3,5,7,9-tetraphenylhexaazaacridine: A highly stable, weakly antiaromatic species with 16 π electrons

Article abstract of DOI:10.1002/anie.200500831

(Chemical Equation Presented) Zwitterionic structures with very low lying triplet states are the dominant characteristics of the first hexaazaacridine, 3,5,7,9-tetraphenylhexaazaacridine (TPH-acridine, see picture), and the analogous TPH-anthracene. The s

Full text of DOI:10.1002/anie.200500831

Angewandte
Chemie
Antiaromatic Species
DOI: 10.1002/anie.200500831
3,5,7,9-Tetraphenylhexaazaacridine: A Highly
Stable, Weakly Antiaromatic Specieswith
16 p Electrons**
Peter Langer,* Anja Bodtke, Nehad N. R. Saleh,
Helmar Görls, and Peter R. Schreiner*
The efficiency of electrical, magnetic, and optical devices
based on organic materials is closely related to their HOMO–
LUMO energy gap.[1] The structure of tetraphenylhexaaza-
anthracene (TPH-anthracene, 1), first prepared in 1908 as a
stable, highly fluorescent compound,[2] was recently studied
by Wudl and Houk et al.; 1 exists as a zwitterionic cyanine
(resonance structure 1a) that undergoes photoinduced intra-
molecular electron transfer to give a radical pair.[3,4] Wudl and
[*] Prof. Dr. P. Langer
Institut für Chemie, Universität Rostock
Albert-Einstein-Strasse 3a, 18059 Rostock (Germany)
Fax: (+49)381-498-6412
E-mail: peter.langer@uni-rostock.de
and
Leibniz-Institut für Organische Katalyse (IfOK)
an der Universität Rostock e.V.
Albert-Einstein-Strasse 29a, 18059 Rostock (Germany)
Prof. Dr. P. R. Schreiner
Institut für Organische Chemie, Justus-Liebig-Universität
Heinrich-Buff-Ring 58, 35392 Giessen (Germany)
Fax: (+49)641-99-34309
E-mail: prs@chem.uga.edu
and
The University of Georgia, Department of Chemistry
Athens, GA (USA)
E-mail: prs@chem.uga.edu
Dr. A. Bodtke, Dr. N. N. R. Saleh
Institut für Chemie und Biochemie
Ernst-Moritz-Arndt Universität Greifswald
Soldmannstrasse 16, 17487 Greifswald (Germany)
Dr. H. Görls
Institut für Anorganische und Analytische Chemie
Universität Jena
August-Bebel-Strasse 2, 07740 Jena (Germany)
[**] This work was supported by the German state of Mecklenburg-
Western Pomerania (scholarship for A.B.), the Dr. Dr. Gerda-von-
Mach Gedächtnisstiftung (scholarship for N.N.R.S.), and the
Deutsche Forschungsgemeinschaft. P.R.S. was supported at UGA
through the National Science Foundation (CHE-0209827).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or fromthe author.
Angew. Chem. Int. Ed. 2005, 44, 5255 –5259
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5255

Communications
Houk et al. concluded that the apparent antiaromaticity of 1
character (Table 1): a solution of 2 in DMF is orange colored
(lmax = 547 nm) and, despite the low solubility of 2, highly
fluorescent at F(lmax) = 585 nm. In TFA (CF3CO2H), how-
ever, the solubility of 2 is higher owing to protonation to give
is overwhelmed by the dominance of cyanine ion stabilization.
However, the aromatic properties, which can be assessed in a
straightforward manner by, for instance, nucleus-independent
chemical shifts (NICS),[5] were not analyzed further.
Herein, we report the synthesis of what is, to the best of
our knowledge, the first example of a hexaazaacridine,
namely 3,5,7,9-tetraphenylhexaazaacridine (TPH-acridine,
2). By means of a combination of experimental and computa-
tional results, we show that 1 and 2 can be viewed as stable, 16-
p-electron, weakly antiaromatic highly zwitterionic structures
with very low lying triplet states. As neutral molecules are
usually less polarized than their excited states, 2, along with its
carbon analogue 1, is the exception where the state polarity is
reversed so that photoexcitation leads to a less polarized
excited state. As a consequence, normally much higher lying
triplet and singlet states are much closer in energy. It is
conceivable that appropriately substituted acridines can have
open-shell ground states and would be excellent candidates
for organic magnets and data storage.[1]
Table 1: Solvatochromic properties of 2.
Solvent
lmax [nm]
(loge)
Color of solution
DMF
CH3CO2H
conc. HNO3
547 (4.51) orange, strongly fluorescent, poorly soluble
535 (4.57) orange, fluorescent, poorly soluble
–
red, very poorly soluble
conc. H3PO4 540 (4.29) bluish purple, poorly soluble
conc. HClO4
CF3CO2H
–
–
green, very poorly soluble
deep purple, reasonably soluble
conc. H2SO4 662 (4.34) deep green, highly soluble
cationic 2-H+ (deep purple solution). Therefore, most NMR
spectra were recorded in [D1]TFA. Compound 2 is somewhat
soluble in glacial acetic acid, concentrated nitric acid,
phosphoric acid, and perchloric acid to give orange, red,
bluish purple, and green solutions, respectively. Excellent
solubility and strong bathochromism were observed in con-
centrated sulfuric acid (lmax = 662 nm, deep green solution).
Successive dilution of this solution with water or acetone
changes the color from green to blue and purple and finally to
pink and orange (see Supporting Information for quantitative
measurements); this process is reversible. In addition to the
acidity and concentration, the counterion also affects the
absorption of the solution, which suggests the formation of
aggregates.
The synthesis of TPH-acridine (2) essentially follows the
strategy reported for the synthesis of 1 (Scheme 1).[2,3]
The structure of TPH-acridine (2) was independently
confirmed through a crystal structure analysis of the cation 2-
H+, which was prepared by protonation of 2 with hydrochloric
and glacial acetic acids (Figures 1 and 2). Protonated TPH-
Scheme 1. Synthesis of TPH-acridine (2a): a) PhCOCl (2 equiv), NEt3
(2 equiv), CH2Cl2, 14 h, 0–208C, 48%; b) PCl5 (2.9 equiv), toluene,
1.5 h, 208C, then 2 h, reflux, 88%; c) PhNHNH2 (2.7 equiv), n-hexane,
14 h, 0–208C, 90%; d) air, DBU (2.8 equiv), MeOH, 4 d, 208C, 10%.
DBU=diazabicyclo[5.4.0]undec-7-ene.
Figure 1. Resonance structures of protonated TPH-acridine, 2-H+.
acridine is a symmetric molecule that is protonated at the
central nitrogen atom. Analysis of the bond lengths of THP-
acridine indicates that the ionic resonance structure 2a-H+ is
more important than the diradical contributor 2b-H+; this
compares well with 1.
For a better understanding of the electronic nature of 2,
we also computed its properties and compared those to 1. As
Wudl and Houk et al. successfully employed density func-
tional theory (DFT) to examine the structures and energies of
1, we adopted the same approach (B3LYP)[6] but expanded
the basis set from 6-31G* to 6-311 + G** for the energies on
the B3LYP/6-31G*-optimized geometries because we noted a
marked basis set dependence of the singlet–triplet energy
separation (DEST); this is not unexpected for species with
However, the oxidative cyclization of 6 as the key step
proved to be very sluggish and afforded 2 in only 3% yield.
After much experimentation, we found that 2 can be obtained
reproducibly in 10% yield. Both the workup of the final
product and the quality of 6 proved to be important
parameters. In this context, the preparation of (rather
unstable) bis(imidoyl) dichloride 5 in high purity was crucial
and required careful handling.
The deep purple hexaazaacridine 2 is air stable, thermally
stable up to 4008C, and poorly soluble in most organic
solvents. Particularly noteworthy is its strong solvatochromic
5256
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5255 –5259

Angewandte
Chemie
Figure 2. ORTEP plot of protonated TPH-acridine, 2-H+. Bond lengths
[Š] and angles [8]: N1-C2 1.353(3), N1-N2 1.368(3), N2-C1 1.307(3),
N3-C3 1.304(3), N3-C1 1.368(3), N4-C3 1.359(3), N4-C6 1.359(3), C2-
C4 1.390(3), C2-C3 1.441(3); C2-N1-N2 123.28(19), C1-N2-N1
116.34(19), C3-N3-C1 115.8(2), C3-N4-C6 123.7(2), N2-C1-N3
126.2(2), N1-C2-C4 125.4(2), N1-C2-C3 114.46(19), C4-C2-C3 120.2(2),
N3-C3-C2 123.3(2). CCDC 255822 contains the supplementary crystal-
lographic data for this paper. These data can be obtained free of
charge fromthe Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
highly polarized ground and close-lying excited states. Our
final level is therefore B3LYP/6-311 + G**//B3LYP/6-31G*;
zero-point vibrational energy (ZPVE) differences are mar-
ginal and are therefore not applied (the ZPVEs are reported
in the Supporting Information). For the computation of UV
spectra, we utilized time-dependent DFT (TD-B3LYP/6-
311 + G**), which is capable of describing vertical excited
states more reliably than, for example, configuration inter-
action with single excitations (CIS).[7]
Since the crystal structure of TPH-anthracene (1) is
known and the structure of the complete molecule had not
been computed (the phenyl groups had been replaced by
hydrogen atoms in the original publication),[3] we first
optimized the structure of 1 to establish whether our
computational approach would reproduce the structural
details within acceptable errors. We computed 1 in Cs and
C2 symmetry (rotation of one of the N-phenyl groups relates
these two structures) and found the Cs form to be very slightly
preferred (0.3 kcalmolÀ1); the X-ray crystal structure is
asymmetric (C1) as a result of packing effects. The key
structural parameters agree quite well (Figure 3), and we
expect our approach also to give good structural data for
TPH-acridine (2). The phenyl groups in 1 lower DEST by
2.4 kcalmolÀ1 (at B3LYP/6-31G*), but this separation
(16.3 kcalmolÀ1 for the preferred Cs structure) increases to
Figure 3. TPH-anthracene (1) and protonated TPH-acridine, 2-H+:
comparison of selected bond lengths [Š] of the computed structure
(B3LYP/6-31G*) and the crystal structure (in parentheses).
17.1 kcalmolÀ1 when we employ a more polarizable and
diffuse basis set at our reference level. For proper compar-
ison, we also computed protonated TPH-acridine, 2-H+, at
B3LYP/6-31G*; the key structural parameters are in excellent
agreement with the crystal structure data (Figure 3).
A comparison of the measured and computed 13C NMR
data of protonated 2 (the spectrum was recorded in [D1]TFA)
also is favorable (Table 2); the average deviation of the
computed data to lower fields is only about 5 ppm and can be
explained by neglect of solvent effects.
The UV/Vis spectrum is also well reproduced computa-
tionally with a maximum absorption at 506 nm (experimen-
tally at 535 nm in AcOH), which is assigned to an excited state
that corresponds qualitatively to the lowest triplet state (as in
1).[3] By analogy to 1, the fluorescence spectrum is not a
mirror image of the UV/Vis absorption spectrum; the Stokeꢀs
Table 2: Experimental and computed (B3LYP/6-311+G**//B3LYP/6-31G*) 13C NMR chemical shifts (d [ppm], relative to TMS as internal standard)
for 2-H+.
Type
CH
CH
CH
CH
C
CH
CH
CH
C
C
C
C
exptl
theor
Dd
91.1
89.2
1.9
126.3
133.5
7.2
129.4
134.8
5.4
131.2
136.7
5.5
132.1
136.2
4.1
132.5
139.0
6.5
134.0
139.8
5.8
135.4
143.1
7.7
141.2
143.0
1.8
141.5
147.3
5.8
158.3
159.1
0.8
158.6
166.0
7.4
Angew. Chem. Int. Ed. 2005, 44, 5255 –5259
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5257

Communications
shift of 2 is about 26000 cmÀ1, even larger than that of 1
(15000 cmÀ1) and hence about 5 to 40 times the typical
Stokeꢀs shift of typical laser dyes.[8]
Azaacridine 2 also has a Cs singlet ground state with a
DEST of 15.1 kcalmolÀ1; note that this gap is significantly
larger at B3LYP/6-31G* (21.1 kcalmolÀ1). The C2 singlet state
lies 7.9 kcalmolÀ1 higher in energy and displays an even larger
basis set effect on the DEST (20.8 at B3LYP/6-31G* and only
6.2 kcalmolÀ1 at our reference level). Stability tests showed
no instabilities in the wavefunctions of the singlet states; in
other words, the small DEST can be interpreted on the basis of
aromaticity/antiaromaticity (see below). Like anthracene 1,
acridine 2 is highly polar (dipole moment: 8.2 D) and displays
pronounced charge separation. The nitrogen atom in the
central ring is the most negatively polarized (À0.53 according
to a natural bond orbital (NBO) analysis) and is also the
preferred site of protonation (in either the Cs or C2 form);
protonation at N3 and N2 is 3.8 and 18.4 kcalmolÀ1 less
favorable, respectively. This is nicely confirmed by the crystal
structure of 2-H+.
Figure 4. HOMOs of 2 and cyclobutadiene (only one of the two degen-
erate HOMOs is shown).
cyclobutadiene is in the same range (5.9 kcalmolÀ1 at B3LYP/
6-311 + G** and 11.5 kcalmolÀ1 at CCSD(T)/cc-pVDZ//
B3LYP/6-311 + G**)[10] as that of 1 and 2, which lends further
support to the proposal that these polycyclic structures are
indeed stable antiaromatic entities. Although this formally
contradicts the conclusions of Wudl und Houk et al.[3] for
TPH-anthracene that double cyanine ion stabilization of
zwitterion 1a is preferred over the formal 16-p antiaroma-
ticity (1b),[11] we remark that the earlier study computation-
ally examined a smaller model system with a
The computed NICS values in the ring centers (Table 3)
reveal that all three rings of acridine 2 are antiaromatic. The
smaller basis set, which has a large effect on
the energies (see above). However, the two
conclusions are not necessarily mutually
exclusive because the systems under con-
sideration are only weakly antiaromatic and
partially avoid this unfavorable situation
through strong polarization.
In conclusion, TPH-acridine, which we
have synthesized and characterized, repre-
Table 3: NICS values (B3LYP/6-311+G**//B3LYP/6-31G*) of 1, 2, and 2-H+.
NICS value
relative to
ring center
1
2
2-H+
Cyclobutadiene (D4h)
inner
outer
inner
outer
inner
outer
0.0
0.5
1.0
2.0
3.0
4.0
+0.6
À1.1
À2.0
À0.2
+0.2
+0.2
+7.9
+5.6
+2.6
+0.7
+0.2
+0.1
+3.5
+0.9
À1.1
À0.5
À0.2
À0.2
+6.3
+4.0
+1.4
+0.1
À0.1
À0.2
+3.1
+1.5
+0.3
+0.2
0.0
+4.1
+1.9
À0.1
À0.3
À0.2
À0.2
+27.1
+27.9
+19.0
+4.5
+1.2
+0.4
sents the first hexaazaacridine. The struc-
ture and spectroscopic properties of TPH-
acridine and TPH-anthracene were studied
À0.1
same is true for 1 although the degree of antiaromaticity
varies more strongly. As NICS values in the ring centers often
are affected by the s electrons of neighboring bonds, we also
computed the NICS values above the ring centers and
compare these with the extreme case of an antiaromatic
species, namely D4h cyclobutadiene.[9] This analysis shows that
the outer rings of 1, 2, and 2-H+ at 0.5 Š above the ring of the
plane are, with the exception of the inner ring of 1, weakly
antiaromatic. NICS points that are further away from the ring
centers demonstrate that the antiaromatic character
decreases rapidly, in particular for the inner rings; these
NICS values are typical for nonaromatic structures. The
decrease in antiaromaticity, which is much more pronounced
than for cyclobutadiene, can be rationalized by differences in
the ring size and the high polarity of the heterocycles in 1 and
2.
by DFT computations and NICS analysis.
TPH-acridine and TPH-anthracene represent highly zwitter-
ionic structures with very low-lying triplet states. Both
polycyclic structures can be considered to be stable 16-p-
electron, weakly antiaromatic entities. TPH-acridine and
TPH-anthracene are of great theoretical and structural
interest and exhibit exciting physical properties that may be
utilized in organomagnetic applications.
Received: March 6, 2005
Published online: July 20, 2005
Keywords: antiaromaticity · aromaticity · density functional
.
theory · hexaazaacridine · hexaazaanthracene
Hence, 1 and 2 must be considered to be stable, weakly
antiaromatic molecules with highly polar singlet ground
states. Our conclusions are further supported by visual
inspection of the HOMO of 2 (Cs; Figure 4), which is
symmetrically p-antibonding. This is in analogy to the
simplest neutral antiaromatic singlet, the highly unstable
D4h-symmetric cyclobutadiene (it is common knowledge that
it has a triplet ground state). Remarkably, the DEST for
[1] For oligothiophenes with low-lying triplet states see: a) F.
Garnier, Angew. Chem. 1989, 101, 529; Angew. Chem. Int. Ed.
Engl. 1989, 28, 513; b) F. Garnier, A. Yassar, R. Hajlaoui, G.
Horowitz, F. Deloffre, B. Servet, S. Ries, P. Alnot, J. Am. Chem.
Soc. 1993, 115, 8716; c) F. Garnier, R. Hajlaoui, A. Yassar, P.
Srivastava, Science 1994, 265, 1684; d) A. Dodabalapur, L. Torsi,
H. E. Katz, Science 1995, 268, 270; e) A. Dodabalapur, L. J.
Rothberg, A. W. P. Fung, H. E. Katz, Science 1996, 272, 1462; a
low-lying triplet state was predicted for the not-yet-synthesized
5258
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2005, 44, 5255 –5259

Angewandte
Chemie
cyclohepta[def]fluorene: f) U. Grieser, K. Hafner, Chem. Ber.
1994, 127, 2307, and references therein.
[2] P. Pierron, Ann. Chim. Phys. 1908, 15, 269.
[3] K. A. Hutchison, G. Srdanov, R. Hicks, H. Yu, F. Wudl, T.
Strassner, M. Nendel, K. N. Houk, J. Am. Chem. Soc. 1998, 120,
2989.
[4] K. A. Hutchison, K. Hasharoni, F. Wudl, A. Berg, Z. Shuali, H.
Levanon, J. Am. Chem. Soc. 1998, 120, 6362.
[5] a) P. v. R. Schleyer, C. Maerker, A. Dransfeld, H. Jiao, N. J. R.
van Eikema Hommes, J. Am. Chem. Soc. 1996, 118, 6317.
[6] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648; b) C. Lee, W.
Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[7] R. Singh, B. M. Deb, Phys. Rep. 1999, 311, 48.
[8] T. Clark, J. Chandrasekhar, Isr. J. Chem. 1993, 33, 435.
[9] P. v. R. Schleyer, M. Manoharan, Z.-X. Wang, B. Kiran, H. Jiao,
R. Puchta, N. J. R. van Eikema Hommes, Org. Lett. 2001, 3,
2465.
[10] V. Gogonea, P. v. R. Schleyer, P. R. Schreiner, Angew. Chem.
1998, 110, 2045; Angew. Chem. Int. Ed. 1998, 37, 1945.
[11] Whereas a variety of cationic cyanine dyes are known, stable
anionic cyanines are rare. TPH-acridine (2) represents, with the
exception of 1, the first molecule with a nitrogen-based anionic
cyanine moiety: a) P. F. Gordon, P. Gregory, Organic Chemistry
in Colour, Springer, Berlin, 1983; b) D. R. Waring, G. Hallas,
The Chemistry and Application of Dyes (Eds.: D. R. Waring, G.
Hallas), Plenum, New York, 1990.
Angew. Chem. Int. Ed. 2005, 44, 5255 –5259
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5259