A highly diastereoselective synthesis of α-hydroxy-β-amino acid derivatives via a Lewis acid catalyzed three-component condensation reaction

Article abstract of DOI:10.1021/jo1011762

A very efficient three-component synthesis of a series of syn α-hydroxy-β-amino esters, obtained in high diastereoselection and yield, was realized starting from an aldehyde, benzylamine, and the ketene silyl acetals derived from Ley's lactones. The synthetic protocol was optimized and the above compounds were obtained without the isolation of intermediates. The origin of the observed diastereoselection was investigated through a computational model of the key reaction step.

Full text of DOI:10.1021/jo1011762

pubs.acs.org/joc
A Highly Diastereoselective Synthesis of r-Hydroxy-β-amino
Acid Derivatives via a Lewis Acid Catalyzed Three-Component
Condensation Reaction
Federico Gassa,† Alessandro Contini,† Gabriele Fontana,‡ Sara Pellegrino,† and
Maria Luisa Gelmi*,†
†
ꢀ ꢀ
DISMAB, Sezione di Chimica Organica “A. Marchesini” Facolta di Farmacia, Universita di Milano,
via Venezian 21, 20133 Milano, Italy, and ‡Indena SPA, via Ripamonti 99, 20141 Milano, Italy
marialuisa.gelmi@unimi.it
Received June 16, 2010
A very efficient three-component synthesis of a series of syn R-hydroxy-β-amino esters, obtained in
high diastereoselection and yield, was realized starting from an aldehyde, benzylamine, and the
ketene silyl acetals derived from Ley’s lactones. The synthetic protocol was optimized and the above
compounds were obtained without the isolation of intermediates. The origin of the observed
diastereoselection was investigated through a computational model of the key reaction step.
SCHEME 1. Retrosynthesis of R-Hydroxy-β-amino Acids
Introduction
R-Hydroxy-β-amino acids 1 (Scheme 1) are synthetic targets
of many researchers because they are present in molecules of
great biological interest such as ornicorrugatin, a new lipopep-
tidic siderophore;1 KRI-1314,2 a potent human renin inhibitor
polypeptide; amastatin,3 a tetrapeptide with immunoregula-
tory, antitumor, and antibacterial activity; microginin;4 threo-
β-benzyloxyaspartate (TBOA),5 the first nontransportable
blocker for all subtypes of excitatory amino acid transporters
(EAATs); and most of all, taxane derivatives.6 In this last case,
since the R-hydroxy-β-amino propanoic acid plays an impor-
tant role in the interaction with the bioreceptor as well as in
increasing the bioavailability of the molecule, several SAR
studies were performed including modifications of the skeleton
and of the functional groups linked to the heteroatoms.
As a result of the importance of these amino acids, differ-
ent synthetic methods have appeared over the years based on
the homologation of chiral amino acids, the enantioselective
introduction of amino and hydroxy groups from olefinic
acids, the Staudinger reaction between ketenes and imines to
give β-lactams, that are then hydrolyzed to the corresponding
amino acids, and the aldimine-type coupling with an enolate of
an ester or a ketene acetal.7
(1) Matthijs, S.; Budzikiewicz, H.; Schafer, M.; Wathelet, B.; Cornelis, P.
Z. Naturforsch., C: J. Biosci. 2008, 63, 8–12.
(2) Iizuka, K.; Kamijo, T.; Harada, H.; Akahane, K.; Kubota, T.;
Umeyama, H.; Ishida, T.; Kiso, Y. J. Med. Chem. 1990, 33, 2707–2714.
(3) Blomgren, H.; Wasserman, J. Cancer Lett. 1981, 11, 303–308.
(4) Okino, T.; Matsuda, H.; Murakami, M.; Yamaguchi, K. Tetrahedron
Lett. 1993, 34, 501–504.
(5) Shimamoto, K. Chem. Rec. 2008, 8, 182–199.
(6) (a) Ke, B.; Qin, Y.; Zhao, F.; Qu, Y. Bioorg. Med. Chem. Lett. 2008,
18, 4783–4785. (b) Ganesh, T.; Yang, C.; Norris, A.; Glass, T.; Bane, S.;
Ravindra, R.; Banerjee, A.; Metaferia, B.; Thomas, S. L.; Giannakakou, P.;
Alcaraz, A. A.; Lakdawala, A. S.; Snyder, J. P.; Kingston, D. G. I. J. Med.
Chem. 2007, 50, 713–725. (c) Thoret, S.; Gueritte, F.; Guenard, D.; Dubois, J.
Org. Lett. 2006, 8, 2301–2304. (d) Chen, J.; Kuznetsova, L. V.; Ungreanu,
I. M.; Ojima, I. In Enantioselective Synthesis of β-Amino Acids, 2nd ed.;
Juaristi, E., Soloshonok, V. A., Eds.; John Wiley & Sons, Inc.: Hoboken, NJ,
2005; pp 447-476. (e) Spletstoser, J. T.; Flaherty, P. T.; Himes, R. H.; Georg,
G. I. J. Med. Chem. 2004, 47, 6459–6465. (f ) Liu, C.; Tamm, M.; Notzel,
M. W.; de Meijere, A.; Schilling, J. K.; Kingston, D. G. I. Tetrahedron Lett.
2003, 44, 2049–2052. and references therein. (g) Kingston, D. G. I.; Jagtap,
P. J.; Yuan, H.; Samala, L. In Progress in the Chemistry of Organic Natural
Products; Herz, W., Kirby, G. W., Moore, R. E., Steglich, W., Tamm, C.,
Eds.; Springer-Verlag: Wien & New York, 2002; Vol. 84, pp 53-225.
(h) Georg, G. I.; Cheruvallath, Z. S.; Himes, R. H.; Mejillano, M. R.; Burke,
C. T. J. Med. Chem. 1992, 35, 4230–4237.
Ley’s acetals 28 have attracted our attention since they are
protected R-hydroxy acids efficiently obtained through a cheap
synthesis, in multigram scale, and in enantiopure form.8j It is
well-known that acetals 2 react with different electrophiles
under basic conditions with high diastereoselectivity. Further-
more, the “one-pot” deprotection to reveal both hydroxy and
DOI: 10.1021/jo1011762
r
Published on Web 10/14/2010
J. Org. Chem. 2010, 75, 7099–7106 7099
2010 American Chemical Society

Gassa et al.
JOCArticle
SCHEME 2. Synthesis of Acetal 2b and of Ketene Silyl Acetals 4a,ba
aReagents and conditions: (i) cat. H2SO4, 50 °C, 1 h. (ii) LHMDS, Et3SiCl, THF -78 °C, 10 min, then 25 °C, 20 h.
carboxylic groups is easily performed, affording a large number
of compounds of biological interest.
Despite their different applications, to our knowledge, the
condensation of compounds 2 with imine derivatives 3,
which could be a simple route to access isoserine derivatives
1 (Scheme 1), has not been reported so far.
Syn isomers are generally needed when amino acid moi-
eties 1 are present in compounds of biological interest. In
principle, this stereochemical feature is in contrast with the
results found by the condensation of 2 with an aldehyde,8d,g
which gives as the main isomer the anti dihydroxy acid. So, a
versatile procedure to give isoserine derivatives and to direct
the diastereoselection toward the syn isomer was our challenge.
These synthetic targets have been accomplished by using a very
efficient three-component condensation protocol, promoted
by Lewis acids, between an aldehyde, an amine, and a ketene
silyl acetal derived from Ley’s lactones, without isolation of
intermediates.
2b, but the preparation of the starting 2,3-diethoxybutadiene9
gave poor yield and low reproducibility. The above proce-
dure was modified and a very efficient “one-pot” protocol
was realized starting directly from a mixture of 2,3-butan-
dione, glycolic acid, and ethyl orthoformate operating in
the presence of catalytic H2SO4 at 50 °C (1 h). Compound
2b was directly obtained in 52% yield. (Scheme 2)
In a first experiment the enolate of 2a, generated with
LHMDS in THF at -78 °C,8c was made to react with
N-benzylidene-benzylamine (3a). The reaction was unsuc-
cessful, and only the starting materials were recovered. The
use of different bases (tBuOK or LDA in THF at -78 °C)
failed also.
The Mannich-like reaction of 2a and 3a catalyzed by Lewis
acids was then investigated. According to a known procedure,10
TiCl4 was used, but 2a decomposed. In order to decrease the
acidity of the Lewis acid, both Ti(i-PrO)4 and InCl3 were tested,
1
but only traces of products were detected by H NMR. The
acid-catalyzed condensation of imines and ketene silyl acetals
represents a known alternative synthetic procedure.7k,n,11
Therefore, the ketene silyl acetals 4 were prepared and tested
(Scheme 2).
Results
The preparation of 2a is reported in the literature8c starting
from 2,3-dimethoxybutadiene and glycolic acid in dichloro-
The preparation of the ketene trimethylsilyl acetal of 2a
was reported in literature:8b,12 To achieve a better stability,
the ketene triethylsilyl acetal 4a was prepared in 82% yield
from 2a by generating its enolate with LHMDS in THF at
-78 °C and then by addition of the silylating agent. The
above protocol was modified for an efficient synthesis of
the new compound 4b (85%), i.e., the solution of 2b and
triethylsilylchloride in THF at -78 °C was initially prepared,
and then the base was added (Scheme 2).
Initially, the reaction of 4a and imine 3a was studied by
using several Lewis acids13 in MeCN at -30 °C (1 h, 3a/4a/
cat. 1:1:0.5 ratio) (Scheme 3). As reported in Table 1, the
reaction was not operative in absence of a catalyst (entry 1),
whereas in the presence of a Lewis acid diastereomers 5a and
6a were obtained together with lactone 2a deriving from the
hydrolysis of 4a. The best diastereoselection was found with
methane at 25 °C and using Ph3P HBr as a catalyst. The
3
same procedure could be adopted to obtain the new lactone
(7) (a) Guo, Z.; Shi, T.; Jiang, J.; Yang, L.; Hu, W. Org. Biomol. Chem.
2009, 7, 5028–5033. and references therein. (b) Cariou, C. C. A.; Clarkson,
G. J.; Shipman, M. J. Org. Chem. 2008, 73, 9762–9764. (c) Bodkin, J. A.;
Bacskay, G. B.; McLeod, M. D. Org. Biomol. Chem. 2008, 6, 2544–2553. (d)
Hata, S.; Tomioka, K. Tetrahedron 2007, 63, 8514–8520. (e) Torssell, S.;
Somfai, P. Adv. Synth. Catal. 2006, 348, 2421–2430. (f ) Li, F.; Schwardt, O.;
€
Jaeger, V. Synthesis 2006, 2173–2182. (g) Wang, Y.; He, Q.-F.; Wang, H.-W.;
Zhou, X.; Huang, Z.-Y.; Qin, Y. J. Org. Chem. 2006, 71, 1588–1591. and
references therein. (h) Feske, B. D.; Kaluzna, I. A.; Stewart, J. D. J. Org.
Chem. 2005, 70, 9654–9657. (i) Martin-Zamora, E.; Ferrete, A.; Llera, J. M.;
~
Munoz, J. M.; Pappalardo, R. R.; Fernandez, R.; Lassaletta, J. M. Chem.;
Eur. J. 2004, 10, 6111–6129. (j) Aggarwal, V. K.; Vasse, J.-L. Org. Lett. 2003,
5, 3987–3990. (k) Ha, H.-J.; Ahn, Y.-G.; Woo, J.-S.; Lee, G. S.; Lee,
W. K. Bull. Chem. Soc. Jpn. 2001, 74, 1667–1672. and references therein.
(l) Andersson, P. G.; Guijarro, D.; Tanner, D. J. Org. Chem. 1997, 62, 7364–
7375. (m) Bunnage, M. E.; Davies, S. G.; Goodwin, C. J. J. Chem. Soc.,
Perkin Trans. 1 1994, 2385–2391. (n) Hattori, K.; Miyata, M.-; Yamamoto,
H. J. Am. Chem. Soc. 1993, 115, 1152–1152. (o) Kobayashi, Y.; Takemoto,
Y.; Kamijo, T.; Harada, H.; Ito, Y.; Terashima, S. Tetrahedron 1992, 48,
1853–1868.
(8) (a) Maddes, M. L.; Tackett, M. N.; Watanabe, H.; Brennan, P. E.;
Spilling, C. D.; Scott, J. S.; Osborn, D. P.; Ley, S. V. Angew. Chem., Int. Ed.
2007, 46, 591–597. (b) Ley, S. V.; Dixon, D. J.; Guy, R. T.; Rodriguez, F.;
Sheppard, T. D. Org. Biomol. Chem. 2005, 3, 4095–4107. (c) Ley, S. V.; Diez,
E.; Dixon, D. J.; Guy, R. T.; Michel, P.; Nattrass, G. L.; Sheppard, T. D. Org.
Biomol. Chem. 2004, 2, 3608–3617. (d) Ley, S. V.; Dixon, D. J.; Guy, R. T.;
(10) (a) Periasamy, M.; Suresh, S.; Ganesan, S. S. Tetrahedron: Asym-
metry 2006, 17, 1323–1331. (b) Silveira, C. C.; Vieira, A. S.; Braga, A. L.;
Russowsky, D. Tetrahedron 2005, 61, 9312–9318.
(11) (a) Wang, S.; Matsumura, S.; Toshima, K. Tetrahedron Lett. 2007,
48, 6449–6452. (b) Xiao, J.; Loh, T.-P. Synlett 2007, 815–817. and references
therein. (c) Akiyama, T.; Suzuki, A.; Fuchibe, K. Synlett 2005, 6, 1024–1026.
ꢁ
Palomero, M. A.; Polara, A.; Rodrı
Chem. 2004, 2, 3618–3627. (e) Diez, E.; Dixon, D. J.; Ley, S. V.; Polara, A.;
Rodrıguez, F. Synlett 2003, 8, 1186–1188. (f ) Dixon, D. J.; Ley, V. S.;
´
guez, F.; Sheppard, T. D. Org. Biol.
(d) Jaber, N.; Assie, M.; Fiaud, J.-C.; Collin, J. Tetrahedron 2004, 60, 3075–
3083. (e) Jacobsen, M. F.; Skrydstrup, T. J. Org. Chem. 2003, 68, 7112–7114.
(f ) Saaby, S.; Nakama, K.; Lie, M. A.; Hazell, R. G.; Jorgensen, K. A.
Chem.;Eur. J. 2003, 9, 6145–6154. (g) Allef, P.; Kunz, H. Tetrahedron:
Asymmetry 2000, 11, 375–378. (h) Cozzi,P.G.;DiSimone,B.;Umani-Ronchi,
A. Tetrahedron Lett. 1996, 37, 1691–1694. (i) Mukaiyama, T.; Kashiwagi, K.;
Matsui, S. Chem. Lett. 1989, 1397–1400. (j) Ojima, I.; Inaba, S.; Nagai, M.
Synthesis 1981, 7, 545–547.
(12) Liu, X.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126, 5182–5191.
(13) Kobayashi, S.; Busujima, T.; Nagayama, S. Chem.;Eur. J. 2000, 6,
3491–3494.
´
Rodriguez, F. Org. Lett. 2001, 3, 3753–3755. (g) Dixon, D. J.; Ley, S. V.;
Polara, A.; Sheppard, T. D. Org. Lett. 2001, 3, 3749–3752. (h) Ley, V. S.;
Michel, P. Synlett 2001, 11, 1793–1795. (i) Ley, V. S.; Baeschlin, D. K.;
Dixon, D. J.; Foster, A. C.; Ince, S. J.; Priepke, H. W. M.; Reynolds, D. J.
Chem Rev. 2001, 101, 53–80. (j) Diez, E.; Dixon, D. J.; Ley, S. V. Angew.
Chem., Int. Ed. 2001, 40, 2906–2909.
(9) Pozzo, J.-L.; Clavier, G. M.; Colomes, M.; Bouas-Laurent, H.
Tetrahedron 1997, 53, 6377–6390.
7100 J. Org. Chem. Vol. 75, No. 21, 2010

Gassa et al.
JOCArticle
SCHEME 3. Preparation of Intermediates 5/6 by Mannich-like Condensationa
aReagents and conditions: (i) Method A: InCl3, MeCN, -30 °C, 1 h. (ii) Method B: activated molecular sieves, MeCN, 25 °C, 1 h, then 4a, InCl3,
-30 °C, 1 h.
TABLE 1. Synthesis of 5/6a: Lewis Acid Effect on Diastereoselection
TABLE 2. Synthesis of 5/6a: Solvent and Temperature Effect on
Diastereoselection
entrya
catalyst
5a:6a:2a
entrya
solvent
catalyst/°C
5a:6a:2a
1
2
3
4
5
6
7
8
b
c
ZnCl2
CoCl2
NbCl5
InCl3
SnCl2
Y(OTf)3
PdCl2
Eu(OTf)3
CeCl3
MgBr2
2.8:1:0.7d,e
2.7:1:0.7d,e
2.8:1:0.6d,e
4:1:0.7d,e,f
3.3:1:0.4d,f
2.5:1:1d
1
2
3
4
5
6
7
8
MeCN
DCM
THF
DMA
toluene
DCM
MeCN
MeCN
DCM
THF
InCl3/-30
InCl3/-30
InCl3/-30
InCl3/-30
InCl3/-30
InCl3/-70
InCl3/25
SnCl2/-30
SnCl2/-30
SnCl2/-30
SnCl2/25
4:1:0.7b,c
4.5:1:2.2b,c
2.35:1:1b
d
d
d
1:2:3d,e
2.1:1:0.6b
3.3:1:0.4b
3.6:1:2.9b
4:1:3.1b
2.5:1:1.1b
9
10
11
2.2:1:1.2d,e
1.8:1:5.5d
4.6:1:0.85d,f
9
10
11
aReaction conditions: MeCN, -30 °C, 1 h, 3a/4a/cat., 1:1:05. bReac-
tion time: 24 h. cRecovered starting materials. dDiastereomeric ratio
calculated by 1H NMR. eDiastereomeric ratio calculated by HPLC.
fIsolated compounds.
MeCN
b
aReaction conditions: 1 h, 3a/4a/cat., 1:1:05. Diastereomeric ratio
calculated by 1H NMR. cDiastereomeric ratio calculated by HPLC.
dRecovered starting materials.
stereochemical result appears very interesting since our
challenge was to obtain the syn adducts 7.
InCl3 (entry 5), SnCl2 (entry 6), and MgBr2 (entry 11), which
gave a mixture of 5a and 6a in comparable yield (InCl3, 65%;
SnCl2, 67%; MgBr2, 66%; isolated compounds).
In order to improve the diastereoselection of the process,
further studies were performed by evaluating the effect of the
variation of the solvent, temperature, and the stoichiometry
of both the catalyst and the reagents. According to the above
results, InCl3 and SnCl2 were selected, and their efficacy was
evaluated in different solvents (Table 2). A strong solvent
effect is evident for InCl3 (entries 1-5). Replacing MeCN by
DCM, the diastereomeric ratio increased, but decomposition
of 4a to lactone 2a was also observed even if anhydrous
conditions were employed. The use of a coordinating solvent
such as THF decreased the diastereoselection, while DMA
and toluene were ineffective in the formation of the reaction
products. A similar trend was observed using SnCl2 with the
exception of THF, which increased the diastereoselection
but also the decomposition of the reagent 4a.
The deprotection of the hydroxy and carboxylic functions
was performed at once, according to a known procedure,8d
by using trimethylsilylchloride in methanol. Starting from
pure compounds 5a and 6a, methyl esters 2R*,3S*-7a and
2R*,3R*-8a (Scheme 4) were obtained, respectively, whose
stereochemistry was confirmed by comparing their 1H NMR
spectra with those of the known compounds.7l This allowed
us to unequivocally assign the relative configuration to the
corresponding stereocenter on intermediates 5a and 6a. This
SCHEME 4. Methanolysis of Adducts 5 and 6a
No positive effects on de were found by increasing the
temperature (entries 7 and 11), while when operating at lower
temperature (entry 6) in DCM the reaction failed.
As reported in Table 3, decreasing the stoichiometry of the
catalyst induced a negative effect on the diastereoselection.
Finally, experiments using a 2:1 or 1:2 ratio of 3a/4a were
carried out in MeCN at -30 °C, but similar results were
found without improvement in diastereoselection (5a/6a/2a,
3.6:1:0.5; 4.1:1:2.9, respectively).
In summary, the above results indicated that the best
reaction conditions giving the highest diastereoselection
and minimizing the degradation of 4 are as follows: 1:1 ratio
aReagents and conditions: (i) MeOH/Me3SiCl (0.5 M), 25 °C, 10 min. (ii)
InCl3, MeCN, -30 °C, 1 h, then (i).
J. Org. Chem. Vol. 75, No. 21, 2010 7101

Gassa et al.
JOCArticle
TABLE 3. Synthesis of 5/6a: Influence of Catalyst Stoichiometry on
Diastereoselection
on the starting amine is of importance since a decrease of the
de was found when a N-alkyl substituent (entries 4 and 5) was
used instead of the benzyl one.
entrya
catalyst (eqiv)
5a:6a:2a
4:1:0.7b,c
To evaluate the bulkiness of the Ley’s lactone ether
functions, the ketene triethylsilyl acetal 4b was made to react
with 3a (Scheme 3). Products 5f and 6f were obtained with an
improved diastereoselectivity (de 83%; Table 4, entry 6). The
isolation of compounds 5/6f was quite difficult despite the
1
2
3
4
5
InCl3 (0.5)
InCl3 (0.2)
InCl3 (0.1)
SnCl2 (0.5)
SnCl2 (0.2)
2.6:1:0.6b,c
1.23:1:0.8b
3.3:1:0.4b
2.1:1:0.7b
aReaction conditions: 1 h, 3a/4a: 1:1, MeCN, -30 °C. bDiastereo-
meric ratio calculated by 1H NMR. cDiastereomeric ratio calculated by
HPLC.
1
fact that the H NMR and HPLC analyses of the crude
reaction mixture showed a high purity. Actually, compound
5f was isolated in very low yields (20-30%) after chroma-
tography on silica gel or on neutral alumina, thus indicating
that the ethoxy lactone is unstable under these conditions.
Pure 5f was successfully isolated in 74% yield by using flash
cartridge chromatography. Similar results were obtained
from 4a and imine 3b, and isomers 5g and 6g were obtained
(Table 4, entry 6). The conversion of intermediate 5f into
ester 7a confirmed the assigned stereochemistry (Scheme 4).
The possibility to use imines containing a chiral auxiliary
was evaluated, and both N-benzylidene-(R)-tert-butylsulfin-
amide14 and N-benzylidene-(S)-R-methylbenzylamine (3f)15
were made to react with 4a and 4b, respectively. The first
imine did not give the condensation products. Instead, by
using 3f, the formation of three stereoisomers was detected
by 1H NMR (4:5.4:1). Their purification on silica gel allowed
isolation of the main quasi-enantiomers 5 h/50h, which are
not separable, in good yield (62%), but in low diastereoselection
(de 14%). The third diastereomer 6h was also isolated in 8%
yield (Scheme 5, see Supporting Information for discussion).
The methanolysis of 5 h/50h gave the corresponding isoserine
esters 2R,3S-7h (minor isomer) and 2S,3R-70h (major isomer),
not separable, belonging to the syn series, as confirmed by
literature.7n Accordingly, 6h was transformed into 8h, which
was characterized by 2S,3S absolute configuration.7n In this
of 3a/4a, in the presence of 0.5 molar equiv of InCl3 or SnCl2,
at -30 °C in MeCN.
To verify the generality of this synthetic protocol and to
evaluate the steric effect of the imine substituent on the diaste-
reoselection, a series of imines 3b-e was prepared according to
general procedures. Their reaction with 4a under standard
conditions gave the expected compounds 5b-e and 6b-e
(Scheme 3) as reported in Table 4 (entries 2-5).
TABLE 4. Yields and Diastereomeric Distribution of Compounds 5/6
1
case, a partial epimerization occurred as shown by H NMR
spectrum.
Aiming to ameliorate the synthetic protocol, further ex-
periments were done by reducing the synthetic steps without
intermediate isolation.
The direct transformation of compounds 5/6, derived
from Mannich condensation, to the corresponding esters 7
and 8 was performed. As an example, the condensation of 3a
and 4b followed by methanolysis gave pure diastereomer 7a
(70%) (Scheme 4).
aMethod A (see Experimental Section). bMethod B (see Experimental
Section). cIsolated compounds (mixture of diastereomers). dYield re-
ferred to the isolated main isomer 5. eDiastereomeric ratio calculated via
1H NMR. fDiastereomeric ratio calculated via HPLC.
In general the de ranged between 60% and 69%, and the
substituent on the starting aldehyde seemed to have little
influence on the diastereoselection. Instead, the substituent
SCHEME 5. Adducts 5/50/6h and Enantiopure Isoserine Derivatives 7/70/8ha
aReagents and conditions: (i) InCl3, MeCN, -30 °C, 1 h. (ii) MeOH/Me3SiCl (0.5 M), 25 °C, 10 min.
7102 J. Org. Chem. Vol. 75, No. 21, 2010

Gassa et al.
JOCArticle
A three-component reaction starting from an aldehyde,
benzylamine, and 4a was then studied. Imine 3a was gener-
ated in situ from benzaldehyde in MeCN and in the presence
of molecular sieves. After 1 h, the mixture was cooled to
-30 °C, and then the catalyst and 4a were added. Com-
pounds 5a/6a were isolated with an improved overall yield
and diastereoselection (5/6a: 87%, 4.4:1). In an analogous
way, staring from isovaleraldehyde, via intermediate 3c,
diastereomers 5c/6c (87%, 5.3:1) were obtained.
The final goal was the direct synthesis of esters 7a-g without
any intermediate isolation (Scheme 6). According to the above
protocol, by using a series of aldehydes, benzylamine, and 4b as
starting reagents, intermediates 5/6 were generated. Finally, the
crude reaction mixture was quenched with a MeOH/TMSCl
solution, and the pure main syn isomers 7a-g were isolated in
good yields and diastereoselection.
tion states leading to diastereomers being compared.16g,h This
decision was also driven by our previous experience with
modeling reactions paths where weak interactions were
found to be important in determining the final regiochemical
outcome17 and was verified through preliminary calculations
performed with the B3LYP functional, which provided poor
results compared to experiments.
The R2 substituent did not significantly affect the regio-
chemical outcome, as experimentally evidenced in Table 4.
Indeed, comparable 5/6 ratios were observed for entry 1
(4.1:1, R2=Bn) and entry 5 (3.9:1, R2=Me), evidence also
explained by conformational searches performed on 5a and
6a at the molecular mechanic level, showing that the benzyl
group was located far away from the reaction centers in all
conformers lying in an energy range of 5 kcal/mol from the
most stable one. For the above reasons, the most computa-
tionally comfortable reaction of 3e with 4a was chosen as a
representative model. For simplicity, the Lewis acid was
modeled as an hydrogen cation, thus avoiding the use of
complex basis sets that would be necessary for the correct
treatment of post-transition metals,18 while the triethylsilyl
moiety of 4a was replaced with a trimethylsilyl group to
reduce CPU time and improve convergence. This approx-
imation is not expected to affect the model predictability, as
preliminary calculations conducted by replacing the alkylsi-
lyl moiety with a hydrogen, thus making the location of
transition states easier, provided comparable results in terms
of diastereoselectivity predictions. The modeled reactants,
transition states (TSs), and products will be hereafter re-
ferred as 3e (imine) and M4a, TS-M5e and TS-M6e (where
the “M” indicates that the trimethylsilyl has been used
instead of the triethylsilyl group, as explained above), and
5e (2R,3S stereochemistry) and 6e (2R,3R stereochemistry),
respectively. The solvent was included in both geometry
optimizations and energy calculations through the CPCM
solvation model for MeCN.
Reactants, TSs, and products were fully optimized at the
CPCM-MPW1B95/6-31þG(d,p) level, while single point
energy evaluations were conducted with the more sophisti-
cated 6-311þG(3df,2p) basis set. Several conformations
were optimized for either reactants, TSs, and products, but
only the most favored conformations will be reported and
discussed. The geometry of the located TSs is represented in
Figure 1, and relative energies and relevant distances are
reported in Table 5.
As reported in Table 5, TS-M5e was found to be 3.2 kcal/
mol more stable than TS-M6e, while reaction energies show
that product 6e is thermodynamically favored over 5e by
1.1 kcal/mol. Equilibrium between 5 and 6 could be possible
through a retrocondensation reaction, but the low reaction
temperatures experimentally adopted, together with the
high-energy barrier expected for the retrocondensation,
suggest that the reaction is mainly under kinetic control.
This observation is supported by the fact that conducting the
SCHEME 6. Direct R-Hydroxy-β-amino Esters 7 Synthesis by
a Three-Component Reactiona
aReagents and conditions: (i) activated molecular sieves, MeCN, 25 °C,
1 h, then 4b, InCl3, -30 °C, 1 h. (ii) MeOH/Me3SiCl (0.5 M), 25 °C,
10 min.
Theoretical Investigation of the Reaction Mechanism
The stereochemical result of the condensation of com-
pounds 4 and the imines 3 differs from that found for the
products obtained by the condensation of Ley’s lactone
and aldehydes in basic conditions, where the major dia-
stereomer is the anti adduct, while in our case the major
diastereomer is the syn adduct. Interestingly, the reported
acid-catalyzed condensation of imines and ketene silyl
acetals afforded syn isoserine derivatives7k by using the Z
ketene silyl acetals, which have an opposite geometry with
respect to compounds 4.
To investigate the origin of the observed diastereoselec-
tivity, the reaction mechanism of the key step addition of
compound 4 to imine 3 was investigated through DFT
calculations. Every calculation was performed with the
recently developed MPW1B95 functional,16 as the popular
B3LYP functional evidenced some shortcomings, especially in
estimating reaction barrier heights and in describing van der
Waals interactions, which might be important when two transi-
(14) Liu, G.; Cogan, D. A.; Owens, T. D.; Tang, T. P.; Ellman, J. A.
J. Org. Chem. 1999, 64, 1278–1284.
(15) Rogalska, E.; Belzecki, C. J. Org. Chem. 1984, 49, 1397–1402.
(16) (a) Dybala-Defratyka, A.; Paneth, P.; Pu, J.; Truhlar, D. G. J. Phys.
Chem. A 2004, 108, 2475–2486. (b) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A
2004, 108, 6908–6918. (c) Boese, A. D.; Martin, J. M. L. J. Chem. Phys. 2004,
121, 3405–3416. (d) Zhao, Y.; Pu, J.; Lynch, B. J.; Truhlar, D. G. Phys. Chem.
Chem. Phys. 2004, 6, 673–676. (e) Zhao, Y.; Truhlar, D. G. Phys. Chem.
Chem. Phys. 2005, 7, 2701–2705. (f ) Pu, J.; Truhlar, D. G. J. Phys. Chem. A
2005, 109, 773–778. (g) Lynch, B. J.; Zhao, Y.; Truhlar, D. G. J. Phys. Chem.
(17) (a) Contini, A.; Leone, S.; Menichetti, S.; Viglianisi, C.; Trimarco, P.
J. Org. Chem. 2006, 71, 5507–5514. (b) Borsini, E.; Broggini, G.; Contini, A.;
Zecchi, G. Eur. J. Org. Chem. 2008, 2808–2816. (c) Aversa, M. C.; Barattucci,
A.; Bonaccorsi, P.; Contini, A. J. Phys. Org. Chem. 2009, 22, 1048–1057.
(18) This choice was supported by the experimental observation that the
reaction proceeds with a similar diastereoselection, even though with very
low yield, by using different protic acids (i.e., p-TSA and 1,10-binaphthyl-
2,20-diyl hydrogenphosphate, see Supporting Information) as the catalyst.
ꢁ
A 2005, 109, 1643–1649. (h) Zhao, Y.; Gonzalez-Garcıa, N.; Truhlar, D. G.
´
J. Phys. Chem. A 2005, 109, 2012–2018. (I) Zhao, Y.; Truhlar, D. G. J. Phys.
Chem. A 2005, 109, 5656–5667.
J. Org. Chem. Vol. 75, No. 21, 2010 7103

Gassa et al.
JOCArticle
group (see d4 distance, Figure 1, 2.83 A), but this interac-
tion can only be operative when R1 is aromatic (entries 1,
2, 4-7, Table 4), while a good diastereoselection was also
obtained for entry 3 (5/6c = 5:1, Table 4) where R1=i-Bu.
More reasonably, the lower TS-M5e energy with respect to
TS-M6e is due to the different steric hindrance occurring
at the TS levels. In fact, the d3 distance (which describes
the hindrance occurring between the R3 group of 4a and R1
or R2 for TS-M5 and TS-M6, respectively) measured
between the C5-linked methoxy group and the C1-phenyl
and N-methyl carbons is 4.13 and 3.72 A, respectively.
This observation is experimentally supported by the fact
that by replacing the methoxy with an ethoxy group the
obtained diastereoselection was greatly enhanced (entries 6
and 7, Table 4). Moreover, these findings are also supported by
the modest diastereoselection observed when the chiral imine
3f (R2 = (S)-CH(Me)Ph) was used in the formation of isomers
5h/5h0. The far chiral moiety in the TS-M5e-like TS gave a low
diastereoselection in the syn series. Instead, a single enantiomer
was obtained in the anti series, according to the TS-M6e-
like TS.
FIGURE 1. Structure of TS-M5e and TS-M6e. Displacement vec-
tors for the unique imaginary frequency are shown, together with
relevant distances.
Conclusion
In conclusion, a very efficient three-component synthesis of
a series of 2R*,3S*-R-hydroxy-β-amino esters 7, obtained in
high diastereoselection and yields, was realized. This synthetic
protocol made it possible to obtain the target compounds
while avoiding the isolation of the single intermediates.
The key reaction step has been modeled and is predicted to
be under kinetic control, with the observed diastereoselec-
tion resulting mainly from the steric factors within the two
different TSs.
TABLE 5. Relative Activation and Reaction Energies (kcal/mol) and
a
˚
Selected Distances (A) for TS-M5e and TS-M6e
5e
6e
ΔEq
ΔE
d1
5.0
-25.1
8.2
-26.2
2.23
2.03
d2
d3
2.73
4.13
3.34
3.72
aRelative energies computed as the sum of total free energies in
solution obtained from single point calculations with the 6-311þ(3df,2p)
basis set and ZPE correction resulted from thermochemical analyses
with the 6-31þG(d,p) basis. Reactant’s energy is the sum of isolated
reactants energies. Relative energies have been computed for the bal-
anced reaction by including a water molecule in the reactants and
trimethylsilanol in the products.
Experimental Section
Theoretical Calculations. Products 5e and 6e were initially
obtained from a conformational search performed in vacuum at
the molecular mechanic level using the MMFF94s force field
implemented in MOE.19 All conformers within 3 kcal/mol from
the most favored were optimized at the MPW1B95/6-31þG(d,p)
level of theory,16 and only the most stable for each product was
further considered. Reactants 3e and M4a and transition states
TS-M5e and TS-M6e were then constructed and fully optimized
at the same level of theory. Several conformations and relative
orientations of reacting groups were considered for each TS, but
only the lowest energy structures were further considered.
Vibrational frequencies were computed at the same level of
theory in order to define optimized geometries as minima (no
imaginary frequencies) or TSs (a unique imaginary frequency
corresponding to the vibrational stretching of the forming/
breaking bonds) and to calculate ZPVE and thermochemical
corrections to electronic energies (1 atm, 298.15 K, unscaled
frequencies). Single point calculations were performed at the
MPW1B95/6-311þG(3df,2p). All calculations were performed
in solution (MeCN) using the CPCM solvent model,20 and
basis sets using Cartesian d and f functions were always
requested. The default united atom topological model UA0,
implemented in Gaussian09,21 was adopted for the construction
reaction at higher temperature results in a loss of diastereos-
electivity (see Table 2).
Concerning geometrical parameters, the forming C-C
bond length (d1, Figure 1) resulted 2.23 A and 2.03 A in
the syn TS-M5e and in the anti TS-M6e, respectively. Thus,
according to the Hammond’s postulate, TS-M6e is “later”
(higher activation barrier, shorter length for the forming
bonds) than TS-M5e. Conversely, the d2 distance between
the enolether oxygen and the nitrogen-coordinated Lewis
acid (modeled as a proton) was 2.73 A for TS-M5e and
3.34 A for TS-M6e. Accordingly, as shown in Figure 1,
although both TSs assume a fused six-membered ring bicy-
clic conformation with a cis junction at the C2-C3 carbons
of 4a, the energetically favored TS-M5e adopts a rather
regular “chair-like” conformation, with quite similar d1 and
d2 distances, while the least favored TS-M6e adopts
a distorted “boat-like” conformation with a difference
between d1 and d2 of more than 1 A. This means that in
TS-M5e a stabilizing interaction involving all three
N-H-O atoms occurs, whereas in TS-M6e this interac-
tion is lost. The highest stability of TS-M5e with respect to
TS-M6e could be ascribed to a stabilizing C-H/π inter-
action between the methoxy hydrogens and the phenyl
(19) MOE V. 2009.10; Chemical Computing Group Inc.: Montreal,
Canada, http://www.chemcomp.com.
(20) Cossi, M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem.
2003, 24, 669–681.
(21) Frisch, M. J. et al. Gaussian 09, Revision A.02; Gaussian, Inc.:
Wallingford, CT, 2009.
7104 J. Org. Chem. Vol. 75, No. 21, 2010

Gassa et al.
JOCArticle
of the molecular cavity (an extra sphere was added on the proton
modeling the Lewis acid in TS structures). All quantum-chemi-
cal calculations were performed with the Gaussian09 software
package.
together with diastereomer 6h. Method B. In a two-necked
round-bottomed flask equipped with a magnetic stirring
bar and a nitrogen inlet, the aldehyde (benzaldehyde, 60 μL,
0.59 mmol; isovaleraldehyde, 63 μL, 0.59 mmol) and amine
(benzylamine, 79 μL, 0.59 mmol) were dissolved in MeCN
(1.5 mL) in the presence of molecular sieves (60 mg, activated
at 200 °C in vacuum for 2 h). After 1 h, the reaction mixture was
cooled to -30 °C, and anhydrous InCl3 (65.3 mg, 0.29 mmol)
was added in one portion. After stirring at this temperature for
10 min, a solution of ketene triethylsilyl acetal 4a (180.2 mg,
0.59 mmol) in dry MeCN (1 mL) was added dropwise. The
reaction mixture was stirred for 1 h and worked up as reported
in Method A. Compounds 5/6a and 5/6c were isolated,
respectively.
(3R*,5S*,6S*)-3-(10-N-Benzylamino-10-phenyl-methyl)-5,6-di-
methoxy-5,6-dimethyl[1,4]dioxan-2-one (5a, 6a). Column chro-
matography: AcOEt/cyclohexane, 1:5. Method A: 5a, 52%; 6a,
13%. Method B: 5a, 71%; 6a, 16%.
Data for 10S*-5a. 108 °C (n-pentane/CH2Cl2). IR (KBr) νmax
3372, 1744 cm-1; 1H NMR (CDCl3) δ 7.40-7.22 (m, 10H), 4.31,
4.28 (AB system, J = 2.9, 2H), 3.60, 3.51 (AM system, J = 13.2,
2H), 3.21 (s, 3H), 3.14 (s, 3H), 3.00-2.00 (br, 1H, exch.), 1.47 (s,
3H), 1.37 (s, 3H); 13C NMR (CDCl3) δ 169.2, 140.8, 139.7,
128.7, 128.6, 128.5, 127.2, 105.1, 98.6, 75.7, 63.6, 51.1, 50.1, 49.5,
18.4, 17.2. MS (ESI) m/z 386.0 [M þ 1]þ. Anal. Calcd for C22-
H27NO5: C, 68.55; H, 7.06; N, 3.63. Found: C, 68.30; H, 7.24;
N, 3.50.
Data for 10R*-6a. Pale yellow oil. IR (NaCl) νmax 3318,
1748 cm-1; 1H NMR (CDCl3) δ 7.41-7.26 (m, 10H), 4.54, 4.30
(AB system, J = 2.7, 2H), 3.75, 3.56 (AM system, J = 13.2, 2H),
3.33 (s, 3H), 2.77 (s, 3H), 2.76 (brs, 1H, exch.), 1.37 (s, 3H), 1.36
(s, 3H); 13C NMR (CDCl3) δ 168.1, 140.5, 138.4, 129.7, 128.7,
128.2, 127.8, 127.4, 105.1, 98.6, 74.7, 63.4, 51.5, 49.6, 49.5, 18.3,
17.2. MS (ESI) m/z 386.2 [M þ 1]þ. Anal. Calcd for C22H27NO5:
C, 68.55; H, 7.06; N, 3.63. Found: C, 68.38; H, 7.19; N, 3.51.
General Procedure for the Methanolysis of Compounds 5 and
6. Compound 5a,f,h or 6a,h (0.225 mmol) was dissolved in a 0.5
M solution of TMSCl in MeOH (1.0 mL, 0.5 mmol) under
stirring at 25 °C for 10 min. The solvent was evaporated, and the
residue was crystallized, giving pure compound (7a: 97% from
5a, 95% from 5f; 8a: 80% from 6a; 7/70h: 84% from 5/50h; 8h:
70% from 6h).
One-Pot Preparation of Methyl 3-(Amino)-2-hydroxy-propio-
nate Derivatives. Method C. The reaction between 3a and 4b was
performed according to Method A. The 1H NMR analysis of the
crude reaction mixture showed the presence of the diastereomer
5f and only traces amount of 6f (HPLC: ASCENTIS SI, 3 μm,
150 ꢀ 4.6 mm, 0.8 mL/min, λ = 210 nm, n-hexane/iPrOH, 98:2;
5f/6f, 92:8). The crude reaction mixture was treated with
MeOH/TMSCl according to the above-reported procedure,
and methyl ester derivative 7a (60%) was obtained after recrys-
tallization. A further batch of compound 7a (10%) was isolated
after column chromatography on silica gel (cyclohexane/
AcOEt, 4:1). Total yield: 70%. Method D. Operating as reported
in Method B, starting from the corresponding aldehyde
(0.59 mmol) and benzylamine (79 μL, 0.59 mmol), imines 3
were generated, and then compound 4b (196.4 mg, 0.59 mmol)
was added. The crude reaction mixture containing compounds 5/6
was treated with MeOH/TMSCl according to the above-reported
procedure, and methyl ester derivatives 7a-g were isolated after
flash silica gel column chromatography (7a,f,g: AcOEt/cyclohex-
ane, 1:4; 7b: AcOEt/nhexane, 1:6; 7c: Et2O/n-hexane, 1:4; 7d,e:
AcOEt/cyclohexane, 1:3).
One-Pot Procedure for the Preparation of 5,6-Diethoxy-5,6-
dimethyl[1,4]dioxan-2-one (2b). 2,3-Butandione (4.6 mL, 46.22
mmol) and glycolic acid (3.07 g, 40.40 mmol) were dissolved in
triethyl orthoformate (40 mL), and a catalytic amount of H2SO4
was added. The reaction mixture was heated at 50 °C for 1 h. A
saturated solution of NaHCO3 (10 mL) was added, and the
mixture was extracted with AcOEt (3 ꢀ 30 mL). The organic
layers were dried over Na2SO4 and the crude reaction mix-
ture was chromatographed on silica gel (cyclohexane/AcOEt,
10:1), affording pure compound 2b (52%) after crystallization.
White solid, mp 37 °C (CH2Cl2/pentane, 0 °C). IR (NaCl) νmax
1752 cm-1; 1H NMR (CDCl3) δ 4.32, 4.23 (AB system, J = 16.6,
2H), 3.79-3.69 (m, 2H), 3.58 (q, J = 7.2, 2H), 1.53 (s, 3H), 1.41
(s, 3H), 1.21 (t, J = 6.9, 3H), 1.99 (t, J = 7.0, 3H); 13C NMR
(CDCl3) δ 168.1, 105.2, 97.8, 60.5, 58.8, 57.3, 18.8, 17.9, 15.7,
15.3. MS (ESI) m/z 341.4 [M þ 23]þ. Anal. Calcd for C10H18O5:
C, 55.03; H, 8.31. Found: C, 54.78; H, 8.52.
5,6-Dimethoxy-5,6-dimethyl-([1,4]dioxen-2-yloxy)triethylsilane
(4a). Lactone 2a (3.23 g, 17 mmol) was dissolved in anhydrous
THF (35 mL), and the mixture was cooled to -78 °C under N2.
A solution of LHMDS (4.52 g, 27 mmol) in THF (10 mL)
was added dropwise. After stirring for 10 min, TESCl (4 mL,
26 mmol) was added. The solution was warmed to 25 °C, and
the stirring was continued overnight. THF was removed under
reduced pressure, and n-pentane (50 mL) was added to the
mixture. A solid was formed and filtered through Celite. After
solvent evaporation, the crude compoundwas distilled in vacuum
(120 °C, 0.8 mmHg). Pure compound 4a was obtained (4.26 g,
82%) as a colorless oil. IR (NaCl) νmax 1719, 1149, 740 cm-1; 1H
NMR (CDCl3) δ 5.52 (s, 1H), 3.37 (s, 3H), 3.22 (s, 3H), 1.45 (s,
3H), 1.38 (s, 3H), 1.06-0.94 (m, 9H), 0.75-0.67 (m, 6H); 13C
NMR (CDCl3) δ 143.9, 104.3, 96.6, 90.4, 49.5, 48.6, 17.6, 17.1,
6.6, 4.8. MS (ESI) m/z 327.1 [M þ 23]þ. Anal. Calcd for C14H28-
O5Si: C, 55.23; H, 9.27. Found: C, 55.05; H, 9.10.
5,6-Diethoxy-5,6-dimethyl-([1,4]dioxin-2-yloxy)triethylsilane
(4b). Silyl derivative 4b (4.8 g, 85%) was prepared according to
the above synthetic protocol except for the base that was added
to the mixture of 2b (3.7 g, 17 mmol) and TESCl. Colorless oil
(130 °C, 0.8 mmHg). IR (NaCl) νmax 1721, 1149, 744 cm-1; 1H
NMR (CDCl3) δ 5.50 (s, 1H), 3.68-3.63 (m, 2H), 3.51 (q, J =
7.0, 2H), 1.46 (s, 3H), 1.40 (s, 3H), 1.22-1.03 (m, 6H), 1.02-
0.88 (m, 9H), 0.75-0.63 (m, 4H), 0.55 (q, J = 10.8, 2H);
13C NMR (CDCl3) δ 143.8, 104.3, 100.2, 96.4, 57.4, 56.6,
18.4, 17.9, 15.9, 15.7, 6.6, 4.9. MS (ESI) m/z 355.2 [M þ 23]þ.
Anal. Calcd for C16H32O5Si: C, 57.79; H, 9.70. Found: C, 57.60;
H, 9.58.
Preparation of Compounds 5 and 6. Method A. Imine 3
(0.66 mmol) was dissolved in dry MeCN (1.2 mL) under nitro-
gen and stirring. The reaction mixture was cooled at -30 °C, and
anhydrous InCl3 (73 mg, 0.33 mmol) was added in one portion.
After stirring at this temperature for 10 min, a solution of silyl
derivative 4 (0.66 mmol) in dry MeCN (1 mL) was added
dropwise. In the case of imine 3c, InCl3 was added to the
solution of the two reagents. The reaction mixture was stirred
for 1 h and then quenched with a saturated solution of NaHCO3
(1 mL). The crude material was extracted with AcOEt (3 ꢀ
5 mL), and the collected organic phases were washed with brine
and dried over Na2SO4. The solvent was removed in vacuum,
and the crude material was purified by silica gel flash chroma-
tography. Compounds 5/6f, 5/6g, and 5/6h were chromatog-
raphed by flash cartridge chromatography (SiO2; n-hexane/
Et2O, 7:2; flow, 30 mL/min). In the first two cases, only isomers
5f and 5g were isolated, respectively. A mixture of quasi-
enantiomers 5/50h, which cannot be separated, was isolated
Methyl 3-(Benzylamino)-2-hydroxy-3-phenyl-propionate (7/8a):
de 83%. Data for (2R*,3S*)-7a. 74%. Mp 107 °C (n-pentane/
Et2O), (107-108 °C).7l 1H NMR (CDCl3) δ 7.41-7.22 (m, 10H),
4.26, 3.95 (AX system, J = 4.1, 2H), 3.77, 3.49 (AM system, J =
13.2, 2H), 3.70 (s, 3H), 1.60 (br, 2H, exch.).
J. Org. Chem. Vol. 75, No. 21, 2010 7105

Gassa et al.
JOCArticle
Data for (2R*,3R*)-8a. 10%. Mp 99 °C (n-pentane/CH2Cl2),
(98-99 °C).7l 1H NMR (CDCl3) δ 7.38-7.22 (m, 10H), 4.54,
4.06 (AX system, J = 4.0, 2H), 3.78, 3.61 (AM system, J = 12.8,
2H), 3.60 (s, 3H), 3.00-2.00 (br, 2H, exch.).
Supporting Information Available: Spectroscopic discussion
for compounds 5/6a. Experimental: general and starting mate-
rials. Column chromatography conditions and spectroscopic
characterization of 5/6b-h, 6b-e, 7b-h, and 8h. Further com-
putational details: Cartesian coordinates and energies of the
mentioned stationary points. Complete ref 21. Further discus-
sion for the reaction of 4b with imine 3f. Copies of 1H and 13
C
Acknowledgment. Funding for this work was provided by
Indena S.p.A. A.C. is grateful to the Consorzio Interuniver-
sitario Lombardo per l0Elaborazione Automatica (CILEA)
for computational facilities.
NMR spectra. HPLC spectra for crude reaction mixtures
affording 5a/6a and 5f/6f. This material is available free of
charge via the Internet at http://pubs.acs.org.
7106 J. Org. Chem. Vol. 75, No. 21, 2010