Displacement of the THF solvent molecule from (η5-C5H5)Mn(CO)2THF by simple two electron donor ligands: Evidence for a dissociative mechanism and determination of the Mn-THF bond strength

Article abstract of DOI:10.1016/S0022-328X(98)00918-8

The reaction between CpMn(CO)₂THF (Cp=η⁵-C₅H₅, THF=tetrahydrofuran) and nitrogen containing ligands is studied in THF solution. In all cases the products of the reaction are the known CpMn(CO)₂L complexes (L=piperidine, 4-acetylpyridine). The reaction of the solvated complex with both ligands studied proceeds through a purely dissociative mechanism. In good agreement with previous thermochemical measurements, kinetic analysis yields an average value of 24.0±3.0 kcal mol^-1 for the CpMn(CO)₂-THF bond dissociation energy. The results of the present study clarify the relationship between metal-solvent bond strengths obtained by kinetic methods and those obtained by thermochemical measurements.

Full text of DOI:10.1016/S0022-328X(98)00918-8

Journal of Organometallic Chemistry 572 (1999) 65–71
Displacement of the THF solvent molecule from
(p5-C5H5)Mn(CO)2THF by simple two electron donor ligands: evidence
for a dissociative mechanism and determination of the Mn–THF bond
strength
Jodi E. Coleman, Kim E. Dulaney, Ashfaq A. Bengali *
Department of Chemistry, Dickinson College, Carlisle, PA17013, USA
Received 24 April 1998; received in revised form 15 July 1998
Abstract
The reaction between CpMn(CO)2THF (Cp=p5-C5H5, THF=tetrahydrofuran) and nitrogen containing ligands is studied in
THF solution. In all cases the products of the reaction are the known CpMn(CO)2L complexes (L=piperidine, 4-acetylpyridine).
The reaction of the solvated complex with both ligands studied proceeds through a purely dissociative mechanism. In good
agreement with previous thermochemical measurements, kinetic analysis yields an average value of 24.093.0 kcal mol−1 for the
CpMn(CO)2–THF bond dissociation energy. The results of the present study clarify the relationship between metal–solvent bond
strengths obtained by kinetic methods and those obtained by thermochemical measurements. © 1999 Elsevier Science S.A. All
rights reserved.
Keywords: THF; Dissociative mechanism; Bond strength
1. Introduction
important, therefore, to establish the strength of the
metal–solvent interaction in such molecules. The tech-
nique of photoacoustic calorimetry (PAC) has been
employed successfully to measure metal–solvent bond
strengths [6]. For example, the strength of the interac-
tion between Group 6 metal pentacarbonyls and alkane
solvents such as pentane, cyclohexane, and heptane has
been determined by PAC to be on the order of 8–17
kcal mol−1 [7,8]. Metal–solvent bond strengths have
also been determined using kinetic methods. For exam-
ple, by studying the CO substitution kinetics of
W(CO)5Xe, Weiller was able to provide an estimate of
8.490.2 kcal mol−1 for the strength of the W–Xe
interaction [4]. It should be noted, however, that the
relationship between bond strengths measured by ki-
netic methods and those obtained by thermochemical
studies is at present unclear and further studies are
required.
Coordinatively
unsaturated
transition
metal
organometallic complexes have been implicated as in-
termediates in a variety of chemical reactions including
homogenous and heterogeneous catalytic processes [1].
It is generally well established that in condensed phase,
the vacant site is often occupied by a solvent molecule
[2]. Indeed, even noble gases such as Xe and Kr are
capable of binding to the metal center in such com-
plexes [3–5]. Because solvated organometallic
molecules are often intermediates in chemical reactions,
it is important to elucidate the mechanism of the dis-
placement of the solvent molecule from the coordina-
tion sphere of the metal complex. It is equally
* Corresponding author. Tel.: +1-717-2451975; e-mail: ben-
gali@dickinson.edu.
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(98)00918-8

66
J.E. Coleman et al. / Journal of Organometallic Chemistry 572 (1999) 65–71
2. Experimental
While several mechanistic studies have been per-
formed on the solvated Group 6 metal carbonyls [2]
(M(CO)5solv, M=Cr, Mo, W, and solv=solvent) rel-
atively few studies have been aimed at elucidating the
mechanism by which CpMn(CO)2solv complexes react
with incoming ligands [9,10]. It is well established that
photolysis of CpMn(CO)3 in the gas phase, in inert
gas matrices, and in alkane solvent results in the gen-
eration of the coordinatively unsaturated CpMn(CO)2
species with a high quantum yield [11–13]. In alkane
solvents such as cyclohexane and heptane there is
good evidence that the species formed upon photolysis
is the solvent stabilized CpMn(CO)2solv complex [9].
The study of the ligand substitution mechanism is
especially interesting for CpMn(CO)2solv since the
possibility exists for the involvement of the elusive
ring slipped (p3-Cp)Mn(CO)2solv intermediate in such
reactions. In addition to the general interest in under-
standing the substitution chemistry of these solvated
complexes, studying the displacement of a solvent
molecule from the CpMn(CO)2solv complex is also
important since such complexes are often used as pre-
cursors in the synthesis of CpMn(CO)2L (L=ligand)
molecules. Since thermal substitution of one CO lig-
and in CpMn(CO)3 is not possible [14], synthesis of
CpMn(CO)2L complexes is often accomplished by dis-
placement of the weakly bound THF (THF=
tetrahydrofuran) solvent molecule by L from the pho-
All kinetic studies were conducted using a Hewlett
Packard Diode Array Spectrophotometer (HP 8453)
equipped with a Peltier temperature controller. A typi-
cal kinetic run was performed as follows. A 0.5–1 mM
solution of CpMn(CO)3 in THF was degassed by purg-
ing with pre-dried Ar for approximately 5 min and then
photolyzed for 1 min in a 1×1 cm quartz cuvette using
the filtered output (300 nm long pass filter) of a 150
Watt Xe arc lamp (Xenon Corp.). The ligand was then
added and the reaction followed by monitoring the
decay of the CpMn(CO)2THF complex absorbing at
298 or 530 nm. When 4-acetylpyridine was employed as
the ligand, the reaction was followed by monitoring the
growth of the product CpMn(CO)2(4-acetylpyridine)
complex at 476 nm because the absorption bands of the
CpMn(CO)2THF complex were obscured by those of
the ligand. All runs were performed under pseudo first
order conditions with the ligand concentration being at
least ten times greater than that of the Mn complex.
Kinetic runs were conducted over a 10- to 20-fold range
of ligand concentrations. Observed rate constants, kobs
,
were obtained by fitting plots of the absorbance of
CpMn(CO)2(THF) versus time to single exponential
functions. In all the kinetic runs residual CpMn(CO)3
was also present. However, there was no evidence to
suggest that the parent tricarbonyl participated in the
reaction. Studies employing IR spectroscopy clearly
showed that the height of the CO stretching absorp-
tions of the residual CpMn(CO)3 complex remained
unchanged during the course of the reaction. IR spec-
troscopy, however, was not used to monitor the reac-
tions because the CO absorptions of both the reactant
CpMn(CO)2THF and product CpMn(CO)2L (L=pipe-
ridine, 4-acetylpyridine) overlapped.
tolytically
[15,16].
generated
CpMn(CO)2THF
complex
Tetrahydrofuran was freshly distilled from Na/ben-
zophenone under nitrogen. Piperidine (Aldrich) and
4-acetylpyridine (Acros) were used without further
purification.
We report in this paper a mechanistic study of the
displacement of the THF solvent molecule from
CpMn(CO)2THF by the simple two electron donor
ligands piperidine and 4-acetylpyridine.In both cases
the reaction proceeds through a purely dissociative
pathway involving dissociation of the THF molecule
from the manganese complex followed by reaction of
the resulting CpMn(CO)2 molecule with the incoming
ligand. The kinetic analysis yields an estimate for the
CpMn(CO)2–THF bond strength which is in good
agreement with the value obtained from thermochemi-
cal measurements suggesting that the kinetic method
can provide a useful complement to thermochemical
studies in determining CpMn(CO)2–solv bond
strengths. The results help to clarify the relationship
between metal–solvent bond strengths obtained by ki-
netic methods and those obtained by thermochemical
methods such as photoacoustic calorimetry.
3. Results and discussion
Photolysis of a solution of CpMn(CO)3 in THF
results in the generation of a red solution with ab-
sorbances at 298, 400, and 530 nm (see Fig. 1). These
absorbances are in good agreement with those of the
previously observed CpMn(CO)2THF complex at 400
and 510 nm [10]. At 298 K in the absence of ligand the
CpMn(CO)2THF complex decays over a period of ca. 1
h. As shown in Fig. 2, in the presence of added ligand,
CpMn(CO)2THF exhibits a first order decay and con-
current with the decay of the dicarbonyl a new species
grows in at the same rate. In agreement with previous
studies [10], for L=piperidine the product is observed

J.E. Coleman et al. / Journal of Organometallic Chemistry 572 (1999) 65–71
67
Fig. 1. Difference spectra showing the decay of the CpMn(CO)2THF complex at 298 K in the absence of added ligand. Peaks marked with a ¡
arrow decrease with time and are associated with the CpMn(CO)2THF complex.
at 442 nm and for L=4-acetylpyridine at 476 and 630
nm (see Fig. 3).
(i.e. kobs=k3[L]) then the activation entropy associated
with k3 is quite positive (DS‡=1695 e.u.) and the
activation enthalpy of 2492 kcal mol−1 has a value
equal to the CpMn(CO)2–THF bond dissociation en-
ergy [17]. For an associative mechanism it is expected
that DS‡ will be negative to slightly positive reflecting
the increased order in the transition state [4,18]. Fur-
thermore, since the CpMn(CO)2–THF bond energy has
been determined by photoacoustic calorimetry to be
2492 kcal mol−1 [17], it is expected that if the reaction
proceeded through an associative mechanism DH‡
would be considerably less than this value. Therefore,
the possibility that the reaction has an associative com-
ponent is ruled out (i.e. k3=0).
As shown in Fig. 4a and b, the observed rate con-
stant, kobs, varies linearly with ligand concentration at
all the temperatures studied. A linear dependence of
kobs on [L] is consistent with both a dissociative and
associative mechanism of THF displacement from the
Mn center (Scheme 1). If the reaction occurs through
an associative mechanism (i.e. k1=0) then kobs=k3[L]
and the slope of a plot of kobs versus [L] should yield k3,
the bimolecular rate constant for the reaction. Since the
reaction is studied at different temperatures, an Eyring
plot should yield the activation enthalpy and entropy
for this reaction. The activation parameters obtained
from the Eyring plots are inconsistent, however, with
an associative mechanism of THF displacement by L.
For example, if an associative mechanism is assumed
For a dissociative mechanism, the dependence of kobs
on [L] can be derived by assuming a steady state
concentration of CpMn(CO)2.
k1k2[L]
−1[THF]+k2[L]
kobs
=
(1)
k
Because in the present work [THF]ꢀ[L] and a plot of
kobs versus [L] does not show any deviation from linear-
ity, it is assumed that k−1[THF]ꢀk2[L]. The kobs ex-
pression then reduces to,
k1k2[L]
−1[THF]
kobs
=
(2)
k
and a plot of kobs versus [L]/[THF] yields a slope of
(k1k2)/k−1. Since this reaction was studied at several
temperatures, activation parameters were obtained and
are presented in Table 1 along with values for the rate
constants. The overall enthalpy of activation for this
reaction is composed of the enthalpies of activation for
each step in the reaction mechanism. Thus, from the
Scheme 1.

68
J.E. Coleman et al. / Journal of Organometallic Chemistry 572 (1999) 65–71
Fig. 2. Kinetic traces obtained when CpMn(CO)2(THF) reacts with 0.14 M piperidine at 293 K. Growth of the product was monitored at 442
nm and decay of the CpMn(CO)2(THF) complex was observed at 298 nm. Note that in the case of the product the absorbance values are
multiplied by a factor of five. Solid lines represent single exponential fits to the data.
kcal mol−1 for the CpMn(CO)2–THF bond energy
obtained by thermochemical measurements [17].
slope of the Eyring plot the quantity DH‡1+(DH2‡−
DH−‡ 1) can be determined. Since gas phase studies have
shown that reactions between the unsolvated CpM-
n(CO)2 fragment and various ligands proceed with a
negligible activation barrier [13], the activation en-
thalpies associated with k−1 and k2 are expected to be
insignificant (i.e. DH‡2, DH−‡ 1 ca. 0). Consequently, a
value for DH‡1 can be determined from the Eyring plot
shown in Fig. 5 and this number is a measure of the
CpMn(CO)2–THF bond dissociation energy. Values of
23.392.0 kcal mol−1 for L=piperidine and 24.792
kcal mol−1 for L=4-acetylpyridine (see Table 1) are
consistent with a previously determined value of 2492
Confirmation of the dissociative nature of THF dis-
placement from the Mn center is obtained by compar-
ing the various rate constants and activation enthalpies
for the two ligands studied. It is expected that if the
mechanism is truly dissociative then the activation en-
thalpies will be independent of the nature of the incom-
ing nucleophile. As shown in Fig. 5 and Table 1, the
activation enthalpies are essentially identical for the
two ligands even though there is a significant difference
in the electronic and steric properties of the ligands.
The difference in the rate constants between the ligands
Fig. 3. Difference spectra showing growth of the CpMn(CO)2(pip) complex at 442 nm and decay of the CpMn(CO)2THF complex at 298 and 530
nm.

J.E. Coleman et al. / Journal of Organometallic Chemistry 572 (1999) 65–71
69
Fig. 4. (a) A plot of kobs versus [piperidine]/[THF] at four temperatures for the reaction of CpMn(CO)2(THF) with piperidine in THF solvent.
(b) Plot of kobs versus [4-acetylpyridine]/[THF] for the reaction of the Mn complex with 4-acetylpyridine in THF solvent.
the observed rate constant roughly doubled increasing
from 1.990.2×10−3 to 4.490.4×10−3 −1. While
is attributed to differences in k2, the step involving the
reaction of the unsolvated dicarbonyl with ligand.
Clearly, k2 will be different for 4-acetylpyridine and
piperidine and our results suggest that 4-acetylpyridine
reacts approximately twice as fast with the unsolvated
CpMn(CO)2 fragment than does piperidine.
Further confirmation of the dissociative nature of
the substitution mechanism was obtained by conduct-
ing mixed solvent experiments where the concentration
of the solvent (THF) was varied by addition of cyclo-
hexane, a relatively non-coordinating solvent. Accord-
ing to Eq. (2), a change in the concentration of THF
should affect the observed rate constant (kobs) only if
the mechanism is dissociative. Indeed, at 293 K in the
presence of 0.13 M piperidine, when the concentration
of THF was decreased by a factor of two, the value of
s
these results should be interpreted with some caution
since the nature of the solvent environment is un-
doubtedly changed by the replacement of 50% of the
THF by the non-polar cyclohexane, the results clearly
indicate that the rate of the reaction is a sensitive
function of the solvent characteristics. It is expected
that if the reaction had a significant associative com-
ponent the observed rate would not exhibit such a
strong dependence on the solvent environment.
The close agreement of the kinetic parameters de-
spite the significant electronic and steric differences
between the ligands studied, and the dependence of
the observed rate on the THF concentration, provides
strong evidence for a dissociative mechanism of THF

70
J.E. Coleman et al. / Journal of Organometallic Chemistry 572 (1999) 65–71
Table 1
CpMn(CO)2 complex undergoes very little electronic
and steric reorganization as it reacts with THF. Be-
cause the activation enthalpy for the dissociation of the
Mn–THF bond is not significantly lower than the
Mn–THF bond strength measured by PAC, our results
also imply that in the transition state the CpMn(CO)2
fragment is not stabilized to a significant extent by the
THF solvent molecule. By contrast, the (CO)5W–hep-
tane and (CO)5W–cyclohexane bond strengths deter-
mined by studying the substitution kinetics of the
solvated complexes were approximately 6 kcal mol−1
lower than those obtained by photoacoustic calorimetry
[21]. The difference between the two numbers may be
attributed to partial solvation of the transition state by
the alkane solvent in the kinetic study. In the present
case, however, the excellent agreement between the
CpMn(CO)2–THF bond strength obtained by thermo-
chemical measurements and by the present kinetic
method suggests that the kinetic method can be very
useful in the determination of CpMn(CO)2–solv bond
strengths.
Unlike other solvated organometallic complexes like
(p6-Arene)Cr(CO)2solv and M(CO)5THF (M=Cr, Mo,
W) which react primarily through an interchange or
associative process with donor ligands [22,23], CpM-
n(CO)2THF reacts through a purely dissociative mecha-
nism. This difference in the pathway of solvent
displacement is even more surprising given that in the
case of the manganese complex a ‘ring slip’ of the Cp
ligand from p5 to p3 coordination would allow accom-
modation of the incoming nucleophile without displace-
ment of the solvent molecule. A dissociative pathway of
ligand substitution for the CpMn(CO)2THF complex is
Rate constants obtained for the thermal reaction of CpM-
n(CO)2(THF) with piperidine and 4-acetylpyridinea
Ligand
Solvent
THFc
T (K)
((k1k2)/k−1)×102 (s−1 b
)
4-Acetylpyridine
287
293
298
306
6.890.1
13.690.8
28.090.7
105.695.0
Piperidine
THF
287
293
298
306
3.890.2
10.890.5
18.791.0
53.893.0
a The rate constants were obtained from the slopes of the plots
shown in Fig. 4a,b.
b For ligand=piperidine, DH1‡+(DH2‡−DH−‡ 1)=23.392.0 kcal
mol−1
,
DS‡1+(DS2‡−DS−‡ 1)=+17.795.0 e.u. For ligand=4-
acetylpyridine, DH‡1+(DH2‡−DH−‡ 1)=24.792.0 kcal mol−1, DS1‡+
(DS‡2−DS−‡ 1)=+23.796.0 e.u.
c [THF]=12.3 M.
displacement from the Mn center. A dissociative mech-
anism of solvent displacement from the Mn center is
consistent with previous studies where the substitution
of cis-cyclooctene from CpMn(CO)2(cis-C8H14) by
PPh3 and the displacement of SBu2 from CpM-
n(CO)2(SBu2) by P(OMe)3 proceeded by purely disso-
ciative pathways [19,20].
While the relationship between metal–solvent bond
strengths obtained by kinetic methods and those deter-
mined by thermochemical measurements is not yet well
established, our results do help in clarifying the connec-
tion. The close agreement between the CpMn(CO)2–
THF bond strength obtained from the kinetic and
thermodynamic studies implies that the unsolvated
Fig. 5. An Eyring plot of ln (k%/T) where k%=(k1k2)/k−1 for L=piperidine and 4-acetylpyridine. From the slope of this plot DH‡1 which is
associated with the CpMn(CO)2–THF bond dissociation energy is determined to be 23.392.0 kcal mol−1 for L=piperidine and 24.792.0 kcal
mol−1 for L=4-acetylpyridine (see Table 1).

J.E. Coleman et al. / Journal of Organometallic Chemistry 572 (1999) 65–71
71
consistent with the relative stabilities of the coordina-
tively unsaturated and unsolvated (p6-Arene)Cr(CO)2,
M(CO)5, and CpMn(CO)2 complexes. Gas phase flash
photolysis studies indicate that the CpMn(CO)2 frag-
ment reacts with CO at rates that are roughly an order
of magnitude slower than the corresponding reactions
of the (p6-Arene)Cr(CO)2 and M(CO)5 fragments
[13,24]. Consequently, it is reasonable to expect that the
unsolvated CpMn(CO)2 fragment will have a relatively
longer lifetime in the solution phase compared to the
(p6-Arene)Cr(CO)2 and M(CO)5 fragments. Therefore,
relative to the (p6-Arene)Cr(CO)2 and M(CO)5 frag-
ments displacement of a solvent molecule from the
CpMn(CO)2solv complex is more likely to proceed
through a dissociative pathway than an associative one.
College. We would also like to thank the Whitaker
Foundation for a summer student-faculty research
grant administered by Dickinson College.
References
[1] J.P. Collman, L.S. Hedegus, J.R. Norton, P.G. Finke, Principles
and Applications of Organotransition Metal Chemistry, 2nd ed.,
University Science Books, Mill Valley, CA 1987.
[2] C. Hall, R.N. Perutz, Chem Rev. 96 (1996) 3125.
[3] R.N. Perutz, J.J. Turner, J. Am. Chem. Soc. 97 (1975) 4791.
[4] B.H. Weiller, J. Am. Chem. Soc. 114 (1992) 10910.
[5] R.H. Schultz, A.A. Bengali, M.J. Tauber, B.H. Weiller, E.P.
Wasserman, K.R. Kyle, C.B. Moore, R.G. Bergman, J. Am.
Chem. Soc. 116 (1994) 7369.
[6] J.M. Morse, G.H. Parker, T.J. Burkey, Organometallics 8 (1989)
2471.
[7] G.-L. Leu, T.J. Burkey, J. Coord. Chem. 34 (1995) 87.
[8] G.K. Yang, V. Vaida, K.S. Peters, Polyhedron 7 (1988) 1619.
[9] B.S. Creaven, A.J. Dixon, J.M. Kelly, C. Long, M. Poliakoff,
Organometallics 6 (1987) 2600.
[10] R.S. Hill, M.S. Wrighton, Organometallics 6 (1987) 632.
[11] W.V. Strohmeier, K.Z. Gerlach, Naturforsch., B: Anorg. Chem.,
Org. Chem., Biochem., Biophys. Biol. 15B (1960) 675.
[12] A.J. Rest, J.R. Sodeau, D.J. Taylor, J. Chem. Soc., Dalton
Trans. (1978) 651.
4. Conclusion
The reaction between CpMn(CO)2(THF) and simple
two electron donor ligands L (L=piperidine, 4-
acetylpyridine) proceeds through a dissociative mecha-
nism resulting in the formation of the corresponding
CpMn(CO)2L complex. The kinetic analysis yields an
average value of 24.093.0 kcal mol−1 for the CpM-
n(CO)2–THF bond energy in good agreement with a
value of 2492 kcal mol−1 obtained from previous
thermochemical studies. The results of the present study
provide an important comparison between the activa-
tion enthalpies of solvent displacement from a metal
center and metal–solvent bond strengths obtained by
thermochemical methods.
[13] Y. Zheng, W. Wang, J. Lin, Y. She, K.-J. Fu, J. Phys. Chem. 96
(1992) 7650.
[14] I. Wender, P. Pino (Eds), Organic Synthesis via Metal Car-
bonyls, vol. 1, Wiley Interscience, New York, 1968.
[15] W.A. Herrmann, Chem. Ber. 108 (1975) 486.
[16] W.A. Herrmann, G. Kriechbaum, M.L. Ziegler, P. Wu¨lknitz,
Chem. Ber. 114 (1981) 276.
[17] P.-F. Yang, G.K. Yang, J. Am. Chem. Soc. 114 (1992) 6937.
[18] Atwood, J.D., Inorganic and Organometallic Reaction Mecha-
nisms; Brooks/Cole, Monterey, CA, 1985.
[19] R.J. Angelici, W. Loewen, Inorg. Chem. 6 (1967) 682.
[20] I.S. Butler, T. Sawai, Inorg. Chem. 14 (1975) 2703.
[21] G.R. Dobson, K.J. Asali, C.D. Cate, C.W. Cate, Inorg. Chem.
30 (1991) 4471.
Acknowledgements
[22] S. Wieland, R. van Eldik, Organometallics 10 (1991) 3110.
[23] B.S. Creaven, M.W. George, A.G. Ginzburg, C. Hughes, J.M.
Kelly, C. Long, I.M. McGrath, M.T. Pryce, Organometallics 12
(1993) 3127.
[24] Y. Zheng, W. Wang, J. Lin, Y. She, K.-J. Fu, J. Phys. Chem. 96
(1992) 9821.
This research was supported by grants from the
Camille and Henry Dreyfus Foundation (Award No.
SU-95-047), Research Corporation (Grant No.
CC4352), and start-up funds provided by Dickinson
.