Ligand binding energies in cationic platinum(II) complexes: A quantitative study in the gas phase

Article abstract of DOI:10.1021/om0610039

Quantitative energy-resolved reactive cross-section measurements and DFT calculations are used to investigate the binding energies of various ligands to the metal center in cationic Pt(II) diimine complexes involved in C-H activation reactions. Independent synthesis of isomeric ions establishes the structure of the adducts formed in gas-phase reactions. The order and relative magnitude of a series of ligand binding energies extracted from experimental cross-sections reproduce the trends seen in solution-phase experiments only when the proper transition state model is employed in the deconvolution of the experimental crosssections. The choice of transition state model can be rationalized by qualitative structural arguments as well as quantum chemical calculations of the reaction coordinate for the dissociation of the range of possible ligands.

Full text of DOI:10.1021/om0610039

Organometallics 2007, 26, 1523-1530
1523
Ligand Binding Energies in Cationic Platinum(II) Complexes: A
Quantitative Study in the Gas Phase
Marc-Etienne Moret and Peter Chen*
Laboratorium fu¨r Organische Chemie, Eidgeno¨ssische Technische Hochschule Zu¨rich (ETHZ),
Zu¨rich, Switzerland
ReceiVed October 31, 2006
Quantitative energy-resolved reactive cross-section measurements and DFT calculations are used to
investigate the binding energies of various ligands to the metal center in cationic Pt(II) diimine complexes
involved in C-H activation reactions. Independent synthesis of isomeric ions establishes the structure of
the adducts formed in gas-phase reactions. The order and relative magnitude of a series of ligand binding
energies extracted from experimental cross-sections reproduce the trends seen in solution-phase experiments
only when the proper transition state model is employed in the deconvolution of the experimental cross-
sections. The choice of transition state model can be rationalized by qualitative structural arguments as
well as quantum chemical calculations of the reaction coordinate for the dissociation of the range of
possible ligands.
Scheme 1. C-H Activation Reaction by Diimine Pt(II)
Introduction
Complexesa
The development of efficient catalytic transformations of
alkanes is one of the most prominent challenges for today’s
organometallic chemists, due to both the difficulty of function-
alizing relatively inert C-H bonds and the appealing industrial
perspectives it would open, such as the use of methane and
simple alkanes and arenes as chemical feedstock.1 Following
the early work of Shilov2 and Garnett,3 C-H activation by
platinum(II) complexes was extensively studied with respect
to both mechanism4 and applications.5 In particular, cationic
complexes incorporating chelating diimine ligands have been
found6,7 to activate both methane and benzene under mild
conditions as shown in Scheme 1.
The mechanism of this reaction has been addressed by means
of NMR,6,8-10 UV-vis,10,11 and mass spectroscopic methods.11
Recently, we reported a quantitative study of ligand binding
energies to the [(N-N)PtCH3]+ fragment by energy-resolved
reactive cross-section measurements of the ions of interest.12
Given that both solution-phase and gas-phase studies of the
C-H activation of benzene and methane by Pt(II) diimine
complexes find that ligand exchange is rate-limiting, one can
argue that a careful characterization of this step is a major
component in the further understanding, development, and
a
In this study, Ar ) 2,6-dichlorophenyl.
optimization of C-H activation chemistry. Extending the range
of ligands in our gas-phase thermochemical experiments from
the coordinating solvents, acetonitrile, water, and 2,2,2-trifluo-
roethanol (TFE), to hydrocarbon substrates, benzene, in par-
ticular, we observed what appeared at first to be a thermochem-
ical anomaly, which raised two questions: (i) What is the
structure of the gas-phase adduct prepared by reaction of [(N-
N)PtCH3]+ and benzene? (ii) Do the ligand dissociations proceed
by structurally similar transition states for all of the potential
ligands? Addressing these two questions, we present a quantita-
tive study of the potential energy surface (PES) of the ligand
dissociation step in the C-H activation reaction in the gas phase,
based on both experimental and computational methods.
Experimental Section
* Corresponding author. E-mail: chen@org.chem.ethz.ch.
(1) Labinger, J. A.; Bercaw, J. E. Nature 2002, 417, 507.
(2) Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879.
(3) Garnett, J. L.; Hodges, R. J. J. Am. Chem. Soc. 1967, 89, 4546.
(4) Lersch, M.; Tilset, M. Chem. ReV. 2005, 105, 2471.
(5) Periana, R. A.; Bhalla, G.; W. J. T., III; Young, K. J. H.; Liu, X. Y.;
Mironov, O.; Jones, C. J.; Ziatdinov, V. R. J. Mol. Catal. A 2004, 220, 7.
(6) Heiberg, H.; Johansson, L.; Gropen, O.; Ryan, O. B.; Swang, O.;
Tilset, M. J. Am. Chem. Soc. 2000, 122, 10831.
(7) Johannsson, L.; Ryan, O. B.; Tilset, M. J. Am. Chem. Soc. 1999,
121, 1974.
(8) Wik, B. J.; Lersch, M.; Krivokapic, A.; Tilset, M. J. Am. Chem. Soc.
2006, 128, 2682.
(9) Zhong, H. A.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2002,
124, 1378.
(10) Johansson, L.; Tilset, M.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem.
Soc. 2000, 122, 10846.
(11) Gerdes, G.; Chen, P. Organometallics 2003, 22, 2217.
(12) Hammad, L. A.; Gerdes, G.; Chen, P. Organometallics 2005, 24,
1907.
General Procedures. Diethyl ether was distilled from Na/K
alloy, dichloromethane from CaH2, and 1,1,1-trifluoroethanol (TFE)
from MgSO4 and NaHCO3, all under a dry nitrogen atmosphere.
All chemical manipulations involving platinum complexes were
performed under an inert atmosphere using standard Schlenk and
glovebox techniques unless otherwise stated. H and 13C NMR
1
spectra were recorded on Varian Gemini 300, Varian Mercury 300,
1
and Bruker ARX 300 instruments. H and 13C chemical shifts are
reported in ppm relative to tetramethylsilane, using the residual
solvent proton resonance as internal standard. Elemental analyses
were carried out by the Mikrolabor of the Laboratorium fu¨r
Organische Chemie of ETH Zu¨rich. ArNC(Me)C(Me)N)Ar (Ar )
2,6-dichlorophenyl) (6) and the corresponding (N-N)Pt(CH3)2
complex were prepared as described previously.11 A mixture of
cis- and trans-PtCl2(SMe2)2 was synthesized by a modification of
10.1021/om0610039 CCC: $37.00 © 2007 American Chemical Society
Publication on Web 02/16/2007

1524 Organometallics, Vol. 26, No. 6, 2007
Moret and Chen
a literature procedure.13 All the other chemicals were obtained
commercially and used as received unless otherwise stated.
Complex Syntheses. [(N-N)Pt(CH3)(TFE)]+[BF4]-, ∼5 mM
Solution. One droplet of HBF4‚Et2O was added to a suspension of
(N-N)Pt(CH3)2 (6 mg, 10 µmol) in TFE, resulting in complete
dissolution and a color change from violet to orange, and the
mixture was stirred for 10 min. The solution was used without
further purification for electrospraying 1a.
(presumably elemental Pt) was removed by filtration, and 12 mL
of hexane was slowly added at 0 °C. The solution was left at
-20 °C overnight. Subsequent filtration and washing of the solid
with hexane (2 × 2 mL), followed by a second crystallization from
CH2Cl2/hexane, afforded the product as black microcrystals (92 mg,
37%). On some occasions, separation of the complex from the free
ligand by crystallization was not successful, and the product was
purified by column chromatography over dry, degassed neutral
alumina (30 g) with 1:1 hexane/Et2O as eluant, resulting in pure
product but lower yield (43 mg, 17%). The solid product can be
handled for short periods under air, but should be stored at 4 °C
under an inert atmosphere. Its solutions decompose over several
hours when exposed to air or at room temperature, yielding a black
[(N-N)Pt(C6H5)(TFE)]+[BF4]-, ∼4 mM Solution. Benzene
(0.5 mL, 0.44 g, 5.6 mmol) was added to 2 mL of a ∼5 × 10-3
M
solution of [(N-N)Pt(CH3)(TFE)]+[BF4]- and stirred for 2 h. The
solution was used without further purification for electrospraying
2a and 2b.
[(N-N)Pt(C2H5)(TFE)]+[BF4]-, ∼5 mM Solution. One droplet
of HBF4‚Et2O was added to a suspension of (N-N)Pt(C2H5)2 (7; 6
mg, 10 µmol) in TFE, resulting in complete dissolution and a color
change from violet to orange, and the mixture was stirred for 10
min. The solution was used without further purification for
electrospraying 3a and 4.
1
precipitate (presumably elemental Pt). Mp: g137 °C (dec). H
3
NMR (CD2Cl2): δ 7.55 (d, J(H-H) ) 8 Hz, 4H, ArHm), δ 7.27
(t, 3J(H-H) ) 8 Hz, 2H, ArHp), δ 1.89 (q, 3J(H-H) ) 8 Hz, 2J(195
-
Pt-H) ) 96 Hz, 4H, PtCH2CH3), δ 1.063 (s, 6H, NC(CH3)C), δ
3
4
0.61 (t, J(H-H) ) 8 Hz, J(195Pt-H) ) 84 Hz, 6H, PtCH2CH3).
13C NMR (CD2Cl2): δ 172.9 (CdN), δ 143.4 (Cipso), δ 128.7
(Cmeta), 127.8 (Cpara), 127.6 (Cortho), δ 21.5 (PtCH2CH3), δ 17.5
(NC(CH3)C(CH3)N), δ 0.8 (PtCH2CH3), no J(Pt-C) observed.
Anal. Calcd for C20H22N2Cl4Pt: C, 38.29, H, 3.53, N, 4,47.
Found: C, 38.34, H, 3.82, N, 4.62.
cis/trans-PtCl2(SMe2)2. (This reaction does not require an inert
atmosphere.) To a solution of H2PtCl6‚xH2O (10.8 g, 22 mmol) in
50 mL of water was added solid N2H4‚2HCl (1.16 g, 11 mmol) in
small amounts. The flask containing the dark red solution was then
placed in a 100 °C bath until no more gas evolution was observed
(0.5 h). After cooling to room temperature, a small amount of
platinum black was filtered off, 12 mL (0.16 mol) of dimethyl
sulfide was added, and the pink suspension was heated to 80 °C
until it had completely turned yellow. After cooling, it was extracted
with dichloromethane (3 × 50 mL) and the combined organic
phases were dried over anhydrous magnesium sulfate. Evaporation
of the solvent under reduced pressure yielded the product as a
yellow crystalline solid (8.40 g, 97%). 1H NMR (CDCl3, 300
[(N-N)Pt(C2H5)(CD3CN)]+[BF4]- Solution. A 50 µL sample
of a dilution of HBF4 (50% in water, 19 µL, 0.3 mmol) in TFE-d3
(1 g) was added to a suspension of 7 (6.3 mg, 10 µmol) in TFE-d3
(1 g), resulting in complete dissolution and a color change from
violet to orange. The mixture was stirred for 10 min, and 50 µL of
CD3CN was added. 1H NMR: δ 7.59-7.65 (m,4H, ArHm), δ 7.36-
7.45 (m, 2H, ArHp), δ 2.07 (s, 3H, NC(CH3)C), δ 1.96 (s, 3H,
3
2
NC(CH3)C, other side), δ 1.37 (q, J(H-H) ) 7.5 Hz, J(195Pt-
H) ) 85 Hz, 2H, PtCH2CH3), δ 0.51 (t, 3J(H-H) ) 7.5 Hz, 4J(195
Pt-H) ) 36 Hz, 3H, PtCH2CH3).
-
MHz): δ 2.58 (s, 3J(195Pt-H) ) 49 Hz, SMe2(cis)), δ 2.47 (s, 3J(195
Pt-H) ) 42 Hz, SMe2(trans)), cis/trans ratio 1:2.
-
Mass Spectrometry. Mass spectrometric measurements were
performed on a TSQ 700 instrument, modified as previously
described.12 The complexes 1a-3a and 2b were electrosprayed from
(1-2) × 10-5 M solutions in TFE and 1:4 mixtures of TFE and
H2O or pure H2O respectively.
Pt2(C2H5)4(µ-Me2S)2 (5). A 6.3 mL amount of a 0.46 M solution
of ethyllithium in 9:1 benzene/cyclohexane (2.90 mmol) was added
to a suspension of finely powdered [PtCl2(Me2S)2] (500 mg, 1.37
mmol) in diethyl ether (20 mL) at -15 °C (ice/salt mixture). The
mixture was stirred for 40 min, during which time the yellow solid
disappeared, the liquid phase became dark brown, and a white solid
(presumably LiCl) appeared. (All workup was performed at 0 °C
under argon.) The reaction was quenched with a N2-flushed, cooled
(0 °C) dilution of 1 mL of saturated NH4Cl in 20 mL of water. A
dark solid was filtered off, then the phases were separated, and the
aqueous phase was extracted with 3 × 10 mL of diethyl ether. The
combined organic phases were dried over MgSO4, activated
charcoal (ca. 0.2 g) was added, and the black mixture was filtered
to give a pale yellow solution, which was generally concentrated
to about 20 mL and used without further purification. (Evaporation
of the solvent in Vacuo at 0 °C yielded a tan solid that started turning
brown as it was becoming dry, as well as when it was allowed to
warm to room temperature.) The yield could be estimated from
the integrals of the 1H NMR spectrum of a sample taken from the
subsequent ligand exchange reaction to 0.20 mmol (29%). 1H NMR
Energy-resolved reactive cross-section measurements were ex-
ecuted as reported earlier.12 The ions were thermalized in the first
24-pole region to a temperature of 343 K using argon or a suitable
reaction gas (3-10 mTorr) and then mass selected by the first
quadrupole. They were then reacted with argon in the octopole
collision cell while monitoring the products as a function of collision
energy. Argon was chosen as reaction gas because it allows working
at higher energies in the lab frame in order to move the CID
threshold region away from the region where the ion beam is
truncated.
The “daughter” mode was used for mass selection, and a
retarding potential measurement of the kinetic energy distribution
of the ions was performed before each experiment, yielding roughly
Gaussian distributions with a fwhm between 1.5 and 2.1 eV in the
laboratory frame. The parent and daughter channels were monitored
alternately while the collision offset was scanned. Additionally, for
each measurement, a normal mass scan was performed with a fixed
collision offset to check the gross reaction chemistry and ensure
that no unexpected products appeared. The intensities of the two
channels were scaled so that their ratio matched the ratio of integrals
observed on the corresponding mass scan, and the cross sections
were calculated as described by Ervin et al.14 with a measured
effective path length of 23 ( 5 cm.
3
(C6D6): δ 2.10 (s, J(195Pt-H) ) 17 Hz, 12H, SCH3), δ 1.62 (q,
2
3J(H-H) ) 8 Hz, J(195Pt-H) ) 93 Hz, 8H, PtCH2CH3), δ 1.38
3
4
(t, J(H-H) ) 8 Hz, J(195Pt-H) ) 93 Hz, 12H, PtCH2CH3).
(N-N)Pt(C2H5)2 ((N-N) ) ArNC(Me)C(Me)N)Ar, Ar ) 2,6-
dichlorophenyl) (7). A 225 mg (0.60 mmol) amount of ArNC-
(Me)C(Me)N)Ar (Ar ) 2,6-dichlorophenyl) (6) was added to a
solution of 0.20 mmol of 5 in 20 mL of ether at 0 °C. The yellow
solution was allowed to warm at room temperature and stirred for
4.5 h. (The workup was effected under an argon atmosphere.) The
solvent was removed in Vacuo, the dark green residue was dissolved
in 4 mL of dichloromethane at 0 °C, a small amount of black solid
Extraction of the activation energy for ligand dissociation was
done with Armentrout’s CRUNCH program. For the computation
of the RRKM rate constant needed for the kinetic shift, frequencies
(13) Hill, G. S.; Irwin, M. J.; Levy, C. J.; Rendina, L. M.; Puddephatt,
R. J. Inorg. Synth. 1998, 32, 149.
(14) Ervin, K. M.; Armentrout, P. B. J. Chem. Phys. 1987, 83, 166.

Ligand Binding Energies in Cationic Pt(II) Complexes
Organometallics, Vol. 26, No. 6, 2007 1525
Scheme 2. Gas-Phase Synthesis and CID Reactions of 1b/c
and rotational constants for both the parent molecule and the
transition state are required. These were obtained from DFT
calculations.
converged geometries were first-order saddle points and that the
only negative frequency corresponded to the desired reaction
coordinate.
All energies were ZPE corrected using a scaling factor of 0.9774
as proposed by Scott and Radom22 for B3PW91/6-31G(d). The
drawings were made using the PLUTON module of the program
suite PLATON.23
DFT Calculations. All DFT calculations were performed using
Gaussian 03.15 Geometry optimizations were performed using the
Becke three-parameter hybrid functional16 using the Perdew-
Wang17 nonlocal correlation. The Stuttgart/Dresden18,19 basis set
and effective core potential were used for Pt, along with a 6-31G-
(d,p) basis set on all other atoms. All frequency calculations was
performed on converged geometries and verified as minima. When
convergence problems were encountered, the GDIIS20 algorithm
was used.
Results
Structure of the Adduct of 1a and Benzene. As previously
reported,12 the reaction of ion [(N-N)PtCH3]+ (1a) with
benzene in the gas phase is presumed to generate the rapidly
equilibrating set of ions 1b/c, which upon CID with a rare gas
loses either benzene or methane (yielding the phenyl complex
2a), as depicted in Scheme 2. Assignment of the observed
threshold for dissociation to the depicted chemical process
assumes a defined structure for the ion from which the
dissociation occurs. According to the Curtin-Hammet principle,
the observed reaction energy in a CID threshold measurement
is the energy difference between the lowest kinetically accessible
minimum of the potential energy surface and the transition state
leading to dissociation. So, if the studied ion would equilibrate
with another structure of lower energy on the time scale of the
gas-phase experiment (∼60 µs), this would result in an
overestimation of the actual binding energy.
Preliminary calculations had indicated, in our case, that [(N-
N)PtH(C6H5-CH3)]+ (3c) could have a lower energy than
1b/c, while its dissociation productssnamely, the cationic
platinum hydride complex 3a and toluenescould lie slightly
above the observed products from dissociation of 1b/c. This
would make it possible for 3c, if it were formed, to potentially
dissociate to the same products as 1b/c. Furthermore, as
only the m/z ratio of the parent ion is measured, the hydrido
toluene complex cannot be distinguished from 1b/c in a
simple mass spectrometry experiment. Thus, it was necessary
to check whether they were equilibrating via a potentially
Transition states were computed by first performing a relaxed
potential energy surface scan along a suitable coordinate and
then using the highest energy point as transition state guess for the
QST3 (Synchronous Transit-Guided Quasi-Newton) method21 along
with starting material and products optimized geometries. A
subsequent frequency calculation was used to check that the
(15) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; 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.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; and Pople, J. A.
Gaussian 03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(16) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(17) Perdew, J. P.; Burke, K.; Wang, Y. Phys. ReV. B 1996, 54, 16533.
(18) Bergner, A.; Dolg, M.; Kuechle, W.; Stoll, H.; Preuss, H. Mol. Phys.
1993, 80, 1431.
(19) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer, R. M. J. Phys. Chem 1993,
97, 5852.
(20) Farkas, O.; Schlegel, H. B. J. Chem. Phys. 1999, 111, 10806.
(21) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput.
Chem. 1996, 17, 49.
(22) Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502.
(23) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7.

1526 Organometallics, Vol. 26, No. 6, 2007
Moret and Chen
Scheme 3. Synthesis of (N-N)Pt(C2H5)2 (7) and Generation
of [(N-N)PtH]+ (3a) in the Gas Phase
Figure 1. Daughter spectrum of 3c (top) and 1b/c (bottom) collided
with 100 µTorr of argon at 35 V offset, showing that the parent
ions do not interconvert on the measurement time scale.
Scheme 4. Reactions Studied by CID Threshold
Measurements
plausible reversible C-C activation reaction on the time scale
of our measurements. In order to do so, 3c was prepared from
the gas-phase reaction of the hydride complex cation with
toluene.
First attempts to prepare [(N-N)PtH]+ (3a) from the reaction
of 1a with molecular hydrogen were unsuccessful. No reaction
was observed in the gas phase under various conditions. In TFE
solution, reaction with up to 5 bar of H2 led only to the
appearance of a peak corresponding to the known9 decomposi-
2+
tion ion [(N-N)Pt(CH3)µ-OH]2 in the mass spectrum, ulti-
mately producing Pt black.
The work of Mole et al.24 on platinum(II) complexes with
chelating diphosphines suggested to us that the desired
hydride complex could be prepared by an alternative route (see
Scheme 3) based on a â-hydride elimination reaction from
the ethyl complex 4. This approach proved to be successful.
First, the complex PtCl2(SMe2)2 was reacted with ethyllithium
in diethyl ether at -15 °C to yield the dinuclear complex Pt2-
(C2H5)4(µ-Me2S)2 (5). Compound 5 is a white solid that is
unstable at room temperature under inert gas as well as in
vacuum. It reacts with diimine 6 in diethyl ether to give the
corresponding diethyl complex 7 in moderate yield. The
protolysis of 7 in TFE yielded solutions from which the ethyls
or the isomeric hydrido ethylene complex (4)scould be elec-
trosprayed as the most intense signal.
A 1H NMR spectrum of a solution obtained by protolysis of
7 with HBF4/H2O in TFE-d3 showed that at least 50% of the
ions are present in the [(N-N)Pt(C2H5)(L)]+ form. Addition
of MeCN-d3 to this solution gave one main product, which is
assigned as [(N-N)Pt(C2H5)(CD3CN)]+, showing that no fast
irreversible loss of ethylene occurs in solution at room temper-
ature.
Collision-induced dissociation of ion 4 gave the desired
hydride complex 3a as the main ionic fragment; alternatively,
tuning the tube lens potential at the entrance to the first rf-
multipole to higher values generated 3a without the need of a
subsequent CID step in the collision cell.
3a was then reacted with toluene in the gas phase. Introduc-
tion of toluene in the 24-pole ion guide of our instrument, which
is located before the first mass selection quadrupole, resulted
in the disappearance of the peak of 3a at m/z ) 569 with
concurrent appearance of a new peak corresponding to the
desired toluene adduct 3c at m/z ) 661 (spectra are provided
in the Supporting Information.). This product ion was then mass
selected and collided with argon at a variety of pressures and
collision energies, resulting only in the loss of toluene, as seen
in Figure 1.
If the previously mentioned, reversible C-C activation
reaction were to have taken place, the same set of fast
interconverting ions would be produced from reaction
of 1a with benzene or of 3a with toluene, and the same CID
products should be observed. As is evident from Figure 1,
the CID products are completely different, ruling out the
possibility that the gas-phase reaction of 1a with benzene
produces the hydrido toluene complex 3c. If one were to
believe the computational prediction that 3c represents a lower
energy structure than 1b/c, then one must postulate a high
enough activation barrier for the reversible C-C activation
reaction so as to render interconversion kinetically uncom-
petitive.
CID Threshold Measurements and DFT Calculations.
Energy-resolved reactive cross-section data were acquired as
described earlier12 for the CID reactions depicted in Scheme 2
as well as for the related ions [(N-N)Pt(C6H5)(L)]+ (L ) H2O,
TFE, benzene) (2b-d) and [(N-N)PtH(C6H6)]+ (3b), listed in
Scheme 4.
The phenyl ions 2a,b could be cleanly electrosprayed from
a solution of 2b in TFE that had been treated with benzene for
∼2 h and diluted with TFE and TFE/H2O, respectively. The
benzene adducts, 2d and 3b, and TFE adduct 2c were obtained
from reaction of the appropriate ions with benzene or TFE in
the gas phase.
(24) Mole, L.; Spencer, J. L.; Litster, S. A.; Redhouse, A. D.; Carr, N.;
Orpen, A. G. J. Chem. Soc., Dalton Trans. 1996, 2315.

Ligand Binding Energies in Cationic Pt(II) Complexes
Organometallics, Vol. 26, No. 6, 2007 1527
Figure 2. Calculated structures and energies [kcal/mol] for “endo” (A) and “exo” (B) 2d, its CID products 2a and benzene (D), and the
transition state for the dissociation reaction (C). Selected distances [Å] angles [deg] and torsion angles [deg] (symmetry equivalent
values are averaged): A: Pt1-N1 2.066, Pt1-N2 2.254, Pt1-C1 2.016, Pt1-C2 2.322, Pt1-H1 2.662; N1-Pt1-N2 74.9, N1-Pt1-C1
92.7, N2-Pt1-C2 108.4, C1-Pt1-C2 83.4; N1-C3-C4-N2 0.8, Pt1-C2′-C2-C5 110.5. B: Pt1-N1 2.058, Pt1-N2 2.204, Pt1-C1
2.013, Pt1-C2 2.315, Pt1-H1 2.648, Pt1-Cl1 4.164; N1-Pt1-N2 75.9, N1-Pt1-C1 94.1, N2-Pt1-C2 94.4, C1-Pt1-C2 95.1; N1-
C3-C4-N2 0.5, Pt1-C2′-C2-C5 110.9. C: Pt1-N1 1.997, Pt1-N2 2.146, Pt1-C1 1.998, Pt1-C2 3.266, Pt1-H1 3.201, Pt1-Cl1
2.631; N1-Pt1-N2 78.0, N1-Pt1-C1 99.9, N2-Pt1-C2 90.1, N2-Pt1-Cl1 76.2; N1-C3-C4-N2 11.0. D: Pt1-N1 2.007, Pt1-N2
2.110, Pt1-C1 2.008, Pt1-Cl1 2.391; N1-Pt1-N2 79.1, N1-Pt1-C1 101.1, N2-Pt1-Cl1 83.6, C1-Pt1-Cl1 95.5; N1-C3-C4-N2
16.5.
Thermochemical information was extracted using the
CRUNCH program,25 which fits the dissociation cross section
to eq 1:
including our earlier measurements of the bond dissociation
energies (BDE) of solvents from cationic platinum diimine
methyl complexes.12 On the other hand, the PSL model is not
appropriate for treating dissociations that go through a tight
transition state, especially in the case where a reverse activation
barrier exists. In this case, frequencies computed from a DFT-
optimized transition structure were used when available.28,29
Alternatively, a tight transition state can be modeled by equating
the frequencies of the transition state with the ones from the
parent ion, after removing the frequency that resembles most
the reaction coordinate.26
As CRUNCH treats the kinetic shift correction without fitting
any parameter to the data, additional knowledge about the nature
of the transition state is needed. DFT calculations provided the
necessary information. The geometries of the parent ions, 1b,c,
2b-d, and 3b, and daughter ions, 1a, 2a, and 3a, as well as
the neutral leaving ligands, were optimized using density
functional theory (DFT), using the B3PW91 exchange correla-
tion functional16,17 and the Stuttgart/Dresden18,19 basis set and
effective core potential for Pt with the 6-31G** basis set for all
other atoms.
nσ0
E
i
σ(E) )
g
(1 - e-k(ꢀ+E )τ)(E - ꢀ)n-1dꢀ (1)
∫
∑
i
(
)
E0-Ei
E
i<bu>
where E is the collision energy in the center-of-mass frame, E0
is the reaction threshold at 0 K, σ0 is a fitted scaling factor, n
is an adjustable parameter, Ei are the energies of rovibrational
states with degeneracy gi, and k is the RRKM reaction rate with
a residence time τ (estimated as 6 × 10-5 s) in the collision
cell. For multiple channel fits, CRUNCH uses the following
modification of eq 2:
nσ0,j
ktot(ꢀ + Ei)
E
σj(E) )
g
i
∫
∑
(
)
E0-Ei
E
kj(ꢀ + Ei)
i
(1 - e-k (ꢀ+E )τ)(E - ꢀ)n-1dꢀ (2)
tot
i
where ktot ) ∑kj. In the case of relatively large ions, in which
vibrational energy is spread over many modes, the kinetic shift
correction accounts for a large part of the observed threshold
energy. Thus, an accurate computation of k, based on an
appropriate transition state description, is necessary.
For simple bond dissociation reactions that proceed without
a reverse barrier, the loose orbiting transition state description26
is optimal for the treatment of the kinetic shift. In this models
the so-called phase space limit (PSL) modelsthe centrifugal
barrier is considered as the transition state, and transitional
modes are equated with rotational modes of the products. It
has been used for treating a variety of metal-ligand systems,27
For the benzene complexes, 1b, 2d, and 3b, two minima were
found, both with an η2 coordination mode. One of them (“exo”)
has the benzene ring pointing outside, in the same direction as
the Pt-L bond, whereas in the other structure (“endo”) the
benzene ring lies parallel to the 2,6-dichlorophenyl group of
the ligand. The two structures are shown in Figure 2, labeled A
and B. The “exo” isomer is always found to be the more stable
of the two, by 3.7, 2.5, and 4.5 kcal/mol for 1b, 2d, and 3b
(27) Rodgers, M. T.; Armentrout, P. B. Mass Spectrom. ReV. 2000, 19,
215.
(28) Treating the reactions for which optimized transition state frequen-
cies are available with the tight TS model based on parent ion frequencies
yielded E0 values within the experimental error.
(29) CRUNCH has been shown to be suitable for such a treatment of
tight transitions states: Muntean, F.; Armentrout P. B. J. Phys. Chem. B
2002, 106, 8117.
(25) CRUNCH, version D1, was kindly provided as an executable by
Prof. P. Armentrout.
(26) Rodgers, M. T.; Ervin, K. M.; Armentrout, P. B. J. Chem. Phys.
1997, 106, 4499.

1528 Organometallics, Vol. 26, No. 6, 2007
Moret and Chen
respectively. The energetic order may be rationalized in terms
of steric repulsion between the benzene ligand and the substi-
tuted phenyl group of the diimine moiety in the “endo” form,
resulting in a somewhat longer Pt-N(cis) distance (2.25 versus
2.20 Å). However, DFT has been shown to reproduce dispersion
interactions, and especially π-π stacking, badly: the calculated
face-to-face benzene-benzene potential energy curve is usually
purely repulsive in DFT treatments, while experiment and higher
level calculations show that there is an energy minimum.30 In
our case, this could lead to an overestimate of the energy of
the “endo” isomers, where the benzene ligand and the diimine
phenyl group are close to a face-to-face parallel configuration,
and thus the energy ordering should be taken with some care.
Nevertheless, Johansson et al.10 observed that low-temperature
Figure 3. Relaxed PES scan along the Pt-Cl distance coordinate
for [(N-N)Pt(Ph)(L)]+ complexes, L ) methane (squares), benzene
(diamonds), TFE (triangles), and water (circles). The zero of energy
is set at the energy of [(N-N)Pt(Ph)]+ + L, designated by the star
on the diagram. The point designated with the solid diamond
represents the optimized transition state for the elimination of
benzene.
protolysis of the related [(NMeNMe)Pt(Me)(Ph)] (NMeNMe
)
ArNdC(Me)C(Me)dNAr, Ar ) 2,6-dimethylphenyl) in dichlo-
romethane yields a [(NMeNMe)Pt(Me)(Benzene)]+ cation in
1
which the PtMe H NMR resonance is shielded by the ring
current of the benzene ligand, giving experimental support to
our finding that the “exo” conformer is more stable. These
authors also observe that the two conformers interconvert
quickly on the NMR time scale even at -69 °C. Thus, we
assume fast equilibrium for the gas-phase reactions as well,
presumably via a C-H σ complex that also lies on the C-H
activation pathway, and we assume that the “exo” isomer is
the ground state for the η2-benzene complexes in further
calculations and discussion.
Table 1. Experimental and Calculated Bond Dissociation
([Pt]+ ) [(N-N)Pt]+, N-N ) 7) (energies in kcal/mol,
entropies in cal‚mol-1K-1
)
fragments
TS model
Eq
∆Sq
E0,DFT ∆Eq
0,DFT
0,exp.
1000
CH3[Pt]+ + C6H6
C6H5[Pt]+ + CH4
H[Pt]+ + C6H6
Ph[Pt]+ + C6H6
Ph[Pt]+ + TFE
Ph[Pt]+ + H2O
tight
tight
tight
tight
PSL
PSL
20.1 ( 1.7
17.3 ( 2.1
21.3 ( 1.7
19.2 ( 1.9
28.4 ( 2.5
31.7 ( 2.5
0.5
18.3
14.2
24.1
16.8
23. 6
28.6
21.5
a
25.0
21.5
b
-4.3
-4.0
0.6
7.8
6.1
b
Information about the nature of the transition states was
obtained as follows. First, it was observed in all cases that, in
the energy-minimized structure of the product ions of the type
[(N-N)Pt(R)]+, comprising 1a, 2a, and 3a, the formally
empty coordination site is in fact occupied by one of the
ortho chlorine atoms of the ligand (see Figure 2, D). It
follows that the dissociation reactions of interest should
not be considered as simple bond-breaking reactions, but rather
as an intramolecular associative ligand exchange reaction. Given
that an associative ligand exchange occurs, in principle, through
two steps, which can be more or less synchronous, and
depending on which of these two steps would be actually rate-
limiting, the relevant transition state for loss of a ligand could
be tighter than expected for a simple dissociation. In the extreme
case one could even justify a nonzero barrier for the reverse
reaction.
a Transition state could not be located. b Calculated to be loose transition
states (see Figure 3).
which suggests some synchronicity for the two formal steps in
the associative ligand exchange. Unfortunately, a well-defined
transition structure could not be optimized for the dissociation
of methane from 1c, presumably because the multiple
binding modes of methane produce a very flat potential energy
surface.
Figure 2 shows the computed structures and energies relevant
to the dissociation of benzene from 2d. The calculated transition
structure is quite “product-like”, as seen by the short Pt-Cl
(2.63 Å) and the long Pt-C(benzene) (3.27 Å) distances and
the position of the benzene molecule, which is closer to a
σ-C-H coordination than to the original η2-C-C. This is
expected from the Hammond postulate for a relatively low-lying
transition state (∼4.7 kcal above the products). Taking these
results into account, quantitative thermochemical information
was extracted from the experimental data using the tight
transition state model for the benzene and methane dissociation
reactions and the PSL model for the others. The results are
collected in Table 1. As expected, the calculated activation
entropies are largely positive for systems where the PSL model
is used and small or negative for tight transition states. The
uncertainty on E0 includes contributions from the standard
deviation obtained from at least three replicate data sets,
a 0.15 eV (lab) uncertainty on the energy scale, and the
uncertainty on the transition state model. The latter, defined as
In order to probe for the occurrence of such transition states
in our diimine platinum phenyl system, relaxed structure
energies were computed with decreasing, constrained Pt-Cl
distances starting from the parent ion, as shown in Figure 3. In
principle, one would expect a tight transition state to be found
if the curve obtained this way crosses the energy of the
dissociation products before the daughter ion Pt-Cl distance
is reached. Interestingly this happens only in the case of the
benzene and methane ligands. One can understand this by the
fact that η2-coordinated ligands fill more space in the first
coordination sphere of the metal and, therefore, have to depart
earlier as the incoming ligand gets closer. The existence of a
reverse energy barrier for the dissociation of benzene from 1b,
2d, and 3b was confirmed by optimizing a transition state
31
the effect of a change of (R in the activation entropy on E0
geometry using the quasi-synchronous transit method.21
A
(linearly interpolated from the data fitting with the tight and
loose transition state models), is the main contribution in all
cases.
frequency calculation on these structures confirmed that only
one negative frequency was present for each and that the
corresponding mode consisted mainly of simultaneous shorten-
ing of the Pt-Cl and lengthening of the Pt-benzene distances,
Agreement between calculated and experimental energies is
acceptable (within ∼5 kcal) for all measurements. An extreme
(30) Sato, T.; Tsuneda, T.; Hirao, K. J Chem. Phys. 2005, 123, 104307.
(31) Wenthold, P. G. J. Phys. Chem. A 2000, 104, 5612.

Ligand Binding Energies in Cationic Pt(II) Complexes
Organometallics, Vol. 26, No. 6, 2007 1529
Table 2. Comparison of the PSL and Tight Transition
State Models (energies in kcal/mol, entropies in
case is found for the dissociation of TFE from 2c, where the
calculated energy is 4.8 kcal lower than the experimental value.
This discrepancy could originate from an overestimation of the
experimental value due to a partial tight character of the
transition state: the relaxed PES results shown in Figure 3
indicate that the dissociation of TFE could be an intermediate
case. Thus, this experimental energy should be taken with
care.
cal‚mol-1K-1
)
PSL
tight
∆Sq
1000
fragments
∆Eq
∆Sq
∆Eq
0
0
1000
CH3[Pt]+ + C6H6
C6H5[Pt]+ + CH4
H[Pt]+ + C6H6
Ph[Pt]+ + C6H6
Ph[Pt]+ + TFE
Ph[Pt]+ + H2O
29.5
12.0
20.1
17.3
21.3
19.2
17.6
20.6
0.5
26.2
33.6
29.4
28.4
31.7
4.4
10.8
11.7
7.8
-4.3
-4.0
0.6
-2.8
-3.1
6.1
Discussion
treatment based on ab initio calculations does reproduce this
sequence.32
CID threshold measurement is an attractive method for
obtaining thermochemical information about a variety of orga-
nometallic systems, as the involved species are often ionic and
sometimes only available in very small quantities or produced
in situ. In addition, it makes it possible to perform measurements
on short-lived intermediates that are otherwise hardly accessible.
The data presented here illustrate this possibility: activation
energies could be measured for the dissociations of 1b/c to 1a
and benzene and to 2a and methane, whereas this information
is usually hidden in solution-phase kinetic studies where 1b/c
are intermediates whose overall reactions are a convolution of
multiple steps.
One may argue that, if the dissociation of the benzene adducts
1b/c, 2d, and 3b were to proceed through a tight transition
state, implying a reverse activation barrier, it might be
reasoned that the complexes should hardly form at all in the
gas phase because of that same barrier. It would present a
problem in our experiment where we use the reverse reaction
to synthetize these ions in the gas phase. Consideration of
two aspects solves this apparent contradiction. First, the
postulated tight transition states have a relatively small
enthalpic barrier (less than 5 kcal/mol according to DFT as
shown in Table 1), but exert their effect on the kinetic shift
correction through the higher frequencies. Second, under the
conditions used in the gas-phase ion-molecule preparation of
the ions (24-pole ion guide without an external longitudinal
potential, 3-10 mTorr benzene), the ions undergo a high number
of collisions (.104 according to Monte Carlo simulations as
described in ref 33). Both considerations combine to allow the
gas-phase synthesis to proceed even if the reaction under single-
collision conditions would have a low probability due to an
activation barrier.
The energies shown in Table 1 are activation energies for
the observed dissociations, which can be interpreted as
actual binding energies only in the case of water and TFE
departure where no reverse activation barrier is present. In
both cases, they can be seen as an estimate of the activation
energy for formally dissociative ligand exchange in solution,
assuming small enough medium effects. In contrast, ligand
exchange is usually an associative process in square planar
Pt complexes and has been shown to be associative in the
case of acetonitrile for complexes closely related to 1 and
2. Thus, these values give an approximate upper bound to the
range of possible activation energies for associative ligand
exchange.
However, getting accurate information about “real life”
complexes incorporating large, flexible ligands and a range of
possible reactive pathways requires care with respect to
several aspects. First, the nature of the observed ions should be
carefully established. As mass spectroscopic methods measure
the m/z ratio of the ions of interest, they cannot in general
distinguish between isomers. A common way to overcome
this limitation is to compare the CID patterns of ions from
different preparations or reactions to assess whether they
correspond to the same structure or set of equilibrating
structures. We have shown here how independent synthesis and
CID of a postulated ion, 3c, could be used to rule out a possible
equilibrium that would have interfered with accurate energy
measurements.
Second, an increase in the size of the studied complexes
implies an increased number of vibrational modes over which
the deposited energy is distributed and thus slower microca-
nonical dissociation rates. This causes larger kinetic shift effects,
for which an accurate treatment is critical. The widely used PSL
model relies on the assumption that the transition state is located
at the centrifugal barrier, which can be true only if no reverse
activation barrier is present. While it may often be the case for
simple ligand dissociation reactions, one should keep in mind
that more flexible systems may offer an additional stabilization
to the generated free coordination site, resulting in more complex
intramolecular associative substitution pathways. Such reactions
can exhibit a reverse activation barrier, making the PSL model
inappropriate, as found in the case of benzene and methane
dissociation reactions. The naive assumption that for similar
ions an inaccurate treatment of the kinetic shift would result
only in a systematic errorsstill yielding relevant relative
energiesswould have been wrong in the case of the [(N-N)-
Pt(Ph)(L)]+ series (2b-d). The expected binding energy
sequence (benzene < TFE < water), consistently supported by
our DFT calculations (see Table 1) and solution-phase observa-
tions,10 would not be obtained using a single transition state
model (see Table 2). On the other hand, our differentiated
The ortho chlorine substituent in the diimine ligand has been
found here to assist the departure of the leaving group and
stabilize the free cation. In the case of related Pt(II) complexes
in which the ortho position on the aryl substituent is occupied
by a methyl group instead of a chlorine, we had previously
observed isotope exchange on the methyl group when the
complex was reacted with C6D6. We interpreted the observation
to indicate that a reversible, intramolecular C-H activation had
occurred, which necessarily implies that a precursor C-H
complex, which suggests in turn that the chlorine substituent
in 1 and 2 is not unique in its ability to fill the nominally empty
coordination site on Pt.
(32) An interesting example of a system that can be treated appropriately
only by considering both tight and loose transition states for different
channels can be found in: Muntean, F.; Armentrout P. B. Z. Phys. Chem.
2000, 214, 1035.
(33) Adlhart C.; Hinderling C.; Baumann H.; Chen P. J. Am. Chem. Soc.
2000, 122, 8204.

1530 Organometallics, Vol. 26, No. 6, 2007
Moret and Chen
Acknowledgment. The authors acknowledge support from
the Swiss National Science Foundation and the Research
Commission of the ETH Zu¨rich.
Conclusions
Quantitative thermochemical information was obtained for
species appearing on the potential energy surface of a C-H
activation reaction via CID threshold measurements. Accurate
modeling of the transition states for ligand dissociation is critical
for obtaining reliable data, and DFT calculations were used for
this purpose. In particular, the dissociation reactions were found
to be best described as intramolecular associative ligand
exchange processes, exhibiting tight transition states in the cases
of η2-coordinating hydrocarbon ligands. The results are in good
agreement with computed DFT energies, as well as with
qualitative solution-phase observations.
Supporting Information Available: Mass spectra showing
the gas-phase synthesis of 3c; energy-resolved CID cross-section
data and fitting; optimized structures, harmonic frequencies,
rotational constants, and energy of all calculated minima and
transition states.
OM0610039