Hydridorhodium(I) complexes with amphiphilic polyether phosphines. NMR study and biphasic hydroformylation of 1-octene

Article abstract of DOI:10.1016/S0022-328X(03)00040-8

The formation of hydridorhodium(I) complexes with the amphiphilic phosphines was studied by NMR methods and IR spectroscopy. Two approaches were used: (1) the substitution reaction of triphenylphosphine in RhH(CO)(PPh₃)₃ by an amphiphilic phosphine and (2) a reaction between the precursor Rh(acac) (CO)₂, syngas and the amphiphilic ligand in hydroformylation reaction conditions. Both approaches show the formation of P-coordinated hydridorhodium(I) complexes, and no relevant chelated (P, O) complexes were recognised. The biphasic rhodium-catalysed hydroformylation of 1-octene with the following water-soluble amphiphilic phosphines was studied: 4-(R)C₆H₄ (OCH₂CH₂)_nP(Ph)CH₂CH ₂SO₃Na (R = 2,2,4,4-tetramethylbutyl, n? = 1.4, 5.1, 11.2, R = n-nonyl, n? = 1.6, 5.6, 11.4), and RP(Ph)CH₂CH₂SO₃Na (R = n-octyl, CH₃(OCH₂CH₂) ₂OCH₂CH₂). Ligands with a hydrophobic group and a short polyether chain led to the higher conversion. This result is ascribed to the ability of these ligands to increase the metal concentration in the organic phase.

Full text of DOI:10.1016/S0022-328X(03)00040-8

Journal of Organometallic Chemistry 669 (2003) 172Á181
/
www.elsevier.com/locate/jorganchem
Hydridorhodium(I) complexes with amphiphilic polyether
phosphines
NMR study and biphasic hydroformylation of 1-octene
Antoni Solsona a, Joan Suades a,*, Rene´ Mathieu b
a Departament de Qu´ımica, Edifici C, Universitat Auto`noma de Barcelona, Bellaterra, Barcelona 08193, Spain
b Laboratoire de Chimie de Coordination du CNRS, 205 route de Narbonne, 31077 Toulouse Cedex 4, France
Received 12 November 2002; received in revised form 20 December 2002; accepted 16 January 2003
Abstract
The formation of hydridorhodium(I) complexes with the amphiphilic phosphines 4-(R)C6H4(OCH2CH2)nPPh2 (Rꢀ
/2,2,4,4-
tetramethylbutyl, nꢀ1, n¯ꢀ5) was studied by NMR methods and IR spectroscopy. Two approaches were used: (1) the substitution
/
reaction of triphenylphosphine in RhH(CO)(PPh3)3 by an amphiphilic phosphine and (2) a reaction between the precursor
Rh(acac)(CO)2, syngas and the amphiphilic ligand in hydroformylation reaction conditions. Both approaches show the formation of
P-coordinated hydridorhodium(I) complexes, and no relevant chelated (P, O) complexes were recognised. The biphasic rhodium-
catalysed hydroformylation of 1-octene with the following water-soluble amphiphilic phosphines was studied: 4-
(R)C6H4(OCH2CH2)nP(Ph)CH2CH2SO3Na (Rꢀ
/
2,2,4,4-tetramethylbutyl, n¯ꢀ1:4; 5:1; 11:2; Rꢀn-nonyl, n¯ꢀ1:6; 5:6; 11:4); and
/
RP(Ph)CH2CH2SO3Na (Rꢀn-octyl, CH3(OCH2CH2)2OCH2CH2). Ligands with a hydrophobic group and a short polyether chain
/
led to the higher conversion. This result is ascribed to the ability of these ligands to increase the metal concentration in the organic
phase.
# 2003 Elsevier Science B.V. All rights reserved.
Keywords: Hydridorhodium(I) complexes; Amphiphilic phosphines; Biphasic hydroformylation
1. Introduction
biphasic hydroformylation process can also be applied
to other olefins, but the reaction rates decrease drama-
tically by increasing the alkyl chain of the olefin as a
result of the poor olefin solubility in the aqueous phase
[3]. Several approaches have been proposed to increase
the reaction rate in the biphasic hydroformylation of
higher olefins as the addition of co-solvents [4], triphe-
nylphosphine [5] or conventional surfactants [6]. In
recent years, the use of amphiphilic ligands has been
proposed as a new way to attain this goal [7]. The
ligands designed for this purpose display, in the same
molecule, a long alkyl chain, a hydrophilic group, and
one or more donor atoms. Hence, they can form a
metallic complex with the properties of a surface-active
agent. These compounds act as a bridge between
aqueous and organic phases, and lead to accumulation
of the metallo-surfactants in the interface and to
supramolecular arrangements like micelles or vesicles.
In this way, the reaction rate can be improved because
The synthesis and study of new water-soluble metal
complexes for aqueous and biphasic homogeneous
catalysis is currently one of the most active research
areas in organometallic chemistry [1]. Water-soluble
transition metal complexes have been made using
functionalised ligands that incorporate polar groups.
These hydrophilic complexes can be applied in aqueous
biphasic catalysis, since the catalyst is poorly soluble in
the organic phase and catalyst recovery is easily
rendered by separation of the two phases. An excellent
example of this approach is the biphasic Rhurchemie/
Rhone-Poulenc process for hydroformylation of pro-
pene with the RhH(CO)(TPPTS)3 complex [2]. This
* Corresponding author. Tel.: ꢁ
3101.
E-mail address: joan.suades@uab.es (J. Suades).
/
34-935-81-2893; fax: ꢁ34-935-81-
/
0022-328X/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0022-328X(03)00040-8

A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
173
catalyst is accumulated in a high interface surface
around the substrate. Most of the reported amphiphilic
ligands are functionalised phosphines with hydrophilic
sulphonate groups [8], but sulphated [9], phosphonate
[10], cationic [11] and non-ionic [12] phosphines have
also been reported. In previous papers, we reported the
synthesis of a new family of neutral amphiphilic
arrangements [14]. This result induced us to focus our
attention on the case of non-ionic ligands, and ligands 1
and 2 were retained as model compounds of the
reported amphiphilic polyether ligands. Ligand 1 was
chosen because it is a molecule where the coordination
of the oxygen atom is hindered by electronic and steric
influence of the aryl group bonded to this atom. On the
contrary, ligand 2 is a compound containing polyether
groups which can lead to bidentate (P, O) coordination
modes to the metal [13,15].
The reaction of [HRh(CO)(PPh3)3] with 1 and 2 was
performed in THF with different equivalents of ligand,
and the reaction products were analysed by NMR
spectroscopy. The 1H- and 1H{31P}-NMR spectra
showed unequivocally that a mixture of complexes
with the structures displayed in Scheme 4 was formed
with both ligands. The 1H spectra display in the hydride
region four quadruplets assigned to the hydride of the
four structures coupled with the three phosphorus atoms
of the ligands (Fig. 1). The 1H{31P}-NMR spectra
confirm this assignment since the four signals are now
displayed as four singlets. The assignment of these
phosphines (ligands 1Á
recently, a new group of anionic amphiphilic phosphines
and their palladium metallo-surfactants (ligands 10Á17,
Scheme 2) have also been described [14].
Here, we report: (1) a study of the [HRh(CO)L3]
complexes (Lꢀ
1, 2) by 1H-, 31P{1H}-, 31P{1H, 103Rh}-,
and Ineptnd 31PÁ103Rh{1H}-NMR spectroscopy. Hy-
/9, Scheme 1) [13], and more
/
/
/
dridorhodium complexes were formed in solution from
the substitution reaction with [HRh(CO)(PPh3)3] or by
reaction with [Rh(acac)(CO)2] and CO/H2; and (2) a
study of biphasic rhodium-catalysed hydroformylation
of 1-octene with the amphiphilic ligands 10Á17.
/
2. Results and discussion
signals to the structures AÁD is also consistent with
/
the changes observed in these spectra in the reaction
conducted with one or three equivalents of ligand. Thus,
the intensity of signals assigned to the structures A and
B is enhanced in the reaction with one equivalent of
ligand, and the signal assigned to D is not observed. On
the contrary, in the reaction with three equivalents of
ligand the signals assigned to structures C and D are
more intense and the signal attributed to A is weak. The
31P{1H}-NMR spectroscopy is also in agreement with
the formation of the mixture of hydrides shown in
Scheme 4, although the interpretation of the spectra is
2.1. Hydridorhodium(I) complexes with amphiphilic
phosphines
First, the formation of hydridorhodium complexes
with the amphiphilic ligands 1Á17 was studied by means
/
of the known substitution reaction with HRh(CO)-
(PPh3)3 (Scheme 3). Unfortunately, preliminary results
showed that NMR methods could not supply valuable
information about the reaction products with the
sulphonated ligands 10Á
bands were observed. The formation of broad bands
with the ligands 10Á17 has been previously observed
and can be assigned to the formation of supramolecular
/17 because only very broad
1
/
not so straightforward as with the H-NMR spectra.
Two groups of signals are observed in the 19Á24 and
/
Scheme 1.

174
A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
Scheme 2.
39Á45 ppm regions, which are respectively assigned to
/
the PPh2CH2 and PPh3 phosphorus atoms coordinated
to the metal. The 31P{1H}-NMR spectrum of the
reaction mixture obtained after reaction of three
equivalents of ligand with HRh(CO)(PPh3)3 is complex,
displaying multiple signals as a result of the presence of
different compounds with different phosphorus atoms
and to the coupling with the rhodium nucleus (Fig. 2).
The 31P{1H, 103Rh}-NMR spectrum was helpful to
assign the signals because the absence of coupling with
the rhodium nucleus leads to simplified spectra (Fig. 2).
Scheme 4.
PPh2CH2 groups (19Á
/
24 ppm). This change is a logical
consequence of the presence of more triphenylphosphine
coordinated to the metal after addition of less amphi-
philic phosphine.
Thus, in the 39Á45 ppm region of this spectrum a
/
singlet, a doublet, and a triplet were observed. These
signals can be respectively assigned to the phosphorus
atoms of the PPh3 in complexes A, B and C, which are
respectively coupled with zero, one or two phosphorus
atoms of the PPh2CH2 groups. The signals of the
For the purpose of achieving information about the
hydridorhodium complexes existing when ligands 1 and
2 are employed in the hydroformylation reaction, the
reaction between a solution of the complex [Rh(a-
cac)(CO)2] and ligands 1 and 2 with syngas (20 atm of
1:1 CO/H2) was studied by NMR and IR spectroscopies
(Scheme 5). The reaction was performed by pressurising
a THF solution of the precursor [Rh(acac)(CO)2] and
the corresponding ligand with the syngas to 20 atm. This
pressurisation was performed in an autoclave which was
heated to 50 8C. After reaction, the resulting solution
was evaporated to dryness under vacuum and the
products were analysed by NMR and IR spectroscopies.
In these conditions, the data obtained are in agreement
with the formation of the complex RhH(CO)L3 as a
main product in the reaction of 1 mol of [Rh(a-
cac)(CO)2] with 3 mol of ligands 1 and 2. The 1H-
NMR spectra at room temperature and at low tempera-
PPh2CH2 groups in the 19Á24 ppm region are not so
/
informative because some signals are superposed. How-
ever, the doublet of C is clearly shown, the triplet of B is
displayed as one peak and two shoulders, and finally the
signal of D is observed as a shoulder that is more visible
in the 31P{1H} spectrum as a doublet. The formation of
the four complexes AÁ
/
D is also supported by the
Ineptnd 31PÁ103Rh{1H}-NMR spectrum, showing the
/
four different rhodium nuclei. The most significant
difference (if we compare the 31P{1H}-NMR spectra
obtained after the addition of two equivalents of ligand
with the spectra after the addition of three equivalents of
ligand) is the intensity enhancement of signals in the
region of coordinated triphenylphosphine (39Á
/
45 ppm)
and the decrease of peaks in the region of coordinated
ture (ꢂ60 8C) display a quadruplet as a main signal,
/
Scheme 3.

A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
175
Fig. 1. The 1H-NMR spectrum (CD2Cl2, 233 K) after reaction of HRh(CO)(PPh3)3 with 3 mol of ligand 2. Top: 1H-NMR spectrum; bottom:
1H{31P} spectrum.
Fig. 2. The 31P-NMR spectrum (CD2Cl2, 233 K) after reaction of HRh(CO)(PPh3)3 with 3 mol of ligand 2. Top: 31P{1H}-NMR spectrum; bottom:
31P{1H, 103Rh} spectrum.

176
A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
Scheme 5.
with chemical shift and coupling constants nearly
identical to those assigned to the complexes RhH-
(CO)(1)3 and RhH(CO)(2)3 in the previously described
substitution reaction studies with RhH(CO)(PPh3). The
1H{31P}-NMR spectra corroborate the coupling of the
hydride with three phosphorus nuclei since these signals
are now displayed as singlets. The 31P{1H}-NMR
spectra at low temperature are also consistent with
these data, and they show a doublet as the main signal,
which is also identical to that observed in the substitu-
tion reaction with RhH(CO)(PPh3)3. These spectra also
atom. This group of experiments did not lead to NMR
spectra with important changes in regard to the previous
studies involving the ratio ligand/rhodiumꢀ
1H-NMR spectra in the hydride region exhibit the signal
of the quadruplet attributed to the RhH(CO)L3 (Lꢀ1,
/
3. Thus, the
/
2) complexes, and the 31P{1H}-NMR spectra display the
characteristic doublet of these complexes. Although
some minor peaks and broad bands were detected, no
significant signals were observed at low fields, as would
be expected for the bidentate (P, O) complexes [18].
In conclusion, the complexing studies performed with
ligands 1 and 2 show that the complexes formed with
these ligands are identical to those reported with other
non-functionalised phosphines like PPh3. Consequently,
there is no significant difference between a ligand with
unfavourable bidentate (P, O) coordination mode (1)
and a ligand with favourable bidentate (P, O) coordina-
tion mode (2). A similar coordination behaviour in the
reported amphiphilic ligands with different polyether
chain is an attractive idea because it implies that it could
be possible to modulate the physical properties (lipo-
philic character, solubility, etc.) of the complexes
involved in catalytic processes by means of changes in
the length of the polyether chain without affecting their
bonding mode.
display other weak signals as a peak at ꢂ
/
22/ꢂ23 ppm
/
assigned to the free ligand and a small doublet at 4.6/4.8
ppm, which can be attributed to the presence of small
amounts of dinuclear complexes [16]. The spectra at
room temperature do not show significant changes; the
doublet assigned to RhH(CO)L3 (Lꢀ1, 2) complexes is
/
also the main signal, and the peak of the free ligand is
not observed as a result of a rapid intermolecular
exchange processes at higher temperatures. The Ineptnd
31PÁ103Rh{1H}-NMR spectra at low temperature ex-
/
hibits a unique signal with the distinctive quadruplet
shape of a rhodium nucleus coupled with three phos-
phorus atoms for RhH(CO)L3 (Lꢀ1, 2) complexes. The
/
chemical shift and coupling constants of these signals
are also very similar to the values obtained in the
substitution reaction of Scheme 3. The IR spectra in the
2.2. Two-phase rhodium-catalysed hydroformylation of
n(CÄ/O) region were examined at the reaction times of 6,
1-octene with amphiphilic ligands 10Á17
/
12 and 24 h. These spectra were performed with samples
extracted from the autoclave without posterior evapora-
tion, and all of them are nearly identical, displaying two
bands at 1976 and 1943 cmꢂ1. The position of these
bands is consistent with the reported data for the
RhH(CO)2(PPh3)2 complex [17]. This result is in accor-
dance with the reported well-known equilibrium be-
tween the RhH(CO)2L2 and RhH(CO)L3 complexes
[16].
The reaction of syngas with a THF solution of
[Rh(acac)(CO)2] and ligands 1 and 2 (Scheme 5) was
also studied with 2 mol of ligand for each mole of metal.
This reaction can supply information about the possible
formation of bidentate (P, O) or upper mode of
coordination of the amphiphilic ligands to the metal
In a previous study, we reported the surfactant
properties of ligands 10Á17 and the metallo-surfactant
/
character of their palladium metal complexes [14].
There, we observed remarkable differences between the
surfactant properties of metallo-surfactants with middle
polyether chains (11, 14) regarding ligands with long
polyether chains (12, 15). Indeed, the metallo-surfac-
tants with middle polyether chains lead to micellar
systems where the metal concentration in the interface
is higher than in the compounds with longer polyether
chains. In order to investigate whether these different
surfactant properties can be related to the catalytic
properties of metal complexes with the amphiphilic
ligands 10Á17, studies of two-phase hydroformylation
/

A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
177
of 1-octene with rhodium complexes of 10Á
/
17 were
organic phase. Indeed, the two-phase system after
hydroformylation reaction showed an intense coloration
of the organic phase in the experiments performed with
ligands 10 and 13. Likewise, the sole ligands (16, 17),
which led to a two-phase system after hydroformylation
reaction with the organic phase absolutely colourless,
coincide with the experiments with the lower conversion
results. As mentioned earlier, the interest of biphasic
catalysis is to obtain a catalyst that is poorly soluble in
the organic phase but which can be easily recovered by
separation of the two phases. As a conclusion, the
undertaken. The catalyst solutions were prepared by
mixing a methanol solution of Rh(acac)(CO)2 with a
water solution of the ligand. The experiments were
conducted at the temperature of 80 8C and with a 20
atm pressure of syngas, with a ligand-to-rhodium ratio
of 10. Catalytic results are summarised in Tables 1 and 2
at substrate-to-metal ratios of 500 and 1000, respec-
tively. For the purposes of comparison, experiments
with trisulphonated triphenylphosphine (TPPTS) were
also performed in order to have available catalytic data
of a non-amphiphilic but water-soluble ligand in iden-
tical experimental conditions. In general, in the assays
with high conversion a relevant activity was observed
for alcohol formation [8b]. The yields of nonanols with
respect to the total products are displayed in the last
column. Since we used methanol as a co-solvent, acetals
formation is a possible side reaction between aldehydes
and methanol. This reaction was only significant in the
assays of Table 1 with higher conversion, which
corresponds to TPPTS (3.3% formation of acetals) and
13 (5.6% formation of acetals). Other side reactions such
as hydrogenation and isomerisation were minor, with
yields lower than 1% in most cases, and with the logical
exception of experiments with very low conversion (16
and 17), where the hydrogenation increases to 10% of
products.
The analysis of conversion data displayed in Tables 1
and 2 shows that all catalysts prepared with ligands with
middle (11, 14, 16) or long polyether chains (12, 15)
exhibit poorer conversion results than the assays with
TPPTS. Conversely, the conversion in experiments
performed with ligands with a hydrophobic group and
a short polyether chain (10, 13) is seen to be significantly
higher than with other ligands. In fact, ligand 13 shows
higher conversion than TPPTS in Table 1 and compar-
able data in Table 2. The special activity of ligands 10
and 13 can probably be attributed to the ability of these
ligands to increase the metal concentration in the
demonstrated surfactant properties of ligands 10Á17 do
/
not involve an improvement of the catalytic biphasic
hydroformylation reaction in the studied reaction con-
ditions because the observed activity seems to be related
to the metal transfer into the organic phase.
3. Experimental
3.1. General
All reactions were performed under nitrogen by
standard Schlenk tube techniques. Solvents were pur-
ified by standard procedures. The NMR spectra were
recorded by the Service de RMN du Laboratoire de
Chimie de Coordination on Bruker AC250 and AMX-
400 instruments. All chemical shift values are given in
ppm and are referenced with respect to residual protons
1
in the solvents for H spectra, to solvent signals for 13C
spectra and to external phosphoric acid for 31P spectra.
The hydroformylation reactions were performed in a
125 ml stainless steel autoclave of Autoclave Engineers
equipped with mechanical stirring and temperature
controller. The solution was contained in a glass inlet.
The pressure of the autoclave was kept constant by
means of a regulator, connected to a syngas reservoir.
The evolution of the reaction was monitored by the drop
of pressure in the reservoir. Conversions and regioselec-
Table 1
Two-phase hydroformylation of 1-octene with rhodium catalyst (1-octene/Rhꢀ500)
/
a
Conversion (%)
b
alcohols (%)
c
Regioselectivity (%)
Ligand
Aldehydesꢁ
/
Nonanol (%)
TPPTS
10
11
91
50
25
31
97
28
25
5
97
100
96
80
78
76
72
73
75
70
70
72
0.5
14
6.4
4.3
12
13
95
100
100
100
80
12.7
10.7
2.9
14
15
16
17
Á/
Á/
6
83
Reaction conditions: 25 mmol of 1-octene, 0.05 mmol of [Rh(acac)(CO)2], and 0.50 mmol of ligand; temperature: 80 8C; and reaction time: 24 h.
Converted olefin as percentage of the initial amount.
a
b
Percentage of aldehydes plus alcohols relative to the converted olefin.
Percentage of linear aldehydes plus alcohols.
c

178
A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á181
/
Table 2
Two-phase hydroformylation of 1-octene with rhodium catalyst (1-octene/Rhꢀ1000)
/
a
Conversion (%)
b
alcohols (%)
c
Regioselectivity (%)
Ligand
Aldehydesꢁ
/
Nonanol (%)
TPPTS
10
11
68
42
4
90
98
75
92
98
85
95
80
76
77
70
68
71
71
69
71
Á
Á/
Á/
Á/
/
12
13
12
60
4
3.2
14
15
10
5
17
Á/
Reaction conditions: 25 mmol of 1-octene, 0.025 mmol of [Rh(acac)(CO)2], and 0.25 mmol of ligand; temperature: 80 8C; and reaction time: 24 h.
Converted olefin as percentage of the initial amount.
a
b
Percentage of aldehydes plus alcohols relative to the converted olefin.
Percentage of linear aldehydes plus alcohols.
c
tivity were determined by gas chromatography with a
Hewlett-Packard G1800A, equipped with a capillary
1.28 (s, (CH3)2), 1.65 (s, CH2, 2,2,4,4-tetramethyl-
butyl), 2.7 (b, CH2ÃP), 4.2 (b, CH2O).
31P{1H}-NMR (Tꢀ
298 K; C6D6): ꢂ
/
HP5 column (30 mꢃ/0.25 mm). The reaction products
/
/
4.9 (b,
107.5
1
2
were identified by comparison with authentic samples
and mass spectroscopy. Compounds [RhH(CO)(PPh3)3]
[19] and [Rh(acac)(CO)2] [20] were prepared by pub-
lished procedures.
PPh3), 21.8 (dd, JPRh
Hz, RhH(CO)(PPh3)(PPh2R)2), 22.3 (d, JPRh
ꢀ
/
153.8 Hz, JPP
ꢀ
/
1
ꢀ
/
1
151.4 Hz, RhH(CO)(PPh2R)3), 40.9 (d, JPRh
ꢀ
/
1
153.8 Hz, RhH(CO)(PPh3)3), 41.1 (dd, JPRh
2
151.4 Hz, JPP
ꢀ
/
ꢀ117.2 Hz, RhH(CO)(PPh3)2L). In
/
the 19Á24 ppm region are observed some weak
/
3.2. Complexation studies with [RhH(CO)(PPh3)3]
signals assigned to the coordinated PPh2R groups of
complex RhH(CO)(PPh3)2L. Likewise, some weak
peaks assigned to the coordinated PPh3 of
The following procedure was used in all experiments
a
with ligands
1
and 2. To
solution of
RhH(CO)(PPh3)L2 are observed in the 39Á
region.
c) Reaction with three equivalents of ligand (L/Rhꢀ
3): 1H-NMR (Tꢀ
298 K; C6D6; except phenyl
/45 ppm
[RhH(CO)(PPh3)3] (0.100 g, 0.11 mmol) in THF (30
ml), 0.11, 0.22 or 0.33 mmol of the respective phos-
phines was slowly added with continuous stirring. The
resulting solutions were stirred for 12 h and evaporated
to dryness in vacuum leading to red oils, which were
/
/
resonances): ꢂ9.97 (partially hidden weak signal
/
with quadruplet shape, 2JHP
:15 Hz, RhH(CO)L3),
/
1
used to measure the 31P{1H}- and H-NMR spectra.
2
9.78 (q, JHP
ꢂ
/
ꢀ15.0 Hz, RhH(CO)(PPh3)L2),
/
ꢂ
/
9.53 (broad signal with quadruplet shape,
2JHP
15 Hz, RhH(CO)(PPh3)2L), 0.75 (s,
(CH3)3), 1.28 (s, (CH3)2), 1.65 (s, CH2, 2,2,4,4-
tetramethylbutyl), 2.7 (b, CH2ÃP), 4.2 (b, CH2O).
31P{1H}-NMR (Tꢀ
298 K; C6D6): ꢂ17.0 (b,
PPh2R), ꢂ4.9 (s, PPh3), 21.8 (dd, JPRh 153.8
107.5 Hz, RhH(CO)(PPh3)(PPh2R)2),
3.2.1. Studies with ligand 1
:
/
a) Reaction with one equivalent of ligand (L/Rhꢀ
1H-NMR (Tꢀ
298 K; C6D6; except phenyl reso-
nances): ꢂ9.78 (b, RhH(CO)(PPh3)L2), ꢂ9.53 (br
15 Hz, RhH(CO)(PPh3)2L), ꢂ9.27 (b,
/
1):
/
/
/
/
/
/
1
2
/
ꢀ
/
q, JHP
:
/
/
2
Hz, JPP
ꢀ
/
RhH(CO)(PPh3)3), 0.75 (s, (CH3)3), 1.28 (s,
(CH3)2), 1.65 (s, CH2, 2,2,4,4-tetramethylbutyl),
1
22.3 (d, JPRh
39Á45 (weak signals assigned to the coordinated
PPh3).
ꢀ151.4 Hz, RhH(CO)(PPh2R)3),
/
/
2.67 (b, CH2Ã
31P{1H}-NMR (Tꢀ
PPh3), 19Á23 (weak signals of the coordinated
PPh2 groups of ligand 1), 40.9 (d, JPRh
/P), 4.11 (b, CH2O).
/
298 K; C6D6): ꢂ5.0 (b,
/
/
1
ꢀ
/
153.8
3.2.2. Studies with ligand 2
1
Hz, RhH(CO)(PPh3)3), 41.1 (dd, JPRh
2JPP
117.2 Hz, RhH(CO)(PPh3)2L).
b) Reaction with two equivalents of ligand (L/Rhꢀ
1H-NMR (Tꢀ
298 K; C6D6; except phenyl reso-
nances): ꢂ9.97 (partially hidden weak signal with
quadruplet shape, JHP
ꢀ
/
151.4 Hz,
ꢀ
/
a) Reaction with two equivalents of ligand (L/Rhꢀ
1H-NMR (Tꢀ
298 K; C6D6; except phenyl reso-
nances): ꢂ10.09 (partially hidden weak signal with
quadruplet shape, JHP
/2):
/2):
/
/
/
2
/
:
/
15 Hz, RhH(CO)L3),
15.0 Hz, RhH(CO)(PPh3)L2),
15 Hz, RhH(CO)(PPh3)2L),
9.32 (b, RhH(CO)(PPh3)3), 0.75 (s, (CH3)3), 1.28
(s, (CH3)2), 1.65 (s, CH2, 2,2,4,4-tetramethylbutyl),
2
2
:
/
15 Hz, RhH(CO)L3),
ꢂ
/
9.86 (q, JHP
ꢀ
/
2
2
ꢂ
/
9.78 (q, JHP
2
9.53 (q, JHP
ꢀ
/
15.0 Hz, RhH(CO)(PPh3)L2),
15.0 Hz, RhH(CO)(PPh3)2L),
ꢂ
/
9.58 (br q, JHP
:
/
ꢂ
/
ꢀ
/
ꢂ
/
ꢂ9.27 (b, RhH(CO)(PPh3)3), 0.75 (s, (CH3)3),
/

A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
179
2
2.6 (b, CH2Ã
/
P), 3.1Á
/
4.0 (m, CH2O).
(t(sh), JPP
23.03 (d, 2JPP
(PPh2R)2), 23.38 (s(sh), RhH(CO)(PPh2R)3),
:
/
125 Hz, RhH(CO)(PPh3)2(PPh2R)),
31P{1H}-NMR (Tꢀ
PPh3), 21.5 (dd, JPRh
/
298 K; C6D6): ꢂ
/
4.6 (b,
105.0
ꢀ
/
108.5 Hz, RhH(CO)(PPh3)-
1
2
151.4 Hz, JPP
ꢀ
/
ꢀ
/
1
Hz, RhH(CO)(PPh3)(PPh2R)2), 40.9 (d, JPRh
2
ꢀ
/
43.72 (s, RhH(CO)(PPh3)3), 43.98 (d, JPP
Hz, RhH(CO)(PPh3)2(PPh2R)), 44.11 (t, JPP
109.8 Hz, RhH(CO)(PPh3)(PPh2R)2).
ꢀ115.5
/
1
2
153.9, RhH(CO)(PPh3)3), 41.1 (dd, JPRh
ꢀ
/
151.7
117.2 Hz, RhH(CO)(PPh3)2(PPh2R)).
Some weak signals and shoulders were observed
between 19Á24 ppm and 39Á45 ppm which were
ꢀ
/
2
Hz, JPP
ꢀ
/
Ineptnd 31PÁ103
CD2Cl2): ꢂ
(CO)(PPh2R)3),
RhH(CO)(PPh3)(PPh2R)2),
150.1 Hz, RhH(CO)(PPh3)2(PPh2R)), ꢂ
1JPRh
153.3 Hz, RhH(CO)(PPh3)3).
/
Rh{1H}-NMR (Tꢀ
/
233 K;
1
/
/
/
885.9 (q, JPRh
ꢀ
/
150.1 Hz, RhH-
1
respectively assigned to RhH(CO)(PPh3)2(PPh2R)
and RhH(CO)(PPh3)(PPh2R)2.
ꢂ
/
864.7 (q, JPRh
842.0 (q, JPRh
820.2 (q,
ꢀ
/
150.6 Hz,
1
ꢂ
/
ꢀ
/
b) Reaction with three equivalents of ligand (L/Rhꢀ
2): 1H-NMR (Tꢀ
298 K; C6D6; except phenyl
resonances): ꢂ10.09 (weak signal with quadruplet
shape partially hidden, JHP
(CO)L3), 9.87 (q,
RhH(CO)(PPh3)L2), ꢂ9.58 (br q, JHP
RhH(CO)(PPh3)2L), ꢂ
0.75 (s, (CH3)3), 1.28 (s, (CH3)2), 1.65 (s, CH2,
2,2,4,4-tetramethylbutyl), 2.6 (b, CH2ÃP), 3.1Á4.0
/
/
/
ꢀ
/
/
2
ꢀ15.0 Hz, RhH-
/
3.3. Complexation studies with [Rh(acac)(CO)2]
ꢂ
/
2JHP
ꢀ
/
14.7
Hz,
15 Hz,
2
/
:
/
The following procedure was used in all experiments
with ligands 1 and 2 with the L/Rh ratios of 2 and 3. To
a solution of [Rh(acac)(CO)2] (0.100 g, 0.39 mmol) in
THF (20 ml) was added 0.78 or 1.17 mmol of the
respective phosphine with continuous stirring. The
resulting solution was transferred under nitrogen to an
/
9.32 (b, RhH(CO)(PPh3)3),
/
/
(m, CH2O).
31P{1H}-NMR (Tꢀ
PPh3), 21.5 (dd, JPRh
/
298 K; C6D6): ꢂ
/
4.9 (b,
105.0
1
2
151.4 Hz, JPP
ꢀ
/
ꢀ
/
autoclave, pressurised to 20 atm with syngas (COꢁH2),
/
1
Hz, RhH(CO)(PPh3)(PPh2R)2), 40.9 (d, JPRh
ꢀ
/
heated to 50 8C and allowed under stirring for 24 h.
Next, the system was cooled, depressurised and the
reaction mixture was transferred under nitrogen to an
Schlenk tube. The resulting solutions were evaporated to
dryness in vacuo yielding dark oils, which were used to
measure the 31P{1H}- and H-NMR spectra. The IR
spectra in the n(CÄ
1
153.9, RhH(CO)(PPh3)3), 41.1 (dd, JPRh
ꢀ
/
151.7
117.2 Hz, RhH(CO)(PPh3)2(PPh2R)).
Some weak signals and shoulders were observed
between 19Á24 ppm and 39Á45 ppm which were
2
Hz, JPP
ꢀ
/
/
/
1
respectively assigned to RhH(CO)(PPh3)2(PPh2R)
and RhH(CO)(PPh3)(PPh2R)2.
/O) region were measured at the
c) Reaction with three equivalents of ligand (L/Rhꢀ
2) at 233 K (these spectra were recorded in an
AMX-400 instrument). 1H-NMR (Tꢀ
233 K;
CD2Cl2; except phenyl resonances): ꢂ10.28 (q,
/
reaction times of 6, 12 and 24 h for all experiments. All
recorded spectra display two peaks at 1976 and 1943
cmꢂ1 (RhH(CO)2L2).
/
/
2JHP
15.2 Hz, RhH(CO)(PPh3)L2), ꢂ
Hz, RhH(CO)(PPh3)2L), ꢂ
ꢀ
/
15.6 Hz, RhH(CO)L3), ꢂ
/
10.09 (q, JHP
ꢀ
/
2
3.3.1. Significant NMR data
/
9.85 (q, 2JHP
9.67 (q, JHP
ꢀ14.6
/
2
/
ꢀ13.2 Hz,
/
3.3.1.1. Studies with ligand 1.
RhH(CO)(PPh3)3), 0.75 (s, (CH3)3), 1.37 (s,
(CH3)2), 1.75 (s, CH2, 2,2,4,4-tetramethylbutyl),
a) Reaction with two equivalents of ligand (L/Rhꢀ
1H-NMR (Tꢀ
298 K; C6D6; except phenyl reso-
nances): ꢂ 15.6, RhH(CO)L3), 0.75
10.02 (q, 2JHP
(s, (CH3)3), 1.28 (s, (CH3)2), 1.64 (s, CH2, 2,2,4,4-
tetramethylbutyl), 2.7 (b, CH2ÃP), 4.1 (b, CH2O).
31P{1H}-NMR (Tꢀ
298 K; C6D6): 22.3 (d,
1JPRh
151.4 Hz, RhH(CO)(PPh2R)3). Two unas-
/
2):
2.7 (b, CH2Ã
1H{31P}-NMR (Tꢀ
yl resonances): ꢂ10.28 (s, RhH(CO)L3), ꢂ
RhH(CO)(PPh3)L2), ꢂ9.85 (s, RhH(CO)(PPh3)2L),
9.67 (s, RhH(CO)(PPh3)3), 0.75 (s, (CH3)3), 1.37
(s, (CH3)2), 1.75 (s, CH2, 2,2,4,4-tetramethylbutyl),
2.7 (b, CH2ÃP), 3.9 (b, CH2O).
31P{1H}-NMR (Tꢀ
233 K; CD2Cl2): 22.98 (dd,
/
P), 3.9 (b, CH2O).
233 K; CD2Cl2; except phen-
10.09 (s,
/
/
/
ꢀ
/
/
/
/
/
ꢂ
/
/
ꢀ
/
/
signed broad signals were observed at 18.0 and 28.7
ppm.
/
2
1JPRh
ꢀ
/
152.2 Hz, JPP
(PPh3)(PPh2R)2), 23.38 (d, JPRh
ꢀ
/
108.5 Hz, RhH(CO)-
b) Reaction with three equivalents of ligand (L/Rhꢀ
3): 1H-NMR (Tꢀ
298 K; C6D6; except phenyl
resonances): ꢂ10.04 (q, JHP
(CO)L3), 0.75 (s, (CH3)3), 1.28 (s, (CH3)2), 1.64 (s,
/
1
ꢀ
/
152.6 Hz,
153.8 Hz,
152.6 Hz,
/
1
RhH(CO)(PPh2R)3), 43.81 (d, JPRh
1
RhH(CO)(PPh3)3), 44.05 (dd, JPRh
2
ꢀ
/
/
ꢀ15.0 Hz, RhH-
/
ꢀ
/
2JPP
(dt, JPRh
ꢀ
/
115.5 Hz, RhH(CO)(PPh3)2(PPh2R)), 44.16
2
150.3 Hz, JPP
CH2, 2,2,4,4-tetramethylbutyl), 2.7 (b, CH2Ã
(b, CH2O).
31P{1H}-NMR (Tꢀ
1JPRh
/
P), 4.1
1
ꢀ
/
ꢀ109.8 Hz, RhH(CO)-
/
(PPh3)(PPh2R)2). Some weak signals and shoulders
were observed between 21 and 24 ppm, which were
assigned to RhH(CO)(PPh3)2(PPh2R).
/
298 K; C6D6): 22.3 (d,
ꢀ
/
151.4 Hz, RhH(CO)(PPh2R)3). A very
weak unassigned doublets were observed at 4.8
ppm (1JPRh
156.3 Hz).
31P{1H, 103Rh}-NMR (Tꢀ
/233 K; CD2Cl2): 22.8
ꢀ
/

180
A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á181
/
c) Reaction with three equivalents of ligand (L/Rhꢀ
3) at 213 K (these spectra were recorded in an
AMX-400 instrument): 1H-NMR (Tꢀ
213 K;
CD2Cl2; except phenyl resonances): ꢂ10.19 (q,
2JHP
15.6 Hz, RhH(CO)L3), 0.63 (s, (CH3)3),
1.27 (s, (CH3)2), 1.67 (s, CH2, 2,2,4,4-tetramethyl-
butyl), 2.53 (b, CH2ÃPÃRh), 2.62 (b, CH2ÃP), 3.7
(b, OCH2CH2PÃRh), 4.1 (b, OCH2CH2PÃRh). An
unassigned broad band is observed at ꢂ12.6 ppm.
1H{31P}-NMR (Tꢀ
213 K; CD2Cl2; except phen-
yl resonances): ꢂ10.19 (s, RhH(CO)L3), 0.63 (s,
(CH3)3), 1.27 (s, (CH3)2), 1.67 (s, CH2, 2,2,4,4-
tetramethylbutyl), 2.53 (b, CH2ÃPÃRh), 2.62 (b,
CH2ÃP), 3.7 (b, OCH2CH2PÃRh), 4.1 (b,
OCH2CH2PÃRh). An unassigned broad band is
observed at ꢂ12.6 ppm.
31P{1H}-NMR (Tꢀ
213 K; CD2Cl2): ꢂ
151.4 Hz, RhH(CO)-
/
(CH3)3), 1.29 (s, (CH3)2), 1.67 (s, CH2, 2,2,4,4-
tetramethylbutyl), 2.6 (b, CH2ÃP), 3.1Á4.0 (m,
CH2O). A weak unassigned broad signal was
observed at ꢂ12.5 ppm.
31P{1H}-NMR (Tꢀ
213 K; CD2Cl2): ꢂ
/
/
/
/
/
ꢀ
/
/
/
22.0 (b,
151.1 Hz, RhH(CO)-
(PPh2R)3). A weak unassigned doublet was ob-
1
PPh2R), 23.1 (d, JPRh
ꢀ
/
/
/
/
1
/
/
served at 4.9 ppm (d, JPRh
ꢀ
/
158.0 Hz).
/
Ineptnd 31PÁ103Rh{1H}-NMR (Tꢀ
/
/
213 K;
151.3 Hz, RhH(CO)-
1
886.3 (q, JPRh
/
CD2Cl2): ꢂ
/
ꢀ
/
/
(PPh2R)3).
/
/
3.4. Catalytic hydroformylation of 1-octene: general
procedure
/
/
/
/
A solution of [Rh(acac)(CO)2] in methanol (5 ml) was
prepared with slight heating in order to achieve solid
dissolution. This solution was added to a solution of the
desired amount of the phosphine (0.25/0.50 mmol) in
water (5 ml). The resulting solution was stirred for 10
min and a yellowish-orange solution was obtained.
Next, 1-octene was added (4 ml) and the resulting
mixture was transferred to the evacuated autoclave,
which was pressurised up to 15 atm and heated to 80 8C
under stirring at 50 rpm. When the thermal equilibrium
was reached, the pressure was adjusted to 20 atm and
the stirring to 400 rpm. This moment was defined as
reaction time equal to zero.
/
/
23.0 (s,
1
PPh2R), 23.7 (d, JPRh
(PPh2R)3).
ꢀ
/
Ineptnd 31PÁ103
/
Rh{1H}-NMR (Tꢀ
/
213 K;
150.9 Hz, RhH(CO)-
1
899.8 (q, JPRh
CD2Cl2): ꢂ
/
ꢀ
/
(PPh2R)3).
3.3.1.2. Studies with ligand 2.
a) Reaction with two equivalents of ligand (L/Rhꢀ
1H-NMR (Tꢀ
298 K; C6D6; except phenyl reso-
nances): ꢂ10.12 (q, JHP
0.75 (s, (CH3)3), 1.28 (s, (CH3)2), 1.65 (s, CH2,
2,2,4,4-tetramethylbutyl), 2.6 (b, CH2ÃP), 3.1Á4.0
(m, CH2O).
31P{1H}-NMR (Tꢀ
1JPRh
151.4 Hz, RhH(CO)(PPh2R)3). A weak
unassigned doublet is observed at 4.6 ppm
(1JPRh
156.3 Hz).
b) Reaction with three equivalents of ligand (L/Rhꢀ
3): 1H-NMR (Tꢀ
298 K; C6D6; except phenyl
resonances): 10.12 (q, 15.0 Hz,
2JHP
/2):
/
2
/
ꢀ15.0 Hz, RhH(CO)L3),
/
/
/
Acknowledgements
/298 K; C6D6): 21.9 (d,
ꢀ
/
The authors thank DGICYT (Programa de Promo-
cio´n General del Conocimiento) for financial support.
ꢀ
/
/
/
ꢂ
/
ꢀ
/
References
RhH(CO)L3), 0.75 (s, (CH3)3), 1.28 (s, (CH3)2),
1.65 (s, CH2, 2,2,4,4-tetramethylbutyl), 2.6 (b,
[1] (a) F. Joo, A. Agnes, J. Mol. Catal. A 116 (1997) 3;
(b) B. Cornils, W.A. Herrmann (Eds.), Aqueous-phase Organo-
metallic Catalysis, Wiley/VCH, Weinheim, 1998.
[2] B. Cornils, E. Kuntz, in: B. Cornils, W.A. Herrmann (Eds.),
Aqueous-phase Organometallic Catalysis (Chapter 6.1.1), Wiley/
VCH, Weinheim, 1998.
CH2Ã
31P{1H}-NMR (Tꢀ
1JPRh
151.4 Hz, RhH(CO)(PPh2R)3). A weak
unassigned doublet is observed at 4.6 ppm
(1JPRh
156.3 Hz).
c) Reaction with three equivalents of ligand (L/Rhꢀ
3) at 213 K (these spectra were recorded in an
AMX-400 instrument): 1H-NMR (Tꢀ
213 K;
CD2Cl2; except phenyl resonances): ꢂ10.40 (q,
2JHP
15.5 Hz, RhH(CO)L3), 0.63 (s, (CH3)3),
1.29 (s, (CH3)2), 1.67 (s, CH2, 2,2,4,4-tetramethyl-
P), 3.1Á4.0 (m, CH2O). A weak
unassigned broad signal was observed at ꢂ
ppm.
1H{31P}-NMR (Tꢀ
yl resonances): ꢂ
/
P), 3.1Á/4.0 (m, CH2O).
/298 K; C6D6): 21.9 (d,
ꢀ
/
ꢀ
/
[3] B. Cornils, Angew. Chem. 34 (1995) 1575.
[4] P. Purwanto, H. Delmas, Catal. Today 24 (1995) 135.
[5] R.V. Chaudhari, B.M. Bhanage, R.M. Deshpande, H. Delmas,
Nature (London) 373 (1995) 501.
/
/
[6] (a) B. Fell, C. Schobben, G. Papadogianakis, J. Mol. Catal. A 101
(1995) 179;
/
ꢀ
/
(b) H. Chen, Y.Z. Li, J.R. Chen, P.M. Cheng, Y.E. He, H.J. Li, J.
Mol. Catal. A 149 (1999) 1381;
(c) M. Haumann, H. Koch, P. Hugo, R. Schomacker, Appl.
Catal. A 225 (2002) 239;
butyl), 2.6 (b, CH2Ã
/
/
/
12.5
(d) Y. Zhang, Z. Mao, J. Chen, Catal. Today 74 (2002) 23;
(e) H. Chen, Y. Li, J. Chen, P. Cheng, X. Li, Catal. Today 74
(2002) 131;
/213 K; CD2Cl2; except phen-
/10.40 (s, RhH(CO)L3), 0.63 (s,
(f) C. Yang, X. Bi, Z. Mao, J. Mol. Catal. A 187 (2002) 35;

A. Solsona et al. / Journal of Organometallic Chemistry 669 (2003) 172Á
/181
181
(g) M. Haumann, H. Yildiz, H. Koch, R. Schomacker, Appl.
Catal. A 236 (2002) 173
[12] (a) J. Jiang, Y. Wang, C. Liu, F. Han, Z. Jin, J. Mol. Catal. A 147
(1999) 131;
[7] (a) G. Papadoginakis, R.A. Sheldon, in: B. Cornils, W.A.
Herrmann (Eds.), Aqueous-phase Organometallic Catalysis
(Chapter 3.2.4), Wiley/VCH, Weinheim, 1998;
(b) B. Fell, Tenside. Surfact. Deterg. 35 (1998) 326;
(c) B. Hanson, H. Ding, C.W. Kohlpaintner, Catal. Today 42
(1998) 421.
(b) J. Jiang, Y. Wang, C. Liu, Q. Xiao, Z. Jin, J. Mol. Catal. A
171 (2001) 85.
[13] E. Valls, J. Suades, R. Mathieu, Organometallics 18 (1999)
5475.
[14] E. Valls, A. Solsona, J. Suades, R. Mathieu, F. Comelles, C.
Lo´pez-Iglesias, Organometallics 21 (2002) 2473.
[15] E. Valls, J. Suades, R. Mathieu, J.F. Piniella, A. Alvarez-Larena,
J. Organomet. Chem. 626 (2001) 139.
[8] (a) B. Fell, G. Papadogianakis, J. Mol. Catal. 66 (1991) 143;
(b) H. Ding, B.E. Hanson, T. Bartik, B. Bartik, Organometallics
13 (1994) 3761;
[16] (a) J.M. Brown, A.G. Kent, J. Chem. Soc. Perkin Trans. II (1987)
1597;
(c) H. Ding, B.E. Hanson, J. Bakos, Angew. Chem. Int. Ed. Engl.
34 (1995) 1645;
(b) Z. Freixa, M.M. Pereira, A.A.C.C. Pais, J.C. Bayo´n, J. Chem.
Soc. Dalton Trans. (1999) 3245;
(d) B.E. Hanson, H. Ding, C.W. Kohlpaintner, Catal. Today 42
(1998) 421;
(c) C. Bianchini, H.M. Lee, A. Meli, F. Vizza, Organometallics 19
(2000) 849.
(e) M.S. Goedheijt, B.E. Hanson, J.N.H. Reek, P.C.J. Kamer,
P.W.N.M. van Leeuwen, J. Am. Chem. Soc. 122 (2000) 1650.
´
[9] (a) H. Gulya´s, P. Arva, J. Bakos, Chem. Commun. (1997) 2385;
[17] D. Evans, G. Yagupsky, G. Wilkinson, J. Chem. Soc. A (1968)
2660.
(b) H. Gulya´s, A. Dobo, J. Bakos, Can. J. Chem. 79 (2001) 1040.
[10] (a) T.L. Schull, L.R. Olano, D.A. Knight, Tetrahedron 56 (2000)
7093;
[18] (a) E. Lindner, H. Norz, Chem. Ber. 123 (1990) 459;
(b) M. Alvarez, N. Lugan, R. Mathieu, J. Organomet. Chem. 468
(1994) 249.
(b) S. Bischoff, M. Kant, Catal. Today 66 (2001) 183.
[11] (a) D.J. Brauer, J. Fischer, S. Kucken, K.P. Langhans, O. Stelzer,
N.Z. Weferling, Naturforsch 49b (1994) 1511;
[19] G.W. Parshall, Inorg. Synth. 15 (1974) 59.
[20] A. Solsona, Ph.D. Thesis, Universitat Auto`noma de Barcelona,
2001.
(b) F. Bitterer, S. Kucken, O. Stelzer, Chem. Ber. 128 (1995) 275.