Catalytic reduction of phenyl-conjugated acetylenic halides by nickel(I) salen: Cyclization versus coupling

Article abstract of DOI:10.1002/ejoc.200700655

Cyclic voltammetry and controlled-potential electrolysis were employed to study the catalytic reduction of five phenyl-conjugated haloalkynes by nickel(I) salen electro-generated at carbon cathodes in dimethylformamide containing tetramethylammonium tetrafluoroborate. Electrocatalytic reduction of 7-bromo- and 7-iodo-1-phenyl-1-heptyne affords the carbocyclic product, benzylidenecyclohexane, in up to 41 % yield, whereas under similar conditions reduction of 5-halo-1-phenyl-1-pentyne and 8-bromo-1-phenyl-1-octyne gives benzylidenecyclobutane and benzylidenecycloheptane, respectively, in very low yield (≤1 %). Dimers, alkynes, and alkenynes are other products formed from the phenyl-conjugated haloalkynes. Dimers (diphenylalkadiynes) derived from 5-halo-1-phenyl-1-pentyne and 8-bromo-1-phenyl-1-octyne are obtained in yields ranging from 85 to 93 %, whereas 1,14-diphenyltetradeca-1,13-diyne (the dimer produced from 7-halo-1-phenyl-1-heptyne) is found in yields of 45-51 %. To account for the formation of the various products, a mechanistic scheme that involves phenyl-conjugated alkynyl radicals arising from nickel(I) salen catalyzed cleavage of the carbon-halogen bond of each substrate was formulated. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200700655

FULL PAPER
DOI: 10.1002/ejoc.200700655
Catalytic Reduction of Phenyl-Conjugated Acetylenic Halides by Nickel(I)
Salen: Cyclization versus Coupling
Mohammad S. Mubarak,*[a] Theodore B. Jennermann,[b] Michael A. Ischay,[b] and
Dennis G. Peters*[b]
Keywords: Catalysis / Nickel / Acetylenic halides / Intramolecular cyclization / Radical coupling
Cyclic voltammetry and controlled-potential electrolysis
were employed to study the catalytic reduction of five
phenyl-conjugated haloalkynes by nickel(I) salen electro-
generated at carbon cathodes in dimethylformamide contain-
ing tetramethylammonium tetrafluoroborate. Electrocatalytic
reduction of 7-bromo- and 7-iodo-1-phenyl-1-heptyne af-
fords the carbocyclic product, benzylidenecyclohexane, in up
to 41% yield, whereas under similar conditions reduction of
5-halo-1-phenyl-1-pentyne and 8-bromo-1-phenyl-1-octyne
conjugated haloalkynes. Dimers (diphenylalkadiynes) de-
rived from 5-halo-1-phenyl-1-pentyne and 8-bromo-1-
phenyl-1-octyne are obtained in yields ranging from 85 to
93%, whereas 1,14-diphenyltetradeca-1,13-diyne (the dimer
produced from 7-halo-1-phenyl-1-heptyne) is found in yields
of 45–51%. To account for the formation of the various prod-
ucts, a mechanistic scheme that involves phenyl-conjugated
alkynyl radicals arising from nickel(I) salen catalyzed cleav-
age of the carbon–halogen bond of each substrate was for-
gives benzylidenecyclobutane and benzylidenecyclohep- mulated.
tane, respectively, in very low yield (Յ1%). Dimers, alkynes,
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
and alkenynes are other products formed from the phenyl-
Germany, 2007)
carbocycle in a higher yield than can be obtained by direct
electrolysis. For example, electrogenerated nickel(I) salen
catalytically reduced phenyl-conjugated haloalkynes such as
6-bromo- and 6-iodo-1-phenyl-1-hexyne to give benzyl-
idenecyclopentane in up to 95% yield,[18] whereas direct re-
duction of the same compounds at a mercury pool or retic-
ulated vitreous carbon cathode produced the carbocycle in
maximum yields of only 24[14] and 61%,[17] respectively,
mainly as a consequence of the formation of undesired acy-
clic side products such as alkynes and alkenynes.
Introduction
Reductive intramolecular cyclization of acetylenic ha-
lides has been a subject of long-standing mechanistic and
synthetic interest to organic chemists.[1–11] An excellent il-
lustration of this process is provided by 6-bromo-1-phenyl-
1-hexyne which, when treated with tri-n-butyltin hydride
(Bu3SnH) in the presence of a trace amount of 2,2Ј-
azobis(2-methylpropionitrile) (AIBN) in refluxing benzene
for 36 h, affords benzylidenecyclopentane in essentially
quantitative yield.[4] As a complement to earlier research
on the chemically promoted intramolecular cyclization of
acetylenic halides, we demonstrated in our laboratory that
direct electrochemical reduction of such compounds at both
mercury and carbon cathodes results in cleavage of the car-
bon–halogen bond, which produces an alkyl radical that
cyclizes to form an ylidenecycloalkane.[12–17]
In subsequent work, [{2,2Ј-[1,2-ethanediylbis(nitrilo-
methylidyne)]bis(phenolato)}(N,NЈ,O,OЈ)]nickelate(I), bet-
ter known as nickel(I) salen, electrogenerated at a glassy
carbon electrode in dimethylformamide containing tetra-
alkylammonium salts, was shown to be a useful catalyst for
the reductive intramolecular cyclization of haloal-
kynes.[18,19] Indeed the catalytic method gives the desired
In a recent investigation of the nickel(I) salen catalyzed
reduction of a family of non-phenyl-conjugated acetylenic
halides,[19] it was found that five-membered carbocycles
were formed in higher yield than were the related six- and
four-membered carbocycles. In quantitative terms, 1-bro-
mo- and 1-iodo-5-decyne afforded pentylidenecyclopentane
in yields as high as 75 and 86%, respectively, whereas pen-
tylidenecyclohexane was obtained in a maximum yield of
only 6% from 11-bromo- and 11-iodo-5-undecyne, and pen-
tylidenecyclobutane was formed in no more than 2% yield
from either 1-bromo- or 1-iodo-4-nonyne. This trend in
yields is a reflection of the difference between the stability
of the cyclized radical precursor (phenylmethylenylcyclo-
alkane) of each carbocycle (before abstraction of a hydro-
gen atom from the solvent to form the final product) and
the stability of the corresponding straight-chain radical
[a] Department of Chemistry, The University of Jordan,
Amman 11942, Jordan
·
E-mail: mmubarak@ju.edu.jo
[C4H9CϵC(CH2)n , where n = 3, 4, or 5] formed initially by
[b] Department of Chemistry, Indiana University,
Bloomington, Indiana 47405, USA
E-mail: peters@indiana.edu
the nickel(I) salen-promoted cleavage of the carbon–halo-
gen bond of each parent acetylenic halide.
5346
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5346–5352

Catalytic Reduction by Nickel(I) Salen: Cyclization vs. Coupling
In the present work, cyclic voltammetry and controlled-
potential electrolysis were employed to investigate further
the extent of reductive intramolecular cyclization of five dif-
ferent phenyl-conjugated acetylenic halides when electro-
generated nickel(I) salen was employed as the catalyst. For
the catalytic reductions of 7-bromo- and 7-iodo-1-phenyl-
1-heptyne, a six-membered carbocycle (benzylidenecyclo-
hexane) was obtained in up to 41% yield. In contrast, when
8-bromo-1-phenyl-1-octyne, 5-bromo-1-phenyl-1-pentyne,
and 5-iodo-1-phenyl-1-pentyne were catalytically reduced,
diphenylalkadiynes (dimeric products) were formed in up
to 93% yield, whereas the desired carbocyclic compounds
were found only in very small amounts.
Results and Discussion
Cyclic Voltammetric Behavior of Phenyl-Conjugated
Haloalkynes
Cyclic voltammograms were recorded at 100 mVs–1 for
the direct reduction at a glassy carbon disk electrode of
separate 2.0 m solutions of 5-bromo-1-phenyl-1-pentyne
(1), 5-iodo-1-phenyl-1-pentyne (2), 7-bromo-1-phenyl-1-
heptyne (3), 7-iodo-1-phenyl-1-heptyne (4), and 8-bromo-1-
phenyl-1-octyne (5) in DMF containing 0.10  TMABF4
under an argon atmosphere; representative cyclic voltam-
mograms for 1 and 2 are depicted in Figure 1.
Figure 1. Cyclic voltammograms recorded with a glassy carbon
disk electrode (area = 0.077 cm2) at 100 mVs–1 in air-free DMF
containing 0.10  TMABF4 and (A) 2.0 m 5-bromo-1-phenyl-1-
pentyne and (B) 2.0 m 5-iodo-1-phenyl-1-pentyne. Scans go from
–0.41 to –2.20 to –0.41 V.
As seen in curve A of Figure 1, the three closely spaced
irreversible waves are associated with the reduction of 1.
First, the shoulder at approximately –1.6 V is attributable
to the two-electron cleavage of the carbon–bromine bond.
Second, on the basis of previous research in our labora-
tory,[20–22] the small peak at –1.85 V is undoubtedly due to
reduction of 1-phenyl-1,2-pentadiene, which is formed in
the diffusion layer by electrolytically induced, base-pro-
moted isomerization of 1-phenyl-1-pentyne derived from
the reduction of 1. Third, the prominent peak at –2.05 V is
due to the reduction of 1-phenyl-1-pentyne itself. Addition-
ally, in situations where reduction of the phenyl-conjugated
haloalkyne leads to substantial intramolecular cyclization
to form a carbocyclic product Ϫ as is true for the reduction
of 3 to afford benzylidenecyclohexane Ϫ another peak ap-
pears at –2.18 V that is caused by reduction of the carbocy-
cle. In other respects, cyclic voltammograms for the re-
duction of 3 and 5 are virtually identical to that of 1.
Curve B of Figure 1 shows a cyclic voltammogram for
the reduction of 2. At –1.40 V is a peak that can be assigned
to irreversible two-electron scission of the carbon–iodine
bond; this observation is in close agreement with results
obtained previously under comparable experimental condi-
cleavage of the carbon–iodine bond of 2 occurs to form 1-
phenyl-1-pentyne at less negative potentials, so the alkyne-
to-allene transformation has more time to take place; how-
ever, some of the allene leaves the diffusion layer without
being reduced, which lowers the current. As expected, curve
B of Figure 1 shows a large irreversible cathodic peak at
–2.05 V that can be ascribed to the reduction of 1-phenyl-
1-pentyne. Not surprisingly, a cyclic voltammogram for a
solution of 4 in DMF/0.10  TMABF4 looks almost exactly
like curve B of Figure 1, except that electrolysis of the par-
ent haloacetylene gives rise (through reductive intramolecu-
lar cyclization) to benzylidenecyclohexane, which exhibits a
cathodic peak at –2.18 V.
Cyclic Voltammetric Behavior of Nickel(II) Salen in the
Absence and Presence of a Phenyl-Conjugated Haloalkyne
Displayed in Figures 2 and 3 are cyclic voltammograms
tions for the reduction of 1-iodo-4-nonyne,[19] 1-iodo-5- recorded with a glassy carbon electrode at a scan rate of
decyne,[19]
11-iodo-5-undecyne,[19]
6-iodo-1-phenyl-1- 100 mVs–1 for reduction of 2.0 m nickel(II) salen in DMF
hexyne,[18] and 1-iodo-5-decyne.[16] A barely discernible containing 0.10  TMABF4 in the absence and presence of
shoulder at –1.85 V signals the reduction of 1-phenyl-1,2- a haloalkyne. We show just the region of potentials where
pentadiene, which again arises from the base-catalyzed nickel(II) salen is electroactive. In the absence of a haloal-
isomerization of 1-phenyl-1-pentyne. In this instance, the kyne, the reversible one-electron nickel(II) salen–nickel(I)
current for reduction of the allene is very small because, salen couple is observed (Figures 2 and 3, curves A), for
unlike the situation for 1 (curve A, Figure 1), reductive which the cathodic peak potential (Epc) is –0.95 V, the an-
Eur. J. Org. Chem. 2007, 5346–5352
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5347

M. S. Mubarak, T. B. Jennermann, M. A. Ischay, D. G. Peters
FULL PAPER
odic peak potential (Epa) is –0.86 V, and the peak-current the catalytic reduction of 20.0 m 3, whereas Ipc is 220 µA
ratio (Ipa/Ipc) is close to unity. These observations are con- for the catalytic reduction of 20.0 m 4. Another feature
sistent with the results of a recent study.[19]
of the cyclic voltammograms (Figures 2 and 3) is that the
cathodic peak currents do not increase in proportion to the
concentration of added substrate. This latter behavior, seen
in a number of earlier studies,[18,19,23–25] has been linked to
sluggish regeneration of the precursor of the active catalyst.
In at least some instances, evidence was obtained to suggest
that this loss in catalytic efficiency of the nickel(II) salen–
nickel(I) salen system may be caused by alkylation of the
imino (C=N) bonds of the ligand by an alkyl radical de-
rived from the reductive cleavage of a carbon–halogen bond
of a substrate.[26,27]
Controlled-Potential (Bulk) Electrolyses of Nickel(II) Salen
in the Presence of a Phenyl-Conjugated Haloalkyne
Figure 2. Cyclic voltammograms recorded with a glassy carbon
disk electrode (area = 0.077 cm2) at 100 mVs–1 in air-free DMF
containing 0.10  TMABF4 and (A) 2.0 m nickel(II) salen, (B)
2.0 m nickel(II) salen and 5.0 m 7-bromo-1-phenyl-1-heptyne,
Bulk electrolyses were performed at reticulated vitreous
carbon cathodes in DMF containing 0.10  TMABF4,
(C) 2.0 m nickel(II) salen and 10.0 m 7-bromo-1-phenyl-1-hep- 2.0 m nickel(II) salen, and two different concentrations of
tyne, and (D) 2.0 m nickel(II) salen and 20.0 m 7-bromo-1-
phenyl-1-heptyne. Scans go from –0.45 to –1.13 to –0.45 V.
each haloalkyne; the cathode potential was held at –1.00 V,
which is significantly more positive than that for the direct
reduction of any of the haloalkynes. Listed in Table 1 are
the coulometric n values and product distributions obtained
from these experiments; each entry represents the average
of two separate runs. Coulometric n values (based on the
number of electrons transferred to each molecule of haloal-
kyne) were reproducible to Ϯ0.05; because n is essentially
1, the nickel(I) salen catalyzed reduction of each haloalkyne
must involve one-electron cleavage of a carbon–halogen
bond to give a reactive primary alkyl radical. Yields of indi-
vidual products were in agreement to Ϯ3% absolute, and
none of the parent haloalkyne remained unreduced; within
experimental error, the total yield of all products was close
to 100%, which indicates that we have successfully ac-
counted for all of the original substrate.
As revealed in Table 1, four products were obtained from
the catalytic reduction of each phenyl-conjugated haloal-
Figure 3. Cyclic voltammograms recorded with a glassy carbon
disk electrode (area = 0.077 cm2) at 100 mVs–1 in air-free DMF
containing 0.10  TMABF4 and (A) 2.0 m nickel(II) salen, (B) kyne. Entries 1–4 show that the principal product (Ͼ85%)
2.0 m nickel(II) salen and 5.0 m 7-iodo-1-phenyl-1-heptyne, (C)
2.0 m nickel(II) salen and 10.0 m 7-iodo-1-phenyl-1-heptyne,
and (D) 2.0 m nickel(II) salen and 20.0 m 7-iodo-1-phenyl-1-
heptyne. Scans go from –0.45 to –1.13 to –0.45 V.
derived from both 5-bromo- and 5-iodo-1-phenyl-1-pentyne
was
a
dimeric species (1,10-diphenyldeca-1,9-diyne),
whereas the desired carbocyclic compound (benzylidenecy-
clobutane) was detected in no higher than 1% yield, along
When a haloalkyne is added to the system, the cyclic volt- with small amounts of 1-phenyl-1-pentyne (5–8%) and 1-
ammetric behavior of nickel(II) salen is changed signifi- phenyl-4-penten-1-yne (2–3%).
cantly. As shown in curves B–D of Figures 2 and 3, when a
In contrast to the preceding results, the catalytic re-
stoichiometric excess of either 3 or 4 is combined with nick- ductions of 7-bromo- and 7-iodo-1-phenyl-1-heptyne af-
el(II) salen, the cathodic peak current corresponding to re- forded benzylidenecyclohexane in yields ranging from 33–
duction of nickel(II) salen increases and the anodic current 41% as well as 1,14-diphenyltetradeca-1,13-diyne (45–
due to oxidation of nickel(I) salen disappears; both features 51%), 1-phenyl-1-heptyne (6–9%), and 1-phenyl-6-hepten-
are predictable characteristics of the nickel(I) salen cata- 1-yne (3–6%), as seen from entries 5–8 of Table 1. On the
lyzed reduction of a haloalkyne. Notice that the enhance- basis of the product distributions obtained from 5-halo-1-
ment in the cathodic peak current is more pronounced with phenyl-1-pentynes and 7-halo-1-phenyl-1-heptynes, it is ob-
4 than with 3 owing to the fact that the catalyst–substrate vious that intramolecular cyclization of the catalytically
reaction is faster with the iodoalkyne than with the bromo- formed 1-phenyl-1-heptyn-7-yl radical is greatly favored
alkyne; for example, when cathodic peak currents are com- from stereochemical considerations over ring closure of the
pared for curve D in both Figures 2 and 3, Ipc is 83 µA for 1-phenyl-1-pentyn-5-yl radical.
5348
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5346–5352

Catalytic Reduction by Nickel(I) Salen: Cyclization vs. Coupling
Table 1. Coulometric data and product distributions for catalytic reduction of phenyl-conjugated haloalkynes by nickel(I) salen electro-
generated at reticulated vitreous carbon cathodes held at –1.00 V in DMF containing 0.10  TMABF4.
Entry
Haloalkyne [m][b]
n[c]
Product distribution [%][a]
6
7
8
9
Total
1
2
3
4
5
6
7
8
9
5-bromo-1-phenyl-1-pentyne, 1 (10)
5-bromo-1-phenyl-1-pentyne, 1 (20)
5-iodo-1-phenyl-1-pentyne, 2 (10)
5-iodo-1-phenyl-1-pentyne, 2 (20)
7-bromo-1-phenyl-1-heptyne, 3 (10)
7-bromo-1-phenyl-1-heptyne, 3 (20)
7-iodo-1-phenyl-1-heptyne, 4 (10)
7-iodo-1-phenyl-1-heptyne, 4 (20)
8-bromo-1-phenyl-1-octyne, 5 (10)
8-bromo-1-phenyl-1-octyne, 5 (20)
1.01
0.98
1.02
0.98
1.02
0.96
0.99
0.95
1.00
1.03
Ͻ1[d]
93[g]
88[g]
85[g]
87[g]
45[h]
49[h]
51[h]
48[h]
86[i]
81[i]
8[j]
2[m]
3[m]
2[m]
3[m]
4[n]
4[n]
3[n]
6[n]
2[o]
2[o]
103
98
94
96
92
93
97
104
95
94
1[d]
6[j]
Ͻ1[d]
1[d]
7[j]
5[j]
37[e]
33[e]
35[e]
41[e]
Ͻ1[f]
1[f]
6[k]
7[k]
8[k]
9[k]
7[l]
10
10[l]
6 = carbocycle (benzylidenecycloalkane); 7 = dimer (diphenylalkadiyne); 8 = 1-phenyl-1-alkyne; 9 = 1-phenylalken-1-yne.
[a] % = Yield expressed as the percentage of original haloalkyne incorporated into each product. [b] Each number in parentheses is the
millimolar concentration of haloalkyne. [c] Number of electrons per molecule of haloalkyne. [d] Benzylidenecyclobutane. [e] Benzyl-
idenecyclohexane. [f] Benzylidenecycloheptane. [g] 1,10-Diphenyldeca-1,9-diyne. [h] 1,14-Diphenyltetradeca-1,13-diyne. [i] 1,16-Di-
phenylhexadeca-1,15-diyne. [j] 1-Phenyl-1-pentyne. [k] 1-Phenyl-1-heptyne. [l] 1-Phenyl-1-octyne. [m] 1-Phenyl-4-penten-1-yne. [n] 1-
Phenyl-6-hepten-1-yne. [o] 1-Phenyl-7-octen-1-yne.
For the catalytic reduction of 8-bromo-1-phenyl-1-octyne idenecyclohexane in 35% yield, but attempts to obtain
(Table 1, Entries 9 and 10), the product distribution was benzylidenecycloheptane from the Grignard reagent of 8-
quite similar to those for the analogous 5-halo-1-phenyl-1- iodo-1-phenyl-1-octyne gave only 1-phenyl-1-octyne. By
pentynes; the dimeric species (1,16-diphenylhexadeca-1,15- employing chromium(II) in DMF containing ethylenedi-
diyne) was the major compound obtained (81–86%) and amine as a reducing agent, Crandall and Michaely[8] re-
only small quantities of benzylidenecycloheptane (1%), 1- ported that no carbocycle could be obtained from either 5-
phenyl-1-octyne (7–10%), and 1-phenyl-7-octen-1-yne (2%) bromo-1-phenyl-1-pentyne or 8-iodo-1-phenyl-1-octyne;
were found.
however, depending on the order in which the reactants
were combined, benzylidenecyclohexane was formed in at
least 79% yield from both 7-bromo- and 7-iodo-1-phenyl-
1-heptyne. Treatment of 5-bromo-1-phenyl-1-pentyne and
7-iodo-1-phenyl-1-heptyne with lithium di-n-butylcuprate
was found to afford benzylidenecyclobutane and benzyl-
idenecyclohexane in yields up to 79 and 58%, respectively,
but no carbocycle was obtained from 8-iodo-1-phenyl-1-oc-
tyne.[7]
In general, when compared with Bu3SnH/AIBN, which is
the most common system employed to effect intramolecular
cyclizations based on the intermediacy of alkyl radicals,
electrogenerated nickel(I) salen seems to be a viable alterna-
tive. Among the advantages of the latter reagent are that
electrosyntheses can be conducted at room temperature and
in relatively short time (Ͻ1 h) and that the environmental
nuisance associated with the recovery and disposal of toxic
tin-containing residues is eliminated. Another virtue of the
nickel(I) salen catalyzed synthesis of carbocycles is that the
carbon–halogen bond is effectively reduced at a much less
negative potential, which thereby avoids the formation of
carbanionic intermediates that do not cyclize. Thus,
nickel(I) salen promoted reductive intramolecular cycliza-
tions deserve serious attention as a tool for organic synthe-
sis.
Comparison of the Electrogenerated Nickel(I) Salen and
Other Chemical Reductants as Reagents for Effecting the
Intramolecular Cyclization of Phenyl-Conjugated
Haloacetylenes
It is interesting to compare some chemical procedures
previously employed to carry out the reductive intramolecu-
lar cyclization of phenyl-conjugated acetylenic halides with
respect to the electrochemical method described in the pres-
ent study. Crandall and coworkers investigated the use of
tri-n-butyltin hydride (Bu3SnH),[4] lithium biphenylide,[4]
Grignard chemistry,[5] chromium(II) in aqueous DMF con-
taining ethylenediamine,[8] and lithium di-n-butylcuprate[7]
as reductants to promote the intramolecular cyclization of
haloacetylenes. With the use of Bu3SnH (with AIBN as ini-
tiator), Crandall and Keyton[4] found that 5-bromo-1-
phenyl-1-pentyne underwent no reductive cyclization,
whereas 7-bromo-1-phenyl-1-heptyne was converted into
benzylidenecyclohexane in up to 75% yield depending on
the initial concentration of the starting material. In the
same work,[4] the use of the lithium metal–biphenyl system
led to formation of benzylidenecyclobutane in 26% yield
from 5-bromo-1-phenyl-1-pentyne, and benzylidenecyclo-
hexane was obtained in 15% yield from 7-bromo-1-phenyl-
1-heptyne. When a tetrahydrofuran solution of the Grig-
nard reagent derived from 5-bromo-1-phenyl-1-pentyne was
heated at reflux for 30 days, the desired carbocycle was pro-
duced in 21% yield; a similar experiment with the Grignard
Mechanistic Aspects of the Nickel(I) Salen Catalyzed
Reduction of Phenyl-Conjugated Haloalkynes
Shown in Equations (1) to (6) are the most plausible
reagent of 7-bromo-1-phenyl-1-heptyne afforded benzyl- pathways for formation of the various products of the
Eur. J. Org. Chem. 2007, 5346–5352
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5349

M. S. Mubarak, T. B. Jennermann, M. A. Ischay, D. G. Peters
FULL PAPER
nickel(I) salen catalyzed reduction of a phenyl-conjugated
haloalkyne. Recent investigations in our laboratory have
indicated that the one-electron reduction of nickel(II) salen
can be considered as either a metal- or a ligand-centered
process. First, for the catalytic reduction of 1-haloalkanes
by electrogenerated nickel(I) salen, we discovered that the
imino (C=N) bonds of the ligand are alkylated, which sug-
gests that reduction of nickel(II) salen places significant
electron density on the imino bonds.[26] Second, by using
density functional theory, we determined that for nickel(II)
salen there are two unoccupied molecular orbitals Ϫ one
metal-centered and one ligand-centered Ϫ differing in en-
ergy by 2–3 kcalmol–1 available to accommodate the elec-
tron added to generate the reduced form of nickel(II)
salen.[27] However, for the present discussion, we offer a
mechanistic scheme that involves only the one-electron,
metal-centered reduction of nickel(II) salen to give catalyti-
cally active nickel(I) salen.
(4)
(5)
(6)
Experimental Section
Accordingly, reversible one-electron reduction of nickel-
(II) salen to nickel(I) salen [Equation (1)] is the initial step
in the catalytic process. As quickly as it is generated, nickel-
(I) salen transfers an electron to the carbon–halogen bond
of a phenyl-conjugated acetylenic halide to afford a 1-
phenyl-1-alkyn-ω-yl radical (10) and a free halide ion, and
to regenerate nickel(II) salen [Equation (2)]. Once formed,
Reagents: Each of the following compounds was purchased from
Aldrich Chemical Company and was used without further purifica-
tion: [{2,2Ј-[1,2-ethanediylbis(nitrilomethylidyne)]bis(phenolato)}-
(N,NЈ,O,OЈ)]nickel(II) {nickel(II) salen, 99%}, 1,3-dibromopro-
pane (99%), 1,5-dibromopentane (99%), 1,6-dibromohexane
(96%), 1,3-diiodopropane (99%), 1,5-diiodopentane (97%), phen-
ylacetylene (98%), 1,10-dibromodecane (97%), 1,12-dibromo-
10 can engage in four product-forming processes. First, and dodecane (98%), benzyltriphenylphosphonium chloride (99%), (4-
bromobutyl)triphenylphosphonium bromide (98%), cyclohexanone
(99+%), cycloheptanone (99%), 1-phenyl-1-pentyne (98%), 1.6 
n-butyllithium in hexanes, 2.0  benzylmagnesium chloride in
tetrahydrofuran, and n-decane (99%). We obtained 1-phenyl-1-hep-
tyne (99%) and 1-phenyl-1-octyne (95%) from TCI America. Zero-
grade argon (Air Products) was used to deaerate solutions for all
electrochemical experiments. Dimethylformamide (DMF, ACS
grade, 99.8%) from Fisher Scientific was used without further puri-
fication as the solvent for cyclic voltammetry and controlled-poten-
tial electrolysis; the concentration of residual water in the solvent
was typically 40 m.[28] Tetramethylammonium tetrafluoroborate
(TMABF4, 97%), employed as the supporting electrolyte, was ob-
tained from GFS Chemicals and was recrystallized from water/
methanol and stored in a vacuum oven at 80 °C to remove traces
of water.
most important to the present study, 10 can undergo intra-
molecular cyclization, followed by hydrogen atom abstrac-
tion from the solvent (SH), to give benzylidenecycloalkane
6 [Equation (3)]. Alternatively, 10 can engage in radical
coupling with itself to yield 7 [Equation (4)]. In addition,
10 can abstract a hydrogen atom from the solvent (SH),
prior to intramolecular cyclization, to give 1-phenyl-1-al-
kyne 8 [Equation (5)]. Finally, disproportionation of 10
leads to 8 and 9 [Equation (6)].
(1)
Instrumentation, Electrodes, and Cells: Instrumentation employed
for cyclic voltammetry and controlled-potential electrolyses is de-
scribed elsewhere.[29,30] For cyclic voltammetry, we constructed a
planar, circular working electrode with an area of 0.077 cm2 by
press-fitting a 3-mm-diameter glassy carbon rod (Grade GC-20,
Tokai Electrode Manufacturing Company, Tokyo, Japan) into a
Teflon shroud. Before each cyclic voltammogram was recorded, the
electrode was cleaned with an aqueous suspension of 0.05-µm alu-
mina on a Master-Tex (Buehler) polishing pad, after which the elec-
trode was rinsed ultrasonically in a water bath.
(2)
Reticulated vitreous carbon disks (RVC 2X1–100S, Energy Re-
search and Generation, Inc., Oakland, CA), approximately 2.4 cm
in diameter, 0.4 cm in thickness, and with an approximate geomet-
ric area of 200 cm2, were employed as working electrodes for all
controlled-potential (bulk) electrolyses. Procedures for preparing,
cleaning, and handling of these electrodes are cited in the litera-
ture.[31]
Electrochemical cells for cyclic voltammetry and controlled-poten-
(3) tial electrolysis have also been described previously.[32,33] All poten-
5350
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5346–5352

Catalytic Reduction by Nickel(I) Salen: Cyclization vs. Coupling
tials are quoted with respect to a reference electrode consisting of
a cadmium-saturated mercury amalgam in contact with DMF satu-
rated with both cadmium chloride and sodium chloride; this elec-
trode has a potential of –0.76 V vs. the aqueous saturated calomel
electrode (SCE) at 25 °C.[34,35]
(39.5 g, 0.102 mol) dissolved in dry THF (250 mL). After the solu-
tion was stirred for 1 h, cyclohexanone (12.3 g, 0.125 mol) was
added, and the mixture was heated at reflux for 4 h. The solution
was then filtered and concentrated, and triphenylphosphane oxide
was precipitated by the addition of cold pentane. Pentane was re-
moved under reduced pressure, and the residue was vacuum dis-
tilled to afford benzylidenecyclohexane. MS (70 eV): m/z (%) = 172
(96) [M]+, 143 (15) [M – 29]+, 129 (63) [M – 43]+, 104 (78) [M –
Separation, Identification, and Quantitation of Electrolysis Prod-
ucts: At the end of each controlled-potential electrolysis, the catho-
lyte containing the products was partitioned between diethyl ether
and water, and then the ether phase was dried with anhydrous mag-
nesium sulfate, concentrated by means of rotary evaporation, and
analyzed with the aid of either GC–MS or GC. First, we used GC–
MS to identify all of the electrolysis products through a compari-
son of their gas chromatographic retention times and mass spectra
with those of authentic reference compounds; a Hewlett–Packard
6890N gas chromatograph coupled with a Hewlett–Packard 5973
inert mass-selective detector was employed. Second, the quantita-
tion of products was accomplished with a Hewlett–Packard 5890
Series II gas chromatograph equipped with a flame-ionization de-
tector, a Hewlett–Packard 3392A integrator, and a 30 mϫ0.25 mm
capillary column (J & W Scientific) with a DB-5 stationary phase
consisting of 5% phenylpolysiloxane and 95% methylpolysiloxane.
As described in an earlier publication,[36] peak areas and yields for
the various products were determined with respect to that of an
electroinactive internal standard (n-decane) added in known
amount to the solution at the beginning of each electrolysis.
Throughout this paper, all product yields are expressed as the per-
centage of starting material incorporated into each species.
1
68]+, 91 (68) [C7H7]+, 81 (100) [M – 91]+. The H NMR spectrum
agreed with that recorded previously.[5]
Benzylidenecycloheptane: We employed the procedure outlined in
the preceding paragraph to prepare benzylidenecycloheptane (8-
phenylheptafulvene), except that cycloheptanone (14.0 g,
0.125 mol) was used instead of cyclohexanone. MS (70 eV): m/z (%)
= 186 (93) [M]+, 143 (21) [M – 43]+, 129 (58) [M – 57]+, 115 (49)
[M – 71]+, 104 (100) [M – 82]+, 91 (52) [C7H7]+.
Synthesis of Dimers (Diphenylalkadiynes) (7)
1,10-Diphenyldeca-1,9-diyne: This compound was prepared accord-
ing to a previously published procedure.[19] Mass spectral measure-
ments yielded the following results. MS (70 eV): m/z (%) = 286 (14)
[M]+, 257 (32) [M – 29]+, 243 (24) [M – 43]+, 128 (44) [M – 158]+,
115 (100) [M – 171]+, 91 (40) [C7H7]+. HRMS: calcd. for C22H22
[M]+ 286.1722; found 286.1730.
1,14-Diphenyltetradeca-1,13-diyne: For the preparation of this com-
pound, phenylacetylene (12.0 g, 0.117 mol, 12.9 mL) in dry THF
(80 mL) was added to n-butyllithium (1.6  in hexanes, 66 mL,
0.106 mol), and the mixture was heated at reflux for 3 h. Then,
1,10-dibromodecane (16.0 g, 0.053 mol, 12.0 mL) was added, and
the solution was heated at reflux overnight. After treatment with
water (2 mL), the organic layer was dried with anhydrous magne-
sium sulfate, and the solvent was removed under reduced pressure.
Finally, the residue was vacuum distilled to eliminate unreacted
starting materials and low-boiling side products, after which the
desired product was obtained. MS (70 eV): m/z (%) = 342 (42)
[M]+, 285 (14) [M – 57]+, 209 (29) [M – 133]+, 129 (84) [M –
213]+, 115 (82) [M – 227]+, 91 (100) [C7H7]+. HRMS: calcd. for
C26H30 [M]+ 342.2342; found 342.2339. HRMS: calcd. for C26H29
[M – H]+ 341.2264; found 341.2254.
Synthesis of Phenyl-Conjugated Bromo- and Iodoalkynes (1–5): We
prepared 5-bromo-1-phenyl-1-pentyne (1), 5-iodo-1-phenyl-1-pen-
tyne (2), 7-bromo-1-phenyl-1-heptyne (3), 7-iodo-1-phenyl-1-hep-
tyne (4), and 8-bromo-1-phenyl-1-octyne (5) according to a pre-
viously published procedure[19] involving the addition of phenyl-
acetylene (12.0 g, 0.117 mol, 12.9 mL) in dry THF (80 mL) to n-
butyllithium (1.6  in hexanes, 66 mL, 0.106 mol). This mixture
was heated to reflux and, after cessation of gas evolution (2–4 h),
the appropriate amount of an α,ω-dibromo- or α,ω-diiodoalkane
was injected into the solution. Then, after heating at reflux over-
night, the mixture was hydrolyzed by the addition of water (2 mL).
Finally, the solution was dried with anhydrous sodium sulfate, the
solvent was removed under reduced pressure, and the residue was
vacuum distilled to give the desired product. For each of the five
compounds, the boiling point and 1H NMR spectrum were in
agreement with previously published data.[5]
1,16-Diphenylhexadeca-1,15-diyne: We used the procedure de-
scribed in the preceding paragraph to synthesize an authentic sam-
ple of 1,16-diphenylhexadeca-1,15-diyne. MS (70 eV): m/z (%) =
370 (20) [M]+, 243 (23) [M – 127]+, 231 (40) [M – 139]+, 121 (48)
[M – 249]+, 115 (100) [M – 255]+, 91 (50) [C7H7]+. HRMS: calcd.
for C28H35 [M + H]+ 371.2731; found 371.2733. HRMS: calcd. for
C28H34 [M]+ 370.2667; found 370.2665. HRMS: calcd. for C28H33
[M – H]+ 369.2573; found 369.2577.
Synthesis of Carbocycles (Benzylidenecycloalkanes) (6)
Benzylidenecyclobutane: We obtained benzylidenecyclobutane
through a Wittig reaction by adding n-butyllithium (1.6  in hex-
anes, 60 mL, 0.094 mol) to (4-bromobutyl)triphenylphosphonium
bromide (38.0 g, 0.080 mol) dissolved in anhydrous diethyl ether
(300 mL). After this solution was heated at reflux for 24 h, ad-
ditional n-butyllithium (1.6  in hexanes, 60 mL, 0.094 mol) was
added to the reaction mixture, which was then stirred for 1 h. Next,
freshly distilled benzaldehyde (10.6 g, 0.100 mol) was added, and
the solution was stirred for 1 h. The mixture was then filtered and
concentrated, and triphenylphosphane oxide was precipitated by
the addition of cold pentane. Finally, the pentane extract was con-
centrated and distilled under vacuum to give benzylidenecyclobu-
tane. MS (70 eV): m/z (%) = 144 (42) [M]+, 129 (100) [M – 15]+,
115 (83) [M – 29]+. The boiling point and 1H NMR spectrum
agreed with those recorded previously.[5]
Identification of Minor Products 8 and 9
Phenyl-Conjugated Alkynes (8): We verified the presence of 1-
phenyl-1-pentyne, 1-phenyl-1-heptyne, and 1-phenyl-1-octyne as
electrolysis products by comparing gas chromatographic retention
times and mass spectroscopic data of the suspected compounds
with those of commercially available authentic samples.
Phenyl-Conjugated Eneynes (9): Eneynes, such as 1-phenyl-7-
hepten-1-yne, were identified as minor electrolysis products on the
basis of their gas chromatographic retention times and mass spec-
tra. Retention times for the eneynes are very close to those for the
corresponding alkynes; for example, with our DB-5 capillary col-
umn (initial temperature, 50 °C; initial time, 1 min; rate,
8 °Cmin–1), the retention time for 1-phenyl-1-heptyne was
8.61 min, whereas that for 1-phenyl-7-hepten-1-yne was 8.41 min.
Benzylidenecyclohexane: We synthesized benzylidenecyclohexane
through a Wittig reaction by adding n-butyllithium (1.6  in hex-
anes, 80 mL, 0.125 mol) to benzyltriphenylphosphonium chloride
Eur. J. Org. Chem. 2007, 5346–5352
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5351

M. S. Mubarak, T. B. Jennermann, M. A. Ischay, D. G. Peters
FULL PAPER
[14] B. C. Willett, W. M. Moore, A. Salajegheh, D. G. Peters, J. Am.
Chem. Soc. 1979, 101, 1162–1167.
[15] R.-L. Shao, J. A. Cleary, D. M. La Perriere, D. G. Peters, J.
Org. Chem. 1983, 48, 3289–3294.
[16] R.-L. Shao, D. G. Peters, J. Org. Chem. 1987, 52, 652–657.
[17] M. S. Mubarak, D. D. Nguyen, D. G. Peters, J. Org. Chem.
1990, 55, 2648–2652.
Data for 1-phenyl-4-penten-1-yne: MS (70 eV): m/z (%) = 142 (85)
[M]+, 141 (100) [M – 1]+, 115 (55) [M – 27]+. Data for 1-phenyl-6-
hepten-1-yne: MS (70 eV): m/z (%) = 170 (16) [M]+, 169 (28) [M –
1]+, 155 (53) [M – 15]+, 141 (94) [M – 29]+, 115 (100) [M – 55]+,
91 (26) [C7H7]+. Data for 1-phenyl-7-octen-1-yne: MS (70 eV): m/z
(%) = 184 (4) [M]+, 183 (8) [M – 1]+, 169 (21) [M –15]+, 155 (33)
[M – 29]+, 141 (73) [M – 43]+, 115 (100) [M – 69]+, 91 (25)
[C7H7]+.
[18] M. S. Mubarak, D. G. Peters, J. Electroanal. Chem. 1992, 332,
127–134.
[19] M. A. Ischay, M. S. Mubarak, D. G. Peters, J. Org. Chem.
2006, 71, 623–628.
Acknowledgments
[20] W. M. Moore, D. G. Peters, J. Am. Chem. Soc. 1975, 97, 139–
144.
For summer 2006, T. B. J. received research scholarships from the
Harry G. Day Fund (Department of Chemistry, Indiana Univer-
sity) and the Hutton Honors College of Indiana University.
M. A. I. thanks GlaxoSmithKline (GSK) for an undergraduate fel-
lowship for summer 2005 as well as the Indiana University Science,
Technology, and Research Scholars (IU STARS) program for fi-
nancial support.
[21] J. Z. Stemple, D. G. Peters, J. Org. Chem. 1989, 54, 5318–5323.
[22] J. Z. Stemple, D. G. Peters, J. Electroanal. Chem. 1990, 286, 89–
108.
[23] D. M. Fang, D. G. Peters, M. S. Mubarak, J. Electrochem. Soc.
2001, 148, E464–E467.
[24] P. Vanalabhpatana, D. G. Peters, J. A. Karty, J. Electroanal.
Chem. 2005, 580, 300–312.
[25] P. Vanalabhpatana, D. G. Peters, J. Electrochem. Soc. 2005,
152, E222–E229.
[1] S. A. Kandil, R. E. Dessy, J. Am. Chem. Soc. 1966, 88, 3027–
[26] D. M. Goken, M. A. Ischay, D. G. Peters, J. W. Tomaszewski,
J. A. Karty, J. P. Reilly, M. S. Mubarak, J. Electrochem. Soc.
2006, 153, E71–E77.
[27] P. W. Raess, M. S. Mubarak, M. A. Ischay, M. P. Foley, T. B.
Jennermann, K. Raghavachari, D. G. Peters, J. Electroanal.
Chem. 2007, 603, 124–134.
[28] D. M. La Perriere, W. F. Carroll Jr, B. C. Willett, E. C. Torp,
D. G. Peters, J. Am. Chem. Soc. 1979, 101, 7561–7568.
[29] W. A. Pritts, D. G. Peters, J. Electroanal. Chem. 1995, 380, 147–
160.
[30] L. J. Klein, K. S. Alleman, D. G. Peters, J. A. Karty, J. P. Reilly,
J. Electroanal. Chem. 2000, 481, 24–33.
[31] J. A. Cleary, M. S. Mubarak, K. L. Vieira, M. R. Anderson,
D. G. Peters, J. Electroanal. Chem. 1986, 198, 107–124.
[32] C. E. Dahm, D. G. Peters, Anal. Chem. 1994, 66, 3117–3123.
[33] K. L. Vieira, D. G. Peters, J. Electroanal. Chem. 1985, 196, 93–
104.
3034.
[2] H. R. Ward, J. Am. Chem. Soc. 1967, 89, 5517–5518.
[3] H. G. Richey Jr, A. M. Rothman, Tetrahedron Lett. 1968, 9,
1457–1460.
[4] J. K. Crandall, D. J. Keyton, Tetrahedron Lett. 1969, 10, 1653–
1656.
[5] W. J. Michaely, PhD Thesis, Indiana University, Bloomington,
1971.
[6] J. K. Crandall, W. J. Michaely, J. Organomet. Chem. 1973, 51,
375–379.
[7] J. K. Crandall, P. Battioni, J. T. Wehlacz, R. Bindra, J. Am.
Chem. Soc. 1975, 97, 7171–7172.
[8] J. K. Crandall, W. J. Michaely, J. Org. Chem. 1984, 49, 4244–
4248.
[9] W. F. Bailey, T. V. Ovaska, Tetrahedron Lett. 1990, 31, 627–630.
[10] W. F. Bailey, T. V. Ovaska, J. Am. Chem. Soc. 1993, 115, 3080–
3090.
[34] L. W. Marple, Anal. Chem. 1967, 39, 844–846.
[11] J. K. Crandall, W. J. Michaely, F. Collonges, D. J. Nelson, T. A. [35] C. W. Manning, W. C. Purdy, Anal. Chim. Acta 1970, 51, 124–
Ayers, J. J. Gajewski, Croat. Chem. Acta 1996, 69, 1473–1490.
[12] W. M. Moore, D. G. Peters, Tetrahedron Lett. 1972, 13, 453–
456.
126.
[36] W. A. Pritts, K. L. Vieira, D. G. Peters, Anal. Chem. 1993, 65,
2145–2149.
[13] W. M. Moore, A. Salajegheh, D. G. Peters, J. Am. Chem. Soc.
1975, 97, 4954–4960.
Received: July 16, 2007
Published Online: August 31, 2007
5352
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 5346–5352