The effect of high pressure on the heck reaction - A contribution to a deeper understanding of the mechanism

Article abstract of DOI:10.1002/ejoc.200200720

The influence of high pressure on the Heck reactions of iodobenzene with methyl, ethyl and tert-butyl acrylate, and of both 4-nitrophenyl iodide and 4-nitrophenyl triflate with methyl acrylate, has been studied for the first time by quantitative on-line F

Full text of DOI:10.1002/ejoc.200200720

FULL PAPER
The Effect of High Pressure on the Heck Reaction ؊ A Contribution to a
Deeper Understanding of the Mechanism
[a]
[b]
[b]
[b]
´
Michael Buback, Thomas Perkovic, Stefan Redlich, and Armin de Meijere*
Keywords: Catalysis / Heck reaction / IR spectroscopy / Kinetics / Palladium
The influence of high pressure on the Heck reactions of iodo- Thus, the observed activation volume for the reaction of iodo-
benzene with methyl, ethyl and tert-butyl acrylate, and of
both 4-nitrophenyl iodide and 4-nitrophenyl triflate with
methyl acrylate, has been studied for the first time by quanti-
tative on-line FT-IR spectroscopy. Reaction rates and activa-
tion parameters (∆H‡, ∆S‡, ∆V‡) were determined at temper-
atures between 50 and 120 °C and pressures up to 3000 bar.
A clear influence on the kinetics of the reactions was ob-
served by the nature of the leaving group and by the steric
and electronic features of the acrylate. tert-Butyl acrylate,
which is more electron-rich than methyl acrylate according
to PES measurements, was found to react 1.4-times faster on
average than methyl acrylate and, at elevated pressures,
triflate turned out to be a better leaving group than iodide.
benzene with tert-butyl acrylate (∆V‡ = −12 cm3 mol−1) is
more negative than that of the reaction of iodobenzene with
ethyl acrylate (∆V‡ = −7 cm3 mol−1) or methyl acrylate (∆V‡ =
−5 cm3 mol−1). For the variation of the leaving group in 4-
nitrophenyl iodide and 4-nitrophenyl triflate, the activation
volumes for the reaction of each with methyl acrylate are −9
and −37 cm3 mol−1, respectively. The results suggest that the
rate-determining step of the overall reaction is not the oxidat-
ive addition, but is either the alkene coordination or the sub-
sequent carbopalladation.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
acceleration (retardation) of the reaction by pressure. Par-
tial formation of a bond en route to the TS between atoms
previously not bonded to each other, and separation of elec-
trostatic charges (as, e.g., in solvolysis), which gives rise to
solvent electrostriction, lead to a negative contribution to
the activation volume. Conversely, partial cleavage of a
bond and dispersal of electrostatic charge en route to the
transition state enhance the activation volume.
A few high-pressure studies have been reported for or-
ganometallic reactions in general,[3Ϫ5] and for the pal-
ladium-catalyzed arylation and alkenylation of alkenes —
the well-known Heck reaction[6] — in particular.[4,7Ϫ12]
Trost et al. observed that palladium-catalyzed [3ϩ2] cyclo-
additions are retarded under high pressure,[5] whereas the
rate of Heck reactions is enhanced by high pressure.[7,9Ϫ12]
Reiser et al. were the first to combine high pressure and
palladium catalysis, by which they achieved an enhanced
enantioselectivity,[7] and an increase of turnover numbers.[8]
High pressure also allows alkenyl and aryl chlorides[9] or
bromides[10] to be coupled at lower temperature and with
better yields as compared to the same reactions at ambi-
ent pressure.
The kinetic approach to elucidating the mechanism of a
chemical reaction involves the measurement of reaction
rates and the determination of rate coefficients as a function
of as many chemical and physical variables as possible.
High pressure has been used as an important variable
within an increasing number of studies into the mechanism
of organic and inorganic reactions.[1,2] From the pressure
dependence of the rate coefficients, k, the activation volume,
∆V‡, is derived according to the relation: ∆V‡ ϭ ϪRT
(Ѩln k/Ѩp)T. Under conditions of chemical reaction control,
the partial molar volume of the transition structure may be
estimated from ∆V‡ at known partial molar volume(s) of
the starting materials. If the transition structure (TS) is
more compact than the sum of the reactands, the reaction
is accelerated by pressure and ∆V‡ is negative, whereas a
reaction with a TS that is more expanded than the sum of
the reactands is retarded by high pressure and ∆V‡ is posi-
tive. The more-negative (positive) is ∆V‡, the larger is the
[a]
Institut für Physikalische Chemie der Georg-August-
Universität Göttingen,
Tammannstrasse 6, 37077 Göttingen, Germany
Most transition metal-catalyzed reactions do not occur
in a single step, but consist of a sequence of individual steps
with different types and extents of pressure dependence. It
is, therefore, difficult to predict whether the overall reaction
will be accelerated or retarded by high pressure. For the
Heck reaction, it has been postulated that the formation of
[b]
Institut für Organische Chemie der Georg-August-Universität
Göttingen,
Tammannstrasse 2, 37077 Göttingen, Germany
Fax: (internat.) ϩ 49-(0)551/399475
E-mail: Armin.deMeijere@chemie.uni-goettingen.de
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2003, 2375Ϫ2382
DOI: 10.1002/ejoc.200200720
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2375

´
M. Buback, T. Perkovic, S. Redlich, A. de Meijere
FULL PAPER
the catalytically active Pd0 species, the oxidative addition of valuable for detailed kinetic studies of organic reactions at
an alkenyl or aryl halide to the Pd0 species, and the com- elevated temperatures and high pressures.[14]
plexation of an alkene to the resulting Pd species are accel-
erated by pressure, whereas the reductive elimination and
Results
decomplexation reactions are retarded by pressure, and that
the insertion, migration and deinsertion steps are not as-
sociated with significant pressure effects.[12] According to
the superb mechanistic and kinetic studies by Amatore and
Jutand into palladium-catalyzed reactions,[13] the individual
steps in the catalytic cycle of the Heck reaction under high
pressure conditions with Pd(OAc)2/PPh3/NEt3 as the cata-
lyst system (Scheme 1)[6] have been found to belong to
three categories:
First, the coupling reactions of iodobenzene 12 with dif-
ferent alkyl acrylates 13a,b,c to yield the corresponding cin-
namates 14a,b,c (Scheme 2) were studied as a function of
pressure, between 250 and 3000 bar, at 70 °C and as a func-
tion of temperature, between 50 and 100 °C, at 1500 bar
The concentrations of the reactands, the solvent, and the
catalyst were the same in the three reaction systems. Thus,
any change in the determined parameters must be related
to differences in steric and/or electronic properties of the
acrylates.
Scheme 2. Heck reaction of iodobenzene (12) with selected acry-
lates
The absorptions of the carbonyl stretching fundamentals
of 13 and of 14 were recorded to monitor the progress of
the reactions. A typical series of absorbance spectra ob-
tained during the reaction of 12 with 13c at 1750 bar and
70 °C over a time period of 6.5 h is shown as an example
in Figure 1. The peak maxima in IR spectra of 13c and
14c occur at 1717 and 1704 cmϪ1, respectively. The arrows
indicate the direction of spectral changes with reaction time.
To eliminate any absorbance of the solvent, of the silicon
windows, and of the walls of the internal Teflon cell, the
first spectrum recorded immediately after the system had
reached constant reaction conditions, for each spectral
series taken during a reaction at constant p and T, was sub-
tracted from all subsequently measured spectra. By this
procedure, the spectral sequence from Figure 1 is trans-
Scheme 1. The mechanism of the Heck reaction and the influence
of high pressure on the individual steps of the reaction
(i) The ligand association (A), intramolecular reduction
(B), oxidative addition (C), coordination (E), syn addition
(F) and reductive elimination (J) steps should be pressure-
accelerated. (ii) The dissociation of XϪ (D), syn elimination
(H) and decomplexation (I) should be pressure-retarded.
(iii) The internal rotation (G) with no changes in bonding
or in electrostatic charges, should not be influenced by
pressure.
Reiser et al. determined the activation volume for the
Heck coupling of iodobenzene with 2,3-dihydrofuran, with
the kinetic data being deduced from several separate runs
and monitored the reaction by GC analysis.[11] To achieve
a more accurate kinetic analysis, we have applied quantitat-
ive on-line FT-IR spectroscopy for measuring concen-
trations in the present study into the Heck coupling under
high pressure of three different acrylates to phenyl iodide
and of methyl acrylate to p-nitrophenyl iodide as well as to
Figure 1. Series of spectra recorded during the Heck reaction of
iodobenzene (12) with tert-butyl acrylate (13c) to form the corre-
sponding cinnamate 14c at 1750 bar and 70 °C (for better clarity
only about one third of the ca. 60 spectra actually recorded are
shown); a reaction time of nearly 6.5 h is covered, corresponding
to 7.5 half-lives; the arrows indicate the direction of intensity
p-nitrophenyl triflate. This technique has proven to be very changes with time.
2376  2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2003, 2375Ϫ2382

The Effect of High Pressure on the Heck Reaction
FULL PAPER
Figure 2. Series of difference spectra recorded during the Heck re-
action of iodobenzene (12) with tert-butyl acrylate (13c) to form
the corresponding cinnamate 14c at 1750 bar and 70 °C (for better
clarity only about one third of the actually recorded spectra is
shown)
Figure 3. Plot of integrated absorbance vs. reaction time for the
Heck reaction of iodobenzene (12) with tert-butyl acrylate (13c) to
the corresponding cinnamate 14c at 1750 bar and 70 °C
ferred into the series shown in Figure 2. Positive absorbance
indicates dominant product absorption, whereas negative
absorbance refers to dominant absorption of starting mate-
rial.
Table 1. Experimental pseudo-first-order rate coefficients k, for the
Heck reaction of iodobenzene (12) with tert-butyl acrylate (13c) to
form the cinnamate 14c, derived from the analysis of the decreasing
IR absorbance band of 13c and the increasing band of 14c
In dilute solutions at constant temperature and pressure,
the LambertϪBeer law holds for the CϭO stretching modes
in Figure 1 and Figure 2. The changes in integrated ab-
sorbance of these vibrations are proportional to changes in
concentration of the associated reactand or product species.
To achieve a reliable quantitative analysis, the areas be-
tween the curves in Figure 2 and the horizontal line through
zero absorbance (difference) were determined by integration
over the high-wavenumber band (for analysis of 13c within
the limits from 1711.5 to 1726.4 cmϪ1) and over the low-
wavenumber band (for analysis of 14c within the limits from
1691.5 to 1711.5 cmϪ1).
With the concentration of the aryl iodide and the aryl
triflate, respectively, by far exceeding the concentration of
the acrylate (by a factor of 30), the coupling reaction may
be treated by first-order kinetics with respect to the acrylate
concentration.[15] Fitting the data of the integrated ab-
sorbance vs. time to a first-order rate expression, by means
T ϭ 70 °C
k [10Ϫ4sϪ1
from 13c from 14c
p ϭ 1500 bar
k [10Ϫ4sϪ1
]
]
p [bar]
T [°C]
from 13c
from 14c
250
500
750
ϪϪ
1.67
1.76
1.94
2.14
2.26
2.43
2.41
2.71
2.86
2.93
3.07
ϪϪ
1.46
1.59
1.71
1.87
1.91
2.12
2.33
2.31
2.42
2.59
2.67
50
60
70
80
90
0.49
1.08
2.26
4.50
7.87
12.3
0.43
1.05
1.91
3.28
6.59
11.5
1000
1250
1500
1750
2000
2250
2500
2750
3000
100
wavenumbers, which could not yet be assigned unequivo-
cally. The results obtained from the analysis of the acrylate
absorbances are representative for the Heck reaction and
serve as the basis for the following discussion. For each
experiment the k values derived from the decreasing re-
actand and from the increasing product absorbance match
within experimental uncertainty.[18] A full report on the ex-
perimental data (k values for Heck reactions of iodoben-
zene with methyl and ethyl acrylate) is provided in the Sup-
porting Information (see also footnote on the first page of
this article).
of
a least-squares procedure using the LevenbergϪ
Marquardt algorithm,[16] yields the first-order rate coef-
ficients, k.[17] The plot of integrated absorbance vs. time,
which relates to the primary spectroscopic data of the
experiment in Figure 1 and Figure 2, is presented in Fig-
ure 3. The time required for reaching constant p and T con-
ditions was typically about ten minutes. This time span is
seen as the plateau region at the onset of the plot.
Table 1 summarizes the pseudo-first-order rate coef-
ficients k for the Heck reaction of iodobenzene and methyl
acrylate that were obtained from the spectra.[18] The slight
deviation of the rate coefficients derived from the analysis
of the cinnamate 14c absorbance band from those derived
from the acrylate 13c absorbance band are caused by the
influence of a time-dependent growth of additional ab-
sorbances in the region from ca. 1690 cmϪ1 down to lower
Activation enthalpies ∆H‡ were derived from the Arrhen-
ius plot (lnk vs. TϪ1) [Equation (1)], and activation entrop-
ies ∆S‡ were calculated according to Equation (2).
∆H‡ ϭ ϪR (dln k/dTϪ1)p Ϫ RT
(1)
(2)
∆S‡ ϭ R [A Ϫ ln (kBT/h) Ϫ 1]
Eur. J. Org. Chem. 2003, 2375Ϫ2382
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2377

´
M. Buback, T. Perkovic, S. Redlich, A. de Meijere
FULL PAPER
where R is the gas constant, A is the natural logarithm of ∆V‡ (13c) Ͻ ∆V‡ (13b) Ͻ ∆V‡ (13a); i.e., the largest nega-
the pre-exponential factor in the Arrhenius expression, kB tive value of ∆V‡ is observed for tert-butyl acrylate (13c).
is the Boltzmann constant, and h is the Planck constant.
Table 2. Activation parameters and experimental rate coefficients
k at 70 °C and 1.5 kbar for the Heck reaction of the acrylates
13a,b,c with iodobenzene (12) to form the corresponding cinna-
mates 14a,b,c
Derived from
∆H‡[a]
∆S‡[a]
∆V‡0
[kJ·molϪ1
]
[J·molϪ1·KϪ1
]
[cm3·molϪ1
]
13a
14a
13b
14b
13c
14c
62 Ϯ 6[b]
54 Ϯ 4
54 Ϯ 3
53 Ϯ 5
63 Ϯ 2
62 Ϯ 2
Ϫ(171 Ϯ 16)
Ϫ(195 Ϯ 12)
Ϫ(194 Ϯ 9)
Ϫ(196 Ϯ 15)
Ϫ(167 Ϯ 6)
Ϫ(171 Ϯ 5)
Ϫ(5 Ϯ 1)
Ϫ(4 Ϯ 1)
Ϫ(7 Ϯ 1)
Ϫ(7 Ϯ 1)
Ϫ(12 Ϯ 2)
Ϫ(11 Ϯ 2)
[a]
[b]
1500 bar.
The reported uncertainties in ∆H‡, ∆S‡ and ∆V‡0
Figure 4. Pressure dependence of the experimental rate coefficients
k for the Heck reaction of iodobenzene (12) with the acrylates
13a,b,c to form the corresponding cinnamates 14a,b,c at 70 °C.
Only the rate coefficients of the reactand are shown
refer to standard deviations.
Table 2 summarizes the values of ∆H‡ and ∆S‡ for the
Heck reactions of 12 with 13a,b,c at a constant pressure of
1500 bar and their corresponding values of ∆V‡0 at ambient
pressure, which are each derived from the analysis of both
the corresponding acrylate absorbances and the cinnamate
absorbances. The reported uncertainties in ∆H‡, ∆S‡ and
∆V‡0 refer to standard deviations.
The relationships between lnk and p are not perfectly lin-
ear. Thus, activation volumes ∆V‡0 were determined from
the slope of the curves of lnk vs. p at ambient pressure. The
data in the curves of ln k vs. p were smoothed according to
the procedures described by Asano et al. [Equation (3) and
Equation (5)].[7] The activation volumes extrapolated to am-
bient pressure, ∆V‡0, are found from Equation (4) and Equa-
tion (6), respectively. The value of k0, which refers to the
rate coefficient at ambient pressure, was derived from a
parabolic fit of the data in the curves of lnk vs. p.
Figure 5. Temperature dependence of the experimental rate coef-
ficients k for the Heck reaction of iodobenzene (12) with the acry-
lates 13a,b,c to form the corresponding cinnamates 14a,b,c at 1.5
kbar. Only the rate coefficients of the reactand are shown
ln (k/k0) ϭ a1p ϩ [b1p/(1 ϩ c1p)]
∆V‡0 ϭ Ϫ(a1 ϩ b1) RT
(3)
(4)
(5)
(6)
To investigate the effect of the leaving group on the kine-
tics of the Heck reaction in the condensed phase, we studied
the kinetics of the reactions of aryl iodide 15a and aryl
triflate 15b with methyl acrylate (13a), to yield the cinna-
mate 16 (Scheme 3), in the pressure range from 250 to
3000 bar at 100 °C, and in the temperature range from 80
to 120 °C at 1500 bar. The concentrations of reactands, the
ln (k/k0) ϭ a2p ϩ b2ln (1 ϩ c2p)
∆V‡0 ϭ Ϫ(a2 ϩ b2c2) RT
The pressure and temperature dependences of the rate solvent and the catalyst system were identical for the three
coefficients are plotted in Figure 4 and Figure 5, respec- reaction systems. Thus, any changes in the determined rate
tively. The rate coefficients for methyl acrylate (13a) are parameters must be due to differences in the type of leaving
smaller than for ethyl acrylate (13b) over the entire exper- group. Unfortunately, because of insufficient reactivity of
imental pressure range. The highest rates were measured the aryl bromide, monitoring the Heck reaction of p-nitro-
with tert-butyl acrylate (13c). The activation volumes, de- phenyl bromide with methyl acrylate with the classical cata-
rived according to Equations (3) to (6), follow the order lyst system did not give satisfactory kinetic data in the same
2378
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2375Ϫ2382

The Effect of High Pressure on the Heck Reaction
FULL PAPER
range of pressure and temperature as was used for the iod-
ides and the triflate.
Scheme 3. Variation of the leaving group in the Heck reaction of
aryl derivatives with methyl acrylate (13a).
The experimental pseudo-first-order rate coefficients k of
the Heck reactions in Scheme 3 were obtained by the
method described above. A full report on the experimental
data (values of k for the Heck reactions of methyl acrylate
with p-nitrophenyl iodide and triflate) is provided in the
Supporting Information.
Figure 7. Temperature dependence of the experimental rate coef-
ficients k for the Heck reaction at 1.5 kbar of the aryl iodide 15a
and triflate 15b with methyl acrylate (13a) to form the correspond-
ing cinnamate 16. Only the rate coefficients of the reactand are
shown
Rate coefficients, measured as a function of pressure, are
plotted in Figure 6.
Table 3. Activation parameters and experimental rate coefficients
k at 100 °C and 1.5 kbar for the Heck reaction of the aryl iodide
15a and the triflate 15b with methyl acrylate (13a) to form the
corresponding cinnamate 16
derived from
∆H‡[a]
∆S‡[a]
∆V‡0
[kJ·molϪ1
]
[J·molϪ1·KϪ1
]
[cm3·molϪ1
]
15a/13a
16
15b/13a
16
74 Ϯ 2[b]
75 Ϯ 1
91 Ϯ 3
54 Ϯ 4
Ϫ(168 Ϯ 4)
Ϫ(164 Ϯ 2)
Ϫ(111 Ϯ 7)
Ϫ(208 Ϯ 12)
Ϫ(9 Ϯ 1)
Ϫ(10 Ϯ 2)
Ϫ(37 Ϯ 6)
Ϫ(32 Ϯ 5)
[a]
[b]
1500 bar.
The reported uncertainties in ∆H‡, ∆S‡ and ∆V‡0
refer to standard deviations.
Figure 6. Pressure dependence of the experimental rate coefficients
k for the Heck reaction of the aryl iodide 15a and triflate 15b with
methyl acrylate (13a) to the corresponding cinnamate 16 at 100 °C.
Only the rate coefficients of the reactand are shown
and activation parameters, it was surprising to find that the
pressure- and temperature-dependent rate coefficients for
the coupling of the acrylates 13a,b,c with iodobenzene (12)
to the corresponding cinnamates 14a,b,c (Scheme 2) follow
the order tert-butyl (13c) Ͼ ethyl (13b) Ͼ methyl acrylate
(13a). On average, in the Heck reaction with iodobenzene,
tert-butyl acrylate reacts 1.4-times faster than methyl acry-
late and 1.3-times faster than ethyl acrylate. Apparently, an
electronic effect must override a potential retarding effect
of the larger steric bulk of the alkyl acrylate. Qualitatively,
the electronic difference of the three acrylates must stem
from the difference in the σ-electron-donating ability of a
methyl, an ethyl and a tert-butyl group in the ester moiety.
In fact, the HOMO of an alkyl acrylate is influenced by the
nature of the alkyl group, as is shown by computations[20]
at the B3LYP/6Ϫ31ϩG(d,p) level of theory for the optimiz-
ation of geometries and at the MP2/aug-cc-pvdz level
of theory for the energy single points and photoelectron
spectroscopic (PES) measurements of the lowest
ionization enegies (IE) for the three acrylates 13a,b,c:
At 250 bar, the triflate 15b reacts at the same rate as the
iodide 15a, but above 500 bar the aryl triflate 15b is the
faster-reacting substrate.
The activation volumes were derived according to Equa-
tions (3)Ϫ(6) and follow the order ∆V‡ (15b) Ͻ ∆V‡ (15a),
i.e., the largest negative value of ∆V‡ is derived from the
triflate 15b.
The temperature dependence of the experimental rate co-
efficients at 1500 bar is plotted in Figure 7. The slowest-
reacting substrate is the iodide 15a.
The activation enthalpies and entropies, derived accord-
ing to Equations (1) and (2), as well as the activation vol-
umes, derived according to Equations (3) to (6), are com-
piled in Table 3.
Discussion
IE(13a)calcd.
ϭ
Ϫ10.73 eV (IEexp.
Ϫ10.68 eV (IEexp.
ive alkyl acrylate and their effect on the rate coefficients Ϫ10.66 eV[22]), and IE(13c)calcd. ϭ Ϫ10.59 eV. On the basis
ϭ
Ϫ10.72,[9]
When focussing on the steric requirements of the respect- Ϫ10.77 eV[10]), IE(13b)calcd.
ϭ
ϭ
Eur. J. Org. Chem. 2003, 2375Ϫ2382
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2379

´
M. Buback, T. Perkovic, S. Redlich, A. de Meijere
FULL PAPER
of Koopmans’ theorem, these decreasing values of IE Ϫ12 cm3·molϪ1, respectively) and with the variation of the
correspond to an increase of the HOMO energies, and thus leaving groups in 15a and 15b (from Ϫ9 to Ϫ37 cm3·molϪ1
the reaction rates of the acrylates 13a,b,c correlate with follow the same trend as previously observed for other reac-
their increasing HOMO energies.
tions under high pressure.[1,24] The more sterically con-
)
Under high pressure, a reaction via a dipolar transition gested a reacting system is, the higher is the rate-accelerat-
structure is more accelerated than a reaction proceeding via ing influence of high pressure going along with a more
an apolar transition structure because of the so-called elec- negative activation volume.[2a,2b] The difference between
trostriction effect.[2] Thus, a reaction of a positively charged tert-butyl acrylate (13c) and ethyl acrylate (13b) (∆∆V‡ ϭ
palladium species that certainly would prefer an electron- 5 cm3·molϪ1) is larger than that between methyl acrylate
rich alkene (with a high HOMO) should be enhanced by (13a) and ethyl acrylate (13b) (∆∆V‡ ϭ 2 cm3·molϪ1) be-
high pressure more than a reaction of a neutral palladium cause of the significantly larger increase in bulk on going
complex that would prefer a less-electron-rich double from an ethyl to a tert-butyl group. The same effect is found
bond.[6f] Thus, the observed pressure dependence of the rate for the leaving group variation, where 15b, with the eight-
coefficients for the coupling reactions of the three acrylates atom triflate leaving group, yields the more negative acti-
13a,b,c indicates that the coordination and insertion steps vation volume than the iodide 15b (∆∆V‡ ϭ 28 cm3·molϪ1),
(steps E and F in Scheme 1) under high-pressure conditions which correlates with the steric bulk. The activation volume
take place more favorably via an [R1PdII(PPh3)2ϩ] complex for the Heck reaction of iodobenzene and 2,3-dihydrofuran
(8) than via an [R1PdII(OAc)(PPh3)2] complex (7) as de- with Pd(OAc)2/PPh3 as catalyst system and under pressures
scribed for such reactions at ambient pressure.[13]
The changes in the experimental first-order rate coef- was determined by Reiser et al. to be ∆V‡ ϭ
ficients for the variation of the leaving group in the arenes Ϫ12 Ϯ 2 cm3·molϪ1 [11]
This value fits well into the order
up to 8000 bar and at a constant temperature of 30 °C
.
15a and 15b under pressures above 500 bar follow the order of magnitude of the activation volumes found in this study
iodide 15a Ͻ triflate 15b. This result is a surprising one, for the reactions of iodobenzene with three different alkyl
because it is generally assumed that the rate-determining acrylates.
step in the catalytic cycle of the Heck reaction is usually
It is noteworthy that the activation volumes were found
the oxidative addition and, therefore, one would predict the in a different order and to be much more negative when the
opposite order of reaction rates. Thus, the rate-determining palladacycle introduced by Hermann and Beller et al.[25]
step under high-pressure conditions cannot be the oxidative was used instead of the classical catalyst system. The ob-
addition as has already been pointed out by Reiser et al.[11] served activation volumes for the coupling of iodobenzene
More recently, Amatore and Jutand,[13] as well as de Vries (12) and bromobenzene with methyl acrylate (13a) were
and van Leeuwen et al.,[23] have presented experimental evi- ∆V‡ ϭ Ϫ(21 Ϯ 3) and Ϫ(27 Ϯ 4) cm3·molϪ1, respectively,
dence that, even at ambient pressure, the oxidative addition while the reaction rates did not differ very much for both
does not need to be the rate-determining step in the overall catalyst systems in the pressure range from 500 to
catalytic cycle of the Heck reaction.
3000 bar.[26] The significantly different values of the acti-
The immediately ensuing question then would be: Which vation volume demonstrate that the rates of reaction of
of the steps is the rate determining one? As already pointed bromobenzene depend more strongly on pressure than do
out, the equilibrium between [R1PdII(OAc)(PPh3)2] (7) and those of iodobenzene (12) and, hence, bromobenzene can
[R1PdII(PPh3)2ϩ] (8) (Scheme 1) is shifted towards 8 under be made to react more efficiently by increasing pressure.[6a]
high pressure. Here, the effect of the counterion has to be
No significant trend is observed for the activation en-
taken into account. As Amatore and Jutand have demon- thalpies and entropies (∆H‡, ∆S‡), neither upon variation
strated, the catalytic action cannot be affected by of the alkyl acrylate nor of the leaving group in the aryl
R1PdIIX(PPh3)2 (with X ϭ I or Br), but the only cata- derivative; in fact, the values are nearly the same within
lytically active species in such systems must be experimental uncertainty. This finding confirms the expec-
[R1PdII(OAc)(PPh3)2] (7).[13] Thus, the reaction producing tation that these activation parameters should depend
iodide ions must be more disfavored under high pressure, mainly on the catalyst system used for these reactions,
which enhances the formation of catalytically inactive rather than on the substrates. Amatore and Jutand found
R1PdIII(PPh3)2. The higher rate coefficients for 15b with the values in the same order of magnitude for the catalyst sys-
triflate leaving group are due to the fact that triflates always tem Pd(OAc)2/PPh3, namely ∆H‡ ϭ 65 kJ·molϪ1 and
ϩ
react via the cationic palladium species [R1PdII(PPh3)2
]
∆S‡ ϭ Ϫ94 J·mol·KϪ1 [27]
.
(8).[13] These experimental results suggest that the rate-de-
termining step in the catalytic cycle of the Heck reaction is
connected with the equilibrium between [R1PdII(OAc)-
(PPh3)2] (7) and [R1PdII(PPh3)2ϩ] (8) and should be the co-
ordination (Scheme 1, step E) or insertion of the acrylate
Conclusion
This study on the influence of high pressure on the Heck
(Scheme 1, step F), respectively, as is corroborated by the reaction, by applying on-line FT-IR spectroscopy, revealed
results obtained with the variation of the acrylates.
that the activation volumes ∆V‡ correlate with the degrees
The changes in the activation volumes with the variation of steric congestion for the variation of the alkyl group in
of the acrylates 13a,b,c (from Ϫ5 and Ϫ7 to the acrylates 13a,b,c and of the leaving groups in the aryl
2380
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2375Ϫ2382

The Effect of High Pressure on the Heck Reaction
FULL PAPER
GmbH, Degussa-Hüls AG as well as Hoechst AG for generous
gifts of chemicals, to Dr. Hans Peter Vögele for his advice in carry-
ing out high-pressure FT spectroscopic experiments and in analyz-
ing the spectroscopic data, and to Dr. Burkhard Knieriem for his
careful reading of the final manuscript.
derivative 15a,b, i.e., the more highly congested systems give
values that are more negative, while activation enthalpies
and entropies ∆H‡ and ∆S‡ are virtually independent of
the reactands.
The rate coefficients for the coupling of methyl acrylate
(13a), ethyl acrylate (13b) and tert-butyl acrylate (13c) with
iodobenzene (12) to form the corresponding cinnamates
14a,b,c (Scheme 2) follow the order k(13c) Ͼ k(13b) Ͼ
k(13a), while those for the coupling of the p-nitrophenyl
iodide (15a) and p-nitrophenyl triflate (15b) with methyl
acrylate (13a) to the corresponding p-nitrocinnamate (16)
(Scheme 3) follow the order k(15b) Ͼ k(15a). These orders
are consistent with the notion that the coordination and
insertion steps of the acrylate (Scheme 1, steps E and F) are
rate determining. Thus, under high-pressure conditions, the
rate-determining step of the overall catalytic cycle of the
Heck reaction is not the oxidative addition, which is gener-
ally assumed to be the rate-determining step for such reac-
tions at ambient pressure.
[1] [1a]
K. Matsumoto, A. Sera, T. Uchida, Synthesis 1985, 1Ϫ26.
[1b]
[1c]
[1d]
K. Matsumoto, A. Sera, Synthesis 1985, 999Ϫ1027.
T.
Asano, W. J. le Noble, Chem. Rev. 1978, 78, 407Ϫ489.
R.
van Eldik, T. Asano, W. J. le Noble, Chem. Rev. 1989, 89,
[1e]
549Ϫ688.
A. Drljaca, C. D. Hubbard, R. van Eldik, T. As-
ano, M. V. Basilevsky, W. J. le Noble, Chem. Rev. 1998, 98,
2167Ϫ2289.
[2] [2a]
G. Jenner, Angew. Chem. Int. Ed. Engl. 1975, 14, 137Ϫ155;
Angew. Chem. 1975, 87, 186Ϫ194. [2b] W. J. le Noble, H. Kelm,
Angew. Chem. Int. Ed. Engl. 1980, 19, 841Ϫ858; Angew. Chem.
1980, 92, 887Ϫ904. [2c] W. J. l e N o b l e, High Pressure Chemistry
and Biochemistry (Eds.: R. van Eldik, J. Jonas), D. Reidel Pub-
[2d]
lishing Company, Dordrecht, 1986, Vol. 197, pp. 279Ϫ293.
W. J. l e N o b l e, High Pressure Chemistry and Biochemistry (Eds.:
R. van Eldik, J. Jonas), D. Reidel Publishing Company, Dord-
recht, 1986, Vol. 197, pp. 295Ϫ310.
[3]
[4]
[5]
N. S. Isaacs, Tetrahedron 1991, 47, 8463Ϫ8497.
O. Reiser, Topics Catalysis 1998, 5, 105Ϫ112.
B. M. Trost, J. R. Parquette, A. L. Marquart, J. Am. Chem.
Soc. 1995, 117, 3284Ϫ3285.
Experimental Section
[6] [6a]
Materials: Solvents were dried and distilled prior to use and de-
gassed by applying several freeze-pump and thaw cycles. TLC was
performed on precoated silica gel SIL G/UV254 plates
A. de Meijere, S. Bräse, J. Organomet. Chem. 1999, 576,
[6b]
88Ϫ110.
S. Bräse, A. de Meijere, Metal-catalyzed Cross-
coupling Reactions (Eds.: F. Diederich, P. Stang), Wiley-VCH,
Weinheim, 1998, pp. 99Ϫ166. [6c] A. de Meijere, S. Bräse, Tran-
(MachereyϪNagel
& Co.), and silica gel (0.066Ϫ0.200 mm,
sition Metal-Catalyzed Reactions (Eds.: S. G. Davies, S.-I. Mur-
70Ϫ230 mesh ASTM) (Merck) was used for column chromatogra-
phy.
[6d]
ahashi), Blackwell Science, Oxford, 1998, pp. 99Ϫ131.
A.
de Meijere, F. E. Meyer, Angew. Chem. Int. Ed. Engl. 1994, 33,
[6e]
2379Ϫ2411; Angew. Chem. 1994, 106, 2473Ϫ2506.
R. F.
Pd(OAc)2, PPh3, NEt3, iodobenzene (12), methyl acrylate (13a),
ethyl acrylate (13b), tert-butyl acrylate (13c), and 4-nitrophenyl iod-
ide (15a) are commercially available (Aldrich and Fluka).
Heck, Comprehensive Organic Synthesis (Eds.: B. M. Trost, I.
Fleming), Pergamon Press, Oxford, 1991, Vol. 4, pp. 833Ϫ863.
[6g]
[6f]W. Cabri, I. Candiani, Acc. Chem. Res. 1995, 28, 2Ϫ7.
G. T. Crisp, Chem. Soc. Rev. 1998, 27, 427Ϫ436.
S. Hillers, O. Reiser, Tetrahedron Lett. 1993, 34, 5265Ϫ5268.
S. Hillers, S. Sartori, O. Reiser, J. Am. Chem. Soc. 1996, 118,
2087Ϫ2088.
4-Nitrophenyl trifluoromethanesulfonate (15b) was prepared ac-
[7]
[8]
cording to a published procedure.[28]
Instrumentation: The reactions were performed in an optical high-
pressure cell equipped with windows made from polycrystalline sili-
con. The samples of the reaction mixture, which were all prepared
at the same time and subsequently deep-frozen in liquid nitrogen
prior to use, were contained in sealed thin-walled Teflon bags as
internal cells, one of which was positioned at each time between
the windows of the high-pressure cell. The optical cell, the other
high-pressure equipment, the procedures of preparing, filling and
sealing the internal cells, and the data treatment have been de-
scribed in detail elsewhere.[14] The spectra were recorded with a
Fourier-transform IR spectrometer (BRUKER IFS 66 and IFS 88).
The decadic absorbance should not exceed A ϭ 1.8 to stay within
the linear range of the DTGS detector used in the spectrometer.
This limitation must be taken into account to ensure reliable quan-
titative spectroscopic analysis. Pressures were determined to be
within Ϯ 10 bar. The uncertainties in temperature measurements
were within Ϯ 0.5 K.
[9]
K. Voigt, U. Schick, F. E. Meyer, A. de Meijere, Synlett 1994,
189Ϫ190.
[10] [10a] L. F. Tietze, O. Burkhardt, M. Henrich, Liebigs Ann. Recl.
1997, 5, 887Ϫ891. [10b] L. F. Tietze, O. Burkhardt, M. Henrich,
Liebigs Ann. Recl. 1997, 7, 1407Ϫ1413.
[11]
S. Hillers, O. Reiser, Chem. Commun. 1996, 2197Ϫ2198.
T. Sugihara, M. Takebayashi, C. Kaneko, Tetrahedron Lett.
[12]
1995, 36, 5547Ϫ5550.
[13] [13a]
C. Amatore, A. Jutand, Acc. Chem. Res. 2000, 33,
[13b]
314Ϫ321.
C. Amatore, A. Jutand, J. Organomet. Chem.
1999, 576, 254Ϫ278, and references cited therein.
[14]
[15]
M. Buback, C. Hinton, High Pressure Techniques in Chemistry
and Physics: A Practical Approach (Eds.: W. B. Holzapfel, N.
S. Isaacs), Oxford University Press, Oxford, 1996, p. 51.
The concentrations of the components in a typical reaction
mixture were 12 (1.28
ϫ ϫ
10Ϫ1 mol·kgϪ1), 13c (3.67
10Ϫ3 mol·kgϪ1), Pd[OAc]2 (1.86
ϫ
10Ϫ4 mol·kgϪ1), PPh3
(4.20 ϫ 10Ϫ4 mol·kgϪ1) and NEt3 (4.23 ϫ 10Ϫ3 mol·kgϪ1). It
is important to note that the concentrations have to be calcu-
lated in pressure-independent concentration units, i.e.
mol·kgϪ1, instead of pressure-dependent units like mol·LϪ1
;
see: S. D. Hamann, W. J. le Noble, J. Chem. Educ. 1984, 61,
658Ϫ660.
Acknowledgments
This work was supported by the Deutsche Forschungsgemeinschaft
(Graduiertenkolleg ‘‘Kinetik und Selektivität chemischer Prozesse
in verdichteter fluider Phase’’) and the Fonds der Chemischen In-
dustrie. T. P. is grateful to the DFG-Graduiertenkolleg for a fellow-
ship. The authors are indebted to BASF AG, Bayer AG, Chemetall
[16]
[17]
W. H. Press, B. P. Flammery, S. A. Teukolsky, W. T. Vetterling,
Numerical Recipes, Cambridge University Press, Cambridge,
1989.
The quality of the fitted first-order rate law function to the
experimental values was tested by a χ2 test mode with a confi-
Eur. J. Org. Chem. 2003, 2375Ϫ2382
www.eurjoc.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2381

´
M. Buback, T. Perkovic, S. Redlich, A. de Meijere
FULL PAPER
[24]
dence interval of 95% always yielding correlation coefficients
better than 0.998.
The uncertainty in the rate coefficients is in between 10 to 15%
It has to be mentioned that the interpretation of the activation
volume in fluid phases is still under critical discussion. See: [24a]
[18]
R. A. Firestone, G. M. Smith, Chem. Ber. 1989, 122,
[24b]
as calculated for the standard deviation with the
1089Ϫ1094.
9082Ϫ9086.
R. Schmidt, J. Phys. Chem. A 1998, 102,
LevenbergϪMarquardt algorithm.
T. Asano, T. Okada, J. Phys. Chem. 1984, 88, 238Ϫ243.
The authors are grateful to Alexander Wittkopp, Institut für
[25] [25a] W. A. Herrmann, C. Broßmer, K. Öfele, C.-P. Reisinger, T.
Priermeier, M. Beller, H. Fischer, Angew. Chem. Int. Ed. Engl.
1995, 34, 1844Ϫ1846; Angew. Chem. 1995, 107, 1989Ϫ1992.
[19]
[20]
Organische Chemie der Georg-August-Universität Göttingen,
[25b]
for calculating these values.
M. Beller, H. Fischer, W. A. Herrmann, K. Öfele, C.
[21] [21a]
R. Sustmann, H. Trill, Tetrahedron Lett. 1972, 42,
Broßmer, Angew. Chem. Int. Ed. Engl. 1995, 34, 1846Ϫ1849;
4271Ϫ4274. [21b] H. van Dam, A. Oskam, J. Electron Spectrosc.
Relat. Phenom. 1978, 13, 273Ϫ290.
Angew. Chem. 1995, 107, 1992Ϫ1993.
[26]
M. Buback, T. Perkovic, A. de Meijere, unpublished results.
[22]
[27]
´
The authors are grateful to Prof. Dr. P. Rademacher, Institut
C. Amatore, E. Carre, A. Jutand, M. A. M’Barki, Organomet-
für Organische Chemie der GH Essen, for measuring these val-
ues.
G. P. F. van Strijdonck, M. D. K. Boele, P. C. J. Kamer, J.
allics 1995, 14, 1818Ϫ1826.
A. M. Echavarren, J. K. Stille, J. Am. Chem. Soc. 1987, 109,
5478Ϫ5486.
[28]
[23]
G. de Vries, P. W. N. M. van Leeuwen, Eur. J. Inorg. Chem.
1999, 1073Ϫ1076.
Received December 23, 2002
2382
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2003, 2375Ϫ2382