Combinatorial development of chiral phosphoramidite-ligands for enantioselective conjugate addition reactions.

Article abstract of DOI:10.1039/b200454b

Chiral phosphoramidite ligands embodying bispidine frame work and a binaphthyl phosphoramidite for Cu-catalysed enantioselective conjugate addition reactions were developed employing principles of combinatorial and solid phase chemistry.

Full text of DOI:10.1039/b200454b

Combinatorial development of chiral phosphoramidite-ligands for
enantioselective conjugate addition reactions
Oliver Huttenloch, Eltepu Laxman and Herbert Waldmann*
Max-Planck-Institut für molekulare Physiologie, Abteilung Chemische Biologie, Otto-Hahn-Str. 11, 44227
Dortmund, Germany, and Universität Dortmund, FB 3, Organische Chemie 44221 Dortmund, Germany.
E-mail: herbert.waldmann@mpi-dortmund.mpg.de
Received (in Cambridge, UK) 15th January 2002, Accepted 18th February 2002
First published as an Advance Article on the web 6th March 2002
Chiral phosphoramidite ligands embodying bispidine frame
work and a binaphthyl phosphoramidite for Cu-catalysed
enantioselective conjugate addition reactions were devel-
oped employing principles of combinatorial and solid phase
chemistry.
converted either to the methyl ester 6 or linked into hydroxy-
methyl polystyrene to give polymer 7 with a loading of 0.52
mmol g21. Removal of the Boc groups and phosphitylation with
chlorophosphite 1010 yielded model compounds 8 and 9.
Application of compounds 1, 8 and 9 as ligands in the Cu-
catalyzed addition of ZnEt2 to cyclohexenone proceeded with
comparable enantioselectivity (Scheme 2). Thus, at 230 °C
ligands 1 and 8 yield addition product 12 with 43% ee and 46%
ee, respectively, demonstrating that modification of the 9-posi-
tion in the oxabispidine core does not negatively influence the
efficiency of the stereoselection. Gratifyingly, immobilization
of the ligand on polystyrene and performing the reaction at 0 °C
instead of 230 °C (to ensure sufficient swelling) resulted in
only a minor drop of the enantioselectivity to 39% ee. In order
to reproducibly achieve this result, in the preparation of the
catalyst, Cu(OTf)2 had to be solubilized by addition of 3% DMF
to the solvent CH2Cl2 before exposure to the immobilized
phosphoramidite ligand.
The application of combinatorial principles to the development
of enantioselective catalysed transformations has opened up
entirely new opportunities to the field. In particular, the
approach to synthesize libraries of chiral ligands for metal
atoms on polymeric supports and to apply them as heteroge-
neous catalysts in screening systems has emerged as a powerful
technique for the rapid development of ultimately soluble
ligands for enantioselective catalysis.1,2 The successful use of
this method requires that the polymer-bound catalysts display
the same trends in stereoselection as the corresponding soluble
catalyst systems.2b,c,3
For the enantioselective steering of conjugate addition of
organometallic reagents to a,b-unsaturated carbonyl com-
pounds,4 the use of dialkylzinc reagents in the presence of a Cu-
catalyst and in the presence of chiral phosphoramidite li-
gands4–6 is particularly efficient.
The application of the combinatorial principle of Cu-
catalyzed enantioselective conjugate addition processes has
been pursued only in a single case7 in which solid-phase-
extraction techniques were used for the generation of soluble
sulfonamide ligands. However, the approach described above
has not been applied to enantioselectively catalyzed conjugate
addition reactions.
For synthesizing a bispidine-phosphoramidite library bispidi-
none 14 was generated by double Mannich reaction from 13 and
then converted into olefin 15 by means of a Wittig reaction.
Exchange of the N-benzyl group for a Fmoc-urethane was
accompanied by reduction of the double bond. The resulting
Fmoc/Boc-protected bispidine carboxylic acid was linked to
Here we describe the combinatorial synthesis of a library of
polymer-bound bispidine-derived phosphoramidite ligands and
its evaluation in the Cu-catalyzed enantioselective addition of
dialkylzinc reagents to enones. We have recently shown that
phosphoramidites embodying a binaphthol unit and a bispidine-
derived modulating substituent, e.g. 1, can successfully be
employed for the steric steering of Cu-catalyzed enantiose-
lective conjugate addition reactions.8 In order to find more
efficient catalysts embodying the same underlying structural
motives it was decided to synthesize a library of ligands 2 (Fig.
1) on a polymeric support. To determine if the results to be
expected from investigating such immobilized ligands would be
comparable to the values recorded for soluble ligand 1
compounds 8 and 9 were synthesized as shown in Scheme 1.
Benzyl-protected bispidinone 38,9 was converted into Boc-
masked compound 4 which was subjected to a Wittig reaction
yielding olefin 5. The carboxylic acid moiety was then
Scheme 1 Synthesis of phosphoramidites 8 and 9: (a) ClCOOCH(Cl)CH3,
CH2Cl2, D, 2 h; (b) MeOH, D, 2 h; (c) Boc2O, KOH, H2O, dioxane, rt, 12
h, 87% for three steps; (d) Br2 Ph3P+(CH2)4COOH, KOtBu, THF, 0 °C?rt,
2.5 h, 54%; (e) EEDQ, MeOH, rt, 12 h, 82%; (f) hydroxymethylpolystyrene,
DIC, DMAP, CH2Cl2, 12 h, rt, 71% (0.52 mmol g21), then Ac2O, pyridine,
CH2Cl2; (g) TFA, CH2Cl2, 0.5 h, rt, quant.; [(h) only on solid phase: Et3N,
CH2Cl2]; (i) chlorophosphite 10, Et3N, toluene, 12 h, 80 °C, 41% [for 7: rt];
EEDQ = 2-ethoxy-1-ethoxycarbonyl-1,2-dihydroquinoline, DIC = N,NA-
diisopropylcarbodiimide, DMAP = 4-dimethylaminopyridine.
Scheme 2 Results of the Cu-catalyzed conjugate addition of ZnEt2 to
cyclohexenone in the presence of ligands 1, 8 and 9.
Fig. 1 Structure of ligand 1 and general structure of the ligand library 2.
This journal is © The Royal Society of Chemistry 2002
CHEM. COMMUN., 2002, 673–675
673

View Article Online
hydroxymethylpolystyrene (loading 0.61 mmol g21), and after
selective removal of the Fmoc group the liberated secondary
amine was modified by means of acylation-, (thio)isocyanation,
reductive amination-, alkylation-, phosphorylation- or sulfony-
lation reactions (Scheme 3).
All transformations were investigated in solution with a
model bispidine first and then transferred to the solid phase.
Finally, after removal of the Boc group the second amino group
was converted into a phosphoramidite with substituted chlor-
ophosphites 18. By analogy, phosphoramidites were synthe-
sized embodying an oxa-, thia- or carba-bispidine system or a
piperidine core. The building blocks employed for the library
synthesis are shown in Fig. 2.
After completion of the synthesis the ligands were converted
into the corresponding polymer-bound copper complexes as
described above and investigated by the Cu-catalyzed conjugate
addition of ZnEt2 to cyclohexenone at 0 °C. After the
transformations had reached > 95% conversion (2 h) the
reactions were quenched with 2 M HCl and the enantiomer ratio
was determined without further separation procedures by means
of gas chromatography employing a chiral stationary phase.8
The ligands were screened in two rounds of investigation. In
the first round bispidine core A1 and unsubstituted bina-
phthylphosphite C1 were employed and the modulating sub-
stituents B1–B28 at the second nitrogen atom were varied. The
results are given in Fig. 3. It is clearly visible that the nature of
the second nitrogen has a profound influence on the efficiency
of the stereoselection. Thus, in the presence of the sterically
demanding pivaloyl amide B3 (44% ee) the stereoselectivity
was significantly higher than with an acetamide B1. Weakly
complexing aromatic substituents like the thiophene system
B13 (38% ee) were much better than pyridine heterocycles
(B10, B11), and also N-alkyl groups were not advantageous
(B15–B18). On the other hand, in particular the phosphoryl
(B26, 50% ee) and the 4-tosyl group B20 (56% ee) gave very
encouraging results.
Fig. 3 Results of the first screening round.
Scheme 3 Synthesis of polymer-bound ligands 17: (a) PhCH2NH2,
(HCHO)n, AcOH, MeOH, 65 °C, 4 h, 81%; b) Br2 Ph3P+(CH2)4COOH,
KOtBu, THF, 0 °C ? rt, 2.5 h, 68%; (c) H2, Pd/C, EtOH, 12 h, rt; (d)
FmocCl, NaHCO3, THF, H2O, 12 h, rt, 69% (2 steps); (e) hydrox-
ymethylpolystyrene, DIC, DMAP, CH2Cl2, 12 h, rt, 94% (0.61 mmol g21),
then Ac2O, pyridine, CH2Cl2; (f) piperidine, DMF; (g) introduction of the
N-substituent (see text); (h) TFA, CH2Cl2; (i) Et3N, CH2Cl2; (j) ClP(R2BI-
NOL) 18, Et3N, toluene; DIC = N,NA-diisopropylcarbodiimide, BINOL =
2,2A-dihydroxy[1,1A]binaphthyl.
Based on these results, ligands embodying the most promis-
ing N-substituents were subjected to a second round of
investigation in which the structure of the binaphthol system
was varied (C1–C4). This second round also included the
bispidine analogs A2–A5. The results of the corresponding
enantioselective conjugate additions are shown in Fig. 4.
Several trends are apparent from these reactions. First, for the
pivaloyl amide (A1–B3–C1/C2), raising the steric demand of
the binaphthyl substituent from H to CH3 led to significant
improvement of the enantioselectivity to 67% ee. A similar
observation is made for the thiophenoylamide (A1–B13–C1/
C2) (45% ee). Interestingly this trend is reversed for the ligands
embodying a sulfonyl- (A1–B20–C2, 46% ee) or a diph-
enylphosphoryl group (A1–B26–C2, 41% ee). Pronounced,
albeit different cooperative effects between the two substituents
of the nitrogen atoms are also apparent from further combina-
tions, compare for instance A1–B3–C3/C4 with A1–B13–C3/
C4. We would like to stress that the most advantageous
Fig. 2 Building blocks used in the library synthesis.
CHEM. COMMUN., 2002, 673–675
Fig. 4 Results of the second screening round.
674

View Article Online
Table 1 Comparison of polymer-bound and corresponding soluble
ligands
modulating N-substituent identified in the first screening round,
i.e. the 4-tosyl-group B20 did not emerge as the substituent of
choice after the second round. Rather, the pivaloyl amide B3
which reached only rank four in the first round turned out to be
most advantageous after appropriate combination with the right
binaphthyl unit. Thus, if we had followed the frequently used
strategy of ‘positional scanning’,2a,b optimising each sub-
stituent independently and subsequently carrying only the best
candidate through (instead of several ones as done here), the
most efficient ligand A1–B3–C2 would not have been identi-
fied. This finding is of general relevance to combinatorial ligand
and catalyst development. It shows that the concept of
‘positional scanning’ may not be general and should be applied
with care.
BINOL-
Entrya Ligand type substituent eeb [%] (ligand)
Heterogeneousa
Homogeneousa
eeb [%] (ligand)
1
2
3
4
5
6
Oxa-
Bispidine
N-
Pivaloyl-
N-Toluene
4-Sulfonyl
H
Me
H
Me
H
Me
39 (A2–C1)
45 (A2–C2)
44 (A1–B3–C1)
67 (A1–B3–C2)
56 (A1–B20–C1) 56 (24)
46 (A1–B20–C2) 51 (26)
43 (1)
47 (27, 1 mol%)
Not determined
64 (25)
a Conversion in all cases > 95%; isolated yields > 90%; b Determined by
means of gas chromatography (Lipodex E, Macherey&Nagel); the (R)-
enantiomer was formed predominantly in all cases.
In order to validate the screening of the solid-phase bound
ligands and to compare the recorded results with the corre-
sponding reactions in homogeneous solution, ligands 24–27
were synthesized as shown in Scheme 4.
combinatorial ligand development and screening system em-
ploying immobilized heterogeneous catalysts provides a correct
picture of the corresponding situation for the enantioselective
conjugate addition reactions in homogeneous solution.
This research was supported by the Fonds der Chemischen
Industrie. E. L. is grateful to the Alexander-von-Humboldt
Foundation for a fellowship.
Notes and references
1 Reviews: (a) M. T. Reetz, Angew. Chem., 2001, 113, 292; Angew.
Chem., Int. Ed. Engl., 2001, 40, 284; (b) B. Jandeleit, D. J. Schäfer, T.
S. Powers, H. W. Turner and W. H. Weinberg, Angew. Chem., 1999,
111, 2648; Angew. Chem., Int. Ed. Engl., 1999, 38, 2494; (c) S. Dahmen
and S. Bräse, Synthesis, 2001, 1431.
2 (a) B. M. Cole, K. D. Shimizu, C. A. Krueger, J. P. A. Harrity, M. L.
Snapper and A. H. Hoveyda, Angew. Chem., 1996, 108, 1776; Angew.
Chem., Int. Ed. Engl., 1996, 35, 1668; (b) K. D. Shimizu, B. M. Cole, C.
A. Krueger, K. W. Kuntz, M. L. Snapper and A. H. Hoveyda, Angew.
Chem., 1997, 109, 1782; Angew.Chem., Int. Ed. Engl., 1997, 36, 1704;
(c) M. S. Sigman and E. N. Jacobsen, J. Am. Chem. Soc., 1998, 110,
4901; (d) M. B. Francis and E. N. Jacobsen, Angew.Chem., 1999, 111,
987; Angew. Chem., Int. Ed. Engl., 1999, 38, 937.
3 (a) S. Orlandi, A. Mandoli, D. Pini and P. Salvadori, Angew. Chem.,
2001, 113, 2587; Angew. Chem., Int. Ed. Engl., 2001, 40, 2519; (b) L.
Canali, E. Cowan, H. Deluze, C. L. Gibson and D. C. Sherrington, J.
Chem. Soc., Perkin Trans. 1, 2000, 2055; (c) A. J. Brouwer, H. J. van der
Linden and R. M. J. Liskamp, J. Org. Chem., 2000, 65, 1750.
4 N. Krause and A. Hoffmann-Röder, Synthesis, 2001, 171.
5 A. Alexakis, J. C. Frutos and P. Mangeney, Tetrahedron: Asymmetry,
1993, 4, 2427.
Scheme 4 Synthesis of ligands 24–27: (a) TosNHNH2, TosOH, NaCNBH3,
DMF, 2 h, 100 °C, 48%; (b) H2, Pd/C, EtOH, 12 h, rt; (c) Piv2O, DMAP,
pyridine, 12 h, rt, 71% (21, 2 steps); (d) TosCl, DMAP, pyridine, 12 h, rt,
72% (22, 2 steps); (e) TFA, CH2Cl2, 30 min, rt; (f) ClP(BINOL) 10, Et3N,
toluene, 12 h, rt, 51% (24); (g) PCl3, Et3N, THF, 1 h, 0 °C; then 3,3A-
dimethyl-2,2A-dihydroxy[1,1A]binaphthyl, Et3N, THF, 12 h, rt, 30% (25),
12% (26) and 3% (27), respectively; Piv = pivaloyl, Tos = toluene-
4-sulfonyl.
To this end, bispidinones 19 and 20 were converted into Boc-
protected intermediates 21–23. After removal of the Boc group
the liberated secondary amine was phosphitylated. Phosphor-
amidite 24 was obtained readily by reaction with chlor-
ophosphite 10. In the cases of methyl-substituted phosphor-
amidites 25–27 subsequent treatment of the secondary amine
with PCl3 and the dimethyl-substituted binaphthol gave better
results.
In Table 1 the enantioselectivity recorded in the Cu-catalyzed
conjugate addition of diethylzinc to cyclohexenone in the
presence of soluble ligands 24–27 and 1 is compared with the
values determined in the presence of the corresponding
polymer-bound ligands. In all cases the ee-values are very
similar, the maximum deviation reaching only 5% ee. Thus, the
6 B. L. Feringa, Acc. Chem. Res., 2000, 33, 346.
7 I. Chataigner, C. Gennari, U. Piarulli and S. Ceccarelli, Angew.Chem.,
2000, 112, 953; Angew. Chem., Int. Ed. Engl., 2000, 39, 916.
8 O. Huttenloch, J. Spieler and H. Waldmann, Chem. Eur. J., 2001, 7,
671.
9 P. Arjunan, K. D. Berlin, C. L. Barnes and D. van der Helm, J. Org.
Chem., 1981, 46, 3196.
10 J. Scherer, G. Huttner, M. Büchner and J. Bakos, J. Organomet. Chem.,
1996, 520, 45.
CHEM. COMMUN., 2002, 673–675
675