Double Diastereoselection in Aldol Reactions Mediated by Dicyclohexylchloroborane between L-Erythrulose Derivatives and Chiral Aldehydes. The Felkin-Anh versus Cornforth Dichotomy

Article abstract of DOI:10.1021/jo035159j

Both matched and mismatched diastereoselections have been observed in aldol reactions of the B,B-dicyclohexylboron enolate of a protected L-erythrulose derivative with a range of chiral aldehydes. The stereochemical outcome of reactions with α-methyl alde

Full text of DOI:10.1021/jo035159j

Dou ble Dia ster eoselection in Ald ol Rea ction s Med ia ted by
Dicycloh exylch lor obor a n e betw een L-Er yth r u lose Der iva tives a n d
Ch ir a l Ald eh yd es. Th e F elk in -An h ver su s Cor n for th Dich otom y
J . Alberto Marco,*,† Miguel Carda,‡ Santiago D´ıaz-Oltra,‡ J uan Murga,‡ Eva Falomir,‡ and
Harald Roeper§
Departamento de Qu´ımica Orga´nica, Universidad de Valencia, E-46100 Burjassot, Valencia, Spain,
Departamento de Qu´ımica Inorga´nica y Orga´nica, Universidad J aume I, E-12080 Castello´n, Spain, and
Cargill TDC Food Europe, Cerestar Vilvoorde R&D Centre, Havenstraat 84, B-1800 Vilvoorde, Belgium
alberto.marco@uv.es
Received August 6, 2003
Both matched and mismatched diastereoselections have been observed in aldol reactions of the
B,B-dicyclohexylboron enolate of a protected L-erythrulose derivative with a range of chiral
aldehydes. The stereochemical outcome of reactions with R-methyl aldehydes can be adequately
explained within the Felkin-Anh paradigm. In the case of R-oxygenated aldehydes, however, strict
adherence to this model does not allow for a satisfactory account of the observed results. In such
cases, the Cornforth model provides a much better explanation.
In tr od u ction
class of ketone substrates, the latter reagent has been
found to promote the formation of Z enolates, in contrast
with its previously documented behavior.2,4 We have thus
been able to isolate syn aldols 2 with high stereoselec-
tivity (Scheme 1, diastereoisomeric ratio, dr > 95:5) in
reactions with achiral aldehydes RCHO.5
The aldol reaction is a powerful and general method
for the stereocontrolled construction of carbon-carbon
bonds.1 Among the many enolate types investigated thus
far, boron enolates have proven to be particularly ver-
satile because of their good reactivity and excellent
stereoselectivity.2 In recent years, we have investigated
the outcome of aldol reactions of boron enolates generated
from suitably protected L-erythrulose derivatives such as
1 and dicyclohexylboron chloride, Chx2BCl.3 With this
In these aldol reactions, ketone 1 displays a marked
enantiofacial preference whereby the Re face of its
enolate attacks only the Re face of the aldehyde carbonyl.
This finding raises the question of whether this facial
bias is strong enough to overcome the inherent facial
preferences of the carbonyl group in R-chiral aldehydes
(double diastereoselection).1a-e This would be of great
importance, as aldol reactions of this type should allow
the synthesis of highly functionalized carbon fragments
such as those present in macrolides, polyether antibiotics,
and other naturally occurring, biologically relevant mol-
ecules.6 With this in mind, then, we prepared several
R-chiral aldehydes in both antipodal forms and investi-
gated their reactions with the Z boron enolate of ketone
1. The results of these reactions and the discussion
thereof are the subject of the present paper.
* To whom correspondence should be addressed. Phone: 34-96-
3544337. Fax: 34-96-3544328.
† Universidad de Valencia.
‡ Universidad J aume I.
§ Cargill TDC Food Europe.
(1) (a) Evans, D. A.; Nelson, J . V.; Taber, T. R. Top. Stereochem.
1982, 13, 1-115. (b) Mukaiyama, T. Org. React. 1982, 28, 203-331.
(c) Masamune, S.; Choy, W.; Petersen, J . S.; Sita, L. R. Angew. Chem.,
Int. Ed. Engl. 1985, 24, 1-30. (d) Heathcock, C. H. In Asymmetric
Synthesis; Morrison, J . D., Ed.; Academic Press: Orlando, FL, 1984;
Vol. 3, pp 111-212. (e) Heathcock, C. H. Aldrichim. Acta 1990, 23,
99-111. (f) Comprehensive Organic Synthesis; Trost, B. M., Fleming,
I., Winterfeldt, E., Eds.; Pergamon Press: Oxford, UK, 1993; Vol. 2.
(g) Mekelburger, H. B.; Wilcox, C. S. In Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Winterfeldt, E., Eds.; Pergamon
Press: Oxford, UK, 1993; Vol. 2, pp 99-131. (h) Heathcock, C. H. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Winter-
feldt, E., Eds.; Pergamon Press: Oxford, UK, 1993; Vol. 2, pp 133-
179 and 181-238. (i) Kim, B. M.; Williams, S. F.; Masamune, S. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Winter-
feldt, E., Eds.; Pergamon Press: Oxford, UK, 1993; Vol. 2, pp 239-
275. (j) Rathke, M. W.; Weipert, P. In Comprehensive Organic
Synthesis; Trost, B. M., Fleming, I., Winterfeldt, E., Eds.; Pergamon
Press: Oxford, UK, 1993; Vol. 2, pp 277-299. (k) Paterson, I. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Winter-
feldt, E., Eds.; Pergamon Press: Oxford, UK, 1993; Vol. 2, pp 301-
319. (l) Franklin, A. S.; Paterson, I. Contemp. Org. Synth. 1994, 1, 317-
338. (m) Braun, M. In Houben-Weyl’s Methods of Organic Chemistry,
Stereoselective Synthesis; Helmchen, G., Hoffmann, R. W., Mulzer, J .,
Schaumann E., Eds.; Georg Thieme Verlag: Stuttgart, Germany, 1996;
Vol. 3, pp 1603-1666 and 1713-1735. (n) Mahrwald, R. Chem. Rev.
1999, 99, 1095-1120.
(3) (a) Carda, M.; Murga, J .; Falomir, E.; Gonza´lez, F.; Marco, J . A.
Tetrahedron 2000, 56, 677-683. (b) Murga, J .; Falomir, E.; Carda, M.;
Marco, J . A. Tetrahedron: Asymmetry 2002, 13, 2317-2327. (c) Murga,
J .; Falomir, E.; Gonza´lez, F.; Carda, M.; Marco, J . A. Tetrahedron 2002,
58, 9697-9707.
(4) This stereochemical outcome of aldol reactions mediated by
dicyclohexylboron chloride may be general in ketones bearing R-alkoxy
or R-silyloxy substituents: Murga, J .; Falomir, E.; Carda, M.; Gonza´lez,
F.; Marco, J . A. Org. Lett. 2001, 3, 901-904.
(5) Other ketones structurally related to erythrulose behave in the
same way: (a) Carda, M.; Murga, J .; Falomir, E.; Gonza´lez, F.; Marco,
J . A. Tetrahedron: Asymmetry 2000, 11, 3211-3220. (b) Carda, M.;
Gonza´lez, F.; Sa´nchez, R.; Marco, J . A. Tetrahedron: Asymmetry 2002,
13, 1005-1010.
(6) (a) Recent Progress in the Chemical Synthesis of Antibiotics;
Lukacs, G., Ohno. M., Eds.; Springer, Berlin, Germany, 1990. (b)
Norcross, R. D.; Paterson, I. Chem. Rev. 1995, 95, 2041-2114.
(2) Cowden, C. J .; Paterson, I. Org. React. 1997, 51, 1-200.
10.1021/jo035159j CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/07/2003
J . Org. Chem. 2003, 68, 8577-8582
8577

Marco et al.
SCHEME 1. Ald ol Ad d ition s of th e Bor on En ola te
of 1 to Ach ir a l Ald eh yd es
SCHEME 3. Ald ol Ad d ition s of th e Bor on En ola te
of 1 to Ald eh yd es (R)-3 a n d (S)-3
SCHEME 2. Ch ir a l Ald eh yd es Used in Th is Stu d y
TABLE 1. Ster eoch em ica l Ou tcom e of Ald ol Ad d ition s
of Keton e 1 w ith Ald eh yd es (R)- a n d (S)-3
entry
aldehyde
% yield
dr
1
2
3
4
(S)-3a
(S)-3b
(R)-3a
(R)-3b
72
80
88
60
>95:5a
>95:5a
>95:5b
∼80:20c
a
b
The only diastereoisomer detected was 6a /6b. The only
diastereoisomer detected was 7a . c The major diastereoisomer was
7b.
Resu lts a n d Discu ssion
completely diastereoselective insofar as can be detected
with NMR spectroscopic methods (dr > 95:5). Aldols 6
were thus formed via enolate attack to the Re aldehyde
carbonyl face. For aldehydes (R)-3, the reactions were
also completely diastereoselective when the protecting
group P was silyl; the only stereoisomer detected was 7a ,
again resulting from enolate attack to the Re aldehyde
face. For P ) benzyl, the minor aldol 8b was also formed
(dr ∼ 80:20). It may thus be concluded that the facial
bias of this ketone enolate (attack to aldehyde Re faces)
is strong enough to overcome the inherent facial prefer-
ence of the carbonyl group in R-methyl aldehydes 3, a
fact that enhances the synthetic value of this methodol-
ogy for the preparation of polypropionate fragments.6
These results can be understood within the same
mechanistic framework presented in one of our recent
publications.3c For achiral aldehydes we have proposed
a cyclic six-membered transition state (TS) of the Zim-
merman-Traxler type12,13 where the Z boron enolate of
1 selectively attacks the aldehyde Re face to yield syn
aldols 2 (Scheme 4), supporting this proposal with
theoretical calculations. This TS not only explains the
observed syn 1,2-induction (simple diastereoselection),
related to the participation of a Z enolate, but also the
syn 1,3-relationship (induced diastereoselection) with the
preexisting stereogenic center.1m This 1,3-induction is due
For the purposes described in the preceding section,
five enantiomerically pure aldehydes (3-4, Scheme 2)
were prepared in both antipodal forms according to
standard procedures.7 The derivatives of 3-hydroxy-2-
methylpropanal 3a ,b and lactaldehyde 4a ,b displayed
two different protecting groups (tert-butyldiphenylsilyl,
TPS; benzyl, Bn).8a
The aldol additions were performed under the condi-
tions described in our previous papers.3,8b,9 We started
with the aldol reactions of ketone 1 and R-methyl
aldehydes (R)-3 and (S)-3. The results are shown in
Scheme 3 and Table 1. The reactions with aldehydes (S)-3
were comparatively rapid (total conversion in 5-6 h) and
(7) All chiral aldehydes were prepared according to literature
procedures. Aldehydes (R)-3a and (S)-3a : Roush, W. R.; Palkowitz,
A. D.; Palmer, M. A: J . J . Org. Chem. 1987, 52, 316-318. Aldehyde
(R)-3b: Meyers, A. I.; Babiak, K. A,; Campbell, A. J .; Comins, D. L.;
Fleming, M. P.; Henning, R.; Heuschmann, M.; Hudspeth, J . P.; Kane,
J . M.; Reider, P. J .; Roland, D. M.; Shimizu, K.; Tomioka, K.; Walkup,
R. D. J . Am. Chem. Soc. 1983, 105, 5015-5024. Aldehyde (S)-3b:
Nagaoka, H.; Kishi, Y. Tetrahedron 1981, 37, 3873-3888. Aldehydes
(R)-4a and (S)-4a : Massad, S. K.; Hawkins, L. D.; Baker, D. C. J . Org.
Chem. 1983, 48, 5180-5182. Aldehydes (R)-4b and (S)-4b: Ito, Y.;
Kobayashi, Y.; Kawabata, T.; Takase, M.; Terashima, S. Tetrahedron
1989, 45, 5767-5790. Aldehyde (R)-4c: Ohgo, Y.; Yoshimura, J .; Kono,
M.; Sato, T. Bull. Chem. Soc. J pn. 1969, 42, 2957-2961. Aldehyde (S)-
4c was prepared via Swern oxidation of (R)-1,2-di-O-benzylglycerol,
obtained from L-ascorbic acid as described in: Mikkilineni, A. B.;
Kumar, P.; Abushanab, E. J . Org. Chem. 1988, 53, 6005-6009.
(8) (a) The aldol reactions were also made with aldehydes bearing
a TBS protecting group. The results were identical with those of TPS
protected aldehydes. (b) The stereostructures of the aldol products were
established with the aid of the chemical correlation methodology used
in our previous papers3 and, in one case, via X-diffraction analysis.9
Aldol adducts were reduced in situ with LiBH4 to yield the expected
syn-1,3-diols.10 These were subsequently converted into acetonides,
which were then studied by means of NMR.11 Standard manipulations
of the protecting groups further permitted the preparation of other
cyclic derivatives suitable for similar NMR studies. Descriptions of
these chemical correlations and analytical data for the correlation
products are given in the Supporting Information.
(10) Paterson, I.; Channon, J . A. Tetrahedron Lett. 1992, 33, 797-
800.
(11) Rychnovsky, S. D.; Rogers, B. N.; Richardson, T. I. Acc. Chem.
Res. 1998, 31, 9-17.
(12) Zimmerman, H. E.; Traxler, M. D. J . Am. Chem. Soc. 1957,
79, 1920-1923.
(13) For theoretical studies on boron aldol reactions, see: (a) Li, Y.;
Paddon-Row, M. N.; Houk, K. N. J . Org. Chem. 1990, 55, 481-493.
(b) Goodman, J . M.; Kahn, S. D.; Paterson, I. J . Org. Chem. 1990, 55,
3295-3303. (c) Bernardi, A.; Capelli, A. M.; Gennari, C.; Goodman, J .
M.; Paterson, I. J . Org. Chem. 1990, 55, 3576-3581. (d) Bernardi, A.;
Capelli, A. M.; Comotti, A.; Gennari, C.; Gardner, M.; Goodman, J .
M.; Paterson, I. Tetrahedron 1991, 47, 3471-3484. (e) Bernardi, F.;
Robb, M. A.; Suzzi-Valli, G.; Tagliavini, E.; Trombini, C.; Umani-
Ronchi, A. J . Org. Chem. 1991, 56, 6472-6475. (f) Gennari, C.; Vieth,
S.; Comotti, A.; Vulpetti, A.; Goodman, J . M.; Paterson, I. Tetrahedron
1992, 48, 4439-4458. (g) Vulpetti, A.; Bernardi, A.; Gennari, C.;
Goodman, J . M.; Paterson, I. Tetrahedron 1993, 49, 685-696. See also
ref 3c.
(9) The stereostructure of compound 6a was established by means
of an X-ray diffraction analysis. Crystallographic data (excluding
structure factors) have been deposited at the Cambridge Crystal-
lographic Data Center as Supporting Information with reference
CCDC-210749. Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK [fax
+44(0)-1223-336033 or e-mail deposit@ccdc.cam.ac.uk].
8578 J . Org. Chem., Vol. 68, No. 22, 2003

Aldol Reactions of Erythrulose Derivatives with Chiral Aldehydes
SCHEME 4. TS of th e Ald ol Ad d ition of th e Bor on
En ola te of 1 to Ach ir a l Ald eh yd es
Indeed, all these factors are present in the TSs depicted
in Scheme 6. The reaction of aldehydes (S)-3 yields solely
stereoisomers 6, in contrast to the results predicted by
the Felkin-Anh model (such addition products are often
called, somewhat confusingly, “anti-Felkin” stereoiso-
mers). This can only be explained by assuming a TS such
as TS-2, which is still of the Felkin-Anh-type (enolate
attack anti to the bulky CH2OP group), but in which the
enolate approaches along an unfavorable Bu¨rgi-Dunitz
trajectory that pushes the nucleophile toward the methyl
group, rather than to the hydrogen atom. This negative
feature of the TS is probably of minor importance (see
conclusions). The alternative, nonobserved attack of the
enolate to the aldehyde Si face must take place, under
the assumed avoidance of syn-pentane interactions,
through the non-Anh rotamer TS-1,19 which additionally
shows an unfavorable steric crowding between the diox-
olane ring and one of the bulky boron ligands. This latter
feature can be alleviated by means of bond rotation, but
only at the cost of increasing the dipolar repulsion
between the C-Oenolate and CR-O bonds. It thus seems
that in the present case, stereocontrol is exerted by the
π-facial bias of the chiral enolate (preferred attack to
aldehyde Re faces), which overrides any Felkin-Anh bias
of the chiral aldehyde. This is not surprising as this bias
is never strong (dr usually < 3:1) for R-methyl alde-
hydes.1,16
In contrast, the aldol reaction with aldehydes (R)-3 led
mainly to the Felkin stereoisomer 7, with the same
absolute configuration at the two newly formed stereo-
genic centers as was found in 6 (Scheme 6). This
stereochemical outcome, again the result of preferred
enolate attack to the aldehyde Re face, can be explained
only if the aldol process occurs via the non-Anh rotamer
TS-4. As mentioned above, this negative feature repre-
sents a minor factor and is less destabilizing than the
syn-pentane interaction present in the true Felkin-Anh
rotamer TS-5.17 As in the previous case, enolate attack
to the aldehyde Si face (TS-3) to yield the “anti-Felkin”
stereoisomer 8 would suffer from steric crowding between
the dioxolane ring and one of the B-cyclohexyl groups. It
is worth mentioning here that stereoisomer 8 was formed
as a minor component when P ) benzyl but was not
detected when P ) silyl (Table 1). This indicates that
some remote influence of the â-protecting group may be
present here.20
SCHEME 5. syn -P en ta n e Rep u lsion in th e TS of th e
Ald ol Ad d ition of a Z Bor on En ola te to a n r-Ch ir a l
Ald eh yd e
to the π-facial bias of the chiral enolate, related in turn
to two energetically favorable features of the depicted
TS: the anticoplanar orientation of the C-Oenolate and
CR-O bonds (minimized dipolar repulsion)1e,14 and the
spatial allocation of the dioxolane ring away from the
bulky boron ligands (minimized steric crowding).3c
In the present situation, the chiral aldehydes under
study display diastereotopic carbonyl faces, whose rela-
tive reactivity toward nucleophiles is dictated by a
combination of electronic and steric effects, customarily
summarized in the Felkin-Anh model and its later
refinements (nucleophile attack anti-coplanar to either
the bulkiest aldehyde CR substituent or that having the
lowest lying σ*C-X orbital), including the most favorable
Bu¨rgi-Dunitz trajectory (approach nearer to the smallest
CR substituent, usually an H atom).15,16 For R-methyl
aldehydes such as 3, this model predicts the formation
of the stereoisomer with a syn relationship between the
vicinal hydroxy and methyl groups (as in 7). When the
Felkin-Anh model is applied to the doubly diastereose-
lective aldol additions of erythrulose boron enolates to
aldehydes 3, the essential features of the TS of Scheme
4 have to be combined with those of the model. In
addition, another important feature is the avoidance of
syn-pentane repulsive interactions between the OTBS
group at the enolate CdC bond and one substituent at
the stereogenic R-aldehyde carbon (Scheme 5 shows this
for a dicyclohexylboron enolate; L and M are the large
and medium substituents, respectively, at the aldehyde
R-carbon atom). This feature, very often associated with
aldol TSs involving Z enolates, is believed to be a
dominant stereocontrol element that determines alde-
hyde π-facial selectivities in many situations.1a,17,18
We then investigated the aldol reactions of ketone 1
and R-oxygenated aldehydes (R)-4 and (S)-4. The results
(19) Lodge, E. P.; Heathcock, C. H. J . Am. Chem. Soc. 1987, 109,
3353-3361. The “non-Anh” label refers to transition structures in
which attack takes place anti to a substituent that neither has the
lowest lying σ*C-X orbital (for R-heteroatom-substituted aldehydes) nor
is the sterically bulkiest one (for aldehydes not bearing R-heteroatoms).
See also ref 16.
(14) Van Draanen, N. A.; Arseniyadis, S.; Crimmins, M. T.; Heath-
cock, C. H. J . Org. Chem. 1991, 56, 2499-2506. For another example
that shows the importance of dipole alignment in aldol TSs, see:
Boeckman, R. K., J r.; Connell, B. T. J . Am. Chem. Soc. 1995, 117,
12368-12369.
(15) (a) Che´rest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968,
9, 2199-2204. (b) Anh, N. T. Top. Curr. Chem. 1980, 88, 145-162.
(16) Gawley, R. E.; Aube´, J . Principles of Asymmetric Synthesis;
Pergamon: New York, 1996; Chapters 4 and 5.
(20) Effects of â-protecting groups at either the enolate or the
aldehyde on the degree of diastereoselectivity of aldol reactions have
been previously described. In some cases, such effects were attributed
to various origins (chelation, conformational effects, etc.) while in others
they remained unexplained. See, for example: (a) Paterson, I.; Bower,
S.; Tillyer, R. D. Tetrahedron Lett. 1993, 34, 4393-4396. (b) Evans,
D. A.; Calter, M. A. Tetrahedron Lett. 1993, 34, 6871-6874. (c) Evans,
D. A.; Ng, H. P.; Rieger, D. L. J . Am. Chem. Soc. 1993, 115, 11446-
11459. (d) Martin, S. F.; Lee, W.-C.; Pacofsky, G. J .; Gist, R. P.;
Mulhern, T. A. J . Am. Chem. Soc. 1994, 116, 4674-4688. (e) Roush,
W. R.; Dilley, G. J . Tetrahedron Lett. 1999, 40, 4955-4959. Sometimes,
influence of still more remote protecting groups has been observed:
Roush, W. R.; Bannister, T. D.; Wendt, M. D. Tetrahedron Lett. 1993,
34, 8387-8390.
(17) Roush, W. R. J . Org. Chem. 1991, 56, 4151-4157.
(18) For a lucid analysis of the various factors which may influence
the stereochemical outcome of aldol reactions, see: Lee, C. B.; Wu, Z.;
Zhang, F.; Chappell, M. D.; Stachel, S. J .; Chou, T.-C.; Guan, Y.;
Danishefsky, S. J . J . Am. Chem. Soc. 2001, 123, 5249-5259.
J . Org. Chem, Vol. 68, No. 22, 2003 8579

Marco et al.
SCHEME 6. F elk in -An h TSs for Ald ol Ad d ition s of th e Bor on En ola te of 1 to Ald eh yd es (R)-3 a n d (S)-3
SCHEME 7. Ald ol Ad d ition s of th e Bor on En ola te
of 1 to Ald eh yd es (R)-4 a n d (S)-4
(matched double diastereoselection) to react across a
standard Felkin-Anh TS. However, this is not the case.
As shown in Scheme 8, the Felkin-Anh TS-6 suffers
from steric crowding that arises from a syn-pentane
interaction1a,17 between the enolate OTBS group and the
R group in the aldehyde moiety. If this repulsion is
avoided by means of a CR-CHO bond rotation, a non-
Anh rotamer19 results (TS-7) in which enolate attack
takes place through a trajectory that is anti to the R
group. Although we have commented above that this may
be a factor of minor importance with R-alkyl aldehydes
such as 3, this is not necessarily so for R-heteroatom-
substituted aldehydes.22,23
are given in Scheme 7. The reactions with aldehydes (S)-4
were highly diastereoselective and gave aldols 9 (dr >
95:5), once again through enolate attack to the aldehyde
Re face. In contrast, the reactions of aldehydes (R)-4 were
extremely slow (less than 50% conversion after 12 h),
yielding complex mixtures of 3-4 stereoisomeric aldols
together with extensive decomposition.21
This last result was both unanticipated and disap-
pointing, especially in view of the aforementioned find-
ings and literature precedents.22 For nucleophilic attacks
to R-oxygenated aldehydes of general structure RCH(OP)-
CHO, the Felkin-Anh model predicts the formation of
the stereoisomer with an anti relationship between the
vicinal hydroxy and OP groups, as observed in 9. At first
sight then, it seems that the π-facial bias of aldehydes
(S)-4 should work together with that of the chiral enolate
The reactions of aldehydes (R)-4 are even more difficult
to rationalize within the Felkin-Anh paradigm. Even if
a well-matched Felkin-Anh TS can be proposed in this
(22) Provided that chelation is not involved in the transition state,
achiral enolates react with R-oxygenated aldehydes to yield predomi-
nantly, albeit with variable diastereoselectivity, the Felkin aldols. See
refs 1 and 2. For more recent cases, see, for example: (a) Esteve, C.;
Ferrero`, M.; Romea, P.; Urp´ı, F.; Vilarrasa, J . Tetrahedron Lett. 1999,
40, 5079-5082. (b) Lu, L.; Chang, H.-Y.; Fang, J .-M. J . Org. Chem.
1999, 64, 843-853. However, it is worth noting that Felkin aldols have
been found to predominate in some reactions where chelation is likely
to occur: Grandel, R.; Kazmaier, U.; Rominger, F. J . Org. Chem. 1998,
63, 4524-4528.
(23) For nucleophilic additions to aldehydes bearing R-heteroatoms
other than oxygen, see, for example: (a) Reetz, M. T. Chem. Rev. 1999,
99, 1121-1162 (R-amino aldehydes). (b) Enders, D.; Piva, O.; Burkamp,
F. Tetrahedron 1996, 52, 2893-2908 (R-sulfenyl aldehydes). (c) Enders,
D.; Adam, J .; Klein, D.; Otten, T. Synlett 2000, 1371-1384 (R-silyl
aldehydes). See also: Enders, D.; Burkamp, F. Collect. Czech. Chem.
Commun. 2003, 68, 975-1006.
(21) This was established upon examination of NMR data for the
crude aldol mixtures. In view of this synthetically useless result, we
did not attempt to isolate individual compounds.
8580 J . Org. Chem., Vol. 68, No. 22, 2003

Aldol Reactions of Erythrulose Derivatives with Chiral Aldehydes
SCHEME 8. F elk in -An h TSs for Ald ol Ad d ition s
of th e Bor on En ola te of 1 to Ald eh yd es (R)-4 a n d
(S)-4
SCHEME 10. Cor n for th TSs for Ald ol Ad d ition s of
th e Bor on En ola te of 1 to Ald eh yd es (R)-4 a n d
(S)-4
less crowded face of the CdO function. As shown in
Scheme 9, this model predicts the formation of the same
1,2-anti configurated adduct for simple nucleophiles as
the Felkin-Anh model. It is perhaps for this reason that
the Cornforth model has to a great extent been sup-
planted by the theoretically more grounded Felkin-Anh
paradigm, which has become the standard explanation
for such processes. The contribution of Evans and co-
workers,24 which specifically deals with aldol reactions
assumed to proceed through cyclic, chairlike transition
states of the Zimmerman-Traxler type, is the first case
where both models have been found to predict different
outcomes. The aforementioned authors demonstrated
that although the stereochemical outcome of aldol addi-
tions of the lithium and boron enolates of ethyl isopropyl
ketone to a range of chiral, R-oxygenated aldehydes could
not be conveniently accounted for with the Felkin-Anh
model, they did receive a satisfactory explanation within
the Cornforth paradigm.26
We thus applied Cornforth’s model to the case of
aldehydes 4. A reexamination of the transition structures
proposed above (Scheme 8) reveals that for reactions with
(S)-4, only TS-7 displays the geometry proposed by the
Cornforth model (Scheme 10). In addition, this TS (a
“non-Anh” TS in Heathcock’s nomenclature19) not only
shows none of the other previously mentioned destabiliz-
ing features, but also implies the formation of stereoiso-
mers 9, as experimentally observed. The alternative
rotameric TS-6 also predicts the formation of 9, but
shows an unfavorable syn-pentane repulsion. We can
thus conclude that a Cornforth-type geometry is intrinsi-
cally more favorable in additions to R-heteroatom-
substituted aldehydes than that predicated by the Felkin-
Anh model. This conclusion seems to be bolstered by the
results of the slow and nonstereoselective aldol additions
with aldehydes (R)-4. That the seemingly ideal Felkin-
SCHEME 9. Cor n for th TS of Nu cleop h ilic
Ad d ition s to r-Oxygen a ted Ald eh yd es
case with no unfavorable features (TS-8), the process
turns out to be very slow in practice, as commented
above, and is both nonstereoselective and accompanied
by a great amount of decomposition. It thus appears that
the standard Felkin-Anh model does not provide a
satisfactory explanation for these kinds of aldol additions
to R-oxygenated aldehydes 4.
A very recent publication by Evans and co-workers
called our attention to an alternative explanation.24 To
account for the stereochemical outcome of aldol additions
to aldehydes of general formula R-CH(X)-CHO (X )
electronegative heteroatom), these authors proposed a
resurrection of the classic Cornforth model.16,25 According
to this model, nucleophilic additions to such carbonyl
compounds take place in a conformation in which the Cd
O and C-X dipoles are at maximum opposition (anti-
coplanar, Scheme 9), thus minimizing dipolar repulsion.
Nucleophilic attacks then take place from the sterically
(24) Evans, D. A.; Siska, S. J .; Cee, V. J . Angew. Chem., Int. Ed.
2003, 42, 1761-1765.
(25) Cornforth, J . W.; Cornforth, R. H.; Mathew, K. K. J . Chem. Soc.
1959, 112-127. See also: Paddon-Row, M. N.; Rondan, N. G.; Houk,
K. N. J . Am. Chem. Soc. 1982, 104, 7162-7166.
(26) The Cornforth model, as an alternative to the Felkin-Anh
model, has already been proposed to explain the stereochemical
outcome of additions of allylboranes to R-heteroatom-substituted
aldehydes. See: Brinkmann, H.; Hoffmann, R. W. Chem. Ber. 1990,
123, 2395-2401 and references therein.
J . Org. Chem, Vol. 68, No. 22, 2003 8581

Marco et al.
Anh TS-8 is not able to determine the stereochemical
outcome of the process can be explained if we assume
that its deviation from the Cornforth geometry increases
its energy content to such an extent as to make this
pathway very slow. Attempts to achieve a Cornforth-type
geometry through CR-CHO bond rotation would only
cause a syn-pentane repulsion between the OTBS and
the R group (cf. Scheme 8, TS not depicted). Furthermore,
enolate attack to the Si aldehyde face through a Corn-
forth geometry would lead to TS-9 (Scheme 10), with the
same type of steric crowding observed in both TS-1 and
TS-3. Since this pathway is also expected to be slow, this
explains not only the overall lack of stereoselectivity of
this aldol process, but also its sluggishness and resulting
proclivity to decomposition.
bond rotation relieves this interaction but simultaneously
increases the dipolar repulsion between the C-Oenolate
and CR-O bonds.
(d) The Felkin-Anh π-facial bias is not very strong for
aldehydes bearing only carbon R-substituents. In this
case, stereocontrol is frequently exerted by the chiral
enolate rather than by the aldehyde. This may give rise
to rotameric TSs in which a steric repulsion develops
between the incoming enolate and one aldehyde non-
hydrogen R-substituent due to approach along an unfa-
vorable Bu¨rgi-Dunitz trajectory (TS-2, TS-4).
Taking all these factors into consideration, we propose
that R-methyl aldehydes (S)-3 react with the boron
enolate of 1 to yield aldols 6 selectively through TS-2
whereas aldehydes (R)-3 generate mainly or exclusively
aldols 7 through TS-4 (Scheme 6), in both cases due to
control by the chiral enolate. The R-oxygenated aldehydes
(S)-4 react to yield aldols 9 exclusively through the
Cornforth transition state TS-7 (Scheme 10). Their
enantiomers (R)-3 react sluggishly and nonstereoselec-
tively because the energy of the Cornforth-type TS-9
(Scheme 10) is increased by factor (c). The energy of the
alternative TS-8 is also increased by its deviation from
the Cornforth geometry, i.e. factor (a) prevails. In view
of these and of the recent results described by Evans and
co-workers,24 it seems that the Cornforth model will have
to be revitalized to explain the stereochemical outcome
of aldol additions to R-heteroatom-substituted carbonyl
groups.26-28
Con clu sion s
The Cornforth model has now been applied for the first
time to a doubly diastereoselective aldol reaction.27,28 We
can summarize the results disclosed above within a
unified general concept in which several factors must be
taken into account to predict the stereochemical outcome
of boron aldol reactions of ketones such as 1 with R-chiral
aldehydes. Cyclic transition states of the Zimmerman-
Traxler type are assumed with no chelation issues
involved. In order of energetically decreasing importance,
these factors are the following:
(a) For R-heteroatom-substituted aldehydes, Cornforth
TSs (TS-7) are markedly preferred.
Ack n ow led gm en t . Financial support has been
granted by the Spanish Ministry of Education and
Science (project BQU2002-00468) and by the Fundacio´
Caixa Castello`-University J aume I (project PI-1B2002-
06). J .M. thanks the Spanish Ministry of Science and
Technology for a Ramo´n y Cajal fellowship. S.D.-O.
thanks the Conseller´ıa de Educacio` de la Generalitat
Valenciana for a predoctoral fellowship. The authors
further thank the S.C.S.I.E. at the University of Va-
lencia for mass spectral measurements.
(b) syn-Pentane repulsions between the OTBS group
and one aldehyde non-hydrogen R-substituent are ener-
getically important interactions (TS-5, TS-6) that must
be avoided through C-C bond rotation.
(c) Steric crowding between the dioxolane ring and one
B-cyclohexyl group arises when attack takes place from
the enolate Si face (TS-1, TS-3, TS-9). Suitable C-C
(27) Doubly diastereoselective aldol reactions of chiral enolates with
chiral R-oxygenated aldehydes are not extensively documented in the
literature. Matched and mismatched processes have been reported,
with the full range from total stereocontrol by the enolate to complete
dominance of the aldehyde being observed. See refs 1 and 2 and, for
more recent cases: (a) Nicolaou, K. C.; Piscopio, A. D.; Bertinato, P.;
Chakraborty, T. K.; Minowa, N.; Koide, K. Chem. Eur. J . 1995, 1, 318-
333. (b) Kobayashi, S.; Furuta, T. Tetrahedron 1998, 54, 10275-10294.
(c) Esteve, C.; Ferrero`, M.; Romea, P.; Urp´ı, F.; Vilarrasa, J . Tetrahe-
dron Lett. 1999, 40, 5083-5086. (d) Nicolaou, K. C.; Pihko, P. M.;
Diedrichs, N.; Zou, N.; Bernal, F. Angew. Chem., Int. Ed. 2001, 40,
1262-1265. (e) Forsyth, C. J .; Hao, J .; Aiguade, J . Angew. Chem., Int.
Ed. 2001, 40, 3663-3667.
Note Ad d ed a fter ASAP P ostin g. Page S7 of the
Supporting Information file posted ASAP October 7,
2003, contained errors. A corrected Supporting Informa-
tion file was posted October 9, 2003.
Su p p or tin g In for m a tion Ava ila ble: Description of ex-
perimental procedures; tabulated physical and spectral data
of aldols 6a , 6b, 7a , 9a , 9b, and 9c; description of the chemical
correlations used to establish the configurations of the aldols
and spectral data of several selected correlation products. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(28) Excluded from this discussion are chiral enolates in which the
chirality resides in the ligands bound to the heteroatom. In these cases,
stereocontrol by the chiral auxiliary is usually observed. See, for
example: Gennari, C.; Pain, G.; Moresca, D. J . Chem. Org. 1995, 60,
6248-6249.
J O035159J
8582 J . Org. Chem., Vol. 68, No. 22, 2003