Highly active hydroformylation catalysts: Synthesis, characterisation and catalytic performance

Article abstract of DOI:10.1007/s10562-013-1010-x

The phoszone ligand [(Ph₂P)(bis-3,5-CF₃-Ph)]NN= CH(penta-fluoro-Ph) transformed in liquid CO₂ at room temperature in presence of [Rh(cod)₂]OTf into [Rh(cod)(η²-P,P′- Ph₂POPPh₂)]OTf. Replacing the O-atom in Ph ₂POPPh₂ by a PrN-group leads to the ligand PrN(PPh ₂)₂ acting similarly as a bidentate ligand in [Rh(cod)(η²-P,P′-PrN(PPh₂)₂)]OTf. Hydroformylation of 1-octene with in situ catalysts formed by the ligands with [Rh(cod)₂]OTf showed hydroformylation activities at 50 % conversion of 16,000 h^-1 (PrN(PPh₂)₂/[Rh(cod) ₂]OTf) and 24,000 h^-1 (phoszone/[Rh(cod)₂]OTf), respectively.

Full text of DOI:10.1007/s10562-013-1010-x

Catal Lett (2013) 143:673–680
DOI 10.1007/s10562-013-1010-x
Highly Active Hydroformylation Catalysts: Synthesis,
Characterisation and Catalytic Performance
•
Elisabetta Piras Bernhard Powietzka Frederik Wurst
•
•
•
Doreen Neumann-Walter Hans-Jorg Grutzmacher
•
¨
Thomas Otto Thomas Zevaco Olaf Walter
¨
•
•
Received: 7 March 2013 / Accepted: 14 April 2013 / Published online: 15 May 2013
Ó Springer Science+Business Media New York 2013
Abstract The phoszone ligand [(Ph2P)(bis-3,5-CF3-Ph)]
NN=CH(penta-fluoro-Ph) transformed in liquid CO2 at room
temperature in presence of [Rh(cod)2]OTf into [Rh(cod)(g2-
P,P0-Ph2POPPh2)]OTf. Replacing the O-atom in Ph2POPPh2
by a PrN-group leads to the ligand PrN(PPh2)2 acting simi-
larly as a bidentate ligand in [Rh(cod)(g2-P,P0-PrN(PPh2)2)]
OTf. Hydroformylation of 1-octene with in situ catalysts
formed by the ligands with [Rh(cod)2]OTf showed hydro-
formylation activities at 50 % conversion of 16,000 h-1
(PrN(PPh2)2/[Rh(cod)2]OTf) and 24,000 h-1 (phoszone/
[Rh(cod)2]OTf), respectively.
hydroformylation of 1-octene [11]. Already the introduc-
tion of some fluorinated positions in the ligands of metal
complexes could increase the solubility of the pre-catalyst
in supercritical carbon dioxide [12]. Therefore it seemed
logical to studying the activity of the Rh-complexes of the
phoszone ligand 1 in supercritical carbon dioxide offering
the possibility by the use of this solvent to enhance the
reactivity further [13, 14]. This could have been well the
case as for highly active catalysts in the more phase
hydroformylation mass transfer could become the rate
limiting step which could be overcome by the appli-
cation of supercritical carbon dioxide as the solvent cre-
ating a single phase system [13, 14]. In this publication we
describe how experiments with the phoszone ligand 1 lead
to the development of highly active Rh-catalysts for the
hydroformylation of 1-octene.
Keywords Homogeneous catalysis ꢀ Activity ꢀ
Hydroformylation ꢀ Isomerisation
1 Introduction
Hydroformylation is one of the best studied catalytic
reactions [1, 2]. However, only few highly active catalysts
are known in the literature [3–10]. The phoszone ligand 1
bearing fluorinated side chains (Fig. 1) forms in the pres-
ence of [Rh(cod)2]OTf or [Rh(cod)2]BF4 an in situ
Rh-catalyst with an exceptionally high activity in the
2 Results and Discussion
Preliminary hydroformylation experiments with an in situ
catalyst system containing the ligand 1 and [Rh(cod)2]OTf as
the pre-catalyst showed a very high catalytic activity with
tofs of more than 20,000 h-1 [11]. Our attempts to study the
complex chemistry of the ligand 1 towards Rh-complex
fragments present in [Rh(cod)2]OTf or [Rh(cod)2]BF4 under
reaction conditions turned out to be difficult: A complex
mixture of products was obtained proven by 31P NMR. In
order to reduce the number of phases in the system we
wanted to investigate the reaction in sc CO2 by HP NMR.
Therefore NMR studies in dense carbon dioxide were pre-
pared: two equivalents (0.02 mmol) of the ligand 1, 1 eq. of
[Rh(cod)2]OTf and liquid carbon dioxide (1 ml) were
introduced in a HP NMR tube for further investigations the
next day. Over night single crystals precipitated in the NMR
Dedicated to the memory of Prof. Ivano Bertini.
E. Piras ꢀ B. Powietzka ꢀ F. Wurst ꢀ D. Neumann-Walter ꢀ
T. Otto ꢀ T. Zevaco ꢀ O. Walter (&)
IKFT, Campus Nord, KIT, Hermann-von-Helmholtz-Platz 1,
76344 Eggenstein-Leopoldshafen, Germany
e-mail: olaf.walter@kit.edu
E. Piras ꢀ H.-J. Gru¨tzmacher
Laboratorium fu¨r Anorganische Chemie, Wolfgang-Pauli-Street.
¨
10, ETH Honggerberg, HCI H 131, 8093 Zurich, Switzerland
123

674
E. Piras et al.
tube of other shape than the pure [Rh(cod)2]OTf. The NMR
tube was depressurised and the crystals were isolated and a
single crystal xray diffraction analysis was performed.
The result from the xray analyses showed the constitu-
tion of the complex to be [Rh(cod)(g2-P,P0-Ph2POP-
Ph2)](OTf) (2*OTf, Fig. 2) without any direct coordination
of the ligand 1 to the Rh(cod)-complex fragment but with a
g2-P,P0-coordination of the POP-ligand Ph2POPPh2. Con-
cerning its structural characteristics the geometry of the
Rh-center in the complex cation 2 can be described as
distorted square planar taking the center of the coordinated
CC bonds as their center of gravity. The Rh–P–O–P four
ring is planar (mean r = 0.46 pm from the plane) with a
small P-Rh-P angle of 68.7° at the Rh atom resulting from
the four ring formation. This structural feature is typical for
complexes where a transition metal is embedded in a
4-membered ring with two phosphino-groups coordinating
to the metal bridged over an oxygen atom [15–20].
C13
C19
C32
C25
P2
C26
C27
C31
O1
Rh1
C30
P1
C28
C29
C7
C1
It seems reasonable to assume that 2*OTf and therefore
the POP ligand Ph2POPPh2 in the cation 2 is formed by
partial hydrolysis of the ligand 1 in the coordination sphere
of the metal according to the literature [15–20]; alterna-
tively carbon dioxide could have functioned as the oxygen
source [21]. In the small amount of carbon dioxide present
in this HP NMR experiment the ligand 1 is already trans-
formed in that way that crystallisation of 2*OTf proceeds.
So it should be obvious that under real catalytic conditions
in presence of higher amounts of carbon dioxide and other
chemicals the ligand 1 and Rh-precursor would be trans-
formed even easier. If furthermore the formation of
Ph2POPPh2 proceeds already at RT then it should proceed
at elevated temperature even faster. Therefore it might be
possible that the catalytically active species in the hydro-
formylation of 1-octene involving the phoszone 1 as the
stabilising ligand is in all cases formed by Rh-complexes
stabilised by coordination to a Ph2POPPh2-ligand.
Fig. 2 View to the molecular structure of one of the two independent
cations of 2 in the crystal, H atoms, triflate anion and numbering of
C-atoms in the aromatic rings omitted due to reasons of clarity. Selected
bond lengths [pm] and angles [°]for both independent molecule cations:
Rh-P: 222.83(6), 223.48(6), 223.82(6), 224.79(6); Rh1-C (cod):
224.5(2), 224.9(2), 226.9(2), 227.0(2), 227.2(2), 228.0(2), 228.3(2),
229.0(2); P-O: 166.35(16), 166.71(16), 167.05(16), 167.10(16); P-Rh-
P: 68.71(2), 68.78(2); Rh-P-O: 95.92(6), 96.28(6), 96.52(6), 96.86(6);
P-O-P: 98.38(8), 98.51(8)
Ph2POPPh2 in 2 by an isoelectronical NR-group leading to
ligands of the type RN(PR02)2. These ligands are more
easily accessible and accordingly many of their transition
complexes are known in the literature [23–26] but only
very few Rh-complexes are reported [27, 28]. Function-
alisation enabling e.g. an immobilisation of these ligands
can relatively easily be achieved at the substituent directly
linked to the N-atom [29].
The ligand Ph2POPPh2 itself is not easily available, and
a further functionalisation can be introduced only in the
substituents at the P-atoms. So our idea at this point was
addressed to the application of the isolobal-principle [22]
and therefore to replace the O-atom in the POP-ligand
We synthesised the ligand PrN(PPh2)2 (3) by reaction of
PrNH2 with two equivalents of ClPPh2 in presence of NEt3,
and the ligand could be isolated after purification in 85 %
yield according to [30–32]. Further reaction of one
equivalent of the ligand with [Rh(cod)2]OTf lead to the
nearby quantitative formation of the complex [Rh(cod)(g2-
P,P0-PrN(PPh2)2)](OTf) of which from CDCl3 solution a
single crystal was isolated being consistent with
4*OTf*CDCl3 (Fig. 3).
In the cation 4 the geometry at the Rh center can be
described as distorted square planar due to the g2-P,P0-
coordination of the ligand PrN(PPh2)2. Like the O-atom in
2 the N-atom in 4 is part of the four membered planar Rh–
P–N–P ring (mean deviation from the plane = 0.8 pm).
The overall structural features of 4 show a general high
Fig. 1 The phoszone ligand 1 from the Gru¨tzmacher’s group
123

Highly Active Hydroformylation Catalyst
675
In this paper only the first results on hydroformylation
experiments are presented. The reactions are performed in
pure 1-octene leading to an initial 1-octene concentration
of *6.42 mol/l or in thf/1-octene mixtures so that the in
Table 1 presented concentrations are reached. By this
approach catalytic experiments with very high substrate to
metal ratios of more than 100,000 could be performed.
Comparing the concentration versus time profiles it can be
noted that the system is highly active in both terms of
hydroformylation and isomerisation (Figs. 4, 5, 6, 7, 8, 9).
In the reaction mixtures analysed different octene deriva-
tives and their hydroformylation products are observed.
The intermediates or products produced in most significant
amounts are 2-octene, nonanal, and 2-methyloctanal,
respectively. During the reaction progress all the starting
material (1-octene) is consumed, turnover numbers (ton)
equal to the substrate to metal ratio are obtained in all
experiments (s/m-ratio, Table 1).
C7
C1
C25
N1
P1
P2
C35
Rh1
C36
C37
C13
C19
C34
C31
C32
C38
Fig. 3 View to the molecular structure of 4*OTf*CDCl3 in the
crystal, H atoms, triflate anion, solvent molecules and numbering of
C-atoms in the aromatic rings and the side chains omitted due to
reasons of clarity. Selected bond lengths [pm] and angles [°]: Rh-P:
224.0(1), 225.7(1); Rh1-C (cod): 223.4(3), 224.3(3), 225.8(3),
228.6(3); P-N: 170.7(3), 170.8(3); P-Rh-P: 70.51(3); Rh-P-N:
94.93(9), 95.59(9); P-N-P: 98.95(14)
In the meantime an immobilised catalyst system is
developed and tested [36]. An alternative approach
involving magnetic nano-particles for catalyst separation
has been developed [37]. Reactions in sc CO2 have been
performed [38] as well as more detailed kinetic investiga-
tions on a series of Rh-complexes stabilised with ligands of
the type RN(PPh2)2 [39]. These results will be subject of
future publications; whereas this contribution is focused on
the discovery and the description of first catalytic experi-
ments which are discussed in detail now.
similarity and comparability to the bonding of the ligand
Ph2POPPh2 in the cation 2. Accordingly due to the
embedding of the Rh-atom in the four membered ring
system the P–Rh–P bond angle is with 70.5° small; and
relatively short Rh–N distances are observed (291.2 and
291.5 pm, respectively, versus a Rh–O distance of
293.6 pm in 2) enabling a certain through space interaction
between those atoms [33].
Generally in all experiments the product distribution
obtained can be explained by two parallel ongoing processes
in agreement with the in the literature discussed mechanisms
[40, 41]: isomerisation and hydroformylation leading to
mixtures of isomeric octenes and C9-aldehydes (Fig. 10);
hydrogenation is practically not observed. In all experi-
ments the concentrations versus time profiles show common
behaviours: the concentration of 2-octene in the reaction
mixture increases by the time and - after having reached a
maximum - it decreases in the following period (Table 1;
Figs. 4, 5, 6, 7, 8, 9). Accordingly the 1-octene concentra-
tion decreases rapidly at the beginning of the reactions, it is
consumed by the isomerisation to 2-octene and the hydro-
formylation to nonanal and/or 2-methyloctanal. Once after
significant amounts of 2-octene are present in the reaction
mixture one possible isomerisation product of 2-octene,
namely 3-octene, is formed in observable concentrations.
Hydroformylation proceeds anyway at the same time which
is reflected in the by time increasing concentrations of
nonanal and 2-methyloctanal. After the 1-octene in the
reaction mixture is nearby completely consumed but con-
siderable amounts of 2-octene are still present the hydro-
formylation of 2-octene proceeds and as a result the
2-methyloctanal formation rate becomes predominant over
the one for nonanal. Accordingly for all experiments the
Due to the generally high congruence in the structural
data of the cations 2 and 4 we expect comparable catalytic
features for Rh-complexes stabilised by Ph2POPPh2 or
PrN(PPh2)2. Under the reaction conditions it is in agree-
ment with the literature very likely that from [Rh(co-
d)2]OTf in the presence of Ph2POPPh2 or PrN(PPh2)2 the
catalytically active complexes [Rh(H)(g2-P,P0-Ph2POP-
Ph2)(CO)2], or [Rh(H)(g2-P,P0-PrN(PPh2)2)(CO)2] are
formed, respectively [34, 35]. Accordingly, here hydro-
formylation experiments are presented with an si-situ cat-
alyst formed from the [Rh(cod)2]OTf and PrN(PPh2)2
(Table 1, entries 1–4). Additionally, results are presented
using an in situ catalyst prepared from the phoszone ligand
1 and appropriate amounts of [Rh(cod)2]OTf (Table 1,
entry 5) and the pre-catalyst [(g2-P,P0-PrN(PPh2)2)Rh(a-
cac)] (Table 1, entry 6), respectively. The catalytic exper-
iments are performed at different pressures at 100 °C as at
lower temperatures the catalysts show a considerably lower
activity. Kinetic concentrations versus time profiles are
obtained by sampling followed by GC-fid analyses
(Figs. 4, 5, 6, 7, 8, 9; Table 1).
123

676
E. Piras et al.
Table 1 Comparison of the hydroformylation experiments
Run, fig
p (bar)
c (cat)
(mmol/l)
c (lig)
(mmol/l)
s/m-ratio
tof50 (h-1
)
Snonanal
(%)
S2-mo
(%)
S2-o
(%)
S3-o
(%)
tof50/HF
(h-1
tof50/Iso
(h-1
)
)
1, 4
2, 5
3, 6
4, 7
5, 8
6, 9
40
54
0.0379
0.0339
0.0318
0.0271
0.0270
0.0405
0.045
0.038
0.041
0.037
0.075
0.0405
169,600
189,800
67,300
13,510
19,470
50,500
25,780
37,660
57,000
11.5
22.9
14.6
44.6
45.8
45.1
2.7
7.5
81.5
66.4
76.5
36.2
34.1
35.3
4.3
3.0
4.5
1.2
1.4
0.8
1,920
5,920
11,580
13,510
40,900
9,640
40
4.3
9,545
120
120
125
118,600
119,000
122,700
17.8
18.5
18.6
16,090
24,210
36,300
13,370
20,600
T: 100 °C. s/m-ratio: substrate to metal ratio. All selectivities (S) calculated for 50 % conversion. S2-mo: Selectivity 2-methyloctanal. S2-o
:
Selectivity towards cis and trans 2-octene. S3-o: Selectivity towards cis and trans 3-octene. All not mentioned products are detectable only in
traces. tof50 refers to the activity at 50 % conversion, tof50/HF to the hydroformylation activity, and tof50/Iso to the isomerisation activity at 50 %
conversion. Entries 1–3 carried out in 200 ml of pure 1-octene (c0 = 6.428 mol/l), the other entries are diluted with thf according to s/m-ratio.
Entry 5 has been performed with the phoszone ligand 1, in all other experiments PrN(PPh2)2 (3) as the ligand was used. In entry 6 the complex
[(g2-P,P0-PrN(PPh2)2)Rh(acac)] was used as the pre-catalyst
6
5
4
3
2
1
0
6
5
4
3
2
1
0
0
10
20
30
40
50
0
5
10
15
20
25
t [h]
t [h]
Fig. 4 concentration versus time curve for a hydroformylation
experiment of 1-octene with an situ-catalyst formed from [Rh(co-
d)2]OTf and PrN(PPh2)2. P = 40 bar. c0(cat) = 0.0000379 mol/l,
c0(1-octene) = 6.428 mol/l ([1-octene]/[Rh] (at t = 0 h) = 169,600).
Filled triangle: 1-octene, filled inverse triangle: 2-octene, left
pointing filled triangle: 3-octene, right pointing filled triangle:
4-octene, filled diamond: nonanal, filled circle: 2-methyloctanal,
filled square: 2-ethylheptanal, filled star: 2-propylhexanal
Fig. 5 concentration versus time curve for a hydroformylation
experiment of 1-octene with an situ-catalyst formed from [Rh(co-
d)2]OTf and PrN(PPh2)2. P = 54 bar. c0(cat) = 0.00003387 mol/l,
c0(1-octene) = 6.428 mol/l ([1-octene]/[Rh] (at t = 0 h) = 189,800).
Filled triangle: 1-octene, filled inverse triangle: 2-octene, left
pointing filled triangle: 3-octene, right pointing filled triangle:
4-octene, filled diamond: nonanal, filled circle: 2-methyloctanal,
filled square: 2-ethylheptanal, filled star: 2-propylhexanal
n/i-ratio changes by the time: at 50 % conversion n/i-ratios
are found in the range between 2.5 (entries 4, 5, 6, Table 1)
and 4 (entry 1 but with low hydroformylation activity)
whereas at the end of the reaction (nearby complete olefine
consumption) they are determined in the range between 1
and 2 (Figs. 4, 5, 6, 7, 8, 9).
In the experiments at lower pressures the isomerisation
activity of the catalyst is significantly higher (up to six
times) than its hydroformylation activity (Table 1: entries
1–3, Fig. 4, 5, 6). This is reflected in higher concentration
of 2-octene as the preliminary isomerisation product
compared to the corresponding sum of the concentrations
of the aldehydes formed. Accordingly in these experiments
the concentration of 2-octene in the reaction mixture
remains high over the whole reaction course: in the same
magnitude as the one of n-nonanal. A significant pressure
influence on the selectivity is observed: increasing the
pressure to 120 bar the hydroformylation activity of the
system becomes approximately two times higher than its
isomerisation activity (Table 1). If the pressure is increased
For evaluating the data closer as reference the data at
50 % conversion are taken as they are representative and
not deviated from low or already very high conversions.
tof50 represents the tof at 50 % and with respect to the
selectivity the tof50/HF the hydroformylation activity of the
catalyst at 50 % conversion, and the tof50/iso reflects its
isomerisation activity, all at 50 % conversion, respectively
(Table 1).
123

Highly Active Hydroformylation Catalyst
677
3
2
1
0
2
1
0
0
1
2
3
4
5
6
7
0
1
2
3
4
5
t [h]
t [h]
Fig. 8 concentration versus time curve for a hydroformylation
experiment of 1-octene with an situ-catalyst formed from [Rh(co-
d)2]OTf and the phoszone 1. P = 120 bar. c0(cat) = 0.000027 mol/l,
c0(1-octene) = 3.214 mol/l ([1-octene]/[Rh] (at t = 0 h) = 119,000).
Filled triangle: 1-octene, filled inverse triangle: 2-octene, left
pointing filled triangle: 3-octene, right pointing filled triangle:
4-octene, filled diamond: nonanal, filled circle: 2-methyloctanal,
filled square: 2-ethylheptanal, filled star: 2-propylhexanal
Fig. 6 concentration versus time curve for a hydroformylation
experiment of 1-octene with an situ-catalyst formed from [Rh(co-
d)2]OTf and PrN(PPh2)2. P = 40 bar. c0(cat) = 0.0000318 mol/l,
c0(1-octene) = 2.14 mol/l ([1-octene]/[Rh] (at t = 0 h) = 67,300).
Filled triangle: 1-octene, filled inverse triangle: 2-octene, left
pointing filled triangle: 3-octene, right pointing filled triangle:
4-octene, filled diamond: nonanal, filled circle: 2-methyloctanal,
filled square: 2-ethylheptanal, filled star: 2-propylhexanal
5
4
3
2
1
0
3
2
1
0
0
1
2
3
4
5
6
0
1
2
3
4
5
6
7
t [h]
t [h]
Fig. 7 concentration versus time curve for a hydroformylation
experiment of 1-octene. P = 120 bar with an situ-catalyst formed
from [Rh(cod)2]OTf and PrN(PPh2)2. c0(cat) = 0.0000271 mol/l,
c0(1-octene) = 3.214 mol/l ([1-octene]/[Rh] (at t = 0 h) = 118,600).
Filled triangle: 1-octene, filled inverse triangle: 2-octene, left pointing
filled triangle: 3-octene, right pointing filled triangle: 4-octene, filled
diamond: nonanal, filled circle: 2-methyloctanal, filled square: 2-ethylh-
eptanal, filled star: 2-propylhexanal
Fig. 9 concentration versus time curve for a hydroformylation
experiment of 1-octene. P = 125 bar. Pre-cat.: [(PrN(PPh2)2)Rh(a-
cac)]. c0(cat) = 0.0000405 mol/l, c0(1-octene) = 4.973 mol/l ([1-
octene]/[Rh] (at t = 0 h) = 122,700). Filled triangle: 1-octene, filled
inverse triangle: 2-octene, left pointing filled triangle: 3-octene, right
pointing filled triangle: 4-octene, filled diamond: nonanal, filled
circle: 2-methyloctanal, filled square: 2-ethylheptanal, filled star:
2-propylhexanal
to more than 160 bar the reaction rate drops again. This
pressure influence could be explained by the assumption
that at lower pressure the concentration of syngas in the
reaction mixture is lower and therefore the corresponding
reaction rate involving syngas as the substrate (the hydro-
formylation) is lower as well. So the hydroformylation
raction rate is influenced positively by a pressure increase.
This pressure influence is not seen for the isomerisation, so
its reaction rate does not show a real pressure influence.
These findings are confirmed by all the experiments
reported (Figs. 4, 5, 6, 7, 8, 9); Table 1): e.g. comparing
the first two entries (Table 1) at 40 and 54 bar pressure but
under otherwise identical conditions one can see that the
hydroformylation activity is positively influenced by the
123

678
E. Piras et al.
1-o
2-o
3-o
4-o
O
O
nonanal
2-mo
2-ph
2-eh
O
O
Fig. 10 isomerisation and hydroformylation products of 1-octene, overview. 1-o: 1-octene. 2-o: 2-octene. 3-o: 3-octene. 4-o: 4-octene. 2-mo:
2-methyloctanal. 2-eh: 2-ethylheptanal. 2-ph: 2-propylhexanal
pressure increase whereas the isomerisation activity shows
only a slight increase. This results in a better selectivity
towards the hydroformylation products in the reaction
mixture.
3 Conclusion
We show that the phoszone ligand 1 is easily transformed
into the POP ligand Ph2POPPh2 which has been proven by
xray analysis of the complex [Rh(cod)(g2-P,P0-Ph2POP-
Ph2)](OTf) (2*OTf). We assume that complexes of the type
[Rh(H)(g2-P,P0-Ph2POPPh2)(CO)2] form the catalytically
active species in the hydroformylation of 1-octene when
the phoszone ligand 1 is involved. As POP ligands are not
easily accessible we prepared the corresponding PNP
ligand PrN(PPh2)2 (3), showed its g2-P,P0-coordination
behaviour in the complex [Rh(cod)(g2-P,P0-PrN(PPh2)2)]
(OTf) (4*OTf), and tested it as an in situ system as well in
the hydroformylation of 1-octene. A pressure dependency
in the system PrN(PPh2)2/[Rh(cod)2]OTf is observed
according to its high activity and showing enhanced
hydroformylation activity at pressures up to 140 bar.
The results show that the system PrN(PPh2)2/[Rh(co-
d)2]OTf generally agrees in terms of catalytic selectivity and
activity with the one obtained from the phoszone 1 and
[Rh(cod)2]OTf being a sign for structural and electronic
comparability of the active catalyst. Both systems show high
hydroformylationactivity with tof50 upto25,000 h-1. Activity
can be further increased by changing the pre-catalyst nature
from ionic to neutral: with [(g2-P,P0-PrN(PPh2)2)Rh(acac)] as
pre-catalystahydroformylationtof50 of 36,300 h-1 is obtained
at identical selectivity for the same pressure.
Experiments 3, 4 and 5 (Table 1) are executed in thf/
1-octene mixtures leading to lower 1-octene concentrations
compared to the experiments 1, 2 and 6. This leads to a
different polarity in the system which seems to effect the
activity positively. Accordingly in entry 3 with the highest
thf concentration the highest activity of the whole system is
observed while the selectivity remains comparable to the
experiments performed at the same pressure.
Comparing entries 4, 5 in Table 1 a difference in the
catalyst system can be noted: in entry 4 PrN(PPh2)2 (3) was
used as the stabilising ligand whereas in entry 5 the pure
phoszone ligand 1 was present in the reaction mixture
under otherwise very similar conditions (but with respect
that two moieties of 1 might be transformed into one
molecule of Ph2POPPh2). The presence of the phoszone
ligand 1 instead of the prepared PrN(PPh2)2 (3) leads to
a higher activity of the system. The selectivity in the
Rh-catalysed hydroformylation of 1-octene is in both cases
comparable. This could be explained by the similar struc-
tural features of the two ligands in the Rh-coordination
sphere, while their electronically different features then
cause the different activity.
However, to our opinion the overall similarity of the
catalytical behaviour of the two ligands in the catalysis is a
strong sign that the finally catalytically active complex in
the Rh-catalysed hydroformylation of 1-octene involving
the phoszone 1 is a complex with the Ph2POPPh2 ligand in
the coordination sphere of the transition metal. This was
the motivation of our studies here showing that the ligand
PrN(PPh2)2 in the Rh-catalysed hydroformylation of
1-octene leads to high activity as well.
As the catalyst is not only a hydroformylation but as
well an effective isomerisation catalyst it seems to be
promising to perform as well more extended investigation
on combined isomerisation hydroformylation experiments
with sterically hindered olefins as the substrate. Further-
more immobilisation of the catalyst system is under
research towards the development of a continuously driven
hydroformylation process for higher olefins.
Indeed changing the pre-catalyst from an ionic in situ
system (formed from [Rh(cod)2]OTf and PrN(PPh2)2) to
the neutral pre-catalyst [(g2-P,P0-PrN(PPh2)2)Rh(acac)]
this influences on the activity (entry 6, Table 1; Fig. 9). A
significantly enhanced activity can be obtained with with a
tof50 of *57,000 h-1 realised with the ligand PrN(PPh2)2.
The selectivity remains comparable to all other experi-
ments at this pressure (Table 1).
4 Experimental
4.1 Materials and Methods
All preparations were carried using standard Schlenk
techniques. Preparation of PrN(PPh2)2 (3) was performed
123

Highly Active Hydroformylation Catalyst
679
according to [30]. The catalytic experiments in this publi-
cation were performed in a 300 ml autoclave. Prior to the
experiment the autoclave was pressurised three times with
N2 and the pressure was released in order to inert the
system. The catalyst was filled into the autoclave together
with the substrate and—if desired—some additional sol-
vent and the system was heated up to the reaction tem-
perature while pressurising to the required pressure. The
time when the reaction temperature was reached was
taken as t0. Samples were taken and applied to GC analyses
(HP 5830).
tube. Details of the single crystal analyses see footnote 1.
No further analytical data can be provided.
4.1.2 Preparation of 4*OTf
468.3 mg (1.0 mmol) of [Rh(cod)2](OTf) were dissolved in
20 ml of CH2Cl2, a solution of 427.5 mg (1.0 mmol)
PrN(PPh2)2 in 20 ml of CH2Cl2 was added and the solution
was stirred for 30 min at RT. The solvent was removed in
vacuum and the residual solid washed three times with
10 ml of Et2O and pentane, successively, and dried in
vacuum. 733.1 mg (0.93 mmol, 93 %) of [Rh(cod)(g2-P,
P0-PrN(PPh2)2)](OTf) (4*OTf) were isolated of which
single crystal were obtained from a solution of the 4*OTf
in CDCl3 being of the consistence 4*OTf*CDCl3. Details
of the single crystal analyses see footnote 1.
NMR analyses were performed on a Bruker AVANCE
250 NMR system. XRD measurements were performed on
Siemens SMART CCD 1000 diffractometer with mono-
chromated MoKa-irradiation collecting a full sphere of
data in the h–range from 1.57 to 28.34°. Frames were
collected with an irradiation time of 6 s (2*OTf) or 40 s
(4*OTf*CDCl3) per frame and x–scan technique with
Dx = 0.45°.1 Data were corrected to Lorentz and polari-
sation effects and an empirical adsorption correction with
sadabs [42] was applied. The structures were solved by
direct methods and refined to an optimum R1 value with
SHELX-97 [43]. Visualisation for evaluation was per-
formed with xpma [44] and figures were created with
winray [45]. The structures have been deposited at the
CCDC referring to the CCDC numbers 915597 (2) and
915598 (4).
1H (CDCl3) d/ppm = 7.8–7.4 (m, 20H, aromatic pro-
tons), 5.21 (s, 4H, coordinated cod CH), 2.72 (m, 2H,
NCH2), 2.4 (s, br, 8H, cod CH2), 1.03 (m, 2H,
3
CH2CH2CH3), 0.43 (t, JH–H = 7.2 Hz, 3H, CH3).
13C{1H} (CDCl3) d/ppm = 135.0, 133.6, 130.4, 128.7
(aromatic C), 118.0 (coordinated cod C), 35.7 (NCH2),
28.1 (cod CH2), 27.0 (CH2CH2CH3), 11.3 (CH3).
31P{1H} (CDCl3) d/ppm = 52.1 (d, 1JP–Rh = 136.2 Hz).
C36H39F3NO3P2RhS, calc. (%): C 54.90; H 4.99; F 7.24;
N 1.78; O 6.09; P 7.87; Rh 13.07; S 4.07—found (%): C
55.13; H 5.16; N 1.91, S 4.19.
4.1.1 Preparation of 2*OTf
4.1.3 Preparation of [PrN(PPh2)2Rh(acac)]
Single crystals of [Rh(cod)(g2-P,P0-Ph2POPPh2)](OTf)
(2*OTf, Fig. 2) were obtained overnight from a solution of
0.02 mmol of the ligand 1 and 0.01 mmol of [Rh(co-
d)2]OTf and liquid carbon dioxide (1 ml) inside a HP NMR
258 mg (1.0 mmol) of [(CO)2Rh(acac)] were dissolved in
10 ml of CH2Cl2, a solution of 427 mg (1.0 mmol) PrN(PPh2)2
in 20 ml of CH2Cl2 was added and the solution was stirred for
30 min at RT. The solvent was removed in vacuum and the
residual solid washed three times with 10 ml of Et2O and
pentane, successively, and dried in vacuum. 550 mg
(0.92 mmol, 92 %) of [(PrN(PPh2)2)Rh(acac)] were isolated.
1H (CDCl3) d/ppm = 8.0–7.0 (m, 20H, aromatic pro-
1
Structural details for 2*OTf: Reflections collected/unique/observed
(I [ 2r): 40,467/15,873/12,529 [R(int) = 0.222]; parameters refined:
800; formula C66H65F6O8P4Rh2S2 (two molecules), MM per mole-
cule 747.0 g/mol; T = 200(2) K; triclinic, P-1 (No. 2, Z = 2); a =
1,149.5(1) pm; b = 1,515.5(1) pm; c = 2,049.6(2) pm; a = 71.784(1)°;
b = 78.518(1)°; c = 70.494(1)°; V = 3,179.2(5) 9 106 pm3; q (calc.) =
3
tons), 5.51 (s, 1H, COCHCO), 2.72 (m, JH–H = 8.3 Hz,
3
3JP–H = 5.8 Hz, 2H, NCH2), 1.07 (tq, JH–H = 7.8 Hz,
3
1.561 g/cm3; Absorption coefficient = 0.758 mm-1
;
F(000) =
3JH–H = 7.2 Hz, 2H, CH2CH2CH3), 0.44 (t, JH–H
=
1,522; Goof (F2) = 1.035; crystal size = 0.8 9 0.8 9 0.6 mm3;
index ranges -16 \ h \ 16, -21 \ k \ 20, -28 \ l \ 28; com-
pleteness to h = 28.34: 95.2 %; R1 (I [ 2r) = 0.0316, wR2 =
0.0851 (all data), largest difference peak and hole: 0.836 and
-0.513 9 10-6 pm-3. Structural details for 4*OTf*CDCl3: Reflections
collected/unique/observed: 25,128/9,067/8,250 [R(int) = 0.0317];
parameters refined: 467; formula C37H40Cl3F3NO3P2RhS, MM 906.96
g/mol; T = 200(2) K; monoclinic, Pn (No. 7, Z = 2); a =
1,158.8(3) pm; b = 944.1(3) pm; c = 1,815.7(5) pm; b = 105.465(5)°;
V = 1,914(1) 9 106 pm3; q (calc.) = 1.573 g/cm3; Absorption coef-
ficient = 0.846 mm-1; F(000) = 924; Goof (F2) = 1.014; crystal
size = 0.6 9 0.6 9 0.5 mm3; index ranges -15 \ h \ 15, -12 \
k \ 12, -24 \ l \ 24; completeness to h = 28.34: 97.7%; R1
(I [ 2r) = 0.0307 wR2 = 0.0698 (all data), largest difference peak
7.3 Hz, 3H, CH3).
31P{1H} (CDCl3) d/ppm = 71.6 (d, 1JP–Rh = 122.9 Hz).
C32H34NO2P2Rh, calc. (%): C 61.06; H 5.44; N 2.23; O
5.08; P 9.84; Rh 16.35—found (%): C 62.17; H 5.82; N 2.42.
Acknowledgment The authors would like to express their deepest
thanks to the KIT and the ETH Zu¨rich for the financial and Prof. Dr.
Eckhard Dinjus for the general support of this work.
References
¨
1. Franke R, Selent D, Bornera A (2012) Chem Rev 112:5675
2. Ungvary F (2007) Coord Chem Rev 251:2072
and hole: 0.653 and -0.325 9 10-6 pm-3
.
123

680
E. Piras et al.
3. Van Rooy A, Orij EN, Kamer PCJ, Van Leeuwen PWNM (1995)
Organometallics 14:34
26. Bowen LE, Haddow MF, Orpen AG, Wass DF (2007) Dalton
Trans 1160
4. Van Rooy A, de Bruijn JNH, Roobeek KF, Kamer PCJ, Van
Leeuwen PWNM (1996) J Organomet Chem 507:69
5. Van Leeuwen PWNM, Roobeek CF (1983) J Organomet Chem
258:343
27. Simon-Manso E, Valderrama M (2006) J Organomet Chem
691:380
28. Valderrama M, Contreras R, Boys D (2003) J Organomet Chem
665:7
6. Jongsma T, Challa G, Van Leeuwan PWNM (1991) J Organomet
Chem 421:121
7. Van Rooy A, Orij EN, Kamer PCJ, Van Den Aardweg F, Van
Leeuwen PWNM (1991) J Chem Soc Chem Commun 16:1096
8. Desset SL, Cole-Hamilton DJ (2009) Angew Chem Int Ed
49:1472
9. van der Slot SC, Kamer PCJ, van Leeuwen PWNM, Fraanje J,
Goubitz K, Lutz M, Spek AL (2000) Organometallics 19:2504
10. Magee MP, Luo W, Hersh WH (2002) Organometallics 21:362
11. E. Piras, Dissertation no. 16317 (ETH Zu¨rich, 2005)
12. Dahmen N, Griesheimer P, Makarczyk P, Pitter S, Walter O
(2006) J Organomet Chem 690:1467
29. Posset T, Rominger F, Blu¨mel J (2005) Chem Mater 17:586
30. Ewart G, Lane AP, McKechnie J, Payne DS (1964) J Chem Soc
1543
31. Yip SK, Lam WH, Zhu N, Yam VWW (2006) Inorg Chim Acta
359:3639
32. Balakrishna MS, Teipel S, Pinkerton AA, Cavell RG (2001) Inorg
Chem 40:1802
33. Walter O, Huttner G, Kern R (1996) Z Naturforsch B 51:922
34. C. Claver, P.W. van Leeuwen, Rhodium Catalyzed Hydrofor-
mylation (Springer; 2000)
¨
35. Maura R, Steele J, Vendier L, Arquier D, Bastin S, Urrutigoıty M,
Kalck P, Igau A (2011) J Organomet Chem 696:897
13. Kainz S, Koch D, Baumann W, Leitner W (1997) Angew Chem
Int Ed Engl 36:1628
36. Ionescu C (2009) Katalyse und Katalysatoren zur Hydroformy-
lierung langkettiger Olefine in u¨berkritischem Kohlendioxid.
¨
14. Koch D, Leitner W (1998) J Am Chem Soc 120:13398
15. Wong EH, Prasad L, Gabe EJ, Bradley FC (1982) J Organomet
Chem 236:321
Dissertation, Universitat Heidelberg
37. Gorschinski A, Habicht W, Walter O, Behrens S, Ionescu C,
Powietzka B (in preparation)
16. Bradley FC, Wong EH, Gabe EJ, Lee FL, Lepage Y (1987)
Polyhedron 6:1103
17. Pavlik S, Mereiter K, Puchberger M, Kirchner K (2005) Orga-
nometallics 24:3561
18. Irvine DJ, Cole-Hamilton DJ, Barnes J, Hodgson PKG (1989)
Polyhedron 8:1575
19. Irvine DJ, Glidewell C, Cole-Hamilton DJ, Barnes JC, Howie A
(1991) J Chem Soc Dalton Trans 263:1765
20. Bravo J, Castro J, Garcia-Fontan S, Rodriguez-Martinez MC,
Rodriguez-Seoane P (2006) Eur J Inorg Chem 3028
21. Domide D, Kaifer E, Mautz J, Walter O, Behrens S, Himmel HJ
(2008) Eur J Inorg Chem 3177
22. Hoffmann R (1981) Science 211:95
23. Durap F, Biricik N, Gumgum B, Ozkar S, Ang WH, Fei Z,
Scopelliti R (2008) Polyhedron 27:96
38. Maniut C (2007), Reaktionstechnische Aspekte zur Hydro-
formylierung in u¨berkritischem Kohlendioxid. Dissertation,
¨
Universitat Heidelberg
¨
39. F. Wurst, Uberfu¨hrung der Hydroformylierung mit immobili-
sierten Rhodium-Katalysatoren in ein kontinuierliches Verfahren
¨
im Labormaßstab. (Dissertation, Universitat Heidelberg, to be
concluded in 2013)
40. Heck RF, Breslow DS (1961) J Am Chem Soc 83:4023
41. Evans D, Yagupsky G, Wilkinson G (1968) J Chem Soc A 3133
42. SADABS, Siemens area detector absorption correction pro-
gramme (Siemens, 1997)
¨
43. SHELX-97, G. Sheldrick (University of Gottingen, Germany,
1997)
44. Xpma L, Zsolnai G (1997) Huttner. University of Heidelberg,
Germany
45. Winray R, Soltek G (1997) Huttner. University of Heidelberg,
Germany
24. Gallo V, Mastrorilli P, Nobile CF, Braunstein P, Englert U (2006)
Dalton Trans 8:2342
25. Gumgum B, Biricik N, Durap F, Ozdemir I, Gurbuz N, Ang WH,
Dyson P (2007) J Appl Organomet Chem 21:11
123