Synthesis of neutral rhodium(I) complexes containing a rigid P-O ligand and their use as catalyst precursors for the hydroformylation of 1-hexene

Article abstract of DOI:10.1016/j.molcata.2008.05.013

The first rhodium complexes with the anionic 2-(bis(2-methoxyphenyl)phosphino)benzenesulfonate (L) ligand have been synthesized and characterized. Selected complexes have been used as catalyst precursors for the homogeneous hydroformylation of 1-hexene, which has been carried out in preparative autoclaves and studied under operando conditions by means of both high-pressure NMR (HPNMR) and high-pressure IR (HPIR) spectroscopy. Operando spectroscopic hydroformylation experiments have shown that the catalyst precursors bearing either 1,5-cyclooctadiene or triethylamine as ancillary ligand are rapidly converted into di-carbonyl complexes. Rhodium(I) mono-and di-carbonyl complexes have been shown to be resting states of the catalytic cycle, most likely in equilibrium with other species containing either the zwitterionic ligand HL coordinated in the κ¹-O bonding mode or exclusively carbonyl ligands.

Full text of DOI:10.1016/j.molcata.2008.05.013

Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Synthesis of neutral rhodium(I) complexes containing a rigid P–O ligand and
their use as catalyst precursors for the hydroformylation of 1-hexene
Lorenzo Bettucci, Claudio Bianchini, Andrea Meli, Werner Oberhauser∗
Istituto di Chimica dei Composti Organometallici (ICCOM-CNR) Area di Ricerca CNR di Firenze, via Madonna del Piano 10, 50019 Sesto Fiorentino, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 May 2008
Received in revised form 26 May 2008
Accepted 27 May 2008
Available online 3 June 2008
The first rhodium complexes with the anionic 2-(bis(2-methoxyphenyl)phosphino)benzenesulfonate (L)
ligand have been synthesized and characterized. Selected complexes have been used as catalyst precursors
for the homogeneous hydroformylation of 1-hexene, which has been carried out in preparative autoclaves
and studied under operando conditions by means of both high-pressure NMR (HPNMR) and high-pressure
IR (HPIR) spectroscopy. Operando spectroscopic hydroformylation experiments have shown that the cata-
lyst precursors bearing either 1,5-cyclooctadiene or triethylamine as ancillary ligand are rapidly converted
into di-carbonyl complexes. Rhodium(I) mono-and di-carbonyl complexes have been shown to be resting
states of the catalytic cycle, most likely in equilibrium with other species containing either the zwitterionic
ligand HL coordinated in the ␬1-O bonding mode or exclusively carbonyl ligands.
Keywords:
Hydroformylation
Neutral rhodium(I) complexes
2-(Bis(2-methoxyphenyl)-
phosphino)benzenesulfonic
acid
© 2008 Elsevier B.V. All rights reserved.
High-pressure spectroscopy
1. Introduction
In this paper, we report the synthesis and characterization of the
first rhodium complexes with the anionic L ligand, which include
The study of transition metal complexes supported by mixed-
donor chelate ligands is still an active and fruitful area of chemical
research, especially as regards homogeneous catalysis [1]. Of par-
ticular interest are dissymmetric chelate ligands bearing different
donor atoms, as they have the potential of regioselectively creating
a free coordination site at the metal centre, thus affecting substrate
recognition and activation as well as reaction rates [1,2].
A class of mixed-donor ligands that is attracting increasing
attention in catalytic polymerization reactions in conjunction
with late transition metals is constituted by anionic P–O lig-
ands [3]. Nickel P–O catalysts are effective for the polymerization
and oligomerization of ␣-olefins [3a,4], while some palladium
systems are active for the non-alternating CO-ethylene copoly-
merization [3d,j,k] as well as the copolymerization of ethylene
with functionalized olefins (Scheme 1a and b) [3e–i]. Accom-
plishing these reactions requires the coordination of PdII ions
by anionic P–O ligands generated by the deprotonation of 2-
(bis(2-methoxyphenyl)phosphino)benzenesulfonic acid (HL) and
its derivatives (Scheme 1c). Other C–C bond forming processes that
are successfully catalyzed by PdL+ fragments are the Suzuki and
Heck reactions [3k,l].
[Rh(␬2-O,P-L)(␩4-COD)] (COD = 1,5-cyclooctadiene) (1) [Rh(␬2-
O,P-L)(CO)2] (2) and [Rh(␬2-O,P-L)(CO)(Y)] (Y = PPh3 (3) PAr2(OMe)
(Ar = 2-methoxyphenyl) (4) 1-phospha-3,5,7-triazaadamantane
(PTA, 5) PHAr2 (Ar = 2-methoxyphenyl) (6) NEt3 (7) Py (8)
(Scheme 2).
Some of these complexes have been used as catalyst precur-
sors for the homogeneous hydroformylation of 1-hexene, which
has been carried out in preparative autoclaves and studied
under operando conditions by means of both high-pressure NMR
(HPNMR) and high-pressure IR (HPIR) spectroscopy. The results of
the batch reactions and of the operando spectroscopic experiments
are reported in this paper.
2. Results and discussion
2.1. Synthesis of rhodium(I) complexes
The reaction of the binuclear rhodium(I) complex [RhCl(␩4-
COD)]2 with AgPF6 in a solvent mixture of dichloromethane
and acetone is a well known method to generate the adduct
[Rh(␩4-COD)(acetone)2]PF6 [5]. The reaction of the latter com-
plex with the zwitterionic P–O ligand HL [3d] gave the
neutral micro-crystalline compound 1 in almost quantitative
yield, which was characterized in solution by multinuclear
NMR spectroscopy and in the solid state by both elemen-
∗
Corresponding author. Tel.: +39 0555225284; fax: +39 0555225203.
E-mail address: werner.oberhauser@iccom.cnr.it (W. Oberhauser).
1381-1169/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2008.05.013

58
L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
Scheme 1.
Scheme 2. Synthesis of the rhodium–carbonyl complexes 2–8.
tal analysis and a single crystal X-ray structure determination
(Scheme 2).
The asymmetric coordination of COD to the metal centre in 1 was
1
proved by the presence of two 13C{ H} NMR signals, at 69.98 ppm
(2JP,C = 14.8 Hz) and 104.30 ppm attributable to the sp2-hybridized
carbon atoms trans to the oxygen and phosphorus atom of the P–O
ligand, respectively. Accordingly, the 1H NMR spectrum showed a
high-field shifted broad singlet at 2.81 ppm for the olefin protons
located trans to the oxygen atom, while the olefin protons cis to
the latter atom gave a singlet at 5.36 ppm, which is in line with the
different trans influence of oxygen and phosphorus atoms.
The overall stereochemistry of 1 in the solid state is shown in
the ORTEP drawing of Fig. 1, while selected bond distances and
angles as well as crystallographic data are reported in Tables 1 and 3,
respectively.
The crystal structure of 1 shows the ligand L coordinated to
the metal center in the ␬2-O,P mode. The rhodium coordination
sphere is completed by two C C bonds from COD. The metal is
part of a twisted six-membered ring system, which is character-
ized by a torsion angle of 37.39(24)◦, defined by the atoms Rh(1)
P(1) C(1) and C(2). The value of the P(1)–Rh(1)–O(1) coordination
angle (92.20(2)◦) is in the range of six membered metallarings [6].
Fig. 1. ORTEP plot of 1. Thermal ellipsoids are shown at the 30% probability level.
Hydrogen atoms are omitted, except the one showing an intra-molecular short
distance to the metal atom.

L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
59
Table 1
coordinating CO groups in 2 was unambiguously demonstrated by
◦
˚
Selected bond lengths (A) and angles ( ) for 1
IR spectroscopy, showing two bands at 2097 and 2019 cm−1 (CH2Cl2
1
solution). Consistently, the 13C{ H} NMR spectrum of 2 at room
Rh(1)–P(1)
Rh(1)–O(1)
Rh(1)–C(21)
Rh(1)–C(22)
Rh(1)–C(25)
Rh(1)–C(26)
C(21)–C(22)
C(25)–C(26)
P(1)–Rh(1)–O(1)
S(1)–O(1)–Rh(1)
C(21)–Rh(1)–C(26)
C(22)–Rh(1)–C(25)
2.3127(9)
2.103(2)
2.116(4)
temperature contained two broad signals at 184.58 (CO trans to O)
and 180.43 ppm (CO trans to P), that resolved at −60 ◦C into a dou-
blet at 184.72 ppm (1JRh,C = 55.3 Hz) and a doublet of doublets at
179.60 ppm (2JP,C = 115.9 Hz, 1JRh,C = 65.4 Hz) respectively.
Compound 2, synthesized in situ, was transformed into the
mono-carbonyl derivatives 3–8 by reaction with various nucle-
ophiles (Scheme 2).
2.114(4)
2.232(3)
2.241(3)
1.406(6)
1.357(5)
92.20(6)
125.26(11)
80.99(15)
81.69(14)
The synthesis of compounds 3–8 is straightforward and allowed
for the isolation of micro-crystalline, yellow products in quite sat-
isfactory yields (ca. 70%).
˚
Intramolecular distances (A)
The phosphine or phosphinite complexes 3–6 are characterized
Rh(1)· · ·O(4)
Rh(1)· · ·O(5)
Rh(1)· · ·H(19)
O(3)· · ·H(3)
3.597(3)
5.254(2)
2.876
1
by a 31P{ H} NMR pattern consisting of a doublet of doublets with
2
a large JP,P value, which is typical for two mutually trans phos-
2.415
2
phorus donor atoms. The trans JP,P values are in the range from
319.1 to 373.1 Hz, with 4 and 5 showing the largest and the small-
est value, respectively. The 1JRh,P values of these complexes are, as
expected, in the range from 123.1 to 155.1 Hz [9]. The presence of a
The COD ligand coordinates asymmetrically to the rhodium atom,
due to the different trans influence of the P(1) and O(1) donor atoms
[7]. The Rh–C bonds trans to both donor atoms are thus signifi-
cantly different (2.116(4) and 2.114(4) (trans oxygen) vs. 2.232(3)
and 2.241(3) (trans phosphorus). Consequently, the C(25)–C(26)
1
unique CO ligand in 3–6 was proved by IR and 13C{ H} NMR mea-
surements. A single CO stretching band was invariably observed at
ca. 1983 cm−1, while the 13C{ H} NMR spectra of the 13CO enriched
1
complexes acquired in CDCl3 at room temperature showed a singlet
1
centered at ca. 189 ppm. At −60 ◦C, 3 exhibited a 13C{ H} NMR dou-
˚
bond located trans to the P atom is significantly shorter (1.357(5) A)
blet of triplets at 189.36 ppm (1JRh,C = 77.2 Hz, 2JP,C = 16.9 Hz) while
the spectra of 4 and 5 contained doublets at 189.13 (1JRh,C = 50.3 Hz)
and at 188.35 ppm (1JRh,C = 72.4 Hz), respectively. Only 6 showed,
even at lower temperature, a broad hump centered at 189.19 ppm,
consistent with a very fast Rh–CO exchange rate.
as compared to the C(21)–C(22) bond located cis to the same P atom
˚
(1.406(6) A).
The sp2-hybridized carbon atoms of COD are not equidistant
from the coordination plane defined by the atoms P(1) Rh(1) and
˚
O(1), showing the following deviations in A from the latter plane:
The reaction of 2 with either a stoichiometric amount or an
excess of triethylamine or pyridine gave the neutral mono-carbonyl
complexes 7 and 8, respectively (Scheme 2). Both complexes are
featured by a similar stereochemistry, where the N atom is located
C(21) 0.5863(47); C(22) −0.8099(48); C(25) −0.6281(40) and C(26)
0.7082(41). The spatial disposition of the two 2-methoxyphenyl
groups is such to originate two different rhodium methoxy–oxygen
˚
distances of 3.597(3) and of 5.254(2) A and one intra-molecular
1
trans to the P atom, as indicated by a 13C{ H} NMR study in CDCl3
˚
rhodium ortho-hydrogen interaction (i.e. Rh(1)· · ·H(19)) of 2.876 A
at room temperature with the 13CO-enriched derivatives. Com-
(Fig. 1). Similar metal–hydrogen interactions have been encoun-
tered in other 2-methoxyphenyl modified phosphine complexes
[8].
pound 7 showed a 13C{ H} NMR doublet of doublets at 187.72 ppm
1
(1JRh,C = 85.2 Hz and JP,C = 21.4 Hz) for the coordinated CO, while
2
8 was featured by a doublet at 188.07 ppm (1JRh,C = 79.1 Hz). The
rapid CO-exchange reaction for the latter complex was frozen out
at −60 ◦C. At this temperature, a multiplet at 188.28 ppm was clearly
Complex 1 was transformed into the corresponding di-carbonyl
complex 2 by bubbling CO into a CHCl3 solution of the former com-
plex (Scheme 2). All our attempts to isolate a pure specimen of 2,
even at low temperature, were unsuccessful, which may be due to a
rapid intermolecular CO-exchange process. The presence of two cis
1
visible with the expected multiplicity (dd, JRh,P = 83.6 Hz and
2JP,C = 19.8 Hz). The corresponding IR spectra of 7 and 8 in CH2Cl2
Table 2
Hydroformylation of 1-hexene catalyzed by rhodium catalystsa
Entry
Precursor
Time [h]
Conversion (%)
n-Heptanal (%)
2-Methyl-hexanal (%)
2-Ethyl-pentanal (%)
0.2
2-Hexenes (%)
TOF aldehydes (h−1
)
l/b Ratio
1
2b
3
1
1
1
1
3
3
3
4
5
5
6
7
1
1
2
1
1
1
2
1
1
2
1
1
1
1
1
64.5
90.4
94.5
63.3
50.6
70.9
67.1
60.8
60.3
81.0
80.1
54.1
49.4
99.8
78.0
43.2
53.1
59.2
42.4
35.7
45.7
45.9
41.5
40.4
47.1
50.0
35.7
30.7
51.9
47.1
19.1
26.9
27.2
19.0
14.9
21.6
20.3
19.3
18.7
22.6
22.7
15.9
15.2
34.5
23.5
2.2
10.2
8.1
1.9
0
3.6
0.9
0
1.2
11.3
7.4
2.5
3.5
7.1
7.0
125
160
86
123
101
135
66
122
118
70
145
103
92
2.26
1.96
2.18
2.23
2.40
2.12
2.26
2.15
2.16
2.08
2.20
2.25
2.02
1.27
1.98
4c
5
6b
7
8
9
10
11
12
13d
14
15b
7
Rhe
Rhe
6.3
0.3
185
142
a
Reaction conditions: 0.01 mmol Rh, 30 ml THF, 250 ␮l (2.0 mmol) 1-hexene, 60 ◦C, 42 bar 1:2 CO/H2 at 60 ◦C.
1:1 CO/H2.
Reaction in 30 ml of cyclohexane.
Reaction in the presence of NEt3 (0.04 mmol).
[RhCl(␩4-COD)]2.
b
c
d
e

60
L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
contained a broad CO stretching band at 1970 and 1979 cm−1
respectively.
,
2.2. Catalytic study
The rhodium complexes 1 and 3–7 were employed to catalyze
the hydroformylation of 1-hexene in THF at 60 ◦C. The results of
the catalytic reactions and the experimental conditions are sum-
marized in Table 2.
The catalytic reactions were performed for 1 h in the presence of
a 1:2 CO/H2 mixture at 42 bar and gave mixtures of n-heptanal and
2-methylhexanal with modest regioselectivity in all cases and small
amounts of cis/trans 2-hexene. Performing the experiments in the
presence of a 1:1 CO/H2 gas mixture gave also 2-ethylpentanal and
an increased amount of 1-hexene isomerization products with an
overall product composition and l/b ratio similar to those obtained
with the unmodified rhodium catalyst precursor [RhCl(␩4-COD)]2
(entries 2 and 6 vs. 15). The hydroformylation reaction performed
with the precursor 1 in cyclohexane gave a result comparable to
that obtained in THF (entry 4 vs. 1), ruling out any potential role
of THF in the catalytic cycle such as that of promoting the het-
erolytic splitting of H2 with formation of mono-hydride species
[9a,10].
Although we did not have direct spectroscopic evidence of
catalyst degradation with time, it is very likely that some decom-
position occurs. Indeed, we observed a decrease in the aldehyde
regioselectivity as well as a concomitant increase in the 1-hexene
isomerization on increasing the reaction time from 1 to 2 h for
the reactions catalyzed by 1 and 5 (entries 1 and 9 vs. 3 and 10,
respectively).
Unlike 1 and 5, the catalyst derived from 3, containing the
strongly coordinating co-ligand PPh3, showed comparable alde-
hyde regioselectivity and isomerization rate in the reactions lasting
either 1 h or 2 h (entries 5 and 7) which corroborates the hypoth-
esis of significant decomposition of the catalysts containing a less
coordinating co-ligand than PPh3.
Using 7, that contains a nitrogen coligand such as NEt3 (entry 12)
which, in principle, might favor the heterolytic splitting of H2 [11],
increased neither the catalytic activity nor the regioselectivity for n-
heptanal formation. Interestingly, the addition of an excess of NEt3
to the reaction mixture containing 7 showed a decrease of the l/b-
ratio of the aldehydes (entry 13 vs. 14), indicating a transformation
of the initially active catalytic species (vide infra).
Fig. 2. HPIR study of 1-hexene hydroformylation with precursors 1, 3, 4, 5, 6 and
7 (precursor (0.063 mmol), 1-hexene (12.6 mmol), THF (25 ml), 60 ◦C, CO/H2 (1:2)
(42.0 bar), 1 h, 60 ◦C): (a) precursor 1, (b) precursor 3, (c) precursor 4, (d) precursor 5,
(e) precursor 6, (f) precursor 7 and (g) precursor 7 and NEt3 (4 equiv.); × (unidentified
rhodium species).
dinating ligand NEt3, was rapidly transformed into 2 together with
a small amount of an unidentified rhodium carbonyl species, whose
concentration, however, increased by adding extra NEt3 (trace f vs.
g). Intrigued by this fact, the hydroformylation of 1-hexene by 7
1
was studied by 31P{ H} HPNMR spectroscopy by means of a 10 mm
OD sapphire tube equipped with a titanium valve in a 1:1 (v:v)
mixture of THF and CD2Cl2. A sequence of variable-temperature
1
31P{ H} NMR spectra is reported in Fig. 3.
1
The 31P{ H} NMR spectrum of 7 at room temperature showed a
doublet centered at 21.50 ppm (1JRh,P = 155.1 Hz) (trace a). On addi-
tion of 1-hexene (200 equiv.), no spectral change occurred (trace
b). Upon pressurizing the NMR tube with CO (14 bar), NEt3 was
replaced by CO to yield the di-carbonyl 2 (trace c). At this point, the
tube was pressurized with H2 (28 bar) at room temperature.
1
As a result, a 31P{ H} NMR singlet at −29.30 ppm appeared
which we safely assign to the ammonium salt of the P–O ligand
[HNEt3(L)] (9) (vide infra). On heating the NMR probe at 60 ◦C for
1 h (trace e) the concentration of 9 increased at the expense of that
of 2. A 2/9 ratio of ca. 1 was estimated by signal integration in
1
the 31P{ H} NMR spectrum acquired at room temperature (trace
f) while an analogous HPNMR study performed in the presence of
2.3. Operando spectroscopic studies
In an attempt of intercepting rhodium species, which might
be relevant to rationalize the hydroformylation of 1-hexene, cat-
alyzed by 1–7, HPIR and HPNMR spectroscopic studies were carried
out. The HPIR studies were performed in THF with an approxi-
mately twelve-fold concentration of the precursor. The HPIR cell
was incorporated into a titanium autoclave, equipped with an effec-
tive mechanical stirring and all the other necessary control devices
[12]. The IR spectra were acquired at room temperature and at 60 ◦C.
The reactions were monitored at this temperature for 1 h. A repre-
sentative absorption IR spectrum of each catalytic system studied
is shown in Fig. 2.
Typically, a THF solution of the rhodium precursors (1 or 3–7)
containing the desired amount of 1-hexene was pressurized with a
1:2 CO/H2 gas mixture to a total pressure of 36 bar at room temper-
ature. The autoclave was then heated to 60 ◦C, reaching an internal
pressure of 42 bar. The IR spectra clearly showed the conversion of
1 into the di-carbonyl 2 (trace a), whereas the mono-carbonyl com-
plexes 3–6 remained unchanged (traces b–e). At variance with all
the other precursors investigated, 7, bearing the very weakly coor-
1
Fig. 3. Variable-temperature 31 P{ H} HPNMR study of the hydroformylation of 1-
hexene with 7 (sapphire tube, THF/CD2Cl2 (1:1, v:v) (2 ml), 1-hexene (200 equiv.),
81.01 MHz): (a) 7 at room temperature; (b) addition of 1-hexene, (c) immediately
after pressurizing the tube with CO (14 bar), (d) immediately after pressurizing the
tube with H2 (28 bar) at room temperature, (e) after 1 h at 60 ◦C and (f) after cooling
to room temperature.

L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
61
generating an unmodified rhodium carbonyl catalyst (Fig. 2, trace
f) [13].
Operando HPNMR experiments were carried out also for the
other catalyst precursors investigated in this work. In all cases,
the spectra were poorly informative, as only mono-carbonyl or di-
carbonyl complexes were seen for all reactions lasting 1 h at 60 ◦C.
Even HPNMR experiments carried out in the absence of substrate
showed no additional species.
2.4. Mechanistic considerations
On the basis of the results obtained from the catalytic and
operando experiments, we feel that the hydroformylation of 1-
hexene catalyzed by the neutral rhodium complexes 2–6 is a
complicated process, probably involving different catalytically
active species depending on the CO pressure, reaction time and
nature of the co-ligands.
Unlike many examples of hydroformylation reactions catalyzed
by rhodium diphosphine complexes [14], no evidence of the forma-
tion of rhodium hydride species was obtained, even under operando
conditions on the short timescale of IR spectroscopy (10−13 s).
Furthermore, the catalytic activity does not depend on the Lewis
basicity of the solvent (Table 2, entry 1 vs. 4), which rules out
the participation of the latter in the activation of hydrogen. The
scarce regioselectivity observed for the hydroformylation reac-
tions is consistent with the loss of the P–O chelating structure,
especially on increasing the CO-partial pressure (Table 2, entry 2
vs. 1), as catalysts supported by firmly chelating ligands gener-
ally exhibit a higher selectivity in the linear product [14]. In this
regard, it is worth recalling that a similarly scarce regioselectivity
in the linear aldehyde has been recently obtained with a ␬2-P,O-1ꢀ-
(diphenylphosphino)ferrocenecarboxylate rhodium precursor that
would form a ␬1-P species under hydroformylation conditions [15].
A possible, yet highly speculative, catalytic cycle as shown in
Scheme 3 might comprise a ligand assisted H2-activation, generat-
ing the ␬1-O mono-hydride species (Scheme 3, a) and a closure of
the catalytic cycle either by intramolecular protonation of the acyl
complexes to re-generate the ␬2-P,O mono- or di-carbonyl precur-
sors (c) or by reaction of the acyl complexes with H2 in order to
Fig. 4. ORTEP plot of 9. Thermal ellipsoids are shown at the 30% probability
level. Hydrogen atoms are omitted, except the one attached to N(1) showing a
˚
˚
short intermolecular contact to O(3) of 1.914(21) A. Selected bond distances in [A]:
P(1)–C(1) = 1.8536(14); P(1)–C(7) = 1.8414(15); P(1)–C(14) = 1.8488(15).
an excess of NEt3 (4 equiv.) showed the complete conversion of 2
into 9 just after 1.5 h.
The authentication of 9 was achieved by its independent syn-
thesis and characterization by means of both NMR spectroscopy
and a single crystal X-ray structure determination. An ORTEP
drawing of 9 is presented in Fig. 4 together with selected bond
distances.
In conclusion, on the basis of operando spectroscopic studies,
there is little doubt that the hydroformylation of 1-hexene with 7
is largely biased by the instability of the catalyst, due to the pres-
ence of the strong base NEt3 which sequestrates the HL ligand, thus
Scheme 3. Proposed catalytic cycle operative in the hydroformylation of 1-hexene with the neutral rhodium complexes 2–6.

62
L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
re-generate the ␬1-O mono-hydride species (b), that, in turn, may
originate a P–O free catalyst (d).
[18]. Elemental analyses were performed using a Carlo Erba Model
1106 elemental analyzer. Infrared spectra were recorded on a
FT-IR PerkinElmer BX spectrometer. The HPIR cell, constructed at
ICCOM-CNR, is constituted by a 75 ml autoclave, equipped with
ZnS windows (4 mm thickness, 8 mm diameter, optical path length
0.2 mm) [10c,12].
3. Conclusion
In this work are reported the synthesis and characterization
of a series of neutral rhodium(I) complexes containing the ␬2-
O,P 2-(bis(2-methoxyphenyl)phosphino)benzenesulfonate ligand.
Some of these complexes have been employed as precursors to
catalyze the hydroformylation reaction of 1-hexene in THF in the
presence of a 1:2 CO/H2 mixture. The regioselectivity in n-heptanal
were quite modest, ranging from 68 to 70%. Lower regioselec-
tivity was found for catalytic reactions either lasting 2 h instead
of 1 h or performing the catalytic reactions in the presence of a
1:1 CO/H2 mixture. Operando spectroscopic studies by means of
HPIR and HPNMR techniques proved the absence of Rh-H species
in the catalytic cycle of sufficient lifetime to be detected on the
fast timescale of IR spectroscopy. Operando HPIR and HPNMR spec-
troscopic experiments have shown that either rhodium(I)-mono
or di-carbonyl complexes are resting state of the catalytic cycle,
most likely in equilibrium with other species containing, besides
the eventual extra phosphine ligand, either the zwitterionic ligand
HL in the ␬1-O bonding mode or carbonyl ligands. A rhodium cat-
alyst containing CO ligands, together with the ammonium salt of
the HL ligand, is actually formed by adding NEt3 to the catalytic
mixture.
4.2. Preparation of complexes
1. [Rh(␩4-COD)Cl]2 (91.9 mg, 0.19 mmol) was dissolved in a
deareated (3:1, v:v) solvent mixture of CH2Cl2 and acetone (10 ml).
AgPF6 (98.9 mg, 0.39 mmol) was added to this solution and
the resulting suspension was allowed to stir for 20 min. After-
wards, the suspension was filtered through a small plug of Celite
which was washed with CH2Cl2 (2 ml). Then ligand HL (150.0 mg,
0.37 mmol) was added to the solution, which was allowed to stir
for 1 h, followed by evaporation under vacuum. The yellow–orange
residue was suspended in diethyl ether (10 ml) stirred for 10 min,
then filtered under nitrogen and dried in a stream of nitrogen.
Yield: 215.2 mg (95%). Anal. calcd. for C28H30O5PRhS (612.2): C,
54.93; H, 4.90. Found: C, 54.88; H, 4.82. 1H NMR (300.13 MHz,
CDCl3, 25 ◦C): ı 1.90 (m, 2H, CHHꢀ), 2.01 (m, 2H, CꢀHHꢀ), 2.43
(m, 4H, CHHꢀ + CꢀHHꢀ), 2.81 (br. s, 2H, CH(trans O)), 3.61 (s, 6H,
OCH3), 5.36 (br. s, 2H, CH(trans P)), 6.96–7.56 (m, 10H, Ar),
3
4
8.02 (br. s 1H, o-H-Ar), 8.09 (ddd, JH,H = 7.8 Hz, JP,H = 4.1 Hz,
4JH,H = 0.9 Hz, 1H, m-H-Ar). 13C{ H} NMR (75.49 MHz, CDCl3, 25 ◦C):
1
4
ı 28.80 (d, JP,H = 12.5 Hz, CꢀHHꢀ), 32.67 (s, CHHꢀ), 55.06 (s, OCH3),
2
69.98 (d, JP,C = 14.8 Hz, CH(trans O)), 104,30 (s, CH(trans P)),
1
4. Experimental
110.45–160.39 (Ar). 31P{ H} NMR (121.49 MHz, CDCl3, 25 ◦C): ı 8.09
(d, 1JRh,P = 151.8 Hz).
4.1. General
2. CO was bubbled through a deareated CD2Cl2 (1.5 ml) solu-
tion of 1 (25.0 mg, 0.04 mmol) for 5 min. Then the solution was
transferred into a 5 mm NMR tube, followed by the acquisition
of the NMR data. 1H NMR (300.13 MHz, CDCl3, 25 ◦C): ı 3.84
All reactions and manipulations were carried out under a
nitrogen atmosphere by using standard Schlenk-type techniques.
The solvents were distilled from suitable dehydrating reagents and
were deoxygenated before use. The reagents were used as pur-
chased from Aldrich or Fluka, unless stated otherwise. Compounds
[RhCl(␩4-COD)]2 [16], bis(2-methoxyphenyl)methoxyphosphine
[3d], 2-(bis(2-methoxyphenyl)phosphino)benzenesulfonic acid
[3d], and bis(2-methoxyphenyl)phosphine [17] were prepared
according to the literature methods. Hydroformylation reactions
were performed in a 320 ml stainless steel autoclave, constructed
at ICCOM-CNR (Florence, Italy), equipped with a magnetic drive
stirrer and a Parr 4842 temperature and pressure controller. GC
analyses were performed with a Shimadzu GC 2010 apparatus,
equipped with a SPB-5 Supelco fused silica capillary column
(60 m × 0.25 mm, 0.25 ␮m film thickness) employing toluene as
internal standard, while GC–MS analyses were performed with
a Shimadzu QP2100S apparatus, equipped with a SPB-1 Supelco
fused silica capillary column (30 m × 0.25 mm, 0.25 ␮(m film
thickness). Deuterated solvents for routine NMR measurements
3
(s, 6H, OCH3), 6.94–7.62 (m, 11H, Ar), 8.15 (ddd, JH,H = 7.7 Hz,
1
4JP,H = 4.9 Hz, 4JH,H = 1.2 Hz, 1H, m-H–Ar). 13C{ H} NMR (75.49 MHz,
CDCl3, 25 ◦C): ı 55.09 (s, OCH3), 111.10–160.70 (Ar), 180.43 (br.
s, CO), 184.58 (br. s, CO). 13C{ H} NMR (75.49 MHz, CDCl3,
1
−60 ◦C): ı 55.20 (s, OCH3), 56.20 (s, OCH3), 111.10–160.90 (Ar),
2
1
179.74 (dd, JP,C = 115.9 Hz, JRh,C = 65.4 Hz, CO(trans P)), 184.72 (d,
1JRh,C = 55.3 Hz, CO(trans O)). 31P{ H} NMR (121.49 MHz, CDCl3,
1
1
25 ◦C): ı −8.39 (d, JRh,P = 131.3 Hz). IR (CH2Cl2): ꢀ(C O) 2019 and
2097 cm−1
.
3–8. Compound 1 (200.0 mg, 0.33 mmol) was dissolved in
deareated CH2Cl2 (10 ml). Through this solution was succes-
sively bubbled CO for 5 min and nitrogen for 5 min at room
temperature, followed by the addition of the appropriate lig-
and (0.33 mmol). The obtained solution was then allowed to
stir for 30 min at room temperature, followed by its concen-
tration to a small volume (4 ml) and on addition of diethyl
ether (10 ml) a yellow, micro-crystalline compound precipitated,
which was filtered off, washed with diethyl ether (5 ml) and
dried in a stream of nitrogen. 3: Yield 196.6 mg (75%). Anal.
were dried over activated molecular sieves. 1H, 13C{ H}, 31P{ H}
NMR spectra were obtained on either a Bruker Avance II DRX 300
spectrometer (300.13, 75.49, 121.49 MHz, respectively) or a Bruker
Avance DRX-400 spectrometer (400.13, 100.62 and 161.98 MHz,
respectively). Chemical shifts (ı) are reported in ppm relative
to TMS, referenced to the chemical shifts of residual solvents
1
1
calcd. for
C39H33O6P2RhS (794.3): C, 58.97; H, 4.15. Found:
C, 58.82; H, 4.08. 1H NMR (300.13 MHz, CDCl3, 25 ◦C): ı 3.70
3
(s, 6H, OCH3), 6.95–7.68 (m, 26H, Ar), 8.13 (ddd, JH,H = 7.5 Hz,
4JP,H = 4.9 Hz, 4JH,H = 1.1 Hz, 1H, m-H–Ar). 13C{ H} NMR (75.49 MHz,
1
resonances (1H and 13C NMR) or 85% H3PO4. 13C{ H} NMR spectra
CDCl3, 25 ◦C):
ı
55.09 (s, OCH3), 111.12–160.70 (Ar), 189.37
1
1
of the rhodium carbonyl complexes 2–8 were acquired after an
in situ preparation of each complex with enriched 13CO (i.e. 13C
content >99%). HPNMR experiments were carried out on a Bruker
ACP 200 spectrometer, operating at 200.13 and 81.01 MHz for 1H
(s, CO). 13C{ H} NMR (75.49 MHz, CDCl3, −60 ◦C): ı 55.13 (s,
OCH3), 111.15–160.90 (Ar), 189.36 (dt, 1JRh,C = 77.2 Hz, 2JP,C = 16.9 Hz,
1
CO). 31P{ H} NMR (121.49 MHz, CDCl3, 25 ◦C):
ı
−0.27 (dd,
1JRh,P = 130.1 Hz, JP,P = 327.3 Hz, L), 28.87 (dd, JRh,P = 135.4 Hz,
2JP,P = 327.3 Hz, P). IR (CH2Cl2): ꢀ(C O) 1983 cm−1; IR (KBr): ꢀ(C O)
1975 cm−1. 4: Yield 197.4 mg (74%). Anal. calcd. for C36H35O9P2RhS
(808.3): C, 53.49; H, 4.33. Found: C, 53.35; H, 4.27. 1H NMR
2
1
1
and 31P{ H}, respectively, using a 10-mm sapphire NMR tube
(Saphikon, Milford, NH), equipped with a titanium high-pressure
charging head constructed at the ICCOM-CNR (Florence, Italy)

L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
63
(300.13 MHz, CDCl3, 25 ◦C): ı 3.63 (s, 6H, OCH3), 3.64 (s, 6H,
OCH3), 4.06 (d, JP,H = 13.9 Hz, 3H, POCH3), 6.85–7.73 (m, 19H,
ether (5 ml) was added to the solid residue and the suspension
was stirred for half an hour. Afterwards the solid was filtered off,
washed with diethyl ether (5 ml) and dried in a stream of nitro-
gen. Yield: 97.0 mg (88%). Anal. calcd. for C26H34NO5PS (503.3):
C, 62.04; H, 6.75. Found: C, 62.12; H, 6.85. 1H NMR (300.13 MHz,
CDCl3, 25 ◦C): ı 1.26 (t, 3JH,H = 7.2 Hz, 9H, CH3), 3.15 (q, 3JH,H = 7.2 Hz,
6H, CH2), 3.63 (s, 6H, OCH3), 6.67–7.35 (m, 11H, Ar), 8.24 (ddd,
3
3
4
4
Ar), 8.11 (ddd, JH,H = 7.6 Hz, JP,H = 4.9 Hz, JH,H = 1.1 Hz, 1H, m-
H–Ar). 13C{ H} NMR (75.49 MHz, CDCl3, 25 ◦C): ı 55.09 (s, OCH3),
1
55.48 (s, OCH3), 57.89 (s, POCH3), 110.93–160.76 (Ar), 188.97
(s, CO). 13C{ H} NMR (75.49 MHz, CDCl3, −60 ◦C): ı 55.26 (s,
1
OCH3), 55.80 (s, OCH3), 58.20 (s, POCH3), 110.88–160.28 (Ar),
189.13 (br d, 1JRh,C = 50.3 Hz, CO). 31P{ H} NMR (121.49 MHz, CDCl3,
3JH,H = 7.8 Hz, JP,H = 4.6 Hz, JH,H = 1.4 Hz, 1H, m-H-Ar), 10.20 (br. s,
1
4
4
25 ◦C): ı 0.32 (dd, JRh,P = 123.1 Hz, JP,P = 373.1 Hz, L−), 116.20 (dd,
1H, NH). 13C{ H} NMR (75.49 MHz, CDCl3, 25 ◦C): ı 8.43 (s, CH3),
1
2
1
1JRh,P = 156.7 Hz, 2JP,P = 373.1 Hz, P). IR (CH2Cl2): ꢀ(C O) 1983 cm−1
IR (KBr): ꢀ(C O) 1984 cm−1. 5: Yield 159.2 mg (70%). Anal. calcd.
;
45.99 (s, CH2), 55.34 (s, OCH3), 109.96–160.82 (Ar). 31P{ H} NMR
1
(121.49 MHz, CDCl3, 25 ◦C): ı −28.01 (s).
for C27H30N3O6P2RhS (689.2): C, 47.05; H, 4.35. Found: C, 47.12;
H, 4.40. 1H NMR (400.13 MHz, CDCl3, 25 ◦C): ı 3.73 (s, 6H, OCH3),
4.4. Crystal structure determinations
4.35 (s, 6H, NCH2P), 4.59 (s, 6H, NCH2N), 6.99–7.52 (m, 11H, Ar),
3
4
1
8.21 (dd, JH,H = 5.4 Hz, JP,H = 3.0 Hz, 1H, m-H-Ar). 13C{ H} NMR
(100.62 MHz, CDCl3, 25 ◦C): ı 51.50 (d, 1JP,C = 13.2 Hz, NCH2P), 55.28
(s, OCH3), 73.30 (d, 3JP,C = 5.7 Hz, NCH2N), 111.19–160.57 (Ar), 188.25
Crystals of 1 and 9, suitable for a single crystal X-ray diffrac-
tion analysis were obtained by diffusion of toluene into a layered
CH2Cl2 solution of either compound at room temperature. X-ray
diffraction intensity data were collected at 173 K on Oxford Diffrac-
tion CCD diffractometers with graphite monochromated Mo K˛
(s, CO). 13C{ H} NMR (100.62 MHz, CDCl3, −60 ◦C): ı 51.10 (d,
1
1JP,C = 12.7 Hz, NCH2P), 55.05 (s, OCH3), 56.05 (s. OCH3), 73.30 (br. s,
NCH2N), 111.10–160.16 (Ar), 188.35 (d, 1JRh,C = 72.4 Hz, CO). 31P{ H}
1
˚
˚
radiation (ꢁ = 0.71073 A) and Cu K˛ radiation (ꢁ = 1.54184 A) using
␻-scans for compound 1 and 9, respectively. Cell refinement, data
reduction and empirical absorption correction were carried out
with the Oxford diffraction software and SADABS [19a]. All struc-
ture determination calculations were performed with the WINGX
package [19b] with SIR-97 [19c], SHELXL-97 [19d], and ORTEP-3
programs [19e]. Final refinements based on F2 were carried out
with anisotropic thermal parameters for all non-hydrogen atoms
while the hydrogen atoms were included in calculated positions
and refined using a riding model with isotropic Uiso(H) values
depending on the Ueq(C) of the adjacent carbon atoms. Rele-
vant crystallographic data and structure refinement details are
summarized in Table 3. CCDC-675841 and CCDC-675840 con-
tain the supplementary crystallographic data for compound 1
and 9, respectively. Copies of the data can be obtained free of
charge from the Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, UK (fax: (+44) 1223/336033; e-mail:
deposit@ccdc.cam.ac.uk).
NMR (161.98 MHz, CDCl3, 25 ◦C): ı −46.47 (dd, JRh,P = 127.9 Hz,
1
2JP,P = 319.1 Hz, P), −3.95 (dd, JRh,P = 129.9 Hz, JP,P = 319.1 Hz, L).
IR (CH2Cl2): ꢀ(C O) 1983 cm−1; IR (KBr): ꢀ(C O) 1983 cm−1. 6:
Yield 187.5 mg (73%). Anal. calcd. for C35H33O8P2RhS (778.3): C,
54.01; H, 4.24. Found: C, 53.94; H, 4.17. 1H NMR (400.13 MHz,
CDCl3, 25 ◦C): ı 3.68 (s, 6H, OCH3), 3.76 (s, 6H, OCH3), 6.77
1
2
1
2
(dd, JP,H = 381.3 Hz, JRh,H = 3.4 Hz, 1H, pH), 6.85–7.49 (m, 17H,
Ar), 7.79 (ddd, JP,H = 13.6 Hz, JH,H = 7.2 Hz, JH,H = 1.6 Hz, 2H, o-
3
3
4
3
4
4
H-Ar(P)), 8.16 (ddd, JH,H = 7.6 Hz, JP,H = 8.0 Hz, JH,H = 1.2 Hz, 1H,
m-H-Ar(L)). 13C{ H} NMR (100.62 MHz, CDCl3, 25 ◦C): ı 55.18
1
(s, OCH3), 55.61 (s, OCH3), 110.46–160.77 (Ar), 188.86 (s, CO).
13C{ H} NMR (100.62 MHz, CDCl3, −60 ◦C): ı 55.16 (s, OCH3),
1
55.74 (s, OCH3), 110.20–160.11 (Ar), 189.19 (br. s, CO). 31P{ H}
1
NMR (161.98 MHz, CDCl3, 25 ◦C): ı −26.14 (dd, JRh,P = 133.3 Hz,
1
2JP,P = 340.9 Hz, P), 1.10 (dd, JRh,P = 133.8 Hz, JP,P = 340.9 Hz, L). IR
(CH2Cl2): ꢀ(C O) 1984 cm−1; IR (KBr): ꢀ(C O) 1969 cm−1. 7: Yield
135.8 mg (65%). Anal. calcd. for C27H33NO6PRhS (633.2): C, 51.21;
H, 5.21. Found: C, 51.15; H, 5.12. 1H NMR (300.13 MHz, CDCl3,
1
2
25 ◦C): ı 1.39 (t, JH,H = 7.2 Hz, 9H, CH3), 2.99 (dq, JP,H = 2.4 Hz,
3
4
3JH,H = 7.2 Hz, 6H, CH2), 3.67 (s, 6H, OCH3), 6.94–7.48 (m, 11H,
Table 3
Crystallographic data and structure refinement details for 1 and 9
3
4
4
Ar), 8.14 (ddd, JH,H = 7.8 Hz, JP,H = 4.6 Hz, JH,H = 1.4 Hz, 1H, m-H-
Ar). 13C{ H} NMR (75.49 MHz, CDCl3, 25 ◦C): ı 10.60 (s, CH3),
1
1
9
49.22 (s, CH2), 54.88 (s, OCH3), 110.93–160.31 (Ar), 187.72 (dd,
Formula
Molecular weight
Crystal habit
Crystal size (mm)
Temperature (K)
C
28H30O5PRhS
C26H34NO5PS
503.57
White prism
0.2 × 0.2 × 0.2
173(2)
1JRh,C = 85.2 Hz, JP,C = 21.4 Hz, CO). 31P{ H} NMR (121.49 MHz,
2
1
612.46
CDCl3, 25 ◦C): ı 21.87 (d, JRh,P = 155.1 Hz). IR (CH2Cl2): ꢀ(C O)
1
Yellow prism
0.4 × 0.2 × 0.2
173(2)
1970 cm−1; IR (KBr): ꢀ(C O) 1963 cm−1. 8: Yield 135.8 mg (70%).
Anal. calcd. for C26H23NO6PRhS (611.2): C, 51.09; H, 3.76. Found: C,
50.93; H, 3.65. 1H NMR (400.13 MHz, CDCl3, 25 ◦C): ı 3.75 (s, 6H,
OCH3), 6.97 + 7.12 + 7.34 (m, 8H, Ar(L)), 7.41 (pseudo t, 3JH,H = 8.0 Hz,
˚
Wavelength (A)
0.71069
Orthorhombic
Pbca
1.54184
Monoclinic
P2(1)/n
Crystal system
Space group
3
˚
a (A)
17.475(5)
15.873(5)
18.568(5)
90.000
90.000
90.000
5150(3)
8
1.580
0.844
2512
3.72–28.61
5685
14.6465(2)
11.4819(1)
16.7465(3)
90.000
112.982(2)
90.000
2592.72(6)
4
1.290
1.991
1072
4.80–72.72
5045
5045
2H, m-H-Ar(Py)), 7.51 (m, 3H, Ar(L)), 7.85 (tt, JH,H = 7.7 Hz,
˚
b (A)
4JH,H = 1.5 Hz, 1H, p-H-Ar(Py)), 8.24 (ddd, 3JH,H = 7.8 Hz, 4JP,H = 4.6 Hz,
˚
c (A)
4JH,H = 1.0 Hz, 1H, m-H-Ar(L)), 8.87 (m, 2H, o-H-Ar(Py)). 13C{ H}
1
˛ (◦)
ˇ (◦)
ꢂ (◦)
NMR (100.62 MHz, CDCl3, 25 ◦C): ı 55.17 (s, OCH3), 111.39–160.53
(Ar), 188.07 (d, JRh,C = 79.1 Hz, CO). 13C{ H} NMR (100.62 MHz,
1
1
3
˚
CDCl3, −60 ◦C): ı 55.17 (s, OCH3), 55.78 (s, OCH3), 111.38–160.41
Volume (A )
Z
(Ar), 188.28 (dd, JRh,C = 83.6 Hz, JP,C = 19.8 Hz, CO). 31P{ H} NMR
(161.98 MHz, CDCl3, 25 ◦C): ı 19.69 (d 1JRh,P = 155.0 Hz). IR (CH2Cl2):
1
2
1
ꢃcalc (g cm−3
)
ꢄ (mm−1
F(000)
)
ꢀ(C O) 1979 cm−1; IR (KBr): ꢀ(C O) 1981 cm−1
.
ꢅ range (◦)
Reflections collected
Independent reflections
Refined parameters
R1 (2ꢆ(I)]
4.3. Preparation of 9
5685
329
0.0362
316
0.0390
To a solution of ligand HL (89.8 mg, 0.22 mmol) in CH2Cl2 (3 ml)
was added triethylamine (62.2 ␮l, 0.45 mmol) and the solution
obtained was allowed to stir for 20 min, followed by its concen-
tration to dryness by means of a vacuum pump. Then diethyl
R1 (all data)
0.0575
1.043
0.762/−0.541
0.0411
1.019
0.320/−0.274
Goodness-of-fit on F2
−3
˚
Largest diff. peak and hole (A
)

64
L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
4.5. Batch hydroformylation tests
0.08 mmol) has been carried out, adding NEt3 to the initial (1:1, v:v)
THF/CD2Cl2. solution mixture of 7. HPNMR experiments with com-
pounds 1 and 3–6 were carried out analogously to that described
above for 7.
In a typical experiment, deareated THF (30 ml) and 1-hexene
(250 ␮l, 2.0 mmol) were introduced into an autoclave (320 ml),
which had been previously evacuated by means of a vacuum pump
and charged with the catalyst precursor (0.01 mmol). In those cases
where the catalytic reactions were carried out in the presence of an
excess of NEt3 (0.04 mmol), the reagent was added to the THF solu-
tion. Then the autoclave was charged with a (1:2) CO/H2 gas mixture
to a total pressure of 20 bar at room temperature, followed by heat-
ing it by means of an electric heater to 60 ◦C. Once the desired
reaction temperature was reached, the total pressure of the appro-
priate gas mixture was adjusted to 42 bar and stirring was started
(600 rpm). After the desired reaction time, the autoclave was cooled
to 0 ◦C by means of an ice-acetone bath and unreacted gases were
vented off and the reaction solution was analyzed by GC–MS as
well as by GC after toluene (100 ␮l) had been added to the catalysis
solution.
Acknowledgments
The authors thank Dr. Marco Frediani (University of Florence)
for the assistance in the GC analyses. The Ente Cassa di Risparmio
di Firenze (Italy) is gratefully acknowledged for granting a 400 MHz
spectrometer as well as for supporting the ICCOM-CNR through the
project HYDROLAB. Thanks are due also to the European Commis-
sion for financing the Network of Excellence, IDECAT (NoE contract
no. NMP3-CT-2005-011730).
References
[1] (a) A. Bader, E. Lindner, Coord. Chem. Rev. 108 (1991) 27–110;
(b) C.S. Slone, D.A. Weinberger, C.A. Mirkin, Prog. Inorg. Chem. 48 (1999)
233–350;
4.6. Operando spectroscopic studies
(c) S.D. Ittel, L. Johnson, M. Brookhart, Chem. Rev. 100 (2000) 1169–1203;
(d) V.C. Gibson, S.K. Spitzmesser, Chem. Rev. 103 (2003) 283–315.
[2] P. Braunstein, F. Naud, Angew. Chem. Int. Ed. Engl. 40 (2001) 690–699.
[3] (a) P. Kuhn, D. Se´meril, D. Matt, M.J. Chetcuti, P. Lutz, Dalton Trans. (2007)
515–528;
4.6.1. HPIR experiments
Typically, the catalyst precursor (1, 3, 4, 5, 6 or 7, 0.063 mmol)
was dissolved in deareated THF (25 ml) containing 1-hexene
(1.6 ml, 12.6 mmol). In those cases where the IR study was car-
ried out in the presence of an excess of NEt3 (0.252 mmol) this
latter reagent was added to the THF solution of the catalyst pre-
cursor. The resulting solution was introduced into the HPIR cell
by means of a syringe under a stream of nitrogen, followed by
pressurization with a 1:2 CO/H2 gas mixture to a total pressure of
36 bar at room temperature. The HPIR cell was placed into the IR
spectrometer and stirring was started (1500 rpm), followed by the
acquisition of IR spectra in the region 1800–2300 cm−1 at room
temperature. Afterwards the HPIR cell was heated to 60 ◦C (total
gas pressure of 42 bar) and IR spectra were acquired every 20 min
for 1 h. At the end of the experiment, the HPIR cell was cooled
to room temperature, the gases were vented off and the solu-
tion was analyzed by GC, showing the same regioselectivity of
the aldehydes formed as observed in parallel batch catalytic reac-
tions.
ˇ
(b) P. Steˇpnicˇka, Eur. J. Inorg. Chem. 19 (2005) 3787–3803;
(c) C.M. Reisinger, R.J. Nowack, D. Volkmer, B. Rieger, Dalton Trans. (2007)
272–278;
(d) E. Drent, R. van Dijk, R. van Ginkel, B. van Oort, R.I. Pugh, Chem. Commun.
(2002) 964–965;
(e) E. Drent, R. van Dijk, R. van Ginkel, B. van Oort, R.I. Pugh, Chem. Commun.
(2002) 744–745;
(f) T. Kochi, S. Noda, K. Yoshimura, K. Nozaki, J. Am. Chem. Soc. 129 (2007)
8948–8949;
(g) S. Luo, J. Vela, G.R. Lief, R. Jordan, J. Am. Chem. Soc. 129 (2007) 8946–8947;
(h) S. Liu, S. Borkar, D. Newsham, H. Yennawar, A. Sen, Organometallics 26 (2007)
210–216;
(i) T. Kochi, K. Yoshimura, K. Nozaki, Dalton Trans. (2006) 25–27;
(j) A.K. Hearly, R.J. Nowack, B. Rieger, Organometallics 24 (2005) 2755–2763;
(k) L. Bettucci, C. Bianchini, C. Claver, E.J. Garcia Suarez, A. Ruiz, A. Meli, W. Ober-
hauser, Dalton Trans. (2007) 5590–5602;
(l) T. Schultz, A. Pfaltz, Synthesis 6 (2005) 1005–1011.
[4] (a) W. Keim, Angew. Chem. Int. Ed. Engl. 29 (1990) 235–244;
(b) W. Keim, F.H. Kowaldt, R. Goddard, C. Kru¨ ger, Angew. Chem. Int. Ed. Engl. 17
(1978) 466–467.
[5] (a) J.J. Gambaro, W.H. Hohman, D.W. Meek, Inorg. Chem. 28 (1989) 4154–4159;
(b) R.R. Schrock, J.A. Osborn, J. Am. Chem. Soc. 93 (1971) 2397–2407.
[6] (a) H. Schumann, H. Hemling, V. Ravindar, Y. Badrieh, J. Blum, J. Organomet.
Chem. 469 (1994) 213–219;
(b) A.D. Burrows, R.W. Harrington, M.F. Mahhon, M.T. Palmer, F. Senia, M. Var-
rone, Dalton Trans. (2003) 3717–3726.
[7] (a) J. Cipot, M.J. Ferguson, M. Stradiotto, Inorg. Chim. Acta 359 (2006)
2780–2785;
4.6.2. HPNMR experiments
A solution of compound 7 (12.7 mg, 0.02 mmol) in a deareated
(1:1, v:v) solvent mixture of THF/CD2Cl2 (2 ml) was transferred
under nitrogen into a 10 mm sapphire tube, which was placed into
a NMR probe head at room temperature followed by the acquisi-
(b) Y. Nishihara, K. Nara, Y. Nishide, K. Osakada, Dalton Trans. (2004) 1366–1375;
(c) D. Selent, W. Baumann, R. Kempe, A. Spannenberg, D. Ro¨ttger, K.D. Wiese, A.
Bo¨rner, Organometallics 22 (2003) 4265–4271.
1
tion of a 31P{ H} NMR spectrum at this temperature. Then, the NMR
tube was removed from probe and 1-hexene (500 ␮l, 4.0 mmol) was
[8] (a) C. Bianchini, A. Meli, W. Oberhauser, A.M. Segarra, C. Claver, E.J. Garcia Suarez,
J. Mol. Catal. A: Chem. 265 (2007) 292–305;
added under nitrogen. Afterwards, a 31P{ H} NMR spectrum was
1
(b) I.M. Angulo, E. Bouwman, S.M. Lok, M. Lutz, W.P. Mul, A.L. Spek, Eur. J. Inorg.
Chem. (2001) 1465–1473;
acquired at room temperature, followed by charging the sapphire
tube with CO (14 bar) outside the NMR probe at room tempera-
(c) R.C. van der Drift, E. Bouwman, E. Drent, H. Kooijman, A.L. Spek, A.B. van
Oort, W.P. Mul, Organometallics 21 (2002) 3401–3407;
(d) K.R. Dunbar, J.-S. Sun, J. Chem. Soc. Chem. Commun. (1994) 2387–2388.
[9] (a) C. Bianchini, W. Oberhauser, A. Orlandini, C. Giannelli, P. Frediani,
Organometallics 24 (2005) 3692–3702;
1
ture and acquiring a 31P{ H} NMR spectrum at room temperature.
The NMR tube was then again removed from NMR probe head and
charged with hydrogen (28 bar), in order to have a CO/H2 ratio of
1
1:2 inside the NMR tube, followed by the acquisition of a 31P{ H}
´
(b) M. Rubio, A. Sua´rez, E. Alvarez, C. Bianchini, W. Oberhauser, M. Peruzzini, A.
Pizzano, Organometallics 26 (2007) 6428–6436;
(c) K. Nozaki, N. Sakai, T. Nanno, T. Higashijima, S. Mano, T. Horiuchi, H. Takaya,
J. Am. Chem. Soc. 119 (1997) 4413–4423.
NMR spectrum at room temperature. Afterwards, the sapphire tube
was heated to 60 ◦C in a temperature-interval of 10 ◦C, lasting the
catalytic reaction at 60 ◦C for 1 h. At each reaction temperature
[10] (a) P.G. Jessop, R.H. Morris, Coord. Chem. Rev. 121 (1992) 155–284;
(b) D.M. Heinekey, W.J. Oldham Jr., Chem. Rev. 93 (1993) 913–926;
(c) C. Bianchini, S. Moneti, M. Peruzzini, F. Vizza, Inorg. Chem. 36 (1997)
5818–5825.
[11] J.P. Collman, J.A. Belmont, J.I. Brauman, J. Am. Chem. Soc. 105 (1983) 7288–7294.
[12] B. Heaton, Mechanisms in Homogeneous Catalysis—A Spectroscopic Approach,
Wiley-VCH, Weinheim, 2004.
[13] C. Bianchini, H.M. Lee, A. Meli, F. Vizza, Organometallics 19 (2000) 849–853.
[14] (a) P.W.N.M. van Leeuwen, P.C.J. Kamer, J.N.H. Reek, P. Dierkes, Chem. Rev. 100
(2000) 2741–2769;
1
a
31P{ H} NMR spectrum was acquired and at 60 ◦C, NMR spec-
tra were acquired at a time-interval of 20 min, before the probe
head had been cooled to room temperature. Once the probe head
reached room temperature, the last 31P{ H} NMR spectrum was
1
acquired, followed by the depressurization of the sapphire tube
and the analysis of the NMR solution by GC–MS. An analogous
NMR experiment in the presence of an excess of NEt3 (11.1 ␮l,

L. Bettucci et al. / Journal of Molecular Catalysis A: Chemical 291 (2008) 57–65
65
(b) B. Cornils, W.A. Herrmann (Eds.), Applied Homogeneous Catalysis with
Organometallic Compounds, Wiley-VCH, Weinheim, 1996.
[19] (a) G.M. Sheldrick, SADABS, Program for Empirical Absorption Corrections, Uni-
versity of Go¨ttingen, Go¨ttingen, Germany, 1986;
ˇ
[15] (a) P. Steˇpnicˇka, I. Cı´sarˇova´, J. Chem. Soc. Dalton Trans. (1998) 2807–2811;
(b) L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837–838;
(c) A. Altomare, M.C. Burla, M. Cavalli, G.L. Cascarano, C. Giacovazzo, A. Gagliardi,
G.G. Moliterni, G. Polidori, R. Spagna, J. Appl. Crystallogr. 32 (1999) 115–119;
(d) G.M. Sheldrick, SHELX-97, University of Go¨ttingen, Go¨ttingen, Germany,
1997;
ˇ
(b) A.M. Trzeciak, P. Steˇpnicˇka, E. Mieczyn´ ska, J.J. Zio´ łkowski, J. Organomet.
Chem. 690 (2005) 3260–3267.
[16] G. Giordano, R.H. Crabtree, Inorg. Synth. 19 (1979) 218.
[17] C. Bianchini, G. Lenoble, W. Oberhauser, S. Parisel, F. Zanobini, Eur. J. Inorg. Chem.
(2005) 4794–4800.
(e) M.N. Burnett, C.K. Johnson, ORTEP-3 Report ORNL-6895, Oak Ridge National
Laboratory, Oak Ridge, TN, 1996.
[18] C. Bianchini, A. Meli, A. Traversi, A. Ital. Pat. FI A000025, 1997.