Chromene chromium carbene complexes in the syntheses of naphthopyran and naphthopyrandione units present in photochromic materials and biologically active natural products

Article abstract of DOI:10.1021/ja0568852

The carbene complex 5-(2,2-dimethyl-2H-chromene)methoxylmethylene chromium pentacarbonyl will undergo a benzannulation reaction with phenylacetylene, 1-pentyne, 3-hexyne, and trimethylsilylacetylene to give 7-hydroxy-10-methoxy- 3H-naphtho[2.1-b]pyrans as

Full text of DOI:10.1021/ja0568852

Published on Web 08/05/2006
Chromene Chromium Carbene Complexes in the Syntheses of
Naphthopyran and Naphthopyrandione Units Present in
Photochromic Materials and Biologically Active Natural
Products
Manish Rawat, Victor Prutyanov, and William D. Wulff*
Contribution from the Department of Chemistry, Michigan State UniVersity,
East Lansing, Michigan 48824
Received October 8, 2005; E-mail: wulff@chemistry.msu.edu
Abstract: The carbene complex 5-(2,2-dimethyl-2H-chromene)methoxylmethylene chromium pentacarbonyl
will undergo a benzannulation reaction with phenylacetylene, 1-pentyne, 3-hexyne, and trimethylsilylacety-
lene to give 7-hydroxy-10-methoxy-3H-naphtho[2.1-b]pyrans as the primary product. These compounds
are difficult to obtain pure due to their sensitivity to air. If the benzannulation reaction is performed in
conjunction with protection of the phenol function at C-7, then good to excellent yields of 7-alkoxy-10-
methoxy-3H-naphtho[2.1-b]pyrans are afforded. If the 7-hydroxy products are captured by triflic anhydride,
then the resulting aryl triflate can be used to access 3H-naphtho[2.1-b]pyrans bearing C-7 carbon
substituents. The 7-hydroxy products can be oxidized to 3H-naphtho[2,1-b]pyran-7,10-diones which are
stable. The chromenyl carbene complex reacts with 1,6-bis(triisopropylsilyl)-1,3,5-hexatriyne to give a 2,3-
dihydro-2,2-dimethylbenzo[de]chromene, a product type that has not been seen before in the reaction of
Fischer carbene complexes with alkynes. A mechanism is proposed for this process that involves R,â-
hydride elimination from a chromacyclobutane intermediate. Chromenyl tungsten complexes react with
alkynes to give products that result from cyclization without CO insertion.
I. Introduction
is desired, e.g., for the manufacture of ophthalmic lenses, contact
lenses, solar protection glasses, filters, camera optics, transmis-
sion devices, agrochem films, glazing, decorative objects, or
information storage by optical inscription (coding).2 The proper-
ties of molecules of the type 3 have been actively studied with
such applications in mind. Several SAR studies of these
molecules have resulted in the finding that the nature and the
position of the substituents in the 3H-naphtho[2,1-b]pyrans have
a significant impact on the absorption properties of these
compounds.2 Although a number of patents and publications
on various naphthopyran derivatives have appeared in the past
two decades, detailed studies with alkyl, aromatic, and hetero-
atom substituents at all positions have not been done. This is
likely due, at least in part, to the limited methods to generate
these compounds.
The 3H-naphtho[2,1-b]pyran and related 3H-naphtho[2,1-b]-
pyran-7,10-dione ring systems are important core units in a
number of natural products1 and also in photochromic com-
pounds.2 A number of natural products contain the 3H-naphtho-
[2,1-b]pyran-7,10-dione core. Cannon and co-workers isolated
nine quinones related to this family from the roots of cono-
spermum teretifolium.1b Kimpe has recently isolated compound
5 as a natural product with the naphthopyran core, although it
is oxygenated at the 7- and 10-positions (Scheme 1).1c The
natural product 5 was isolated from the roots of Pentas bussei,
a plant found in Kenya. The decoction of the roots is used as a
remedy for gonorrhea, syphilis, and dysentery. Conocurvone is
a unique natural product in that it contains three naphthopyr-
andione units, and this fact may be related to its remarkable
anti-HIV activity.1a 3H-Naphtho[2,1-b]pyrans are known to
exhibit photochromic properties which occur with the photo-
induced electrocyclic ring-opening to the ortho-quinone methide
4.2 Photochromic compounds have found wide applications in
which a sunlight-induced reversible color change or darkening
II. Background
Most of the studies of the effects of substitution of the
naphthopyran nucleus on the photochromic behavior has been
done at the positions 1 through 6.2 We envisioned that a family
of naphthopyrans with substitution in positions 7-10 could be
accessed through the reaction of a 5-chromenyl chromium
carbene complex of the type 7 with alkynes (Scheme 2).3 This
reaction would be expected to generate the naphthopyran 8 with
an alkoxy group in the 10-position and a hydroxy group in
position 7. Positions 8 and 9 could be introduced by proper
choice of the alkyne. For terminal alkynes, the regiochemical
incorporation of the alkyne in reactions with carbene complexes
(1) (a) Decosterd, L. A.; Parsons, I. C.; Gustafson, K. R.; Cardellina, J. H., II;
McMahon, J. B.; Cragg, G. M.; Murata, Y.; Pannell, L. K.; Steiner, J. R.;
Clardy, J.; Boyd, M. R. J. Am. Chem. Soc. 1993, 115, 6673-6679. (b)
Cannon, J. R.; Joshi, K. R.; McDonald, I. A.; Retallack, R. W.; Sierakowki,
A. F.; Wong, L. C. H. Tetrahedron Lett. 1975, 16, 2795-2798. (c) Bukuru,
J. F.; Van, T. N.; Puyvelde, L. V.; Mathenge, S. G.; Mudida, F. P.; Kimpe,
N. D. J. Nat. Prod. 2002, 65, 783-785.
(2) Gemert B. V. Benzo and Naphthopyrans (Chromenes). In Organic
Photochromic and Thermochromic Compounds; Crano, J. C., Guglielmetti,
R., Eds.; Plenum Press: New York, 1999; Vol. 1, Chapter 3.
9
11044
J. AM. CHEM. SOC. 2006, 128, 11044-11053
10.1021/ja0568852 CCC: $33.50 © 2006 American Chemical Society

Chromene Chromium Carbene Complexes
A R T I C L E S
Scheme 1
Scheme 3
rated next to the phenol function.5 In some instances, electronics
can be used to control the regiochemical incorporation of the
alkyne.6 The benzannulation product 8 could serve as a direct
source of the naphthopyrandione core 9 and, furthermore, would
provide the additional flexibility for the preparation of a variety
of naphthopyran derivatives substituted at position 7 via various
coupling reactions of the triflate 10.
Scheme 2
While the reaction of chromenyl complex 7 with simple
alkynes should be able to provide access to naphthopyrandiones
of the type 9, the reaction of complex 7 with conjugated triynes
of the type 13 has the potential of providing rapid access to
tris-naphthopyrandiones of the type 12 and, if this is the case,
then a rather straightforward route for the total synthesis of
Conocurvone 6 (Scheme 3). We have recently reported the first
examples of the reaction of carbene complexes with conjugated
triynes and were surprised to find that while the reaction of
one equivalent of the cyclohexenyl complex 18 with triyne 13
(R ) Si(i-Pr)3) gave the expected benzannulation product 17,
the same reaction of the phenyl complex 15 gave the furan
product 16 where CO insertion had occurred but not cyclization
to the phenyl ring.7 Both reactions did occur by the expected
selective reaction with the less sterically hindered central alkyne
unit of the triyne 13 (R ) Si(i-Pr)3), although other triynes did
not show this regioselectivity. Based on these limited results, it
might be expected that the chromenyl complex 7 might behave
more like the phenyl complex 15 than the cyclohexenyl complex
18 and thus not be able to provide for a direct route to tris-
quinones of the type 12. Nonetheless, our limited experience
with the reactions of carbene complexes with triynes suggested
that surprise is the norm and not the exception. Indeed, this
proved to be the case with the chromenyl complex 7. The results
is known to be very high with a single isomer produced where
the alkyne substituent would be adjacent to the phenol function
(C-8).4 For internal alkynes, the regiochemical control is usually
determined by the relative size of the substituents. For example,
1-phenyl-1-propyne has been observed to give a 40:1 mixture
of regioisomers with the phenyl group preferentially incorpo-
(3) For recent reviews on carbene complexes in organic chemistry, see: (a)
Wulff, W. D. In ComprehensiVe Organometallic Chemistry II; Abel, E.
W., Stone, R. G. A., Wilkinson, G., Eds.; Pergemon Press: 1995; Vol. 12,
pp 469-547. (b) Hegedus, L. S. Tetrahedron 1997, 53, 4105-4128. (c)
de Meijer, A.; Schirmer, H.; Duetsch, M. Angew. Chem., Int. Ed. 2000,
39, 3964-4002. (d) Do¨tz, K. H.; Tomuschatt, P. Chem. Soc. ReV. 1999,
28 187-198. (e) Herndon, J. W. Coord. Chem. ReV. 1999, 181, 177-242.
(f) Do¨rwald, F. Z. Metal Carbenes in Organic Synthesis; Wiley-VCH:
Weinheim, New York, 1999.
(4) (a) Wulff, W. D.; Tang, P. C.; McCallum, J. J. Am. Chem. Soc. 1981, 103,
7677-7678. (b) Do¨tz, K. H.; Mu¨hlemeier, J.; Schubert, U.; Orama, O. J.
Organomet. Chem. 1983, 247, 187-201. (c) Yamashita, A.; Toy, A.
Tetrahedron Lett. 1986, 27, 3471-3473.
(5) Waters, M. L.; Bos, M. E.; Wulff, W. D. J. Am. Chem. Soc. 1999, 121,
6403-6413.
(6) (a) Chamberlin, S.; Wulff, W. D.; Waters, M. L. J. Am. Chem. Soc. 1994,
116, 3113-3114. (b) Davies, M. W.; Johnson, C. N.; Harrity, J. P. A. J.
Chem. Soc., Chem. Commun. 1999, 2107-2108.
(7) Jiang, M. X.-W.; Rawat, M.; Wulff, W. D. J. Am. Chem. Soc. 2004, 126,
5970-5971.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11045

A R T I C L E S
Scheme 4
Rawat et al.
Table 1. Directed Metalation/Bromination of Aminal 28a
t-BuLi
ratio
29:30b
yield (%)
29
30c
entry
(equiv)
T1
(
°C)
T2
(°
C)
+
1
2
3
4
5
6
7
8
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
rt
rt
rt
rt
14
0
rt
rt
rt
rt
2:1
3.3:1
3.3:1
35
48
41
e10d
43
14 to rt
0 to rt
-20 to rt
-70 to rt
3.3:1
5.6:1
5.6:1
50:1
52
72
69
-20
-40
a All reactions were at 0.1 M in 28 and used 3.0 equiv of bromination
agent. b Determined in isolated mixture. c Isolated yield of the mixture.
d e10% yield was observed with either NBS or bromine as bromination
agent.
of studies described herein show that while the reaction of
complex 7 with simple alkynes does in fact provide for a facile
entry to naphthopyrans and naphthopyrandiones, the reaction
of the chromenyl complex 7 with conjugated triynes led to an
unprecedented reaction involving alkyne insertion and a sub-
sequent addition/rearrangement process involving the double
bond of the pyran ring.
Scheme 5
III. Results and Discussion
Synthesis of the Chromium Cromenyl Carbene Complex
20. Fischer carbene complexes with aryl substituents are
typically made from the corresponding aryl halides.3 In the case
of carbene complex 20, this will require the aryl halide 21
(Scheme 4). A three-step synthesis of 21 has been reported that
begins with the regioselective alkylation of 3-bromophenol with
prenyl bromide,8 but in our hands, this reaction under the
reported conditions gave an inseparable mixture of products
which included mono- and dialkylated species. We thus decided
to evaluate new approaches for the synthesis of the chromenyl
bromide 21 (Scheme 4). One involves the construction of the
pyran ring by an acid-catalyzed cyclization of the allylic alcohol
22 which in turn should be accessible from the aldehyde 23.
The second approach begins with the base-catalyzed cyclization
of 1,3-cyclohexanedione 27 with 3-methyl-2-butenal.
The aldehyde 23 that is needed for the synthesis of carbene
complex 20 as outlined in Scheme 4 has previously been
prepared by Couture and co-workers.9 Their synthesis involved
the directed-metalation of imidazolidine 28 and then bromination
with dibromotetrachloroethane followed by acid workup to
provide 2-bromoanisaldehyde 29 in 79% yield (Table 1). When
this reaction was repeated under the reported conditions, a 2:1
mixture of 2-bromoanisaldehyde and 2-chloroanisaldehyde was
obtained in 35% yield (entry 1). It was found that the ratio of
29:30 showed a noticeable dependence on the temperature of
the reaction (Table 1). Optimal conditions required ortho-
lithiation at -40 °C and the addition of the brominating agent
at -78 °C which gave a 69% yield of 29 (entry 8) and only a
trace amount of 30 (29:30 ) 50:1). Final conversion to aldehyde
23 was accomplished by demethylation of 29 with boron
tribromide which provided 2-bromosalicylaldehyde 23 in excel-
lent yield (94% yield).
Completion of the synthesis of carbene complex 20 requires
the acid-catalyzed cyclization of the allylic alcohol 22, a process
that is known for related allylic alcohols.10 The addition of 2
equiv of 2-methyl-1-propenyllithium to the hydroxybenzalde-
hyde 23 produced a mixture of the desired allylic alcohol 22
and the dienyl arene 31 (Scheme 5). With no purification, the
crude reaction mixture was subjected to acid-catalyzed cycliza-
tion which afforded a 5:1 mixture of the desired bromochromene
21 and of the 4-substituted chromene 32 which was obtained
as a mixture of olefin isomers. Further purification by Kugelrohr
distillation improved the ratio of 21:32 to 40:1 and provided
21 in 60% overall yield from aldehyde 23. The synthesis of the
carbene complex 20 was achieved by the standard Fischer
procedure as indicated in Scheme 5. Treatment of 21 with 2
equiv of tert-butyllithium and subsequent reaction with chro-
mium carbonyl and finally methylation with Meerwein’s salt
gave carbene complex 20 in 81% yield as a red crystalline solid.
The chromenyl chromium carbene complex 20 is remarkably
stable for an aryl chromium complex and could be separated
from the small amount of carbene complex generated from the
(8) Nicolaou, K. C.; Pfefferkorn, J. A.; Roecker, A. J.; Cao, G.-Q.; Barluenga,
S.; Mitchell, H. J. J. Am. Chem. Soc. 2000, 122, 9939-9953.
(9) Couture, A.; Deniau, E.; Grandclaudon, P.; Christophe, H. J. Org. Chem.
1998, 63, 3128-3132.
(10) (a) Talley, J. J. Synthesis 1983, 845-846. (b) Cruz-Almanza, R.; Perez-
Flores, F.; Cardenas, J.; Vazquez, C. Fuentes, A. Synth. Commun. 1994,
24, 1009. (c) Chauder, B. A.; Kalinin, A. V.; Snieckus, V. Synthesis 2001,
140-144.
9
11046 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

Chromene Chromium Carbene Complexes
A R T I C L E S
Table 2. Pd Catalyzed Triflate/Distannane Coupling of 24a
Scheme 6
ratio
entry
solvent
temp (
°
C)
time (h)
% conv
34:35
1
2
3
4
5
6
7
THF
60
105
105
110
120
160
110
48
8
48
96
8
3b
10
77
100
2
100
100
dioxane
dioxane
dioxane
dibutyl ether
DMF
47:53
90:10
0.2
110
26:74
30:70
bromide 32 by silica gel chromatography. In this way, carbene
complex 20 could be obtained in pure form in six steps from
o-anisaldehyde in 24% overall yield.
DMF
a Unless otherwise specified, all reactions were carried out at 0.1 M in
24 with a ratio of reagents: 24/(Me3Sn)2/Pd(PPh3)4/dppf/LiCl ) 1:0.9:0.02:
0.45:6. b dppf not used.
An alternative approach to the synthesis of the carbene
complex 20 begins with cyclohexan-1,3-dione and aldehyde 26
(Scheme 6). The base-catalyzed condensations of 1,3-diones and
R,â-unsaturated aldehydes are known to give pyrans.11 The
reaction of cyclohexan-1,3-dione with prenyl aldehyde in the
presence of ammonium salt of 1,2-diaminoethane gives the
pyran 25 in 78% yield. Subsequent DDQ oxidation gave the
benzopyran 33 which was then treated with triflic anhydride to
give the aryl triflate 24. The key conversion of the aryl triflate
24 to the bromide 21 was planned to utilize chemistry we had
previously developed for the conversion of aryl triflates to aryl
halides via a two-step sequence involving coupling of the aryl
triflate with a distannane and then conversion of the aryl
stannane to an aryl bromide with NBS or bromine.12 An
alternative method for the conversion of 24 to 21 involves the
reaction of triflate 24 with tributyl stannyl cuprate;13 however,
a variety of conditions were explored, but all failed to give any
coupling product and instead this reaction only resulted in the
recovery of triflate 24 and/or phenol 33.
The palladium-catalyzed coupling of aryl triflate 24 with
hexamethylditin was a slow reaction and gave mixtures of the
desired aryl stannane 34 and the proto-destannylated chromene
35 (Table 2). Very little reaction was seen in THF at 60 °C,
and thus it was necessary to raise the reaction temperature. The
solvent was thus changed to 1,4-dioxane, and it was also found
beneficial to add bis-(diphenylphosphinyl)ferrocene (dppf) to
provide a stable palladium complex at the higher temperatures.
Dioxane was the superior solvent as dibutyl ether gave consider-
able precipitation of palladium black and DMF gave substantial
amounts of the reduction product 35. At the optimal conditions
of 110 °C for 96 h, the reaction went to completion and gave
a 90:10 ratio (by GC) of the desired stannane 34 to the reduced
product 35 along with a small amount of a third unidentified
product. Purification of the stannane 34 was not possible by
silica gel chromatography as the desired product 34 coeluted
with the reduced product 35 and with the triphenylphosphine
from the catalyst. It was therefore found most convenient to
directly treat the crude reaction mixture containing the stannane
34 with NBS in THF at room temperature for 1 h which gave
the bromochromene 21 in pure form in 75% overall yield from
the triflate 24. This alternate route for the synthesis of carbene
complex 20 provides the complex in six steps in 24% overall
yield, a situation identical with the first route involving aldehyde
23 (Scheme 4). The choice between the two thus comes down
to the cost of reagents, and by this criteria, the first route is the
most desirable.
Reaction of the Chromenyl Carbene Complex 20 with
Acetylenes. Our studies of the benzannulation reactions of the
chromenyl carbene complex 20 with alkynes were initially quite
disappointing. The reaction of complex 20 with phenylacetylene
was performed under typical conditions. The reaction was
carried out in benzene at 50 °C until the carbene complex was
consumed (24 h) at which point the crude reaction mixture was
stirred in air for 12 h at room temperature to allow for the
oxidative decomplexation of any chromium tricarbonyl fragment
from the product. Surprisingly, neither the expected phenol 36a
or its corresponding quinone 37 could be detected in more than
trace amounts in the crude reaction mixture. The same reaction
of complex 20 with 1-pentyne gave a crude reaction mixture
that did not contain any of the expected phenol 38a and the
only product that could be purified from the mixture was the
quinone 39 in 16% yield. The failure to isolate any of the
phenols 36a and 38a was surprising, since, in our experience,
it is rare that phenol products from the reactions of Fischer
carbene complexes and alkynes are air-sensitive enough that
they would not survive exposure to air during the workup of
these reactions. It is even more unusual that the exposure of
these phenols to air did not lead to the formation of substantial
amounts of the corresponding quinones 37 or 39.
If the failure of the reactions shown in Scheme 7 is due to
the sensitivity of the phenols 36a and 38a to air, then a possible
solution might be able to protect the phenols in-situ which would
lead to the isolation of the aryl ethers 40 (Scheme 8). The
primary product from the reaction of chromium carbene complex
20 with alkynes is expected to be the chromium tricarbonyl
complex 40a (P ) H) with the chromium complexed to the
newly formed benzene ring. Normally, these chromium tricar-
bonyl complexed phenols are air sensitive, and the chromium
tricarbonyl group is quickly lost when the reaction mixture is
opened to air. Since chromium tricarbonyl complexes of aryl
(11) (a) Tietze, L.-F.; Kiedrowski, G.; Berger, B. Synthesis 1982, 683-684.
(b) Schuda, P. F.; Price, W. A. J. Org. Chem. 1987, 52, 1972-1979.
(12) Wulff, W. D.; Peterson, G. A.; Bauta, W. E.; Chan, K.-S., Faron, K. L.;
Gilbertson, S. R.; Kaesler, R. W.; Yang, D. C.; Murray, C. K. J. Org.
Chem. 1986, 51, 277-279.
(13) Gilbertson, S. R.; Challener, C. A.; Bos, M. E.; Wulff, W. D. Tetrahedron
Lett. 1988, 29, 4795-4798.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11047

A R T I C L E S
Scheme 7
Rawat et al.
Table 3. Benzannulation of Complex 20 with Phenylacetylenea
protecting
reagent
entry
methodb
product
P
% yield 36c
1
2
3
4
5
6
7
8
A
A
A
A
A
B
B
A
A
A
A
none
36a
36b
36b
36b
36c
36c
36c
36c
36d
36d
36e
H
trace
27
TBSCl
TBSCl
TBSOTf
TMSCl
TMSCl
TMSCl
TMSOTf
Ac2O
TBS
TBS
TBS
TMS
TMS
TMS
TMS
Ac
45d
57e
65
25f
62
25
9
10
11
57
Ac2O
MOMCl
Ac
MOM
57g
60
a Unless otherwise specified, all reactions were in benzene at 0.05 M in
20 for 24 h at 50 °C. Ratio of reagents: 20/phenylacetylene/protecting
reagent/Hunig’s base ) 1:2:3:5. b See Scheme 8. c Isolated yield. After silica
e
gel chromatography. d Reaction at rt for 6 d. Reaction at 60 °C; 36b was
isolated as a mixture and its yield was determined by 1H NMR. f Not
freeze-pump-thaw deoxygenated after Hunig’s base and TMSCl were
added. g DMAP (5 mol %) also added.
Scheme 8
protected phenol 40 without the chromium tricarbonyl group.
Concurrent and sequential protocols have been developed for
in situ protection (Scheme 8), and both were evaluated for
complex 20.
The reaction of the chromenyl carbene complex 20 with
phenylacetylene was examined with a number of different
electrophiles under both the concurrent and sequential proce-
dures (Table 3). All the reactions in Table 3 were performed in
benzene at 0.05 M in 20 with 2 equiv of alkyne, 3 equiv of
protecting agent, and 5 equiv of Hunig’s base. To ensure the
absence of oxygen, each reaction (except where specified) was
deoxygenated by the freeze-pump-thaw method at the begin-
ning of the reaction (three cycles) and after addition of the
Hunig’s base and protecting agent (two cycles, sequential
method). The optimal silylating agent proved to be trimethylsilyl
chloride which provided the silylated phenol 36c in 65% yield
as the metal-free arene (entry 5). The reaction of Fischer carbene
complexes with alkynes can produce a large number of different
products (dozens) depending on the substrates and conditions.3
Thus, the isolation of the protected phenols 36 demonstrates
that the major pathway for the reaction of the chromenyl
complex 20 leads to the normal phenol product and is supportive
of the contention that the phenol 36 is sensitive to air under the
reaction conditions. This is also supported by the lower yield
of 36c observed under the sequential protocol where the reaction
mixture is not deoxygenated after the addition of Hunig’s base
and protecting agent (entry 6 vs entry 7). This result could be
consistent with an air sensitivity of the phenol 36a or to the air
sensitivity of the chromium tricarbonyl complex of 36a. It is
surprising that this reaction is so sensitive to air, since many
reactions of carbene complexes with alkynes can be performed
without freeze-thaw deoxygenation resulting in no significant
loss in yield.15 Similar yields of the protected phenol 36 could
ethers are air stable, it might be expected that in situ protection
of the phenol function produced in these reactions would result
in the isolation of the protected phenol chromium tricarbonyl
complex 40a (P * H). However, in nearly all known examples
of in situ protection of phenol chromium tricarbonyl complexes
produced from the reaction of aryl carbene complexes, the
chromium tricarbonyl group is lost.14 It should be added that,
in contrast to the reaction of aryl complexes, the in situ
protection of the phenols produced from the reaction of alkenyl
complexes occurs with the isolation of the chromium tricarbonyl
complexed protected phenols in high yields.14 This dichotomy
is apparently due to the greater ease of displacement of η6-
naphthalene ligands compared with η6-benzene ligands. Thus,
based on what is known in the literature, the reaction of the
chromenyl complex 20 with alkynes with in situ protection of
the phenol function is expected to lead to the isolation of the
(14) (a) Fogel, L.; Hsung, R. P.; Wulff, W. D.; Sommer, R. D.; Rheingold, A.
L. J. Am. Chem. Soc. 2001, 123, 5580-5581. (b) Chamberlin, S.; Wulff,
W. D.; Bax, B. Tetrahedron 1993, 49, 5531-5547.
(15) Bos, M. E.; Wulff, W. D.; Miller, R. A.; Chamberlin, S.; Brandvold, T. A.
J. Am. Chem. Soc. 1991, 113, 9293-9319.
9
11048 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

Chromene Chromium Carbene Complexes
A R T I C L E S
Table 4. Solvent and Temperature Effects on Reaction of 20a
Scheme 9
entry
temp (
°
C)
solvent
% yield 36cb
1
2
3
4
5
6
50
50
50
75
100
125
THF
50
55
65
67
68
68
CH3CN
benzene
benzene
benzene
benzene
agent and base (entry 3 vs 4). Very efficient trapping was also
observed from the reactions with 3-hexyne and trimethylsily-
lacetylene giving the protected phenols 41c and 42c in 96%
and 89% yields, respectively. All of the trimethysilylaryl ethers
in Tables 4 and 5 are stable to silica gel except 41c. In this
case, the reaction of 20 and 3-hexyne was very clean, and simple
filtration through Celite gave essentially pure material. The TBS-
protected analogue 41b obtained from the reaction of 20 with
3-hexyne in the presence of TBSCl is stable to silica gel and
can be obtained in pure form. In contrast, the trapping of the
reactions of 1-pentyne and trimethylsilylacetylene with TBSCl
is not clean giving 38b along with some of the quinone 39 for
the former and a very complex reaction mixture with the latter.
Electrophiles other than silyl halides can be used as demon-
strated by the isolation of the aryl triflate 38f in 69% from the
reaction of 20 with 1-pentyne and also by the isolation of the
MOM ether 42e in 85% yield from the reaction of 20 with
trimethylsilylacetylene.
As outlined in Scheme 2, we anticipated that the triflation of
the phenol function in 8 may allow for the introduction of carbon
substituents in the 7-position of the 3H-naphtho[2,1-b]pyrans
1. Thus a significant finding in this regard is that the phenol
obtained from the reaction of carbene complex 20 with
1-pentyne could be trapped with triflic anhydride to give the
aryl triflate 38f in 69% yield. It is curious that the trapping of
the phenol from this reaction with triflic anhydride was
successful under sequential (Method B) but not concurrent
(Method A) protocols (Table 5, entries 6 and 7). Trimethylsilyl
chloride gives approximately the same yields under either
reaction method for either 1-pentyne (Table 5, entries 2 and 4)
or phenylacetylene (Table 3, entries 5 and 7). The origin of
this effect is not understood at this time. There was some
concern with the prospect of success for the Suzuki coupling
of the aryl triflate 38f given the fact that the triflate function is
flanked by substituents on either side. This concern was
misplaced as triflate 38f underwent Suzuki coupling with
phenylboronic acid under standard conditions16 to give the
7-phenyl-substituted naphthopyran 43 in 69% yield (Scheme
9).
a Unless otherwise specified, all reactions were 0.05 M in 20 for 24 h.
Ratio of reagents: 20/phenylacetylene/TMSCl/Hunig’s base ) 1:2:3:5.
b Isolated yield after silica gel chromatography.
Table 5. Benzannulation of Complex 20 with Acetylenesa
protecting
reagent
% yield
entry
methodb
R8
R9
product
P
38, 41, 42c
1
2
3
4
5
6
7
8
none
n-Pr
n-Pr
n-Pr
n-Pr
n-Pr
n-Pr
n-Pr
Et
H
H
H
H
H
H
H
38a
38c
38c
38c
38b
38f
38f
H
16d
97
50e
89
88f
g
69
96h
85
89
g
A
B
B
A
A
B
A
A
A
A
A
A
TMSCl
TMSCl
TMSCl
TBSCl
Tf2O
TMS
TMS
TMS
TBS
Tf
Tf2O
Tf
TMSCl
TBSCl
TMSCl
TBSCl
MOMCl TMS
MOMCl TMS
Et 41c
Et 41b
H
H
H
H
TMS
TBS
TMS
TBS
MOM
MOM
9
Et
TMS
TMS
10
11
12
13
42c
42b
42e
42e
85
70i
a Unless otherwise specified, all reactions were in benzene at 0.1 M in
20 for 24 h at 50 °C. Ratio of reagents: 20/acetylene/protecting reagent/
Hunig’s base ) 1:2:3:5. b See Scheme 8. c Isolated yield after silica gel
chromatography. d Only 16% yield of quinone 39 was isolated. e Not
freeze-pump-thaw deoxygenated after the addition of base and protecting
reagent. f Includes a 10% yield of the quinone 39. g Complex mixture
observed which was not analyzed. h Isolated by filtration through Celite.
i Reaction at rt for 5 d.
be obtained with acetic anhydride and chloromethylmethyl ether
as trapping agents. The structure of 36d was confirmed by X-ray
diffraction. The reactions of Fischer carbene complexes with
alkynes can be sensitive to solvent and temperature.15 However,
as indicated by the data in Table 4, the reaction of complex 20
with phenylacetylene is not at all sensitive to temperature and
only slightly sensitive to solvent with benzene giving slightly
higher yields than THF or acetonitrile.
The trapping of the phenol generated from the reaction of
complex 20 with 1-pentyne was much more efficient than with
phenylacetylene giving a 97% yield of the TMS ether 38c (Table
5). The air sensitivity of the intermediates in this reaction is
thus clear by comparison of this result with that from the reaction
with no trap (entry 1 vs 2) and with the reaction in which
deoxygenation was not performed after the addition of protecting
Deprotection of the trimethylsilyl ether 38c was performed
to more properly determine the stability of the phenol 38a.
Treatment of 38c with TBAF in THF at 0° for 30 min gave the
phenol 38a in 75% yield after purification by silica gel
chromatography (Scheme 10). Although phenol 38a was
obtained in relatively pure form, it contained a small amount
(16) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457-2483.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11049

A R T I C L E S
Scheme 10
Rawat et al.
Scheme 12
Scheme 11
Scheme 13
of impurity and decomposed at a rate such that the impurity
could not be removed with repeated chromatographic purifica-
tions. Oxidation of phenol 38a with ceric ammonium nitrate
(CAN) gave the quinone 39 in 56% yield. The quinone 39 was
stable and robust to both air and light and was completely
characterized. Given the sensitivity of the phenol 38a, the best
method for access to the quinone 39 was treatment of the
trimethylsilyl ether 38c with TBAF and then treatment of the
crude phenol 38a with CAN which results in a 66% yield of
the quinone 39.
The direct conversion of the carbene complex 20 to the
quinone 39 was examined with and without in situ generation
of the trimethylsilyl ether 38a (Scheme 11). After the reaction
of complex 20 and 1-pentyne was carried out at 50° for 24 h in
the presence of trimethylsilyl chloride and Hunig’s base, the
reaction was worked up to provide the crude silyl ether 38c.
This silyl ether was not purified but directly treated with TBAF,
and then the resultant phenol 38a was directly oxidized to the
quinone 39 without purification. The overall yield for this
conversion of complex 20 to quinone 39 was 65% yield which
is essentially identical with the overall yield that was observed
when the silyl ether 38c was isolated and purified (97% × 66%
) 64%). This is to be compared with the direct conversion of
carbene complex 20 to the quinone 39 without in situ trapping
of phenol 38a with trimethylsilyl chloride which is only
marginally less effective (58%).
ylsilyl ether 36c is isolated and purified (65% × 32% ) 21%).
This lower yield of quinone 37 compared with quinone 39
suggests that the phenol 36a from phenylacetylene is more
sensitive to decomposition than the phenol 38a from 1-pentyne.
It further suggests that it may be possible to optimize the yield
of quinones from these reactions if the process is optimized for
the oxidizing agent especially when considering that the
reactions with 1-pentyne, 3-hexyne, and trimethysilyl acetylene
give high yields of the trapping products with trimethylsilyl
chloride (Table 5).
Reaction of Chromenyl Carbene Complex 20 with Triynes.
Based on our limited studies on the reaction of Fischer carbene
complexes with conjugated triynes, the reaction of the chromenyl
carbene complex 20 with the bis-silyl substituted triyne 44 would
be expected to give either the phenol product 45 or the furan
product 46 (Scheme 13).7 As summarized in Scheme 3, the
reaction of the phenyl complex 15 with 44 gives the furan 15,
whereas the reaction of the cyclohexenyl complex 18 gives the
phenol 17 in 70% yield. Thus based on these observations and
the fact that the chromenyl complex 20 is an aryl-substituted
carbene complex, we would expect that the furan 46 would be
the major product formed from the reaction of complex 20 with
triyne 44.
Deprotection of the silyl ether 36c obtained from the reaction
of complex 20 with phenylacetyelene gave the phenol 36a in
65% yield (Scheme 12). Like the phenol 38a, the phenol 36a
was relatively unstable and was not fully characterized but rather
oxidized with CAN to give the quinone 37 in 50% yield. The
quinone 37 could be obtained directly from the reaction of the
carbene complex 20 and phenylacetylene in the absence of any
trapping agent in 25% yield if the crude reaction mixture is
oxidized with CAN. This is essentially the same overall yield
from carbene complex 20 that is observed when the trimeth-
The reaction of carbene complex 20 with triyne 44 gave
neither the phenol 45 nor the furan 46 as was expected. Instead,
this reaction lead to the formation of the naphthalene deriative
47a that results from incorporation of the triyne and then
9
11050 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

Chromene Chromium Carbene Complexes
A R T I C L E S
Scheme 14
the carbene complex 20 and addition to the central alkyne unit
in triyne 44 would be expected to give the h1,h3-vinyl carbene
complex 50 as either the E- and Z-isomer or as a mixture. It is
also possible that the E- and Z-isomers of 50 are in equilibrium
with respect to product formation.17c The phenol product 45
can only arise from the E-isomer of 50 via the ketene complex
51, and previous work18 suggests that the furan product 46 arises
from the Z-isomer of 50 via the ketene complexes 52. For the
reactions of carbene complexes and alkynes in general, one of
the most common side-products that arises from the E-isomer
of the vinyl carbene complexed intermediate is an indene
product, and in the case of the triyne 44, this would be the indene
derivative 54. This product was not detected in the reaction of
complex 20 with triyne 44, nor was the phenol 45 or the furan
46 observed from this reaction. The alkene addition product
47a appears to derive from the chromacyclobutane intermediate
56 that would be expected from an intramolecular [2 + 2]
cycloaddition of the chromium-carbon double bond in 55 with
the alkene in the chromene ring. The formation of a chroma-
cyclobutane has been widely envoked in the reaction of carbene
complexes with alkenes as the penultimate intermediate in the
formation of cyclopropanes via reductive elimination.19 No
cyclopropane product was observed in the reaction of 20 with
44, although the possibility exists that the alkene insertion
product 47a is a secondary product of the reaction resulting
from an acid-catalyzed isomerization of 57. Although relatively
rare, chromacyclobutane intermediates have been reported to
undergo â-hydride elimination and then reductive elimination
of hydride to give an alkene product instead of cyclopropane
products.20 This type of process could account for the formation
of the alkene insertion product 47a via reductive elimination
from the chromium(II) hydride intermediate 58.
Table 6. Reaction of Complex 20 with Triyne 44a
entry
solvent
% yield 47ab
1
2
3
4
5
6
7
benzenec
benzened
benzene
CH2Cl2
THF
74
71
80
88
75
CH3CNe
DMFe
a Unless otherwise specified all reactions were carried out at 0.05 M in
20 with 1.0 equiv of 44. b Isolated yields after purification by silica gel
chromatography. c TMSCl (2 equiv) and Hunig’s base (4 equiv) were used.
d MOMCl (2 equiv) and Hung’s base (4 equiv) were used. e Complex
mixture observed which contained at most a trace of 47a.
The reaction manifold outlined in Scheme 15 was also
explored with quantum calculations. The geometries of the
reactants, intermediates, and products were fully optimized with
the Spartan 5.1.3 program by the semiempirical PM3tm method.
The PM3tm optimized geometries were subjected to BP86
single-point (SP) calculations with the DN* basis set as
implemented in the Spartan program. The details of these
calculations can be found in the Supporting Information, and
the results are consistent with experiment. The formation of all
four products, 45, 46, 54, and 47a, are all predicted to be
exothermic with 47a as the most favored. The E-isomer of the
vinyl carbene intermediate 50 was found to be more stable than
the Z-isomer. Of the three intermediates that could eminate from
(E)-50, all of the intermediates on the pathway to 47a, namely
55, 56, 57, and 58, were lower in energy than either 51 or 53.
cyclization to the double bond of the chromene ring (Scheme
14). Two regioisomers of 47a would also be possible from this
reaction which would result from the different modes of
incorporation of the triyne. The regioisomers 48 and 49 were
ruled out as possibilities with the cleavage of the silyl groups
in the product from the reaction. Only isomer 47a could give a
desilylated product that had two different acetylene protons.
Treatment of 47a with TBAF gave the bis-alkyne 47b in 68%
yield with two alkynyl proton singlets at δ ) 3.55 and 3.56
ppm.
The reactions of complex 20 and triyne 44 were initially
performed in the presence of trapping agents in an effort to trap
the phenol 45 with the thought that it may be unstable like 38a
and 36a. The reaction was run in benzene in the presence of
base and trimethylsilyl chloride and methyl chloromethyl ether
(Table 6, entries 1 and 2). The only product that was observed
to be silica gel mobile in these reactions was the alkene addition
product 47a. The yield of 47a was higher if the trapping agents
were omitted (Table 6, entry 3). The best yield of the addition
product 47a was found in methylene chloride (88%), whereas
the use of the polar coordinated solvents acetonitrile and DMF
lead to a complex mixture of products in which 47a was present
in at most trace amounts.
(17) For references to mechanistic issues and leading references, see: (a)
Gleichmann, M. M.; Do¨tz, K. H.; Hess, B. A. J. Am. Chem. Soc. 1996,
118, 10551-10560. (b) Torrent, M.; Duran, M.; Sola, M. J. Am. Chem.
Soc. 1999, 121, 1309-1316. (c) Waters, M. L.; Bos, M. E.; Wulff, W. D.
J. Am. Chem. Soc. 1999, 121, 6403-6413. (d) Barluenga, J.; Aznar, F.;
Gutierrez, I.; Martin, A.; Garcia-Granda, S.; Llorca-Baragano, M. A. J.
Am. Chem. Soc. 2000, 122, 1314-1324.
(18) (a) McCallum, J. S.; Kunng, F.-A.; Gilbertson, S. R.; Wulff, W. D.
Organometallics 1988, 7, 2346-2360. (b) Parlier, A.; Rudler, M.; Rudler,
H.; Goumont, R.; Daran, J.-C.; Vaissermann, J. Organometallics 1995, 14,
2760-2774.
(19) For reviews, see: (a) Doyle, M. P. In ComprehensiVe Organometallic
Chemistry II; Able, E. W., Stone, F. G. A., Wilkinson, G., Eds.;
Pergamon: Oxford, 1995; Vol 12, p 387. (b) Brookhart, M.; Studabaker,
W. B. Chem. ReV. 1987, 87, 411-432.
Mechansitic Discussion. A possible mechanism17 is shown
for the formation of the alkene addition product 47a in Scheme
15 along with the other products that could have arisen from
this reaction, the furan 46, the phenol 45, the indene 54, and
the cyclopropane 57. Loss of a carbon monoxide ligand from
(20) (a) Wienand, A.; Reissig, H.-U. Angew. Chem., Int. Ed. Engl. 1990, 29,
1129. (b) Hwu, C.-C.; Wang, F.-C.; Yeh, M.-C. P.; Sheu, J.-H. J.
Organomet. Chem. 1994, 474, 123-128. (c) Barluenga, J.; Gonzalez, R.;
Fananas, F. J. Organometallics 1997, 16, 4525-4526. (d) Woodgate, P.
D.; Sutherland, H. S. J. Organomet. Chem. 2001, 628, 155-168.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11051

A R T I C L E S
Scheme 15
Rawat et al.
Scheme 16
The results show that the â-hydride elimination pathway via
58 is preferred over the cyclopropanation pathway via 57 by
∼12 kcal/mol. Although the energy of (E)-50 is lower than that
for (Z)-50, the second lowest pathway is that to the furan product
46, since the intermediate 52 is lower in energy than either the
intermediates 51 or 53.
Reaction of the Dihydrochromenyl Complex 60 with
Alkynes. The reaction of the carbene complex 20 with several
internal and terminal alkynes proceeds to give the normal
benzannulated product (Table 5), and thus the reaction with the
triyne 44 to give the alkene inserted product 47a was quite
unexpected and thwarted the idea for a straightforward access
to conocurvone (Scheme 3). If the double bond in complex 20
was removed, the formation of the alkene insertion product
would be obviated and then the question would revert to whether
the reaction with triyne 44 would give the normal phenol product
or the furan product (Scheme 13). The dihydrochromene carbene
complex 60 was prepared as outlined in Scheme 16. The double
bond in the bromochromene 21 was selectively reduced in the
presence of the aryl bromide with rhodium on alumina21 to give
the aryl bromide 59 in 92% yield. As expected the dihydro-
chromene carbene complex 60 will react with 1-pentyne in the
presence of TBSCl as trapping agent to give the normal
benzannulated product 61 in 77% yield. The key issue is the
reaction with the triyne 44 which was found to proceed to
cleanly give the furan product 62. This result is consistent with
the calculations described above and adds to the growing
evidence that aryl complexes are selective for furan products
(21) Nakazaki, A.; Sharma, U.; Tius, M. A. Org. Lett. 2002, 4, 3363-3366.
9
11052 J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006

Chromene Chromium Carbene Complexes
A R T I C L E S
Scheme 17
64 in 91% yield. Furthermore, it was found that the tungsten
complex 63 would not react with the triyne 44 to give significant
amounts of the alkene insertion product 47a. There is no reaction
at 90 °C, and at 130 °C the triyne 44 (1 equiv) was completely
consumed but the only silica gel mobile products observed were
47a (3%) and a 30% recovery of the carbene complex 63. This
suggests that the tungsten complex 63 is oligiomerizing the
triyne 44, a process that has been often noted in the reaction of
tungsten carbene complexes with alkynes.23
IV. Conclusion
It has been shown that the reactions of chromenyl carbene
complexes of chromium with alkynes can provide for a direct
entry to 3H-naphtho[2,1-b]pyrans and 3H-naphtho[2,1-b]pyran-
7,10-diones that are highly substituted in the C-ring. This
chemistry should be useful in the synthesis of natural products
and photochromic materials containing these two rings systems.
The degree and nature of the substituents at the C7-C10
positions in the C-ring can be controlled by the nature of the
alkyne and by post-benzannulation modification of the oxygen
substituent at C-7 (and thus presumably also at C-10). Although
it was found that the reaction of chromenyl complexes with
conjugated triynes will not be useful for the construction of the
central quinone unit in conocurvone (it may still be useful for
the two end quinone units), this reaction, nonetheless, proved
to be quite interesting as it occurs to give a product type that
has not been previously observed for the reactions of Fischer
carbene complexes with alkynes. This reaction leads to a non-
CO insertion benzannulation onto the remote double bond
presence in the carbene complex. A mechanism is proposed for
this process that involves a â-hydride elimination from a
chromacyclobutane intermediate. Synthetic applications of the
reactions with chromenyl carbene complexes with alkynes will
be reported in due course.
in their reactions with triyne 44 and alkenyl complexes are
selective for the phenol products.7
Reaction of the Tungsten Chromenyl Complex 63 with
Alkynes. The formation of the alkene insertion product 47a from
complex 20 and triyne 44 occurs without the insertion of a
carbon monoxide ligand. When the alkene in 20 is removed,
i.e., complex 60, then CO insertions occurs and the phenol
product 61 is formed upon reaction with 1-pentyne. It thus
appears that CO insertion is interrupted and prevented by
coordination of the metal to the alkene and the formation of
the intermediate 55 (Scheme 15). In an effort to generalize this
alkene insertion process and to expand its scope from triynes
to simple alkynes, we decided to investigate the reactions of
the tungsten complex 63, the analogue of chromium complex
20. The reason is that, relative to chromium complexes, tungsten
complexes are well-known to have much less of a propensity
for producing CO inserted products in their reactions with
alkynes.22 Thus, the tungsten should alter the competition
between the CO insertion and alkene coordination pathways in
the tungsten intermediate corresponding to 50-E in Scheme 15
in favor of alkene coordination and thus the formation of the
alkene insertion product 47a. While the reaction of the tungsten
complex 63 with 3-hexyne did not produce any CO insertion
products, the alkene insertion pathway did not compete with
indene formation that derives from a cyclization to the arene
ring in intermediate 50-E to give the indene product 54 perhaps
via the intermediacy of the metallocycle 53. The reaction of
tungsten complex 63 with 3-hexyne gives the indene product
Acknowledgment. This work was supported by a grant from
the National Institutes of Health (GM 33589).
Supporting Information Available: Procedures for the prepa-
ration of new compounds, characterization data for all new
compounds, and details on QM calculations of the reaction of
complex 20 with triyne 44. This material is available free of
charge via the Internet at http://pubs.acs.org.
JA0568852
(23) (a) Foley, H. C.; Strubinger, L. M.; Targos, T. S.; Geoffroy, G. L. J. Am.
Chem. Soc. 1983, 103, 3064. (b) Katz, T. J.; Lee, S. J. J. Am. Chem. Soc.
1980, 102, 422-424.
(22) Wulff, W. D.; Bax, B. M.; Brandvold, T. A.; Chan, K. S.; Gilbert, A. M.;
Hsung, R. P.; Mitchell, J.; Clardy, J. Organometallics 1994, 13, 102-126.
9
J. AM. CHEM. SOC. VOL. 128, NO. 34, 2006 11053