Diastereoselective Darzens condensations

Article abstract of DOI:10.1016/j.tetlet.2007.03.010

N,N-Diphenyl-α-haloacetamides undergo Darzens condensations with aldehydes under heterogeneous reaction conditions in the presence of a metal hydroxide base. By appropriate choice of solvent, base and halide, very high diastereoselectivities favouring for

Full text of DOI:10.1016/j.tetlet.2007.03.010

Tetrahedron Letters 48 (2007) 2961–2964
Diastereoselective Darzens condensations
Thierry J. R. Achard,a Yuri N. Belokon,b Jamie Hunt,a
Michael Northa,* and Francesca Pizzatoa
aSchool of Natural Sciences, Bedson Building, Newcastle University, Newcastle Upon Tyne, NE1 7RU, UK
bA.N. Nesmeyanov Institute of Organo-Element Compounds, Russian Academy of Sciences, 119991 Moscow,
Vavilov 28, Russian Federation
Received 16 January 2007; revised 12 February 2007; accepted 1 March 2007
Available online 4 March 2007
Abstract—N,N-Diphenyl-a-haloacetamides undergo Darzens condensations with aldehydes under heterogeneous reaction condi-
tions in the presence of a metal hydroxide base. By appropriate choice of solvent, base and halide, very high diastereoselectivities
favouring formation of the cis- or trans-epoxide can be obtained.
Ó 2007 Elsevier Ltd. All rights reserved.
Epoxides are particularly versatile synthetic intermedi-
ates which can readily be converted into a wide range
of polyfunctional compounds. A useful method for the
synthesis of a,b-epoxy-carbonyl compounds and related
compounds is the Darzens condensation between a car-
bonyl compound and an a-halo-carbonyl compound (or
related species).1 This reaction has been known for over
100 years, but only recently has progress been made in
controlling the relative and absolute stereochemistry of
the reaction. Thus, Arai et al. have reported that the dia-
stereoselectivity obtained using N,N-diphenyl a-chloro-
acetamide 1a and 4-tert-butylbenzaldehyde in the
presence of tetrahexylammonium bromide was base
dependent, giving largely cis-epoxide 2a when potassium
hydroxide was used as base, but exclusively the trans-
epoxide 2b when lithium hydroxide was used as base.2
O
O
CONPh2
CONPh2
Ar
Ar
O
O
2a
2b
(i)
+
X
NPh2
Ar
H
1a: X = Cl
1b: X = Br
1c: X = I
Scheme 1. Reagents and conditions: (i) MOH (solid), solvent, 4–24 h,
rt.
pounds and undergo regioselective ring-opening reac-
tions.5,6 In this Letter, we show that the nature of the
halide, base and solvent are all important factors and that
by choice of appropriate conditions, either cis- or trans-
epoxides can be obtained with high diastereoselectivity.
As a prelude to a project aimed at developing metal(sa-
len) complexes as asymmetric catalysts for the Darzens
condensation, we investigated in more detail the factors
that determine the diastereoselectivity of the Darzens
condensation of substrates 1a–c with carbonyl com-
pounds (Scheme 1). Substrates 1a–c were chosen3,4 due
to the literature precedent for the use of substrate 1a in
diastereoselective Darzens condensations,2 use of sub-
strates 1a–b in enantioselective Darzens condensations5
and the known versatility of epoxy amides which can
readily be converted into other epoxy-carbonyl com-
An initial study was carried out using benzaldehyde as
the carbonyl compound and dichloromethane as solvent
with no added phase transfer catalyst. The results of this
study using substrates 1a–c and powdered sodium,
potassium or rubidium hydroxide as base are reported
in Table 1. Other bases (lithium hydroxide, potassium
bicarbonate, sodium acetate, potassium tert-butoxide)
gave no significant reaction under these conditions.
a-Chloroamide 1a underwent Darzens condensations in
the presence of sodium or potassium hydroxide (Table
1, entries 1 and 2) to give selectively trans-epoxide 2b,
with sodium hydroxide giving a particularly good dia-
stereoselectivity (Table 1, entry 1). Rubidium hydroxide
*
Corresponding author. Tel.: +44 191 222 7128; fax: +44 870 131
3783; e-mail: michael.north@ncl.ac.uk
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.03.010

2962
T. J. R. Achard et al. / Tetrahedron Letters 48 (2007) 2961–2964
Table 1. Darzens condensation between a-haloamides 1a–c and
benzaldehyde in dichloromethane
O
O
O
+
M
ClCH2CONPh2 / MOH
Ar
thermodynamically
controlled aldol
Entry Haloamide Base
Time (h) cis/trans Yield (%)
Cl
Ar
H
NPh2
ratioa
H
Br(I)
reaction
1
2
3
4
5
6
7
8
9
1a
1a
1a
1b
1b
1b
1c
1c
1c
NaOH
KOH
RbOH
NaOH 24
KOH
4
4
4
1:16.0
1:1.8
1.4:1
1.5:1
2.4:1
2.0:1
1.2:1
2.3:1
2.2:1
68
73
71
73
78
76
51
62
80
Ar
Cl
Ph2N
OM
kinetically controlled
aldol reaction
MO
MO
CONPh2
CONPh2
4
4
RbOH
Ar
Br(I)
MO
Ar
CONPh2
NaOH 24
KOH 24
RbOH 24
MO
CONPh2
Br(I)
fast
Ar
Cl
a Determined by 1H NMR analysis of the crude mixture of epoxides 2a
and 2b.
slow
O
O
Ar
CONPh2 Ar
CONPh2
2a
2b
also induced the Darzens condensation of substrate 1a,
but gave cis-epoxide 2a as the major product, though
with low diastereoselectivity (Table 1, entry 3). All of
these reactions were complete in 4 h. In contrast, sub-
strates 1b and 1c gave cis-epoxide 2a as the major prod-
uct with all three bases (Table 1, entries 4–9). The best
diastereoselectivity with substrate 1b was 2.4:1, obtained
using potassium hydroxide as base (Table 1, entry 5).
Almost identical diastereoselectivities could be obtained
using a-iodoamide 1c and either potassium or rubidium
hydroxide as base (Table 1, entries 8 and 9). In the latter
case, a control experiment showed that the diastereo-
selectivity was not time dependent as a 2.3:1 ratio of
2a:2b was observed after a reaction time of 4 h.
Scheme 2.
controlled aldol reaction then occurs through a chelated
transition state to give the syn-aldol product which on
cyclization forms the trans-epoxide as shown in Scheme
2. The effect of the metal counter-ion is also explained
on this basis. Thus, for reactions involving a-chloro-
amide 1a, a sodium ion will give a tighter chelated
transition state, resulting in a higher diastereoselectivity.
In contrast, for reactions involving a-bromoamide 1b or
a-iodoamide 1c, the greater degree of ionic character in
the potassium or rubidium oxygen bond will increase the
tendency for the enolate to be electrostatically repelled
by the polarized aldehyde bond, thus favouring the
kinetic aldol product which leads to the cis-epoxide.
The above results suggested that under these reaction
conditions, the initial deprotonation of a-haloamide
1a–c was the rate limiting step of the reaction. Thus,
as the acidity of the substrate decreased from 1a to 1b
to 1c, the chemical yield decreased and/or the reaction
time increased; whilst increasing the strength of the base
from lithium through sodium and potassium to rubid-
ium hydroxide generally increased the chemical yield
and/or reduced the reaction time.
On the basis of the results shown in Table 1 and the
mechanistic analysis shown in Scheme 2, the use of
sodium hydroxide with a-chloroamide 1a was optimal
for the synthesis of trans-epoxide 2b (Table 1, entry 1)
whilst the use of potassium hydroxide with a-bromo-
amide 1b was optimal for the synthesis of cis-epoxide
2a (Table 1, entry 5). These combinations thus provided
the starting point for a subsequent study to determine
the most appropriate solvent for the reaction. The
results of this investigation are presented in Table 2.
The inversion of diastereoselectivity observed with so-
dium or potassium hydroxide on changing the leaving
group from chloride to bromide or iodide (Table 1, com-
pare entries 1 and 2 with entries 4, 5, 7 and 8) appears to
be unprecedented in previous work on the Darzens con-
densation. The results are however consistent with the
established mechanism for the Darzens condensation,
which involves an aldol reaction followed by an intra-
molecular substitution reaction.1 Depending upon the
substrate and reaction conditions, the initial aldol reac-
tion may be reversible or irreversible. For reactions
involving a-bromoamide 1b and a-iodoamide 1c, the
results are consistent with a kinetically controlled, irre-
versible aldol reaction which occurs through a non-che-
lated transition state to give the anti-aldol intermediate
which undergoes a rapid intramolecular cyclization to
produce the cis-epoxide as shown in Scheme 2. In con-
trast, for reactions involving a-chloroamide 1a, the
intramolecular cyclization step is slower than the retro-
aldol reaction and this allows the aldol reaction to
equilibrate prior to cyclization. The thermodynamically
Reactions carried out in non-polar solvents such as tol-
uene, ether or THF7 usually failed to go to completion
and gave a mixture of the desired epoxides 2a,b and
the intermediate aldol products (Table 2, entries 1–6).
These reactions also tended to exhibit low diastereose-
lectivities, though the use of a-chloroamide 1a with
sodium hydroxide in toluene was exceptional in this
respect (Table 2, entry 1). Methanol was a satisfactory
solvent for reactions involving a-chloroamide 1a, but
not for those using a-bromoamide 1b, due to a compet-
ing methanolysis of the bromide (Table 2, entries 7 and
8).
However, the use of polar aprotic solvents gave more
positive outcomes. In both acetonitrile8 and DMF, all
of the reactions favoured the formation of cis-epoxide
2a (Table 2, entries 9–26). Thus, whereas the combina-

T. J. R. Achard et al. / Tetrahedron Letters 48 (2007) 2961–2964
2963
Table 2. Darzens condensation between a-haloamides 1a,b and
benzaldehyde in various solvents
ratio of 6:1 (Table 2, entry 9) and in DMF, exclusively
cis-epoxide 2a was obtained (Table 2, entry 18). Use of
other bases with substrate 1a in acetonitrile or DMF
also resulted in the selective formation of cis-epoxide
2a, though with lower diastereoselectivity (Table 2, en-
tries 12, 15, 21 and 24). With substrate 1b, all three bases
showed enhanced diastereoselectivity for the formation
of cis-epoxide 2a in both acetonitrile and DMF com-
pared to reactions carried out in dichloromethane (com-
pare Table 1, entries 4–6 with Table 2, entries 10, 13, 16,
19, 22 and 25). However, the use of potassium hydroxide
in either acetonitrile or DMF gave exceptional diastereo-
selectivities in favour of cis-epoxide 2a (Table 2, entries
13 and 22). Reactions involving a-iodoamide 1c were
also cis-selective in acetonitrile and DMF (Table 2, en-
tries 11, 14, 17, 20, 23 and 26) and gave exclusively
cis-epoxide 2a when rubidium hydroxide was used as
base in acetonitrile (Table 2, entry 17) or potassium
hydroxide was used as base in DMF (Table 2, entry 23).
Entry Haloamide Base
Solvent cis/trans
Conversionb
(%)
ratioa
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1a
1b
1a
1b
1a
1b
1a
1b
1a
1b
1c
1a
1b
1c
1a
1b
1c
1a
1b
1c
1a
1b
1c
1a
1b
1c
NaOH Toluene Only trans 71c
KOH
NaOH Et2O
KOH Et2O
NaOH THF
KOH THF
NaOH MeOH
KOH MeOH
Toluene 1.6:1
97
1:2.1
1.4:1
81c
100
0c
92c
98
0
1:6.4
2:1
NaOH MeCN
NaOH MeCN
NaOH MeCN
6.0:1
7.3:1
8.8:1
3.2:1
22.8:1
5.0:1
5.2:1
100
100
100
100
100
100
100
100
98
KOH
KOH
KOH
MeCN
MeCN
MeCN
RbOH MeCN
RbOH MeCN
RbOH MeCN
NaOH DMF
NaOH DMF
NaOH DMF
6.3:1
Only cis
Only cis
2.6:1
100
100
83
In polar aprotic solvents, it is likely that the initial aldol
reaction always occurs through a non-chelated transi-
tion state which accounts for the cis-selectivity that is
observed in all of these reactions as shown in Scheme
2. In many cases, the diastereoselectivity increased from
substrate 1a through 1b to 1c (Table 2, entries 9–11, 15–
17 and 21–23), consistent with the better leaving group
increasing the rate of the intramolecular substitution
reaction and hence decreasing the reversibility of the al-
dol reaction. However, there are also exceptions to this
rule where specific combinations of base and substrate
gave high diastereoselectivities.
1.8:1
KOH
KOH
KOH
RbOH DMF
RbOH DMF
RbOH DMF
DMF
DMF
DMF
3.7:1
100
100
69
100
100
61
Only cis
Only cis
3.5:1
16.6:1
4.4:1
a Determined by 1H NMR analysis of the crude mixture of epoxides 2a
and 2b.
b Reactions involving NaOH took 24 h at room temperature, reactions
involving KOH or RbOH were complete after 4 h at room temper-
ature except for entry 26 which required 24 h.
c The intermediate aldol product was also detected in these reactions.
On the basis of these results, it can be concluded that
highly diastereoselective syntheses of cis-epoxide 2a
can be achieved with each of substrates 1a–c by choice
of appropriate base and solvent. However, the use of
substrate 1b and potassium hydroxide in acetonitrile
(Table 2, entry 13) was particularly experimentally
tion of substrate 1a and sodium hydroxide had given a
cis/trans ratio of 1:16 in dichloromethane (Table 1, entry
1), the same combination in acetonitrile gave a cis/trans
Table 3. Darzens condensation between a-haloamides 1a,b and various carbonyl compounds
Entry
Haloamide
Aldehyde
Base
cis/trans ratioa
Yieldb (%)
81
1
2
3
4
5
6
7
8
9
10
11c
12
13
14
15
1a
1b
1a
1b
1a
1b
1a
1b
1a
1b
1b
1a
1b
1a
1b
2-MeC6H4CHO
2-MeC6H4CHO
3-MeC6H4CHO
3-MeC6H4CHO
4-MeC6H4CHO
4-MeC6H4CHO
4-MeOC6H4CHO
4-MeOC6H4CHO
4-O2NC6H4CHO
4-O2NC6H4CHO
4-O2NC6H4CHO
Me3CCHO
NaOH
KOH
NaOH
KOH
NaOH
KOH
NaOH
KOH
NaOH
KOH
KOH
NaOH
KOH
NaOH
KOH
1:28.9
12.5:1
1:1.4
2.7:1
1:3.0
21.1:1
1:5.4
Only cis
1:4.8
1:1
1.7:1
100
84
100
72
100
66
93
78
100
72
Only trans
4.9:1
1:2.7
99d
Me3CCHO
PhCOMe
PhCOMe
99d
93d
0d
a Determined by 1H NMR analysis of the crude mixture of epoxides 2a and 2b.
b All reactions involving substrate 1a were carried out in dichloromethane, whilst reactions involving substrate 1b were conducted in acetonitrile.
Reaction time 24 h when NaOH was used as base and 4 h when KOH was used as base except where stated. All reactions were carried out at room
temperature.
c Reaction carried out in dichloromethane.
d Reaction time 3 days.

2964
T. J. R. Achard et al. / Tetrahedron Letters 48 (2007) 2961–2964
convenient. In contrast, the optimal yield and diastereo-
selectivity in the synthesis of trans-epoxide 2b (1:16.0)
was obtained using chloroamide 1a in dichloromethane
(Table 1, entry 1). These two sets of conditions were
then utilized with a range of carbonyl compounds to
demonstrate the generality of the process (Table 3).
In conclusion, we have shown that the diastereoelectiv-
ity of the Darzens condensation between a-haloamides
1a–c and aldehydes is significantly influenced by both
the base and the solvent. Reactions involving a-chloro-
amide 1a and sodium hydroxide in dichloromethane
usually exhibit good to excellent trans-selectivity, whilst
use of a-bromoamide 1b and potassium hydroxide in
acetonitrile gives high cis-selectivity.
The Darzens condensation of chloroamide 1a with var-
ious aromatic aldehydes induced by sodium hydroxide
was always found to be diastereoselective in favour of
trans-epoxide 2b (Table 3, entries 1, 3, 5, 7 and 9). 2-
Methylbenzaldehyde was a particularly good substrate
for this reaction (Table 3, entry 1), giving a 1:28.9 ratio
of epoxides 2a and 2b in high yield. The high diastereo-
selectivity observed in this case is probably due to steric
effects as 3-methylbenzaldehyde exhibited very little dia-
stereoselectivity (Table 3, entry 3) and 4-methylbenz-
aldehyde gave epoxides 2a and 2b in only a 1:3 ratio
(Table 3, entry 5). Both the electron rich 4-methoxy-
benzaldehyde (Table 3, entry 7) and the electron
deficient 4-nitrobenzaldehyde (Table 3, entry 9) gave
around a 1:5 ratio of epoxides 2a and 2b suggesting that
the electronic nature of the aldehyde had little influence
in this case. These results confirm that a diastereoselec-
tive synthesis of trans-epoxides 2b can be achieved by
use of solid sodium hydroxide in dichloromethane.
Acknowledgements
The authors thank the EU Descartes prize research fund
for a studentship (to T.J.R.A.). Mass spectra were
recorded by the EPSRC national mass spectrometry
service at the University of Wales, Swansea, and use of
the EPSRC’s Chemical Database Service at Daresbury
is also gratefully acknowledged.
References and notes
1. For a review of the Darzens reaction see: Rosen, T. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I.,
Eds.; Pergamon: Oxford, 1991; Vol. 2, pp 409–439.
2. Arai, S.; Suzuki, Y.; Tokmaru, K.; Shioiri, T. Tetrahedron
Lett. 2002, 43, 833–836.
3. Substrates 1a and 1b were prepared by literature proce-
dures: Benkli, K.; Demirayak, S.; Gundogdu-Karaburun,
N.; Kiraz, N.; Iscan, G. Indian J. Chem., Sect. B 2004, 174–
179; Aquino, C. J.; Armour, D. R.; Berman, J. M.;
Birkemo, L. S.; Carr, R. A. E.; Croom, D. K.; Dezube,
M., ; Dougherty, R. W., Jr.; Ervin, G. N.; Grizzle, M. K.;
Head, J. E.; Hirst, G. C.; James, M. K.; Johnson, M. F.;
Miller, L. J.; Queen, K. L.; Rimele, T. J.; Smith, D. N.;
Sugg, E. E. J. Med. Chem. 1996, 39, 562–569.
4. Synthesis of 1c: Solutions of 1a (1.5 g, 6.1 mmol) in acetone
(16 ml) and sodium iodide (1.0 g, 6.5 mmol) in acetone
(8.5 ml) were mixed and heated under reflux for 10 min.
The solution was allowed to cool to room temperature and
filtered through Celite. The filtrate was again heated under
reflux for 10 min, cooled to room temperature and filtered
through Celite. The solvent was removed in vacuo and the
yellow residue was crystallized from CH2Cl2 to give
compound 1c (1.45 g, 70%) as a yellow solid. mp 116–
119 °C; dH (300 MHz, CDCl3) 3.71 (2H, s), 7.3–7.5 (10H,
br); dC (75 MHz, CDCl3) À1.2 (CH2), 127.5 (ArC), 129.5
(ArCH), 129.6 (ArCH), 142.7 (ArCH), 168.0 (C@O); m/z
(nanospray): Found 338.0036; C14H13NOI (MH+) requires
338.0036.
For Darzens reactions involving a-bromoamide 1b, the
use of potassium hydroxide as base generally resulted in
preferential formation of cis-amide 2a (Table 3, entries
2, 4, 6, 8, 10 and 11). Both 2-methyl and 4-methylbenz-
aldehydes gave epoxide 2a with excellent diastereoselec-
tivity (Table 3, entries 2 and 6), whilst the 3-methyl
isomer reacted with much lower diastereoselectivity
(Table 3, entry 4). 3-Methylbenzaldehyde was also the
least diastereoselective substrate with chloroamide 1a
(Table 3, entry 3). Unlike the corresponding reactions
with chloroamide 1a, electronic effects had a pronounced
influence on reactions involving bromoamide 1b. Thus,
4-methoxybenzaldehyde reacted with 1b to give only
cis-epoxide 2a (Table 3, entry 8), whilst under the same
conditions, 4-nitrobenzaldehyde gave a 1:1 mixture of
epoxides 2a and 2b (Table 3, entry 10). Some diastereo-
selectivity in favour of epoxide 2a could be restored in
this case by changing the solvent from acetonitrile to
dichloromethane (Table 3, entry 11).
Aliphatic aldehydes were also investigated as substrates.
Pivaldehyde was a good substrate (Table 3, entries 12
and 13), giving exclusively trans-epoxide with a-chloro-
amide 1a, and a good selectivity in favour of cis-epoxide
with a-bromoamide 1b. However, aliphatic aldehydes
with a-protons (cyclohexane carboxaldehyde and non-
anal) failed to give any epoxide-containing products,
presumably due to competing aldol reactions. Finally,
acetophenone reacted with substrate 1a and sodium
hydroxide as expected to give predominantly trans-
epoxide 2b (Table 3, entry 14), but failed to react with
substrate 1b (Table 3, entry 15), possibly due to compet-
ing hydrolysis of the alkyl bromide.
5. Arai, S.; Tokumaru, K.; Aoyama, T. Tetrahedron Lett.
2004, 45, 1845–1848.
6. Aggarwal, V. K.; Charmant, J. P. H.; Fuentes, D.; Harvey,
J. N.; Hynd, G.; Ohara, D.; Picoul, W.; Robiette, R.;
Smith, C.; Vasse, J.-L.; Winn, C. L. J. Am. Chem. Soc.
2006, 128, 2105–2114.
7. THF has previouly been reported to be the solvent of
choice to obtain cis-epoxides from Darzens condensations:
Wang, Z.; Xu, L.; Mu, Z.; Xia, C.; Wang, H. J. Mol. Catal.
A 2004, 218, 157–160.
8. For a previous report on the use of acetonitrile to obtain
trans-selectivity in Darzens condensations using ethyl a-
chloroacetamide see: Wang, Z.-T.; Xu, L.-W.; Xia, C.-G.;
Wang, H. Q. Helv. Chim. Acta 2004, 87, 1958–1962.