The adverse effect of benzannelation on the aromaticity of oxocinyl anion: A combined experimental and theoretical study

Article abstract of DOI:10.1002/hlca.200900430

A synthesis of 2H-1-benzoxocin from readily available compounds was accomplished. The potentially 'aromatic' π-excessive systems 2H-1-benzoxocinyl and 6H-dibenz[b,f]oxocinyl anions were generated from their corresponding conjugate acid precursors 7c and 8, respectively. It was found that 2H-1-benzoxocinide 3d lacks the type of π-frame stability associated with the parent 2H-oxocinide 1d and that the dibenzo analog 5b is more unstable than 3d. Both 3d and 5b undergo rapid structural reorganization to form their corresponding stable isomeric anions. We were able to characterize the proton-quenched products of these anions as the ring-opened structures 15 and 18, respectively. ¹H-NMR and an ab initio calculation at the 6-31g* level indicated that, unlike the 'aromatic' parent 2H-oxocinide 1d and the aza analog 3c, 3d incorporates a non-planar oxocinyl ring in which the negative charge is primarily localized on the pentadienyl moiety of the ring, but also partial delocalization of π-electron density onto the benzene ring occurs.

Full text of DOI:10.1002/hlca.200900430

1532
Helvetica Chimica Acta – Vol. 93 (2010)
The Adverse Effect of Benzannelation on the Aromaticity of Oxocinyl Anion:
A Combined Experimental and Theoretical Study
by Hamid S. Kasmai*a), Xiaodong Wanga), Hanh-Nhon Doana), Josef F. Femiab), Thad M.
Yablonskyb), Mary E. Readb), and David T. Duttonb)
a) Department of Chemistry, East Tennessee State University, Johnson City, TN 37614, USA
(phone: þ 1-423-4394367; fax: þ 1-423-4395835; e-mail: kasmai@etsu.edu)
b) Department of Chemistry, Hamilton College, Clinton, NY 13323, USA
A synthesis of 2H-1-benzoxocin from readily available compounds was accomplished. The
potentially ꢀaromaticꢁ p-excessive systems 2H-1-benzoxocinyl and 6H-dibenz[b,f]oxocinyl anions were
generated from their corresponding conjugate acid precursors 7c and 8, respectively. It was found that
2H-1-benzoxocinide 3d lacks the type of p-frame stability associated with the parent 2H-oxocinide 1d
and that the dibenzo analog 5b is more unstable than 3d. Both 3d and 5b undergo rapid structural
reorganization to form their corresponding stable isomeric anions. We were able to characterize the
proton-quenched products of these anions as the ring-opened structures 15 and 18, respectively. 1H-NMR
and an ab initio calculation at the 6-31g* level indicated that, unlike the ꢀaromaticꢁ parent 2H-oxocinide
1d and the aza analog 3c, 3d incorporates a non-planar oxocinyl ring in which the negative charge is
primarily localized on the pentadienyl moiety of the ring, but also partial delocalization of p-electron
density onto the benzene ring occurs.
Introduction. – The synthesis and study of aromatic and heteroaromatic compounds
constitute a fundamentally significant part of basic research in the field of organic
chemistry. In addition to the importance of these studies in the development of basic
theoretical and experimental concepts [1], the applied areas such as material sciences
[2] and biomedical chemistry [3] have found practical applications of this field of
research.
Our work on the study of heteroaromatic compounds has focused on the generation
and characterization of benzo-fused analogs of bis-p-excessive heterocinyl anions 1b –
1d, which are iso-p-electronic with the aromatic 10p-electron cyclooctatetraene
dianion 1a [4] (Fig. 1).
In 1987, Prinzbach and co-workers [5] reported that the 4-cyano-substituted analogs
2a – 2c are also planar aromatic 10p anions. Theoretically, the development of ꢀaromatic
characterꢁ [6] in the (4n þ 2)p ribbons of 1a – 1d and 2a – 2c may be discussed in terms
of structural demand imposed by p-electron mobility on the corresponding eight-
membered rings, i.e., in terms of factors that primarily influence the transformation of
the flexible molecular frame from buckled to flat, energetics of which are governed by
the well-known free-energy relationship (Eqn. 1).
DG0 (B ! P) ¼ DHp0 þ DH0s ꢀ T DS0.
(1)
ꢂ 2010 Verlag Helvetica Chimica Acta AG, Zꢃrich

Helvetica Chimica Acta – Vol. 93 (2010)
1533
Fig. 1. Heterocinyl anions
Within a specified direction of change, i.e., buckled (B) to planar (P), DHp0 (a negative
quantity for (4n þ 2)p systems) denotes the stabilizing influence of electron delocal-
ization, DHs0 (a positive quantity) represents the destabilizing effect of sigma strain,
and DS0 (a negative quantity) describes the loss of molecular flexibility attending the
change from buckled to planar. In the case of the five-membered-ring heterocycles
pyrrole and furan which are skeletally less flexible and strain-free, the DHp0 term is the
controlling factor in the development of the aromatic character. For the eight-
membered rings in 1a – 1d and 2a – 2c, the effects of the DHp0 and DH0s ꢀ T DS0 terms
oppose each other, and the former has to exert a controlling influence for the system to
develop a planar delocalized frame.
In a closer examination, one observes that, while the extent of the development of
ꢀaromatic characterꢁ in the five-membered-ring heterocycles follows the decreasing
order of pyrrole > furan due to the electronegativity effect of the heteroatom [7], the
extent of heteroatomic lone-pair participation in cyclic delocalization in heterocinyl
anions 2a – 2c does not appear to play a significant role in the development of
aromaticity. Thus 2c, carrying the electronegative heteroatom O sustains a diatropic
character as does 2a, which contains the electron-rich-alkyl substituted heteroatom N.
This phenomenon may be explained in terms of Eqn. 1, where the stabilizing influence
of (4n þ 2)p-electron delocalization (DHp0 ) in 2a – 2c more than compensates for the
combined destabilization introduced by the electron-repulsion component, skeletal
strain, and reduced molecular flexibility (DHs0 ꢀ T DS0).
We have been interested in studying the effect of further perturbation caused by
benzo-fusion on the development of aromaticity in eight-membered-ring heterocinyl
anions. In the case of the cyclooctatetraene dianion 1a, it has been documented that the
p-electron delocalization persists under the adverse effect of benzo-fusion, and in fact
benzocyclooctene dianion 3a [8] and the dibenzo analog 4a [9] (Fig. 2) were found to
be 14p- and 18p-electron diatropic systems, respectively. On the other hand, in the case
of anions 1b – 1d, the persistence of the aromatic character under the benzo-fusion is
expected to modulate with the electronegativity of the heteroatomic center. In fact, our
earlier studies indicate that while the monobenzo members 3b and 3c are planar
diatropic anions [10], in the dibenzo series [10], only 4b containing an N-atom with a
higher lone-pair mobility is found to have a flattened central ring with a diatropic
character, whereas the N-methyldibenzoazocinyl anion 5a [11] distinctly lacks the type
of p-frame stability normally associated with ꢀaromaticꢁ systems.

1534
Helvetica Chimica Acta – Vol. 93 (2010)
Fig. 2. Benzo- and Dibenzoheterocinyl Anions 3 – 6
The central question of how does the 10p-electron oxocinyl anion 1d compare with
benzene in terms of p-electron delocalization, and what effect will the perturbation of
benzo-fusion have on its p-delocalization largely remains unanswered to date. We have
shown that a nonlinear dibenzo-fused derivative 6 does not contain a diatropic central
8-membered ring [12]. The lack of aromacity in 6 is presumably due to a periphery-
induced (ꢀperiꢁ H-H) [13] skeletal inability of the central ring to adopt a planar
geometry.
In this work, we describe the synthesis of a conjugate-acid precursor of the anion
2H-1-benzoxocinide (3d), namely 2H-1-benzoxocin (7c) and the generation and NMR
characterization of 3d. The generation of a linearly dibenzo-fused oxocinyl anion,
namely 6H-dibenz[b,f]oxocinide (5b), from its precursor 8 [14] (Fig. 3), will also be
included, and the chemical reactivities of anions 3d and 5b will be discussed. Finally, the
results of an ab initio calculation at the 6-31g* level on the anions 1c and 1d, and on the
benzo-fused analogs 3c and 3d along with the corresponding conjugate-acid precursors
shown in the Table (cf. below) will be presented and discussed.
Fig. 3. 2H-1-Benzoxocin-2-ide (3d), 2H-1-benzheterocins 7a – 7c, and 6H-dibenz[b,f]oxocin (8)
Results and Discussion. – Syntheses. The synthesis of 2H-1-benzoxocin (7c) is
shown in Scheme 1. The preparation of the key intermediate 12 is based on the
procedures prescribed by Schweizer and co-workers [15] with some modifications to
improve the yields. In the reported procedures, the phosphonium salt 11 (obtained from
9 via 10) was treated with EtONa in DMF to yield 12 along with an isomeric compound
(95 :5 ratio) in 19% yield. In our procedure, 12 was obtained in 54% yield by slow
addition of the slurry of excess t-BuOK in DMF to a refluxing dilute solution of 11. It is
noteworthy to point out that a dihydro derivative of 12, namely, (þ)-helianane [16a], a
marine natural product (synthesis reported in 1974 [16b]), as well as the sunflower-
derived close relative heliannuol A [17] have been investigated for their bioactivity.

Helvetica Chimica Acta – Vol. 93 (2010)
Scheme 1. Synthesis of 2H-1-Benzoxocin (7c)
1535
a) Br(CH2)4Br, OHꢀ. b) Ph3P, AcOEt. c) t-BuOK, DMF, reflux. d) NBS, CCl4, hn. e) DBU, NaI, MeCN, 08.
Allylic bromination of 12 with N-bromosuccinimide (NBS) under light catalysis
yielded 13 in a quantitative yield, and chromatography at ꢀ 208 afforded pure 13 in
83% yield. The elimination of HBr from 13 was best achieved by treatment with 1,8-
diazabicyclo[5.4.0]undec-7-ene (DBU) in the presence of NaI in MeCN at 08. The
crude product was chromatographed at ꢀ 208 to afford pure 7c in 34% yield. This
compound is thermally unstable at room temperature and had to be stored in a freezer.
The synthesis of 6H-dibenz[b,f]oxocin (8), the precursor for anion 5b, was
accomplished according to the published method [14].
Generation and Characterization of Anions. 2H-1-Benzoxocin-2-ide (3d). Exposure
of the conjugate-acid precursor 7c to KNH2/NH3 (liq.) at ꢀ 788 followed by warming to
08, which was necessary to obtain a homogeneous solution for the NMR measurement,
1
resulted in the formation of an intensely colored solution. The recorded H-NMR
spectrum (Fig. 4) was consistent with the structure of anion 3d.
1
To assess the diatropicity of 3d, we compared the key features of the H-NMR
spectrum of 3d with those of 2c and 3c and with those of their corresponding conjugate
acids 7c, 14, and 7b, respectively (Fig. 4). Notable in these comparisons are:
1) On passing from the conjugate-acid progenitors to the corresponding anions, the
observed NMR-chemical-shift changes (in ppm) i) for the H-atoms of the former sp3 C-
atoms (2c (HꢀCc): Dd ¼ þ2.31; 3c (HꢀCa): Dd ¼ þ3.4; 3d (HꢀCa): Dd ¼ ꢀ0.31) and
ii) for the key HꢀCa of 2c Dd ¼ þ0.62) and HꢀCc of 3c (Dd ¼ ꢀ0.02) and 3d (Dd ¼
ꢀ2.9) are indicative of substantial localization of the negative charge on the CaꢀCc
allylic section of the pentadienyl moiety of the eight-membered ring in 3d, in contrast to
2c [6] and 3c [10], which were recognized as delocalized 10p- and 14p-electron systems,
respectively.
2) While the average chemical shift of the ring H-atoms in the parent anion 2c (d
6.1) is 0.15 ppm higher than the corresponding H-atoms of the conjugate-acid precursor
(d 5.96), in the case of 3d, the average chemical shift of the eight-membered-ring H-
atoms (d 5.11) is 0.80 ppm lower than that of 7c (d 5.91 for 4 olefinic H-atoms). The
comparison of the average chemical shift of the eight-membered-ring H-atoms indicate
that, while in 2c the expected high-field shift of the H-signals as a result of higher
electron density in the anion, compared to its conjugate acid, is offset by a diamagnetic-
ring-current effect, such an effect does not appear to be present in 3d.
3) In contrast to the N-analog 3c, where there is a downfield shift of NMR signals
for some of the benzo H-atoms [10] on the transformation of 7b to 3c, we observed an

1536
Helvetica Chimica Acta – Vol. 93 (2010)
Fig. 4. 1H-NMR Data (d in ppm, J in Hz) for Anions 2c, 3c, 3d, and the corresponding conjugate-acid
precursors 14, 7b, and 7c. For convenience, the H-atoms at the cyclic structures are represented by their
bonds (without attached H).
upfield shift of these signals in going from 7c (d 6.96 – 7.13) to 3d (d 6.18 – 6.68). This
upfield shift of aromatic H-atoms is typical of systems such as diphenylmethanide [18]
and the atropic dibenzo member of the oxocinide family, namely anion 6 [12]. Thus, the
NMR data presented above are not consistent with a diamagnetic 10p- or 14p-system,
and 3d must largely be regarded as a conventional system, i.e., one in which the
negative charge is primarily localized in the pentadienyl moiety of the eight-membered
ring, but also partial delocalization of p-electron density onto the benzo moiety has
occurred.
From a chemical view point, it is noteworthy that anion 3d is thermally less stable
than the cyano-substituted parent anion 2c (stable at room temperature for several
hours [6]). This instability did not allow us to obtain a clean 13C-NMR spectrum of 3d
nor was it possible to obtain clean identifiable product(s) from the H2O quench
experiments at 08.
Specifically, when the NMR solution of 3d was gradually warmed from 08 to ca. 308
(NMR-probe temp.), a gradual change in the observed NMR spectrum occurred right
after a few scans. Continued recording of the spectrum led to one which did not exhibit
any quantitative or qualitative change at probe temperature. This transformation of
anion 3d to another stable species occurred with an estimated half-life of 20 min.
1
Although from the recorded H- and 13C-NMR spectra we were not able to assign a

Helvetica Chimica Acta – Vol. 93 (2010)
1537
structure for the rearranged anion, we could identify the product isolated after
quenching with H2O as compound 15 (38% yield), a pale yellow solid (Scheme 2).
Scheme 2. Generation, Rearrangement, and H2O-Quench Product of 3d
a) KNH2, liq. NH3, ꢀ 788. b) 08. c) 08 ! ca. 308. d) H2O, Et2O.
The characterization of 15 was achieved by careful analysis of its mass spectrum as
well as 1H- and 13C-NMR and IR spectra. The key features of the NMR and IR spectra
included: 1) In the 1H-NMR, the high-field d at d 4.8 (J(He,He) ¼ 6.6 Hz) is assigned to
2 HꢀCe. HꢀCc appears as a m at d 5.85 – 6.25, and signals due to Ha and Hb appear as
ms along with aromatic H-atoms at d 6.49 – 7.56. The OH H-atom resonance coincides
with that of the 2 HꢀCe at d 4.8. Addition of the shift reagent [EuIII(fod)3] causes a
downfield shift of the OH signal as expected. The chemical shifts of Ha to He, however,
remain unchanged, presumably due to the large steric effect [19] present in the
resulting complex with the shift reagent. 2) In the 13C-NMR spectrum, the resonances
at d 77.36 (t, J(C,H) ¼ 168.3 Hz) and 95.15 (d, J(C,H) ¼ 166.0 Hz) are assigned to Ce
and Cc, respectively. The Ca and Cb resonances are observed at d 116.5 (d, J(C,H) ¼
155.8 Hz) and/or 121.83 (d, J(C,H) ¼ 156.3 Hz), and the allenic Cd gives rise to a
typical downfield-shifted signal at d 214.8 [20]. 3) In the IR spectrum, frequencies at
1928 (C¼C¼C stretching) and 858 cmꢀ1 (¼CH2 bending) are clear indications of the
presence of an allene functionality [21] in 15. The proposed (E) configuration of the
olefinic bond in 15 also follows from the IR absorption at 969 cmꢀ1 due to the out-of-
plane deformation of the trans-positioned CꢀH bonds [21].
It is noteworthy that the thermal reorganization of 3d, as described above, took a
different pathway compared to that of the more stable diatropic N-atom counterpart
3b, which occurred under the same condition according to Scheme 3 [10].
Scheme 3. Rearrangement of 1-Methyl-1-benzazocin-2(1H)-ide Anion (3b)
Finally, to complement our experimental results on the nature of anion 3d, we
carried out an ab initio MO calculation at the restricted Hartree – Fock SCF level by
using a 6-31 G* basis set [22] on the conjugate-acid precursors 1d · Hþ, 7b, and 7c and
the corresponding anions 1d, 3c, and 3d, respectively. We might add that i) this level of
quantum-mechanical calculation has been assessed as good for obtaining the
equilibrium geometries of medium-sized organic molecules containing heteroatoms
[23], and ii) in the discussion of the calculated bond-length alternation as a structural

1538
Helvetica Chimica Acta – Vol. 93 (2010)
criterion of aromaticity [24a] (see below), we note that according to Kertesz and co-
workers [24b], in general, the Hartree – Fock calculations tend to overestimate the
degree of bond-length alternation in both conjugated molecules and extended systems.
Our calculated structural parameters for the above-mentioned molecules are
provided in the Table. The numbering of atoms and angles referred to therein as well as
the calculated Mulliken atomic charges [25] are shown in Fig. 5.
Focusing on the crucial structural requirements of aromaticity in cyclic systems,
namely, i) achieving a planar geometry which allows maximum delocalization of p-
electrons, and ii) achieving bond equalization in the ring, our calculated results clearly
indicate that 1) the 2H-oxocin ! 2H-oxocinide transition is accompanied by a
geometry transformation from a half-chair-like geometry to a planar one (Fig. 6),
Table. Calculated Structural Parametersa) for 14, 1d, 7b, 3c, 7c, and 3db)
X ¼ O
1d · Hþ
X ¼ NH
7b
X ¼ O
7c
1d
3c
3d
XꢀC(1)
1.336
1.347
1.348
1.424
1.383
1.383
1.424
1.348
1.347
–
–
–
–
–
133.6
135.2
138.9
131.9
134.6
131.9
138.9
135.2
–
1.394
1.347
1.467
1.426
1.401
1.380
1.418
1.382
1.332
1.436
1.365
1.411
1.367
1.434
1.351
1.366
1.412
1.438
1.394
1.380
1.438
1.329
1.373
1.416
1.379
1.387
1.388
1.380
C(1)ꢀC(2)
1.330
1.476
1.329
1.470
1.323
1.504
1.414
–
–
–
–
–
1.404
1.496
1.326
1.474
1.322
1.511
1.450
1.401
1.375
1.388
1.373
1.405
1.399
1.493
1.327
1.470
1.322
1.503
1.408
1.406
1.373
1.390
1.372
1.399
C(2)ꢀC(3)
C(3)ꢀC(4)
C(4)ꢀC(5)
C(5)ꢀC(6)
C(6)ꢀC(7)
C(7)ꢀX
C(2)ꢀC(8)
C(8)ꢀC9)
C(9)ꢀC(10)
C(10)ꢀC(11)
C(11)ꢀC(1)
C(1)ꢀXꢀC(7)
119.6
125.1
135.4
123.8
123.7
XꢀC(7)ꢀC(6)
C(7)ꢀC(6)ꢀC(5)
C(6)ꢀC(5)ꢀC(4)
C(5)ꢀC(4)ꢀC(3)
C(4)ꢀC(3)ꢀC(2)
C(3)ꢀC(2)ꢀC(1)
C(2)ꢀC(1)ꢀX
113.1
124.2
125.1
126.6
130.9
134.5
133.0
–
115.4
123.8
122.9
126.0
130.1
128.9
126.8
162.7
ꢀ 54.2
89.3
140.0
134.5
131.5
136.0
137.4
131.4
133.9
180.0
ꢀ 0.21
0.21
113.0
123.0
123.3
126.4
131.3
130.2
128.9
ꢀ 175.9
68.7
123.7
130.1
129.0
138.5
134.0
127.2
120.9
ꢀ 103.3
–90.0
8.3
23.6
3.8
ꢀ 6.2
ꢀ 22.0
ꢀ 3.8
82.2
C(11)ꢀC(1)ꢀXꢀC(7)
C(1)ꢀXꢀC(7)ꢀC(6)
XꢀC(7)ꢀC(6)ꢀC(5)
C(7)ꢀC(6)ꢀC(5)ꢀC(4)
C(6)ꢀC(5)ꢀC(5)ꢀC(3)
C(5)ꢀC(4)ꢀC(3)ꢀC(2)
C(4)ꢀC(3)ꢀC(2)ꢀC(1)
C(3)ꢀC(2)ꢀC(1)ꢀX
C(2)ꢀC(1)ꢀXꢀC(7)
C(8)ꢀC(2)ꢀC(3)ꢀC(4)
73.6
0.13
ꢀ 92.6
ꢀ 0.07
ꢀ 93.0
2.8
0.08
ꢀ 1.3
ꢀ 51.1
ꢀ 9.9
54.4
0.11
1.5
43.7
4.9
ꢀ 39.9
ꢀ 1.74
ꢀ 1.68
–
ꢀ 0.03
ꢀ 0.08
ꢀ 0.03
ꢀ 0.08
0.00
ꢀ 0.11
ꢀ 0.2
0.2
0.0
0.0
46.9
9.4
ꢀ 48.0
ꢀ 2.9
6.5
4.53
ꢀ 21.1
–
ꢀ 126.5
ꢀ 179.8
132.9
161.6
a) Bond length [ꢄ], angles [8], and dihedral angle [8], calculated by ab initio Hartree – Fock SCF with the
6-31 G* basis set. b) See Fig. 5 for structures, atom numbering, and definitions of angles.

Helvetica Chimica Acta – Vol. 93 (2010)
1539
Fig. 5. Mulliken atomic-charge distribution a) of 1d · Hþ and 1d, and b) of 7b, 3c, 7c, and 3d. Arbitrary
atom numbering.
that 2) in the benzo-fused series, while 3c is nearly planar, the oxygen analog 3d
contains a fairly buckled eight-membered-ring moiety with the O-atom lying out-of-
plane of the molecule (Fig. 6), and that 3) benzo-fusion of 2H-oxocin does not affect
the bond lengths of the oxocin ring significantly (the average bond-length alternation
among eight bonds of the ring changes by 0.002 ꢄ on going from 1d · Hþ to 7c), while
the conversion of 1d to 3d is accompanied by a significant bond alternation in the
oxocinide moiety (the average bond-length alternation of the eight-membered ring
changes by 0.010 ꢄ).
Fig. 6. Calculated equilibrium geometries of anions 1d, 3c, and 3d, and their respective conjugate acids
1d · Hþ, 7b, and 7c

1540
Helvetica Chimica Acta – Vol. 93 (2010)
We have also determined the charge distribution in 3c and 3d in the form of
electrostatic-potential (ESP) surfaces (Fig. 7). While the calculated electronic
structure indicates a substantially uniform charge distribution over the azocinyl moiety
in 3c, with some of the extra negative charge remaining on the anionic N-atom, there is
significant electron density remaining on the sp2-hybridized O-atom in 3d, and partial
charge distributions on the pentadienyl moiety as well as on the benzo moiety are
indicated, which are all consistent with our NMR studies discussed above.
Fig. 7. Electrostatic-potential surfaces of 3c (left) and 3d (right) at the 0.002 electron a.u.ꢀ3 density iso-
contour levels of ꢀ 966.9 to ꢀ 511.6 kJ molꢀ1 and ꢀ 596.3 to ꢀ 210.8 kJ molꢀ1, respectively
In conclusion, based on the magnetic criteria [26] (1H-NMR chemical-shift studies)
as well as the calculated structure (geometry and bond-alternation criteria [26])
discussed above, we conclude that the anion 2H-benzoxocinide (3d) lacks complete p-
electron delocalization over the periphery of the eight-membered-ring moiety, and
therefore, it may be classified as containing a nonaromatic (atropic) oxocinide ring. We
attribute this loss of aromaticity in the oxocin moiety to the benzo-fusion, which
perturbs the system to an extent that the heteroatom (O) electronegativity becomes a
deciding factor in development of a completely delocalized oxocinide anion.
6H-Dibenz[b,f]oxocin-6-ide (5b). Exposure of the conjugate-acid precursor 8
dissolved in (D8)THF to KNH2/NH3 (liq.) at ꢀ 788 followed by warming to 08 resulted
in an intensely colored solution. An immediate 1H-NMR scan revealed the presence of
an anion other than the expected anion 5b. The observed spectrum did not change
qualitatively or quantitatively with time indicating the presence of a stable species,
which was characterized as the rearranged dianion 17 on the basis of 1H-NMR
absorptions (a series of ms at d 6.0 – 7.5) and its proton-quenched product 18. In an
attempt to confirm the formation of 5b prior to its rearrangement, 8 was treated with
KNH2/NH3 (liq.) at ꢀ 788 in THF, and the resulting mixture was warmed to ꢀ 358,
stirred at this temperature for 1 h, and then quenched by dropwise addition of MeOH/
Et2O at ꢀ 358. The isolated product was characterized to be exclusively compound 18
(Scheme 4).
Characterization of 18, a yellow solid, was established by its mass spectrum (m/z
208 (100%, Mþ)), 1H-NMR (signals due to aromatic H-atoms, a s at d 3.86 assigned to
CH2 H-atoms, and a broad absorption at d 5.1 – 5.6 due to the OH group, which
disappeared in the presence of added D2O), 13C-NMR (absorptions due to a benzene
ring plus a t at d 40.93 with J(C,H) ¼ 125.3 Hz due to the CH2 group), and IR spectrum
(3519 cmꢀ1 (OꢀH stretching) and 2922 cmꢀ1 (CH2 stretching)).
Mechanistically, the reorganization of 5b to 17 is quite likely intermediated by
formation of compounds 19, 20, and 21 as shown in Scheme 4 and is expected to be

Helvetica Chimica Acta – Vol. 93 (2010)
1541
Scheme 4. Generation and Thermal Reorganization of Anion 5b to 17 and Formation of 18
a) KNH2, liq. NH3, THF, ꢀ 788. b) ca. 358. c) MeOH, Et2O, ꢀ 358.
driven by the formation of the aromatically stabilized indenyl anion moiety present in
17. This proposed mechanism carries credence from the observed rearrangement of the
benzazocinyl anion 3b shown in Scheme 3 above.
It is worth noting that in the case of the more stable dibenzazocinyl analog 5a [11],
the rearrangement took a distinctly different pathway as shown in Scheme 5.
Scheme 5. Rearrangement of 5a [11]
In conclusion, whereas the failure to directly observe the desired anion 5b precludes
the determination of the presence or absence of ring diatropicity, the observed rapid
rearrangement into 17 indicates that: i) The linearly dibenzo-fused oxocinyl anion 5b
clearly is less stable than the monobenzo-fused anion 3d; ii) 5b also lacks the type of p-
frame stability normally associated with ꢀaromaticꢁ systems, and iii) the linearly
dibenzo-fused oxocinyl anion is less stable than the nonlinear analog 6 [12].
Experimental Part
General. All chemicals were purchased from Aldrich or Fischer Scientific and used without
purification unless stated otherwise. DMFand MeCN were refluxed over CaH2 and distilled prior to their
use. THF was refluxed over Na and benzophenone until a blue-violet color persisted and distilled before
use. Compounds 11 [15] and 12 [15] were prepared according to the previously published procedures:
Column chromatography ¼ CC. IR Spectra: Matteson-Polaris FT-IR spectrometer; ˜n in cmꢀ1. 1H- (90 and
500 MHz) and 13C-NMR (22.5 and 125 MHz) Spectra: JOEL-FX-90Q and Bruker-DRX-500 spectrom-
eters, resp.; CDCl3 solns.; d in ppm rel. to Me4Si as internal standard, J in Hz; spectra of anions were
recorded in liq. NH3 containing (D8)THF as an internal standard, and the d(H) and d(C) were measured
from the central signal of the lowfield m of (D8)THF (d(H) 3.70 and d(C) 67.9 with respect to SiMe4) and

1542
Helvetica Chimica Acta – Vol. 93 (2010)
reported rel. to Me4Si. MS: JOEL sector-type spectrometer; in m/z (with the lowest isotope mass) (rel.
%). The theoretical calculations were carried out with the program Spartan ꢁ06 [27].
3,4-Dihydro-2H-1-benzoxocin (12) [15]. Modified Procedure: To a refluxing soln. of phosphonium
bromide 11 (2.0 g, 3.85 mmol) in dry DMF (20 ml) under N2 was added a slurry of t-BuOK (0.60 g,
5 mmol) in DMF (13 ml) within 3.0 h. After reflux for an additional hour, the mixture was cooled to r.t.
and poured into dist. H2O (15 ml). The org. layer was extracted with Et2O (3 ꢁ 50 ml) and the combined
Et2O layer washed with H2O (50 ml), dried (anh. MgSO4), and concentrated. The obtained sticky white
solid (1.48 g) was triturated with AcOEt/petroleum ether (PE) 10 :90 which dissolved 0.427 g of the
crude mixture. The decanted soln. was subjected to CC (25 ꢁ 2 cm column, SiO2 (230 – 400 mesh), packed
and eluted with Et2O/PE 10 :90): 12 (309 mg, 50%). Pale oil. 1H-NMR: identical with the published one
[15].
4-Bromo-2,3-dihydro-2H-1-benzoxocin (13). To a soln. of 12 (0.50 g, 3.12 mmol) in freshly dist. CCl4
(3.7 ml) was added N-bromosuccinimide (NBS; 0.61 g, 3.43 mmol). The resulting soln. was heated to
reflux under N2 and irradiated with direct light from a 200 W light bulb for 5.5 h. The mixture was cooled
to r.t., the precipitated succinimide filtered off, and the filtrate evaporated at 08 to afford 720 mg (96%)
of a yellowish oil. The oil was subjected to CC (jacketed column, SiO2 (30 g, 230 – 400 mesh) wet-packed
at 08 under N2; the column temp. was lowered to ꢀ 208 before elution with Et2O/PE 5 :95. Three fractions
(20 ml), each collected at 08 and under N2 were combined and evaporated at 08: 13 (597 mg, 83%).
Colorless oil. 1H-NMR (90 MHz, CDCl3): 2.24 (m, 2 H); 4.20 (m, 1 H); 4.50 (m, 1 H); 5.02 (m, 1 H); 5.96
(dd, J ¼ 12.2, 5.1, 1 H); 6.27 (d, J ¼ 12.2, 1 H); 7.1 (m, 4 H). 13C-NMR (22.5 MHz, CDCl3): 33.3 (t, J ¼
131.7); 49.8 (d, J ¼ 150.2); 70.9 (t, J ¼ 144.5); 122.4 (d, J ¼ 166.0); 123.7 (d, J ¼ 168.4); 125.6 (d, J ¼
157.5); 128.1(s); 129.8 (d, J ¼ 168.4); 131.0 (d, J ¼ 167.2); 131.8 (d, J ¼ 166.0); 154.9 (s).
2H-1-Benzoxocin (7c). A mixture of 13 (180 mg, 0.75 mmol), 1,8-diazabicyclo[5.4.0]undec-7-ene
(DBU; 230 mg, 1.51 mmol), and NaI (900 mg, 6.0 mmol) in freshly dist. MeCN (11 ml) was stirred at 08
under N2 for 8 h. H2O (20 ml) was added to the resulting mixture, the aq. layer extracted with Et2O (2 ꢁ
15 ml), the combined Et2O layer successively washed with 5% H2SO4 soln. (15 ml), 5% NaHCO3 soln.
(15 ml), aq. sat. NaCl soln. (15 ml), and H2O (15 ml), dried (MgSO4), and concentrated at 08: 80 mg
(67%) of a yellow oil. The crude oil was subjected to CC (jacketed column, SiO2 (25 g, 230 – 400 mesh)
wet-packed with Et2O/PE 5 :95 at 08 under N2 and eluted with the same solvent at ꢀ 208): 7c (40 mg,
34%). Colorless oil. 1H-NMR (500 MHz, CDCl3): 4.72 (d, J ¼ 3, 2 H); 5.62 (dt, J ¼ 11.6, 3, 1 H); 5.81 (dd,
J ¼ 11.6, 6.1, 1 H); 5.94 (dd, J ¼ 12.2, 6.1, 1 H); 6.25 (d, J ¼ 12.2, 1 H); 6.95 (d, J ¼ 6.7, 1 H); 6.96 (dd, J ¼
6.7, 7.3, 1 H); 7.05 (d, J ¼ 7.9, 1 H); 7.13 (dd, J ¼ 7.9, 7.3, 1 H). 13C-NMR (125 MHz, CDCl3): 75.7 (1 C);
116.1 (1 C);116.4 (1 C); 120.2 (1 C); 121.2 (1 C); 123.9 (1 C); 124.3 (1 C); 126.6 (1 C); 129.3 (1 C); 136.6
(1 C); 153.1 (1 C). MS: 158 (100, Mþ).
Generation of Anion 3d. A medium-wall 9 in. NMR tube containing KNH2 in liq. NH3 (prepared
from K (80 mg, 2.0 mmol), liq. NH3 (0.5 ml), and a small crystal of FeCl3) maintained at ꢀ 788 was
charged with a soln. of 7c (90 mg, 0.57 mmol) in (D8)THF (ca. 0.5 ml). A glass-wool plug was pushed
about half way down into the tube, and the tube was carefully sealed under room pressure. The tube then
was placed in a dry ice/acetone bath at ꢀ 308 for a few minutes, and then it was inverted into a ꢀ 788
bath, allowing the dark soln. to filter through the glass-wool plug. The tube was warmed to ca. 358 for ca.
5 min and then placed in an ice-bath for 5 min and then the NMR spectrum recorded immediately at the
probe temp. 1H-NMR (90 MHz, liq. NH3): 3.72 (dd, J ¼ 11.2, 8.1, 1 H); 4.41 (d, J ¼ 11.2, 1 H); 5.51 (dd,
J ¼ 8.1, 5.2, 1 H); 5.72 (t, J ¼ 11.2, 1 H); 6.20 (d, J ¼ 5.2, 1 H); 6.18 and 6.68 (m, 4 H).
Water-Quench Product of the Rearranged Anion: 2-(Penta-1,3,4-trien-1-yl)phenol (15). The NMR
tube containing anion 7c in liq. NH3 (prepared from 90 mg of 8c as described above), was warmed to r.t.
and then to ꢀ 358 and cut open. The contents were poured into a degassed mixture of Et2O (100 ml) and
H2O (30 ml) at 08 under N2. The mixture was stirred for 0.5 h, the org. layer washed with H2O (2 ꢁ 25 ml)
and sat. NaCl soln. (2 ꢁ 20 ml), dried (anh. MgSO4), and concentrated at 08, and the crude yellow oil
(80 mg, 89%) subjected to CC (SiO2 (25 g, 230 – 400 mesh), wet-packed with PE, elution with Et2O/PE
20 :80): 15 (30 mg, 43%). Pale solid. M.p. 119 – 1218. IR (KBr): 3369s (OH), 1928s (C¼C¼C stretching),
969s (trans CH¼CH), 858m (¼CH2 bending). 1H-NMR (90 MHz, CDCl3): 4.80 (d, overlapping OH
signal, J ¼ 6.6, 2 HꢀCe, OH); 5.85 – 6.25 (m, HꢀCc); 6.49 – 7.56 (m, HꢀCa, HꢀCb, 4 arom. H). 13C-NMR
(22.5 MHz, CDCl3): 77.36 (t, J ¼ 126.8, Ce); 95.15 (d, J ¼ 166.0, Cc); 116.50 (d, J ¼ 155.8, 1 C); 121.83 (d,

Helvetica Chimica Acta – Vol. 93 (2010)
1543
J ¼ 156.3, 1 C); 125.39 (d, J ¼ 156.2, 1 C); 126.28 (d, J ¼ 156.3, 1 C); 128.06 (d, J ¼ 156.3, 1 C); 128.95 (d,
J ¼ 161.2, 1 C); 131.17 (s, 1 C), 153.85 (s, 1 C); 214.76 (s, Cd). MS: 158 (93, Mþ), 141 (47, [M ꢀ 17]þ), 77
(100, [M ꢀ 18]þ).
2-(1H-Inden-2-yl)phenol (18). To a soln. of KNH2 in liq. NH3 prepared from K (180 mg, 4.0 mmol), a
few crystals of FeCl3, and liq. NH3 (30 ml) was added at ꢀ 788 under N2 a soln. of 6H-dibenz[b,f]oxocin
(8; 208 mg, 1.0 mmol) in anh. THF (7 ml). The resulting soln. was gradually warmed to ꢀ 358 and stirred
at ꢀ 358 for 1 h. A mixture of MeOH (30 ml) and Et2O (100 ml) was added dropwise to the mixture. The
org. layer was washed with H2O (2 ꢁ 25 ml) and sat. NaCl soln. (2 ꢁ 20 ml), dried (anh. MgSO4), and
concentrated at 08. The obtained yellow oil (190 mg, 91%) was subjected to CC (SiO2 (25 g, 230 – 400
mesh), Et2O/PE 20 :80): 18 (40 mg, 21%). Yellow solid. M.p. 170 – 1738. IR (CHCl3): 3519 (OH), 2922
1
(CH2). H-NMR (90 MHz, CDCl3): 3.86 (s, 2 H); 5.14 – 5.56 (m, 1 H); 6.79 – 7.55 (br., 9 H). 13C-NMR
(22.5 MHz, CDCl3): 42.36 (1 C); 117.2 (1 C); 121.8 (1 C); 122.1 (1 C); 124.5 (1 C); 125.9 (1 C); 127.6
(1 C); 129.5 (1 C); 129.7 (1 C), 130.9 (1 C); 131.9 (1 C); 143.7 (1 C); 144.0 (1 C); 146.3 (1 C); 154.2 (1 C).
MS: 208 (100, Mþ); 115 (46, [M ꢀ 93]þ).
REFERENCES
[1] E. Hꢃckel, Z. Phys. A Hadron. Nucl. 1931, 70, 204; E. Hꢃckel, Z. Phys. A Hadron. Nucl. 1931, 72,
310; E. Hꢃckel, Z. Phys. A Hadron. Nucl. 1932, 76, 628; E. Hꢃckel, Z. Elektrochem. 1937, 43, 752;
H. C. Longuet-Higgins, L. Salem, Proc. R. Soc. 1960, A257, 445; J. A. Elvidge, L. M. Jackman, J.
Chem. Soc. 1961, 859; F. Sondheimer, Pure Appl. Chem. 1964, 859; M. J. S. Dewar, Tetrahedron 1966,
8 (Suppl.), 75; H. E. Zimmerman, J. Am. Chem. Soc. 1966, 88, 1564; H. E. Zimmerman, J. Am.
Chem. Soc. 1966, 88, 1566; ꢀAromaticityꢁ, Special Publication No. 21, the Chemical Society, London,
1967; A. Julg, P. FranÅois, Theor. Chim. Acta 1967, 7, 249; R. B. Woodward, R. Hoffmann, Angew.
Chem. 1969, 81, 797; R. B. Woodward, R. Hoffmann, Angew. Chem., Int. Ed. 1969, 8, 781;
ꢀNonbenzenoid Aromaticsꢁ, Ed. J. P. Snyder, Academic Press, New York, 1971, Vol. II; R. C.
Haddon, V. R. Haddon, L. M. Jackman, Top. Curr. Chem. 1971, 16, 103; M. J. S. Dewar, Angew.
Chem., Int. Ed. 1971, 10, 761; H. E. Zimmerman, Acc. Chem. Res. 1971, 4, 272; E. Clar, ꢀThe
Aromatic Sextetꢁ, J. Wiley, London, 1972; F. Sondheimer, Acc. Chem. Res. 1972, 5, 81; ꢀTopics in
Nonbenzenoid Aromatic Chemistryꢁ, Eds. T. Nozoe, R. Breslow, K. Hafner, S. Ito, I. Murata,
Hirokawa, Tokyo, 1973, Vol. I; B. A. Hess Jr., L. J. Shaad, J. Am. Chem. Soc. 1973, 95, 3907; M. J.
Cook, A. R. Katritzky, P. Linda, in ꢀAdvances in Heterocyclic Chemistryꢁ, Eds. A. R. Katritzky, A. J.
Boulton, Academic Press, New York, 1974, Vol. 17, p. 256; A. G. Anastassiou, Acc. Chem. Res. 1976,
9, 453; A. G. Anastassiou, H. S. Kasmai, in ꢀAdvances in Heterocyclic Chemistryꢁ, Eds. A. R.
Katritzky, A. J. Boulton, Academic Press, New York, 1978, Vol. 23, p. 55; S. S. Shaik, P. C. Hiberty, J.
Am. Chem. Soc. 1985, 107, 3039; P. C. Hiberty, S. S. Shaik, J.-M. Lefour, G. Ohanessian, J. Org. Chem.
1985, 50, 4657; P. J. Garratt, ꢀAromaticityꢁ, J. Wiley & Sons, New York, 1986; A. T. Balaban, M.
Bancin, V. Ciorba, ꢀAnnulenes, Benzo-, Hetero-, Homo Derivatives, and their Valence Isomersꢁ,
´
CRC Press, Boca Raton, 1987, Vol. I; P. von Rague Schleyer, C. Maerker, A. Dransfield, H. Jiao,
N. J. R. van Eikema Hommes, J. Am. Chem. Soc. 1996, 118, 6317; ꢀComprehensive Heterocyclic
Chemistryꢁ, Eds. A. R. Katritzky, C. W. Rees, E. Scriven, Pergamon/Elsevier, Oxford, 1984 – 1996,
18 Vols.; A. F. Pozharskii, A. T. Soldatenkov, A. R. Katritzky, ꢀ Heterocycles in Life and Societyꢁ,
Wiley, New York, 1997.
[2] R. C. Haddon, L. F. Schneemeyer, J. V. Waszczak, S. H. Glarum, R. Tycko, G. Dabbagh, A. R.
Kortan, A. J. Muller, A. M. Mujsce, M. J. Rosseinsky, S. M. Zahurak, A. V. Makhija, F. A. Thiel, K.
Raghavachari, E. Cockayne, V. Elser, Nature (London, U.K.) 1991, 350, 46; R. C. Haddon, Science
(Washington, DC, U.S.) 1993, 261, 1545; J.-i. Aihara, S. Takata, J. Chem. Soc., Perkin Trans. 2 1994, 1,
65; J. Aihara, T. Yamabe, H. Hosoya, Synth. Met. 1994, 64; M. Bꢃhl, A. Hirsch, Chem. Rev. 2001, 101,
1153; G. Van Lier, P. W. Fowler, F. De Proft, P. Geerlings, J. Phys. Chem. A 2002, 106, 5128; L. Liu,
C. S. Jayanthi, S. Y. Wu, Chem. Phys. Lett. 2002, 357, 91; D. Astruc, in ꢀModern Arene Chemistryꢁ,
Ed. D. Astruc, Wiley – VCH, Weinheim, 2002, p. 1; J.-i. Aihara, Bull. Chem. Soc. Jpn. 2003, 76, 103;
R. Taylor, Phys. Chem. Chem. Phys. 2004, 6, 328; Z. Chen, R. B. King, Chem. Rev. 2005, 105, 3613; I.
Lukovits, F. H. Karman, P. M. Nagy, E. Kalman, Croat. Chem. Acta 2007, 80, 233.

1544
Helvetica Chimica Acta – Vol. 93 (2010)
[3] N. V. Subba Rao, Khim. Geterotsikl. Soedin. 1977, 291; A. K. Debnath, R. L. Lopez de Compadre, G.
Debnath, A. J. Shusterman, C. Hansch, J. Med. Chem. 1991, 34, 786; A. Bomhard, M. Reiffen, J.
Heider, M. Psiorz, C. Lillie, J. Med. Chem. 1991, 34, 942; G. H. Posner, Z. Li, M. C. White, V.
Vinader, K. Takeuchi, S. E. Guggino, P. Dolan, T. W. Kensler, J. Med. Chem. 1995, 38, 4529; D. L.
Moody, M. Dyba, T. Kosakowska-Cholody, N. I. Tarasova, C. J. Michejda, Bioorg. Med. Chem. Lett.
2007, 17, 2380; T. Asaki, T. Hamamoto, Y. Sugiyama, K. Kuwano, K. Kuwabara, T. Niwa, Bioorg.
Med. Chem. Lett. 2007, 17, 6588.
[4] T. J. Katz, J. Am. Chem. Soc. 1960, 82, 3784; T. J. Katz, J. Am. Chem. Soc. 1960, 82, 3785; J. H.
Noordlik, T. E. M. van der Hark, J. J. Mooji, A. K. K. Klaassen, Acta Crystallogr., Sect. B 1974, 30,
833.
[5] M. Fletschinger, B. Zipperer, H. Fritz, H. Prinzbach, Tetrahedron Lett. 1987, 28, 2517.
[6] G. M. Badger, ꢀAromatic Character and Aromaticityꢁ, Cambridge University Press, Cambridge,
1969; D. Lewis, D. Peters, ꢀFacts and Theories of Aromaticityꢁ, Macmillan, London, 1975.
´
[7] F. Bernardi, A. Bottoni, A. Venturini, J. Mol. Struct. (Theochem) 1988, 163, 173; P. v. Rague
Schleyer, H. Jiao, B. Goldfuss, P. K. Freeman, Angew. Chem., Int. Ed. 1995, 34, 337.
[8] A. G. Anastassiou, E. Reichmanis, Angew. Chem., Int. Ed. 1974, 13, 404.
[9] T. J. Katz, M. Yoshida, L. C. Siew, J. Am. Chem. Soc. 1965, 87, 4516.
[10] A. G. Anastassiou, H. S. Kasmai, D. Hauger, Angew. Chem., Int. Ed. 1982, 21, 791.
[11] A. G. Anastassiou, H. S. Kasmai, D. Hauger, J. Chem. Soc., Chem. Commun. 1981, 646.
[12] A. G. Anastassiou, H. S. Kasmai, Angew. Chem., Int. Ed. 1980, 19, 393.
[13] H. E. Simmons, J. K. Williams, J. Am. Chem. Soc. 1964, 86, 3222.
[14] B. Begasse, M. Le Corre, Synthesis 1981, 197.
[15] E. E. Schweizer, T. Minami, S. E. Anderson, J. Org. Chem. 1974, 39, 3038.
[16] a) B. Harrison, P. Crews, J. Org. Chem. 1997, 62, 2646; b) S. K. Sabui, R. V. Venkateswaran,
Tetrahedron Lett. 2004, 45, 9653.
[17] F. A. Macias, R. M. Varela, A. Torres, J. M. G. Molinillo, F. R. Fronczek, Tetrahedron. Lett. 1993, 34,
1999; F. A. Macias, J. M. G. Molinillo, R. M. Varela, A. Torres, F. R. Fronczek, J. Org. Chem. 1994,
59, 8261.
[18] A. G. Anastassiou, H. S. Kasmai, Angew. Chem., Int. Ed. 1980, 19, 43.
[19] A. F. Cockerill, G. L. O. Davies, R. C. Harden, D. M. Rackham, Chem. Rev. 1973, 73, 553.
[20] N. Bhuvaneswari, C. S. Venkatachalam, K. K. Balasubramanian, J. Chem. Soc., Chem. Commun.
1994, 1177.
[21] V. M. Parikh, ꢀAbsorption Spectroscopy of Organic Moleculesꢁ, Addison-Wesley, Reading, 1974,
p. 59.
[22] J. S. Binkley, J. A. Pople, W. J. Hehre, J. Am. Chem. Soc. 1980, 102, 939; W. J. Hehre, L. Radom, P. v.
´
Rague Schleyer, J. A. Pople, ꢀAb Initio Molecular Orbital Theoryꢁ, J. Wiley & Sons, New York, 1986.
[23] W. J. Hehre, ꢀA Guide to Molecular Mechanics and Quantum Chemical Calculationsꢁ, 2nd edn.,
Wavefunction, Inc., Irvine, CA, 2006, Chapt. 5.
´
[24] a) T. M. Krygowski, M. K. Cyranski, Chem. Rev. 2001, 101, 1385; b) M. Kertesz, C. H. Choi, S. Yang,
Chem. Rev. 2005, 105, 3448.
[25] R. S. Mulliken, J. Chem. Phys. 1955, 23, 1833; R. S. Mulliken, J. Chem. Phys. 1955, 23, 1841; R. S.
Mulliken, J. Chem. Phys. 1955, 23, 2238; R. S. Mulliken, J. Chem. Phys. 1955, 23, 2343.
[26] V. I. Minkin, M. N. Glukhovtsev, B. Y. Simkin, ꢀAromaticity and Antiaromaticity, Electronic and
Structural Aspectsꢁ, J. Wiley & Sons, New York, 1994, Chapt. 2, p. 6; T. M. Krygowski, M. K.
˜
Cyranski, Z. Czarnocki, G. Hꢅfelinger, A. R. Katritzky, Tetrahedron 2000, 56, 1783.
[27] Spartanꢁ06, Wavefunction, Inc., Irvine, CA, 2006.
Received December 2, 2009