1,2- versus 1,4-addition to α-donor-cyclopenten-1-ones; a comparison of calculated and experimental data

Article abstract of DOI:10.1039/a905444h

A series of α-donor-cyclopenten-1-ones, 1 (α-donor: X = H, Cl, Br, OMe, pyrrolidin-1-yl, SPh and SePh) were treated with methyllithium, lithium dimethylcuprate and cyanomethyllithium in tetrahydrofuran at -78°C. The reactions between 1 and methyllithium gave 1,2-adducts in 28-75% yield. No trace of the corresponding 1,4-adduct was observed. By treatment of 1 with lithium dimethylcuprate only 1,4-addition was obtained in 51-76% yield, except for X = OMe or pyrrolidin-1-yl, which gave no reaction at all. In the reactions between 1 and cyanomethyllithium the 1,2-adducts were the main products (43-60%), but varying amounts of 1,4-adducts (8-30%) were also formed. The products formed have been fully characterized. Net atomic charges obtained by density functional theory (DFT) calculations of the lithiated cations of 1 with X = H, Cl, Br or OMe showed very small differences between the substituted and unsubstituted systems for all X, except for X = OMe. In this case the C² position was positive rather than negative. The C³ position was found to be significantly more negative. The calculated relative energies of the methylated and cyanomethylated 1,2- and 1,4-adducts with X = H, Cl, Br and OMe indicated that the 1,4-adducts were thermodynamically favoured by 8 to 12 kcal mol^-1 for all but one case. In the cyanomethylation of 1 with X = OMe the 1,2-adduct was predicted to be favoured by 12 kcal mol^-1. Thus, the predominance of 1,2-attack in the reactions involving methyllithium and cyanomethyllithium cannot be explained by the relative thermodynamic stabilities of the products.

Full text of DOI:10.1039/a905444h

View Article Online / Journal Homepage / Table of Contents for this issue
1,2- versus 1,4-Addition to ꢀ-donor-cyclopenten-1-ones;
a comparison of calculated and experimental data†
Hanna S. H. Gautun, Rolf Carlson* and Per N. Skancke
Department of Chemistry, Faculty of Science, University of Tromsø, NO-9037 Tromsø,
Norway
Received (in Lund, Sweden) 23rd April 1999, Accepted and transferred from Acta Chem.
Scand. 1st July 1999
A series of α-donor-cyclopenten-1-ones, 1 (α-donor; X = H, Cl, Br, OMe, pyrrolidin-1-yl, SPh and SePh) were treated
with methyllithium, lithium dimethylcuprate and cyanomethyllithium in tetrahydrofuran at Ϫ78 ЊC. The reactions
between 1 and methyllithium gave 1,2-adducts in 28–75% yield. No trace of the corresponding 1,4-adduct was
observed. By treatment of 1 with lithium dimethylcuprate only 1,4-addition was obtained in 51–76% yield, except
for X = OMe or pyrrolidin-1-yl, which gave no reaction at all. In the reactions between 1 and cyanomethyllithium
the 1,2-adducts were the main products (43–60%), but varying amounts of 1,4-adducts (8–30%) were also formed.
The products formed have been fully characterized. Net atomic charges obtained by density functional theory
(DFT) calculations of the lithiated cations of 1 with X = H, Cl, Br or OMe showed very small differences between
the substituted and unsubstituted systems for all X, except for X = OMe. In this case the C2 position was positive
rather than negative. The C3 position was found to be significantly more negative. The calculated relative energies
of the methylated and cyanomethylated 1,2- and 1,4-adducts with X = H, Cl, Br and OMe indicated that the
1,4-adducts were thermodynamically favoured by 8 to 12 kcal molϪ1 for all but one case. In the cyanomethylation
of 1 with X = OMe the 1,2-adduct was predicted to be favoured by 12 kcal molϪ1. Thus, the predominance of
1,2-attack in the reactions involving methyllithium and cyanomethyllithium cannot be explained by the relative
thermodynamic stabilities of the products.
α-donor-α,β-unsaturated ketones have been used in conjugated
Introduction
addition reactions. Kowalski et al.2a have reported selective
As part of a larger project on the synthetic use of α-donor-α,β-
1,4-addition to 2-bromocyclopent-2-en-1-one, 1c. They iso-
unsaturated ketones, cyclopentenones have been of particular
lated the 1,4-adduct as the corresponding acetate in 73% yield.
interest as potentially useful synthetic intermediates. Especially,
nucleophilic conjugate addition to such substrates would be a
Other cyclic α-bromo-α,β-unsaturated ketones have also been
reported to undergo selective 1,4-addition of allyllithium,
useful reaction.1 The presence of an electron donating substitu-
allylmagnesium bromide and lithium dialkylcuprates.2b,c The
ent in the α-position of the cyclopentenone as shown in Scheme
yields obtained were generally good, ≥74%. In a paper by
1 would make it possible to regenerate the initial double bond.
Fang3a 2-phenylthiocyclopent-2-en-1-one, 1f, was treated with
2-methylmalonic methyl and ethyl esters and ethyl propionate to
give selective 1,4-addition. In the latter case addition of HMPA
to the enolate was required to suppress the competing 1,2-
addition reaction. The isolated yields varied from 45 to 95%.
Monteiro3b has applied 1,4-addition of Me2CuLi to 1f as a key
step in the synthesis of ( )-methyl dehydrojasmonate. Liotta
and coworkers4 have subjected various cyclic α-phenylselenyl
enones (e.g. 1g) to several organocuprates and obtained excel-
lent 1,4-selectivities and yields. Addition reactions between
cyclic α-ketoenamines (e.g. pyrrolidin-2-ylcyclopent-2-en-1-
one, 1e) and diazenes have been reported by Felluga et al.5 to
take place at the β-position. These reactions may, however, be
regarded as analogues to the Diels–Alder reactions. Finally,
Scheme 1 α-Donor-cyclopentenones as electrophiles in conjugate
selective 1,4-addition of organocuprates to (S)-2-(p-tolyl-
addition reactions.
sulfinyl)cycloalk-2-en-1-ones has been reported by Posner et al.6
Although the sulfinyl group is generally regarded as electron
accepting, it may also function as a donor of its lone pair
electrons.
Even though the above mentioned reports of successful con-
jugate addition to cyclic α-donor-α,β-unsaturated ketones have
The product formed could then be used in a second conjugate
addition. Thus, 1 would constitute a synthetic equivalent for the
unknown 2-oxocyclopent-1-yne, 2, illustrated in Scheme 1.
There are only a few examples in the literature where cyclic
been available for quite some time, the application of this group
† Spectroscopic data and optimized geometries for a selection of
of compounds is rather scarce. In order to shed some light on
compounds are available as supplementary data. For direct electronic
the scope and limitations of these compounds as electrophiles
access see http://www.rsc.org/suppdata/p2/1999/1835, otherwise avail-
in conjugate addition reactions, we have treated a series of
able from BLDSC (SUPPL. NO. 57606, pp. 28) or the RSC Library.
α-donor-cyclopenten-1-ones, 1, with various substituents X
[X = H, Cl, Br, OMe, pyrrolidin-1-yl (Pyl), SPh and SePh] with
See Instructions for authors available via the RSC web page (http://
www.rsc.org/authors).
J. Chem. Soc., Perkin Trans. 2, 1999, 1835–1840
1835

View Article Online
Table 1 Absolute energies (E/hartree) and relative energies (∆E/kcal molϪ1) estimated from Mulliken population analysis. ∆E for 9a–d has been
defined as the reaction energy for 1a–d ϩ Liϩ→9a–d. ∆E for 10 (11) a–d has been defined as the reaction energy for 1a–d ϩ 7→10 (11) a–d. ∆E for 12
(13) a–d has been defined as the reaction energy for 1a–d ϩ 8→12 (13) a–d
Species
ϪEa
ϪEb
Ϫ∆Ea
Ϫ∆Eb
ZPEb
Ϫ∆G(solv.)b
Liϩ
:CH3
[:CH2CN]
1a
b
c
7.28491
7.28494
39.82371
132.16970
269.41370
729.03540
2842.9566
383.96250
47.41397
39.80385
132.11228
269.30939
283.64702
281.86305
383.81389
47.39779
61.8
55.8
55.4
82.0
4.1
3.7
3.9
3.0
6.0
d
7
8
139.64702
276.68843
291.01801
3289.23930
391.21073
316.78435
331.12990
329.34882
431.30515
316.78434
331.15138
329.37009
431.31860
409.01291
423.35603
421.57486
523.53305
409.01668
423.36895
421.58793
523.51868
21.5
63.3
57.0
56.6
84.1
87.5
81.3
81.0
108.0
87.5
82.4
82.1
108.4
86.4
80.3
80.0
107.0
87.2
81.2
81.0
107.3
9.7
9a
b
c
d
10a
b
c
d
11a
b
c
d
12a
b
c
d
276.78792
736.40830
2850.3320
391.35593
316.89439
776.52727
2890.44943
431.45900
316.91054
776.55357
2890.47615
431.47602
56.1
55.2
56.8
68.1
41.9
48.9
49.4
51.8
52.1
65.4
66.2
62.5
59.0
56.6
57.3
70.2
56.1
55.9
55.2
58.7
48.4
69.4
68.6
67.0
35.5
41.5
40.7
45.3
37.8
49.6
48.8
36.2
46.0
43.0
43.1
42.8
8.1
6.7
6.6
8.2
10.9
7.1
7.0
6.8
6.7
8.0
5.7
7.6
12.2
8.6
8.8
13a
b
c
d
11.8
a Calculation level B3LYP/6-311G*. b Calculation level B3LYP/LanL2DZ.
three different nucleophiles (MeLi, Me2CuLi and LiCH2CN).
In these reactions both 1,2- and 1,4-additions were anticipated
as depicted in Scheme 2.
Scheme 2 Nucleophilic addition to α-donor-cyclopentenones 1a–g.
Scheme 3 Compounds included in the DFT calculations.
The results obtained were evaluated in the light of theoretical
calculations of the net charge distributions and frontier orbitals
for 1 with X = H, Cl, Br and OMe, and the relative energies of
the lithiated 1,2- and 1,4-adducts obtained by reaction of 1a–d
with MeLi and NCCH2Li. The theoretical calculations were
carried out to shed some light on the structures and relative
stabilities of the starting materials and products. A more
thorough study of reaction pathways is beyond the scope of the
present work.
largest differences in geometries. Only the key geometrical
parameters showing the general shape of the molecules, are
included.
In this study we have tacitly assumed that the additions of
MeLi and NCCH2Li to the cyclopentenones occur as two-step
reactions in which the Liϩ first adds, creating a cation which is
susceptible to a nucleophilic attack by the methyl and cyano-
methyl anions, respectively.
The relative energies of the cations 9a–d obtained by the two
basis sets show that the addition of Liϩ to the cyclopentenones
is strongly exothermic, as expected. Furthermore, the predicted
exothermicity is found to be largely independent of the sub-
stituent X, except for the OMe group which gives a larger reac-
tion energy. This is easily explained as being due to a favourable
position of the positively charged Li atom in the attractive field
of two oxygen atoms. This increased stability of the cation is
clearly demonstrated by the optimized structures of 9a and 9d.
In the former the C1–O–Li arrangement is linear with an O–Li
bond of 1.73 Å, whereas in the latter there is a large deviation
from linearity, and a lengthening of this bond to 1.89 Å with an
Computational studies
The compounds included in the computational studies are
shown in Scheme 3. The absolute and relative energies for all
the species included in our study are given in Table 1, where
also zero-point energy values and solvent effects using the
LanL2DZ basis are included. Numbering of the different
species is given in Scheme 3.
The optimized geometries of a selection of the molecules are
available as supplementary data.† We have chosen to present
the molecules for which X = H and OMe as these display the
1836
J. Chem. Soc., Perkin Trans. 2, 1999, 1835–1840

View Article Online
Table 2 Net atomic charges for the cations 9a–d estimated from
Mulliken population analysis based on B3LYP/LanL2DZ
Table 3 Treatment of 1a–g with methyllithium in THF, at Ϫ78 ЊC
1:3:4 (GLC of
reaction, 1 h)
Crude
(%)
Isol. prod.
(%)
9
qO
qC1
qC2
qC3
1
X
a
a
b
c
Ϫ0.60
Ϫ0.50
Ϫ0.49
Ϫ0.40
0.37
0.14
0.49
0.26
Ϫ0.28
Ϫ0.18
Ϫ0.29
0.30
Ϫ0.09
Ϫ0.10
Ϫ0.07
Ϫ0.23
a
b
c
d
e
f
H
Cl
Br
OMe
Pyl
SPh
SePh
3.1:96.6:0
23.1:76.9:0
17.2:82.0:0
13.5:86.5:0
67.4:20.5:0
13.2:86.7:0
20.5:63.6:0
—
—
82.5
75.9
67.5
—
94.4
83.2
28.0
35.8
b
d
—
—
b
75.6
59.7
g
a Compound 3a was identified by GLC-MS of the reaction mixture.
b Compounds 4e–d were identified by GLC-MS of the reaction mixture,
but were hydrolysed to 2-hydroxy-2-methylcyclopentanone (14) during
work-up.
the latter member of the pairs, that is the 1,4-adducts. This is
independent of whether the nucleophile is MeLi or NCCH2Li,
with the notable exception of 12d, which has a lower energy
than the isomer 13d. In the case of MeLi the energy differences
between the two isomeric forms of the reaction product were
estimated to be around 10–12 kcal molϪ1 in favour of the 1,4-
adducts, irrespective of the nature of the substituent at C2. For
NCCH2Li the corresponding differences for the substituted
compounds were found to be around 8 kcal molϪ1 in favour of
the 1,4-adducts, except for the methoxy derivative where this
difference was around 9 kcal molϪ1 in favour of the 1,2-adduct.
So, regarding the thermodynamic stability of the reaction
products alone would suggest that addition to C3 would be
favoured as compared to addition to C1. An exception was
found for the addition of NCCH2Li to the methoxy derivative
9d, which gives a product 12d having a very favourable attract-
ive interaction involving Li, the two oxygen atoms and the
nitrile nitrogen. The solvent effects estimated as described
above are found to be roughly equal for the 1,2- and 1,4-
adducts and a given substituent. This implies that the conclu-
sions reached above on the basis of the theoretical calculations
would be valid also in a THF solution.
Fig. 1 LUMO ϩ 1 orbitals in the cations 5a–d from RHF/3-21G*
calculations.
additional bonding interaction to the methoxy group resulting
in an O–Li distance of 1.98 Å.
The calculated atomic net charges in the cations 9a–d based
on a Mulliken population analysis are shown in Table 2. The
data reveal a charge distribution leading to a pronounced posi-
tive charge on the carbon atom (C1) adjacent to the oxygen, a
significant negative charge on C2, except for in 9d where this
atom has a positive charge due to the proximity of the methoxy
oxygen. The atom C3 has a negative, but very small net charge
in 9a–c. In 9d the negative charge is rather significant. These
charges indicate that the most susceptible site for a nucleophilic
attack on the cations 9 would be C1.
Fig. 1 depicts a picture of the second lowest unoccupied
(LUMO ϩ 1) orbitals obtained for the geometry optimized
cations 9 by SCF calculations using a 3-21G* basis set.7 The
reason for this choice of orbitals is that the LUMO in the
cations is for obvious reasons strictly located on the Li atom
from which the nucleophilic anions have escaped.
Using frontier orbitals rather than net charges the shape and
extension of the second lowest unoccupied orbital (LUMO ϩ
1) in the cations, 9a–d, would be expected to give information
on the most favoured region for a nucleophilic attack. As shown
in Fig. 1 these orbitals, which are independent of the substitu-
ent X, are dominated by the π-orbital confined to the carbonyl
oxygen atom and the virtually planar unsaturated part of the
ring. These orbitals have a node between oxygen and C1, and
another one between C2 and C3. The middle region of the
orbital, located between C1 and C2 is oriented more towards C1
than C2, thus making this site a likely one for nucleophilic
attack. The figure also shows a dominant part of the orbital
centred around C3, making this centre competitive with C1 for
addition of an anion.
Experimental studies
The α-donor substituted cyclopentenones 1b and 1d–g were
prepared according to literature procedures.2a,3b,4,8,9 Compound
1c was prepared by direct chlorination of cyclopent-2-en-1-one
in 37% yield. These model substrates were treated with methyl-
lithium, lithium dimethylcuprate and cyanomethyllithium in
THF at Ϫ78 ЊC as depicted in Scheme 2. The product com-
position of the reaction mixtures was analysed by GLC after
1 h, and the reactions worked up after 1.5 h. The results of
treatment of 1a–g with MeLi are given in Table 3.
As expected, only 1,2-adducts (3a–g) were obtained. No
traces of 1,4-adducts (4a–g) were observed by GLC analysis
of the reaction mixtures or of the crude products. For 1d the
1,2-adduct 3d was isolated as a crude product and identified
1
by H NMR and GLC-MS. However, 3d was hydrolysed after
5–10 min at room temperature to 2-methyl-2-hydroxycyclo-
pentanone, 14, which quickly was dehydrated to give 2-
methylcyclopent-2-en-1-one, 15, which was identified by GC
1
co-chromatography with an authentic sample, H NMR, 13C
NMR and GLC-MS. The final product 15 was, however not
isolated due to its volatility. Similarly, 3e was identified as the
only product from 1e by GLC-MS of the reaction mixture. The
enamine function was hydrolysed within minutes at room
temperature to give 2-hydroxy-2-methylcyclopentanone, 14.
With the exception of 1a, relatively large amounts of starting
material remained unreacted, even after prolonged reaction
times (12 h). GLC analysis of experiments performed in the
presence of an internal standard showed a “loss of mass” up to
40%. This indicated that MeLi also reacted as a base, abstract-
ing an α-proton as a side reaction. The low isolated yields for
Comparison of the relative energies (uncorrected for ∆ZPE)
for the isomer pairs 10a–d and 11a–d (Nu: MeLi), and 12a–d
and 13a–d (Nu: NCCH2Li), shows a significant preference for
J. Chem. Soc., Perkin Trans. 2, 1999, 1835–1840
1837

View Article Online
Table 4 Treatment of 1a–g with lithium dimethylcuprate in THF, at
Ϫ78 ЊC
Table 5 Treatment of 1a–g with cyanomethyllithium in THF, at
Ϫ78 ЊC
1:3:4 (GLC of
reaction, 1 h)
Crude
(%)
Isol. prod.
(%)
cis:trans
1:5:6 (GLC of
reaction, 1 h)
Crude
(%)
Isol. prod.
(%)
cis:trans
1
X
(1H NMR)
1
X
(1H NMR)
a
a
b
c
d
e
f
H
Cl
Br
OMe
Pyl
SPh
SePh
0:0:100
0:0:100
0:0:100
100:0:0
100:0:0
0:0:100
0:0:100
<60
—
—
a
b
c
H
0:67.8:32.2
0:92.1:8.9
0:98.5:1.5
80
5a; 43.3
6a; 30.0
5b; 47.5
6b; 8.9
5c; 46.0
6c; 0
6d; 60.1
16; 74.8a
5f; 57.1
6f; 16.2
5g; 48.0
6g; 14.1
—
—
—
2.9:1
—
—
—
—
—
87.6
87.3
—
—
80.4
81.8
53.6
53.3
—
3.9:1
3.2:1
—
—
7.8:1
8.3:1
Cl
Br
84.0
70.5
—
51.3
76.1
g
d
e
f
OMe
Pyl
SPh
0:100:0
5.2:94.8:0
1.3:80.3:18.0
90.7
79.7
79.9
a Compound 4a was identified by GLC-MS of the reaction mixture and
GLC co-chromatography with an authentic sample.
6.1:1
—
5.6:1
g
SePh
4.7:68.2:16.4
78.5
3b, c and g may be caused by reactions on the silica gel used for
purification.
a Compound 5e hydrolysed to 2-cyanomethyl-2-hydroxycyclopentan-
one (16) upon work-up.
The calculated relative energies for the lithiated products
10a–d and 11a–d indicated that the 1,4-adducts would be the
thermodynamically most stable product. However, in the reac-
tions of 1a–g with MeLi only 1,2-adducts were obtained. This
indicates that the products are formed under kinetic control. A
thorough understanding of the experimentally obtained data
would, however, require estimation of all activation energies for
the possible steps involved in the total reaction sequence from
starting material to products. If the barrier to dissociation of
MeLi in the field of the dissolved cyclopentenones is larger than
the barrier to a concerted addition in the carbonyl region of the
molecules, the reaction products from 1,2-addition, 3a–g,
would be favoured. If, on the other hand, the dissociation bar-
rier of MeLi is low, the methyl anion would be able to orient
itself in the field of the cations in a second step of the reaction,
making 1,4-addition possible.
The results obtained by treatment of 1a–g with Me2CuLi
are given in Table 4. Only 1,4-adducts were obtained, with the
exception of 1d and 1e, which were found to be unreactive
towards Me2CuLi. The crude yields were fairly good and the
purities above 90% by 1H NMR. However, the yields after flash
chromatography were only moderate.
The mechanism for the addition of dialkyl cuprates to conju-
gated carbonyl compounds has been investigated by several
authors.10 The generally accepted mechanism today involves the
following steps: first lithium coordinates to the carbonyl oxy-
gen, copper coordinates to the double bond. Copper then adds
as copper() to the C3 position and thereby directs one methyl
group into this position, rather than the more electrophilic C1
position. When the electron donating group, X, was OMe or
Pyl, 1d–e, no reaction took place. We interpret these findings as
a result of the increased negative charge in the C3 position of
the starting materials, thus disfavouring the transfer of the
methyl group. It may also be that the donor group X competes
with the double bond in the coordination to copper.
amount of 1,4-adduct, 6a–g, formed in the reactions decreased
in the following order of X: H (32%) > SPh, SePh (18–
15%) > Cl (9%) > Br (2%) > OMe, pyrrolidinyl (0%). The net
charge distribution of the cation 9d (X = OMe) also showed a
clear negative charge on the C3 atom. The relative energies for
12d and 13d showed that the 1,2-adduct 12d was the thermo-
dynamically favoured one. Both of these factors strongly dis-
favour 1,4-addition. If we assume that the lone pair electrons of
the pyrrolidinyl substituent are in conjugation with the double
bond system of the cyclopentenone ring, the influence of the
lone pair electrons may be expected to be very similar to the
influence of the OMe group on the charge distributions of 9e
and the relative energies of 12e and 13e. In the cases where X
was Cl and Br very little 1,4-addition was obtained. For these
compounds the net charge distributions did not differ much
from the unsubstituted system. The relative energies showed
that the 1,4-adducts were thermodynamically preferred. The
reason for the low share of 1,4-addition might therefore be
ascribed to an increase in the ability of X to coordinate to the
nucleophile as X becomes more electronegative and polarizable.
Such a coordination might influence the direction of the attack
in favour of 1,2-addition.
Conclusion
By treatment of 1a–g with methyllithium the 1,2-adducts 3a–g
were formed in 28–75% yield. No traces of the corresponding
1,4-adducts 4a–g were observed, indicating that the products
were formed under kinetic control. The reactions between 1a–g
and lithium dimethylcuprate gave only the 1,4-adducts 4a–c and
4f–g in 51–76% yield. No reaction was observed for 1d and
1e (X = OMe and pyrrolidin-1-yl). Treatment of 1a–g with
cyanomethyllithium predominantly gave the 1,2-adducts 12a–g
(43–60% yield). However, formation of the corresponding 1,4-
adducts 13a–c and 13f–g (8–30% yield) was also observed
with the amount of 1,4-addition decreasing in the following
order: H (32%) > SPh, SePh (18–15%) > Cl (9%) > Br (2%) >
OMe, pyrrolidinyl (0%).
The ratio of cis:trans 1,4-adducts were in all cases in favour
of the cis isomer. A tentative way to rationalize these observ-
ations is to involve steric effects. Clarification of this issue
would, however, require extensive computational studies reach-
ing far beyond the scope of this project.
The reactions between 1a–g and NCCH2Li also gave pre-
dominantly 1,2-adducts, but various amounts of 1,4-adducts,
6a–c, 6f–g were also obtained, as shown in Table 5. In the
cases where X was OMe and Pyl, 1d and 1e, only 1,2-addition
was obtained. For the reaction with 1e, the isolated product
was 2-cyanomethyl-2-hydroxycyclopentanone, 16, formed by
hydrolysis of the enamine during work-up. GLC of reaction
mixtures run with internal standard did not indicate any major
“loss of mass”, except for the reaction involving 1g.
The calculated net atomic charges of 9a–d obtained by DFT
calculations showed only minor differences for the α-donor-
substituted systems 9b and 9c (X = Cl and Br) compared to the
unsubstituted compound, 9a. Compound 9d (X = OMe) dif-
fered, however, in the C2 position, which was found to be posi-
tively charged rather than negatively as for the others. The C3
position in 9d was found to be significantly negative relative to
the three other cyclopentenones 9a–c. The relative energies
estimated for the 1,2- and the 1,4-adducts from addition of
methyllithium or cyanomethyllithium pointed out the 1,4-
adduct as thermodynamically favoured. An exception was
found for the addition of cyanomethyllithium to 1d (X = OMe),
The crude yields were acceptable, but the yields of the puri-
fied products were somewhat lower than expected. Also in these
reactions, the cis isomer was the favoured 1,4-adduct. The
1838
J. Chem. Soc., Perkin Trans. 2, 1999, 1835–1840

View Article Online
where the 1,2-adduct, 12d, was the thermodynamically most
stable one.
1H NMR δ 2.52–2.54 (2H, m), 2.67–2.70 (2H, m), 7.58 (1H, t,
J 2.93); 13C NMR δ 25.9, 33.0, 136.3, 157.3, 201.1; MS m/z 118
(M ϩ 2, 4%), 116 (M, 12), 88 (8), 73 (8), 62 (9), 61 (9), 60 (21),
53 (49), 52 (9), 51 (26), 50 (29), 49 (18), 48 (10), 47 (24).
For synthetic purposes conjugate 1,4-addition is the most
desirable reaction. The estimated relative energy calculations
showed that the 1,4-adducts were thermodynamically most
stable. This points to a future development of the experimental
procedures favouring the 1,4-addition. A possible lead may be
to explore electrophilic catalysis in reaction with nucleophiles
giving mixtures of 1,2- and 1,4-addition, such as cyanomethyl-
lithium.
General procedure for treatment of ꢀ-donor-cyclopent-2-en-1-
ones with methyllithium
A solution of MeLi (1.65 M in diethyl ether, 1.0 equiv.) in THF
(2 ml) was cooled to Ϫ78 ЊC. To this mixture was added a cold
(Ϫ78 ЊC) solution of 1a–g (0.5 mmol) in THF (3 ml) via can-
nula. The reaction was stirred at Ϫ78 ЊC for 1 h. At this time, a
sample (10 µl) for GC analysis was withdrawn. After 1.5 h total
reaction time the mixture was lifted out of the cooling bath and
water added (3 ml) immediately. The mixture was stirred at
room temperature for 10 min and then extracted with CH2Cl2
(3 × 5 ml). The combined organic layers were washed with sat.
NaCl solution (4 ml) and dried (MgSO4). Evaporation of the
solvent under reduced pressure gave a crude product which was
Experimental
All reactions were carried out under a nitrogen atmosphere.
The chemicals were commercial products of p.a. quality and
used directly as delivered unless otherwise stated. Copper
iodide was purified by treatment with potassium iodide as
described by Taylor.10b All solvents were dried by published
procedures.19 Melting points were determined on a Büchi 535
apparatus and are uncorrected. TLC was performed on Merck
5554 Fertigplatten, DC-Alufolien, Kieselgel 60254, using UV
light at 254 nm and 5% alcoholic molybdophosphoric acid for
detection. Flash chromatography was carried out using Merck’s
Kieselgel 60 (230–400 mesh). GLC analyses were performed on
a Varian 3300 chromatograph equipped with split injector, FID
detector and a Varian 4400 integrator. Two capillary columns
were applied; Chrompack CP sil 8 CB (25 m) and Supelco SPB-
20 (30 m). The elemental analyses were performed by the Insti-
tute of Chemical Technology in Prague, Czech Republic. Exact
mass measurements were carried out on an AEI MS-902 double
focusing mass spectrometer (Nier-Johnson geometry) and a
Fisons Instrument VG Prospec mass spectrometer, by peak
matching with perfluorokerosene as standard for mass refer-
ences. IR spectra were recorded on a 470 IR Shimadzu spec-
trometer and are reported as wave number, ν. 1H and 13C NMR
spectra were recorded on a JEOL JNM-EX 400 FT NMR
SYSTEM, using CDCl3 as solvent and reference. Chemical
shifts are given in ppm and J values in Hz. Mass spectra were
registered on a Hewlett Packard 5890 Series II GC spec-
trometer equipped with a Chrompack CP sil 5 CB capillary
column (25 m), followed by a VG Quattro mass spectrometer.
The spectra were obtained at 220 ЊC and 70 eV. Signals of
intensity <5% have been omitted.
1
analysed by GLC and H NMR, and subsequently purified by
flash chromatography.
The product composition and yields obtained by treatment
of 1a–g with MeLi are given in Table 3. Full spectroscopic
documentation of the isolated products are available as sup-
plementary data.†
General procedure for treatment of ꢀ-donor-cyclopent-2-en-1-
ones with lithium dimethylcuprate
A suspension of CuI (2 equiv.) in THF (6 ml) was cooled to
Ϫ15 ЊC and a solution of MeLi in diethyl ether (1.71 M, 4
equiv.) was added. The mixture was stirred at Ϫ15 ЊC until all
the yellow, solid material was dissolved. The resulting clear,
colour free cuprate solution was cooled to Ϫ78 ЊC and added to
a cold (Ϫ78 ЊC) solution of 1a–g (0.6 mmol) in THF (3 ml) via
cannula. The reaction was stirred at Ϫ78 ЊC for 1 h, at which
time a sample (10 µl) was withdrawn for GLC analysis. After a
total reaction time of 1.5 h, the reaction was lifted out of the
cooling bath and aqueous NH3 (2 M, 3 ml), aqueous NH4Cl
(10 weight%, 3 ml) and diethyl ether (10 ml) were added.
The mixture was stirred vigorously in air at room temperature
until all precipitated material had dissolved. The phases were
separated and the deep blue aqueous layer was extracted with
diethyl ether (2 × 10 ml). The combined organic layers were
washed with sat. NaCl solution (10 ml) and dried (MgSO4).
Evaporation of the solvent under reduced pressure afforded a
crude product which was analysed by GLC and 1H NMR prior
to purification by flash chromatography.
Preparation of ꢀ-donor-cyclopenten-1-ones
2-Bromocyclopent-2-en-1-one2a (1b), 2-methoxycyclopent-
2-en-1-one16 (1d), 2-(pyrrolidin-1-yl)cyclopent-2-en-1-one17 (1e),
2-phenylthiocyclopent-2-en-1-one3b (1f) and 2-phenylseleno-
cyclopent-2-en-1-one4 (1g). These were prepared according to
literature procedures.
The product composition and yields obtained by treatment
of 1a–g with lithium dimethylcuprate are given in Table 4. Full
spectroscopic documentation of the isolated products are avail-
able as supplementary data.†
General procedure for treatment of ꢀ-donor-cyclopent-2-en-1-
ones with cyanomethyllithium
2-Chlorocyclopent-2-en-1-one (1c). A solution of cyclopent-
2-en-1-one (2.00 ml, 64.56 mmol) in dry CH2Cl2 (50 ml) was
cooled to Ϫ78 ЊC and Cl2 was bubbled through until the reac-
tion turned yellow (2–3 min). The reaction was stirred at
Ϫ78 ЊC until it became colour free (5 min) and then Cl2 was
bubbled through until the yellow colour reappeared (1–2 min).
After stirring at Ϫ78 ЊC for another 30 min to the reaction was
added aqueous NaOH (30%, 30 ml) and diethyl ether (30 ml).
The mixture was stirred vigorously at room temperature for 1 h.
The phases were separated and the aqueous layer extracted with
diethyl ether (2 × 30 ml). The combined organic layers were
washed with aqueous Na2CO3 (3 × 50 ml), water (2 × 50 ml)
and sat. NaCl solution (50 ml). Drying (MgSO4) and evapor-
ation of the solvent under reduced pressure afforded 1c (1.08 g,
37% yield, purity > 99% by GLC) as a light yellow oil. HRMS
Found M 116.0025. Calc. for C5H5ClO: 116.0029; IR νmax (neat)
3410w, 3070w, 2920m, 1715s, 1595s, 1435m, 1400m, 1290s,
A solution of acetonitrile (1 equiv.) in THF (2 ml) was cooled to
Ϫ78 ЊC and a solution of methyllithium in diethyl ether (1.65
M, 1 equiv.) was added. The mixture was stirred for 15 min. A
precooled solution (Ϫ78 ЊC) of 1 (0.5 mmol) in THF (3 ml) was
then transferred into the cyanomethyllithium solution via can-
nula. The reaction was kept stirring at Ϫ78 ЊC and monitored
by GLC after 1 h. After 1.5 h the reaction was lifted out of the
cooling bath, aqueous ammonium chloride (sat., 1 ml) was
added and the mixture was stirred vigorously for 15 min. The
organic solvent was evaporated before the residue was treated
with more aqueous ammonium chloride (sat., 3 ml) and
extracted with ethyl acetate (3 × 5 ml). The combined organic
layers were washed with sat. NaCl solution (4 ml) and dried
(MgSO4). Removal of the solvent under reduced pressure gave
a crude product which was analysed by GC and 1H NMR
before it was purified by flash chromatography.
1235m, 1165m, 1015m, 995m, 950s, 915w, 780m, 750m cmϪ1
;
J. Chem. Soc., Perkin Trans. 2, 1999, 1835–1840
1839

View Article Online
1994, 59, 7924; (c) T. Kametani, K. Suzuki, H. Nemeto and
The product composition and yields obtained by treatment
of 1a–g with cyanomethyllithium are given in Table 5. Full
spectroscopic documentation of the isolated products is avail-
able as supplementary data.†
K. Fukumoto, J. Org. Chem., 1979, 44, 1036.
3 (a) J.-M. Fang, J. Org. Chem., 1982, 47, 3464; (b) H. J. Monteiro,
J. Org. Chem., 1977, 42, 2324.
4 D. Liotta, M. Saindane, C. Barnum and G. Zima, Tetrahedron, 1985,
41, 4881.
5 F. Felluga, P. Nitti, E. Ruocco and C. Russo, Gazz. Chim. Ital., 1997,
127, 387.
Computational method
The species included in the computational studies were all fully
optimized by density functional theory (DFT) calculations,11
using Becke’s 3-parameter hybrid exchange functional12 and
the non-local correlation functional of Lee, Yang and Parr
(LYP). Calculations using the triple zeta basis 6-311G*13 and a
smaller one comprising a Dunning double-zeta14 for all first
row atoms, and the Los Alamos ECP plus double zeta for the
heavier atoms15 (LanL2DZ) showed rather small differences
in predicted geometries, vibrational frequencies and relative
energies for some of the species chosen as test compounds.
Accordingly, we adopted the less expensive LanL2DZ basis
consistently throughout the calculations. Solvent effects
simulating THF as solvent were estimated using the polarized
continuum model of Tomasi.16,17 The program package
Gaussian 9418 was used throughout all the calculations.
6 (a) G. H. Posner and M. Hulce, Tetrahedron Lett., 1984, 25, 379;
(b) G. H. Posner, T. P. Kogan and M. Hulce, Tetrahedron Lett., 1984,
25, 383.
7 W. J. Pietro, M. M. Francl, W. J. Hehre, D. J. Defrees, J. A. Pople and
J. S. Binkley, J. Am. Chem. Soc., 1982, 104, 5039.
8 A. A. Ponaras and Y. Meah, Tetrahedron Lett., 1986, 27, 4953.
9 K. Sato, S. Inoue, T. Kifagawa and T. Takahashi, J. Org. Chem.,
1973, 38, 551.
10 (a) S. Mori and E. Nakamura, Chem. Eur. J., 1999, 5, 1534; (b)
J. K. Taylor, Organocopper Reagents, Oxford University Press,
Oxford, 1994.
11 J. M. Seminario (Ed.), Recent Developments and Applications of
Modern Density Functional Theory, in the series Theoretical and
Computational Chemistry, vol. 4, Elsevier, Amsterdam, 1996.
12 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
13 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 23, 785.
14 T. H. Dunning, Jr. and P. J. Hay, in Modern Theoretical Chemistry,
ed. H. F. Schäfer, III, Plenum, New York, 1976, pp. 1–28.
15 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82, 299.
16 S. Miertus, E. Scrocco and J. Tomasi, J. Chem. Phys., 1981, 55, 117.
17 S. Miertus and J. Tomasi, J. Chem. Phys., 1982, 56, 239.
18 Gaussian 94, Revision D.4, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
P. M. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman,
T. Keith, G. A. Peresson, J. A. Montgomery, K. Raghavachari,
M. A. Al-Laham, V. G. Zakrzewski, J. V. Ortiz, B. Foresman,
J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe,
C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres,
E. S. Replogle, R. Gomperts, R. L. Martin, D. J. Fox, J. S. Binkley,
D. J. Defrees, J. Baker, J. P. Stewart, M. Head-Gordon, C. Gonzalez
and J. A. Pople, Gaussian, Inc., Pittsburgh, PA, 1995.
19 D. D. Perrin and W. L. F. Armarago, Purification of Laboratory
Chemicals, 3rd edn., Pergamon Press, Oxford, 1988.
Acknowledgements
The authors wish to thank The Norwegian Research Council
for financial support, Professor Anne Skancke for performing
the calculations of the (LUMO ϩ 1) orbitals, and Mr Bjørn
Olsrød at the Department of Organic Chemistry, Norwegian
University of Science and Technology and Mrs Hilde Røise at
the Department of Chemistry, University of Oslo for obtaining
HRMS data.
References
1 P. Perlmutter, Conjugate Addition Reactions in Organic Chemistry,
Pergamon Press Ltd., Exeter, 1992.
2 (a) C. J. Kowalski, A. E. Weber and K. W. Fields, J. Org. Chem.,
1982, 47, 5088; (b) H. J. Klemeyer and L. A. Paquette, J. Org. Chem.,
Paper 9/05444H
1840
J. Chem. Soc., Perkin Trans. 2, 1999, 1835–1840