A R T I C L E S
Fu et al.
It may be noted that the calculated geometries of the planar
ylides resemble those of polyalkenes. Cyclooctatetraene, which
1
is nonplanar, has δ 13C of 131.5 and δ H of 5.6910 and C-C
bond lengths of 1.334 and 1.462 Å, giving ∆R of 0.128.11 We
have obtained NICS(0) and NICS(1) values for the nonplanar
structure as 5.74 and 2.30, respectively, where NICS(0) is taken
as the center of the molecule and NICS (1) is 1 Å above this
(Table 2). This calculation indicates that nonplanar cyclooc-
tatetraene is nonaromatic, as expected. However, the ylides
considered here have the geometric properties of planar non-
aromatic nonconjugated polyenes, and this may reasonably be
attributed to a significant contribution from antiaromaticity.
In summary, observable azacyclobutenone ylides 2, 11, 14,
17, 19, and 21 have been prepared in solution at ambient
temperatures and characterized by their distinctive IR and UV
spectra. The NMR spectra of the longer-lived triisopropylsilyl-
substituted species 21 and calculated NICS values and bond
alternations for 2, 11, and 21 indicate substantial antiaromatic
character for these species. Benzocyclobutenone enolate (24)
and cyclobutenone enolate (26) share some of these properties
with the ylides and with benzocyclobutadiene (23).
Figure 5. MP2/6-31G(d) calculated7a bond distances (Å) of cyclobutenone
enolate (26).
clobutenone enolate (26) has been found to be strongly
destabilized (as indicated by the energy change in eq 10
calculated at the MP2/6-31G(d) level) and a calculated geometry
(Figure 5) described as having allylic anion character with the
C2 and C4 hydrogens bent out of plane by 43°.7a We calculate
essentially the same geometry for 26 using DFT methods, and
calculate a NICS(1) value of 8.59 (Table 2) indicative of
significant antiaromatic character for 26. This may be compared
to the value of 13.3 for cyclobutadiene (Table 2).
Another criterion of antiaromaticity is bond alternation, and
the B3LYP calculated geometries of ylides 2, 11, and 21 and
model compounds are shown in Table S-3 (Supporting Informa-
tion). Values of the difference ∆R between the longest and
shortest bonds in each ring provide a measure of bond
alternation, and as shown in Table 2, these are larger in 2, 11,
and 21 than in benzocyclobutadiene (23)8 and are also larger
than in naphthalene (0.047 Å at 205 K).6c Heteroatoms in 2,
11, and 21 affect these distances, but the differences are
Experimental Section
Diazo ketones 11 and 13,4b and triazole 4 (R ) 2-pyridyl),4c,d are
known compounds prepared by the reported procedures. UV-visible
LFP experiments used a Xe-light source (150 W), spectrograph,
photomultiplier, and ICCD camera, with 248 nm KrF and 308 nm XeCl
excimer lasers, and a 266 and 355 nm Nd:YAG laser. Time-resolved
IR experiments were conducted in Japan with a dispersive-type IR
spectrometer utilizing 266 nm pulses of a Nd:YAG laser (15 Hz
repetition rate, 2 mJ per pulse) as the excitation source. A reservoir of
sample solution was continuously circulated between two calcium
fluoride plates with 2 mm path length. Computations were carried out
using Gaussian 98.12
3-Diazoacetylpyridazine (10). To 3-pyridazinylcarboxylic acid4a
(765 mg, 6.2 mmol) stirred in THF (15 mL) and cooled in an ice bath
under argon was added Et3N (0.945 mL, 6.7 mmol) and isobutyl
chloroformate (0.805 mL, 6.2 mmol). The mixture was stirred 1 h with
cooling, and then added to a cold ether solution of excess diazomethane,
left cold for 1 h, and allowed to warm to room temperature overnight.
The ether solution was washed with saturated NaCl solution and dried
over anhydrous Na2SO4; after evaporation of the solvent, the resulting
solid was chromatographed (alumina, CH2Cl2/hexanes, 1:1 v/v) to give
1
consistent with the other evidence from H NMR and NICS
calculations. The ∆R values for enolate 23 are a little greater
than for 21 for the 6-ring, and a little less for the 4-ring, but the
6-ring values are significantly larger than that for pyridine. Thus
the ∆R values indicate a significant lack of aromaticity in the
4-membered rings of both the ylides 2, 11, and 21, as well as
in benzocyclobutadiene 22 and the enolate 23, and more
aromatic character in the 6-membered rings than in the
4-membered, but less than in benzene or pyridine.
The question of aromatic and antiaromatic character in the
azacyclobutenone ylides 2, 11, 14, 17, 19, and 21 is closely
related to similar considerations regarding cyclobutadiene, and
the continuing great interest in the latter species has recently
been considered in a thoughtful essay.9 There it is noted
“Various criteria have been proposed over time to judge whether
a molecule is aromatic or not... and it was generally assumed
that several of these criteria have to be met for a molecule to
qualify for this distinction.” The magnetic criteria, includ-
ing NICS calculations, indicate antiaromaticity for cyclo-
butadiene, but energetic considerations are more problematic.9
For the ylides considered here the criteria used are the
1
10 (435 mg, 48%) as opaque crystals, mp 127-131 °C dec. H NMR
(400 MHz, CDCl3) δ 9.32 (dd, J ) 1.6, 4.9 Hz, 1H), 8.22 (dd, J )
1.6, 8.4 Hz, 1H), 7.67 (dd, J ) 4.9, 8.4 Hz, 1H), 6.96 (s, 1H). 13C
NMR (100 MHz, CDCl3) δ 184.1, 155.2, 153.7, 127.6, 124.3, 55.0. IR
(CHCl3) 2112, 1634 cm-1. UV-vis (CH3CN) λmax (ꢀ) 217 nm (9.8 ×
103), 307 nm (8.1 × 103). EI-MS m/z 148 (11, M+), 120 (60, [M -
N2]+), 92 (100), 69 (19), 65 (49). HR-EI-MS m/z calcd for C6H4N4O
148.0385, found 148.0389.
2-Pyridyl 1-Diazoethyl Ketone (16). 2-Pyridylcarboxylic acid (0.2
g, 1.62 mmol) in THF (10 mL) was cooled in an ice bath, Et3N (0.25
mL, 1.78 mmol) and isobutyl chloroformate (0.21 mL, 1.62 mmol)
were added, and the mixture stirred for 1 h. A solution of diazoethane
prepared from N-nitroso-N-ethylurea (1.7 g, 14.5 mmol) was added to
the cold solution, which was left standing for 1 h, and then allowed to
warm to room temperature overnight. Saturated NaCl solution was
added and extracted three times with ether. The combined ether layers
1
measured H chemical shifts, and the calculated bond alterna-
tions and NICS parameters, and all of these are consistent
with significant antiaromatic character. While computations
indicate that ylide 2 is strongly destabilized, any relevance of
this result to the antiaromatic character of 2 and its derivatives
is difficult to quantify and beyond the scope of this investigation.
(10) van Eikema Hommes, N. J. R.; Clark, T. J. Mol. Model. 2005, 11, 175-
(8) The calculated geometry of 22 at the ab initio MP2 level (ref 6a) gives ∆R
values of 0.146 for the 4-membered ring and 0.065 for the 6-membered
ring.
(9) Bally, T. Angew. Chem., Int. Ed. 2006, 45, 6616-6619.
185.
(11) Schro¨der, G. Cyclooctatetraene; Verlag Chemie: Weinheim, 1965.
(12) Frisch, M. J.; et al. Gaussian 98, Revision A.9; Gaussian, Inc.: Pittsburgh,
PA, 1998.
9
6214 J. AM. CHEM. SOC. VOL. 129, NO. 19, 2007