Elaboration of Diaryl Ketones into Naphthalenes Fused on Two or Four Sides: A Naphthoannulation Procedure

Article abstract of DOI:10.1021/ja038254i

Transition metal-catalyzed double ring closures of 1,1-diaryl-2,2- diethynylethylenes yield polycyclic aromatic hydrocarbons and heterocycles that contain a newly formed naphthalene ring system embedded in a larger polycyclic network. The diynes required for this procedure are readily synthesized from diaryl ketones by the Corey-Fuchs olefination and subsequent Sonogashira coupling with trimethylsilylacetylene followed by desilylation. This procedure provides easy access to new compounds such as 3,11-di-tert-butyl[4]helicene and 1,8,9-perinaphthothioxanthene. Double naphthoannulation of 9,10-anthraquinone by this procedure closes four new benzene rings in a single operation to give coronene, although the yield in this case is presently low.

Full text of DOI:10.1021/ja038254i

Published on Web 02/24/2004
Elaboration of Diaryl Ketones into Naphthalenes Fused on
Two or Four Sides: A Naphthoannulation Procedure
Patrick M. Donovan and Lawrence T. Scott*
Contribution from the Department of Chemistry, Merkert Chemistry Center, Boston College,
Chestnut Hill, Massachusetts 02467-3860
Received September 1, 2003; E-mail: lawrence.scott@bc.edu
Abstract: Transition metal-catalyzed double ring closures of 1,1-diaryl-2,2-diethynylethylenes yield polycyclic
aromatic hydrocarbons and heterocycles that contain a newly formed naphthalene ring system embedded
in a larger polycyclic network. The diynes required for this procedure are readily synthesized from diaryl
ketones by the Corey-Fuchs olefination and subsequent Sonogashira coupling with trimethylsilylacetylene
followed by desilylation. This procedure provides easy access to new compounds such as 3,11-di-tert-
butyl[4]helicene and 1,8,9-perinaphthothioxanthene. Double naphthoannulation of 9,10-anthraquinone by
this procedure closes four new benzene rings in a single operation to give coronene, although the yield in
this case is presently low.
Introduction
Polycyclic aromatic hydrocarbons (PAHs) and related com-
pounds that contain one or more heteroatoms (primarily O, N,
or S) pervade many branches of chemistry and the allied
sciences.1 They abound, for example, in the soot and smoke
from burning coal and from the incomplete combustion of other
fossil fuels, wood and tobacco;2 environmental chemists are
especially concerned about compounds of this class that exhibit
carcinogenic activity, e.g., benzo[a]pyrene.3 In a completely
different arena, much of the molecular photophysics that is
known today has come from studies on the absorption and
emission properties of polycyclic aromatic π-systems.4 Ques-
tions about the aromaticity of PAHs and related heterocycles
have occupied theoreticians for decades,5 and new questions
keep emerging.6 In these and numerous other branches of chem-
istry, laboratory syntheses of polycyclic aromatic compounds
have been crucial to the advancement of our understanding. No
scientific endeavors have stimulated as much interest in the
development of new methods for the synthesis of such com-
pounds, however, as the recent discovery of fullerenes and the
Figure 1. A new strategy for constructing naphthalenes.
current quest for new carbon-rich materials related to C60, graph-
ite, and nanotubes.7 We report here a procedure for synthesizing
the bicyclic naphthalene ring system in an unconventional man-
ner that annulates both of the new benzene rings onto preexisting
arenes.
In the course of studies on the reactivity of large PAHs, we
needed a method to build fused naphthalenes starting from diaryl
ketones (Figure 1).
To the best of our knowledge, the only previous construction
of a fused naphthalene ring system by a strategy similar to this
is the one reported by Zimmermann et al. in their novel synthesis
of corannulene (Scheme 1).8 In that work, flash vacuum
pyrolysis (FVP) of the bis(trimethylsilylethynyl)olefin 1a on a
20-40 mg scale gave 1.5-3.0 mg of corannulene (ca. 15%
yield). Attempts by Zimmermann et al. to investigate FVP of
the desilylated compound (1b) were thwarted, unfortunately,
by rapid polymerization of the parent hydrocarbon.
(1) Harvey, R. G. Polycyclic Aromatic Hydrocarbons; Wiley-VCH: New York,
1997.
(2) (a) Bennett, D. A.; Britt, P. F.; Buchanan, A. C., III; Ilgner, R. H.; Jenkins,
R. A. Abstracts of Papers, 226th ACS National Meeting, New York,
September 2003; American Chemical Society: Washington, DC, 2003;
ANYL-110. (b) Oanh, N. T. K.; Reutergrdh, L. B.; Dung, N. T. EnViron.
Sci. Technol. 1999, 33, 2703-2709. (c) Mastral, A. M.; Callen, M.; Murillo,
R.; Garcia, T. Fuel 1998, 77, 1513-1516. (d) Freeman, D. J.; Cattell, F.
C. R. EnViron. Sci. Technol 1990, 24, 1581-1585.
(3) (a) Polycyclic Hydrocarbons and Carcinogenesis; Harvey, R. G., Ed.; ACS
Symposium Series No. 283; American Chemical Society: Washington, DC,
1985. (b) DeMarini, D. M.; Landi, S.; Tian, D.; Hanley, N. M.; Li, X.; Hu,
F.; Roop, B. C.; Mass, M. J.; Keohavong, P.; Gao, W.; Olivier, M.; Hainaut,
P.; Mumford, J. L. Cancer Res. 2001, 61, 6679-6681.
(4) (a) Malkin, J. Photophysical and Photochemical Properties of Aromatic
Compounds; CRC: Boca Raton, FL, 1993. (b) Birks, J. B. Photophysics
of Aromatic Molecules; Wiley-Interscience: New York, 1970.
(6) (a) Moran, D.; Stahl, F.; Bettinger, H. F.; Schaefer, H. F., III; Schleyer, P.
v. R. J. Am. Chem. Soc. 2003, 125, 6746-6752. (b) Schleyer, P. v. R.;
Nyulaszi, L.; Karpati, T. Eur. J. Org. Chem. 2003, 1923-1930. (c)
Havenith, R. W. A.; Jiao, H.; Jenneskens, L. W.; Van Lenthe, J. H.; Sarobe,
M.; Schleyer, P. v. R.; Kataoka, M.; Necula, A.; Scott, L. T. J. Am. Chem.
Soc. 2002, 124, 2363-2370. (d) Schulman, J. M.; Disch, R. L. J. Am. Chem.
Soc. 1996, 118, 8470-8474. (e) Zhou, Z. J. Phys. Org. Chem. 1995, 8,
103-107.
(7) (a) Taylor, R. Lecture Notes on Fullerene Chemistry: A Handbook for
Chemists; Imperial College Press: London, 1999. (b) Simpson, C. D.;
Brand, J. D.; Berresheim, A. J.; Przybilla, L.; Rader, H. J.; Mu¨llen, K.
Chemistry 2002, 8, 1424-1429. (c) Haddon, R. C., Ed. Acc. Chem. Res.
2002, 35(12): special issue on Carbon Nanotubes.
(8) (a) Zimmermann, G.; Nuechter, U.; Hagen, S.; Nuechter, M. Tetrahedron
Lett. 1994, 35, 4747-4750. (b) For evidence that thermal cyclizations of
such 1,1-diaryl-2,2-dialkynylethylenes involve the intermediacy of “isoben-
zenes,” see: Hopf, H.; Berger, H.; Zimmermann, G.; Nuchter, U.; Jones,
P. G.; Dix, I. Angew. Chem. 1997, 109, 1236-1239; Angew. Chem., Int.
Ed. Engl. 1997, 36, 1187-1190.
(5) (a) Schleyer, P. v. R., Ed.; Chem. ReV. 2001, 101(5): special issue on
Aromaticity. (b) Minkin, V. I.; Glukhovtsev, M. N.; Simkin, B. Y.
Aromaticity and Antiaromaticity; Wiley: New York, 1994. (c) Garratt, P.
J. Aromaticity; John Wiley Sons: New York, 1986. (d) Lewis, D.; Peters,
D. Facts and Theories of Aromaticity; Crane Russak: New York, 1975.
9
3108
J. AM. CHEM. SOC. 2004, 126, 3108-3112
10.1021/ja038254i CCC: $27.50 © 2004 American Chemical Society

Naphthoannulation of Diaryl ketones
A R T I C L E S
Scheme 1
new ring system can be found in the Supporting Information;
the molecule is planar, as expected, with long C-S bonds, as
shown in Scheme 2.
The ruthenium-catalyzed cycloisomerization reaction of
1-aryl-1-buten-3-ynes has previously been shown by deuterium-
labeling studies to involve initial formation of a ruthenium-
vinylidene.11 Merlic has also provided evidence that the
subsequent carbon-carbon bond-forming step exhibits the
electronic characteristics of an electrophilic aromatic substitu-
tion;11 however, the final steps of the mechanism remain the
subject of speculation (Scheme 3).
Our synthesis puts together analogous 1,1-diethynylethylenes
in a different way and then cyclizes them by solution methods,
rather than by gas-phase FVP.
Scheme 3
Results and Discussion
Our synthesis of the previously unknown 1,8,9-perinaph-
thothioxanthene (2, Scheme 2) illustrates the four steps we use
Scheme 2
Three aspects of the particular reaction sequence shown in
Scheme 2 deserve further comment. First, we found the Corey-
Fuchs olefination of thioxanthenone (3) to be rather sluggish,
presumably owing to the special stability of the “aromatic”
heteroannulenone ring system.12 Fortunately, forcing conditions
(150 °C/sealed vessel, 48-72 h) produced the 1,1-dibromoalk-
ene (4) without complications. Second, alkene 4 was observed
to capture adventitious acid (e.g., the traces of acid in CDCl3)
to produce what we believe to be the thiopyrylium ion 7
(eq 1).13
to accomplish the overall naphthoannulation outlined in Figure
1. It begins with a Corey-Fuchs olefination9 of the starting
ketone, thioxanthenone (3). The resulting 1,1-dibromoalkene (4)
is then transformed into the corresponding 1,1-bis(trimethylsi-
lylethynyl)alkene (5) by a double Sonogashira coupling10 with
trimethylsilylacetylene; subsequent desilylation with tetra-n-
butylammonium fluoride provides the parent diyne (6). To make
two new benzene rings out of the two 1-aryl-1-buten-3-yne
moieties in 6 we employed the ruthenium-catalyzed cycloi-
somerization protocol developed by Merlic.11 Thus, slow
addition of 6 to a solution of (Ph3P)Ru(cymene)Cl2 (0.2 equiv)
in refluxing 1,2-dichloroethane gives 1,8,9-perinaphthothiox-
anthene (2) in 95% yield. An X-ray crystal structure of this
A set of small, low field signals arising from the thiopyrylium
ion 714 can be seen in most NMR spectra of 4, but neutralization
regenerates the alkene with no deleterious consequences.15
Finally, we found the unprotected parent diyne (6) to be rather
(12) (a) Bouanani, H.; Gayoso, J. Bull. Soc. Chim. Fr. 1974, 545-552. (b) ref
5d, pp 79-81 and references therein.
(13) For references to pyrylium and thiopyrylium ions, see: (a) Balaban, A. T.;
Dinculescu, A.; Dorofeenko, G. N.; Fischer, G. V.; Kablik, A. V.;
Mezheritskii, V. V.; Schroth, W. AdVances in Heterocyclic Chemistry;
Supplement 2: Pyrylium Salts: Synthesis, Reactions, and Physical Proper-
ties; Academic Press: New York, 1982. (b) Reference 5c, Chapter 8.
(14) Small, but unobscured, 1H NMR signals for thiopyrylium ion 7 (400 MHz,
CDCl3): 7.98 (dm, J ) 9.6 Hz), 7.41-7.51 (m), 6.68 (s, CHBr2).
(15) Further evidence for the thiopyrylium ion 7 follows from the similarity of
its behavior to that of the oxygen analogue (ref 17).
(9) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769-3772.
(10) (a) Sonogashira, K. In Handbook of Organopalladium Chemistry for
Organic Synthesis; Negishi, E., Ed.; Wiley: Hoboken, NJ, 2002; Vol. 1,
pp 493-529. (b) Sonogashira, K. J. Organomet. Chem. 2002, 653, 46-
49. (c) Sonogashira, K. In Metal-Catalyzed Cross-Coupling Reactions;
Diederich, F., Stang, P. J., Eds.; Wiley-VCH Verlag GmbH: Weinheim,
Germany, 1998, pp 203-229.
(11) Merlic, C. A.; Pauly, M. E. J. Am. Chem. Soc. 1996, 118, 11319-11320.
9
J. AM. CHEM. SOC. VOL. 126, NO. 10, 2004 3109

A R T I C L E S
Donovan and Scott
Scheme 4
in high yield, but the absence of isomeric side products in the
latter makes it the more attractive choice. Better conditions for
the cyclization of 13 must be found, obviously, to make this
synthesis of 14 practical; however, we did succeed in preparing
enough pure 14 to obtain an X-ray crystal structure (planar;
see Supporting Information) and to record the first 1H and 13
C
NMR spectra of this previously uncharacterized compound.19
Factors other than electronic deactivation by the meta oxygen
atom that may also contribute to the low yield for cycloisomer-
ization of 13 are discussed in the following section.
Finally, in an effort to expand the power of this naphthoan-
nulation strategy, we applied the four-step procedure to an-
thraquinone (15). Gratifyingly, the conversions to 16, 17, and
18 all worked satisfactorily, and the final closure with (Ph3P)-
Ru(cymene)Cl2 (0.2 equiv) does indeed produce coronene (19).21
The yield in the final 4-fold cycloisomerization, unfortunately,
is only in the 15-20% range at the present stage of optimization.
labile and prone to easy polymerization (cf 1b8a). Diyne 6 can
be prepared in essentially quantitative yield by desilylation of
5 and survives long enough at room temperature for routine
spectroscopic characterization, but we typically use it im-
mediately for the next step. The 95% yield reported above for
the synthesis of 1,8,9-perinaphthothioxanthene (2) actually
corresponds to the overall yield of 2 for the two steps from 5.
When the same four-step reaction sequence is carried out on
acyclic diaryl ketones, e.g., benzophenone (8a) and 4,4′-di-tert-
butylbenzophenone (8b), [4]helicenes (9) are produced (Scheme
4). Despite the steric strain of the final products, the yields for
the cyclization steps in these syntheses are still quite decent
(60% in both cases, over two steps from the corresponding
silylated diynes).
Xanthenone (10), like thioxanthenone (3), is sluggish in the
Corey-Fuchs olefination, and the resulting dibromoalkene (11)
is even more susceptible to protonation by traces of acid than
is its sulfur counterpart (4).12,13,16,17 Nevertheless, the corre-
sponding 1,1-bis(trimethylsilylethynyl)alkene (12) can be readily
obtained as before and cleanly desilylated to the parent diyne
(13). At this point, however, we encountered a serious short-
coming of the ruthenium-catalyzed cycloisomerization reaction.
In keeping with Merlic’s view that the carbon-carbon bond-
forming step involves electrophilic attack on the aromatic ring11
and the fact that ether oxygen atoms strongly deactivate positions
meta to their point of attachment,18 exposure of 13 to solutions
of (Ph3P)Ru(cymene)Cl2 under a variety of conditions promoted
polymerization of the diyne and gave no more than 10% yield
of cyclized product (14).19 The reaction was further complicated
by the formation of a second, inseparable, product that has the
same molecular weight as 14 (GC-MS).
Limitations of the Transition Metal-Catalyzed Cycloi-
somerizations. The low yields associated with cyclizations meta
to an ether oxygen (13 f 14) and with the ambitious double
naphthoannulation of anthraquinone (18 f 19) have already
been mentioned above. One additional system we investigated
that fails completely in the final cycloisomerization step is
fluorenone (20). As usual, the conversions to the requisite
intermediates (21, 22, and 23) all worked satisfactorily, but none
of the catalysts examined in this case promoted double cycliza-
tion to the known hydrocarbon 24.22 We do not see even the
product of one cyclization; only polymers are obtained.
To contend with this problem, we explored the use of other
catalytic systems introduced recently by Fu¨rstner for the
cyclization of o-ethynylbiphenyls to phenanthrenes (eq 2).20
The factors that determine the success or failure of these
cycloisomerizations are all kinetic in origin and not thermody-
namic. Even for the cycloisomerizations that give quite strained
products (e.g., 23 f 24), the final ring closure is calculated to
(16) Small, but unobscured, 1H NMR signals for pyrylium ion 11 (400 MHz,
CDCl3): 8.46 (dd, J ) ca 8.8, 2.0 Hz), 7.48 (dd, J ) ca 8.8, 2.0 Hz), 6.63
(s, CHBr2).
(17) Addition of CH3OH to the NMR tube causes the signals for the small
amount of pyrylium ion derived from 11 to disappear, and new signals
corresponding to those of a methyl ether grow in.
(18) Taylor, R. Electrophilic Aromatic Substitution; Wiley: New York, 1990;
pp 47, 88, and later chapters.
(19) Naphtho[2,1,8,7-klmn]xanthene (14) was first prepared in 1941 from coal
tar pitch by Kruber, O. Chem. Ber. 1941, 74, 4B, 1688-1692, but no
spectroscopic characterization has subsequently been reported.
(20) Fu¨rstner, A.; Mamane, V. J. Org. Chem. 2002, 67, 6264-6267.
Thus, it was found that catalytic PtCl2 in hot toluene will also
convert the 1,1-diethynylalkene 13 into 1,8,9-perinaphthoxan-
thene (14), albeit only in about 15% yield. Neither the
Ru-catalyzed reaction nor the Pt-catalyzed alternative proceeds
9
3110 J. AM. CHEM. SOC. VOL. 126, NO. 10, 2004

Naphthoannulation of Diaryl ketones
A R T I C L E S
vacuum pyrolysis in this case undoubtedly stems from the
increased magnitude of out-of-plane distortions at elevated
temperatures.26
Figure 2. Distances for (a) first closure (b) second closure.
be exothermic by more than 40 kcal/mol, and the first cyclization
is calculated to be exothermic by more than 60 kcal/mol.23
Apparently, bimolecular processes that lead to higher-molecular
weight polymers dominate the reactions whenever some struc-
tural feature slows down the desired cyclizations.
Conclusions
Without a good picture for the transition state of the rate-
limiting step in these transition metal-catalyzed cycloisomer-
izations it is difficult to explain why some naphthoannulations
work so well (e.g., 6 f 2) while others fail. We have noticed,
however, that there is an inverse correlation between the success
of the naphthoannulation and the calculated distance23 that
separates the terminal acetylenic carbon atom in the starting
material from the aryl carbon atom to which it will become
bonded (distances a and b in Figure 2).24,25
A four-step procedure is reported for the transformation of
diaryl ketones into polycyclic aromatic hydrocarbons and
heterocycles that contain a newly formed, fused naphthalene
ring system. The transition metal-catalyzed double ring closures
of 1,1-diaryl-2,2-diethynylethylenes in the final step appear to
work best for those compounds in which the distances between
the atoms to be joined are less than about 3.35-3.40 Å.
Experimental Section
In the most successful cycloisomerization reported here (6
f 2), those distances are 3.34 and 3.33 Å, respectively, whereas
in the barely viable oxygen analogue (13 f 14), those distances
are 3.35 and 3.42 Å, respectively. The demands in the first step
are comparable in these two cases, but the second step in the
oxygen heterocycle may be too slow to compete effectively with
polymerization. In the double naphthoannulation of anthraquino-
ne (18 f 19), the distances for the four cyclization steps,
following the energetically most favorable pathway, are 3.22,
3.32, 3.39, and 3.39 Å, respectively. The longest distances in
this case fall between those in the sulfur heterocycle (6) and
those in the oxygen heterocycle (13), and the overall yield
likewise falls between those two extremes. In the naphthoan-
nulation of fluorenone (23 f 24), the distances for the two
cyclization steps are much longer (3.46 and 3.80 Å, respec-
tively), and the reaction fails completely.
From these admittedly sparse data, we tentatively conclude
that cycloisomerizations of the sort reported here may be limited
to those in which the distances between the atoms to be joined
are less than about 3.40 Å. In the benzophenone systems, which
successfully lead to [4]helicenes (Scheme 4), relatively free
rotation of the phenyl groups brings the atoms to be joined well
within 3.40 Å of one another. In light of this empirical “distance
rule,” it is no surprise that diyne 25 fails to cyclize to
corannulene under the transition metal-catalyzed cycloisomer-
ization conditions reported here (eq 3). The success of flash
General procedures for the four steps of the naphthoannulation
sequence are reported below, using the transformations illustrated in
Scheme 2 as examples. Full details for all reactions can be found in
the Supporting Information.
General Procedure 1: Corey-Fuchs Olefination. An oven-dried
sealable 150-mL pressure vessel was charged with thioxanthen-9-one
(3, 1.875 g, 8.833 mmol), carbon tetrabromide (5.885 g, 17.74 mmol),
and a magnetic stirring bar. The vessel was purged with nitrogen gas,
50 mL of anhydrous benzene was added, and the mixture was stirred
for 5 min. Triphenyl phosphine (9.313 g, 35.51 mmol) was then added,
and the vessel was sealed. The reaction mixture was heated at 150 °C
for 44 h in a wax bath with vigorous stirring. The vessel was cooled
to room temperature before opening, and the contents were rinsed into
a round-bottom flask with dichloromethane. The crude product was
then preadsorbed onto neutral alumina and chromatographed on an
alumina column with hexanes as the eluent. The hexane fractions were
dried over magnesium sulfate and concentrated to dryness on a rotary
evaporator to give 1.739 g (53% yield) of 9-(dibromomethylene)-9H-
1
thioxanthene (4) as an off-white powder that was pure by H NMR
analysis: mp 117-119 °C. 1H NMR (400 MHz, CDCl3) δ (ppm) 7.78
(dm, J ) 7.2 Hz, 2H), 7.51 (dm, J ) 7.6 Hz, 2H), 7.30 (td, J ) 7.4 Hz
and J ) 1.8 Hz, 2H), 7.26 (td, J ) 7.6 Hz and J ) 1.6 Hz, 2H). 13C
NMR (125 MHz, CDCl3) δ 140.49, 135.49, 134.11, 129.37, 127.91,
127.22, 126.08, 92.409. Anal. Calcd. for C14H8Br2S: C, 45.68; H, 2.19.
Found: C, 45.89; H, 1.89.
General Procedure 2: Sonogashira Coupling. Dibromide 4
(1.1025 g, 2.995 mmol), PdCl2(PPh3)2 (226 mg, 0.321 mmol), and
copper(I) iodide (145 mg, 0.764 mmol) were added to a sealable vessel
that contained 30 mL of triethylamine. The solution was stirred and
purged with nitrogen for 5 min. Trimethylsilyl acetylene (1.95 mL,
13.8 mmol) was then added by syringe, and the vessel was sealed.
The reaction mixture was heated at 95 °C in a wax bath with stirring
for 21 h. The vessel was then cooled to room temperature and opened,
and the contents were diluted with 100 mL of dichloromethane. The
resulting solution was washed successively with two 100-mL portions
of saturated aqueous ammonium chloride solution and two 100-mL
portions of distilled water. The organic layer was then dried over
magnesium sulfate, and the crude product was preadsorbed onto silica
gel. The sample was chromatographed on a silica gel column with a
(21) Coronene (19): Chem. Abstr. Registry No. 191-07-1. Available from
Aldrich Chemical Co. (catalog no. C8,480-1).
(22) Benzo[ghi]fluoranthene (24): Chem. Abstr. Registry No. 203-12-3. Avail-
able from Fluka Chemical Co. (catalog no. BCR-139).
(23) Calculations were performed at the B3LYP/6-31G*//AM1 level of theory,
using the Spartan package of programs (Linux v. 02: Wavefunction, Irvine,
CA).
(24) A more accurate estimation of relative transition-state geometries would
be given by comparisons among the proposed vinylidene intermediates
(Scheme 3); however, geometry calculations on transition metal-vinylidene
complexes are less reliable, and the trend should be qualitatively no
different.
(25) Distance-reactivity relationships of this sort have been noted for other types
of cyclization reactions; see, for example: (a) Nicolaou, K. C.; Dai, W.
M. Angew. Chem., Int. Ed. Engl. 1991, 30, 1387-1416. (b) Nicolaou, K.
C.; Smith, A. L. Acc. Chem. Res. 1992, 25, 497-503. (c) Chen, W.-C.;
Zou, J.-W.; Yu, C.-H. J. Org. Chem. 2003, 68, 3663-3672. (d) Gaffney,
S. M.; Capitani, J. F.; Castaldo, L.; Mitra, A. Int. J. Quantum Chem. 2003,
95, 706-712.
(26) (a) Scott, L. T.; Hashemi, M. M.; Meyer, D. T.; Warren, H. B. J. Am.
Chem. Soc. 1991, 113, 7082-7084. (b) Scott, L. T.; Cheng, P.-C.; Hashemi,
M. M.; Bratcher, M. S.; Meyer, D. T.; Warren, H. B. J. Am. Chem. Soc.
1997, 119, 10963-10968.
9
J. AM. CHEM. SOC. VOL. 126, NO. 10, 2004 3111

A R T I C L E S
Donovan and Scott
2% ethyl acetate:98% hexane solution until a brown-yellow band eluted.
The fractions containing the product were then combined and concen-
trated to dryness on a rotary evaporator to give 895 mg (74% yield) of
9-[bis(trimethylsilylethynyl)methylene)-9H-thioxanthene (5) as a brown
solid which was pure by 1H NMR analysis: mp 199-201 °C. 1H NMR
(400 MHz, CDCl3) δ (ppm) 8.11 (dm, J ) 8.0 Hz, 2H), 7.46 (dm, J )
6.8 Hz, 2H), 7.27 (td, J ) 7.2 Hz and J ) 1.6 Hz, 2H), 7.23 (td, J )
7.6 Hz and J ) 1.6 Hz, 2H), 0.17 (s, 18H). 13C NMR (125 MHz, CDCl3)
δ 148.61, 133.59, 133.47, 129.68, 128.28, 126.24, 125.37, 103.18,
102.50, 99.18, -0.21. HRMS ESI (m/z): [M + Na]+ calcd for C24H26-
NaSi2S, 425.1191: found 425.1196.
heated to 75 °C under nitrogen in a three-neck flask equipped with a
reflux condenser and a septum. The solution of 6 was then added at a
rate of 0.5 mL/h with a syringe pump. Heating was continued for an
additional 13 h after completion of the addition, for a total reaction
time of 25 h. The reaction mixture was then cooled to room temperature,
diluted with dichloromethane, and filtered through a short pad of silica
gel, which was washed with several additional portions of dichlo-
romethane. The crude product was then adsorbed onto silica gel and
subjected to column chromatography on silica gel with hexanes as the
eluent. The first yellow band was collected. The fractions were
combined, dried over magnesium sulfate, and concentrated to dryness
on a rotary evaporator to give 190 mg of naphtho[2,1,8,7-klmn]-
thioxanthene (2) as yellow needles (95% yield over 2 steps), which
were pure by 1H NMR analysis: mp 149-151 °C. 1H NMR (400 MHz,
CDCl3) δ (ppm) 7.64 (s, 4H), 7.44 (dd, J ) 8.0 Hz and J ) 1.2 Hz,
2H), 7.28 (t, J ) 7.2 Hz, 2H), 7.16 (dd, J ) 7.2 Hz and J ) 1.2 Hz,
2H). 13C NMR (125 MHz, CDCl3) δ 134.12, 132.02, 128.38, 128.06,
127.61, 127.26, 126.85, 126.36, 124.87, 120.96. HRMS ESI (m/z):
[M]+ calcd for C18H10S, 258.0503: found 258.0498.
General Procedure 3: Desilylation. The bis(trimethysilyl)diyne 5
(314 mg, 0.780 mmol) was dissolved in 5 mL of dichloromethane. A
separate solution of tetrabutylammonium fluoride trihydrate (1.012 g,
3.208 mmol) in 5 mL of dichloromethane was also prepared. The two
solutions were combined, and the resulting mixture was stirred at
ambient temperature. After TLC showed no remaining starting material
(ca 15 min), the solution was filtered through a short pad of silica gel,
which was subsequently washed with additional dichloromethane. The
combined dichloromethane solutions were dried over magnesium sulfate
and then concentrated nearly to dryness on a rotary evaporator, taking
care not to apply excessive heat. The unprotected 9-(diethynylmeth-
ylene)-9H-thioxanthene (6) is sensitive to heat and polymerizes readily,
and thus, no attempt was made to obtain an accurate mass. According
X-ray Structures. Four X-ray crystal structures on compounds
reported here have been deposited at the Cambridge Crystallographic
Data Centre: 2 (CCDC 226047), 9b (CCDC 226048), 14 (CCDC
226049), 18 (CCDC 226050).
1
to H NMR analysis, 6 was formed with complete conversion as the
Acknowledgment. We thank Jarrod Blank for determining
X-ray crystal structures for compounds 2, 9b, 14, and 18.
Financial support from the National Science Foundation is
gratefully acknowledged.
only product and in essentially quantitative yield. After the NMR
analysis, this product was immediately subjected to the next reaction
without further purification. 1H NMR (400 MHz, CDCl3) δ (ppm) 8.08
(m, 2H), 7.50 (m, 2H), 7.30 (m, 4H), 3.12 (s, 2H).
General Procedure 4: Double Cyclization. The crude product from
the previous reaction (6) was diluted to a volume of 6 mL with 1,2-
dichloroethane, and the resulting solution was drawn into a 10 mL
syringe equipped with a long needle. A separate solution of RuCl2-
(PPh3)(η6-p-cymene)27 (141 mg, 0.248 mmol) and ammonium hexafluo-
rophosphate (84 mg, 0.51 mmol) in 16 mL of 1,2-dichloroethane was
Supporting Information Available: Experimental details for
all the reactions reported herein, analytical and spectral char-
acterization data for all new compounds (PDF), and crystal-
lographic information for compounds 2, 9b, 14, and 18 (CIF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
(27) Bennett, M. A. S., Anthony K. J. Chem. Soc., Dalton Trans. 1974, 2, 233-
241.
JA038254I
9
3112 J. AM. CHEM. SOC. VOL. 126, NO. 10, 2004