Catalytic cycloisomerization of unsaturated organoiodides

Article abstract of DOI:10.1021/jo981991q

Catalytic quantities of phenyllithium (PhLi) have been found to initiate novel 5-exo cycloisomerization of a variety of structurally diverse unsaturated organoiodides. The isomerization reaction appears to be a process of broad synthetic utility for the preparation of iodomethyl-substituted five-membered rings. Primary, secondary, tertiary, or aryl iodides tethered to a suitably positioned carbon-carbon π-bond are converted cleanly to their cyclic isomers in good to excellent yield (i.e., 70-90%) by simply allowing a hydrocarbon-MTBE solution of the iodide to stand in the presence of a small quantity of PhLi at an appropriate temperature. The mechanism of the cycloisomerization was found to be substrate dependent: unsaturated aryl and primary alkyl iodides undergo isomerization via a three-step cascades (eqs 1- 3) mediated by two reversible lithium-iodine exchange reactions bracketing an irreversible 5-exo cyclization of an unsaturated organolithium; unsaturated secondary and tertiary alkyl iodides apparently isomerize via a radical- mediated atom transfer process initiated by homolytic fragmentation of the ate-complex generated upon attack of PhLi on the iodine atom of the substrate.

Full text of DOI:10.1021/jo981991q

9960
J . Org. Chem. 1998, 63, 9960-9967
Ca ta lytic Cycloisom er iza tion of Un sa tu r a ted Or ga n oiod id es
William F. Bailey* and Matthew W. Carson
Department of Chemistry, The University of Connecticut, Storrs, Connecticut 06269-3060
Received October 2, 1998
Catalytic quantities of phenyllithium (PhLi) have been found to initiate novel 5-exo cycloisomer-
ization of a variety of structurally diverse unsaturated organoiodides. The isomerization reaction
appears to be a process of broad synthetic utility for the preparation of iodomethyl-substituted
five-membered rings. Primary, secondary, tertiary, or aryl iodides tethered to a suitably positioned
carbon-carbon π-bond are converted cleanly to their cyclic isomers in good to excellent yield (i.e.,
70-90%) by simply allowing a hydrocarbon-MTBE solution of the iodide to stand in the presence
of a small quantity of PhLi at an appropriate temperature. The mechanism of the cycloisomerization
was found to be substrate dependent: unsaturated aryl and primary alkyl iodides undergo
isomerization via a three-step cascade (eqs 1-3) mediated by two reversible lithium-iodine exchange
reactions bracketing an irreversible 5-exo cyclization of an unsaturated organolithium; unsaturated
secondary and tertiary alkyl iodides apparently isomerize via a radical-mediated atom transfer
process initiated by homolytic fragmentation of the ate-complex generated upon attack of PhLi on
the iodine atom of the substrate.
We recently reported that the reversible nature of the
more complex than that suggested by the three-step
lithium-iodine exchange reaction may be exploited to
effect clean isomerization of 6-iodo-1-hexene to (iodo-
methyl)cyclopentane upon treatment of the unsaturated
alkyl iodide with a catalytic quantity of phenyllithium
(PhLi).1 This novel cycloisomerization apparently involves
three discrete steps as illustrated below (eqs 1-3).1 Thus,
the unsaturated alkyllithium initially generated by re-
versible exchange2,3 between the iodide precursor and
PhLi (eq 1) undergoes irreversible cyclization (eq 2)4,5
prior to reconversion to a cyclic iodide and regeneration
of PhLi (eq 3). Provided that the final equilibrium
represented by eq 3 lies far to the right, the net isomer-
ization is catalytic in PhLi.
sequence outlined in eqs 1-3.
Prompted by the potential synthetic utility of a process
that serves to convert an unsaturated organoiodide to its
cyclic isomer in a highly atom-economical fashion,6 we
have investigated the scope of PhLi-initiated cycloisomer-
ization as a route to iodomethyl-substituted five-mem-
bered rings. Herein we report that such isomerizations
appear to be of general utility: a catalytic quantity of
PhLi is indeed capable of effecting cycloisomerization of
a variety of structurally diverse unsaturated alkyl and
aryl iodides. However, as detailed below, the mechanism
of the process is substrate dependent and it is sometimes
Resu lts a n d Discu ssion
Ar yl Iod id es Bea r in g a P en d a n t Un sa tu r a tion .
The possibility of effecting cycloisomerization of an aryl
iodide with a pendant carbon-carbon π-bond was ex-
plored with N,N-diallyl-2-iodoaniline (1) as a typical
substrate. Prior art has demonstrated that the organo-
lithium derived from 1 undergoes clean, rapid 5-exo
cyclization at temperatures above ∼ -20 °C in the
presence of N,N,N′N′-tetramethylethylenediamine (TME-
DA) to deliver [(1-allyl-3-indolinyl)methyl]lithium in high
yield.7 It might be anticipated that 1 would be a good
substrate for PhLi-initiated cycloisomerization via the
exchange-mediated sequence discussed above: not only
is the initial exchange equilibrium between PhLi and 1
(eq 1) likely to be more favorable than that between PhLi
and an alkyl iodide,2,3 the [(1-allyl-3-indolinyl)methyl]-
lithium generated upon cyclization (eq 2) may be con-
verted to product by exchange with either PhI (eq 3) or
the starting material (1).
(1) Bailey, W. F.; Carson, M. W. J . Org. Chem. 1998, 63, 361.
(2) Applequist, D. E.; O’Brien, D. F. J . Am. Chem. Soc. 1963, 85,
743.
(3) (a) Wakefield, B. J . The Chemistry of Organolithium Compounds;
Pergamon Press: New York, 1974. (b) Bailey, W. F.; Patricia, J . J . J .
Organomet. Chem. 1988, 352, 1. (c) Wakefield, B. J . Organolithium
Methods; Pergamon Press: New York, 1988. (d) Wardell, J . L. In
Comprehensive Organometallic Chemistry; Wilkinson, G., Ed.; Perga-
mon Press: New York, 1982; Vol. 1, p 43. (e) J ones, R. G.; Gilman, H.
Chem. Rev. 1954, 54, 835. (f) Gilman, H.; J ones, R. G. Org. React. (N.
Y.) 1951, 6, 339.
(4) Bailey, W. F.; Patricia, J . J .; DelGobbo, V. C.; J arret, R. M.;
Okarma, P. J . J . Org. Chem. 1985, 50, 1999.
(5) Bailey, W. F.; Khanolkar, A. D.; Gavaskar, K.; Ovaska, T. V.;
Rossi, K.; Thiel, Y.; Wiberg, K. B. J . Am. Chem. Soc. 1991, 113, 5720
and references therein.
As expected, cycloisomerization of 1 to give 1-allyl-3-
iodomethylindoline (2) was readily accomplished, as
illustrated below, by simply allowing an approximately
(7) (a) Zhang, D.; Liebeskind, L. S. J . Org. Chem. 1996, 61, 2594.
(b) Bailey, W. F.; J iang, X.-L. J . Org. Chem. 1996, 61, 2596. (c) Bailey,
W. F.; Carson, M. W. Tetrahedron Lett. 1997, 38, 1329.
(6) (a) Trost, B. M. Science 1991, 254, 1471. (b) Trost, B. M. Angew.
Chem., Int. Ed. Engl. 1995, 34, 259.
10.1021/jo981991q CCC: $15.00 © 1998 American Chemical Society
Published on Web 12/04/1998

Catalytic Cycloisomerization
J . Org. Chem., Vol. 63, No. 26, 1998 9961
Sch em e 1
0.1 M solution of 1 in dry and deoxygenated n-pentane-
MTBE (9:1 by vol) to stand under argon in the presence
of 10 mol % of both PhLi and TMEDA for 1 h at 0 °C. GC
analysis of the crude reaction mixture revealed that 2
had been produced in 87% yield; the somewhat labile
iodide was isolated as analytically pure material in 70%
yield. It should be noted, as detailed elsewhere,1 although
the conversion of 1 to 2 requires, in principle, only a trace
of PhLi to initiate the reaction sequence (eqs 1-3), it is
our experience that it is exceedingly difficult, in practice,
to avoid premature termination of an exchange-mediated
cycloisomerization through inadvertent quench of
organolithium intermediates by proton abstraction from
solvent or adventitious moisture when very small quanti-
ties of initiator are used.
approximately 70/30 mixture of 2 and 1-allyl-3-methyl-
indole. Not surprisingly, a longer reaction time or a larger
initial concentration of PhLi leads to a larger proportion
of indole in the product mixture. The indole is most likely
generated, as summarized in Scheme 1, by base-catalyzed
isomerization of the 3-methylene species produced upon
loss of HI from 2. As a practical matter, dehydrohaloge-
nation of the product is not a major side reaction when
a catalytic quantity of PhLi-TMEDA is used, as de-
scribed above, to initiate the cycloisomerization of 1 to
2; the favorable position of the initial exchange equilib-
rium leading to 3 (eq 1), followed by the rapid unimo-
lecular cyclization of 3 (eq 2) in the presence of TMEDA
at 0 °C,7 allows the cycloisomerization sequence to
proceed essentially to completion before the inherently
slower bimolecular elimination reaction becomes a prob-
lem.
Dehydrohalogenation can be a potentially serious
complication when the cyclization step (eq 2) of the
exchange-mediated cycloisomerization is sluggish. To the
extent that the alkyl iodide product is consumed by
reaction with an organolithium, the dehydrohalogenation
also removes a portion of the initiator and/or the initially
generated aryllithium needed to complete the cyclo-
isomerization. In short, dehydrohalogenation serves to
terminate the isomerization sequence (eqs 1 and 3) when
the cyclization step (eq 2) is slow.
The PhLi-initiated isomerization of 1 to 2 is undoubt-
edly mediated by the exchange-cyclization-exchange
cascade outlined above (eqs 1-3). Indeed, as shown
below, addition of a full molar equivalent of PhLi to a
solution of 1 in n-pentane-MTBE (9:1 by vol) effects
virtually complete exchange within minutes at -40 °C
to give iodobenzene and [2-(N,N-diallylamino)phenyl]-
lithium (3); quench of such a reaction mixture with
MeOH affords N,N-diallylaniline and iodobenzene in
essentially quantitative yield. Apparently, the ortho-
amino substituent present in 3, which is known to
stabilize an adjacent carbon-metal bond,8 is responsible
for this highly one-sided exchange equilibrium. As ex-
pected,7 no cyclic product was detected when the ex-
change between 1 and PhLi was established in the
absence of TMEDA at temperatures below -20 °C.
J ust such behavior is apparent in the PhLi-initiated
conversion of 2-iodo-1-(3-butenyl)benzene (4) to 1-iodo-
methylindan (5) illustrated below. Cyclization of [2-(3-
butenyl)phenyl]lithium, the organolithium derived from
4, which was first reported over a decade ago by Woosley
and co-workers,9 is considerably less facile than is the
ring closure of 3 discussed above. Moreover, the lithium-
iodine exchange equilibrium (eq 1) between 4 and PhLi
to give [2-(3-butenyl)phenyl]lithium and PhI is much less
favorable than the comparable exchange involving 1.10
As a result, the cycloisomerization of 4 to 5 requires more
forcing conditions (elevated temperature and consider-
ably more initiator) than does the conversion of 1 to 2.
Thus, when a solution of 4 in scrupulously dry, oxygen-
free cyclohexane-MTBE (9:1 by vol) containing 25 mol
% of both PhLi and TMEDA was heated at 60 °C for 7 h
under an atmosphere of pure argon, the product mixture
consisted of 5 (43%) as well as 3-methylindene (6, 17%),
4-phenyl-1-butene (7, 13%), and recovered 4 (27%).
Clearly, the dehydrohalogenation reaction, which is
unavoidable at the elevated temperature needed to effect
Although cycloisomerization of 1 to 2 may be ac-
complished at temperatures as low as -40 °C in the
presence of a full equivalent of PhLi and TMEDA,
significant benefit derives from conducting the transfor-
mation, albeit more slowly, using only a catalytic quantity
of the aryllithium initiator at 0 °C. While a higher initial
concentration of PhLi leads to a faster overall isomer-
ization, the presence of excess organolithium-TMEDA
in such reaction mixtures may result in dehydrohaloge-
nation of the alkyl iodide product (2). For example, as
illustrated in Scheme 1, treatment of a solution of 1 in
n-pentane-MTBE (9:1 by vol) with 1 equiv of PhLi and
2 equiv of TMEDA at -20 °C leads within 3 min to
consumption of the starting iodide and formation of an
(9) (a) Ross, G. A.; Koppang, M. D.; Bartak, D. E.; Woolsey, N. F. J .
Am. Chem. Soc. 1985, 107, 6742. (b) Koppang, M. D.; Ross, G. A.;
Woolsey, N. F.; Bartak, D. E. J . Am. Chem. Soc. 1986, 108, 1441.
(8) (a) Snieckus, V. Chem. Rev. 1990, 90, 879. (b) Gschwend, H. W.;
Rodriguez, H. R. Org. React. (N. Y.) 1979, 1.

9962 J . Org. Chem., Vol. 63, No. 26, 1998
Bailey and Carson
cyclization of the aryllithium derived from 4, is respon-
sible for formation of both 6 and 7. Perhaps more
importantly, no active organolithium remains in the
reaction mixture after 7 h at 60 °C and the isomerization
sequence is effectively terminated. Of course, additional
PhLi may be used to complete the consumption of 4 but
this expedient also results in consumption of product
iodide (5) by conversion to 6. Although the PhLi-initiated
cycloisomerization of 4 is slow and results in modest
yields of 5, the transformation most likely involves the
three-step, exchange-mediated sequence outlined above
(eqs 1-3).
(viz., trans/cis > 30),11-13 it seems improbable that the
PhLi-initiated conversion of 8 to 9 involves cyclization
of the organolithium. The cis-rich stereochemistry of the
product mixture is, however, highly suggestive of the
intermediacy of a 1-methyl-5-hexenyl radical.14,15 Indeed,
the stereochemical outcome of a reaction mediated by
cyclization of the 1-methyl-5-hexenyl radical is well-
known: kinetic studies by Ingold’s group have demon-
strated that the product composition is temperature
dependent and Arrhenius parameters characterizing both
the cis- and trans-modes of ring closure are available.15
In view of the apparent involvement of the 1-methyl-5-
hexenyl radical in the PhLi-initiated cycloisomerization
of 8, the isomeric composition of the 2-iodomethyl-1-
methylcyclopentane product (cis-9 and trans-9) should be
temperature dependent and quantitatively predictable.16
Agreement of experimental cis/trans ratios with product
composition predicted on the assumption that stereo-
chemistry is controlled by cyclization of the 1-methyl-5-
hexenyl radical would provide strong evidence for the
putative radical intermediate. To this end, a series of
experiments were conducted in which solutions of 8 in
hydrocarbon-MTBE (9:1 by vol) were treated with PhLi
at several temperatures between -20 and 60 °C. The
results of these experiments are summarized in Table 1.
At the outset it should be noted that the isomerizations
reported in Table 1 were conducted by using a full molar
equivalent of PhLi so as to increase the overall rate of
the cycloisomerization: analogous results were obtained
when catalytic quantities of PhLi were employed for
several of the experiments but reaction times were then
much longer. Cursory inspection of the data presented
in Table 1 demonstrates that there is a very good
correlation between the observed isomeric composition
and that calculated on the assumption that the cyclo-
isomerization reaction is mediated by cyclization of the
1-methyl-5-hexenyl radical. The small variation in the
Un sa tu r a ted Secon d a r y Alk yl Iod id es. In light of
the facile PhLi-initiated cycloisomerization of 6-iodo-1-
hexene and related systems bearing a primary alkyl
iodide,1 it was of interest to determine whether a second-
ary alkyl iodide incorporating a 5-hexenyl unit would
behave analogously. At the inception of the exploratory
phase of this study it was recognized that lithium-iodine
exchange between an aryllithium and a secondary alkyl
iodide to give a secondary alkyllithium and an aryl iodide
is a highly unfavorable process;2,3 conventional wisdom
suggests that the more probable outcome from reaction
of a secondary halide with an organolithium would be
elimination.3 Thus, it seemed unlikely that the exchange-
mediated cascade (eqs 1-3) would prove useful for the
cycloisomerization of a secondary substrate.
Despite these reservations, the reaction of 6-iodo-1-
heptene (8) with PhLi was investigated. Remarkably, and
quite unexpectedly, treatment of a 0.5 M solution of 8 in
n-pentane-MTBE (9:1 by vol) with as little as 3 mol %
of PhLi for 22 h at room temperature effects almost
complete isomerization of the substrate to 2-iodomethyl-
1-methylcyclopentane (cis-9 and trans-9); as shown be-
low, the cyclic product was isolated in 89% yield as a 3.3/1
mixture of the cis and trans isomers. Given the small
amount of PhLi needed to initiate the cycloisomerization
depicted below, it is important to note that control
experiments, conducted in the absence of PhLi but under
otherwise identical conditions, demonstrated that there
is no reaction in the absence of the aryllithium. Moreover,
the cycloisomerization of 8 proceeds at a comparable rate
both in the dark as well as in ambient light.
(11) Bailey, W. F.; Nurmi, T. T.; Patricia, J . J .; Wang, W. J . Am.
Chem. Soc. 1987, 109, 2442.
(12) Ashby, E. C.; Pham, T. N. J . Org. Chem. 1987, 52, 1291.
(13) It might be noted that other 6-hepten-2-yl organometallics also
cyclize with a high degree of preference for the trans stereochemistry.
Cyclization of Grignard reagents derived from 6-halo-1-heptenes
invariably produce a preponderance of the trans-isomer (trans/cis >
10; for a review, see: Hill, E. A. J . Organomet. Chem. 1975, 91, 123)
and ring closure of 1-methyl-5-hexenylsodium has been reported to
give the trans-product (see: Garst, J . F.; Hines, J . B., J r. J . Am. Chem.
Soc. 1984, 106, 6433).
In view of the fact that 1-methyl-5-hexenyllithium, the
organolithium derived from 8, is known to cyclize with a
very distinct preference for the trans-stereochemistry3b
(14) Brace, N. O. J . Org. Chem. 1967, 32, 2711.
(15) Lusztyk, J .; Maillard, B.; Deycard, S.; Lindsay, D. A.; Ingold,
K. U. J . Org. Chem. 1987, 52, 3509.
(10) In striking contrast to the favorable exchange equilibrium
established when N,N-diallyl-2-iodoaniline (1) is treated with PhLi,
the lithium-iodine exchange between 4 and PhLi gives very little [2-(3-
butenyl)phenyl]lithium and PhI. The apparent equilibrium constant
(Kobs) for this exchange [2-iodo-1-(3-butenyl)benzene (4) + PhLi h [2-(3-
butenyl)phenyl]lithium + PhI] in n-pentane-MTBE (9:1 by vol) was
determined at -40 °C following the method introduced by Applequist
and O’Brien2 and detailed in our previous report.1 The unexpectedly
small apparent equilibrium constant for the exchange between 4 and
PhLi (Kobs ) (8 ( 5) × 10 -3) may have a profound effect on the overall
rate of the exchange-mediated cycloisomerization of 4 to 5 since there
is very little [2-(3-butenyl)phenyl]lithium generated when PhLi is used
to initiate the isomerization. In a larger sense, the differing behavior
of aryl iodides 1 and 4 when treated with PhLi at low temperature
suggests that there is much yet to be learned about the factors affecting
lithium-halogen exchange equilibria.
(16) The rapid 5-exo cyclization of the 1-methyl-5-hexenyl radical
is highly cis-selective. Ingold’s group has studied the kinetics of this
cyclization over a broad range of temperatures (ca. -30 °C to 100 °C)
and Arrhenius parameters for both cis- and trans-modes of ring closure
are available from this work.15 The “Best Arrhenius Parameters”
reported in Table 2 of ref 15 for cyclization of the 1-methyl-5-hexenyl
radical to give the cis-isomer (viz., cis: log A ) 9.79 ( 0.24; Ea ) 6.50
( 0.26 kcal/mol) and the trans-isomer (viz., trans: log A ) 9.92 ( 0.26;
Ea ) 7.44 ( 0.29 kcal/mol) may be used to predict the relative
proportions of cis and trans product that would result at a given
temperature from a radical-mediated cycloisomerization of 6-iodo-1-
heptene (8) initiated by PhLi. On the reasonable assumption that
cyclization of the 1-methyl-5-hexenyl radical is irreversible under the
conditions used for the cycloisomerization, the isomeric composition
of the product mixture is given by: cis/trans ) kc/kt ) 0.741 × e473/T
.
The predicted product ratios appear in the last column of Table 1.

Catalytic Cycloisomerization
J . Org. Chem., Vol. 63, No. 26, 1998 9963
Ta ble 1. Cycloisom er iza tion of 6-Iod o-1-h ep ten e (8)a
ated atom transfer cyclization process.18,19 This well-
characterized chain reaction may be initiated by even
trace amounts of the 1-methyl-5-hexenyl radical (vide
infra).18,19 The effectively irreversible18 iodine atom trans-
fer cyclization sequence, which is depicted in abbreviated
form below, would serve to rapidly convert the unsatur-
ated secondary alkyl iodide to cyclic product.
products, % yieldb
temp, time,
cis/ predicted
transc cis/transd
entry
°C
h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
-20e
3
6
9
0.33
1
3
27.2
12.7
8.4
24.4
6.4
1.7
3.5
1.6
0.6
0.3
0.7
1.2
2.0
2.9
3.1
4.3
4.8
7.4
6.1
7.7
5.4
10.3
14.4
71.0
84.7
88.6
73.3
93.3
95.2
86.3
91.9
84.8
93.0
92.9
88.5
87.0
80.6
4.5
4.7
4.6
3.7
3.9
3.9
3.4
3.4
3.5
3.1
3.1
2.9
3.0
2.9
4.8
4.2
3.7
Generation of the 1-methyl-5-hexenyl radical from 8
in more traditional ways has been demonstrated to effect
cyclization of 8 to 9 by both Brace (who used AIBN to
initiate the conversion)14 and Curran’s group (initiation
with hexabutylditin and sunlamp irradiation).19 In light
of these precedents, it seems quite reasonable to propose
that the PhLi-initiated transformation of 8 to 9 discussed
above is the result of inadvertent initiation of the atom
transfer cyclization chain by a phenyl radical or a radical
impurity in the PhLi reagent. While one cannot exclude
such a possibility, there are two additional experimental
observations, admittedly anecdotal, that mitigate against
this supposition. (1) The PhLi-initiated rearrangement
of 8 to 9 does not seem to be affected, either in terms of
rate or product composition, by the source of the initia-
tor: both commercial (Aldrich, FMC) PhLi in cyclohex-
ane-diethyl ether (produced from chlorobenzene) and the
reagent prepared in situ from bromobenzene in pure
diethyl ether by low-temperature lithium-halogen ex-
change with t-BuLi20 give entirely analogous results. (2)
The presence of 1 equiv of cumene, an effective scavenger
of the phenyl radical,21 has no effect on the isomerization
of 8 to 9: allowing a 0.1 M solution of both 8 and cumene
in cyclohexane-MTBE (9:1 by vol) containing 1 equiv of
PhLi to stand for 1 h at 20 °C afforded 9 (93.6%, cis/trans
) 3.4) and 10 (5.5%) in good accord with the result
obtained in the absence of cumene (cf. Table 1, entry 8).
Be that as it may, the behavior of the 3-oxa analogue
of 8, 3-(2-iodopropyloxy)propene (11), upon treatment
with PhLi demonstrates that the role of the organo-
lithium initiator is far from trivial. At the outset, it was
expected that 11 would undergo reaction with PhLi in a
fashion analogous to 8 since the secondary radical derived
from 11 is known to cyclize some 10 times more rapidly
than does the 1-methyl-5-hexenyl radical derived from
8.22 In that event, addition of 1 M equiv of PhLi to an
approximately 0.1 M solution of 11 in cyclohexane-
MTBE (9:1 by vol) at 0 °C resulted in complete consump-
tion of both 11 and PhLi within 15 min, but there was
no evidence of the expected cycloisomerization product.
Rather, as illustrated below, reaction of 11 with PhLi
affords essentially 100% yields of the lithium salt of allyl
alcohol (assayed as the alcohol), propene (assayed by GC-
0
20 0.5
1
2
40 0.25
0.5
60 0.10
0.15
2.5
3.4
3.1
1.8
0.3
0.33
a
Unless otherwise indicated, 0.1 M solutions of 6-iodo-1-heptene
(8) in cyclohexane-MTBE (9:1 by vol) containing 1 M equiv of PhLi
were stirred under an atmosphere of argon at the specified
temperature for the indicated time before the addition of water
b
or methanol. Yields were determined by capillary GC by using
internal standards (typically n-heptane and n-decane) and cor-
rection for detector response; stereoisomers were assumed to have
identical detector response. c Ratio of cis-9 + cis-10/trans-9 +
d
trans-10. Proportions of cis and trans isomeric products expected
from cyclization of the 1-methyl-5-hexenyl radical at the temper-
ature of the experiment; see ref 16. e A solvent system composed
of n-pentane-MTBE (9:1 by vol) was used for experiments
conducted at -20 °C.
experimental ratios at a given nominal temperature are
most likely the result of systematic error in our temper-
ature control. Nonetheless, the results of these experi-
ments provide compelling evidence, albeit indirect, for
generation and cyclization of a secondary radical in the
course of the isomerization of 8.
The results summarized in Table 1 also serve to
demonstrate that the PhLi-initiated conversion of 8 to 9
is a very clean process. The absence (i.e., <1%) of any
products attributable to dehydrohalogenation is note-
worthy: despite the fact that a full equivalent of PhLi
was used in these conversions, there is no evidence of
HI elimination from either the secondary iodide (8) or
the product (9). Indeed, the only byproducts detected in
the reaction mixtures were small and variable amounts
of 1,2-dimethylcyclopentanes (cis-10 and trans-10) de-
rived from formal reduction of the product iodide.17
Moreover, very little of 10 is produced when the isomer-
ization is conducted at lower temperatures (Table 1,
entries 1-6), while at higher temperatures, the isomer-
ization of 8 to 9 is complete well before significant
quantities of 10 are generated (Table 1, entries 7-14).
Given that the 1-methyl-5-hexenyl radical is an inter-
mediate in the PhLi-initiated transformation of 8 to 9,
the rearrangement most likely involves a radical-medi-
(18) (a) Curran, D. P.; Chen, M.-H.; Kim, D. J . Am. Chem. Soc. 1986,
108, 2489. (b) Newcomb, M.; Curran, D. P. Acc. Chem. Res. 1988, 21,
206. (c) Curran, D. P. Synthesis 1988, 417 and 489. (d) Curran, D. P.;
Fevig, T. L.; J asperse, C. P. Chem Rev. 1991, 91, 1237. (e) Curran, D.
P.; Porter, N. A.; Giese, B. Stereochemistry of Radical Reactions; VCH
Publishers: New York, 1995.
(19) Curran, D. P.; Kim, D. Tetrahedron Lett. 1986, 27, 5821.
(20) Bailey, W. F.; Punzalan, E. R. J . Org. Chem. 1990, 55, 5404.
(21) Kryger, R. G.; Lorand, J . P.; Stevens, N. R.; Herron, N. R. J .
Am. Chem. Soc. 1977, 99, 7589 and references therein.
(22) Beckwith, A. L. J .; Blair, I.; Phillipou, G. J . Am. Chem. Soc.
1974, 96, 1613.
(17) It might be noted in this connection that lithium-iodine
exchange between the primary alkyl iodide products, cis-9 and trans-
9, and excess PhLi present in reaction mixtures may account for small
quantities of 1,2-dimethylcyclopentanes (cis-10 and trans-10) detected
in these experiments. Previous work has demonstrated, however, that
the apparent equilibrium constant for such an exchange is negligibly
small.1

9964 J . Org. Chem., Vol. 63, No. 26, 1998
Bailey and Carson
Sch em e 2
Such C-I bond homolysis may well be a relatively rare
event but, as illustrated in Scheme 2 and discussed
explicitly above, formation of even trace amounts of the
1-methyl-5-hexenyl radical will invariably result in a very
rapid radical-mediated atom transfer cyclization to give
cis-rich 9.18,19
The intermediacy of a 10-I-2 complex may well be a
common feature of all reactions of PhLi with unsaturated
organoiodides. This assumption serves to unify the
ostensibly dissimilar mechanisms of PhLi-initiated cy-
cloisomerization of a variety of substrates. The exchange-
mediated cycloisomerizations (eqs 1-3) of unsaturated
aryl iodides such as 1 and 4, as well as primary alkyl
iodides such as 6-iodo-1-hexene,1 most likely involve
reversible generation of ate-complexes in the course of
the lithium-iodine interchange steps (i.e., 12, eq 4).3,23,24
When the exchange between PhLi and the iodide sub-
strate is highly unfavorable, as is certainly the case for
secondary alkyl iodides such as 8,2,3 the ate complex may
fragment homolytically to initiate a radical-mediated
isomerization. In short, from the prospective of mecha-
nism, all of the cycloisomerizations discussed above (as
well as those to be introduced below) may be viewed as
proceeding from a common elementary stepsreversible
attack of the PhLi on the iodine atom of the substrate.
Regardless of the mechanistic details, it is clear that
even small amounts of PhLi are sufficient to initiate
cycloisomerization of suitably constituted organoiodides.
The process is not confined to substrates bearing a
pendant olefinic unit: as illustrated below, treatment of
an approximately 0.5 M solution of 7-iodo-2-octyne (13)
in dry, deoxygenated cyclohexane-MTBE (9:1 by vol)
with 15 mol % of PhLi for 18 h at 50 °C effects clean
5-exo cycloisomerization of the acetylenic alkyl iodide.
The cyclic product (14) was isolated in 91% yield as a
mixture of diastereoisomers (Z/ E ) 1.4).27 This efficient
isomerization is almost certainly the result of a PhLi-
initiated, radical-mediated atom transfer process18,19 and
the mixture of diastereoisomers that results is consistent
with cyclization of the 1-methyl-5-heptynyl radical.28
MS), and iodobenzene. This unexpected result is totally
inconsistent with inadvertent initiation of an atom
transfer cyclization by phenyl radical or trace impurities
in the reaction medium since such a scenario would have
resulted in cycloisomerization of 11 rather than the
observed fragmentation. Indeed, the fragmentation of 11
upon treatment with PhLi is characteristic of generation
of an electron-rich center at the iodine-bearing carbon
and expulsion of the allyloxy nucleofuge from the â-posi-
tion.
The disparate behavior of substrates 8 and 11 upon
treatment with PhLi may be reconciled if one postulates,
as depicted in Scheme 2, that the initial event in each
instance involves reversible attack of the organolithium
on the iodine atom of the substrate to give a 10-I-2 “ate-
complex”. J ust such a species has been implicated as an
intermediate in the lithium-iodine exchange reaction
(12, eq 4);3,23,24 the iodine ate complex generated in the
reaction of iodobenzene with PhLi (12, R ) R′ ) Ph) has
been detected spectroscopically25 and its perfluoro ana-
logue (12, R ) R′ ) C6F5) has been crystallized as a stable
TMEDA adduct.26
R-I + Li-R′ h [R-I-R′]-Li+ h R-Li + I-R′ (4)
It seems reasonable to expect rapid expulsion of the
allyloxy anion from the electron-rich ate-complex gener-
ated from 11 (Scheme 2) prior to completion of the
exchange reaction and this fragmentation nicely accounts
for the behavior of this substrate when treated with PhLi.
The behavior of 8 when treated with PhLi is also
consistent with reversible generation of an ate-complex
if it is realized that the complex need not lead to an
exchange: indeed, as noted above, there is no evidence
for the production of 1-methyl-5-hexenyllithium in the
reaction of 8 with PhLi. Rather, the observed chemistry
is consistent with homolytic fragmentation of the long,26
and apparently weak, carbon-iodine bond in the putative
ate-complex derived from 8 to give the 1-methyl-5-
hexenyl radical and the radical anion of iodobenzene.
Un sa tu r a ted Ter tia r y Alk yl Iod id es. Prior art
suggests that treatment of a tertiary alkyl iodide with
an organolithium should result in dehydroiodination of
the iodide.3 The lithium-iodine exchange between a
primary alkyl iodide and t-BuLi is, in fact, driven to
completion by rapid reaction of excess of t-BuLi with the
cogenerated t-BuI.20 However, in view of the ease with
which tertiary alkyl radicals participate in atom transfer
(27) The configuration of the major diastereoisomer (Z-14) produced
upon cycloisomerization of 13 was established by conversion of the
vinylic iodides to a mixture of E- and Z-1-ethylidene-2-methylcyclo-
pentane with retention of configuration at the vinylic carbon (2.2 t-BuLi
in n-pentane-MTBE (9:1 by vol) at -78 °C for 10 min followed by
MeOH quench). The major product of the cycloisomerization (Z-14)
gave E-1-ethylidene-2-methylcyclopentane whose NMR spectrum was
identical to that reported for an authentic sample (see: Negishi, E.-I.;
Swanson, D. R.; Cederbaum, F. E.; Takahashi, T. Tetrahedron Lett.
1987, 28, 917).
(23) Wittig, G.; Scho¨llkopf, U. Tetrahedron 1958, 3, 91.
(24) J edlicka, B.; Crabtree, R. H.; Siegbahn, P. E. M. Organometal-
lics 1997, 16, 6021.
(25) (a) Reich, H. J .; Phillips, N. H.; Reich, I. L. J . Am. Chem. Soc.
1985, 107, 4101. (b) Reich, H. J .; Green, D. P.; Phillips, N. H. J . Am.
Chem. Soc. 1989, 111, 3444. (c) Reich, H. J .; Green, D. P.; Phillips, N.
H. J . Am. Chem. Soc. 1991, 113, 1414.
(28) It might be noted that 5-exo cyclization of acetylenic alkyllithi-
ums is a stereoselectively syn-process, see: Bailey, W. F.; Ovaska, T.
V. J . Am. Chem. Soc. 1993, 115, 3080.
(26) Farnham, W. B.; Calabrese, J . C. J . Am. Chem. Soc. 1986, 108,
2449.

Catalytic Cycloisomerization
J . Org. Chem., Vol. 63, No. 26, 1998 9965
cyclization reactions,18,19 and given the apparent ability
of PhLi to initiate such processes, it seemed worthwhile
to attempt cycloisomerization of an unsaturated tertiary
iodide. The base sensitive 6-iodo-6-methyl-1-heptene (15)
system, used by Curran and Kim in their pioneering
studies of the radical-mediated atom transfer cycliza-
tion,19 was chosen as a representative substrate.
of commercial solutions (Aldrich) of phenyllithium (PhLi) in
cyclohexane-diethyl ether (7:3 by vol) were determined im-
mediately prior to use by the method of Watson and East-
ham.29 Product mixtures were analyzed by capillary GC on
one of the following columns: (1) a 25-m × 0.2-mm × 0.33-µm
film thickness, cross-linked, phenyl methyl (5%) silicone
column or (2) a 25-m × 0.2-mm × 0.33-µm film thickness,
cross-linked, methyl silicone gum column.
As shown below, simply stirring a 0.5 M solution of 15
in scrupulously dry and deoxygenated n-pentane-MTBE
(9:1 by vol) containing as little as 3 mol % of PhLi for 2
h at room temperature produces 2-iodomethyl-1,1-di-
methylcyclopentane (16) in 82% isolated yield. This result
leaves little doubt that PhLi-initiated cycloisomerization
of unsaturated organoiodides is of broad utility.
Standard procedures were employed for the preparation of
6-hepten-2-ol from 4-pentenyllithium20 and acetaldehyde in
90% yield; the known alcohol30 was then converted, via the
mesylate, to the known 6-iodo-1-heptene (8)30 in 62% yield. In
an analogous fashion, the known 2-methyl-6-hepten-2-ol15,31
was prepared from 4-pentenyllithium20 and acetone in 79%
yield.
N,N-Dia llyl-2-iod oa n ilin e (1). A mixture of 4.81 g (22.0
mmol) of 2-iodoaniline, 16.0 mL (0.185 mol) of allyl bromide,
9.26 g (87.3 mmol) of sodium carbonate, and 100 mL of DMF
was heated at reflux for 4 h. The mixture was allowed to cool
to room temperature, inorganic salts were removed by filtra-
tion, and the solid was washed with a generous portion of
diethyl ether (ca. 100 mL). The combined filtrate and washings
were poured into 200 mL of water, the layers were separated,
and the aqueous layer was extracted with three 60-mL
portions of diethyl ether. The combined organic extracts were
washed with water, dried (MgSO4), and concentrated by rotary
evaporation. Ku¨gelrohr distillation of the residue afforded 5.83
g (89%) of the iodide: bp (bath temp) 105 °C (0.3 mm); 1H NMR
(CDCl3) δ 3.60-3.62 (m, 4 H), 5.07-5.17 (m, 4 H), 5.78-5.84
(m, 2 H), 6.75 (m, 1 H), 7.00 (dd, J ) 7.98, 1.52 Hz, 1 H), 7.23-
7.27 (m, 1 H), 7.84 (dd, J ) 7.88, 1.56 Hz, 1 H); 13C NMR
(CDCl3) δ 56.14, 117.71, 124.21, 125.60, 125.89, 128.44, 134.82,
139.95, 151.78. Anal. Calcd for C12H14NI: C, 48.18; H, 4.72;
N, 4.68. Found: C, 48.15; H, 4.88; N, 4.84.
Con clu sion s
The chemistry detailed above represents a considerable
extension in scope of the novel PhLi-initiated cyclo-
isomerization of 5-hexenyl iodides disclosed in our pre-
liminary report.1 The results presented herein demon-
strate that primary, secondary, tertiary, or aryl iodides
tethered to a suitable positioned carbon-carbon π-bond
may be converted to their cyclic isomers in good to
excellent yield by simply stirring a solution of the iodide
in the presence of a catalytic quantity of PhLi. The
mechanism of the isomerization was found to be sub-
strate dependent: the PhLi-initiated 5-exo cycloisomer-
ization of aryl and (presumably) primary iodides involves
a three-step cascade (eqs 1-3) mediated by a series of
lithium-iodine exchange equilibria (eq 1 and 3) bracket-
ing an irreversible cyclization step (eq 2); isomerization
of secondary and tertiary substrates apparently entails
a radical-mediated atom transfer process initiated by
homolytic fragmentation of an ate-complex. Both mecha-
nistic scenarios share an ate-complex intermediate or
transition state as a common feature.
1-(3-Bu ten yl)-2-iod oben zen e (4). A solution of 4.63 g (21.9
mmol) of 2-bromo-1-(3-butenyl)benzene9a in 40 mL of diethyl
ether and 70 mL of n-pentane was cooled to -78 °C and 28.0
mL of a 1.73 M solution of t-BuLi (48.4 mmol) in pentane were
added dropwise over a 10-min period. The resulting mixture
was stirred at -78 °C for an additional 10 min and then was
allowed to warm and stand at room temperature for ca. 1 h to
remove residual t-BuLi.20 The cloudy, yellow solution was
recooled to -78 °C and a saturated solution of iodine in diethyl
ether was added dropwise until an orange color persisted in
the reaction mixture. The orange mixture was stirred at -78
°C for 5 min before the addition of 50 mL of 10% aqueous
sodium thiosulfate, the cooling bath was then removed, and
the slurry was allowed to warm to room temperature. The
resulting mixture was washed with 50 mL of water, dried
(MgSO4), and concentrated under reduced pressure. Ku¨gelrohr
distillation of the residue gave 5.15 g (91%) of the iodide: bp
In short, PhLi-initiated 5-exo cycloisomerization, de-
picted below, appears to be a process of broad synthetic
utility for the preparation of iodomethyl-substituted five-
membered rings. The isomerization reaction is operation-
ally simple, approaches near total atom economy, and
appears to be a general phenomenon for a variety of
structurally diverse unsaturated organoiodides.
1
(bath temp) 90 °C (1 mm) [lit.32 bp 130-132 °C (19 mm)]; H
NMR (CDCl3) δ 2.30-2.37 (m, 2 H), 2.77-2.81 (m, 2 H), 4.98-
5.09 (m, 2 H), 5.83-5.93 (m, 1 H), 6.87 (td, J ) 7.58, 1.89 Hz,
1 H), 7.19 (dd, J ) 7.58, 1.89 Hz, 1 H), 7.25 (td, J ) 7.58, 1.27
Hz, 1 H), 7.80 (dd, J ) 7.58, 1.27 Hz, 1 H); 13C NMR (CDCl3)
δ 34.13, 40.22, 100.54, 115.22, 127.72, 128.24, 129.43, 137.49,
139.46, 144.29.
1-(2-P r op en oxy)-2-p r op a n ol. Neat allyl alcohol, 13.5 mL
(0.198 mol), was slowly added to a suspension of 4.80 g (0.200
mol) of oil-free sodium hydride in 100 mL of dry THF and 50
mL of dry HMPA and the resulting mixture was heated at
reflux for 3 h. The alcoholate solution was allowed to cool to
room temperature and 5.72 g (0.130 mol) of propylene oxide
(freshly distilled from CaSO4) was then added over a 20-min
period. The resulting mixture was stirred for 30 min at room
Exp er im en ta l Section
Gen er a l P r oced u r es. Spectroscopic and chromatographic
procedures, methods used for the purification of reagents and
solvents, and precautions regarding the manipulation of
organolithiums have been previously described.5 It is to be
noted that all reagents (starting halides, internal standard,
and solvents) used in the PhLi-initiated cycloisomerizations
that had not been freshly distilled under argon were rendered
essentially oxygen-free before use by bubbling dry, deoxygen-
ated argon gas through the neat liquid for at least 5 min before
use: failure to follow this rather stringent protocol may result
in incomplete isomerization due to consumption of PhLi
through reaction with adventitious oxygen. The concentrations
(29) Watson, S. C.; Eastham, J . F. J . Organomet. Chem. 1967, 9,
165.
(30) Ashby, E. C.; DePriest, R. N.; Goel, A. B.; Wenderoth, B.; Pham,
T. N. J . Org. Chem. 1984, 49, 3545.
(31) (a) Colonge, J .; Lasfargues, P. Bull. Soc. Chim. Fr. 1962, 177.
(b) Surzur, J .-M.; Teissier, P. Bull. Soc. Chim. Fr. 1970, 3060.
(32) Beckwith, A. L. J .; Gara. W. B. J . Chem. Soc., Perkin Trans. 2
1975, 795.

9966 J . Org. Chem., Vol. 63, No. 26, 1998
Bailey and Carson
temperature then heated at reflux for 4 h before cautious
addition of 50 mL of water and 75 mL of 3 M aqueous
hydrochloric acid. The aqueous layer was extracted with
diethyl ether, and the combined organic extracts were washed
successively with water, aqueous sodium bicarbonate, and
brine, then dried (MgSO4), and concentrated under reduced
pressure. Vacuum distillation of the residue afforded 4.17 g
(28%) of the known alcohol:33 bp 66-70 °C (25 mm); 1H NMR
(CDCl3) δ 1.11-1.13 (m, 3 H), 2.37 (br, s, 1 H), 3.19-3.24 (m,
1 H), 3.39-3.43 (m, 1 H), 3.95-4.01 (m, 3 H), 5.14-5.28 (m, 2
H), 5.83-5.93 (m, 1 H); 13C NMR (CDCl3) δ 18.59, 66.26, 72.05,
75.69, 117.08, 134.44.
3-(2-Iod op r op yloxy)p r op en e (11). Following the general
procedure of Crossland and Servis,34 4.00 g (34.5 mmol) of 1-(2-
propenoxy)-2-propanol was converted to its mesylate. The
crude mesylate was added to a solution of 15.5 g (0.104 mol)
of anhydrous sodium iodide in 150 mL of dry acetone and the
resulting mixture was stirred for 36 h at room temperature
and then for 10 h at gentle reflux. Standard workup yielded
1.59 g (20% from the alcohol) of the iodide: 1H NMR (CDCl3)
iodide:1H NMR (CDCl3) δ 1.51-1.75 (m, 6 H), 1.84-1.89 (m,
1 H), 1.90 (d, J ) 6.84 Hz, 3 H), 2.10-2.14 (m, 2 H), 4.11-
4.19 (m, 1 H); 13C NMR (CDCl3) δ 3.42, 17.85, 28.92, 29.06,
29.62, 41.83, 76.00, 78.37; HRMS calcd for C8H13I 236.0062,
found 236.0061.
6-Iod o-6-m eth yl-1-h ep ten e (15). Neat bis(trimethylsilyl)-
trifluoroacetamide (BSTFA), 5.90 mL (22.2 mmol), was added
dropwise over ∼5 min period to a solution of 2.37 g (18.5 mmol)
of 2-methyl-6-hepten-2-ol15,31 in 40 mL of dry DMF. The
mixture was stirred at 95-100 °C for 5 h and excess BSTFA
was then destroyed by the addition of 40 mL of water. The
aqueous layer was extracted with three 10-mL portions of
pentane; the combined organic extracts were washed with
water and brine, dried (MgSO4), and concentrated under
reduced pressure. Ku¨gelrohr distillation of the residue afforded
3.00 g (81%) of 2-m eth yl-2-tr im eth ylsilyloxy-6-h ep ten e: bp
(bath temp) 105-110 °C (25 mm); 1H NMR (CDCl3) δ 0.08 (s,
9 H), 1.18 (s, 6 H), 1.23-1.25 (m, 2 H), 2.00-2.03 (m, 4 H),
4.90-5.01 (m, 2 H), 5.75-5.85 (m, 1 H); 13C NMR (CDCl3) δ
2.60, 23.74, 29.84, 34.24, 44.34, 73.91, 114.24, 139.12. The silyl
ether was converted to the title iodide in portions as needed.
Neat iodotrimethylsilane (freshly distilled from copper pow-
der), 0.710 mL (4.99 mmol), was added dropwise via syringe
over a 5-min period to a solution of 0.791 g (3.95 mmol) of the
silyl ether in 10 mL of dry benzene and the resulting yellow
solution was stirred for 30 min at room temperature. The
contents of the flask were then partitioned between 30 mL of
water and 30 mL of pentane, the aqueous layer was extracted
with three 5-mL portions of pentane, and the combined pink
organic extracts were washed successively with cold 10-mL
portions of saturated, aqueous sodium bicarbonate, 10% aque-
ous sodium thiosulfate, water, and brine, then dried (MgSO4),
and concentrated by rotary evaporation to give 0.861 g (92%)
of the known19 title iodide: 1H NMR (CDCl3) δ 1.55-1.61 (m,
4 H), 1.90 (s, 6 H), 2.06-2.10 (m, 2 H), 4.93-5.04 (m, 2 H),
5.77-5.84 (m, 1 H); 13C NMR (CDCl3) δ 27.72, 33.34, 38.05,
49.86, 52.24, 114.90, 138.31. The base-labile iodide was stored
in a freezer under argon and used within a few hours of
preparation.
Cycloisom er iza tion of N,N-Dia llyl-2-iod oa n ilin e (1) to
1-Allyl-3-iod om eth ylin d olin e (2). A 25-mL, one-necked,
round-bottomed flask, equipped with a Teflon-coated magnetic
stir bar and capped with a septum, was flame dried under an
atmosphere of argon. The flask was charged with 0.294 g
(0.983 mmol) of N,N-diallyl-2-iodoaniline (1), 0.260 mL of a
0.393 M solution of TMEDA (0.102 mmol) in MTBE, 9.0 mL
of n-pentane, and 1.0 mL of MTBE. The solution was cooled
to 0 °C and 0.060 mL of a 1.65 M solution of PhLi (0.099 mmol)
in cyclohexane-diethyl ether (7:3 by volume) was added in
one portion. The resulting light yellow solution was stirred at
0 °C for 1 h under a positive pressure of argon before the
addition of water (1 mL). The contents of the flask were
partitioned between 25 mL of pentane and 10 mL of water,
the layers were separated, and the organic layer was washed
with brine, dried (MgSO4), and concentrated under reduced
pressure. Purification of the residue by flash chromatography
on silica gel (1% ethyl acetate-2% triethylamine-hexanes)
gave 0.204 g (70%) of the title iodide: Rf ) 0.29; 1H NMR
(CDCl3) δ 3.16-3.23 (m, 2 H), 3.44-3.60 (m, 3 H), 3.69-3.71
(m, 2 H), 5.18-5.29 (m, 2 H), 5.82-5.91 (m, 1 H), 6.48 (d, J )
7.89 Hz, 1 H), 6.66 (dt, J ) 7.38, 0.96 Hz, 1 H), 7.07-7.24 (m,
2 H); 13C NMR (CDCl3) δ 10.04, 43.75, 51.33, 59.95, 107.84,
117.58, 117.67, 123.99, 128.68, 131.03, 133.58, 151.92; HRMS
calcd for C12H14NI 299.0171, found 299.0179.
δ 1.88 (d, J ) 6.87 Hz, 3 H), 3.41 (A portion of ABX, J AB
10.34 Hz, J AX ) 7.34 Hz, 1 H), 3.71 (B portion of ABX, J AB
)
)
10.34 Hz, J BX ) 6.02 Hz, 1 H), 4.01-4.05 (m, 2 H), 4.15-4.23
(m, 1 H), 5.16-5.30 (m, 2 H), 5.84-5.93 (m, 1 H); 13C NMR
(CDCl3) δ 24.29, 24.74, 71.81, 77.15, 117.35, 134.38; HRMS
calcd for C6H11OI 225.9854, found 225.9852.
6-Hep tyn -2-ol. A suspension of 8.65 g (0.360 mol) of oil-
free sodium hydride in 125 mL of 1,3-diaminopropane (freshly
distilled from sodium) was stirred at 70 °C for 1.5 h. The brown
mixture was then cooled to 55 °C and 5.02 g (44.7 mmol) of
4-heptyn-2-ol in 25 mL of 1,3-diaminopropane were added
dropwise. The resulting reddish-brown mixture was stirred at
55 °C for 3.5 h and then cooled to 0 °C before cautious,
successive addition of 50 mL of water and 300 mL of 10%
aqueous hydrochloric acid. The brown solution was then
continually extracted with 350 mL of diethyl ether for 22 h;
the ethereal extract was washed with water, dried (MgSO4),
and concentrated by rotary evaporation. Ku¨gelrohr distillation
of the orange residue afforded 2.70 g (54%) of the alcohol: bp
(bath temp) 80-90 °C (13 mm) [lit.35 bp 75-77 °C (12 mm)];
1H NMR (CDCl3) δ 1.13 (d, J ) 6.18 Hz, 3 H), 1.47-1.59 (m,
4 H), 1.90 (t, J ) 2.67 Hz, 1 H), 2.10 (br, s, 1 H), 2.13-2.17
(m, 2 H), 3.72-3.77 (m, 1 H); 13C NMR (CDCl3) δ 18.22, 23.35,
24.54, 30.02, 67.24, 68.39, 84.26.
6-Octyn -2-ol. A solution of lithium amide in ammonia was
prepared from 0.62 g (89 mmol) of lithium rod and 50 mg of
ferric nitrate nonahydrate in 100 mL of liquid ammonia at
-33 °C and 4.65 g (41.4 mmol) of 6-heptyn-2-ol was added
dropwise over a 20 min period. Neat iodomethane, 7.49 g (52.7
mmol), was then added and the ammonia was allowed to
evaporate overnight. Aqueous ammonium chloride solution (5
g/80 mL) was added to the resulting semisolid, the mixture
was extracted with five 40-mL portions of diethyl ether, and
the combined extracts were dried (MgSO4) and concentrated
under reduced pressure. Ku¨gelrohr distillation of the residue
yielded 4.28 g (82%) of the alcohol: bp 103-106 °C (12 mm)
[lit.36 bp 92 °C (10 mm)]; 1H NMR (CDCl3) δ 1.16 (d, J ) 6.02
Hz, 3 H), 1.46-1.58 (m, 5 H, including OH), 1.74 (t, J ) 2.56
Hz, 3 H), 2.09-2.14 (m, 2 H), 3.78 (apparent six-line pattern,
J ) 6.02 Hz, 1 H); 13C NMR (CDCl3) δ 3.33, 18.60, 23.42, 25.17,
38.31, 67.57, 75.70, 78.92.
7-Iod o-2-octyn e (13). Following the general procedure of
Crossland and Servis,34 4.10 g (32.5 mmol) of 6-octyn-2-ol was
converted to its mesylate. The crude mesylate was added to a
solution of 14.6 g (97.3 mmol) of anhydrous sodium iodide in
150 mL of dry acetone and the resulting mixture was stirred
for 22 h at room temperature and then for 3 h at gentle reflux.
Standard workup gave 5.99 g (78% from the alcohol) of the
Cycloisom er iza tion of 1-(3-Bu ten yl)-2-iod oben zen e (4).
A solution of 0.333 g (1.29 mmol) of 1-(3-butenyl)-2-iodoben-
zene (4) in 13.5 mL of n-pentane and 1.5 mL of MTBE,
containing 0.820 mL of a 0.391 M solution of TMEDA (0.320
mmol) in MTBE, was warmed to 60 °C under argon and 0.190
mL of a 1.71 M solution of PhLi (0.325 mmol) in cyclohexane-
diethyl ether (7:3 by volume) was added in one portion. The
resulting light yellow solution was stirred at 60 °C for 7 h
under a positive pressure of argon before the addition of water
(1 mL). The organic layer was washed with water, dried
(MgSO4), and concentrated under reduced pressure. GC analy-
(33) Kablaoui, N. M.; Buchwald, S. L. J . Am. Chem. Soc. 1996, 118,
3182.
(34) Crossland, R. K.; Servis, K. L. J . Org. Chem. 1970, 35, 3195.
(35) Peterson, P. E.; Kamat, R. J . J . Am. Chem. Soc. 1969, 91, 4521.
(36) Hanack, M.; Harding, C. E.; Derocque, J .-L Chem. Ber. 1972,
105, 421.

Catalytic Cycloisomerization
J . Org. Chem., Vol. 63, No. 26, 1998 9967
sis on column (1) using temperature programming (100 °C for
5 min, 10 °C/min to 250 °C for 10 min), as well as GC/MS
analysis, revealed that the product mixture contained, in order
of elution, (3-butenyl)benzene (7, 13%), 3-methylindene (6,
17%), 1-(3-butenyl)-2-iodobenzene (4, 27%), and 1-iodometh-
ylindan37 (5, 43%). Products were identified by comparison of
their mass spectra with those reported for authentic samples.37,38
Cycloisom er iza tion of 6-Iod o-1-h ep ten e (8) to cis- a n d
tr a n s-2-Iod om eth yl-1-m eth ylcyclop en ta n e (9). A solution
of 1.12 g (5.00 mmol) of 6-iodo-1-heptene (8) in 9.0 mL of
n-pentane and 1.0 mL of MTBE was stirred at room temper-
ature under argon and 0.075 mL of a 1.93 M solution of PhLi
(0.14 mmol) in cyclohexane-diethyl ether (7:3 by volume) was
added in one portion. The resulting yellow solution was then
stirred at 22 °C for 22 h under a positive pressure of argon.
Water (1 mL) was added to the reaction mixture, the layers
were separated, and the organic layer was washed with brine
and dried (Na2SO4). Evaporation of the solvent under reduced
pressure afforded 0.995 g (89%) of the known iodides14 as a
mixture of isomers (cis/ trans ) 3.3). The diastereoisomers
were separated by preparative GC on a 10-ft, 10% SE-30 on
Anakrom A (60/80 mesh) column at 122 °C and were identified
on the basis of the following spectroscopic properties. cis-2-
Iodom eth yl-1-m eth ylcyclopen tan e (cis-9): 1H NMR (CDCl3)
δ 0.80 (d, J ) 7.12 Hz, 3 H), 1.28-1.40 (m, 2 H), 1.59-1.84
(m, 4 H), 2.07-2.17 (m, 1 H), 2.21-2.31 (m, 1 H), 3.33 (A
portion of ABX, J AB ) 9.47 Hz, J AX ) 8.08 Hz, 1 H), 3.39 (B
portion of ABX, J AB ) 9.47 Hz, J BX ) 7.84 Hz, 1 H); 13C NMR
(CDCl3) 9.51, 14.07, 22.79, 30.79, 33.37, 36.74, 46.76. tr a n s-
2-Iod om eth yl-1-m eth ylcyclop en ta n e (tr a n s-9): 1H NMR
(CDCl3) δ 0.97 (d, J ) 6.24 Hz, 3 H), 1.20-1.37 (m, 2 H), 1.46-
1.62 (m, 4 H), 1.86-1.92 (m, 2 H), 2.97 (A portion of ABX, J AB
GC/MS analysis, revealed that the reaction had produced allyl
alcohol (∼100%), iodobenzene (∼100%), and propene (charac-
terized by GC/MS). Reaction products were identified by
comparison of their GC retention times and mass spectra with
those of authentic samples; the yields of iodobenzene and allyl
alcohol were corrected for detector response under the condi-
tions of the analysis using samples of pure product and
standard.
Cycloisom er iza tion of 7-Iod o-2-octyn e (13) to Z- a n d
E-2-(1-Iod oeth ylid en e)-1-m eth ylcyclop en ta n e (14). A so-
lution of 0.576 g (2.44 mmol) of 7-iodo-2-octyne (13) in 4.5 mL
of cyclohexane and 0.5 mL of MTBE was warmed to 50 °C
under argon and 0.260 mL of a 1.40 M solution of PhLi (0.364
mmol) in cyclohexane-diethyl ether (7:3 by volume) was added
in one portion. The resulting light yellow solution was stirred
at 50 °C for 18 h under a positive pressure of argon. Water (1
mL) was then added to the reaction mixture and the organic
layer was washed with brine and dried (Na2SO4). Removal of
the solvent under reduced pressure gave 0.530 g (91%) of the
vinyl iodides as a mixture of diastereoisomers (Z/ E ) 1.4):27
1H NMR (CDCl3, mixture of diastereoisomers) δ 0.97 (d, J )
7.16 Hz, 3 H), 1.03 (d, J ) 7.79 Hz, 3 H), 1.44-1.90 (m, 8 H),
2.19-2.47 (m, 10 H), 2.61-2.68 (m, 1 H), 2.79-2.86 (m, 1 H);
13C NMR (CDCl3, mixture of diastereoisomers) δ 18.30, 19.34,
23.01, 25.10, 30.12, 30.45, 30.65, 33.55, 36.19, 37.19, 40.73,
45.09, 86.36, 89.05, 152.61, 153.60. Anal. Calcd for C8H13I: C,
40.70; H, 5.55; I, 53.75. Found: C, 40.85; H, 5.55.
Cycloisom er iza tion of 6-Iod o-6-m eth yl-1-h ep ten e (15)
to 2-Iod om eth yl-1,1-d im eth ylcyclop en ta n e (16). A solu-
tion of 0.835 g (3.52 mmol) of 6-iodo-6-methyl-1-heptene (15)
in 6.3 mL of n-pentane and 0.7 mL of MTBE was stirred at
room temperature under argon and 0.080 mL of a 1.93 M
solution of PhLi (0.15 mmol) in cyclohexane-diethyl ether (7:3
by volume) was added in one portion. The resulting light yellow
solution was stirred at 22 °C for 2 h under a positive pressure
of argon before the addition of 1 mL of water. The organic layer
was washed with brine, dried (Na2SO4), and concentrated
under reduced pressure to yield 0.686 g (82%) of the known19
iodide: 1H NMR (CDCl3) δ 0.75 (s, 3 H), 1.01 (s, 3 H), 1.33-
1.63 (m, 4 H), 1.84-1.92 (m, 2 H), 2.07-2.16 (m, 1 H), 2.56 (A
portion of ABX, J AX ) 11.52 Hz, J AB ) 9.22 Hz, 1H), 3.64 (B
portion of ABX, J AB ) 9.22 Hz, J BX ) 3.76 Hz, 1 H); 13C NMR
(CDCl3) δ 9.02, 20.17, 21.18, 28.14, 32.55, 41.87, 42.77, 52.93.
) 9.51 Hz, J AX ) 7.58 Hz, 1 H), 3.49 (B portion of ABX, J AB
)
9.51 Hz, J BX ) 4.02 Hz, 1 H); 13C NMR (CDCl3) δ 13.84, 19.09,
22.96, 33.69, 35.27, 40.75, 49.47.
R ea ct ion of 3-(2-Iod op r op yloxy)p r op en e (11) w it h
P h Li. A solution of 67.7 mg (0.299 mmol) of 3-(2-iodoprop-
yloxy)propene (11) in 2.7 mL of cyclohexane and 0.3 mL of
MTBE containing 46.3 mg (0.326 mmol) of n-decane (internal
standard) was cooled to 0 °C, and 0.180 mL of a 1.93 M solution
of PhLi (0.347 mmol) in cyclohexane-diethyl ether (7:3 by
volume) was added dropwise. The resulting solution was
stirred at 0 °C for 15 min before addition of just enough
saturated, aqueous ammonium chloride to form a clear organic
layer, which was dried (MgSO4). GC analysis on column (1)
using temperature programming (30 °C for 16 min, 50 °C/min
to 100 °C for 1 min, 10 °C/min to 250 °C for 5 min), as well as
Ack n ow led gm en t. We are grateful to Professor
Dennis P. Curran for providing details on the prepara-
tion of 6-iodo-6-methyl-1-heptene (15). This work was
supported by a grant from the Connecticut Department
of Economic Development.
(37) Beckwith, A. L. J .; Meijs, G. F. J . Org. Chem. 1987, 52, 1922.
(38) Heller, S. R.; Milne, G. W. A. EPA/ NIH Mass Spectra Data
Base; U. S. Government Printing Office: Washington, 1978.
J O981991Q