Mechanistic investigations on the highly stereoselective allylation of aldehydes with a norpseudoephedrine derivative

Article abstract of DOI:10.1021/ja974390+

Reaction of aliphatic aldehydes 1 and allylsilane 3 at -78 °C in the presence of the norpseudoephedrine derivative 2a and catalytic amounts of trifluoromethanesulfonic acid trimethylsilyl ester gives the homoallylic ethers 4 with >98% diastereoselectivity

Full text of DOI:10.1021/ja974390+

4276
J. Am. Chem. Soc. 1998, 120, 4276-4280
Mechanistic Investigations on the Highly Stereoselective Allylation of
Aldehydes with a Norpseudoephedrine Derivative
Lutz F. Tietze,* Christian Wulff, Christoph Wegner, Ansgar Schuffenhauer, and
Kai Schiemann
Contribution from the Institut fu¨r Organische Chemie der Georg-August UniVersita¨t, Tammannstrasse 2,
D-37077 Go¨ttingen, Germany
ReceiVed December 29, 1997
Abstract: Reaction of aliphatic aldehydes 1 and allylsilane 3 at -78 °C in the presence of the norpseudo-
ephedrine derivative 2a and catalytic amounts of trifluoromethanesulfonic acid trimethylsilyl ester gives the
homoallylic ethers 4 with >98% diastereoselectivity. The ethers 4 can be transformed into the corresponding
homoallylic alcohols 5 having an enantiomeric excess >98% using sodium in liquid ammonia. On-line NMR
spectroscopy indicates that the mixed acetal 7 and the oxazolidinium ion 12 are intermediates in the formation
of 4. At higher temperature proton transfer occurs to give the oxazolidinium ion 18, which does not react
with allylsilane, but gives the oxazolidine 17 on aqueous workup. Protonation of 17 with trifluoromethane-
sulfonic acid leads to 18 but not to the oxazolidinium ion 12. Thus, an allylation starting from 17 is not
possible.
The allylation of aliphatic aldehydes 11 with allylsilane 3 in
Scheme 1
the presence of the norpseudoephedrine derivative 2a or ent-
2a and catalytic amounts of trifluoromethanesulfonic acid
trimethylsilyl ester (TMSOTf) is a very powerful procedure for
the synthesis of enantiopure homoallylic alcohols 5 (Scheme
1).2,3 Under these conditions nearly all aliphatic aldehydes can
be transformed into the corresponding homoallylic ethers 4 with
a selectivity of >99:1. Subsequent reductive removal of the
chiral auxiliary gives the enantiopure homoallylic alcohols 5
together with the oxazolidine 17 as a side product. In similar
fashion, ketones such as ethyl methyl ketone can be transformed
into the corresponding tertiary homoallylic ethers with an
astonishing diastereoselectivity (ds) ) 96:4. However, in this
case, the opposite configuration of the newly formed stereogenic
center is found and the formation of an oxazolidine as a
byproduct is not observed.4
The high stereocontrol in these reactions is quite surprising
because the key bond-forming step takes place in an acyclic
system and chelation can be excluded due to the use of the
monodentate Lewis acid TMSOTf for aldehydes and the
Brønsted acid trifluoromethanesulfonic acid (TfOH) for ketones.
In this paper, we describe our investigations on the mechanism
of the allylation of aldehydes using on-line NMR spectroscopy
and model reactions of possible intermediates. In addition, the
influence of the catalyst and the reaction temperature on the
selectivity and the yield of the allylation are described.
For the allylation of aldehydes 1, it can be assumed that at
first an oxenium ion (6) is formed, which reacts with the
norpseudoephedrine derivative 2a to give the mixed acetal 7
(Scheme 2).
This can undergo a cyclization to yield the oxazolidinium
ions 11 or 12 either via formation of the oxenium ion 9 and 10,
respectively, or by direct substitution of 7. The final step is
the irreversible substitution of the allylsilane 3 at C-3 of 11 or
12 to give the homoallylic ether 4 with inversion of configu-
ration of the stereogenic center as in 11 or 12, which is in
agreement with the experimental results. The high stereo-
selectivity in the synthesis of 4 makes it quite unlikely that a
direct addition of 3 to 9 or 10 takes place to give 4.
Furthermore, we assume that 12, not 11, is the intermediate in
the described process; two facts led to this assumption. First,
(1) (a) Tietze, L. F.; Do¨lle, A.; Schiemann, K. Angew. Chem. 1992, 104,
1366; Angew. Chem., Int. Ed. Engl. 1992, 31, 1372. (b) Tietze, L. F.;
Schiemann, K.; Wegner, C.; Wulff, C. Chem. Eur. J. 1996, 2, 1164.
(2) For a review, see: Yamamoto, Y.; Asao, N. Chem. ReV. 1993, 93,
2207.
(3) For selective allylations of aldehydes, see also: (a) Stu¨rmer, R.;
Hoffmann, R. W. Synlett 1990, 759. (b) Brown, H. C.; Randad, R. S.; Bhat,
K. S.; Zaidlewicz, M.; Racherla, U. S. J. Am Chem. Soc. 1990, 112, 2389.
(c) Roush, W. R.; Banfi, L. J. Am Chem. Soc. 1988, 110, 3979. (d) Corey,
E. J.; Yu, C.-M.; Kim, S. S. J. Am. Chem. Soc. 1989, 111, 5495. (e) Short,
R. P.; Masamune, S. J. Am. Chem. Soc. 1989, 111, 1892. (f) Hafner, A.;
Duthaler, R. O.; Marti, R.; Rihs, G.; Rothe-Streit, P.; Schwarzenbach, F. J.
Am. Chem. Soc. 1992, 114, 2321. (g) Bru¨ckner, R.; Weigand, S. Chem
Eur. J. 1996, 2, 1077. (h) Gauthier, D. E.; Carreira, E. M. Angew. Chem.
1996, 108, 2521. (i) Costa, A. L.; Piazza, M. G.; Tagliavini, E.; Trombini,
C.; Umani-Ronchi, A. J. Am. Chem. Soc. 1993, 115, 7001. (j) Keck, G. E.;
Tarbet, K. H.; Geraci, L. S. J. Am. Chem. Soc. 1993, 115, 8467. (k) Ishihara,
K.; Mouri, M.; Gao, Q.; Maruyama, K.; Furuta, K.; Yamamoto, H. J. Am.
Chem. Soc. 1993, 115, 11490.
(4) (a) Tietze, L. F.; Schiemann, K.; Wegner, C. J. Am. Chem. Soc. 1995,
117, 5851. (b) Tietze, L. F.; Wegner, C.; Wulff, C. Synlett 1996, 471.
S0002-7863(97)04390-4 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/28/1998

Highly StereoselectiVe Allylation of Aldehydes
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4277
Scheme 2
Scheme 5
Table 1. Reaction of Aldehydes 1 with 2a and TMSOTf
products: yield (%)
TMSOTf
(equiv)
aldehyde 1: R
react. time
2a
2b
17a-c
a: CH3
1.5
0.1
0.1
0.1
0.1
1.1
1.5
14 h
30 s
5 min
15 min
60 min
60 min
14 h
<1
a: 99
b: <1
b: <1
b: <1
b: 5
b: 88
c: 46
b: C2H5
b: C2H5
b: C2H5
b: C2H5
b: C2H5
c: C5H11
51
30
21
16
12
a
49
70
79
79
<1
a
Scheme 3
Scheme 4
a Yield not determined.
(Scheme 5). The relative configuration of 17a has been
confirmed by X-ray analysis and that of 17a-c by NOESY
spectra. Thus, it was thought that the stereochemistry of the
resulting homoallylic ethers 4 might easily be explained by
attack of allylsilane 3 at C-2 of the protonated N,O-acetal 12
by an SN2 process. However, reaction of isolated 17a with
allylsilane 3 in the presence of a strong acid or TMSOTf did
not lead to the desired homoallylic ether 4a. To clarify this
inconsistency and to get a better insight into the mechanism of
the allylation, we performed on-line NMR spectroscopy of the
reaction at various temperatures. In addition, the oxazolidines
17a-c were prepared and investigated by NMR spectroscopy
in the presence of acid and TMSOTf. For the synthesis of
oxazolidines 17a-c, the aldehydes 1a-c were treated with the
norpseudoephedrine derivative 2a in dichloromethane in the
presence of TMSOTf at -78 °C. Using more than 1 equiv of
TMSOTf, the oxazolidines 17a-c were formed in 99%, 88%,
and 46% yields, respectively, as single diastereomers. However,
with only 0.1 equiv of TMSOTf, the desilylated norpseudo-
ephedrine derivative 2b was the main product (Table 1). The
13C NMR spectra of 17a-c show two equilibrating species with
a coalescence temperature of about 25 °C. At -78 °C separated
signals were observed for the two isomers, whereas at 100 °C,
only one resonance per carbon was found.
it is known5 that protonated amides 14 preferentially exist as
their iminium-ion tautomers 13 (Scheme 3).
In addition, we have performed PM3-calculations, which
show that the heat of formation of the iminium-ion tautomers
(E)-12a and (Z)-12a is more than 6 kcal/mol lower than that of
the corresponding protonated amides (R)-11a and (S)-11a.
Thus, in the allylation a transition state of type 16 should be
more favorable than that of type 15. However, the calculations
do not allow to distinguish between the two double bond isomers
(E)-12a and (Z)-12a since they differ by only 0.3 kcal/mol
(Scheme 4).
The formation of 2b in the reaction of 1a and 2a with 0.1
equiv of TMSOTf could be explained by hydrolysis of 2a during
workup. However, this is unlikely since pure 2a is stable under
the reaction conditions. This would also not explain the
increased yield of 2b with prolonged reaction time. It must
therefore be assumed that the mixed acetal 7 is formed as a
first intermediate, which is stable in the absence of TMSOTf
and forms 2b via 8 during aqueous workup. A similar
intermediate resembling 7 was recently isolated by Polt in the
reaction of esters with DIBAH and (trimethylsilyl)imidazole.6
A second indication of the intermediacy of the oxazolidinium
ion 12 is the occurrence of the oxazolidines 17 as side products
(5) Fraenkel, G.; Franconi, C. J. Am. Chem. Soc. 1960, 82, 4478.

4278 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Tietze et al.
In the presence of excess TMSOTf 7 irreversibly gives the
oxazolidinium ion 12; this assumption is confirmed by our
investigations showing that the yield of the isolated oxazolidine
17 corresponds directly to the amount of TMSOTf added (Table
1). Formation of 12 from 7 can proceed via silylation at the
oxygen of the trimethylsilyloxy moiety followed by elimination
of hexamethyldisiloxane and ring-closure either by a concerted
process or via 10 as intermediate.7 Deprotonation of the ion
12 during aqueous workup would give the oxazolidine 17.
Scheme 6
Other methods for the synthesis of the oxazolidines,8 e.g.,
17a,d such as transacetalization of the dimethyl acetals of
acetaldehyde or o-bromobenzaldehyde in the presence of
catalytic amounts of the pyridinium salt of p-toluenesulfonic
acid were less effective; only 12% of 17a and traces of the
unstable 17d were obtained. Using the Lewis acid BF3‚OEt2
in toluene, no reaction occurred.
To prove the hypothesis that the oxazolidinium ion 12 is the
intermediate which leads to the homoallylic ethers 4, 19F and
13C NMR spectra were taken on reaction mixtures containing
2.0 equiv of acetaldehyde, 1.0 equiv of 2a, and 1.0 equiv of
TMSOTf at -78 °C in CH2Cl2/CD2Cl2. In the 19F NMR
spectra, two strong resonances of identical intensity at δ )
-75.14 and -75.31 were observed in addition to the signals of
the oxazolidine 17a at δ ) -69.81 and δ ) -70.80, of 2a at
δ ) -75.59, of TMSOTf at δ ) -77.11 and of TfOH (small
signal) at δ ) -78.46. We assume that these new signals
correspond to the two trifluoromethyl groups of the oxazoli-
dinium salt 12. This was confirmed by the 13C NMR data with
signals at δ ) 50.90 for C-4, δ ) 78.00 for C-5, δ ) 94.85 for
C-2, and δ ) 115.1 as well as δ ) 153.3 for the COCF3 group.
These signals are clearly different from those found for 18. It
can also be excluded from these data that the observed
intermediate is the O-silyl oxonium triflate derived from 7, since
simple monosilyl acetals6 show resonances at around δ ) 100
for the acetal carbon and for the oxonium ion a downfield shift
should be found. This is not in accordance with the signal at
δ ) 94.85 observed in the 13C NMR spectrum of the reaction
mixture. In addition, the oxonium ion of 7 would not lead to
the oxazolidine 17 by quenching with triethylamine.
The observed chemical shift values of 12 and 18 are worthy
of comment. A comparison of the 13C NMR shift values of
trialkylamines and of the corresponding tetraalkylammonium
bromides shows that the resonances of the R-carbons of the ions
appear downfield (∆δ ) 3-5 ppm), those of the â-carbons
upfield (∆δ ) 6-8 ppm) and those of the γ-carbons downfield
again (∆δ ) 1-2 ppm).9 Therefore, a protonation of the
nitrogen or oxygen as in 11, 12, and 18 should lead to a strong
downfield shift of the resonances of the R-carbons, whereas for
the â-carbons a strong upfield shift should be observed.
However, such an observation would only be true for separated
ion pairs. For contact ion pairs, the opposite is found due to
the effect of the anion10 as described by Pattaroni and
Lauterwein10a for a similar problem. Thus, we assume for 12
and 18 the existence of the contact ion pairs CIP-12 and CIP-
18, which is in accordance with the NMR data (Scheme 6).
Compound 12 can exist as two double-bond isomers.
Although PM3 calculations on the individual separated ions did
not allow us to rule out one of the two double-bond isomers,
the NMR data clearly indicate that the contact ion pair exists
preferentially as a single isomer since only one set of signals is
Having the pure oxazolidines 17a-c in hand, we attempted
to transform them into the homoallylic ethers 4 with allylsilane.
As an example 17a was reacted under various conditions; thus,
different amounts of TfOH and a mixture of TfOH and TMSOTf
with and without hexamethyldisiloxane were employed. In all
cases, 17a was recovered unchanged. Finally, 17a was added
to a reaction mixture of hexanal and 2a in the presence of
TMSOTf followed by addition of allylsilane. Again 17a was
recovered unchanged together with the homoallylic ether 4c
derived from hexanal. From these experiments, it can clearly
be deduced that oxazolidines 17 are not intermediates in the
described allylation of aldehydes. However, there is of course
the possibility that both the oxazolidines and the homoallylic
ethers are formed from the same intermediates. Therefore, the
behavior of the oxazolidine 17a in the presence of 1 equiv of
TfOH and TMSOTf was studied by on-line 13C NMR spec-
troscopy at -78 °C. In the presence of TMSOTf, the carbon
resonances did not change, but a significant shift of the
resonances was observed using TfOH. The signals of C-5 of
17a were shifted upfield by about 2 ppm from δ ) 85.68 to
83.59 (major amide isomer) and from δ ) 84.25 to 82.24 (minor
amide isomer) and those of C-2 by 0.44 ppm upfield from δ )
88.56 to 88.12 (major amide isomer) and downfield by about 1
ppm from δ ) 86.98 to 87.94 (minor amide isomer). In
contrast, the C-4 signals were only slightly changed from δ )
60.16 to 59.98 (major amide isomer) and from δ ) 61.51 to
61.99 (minor amide isomer). The 13C resonances observed for
the COCF3 group are nearly identical with those of the
oxazolidine 17a.
This clearly indicates that in the reaction of 17a with TfOH
neither 11 nor 12 was formed, and the protonation must have
taken place at the ring oxygen to give 18. A nucleophilic
substitution at C-2 of the protonated oxazolidine 18 with
allylsilane 3 to give 4 seems not to be possible under the reaction
conditions, presumably due to the low activity of the hydroxyl
moiety as a leaving group. Thus, the ring oxygen and not the
carbonyl oxygen or the nitrogen seems to be the position of the
highest basicity in 17. Indeed, the reduced nucleophilicity of
the carbonyl oxygen is an essential requirement for the allylation
of the aldehydes, since reaction of 1 and 2 having an N-acetyl
instead of the N-trifluoroacetyl moiety did not lead to the desired
product.4b
(6) Sames, D.; Liu, Y.; DeYoung, L.; Polt, R. J. Org. Chem. 1995, 60,
2153.
(7) Tsunoda, T.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980, 21, 1357.
(8) (a) Bernardi, A.; Cardani, S.; Pilati, T.; Poli, G.; Scolastico, C.; Villa,
R. J. Org. Chem. 1988, 53, 1600. (b) Hoppe, I.; Hoppe, D.; Wolff, C.;
Egert, E.; Herbst, R. Angew. Chem. 1989, 101, 65; Angew. Chem., Int. Ed.
Engl. 1989, 28, 67.
(9) Hart, J. D.; Ford, W. T. J. Org. Chem. 1974, 39, 363.
(10) (a) Pattaroni, C.; Lauterwein, J. HelV. Chim. Acta 1981, 64, 1969.
(b) Ogino, J.-I.; Suezawa, H.; Hirota, M. Chem. Lett. 1983, 889. (c)
Ilczyszyn, M.; Ratajczak, H.; Skowronek, K. Magn. Reson. Chem. 1988,
26, 445. (d) Andres, C.; Delgado, M.; Pedrosa, R. Synth. Commun. 1992,
22, 829. (e) Paz-Sandoval, M. A.; Santiesteban, F.; Contreras, R. Magn.
Reson. Chem. 1985, 23, 428.

Highly StereoselectiVe Allylation of Aldehydes
J. Am. Chem. Soc., Vol. 120, No. 18, 1998 4279
Table 2. Allylation of Aldehyde 1a with Different Catalysts and
at Different Temperatures
in the presence of TMSOTf which reacts with 2a to give the
mixed acetal 7 and 1 equiv of TMSOTf. In the next step, 7 is
transformed into the oxazolidinium salt 12 using 1 equiv of
TMSOTf. At temperatures above -80 °C, proton transfer in
12 occurs to give the isomeric oxazolidinium ion 18 as a single
diastereomer as detected by on-line NMR spectroscopy. The
ion 18 does not react with the allylsilane 3, but yields the
oxazolidine 17 after aqueous workup. At temperatures below
-80 °C, 12 is stable and reacts with the allylsilane 3 to give
the desired homoallylic ether 4 in an SN2-type fashion with
inversion of the configuration at C-2 of 12. In addition, 1 equiv
of TMSOTf is formed, so that TMSOTf is necessary only in
catalytic amounts. The lower selectivity of the allylation of
aromatic aldehydes 1 (R ) Ar) using this method may be due
to a stabilization of the open oxenium ion 10, which is in
equilibrium with 12 being controlled by the character of the
substituent R and which reacts with 3 in a nonstereoselective
fashion.12 Also of interest is the result that the diastereoselec-
tivity of the allylation decreases dramatically (52-78% dr) using
the ephedrine derivative 20.1 Thus, the highly stereoselective
formation of 12 is controlled by both stereogenic centers C-1
and C-2 in 2a. This is in clear contrast to the allylation of
ketones,4 where such a dependence was not observed. Due to
this fact and the opposite configuration of the newly formed
stereogenic center in the resulting homoallylic ether, the
intermediacy of an oxazolidinium ion of type 12 is very unlikely
in the allylation of ketones. We are aware that the proposed
mechanism is somehow provocative since it assumes that at
low temperature the SN2 displacement at 12 by allyl silane is
faster than the proton transfer. However, all obtained facts are
in clear accordance with this mechanism.
yield (%)
temp
catalyst
(°C)
time
2a
2b
4a
17a
TMSOTf
TfOH
TMSOTf
TfOH
TMSOTf
TMSB(OTf)4
TMSOTf
TMSB(OTf)4
TMSOTf
TMSB(OTf)4
-78
-78
-78
-78
-78
-78
-60
-60
-50
-50
30 sa
14
15
17
17
15
62
15
51
67
23
65
32
66
65
11
19
7
4
<1
3
<1
8
23
48
51
42
55
30 sa
5 mina
5 mina
12 hb
12 hb
12 hb
12 hb
12 hb
12 hb
26
12
41
30
51
24
21
a Time after the addition of allylsilane 3 (2.0 equiv); before addition
of 3, 1a (2.0 equiv), 2a (1.0 equiv), and the catalyst (0.1 equiv) were
kept for 1 h. b Total reaction time.
found. Due to the strong shielding of C-4, it is most reasonable
that the configuration of the double bond is E. Thus, H-bonding
to the triflate anion leads to an interaction between the triflate
and the C-2 methyl group, resulting in a shielding of C-4 with
the consequence of an upfield shift of the C-4 resonance. In
18, the situation is different because the phenyl and methyl group
at C-5 and C-2, respectively, are situated on the same side of
the molecule. Protonation can now take place from the R- or
â-side, influenced by a 1,3 or a 1,2 interaction. However, in
both cases a similar interaction between the triflate and the two
R-substituents would exist, resulting in a similar shielding of
C-2 and C-5, which is observed experimentally.
Though in the formation of 4 the intermediacy of the ion 12
seems to be most likely, it is also possible that the oxygen of
the carbonyl moiety of the trifluoroacetamide group in 10 acts
as a nucleophile to give a seven-membered ion 19. The ion 19
cannot be excluded on grounds of the NMR data. If it is
assumed that the formation of the seven-membered ring occurs
diastereoselectively, then the observed stereocontrol in the
allylation step could be explained. However, the preferred
formation of a seven-membered over a five-membered ring as
well as a diastereoselective cyclization to 19 is highly unlikely.
Furthermore, the diastereoselective formation of 17 from 19
cannot easily be explained.
Experimental Section
Instrumentation. Multiplicities of 13C NMR peaks were determined
with the APT pulse sequence. Mass spectra were measured at 70 eV.
UV-vis spectra [λmax, nm (log ꢀ)] were taken in CH3CN. IR spectra
were recorded as KBr pellets or as films. Melting points are corrected.
Materials. All solvents were dried and distilled prior to use. All
compounds were obtained enantio- and diastereomerically pure.
Transformations and NMR studies were performed in flame-dried
glassware under inert gas atmosphere.
General Procedure I. Preparation of Oxazolidines 17. To a
solution of 1 (2.00 mmol) and 2a (1.00 mmol) in CH2Cl2 (5 mL) was
added TMSOTf (1.10 mmol) dropwise with stirring at -78 °C, and
stirring was continued for 1-24 h at -78 °C. The reaction was
quenched by addition of aqueous triethylamine (2.00 mL) at -78 °C.
The solution was washed with water (10 mL), and the aqueous phase
was extracted with CH2Cl2 (3 × 3.0 mL). The combined organic phases
were dried (Na2SO4), the solvent was evaporated, and the crude product
was purified by column chromatography on silica gel (tert-butyl methyl
ether/petroleum ether, 1:10).
11
For the allylation of aldehydes, TMSOTf and TMSB(OTf)4
were used as catalysts. Both compounds can contain consider-
able amounts of TfOH. Therefore, the reactions were also
performed with pure TfOH. The results clearly show that TfOH
is much less suitable as a catalyst than TMSOTf. In addition,
the use of the very strong Lewis acid TMSB(OTf)4 gives no
advantage over TMSOTf which is commercially available and
therefore preferable (Table 2).
The time and temperature dependence of the transformation
revealed two important features. First, the homoallylic ethers
4 are formed very fast after the addition of the allylsilane 3.
Second, with increasing temperature, the yield of the homoallylic
ethers 4 decreases dramatically whereas the amount of the
oxazolidine 17 increases. This indicates that an irreversible
proton transfer in 12 occurs to give 18 at higher temperatures,
whereas at temperatures below -80 °C, this process must be
rather slow.
General Procedure II. On-line NMR Studies of the Reaction
Mixture. To a solution of 1 (2.0 equiv) and 2a (1.0 equiv) in CH2Cl2
(450 µL) in an NMR tube were added carefully TMSOTf (1.0 equiv)
and CD2Cl2 (150 µL) at -98 °C. Measurement of the spectra was
started immediately at a temperature of -78 °C.
(2S,4R,5R)-2,4-Dimethyl-5-phenyl-3-(trifluoroacetyl)oxazoli-
dine (17a). Reaction of 1a and 2a according to general procedure I
gave 273 mg (99%) of 17a as a white solid: mp 40 °C; Rf 0.53 (tert-
butyl methyl ether/petroleum ether, 1:3); [R]20D -11.0° (c 1.00, CHCl3);
1
UV 191 (4.222), 208 (3.779); IR (KBr) 3068, 3034, 1688; H NMR
The experimental results and the on-line NMR investigations
allow us to now present the following mechanism for the highly
stereoselective allylation of aldehydes 1 to give the homoallylic
ethers 4. In the first step, the oxenium ion 6 is formed from 1
(12) (a) Mori, I.; Ishihara, K.; Flippin, L. A.; Nozaki, K.; Yamamoto,
H.; Bartlett, P. A.; Heathcock, C. A. J. Org. Chem. 1990, 55, 6107. (b)
Denmark, S. E.; Willson, T. M. J. Am. Chem. Soc. 1989, 111, 3475. (c)
Denmark, S. E.; Willson, T. M.; Almstead, N. G. J. Am. Chem. Soc. 1989,
111, 9258. (d) Denmark, S. E.; Almstead, N. G. J. Am. Chem. Soc. 1991,
113, 8089.
(11) Davis, A. P.; Jaspars, M. Angew. Chem. 1992, 104, 475; Angew.
Chem., Int. Ed. Engl. 1992, 31, 470.

4280 J. Am. Chem. Soc., Vol. 120, No. 18, 1998
Tietze et al.
(200 MHz, CDCl3) δ 1.43 (d, J ) 6.0 Hz, 3 H), 1.66 (d, J ) 5.1 Hz,
3 H), 4.10 (br, 1 H), 4.58 (br d, J ) 6.4 Hz, 1 H), 5.55 (br, 1 H),
7.33-7.46 (m, 5 H); 13C NMR (125.7 MHz, CD2Cl2, -78 °C), major
amide isomer δ 16.7, 18.2, 60.2, 85.7, 88.6, 115.2 (q, 1JCF ) 287 Hz),
127.0, 128.4, 128.8, 135.5, 154.7 (q, 2JCF ) 38 Hz), minor amide isomer
(br, 1 H), 7.30-7.50 (m, 5 H); 13C NMR (125.7 MHz, CD2Cl2, -78
°C), major amide isomer δ 13.9, 16.5, 22.2, 22.5, 31.3, 31.1, 60.5,
85.6, 91.2, 115.2 (q, 1JCF ) 287 Hz), 127.1, 128.4, 128.9, 135.4, 154.6
2
(q, JCF ) 38 Hz), minor amide isomer δ 13.9, 13.3, 21.8, 22.5, 30.0,
1
33.8, 61.9, 84.3, 89.5, 115.3 (q, JCF ) 288 Hz), 127.0, 128.3, 128.8,
1
2
135.3, 154.6 (q, JCF ) 38 Hz); 13C NMR (50.3 MHz, C2D2Cl4, 100
δ 13.39, 21.16, 61.51, 84.25, 86.98, 115.3 (q, JCF ) 288 Hz), 126.9,
2
128.4, 128.8, 135.3, 154.4 (q, JCF ) 38 Hz); 13C NMR (50.3 MHz,
°C) δ 13.74, 17.21, 22.35, 23.28, 31.39, 33.36, 60.54, 86.39, 91.53,
C2D2Cl4, 100 °C) δ 17.41, 20.78, 60.32, 86.28, 88.34, 116.1 (q, 1JCF
)
116.1 (q, 1JCF ) 288 Hz), 126.8, 128.8, 129.0, 137.3, 154.6 (q, 2JCF
)
288 Hz), 126.8, 128.8, 129.0, 137.4, 154.3 (q, 2JCF ) 38 Hz); 19F NMR
(470.3 MHz, CH2Cl2/CD2Cl2 (3:1), -80 °C) δ -70.81 (major amide
isomer), -69.82 (minor amide isomer); MS m/z 273 (2, M+), 230 (6,
M+ - C(CH3)O), 167 (100, M+ - PhCHO). Anal. Calcd for C13H14F3-
NO2 (273.3): C, 57.14; H, 5.16. Found: C, 57.23; H, 4.97.
38 Hz); MS m/z 329 (0.2, M+), 258 (100, M+ - C5H11), 230 (25, M+
- C6H13O); EI HRMS m/e 329.1602, C17H22F3NO2 requires 329.1603.
On-line 13C NMR Study of the Protonation of 17a. To a solution
of 17a (50.0 mg, 180 µmol, 1.0 equiv) in CD2Cl2 (550 µL) in an NMR
tube was added TfOH (17.0 µL, 180 µmol, 1.0 equiv) carefully at -78
°C, and a 13C NMR spectrum of the mixture was immediately taken at
-78 °C: 13C NMR (125.7 MHz, CD2Cl2, -78 °C), major amide isomer
δ 18.9, 19.7, 60.0, 83.6, 88.1, 114.9 (q, 1JCF ) 285.9 Hz), 124.7, 126.5,
128.6, 135.8, 155.4 (q, 2JCF ) 40.3 Hz), minor amide isomer (different
signals) δ 15.5, 19.5, 62.0, 82.2, 87.9, 115.1 (q, 1JCF ) 286 Hz), 155.3
(2S,4R,5R)-2-Ethyl-4-methyl-5-phenyl-3-(trifluoroacetyl)oxa-
zolidine (17b). Reaction of 1b and 2a according to general procedure
I gave 253 mg (88%) of 17b as a colorless oil: Rf 0.54 (tert-butyl
methyl ether/petroleum ether, 1:5); [R]20D -32.7° (c 0.75, CHCl3); UV
191 (4.179), 209 (3.738); IR (film) 3090, 3068, 3036, 1686; 1H NMR
(200 MHz, CDCl3) δ 1.00 (br t, J ) 8.2 Hz, 3 H), 1.43 (d, J ) 7.5 Hz,
3 H), 2.05 (m, 2 H), 3.97 (m, 1 H), 4.53 (d, J ) 8.0 Hz, 1 H), 5.49 (m,
1 H), 7.34-7.48 (m, 5 H); 13C NMR (50 MHz, CDCl3), major amide
2
(q, JCF ) 41 Hz).
On-line 13C and 19F NMR Studies of the Reaction of 1a with 2a.
For 13C NMR studies, 1a (18.0 µL, 310 µmol, 2.0 equiv), 2a (50.0
mg, 160 µmol, 1.0 equiv), and TMSOTf (31.0 µL, 160 µmol, 1.0 equiv)
were used. For 19F NMR studies 1a (5.3 µL, 76 µmol, 2.0 equiv), 2a
(12.0 mg, 38.0 µmol, 1.0 equiv), and TMSOTf (7.4 µL, 38 µmol, 1.0
equiv) were used: 13C NMR (125.7 MHz, CH2Cl2/CD2Cl2, 3:1, -78
1
isomer δ 7.1, 18.0, 24.0, 60.6, 86.5, 92.3, 115.8 (q, JCF ) 288 Hz),
127.0, 128.7, 129.1, 136.5, 155.5 (q, 2JCF ) 38 Hz), minor amide isomer
(different signals) δ 14.3, 28.2, 61.86, 85.53, 90.76; MS m/z 287 (0.4,
M+), 258 (100, M+ - CH3CH2), 230 (51, M+ - CH3CH2CO), 181
(57, M+ - PhCHO); EI HRMS m/e 287.1133, C14H16F3NO2 requires
287.1133.
1
°C) δ 14.6, 19.5, 50.9, 78.0, 94.9, 115.0 (q, JCF ) 287.5 Hz), 126.6,
128.1, 128.6, 136.5, 157.3; 19F NMR (470.3 MHz, CH2Cl2/CD2Cl2, 3:1,
(2S,4R,5R)-4-Methyl-2-pentyl-5-phenyl-3-(trifluoroacetyl)oxa-
zolidine (17c). Reaction of 1c and 2a according to general procedure
I gave 137 mg (46%) of 17c as a colorless oil: Rf 0.58 (tert-butyl
methyl ether/petroleum ether, 1:3); [R]20D -39.7° (c 0.75, CHCl3); UV
209 (3.661); IR (film) 3068, 3036, 1688; 1H NMR (200 MHz, CDCl3)
δ 0.90 (br, 3 H), 1.20-1.60 (m, 6 H), 1.42 (d, J ) 6.0 Hz, 3 H), 1.86
(br, 1 H), 2.12 (br, 1 H), 4.00 (br, 1 H), 4.53 (d, J ) 7.8 Hz, 1 H), 5.50
-78 °C) δ -75.31, -75.14.
Acknowledgment. We thank the Volkswagen Stiftung and
the Fonds der Chemischen Industrie for generous support.
JA974390+