Di(9-anthryl)carbene revisited. Product analysis and spectroscopic studies

Article abstract of DOI:10.1246/bcsj.77.1509

Irradiation (λ > 300 nm) of di(9-anthryl)diazomethane (1) in degassed benzene gave the trimer (3) of dianthrylcarbene (2) as the main product along with a small amount of a tetramer (4). The structures of those oligomers were characterized as caged compounds produced as a result of three- and four-fold coupling at the 10 and 10′ positions of 2. Photolysis of 1 in the presence of oxygen with short wavelength UV light resulted in the formation of di(9-anthryl) ketone (5) and anthraquinone (6), while similar irradiation with a longer wavelength light gave 5 and 9-anthryl 9-anthracenecarboxylate (9), which was found to give 6 upon irradiation with a short wavelength light in the presence of oxygen. On the other hand, the photolysis of 4 gave 3 as the main product, while 3 was almost completely photostable under similar conditions. However, irradiation of both 3 and 4 in the presence of oxygen gave 6. In order to elucidate the reactions observed, we carried out spectroscopic studies (matrix EPR and UV/vis spectroscopies as well as laser flash photolysis) with 1, 3, and 4. These studies revealed that 3 generated a trimer diradical (10) formed as a result of C-C bond cleavage upon irradiation and that 10 reformed 3 almost exclusively, while similar photolysis of 1 and 4 generated the same two transient species, 10 and a dimer diradical. It is proposed that triplet carbene 2 equilibrates with the dimer diradical 2-2 and that the main decay pathway of this mixture is to generate the trimer diradical 10 while dimer diradical 2-2 undergoes coupling to produce the tetramer 4. The structure of the triplet carbene in equilibrium with diradicals is also discussed.

Full text of DOI:10.1246/bcsj.77.1509

ꢀ 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 1509–1522 (2004) 1509
Di(9-anthryl)carbene Revisited. Product Analysis
and Spectroscopic Studies
Ken-ichi Yoshida, Eiji Iiba, Yoshinori Nozaki, Katsuyuki Hirai, Yasutake Takahashi,
y
ꢀ
Hideo Tomioka, Cheng-Tung Lin,1; and Peter P. Gaspar1
Chemistry Department for Materials, Faculty of Engineering, Mie University, Tsu, Mie 514-8507
1Department of Chemistry, Washington University, St. Louis, MO 63130-4899, USA
Received January 21, 2004; E-mail: tomioka@chem.mie-u.ac.jp
Irradiation (ꢀ > 300 nm) of di(9-anthryl)diazomethane (1) in degassed benzene gave the trimer (3) of dianthrylcar-
bene (2) as the main product along with a small amount of a tetramer (4). The structures of those oligomers were char-
acterized as caged compounds produced as a result of three- and four-fold coupling at the 10 and 100 positions of 2. Pho-
tolysis of 1 in the presence of oxygen with short wavelength UV light resulted in the formation of di(9-anthryl) ketone (5)
and anthraquinone (6), while similar irradiation with a longer wavelength light gave 5 and 9-anthryl 9-anthracenecarbox-
ylate (9), which was found to give 6 upon irradiation with a short wavelength light in the presence of oxygen. On the other
hand, the photolysis of 4 gave 3 as the main product, while 3 was almost completely photostable under similar conditions.
However, irradiation of both 3 and 4 in the presence of oxygen gave 6. In order to elucidate the reactions observed, we
carried out spectroscopic studies (matrix EPR and UV/vis spectroscopies as well as laser flash photolysis) with 1, 3, and 4.
These studies revealed that 3 generated a trimer diradical (10) formed as a result of C–C bond cleavage upon irradiation
and that 10 reformed 3 almost exclusively, while similar photolysis of 1 and 4 generated the same two transient species, 10
and a dimer diradical. It is proposed that triplet carbene 2 equilibrates with the dimer diradical 2-2 and that the main decay
pathway of this mixture is to generate the trimer diradical 10 while dimer diradical 2-2 undergoes coupling to produce the
tetramer 4. The structure of the triplet carbene in equilibrium with diradicals is also discussed.
Triplet carbenes are usually characterized by electron para-
magnetic resonance (EPR) spectroscopy.1,2 The EPR signals
can be analyzed in terms of zero-field splitting (ZFS) parame-
ters, D and E values, which provide information on the molecu-
lar and electronic structure of triplet carbenes. The D value is
related to the average separation between the unpaired elec-
trons, and the E value is a measure of the difference of the mag-
nitude of the dipolar interaction along the x and y axes. More
plainly, the more the two electrons are delocalized in conjugat-
ed ꢁ systems, the smaller the D value will be. On the other
hand, increasing the bond angle at the carbene center leads to
a smaller value for E.
Among the many triplet arylcarbenes, triplet di(9-anthryl)-
carbene (2)3–5 is unique in that it shows the smallest D and E
values ever reported for a triplet diarylcarbene. This indicates
that it has an almost linear geometry at the carbene center, with
extensive delocalization of the unpaired electrons onto the at-
tached anthryl groups, whose planes are perpendicular to each
other. The extensive delocalization is expected to stabilize this
carbene thermodynamically, while the perpendicular geometry
of the anthryl groups stabilizes the carbene center kinetically
through steric shielding by the four peri-hydrogens. That is,
both the electronic structure and sterics of the molecule seem
ideal for the formation of a stable triplet carbene.
have been published concerning this carbene, and extensive
characterization of this species and its chemistry still is not
available.
Triplet di(9-anthryl)carbene (32) was first generated in 1971
by photolysis of the corresponding diazo precursor (1) and
characterized by EPR spectroscopy in a dianthryl ketone ma-
trix, which revealed relatively small values for both E and D,
i.e., 0.0011 and 0.113 cmꢁ1, respectively, at 4 K (Scheme 1).3
Annealing the matrix allows the carbene to adopt its minimum
energy structure, which is almost completely linear, i.e., the
value of E drops to zero (Scheme 1). It has been reported that
the half-life of 32 in the matrix was 7 min at 200 ꢂC; however,
it was also concluded that the stability was due to the rigidity of
the environment and not to an intrinsic lack of reactivity. A sol-
3
id solution of 2 in viscous glass matrices such as 2-methyl-
tetrahydrofuran was stable at 77 K, but, upon warming with a
consequent thawing of the matrix, 32 disappeared irreversibly.3
The reactivity of 32 in solution at ambient temperatures was
then studied in 1988 in more detail, and it was found that 32 un-
derwent a bimolecular self-reaction at the diffusion-controlled
limit. Singlet carbene 2 was not quenched by a typical trapping
reagent for a singlet carbene, i.e., methanol, but the triplet was
quenched by oxygen with a rate constant of 1 ꢃ 106 Mꢁ1 sꢁ1
,
nearly 5 orders of magnitude smaller than the rate constants
for many other triplet carbenes. Product studies showed that
32 reacted with O2 to form dianthryl ketone and anthraquinone
but formed the carbene dimer in the absence of O2, thus sup-
porting the interpretation of the reaction kinetics.4 As part of
In spite of these highly interesting findings, only two papers
y Current address: Department of Chemistry, Box 864, Tunghai
University, Taichung 400, Taiwan
Published on the web August 6, 2004; DOI 10.1246/bcsj.77.1509

1510 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
N2
hν
.
.
.
.
.
- N2
2B
2A
1
Scheme 1.
hν
.
.
+
1
- N2
3
4
4
3
Scheme 2.
our continued attempts to isolate a triplet carbene,6 the chemis-
try of dianthrylcarbene was investigated by product analysis
and spectroscopy.
very similar to those observed for the trimer 3, but the mass
spectrum showed a peak at m=z 1466.2 (10.5%), corresponding
to the M þ 1 ion calculated for the tetramer, along with 1098.5
(12.2%) and 732.2 (63.2%) corresponding to the fragments
formed as a result of the loss of one and two carbene units, re-
spectively. The structure of the tetramer which fits those data
best is that of (4), formed as a result of four-fold coupling at
the 10 and 100 positions of 2 (Scheme 2).
The formation of the oligomers (3 and 4) is not quenched
even in the presence of typical carbene trapping agents.8 Thus,
similar irradiations in the presence of methanol, 1,4-cyclohex-
adiene, 2-methyltetrahydrofuran, and phenylsilane again pro-
duced the oligomers in similar corresponding yields. No evi-
dence was found for any reaction with these trapping agents
at either the divalent carbon atom or the 10 and 100 positions.
Neither the methyl ether, which is expected to be formed as a
result of insertion of the carbene into an O–H bond of the alco-
hol,7 nor a reduction product such as di(9-anthryl)methane or
9,90-di(9-anthrylidene)allene, which might have been formed
as a result of H atom abstraction at the divalent carbon or at
the 10,100-positions, respectively, was detected. However, di-
anthrylcarbene reacted with thiophenol in a dark at room tem-
perature to give di(9-anthryl)methyl phenyl sulfide.
The ratio of the oligomers appears to be sensitive to the re-
action conditions. When a frozen glass containing 2:5 ꢃ 10ꢁ3
M 1 dissolved in 2-methyltetrahydrofuran was irradiated at
77 K until no 1 remained and then thawed, 3 was still the major
product, but chromatographic and spectroscopic analysis indi-
cated the presence of an increased amount of a higher oligomer
of 2.
However, when the irradiation was carried out in a non-de-
gassed benzene solution, two oxidation products, i.e., dianthryl
ketone (5) and anthraquinone (6), were formed at the almost
complete expense of oligomer formation (Scheme 3). Although
the formation of the ketones in the photolysis of diaryldiazo-
methanes in the presence of oxygen via the intermediacy of
the related carbonyl oxides (e.g., 7) followed by the self-reac-
tion has been demonstrated,9 the formation of anthraquinone
(6) cannot be explained easily. The possibility that 6 is pro-
Results
Product Analysis Studies. Photolysis of Dianthryldiazo-
methane: Irradiation (ꢀ > 300 nm) of di(9-anthryl)diazo-
methane (1, 5 ꢃ 10ꢁ3 M) in degassed benzene caused rapid
fading of the orange diazo color with copious evolution of ni-
trogen gas to give a pale-yellow solution. Gel permeation chro-
matography of the photomixture resulted in the isolation of
white crystals (mp 332–333 ꢂC) in 55% yield. The spectroscop-
ic data suggested that the trimer of dianthrylcarbene (2), formed
as a result of 10–100 coupling, is the most likely structure. The
mass spectrum (FAB MS) showed a peak at m=z 1099.1
(19.1%), which corresponds to the M þ 1 ion calculated for
three di(9-anthryl)carbene units. Additional peaks were ob-
served at m=z 732.1 (19.6%) and 366 (100%), corresponding
to the fragments formed as a result of the successive loss of a
carbene unit from the trimer. Thus, the data clearly suggest that
the main product is the trimer of di(9-anthryl)carbene rather
than a dimer, as suspected earlier.4 The structure of the trimer
which explains the NMR data best is the one (3) formed as a
result of three-fold coupling at the 10 and 100 positions of
di(9-anthryl)carbenes (Scheme 2). 1H NMR (500 MHz) showed
the presence of eight aromatic proton signals ranging from 5.91
to 7.79 ppm, plus one signal of a proton at a benzylic carbon
atom at 4.82 ppm. The 13C NMR (126 MHz) displayed a total
of 12 sp2 signals between 136.42 and 125.46 ppm. In addition,
two signals were observed at 208.34 and 111.10 ppm, most
probably assignable to the central and terminal carbons on an
allenic subunit, respectively, and to one sp3 carbon at 55.06
ppm.
The X-ray crystallographic analysis of 3 revealed that the
compound has a complicated structure and a flattened hexago-
nal bipyramidal shape.5c
A minor product was isolated from a higher-molecular-
weight fraction as white crystals (mp 222–223 ꢂC) in 12%
yield. The 1H and 13C NMR spectra of the minor product were

K. Yoshida et al.
Bull. Chem. Soc. Jpn., 77, No. 8 (2004) 1511
3
O
O
hν
TEMPO
hν /∆
hν
+
O
O
O2
(>300 nm)
hν
O
O2
hν
hν
O
O2
TEMPO
1
5
6
5
4
6
hν / O2
(>520 nm)
Scheme 4.
hν
O2
hν
TEMPO
O
(>300nm)
hν
O
O2
5
9
-
O
+
O
O
O
7
8
Scheme 3.
duced by the photo-oxidation of 5 is precluded by the finding
that 5 is almost completely stable under the photolysis condi-
tions. It is unlikely, however, that the quinone is formed by a
completely different pathway from that producing the ketone.
It has been shown that carbonyl oxides undergo rearrangement,
either thermally or photochemically, to give dioxiranes (e.g., 8)
and then esters (e.g., 9) in competition with the oxygen transfer
to form ketones.9 Thus, the ester, i.e., 9-anthryl 9-anthracene-
carboxylate (9), is proposed to be the most likely precursor
for quinone 6. Indeed irradiation of ester (9) under essentially
the same conditions as employed for the photolysis of 1 gave
the quinone (6) almost exclusively. Moreover, the ester (9)
was detected when the irradiation of 1 was carried out with
longer wavelength light (ꢀ > 520 nm) and decomposed to give
6 upon further irradiation with shorter wavelength light
(ꢀ > 300 nm) (Scheme 3).
Fig. 1. ESR spectra of di(9-anthryl)carbene (32) in 2-meth-
yltetrahydrofuran (2-MTHF) matrix (a) at 77 K and (b) at
110 K.
anthraquinone (6) as the main product. More interestingly, it
reacted with TEMPO under irradiation (ꢀ > 300 nm) to form
dianthryl ketone (5) exclusively (Scheme 4).
The tetramer 4 showed completely different reactivity upon
heating and irradiation. Thus, irradiation of 4 in degassed C6D6
produced the trimer 3 almost exclusively. The tetramer 4 was
also thermally labile; it started to decompose already at 50
ꢂC in degassed C6D6, then rapidly and completely decomposed
at 100 ꢂC to form the trimer 3 as the only identifiable product
among a large amount of tarry compounds. However, the tetra-
mer exhibited essentially the same reactivity as that of the trim-
er toward O2 and TEMPO; it gave, upon irradiation in the pres-
ence of TEMPO or O2, 5 or 6, respectively (Scheme 4).
Spectroscopic Studies in Frozen Media at Low Temper-
ature. In order to gain more direct information about the struc-
ture and reactivity of 2, the following spectroscopic studies
were carried out.
Another reagent which can trap 2 with moderate efficiency
was a persistent nitroxide radical. Thus, irradiation of 1 in
degassed benzene containing 2,2,6,6-tetramethylpiperidine-1-
oxyl (TEMPO) gave dianthryl ketone (5) and tetramethyl-
piperidine in equimolar amounts (Scheme 3).10 It has been
pointed out that the reaction of triplet diphenylcarbene with a
persistent nitroxide is spin-allowed and over 100 kcal/mol exo-
thermic. Indeed, the reaction proceeds smoothly to give benzo-
phenone and piperidine. The absolute rate constant was deter-
mined to be 2:7 ꢃ 108 Mꢁ1 sꢁ1 in acetonitrile, which is well be-
low the diffusion limit and much slower than the corresponding
Matrix Photolysis of 1: Irradiation of 1 (2:5 ꢃ 10ꢁ3 M) in
2-methyltetrahydrofuran (2-MTHF) at 77 K gave EPR signals
with typical fine structure patterns for unoriented triplet species
(Fig. 1a). The spectrum showing unresolved x and y lines and a
rather intense line at 152 mT corresponding to ꢀms ¼ 2 is very
similar to that recorded for triplet dianthrylcarbene in a dian-
reaction with molecular oxygen (kO ¼ 5 ꢃ 109 Mꢁ1 sꢁ1).9,11
2
Photolysis of a Trimer and a Tetramer: In spite of its
sterically highly congested structure, the trimer 3 exhibited sur-
prising stability. It remꢂained unchanged both upon heating in
degassed C6D6 at 140 C for 2 h and upon irradiation in de-
gassed benzene with light (ꢀ > 300 nm) for 2 h. However, it
decomposed in air either thermally or photochemically to give
thryl ketone matrix3 and, hence, can be safely assigned to 2.
3
Very small D and E values (D ¼ 0:1038 cmꢁ1 and E ¼
0:0000 cmꢁ1) calculated from the spectra are also in good
agreement with those reported for 32. The EPR signals not only

1512 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
were stable at this temperature for hours but also survived
warming up to 130 K. However, when the samples were grad-
ually warmed, the signals became sharp at around 90 K, a small
shift in the z lines was observed, and a small but distinct
decrease in the D value is noted (D ¼ 0:089 cmꢁ1 and E ¼
0:0000 cmꢁ1) (Fig. 1b).
A decrease in the D and E values upon annealing the matrix
has often been observed for sterically congested triplet diaryl-
carbenes and is usually interpreted in terms of geometrical
changes.2c,12 Thus, when a carbene is generated in rigid matri-
ces at very low temperature, it should have the geometry dictat-
ed by that of a precursor. Even if the thermodynamically most
stable geometry of the carbene is different from that at birth, the
rigidity of the matrices prevents the carbene from achieving its
minimum-energy geometry. In the present case, this state is
3
3
now termed 2(nascent) or 2n. When matrices are softened
3
Fig. 2. UV/vis spectra obtained by irradiation of 1 in
MTHF. (a) Spectra of 1 in 2-MTHF at 77 K; (b) same sam-
ple after 5 min irradiation (ꢀ > 350 nm); (c) same sample
after warming to 85 K; (d) same sample after warming to
115 K; (e) same sample after warming the matrix to room
temperature and refreezing to 77 K.
by annealing, the carbene undergoes (e.g., 2n) relaxation to
the preferred geometry probably to gain relief from steric com-
pression. It has been shown that the ZFS parameters of sterical-
ly unperturbed diphenylcarbenes show little sensitivity toward
the temperature of the matrix.13 The rather small D values cou-
pled with the essentially zero E value observed for 32 indicate
that there is extensive delocalization of the unpaired electrons
into the anthryl portions and that the carbene has a substantial
degree of allenic character. A small but distinct decrease in the
D value upon annealing the matrix can thus be interpreted as
indicating that the triplet carbene relaxed to a nearly linear ge-
ometry, in which the unpaired electrons are further delocalized
shifted very little with a slight increase in the intensity
(Fig. 2c). At 115 K, the band at 448 nm shifted further to a
shorter wavelength, 433 nm, with a significant decrease in in-
tensity, while the band at 359 nm again shifted very little with
a decrease in intensity (Fig. 2d). Upon further thawing, all of
those bands disappeared to form a featureless band with an ap-
parent maximum at 294 nm (Fig. 2e), which is essentially the
same as the spectrum of the final product, e.g., 3. These obser-
vations suggest that at least three transient species including 32
3
into anthryl portions. This state is now termed 2(relaxed) or
32r. Although the E values must accordingly have changed in
this case, their small magnitudes did not allow us to observe
these changes.
3
were involved, i.e., nascent triplet carbene 2n (359, 450 nm),
Another interesting feature of the EPR spectra is the fact that
peaks at around 333.7 mT became stronger at the expense of the
signals attributable to 2r as the matrix temperature increased
species A (359, 448 nm), and B (361, 433 nm). These changes
were not reversed even when the samples were recooled.
The EPR studies clearly showed that the signals changed
from the nascent, less linear geometries to the relaxed, almost
3
(Fig. 1b). Usually, these peaks are observed in the photolysis
of diazo compounds in organic matrices at low temperatures
and are assigned to ‘‘radical impurities’’ formed under the
photochemical conditions.2 In the present case, however, the
new signals increased at the expense of the signals due to
32r; hence, they are ascribable to a doublet species produced
3
linear ones 2r as the temperature was increased from 77 to
3
110 K in 2-MTHF. Therefore, the initial changes (from 2n
to A) are interpreted in terms of the geometrical changes lead-
ing to relaxed triplet carbene 32r, associated with viscosity
changes rather than intermolecular reactions, e.g., hydrogen ab-
straction, with matrices. These spectral changes associated with
geometrical changes have often been observed in similar spec-
troscopic studies of sterically congested diarylcarbenes.15 The
successive changes (from A ¼ 32r to B) are explained in terms
3
as a result of a thermal reaction of 2r.
When the irradiation of 1 (1:2 ꢃ 10ꢁ4 M) in a 2-MTHF glass
at 77 K was monitored by UV/vis spectroscopy, a new absorp-
tion band at 359 nm and a rather broad band at 450 nm appeared
at the expense of the original absorption due to 1 (Fig. 2b).14
The glassy solution did not exhibit any spectral change for sev-
eral hours when kept at this temperature. However, when it was
allowed to warm to room temperature and then cooled to 77 K,
the characteristic bands disappeared (Fig. 2e). On the basis of
these observations coupled with EPR data (vide supra), the
absorption spectrum can be attributed to the nascent triplet
dianthrylcarbene (32n).
3
of the chemical reactions of 2r rather than the geometrical
changes since the EPR signals due to a doublet species domi-
nated under these conditions at almost the complete expense
3
of the original signals due to 2r.
Support for the assignment was obtained when the spectral
changes were monitored using other organic glasses of different
viscosity. For instance, in a viscous matrix, i.e., triacetin matri-
ces (ꢂ=ꢂ3-MP ¼ 1:2 ꢃ 1045),16 the initial, second, and final
spectral changes were observed at 190, 220, and 230 K, respec-
tively (Fig. 3). In marked contrast, when soft matrices were
used, the initial photoproduct at 77 K was very similar to the
secondary product formed upon annealing the initial product
in the hard matrix, i.e., 2-MTHF (ꢂ=ꢂ3-MP ¼ 6:9 ꢃ 105).16
Thus, irradiation of 1 in 3-methylpentane–isopentane (3-MP–
When the spectral changes of the matrix were more carefully
monitored as a function of temperature, an interesting feature
appeared. For example, when the matrix was warmed gradually
to 85 K, a broad absorption band at 450 nm changed rather dra-
matically to a sharp and strong band with its maximum shifting
to a shorter wavelength at 448 nm, while the band at 359 nm

K. Yoshida et al.
Bull. Chem. Soc. Jpn., 77, No. 8 (2004) 1513
16
IP, 1:1, ꢂ=ꢂ3-MP ¼ 6:3 ꢃ 10ꢁ4
)
matrices at 77 K resulted in
signals were observed (Fig. 5a). When the irradiation was
monitored by UV/vis spectroscopy, new absorption bands with
maxima at 361 and 433 nm were detected (Fig. 5b). The EPR
signals and the absorption disappeared upon thawing the ma-
trix, but reappeared upon irradiation at 77 K. Those spectral
changes could be repeated several times. Moreover, analysis
of the spent solution indicated that no new product was detect-
ed. These observations suggest that the transient species is like-
ly to be diradical (10) produced as a result of cleavage of a C–C
singlet bond connecting the dihydroanthracene unit in 3
(Scheme 5). Essentially the same spectroscopic changes were
observed in the photolysis of the tetramer 4 under identical con-
ditions. However, in this case, the formation of a trimer was
confirmed in the spent solution (Schemes 4 and 5). Taking into
account the fact that 3 and 4 are composed of the same unit, i.e.,
two dihydroanthracene units connected by an allene bond, it is
the appearance of two sharp bands at 357 and 445 nm without
showing a broad band centered at around 450 nm. The spectra
started to change already at 85 K to form the 432 nm species
(Fig. 4).
Matrix Photolysis of Oligomers 3 and 4: In order to gain
more insights into the nature of the other transient species de-
tected in the photolysis of 1 in a matrix at low temperature,
we also irradiated the final stable products, i.e., 3 and 4, and
studied them spectroscopically under identical conditions. Irra-
diation of 3 in a 2-MTHF matrix at 77 K gave EPR signals cen-
tered at 333 mT ascribable to a doublet species, but no triplet
Fig. 3. UV/vis spectra obtained by irradiation of 1 in triace-
tin matrices. (a) Spectra of 1 in triacetine at 77 K; (b) same
sample after 5 min irradiation; (c) same sample after thaw-
ing at 190 K; (d) same sample after thawing at 220 K; (e)
same sample after thawing at 230 K.
Fig. 4. UV/vis spectra obtained by irradiation of 1 in 3-
methylpentane–isopentane matrices. (a) Spectra of 1 in 3-
methylpentane–isopentane at 77 K; (b) same sample after
5 min irradiation; (c) same sample after thawing at 85 K.
ESR
(A)
0.11
0.20
0.30
0.40
0.50
0.56
Magnetic Field / T
0.4
361 nm
(B)
UV / vis
433 nm
0.3
0.2
0.1
0.0
300
400
500
600
Wavelength / nm
Fig. 5. (a) ESR spectra obtained by irradiation of 3 in 2-MTHF at 77 K. (b) UV/vis spectra obtained by irradiation of 3 in 2-MTHF
at 77 K.

1514 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
.
hν
.
.
.
.
.
.
.
3
O2
10
6
O2
.
hν
.
4
3
2
11
Scheme 5.
(a)
250 300
400
500
600 (nm)
0.09
0.20
0.40
0.59 (T)
(b)
(c)
250 300
400
500
600 (nm)
0.09
0.20
0.40
0.59 (T)
250 300
400
500
600 (nm)
0.09
0.20
0.40
0.59 (T)
UV / vis
ESR
Fig. 6. UV/vis and ESR spectra obtained by irradiation of 1 in (a) 2-methyltetrahydrofuran, (b) 3-methylpentane, and (c) 3-methyl-
pentane–isopentane matrices at 77 K.
reasonable to expect that the diradicals (10, 11) formed as a re-
sult of C–C bond cleavage should have a very similar spectro-
scopic feature.
to the ratio of the 457-nm broad bands to the 448-nm sharp
bands in the UV/vis spectra. Thus, in triacetin and 2-MTHF
matrices, where the doublet radical signals were essentially
not observed at 77 K, only broad 457-nm bands appeared at
77 K; in the 3-MP–IP (1:1) matrix, on the contrary, where
the sharp 445-nm band formed initially even at 77 K, no car-
bene signals were observed in EPR. The 3-MP matrix probably
lies between the two extremes; the 457 and 448-nm bands were
apparent in the UV/vis spectra as were the radical and carbene
signals, which were observed in EPR (Fig. 6). Detailed analysis
of the thermal changes of triplet carbene 2 using EPR spectros-
A comparison of the spectra observed in the photolysis of 3
and 4 with those obtained in the irradiation of 1 suggests that
the spectroscopic features of the diradicals formed upon photol-
ysis of 3 and 4 are very similar to those of the species B pro-
3
duced from the relaxed triplet carbene 2r upon annealing the
matrix. Careful comparison of the results from both techniques,
UV/vis and EPR, suggests that the ratio of the triplet carbene
signals to the ‘‘radical’’ doublet signals roughly corresponds

K. Yoshida et al.
Bull. Chem. Soc. Jpn., 77, No. 8 (2004) 1515
copy also supports this assignment. Thus, when the triacetine
matrix containing triplet 2 was gradually thawed, the character-
istic triplet signals started to decrease with a concomitant in-
crease in the intensity of the diradical signals at around 200
K, the same temperature at which the broad 455-nm bands start-
ed to decay to form the sharp 445-nm bands in the UV/vis spec-
trum. These observations suggest that the broad 457-nm bands
that initially formed at 77 K in hard matrices are assignable
to the triplet carbene 2, while the sharp 445-nm bands formed
either in the soft matrices at 77 K or upon thawing of the hard
matrices are obviously not attributable to the geometrical iso-
mers of the triplet carbene.
Laser Flash Photolysis (LFP) Studies in Solution at Room
Temperature. The LFP of Diazomethane 1: Laser flash
photolysis (LFP) of 1 (1:6 ꢃ 10ꢁ4 M) in a degassed benzene
solution at room temperature with a 10 ns, 10 mJ, and 308-
nm pulse from a XeCl laser produced a transient species show-
ing strong, rather sharp bands at 362 and 435 nm, which ap-
peared coincidently with the pulse and disappeared within 1
ms (Fig. 7). Due to overlap with the absorption maxima of
the diazo precursor 1, the samples are not sufficiently transpar-
ent for adequate monitoring in the 300–330-nm region. Howev-
er, the two bands decayed in a similar manner, in which two
decay modes were involved. The first 80% of the decay was
fit to a first-order decay (k ¼ 4:3 ꢃ 105 sꢁ1, ꢃ ¼ 2:3 ms), while
the residual slower decay was found to be of the second order
(2k="l ¼ 6:5 ꢃ 104 sꢁ1). This observation indicates that two
transients with considerably different reactivities but very
similar absorption bands are actually involved.
and 4. The rather poor resolution of the spectra obtained by
LFP as opposed to that in the matrix, along with the similarity
of the spectroscopic shapes of the two species, did not allow us
to make a definite assignment of the species simply on the basis
of the shape of the spectra.
The LFP of 1 in the presence of oxygen gave somewhat dif-
ferent results (Fig. 8). Thus, the LFP of a non-degassed solution
of 1 again produced essentially the same transient species as
observed in the LFP of the degassed solution of 1. However,
there are two observations that should be noted. First, the decay
rate of the transients was found to increase as a function of the
oxygen concentration, indicating that the transient species are
quenched by oxygen. The plot of the decay rate as a function
of the oxygen concentration in the range of 0.1 to 10 mM is lin-
ear. The slope of this plot yields a quenching rate constant by
oxygen, kO ¼ 4:0 ꢃ 108 Mꢁ1 sꢁ1, and the intercept yields k0 ¼
2
4:1 ꢃ 105 sꢁ1 (inset in Fig. 8). Second, the optical yields of the
species were notably decreased as the concentration of oxygen
in the solution increased. This suggests that a precursor leading
to the observed species is also trapped by oxygen. The product
studies in Scheme 3 indicated that the photolysis of 1 in the
presence of oxygen produced di(9-anthryl) ketone and 9-an-
thryl 9-anthracenecarboxylate, which are most probably de-
rived from the reaction of triplet dianthrylcarbene with oxygen,
most probably by way of the oxides 7. Therefore, it is likely that
the precursor to the observed species is triplet carbene 32. It is
well documented that diarylcarbenes with triplet ground states
are readily trapped by oxygen to give the corresponding diaryl
ketone oxides, which exhibit rather broad bands in the visible
region and, hence, are relatively easily detected by LFP.9,11
However, all attempts to detect the oxides 7 were unsuccessful.
It may be that the oxide is not long-lived enough to be ob-
served, at least to the limit of the time resolution (ꢄ50 ns) of
our LFP system or that the absorption bands are too broadened
and/or too red-shifted to be detected to the limit of detection of
our system.
The transient spectra were compared with those observed in
the photolysis of 1 in a matrix at low temperature. The transient
band observed by LFP is obviously different from the broad
band centered at 450 nm (Fig. 2b) that is ascribable to the nas-
3
cent triplet carbene 2n observed in a rigid matrix at low tem-
perature. The former band is, however, very similar to those
assigned as either a relaxed, more linear triplet carbene 32r
(Fig. 2c) or as a diradical (10 and 11) species (Fig. 3b) gener-
ated as a result of the C–C bond cleavage of the oligomers 3
The transient species were not trapped at all by 1,4-cyclo-
hexadiene, which is also known as an equally efficient trapping
1.1
435 nm
N2
k = 4.3 ×10 5 s-1
τ = 2.3 µs
308 nm
0.2 µs
1
PhH
0.8
0.5
3
5
20
362 nm
10 µs
100
2k/
ε
l = 6.5 × 104 s-1
400 µs
0.2
-0.1
300
350
400
450
500
550
Wavelength / nm
Fig. 7. Absorption spectra of the transient products obtained upon 308-nm excitation of 1 in degassed benzene recorded after 0.2, 1,
3, 5, 20, and 100 ms. Inset shows the decay of the transient products monitored at 435 nm.

1516 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
0.750
10
5
0.0 mM
•
N2
308 nm
•
•
O2 / PhH
•
0.558
0.365
0.173
-0.020
ko = 4.1 × 105 s-1
•
kO = 4.0 × 108 M-1s-1
2
γ = 0.987
0.13 mM
0
0
5
10
[ O2 ] / 10-4
M
0.43 mM
2.1 mM
2.5
-2.5
7.5
12.5
17.5
22.5
Time / µs
Fig. 8. Effect of increasing oxygen concentration on the decay of transient following pulse irradiation in benzene at 20 ꢂC, monitored
at 435 nm. Inset shows plot of decay rate as a function of oxygen concentration.
0.55
435 nm
308 nm
.
3
PhH
5
-1
k = 3.8 × 10 s
1 µs
0.40
0.25
0.10
-0.05
τ = 2.6 µs
2
3
5
362 nm
10 µs
10
330
350
400
450
500
550
Wavelength / nm
Fig. 9. Absorption spectra of the transient products obtained upon 308-nm excitation of 3 in degassed benzene recorded after 1, 2, 3,
5, and 10 ms. Inset shows the decay of the transient products monitored at 435 nm.
reagent for triplet carbene.17 Thus, the LFP of 1 in the presence
of 1,4-cyclohexadiene gave the same transient absorption
bands, which decayed in a similar manner as in degassed ben-
zene. It is also well known that diarylcarbenes generated in
good hydrogen donors, such as the diene, undergo H-abstrac-
tion leading to the corresponding diarylmethyl radicals, which
usually show transient absorption at longer wavelengths than
that of the precursor carbene and are, hence, easily detected.18
The LFP of 1 in methanol also gave the same results. Those
spectroscopic observations, along with the product analysis
studies, which show that the formation of 3 and 4 is not
quenched at all either by methanol or by the diene, suggest that
the transient species is not a triplet carbene but is likely an
intermediate leading to the final products, 3 and 4.
ramer (4) were performed. The LFP of 3 in degassed benzene
produced a transient species (Fig. 9) showing essentially the
same bands as observed in the matrix photolysis of 3 (Fig. 5).
The bands were also similar to those observed in the LFP of
1. However, all the transient bands exhibited very clean first or-
der kinetic decay, and the decay rate was essentially the same
(k1 ¼ 3:8 ꢃ 105 sꢁ1, ꢃ ¼ 2:6 ms) as that of the fast-decaying
species observed during the LFP of 1. The product analysis data
indicate that the trimer undergoes C–C bond cleavage to form a
diradical (Scheme 5), which undergoes recombination almost
exclusively in the absence of a trapping reagent but forms ox-
idation products in the presence of oxygen. This data along with
the matrix spectroscopy observations mean that the most likely
transient species is the trimer diradical (10) formed as a result
of fission of the sp3 C–C bond of the dihydroanthryl portion.
This in turn suggests that the major species observed in LFP
The LFP of Oligomers: In order to identify the origins of
the intermediate, the LFP studies of the trimer (3) and the tet-

K. Yoshida et al.
Bull. Chem. Soc. Jpn., 77, No. 8 (2004) 1517
Fig. 10. Absorption spectra of the transient products obtained upon 308-nm excitation of 4 in degassed benzene recorded after 1, 2,
10, and 100 ms. Inset shows the decay of the transient products monitored at 435 nm.
Discussion
of 1 can be assigned as 10.
The LFP of 4 gave somewhat different results. Absorption
bands obtained in the LFP of 4 were essentially identical to
those observed in the LFP of 1 and 3, but the bands decayed
similarly to those from 1 rather than to those from 3 (Fig. 10).
Thus, most of the bands decayed by first-order kinetics
(k1 ¼ 3:9 ꢃ 105 sꢁ1, ꢃ ¼ 2:6 ms), while the remaining minor
component decayed very slowly in a second-order fashion
(2k2="l ¼ 1:9 ꢃ 105 sꢁ1). This means that the LFP of 1 and 4
results in the generation of the same two species.
The fast-decaying main species observed in the LFP of both
1 and 3 can be safely assigned to a trimer diradical 10 if one
compares its spectroscopic features and decay kinetics with
those of the LFP of a trimer 3. The assignment is in agreement
with the results of a product analysis, which showed that the
main product obtained in the photolysis of both 1 and a tetramer
is a trimer. What is this slow-decaying minor species observed
in the LFP of tetramer (and also of 1), then? Since a tetramer is
formed in both cases, the most likely species is a precursor
leading to the tetramer. As this species decays by second-order
kinetics, a dimer diradical (2-2) can be assumed as a possible
intermediate (Scheme 6).
Photolysis of di(9-dianthryl)diazomethane (1) resulted in the
formation of the trimer (3) as a main product along with a small
amount of the tetramer (4), both of which are most probably
formed by way of the corresponding carbene (32). The struc-
tures of the products strongly reflect the unique nature of the
triplet carbene (32), that is, an almost linear and perpendicular
molecular structure with extensive delocalization of the un-
paired electrons onto the two anthryl rings (Scheme 1). Due
to this unusual structure, the carbene has an unexpectedly short
life time, much shorter than that of even diphenylcarbene (see
below), although it has an ideal molecular and electronic struc-
ture for a stable triplet carbene.
The unusual reactivity of this carbene is also noted in its
reaction with typical carbene traps. Thus, the triplet state 32 is
intercepted with a very efficient trapping reagent, oxygen and
TEMPO, but not by a good hydrogen donor, such as 1,4-cyclo-
hexadiene. The singlet state is not trapped even with a very
efficient quencher, i.e., methanol. Such observations are com-
pletely unprecedented in the chemistry of diarylcarbenes.1
We are not able to detect triplet carbene 32 itself by LFP in
solution at room temperature mainly because of the very simi-
lar nature of the absorption bands for the species formed, e.g.,
32 (nascent and relaxed), diradicals (dimer and trimer), and also
because of the rather rapid decay process from 32 to diradicals
species. There are subtle differences between 32n and the oth-
ers, which can only be discriminated by UV/vis in a hard ma-
trix at low temperature, in which the resolution is better than in
LFP at room temperature. However, one can estimate a rough
lifetime of the carbene by the Stern–Volmer treatment of
trapping experiments with oxygen (Fig. 8), assuming that the
precursor leading to the species observed is a triplet carbene
1
hν -N2
.
.
.
.
.
.
.
3
3
2d
2 - 2
(either 2n or 2r). Thus the ratio of the difference in optical
density (ꢀOD) as a function of the oxygen concentration is
expressed as:
hν
hν
4
3
10
ꢀOD0=ꢀOD ¼ 1 þ kO ꢃ½O2ꢅ
ð1Þ
2
Scheme 6.

1518 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
dimer diradical 2-2, which is in equilibrium with 2d. The sub-
sequent events occurring from the equilibrated mixture are
essentially the same as those observed in the photolysis of 1.
The lack of formation of the tetramer 4 in the photolysis of 3
and the unimolecular decay of the trimer diradical 10 in the
LFP of 3 suggest that 10 undergoes coupling to 3 exclusively,
without decaying to a mixture of a carbene and a dimer dirad-
ical. This means that the coupling of 2d and 2-2 leading to 10 is
essentially an irreversible pathway.
Thus, the apparently puzzling reactions involved in the pho-
tolysis of 1, 3, and 4 can be explained by the pathway shown in
Scheme 6, at least qualitatively. The exact processes of the cou-
pling to form the trimer diradical 10 from the mixture of 2d and
2-2 and the tetramer 4 from the dimer diradical 2-2 are not clear
yet. Moreover, we do not know the reason why 4 undergoes
photolysis to generate the dimer diradical 2-2 (in equilibrium
with 2d) presumably by two C–C bond cleavage, while 3 under-
goes one C–C bond cleavage to generate the corresponding
diradical 10.
An interesting issue related to Scheme 6 is the nature of 2d
regenerated by the photolysis of the tetramer 4. The finding that
the irradiation of 4 in the presence of TEMPO produced di-
anthryl ketone, which is also produced in the photolysis of 1 un-
der similar conditions, suggests that a triplet carbene 32 may be
generated from 4. However, the spectroscopic studies do not
provide definitive evidence to support this process. For in-
stance, the irradiation of 4 in a matrix indicated EPR signals
due to a doublet species without showing those ascribable to
the triplet carbene 32. This may simply mean that a dimer dir-
adical 2-2 cannot be in equilibrium with a triplet carbene 32 in a
matrix.
Dynamic EPR and UV/vis studies indicate that a nascent
triplet carbene 2n with a bent structure is observed but it re-
laxed to a more linear one 2r as the matrix is thawed and soft-
ened. The EPR spectra upon photolysis of 1 clearly indicate that
the latter species still retains the characteristic feature as a trip-
let carbene. As the matrix temperature is further raised, the trip-
let EPR signals were found to collapse to a doublet signal. UV/
vis spectral changes correlate well with EPR changes. Assum-
ing that the monomer diradical is different from the triplet car-
bene, the doublet species can be assigned to the monomer dir-
adical 2d. It may be that there is no spin interaction between the
two unpaired electrons on the 10 and 100 positions in the mono-
mer diradical 2d, while two unpaired electrons are still interact-
ing with each other in the resonance structure of the triplet car-
bene (2A in Scheme 1). Since these changes are not reversed
upon recooling of the matrix, the two species, the relaxed triplet
carbene 2r and the monomer diradical 2d, can be regarded as
different species that are not interconvertible.22,23 The triplet
carbene detected in the matrix photolysis at low temperature
is not detected in LFP at ambient temperature. This means that
the triplet carbene (2n and 2r) decays to a monomer diradical 2d
very rapidly in solution at room temperature (within 1 ms).
The overall view of the reaction is summarized in Scheme 7
(here, intermediates are indicated in parentheses). There are
still unclear issues. For instance, the mechanism of the forma-
tion of 4 from 2, especially the role of 11, the difference in the
reactivity between 10 and 11, mechanism of the reaction of
triplet carbene and diradicals with oxygen and TEMPO and
Fig. 11. Stern–Volmer analysis of the yield of transient
product as a function of oxygen.
where kO is the quenching rate constant of 32 by oxygen and ꢃ
2
is the lifetime of 32 in the absence of oxygen. A plot of ꢀOD0/
ꢀOD versus the concentration of oxygen is linear (Fig. 11), and
the slope is equal to kO ꢃ (¼ 5:7 ꢃ 103 Mꢁ1). Assuming that
2
kO is a diffusion rate constant (¼ 1010
M
ꢁ1 sꢁ1 in benzene at
2
20 ꢂC),11 one can estimate kd to be 1:8 ꢃ 106 sꢁ1, ꢃ ¼ 0:56 ms.
The mechanism of formation of the trimer 3 and the tetramer
4 from 32 is problematic. One possibility was that those oligo-
mers are formed by the reaction of the carbene 32 with the pre-
cursor diazomethane 1. For instance, carbene 32 may attack the
diazomethane 1 at position 10 with subsequent extrusion of
nitrogen, resulting in the formation of a dimer diradical, which
then may undergo further reaction with 1. This possibility could
be checked by examining the effect of the concentration of 1 on
the rate of the light-induced decay of 1. Thus, two samples con-
taining a considerably different concentration of 1 were irradi-
ated under identical conditions. However, essentially no de-
pendence of the rate on the concentration of 1 was observed.19
Moreover, the quantum yield of the photodecomposition of 1
was determined to be 0.43.20 These observations clearly indi-
cate that the photodecomposition of 1 does not involve a chain
process at least under these conditions (at 10ꢁ3 M).
LFP studies indicate that the trimer diradicals (10) are gen-
erated as a main initial product in the photolysis of diazo pre-
cursor 1. A convenient way to explain the formation of the dir-
adical 10 as a major product is to assume that a monomeric spe-
cies 2d (see below) formed from 1 is in equilibrium with the di-
mer diradical 2-2 and that the mixture decays rapidly to form
the trimer diradical 10, which eventually leads to the trimer
3. As a trimer diradical was observed as the major transient
product at the initial stage of the LFP of 1, this decay pathway
is assumed to be very rapid and efficient.14,21 The formation of
the tetramer 4 from a triplet carbene can then be explained in
terms of the dimerization of a dimer diradical, which is in equi-
librium with 2d. As the tetramer diradical (11) could not be de-
tected by LFP of either 1 or 4, and given that the dimer diradical
decays by second-order kinetics (to form 4), it is likely that the
dimer diradical undergoes coupling directly to form 4 without
recourse to the tetramer diradical 11. This may explain why this
process is much slower than other processes.
The mechanism outlined in Scheme 6 also explains the LFP
and product analysis data on the trimer 3 and the tetramer 4.
Presumably, the tetramer generates, upon photoexcitation, a

K. Yoshida et al.
Bull. Chem. Soc. Jpn., 77, No. 8 (2004) 1519
732.2 (½M ꢁ 2 ꢃ 2ꢅþ, 63.2%); 1H NMR (500 MHz, CDCl3) ꢄ
7.89 (d, J ¼ 7:35 Hz, 8H), 7.47 (dd, J ¼ 6:61, 6.98 Hz, 8H),
7.28 (overlapping with signals of CDCl3), 7.19 (dd, J ¼ 7:35,
7.53 Hz, 8H), 7.09 (dd, J ¼ 6:79, 7.54 Hz, 8H), 6.96 (d, J ¼
6:79 Hz, 8H), 6.84 (dd, J ¼ 6:98, 6.99 Hz, 8H), 5.90 (d, J ¼
7:72 Hz, 8H), 5.00 (s, 8H); 13C NMR (126 MHz, CDCl3) ꢄ
209.82, 137.46, 131.69, 130.39, 129.66, 129.05, 128.89, 128.13,
127.58, 127.20, 126.49, 126.24, 125.90, 111.8, 54.9; IR (KBr) ꢅ
(cmꢁ1) 3056 (m), 3024 (m), 2880 (m), 1598 (m), 1476 (s), 1444
(s), 1291 (m), 1155 (w), 1117 (m), 1046 (w), 768 (vs), 640 (s).
When a solution of 1 (5 mg) in nondegassed benzene (5 mL) was
irradiated (ꢀ > 300 nm) for 30 min, anthraquinone and di(9-an-
thryl) ketone were obtained in a roughly 1:1 ratio. When a similar
irradiation was carried out with long wavelength light (ꢀ > 520
nm) for 1.5 h, 9-anthryl 9-anthracenecarboxylate (9) was isolated
along with di(9-anthryl) ketone in a roughly 1:1 ratio, no anthra-
quinone being formed in this case.
5 + 9
O2
[32n]
5
TEMPO
hν
1
[32r]
- N2
[2d]
[2-2]
O2
O2
[10]
6
5
[11?]
TEMPO
TEMPO
hν
hν
3
4
9-Anthryl 9-Anthracenecarboxylate (9):
1H NMR (500
Scheme 7.
MHz, CDCl3) ꢄ 8.74 (s, 1H), 8.71 (d, J ¼ 8:85 Hz, 2H), 8.51 (s,
1H), 8.30 (d, J ¼ 7:93 Hz, 2H), 8.17 (d, J ¼ 7.93 Hz, 2H), 8.12
(d, J ¼ 7:32 Hz, 2H), 7.69 (t, J ¼ 7:01 Hz, 2H), 7.61 (t, J ¼
8:70 Hz, 2H), 7.54–7.58 (m, 4H); IR (KBr) ꢅ (cmꢁ1) 1161,
1727; UV/vis (CHCl3) 351, 369, 390 nm.
so on. Further studies to solve the remaining issues are in
progress.
Experimental
General Methods. IR spectra were measured on a JASCO A-
100 recording spectrometer and the mass spectra were recorded on
a JEOL MS route JMS-600W.
1H and 13C NMR spectra were determined on either JEOL JNM-
EX 270 or JNM-500 spectrometers. UV/vis spectra were recorded
on a JASCO CT-560 Grating Monochromator.
Thin-layer chromatography was done on a Merck Kieselgel 60
PF 254. Column chromatography was carried out on silica gel
(ICN for dry column chromatography). HPLC and GPC were
undertaken with a JASCO 800 chromatograph equipped with a
UVIDEC-100-II UV/vis detector using a Fine pack C18-T5
column (4:6 ꢃ 25 cm) and a Shodex GPC H-2001 (20 mm ꢃ
50 cm) column, respectively. Unless otherwise noted, all the
reagents employed in this study are commercial products and were
used after standard purification.
Photolysis of 9-Anthryl 9-Anthracenecarboxylate (9). Irra-
diation (ꢀ > 300 nm) of the ester (9, 1 mg) in non-degassed ben-
zene (2 mL) for 15 min gave anthraquinone (ꢄ80%) as the main
product.
Photolysis of Oligomers (3 and 4). A solution of tetramer (2
mg) in C6D6 (0.4 mL) was placed in a Pyrex NMR tube and the
sample was degassed and flame-sealed under reduced pressure. Ir-
radiation of the sample with a high-pressure, 300-W mercury lamp
1
at room temperature was monitored by H NMR, which indicated
that characteristic signals due to the trimer appeared at the expense
of those due to the tetramer, no other signals being noted. Irradia-
tion of the trimer under essentially the same conditions showed al-
most no change in the spectrum even after prolonged irradiation.
However similar irradiation of either tetramer or trimer in a non-
degassed benzene resulted in a rapid disappearance of the starting
oligomers with concomitant formation of anthraquinone as the
main product (ꢄ80% yield). Similar irradiation of oligomer (5
mg) in degassed benzene (5 mL) in the presence of TEMPO (4.7
mg) for 2 h gave dianthryl ketone as a main product (50% yield).
Irradiation of Di(9-anthryl)diazomethane in a Mixture of
Tetrahydrofuran and Methanol. A solution of 200 mg (0.50
mmol) di(9-anthryl)diazomethane 1 in 100 mL THF–methanol
(2:3) was divided among five 18 mm o.d. Pyrex tubes. The samples
were degassed, sealed, and irradiated for 15 min, using a medium-
pressure mercury lamp filtered with uranium glass. The product
analysis by thin-layer chromatography showed that only two prod-
ucts were formed (Rf ¼ 0:67 and 0.61, elution with benzene). The
residue was triturated with acetone giving as an off-white powder
130 mg of 3, (Rf ¼ 0:67). Pure compound 3 (off-white powder,
106 mg, 53% yield) was obtained by chromatography on silica
gel with 3:1 by volume hexane–benzene eluent, mp 330 ꢂC.
1H NMR (100 MHz, C6D6) ꢄ 6.10–7.40 (m, 48H) 4.65 (s, 6H);
13C NMR (25.0 MHz, C6D6, completely decoupled and off-reso-
nance decoupled) ꢄ 208.6, 124.3–136.9, 111.7, 55.26 (d, J ¼ 60
Hz); IR (KBr) ꢅ (cmꢁ1) 3060 (w), 3020 (w), 2890 (w), 1600
(w), 1480 (s), 1450 (s), 1294 (m), 1170 (m), 1050 (m), 920 (m),
770 (s); MW calcd for (C29H18)3: 1098 g/mol, found: 1109 g/
mol; elemental analysis, calcd for (C29H18)3: C, 95.05, H,
4.95%, found: C, 94.48, H, 5.51%.
Photolysis of Diazomethane (1). A solution of diazomethane
(1,3 40 mg) in anhydrous benzene (15 mL) was placed in a Pyrex
tube. The sample was degassed using a minimum of four freeze–
degas–thaw cycles at a pressure near 10ꢁ5 Torr, flame-sealed under
reduced pressure and irradiated with a high-pressure, 300-W mer-
cury lamp at room temperature until all the diazo compound was
destroyed. The irradiation mixture was then concentrated on a
rotary evaporator below 20 ꢂC and individual components were
isolated by repeated chromatography on a gel permeation column
using CHCl3 as an eluent to afford the following two productsꢂ:
Trimer (3): 22 mg (55%); white solid; mp 332.4–333.1 C;
FABMS m=z 1099.1 (M + 1, 19.1%), 732.1 ([M ꢁ 2]þ, 19.6%),
þ
1
366 (½M ꢁ 2 ꢃ 2ꢅ , 100%); H NMR (500 MHz, CDCl3) ꢄ 7.79
(d, J ¼ 7:63 Hz, 6H), 7.43 (t, J ¼ 7:02 Hz, 6H), 7.29 (t, J ¼
7:94 Hz, 6H), 7.21 (d, J ¼ 7:01 Hz, 6H), 7.15 (t, J ¼ 7:02 Hz,
6H), 6.83 (t, J ¼ 7:94 Hz, 6H), 6.34 (d, J ¼ 7:94 Hz, 6H), 5.91
(d, J ¼ 7:02 Hz, 6H), 4.82 (s, 6H); 13C NMR (126 MHz, CDCl3)
ꢄ 208.34, 136.42, 131.83, 131.75, 130.73, 129.86, 129.23,
127.60, 127.37, 127.00, 126.44, 125.88, 125.46, 111.10, 55.06;
IR (KBr) ꢅ (cmꢁ1) 3056 (m), 3024 (m), 2880 (m), 1598 (m),
1476 (s), 1444 (s), 1291 (m), 1155 (w), 1117 (m), 1046 (w), 768
(vs), 640 (s).
ꢂ
Tetramer (4): 5 mg (12%); white solid; mp 222.4–223.4 C;
FABMS m=z 1466.2 (M þ 1, 10.5%), 1098.5 (½M ꢁ 2ꢅþ, 12.2%),

1520 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
Di(9-anthryl) ketone 5 (Rf ¼ 0:61) was obtained from the ace-
tone layer and isolated upon chromatography under the same con-
ditions as those employed for 3 (29 mg, 15% yield), mp 264 ꢂC (lit.
264–266 ꢂC).24 MS (EI, 70 eV) m=e (relative intensity) 383 (7.7),
382 (M, 27), 381 (8), 353 (16), 352 (7), 205 (60), 178 (18), 177
(100), 176 (88), 170 (30).
Irradiation of Di(9-anthryl)diazomethane in 2-Methyl-
tetrahydrofuran. A solution of 0.2 g (0.5 mmol) 1 in 32 mL
2-methyltetrahydrofuran was irradiated as previously for 15 min,
at which time the reddish solution turned yellow. Evaporation of
the solvent gave 180 mg of crude product whose thin-layer chro-
matographic analysis revealed the same two products as had been
formed in the THF–methanol mixture described above. Trituration
with acetone gave a 154 mg residue of 3 as a white powder. This
was further purified by preparative thin-layer chromatography
and chromatography on silica gel as previously (128 mg, 65%
yield). The off-resonance decoupled 13C NMR and IR spectra were
identical with those of previously isolated 3. Dianthryl ketone 5
was isolated in 13% yield by chromatography of the residue recov-
ered from the solvent layer in the trituration step.
bar and the whole system was de-aerated by bubbling argon for
15 min. The cuvette was tightly capped and irradiated under stir-
ring with 366 nm light obtained from a Hitachi fluorophotometer
(F-4500) for 30 min. The decomposition percentage of 1 was meas-
ured at several wavelengths to be 29.6% under these conditions.
The radiant flux was determined by irradiating a toluene solu-
tion of E-ꢆ-[2-(2,5-dimethyl-3-furyl)ethylidene]-ꢆ0-(isopropyli-
dene)succinic anhydride (5:0 ꢃ 10ꢁ3 M) under essentially identi-
cal conditions to be 1:06 ꢃ 1016 photons/min. The quantum yield
was then determined based on those data.25
Determination of the Quantum Yield of Di(9-anthryl)diazo-
methane Photolysis in a 2:15 Â 10À2 M Solution (1:1 THF–
Phenylsilane). An actinometer solution was prepared by mixing
20 mL of an aqueous solution of uranyl acetate (2:09 ꢃ 10ꢁ2 M)
with 20 mL of an aqueous solution of oxalic acid 1:94 ꢃ 10ꢁ2
M). The solution was irradiated with the medium-pressure Hg lamp
for exactly 60 min. The number of Einsteins absorbed by the solu-
tion, 3:81 ꢃ 10ꢁ4, was calculated from the reported quantum yield
for destruction of oxalic acid, 0.57,26 and the number of moles of
oxalic acid decomposed as determined by permanganate titration.
The irradiation was repeated using a solution of 1 (340 mg, 0.86
mmol) in 40 mL of a mixture of THF and phenylsilane 1:1 by vol-
ume, irradiated in the same photolysis tube until the reddish color
of the reaction mixture turned yellow. After concentration of the
reaction mixture, an IR spectrum of the residue showed that 1
was entirely consumed. The quantum yield was calculated from
the moles of 1 reacted and the number of Einsteins absorbed to
be 9.4.
Irradiation of Di(9-anthryl)diazomethane in a Mixture of
Tetrahydrofuran and Phenylsilane. A solution of 0.2 g (0.5
mmol) 1 in 32 mL of tetrahydrofuran-phenylsilane (5:2) was irra-
diated as previously. The residue, after washing with 3 mL acetone,
afforded 3 in 70% isolated yield upon column chromatography. 5
was isolated in 12% yield from the acetone.
Reaction of Di(9-anthryl)diazomethane with Thiophenol.
To a solution of 0.2 g (0.5 mmol) 1 in 40 mL THF was added 10
mL thiophenol. The reddish color of the reaction mixture turned
to yellow-green as the solution was stirred. An IR spectrum of
the residue after evaporation of volatiles indicated the complete de-
pletion of 1. The product, isolated by column chromatography, was
identified as di(9-anthryl)methyl phenyl sulfide (166 mg, 70%
yield), mp 97 ꢂC. 1H NMR (60 MHz, CDCl3) ꢄ 5.24 (s, 1H,
PhSCHAn2), 6.2–8.7 (m, 23H, ArH); 13C NMR (25.0 MHz,
CDCl3, off-resonance decoupled) ꢄ 124.33–138.56 (19 signals,
ESR Measurements. The diazo compound was dissolved in 2-
methyltetrahydrofuran (10ꢁ3 M), and the solution was degassed in
a quartz cell by three freeze–degas–thaw cycles. The sample was
cooled in an optical transmission EPR cavity at 77 K and then ir-
radiated with a Wacom 500-W Xe lamp using a Pyrex filter. ESR
spectra were measured on a JEOL JES TE 200 spectrometer
(X-band microwave unit, 100 kHz field modulation). The signal
positions were read by the use of a gaussmeter.
ArC), 54.83 (d, J ¼ 79:3 Hz, PhSCHAn2); IR (KBr) ꢅ (cmꢁ1
)
Low-Temperature UV/Vis Spectra. Low-temperature spec-
tra at 77 K were obtained by using an Oxford variable-temperature
liquid-nitrogen cryostat (DN 1704) equipped with a quartz outer
window and a sapphire inner window. The sample was dissolved
in dry 2-MTHF, placed in a long-necked quartz cuvette of 1-mm
path length, and degassed thoroughly by repeated freeze–degas–
thaw cycles at a pressure near 10ꢁ5 Torr. The cuvette was flame-
sealed under reduced pressure, placed in the cryostat, and cooled
to 77 K. The sample was irradiated for several minutes in the spec-
trometer with a Halos 300-W high-pressure mercury lamp using a
Pyrex filter, and the spectral changes were recorded at appropriate
time intervals. The spectral changes upon thawing were also moni-
tored by carefully controlling the matrix temperature with an
Oxford Instrument Intelligent Temperature Controller (ITC 4).
Flash Photolysis. All flash measurements were made on a Uni-
soku TSP-601 flash spectrometer. Three excitation light sources
were used, depending on the precursor absorption bands and the
lifetime of the transient species. They were (i) a cylindrical 150-
W Xe flash lamp (100 J/flash with 10-ms pulse duration), (ii) a
Quanta-Ray GCR-11 Nd:YAG laser (355 nm pulses of up to 40
mJ/pulse and 5–6-ns duration; 266 nm pulses of up to 30 mJ/pulse
and 4–5-ns duration), and (iii) Lamda Physik LEXTRA XeCl ex-
cimer laser (308 nm pulses of up to 200 mJ/pulse and 17-ns dura-
tion). The beam shape and size were controlled by a focal length
cylindrical lens.
3052 (w), 2938 (w, br), 1480 (m), 1475 (m), 882 (m), 840 (m),
770 (s); MS (EI, 70 eV) m=e (relative intensity) 476 (M, 4.1),
475 (6.9), 367 (100), 365 (49), 363 (41), 353 (21), 352 (29), 339
(23), 289 (39), 220 (45), 218 (100), 189 (56), 108 (88). 5 (43
mg, 23% yield) was also isolated by column chromatography.
Irradiation of Di(9-anthryl)diazomethane in a 2-Methyl-
tetrahydrofuran Matrix. Five 18 mm o.d. Pyrex tubes were fill-
ed with aliquots of a 2:3 ꢃ 10ꢁ3 M solution of 100 mg (0.25 mmol)
1 in 100 mL 2-MTHF. The samples were irradiated for 15 min with
a medium-pressure Hg lamp at 77 K in a 27 cm o.d. Pyrex Dewar
flask. The reddish matrix turned to deep green in the course of the
irradiation; when allowed to warm slowly to room temperature, it
formed a yellow solution. Evaporation of volatiles afforded a
brown solid whose IR spectrum indicated that no 1 remained. In
contrast to the solution irradiations, the low solubility of the crude
product (acetone, THF, benzene, CS2, CHCl3) made purification
difficult. Washing with acetone gave 42 mg of a pale-yellow prod-
uct that thin-layer chromatography revealed to be 3 (Rf ¼ 0:67)
contaminated with
a long tail of unseparated compounds
(Rf ¼ 0:30{0:72). 1H and 13C NMR and IR spectra were very sim-
ilar to those of pure 3. From the acetone wash solution 5 was iso-
lated in 14% yield.
Quantum Yield Measurement. A sample of 3:33 ꢃ 10ꢁ4
M
solution of the diazomethane 1 in toluene (3 mL) was placed in
1 cm path-length cuvette equipped with a screw cap and a stirrer
A Hamamatsu 150-W xenon short arc lamp (L 2195) was used

K. Yoshida et al.
Bull. Chem. Soc. Jpn., 77, No. 8 (2004) 1521
as the probe source, and the monitoring beam, guided using an op-
tical fiber scope, was arranged in an orientation perpendicular to
the excitation source. The probe beam was monitored with a
Hamamatsu R2949 photomultiplier tube through a Hamamatsu
S3701-512Q MOS linear image sensor (512 photodiodes used).
Timing of the excitation pulse, the probe beam, and the detection
system was achieved through an Iwatsu Model DS-8631 digital
synchro scope which was interfaced to a NEC 9801 RX2 computer.
This allowed for rapid processing and storage of the data and pro-
vided printed graphic capabilities. Each trace was also displayed
on a NEC CRT N5913U monitor.
Each sample was placed in a long-necked Pyrex tube which had
a side arm connected to a quartz fluorescence cuvette and degassed
using a minimum of four freeze–degas–thaw cycles at a pressure
near 10ꢁ5 Torr immediately prior to being flashed. The sample sys-
tem was flame-sealed under reduced pressure, and the solution was
transferred to the quartz cuvette which was placed in the sample
chamber of the flash spectrometer. A cell holder block of the sam-
ple chamber was equipped with a thermostat and allowed to come
to thermal equilibrium. The concentration of the sample was ad-
justed so that it absorbed a significant portion of the excitation
light.
Acc. Chem. Res., 30, 315 (1997). b) H. Tomioka, ‘‘Advances in
Carbene Chemistry,’’ ed by U. Brinker, JAI Press, Greenwich,
CT (1998), Vol. 2, pp. 175–214. c) H. Tomioka, ‘‘Carbene Chem-
istry,’’ ed by G. Bertrand, Fontis Media S. A., Lausanne (2002),
pp. 103–152.
7
See for reviews: a) W. Kirmse, ‘‘Advances in Carbene
Chemistry,’’ ed by U. Brinker, JAI Press, Greenwich, CT (1994),
pp. 1–57. b) W. Kirmse, ‘‘Advances in Carbene Chemistry,’’ ed
by U. Brinker, JAI Press, Greenwich, CT (2001), Vol. 3, pp. 1–51.
8
For reviews, see: a) M. S. Platz, ‘‘Kinetics and Spectrosco-
py of Carbenes and Biradicals,’’ Plenum, New York (1990). b) J. E.
Jackson and M. S. Platz, ‘‘Advances in Carbene Chemistry,’’ ed by
U. Brinker, JAI Press, Greenwich, CT (1994), Vol. 1, pp. 87–160.
See also, c) G. L. Closs and B. E. Rabinow, J. Am. Chem. Soc., 98,
8190 (1976).
9
See for review: a) W. Sander, Angew. Chem., Int. Ed. Engl.,
29, 344 (1990). b) J. C. Scaiano, W. G. McGimpsey, and H. L.
Casal, J. Org. Chem., 54, 1612 (1989). c) W. Bunnelle, Chem.
Rev., 91, 336 (1991). d) W. Sander, K. Schroeder, S. Muthusamy,
A. Kirschfeld, W. Kappert, R. Boese, E. Kraka, C. Sosa, and D.
Cremer, J. Am. Chem. Soc., 119, 7265 (1997).
10 H. L. Casal, N. H. Werstiuk, and J. C. Scaiano, J. Org.
Chem., 49, 5214 (1984).
11 a) T. Sugawara, H. Iwamura, H. Hayashi, A. Sekiguchi,
W. Ando, and M. T. H. Liu, Chem. Lett., 1983, 1261. b) M. L.
Casal, S. E. Sugamori, and J. C. Scaiano, J. Am. Chem. Soc.,
106, 7623 (1984). c) H. L. Casal, M. Tanner, N. H. Werstiuk,
and J. C. Scaiano, J. Am. Chem. Soc., 107, 4616 (1985). d) R. L.
Barcus, L. M. Hadel, L. J. Johnston, M. L. Platz, T. G. Savino,
and J. C. Scaiano, J. Am. Chem. Soc., 108, 3928 (1986). e) Y.
Fujiwara, Y. Tanimoto, M. Itoh, K. Hirai, and H. Tomioka, J.
Am. Chem. Soc., 109, 1942 (1987).
The authors at Mie University are grateful to the Ministry of
Education, Culture, Sports, Science and Technology for sup-
port of this work through a Grant-in-Aid for Scientific Research
for Specially Promoted Research (No. 12002007). The Wash-
ington University work received financial support from the
National Science Foundation, under grants CHE-7520163 and
CHE-8019638.
We also thank Professor C. Trindle, University of Virginia
for disclosing his data prior to its publication.
12 a) H. Tukada, T. Sugawara, S. Murata, and H. Iwamura,
Tetrahedron Lett., 27, 235 (1986). b) A. S. Nazran, F. J. Gabe,
Y. LePage, D. J. Northcott, J. M. Park, and D. Griller, J. Am. Chem.
Soc., 105, 2912 (1983). c) A. S. Nazran and D. Griller, J. Chem.
Soc., Chem. Commun., 1983, 850. d) B. C. Gilbert, D. Griller,
and A. S. Nazran, J. Org. Chem., 50, 4738 (1985). e) A. S. Nazran,
F. L. Lee, Y. LePage, D. J. Northcott, J. M. Park, and D. Griller,
J. Phys. Chem., 88, 5251 (1984). f) H. Tomioka, T. Watanabe,
K. Hirai, K. Furukawa, T. Takui, and K. Itoh, J. Am. Chem. Soc.,
117, 6376 (1995).
References
1
For reviews of general reactions of carbenes, see: a)
W. Kirmse, ‘‘Carbene Chemistry,’’ 2nd ed, Academic Press,
New York (1971). b) ‘‘Carbenes,’’ ed by R. A. Moss and M. Jones,
Jr., Wiley, New York (1973 and 1975), Vols. 1 and 2. c) ‘‘Carbene-
(oide), Carbine,’’ ed by M. Regitz, Thieme, Stuttgart (1989).
2
See for reviews of the EPR spectra of triplet carbenes: a)
W. Sander, G. Bucher, and S. Wierlacher, Chem. Rev., 93, 1583
(1993). b) A. M. Trozzolo and E. Wasserman, ‘‘Carbenes,’’ ed
by M. Jones, Jr. and R. A. Moss, Wiley, New York (1975),
Vol. 2, pp. 185–206. c) H. Tomioka, ‘‘Advances in Strained and
Interesting Organic Molecules,’’ ed by B. Halton, JAI Press,
Greenwich, CT (2000), Vol. 8, pp. 83–112.
13 A. M. Trozzolo, E. Wasserman, and W. A. Yager, J. Chim.
Phys. Phys.-Chim. Biol., 61, 1003 (1964).
14 a) W. Adam, K. Schneider, M. Stapper, and S. Steenken, J.
Am. Chem. Soc., 119, 3280 (1997). b) K. Tokumura, N. Mizukami,
M. Udagawa, and M. Itoh, J. Phys. Chem., 90, 3873 (1986).
15 a) H. Tomioka, T. Watanabe, K. Hirai, K. Furukawa, T.
Takui, and K. Itoh, J. Am. Chem. Soc., 117, 6376 (1995). b) H.
Tomioka, T. Watanabe, M. Hattori, N. Nomura, and K. Hirai,
J. Am. Chem. Soc., 124, 474 (2002).
3
E. Wasserman, V. J. Kuck, W. A. Yager, R. S. Hutton, F. D.
Greene, V. P. Abegg, and N. M. Weinshenker, J. Am. Chem. Soc.,
93, 6335 (1971).
4
D. J. Astles, M. Girard, D. Griller, R. J. Kolt, and D. D. M.
Wayner, J. Org. Chem., 53, 6053 (1988).
For studies of other anthrylcarbenes, see: a) H. Tomioka,
16 Viscosity of 3-MP, ꢂ, relative to 2-MTHF at 77 K is report-
ed to be 1:44 ꢃ 10ꢁ6. See, S. L. Murov, I. Carmichael, and G. L.
Hug, ‘‘Handbook of Photochemistry,’’ Marcel Dekker, New York
(1993).
17 M. V. Encinas and J. C. Scaiano, J. Am. Chem. Soc., 103,
6393 (1981).
18 a) I. Moritani, S. Murahashi, H. Ashitake, K. Kimura, and
H. Tsubomura, J. Am. Chem. Soc., 90, 5918 (1968). b) P. G.
Grasse, B.-E. Brauer, J. J. Zupancic, K. J. Kaufmann, and G. B.
Schuster, J. Am. Chem. Soc., 105, 6933 (1983). c) T. Sugawara,
H. Iwamura, H. Hayashi, A. Sekiguchi, W. Ando, and M. T. H.
5
J. Nakajima, H. Mizuno, E. Iiba, and K. Hirai, Can. J. Chem.,
77, 1066 (1999). b) H. Itakura and H. Tomioka, Org. Lett., 2,
2995 (2000). c) Y. Takahashi, M. Tomura, K. Yoshida, S. Murata,
and H. Tomioka, Angew. Chem., Int. Ed., 39, 3478 (2000). d) H.
Tomioka, E. Iwamoto, H. Itakura, and K. Hirai, Nature, 412,
6847 (2001). e) E. Iwamoto, K. Hirai, and H. Tomioka, J. Am.
Chem. Soc., 125, 14664 (2003).
6
Review for persistent triplet carbenes, see: a) H. Tomioka,

1522 Bull. Chem. Soc. Jpn., 77, No. 8 (2004)
Di(9-anthryl)carbene Revisited
Liu, Chem. Lett., 1983, 1259. d) S. C. Lapin, B. E. Brauer, and G.
B. Schuster, J. Am. Chem. Soc., 106, 2092 (1984). e) L. M.
Hadel, M. S. Platz, and J. C. Scaiano, J. Am. Chem. Soc., 106,
283 (1984). f) G. W. Griffin and K. A. Horn, J. Am. Chem. Soc.,
109, 4491 (1987). g) S. H. Doss, A. A. Abdel-Wahab, E. M.
and closed shell (s2) singlet/open shell (sp) singlet in diarylcar-
benes and conclude that dianthrylcarbene is considered a biradical
because it has open shell (sp) singlet states only 1.5 kcal/mol
above the triplet state, while the closed shell (s2) singlet lies 15
kcal/mol above the triplet. See, C. Trindle, J. Org. Chem., in press.
23 The structure of sterically congested triplet carbenes is flex-
ible as evidenced by changes in the zero-field splitting parameters
with change in the rigidity of the matrix,2c and hence we assume
that those carbenes can exist in a perturbed structure on the soft en-
ergy surface. This has been shown by in situ X-ray crystallographic
observation of triplet bis(2,4,6-trichlorophenyl)carbene trapped in
a crystal. In this case, the observed structure of triplet carbene is
different from the DFP optimized one and is corresponding to a
non-global minimum on very shallow potentials. Such perturbation
is ascribable to the steric effect in a crystalline lattice. See, M.
Kawano, K. Hirai, H. Tomioka, and Y. Ohashi, J. Am. Chem.
Soc., 123, 6904 (2001).
24 J. G. Jewett, Doctoral Dissertation, Massachusetts Institute
of Technology (1962); V. P. Abbeg, Doctoral Dissertation, Massa-
chusetts Institute of Technology (1967); N. M. Weinshenker,
Doctoral Dissertation, Massachusetts Institute of Technology
(1968).
25 H. G. Heller and J. R. Langan, J. Chem. Soc., Perkin Trans.
2, 1981, 341.
26 W. Kirmse and L. Horner, Justus Liebigs Ann. Chem., 625,
34 (1959).
Fruhof, H. Durr, I. R. Gould, and N. J. Turro, J. Org. Chem., 52,
434 (1987).
¨
19 Griller et al. (Ref. 4) reported that the rate of photolysis of 1
is more than ten times as great at 10ꢁ2 M concentration than at
10ꢁ4 M.
20 At a much higher concentration of 2:15 ꢃ 10ꢁ2 M in 2-
methyltetrahydrofuran, a quantum yield for the disappearance of
1 upon irradiation with the uranium glass-filtered radiation of a
medium pressure Hg lamp of 9.4 was estimated by uranyl oxalate
actinometry.
21 Similar coupling reactions are known in the reaction of
analogous radicals such as 9-oxo- and 9,9-dimethyl dihydroanthryl
radicals. a) D. Griller, J. A. Martinho Simoes, P. Mulder, B. A.
Smith, and D. D. M. Wayner, J. Am. Chem. Soc., 111, 7872
(1989). b) T. Nagaoka, A. B. Berinstain, D. Griller, and D. D.
M. Wayner, J. Org. Chem., 55, 3707 (1990). c) C. Herberg,
H.-D. Beckhaus, and C. Ruchardt, Chem. Ber., 127, 2065 (1994).
See also Ref. 14.
22 DFT studies of a series of diarylcarbenes including di-
anthrylcarbene have been carried out by C. Trindle. The studies
deal with the carbene-biradical continuum in terms of spin distribu-
tion, central angle, and the energy difference between triplet states