Comparative investigation of the regioselectivity in styrene and α-methylstyrene hydroalkoxycarbonylation as a function of palladium catalyst structure

Article abstract of DOI:10.1016/S0022-328X(99)00225-9

Catalytic pathways of the styrene and α-methyl-styrene hydroalkoxycarbonylation in the presence of Pd(PPh₃)₂Cl₂ and Pd(PPh₃)₂Cl₂/SnCl₂ catalyst precursors have been suggested. As a method, deuterium labelling with EtOD has been applied and it resulted in mixtures of mono- and polydeuterated reaction products, detected and determined by NMR methods. Comparative elucidation of the mechanisms governing these systems does support the assumption that the hydrido route is operative. The different behaviour of the metal-alkyl intermediates accounts for the observed strong influence of catalyst and substrate structure on regioselectivity.

Full text of DOI:10.1016/S0022-328X(99)00225-9

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 586 (1999) 85–93
Comparative investigation of the regioselectivity in styrene and
a-methylstyrene hydroalkoxycarbonylation as a function of
palladium catalyst structure
a
b
b,
Csilla Benedek , Szil a´ rd Tor o¨ s , B a´ lint Heil *
3
a
Research Group for Petrochemistry of the Hungarian Academy of Sciences, PO Box 158, H-8201 Veszpr e´ m, Hungary
b
Department of Organic Chemistry, Uni6ersity of Veszpr e´ m, PO Box 158, H-8201 Veszpr e´ m, Hungary
Received 10 November 1998; received in revised form 22 March 1999
Dedicated to: Professor L a´ szl o´ Mark o´ on the occasion of his 70th birthday, in recognition of his outstanding service to organometallic
chemistry.
Abstract
Catalytic pathways of the styrene and a-methyl-styrene hydroalkoxycarbonylation in the presence of Pd(PPh ) Cl and
3
2
2
Pd(PPh ) Cl /SnCl catalyst precursors have been suggested. As a method, deuterium labelling with EtOD has been applied and
3
2
2
2
it resulted in mixtures of mono- and polydeuterated reaction products, detected and determined by NMR methods. Comparative
elucidation of the mechanisms governing these systems does support the assumption that the hydrido route is operative. The
different behaviour of the metal–alkyl intermediates accounts for the observed strong influence of catalyst and substrate structure
on regioselectivity. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Carbonylation; Palladium; Mechanism; Catalysis
1
. Introduction
insertion, yielding the final product upon nucleophilic
attack of the alkanol. According to the ‘carboalkoxy’
mechanism [8,9], a Pd–acyl species is formed by the
attack of the alcohol on Pd–CO. Insertion of the olefin
into the Pd–COOR bond is followed by protonolysis of
the metal–alkyl s-bond to give the corresponding ester
and regeneration of the active form. These two mecha-
nisms are claimed to be operative depending on the
reaction conditions and substrates [10]. We have shown
The palladium–monophosphine-catalysed
hy-
droalkoxycarbonylation of vinyl aromatic olefins has
attracted great interest due to its importance in obtain-
ing non-steroidal anti-inflammatory drugs [1]. The reac-
tion is known to be characterised by high a-selectivities
[2], i.e. the formation of the branched ester. Adopting
Knifton’s [3] catalytic system, addition of SnCl leads
2
to a prevalence of the linear product. At the same time,
vinylidene olefins such as a-methylstyrene, give the
linear derivative [4] selectively under both reaction con-
ditions. A satisfactory explanation of the regioselectivi-
ties is still to be given in the framework of the two
accepted mechanisms [5] for hydroalkoxycarbonylation
of olefins.
previously [11] that in the Pd(PPh ) Cl /SnCl catalysed
styrene hydroalkoxycarbonylation the first one is
acting.
3 2 2 2
In order to find out the pathways governing the
systems studied and to determine the influence of sub-
strate and catalyst structure on regioselectivities, deu-
terioalkoxycarbonylation of styrene and a-methyl-
styrene has been carried out. In the present paper we
report a detailed investigation of the reaction per-
formed at two temperatures and partial substrate con-
versions. The labelled species determined by NMR
The first one, the ‘hydrido’ route [3,6,7], involves
olefin insertion into the Pd–H bond followed by CO
*
Corresponding author. Tel.: +36-88-422022; fax: +36-88-
1
2
13
(
H-, H-, C-) and MS techniques throw light on the
427492.
E-mail address: heil@almos.vein.hu (B. Heil)
catalytic routes of the reaction.
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00225-9

86
C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
2. Results and discussion
tion calculated previously for the major carbonylated
product by means of H. The data (Table 1) brought us
1
2
.1. Deuterioalkoxycarbonylation experiments
to the conclusion that the total amount of isotope
incorporation increases by increasing the reaction tem-
perature in both systems, the overall deuterium content
Deuterioalkoxycarbonylation of styrene (1) and a-
methylstyrene (2) (Scheme 1) has been carried out at
00 and 130°C and 40 atm of CO with Pd(PPh ) Cl
being greater in the SnCl modified system.
2
1
The results of our investigation show that variable
deuterium incorporation occurs both in the substrate
and in the reaction products when deuterioalkoxycar-
bonylation of styrene is carried out in the presence of
catalyst precursors A or B. We have already reported
[11] the accommodation of these species for the latter
system in terms of the routes described in Scheme 2.
Thus, the formation of the two Pd–alkyl intermediates
n% and b% is reversible. Undergoing b-hydride elimina-
3
2
2
(
A) and Pd(PPh ) Cl /SnCl (B) as catalytic precursors
3
2
2
2
in toluene. The reaction has been stopped in each case
at about 75% substrate conversion determined unam-
biguously together with the ratio of the isomers by
GLC analysis. Purified samples of the reaction products
were obtained by column chromatography on silica gel.
The crude reaction mixtures and the purified samples
1
of isomeric esters were analysed by MS and by H-,
1
1
2
13
H-, C-NMR spectroscopy in order to determine the
tion (controlled by the kinetic deuterium isotope effect
[14]), they give the Pd–hydride p-complexes I and II. In
complex I, the insertion of the olefin into the Pd–H
bond could regenerate the starting linear palladium–
extent and the position of deuterium incorporation in
the molecules.
2.2. Deuterioalkoxycarbonylation of styrene (1)
alkyl n% yielding the corresponding linear 3-deuterated
1
ester (3-d -3) or produce the isomeric branched palla-
1
The rapid and complete identification of the deuter-
dium–alkyl intermediate b%, from which the 2-deuter-
2
ated species present in solution has been accomplished
by the inspection of the H-NMR spectra [12]. Fig. 1
ated branched ester (2-d -4) can be formed.
Analogously, complex II gives rise to the linear palla-
1
2
shows one of these recorded from the reaction mixture
formed at 100°C and 75% conversion in toluene (with
residual EtOD removed) in the presence of
Pd(PPh ) Cl (A) (Fig. 1(a)) and Pd(PPh ) Cl /SnCl
dium–alkyl intermediate n% and so to the linear 2-
2
deuterated ester 2-d -3 or regenerates the branched
1
intermediate b% leading to the branched 3-deuterated
1
ester 3-d -4. The overall process is, therefore, the iso-
3
2
2
3 2
2
2
1
(
B) (Fig. 1(b)) as catalytic precursors. The main groups
merization of the starting branched metal–alkyl inter-
of signals are those arising from deuterio-styrenes,
ethyl-deuterio-3-phenylpropanoate (3) and ethyl-deu-
terio-2-phenylpropanoate (4). In Fig. 1(a) the spectrum
shows only the signals corresponding to the methyl
group of the branched product (which is obtained in
mediate b% to the linear isomeric intermediate n% and,
1
2
conversely, the isomerization of n% to b%. On the other
1
2
hand, p-complexes I and II can undergo an intermolec-
ular exchange process with unlabelled styrene, produc-
ing
monodeuteriostyrenes
2-d -1
and
1-d -1,
1
1
9
7.4% regioselectivity) at 1.53 ppm and those corre-
respectively and the unlabelled p-complex III. This last
one can generate the corresponding linear and
branched esters (3 and 4) through the linear (n) or
branched (b) unlabelled metal–alkyls. Once labelled,
styrene species may undergo further reactions resulting
in di- and oligo-deuterated derivatives.
sponding to the two geometric isomers (E)- and (Z)-1-
deuterio-vinylbenzene (at 5.28 and 5.75 ppm,
respectively). Isotope incorporations can be observed
neither in the positions situated a to the phenyl ring nor
in the aromatic region.
The spectrum of the reaction mixture obtained with
system B is more complex. The signals at 2.50 ppm and
Surprisingly, when the reaction was carried out with
catalyst A, no deuterium incorporation could be de-
tected on the carbon atom bound to the phenyl ring
both in the recovered substrate and the branched
product even at temperatures as high as 130°C. This
2.90 ppm can be ascribed to the methylene groups of
the linear ester (formed in 74.5% selectivity) situated a
and b to the ethoxycarbonyl moiety, respectively. The
resonance at 3.66 ppm corresponds to the methine
group of the branched product while that at 6.70 ppm
corresponds to 2-deuteriovinylbenzene.
The MS analysis [13] of the crude mixtures proved
the presence of mono-, di- and trideuterated styrenes as
well as mono-, and polydeuterated esters, these latter
13
ones also being detected by C-NMR [11]. The deu-
terium content at each of the labelled carbon atoms has
2
been determined from the H-NMR spectra (after elim-
inating the residual d-ethanol and part of the toluene)
on the basis of composition (GLC) and the incorpora-
Scheme 1. Deuterioalkoxycarbonylation reaction of styrene and a-
methylstyrene.

C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
87
fact indicates that, under the conditions given, the
reactions leading to palladium–alkyl intermediates n1%
and n%, do not take place.
2
Some hypotheses have been advanced to explain
the marked difference in regioselectivity determined
by the presence of SnCl . According to Knifton [3],
2
the moderately basic PPh raises the hydridic charac-
3
ter of the active species, thus promoting both anti-
Markownikov addition to the olefin and formation of
a linear palladium–alkyl. On the other hand, compet-
ing electronic factors [15–17] would result in the pref-
erential formation of the secondary metal–alkyl
intermediate stabilised by the p-benzylic interaction of
the phenyl group with Pd. It seems likely that the
−
SnCl3 ligand provides a particularly sterically hin-
dered catalyst [3] in which steric constraints are pow-
erful enough to overcome the unfavourable electronic
effects.
It would be interesting to consider the ‘kinetic ef-
−
fect’ of SnCl3 described for the closely related Pt-
containing intermediates [18] where fast b-hydride
elimination of olefin from a secondary s-complex,
possessing a Pt–Sn bond, takes place. In our systems
the lifetime of the analogous branched Pd–Sn species
may be too short to allow CO insertion to compete
with hydride elimination.
2.3. Deuterioalkoxycarbonylation of h-methylstyrene
The reaction mixtures of a-methylstyrene were also
2
investigated by H-NMR spectroscopy; the identifica-
tion of the deuterated species made possible by the
1
2
similarity of the chemical shifts of the H and
H
2
nuclei. In Fig. 2(a) the H-NMR spectrum of a reac-
tion mixture synthesised with the catalyst A is re-
ported. The signals of the linear ester (formed in
8
2.7% regioselectivity) are centred at 1.30 ppm (corre-
sponding to the methyl group), at 2.55 ppm (corre-
sponding to the methylene group) and at 3.27 ppm
(
1
corresponding to the methine group). The signal at
.55 ppm denotes D incorporation at the methyl
groups of the branched ester. The two resonances of
the olefinic region (5.18 and 5.55 ppm) are attributed
to the E- and Z-1-deuterio-2-methyl vinylbenzene,
while the D incorporation in the methyl group of the
unconverted substrate is detectable at 2.15 ppm.
The same signals are also present in the spectrum
of the reaction mixture formed with catalyst B (Fig.
2
(b)). However, the extent of isotope incorporation is
rather different.
MS analyses of the reaction mixtures proved again
the presence of mono-, di- and trideuterated olefins
together with mono-, and polydeuterated esters (de-
13
tected also by C-NMR, see Section 4). The deu-
terium content has been determined similarly as
described for styrene. The data acquired (Table 2)

88
C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
Fig. 1. (a) 2H-NMR spectrum (46 MHz, 20°C, toluene) of the reaction products obtained by deuterioalkoxycarbonylation of styrene at 100°C,
2
7
5% conversion with catalyst A. (b) H-NMR spectrum (46 MHz, 20°C, toluene) of the reaction products obtained by deuterioalkoxycarbonyla-
tion of styrene at 100°C, 75% conversion with catalyst B.
show the same temperature dependence of isotope in-
corporation, while the total deuterium content is
p-complex IV% with a Pd–D species, a primary (p%) and
1
a tertiary (t%) palladium alkyl may be formed. Interme-
1
greater in the absence of SnCl . The pathways leading
diate p% can lead only to the corresponding labelled
2
1
to these species are presented in Scheme 3.
According to this, when a-methylstyrene (2) gives the
linear ester 3-d -5 because no b-hydride elimination is
1
possible. On the other hand, according to our data, CO

C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
89
insertion practically does not occur at the tertiary
metal–alkyl as this would produce the branched ester
which is formed in very low amounts. However, it
undergoes b-hydride elimination resulting in the two
p-complexes V and VI. In a subsequent addition V
leads to p%, and so the linear product 2-d -5, while VI
CO insertion leading to the corresponding branched
ester is probably sterically hindered. To the best of
our knowledge, the existence of this intermediate has
not been evidenced yet in hydroesterification of
olefinic substrates, although it has been shown to be
present in the related rhodium-catalysed hydroformy-
lation reaction [19].
2
1
gives rise to p% and the derived ester 4-d -5. After
3
1
intermolecular exchange processes with the starting
olefin, in addition to complex IV which gives unla-
belled 5, 1-d -2 and 3-d -2 are formed.
3. Conclusions
1
1
The large numbers of mono- and polydeuterated
species supports, analogous to the reaction of styrene,
the dominance of the hydrido route. Moreover, the
presence of 1- and 3-labelled a-methylstyrene, to-
gether with the 2- and 4-deuterated linear esters (de-
tected already at 20% conversion) clearly evidences
the formation of the tertiary alkyl–palladium interme-
diate (t%). The ratio D(C )/[D(C )+D(C )] (see Table
The composition of the rather complex reaction
mixtures was successfully elucidated by MS analysis
and combined NMR methods. The labelled deriva-
tives detected can be deduced in terms of the reaction
pathways described and support the ‘hydrido’ mecha-
nism operating in the systems studied.
At the same time, the results obtained in our ex-
periments throw light on the different behaviour of
alkyl–palladium intermediates determined by catalyst
and substrate structure. Transforming styrene with
1
b
a
g
2) is less than one in each of the reactions which
indicates that the tertiary species is favoured over the
primary one, but the first undergoes mainly b-hydride
elimination followed by intermolecular exchange with
unlabelled olefin or consecutive isomerization. The
Pd(PPh ) Cl /SnCl
as precursor ethyl-3-phenyl-
propanoate is the major product. The formation of
3
2
2
2
Scheme 2. Proposed pathways for deuteration of styrene and the derived deuterated isomeric esters.

90
C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
both linear and branched Pd–alkyls is reversible and,
as a consequence, D-labelling occurs at all carbon
atoms of the side-chains of substrate and esters.
In the absence of SnCl , when ethyl-2-phenyl-
2
propanoate is formed regioselectively, no isotope was
found in the positions adjacent to the phenyl ring.
So, the routes leading to the linear metal–alkyls must
be blocked.
Using a-methylstyrene as starting material, in both
catalytic systems ethyl-3-phenylbutanoate was ob-
tained predominantly and no significant difference in
the deuterium distribution could be detected. For the
first time, evidence has been given for the existence of
the tertiary alkyl–palladium species. This intermedi-
ate, although favoured over the primary one, in its
formation undergoes overwhelmingly b-hydride elimi-
nation, since releasing of the derived branched ester is
probably sterically hindered. These findings reveal the
different behaviour of primary, secondary and tertiary
alkyl–palladium intermediates under hydroalkoxycar-
bonylation conditions.
4
. Experimental
Toluene was distilled from sodium under an argon
atmosphere. d-Ethanol was purchased from Fluka.
The compound Pd(PPh ) Cl was prepared as re-
3
2
2
ported in the literature [20]. Anhydrous SnCl2 was
obtained by allowing SnCl ·2H O to react with acetic
2
2
anhydride and washing with ether. GLC analyses
were performed on a HP 5830 gas chromatograph
equipped with SPB-1 columns (30 m×1 mm film
depth) and a flame ionization detector, helium was
used as the carrier gas. Column chromatography was
carried out on a silica gel column using hexane, ben-
zene, chloroform, dichloromethane and acetone as
eluents.
Mass spectra were measured with a VG ZAB-SEQ
instrument at 10–12 eV ionization connected to a
HP-5890A gas chromatograph with CP-Sil-8 CB
columns (25 m×0.25 mm film depth) using helium as
carrier gas.
1
2
13
The H-, H- and C-NMR spectra were recorded
on a Varian Unity 300 spectrometer, operating at
3
00, 46 and 75.4 MHz, respectively. Chemical shifts
2
of the H-NMR measurements were determined by
reference to CDCl as external standard.
3
4.1. Deuterioalkoxycarbonylation of styrene and h-
methylstyrene: general procedure
The catalyst (75 mg, 0.11 mmol Pd(PPh ) Cl (A)
3
2
2
or 75 mg, 0.11 mmol Pd(PPh ) Cl ; 55 mg, 0.29
3
2
2

C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
91
2
Fig. 2. (a) H-NMR spectrum (46 MHz, 20°C, toluene) of the reaction products obtained by deuterioalkoxycarbonylation of a-methylstyrene at
2
1
00°C, 75% conversion with catalyst A. (b) H-NMR spectrum (46 MHz, 20°C, toluene) of the reaction products obtained by deuterioalkoxycar-
bonylation of a-methylstyrene at 100°C, 75% conversion with catalyst B.
mmol anhydrous SnCl ; 50.1 mg, 0.19 mmol PPh (B))
standard. d-Ethanol and the solvent were removed un-
2
3
2
1
was introduced to a 50 ml stainless steel autoclave.
After sealing, a solution of olefin (0.37 ml, 3.2 mmol
styrene or 0.42 ml, 3.2 mmol a-methylstyrene), d-
ethanol (1.7 ml, 28.9 mmol) in toluene (10 ml) was
introduced by suction. Then, carbon monoxide was
charged (40 atm at room temperature) and the mixture
was magnetically stirred at the reaction temperature
until the conversion desired. After cooling, the reaction
mixture was siphoned out, and GLC was used to
determine the composition with toluene as internal
der vacuum before H-, H-NMR and MS analyses.
Isomeric esters were separated by column chromatogra-
phy and studied further.
13
1
4.2. C{ H}-NMR data for labelled h-methylstyrene
deri6ati6es 5 and 6
2
2
2
4.2.1. Ethyl-3-phenyl-[2- H, 3- H, 4- H]-butanoate (5)
NMR l 13.98 (CH –CH ), 36.37 (s, CH–CH3),
H–CH D, Z =4.7 Hz), 36.24 (s, CH–
6
6
2
3
2
36.32 (s, C
6
6
2

92
C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
Scheme 3. Proposed pathways for deuteration of a-methylstyrene and the derived deuterated isomeric esters.
CHD ), 36.17 (s, C
6
H–CD ); 42.90 (s, C
6
H –CH), 42.80
References
2
3
2
2
1
(
s, C
6
H –CD, Z =7.8 Hz), 42.56 (t, C
6
HD–CH, Z =
HD–CD, Z =7.1,
2
1
CD
2
[
[
[
1] H.S. Yun, K.H. Lee, J.S. Lee, J. Mol. Cat. A.: Chem. 95 (1995)
11.
2] K. Bittler, N. Kutepow, D. Neubauer, Angew. Chem. Int. Ed.
Engl. 80 (1968) 352.
1
J
1
9.7, J =19.7 Hz), 42.46 (t, C
6
1
CD
=19.7 Hz), 60.20 ( C6 H –CH ), 125.45, 126.27,
2 3
26.54, 128.30, 128.40 (phenyl ring).
3] J.F. Knifton, J. Org. Chem. 41 (1976) 2885.
2
2
4
.2.2. Ethyl-2-([ H]-methyl)-2-phenyl-[3- H]-propan-
[4] G. Consiglio, M. Marchetti, Chimia 30 (1976) 26.
[5] J. Tsuji, Palladium Reagents and Catalysts, Wiley, New York,
oate (6)
NMR l 14.10 (CH –C
1
995, pp. 510–511.
6
H ), 26.45 (s, C–C6 H ), 26.42
3 3
2
[
[
6] G. Cavinato, L. Toniolo, J. Organomet. Chem. 398 (1990) 187.
7] R. Bardi, A. Del Pra, A.M. Piazessi, L. Toniolo, Inorg. Chim.
Acta 35 (1979) L345.
1
(
s, C– C6 H ), 26.17 and 26.20 (t, CH D, Z =22.0,
3 2
1
1
CD
1
J
1
1
=19.8 Hz), 25.78 and 25.80 (q, Z =25.7, J
=
CD
8.6 Hz), 60.70 (C
28.30, 128.40 (phenyl ring).
6
H –CH ), 125.45, 126.52, 126.54,
[8] D.M. Fenton, J. Org. Chem. 38 (1973) 3192.
[9] R.F. Heck, J. Am. Chem. Soc. 94 (1972) 2712.
2
3
[
[
[
[
[
[
10] K.H. Shaughnessy, R.M. Waymouth, Organometallics 16 (1996)
1
001.
´
11] Cs. Benedek, G. Szalontai, A. G o¨ m o¨ ry, Sz. T o˜ r o¨ s, B. Heil, J.
Organomet. Chem. 579 (1999) 147.
12] G. Uccello-Barretta, R. Lazzaroni, R. Settambolo, P. Salvadori,
J. Organomet. Chem. 417 (1991) 111.
13] A. Raffaelli, S. Pucci, R. Settambolo, G. Uccello-Barretta, R.
Lazzaroni, Organometallics 10 (1991) 3892.
14] J. Evans, J. Schwartz, P.W. Urquhart, J. Organomet. Chem. 81
Acknowledgements
The authors wish to thank L. Koll a´ r (Jannus Panno-
nius University, P e´ cs) for his stimulating interest and
helpful discussions and to Zs. Cs o´ k (University of
Veszpr e´ m) for the NMR measurements. This work was
supported by the Hungarian National Science Founda-
tion (OTKA T-016329).
(
1974) C37.
15] S. Toda, M. Miyamoto, H. Kinoshita, K. Inomata, Bull. Chem.
Soc. Jpn. 64 (1991) 3600.

C. Benedek et al. / Journal of Organometallic Chemistry 586 (1999) 85–93
93
[
[
16] D.L. Reger, P.J. McElligott, J. Organomet. Chem. 216 (1981)
C12.
17] R. Lazzaroni, A. Raffaelli, R. Settambolo, S. Bertozzi, G. Vit-
ulli, J. Mol. Catal. 50 (1989) 1.
[18] T. K e´ gl, L. Koll a´ r, L. Radics, Inorg. Chim. Acta 265 (1997) 249.
[19] R. Lazzaroni, G. Uccello-Barretta, S. Scamuzzi, R. Settambolo,
A. Caiazzo, Organometallics. 15 (1996) 4657.
[20] H. Itatani, J.C. Bailar, J. Am. Oil Chem. Soc. 44 (1967) 147.
.
.