Lewis acid-aldehyde-solvent interactions. A computational approach for determining the critical amount of THF molecules necessary for achieving a high enantioselection in chiral oxazaborolidinone-promoted asymmetric aldol reactions

Article abstract of DOI:10.1016/j.tetlet.2004.12.074

The effect of THF as a solvent on the enantioselectivity of oxazaborolidinone-promoted asymmetric aldol reactions of aldehydes with silylketene acetals was investigated. The use of 4-5 M equiv of THF, relative to the chiral borane, was required to achieve

Full text of DOI:10.1016/j.tetlet.2004.12.074

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 1211–1215
Lewis acid–aldehyde–solvent interactions.
A computational approach for determining the critical amount of
THF molecules necessary for achieving a high enantioselection
in chiral oxazaborolidinone-promoted asymmetric aldol reactions
Ryoji Fujiyama,* Kazuki Goh and Syun-ichi Kiyooka
Department of Material Science, Faculty of Science, Kochi University, Akebono-cho, Kochi 780-8520, Japan
Received 23 August 2004; revised 29 November 2004; accepted 10 December 2004
Available online 8 January 2005
Abstract—The effect of THF as a solvent on the enantioselectivity of oxazaborolidinone-promoted asymmetric aldol reactions of
aldehydes with silylketene acetals was investigated. The use of 4–5 M equiv of THF, relative to the chiral borane, was required
to achieve a high enantioselectivity. A solvent-modeling study, based on ab initio calculations of intermolecular interactions,
revealed the existence of an extended hydrogen bonding network in the resulting assembly, which was composed of THF molecules,
the aldehyde, and the oxazaborolidinone. The model rationally provides a spatially suitable active site for controlling the stereo-
chemical outcome of the reaction.
Ó 2004 Elsevier Ltd. All rights reserved.
Chiral oxazaborolidinone-promoted asymmetric aldol
reactions of aldehydes with silylketene acetals are widely
used in the construction of macrolide skeletons, having
O
O
B
TsN
O
(S)-1
B
H
1,3-polyhydroxy stereocenters, because of the high
resulting enantioselectivity.1 Stoichiometric reaction
conditions in CH2Cl2 are typically used in the labora-
tory in reproducing the level of enantioselectivity.2 The
high facial selectivity can be explained satisfactorily by
assuming a transition state model, as shown in square
brackets in Scheme 1. Nontraditional hydrogen bonding
between the ring oxygen and the formyl hydrogen, as
proposed by Corey, plays an important role in effec-
tively fixing the face being attacked by a silyl nucleophile
along with the cooperative coordination of the borane
to the aldehyde carbonyl function.3 A theoretical study
with coordination models between aldehydes and the re-
lated simplified boranes indicated that a different con-
former has a lower energy minima.4 In spite of these
previous studies, questions concerning this high selectiv-
ity continue to remain because the front side of the chi-
ral borane is extensively exposed. The use of THF was
investigated in the course of a search for a more appro-
priate solvent to replace CH2Cl2 in the reaction from an
O
O
Nu
H
N
H
OTMS
H
(1 equiv.)
Ts
PhCHO
+
OEt Solvent/ -78ºC/ 3 h
2
Ph
(OTMS)
Si-facial selectivity
OH
O
Ph
OEt
3
Scheme 1.
environmental point of view, and, as a result, a remark-
able solvent effect was found. We propose herein an
explanation of the origin of the high enantioselectivity
in asymmetric aldol reactions that require a critical
amount of THF. Our conclusions are based on ab initio
calculations of an appropriate model system and involve
the participation of THF molecules.
*
Corresponding author. Tel.: +81 88 844 8300; fax: +81 88 844
8359; e-mail: fujiyama@cc.kochi-u.ac.jp
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.12.074

1212
R. Fujiyama et al. / Tetrahedron Letters 46 (2005) 1211–1215
In preliminary studies5 of enantioselective Diels–Alder
reactions with chiral oxazaborolidinones, Helmchen
pointed out that ca. 12 equiv of THF are initially in-
volved in a typical procedure for the preparation of
the oxazaborolidinones as the result of adding BH3ÆTHF
(1 M equiv in THF) to a suspension of N-arylsulfonyl-a-
amino acid in CH2Cl2 and that the quantity of the donor
additive is necessary for achieving a high enantioselec-
tivity in the reaction and the lowering of selectivity in
reactions without such a donor solvent is caused by an
association of the catalysts used, that is, a unique species
multiplicity.6a In the case of the chiral oxazaborolidi-
none-catalyzed 1,3-dipolar cycloaddition of a nitrone
to ketene acetals, contrary to the results of the above
Diels–Alder reaction, a high degree of enantioselection
could be achieved by excluding donor solvents (such
as THF) from the reaction mixture.6b We first confirmed
THF solvent effects in the chiral oxazaborolidinone-pro-
moted asymmetric aldol reaction of benzaldehyde with
silylketene acetal 2 (Table 1): the reaction, under normal
conditions, gave the aldol with 88% ee in the presence of
(S)-1, prepared from N-p-toluenesulfonyl-(S)-valine and
a 1 M THF solution of BH3ÆTHF in CH2Cl2, while the
reaction without THF resulted in a low enantioselection
of 28% ee. No deuterium solvent (THF-d8) effects on the
selectivity were observed. Further, no deuterium isotope
effects were observed in control reactions using deute-
rium benzaldehyde (CDO). Thus, the THF molecules
in question do not appear to participate in bond-break-
ing and -forming processes at the transition state, where
the selectivity is determined.7 Therefore, the effects of
THF on selectivity might be attributable to the forma-
tion of a type of solvent cluster by the coordination of
THF molecules around the borane in the reaction. The
corresponding reaction in using only THF as the solvent
proceeded smoothly to give the aldol product of 82% ee
in high yield. This small but significant decrease in selec-
tivity indicates that a large excess of THF disturbs the
optimal structure of the solvent cluster slightly. How-
ever, THF is clearly a good substitute solvent for
CH2Cl2 in this aldol reaction. In addition, when steri-
cally bulky ether solvents, which allow free rotation
around the C–O–C bond, for example, diisopropyl
ether, diphenyl ether, and dibenzyl ether, were used, sig-
nificant reductions in enantioselectivity were observed
(20–30%). THF is the most advantageous solvent for
such reactions among these ethers, presumably because
of its inherent compact cyclic structure.
Control experiments on the proportion of THF to (S)-1
were performed in order to determine the precise quan-
tity of THF required to achieve a high enantioselectivity
in the reaction. Unless otherwise noted, the asymmetric
aldol reactions described herein were carried out at
ꢀ78 °C for 3 h under stoichiometric conditions because
accurate ratios of the borane–aldehyde complex to THF
molecules must be correctly determined.2 An interesting
behavior related to the dependence of enantioselectivity
on the quantity of THF present was found in the reac-
tion of benzaldehyde with 2 to give aldol adduct 3, as
shown in Figure 1. Initially, (S)-1 was prepared under
THF-free conditions from BH3ÆS(CH3)2 in CH2Cl2.
Under THF-free conditions, the reaction resulted in a
low enantioselectivity of only 28% ee. The complex
(S)-1 was prepared in situ with an aliquot of THF added
into the starting CH2Cl2 solution in each run.8 When the
molar equivalent of THF molecule approached 5, the lev-
els of enantioselectivity became constant at 86–88% ee.
Table 1. A typical THF solvent effect in the chiral oxazaborolidinone-
promoted asymmetric aldol reaction of benzaldehyde with 2 in the
presence of a stoichiometric amount of (S)-1 (Scheme 1)
Entry Promoter formation conditions
% ee Yield (%)
1
2
3
(S)-1, prepared from BH3ÆTHF
in CH2Cl2
88
76
75
85
(S)-1, prepared from BH3ÆS(CH3)2 28
in CH2Cl2
(S)-1, prepared from BH3ÆTHF
in THF
82
100
90
80% ee
80% yield
80
88% ee
82% ee
86% ee
76% yield
85% yield
85% yield
70
60
63% ee
81% yield
50
40
30
20
10
28% ee
75% yield
0
12
0
5
10
THF only
THF (molar equiv.)
Figure 1. Enantioselectivity as a function of the quantity of THF in the reaction of benzaldehyde with 2 to give aldol adduct 3 (Ref. 8).

R. Fujiyama et al. / Tetrahedron Letters 46 (2005) 1211–1215
1213
Thus, the presence of only a 4–5 M equiv of THF appears
to be the critical quantity necessary for satisfying the
inherent enantioselectivity. Presumably, this quantity of
THF permits the formation of an effective solvent clus-
ter around the chiral borane–aldehyde complex. This
interesting dependence of enantioselectivity on THF
concentration in the borane-promoted aldol reaction is
the first quantitative example indicating the importance
of Lewis acid–solvent interactions.
cule, were performed in order to estimate the nature
of the interactions of one THF molecule with A and B
at the HF/3-21G* level. As expected, the optimization
of the assemblies formed from the complexes (A or B)
and one THF molecule indicates a nucleophilic interac-
tion of the oxygen of THF to the borane-activated car-
bonyl carbon of the aldehyde. The assemblies are not
adequate as a precise model because the nucleophilic
interaction of the oxygen atom of one THF molecule
to the coordinated carbonyl carbon completely prevents
the approach of the silyl nucleophile in the subsequent
asymmetric aldol reaction process. On the other hand,
the optimization of the assembly formed from A and
one THF molecule also resulted in the geometry of
structure A0 with a unique additional hydrogen bonding
of the formyl hydrogen to the oxygen of THF where the
THF molecules clearly surround the chiral oxazaboro-
lidinone–aldehyde complex and might control the effi-
cient approach of a silyl nucleophile to the core center
involving a boron atom. At this stage, however, it ap-
peared to be difficult to obtain additional new experi-
mental information concerning the Lewis acid (the
borane)–Lewis base (THF) interaction. A computa-
tional approach represents an alternative to resolve such
an ambiguous problem. Thus, calculations were per-
formed to obtain a rational model for the THF solvent
effects. In the calculation, acetaldehyde was used as a
model aldehyde in order to reduce the time required
for the computation. Two structures (A and B), opti-
mized in advance, were selected as valid models for the
borane (S)-1-aldehyde complex, as shown in Figure 2.
Structure A has been frequently employed to explain
the enantioselectivity in our aldol reaction.1 The facial
selectivity is thought to be enhanced by hydrogen bond-
ing of the formyl hydrogen to the ring oxygen3 and the
increasing importance of a cooperative C–Hꢁ ꢁ ꢁO inter-
action of the aldehydic proton (formyl hydrogen) with
the oxygen of other substrates also should be consid-
ered.9 Structure B, in which a hydrogen bonding be-
tween the hydrogen of B–H and the formyl hydrogen
is present, was selected as an alternative because it has
been calculated to be 3–4 kcalmolꢀ1 lower than struc-
˚
length of the hydrogen bonding is 2.096 A.
An assembly was computed as spatially plausible model,
consisting of the borane–aldehyde complex A and four
THF molecules in ascertaining the minimal structure
of the THF cluster around the complex at the B3LYP/
6-31G** level. The optimized geometry is clearly dem-
onstrated by the important interatomic distances in Fig-
ure 3. In the geometry, four THF molecules surround
the borane–aldehyde complex. Two THF molecules
(THF #3 and THF #4) are located above the top side
of the oxazaborolidinone ring with a low stabilization
energy of 2–3 kcalmolꢀ1. One THF molecule (THF
THF #3
THF #4
ture A at the B3LYP/6-31G** level by Salvatella and
4
´
Ruiz-Lopez. Preliminary calculations of the assemblies,
consisting of the complexes (A or B) and one THF mole-
2.466Å
2.388Å
Me
O
Me
Me
O
Me
1.692Å
O
O
Me
O
Ts
B
H
N
B
O
2.279Å
H
H
N
H
Ts
H
B
A
Me
THF #1
THF #2
Me
Me
O
O
Me
Me
Me
O
O
O
Ts
O
O
O
Me
N
THF
O
H
N B
H
N
H
B
H
Me
THF
H
Ts
H
B'
O
Ts
B
Me
A'
O
H
H
THF
Figure 2. Complex A is our model leading to a satisfactory explanation
of the stereochemical outcome in the asymmetric reaction. An
´
alternative complex B is the Ruiz-Lopez model having the lowest
O
THF
Me
binding energy in a simplified system. The geometry of the assemblies
A0 and B0 from the complexes (A and B) and one THF molecule was
obtained by preliminary HF/3-21G* calculations.
Figure 3. The geometry optimized from the oxazaborolidinone–
acetaldehyde complex and four THF molecules by a B3LYP/6-
31G** set and intermolecular hydrogen-bonding interactions.

1214
R. Fujiyama et al. / Tetrahedron Letters 46 (2005) 1211–1215
#1) is present within a cavity formed by the ring substit-
uents at the bottom side and blocks the approach of the
silyl nucleophile leading to the opposite facial selectivity.
The stabilization energy of borane–aldehyde complex
and this THF was calculated to be 8 kcalmolꢀ1. The
most important THF molecule (THF #2) is situated to
the right of the borane–aldehyde complex with a high
stabilization energy of 8 kcalmolꢀ1. As expected, this
THF molecules (THF #2) plays a specific role in the
enatioselectivity. The originally assumed hydrogen
bonding between the formyl hydrogen and the ring oxy-
gen of the complex was exactly reproduced and the bond
References and notes
1. (a) Kiyooka, S.-i.; Shahid, K. A.; Goto, F.; Okazaki, M.;
Shuto, Y. J. Org. Chem. 2003, 68, 7967–7978; (b)
Kiyooka, S.-i. Tetrahedron: Asymmetry 2003, 14, 2897–
2910; (c) Shahid, K. A.; Li, Y.-N.; Okazaki, M.; Shuto, Y.;
Goto, F.; Kiyooka, S.-i. Tetrahedron Lett. 2002, 43, 6373–
6376; (d) Shahid, K. A.; Mursheda, J.; Okazaki, M.;
Shuto, Y.; Goto, F.; Kiyooka, S.-i. Tetrahedron Lett.
2002, 43, 6377–6381, and 2003, 44, 1519–1520; (e)
Kiyooka, S.-i.; Shiinoki, M.; Nakata, K.; Goto, F.
Tetrahedron Lett. 2002, 43, 5377–5380; (f) Kiyooka, S.-i.;
Shahid, K. A. Bull. Chem. Soc. Jpn. 2001, 74, 1485–1495;
(g) Kiyooka, S.-i.; Hena, M. A.; Yabukami, T.; Murai, K.;
Goto, F. Tetrahedron Lett. 2000, 41, 7511–7516; (h)
Kiyooka, S.-i.; Goh, K.; Nakamura, Y.; Takesue, H.;
Hena, M. A. Tetrahedron Lett. 2000, 41, 6599–6603; (i)
Kiyooka, S.-i.; Shahid, K. A. Tetrahedron Lett. 2000, 41,
2633–2637; (j) Kiyooka, S.-i.; Shahid, K. A.; Hena, M. A.
Tetrahedron Lett. 1999, 40, 6447–6449; (k) Kiyooka, S.-i.;
Hena, M. A.; Goto, F. Tetrahedron: Asymmetry 1999, 10,
2871–2879; (l) Kiyooka, S.-i.; Hena, M. A. J. Org. Chem.
1999, 64, 5511–5523; (m) Kiyooka, S.-i.; Maeda, H.;
Hena, M. A.; Uchida, M.; Kim, C.-S.; Horiike, M.
Tetrahedron Lett. 1998, 39, 8287–8290; (n) Hena, M. A.;
Terauchi, S.; Kim, C.-S.; Horiike, M.; Kiyooka, S.-i.
Tetrahedron: Asymmetry 1998, 9, 1883–1890; (o) Kiyooka,
S.-i.; Maeda, H. Tetrahedron: Asymmetry 1997, 8, 3371–
3374; (p) Kiyooka, S.-i.; Yamaguchi, T.; Maeda, H.; Kira,
H.; Hena, M. A.; Horiike, M. Tetrahedron Lett. 1997, 38,
3553–3556; (q) Kiyooka, S.-i. Rev. Heteroatom Chem.
1997, 17, 245–270; (r) Kiyooka, S.-i.; Kira, H.; Hena, M.
A. Tetrahedron Lett. 1996, 37, 2597–2600; (s) Kiyooka,
S.-i.; Hena, M. A. Tetrahedron: Asymmetry 1996, 7, 2181–
2184; (t) Kiyooka, S.-i.; Kaneko, Y.; Kume, K. Tetrahe-
dron Lett. 1992, 33, 4927–4930; (u) Kiyooka, S.-i.;
Kaneko, Y.; Komura, M.; Matsuo, H.; Nakano, M. J.
Org. Chem. 1991, 56, 2276–2278.
˚
length was found to be 2.388 A. Furthermore, an addi-
tional and interesting intermolecular interaction be-
tween the formyl hydrogen and the oxygen of THF
˚
was found to be 2.279 A, which is shorter than the for-
mer one. This suggests that the newly found hydrogen
bonding is stronger. The structure is additionally stabi-
lized by the additional interaction between the carbonyl
oxygen of the complex and one of THF hydrogens
˚
(2.466 A). These Hꢁ ꢁ ꢁO distances are well below the
˚
sum of the van der Waals radii of 2.72 A (H = 1.20 A
˚
10
˚
and O = 1.52 A). Thus, the resulting assembly is elec-
trostatically stabilized by three cooperative C–Hꢁ ꢁ ꢁO
hydrogen bonding interactions along with the original
Lewis acid–base interaction. The THF molecule (THF
#2) occupying such a position might be able to play a
role in controlling the approach of the silyl nucleophile.
That is to say, the front-side space formed by the bor-
ane–aldehyde complex and the THF molecule would
be maintained at the sequential transition state assembly
suitable for achieving a high enantioselectivity. The
THF cluster model obtained is consistent with the experi-
mental results presented above.
2. The asymmetric aldol reaction proceeded smoothly in
In conclusion, the use of 4–5 equiv of THF molecules at
the stage of oxazaborolidinone-catalyst formation was
found to be essential for satisfying the enantioselectivity
in oxazaborolidinone-promoted asymmetric aldol reac-
tions of aldehydes with silylketene acetals. The solvent
participates stereochemically in the tuning of an optimal
chiral field at the active site. A spatially plausible model
for the assembly of the borane–aldehyde complex with
four THF molecules was calculated and resulted in a
rational interpretation of the experimentally observed
THF solvent effects. This modeling study in searching
for the role of the THF solvent in question is an example
that justifies the utility of ab initio calculations in deter-
mining the nature of such solvent effects in considering
reactivity and selectivity in various types of reactions
involving Lewis acids.
nitroethane under semi-catalytic conditions with
a
20 mol % loading of the borane,1s but the level of
enantioselectivity sometimes varied depending on the
conditions used in purifying the solvent.
3. (a) Corey, E. J.; Lee, T. W. Chem. Commun. 2001, 1321–
1329; (b) Corey, E. J.; Rohde, J. J. Tetrahedron Lett. 1997,
38, 37–40; (c) Corey, E. J.; Barnes-Seeman, D.; Lee, T. W.
Tetrahedron Lett. 1997, 38, 1699–1702.
´
4. Salvatella, L.; Ruiz-Lopez, M. F. J. Am. Chem. Soc. 1999,
121, 10772–10780.
5. (a) Takasu, M.; Yamamoto, H. Synlett 1990, 194–196; (b)
Sartor, D.; Saffrich, J.; Helmchen, G. Synlett 1990, 197–
198.
6. (a) Sartor, D.; Saffrich, J.; Helmchen, G.; Richards, C. J.;
Lambert, H. Tetrahedron: Asymmetry 1991, 2, 639–642;
(b) Seerden, J.-P. G.; Kuypers, M. M. M.; Scheeren, H. W.
Tetrahedron: Asymmetry 1995, 6, 1441–1450.
7. (a) Isotopes in Organic Chemistry: Secondary and Solvent
Isotope Effects; Buncel, E., Lee, C. C., Eds.; Elsevier
Science: Oxford, 1987; (b) Reichardt, C. Solvents and
Solvent Effects in Organic Chemistry; VCH Publishers:
Weinheim, 1988.
Method of calculation: The calculations were performed
with a Gaussian 98 rev. A.11, using the HF and B3LYP
methods with 3-21G* and 6-31G** basis sets for geom-
etry optimizations.11
8. Experimental procedure: To a solution of N-p-toluene-
sulfonyl-(S)-valine (326 mg, 1.2 mmol) in CH2Cl2 (10 mL)
was added an aliquot of THF (0.1, 0.2, 0.5, and 1.0 mL).
A 1 M solution of BH3ÆS(CH3)2 in CH2Cl2 (1.0 mL,
1.0 mmol) was added over 5 min at 0 °C and the resulting
solution was stirred for 30 min. At ꢀ78 °C, benzaldehyde
(1 mmol, 0.5 mL of a CH2Cl2 solution) and silylketene
acetal 2 (1 mmol, 0.5 mL of a CH2Cl2 solution) were
Acknowledgements
This work was supported by a Grant-in-Aid for Scien-
tific Research from Japan Society for the Promotion
of Science.

R. Fujiyama et al. / Tetrahedron Letters 46 (2005) 1211–1215
1215
successively added and the reaction mixture was then
stirred for 3 h. After the usual workup, the aldol adduct 3
was obtained in high yields (80–90%). The % ee was
determined by chiral HPLC analysis using a Daicel
Chiralcel OD-H column.
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.; Raghav-
achari, 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, 2002.
9. (a) Raveendran, P.; Wallen, S. L. J. Am. Chem. Soc. 2002,
124, 12590–12599; (b) Blatchford, M. A.; Raveendran, P.;
Wallen, S. L. J. Am. Chem. Soc. 2002, 124, 14818–
14819.
10. Bondi, A. J. Phys. Chem. 1964, 68, 441–451.
11. (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652; (b)
Gaussian 98, Revision A.11: 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.;