Diastereoselective darzens condensations of α-haloamides: Influence of aryl substituents on diastereoselectivity

Article abstract of DOI:10.2174/157017809789869465

N, N-Diaryl α-haloamides undergo Darzens condensations with aldehydes induced by metal hydroxides. The diastereoselectivity of epoxide formation is strongly influenced by the electronic properties of the arylamide, which can be rationalised on the basis of the acidity of the substrate hydrogens and hence on reaction occurring under kinetic or thermodynamic control.

Full text of DOI:10.2174/157017809789869465

552
Letters in Organic Chemistry, 2009, 6, 552-556
Diastereoselective Darzens Condensations of ꢀ-Haloamides: Influence of
Aryl Substituents on Diastereoselectivity
Michael North* and Francesca Pizzato
School of Chemistry and University Research Centre in Catalysis and Intensified Processing, Bedson Building,
Newcastle University, Newcastle Upon Tyne, NE1 7RU, UK
Received March 05, 2009: Revised July 02, 2009: Accepted July 07, 2009
Abstract: N,N-Diaryl ꢀ-haloamides undergo Darzens condensations with aldehydes induced by metal hydroxides. The
diastereoselectivity of epoxide formation is strongly influenced by the electronic properties of the arylamide, which can be
rationalised on the basis of the acidity of the substrate hydrogens and hence on reaction occurring under kinetic or
thermodynamic control.
Keywords: Darzens condensation, metal hydroxide, aldehyde, proton-acidity, diastereoselectivity, epoxide.
O
INTRODUCTION
The Darzens condensation [1] between an ꢀ-
R
CONPh2
halocarbonyl compound and an aldehyde or ketone is a
potentially useful way of synthesising epoxy-carbonyl
compounds (Scheme 1). Although this reaction has been
known since the 19th century, its use in synthesis has been
limited due to lack of control in the stereoselectivity with
which the two new stereocentres are created: leading to
mixtures of enantiomers and diastereomers. We recently
reported [2, 3] that N,N-diphenylamides (1a-c) undergo
Darzens condensations under mild heterogeneous conditions
in the presence of an aldehyde and solid metal hydroxide
base. The diastereoselectivity of these reactions was
determined by an interplay of the leaving group ability of the
halide, the basicity of the metal hydroxide and the solvent
polarity, allowing either the cis- or trans-epoxide to be
obtained with high diastereoselectivity by choice of
appropriate reaction conditions. Other workers have also
reported the influence of base [4] and solvent [5, 6] on
Darzens condensations.
cis-epoxide formed
via kinetic enolate
k3
X
O
O
M O
R
CONPh2
M
R
X
NPh2
k2
RCHO
k-2
OM
O
k1
X
+ MOH
X
NPh2
k-1
NPh2
H
1a: X = Cl
1b: X = Br
1c: X = I
2a
X = Cl
O
OM
O
O
R
R2
R1
H
+
base
X
NPh2
R
R2
R1
O
Cl
2b
Scheme 1.
k2'
RCHO
k-2'
Our results could be rationalised on the basis of kinetic
versus thermodynamic control in the aldol step of the
reaction mechanism. Thus, the initially formed enolate 2a
will have the Z-configuration for steric reasons (the large
NPh2 group is cis to the hydrogen) and also for electronic
reasons since the metal can chelate to the halide (Scheme 2).
A chelation controlled aldol reaction involving enolate 2a
will place the halide in an axial position in the transition
O
O
M O
R
CONPh2
M
R
X
Cl
NPh2
k3'
O
R
CONPh2
*Address correspondence to this author at the School of Chemistry and
University Research Centre in Catalysis and Intensified Processing, Bedson
Building, Newcastle University, Newcastle Upon Tyne, NE1 7RU, UK; Tel:
+44 (0)191 2227128; Fax: +44 (0)870 1313783;
trans-epoxide formed
via thermodynamic enolate
Scheme 2.
E-mail: Michael.north@ncl.ac.uk
1570-1786/09 $55.00+.00
© 2009 Bentham Science Publishers Ltd.

Diastereoselective Darzens Condensations
Letters in Organic Chemistry, 2009, Vol. 6, No. 7
553
Y
state, leading to the syn-aldol product and hence to the cis-
epoxide. A good leaving group (bromide or iodide) and a
polar solvent (acetonitrile or DMF) will increase the rate of
epoxide formation (k3) relative to the rate of the retro-aldol
reaction (k-2) and so favour this pathway leading to
predominant formation of the cis-epoxide. It is also possible
that the cis-epoxide is formed by a non-chelated transition
state or via a boat transition state to avoid the necessity for
the halide to be in an axial position in the transition state.
Both of these possibilities also lead to preferential formation
of the cis-epoxide. In contrast, use of a poorer leaving group
(chloride) in a less polar solvent (toluene or dichloro-
methane) decreases the rate of cyclisation so that k-2 is
greater than k3. This allows the initial aldol product to
equilibrate with enolate 2a and also allows Z-enolate 2a to
equilibrate with E-enolate 2b. The latter can also undergo a
chelation controlled aldol reaction leading to the anti-aldol
product and thus to the trans-epoxide. Since in this process,
the chelated transition state has all of the substituents
equatorial, it is thermodynamically preferred and hence the
trans-epoxide will predominate whenever the rate of aldol
equilibration is greater than the rate of cyclisation (k-2 and k2’
greater than k3’).
O
a: X = Cl; b: X = Br
1a,b: Y = Z = H
X
3a,b: Y = Z = NO2
4a,b: Y = H, Z = NO2
5a,b: Y = H, Z = OMe
6a,b: Y= Z = OMe
7a,b: Y = OMe, Z = NO2
N
Z
PhCHO / MOH
CH2Cl2, RT
Y
Y
Ph
O
O
Ph
O
O
N
N
Z
+
Z
8a: Y = Z = H
9a: Y = Z = H
8b: Y = Z = NO2
8c: Y = H, Z = NO2
8d: Y = H, Z = OMe
8e: Y= Z = OMe
9b: Y = Z = NO2
9c: Y = H, Z = NO2
9d: Y = H, Z = OMe
9e: Y= Z = OMe
Based on this analysis, we reasoned that the electronic
nature of the aryl rings in N,N-diarylamides could also have
an influence on the diastereoselectivity of the Darzens
condensations. Electron-donating substituents attached to the
aromatic rings should reduce the acidity of the ꢀ-protons,
thus destabilising formation of enolate 2 (reduce k1 and k-2)
and encouraging the aldol reaction to occur under kinetically
controlled conditions, leading to the cis-epoxide. On the
other hand, electron-withdrawing substituents should
increase the acidity of the ꢀ-protons, thus stabilising enolate
2 and facilitating the retro-aldol reaction (increase k-2) so that
the product is formed via the thermodynamically controlled
chelated enolate, leading to the trans-epoxide. In this Letter
we report the results of this investigation.
8f: Y = OMe, Z = NO2
9f: Y = OMe, Z = NO2
Scheme 3.
Y
Y
O
ClCH2COCl
or
BrCH2COBr
CH2Cl2, Et3N
or neat
X
HN
N
39-98%
Z
Z
10: Y = Z = NO2
a: X = Cl; b: X = Br
3a,b: Y = Z = NO2
4a,b: Y = H, Z = NO2
5a,b: Y = H, Z = OMe
6a,b: Y= Z = OMe
RESULTS AND DISCUSSION
11: Y = H, Z = NO2
12: Y = H, Z = OMe
13: Y= Z = OMe
For this study, amides 3-7a,b were studied relative to
N,N-diphenylamides 1a,b using benzaldehyde as the enolate
acceptor and with a range of solid alkali metal hydroxides
(NaOH, KOH, RbOH and CsOH) as bases in
dichloromethane solvent at room temperature to give
epoxides 8a-f and 9a-f as shown in Scheme 3 [7].
Compounds 4-6 were prepared by reaction of diaryl amines
11-13 with chloroacetyl chloride and bromoacetyl bromide
in dichloromethane in the presence of triethylamine for 16
hours to give 4a-6a and 4b-6b respectively as shown in
Scheme 4 [8]. For compounds 3a,b and 7a,b, this synthesis
was modified by running the reaction in the absence of
solvent and triethylamine in view of the low nucleophilicity
of nitroamines 10 and 14. Amines 10-14 are known
compounds [9-12] which were prepared by an amino acid
promoted Ullmann reaction [13] involving 4-methoxyaniline
(to form compounds 12-13) or an SNAr reaction involving 4-
nitrofluorobenzene (to form diamines 10, 11 and 14) [9].
14: Y = OMe, Z = NO2
7a,b: Y = OMe, Z = NO2
Scheme 4
also the least reactive with reaction times of 65 hours being
required to produce high conversions. This is consistent with
the low acidity of the hydrogens in 6a resulting in a low
value for rate constant k1. Removal of one of the methoxy
groups to give substrate 5a results in a significant increase in
reactivity (reaction times 16 hours), though this is still
significantly less reactive than unsubstituted substrate 1a. It
is also apparent from Table 1 that for any base, the
percentage of trans epoxide 9a-e formed increases as the
substrate becomes more electron-deficient. Indeed, the 4-
nitrophenyl containing substrates (3a and 4a) give
exclusively trans-epoxide 9d,e with sodium and potassium
hydroxide as bases. However, the conversions obtained with
substrate 3a which contains two 4-nitrophenyl groups are
much higher than those obtained with substrate 4a (reaction
Table 1 compares the results obtained with chloroamides
3a-7a with those obtained using N,N-diphenyl chloroamide
1a. As expected, the most electron-rich substrate (6a) was

554 Letters in Organic Chemistry, 2009, Vol. 6, No. 7
North and Pizzato
Table 1. Comparison of Chloroamides 1a and 3a-7a in the Darzens Condensation with Benzaldehydea
Substrate
NaOH
KOH
RbOH
CsOH
6ab
5ac
1ad
4ac
3ac
7ac
100 (55% trans)
96 (71% trans)
68 (94% trans)
49 (100% trans)
95 (100% trans)
54 (77% trans)
64 (55% trans)
81 (52% trans)
73 (64% trans)
48 (100% trans)
90 (100% trans)
65 (48% trans)
80 (40% trans)
0
28 (38% trans)
18 (33% trans)
87 (38% trans)
17 (79% trans)
35 (100% trans)
62 (31% trans)
71 (42% trans)
70 (91% trans)
82 (100% trans)
44 (50% trans)
aIn each cell, the conversion is given first followed in brackets by the percentage of trans-epoxide 9 in the cis / trans mixture prior to purification as determined by 1H NMR
spectroscopy. bReaction time 65 hours. cReaction time 16 hours. dReaction time 4 hours.
time of 16 hours in each case to allow comparison). This is
also consistent with the mechanism shown in Scheme 2 since
both k1 and k-2 will increase as the substrate becomes more
electron-deficient, resulting in higher chemical yields and a
greater amount of reaction through the thermodynamically
controlled aldol transition state, resulting in a greater
proportion of trans-epoxide 9 being formed. For any of
substrates 1a and 3a-7a, the highest percentage of trans-
epoxide was formed using sodium hydroxide.
nature of the base counterion. For reactions involving
methoxy-substituted amides (6b and 5b), the reactions all
give predominantly cis-epoxide 8, though there is no
significant change in diastereoselectivity compared to the
results obtained with unsubstituted bromoamide 1b. Thus it
seems that the methoxy-substituents do not decrease the
value of k-2 below the value observed for bromoamide 1b.
Substrates 4b and 3b which contain electron-withdrawing
nitro groups still give predominantly trans-epoxide 9, though
the percentage of trans-epoxide is (with the exception of
substrate 3b and sodium hydroxide) always lower than that
obtained from the corresponding chloroamides (4a and 3a).
It is notable that the most electron-deficient substrate 3b
when treated with sodium hydroxide gives exclusively the
trans-epoxide product 9b in high yield, suggesting that in
this case, even the good leaving group does not result in k3
being large enough to prevent equilibration of enolates 2a,b.
The results obtained with substrate 7b which contains both a
4-nitro and a 4-methoxy substituted aromatic ring are again
roughly comparable with those obtained using substrate 5b
which contains just the methoxy group, indicating that the
influence of the methoxy substituent is again greater than
that of the nitro group.
The diastereoselectivities obtained with substrate 7a
which contains both a 4-nitro and a 4-methoxy substituted
aromatic ring are comparable to those of substrate 5a which
contains just the 4-methoxy substituent, indicating that the
influence of the methoxy substituent is greater than that of
the nitro group. The conversions achieved using substrate 7a
with rubidium and caesium hydroxides were however much
greater than those achieved using compound 5a which may
reflect the acidifying influence of the nitro group on the
protons adjacent to the carbonyl, thus increasing k1.
Table 2 gives the same comparison for bromoamide
substrates 1b and 3b-7b. In this case, the trends are less
pronounced. The conversions are extremely variable and one
reason for this may be the competing SN2 reaction between
hydroxide and the bromo-substrates. A bromo-substituent is
less electron-withdrawing than a chloro-substituent, which
may explain why the conversions are generally lower in the
bromo series (compare Table 1 with Table 2) since k1 will be
smaller in the former case. In general, the nature of the
To investigate the need for two aromatic rings within the
Darzens condensation substrates, known N-phenyl-N-alkyl
haloamides 15-20 [14-18] and N,N-dialkyl haloamides 21-26
[19-23] were prepared and used as substrates for Darzens
condensations to give epoxides 27a-f and 28a-f [24-27]
(Scheme 5) with the results being tabulated in Table 3.
Substrates 15-20 showed no variation in diastereoselectivity
between the substrates with bromide and chloride leaving
substrate has
a
much greater influence on the
diastereoselectivity of the Darzens condensation than the
Table 2. Comparison of Bromoamides 1b and 3b-7b in the Darzens Condensation with Benzaldehydea
Substrate
NaOH
KOH
RbOH
CsOH
6bb
5bc
1b
100 (48% trans)
79 (55% trans)
73d (40% trans)
35 (74% trans)
89 (100% trans)
27 (64% trans)
49 (28% trans)
58 (45% trans)
78d (29% trans)
45 (74% trans)
73 (60% trans)
53 (40% trans)
34 (33% trans)
57 (43% trans)
76 (33% trans)
39 (56% trans)
84 (71% trans)
73 (50% trans)
39 (33% trans)
90 (30% trans)
79 (29% trans)
60 (89% trans)
59 (60% trans)
50 (42% trans)
4bc
3bc
7bc
aIn each cell, the conversion is given first followed in brackets by the percentage of trans-epoxide 9 in the cis / trans mixture prior to purification as determined by 1H NMR
spectroscopy. bReaction time 65 hours. cReaction time 16 hours. dReaction time 4 hours.

Diastereoselective Darzens Condensations
Letters in Organic Chemistry, 2009, Vol. 6, No. 7
555
R1
O
O
PhCHO / MOH
N
R2
R1 CH2Cl2, RT, 16 h
Ph
Ph
X
N
O
R2
27a-f
+
O
15: X = Cl, R1 = Ph, R2 = Me
16: X = Br, R1 = Ph, R2 = Me
17: X = Cl, R1 = Ph, R2 = Et
18: X = Br, R1 = Ph, R2 = Et
19: X = Cl, R1 = Ph, R2 = iPr
20: X = Br, R1 = Ph, R2 = iPr
21: X = Cl, R1 = R2 = Me
22: X = Br, R1 = R2 = Me
23: X = Cl, R1 = R2 = Et
R1
N
R2
O
28a-f
a: R1 = Ph, R2 = Me
b: R1 = Ph, R2 = Et
c: R1 = Ph, R2 = iPr
d: R1 = R2 = Me
e: R1 = R2 = Et
24: X = Br, R1 = R2 = Et
25: X = Cl, R1 = R2 = iPr
26: X = Br, R1 = R2 = iPr
f: R1 = R2 = iPr
Scheme 5.
Table 3. Comparison of Haloamides 15-26 in the Darzens Condensation with Benzaldehydea
Substrate
NaOH
KOH
RbOH
15
16
17
18
19
20
21
22
23
24
25
26
100 (50% trans)
89 (50% trans)
89 (50% trans)
88 (48% trans)
84 (48% trans)
87 (48% trans)
43 (50% trans)
26 (58% trans)
21 (50% trans)
16 (43% trans)
11 (45% trans)
0
97 (50% trans)
85 (50% trans)
95 (40% trans)
89 (40% trans)
96 (40% trans)
82 (40% trans)
71 (50% trans)
36 (55% trans)
28 (45% trans)
17 (48% trans)
31 (43% trans)
0
92 (48% trans)
78 (45% trans)
87 (37% trans)
77 (40% trans)
95 (34% trans)
87 (40% trans)
64 (50% trans)
8 (67% trans)
41 (43% trans)
0
68 (38% trans)
0
aIn each cell, the conversion is given first followed in brackets by the percentage of trans-epoxide 28 in the cis / trans mixture prior to purification as determined by 1H NMR
spectroscopy. In each case the reaction time was 16 hours.
groups. In most cases an almost 1:1 ratio of cis- and trans-
epoxides was formed, though in every case the amount of
trans-epoxide 28a-c formed decreases slightly as the size of
the base counterion increases. The chemical yields of
epoxides 27d-f and 28d-f were significantly lower than those
obtained from N-aryl haloamides which reflects the lower
acidity of the hydrogens in substrates 21-26 since the lone
pair of electrons on the nitrogen atom no longer has an
aromatic ring to delocalise into and so is more delocalised
into the carbonyl bond. The low yields are particularly
apparent for bromoamide substrates 22, 24 and 26 and in
these cases an SN2 reaction of hydroxide on the alkyl
bromide may compete with the desired deprotonation due to
the low acidity of the hydrogens. The Darzens reactions of
substrates 21-25 show little or no diastereoselectity and no
discernable trends as the nature of the base counterion is
varied.
CONCLUSIONS
The electronic nature of the aromatic rings in N,N-diaryl
ꢀ-haloamides have
a
pronounced influence on the
diastereoselectivity of Darzens condensations. Electron-
withdrawing nitro substituents favour formation of the trans-
epoxide through a chelated, thermodynamically controlled
aldol reaction, whilst electron-donating methoxy substituents
enhance the formation of cis-epoxide, especially from ꢀ-
bromo substrates. When both 4-nitro and 4-methoxy
substituents are present in the amide, the methoxy group has
a greater influence on the diastereoselectivity than the nitro
group.
If one of the N-aryl groups is changed to an alkyl group,
then the influence of the halide on the diastereoselectivity of
the Darzens condensation disappears and the reactions show
no diastereoselectivity or a small preference for formation of

556 Letters in Organic Chemistry, 2009, Vol. 6, No. 7
North and Pizzato
[10]
[11]
[12]
For compound 11 see: Kotsuki, H; Kobayashi, S.; Matsumoto, K.;
Suenaga, H.; Nishizawa, H. Synthesis, 1990, 1147.
For compounds 12 and 13 see: Kelkar, A.A.; Patil, N.M.;
Chaudhari, R.V. Tetrahedron Lett., 2002, 43, 7143.
For compound 14 see: Wolfe, J.P.; Tomori, H.; Sadighi, J.P.; Yin,
J.; Buchwald, S.L. J. Org. Chem., 2000, 65, 1158.
Zhang, H.; Cai, Q.; Ma, D. J. Org. Chem., 2005, 70, 5164.
For compounds 15 and 17 see: Hennessy, E.J.; Buchwald, S.L. J.
Am. Chem. Soc., 2003, 125, 12084.
For compound 16 see: McAllister, L.A.; Turner, K.L.; Brand, S.;
Stefaniak, M.; Procter, D.J. J. Org. Chem., 2006, 71, 6497.
For compound 18 see: Trzupek, L.S.; Stedronsky, E.R.;
Whitesides, G.M. J. Org. Chem., 1972, 37, 3300.
cis-epoxides. N,N-Dialkyl haloamides are poor substrates for
Darzens condensations with solid metal hydroxides as bases,
giving low yields and little or no diastereoselectivity.
[13]
[14]
ACKNOWLEDGEMENT
Mass spectra were recorded by the EPSRC national mass
spectrometry service at the University of Wales, Swansea.
[15]
[16]
[17]
[18]
REFERENCES
For compound 19 see: Phillips, W.G.; Ratts, K.W. J. Org. Chem.,
1972, 37, 3818.
[1]
For a review of the Darzens reaction see: Rosen, T. In
For compound 20 see: Henke, B.R.; Aquino, C.J.; Birkemo, L.S.;
Croom, D.K.; Dougherty, R.W.; Ervin, G.N.; Grizzle, M.K.; Hirst,
G.C.; James, M.K.; Johnson, M.F.; Queen, K.L.; Sherrill, R.G.;
Sugg, E.E.; Suh, E.M.; Szewczyk, J.W.; Unwalla, R.J.; Yingling,
J.; Willson, T.M. J. Med. Chem., 1997, 40, 2706.
For compound 21 see: Miyazawa, T.; Tanaka, K.; Ensatsu, E.;
Yanagihara, R.; Yamada, T. J. Chem. Soc. Perkin Trans. 1, 2001,
87.
For compounds 22 and 26 see: Pospíꢁil, J.; Potáꢂek, M.
Tetrahedron, 2007, 63, 337.
For compound 23 see: Landor, P.D.; Landor, S.R.; Odyek, O. J.
Chem. Soc. Perkin Trans. 1, 1977, 93.
For compound 24 see: Hama, T.; Culkin, D.A.; Hartwig, J.F. J. Am.
Chem. Soc., 2006, 128, 4976.
For compound 25 see: Speziale, A.J.; Hamm, P.C. J. Am. Chem.
Soc., 1956, 78, 2556.
For compounds 27d,e see: Chan, S.O.; Wells, E.J. Can. J. Chem.,
1967, 45, 2123.
For compounds 28d,e see: Concellón, J.M.; Huerta, M.; Bardales,
E. Tetrahedron, 2004, 60, 10059.
Comprehensive Organic Synthesis; Trost, B.M.; Fleming, I. Eds.;
Pergamon: Oxford, 1991; Vol. 2, pp. 409-439.
[2]
[3]
[4]
[5]
[6]
[7]
Achard, T.J.R.; Belokon’, Y.N.; Hunt, J.; North, M.; Pizzato, F.
Tetrahedron Lett., 2007, 48, 2961.
Achard, T.J.R.; Belokon’, Y.N.; Ilyin, M.; Moskalenko M.; Hunt,
J.; North, M.; Pizzato, F. Tetrahedron Lett., 2007, 48, 2965.
Arai, S.; Suzuki, Y.; Tokmaru, K.; Shioiri, T. Tetrahedron Lett.,
2002, 43, 833.
Wang, Z.; Xu, L.; Mu, Z.; Xia, C.; Wang, H. J. Mol. Catal. A,
2004, 218, 157.
Wang, Z.-T.; Xu, L.-W.; Xia, C.-G.; Wang, H.Q. Helv. Chim. Acta,
2004, 87, 1958.
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
To a stirring mixture of a haloacetamide (0.7 mmol) and a metal
hydroxide base (1.14 mmol) in CH2Cl2 (4 mL) was added
benzaldehyde (64 mg, 0.6 mmol). After being stirred for the
appropriate time, the reaction was quenched with 1 N HCl (1 mL)
and the aqueous layer was extracted with EtOAc (3ꢀ2 mL), washed
with brine (6 mL) and dried (Na2SO4). The solvent was removed in
1
vacuo and the crude product analysed by H NMR spectroscopy to
determine the diastereomeric ratio. Purification by flash column
chromatography (hexane / EtOAc) then gave the cis- and trans-
epoxides.
For compound 27f see: Concellón, J.M.; Bardales, E.; Concellón,
C.; García-Granda, S.; Díaz, M.R. J. Org. Chem., 2004, 69, 6923.
For compound 28f see: Meth-Cohn, O.; Moore, C.; Taljaard, H.C.
J. Chem. Soc. Perkin Trans. 1, 1988, 2663.
[8]
[9]
Of compounds 3a,b to 7a,b, only compound 4a has been reported
before: Wadia, M.S.; Patil, D.V. Synth. Commun., 2003, 33, 2725.
For compound 10 see: Gorvin, J.H. J. Chem. Soc. Perkin Trans. 1,
1988, 1331.