Synthesis of fused polycycles by 1,4-palladium migration chemistry

Article abstract of DOI:10.1021/jo048788h

Novel palladium migration/arylation methodology for the synthesis of complex fused polycycles has been developed, in which one or more sequential Pd-catalyzed intramolecular migration processes involving C-H activation are employed. The chemistry works best with electron-rich aromatics, which is in agreement with the idea that these palladium-catalyzed C-H activation reactions parallel electrophilic aromatic substitution.

Full text of DOI:10.1021/jo048788h

Synthesis of Fused Polycycles by 1,4-Palladium Migration
Chemistry
Qinhua Huang, Marino A. Campo, Tuanli Yao, Qingping Tian, and Richard C. Larock*
Department of Chemistry, Iowa State University, Ames, Iowa 50011
larock@iastate.edu
Received July 15, 2004
Novel palladium migration/arylation methodology for the synthesis of complex fused polycycles
has been developed, in which one or more sequential Pd-catalyzed intramolecular migration
processes involving C-H activation are employed. The chemistry works best with electron-rich
aromatics, which is in agreement with the idea that these palladium-catalyzed C-H activation
reactions parallel electrophilic aromatic substitution.
SCHEME 1
Introduction
The ability of palladium to activate C-H bonds has
been used extensively in organic synthesis in recent years
due to the wide variety of reactions this metal will
catalyze.1 For instance, catalytic amounts of Pd salts have
been used to effect the addition of C-H bonds of electron-
rich arenes to alkenes and alkynes and to effect carbonyl-
ation.1a,2-4 We have previously reported the synthesis of
9-benzylidene-9H-fluorenes by Pd-catalyzed intramolecu-
lar C-H activation involving the rearrangement of
organopalladium intermediates derived from aryl halides
and internal alkynes.3b Similarly, intramolecular C-H
activation in organopalladium intermediates derived
from o-halobiaryls leads to a 1,4-palladium migration
(Scheme 1).3a,5 We have already shown that such arylpal-
ladium intermediates can be trapped by Heck and alkyne
annulation reactions.5 We have recently reported that
this aryl-to-aryl palladium migration process, followed
by arylation, provides a novel, new route to a wide variety
of carbocycles and heterocycles.6 Herein, we wish to
report further details on this aryl-aryl migration and
also vinylic-aryl migration chemistry, followed by in-
tramolecular arylation.
Our strategy involves palladium C-H activation and
1,4-palladium migration within a biaryl, which generates
key arylpalladium intermediates which subsequently
undergo C-C bond formation by intramolecular arylation
producing fused polycycles (Scheme 2). This process
represents a very powerful new tool for the preparation
of complex molecules, which might be difficult to prepare
by any other present methodology.
(1) (a) Ritleng, V.; Sirlin, C.; Pfeffer, M. Chem. Rev. 2002, 102, 1731.
(b) Li, C.-J. Acc. Chem. Res. 2002, 35, 533. (c) Catellani, M.; Chiusoli,
G. P. J. Organomet. Chem. 1983, 250, 509. (d) Oyamada, J.; Jia, C.;
Fujiwara, Y. Chem. Lett. 2002, 2. (e) Oyamada, J.; Jia, C.; Fujiwara,
Y. Chem. Lett. 2002, 380. (f) Jia, C.; Piao, D.; Kitamura, T.; Fujiwara,
Y. J. Org. Chem. 2000, 65, 7516.
(2) For reviews, see: (a) Fujiwara, Y.; Jia, C. Handbook of Orga-
nopalladium Chemistry for Organic Synthesis; John Wiley & Sons; New
York, 2002; Vol. 2, p 2859. (b) Jia, C.; Kitamura, T.; Fujiwara, Y. Acc.
Chem. Res. 2001, 34, 633. (c) Catellani, M. Synlett 2003, 298. (d) Dyker,
G. Chem. Ber. Recl. 1997, 130, 1567.
(3) For palladium migration, see: (a) Karig, G.; Moon, M.-T.;
Thasana, N.; Gallagher, T. Org. Lett. 2002, 4, 3115. (b) Tian, Q.; Larock,
R. C. Org. Lett. 2000, 2, 3329. (c) Ca´mpora, J.; Lo´pez, J. A.; Palma, P.;
Valerga, P.; Spillner, E.; Carmona, E. Angew. Chem., Int. Ed. 1999,
38, 147. (d) Markies, B. A.; Wijkens, P.; Kooijman, H.; Spek, A. L.;
Boersma, J.; van Koten, G. J. Chem. Soc., Chem. Commun. 1992, 1420.
(e) Catellani, M.; Chiusoli, G. P. J. Organomet. Chem. 1985, 279, 225.
(f) Huang, Q.; Fazio A.; Dai G.; Campo, M. A.; Larock, R. C. J. Am.
Chem. Soc. 2004, 126, 7460.
Results and Discussion
To obtain an optimum set of reaction conditions for
palladium migration, we have reinvestigated the pal-
ladium-catalyzed transformation of 1-iodo-1,2,2-triphen-
ylethene (1) to 9-benzylidene-9H-fluorene (2) as our
model system3b (Table 1). While this system may not be
the most obvious for a study of aryl-to-aryl Pd migrations,
we had previously accumulated substantial data on this
system. To begin with, we carried out this reaction using
our previously reported conditions3b and obtained the
desired compound 2 in a 73% yield, along with triphen-
ylethene (3) in a 15% yield (entry 1).
(4) For rhodium migration, see: Oguma, K.; Miura, M.; Satoh, T.;
Nomura, M. J. Am. Chem. Soc. 2000, 122, 10464.
(5) Campo, M. A.; Larock, R. C. J. Am. Chem. Soc. 2002, 124, 14326.
(6) Campo, M. A.; Huang, Q.; Yao, T.; Tian, Q.; Larock, R. C. J. Am.
Chem. Soc. 2003, 125, 11506.
10.1021/jo048788h CCC: $27.50 © 2004 American Chemical Society
Published on Web 10/26/2004
J. Org. Chem. 2004, 69, 8251-8257
8251

Huang et al.
SCHEME 2
TABLE 1. Palladium-Catalyzed Cyclization of 1-Iodo-
1,2,2-triphenylethene (1) to 9-Benzylidenefluorene (2)a
and DMSO (entries 5-7). DMSO gave none of the desired
fluorene 2 or any reduction product 3. The amount of
reduction was more or less the same in the other solvents.
Thus, we continued our investigation using DMF as the
reaction solvent.
We also examined the effect of various bases on the
yields of 2 and 3, including organic bases, such as
pyridine and diisopropylethylamine (entries 8 and 9).
These bases were ineffective in promoting the reaction
and TLC analysis of the reaction mixtures indicated only
the presence of the starting vinylic halide 1. The use of
Na2CO3 as the base provided 2 in a 70% yield, but we
also obtained reduced product 3 in a 26% yield (entry
10). The base NaHCO3 provided a 65% yield of 2 and a
23% yield of 3 (entry 11). We believe that the solubility
of these bases in the reaction mixture may be playing a
critical role in determining the outcome. Thus, once again
we used Na2CO3 as the base, but this time we added 1
equiv of NaI, which is completely soluble in DMF, as an
additive to promote a sodium common ion effect intended
to make Na2CO3 less soluble in the reaction mixture.
Indeed, this experiment revealed that under such reac-
tion conditions only the reduced product 3 was produced
in a 47% yield (entry 12). None of the desired product 2
was observed (compare entries 10 and 12). Although
there may be a number of other effects going on under
these reaction conditions, it seemed logical to assume that
the yield of the reaction would improve by using more
soluble inorganic bases. Thus, the use of Cs2CO3, which
presumably has better solubility than other alkali car-
bonates in DMF,7 provided a slightly higher 74% yield
of compound 2, along with a 22% yield of the reduced
product 3 (entry 13). Similarly, the use of very soluble
CsOAc as the base provided a 90% yield of the desired
product 2, along with 4% of the reduced product 3 after
2 d (entry 14). We subsequently found that cesium
pivalate (CsO2CCMe3), unlike any other previously stud-
ied base, was completely soluble in DMF at 100 °C. In
this case, we obtained the desired compound 2 in an
impressive 96% yield, along with only a small amount of
the reduced product 3 (4%) after only 1 d (entry 15).
Clearly, cesium pivalate is far superior as a base in this
reaction, and its high solubility in DMF seems to be one
of the reasons to explain this phenomena. To illustrate,
we carried out the reaction of 1 under conditions identical
with those described in entry 15, but we used DMA
%
%
ligand
chloride
time yield yield
entry
(mol %)
base
source solvent (d) of 2 of 3
1
2
3
4
5
6
7
8
9
PPh3 (10)
P(o-tol)3 (10) NaOAc
NaOAc
TBACb DMF
TBAC DMF
TBAC DMF
DMF
3
3
3
1
1.5
1
1
1
1
1
1
1
1
2
1
1
1
1
73 15
75 20
79 15
47 47
52 48
46 46
dppmc(5)
dppm (5)
dppm (5)
dppm (5)
dppm (5)
dppm (5)
dppm (5)
NaOAc
NaOAc
NaOAc
NaOAc
NaOAc
pyridine
i-Pr2NEt
Na2CO3
NaHCO3
DMA
NMP
DMSO
DMF
DMF
10 dppm (5)
11 dppm (5)
12 dppm (5)
13 dppm (5)
14 dppm (5)
15 dppm (5)
16 dppm (5)
17 dppef(5)
18 dppm (5)
DMF
70 26
65 23
47
DMF
d
Na2CO3
DMF
Cs2CO3
DMF
74 22
CsOAc
DMF
90
96
4
4
CsO2CCMe3
CsO2CCMe3
CsO2CCMe3
n-Bu4NOAc
DMF
DMA
59 17e
90 10
DMF
DMF
<10
a The reaction was run with 0.25 mmol of 1-iodo-1,2,2-triphen-
ylethene (1), 5 mol % of Pd(OAc)2 and 4 mL of solvent at 100 °C.
b TBAC ) n-Bu4NCl. c Dppm ) 1,1-bis(diphenylphosphino)methane.
d 1 equiv of NaI was added. e 24% of 1 was recovered. f Dppe )
1,2-bis(diphenylphosphino)ethane.
The first variable to be examined was the ligand on
palladium. We examined phosphines other than PPh3.
Entries 2 and 3 indicate that P(o-tol)3 and CH2(PPh2)2
(dppm) are superior to PPh3 in affording higher yields of
the fluorene 2. However, the reactions with all of these
phosphine ligands required 3 d to reach completion. To
shorten this relatively lengthy reaction time, we elimi-
nated n-Bu4NCl (TBAC) and found that the reaction was
complete after only 1 d (entry 4). It is likely that
coordination of chloride to palladium increases the stabil-
ity and electronic density of the palladium intermediate,
and thus lowers its reactivity toward the C-H bond,
which subsequently slows the Pd migration. Unfortu-
nately, this led to a much lower yield of the desired
compound 2 (47%), and the yield of reduced product 3
increased to 47%. To investigate whether DMF was the
hydride source for the reduction of 1 to 3, we carried out
this reaction using other solvents, such as DMA, NMP,
(7) Satoh, T.; Inoch, J.; Kawamura, Y.; Miura, M.; Nomura, M. Bull.
Chem. Soc. Jpn. 1998, 71, 2239.
8252 J. Org. Chem., Vol. 69, No. 24, 2004

Synthesis of Fused Polycycles
instead of DMF as the solvent, in which cesium pivalate
is not completely soluble. Under these conditions, we
obtained the desired compound 2 in a relatively low 59%
yield, along with the reduced product 3 in a 17% yield
(entry 16). Twenty four percent of the starting vinylic
iodide 1 was also obtained. Finally, to test whether dppm
was indeed critical to this reaction, we carried out the
transformation using another chelating phosphine ligand,
namely 1,2-bis(diphenylphosphino)ethane (dppe), and we
obtained compound 2 in a 90% yield, along with reduced
product 3 in a 10% yield (entry 17).
FIGURE 1. Unfavorable seven-membered-ring palladacycle
intermediates.
chlorophenoxy)-2-iodobiphenyl (8) under our standard
conditions, while leaving the chloro functionality intact
(entry 3).
As a result of this optimization work, our optimal set
of reaction conditions for this transformation are those
listed in entry 15 of Table 1. Notice that the newly
developed reaction conditions catalyze the transformation
of 1 to 2 in high yield and much shorter reaction time
than our earlier reported procedure (entry 1).3b The
variable most critical to the success of this process
appears to be the highly soluble cesium pivalate base.
Surprisingly, the use of n-Bu4NOAc as the base, which
is also completely soluble in DMF under reaction condi-
tions identical with those described in entry 15, failed to
promote this reaction, affording only trace amounts of
the desired product 2 after 1 d (entry 18). Thus, not only
the solubility, but also the exact nature of the base
appears critical in determining the yield of fluorene. It
is interesting to note that the work of Buchwald, Hartwig,
and Fu has demonstrated that steric congestion imposed
on palladium by bulky, electron-rich ligands facilitates
both the oxidative addition and reductive elimination
steps involving palladium, and gives rise to more effective
catalyst systems.8 However, nothing is apparently known
about the effects of using a sterically hindered base, such
as pivalate, in palladium chemistry, and whether it may
give similar results to those obtained with bulky ligands.
With an apparently “optimal” set of reaction conditions
for palladium migration chemistry at our disposal, we
proceeded to study the sequential Pd-catalyzed migration/
arylation of various 3′-substituted 2-iodobiphenyls (Table
2). We began by allowing 3′-benzyl-2-iodobiphenyl (4) to
react under our standard reaction conditions at 100 °C,
but after 2 d this substrate failed to react. However, by
simply increasing the reaction temperature to 110 °C,
we were able to obtain the desired compound 5 in a 40%
yield (entry 1). The disappointingly low yield obtained
with this substrate might be explained by the poor
reactivity of the benzyl moiety as an intramolecular trap.
To test this idea, we carried out the reaction with the
more electron-rich 2-iodo-3′-phenoxybiphenyl (6) and
obtained the desired 4-phenyldibenzofuran (7) in an
impressive 89% yield (entry 2). Clearly, these results
indicate that the electron-rich oxygen-substituted phenyl
ring is superior as an arene trap. Our finding that
electron-rich arenes are superior to electron-neutral
arene traps is consistent with literature reports indicat-
ing that the ease of C-H activation by palladium
parallels electrophilic aromatic substitution.9 Similarly,
we have been able to selectively obtain 3-chloro-5-
phenyldibenzofuran (9) in an 82% yield from 3-(p-
Motivated by the ease of preparation of the following
starting materials and by the knowledge that electron-
rich arenes are apparently superior as intramolecular
traps for our arylpalladium intermediates, we synthe-
sized the indole derivatives 10 and 12. To our great
satisfaction, compound 10 smoothly underwent the de-
sired reaction, producing the relatively strained isoindo-
loindole 11 in a 70% yield (entry 4). Surprisingly,
compound 12 produced the strained and sterically con-
gested 2-methylisoindoloindole 13 in a comparable 71%
yield (entry 5).
We next examined the possibility of using an intramo-
lecular arylation to form six-membered rings. Unfortu-
nately, 3-(2-iodophenyl)benzyl phenyl ether (14) failed to
react under our standard reaction conditions. Even at 110
°C, the reaction was sluggish, so the temperature was
increased to 120 °C, in which case the reaction was
complete after 2 d. Unfortunately, an inseparable 60:40
mixture of the desired compound 15 and the reduced
product 16 was obtained in a 75% overall yield (entry 6).
Clearly, the formation of a six-membered ring is not as
favorable as five-membered ring formation (compare
entries 2 and 6). This might be due to the difficulty in
forming a seven-membered-ring palladacycle (Figure 1).
We proceeded to investigate the sequential migration/
arylation reaction of more complex polyaromatic com-
pounds. In theory, 2-iodo-1-phenylnaphthalene (17) should
afford fluoranthene (18) using our methodology. Mecha-
nistically, the palladium must first undergo a 1,4-
palladium migration from the 2-position of the naphtha-
lene to the ortho position of the phenyl substituent,
followed by arylation at the 8-position of the naphthalene
(Scheme 3). Although the reaction did not proceed at 100
°C, at 110 °C compound 17 produced the desired com-
pound 18 in an 81% yield (entry 7). Similarly, 2-bromo-
1-phenylnaphthalene (19) produced the desired fluoran-
thene (18) in a 70% yield, indicating that this aryl
bromide also undergoes the desired transformation, but
in a somewhat lower yield and a longer reaction time.
Another interesting example of this migration/aryla-
tion chemistry involves the rearrangement of easily pre-
pared 9-iodo-10-arylphenanthrenes10 to benz[e]acephen-
anthrylenes (entries 9-12). In this case, the palladium
migrates from the 9-position of the phenanthrene to the
ortho position of the aryl substituent, followed by cy-
(9) For mechanistic studies suggesting electrophilic species in
intramolecular C-H activation processes employing palladium, see:
(a) Martin-Matute, B.; Mateo, C.; Cardenas, D. J.; Echavarren, A. M.
Chem. Eur. J. 2001, 7, 2341 and references therein. (b) Catellani, M.;
Chiusoli, G. P. J. Organomet. Chem. 1992, 425, 151 and references
therein.
(8) (a) Old, D. W.; Harris, M. C.; Buchwald, S. L. Org. Lett. 2000, 2,
1403. (b) Mann, G.; Incarvito, C.; Rheingold, A. L.; Hartwig, J. F. J.
Am. Chem. Soc. 1999, 121, 3224. (c) Littke, A. F.; Fu, G. C. J. Am.
Chem. Soc. 2001, 123, 6989.
(10) Yao, T.; Campo, M. A.; Larock, R. C. Org. Lett. 2004, 6, 2677.
J. Org. Chem, Vol. 69, No. 24, 2004 8253

Huang et al.
TABLE 2. Sequential 1,4-Palladium Migration, Followed by Intramolecular Arylationa
a The reaction was carried out under the following standard conditions employing 0.25 mmol of the aryl halide, 5 mol % of Pd(OAc)2,
5 mol % of dppm, and 2 equiv of CsO2CCMe3 in DMF (4 mL) at 100 °C unless otherwise noted. b The yield in parentheses corresponds to
the GC yield of product in which the C-I bond has been reduced to a C-H. c The reaction was run at 110 °C. d The reaction temperature
was increased to 120 °C. e The yield was determined by 1H NMR spectroscopy.
8254 J. Org. Chem., Vol. 69, No. 24, 2004

Synthesis of Fused Polycycles
SCHEME 3
clization onto the 1-position of the phenanthrene. Indeed,
the reaction of 9-iodo-10-phenylphenanthrene (20) under
our standard reaction conditions at 110 °C produced the
desired benz[e]acephenanthrylene (21) in a 78% yield
(entry 9). We proceeded to investigate electronic effects
in this phenanthrene reaction by looking at different
substituents on the phenyl moiety. As expected, the use
of an electron-donating methoxy group in compound 22
gave a good yield (71%) of the corresponding benz[e]-
acephenanthrylene 23, although the yield was slightly
lower than that of the parent system (entry 10). As
expected, the introduction of an electron-withdrawing
CO2Et group in the para position of the phenyl substitu-
ent was detrimental to the reaction, producing compound
25 in only a 50% yield. All of these phenanthrene
reactions gave approximately a 20% yield of the corre-
sponding reduction product.
We have also studied the regioselectivity of the migra-
tion by using an m-tolyl moiety in the 10-position of the
9-iodophenanthrene (entry 12). Compound 26 has two
available positions for palladium migration, the more
sterically congested neighboring 2-position or the remote
6-position of the phenyl ring. The palladium-catalyzed
cyclization of compound 26 generated compound 27
exclusively in a 56% yield, alongside a significant amount
of reduction product (37%). This result indicates that
palladium migration occurs exclusively onto the less
sterically congested 6-position of the phenyl moiety and
that the presence of a methyl group apparently com-
pletely inhibited migration to the more hindered 2-posi-
tion or at least cyclization of that intermediate to the
corresponding polycyclic product.
We next tried to prepare the more strained fused
thiophene 29 from phenanthrene 28. Unfortunately, this
reaction led to a very complex mixture, which produced
none of the desired compound 29. Besides the unfavor-
able ring strain associated with the final product 29,
intramolecular sulfur chelation of the intermediate 10-
(thiophen-2-yl)phenanthren-9-ylpalladium iodide might
be inhibiting the palladium migration step (Figure 2).
The relatively electron-rich benzo[e]phenanthrene 30
also underwent the migration/arylation reaction, produc-
ing the highly conjugated hexacyclic compound 31 in a
65% yield (entry 14). Unfortunately, compound 32 failed
to generate the desired hexacycle 33 under our reaction
conditions at 110 °C (entry 15). Only reduction product
was isolated in an 80% yield. This example once again
FIGURE 2. Intramolecular palladium chelation by sulfur.
indicates that intramolecular cyclization to form a six-
membered ring is apparently unfavorable.
Having studied the palladium-catalyzed transforma-
tion of a variety of polycyclic aromatic halides, we
switched our attention to heterocyclic aromatic com-
pounds. To begin with, we carried out the reaction of
3-iodo-4-phenylquinoline (34) under our standard reac-
tion conditions at 100 °C, but after 2 d this substrate
failed to react. By simply increasing the reaction tem-
perature to 110 °C, we obtained the desired indeno[1,2,3-
de]quinoline (35) in a 54% yield. Again the modest yield
obtained with this electron-deficient substrate is consis-
tent with our previous observations that electron-
deficient substrates do not perform as well as more
electron-rich substrates (compare entries 7 and 16). We
also allowed 2-iodo-1-methyl-3-phenylindole (36) to react
under our standard reaction conditions at 110 °C, but
failed to obtain the desired tricyclic compound 37 (entry
17). A similar negative result was obtained with iodoin-
dole 38 (entry 18). The poor results obtained with
substrates 36 and 38 can be explained in terms of the
unfavorable ring strain of the corresponding products 37
and 39. In addition, we have previously established, using
Heck trapping experiments, that palladium prefers to
reside on the indole moiety in such substrates.5
To confirm our suspicion that the palladium prefers
to migrate to a more electron-rich position, because of
the relatively easy activation of an electron-rich C-H
bond,5,9 compound 40 was allowed to react under our
migration conditions and indole 41 was produced in a
92% yield in 1 d at 100 °C (entry 19). From the results of
entries 1 and 19, it appears that the high efficiency of
palladium migration to a relatively electron-rich ring
allows the sequential migration/arylation to proceed
smoothly at a lower temperature and in a shorter
reaction time, although the benzyl group does not appear
to be a particularly good arylating agent. When a
methoxy group was introduced into the meta position of
the benzyl group, this migration/arylation reaction af-
J. Org. Chem, Vol. 69, No. 24, 2004 8255

Huang et al.
SCHEME 4
SCHEME 5
tacyclic compound 50 in an 80% yield (entry 23). On the
other hand, treating the electron-deficient 3-iodo-4-
phenylisocoumarin (51) under our standard reaction
conditions gave a complex mixture, and we failed to
isolate any of the desired tricyclic compound 52 (entry
24). This disappointing result was not unexpected, since
our previous experience with compound 51 has indicated
that palladium easily catalyzes its decomposition. Our
last attempt to generate polycycles from vinylic iodides
involved the use of isoquinolone 53 (entry 25). This
substrate suffers the disadvantage that the intramolecu-
lar arylation step requires the formation of a six-
membered ring. As expected, the reaction of substrate
53 under our standard reaction conditions at 110 °C
failed to produce the desired pentacyclic product 54. After
5 d of reaction, we were only able to isolate the reduction
product N-phenyl-3-phenylisoquinolone12 in a 12% yield.
A mechanistically interesting question is whether the
arylpalladium intermediate can migrate more than once
and still effect synthetically useful chemistry. To examine
this possibility, 2-iodo-5-phenoxybiphenyl (55) was al-
lowed to react under our migration conditions and double
migration product 4 was isolated in an 88% yield (entry
26 and Scheme 5). Mechanistically, the palladium first
inserts into the aryl iodide bond to form intermediate A,
which migrates to the phenyl unit by through-space C-H
activation. The metal moiety in the first migration
intermediate B can either return to the original aromatic
ring in the position from which it originally migrated (A)
or migrate to the position ortho to the phenoxy group (C),
where it can be trapped by arylation. Note that the yield
for this double migration chemistry is very similar to that
from the single migration chemistry (entry 2) and the
success of this double palladium migration indicates that
multiple migration processes are entirely feasible.
forded a mixture of indoles 43 and 44 in 78% and 12%
yields, respectively (entry 20). This result is consistent
with our previous observation (see entry 12) that the
palladium intermediate is more likely to form the final
carbon-carbon bond at the less hindered position of the
aryl terminus. When compound 45, which has an electron-
deficient aryl terminus, was allowed to react under the
standard migration conditions, product 46 was isolated
in 33% yield (entry 21). This is consistent with electron-
deficient termini giving lower yields.
We have also examined the possibility of trapping
palladium migration species by alkynes. Thus, we have
carried out the palladium-catalyzed sequential migration/
alkyne insertion/arylation of aryl halide 47 in the hope
that the arylpalladium intermediate generated by a 1,4-
Pd shift via through-space C-H activation could be
trapped by alkyne insertion-annulation chemistry de-
scribed earlier by us (Scheme 4).11 The reaction was
carried out under our standard migration conditions and
carbazole 48 was isolated in a 65% yield (entry 22).
Importantly, this reaction was complete in 0.5 d at 100
°C, consistent with the particularly facile migration of
Pd to the electron-rich indole ring system. Although we
cannot rule out the possibility that this reaction is
proceeding by direct endocyclic addition of the initial
arylpalladium species to the alkyne triple bond and
subsequent ring closure onto the indole to give product
48, this seems unlikely since exocyclic addition is more
common. This successful alkyne insertion chemistry
suggests that there is the exciting possibility of trapping
aryl- and other organopalladium intermediates generated
by a 1,4-Pd shift by many other synthetically useful
palladium methodologies, such as amination and annu-
lation. We are presently examining such possibilities.
While our efforts have focused on synthesizing poly-
cyclic compounds in which the key 1,4-palladium shift
occurs from an aryl to another aryl position, we wanted
to establish that our methodology could also make use
of a vinylic to aryl palladium migration to generate the
key intermediate for the intramolecular arylation step.
We have already shown one example of such a transfor-
mation in converting compound 1 to 2 (Table 1). Another
illustration of this process involves the use of 9-iodo-10-
phenyldibenz[b,f]oxepine (49). The reaction of this rela-
tively electron-rich substrate produced the desired pen-
One attempt to extend this chemistry to a triple
migration process has been unsuccessful. The reaction
of iodonaphthalene 56 under our standard migration
conditions afforded none of the desired triple migration
product 57, producing instead the reduction product in
62% yield (entry 27). The reason for this failure to afford
compound 57 is indicated in Scheme 6. While we would
anticipate that intermediates D and E should be easily
formed, we believe that the problem lies in getting the
relatively unhindered species E to migrate the palladium
to the more hindered position present in F. Instead the
(12) The spectral properties for this compound were identical with
those previously reported: Hellwinkel, D.; Goeke, K. Synthesis 1995,
1135.
(11) Larock, R. C.; Doty, M. J.; Tian, Q.; Zenner, J. M. J. Org. Chem.
1997, 62, 7536.
8256 J. Org. Chem., Vol. 69, No. 24, 2004

Synthesis of Fused Polycycles
SCHEME 6
Experimental Section
Compounds 19,14 20,14 36,14 and 4914 were prepared accord-
ing to previous literature procedures.
Representative Procedure for the Palladium-Cata-
lyzed Migration Reactions. The appropriate aryl iodide
(0.25 mmol), Pd(OAc)2 (2.8 mg, 0.0125 mmol), 1,1-bis(diphen-
ylphosphino)methane (dppm) (4.8 mg, 0.0125 mmol), and CsO2-
CCMe3 (CsPiv) (0.117 g, 0.5 mmol) in DMF (4 mL) were stirred
under Ar at 100 °C for the specified period of time. The reaction
mixture was allowed to cool to room temperature, diluted with
diethyl ether (35 mL), and washed with brine (30 mL). The
aqueous layer was reextracted with diethyl ether (15 mL). The
organic layers were combined, dried (MgSO4), and filtered, and
the solvent was removed under reduced pressure. The residue
was purified by flash chromatography on silica gel.
1-Phenyldibenzofuran (7). Compound 6 (93.0 mg, 0.25
mmol) or compound 55 (93.0 mg, 0.25 mmol) was allowed to
react under our standard reaction conditions for 1 d. The
reaction mixtures were chromatographed with 30:1 hexane/
EtOAc to afford 54.4 mg (89%) (entry 2, Table 2) or 54.0 mg
(88%) (entry 8, Table 2) of the indicated compound 7, respec-
tively, as a white solid: mp 62-63 °C (lit.13 mp 63-64 °C); 1H
NMR (CDCl3) δ 7.10-7.14 (m, 1H), 7.24-7.26 (m, 1H), 7.37-
7.42 (m, 1H), 7.46-7.64 (m, 9H); 13C NMR (CDCl3) δ 110.5,
111.6, 121.8, 122.3, 122.5, 123.9, 124.0, 127.1, 127.1, 127.9,
128.6, 129.0, 138.0, 140.0, 156.4, 156.5. The other spectral
properties were identical with those previously reported.13
palladium presumably migrates back to the less hindered
position present in intermediate D. This is consistent
with our previous observation in entries 12 and 20 that
palladium is more likely to migrate to or form a new C-C
bond at a less hindered position. Therefore, the palladium
intermediates only equilibrate between D and E, and
eventually produce the reduced 1,3-diphenylnaphthalene.
Acknowledgment. We gratefully acknowledge the
National Science Foundation and the donors of the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society, for partial support of this re-
search and Johnson Matthey Inc. and Kawaken Fine
Chemicals Co. Ltd. for generous donations of palladium
salts.
Conclusions
In conclusion, we have developed novel methodology
for the synthesis of complex fused polycycles employing
two or more sequential Pd-catalyzed intramolecular
processes involving C-H activation. This methodology
exploits relatively facile aryl-to-aryl and vinylic-to-aryl
palladium migrations, followed by intramolecular aryla-
tion to prepare a wide variety of carbocycles and hetero-
cycles. This chemistry works best with electron-rich
aromatics, which is in agreement with the idea that these
palladium-catalyzed C-H activation reactions parallel
electrophilic aromatic substitution. The success of our
double palladium migration for the conversion of biphenyl
55 to dibenzofuran 4 indicates that multiple migration
processes can be employed to produce novel new routes
to a variety of polycycles. Finally, our demonstration that
this chemistry is applicable to alkyne insertion processes
as well opens up still further unique routes to polycylic
products.
Supporting Information Available: The preparation and
characterization of the starting materials 4, 6, 8, 10, 12, 14,
17, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 45, 47, 51, 53, 55,
and 56; the characterization data for compounds 5, 7, 9, 11,
13, 15, 16, 18, 21, 23, 25, 27, 31, 35, 41, 43, 44, 46, 48, and
50; and copies of 1H NMR and 13C NMR spectra for compounds
4-18, 21-28, 30-32, 34-36, 38, 40-48, 50-51, 53, 55, and
56 are available in the Supporting Information. This material
is available free of charge via the Internet at http://pubs.acs.org.
JO048788H
(13) McCall, E. B.; Neale, A. J.; Rawlings, T. J. J. Chem. Soc. 1962,
4900.
(14) (a) Campo, M. A.; Larock, R. C. Org. Lett. 2000, 2, 3675. (b)
Campo, M. A.; Larock, R. C. J. Org. Chem. 2002, 67, 5616.
J. Org. Chem, Vol. 69, No. 24, 2004 8257