Solid-phase parallel synthesis of phosphite ligands

Article abstract of DOI:10.1021/ol703070x

Various routes for the synthesis of polymer-bound phosphites and phosphoramidites have been investigated. In the presence of a suitable activator the supported phosphoramidites react cleanly with alcohols to give the corresponding monodentate phosphite ligands in solution. We have applied this novel solid-phase route in the parallel synthesis of several monodentate chiral and achiral phosphite ligands.

Full text of DOI:10.1021/ol703070x

ORGANIC
LETTERS
2008
Vol. 10, No. 5
989-992
Solid-Phase Parallel Synthesis of
Phosphite Ligands
Bert H. G. Swennenhuis, Ruifang Chen, Piet W. N. M. van Leeuwen,
Johannes G. de Vries, and Paul C. J. Kamer*
Van ‘t Hoff Institute of Molecular Sciences, UniVersity of Amsterdam,
The Netherlands, DSM Pharmaceutical Products-AdVanced Synthesis, Catalysis &
DeVelopment, P.O. Box 18, 6160 MD Geleen, The Netherlands, and EaStCHEM,
School of Chemistry, UniVersity of St. Andrews, St. Andrews, Fife, Scotland KY16 9ST
pcjk@st-andrews.ac.uk
Received December 20, 2007
ABSTRACT
Various routes for the synthesis of polymer-bound phosphites and phosphoramidites have been investigated. In the presence of a suitable
activator the supported phosphoramidites react cleanly with alcohols to give the corresponding monodentate phosphite ligands in solution.
We have applied this novel solid-phase route in the parallel synthesis of several monodentate chiral and achiral phosphite ligands.
In the past decade (chiral) phosphite and phosphoramidite
ligands have found wide applications in asymmetric cataly-
sis.1,2 In several transition-metal-catalyzed (asymmetric)
reactions these ligands proved superior with respect to
activity and enantioselectivity. Most notable is the successful
application of monodentate ligands in the rhodium-catalyzed
hydrogenation of alkenes and carbon-element double bonds.2
The high-speed generation of chemical libraries offered by
solid-phase organic synthesis is highly efficient, as workup
and purification can be achieved by simple washing and
filtration. Moreover the site-isolation, resulting from the
immobilization of the reactant on the support, can stabilize
reactive intermediates and thus reduce byproduct formation.
Polymer-supported ligands allow the recycling of the catalyst,
and as a result of the swelling of the polymer, good mixing
with the reactants is maintained.3 Several groups have
prepared polystyrene-supported phosphites and phosphor-
amidites.4 Waldmann and co-workers have shown that
polymer-bound phosphoramidites in copper-catalyzed enan-
tioselective conjugate addition reactions mirrored the per-
(1) See for instance: (a) Baker, M.; Pringle, P. G. J. Chem. Soc. Chem.
Commun. 1991, 1292. (b) Buisman, G. J. H.; Kamer, P. C. J.; van Leeuwen,
P. W. M. N. Tetrahedron: Asymmetry 1993, 4, 1625. (c) Seebach, D.;
Hayakama, M.; Sakaki, J.; Schweizer, B. Tetrahedron 1993, 49, 1711. (d)
Alexakis, S.; Burton, J.; Vastra, J.; Benhaim, C.; Fournioux, X.; van den
Heuvel, A.; Leveque, J.-M.; Maze, F.; Rosset, S. Eur. J. Org. Chem. 2000,
4011-4027. (e) Boele, M. D. K.; Kamer, P. C. J.; Lutz, M.; Spek, A. L.;
de Vries, J. G.; van Leeuwen, P. W. N. M.; van Strijdonck, G. P. F. Chem.
Eur. J. 2004, 10, 24, 6232-6246.
(2) (a) Minnaard, A. J.; Feringa, B. L.; Lefort, L.; de Vries, J. G. Acc.
Chem. Res. 2007, 40, 1267-1277. (b) Reetz, M. T.; Mehler, G. Angew.
Chem., Int. Ed. 2000, 39, 3889-3890. (c) Ansell, J.; Wills, M. Chem. Soc.
ReV. 2002, 31, 259-268.
(3) (a) McNamara, C. A.; Dixon, M. J.; Bradley, M. Chem. ReV. 2002,
102, 3275-3300. (b) Bergbreiter, D. E. Chem. ReV. 2002, 102, 3345-
3384. (c) Leadbeater, N. E.; Marco, M. Chem. ReV. 2002, 102, 3217-
3274. (d) Clapham, B.; Reger, T. S.; Janda, K. D. Tetrahedron 2001, 57,
4637-4662.
(4) (a) Jiang, Z.-D.; Meng, Z.-H. Chin. J. Chem. 2007, 25, 542-545.
(b) Chen, W.; Roberts, S. M.; Whittall, J. Tetrahedron Lett. 2006, 47, 4263-
4266. (c) Hu, X.-P.; Huang, J.-D.; Zeng, Q.-H.; Zheng, Z. Chem. Commun.
2006, 293-295. (d) Doherty, S.; Robins, E. G.; Pal, I.; Newman, C.;
Hardacre, C.; Rooney, D.; Mooney, D. A. Tetrahedron: Asymmetry 2003,
14, 1517-1527. (e) Mandoli, A.; Calamante, M.; Feringa, B. L.; Salvadoria,
P. Tetrahedron: Asymmetry 2003, 14, 3647-3650. (f) Huttenloch, O.;
Laxman, E.; Waldmann, H. Chem. Eur. J. 2002, 8, 4767-4780.
10.1021/ol703070x CCC: $40.75
© 2008 American Chemical Society
Published on Web 02/01/2008

formance of the soluble analogues, albeit with lower activities
and selectivites.4f Lefort and de Vries reported the parallel
synthesis of a library of 96 phosphoramidite ligands in
solution and its application in asymmetric hydrogenation.5
Combinatorial solid-phase techniques can thus speed up both
the synthesis and the testing of libraries of ligands, provided
that efficient synthetic protocols to structural diversity exist.
The reported phosphoramidites and phosphites are, however,
almost exclusively binol (1,1′-binaphthalene-2,2′-diol)-based,
prepared by reacting the corresponding chlorophosphites
with, respectively, amino- or hydroxy-functionalized resins
(Scheme 1).
base in DCM, toluene, or THF as a solvent, allowed the
straightforward synthesis of a variety of (chiral) immobilized
phosphoramidites and phosphites (Scheme 1). Figure 2
Scheme 1. Synthesis of Solid-Phase Ligands
Figure 2. Diols used in the synthesis of supported ligands.
depicts the diols applied in the synthesis of the supported
ligands.6 The presence of phosphorus-based compounds on
the resin was monitored with 31P gel-phase NMR,7,8 and the
observed chemical shifts were in accordance with their
homogeneous analogues. The bis-aryl-based phosphoramid-
ites supported on resin 1 and 2 showed two fairly broad
signals around 148 and 139 ppm, respectively, in a 10:6 (1)
or 10:2 ratio (2).9,10 On the basis of the chemical shifts, the
downfield signals are assigned to the anticipated phosphora-
midites, while the upfield signals remain unidentified,
although the values suggest the presence of supported
phosphites.11
Here we report efficient routes for the solid-phase synthesis
of monodentate phosphoramidites and phosphites, as well
as the application of resin-bound phosphoramidites as
intermediates in the combinatorial synthesis of (chiral)
homogeneous monodentate phosphite ligands. Figure 1
This thus suggests the presence of a second type of
functionality, possibly an alcohol, on the resin. The primary
amine-functionalized TentaGel resin 3 yielded S-binol (b)-
based phosphoramidite 3b, but it proved very prone to
hydrolysis; next to the main signal at 152 ppm, several
signals between 15 and -8 ppm were observed (∼40%).
Phosphoramidites supported on pyrrolidine-functionalized 4
showed a higher stability. The ligands were formed in good
purity, with small amounts (<2%) of hydrolysis products
as the major byproduct. The R-binol (a), bisphenol (c) and
the aliphatic taddol (1,1,4,4-tetraphenyl-2,3-O-isopropyl-
idene-L-threitol, f) based phosphoramidites showed single
resonances at 152 (4a), 151 (4c), and 143 ppm (4f). Applying
hydroxy-functionalized resin 5 in this reaction yielded
supported phosphites cleanly, displaying typical phosphite
resonances in the 31P NMR (128-146 ppm). Since the
purification of the resins can be achieved by a simple
washing procedure, the use of an excess of partially
hydrolyzed (∼30%) chlorophosphite does not compromise
the purity of the resulting immobilized ligands.
Figure 1. Resins used in this study. The balls represent the
polystyrene supports (DVB-PS), and the waved lines (∼∼∼)
represent poly(ethylene glycol) linkers (TentaGel); 1 is N-methyl
aminomethylated polystyrene, 2 is aminomethylated polystyrene,
3 is TentaGel-S-NH2, 4 is 3 functionalized with a proline moiety,
6 is 3 functionalized with a phenol moiety, and 5 is 4-hydroxyphenol
polymer-bound.
depicts the resins used in this study, which were prepared
by standard peptide coupling reactions (Scheme 2).
Scheme 2. Synthesis of Resin 4
(6) The supported ligands are referred to with a code, in which the
numbers indicate the support, and the letters refer to the diol or alcohol
moiety.
(7) Bardella, F.; Eritja, R.; Pedroso, E.; Giralt, E. Bioorg. Med. Chem.
Lett. 1993, 3, 2193-2196.
(8) The use of THF resulted in considerably sharper signals than C6D6,
which we attribute to an increased swelling. Similarly TentaGel gave sharper
signals than unmodified polystyrene.
Reaction of the resins with an excess (1.2-3 equiv) of
chlorophosphites, in the presence of a tertiary amine as a
(9) For 1f a similar pattern was observed at 140 and 132 ppm.
(10) During the preparation of this manuscript Jiang et al. reported the
synthesis of 1a and 1f.4a
(5) Lefort, L.; Boogers, J. A. F.; de Vries, A. H. M.; de Vries, J. G.
Org. Lett. 2004, 6, 1733-1735.
990
Org. Lett., Vol. 10, No. 5, 2008

The synthesis of chlorophosphites from phenols and PCl3
is not trivial, as it frequently leads to mixtures containing
additional species such as dichlorophosphites and phosphites.
We therefore investigated solid-phase routes that can over-
come these problems. We anticipated that by pushing the
equilibrium toward the formation of a mixture of chloro-
phosphite and phosphite, the formation of dichlorophosphite,
and thus of resin-bound byproducts, could be repressed.
Indeed reacting PCl3 with 2.2 equiv of 2-phenylphenol (j),
which should lead to a chlorophosphite/phosphite ratio of
77:23, and the subsequent reaction with the hydroxy-
functionalized resin 5 lead to the clean formation of 5j2
(Scheme 3). Next we investigated the use of resin bound
Inspired by the efficiency of fully automated DNA synthe-
sis,12 we investigated the potential of our phosphoramidites
as synthetic intermediates for phosphites. By applying
phosphoramidites linked to the solid support via the amido
functionality, acid-catalyzed substitution of the P-N bond
by alcohols should thus yield monophosphites liberated from
the resin (Scheme 5). This route would therefore allow the
combinatorial solid-phase synthesis of phosphites. First we
Scheme 5. Solid-Phase Phosphoramidites as Intermediates in
the Synthesis of Phosphites
Scheme 3. Use of Phosphite/Chlorophosphite Mixture
investigated the efficiency of several activators. In the
absence of an acid activator, no reaction was observed for
any of the phosphoramidites with methanol. 1H-Tetrazole,
universally applied as activator in DNA synthesis, proved
inactive for the di-aryl phosphoramidites 4a and 4c, whereas
with the aliphatic 4f it gave rise to approximately 10%
conversion in 12 h (Table 1, entry 1). Other activators like
dichlorophosphites. Reaction of 5 with an excess PCl3 leads
to the clean formation of the corresponding dichlorophosphite
(7), as indicated by a single resonance at 180 ppm observed
with 31P gel-phase NMR. Subsequent reaction of 7 with
2,4,6-trimethylphenol (k) yielded the corresponding phos-
phite 5k2 as the sole product. The approach is not limited to
phenols; reaction with bisphenol (c) yielded 5c.
Table 1. Combinatorial Solid-Phase Synthesis of Phosphites
alcohol
(1.5 equiv)
activator
(1 equiv)
time
(min)
yielda
(%)
entry
resin
To allow maximum structural diversity (“bottom up”
approach), we explored routes to synthesize solid-phase
phosphoramidites containing two different alcohol moieties
(Scheme 4). After reaction of 5 with Et2NPCl2, a single
1
2
3
4
5
6
7
8
4f
4f
4f
4a
4d
1f
1f
1c
MeOH
MeOH
(-)-menthol
BnOH
MeOH
MeOH
MeOH
MeOH
tetrazole
TAMA
TAMA
TAMA
TAMA
TAMA
TAMA
TAMA
720
30
60
60
30
30
30
30
10
82
84
84
91
88
85b
0
Scheme 4. Solid-Phase Synthesis of Phosphoramidites
a Yields based on the loading of 4 and 1, respectively. b 1 recovered
from entry 6 was applied as support for 1f.
N-methyl-imidazolinium trifluoroacetate and pyridinium-
tetrafluoroborate were tested but proved unsatisfactory, as
conversion to product was slow or incomplete and the
prolonged reaction times promoted side product formation
via oxidation and hydrolysis. N-Methylanilinium trifluoro-
acetate13 (TAMA) (1 equiv) was found to be the most suitable
activator, resulting in fast product formation (within 30 min)
in good yields (>80%) for both the aliphatic and the bis-
aryl phosphoramidites immobilized on resin 4 (entries 2-5).
A short plug filtration over silica proved sufficient to remove
the TAMA. Applying catalytic amounts of TAMA slowed
down the reaction considerably.
resonance at 168 ppm was observed with 31P gel-phase NMR
spectroscopy, which we assign to the supported chlorophos-
phoramidite (8, Route A). Subsequent reaction with the
sodium salt of 2-phenylphenol or 2-(pyridine-2-yl)ethanol
yielded the corresponding racemic phosphoramidites (9 and
10) as the sole products, displaying characteristic resonances
at 143 and 146 ppm, respectively. Alternatively homogeneous
in situ formed chlorophosphoramidites based on 2-tert-butyl-
4-methylphenol were reacted with resin 6, leading to the
clean formation of the supported phosphoramidites (Route
B). By variation of the secondary amine applied in the
synthesis (i-Pr2HN or pyrrolidine), two different racemic
phosphoramidites (11 and 12) were synthesized in situ.
(11) Pen˜a, D.; Minnaard, A. J.; de Vries, J. G.; Feringa, B. L. J. Am.
Chem. Soc. 2002, 124, 14552-14553.
(12) Beaucage, S. L.; Iyer, R. P. Tetrahedron 1992, 48, 2223-2311.
This process is based on the protonation of the nitrogen of a phosphoramidite
by an acid activator, followed by reaction with an alcohol to give the
corresponding phosphite
(13) Fourrey, J.-L.; Varenne, J. Tetrahedron Lett. 1984, 25, 4511-4514.
Org. Lett., Vol. 10, No. 5, 2008
991

We applied the optimized reaction conditions to in situ
generated resins 4a-4f in the synthesis of 20 (chiral)
monophosphites using various alcohols. This approach proves
the potential for a parallel synthesis of a monodentate
phosphite library. All phosphites were obtained in high yields
(82-91%) and purities (>97%). Both primary alkyl and
more sterically hindered alcohols can be applied, although
the latter need slightly longer reaction times to reach full
conversion (entries 3 and 4). To find the optimal combination
of stability and reactivity for phosphite release, we also tested
supported phosphoramidites with different amido groups. For
the commercially available N-methylaminomethyl resin (1),
phosphite formation depended highly on the diol moiety. The
aliphatic taddol based phosphoramidite 1f gave the corre-
sponding monophosphites in good yields (80-89%, entry
6). This result indicates that the presence of the second
functionality on the resin, as observed by 31P NMR (vide
supra), does not hamper the overall reaction significantly.
Resin 1 could be successfully reused as support for the solid-
phase synthesis of taddol-based phosphites, with only a slight
drop in yield between successive runs (88% to 85%, entries
6-7), indicating the regeneration of the amino functionality
on the resin. No products were observed when phosphora-
midites 1a and 1c were applied (entry 8).14 We attribute the
difference in reactivity to an insufficient basicity of the
N-methyl-benzylamine moiety (pKa ) 9.7) in bis-aryl phos-
phoramidites. Since the pyrrolidine linker in 4 has a higher
basicity (pKa ) 11.3), it yields reactive phosphoramidites
with both aliphatic and aromatic diols. The use of more acidic
activators in combination with 1a or 1c led to mainly
phosphite hydrolysis products. Homogeneous model com-
pounds showed similar behavior; bisphenol, binol, and taddol
phosphoramidites with a pyrrolidine amido substituent gave
the corresponding phosphites, whereas for N-methyl-benzyl-
amine and dimethylamino based phosphoramidites, only the
taddol derivative was found to be reactive (Scheme 6).
Scheme 6. Effect of the Leaving Group in Di-aryl
Phosphoramidites
In summary, facile synthetic routes to structural diverse
solid-phase (chiral) monodentate phosphite and phosphor-
amidite ligands have been developed. In addition the
polymer-supported phosphoramidites proved valuable inter-
mediates for the combinatorial synthesis of homogeneous
(chiral) monodentate phosphites. We found that the acidity
of the activator, the leaving group capacity of the amino-
functionalized resin, and the nature of the resin and the diol
are all important. The easy preparation of the solid-phase
phosphoramidites and the excellent yields allowed the
preparation of a ligand library in a very efficient manner. A
report on the performance of the solid-phase ligands in
asymmetric transition-metal-catalyzed reactions is in prepara-
tion.
Acknowledgment. This work was supported by the NWO
Combinatorial Chemistry program and DSM.
Supporting Information Available: Analytical data and
experimental procedures for the synthesis of the resins, the
solid-phase ligands and the homogeneous phosphites. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(14) Brill, W. K.-D.; Nielsen, J.; Caruthers, M. H. J. Am. Chem. Soc.
1991, 113, 3972-3980.
OL703070X
992
Org. Lett., Vol. 10, No. 5, 2008