Aldolases as leitmotiv - A nonenzymatic domino aldol reaction triggered by zinc bisenolate and polyenolate (ate) complexes

Article abstract of DOI:10.1055/s-2001-18766

A new zinc(II) mediated sequential aldol reaction is described for the diastereoselective one-pot synthesis of tetrahydropyran-2,4-diols. The dependence of the yield on the ratio Zn²⁺/enolate proposes that zinc(II) polyenolate and ate complexes may be involved.

Full text of DOI:10.1055/s-2001-18766

1992
LETTER
Aldolases as Leitmotiv – A Nonenzymatic Domino Aldol Reaction Triggered
by Zinc Bisenolate and Polyenolate (ate) Complexes
Domino
Aldol
R
eactio
i
nTrigg
c
e
red byZin
h
c
Bisenolateael Schmittel,* Manas K. Ghorai
FB 8 (Chemie - Biologie), Organische Chemie I, Universität Siegen, Adolf Reichwein Str., 57068 Siegen, Germany
Fax +49(271)7403270; E-mail: schmittel@chemie.uni-siegen.de
Received 31 August 2001
This paper is dedicated to Prof. Dr. M. Christl on the occasion of his 60th birthday
resultant clear solution was subsequently treated with a
Abstract: A new zinc(II) mediated sequential aldol reaction is de-
scribed for the diastereoselective one-pot synthesis of tetrahydropy-
ran-2,4-diols. The dependence of the yield on the ratio Zn2+/enolate
proposes that zinc(II) polyenolate and ate complexes may be in-
volved.
stoichiometric amount of benzaldehyde (r.t., 2 h) yielding
after aqueous work-up 1a as a single diastereomer besides
2a and 3a (see Table 1 and Table 2).10
Table 1 Reaction of various Aldehydes with Propiophenone Eno-
late in the Presence of ZnBr2
Key words: aldol reactions, domino reactions, zinc, stereoselective
synthesis, hemiacetal
O
0.5 eq. ZnBr2
[THF]
RCHO
LDA
[THF]
Me
[THF]
Ph
Aldolases are attractive enzymes for synthetic chemistry
as they catalyze a variety of C–C bond formation and
cleavage reactions, most often with exquisite stereochem-
ical control. Two categories are known: class I aldolases
use Schiff base formation with an active-site lysine
whereas class II enzymes require a divalent metal ion, in
particular zinc. While both protocols have been used fre-
quently for monoaldol formation, the deoxyribose phos-
phate aldolase (DERA, class I aldolase, see Equation)1
and its combination with RAMA2 are able to even cata-
lyze a sequential aldol-aldol addition by reacting two
equivalents of an enolizable carbonyl compound with an
aldehyde as electrophile to afford tetrahydropyran-2,4-di-
ols that are a part of various important natural products.3
48 °C, 2 h
−40 °C, 1 h
−40 °C, 30 min
25 °C, 1 h
OH
OH
O
O
OH
Ph
O
R
+
+
Ph
R
Ph
R
Me
Me
Me
Me
Ph
3a-i
1a-i
2a-i
Aldehyde RCHO
R
Yielda of Yielda of
1a–ib [%] 1a–ic [%]
a
benzaldehyde
H5C6-
50
75
62
b
dimethylaminobenzal- 4-Me2NC6H4- 37
dehyde
c
d
e
f
methoxybenzaldehyde 4-MeOC6H4-
44
45
42
69
82
R
O
OH
OH OH
O
DERA
fluorobenzaldehyde
furfural
4-FC6H4-
furfuryl-
RCHO
R
2 H3CCHO
OH
cinnamaldehyde
C6H5CH=CH- 10
Equation
g
anthracenecarbalde-
hyde
9-C10H9-
40
It is well known that in class II aldolases4 a zinc ion plays
a pivotal role to activate the deprotonation of the carbonyl
components and to bind the enol.5 This simple picture and
our recent success with titanium bisenolates6 suggested to
treat in a biomimetic fashion two equivalents of an enolate
with an appropriate electrophile in the presence of Zn2+ as
Lewis acid, despite the fact that zinc(II) has a poor history
in aldol reactions,7,8 and that the naked Zn2+ ion makes a
poor model for zinc aldolases.9
h
i
nitrobenzaldehyde
4-NO2C6H4-
20
10
46
iso-butyraldehyde
i-Pr
a Yield is based on aldehyde.
b Using one equivalent of 4.
c Using two equivalents of 4.
According to NMR- and X-ray investigations11 all large
substituents in 1a occupy equatorial positions with only
the hydroxyl groups being placed in axial positions.12
Hence, in this reaction 5 new stereogenic centers are gen-
erated in a highly diastereoselective manner giving rise to
only 1 out of 16 possible stereoisomers.
To test our idea we prepared zinc bisenolate 4 from pro-
piophenone by deprotonation with LDA in THF and sub-
sequent reaction with 0.5 equivalents of ZnBr2. The
Synlett 2001, No. 12, 30 11 2001. Article Identifier:
1437-2096,E;2001,0,12,1992,1994,ftx,en;G19101ST.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0936-5214

LETTER
Domino Aldol Reaction Triggered by Zinc Bisenolate
1993
Table 2 Varying the Ratio of Propiophenone Enolate to ZnBr2 in
Presence of Benzaldehyde at 48 °C
To improve the yields of the domino aldol product 1a the
reaction was studied at various temperatures, i.e. 0 °C,
25 °C, 48 °C and 67 °C. While refluxing the reaction mix-
ture at 67 °C led exclusively to the formation of monoal-
dol condensation product 3a, addition of benzaldehyde at
room temperature followed by heating of the reaction
mixture at 48 °C for 2 hours provided the highest yield of
1a (50%, see Table 1). Hence, a number of additional al-
dehydes was studied using the optimized conditions
(Table 1).
Entry PhCHO Propio- ZnBr2
1a
2a
3a
[equiv]
phenone [equiv]
enolate
[%]
[%]
[%]
[equiv]
1
2
3
4
5
6
7
1
1
1
1
1
1
1
4
4
3
2
2
2
2
2
75
36
64
50
32
2
14
35
15
31
43
36
23
1
10
5
1
1
A straightforward mechanistic proposal for tetrahydropy-
ran-2,4-diol formation is depicted in Scheme 1. At first, a
zinc bisenolate 4 is formed that undergoes an aldol reac-
tion with the aldehyde. The resultant zinc aldolate 5 sub-
sequently is attacked intramolecularly by the second
enolate. Finally, zinc bound tetrahydropyran-2,4-diol 7 is
afforded by intramolecular hemiacetal formation. Under
thermodynamic control all large substituents (methyl,
phenyl, R) are placed in the equatorial position.
1
7
0.75
0.5
0.25
9
19
7
–
equilibrium is shifting from the domino product to the
monoaldol products 2a and 3a.
Ph
The combined preparative results in Table 2 are indeed in
line with a dimeric species having only one zinc biseno-
late reacting to 1. One could argue that yields 50% at an
enolate to Zn2+ ratio of 2:1 (Table 2, entry 4) are due to in-
complete enolate formation under our conditions, in par-
ticular since doubling the amount of zinc bisenolate leads
to a sizeable increase in the yield of 1a (75%, see Table 2,
entry 1). However, three arguments can be raised against
such a hypothesis: (i) Enolate formation from propiophe-
none proved to be quantitative in trapping reactions with
trimethylsilylchloride. (ii) Doubling the amount of eno-
late while keeping the amount of Zn2+ constant led to a de-
crease in yield (36% of 1a, see Table 2, entry 2). (iii)
Keeping the enolate to Zn2+ ratio at 4:1 but now decreas-
ing its overall amount by 50% (see Table 2, entry 6) en-
tailed a sharp drop of the yield to 2%. Hence, there is no
linear correlation of yield of 1a and the amount of enolate.
O
Zn
O
1. aldol
addition
+
RCHO
Zn
O
R
O
Ph
2
Ph
4
5
Zn
O
hemiacetal
formation
O
O
Ph
2. aldol
addition
Ph
R
6
Zn
O
O
OH
OH
hydrolysis
Ph
O
R
Ph
O
R
Ph
Ph
7
1
Scheme 1
Strong evidence for dimeric species emerges from the re-
sults of entries 2–4 in Table 2 as the yield changes from
50% (aldehyde:enolate:Zn2+ = 1:2:1) to 64% (1:3:1) and
to 36% (1:4:1). Such a trend is most readily understood
with structural changes of the reactive zinc(II) enolate
species; i.e. a switch from a dimeric zinc bisenolate 8 and/
or 9 (enolate:Zn2+ = 2:1) to a monomeric ate complex 10
(3:1) and finally to a zinc(II) tetrakisenolate dianion 11
(4:1) (Scheme 2). The latter species is expected to show a
reduced reactivity for the sequential aldol reaction as co-
ordination of the monoaldolate to the zinc(II) to activate
the carbonyl for the second aldol reaction is certainly im-
peded.
However, following the mechanism proposed in
Scheme 1 it is astounding that only yields of 50% (with
1 equiv of ZnBr2/2 equiv of enolate) are obtained, which
led us to consider the structure of the putative zinc biseno-
late in solution. From solid state investigations on related
zinc alkoxides13 and zinc enolate14 one has to assume that,
in solution, dimeric species should prevail with THF act-
ing as additional ligand.
To investigate the role of zinc enolate aggregation we var-
ied the amount of propiophenone enolate with regard to
zinc(II) to 3:1 which increased the yield of 1a (64%;
Table 2, entry 3). Further increase of the propiophenone
enolate/zinc(II) ratio to 4:1 (entry 2), however, ensued in
a sharp decline in the yield. Moreover, the yield of 1a de-
cayed rapidly (entries 5–7) when the molar amount of al-
dehyde exceeded the amount of zinc(II). While, at first,
this finding indicates that a catalytic route cannot be real-
ized, it also makes clear that with excess aldehyde the
Overall, the mechanistic investigation shows that the yield
of 1 is coupled in a complex manner to the amount of eno-
late applied. Yields can be increased beyond those in the
standard protocol (Zn2+:enolate:aldehyde = 1:2:1) in two
ways, both of which are in accordance with dimeric zinc
species in solution: (i) Use of an enolate:Zn2+ ratio of 3:1
to operate via ate complexes, or (ii) doubling the amount
Synlett 2001, No. 12, 1992–1994 ISSN 0936-5214 © Thieme Stuttgart · New York

1994
M. Schmittel, M. K. Ghorai
LETTER
(6) Schmittel, M.; Ghorai, M. K.; Haeuseler, A.; Henn, W.; Koy,
T.; Söllner, R. Eur. J. Org. Chem. 1999, 2007.
(7) (a) Kennedy, G.; Rossi, T.; Tamburini, B. Tetrahedron Lett.
1996, 37, 7441. (b) Lee, H. K.; Kim, J.; Pak, C. S.
Ph
Ph
O
O
O
O
O
O
O
O
O
O
Zn
Zn
Zn
Ph
Zn
Ph
Ph
Ph
O
Tetrahedron Lett. 1999, 40, 6267.
O
(8) Some recent developments are nevertheless very promising:
(a) Trost, B. M.; Ito, H. J. Am. Chem. Soc. 2000, 122, 12003.
(b) Trost, B. M.; Ito, H.; Silcoff, E. R. J. Am. Chem. Soc.
2001, 123, 3367.
Ph
Ph
9
8
(9) Ruf, M.; Weis, K.; Brasack, I.; Vahrenkamp, H. Inorg.
Chim. Acta 1996, 250, 271.
2
Ph
Ph
(10) General Procedure: A solution of diisopropylamine (1.26
mL, 9.0 mmol) in THF (30 mL) was treated at 0 °C with a
solution of n-butyllithium (3.0 mL, 2.5 M in hexane, 7.5
mmol) and stirred for 15 min. After cooling to –40 °C
propiophenone (1.01 mL, 7.50 mmol) was added and the
mixture was stirred at –40 °C for 1 h. Then, the appropriate
amount of zinc dibromide (usually 3.75 mmol) was added.
The yellow reaction mixture was stirred for 30 min at –40 °C
and for 1 h at r.t. Now, it was treated with a solution of the
aldehyde (3.75 mmol) in THF (30 mL) and the temperature
was raised to 48 °C (see Table 1), where it was stirred for 2
h. The reaction was quenched with sat. aq NaHCO3 (50 mL).
The layers were separated and the aq layer was extracted
three times with diethylether. The combined organic layers
were washed with sat. aq NaCl and dried with Na2SO4.
(11) The structural identification of the substrates is based on
extensive NMR investigations (C,H- and H,H-COSY as well
as NOESY), which will be discussed in more detail in the
full paper. A X-ray analysis solved recently is in agreement
with our assignment, s. Engelen, B.; Panthöfer, M. personal
communication.
O
O
O
O
Zn
Zn
Ph
O
Ph
O
O
O
Ph
Ph
Ph
10
11
Scheme 2
of zinc bisenolate (see Table 1, right column) to make up
for one unreactive bisenolate caught up in the dimeric Zn2
species.
With regard to synthetic applications, the present ap-
proach is a good alternative to the DERA/RAMA1,2 cata-
lyzed reaction that has only a limited substrate tolerance.
Although not catalytic in nature, various aromatic and al-
iphatic aldehydes, even containing strongly coordinating
substituents and bulky groups, can be treated with zinc(II)
bisenolate to afford the highly substituted tetrahydropyr-
an-2,4-diols as one single diastereomer.
(12) 1c: 1H NMR (200 MHz, CDCl3): = 0.45 (d, 3J = 7.0 Hz,
3 H), 0.60 (d, 3J = 7.1 Hz, 3 H), 2.34 (dq, 3J = 10.8 Hz,
3J = 7.1 Hz, 1 H), 2.36 (qd, 3J = 7.0 Hz, 4J = 1.3 Hz, 1 H),
3.81 (s, 3 H), 3.90 (s, 1 H), 3.95 (s, 1 H), 5.02 (d,
In conclusion, we have developed a new strategy using
classical, nonbiological zinc(II) chemistry to perform a
sequential double aldol reaction leading to heavily substi-
tuted tetrahydropyran-2,4-diols in a highly stereoselective
manner. Control experiments suggest that dimeric zinc(II)
species may play a pivotal role in the process.
3J = 10.8 Hz, 1 H), 6.90 (d 3J = 9.6 Hz, 2 H), 7.19–7.52 (m,
9 H), 7.60–7.82 (m, 3 H); 1d: 1H NMR (200 MHz, CDCl3):
=0.45 (d, 3J = 7.1 Hz, 3 H), 0.60 (d, 3J = 7.1 Hz, 3 H), 2.35
(dq, 3J = 10.1 Hz, 3J = 7.1 Hz, 1 H), 2.36 (qd, 3J = 7.1 Hz,
4J = 1.1 Hz, 1 H), 3.90 (s, 1 H), 3.99 (s, 1 H), 5.06 (d,
3J = 10.1 Hz, 1H), 7.08 (m, 2 H), 7.19–7.52 (m, 9 H), 7.61–
7.70 (m, 3 H); 1e: 1H NMR (200 MHz, CDCl3): = 0.49 (d,
3J = 6.7 Hz, 3 H), 0.56 (d, 3J = 7.0 Hz, 3 H), 2.32 (q,
3J = 7.0 Hz, 1 H), 2.70 (dq, 3J = 11.0 Hz, 3J = 6.7 Hz, 1 H),
3.82 (s, 1 H), 4.02 (s, 1 H), 5.14 (d, 3J = 11.0 Hz 1 H), 6.35
(m, 2 H), 7.23–7.43 (m, 8 H), 7.65 (m, 3 H); 1f: 1H NMR
(200 MHz, CDCl3): = 0.45 (d, 3J = 7.1 Hz, 3 H), 0.58 (d,
3J = 6.9 Hz, 3 H), 2.35 (dq, 3J = 10.4 Hz, 3J = 7.1 Hz 1 H),
2.36 (qd, 3J = 6.9 Hz, 4J = 1.3 Hz 1 H), 3.66 (s, 1 H), 3.81 (s,
1 H), 4.61 (dd, 3J = 10.4 Hz, 3J = 7.1 Hz 1 H), 6.23 (dd,
3J = 15.8 Hz, 3J = 7.1 Hz), 6.63 (d, 3J = 15.8 Hz 1 H), 7.12–
7.32 (m, 12 H), 7.53–7.60 (m, 3 H); 1g: 1H NMR (200 MHz,
CDCl3): = 0.32 (d, 3J = 6.8 Hz, 3 H), 0.72 (d, 3J = 7.2 Hz,
3 H), 2.67 (dq, 3J = 10.8 Hz, 3J = 6.8 Hz 1 H), 3.65 (qd,
3J = 7.2 Hz, 4J = 1.3 Hz 1 H), 4.13 (d, 4J = 1.7 Hz 1 H), 4.27
(s, 1 H), 6.9 (d, 3J = 10.8 Hz 1 H), 7.36–7.41 (m, 8 H) 7.42–
7.58 (m, 2 H), 7.67–7.80 (m, 4 H), 7.97–8.07 (m, 2 H), 8.44
(s, 1 H), 8.51 (d, 3J = 8.6 Hz, 1 H), 9.41 (d, 3J = 8.8 Hz, 1 H).
(13) (a) Olmstead, M. M.; Power, P. P.; Shoner, S. C. J. Am.
Chem. Soc. 1991, 113, 3379. (b) Parvez, M.; Bergstresser,
G. L.; Richey, H. G. Jr. Acta Crystallogr. 1992, C48, 641.
(c) Kitamura, M.; Yamakawa, M.; Oka, H.; Suga, S.;
Noyori, R. Chem.-Eur. J. 1996, 2, 1173. (d) Tesmer, M.;
Müller, B.; Vahrenkamp, H. Chem. Commun. 1997, 721.
(14) van Vliet, M. R. P.; van Koten, G.; Buysingh, P.; Jastrzebski,
J. T. B. H.; Spek, A. L. Organometallics 1987, 6, 537.
Acknowledgement
We are indebted to the DFG and the Fonds der Chemischen Indu-
strie for support. M. K. G. thanks the Alexander von Humboldt Stif-
tung, Germany, for the award of a research fellowship.
References
(1) (a) Gijsen, H. J. M.; Wong, C.-H. J. Am. Chem. Soc. 1994,
116, 8422. (b) Gijsen, H. J. M.; Wong, C.-H. J. Am. Chem.
Soc. 1995, 117, 7585.
(2) Gijsen, H. J. M.; Wong, C.-H. J. Am. Chem. Soc. 1995, 117,
2947.
(3) (a) Hochlowski, J. E.; Faulkner, D. J.; Matsumoto, G. K.;
Clardy, J. J. Am. Chem. Soc. 1983, 105, 7413. (b) Ziegler,
F. E.; Becker, M. R. J. Org. Chem. 1990, 55, 2800.
(4) Kimura, E.; Gotoh, T.; Koike, T.; Shiro, M. J. Am. Chem.
Soc. 1999, 121, 1267.
(5) (a) Wong, C.-H.; Whitesides, G. M. Enzymes in Organic
Chemistry; Fessner, W. D., Ed.; Pergamon: Oxford, 1994,
Chap. 4. (b) Walter, C. Top. Curr. Chem. 1997, 184, 97.
(c) Petersen, M.; Zannetti, M. T.; Fessner, W. D. Top. Curr.
Chem. 1997, 186, 97. (d) Fessner, W. D. Curr. Opin. Chem.
Biol. 1998, 2, 85.
Synlett 2001, No. 12, 1992–1994 ISSN 0936-5214 © Thieme Stuttgart · New York