Organic photochemical rearrangements of triplets and zwitterions; mechanistic and exploratory organic photochemistry

Article abstract of DOI:10.1002/ejoc.200700456

There has been some controversy about the mechanisms of reactions of enones and dienones. The question has been whether a zwitterion or diradical is involved as the intermediate. In the case of ground state species it has not been recognized that zwitterions may have diradical character. In triplet reactions there is the question of whether the rearrangements take place at the initial T1 stage or subsequently as S0 ground state zwitterions or diradicals. We now have some new rearrangements bearing on these questions. In some cases it is the triplet which rearranges, while in others it is the SO zwitterion. The zwitterion vs. diradical nature of the S0 species has been assessed by both experimental as well as theoretical means. Similarly, both experiment and computational theory have been used to determine at what stage the triplet rearrangements occur and to determine the characteristics of the rearranging species. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200700456

FULL PAPER
DOI: 10.1002/ejoc.200700456
Organic Photochemical Rearrangements of Triplets and Zwitterions;
Mechanistic and Exploratory Organic Photochemistry[‡]
Howard E. Zimmerman*[a] and Vijay Suryanarayan[a]
Keywords: Zwitterions / Diradicals / Photochemistry / Rearrangements / Computation
There has been some controversy about the mechanisms of
reactions of enones and dienones. The question has been
whether a zwitterion or diradical is involved as the interme-
diate. In the case of ground state species it has not been rec-
ognized that zwitterions may have diradical character. In
triplet reactions there is the question of whether the re-
arrangements take place at the initial T1 stage or sub-
sequently as S0 ground state zwitterions or diradicals. We
now have some new rearrangements bearing on these ques-
tions. In some cases it is the triplet which rearranges, while
in others it is the S0 zwitterion. The zwitterion vs. diradical
nature of the S0 species has been assessed by both experi-
mental as well as theoretical means. Similarly, both experi-
ment and computational theory have been used to determine
at what stage the triplet rearrangements occur and to deter-
mine the characteristics of the rearranging species.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Introduction
A rearrangement of 2,5-cyclohexadienones [note Equa-
It was as early as 1961 that photochemical rearrange- tion (1); the circle-dot-y notation designates electrons in the
ments beginning as triplets on excitation were postulated to sp, π, and py orbitals, respectively[1b]]. Still another example
rearrange via zwitterions.[1] This first example was the Type was the Type B bicyclic photochemical reaction of bicy-
(1)
(2)
clo[3.1.0]hex-3-en-2-ones,[2] as illustrated in Equation (2). In
[‡] Paper 283. Part 281: H. E. Zimmerman, O. Mitkin, J. Am. both cases the initial triplet T1 undergoes skeletal re-
Chem. Soc. 2006, 7, 12743–12749. Part 282: H. E. Zimmerman,
arrangement prior to the penultimate intersystem crossing
to ground state S0.
In the Type A example the triplet undergoes β,β-bonding
and only then undergoes intersystem crossing to the ground
state zwitterion 2Z. In the case of the Type B bicyclic re-
Pure Appl. Chem. 2006, 78, 2193–2203.
[a] Department of Chemistry, University of Wisconsin – Madison,
Madison, Wisconsin 53706, USA
E-mail: zimmerman@chem.wisc.edu
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2007, 4091–4102
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4091

H. E. Zimmerman, V. Suryanarayan
FULL PAPER
arrangement [Equation (2)], the triplet undergoes scission Michael addition of benzyl Grignard reagent to the
of the internal three-membered ring bond, followed by in- Knoevenagel reactant 11, followed by basic hydrolysis in
tersystem crossing to afford zwitterion 4Z. This species then two steps, then cyclization with polyphosphoric acid, and
undergoes a pinacolic rearrangement and aromatization.[2] NBS bromination.
The experimental evidence supporting these early sugges-
tions is considered subsequently along with our discussion 2,3-diaryl-substituted naphthol which proved to be mechan-
of the present research and further relevant literature. istically interesting. Thus, this was synthesized as outlined
Finally, our preliminary photochemical efforts led to a
Interestingly, there is a common feature in these triplet in Scheme 3.
rearrangements, namely one initial bond-altering step of the
excited state which then leads the excited state onward and
Initial Efforts – the Photochemistry of the Benzobicyclic
Ketones
only subsequently to ground state. However, the involve-
ment of zwitterions is not without controversy. Thus, it has
been suggested that the species formed are diradicals and
not zwitterions.[3]
The theme of the present research was development of
related rearrangements in new systems and exploration of
their mechanisms using both experimental and computa-
tional approaches.
The photochemistry of the unsubstituted benzobicy-
clohexenone 9 was studied initially with the idea that scis-
sion of the internal bond would lead to a zwitterion or di-
radical of present interest. As noted in Equation (3), pho-
tolysis of bicyclic compound 9 in benzene led to the
known[4] 2,3-diphenylnaphthol (26), the known[5] 3,4-di-
phenylnaphthol (27) and photoproduct 25 whose structure
was determined spectroscopically (1H NMR, 13C NMR,
HRMS and IR; note the Experimental Section). In initial
runs with short irradiation times, the ratio of 26/25 was
close to 40:60 in slight favor of 25; and the sum of these
Results
Synthesis of Reactants and Potential Photoproducts
The first reactants of interest were the diaryl tricyclic sys- remained in a 75:25 ratio to 27.
tems 9, 10a, and 10b. These were prepared by reaction of
Interestingly, the ratio of these photoproducts was de-
diaryldiazomethanes with indenone. The reaction is shown pendent on the photolysis time with naphthalenone 25 dis-
in Scheme 1. In the case of (p-cyanophenyl)(phenyl)diazo- appearing on extended photolysis with an increase of for-
methane it provided both the exo and endo stereoisomers mation of 2,3-diphenylnaphthol (26) relative to 3,4-di-
with little selectivity. Details are given in the Experimental phenylnaphthol (27).
Section.
When the photolysis was repeated in methanol, the prod-
Also of interest was the bromotetralone 17. This was syn- ucts obtained were 3,4-diphenylnaphthol (27) (44%) along
thesized using the sequence in Scheme 2. This involved with 25 (2.4%) and methyl ester 28 (42%) as depicted in
Scheme 1. Synthesis of the benzobicyclic reactants.
Scheme 2. Synthesis of a bromotetralone of interest.
4092
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 4091–4102

Organic Photochemical Rearrangements of Triplets and Zwitterions
FULL PAPER
Scheme 3. Synthesis of a 2,3-disubstituted naphthol.
(3)
(4)
Equation (4). The dependence on extent of irradiation and as outlined in Figure 1. That cyclohexadiene (ET = 53 kcal/
on solvent is considered in the Discussion Section (vide mol) nicely quenched the reaction of benzobicyclic ketone
infra).
9 in a linear fashion established the triplet multiplicity. The
The Multiplicity of the Benzobicyclic Reactions
With these reactions in mind, the question remained
whether the multiplicity of the reactions could be assumed
to be the same as in the simpler, bicyclic reactions studied
earlier [e.g. Equation (2)] where triplets were involved. In
the present case, the reactant has a benzoyl chromophore
and thus a triplet energy near that of acetophenone (ET
=
74 kcal/mol). Thus, we used cyclohexadiene quenching runs Figure 1. Inverse of conversion vs. quencher concentration.
Eur. J. Org. Chem. 2007, 4091–4102
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4093

H. E. Zimmerman, V. Suryanarayan
FULL PAPER
slope gives A(kq/kr) where kr is the triplet reaction rate, kq was of interest to ascertain whether this species could be
is the quenching rate, and A represents light absorbed in generated independently without light and whether the re-
millieinsteins per reaction time.
gioselectivity would be the same or different as found in
the photochemistry. The “Favorskii-like” generation of the
zwitterionic oxyallyl species employed the reaction of
bromo ketone 17 with base. The ratio of the regioisomeric
pair, 29a and 29b, proved to be the same as in the photolysis
of bicyclic compounds 10a and 10b.
Evidence on Zwitterion vs. Diradical Migrations
Having observed these rearrangements of the unsubsti-
tuted benzobicyclohexenone, we proceeded to investigate
the behavior of the counterparts 10a and 10b in which one
aryl group was para-substituted with a cyano moiety. In
methanol both of the known[6] 3/4-(p-cyanophenyl)-4/3-
phenylnaphthols were formed with the 3-(p-cyanophenyl)-
4-phenyl isomer 29a predominating.
Computational Aspects and Methodology – the Type A
Species
In parallel with the experimental photochemical behav-
The product ratio of 29a/29b (65:35) was independent of ior, we proceeded to investigate the chemistry computation-
the reactant stereochemistry. Additionally, the methoxycar- ally, using the dimethyl analog of zwitterions 2Z, 4Z, and
bonyl ring-opened product 30 was observed as a 50% prod- 35Z. Our computations utilized a 6-31g* basis set along
uct; see Equation (5).
with density functional B3LYP, CASSCF and HF-MP2
In dramatic contrast, a different reaction course was ob- methodology. In the case of CASSCF the active space re-
served in benzene [note Equation (6)]. Here naphthalenone quired was determined as (8,8) for the monocyclics and
31 and (independently synthesized, vide supra) 3-(p-cya- (10,10) for the benzo derivatives, using the Nemukhin–
nophenyl)-2-phenylnaphthol (24) were encountered. On ex- Weinhold approach in which bonds and lone-pairs involved
tended photolysis, the naphthol became the only product. in excitation were computed, then selected and, finally, used
That the naphthalenone 31 was a reaction intermediate as a basis. Since CASSCF does not include correlation ener-
leading to 24 was suggested by its independent photolysis gies, for interests’ sake these were obtained by extractions
which afforded 24.
from density functional computations. The purpose was to
Still another approach was pursued. The reactions in determine if the differences were of the correct order of
methanol of the diastereomeric bicyclic compounds 10a magnitude.
and 10b to afford naphthols 29a and 29b (vida supra)
Rather than using synthetic simulations of solvent, a di-
clearly proceeded via an intermediate with the central three- rect approach was employed. Thus, in the oxygen plane, two
membered ring bond severed. One possibility was a six- methanol molecules were inserted with the hydrogen atoms
membered ring oxyallyl species [see Equation (7)]. Hence, it oriented as to aim at the py orbital lobes. In the case of four
(5)
(6)
(7)
4094
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 4091–4102

Organic Photochemical Rearrangements of Triplets and Zwitterions
FULL PAPER
Table 1. B3LYP results for Type A zwitterion 2Z.
Species
Relative energy
(Hartrees)[a]
Relative energy
Relative energy
(kcal/mol)
Correlation energy
(Hartrees)
(minus MeOH)[a]
Zwitterion, no methanol molecules
Zwitterion, two methanol molecules
Zwitterion, four methanol molecules
Methanol
0
0
0.00000
26.1040
46.5605
–
2.8404
4.393
5.948
–
–231.4704
–462.9318
–115.7144
0.0416[b]
0.0742[c]
–
[a] –385.9737 Hartrees as the zero. [b] 231.4288 per 2 MeOH molecules. [c] 462.8576 per 4 MeOH molecules.
Table 2. CASSCF results for Type A zwitterion 2Z.
Species
Relative energy
(Hartrees)[a]
Natural occupation analysis
HOMO
LUMO
Zwitterion, no methanols molecules
Zwitterion, two methanol molecules
Zwitterion, four methanol molecules
0
1.59887
1.71439
1.83586
0.34516
0.27879
0.15107
230.0339
460.1333
[a] –383.59978 Hartrees as the zero.
Table 3. CASSCF results: monocyclic Type B and benzobicyclic Type B species.
Species
Relative energy
(Hartrees)
Natural occupation analysis
HOMO
LUMO
Type-B zwitterion 4Z
–383.60871
–536.32070
–996.40054
1.64538
1.44264
1.78910
0.34958
0.55675
0.19627
Benzo zwitterion 35Z
Benzo zwitterion 35Z, 4 methanol molecules
methanol molecules, these solvent molecules were added terms of electronic excited state structures. Thus, the
and placed above and below the molecular plane. With two mechanism in Equation (9) [see also Equation (1)] is one we
and with four solvent molecules, geometry optimization was presented in our early series which correlated the known
performed. The in-plane methanol molecules did not move photochemical reactions with their electronically excited
appreciably, while those above and below the molecular structures. Equation (9) provides the example of the Type
plane moved slightly more.
A cyclohexadienone rearrangement.
All of the computational results are summarized in
Tables 1, 2, and 3.
Discussion
Background
(9)
Our interest in photochemistry involving zwitterionic re-
arrangements goes back decades.[1] An early example of the
Type A rearrangement was the Santonin-to-Lumisantonin
rearrangement as shown in Equation (8).
Similarly, our mechanism for the photochemistry of bicy-
clo[3.1.0]hex-2-en-3-ones involved enolate zwitterions as
noted in Equation (10) [note also Equation (2)].
(8)
(10)
The structure of Santonin had been established by
Woodward and Yates.[7] That of Lumisantonin derived from
However, there has been some controversy about whether
the work of Barton et al.[8] It was at that point that we first these species really are zwitterions or diradicals. A reason-
suggested the intervention of zwitterions in a mechanism able definition of zwitterions relies on whether one has a
leading to Lumisantonin. Actually, this was a time when largely closed shell species. If the principal configurations
photochemical reactions had not yet been described in mainly have one electron promoted to an antibonding MO
Eur. J. Org. Chem. 2007, 4091–4102
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4095

H. E. Zimmerman, V. Suryanarayan
FULL PAPER
while one remains in the bonding set, we would term the Photochemistry of the Benzobicylic Ketones – the Diphenyl
species a diradical. Conversely, with the virtual orbitals System
largely vacant and the bonding orbitals close to doubly oc-
Perhaps based on intuition or based on thoughts of pos-
cupied, the species would be termed a zwitterion.
sible mechanisms, the two reaction intermediates 25 and 34
In the ground state (i.e. S0), depending on the extent of
were considered as involved in much of the photochemistry
single electron promotion, one might consider terming the
of the benzobicyclic ketones with ketene 34 being a precur-
species a zwitterion or a diradical. But these correspond to
sor to diphenylnaphthalenone 25 [Equation (11)].
the same S0 ground state. In valence bond terminology one
is merely dealing with resonance structures of one species.
In the case of triplets one clearly has a diradical.
For the Type A rearrangement, there has been some
question whether the species shown are actually intermedi-
ates or are merely points on the reaction hypersurface.[3c] In
a very pretty computational study Robb et al.[3c] has veri-
fied that our original pathway shown in Equation (1) is
(11)
correct. Somewhat earlier, using an AM1 computation,
In fact, the photolysis of benzobicyclic ketone 9 as shown
Kletskii et al.[9] also confirmed the nature of species in- in Scheme 4 afforded (inter alia) the two photoproducts 25
volved. Discussion of our own computations on our sys- and 26. A complete mechanism for the photolysis of benzo-
tems is delayed to a section below.
bicyclic ketone 9 is given in Scheme 4.
But, perhaps, the most convincing evidence tends to de-
As noted above, 2,2-diphenylnaphthalenone (25), 2,3-di-
rive from experimental observations. Thus, our early studies phenylnaphthol (26), and 3,4-diphenylnaphthol (27) were
revealed that the Type A rearrangements have very high isolable photoproducts, while the diphenyl ketene 34 was
quantum yields, often approaching unity.[10] This signifies not detectible. However, evidence for the intermediacy of
that there are no efficient conical intersections or points of this ketene is present (vide infra). It is seen that the 3,4-
triplet intersystem crossing occurring prior to the transition diphenylnaphthol arises by scission of internal bond a,
state being reached. Otherwise, the quantum yields would while 2,2-diphenylnaphthalenone (25) and 2,3-diphenyl-
be low as a consequence of the reacting species being re- naphthol (26) arise from fission of bonds b and c. Addition-
turned to reactant ground state. A second piece of evidence ally, 2,3-diphenylnaphthol proved to be a secondary pho-
was the generation of the zwitterion Z (S0) by a ground toproduct, being formed from 2,2-diphenylnaphthalenone.
state “Favorskii-like” approach generating the zwitterion Its formation from zwitterion 35 S0 is mechanistically ruled
directly without the use of light.[11]
out by the loss of aromaticity if a phenyl group were to
Still another example involving zwitterions is seen in the bond to C-2. It did prove possible to trap the intermediate
preferential phenyl migration in the photolysis of 6,6-diaryl- ketene 34 by photolysis in methanol [note Equation (4)].
bicyclo[3.1.0]hex-3-en-2-ones [refer to Equation (10)].[12] In This intermediate proceeded to undergo pericyclic six-mem-
this reaction the internal three-membered ring bond is bered ring closure as indicated in Equation (11). Thus, in
severed to afford a divalent species. If this species were a the photochemistry there were two distinct routes – A and
diradical, the cyanophenyl group would migrate in prefer- BC – encountered (note Scheme 4). The partition proved to
ence, but does not.[12] Additionally, the effect of acid on the depend on the solvent employed.
photorearrangements of these species was highly suggestive
Interestingly, path A occurs to a greater extent in meth-
anol, while Path BC is favored by use of benzene. To the
of a zwitterion.[2a]
Scheme 4. Postulated mechanisms in the photolysis of benzobicyclic ketone 9.
4096
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 4091–4102

Organic Photochemical Rearrangements of Triplets and Zwitterions
FULL PAPER
Scheme 5. Phenyl migration preference signifies zwitterionic character; dependence of migrations on zwitterionic vs. diradical character.
extent that the carbonyl moiety has its py orbital hydrogen- photolysis of the parent cyano-benzobicyclohexenone 10 in
bonded as in methanol, the n-π* species utilizes its π-sys- benzene. In the photolysis in methanol the open-chain ester
tem, with the one added electron, to break the internal 30 was also isolated. This clearly results from the trapping
three-membered ring bond. Conversely, in absence of hy- of ketene 40 (Scheme 6); see Equation (5). The significance
drogen bonding as in benzene, the py orbital interacts with of formation of naphthalenone 31 is that this then accounts
bond b (note Scheme 4), while the π-system interacts with for formation of 3-(p-cyanophenyl)-2-phenylnaphthol (24),
bond c. This is reminiscent of our earlier comment[1] that since the latter is just a di-π-methane-like product resulting
n-π* excited states exhibit dual sources of reactivity – that from migration of one of the aryl groups at C-2 to the styryl
arising from the singly occupied py orbital and that re- moiety at C-3.
sulting from the reactivity of the extra π-electron.
Photochemistry of the Benzobicylic Ketones – the
Cyanophenyl System
With the diphenyl-substituted benzobicyclic system giv-
ing general mechanistic outlines, p-cyano substitution pro-
vided further insight. The reaction of the diastereomeric 6-
(p-cyanophenyl)-6-phenylbenzobicyclo[3.1.0]hexanones 10a
and 10b led to the 3,4-diarylnaphthols 29a and 29b in com-
petition with formation of ketene 40 [see Equation (5) and
Scheme 6. Reaction of the intermediate ketene 40.
Scheme 6]. Interestingly the phenyl migration product 3-(p-
cyanophenyl)-4-phenylnaphthol (29a) was the major isomer
being formed in a ratio of 65:35 relative to the cyanophenyl
migration isomer. This is understood if the center to which
the species undergoing aryl migration is electron-deficient
rather than odd-electron in character; see Scheme 5. A re-
Computational Aspects of the Diradical – Zwitterion
Problem
In view of the controversy[3] about the mechanism of the
lated and important point is that the same ratio of 29a/29b reactions involving diradical-zwitterion intermediates com-
resulted independent of the reactant diastereomer, 10a or putations were undertaken. The case of the Type B bicyclic
10b, utilized, thus indicating a common intermediate such rearrangement is typical. Most of the computational results
as 38.
have been of the “gas-phase” variety, sometimes with a gen-
Hence, the direct generation of the ground state (i.e. S0) eralized dielectric being imposed, but with an explicit mole-
zwitterion 38, utilizing the first step of a Favorskii reaction cule of water being introduced in one case.
as described in Equation (7), is particularly meaningful.
Probably the most reliable assessment of the electronic
The same products – 3-(p-cyanophenyl)-4-phenylnaphthol nature of the molecular wavefunction is natural orbital
(29a) and 4-(p-cyanophenyl)-3-phenylnaphthol (29b) – as in analysis.[13] This affords the electron occupation of the
the photochemistry were formed and in the same 80:20 ra- bonding and antibonding natural orbital molecular orbit-
tio (within experimental error).
als. We recognize that these natural molecular orbitals rep-
The ratio was independent of the bromo ketone stereo- resent the inclusion of the various configurations in terms
chemistry. This ground state approach provides further evi- of scf MOs. A species which could be termed a true diradi-
dence for the intermediate zwitterion species 38.
cal will have occupations of the bonding and antibonding
In addition to the naphthol photoproducts 2-(p-cya- MOs as 1:1 as in S1. A true closed shell zwitterion will have
nophenyl)-2-phenylnaphthalenone (31) was isolated from a HOMO occupation of 2:0 as in S0.
Eur. J. Org. Chem. 2007, 4091–4102
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4097

H. E. Zimmerman, V. Suryanarayan
FULL PAPER
Figure 2. Natural orbital occupancies.
Figure 3. Electron densities and charges in the π-system.
If we now refer to Tables 1, 2, and 3, as summarized in ceed by mechanisms in which the penultimate step is re-
Figure 2, we note that the antibonding natural MOs are arrangement of a zwitterion. This evidence is both experi-
weakly populated, while the electron density in the highest mental and computational. The experimental evidence rests
energy bonding MOs have electron densities approaching on (1) a preference of phenyl groups to migrate compared
two. Further, as a test, in the case of the Type A species, with p-cyanophenyl counterparts, and (2) generation of the
methanol molecules have been explicitly put in proximity of same ground state species from a Favorskii-like generation.
the carbonyl oxygen atom. The first two were put in the Additionally, in this study a number of new transformations
molecular plane, and the next two above and below the mo- have been discovered.
lecular plane. It is seen (note Figure 2) that as methanol
solvation is increased, the occupancy of the virtual orbitals
diminishes, and the bonding MOs approach an occupancy
of two as in a closed shell.
Experimental Section
General Procedures: Melting points were determined in open capil-
laries and are uncorrected. Column chromatography was per-
formed on silica gel (60–80 mesh) mixed with 1% of Sylvania 2282
(green) phosphor and slurry-packed into quartz columns to allow
A further result of interest is the effect of solvent, here
methanol, on the diradical character of the species. Here we
select the Type A species to illustrate the point. Thus, using
the Weinhold NBO analysis to give bond orbital densities,
we find (Figure 3) carbon atoms 2 and 6 to be electron-
deficient with less than unit density while the oxygen p-π
orbital is electron-rich. The effect of methanol molecules
clustering around the oxygen atom enhances the polariza-
tion dramatically. The carbonyl carbon atom is minimally
electron-deficient; considerable density has been fed to the
antibonding π* bond. Overall we see with four methanol
molecules a total of 1.4 electrons being transferred to the
oxygen p-π orbital. The charges are also listed in Figure 3.
Interestingly, the Type A rearrangement occurs nicely in po-
lar solvents while in benzene it becomes a minor process.
We also need to note that there clearly is some diradical
character in the oxyallyl moiety, but this is minimal and
overridden by the zwitterionic nature of the species. Ad-
ditionally, we conclude that “gas-phase” computations
exaggerate this diradical character of the zwitterions. One
can best term these species as “zwitterionic with some di-
radicaloid” character.
monitoring with a hand-held UV lamp. H and 13C NMR spectra
1
were recorded at 300 MHz and 75.4 MHz, respectively, with TMS
as an internal standard. Photolysis experiments were carried out in
an enclosed box with a 400 W medium-pressure mercury lamp in
a quartz immersion well with a 0.2  copper sulfate filter solution
and under nitrogen.
Synthesis of Inden-1-one (6): A useful and simpler method was em-
ployed for the synthesis of inden-1-one.[14] To a stirred solution of
2.09 g (10 mmol) of 3-bromoindan-1-one[15] in 20 mL of diethyl
ether, 3.03 g (30 mmol) of triethylamine was added dropwise at
room temperature over 10 min. The hydrobromide salt separated
out during addition. The reaction mixture was stirred for further
1 h and quenched with cold water. The organic layer was separated,
washed well with water, brine and dried with anhydrous sodium
sulfate. Removal of ether under vacuum afforded 1.0 g (77.0%) of
6 as pale yellow oil which was pure enough (purity Ͼ99%) to be
1
used directly for further reactions. H NMR (CDCl3): δ = 5.89 (d,
J = 6.0 Hz, 1 H), 7.06 (d, J = 6.9 Hz, 1 H), 7.23 (t, J = 6.9 Hz, 1
H), 7.34 (t, J = 6.9 Hz, 1 H), 7.43 (d, J = 6.9 Hz), 7.57 (d, J =
6.0 Hz, 1 H) ppm.
Synthesis of (p-Cyanophenyl)(phenyl)diazomethane (8):[16] A mixture
of 3.5 g (69 mmol) of hydrazine hydrate, 50 mL of absolute ethanol
and 10 g of anhydrous sodium sulfate was heated to reflux. A solu-
tion of 718 mg (3.47 mmol) of p-cyanobenzophenone in 50 mL of
Conclusions
The evidence is that the photochemical processes in the
Type A, the bicyclic Type B, and the benzo Type B – pro- ethanol and 20 mL of benzene was added over 8 h, at which point
4098 Eur. J. Org. Chem. 2007, 4091–4102
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Organic Photochemical Rearrangements of Triplets and Zwitterions
FULL PAPER
the reaction mixture had a light yellow color. The mixture was
refluxed overnight. The solution was filtered under gravity to re-
move the sodium sulfate and the solvent removed under vacuum.
The residual solid, on crystallization from absolute ethanol, af-
forded p-cyanobenzophenone hydrazone as light yellow needles,
230 mg (30.2%); m.p. 165 °C. 1H NMR (CDCl3): δ = 5.41 (s, 2 H),
Irradiation of 1,1-Diphenyl-1a,6a-dihydro-1H-cyclopropa[a]inden-6-
one (9) in Benzene: A solution of 250 mg (0.84 mmol) of 9 in
250 mL of benzene was purged with nitrogen for 1 h. The irradia-
tion was then carried out with a 0.2  CuSO4 filter solution for
10 min. TLC indicated the absence of the starting material. The
solvent was removed under vacuum and the residue subjected to
7.24–7.84 (m, 9 H) ppm. 13C NMR (CDCl3): δ = 113.28, 118.48, column chromatography (hexane, hexane/ethyl acetate, 19:1, 9:1) to
126.52, 128.60, 128.79, 130.21, 133.44, 137.69, 138.26, 146.99 ppm.
HRMS: calcd. for C14H11N3 221.0953 (monoisotopic); found
221.0954. To a solution of 230 mg (1.04 mmol) of p-cyanobenzo-
phenone hydrazone in 20 mL of dry diethyl ether was added 1.0 g
of anhydrous sodium sulfate, 906 mg (4.23 mmol) of yellow mercu-
ric oxide and 0.50 mL of saturated ethanolic potassium hydroxide.
The reaction mixture was stirred in the dark overnight (care was
taken to avoid light by covering the flask with aluminium foil).
Filtration under gravity with care to exclude light left a deep orange
solution which was concentrated in vacuo to afford 170 mg (74.8%)
of 8. This was crystallized from diethyl ether/hexane (80:20) as vi-
afford the following products:
Fraction 1. 2,3-Diphenyl-1-naphthol (26): Crystallized from hexane,
Yield: 70 mg (28.2%), m.p. 129 °C (ref.[4] m.p. 128–130 °C).1H
NMR (CDCl3): δ = 5.70 (s, 1 H), 7.13–8.35 (m, 15 H) ppm.
Fraction 2. 2,2-Diphenyl-2H-naphthalen-1-one (25): Crystallized
from hexane/pentane, 1:1, Yield: 115 mg (46.3%), m.p. 124 °C. H
1
NMR (CDCl3): δ = 6.52 (d, J = 9.6 Hz, 1 H), 6.76 (d, J = 9.6 Hz,
1 H), 7.25–7.58 (m, 13 H), 8.06 (d, J = 8.4 Hz, 1 H) ppm. 13C
NMR (CDCl3): δ = 63.18, 123.81, 127.50, 127.57, 127.94, 128.07,
128.23, 128.63, 128.66, 128.88, 129.09, 129.34, 129.53, 134.90,
137.90, 139.06, 142.71, 200.21 ppm. HRMS: calcd. for C22H16O
296.1201; found 296.1194. The IR spectrum contained a carbonyl
stretching peak at 1684 cm–1.
1
olet needles; m.p. 109 °C. H NMR (CDCl3): δ = 7.24–7.61 (m, 9
H) ppm. 13C NMR (CDCl3): δ = 107.68, 119.41, 123.76, 126.97,
127.08, 127.30, 129.81, 132.53, 133.01, 136.06 ppm. HRMS: calcd.
for C14H9N3 219.0796; found 219.0791.
Fraction 3. 3,4-Diphenyl-1-naphthol (27): Yield: 60 mg (24.1%),
1
Synthesis of 1,1-Diphenyl-1a,6a-dihydro-1H-cyclopropa[a]inden-6-
one (9): To a solution of 1.32 g (10.19 mmol) of indenone in 20 mL
of benzene and 20 mL of chloroform was added a solution of 2.23 g
(11.5 mmol) of diphenyldiazomethane in 20 mL of benzene
through a hypodermic needle over 4 h. A gentle reflux was main-
tained during the addition. After the addition was complete, the
reaction mixture was further refluxed for 3 h and stirred at room
temperature overnight. The reaction mixture was diluted with
water. The organic phase was dried and concentrated in vacuo to
afford a semisolid which on column chromatography (hexane/ethyl
acetate, 9:1) and crystallization from benzene/hexane (1:1) afforded
m.p. 142 °C (ref.[5] m.p. 142–143 °C). H NMR (CDCl3): δ = 5.40
(s, 1 H), 6.92 (s, 1 H), 7.09–7.64 (m, 13 H), 8.27 (d, J = 9.0 Hz, 1
H) ppm.
Irradiation of 1-(p-Cyanophenyl)-1-phenyl-1a,6a-dihydro-1H-cyclo-
propa[a]inden-6-one (10) in Benzene:
(0.77 mmol) of the diastereomer isomer 10a in 250 mL of benzene
was purged with nitrogen for 1 h. The irradiation was then carried
A solution of 250 mg
1
out with a 0.2  CuSO4 filter solution for 10 min. The H NMR
spectrum showed a 60% conversion to products. The solvent was
removed under vacuum and the residue subjected to column
chromatography (hexane, hexane/ethyl acetate, 19:1) to afford the
following products:
1
1.1 g (36.5%) of pure 9, m.p. 130 °C. H NMR (CDCl3): δ = 3.20
(d, J = 5.1 Hz, 1 H), 3.68 (d, J = 5.1 Hz, 1 H), 7.01–7.54 (m, 14
H) ppm. 13C NMR (CDCl3): δ = 36.04, 41.70, 59.80, 123.94,
126.34, 126.34, 126.99, 127.20, 127.33, 128.34, 128.820, 131.63,
133.89, 136.69, 136.95, 144.51, 150.94, 200.78 ppm. HRMS: calcd.
for C22H16O 296.1202; found 296.1189.
Fraction 1. 3-(p-Cyanophenyl)-2-phenyl-1-naphthol (24): Recrys-
tallized from hexane/diethyl ether, 9:1, Yield 90 mg (36.4%), m.p.
1
177 °C. H NMR (CDCl3): δ = 5.74 (s, 1 H), 7.18–7.57 (m, 12 H),
7.85 (m, 1 H), 8.33 (m, 1 H) ppm. 13C NMR (CDCl3): δ = 110.54,
119.24, 120.57, 121.45, 122.94, 124.06, 126.49, 127.63, 127.99,
128.61, 129.86, 130.70, 131.44, 131.81, 133.85, 134.91, 137.99,
146.57, 149.07 ppm. HRMS: calcd. for C23H15NO 321.1154; found
321.1144.
Synthesis of 1-(p-Cyanophenyl)-1-phenyl-1a,6a-dihydro-1H-cyclo-
propa[a]inden-6-one (Diastereomers 10a and 10b): To a solution of
1.32 g (10.19 mmol) of indenone in 20 mL of benzene and 20 mL
of chloroform was added a solution of 2.52 g (11.5 mmol) of 8 in
20 mL of benzene through a hypodermic needle over 4 h. A gentle
reflux was maintained during the addition. After the addition was
complete, the reaction mixture was further refluxed for 3 h and
stirred at room temperature overnight. The reaction mixture was
diluted with water. The organic phase was dried and concentrated
in vacuo to afford a semisolid. Column chromatography (hexane/
ethyl acetate, 9:1) afforded a mixture of the diastereomers which
were separated after repeated crystallizations from hexane/ethyl
acetate (8:2). 10a: Yield: 700 mg (21.4%); m.p. 170 °C. 1H NMR
(CDCl3): δ = 3.22 (d, J = 5.1 Hz, 1 H), 3.71 (d, J = 5.1 Hz, 1 H),
7.06–7.54 (m, 13 H) ppm. 13C NMR (CDCl3): δ = 35.66, 41.29,
59.41, 111.16, 118.72, 124.43, 126.44, 127.03, 127.90, 128.02,
129.23, 132.25, 132.28, 134.51, 136.20, 143.03, 150.50, 200.32 ppm.
Fraction 2. 2-(p-Cyanophenyl)-2-phenyl-2H-naphthalen-1-one (31):
Yield 85 mg (34.4%), m.p. 141 °C. H NMR (CDCl3): δ = 6.45 (d,
1
J = 9.8 Hz, 1 H), 6.82 (d, J = 9.8 Hz, 1 H), 7.08–7.64 (m, 12 H),
8.06 (d, J = 6.9 Hz, 1 H) ppm. 13C NMR (CDCl3): δ = 63.33,
111.51, 119.10, 125.01, 127.93, 128.13, 128.15, 128.29, 129.93,
129.16, 130.15, 130.71, 131.66, 131.71, 135.46, 137.50, 141.49,
148.53, 199.15 ppm. HRMS: calcd. for C23H15NO 344.1051 [M +
Na]+; found 344.1047 [M + Na]+.
Under similar experimental conditions, the irradiation of the dia-
stereomer 10b afforded 24 and 31 with a yield of 35.1% and 36.8%,
respectively.
Irradiation of 1,1-Diphenyl-1a,6a-dihydro-1H-cyclopropa[a]inden-6-
HRMS: calcd. for C23H15NO 321.1154; found 321.1142. 10b: one (9) in Methanol: A solution of 250 mg (0.84 mmol) of the tricy-
1
Yield: 640 mg (19.57%); m.p. 179 °C. H NMR (CDCl3): δ = 3.17
(d, J = 5.4 Hz, 1 H), 3.67 (d, J = 5.4 Hz, 1 H), 7.06–7.60 (m, 13
H) ppm. 13C NMR (CDCl3): δ = 36.32, 41.67, 58.24, 111.15, 118.91
124.23, 126.48, 127.52, 127.70, 127.82, 128.78, 131.80, 132.72,
134.28, 136.49, 149.34, 150.31, 199.93 ppm. HRMS: calcd. for
C23H15NO 321.1154; found 321.1147.
clic compound 9 in 250 mL of methanol was purged with nitrogen
for 1 h. The irradiation was then carried out with a 0.2  CuSO4
filter solution for 10 min. The 1H NMR indicated 10% of unre-
acted starting material. The solvent was removed under vacuum
and the residue subjected to column chromatography (hexane, hex-
ane/ethyl acetate, 19:1) to afford the following products:
Eur. J. Org. Chem. 2007, 4091–4102
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4099

H. E. Zimmerman, V. Suryanarayan
FULL PAPER
Fraction 1. Methyl 2-(3,3-Diphenylallyl)benzoate (28): Yield 115 mg
(41.8%), pale yellow oil, 1H NMR (CDCl3): δ = 3.83 (s and d
merged, 5 H), 6.26 (t, J = 7.8 Hz 1 H), 7.19–7.44 (m, 13 H), 7.89
(d, 1 H) ppm. 13C NMR (CDCl3): δ = 34.48, 52.20, 126.30, 125.96,
127.21, 127.58, 128.05, 128.29, 128.50, 130.18, 130.81, 130.85,
132.29, 140.09, 142.62, 142.65, 142.74, 168.37 ppm. HRMS: calcd.
for C23H20O2 351.1361 [M + Na]+; found 351.1375 [M + Na]+.
139.25, 141.38, 144.37, 151.65 ppm. HRMS: calcd. for C23H15NO
321.1154; found 321.1141.
Under similar conditions, the irradiation of the diastereomer 10b
in methanol gave identical results with an experimental error of
Ϯ2%.
Synthesis of Ethyl 3-Benzyl-3-(p-chlorophenyl)-2-cyano-3-phenylpro-
pionate (12): A solution of 26.0 g (85 mmol) of ethyl α-cyano-β-
phenyl-β-(p-chlorophenyl)acrylate (11)[17] in 30 mL of dry THF
was added over 45 min at room temperature to a Grignard solution
prepared from 10.37 g (90 mmol) of benzyl chloride and 2.2 g
(90 mmol) of magnesium turnings in 25 mL of dry diethyl ether.
When the spontaneous reflux had subsided the reaction mixture
was stirred and heated to reflux for another 3 h. The mixture was
then poured onto cracked ice, acidified with 20% hydrochloric acid,
extracted with ethyl acetate, dried and concentrated under vacuo.
The crude product thus obtained was washed with diethyl ether to
remove unreacted 11. Crystallization from ethanol afforded 14.0 g
(40.8%) of 12 as a mixture of inseparable diastereomers, m.p.
125 °C. 1H NMR (CDCl3): δ = 0.80–0.953 (m, 3 H), 3.07–3.14 (m,
1 H), 3.80–3.99 (m, 3 H), 4.30 (s, 1 H), 6.56–6.64 (m, 2 H), 6.72–
6.79 (m, 2 H), 7.04–7.46 (m, 10 H) ppm. 13C NMR (CDCl3): δ =
13.66, 25.84, 44.41, 54.50, 62.87, 116.74, 127.98, 128.01, 128.06,
128.18, 128.33, 128.38, 129.01, 129.21, 130.61, 130.76, 131.30,
131.35, 133.47, 134.01, 135.51, 135.54, 141.11, 164.96, 165.06 ppm.
HRMS: calcd. for C25H22ClNO2 403.1339; found 403.1330.
Fraction 2. 2,2-Diphenyl-2H-naphthalen-1-one (25): Yield 6 mg
(2.4%).
Fraction 3. 3,4-Diphenyl-1-naphthol (27): Yield 109 mg (43.9%).
Irradiation of 1,1-Diphenyl-1a,6a-dihydro-1H-cyclopropa[a]inden-6-
one (9) in Methanol with 1,3-Cyclohexadiene as Quencher: Three
different experiments were performed under identical photolysis
conditions. The conversion to products in all cases was determined
1
by H NMR (CDCl3).
(a) A solution of 11.8 mg (0.04 mmol) of 9 in 80 mL of methanol
and 120 mg of 1,3 cyclohexadiene (1.5 mmol, 0.018  solution) was
purged with nitrogen for 1 h. Irradiation was carried out in a 0.2 
CuSO4 filter solution for 10 min, and the resulting solution was
1
concentrated under vacuum. H NMR of the residue showed 50%
of starting ketone 9.
(b) A solution of 11.8 mg (0.04 mmol) of 9 in 80 mL of methanol
and 316 mg of 1,3 cyclohexadiene (50 mmol, 0.049  solution) was
purged with nitrogen for 1 h. After irradiation for 10 min, the re-
1
sulting solution was concentrated under vacuum. H NMR of the
Synthesis of 3-Benzyl-3-(p-chlorophenyl)-3-phenylpropionitrile (13):
A solution of 10 g (24.81 mmol) of 12, 1.38 g (24.81 mmol) of po-
tassium hydroxide and 40 mL of ethylene glycol was heated under
reflux under nitrogen for 3 h. It was then cooled, diluted with water
and acidified with 20% hydrochloric acid to afford a colorless solid.
Filtration and crystallization from ethanol afforded 8.0 g (97.4%)
residual semisolid revealed the presence of 93% unconverted start-
ing material 9.
(c) A solution of 11.8 mg (0.04 mmol) of 9 in 80 mL of methanol
was purged with nitrogen for 1 h. After irradiation for 10 min, the
1
resulting solution was concentrated under vacuum. H NMR con-
firmed the presence of 12% of unconverted ketone 9.
1
of nitrile as colorless plates, m.p. 134 °C. H NMR (CDCl3): δ =
Irradiation of 1-(p-Cyanophenyl)-1-phenyl-1a,6a-dihydro-1H-cyclo-
2.93 (s, 1 H), 3.58 (s, 1 H), 6.64 (m, 2 H), 7.04–7.35 (m, 14 H) ppm.
13C NMR (CDCl3): δ = 28.34, 44.63, 49.39, 118.05, 127.18, 127.48,
127.87, 128.19, 128.62, 128.66, 129.40, 130.56, 133.22, 135.66,
143.77, 144.46 ppm. HRMS: calcd. for C22H18ClN 331.1128; found
331.1122.
propa[a]inden-6-one (10) in Methanol:
A solution of 250 mg
(0.77 mmol) of the diastereomer 10a in 250 mL of methanol was
purged with nitrogen for 1 h. The irradiation was then carried out
with a 0.2  CuSO4 filter solution for 10 min. The 1H NMR
showed a complete conversion of starting ketone to products. The
solvent was removed under vacuum and the residue subjected to
column chromatography (hexane, hexane/ethyl acetate, 19:1, 9:1,
and 6:1) to afford the following products:
Synthesis of 3-Benzyl-3-(p-chlorophenyl)-3-phenylpropionic Acid
(14): A solution of 8.0 g (24.13 mmol) of 13 was refluxed with
4.05 g (72.39 mmol) of KOH in 30 mL of ethylene glycol under
nitrogen for 6 h. The cooled reaction mixture was poured onto
crushed ice and acidified with 20% hydrochloric acid to pH = 2–
3. Crude 14 was collected by filtration and crystallized from hexane
Fraction 1. Methyl 2-[3-(p-Cyanophenyl)-3-phenylallyl]benzoate
(30): Yield 137 mg (50.5%), colorless oil. 1H NMR (CDCl3): δ =
3.81 (d, J = 7.5 Hz, 2 H), 3.85 (s, 3 H), 6.33 (t, J = 7.5 Hz, 1 H),
7.14–7.44 (m, 11 H), 7.67 (d, J = 6.0 Hz, 1 H), 7.91 (d, J = 7.8 Hz,
1 H) ppm. 13C NMR (CDCl3): δ = 30.36, 52.10, 110.37, 119.65,
126.23, 127.64, 128.24, 128.46, 128.68, 129.06, 129.28, 130.48,
130.91, 131.13, 132.05, 132.49, 133.53, 138.86, 143.00, 167.46 ppm.
HRMS: calcd. for C24H19NO2 353.1416; found 353.1428.
1
as colorless plates, 7.8 g (92.6%), m.p. 175 °C. H NMR (CDCl3):
δ = 2.99 (s, 2 H), 3.65 (s, 2 H), 6.61 (d, J = 8.1 Hz, 2 H), 7.00–7.28
(m, 12 H) ppm. 13C NMR (CDCl3): δ = 41.12, 43.68, 49.31, 126.63,
126.77, 127.81, 128.20, 128.22, 129.72, 131.11, 132.41, 137.14,
145.71, 146.47, 176.93 ppm. HRMS: calcd. for C22H19ClO2
349.0996 [M – H]–; found 349.1005 [M – H]–.
Fraction 2. 3-(p-Cyanophenyl)-4-phenyl-1-naphthol (29a): Yield
81 mg (32.0%), m.p. Ͼ250 °C. 1H NMR (CDCl3): δ = 5.51 (s, 1
H), 6.87 (s, 1 H), 7.11–7.65 (m, 12 H), 8.27 (d, J = 8.4 Hz, 1 H)
ppm. 13C NMR (CDCl3): δ = 109.51, 109.59, 119.05, 122.30,
124.72, 124.99, 126.49, 126.73, 127.95, 129.24, 130.76, 131.25,
131.86, 133.68, 136.76, 138.62, 147.49, 153.00 ppm. HRMS: calcd.
for C23H15NO 320.1076 [M – H]; found 320.1063 [M – H].
Synthesis of 3-(p-Chlorophenyl)-3-phenyl-1,2,3,4-tetrahydronaphtha-
lene-1-one (15): A mixture of 5.0 g (14.25 mmol) of 14 and 20.0 g
of polyphosphoric acid, preheated at 110 °C for 30 min, was stirred
with heating at the same temperature for 6 h. After cooling, it was
poured onto cracked ice and extracted with ethyl acetate (3 times).
The combined extracts were dried and concentrated under vacuum
to afford the crude product. Column chromatography (ethyl ace-
tate/hexane, 10:90) furnished 2.6 g (54.8%) of pure 15 as pale yel-
Fraction 3. 4-(p-Cyanophenyl)-3-phenyl-1-naphthol (29b): Yield
44 mg (17.6%), m.p. 208 °C. 1H NMR (CDCl3): δ = 5.654 (s, 1 H),
6.93 (s, 1 H), 7.05–7.57 (m, 12 H), 8.30 (d, J = 8.4 Hz, 1 H) ppm.
low semisolid. IR (CDCl ): ν = 2969, 1694 cm–1. 1H NMR
˜
3
(CDCl3): δ = 3.41 (s, 2 H), 3.74 (s, 2 H) ppm. 7.11–7.507 (m, 12
13C NMR (CDCl3): δ = 110.76, 111.15, 119.29, 122.17, 123.94, H), 7.92 (d, 1 H, J = 7.5 Hz). 13C NMR (CDCl3): δ = 42.42, 48.60,
125.71, 126.16, 127.11, 127.67, 128.23, 130.17, 131.94, 133.04,
51.16, 126.83, 126.96, 127.10, 127.22, 128.702, 128.83, 129.07,
4100
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 4091–4102

Organic Photochemical Rearrangements of Triplets and Zwitterions
FULL PAPER
132.50, 132.56, 134.40, 141.89, 144.53, 145.819, 196.75 ppm.
HRMS: calcd. for C22H17ClO 332.0968; found 332.0957.
Synthesis of 3-Benzyl-2-(p-chlorophenyl)-2-phenylpropanoic Acid
(20): A solution of 1.65 g (5.0 mmol) of 19 and 2.75 g (2.0 mmol) of
KOH was refluxed in 10 mL of ethylene glycol for 5 h. The reaction
mixture was then poured onto ice/water and then acidified with
10% HCl to pH = 2–3. The crude product was filtered and dried
under vacuum. Recrystallization from hexane/diethyl ether (1:1) af-
Synthesis of 3-(p-Cyanophenyl)-3-phenyl-1,2,3,4-tetrahydronaph-
thalen-1-one (16): A solution of 2.0 g (6.0 mmol) of 15 and 2.16 g
(24.0 mmol) of cuprous cyanide in 10 mL of N-methylpyrrolidone
was refluxed with stirring under nitrogen for 24 h. The reaction
mixture was cooled to room temperature, diluted with 10 mL of
10% ammonium hydroxide solution and extracted with 30 mL of
ethyl acetate (3 times). The combined organic layers were washed
with water, dried and concentrated under vacuum. The crude prod-
uct on column chromatography (ethyl acetate/hexane, 1:6) afforded
16 as a pale yellow solid which was crystallized from hexane/diethyl
ether (1:1), yield 580 mg (30.0%), m.p. 223–225 °C (dec.). 1H NMR
(CDCl3): δ = 3.44 (s, 2 H), 3.78 (s, 2 H), 7.15–7.72 (m, 12 H), 7.93
(d, J = 8.1 Hz, 1 H) ppm. 13C NMR (CDCl3): δ = 42.10, 49.23,
50.82, 110.71, 118.70, 126.93, 127.19, 127.45, 128.11, 129.04,
132.45, 132.53, 134.61, 141.37, 144.37, 144.92, 151.45, 196.18 ppm.
HRMS: calcd. for C23H17NO 323.1310; found 323.1318.
1
forded 1.47 g (84.0%) of the pure carboxylic acid, m.p. 141 °C. H
NMR (CDCl3): δ = 2.48–2.52 and 3.147 (m, 1 H), 2.72–2.83 (m, 1
H), 3.52–3.60 (m, 1 H), 3.84–3.91 (t, 1 H), 6.66–6.72 (m, 2 H),
6.89–7.53 (m, 12 H) ppm. 13C NMR (CDCl3): δ = 39.95, 41.90,
50.53, 57.46, 58.01, 126.18, 126.35, 127.71, 128.19, 128.24, 128.27,
128.41, 128.58, 128.61, 128.88, 128.96, 129.13, 129.31, 129.37,
129.87, 130.06, 132.08, 132.65, 136.31, 136.73, 139.01, 139.06,
139.22, 139.96, 178.08, 179.45 ppm. HRMS: calcd. for C22H19ClO2
349.0995 [M – H]–; found 349.0982 [M – H]–.
Synthesis of 3-(p-Chlorophenyl)-2-phenyl-1,2,3,4-tetrahydronaph-
thalen-1-one (21): A mixture of 1.25 g (3.5 mmol) of 20 and 7.0 g
of polyphosphoric acid was heated for 6 h. The reaction mixture
was poured onto ice/water and extracted with ethyl acetate. The
organic layer was dried and concentrated under vacuum to afford
the crude product which was purified by column chromatography
(hexane/ethyl acetate, 9:1) as colorless shining needles, yield 860 mg
(74.1%), m.p. 174 °C. 1H NMR (CDCl3): δ = 3.17–3.24 (dd, J = 4,
16 Hz, 1 H), 3.35–3.450 (m, 1 H), 3.65–3.75 (dt, J = 4, 12 Hz, 1
H), 3.93–3.97 (d, J = 12 Hz, 1 H), 6.95–7.58 (m, 12 H), 8.09–8.10
(d, J = 7.8 Hz, 1 H) ppm. 13C NMR (CDCl3): δ = 47.94, 61.18,
68.37, 127.04, 127.43, 128.17, 128.54, 128.79, 128.80, 129.03,
129.05, 129.46, 132.69, 134.14, 138.40, 141.08, 142.59, 197.88 ppm.
HRMS: calcd. for C22H17ClO 355.0866 [M+ + Na]; found 355.0866
[M+ + Na].
Synthesis of 4-Bromo-3-(p-cyanophenyl)-3-phenyl-1,2,3,4-tetra-
hydronaphthalen-1-one (17): A solution of 484 mg (1.5 mmol) of 16,
264 mg (1.5 mmol) of NBS, 5.0 mg (0.03 mmol) of AIBN in 10 mL
of carbon tetrachloride was refluxed for 5 h. After cooling the reac-
tion mixture to room temperature, the solution was filtered. The
filtrate on concentration under vacuum afforded crude 17 which
was crystallized from hexane/diethyl ether (9:1), yield 450 mg
(74.5%), m.p. 91 °C. 1H NMR (CDCl3): δ = 3.76 (m, 1 H), 3.95
(m, 1 H), 6.29 (m, 1 H), 7.06–7.97 (m, 13 H) ppm. MS: calcd. for
C23H16BrNO – Br 322.3; found 322.1; m/z = 277.1, 244.1, 219.0,
194.0, 167.0, 149.0, 118.0, 83.9, 68.0.
Reaction of 4-Bromo-3-(p-cyanophenyl)-3-phenyl-1,2,3,4-tetra-
hydronaphthalen-1-one (17) with Potassium tert-Butoxide. Genera-
tion of the Zwitterion in the Dark: A solution of 402 mg (1.0 mmol)
of bromo compound 17 in 7.0 mL of tert-butyl alcohol and 112 mg
(1.0 mmol) of potassium tert-butoxide was stirred in dark for 12 h.
The reaction mixture was then poured onto crushed ice and acidi-
fied with 10% HCl to pH = 2–3. The solid which separated was
extracted with ethyl acetate, and the organic phase was dried with
anhydrous sodium sulfate. Column chromatography using hexane
and hexane/ethyl acetate (19:1, 9:1 and 6:1) afforded the following
products:
Synthesis of 3-(p-Cyanophenyl)-2-phenyl-1,2,3,4-tetrahydronaph-
thalen-1-one (22): A mixture of 800 mg (2.4 mmol) of 21 and
883 mg (9.9 mmol) of cuprous cyanide in 10 mL of N-methylpyrrol-
idone was refluxed under nitrogen for 24 h. After cooling to room
temperature, the reaction mixture was diluted with 15 mL of 10%
ammonium hydroxide solution and extracted with ethyl acetate.
The organic phase was washed well with water, dried and concen-
trated under vacuum. The residual semisolid was chromatographed
(hexane/ethyl acetate, 9:1) to afford 201 mg (25.9%) of pure 22,
1
m.p. 192 °C. H NMR (CDCl3): δ = 3.15–3.25 (dd, J = 4, 16 Hz,
1 H), 3.38–3.42 (m, 1 H), 3.77–3.78 (dt, J = 4, 12 Hz, 1 H), 3.96
(d, J = 12 Hz, 1 H), 6.93–7.59 (m, 12 H), 8.10 (d, J = 7.8 Hz, 1 H)
ppm. 13C NMR (CDCl3): δ = 37.41, 48.68, 60.84, 110.99, 118.95,
127.28, 127.64, 18.26, 128.57, 128.68, 128.79, 129.36, 132.53,
134.29, 137.96, 142.036, 147.85, 197.27 ppm. HRMS: calcd. for
C23H17NO 323.1310; found 323.1308.
Fraction 1. 1-(p-Cyanophenyl)-1-phenyl-1a,6a-dihydro-1H-cyclopro-
pa[a]inden-6-one (Diastereomers 10a and 10b): Yield 178 mg
1
(55.4%). Ratio of 10a/10b = 60:40 (by H NMR).
Fraction 2. 3-(p-Cyanophenyl)-4-phenyl-1-naphthol (29a): Yield
114 mg (35.5%).
Synthesis of 2-Bromo-3-(p-cyanophenyl)-2-phenyl-1,2,3,4-tetrahhyd-
ronaphthalen-1-one (23): To a solution of 484 mg (1.5 mmol) of 22
in 10 mL of acetic acid was added dropwise a solution of 120 mg
(1.5 mmol) of bromine in 5.0 mL of acetic acid over a period of
10 min. The reaction mixture was stirred at ambient temperature
for 2 h and then poured onto crushed ice. The yellow product was
filtered and crystallized from hexane as a pale yellow solid, yield
480 mg (79.8%), m.p. 107–109 °C. 1H NMR (CDCl3): δ = 3.07–
3.08 (m, 1 H), 3.53–3.57 (dd, 1 H), 3.58–3.79 (m, 1 H), 6.81–6.84
(d, 2 H), 7.07–7.60 (m, 10 H), 8.27–8.28 (d, 1 H) ppm. 13C NMR
(CDCl3): δ = 32.94, 53.12, 57.22, 110.53, 124.96, 126.69, 126.90,
127.25, 127.69, 128.00, 128.78, 128.82, 129.37, 130.32, 130.48,
133.32, 136.44, 199.13 ppm. MS: calcd. for C23H16BrNO – HBr
321.3; found 321.2; m/z = 291.0 232.1, 187.1 172.2, 149.1, 118.1,
85.1, 71.1.
Fraction 3. 4-(p-Cyanophenyl)-3-phenyl-1-naphthol (29b): Yield
29 mg (9.0%).
Synthesis of 3-Benzyl-2-(p-chlorophenyl)-2-phenylpropanenitrile
(19): To a freshly prepared solution of benzylmagnesium chloride
(12 mmol) in diethyl ether was added 2.39 g (10 mmol) of 18[18] in
10 mL of anhydrous THF over 15 min and the mixture refluxed in
an oil bath for 4 h. The mixture was then cooled, poured onto
crushed ice and the acidified with cold 10% HCl to pH = 2–3.
Column chromatography (ethyl acetate/hexane, 1:19) and crystalli-
zation from hexane afforded 2.4 g (72.5%) of the pure product,
m.p. 134 °C. 1H NMR (CDCl3): δ = 3.11–3.26 (m, 3 H), 4.03–4.05
(d, J = 5.1 Hz, 1 H), 6.92–6.96 (m, 4 H), 7.19–7.33 (m, 10 H) ppm.
13C NMR (CDCl3): δ = 39.84, 42.66, 52.50, 119.18, 127.13, 128.11,
128.32, 128.65, 128.97, 129.05, 129.13, 130.10, 133.68, 134.21,
137.19, 138.62 ppm. HRMS: calcd. for C22H18ClN 331.1128; found
331.1127.
Synthesis of 3-(p-Cyanophenyl)-2-phenylnaphthol (24): A solution of
400 mg (1.2 mmol) of 23, 93 mg (1.2 mmol) of lithium carbonate,
Eur. J. Org. Chem. 2007, 4091–4102
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4101

H. E. Zimmerman, V. Suryanarayan
FULL PAPER
[6] H. E. Zimmerman, R. C. Hahn, H. Morrison, M. C. Wani, J.
Am. Chem. Soc. 1965, 87, 1138–1140.
and 24 mg (0.3 mmol) of lithium bromide in 5.0 mL of dimethyl-
formamide was heated at 100 °C for 4 h. The progress of reaction
was monitored by TLC. When all the starting material was con-
sumed, the reaction mixture was poured onto crushed ice and acidi-
fied with 20% hydrochloric acid. The crude product was filtered,
washed with water and crystallized twice from hexane/diethyl ether
(9:1) as pale yellow crystals, yield 298 mg (77.4%), m.p. 177 °C.
[7] R. B. Woodward, P. Yates, J. Am. Chem. Soc. 1963, 85, 553–
557.
[8] a) D. H. R. Barton, P. De Mayo, M. Shafiq, J. Chem. Soc.
1958, 140–145; b) D. H. R. Barton, P. T. Gilhan, J. Chem. Soc.
1960, 4596–4599.
[9] E. B. Gluz, M. E. Kletskii, Russ. J. Org. Chem. 1995, 31, 882–
886.
Supporting Information (see footnote on the first page of this arti-
cle): 1H and 13C NMR spectra of compounds, methodology and
input files.
[10] H. E. Zimmerman, J. S. Swenton, J. Am. Chem. Soc. 1967, 89,
906–912.
[11] a) H. E. Zimmerman, D. Döpp, P. S. Huyffer, J. Am. Chem.
Soc. 1966, 88, 5352–5353; b) H. E. Zimmerman, D. S. Crum-
rine, J. Am. Chem. Soc. 1968, 90, 5612–5614; c) H. E. Zimmer-
man, D. S. Crumrine, D. Döpp, P. S. Huyffer, J. Am. Chem.
Soc. 1969, 91, 434–445; d) T. H. Brennan, R. K. Hill, J. Am.
Chem. Soc. 1968, 90, 5614–5615.
Acknowledgments
Support of this research by the National Science Foundation is
gratefully acknowledged with special appreciation for its support
of basic research.
[12] H. E. Zimmerman, J. O. Grunewald, J. Am. Chem. Soc. 1967,
89, 5163–5172.
[13] a) J. Michl, V. Bonancic-Koutecky, Tetrahedron 1988, 44, 7559–
7585; b) V. Bonancic-Koutecky, J. Koutecky, J. Michl, Angew.
Chem. Int. Ed. Engl. 1987, 26, 170–189; c) A. R. Matlin, P. M.
Lahti, D. Appella, A. Straumanis, S. Lin, H. Patel, K. Jin, K. P.
Schreiber, J. Pauls, P. Raulerson, J. Am. Chem. Soc. 1999, 121,
2164–2173; d) Electronic Aspects of Organic Photochemistry
(Eds.: J. Michl, V. Bonancic-Koutecky), Wiley, New York,
1990, pp. 174–175; e) P. V. Alston, D. D. Shillady, C. Trindle,
J. Am. Chem. Soc. 1975, 97, 469–476.
[14] a) D. C. C. Smith, P. H. Lacy, J. Chem. Soc. C 1971, 1, 41–43;
b) H. H. Szmant, R. Nanjundiah, J. Org. Chem. 1978, 43,
1835–1837.
[15] L. Minuti, A. Taticchi, A. Marrocchi, D. Lanari, A. Broggi, E.
Gacs-Baitz, Tetrahedron: Asymmetry 2003, 14, 481–487.
[16] R. W. R. Humphreys, D. R. Arnold, Can. J. Chem. 1977, 55,
2286–2291.
[1] a) H. E. Zimmerman, Seventeenth National Organic Sympo-
sium of the American Chemical Society, Abstracts, Bloom-
ington, Indiana, 1961, pp. 31–41; b) H. E. Zimmerman, D. I.
Schuster, J. Am. Chem. Soc. 1962, 84, 4527–4540; c) H. E. Zim-
merman, in Advances in Photochemistry (Eds.: W. A. Noyes, Jr.,
G. S. Hammond, J. N. Pitts, Jr.), Interscience, New York, 1963,
vol. 1, pp. 183–208.
[2] a) H. E. Zimmerman, J. Nasielski, R. Keese, J. S. Swenton, J.
Am. Chem. Soc. 1966, 88, 4895–4903; b) H. E. Zimmerman,
J. O. Grunewald, J. Am. Chem. Soc. 1967, 89, 3354–3356.
[3] a) I. Gomez, O. Santiogo, M. Reguero, A. Riera, A. Sole, J.
Am. Chem. Soc. 2002, 124, 15375–15384; b) I. Gomez, M. Re-
guero, Mol. Phys. 2006, 104, 889–901; c) M. Reguero, F. Ber-
nardi, M. Olivucci, M. A. Robb, J. Org. Chem. 1997, 62, 6897–
6902.
[17] A. R. Katritzky, S. K. Nair, R. M. Witek, S. M. Hutchins, AR-
KIVOC 2003, 5, 9–18.
[18] S. Wawzonek, E. M. Smolin, Org. Synth., Coll. Vol. 1953, III,
715.
[4] F. Bergmann, J. Szmuszkovic, J. Am. Chem. Soc. 1946, 68,
1662.
[5] C. R. Hauser, M. T. Tetenbaum, J. Org. Chem. 1958, 23, 233–
235.
Received: May 21, 2007
Published Online: July 3, 2007
4102
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 4091–4102