Determination of ruthenium-phosphorus bond dissociation energies by ES-FTICR mass spectrometry

Article abstract of DOI:10.1021/om020308u

Gas-phase activation energies were determined for the dissociation of one and two phosphines from the cationic ruthenium complex by the combined techniques of ES introduction and Fourier ion cyclotron resonance (FTICR) mass spectrometry (MS). The calculat

Full text of DOI:10.1021/om020308u

5688
Organometallics 2002, 21, 5688-5691
Deter m in a tion of Ru th en iu m -P h osp h or u s Bon d
Dissocia tion En er gies by ES-F TICR Ma ss Sp ectr om etr y
J ohn B. Westmore,* Lisa Rosenberg,*,† and Timothy S. Hooper
Department of Chemistry, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2
Gary D. Willett and Keith J . Fisher
School of Chemistry, University of New South Wales, Sydney,
New South Wales 2052, Australia
Received April 17, 2002
Sch em e 1
Summary: Gas-phase activation energies have been es-
timated for the dissociation of one and two phosphines
from the cationic ruthenium complex [(η5-indenyl)Ru-
(HPPh2)3]+ (1) by the combined techniques of ES intro-
duction and Fourier transform ion cyclotron resonance
(FTICR) mass spectrometry (MS). These values are 16.6
( 2.3 and 34.6 ( 2.3 kcal mol-1, respectively.
Phosphine dissociation is critical to many organo-
metallic reaction sequences. Metal-phosphine bond
strengths are useful in the evaluation of possible mech-
anisms, particularly in catalysis, where important
elementary steps can occur too quickly for in situ
measurement of their thermal parameters. Despite the
development of a range of techniques for probing metal-
ligand bond energy information, both in solution and
in the gas phase, the amassed library of organometallic
bond energies remains extremely limited.1 Among ther-
modynamic data collected, bond energies for metal
complexes containing primary (1°) and secondary (2°)
phosphines are particularly rare.
Fourier transform ion cyclotron resonance mass spec-
trometry (FTICR-MS) provides extremely high mass
resolution2 and is a valuable tool for both qualitative
and quantitative analysis of bond activation/reactions
of ions in the gas phase.3 This technique has been
applied exhaustively to the study of ions generated from
volatile precursors in the ICR cell (i.e. at very low
pressures). Currently, the coupling of high-pressure
introduction techniques such as electrospray (ES) with
FTICR is of intense interest in analytical mass spec-
trometry.4 The application of such coupled MS tech-
niques to analysis of nonvolatile, solution-stable species
is gradually expanding from the study of biological and
organic molecules into the realm of inorganic chemis-
try.5 Here we report the combined use of ES and FTICR-
MS techniques to examine the fragmentation patterns
of the cationic ruthenium(II) complex [(η5-indenyl)Ru-
(HPPh2)3]+ (1) and to estimate energies for the dissocia-
tion of ruthenium-phosphorus bonds.
We prepared complex 1 during our ongoing investiga-
tion of the reactions of (η5-indenyl)Ru(PPh3)2Cl (2) with
1° and 2° phosphines, in the context of catalytic P-H
activation. Preliminary studies indicate some activity
of 2 toward the hydrophosphination of acrylonitrile by
2° phosphines. The complex 1‚Cl- results from the
addition of 3 equiv or more of diphenylphosphine to
complex 2 (Scheme 1).6 Monitoring of this reaction by
31P{1H} NMR spectroscopy indicates that a neutral com-
plex, (η5-indenyl)Ru(PPh3)(HPPh2)Cl (3),7 resulting from
substitution of one of the two triphenylphosphine ligands
in 2, begins to form immediately upon addition of
diphenylphosphine to 2. Intermediate 3 remains in high
concentration during the (much slower) formation of 1.8
Preliminary reactivity studies indicate that 1 will slowly
convert to 3 in the presence of triphenylphosphine.9
The positive charge on complex 1, along with its
stability, made it straightforward to study by electro-
spray mass spectrometry (ES-MS).10 Following well-
established procedures for using ES-MS to determine
relative labilities of ligands in metal complexes,11 we
varied the ES cone voltage12 to either obtain the parent
* To whom correspondence should be addressed. E-mail: J .B.W.,
westmor@cc.umanitoba.ca; L.R., lisarose@uvic.ca. Fax: J .B.W., 204-
474-7608; L.R., 250-721-7147.
† Present address: Department of Chemistry, University of Victoria,
P.O. Box 3065, Victoria, British Columbia, Canada V8W 3V6.
(1) (a) Marks, T. J ., Ed. Bonding Energetics in Organometallic
Compounds; American Chemical Society: Washington, DC, 1990; Vol.
428. (b) Nolan, S. P. In Encyclopedia of Inorganic Chemistry; King, R.
B., Ed.; Wiley: New York, 1994; p 307.
(5) See for example: (a) Henderson, W.; Nicholson, B. K.; McCaffrey,
L. J . Polyhedron 1998, 17, 4291. (b) Hasselgren, C.; Fisher, K.; J agner,
S.; Dance, I. Chem. Eur. J . 2000, 6, 3671. (c) Satterfield, M.; Brodbelt,
J . S. Inorg. Chem. 2001, 40, 5393. (d) Sautter, A.; Schmid, D. G.; J ung,
G.; Wu¨rthner, F. J . Am. Chem. Soc. 2001, 123, 5424.
(2) (a) Marshall, A. G.; Hendrikson, C. L.; J ackson, G. S. Mass
Spectrom. Rev. 1998, 17, 1. (b) Amster, I. J . J . Mass Spectrom. 1996,
31, 1325.
(3) Freiser, B. S. In Bonding Energetics in Organometallic Com-
pounds; Marks, T. J ., Ed.; American Chemical Society: Washington,
DC, 1990; Vol. 428, p 55.
(6) See the Supporting Information for details of characterization
and properties of 1.
2
(4) McLuckey, S. A.; Van Berkel, G. J .; Glish, G. L. Anal. Chem.
1994, 66, 689A.
(7) 31P{1H} NMR for 3 (121.5 MHz, CDCl3, δ): 50.9 (d, J PP ) 45
Hz, HPPh2). 47.5 (d, PPh3).
10.1021/om020308u CCC: $22.00 © 2002 American Chemical Society
Publication on Web 11/27/2002

Communications
Organometallics, Vol. 21, No. 26, 2002 5689
metal complex ion or promote ligand dissociation. At the
minimum cone voltage, 80 V, the only ion type present
is the intact cation 1 (major isotopic peak m/z 775). As
the cone voltage is increased incrementally to 300 V,
peaks appear due to first [(C9H7)Ru(HPPh2)2]+ (4, m/ z
589) and then [(C9H7)Ru(HPPh2)]+ (5, m/ z 403), which
correspond to dissociation of one and two HPPh2 ligands,
respectively, from 1. At cone voltages above 140 V, there
is also a small peak corresponding to the chloride-bound
dimer13 [{(C9H7)Ru(HPPh2)3}2(Cl)]+ (6). The remaining
phosphine in 5 is resistant to Ru-P bond cleavage, but
at higher cone voltages an ion formed by loss of C6H6
from 5 appears (major peak m/z 325). The (η5-indenyl)-
Ru core is apparently extremely robust: in these experi-
ments, no ions were detected that could be attributed
to indenyl-free ruthenium.
translational energy of the reactant ion and of the mass
of the target relative to the total mass of reactant ion
and target.16 For the cell and conditions used in our
experiments the collision energy is given by eq 1, in
Ecom ) (1.20607 × 107)[mtarget
/
(mion + mtarget)][(zVpptrfS/d)2/mion] eV/collision (1)
which the only two variable parameters are Vpp (the
peak-to-peak voltage of the radio frequency pulse used
to accelerate the parent ion) and trf (the duration of this
radio frequency pulse).17 In our experiments, ions enter-
ing the ICR cell18 were translationally energized by a
constant Vpp of 80 V, of controlled trf, and were subse-
quently allowed to collide with N2 at a pressure of 1 ×
10-8 mbar (as recorded by the manifold ion gauge). Since
Having established possible fragmentation products
for 1, we further examined its behavior in the gas phase
by placing the intact ion (produced from ES at a cone
voltage of 80 V) in an FTICR cell. We performed
collision-induced dissociation (CID) experiments,14 in
which a radio frequency pulse is applied to the ICR cell
plates to increase the translational energies of mass-
selected ions (e.g. intact 1). These ions are then allowed
to collide with a target gas for a defined period of time
(CID delay), during which fragmentation may occur if
the collisions are sufficiently energetic; resulting frag-
ment ions can then be identified. The yields of fragment
ions can be increased by increasing the translational
energies of the intact ions or by increasing the number
of collisions. Alternatively, if we can arrange for no more
than a single collision between a reactant ion and the
target gas during the CID delay, it should be possible
to measure the threshold energy for fragmentation by
studying collisions at steadily increasing collision ener-
gies.15 We used this latter technique to estimate Ru-P
bond dissociation energies involved in the fragmentation
of 1 to give the ions [(η5-indenyl)Ru(HPPh2)2]+ (4) and
[(η5-indenyl)Ru(HPPh2)]+ (5).
V
pp was held constant, the collision impact was directly
controlled through variation of trf.
The frequency of collisions was estimated from eq 2,
which is based on a previously derived equation15 for
the average time between collisions, tcollision, and adapted
for our experimental conditions.19 For trf ) 50 µs (the
1/tcollision ) (rion + rtarget)2PzVpptrfS/[(8.983 ×
10-37)Tmiond] (2)
approximate midpoint of the range of excitation times
at which dissociation was detected by the mass spec-
trometer), under the conditions used in our experiments,
the collision frequency is 2.60 collisions/s.20 Our experi-
ments21 indicated that a CID delay of 0.05 s should give
sufficient numbers of single collisions (∼0.13 per ion)
for measurable product ion yields,22 while minimizing
the number of double or multiple collisions.23 We
therefore used CID data from experiments with this
delay time to estimate the threshold energies for phos-
phine dissociation from 1.
Threshold energies for dissociation of one and two
phosphines from 1 were determined by varying its
collision impact with the target gas N2. The relevant
energy for collision, in an ICR cell, of a reactant ion such
as 1 with a stationary target such as N2 is the center-
of-mass energy, Ecom. This is a function both of the
Mass spectra of the ionic collision products showed
that at sufficiently high collision energies loss of one or
two phosphine ligands occurs, as in the ES experiments
carried out at cone voltages greater than or equal to 100
(16) Shukla, A. K.; Futrell, J . H. Mass Spectrom. Rev. 1993, 12, 211.
(17) In eq 1, mtarget and mion are the masses (in atomic mass units)
of the target molecule and reactant ion, respectively, z is the number
of charges on the reactant ion, S ()0.814) is the geometry factor for
the 6 cm × 6 cm cylindrical ICR cell, and d is the diameter of the cell
(0.060 m).
(18) A correlated sweep centered on m/ z 775 was used to ensure
that only intact 1 entered the ICR cell. See the Supporting Information
for details.
(19) In eq 2, rion and rtarget are the average radii (in m) of the ion
and target molecule, P (in atm) is the pressure of the target gas, T (K)
is the temperature of the cell/target gas, and the other symbols are as
defined above.
(20) For the calculation of collision frequency, mion ) 775, rion + rtarget
) 1 × 10-9 m, P ) 10-11 atm, z ) 1, Vpp ) 80 V, S ) 0.814, T ) 300
K, and d ) 0.06 m. Uncertainty in the result arises from estimates
(see Supporting Information) of the effective collision radii and the
efficiency of energy transfer in the collisions, since some of the collisions
will be glancing (a minority, based on the relative sizes of the projectile
ion and the target molecule). However, the largest uncertainties stem
from the potential error in the measured N2 pressure in the collision
cell (vide infra).
(8) This route to the tris-substituted 1 must also include substitu-
tion, in 3, of either the remaining PPh3 or the chloride, yielding (η5-
indenyl)Ru(HPPh2)2Cl or [(η5-indenyl)Ru(PPh3)(HPPh2)2]+Cl-, respec-
tively, but so far neither of these bis(diphenylphosphine) intermediates
has been observed by NMR spectroscopy.
(9) Friesen, D. M.; Rosenberg, L. Unpublished results.
(10) Techniques do exist for manipulating air- and moisture-
sensitive compounds in ES experiments and also for activating neutral
species for electrospray. (See ref 5a and references therein.)
(11) (a) Henderson, W.; Fawcett, J .; Kemmitt, R. D. W.; Russell, D.
R. J . Chem. Soc., Dalton Trans. 1995, 3007. (b) Hori, H.; Ishitani, O.;
Koike, K.; Takeuchi, K.; Ibusuki, T. Anal. Sci. 1996, 12, 587. (c) Hori,
H.; J ohnson, F. P. A.; Koike, K.; Takeuchi, K.; Ibusuki, T.; Ishitani, O.
J . Chem. Soc., Dalton Trans. 1997, 1019.
(12) Cone voltage refers to the potential difference between the
electrospray nozzle and the skimmer, through which ions pass for mass
detection. This value is also referred to as the nozzle-to-skimmer
potential difference (PD).
(13) The dimeric structure is supported by the isotopic pattern
observed.
(14) (a) Burnier, R. C.; Cody, R. B.; Freiser, B. S. J . Am. Chem. Soc.
1982, 104, 7436. (b) Cody, R. B.; Burnier, R. C.; Freiser, B. S. Anal.
Chem. 1982, 54, 96. (c) Ions generated by electrospray are de-energized
in a hexapole ion trap prior to injection into the ICR cell.
(15) Hop, C. E. C. A.; McMahon, T. B.; Willett, G. D. Int. J . Mass
Spectrom. Ion. Processes 1990, 101, 191.
(21) See the Supporting Information for details of the choice of CID
delay times.
(22) More detailed analysis (vide infra) indicates that this estimate
is too low but confirms that single-collision conditions do predominate
for a CID delay of 0.05 s.
(23) Kim, M. S. Int. J . Mass Spectrom. Ion Phys. 1983, 50, 189.

5690 Organometallics, Vol. 21, No. 26, 2002
Communications
energy for loss of one phosphine ligand from 1 is 69.5 (
10 kJ mol-1 (16.6 ( 2.3 kcal mol-1) and for dissociation
of two ligands is 145 ( 10 kJ mol-1 (34.6 ( 2.3 kcal
mol-1).
In the context of the potential participation of com-
plexes 1-3 in catalysis, the dissociation energy calcu-
lated for loss of just one phosphine from 1, to give the
16-electron cation 4, is most relevant. Substitution of
this first phosphine is critical for further reactivity at
the (indenyl)Ru core in 1, to pull this complex (back)
into a catalytic cycle, and previous studies of 2 have
shown phosphine substitution reactions at this (inde-
nyl)Ru fragment to proceed by dissociative mecha-
nisms.29 Our estimated Ru-PPh2H dissociation energy
of 16.6 kcal mol-1 falls between values for a Ru-PPh3
bond (BDE ) 14.0 kcal mol-1) and a Ru-PPh2Me bond
(BDE ) 19.7 kcal mol-1) calculated from solution
calorimetry studies of a series of substitution reactions
of phosphines at the neutral Cp*RuCl core.30 The strong
dependence of phosphine basicity (as gauged by pKb
values)31 on the inductive effects of substituents at
phosphorus32 suggests that diphenylphosphine should
bind less strongly than either PPh2Me or PPh3. How-
ever, for 3° phosphines, at least, Ru-P bond enthalpies
at Cp*RuCl correlate much more strongly with the
relative bulkiness of the phosphine than with electronic
features. In this context our calculated value might be
expected to be somewhat higher than the 19.7 kcal
mol-1 reported for Ru-PPh2Me. Our estimated value
may represent a balance point for electronic and steric
parameters that is unique to 2° phosphines (by com-
parison with 3° phosphines). This can only be verified
by collecting similar data for analogous cationic com-
plexes of a broad range of secondary phosphines.
The Cp*Ru-P bond energies determined from the
measured enthalpies of a series of substitution reactions
apply only within that system under the conditions used
(e.g. solvent, temperature). In contrast, our estimated,
gas-phase dissociation energies are absolute. There are
concerns over inaccuracies in CID-derived bond dis-
sociation energy data, which may arise from the variety
of (unknown) vibrational and rotational excited states
involved before and after an ion/molecule collision.33 In
this context and given the inherent lack of dynamic
range and sensitivity of the ICR technique relative to,
for example, ion beam methods,34 our estimated Ru-P
BDE values must be considered minimum values.
Nonetheless, the ES-FTICR-CID technique has the
potential for rapid determination of useful bond dis-
sociation energies in organometallic complexes. Ideally,
calorimetry and ES-MS techniques such as FTICR-CID
could provide complementary data, allowing correlation
F igu r e 1. Normalized product ion to precursor ion ratios
from CID experiments on 1 (m/ z 775).
V. The normalized abundance24 of each of the product
ions 4 and 5, relative to that of the parent ion 1, was
plotted as a function of the calculated collision energies,
as shown in Figure 1.25 The plots in Figure 1 appear to
have extended linear regions, as is typical above thresh-
old in collision experiments,26 despite the probable
increasing importance of multiple collisions as trf in-
creases. Although the random thermal motion of the
target molecules and the parent ions prior to accelera-
tion can introduce a spread in the center-of-mass
collision energy, in these experiments the thermal
motion of the parent ion (1) prior to acceleration is
negligible in this context.27 However, the thermal mo-
tion of the target molecule (N2) does influence the
computed threshold curves shown in Figure 1, producing
curvature prior to the beginning of the linear portion.
Thus, extrapolation of the linear portion back to a zero
abundance of product ions does not provide a true value
of Ecom(threshold) but instead gives Ecom(threshold) -
3γkT, where γ ) mi/(mi + mt), k is the Boltzmann
constant, and T is the temperature of the target gas. In
our experiments this correction term is small: at 300
K, with mi ) 775 and mt ) 28, we find 3γkT ) 0.075
eV/molecule ()1.8 kcal/mol ) 7.2 kJ /mol). The plots for
dissociation of one and two phosphine ligands from 1
extrapolate to axis values of 0.642 ( 0.095 and 1.428 (
0.087 eV, respectively.28 Adding the 0.075 eV correction
term to these values gives Ecom(threshold) values of 0.72
( 0.10 eV and 1.50 ( 0.10 eV for the dissociation pro-
cesses, respectively. Thus, our estimated dissociation
(24) To account for the increase in collision frequency with trf, the
ion ratios plotted in Figure 1 were normalized to the ratio at trf ) 50
µs by multiplying the ratio at trf ) 50 µs by 50/x, where x is the
excitation time for the measurement.
(25) These data imply a collision probability of 0.80, rather than
the 0.13 estimated from eq 2. This may be attributed to a N2 pressure
in the ICR cell ≈6 times higher than that recorded by the manifold
ion gauge. See ref 22 and the Supporting Information for details of
the calculation of collisional probabilities for these Poisson distribu-
tion(2s6. ) (a) Graul, S. T.; Squires, R. R. Int. J . Mass Spectrom. Ion
Processes 1987, 81, 183. (b) Graul, S. T.; Squires, R. R. J . Am. Chem.
Soc. 1989, 111, 892. (c) Kinter, M. T.; Bursey, M. M. J . Am. Chem.
Soc. 1986, 108, 1797. (d) Marinelli, P. J .; Squires, R. R. J . Am. Chem.
Soc. 1989, 111, 4101.
(29) Gamasa, M. P.; Gimeno, J .; Gonzalez-Bernardo, C.; Martin-
Vaca, B. M.; Monti, D.; Bassetti, M. Organometallics 1996, 15, 302.
(30) Luo, L.; Nolan, S. P. Organometallics 1994, 13, 4781. These
experiments gave Ru-PR3 dissociation energies ranging from 9.4 to
23.8 kcal mol-1
.
(31) Quin, L. D. A Guide to Organophosphorus Chemistry; Wiley:
New York, 2000.
(32) Estimated electronic parameters for phosphines indicate a
decrease in ligand donor ability with variation of substituents in the
order Me > Ph > H: Tolman, C. A. J . Am. Chem. Soc. 1970, 92, 2953.
(33) The similarity of our BDE values to those measured by solution
calorimetry for Cp*Ru systems is reassuring on this count. Inaccuracies
arising from a range of accessible vibrational/rotational states may be
insignificant on the scale of BDE being measured and within the levels
of precision possible for the FTICR technique.
(27) Chantry, P. J . J . Chem. Phys. 1971, 55, 2746.
(28) See the Supporting Information for details of the linear regres-
sion analysis.
(34) Armentrout, P. B. J . Am. Soc. Mass Spectrom. 2002, 13, 419.

Communications
Organometallics, Vol. 21, No. 26, 2002 5691
of metal-ligand bond strength trends, established by
solution calorimetry, with absolute (if minimum) bond
strengths provided by mass spectrometric studies of
intact, solution-stable organometallic complexes.
Ack n ow led gm en t. This work was supported by
grants from the Natural Sciences and Engineering
Research Council of Canada, the University of Mani-
toba, and the Australian Research Council.
We have recently prepared cationic analogues of 1,
which contain other secondary phosphines or mixtures
of two primary phosphines and a single triphenylphos-
phine. Our continued study of this system of complexes
using combined ES-FTICR-MS will allow us to create a
useful library of minimum Ru-P bond dissociation ener-
gies and to probe, in the mixed-phosphine complexes,
the relative binding affinities of 1° and 3° phosphines.
Su p p or tin g In for m a tion Ava ila ble: Text giving full
synthetic and mass spectrometric experimental details and
figures giving descriptions of the equipment used in these
experiments. This material is available free of charge via the
Internet at http://pubs.acs.org.
OM020308U