Interconversion of MeReO(dithiolate)(NC5H4-X) and MeReO(dithiolate)(PAr3) complexes: The equilibrium constants follow the Hammett equation but the rate constants do not

Article abstract of DOI:10.1039/b306634g

Equilibration occurs among the species MeReO(dithiolate)Py, MeReO(dithiolate)(PZ₃), Py, and PZ₃ where the chelating dithiolate ligand is 1,2-ethanedithiol (edt) or 1,3-propanedithiol (pdt), Py stands for NC₅H₄-4-X and PZ₃ for both P(C ₆H₄-4-Y)₃ and P(alkyl)_n(Ph) _3-n. Equilibrium constants in the pdt series were evaluated directly; values of K generally favor phosphane coordination and range from 4.8 × 10^-2 (X = NMe₂, Y = Cl) to 3.2 × 10 ⁴ (X = CN, Y = OMe). The values of K are well correlated by the Hammett equation with ρ_X^K = 2.7(3) and ρ _Y^K = -2.0(3). Kinetic data were determined with the stopped-flow method for 65 reactions of the edt and pdt complexes, and resolved into forward and reverse components by use of the equilibrium constants. Values of k_for, deviate markedly from Hammett behavior, especially along any series with a given X substituent, where plots of log k_for against 3σ_Y take on a V-shaped appearance. This pattern has been interpreted in terms of a two step mechanism for ligand substitution reactions of these complexes. The rate constants for those phosphanes that are the better Lewis bases are governed by Re-P bond formation. The rate constants for those phosphanes that are weaker Lewis bases, on the other hand, are governed by the second step in which an initial ψ-octahedral complex rotates towards a transition state that is an approximate trigonal prism. In so doing, the prior Re-P interaction is weakened, which gives rise to an increase in log k_for with σ_Y.

Full text of DOI:10.1039/b306634g

View Article Online / Journal Homepage / Table of Contents for this issue
Interconversion of MeReO(dithiolate)(NC H -X) and
5
4
MeReO(dithiolate)(PAr ) complexes: The equilibrium constants
3
follow the Hammett equation but the rate constants do not†
Xiaopeng Shan and James H. Espenson*
Ames Laboratory and Department of Chemistry, Iowa State University, Ames, Iowa 50011,
USA; Fax: ϩ01-515-294-5233; Tel: ϩ01-515-294-5730
Received 11th June 2003, Accepted 28th July 2003
First published as an Advance Article on the web 15th August 2003
Equilibration occurs among the species MeReO(dithiolate)Py, MeReO(dithiolate)(PZ ), Py, and PZ where the
3
3
chelating dithiolate ligand is 1,2-ethanedithiol (edt) or 1,3-propanedithiol (pdt), Py stands for NC H -4-X and PZ
5
4
3
for both P(C H -4-Y) and P(alkyl) (Ph) . Equilibrium constants in the pdt series were evaluated directly; values
6
4
3
n
3Ϫn
Ϫ2
4
of K generally favor phosphane coordination and range from 4.8 × 10 (X = NMe , Y = Cl) to 3.2 × 10 (X = CN,
2
K
K
Y = OMe). The values of K are well correlated by the Hammett equation with ρX = 2.7(3) and ρY = Ϫ2.0(3). Kinetic
data were determined with the stopped-flow method for 65 reactions of the edt and pdt complexes, and resolved into
forward and reverse components by use of the equilibrium constants. Values of kfor deviate markedly from Hammett
behavior, especially along any series with a given X substituent, where plots of log kfor against 3σ take on a V-shaped
Y
appearance. This pattern has been interpreted in terms of a two step mechanism for ligand substitution reactions of
these complexes. The rate constants for those phosphanes that are the better Lewis bases are governed by Re–P bond
formation. The rate constants for those phosphanes that are weaker Lewis bases, on the other hand, are governed by
the second step in which an initial ψ-octahedral complex rotates towards a transition state that is an approximate
trigonal prism. In so doing, the prior Re–P interaction is weakened, which gives rise to an increase in log kfor with σY.
Introduction
which also satisfies the principle of microscopic reversibility.
14
This unusual aspect was explored in our earlier research, and
it will be amplified in the Discussion section, on the basis of the
new results and their mechanistic analysis.
Ligand substitution reactions of five-coordinate, square-
pyramidal methyl(oxo)rhenium() dithiolate complexes can be
represented by the equation
1
,2
Two examples of this reaction have been explored previously.
The seemingly simpler one is pure exchange itself, between
i
j
j
i
MeReO(dt)L ϩ L = MeReO(dt)L ϩ L
(1)
i
i
MeReO(edt)Py and Py . These are rapid reactions that were
studied by NMR line broadening. In every case, pyridine
exchange follows second-order kinetics and proceeds with an
In this notation, dt represents 1,2-ethanedithiolate (edt), 1,3-
propanedithiolate (pdt), or 2-(mercaptomethyl)thiophenolate
15
associative mechanism. A more telling example of ligand
substitution (which further established that the Py exchange
process does not proceed in a single-step) is provided by
reactions between MeReO(mtp)(PZ ) and PY , where Z and
i
j
(
mtp) and the ligands L and L may be, for example, phosphines
3–8
or pyridines. Structural formulas are shown in Chart 1.
Our interest in this family of reactions has been sustained by
two factors. First, they are intimately involved in the mech-
anism of rhenium-catalyzed oxygen–atom transfer reactions
3
3
Y include aryl and alkyl groups. In this case substitution
proceeded in two steps with the intervention of easily detected
2
,9–13
(
YO ϩ X
Y ϩ XO).
Second, they present a fundamental
concentrations of the intermediate [MeReO(mtp)(PY )]*, a
3
14
issue of mechanism that appears unique to this structure, in
that no direct route is available to the symmetric transition state
that is necessary for an associative mechanism for substitution
geometric isomer of the final product.
In the present study we have undertaken a study of the
reactions of complexes bearing edt and pdt ligands, to avoid
sequential reactions of geometrical isomers. Important as this
aspect was in the mtp case, the complicating factor of observ-
able sequential reactions of geometrical isomers can be
eliminated by use of a symmetric dithiolate. This study thus
†
Electronic supplementary information (ESI) available: Additional
plots of kinetic data and analysis. See http://www.rsc.org/suppdata/dt/
b3/b306634g/
Chart 1 Structural formulas of methyl(oxo)rhenium() dithiolates.
3
612
D a l t o n T r a n s . , 2 0 0 3 , 3 6 1 2 – 3 6 1 6
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3

View Article Online
entailed direct equilibrium and kinetic measurements on the
reaction
MeReO(dt)(Py) ϩ PZ = MeReO(dt)(PZ ) ϩ Py (2)
3
3
where PZ denotes a generalized phosphane (Z = aryl, alkyl;
3
often P(C H -4-Y) ), Py a generalized pyridine, NC H -4-X,
6
4
3
5
4
and the dithiolates edt or pdt. The symmetry of the inter-
mediate precludes a second step being detectable, [MeReO-
(
dt)(PY ]* and [MeReO(dt)(PY )] being enantiomers not
3
3
geometrical isomers. Nonetheless, there will still be two steps
in reaching the transition state. They respond differently to
substituent groups X and Y on the ligands, however, which
gives rise to non-Hammett behavior as one step or the other
becomes rate-controlling.
Experimental
General
Methyl(oxo)rhenium() dithiolate dimers, {MeReO(edt)} and
2
{
MeReO(pdt)} , were synthesized according to literature
2
4
procedures from 1,2-ethanedithiol (edtH ) or 1,3-propane-
2
dithiol (pdtH ) and {MeReO(SPh) } . The latter compound
2
2 2
1
6
was prepared from MeReO and PhSH, as reported. Pyridines
3
and phosphanes were purchased (Aldrich or Strem) and used as
received. Benzene (ACS grade, Aldrich) was used as the solvent
for UV-visible spectroscopy. Solutions of either MeReO(dt)-
(
NC H -4-X) complex in C H or C D6 were prepared by
5 4 6 6 6
adding an excess of that pyridine in a solution of the dimer. An
excess of pyridine was maintained to ensure that the no dimeric
Fig. 1 Upper: Stopped flow kinetic data showing repetitive scans of
reaction 1 with these conditions: 0.05 mM MeReO(edt)(NC H ), 10.0
6,17
5
5
compound remained.
The mononuclear complexes were
mM PPh and 10.0 mM Py in C H at 25 ЊC. Data were collected over a
3
6
6
identified by NMR spectroscopy in comparison with spectra
total time of 5.0 s at intervals of 0.0125 s; for sake of clarity only scans
at 0.1 s intervals are shown. The inset shows a fit to first-order kinetics
of the data extracted at 367 nm. Lower: The results of a global fit of the
kinetic data by the singular value decomposition method, showing
the spectra of reactant and product separately resolved by this analysis
and an inset in which the spectra have been resolved into their reactant
and product components, expressed as molar ratios.
4
determined previously.
Equilibrium studies
Values of K1 were determined by UV-visible spectrophoto-
metric titration only for the pdt complexes. A concentration in
the range of 10–500 mM of the given Py ligand was placed in a
1
cm path UV-visible cell containing 0.05 mM {MeReO(pdt)}2.
of pyridine was used to prevent formation of the parent dimer,
The spectrum was recorded using a Shimadzu UV 3101PC
spectrophotometer after the addition of each 10 µL portion of
{MeReO(edt)} . The rapid-scan spectra for this particular
experiment are shown in Fig. 1. Reaction (2) proceeds nearly to
completion with these concentrations.
2
a solution of PAr . This procedure was repeated until the
3
spectra remained unchanged with further addition of PAr .
3
18
The software PSEQUAD was used to analyze the multi-
wavelength absorbance data to obtain the equilibrium constant.
These calculations are based on the net absorbance change
Results and discussion
Equilibrium constants
from beginning to end of the addition of PAr . Typically,
3
Values of K were determined for 11 reactions of the pdt
absorbance readings were recorded at 10 different wavelengths
in the 400–600 nm range; this set of values was used in a global
fit to extract the equilibrium constant for selected pyridine and
phosphane derivatives.
complexes for 10 pairs of Py and PZ derivatives: (X = NMe ,
3
2
Y = OMe), (X = NMe , Y = Me), (X = OMe, Y = Me), ( X =
2
OMe, Y = H), (X = Me, Y = H), (X = Me, Y = F), (X = H, Y= F),
(
X = H, Y = Cl), (X = CHO, Y = Cl), (X = Ac, Y = Cl), (X = CN,
Y = Cl). The values of K, which have been determined to a
precision of ca. 10%, are given in Table 1. The remaining 24
values of K are redundant and were calculated directly or
indirectly from these eleven experimental values; of course
the calculated ones will have higher error limits owing to the
propagation of errors. The general formula is as given in
eqn. (4) for the equilibrium
Kinetics
Rate constants for reaction (2) were obtained by the use of
an OLIS stopped-flow instrument equipped with a Rapid
Scanning Monochromator system set for the range 350–500
nm. The absorbance readings monitor the decrease of absorb-
ance at 350–395 nm, in the particular instance of C H N and
5
5
PPh , where ε
> εRePAr3 and the simultaneous increase at
3
RePy
3
95–415 nm, where the reverse inequality holds. A persistent
[
Re]Py ϩ PAr = [Re]PAr3B ϩ PyA
(3)
A
3B
isosbestic point for this pair was found at 395 nm which showed
that no intermediate attains an appreciable concentration
during the reaction. The same was true for the other pyridine-
phosphine combinations, save for small differences in the wave-
lengths cited. Circulating thermostated water was used to
control the temperature at 25.0 ± 0.2 ЊC.
(
4)
This particular equilibrium illustrates the procedure:
A typical experiment involved mixing equal volumes of
benzene solutions containing 20.0 mM PPh3 and 0.1 mM
MeReO(edt)(NC H ) containing 20.0 mM C H N. The excess
[Re]NC H CHO ϩ P(C H Me) =
5
4
6
4
3
[Re]P(C H Me) ϩ NC H CHO (5)
5
5
5
5
6
4
3
5
4
D a l t o n T r a n s . , 2 0 0 3 , 3 6 1 2 – 3 6 1 6
3613

View Article Online
a
Table 1 Equilibrium constants for reactions between MeReO(pdt)(NC H -4-X) and P(C H -4-Y) in C H at 25 ЊC
5
4
6
4
3
6
6
K
X↓ Y
OMe
Me
H
F
Cl
NMe2
OMe
Me
8.2 ± 0.8
300
340
1300
(4700)
(3700)
(32000)
2.1 ± 0.2
0.40
15 ± 1
17 ± 1
63
230
180
0.14
5.2
5.9 ± 0.6
22 ± 2
80
62
0.048
1.8
2.0
78 ± 7
88
330
1200
930
8000
H
7.4 ± 0.7
CHO
C(O)Me
CN
27 ± 2
21 ± 2
1500
530
180 ± 20
a
Experimental values are given in boldface; the others are redundant values of K calculated from them according to eqns. (4)–(6) as explained in the
text. Standard deviations of the calculated entries are obtained by the law of propagation of errors. Values enclosed parenthetically have errors nearly
as large as the entries themselves.
Ϫ1
Ϫ1
The equilibrium constant for reaction (3), K(CHO,Me) is
arrived at in these steps:
Table 2 Rate constants (kfor/L mol
s ) for reactions between
a
MeReO(pdt)(NC H -4-X) and P(4-YC H ) in C H at 25 ЊC
5
4
6
4
3
6
6
X↓ Y
OMe
Me
H
F
Cl
(
6)
NMe2
OMe
Me
—
—
91 ± 1
82 ± 1
114 ± 2
235 ± 2
201 ± 2
477 ± 5
28.8 ± 0.3
78 ± 1
71 ± 1
103 ± 1
171 ± 2
154 ± 2
401 ± 6
—
—
157 ± 2
137 ± 2
190 ± 2
368 ± 5
340 ± 7
760 ± 20
137 ± 2
121 ± 2
166 ± 2
162 ± 3
166 ± 4
251 ± 3
156 ± 2
136 ± 2
182 ± 3
179 ± 5
197 ± 3
281 ± 3
where the standard deviation was calculated from those of
the components. In Table 1, 11 values of K derive from experi-
ments, 9 can be calculated from three experimental values, and
the remaining three by combining experimental and calculated
H
CHO
C(O)Me
CN
values. Naturally, the latter group have the largest error limits.
a
Values of kfor were obtained by fitting k to eqn. (10) with K from
e
4
Not surprisingly, the largest K is 3.2 × 10 for NC H CN,
5
4
Table 1.
which has the most electron-withdrawing substituent, in
combination with P(C H OMe) , with the most electron
6
4
3
releasing substituent. Likewise, the smallest K value is 0.048 for
in eqn. (10) does not contribute. For the case where krev does
contribute, K could be determined, as given in Table 4.
Although these values have considerable uncertainties associ-
ated with them, the trend is clear that Kedt > Kpdt. Perhaps this is
so because the five-membered ring of edt is nearly co-planar,
whereas the six-membered with pdt adopts a chair conform-
ation that makes ligand attachment more difficult, especially for
large phosphines.
the reaction between [Re]NC H NMe and P(C H Cl) .
5
4
2
6
4
3
Kinetic data
Reaction (2) was studied with an excess of the pyridine present,
Ϫ1
usually at a concentration of 10 mmol L , to prevent compli-
cations arising from the presence of the dimer {MeReO(pdt)}2,
were the following equilibrium to proceed to the right to any
substantial extent,
Reactions of a selected group of other phosphanes were also
studied, with the results presented in Table 5. Presented along
with the rate constants are the cone angles and pK values of
2
MeReO(pdt)Py = {MeReO(pdt)} ϩ 2 Py
(7)
a
2
19
the conjugate acids of the phosphanes. Although PMePh and
2
P(C H -4-OMe) have the same value of pK , the rate constant
The phosphanes were also used in excess over rhenium, such
6
4
3
a
of ligand substitution of the former is 55 times faster than that
of the latter owing to the effect of steric hindrance, as indicated
by a bigger cone angle, 145Њ.
that both [Py] and [PAr ] remained constant in any experiment.
3
The rate of approach to equilibrium is reaction (2) can be
written as follows, with (Re) standing for MeReO(pdt)
Temperature profiles. The temperature dependence of the
(
8)
reaction between MeReO(edt)Py and PPh was studied over the
3
temperature range of 283–323 K. Activation parameters were‡
calculated from the least-squares fit of ln(k/T) against 1/T; ∆H
The rate constant for equilibration thus becomes
Ϫ1
‡
Ϫ1
Ϫ1
=
24.0 ± 0.5 kꢀ mol and ∆S = –124 ± 2 ꢀ mol K . The
large negative activation entropy indicates an associative
k = k [PAr ] ϩ krev[Py]
(9)
e
for
3
20–23
pathway,
which is compatible with earlier findings which
have demonstrated that the rate depends directly on the identity
and concentration of the incoming ligand.
With incorporation of K from Table 1, this expression is
obtained
14
Ϫ1
ke = kfor × {[PAr ] ϩ [Py]K }
(10)
Substituent effects
3
Ϫ1
Equilibrium constant. One must take into account the effects
Thus a plot of k against {[PAr ] ϩ [Py]K } represents
e
3
of variations of substituents X on Py and Y on PAr , which act
3
the equation of a straight line that passes through the origin.
The least-squares fit to eqn. (10) gave values of kfor that are
summarized in Table 2.
Reactions of MeReO(edt)Py required a different analysis
because K was not independently evaluated. Eqn. (10) was also
used for the data analysis, but with both kfor and K treated as
adjustable parameters. The values of kfor are given in Table 3.
The equilibrium constant is often so large that the second term
in concert. According to studies of multiple-substituent effects
24–27
in organic chemistry,
a correlation was made using the
Hammett substituent constants, taken from the compilation of
28
Brown and Okamoto for the substituent constants of X and
Y in the overall expression:
K
K
log K(X,Y) = log Kref ϩ ρX × σ ϩ ρ × (3σY) (11)
X
Y
3
614
D a l t o n T r a n s . , 2 0 0 3 , 3 6 1 2 – 3 6 1 6

View Article Online
Ϫ1 Ϫ1
Table 3 Rate constants (kfor/L mol
s
) for reactions between MeReO(edt)(NC H -4-X) and P(4-YC H ) in C H at 25 ЊC
5
4
6
4
3
6
6
X↓ Y
OMe
Me
H
F
Cl
a
a
b
b
NMe2
OMe
Me
91 ± 1a
227 ± 5a
205 ± 6
50.5 ± 0.7
39.9 ± 0.4
150 ± 20
—
a
b
b
a
120 ± 2a
96 ± 2b
143 ± 2b
186 ± 1b
148 ± 5
160 ± 10
111 ± 2a
152 ± 2a
284 ± 3
91 ± 1b
128 ± 2a
208 ± 2
133 ± 2a
179 ± 2a
168 ± 2
a
b
H
300 ± 20a
a
CHO
C(O)Me
CN
530 ± 20
189 ± 4a
a
a
a
a
480 ± 10a
258 ± 6a
195 ± 2a
174 ± 2a
211 ± 2a
251 ± 3
1190 ± 40
591 ± 6
480 ± 8
279 ± 8
a
Ϫ1
b
Ϫ1
Values of k were analyzed according to eqn. (10), with the term [Py]K omitted. The term [Py]K of eqn. (10) was retained.
e
Table 4 Equilibrium constants for reactions of MeReO(edt)(NC H -
In a second treatment, K(H,H) was fixed at the experimental
5
4
a
4
-X) with P(C H -4-Y) at 25 ЊC in benzene
K
6
4
3
value, which resulted in these reaction constants, ρX = 2.7 ±
K
0
.3 and ρY = Ϫ2.0 ± 0.3, which are not significantly different.
X↓ Y
H
F
Cl
The reaction constants reflect the entering and leaving groups.
Including the calculated but no less valid values of K in Table 1,
Me N
2.5 ± 0.2
7 ± 4
36 ± 29
—
3.0 ± 1.8
18 ± 5
38 ± 12
—
—
—
11 ± 4
8 ± 4
2
the entire range of equilibrium constants spans a factor of
MeO
Me
H
5
6
.7 × 10 , which indicates that the difference between reactant
and product is quite sensitive to substituents.
a
From fitting of kinetic data to eqn. (10).
Rate constants. The entries in Tables 2 and 3 indicate that
factors other than simple electronic effects on a single rate
constant come into play. For one thing, the span of rate
constant values for the same X, Y substituent pair is consider-
ably narrower than that for K. For example, values of kfor in
Table 2 vary by but a factor of 30, as compared to a range
Table 5 Rate constants of reactions between MeReO(edt)(NC H )
5
5
and other phosphanes in C H at 25 ЊC
6
6
Ϫ1 Ϫ1
PZ3
θ/Њ
pKa
k/L mol
s
4
2
of 8 × 10 in K, the same ligands considered. Further, the span
PCyPh2
153
162
136
145
107
5.05
—
4.57
4.57
2.60
2.30 × 10
3.01 × 10
1.42 × 10
2.57 × 10
1
4
2
PCy Ph
of the rate constants with variation of the Y substituent is
narrower than that for the variation of the X substituent, which
reverses the trends for the equilibrium constants.
Even more strikingly, the kfor values show unusual trends.
The small reversal of the ordering of rate constants in each
column of Table 2 is barely significant. But the falling and
rising of the values in each row with variation of Y is rather
pronounced. The same pattern held for every Y group. Put
2
PMePh2
P(C H -4-OMe)
6
4
3
4
P(MeO)3
> 4 × 10
The factor σY includes a multiplying factor of three for
P(C H Y) . The fitted reaction constants for K are symbolized
6
4
3
K
K
by ρX and ρY . For this analysis only the 11 experimental data
from Table 1 were used, because the remaining values of K were
calculated ones whose inclusion would not enhance the
analysis. The reaction constants were obtained by use of the
another way, a plot of log kfor versus σ , for any given substituent
Y
X on Py, shows a distinctive V-shape. Fig. 3 depicts these
nonlinear Hammett plots for a few of the reactions.
29
least-square program SCIENTIST which allows for the use
K
of two independent variables from the fit, ρ = 2.2 ± 0.3 and
X
K
ρY = Ϫ1.5 ± 0.3. These parameters have opposite signs,
of course, because pyridine is coordinated to the reactant,
phosphane to the product. In this calculation log K(H,H) was
also allowed to float, the refined value being 1.58 ± 0.09, as
compared to the experimental value 1.80 ± 0.11. ꢁiven the
error, the internal consistency is adequate, as illustrated in
Fig. 2 by a plot of log (Kcalcd vs. Kexp) which has a 45Њ slope.
Fig. 3 Hammett plots of log k for the reactions of MeReO(dt)-
[
(
NC H -4-CHO) with different P(C H -4-Y) compounds for dt = edt
5
4
6
4
3
hatched squares) and pdt (filled squares) showing the non-linear
relationship indicative of a sequential-step mechanism.
In the Introduction, a complex mechanism was mentioned.
Our depiction of that mechanism is given in Scheme 1, involving
multiple steps: PZ approaches the rhenium from the vacant
3
axial position to form a six-coordinate intermediate A; a Bailar
twist transfers A to its geometric isomer B; Py departs from
30,31
B to generate the final product.
The coordination number
of rhenium complexes can easily vary from five to six. Bailar
twist has been already proved to be a common isomerization
pathway for certain six-coordinate transition metal com-
Fig. 2 Illustrating the agreement between experimental (Table 1) and
Hammett-correlated values of the equilibrium constants for reactions
between MeReO(pdt)(NC H -4-X) and P(C H -4-Y) . The least-
5
4
6
4
3
32–37
squared line has a slope of 1.01.
plexes.
The justification for such an elaborate scheme, and
D a l t o n T r a n s . , 2 0 0 3 , 3 6 1 2 – 3 6 1 6
3615

View Article Online
under Contract W-7405-Eng-82 with Iowa State University of
Science and Technology.
References
1
2
3
ꢁ. Lente and ꢀ. H. Espenson, Inorg. Chem., 2000, 39, 4809–4814.
Y. Wang and ꢀ. H. Espenson, Inorg. Chem., 2002, 41, 2266–2274.
ꢀ. ꢀacob, I. A. ꢁuzei and ꢀ. H. Espenson, Inorg. Chem., 1999, 38,
1
040–1041.
ꢀ. H. Espenson, X. Shan, Y. Wang, R. Huang, D. W. Lahti, ꢀ. Dixon,
. Lente, A. Ellern and I. A. ꢁuzei, Inorg. Chem., 2002, 41,
583–2591.
ꢀ. ꢀacob, I. A. ꢁuzei and ꢀ. H. Espenson, Inorg. Chem., 1999, 38,
266–3267.
ꢀ. ꢀacob, ꢁ. Lente, I. A. ꢁuzei and ꢀ. H. Espenson, Inorg. Chem.,
999, 38, 3762–3763.
4
ꢁ
2
5
6
7
8
9
3
1
ꢁ. Lente, ꢀ. ꢀacob, I. A. ꢁuzei and ꢀ. H. Espenson, Inorg. React.
Mech., 2000, 2, 169–177.
ꢁ. Lente, X. Shan, I. A. ꢁuzei and ꢀ. H. Espenson, Inorg. Chem.,
2
000, 39, 3572–3576.
Y. Wang and ꢀ. H. Espenson, Org. Lett., 2000, 2, 3525–3526.
1
1
1
0 Y. Wang, ꢁ. Lente and ꢀ. H. Espenson, Inorg. Chem., 2002, 41,
272–1280.
1 ꢀ. Arias, C. R. Newlands and M. M. Abu-Omar, Inorg. Chem., 2001,
0, 2185–2192.
2 S. Bhattacharyya, I. Chakraborty, B. K. Dirghangi and A.
Chakravorty, Inorg. Chem., 2001, 40, 286–293.
1
4
1
3 K. P. ꢁable and E. C. Brown, Organometallics, 2000, 19, 944–946.
4 D. W. Lahti and ꢀ. H. Espenson, J. Am. Chem. Soc., 2001, 123,
6014–6024.
1
Scheme 1 Two step mechanism for ligand substitution.
1
5 ꢀ. H. Espenson, X. Shan, D. W. Lahti, T. M. Rockey, B. Saha and
A. Ellern, Inorg. Chem., 2001, 40, 6717–6724.
the data that support it (which include the spectroscopic
14
16 W. A. Herrmann, R. M. Kratzer and R. W. Fischer, Angew. Chem.,
Int. Ed., 1997, 36, 2652–2654.
detection of intermediate A) have been presented earlier.
iven this mechanism as a premise, the expression for the rate
constant is
ꢁ
1
7 ꢁ. Lente, I. A. ꢁuzei and ꢀ. H. Espenson, Inorg. Chem., 2000, 39,
311–1319.
1
1
8 L. Zékány, I. Nagypál, Computational Methods for the
Determination of Formation Constants, ed. D. Leggett, New York,
(
12)
1
985 . We are grateful to ꢁábor Lente for assistance in its
implementation.
1
2
2
2
9 P. B. Dias, M. E. Minas de Piedade and ꢀ. A. Martinho Simoes,
Coord. Chem. Rev., 1994, 135, 737–807.
0 C. Kayran, F. Kozanoglu, S. Ozkar, S. Saldamli, A. Tekkaya and
C. ꢁ. Kreiter, Inorg. Chim. Acta, 1999, 284, 229–236.
1 A. Lugovskoy, R. Paur-Afshari and R. H. Schultz, J. Phys. Chem. A,
There are two limiting forms of eqn. (1). One identifies kfor as
k , which will be the case when k ӷ k . In that case the first
1
2
Ϫ1
step is rate-controlling, and the second much more rapid in
comparison. Whether the limit is actually realized is not the
major issue; if it is, or nearly is, the rate will be governed by
bond-making between Re and the incoming phosphane ligand.
As such, the “ordinary” nucleophilic effect will be seen, and
the more electron-donating Y is, the faster the reaction. This is
evidently the case for Y = MeO and Me, for example.
2
000, 104, 10587–10593.
2 Q. Meng, Y. Huang, W. ꢀ. Ryan and D. A. Sweigart, Inorg. Chem.,
1992, 31, 4051–4052.
23 ꢀ.-K. Shen, ꢁ. ꢀ. Kubas and A. L. Rheingold, Inorg. Chim. Acta,
995, 240, 99–104.
1
2
4 A. Argile and M. F. Ruasse, J. Org. Chem., 1983, 48, 209–214.
2
5 ꢀ. E. Dubois, M. F. Ruasse and A. Argile, J. Am. Chem. Soc., 1984,
The other limit has kfor = (k /k )k , in which case the first
1
Ϫ1
2
1
06, 4840–4845.
step is at equilibrium and the second is rate-controlling. When
2
6 I. Lee, Y. K. Park, C. Huh and H. W. Lee, J. Phys. Org. Chem., 1994,
7, 555–560.
P(C H -4-Y) is less nucleophilic, as with Y = F and Cl, then
6
4
3
the first step will feature a small binding constant, k /k Ӷ 1;
27 I. Lee, J. Phys. Org. Chem., 1994, 7, 448–454.
1
Ϫ1
2
2
3
8 H. C. Brown and Y. Okamoto, J. Am. Chem. Soc., 1958, 60, 4979.
moreover, the weakly-bound phosphine will leave more rapidly
9 Scientist, Micromath Software, 2.0 edn., 1995.
than the turnstile rotation step; that is, kϪ1 ӷ k . Because k2
2
0 M. R. Cook, W. A. Herrmann, P. Kiprof and ꢀ. Takacs, J. Chem.
Soc., Dalton Trans., 1991, 797–804.
governs the substituent effect under those circumstances, the
reaction constant will show a rising effect with σY as co-
ordination of the bulky phosphane group weakens as the
molecule approaches the configuration of a trigonal prism
because the ligands must rotate past one another. Again, the
equilibrium constants for the same reaction conform to an
ordinary Hammett relationship because only the initial and
final states of the system come under consideration, not the
special intermediate.
31 W. A. Herrmann, F. E. Kuehn and P. W. Roesky, J. Organomet.
Chem., 1995, 485, 243–251.
32 F. Abugideiri, ꢀ. C. Fettinger, B. Pleune, R. Poli, C. A. Bayse and
M. B. Hall, Organometallics, 1997, 16, 1179–1185.
3
3 S. Alvarez, M. Pinsky and D. Avnir, Eur. J. Inorg. Chem., 2001,
499–1503.
4 C. A. Bayse, M. Couty and M. B. Hall, J. Am. Chem. Soc., 1996, 118,
916–8919.
1
3
8
35 L. Kakaliou, W. ꢀ. Scanlon, B. Qian, S. W. Baek, M. R. Smith, III
and D. H. Motry, Inorg. Chem., 1999, 38, 5964–5977.
3
6 N. A. P. Kane-Maguire, T. W. Hanks, D. ꢁ. ꢀurs, R. M. Tollison,
A. L. Heatherington, L. M. Ritzenthaler, L. M. McNulty and
H. M. Wilson, Inorg. Chem., 1995, 34, 1121–1124.
7 M. Rahim, N. ꢀ. Taylor, S. Xin and S. Collins, Organometallics, 1998,
17, 1315–1323.
Acknowledgements
This research was supported by the U.S. Department of Energy,
Office of Basic Energy Sciences, Division of Chemical Sciences
3
3
616
D a l t o n T r a n s . , 2 0 0 3 , 3 6 1 2 – 3 6 1 6