Coupling of Fischer carbene complexes with conjugated enediynes featuring radical traps: Novel structure and reactivity features of chromium complexed arene diradical species

Article abstract of DOI:10.1016/j.jorganchem.2008.08.003

The reaction of Fischer carbene complexes with conjugated enediynes that feature a pendant alkene group has been examined. The reaction proceeds through carbene-alkyne coupling to generate an enyne-ketene intermediate. This intermediate then undergoes Moore cyclization to generate a chromium complexed arene diradical, which then undergoes cyclization with the pendant alkene group. The radical cyclization prefers the 6-endo mode unless radical-stabilizing groups are present to favor the 5-exo mode. The intermediate diradical species were evaluated computationally in both the singlet and triplet configurations. Arene triplet diradicals feature minimal spin density at oxygen and delocalization to chromium. The 6-endo cyclization product was kinetically and thermodynamically favored.

Full text of DOI:10.1016/j.jorganchem.2008.08.003

Journal of Organometallic Chemistry 693 (2008) 3337–3345
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Coupling of Fischer carbene complexes with conjugated enediynes featuring
radical traps: Novel structure and reactivity features of chromium
complexed arene diradical species
Yi Zhang a, Tareq Irshaidat a, Haixia Wang b, Kris V. Waynant a, Haobin Wang a, ,1, James W. Herndon a, ,2
*
*
a Department of Chemistry and Biochemistry, New Mexico State University, MSC 3C, Las Cruces, NM 88003, USA
b Department of Chemistry and Biochemistry, University of Maryland, College Park, MD 20742-2021, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 30 June 2008
Received in revised form 4 August 2008
Accepted 4 August 2008
Available online 12 August 2008
The reaction of Fischer carbene complexes with conjugated enediynes that feature a pendant alkene
group has been examined. The reaction proceeds through carbene–alkyne coupling to generate an eny-
ne–ketene intermediate. This intermediate then undergoes Moore cyclization to generate a chromium
complexed arene diradical, which then undergoes cyclization with the pendant alkene group. The radical
cyclization prefers the 6-endo mode unless radical-stabilizing groups are present to favor the 5-exo
mode. The intermediate diradical species were evaluated computationally in both the singlet and triplet
configurations. Arene triplet diradicals feature minimal spin density at oxygen and delocalization to chro-
mium. The 6-endo cyclization product was kinetically and thermodynamically favored.
Ó 2008 Elsevier B.V. All rights reserved.
Keywords:
Fischer carbene
Conjugated enediyne
Diradicals
Radical cyclization
.
1. Introduction
insertion to produce the chromium-complexed enyne ketene spe-
cies G [7], which can then undergo the Moore cyclization to afford
The generation and reactivity of aromatic diradical species ema-
nating from the cyclization of highly conjugated and unsaturated
alkyne derivatives has been a very intense area of research since
the discovery enediyne anticancer agents [1]. The three major reac-
tions of this type are the Bergman cyclization of conjugated
enediynes, the Myers–Saito cyclization of enyne–allenes, and the
Moore cyclization of enyne–ketenes, depicted in Scheme 1 (con-
version of B–C in Scheme 1). This paper is focused on the Moore
cyclization [2].
The Moore cyclization is limited by the paucity of reaction pro-
cesses that generate enyne ketene intermediates. The most com-
mon process involves thermolysis of 4-alkynyl-2-cyclobutenones
(A). Alternative but less-commonly employed methods include
arene-complexed diradical species H [8]. An alternative method for
successful generation of related chromium-complexed arene
diradicals involves the coupling of amino-alkynylcarbene com-
plexes with alkynes [9]. In this case the diradical species were suc-
cessfully quenched with hydrogen atom donors. The trapping of
the diradical species through intramolecular hydrogen atom
abstraction processes and a single example of trapping through a
5-exo radical cyclization event have previously been reported [8].
In this paper additional examples of the alkene radical trapping
reaction and an important structural correction are reported,
which reveal interesting and unique features for diradicals gener-
ated in the coordination sphere of chromium.
thermolysis of appropriate
a-diazo ketones [3] or thermolysis of
2. Results and discussion
appropriate azidobenzoquinones [4], and ene reactions of eth-
oxyenediynes [5]. A potentially excellent source of enyne–ketenes
is the coupling of Fischer carbene complexes (E) with readily-avail-
able conjugated enediynes (D) [6]. Coupling of Fischer carbene
complexes with enediyne D should proceed through carbene al-
kyne coupling at the less hindered alkyne unit followed by CO
2.1. Experimental results
General synthetic approaches to the conjugated enediynes em-
ployed in this study are depicted in Scheme 2. The key step in this
process is selective halogen metal exchange in tribromodiene 2 [8],
easily obtained through Wittig reaction of the bromo-aldehyde 1
[10]. This process is selective for conversion of the gem dib-
romoalkene group to the enyne-bromide. Subsequent Sonogashira
coupling, Swern oxidation and Wittig reaction affords the enediy-
nes used in this study.
* Corresponding authors. Fax: +1 5056462649.
E-mail addresses: haobin@nmsu.edu (H. Wang), jherndon@nmsu.edu (J.W.
Herndon).
1
Computational studies at New Mexico State University.
Experimental studies at New Mexico State University.
2
0022-328X/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2008.08.003

3338
Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
bons are CH2 groups, and three are CH groups. Another important
.
O
finding since the original report is that the yield of 16a relative to
the naphthalene structure 15a is highly variable between experi-
mental runs. Longer reflux times lead to increased amounts of
15a, which suggests that excessive reflux time causes a slow dehy-
drogenation of intermediate 12a to 11a [12].
O
R
O
R
.
R
C
B
A
The reaction of analogous carbene complexes and enediynes
where the trapping alkene group is more substituted are depicted
in Table 1. In all of the examples in Table 1, separation of 15 and 16
proved to be very difficult, and the ratio of 15:16 was not constant.
Treatment of enediyne 6b with methylcarbene complex 7a led to
the dihydronaphthalene derivative 16b accompanied by minor
amounts of 15b. In this case no structural ambiguity exists for
13b since the methyl group originating at the trapping alkene
(d1.22, doublet, J = 6.0 Hz in the product 16b) and the alkene pro-
tons (d5.97, broad singlet integrating for 2 H’s in product 16b)
are clearly consistent only with 16b. The use of chlorobenzene as
solvent led to a slight improvement in the yield of 16b. Chloroben-
zene is a much poorer hydrogen atom donor compared to dioxane,
however is not the optimal solvent for carbene-terminal alkyne
coupling reactions that require CO insertion [13]. A similar reaction
OMe
R'
Cr(CO)5
+
Cr(CO)4
R'
OMe
E
F
R
D
R
OMe
R'
OMe
R'
O
Cr(CO)3
(CO)3Cr
.
O
R
G
H
R
.
Scheme 1.
Cr(CO)5
Me
OMe
In the original report of this reaction (Scheme 3) [8] involving
the coupling of enediyne 6a with Fischer carbene complex 7a fol-
lowed by acid treatment, the structures were assigned to the prod-
ucts based on the isolation and full characterization (1H NMR,
broad band-decoupled 13C NMR, IR, EI-MS, HRMS) for the major
product 13a and GC–MS and 1H NMR analysis of the crude reaction
mixture for the assignment of 14a and 15a. The reaction initially
afforded enol ethers 9a–12a, which cyclized to afford the furan
derivatives upon acid treatment. As more examples of this type
of reaction were acquired, it became apparent that the structure
of the major product 13a was likely misassigned. The isomeric
unconjugated dihydronaphthalene structure 16a was determined
to be a better structural fit to the major isomer [11]. Compound
16a results from a 6-endo radical cyclization rather than a 5-exo
radical cyclization (see Section 2.2). The most important evidence
is the DEPT C-13 spectrum, which shows that none of the sp2 car-
7a
6a
OMe
Me
(CO)3Cr
.
O
.
8a
OMe
Me
OH
OMe
OMe
Me
OH
OMe
Me
Me
OH
OH
CBr4 / Zn /
PPh3
2 eq n-BuLi,
Br
CHO
Br
then TBSCl
(CO)3Cr
12a
(CO)3Cr
(CO)3Cr
Me
(CO)3Cr
Br
Br
2
10a
H+
11a
1
9a
H+
H+
TBS
OH
H+
TBS
DMSO / Et3N /
ClCOCOCl
Me
O
Me
O
Me
O
Me
O
Br
PdCl2 / CuI / PPh3
Et2N- -Pr
3
i
4
HO
TBS
13a
Me
15a
16a
14a
R
1. Ph3P=CHR
2. TBAF / THF
Reported in J. Org. Chem. 1998, 63, 4562-4563:
13a - 54%, 14a - 13%, 15a - 4%, 16a - not formed
6
a R = H
Should be amended to:
5
OHC
b R = Me
13a - not formed, 14a - < 5% 15a - highly variable yield, 16a -
c R = COOMe
major product in 30-54% yield
Scheme 2.
Scheme 3.

Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
3339
Table 1
Coupling of diene–diynes with Fischer carbene complexes
R2
O
R2
O
R2
O
OMe
OH
Cr(CO)5
R1
1.
7
R2 OMe
2.
TsOH or I2
6
X
R1
R1
14
15
16
17
R1
Entrya
Reactants
R1
R2
X
Solvent
Yield 14(%)
Yield 15b (%)
Yield 16b (%)
1 (a-series)
2 (b-series)
3 (b-series)
4 (c-series)
5 (c-series)
6 (d-series)
6a + 7a
6b + 7a
6b + 7a
6b + 7b
6b + 7b
6c + 7a
H
Me
Me
Me
Me
COOMe
Me
Me
Me
Ph
Ph
Me
N/A
Dioxane
Dioxane
Ph–Cl
Ph–Cl
Benzene
Dioxane
Trace
14c
13
4
Trace
21
0
54c
40
45
45
0
N/A
N/A
N/A
N/A
H
48
0
a
Substituent identifiers for adduct compounds 9–17 and reactive intermediates 18–21 correlate with series letters adjacent to Entry numbers.
b
The yield of 15 and 16 refers to the isolation of a single chromatography fraction containing both 15 and 16; the relative proportion was determined through the 1H NMR
spectra.
c
Yield varies widely in experimental runs, however total yield of 15a + 16a is consistently about 60%.
OMe
Me
OH
OMe
Me
(CO)3Cr
(CO)3Cr
R-H
.
O
OMe
Me
(CO)3Cr
.
O
R
18
R
10
.
.
OMe
OMe
Me
8
R
Me
OH
(CO)3Cr
(CO)3Cr
.
O
19
R
.
12
OMe
Me
OMe
Me
MeO
Me
O
O
R
O
R
H
H
(CO)3Cr
(CO)3Cr
R
(CO)3Cr
20
H
21
Scheme 4.
process occurred using the phenylcarbene complex 7b. A notewor-
thy observation in this reaction is that no product consistent with
the Dötz reaction product 17 could be isolated from this reaction.
This same reaction in benzene led to the surprisingly selective for-
mation of the naphthalene derivative 15c, however in low yield. A
similar type of reaction was examined using electron-deficient al-
kenes as the radical trap (entry 6). In these cases, the exclusive
reaction products are in fact the reduced 5-exo cyclization product
14d.
2.2. Discussion of experimental results
A reasonable mechanism for the formation of the observed
products is depicted in Scheme 4. After formation of diradical
intermediate 8 via the Moore cyclization, radical cyclization can
occur in either the 5-exo or the 6-endo mode. Cyclization in the
5-exo mode affords intermediate 10, the precursor to benzofuran
derivative 14, after transfer of two hydrogen atoms from the sol-
vent. Radical cyclization in the 5-exo mode is typically favored

3340
Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
for arene radicals [14] and for arene-complexed monoradicals [15].
We are not aware of efforts to trap Moore cyclization-derived
diradicals through radical cyclization, however they have been
trapped through intramolecular hydrogen atom abstraction
[4,16]. Successful trapping of related Bergman cyclization and
Myers cyclization derived diradicals through 5-exo cyclization
has also been reported, however in most of these cases the trap-
ping alkene is a conjugated ester, and reduced products analogous
to 14 are typically formed when these reactions are performed in
the presence of hydrogen atom donors [17]. In simple radical cycli-
zation reactions, the observation of 6-endo products has typically
been attributed to 5-exo-cyclization followed by the neophyll rear-
rangement process [18]. A similar pathway involving the 6-endo
radical cyclization pathway leads to intermediate 12, the precursor
to compounds 15 and 16. The 5-exo-derived diradical intermediate
18 undergoes hydrogen abstraction, while the 6-endo-derived
intermediate 19 undergoes internal hydrogen transfer. A likely
explanation for the diverging pathway is due to interaction of
the radical species with chromium in intermediate 19. No such
interaction exists for intermediate 18 due to the anticipated stere-
oselectivity, which places the chromium tricarbonyl unit and the
radical substituent on opposite faces of the molecule due to steric
interaction between the chromium tricarbonyl and the –CHRÅ
groups [19]. In intermediate 19, interaction of the radical centers
would result in 18-electron complex 20. Formation of compound
12 from intermediate 20 can be viewed as a b-hydride elimination
leading to intermediate 21 followed by reductive elimination and
enolization. This process would result in the observed non-conju-
gated double bond isomer since this isomer arises from the only
b-hydrogen syn to chromium in 20. Significant amounts of the re-
duced 5-exo product 14 were observed only in the case where R is
the electron-withdrawing and radical-stabilizing group COOMe.
This is likely due to the stabilization of the 5-exo cyclization de-
rived radical by the COOMe group.
benzylic and homobenzylic positions have been reported [20]. This
study suggested that arene complexation has no major effect on
the stability of homobenzylic monoradicals. Hypothetical radical
intermediates 18–19 and 24–25 are in fact arene complexed
homobenzylic radicals, however these compounds contain a sec-
ond unpaired electron.
Density function theory (DFT) was used to investigate the inter-
mediates 22–25 and their respective transition states. All calcula-
tions were performed with the quantum chemical program
package Gaussian 03 [21]. The B3LYP hybrid functional, which in-
cludes the Becke three-parameter exchange [22] and the Lee, Yang,
and Parr correction functionals [23], was used in the DFT calcula-
tion. The SDD [24] basis set was employed for chromium whereas
*
the standard 6-31G basis set was used for all other atoms. All of
the diradical species were evaluated as both singlet and triplet
diradicals, and in some cases higher spin configurations were also
considered. The resulting conformations and their respective ener-
gies are depicted in Fig. 1 and Table 2. The reaction processes are
displayed on a potential energy diagram in Schemes 6 and 7. Se-
lected spin densities and bond lengths are depicted in Table 3.
The enyne–ketene complex 22 exists as a diamagnetic low spin
complex, which fits related experimental observations where NMR
spectra are routinely acquired for various polyene species coordi-
nated to the Cr(CO)3 metal–ligand system. The alkyne group is
tilted toward chromium (C5-6-7 angle = 161°, see Fig. 1 for atom
numbers), possibly due to the 16-electron configuration at chro-
mium. Higher spin analogues of the enyne–ketene complex 22
were also evaluated. Both the triplet state (+11.4 kcal/mol relative
to singlet) and quintet state (+23.3 kcal/mol relative to singlet)
were evaluated. In both of these higher spin configurations the spin
is concentrated at the chromium atom and not delocalized to any
atoms of the polyene ligand.
The initially-formed arene diradical 23 can exist in either the
singlet or triplet configuration, depending on whether the individ-
ual radical species are spin-paired (singlet) or possess the same
spin orientation (triplet). The triplet configuration is more stable
than the singlet configuration by 8.5 kcal/mol, however the singlet
configuration is presumably the species initially produced from
singlet enyne–ketene complex 22. The spin density of the triplet
diradical exists predominantly at chromium and the aromatic car-
bon designated as C6. Virtually none of the spin density resides at
the phenolic oxygen (O1), which is in contrast to the phenoxy rad-
ical. The C6–Cr bond distance is considerably shorter in the singlet
diradical (1.89 Å) compared to the triplet diradical (2.23 Å). The C–
C bond angles at C6 are also different in the singlet (C5-6-7 an-
gle = 133.4°) and triplet (C5-6-7 angle = 127.5°) configurations.
2.3. Results and discussion (computational)
The processes depicted in Scheme 4 were evaluated computa-
tionally to gain insight into the rationale for the observed reaction
pathways and to better understand the role of chromium in the
stabilization and reactivity of diradical intermediates. The reaction
pathway under investigation involves the transformation of eny-
ne–ketene complex 22 (Scheme 5) into cyclized diradicals (24
and 25). Optimized structures for reactive intermediates 22–25
are depicted in Fig. 1. Theoretical and experimental investigations
of arene–chromium complexes that contain a mono-radical at the
.
.
Cr(CO)3
Cr(CO)3
Cr(CO)3
Cr(CO)3
5
4
3
2
1
.
6
7
.
O
.
O
O
.
O
8
24
9
23a
23
10
22
11
.
.
Cr(CO)3
Cr(CO)3
O
O
.
25
.
26
Scheme 5.

Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
3341
Fig. 1. Energy-minimized conformations of key reactive intermediates.
Intermediate diradical 23 is best represented as the
g
5-cyclohexa-
1.23 Å in the triplet) is indicative of substantial double bond char-
acter [25]. By comparison, the corresponding calculated C–O bond
lengths are considerable longer in the optimized structures for
phenol (1.42 Å) and the simple phenoxy radical (1.30 Å), compared
with 1.24–1.27 Å in stable phenoxy radicals [26]. In the simple
phenoxy radical the spin density is also concentrated at oxygen
(0.51).
dienyl-Cr(I) resonance form 23a rather than the
g
6-arene complex,
regardless of the spin orientation of the radical species. This is also
the more appropriate description for related diradicals 24 and 25.
In 23–25 (all spin configurations) the C2–Cr distance is consider-
ably longer that the C–Cr distance for all of the other aromatic ring
carbons. The O1–C2 bond length of 23a (1.22 Å in the singlet,

3342
Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
Table 2
The Moore cyclization event (conversion of ketene complex 22 to
arene diradical complex 23a) occurs with an activation free energy
of +9.6 kcal/mol. Transition state calculations were conducted only
for the conversion of 22 to the singlet form of 23a since 22 is a sin-
glet species. The reaction pathway involving the singlet diradical
species can result only in the 6-endo product due to the facile con-
version of the 5-exo product to the 6-endo product through an en-
ergy barrier lower than that required to recognize a local minimum
in a geometry optimization. If the reaction pathway proceeds
exclusively through spin-paired diradicals, the 6-endo pathway is
the only allowed reaction pathway, regardless of kinetic prefer-
ences in the radical cyclization mode. This process is highly exer-
Relative energies (kcal/mol) for intermediates 22–25 and key transition states
¼ ¼
TS’s DG (DH )
Relative energies – G (H)
22
0 (0)
22 to 23a-s 9.6 (8.5)
23a-s to 25-s 5.2 (3.1)
23a-t to 24-t 5.4 (3.2)
23a-t to 25-t 3.6 (1.2)
23a-s
23a-t
24-t
25-s
25-t
+0.7 (ꢀ1.1)
ꢀ7.8 (ꢀ8.0)
ꢀ31.6 (ꢀ33.9)
ꢀ37.2 (ꢀ43.9)
ꢀ38.7 (ꢀ41.1)
The 5-exo cyclization product 24 was successfully optimized
only in the triplet configuration. All attempts to optimize the 5-
exo cyclization product in the singlet configuration led to the 6-
endo cyclization product 25-s. Intermediate species during the
progression of the optimization of 24-s bear resemblance to the
structure 26 depicted in Scheme 5. The triplet diradical 24-t prefers
the exo configuration of the RCHꢀ2 group relative to the Cr(CO)3
unit. The spin density is concentrated at chromium and at C11.
The 6-endo radical cyclization product 25 was successfully opti-
mized in both the singlet and triplet configurations, which are very
gonic (
activation of 5.2 kcal/mol.
D
G = ꢀ37.9 kcal/mol) and occurs with a free energy of
The pathway employing triplet intermediates is depicted in
Scheme 7. Intersystem crossing from initially-formed intermediate
23a-s leads to 23a-t in a process that is exergonic (
D
G = ꢀ8.5 kcal/
mol). The initially-formed radical can react at the pendant alkene
through either the 5-exo mode to afford intermediate diradical
24-t or through the 6-endo reaction mode to afford diradical 25-
t. Both of the hypothetical reaction products and transition states
were optimized. The 5-exo trig cyclization event is highly exer-
similar in energy (
D
G singlet ? triplet = ꢀ1.5 kcal/mol). The sin-
glet configuration features a very close contact between chromium
and the radical site C10 (2.43 Å), which thus supports the interme-
diacy of compound 20 in Scheme 4. Bond lengths for chromium
bound to sp3 carbons obtained from X-ray structures vary from
2.26 to 2.34 Å [27]. The spin density in the triplet diradical 25-t
is concentrated at the radical site C10 and at chromium. The 6-
endo radical cyclization product 25-t is substantially more stable
gonic (
D
G = ꢀ23.8 kcal/mol) and occurs at a moderate free energy
of activation (5.4 kcal/mol). The 6-endo trig cyclization occurs with
a similar activation energy (
modynamically more stable product (
version of 23a-t to 25-t). If the triplet reaction pathway is preferred
the 6-endo and 5-exo radical cyclization pathways are expected to
be competitive. The 6-endo pathway is slightly preferred kineti-
cally and strongly preferred thermodynamically.
¼
D
G
= 3.6) however results in the ther-
G = ꢀ30.9 kcal/mol for con-
D
than the 5-exo cyclization product 24-t (
24-t to 25-t).
D
G = ꢀ7.1 kcal/mol for
The hypothetical reaction pathways, including transition states
for all singlet–singlet and triplet–triplet interconversions, are de-
picted in the potential energy diagrams in Schemes 6 and 7. The
pathway through singlet intermediates is depicted in Scheme 6.
In summary, the calculations predict that: (1) a reaction that
proceeds through the singlet diradical intermediates should result
in 6-endo cyclization derived products, and (2) a reaction that pro-
ceeds through triplet diradical intermediates could lead to either
10
TS 22 to 23a-s
TS 23a-s
0
to 25-s
.
(CO)3Cr
Cr(CO)3
.
Cr(CO)3
10
O
.
O
.
O
26
22
singlet
23a-s
20
30
(CO)3
.
40
Cr(CO)3
Cr
O
O
25-s*
25-s
.
Scheme 6.

Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
3343
10
0
TS 23a-t
to 24-t
.
Cr(CO)3
TS 23a-t
to 25-t
.
O
10
20
30
.
Cr(CO)3
singlet
23a-s
.
.
O
Cr(CO)3
.
23a-t
O
24-t
.
Cr(CO)3
40
O
.
25-t
Scheme 7.
Table 3
Selected spin densities (S.D)a and bond lengths (Å) for intermediates 22–25
Compound
S.D. @ Cr
S.D. @ C6
S.D. @ C10
S.D.@ C11
S.D.@ O1
C2–O1
C2–Cr
C3–Cr
C4–Cr
C5–Cr
C6–Cr
C7–Cr
C10–Cr
C11–Cr
22
23a-s
23a-t
24-t
25-s
25-t
1.22
1.23
1.23
1.23
1.23
2.69
2.62
2.58
2.65
2.55
2.36
2.30
2.28
2.35
2.29
2.30
2.25
2.23
2.27
2.27
2.18
2.23
2.25
2.23
2.25
1.89
2.23
2.33
2.18
2.29
2.26
2.35
2.36
2.27
2.33
1.13
1.06
0.82
0.06
0.04
1.07
2.43
3.99
3.20
3.44
1.02
1.02
0.05
a
If blank, the spin density is less than 0.1.
the 5-exo- or 6-endo-derived products in a kinetically-competitive
process where the 6-endo product is strongly favored thermody-
namically. A limitation in all of these calculations is that the acti-
vation energy for singlet to triplet diradical interconversions can
not be calculated. If the radicals freely interconvert between sin-
glet and triplet configurations, any 5-exo-derived product 24-t
could easily convert to the 6-endo-derived product 25-s through
a pathway involving conversion to the singlet 24-s followed by
spontaneous conversion to the singlet 6-endo product 25-s. The
free energy of the triplet diradical 25-t is very similar to that of
lizing ester groups are positioned to stabilize the 5-exo-derived
free radical intermediate. The reaction pathway was evaluated
computationally using both triplet and singlet configurations of
the diradicals species.
4. Experimental [28]
4.1. General procedure for the coupling of carbene complexes with
diene–diynes
the singlet diradicals 25-s (
D
G = ꢀ1.5 kcal/molꢀ1 for conversion
An 0.03–0.05 M dioxane or chlorobenzene solution of the
enediyne or dienyne (1.0 eq.) and the carbene complex (1.2 eq.)
was heated to 100 °C for 24 h using an oil bath heated to 100 °C.
The reaction mixture was then allowed to cool to room tempera-
ture and the solvent was removed on a rotary evaporator. Hexane
(15 mL) was added and the resulting green suspension was filtered
through Celite. Iodine (1.5 eq.) was added to the filtrate and this
solution was stirred at room temperature for 24 h. The reaction
mixture was poured into aqueous sodium thiosulfate solution in
a separatory funnel and the aqueous layer was extracted two times
with hexane. The combined hexane layers were washed with
of 25-s to 25-t) however the enthalpy is actually lower for the sin-
glet diradicals (DH = +2.8 kcal/mol for 25-s to 25-t).
3. Conclusions
In summary, the formation of arene diradicals through Moore
cyclization of chromium carbene-derived enyne ketenes followed
by subsequent radical cyclization reactions to pendant alkene
groups has been demonstrated. The reaction proceeds predomi-
nantly through the 6-endo cyclization mode unless radical-stabi-

3344
Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
saturated aqueous sodium chloride solution and dried over sodium
sulfate. The solvent was removed on a rotary evaporator and the
crude residue was purified by chromatography on silica gel using
19:1 hexane: ethyl acetate as the eluent.
J = 1.0 Hz), 2.23 (quintet, 2H, J = 7.7 Hz); 13C NMR (CDCl3): d154.5,
152.3, 135.3, 134.6, 133.4, 132.9, 128.1, 127.4, 124.5, 118.8,
102.1, 36.3, 31.5, 25.1, 24.5, 14.3.
4.4. Coupling of methylcarbene complex 7a with ethylidene enediyne
6b in chlorobenzene (Result in Table 1, Entry 3)
4.2. Coupling of methylcarbene complex 7a with simple enediyne 6a
(Result in Table 1, Entry 1)
The general procedure was followed using methylcarbene com-
plex 7a (0.458 g, 1.833 mmol) and dienediyne 6b (0.270 g,
1.466 mmol) and using chlorobenzene (30 mL) as solvent. Final
purification using flash column chromatography (silica gel/9:1
hexanes: ethyl acetate) yielded a single fraction (0.170 g, 49%
yield) that revealed to be a 9:1 mixture of compounds 16b (45%
yield) and 15b (4% yield). The spectral data were identical to the
same products obtained in dioxane (previous experiment).
The general procedure was followed using methylcarbene com-
plex 7a (0.198 g, 0.7906 mmol) and dienediyne 6a (0.112 g,
0.6588 mmol). The crude product was treated with p-toluenesul-
fonic acid monohydrate (0.0125 g, 0.06590 mmol) in dichloro-
methane (30 mL) for 12 h at room temperature. Purification
using preparative TLC (eluent: hexanes) yielded a single fraction
(0.105 g, 71% yield) which was revealed to be a 76:5:19 mixture
of benzofuran-alkene 16a (54% yield), benzofuran-alkane 13a (3%
yield), and naphthofuran 15a (14% yield). A second preparative
TLC was performed, and a pure sample of the major product, al-
kene 16a, was obtained by cutting the center of the band. Benzofu-
ran-alkene 16a: 1H NMR (CDCl3): d6.24 (s, 1H), 5.95 (br s, 2H), 3.55
(m, 2H), 3.30 (m, 2H), 2.99 (t, 2H, J = 7.3 Hz), 2.82 (t, 2H, J = 7.3 Hz),
2.43 (s, 3H), 2.15 (quintet, 2H, J = 7.3 Hz); 13C NMR (CDCl3): d154.4
(quat C), 152.5 (quat C), 136.0 (quat C), 132.1 (quat C), 124.8 (quat
C), 124.4 (CH), 123.6 (CH), 122.8 (quat C), 114.8 (quat C), 101.2
(CH), 31.4 (CH2), 31.0 (CH2), 27.4 (CH2), 25.1 (CH2), 23.7 (CH2),
14.2 (CH3); Mass Spec (El): 224 (M, 100), 223 (30), 209 (18), 196
(22), 185 (32), 182 (14), 181 (67), 179 (11), 178 (15), 167(14),
4.5. Coupling of phenylcarbene complex 7b with ethylidene enediyne
6b in chlorobenzene (Result in Table 1, Entry 4)
The General Procedure was followed using phenylcarbene com-
plex 7b (0.622 g, 1.994 mmol) and dienediyne 6b (0.306 g,
1.661 mmol) and using chlorobenzene (30 mL) as solvent. Final
purification using flash column chromatography (silica gel/9:1
hexanes:ethyl acetate) yielded a single fraction (0.233 g, 47% yield)
identified as a 19:1 mixture of benzofuran 16c (45% yield) and
naphthofuran 15c (2% yield). Further purification by preparative
TLC yielded benzofuran 16c free of naphthofuran 15c. Benzofuran
16c: 1H NMR (CDCl3): d7.85 (d, 2H, J = 8.0 Hz), 7.38 (m, 3H), 6.93
(s, 1H), 6.00 (br s, 2H), 3.74 (m, 1H), 3.49 (m, 2H), 3.06 (t, 2H,
J = 8.0 Hz), 2.94 (t, 2H, J = 8.0 Hz), 2.18 (m, 2H), 1.26 (d, 3H,
J = 7.0 Hz); 13C NMR (CDCl3): d156.3, 152.5, 135.9, 132.6, 131.8,
131.0, 128.6, 128.0, 128.5, 126.9, 124.6, 122.6, 115.5, 100.0, 32.4,
31.2, 29.7, 25.5, 23.8, 22.3; Mass Spec (El): 300 (M, 83), 298 (18),
286 (25), 285 (100), 257 (30), 229 (18), 154 (21), 105 (15),
91(13); HRMS calc. for C22H20O 300.1529. Found 300.1514. For
15c spectral data, see the next experiment.
166 (17), 165 (31), 153 (11), 152 (18); HRMS calc. for C16H16
O
224.12012. Found 224.11922. Naphthofuran 15a: 1H NMR: d8.27
(d, 1H J = 7.8 Hz), 7.85 (d, 1H J = 7.3 Hz), 7.51 (m, 2H), 6.45 (s,
1H), 3.30 (t, 2H, J = 7.4 Hz), 3.19 (t, 2H, J = 7.4 Hz), 2.59 (s, 3H),
2.34 (quintet, 2H, J = 7.4 Hz); 13C NMR: d156.5 (quat C), 150.1 (quat
C), 134.0 (quat C), 133.9 (quat C), 127.7 (quat C), 125.2 (CH), 125.0
(CH), 124.6 (CH), 122.5 (quat C), 120.5 (CH), 120.4 (quat C), 102.5
(CH), 32.5 (CH2), 31.4 (CH2), 24.8 (CH2), 14.5 (CH3); MS (EI): 222
(M, 100), 207 (MꢀCH3, 9), 193 (Mꢀethyl, 6), 179 (Mꢀacetyl, 41);
HRMS (ESI) calc. for C16H15O (MH+) 223.1123. Found 223.1112.
Spectral evidence for benzofuran-alkane 13a: 1H NMR (CDCl3):
The crude 1H NMR spectrum features peak at d 1.23 (d, J = 6.9)
which is consistent with the nonallylic methyl of 14a; GCMS (EI):
226 (M, 34), 211 (MꢀCH3, 100), 198 (Mꢀethylene, 3), 183
(Mꢀacetyl, 16).
4.6. Coupling of phenylcarbene complex 7b with ethylidene enediyne
6b in benzene solvent (Result in Table 1, Entry 5)
The General Procedure was followed using phenylcarbene com-
plex 7b (0.563 g, 1.805 mmol) and dienediyne 6b (0.277 g,
1.504 mmol) and using benzene (30 mL) as solvent. The reaction
was conducted at the reflux temperature of benzene. Final purifica-
tion using flash column chromatography (silica gel/20:1 hex-
anes:ethyl acetate) yielded compound 15c (0.091 g, 21% yield).
Compound 15c: 1H NMR (CDCl3): d8.28 (d, 1H, J = 8.0 Hz), 7.96
(d, 2H, J = 8.0 Hz), 7.39 (m, 4H), 7.07 (s, 1H), 3.66 (t, 2H,
J = 8.0 Hz), 3.18 (t, 2H, J = 8.0 Hz), 2.91 (s, 3 H), 2.27 (quintet, 2H,
J = 8.0 Hz); 13C NMR (CDCl3): d157.0, 155.4, 137.2, 136.6, 135.9,
134.9, 133.1, 131.1, 129.8, 128.9, 128.2, 124.7, 119.3, 101.1, 36.4,
31.6, 25.0, 24.5; UV: kMAX = 242, 262, 284, 294, 314, 328, 348,
364 (nm) Mass Spec (El): 298 (M, 23), 105 (100), 91(37), 76 (46),
51(13); HRMS calc. for C22H18O 298.1358. Found 298.1364.
4.3. Coupling of methylcarbene complex 7a with ethylidene enediyne
6b in dioxane (Result in Table 1, Entry 2)
The general procedure was followed using methylcarbene com-
plex 7a (0.458 g, 1.833 mmol) and dienediyne 6b (0.270 g,
1.466 mmol) and using dioxane (30 mL) as solvent. Final purifica-
tion using flash column chromatography (silica gel/9:1 hexanes:
ethyl acetate) yielded a single fraction (0.185 g, 53% yield) identi-
fied as a 3:1 mixture of benzofuran 16b (40% yield) and naphthofu-
ran 15b (13% yield). Further purification by preparative TLC and
cutting the slowest one third of the band yielded pure 16b. Napht-
hofuran 15b could not be obtained free of benzofuran 16b. Benzo-
furan 16b: 1H NMR (CDCl3): d6.26 (s, lH), 5.97 (m, 2H), 3.60–3.35
(m, 3H), 2.99 (t, 2H, J = 8.0 Hz), 2.91 (t, 2H, J = 8.0 Hz), 2.43 (s,
3H), 2.12 (quintet, 2H, J = 8.0 Hz), 1.22 (d, 3H, J = 6.0 Hz); 13C
NMR (CDCl3): d154.6, 150.1, 135.5, 132.9, 131.9, 131.2, 122.6,
115.0, 101.2, 32.3, 31.3, 31.2, 25.5, 23.8, 22.4, 14.2; IR (neat):
3025, 2943, 2841, 1594, 1427 cmꢀ1; MS (EI): 238 (M, 50), 223
(100), 211 (18), 195 (21); HRMS: Calc. for C17H18O 238.1358. Found
238.1370. Naphthofuran 15b (mixture with 16b – peaks for 16b
omitted): 1H NMR (CDCl3): d8.12 (d, 1H, J = 8.0 Hz), 7.32 (t, 1H,
J = 8.0 Hz), 7.19 (d, 1H, J = 8.0 Hz), 6.40 (q, 1H, J = 1.0 Hz), 3.09 (t,
2H, J = 7.7 Hz), 2.96 (t, 2H, J = 7.7 Hz), 2.88 (s, 3H), 2.55 (d, 3H,
4.7. Coupling of carbene complex 7a with dienediyne methyl ester 6c
(Result in Table 1, Entry 6)
The General Procedure was followed using methylcarbene com-
plex 7a (0.307 g, 1.227 mmol) and dienediyne 6c (0.214 g,
0.982 mmol) and using dioxane (22.5 mL) as solvent. Final purifica-
tion using flash column chromatography (silica gel/4:1 hex-
anes:ethyl acetate) yielded compound 14d (0.133 g, 48% yield). 1
H
NMR (CDCl3): d6.27 (s, 1H), 3.70 (s, 3H), 3.70 (m, 1H, line-
width = 19.1 Hz), 3.20–2.70 (m, 8H), 2.43 (s, 3H), 2.40 (dt, 2H,
J = 13.0, 5.0 Hz), 2.18 (quintet, 2H, J = 8.0 Hz); 13C NMR (CDCl3):

Y. Zhang et al. / Journal of Organometallic Chemistry 693 (2008) 3337–3345
3345
[7] For a recurring annual series on the reactivity of carbene complexes, see; J.W.
Herndon, Coord. Chem. Rev. 251 (2007) 1158–1258. earlier editions of the
series.
[8] For a preliminary account of this work, see; J.W. Herndon, H. Wang, J. Org.
Chem. 63 (1998) 4562–4563.
[9] A. Rahm, W.D. Wulff, J. Am. Chem. Soc. 118 (1996) 1807–1808.
[10] K.K. Wang, B. Liu, Y. Lu, Tetrahedron Lett. 36 (1995) 3785–3788.
[11] The incorrect assignment was based on accidental isochrony attributed to the
two alkene protons of 13a, coupled with a preponderance of evidence that
suggests radical cyclizations favor 5-exo processes over 6-endo processes. See:
Curran, D.P. Synthesis 1988, 417–439 and Synthesis 1988, 489–513.
[12] For oxidative dehydrogenation by chromium(0) species, see; S.U. Tumer, J.W.
Herndon, L.A. McMullen, J. Am. Chem. Soc. 114 (1992) 8394–8404.
[13] M.E. Bos, W.D. Wulff, R.A. Miller, S. Chamberlin, T.A. Brandvold, J. Am. Chem.
Soc. 113 (1991) 9293–9319.
[14] (a) For recent examples M.A. Grundl, D. Trauner, Org. Lett. 8 (2006) 23–25;
(b) M.C. Elliott, N.N.E. El Sayed, Tetrahedron Lett. 46 (2005) 2957–2959;
(c) N. Kurono, E. Honda, F. Komatsu, K. Orito, M. Tokuda, Tetrahedron 60
(2004) 1791–1801;
H. Ishibashi, T. Kobayashi, D. Takamasu, Synlett (1999) 1286–1288.
[15] E.P. Kündig, H. Ratni, B. Crousse, G. Bernardinelli, J. Org. Chem. 66 (2001)
1852–1860.
[16] Y. Zhang, J.W. Herndon, Tetrahedron 56 (2000) 2175–2184.
[17] J.W. Grissom, D. Klingberg, D. Huang, B.J. Slattery, J. Org. Chem. 62 (1997) 603–
626.
[18] A.N. Abeywickrema, A.L.J. Beckwith, S. Gerba, J. Org. Chem. 52 (1987) 4072–
4078.
d173.4, 154.9, 152.2, 137.3, 134.1, 133.3, 124.6, 122.8, 101.3, 51.5,
41.1, 38.6, 31.7, 30.7, 30.2, 26.9, 25.7, 14.0; IR (neat): 2948 (s),
2847 (m), 1739 (s), 1435 (m), 1213 (s), 1158 (s) cmꢀ1; MS (EI): 284
(M, 30), 211 (100); HRMS: Calc. for C18H20O3 284.1412. Found
284.1414.
Acknowledgements
This research was supported by the Petroleum Research Fund,
Administered by the American Chemical Society (JWH), the SCORE
program of NIH (S06 GM008136) (JWH), and National Science
Foundation CAREER award CHE-0348956 (HW). We are grateful
to a reviewer for insightful comments.
Appendix A. Supplementary material
General experimental and synthetic procedures for the prepara-
tion of diene–diynes 6a–c and copies of 1H NMR and 13C NMR
spectra for compounds 6a–c, 14d, 15a, 15c, 16a–c. Atomic coordi-
nates, energies, and spin densities for all of the energy-minimized
structures in Fig. 1 plus relevant transition states. Supplementary
data associated with this article can be found, in the online version,
at doi:10.1016/j.jorganchem.2008.08.003.
[19] For a review of arene–Cr(CO)3 complexes see; M. Uemura, Org. React. 67
(2006) 217–657.
[20] C.A. Merlic, Bruce N. Hietbrink, K.N. Houk, J. Org. Chem. 66 (2001) 6738–
6744.
[21] Frisch, M.J. et al, GAUSSIAN 03, revision C.02; Gaussian, Inc.: Wallingford, CT,
2004.
References
[22] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[1] For a recent review, see; U. Galm, M.H. Hager, S.G. Van Lanen, J. Ju, J.S. Thorson,
B. Shen, Chem. Rev. 105 (2005) 739–758.
[2] (a) Recent Examples; M. Zora, M. Kokturk, T. Eralp, Tetrahedron 62 (2006)
10344–10351;
[23] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[24] (a) M. Dolg, U. Wedig, H. Stoll, H. Preuss, J. Chem. Phys. 86 (1987) 866;
(b) D. Andrae, U. Haussermann, M. Dolg, H. Stoll, H. Preuss, Theor. Chim. Acta
77 (1990) 123.
(b) J.E. Ezcurra, K. Karabelas, H.W. Moore, Tetrahedron 61 (2005) 275–286;
(c) P. Wipf, C.R. Hopkins, J. Org. Chem. 64 (1999) 6881–6887;
(d) Y. Xiong, H. Xia, H.W. Moore, J. Org. Chem. 60 (1995) 6460–6467;
(e) L.M. Gayo, M.P. Winters, H.W. Moore, J. Org. Chem. 57 (1992) 6896–6899.
[3] A. Padwa, U. Chiacchio, D.J. Fairfax, J.M. Kassir, A. Litrico, M.A. Semones, S.L. Xu,
J. Org. Chem. 58 (1993) 6429–6437.
[25] The value for the C–O bond in the simple ketone benzophenone is 1.23 Å; E.B.
Fleischer, N. Sung, S. Hawkinson, J. Phys. Chem. 72 (1968) 4311–4312.
[26] (a) V.W. Manner, T.F. Markle, J.H. Freudenthal, J.P. Roth, J.M. Mayer, Chem.
Commun. (2008) 256–258;
(b) D.E. Williams, Mol. Phys. 16 (1969) 145–151.
[27] (a) L. Weber, D. Wewers, R. Broese, Chem. Ber. 118 (1985) 3570–3578;
(b) P.G. Cozzi, P. Veya, C. Floriani, F.P. Rotzinger, A. Chiesi-Villa, C. Rizzoli,
Organometallics 14 (1995) 4092–4100;
[4] N.V. Nguyen, K. Chow, J.O. Karlsson, R.J. Doedens, H.W. Moore, J. Org. Chem. 51
(1986) 419–420.
[5] A. Tarli, K.K. Wang, J. Org. Chem. 62 (1997) 8841–8847.
[6] J.W. Grissom, G.U. Gunawardena, D. Klingberg, D. Huang, Tetrahedron 52
(1996) 6453–6518.
(c) H.G. Raubenheimer, M.W. Esterhuysen, G. Frenking, A.Y. Timoshkin, C.
Esterhuysen, U.E.I. Horvath, Dalton Trans. (2006) 4580–4589.
[28] For a General Experimental, see the Supplementary Material.