Effect of allylic CH3-nFn groups (n = 1-3) on π-facial diastereoselection

Article abstract of DOI:10.1021/ol016401g

(equation presented) Michael addition of various enolates toward γ-CH_3-nF_n-α,β-unsaturated ketones (n = 1-3) was proven to smoothly furnish the 1,4-adducts with high si face selectivities which monotonously decreased by reduction in the number of fluorines. Although the Felkin-Anh model correctly anticipates the present stereochemical outcome only with E-acceptors, the hyperconjugative stabilization of transition states by electron donation from the allylic substituents (the Cieplak rule) successfully explains the π-facial preference of both acceptors at least in a qualitative level.

Full text of DOI:10.1021/ol016401g

ORGANIC
LETTERS
2001
Vol. 3, No. 18
2915-2918
Effect of Allylic CH F Groups
(n ) 1−3) on π-Facial
Diastereoselection
3-n n
,†
Takashi Yamazaki,* Tatsuro Ichige,† Satoshi Takei,† Seiji Kawashita,†
Tomoya Kitazume,† and Toshio Kubota‡
Department of Bioengineering, Tokyo Institute of Technology, 4259 Nagatsuta-cho,
Midori-ku, Yokohama 226-8501, Japan, and Department of Materials Science,
Ibaraki UniVersity, Nakanarusawa 4-12-1, Hitachi, Ibaraki 316-8511, Japan
tyamazak@bio.titech.ac.jp
Received July 5, 2001
ABSTRACT
Michael addition of various enolates toward γ-CH3-nFn-r,â-unsaturated ketones (n ) 1−3) was proven to smoothly furnish the 1,4-adducts
with high si face selectivities which monotonously decreased by reduction in the number of fluorines. Although the Felkin−Anh model correctly
anticipates the present stereochemical outcome only with E-acceptors, the hyperconjugative stabilization of transition states by electron
donation from the allylic substituents (the Cieplak rule) successfully explains the π-facial preference of both acceptors at least in a qualitative
level.
Stereoselective construction of organic molecules has been
one of the central topics in synthetic organic chemistry, and
a large number of challenges have been encountered in this
field, especially in the diastereotopic π-facial discrimination
at sp2 hybridized carbon atoms.1 One of the most significant
factors for successfully obtaining a satisfactory outcome is,
of course, to fix substrate conformations in an effective
manner so as to attain maximum discrimination between both
π faces by the steric and/or electrostatic effects of neighbor-
ing allylic substituents.
During the course of study on the development of novel
preparation methods to access fluorine-containing materials
in a stereoselective manner,2 we have recently revealed
highly diastereoselective Ireland-Claisen rearrangements
Scheme 1a
a (a) LDA/THF; (b) TMSCl; (c) 2.5 mol % PdCl2‚(PhCN)2; (d)
reflux.
(Scheme 1).3 When a THF solution of ketene silyl acetal 24
was refluxed for 6 h in the presence of a Pd catalyst,5 the
carboxylic acid 3 was produced with 88% syn selectivity.
Taking into account the fact that the steric size of a CF3
† Tokyo Institute of Technology.
‡ Ibaraki University.
(1) Gung, B. W.; le Noble, B. Guest editors for the special thematic
issue on diastereoselection. Chem. ReV. 1999, 99, 1067.
(2) (a) Yamazaki, T.; Umetani, H.; Kitazume, T. Isr. J. Chem. 1999, 39,
193. (b) Yamazaki, T.; Shinohara, N.; Kitazume, T.; Sato, S. J. Fluorine
Chem. 1999, 97, 91. (c) Yamazaki, T.; Hiraoka, S.; Kitazume, T.
Tetrahedron: Asymmetry 1997, 8, 1157.
(3) Yamazaki, T.; Shinohara, N.; Kitazume, T.; Sato, S. J. Org. Chem.
1995, 60, 8140.
(4) The independent reaction demonstrated that the E,Z ratio of 2 was
92:8 in favor of the E isomer. Details will be published elsewhere.
10.1021/ol016401g CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/16/2001

group is regarded to be between i-Pr and i-Bu moieties,6 this
relatively high syn preference should stem from the electro-
static environmental difference between two π-faces. This
intriguing example encouraged us to investigate other
systems for confirming its generality, and we at first selected
Michael addition reactions to substrates with fluorine-
containing methyl groups at their allylic positions. In this
communication were described the experimental results of
enolate-Michael addition to (E)-γ-CH3-nFn-R,â-unsaturated
ketones 6 (n ) 1-3) as the representative substrates, which
as expected showed a clear correlation between the number
of fluorine atoms and diastereofacial selectivities, mono-
tonously decreasing the latter by successive reduction of the
former.7
Emmons (HWE) reaction. For the synthesis of the corre-
sponding F3 molecule, direct reaction of commercially
available 3,3,3-trifluoro-2-methylpropionaldehyde10 with the
same phosphonate produced E-6c in 64% yield. On the other
hand, construction of the difluorinated counterpart E-6b was
found not to be straightforward. Our first synthetic plan was
to start with the SN2′-type substitution of fluorine in 9 (as
the sodium salt) for the hydride11 and then hydrogenation
after derivatization into the corresponding ester. However,
because of the detection of an appreciable amount of
impurities, we have devised a new route involving depro-
tonation of the readily available 10c from 9 as the key step.12
Close examination of the reaction conditions proved that slow
addition of LDA in THF to 10c dissolved in the same solvent
at -78 °C was quite effective in suppressing the possible
formation of undesirable byproducts, and the subsequent
hydrogenolysis with 10% Pd/C under 0.5 MPa pressure of
hydrogen allowed us to isolate the desired 10b in 60% total
yield from 10c. Its transformation into E-6b was attained
successfully by the similar HWE protocol. In addition to the
three types of E-6 thus obtained, the corresponding Z-6c was
also prepared for comparison. Several trials indicated that
(diphenyl-phosphono)acetate13 was the reagent of choice, and
the intermediary R,â-unsaturated carboxylic acid 8 was
isolated in 86% overall yield as the sole Z-isomer.14 Z-6c
was finally obtained by tert-BuMgCl reaction with the
corresponding mixed anhydride in 24% yield.
Preparation of variously fluorinated Michael acceptors 6
was performed as depicted in Scheme 2. Thus, after
Scheme 2a
With four requisite acceptors E-6a-c and Z-6c in hand,15
enolate-Michael addition reactions were carried out with three
representative lithium enolates from propiophenone, ethyl
(methylthio)acetate,16 and N,N-dimethyl-propionamide whose
results are summarized in Table 1. As a general trend,
chemical yields as well as diastereomeric ratios of products
have a qualitatively proportional relationship with the number
of fluorine atoms in the acceptors E-6. Theoretical molecular
orbital calculation17 of E-6c suggested its LUMO level of
-2.055 eV at the B3LYP/6-31+G* level of theory, about
0.43 eV lower than the corresponding nonfluorinated coun-
terpart, would be at least in part responsible for the higher
reactivity of E-6c compared to the others.
a (a) Et2NCF2CHFCF3/CH2Cl2; (b) DIBAL/Et2O; (c) (EtO)2P(O)-
CH2-C(O)t-Bu, n-BuLi/Et2O; (d) (PhO)2P(O)CH2CO2Et, NaH/
THF; (e) NaOH/THF-H2O; (f) AcCl, Et3N/CH2Cl2; (g) t-BuMgCl/
THF; (h) C6H4(COCl)2; (i) PhCH2CH2OH, pyridine/CH2Cl2; (j) 10%
Pd/C/ MeOH; (k) LDA/THF.
Although all possible four diastereomers were formed
when E-6 was treated with the amide enolate, only two
isomers were observed for the adducts 11 and 12 from the
ketone and ester enolates, respectively. The isomeric acceptor
Z-6c was apparently less reactive, possibly as a result of its
fluorination of the hydroxy group in 4,8 partial reduction
followed by in situ condensation of the resultant hemiacetal
intermediate9 furnished the monofluorinated acceptor E-6a
in 30% total yield by way of the Ho¨rner-Wadsworth-
(9) Lanier, M.; Haddach, M.; Pastor, R.; Riess, J. G. Tetrahedron Lett.
1993, 34, 2469 and references therein.
(10) Distillation with a few drops of BF3‚OEt2 was required just prior
to use for reproducible results.
(11) Fuchikami, T.; Shibata, Y.; Suzuki, Y. Tetrahedron Lett. 1986, 27,
3173.
(5) Overman, L. E.; Renaldo, A. F. J. Am. Chem. Soc. 1990, 112, 3945.
(6) (a) MacPhee, J. A.; Panaye, A. Dubois, J.-E. Tetrahedron 1978, 34,
3553. (b) Kitazume, T.; Yamazaki, T. Experimental Methods in Organic
Fluorine Chemistry; Kodansha, Gordon, and Breach Science Publisher:
Tokyo, 1998; Chapter 1. See also: Be´guin, C. G.; Schlosser, M. In
Enantiocontrolled Synthesis of Fluoro-Organic Compounds; Soloshonok,
V., Ed.; John Wiley and Sons: New York, 1999; pp 601 and 613,
respectively.
(7) For the discussion of the number of fluorines and stereoselectivity,
see: Soloshonok, V. A.; Kacharov, A. D.; Avilov, D. V.; Ishikawa, K.;
Nagashima, N.; Hayashi, T. J. Org. Chem. 1997, 62, 3470.
(8) O’Hagan, D. J. Fluorine Chem. 1989, 43, 371.
(12) Recently, preparation of 3,3-difluoroacrylate was reported; see:
Botteghi, C.; Paganelli, S.; Sbrogio`, F.; Zarantonello, C. Tetrahedron Lett.
1999, 40, 8435 and references therein.
(13) Ando, K. J. Org. Chem. 1999, 64, 8406.
(14) Initial E:Z ratio was 3:97 at the HWE step (determined by 19F NMR)
while the minor isomer was removed after hydrolysis.
(15) Throughout the text, the configuration of the CH3-nFn-attached
carbon in 6 is conveniently fixed as S for the simpler stereochemical
discussion despite employment of their racemic forms except for the case
of E-6a.
(16) This donor was selected because no 1,4- or 1,2-adducts were
provided when ethyl propionate was used.
2916
Org. Lett., Vol. 3, No. 18, 2001

Scheme 3a
Table 1. Enolate-Michael Addition Reactions toward 6a
n
R1
Ph
R2
product yield (%) diastereomeric ratiob
3
2
1
3c
Me
11c
11b
11a
95.0
80.5
76.3
0d
2.8:97.2
7.6:92.4
17.6:82.4
a (a) n-Bu3SnH, cat. AIBN/PhH; (b) LDA/THF; (c) HMPA, MeI;
(d) LiOH/THF-H2O; (e) DAST; (f) PhMgBr/THF; (g) t-BuC(O)Cl,
pyridine/ CH2Cl2; (h) 50% HNMe2 aq.
3
2
1
3c
OEt
SMe
12c
12b
12a
12c
quant
86.0
76.3
45.9
7.7:92.3 [1 isomer]
15.1:84.9 [16.7:83.3]
30.9:69.1 [26.3:73.7]
<1.0:>99.0
3
2
1
3c
NMe2 Me
13c
13b
13a
14ce
87.5
74.4
78.8
87.6
3.6: 5.1:30.4:60.9
5.0:13.8:30.9:50.3
15.6:16.3:34.0:34.1
2.2:16.6:28.4:52.8
At the next stage, methylation of the ketoester 15c obtained
above was carried out, and 16c was produced as a 74:26
diastereomer mixture, which without further purification was
independently converted into 17c and 18c with almost
complete retention of the original stereochemical integrity
(Scheme 3). Spectroscopic comparison of 18c with the
Michael adduct 13c demonstrated that the major and minor
18c corresponded to the major and the second major isomers
of 13c, respectively. On the other hand, ready chromato-
graphic separation of the major 13c enabled us to determine
its relative stereochemical relationship as (2R*,3R*,5S*) by
the crystallographic technique.18 These data led to the
conclusion that the major and the second major isomers of
13c should possess (2R*,3R*,5S*) and (2S*,3R*,5S*) ste-
reochemistries, respectively, and the minor two isomers
should be (3S*,5S*). On the other hand, while the major 17c
was proven to be identical to the major 11c, the correspond-
ing minor isomers were different from each other, suggesting
that two isomers of 11c were yielded as the result of the
opposite diastereofacial selection.
The unambiguous stereochemical assignment for the CF3
adducts 11c-13c and analogous consideration for the others
demonstrated a good to excellent level of si-face selection
in every instances. These results are summarized in Table 2
along with some parameters of interest in regard to fluorine-
possessing methyl groups.
For the reactions with E-6 where R1 ) t-BuC(O) and
R2 ) H, the Felkin-Anh (FA)-type transition state models
are usually employed for the explanation of the product
stereochemistry. As shown in Figure 1, because of the
undesired steric repulsive interaction between incoming
nucleophiles and the allylic substituents, TS-FA-si is re-
garded to be more preferable than TS-FA-re. Moreover, an
increase in the number of fluorine atoms increases the steric
size of CH3-nFn groups as well as their electron-withdrawing
ability (see Table 2), and thus TS-FA-si preference becomes
a Two equivalents of an enolate, prepared from a carbonyl compound
and LDA in THF at -78 °C, was reacted with an acceptor at -78 °C for
2 h. b In the bracket is shown the diastereomeric ratio of 15 observed after
removal of the methylthio group from 12. c Z-acceptor was employed. d No
reaction occurred. e The corresponding 1,2-adduct was obtained as sole
product in the diastereomeric ratio depicted above.
severe steric congestion, and complete recovery of the
starting material was observed for the reaction with the
propiophenone enolate. On the other hand, high regioselec-
tivity was noticed for the amide enolate, leading to the
formation of the undesired 1,2-adduct in high yield. The ester
enolate, different from other two enolates, realized the
conjugate addition to afford 12c basically as a single
diastereoisomer, which was proven to be identical to the
major isomer prepared from E-6c by spectroscopic com-
parison.
Stereochemical clarification was carried out as shown in
Scheme 3. At first, n-Bu3SnH-mediated removal of a MeS
group was performed for obtaining direct information on the
π-facial selectivity. This procedure for 12c yielded 15c as a
single stereoisomer, while the corresponding mono- and
difluorinated adducts, 12a and 12b, furnished 15a and 15b
as diastereomeric pairs with the same isomeric ratios,
respectively. Thus, 12c contained two epimeric stereoisomers
at C2, but two isomers of 12a and 12b were formed as the
result of the different π-facial selection.
(17) Gaussian 98, Revision A.7; Frisch, M. J.; Trucks, G. W.; Schlegel,
H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A., Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam,
J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.; Replogle,
E. S.; Pople, J. A. Gaussian, Inc.: Pittsburgh, PA, 1998.
(18) Crystallographic data for the major diastereomer of 13c has been
deposited at the Cambridge Crystallographic Data Center and allocated the
deposition number CCDC 166764.
Org. Lett., Vol. 3, No. 18, 2001
2917

to play a central role in stabilizing transition states by electron
donation from the allylic substituents to the incipient
antibonding orbital.21,22 In connection to the fact that Michael
additions have been reported to be generally exothermic23
and their TSs should be early and reactant-like,24 the two
major models to be considered would be TS-C-si and -re,
both stemming from the energetically most favorable con-
formation with the smallest hydrogen atom located at the
inside position. Then, TS-C-si, where the electron-donating
methyl group occupies the antiperiplanar position with
respect to the incoming nucleophiles, would be more
preferable than TS-C-re, almost identical in appearance to
TS-FA-re, with an electron-withdrawing fluorine-containing
methyl group anti to the nucleophiles. Moreover, increasing
the number of fluorines will increase the gap in electron-
donating ability between Me and CH3-nFn groups, which also
reasonably dictates the observed diastereoselectivity varia-
tion. The 3JH-H values shown in Table 2, coupling constants
between vinylic H â to the carbonyl group and the allylic
H, demonstrated a quite interesting trend, and constant
addition of fluorine affects the monotonic increase of the
3JH-H values. This is understood as the result of the larger
population of such conformations with inside hydrogen as
TS-FA-re, TS-C-re, and TS-C-si, and considering that the
transition states really resemble the substrates, it is the
Cieplak rule, at least in the present system, which consistently
explains the observed π-face selective nucleophilic attack
of enolates from the face where bulkier F-containing methyl
groups are located.25
Table 2. Relationship between Steric Bulkiness of CH3-nFn
Groups vs Diastereofacial Selectivities
si face selectivity (%)d
b
n
steric factora pKa
3J H-H (Hz)c ketone ester amide
3
2
1
0
-1.90
-1.47
-1.32
-1.12
0.00
0.23
1.24
2.59
4.76
3.75
8.5e
7.4
7.0
6.9f
6.5f
97.2 100
91.3
>92.4
>82.4
84.9 <81.2
69.1 <68.1
Hg
a The larger negative value expresses its bulkier steric requirement. See
ref 6a. Representative values for the nonfluorinated alkyl groups are as
follows.: Et, -1.20; n-Pr, -1.43; i-Pr, -1.60; i-Bu, -2.05; sec-Bu, -2.12.
b The values for CH3-nFnCO2H as an index for estimation of the electron-
withdrawing ability of the fluorine-containing alkyl groups. c Coupling
constants between allylic H and carbonyl â-H. d Inequality signs were used
since the relative stereochemistries were not determined yet. e A value of
9.7 Hz was observed for the corresponding Z-6c. f See ref 19. g H instead
of a CH3-nFn group.
more pronounced to attain the better diastereoselectivity. This
is totally in accord with the experimental results discussed
Supporting Information Available: Experimental pro-
cedure for the preparation of acceptors E-6a-c and Z-6c, as
well as their reactions with enolates to yield the correspond-
ing adducts 11-13. This material is available free of charge
via the Internet at http://pubs.acs.org.
Figure 1. The Felkin-Anh and Cieplak models.
above. However, this is not the case for the corresponding
Z-6c: significant steric disturbance between R2 (t-BuC(O))
and the allylic Me moiety allows TS-FA-re to be the
alternative route, resulting in the erroneous prediction of the
diastereoisomers actually obtained if nucleophiles are as-
sumed to approach from the less hindered side.20
OL016401G
(21) Cieplak, A. S. Chem. ReV. 1999, 99, 1265.
(22) Cieplak, A. S. J. Am. Chem. Soc. 1981, 103, 4540.
(23) (a) Wong, S. S.; Paddon-Row: M. N.; Li, Y.; Houk, K. N. J. Am.
Chem. Soc. 1990, 112, 8679. (b) Tatsukawa, A.; Kawatake, K.; Kanemasa,
S.; Rudzinski, J. M. J. Chem. Soc., Perkin Trans. 1 1994, 2525. See also:
Bernardi, A.; Capelli, A. M.; Cassinari, A.; Comotti, A.; Gennari, C.;
Scolastico, C. J. Org. Chem. 1992, 57, 7029.
In constrast, instead of the the FA model, the Cieplak (C)
rule is also applicable where hyperconjugation is considered
(19) Oare, D. A.; Henderson, M. A.; Sanner, M. A.; Heathcock, C. H. J.
Org. Chem. 1990, 55, 132.
(20) In the case of nonfluorinated systems, the Felkin-Anh model for
the E-acceptors and the 1,3-allylic strain concept for the Z-acceptors have
been successfully employed for the explanation of the product stereochem-
istry; see, Mengel, A.; Reiser, O. Chem. ReV. 1999, 99, 1191.
(24) Seeman, J. I. Chem. ReV. 1983, 83, 83.
(25) One reviewer suggested that Li‚‚‚F chelation is also applicable for
the explanation of the present results, but this was ruled out here because
of our previous experience on diastereoselectivity independent from the
number of fluorines; see: Yamazaki, T.; Ando, M.; Kitazume, T.; Kubota,
T.; Omura, M. Org. Lett. 1999, 1, 905.
2918
Org. Lett., Vol. 3, No. 18, 2001