A new and highly effective organometallic approach to 1,2-dehalogenations and related reactions

Article abstract of DOI:10.1016/j.jorganchem.2007.05.039

We investigated the reductive elimination of several functionalized and non-functionalized vic-dibromides with 1,2-diphenyl-, 1,1,2,2-tetraphenyl- and 1-phenyl-2-(2-pyridyl)-1,2-disodioethane. The reaction, involving some of the less expensive organic and inorganic reagents, proceeds under mild conditions, and is tolerant of a variety of functional groups. Extension of this procedure to similar 1,2-disubstituted compounds was also investigated. Reductive eliminations run on stereochemical probe compounds strongly suggest that this reaction proceeds via a "single electron" reductive elimination reaction pathway.

Full text of DOI:10.1016/j.jorganchem.2007.05.039

Journal of Organometallic Chemistry 692 (2007) 3892–3900
www.elsevier.com/locate/jorganchem
A new and highly effective organometallic approach to
1,2-dehalogenations and related reactions
a,
a
a
a
b
Ugo Azzena *, Mario Pittalis , Giovanna Dettori , Luisa Pisano , Emanuela Azara
a
`
Dipartimento di Chimica, Universita di Sassari, via Vienna 2, I-07100 Sassari, Italy
Istituto di Chimica Biomolecolare del C.N.R., sez. di Sassari, traversa la Crucca 3, regione Baldinca, I-07100 Sassari, Italy
b
Received 19 April 2007; received in revised form 21 May 2007; accepted 21 May 2007
Available online 2 June 2007
Dedicated to Professor Miguel Yus on the occasion of his 60th birthday.
Abstract
We investigated the reductive elimination of several functionalized and non-functionalized vic-dibromides with 1,2-diphenyl-, 1,1,2,2-
tetraphenyl- and 1-phenyl-2-(2-pyridyl)-1,2-disodioethane. The reaction, involving some of the less expensive organic and inorganic
reagents, proceeds under mild conditions, and is tolerant of a variety of functional groups. Extension of this procedure to similar
1,2-disubstituted compounds was also investigated. Reductive eliminations run on stereochemical probe compounds strongly suggest
that this reaction proceeds via a ‘‘single electron’’ reductive elimination reaction pathway.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Sodium; vic-Diorganometals; Reduction; Elimination
1. Introduction
dehalogenation procedures to the abiotic degradation of
organohalide contaminants [9].
Reductive dehalogenation of vic-dihalides is a useful and
long time known reaction, and many reagents have been
developed to achieve this goal [1]. Indeed, halogenation/
dehalogenation of alkenes has been employed either to pro-
tect double bonds towards oxidations, reductions or elec-
trophilic additions [2], or as a useful way to achieve
alkene purification [3]. Furthermore, reductive dehalogen-
ation allows the employment of 1,2-dihaloethenes as syn-
thetic equivalents of acetylene in cycloaddition reactions
[4–6]. Recently, the dehalogenation procedure was success-
fully employed in the synthesis of isotopically labelled com-
pounds, like [1-14C]-E-isomers of fatty acids [7] and
[1-79Br]-1-bromo- or [1-37Cl]-1-chloro-2-fluoroethylene [8].
Finally, it is worth mentioning the application of reductive
We recently reported that reaction of 1,2-diarylethenes
with Na metal in dry THF allows the generation of a wide
array of 1,2-diaryl-1,2-disodioethanes, whose reaction with
1,3-dichloropropane led to a highly stereoselective synthe-
sis of trans-1,2-diaryl-substituted cyclopentanes [10]. At
odds with these results, reaction of 1,2-diaryl-1,2-disodio-
ethanes with 1,2-dihaloethanes affords the parent trans-
1,2-diarylethenes, probably with concurrent formation of
ethene [11]. Although the ability of delocalized polycarba-
nions to act either as nucleophiles or reducing agents
towards alkyl halides is well known [12], the latter reaction
was never synthetically exploited.
Following our interest in the development of efficient
alkali metal-mediated reaction procedures, and with the
aim to expand the synthetic utility of these interesting
vic-diorganometallics, we investigated the reductive
dehalogenation of several vic-disubstituted compounds
*
Corresponding author. Tel.: +39079229549; fax: +39079229559.
E-mail address: ugo@uniss.it (U. Azzena).
0022-328X/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2007.05.039

U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
3893
diorganometal can be stopped as soon as a faint red colour
persists in the reaction mixture.
Besides the desired unsaturated reaction products, reac-
tions run with the disodium-derivative 1a (or 1b) led to the
formation of trans-stilbene (or tetraphenylethene), as well
as to the formation of minor amounts of 1,2-diphenyle-
thane (or 1,1,2,2-tetraphenylethane), the product of pro-
tonation of 1a (or 1b). In a similar way, trans-stilbazole
is the main by-product of reductions run with the pyridine
derivative 1c, usually accompanied by minor amounts of
2-phenethylpyridine (i.e., the product of protonation of
1c); the relative basicity of these pyridine derivatives
allowed their easy separation form the reaction mixture
upon acid washings with 1 N HCl.
Scheme 1. Reductive eliminations of vic-disubstituted compounds. 1a,
Ar = Ph, R = H; 1b, Ar = R = Ph; 1c, Ar = 2-C5H4N, R = H; R1, R2,
X = see text and Tables 1–3.
with 1,2-diphenyl-, 1,1,2,2-tetraphenyl- and 1-phenyl-2-(2-
pyridyl)-1,2-disodioethane (Scheme 1). A preliminary com-
munication on this topic already appeared [13].
2. Results and discussion
2.1. Reductive elimination reactions
Results obtained in the reductive elimination of several
simple vic-disubstituted compounds (Chart 1) are reported
in Table 1.
0.1–0.2 M deep red solutions of 1,2-diaryl-1,2-disodioe-
thanes (1a–c) were obtained by the reaction of trans-stil-
bene, tetraphenylethene or trans-stilbazole, respectively,
with an excess of Na metal in dry THF, under Ar [10].
Yields of the diorganometals were determined by quench-
ing samples of the resulting solutions with D2O, followed
Reaction of meso-1,2-dibromo-1,2-diphenylethane (2a)
with 1.1 equiv. of dianion 1a led, within 10 min at 0 ꢁC,
to the recovery of a reaction mixture containing trans-stil-
bene (3a) as the major reaction product, as well as a minor
amount of 1,2-diphenylethane (Table 1, entry 1). Running
this reaction without removing the excess of the metal, to
check the feasibility of an operatively simpler reaction pro-
cedure, led to a comparable result.
Under similar reaction conditions, the dibromo deriva-
tive 2b (as a 9:1 erythro/threo mixture of diastereoisomers)
was quantitatively converted into trans-anetole (3b), which,
however, was inseparable from trans-stilbene (Table 1,
entry 2).
Therefore, dehalogenation of 2b was performed in the
presence of 1.2 equiv. of dianion 1c: after aqueous work
up and acid washings to separate basic derivatives, trans-
anetole (3b) was recovered in almost quantitative yield
(Table 1, entry 3) [15].
The procedure was successfully applied to vic-dichlo-
rides: in a first experiment (not reported in Table 1), treat-
ment of dianion 1a with 1 equiv. of 1,2-dichloroethane led
to the quantitative recovery of stilbene. According to this
preliminary finding, reduction of 1,2-dichlorododecane
1
by H NMR analysis of crude reaction mixtures, to deter-
mine the percentage of deuterium incorporation in the
arylmethyl positions. The same procedure was employed
to determine that solutions of organometals 1a–c can be
stored under Ar in a freezer, for at least one week, without
detrimental effects.
Solutions of 1 were decanted from excess metal immedi-
ately prior to use, a procedure allowing a simple way to
recycle the excess of the metal [14].
Reductive eliminations were carried out by adding a
solution of the appropriate vic-disubstituted compound 2
to a chilled solution of 1a–c, followed by stirring the reac-
tion mixture at the same temperature before aqueous work
up. Inverse addition of the organometallic reagent to the
substrate is usually equally effective. It is worth noting that
the last procedure is particularly effective for those reac-
tions requiring stoichiometric amounts of the reducing
agent. Indeed, under these conditions, addition of the
Chart 1. Vic-disubstituted compounds 2a–2f, and resulting alkenes 3a–3e.

3894
U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
Table 1
Reaction of 2-bromoethyl methyl ether with an excess of
dianion 1a led to the recovery of a relatively complex reac-
tion mixture containing, besides a major amount of stil-
bene (68% by GC/MS), several unidentified by-products.
Similar results were obtained reacting 2-bromoethyl methyl
ether with 1b (not reported in Table 1).
At variance with these preliminary results, satisfactory
results were obtained in reactions leading to the formation
of styrene derivatives (3a and 3d, Table 1, entries 5 and 6),
or involving a phenoxide as a leaving group (3e, Table 1,
entry 7) [16]. These results strongly suggest that either the
formation of a conjugated alkene or the presence of an effi-
cient oxygen-leaving group is necessary to obtain good
results in the reductive elimination of a-haloethers.
Table 2 reports a second set of results obtained in the
reduction of several vic-dibromides bearing an acidic pro-
ton (compounds 2g–2j, Chart 2), or leading to the forma-
tion of an acidic reaction product (2k, Chart 2).
Reduction of 10,11-dibromoundecanoic acid (2g) with
an excess of 1a, during 10 min at 0 ꢁC, followed by aqueous
work up and chromatographic purification of the reaction
product, led to the formation of 10-undecenoic acid (3f) in
80% yield (Table 2, entry 1).
Application of a similar procedure to diastereoisomeric
threo- and erythro-13,14-dibromodocosanoic acid (2h)
allowed us to shed more light on the stereochemical out-
come of our reductive dehalogenation procedure.
Reductive elimination of compounds 2a–2f
Entry Dianion
(equiv.)
Substrate
T
t
Product Yield
(%)a
(ꢁC) (min)
1
2
1a (1.1)
1a (1.2)
meso-2a
2b, erythro/
threo = 9:1
2b, erythro/
threo = 9:1
2c
0
0
10
10
3a
3b
92b,c
>95b
3
1c (1.2)
0
10
3b
91
4
5
6
7
1c (1.3)
1c (1.3)
1c (1.6)
1b (1.1)d
0
0
0
0
60
60
60
10
3c
3a
3d
3e
>95b
90
70
92
erythro-2d
2e
2f
a
b
c
Yields determined on isolated products, unless otherwise indicated.
As determined by 1H NMR of crude reaction mixture.
3a is the product of dehalogenation of meso-2a, and the product of
oxidation of 1a; 8% of 1,2-diphenylethane was also recovered.
d
Inverse addition, see text.
(2c) with dianion 1c afforded 1-dodecene (3c) in satisfac-
tory yields (Table 1, entry 4).
We next investigated the reduction of several simple
a-haloethers.
Table 2
Reductive eliminations of compounds 2g–2k
Entry Dianion Substrate
(equiv.)
T
t
Product
Yield
(%)a
(ꢁC) (min)
1
2
3
4
1a (2.0) 2g
0
0
10
10
3f
80
Indeed, reaction of threo-13,14-dibromodocosanoic
acid, threo-2h, with 1.6 equiv. of 1c at 0 ꢁC, afforded the
corresponding unsaturated carboxylic acid, 3g, as a Z/
E = 35:65 mixture of diastereoisomers (Table 2, entry 2),
whilst a similar reaction, run at ꢀ80 ꢁC, afforded the same
compound as a Z/E = 70:30 mixture of diastereoisomers
(Table 2, entry 3).
1c (1.6)
1c (1.6)
1c (1.6)
threo-2h
threo-2h
erythro-2h
3g, Z:E = 35:65b 90
3g, Z:E = 70:30b 90
ꢀ80 10
0
10
3g,
92
Z:E = < 5: > 95b
5
6
7
8
1a (2.0) 2i
0
0
10
10
3h
70
65
65
>95c
1a (2.0) threo-2j
1a (2.0) threo-2j
3i, Z:E = 70:30c
3i, Z:E = 80:20c
3j
ꢀ80 10
1c (1.6)
2k,
0
10
On the contrary, reductive dehalogenation of erythro-
13,14-dibromodocosanoic acid, erythro-2h, run at 0 ꢁC in
the presence of 1.6 equiv. of 1c, afforded almost exclusively
E-3g (Table 2, entry 4). The reliability of these stereochem-
ical results was assured by observing the stereochemical
Z:E = 23:77
a
Yields determined on isolated products, unless otherwise indicated.
As determined by GC-MS analyses of the correspondig methyl ester,
b
obtained by reacting crude reaction mixtures with diazomethane.
As determined by 1H NMR of crude reaction mixtures.
c
Chart 2. Functionalized vic-disubstituted compounds 2g–2k, and resulting elimination products 3f–3j.

U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
3895
Scheme 2. Synthesis of methyl 10-undecenyl ether, 4.
stability of acid Z-3g under the above reported reaction
conditions.
almost quantitative yield (Table 2, entry 8), thus suggesting
deprotonation at C(2) of the reaction product.
A similar procedure was successfully applied to the
reductive dehalogenation of dibromoalcohols: reaction of
10,11-dibromoundecan-1-ol (2i) with 2 equivs. of 1a led,
after aqueous work up, to the recovery of 10-undecen-1-
ol (3h) in 70% yield (Table 2, entry 5). The last reaction
can be successfully run without removing the excess of
the metal (not reported in Table 2).
Our procedure was successfully extended to several
vic-disubstituted compounds bearing a carbonyl substitu-
ent (Chart 3). These results are reported in Table 3.
Debromination of amide 2l with 1 equiv. of 1a, during
10 min at ꢀ80 ꢁC, followed by aqueous work up and chro-
matographic purification of the reaction product, led to the
A further debromination of alcohol 2i, run according to
the general procedure, was quenched with 2.2 equivs. of
CH3I thus allowing, after aqueous work up and acid wash-
ings with 1 N HCl, the recovery of methyl 10-undecenyl
ether (4) in 65% yield (Scheme 2).
Table 3
Reductive eliminations of compounds 2l–p
Entry Dianion
(equiv.)
Substrate
T
t
Product Yield
(%)a
(ꢁC)
(min)
Reaction of threo-3,4-dibromooctan-1-ol, threo-2j, with
2 equivs. of 1a afforded, after 10 min at 0 ꢁC, a 70:30
Z/E mixture of the corresponding unsaturated alcohol, 3i
(Table 2, entry 6). It is worth noting that a similar reaction
run at ꢀ80 ꢁC led to the formation of the unsaturated alco-
hol 3i, as a 80:20 Z/E mixture of diastereoisomers (Table 2,
entry 7). Comparable results were obtained running the
reductive debromination of threo-2j with the disodium
derivative of stilbazole (1c) (not reported in Table 2).
We finally investigated the dehalogenation of 1,2-dib-
romo-1-phenylethene (2k) (as a 23:77 Z/E mixture of dia-
stereoisomers); a first reaction, run with 1 equiv. of 1c,
led to the recovery of a reaction mixture containing an
almost 1:1 mixture of phenylacetylene (3j) and starting
material (not reported in Table 2). Increasing the relative
amount of 1c to 1.6 equivs. led to recover acetylene 3j in
1
2
3
4
5
6
7
8
9
1a (1.0)
1c (1.0)
2l
ꢀ80
ꢀ80
ꢀ80
ꢀ20
0
ꢀ40
ꢀ40
0
10
60
3k
3l
90
2m
2m
2m
2n
58b,c
80
1c (1.1)d
1b (1.1)d
1a (1.1)d
1a (1.1)d
1a (1.1)d
1b (1.1)d
1b (1.1)d
30e
30e
30e
30e
30e
30e
30e
3l
3l
70f
36b
90
3m
3m
3m
3m
3ng
2n
2o
90
2o
2pg
>95c
94
ꢀ80
a
b
c
Yields determined on isolated products, unless otherwise indicated.
As determined by 1H NMR of crude reaction mixture.
37% of 2m and 5% of alcohol 3h were also detected.
Inverse addition, see text.
d
e
Time needed to perform a slow addition of the reducing agent to the
substrate.
f
7% of alcohol 3h was also detected.
Z = Benzyloxycarbonyl.
g
Chart 3. Functionalized vic-disubstituted compounds 2l–2p, and resulting elimination products 3k–3n.

3896
U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
formation of amide 3k, in almost quantitative yield (Table
3, entry 1).
As an additional advantage, all reactions leading to the
formation of acid 3m were easily purified by acid-basic
washings.
We next investigated the reaction of 10,11-dibromoun-
decyl acetate (2m) with 1 equiv. of dianion 1c at ꢀ80 ꢁC:
under these reaction conditions, aqueous work up and acid
washing (1 N HCl) to separate basic derivatives, allowed to
recover, besides the desired unsaturated ester 3l and some
unreacted starting material, a minor amount of 11-unde-
cen-1-ol (3h) (Table 3, entry 2).
The above described procedure was successfully applied
to the regioselective deprotection of the 2-iodoethyl ester of
N-benzyloxycarbonylproline (2p). Indeed, inverse addition
of dianion 1b to a THF solution of compound 2p, chilled at
ꢀ20 ꢁC, allowed the recovery of Z-proline, 3n, in 62% iso-
lated yield (not reported in Table 3), whilst a similar reac-
tion, run at ꢀ80 ꢁC, afforded the desired product in almost
quantitative yield (Table 3, entry 9).
Interestingly, alcohol 3h was the main product (63%, as
1
determined by H NMR) of a reaction run in the presence
of 1.7 equivs. of the reducing agent (not reported in Table 3).
This behaviour can be rationalized by assuming that
dianion 1c is acting towards the dibromoester 2m, both
as a reducing agent, leading to the formation of dehalogen-
ated products, as well as a nucleophile towards the ester
moiety, eventually leading to the formation of alcohol 3h.
Accordingly, the desired unsaturated ester 3l was recovered
in 80% yield, by reacting dibromide 2m with 1.1 equiv. of
1c, under inverse addition reaction conditions at ꢀ80 ꢁC
(Table 3, entry 3).
Finally, we found it possible to selectively dehalogenate
ester 2m under milder reaction conditions. Indeed, the
reduction of 2m was successfully performed at ꢀ20 ꢁC,
with the sterically hindered, hence poorly nucleophilic,
dianion 1b. Under inverse addition reaction conditions,
the unsaturated ester 3l was recovered in 70% isolated yield
(Table 3, entry 4).
Interestingly, under the above reported reaction condi-
tions (Table 3, entries 2–4), the reductive dehalogenation
of ester 2m never led to the formation of the corresponding
acyloin, i.e., the product of the reaction of an alkyl (or aryl)
ester with Na metal in THF [17].
The relative stability of the ester group under the above
reported reaction conditions, as well as the possibility to
obtain the reductive elimination of vic-halide-oxygen
disubstituted compounds (Table 1, entries 5–7), prompted
us to investigate the reductive deprotection of b-haloethyl
esters [18].
The reaction of the 2-iodoethyl ester of benzoic acid (2n)
with 1.1 equivs. of dianion 1a was investigated as a function
of reaction temperature. A reaction run at 0 ꢁC, under
inverse addition reaction conditions, led to the recovery of
benzoic acid, 3m in modest yield, probably due to a compet-
itive nucleophilic reaction of the dianion with the ester group
(Table 3, entry 5). A much better result was obtained per-
forming the reaction at ꢀ40 ꢁC; indeed, under these reaction
conditions, the desired acid was recovered in 90% isolated
yield (Table 3, entry 6). A similar result was obtained per-
forming this reaction at ꢀ80 ꢁC (not reported in Table 3).
Besides the 2-iodoethyl ester 2n, also the less reactive
2-bromoethyl ester of benzoic acid [19], 2o, was efficiently
deprotected under the above reported reaction conditions;
indeed, this ester was almost quantitatively converted into
benzoic acid, 3m, either by reduction with the disodium deriv-
ative 1a, at ꢀ40 ꢁC (Table 3, entry 7), or by reduction with the
poorly nucleophilic dianion 1b, at 0 ꢁC (Table 3, entry 8).
2.2. Reaction mechanism
Alternative reductive dehalogenations mechanisms are
usually described as ‘‘single electron’’ or as ‘‘concerted two
electron’’ reaction pathways [20]. More recently, the ‘‘single
electron’’ pathway was described as an ‘‘outer-sphere’’ redox
process, whilst the ‘‘concerted two electron’’ pathway was
described as an ‘‘inner-sphere’’ redox process [1i].
As recently emphasized [9a], it is possible to discriminate
between these mechanisms by examining the reduction of a
series of stereochemical probe compounds. Indeed, a ‘‘con-
certed two electron’’ reaction pathway should proceed with
anti stereospecificity (E2-like process, also termed1i E2(R)
process), with erythro- and meso-starting materials leading
to the formation of E-reaction products, and dl- and threo-
compounds being reduced to the corresponding Z-alkenes.
On the contrary, ‘‘single electron’’ reductive elimina-
tions are characterized by a great variety of stereochemical
results. Indeed, the stereochemical outcome of this reduc-
tion pathway strongly depends on the relative configura-
tional stability of reaction intermediates, as well as on
their chemical behaviour [1b,1i,1l,9a]. Interestingly, reduc-
tive dehalogenations promoted by radical anions and car-
banions are usually considered to proceed via ‘‘single
electron’’ reaction pathways [1b], mostly in the case of car-
banions with a negative redox potential [1i].
From this point of view, we observed anti stereospecific-
ity in the reductive eliminations of meso-2a (Table 1, entry
1), erythro-2d (Table 1, entry 5) and erythro-2h (Table 2,
entry 4), and a lack of stereospecifity in the reductive
dehalogenations of 2b, employed as a 9:1 erythro/threo
mixture of diastereoisomers (Table 1, entries 2 and 3), as
well as of threo-2h (Table 2, entries 2 and 3) and threo-2j
(Table 2, entries 6 and 7).
According to these stereochemical findings, a ‘‘concerted
two electron’’ reaction pathway, i.e., an ‘‘inner-sphere’’
redox process, can be ruled out, and the actual mechanism
should involve an ‘‘outer-sphere’’ redox process, leading to
the formation of transient species behaving in a non-stereo-
specific way.
We therefore suggest that our reductive eliminations
proceed via a ‘‘single electron’’ reaction pathway, as
depicted in Eqs. (1)–(4).
A first single electron transfer (SET) from the dianion to
the vic-disubstituted substrate generates two radical anions

U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
3897
(Eq. (1)); the halide-substituted radical anion immediately
(or contemporary to the SET step [21]) loses a halide anion
leading to the formation of a radical (Eq. (2)); a second
SET, from the 1,2-diaryl-substituted radical anion to the
radical, afforded the 1,2-diarylethene and a b-substituted
carbanion (Eq. (3)); the last one is finally transformed into
the corresponding alkene (Eq. (4)).
reduction of stilbene dibromides led to the formation of
complex reaction mixtures. Other methods, involving the
employment of Na in liquid NH3 [26] or the K-graphite
intercalate C8K [27], appear less attractive form a practical
point of view.
Furthermore, depending on the nature of the substrate
to be reduced, the arylalkene employed to generate the
diorganometals can often be easily separated from the reac-
tion product by acid–base washings, thus avoiding labori-
ous work-up procedures.
It is also worth noting that our procedure is tolerant of
several functional groups (alcohols, carboxylic acids, alky-
nes, esters, amides and urethanes), and shows promising
applications as a new and selective protocol to be employed
in the reductive removal of protecting groups.
From this point of view, the deprotection of benzoic
acid 2-bromoethyl ester (2o) (Table 3, entries 7 and 8)
appears as a particularly interesting reaction. Indeed, it
has been reported that reductive removal of 2-bromoethyl
ester is a problematic reaction, and extremely slow and
inefficient when Zn, or even SmI2 [19], are employed as
reducing agents.
ð1Þ
ð2Þ
ð3Þ
ð4Þ
As already argued, our stereochemical results can be ratio-
nalized taking into account, besides a lack of stereospecific-
ity in reactions leading to the formation of both the
haloalkyl radical and carbanion, the possibility for these
intermediates to undergo rotameric relaxation [22], thus
favouring preferential or exclusive formation of E-reaction
products.
The results obtained running the reductive dehalogen-
ations of acid threo-2h (Table 2, entries 3 and 4) and alcohol
threo-2j (Table 2, entries 6 and 7) at different temperatures,
further support this hypothesis. In both cases, a higher per-
centage of the Z stereoisomer of the reaction product was
obtained running each reduction at the lower temperature
investigated, i.e., under conditions which should signifi-
cantly slow down rotameric relaxation of reactive interme-
diates [1l,9a].
As a conclusion, we can say that 1,2-diaryl-1,2-disodioe-
thanes can be considered as synthetic equivalents of an
activated, selective, and highly reactive form of Na metal,
functioning under homogeneous reaction conditions.
4. Experimental
4.1. General
Boiling and melting points are uncorrected; the air bath
temperature on bulb-to-bulb distillation are given as boil-
ing points. Starting materials were of the highest commer-
cial quality and were purified by distillation or
recrystallization immediately prior to use. THF was dis-
tilled from Na/K alloy under N2 immediately prior to
3. Concluding remarks
1
use. H NMR spectra were recorded at 300 MHz and 13C
Our results show that 1,2-diaryl-1,2-disodium ethanes
are effective reductive elimination reagents presenting sev-
eral useful and interesting features.
A first remark concerns their easy preparation from
readily available starting materials. Indeed, Na is by far
one of the less expensive inorganic reagents, whilst trans-
stilbene and tetraphenylethene are relatively inexpensive
organic starting materials. Concerning trans-stilbazole, it
can be prepared in a multigram scale via a Wittig-type reac-
tion, as already described [10].
NMR spectra were recorded at 75 MHz in CDCl3 (unless
otherwise indicated) with SiMe4 as internal standard on a
Varian VXR 300 spectrometer. Deuterium incorporation
was calculated by monitoring the 1H NMR spectra of
crude reaction mixtures, as described in Ref. [10]. IR spec-
tra were recorded on pure samples on a FT-IR Jacso 480
P. GC-MS spectra were measured at 70 eV on a Trace
GC-Polaris Q spectrometer. Flash chromatography was
performed on Merck silica gel 60 (40–63 lm), and TLC
analyses on Macherey–Nagel silica gel pre-coated plastic
sheets (0.20 mm). Elemental analyses were performed by
the microanalytical laboratory of the Dipartimento di Chi-
A comparison with known alkali metal-based reductive
elimination reagents, shows some of the advantages of our
procedure.
Indeed, bulk Li [23] or Na [24] metal in THF are usually
employed in large excess and for relatively long reaction
times, whilst our reagents are employed in stoichiometric
amounts (or in a slight excess), and drive reactions to com-
pletion within a very short reaction time, under mild condi-
tions. Na naphthalenide [1m,25] is a particularly effective
reagents towards aliphatic (or alicyclic) dihalides, but
`
mica, Universita di Sassari.
4.2. Starting materials
Starting materials 2 are either known compounds, pre-
pared according to standard procedures and characterized
according to the literature [28], or commercially available

3898
U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
(2a, 2f, 2o). The stereochemistry of starting materials was
assigned by comparison with commercially available sam-
ples and/or literature data [28]; the trans stereochemistry
of amide 2l was assigned by comparison with the stereo-
chemistry of the corresponding N-carboxamide and N-car-
bonyl chloride, prepared under similar reaction conditions
[29].
was added in one portion, and the resulting mixture was
stirred overnight at 0 ꢁC in the dark. The reaction was
quenched by the addition of oxalic acid (0.11 g, 1.2 mmol)
dissolved in THF (2 mL), stirred at rt for 30 min, filtered,
and the filter cake washed with CH2Cl2 (3 · 10 mL). The
combined organic phases were collected, washed with 1 N
HCl (2 · 10 ml), brine (10 mL), dried (MgSO4), and the sol-
vent evaporated. The resulting oil was purified by flash
chromatography (AcOEt/petroleum ether 6:4), to give the
pure product (2.37 g, 5.9 mmol, 74%), which was character-
ized as follows: light yellow oil; Rf = 0.65 (AcOEt/petro-
leum ether 6:4); IR (neat): 1748, 1707 cmꢀ1; dH mixture of
rotamers 1.80–2.20 (2H, m, CH2), 2.18–2.36 (1H, m, CH),
3.20–3.50 (5H, m, 2· CH2, CH), 4.18–4.41 (3H, m,
NCHCO, CH2O), 4.87–5.16 (2H, m, CH2Ph), 7.21–743
(5H, m, Ph); dC (mixture of rotamers) 2.4, 2.7, 23.1, 23.9,
29.5, 30.5, 46.2, 46.8, 58.4, 59.0, 64.7, 64.7, 66.1, 127.4,
127.5, 127.8, 127.9, 128.4, 128.5, 136.7, 136.9, 153.5,
154.0, 171.7, 172.0. Anal. Calc. for C15H18INO4: C, 44.68;
H, 4.50; N, 3.47. Found: C, 44.57; H, 4.76; N, 3.36%.
Other products were synthesized and characterized as
follows.
4.2.1. Synthesis of threo-3,4-dibromoocta-1-ol (threo-2j)
A solution of (Z)-3-octen-1-ol (2.00 g, 15.6 mmol) in dry
CCl4 (20 mL) was placed in a two-necked flask, equipped
with reflux condenser, CaCl2 tube and magnetic stirrer,
and chilled to ꢀ5 ꢁC. To this solution was added dropwise
a solution of Br2 (0.8 mL, 2.50 g, 15.6 mmol) dissolved in
CCl4 (2 mL). The mixture was stirred 30 min at ꢀ5 ꢁC, then
2 h at r.t. The reaction mixture was evaporated, and the
resulting brown oil was purified by vacuum distillation, to
give the pure product (3.00 g, 9.9 mmol, 63%), which was
characterized as follows: light yellow oil, b.p. 120–125 ꢁC/
1 mmHg; IR (neat): 3345 cmꢀ1; dH 0.93 (3H, t, J =
7.2 Hz, CH3), 1.24–2.38 (9H, m, 4· CH2, OH), 3.79–3.95
(2H, m, CH2O), 4.20–4.27 (1H, m, CHBr), 4.46–4.54 (1H,
m, CHBr); dC 13.9, 21.9, 29.8, 35.1, 37.8, 55.6, 59.6, 60.2.
Anal Calc. for C8H16Br2O: C, 33.36; H, 5.60. Found: C,
33.25; H, 5.79%.
4.3. Generation of diorganosodium reagents 1a–c
Deep red solutions (0.1 M) of 1a and 1c were prepared
by the reaction of freshly cut Na metal with trans-stilbene,
or trans-stilbazole, respectively in dry THF, as described in
Ref. [10]. These solutions were drained from excess metal
under an Ar atmosphere, immediately prior to use.
4.2.2. Synthesis of 10,11-dibromoundecyl acetate (2m)
A solution of 10,11-undecenyl acetate, 3l, (4.0 g,
18.8 mmol) in dry CCl4 (20 mL) was placed in a two-
necked flask, equipped with reflux condenser, CaCl2 tube
and magnetic stirrer, and chilled to ꢀ5 ꢁC. To this solution
was added dropwise a solution of Br2 (0.97 mL, 3.01 g,
18.8 mmol) dissolved in CCl4 (2 mL). The mixture was stir-
red 30 min at ꢀ5 ꢁC, then 40 min at r.t. The reaction mix-
ture was evaporated, and the resulting brown oil was
purified by vacuum distillation, to give the pure product
(5.0 g, 13.4 mmol, 72%), which was characterized as fol-
lows: light yellow oil, b.p. 150–155 ꢁC/1 mmHg; IR (neat):
1737 cmꢀ1; dH: 1.27–1.40 (14 H, m, 7· CH2), 1.70–1.84
(1H, m, CH), 2.05 (3H, s, CH3), 2.05–2.20 (1H, m, CH),
3.63 (1H, t, J = 9.9 Hz, CHBr), 3.85 (1H, dd, J = 10.2,
4.5 Hz, CHBr), 4.05 (2H, t, J = 6.6 Hz, CH2O), 4.12–4.22
(1H, m, CHBr); dC: 21.0, 25.8, 26.7, 28.5, 28.7, 29.1, 29.2,
29.3, 35.9, 36.3, 53.1, 64.6, 171.2. Anal. Calc. for
C13H24Br2O2: C, 41.96; H, 6.50. Found: C, 41.71; H,
6.79%.
4.3.1. Generation of 1,1,2,2-tetraphenyl-1,2-disodioethane
(1b)
Two to three pieces of freshly cut Na metal (0.14 g,
4.5 mg atoms) were placed under Ar in a 50 mL two-
necked flask equipped with reflux condenser and magnetic
stirrer, and were suspended in dry THF (10 mL). 1,1,2,2-
Tetraphenylethene (0.50 g, 1.5 mmol), dissolved dry THF
(5 mL), was added dropwise, and the resulting mixture
was vigorously stirred at rt during 4 h. Yield of dianion
1b (>95%) was determined by quenching an aliquot of
the resulting deep violet solution with D2O, and determin-
ing deuterium incorporation (tipically quantitative) as
described in Ref. [10]. The excess of the metal was with-
drawn, under an Ar atmosphere, immediately before use.
4.4. General reductive elimination procedure
To 10 ml of a 0.1 M solution of 1a–1c (1 mmol) chilled
at the temperature reported in Tables 1–3 was added a
solution of the appropriate vic-disubstituted compound 2
(0.8–0.5 mmol), dissolved in 3 mL of dry THF. Inverse
addition involves addition of the organometallic solution
to a chilled solution of compounds 2, under otherwise iden-
tical reaction conditions. After stirring for 10 min (except
when otherwise indicated), the mixture was quenched by
slow dropwise addition of H2O (15 mL), the cold bath
removed, and the resulting mixture extracted (3 · 10 mL)
with an organic solvent (Et2O for reactions run with
4.2.3. Synthesis of N-benzyloxycarbonylproline 2-iodoethyl
ester (2p)
A solution of Z-proline (2.00 g, 8.0 mmol) in dry CH2Cl2
(60 mL) was placed under N2 in a two-necked flask,
equipped with reflux condenser and magnetic stirrer, and
chilled to 0 ꢁC. To this solution, were added 2-iodoethanol
(0.75 ml, 1.65 g, 9.6 mmol) and pyridine (1.3 mL, 1.27 g,
16.6 mmol). After 10 min stirring, DCC (1.82 g, 8.8 mmol)

U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
3899
[6] For a review on synthetic equivalents of acetylene in cycloaddition
reactions, see: De O. Lucchi, G. Modena, Tetrahedron 40 (1984)
2585–2632.
[7] D. Georgin, F. Taran, C. Mioskowski, Chem. Phys. Lipids 125
(2003) 83–91.
[8] (a) A. Baldan, A. Tassan, Synth. Commun. 35 (2005) 1447–1453;
(b) A. Baldan, Synth. Commun. 34 (2004) 1229–1235.
[9] (a) L.A. Totten, U. Jans, A.L. Roberts, Environ. Sci. Technol. 35
(2001) 2268–2274, and references therein;
(b) E.C. Butler, K.F. Hayes, Environ. Sci. Technol. 34 (2000) 422–429.
[10] U. Azzena, G. Dettori, C. Lubinu, A. Mannu, L. Pisano, Tetrahedron
61 (2005) 8663–8668.
[11] (a) Ethene is a reasonable co-product of this reaction, but its
formation was never ascertained J.W.B. Reesor, J.G. Smith, G.F.
Wright, J. Org. Chem. 19 (1954) 940–956;
(b) U. Azzena, G. Dettori, M.V. Idini, L. Pisano, G. Sechi,
Tetrahedron 59 (2003) 7961–7966.
[12] (a) C.E. Barry III, R.B. Bates, W.A. Beavers, F.A. Camou, B.
Gordon III, H.F.-J. Hsu, N.S. Mills, C.A. Ogle, T.J. Siahaan, K.
Suvannachut, S.R. Taylor, J.J. White, K.M. Yager, Synlett (1991)
207–212, and references therein;
dianions 1a and 1 c; AcOEt for reactions run with dianion
1b). The organic phase was washed with brine (10 mL)
then, in case of reactions with 1 c, with 1 N HCl
(3 · 10 mL), dried (Na2SO4) and the solvent evaporated.
Crude products were purified by flash chromatography
(Petroleum ether/AcOEt). However, all products obtained
from reactions run with dianion 1c (i.e., 3a–d, 3g, 3j, 3l) as
well as phenol 3e and carboxylic acids 3f, 3m and 3n can be
efficiently purified by acid–base washings.
Reaction products 3 and 4 were recognized by compar-
ison with commercially available samples (3a–c, 3f, Z-3g,
3h, Z-3i, 3j) or with literature data [30]. When necessary,
the stereochemistry of reaction products was assigned by
comparison with commercially available samples and/or
literature data [30]. Diastereoisomeric mixtures of acid
3g, as obtained according to the general reductive elimina-
tion procedure, were dissolved in CH2Cl2 and treated with
diazomethane, according to the literature [31]; the resulting
mixtures, containing the corresponding methyl esters, were
analyzed by GC-MS to determine their diastereoisomeric
compositions.
(b) J.G. Smith, E. Oliver, T.J. Boettger, Organometallics 2 (1983)
1577–1582;
(c) J.G. Smith, J.R. Talvitie, A.R.E. Eix, J. Chem. Soc., Perkin
Trans. 1 (1975) 1474–1479.
[13] U. Azzena, M. Pittalis, G. Dettori, S. Madeddu, E. Azara, Tetrahe-
dron Lett. 47 (2006) 1055–1058.
[14] The residual metal is usually recovered as a single piece, which can be
washed with a dry solvent (Et2O or THF), weighed, and recycled to a
successive reaction.
Acknowledgement
`
Financial support from the Universita di Sassari (Fondo
[15] Accordingly, trans-stilbazole can be recovered and purified by flash-
chromatography (AcOEt/Petroleum Ether = 4:6) in 60–65% yield.
[16] (a) 2-Allyloxyphenol, 3e, is a pyrocathecol monoprotected with a
group known to be unstable under reductive electron transfer reaction
conditions. J.J. Eisch, A.M. Jacobs, J. Org. Chem. 28 (1963) 2145–
2146;
(b) E. Alonso, D.J. Ramon, M. Yus, Tetrahedron 53 (1997) 14355–
14368.
[17] M. Makosza, K. Grela, Synlett (1997) 267–268, and references
therein.
[18] T.W. Greene, P.G. Wuts, Protective Groups in Organic Synthesis,
third ed., Wiley, New York, 1999, pp. 397–398.
[19] A.J. Pearson, K. Lee, J. Org. Chem. 59 (1994) 2257–2260, and
references therein.
[20] See Ref. [1b] for a comparison between these mechanisms. See Ref.
[1i] for an in-depth discussion on competitive ‘‘inner-‘‘ and ‘‘outer-
sphere’’ redox processes.
di Ateneo) is gratefully acknowledged.
References
[1] (a) For a list of reagents employed in reductive dehalogenation, see:
R.C. Larock, Comprehensive Organic Transformation, VCH Pub-
lishers, New York, 1989, pp. 133–135;
(b) For a review, see: E. Baciocchi, in: S. Patai, Z. Rappoport (Eds.),
The Chemistry of Functional Groups, Supplement D, Part I, Wiley,
New York, 1983, pp. 161–201;
´
(c) For a selection of relevant literature, see: J.M. Concellon, H.
´
Rodrıguez-Solla, Chem. Soc. Rev. 33 (2004) 599–609;
(d) B.C. Ranu, A. Das, A. Hajra, Synthesis (2003) 1012–1014;
(e) T.S. Butcher, M.R. Detty, J. Org. Chem. 63 (1998) 177–180;
(f) Y. Takagughi, A. Hosokawa, S. Yamada, J. Motoyoshiya, H.
Aoyama, J. Chem. Soc., Perkin Trans. 1 (1998) 3147–3149;
(g) C. Malanga, S. Mannucci, L. Lardicci, Tetrahedron 54 (1998)
1021–1028;
(h) R. Yanada, N. Negoro, K. Yanada, T. Fujita, Tetrahedron Lett.
37 (1996) 9313–9316;
(i) T. Lund, C. Bjørn, H.S. Hansen, A.K. Jensen, T.K. Thorsen,
Acta Chem. Scand. 47 (1993) 877–884;
(j) D. Landini, L. Milesi, M.L. Quadri, F. Rolla, J. Org. Chem. 49
(1984) 152–153;
(k) D. Savoia, E. Tagliavini, C. Trombini, A. Umani-Ronchi, J. Org.
Chem. 47 (1982) 876–879;
(l) J.F. Garst, J.A. Pacifici, V.D. Singleton, M.F. Ezzel, J.I. Morris,
J. Am. Chem. Soc. 97 (1975) 5242;
´
[21] C.P. Andrieux, A. Le Gorande, J.-M. Saveant, J. Am. Chem. Soc.
114 (1992) 6892–6904, and references therein.
[22] a-Bromine-substituted radical intermediates are considered to be
configurationally stabilized by bromine bridging. See Ref. [9a].
[23] J. Sicher, M. Havel, M. Svoboda, Tetrahedron Lett. 9 (1968) 4269–
4272.
[24] Na in THF: H.O. House, R.S. Ro, J. Am. Chem. Soc. 80 (1958) 182–
187.
[25] Na naphthalenide in 1,2-dimethoxyethane H.R. Pfaendler, F.X.
Muller, Synthesis (1992) 350–352, and references therein.
¨
[26] Na in liq. NH3: E.L. Allred, B.R. Beck, K.J. Voorhees, J. Org. Chem.
39 (1974) 1426–1427.
[27] K graphite, C8K: M. Rabinovitz, B. Tamarkin, Synth. Commun. 14
(1984) 377–379.
[28] (a) Literature references to starting materials: 2b: R.C. Fahey, H.-J.
Schneider, J. Am. Chem. Soc. 90 (1968) 4429–4434;
(b) 2c: S.J. Lapporte, L.L. Ferstandig, J. Org. Chem. 26 (1961) 3681–
3685;
(m) W. Adam, J. Arce, J. Org. Chem. 37 (1972) 507–508.
[2] D.W. Young, in: J.F.W. McOmie (Ed.), Protective Groups in
Organic Chemistry, Plenum Press, London, 1973, pp. 309–320.
[3] See, for example: L.F. Fieser, M. Fieser, Advanced Organic Chemistry,
Reinhold Publishing Corporation, New York, 1961, pp. 163–165.
[4] [2+2] cycloadditions: H.O. House, T.H. Cronin, J. Org. Chem. 30
(1965) 1061–1070.
(c) 2d: see Ref. [24];
(d) 2e: S.D. Higgins, C.B. Thomas, J. Chem. Soc. Perkin Trans. 1
(1982) 235–242;
[5] [4+2] cycloadditions: S.J. Cristol, N.L. Hause, J. Am. Chem. Soc. 74
(1952) 2193–2197.

3900
U. Azzena et al. / Journal of Organometallic Chemistry 692 (2007) 3892–3900
(e) 2g: P. Vinczer, G. Baan, Z. Juvancz, L. Novak, C. Szantay, C.
Synth. Commun. 15 (1985) 1257–1270;
(f) threo- and erythro-2h: T. Maruyama, Proc. Imp. Acad. 10 (1934)
467–469, Chem. Zentralbl. 106 (1935) 2358; Chem. Abstr. 29 (1935)
[30] (a) Literature references to reaction products: 3d: S.D. Higgins, B.C.
Thomas, J. Chem. Soc., Perkin Trans. 1 (1982) 235–242;
(b) 3e: G. Foulard, T. Brigaud, C. Portella, Tetrahedron 52 (1996)
6187–6200;
1062;
(c) (E)-3g: G. Rankoff, Chem. Ber. 63 (1930) 2139–2142;
(d) (E)-3i: J.R. Vyvyan, C.L. Hoist, A.J. Johnson, C.M. Schwenk, J.
Org. Chem. 67 (2002) 2263–2265;
(g) 2i: C.E. Bishop, G.W. Morrow, J. Org. Chem. 48 (1983)
657–660;
(h) 2k: J.K. Espenson, Z. Zhu, T.H. Zauche, J .Org. Chem. 64 (1999)
1191–1196;
(e) 3k: T. Ohta, N. Miyata, M. Hirobe, Chem. Pharm. Bull. 29 (1981)
1221–1230;
(i) 2l: B.P. Das, D.W. Boykin, J. Med. Chem. 14 (1971) 56–58;
(j) 2n: F. Foubelo, F. Lloret, M. Yus, Tetrahedron 50 (1994) 5131–
5138.
(f) 3l: C.S. Marvel, W.E. Garrison, J. Am. Chem. Soc. 81 (1959)
4737–4744;
(g) 4: C.D. Hurd, W.H. Saunders, J. Am. Chem. Soc. 74 (1952) 5324–
5329.
[29] G. Bellucci, C. Chiappe, F. Marioni, F. Marchetti, J. Chem. Soc.,
Perkin Trans. 2 (1992) 637–642, and references therein.
[31] P. Lombardi, Chem. Ind. London (1990) 708.