Platinum-alkyl-B(C6F5)3 (or BF3) 'in situ' systems as tin(II) halide-free enantioselective hydroformylation catalysts

Article abstract of DOI:10.1016/j.jorganchem.2008.01.018

The dialkyl/diaryl-platinum complexes (Pt(CH₃)₂(bdpp); PtPh₂(bdpp) and Pt(2-Thioph)₂(bdpp), where bdpp stands for (2S,4S)-2,4-bis(diphenylphosphino)pentane) were reacted either with B(C₆F₅)

Full text of DOI:10.1016/j.jorganchem.2008.01.018

Available online at www.sciencedirect.com
Journal of Organometallic Chemistry 693 (2008) 1127–1135
www.elsevier.com/locate/jorganchem
Platinum–alkyl–B(C6F5)3 (or BF3) ‘in situ’ systems as tin(II)
halide-free enantioselective hydroformylation catalysts
a
b
a,
*
´
´ ´
´
´
´
´
´
Laszlo Janosi , Tamas Kegl , Laszlo Kollar
a University of Pe´cs, Department of Inorganic Chemistry, H-7624 Pe´cs, P.O. Box 266, Hungary
b University of Pannonia, Institute of Chemistry, H-8200 Veszpre´m, Hungary
Received 14 September 2007; received in revised form 29 December 2007; accepted 3 January 2008
Available online 17 January 2008
Abstract
The dialkyl/diaryl-platinum complexes (Pt(CH3)2(bdpp); PtPh2(bdpp) and Pt(2-Thioph)2(bdpp), where bdpp stands for (2S,4S)-2,4-
bis(diphenylphosphino)pentane) were reacted either with B(C6F5)3, BPh3 or BF3. In the presence of PPh3 or carbon monoxide cationic
species with a general formulae [PtR(L)(bdpp)]+ (L = PPh3, CO) were formed exclusively. The ability of boron additives to provide
vacant coordination site at the platinum made these systems suitable as hydroformylation catalysts. Enantioselective hydroformylation
of styrene was carried out in the presence of in situ catalysts formed from Pt(alkyl/aryl)2(bdpp) and B(C6F5)3 or BF3. Moderate e.e-s
depending strongly on the structure of the catalytic precursor have been obtained. DFT/PCM calculations reveal an SN2-type reaction
mechanism for the alkyl/aryl ligand abstraction with a notably lower activation barrier for BF3.
Ó 2008 Elsevier B.V. All rights reserved.
Keywords: Platinum-methyl complexes; Enantioselective hydroformylation; Triarylborane; Boron trifluoride
1. Introduction
ibuprofen as nonsteroidal antiimflammatory drug) made
them interesting as catalysts of synthetically important
reactions.
Among the carbonylation reactions hydroformylation is
considered as a homogeneous catalytic process of the high-
est industrial importance [1]. The cobalt and rhodium con-
taining catalytic systems have successfully been used even
in industrial scale [2] and have been investigated mechanis-
tically in details [2]. In the 1970s the platinum–phosphine–
tin(II) halide systems were discovered to act as promising
hydroformylation catalysts as well [3]. The application of
various types of chiral diphosphines resulted in excellent
enantioselectivities in platinum-catalysed hydroformyla-
tion [4–10]. Although their activity were behind the best
rhodium catalysts, the high regioselectivities obtained with
1,1-disubstituted olefins [11] and the facile route towards
optically active 2-aryl-propanal derivatives, the direct pre-
cursors of 2-phenyl-propionic acid derivatives [12–15] (like
The platinum-containing systems have been investigated
also mechanistically. Most of these investigations were
focused on the necessity of tin(II) chloride in the catalytic
system [16,17]. The crucial role of the trichlorostannato
ligand as a facile leaving group, formed upon ‘carbene-like’
insertion of tin(II) chloride into the platinum–chloride
bond, has been shown by HP NMR investigations as well
[18].
Quantum-chemical investigations concerning the forma-
tion of platinum–hydride bond [19,20], the olefin insertion
into the Pt–H bond [21–27] and the carbon monoxide inser-
tion into the platinum–alkyl bond and the isomerization
reactions before and after the insertion [28,29] have been
carried out as well.
In this paper, the role of boron-trifluoride, perfluoro-tri-
phenylborane and triphenylborane in forming a vacant
coordination site at platinum from Pt(alkyl/aryl)2(diphos-
phine) precursors will be discussed. Their potential as
*
Corresponding author. Tel.: +36 72 327622 4153; fax: +36 72 327622
4680.
´
E-mail address: kollar@ttk.pte.hu (L. Kollar).
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.01.018

1128
L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
hydroformylation catalytic systems in the hydroformyla-
tion of styrene will be discussed as well.
2. Results and discussion
2.1. Reaction of PtR2(bdpp) (R = CH3, Ph, 2-Thioph)
complexes with boron additives
The starting dialkyl/diaryl-complexes (Pt(CH3)2(bdpp)
(1); PtPh2(bdpp) (2) and Pt(2-Thioph)2(bdpp) (3)) under-
went complete alkyl/aryl ligand abstraction upon addition
of B(C6F5)3 or BF3. Rather complicated mixtures of equil-
ibrating species were obtained in the absence of a mono-
dentate donor ligand resulting in wide signals of half line
with of ca. 120 Hz. On the basis of NMR details the equi-
librium of dinuclear species 4a, 4b and 5 are supposed (runs
1, 2 and 8) (Scheme 1). Although the methyl-bridged dinu-
clear platinum complexes can be considered as rather spec-
ulative, it has to be noted that the DFT calculations
revealed a genuine energy minimum for the cation of 5
(Fig. 1) with C2 symmetry. The most common ways of met-
als being bridged by methyl groups are the (almost) sym-
metrically bridged geometry (A) [30,31] involving a three-
centre M–C–M interaction, and the asymmetrically
bridged agostic geometry (B), in which the methyl carbon
is r-bound to one metal while being involved in a three-
centre M–C–H agnostic interaction with the second metal
(Scheme 2). For late transition metal complexes geometry
B is far more common [32,33].
Fig. 1. Optimised geometry of the [Pt2(PH3)4(l-CH3)2]2+ cation of 5.
Bond distances are given in A.
˚
H
H
C
H
H
H
H
C
M
M
M
M
A
B
Scheme 2. The bonding modes of methyl groups as bridging ligands.
[PtR(PPh3)(bdpp)]+ (1a, 2a, 3a) (Scheme 3) and [Pt(CH3)
(CO)(bdpp)]+ (1b) [18] complexes (Scheme 4) have been
obtained in the presence of PPh3 and carbon monoxide,
respectively. Similar phenomenon of providing a vacant
However, upon addition of a compound as potential
ligand, cationic species of well-defined structure with a gen-
eral formulae [PtR(L)(bdpp)]+ were obtained. This way,
2+
P
P
CH3
Pt
-
Pt
P
P
B(C6F5)3(CH3)
2
H3C
4a
2+
P
P
CH3
Pt
P
P
CH3
CH3
-
-
Pt
P
CH3
B(C6F5)3(CH3)
2
Pt
+ 2 B(C6F5)3
2
P
1
4b
2+
CH3
Pt
P
P
P
-
B(C6F5)3(CH3)
Pt
2
P
CH3
5
Scheme 1. The formation of dinuclear platinum complexes.

L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
1129
+
P
P
R
P
P
R
R
-
Pt
Pt
+ PPh3 + B(C6F5)3
B(C6F5)3R
PPh3
1
2
3
(R=Me)
1a (R=Me)
2a (R=Ph)
(R=Ph)
(R=2-Thioph)
3a (R=2-Thioph)
Scheme 3. The formation of [Pt(alkyl/aryl)(PPh3)(bdpp)]+ complexes.
+
CH3
CO
P
P
P
P
CH3
CH3
B(C6F5)3(CH3) -
Pt
Pt
+ CO + B(C6F5)3
1
1b
Scheme 4. The formation of [Pt(CO)(CH3)(bdpp)]+ complex.
coordination site was achieved by using PtCl2(bdpp) cata-
lyst precursor and tin(II) chloride co-catalyst. A trichloro-
stannato ligand with good leaving group properties was
formed via insertion of tin(II) chloride into the Pt–Cl bond
(vide infra) [18].
‘PtRP3’-type ionic complexes, 1a and 2a, containing methyl
and phenyl ligands, respectively, were successfully isolated
as solid substances. B(C6F5)3 was used as methyl/phenyl
abstracting reagent forming the corresponding borate
counterion (see Section 3).
The reactions resulted in the above cationic species
regardless of the Pt/boron additive ratio. The same types
of complexes were obtained both in the presence of 1 and
2 equivalents of PPh3 as well (Table 1, runs 3, 6, 9, 11,
12 and runs 5 and 10, respectively). The boron additives
(B(C6F5)3 or BF3) were added in 1/1 ratio to PPh3, i.e.
either in 1/1 or 2/1 ratio to platinum. The results can be
rationalised as the abstraction of the ‘second’ alkyl/aryl
ligand could not be achieved even using the stoichiometri-
cally necessary amount of boron reagent. Two of these
The analogous square-planar carbonyl complex 1b was
formed when the reaction was carried out in the absence
of PPh3 under atmospheric carbon monoxide (run 4).
Interestingly, the formation of a further complex (6) has
been observed by using 3 as starting complex probably due
to ligand exchange reaction (run 7). The exchange of the
two aryl ligands, 2-Thioph and C6F5, is supposed, that
resulted in the formation of the [Pt(C6F5)(PPh3)(bdpp)]+
cation. Triphenyl borane proved to be the less powerful
additive related to methyl ligand abstraction (run 13).
Table 1
Reactions of PtR2{(S,S)-bdpp} complexes with boron additives (B(C6F5)3, BF3 or BPh3)a
Run
Complex
R. time (h)
Boron-additive
PPh3 (eq.)
Composition of the reaction mixture (%)b
[PtR(bdpp)(PPh3)]+
[PtR(CO)(bdpp)]+
Other species
1
2
3
4
5
6
7
8
9
10
11
12
13
a
1
0.25
0.25
0.25
0.5
0.25
0.25
0.25
0.25
0.25
0.25
0.25
0.25
120
1 B(C6F5)3
2 B(C6F5)3
1 B(C6F5)3
1 B(C6F5)3
2 B(C6F5)3
1 B(C6F5)3
1 B(C6F5)3
1 BF3
1 BF3
2 BF3
1 BF3
1 BF3
–
–
1
–
2
1
1
–
1
2
1
1
1
–
–
–
–
–
100 (4, 5)
100 (4, 5)
1
1
100 (1a)
–
100 (1a)
97 (2a)
31 (3a)
–
100 (1a0)
100 (1a0)
100 (20a0)
96 (3a )
64 (1a00)
1c
1
95 (1b)
5 (n.d.)
–
–
–
–
–
–
–
–
–
2
3 (n.d.)
69 (6)
100 (4, 5)
3
1
1
1
2
3
1d
4 (n.d.)
1 BPh3
Reaction conditions (unless otherwise stated): CDCl3; room temperature; Ar atmosphere; practically total conversions have been obtained; n.d.: the
structure could not be determined because of the lack of the 1J(Pt,P) coupling constants (that of the platinum satellites).
b
Determined by 31P NMR.
The reaction was carried out under atmospheric carbon monoxide.
c
d
Thirty six percentage of the starting complex (1) was not converted.

1130
L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
2.2. 31P NMR characterization of the platinum-bdpp
complexes formed upon addition of boron additives
methyl, phenyl and 2-thiophenyl series (1a–1a00, 2a–2a0
and 3a–3a0, respectively). The differences both in chemical
shifts ( 0.1 ppm) and 1J(Pt,P) coupling constants
( 3 Hz) fall within the range of error of determination
(Table 2).
All of the ‘PtP3’-type cationic complexes (1a, 1a0, 1a00,
2a, 2a0, 3a and 3a0) can be easily identified by the
2
1J(Pt,P) and J(P,P) coupling constants (Table 2). They
are of diagnostic value showing 2700–2800 Hz or 1700–
2000 Hz for phosphorus possessing phosphine or alkyl/aryl
ligand in trans position, respectively. As for the 31P NMR
patterns, each phosphorus attached to platinum gives three
signals at a ratio of 1/4/1 due to the natural abundance of
33.8% of 195Pt (I = 1/2) isotope. It results in a central line
due to phosphorus nuclei bound to platinum nuclei other
than 195Pt. The central lines are flanked by platinum satel-
lites, i.e. a doublet as a result of coupling between 195Pt and
31P nuclei both with I = 1/2. It has to be added that the
2.3. Hydroformylation of styrene in the presence of ‘in situ’
complexes formed from PtR2(bdpp) and boron-containing
additives
Styrene (7) as a model substrate was reacted in the pres-
ence of in situ catalysts, formed from one of the PtR2(bdpp)
precursors (1, 2 or 3) and boron additives (B(C6F5)3 or
BF3), with CO/H2 (1/1) at 100 °C and at a pressure of
80 bar (Table 3). In addition to the formation of the formyl
regioisomers 8 and 9 that of the hydrogenation product 10
is also expected.
phosphorus–phosphorus coupling constants (2Jcis and tJrans
)
are of diagnostic value as well, being in the range of 19–
29 Hz and 327–378 Hz, respectively.
It can be stated that neither the chemical shifts nor the
coupling constants show strong dependence on the borate
counterion formed by aryl abstraction from the complex.
Practically the same data have been obtained for the
CO=H2
PhCH ¼ CH2
!
7
PhCHðCHOÞCH3 þ PhCH2CH2CHO þ PhCH2CH3
9
10
8
ð1Þ
Table 2
31P NMR data of the platinum-{(S,S)-bdpp} complexes formed upon addition of B(C6F5)3, BPh3 or BF3 to the corresponding PtR2{(S,S)-bdpp}
complexesa
Complexb
dPcA
dPcB
dPcC
1J(Pt,PA) 1J(Pt,PB) 1J(Pt,PC) 2J(PA,PB) 2J(PA,PC) 2J(PB,PC)
(ppm) (ppm) (ppm) (Hz)
(Hz)
(Hz)
(Hz)
(Hz)
(Hz)
[Pt(CH3)(bdpp)(PPh3)]+ 1a, 1a0, 1a00
[Pt(Ph)(bdpp)(PPh3)]+ 2a, 2a0
29.4
21.0
26.3
15.4
17.7
14.6
13.0
8.9
22.6
16.5
17.7
–
2787
2792
2718
3141
2565
1848
1688
1990
1557
2017
2738
2744
2597
–
19.7
19.0
19.6
33.5
19.7
378
372
327
–
28.8
26.4
27.0
–
[Pt(2-Thioph)(bdpp)(PPh3)]+ 3a, 3a0
[Pt(CH3)(CO)(bdpp)][B(C6F5)3(CH3)] 1b
[Pt(C6F5)(bdpp)(PPh3)][B(C6F5)3(2-Thioph)] 6 19.8
15.1
17.6
2643
385
25.5
a
Spectra were measured in CDCl3 (room temperature).
1a0, 2a0, 3a0 and 1a00 stands for complexes possessing [BF3(CH3)]ꢀ, [BF3(Ph)]ꢀ, [BF3(2-Thioph)]ꢀ and [BPh3(CH3)]ꢀ counterion, respectively (for parent
b
compounds see Scheme3).
c
PB trans to aryl/alkyl ligand, PA trans to PC (PPh3) or CO.
Table 3
a
Hydroformylation of styrene (7) in the presence of PtR2{(S,S)-bdpp} + B(C6F5)3 (or BF3) in situ catalysts
Entry
Catalyst
Temperature (°C)
Conversion (%)
RCb
Rcbr
e.e.d (%)
1
2
3
4
5
6
7
8
a
1 + B(C6F5)3
2 + B(C6F5)3
3 + B(C6F5)3
1 + B(C6F5)3
1 + B(C6F5)f3
1 + BF3
100
100
100
50
100
100
100
100
>98
>98
>98e
88
>98e
>98
>98
>98
98
87
97
46
98
96
>98
>98
15
34
24
41
50
10
25
30
25 (R)
60 (R)
24 (R)
23 (R)
19 (R)
23 (R)
34 (R)
29 (R)
1 + PPh3 + BFg3
2 + PPh3 + BFg3
Reaction conditions (unless otherwise stated): 0.01 mmol PtR2(bdpp); 0.01 mmol B(C6F5)3 or BF3; 1 mmol styrene; solvent toluene;
p(CO) = p(H2) = 40 bar, r. time: 24 h.
b
Chemoselectivity (mol 8 + mol 9)/(mol 8 + mol 9 + mol 10) ꢁ 100.
c
Regioselectivity (mol 8)/(mol 8 + mol 9) ꢁ 100.
d
Determined by chiral-GC.
Partial reduction of the aldehydes (8 and 9) towards the corresponding alcohols (8a and 9a).
p(CO) = 40 bar, p(H2) = 80 bar.
One equivalent of PPh3 added.
e
f
g

L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
1131
The carbonylation activity of most of the platinum cat-
alysts used up till now is due to the presence of the tin(II)
halide, mostly tin(II) chloride, by forming trichlorostanna-
to ligand upon insertion of tin(II) chloride into the Pt–Cl
bond. The SnCl3 ligand can undergo facile dissociation
providing vacant coordination site for e.g. carbon monox-
ide or alkene [18].
reactivity of the PtR2(bdpp) precursors, the involvement
of five-coordinate species like PtR(H)(bdpp)(CO) or
PtR(R0)(bdpp)(CO) (R0 stands for the ‘second’ alkyl ligand
formed in the insertion of the alkene substrate into Pt–H
bond).
It has to be added that the regioselectivities towards
branched aldehyde are slightly modified by the consecutive
reduction of the aldehydes (entries 3 and 5), because hydro-
genation of the linear aldehyde (9) is favoured over the
branched one (8) resulting in the corresponding alcohols
3-phenyl-propanol (9a) and 2-phenyl-propanol (8a),
respectively. Therefore, the 8 to 9 ratio has slightly
increased due to this side reaction.
Low to moderate enantioselectivities have been obtained
by the variation of the structure of the catalytic precursors
and that of the reaction conditions. Also these results sug-
gest (vide supra) that the application of different precursors
leads to catalytic intermediates of different reactivities. The
application of 2 (R = Ph) as catalyst precursor resulted
undoubtedly in the highest e.e. by the dominance of the
R enantiomer (entry 2).
The basic idea of using Pt–alkyl/aryl complexes and
boron additives was to generate a platinum-containing
‘tin(II) halide-free’ hydroformylation catalyst. As has been
proven above, the addition of the boron additives, espe-
cially B(C6F5)3 and BF3, to the PtR2(bdpp) type complexes
could provide a vacant coordination site, which could be
occupied either by a phosphine in 1a, 2a, 3a or even by car-
bon monoxide forming a catalytically relevant species 1b.
According to the above considerations, catalytically
active platinum-containing hydroformylation catalysts,
formed in situ, have been obtained. All the three precur-
sors, 1, 2 and 3 with B(C6F5)3 formed active catalysts
(entries 1–3) resulting in nearly complete conversion of sty-
rene at 100 °C. Slightly lower conversion has been obtained
at 50 °C (entry 4). The chemoselectivity towards aldehyde
formation was varied between 87% and 99% when the reac-
tion was carried out at 100 °C. It is worth noting that the
hydrogenation of the substrate towards ethylbenzene (10)
took place to a very low extent even under increased hydro-
gen partial pressure (entry 5). Carrying out the reaction at
50 °C, unexpectedly low aldehyde selectivity has been
obtained (entry 4). It refers to catalytic intermediates com-
pletely different from those acting as catalytically active
complexes in ‘conventional’ platinum–phosphine–tin(II)
halide system. The latter ones show an opposite tempera-
ture dependence: the chemoselectivity towards aldehydes
is slightly increased by decreasing the temperature.
As for the regioselectivity of hydroformylation, the pre-
vailing formation of the linear aldehyde (9) was observed in
all cases. The formation of the branched aldehyde (8) was
favoured by low temperature (entry 4) and by increased
hydrogen partial pressure (entry 5). The regioselectivity
of hydroformylation is also influenced by the alkyl/aryl
group of the precursor. The presence of the methyl ligand
in the catalytic intermediates results in the lowest branched
regioselectivity, while that of the phenyl ligand in the high-
est in the complex precursor series (compare entries 1, 2
and 3, as well as 7 and 8). The selectivity differences refer
to different catalytic intermediates, i.e. the presence of the
alkyl/aryl or that of the corresponding acyl ligand, formed
by carbon monoxide insertion, in the coordination sphere.
However, it queries the exclusive formation of square-pla-
nar platinum complexes generally supposed to act as cata-
lytic intermediates. If three coordination sites are blocked
by bdpp and the alkyl (or aryl) ligand as spectator ligands,
two further positions are necessary e.g. for the insertion of
alkene to platinum–hydride bond or that of carbon monox-
ide to platinum–alkyl bond. Therefore, a reaction mecha-
nism different to that in the presence of tin(II)chloride
co-catalyst [18] might be operative. On the basis of the
2.4. Electronic structure of B(C6F5)3 and BF3
In order to have a deeper insight into the Lewis acid
characters of the boron additives NBO studies have been
carried out on the BP86-optimised structures of B(C6F5)3
(12) and BF3 (13). Both boron compounds are featured
by an unfilled one-centre nonbonded orbital almost
entirely formed from the pz atomic orbital. These one-cen-
tre orbitals, however have some occupancies (0.253 for 12
and 0.386 for 13) caused by a donor–acceptor interaction
with a C–C p-type orbital or with one of the lone pairs
of fluorine, respectively. The boron atom carries a positive
charge in both cases, more positive in BF3 as seen in Fig. 2.
The r-bond with the neighbouring atom is also more
polarized; the higher electronegativity of fluorine is
reflected in the higher polarization coefficient, the percent-
age of the B–F natural bond orbital on F is 82%, while the
B–C bonds are polarized in a 68–32% ratio towards the
ipso carbon atoms.
2.5. Computational studies on the alkyl ligand abstraction
mechanism
The mechanism of the alkyl abstraction from the plati-
num-dialkyl-diphosphine complex was modelled starting
with the cis-[Pt(PH3)2(CH3)2] complex (11), as precursor.
The reaction takes place via transition states 14TS and
15TS, with one characteristic imaginary frequency each
of 230i cmꢀ1 and 202i cmꢀ1, respectively (Fig. 3). In
14TS the breaking Pt–C and the forming C–F bonds tend
to be collinear, while in 15TS they have an angle of 154°.
BF3 reacts with the dimethyl complex 11 in a significantly
faster reaction, as the free energy of activation is
6.8 kcal/mol, while DGzsolv ¼ 19:9 kcal=mol during the reac-
tion of 11 with 12 in agreement with the higher Lewis-acid

1132
L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
Fig. 2. Optimised geometries of cis-[Pt(PH3)2(CH3)2] complex (11),
˚
Fig. 3. Computed geometries of the transition states of the methyl
abstraction step.
B(C6F5)3 (12) and BF3 (13). Bond distances are given in A, natural
charges are written in italics.
character of BF3. After the C–Pt bond cleavage tightly
bound adducts are formed (Fig. 4) in an endothermic reac-
tion with the reaction free energy being higher (16.4 kcal/
mol) in case of B(C6F5)3, than for BF3 (2.2 kcal/mol).
The adducts are stabilized by a loose coordination of one
methyl hydrogen (in 16) or two methyl hydrogens (in 17)
to the platinum atom, which is reflected in the higher disso-
ciation energy of the latter one. In the final stage of the
reaction the adducts dissociate in an endothermic reaction
into the coordinatively unsaturated platinum-methyl cation
(18) with the [B(C6F5)3(CH3)]ꢀ (19) or [BF3(CH3)]ꢀ (20)
ions, respectively (Fig. 5). The free energy needed to disso-
ciate 16 into 18 and 19 is notably lower (21.8 kcal/mol)
than that in case of 17 (36.1 kcal/mol) due to the stronger
stabilizing interactions within the adduct formed from BF3.
The free energy profiles of the two reactions are depicted in
Fig. 6.
methyl group to iodine in Pt–diphosphine complexes [34].
The free energy profile of the reaction of 12 and 13 with
11 shows significant difference both in the activation bar-
rier of the substitution and in the relative energy of the
resulted adduct. These differences, however, are not
reflected in the catalytic activity probably due to the slower
formation of the catalytically active hydrido species in the
following step. Additional computational work is in pro-
gress aiming at the detailed mechanism of the platinum-
catalysed hydroformylation as well as the mechanism of
the formation of the catalytically active species.
3. Experimental
3.1. General
The methyl abstraction reactions discussed in this study
reveal formally an SN2-type mechanism taking place on
one of the methyl groups as the reaction centre. This mech-
anism is similar to that reported for the substitution of
The starting dialkyl/diaryl-complexes (Pt(CH3)2(bdpp)
(1); PtPh2(bdpp) (2) and Pt(2-Thioph)2(bdpp) (3)) were syn-
thesised as described previously [35]. (S,S)-bdpp was pur-
chased from Strem.

L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
1133
Fig. 5. Optimised geometries of cis-[Pt(PH3)2(CH3)]+ cation (18), and the
[B(C6F5)3(CH3)]ꢀ (19) and [BF3(CH3)]ꢀ (20) anions, respectively.
3.3. Synthesis of [PtPh{(S,S)-
bdpp}(PPh3)][BPh(C6F5)3] 2a complex
To
a
degassed solution of 78.9 mg (0.1 mmol)
PtPh2{(S,S)-bdpp} (2) and 26.2 mg (0.1 mmol) PPh3 in
10 mL benzene a degassed solution of 51.1 mg (0.1 mmol)
B(C6F5)3 in 5 mL benzene is added under argon. The reac-
tion mixture is stirred at room temperature for 3 h. The sol-
vent was distilled off and a white powder-like precipitate
was formed. It was washed with hexane and dried in
vacuum.
Yield: 55%. Anal. Calc. for C77H55P3F15BPt (1564.07):
C, 59.13; H, 3.54. Found: C, 59.02; H, 3.39%. For NMR
see Table 2.
Fig. 4. Optimised geometries of the ion pairs/adducts resulted in the
methyl abstraction step.
Toluene was distilled and purified by standard methods
and stored under argon. All reactions were carried out
under argon using standard Schlenk techniques.
1
The 31P and H NMR spectra were taken on a Varian
Inova 400 spectrometer operating at 400.13 MHz (1H)
and 161.89 MHz (31P). Chemical shifts are reported in
ppm relative to 85% H3PO4 or TMS (downfield) for 31P
3.4. Hydroformylation experiments
1
and H NMR, respectively.
In a typical experiment a solution of 0.01 mmol of
PtR2(bdpp) (1, 2 or 3) and 0.01 mmol of boron additive
(B(C6F5)3, BPh3 or BF3) in 5 mL toluene containing 1 mmol
of styrene was transferred under argon into a 100 mL stain-
less steel autoclave. The reaction vessel was pressurized to
80 bar total pressure (CO/H2 = 1/1) and placed in an oil
bath and the mixture was stirred with a magnetic stirrer
for the given reaction time. The pressure was monitored
throughout the reaction. After cooling and venting of the
autoclave, the pale yellow solution was removed and imme-
diately analysed by GC–MS and chiral GC.
3.2. Synthesis of [PtMe{(S,S)-
bdpp}(PPh3)][BMe(C6F5)3] 1a complex
To
a
degassed solution of 66.5 mg (0.1 mmol)
PtMe2{(S,S)-bdpp} (1) and 26.2 mg (0.1 mmol) PPh3 in
10 mL benzene a degassed solution of 51.1 mg (0.1 mmol)
B(C6F5)3 in 3 mL benzene is added under argon. The reac-
tion mixture is stirred at room temperature for 1 h. The sol-
vent was distilled off and a white powder-like precipitate
was formed. It was washed with hexane and dried in
vacuum.
3.5. Computational details
Yield: 62%. Anal. Calc. for C67H51P3F15BPt (1439.93):
C, 55.89; H, 3.57. Found: C, 56.01; H, 3.49%. For NMR
see Table 2.
Full geometry optimisations have been performed at the
density functional level of theory without any symmetry

1134
L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
Fig. 6. Computed free energy profile of the proposed ligand abstraction mechanism calculated at BP86/LANL2DZ level of theory including solvation
effects.
[2] I. Ojima, C.-Y. Tsai, M. Tzamarioudaki, D. Bonafoux, The hydro-
constraints using the GAUSSIAN 03 suite of programs [36] and
the LANL2DZ basis set for P and Pt, with the correspond-
formylation reaction, in: L. Overman et al. (Eds.), Organic Reactions,
J. Wiley & Sons, 2000, pp. 1–354.
ing effective core potential (relativistic for Pt) [37] and the
[3] I. Schwager, J.F. Knifton, J. Catal. 45 (1976) 256.
valence double-f basis set of Dunning and Hay [38]. The
[4] G. Consiglio, P. Pino, L.I. Flowers, C.U. Pittmann Jr., J. Chem.
density fitting basis sets were generated automatically from
the AO primitives by the GAUSSIAN 03 program. The station-
ary points were characterized by frequency calculations in
order to verify that they have zero imaginary frequency
for equilibrium geometries and one imaginary frequency
for transition states. The NBO analyses were carried out
on the stationary points using the NBO 3.1 program [39]
as implemented in Gaussian. For all the calculations the
gradient-corrected exchange functional developed by Becke
[40] was utilised in combination with a correlation func-
tional developed by Perdew [41] and denoted as BP86.
To estimate the effect of the solvent, single-point calcu-
lations on the gas-phase optimised structures were carried
out using the polarized continuum model (PCM) [42] using
the dielectric constant e0 = 2.38 for toluene. PCM method
has been used for numerous studies involving transition
metal complexes with reasonable accuracy [43,44]. Ther-
mochemistry corrections were taken from gas-phase fre-
quency calculations at 298.15 K and 1 atm.
Soc., Chem. Commun. (1983) 612.
[5] P. Haelg, G. Consiglio, P. Pino, J. Organomet. Chem. 296 (1985)
281.
[6] G. Consiglio, F. Morandini, M. Scalone, P. Pino, J. Organomet.
Chem. 279 (1985) 193.
[7] G. Parrinello, J.K. Stille, J. Am. Chem. Soc. 109 (1987) 7122.
´
´
[8] L. Kollar, J. Bakos, I. Toth, B. Heil, J. Organomet. Chem. 370 (1989)
257.
[9] G. Consiglio, S.C.A. Nefkens, A. Borer, Organometallics 10 (1991)
2046.
´
[10] I. Toth, I. Guo, B. Hanson, Organometallics 12 (1993) 848.
´
[11] L. Kollar, G. Consiglio, P. Pino, J. Organomet. Chem. 330 (1987)
305.
[12] C. Botteghi, S. Paganelli, A. Schionato, M. Marchetti, Chirality 3
(1991) 355.
[13] C. Botteghi, M. Marchetti, S. Paganelli, in: M. Beller, C. Bolm
(Eds.), Transition Metals for Organic Synthesis, vol. 2, Wiley–VCH,
Weinheim, 1998, p. 25ff.
´
[14] S. Gladiali, J.C. Bayon, C. Claver, Tetrahedron: Asymmetry 6 (1995)
1453.
[15] F. Agbossou, J.-F. Carpentier, A. Mortreux, Chem. Rev. 95 (1995)
2485.
[16] M. Gomez, G. Muller, D. Sainz, J. Salez, Organometallics 10 (1991)
4036.
References
[17] A. Scrivanti, C. Botteghi, L. Toniolo, A. Berton, J. Organomet.
Chem. 344 (1988) 261.
´
´
´
[18] I. Toth, C.J. Elsevier, T. Kegl, L. Kollar, Inorg. Chem. 33 (1994)
[1] C.D. Frohning, C.W. Kohlpainter, in: B. Cornils, W.A. Herrmann
(Eds.), Applied Homogeneous Catalysis with Organometallic Com-
pounds, vol. 1, VCH, Weinheim, 1996, p. 29ff.
5708.
[19] P.J. Hay, New J. Chem. 15 (1991) 735.

L. Ja´nosi et al. / Journal of Organometallic Chemistry 693 (2008) 1127–1135
1135
[20] T. Matsubara, F. Maseras, N. Koga, K. Morokuma, J. Phys. Chem.
100 (1996) 2573.
[21] B.B. Coussens, F. Buda, H. Overing, R.J. Meier, Organometallics 17
(1998) 795.
[22] N. Koga, K. MorokumaTheoretical Aspects of Homogeneous
Catalysis, vol. 18, Kluwer Academic Publishers, Dordrecht, Boston,
London, 1995, p. 65.
[31] M. Ganesan, F.P. Gabbai, Organometallics 23 (2004) 4608.
[32] C.P. Casey, P.J. Fagan, W.H. Miles, J. Am. Chem. Soc. 104 (1982)
1134.
[33] J.R. Wigginton, S.J. Trepanier, R. McDonald, M.J. Ferguson, M.
Cowie, Organometallics 24 (2005) 6194, and references cited therein.
´
´
[34] T. Kegl, L. Kollar, J. Organomet. Chem. 692 (2007) 1852.
´
´
[35] L. Janosi, L. Kollar, P. Macchi, A. Sironi, J. Organomet. Chem. 691
[23] S. Sakaki, M. Ogawa, Y. Musashi, T. Arai, J. Am. Chem. Soc. 116
(1994) 7258.
[24] S. Creve, H. Overing, B.B. Coussens, Organometallics 18 (1999)
1967.
[25] D.L. Thorn, R. Hoffmann, J. Am. Chem. Soc. 100 (1978) 2079.
[26] N. Koga, S.Q. Jin, K. Morokuma, J. Am. Chem. Soc. 110 (1988)
3417.
[27] W.R. Rocha, W.B. Almeida, Organometallics 17 (1998) 1961.
[28] S. Sakaki, K. Kitaura, K. Morokuma, K. Ohkubo, J. Am. Chem.
Soc. 105 (1983) 2280.
(2006) 2846.
[36] GAUSSIAN 03, Revision B.05, Gaussian Inc., Pittsburgh, PA, 2003.
[37] W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284;
W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 299.
[38] T.H. Dunning Jr., P.J. Hay, in: H.F. SchaeferIII (Ed.), Modern
Theoretical Chemistry, vol. 3, Plenum, New York, 1976, pp. 1–28.
[39] E.D. Glendening, A.E. Reed, J.E. Carpenter, F. Weinhold, NBO
Version 3.1.
[40] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
[41] J.P. Perdew, Phys. Rev. B 33 (1986) 8822.
[29] W.R. Rocha, W.B. Almeida, J. Comput. Chem. 21 (2000) 668.
[30] Z. Yu, J.M. Wittbrodt, M.J. Heeg, H.B. Schlegel, C.H. Winter, J.
Am. Chem. Soc. 122 (2000) 9338.
[42] S. Miertus, E. Scrocco, J. Tomasi, Chem. Phys. 55 (1981) 117.
[43] W. Zierkiewicz, T. Privalov, Dalton Trans. (2006) 1867.
[44] D. Tang, S. Qin, Z. Su, C. Hu, Organometallics 26 (2007) 33.