Ligand substitution kinetics in M(CO)4(η2:2-norbornadiene) complexes (M=Cr, Mo, W): Displacement of norbornadiene by bis(diphenylphosphino)alkanes

Article abstract of DOI:10.1016/S0022-328X(99)00458-1

The thermal substitution kinetics of norbornadiene (NBD) by bis(diphenylphosphino)alkanes (PP), (C₆H₅)₂P(CH₂)_nP(C ₆H₅)₂ (n=1, 2, 3) in M(CO)₄(η^2:2-NBD) complexes (M=Cr, Mo, W), were studied by quantitative FT-IR spectroscopy. The reaction rate exhibits first-order dependence on the concentration of the starting complex, and the observed rate constant depends on the concentration of the leaving NBD ligand and on the concentration and the nature of the entering PP ligand. In the proposed mechanism there are two competing initial steps: an associative reaction involving the attachment of the entering PP ligand to the transition metal center and a dissociative reaction involving the stepwise detachment of the diolefin ligand from the transition metal center. A rate law is derived from the proposed mechanism. The activation parameters are obtained from the evaluation of the kinetic data. It is found that at higher concentrations of the entering ligand, the associative path is dominant, while at lower concentrations the contribution of the dissociative path becomes significant. Both the observed rate constant and the activation parameters show noticeable variation with the chain length of the diphosphine ligand.

Full text of DOI:10.1016/S0022-328X(99)00458-1

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 590 (1999) 208–216
Ligand substitution kinetics in M(CO)4(h2:2-norbornadiene)
complexes (M=Cr, Mo, W): displacement of norbornadiene by
bis(diphenylphosphino)alkanes
8
Ays¸in Tekkaya, Saim Ozkar *
Department of Chemistry, Middle East Technical Uni6ersity, 06531 Ankara, Turkey
Received 8 June 1999; accepted 29 July 1999
Abstract
The thermal substitution kinetics of norbornadiene (NBD) by bis(diphenylphosphino)alkanes (PP), (C6H5)2P(CH2)nP(C6H5)2
(n=1, 2, 3) in M(CO)4(h2:2-NBD) complexes (M=Cr, Mo, W), were studied by quantitative FT-IR spectroscopy. The reaction
rate exhibits first-order dependence on the concentration of the starting complex, and the observed rate constant depends on the
concentration of the leaving NBD ligand and on the concentration and the nature of the entering PP ligand. In the proposed
mechanism there are two competing initial steps: an associative reaction involving the attachment of the entering PP ligand to the
transition metal center and a dissociative reaction involving the stepwise detachment of the diolefin ligand from the transition
metal center. A rate law is derived from the proposed mechanism. The activation parameters are obtained from the evaluation
of the kinetic data. It is found that at higher concentrations of the entering ligand, the associative path is dominant, while at lower
concentrations the contribution of the dissociative path becomes significant. Both the observed rate constant and the activation
parameters show noticeable variation with the chain length of the diphosphine ligand. © 1999 Elsevier Science S.A. All rights
reserved.
Keywords: Substitution; Diphosphinoalkanes; Norbornadiene; Chromium; Molybdenum; Tungsten
concentrations of both leaving COD and entering PP
ligands. Both the observed rate constant and the activa-
tion parameters show little variation with the chain
length of the diphosphine ligand, but significant
changes are observed as the central metal atom is
varied. Here, we report our extended study on the
displacement kinetics of norbornadiene, NBD, from
M(CO)4(h2:2-NBD) (M=Cr (1A), Mo (2A), W (3A))
by bis(diphenylphosphino)alkanes (PP), (C6H5)2-
P(CH2)nP(C6H5)2 (n=1 (B), 2 (C), 3 (D)) with the aim
to investigate the dependence of the reaction kinetics
and the mechanism on the nature of the leaving diene
ligand.
1. Introduction
The thermal substitution kinetics of 1,5-cyclooctadi-
ene (COD) by bis(diphenylphosphino)alkanes (PP),
(C6H5)2P(CH2)nP(C6H5)2 (n=1, 2, 3) in M(CO)4(h2:2
-
COD) complexes (M=Cr, Mo, W), were recently re-
ported [1]. It was found that the weakly bound COD
ligand is replaced by PP at an observable rate in the
temperature range 35–90°C to form the known com-
plexes [M(CO)4(PP)] as the final product [2]. The kinet-
ics of this thermal substitution reaction were studied by
quantitative FT-IR spectroscopy. A rate law was
derived from the proposed mechanism in which the rate
determining step is the cleavage of one of two
metalꢀolefin bonds. The reaction rate exhibits first-or-
D
M(CO)4(h2:2-NBD)+ PP +NBD“M(CO)4(PP)
1Aꢀ3A
BꢀD
1Bꢀ3D
der dependence on the concentration of [M(CO)4(h2:2
COD)], and the observed rate constant depends on the
-
M=Cr, Mo, W; PP=(C6H5)2P(CH2)nP(C6H5)2;
n=1, 2, 3
Quantitative FT-IR spectroscopy was used to study
the kinetics of these ligand displacement reactions.
* Corresponding author. Fax: +90-312-2101280.
8
E-mail address: sozkar@metu.edu.tr (S. Ozkar)
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00458-1

8
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
209
Monitoring the intensities of the absorption bands of
the CO stretching vibrations for both the reactant and
product enables one to determine the changes in their
concentrations, thus to follow the reaction precisely,
and to check the material balance and the yield
throughout the reaction.
material precipitated is removed by filtration. The re-
maining orange solution is allowed to stand at −35°C
for 1 day for crystallization. The orange crystals are
dried in vacuum.
2.2. Tetracarbonyl-bis(diphenylphosphino)alkane-
metal(0)
2. Experimental
Equimolar amounts (ca. 1 mmol) of M(CO)4(h2:2
-
NBD) (1A–3A) and PPꢁPh2P(CH2)nPPh2 (n=1, 2, 3)
(B–D) were dissolved in toluene (50 ml) and refluxed
for several minutes. Cooling the solution to −35°C
yields pale yellow crystals of M(CO)4(PP) (1B–3D)
which are separated from the supernatant liquid by
decantation and dried under vacuum and identified by
13C- and 31P-NMR spectroscopy [6].
All of reactions and manipulations of the complexes
were carried out either in vacuum or under a dry and
oxygen free nitrogen atmosphere. Solvents were dis-
tilled after refluxing over metallic sodium or phospho-
rous pentoxide for 3–4 days and stored under nitrogen
until used. Hexacarbonylchromium(0), hexacarbonyl-
molybdenum(0), hexacarbonyltungsten(0), norbornadi-
ene (A), bis(diphenylphosphino)methane (B), 1,2-bis-
(diphenylphosphino)ethane (C), and 1,3-bis(diphenyl-
phosphino)propane (D) were purchased from Aldrich,
Dorset, UK. The thermal reactions and other treat-
ments of organometallic compounds such as purifica-
tion and crystallization were followed by taking IR
spectra at appropriate time intervals. NMR spectra
were recorded on a Bruker DPX 400 spectrometer
(400.132 MHz for 1H, 161.996 MHz for 31P, and
100.613 MHz for 13C). TMS was used as internal
2.3. Kinetic measurements
All kinetic measurements were conducted on toluene
solutions in a Specac Variable Temperature IR-Liquid
Cell with a 0.129 mm path length and calcium fluoride
windows. Extinction coefficients of M(CO)4(h2:2-NBD)
(1A–3A) and M(CO)4(PP) (1B–3D) were determined
from toluene solutions of the pure complexes by plot-
ting the absorbance of the wCO IR band with the
highest frequency versus the concentration of the re-
spective complex in the range 5×10−4 to 2×10−3
mol l−1 (Table 1). Using these values, the concentra-
tions of the complexes during the reactions of 1A–3A
with PP could be determined from the measured IR
absorbances at any stage of the reactions. Thus the
material balance could also be checked at any point of
conversion.
1
reference for H- and 13C-NMR chemical shifts. Phos-
phoric acid (85%) in a capillary tube was used as
external reference for 31P-NMR chemical shifts. In-
frared spectra were recorded from solutions on a
Perkin–Elmer 16 PC FT-IR spectrometer. Photochemi-
cal reactions were carried out in an immersion-well
apparatus [3] (solidex glass, u\280 nm) by using a
Hanau TQ 150 high-pressure mercury lamp, which was
cooled by circulating water or cold methanol. A circu-
lating thermostat bath (Heto CB 11e) was used to
provide a constant temperature during the kinetic mea-
surements. Ethylene glycol was circulated from the
thermostat through the heating jacket of the reaction
vessel.
The M(CO)4(h2:2-NBD) complexes were added to the
solutions of the diphosphine in toluene at low tempera-
ture to avoid any reaction. The initial concentration
[C0] of the complexes was around 0.01 mol l−1. A
sample of this solution was transferred into the Specac
Variable Temperature IR-Liquid Cell P/N 21500, which
is controlled by an Automatic Temperature Controller
P/N 20120, preheated to the reaction temperature. The
IR spectra were recorded periodically as the reaction
proceeded at constant temperature. The rates of the
reaction were determined by following the disappear-
ance of the highest-frequency wCO band of the educt,
since it was the only distinct band that does not overlap
with the bands of the products. In this way, quantita-
tive data on the rates of reactions were collected easily
and accurately. The graphical evaluation of the data
provides the observed rate constants. In order to study
the dependence of the rate on the leaving NBD and
entering PP, the kinetic experiments were performed by
varying the concentrations of PP (0.01–1.4 mol l−1) or
NBD (0.0–0.1 mol l−1) in the solutions of 1A–3A
(C0=0.009 mol l−1) at 40°C (M=Cr), 45°C (M=
2.1. Tetracarbonyl(p2:2-norbornadiene)metal(0)
Tetracarbonyl(h2:2-norbornadiene)chromium(0), Cr-
(CO)4(h2:2-NBD) (1A), and tetracarbonyl(h2:2-norbor-
nadiene)molybdenum(0), Mo(CO)4(h2:2-NBD) (2A),
were prepared by refluxing a solution of hexacarbonyl-
metal(0) and NBD in isooctane as described in the
literature [4]. The reaction solution is evaporated to
dryness. Crystallization from n-hexane solution yields
yellow crystals of the complexes. Tetracarbonyl(h2:2
-
norbornadiene)tungsten(0), W(CO)4(h2:2-NBD) (3A),
was prepared by refluxing W(pip)2(CO)4 and NBD in
n-hexane as described in the literature [5]. The reaction
mixture is cooled down to 0°C and the polymeric

8
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
210
Table 1
The CO stretching frequencies (cm−1) for the complexes and molar extinction coefficient (l mol−1 cm−1) for the highest-frequency band in toluene
Complex
m×10−3 (l mol−1 cm−1) for A1(2)
A1(2)
A1(1)
B1
B2
Cr(CO)4(h2:2-NBD) (1A)
Mo(CO)4(h2:2-NBD) (2A)
W(CO)4(h2:2-NBD) (3A)
Cr(CO)4(DPPM) (1B)
Mo(CO)4(DPPM) (2B)
W(CO)4(DPPM) (3B)
Cr(CO)4(DPPE) (1C)
Mo(CO)4(DPPE) (2C)
W(CO)4(DPPE) (3C)
Cr(CO)4(DPPP) (1D)
Mo(CO)4(DPPP) (2D)
W(CO)4(DPPP) (3D)
3.4
2.4
1.9
3.3
3.5
4.5
3.0
3.1
1.3
4.5
4.2
1.9
2026
2039
2038
2012
2012
2012
2020
2017
2023
2015
2015
2016
1935
1948
1948
1945
1903
1901
1902
1917
1908
1908
1907
1899
1896
1893
1891
1890
1889
1897
1897
1922
1930
1923
1929
1928
1936
1935
1930
1933
1890
1883
1888
1901
Mo), and 90°C (M=W). The activation parameters
were determined from the temperature dependence of
the observed rate constant in the range 5–90°C.
The substitution reaction of 2A with DPPM in ten-
fold excess at 20°C was repeated ten times under identi-
cal conditions to check the reproducibility and to
calculate the indeterminate errors for the kinetic data.
The error in the observed rate constant values was
found to be less than 92%.
the highest-frequency bands of the reactant and
product remain well resolved during the whole reaction.
These two IR bands are selected to follow the con-
sumption of the reactant and the growth of the
product, respectively.
The concentration versus time graph for the course
of the thermal substitution reaction shows an exponen-
tial decay for the starting complex and an exponential
growth for the product (Fig. 2(a)). The logarithmic plot
of the concentration of the reactant against time gives a
straight line for the aforementioned reaction even at
low concentrations of bis(diphenylphosphino)-alkane
(Fig. 2(b)). This indicates that the displacement of
NBD from 1A–3A) by PP obeys pseudo-first-order
kinetics over at least four half-lives. The slope of the
straight line gives the observed rate constant, kobs [s−1],
for the pseudo-first-order thermal substitution reaction.
The kinetic data for the NBD substitution in
M(CO)4(h2:2-NBD) by the diphosphinoalkanes can be
interpreted by considering a multi-step mechanism sim-
3. Results and discussion
The IR absorption spectra of both the M(CO)4(h2:2
-
NBD) and M(CO)4(PP) complexes taken in toluene
show four absorption bands in the CO stretching region
indicating a cis arrangement of four CO groups in the
pseudo-octahedral coordination sphere of the metal
(Table 1). Thus, the M(CO)4 moiety in the complexes
has a local C26 symmetry with 2A1+B1+B2 CO
stretching modes [7]. The complexes 2A, 3A, 3C, and
2D give only three IR CO stretching bands indicating
that the two of the four expected bands are accidentally
degenerate.
The kinetics of the thermal substitution of NBD
from 1A–3A by bis(diphenyl-phosphino)alkanes were
followed by using the quantitative FT-IR spectroscopy.
In the wCO region of the IR spectrum, the absorption
bands of 1A–3A are gradually replaced by the new
bands of the corresponding product M(CO)4(PP) in the
course of the displacement reaction. The gradual
change in the IR spectrum during the reaction of
Mo(CO)4(h2:2-NBD) (2A) with DPPM at 20°C is de-
picted in Fig. 1 as an illustrative example. The observa-
tion of narrow isosbestic points indicates
a
straightforward conversion of the reactant into the
product without side or subsequent reactions [8]. In
other words, there are only two absorbing species
throughout the reaction. Another point emerging from
the inspection of the IR spectra at a first glance is that
Fig. 1. The FT-IR spectra taken in the course of the thermal
substitution reaction of 2A with DPPM at 20°C with a time interval
of ca. 20 min. [DPPM]0/[2A]0=10.

8
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
211
Heating solutions of the starting complexes 1A–3A in
the presence of equimolar amounts of PP under a CO
atmosphere at the same temperature gives rise to the
predominant formation of M(CO)4(PP), together with
M(CO)5(PP) and M(CO)6 in low concentrations (less
than 15 and 5%, respectively). The observation of pen-
tacarbonyl-diphosphinechromium(0) [9] in the latter ex-
periment is a compelling evidence for the formation of
intermediates in which the diphosphine is bound to the
metal in monodentate fashion. It indicates further that
the CO detachment does not occur as an initial step.
Working in the presence of ten-fold excess of PP,
M(CO)4(PP) is the sole product.
Thermal substitution reaction of 1A with DPPE
(one-fold) was also performed in the presence of one-
or two-fold P(OCH3)3 at 85°C. The following five
complexes could be detected in the reaction mixture:
cis-[Cr(CO)4(P(OCH3)3)2], trans-[Cr(CO)4(P(OCH3)3)2],
[Cr(CO)4(DPPE)] (1C), cis-[Cr(CO)4(P(OCH3)3)(DP-
PE)], and trans-[Cr(CO)4(P(OCH3)3)(DPPE)]. The most
distinguishable CO-stretching frequencies for these
complexes were 2023 cm−1 for cis-[Cr(CO)4(P-
(OCH3)3)2], 1912 cm−1 for trans-[Cr(CO)4(P(OCH3)3)2],
2016 cm−1 for cis-[Cr(CO)4(P(OCH3)3)(DPPE)], 1905
cm−1 for trans-[Cr(CO)4(P(OCH3)3)(DPPE)], and 2020
cm−1 for 1C [10,11]. The same reaction was performed
by using 3A instead of 1A in order to reduce the
number of the products as tungsten prefers only cis-iso-
mers. Substitution reaction of 3A with DPPE (one-fold)
was performed in the presence of one- or two-fold
P(OCH3)3 at 65°C. The IR spectrum at the end of the
reaction shows the presence of only three different
products. These are cis-[W(CO)4(P(OCH3)3)2], cis-
[W(CO)4(P(OCH3)3)(DPPE)], and [W(CO)4(DPPE)]
(3C). The most distinguishable CO-stretching frequen-
cies for these complexes were 2032 cm−1 for cis-
[W(CO)4(P(OCH3)3)2], 2026 cm−1 for cis-[W(CO)4(P-
(OCH3)3)(DPPE)], and 2023 cm−1 for 3C [10,11]. Al-
though the ligand exchange reactions of cis-[W(CO)4L2]
complexes with L% (both L and L% are monodentate
phosphine ligands) are known at high temperatures
[12], no detectable change was observed upon heating
the solution of [W(CO)4(DPPE)] and P(OCH3)3 to
65°C. Thus, the formation of cis-[Cr(CO)4(P-
(OCH3)3)(DPPE)], trans-[Cr(CO)4(P(OCH3)3)(DPPE)],
and cis-[W(CO)4(P(OCH3)3)(DPPE)] in the aforemen-
tioned experiments indicates a stepwise displacement of
NBD in the thermal substitution reaction of
M(CO)4(h2:2-NBD) instead of a CO detachment.
Fig. 2. (a) The normalized concentration vs. time plot for the
substitution reaction of 2A (ꢀ) with DPPM at 15°C forming 2B (ꢁ),
[PP]0/[2B]0=10; (b) plot of the first-order kinetics for the same
reaction.
ilar to the one proposed for the COD substitution in
the M(CO)4(h2:2-COD) complexes by PP [1]. The reac-
tion mechanism is complicated, because the thermal
reaction is a ligand exchange reaction, where one has to
consider the effect of both the bidentate leaving and
bidentate entering ligands on the reaction mechanism.
Furthermore, additional experiments are required to
collect evidence for the intermediates in the proposed
mechanism. Therefore, in what follows, all the kinetic
data and the other experimental evidence will be dis-
cussed before giving the proposed mechanism, which
will then be shown to be in agreement with the experi-
mental observations.
A series of experiments was performed to check
whether a CO detachment would occur as an initial
step and to obtain evidence for the formation of possi-
ble intermediates. First of all, none of the three starting
complexes 1A–3A gives any reaction in the absence of
entering ligand upon heating to the temperature at
which the ligand substitution takes place at an appre-
ciable rate. Consequently, one can assume that the
observed rate constant (kobs) is zero at zero concentra-
tion of entering ligand (DPPM, DPPE or DPPP). In a
similar reaction under 1 atm CO at the same tempera-
ture, the starting complex is completely converted into
the corresponding hexacarbonylmetal(0) within 2–4 h.
The concentration of the entering ligand (DPPM,
DPPE, and DPPP) plays an important role in the rate
of the reaction via the observed rate constant (Table 2,
Fig. 3). kobs gradually increases with the increasing
concentration of the entering ligand. At high concentra-
tions of the entering ligand, the increase in the observed

8
212
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
Table 2
Table 3
The observed rate constants (kobs) at various concentrations of enter-
ing ligand (DPPM, DPPE and DPPP) for the substitution of norbor-
nadiene in 1A at 85°C a
The observed rate constants (kobs) at various concentrations of leav-
ing ligand (NBD) for the substitution of norbornadiene in 1A at
85°C a
[PP]/[1A]
k
obs×104 (s−1
)
[NBD]/[1A]
k )
obs×104 (s−1
DPPM
DPPE
DPPP
0
1
2
4
6
2.97
2.18
1.97
1.70
1.70
1.68
10
20
40
60
100
150
3.60
8.12
13.8
16.6
22.0
–
6.38
13.8
25.7
41.0
54.5
65.8
11.8
25.5
39.2
45.2
63.9
69.8
10
a [DPPE]0=[1A]0=9.00 mM.
a [1A]0=9.00 mM.
complexes by diphosphines [1]. The increasing order of
the observed rate constant, DPPMBDPPEBDPPP,
nicely correlates with the basicity of the chelating
diphosphines [14] and the metalꢀdiphosphine bond
strength [15].
rate constant is retarded as if it were approaching a
saturation limit. However, it has been realized that this
is due to the limited solubility of the bis(diphenylphos-
phino)alkane in toluene even at high temperature of the
measurement. Inspection of Fig. 3 shows further that
the observed rate constant, kobs, depends not only on
the concentration, but also on the nature of the enter-
ing ligand. This observation implies that the substitu-
tion reaction is associative in nature [13]. This is one of
the experimental results implying that an associative
first step should be incorporated into the mechanism,
proposed for the COD substitution in the same kind of
The effect of the leaving group on the rate of the
thermal substitution reaction was also studied by vary-
ing the concentration of NBD but keeping the [DPPE]/
[1A] ratio constant at one (Table 3). The plot of kobs
versus the concentration of NBD (Fig. 4) is curved
downward with a positive intercept, indicating the pres-
ence of a complex mechanism in which increasing con-
centration of NBD causes a decrease in kobs [16]. This
clearly suggests that the substitution rate of NBD is
retarded by the presence of NBD in the solution. In
other words, the starting material, 1A, is stabilized by
the presence of excess NBD. This observation is impor-
tant in proposing a mechanism for the thermal ligand
substitution reaction. As the concentration of NBD
goes to infinity, kobs approaches a limiting value, which
can be obtained from Fig. 4.
At this point one can give the proposed mechanism
that will be in agreement with the experimental results
for the thermal NBD substitution in M(CO)4(h2:2
-
NBD) by PP (Scheme 1). The mechanism essentially
resembles the one proposed for the stepwise displace-
ment of COD from M(CO)4(h2:2-COD) by PP [1],
where the initial, rate-determining step is the cleavage
of one of the metalꢀolefin bonds [17] and all the
possible steps are considered in the substitution reac-
tion. The only difference between the two mechanisms
is the incorporation of an associative initial step into
the mechanism proposed for the thermal NBD substitu-
tion in M(CO)4(h2:2-NBD) by PP (Scheme 1). Thus,
there are two initial steps, which are competing with
each other. One is an associative reaction involving the
attachment of entering PP ligand to the transition metal
center. The other is a dissociative reaction involving the
stepwise detachment of the diolefin ligand. In the disso-
ciative path, the intermediate I may produce the start-
ing complex by reattachment of the free CꢁC bond of
NBD, or bind irreversibly a PP ligand, yielding the
Fig. 3. Variations in the observed rate constants (kobs) with the
concentration of entering ligand for the substitution of norbornadiene
in 1A at 85°C. [1A]0=9 mM.

8
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
213
Fig. 4. Variations in the observed rate constants (kobs) with the concentration of NBD for the substitution of NBD in 1A with DPPE at 85°C.
[1A]0=[DPPE]0=9.00 Mm.
intermediate III in which both of the potentially biden-
tate ligands (NBD and PP) are bonded to the transition
metal center in monodentate fashion. Furthermore,
complete detachment of NBD from I can take place to
yield a 14 electron species (II), which is very unstable
and reacts very fast with PP to form the complexes
1B–3D. The intermediate III will also produce the
complexes 1B–3D upon detachment of NBD, followed
by the fast ring-closure of the PP ligand.
The proposed mechanism (Scheme 1) comprises
a series of steps and therefore involves several inter-
mediates. The observation of Cr(CO)5(DPPE), cis-
and trans-[Cr(CO)4(P(OCH3)3)(DPPE)], and cis-
[W(CO)4(P(OCH3)3)(DPPE)] as products of the thermal
substitution of NBD from M(CO)4(h2:2-NBD) by PP in
the presence of P(OCH3)3 provides a compelling evi-
dence for the formation of III as intermediate in the
reaction mechanism given in Scheme 1. Please recall
that the results of our experiments under CO atmo-
sphere and in the presence of P(OCH3)3 indicate a
stepwise displacement of NBD in the thermal substitu-
tion reaction of M(CO)4(h2:2-NBD) instead of CO de-
tachment. The intermediate I is known from the flash
photolysis and the low-temperature matrix isolation
studies in the literature [18]. There is no evidence for
the formation of cis-M(CO)4(h2-NBD)2 from the inter-
mediate I upon reversible coordination of a second
NBD molecule, even though similar complexes have
already been isolated [19]. However, such an intermedi-
ate would not affect the rate expression and is not
shown in the Scheme 1.
RATE = kobs[M]
(1)
kobs = k1 + k2[PP]
k1k−1
k−1 + k4 + k3[PP] −
−
k4
k6[PP]
−4[NBD]
1 +
k
(2)
where [M] denotes the concentration of the starting
M(CO)4(h2:2-NBD) complex. It is obvious that at very
high concentrations of the entering ligand, the observed
rate constant will be converted to a two-term rate
By applying the well-known steady-state approxima-
tion for intermediates I, and II, the following rate law
can be derived for the thermal NBD substitution in
M(CO)4(h2:2-NBD) by PP:
Scheme 1.

8
214
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
Table 4
Table 6
The observed rate constants obtained for the substitution reaction of
The observed rate constants obtained at different temperatures for
starting complexes 1A–3A at different temperatures a
1A a
Starting complex
T (K)
k
obs×104 (s−1
)
T (K)
k )
obs×104 (s−1
DPPM
DPPE
DPPP
348
353
358
363
1.13
1.87
2.97
4.58
1A
2A
3A
348
353
358
363
278
283
289
293
323
328
333
338
2.00
3.09
3.81
5.33
4.94
7.76
14.2
18.3
1.91
2.16
2.88
4.24
3.67
5.73
6.38
9.29
11.6
17.8
28.9
37.4
3.77
5.58
7.10
10.4
6.60
8.10
11.8
13.2
14.8
23.8
32.2
45.9
5.68
8.17
9.45
11.5
a [DPPE]0/[1A]0=1, [1A]0=9.00 mM.
of kobs to get at least some clue about the nature of
mechanism. For this purpose, thermal substitution re-
actions were performed at 75, 80, 85, and 90°C for 1A,
5, 10, 15, and 20°C for 2A, and 50, 55, 60, and 65°C for
3A in the presence of ten-fold excess of diphosphine
ligand to provide pseudo-first-order conditions. The
observed rate constants obtained for the substitution of
NBD in 1A–3A with three different entering ligands at
the given temperatures are listed in Table 4. The en-
thalpy of activation, DH‡, and the entropy of activa-
tion, DS‡, are calculated from these data by using the
Eyring equation and given in Table 5. The large nega-
tive values of the entropy of activation for all the
substitution reactions imply an associative mechanism
in the transition state [21]. If some associated species
were a reactive form of the complex, an equilibrium
between the free complex and this species would result
in a ligand dependence of the rate constants at low
ligand concentrations [22]. In other words, the depen-
dence of the rate on the ligand type and concentration
can be easily explained by a displacement mechanism,
which proceeds via the formation of a seven coordi-
nated complex in the transition state [23].
a [PP]0/[M]0=10.
constant, i.e. lim[PP]“ꢀkobs=k1+k2[PP]. Although
such an approximation would simplify the system enor-
mously, the limited solubility of diphosphine in toluene
does not allow very high PP concentrations. The rate
expression given above can be used to explain all the
experimental results at the available concentrations of
diphosphine.
In a multistage reaction the rate-limiting step is that
for which the rate of conversion of molecules is the
lowest; on this definition it is not necessarily the step
with the smallest rate constant. Rate-determining step is
often used synonymously, but it is a little misleading,
because in general, the rate of a multistage reaction is a
composite of the rates of all the individual steps and is
not determined by any single one of them [20]. In other
words, more than one step may limit the reaction rate.
Activation parameters are the basis for the determina-
tion of the nature of a mechanism [13]. However, the
complexity of the mechanism proposed for the displace-
ment reactions of 1A–3A with the entering ligands,
unfortunately, does not allow us to determine the rate
constants for the individual steps. Although interpreta-
tion of activation parameters for such rate regimes is
always open to debate, we determined the activation
enthalpy and entropy from the temperature dependence
The NBD substitution in 1A was carried out at four
different temperatures in the presence of one-fold
DPPE for a better appreciation of the effect of entering
ligand concentration on the activation parameters.
From the observed rate constants at these temperatures
(Table 6) the activation enthalpy and activation entropy
values were found to be DH‡=95 kJ mol−1 and
DS‡= −49 J mol−1
K
−1, respectively. Comparing to
the values in Table 5 (DH‡=63 kJ mol−1 and DS‡=
−135 J mol−1 −1), one observes a decrease in both
the activation enthalpy and entropy when the entering
K
Table 5
The activation enthalpy and the activation entropy values for the substitution of NBD in 1A–3A a
DH‡ (kJ mol−1
)
DS‡ (J mol−1 K−1
)
1A
2A
3A
1A
2A
−9894
−11794
−13195
3A
DPPM
DPPE
DPPP
6392
5892
4992
5892
5192
4792
4692
5792
3892
−13595
−14696
−16796
−17696
−13595
−18896
a [PP]0/[M]0=10, [M]0=9.00 mM.

8
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
215
ligand concentration was increased ten times. This re-
sult indicates that the contribution of the dissociative
path becomes noticeable at lower concentrations of
entering ligand. This is also indicative of the competi-
tion between the associative and dissociative paths in
the proposed mechanism (Scheme 1). At higher concen-
tration of entering ligand, the associative path is domi-
nant, while at lower concentrations the contribution of
the dissociative path becomes significant.
compensation temperature. By using the data in Table
5, DH‡ versus DS‡ plots are drawn for the substitution
reaction of 1A–3A with DPPM, DPPE or DPPP (Fig.
5). The slopes of the lines give Tc values of 447, 359,
and 328 K for the respective reaction series with a
correlation coefficient 0.998, 0.997, and 0.974,
respectively.
Taking together all the experimental data on the
substitution of NBD in M(CO)4(h2:2-NBD) given above
and the previously reported results on the substitution
of COD in M(CO)4(h2:2-COD) with diphosphinoal-
kanes [1], one can understand the effect of the leaving
bidentate ligand on the reaction kinetics and mecha-
nism. From the mechanistic standpoint, substitution of
the ligands NBD in M(CO)4(h2:2-NBD) (M=Cr, Mo,
W) and COD in M(CO)4(h2:2-COD) (M=Mo and W)
(1–2) has associative nature in the transition states,
whereas substitution of COD in Cr(CO)4(h2:2-COD)
has a dissociative character. Reason for this behavior
can be sterical crowding around the small chromium
atom. As a bidentate ligand, 1,5-cyclooctadiene requires
larger coordination volume than norbornadiene, a
point, which can be figured out from the structural data
of the complexes of both ligand [26,28]. The term
coordination volume comprises the bite angle as well as
the space required by the ligand in the coordination
sphere of the metal. When diphosphine is substituted in
place of COD in Cr(CO)4(h2:2-COD) it will be of
dissociative nature in transition states. From the kinetic
point of view, one can compare the substitution rates of
diene in the tetracarbonyl(h2:2-diene)metal(0) complexes
with diphosphinoalkanes. In terms of the diene replace-
ment rate, Mo(CO)4(h2:2-NBD) is found to be the most
labile complex, while W(CO)4(h2:2-COD) is the least
labile one in the presence of bidentate diphosphines.
The lability of NBD in the M(CO)4(h2:2-NBD) com-
plexes with respect to substitution with diphosphine
increases in the order CrBWBMo. This order
changes as WBMoBCr for COD substitution in
M(CO)4(h2:2-COD). Cr(CO)4(h2:2-NBD) is found to be
In addition, in a dissociative mechanism, the rate-
limiting step involves only bond breaking as in
MꢀNBD bond, whereas in an associative mechanism,
the rate-determining step involves both bond breaking
in MꢀNBD and bond making in M-(entering ligand)
[24]. It is expected that the enthalpy of activation, DH‡,
would approach the MꢀNBD bond energy for a pre-
dominantly dissociative process and be rather indepen-
dent of the nature of the entering ligand, whereas for an
associative process DH‡ is expected to be smaller than
MꢀNBD bond energy [25]. The metalꢀligand bond
strength is known only for 2A, 2B, 2C, and 2D. The
MoꢀNBD bond energy is known to be 111.6 kJ mol−1
in literature [15] and the activation enthalpies for NBD
substitution in 2A are found to be DH‡=58, 51 and 47
kJ mol−1 for DPPM, DPPE, and DPPP, respectively.
Thus, this also clearly indicates that substitution reac-
tion of 2A with DPPM, DPPE or DPPP has an associa-
tive mechanism in the transition state. The bond-energy
measurement of 1A and 3A was not found in the
literature. However, according to the activation data,
the same result could be arrived at for the substitution
reaction of 1A and with 3A DPPM, DPPE and DPPP.
The starting complexes 1A–3A are assumed to have
the same structure [26]. The diphosphines essentially act
as bidentate ligands. Because of these similarities in the
structure of starting complexes and in the ligating be-
havior of the diphoshines, a common mechanism is
expected to be valid in each of the M(CO)4(h2:2-NBD)
complex series. Variation in the rate within a reaction
series may be caused by changes in the enthalpy or
entropy of activation. If these quantities vary in a
parallel fashion, compensating each other to produce
minor changes in the free energy of activation of the
process under investigation, thus a plot of DH‡ versus
DS‡ will be linear. This is a consequence of the ex-
trathermodynamic relationship [27], DH‡=TcDS‡. This
linear relationship between entropy and enthalpy of
activation is also referred to as the isokinetic relation-
ship or the compensation effect, and considered as a
test for reaction series with a common mechanism. The
slope of the linear plot of DH‡ versus DS‡ is Tc, the
compensation or isokinetic temperature, at which all
the reaction represented on the line occur at the same
rate. The existence of an isokinetic plot for a reaction
series implies that the relative order of rates for all the
reactions in the series is simply inverted on passing the
Fig. 5. Plot of DH‡ vs. DS‡ for the NBD substitution in 1A–3A by
DPPM, DPPE, and DPPP.

8
A. Tekkaya, S. Ozkar / Journal of Organometallic Chemistry 590 (1999) 208–216
216
[9] J.A. Connor, J.P. Day, E.M. Jones, G.K. McEwen, J. Chem.
Soc. Dalton Trans. (1973) 347.
the most inert complex with respect to the NBD substi-
tution. This can be ascribed to the comfortable coordi-
nation of the NBD, which requires less space than the
COD around the central chromium atom. With respect
to the COD substitution, W(CO)4(h2:2-COD) is found
to be the most inert complex which might be attributed
to the size compatibility of the diene ligand and the
central atom.
8
[10] C. Kayran, S. Ozkar, W.I.M. Sultan, J. Chem. Soc. Dalton
Trans. (1994) 2239.
[11] D.T. Dixon, J.C. Kola, J.A.S. Howell, J. Chem. Soc. Dalton
Trans. (1984) 1307.
[12] W.A. Schenk, J. Organomet. Chem. 184 (1980) 195.
[13] F. Basolo, R.G. Pearson, Mechanisms of Inorganic Reactions,
second ed., Wiley, New York, 1967.
[14] J.R. Sowa, R.J. Angelici, Inorg. Chem. 30 (1991) 3534.
[15] S.L. Mukerjee, S.P. Nolan, C.D. Hoff, R.L. de la Vega, Inorg.
Chem. 27 (1988) 81.
Acknowledgements
[16] S. Zhang, I.H. Wang, P.H. Werner, C.B. Dobson, G.R. Dobson,
Inorg. Chem. 31 (1992) 3482.
[17] D.T. Dixon, P.M. Burkinshaw, J.A.S., Howell, J. Chem. Soc.
Dalton Trans. (1980) 2237.
Support of this work by Volkswagen Stiftung,
8
8
TUBA, and TUBITAK under Grant no. TBAG-1226 is
gratefully acknowledged.
[18] (a) S. Zhang, G.R. Dobson, Inorg. Chem. 29 (1990) 598. (b)
F.-W. Grevels, J, Jacke, P. Betz, C. Kru¨ger, Y.H. Tsay,
Organometallics 8 (1989) 293. (c) W.F. Grevels, J, Jacke, W.E.
Klotzbu¨cher, K, Schaffner, R.H. Hooker, A.J. Rest, J.
Organomet. Chem. 382 (1990) 201.
References
[19] W.F. Grevels, J Jacke, W.E. Klotzbu¨cher, F. Mark, C. Kru¨ger,
8
C.W. Lehmann, S. Ozkar, K Schaffner, V. Skibbe, Organometal-
lics 18 (1999) 3278.
8
[1] (a) C. Kayran, F. Kozanog˘lu, S. Ozkar, S. Saldamli, A. Tek-
kaya, C.G. Kreiter, Inorg. Chim. Acta 284 (1999) 229. (b) C.
[20] C.A. Bunton, Nucleophilic Substitution of a Saturated Carbon
Atom, Elsevier, London, 1963.
8
8
8
Kayran, E.Okan, S. Ozkar, O. Oztu¨rk, S. Saldamli, A. Tekkaya,
C.G. Kreiter, Pure Appl. Chem. 69 (1997) 193. (c) A. Tekkaya,
C. Kayran, S. Ozkar, C.G. Kreiter, Inorg. Chem. 33 (1994) 2439.
[21] R.J. Angelici, J.R. Graham, J. Am. Chem. Soc. 87 (1965) 5586.
[22] J.R. Graham, R.J. Angelici, J. Am. Chem. Soc. 87 (1965) 5590.
[23] (a) R.J. Angelici, J.R. Graham, J. Am. Chem. Soc. 88 (1966)
3658. (b) R.J. Angelici, J.R. Graham, Inorg. Chem. 6 (1967)
2082.
8
[2] (a) J. Chatt, H.R., Watson, J. Chem. Soc. (1961) 4980. (b) K.K.
Cheung, T.F. Lai, K.S. Mok, J. Chem. Soc. A (1971) 1644.
[3] F.-W. Grevels, J.G.A. Reuvers, J. Takats, Inorg. Synth. 24
(1986) 176.
[4] (a) H. Werner, R. Prinz, Chem. Ber. 100 (1967) 265. (b) M.A.
Bennet, L. Pratt, G. Wilkinson, J. Chem. Soc. (London) 75
(1961) 2037.
[5] D.J. Darrensbourg, R.L. Kump, Inorg. Chem. 17 (1978) 2680.
[6] G.T. Andrews, I.J. Colquhoun, W. McFarlane, Polyhedron 2
(1983) 783.
[7] P.S. Braterman, Metal Carbonyl Spectra, Academic, London,
1975.
[24] F. Basolo, Polyhedron 9 (1990) 1503.
[25] K.J. Laidler, Pure Appl. Chem. 53 (1981) 753.
[26] W.F. Grevels,
J Jacke, P. Betz, C. Kru¨ger, Y.H. Tsay,
Organometallics 8 (1989) 293.
[27] (a) M.V. Twigg, Mechanisms of Inorganic and Organometallic
Reactions, Plenum, New York, 1994. (b) J.E. Leffler, E. Grun-
wald, Rates and Equilibria of Organic Reactions, Wiley, New
York, 1963.
8
8
[28] D. Ulku¨, M.N. Tahir, S. Ozkaˆr, Acta Crystallogr. C53 (1997)
185.
[8] B.S. Creaven, F.-W. Grevels, C. Long, Inorg. Chem. 28 (1989)
2231.
.