CH4), 547 (-(NN)PtMe2 - CH4 - CuCla), 437
(-(NN)PtMe2 - PtMe2)
5c
748 707 (-MeCN), 481 (-MeCN - PtMe2)
14 1308 1081 (w, -PtMe2), 707 (-(NN)PtMe2), 481 (w,
-(NN)PtMe2 - PtMe2)
8c
6b 1510 787 (-(NN)PtPh2)
9c
828 787 (-MeCN)
872 831 (-MeCN)
7b 1554 831 (-(NN)PtPh2)
a Formed by activation of a C-Cl bond of the diimine ligand.
3 and CuOTf or AgOTf allowed the clean observation of
solvated 1:1 complexes which we presume have the structures
4c, 5c, 8c, and 9c (Scheme 4). Additionally, weaker signals
corresponding to the 2:1 complexes 13, 14, 6b, and 7b also
appeared in the corresponding spectra. Collision induced dis-
sociation (CID) spectra were recorded for each individual ion,
and the resulting fragments are listed in Table 2.
The main dissociation pathway for the 1:1 cations 4c, 5c, 8c
and 9c is the loss of the coordinated acetonitrile, followed by
further fragmentation in the case of 4c (Figure 12). While this
may appear unsurprising, it gives some information about the
strength of the PtfM bonds, showing that they are in all cases
stronger than the MeCN-M coordination bonds. For compari-
son, the gas-phase dissociation energy of an acetonitrile ligand
from the linear [(MeCN)2M]+ cations have been experimentally
determined as 57.0 kcal/mol for M ) Cu60 and 34.6 kcal/mol
for M ) Ag.61
Figure 12. Top: ESI-MS spectrum of a solution of 2 and CuOTf in MeCN;
experimental (black) and calculated (white) isotope patterns. Middle: CID
spectrum of 4c. Bottom: CID spectrum of 13. Peak assignment: A:
{[(NN)PtMe2]Cu}+, B: A - CH4, C: B - CuCl, D: [(NN)Cu]+.
tion of the 2:1 complexes 13 and 14 is more complex (Figure
12, Table 2). In addition to the expected loss of the [(NN)PtMe2]
fragment 2 and subsequent fragmentation, cations 13 and 14
can also lose a naked PtMe2 neutral fragment; for the Cu(I)
containing cation 13, this is even the main CID channel (Figure
12). The structures of the cationic fragments cannot be
unambiguously determined, but the simplest possible structures
would be 15 and 16 (Scheme 4), in which one PtfM bond is
conserved and the second R-diimine ligand is coordinated to
the coinage metal M. While this process in itself is of limited
interest, its observation shows that d8-d10 bonds can be strong
enough to support ligand scrambling processes between two
metals in the gas phase. This suggests that, with the right ligand
When 3 is the ligand, the 2:1 complexes 6b and 7b cleanly
lose one [(NN)PtPh2] unit upon CID. In contrast, the fragmenta-
(60) Vitale, G.; Valina, A. B.; Huang, H.; Amunugama, R.; Rodgers, M. T.
J. Phys. Chem. A 2001, 105, 11351–11364.
(61) Shoeib, T.; El aribi, H.; Siu, K. W. M.; Hopkinson, A. C. J. Phys.
Chem. A 2001, 105, 710–719.
9
5682 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
Table 3. Comparison between Experimental (XRD) and Calculated
in the more complex tetranuclear structures of type
[(baimp)PtPh2]2Cu2 (baimp ) 3,5-bis(4-aryliminoacetyl)-4-
methylpyrazolate).37
(BP86/6-31+G(d);Cu,Ag,Pt:SDD) Values for Selected Distances
(Å)
4a
5b
6a
In the second structure, labeled 8a2, the binding mode is the
one a found in the 2:1 complex 6a and in the 2:6 adduct 12:
the copper atom has close contacts (2.098 and 2.128 Å) with
the ipso carbon atoms of both phenyl ligands, and the Pt-Cu
axis makes an angle of 27° with the plane of the ligands around
the platinum atom. The Pt-Cu distance (2.654 Å), however, is
remarkably short, and should allow for a significant metal-metal
interaction. The longer Pt-Cu bond length (2.758 Å) calculated
for the 2:1 cation of 6a is likely due to steric repulsion between
the two [(NN)PtPh2] moieties.
exp
calc
exp
calc
exp
calc
Pt-Ma
2.399
2.438
2.829
2.910
2.354
2.829
2.970
2.039
2.043
2.117
2.118
2.917
2.930
2.328
2.915
2.959
2.057
2.057
2.181
2.183
2.725
2.758
M-Oa
1.907
2.929
3.003
2.032
2.039
2.110
2.100
1.888
3.131
3.135
2.050
2.050
2.176
2.175
-
-
M-Ca,b
2.138
2.336
2.022
2.033
2.082
2.106
2.164
2.361
2.046
2.061
2.152
2.164
Pt-C
Pt-N
a M ) Cu for 4a and 6a, M ) Ag for 5b. b No chemical bond for 4a
and 5b.
Two similar isomers are found for the silver complex 9a.
However, some noticeable differences are found between
structures 9a2 and 8a2. The silver atom in 9a2 has a short
contact to only one of the phenyl ipso carbon atoms (2.321 vs
2.640 Å), but it is also close to one of the ortho carbon atoms
(2.566 Å). Thus, the Ag(I) center in 9a2 is coordinated by a
single phenyl group in a distorted η2 fashion, while the Cu(I)
center in 8a2 experiences η1 coordination from both phenyl
groups in an almost symmetrical way. Another difference is
the fact that the Ag-Pt distance in 9a2 (3.059 Å) is significantly
larger than that in 9a1 (2.847 Å), reflecting a relatively weak
metal-metal interaction.
combination, a process of fundamental interest like transmeta-
lation could be studied on related complex ions.
DFT Calculations. The various experimental investigations
described in the previous sections were complemented by
theoretical studies using density functional theory. These
calculations were performed with two main goals. The first one
was to clarify the structure of compounds whose existence was
implied by solution phase data but for which no XRD structure
could be obtained, namely the 1:1 adduct 8a and its silver
analogue 9a. The second objective was to get a more precise
description of the bonding in the various compounds described
in this work.
The geometries of the 1:1 adducts 4a, 5a, 8a, and 9a, as well
as the dimeric structure 5b and the 2:1 cations of 6b and 7b,
were optimized at the BP86 level with a Pople-type 6-31+G(d)
basis set on main group elements and the Stuttgart-Dresden62,63
effective core potentials and basis sets on transition metals. The
overall geometries of compounds characterized by XRD are well
reproduced by our calculations. The comparison between
selected experimental and calculated bond lengths (Table 3)
shows that DFT tends to slightly overestimate the metal-ligand
and metal-metal bond lengths, but calculated values are always
within 4% of the experimental ones. The ability of this DFT
method to predict the length of the unsupported PtfM bonds
in 4a and 5b is remarkable given its generally poor treatment
of dispersion forces.1 While error cancelation may be partially
responsible for the good performance of DFT, we tentatively
attribute it to the strong donor-acceptor character of these
PtfM bonds as compared to usual van der Waals or metallo-
philic bonds.
The difference between copper(I) and silver(I) is also apparent
in the computed relative energies of the two isomers (Figure
13): structure 8a2 is predicted to be more stable than 8a1, while
9a2 is higher in energy than 9a1. The simplest explanation to
these observations is based on the different ionic radii of Ag(I)
and Cu(I). The smaller Cu(I) cation fits well in the cavity formed
by the two phenyl ligands and the Pt(II) center, and it is able to
maximize its interactions with these three groups in structure
8a2. On the other hand, the Ag(I) cation is too voluminous for
this cavity and preferentially binds to only one phenyl group in
an η2 fashion and to the platinum center, which is most
efficiently achieved in structure 9a1.The binding mode found
in 9a1 leaves more accessible space around the silver(I) ion,
which may explain the observation of the dimer 9b in solution.
Some insight into the nature of the interaction of the
organometallic platinum complexes 2 and 3 with Lewis-acidic
metals is gained by considering their computed Kohn-Sham
(KS) orbital structures (Figure 14). In compound 2, the highest
2
occupied molecular orbital (HOMO) is the dz orbital of the
platinum atom, as expected for a square planar d8 transition
metal complex. The lowest unoccupied molecular orbital
(LUMO) is a π* orbital of the R-diimine ligand, which is
consistent with the assignment of the first band in its UV-vis
spectrum to 1 (d(Pt)fπ*) 1MLCT transitions (Vide supra). The
situation is slightly different for the diphenyl complex 3. In this
case, there is an important interaction between the filled π
orbitals of the phenyl ligands and the dxy orbital of the metal.
As a result, the HOMO is an antibonding combination of these
For the 1:1 adduct 8a we found two structures involving two
different coordination modes of the [(NN)PtPh2] moiety (Figure
13). In the first one, labeled 8a1, the Cu(I) cation makes a
relatively short (2.668 Å) bond with the Pt(II) center, and is
coordinated in an η2 fashion to only one of the phenyl groups.
The coordination to Cu(I) induces a lengthening of the corre-
sponding Pt-C bond (2.048 vs 2.022 Å) and a distortion of the
square planar geometry around the platinum center: the Pt-C
bond makes an angle of 16° with the plane containing the three
other ligands. While not optimal, the angle of 64° between the
Pt-Cu axis and the ligand plane should allow for a significant
2
orbitals, and the dz orbital is found at the HOMO-1 level.
The mapping of the electric potential (MEP) at the surface
of molecules 2 and 3 is also represented in Figure 14. In accord
with the KS orbital structure, compound 2 exhibits a single zone
of strongly negative potential above the Pt(II) center. For
compound 3, the most negative region is found in between the
two phenyl groups, but the space above the platinum atom also
experiences a distinctly negative potential. Comparison with the
experimental (XRD) and calculated structures of the MOTf
adducts (M ) Ag, Cu) shows that the MEP strongly correlates
2
donation of the dz of the Pt(II) center to the Cu(I) atom.
Interestingly, this binding mode is very similar to the one found
(62) Bergner, A.; Dolg, M.; Kuechle, W.; Stoll, H.; Preuss, H. Mol. Phys.
1993, 80, 1431.
(63) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer, R. M. J. Phys. Chem. 1993,
97, 5852.
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5683

A R T I C L E S
Moret and Chen
Figure 13. Calculated structures (BP86/6-21+G(d);Cu,Pt,Ag:SDD) and binding energies (BP86/TZ2P) [kcal/mol] for the 1:1 adducts 4a, 5a, 8a, and 9a.
Colors: H light gray, C dark gray, N dark blue, O red, F light blue, S yellow, Cl green, Cu orange, Ag brown, Pt purple. Selected bond lengths [Å] (in cases
where the two C (N) atoms are not equivalent, the value corresponding to the atom which is closest to the coinage metal is marked with *): 4a: Pt-Cu 2.438;
Pt-C 2.050, 2.050; Pt-N 2.175, 2.176; Cu-O 1.888. 5a: Pt-Ag 2.661; Pt-C 2.050, 2.050; Pt-N 2.170, 2.175; Ag-O 2.199. 8a1: Pt-Cu 2.668; Pt-C
2.048*, 2.022; Pt-N 2.198*, 2.138; Cu-O 1.918; Cu-Cipso 2.115; Cu-Cortho 2.155. 8a2: Pt-Cu 2.654; Pt-C 2.059*, 2.062; Pt-N 2.139*, 2.141; Cu-O
1.987; Cu-Cipso 2.098*, 2.128; Cu-Cortho 2.466*, 2.574. 9a1: Pt-Ag 2.847; Pt-C 2.036*, 2.022; Pt-N 2.203*, 2.149; Ag-O 2.185; Ag-Cipso 2.516;
Ag-Cortho 2.434. 9a2: Pt-Ag 3.059; Pt-C 2.048*, 2.041; Pt-N 2.161*, 2.137; Ag-O 2.228; Ag-Cipso 2.321*, 2.640; Ag-Cortho 2.566*, 2.709.
Therefore, structures 4a and 5a can be seen as isolobal
analogues of the pentacoordinate Pt(IV) hydride complexes that
are thought to be intermediates in the protolysis of Pt(II)
dialkyl41,64 and diaryl39,41,65 complexes. On the other hand, 8a
and 9a resemble more the transition state for the reductive
elimination of benzene from the Pt(IV) intermediate. This
difference between d10 metal cations and H+ can be understood
from the fact that the former are softer Lewis acids, allowing
for a stronger interaction with the π electrons of the phenyl
rings.
A semiquantitative estimation of the different contributions
to the computed bond dissociation energies (BDEs) was obtained
by performing an energy partitioning analysis66 on the computed
structures for the 1:1 adducts 4a, 5a, 8a and 9a. The total
binding energy ∆E of the [(NN)PtR2] and MOTf fragments is
decomposed into the preparation energy, ∆Eprep, and the
interaction energy, ∆Eint. ∆Eprep corresponds to the energy
needed to deform the fragments from their optimal geometry
Figure 14. Frontier Kohn-Sham orbitals and mapping of the electric
to their geometry in the complex. ∆Eint is further split into the
potential (MEP) from -0.055 (red) to +0.055 (blue), projected on an
Pauli repulsion term, ∆EPauli, the electrostatic term, ∆Eelstat, and
the orbital interaction term, ∆Eorb. Additionally, the extent to
which charge transfer contributes to the binding was estimated
by computing the Hirshfeld89 charge on the [(NN)PtR2] fragment
in all complexes.
isodensity surface for compounds 2 (left) and 3 (right)
with the observed binding modes. In all complexes of the
dimethyl complex 2 the coinage metal is found right on top of
the square planar Pt(II) center, while in most complexes of the
diphenyl complex 3 η1 or η2 coordination of the phenyl groups
is preferred. The 1:2 adduct 11b offers an example of a structure
where both positions are occupied. Thus, a simple electrostatic
model is able to predict the locations of the Cu(I) and Ag(I)
metals in the corresponding adducts, which suggests that these
interactions are mainly electrostatic in nature, the coinage metal
cation acting like a “fat proton”.
(64) Wik, B. J.; Ivanovic-Burmazovic, I.; Tilset, M.; van Eldik, R. Inorg.
Chem. 2006, 45, 3613–3621.
(65) Ong, C. M.; Jennings, M. C.; Puddephatt, R. J. Can. J. Chem. 2003,
81, 1196–1205.
(66) Bickelhaupt, F.; Baerends, E. Kohn-Sham Density Functional Theory:
Predicting and Understanding Chemistry. In ReV. Comp. Chem., Vol.
15; Lipkowitz, K. B.; Boyd, D. B. , Eds.; Wiley-VCH: 2000.
9
5684 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
Table 4. Energy Partitioning Analysis [kcal/mol] and Fragment
Scheme 5
Hirshfeld Charges for the Calculated Structures of 4a, 5a,8a, and
9a
4a
5a
8a1
8a2
9a1
9a2
∆EPauli
∆Eelstat
∆Eorb
57.05
45.27 103.38
129.97
75.81
80.10
-62.18 -50.97 -94.9
-43.89 -30.54 -64.28
-49.02 -36.24 -55.82
5.24
-43.78 -35.22 -42.98
0.198 0.195 0.126
-105.50 -73.78 -70.25
-75.75 -45.92 -44.27
-51.29 -43.90 -34.43
6.13
-45.15 -31.05 -29.47
0.067 0.158 0.118
a
∆Eint
∆Eprep
1.02
12.85
12.84
4.95
slightly better yield of 3 (60%), and decomposition leading to
free ligand 1 was less important (1%).
∆Eb
q((NN)PtR2)
Monitoring these reactions in situ by 1H NMR spectroscopy
proved unfeasible due to the number of transient species and
the formation of a range of MOTf adducts with different
stoichiometry, but nevertheless allowed the detection of the
produced methane as a mixture of isotopomers. Mechanistically
relevant information could be obtained from ESI-MS analysis
of samples of the reaction mixture diluted in acetonitrile. Figure
15 shows a spectrum obtained from a solution of 4a in benzene
after a reaction time of 17 h at 50 °C. This spectrum indicates
the presence in the reaction mixture of starting material 2 and
product 3, as well as the half-reacted intermediate [(NN)Pt-
(Me)(Ph)] (19), observed as their solvated Cu+ complexes.
Additionally, the peaks at m/z 613 and 654 arise from the Pt(IV)
fragment [(NN)PtMe3]+, presumably formed by the facile
dissociation of the triflate complex [(NN)PtMe3(OTf)] (17). The
peak at m/z 623 comes from the acetonitrile-solvated complex
[(NN)Pt(Me)(NCMe)]+, which we attribute to the reaction of
the weakly bound triflate complex [(NN)Pt(Me)(OTf)] (18) with
acetonitrile. Finally, the presence of free ligand 1 is confirmed
by the peak at m/z 478, which corresponds to its Cu(I) complex
[(NN)Cu(NCMe)]+. Workup of this partially reacted mixture
and analysis of the residue by 1H NMR confirmed the presence
of 1 (12%), 2 (27%), 19 (18%), and 3 (43%). The triflate-
containing species 17 and 18 were not observed and were
probably retained on the alumina column together with copper(I)
triflate.
a ∆Eint ) ∆EPauli + ∆Eelstat + ∆Eorb b ∆E ) ∆Eint + ∆Eprep
. .
The contribution of orbital interaction to the sum of the
attractive terms ranges from 37.5% for 5a to 41.8% for 8a2.
Thus, the binding in all complexes is predominantly electrostatic,
but orbital contributions are not negligible. It should be noted,
however, that the term ∆Eorb also includes polarization effects,
and thus the actual contribution of donor-acceptor orbital
interaction is expected to be lower than ∆Eorb. The computed
partition of the attractive terms is similar to what was found by
Nechaev et al.67 for simple ethylene and acetylene complexes
of coinage metal cations, where the orbital interaction accounts
for 40.6-44.2% of the attraction.
The computed charges of +0.198 and +0.195 on the
[(NN)PtMe2] fragments of 4a and 5a indicate an important
charge transfer character of the d8-d10 interactions in these
compounds, in agreement with their description as PtfM dative
bonds. Structures 8a2 (q((NN)PtPh2) ) +0.067) and 9a2
(q((NN)PtPh2) ) +0.118) exhibit less charge transfer, which
is probably compensated by the polarization of the π-electron
density of the phenyl groups. An intermediate situation is found
for structures 8a1 (q((NN)PtPh2) ) +0.126) and 9a1 (q((N-
2
N)PtPh2) ) +0.158), in which some interaction of the dz orbital
of Pt(II) with the coinage metal is expected due to its location
above the plane of the ligands.
To check whether the activation of 2 toward C-H bond
breaking could proceed through a reversible transfer of one
methyl group to Cu(I) or Ag(I),34 we conducted methyl
scrambling experiments. A benzene solution containing ap-
proximately equal amounts of 2 and [(NN)Pt(CD3)2] (2-d6), and
an excess CuOTf, was stored at room temperature for 7 days.
ESI-MS analysis (see Supporting Information) revealed the slow
formation of the mixed species [(NN)Pt(CH3)(CD3)] (2-d3), with
concomitant C-H activation reaction leading to 3. However,
no methyl scrambling was observed under the same conditions
with AgOTf, although the C-H activation reaction was found
The two binding modes of the diphenyl complex 3 also differ
by the magnitude of the preparation energy term, which is 2-3
times higher for 8a1 and 9a1 as for 8a2 and 9a2. This difference
is due to the out-of-plane bending of one of the Pt-C bonds
that is required to maximize the interaction between the coinage
metal and both the Pt(II) center and the phenyl ligand.
Another prominent feature of the data presented in Table 4
is the fact that PtfM bonds are predicted to be significantly
stronger for M ) Cu(I) than for M ) Ag(I). This may appear
surprising in view of the small number of reported compounds
incorporating PtfCu bonds, as compared to the large body of
literature dealing with PtfAg bonds. These observations
illustrate the fact that trends in the strength of a particular bond
do not necessarily correlate with the overall stability of the
corresponding compounds. In fact, compound 4a, for example, is
thermally less stable than 5a in solution due to its easier dispro-
portionation into Cu(0), Pt(II) and Pt(IV) products (Vide infra).
C-H Activation Reactivity. In order to probe for the reactivity
of compounds incorporating PtfM bonds in relation with C-H
bond activation, we studied the thermolysis of complexes 4a
and 5a in benzene (Scheme 5). Heating a solution of 4a prepared
in situ in benzene for 3 days at 50 °C, followed by filtration
over a short column of activated alumina, yielded 41% of the
double C-H activation product 3 and 4% of free ligand 1. The
same treatment applied to the silver(I) complex 5a afforded a
Figure 15. ESI-MS spectrum obtained from a reaction mixture containing
2 and one equivalent of CuOTf in benzene after 17 h at 50 °C, diluted in
acetonitrile.
(67) Nechaev, M. S.; Rayon, V. M.; Frenking, G. J. Phys. Chem. A 2004,
108, 3134–3142.
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5685

A R T I C L E S
Moret and Chen
Scheme 6
to proceed. Furthermore, the methyl-transfer pathway does not
account for the appearance of Pt(IV) species in the ESI-MS
spectra, and thus an alternative mechanism has to be involved.
We propose the mechanism presented in Scheme 6. The
reaction is initiated by one-electron oxidation of the Pt(II)
complex 2 by the coinage metal triflate (Scheme 6(I)), which
leads to a postulated transient Pt(III) intermediate that dispro-
portionates into the Pt(IV) complex 17 and the activated Pt(II)
complex 18. This initiation reaction consumes only a small
fraction of the starting material. The next reaction is the actual
double C-H activation of benzene promoted by species 18
(Scheme 6(II)). The labile triflate anion is exchanged for a
benzene molecule to form the cationic intermediate A, which
undergoes oxidative addition of the Ph-H bond to afford the
Pt(IV) intermediate B. Reversible deprotonation of B accounts
for the observation of the mixed complex 19. Alternatively, B
can undergo reductive elimination of methane followed by
complexation of a second benzene molecule to produce inter-
mediate C. A second oxidative addition step yields the Pt(IV)
hydride complex D, which is able to transfer a proton to the
starting material 2, releasing the product 3. Reductive elimina-
tion of methane and its displacement by a benzene molecule
regenerates intermediate A, allowing for the production of
several molecules of 3 for one molecule of 18 generated in the
initiation step I. It should be noted that the proton transfer
reaction D + 2 f 3 + E does not have to proceed in a single
step, but may also involve protonation of the triflate anion by
D to transiently form triflic acid, which would then protonate 2
to form intermediate E.
of this pathway for 2 was shown by oxidizing 2 with one
equivalent of ferrocenium triflate in acetone-d6, affording
equimolar amounts of 17 and 18. The assignment of the products
1
was made by comparison of the H NMR spectrum with that
of independently synthesized 17 and that of 18 generated in
situ from [(NN)PtMeCl] (20) and AgOTf. To prove that one-
electron oxidation chemistry is sufficient to explain the observed
double C-H activation reaction, we heated a benzene solution
containing compound 2 and 10 mol% of ferrocenium triflate to
1
50 °C for 3 days. H NMR analysis of the reaction mixture
revealed that nearly quantitative formation of the diphenyl
complex 3 had occurred. Additionally, reduced ferrocene and
a small amount of the Pt(IV) trimethyl complex 17 were
observed, indicating that the latter is not responsible for the
C-H activation chemistry.
The methyl scrambling observed for the copper(I) complex
4a can be explained by the formation of trace amounts of the
Pt(IV) complex 17 Via the redox reaction depicted in Scheme
6(I). Despite its low solubility in benzene, compound 17 was
found to be able to transfer methyl groups to 2-d6, presumably
Via an SN2-type reaction of the nucleophilic Pt(II) center of 2
and an electrophilic methyl group of 1770,71 (Scheme 7). This
reaction is expected to proceed only for the free complex 2 in
equilibrium with 4a, since complexation to a Lewis-acidic metal
should strongly decrease the nucleophilicity of the Pt(II) center.
The fact that no methyl scrambling is observed for the silver
complex 5a is thus attributed to its stronger association constant
(Vide supra).
Related double C-H activation reactions have been observed
for anionic complexes of type [(LL)PtMe2]- (LL ) dimethyl-
bis(2-pyridyl)borate,72 diphenylbis(1-pyrazolyl)borate73), and for
a homodinuclear Pt(II) complex.74 The proposed mechanisms
The one-electron oxidation chemistry proposed for the
activation step (I) parallels that observed upon chemical or
electrochemical oxidation of closely related complexes in
acetonitrile.58,68,69 Furthermore, a related process has been
observed in the reaction of [(bpy)PtMe2] with CuCl, yielding
[(bpy)PtMe3Cl], [(bpy)PtMeCl] and Cu(0).34 The availability
(70) Aye, K. T.; Canty, A. J.; Crespo, M.; Puddephatt, R. J.; Scott, J. D.;
Watson, A. A. Organometallics 1989, 8, 1518–1522.
(71) Hill, G. S.; Yap, G. P. A.; Puddephatt, R. J. Organometallics 1999,
18, 1408–1418.
(68) Wik, B. J.; Tilset, M. J. Organomet. Chem. 2007, 692, 3223–3230.
(69) Johansson, L.; Ryan, O. B.; Rømming, C.; Tilset, M. Organometallics
1998, 17, 3957–3966.
(72) Khaskin, E.; Zavalij, P. Y.; Vedernikov, A. N. J. Am. Chem. Soc.
2006, 128, 13054–13055.
(73) Thomas, C. M.; Peters, J. C. Organometallics 2005, 24, 5858–5867.
9
5686 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
Scheme 7
Experimental Section
for these reactions parallel the one depicted in Scheme 6(II),
the main difference being that the active Pt(II) methyl species
is generated by protolysis72-74 or methide abstraction73 from
the dimethyl complexes, whereas here it is produced by one-
electron oxidation.
General Procedures. Solvents were obtained commercially in
p.a. quality and used as received. For reactions involving metal
complexes, solvents were distilled under nitrogen over sodium
(hexane, toluene), potassium (tetrahydrofuran, benzene), CaH2
(acetonitrile, dichloromethane). All chemical manipulations involv-
ing metal complexes were performed under an inert atmosphere
using standard Schlenk and glovebox techniques unless otherwise
stated. Deuterated solvents were dried through a short column of
Conclusions
The electron-poor R-diimine ligand 1 was found to be efficient
for the stabilization of structurally simple compounds formed
by the interaction of organoplatinum(II) complexes with monov-
alent coinage metal triflates, without the need for an extra
stabilizing or bridging ligand.
activated alumina immediately before use. H and 13C chemical
1
shifts are reported in ppm relative to tetramethylsilane, using
residual solvent proton and 13C resonances as internal standards.
Multiplicities are indicated s (singlet), d (doublet), t (triplet), dd
(double doublet), while apparent singlets, doublets and triplets are
indicated by “s”, “d”, and “t”, respectively. Elemental analyses were
carried out at the Mikrolabor of the Laboratorium fu¨r Organische
Chemie of ETH Zu¨rich. Organolithium reagents were titrated using
the procedure from Watson and Eastham.75 Precursors cis-
PtPh2(SMe2)237 and trans-(Me2S)2Pt(Me)Cl,76 R-diimine ligand 154
and complex 254 were prepared according to literature procedures.
The diphenylplatinum(II) complex 3 was synthesized by a modi-
fication of the procedure by Gerdes.56 All other chemicals were
obtained commercially and used as received.
The dimethylplatinum complex 2 binds to the coinage metal
through short, unsupported PtfM bonds in the solid phase, in
solution, and in the gas phase. The CuOTf adduct 4b exhibits
the shortest Pt-Cu contact known to date. DFT calculations
shows that these interactions are essentially electrostatic in
nature, but with a significant contribution from orbital interac-
tions. In particular, a significant amount of charge transfer is
computed, in accord with the observation of an important
1
downfield shift of the ligand methyl H NMR resonance and
1
the blue shift of the MLCT bands in the UV-vis absorbance
(NN)PtPh2 (3). A mixture of 1 (897 mg, 1.89 mmol) and cis-
PtPh2(SMe2)2 (711 mg, 1.90 mmol) in toluene (20 mL) was stirred
for 1 day. The solvent was then removed in Vacuo, and toluene
(20 mL) was added. This cycle was repeated three times, and the
mixture was stirred for 3 days. Removal of the solvent in Vacuo,
washing of the residue with toluene (3 × 5 mL), and drying in
Vacuo yielded the product as a purplish-gray powder (1.21 g, 88%).
1H NMR (CD2Cl2, 300 MHz): δ 7.19 (d, 3J(H-H) ) 8.0 Hz, 4H,
spectra upon coordination. Thermolysis of the complexes in
benzene produces the diphenyl complex 3 in moderate yield
through a double C-H activation reaction. We propose that the
mechanism of these reactions is similar to known double C-H
activation processes,72-74 except for the fact that in our case
the active cationic Pt(II) species is generated through an
oxidative process rather than protolysis or methide abstraction.
Electrostatic considerations based on DFT calculations predict
a more complex binding situation for the diphenylplatinum
complex 3: the metal cation is able to bind either to the platinum
center through a PtfM bond or to the phenyl groups. In the
solid state, we found that the latter is preferred in the homoleptic
2:1 complexes 6a and 7a as well as in the 2:6 complex 12,
while both binding sites are occupied in the 1:2 complex 11b.
Spectroscopic studies in solution revealed the existence of
several complexes of 3 with coinage metal triflates, namely the
2:1 complexes 6a and 7a, the 1:1 complexe 8a, the 2:2 complex
9b, and possibly the 1:2 complex 10a with copper(I) triflate at
higher concentrations.
NArH3,5), 7.00-6.93 (m, 2H, NArH4), 6.95 (dd, J(H-H) ) 7.8
3
Hz, 5J(H-H) ) 2.2 Hz, 3J(Pt-H) ) 70.9 Hz, 4H, PtArH2,6),
6.61-6.49 (m, 4H, PtArH3,5), 6.44 (tt, 3J(H-H) ) 7.2 Hz, 5J(H-H)
) 1.4 Hz, 2H, PtArH4), 1.64 (s, 6H, CH3). Anal.: calcd for
C28H22N2Cl4Pt: C 46.49, H 3.07, N 3.87. Found: C 46.21, H 3.15,
N 3.65.
[(NN)PtMe2]CuOTf (4a). A mixture of 2 (50 mg, 83 µmol)
and CuOTf ·1/2C7H8 (24 mg, 93 µmol) in toluene (5 mL) was stirred
for 30 min. A small amount of brown precipitate was removed by
filtration, yielding a dark-red solution. Hexane (10 mL) was added,
and the solution was stored at -20 °C for 7 days, during which
time dark red needles formed. Solvent removal by decantation,
washing with hexane (4 × 2 mL) and drying in Vacuo yielded the
product as dark, brownish-red needles (31 mg, 46%). Anal.: calcd
for C19H18N2O3F3SCl4CuPt: C 28.11, H 2.23, N 3.45. Found: C
28.51, H 2.30, N 3.45. Crystals suitable for XRD were obtained at
-35 °C from a freshly prepared solution of 4a in 1:1 benzene/
hexane; benzene precipitation initiated the crystallization, and the
crystals that formed on the walls of the flask were mechanically
separated from the solid benzene.
These results demonstrate the ability of simple organoplatinum
compounds to act as ligands for Lewis-acidic metals. The physical
properties of the platinum complexes change markedly upon
coordination to monovalent coinage metal cations, and the Pt-M
bonds have been found to be strong enough to support ligand
exchange processes. Thus, we anticipate that complexes similar
to the ones described in this study might allow for the observation
of novel chemical processes based on d8-d10 interactions. This is
the subject of further investigations in our laboratories.
(75) Watson, S. C.; Eastham, J. F. J. Organomet. Chem. 1967, 9, 165–
168.
(76) Hill, G. S.; Irwin, M. J.; Levy, C. J.; Rendina, L. M.; Puddephatt,
R. J. Inorg. Synth. 1998, 32, 149–152.
(74) Song, D.; Jia, W. L.; Wang, S. Organometallics 2004, 23, 1194–1196.
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5687

A R T I C L E S
Moret and Chen
[(NN)PtMe2]2Ag2(OTf)2 ·2C6H6 (5b·2C6H6). A mixture of 2 (60
mg, 0.1 mmol) and AgOTf (26 mg, 0.1 mmol) was stirred in
benzene (7 mL) for 1 h, during which time a small amount of a
dark-red precipitate appeared. The mixture was concentrated to 2
mL and filtered. Washing with benzene (3 × 1 mL) and drying in
Vacuo afforded the product as a purple powder containing two
equivalents of benzene (60 mg, 64%). Anal: calcd for
C38H36N4O6F6SCl8Ag2Pt2 ·2C6H6: C 32.14, H 2.59, N 3.00. Found:
C 31.62, H 2.50, N 2.89. Crystals suitable for XRD were obtained
from vapor diffusion of hexane into a benzene solution of 2 in the
presence of a large excess of AgOTf. Solvated AgOTf crystallized
as white needles at the same time.
trans-(Me2S)2Pt(Me)Cl (198 mg, 0.53 mmol) were dissolved in THF
(20 mL). The flask was heated, resulting in a color change from
yellow to dark purple, and the temperature was adjusted so that
THF slowly distilled (∼20 mL/h). Heating was continued over a
period of 12 h, and THF was regularly added to keep the reaction
volume above 10 mL. Filtration, washing with THF (3 × 2 mL)
and drying in Vacuo afforded the product as a dark-purple,
microcrystalline solid (196 mg, 60%). 1H NMR (CD2Cl2, 300
MHz): δ 7.59 (2H), 7.53 (2H), 7.34 (1H), 7.29 (1H) (2 × A2B,
CCH3), 1.19 (s, 4J(Pt-H) ) 9.3 Hz, 3H, CCH3), 0.94 (s, 2J(Pt-H)
) 80.8 Hz, 3H, PtCH3). 13C NMR (CD2Cl2, 75 MHz): δ 176.2
(C)NAr), 174.8 (C)NAr), 142.2 (CAr), 142.0 (CAr), 129.7 (CAr),
129.24 (CAr), 129.18 (CAr), 128.9 (CAr), 128.7 (CAr), 127.9 (CAr),
21.6 (CCH3), 20.3 (CCH3), -13.4 (PtCH3). Anal.: calcd for
C17H15N2Cl5Pt: C 32.95, H 2.44, N 4.52. Found: C 32.64, H 2.50,
N 4.51.
3
3J(H-H) ) 8.2 Hz, J(H-H) ) 8.2 Hz, NArH3,4,5), 1.72 (s, 3H,
{[(NN)PtPh2]2Cu}+TfO- (6a). A mixture of 3 (100 mg, 0.138
mmol) and CuOTf ·1/2C7H8 (20 mg, 0.077mmol) in dichlo-
romethane (4 mL) was stirred for 10 min and filtered. Toluene (10
mL) was added to the resulting dark-purple solution, which was
slowly concentrated under a flow of argon over 18 h. Filtration,
washing with toluene (3 × 2 mL) and drying in Vacuo yielded the
6a as black needles suitable for XRD (60 mg, 52%). Anal: calcd
for C57H44N4O3F3SCl8CuPt2: C 41.26, H 2.67, N 3.38. Found: C
41.45, H 2.64, N 3.38.
(NN)Pt(Me)(O3SCF3) (18). A mixture of 20 (12 mg, 19 µmol)
and AgOTf (5 mg, 19 µmol) in acetone-d6 was stirred for 5 min,
during which time its color changed from dark purple to orange. A
slightly grayish precipitate (AgCl) was removed by filtration, and
the resulting orange solution was characterized by 1H NMR
spectroscopy. 1H NMR (CD2Cl2, 300 MHz): δ 7.78 (2H), 7.76 (2H),
{[(NN)PtPh2]2Ag}+TfO- (7a). A mixture of 3 (50 mg, 69 µmol)
and AgOTf (10 mg, 39 µmol) in dichloromethane (1.5 mL) was
stirred for 2 min. Filtration and washing with dichloromethane (2
× 0.5 mL) yielded a deep-purple solution, to which toluene (4 mL)
was added. Crystallization started after a few minutes, and the
solution was kept at -20 °C overnight. Filtration, washing with
toluene (3 × 1 mL), and drying in Vacuo yielded the product as
dark-purple plates (36 mg, 61%). Anal: calcd for
C57H44N4O3F3SCl8AgPt2: C 40.18, H 2.60, N 3.29. Found: C 40.28,
H 2.72, N 3.40. Crystals suitable for XRD were obtained upon slow
concentration of a solution of 7b in THF/benzene or DCM/toluene.
{[(NN)PtPh2]Ag2(OTf)2(C6H6)}n ·2C6H6 (11b ·2C6H6). A solu-
tion of 3 (20 mg, 28 µmol) and AgOTf (20 mg, 78 µmol) in 1:1
dichloromethane/benzene was slowly concentrated under a flow of
argon, resulting in the formation of red crystals suitable for XRD.
Washing with hexane and drying in Vacuo yielded the compound
as a red solid (31 mg, 75%). The crystals lost about 0.5 equiv of
benzene upon drying, as shown by elemental analysis. Anal.: calcd
for C36H28N2O6F6S2Cl4Ag2Pt·1.5C6H6: C 37.73, H 2.60, N 1.96.
Found: C 37.89, H 2.57, N 1.90.
{[(NN)PtPh2]2Cu6(OTf)6}·6CD2Cl2 (12·6CD2Cl2). A mixture
of 3 (7 mg, 10 µmol) and CuOTf ·1/2C6H6 (10 mg, 40 µmol) in
CD2Cl2 (0.7 mL) was stirred for 5 min and the excess CuOTf ·1/
2C6H6 was removed by filtration, resulting in a dark-purple solution
which was transferred into a Young-type NMR tube. The solution
was left standing at 20 °C for 3 h and then cooled at 4 °C for 1 h,
causing the crystallization a small amount of 12 as dark-purple
needles which were suitable for XRD. No clean elemental analysis
was obtained due to the concomitant formation of a grayish
precipitate (presumably CuOTf) that could not be separated from
12.
(NN)Pt(Me)3(CF3SO3) (17). Methyl triflate (22 µL, 0.20 mmol)
was added to a dark blue solution of 2 (100 mg, 0.167 mmol) in
dichloromethane (5 mL), resulting in a rapid color change to dark
orange. The resulting solution was layered with hexane (10 mL).
After 2 days, filtration, washing with hexane (3 × 1 mL) and drying
in Vacuo yielded the product as brown microcrystals (95 mg, 75%).
1H NMR (CD2Cl2, 300 MHz): δ 7.56, 7.31 (A2B, 3J(H-H) ) 8.2
Hz, 6H, NArH3,4,5), 2.46 (s, 6H, CCH3), 1.04 (s, 2J(Pt-H) ) 74.9
Hz, 9H, PtCH3). 13C NMR (CD2Cl2, 75 MHz): δ 180.2 (br,
C)NAr), 139.5 (CAr), 129.7-129.3 (br m, CAr), 127.7 (CAr), 23.4
(CCH3), 23.3 (CCH3), -4.1-5.6 (br, PtCH3). Anal.: calcd for
C20H21N2O3F3SCl4Pt: C 31.47, H 2.77, N 3.67. Found: C 31.26, H
2.88, N 3.54. Crystals suitable for XRD were obtained from vapor
diffusion of hexane into a freshly prepared solution of 17 in 1:1
dichloromethane/hexane.
3
3
7.59 (1H), 7.54 (1H) (2 × A2B, J(H-H) ) 8.2 Hz, J(H-H) )
8.2 Hz, NArH3,4,5), 2.32 (s, 3H, CCH3), 2.15 (s, J(Pt-H) ≈ 11
4
2
Hz, 3H CCH3), 0.57 (s, J(Pt-H) ) 74.8 Hz, 3H, PtCH3).
Thermolysis of 4a and 5a. Equimolar amounts of 2 and AgOTf
or CuOTf ·1/2C7H8 were dissolved in benzene (3 mL) and heated
to 50 °C for 3 days in a closed vessel. The resulting mixture was
passed over a short column of aluminum oxide. The undissolved
solids were taken in dichloromethane and passed over the same
column, which was thoroughly washed with dichloromethane until
the washings became colorless. Solvent was evaporated under
reduced pressure from the resulting black solution, and the residue
1
was weighed and analyzed by H NMR.
UV-Vis Spectroscopy. UV-vis spectra were measured on a
Hitachi U-2010 instrument. Samples were prepared in a nitrogen-
filled glovebox and introduced into Teflon-stoppered quartz cells
with a path length of 1 cm. Measurements involving 2 were
performed in benzene, while those involving 3 were conducted in
1:1 dichloromethane/benzene to ensure solubility of the reactants
and products. Samples containing AgOTf were prepared by
appropriate dilution of ∼2 mM solutions of 2, 3 and AgOTf. Due
to the low solubility of CuOTf, mother solutions were prepared by
addition of a weighed amount of CuOTf ·1/2C7H8 to 5 mL of a ∼2
mM solution of 2 or 3. In the case of 2, this solution is only
moderately stable at room temperature and was used within 3 h.
PFG NMR Spectroscopy. Diffusion coefficients were measured
by pulsed-field-gradient NMR (PFG NMR) on ∼14 mM solutions
in CD2Cl2 at 25 °C using the pulse sequence of Wu et al.,77 which
is based on stimulated echo (STE)78 and uses longitudinal-eddy-
current delay (LDE)79 and bipolar gradient pulses. The absolute
gradient strength was calibrated against 1% H2O in D2O, which
has a diffusion coefficient of 18.7 × 10-10 m2 s-1. Diffusion
coefficients (D) were extracted using the TOPSPIN80 program.
Hydrodynamic volumes were calculated as described by Mac-
chioni et al.,81 using a modified Stokes-Einstein relation:
kT
D )
(1)
c(rsolv, rH)πηrH
where η ) 4.23·10-4 Pa·s is the viscosity of dichloromethane, rH
the hydrodynamic radius of the solute, rsolv ) 2.49 Å the
(77) Wu, D.; Chen, A.; Johnson, C. S. J. Magn. Reson. A 1995, 115, 260–
264.
(78) Tanner, J. E. J. Chem. Phys. 1970, 52, 2523–2526.
(79) Gibbs, S. J.; Johnson, C. S. J. Magn. Reson. 1991, 93, 395–402.
(80) TOPSPIN, 2.1.1; Bruker: Rheinstetten, 2008.
(81) Macchioni, A.; Ciancaleoni, G.; Zuccacia, C.; Zuccacia, D. Chem.
Soc. ReV. 2008, 37, 479–489.
(NN)Pt(Me)Cl (20). In a round bottomed Schlenk flask fitted
with a Vigreux distillation apparatus, 1 (200 mg, 0.53 mmol) and
9
5688 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009

Organoplatinum(II) Interactions with Coinage Metal Triflates
A R T I C L E S
Table 5. Crystallographic Data for Compounds 4a, 5b, and 6a
4a
5b
6a
chem formula
FW
C19H18Cl4CuF3N2O3PtS
811.86
C50H48Ag2Cl8F6N4O6Pt2S2
1868.56
C57H44Cl8CuF3N4O3Pt2S
1659.4
cryst syst
space group
a [Å]
b [Å]
c [Å]
triclinic
P1
triclinic
P1
monoclinic
C2/c
19.8903(4)
18.8114(5)
17.0905(3)
90.00
110.0366(13)
90.00
6007.6(2)
4
j
j
9.2087(4)
10.9551(5)
14.8595(8)
110.493(3)
92.371(3)
108.700(2)
1309.98(11)
2
2.058
776
6.679
173
10.6556(12)
12.1622(13)
12.6936(13)
110.231(12)
105.478(11)
90.209(10)
1479.2(3)
1
2.098
896
5.868
220
R [deg]
ꢀ [deg]
γ [deg]
V [Å3]
Z
Dcalcd [g cm-3
F(000)
]
1.835
3208
5.443
203
µ [mm-1
]
temp. [K]
wavelength [Å]
measd rflns
0.71073
14635
0.71070
12197
0.71073
18502
unique rflns
5959
6777
6866
data/restraints/param
R(F) (I > 2σ(I))
wR(F2) (I > 2σ(I))
GOF
5959/0/311
0.0816
0.2013
6777/0/366
0.0302
0.0786
6866/7/356
0.0381
0.1085
1.174
1.026
0.860
Table 6. Crystallographic Data for Compounds 11b and 12
11b 12
hydrodynamic radius of dichloromethane, and c(rsolv,rH) a correction
factor that accounts for the relative size of the solvent and solute
molecules. We used the expression of c(rsolv,rH) introduced by
Chen:82
chem formula
FW
cryst syst
space group
a [Å]
C48H40Ag2Cl4F6N2O6PtS2 C34H28Cl10Cu3F9N2O9PtS3
1471.59
orthorhombic
Pnma
1616.035
triclinic
P1
6
rsolv
(
j
c )
(2)
26.1863(5)
19.3506(3)
10.1987(2)
90.00
90.00
90.00
12.6180(2)
12.8711(3)
17.6498(4)
77.4279(11)
80.5915(10)
69.5194(9)
2608.91(9)Å3
2
1 + 0.695
2.234
b [Å]
c [Å]
)
rH
R [deg]
ꢀ [deg]
γ [deg]
V [Å3]
Z
ESI-MS. ESI-MS spectra were recorded on a ThermoFinnigan
TSQ Quantum instrument. Ions with masses above 1500 were
studied on a Finnigan TSQ700 instrument. CID spectra were
measured using argon as collision gas.
5167.9(2)
4
Dcalcd [g cm-3
F(000)
]
1.891
2.057
{[(NN)PtR2]M(NCMe)}+ and{ [(NN)PtR2]2M}+ (R ) Me, Ph;
M ) Cu, Ag) were electrosprayed from solutions freshly prepared
as follows: [(NN)PtR2] (∼1 mg) and MOTf (∼1 mg) were dissolved
in acetonitrile (2 mL), and the resulting solution was diluted 20
times.
2864
1564
µ [mm-1
]
3.809
4.59
temp. [K]
223
223
wavelength [Å]
measd rflns
unique rflns
0.71073
23572
5988
0.71073
36432
11895
X-ray Crystallography. Relevant details about the structure
refinements are given in Tables 5 and 6, and selected geometrical
parameters are included in the captions of the corresponding figures.
Data collection was performed on a Bruker-Nonius Kappa-CCD
(graphite monochromator, Mo KR). The structures were solved by
direct methods83 and refined by full-matrix least-squares analysis.84
All non-H atoms were refined anisotropically unless otherwise stated
in the figure caption. Hydrogen atomic positions are based on
data/restraints/param 5988/0/332
R(F) (I > 2σ(I)) 0.0517
wR(F2) (I > 2σ(I)) 0.1608
GOF 1.079
11895/0/642
0.0522
0.1432
1.033
tions performed with the Gaussian03 program. Due to the flat
character of the potential energy surfaces encountered, the GDIIS
algorithm was used systematically and “tight” convergence criteria
were imposed. In one case (structure 9a2), “tight” convergence
could not be achieved, and only convergence with the normal
criteria was obtained. A full frequency calculation was performed
on each structure to ensure that they are true minima.
-
stereochemical considerations. The disordered CF3SO3 anions in
structures 6a and 11b were refined over two positions.
DFT Calculations. Geometry optmizations were performed
using the Gaussian03 package.85 The BP86 exchange-correlation
functional was employed. The Stuttgart/Dresden62,63 basis set and
effective core potential were used for transition metals, along with
a 6-31+G(d) basis set on all other atoms. Diffuse functions on main
group elements were found to be necessary for an accurate
description of structure 7a and thus were included for all calcula-
Energy partitioning analyses66 were performed with the ADF2008
package86-88 on geometries optimized as described above. The
BP86 exchange-correlation functional was used, and scalar rela-
tivistic effects were included Via the zeroth order regular ap-
(82) Chen, H.-C.; Chen, S.-H. J. Phys. Chem. 1984, 88, 5118–5121.
(83) Altomare, A.; Burla, M.; Camalli, M.; Cascarano, G.; Giacovazzo,
C.; Guagliardi, A.; Moliterni, A.; Polidori, G.; Spagna, R. J. Appl.
Crystallogr. 1999, 32, 115–119.
(84) Sheldrick, G. M. SHELXL-97 Program for the Refinement of Crystal
Structures; University of Goettingen: Germany, 1997;
(85) Frisch, M. J.; et al. Gaussian 03, Revision D.01; Gaussian, Inc.:
Wallingford, CT, 2004.
(86) te Velde, G.; Bickelhaupt, F. M.; van Gisbergen, S. J. A.; Fonseca
Guerra, C.; Baerends, E. J.; Snijders, J. G.; Ziegler, T. J. Comput.
Chem. 2001, 22, 931–967.
(87) Fonseca Guerra, C.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor.
Chem. Acc. 1998, 99, 391.
(88) ADF2008. 01, SCM, Theoretical Chemistry.Vrije Universiteit: Am-
sterdam, The Netherlands, 2008; http://www.scm.com.
9
J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009 5689

A R T I C L E S
Moret and Chen
proximation (ZORA) formalism. A TZ2P basis set was used and
the electrons up to 1s for C, N, O, F, 2s for S and Cl, 3s for Cu,
4s for Ag, and 5d for Pt were treated within the frozen core
approximation. The total bond energy is divided in two parts (eq
3):
cuppied orbitals of the other) and polarization contributions (the
interactions between empty and filled orbitals of a single fragment
caused by the presence of the other). The charges associated to
each fragment were computed by the Hirshfeld89 method.
Acknowledgment. We acknowledge support from the Swiss
National Foundation. We thank Dr. B. Schweizer and Mr. P. Seiler
for the XRD structure determinations and Dr. M.-O. Ebert for
assistance with the PFG NMR diffusion measurements.
∆E ) ∆Eprep + ∆Eint
(3)
where ∆Eprep (positive) is the energy required to deform the
fragments from their optimized geometries to the geometry they
adopt in the complex, and ∆Eint (negative) is the actual change in
energy that occurs when the fragments are combined together in
the complex. ∆Eint can be divided further according to eq 4:
Supporting Information Available: X-ray crystallographic
data in CIF format, ORTEP representations of the structures of
7a and 17, UV-vis data relative to the solvatochromism of 2
and 3, 1H NMR data relative to the interaction of 3 and AgOTf,
experimental and fitted isotope patterns for methyl scrambling
experiments, XYZ coordinates for DFT-optimized geometries,
and complete reference 85 (ref 2 in SI).This material is available
free of charge via the Internet at http://pubs.acs.org.
∆Eint ) ∆EPauli + ∆Eelstat + ∆Eorb
(4)
where ∆EPauli (positive) accounts for the destabilizing interactions
between occupied orbitals and can be related to steric repulsion,
∆Eelstat (negative) corresponds to the electrostatic interaction
between the prepared fragments before relaxation of the electron
density, and ∆Eorb (negative) is the change in energy occurring when
the electron density is allowed to relax to the minimal energy. This
last term contains both charge transfer (the donor-acceptor
interactions between occupied orbitals of one fragment and unno-
JA900449Y
(89) Hirshfeld, F. L. Theor. Chim. Acta 1977, 44, 129–138.
9
5690 J. AM. CHEM. SOC. VOL. 131, NO. 15, 2009