Post on 10-Jun-2020
Investigations on rhodium-catalyzed asymmetrichydroformylationCitation for published version (APA):Zijp, E. J. (2007). Investigations on rhodium-catalyzed asymmetric hydroformylation. Eindhoven: TechnischeUniversiteit Eindhoven. https://doi.org/10.6100/IR627570
DOI:10.6100/IR627570
Document status and date:Published: 01/01/2007
Document Version:Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:
• A submitted manuscript is the version of the article upon submission and before peer-review. There can beimportant differences between the submitted version and the official published version of record. Peopleinterested in the research are advised to contact the author for the final version of the publication, or visit theDOI to the publisher's website.• The final author version and the galley proof are versions of the publication after peer review.• The final published version features the final layout of the paper including the volume, issue and pagenumbers.Link to publication
General rightsCopyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright ownersand it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, pleasefollow below link for the End User Agreement:www.tue.nl/taverne
Take down policyIf you believe that this document breaches copyright please contact us at:openaccess@tue.nlproviding details and we will investigate your claim.
Download date: 17. Jun. 2020
Investigations on Rhodium-Catalyzed Asymmetric Hydroformylation
PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de Rector Magnificus, prof.dr.ir. C.J. van Duijn, voor een commissie aangewezen door het College voor Promoties in het openbaar te verdedigen op dinsdag 3 juli 2007 om 16.00 uur door Eric Jurriën Zijp geboren te Sleeuwijk
Dit proefschrift is goedgekeurd door de promotor: prof.dr. D. Vogt Copromotor: dr. H.C.L. Abbenhuis
Investigations on Rhodium-Catalyzed Asymmetric Hydroformylation / by Eric Jurriën Zijp
Eindhoven : Eindhoven University of Technology, 2007
A catalogue record is available from the Eindhoven University of Technology Library
ISBN: 978-90-386-1055-9
Omslag: Oranje Vormgevers Eindhoven, naar een idee van Henrike Klein Ikkink
Druk: Universiteitsdrukkerij, Technische Universiteit Eindhoven
Copyright © 2007 Eric J. Zijp
Table of Contents
Table of Contents
1
19
51
69
87
103
105
107
111
Chapter 1 Introduction and Scope
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their
Coordination Behavior
Chapter 3 DFT Study into Models of Bisaminophosphine
Ligands
Application of Bisaminophosphine Ligands in
Rh-Catalyzed Asymmetric Hydrogenation
Chapter 4 Application of Bisaminophosphine Ligands in Rh-
Catalyzed Asymmetric Hydroformylation
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-
Catalyzed Asymmetric Hydroformylation
Summary
Samenvatting
Dankwoord
Curriculum Vitae
Chapter 1
Introduction and Scope
Asymmetric hydroformylation is considered as a
transformation of high potential for the fine chemical industry
since decades. This is due to the role aldehydes and derived
building blocks play in organic synthesis. However, few ligand
systems meet the requirements for industrial applications in
terms of selectivity and activity, though significant progress has
been made in recent years. New classes of ligands and deeper
understanding of the stereoselective step obtained by
spectroscopic studies and theoretical investigations have
brought this reaction close to the edge of real application. More
theoretical insight is necessary to be able to make the final step.
Chapter 1 Introduction and Scope
1.1 Introduction
Since the discovery by Roelen in 1938[1] regioselective hydroformylation (Eq. 1) has
developed into an important industrial process for the production of linear aldehydes
(softeners for plastics and detergent industry). An industrial application of the
asymmetric hydroformylation reaction, leading to the branched aldehyde, however still
does not exist. The expected potential of the reaction can be illustrated by the interest
shown by academia as well as industry, resulting in many publications including
reviews[2-5] and an excellent book by van Leeuwen and Claver[6] on the matter.
R R
CHO
RCHO+
branched linear
CO/H2
cat. *
(1)
Here we emphasize on the possible applications of asymmetric hydroformylation, a
historic overview on the catalyst-systems and ligands used for the reaction and the
knowledge of the important factors determining the ability of a ligand to induce
stereoselection in this transformation.
1.2 Application
Asymmetric hydroformylation potentially could be an important reaction for the
synthesis of chiral aldehydes as intermediates in drug synthesis. A recent review
illustrates that several substrate classes can be selectively hydroformylated to the
branched product.[7] Either the tendency of substrates to chelate to the catalyst (e.g.
vinyl acetate) or the electronic preference for branched aldehyde formation (e.g.
vinylarenes) can be exploited by the synthetic chemist. The branched selective
hydroformylation of unfunctionalized alkenes remains a challenge, as well as the regio-
control within internal alkenes to form one branched aldehyde over the other branched
product.
Other emerging technologies in industry are the combination of biocatalysis and
hydroformylation (e.g. biocatalysts provide access to chiral substrates for
hydroformylation)[8] or the combination of chemocatalytic reactions (e.g. tandem
2
Chapter 1 Introduction and Scope
hydroformylation / reductive amination and asymmetric allylic alkylation providing
chiral alkenes for hydroformylation).[9,10]
1.3 Platinum/Tin Catalyzed Asymmetric Hydroformylation
The first studies applying platinum/tin as the catalyst in asymmetric hydroformylation
appeared not long after the introduction of the rhodium catalyzed analogue.[11-13]
Mainly diphosphines were applied (Figure 1). The use of DIOP (1) in asymmetric
hydroformylation of styrene gave mainly linear aldehyde (b/l = 0.3) and a maximum ee
of 22%.[14] Modification of the ligand to BDP-DIOP (2) gave somewhat better
selectivity to the branched aldehyde and higher ee's up to 65%.[15]
The disadvantages of the platinum/tin based systems are
• The low selectivity to the branched product
• The need of an excess of (poisonous) SnCl2 with respect to platinum
• Racemization of the product aldehydes caused by the Lewis acid SnCl2
• High amount of alkene hydrogenation
• Undesired alkene isomerization
• Low activity compared to rhodium-based catalysts
O
O
PPh2
PPh2
O
O
P
P
NH
PPh2PPh2
PPh2 PPh2
1 2 3
4 5
PPh2
PPh2
Figure 1 Chiral diphosphine ligands used in the Pt/Sn-catalyzed asymmetric hydroformylation. DIOP
(1), BDP-DIOP (2), PPM (3), BDPP (4), BINAP (5).
3
Chapter 1 Introduction and Scope
The racemization of the product aldehydes could be suppressed by trapping the chiral
aldehydes as the acetal by using triethyl orthoformate as the co-solvent. Although the
speed of reaction was much lower than in benzene, the enantioselectivity induced by
PPM (3) as ligand improved from 70% to >96%.[16-17] Meessen et al and Van Duren et
al. showed that the excess of tin(II)chloride could be omitted while retaining a
sufficient activity by the use of large biteangle ligands. The preformation of the catalyst
proved also to be effective with only one equivalent of the tin-source.[18]
Other ligands were developed to gain more insight in the mechanism governing the
enantioselectivity. Kollár et al. showed for the first time that the direction of
enantioselectivity can change sign with increasing temperature in a study with BDPP
(4) as the chiral ligand employed.[19] This was explained by a conformational change of
the six-membered chelate ring. The concept of enantio-inversion was later studied in
more detail with BINAP (5). Using dynamic NMR studies it was shown that this effect
was caused by changes in the conformation of the ligand.[20] At higher temperature
these changes altered the geometry around the platinum atom, which in turn caused the
insertion of the substrate (styrene) to take place from the other enantioface.
Casey et al. studied this phenomenon by performing deuterioformylation.[21] The
enantioselectivity turned out not to be fully determined until the final hydrogenolysis of
the platinum acyl intermediate.
1.4 Rhodium catalyzed asymmetric hydroformylation[7]
The bidentate ligands used in asymmetric hydroformylation were traditionally the
diphosphines developed for asymmetric hydrogenation, i.e. ligands for square planar
complexes. The enantiomeric excesses obtained for rhodium catalysts with these
ligands remained below 60%.
4
Chapter 1 Introduction and Scope
Rh
H
COL
LR
Rh
H
CO
LL
Rh
H
COL
LCO
L
Rh
H
CO
LCO
CHO
R
R
L
Rh
H
CO
L
Rh
COL
LCO
R
Rh
COL
LCO
R
Rh
CO
LL
O R
L
Rh
O R
LCO H
H
R
CHO
towards linear aldehyde
branchedaldehyde
Scheme 1 Catalytic cycle for the branched selective hydroformylation of a terminal linear alkene.
Rhodium catalyzed hydroformylation however requires trigonal bipyramidal
complexes, as is shown in the proposed catalytic cycle for the rhodium-catalyzed
hydroformylation of a linear alkene (Scheme 1) and in more detail for a bidentate
ligand in Scheme 2.
Rh CO
H
CO
P
P PP
Rh CO
HOC
ee ea Scheme 2 Equatorial/equatorial ee and equatorial/axial ea coordination modes in trigonal bipyramidal
Rh-complexes.
The equilibrium between the ee and ea coordinated bidentate ligand to the rhodium
center is an important factor in the capacity of a ligand to induce stereodifferentiation.
This will be shown for different successful classes of ligands (vide §1.5). This resting
state of the catalyst is usually detected using in situ spectroscopic techniques and is
therefore often the target in molecular modelling studies. For 1-alkenes the rate limiting
step could be the hydrogenolysis of the rhodium acyl intermediate.[6]
5
Chapter 1 Introduction and Scope
One of the CO molecules is released from this trigonal bipyramidal complex. Then
alkene coordination followed by insertion in the Rh-H bond gives either a linear or
branched rhodium alkyl species. CO coordination followed by migratory insertion gives
either the branched or the linear rhodium-acyl species. These complexes undergo
hydrogenolysis to release aldehydes and regenerate the Rh-H catalyst. All the steps in
the rhodium-catalyzed hydroformylation can be reversible, with the exception of the
final hydrogenolysis.
The steric constraints of the ligands employed in hydroformylation are different
because of this altered geometry for the Pt/Sn and Rh catalysts. If the alkene
coordinates in the equatorial plane (in the rhodium catalyst), the interaction of the
substrate with the ligand is weaker than in a square planar intermediate (platinum),
certainly if the bidentate phosphine is also coordinated in the equatorial plane.
1.5 Ligands
In the 1990’s three new types of ligands have led to high ee’s in the asymmetric
hydroformylation of styrene, namely diphosphites, the mixed phosphine-phosphite
ligand BINAPHOS and the modular AMPP ligands. In the 2000’s the understanding of
the enantioselective step in the reaction rose through a theoretical investigation of the
latter ligand type applying QM/MM methods. The world-record holder concerning ee
for styrene is a hybrid phosphine-phosphoramidite ligand adapted from BINAPHOS
and was just developed in 2006. Bis-phosphacyclic ligands originally designed for
asymmetric hydrogenation proved also to induce high enantioselectivities in the
asymmetric hydroformylation of several substrates, albeit with rather low activities.
These ligand systems will be discussed in more detail (vide infra).
1.5.1 Diphosphites
Babin and Whiteker at Union Carbide reported the asymmetric hydroformylation of
various alkenes with ee’s up to 90% using bulky diphosphites (6) derived from
(2R,4R)-pentane-2,4-diol (figure 2).[22]
6
Chapter 1 Introduction and Scope
OP
O
O
tBu
tBu
R
R
OP
O
O
tBu
tBu
R
R6
Figure 2 Chiral diphosphite (6) from Union Carbide.
Investigations on the bridge-lengths of other diphosphites based on commercially
available chiral diols showed some distinct features: when the bridging backbone
consists of two carbon atoms the ligands adopt the e-a coordination and low ee’s are
obtained. Ligands having four carbon atoms in the bridge lead to the anticipated e-e
configuration, but also here the ee’s are low. Only when ligands with three carbon
atoms between the phosphite moieties are applied the bidentate ligand sits in the
trigonal plane of the complex and is rigid and bulky enough to lead to high ee’s.[23,24]
Also ligands based on sugar backbones give good ee’s if a three-carbon bridge is
involved.[25]
When the bisphenol groups of the diphosphites were replaced by bisnaphthols a
matched and a mismatched combination was found. By increasing the steric bulk on the
3- and 3’-positions of the bisnaphthol the ee’s increased, with an optimum reached for
SiMe3 groups.[26]
OO
PO
tBu
tBu
OOP
O
tBu
tBu
7
Figure 3 Kelliphite (7).
Although (S,S)-Kelliphite (7) (Figure 3) has 4 carbons between the phosphite moieties
it forms the best performing diphosphite ligand concerning regioselectivity in the
asymmetric hydroformylation of vinyl acetate. Using a beautiful example of a
multisubstrate screening Whiteker and coworkers[27] tested several phosphite-based
7
Chapter 1 Introduction and Scope
ligands on their performance with the benchmark substrates. An unprecedented b/l of
125 was reached along with an excellent ee of 88 % (at 25 °C).
1.5.2 BINAPHOS[28-32]
The benchmark ligand system for years has been the mixed phosphine-phosphite
BINAPHOS (8) (see figure 4). Where C1 symmetric ligands were generally avoided in
favor of C2 symmetric ligands (e.g. the diphosphites, vide supra), in this case the
dissymmetric substitution pattern proved to be very efficient. Most research on ligands
for the asymmetric hydroformylation reaction after this discovery was aimed on the
development of new C1 symmetric catalyst modifiers.
O
O
PO
PPh2
8
Figure 4 Takaya’s (R,S)-BINAPHOS (8).
The shown configuration (R for the bisnaphthyl unit, S for the bisnaphthol unit) was the
most efficient ligand in a larger series. The absolute configuration of the bisnaphthyl
determines the absolute configuration of the product, in the matched pair the used
bisnaphthol in the phosphite moiety has the opposite configuration. At 60 °C both
groups work together and 94% ee is obtained for styrene, when the mismatched pair is
used only 25% ee is reached. The rationalization of this phenomenon lies in the fact the
(R,S)-ligand has the tendency to bring the phosphorus donor atoms in closer proximity,
likewise in the free ligand.
HP-NMR measurements show that the RhH(CO)2(R,S-BINAPHOS(8)) is a single
species in toluene-d8 under CO atmosphere. The more σ-donor phosphine P atom sits in
the plane with the CO ligands whereas the more π-accepting phosphite P atom sits
apical to the hydride. No apical-equatorial interchange is observed at any temperature.
8
Chapter 1 Introduction and Scope
This unique dissymmetric environment in a single catalytically active species seems to
be an important factor in the high enantioselectivity obtained.
1.5.3 AMPP
Aminophosphane phosphinite (AMPP) ligands were developed in the 1980’s and used
for enantioselective hydrogenation.[33] In 2000 a new class of AMPP ligands (9),
bearing a stereogenic P atom of the aminophosphane moiety, was synthesized by Vogt
and coworkers (see figure 5). High enantioselectivities were reached when they were
applied in the asymmetric hydroformylation of vinylarenes.[34] O
MeN
Ar2PPh
Me
PPh
R9
Figure 5 Modular class of AMPP ligands (9) with stereogenic phosphorus atom.
The stereogenic P atom seemed to be essential for obtaining good ee’s. But also here
the coordination of the ligands in the catalytic resting state was important. The ligands
which give high ee all coordinate in a stable equatorial/axial manner, with the
aminophosphane moiety in the axial position. This could be concluded from combined
HP-IR and HP-NMR studies.
The origin of stereoinduction of the AMPP ligands was investigated by combining DFT
and QM/MM calculations.[35] Alkene insertion into the rhodium-hydride bond was the
selectivity-determining step, and not the alkene coordination. Different weak non-
bonding interactions of styrene with the substituents of the stereogenic phosphorus
atom in an axial position are responsible for stereodifferentiation.
1.5.4 Phosphine-phosphoramidite[36]
Zhang and coworkers just recently synthesized the mixed phosphine-phosphoramidite
ligand 10 (R,S)-Yanphos (figure 6) from the chiral synthon NOBIN (2-amino-2’-
hydroxy-1,1’-binaphthyl) and compared the space-filling and stick models (based on
CAChe MM2 calculations) of the active intermediates in hydroformylation reactions
with the ones applying BINAPHOS. Due to the crowded N-substituent the
RhH(CO)2(10) complex provides a deeper and more closed chiral pocket than the
9
Chapter 1 Introduction and Scope
corresponding RhH(CO)2(BINAPHOS) complex. Besides this the complex is
conformationally more rigid. The envision was this could lead to higher asymmetric
induction.
O
ON
PPh2
P
10 Figure 6 Zhang’s hybrid phosphine-phosphoramidite ligand 10.
Ligand 10 proved indeed to give a better performance in terms of ee and conversion
under mild conditions than BINAPHOS for a large number of vinylarenes and vinyl
acetate as the substrates. It has to be noted that not all applied conditions were similar
to the optimized conditions for BINAPHOS. For vinyl acetate the results matched the
previous best result using diazaphospholane ligand (11).[37]
A big advantage of the applied ligand systems seems to be the absence of racemization
of the produced aldehydes which does occur when BINAPHOS is applied, even before
full conversion.
Future developments are aimed at structural variation of the N-substituent of the
phosphine-phosphoramidite ligand for application in asymmetric hydroformylation as
well as other homogeneously metal-catalyzed transformations.
1.5.5 Bis-phosphacycles[38]
Bis-phosphacycles (figure 7), which were known to act as successful ligands in
asymmetric hydrogenation reactions, proved also to be efficient in inducing high
enantioselectivities in asymmetric hydroformylation of several substrates. The
assumption that good hydrogenation ligands are not efficient hydroformylation ligands
is therefore not true. The application of TangPhos (13) in the hydroformylation of
norbornylene was in fact already shown in 2005.[39]
10
Chapter 1 Introduction and Scope
P
H
tBu
PH
tBu
PP
Ph
PhPh
Ph
N
N
O
O
P
O
O
NH
HN
Ph
Ph
N
N
O
O
P
O
O
NH
NH
Ph
Ph
P
tBuH
P
tBuH
11 12
13 14 Figure 7 Bis-phosphacyclic ligands. Diazaphospholane (11), (S)-Binapine (12), (S,S,R,R)-Tangphos (13),
(R,R)-Ph-BPE (14).
Diazaphospholane (11) is the most efficient ligand for the hydroformylation of vinyl
acetate (96% ee) whereas (S)-Binapine (12) induced the highest ee ever reported (94%)
for allyl cyanide. The optimal working conditions should be carefully determined for
individual ligands and substrates. For instance Tangphos (13) should not be used in
excess with respect to rhodium. The non-coordinating anion (acac)- in the formed
[Rh(Tangphos)2]+[acac]- proved to racemize the chiral products, whereas Ph-BPE (14)
did not show the same behavior and is best used in a 2-fold excess. The activities and
regioselectivities were generally low compared to the hybrid ligands Yanphos and
BINAPHOS. The introduction of electronwithdrawing substituents in close proximity
of the phosphorus atoms may overcome this disadvantage. The common elements in
this class of ligands are the phosphacyclic motif and the two-carbon bridging group
between the phosphorus atoms. Wills’ successful bis-diazaphospholidine ligand
ESPHOS (15, see figure 8), which is only suitable for vinyl acetate (ee = 90%, b/l =
16:1) also belongs to this selection[40]
PP
N
NN
N
Ph
Ph
15 Figure 8 Bis-diazaphospholidine ligand ESPHOS (15).
11
Chapter 1 Introduction and Scope
1.5.6 Monodentate phosphorus-based ligands
Due to the low enantioselectivities and despite the good regioselectivities obtained in
the 1970's [41] the use of monodentate chiral phosphorus-based ligands was virtually
abandoned in favor of the more efficient bidentate ligands.[42]
Some literature examples employing monodentate chiral phosphorus-based ligands (see
Figure 9) other than phosphines are the TADDOL-based phosphonite ligand (16) of
Seebach et al. [43] and the diazaphospholidine (17) of Wills and coworkers.[40] The
phosphonite of Seebach induces in the asymmetric hydroformylation of styrene very
good b/l ratios of about 20 and an ee of 20% where the diazaphospholidine shows a
limited b/l ratio and virtually no ee. The bisnaphthyldiamine based diazaphospholidine
(18) of Reetz and coworkers gave the highest ee for styrene (up to 37%).[44]
O
O
O
O
Ph Ph
Ph Ph
P
MeO
PN
N
Ph
N
N
O
O
P
O
O
NH
NH
Ph
Ph
O
OP N
Ph
Ph
N
NP X
R
R
O
OP O
R
R
16 17 18
19 20 21
Figure 9 Monodentate chiral ligands 16-21 employed in Rh-catalyzed asymmetric hydroformylation.
Monodentate phosphoramidites were successfully applied in asymmetric hydrogenation
of various substrates owing to their modular nature and ease of preparation allowing for
the construction of large libraries of ligands to be used in fast identification of new and
efficient catalysts. This prompted a feasibility study of these versatile compounds as
chiral ligands in the rhodium-catalyzed asymmetric hydroformylation of the benchmark
12
Chapter 1 Introduction and Scope
substrates styrene and vinyl acetate. Good activities, chemoselectivities and
regioselectivities were generally achieved. The best ee however reached only a
moderate value of 27% when ligand 19 was used.[45] The highest activities were found
when the phosphites of Whiteker and coworkers (20) were applied[46] and very high
regioselectivities (b:l = 39) were obtained with the phosphacyclic ligand (21) developed
by Clark et al.[37]
22.1 - 5% 22.2 - 14% 22.3 - 68%
O
OP N O
OP N
O
OP N
tBu
tBu
Figure 10 Phosphoramidite ligand series 22.1-22.3 with obtained ee when employed in asymmetric
hydroformylation of allyl cyanide at 60 ºC (L:Rh = 3:1) in benzene.
Best results in terms of enantioselectivity for allyl cyanide as the substrate were
obtained with the phosphoramidites 22 by Ojima and coworkers (see Figure 10).[47] In
the ligand series 22.1-22.3 the importance of the 3- and 3'-positions was once more
confirmed. Upon lowering the temperature to 25 ºC 80% ee was obtained in toluene,
although complete conversion was only reached after 74h. A solvent screening
confirmed that toluene was the best solvent among those tested (THF, dichloromethane
and MeOH).
The authors performed a molecular modelling study (Spartan; MM2/PM3) to try to
understand the dramatic effect by using the bulkiest ligand 22.3. It was shown that due
to the steric repulsion the two ligands should coordinate in the equatorial-equatorial
positions of the trigonal-bipyramidal Rh-complex.
13
Chapter 1 Introduction and Scope
1.6 Conclusions
Asymmetric hydroformylation has become a more competitive transformation for fine-
chemical industry. The development of more selective and active ligand systems,
applicable for more than a single substrate has taken a high flight. Commercially
available ligands for asymmetric hydrogenation proved to form efficient
hydroformylation catalysts under the right conditions determined by using HTE
techniques.
1.7 Perspective
There is a lot to gain by spectroscopic analysis of hydroformylation catalysts under
working conditions, which is only occasionally done so far. A better understanding of
the stereoselective step could be acquired by theoretical investigations on sufficient
high level of modelsystems of the most successful systems. Comparison of different
catalyst-systems is often difficult since experimental conditions are critical and hardly
ever identical; a comprehensive database with all available data on the asymmetric
hydroformylation reaction would prove invaluable in seeing through all quantitative
structure performance relationships.
14
Chapter 1 Introduction and Scope
1.8 Aim and Scope of Thesis
In this thesis the development of new hydroformylation catalysts is presented. Newly
synthesized ligands and their transition metal complexes were analyzed carefully via
spectroscopic means, X-ray analyses and DFT calculations. Application in asymmetric
hydrogenation and hydroformylation reactions showed the potential in catalysis. The
catalysts were monitored under working conditions to determine the coordination-
modes. The obtained data can be used for a better understanding of the reaction and the
development of new generations of catalysts.
In Chapter 2 the successful synthesis of a series of symmetrically and non-
symmetrically substituted chiral bisaminophosphine ligands following two modular
routes applying easy purification procedures is shown. X-ray analyses of both free
ligands and mononuclear cis-coordinated transition metal complexes thereof indicated
the trigonal planar geometry of the nitrogen atoms, as well as a P-N bond with double
bond character.
Chapter 3 contains DFT calculations performed on model compounds for
bisaminophosphine ligands to analyze the geometries and charge distributions. The
computed structure of a simplified cis-Pd complex of a bidentate bisaminophosphine
ligand gives valuable information on the coordination behavior. Catalysts generated in
situ from [Rh(cod)2]BF4 and bisaminophosphine ligands were used in the asymmetric
hydrogenation of methyl Z-acetylaminocinnamate with ee’s up to 91%.
The application of C2-symmetric bisaminophosphine ligands in the Rh-catalyzed
asymmetric hydroformylation of prochiral alkenes is described in Chapter 4. HP-NMR
studies indicated that equatorial - equatorial is the preferred coordination mode, which
could be confirmed by HP-IR spectroscopy.
In Chapter 5 hybrid Me-BINOLane ligands are described which form active catalysts
in the Rh-catalyzed asymmetric hydroformylation of styrene. Branched/linear ratio’s
higher than 20 were obtained. The found ee’s depend mostly on the atropisomeric
element in the phosphonite part of the ligand and reach values just over 50%.
15
Chapter 1 Introduction and Scope
1.9 References
[1] O. Roelen (to Ruhrchemie AG) German patent 849548, 1938. [2] F. Agbossou, J. -F. Carpentier, A. Mortreux, Chem. Rev. 1995, 95, 2485. [3] S. Gladiali, J. C. Bayón, C. Claver, Tetrahedron Asymmetry, 1995, 6, 1453. [4] B. Breit, W. Seiche, Synthesis, 2001, 1. [5] C. Claver, M. Diéguez, O. Pàmies, S. Castillón, Top. Organomet. Chem. 2006, 18, 35. [6] C. Claver, P. W. N. M. van Leeuwen, Rhodium Catalyzed Hydroformylation (Eds. C. Claver, P.
W. N. M. van Leeuwen), Kluwer-CMC, Dordrecht, 2000, pp 107. [7] M. L. Clarke, Curr. Org. Chem. 2005, 9, 701. [8] E. D. Daugs, W.-J. Peng, C. L. Rand (to DOW Technologies Inc.) World patent WO
2005110986 A1, 2004. [9] J. R. Briggs, J. Klosin, G. T. Whiteker, Org. Lett. 2005, 7, 4795. [10] M. C. J. Harris, M. Jackson, I. C. Lennon, J. A. Ramsden, H. Samuel, Tetrahedron Lett. 2000,
41, 3187. [11] G. Consiglio, P. Pino, Helv. Chim. Acta, 1976, 59, 642. [12] Y. Kawabata, T. M. Suzuki, I. Ogata, Chem. Lett. 1978, 4, 361. [13] P. Haelg, G. Consiglio, P. Pino, J. Organomet. Chem. 1985, 296, 281. [14] C. U. Pittmann, Y. Kawabata, L. I. Flowers, J. Chem. Soc., Chem. Commun. 1982, 473. [15] G. Parinello, R. Deschenaux, J. K. Stille, J. Org. Chem. 1986, 51, 4189. [16] G. Parinello, J. K. Stille, J. Am. Chem. Soc. 1987, 109, 7122. [17] J. K. Stille, G. Parinello (to Colorado State University Research Foundation), World Pat.
88/08835, 1988. [18] a) P. Meessen, D. Vogt, W. Keim, Organometallics, 1998, 551, 165. b) R. van Duren, Platinum
Catalyzed Hydroformylation, PhD thesis, University of Eindhoven, 2004. c) R. van Duren, L. L.
J. M. Cornelissen, J. I. van der Vlugt, J. P. J. Huijbers, A. M. Mills, A. L. Spek, C. Müller, D.
Vogt, Helv. Chim. Acta, 2006, 89, 1547. [19] L. Kollár, J. Bakos, I. Tóth, B. Heil, J. Organomet. Chem. 1988, 350, 277. [20] L. Kollár, P. Sándor, G. Szalontai, J. Mol. Cat. 1991, 67, 191. [21] C. P. Casey, S. C. Martins, M. A. Fagan, J. Am. Chem. Soc. 2004, 126, 5585. [22] J. E. Babin, G. T. Whiteker (to UCC) World Pat. 93/03839, 1993. [23] G. J. H. Buisman, E. J. Vos, P. C. J. Kamer, P. W. N. M. van Leeuwen, J. Chem. Soc. Dalton
Trans. 1995, 409. [24] G. J. H. Buisman, P. C. J. Kamer, P. W. N. M. van Leeuwen, Tetrahedron Asymmetry, 1993, 4,
1625. [25] see for example a) M. Diéguez, O. Pàmies, A. Ruiz, S. Castillón, C. Claver, Chem. Eur. J. 2003,
7, 3086; b) M. Diéguez, A. Ruiz, C. Claver, Dalton Trans. 2003, 2957. [26] G. J. H. Buisman, L. A. van der Veen, A. Klootwijk, W. G. J. de Lange, P. C. J. Kamer, P. W.
N. M. van Leeuwen, D. Vogt, Organometallics, 1997, 16, 2929. [27] C. J. Cobley, J. Klosin, C. Qin, G. T. Whiteker, Org. Lett. 2004, 6, 3277.
16
Chapter 1 Introduction and Scope [28] N. Sakai, S. Mano, K. Nozaki, H. Takaya, J. Am. Chem. Soc. 1993, 115, 7033. [29] N. Sakai, K. Nozaki, H. Takaya, J. Chem. Soc. Chem. Comm. 1994, 395. [30] T. Higashijima, N. Sakai, K. Nozaki, H. Takaya, Tetrahedron Lett. 1994, 35, 2033. [31] T. Horiuchi, T. Ohta, K. Nozaki, H.Takaya, Chem. Comm. 1996, 155. [32] T. Horiuchi, T. Ohta, E. Shirakawa, K. Nozaki, H.Takaya, J. Org. Chem. 1997, 62, 4285. [33] a) M. Petit, A. Mortreux, F. Petit, G. Buono, G. Pfeiffer, New. J. Chem. 1983, 7, 583. b) E.
Cesarotti, A. Chiesa, G. D’Alfonso, Tetrahedron Lett. 1982, 23, 2995. c) G. Pracejus, H.
Pracejus, J. Mol. Catal. 1984, 24, 227. [34] R. Ewalds, E. B. Eggeling, C. H. Alison, P. C. J. Kamer, P. W. N. M. van Leeuwen, D. Vogt,
Chem. Eur. J. 2000, 6, 1496. [35] J. J. Carbó. A Lledós, D. Vogt, C. Bo, Chem. Eur. J. 2006, 12, 1457. [36] Y. Yan, X. Zhang, J. Am. Chem. Soc. 2006, 128, 7198. [37] T. P. Clark, C. R. Landis, S. L. Freed, J. Klosin, K. A. Abboud, J. Am. Chem. Soc. 2005, 127,
5040. [38] A. T. Axtell, J. Klosin, K. A. Abboud, Organometallics, 2006, 25, 5003. [39] J. Huang, E. Bunel, A. Allgeier, J. Tedrow, T. Storz, J. Preston, T. Correll, D. Manley, T.
Soukup, R. Jensen, R. Syed, G. Moniz, R. Larsen, M. Martinelli P. J. Reider [40] S. Breeden, D. J. Cole-Hamilton, D. F. Foster, G. J. Schwarz, M. Wills, Angew. Chem. Int. Ed.
2000, 39, 4106. [41] a) I. Ogata, Y. Ikeda, Chem. Lett. 1972, 487. b) M. Tanaka, Y. Watanabe, T.-A. Mitsudo, K.
Yamamoto, Y. Takegami, Chem. Lett. 1972, 483. [42] F. Lagasse, H. B. Kagan, Chem. Pharm. Bull. 2000, 48, 315. [43] J. -I. Sakaki, W. B. Schweizer, D. Seebach, Helv. Chim. Acta, 1993, 76, 2644. [44] M. T. Reetz, H. Oka, R. Goddard, Synthesis, 2003, 1809. [44] L. Panella, Phosphoramidite Ligands and Artificial Metalloenzymes in Enantioselective
Rhodium-Catalysis, PhD thesis, Rijksuniversiteit Groningen, 2006. [46] A. T. Axtell, C. J. Cobley, J. Klosin, G. T. Whiteker, A. Zanotti-Gerosa, K. A. Abboud, Angew.
Chem. Int. Ed. 2005, 44, 5834. [47] Z. Hua, V. C. Vassar, H. Choi, I. Ojima, Proc. Natl. Acad. Sci. USA, 2004, 101, 5411.
17
Chapter 2
Synthesis of Bisaminophosphine Ligands and
Their Coordination Behavior
The versatile modular synthesis of novel symmetrically and
non-symmetrically substituted bisaminophosphine ligands is
described. Investigation of the molecular structures showed the
trigonal planar geometry of the nitrogen atoms and a
significant contribution of π-bonding to the P-N bond.
Upon complexation to late transition metals mononuclear cis-
coordination was mainly found.
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
2.1 Introduction
The coordination chemistry of diphosphine ligands with a variety of transition metals
is widely studied and several types of coordination modes have been established over
the years.[1] Numerous families of novel (chiral) ligands have been synthesized,[2] with
emphasis on cis-chelating properties to form monomeric metal complexes. Besides
ligand design based on desired behavior towards transition metal complexes and the
catalytic activity of such systems, the approach of modular design and availability of
cheap resources has gained significant importance.
Especially in asymmetric catalysis such a modular approach is highly desirable, since
full understanding of the factors governing the enantioselectivity during the catalytic
cycle is often lacking and the availability of tunable ligand families would greatly
enhance the generation of data leading to new insights. We have therefore set out to
explore new chiral diamines as chiral auxiliaries, since they form a class of hitherto
neglected ligand backbones.
N
NPPh2
PPh2
N
N
H
H
PPh2
PPh2
N PN
o-An
Ph
P
P
N
N
N
N
PhPh
i
ii
iii iv
Figure 1 Aminophosphine ligand systems based on substituted heteroatoms: Piperazine (i) and 1,2-
diaminobenzene (ii), developed by the group of Woollins3 and the ligands SEMI-ESPHOS (iii) and
ESPHOS (iv) reported by Wills et al.10
The synthesis and limited use of heteroatom substituted phosphines (Figure 1) and
their transition metal complexes has received quite some attention lately,[3-5] due to
the search for new structural diversity and catalytic activities. However, little has
appeared on the use of chiral diamines as backbone structures for phosphorus ligands,
although some reports described their application in the asymmetric hydrogenation of
20
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
various substrates.[6-9] Wills et al. have described the synthesis and application of the
so-called ESPHOS ligand system iv (Figure 1) in the rhodium-catalyzed asymmetric
hydroformylation of vinylacetate, with high enantioselectivities, although commercial
availability of the chiral diamine used is limited.[10] Generally chiral amines are
widely available nowadays due to heavy industrial investments in commercially
viable synthetic intermediates and specialty chemicals. Therefore the application of
chiral amines to build up the chiral backbone could be a viable approach.
In this chapter we report on the synthesis of novel chiral bidentate aminophosphine
ligands modularly constructed from chiral amines, both symmetrically and non-
symmetrically substituted, together with a study of their coordination chemistry
towards the transition metals palladium, platinum and rhodium.
2.2 Results
2.2.1 Symmetrically Substituted Bisaminophosphines
ClPPh2
OO
BrBr
NH
R1
R2NH
R1
R2PPh2PPh2
R1
R2N
R1
R2N
A
B
4 equiv chiral amine
2 equiv KOH (aq)
2 equiv chiral amine
2 equiv LiAlH4
NEt3
Scheme 1 Generic synthesis routes A and B to bisaminophosphine ligands.
The authors from references 6-9 chose condensation on diethyloxalate followed by
reduction of the corresponding diamide to synthesize their class of ligands. However,
in this protocol the reduction of the amide bonds is often not trivial. This method
required a case by case optimization making this route unsuitable for a parallel
approach.
We prepared the chiral bidentate aminophosphine ligands L1-L9 following two other
synthetic methodologies (scheme 1). This enables a modular construction of the
ligands in the way that there might be an alternative route if the first would fail.
21
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
Procedure A follows the two-step approach where the chiral amine moieties are
introduced in the backbone of the ligand by a nucleophilic SN2 substitution on a
dihaloalkane, generally dibromoethane, to form two secondary amines with a two-
carbon bridge spacer obtained as their hydrohalides. After treatment with a strong
base the diamine is purified by distillation. These intermediate diamines may be used
as ligands in e.g. asymmetric hydrosilylation of prochiral ketones.[11] By reaction of
the diamine with ClPPh2 in the presence of NEt3 to scavenge the liberated
hydrochloric acid, the corresponding diphosphines were obtained in good overall
yields. During the first step of this synthetic route an excess of chiral amine is used (4
equiv) but no solvent. Recovery of the precious chiral amine is possible during the
distillative workup, which is most efficiently achieved when the synthesis is
performed at larger scale (min. 50 mmol). Variation of the number of carbons in the
bridge can be achieved (Cn with n >1).
On a smaller scale, if the desired amount of ligand is lower or the chiral amine
building block is more expensive or less available, procedure B may be appropriate.
Here condensation of glyoxal with two equivalents of the appropriate chiral amine
gave the corresponding diimine in high yield. Reduction of the diimine with LiAlH4
and subsequent reaction of the amino-functionalities with ClPPh2 in the presence of
NEt3 afforded the bisaminophosphines in good yield. Washing with acetonitrile
afforded the pure compound without the need for further purification. We have to
stress that the intermediate diimines can be used as ligands themselves if applied to
e.g. semihydrogenation of alkynes.[12] A disadvantage of this second route may be the
limitation to ligands with a 2 carbon bridge. Note that both routes should enable the
introduction of other chlorophosphine compounds without much alteration.
22
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
Table 1 List of the prepared bisaminophosphines L1 – L9, the followed methodology for each ligand
and few remarks on the synthesis.
Ligand code
Ligand Structure
Procedure A / B
Remarks
L1 C2-spea-PPh2 NN
PPh2 PPh2 A
L2 C2-rpea-PPh2 NN
PPh2 PPh2 A
L3 C2-paea-PPh2 NN
PPh2 PPh2OMeMeO
A introduction phosphorus low yield
L4 C2-ppa-PPh2 NN
PPh2 PPh2 A
L5 C2-1nea-PPh2 NN
PPh2 PPh2
A diamine not purified by distillation
L6 C2-Cyea-PPh2 NN
PPh2 PPh2 B intermediate diamine very mobile oil
L7 C2-thna-PPh2 NNPPh2PPh2
B intermediate diimine not isolated
L8 C3-rpea-PPh2 N N
PPh2PPh2
A use of dibromopropane
L9 C3-ppa-PPh2 N N
PPh2PPh2
A use of dibromopropane
To expand the family of bisaminophosphine ligands to a smaller bridge size the
heteroatom bridge N-Si-N was considered, since a large variety of commercially
available dichlorosilane compounds exists, which can be converted into the
corresponding diaminosilane intermediates by a simple condensation step. The
condensation of the N-Si-N compounds with ClPPh2 imposed an unforeseen problem
due to the labile character of the Si-N bond in acidic environment. During the
attempts to synthesize N,N'-bis[(R)-(α)-methylbenzylamine)-N,N'-bis-
(diphenylphosphino)diaminodimethylsilane 2, unwanted diphosphazane 3 was
obtained in high yield with respect to ClPPh2 (scheme 2).
Workers in the group of Woollins did succeed in the synthesis of the PNSiNP
sequence, following the synthetic approach where aminophosphine units are
deprotonated and coupled to the dichlorosilane compound. Their coordination
behavior to transition metals was studied, where even exploitation of the labile N-Si-
N linkage was envisioned.[13]
23
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
NHNH
Si
NN
Si
PPh2 PPh2
NPPh2
PPh2
ClPPh2
NEt3
ClPPh2
NEt31
2
3
Scheme 2 Attempted synthesis of PNSiNP sequence 2 and obtained diazaphosphane 3.
The structures of the prepared ligands including the determination of the absolute
configuration of the stereogenic carbon atom could be confirmed by the molecular
structures of two ligands for which crystals suitable for X-ray analyses could be
grown.
Figures 2 and 3 show the structures of L6 and L7 respectively and tables 2 and 3
contain selected bond lengths, distances and angles of the ligands. Both ligands
possess C2 symmetry, the geometry around the nitrogen atoms is trigonal planar with
the sum of angles around N close to 360º and P-N distances around 1.68 Å indicating
considerable double-bond character (via a π-interaction) between the P and N atom.
P1
N1
C1
C2
N2
C4P2
Figure 2 ORTEP representation of ligand L6. Displacement ellipsoids are drawn at 50% probability
level. Hydrogen atoms are omitted for clarity.
24
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
Table 2 Selected bond lengths, distances and angles for ligand L6.
Bond lengths (Å) P1-N1 1.6793(16) P2-N2 1.6835(16) P1-P2 (dist) 6.379 N1-N2 (dist) 3.815
Angles (º) P1-N1-C1 122.97(13) P1-N1-C3 116.40(13) C1-N1-C3 120.17(15) P2-N2-C2 124.20(12) P2-N2-C4 118.66(13) C2-N2-C4 116.58(15) Sum angles N1 359.5 sum angles N2 359.4
C3P1
C1C2 N2N1
P2C13
Figure 3 ORTEP representation of ligand L7. Displacement ellipsoids are drawn at 50% probability
level. Hydrogen atoms are omitted for clarity.
Table 3 Selected bond lengths, distances and angles for ligand L7.
Bond lengths (Å) P1-N1 1.6887(21) P2-N2 1.6874(18) P1-P2 (dist) 6.407 N1-N2 (dist) 3.450
Angles (º) P1-N1-C1 123.32(14) P1-N1-C13 122.41(16) C1-N1-C13 113.41(19) P2-N2-C2 124.16(14) P2-N2-C3 119.87(14) C2-N2-C3 115.94(16) Sum angles N1 359.1 sum angles N2 360.0
Bisaminophosphine L4 was oxidized by 30% aqueous H2O2 to check if the
phenomena of a virtually flat geometry around N and a P-N bond with double bond
character would also occur after oxidation. X-ray analysis revealed the unambiguous
molecular structure of the ligand (Figure 4), and showed the slightly distorted trigonal
planar geometry of the nitrogen atoms and a more pronounced P-N double bond (1.66
Å), presumably due to the greater donation of the N lone pair to the more electron
poor phosphorus atom (an upfield shift of 12.1 ppm in 31P NMR was observed upon
25
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
oxidation compared to the non-oxidized ligand). Detailed information on bond
lengths, distances and angles for ligand oxide 4, is listed in Table 4.
P2C12
N1 O2O1 C2N2C1
P1C3
Figure 4 ORTEP representation of ligand L4 oxide 4. Displacement ellipsoids are drawn at 50%
probability level. Hydrogen atoms are omitted for clarity.
Table 4 Selected bond lengths, distances and angles for ligand L4 oxide 4.
Bond lengths (Å) P1-N1 1.661(3) P2-N2 1.652(3) P1-O1 1.477(3) P2-O2 1.477(3) P1-P2 (dist) 6.472 N1-N2 (dist) 3.787
Angles (º) O1-P1-N1 113.03(16) O2-P2-N2 111.10(17) P1-N1-C1 120.5(2) P1-N1-C12 123.0(3) C1-N1-C12 114.4(3) P2-N2-C2 121.7(2) P2-N2-C3 116.3(3) C2-N2-C3 118.6(3) sum angles N1 357.9 sum angles N2 356.6
Besides investigation of molecular structures and therewith structural features of the
prepared ligands, their electronic parameters were studied. A simple and efficient
method to evaluate the σ-donor character and hence the basicity of a phosphine
moiety is to measure the magnitude of the coupling constant 1JSe-P in the 31P NMR
spectrum of the 77Se isotopomer of the corresponding diphenylphosphine
selenide.[14,15] An increase in the coupling constant is indicative of increasing s-
character of the phosphorus lone-pair and hence of lower basicity. Ligands L1 and L9
were reacted with elemental selenium for 30 minutes in toluene at 70 °C. In the 31P
NMR spectrum of the corresponding selenides 5 and 6 in CH2Cl2, a singlet was found
at δ 70.3 ppm and 69.1 ppm, respectively. Both signals were flanked by two 77Se-
satellites, and the coupling constants JSe-P were 752 Hz and 750 Hz, respectively. This
shows that both ligands are essentially identical in their electronic character of the
26
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
phosphorus moieties. These values are in agreement with the few available literature
data on aminophosphines.[3a,4] The JSe-P values are higher than for corresponding
diarylphosphines, which implies lower basicity of the phosphine moiety, due to higher
electronegativity of the adjacent nitrogen atom.[16]
2.2.2 Non-Symmetrically Substituted Bisaminophosphines
Besides the extension of the set of modular bisaminophosphines and a study into their
coordination behavior we were curious if we could study the influence of an
individual aminophosphine moiety on the properties of the ligand. Hence we were
aiming at a versatile synthetic route which would enable the preparation of non-
symmetrically substituted chiral bisamines, which could be converted into the
corresponding bisaminophosphines by the application of a similar condensation with
chlorodiphenylphosphine as mentioned in paragraph 2.2.1.
First attempts consisted of a combination of a nucleophilic substitution and a
condensation, performed with two amines of choice on the same backbone (scheme
3).
i
ii
ClO
O
OBr
RNH2
RNH2
O
ONHR deprotection
Scheme 3 Attempted syntheses towards non-symmetrically substituted chiral diamines.
Reaction i) did not yield any identifiable products, the observed black color probably
indicating the formation of polymeric materials. Nucleophilic substitution on the
commercially available bromoalkane (ii), leaving the 1,3-dioxalane protected
aldehyde intact, nicely afforded the desired compounds applying different chiral
amines, but subsequent deprotection of the aldehyde did not occur under normally
successful mildly acidic conditions.
2-Chloroethanol however could serve as the basis of the synthesis of non-
symmetrically substituted chiral diamines (scheme 4).
27
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
NCl
H HClN
OH
HClOH
NH2
SOCl2KOH
N NR
H H
RNH2
KOH
ClPPh2
NEt3
N
PPh2
N
PPh2
R
L10 R =
L12 R =
L11 R =
7
8
9
Scheme 4 Synthetic procedure for non-symmetrically substituted bisaminophosphine ligands L10-L12.
A nucleophilic substitution in neat (R)-α-methylbenzylamine gave after KOH
treatment and recovery of the excess amine the chiral aminoalcohol 7. Conversion of
the alcohol function into the chloride leaving group by a reaction with thionylchloride,
meanwhile leaving the amine function protected as the hydrochloride, yields a
versatile building block which is stable in air for months. The introduction of the
second amine is again performed with a small excess of amine at a high temperature
to ensure homogeneity and the chiral, non-symmetrically substituted diamine can be
obtained 9. Conversion into the corresponding bisaminophosphine is done without
laborious purification of the diamine (only basic workup and removal of amine in
vacuo is required), by the reaction with chlorodiphenylphosphine in the presence of
triethylamine.
In the above described manner 3 ligands were synthesized, all based on the
intermediate aminoalcohol created from (R)-α-methylbenzylamine. The C1-symmetric
ligands L10 and L11 were obtained by respectively using achiral benzylamine and
(R)-α-ethylbenzylamine as the second amine. Curiosity driven the Cs-symmetric
ligand L12 was made by applying (S)-α-methylbenzylamine to obtain a ligand
possessing a mirror plane thus being a meso compound.
Theoretically this route would enable the synthesis of a large family of ligands, with
different steric and electronic properties on both aminophosphine moieties and
tunable stereogenic information.
28
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
2.2.3 Coordination Chemistry
The coordination behavior of the above described symmetrically substituted chiral
amine-based diphosphine ligands was studied with the representative ligand 1 towards
palladium, platinum and rhodium precursors.
Preparation of palladium(II) complexes C1 and C2.
Reaction of [PdCl2(cod)] with L1 for 2 hours in CH2Cl2 at room temperature resulted
in a yellow solid complex C1 (figure 5 (M = Pd, X = Cl)), for which the 31P NMR
spectrum showed a singlet at δ 87.3 ppm. Little structural information can be deduced
from this chemical shift however. The ligand backbone skeleton (viz. size, flexibility)
is analogous to the well studied ligand dppb, for which only the cis-isomer is
reported.[17]
NN
PPh2PPh2
MCl X
Figure 5 Prepared and studied complexes of ligand L1, C1 (M = Pd, X = Cl), C2 (M = Pd, X = Me)
and C3 (M = Pt, X = Cl).
To further elucidate the structure of complex C1, ligand L1 was reacted with
[PdCl(CH3)(cod)] to give complex C2 as a micro-crystalline yellow solid, (figure 5
(M = Pd, X = Me)), Characterization of this species in solution by 31P NMR
spectroscopy showed an AB system with two doublets at δ 91.3 ppm and 81.0 ppm
with coupling constants JP-P of 28 Hz, while in the 1H NMR spectrum a doublet of
doublets was present at δ 0.47 ppm for the methyl ligand at palladium. Both
observations clearly indicate the sole formation of cis-[PdCl(CH3)(L1)],[18] with the
downfield doublet at δ 91.3 ppm corresponding to the phosphine trans to chloride and
the upfield doublet at δ 81.0 ppm to the phosphine trans to the methyl ligand.
Furthermore, in the 1H NMR spectrum two triplets at δ 4.26 ppm and 4.41 ppm were
found for the inequivalent CH2-groups in the backbone. Both complexes C1 and C2
were surprisingly very soluble in acetonitrile, but we could obtain single crystals by
29
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
slow diffusion of hexanes into a CH2Cl2 solution of complex C1. The molecular
structure of this compound was unequivocally determined by X-ray crystallography.
Figure 6 shows the structure of C1. Table 5 contains data on selected bond lengths
and angles. The asymmetric unit cell contained two independent molecules that
differed mainly in the orientation of the phenyl rings on the phosphorus atoms. For
clarity only one residue molecule is shown.
C1 C4C2C3
N2N1
P2P1
Pd1
Cl2Cl1
Figure 6 ORTEP representation of the first of two independent molecules of complex C1, cis-
[PdCl2(L1)]. Displacement ellipsoids are drawn at 50% probability level. Hydrogen atoms are omitted
for clarity.
Table 5 Selected bond lengths, distances and angles for complex C1, cis-[PdCl2(L1)].
Bond lengths (Å) Pd1-P1 2.2681(6) Pd1-P2 2.2549(5) Pd1-Cl1 2.3507(5) Pd1-Cl2 2.3489(6) P1-N1 1.6811(18) P2-N2 1.6591(19) P1-P2 (dist) 3.381 N1-N2 (dist) 3.089
Angles (º) P1-Pd1-P2 96.74(2) Cl1-Pd1-Cl2 90.19(2) Cl1-Pd1-P1 83.36(2) Cl1-Pd1-P2 173.66(2) Cl2-Pd1-P1 173.41(2) Cl2-Pd1-P2 89.52(2) P1-N1-C1 116.56(14) P1-N1-C3 125.83(14) C1-N1-C3 113.91(17) P2-N2-C2 119.39(14) P2-N2-C4 123.21(14) C2-N1-C4 116.04(17) sum angles N1 356.3 sum angles N2 358.6
The geometry around the palladium atom in complex C1, cis-[PdCl2(L1)], is slightly
distorted square planar, with the aminophosphine moieties coordinated in a mutual
cis-fashion, in agreement with the spectroscopic data. The distortion is evident from
the bite angle P1-Pd1-P2 of 97°, which led to P1-Pd1-Cl1 and P1-Pd1-Cl2 angles of 83°
and 173°, respectively. The seven-membered chelate ring has a boat or open-envelope
conformation. Values found for the Pd-P, Pd-Cl and P-N bond lengths were within
30
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
the ranges reported for similar aminophosphine complexes.[3,19-21] The two P-N bond
lengths were 1.659 Å for P2-N2 and 1.681 Å for P1-N1. This is again a clear indication
of considerable double-bond character (via a π-interaction) between the P and N
atoms. The ethane backbone is severely twisted with a torsion angle N1-C1-C2-N2 of -
66°. Notably, the geometry around the nitrogen atoms N1 and N2 is trigonal planar,
with bond angles of between 116.0° and 125.8°. The intramolecular P1-P2 distance
was 3.38 Å while the N1-N2 distance was 3.09 Å.
Preparation of dichloroplatinum(II) complexes C3-C5.
Reaction of [PtCl2(cod)] with ligand L1 resulted in the straightforward formation of a
white solid, complex C3, (figure 5 (M = Pt, X = Cl)). Characterization of the species
present in solution by 31P NMR spectroscopy showed a single peak at δ 62.1 ppm,
flanked by 195Pt satellites and a coupling constant JPt-P of 4151 Hz. This latter value is
a clear indication of cis-coordination of the diphosphine to the platinum center,
yielding cis-[PtCl2(L1)].[22] The chemical shift is remarkably high for a diphosphine-
based cis-platinum complex and reflects the electronic influence of the amino-groups.
It is virtually similar to a silsesquioxane-based diphosphinite Pt-complex developed
previously in our group.[23] When the same reaction was performed with the more
flexible ligand L9 an off-white solid was obtained, C4, for which the 31P NMR
spectrum showed a singlet at 60.5 ppm, together with 195Pt satellites and a coupling
constant JPt-P of 4285 Hz, suggesting formation of cis-[PtCl2(L9)]. The clear
preference of these ligands to coordinate in the cis-geometry is also visible for most
crowded ligand L5, after its reaction with the metal precursor complex C5 was
obtained, the signals in the 31P NMR spectrum (δ 60.4, JPt-P = 4120 Hz) indicating that
the obtained material was indeed cis-[PtCl2(L5)].
Preparation of chlorocarbonylrhodium(I) complex C6.
Grimblot et al. have previously reported X-ray photoelectron and IR spectroscopy
studies on RhCl(CO)-complexes with various phosphine ligands, including
aminophosphines, and their correlation with the results obtained in rhodium-catalyzed
hydroformylation of 1-hexene.[24] Their initial studies on the influence of the
(amino)phosphine ligand on the CO stretching frequency in the corresponding Rh-
31
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
complex led to the conclusion that all tested ligands led to trans-complexes except for
dppe (bis(diphenylphosphino)ethane) as applied ligand.
Upon addition of ligand L1 to a CH2Cl2-solution of [RhCl(CO)2]2, the solution
immediately turned yellow. After two hours of reaction at room temperature, removal
of volatiles left a clear yellow microcrystalline solid, complex C6.
The 31P NMR spectrum for this complex showed a mixture of the cis and trans-
isomers in a ratio of 34:66 in favor of the trans-isomer. This trans-complex was
characterized by a doublet at δ 81.0 ppm, and a coupling constant of JRh-P 133 Hz,
which is a typical value for diphenylphosphine ligands. The cis-complex appeared as
a set of two doublets of doublets at δ 99.6 ppm and δ 75.3 ppm. The respective
coupling constants JRh-P were 180 Hz for the P trans to the Cl ligand and 133 Hz for
the P trans to the CO ligand, while the JP-P was 33 Hz. The related FT-IR-spectrum
showed an absorption band in the carbonyl region at νCO 1968 cm-1, which is in the
range found for complexes with σ-donor ligands at the Rh center.[25-27]
2.3 Conclusions
We showed the successful synthesis of a series of symmetrically substituted chiral
bisaminophosphine ligands following two modular routes applying easy purification
procedures. X-ray analyses indicated the trigonal planar geometry of the nitrogen
atoms, as well as a P-N bond with double bond character. A generally applicable
reaction scheme was designed and used for the construction of non-symmetrically
substituted chiral bisaminophosphine ligands. Coordination towards palladium,
platinum and rhodium precursors showed a preference for mononuclear cis-
coordination.
2.4 Perspective
More commercially available chiral amines could be incorporated in the
bisaminophosphine ligands, both the symmetrically as non-symmetrically substituted.
To increase rigidity the backbone could be derived from 1,2-dibromocyclohexane and
32
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
the sequential direct amination of 1,2-dibromobenzene with several amines would
open up a whole new class of fine-tunable ligands after condensation with
chlorophosphines (figure 7).
N
N
PPh2
PPh2
N
NPPh2
PPh2
Figure 7 Envisioned more rigid bisaminophosphine ligand classes.
2.5 Acknowledgements
Part of this work has been published, Eric J. Zijp, Jarl Ivar van der Vlugt, Duncan M.
Tooke, Anthony L. Spek and Dieter Vogt, Dalton Transactions, 2005, 512-517.
Avantium Technologies, National Research School Combination for Catalysis and the
Netherlands Organization for Scientific Research (NWO) are kindly acknowledged
for financial support. BASF A.G. is thanked for a kind gift of chiral amines and
Umicor Co. for a loan of precious metals. Bart van As and Roser Bartra Vallverdu
(Erasmus Exchange Program) supported the experimental work in this chapter during
their graduation work and research stage respectively, for which we are indebted.
2.6 Experimental Section
All manipulations were carried out under argon using standard Schlenk techniques.
Chemicals were purchased from Acros, Aldrich, Lancaster or VWR and used as
received or distilled from CaH2 before use. (R)-α-ethylbenzylamine, (R)-1-(4-
methoxyphenyl)ethylamine and (S)-(1-(1-naphthyl))-ethylamine were received as a
kind gift from BASF AG. Solvents were either taken HPLC-grade from an argon-
flushed column, packed with aluminum oxide, or distilled under argon prior to use
over an appropriate drying agent. NMR spectra were recorded at room temperature on
a Varian Mercury 400 MHz spectrometer. Chemical shifts are given in ppm and
spectra are referenced to CDCl3 (1H, 13C{1H}) or 85% H3PO4 (31P{1H}). FT-IR
33
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
34
spectra were taken on an AVATAR E.S.P. 360 FTIR spectrometer. PdCl2(cod),[28]
PdCl(CH3)(cod),[29] and PtCl2(cod)[30] were prepared according to literature
procedures.
N,N'-bis[(S)-α-methylbenzyl]-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine (L1):
N,N
N,N'-bis[(R)-α-methylbenzyl]-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine (L2):
,2-
diamine
'-bis[(S)-α-methylbenzyl]-ethane-1,2-diamine was
prepared by following a procedure by Mimoun et
al.[31] To a flask containing 4 equivalents of (S)-α-
methylbenzylamine (46.7 g, 385 mmol) at 130 °C was added dropwise 1,2-
dibromoethane (17.0 g, 91 mmol). After 1 h at elevated temperature the mixture was
cooled to 80 °C and a 4M aqueous solution of KOH (28.4 g, 506 mmol) was added.
After extraction of the mixture with ethylacetate and concentration, the mixture was
fractionally distilled at 2.5 mbar and N,N'-bis[(S)-α-methylbenzyl]-ethane-1,2-
diamine (18.3 g, 68 mmol) was obtained at 155 °C as a colorless oil in 75% yield.
Spectral properties were similar as described in literature.
L1 was prepared by adding N,N'-bis[(S)-α-methylbenzyl]-ethane-1,2-diamine (2.51 g,
9.45 mmol) dropwise to a mixture of triethylamine (2.5 g, 25 mmol) and
chlorodiphenylphosphine (4.22 g, 19.1 mmol) in diethylether. The produced
ammoniumsalts are removed from the suspension by filtration and the mixture is
concentrated, forming a white solid. After stripping with hexanes and recrystallization
from hot acetonitrile the analytically pure compound L1 (4.21 g, 6.61 mmol) was
obtained as a white semi-crystalline solid in 70% yield. Spectral properties were
similar as described in literature.[6c]
N,N'-bis[(R)-α-methylbenzyl]-ethane-1 was prepared by following the procedure
described for L1. Starting from R-α-
methylbenzylamine (43.2 g, 356 mmol) and 1,2-dibromoethane (16.7 g, 89 mmol)
N,N'-bis[(R)-α-methylbenzyl]-ethane-1,2-diamine (18.9 g, 70 mmol) was obtained at
155 °C as a colorless oil in 79% yield. Spectral properties were similar as described in
literature.
NN
PPh2 PPh2
NN
PPh2 PPh2
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
35
Compound L2 was prepared by following the procedure described for L1 starting
from N,N'-bis[(R)-α-methylbenzyl]-ethane-1,2-diamine (2.72 g, 10.1 mmol),
triethylamine (2.5 g, 25 mmol) and chlorodiphenylphosphine (4.47 g, 20.3 mmol). L2
(4.78 g, 7.51 mmol) was obtained as a white semi-crystalline solid in 74% yield.
Spectral properties were similar as described in literature.[7]
N,N'-bis[(R)-1-(4-methoxyphenyl)-ethylamine]-N,N'-bis-(diphenylphosphino)-ethane-
1,2-diamine (L3):
Following the procedure described for L1,
starting from (R)-1-(4-methoxy
phenyl)ethylamine (55.1 g, 364 mmol) and
1,2-dibromoethane (16.91 g, 90 mmol)
N,N'-bis[(R)-1-(4-methoxyphenyl)-ethyl]-ethane-1,2-diamine (28.14 g, 86 mmol) was
obtained in 95% yield after removal of all volatiles.
Starting from N,N'-bis[(R)-1-(4-methoxyphenyl)-ethyl]-ethane-1,2-diamine (3.14 g,
9.6 mmol), chlorodiphenylphosphine (4.29 g, 19.4 mmol) and triethylamine (2.5 g, 25
mmol), L3 (1.14 g, 1.67 mmol) was obtained as a white solid in 17% yield with
similar spectral properties as described elsewhere.[32]
N,N'-bis[(R)-α-ethylbenzyl]-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine (L4):
arting
ompound L4 was prepared by following the procedure described for L1 starting
Following the procedure described for L1, st
from (R)-α-ethylbenzylamine (53.49 g, 441 mmol)
and 1,2-dibromoethane (18.97 g, 101 mmol).
Distillation at 1.6 mbar at 174 °C afforded N,N'-bis[(R)-α-ethylbenzyl]-ethane-1,2-
diamine (19.31 g, 71.9 mmol) in 71% yield.
C
from N,N'-bis[(R)-α-ethylbenzyl]-ethane-1,2-diamine (2.91 g, 10.8 mmol),
triethylamine (2.5 g, 25 mmol) and chlorodiphenylphosphine (4.85 g, 22.0 mmol). L4
(5.78 g, 9.09 mmol) was obtained as a white semi-crystalline solid in 84% yield.
NN
PPh2 PPh2
NN
PPh2 PPh2OMeMeO
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
36
NN
PPh2 PPh2
1H NMR (CDCl3) δ 6.8-7.4 (m, 30H, Ph), 3.43 (dt, 2H, CHCH2CH3, 3JP-H = 16.8 Hz,
, Ph), 140.5, 139.3 (2d, Ph, 1JP-C = 67 Hz), 132.3, 131.5
,N'-bis[(S)-(1-(1-naphthyl)-ethyl)]-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine
N'-bis[(S)-(1-(1-naphthyl)-ethyl)]-ethane-1,2-
dia
the diamine was obtained in 95% y
(CDCl3) δ 8.15 (m, 2H, Ph), 7.84 (m, 2H, Ph), 7.78 (d, 2H, 3JH-H = 8.1 Hz),
MR (CDCl3) δ 141.5, 134.0, 131.6, 129.2, 127.4, 126.0, 125.9, 125.6, 124.0,
ollowing the (modified) procedure for L1, starting from N,N'-bis[(S)-(1-(1-
3JH-H = 7.3 Hz), 2.63 (m, 4H, CH2N), 1.94 (m, 4H, CHCH2CH3), 0.73 (t, 6H,
CHCH2CH3, 3JH-H = 7.3 Hz). 13C NMR (CDCl3) δ 143.6 (s
(2d, Ph, 2JP-C = 21 Hz), 128.2, 128.1, 128.0, 128.0, 127.8, 127.7, 126.8 (7s, Ph), 67.1
(d, CHCH2CH3, 2JP-C = 24 Hz), 50.3 (d, CH2N, 2JP-C = 8 Hz), 28.5 (d, CHCH2CH3, 2JP-C = 18 Hz), 12.0 (CHCH2CH3). 31P NMR (CDCl3) δ 45.6.
N
(L5):
N,
mine was prepared following the procedure for
L1, starting from (S)-(1-(1-naphthyl))-ethylamine
(15.1 g, 88 mmol) and 1,2-dibromoethane (8.3 g,
44 mmol). After removal of the access of amine
ield as a brownish syrup (15.47 g, 42 mmol) which
solidified upon standing.
1H NMR
7.67 (d, 2H, Ph, 3JH-H = 7.5 Hz), 7.45-7.55 (m, 8H, Ph), 4.53 (q, 2H, CHCH3, 3JH-H =
6.6 Hz), 2.68 (s, 4H, CH2CH2), 2.02 (bs, 2H, NH), 1.53 (d, 6H, CHCH3, 3JH-H = 6.6
Hz). 13C N
123.2 (10s, Ph), 53.9 (CHCH3), 47.7 (CH2NH), 24.0 (CHCH3).
F
naphthyl)ethyl)]-ethane-1,2-diamine (1.95 g, 5.3 mmol) in 20 mL toluene,
chlorodiphenylphosphine (2.38 g, 10.8 mmol) and 2 mL triethylamine (1.45 g, 14.3
mmol) in 30 mL toluene and a reaction overnight at 80 °C, L5 was obtained as a
white powder (2.33 g, 3.2 mmol) in 60% yield.
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
37
1H NMR (CDCl3) δ 7.85 (m, 4H, Ph), 7.65 (m, 2H, Ph), 7.47 (m, 2H, Ph), 7.36 (m,
2H, Ph), 7.08-7.30 (m, 20 H, Ph), 6.96 (m, 4H, Ph), 4.72 (dq, 2H, CHCH3, 3JP-H = 13
Hz; 3JH-H = 6.6 Hz), 2.69 (s, 4H, CH2CH2), 1.42 (d, 6H, CHCH3, 3JH-H = 6.6 Hz). 13C NMR (CDCl3) δ 140.4 (s, Ph), 140.4, 139.6 (dd, Ph, 1JP-C = 85 Hz), 134.1, 132.4
(dd, Ph, 1JP-C = 85 Hz), 132.4 (d, Ph, 2JP-C = 40 Hz), 131.6, 128.9, 128.4 (3s, Ph),
128.1, 128.0 (2d, Ph, 3JP-C = 6 Hz), 127.8, 125.8, 125.5, 125.3, 124.6, 56.6 (d,
CHCH3, 2JP-C = 31 Hz), 49.9 (d, CH2N, 2JP-C = 9 Hz), 21.0 (d, CHCH3, 3JP-C = 10 Hz). 31P NMR (CDCl3) δ 55.9.
N,N'-bis[(S)-(1-cyclohexylethyl)]-N,N ine (L6): 4] a
o a solution of the diimine (1.79 g, 6.5 mmol) in 25 mL diethylether was added in
(CDCl3) δ 2.83 (dq, 2H, CHCH3, 3JH-H = 5.4 Hz), 2.64 (m, 2H, CH2NH),
3), 47.1 (CH2NH), 42.8, 29.9, 28.2, 26.8, 26.6, 26.5
re described for L1 starting from N,N'-bis[(S)-(1-
'-bis-(diphenylphosphino)-ethane-1,2-diam
Following a procedure of Weber et al. [3
solution of (S)-(1-cyclohexyl)ethylamine (12.1 g,
95 mmol) in 50 mL hexanes was added to a 40
wt% aqueous solution of glyoxal (47 mmol). After 30 minutes of reaction the phases
were separated and the water layer was extracted with hexanes. After drying over
MgSO4, concentration in vacuo afforded the corresponding diimine as a white powder
(12.3 g, 44 mmol, 95%) with similar spectral properties as described in literature and
which was used without any purification.
T
portions LiAlH4 (1.05 g, 27.7 mmol). After one hour the excess of LiAlH4 was
neutralized by carefully adding water. After drying over MgSO4, extraction with
diethylether and concentration the corresponding diamine was obtained in quantitative
yield as a colorless mobile oil (1.82 g, 6.5 mmol).
1H NMR
2.46 (m, 2H, CH2NH), 2.04 (bs, 2H, NH), 1.73 (m, 10H, Cy), 1.20 (m, 12H, Cy), 1.06
(d, 6H, CHCH3, 3JH-H = 5.4 Hz). 13C NMR (CDCl3) δ 58.0 (CHCH
(6s, Cy), 16.9 (CHCH3).
Following the procedu
cyclohexylethyl)]-ethane-1,2-diamine (1.03 g, 3.67 mmol), chlorodiphenylphosphine
(1.65 g, 7.48 mmol) and triethylamine (1.5 mL, 10.7 mmol) the analytically pure
NN
PPh2 PPh2
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
38
compound L6 (1.98 g, 3.05 mmol) was obtained as a white semi-crystalline solid in
83% yield.
1H NMR (CDCl3) δ 7.12-7.22 (10H, PPh2), 2.64 (dm, 4H, CH2N, 3JP-H = 36.2 Hz),
.0 (2d, Ph, 2JP-C
2: C, 77.75; H, 8.39; N, 4.32. Found: C, 77.53; H, 8.53; N,
,N'-bis[(R)-(1,2,3,4-tetrahydro-1-naphthylamine)]-N,N'-bis-(diphenylphosphino)-
ed procedure described for L6
fi
(CDCl3) δ 7.3-7.2 (m, 28H, Ph), 4.04 (m, 2H, CH), 2.81 (m, 4H), 2.64 (m,
NNPPh2PPh2
2.23 (m, 2H, CHCH3), 1.89 (m, 2H, Cy), 1.68 (m, 8H, Cy), 1.50 (m, 2H, Cy), 1.17
(m, 6H, Cy), 1.06 (d, 6H, CHCH3, 3JH-H = 6.3 Hz), 0.75 (m, 4H, Cy). 13C NMR (CDCl3) δ 140.7, 140.5 (2d, Ph, 1JP-C = 48 Hz), 132.4, 132
= 21 Hz), 128.0 (d, Ph, 4JP-C = 4 Hz), 127.9 (d, Ph, 3JP-C = 6 Hz), 61.8 (d, CHCH3, 2JP-
C = 26 Hz), 49.3 (d, CH2N, 2JP-C = 26 Hz), 42.8 (d, CHCHCH3, 3JP-C = 10 Hz), 31.2,
30.0, 26.4, 26.2, 26.0 (5s, Cy), 19.2 (d, CHCH3, 3JP-C = 11 Hz). 31P NMR (CDCl3) δ 46.0.
Anal. Calcd. for C42H54N2P
4.24.
N
ethane-1,2-diamine (7)
Following the modifi
compound L7 was prepared starting from (R)-1,2,3,4-
tetrahydro-1-naphthylamine (9.1 g, 62 mmol), aqueous
40% glyoxal (3.5 mL, 31 mmol) the intermediate
ltration and washing as an off-white solid in 87% yield
(8.5 g, 27 mmol) which was used without purification. Lithiumaluminiumhydride (2.2
g, 58 mmol) was added in portions to a solution of diimine (8.2 g, 26 mmol) in
diethylether. After neutralization with water and extraction with diethylether the
intermediate crude diamine was obtained. Chlorodiphenylphosphine (2.3 mL, 12.8
mmol) was added to a solution of the diamine (2.02 g, 6.3 mmol) in diethylether
containing triethylamine (2 mL, 14.3 mmol). Usual workup afforded L7 in 78% yield
(3.38 g, 4.9 mmol). Clear rectangular crystals suitable for X-ray analysis were
prepared by crystallization from dichloromethane/hexanes (1/3).
diimine was obtained after
1H NMR
4H) 1.76 (m, 4H), 1.56 (m, 4H).
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
13C NMR (CDCl3) δ 140.2, 139.9, 139.3, 137.9, 132.1 (d, Ph, 3JP-H = 10 Hz), 128.7,
128.2, 127.9 (d, Ph, 4JP-H = 5.7 Hz), 126.3, 125.3, 60.5 (d, CHN, 2JP-C = 21 Hz), 49.2
(d, CH2N, 2JP-C = 8 Hz), 30.3, 29.5, 21.6. 31P NMR (CDCl3) δ 55.6.
N,N'-bis[(R)-α-methylbenzyl]-N,N'-bis-(diphenylphosphino)-propane-1,3-diamine (L8)
N,N'-bis[(R)-α-methylbenzyl]-propane-1,3-
diamine was prepared by following the
procedure described for L1. Starting from R-
α-methylbenzylamine (57.14 g, 423 mmol) and 1,3-dibromopropane (18.17 g, 97
mmol) N,N'-bis[(R)-α-methylbenzyl]-propane-1,3-diamine (19.06 g, 64 mmol) was
obtained at 151 °C (2.2 mbar) as a colorless oil in 66% yield.[33]
N N
PPh2PPh2
1H NMR (CDCl3) δ 7.3-7.2 (m, 10H, Ph), 3.72 (q, 2H, CHCH3, 3JH-H = 6.6 Hz), 2.51
(m, 4H, CH2CH2CH2), 1.61 (m, 2H, CH2CH2CH2), 1.54 (bs, 2H. NH), 1.34 (d, 6H,
CHCH3, 3JH-H = 6.6 Hz). 13C NMR (CDCl3) δ 146.1, 128.6, 127.0, 126.8 (4s, Ph), 58.7 (CHCH3), 47.7
(CH2NH), 30.6 (CH2CH2CH2) 24.0 (CHCH3).
Compound L8 was prepared by following the procedure described for L1 starting
from N,N'-bis[(R)-α-methylbenzyl]-propane-1,3-diamine (2.53 g, 8.96 mmol),
triethylamine (2.5 g, 25 mmol) and chlorodiphenylphosphine (3.84 g, 17.5 mmol). L8
(3.28 g, 5.04 mmol) was obtained as a white semi-crystalline solid in 56% yield.
1H NMR (CDCl3) δ 7.3-7.2 (m, 16H, Ph), 7.04 (m, 4H, Ph), 3.93 (dq, 2H, CH), 2.41
(m, 4H, CH2CH2CH2), 1.51 (m, 6H, CHCH3), 0.82 (m, 2H, CH2CH2CH2).
13C NMR (CDCl3) δ 144.8, 140.0 (dd, Ph, 2JP-C = 14 Hz), 132.5 (dd, Ph, 2JP-C = 20
Hz), 128.2, 128.1, 127.9, 127.8, 127.8, 127.4 (d, Ph, 3JP-C = 2 Hz), 126.8, 59.1 (d,
CHCH3, 2JP-C = 26 Hz), 47.1 (d, CH2N, 2JP-C = 8 Hz), 29.2 (CH2CH2CH2), 21.8 (d,
CH2N, 2JP-C = 18 Hz, CHCH3). 31P NMR (CDCl3) δ 48.0.
39
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
N,N'-bis[(R)-α-ethylbenzyl]-N,N'-bis-(diphenylphosphino)-propane-1,3-diamine (L9):
Following the procedure described for L1,
starting from (R)-α-ethylbenzylamine (49.07
g, 363 mmol) and 1,3-dibromopropane (17.35
g, 86 mmol). Distillation at 1.5 mbar at 165
°C afforded N,N'-bis[(R)-α-ethylbenzyl]-propane-1,3-diamine (16.33 g, 53 mmol) in
61% yield.[31]
N N
PPh2PPh2
1H NMR (CDCl3) δ 7.2-7.4 (10H, Ph), 3.43 (t, 2H, CHCH2CH3, 3JH-H = 6.2 Hz), 2.49
(m, 4H, CH2NH), 1.55-1.80 (m, 8H, CH2CH2CH2; CHCH2CH3; NH), 0.79 (t, 6H,
CHCH2CH3, 3JH-H = 7.3 Hz). 13C NMR (CDCl3) δ 144.4, 128.5, 127.5, 127.1 (4s, Ph), 65.5 (CHCH2CH3), 46.8
(CH2NH), 31.2 (CHCH2CH3), 30.4 (CH2CH2NH), 11.1 (CHCH2CH3).
Starting from N,N'-bis[(R)-α-ethylbenzyl]-propane-1,3-diamine (2.64 g, 8.5 mmol),
chlorodiphenylphosphine (3.83 g, 17.4 mmol) and triethylamine (2.4 g, 24 mmol),
compound L9 (5.14 g, 7.57 mmol) was obtained as a white solid in 89% yield.
1H NMR (CDCl3) δ 7.40-7.34 (m, 10H, Ph), 7.28-7.14 (m, 16H, Ph), 6.91-6.85 (m,
4H, Ph), 3.43 (dt, 2H, CHCH2CH3, 3JP-H = 16.1 Hz, 3JH-H = 6.2 Hz), 2.40 (m, 4H,
CH2N), 2.20 (m, 4H, CHCH2CH3), 0.81 (t, 6H, CHCH2CH3, 3JH-H = 7.7 Hz), 0.72 (m,
2H, CH2CH2CH2). 13C NMR (CDCl3) δ 143.4 (s, Ph), 140.1, 139.9 (2d, Ph, 1JP-C = 67 Hz), 132.6, 131.5
(2d, Ph, 2JP-C = 21 Hz), 128.3, 128.0, 128.0, 127.7, 127.6, 127.5, 126.8 (7s, Ph), 66.3
(d, CHCH2CH3, 2JP-C = 26 Hz), 47.2 (d, CH2N, 2JP-C = 11 Hz), 28.5 (CHCH2CH3),
28.2 (CH2CH2CH2), 11.8 (CHCH2CH3). 31P NMR (CDCl3) δ 45.8.
Anal. Calcd. for C45H48N2P2: C, 79.62; H, 7.13; N, 4.13. Found: C, 79.30; H, 7.09; N,
4.03.
40
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
41
N,N'-bis[(R)-(α)-methylbenzylamine)-diaminodimethylsilane (1)
(R)-α-methylbenzylamine (18.8 g, 155 mmol) was
added to a solution of dichlorodimethylsilane (9.43
mL, 78 mmol) and triethylamine (21.5 mL, 155
mmol) in diethylether. After removal of the ammoniumsalts by filtration,
concentration and distillation at 180°C and 1.5 mbar 1 was obtained as a clear liquid
in 89 % yield (20.6 g, 69 mmol).
1H NMR (CDCl3) δ 7.16-7.30 (m, 10H, Ph), 4.08 (m, 2H, CHCH3, 3JH-H = 6.6 Hz),
1.35 (d, 6H, CHCH3, 3JH-H = 6.6 Hz), 0.96 (bm, 2H, NH), -0.08 (s, 6H, (Si(CH3)2), 1JSi-H = 3.3 Hz). 13C NMR (CDCl3) δ 149.6, 128.3, 126.3, 126.0 (4s, Ph), 51.0 (CHCH3), 28.1
(CHCH3), -0.4 (Si(CH3)2).
Attempted synthesis of N,N'-bis[(R)-(α)-methylbenzylamine)-N,N'-bis-
(diphenylphosphino)diaminodimethylsilane (2)
Following the procedure of L1 starting from N,N'-
zylamine)-diaminodimethylsilane
1 (1.77 g, 5.9 mmol), chlorodiphenylphosphine (2.65,
12.0 mmol) and triethylamine (2.5 mL, 18.0 mmol) in
diethylether (25 mL). No immediate reaction occurred, after 48 h at reflux
temperature ammoniumsalts were filtered off and crude 31P NMR showed full
conversion of the chlorodiphenylphosphine. After workup and recrystallization from
boiling acetonitrile diazaphosphane N,N-bis-(diphenylphosphino)-(R)-(α)-
methylbenzylamine 3 was obtained in 39% yield (calculated from N,N'-bis[(R)-(α)-
methylbenzylamine)-diaminodimethylsilane 1) as a white crystalline powder (1.13 g,
2.31 mmol) with similar spectral properties as described in literature.[33]
bis[(R)-(α)-methylben
,N'-bis[(R)-α-ethylbenzyl]-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine oxide (4): N
L4 (100 mg, 151 μmol) was dissolved in 5 mL THF
and an excess of 30% H2O2 (aq) was added. The
mixture was vigorously stirred for 16 h at room
temperature. After removal of the THF in vacuo the
NN
PPh2 PPh2O O
NHNH
Si
NN
Si
PPh2 PPh2
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
42
aqueous mixture was extracted with CH2Cl2 and after drying over MgSO4 4 was
obtained as a white solid. Yield: 89% (94.1 mg, 134 μmol).
31P NMR (CDCl3) δ 33.5.
,N'-bis[(S)-α-methylbenzyl]-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine
L1 (85.9 mg, 135 μmol) was dissolved in 5 mL
).
.33; P, 3.53. Found: C, 63.55; H, 5.41;
,N'-bis[(R)-α-ethylbenzyl]-N,N'-bis-(diphenylphosphino)-propane-1,3-diamine selenide
Following the same procedure as for compound
P NMR (CH2Cl2) δ 69.1 (s, JSe-P = 750 Hz).
.78; P, 3.35. Found: C, 64.74; H, 5.82;
-((R)-α-methylbenzyl)-2-aminoethanol (7)
g, 81 mmol) was added dropwise to
c
NOH
H
N
selenide (5):
toluene and excess black selenium was added. The
reaction mixture was stirred for 30 minutes at 70 °C.
Filtration to remove unreacted selenium by cannula
the filtrate to dryness, leaving 5 as a white solid.
Yield: 95% (101.8 mg, 128.2 μmol). 31P NMR (CH2Cl2) δ 70.3 (s, JSe-P = 752 Hz
was followed by evaporation of
Anal. Calcd. for C42H42N2P2Se2: C, 63.48; H, 5
P, 3.58.
N
(6):
5, ligand L9 (92.1 mg, 136 μmol) was converted
to selenide 6 in a yield of 98% (111.3 mg, 133.0
μmol).
31
Anal. Calcd. for C45H48N2P2Se2: C, 64.59; H, 5
P, 3.38.
N
2-Chloroethanol (6.5
(R)-α-methylbenzylamine (24.6 g, 203 mmol) at 100°C.
After cooling to 80°C aqueous 4M KOH was added and the
ted with CH2Cl2. After washing with water the excess (R)-α-organic phase was extra
NN
PPh2 PPh2Se Se
N N
PPh2PPh2Se Se
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
43
methylbenzylamine could be recovered by distillation and 7 was obtained as an clear
oil in 76% yield (10.16 g, 61 mmol).
1H NMR (CDCl3) δ 7.2-7.4 (m, 5H, Ph), 3.77 (q, 1H, CHCH3, 3JH-H = 6.6 Hz), 3.58
l3) δ 145.3, 128.5, 127.0, 126.5 (4s, Ph), 61.2 (CH2OH), 58.1
-(2-chloroethyl)-(R)-α-methylbenzylamine hydrochloride (8)
ol 7 (3.3 g, 20 mmol) in
precipitated, after filtrat
(D2O) δ 7.37 (m, 5H, Ph), 4.37 (q, 1H, CHCH3, 3JH-H = 6.9 Hz), 3.66 (m, 3
CH3),
-(benzyl)-N’-((R)-α-methylbenzyl)-N,N'-bis-(diphenylphosphino)-ethane-1,2-diamine
To N-(2-chloroethyl)-α-methylbenzylamine hydro
ents benzylamine (3 mL, 27 mmol) and the
nt
1 starting from crude N-(benzyl)-N’-(R)-α-
ethylbenzyl-ethane-1,2-diamine (0.76 g, 2.99 mmol), chlorodiphenylphosphine
(m, 2H, CH2OH) 2.62 (m, 2H, CH2N), 2.28 (bs, 2H, NH / OH), 1.37 (d, 3H, CHCH3, 3JH-H = 6.6 Hz). 13C NMR (CDC
(CHCH3), 49.0 (NHCH2), 24.1 (CHCH3).
N
N-((R)-1-methylbenzyl)-2-aminoethan
50 mL chloroform was reacted with 3 equivalents thionyl
chloride (7.5 g, 63 mmol) and the mixture was refluxed for 2
h. After cooling to room temperature a white solid
ion and washing with diethylether 8 was obtained as a white
solid in 85% yield (3.7 g, 17 mmol). 1H NMR
2H, CH2Cl), 3.00-3.15 (m, 2H, NHCH2), 1.57 (d, 3H, CHCH3, JH-H = 6.9 Hz). 13C NMR (D2O) δ 137.7, 132.2, 131.8, 130.0 (4s, Ph), 61.0 (NHCH2), 49.3 (CH
41.5 (CH2Cl), 20.6 (CHCH3).
N
(L10)
chloride 8 (1.80 g, 8.2 mmol) was added 3
equival
mixture was heated to 175 °C u il it became homogeneous. After treatment with
aqueous 45% KOH and removal of the excess of benzylamine in vacuo the crude
diamine was obtained in 68% yield (1.42 g, 5.6 mmol) which was used as such.
Following the procedure for L
m
NCl
H HCl
N
PPh2
N
PPh2
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
44
(1.41 g, 6.39 mmol) and triethylamine (1 mL, 7.5 mmol) L10 was obtained as a white
solid in 89% yield (1.66 g, 2.66 mmol).
1H NMR (CDCl3) δ 6.95-7.4 (m, 30H, Ph), 4.01 (dq, 1H, CHCH3, 3JH-H = 7.0 Hz),
.90 (dd, 2H, PhCH2N, 3JP-H = 7.7 Hz, 2JH-H = 1.9 Hz), 2.6-2.9 (m, 4H, CH2CH2), 1.57
lphosphino)-ethane-
,2-diamine (L11)
l-
benzylamine hydrochloride 8 (1.2 g, 5.5 mmol) in
6
and additional (R)-α-ethylbenzyl
4.3 mmol), chlorodiphenylphosphine
), 3.48 (dq, 1H, CHCH3, 3JH-H = 7.0 Hz), 3.35
t, 1H, CHCH2CH3, 3JH-H = 7.3 Hz), 2.4-2.8 (m, 4H, CH2CH2), 1.95 (m, 2H,
o)-ethane-
-
chloroethyl)-(R)-α-methylbenzylamine hydrochloride
N
PPh2PPh2
3
(d, 3H, CHCH3, 3JH-H = 7.0 Hz). 31P NMR (CDCl3) δ 64.6 (benzyl), 48.9 ((R)-α-methylbenzyl)
N-((R)-α-ethylbenzyl)-N’-((R)-α-methylbenzyl)-N,N'-bis-(dipheny
1
To a solution of N-(2-chloroethyl)-(R)-α-methy
N
20 mL DME at 80 °C was added 3 equivalents (R)-
mmol). The precipitated white solid was filtered off
amine (4 mL, 26 mmol) was added. After treatment
with aqueous 45% KOH, extraction with dichloromethane and removal of the excess
of (R)-α-ethylbenzylamine in vacuo the crude diamine was obtained in 79% yield
(1.22 g, 4.3 mmol) which was used as such.[35]
Following the procedure for L1 starting from crude N-((R)-α-methylbenzyl)-N’-(R)-
α-methylbenzyl-ethane-1,2-diamine (1.22 g,
α-ethylbenzylamine (1.84 g, 13.
(1.89 g, 8.6 mmol) and triethylamine (1 mL, 7.5 mmol) L11 was obtained as a white
solid in 69% yield (1.93 g, 2.97 mmol).
1H NMR (CDCl3) δ 6.8-7.4 (m, 30H, Ph
(d
CHCH2CH3), 1.35 (d, 3H, CHCH3, 3JH-H = 7.0 Hz), 0.70 (t, 3H, CHCH2CH3, 3JH-H =
7.3 Hz). 31P NMR (CDCl3) δ 47.8 ((R)-α-methylbenzyl), 45.1 ((R)-α-ethylbenzyl)
N-((R)-α-methylbenzyl)-N’-((S)-α-methylbenzyl)-N,N'-bis-(diphenylphosphin
1,2-diamine (L12)
Following the procedure for L10, starting from N-(2NN
PPh2PPh2
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
45
(1.59 g, 7.30 mmol) and (S)-α
reaction of the crude intermediate diam
chlorodiphenylphosphine (1.58
PdCl2(cod) (32.9 mg, 115.2 μmol) and L1 (73.4 mg,
115.3 μmol) were dissolved in 5 mL CH2Cl2 and
ours at ambient temperature.
w
to leave complex C1 as a pure
Layering with CH2Cl2/CH3CN un
7.55 (t, 1H, Ph, 3JH-H = 6.8 Hz), 7.49 (d, 1H, Ph,
H-H = 5.6 Hz), 7.41 (t, 6H, Ph, 3JH-H = 7.2 Hz), 7.30 (d, 6H, Ph, 3JH-H = 7.6 Hz), 6.99
Following the same procedure as for complex C1,
but starting from PdCl(CH3)(cod) (10.5 mg, 39.6
μmol) and L1 (27.2 mg, 42.7 μmol), complex C2
NN
PPh2PPh2
PdCl Cl
-methylbenzylamine (3.80 g, 31.3 mmol) gave after
ine (0.97 g, 3.59 mmol) with
g, 7.20 mmol) in the presence of triethylamine (1.82
g, 17.9 mmol) L12 as a white solid in 67% yield (1.53 g, 2.41 mmol) with the same
spectral properties as the intermediate diamine for L1.
cis-[PdCl2(L1)] C1
stirred for 12 h
Solvents were then evaporated in vacuo. After that
ere removed by stripping 2 times with 5 mL CH2Cl2
yellow solid in 96% yield (90.1 mg, 110.6 μmol).
der slight argon flow gave yellow rectangular single
crystals, suitable for X-ray analysis.
1H NMR (CDCl3) δ 7.99 (pq, 2H, Ph, 4JH-H = 4.0 Hz, 3JH-H = 11.2 Hz), 7.70 (pq, 2H,
Ph, 4JH-H = 4.0 Hz, 3JH-H = 11.2 Hz),
the remaining traces of solvent
3J
(t, 4H, Ph, 3JH-H = 7.2 Hz), 6.90 (dd, 4H, Ph, 3JH-H = 7.6 Hz, 4JH-H = 1.2 Hz), 4.27 (m,
2H, CH), 3.59 (pq, 2H, CH2, 3JH-H = 6.8 Hz, 3JH-H = 11.2 Hz), 3.05 (pq, 2H, CH2, 4JH-
H = 4.0 Hz, 3JH-H = 11.2 Hz), 0.82 (d, 6H, CH3, 3JH-H = 6.8 Hz). 31P NMR (CDCl3) δ 87.3 (s).
Anal. Calcd. for C42H42Cl2N2P2Pd: C, 61.97; H, 5.20; N, 3.44. Found: C, 62.03; H,
5.24; N, 3.48.
cis-[PdCl(CH3)(L1)] C2
was obtained as a pure yellow solid. Yield: 94%
(29.6 mg, 37.2 μmol).
NN
PdCl Me
PPh2PPh2
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
46
1H NMR (CDCl3) δ 7.87 (dt, 2H, Ph, 3JH-H = 8.8 Hz, 4JH-H = 1.6 Hz), 7.83 (m, 2H,
h), 7.63 (ddd, 2H, Ph, 3JH-H = 11.2 Hz, 4JH-H = 1.2 Hz), 7.54 (ddd, 2H, Ph, 3JH-H =
3
PtCl2(cod) (36.3 mg, 97.0 μmol) and L1 (66.3
ere dissolved in 5 mL CH2Cl2
n
as a white powder. Yield: 92% (80.6
(dd, 4H, 3JH-H = 8.0 Hz), 7.49
, 4H), 7.40 (m, 6H), 7.42 (t, 8H, 3JH-H = 6.8 Hz), 7.32 (t, 6H, 3JH-H = 6.8 Hz), 6.90
(L9) C4
Following the same procedure as for complex
and L9 (75.0 mg, 110.5 μmol)
NN
PPh2PPh2
PtCl Cl
P
11.2 Hz, 4JH-H = 1.2 Hz), 7.45 (d, 5H, Ph, 4JH-H = 2.0 Hz), 7.35 (d, 5H, Ph, J1 = 7.6
Hz), 7.32 (m, 2H, Ph), 7.28 (m, 2H, Ph), 7.22 (dd, 4H, Ph, 3JH-H = 9.2 Hz, 4JH-H = 2.0
Hz), 7.16 (dd, 2H, Ph, 3JH-H = 7.2 Hz, 4JH-H = 2.8 Hz), 6.73 (dd, 2H, Ph, 3JH-H = 8.0
Hz, 4JH-H = 1.6 Hz), 4.41 (t, 2H, CH2), 4.26 (t, 2H, CH2), 3.35 (m, 2H, CH), 1.01 (d,
3H, CH3, 3JH-H = 7.6 Hz), 0.73 (d, 3H, CH3, 3JH-H = 6.8 Hz), 0.47 (dd, 3H, Pd(CH3), 3JH-H = 7.6 Hz, 3JH-H = 4.4 Hz). 31P NMR (CDCl3) δ 91.3 (d, JP-P = 28 Hz, P trans to Cl), 81.0 (d, JP-P = 28 Hz, P
trans to CH3).
cis-[PtCl2(L1)] C
mg, 104.1 μmol) w
and stirred for 2 hours at r.t. Then the solvent was
removed in vacuo. After that the remaining traces
g 2 times with 5 mL hexanes to leave complex C3
mg, 89.3 μmol).
1H NMR (CDCl3) δ 7.99 (dd, 4H, 3JH-H = 8.0 Hz), 7.73
of solvent were removed by strippi
(m
(dd, 4H, 3JH-H = 7.6 Hz, 4JH-H = 1.2 Hz), 4.29 (t, 2H, CH, 3JH-H = 7.2 Hz), 3.60 (dt, 4H,
CH2, 3JH-H = 14.4 Hz, 4JH-H = 2.8 Hz), 3.01 (t, 4H, CH2, 4JH-H = 13.6 Hz), 0.80 (d, 6H,
CH3, 3JH-H = 6.8 Hz). 31P NMR (CDCl3) δ 62.1 (s, JPt-P = 4151 Hz).
cis-PtCl2
C3, but using lig
and PtCl2(cod) (35.3 mg, 94.3 μmol) complex
C4 was obtained in a yield of 95% (84.7 mg,
89.6 μmol).
NN
PPh2PPh2
PtCl Cl
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
47
1H NMR (CDCl3) δ 8.24 (pq, 4H, 4JH-H = 4.4 Hz, 3JH-H = 6.8 Hz), 7.96 (pq, 4H, 4JH-H
4.4 Hz, 3JH-H = 6.8 Hz), 7.51 (t, 14H, 3JH-H = 7.2 Hz), 7.26 (t, 4H, 3JH-H = 2.4 Hz),
Following the same procedure as for complex C3,
d L5 (17 mg, 23 μmol) and
[Rh(μ-Cl)(CO)2]2 (56.9 mg,
146.3 μmol) and L1 (186.4 mg,
6
microcrystalline solid.
.6 (dd, cis, P trans to Cl, JRh-P = 180 Hz, JP-P = 33 Hz), 81.0 (d,
ans, JRh-P = 133 Hz), 75.3 (dd, cis, P trans to CO, JRh-P = 133 Hz, JP-P = 33 Hz)
3; H,
NN
PPh2PPh2
PtCl Cl
=
6.99 (t, 4H, 3JH-H = 3.6 Hz), 3.83 (t, 2H, NCH, 3JH-H = 8.4 Hz), 3.00 (m, NCH2), 2.67
(m, 2H, NCH2), 1.76 (m, 2H, CH2CH2CH2), 1.35 (m, 4H, CH2CH3), 0.30 (t, 6H, CH3, 3JH-H = 7.6 Hz). 31P NMR (CDCl3) δ 60.5 (s, JPt-P = 4285 Hz).
cis-PtCl2(L5) C5
but using ligan
PtCl2(cod) (8 mg, 21 μmol) complex C5 was
obtained as a white solid.
= 4120 Hz).
31P NMR (CDCl3) δ 60.4 (s, JPt-P
[Rh(Cl)(CO)(L1)] C6
292.7 μmol) were stirred in 10
mL of CH2Cl2 for 16 hours,
giving a light yellow solution.
was obtained as a bright-yellow
31P NMR (CDCl3) δ 99
After removal of the solvent in vacuo, complex C
tr
FTIR (ATR mode, solid, cm-1): ν 1967.5 (Rh(CO)).
Anal. Calcd. for C43H42ClN2OP2Rh: C, 64.31; H, 5.27; P, 3.49. Found: C, 64.1
5.37; P, 3.55.
NN
PPh2PPh2
RhCl CO
RhP CO
PCl
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
Crystal structure determinations
Intensity data for the molecular structures L6, L7, 5 and C1 were collected using
graphite-monochromated Mo-Kα radiation, on a Nonius Kappa CCD diffractometer
(see Table 6 for experimental details). An semi-empirical absorption correction based
on multiple measurements was applied using SADABS.[37a] The structure was solved
by automated Patterson methods using DIRDIF,[37b] and refined on F2 using
SHELXL97.[37c] Structure validation and molecular graphics preparation were
performed with the PLATON package.[38]
Table 6: Selected crystallographic data for molecular structures L6, L7, 5 and C1.
L6 L7 5 C1 formula C42H54N2P2 C46H46N2P2 C44H46N2O2P2 C42H42Cl2N2P2Pd FW (g mol-1) 648.81 688.79 696.77 814.04 crystal system triclinic monoclinic triclinic monoclinic space group P1 (no. 1) P21 (no. 4) P1 (no. 2) P21 (no. 4) a (Å) 8.6982(3) 15.2158(14) 11.5759(9) 11.3330(10) b (Å) 10.0730(6) 8.1986(4) 12.0731(10) 26.0777(10) c (Å) 11.6643(7) 15.7910(13) 16.6081(11) 12.6339(10) α (º) 69.066(4) 90 93.340(7) 90 β (º) 79.067(3) 103.775(7) 109.111(7) 92.7710(10) γ (º) 74.770(5) 90 117.831(6) 90 V (Å3) 915.79 1913.2 1876.6(3) 3729.4(5) Z 1 1 2 4 dcalc (g cm-3) 1.176 1.196 1.233 1.450 μ (Mo-Kα) (mm-1) 0.150 0.148 0.155 0.760 F(000) 350 732 740 1672 crystal size (mm) 0.20x0.2x0.3 0.24x0.30x0.48 - - T (K) 150 150 150 150 total reflections 23754 54708 - - unique reflections 7213 8737 - - R(nt) 0.029 0.029 - - wR2 (F2) 0.0782 0.1202 - - λ (Å) 0.71073 0.71073 0.71073 0.71073 R1 (F) 0.0317 0.0456 - -
48
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior
2.7 References [1] C. A. Bessel, P. Aggarwal, A. C. Marschilok, K. J. Takeuchi, Chem. Rev. 2001, 101, 1031. [2] For a recent review on chiral phosphorus ligands, see: W. Tang, X. Zhang, Chem. Rev. 2003,
103, 3029. [3] a) M. Rodriguez i Zubiri, M. L. Clarke, D. F. Foster, D. J. Cole-Hamilton, A. M. Z. Slawin, J.
Woollins, J. Chem. Soc., Dalton Trans. 2001, 969; b) A. M. Z. Slawin, M. Wainwright, J. D.
Woollins, J. Chem. Soc., Dalton Trans. 2002, 513. [4] a) M. S. Balakrishna, V. Sreenivasa Reddy, S. S. Krishnamurthy, J. F. Nixon, J. C. T. R.
Burckett St. Laurent, Coord. Chem. Rev. 1994, 129, 1 and references therein; b) M. S.
Balakrishna, M. G. Walawalker, J. Organomet. Chem. 2001, 628, 76. [5] M. P. Magee, H.-Q. Li, O. Morgan, W. H. Hersh, Dalton Trans. 2003, 387. [6] a) M. Fiorini, G. M. Giongo, F. Marcati, W. Marconi, J. Mol. Cat. 1976, 1, 451; b) M. Fiorini,
F. Marcati, G. M. Giongo, J. Mol. Cat. 1977/78, 3, 385; c) M. Fiorini, F. Marcati, G. M.
Giongo, J. Mol. Cat. 1978, 4, 125; d) M. Fiorini, G. M. Giongo, J. Mol. Cat. 1979, 5, 303. [7] G. Pracejus, H. Pracejus, Tetrahedron Lett. 1977, 39, 3497. [8] K. Kashiwabara, K. Hanaki, J. Fujita, Bull. Chem. Soc. Jpn. 1980, 53, 2275. [9] R. Guo, X. Li, J. Wu, W. H. Kwok, J. Chen. M. C .K. Choi, A. S. C. Chan, Tetrahedron Lett.
2002, 43, 6803. [10] a) S. Breeden, M. Wills, J. Org. Chem. 1999, 64, 9735; b) S. Breeden, D. J. Cole-Hamilton, D.
F. Foster, G. J. Schwarz, M. Wills, Angew. Chem. Int. Ed. 2000, 39, 4106; c) J. Ansell, M.
Wills, Chem. Soc. Rev. 2002, 31, 259. [11] V. M. Mastrano, L. Quintero, C. A. de Parrodi, E. Juaristi, P. J. Walsh, Tetrahedron, 2004, 60,
1781. [12] M. W. van Laren, C. J. Elsevier, Angew. Chem. Int. Ed. 1999, 38, 3715. [13] a) M. L. Clarke, A. M. Z. Slawin, J. D. Woollins, Phosphorus, Sulfur and Silicon, 2001, 169,
5; b) S. M. Aucott, M. L. Clarke, A. M. Z. Slawin, J. D. Woollins, J. Chem. Soc., Dalton
Trans. 2001, 972. [14] P. W. Dyer, J. Fawcett, M. J. Hanton, R. D. W. Kemmitt, R. Padda, N. Singh, Dalton Trans.
2003, 104. [15] S. Jeulin, S. Duprat de Paule, V. Ratovelomanana-Vidal, J. P. Genêt, N. Champion, P. Dellis,
Angew. Chem. Int. Ed. 2004, 43, 320 and references therein. [16] R. P. Pinnell, C. A. Megerle, S. L. Manatt, P. A. Kroon, J. Am. Chem. Soc. 1973, 95, 977. [17] V. D. Makhaev, Z. M. Dzhabieva, S. V. Konovalikhin, O. A. D’Yachenko, G. P. Belov,
Koord. Khim. 1996, 22, 598. [18] a) G. P. C. M. Dekker, C. J. Elsevier, K. Vrieze, P. W. N. M. van Leeuwen, Organometallics
1992, 11, 1937; b) M. A. Zuideveld, B. H. G. Swennenhuis, M. D. K. Boele, Y. Guari, G. P.
F. van Strijdonck, J. N. H. Reek, P. C. J. Kamer, K. Goubitz, J. Fraanje, M. Lutz, A. L. Spek,
P. W. N. M. van Leeuwen, J. Chem. Soc. Dalton Trans. 2002, 2308. [19] A. D. Burrows, M. F. Mahon, M. T. Palmer, J. Chem. Soc., Dalton Trans. 2000, 1669.
49
Chapter 2 Synthesis of Bisaminophosphine Ligands and Their Coordination Behavior [20] A. M. Z. Slawin, J. D. Woollins, Q. Zhang, Inorg. Chem. Commun. 1999, 2, 386. [21] J. Bravo, C. Cativiela, J. E. Chaves, R. Navarro, E. P. Urriolabeitia, Inorg. Chem. 2003, 42,
1006. [22] C. J. Cobley, P. G. Pringle, Inorg. Chim. Acta 1997, 265, 107. [23] J. I. van der Vlugt, M. Fioroni, J. Ackerstaff, R. W. J. M. Hanssen, A. M. Mills, A. L. Spek,
A. Meetsma, H. C. L. Abbenhuis, D. Vogt, Organometallics, 2003, 22, 5697. [24] a) J. Grimblot, J. P. Bonnelle, A. Mortreux, F. Petit, Inorg. Chim. Acta 1979, 34, 29; b) J.
Grimblot, J. P. Bonnelle, C. Vaccher, A. Mortreux, F. Petit, G. Pfeiffer, J. Mol. Cat. 1980, 9,
357. [25] P. Suomalainen, S. Jääskeläinen, M. Haukka, R. H. Laitinen, J. Pursiainen, T. A. Pakkanen,
Eur. J. Inorg. Chem. 2000, 2607. [26] M. J. Atherton, K. S. Coleman, J. Fawcett, J. H. Holloway, E. G. Hope, A. Karaçar, L. A.
Peck, G. C. Saunders, J. Chem. Soc., Dalton Trans. 1995, 4029. [27] K. G. Moloy, J. L. Petersen, J. Am. Chem. Soc. 1995, 117, 7696. [28] D. R. Drew, J. R. Doyle, Inorg. Synth. 1990, 28, 346. [29] F. T. Ladipo, G. K. Anderson, Organometallics, 1994, 13, 303. [30] H. C. Clark, L. E. Manzer, J. Organomet. Chem. 1973, 59, 411. [31] H. Mimoun, J. Y. de Saint Laumer, L. Giannini, R. Scopelliti, C. Floriani, J. Am. Chem. Soc.
1999, 121, 6158. [32] L. Xueliang, Z. Suizhi, G. Hefu, Huaxue Shiji, 1994, 16, 132. [33] O. Equey, A. Alexakis, Tetrahedron Asymmetry, 2004, 15, 1069. [34] L. Weber, A. Rausch, H. B. Wartig, H. -G. Stammler, B. Neumann, Eur. J. Inorg. Chem.
2002, 2438. [35] R. P. Kamalesh Babu, S. S. Krishnamurthy, M. Nethaji, Tetrahedron Asymmetry, 1995, 6,
427. [36] S. M. Ludeman, D. L. Bartlett, G. Zon, J. Am. Chem. Soc. 1979, 44, 1163. [37] a) SADABS, Bruker AXS, Karlsruhe, Germany, 2003; b) P. T. Beurskens, G. Admiraal, G.
Beurskens, W. P. Bosman, S. García-Granda, R. O. Gould, J. M. M. Smits, C. Smykalla,
DIRDIF99 program system; University of Nijmegen, The Netherlands, 1999; c) G. M.
Sheldrick, SHELXL97; University of Göttingen, Germany, 1997. [38] A. L. Spek, J. Appl. Cryst. 2003, 36, 7.
50
Chapter 3
DFT Study into Models of
Bisaminophosphine Ligands
Application of Bisaminophosphine Ligands in
Rh-Catalyzed Asymmetric Hydrogenation
DFT calculations were performed on model compounds for
bisaminophosphine ligands to analyze the geometries and
charge distributions. The computed structure of a simplified
cis-Pd complex of a bidentate bisaminophosphine ligand gives
valuable information on the coordination behavior.
Catalysts generated in situ from [Rh(cod)2]BF4 and
bisaminophosphine ligands perform efficiently in the
asymmetric hydrogenation of (Z)-N-acetylaminocinnamate with
ee’s up to 91%. The individual contributions of
aminophosphine moieties are recognized.
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
52
MeO
AcO
COOH
NHAc
COOMe
NHAc
[Rh(cod)(DiPamp)]BF4 / H2
deprotection
HO
HO
COOH
NH2
iv v
3.1 Introduction
The demand for enantiomerically pure compounds especially for pharmaceuticals and
agrochemicals paved the way for asymmetric hydrogenation to become one of the most
studied and efficient methods to produce chiral compounds.[1] The breakthrough in this
field was initiated by Knowles and ultimately led to the discovery of the C2-symmetric
chelating compound DiPAMP (Figure 1) as a very efficient ligand the Rh-catalyzed
asymmetric hydrogenation of dehydroamino acids.[2]
PP
MeO
OMe
i ii iii
O
O
PPh2
PPh2
P
P
Figure 1 Breakthrough ligands for asymmetric hydrogenation: DiPamp (i), DIOP (ii) and DuPhos (iii).
The industrial production of L-DOPA (Parkinson’s Disease) by Monsanto emphasized
the possibility of a practical synthesis employing the developed technology (Figure
2).[3] For this work Knowles was awarded the Nobel Prize in 2001.[4] Other important
ligands in this reaction over the years are DIOP (Kagan)[5] and DuPhos (Burk).[6]
(Figure 1)
Figure 2 Monsanto’s L-DOPA (iv) process and a model substrate (v) for the asymmetric transformation.
Especially the fine chemicals industry has a strong interest in the development of new
generic classes of chiral ligands. Also in academia there is a continuing interest in this
subject aiming at deeper insight. This is reflected by the large number of publications
appearing every year. The two dedicated issues in Advanced Synthesis & Catalysis in
2003 on the subject of catalytic hydrogenation underlines this.[7]
Recently the focus shifted towards the application of monodentate ligands, mainly
phosphoramidites and phosphites (Figure 3) with a strong interest from industry due to
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
a simple synthesis.[8] Here high-throughput experimentation techniques come into play,
especially since coworkers from DSM together with the group of Feringa and the
research group of Reetz independently discovered that combinations of chiral ligands or
even the application of chiral and an additional achiral ligand could increase the
performance of individual systems significantly.[9] The number of experiments to be
performed raises exponentionally with all variables, surely if one takes into account that
each substrate has its own optimal catalytic system (ligand, metal-(precursor) and
physical conditions). A rational approach is therefore highly desirable.
O
OP O
O
OP NMe2
vi vii Figure 3 Parent monodentate BINOL-based phosphoramidite (vi) and phosphite (vii) ligands for
asymmetric hydrogenation.
Many attempts are made to quantify ligand properties and their performance in
catalysis. Two of the earliest and most famous are the quantifications of electronic
effects and steric effects by the introduction of the respective Tolman factors χ and
θ.[10] These parameters were applied to monodentate phosphorus ligands. For bidentate
phosphorus ligands the natural bite angle βn was introduced by Casey et al.[11] and
further developed by Van Leeuwen who correlated βn of series of ligands to their
performance in various homogeneously catalyzed transformations.[12] Nowadays still
high demands in computational power are required to assess ligand properties in order
to ultimately come to de novo design of ligands for a specific reaction, substrate and
regio- and stereospecifity, which still remains an elusive goal.[13]
In the late 1970’s symmetrically substituted bisaminophosphine ligands were developed
and used in the Rh-catalyzed asymmetric hydrogenation of functionalized alkenes.[14-17]
53
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
N
PEt2
N
PEt2
N N
PPh2PPh2
N NMe Me
PPh2 PPh2
MeN N
PPh2
Me
PPh2
Ph Ph
viii ix
x xi
Figure 4 PNNP ligands for Rh-catalyzed asymmetric hydrogenation.
Figure 4 shows some of the reported PNNP ligands; viii and ix would form 7-
membered rings with the stereogenic carbons outside the ring when complexed to Rh
while x and xi have the stereogenic information within this chelate ring.
Ligands viii-xi all performed distinctly different in catalysis and in order to come to a
better understanding of the catalytic system the expansion of this applied set of ligands
is desired. When the individual contributions of the two aminophosphine moieties in
the ligands could be investigated by an independent variation of the two chiral amines
used this would be a powerful additional tool. A fine attempt was made by Roucoux et
al.[18], however they only used commercially available non-symmetric diamines for
their purpose and were therefore limited in the number of variations.
In this chapter we present DFT calculations on model compounds mimicking the
studied bisaminophosphine ligands, to come to a better understanding of the relevant
electronic and geometric parameters of our system. The bisaminophosphine ligands
described in Chapter 2, both symmetrically as non-symmetrically substituted, were
applied in the asymmetric hydrogenation of methyl Z-acetylaminocinnamate.
54
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
3.2 Results
3.2.1 DFT Calculations on model compounds.
Models. Phosphines substituted with aryl and alkyl groups typically show mainly σ-
donor character. In order to compare the electronic properties of the bisaminophosphine
ligands from Chapter 2 with ordinary phosphines, we calculated the electron densities
on the phosphorus atoms of the four model compounds PPh3 (I), CH3PPh2 (II),
Ph2PN(CH3)2 (III) and Ph2P(pyrrole) (IV), depicted in Figure 5, by using DFT
methods. For computational purposes monodentate analogues were considered. This
allowed us to avoid oversimplification by the commonly used PH2-group, which has
very little relevance to the actual systems and is normally chosen for obvious
restrictions by computation time, and to use the more realistic PPh2-group instead.
Using the bidentate equivalent of III, model compound V, depicted in Figure 6, we also
investigated the corresponding complex (VI), cis-[PdCl2(V)]. Here methyl groups were
introduced on the phosphorus moieties instead of phenyl groups as a compromise
arising from computational limitations.
PPh3 CH3PPh2N
CH3 CH3
PPh2N
PPh2
I II III IV Figure 5 Model compounds I to IV as used in the DFT calculations.
NN
PMe2Me2P
CH3CH3NN
PMe2Me2P
CH3CH3
PdCl Cl
V VI
Figure 6 Illustration of model compounds V and VI (cis-[PdCl2(V)]), and the optimized structure for VI,
calculated by DFT.
Geometries. Selected geometric parameters (bond lengths and angles) obtained for the
optimized geometries of model compounds I-IV are listed in Table 1. The P-Cα,Ph
distance was constant for the complete series at 1.85 Å, while significant differences
were found for the P-N distance. In case of the aminophosphine III this bond length
55
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
was 1.73 Å, which is in good agreement with the values found for such bonds in
compounds 6 and 7 (1.68 Å, vide Chapter 2) while in the pyrrolyl-based compound IV
this bond length was calculated to be 1.76 Å. This is in close agreement with the
experimental data provided by Atwood et al.[19], being 1.71 Å. The only angle
comparable between the aforementioned models is the Cα,Ph-P-Cα,Ph’, which shows a
nearly constant value. Little differences exist between the internal angles Cα,Ph-P-CCH3
for compound II and Cα,Ph-P-N for compound III, which is due to the different
orientation of the two phenyl groups. These are reciprocally in trans position, towards
the methyl and N(CH3)2 group.
Table 1 Selected bond lengths and angles for the optimized geometries of model compounds I to IV and
Pd-complex VI.a
PPh3 (I)
CH3PPh2 (II)
(CH3)2NPPh2 (III)
Ph2P(pyr) (IV)
cis-PdCl2(V) (VI)
Bond lengths (Å) -Cα, Ph 1.85 P-Cα, Ph 1.85 P-Cα, Ph 1.85 P-Cα, Ph 1.85 Pd-Cl 2.40
P-CCH3 1.86 P-N 1.73 P-N 1.76 Pd-P 2.35
N-CCH3 1.46 P-N 1.72
P-CCH3 1.84
N-CCH3 1.47
P-P 3.85 Bond Angles (°)a
Cα,Ph-P-Cα, Ph Cα, Ph-P-Cα, Ph 101.7 Cα, Ph-P-Cα, Ph 102.0 Cα, Ph-P-Cα, Ph 102.6 Cl1-Pd-Cl2 87.5 102.6 Cα, Ph-P-CCH3
102.5 Cα, Ph-P-N 101.5 Cα, Ph-P-N 100.7 Cl1-Pd-P1 81.5 Cα, Ph-P-CCH3 99.9 Cα, Ph-P-N 105.8 Cα, Ph-P-N 102.4 Cl2-Pd-P2 81.5 N-P-CCH3 115.6 Cl1-Pd-P2 168.9 Cl2-Pd-P1 168.9 Pd- P1-N1 121.2 Pd- P1-N2 121.2 P1-Pd-P2 109.5
a Repeated entries are referred to the angles estimated on the two different Ph or methyl groups.
Also listed in Table 1 are the geometric parameters of the corresponding palladium
complex VI of model compound V. It is striking that the computed values for the bond
lengths for Pd-complex VI are slightly higher than found in complex C1 (Chapter 2).
This might originate from the applied basis set 6-31G and B3LYP but could also be an
effect of the obvious different substituents on the phosphorus atom, i.e. methyl-units
instead of phenyl groups. As for the correlation between the calculated and
experimentally determined geometry of complex C1, the overall general features are
comparable, especially with regard to the intramolecular P-P distance (3.85 Å vs. 3.83
56
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
Å) and some of the angles (e.g. Pd-P1-N1). The angles found for the optimized
calculated geometry of VI deviate from experimentally determined values, with an
overestimation of the bite angle P1-Pd-P2 (Δ = ~12°) and consequently the Cl1-Pd-P2
angle falls short in the computed complex. This might well be related to the simplified
model structure used, neglecting also the chiral substituents on the nitrogen atoms.
Charges. The electron distributions in the model compounds were analyzed through an
electrostatic charges analysis. Although atomic charges are not an observable in
quantum mechanics, they are appropriate to get an idea on the electron distribution.
Different schemes and algorithms can be employed. In this study the Mulliken
population analysis was considered.[20] This method assigns charges by partitioning the
orbital overlap evenly between the two atoms that are involved. In Table 2, the
Mulliken atomic charges are reported for the four model compounds I-IV and Pd-
complex VI.
Table 2 Selected Mulliken Atomic Charges for model compounds I to IV and complex VI.a,b
PPh3 (I)
CH3PPh2 (II)
(CH3)2NPPh2 (III)
Ph2P(pyr) (IV)
cis-[PdCl2(V)] (VI)
Atomic Charges
P-Cα, Ph +0.186 P-Cα, Ph +0.352 P-Cα, Ph +0.491 P-Cα, Ph +0.562 Pd -0.967
P-N -0.421 P-N -0.321 P1-N1 +1.09
P1-N1 -0.304
P2-N2 +1.095
P2-N2 -0.304
Cl1 -0.258
Cl2 -0.258
a Electron Units (charge of electron is equal to -1) b Atoms considered in the Mulliken Population
analysis are in italics.
The charge on the phosphorus atoms is lowest in case of PPh3 (I), and the value
increases going along compounds II, III to IV. Therefore the introduction of a nitrogen
atom unequivocally raises the positive charge on the P atom, with a stronger effect
when the π-acidic pyrrole moiety is incorporated rather than a tertiary amine.
57
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
With respect to the palladium-complex VI there is strong indication of a nearly net -1
charge on the metal, with the P atoms bringing a +1 charge. Compound III bears
closest resemblance to the model ligand V used in compound VI. Comparing the net
charges on the free ligand with those present in the metal complex, the P atom
undergoes a dramatic change increasing the net positive charge of 0.5, while the charge
on the nitrogen atom decreases to a negative charge of -0.2. This can be interpreted as a
strong backdonation of the P atoms towards the Pd atom, especially if a net charge of -
0.2 is present on a chlorine atom.
3.2.2 Rh-catalyzed asymmetric hydrogenation of methyl Z-acetylaminocinnamate
The bisaminophosphine compounds described in Chapter 2 were employed as ligands
in the asymmetric hydrogenation of methyl (Z)-N-acetylaminocinnamate, one of the
benchmark substrates to assess the performance of a given chiral modifier in this
reaction (Eq. 1). The standard metal precursor [Rh(cod)2]BF4 was used, which during
catalysis is converted in-situ to [Rh(cod)L]BF4 while one of the cod (cyclooctadiene)
groups is lost and hydrogenated.
COOCH3
NHAc
H2
[Rh(cod)L]BF4
COOCH3
NHAc
*(1)
First the optimal solvent for our system was determined from a small set of commonly
used solvents in hydrogenation reactions, namely methanol, ethyl acetate and
dichloromethane. The novel ligand L7 (figure 7) was the ligand of choice for the
screening. Besides, for one entry an excess of ligand to metal (2.2 equiv) was used to
check if the performance would be significantly different. The obtained results under
typical conditions are summarized in Table 3.
58
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
Table 3 Initial solvent and stoichiometry screening for asymmetric hydrogenation of methyl (Z)-N-
acetylaminocinnamate results with [Rh(cod)L]BF4a
entry solvent equiv L TOF (mol mol-1 h-1)b conv (%)c ee (%)d 1 MeOH 1.1 73 100 40 (S) 2 CH2Cl2 1.1 37 100 66 (S) 3 EtOAc 1.1 31 100 85 (S) 4 EtOAc 2.2 34 100 85 (S)
a Reaction conditions: 0.011 mmol L7 (R) ; 0.01 mmol [Rh(cod)2]BF4 ; 5 mL solvent ; H2 atmosphere
1.1 bar ; T = 25 ºC; 1 mmol (Z)-N-acetylaminocinnamate. b Turn Over Frequency, average conversion of (Z)-N-acetylaminocinnamate over first hour. c Percent conversion of (Z)-N-acetylaminocinnamate after 18 h. d Enantiomeric excess determined by chiral GC, absolute configurations given in parentheses.
NNPPh2PPh2
NN
PPh2 PPh2
NN
PPh2 PPh2
NN
PPh2 PPh2OMeMeO
NN
PPh2 PPh2
NN
PPh2 PPh2
NN
PPh2 PPh2
L1 L2
L3 L4
L5 L6
NN
PPh2 PPh2
NN
PPh2 PPh2
NN
PPh2 PPh2
L7 L10
L11 L12
Figure 7 Bisaminophosphine ligands L1-L7 and L10-L12 used in asymmetric hydrogenation.
Obviously the performance of our catalytic system is greatly influenced by the solvent
used. The initial TOF is highest for MeOH, but in this solvent the ee was lowest. This
may be caused by a degree of degradation of the ligand by the protic solvent, which is
59
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
partially described by Pracejus et al.,[15] although with another Rh-precursor ([Rh(μ-Cl)
(C2H4)2]2) methanol outperforms benzene as the solvent in both stereospecifity as
activity in that study. Entries 2 and 3 show equal TOF’s for dichloromethane and ethyl
acetate but for ethyl acetate the highest ee was obtained (85%), which might be caused
by a beneficial participation of the carboxylic function of this ester. A similar run with
double the amount of ligand (entry 4) showed comparable activity and selectivity.
Independent of the solvent the major product obtained has the S configuration.
The ligands shown in figure 7 were now applied in the reaction using ethyl acetate, the
solvent in which the highest ee was obtained for ligand L7. The results are summarized
in Table 4.
Table 4 Asymmetric hydrogenation of (Z)-N-acetylaminocinnamate results with [Rh(cod)L]BF4
a
Entry Ligand L TOF (mol mol-1 h-1)b conv (%)c ee (%)d 1 L1 50 >99 85 (R) 2 L2 43 >99 85 (S) 3 L3 65 >99 85 (S) 4 L4 32 91 79 (S) 5 L5 85 >99 91 (R) 6 L6 93 >99 16 (R) 7 L7 31 >99 85 (S)
non symmetrically substituted ligands 8 L10 58 >99 35 (S) 9 L11 39 >99 84 (S) 10 L12 48 >99 0
a Reaction conditions: 0.011 mmol ligand ; 0.01 mmol [Rh(cod)2]BF4 ; 5 mL EtOAc ; H2 atmosphere 1.1
bar ; T = 25 ºC; 1 mmol (Z)-N-acetylaminocinnamate. b Turn Over Frequency, average conversion of (Z)-N-acetylaminocinnamate over first hour. c Percent conversion of (Z)-N-acetylaminocinnamate after 18 h. d Enantiomeric excess determined by chiral GC, absolute configurations given in parentheses.
The first striking observation is that all ligands but L6 perform equally well when ee is
concerned. In a small range of ee’s on average 85% ee is obtained, in any case the
major product being the enantiomer with the sign of rotation opposite of the sign of
rotation of the amine used in the synthesis of the ligand. The exception L6 gives only a
low ee of 16% (entry 5) but remarkably it is also the fastest ligand in the series with an
initial TOF of 93 on average over the first hour. Best ligand overall is L5 based on the
60
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
most crowded 1-naphthylethyl moiety, which gives a good TOF and the highest ee of
91% (entry 4).
L6 is the only ligand which does not possess an aromatic ring in the stereogenic amine
part, possibly making the phosphorus atoms less crowded and subsequently the Rh
center more open for the incoming substrate. Possibly the absence of π-π interactions
between ligand and substrate also plays a role. These differences may be the cause for
the observed highest activity for L6, although activities do not reach industrially viable
values (>500 mol mol-1 h-1) but it should be noted that the applied pressure is only 1.1
bar while in commercial processes the use of higher pressures is common practice.
C2-rpea-ppa-PPh2 (L11) is a pseudo C2-symmetric ligand and gives the highest ee for
the three non symmetrically substituted ligands. The ligand with one achiral element
(L10) fails to block one specific quadrant which you could identify in the quadrant
diagram model.[2] This model occupied and vacant quadrants indicate areas of
maximum and minimum repulsive interactions between parts of the catalyst and the
prochiral substrate. Therefore is is not surprising the ligand L10 generated limited
chiral induction.
Figure 8 31P{1H} spectra of in-situ generated [Rh(cod)(bisaminophosphine)]BF4 complexes of ligand L5
(top) and ligand L10 (bottom).
A means to investigate the electronic properties of the bisaminophosphine ligands is to
measure the NMR-spectra of the corresponding complexes to assess the chemical shifts
61
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
(δ 31P{1H} (ppm)) and the coupling constant of the Rh-P bond (1JRh-P (Hz)). This has
been done for the various ligands, including the non symmetrical ligands; the selected
data from the obtained spectra are summarized in Table 5. For illustration purposes
NMR-spectra of one symmetrical (L5) and one non-symmetrical ligand (L10) are
shown in Figure 8.
It was confirmed that the complexes derived by in-situ mixing the ligand and the metal
precursor [Rh(cod)2]BF4 in CDCl3 are the same as a complex synthesized on a larger
scale, isolated by precipitation and purified by crystallization (entry 0 vs. entry 1).
All complexes show comparable values, chemical shifts δ 31P around 84 ppm, and
coupling constants around 164 Hz. Both values are in the expected ranges for a cationic
Rh(I) complex bearing a diene and an electron-poor diphosphorus ligand. No
quantitative relationship can be deduced from these numbers.
Table 5 NMR study [Rh(cod)(bisaminophosphine)]BF4 complexes.
Entry Ligand L δ 31P{1H} (ppm) 1JRh-P (Hz) JP-P (Hz) 0 L1 a 83.7 163.6 - 1 L1 83.8 162.4 - 2 L2 83.7 163.6 - 3 L3 85.7 163.6 - 4 L4 85.6 166.0 - 5 L5 84.0 161.1 - 6 L6 81.1 161.1 - 7 L7 82.9 163.6 -
non symmetrical ligands -
8 L10 89.6 , 77.5 159.9 , 164.7 24.4 9 L11 85.0 , 84.0 163.6 , 163.6 22.7 10 L12 84.9 162.5 -
a preformed and isolated complex
For the non symmetrically substituted ligands (entries 8-10) the picture changes, since
the phosphorus atoms are not equivalent anymore and coupling between the phosphorus
atoms occurs. For L10 the difference is significant (upfield shift of ca. 12 ppm for the
benzylaminophosphine moiety compared to the phenylethylaminephosphine group
(entry 8)). This proves the possibility of a new concept in designing and synthesizing
62
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
bisaminophosphine ligands with different electronic and steric properties for the two
donating groups.
3.3 Conclusions
From the DFT calculations the role of the nitrogen in P-N bond containing compounds
becomes more clear; the nitrogen increases the positive charge on the phosphorus atom,
which is indicative for a degree of π-bonding in the P-N bond. Upon complexation with
palladium the charge on P increases significantly.
The application of bisaminophosphine ligands in the Rh-catalyzed asymmetric
hydrogenation of (Z)-N-acetylaminocinnamate gives ee’s up to 91% and full conversion
under ambient conditions in the donating solvent ethyl acetate.
3.4 Perspective
The increase in computational power will enable the modeling of more realistic
compounds without concessions due to complexity. Or alternatively one may choose
higher level calculations to yet a better understanding of electronic interactions in
ligand, their transition metal complexes or even essential transformations during a
catalytic cycle.
The commercial availability of a wide range of chiral amines opens up extra
possibilities to find ligands with extraordinary effects in catalysis. The applied types of
bisaminophosphines may be used, and are being used, for various other homogeneously
catalyzed transformations; an example is the nickel-catalyzed alkylation of allylic
acetates.[21] Rhodium in combination with this type of bisaminophosphine ligands is
also effective in the asymmetric hydrogenation of activated ketones[18] The new
symmetrical and non symmetrical substituted bisaminophosphines reported here can
contribute to a better insight into and performance of these reactions, since their
modular and easy construction allows for ligand fine tuning in an automated synthesis
and testing setup.
63
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
3.5 Acknowledgements
Avantium Technologies is kindly acknowledged for financial support, Umicor Co is
thanked for the generous loan of precious metals. All DFT calculations were performed
by Marco Fioroni. We are indebted to Ton Staring for valuable technical assistance.
3.6 Experimental Section
General
All manipulations were carried out under argon using standard Schlenk techniques.
Chemicals were purchased from Acros, Aldrich, Lancaster or VWR and used as
received or distilled from CaH2 before use. Solvents were either taken HPLC-grade
from an argon-flushed column, packed with aluminum oxide, or distilled under argon
prior to use over an appropriate drying agent. NMR spectra were recorded at room
temperature on a Varian Mercury 400 MHz spectrometer. Chemical shifts are given in
ppm and spectra are referenced to CDCl3 (1H) or 85% H3PO4 (31P{1H}). All described
ligands were prepared following procedures described in Chapter two. (Z)-N-
acetylaminocinnamate was kindly synthesized by Gabriela Ionescu.[22] [Rh(cod)2]BF4
was synthesized following literature procedures[23] and kept under Ar.
Hydrogenation of (Z)-N-acetylaminocinnamate
A Schlenk tube was charged with 1 mmol substrate ((Z)-N-acetylaminocinnamate),
0.01 mmol catalyst precursor [Rh(cod)2]BF4 and ligand (0.011mmol) in 5 mL of the
appropriate solvent. After three H2 purges the reaction mixture was stirred at 298 K
under a constant H2 atmosphere (1.1 bar). Samples were taken under an outflow of H2
gas. The conversion was determined on a 50 m PONA (HP) column (carrier gas 150
kPa He, FID detector). For the ee measurement an L-Chiralsil Val column (carrier gas
120 kPa He, FID detector) was used.
NMR studies on [Rh(cod)(bisaminophosphine)]BF4 complexes
[Rh(cod)2]BF4 (4.0 mg, 9.9 µmol) and 1.0 equiv of the appropriate ligand were stirred
in 0.6 mL of CDCl3 at room temperature for 30 min. The solution was transferred to an
NMR tube and the locked 31P NMR spectrum was recorded.
64
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation
Computational Methods
For all the presented calculations, the Gaussian98 series of computer programs have
been used.[24]
Density Functional Methods
Standard computational methods based on the Density Functional Theory (DFT) have
been employed.[25] The used functional is the three-parameter exchange functional of
Becke[26] together with the correlation functional of Lee, Yang and Parr (B3LYP).[27]
For the P, C, Cl and H atoms the basis set used is the Pople style basis set 6-31G[28]
with diffuse (+) s- and p- functions added on the heavy atoms[29] and polarization
function[30] (d, p), adding one d function on the heavy atoms and one p function on the
hydrogens [6-31+G(d, p)]. For the transition metal palladium, the LanL2DZ Hay-Wadt
relativistic small-core effective core potential (ECP) and the corresponding basis set,
split valence double-ζ, were used.
The geometries of all the model compounds have been fully optimized using analytical
gradients technique at the B3LYP level of theory previously cited. No symmetry
constraints have been introduced. The optimized stationary points have been confirmed
through an harmonic vibrational analysis (B3LYP level), using analytical or numerical
differentiation of the obtained analytical energy first derivative. Energy calculations
were performed at the same level of the geometry optimization, including the zero-
point vibrational energy correction, applying the harmonic oscillator approximation.
65
Chapter 3 DFT Study into Models of Bisaminophosphine Ligands
3.7 References
[1] W. Tang, X. Zhang, Chem. Rev. 2003, 103, 3029 and references therein. [2] a) B. D. Vineyard, W. S. Knowles, M. J. Sabacky, G. L. Bachman, O. J. Weinkauff. J. Am.
Chem. Soc. 1977, 99, 5946. b) W. S. Knowles, Acc. Chem. Res. 1983, 16, 106. [3] W. S. Knowles, J. Chem. Educ. 1986, 63, 222. [4] W. S. Knowles, Angew. Chem. Int. Ed. 2002, 41, 1998. [5] H. B. Kagan, Chem. Commun. 1971, 481. [6] W. A. Nugent, T. V. RajanBabu, M. J. Burk, Science, 1993, 259, 479. [7] various authors, Adv. Synth. Catal. 2003, issues 1-2. [8] see for example a) M. v. d. Berg, A. J. Minnaard, R. M. Haak, M. Leeman, E. P. Schudde, A.
Meetsma, B. L. Feringa, A. H. M. de Vries, C. E. P. Maljaars, C. E. Willans, D. Hyett, J. A. F.
Boogers, H. J. W. Hendrickx, J. G. van de Vries, Adv. Synth. Catal. 2003, 345, 308. b) M. T.
Reetz, J.-A. Ma, R. Goddard, Angew. Chem. Int. Ed. 2005, 44, 2962. [9] a) D. Peña, A. J. Minnaard, J. A. F. Boogers, A. H. M. de Vries, J. G. de Vries, B. L. Feringa,
Org. Biomol. Chem. 2003, 1, 1087. b) M. T. Reetz, T. Sell, A. Meiswinkel, G. Mehler, Angew.
Chem. Int. Ed. 2003, 42, 790. [10] a) C. A. Tolman, J. Am. Chem. Soc. 1970, 92, 2953; b) C. A. Tolman, J. Am. Chem. Soc. 1970,
92, 2956 ; c) C. A. Tolman, Chem. Rev. 1977, 77, 313. [11] C. P. Casey, G. T. Whiteker, Israel J. Chem. 1990, 30, 299. [12] P. W. N. M. van Leeuwen, P. C. J. Kamer, J. N. H. Reek, P. Dierkes, Chem. Rev. 2000, 100,
2741 and references therein. [13] MM/SEQM: K. D. Cooney, T. R. Cundari, N. W. Hoffman, K. A. Pittard, M. D. Temple, Y.
Zhao, J. Am. Chem. Soc. 2003, 125, 4318; AMS model: K. Angermund, W. Baumann, E.
Dinjus, R. Fornika, H. Görls, M. Kessler, C. Krüger, W. Leitner, F. Lutz, Chem. Eur. J. 1997, 3,
755; DFT: S. A. Decker, Organometallics, 2001, 20, 2827 and F. Delbecq, V. Guiral, P. Sautet,
Eur. J. Org. Chem. 2003, 2092. [14] a) M. Fiorini, G. M. Giongo, F. Marcati, W. Marconi, J. Mol. Cat. 1976, 1, 451; b) M. Fiorini,
F. Marcati, G. M. Giongo, J. Mol. Cat. 1978, 3, 385; c) M. Fiorini, F. Marcati, G. M. Giongo, J.
Mol. Cat. 1978, 4, 125; d) M. Fiorini, G. M. Giongo, J. Mol. Cat. 1979, 5, 303. [15] G. Pracejus, H. Pracejus, Tetrahedron Lett. 1977, 39, 3497. [16] K. Kashiwabara, K. Hanaki, J. Fujita, Bull. Chem. Soc. Jpn. 1980, 53, 2275. [17] R. Guo, X. Li, J. Wu, W. H. Kwok, J. Chen. M. C. K. Choi, A. S. C. Chan, Tetrahedron Lett.
2002, 43, 6803. [18] A. Roucoux, I. Suisse, M. Devocelle, J.-F. Carpentier, F. Agbossou, A. Mortreux, Tetrahedron:
Asymmetry, 1996, 7, 379. [19] J. L. Atwood, A. H. Cowley, W. E. Hunter, S. K. Mehrotra, Inorg. Chem. 1982, 21, 1354. [20] R. S. Mulliken, J. Chem. Phys. 1955, 23, 1833. [21] H. Bricout, J.-F. Carpentier, A. Mortreux, Tetrahedron Lett. 1996, 37, 6105. [22] S. Gladiali, L. Pinna, Tetrahedron Asymmetry, 1991, 2, 623.
66
Chapter 3 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydrogenation [23] T. G. Schenck, J. M. Downes, C. R. C. Milne, P. B. Mackenzie, H. Boucher, J. Whelan, B.
Bosnich, Inorg. Chem. 1985, 24, 2334. [24] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A.
D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q.
Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J.
Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R.
Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C.
Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L. Andres,
M. Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian 98, Revision A.3 (by Gaussian, Inc.),
1998, Pittsburgh. [25] R. G. Parr, W. Yang, in Density Functional Theory of Atoms and Molecules, R. G. Parr, and W.
Yang, (Eds.): Oxford Science Publications, 1989, Oxford. [26] A. D. Becke, J. Chem. Phys. 1993, 95, 5648. [27] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B, 1988, 37, 785. [28] W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56, 2257. [29] M. J. Frisch, J. A. Pople, J. S. Binkley, J. Chem. Phys. 1984, 80, 3265. [30] R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys. 1980, 72, 650.
67
Chapter 4
Application of Bisaminophosphine Ligands in
Rh-Catalyzed Asymmetric Hydroformylation
C2-symmetric bisaminophosphine ligands were applied in the
Rh-catalyzed asymmetric hydroformylation of prochiral
alkenes. For styrene the branched/linear ratio reached 12, the
ee remained limited to 12%. Vinyl acetate was hydroformylated
more efficiently: the desired branched product was observed in
high selectivity with a branched/linear ratio up to 50. The ee’s
were medium with a maximum of 51%. HP-NMR studies
indicated that equatorial - equatorial is the preferred
coordination mode, which could be confirmed by HP-IR
spectroscopy.
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
70
R R
CHO
RCHO+
branched linear
CO/H2
cat. *
4.1 Introduction
Hydroformylation or oxo-synthesis is the atom-efficient process in which an aldehyde
is produced from an alkene via the catalyzed addition of carbon monoxide and
dihydrogen (Eq. 1).
(1)
Propene is the most important substrate in industrial applications, since annually bulk
amounts of n-butanal are converted among others to plasticizer alcohols for the
polymer industry.[1-3] The regioselectivity is here a very important parameter; the linear
aldehyde being the desired product.[4]
The branched product is desired in the asymmetric hydroformylation of prochiral
substrates, here the carbon skeleton of an alkene is extended by one carbon atom and a
stereocenter is created. The thus formed optically active aldehydes are of high synthetic
utility in organic synthesis.[5] These chiral molecules, in enantiomerically pure form, are
valuable precursors for drugs, agrochemicals and food additives.[4] An example lies in
the asymmetric hydroformylation of allyl cyanide described by De Vries et al.[6] The
product 2-methyl-3-cyanopropanal can be converted by a hydrogenation step to 3-
methyl-4-aminobutanol which is used as a building block for a new Tachykinin NK1
receptor antagonist[7]
A real breakthrough occurred in this field with the discovery of the Rh/BINAPHOS (i)
catalyst systems by Takaya et al.[8] Since then, new active chiral ligands such as
aminophosphine phosphinites (ii)[9] and diphosphites (iii)[10,11] have been developed,
see figure 1. Also Pt/Sn-based systems were considered and successfully applied.[12]
However, hydroformylation has yet not been used in organic synthesis on a frequent
basis. Simultaneous control of regio- and enantioselectivity, while maintaining
sufficient activity is the big challenge to be addressed.
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
O
O
PO
PPh2
O
MeN
Ar2PPh
Me
PPh
R
i ii
OOPP
O
O
tBu
tBu
O
O
tBu
tBu
R
R
R
R
P
O
O
tBu
tBu
R
R
OO
OP
O
O
tBu
tBu
R
R
O
iii iv Figure 1 Ligands for asymmetric hydroformylation: BINAPHOS (i), AMPP (ii), diphosphites (iii, iv).
Recently progress has been made in the field with the development of new chiral
AMPP (aminophosphine phosphonite) ii ligands in the group of Vogt.[9] Although the
BINAPHOS system remains the benchmark catalyst in asymmetric hydroformylation,
the AMPP ligand family provides enormous potential for variation and ligand fine-
tuning. A very recent theoretical investigation by Carbó et al. gave more insight in
these systems and will potentially lead to more successful ligands in due time.[13] Sugar
based diphosphites (e.g. iv) give a tremendous number of successful ligands from the
chiral pool.[11]
All ligands which provide high enantioselectivities have one common characteristic:
the ligands coordinate in the hydrido rhodium complexes in a specific mode. Either in
the equatorial/equatorial (ee) manner (diphosphites) or the equatorial/axial (ea) manner
(BINAPHOS/AMPP) in the trigonal bipyramidal.[14] In the latter cases the stronger π-
acceptor phosphorus atom occupies the axial position, trans to the hydride, while the
equatorial position of the complex is occupied by the stronger σ-donor phosphorus
atom.
71
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
Rh CO
H
CO
P
P PP
Rh CO
HOC
ee ea
Rh RhCO
PPP
P
OC
CO
CO
Scheme 1 Equatorial/equatorial ee and equatorial/axial ea coordination modes in trigonal bipyramidal
Rh-complexes. Righthand side: catalytically inactive bridged dinuclear species.
In our view a neglected ligand type in asymmetric hydroformylation may be found in
the bisaminophosphine ligands. These ligands, easily constructed from commercially
available chiral amines and chlorophosphines (see Chapter 2), provide the desired
modularity and therefore could be fine-tuned for different enantioselective
transformations. Especially the possibility of independent variation of the two amine
groups, and thereby the two aminophosphine moieties, allows to tune electronic and
steric properties. This is in contrast with the BINAPHOS ligand (i) in Figure 1 which
suffers from a tedious multi-step synthetic route.
v
PP
N
NN
N
Ph
Ph
Figure 2 ESPHOS ligand (v) for asymmetric hydroformylation of vinyl acetate.
A successful ligand based on chiral diamines, although in the form of a bis-
diazaphospolidine, was developed by Breeden et al.[15] This ESPHOS ligand (v) (figure
2) was applied in the asymmetric hydroformylation of vinyl acetate, with high
enantioselectivities, however with styrene as the substrate the product was virtually
racemic. The limited commercial availability of the used chiral diamine and the not
straightforward synthesis of the 1,2 substitution pattern on the phenyl backbone makes
the synthesis less suited for modular approaches.
Here we present the first application of bisaminophosphine ligands, synthesized in
Chapter 2 in the asymmetric hydroformylation of alkenes. The coordination of the
ligand to the rhodium center in the resting state of the catalyst under syngas conditions
was investigated using in-situ HP-NMR and HP-IR spectroscopy techniques.
72
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
4.2 Results
4.2.1 Catalysis
Styrene, being a generally accepted and widely used benchmark substrate for the
asymmetric hydroformylation reaction, was selected as the first substrate see Eq. 2.
CO/H2
Rh/L
CHO
CHO
+*
(2)
The catalysts were prepared for the screening by in situ mixing the ligands with the
metalprecursor Rh(acac)(CO)2 in a ratio 2:1 and heating in an autoclave under typical
reaction conditions (60°C, 20 bar syngas (1:1 CO/H2)). Subsequently the substrate with
internal standard was injected under pressure and the mixture was allowed to react for
15 h. After workup and derivatization to the trifluoro acetic ester the products were
analyzed by (chiral) GC. The results are presented in Table 1.
Table 1 Selected results of asymmetric hydroformylation of styrene performed by bisaminophosphine /
Rh(acac)(CO)2 catalyst systems.a
Entry Ligand conv (%) b b/l c ee (%) d
1 L1 57 3.6 -2 2 L2 66 3.6 2 3 L3 e 20 2.4 2 4 L4 22 7.2 9 5 L5 34 11.5 -1 6 L6 92 4.2 -12 7 L7 61 4.6 5
a Reaction conditions: T = 60°C; p = 20 bar (1:1 CO/H2); solvent: toluene, [Rh] = 0.46 mM; Rh:S =
1300; L:Rh = 2; preformation t = 1h; t = 15h. b Percent conversion of styrene after 15h. c Branched /
linear ratio. d Enantiomeric excess determined by chiral GC. e T = 40°C.
All ligands give active catalysts in this enantioselective conversion and the desired
branched product is in any case the dominant regioisomer. Only the ligand based on 1-
naphthylethylamine L5, however, induces the excellent branched/linear ratio which is
generally expected for styrene and similar vinylarenes (entry 5).[16] The steric crowding
around the Rh center is expected to be highest for this catalyst complex. The highest
conversion and simultaneously the highest enantioselectivity was found for ligand L6,
based on the cyclohexyl moiety, but the overall performance is still far from good.
73
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
O
O
Me
CO/H2
Rh/L O
CHOO
Me OCHO
O
Me+
* (3)
The second substrate for the asymmetric hydroformylation is vinyl acetate, see Eq. 3.
The branched product of this conversion is 2-acetoxypropanal which is a precursor for
the synthesis of hydroxy-amino acid threonine in the Strecker synthesis. The product
can also be converted to 2-hydroxypropanal, a useful intermediate in the synthesis of
steroids, pheromones, antibiotics and peptides.[17] Here the reaction mixture can be
analyzed directly by GC after workup without an additional derivatization step. Results
are given in Table 2.
Table 2 Selected results of asymmetric hydroformylation of vinyl acetate performed by
bisaminophosphine / Rh(acac)(CO)2 catalyst systems.a
Entry Ligand conv (%)b b/l c ee (%)d
1 L1 70 3.2 4 2 L2 77 3.6 -5 3 L3 nd nd nd 4 L4 71 >50 12 5 L5 82 17 32 6 L5e 65 20 51 7 L6 68 4.2 18 8 L7 64 2.4 20
a Reaction conditions: T = 60°C; p = 20 bar (1:1 CO/H2); solvent: benzene, [Rh] = 0.54 mM; Rh:S =
1400; L:Rh = 2; preformation t = 1h; t = 20h. b Percent conversion of vinyl acetate after 20h. c Branched /
linear ratio. d Enantiomeric excess determined by chiral GC. e T = 40°C. f nd = not determined.
The use of bisaminophosphine ligands for the asymmetric hydroformylation of vinyl
acetate proves to be an efficient one. All applied ligands give more than 60%
conversion and the desired branched product is in some cases the only product
observed. The ethyl substituents on the α-positions with respect to the nitrogen atoms
in the case of L4 seems to be steering the regioselectivity to a great extend providing an
excellent regioselectivity of >50. The enantioselectivity is highest for the ligand with
the biggest aryl substituent 1-naphthyl (L5) used in entries 5 and 6. Lowering the
reaction temperature to 40 °C raises the ee above 50 % which remains only a fair result.
74
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
For both substrates styrene and vinyl acetate the chirality-inducing elements seem to be
too far of the metal center, or the backbones are too flexible to force the incoming
alkene in one specific geometry allowing the formation of both enantiomers of the
product. It has to be stressed that only a small excess of bidentate ligand to rhodium is
used (2:1) for both substrates. In order to obtain higher enantiomeric excesses it might
be essential to apply a larger excess of ligand (4 equiv with respect to rhodium). Also at
lower ligand to metal ratios [HRh(CO)4] can be formed.
In order to evaluate and quantify the differences in performance of the
bisaminophosphine ligands in the asymmetric hydroformylation of styrene and vinyl
acetate, HP-NMR and HP-IR could be valuable. These tools give insight into the
coordination mode of the used ligands under (pseudo-)reaction conditions. The
postulation that only one specific coordination mode may be present in the catalyst to
give an efficient enantioselective catalyst can be checked.
4.2.2 HP-NMR
NNPPh2PPh2
NN
PPh2 PPh2
NN
PPh2 PPh2OMeMeO
NN
PPh2 PPh2
NN
PPh2 PPh2
NN
PPh2 PPh2
L1 L2
L3 L4
L5 L6
N
PPh2
N
PPh2
L7 L8
NN
PPh2 PPh2
Figure 3 Bisaminophosphine ligands L1-L8 applied in HP-NMR studies.
75
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
The ligands shown in figure 3 were reacted with one equivalent of the metal precursor
Rh(acac)(CO)2 in toluene-d8. After stirring for 5 minutes the clear yellow solution was
transferred to a 10 mm sapphire NMR tube where it was subjected to 20 bar 1:1 CO:H2
syngas pressure. The tube was subsequently heated to 60°C for 1 h. After this
preformation the pressure was carefully released and the solution was quickly
transferred to a routine 5 mm NMR tube for spectra to be recorded at room
temperature. Selected parameters are presented in Table 3.
Table 3 Selected NMR data on bisaminophosphine / Rh(acac)(CO)2 catalyst system preformed with
CO/H2.
Ligand δ 31P{1H} (ppm) δ 1H (ppm) 1JRh-P (Hz) 1J Rh-H (Hz) 2J P-H (Hz)
L1 / L2 88.6 -9.18 144.0 9.1 13.7 L3 88.1 -9.15 142.8 6.7 15.6 L4 88.8 -9.30 142.8 6.7 13.7 L5 89.3 -9.11 147.7 5.5 6.8 L6 87.4 -9.24 144.0 6.1 11.9 L7 88.0 -8.90 140.4 6.7 16.8 L8 100.1 -9.27 152.6 4.0 -
The 31P{1H} spectra show a doublet at around 88 ppm with a 1JRh-P (Hz) coupling of
around 145 Hz. This seems to indicate that the two phosphorus atoms are magnetically
equivalent and therefore coordinate predominantly in the equatorial - equatorial
coordination mode, although the complexes are dynamic. Figure 4 shows the 31P{1H}-
NMR spectrum of the representative catalyst system based on ligand L7.
ppm
Figure 4 31P{1H}-NMR spectrum of the representative catalyst system based on ligand L7.
76
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
In the hydride region a clear doublet of triplets was visible around –9 ppm for all
ligands with a C2 bridge but L5, vide infra. Figure 5 shows the 1H-NMR spectrum in
the hydride region of the representative ligand L7.
ppm
Figure 5 1H-NMR spectrum of the hydride region of the representative catalyst system based on ligand
L7.
The values of 1J Rh-H range from 5.5 to 9.1 Hz and the 2J P-H coupling constants vary
between 11.9 and 16.8 Hz. These numbers do not give a clear indication weather the
seemingly obvious equatorial - equatorial coordination mode is the major or the only
species in solution, since it is known that the species may interchange on the NMR
timescale giving rise to average coupling constants.
Ligand L5 based on 1-naphthyl has a distinctly different appearance in 1H NMR. From
NMR simulation it could be deduced that the 1J Rh-H and the 2J P-H differ that little in
magnitude (5.5 Hz vs. 6.8 Hz) in this catalyst system that the signal appears as a pseudo
quartet. Also from these coupling constants it is obvious that L5 coordinates
predominantly in the equatorial - equatorial mode. An equatorial - axial relationship
would lead to larger values for the found coupling constants as found for related
diphosphite systems.[18] No indication is found that bridged species (Scheme 1) exist
under these (concentrated) conditions and therefore are also thought to be absent during
the hydroformylation experiments.[19]
77
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
The only measured ligand with a C3 bridge, L8, surprisingly only shows a doublet in
the hydride region, which could imply fast exchange of the phosphorus atoms. The 31P{1H}-NMR spectrum gives a doublet more downfield than for the other ligands in
this series and the 1JRh-P coupling is significantly larger. This could indeed indicate a
larger contribution of ea coordinated species in solution.
To validate the assumption that after releasing pressure the obtained complexes in
solution are stable enough to be measured in the way described above the 1H-NMR
spectrum of the ligand L5 were also measured in the 10 mm sapphire NMR tube after
the same preformation on a Bruker 200 MHz under pressure. Figure 6 shows the
obtained spectrum (hydride region).
-8.4 -8.8 -9.2 -9.6 ppm
Figure 6 1H-NMR spectrum of L5 / Rh(acac)(CO)2 catalyst system under syngas conditions (hydride
region).
The measured chemical shifts and coupling constants were identical to the values
obtained for the system after releasing pressure. We believe that analogously the other
catalyst systems would show equal behavior and therefore that the comparison of
spectral parameters is justified.
The preformation of the catalyst complex is also followed over time at room
temperature under 20 bar syngas (1:1 CO/H2) pressure, see figure 7. The first spectrum
is taken after 21 minutes. After 7h the doublet of triplets pattern is already visible.
Every 7 hours another spectrum is recorded and it shows that after 42 hours the
preformation is complete. Compared to the preformation at 60°C no spectral
differences exist, only the rate of the reaction (full conversion in less than one hour) is
as expected slower. Note that the preformation conditions in the latter experiments are
equal to the conditions applied during the hydroformylation experiments, besides the
higher concentrations for the sake of sensitivity of the spectrometer. This confirms that
78
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
the 1 hour preformation time during regular hydroformylation experiments is likely to
be sufficient for a full transformation to the catalytic resting state.
56 h 49 h 42 h 35 h 28 h 21 h 14 h 7 h 21 min
δ (ppm)
Figure 7 HP-NMR of the L5 / Rh(acac)(CO)2 catalyst system followed over time, hydride region.
4.2.3 HP-IR
For additional information on the coordination behavior close to reaction conditions,
with regard to catalyst concentration and pressure, HP-IR is used. The ligand is
dissolved in cyclohexane and stirred with one equivalent of Rh(acac)(CO)2. The
solution is transferred to an autoclave equipped with IR-transparent windows (ZnS) and
a dedicated FT-IR machine where it is pressurized to 20 bar syngas (1:1 H2/CO) at 60
°C. The observed signals in the carbonyl stretching region are listed in Table 4. It is
apparent that for the measured C2 bridged ligands L2 and L5 there is only one complex
present in solution with equatorial – equatorial coordinated ligand since two bands
with equal intensities are found. This is similar to the findings of van der Vlugt et al.[20]
with sterically constrained diphosphonites, while with Xantphos type ligands mixtures
of ee and ea coordinated species are obtained which are in fast equilibrium, which is
reported by van der Veen et al. [21]
79
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
Table 4 Selected HP-IR data of the νCO on preformed Rh(acac)(CO)2 catalyst systems
Ligand ν1 (cm-1) ν2 (cm-1)
L2 1953 1984 L5 1941 1955 L8 1963 , 1954 , 1942 1983
The C3 bridged ligand L8 shows more bands with different intensities in the IR-
spectrum. This indicates that more than one species are present. This is consistent with
the findings in the HP-NMR experiments. The position of the bands in the spectrum are
indicative for the electron density on the Rh atom in the catalyst complex. The lower
wavenumbers obtained for the ligand L5 would indicate more donation from the
phosphorus atoms to the Rh center than ligand L2.
4.3 Conclusions
The tested bisaminophosphine ligands give active catalysts in the asymmetric
hydroformylation of styrene and vinyl acetate when combined with Rh(acac)(CO)2 as
the metal precursor. Where branched/linear ratio reach excellent values, namely up to
12 for styrene and >50 for vinyl acetate, the enantioselectivities remained low with a
maximum of 51 % for vinyl acetate. HP-NMR experiments show that all ligands
coordinate predominantly in the equatorial-equatorial coordination mode in the trigonal
bipyrimidal resting state of the catalyst. For the 1-naphthyl derived bisaminophosphine
ligand L5 HP-IR proves this is the only coordination mode. However, the ability to
differentiate between stereoisomers apparently remains low.
4.4 Perspective
Considering the requirements for efficient stereocontrol, the obtained low ee makes
sense. First one specific coordination mode of the ligand in the active catalyst has to be
selected, which is equatorial-equatorial in this case applying the bisaminophosphine
ligands. The precondition of coordination mode is now met. Then interactions of the
substituents of the chiral ligand with the substrate (during coordination in the transition
state) have to control the stereoselectivity.
80
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
Due to the fact there is a lot of space in the plane of an ee-coordinated species, the
ligand has to bear very bulky groups for efficient interaction. This is reflected by the
fact that all successful systems relying on an ee-coordinated ligand, reported so far, are
very bulky ligands like the diphosphites described by Babin and Van Leeuwen.[10] As a
consequence, the rigidity and the bulk of the bisaminophosphine ligands should be
increased, preferably based on more structural information on the systems derived from
molecular structures obtained by X-ray diffraction and or computer modeling studies.
For instance the data presented in Chapters 2 and 3 could prove to be valuable.
With the demonstrated ability to control the coordination mode effectively, the
bisaminophosphine ligands have a high potential to reach higher ee’s in asymmetric
hydroformylation, especially with view on their modular construction.
4.5 Acknowledgements
Part of this work has been published (Eric J. Zijp, Jarl Ivar van der Vlugt, Duncan M.
Tooke, Anthony L. Spek and Dieter Vogt, Dalton Transactions, 2005, 512-517).
Avantium Technologies is kindly acknowledged for financial support, Umicor Co. is
thanked for the generous loan of precious metals. Brahim Mezzari is gratefully
acknowledged for technical assistance during the HP-NMR experiments and Pieter
Magusin for discussions on this part of the research. Ruben van Duren is thanked for
the aid on the HP-IR experiments.
4.6 Experimental Section
General
All manipulations were carried out under argon using standard Schlenk techniques.
Syngas (1:1 CO/H2) was bought from Hoekloos. Solvents were either taken HPLC-
grade from an argon-flushed column, packed with aluminum oxide, or distilled under
argon prior to use over an appropriate drying agent. NMR spectra were recorded at
room temperature on a Varian Mercury 400 MHz spectrometer, the HP-NMR spectra
were recorded on a Bruker 200 MHz spectrometer. Chemical shifts are given in ppm
and spectra are referenced to CDCl3 (1H) or 85% H3PO4 (31P{1H}). High-pressure
infrared measurements were performed on a Shimadzu Fourier Transform Infrared
81
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
Spectrophotometer FT-IR 8300 with a Michelson interferometer. The spectra were
processed with Hyper-IR software provided by Shimadzu Corp. An in-house build
autoclave after design by Van Leeuwen et al.22 made from stainless steel (SS 316) with
a volume of 50 mL, equipped with a temperature controller and a pressure transducer
was placed in the spectrophotometer. The infrared beam was led through ZnS windows
(transparent up to 700 cm-1, 10 mm internal diameter) and an effective path length of
0.4 mm. Stirring was performed with a mechanical stirrer equipped with a Rushton-
type stirrer. Gas chromatographic analyses were done on a Shimadzu 17A or a Carlo
Erba (Vega Serie 2) apparatus. The reaction mixtures obtained from the asymmetric
hydroformylation of styrene were analyzed on a 25 m Ultra 2 column (carrier gas 100
kPa N2, FID detector) and for vinyl acetate on a 50 m PONA column (carrier gas 150
kPa He, FID detector). The enantiomeric excess for 2-phenylpropanal was determined
after reduction of the aldehyde and subsequent esterification to the corresponding
trifluoro acetate on a 25 m Lipodex E capillary column (carrier gas 50 kPa H2, FID
detector), 2-acetoxypropanal was analyzed without derivatization on a 50 m FS-
Cyclodex β I/P column (carrier gas 140 kPa He, FID detector).
Hydroformylation experiments
Caution! The hydroformylation experiments are performed with syngas (1:1 = CO/H2)
which is extremely poisonous. Accidents may be lethal. When working with carbon
monoxide a sensitive personal detector should be carried and all experiments are to be
performed in a well ventilated fumehood equipped with a detector, maintaining the CO
concentration in the fumehood below the MAC-value.
Hydroformylation of styrene
Hydroformylation experiments were carried out in home made 75 mL stainless steel
autoclaves, equipped with a glass inner beaker and a magnetic stirrer. The temperature
was controlled by an internal thermocouple. In a typical hydroformylation experiment
the autoclave was heated up to 60 °C and dried under vacuum for 1 h. After cooling the
catalyst precursor (13.1 μmol in 5 mL toluene) is introduced by syringe, rinsing the
Schlenk tube with an additional 5 mL solvent. Likewise the appropriate ligand (26.2
μmol) is added after which the autoclave is purged with syngas, pressurized to 20 bar
and heated to the reaction temperature for the duration of the preformation time (1 h).
82
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
Then a freshly prepared stock solution of styrene ((2.0 mL, 17.4 mmol) filtered over
neutral, activated alumina), internal standard n-decane (1.0 mL, 5.1 mmol) and 5 mL
toluene was added under pressure. The total mixture was allowed to react for 15 h. The
autoclave was then cooled, depressurized and vented with argon. The reaction mixture
was transferred and distilled quantitatively to remove catalyst and excess of ligand. A
sample was analyzed for conversion and regioselectivity. For the ee determination a
part of the mixture was dropped into a suspension of LiAlH4 in Et2O and after 1 h
quenched with water. The mixture was extracted and dried over NaSO4 and evaporated
to dryness under reduced pressure. The residue was dissolved in CH2Cl2 and treated
with 2 equiv of trifluoro acetic acid anhydride. After evaporation to dryness under
reduced pressure a sample of the resulting trifluoro acetate (20 μL) was dissolved in
CH2Cl2 and analyzed by chiral GC.
Hydroformylation of vinyl acetate
A similar procedure as for the hydroformylation of styrene was used. Vinyl acetate (1.5
mL, 16.3 mmol), internal standard ethyl propionate (0.5 g, 4.9 mmol) and benzene as
solvent were applied. After the catalytic experiment the reaction mixture was distilled
under vacuum in order to remove catalyst and excess of ligand. The composition of the
mixture was measured directly by GC without further workup.
HP-NMR experiments
A 10mm outer diameter sapphire NMR tube was filled with a solution of
Rh(acac)(CO)2 (5.0 mg, 19.4 mmol), 1.1 equiv ligand ((21.3 mmol) small excess for
referencing to free ligand) and toluene-d8 (1.5 mL). The tube was purged three times
with syngas and pressurized to 20 bar. The tube was then brought to the desired
temperature and spectra were recorded over time.
HP-IR experiments
The autoclave was flushed with argon for at least an hour. Then a solution of
Rh(acac)(CO)2 (4.9 mg, 19 mmol) and ligand (20 mmol) in 15 mL cyclohexane were
introduced under argon outflow. The equipment was flushed three times with syngas
and afterwards brought to the desired temperature and syngas pressure. Spectra were
recorded after 1 hour preformation.
83
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
4.7 References
[1] D. Selent, K. D. Wiese, D. Rotger, A. Börner, Angew. Chem. Int. Ed. 2000, 39, 1639. [2] C. D. Frohning, C. W. Kohlpaintner, Applied Homogeneous Catalysis with Organometallic
Compounds, Vol 1, Wiley-VCH, Weinhein, 1996, 3. [3] W. A. Herrmann, B. Cornils, Angew. Chem, 1997, 109, 1074. [4] a) B. Cornils, W. A. Herrmann, Ed., Applied Homogeneous Catalysis with Organometallic
Compounds, Vol 1, Wiley-VCH, Weinhein, 2002, 31; b) P. W. N. M. van Leeuwen, C. Claver,
Ed., Rhodium Catalyzed Hydroformylation, Kluwer-CMC, Dordrecht, 2001. [5] a) F. Agbossou, J. -F. Carpentier, A. Mortreux, Chem. Rev. 1995, 95, 2485; b) B. Breit, W.
Seiche, Synthesis, 2001, 1. [6] J. G. de Vries, M. M. H. Lambers-Verstappen, Adv. Synth. Catal, 2003, 345, 478. [7] a) H. Natsugari, Y. Ikeura, I. Kamo, T. Ishimaru, Y. Ischichi, A. Fujishima, T. Tanaka, F.
Kasahara, M. Kawada, T. Doi, J. Med. Chem, 1999, 42, 3982; b) Y. Ikeura, T. Ishimaru, T. Doi,
M. Kawada, A. Fujishima, H. Natsugari, Chem. Commun. 1998, 2141. [8] N. Sakai, S. Mano, K. Nozaki, H. Takaya, J. Am. Chem. Soc. 1993, 115, 7033. [9] R. Ewalds, E. B. Eggeling, C. H. Alison, P. C. J. Kamer, P. W. N. M. van Leeuwen, D. Vogt,
Chem. Eur. J. 2000, 6, 1496. [10] a) J. E. Babin, G. T. Whiteker, WO 93/03830, 1992. b) G. J. H. Buisman, L. A. van der Veen,
A. Klootwijk, W. G. J. de Lange, P. C. J. Kamer, P. W. N. M. van Leeuwen, Organometallics,
1997, 16, 2929. [11] M. Diéguez, O. Pàmies, A. Ruiz, S. Castillón, C. Claver, Chem. Eur. J. 2003, 7, 3086. [12] a) J. K. Stille, H. Su, P. Brechot, G. Parinello, L. S. Hegedus, Organometallics, 1991, 10, 1183;
b) R. van Duren, Platinum Catalyzed Hydroformylation, PhD thesis, Eindhoven University of
Technology, 2004. [13] J. J. Carbó, A. Lledós, D. Vogt, C. Bo, Chem. Eur. J. 2006, 12, 1457. [14] R. Ewalds, Asymmetrische Hydroformylierung mit Phosphor-chiralen Aminophosphin
phosphinit-Liganden, PhD thesis, RWTH Aachen, 1997. [15] a) S. Breeden, M. Wills, J. Org. Chem. 1999, 64, 9735; b) S. Breeden, D. J. Cole-Hamilton, D.
F. Foster, G. J. Schwarz, M. Wills, Angew. Chem. Int. Ed. 2000, 39, 4106; c) J. Ansell, M.
Wills, Chem. Soc. Rev. 2002, 31, 259. [16] T. J. Kwok, D. J. Wink, Organometallics, 1993, 12, 1954. [17] C. F. Hobbs, W. S. Knowles, J. Org. Chem. 1981, 46, 4422. [18] a) C. B. Dieleman, P. C. J. Kamer, J. N. H. Reek, P. W. N. M. van Leeuwen, Helv. Chim. Acta,
2001, 84, 3269; b) G. J. H. Buisman, L. A. van der Veen, P. C. J. Kamer, P. W. N. M. van
Leeuwen, Organometallics, 1997, 16, 5681. [19] A. Castellanos-Páez, S. Castillón, C. Claver, P. W. N. M. van Leeuwen, W. G. J. de Lange,
Organometallics, 1998, 17, 2543. [20] J. I. van der Vlugt, R. Sablong, P. C. M. M. Magusin, A. M. Mills, A. L. Spek, D. Vogt,
Organometallics, 2004, 23, 3177.
84
Chapter 4 Application of Bisaminophosphine Ligands in Rh-Catalyzed Asymmetric Hydroformylation
[21] L. A. van der Veen, P. H. Keeven, G. C. Schoenmaker, J. N. H. Reek, P. C. J. Kamer, P. W. N.
M. van Leeuwen, M. Lutz, A. L. Spek, Organometallics, 2000, 19, 872. [22] A. van Rooy, Rhodium Catalysed Hydroformylation with Bulky Phosphites as Modifying
Ligands, PhD thesis, Universiteit van Amsterdam, 1995.
85
Chapter 5
Phosphonite-Phospholane Ligands Applied in
Rh-Catalyzed Asymmetric Hydroformylation
Mixed phosphonite-phospholane ligands are effective when
applied in Rh-catalyzed asymmetric hydroformylation of
styrene. Branched/linear ratio’s higher than 20 were obtained
while the ee reached a moderate 55%. The dependence of the
ligand performance on pressure, temperature, and ligand
concentration was studied. NMR studies did not reveal the
coordination mode of the ligands in the trigonal bipyramidal
resting state of the catalytic cycle.
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
5.1 Introduction
The discovery and development of new successful ligands for asymmetric
homogeneous catalysts is usually a very tedious process most of the time only governed
by rules of thumb and accompanied by a lot of trial and error. However, a huge variety
of chiral ligands has been developed for certain reactions, e.g. for hydrogenation or
allylic substitution, that have never been tested for other catalytic transformations. The
potential of this existing pool of chiral ligands should not be underestimated although
results might come as a surprise, valuable new insight might be generated. Interesting
results on asymmetric hydroformylation were reported very recently by Abboud et al.
applying phosphacyclic ligands that were developed for asymmetric hydrogenations.[1]
A similar approach was followed in this study, applying phosphonite-phospholane
ligands in the Rh-catalyzed asymmetric hydroformylation of styrene.
In a joint effort of CIBA SC and A. Salzer at RWTH Aachen, a flexible approach to the
synthesis of different families of bidentate phosphorus ligands was followed for
application in asymmetric catalysis. The asymmetric hydrogenation of dehydration acid
derivatives, enamides, and itaconates proceeded with ee values of up to 98.7 %.[2]
Evaluating the structure of the mixed phosphonite-phospholanes (see Figure 1) in that
study revealed attractive features which often proved to be valuable if applied in
asymmetric hydroformylation of alkenes: The ligands consist of a rigid ligand scaffold
with mixed phosphorus functionalities expected to allow for a predominant ea
configuration of the ligand in the trigonal bipyramidal resting state of the catalyst.
Besides this the stereogenic information is close to the phosphorus for the phospholane
part[3] and the generally very effective atropisomeric bisnaphthol[4] is included in the
phosphonite moiety of the ligand.
Those observations prompted us to apply the ligands in the asymmetric
hydroformylation of styrene. Chapter 1 of this thesis contains an introduction to the
field of asymmetric hydroformylation, including description of the successful ligands,
studies into the mechanism of stereoselection and possible applications in real-life
chemistry.
88
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
89
O
OP
S
P
O
OP
S
P
I II Figure 1 Applied BINOLane ligands I and II.
Here we present the successful application of the BINOLane ligands in the asymmetric
hydroformylation of styrene. Investigations of the coordination mode of the ligands in
the trigonal bipyramidal resting state of the catalyst were undertaken.
5.2 Results
5.2.1 Catalysis
Styrene, being a generally accepted and widely used benchmark substrate for the
asymmetric hydroformylation reaction, was selected as the substrate (see Eq. 1).
CO/H2
Rh/L2
CHO
CHO
+*
(1)
The catalysts were prepared for this thorough screening by in situ mixing the ligands
with the metal precursor Rh(acac)(CO)2 in a ratio 2:1 and heating in a AMTEC SPR16
reactor under typical reaction conditions (60°C, 20 bar (1:1 CO/H2)) for 1 hour.
Subsequently the substrate with internal standard was injected and the mixture was
allowed to react under the indicated reaction conditions while measuring the gas
uptake. After workup the product distribution was analyzed by (chiral) GC directly or,
alternatively, after derivatization to the trifluoro acetic ester. Both methods gave
identical values for the ee. The results of the first screening are presented below (Table
1).
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
Table 1 Selected results of initial ligand screening under standard conditions.a
Ligand Conversion b b/l c ee (%) d I 100 21 31 (S) II 27 21 37 (R)
a Reaction conditions: T = 60 °C; p = 20 bar (1:1 CO/H2); solvent: 4 mL toluene; 2 mL styrene, 1 mL
decane internal standard; 14.3 μmol Rh(acac)(CO)2; L:Rh = 2; preformation t = 1h; reaction t = 24h. b
Percent conversion of styrene after 24h c Branched / linear ratio. d Enantiomeric excess determined by
chiral GC.
The catalyst based on BINOLane I, with the (R) enantiomer of BINOL in the
phosphonite moiety, gives full conversion of styrene in the applied time. The
branched/linear ratio is excellent with 21 and a reasonable ee of 31% is reached. With
BINOLane II, with an opposite configuration of the BINOL part of the ligand, the
conversion after 24 hours is still low and the branched/linear ratio is also 21. The
obtained branched/linear ratios can compete with the numbers obtained for settled
hybrid ligands like AMPP (20-40) [5] and BINAPHOS (7-12).[4a] The ee values of the
styrene hydroformylation products induced by the BINOLane ligands however, are
moderate compared to the renowned ligands AMPP (46%-75%)and BINAPHOS (85%-
94%).
It seems that the absolute configuration of the BINOL moiety determines the absolute
configuration of the major product. The stereogenic centers of the phospholane ring
form an inefficient matched or miss-matched pair in terms of enantioselectivity.
To determine the stability of the catalyst we followed the reaction over time by taking
samples on predetermined times, while measuring the gas uptake. The samples were
analyzed on b/l ratio and enantioselectivity (see Table 2 for details and Figure 2 for a
graphic representation).
90
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
91
0 5 10 15
0
50
100
150
200
250
Gas
upt
ake
(mL)
0
10
30
40
20
Gas uptake ee (R) b/l
Table 2 b/l ratio’s and ee’s followed over time for the respective ligands (R,R)-Me-BINOLane (I) and
(S,R)-Me-BINOLane (II).a
time (h) b/l (I) b ee (S) (I) (%) c b/l (II) b ee (R) (II) (%) c 0,37 22,8 32 19,2 38 0,71 23,5 32 19,6 36 1,04 23,9 32 19,6 35 1,87 24,6 31 20,4 36 3,21 24,8 33 20,2 37 5,21 24,3 33 20,9 36 9,55 23,0 33 21,5 37 15,3 22,6 33 21,8 38
a Reaction conditions: T = 60 °C; p = 20 bar (1:1 CO/H2); solvent: 4 mL toluene; 2 mL styrene, 1 mL
decane internal standard; 14.3 μmol Rh(acac)(CO)2; L:Rh = 2; preformation t = 1h; reaction t = 16h; 100
μL samples taken over time b Branched / linear ratio c Enantiomeric excess determined by chiral GC.
Figure 2a) upper and 2b) lower. Gas uptake, b/l ratio and ee followed over time for the respective
ligands (R,R)-Me-BINOLane (I) and (S,R)-Me-BINOLane (II).
0 5 10 150
100
200
300
400
500
600
700 Gas uptake ee (S) b/l
Time (h)
Gas
upt
ake
(mL)
0
25
50
75
100
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
For both ligands I and II the ee’s are virtually constant. The b/l ratio’s seem to increase
slowly over time, maybe dropping marginally when the conversion is full (at extended
reaction times for ligand I). This indicates that the catalyst is stable over time, but the
preformation time could be extended (or the reaction conditions during preformation
intensified) to ensure full conversion to the catalyst resting state prior to substrate
injection.
To assess the dependencies of the performance of the system on applied pressure,
temperature, and stoichiometry these parameters were systematically varied.
Firstly the applied pressure was varied from 10-40 bar syngas (1:1 CO/H2). The
obtained data are gathered in Table 3 and 4.
Table 3 b/l ratio and ee obtained for different applied pressures by using ligand (R,R)-Me-BINOLane (I).a
bar conversion (%) b b/l c ee (S) d 10 100 18 26 20 100 19 30 30 99 24 33 40 99 25 32
a Reaction conditions: T = 60 °C; p = x bar (1:1 CO/H2); solvent: 4 mL toluene; 2 mL styrene, 1 mL
decane internal standard; 14.3 μmol Rh(acac)(CO)2; I:Rh = 2; preformation t = 1h; reaction t = 24h; b
Percent conversion of styrene after 24h c Branched / linear ratio d Enantiomeric excess determined by
chiral GC. Table 4 b/l ratio and ee obtained for different applied pressures by using ligand (S,R)-Me-BINOLane (II).a
bar conversion (%) b b/l c ee (R) d 10 13 18 36 20 27 21 36 30 32 22 36 40 36 22 37
a Reaction conditions: T = 60 °C; p = x bar (1:1 CO/H2); solvent: 4 mL toluene; 2 mL styrene, 1 mL
decane internal standard; 14.3 μmol Rh(acac)(CO)2; II:Rh = 2; preformation t = 1h; reaction t = 24h; b
Percent conversion of styrene after 24h c Branched / linear ratio d Enantiomeric excess determined by
chiral GC.
92
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
93
0 5 10 15 200
100
200
300
400
500
10 bar 20 bar 30 bar 40 bar
Time (h)
Gas
Upt
ake
(mL)
0
10
20
30
Conversion (%
)
Figure 3 Gas uptake followed over time for different applied pressures by using ligand (II).
For increasing syngas pressure the conversion, the ee, and the b/l ratio all seem to go
up, albeit with small numbers. This is in contrast with the AMPP ligands where a
negative influence of the pressure on the enantioselectivity was observed.[5] Coworkers
of DOW Pharma found for a range of ligands applied for styrene, allyl cyanide and
vinyl acetate that all enantioselectivities were unaffected by changing pressure. The
regioselectivity for the styrene increased with increasing pressure, where the vinyl
acetate products were obtained with a lower regioselectivity.[6]
The second investigated parameter was the reaction temperature which was varied from
25-120 ºC. Tables 5 and 6 show the selected results. During these investigations the
applied temperature during the 1h catalyst preformation was constant with 60 ºC but the
reaction times varied.
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
Table 5 b/l ratio and ee obtained for different reaction temperatures applying ligand (R,R)-Me-
BINOLane (I). a
Temperature (°C) b/l b ee (S) (%) c 25 32 32 40 32 31 60 21 31 80 12 21 100 5 3 120 3 0
a 14.3 μmol Rh(acac)(CO)2, 2 equiv I, 4 mL toluene, 2 ml styrene, 1 mL decane, p = 20 bar CO:H2 (1:1),
1h preformation b Branched / linear ratio c Enantiomeric excess determined by chiral GC.
Table 6 b/l ratio and ee obtained for different reaction temperatures applying ligand (S,R)-Me-
BINOLane (II). a
Temperature (°C) b/l b ee (R) (%) c 25 29 55 40 28 52 60 21 37 80 14 28 100 6 8 120 4 2
a 14.3 μmol Rh(acac)(CO)2, 2 equiv II, 4 mL toluene, 2 ml styrene, 1 mL decane, p = 20 bar CO:H2 (1:1),
1h preformation b Branched / linear ratio c Enantiomeric excess determined by chiral GC.
At lower temperature the selectivity of the reaction, both in terms of b/l ratio as ee were
maximum, reaching 97% selectivity of the branched product for ligand I and more than
50% enantiomeric excess for ligand II. Logically the rate of reaction is lowest in these
cases. At higher reaction temperature the undesired polymerization of styrene plays a
significant role, besides a higher degree of degradation of the catalyst under these harsh
conditions.
The last parameter that was varied was the Rh/L(I) ratio. Table 7 gives the relevant
numbers and a graphic representation is shown in Figure 4.
94
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
95
Table 7 Conversion, b/l ratio and ee obtained for several stoichiometries of ligand (R,R)-Me-BINOLane
(I) to Rh. a
x equiv ligand (I) Conversion (%) b/l ee (S) (%) 0 100 2 0 1 100 8 12 2 31 20 22 3 12 22 29 4 0 - - 5 0 - -
a 14.3 μmol Rh(acac)(CO)2, x equiv (I), 4 mL toluene, 2 ml styrene, 1 mL decane, T = 60 °C, p = 20 bar
CO:H2 (1:1), 1h preformation, 23h reaction. b Percent conversion of styrene after 23h c Branched / linear
ratio c Enantiomeric excess determined by chiral GC.
0 5 10 15 200
250
500
750
1000 1 equiv 2 equiv 3 equiv 4 equiv
Time (h)
Gas
Upt
ake
(mL)
Figure 4 Gas uptake followed over time for selected stoichiometries of ligand (R,R)-Me-BINOLane (I)
to Rh.
It can be seen that the fastest reactions are observed with the smallest amounts of
ligand. The metal precursor Rh(acac)(CO)2 without additional ligand is under the
applied conditions active as hydroformylation catalyst, albeit with very low selectivity
towards the desired branched product. Also with one equivalent ligand the rate is high
but now the selectivity rises fast. With 2 equivalents of ligand the optimal amount is
reached. The full potential of the ligand in steering the reaction is used and at higher
ratio’s of ligand to rhodium the reaction becomes too slow.
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
PCO
Rh CO
HOC
P
Rh CO
H
CO
OC
OC
P^P
CO CO
a b c
PP
Rh CO
HOC
Figure 5 Complex equilibria dependent on ligand concentration.
The reasons for this are the equilibria between non-coordinated, semi-coordinated and
coordinated complexes (Figure 5). If a larger amount of available Rh is present as
complexes a or b the reaction would become less selective. A higher concentration of
ligand would shift the equilibria more to the most selective coordinated complex c. An
even larger amount of ligand could force the formation of (inactive) complexes with
more than one coordinated ligand, which is undesirable.
A higher ratio of ligand to rhodium not always affects the catalysis in a positive way.
For instance sugar-derived diphosphite ligands are normally employed in a 1:1 ratio
and an increase does not affect the catalysis much at all.[7]
5.2.2 NMR Studies
Two small tests were done to check the coordination of the ligand to Pd and Rh. When
equimolar amounts of (R,R)-Me-BINOLane (I) were reacted with either Pd(cod)Cl2 or
[Rh(cod)2]BF4 in CDCl3 instantaneous reactions were observed. Obtained NMR-data
can be found in Table 8.
Table 8 NMR-data of metal-complexes of (R,R)-Me-BINOLane (I).
Metal precursor δ (ppm) phospholane
δ (ppm) phosphonite
2J(P-P) (Hz) 1J(P-M)
Pd(cod)Cl2 80.8 140.8 16.7 - [Rh(cod)2]BF4 65.0 163.9 34.6 148.3 - 233.5
The obtained values suggest a mononuclear cis-coordination for both complexes, as is
found for a related cationic complex described by the coworkers of CIBA SC (Figure 6,
R = ethyl).[2a]
96
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
97
S
PR R
P
R
R
Figure 6 Bisphospholane ligand by CIBA.
In a comparable manner as described in Chapter 4 for the bisaminophosphine ligands
the coordination behavior of the here used Me-BINOLane ligands was investigated to
reveal the structure of the trigonal bipyramidal resting state of the hydroformylation
catalyst. Equimolar amounts of ligand and Rh(acac)(CO)2 were dissolved in deuterated
toluene in a 10mm sapphire NMR tube and pressurized to 20 bar of syngas. After 1
hour preformation time at elevated temperature 60°C the tube was cooled to room
temperature and measured on a Bruker 200 MHz NMR machine.
Under the applied conditions however, no signal in the hydride region was obtained.
Strategies to lengthen relaxation times, to widen the spectral width, to increase
concentrations and to prolong measuring times did not result in any information on the
coordination behavior. Maybe at room temperature the coalescence is reached for a
fluxional process in the complexes, thus resulting in a non-appearing signal. No
variation in temperature was attempted.
Comparing the structure of the ligands to a bisphospholane ligand based on the same
benzo[b]thiophene scaffold (figure 6, with R = ethyl) where a bite angle of 85° P-Rh-P
is found in a cationic Rh complex the expected coordination mode in the trigonal
bipyramidal resting state of the catalyst would be equatorial-axial. In this coordination
mode the ideal 90° angle is closely resembled. The different electronic properties of the
phospholane and the phosphonite part present in the ligand used in our study could
ensure a preferential coordination and thus the respectable enantioselectivities achieved
in this study.
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
5.3 Conclusions
The Me-BINOLane ligands form active catalysts for the Rh-catalyzed asymmetric
hydroformylation of styrene and the regioselectivity for the branched product is high
with b/l ratio’s over 20. The ee’s obtained depend mostly on the atropisomeric element
in the phosphonite part of the ligand and reach values just over 50%. Application of 2
equivalents of ligand with respect to rhodium is most efficient in a compromise
between activity and enantioselectivity.
5.4 Perspective
The ee’s obtained in this study for the asymmetric hydroformylation of styrene can not
compete with the numbers obtained with a variety of other ligands reported elsewhere.
Since the coordination mode of the ligands under reaction conditions was not disclosed
it remains speculation if preferential coordination is reached. The synthetic strategy
however allows for independent variation of the ligands in both the phospholane as the
phosphonite part. Substitution on the 2- and 2’-positions of the used bisnaphthol often
creates a more stereoselective ligand, as could be the use of ethyl- or propyl-
substituents on the phospholane ring. Other substrates like vinyl acetate or allyl cyanide
could be used to check the efficacy of the ligands in their enantioselective conversion.
The presence of the sulfur heteroatom may act as a possibility to electronically modify
the backbone (e.g. by coordination to early transition metals) or as a means to anchor
the ligands to a support, thus allowing for recycling of the ligand (or catalyst).
5.5 Acknowledgements
Avantium Technologies is kindly acknowledged for financial support, Umicor Co. is
thanked for the generous loan of precious metals. CIBA SC is thanked for putting a
sample of the used ligands to our disposal and we are especially grateful to Ulrich
Berens for detailed help on the synthetic procedures. Leandra Cornelissen is
acknowledged for the blood, sweat and tears shed during the syntheses of the ligands.
Ton Staring is gratefully acknowledged for his skillful technical assistance during the
98
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
99
chromatographic analyses. And finally Christian Müller is thanked for his kind help
during the AMTEC runs.
5.6 Experimental Section
General
All manipulations were carried out under argon using standard Schlenk techniques.
Chemicals were purchased from Acros, Aldrich, Lancaster or VWR and used as
received or distilled from CaH2 before use. Syngas (CO:H2 (1:1)) was bought from
Hoekloos. Solvents were either taken HPLC-grade from an argon-flushed column,
packed with aluminum oxide, or distilled under argon prior to use over an appropriate
drying agent. The NMR spectra were recorded on a Bruker 200 MHz spectrometer. Gas
chromatographic analyses were done on a Shimadzu 17A or a Carlo Erba (Vega Serie
2) apparatus. The reaction mixtures obtained from the asymmetric hydroformylation of
styrene were analyzed on a 25 m Ultra 2 column (carrier gas 100 kPa N2, FID detector).
The enantiomeric excess in the product 2-phenylpropanal was determined after
reduction of the aldehyde and subsequent esterification to the corresponding trifluoro
acetate on a 25 m Lipodex E capillary column (carrier gas 50 kPa H2, FID detector), or
without derivatization on a Supelco Betadex column (carrier gas 60 kPa He, FID
detector).
Hydroformylation experiments
Caution! The hydroformylation experiments are performed with syngas (1:1 = CO/H2)
which is extremely poisonous. Accidents may be lethal. When working with carbon
monoxide a sensitive personal detector should be carried and all experiments are to be
performed in a well ventilated fume hood equipped with a detector, maintaining the CO
concentration in the fume hood below the MAC-value.
Hydroformylation of styrene
Hydroformylation experiments were carried out in an AMTEC SPR16 machine. Before
use the reactors were heated up to 60 °C and dried under vacuum for 1 h. After cooling
the catalyst is introduced by syringe after which the autoclaves are purged with syngas,
pressurized to 20 bar and heated to the reaction temperature for the duration of the
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
preformation time (1 h). After cooling and lowering the pressure a freshly prepared
stock solution of styrene ((2.0 mL, 17.4 mmol) and internal standard n-decane (1.0 mL,
5.1 mmol) was added. The mixture was then brought to the desired temperature and
pressure and it was allowed to react. The autoclaves were then cooled, depressurized
and vented with argon. From the contents a sample was analyzed for conversion and
regioselectivity on a Ultra column. For the ee determination 1 mL of the mixture was
dropped into a suspension of 150 mg LiAlH4 in Et2O and after 1 h quenched with
water. The mixture was extracted and dried over NaSO4 and evaporated to dryness
under reduced pressure. The residue was dissolved in CH2Cl2 and treated with 0.5 mL
of trifluoro acetic anhydride. After evaporation to dryness under reduced pressure a
sample of the resulting trifluoro acetate (20 μL) was dissolved in CH2Cl2 and analyzed
on a Lipodex column. Alternatively the ee could be determined directly on a Supelco
Betadex column.
NMR experiments
A 10mm outer diameter sapphire NMR tube was filled with a solution of
Rh(acac)(CO)2 (5.0 mg, 19.4 mmol), 1.1 equiv ligand (21.3 mmol, small excess for
referencing to free ligand) in toluene-d8 (1.5 mL). The tube was purged three times with
syngas and pressurized to 20 bar. The tube was then brought to the desired temperature.
After 1 hour the pressure was released and the contents were quickly transferred to a 5
mm NMR tube and directly measured or analyzed on a Bruker 200 MHz NMR machine
without prior release of pressure.
100
Chapter 5 Phosphonite-Phospholane Ligands Applied in Rh-Catalyzed Asymmetric Hydroformylation
101
5.7 References
[1] A. T. Axtell, J. Klosin, K. A. Abboud, Organometallics, 2006, 25, 5003. [2] a) U. Berens, U. Englert, S. Gwyser, J. Runsink, A. Salzer, Eur. J. Org. Chem. 2006, 2100; b)
U. Berens to Solvias A.G., WO 03/031456 A2. [3] e.g. Me-DuPhos; M. J. Burk, J. E. Feaster, W. A. Nugent, R. L. Harlow, J. Am. Chem. Soc.
1993, 115, 10125.
[4] For successful applications of the bisnaphthol unit in asymmetric hydroformylation see e.g.
(R,S)-BINAPHOS a) N. Sakai, S. Mano, K. Nozaki, H. Takaya, J. Am. Chem. Soc. 1993, 115,
7033 or (R,S)-Yanphos; b) Y. Yan, X. Zhang, J. Am. Chem. Soc. 2006, 128, 7198. [5] R. Ewalds, E. B. Eggeling, C. H. Alison, P. C. J. Kamer, P. W. N. M. van Leeuwen, D. Vogt,
Chem. Eur. J. 2000, 6, 1496. [6] A. T. Axtell, C. J. Cobley, J. Klosin, G. T. Whiteker, A. Zanotti-Gerosa, K. A. Abboud, Angew.
Chem. Int. Ed. 2005, 44, 5834. [7] M. Diéguez, O. Pàmies, A. Ruiz, S. Castillón, C. Claver, Chem. Commun., 2000, 1607.
Summary
Summary
Investigations on Rhodium-Catalyzed Asymmetric Hydroformylation
In asymmetric metal catalysis it is of special value to have chiral ligand classes
available which allow for a straight forward and highly modular, maybe even
automated synthesis and variability of the molecular structure. From an academic point
of view this is important in order to study in detail the structure-performance relations
in order to generate basic understanding of stereoselection mechanisms and ultimately
derive at a rational design of new catalysts for a given synthetic problem. From an
industrial point of view readily available ligand libraries allow for a rapid sceening and
optimization of a catalyst for a given substrate, as especially in fine chemicals business
the given time for development is extremely short.
Theoretical insight, next to structure-performance relations, is obtained by studying the
coordination behavior of the ligands in catalytically active species by means of
structural analysis and in situ spectroscopic investigations. This can provide valuable
data for further theoretical studies on a high level.
Chapter one gives an introduction in asymmetric hydroformylation. Starting with a
historical overview and ending with the state-of-the-art ligand systems that give the
currently most active and selective catalysts. High-Throughput-Experimentation,
theoretical investigations, and spectroscopic studies are identified as the important
elements leading to success.
In Chapter two the versatile modular synthesis of novel symmetrically and non-
symmetrically substituted bisaminophosphine ligands is described. Molecular structures
of the ligands and complexes thereof revealed a trigonal planar geometry of the
nitrogen atoms bound to the phosphorus donor atom, resulting from a significant
contribution of π-bonding to the P-N bond.
103
Summary
DFT calculations were performed on model compounds for bisaminophosphine ligands
to analyze the geometries and charge distributions, which are discussed in Chapter
three. The computed structure of a simplified cis-Pd complex of a bidentate
bisaminophosphine ligand gives valuable information on the coordination behavior.
Application of catalysts generated in situ from [Rh(cod)2]BF4 and bisaminophosphines
in the asymmetric hydrogenation of methyl (Z)-N-acetylaminocinnamate gave ee’s of
up to 91%. The contributions to stereoselection of individual aminophosphine moieties
are recognized.
Chapter four shows that the bisaminophosphine ligands form effective catalysts in the
Rh-catalyzed asymmetric hydroformylation of prochiral alkenes. The regioselectivities
for styrene and vinyl acetate were very good, while the enantioselectivities however
stay low with 12% and 51% respectively. HP-NMR studies indicated that equatorial -
equatorial coordination mode in the catalytic resting state is preferred for these ligands,
as was confirmed by HP-IR spectroscopy.
Finally in Chapter five mixed phosphonite-phospholane ligands are presented. They
are effective when applied in the Rh-catalyzed asymmetric hydroformylation of
styrene. Branched/linear ratio’s higher than 20 were obtained and the ee reached a
moderate 55%. NMR studies did not reveal the coordination mode of the ligands in the
trigonal bipyrimidal resting state of the catalytic cycle. The dependency of the catalyst
performance on the parameters temperature, pressure and L/Rh ratio were determined.
104
Samenvatting
Samenvatting
Onderzoek in Rhodium-Gekatalyseerde Asymmetrische Hydroformylering
In asymmetrische homogene metaalkatalyse is het van extreem belang om klassen van
chirale liganden beschikbaar te hebben die eenvoudig en modulair opgebouwd zijn en
wellicht zelfs geautomatiseerd gesynthetiseerd kunnen worden. Voor de academische
wereld is dit belangrijk om gedetailleerd structuur-prestatie relaties te bestuderen om
kennis te vergaren over de mechanismen van stereoselectie om uiteindelijk uit te komen
bij het rationele ontwerp van nieuwe katalysatoren voor een bepaald synthetisch
probleem. Vanuit het oogpunt van de chemische industrie zullen beschikbare
databanken aan liganden bijdragen aan snelle screening en optimalisatie van
katalysatoren voor een gegeven substraat, wat belangrijk is aangezien vooral in de fijn-
chemische industrie de beschikbare tijd voor ontwikkeling erg kort is.
Theoretisch inzicht, naast structuur-prestatie relaties, wordt verkregen door het
coördinatie-gedrag van liganden in de katalytisch actieve deeltjes te bestuderen door
structuur analyses en in situ spectroscopische technieken. Dit kan belangrijke data
opleveren voor verdere theoretische studies op een hoog niveau.
Hoofdstuk één geeft een introductie over asymmetrische hydroformylering, startend
met een historisch overzicht en eindigend met de toonaangevende ligandsystemen die
op dit moment de meest actieve en selectieve katalysatoren vormen. High-Throughput-
Experimentation, theoretische beschouwingen en spectroscopische onderzoeken
worden genoemd als belangrijke factoren die hierbij tot succes kunnen leiden.
In Hoofdstuk twee is de modulaire synthese van nieuwe symmetrisch en niet-
symmetrisch gesubstitueerde bisaminofosfine liganden beschreven. Kristalstructuren
van liganden en complexen daarvan lieten de trigonaal planaire geometrie zien van de
stikstof atomen die gebonden zijn aan de fosfor donoratomen. Dit is het gevolg van een
substantiële bijdrage van een π-binding aan de P-N binding.
105
Samenvatting
Er zijn Density Functional Theory berekeningen op model verbindingen voor
bisaminofosfine liganden uitgevoerd om geometrieën en ladingverdelingen te
analyseren zijn uitgevoerd. Dezen worden in Hoofdstuk drie gepresenteerd. De
berekende structuur van een gesimplificeerd cis-Pd complex van een bidentate
bisaminofosfine ligand gaf belangrijke informatie over het coördinatiegedrag. De
toepassing van katalysatoren in-situ gegenereerd van [Rh(cod)2]BF4 en
bisaminophosphines in de asymmetrische hydrogenering van methyl (Z)-N-
acetylaminocinnamaat gaf ee’s tot 91%. De bijdragen van de individuele aminofosfine
delen aan de stereoselectie zijn onderkend.
Hoofdstuk vier laat zien dat bisaminofosfine liganden effectieve katalysatoren vormen
in de Rh-gekatalyseerde asymmetrische hydroformylering van prochirale alkenen. De
regioselectiviteiten van zowel styreen als vinyl acetaat waren erg goed, de ee’s bleven
echter laag met respectievelijk 12% en 51%. NMR studies onder verhoogde druk gaven
aan dat equatoriaal – equatoriaal de geprefereerde coördinatie vormt voor de liganden,
wat bevestigd werd door hoge druk IR-spectroscopie.
Tenslotte worden in Hoofdstuk vijf gemixte fosfoniet-fosfolaan liganden
gepresenteerd. Ze zijn effectief wanneer ze toegepast worden in de Rh-gekatalyseerde
asymmetrische hydroformylering van styreen. Iso/n ratio’s hoger dan 20 werden
behaald en de ee bereikte een matige 55%. NMR studies konden de coördinatie van de
liganden in de trigonaal bipyrimidale slapende toestand van de katalysator in de
katalytische cyclus niet verhelderen. De afhankelijkheid van de prestaties van het
systeem van de temperatuur, de druk en de ligand tot rhodium verhouding zijn bepaald.
106
Dankwoord
Dankwoord
Een proefschrift moet je in je eentje afronden, maar de inhoud ervan komt tot stand
dankzij de inbreng van velen. Hier is een goede gelegenheid om al die personen
nogmaals te bedanken.
Je moet niet starten met bouwen zonder een goed fundament. Dat is gelegd op de UvA
waar ik mijn studie Scheikunde heb afgesloten in de groep van Kees Elsevier. In die
onderzoeksgroep werd ik gegrepen door de sfeer; de sfeer van fundamentele
wetenschap en de sfeer tussen de aio’s onderling waar je moeiteloos in opgenomen
werd. Met name wil ik Jeroen en Boke noemen als personen waar ik veel van geleerd
heb.
Aangestoken door het virus begon ik als promovendus in de homogene katalyse in de
groep van Dieter Vogt. Beste Dieter, onze onderlinge communicatie is niet altijd
vlekkeloos verlopen. Maar ik denk dat ik voor beiden spreek als ik zeg dat we altijd het
beste met elkaar voor hadden en met het onderzoek wat moest leiden tot dit
proefschrift. Geduld is een schone zaak gebleken. Copromoter Erik, je bent een man
met een on-conventionele manier van begeleiden en aanpak. Dat is je grote kracht maar
laat het niet je valkuil vormen! Ook de overige leden van mijn promotiecommissie wil
ik hierbij danken voor hun input.
De sfeer in de groep was ongeëvenaard. Iedereen had zo zijn of haar unieke bijdrage
aan het geheel.
Jarl, senior aio, work-alholic 1e klas, altijd goed voor serieuze gesprekken en culturele
of sociale uitjes: zonder jou zou dit nooit gelukt zijn... Als ik één iemand ken die ik
succes gun in zijn carrière ben jij het, en het gaat je nog lukken ook!
Gaby, je klaagzang was meestal duidelijk hoorbaar, maar je hebt het ook niet
gemakkelijk gehad.
107
Dankwoord
Niek, mijn gewaardeerde kamergenoot, door schade en schande heb ik van je kunnen
leren. Concerten, films, borrels; we gingen echt overal heen. Ook na de TU blijf ik je
achtervolgen; je bent als collega niet weg te denken.
Ruben, mijn andere gewaardeerde kamergenoot, was altijd hoorbaar vrolijk aanwezig.
Je bent een man van staal en je organisatietalent en gastvrijheid kent geen weerga.
Oeteldonk is niet hetzelfde zonder jou!
Michiel, samen hebben we wel aan onze onderwijstaken voldaan, het was een
genoegen! Dankzij de gezamelijke sappel- en karnemelksessies bleven onze vitamine C
en calcium gehaltes goed op peil.
Mabel, ik vergeet ons tripje naar Venetië nooit meer en zonder jou had ik de Spaanse
filmcultuur nooit ontdekt.
Jos, man zonder compromissen, veel daden en weinig woorden: respect!
Gijs, je bent een persoon vol eigen-aardigheden, een lopend Handbook of Chemistry
and Physics. Zorg dat je dat boekje afschrijft...
Kathi en Laura, de knappe dames in STW 4.36; jullie hebben een speciaal plekje in
mijn hart veroverd, ik zal jullie voortgang op de voet blijven volgen en bewaken.
Ook van de oude Silly’s (Mark, Tessa en Rob), de overige jongere garde (Leandra,
Bart, Patrick en Michèle), delegaties aan Post-Docs (met name Sam, Rafaël, Marije,
Marco, Daniël en Vincent) en studenten (met name afstudeerder Bart, Erasmus-stagaire
Roser, Jeroen v. B., Jos en Saskia) kon ik veel leren of kon ik kennis aan overdragen.
Christian als goedlachse hardwerkende UD was tenslotte onmisbaar. Allen bedankt!
Daarnaast hebben vele andere SKA-ers de tijd onvergetelijk gemaakt bij lunches,
koffiepauzes, de borrel, de FORT, film, Tourpoul, Niok Soccer Cup, etc. Ik denk met
name aan Bouke, Chrétien, Joost, Sander, Thijs en Tiny. Ook overige OBP-ers (oa vele
secretaresses, good-old Ton Staring, vrolijke Wout(er) van Herpen en de mentale
ondersteuning van Wilma en vooral Annemieke) bleken vaak onmisbaar.
108
Dankwoord
Er was ook een leven naast de TU. Vele, vele uren heb ik doorgebracht bij Unicum, met
afstand de gezelligste tafeltennisvereniging van het land: veelal om te tafeltennissen (ik
bedank mijn teamleden Bert, Erik en Truus voor het rondcrossen met mij in de regio)
maar ook zeker voor de gezelligheid na vrij spelen, competitie of bij het 1e team, het
landskampioenschap van de jongens in Middelburg en vele feestavonden waren
onvergetelijk. Het was maar goed dat het op loopafstand van huis was!
Dat huis was mijn kamer op het Villapark, dat ik al die jaren trouw ben gebleven. Dat
kan ook niet anders met een hospita als mevrouw Vermulst. U gunde me alle privacy en
vrijheid gunt in uw huis en tuin, naast goede gesprekken met een kop koffie en altijd
een koek.
Ik begon over een fundament. Het echte fundament is uiteraard gelegd door mijn
ouders. Ook oma en grote broer Robin zijn als familie altijd dichtbij. Dankzij jullie
allen, door een goede opvoeding en onvoorwaardelijke steun denk ik een integer
persoon te zijn geworden, met een goede set normen en waarden. En is dat naast liefde
niet wat feitelijk het belangrijkste is in het leven?
Bij liefde denk ik meteen aan mijn lieve Henrike. Je hebt veel geduld moeten
opbrengen, maar uiteindelijk is er echt een einde aan gekomen. Nu kunnen we verder
bouwen aan onze toekomst. Zullen we samen proberen ervoor te zorgen dat de beren
weggaan?
Terugkijkend: ik heb nergens spijt van, maar sommige zaken zou ik nu anders (beter?)
aanpakken. Zou ik weer gaan promoveren? Jazeker!!
Eric
109
Curriculum Vitae
Curriculum Vitae
Eric Zijp werd op 24 november 1976 in Sleeuwijk geboren. Op
OSG de Meergronden te Almere behaalde hij in 1995 zijn VWO
diploma, waarna hij aan de Universiteit van Amsterdam begon
aan de studie Scheikunde. In de groep Anorganische Chemie van
Prof. dr. Kees Vrieze en Prof. dr. Kees Elsevier werden onder
supervisie van Dr. Jeroen Diederen palladium gekatalyseerde
ringsluitingen bestudeerd. In dezelfde groep werden tevens oxidatieve addities aan
Rh(I)-Terpy*-complexen onderzocht met als coach Dr. Boke de Pater. In 2000
behaalde hij zijn diploma en begon als promovendus in de groep Homogene Katalyse
en Coördinatie Chemie onder leiding van Prof. dr. Dieter Vogt aan de Technische
Universiteit Eindhoven, wat uiteindelijk resulteerde in dit proefschrift. Een jaar
werkte hij gedetacheerd vanuit Yacht op projectbasis bij DSM, afdeling LS-ASC&D.
Momenteel is hij werkzaam als Chemist bij ChemShop in Weert.
Eric Zijp was born on the 24th of November 1976 in Sleeuwijk, the Netherlands. He
graduated from OSG de Meergronden in Almere in 1995 and shortly after started his
chemistry studies at the University of Amsterdam. In the group Inorganic Chemistry
of Prof. dr. Kees Vrieze and Prof. dr. Kees Elsevier he studied under guidance of Dr.
Jeroen Diederen on palladium catalyzed ring-annulations. In the same group oxidative
additions to Rh(I)-Terpy*-complexes were investigated, herein coached by Dr. Boke
de Pater. In 2000 he received his diploma and started as PhD-student in the group
Homogeneous Catalysis and Coordination Chemistry of Prof. dr. Dieter Vogt at the
Eindhoven University of Technology ultimately resulting in this thesis. For one year
he worked for Yacht at DSM, LS-ASC&D department. Currently he is employed as
Chemist at ChemShop in Weert.
111