An investigation into the reactivity of organometallic noble gas complexes: A time-resolved infrared study in supercritical noble gas and alkane solution at room temperature

Article abstract of DOI:10.1021/jp993736v

A study investigated the effect of cyclopentadienyl ring substituents on the reactivity of the Group 7 half-sandwich complexes, (η⁵-C₅R₅)M(CO)₂L (M = Mn and Re; R = H, Me and Et (Mn only); L = Kr and Xe) toward CO in supercritical fluid solution at room temperature. The steric bulk of the three types of ring substituent steadily increased in the order H < Me < Et, and the Me and Et substituents had nearly identical electron direct effects. Room temperature CO substitution kinetics indicated that for both noble gases, the H and Me substituted complexes, (η⁵-C₅R'₅)Mn(CO)₂Xe (R' = H and Me) and (η⁵-C₅R'₅)Mn(CO)₂Kr had very similar reactivity toward CO, while (η⁵C₅Et₅)Mn(CO)₂L (L = Xe and Kr) were approximately twice as reactive. Experiments with the manganese xenon complexes suggested that the noble gas complexes react with CO in supercritical solution via a dissociative mechanism. This reaction mechanism was further evidenced by temperature dependence studies where the enthalpies of activation were calculated for these noble gas complexes and the analogous alkane complexes.

Full text of DOI:10.1021/jp993736v

4300
J. Phys. Chem. A 2000, 104, 4300-4307
An Investigation into the Reactivity of Organometallic Noble Gas Complexes: A
Time-Resolved Infrared Study in Supercritical Noble Gas and Alkane Solution at Room
Temperature
David C. Grills, Xue Z. Sun, Gavin I. Childs, and Michael W. George*
School of Chemistry, UniVersity of Nottingham, UniVersity Park, Nottingham NG7 2RD, U.K.
ReceiVed: October 20, 1999; In Final Form: February 18, 2000
A series of Group 7 organometallic noble gas complexes, (η5-C5R5)M(CO)2L (M ) Mn and Re; R ) H, Me.
and Et (Mn only); L ) Kr and Xe) have been characterized in supercritical noble gas solution at room
temperature using fast time-resolved infrared spectroscopy. The kinetics and activation parameters for the
reaction of these complexes with CO were measured and compared to the analogous alkane complexes, (η5-
C5H5)M(CO)2(n-heptane) (M ) Mn and Re). Evidence obtained from values of the activation enthalpies and
experiments, in which the lifetimes of (η5-C5R5)Mn(CO)2Xe (R ) H, Me, and Et) were measured as a function
of [CO] at a constant [CO]/[Xe] ratio, suggests that the reaction of the noble gas complexes with CO occurs
mainly via a dissociative or dissociative interchange mechanism.
Introduction
co-workers studied the bonding of Xe and Kr atoms to M(CO)5
fragments (M ) Cr and Mo (Xe only) and W) in the gas phase
using TRIR spectroscopy.15 Pyykko¨ performed a theoretical
study of the gold(I) cation complexes, AuL+ and LAuL+ (L )
He, Ne, Ar, Kr, and Xe),16 and found the Au-Xe BDEs to be
87.8 kJ mol-1 in AuXe+ and 108.5 kJ mol-1 in XeAuXe+, with
the Au-Xe bonds having considerable covalent character. More
recently Schro¨der and Pyykko¨ performed a combined experi-
mental and high level theoretical investigation of cationic gold-
(I) complexes and the predicted Au-Xe BDE in AuXe+
Organometallic noble gas and alkane complexes have been
known for over 20 years.1 Perutz and Turner2 used matrix
isolation to show that M(CO)5L (M ) Cr, Mo and W; L ) Ar,
Kr, Xe and CH4) could be generated following UV photolysis
of M(CO)6 in low temperature (12 K) matrices. Matrix isolation
has since become a powerful technique for the characterization
of organometallic noble gas complexes including [Fe(CO)4Xe],3
[Mn(CO)5Kr],4 [Fe(CO)5Kr]+,4 [Cr(CO)3(dfepe)L] (dfepe )
(C2F5)2PCH2CH2P(C2F5)2; L ) Ar and Xe),5 [Ru(CO)2-
(PMe3)2L] (L ) Ar and Xe)6 and [Ru(CO)2(dmpe)L] (dmpe )
Me2PCH2CH2PMe2; L ) Ar and Xe).7 Although matrix isolation
permits the characterization of unstable species at low temper-
ature, it does not provide any kinetic information quantifying
the reactivity of these.
increased to 126.8 kJ mol-1 17
.
We have recently shown that organometallic noble gas
complexes can be observed in fluid solution at room temperature
following irradiation of metal carbonyls in supercritical noble
gas solution. Using TRIR spectroscopy, we characterized
M(CO)5L (M ) Cr, Mo, and W; L ) Ar (W only), Kr and
Xe)18 and (η5-C5H5)Re(CO)2L (L ) Kr and Xe).19 The rate
constant for the reaction of these complexes with CO increases
as follows: Xe < Kr < Ar; Re , W < Mo ≈ Cr. Surprisingly,
we also found that the rate constants for the Xe complexes,
LmM(CO)x-1(Xe) were very similar to those for the correspond-
ing alkane complexes, LmM(CO)x-1(n-heptane).
The stability of (η5-C5H5)Re(CO)2(alkane) is significant, and
Ball and co-workers have recently been able to characterize
(η5-C5H5)Re(CO)2(cyclopentane) at low temperature by 1H
NMR spectroscopy.20 The factors governing the reactivity of
unstable alkane complexes have been investigated. Johnson et
al. found that the reactivity of (η5-C5R5)Mn(CO)2(n-heptane)
(R ) H, Me and Et) toward small molecules such as CO, N2,
and H2 in n-heptane solution steadily increased in the order H
< Me < Et while the reactivity toward PPh3 did not change.21
These results indicated that steric rather than electronic factors
governed the reactivity of these alkane complexes. Long and
co-workers22 found that the reactivity of (η6-C6R6)Cr(CO)2-
(cyclohexane) (R ) H, Me and Et) toward CO in cyclohexane
also increased in the order H < Me < Et. Changing the solvent
from cyclohexane to longer and longer linear alkanes resulted
in an increased rate of reaction between (η6-arene)Cr(CO)2-
Nanosecond flash photolysis experiments8 showed that Cr-
(CO)5(cyclohexane) is formed within 50 ns following photolysis
of Cr(CO)6 in cyclohexane solution, and subsequent ultrafast
transient spectroscopy9,10 has shown that the solvation occurs
within the first picosecond following UV irradiation. 16-Electron
intermediates generated by flash photolysis of the metal
carbonyls usually form complexes with a solvent molecule. The
only exception to this was reported by Bergman and co-
workers11 where (η5-C5R5)Co(CO) (R ) H and Me) appeared
not to interact with the solvent following flash photolysis of
(η5-C5R5)Co(CO)2 in liquefied noble gas solution.
Organometallic noble gas complexes have also been observed
in solution. Simpson et al.12 showed that Cr(CO)5Xe could be
detected in solution at low temperature using FTIR spectroscopy;
either in liquefied Xe (lXe) at - 98 °C or liquefied Kr (lKr)
doped with Xe at -122 °C. Weiller extended this work using
rapid scan FTIR13 to identify M(CO)5L (M ) Cr and W; L)
Kr and Xe) in liquefied noble gas solution at low temperature,
and Bergman and co-workers used a similar approach to detect
(η5-C5Me5)Rh(CO)L (L ) Kr and Xe).14
Organometallic noble gas complexes have been observed in
the gas phase at room temperature. For example, Weitz and
* E-mail: Mike.George@nottingham.ac.uk
10.1021/jp993736v CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/19/2000

Reactivity of Organometallic Noble Gas Complexes
J. Phys. Chem. A, Vol. 104, No. 18, 2000 4301
(alkane) and CO. The enthalpy of activation, ∆H‡, for this
reaction did not vary with substituent or alkane (24 ( 2 kJ
mol-1) and was only half that of the Cr-cyclohexane bond
energy (53 ( 9 kJ mol-1) determined previously by photo-
acoustic calorimetry (PAC).23 They concluded that the variation
in reactivity of the (η6-arene)Cr(CO)2(alkane) photoproduct
toward CO can be explained by changes in the entropy of
activation, ∆S‡, and not by variations in the interaction energy
between the coordinatively unsaturated photofragment and the
solvent. An interchange mechanism was proposed for the
displacement of alkane by CO. More recently, Long and co-
workers24 performed a systematic investigation into the reactivity
of M(CO)5(alkane) (M ) Cr, Mo and W) and (η6-arene)Mo-
(CO)2(alkane) toward CO in alkane solution. For these systems,
they also found that ∆H‡ remained constant, and an interchange
mechanism was proposed for the Cr systems, with the rate
constant, k2, depending on the ∆S‡ expression. For the Mo and
W systems however, the nature of the solvent had much less of
an effect on k2 and a more associative mechanism was proposed.
In this paper, we have investigated the effect of cyclopen-
tadienyl ring substituents on the reactivity of the Group 7 half-
sandwich complexes, (η5-C5R5)M(CO)2L (M ) Mn and Re; R
) H, Me and Et (Mn only); L ) Kr and Xe) toward CO in
supercritical fluid solution at room temperature.
Figure 1. (a) FTIR spectrum of (η5-C5H5)Mn(CO)3 in scXe (1395 psi,
25 °C) in the presence of CO (30 psi). (b) TRIR spectrum obtained
500 ns after photolysis (355 nm) of this solution.
C5Et5)Mn(CO)3 was a gift from Dr. A. G. Ginzburg. n-heptane
was distilled over CaH2 prior to use.
Experimental Section
The Nottingham laser TRIR apparatus has been described in
detail elsewhere.25 In these experiments the photochemical
reactions were initiated by a pulsed Nd:YAG laser (Quanta-
Ray GCR-12), and a cw IR diode laser (Mu¨tek MDS 1100)
was used to monitor the transient IR absorptions. The IR spectra
are built up on a “point-by-point” basis by repeating this
measurement at different infrared frequencies. The stainless steel
cell used for supercritical TRIR measurements has been
described previously for conventional spectroscopic monitor-
ing.26 For TRIR experiments, the cells were used with CaF2
windows, and a pressure transducer (RDP Electronics). Heating
was achieved with an external aluminum jacket containing
cartridge heaters. Kr and Xe (BOC, research grade), CO (Air
Products, premier grade), (η5-C5H5)Mn(CO)3 (Aldrich), (η5-
C5H5)Re(CO)3 (Strem), and (η5-C5Me5)Re(CO)3 (Strem) were
used as supplied. (η5-C5Me5)Mn(CO)3 and (η5-C5H4Et)Mn(CO)3
were synthesized according to literature procedures27,28 and (η5-
Results and Discussion
Characterization and Reactivity of (η5-C5R5)M(CO)2Xe
(M ) Mn and Re; R ) H, Me and Et (Mn only)) in scXe.
Figure 1(a) shows the FTIR spectrum of (η5-C5H5)Mn(CO)3
dissolved in scXe (1395 psi, 25 °C) in the presence of CO (30
psi). Figure 1(b) shows the TRIR spectrum of this solution
obtained 500 ns after photolysis (355 nm). The parent absorp-
tions are bleached and two new absorptions are produced which
can readily be assigned to (η5-C5H5)Mn(CO)2Xe by comparison
with previous matrix isolation results29 and TRIR experiments21
(see Table 1). The formation of (η5-C5H5)Mn(CO)2Xe is
reversible with (η5-C5H5)Mn(CO)2Xe decaying (observed rate,
kobs ) 2.5 (( 0.3) × 105 s-1) to reform (η5-C5H5)Mn(CO)3 (kobs
) 2.4 (( 0.2) × 105 s-1). This pseudo-first-order rate constant
depends linearly on CO concentration, affording the second-
order rate constant, kCO for the reaction of (η5-C5H5)Mn(CO)2-
TABLE 1: ν(C-O) Infrared Band Positions (cm-1) of (η5-C5R5)M(CO)2L Species (M ) Mn and Re; R ) H, Me and Et (Mn
only); L ) CO, n-heptane, Kr and Xe)a
R
complex
conditions
H
Me
Et
(η5-C5R5)Mn(CO)3
(η5-C5R5)Mn(CO)3
(η5-C5R5)Mn(CO)3
(η5-C5R5)Mn(CO)3
(η5-C5R5)Mn(CO)2
(η5-C5R5)Mn(CO)2(hept)
(η5-C5R5)Mn(CO)2Kr
(η5-C5R5)Mn(CO)2Xe
matrix/20 K
n-heptane/298 K
scKr/298 K
2033, 1951b
2029, 1947d
2035, 1956
2032, 1952
1972, 1903b
1962, 1894d
1973, 1908
1970, 1903
2008, 1923c
2008, 1925d
2015, 1936
2013, 1932
1932, 1863c
1942, 1873d
1948, 1884
1945, 1882
2008, 1927d
2014, 1935
2011, 1931
scXe/298 K
matrix/20 K
n-heptane/298 K
scKr/298 K
1942, 1873d
1950, 1885
1945, 1880
scXe/298 K
(η5-C5R5)Re(CO)3
(η5-C5R5)Re(CO)3
(η5-C5R5)Re(CO)3
(η5-C5R5)Re(CO)3
(η5-C5R5)Re(CO)2
(η5-C5R5)Re(CO)2(hept)
(η5-C5R5)Re(CO)2Kr
(η5-C5R5)Re(CO)2Xe
matrix/20 K
n-heptane/298 K
scKr/298 K
2029, 1934e
2031, 1940g
2038, 1952g
2035, 1946g
1947, 1879e
1954, 1890g
1966, 1903g
1957, 1894g
2015, 1921f
2020, 1933
2016, 1927
1945, 1879f
scXe/298 K
matrix/20 K
n-heptane/298 K
scKr/298 K
1944, 1882
1936, 1873
scXe/298 K
a This work unless referenced. b Reference 29. c Reference 34. d Reference 21. e Reference 35. f Reference 36. g Reference 19.

4302 J. Phys. Chem. A, Vol. 104, No. 18, 2000
Grills et al.
TABLE 2: Values of kCO for Reaction of the Group 7 Noble
Gas and n-heptane Complexes, (η5-C5R5)M(CO)2L (M ) Mn
and Re; R ) H, Me and Et (Mn only); L ) n-heptane, Kr
and Xe) with CO at 298 K
kCO (dm3
pressure
(psi)b
a
M
R
L
conditions
mol-1 s-1
)
Mn
Mn
Mn
Re
H
Me
Et
n-heptane
n-heptane
n-heptane
n-heptane
n-heptane
n-heptane
n-heptane
n-heptane
8.1 × 105
1.6 × 106
5.4 × 106
2.1 × 102
H
Mn
Mn
Mn
Re
H
Me
Et
H
Me
Xe
Xe
Xe
Xe
Xe
scXe
scXe
scXe
scXe
scXe
1.6 × 106
1.8 × 106
2.9 × 106
4.8 × 103
6.0 × 103
2180
1720
2120
1935
2570
Re
Mn
Mn
Mn
Re
H
Me
Et
H
Me
Kr
Kr
Kr
Kr
Kr
scKr
scKr
scKr
scKr
scKr
7.2 × 107
7.5 × 107
1.6 × 108
8.1 × 106
7.8 × 106
3290
3360
4080
3500
3080
Re
a (10%. b (10 psi.
Figure 2. TRIR spectra obtained after photolysis (355 nm) of (a) (η5-
C5H5)Mn(CO)3 in scXe (1395 psi, 25 °C) in the presence of CO (30
psi) (b) (η5-C5Me5)Mn(CO)3 in scXe (1420 psi, 25 °C) in the presence
of CO (15 psi) and (c) (η5-C5Et5)Mn(CO)3 in scXe (1430 psi, 25 °C)
in the presence of CO (30 psi).
Figure 4. TRIR spectra obtained after photolysis (266 nm) of (a) (η5-
C5H5)Re(CO)3 in scXe (1720 psi, 25 °C) in the presence of CO (60
psi) and (b) (η5-C5Me5)Re(CO)3 in scXe (2600 psi, 25 °C) in the
presence of CO (80 psi).
Figure 3. (a) TRIR decay traces recorded at wavenumbers corre-
sponding to (η5-C5R5)Mn(CO)2Xe (R ) H, Me and Et), after photolysis
(355 nm) of (η5-C5R5)Mn(CO)3 in scXe at 25 °C in the presence of
CO (75 psi). The traces were recorded in three separate experiments
and have been normalized in intensity. (b) Plots of the observed rate
constant, kobs versus [CO] for the reaction of (η5-C5R5)Mn(CO)2Xe (R
) H (b), Me (9) and Et (2)) with excess CO in scXe at 25 °C.
Et5)Mn(CO)2Xe are similarly shifted relative to those of (η5-
C5H5)Mn(CO)2Xe. The formation of (η5-C5Me5)Mn(CO)2Xe
and (η5-C5Et5)Mn(CO)2Xe is reversible.
The reactivity of (η5-C5R5)Mn(CO)2Xe toward CO in scXe
follows the order R ) Et > Me ≈ H (see Figure 3 and Table
2). This is clearly not the same as the reactivity of the
corresponding alkane complexes, (η5-C5R5)Mn(CO)2(n-heptane),
where the reactivity was found21 to increase in the order H <
Me < Et (Table 2). We have investigated further the similarity
in the reactivity of (η5-C5H5)Mn(CO)2Xe and (η5-C5Me5)Mn-
(CO)2Xe by comparing the reactivity of (η5-C5H5)Re(CO)2Xe
and (η5-C5Me5)Re(CO)2Xe in scXe.
Xe with CO in scXe (kCO ) 1.6 (( 0.2) × 106 dm3 mol-1 s-1).
Similarly, TRIR experiments with (η5-C5Me5)Mn(CO)3 and (η5-
C5Et5)Mn(CO)3 permit the characterization of (η5-C5Me5)Mn-
(CO)2Xe and (η5-C5Et5)Mn(CO)2Xe, respectively (see Figure
2 and Table 1). The ν(C-O) band positions of (η5-C5Me5)Mn-
(CO)3 and (η5-C5Et5)Mn(CO)3 are very similar and are shifted
down in wavenumber relative to (η5-C5H5)Mn(CO)3, reflecting
the higher electron density on the Mn center in these complexes.
The ν(C-O) bands of (η5-C5Me5)Mn(CO)2Xe and (η5-C5-
We have previously characterized and measured the reactivity
of (η5-C5H5)Re(CO)2Xe in scXe.19 Figure 4(b) shows the TRIR
spectrum obtained 15 µs after photolysis (266 nm) of (η5-C5-

Reactivity of Organometallic Noble Gas Complexes
J. Phys. Chem. A, Vol. 104, No. 18, 2000 4303
Me5)Re(CO)3 in scXe (2600 psi, 25 °C) in the presence of CO
(80 psi). As for the analogous Mn system, the parent absorptions
of (η5-C5Me5)Re(CO)3 are shifted down in frequency by ca. 20
cm-1 relative to those of (η5-C5H5)Re(CO)3 (Table 2). The two
new absorptions that are produced are similarly shifted relative
to those of (η5-C5H5)Re(CO)2Xe and are assigned to (η5-C5-
Me5)Re(CO)2Xe. We find that (η5-C5H5)Re(CO)2Xe and (η5-
C5Me5)Re(CO)2Xe have a similar reactivity toward CO (Table
2).
All of the TRIR experiments were carried out in the presence
of excess CO under pseudo-first-order conditions. A detailed
rate analysis for the decay of noble gas complexes in liquid
xenon has been performed by Weiller13 for W(CO)5Xe and
Bergman et al.14 for (η5-C5Me5)Rh(CO)Xe. We have used a
similar approach to obtain expressions for the observed rate of
decay, kobs, of the xenon complexes Cp′M(CO)2Xe (Cp′ ) (η5-
C5H5), (η5-C5Me5), and (η5-C5Et5); M ) Mn and Re) in their
reaction with CO in scXe solution. Plots of kobs versus [CO]
for all of the complexes studied were linear with nonzero
intercepts. The gradient, kCO, is equal to the second-order rate
constant for reaction of Cp′M(CO)2Xe with CO.
Measurement of kobs as a Function of [CO] at Constant
[CO]/[Xe] Ratio. In the condensed phase, such as lXe and
alkane solution, the concentration of solvent cannot be altered.
However, supercritical fluids allow the measurement of kobs as
a function of Xe concentration. For a dissociative mechanism,
kobs will depend on the concentration of the leaving ligand, Xe,
kobs 1/[Xe]
An associative process would exhibit no such dependence on
[Xe]. Therefore, an experiment in which kobs is measured as a
function of [Xe] should provide an insight into the nature of
the mechanism.
We have previously probed the CO substitution kinetics of
W(CO)5(CO2) in scCO2.18 We measured kobs as a function of
[CO] while holding the [CO]/[CO2] ratio constant. The rate
equations are analogous to those described here and so it was
predicted that under these conditions kobs would only decrease
with decreasing density if the reaction with W(CO)5(CO2) with
CO were partially associative. In fact, kobs was found to increase
slightly with decreasing density, implying that associative
processes were not significant. The second experiment involved
measuring the lifetime of W(CO)5(CO2) (1/kobs) as a function
of [CO2] at constant [CO]. For a dissociative process, 1/kobs
[CO2] and we were able to show that 1/kobs exactly follows the
change in CO2 density with pressure, therefore demonstrating
that the reaction of W(CO)5(CO2) with CO in scCO2 is
predominantly a dissociative process.
For a dissociatiVe mechanism in scXe, the following rate
equations would apply for the reaction of Cp′M(CO)2Xe with
CO:
Cp′M(CO)2Xe yk1z Cp′M(CO)2 + Xe
k-1
k2
Figure 5 shows how kobs, for the reaction of (η5-C5R5)Mn(CO)2-
Xe (R ) H, Me and Et) with CO in scXe, varies with increasing
[CO] while maintaining a constant [CO]/[Xe] ratio. The
expression for kobs, assuming both dissociative and associative
mechanisms, is given below
Cp′M(CO)2 + CO
98 Cp′M(CO)3
Application of the steady-state approximation to the intermedi-
ate, Cp′M(CO)2, results in the following expression for kobs
where k1 is the elementary rate constant for dissociation of Xe
,
k1[CO]
k1k2[CO]
k1[CO]
kobs
∼
+ ka[CO]
kobs
)
∼
[Xe]
k-1[Xe] + k2[CO] [Xe]
Therefore, if both dissociative and associative pathways are
operating simultaneously, a plot of kobs versus [CO] at constant
[CO]/[Xe] ratio should produce a straight line with gradient
equal to ka and intercept equal to k1[CO]/[Xe]. If the slope is
zero then associative pathways can be assumed to be insignifi-
cant. Figure 5 shows that at a constant [CO]/[Xe] ratio, kobs
shows no appreciable dependence on [CO] for the reaction of
(η5-C5R5)Mn(CO)2Xe with CO. From these experiments we
conclude that the contribution from an associative mechanism
for the reaction of (η5-C5R5)Mn(CO)2Xe with CO in scXe is
not significant and the reaction proceeds via either a purely
dissociative or a dissociative interchange mechanism. A more
definitive experiment would be to measure 1/kobs as a function
of [Xe] at constant [CO]. In this case a dissociative mechanism
should produce a linear plot of 1/kobs versus [Xe], while an
associative or interchange mechanism would have no slope. The
amount of xenon required for this was prohibitive and these
experiments have not been conducted.
This expression was reduced using the following approxima-
tion: k-1[Xe] . k2[CO]. Ligand recombination rates are
generally very fast. For example, the rate of reaction of CpMn-
(CO)2 with CO has been measured30 in the gas phase to be 6 ×
1011 cm3 mol-1 s-1. Therefore, it is likely that in supercritical
solution, Xe and CO will compete to react with Cp′M(CO)2 at
near diffusion-controlled rates (k-1 and k2, respectively). The
diffusion-controlled rate constant, kd can be estimated in scXe
using the Stokes-Einstein relation, giving k-1 ∼ k2 ∼ 5 × 1010
dm3 mol-1 s-1. Thus, since [Xe] ∼ 70 × [CO] under the
conditions of this experiment, the above approximation is
reasonable.
For an associatiVe mechanism, the following rate equation
would apply, where ka is the bimolecular associative rate
constant
ka
Cp′M(CO)2Xe + CO 98 Cp′M(CO)3 + Xe
Determination of Activation Parameters for the Reaction
of (η5-C5R5)M(CO)2Xe (M ) Mn and Re; R ) H, Me and
Et (Mn only)) with CO in scXe. For a given metal/ligand
system the BDE for a metal-Xe bond is expected to be
significantly less than that of the corresponding metal-heptane
bond. For example, the W-heptane BDE in W(CO)5(n-heptane)
has been determined31 using PAC (62.8 kJ mol-1) and is
approximately two times the W-Xe BDE in W(CO)5(Xe),
estimated by TRIR experiments in the gas phase at room
temperature15 (34.3 ( 4.2 kJ mol-1) and in lXe at cryogenic
This leads to the expression
kobs ) ka[CO]
Therefore, under the conditions of this experiment, solely from
the room-temperature CO substitution kinetics, it is impossible
to distinguish between the two extreme forms of mechanism,
since they both predict a linear dependence of kobs on [CO], as
was observed.

4304 J. Phys. Chem. A, Vol. 104, No. 18, 2000
Grills et al.
Figure 5. The CO concentration dependence of the observed rate constant, kobs for the reaction of (η5-C5R5)Mn(CO)2Xe (R ) H, Me and Et) with
excess CO in scXe, while maintaining a constant [CO]/[Xe] ratio.
temperatures13 (35.1 ( 0.8 kJ mol-1). We have shown18 that
the rate of reaction of W(CO)5Xe with CO in scXe is only ca.
2.5 times the rate for the corresponding reaction of W(CO)5-
(n-heptane) in n-heptane solution. The activation energy, Ea,
for the reaction of W(CO)5Xe with CO in scXe is almost
identical to the estimated W-Xe BDE. However, Long and co-
workers24 have shown that ∆H‡ for the reaction of W(CO)5(n-
heptane) with CO (20 ( 2 kJ mol-1) is less than half the
W-heptane BDE (62.8 kJ mol-1). These results are consistent
with the alkane complex reacting with CO via an associative
mechanism, while the Xe complex is reacting mainly through
a dissociative pathway.
We also find that (η5-C5R5)Mn(CO)2Xe and (η5-C5R5)Mn-
(CO)2(n-heptane) (R ) H, Me and Et) complexes have very
similar reactivity toward CO. We have therefore determined
the activation parameters for the reactions of (η5-C5H5)Mn(CO)2-
(n-heptane) and (η5-C5R5)Mn(CO)2Xe (R ) H, Me and Et) with
CO in n-heptane and scXe solution, respectively. In the
supercritical fluid environment, CO and Xe are completely
miscible over a wide temperature range, which greatly simplifies
variable temperature studies. Figure 6 shows Arrhenius and
Figure 6. (a) Arrhenius plots and (b) Eyring plots for the reaction of
Eyring plots from these experiments (see Table 3 for values of
(η5-C5R5)Mn(CO)2Xe with CO in scXe solution, R ) H (b), Me (9)
∆H‡). It is clear that for (η5-C5R5)Mn(CO)2Xe, all three
and Et (2) and (η5-C5H5)Mn(CO)2(n-heptane) with CO in n-heptane
complexes have a similar ∆H‡ and that this is similar to the
solution ([).
value obtained for the reaction of (η5-C5H5)Mn(CO)2(n-heptane)
with CO in n-heptane solution. The ∆H‡ value for (η5-
C5R5)Mn(CO)2Xe represents a lower limit for the Mn-Xe BDE.
Early PAC experiments by Yang and co-workers32 estimated
the Mn-heptane BDE, 8-9 kcalmol-1. The precise determi-
nation of the Mn-heptane BDE from PAC data requires the
Mn-CO BDE. This value was not known and Yang and co-
workers estimated the Mn-alkane binding energy from the
enthalpy of activation for the reaction of (η5-C5H5)Mn(CO)2-
(n-heptane) with cis-cyclooctene and the enthalpy of activation
for the reaction of (η5-C5H5)Mn(CO)2(cis-cyclooctene) with
PPh3. They assumed that the reaction was purely dissociative
and used these values to calculate the Mn-heptane BDE, 8-9
kcal mol-1. However, Burkey has pointed out that this is a lower
limit for the Mn-heptane BDE since the reaction may be an
associative process with respect to the alkane.33 Indeed, Long
and co-workers22,24 have shown that many similar reactions for
(η6-C6H6)M(CO)2(alkane) (M ) Cr, Mo) do not proceed by a
dissociative pathway and the enthalpy of activation for the
reaction of (η6-C6H6)M(CO)2(alkane) with CO is only half the
M-alkane BDE. The ∆H‡ determined for reaction of (η5-C5H5)-
Mn(CO)2(n-heptane) with CO is significantly lower than other
first row transition metal alkane BDEs. For example, the Cr-
heptane BDE determined24 by PAC is 50 ( 5 kJ mol-1. We
would expect that the Mn-heptane BDE . Mn-Xe BDE, and
this suggests that our value of ∆H‡ is not a good estimate of
the Mn-heptane BDE.

Reactivity of Organometallic Noble Gas Complexes
J. Phys. Chem. A, Vol. 104, No. 18, 2000 4305
TABLE 3: Activation Enthalpies for the Reaction of
(η5-C5R5)M(CO)2L (M ) Mn and Re; R ) H, Me and Et
(Mn only); L ) n-heptane, Kr and Xe) with CO
a
M
R
L
conditions
∆H‡ (kJ mol-1
)
Mn
Mn
Mn
Mn
H
H
Me
Et
n-heptane
Xe
Xe
n-heptane
scXe
scXe
36
30
34
28
Xe
scXe
Re
Re
Re
H
H
Me
n-heptane
Xe
Xe
n-heptane
scXe
scXe
46
47
44
Re
Me
Kr
scKr
31
a (2.
An indication of the nature of the mechanism would come
from the values of ∆S‡ derived from the Eyring plots.
Unfortunately, the large errors in these values obtained from
the intercept of the Eyring plots prevent such an analysis.
However, the fact that the (η5-C5R5)Mn(CO)2Xe complexes are
reacting with CO in scXe by a dissociative mechanism is
strongly supported by the experiments described earlier in which
kobs was measured as a function of [CO] at a constant [CO]/
[Xe] ratio. These results are consistent with (η5-C5H5)Mn(CO)2-
(n-heptane) and (η5-C5R5)Mn(CO)2Xe reacting with CO by
different mechanisms, and we tentatively suggest that in
n-heptane (η5-C5H5)Mn(CO)2(n-heptane) reacts with CO by an
associative or associative interchange mechanism and in scXe
(η5-C5R5)Mn(CO)2Xe reacts with CO by a dissociative or
dissociative interchange mechanism.
We have also made temperature dependent measurements for
the reaction of (η5-C5H5)Re(CO)2(n-heptane) with CO in
n-heptane solution and (η5-C5R′5)Re(CO)2Xe (R′ ) H and Me)
with CO in scXe. We have already shown that (η5-C5H5)Re-
(CO)2(n-heptane) and (η5-C5R5)Re(CO)2Xe have similar reac-
tivity toward CO (see Table 2). We also find that (η5-
C5H5)Re(CO)2Xe and (η5-C5H5)Re(CO)2(n-heptane) have very
similar ∆H‡ for reaction with CO. The ∆H‡ values are higher
than for the corresponding Mn complexes, reflecting the lower
reactivity of the Re complexes toward CO. The Re-heptane
BDE has not been estimated by PAC but since Re is a third
row transition metal we expect that the Re-heptane BDE will
be greater than or equal to the Mn-heptane BDE. ∆H‡ for the
reaction of (η5-C5H5)Re(CO)2(n-heptane) with CO is greater
than for the corresponding reaction of the Mn-alkane complex;
however, it is still significantly smaller than the W-heptane
BDE determined31 by PAC (62.8 kJ mol-1). We again suggest
that ∆H‡ obtained for the reaction of (η5-C5H5)Re(CO)2(n-
heptane) with CO is less than the Re-heptane BDE. Although
we have not measured the reaction of (η5-C5H5)Re(CO)2(Xe)
as a function of [CO] at constant [CO]/[Xe] ratio, we tentatively
suggest that the reaction of (η5-C5H5)Re(CO)2(n-heptane) with
CO in n-heptane occurs mainly by an associative or associative
interchange mechanism while η5-C5R′5)Re(CO)2Xe reacts with
CO in scXe predominantly by a dissociative or dissociative
interchange mechanism.
Figure 7. (a) FTIR spectrum of (η5-C5H4Et)Mn(CO)3 in scXe (1350
psi, 25 °C) in the presence of CO (30 psi). Note the small contamination
by (η5-C5H5)Mn(CO)3 (*) and (η5-C5H3Et2)Mn(CO)3 (•). (b) TRIR
spectrum obtained 500 ns after photolysis (355 nm) of this solution.
terminal C-H bond of an Et group in scXe. This would require
that in n-heptane, the stronger Mn-heptane bond displaces the
agostic interaction, i.e., the reactive species is (η5-C5Et5)Mn-
(CO)2 in scXe and (η5-C5Et5)Mn(CO)2(n-heptane) in n-heptane.
We have therefore synthesized28 (η5-C5H4Et)Mn(CO)3. This
molecule has only one Et substituent on the cyclopentadienyl
ring. If an agostic interaction were responsible for the increased
reactivity of (η5-C5Et5)Mn(CO)2Xe, then irradiation of this
complex in scXe would be expected to form a similar agostic
intermediate. The rate constant for reaction of (η5-C5H4Et)Mn-
(CO)2 with CO would be similar to that for (η5-C5Et5)Mn(CO)2.
Figure 7a shows the FTIR spectrum of (η5-C5H4Et)Mn(CO)3
in scXe (1350 psi, 25 °C) in the presence of CO (30 psi). The
parent ν(C-O) bands (at 2027 and 1946 cm-1) are shifted down
in frequency by ca. 4 cm-1 relative to those of (η5-C5H5)Mn-
(CO)3 in scXe. Figure 7b shows the TRIR spectrum obtained
500 ns after photolysis (355 nm) of this solution. Depletion of
the bands due to the starting material is accompanied by the
production of two new bands at 1961 and 1895 cm-1, which
are assigned to (η5-C5H4Et)Mn(CO)2Xe by comparison with the
TRIR spectrum of (η5-C5H5)Mn(CO)2Xe in scXe.
We now return to the reactivity of (η5-C5Et5)Mn(CO)2Xe with
CO in scXe, which is approximately two times greater than that
of (η5-C5H5)Mn(CO)2Xe and (η5-C5Me5)Mn(CO)2Xe. There is
not a steady increase in reactivity as was observed21 for (η5-
C5R5)Mn(CO)2(n-heptane) (R ) H, Me and Et) by Johnson et
al.
One possibility is that the greater length of the Et substituents
in (η5-C5Et5)Mn(CO)2Xe, compared with the H and Me ring
substituents in (η5-C5R′5)Mn(CO)2Xe, permits an agostic in-
teraction to occur between the manganese center and the
A plot of kobs versus [CO] for the reaction of (η5-C5H4Et)-
Mn(CO)2Xe with CO in scXe is shown in Figure 8. The linear
fit to this graph affords the bimolecular rate constant for reaction
with CO, kCO ) 8.0 (( 0.8) × 105 dm3 mol-1 s-1. Interestingly,
this value is less than that calculated for each of the (η5-
C5R5)Mn(CO)2Xe (R ) H, Me and Et) complexes discussed
previously (see Table 2) and therefore suggests that the increased
reactivity of (η5-C5Et5)Mn(CO)2Xe toward CO in scXe is not
due to an agostic interaction.

4306 J. Phys. Chem. A, Vol. 104, No. 18, 2000
Grills et al.
both the electronic and steric effects of the ring substituents on
reactivity to be investigated. Room temperature CO substitution
kinetics showed that for both noble gases, the H and Me
substituted complexes, (η5-C5R′5)Mn(CO)2Xe (R′ ) H and Me)
and (η5-C5R′5)Mn(CO)2Kr have very similar reactivity toward
CO, whereas (η5-C5Et5)Mn(CO)2L (L ) Xe and Kr) are
approximately twice as reactive. Experiments were conducted
with the manganese xenon complexes in which the observed
rate of decay was measured as a function of [CO] at a constant
[CO]/[Xe] ratio. These suggested that the noble gas complexes
react with CO in supercritcal solution via a mechanism that is
dissociative in nature. Further evidence for the reaction mech-
anism was obtained from temperature dependence studies in
which the enthalpies of activation were calculated for these noble
gas complexes and the analogous alkane complexes, (η5-
C5H5)M(CO)2(n-heptane) (M ) Mn and Re). The value of ∆H‡
for the reaction of (η5-C5H5)M(CO)2Xe (M ) Mn or Re) with
CO in scXe was found to be very similar to the analogous
reaction of (η5-C5H5)M(CO)2(n-heptane) in n-heptane solution.
The value of ∆H‡ for (η5-C5R′5)M(CO)2Xe represents a lower
limit for the M-Xe BDE. We suggest that (η5-C5H5)Mn(CO)2-
(n-heptane) and (η5-C5R5)Mn(CO)2Xe are reacting with CO by
different mechanisms. Further studies are currently underway
involving different alkanes and ring substituents in order to
understand fully the mechanisms involved.
Figure 8. Plot of the observed rate constant, kobs versus [CO] for the
reaction of (η5-C5H4Et)Mn(CO)2Xe with excess CO in scXe at 25 °C.
Comparison of the rate constants for the reaction of (η5-C5-
Et5)Mn(CO)2Xe and (η5-C5Et5)Mn(CO)2Kr with CO in scXe
and scKr, respectively, provides further insight into the possible
role of an agostic interaction in the reaction of (η5-C5Et5)Mn-
(CO)2L (L ) Xe and Kr) with CO. If an agostic interaction
was formed following CO dissociation, then the rates of reaction
of (η5-C5Et5)Mn(CO)2Xe in scXe and (η5-C5Et5)Mn(CO)2Kr in
scKr should be very similar.
Acknowledgment. We thank Professors J. J. Turner and.
M. Poliakoff for helpful discussions and Mr. K. Stanley and
Mr. M. Guyler for their technical help. We are grateful to
EPSRC, the Isle of Man Government (studentship D.C.G.) and
the University of Nottingham (studentship X.Z.S.) for financial
support.
The Reactivity of (η5-C5R5)M(CO)2Kr (M ) Mn and Re;
R ) H, Me and Et (Mn only)) in scKr. The krypton
complexes, (η5-C5R5)M(CO)2Kr (M ) Mn and Re; R ) H, Me
and Et (Mn only)) have been characterized in scKr (Table 1)
and their room temperature CO substitution kinetics have been
determined (Table 2). The rates are significantly higher than
those found for the corresponding xenon complexes. More
importantly, the trend in reactivity of the krypton complexes is
very similar to that observed for the xenon complexes
References and Notes
(1) Hall, C.; Perutz, R. N. Chem. ReV. 1996, 96, 3125.
(2) Perutz, R. N.; Turner, J. J. J. Am. Chem. Soc. 1975, 97, 4791.
(3) Poliakoff, M. Chem. Soc. ReV. 1978, 7, 527.
(4) Fairhurst, S. A.; Morton, J. R.; Perutz, R. N.; Preston, K. F.
Organometallics 1984, 3, 1389.
(5) Brookhart, M.; Chandler, W.; Kessler, R. J.; Liu, Y.; Pienta, N. J.;
Santini, C. C.; Hall, C.; Perutz, R. N.; Timney, J. A. J. Am. Chem. Soc.
1992, 114, 3802.
(η5-C5H5)Mn(CO)2Kr ≈
(η5-C5Me5)Mn(CO)2Kr < (η5-C5Et5)Mn(CO)2Kr
(6) Mawby, R. J.; Perutz, R. N.; Whittlesey, M. K. Organometallics
1995, 14, 3268.
(7) Whittlesey, M. K.; Perutz, R. N.; Virrels, I. G.; George, M. W.
Organometallics 1997, 16, 268.
(η5-C5H5)Re(CO)2Kr ≈ (η5-C5Me5)Re(CO)2Kr
Activation parameters have been determined for the reaction
of (η5-C5Me5)Re(CO)2Kr with CO (∆H‡ ) 31.0 ( 3.0 kJ
mol-1). This ∆H‡ value represents a lower limit for the Re-Kr
BDE. (η5-C5Et5)Mn(CO)2Kr was found to be approximately 50
times more reactive than (η5-C5Et5)Mn(CO)2Xe. This suggests
that an agostic interaction is not occurring. It appears that there
is no significant effect (electronic or steric) on the reactivity
toward CO for the series of group 7 cyclopentadienyl noble
gas complexes by changing between (η5-C5H5) and (η5-C5Me5).
There is an effect on the reactivity when (η5-C5Me5) is changed
to (η5-C5Et5). This suggests that steric factors are responsible
for this difference.
(8) Kelly, J. M.; Hermann, H.; Koerner von Gustorf, E. J. Chem. Soc.,
Chem. Commun. 1973, 105.
(9) Simon, J. D.; Xie, X. L. J. Phys. Chem. 1986, 90, 6751.
(10) Xie, X. L.; Simon, J. D. J. Am. Chem. Soc. 1990, 112, 1130.
(11) Bengali, A. A.; Bergman, R. G.; Moore, C. B. J. Am. Chem. Soc.
1995, 117, 3879.
(12) Simpson, M. B.; Poliakoff, M.; Turner, J. J.; Maier, W. B., II.;
McLaughlin, J. G. J. Chem. Soc., Chem. Commun. 1983, 1355.
(13) Weiller, B. H. J. Am. Chem. Soc. 1992, 114, 10910.
(14) Weiller, B. H.; Wasserman, E. P.; Moore, C. B.; Bergman, R. G.
J. Am. Chem. Soc. 1993, 115, 4326.
(15) Wells, J. R.; Weitz, E. J. Am. Chem. Soc. 1992, 114, 2783.
(16) Pyykko¨, P. J. Am. Chem. Soc. 1995, 117, 2067.
(17) Schro¨der, D.; Schwarz, H.; Hrusa´k, J.; Pyykko¨, P. Inorg. Chem.
1998, 37, 624.
(18) Sun, X. Z.; George, M. W.; Kazarian, S. G.; Nikiforov, S. M.;
Poliakoff, M. J. Am. Chem. Soc. 1996, 118, 10525.
(19) Sun, X. Z.; Grills, D. C.; Nikiforov, S. M.; Poliakoff, M.; George,
M. W. J. Am. Chem. Soc. 1997, 119, 7521.
(20) Geftakis, S.; Ball, G. E. J. Am. Chem. Soc. 1998, 120, 9953.
(21) Johnson, F. P. A.; George, M. W.; Bagratashvili, V. N.; Veresh-
chagina, L. N.; Poliakoff, M. MendeleeV Commun. 1991, 26.
(22) Creaven, B. S.; George, M. W.; Ginzburg, A. G.; Hughes, C.; Kelly,
J. M.; Long, C.; McGrath, I. M.; Pryce, M. T. Organometallics 1993, 12,
3127.
Conclusions
In this paper we have described the characterization of the
new organometallic noble gas complexes (η5-C5R5)M(CO)2L
(M ) Mn and Re; R ) H, Me and Et (Mn only); L ) Kr and
Xe) in supercritical noble gas solution at room temperature. The
steric bulk of the three types of ring substituent increases steadily
in the order H < Me < Et, and the Me and Et substituents
have almost identical electron directing effects. This has allowed
(23) Morse, J. M., Jr.; Parker, G. H.; Burkey, T. J. Organometallics
1989, 8, 2471.

Reactivity of Organometallic Noble Gas Complexes
J. Phys. Chem. A, Vol. 104, No. 18, 2000 4307
(24) Breheny, C. J.; Kelly, J. M.; Long, C.; O’Keeffe, S.; Pryce, M. T.;
Russell, G.; Walsh, M. M. Organometallics 1998, 17, 3690.
(25) George, M. W.; Poliakoff, M.; Turner, J. J. Analyst 1994, 119, 551.
(26) Poliakoff, M.; Howdle, S. M.; Kazarian, S. G. Angew. Chem., Int.
Ed. Eng. 1995, 34, 1275.
(27) Bernal, I.; Korp, J. D.; Herrmann, W. A.; Serrano, R. Chem. Ber.
1984, 117, 434.
1992, 96, 7650.
(31) Burney, D. P.; Burkey, T. J. Abs. Paps. Am. Chem. Soc. 1994, 207,
19.
(32) Klassen, J. K.; Selke, M.; Sorensen, A. A.; Yang, G. K. J. Am.
Chem. Soc. 1990, 112, 1267.
(33) Burkey, T. J. J. Am. Chem. Soc. 1990, 112, 8329.
(34) Hill, R. H.; Wrighton, M. S. Organomet. 1987, 6, 632.
(35) Bitterwolf, T. E.; Lott, K. A.; Rest, A. J.; Mascetti, J. J. Organomet.
Chem. 1991, 419, 113.
(28) Nesmeyanov, A. N.; Anisimov, K. N.; Valueva, Z. P. IzV. Akad.
Nauk SSSR, Otdel. Khim. Nauk. 1961, 1780.
(29) Rest, A. J.; Sodeau, J. R.; Taylor, D. J. J. Chem. Soc., Dalton Trans.
1978, 651.
(30) Zheng, Y.; Wang, W.; Lin, J.; She, Y.; Fu, K.-J. J. Phys. Chem.
(36) Ball, R. G.; Campen, A. K.; Graham, W. A. G.; Hamley, P. A.;
Kazarian, S. G.; Ollino, M. A.; Poliakoff, M.; Rest, A. J.; Sturgeoff, L.;
Whitwell, I. Inorg. Chim. Acta 1997, 259, 137.