Polymer-supported ferrocenyl oxazolines for the catalyzed highly enantioselective phenyl transfer to aldehydes

Article abstract of DOI:10.1016/S0960-894X(02)00271-8

A new chiral MeO-PEG-supported ferrocenyl oxazoline was synthesized and successfully employed in the enantioselective phenyl transfer to aldehydes. The products were obtained in high yields and with excellent enantioselectivities (up to 97% ee). Furthermo

Full text of DOI:10.1016/S0960-894X(02)00271-8

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1795–1798
Polymer-Supported Ferrocenyl Oxazolines for the Catalyzed
Highly Enantioselective Phenyl Transfer to Aldehydes
Carsten Bolm,* Nina Hermanns, Arno Claßen and Kilian Muniz
Institut fur Organische Chemie, RWTH Aachen, Prof.-Pirlet-Str. 1, D-52056 Aachen, Germany
¨
Received 5 December 2001; accepted 2 April 2002
Abstract—A new chiral MeO-PEG-supported ferrocenyl oxazoline was synthesized and successfully employed in the enantio-
selective phenyl transfer to aldehydes. The products were obtained in high yields and with excellent enantioselectivities (up to 97% ee).
Furthermore, the recovery of the ferrocene was facile, and catalyst efficiency was maintained in subsequent reactions. # 2002
Elsevier Science Ltd. All rights reserved.
The diarylmethane moiety is essential for the physio-
logical activity of many organic compounds.1 Hence,
chiral diarylmethanols are useful intermediates in the
synthesis of biologically active substrates, which have
been used as drugs possessing antihistaminic, anti-
cholinergic, local-anesthetic and laxative properties
(Scheme 1).
polymer matrix and the difficulty to compete with a
rapid and undesired background reaction, respectively,
the application of fully soluble enlarged ligands appears
particularly attractive. In those systems the immobilized
ligands mimic the properties of the corresponding low-
molecular weight counterparts, since this methodology
provides homogeneous reaction conditions.
Diarylmethanols may be prepared in an enantioselective
manner via the asymmetric addition of a suitable
organometallic phenyl transfer reagent to aromatic
aldehydes. Recently, we developed various highly
enantioselective catalytic systems for this transfor-
mation, which, as a common feature, all involve the use
of organozinc reagents.2,3 One of them utilizes ferro-
cenyl oxazoline 4 and a zinc species formed in situ
from ZnPh2 and ZnEt2. Excellent enantioselectivities of
up to 98% ee have been obtained for a large variety of
aldehydes (Scheme 2).
Scheme 1. Biologically active diarylmethanol derivatives.
Generally, a purification by column chromatography
must be employed for the isolation of product and for
the recovery of ferrocenyl oxazoline 4. With the inten-
tion to facilitate the workup as well as the recovery and
reuse of the ferrocene in subsequent catalyses, we have
now focused on an immobilization of the ferrocene by
anchoring it onto polymeric support.4 Since the use of
insoluble polymers often results in low catalytic activity
and enantioselectivity due to the restrictions of the
Scheme 2. Enantioselective phenyl transfer from an in situ generated
organozinc reagent to aldehydes using a catalyst derived from ferro-
cenyl oxazoline 4.
*Corresponding author. Tel.: +49-241-809-4675; fax: +49-241-809-
2391; e-mail: carsten.bolm@oc.rwth-aachen.de
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00271-8

1796
C. Bolm et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1795–1798
Scheme 4. Synthesis of polymer-supported ferrocenes 11 and 13.
Reagents and conditions: (a) Py, DMAP, 50%; (b) DCC, DMAP,
DCM, 95%.
the amide, and ring closure of the tosylated alcohol.
Finally, directed ortho-metallation with s-butyllithium
and subsequent quenching of the resulting anion with
benzophenone followed by standard deprotection of the
hydroxyl group gave ferrocene 9 in 78% overall yield.
Scheme 3. Synthesis of ferrocenyl oxazoline 9. Reagents and condi-
tions: (a) nBuLi, TMEDA, Et2O, then Bu3SnCl, 82%; (b) nBuLi,
THF, then DMF, 80%; (c) LiAlH4, Et2O, 98%; (d) NaH,
Br(CH2)6OTBS, THF, 82%; (e) nBuLi, THF, then (MeO)2CO, 72%;
(f) NaOH, NaBH4, tBuOH, H2O, 98%; (g) C6F5OH, DCC, DCM;
(h) (S)-tert-leucinol, NEt3, DMF, 94%; (i) TsCl, NEt3, DMAP, DCM,
.
94%; (j) sBuLi, THF, then Ph2CO, 80%; (k) HF Py, NEt3, MeCN,
98%.9
First, ferrocene 9 was bound to insoluble trityl chloride
resin 1010 to give 11 by selective ether formation.
Unfortunately, this coupling proceeded in only 50%
yield, as determined by gravimetrical analysis. In con-
trast, the attachment of 9 onto MeO-PEG 12 gave 13 in
95% yield (determined by gravimetrical and elemental
analysis) (Scheme 4).11
Several catalytically active polymers have been reported
that perform with very high enantioselectivities in the
catalytic enantioselective addition of diethylzinc to both
aromatic and aliphatic aldehydes.5,6 However, in the
field of asymmetric phenyl transfer from organozinc
reagents to aldehydes, to the best of our knowledge,
only Pu et al.3b reported a polymeric system. Even
though those polybinaphthyl ligands gave rise to diaryl-
carbinols with high enantiomeric excesses, the reaction
conditions were far from being optimal since high cata-
lyst loadings, low temperatures and slow addition of the
reacting components via a syringe pump were required.
The polymer supported ferrocenes 11 and 13 were then
employed as chiral ligands in enantioselective phenyl
and ethyl addition reactions to aldehydes. As test sub-
strates p-chloro-benzaldehyde and benzaldehyde were
used. The results are summarized in Table 1.
The insoluble resin-bound ferrocene 11 proved to be
unsuitable for the catalysis of the asymmetric phenyl
transfer reaction, and only racemic product was
obtained in the addition of the phenylzinc reagent to
p-chlorobenzaldehyde (Table 1, entry 1). Obviously, the
catalytic reaction is too slow, and most of the product
derives from the fast uncatalyzed background reaction
which affords a racemate. In contrast, the addition of
diethylzinc to benzaldehyde in the presence of ferrocene
11 gave rise to the addition product with 87% ee. In this
case it is known that the uncatalyzed background reac-
tion is slow, and therefore it is not able to compete with
Herein, we describe the preparation of new polymer-
supported ligands based on ferrocenyl oxazoline 4 and
their catalytic properties in the enantioselective phenyl
transfer to aldehydes. For including both soluble, as
well as insoluble supports we chose the application of
commercially available polyethylene glycol monomethyl
ether [MeO-PEG-OH (MPEG); MW=5000] and a trityl
chloride resin, respectively.
Ferrocenyl oxazoline 9, which bears a linker at the 10-
carbon, was chosen as key-intermediate for the immo-
bilization.7 The different substitution pattern of the two
cyclopentadienyl rings was accomplished by sequen-
tial transmetalation/substitution reactions of 1,10-bis
(tributylstannyl)ferrocene8 (Scheme 3). Thus, aldehyde 5
was obtained in 80% yield after monolithiation, trap-
ping with DMF, and subsequent hydrolysis. After
reduction of 5 to the corresponding alcohol, the TBS-
protected linker chain was attached by ether formation,
giving rise to ferrocene 6 in 80% overall yield. Trans-
metalation of the remaining tributyltin moiety of 6 with
n-butyllithium, followed by reaction of the resulting
anion with dimethyl carbonate and subsequent hydro-
lysis of the intermediate methylester afforded acid 7. The
oxazolinyl substituent was introduced by carboxy group
activation via the pentafluorophenyl ester, formation of
Table 1. Enantioselective ethyl/phenyl transfer to benzaldehyde/
p-chlorobenzaldehydea
Entry
Substrate
Zinc reagent
Ferrocene
eeb [%]
1
2
3
4
4-Cl–C6H4–CHO
C6H5--CHO
4-Cl–C6H4–CHO
C6H5–CHO
ZnPh2/ZnEt2
ZnEt2
ZnPh2/ZnEt2
ZnEt2
11
11
13
13
0(97)
87 (93)c
97 (97)
86 (93)c
aReactions were carried out with 10mol% of ferrocene in toluene at
10 ꢀC. The products were obtained in good to quantitative yields.
bDetermined by HPLC using a chiral stationary phase. Values in par-
entheses refer to results from catalyses with ferrocenyl oxazoline 4.
cResult in parentheses refers to a catalysis with 5 mol% of ferrocenyl
oxazoline 4 at 0 ꢀC.

C. Bolm et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1795–1798
Table 2. Enantioselective phenyl transfer using recycled ferrocene 13a
1797
Acknowledgements
Entry
Substrate
Cycle
Yield (%)
eeb (%)
We are grateful to the Fonds der Chemischen Industrie
and to the Deutsche Forschungsgemeinschaft (DFG)
within the Collaborative Research Center (SFB) 380
‘Asymmetric Synthesis by Chemical and Biological
Methods’ for financial support.
1
2
3
4
5
4-Cl–C6H4–CHO
4-Cl–C6H4–CHO
4-Cl–C6H4–CHO
4-Cl–C6H4–CHO
4-Cl–C6H4–CHO
1
2
3
4
5
8097
75
81
97
96
95
95
8095
aReactions were carried out with 10mol% of catalyst in toluene at
10 ꢀC.
bDetermined by HPLC using a chiral stationary phase.
References and Notes
1. (a) For examples see: Meguro, K.; Aizawa, M.; Sohda, T.;
Kawamatsu, Y.; Nagaoka, A. Chem. Pharm. Bull. 1985, 33,
3787. (b) Toda, F.; Tanaka, K.; Koshiro, K. Tetrahedron:
Asymmetry 1991, 2, 873. (c) Botta, M.; Summa, V.; Corelli, F.;
Di Pietro, G.; Lombardi, P. Tetrahedron: Asymmetry 1996, 7,
1263. (d) Stanchev, S.; Rakovska, R.; Berova, N.; Snatzke, G.
Tetrahedron: Asymmetry 1995, 6, 183.
the catalytic system. However, compared to the result
obtained in a catalysis with low-molecular ferrocenyl
oxazoline 4, the enantioselectivity in the run with 11 was
significantly lower (Table 1, entry 2).12
2. (a) Bolm, C.; Muniz, K. Chem. Commun. 1999, 1295. (b)
Bolm, C.; Hermanns, N.; Hildebrand, J. P.; Muniz, K. Angew.
Chem., Int. Ed. 2000, 39, 3465. (c) Bolm, C.; Kesselgruber, M.;
Hermanns, N.; Hildebrand, J. P. Angew. Chem., Int. Ed. 2001,
39, 1488. (d) Bolm, C.; Kesselgruber, M.; Grenz, A.; Her-
manns, N.; Hildebrand, J. P. New J. Chem. 2001, 25, 13. (e)
Bolm, C.; Hermanns, N.; Kesselgruber, M.; Hildebrand, J. P.
J. Organomet. Chem. 2001, 624, 157. (f) For a recent review on
catalyzed asymmetric arylation reactions: Bolm, C.; Hildeb-
rand, J. P.; Muniz, K.; Hermanns, N. Angew. Chem., Int. Ed.
2001, 40, 3284.
3. (a) For other catalyzed phenyl zinc additions to aldehydes,
see: Dosa, P. I.; Ruble, J. C.; Fu, G. C. J. Org. Chem. 1997,
62, 444. (b) Huang, W.-S.; Hu, Q.-S.; Pu, L. J. Org. Chem.
1999, 64, 7940. (c) Huang, W.-S.; Pu, L. Tetrahedron Lett.
2000, 41, 145 and references therein..
When the MeO-PEG-supported ferrocene 13 was
employed in the title reaction an excellent enantio-
selectivity was observed. The addition product of the
phenylzinc reagent onto p-chloro-benzaldehyde was
obtained in quantitative yields and with 97% ee (Table
1, entry 3). It is noteworthy, that this result is identical
to the one obtained with low-molecular ferrocenyl oxa-
zoline 4. Furthermore, for this remarkably high
enantioselectivity neither an increase of catalyst loading
nor a slow substrate addition was required.
Unfortunately, in the case of the diethylzinc addition to
benzaldehyde, a catalysis with the immobilized ligand
13 did not reach the enantioselectivity achieved with
ferrocenyl oxazoline 4. Here, the addition product was
only obtained with 86% ee.
4. (a) DeVos, D. E., Vankelecom, I. F. J., Jacobs, P. A., Eds.
Chiral Catalyst Immobilization and Recycling; Wiley-VCH:
Weinheim, 2000. (b) Clapham, B.; Reger, T. S.; Janda, K. D.
Tetrahedron 2001, 57, 4637. (c) Brase, S.; Dahmen, S. Syn-
thesis 2001, 1431.
¨
In order to gain insight into the recyclability of the
MeO-PEG-bound ferrocene 13, studies on the recovery
and reuse were performed (Table 2). To our delight, we
found that the excellent enantioselectivity was retained
throughout successive addition reactions. Even after
consecutive use of 13 in five catalytic cycles, the product
from the phenyl transfer to p-chlorobenzaldehyde was
obtained with an excellent ee of 95%. Furthermore it
should be noted that both the reaction protocol as well
as the separation and recycling of the catalyst were
simple and efficient.13
5. Pu, L.; Yu, H.-B. Chem. Rev. 2001, 101, 757 and references
cited therein..
6. (a) For reviews on soluble polymer-supported catalysts, see:
Wentworth, H. P., Jr; Janda, K. D. Chem. Commun. 1999,
1917. (b) Pu, L. Chem. Eur. J. 1999, 5, 2227. (c) Bergbreiter,
D. E. Catal. Today 1998, 42, 389. (d) Bolm, C.; Gerlach, A.
Eur. J. Org. Chem. 1998, 21.
7. (a) For the use of other supported ferrocenes in asymmetric
catalysis, see: Kollner, C.; Pugin, B.; Togni, A. J. Am. Chem.
¨
Soc. 1998, 120, 10274. (b) Blaser, H.-U.; Spindler, F.; Jacob-
sen, E. N.; Pfaltz, A.; Yamamoto, H. Comprehensive Asym-
metric Catalysis, Vol. 3; Springer: Berlin, 1999; p 1427. (c)
Pugin, B. Chimia 2001, 55, 719. (d) Gotov, B.; Toma, S.;
Macquarrie, D. J. New J. Chem. 2000, 24, 597. (e) Johnson,
B. F. G.; Raynor, S. A.; Shephard, D. S.; Mashmeyer, T.;
Thomas, J. M.; Sankar, G.; Bromley, S.; Oldroyd, R.; Glad-
den, L.; Mantle, M. D. Chem. Commun. 1999, 1167.
In conclusion, we have reported the syntheses of
new polymer-supported ferrocenes and their use in
asymmetric C–C-bond formation by phenyl-to-aldehyde
addition reactions. The results from the enantioselective
catalysis demonstrate that the soluble MeO-PEG-bound
ferrocene 13 exhibits excellent catalytic activity and
enantioselectivity. Furthermore, 13 can easily be recov-
ered and reused in successive catalytic additions,
maintaining excellent enantioselectivity even after five
cycles. To the best of our knowledge, this is the
first example of a MeO-PEG-supported ligand in the
asymmetric addition of organozinc reagents to alde-
hydes. We have thus devised a suitable way to generate
chiral arylphenylmethanols in an asymmetric catalytic
fashion.
8. (a) Sammakia, T.; Latham, H. A. J. Org. Chem. 1996, 61,
1629. (b) Wright, M. E. Organometallics 1990, 9, 853.
1
9. Selected analytical data. 5: H NMR (CDCl3, 300 MHz) d
0.89–0.96 (m, 9H), 0.99–1.08 (m, 6H), 1.28–1.42 (m, 6H),
1.48–1.60(m, 6H), 4.09 (s, 2H), 4.46 (s, 2H), 4.51 (s, 2H), 4.72
(s, 2H), 9.95 (s, 1H). 13C NMR (CDCl3, 75 MHz) d 10.22,
13.70, 27.34, 29.13, 69.45, 71.35, 72.10, 73.12, 75.92, 79.20,
1
193.24. 6: H NMR (CDCl3, 300 MHz) d 0.04 (s, 3H), 0.05 (s,
3H), 0.89 (s, 9H), 0.89 (s, 9H), 0.99–1.05 (m, 6H), 1.26–1.42
(m, 10H), 1.44–1.62 (m, 10H), 3.39 (t, J=6.7 Hz, 2H), 3.58 (t,
J=6.6 Hz, 2H), 3.96 (t, J=1.7 Hz, 2H), 4.06 (t, J=1.7 Hz,
2H), 4.17 (t, J=1.7 Hz, 2H), 4.25 (s, 2H), 4.28 (t, J=1.7 Hz,

1798
C. Bolm et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1795–1798
2H). 13C NMR (CDCl3, 75 MHz) d ꢁ5.26, 10.26, 13.72, 18.36,
25.68, 25.98, 27.43, 29.21, 29.73, 32.83, 63.20, 68.43, 69.10,
69.19, 69.38, 70.00, 70.89, 74.80, 83.42. 7: mp 46–47 ꢀC. 1H
NMR (CDCl3, 300 MHz) d 0.04 (s, 6H), 0.89 (s, 9H), 1.25–
1.40(m, 4H), 1.45–1.60(m, 4H), 3.42 (t, J=6.6 Hz, 2H), 3.58
(t, J=6.6 Hz, 2H), 4.24 (t, J=1.9 Hz, 2H), 4.24 (s, 2H), 4.29
(t, J=1.9 Hz, 2H), 4.43 (t, J=1.9 Hz, 2H), 4.81 (t, J=1.9 Hz,
2H). 13C NMR (CDCl3, 75 MHz) d ꢁ5.24, 18.38, 25.64, 26.00,
29.64, 32.80, 63.25, 67.99, 70.51, 71.06, 71.13, 72.42, 85.38,
177.09. IR (KBr) ꢀ~=2931, 2858, 1653, 1475, 1288, 1104, 834
cmꢁ1. MS (70eV) m/z (%): 474 (M, 3), 55 (100). Anal. calcd
for C24H38FeO4Si: C, 60.75; H, 8.07. Found C, 60.82; H, 8.01.
8: 1H NMR (CDCl3, 300 MHz) d 0.03 (s, 6H), 0.89 (s, 9H),
1.26–1.37 (m, 4H), 1.45–1.59 (m, 4H), 3.38 (t, J=6.6 Hz, 2H),
3.58 (t, J=6.6 Hz, 2H), 3.90(dd, J=7.8, 10.0 Hz, 1H), 4.14–
4.18 (m, 3H), 4.22–4.25 (m, 5H), 4.27–4.30(m, 2H), 4.70–4.73
(m, 2H). 13C NMR (CDCl3, 75 MHz) d ꢁ5.26, 18.33, 25.64,
25.98, 26.00, 29.69, 32.80, 33.57, 63.16, 68.36, 69.49, 69.87,
69.95, 70.09, 70.59, 70.62, 70.74, 70.89, 70.97, 75.99, 84.89,
167.45; IR (CHCl3) ꢀ=3408, 3086, 3059, 2935, 2863, 1651,
1448, 756 cmꢁ1. MS (70eV) m/z (%): 623 (M, 56), 624 (M+1,
24), 411 (100). Anal. calcd for C37H45FeNO4: C, 71.26; H,
7.27; N, 2.25. Found C, 71.16; H, 7.33; N, 2.70.
~
10. Trityl chloride resin, polymer matrix: copoly(styrene-1%
DVB), 200–400 mesh, Novabiochem.
11. Elemental analysis indicated a slightly lower yield as was
determined by gravimetrical analysis.
12. Use of 5 mol% of ferrocenyl oxazoline 4 in the diethylzinc
addition to benzaldehyde gave the product with 93% ee.
13. Experimental: Typical procedure for the addition of the
phenylzinc reagent (obtained from ZnPh2 and ZnEt2) to alde-
hydes using MeO-PEG-supported ferrocene 13: In a glovebox,
a well-dried Schlenk flask was charged with diphenylzinc (39
mg, 0.176 mmol). The flask was sealed and removed from the
glovebox. Freshly distilled toluene (3 mL) was added followed
by diethylzinc (58 mL, 0.58 mmol). After stirring the mixture
for 30min at room temperature, MeO-PEG-supported ferro-
cene 13 (137 mg, 0.025 mmol) was added, and the resulting
solution was then cooled to 10 ꢀC. Stirring was continued for
additional 10min at this temperature, and the aldehyde (0.25
mmol) was then added directly in one portion. The Schlenk
flask was sealed, and the reaction mixture was stirred at 10 ꢀC
overnight. Ferrocene 13 was precipitated by addition of 50mL
diethyl ether and then recovered by filtration. The ether filtrate
was washed with water. Drying with MgSO4 and evaporation
of the solvent under reduced pressure gave the product in
good to quantitative yields. The enantiomeric excess of the
product was determined by HPLC analysis after filtration over
a small pad of silica. The recovered catalyst was used for sub-
sequent reactions without further purification.
165.53. IR (kap) ꢀ~=2952, 2933, 2858, 1660, 1114, 1096 cmꢁ1
.
MS (70eV) m/z (%): 555 (M, 68), 556 (M+1, 29), 340 (100).
Anal. calcd for C30H49FeNO3Si: C, 64.85; H, 8.89; N, 2.52.
Found C, 64.45; H, 9.05; N, 2.82. 9: [a]2D0=ꢁ286.5ꢀ (c 1.15,
CH2Cl2). 1H NMR (CDCl3, 300 MHz) d 0.96 (s, 9H), 1.30–
1.38 (m, 4H), 1.48–1.70(m, 4H), 3.34 (br s, 2H), 3.48–3.56 (m,
1H), 3.58–3.66 (m, 2H), 3.73 (s, 1H), 3.95–4.17 (m, 4H), 4.22
(s, 1H), 4.67 (s, 1H), 7.12 (s, 5H), 7.22–7.26 (m, 1H), 7.28–7.35
(m, 2H), 7.50–7.58 (m, 2H), 9.11 (s, 1H). 13C NMR (CDCl3,
75 MHz) d 25.56, 25.94, 26.26, 29.60, 32.67, 32.91, 62.82,
66.43, 68.26, 68.56, 69.95, 71.19, 71.64,71.90, 75.07, 75.26,
84.86, 100.62, 126.08, 126.43, 126.99, 127.62, 145.84, 148.93,