Attempted generation of the potentially aromatic 6,7-diphenyldibenzo[e,g] [1,4]diazocine dianion leads with profound rearrangement to the isomeric n-(2-amino-1,2-diphenylethenyl)carbazole dianions

Article abstract of DOI:10.1021/ol901503s

The attempted generation of the potentially aromatic 6,7-diphenyldibenzo[e, g][1,4]diazocine dianion from sodium in THF leads with profound rearrangement to the Isomeric N-(2-amino-1,2-diphenylethenyl)carbazole dianions and, after hydrolysis, to a 55:45 m

Full text of DOI:10.1021/ol901503s

ORGANIC
LETTERS
2009
Vol. 11, No. 18
4060-4063
Attempted Generation of the Potentially
Aromatic
6,7-Diphenyldibenzo[e,g][1,4]diazocine
Dianion Leads with Profound
Rearrangement to the Isomeric
N-(2-Amino-1,2-diphenylethenyl)carbazole
Dianions
John J. Eisch,*,† Renuka N. Manchanayakage,† and Arnold L. Rheingold‡
Department of Chemistry, State UniVersity of New York at Binghamton, Binghamton,
New York 13902-6000, and Department of Chemistry and Biochemistry, UniVersity of
California, San Diego, La Jolla, California 92093-0332
jjeisch@binghamton.edu
Received July 1, 2009
ABSTRACT
The attempted generation of the potentially aromatic 6,7-diphenyldibenzo[e,g][1,4]diazocine dianion from sodium in THF leads with profound
rearrangement to the isomeric N-(2-amino-1,2-diphenylethenyl)carbazole dianions and, after hydrolysis, to a 55:45 mixture of Z- and E-isomers
of N-(2-amino-1,2-diphenylethenyl)carbazoles. These isomers were separated and their individual structures determined by X-ray diffraction.
Since treatment of the starting diazocine first with tert-butyllithium and then with water also yielded the same Z- and E-isomeric mixture,
electron transfer reduction is clearly involved.
Our ongoing interest in the chemical properties of proposed1
and actual2 cyclic conjugated imines has led us to examine
the electron-transfer reduction of dibenzo[e,g][1,4]diazocines
such as 1, having a nonplanar eight-membered ring3 and
having received previous attention as a heterocyclic model
for cyclooctatetraene.4,5 The central question of our study
has been whether the eight-membered ring of 1 might
become planar upon the addition of two electrons (2) because
of the potential aromaticity of such a dianion (Scheme 1).
Such a structural change would parallel the known behavior
of nonplanar cyclooctatraene, which forms a planar dianion
upon two-electron reduction.6 Previous studies of 1 seemed
to be in accord with such a reduction, since the reduction of
† State University of New York at Binghamton.
‡ University of California, San Diego.
(4) Allinger, N. L.; Youngdale, G. A. J. Org. Chem. 1959, 24, 306–
308.
(1) Eisch, J. J.; Chan, T. Y.; Gitua, J. N. Eur. J. Org. Chem. 2008, 39,
392–397.
(5) Allinger, N. L.; Szkrybalo, W.; DaRooge, M. A. J. Org. Chem. 1963,
28, 3007–3009.
(2) Eisch, J. J.; Abraham, T. Tetrahedron Lett. 1976, 1647–1650.
(3) Finder, C. J.; Newton, M. G.; Allinger, N. L. J. Chem. Soc., Perkin
Trans. 2 1973, 1928–1932.
(6) (a) Katz, T. J. J. Am. Chem. Soc. 1960, 82, 3784. (b) Goldberg,
S. Z.; Raymond, K. N.; Harmon, C. A.; Templeton, D. H. J. Am. Chem.
Soc. 1974, 96, 1348.
10.1021/ol901503s CCC: $40.75
Published on Web 08/24/2009
 2009 American Chemical Society

1 by sodium amalgam in ethanol was interpreted to have
produced the 5,6,7,8-tetrahydro derivative 3 of 1 (assumed
trans-6,7-diphenyl), based upon elemental analyses and
infrared spectra data4,7,8 If 3 were indeed formed in such a
reduction, 2 would seem to be a necessary precursor (but
cf. infra4).
Scheme 1
Figure 1. Thermal ellipsoid (30%) diagrams for the Z-isomer of
N-(2-amino-1,2-diphenylethenyl)carbazole (5a) (orthorhombic, pale
yellow block-type crystals, mp 166-167 °C) and the E-isomer of
N-(2-amino-1,2-diphenylethenyl)carbazole (5b) (triclinic, colorless
needles, mp 124-125 °C). Selected bond lengths (Å) and angles
of 5a: N(1)-C(12) 1.388(2), N(1)-C(1) 1.390(2), N(1)-C(13)
1.436(2),N(2)-C(20)1.387(2),C(13)-C(20)1.356(2),C(12)-N(1)-C(1)
108.22(13), C(12)-N(1)-C(13) 126.04(14), C(1)-N(1)-C(13)
124.08(14),N(1)-C(1)-C(2)128.55(15),N(1)-C(1)-C(6)109.56(14),
N(1)-C(13)-C(14) 115.13(14), C(13)-C(20)-N(2) 122.47 (16).
5b: N(1)-C(12) 1.396(4), N(1)-C(1) 1.396(4), N(1)-C(13) 1.440(4),
N(2)-C(20) 1.378(4), C(13)-C(20) 1.354(5), C(12)-N(1)-C(1)
108.4(3), C(12)-N(1)-C(13) 127.0(3), C(1)-N(1)-C(13) 124.6(3),
N(1)-C(1)-C(2)128.4(3),N(1)-C(1)-C(6)109.5(3),N(1)-C(13)-C(14)
115.3(3), C(13)-C(20)-N(2) 122.3(3).
We therefore undertook the reduction of 1 with sodium
metal in anhydrous THF at 25 °C. Hydrolysis of the reaction
solution, after 5 h and after separation of the residual sodium,
showed complete consumption of 1 and gave an 82% yield
of a colorless powder, whose mass spectrum (70 eV) had
the appropriate parent peak of 360 for structure 4 but
whose infrared spectrum displayed two sharp N-H
stretching bands at 3378 and 3480 cm-1. Such bands are
inconsistent with the secondary amine of structure 4
having a symmetry plane but instead would require a
primary amine 5. Amine 5 was recrystallized from 95%
ethanol to yield as separate crystals pale yellow cubes of
5a, mp 166-167 °C, and colorless needles of 5b, mp
166-167 °C, in a 55:45 ratio (Figure 1, Scheme 2).
Picking apart such crystals9 under low-power (30×)
microscopic magnification and determining their individual
structures by X-ray diffraction showed that 5a and 5b are
the Z- and E-isomers, respectively, of N-(2-amino-1,2-
diphenylethenyl)carbazole (Figure 1).
Scheme 2
Although there remains no doubt of the structures of 5a
and 5b in the solid state and of the isolated solid 5 as
therefore a mixture of 5a and 5b, dissolution of the individual
solids, 5, 5a, and 5b, in CDCl3 and NMR spectral measure-
Scheme 3
1
ments give an unexpected result. The H and 13C NMR
spectra of the three samples were superimposably identical
and the decoupled 13C spectrum of each displayed exactly
16 singlets. One interpretation of these spectral data is that
in solution CDCl3 sets up an equilibration between 5a and
5b and that D-bonding of CDCl3 with the amino group of
either isomer favors the solvated E-isomer 6b over Z-isomer
6a for steric reasons (Scheme 3).10
Chemical confirmation of structures 5a and 5b was
obtained through the quantitative hydrolysis of their mixture
(7) Ta¨uber, E. Ber. Dtsch. Chem. Ges. 1892, 25, 3287
(8) Ta¨uber, E. Ber. Dtsch. Chem. Ges. 1893, 26, 1703
(9) A.L.R. scrutinized under magnification the crystalline sample of 5
intended for XRD analysis and was able to discern the two distinct crystal
habits and to separate crystals of 5a and 5b for individual XRD analysis.
.
.
(10) The complete 1H and 13C NMR spectra of 5, 5a, and 5b in CDCl3
are available for inspection in the Supporting Information.
Org. Lett., Vol. 11, No. 18, 2009
4061

in refluxing acetic acid with 5 N HCl to yield the known
ketone 7 (Scheme 4).4 The minor products isolated from the
original reduction of 1 (Scheme 1) were carbazole (8, 9%)
and benzyl phenyl ketone (9, 9%), both of which can be
viewed as further C-N cleavage products of the dianions
of 5a and 5b, followed by hydrolysis.
is ample precedent for such intramolecular formation and
coupling of bis-radical anions by alkali metals. For example,
treatment of o-diphenylbenzene (14) with lithium metal in
THF at 25 °C and hydrolysis leads via 15a to tetrahydro
derivative 15b, whose subsequent dehydrogenation with 2,3-
dichloro-5,6-dicyanobenzoquinone (DDQ) yields 65% of
triphenylene (16) (Scheme 6).11 In an analogous manner,
1,1′-binaphthyl is transformed into perylene in 55% yield.11
Scheme 4
Scheme 6
From these surprising findings it follows that the previ-
ously reported tetrahydro derivative does not have proposed
structure 3, but rather that of an N-(2-amino-1,2-diphenyl-
ethyl)carbazole derivative (10) of 5. In better accord with
such a primary amine structure for 10 is the reported presence
of two infrared stretching bands at 3425 and 3333 cm-1.4
Furthermore, that such a rearrangement can be effected
by electron-transfer agents other than metals received cor-
roboration by treating 1 in THF at -78 °C with tert-
butyllithium. After bringing such a mixture to 25 °C and
hydrolyzing, again the principal products were 5a and 5b.
This reagent in pentane has been shown to effect by electron
transfer the reductive dimerization of diphenylacetylene into
(E,E)-1,4-dilithio-1,2,3,4-tetraphenyl-1,3-butadiene.12a,b,13
This last experiment provides insight into the original work
of Allinger and co-workers,4 who observed the extraordinary
transformation of 1 into 7 by two consecutive reactions: first,
The formation of disodium salts 13a and 13b in the course
of the sodium reduction of 1 appears to be best explained
by the intramolecular coupling of the asterisked N- and
C-centered radicals in the nonplanar bis-radical anion 11 to
yield 12, which undergoes anionic elimination and aroma-
tization to produce the anions 13a and 13b (Scheme 5). There
(11) Eisch, J. J.; Kovacs, C. A. In Polynuclear Aromatic Compounds;
Ebert, L. B., Ed.; Advances in Chemistry Series 217; American Chemical
Society: Washington, D.C., 1988; Chapter 6, pp 96, 103-4, where
experimental procedure and discussion are available.
Scheme 5
(12) Such a reductive dimerization of this alkyne by tert-butyllithium
is readily understood as an electron-transfer formation of Li+[PhCdCPh]-
and coupling of such radical-anions. (a) Mulvaney, J. E.; Garlund, Z. G.;
Garlund, S. L. J. Am. Chem. Soc. 1963, 85, 3897–3898. (b) The reaction
between 1 in THF at -78 °C and a pentane solution of tert-butyllithium
(1.7 M, Aldrich) was conducted on a 1.50 mM scale of 1 and a 3-fold
molar excess of tert-butyllithium for 1 h, and then the reaction mixture
was brought to and maintained at 25 °C for 3 h before quenching with wet
diethyl ether. Usual workup and 1H NMR analysis showed the typical
absorptions of 5a and 5b in CDCl3 solution (cf. Scheme 3). All of these
operations occurred without incident since the tert-butyllithium in pentane
and the reaction mixture were maintained under an atmosphere of anhydrous,
deoxygenated argon. Later in 2008, a fatal accident occurred at UCLA when
a chemistry research assistant died from injuries sustained while conducting
an organic synthesis procedure employing tert-butyllithium in pentane:
Chem. Eng. News, Online, Latest News, Jan 22, 2009. The details of the
experiment and some analysis of the circumstances of the tragic accident
have now been reported: Chem. Eng. News, Aug 3, 2009. We advise anyone
considering experimentation with tert-butyllithium to study the foregoing
reports and to employ instead tert-butyllithium in less volatile solvents,
such as cyclohexane or heptane, and to work under an argon atmosphere.
Appropriate safety measures for working with air- and moisture-sensitive
4062
Org. Lett., Vol. 11, No. 18, 2009

the treatment of 1 with LiAH4 in refluxing THF, followed
by hydrolysis, and second, heating the resulting crude product
with aqueous HCl and acetic acid. To explain such a
surprising formation of 7, these workers4 proposed successive
hydride attack on 1 followed by intramolecular nucleophilic
amide attack and hydrolysis. Since our results show that
attack of tert-butyllithium on 1 can produce 5a and 5b by
an electron-transfer process, we suggest that LiAlH4 in THF
can similarly achieve electron-transfer reduction of 1 into
5a and 5b. Previous workers have proposed electron-transfer
processes for reductions with LiAlH4.14
Acknowledgment. This work is dedicated to the memory
of Glen Allan Russell (1925-1998), late Distinguished
Professor of Chemistry, Iowa State University, outstanding
mentor and pioneer of organic electron-transfer reaction
mechanisms. Support for this research stems from a Senior
Scientist Award (J.J.E.) made by the Alexander von Hum-
boldt Foundation, Bonn, Germany.
Supporting Information Available: Crystallographic
(CIF) data for structures 5a and 5b in the solid state; criteria
of purity of starting materials and separated final products
for key transformations; experimental procedures; and copies
Finally, this work suggests that diazocine 1 may prove to be
the most effective electron trap for the detection of electron
transfer in organometallic reactions with organic substrates.15
1
of the fully displayed H and 13C NMR spectra of the
following solutions in CDCl3: (1) starting diazocine, com-
pound 1; (2) isolated mixture of 5a and 5b, compound 5;
(3) Z-isomer, mp 166-167 °C, compound 5a; (4) E-isomer,
mp 124-125 °C, compound 5b; and (5) compound 7. This
material is available free of charge via the Internet at
http://pubs.acs.org.
compounds are discussed extensively in the following reference: Eisch, J. J.
Organometallic Syntheses; Academic Press: New York, 1981; Vol. 2, Part
I, pp 1-81, in detail and Part II, pp 87-191, passim.
(13) Another apparent electron-transfer cyclization by n-butyllithium in
THF is the transformation of 2-(2,2-diphenyl-1-ethenyl)biphenyl into
9-(diphenylmethyl)fluorene in 25% yield. Eisch, J. J.; Chobe, P. Polynuclear
Aromatic Compounds; Ebert, L. B., Ed.; Advances in Chemistry Series,
No. 217; American Chemical Society: Washington, D.C., 1988; pp 97-
105.
OL901503S
(14) In such a precedent for the electron-transfer reduction by LiAlH4
in THF we offer the pertinent study of the reduction of alkyl iodides: Ashby,
E. C.; Pham, T. N.; Amrollah-Madjdabadi, A J. Org. Chem. 1991, 56, 1596–
1603.
(15) Eisch, J. J. Doetschman, D. C. UV spectral and EPR studies in
progress.
Org. Lett., Vol. 11, No. 18, 2009
4063